U R N A LO F

J A IN E E 7 6 (6 )1 1 6 3 -1 4 3 2 (1993)
IS S N 1060-3271
 M VCOSEP # 2 2 4
g re m o v e s c a r b o h y d ra te s

M
ULTIFUNCTIONAL | p r o t e in s

~> p i g m e n ts
fa ts

fro m
a n d

fe e d /fo o d

CLEANING §
5
e x tra c ts

1 0 s e c o n d s
in ju s t

(MFC)COLUM N
M y c o to x in A p p lic a tio n s
CLEANING COLUMN

ns ▲= Impurities • 2 0 m in . H P L C of
-f = M ycotoxins
A fla to x in s

• 3 0 m in . S im u ltan e o u s
T L C of A fla to x in , D O N ,
a n d Zearalenone

• G C of Trichothecenes
(D O N , T-2 Toxin, etc.)

PURIFIED
EXTRACT A sk a b o u t ou r w o r k s h o p s .

CIRCLE 88 ON READER SERVICE CARD

R Ö M E R L A B S , IN C .
1301 STYLE M ASTER DRIVE
UNIO N, M O 6 3 0 8 4
SAMPLE EXTRACT PHONE (314) 583-8600 FAX (314) 583-6553
- + % = - X - + o/o = - X - + o/0 = - X - +
X
A G u id e f o r U n d e r s ta n d in g a n d A p p l y i n g B a sic
+
B io c h e m ic a l C a lc u la tio n s ...
%
X

+
T h is p u b licatio n will help te c h n ic ia n s
B a s ic and stu d e n ts to u n d e rsta n d an d apply
%
basic biochem ical ca lc u latio n s using
X C a lc u la t io n s sim ple an d c o n siste n t tech n iq u e s.
B iochem ical ca lc u latio n s are needed
+
at som e stage of m ost lab o ra to ry
%
f o r C h e m ic a l
ex p erim en ts— th e types of calculations X
vary b u t often in clu d e basic tasks
& B io lo g ic a l +
su ch as calc u latin g th e a m o u n t of
a re ag en t in an assay m edium ;
%
A n a ly s is c o n v e rtin g raw ex p e rim en tal data
X in to c o n c e n tra tio n s, a m o u n ts, and
activ ities; in te rc o n v e rtin g u n its of
+ Bassey J.S. Efiok %
m easu re; an d ca lc u latin g variables or
c o n sta n ts for biochem ical re actio n s. X

+
W h e th e r you need a review of how to
apply general p rin cip les to specific
%
q u an titativ e problem s or sim ply a less
X com plex te c h n iq u e — th is book will fill
th e need. Q uick referen ce guides to
+ %
ab b rev iatio n s, SI u n its, a n d selected
eq u a tio n s, and an index to th e X
p ractical exam ples is provided.
+

% CONTENTS:
C a lc u la tin g fo r A m o u n t, C o n c e n tr a tio n , a n d
X P r e p a r a tio n o f R e a g e n ts ; B u ffe rs: P r in c ip le s ,
C a lc u la tio n s a n d P r e p a ra tio n ; S p e c tro p h o to ­
+ m e tr y : B a s ic P r in c i p le s a n d Q u a n tita tiv e
%
A p p lic a tio n s ; E n z y m e A ss a y s a n d A c tiv ity ;
R a d io a c tiv ity a n d R e la te d C a lc u la tio n s ; M is c e l­
X
la n e o u s E x a m p le s ; A p p e n d ic e s a n d R e f e r e n c e s .
* +
A p p ro x . 2 0 0 p a g e s. J u ly 1 9 9 3 . S o ftb o u n d .
I S B N 0 - 9 3 5 5 8 4 - 5 1 -X . A p p e n d ic e s . $ 3 9 .0 0 in
% N o r th A m e ric a (U S A , C a n a d a , M e x ic o ), $ 4 7 .0 0

T O O R D E R : sen d y o u r n am e a n d ad d re ss, o u ts id e N o r th A m e ric a . A O A C I N T E R ­

X N A T IO N A L M e m b e rs : s u b tr a c t 1 0 % d is c o u n t.
a n d p a y m e n t (A O A C IN T E R N A T IO N A L

+ a c c e p ts U S fu n d s o n U S b a n k s o n ly ) o r %
c re d it c a r d s : V IS A , M a s te rC a rd o r D in e rs .
W h e n p a y in g b y c r e d it c a r d p le a s e in c lu d e : X
ty p e o f c re d it c a rd , c a rd n u m b e r, e x p ira tio n
d a te a n d y o u r s ig n a tu re . +

%
S E N D T O : A O A C IN T E R N A T IO N A L -J,
1 9 7 0 C h a in B rid g e R o a d , D e p t. 0 7 4 2 ,
M i
M cL ea n , V A 2 2 1 0 9 -0 7 4 2 .
A O A C
X C r e d it c a r d o r d e r s m a y a ls o b e p la c e d IN TER N A TIO N A L
+
by p h o n e : + 1 -7 0 3 -5 2 2 -3 0 3 2 , o r FA X
The Scientific Association Dedicated to Analytical Excellence
%
+ 1 -7 0 3 -5 2 2 -5 4 6 8 .

+ - X - = o/o + - X -H = o/o + - X - = % + -
 A O A C
IN T E R N A T IO N A L

T h e U S . E P A M a n u a l

o f C h e m ic a l M e th o d s

f o r P e s tic id e s a n d D e v ic e s ,

S e c o n d E d itio n

C h a r le s J. S taffo rd , E v e re tt S. G re e r, A d ria n W. B u rn s , D e a n F. Hill, E d ito rs

T h e U .S. EPA t o b e t h e m o s t s u i t a b l e a n d , in
M anual o f s o m e c a s e s , th e o n ly m e th o d s
C h e m ic a l a v a ila b le fo r a p a r tic u la r
M e th o d s f o r fo rm u la tio n .
P e s tic id e s a n d
T h is n e w ly r e v is e d e d itio n o ffe rs
D e v i c e s is a
a n u p d a t e d f o r m a t a n d 18 n e w
c o m p e n d iu m
m e th o d s . S o m e m e th o d s p r e s e n t
o f c h e m ic a l
in t h e p r e v i o u s e d i t i o n a n d u p ­
m e t h o d s fo r
d a te s h a v e b e e n e lim in a te d , s u c h
t h e a n a l y s i s o f p e s t i c i d e s in t e c h ­
a s th o s e fo r p e s tic id e s th a t a r e
n ic a l m a te ria ls , c o m m e r c ia l p e s ti­
n o lo n g e r re g is te r e d a n d th o s e
c id e fo r m u la tio n s a n d d e v ic e s . T h e
fo r w h ic h a n e q u iv a le n t p r o ­
m a n u a l c o n ta in s 2 8 7 m e th o d s th a t
c e d u r e e x i s t s in O ffic ia l M e t h o d s
h a v e b e e n c o n trib u te d b y fe d e ra l
o f A n a ly s is o f th e A O A C . T h e
a n d s ta te a g e n c ie s a n d p riv a te
r e s u l t is a c o n c i s e , u p - t o - d a t e
in d u s try .
m a n u a l d e s i g n e d t o s e r v e a ll
A lth o u g h n o t c o lla b o r a tiv e ly a n a l y t i c a l s c i e n t i s t s i n v o l v e d in
t e s t e d o f f ic ia l A O A C m e t h o d s , p e s tic id e s a n d d e v ic e s .
S e c o n d e d itio n . A p p r o x im a te ly 7 9 0 p a g es. 1992.
m o s t h a v e b e e n v a l i d a t e d in
3 - h o le d r i l l w it h b in d e r IS B N 0 - 9 3 5 5 8 4 -4 7 - 1 .
e ith e r EPA o r s ta te la b o ra to rie s . $ 1 3 8 . 0 0 in N o rth A m e ric a fU S A , C a n a d a , M e x ic o ),
$ 1 6 2 . 0 0 o u ts id e N o rth A m e ric a . M e m b e rs s u b tr a c t
T h e s e p r o c e d u r e s a r e b e lie v e d 10% d is c o u n t.

To order: Send your nam e and address and paym ent. AOAC International accepts checks (US
funds on US banks only, please) and VISA, MasterCard or Diners credit cards. When paying by
credit card, please include: type of card, card num ber, expiration date and your signature.
Send to: AOAC International - J, 1970 Chain Bridge Road, Dept. 0742, McLean, VA 22109-0742.
Credit card ord ers m ay a ls o b e p la ced by p h o n e +1 (703) 5 2 2 -3 0 3 2 , or FAX +1 (703) 5 2 2 -5 4 6 8 .
 J O U R N A L O F
' * - iT v ^ V ’ Ï " <™rj

A O A C T a b le o f C o n t e n t s

INTERNATIONAL
A Peer-Reviewed Journal Publishing N ovem ber/D ecem ber 1993 Vol. 76, No. 6 JA IN EE 7 6 (6 )1 1 6 3 -1 4 3 2 (1 9 9 3 )

Basic and Applied Research in the

Analytical Sciences Related to Foods,
Drugs, Agriculture, and the Environment 161A B o o k s in B r ie f
Ê0È - f 5R >k " 'A & ’ ™ -> I 163A N e w P ro d u c ts
E dito rial B o ard
168A F o r Y o u r In fo r m a tio n

Jam es T. Tanner 175A T h a n k s to R e v ie w e r s

C h a ir m a n a n d E d ito r - in - C h ie f
180A R e p o r t o f t h e A O A C IN T E R N A T IO N A L T a s k F o r c e o n M e t h o d s
Dieter Arnold
fo r N u tr ie n t L a b e lin g A n a ly s e s
Jam es Ault
Franklin E. Barton II 202A C u m u la tiv e A u th o r In d e x
Richard H. Collier
Richard A. Durst
J. Richard Gorham
Anant V. Jain
Jam es Lichtenberg A G R IC U L T U R A L M A T E R IA L S
W. Harvey N ewsom e
Joseph Sherma 1163 L i q u id C h r o m a to g r a p h ic D e te r m in a tio n o f M e le n g e s t r o l A c e ta te in F e e d s
David L. Terry
Harold M. Campbell and François Sauvé

J o urna l E ditors 1168 R a p i d D e te r m in a tio n o f T r y p to p h a n in I n ta c t P r o te in s b y D e r iv a tiv e

Agricultural Materials S p e c tro p h o to m e try

R odney J. Noel Dimitrios J. Fletouris, Nickos A. Botsoglou, Georgios E. Papageorgiou, and

Drugs, Cosmetics, Forensic Materials Antonios J. Mantis
Joseph W estheim er
Food Chemical Contaminants 1174 E v a lu a tio n o f C u rre n t A O A C F e rtiliz e r S a m p le P re p a ra tio n R e q u ire m e n ts

Albert E. Pohland Natalie Newlon, Candace Cox-Trout, and Peter Kane

Residues and Trace Elements
Joseph Sherm a 1182 E v a lu a tio n o f N e w F la m e P h o to m e tr ic In s tru m e n ta tio n to M e e t R e q u ire m e n ts

Food Composition and Additives o f A O A C O f f ic ia l M e th o d f o r P o ta s s iu m in F e r tiliz e r s

Jam es F. Lawrence Natalie Newlon

Chemical Contaminants Monitoring
Jam es D. M acNeil 1187 R e v e rs e d -P h a s e L iq u id C h r o m a to g ra p h ic M e th o d fo r th e S im u lta n e o u s

Statistical Analysis D e te r m in a tio n o f B e n o m y l a n d M e th y l lf f - B e n z im id a z o l- 2 - y lc a r b a m a te

G eorge W . Latimer, Jr in W e tta b le P o w d e r F o r m u la tio n s

Food Biological Contaminants Raj P. Singh and Mikio Chiba

G eorge J.Jackson
C H E M IC A L C O N T A M IN A N T S M O N IT O R IN G

J o ur n a l S taff
_ ■■ . • ; 1193 T h e C h in e s e T o ta l D ie t S tu d y in 1 9 9 0 . P a r t I. C h e m ic a l C o n ta m in a n t s
Publications Director: Krystyna A. Mclver Junshi Chen and Junquan Gao
Managing Editor: Robert E. Rathbone
Assistant Editor: Scott Hofmann-Reardon
1206 T h e C h i n e s e T o t a l D i e t S t u d y i n 1 9 9 0 . P a r t II . N u t r i e n t s
Production: Thomas McMullan
Advertising: Christine Barger Junshi Chen and Junquan Gao
Subscriptions: Mark Stewart

A OAC B oard O F D irectors

President: Arvid Munson
President-Elect: Alan R. Hanks
Secretary/Treasurer: P. Frank Ross Copyright, 1993, by AO AC International, Inc. All rights reserved. The Journal o f AOAC
Immediate Past International (ISSN 1060-3271) is published bimonthly by AO AC International, Suite 400,
President Henry B.S. Conacher 2200 Wilson Blvd, Arlington, VA 22201-3301 USA. Subscriptions for Vol. 76 (1993) are:
Directors: Paul R. Beljaars Members, $120.00 in U.S., $158.00 outside U.S.; Nonmembers, $175.00 in U.S., $213.00
George H. Boone outside U.S. Second class postage paid at Arlington, VA, and additional mailing offices. Ab­
Claire A. Franklin stracted or indexed in A G RICO LA , Analytical Abstracts, Biological Abstracts, Cambridge Sci­
Eugene J. Klesta entific Abstracts, Chemical Abstracts, Current Advances in Ecological Sciences, Current
Raymond Matulis Contents, and others. Printed in USA.
Albert E. Pohland POSTMASTER: Send address changes to AO AC International, Suite 400, 2200 Wilson
• V.■Executive Director: Ronald R. Christensen Blvd, Arlington, VA 22201-3301 USA.

ut I J 11 u m .U V lI U lfm

à D. R. 253'/
 1213 S u r v e y o f B e n z e n e in F o o d s b y U s in g H e a d s p a c e C o n c e n tr a tio n T e c h n iq u e s
a n d C a p illa r y G a s C h ro m a to g ra p h y
Timothy P. McNeal, Patricia J. Nyman, Gregory W. Diachenko, and Henry
C. Hollifield

1220 P e s t i c i d e R e s i d u e s in C o m p o s i t e d M i l k C o l l e c t e d T h r o u g h t h e U .S .
P a s te u r iz e d M ilk N e tw o rk
William J. Trotter and Richard Dickerson

1225 S u r v e y o f T o ta l T e tr a h y d r o p h th a lim id e in B a b y F o o d s U s in g B o th
E n z y m e -L in k e d Im m u n o s o rb e n t A ss a y a n d G a s C h ro m a to g ra p h y /M a s s
S p e c tr o m e tr y : A C o m p a ra tiv e S tu d y
Jupiter M. Yeung and W. Harvey Newsome

DRUGS, COSMETICS, FORENSIC MATERIALS

1230 L i q u id C h r o m a to g r a p h ic D e te r m in a tio n o f M o x id e c tin R e s id u e s in C a ttle

T i s s u e s a n d C o n f ir m a tio n in C a t tle F a t b y L i q u id C h r o m a to g r a p h y /M a s s
S p e c tro m e try
Alisa Khunachak, Adrian R. Dacunha, and Steven J. Stout

1235 S im u lta n e o u s D e te r m in a tio n o f N itr o f u r a z o n e a n d F u r a z o lid o n e in S h r im p

(.Penaeus vannamei) M u s c l e T is s u e b y L iq u id C h r o m a to g r a p h y w ith U V
D e te c tio n
Heidi S. Rupp, Robert K. Munns, and Austin R. Long

FOOD BIOLOGICAL CONTAMINANTS

1240 Bacteriological Analytical Manual C u l t u r e M e t h o d f o r

E f fe c tiv e n e s s o f th e th e
Shigella sonnei f r o m S e l e c t e d F o o d s
R e c o v e ry o f
Geraldine A. June, Patricia S. Sherrod, R. Miguel Amaguana, Wallace H.
Andrews, and Thomas S. Hammack

1248 L i q u id C h r o m a to g r a p h ic D e te r m in a tio n o f A f la to x in M i in M ilk P o w d e r U s i n g

Im m u n o a f fin ity C o lu m n s f o r C le a n u p : In te rla b o r a to r y S tu d y
Louis G.M.Th. Tuinstra, Arie H. Roos, and John M.P van Trijp

FOOD CHEMICAL CONTAMINANTS

1255 S e m ia u to m a tic D e te r m in a tio n o f F u r a n ic A ld e h y d e s in F o o d a n d

P h a rm a c e u tic a l S a m p le s b y a S to p p e d -F lo w In je c tio n A n a ly s is M e th o d
A. Espinosa-Mansilla, A. Muñoz de la Peña, and F. Salinas

1262 S p e c ia tio n A n a ly s is o f O rg a n o le a d C o m p o u n d s in W in e b y C a p illa r y G a s

C h ro m a to g ra p h y /M ic r o w a v e -I n d u c e d -P la s m a A to m ic E m is s io n
S p e c tro m e try
Ryszard Lobinski, Joanna Szpunar-Lobinska, Freddy C. Adams,
Pierre-Louis Teissedre, and Jean-Claude Cabanis

1268 D e te r m in a tio n o f V o la tile C h e m ic a ls R e l e a s e d f r o m M ic r o w a v e - H e a t-

S u s c e p to r F o o d P a c k a g in g
Timothy P. McNeal and Henry C. Hollifield

FOOD COMPOSITION AND ADDITIVES

1276 M e th o d M o d ific a tio n f o r L iq u id C h ro m a to g ra p h ic D e te r m in a tio n o f T h ia m in e ,

R ib o f la v in , a n d P y r id o x in e in M e d ic a l F o o d s
G. William Chase, William 0 . Landen, Jr, Abdel-Gawad M. Soliman, and
Ronald R. Eitenmiller

1280 P a rtia l A u to m a tio n o f M ic r o b io lo g ic a l A s s a y s f o r V ita m in s

TO. Richard Haggett, Alan R. Matheson, and Michelle Harnett

158 A J o u rn a l O f A O A C In ternational V o l . 76, N o. 6 ,1 9 9 3

1289 D eterm in a tio n o f M e n th o l in H o n ey b y G a s C h ro m atog rap h y
Maojuan Li, Donald L. Nelson, and Peter Sporns

1295 C u p rim etric A ss a y o f C a se in H y d ro ly sates

M.P.C. Silvestre, E. Lati, C. Dauphin, and M. Hamon

1300 A Q u an titativ e L in ea rity E v a lu a tio n M eth o d fo r In frared M ilk A n a ly z e rs

Erika B. Smith, David M. Barbano, Joanna M. Lynch, and J. Richard
Fleming

RESIDUES AND TRACE ELEMENTS

1309 A n a ly te S ta b ility Stu d y o f A -M e th y lc a rb a m a te P e stic id e s in B e e f and P oultry

L iv e r T issu e s b y L iq u id C h ro m ato g rap h y
M. SherAIi, Jerry D. White, Ralph S. Bakowski Evan T. Phillippo, and
Richard L. Ellis

1317 C o m p a riso n o f G e l P erm ea tio n C h rom atog rap h y , S w e e p C o d istilla tio n , and
F lo ris il C o lu m n A d so rp tio n C h ro m ato g rap h y as S a m p le C lean u p
T ech n iq u es fo r th e D eterm in a tio n o f O rg a n o ch lo rin e P e stic id e R e sid u e s
in A n im a l F a ts
Paul Armishaw and Roderick G. Millar

1323 D e te rm in a tio n o f P araq u at and D iq u a t in L o w -M o istu r e F o o d C ro p s U sin g

S ilic a C o lu m n C lea n u p and L iq u id C h ro m ato g rap h y w ith U V D e te c tio n
Tina M.P. Chichila and Dalia M. Gilvydis

1329 In terla b o ra to ry Stu d y o f a T h e rm o sp ra y -L iq u id C h ro m a to g ra p h ic/M a ss

S p e c tro m e tric M e th o d fo r S e le c te d A -M e th y l C arb am ates, A -M e th y l
C a rb a m o y lo x im e s, and Su b stitu ted U re a P e stic id e s
Viorica Lopez-Avila and Tammy L. Jones

1344 C o n firm a tio n o f Id en tities o f P ro p y len e and E th y le n e G ly c o ls in A n c h o v ie s by

T a n d em M a s s Sp ectro m etry
Jean E. Matusik, Paul P. Eilers, Ellen M. Waldron, Steve M. Conrad, and
James A. Sphon

1348 D e te rm in a tio n o f P yreth ro id R e sid u e s in V e g e ta b le s, F ru its, G ra in s, B e a n s , and

G reen T e a L e a v e s : A p p lica tio n s to P yreth ro id R e sid u e M o n ito rin g
Stu d ies
Yumiko Nakamura, Yasuhide Tonogai, Yukari Tsumura, and Yoshio Ito

1362 D e te rm in a tio n o f F o rm e ta n a te H y d ro ch lo rid e in S e le c te d F ru its b y

C o u p le d -C o lu m n C a tio n E x c h a n g e L iq u id C h ro m atog rap h y
Richard A. Niemann

1369 R a p id G a s C h ro m a to g ra p h ic M eth o d fo r the M u ltiresid u e S c re e n in g o f F ru its

and V eg eta b les fo r O rg a n o ch lo rin e and O rg an o p h osp h ate P e stic id e s
Harry M. Pylypiw, Jr

1374 In d u ctiv ely C o u p led P la s m a A to m ic E m iss io n S p e ctro m e tric D eterm in a tio n o f

T in in C a n n e d F o o d
Hidenobu Sumitani, Sachiko Suekane, Aya Nakatani, and Kiyoaki Tatsuka

1378 D e te rm in a tio n o f U ltra tra ce L e v e ls o f L e a d in In fa n t F o rm u la b y Iso to p e

D ilu tio n In d u ctiv e ly C o u p led P la s m a M a s s Sp e ctro m e try
Joseph J. Thompson

J o u rn a l O f A O A C In tern a tio na l V o l . 76, N o. 6 ,1 9 9 3 159A

 TECH N IC A L COM M UNICATIONS

1385 F lu o r o m e tr ic D e te r m in a tio n o f T o ta l a n d B o u n d S u lf ite in W in e b y

/V -(9 -A c rid in y l)m a le im id e
Kazuaki Akasaka, H iroshi Ohrui, H iroshi Meguro, Tateo Suzuki, and
Harufum i M iw a

1389 D e te r m in a tio n o f S u lf u r D io x id e in W in e s a n d B e v e r a g e s b y F lo w I n je c tio n

A n a ly s is w ith R e d u c tiv e A m p e r o m e tr ic D e te c tio n a n d E le c tr o ly tic
C le a n u p
Terence J. Cardwell, Robert W. Cattrall, Guo Nan Chen, Geoffrey R.
Scollary, and Ian C. Ham ilton

1394 T h in - L a y e r C h r o m a to g ra p h ic D e te c tio n o f O r g a n o p h o s p h o r u s In s e c tic id e s

C o n ta in in g a N itro p h e n y l G ro u p
Vitthal B. P a til and M u rlid h a r S. Shingare

1396 L i q u id C h r o m a to g r a p h ic M e th o d f o r D e te r m in in g th e P e r c e n t o f O le s tr a in
L ip id S a m p le s
Daniel H. Tallmadge and Peter Y.T. Lin

1400 U s e o f S ta n d a rd A d d itio n s to D ia g n o s e M a tr ix E ffe c ts in th e A to m ic

A b s o r p tio n S p e c tro p h o to m e tric A n a ly s is o f F e e d B le n d e d f o r a
R o x a r s o n e C o m b in a tio n S tu d y
Gregory K. Webster and Loretta A. Hearne

1993 INDEX

1405 A u th o r In d e x

1416 S u b je c t In d e x

160A J o u rn a l O f A O A C I n ternational Y o l . 76, N o. 6 ,1 9 9 3

 r o b o t s c o u p e
B o o k s i n B r i e f

New Batch
Processing
H andbook for Derivatives for Chro­ bols, and units. I; is a unified reference
m atography. Edited by Karl Blau and
John M. Halket. Published by John
source for all authors of scientific publi­
cations.
Equipment
Wiley & Sons, Inc., 1 Wiley Dr, Somer­
set, NJ 08875-1272,1993.369 pp. Price: Directory o f R esearch & Education in
$95.00. ISBN 047192699X. Food Science, Technology & Engi­
neering. Sponsored by the International
The new edition of this handbook is de­ Food Information Service, the European
signed to be more comprehensive and Federation of Food Science and Tech­
systematic, and to reflect recent devel­ nology, and the International Union of
opments and advances in chemical deri- Food Science and Technology. Publish­
vatization, in chromatographic methods ed by IF1S Publishing, Lane End House,
and in analytical instrumentation. It is a Shinfield, Reading, RG2 9BB, UK,
practically oriented manual for labora­ 1993. Price: UK: £130.00, US: $225.00.
tory use. The topics of interest include:
advances in silylation and alkylation, Ideal for R&D,
Now all the relevant information on re­
derivative formation by ketone-base
search and education in all aspects of
QC, scientists,
condensation, colored UV-absorbing
food science, technology, and engineer­
engineers, sam ­
derivatives and fluorescent derivatives, ple preparation
ing is available in one place. This new
and derivatization for chromatographic for analysis
directory lists the research interests of
resolution of optically active com­ and pilot plant
the major players in Europe and North
pounds. Many techniques are discussed,
America; details undergraduate and production,
in particular: analysis by soft ionization
graduate courses; tells you who is avail­ these batch
mass spectrometry, supercritical fluid
chromatography, gas chromatogra- able for consultancy work and in which processors are
phy/mass spectrometry, and post chro­ countries they have recent expertise; used for mixing light to
matographic derivatization. The final shows you where networks are being set heavy viscosity matter,
chapter is a general view of the practical up; lists journals and other regular pub­ powders, liquids with solids
aspects, including safety, involved in lications; covers short courses and or powders. Rugged 1/2 to
derivatization chemistry. sources of funding; and indicates who to 15 HP m otors provide m ix­
contact for admission. It covers Europe, ing, blending, homogenizing
including the former USSR, and North and em ulsification in a con­
Reporting Experim ental Data. Edited
by Howard J. White, Jr. Published by the America, including Canada, the United trolled environm ent vessel.
American Chemical Society, 1155 16th States, anc Mexico, and has entries from Size reduction can be
over 500 universities, companies, and
St, NW, Washington, DC 20036, 1993. achieved during the mixing
360 pp. Price: U.S. & Export: $89.95. institutions undertaking research in food
cycle. Mixing and size
ISBN 0-8412-2529-X. science, technology, or engineering, or
reduction can be achieved
providing education and training in
independently or sim ulta­
these fields. It is designed to be easy to
This practical guide will tell you every­
use and includes indexes by institutions,
neously. Som e benchtop
thing you need to know about the proper models can also be
as well as by areas of scientific exper­
handling of numeric data. Compiled converted to continuous
from widely scattered sources, this de­ tise. Entries are grouped alphabetically
by country. feed for size reduction.
finitive reference volume offers infor­
mation on coordinating data from many For more information
disciplines. This volume gathers the rec­ Taulbee’s Pocket Com panion: U.S.
call Product Mgmt. Dept.,
ommendations for publication of data FD A and EPA G LPs in Parallel. Ed­
ited by Stephanie M. Taulbee and
601-956-3216 or
for measurements of a physical chemi­
Robert S. DeWoskin. Published by In-
FAX 601-956-5758.
cal nature together with current IUPAC
CIRCLE 101 ON READER SERVICE CARD
recommendations for quantities, sym­ terpharm Press, 1358 Busch Pkwy, Buf-

Jo u r n a l O f A O A C I n t e r n a t io n a l V o l . 76 N o. 6 ,1 9 9 3 161A
Books in Brief

falo Grove, IL 60089,1993. Price: Five- procedures for the analytical work itself FD A News: D rugs, C osm etics, D e­
Pack/$59.00. is now expected to become a require­ vices and Biologies. A monthly news­
Taulbee’s Pocket Companion is an easy- ment for confidence in laboratories and letter sponsored by the Association of
to-carry, concise reference to the full for the acceptance of results. In the fu­ Food and Dmg Officials. Edited by Y.H.
text of the U.S. Environmental Protec­ ture, laboratory accreditation—as an Hui. Published by the Technomic Pub­
tion Agency and the U.S. Food and Drug important instrument for the dairy in­ lishing Co., 851 New Holland Ave, Box
Administration GLPs made especially dustry to strength confidence in labora­ 3535, Lancaster, PA 17604,1993. Price:
for those who have quality assurance, tory results—will only be possible on One year: $185.00. ISBN 1069-5109.
regulatory affairs, marketing, or scien­ the basis of such quality assurance prin­
tific responsibilities for products or stud­ ciples. Aimlytical Quality Assurance For many years, the Association of Drug
ies having to comply with more than one and Good Laboratory Practice in Dairy Officials (AFDO) has spent an enor­
U.S. regulatory agency. It contains the Laboratories covers basic needs and mous amount of effort to assist the food
verbatim text of the FDA and EPA concepts, general requirements, and in- and drug industries in complying with
(TSCA and FTFRA) GLPs, fully refer­ tralaboratory and interlaboratory aspects federal and state regulations To enhance
enced and guaranteed to be the most re­ of analytical quality assurance for dairy its objective of communicating cmcial
cent versions available. It is painstak­ institutes, central testing laboratories, information quickly AFDO has devel­
ingly arranged so that the text of the and dairy factory laboratories. The semi­ oped a unique newsletter. This newslet­
different GLP regulations and the appli­ nar was sponsored by AOAC INTER­ ter, FDA News: Drugs, Cosmetics, De­
cable requirements for each topic can be NATIONAL, the Commission of the vices and Biologies, is designed to keep
compared side-by-side. the affected industries information of the
European Communities, the Interna­
tional Dairy Federation, and the Ger­ latest legal, regulatory, and technical de­
A nalytical Q uality Assurance and
many National Committee of IDF. velopments regarding their products and
Good Laboratory Practice in D airy services. Each month this newsletter
Laboratories. Published by the Interna­ will review numerous FDA documents
International GLPs. Edited by Robert
tional Dairy Federation, 41, Square Ver- and summarize in 1 2 pages items that
S. DeWoskin and Stephanie M. Taulbee.
gote, B-1040 Brussels, Belgium, 1993. are cmcial to industries doing business
Published by Interpharm Press, 1358
429 pp. Price: 3900 BF. ISBN 92-9098- in drugs, cosmetics, devices, and bi­
Busch Pkwy, Buffalo Grove, IL 60089,
0 1 0 -2. ologies. Scanning this letter takes about
1993.444 pp. Price: $198.00.
1 0 min, which will replace the normal

The proceedings of an international time block you spend each month on

International GLPs is a guide to U.S., browsing government documents. Since
seminar held in Sonthofen, Germany,
Japanese, and OECD good laboratory the original sources are stated, you will
May 18-20, 1992, Analytical Quality
Assurance and Good Laboratory Prac­ practice regulations, standards, codes, definitely save time if you wish to con­
tice in Dairy Laboratories examines the and guidelines. In addition to unique ta­ sult the original sources. Also, each issue
role of quality assurance in the compara­ bles comparing the texts of the United will increase the latest data from 5 cen­
bility and reliability of dairy laboratory States, Japanese, and OECD GLPs, it in­ ters with the FDA (CBER, CDER,
measurements. At first glance, such a cludes a discussion of the history and CDRH, CVM, AND NCTR) and list
general topic may seem like a small part harmonization of GLPs, a detailed look names of companies having received a
of the increasingly sophisticated world at the critical differences among national warning letter from FDA during one
of laboratory research. However, experi­ GLPs for compliance programs, the ver­ specific month. The newsletter will in­
ence in many interlaboratory studies at batim text of 7 different GLPs, and more clude essential information form major
national and international levels has — all in one convenient volume. It is an and minor documents, such as Federal
demonstrated that besides standardized essential resource for regulatory affairs Register, FDA Compliance Policy
and validated methods, analytical qual­ professionals, quality assurance audi­ Guides, Import Alerts, FDA Medical
ity assurance plays a key role for the re­ tors, and all others involved in multi-na­ Bulletin, FDA Veterinarian, FDA In­
liability of laboratory results. Introduc­ tional research, testing, or marketing of spection Operations Manual, and many
tion of systematic quality assurance pharmaceuticals and bulk chemicals. more. ■

162A J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 N o . 6 ,1 9 9 3
N e w P r o d u c t s

New ReactIR MP Mobile a wide range of conditions and tempera­ the new 3M product offers fast non­
Reaction Monitoring System tures without effecting chemical equili­ toxic coupling, high capacity, low ligand
Expands the Use of FTIR from bria. Choice of reaction sampling tech­ leaching, and minimal non-specific
the Laboratory to the Plant nology includes transmission cells, binding. 3M Emphaze medium is stable
internal reflectance insertion probe for in a wide range of cleaning and regen­
Applied Systems Inc has introduced Re­ in-line analysis, and a flow cell for on­ eration solutions. It can withstand pres­
actIR MP, the latest addition to the com­ line analysis of slip streams. Applied sure up to 100 psi, has a linear velocity
pany’s line of high performance reaction Systems Inc. up to 3000 cm/h, a typical particle size
analysis systems for the process indus­ Circle No. 380 on reader service card. of 60 microns, and a maximal pore size
try. ReactIR MP is a modular and mobile greater than 1000 Angstroms. 3M.
reaction monitoring system that can eas­ New Biosupport Medium Circle No. 381 on reader service card.
ily be moved from one location to an­ Available for Affinity
other and quickly adapted to accommo­ Chromatography Processes New Integrated Spectrometer
date different sampling requirements. System Available
ReactIR MP is a fully integrated system 3M Emphaze Biosupport medium AB 1
designed for in-situ, real-time monitor­ is now available from 3M Bioapplica­ Chromex, Inc announced availability of
ing and analysis of chemical reactions. It tions. 3M Emphaze medium is pre-acti­ the new integrated Raman One spec­
provides reaction survey, end point, and vated, making it especially well suited trometer system. Specifically designed
kinetic information in real-time with no for immunoaffinity, proprietary ligand, for broad applications in research, in­
sample handling or preparation. The MP and enzyme immobilization applica­ dustry, and education, the instrument is
is capable of analyzing chemistry under tions. Based on an azlactone copolymer, compact, easy to use, and extremely cost

ARL

Sim ply th e best KEVEX

VG INSTRUMENTS

replacem ent for

Kjeldahl. • 1 2 5 p o sitio n a u to sa m p ler

• Results in < 3 m in u tes

T h e N A 2 0 0 0 p ro v id e s a safe

re p la c e m e n t fo r y o u r N itro g e n /P ro te in

a n a ly s is b y e lim in a tin g th e n e e d fo r a c id s

a n d to x ic c h e m ic a ls. T h e sim p lic ity o f th e

CIRCLE 102 ON READER
d esig n , c o m p le te SERVICE CARD

a u to m a tio n a n d
F is O N s
re lia b ility m a k e s I Instrun
Instruments
th e N A 2 0 0 0 th e
Fisons instruments Inc.
32 Commerce Center
b e st N itro g e n / Cherry Hill Drive
Danvers. MA 01923 USA
P ro te in a n a ly z er. Tel: 508 524-1000
New Products

co m p e titiv e . T h e R am an O n e integrates 2 o n -screen p lottin g d e v ic e s , 8 critical m L o f g ly p h o sa te R estore to th e guard or

the fo llo w in g co m p o n en ts in to a c o m ­ referen ce charts and tab les, printout re­ co lu m n e lim in a tes traditional tim e-co n ­
pact d esk top sp ectroscop y system : fast sou rces, b ibliography, and an exten d ed su m in g and le s s-e ffe c tiv e ch ela tio n pro­
(F /4 ) sin g le stage im a g in g spectrograph, g lo ssa ry o f im portant H P L C term s. S A ­ ced u res to sa v e the an alytical colu m n .
built-in M R d io d e laser, therm o-electri- V A N T A u d io v isu a ls. P ick erin g L aboratories.
ca lly c o o le d 16 bit C C D array detector, C ircle N o . 3 8 3 on reader se r v ic e card. C ircle N o . 3 8 5 on reader se r v ic e card.
fiber op tic probe in a ligh t-tigh t sam p le
cham ber, C h rom sp ec R am an softw are New Dataplate Model 300 Cel-Line Offers HTC Slides on
for total sp ectrom eter control from M S - Constant Temperature Micropure Optical Quality Glass
W in d o w s softw are w ith spectrum e x ­ Circulators Feature Simple
port to op tion al G ram s softw are and Digital Controls T h e n ew H T C (H ea v y T eflon C oated )
PLS routin es for ch em om etric an alysis. printed m ic r o sc o p e slid es o n M icroP ure
C o n v en ien t R am an O n e sp ectroscop y P M C Industries, Inc. has ju st introduced op tical w h ite g la ss fro m C e l-L in e perm it
a llo w s users to sam p le liq u id s and solid s its n e w D ataplate M o d el 3 0 0 sim p le m ore accurate sp e c im e n evalu ation . M i-
from b ottles, curvets, and sea led vials. It digital con stan t tem perature circulators croP ure G la ss con tain s 47% le ss iron
features e a sy set-up for rapid acq u isition fo r u se in b iotech , clin ic a l, m ed ical, than w h ite w ater slid es, and transm its
o f R am an spectra. T h e un iq u e c o n fig u ­ ch e m ic a l, and other laboratories. T h e lig h t and co lo r w ith the least am ou n t o f
ration o f the R am an O n e p rovid es the M o d e l 3 0 0 features sim p le d igital c o n ­ distortion. H T C slid es are the o n ly
user w ith the cap ab ility to perform quan­ trol o f all param eters, m o d e m m icro­ printed slid es m anufactured w ith M i-
titative a n a ly sis on liquid sam p les. p rocessor e lec tro n ics, and P ID tem pera­ croP ure op tica l q u ality g la ss, and are
C h rom ex, Inc. ture con trol. T em perature readout is to av a ila b le e x c lu s iv e ly from C el-L in e.
C ircle N o . 3 8 2 o n reader serv ice card. 0.1 °C or °F, w ith lo w w ater le v e l cu toff, T h e proprietary F1TC co a tin g h e lp s pro­
over-tem perature protection, and other pel liq u id s aw a y from the m ask to ensure
HPLC Calculator Program sa fety features, in clu d in g heater drive greater cle a n lin e ss o f the slid e , prevents
and e lec tro n ics failu re c u to ffs. U n its are cross contamination, and increases m i­
av a ila b le in 100, 115, and 230V A C , croliter capacities. C el-Line A ssociates, Inc.
A n ew W in d o w s H P L C so ftw a re pro­
5 0 /6 0 H z, 1100 w atts, 10-am p. P M C In­ C ircle N o . 3 8 6 on reader serv ice card.
gram , con tain in g a c o lle c tio n o f labora­
dustries, Inc.
tory calcu lators and oth er u sefu l u tilities,
is n o w a vailab le from S A V A N T A u d io ­ C ircle N o . 3 8 4 on reader serv ice card. Cell Robotics Introduces New
v isu a ls, Inc. T h e program , HPLC Calcu­ Motorized Microscope Stage
lations Assistant and Reference Tools, is Glyphosate Column
u sefu l to a n y o n e w ork in g in a laboratory Decontaminant C ell R o b o tics, Inc. (C R I) an n ou n ced
w h o n eed s to do various H P L C ca lc u la ­ in troduction o f th e a ll-n e w S m artStage
tions. T h e program features a m aster G ly p h o sa te h erb icid e an alysis by p o st­ m otorized m ic r o sc o p e stage. D e sig n e d
P a g e S electo r w h ich a llo w s the operator co lu m n H P L C (E P A M eth o d 5 4 7 ) e m ­ to rep lace m anual stages in standard up­
to c h o o s e any o f 33 b uilt-in fu n ction s. A p lo y s a c a tio n -e x c h a n g e colu m n . P o ly ­ right and in verted m icro sco p es, the
s w e e p o f the m o u se cursor prod u ces a v a len t m etal io n s from sa m p les, u su ally S m artStage co m b in es p recision (0.1 m i­
pop -u p w in d o w d escrib in g ea ch but­ iron, accu m u late in the ca tio n -ex ch a n g e cron resolu tion ), co m fo rt and afford ab il­
to n ’s fu n ction s. Individual calcu lators resin ca u sin g seriou s d egradation o f c o l­ ity to m ake accurate scan n in g o f slid es
related to p eak s, band spreading, resolu ­ um n p erform ance. A s little as 2 0 0 n m o le and sa m p les q u ick , sim p le, and ea sily
tion , co lu m n s, elu en ts, d etection , and o f ferric ion can c a u se the gly p h o sa te repeatable. F eaturing a 7 5 m m x 5 0 m m
quantitation are con tain ed in th e pro­ peak to d isappear co m p letely . F or its range o f m otion a lo n g the X - a x is and Y -
gram . F or e x a m p le, calculators in the g ly p h o sa te an alysis sy ste m s, P ick erin g a x is, resp ectively, the Sm artS tage a c­
“P eak s” sectio n d eterm ine retention L aboratories in clu d es a guard colu m n co m m o d a tes 3 standard m ic r o sc o p e
tim e, peak v elo city , retention v o lu m e, that con tain s the sa m e cation resin as the slid es sim u ltan eou sly. Petri d ish e s, c u l­
se lectiv ity , p lates, H ETP, and a sy m m e ­ an alytical co lu m n . B e c a u se th e guard ture flask s, and oth er d e v ic e s m ay a lso
try. In operation, users m ay input par­ w ill trap all o f the m etal ion s, it sh ou ld b e u sed. A n op tion al m otorized fo cu s
ticular param eters particular to their be the o n ly co lu m n requiring restora­ adjust a llo w s fu ll 3 -a x is con trol o f all pa­
o w n exp erim en ts. T h e softw are in clu d es tion . T h e ap p lication o f as little as 6 - 1 0 ram eters u sin g a sin g le trackball inter-

164A Jo u rn a l O f AOAC In t e r n a t io n a l V o l . 76, No. 6 1 993

New Products

fa c e and p rovid es sen sitiv ity equ al to or B io -S p in 6 c o lu m n s h a v e an ex c lu sio n co lu m n s h a v e an e x c lu s io n lim it o f ap­

better than the m ech a n ica l fo c u s it re­ lim it o f ap p roxim ately 5 b ase pairs, or p roxim ately 2 0 b a se pairs, or 4 0 0 0 0 d al­
p la ces. A p o w erfu l m icrop rocessor- 6 0 0 0 d alton s for p roteins. B io -S p in 3 0 ton s for p roteins. T h e u n iq u e d esig n o f
b a sed controller, h o u se d w ith in the X - Y
stage, co n n e c ts to the trackball d e v ic e
and a h an d h eld R S 2 3 2 term inal (or a p e ­
ripheral com p u ter w ith a standard c o m ­
m u n ication s p a ck a g e m a y b e su b sti­ N U T R IT IO N NUTRITION INFORMATION
PER SERVING
tuted). B e c a u se the S m a rts tage
Serving Size:
responds to an e x te n s iv e repertoire o f
sim p le text co m m a n d s, n o sp ecia l h o st
softw are is required. T h e S m a rts tage al­
LABEL 4 ounces condensed
1 cup (250 g) as prepared
Servings per container: 2 3/4
lo w s u se o f either ab solu te or relative
coord in ates fo r X - Y m o tio n . A n y refer­ AN SW ERS Calories
Calories from fat
170
30
e n c e lo ca tio n can b e se le c te d b y m o v in g Fat 3 g
to that p o in t u sin g p o sitio n in g c o m ­ N u tritio n l a b e l i n g u n d e r th e Saturated fat 19
m ands from the term inal or b y u sin g the n e w r e g u la tio n s w ill be m o re Cholesterol 5 mg
trackball and th en en terin g a sin g le char­ d e m a n d in g . M ed a llio n L a b o ra to rie s Sodium 1,000 mg
acter com m an d . S ca n n in g m atrixes are
c a n h elp m e e t th o se d em a n d s. W e
Carbohydrate 27 g
d efin ed b y th e user, orien ted b y co lu m n
a s s is t y o u in s e t tin g u p a n a n a ly sis
Fiber 3g
or row . T h e s iz e o f th e fie ld s w ith in the Protein 9g
p ro g ra m th at keeps y o u r p rod u ct
m atrix are a lso u ser-d efin ed , after w h ic h PERCENT OF DAILY VALUE
the scan n in g p ro cess m ay b e m an u a lly
a n d la b e ls c u r r e n t w it h F D A a n d
Vitamin A t
or au tom atically con trolled . L o ca tio n U S D A p o lic y .
Vitamin C 2
coord in ates an d d ista n ces b e tw e e n lo c a ­ Y o u g e t e x p e rt c o n s u lta tio n , fast
Calcium 4
tion s can b e v ie w e d o n the term inal d is­ tu rn a ro u n d an d q u ality te stin g — Iron 8
play. C e ll R o b o tics, Inc. fu ll s e r v ic e f r o m in itia l in g r e d ie n t t Contains less than 2% of
C ircle N o . 3 8 7 o n reader serv ice card. t e s t s to f in a l p r o d u c t la b e l. the daily value of the nutrient
M e d a llio n a n a ly z e s fo o d s for
Rapid Nucleic Acid and Protein a d d itiv es, re s id u e s o r to x in s.
Purification with New BioSpin M ic ro sco p y an d m ic ro -
Columns b io lo g ica l tests are g
a lso p erfo rm e d .
N e w B io -S p in 6 and B io -S p in 3 0 C a ll M e d a llio n fo r
ch rom atography co lu m n s are p re-p ack ­ th e a n s w e rs y o u n eed . J
aged , rea d y -to -u se spin c o lu m n s fo r 1-8 0 0 -2 4 5-56 15 i
rapid and e ffic ie n t d esa ltin g o f b io ­
m o le c u le s. In m in u tes, B io -S p in c o l­
N u tr itio n
um n s w ill clea n up and purify probe,
la b e lin g
n ick translation, p lasm id , and a m p lifica ­ in c a p a b l e
tion reaction m ixtu res. T h ey w ill a lso h a n d s.

d esa lt and cle a n up la b eled and unla-

b eled protein and p ep tid e preparations, M e d a llio n
or r e m o v e un b ou n d d y e s in d y e b in d in g L a b o r a t o r i e s
assa y s. C o m p le te ly au toclavab le, th ese
9000 Plymouth Avenue
d isp o sa b le co lu m n s are p re-p ack ed w ith
Minneapolis, Minnesota 55427
p re -sw o lle n B io -G e l p o ly a cry la m id e g e l
Phone 612-540-4453
filtration b ea d s, a n on -carbohydrate m a ­
trix for op tim u m reco v ery o f sam p le.
CIRCLE 86 ON READER SERVICE CARD
J ournal O f A O A C International V ol . 76, N o. 6 1993 165A
 Be Sure Your Laboratory Uses the Best,
Most Current Analytical Methods Available...

Official M ethods o f A n alysis

o f the
15th Edition

Save Time and Money, Achieve M ore ed ition s, th ese are p u b lish ed in an n u al In Volum es I & II:
Accurate A n a lysis, and Im prove su p p lem en ts w h ich are sen t, at no
A ppendixes: Standard S olu tion s and
Y our Lab ora tory’s Reputation addition al c o st to p u rch asers of the
Certified Reference M aterials, Laboratory
m ost recen t edition .
The dem on strated reliab ility of AOAC Safety, R eference Tables. Su b ject and
m eth od s prom pts th eir u se w h en ever a Thus, AOAC’s OFFICIAL METHODS OF M ethod N um ber In d exes.
need for a n a ly sis arises in regulatory, ANALYSIS n ot only brings you up to
A O A C -F or over 100 years, the m ission
research, and quality control activities. date, it keeps you there!
of the A ssociation of Official A nalytical
C hem ists, m icrobiologists, and other C hem ists has b een to provide th e a n a ­
sc ie n tists dep en d upon Official M ethods CONTENTS lytical sc ie n c e com m u n ity w ith fully
of A nalysis of th e AOAC b eca u se th ey validated standard m eth od s for the
Volume I-A gricultural Chemicals,
know the reliab ility and reproducibility chem ical and biological analysis of foods,
Contaminants, Drugs
of AOAC m eth od s have been d em o n ­ Agricultural Liming Materials drugs, cosm etics, p esticid es, fertilizers,
strated through thorough testin g by Fertilizers feeds, hazardous su b sta n ces, air, water,
interlaboratory collaborative stu d ies Plants soil, and any oth er p roducts and sub­
in governm ent, industry, and u n iversity Feeds sta n ces affecting the public h ealth
laboratories. Drugs in Feeds
and safety, the econ om ic p rotection of
Disinfectants
The authority of th e m eth od s is further th e consum er, or the q u ality of the
Pesticide Form ulations
e n h a n ced by AOAC collab oration w ith Hazardous Substances environm ent.
oth er w orldw ide organizations. Metals & Other Elem ents 1200 Pages. 1990. 2 Volumes. Hard­
Pesticide & Industrial Chemicals bound, 185 illus, ISBN 0-935584-42-0.
Take Advantage o f A O A C M ethods Waters; Salt Binders for supplements are shipped
A vaila b ility and Ease-of-Use Microchemical Methods with the first supplement. $281.00 in con­
Radioactivity tinental North America (USA, Canada,
Over 1,800 AOAC m eth od s are ready for Veterinary Analytical Toxicology
Mexico). $312.00 outside continental
your im m ed iate u se. The easy-to-follow , Cosmetics
North America. Members: subtract 10%
step-by-step form at sp ecifies all reagents Extraneous Materials
discount. Includes 5 annual supplements
and apparatus. And altern ative m ethods Microbiological Methods
Drugs plus a looseleaf binder for their storage.
are often provided to accom odate a w ide
range of laboratory cap ab ilities. Drug & Feed Additives in Animal Tissues
Forensic Methods
To obtain your copy of OFFICIAL METHODS
Each m ethod is referen ced for back­
Volume II-F ood Composition OF ANALYSIS, send your order w ith the
ground and d evelop m en t m aterial, and
Baking Chemicals following: your name and address, a check
con tain s C hem ical Abstracts S ervice Distilled Liquors or Mastercard, VISA, or Diners credit card
(CAS) num bers w here applicable. Both Malt Beverages information (nam e of card, card number, and
ch em ical and com m on n am es are Wines expiration date) and your signature to
provided for all drugs and p esticid es. Nonalcoholic Beverages
Coffee & Tea
Com ply M ore E a sily w ith Cacao Bean
Governm ent Regulations! Cereal Foods
Dairy Products

AOAC
Sin ce th eir in cep tio n in 1884, AOAC Eggs & Egg Products â i m
m eth od s have b een accorded preferred Fish & Other Marine Products
statu s in the courts and are often Flavors
referred to in n ational, state, and Fruits & Fruit Products IN TERN ATIO N AL
Gelatin, D essert Preparations
provincial regulations and governm ental
Meat & Meat Products
and com m ercial sp ecification s.
Nuts & Nut Products
at AOAC-J, 1970 Chain Bridge Road,
Oils & Fats
Be Confident You’re U sin g the Best, Vegetable Products Dept. 0742, M cLean, VA 22109-0742
M o s t C u rren t M ethods Available Spices & Other Condiments (7 0 3 ) 522-3032, FAX (7 0 3 ) 522-5468
Sugars & Sugar Products
As analytical n eed s change and as Vitamins
know ledge and tech n iq u es advance, new Color Additives All orders must be prepaid. Please make checks pay­
and revised m eth od s are con tin u ally Food Additives, Direct & Indirect able to AOAC, U.S. funds on U.S. banks only. Credit
being valid ated by AOAC. B etw een Natural Poisons card orders may be placed by mail, telephone, or FAX.
New Products

the B io -S p in chrom atograp h y co lu m n strip in the m eter. In a m atter o f seco n d s p recision . D e sig n e d for routine ap p lica­
en a b les it to be u sed w ith any 1.5 m L the results are d isp layed d irectly in c o n ­ tion s, fu ll sca le A F M scan ran ges are s e ­
m icro test tube or 12 x 75 m m test tube centration units. T h e R eflectoq u an t m e ­ lecta b le from 5 to 7 5 m icrom eters w ith
in m ost tabletop clin ical or lo w sp eed ter fits in the p a lm o f you r hand, and pro­ m easu rem en t p r e c isio n to <1 A n gstrom .
flo o r m od el c en trifu g es w ith sw in g in g v id es a broad range o f inorganic and A n X -Y sam p le p o sitio n in g stage is
b u ck et rotors. B io -S p in co lu m n s are a lso organic test m ethods: ascorb ic acid, standard, and th e sy ste m is com p atib le
a vailab le em p ty for y ou r o w n sp in ch ro­ chrom ate, cop p er, lead , m olyb d en u m , w ith op tical m ic r o sc o p e s from popular
m atography and gravity flo w ap p lica­ nitrate, nitrite, p eracetic acid, p eroxid e, ven dors. T h e P ersonal A F M can b e e x ­
tions. B io -R a d Laboratories. pH . and m ore. E M S c ie n c e . panded to in clu d e S T M and U H V /S T M
C ircle N o . 3 8 8 on reader serv ice card. C ircle N o . 3 8 9 on reader serv ice card. capability. T he unique T m eC ontour servo
control circuitry m in im izes overshoot and
Introducing the Reflectoquant Personal AFM Microscope undershoot w h en scanning over sharp fea­
Analysis System tures. S cans are autom atically linearized
B u rleig h Instrum ents in trod u ced the w ith active feedback during the scan so
E M S c ie n c e has introduced th e R e fle c ­ n e w lo w c o st and ea sy to u se Personal that all im ages are undistorted and all
toquant A n a ly sis S y stem for test strip, A F M m icro sco p e that can q u ick ly im ­ m easurem ents are accurate and reliable.
in clu d in g colorim etric, tests. To u se the a g e the su rface o f virtually any solid or The Burleigh Personal A F M is supplied
R e flecto q u a n t sy stem , load th e test se m i-so lid m aterial w ith ou t sam p le com p lete w ith W in d ow s b ased True-Im-
m eth od from a b ar-cod e, dip a test strip preparation, and p rovid es subnanom eter age software. B urleigh Instruments, Inc.
in the sam p le, and insert th e e x p o se d im a g e resolu tion and 3 D m easu rem en t C ircle N o . 3 9 0 on reader serv ice card.

Ç 7A & /O tft/y

y tO tA fâ 'S T n te r /ia tia r ia /

For Analytical Science that Solves Today’s Problems and Anticipates Tomorrow’s Needs
Symposia: Other Symposia:
A nalytical M e th o d o lo g y fo r T o d a y ’s Issues: ■ E nvironm ental Issues A ssociated w ith Fertilizers
■ A nalytical M eth o d s for H erbal P roducts ■ Chem ical Analyses R elated to Oil Spills
■ Analytical M eth o d s for the D eterm ination o f
E co n o m ic A dulteration (all foods) Forums:
■ M ethods o f A nalysis for Processed Seafood
■ F orum o n A ntibiotics a n d D ru g s in Feeds
Products, inclu d in g D eco m p o sitio n
■ F o ru m b y th e T echnical C o m m ittee for Juice and
■ A nalytical M eth o d s for F erm entation Processes
Juice P roducts
L ab M a n a g e m e n t S eries:
■ L aboratory Safety
■ A utom ated M ethods Open Forum
T e c h n o lo g y o f th e F u tu re : Laboratory Equipment Exposition
■ Capillary E lectrophoresis
Modern L aboratory Workshop
W iley A w ard S y m p o siu m :
■ T o be determ ined by 1994 W iley A w ard W inner Short Courses
O ld F rie n d s, N ew E n e m ie s S eries: Plus ... more than 200 poster sessions
■ Biom edical W aste T reatm ent T echnology and
A ssessm ent M ethodology, Inclu d in g R egulatory
R oundtable as p art o f the Sym posium

F or m ore inform ation, c o n ta ct A O A C IN T E R N A T IO N A L The Scientific Association A O A C

M eetings D epartm ent: +1-703-522-3032 o r F a x +1-703-522-5468 Dedicated to Analytical Excellence INTEIiNAIK>NAL

Jo u rn a l O f AOAC In t e r n a t io n a l Vol. 76, No. 6 1993 167A

F o r Y ou r In fo r m a tio n
Meetings ture, N utrient C o m p o sitio n Laboratory, w a s h on ored as b ein g o n its w a y to b ein g
B e lts v ille , M D 2 0 7 0 5 , telep h o n e +1 the first su b sectio n o f A O A C IN T E R ­
(3 0 1 ) 5 0 4 -8 9 2 7 . N A T IO N A L by its parent sectio n , the
November 16-18, 1993: A O A C
B oard o f D irectors M eetin g , A O A C , A r­ May 2-M, 1994: A O A C N ortheast S ec­ E urope S ectio n , represented b y E urope

lin gton , V A . C ontact: F a y e H. N orth, tion M eeting, Durham , N H . Contact: E d­ S ectio n Secretary-T reasurer E llen Jan

A O A C , 2 2 0 0 W ilso n B lv d , S u ite 4 0 0 , m ond J. Baratta, U .S . F ood and D rug A d ­ deV ries. R ep resen tatives fro m P oland,

A rlin gton , V A 2 2 2 0 1 -3 3 0 1 , +1 (7 0 3 ) ministration, 109 H olton St, W inchester, H ungary, and the C z e c h and S lo v a k R e ­

5 2 2 -3 0 3 2 . M A 0 1 8 9 0 , telephone +1 (6 1 7 ) 7 2 9 -5 7 0 0 . p u b lics adopted b y la w s and u nani­

November 25, 1993: A O A C M id - June 13-15, 1994: A O A C M id w e st m o u sly e le c te d an e x e c u tiv e c o m m it­

C anada S e c tio n M eetin g , M anitoba, S e c tio n M e e tin g , C olu m b ia, M O . C on ­ tee— step s required b efo re o fficia l

C anada. C ontact: M ark G . T orchia, St. tact: G eorge R ottin gh au s, U n iv ersity o f chartering. A recep tion fo llo w e d , h osted

B o n ifa c e G eneral H osp ital, D epartm ent M isso u ri-C o lu m b ia , V eterinary M ed ica l by P ragu e’s E c o lo g ic a l Institute and

o f Surgery, 4 0 9 Tache A ve, W in n ip eg, D ia g n o stic L aboratory,PO B o x 6 0 2 3 , sp on sored b y the E urope S e c tio n . T h e

M B , R 2 H 2 A 6 , C anada, telep h o n e +1 C olu m b ia, M O 6 5 2 1 1 , telep h o n e +1 celebration provid ed a backdrop for the
(2 0 4 ) 2 3 7 -2 5 6 8 . (7 0 3 ) 5 2 2 -3 0 3 2 . m eetin g, w h ich a lso featured a scien tific

February 2-3, 1994: A O A C S ou th ­ June 1994: A O A C P a c ific N orth w est program o n recent strategies on quality
S e c tio n M eetin g , O ly m p ia , W A . C o n ­ assurance in fo o d a n a ly sis.
ea st U S A S ectio n M eetin g , A tlanta, G A .
tact: Isabel C . C ham berlain, U .S . E n v i­ T h e n ex t step s in the estab lish m en t o f
C ontact: D o u g H ite, T en n essee D epart­
ronm en tal P rotection A g e n c y , R eg io n the group as an o ffic a l su b sectio n o f
m en t o f A gricu ltu re, T ech n ical S erv ice s,
10, 7411 B e a c h D r E ast, Port Orchard, A O A C IN T E R N A T IO N A L are form al
P O B o x 4 0 6 2 7 M elr o se Station, N a sh ­
W A 9 8 3 6 6 , telep h o n e +1 (7 0 3 ) 5 2 2 - approval b y th e E x e c u tiv e C o m m ittee o f
v ille , T N 3 7 2 0 4 , telep h o n e +1 (6 1 5 )
3 6 0 -0 3 1 4 . 3032. the E urope S ectio n and chartering b y the

February 23-25, 1994: A O A C P a­ September 12-15, 1994: T h e 108th B oard o f D irectors o f A O A C IN T E R ­

AOAC IN T E R N A T IO N A L A nn u al N A T IO N A L , both e x p e c te d to take
c ific S o u th w est S e c tio n M eetin g , N apa,
M e e tin g and E x p o sitio n , Portland, O R. p la ce w ith in the n ext fe w m onths.
C A . C ontact: Terry Jackson, C alifornia
C on tact M argaret R id g ell, A O A C , 2 2 0 0 Jana H a jslo v a , U n iv ersity o f Prague,
D ep artm en t o f F o o d and A griculture,
W ilso n B lv d , S u ite 4 0 0 , A rlin gton , V A w h o w a s e lec ted the C entral E urope
3 2 9 2 M e a d o w v ie w R d, S acram en to, C A
9 5 8 2 3 , telep h o n e +1 (9 1 6 ) 2 6 2 -1 4 3 4 . 2 2 2 0 1 -3 3 0 1 , +1 (7 0 3 ) 5 2 2 -3 0 3 2 . S u b sectio n presid en t, ex p la in e d that a
March 20-23, 1994: A O A C S outh- September 29-30, 1994: A O A C steerin g group d e c id e d on th e form ation
E urope S e c tio n , N yon, S w itzerland. o f a su b section rather than a separate
w e st-U S A S e c tio n M eetin g , S an A n to ­
C ontact: T. R ih s, S w iss Federal R e ­ sectio n , at its first m eetin g in N o v e m b e r
n io , T X . C ontact: Jam es E. B althrop, O f­
search Station s for A n im a l P roduction, 19 9 2 , b e c a u se “E urope is n o w o n e, and
fic e o f th e T exas State C h em ist, P O B o x
3 1 6 0 , C o lle g e Station , T X 7 7 8 4 1 , te le ­ C H -1 7 2 5 P o sie u x , S w itzerlan d , te le ­ sh ou ld n ot again be sp lit in E ast and
p h o n e +1 (4 0 9 ) 8 4 5 -1 1 2 1 . p h o n e +41 3 7 8 7 7 111. W est. T h e se d iv isio n s are b eh in d u s,”
March 21-25, 1994: A O A C Short said H ajslova, “w e w o u ld lik e to w ork
C ou rses, S an A n to n io , T X . T opics: Q A Scientists from Poland, w ith our c o lle a g u e s in all o f E u rop e.”
for A n a ly tica l L aboratories, ISO 9 0 0 0 , Hungary, the Czech and Slovak T h e steerin g group, co n sistin g o f
S tatistics for M e th o d o lo g y D e v e lo p ­ Republics, and Slovenia J o se f K alas o f B ratislava, S lo v a k R e ­
m ent, A n a ly tic a l M eth od D e v e lo p m e n t Organize First AOAC Subsection pu b lic, and Jana H a jslo v a and V la d im ir
(tentative), G o o d L ab P ractices (tenta­ K ocou rek , both o f P rague, C z e c h R e ­
tiv e ). C ontact: C arol R o u se , A O A C IN ­ A p rop osed C entral E urope S u b section , pu b lic, reported that H ungary, P oland,
T E R N A T IO N A L , 2 2 0 0 W ilso n B lvd , co n sistin g of H ungary, P olan d , the the C z e c h and S lo v a k R e p u b lics, and
S u ite 4 0 0 , A rlin gton , V A 2 2 2 0 1 -3 3 0 1 , C zech and S lo v a k R ep u b lics, and S lo v e n ia are all b u ild in g up their e c o n o ­
telep h o n e +1 (7 0 3 ) 5 2 2 -3 0 3 2 . S lo v en ia , w a s organ ized at a m eetin g in m ie s, and. th erefore, share co m m o n
April 17-21, 1994: Sixth Interna­ Prague, C z e c h R ep u b lic, S ep tem b er 1 3 - p rob lem s and interests in the laboratory.
tion al S y m p o siu m on B io lo g ic a l and 1 4 ,1 9 9 3 . T h e m o st urgent p rob lem is the e c o n ­
E n viron m en tal R eferen ce M aterials T h e A O A C C entral E urop e S u b se c ­ o m y o f th ese cou n tries, w h ic h puts a m a­
(B E R M -6 ), K ona, HI. C ontact: W ayn e tion o f th e E urope S e c tio n o f A O A C I N ­ jo r lim it o n scien tific w ork and e x p a n ­
R. W olf, U .S . D ep artm en t o f A g ricu l­ T E R N A T IO N A L , as o ffic ia lly nam ed. sion , said M argreet L auw aars, A O A C

168A Jo u rn a l Of AOAC I n t e r n a t io n a l Vol. 76, No. 6, 1993

For Your Information

E uropean R ep resen tative. T h is problem ■ President: Jana H ajslova, Prague, AOAC Nutrition Labeling Book
w ill a lso a ffect the C entral E urop e S u b ­ C z e c h R ep u b lic Published 1
sectio n , certainly w h ere hard currency is ■ P resident-E lect: A n to n K ocan ,
n eed ed to p ay for A O A C m em b ersh ip B ratislava, S lo v a k R ep u b lic
W ith the M a y 1 9 9 4 d ea d lin e for c o m p li­
and ed u cation p rogram s, su ch as A O A C ■ Secretary-Treasurer: V lad im ir
an ce w ith the N utrition L a b elin g and
short cou rses. E v e n the organ ization o f K ocou rek , P rague, C z e c h R ep u b lic
E ducation A c t (N L E A ) nearing, A O A C
interlaboratory stu d ies is n ot e a sily real­ ■ M em bers: IN T E R N A T IO N A L recen tly p u b lish ed
iz e d in C entral E urope, b e c a u se m o st • Ilo n a B o r o s, H ungary a b o o k d e sig n e d to h elp an alysts select
laboratories w ill o n ly participate for f i­ • M ie c z y s la w O b ied zy n sk i, the proper O fficia l M eth o d fo r fo o d la­
n an cial com p en sation . P olan d b elin g.
A O A C w a s rep resented at the organ­ • M aria V aradi, H ungary R e le a se d O ctob er 2 2 , Methods of
iza tio n a l m eetin g by L auw aars and • P a v e l F arkas, S lo v a k R ep u b lic Analysis fo r Nutrition Labeling con tain s
Jam es F. L aw ren ce, H ea lth and W elfare • O n e addition al m em b er from 281 A O A C O ffic ia l M eth o d s appropri­
C anada, w h o a ls o p resen ted a paper dur­ S lo v e n ia is to b e nam ed. ate for nutrition la b elin g , w h ic h are
in g the sc ie n tific p rogram o n q u ality as­ T h e n e x t m e e tin g o f the C entral grou p ed b y an alyte for e a sy referen ce.
surance and interlaboralory stu d ies in E urop e S u b sectio n w ill be h eld in B ra­ T h e b o o k a ls o con tain s chapters w ritten
trace an alysis. tislava, S lo v a k R ep u b lic, O ctob er 1994. b y exp erts in the fie ld c o v e r in g variou s
T h e e x e c u tiv e c o m m ittee o f the C en ­ T h e top ic o f the sc ie n tific program w ill to p ics pertinent to nutrition lab elin g.
tral E urope S u b se c tio n e le c te d for b e laboratory quality assurance, refer­ A c co rd in g to co-E d itor D arryl S u lli­
1 9 9 3 -1 9 9 4 is as fo llo w s: e n c e m aterials, and accreditation. van, H a zleto n W isc o n sin , “T o date,

H elp Y ou r P rofession
N om in ate a D eserving In d ivid u a l f o r the
1995 H arvey W. W iley A w a rd
P le a s e d o y o u r p a r t to e n s u r e th a t s c ie n tif ic d e d i­
W e n e e d y o u r h e l p !I
c a tio n a n d e x c e lle n c e d o e s n o t g o u n r e c o g n iz e d .

E a c h y e a r A O A C p resents the H a r v e y W . W ile y N o m in a te so m e o n e to d a y !

A w a r d fo r the D e ve lo p m e n t o f A n a ly tic a l M e t h o d s
to a n o u t s ta n d in g s c ie n tis t o r s c ie n tific te a m fo r M a r c h 3 1 s t is t h e c l o s i n g d a t e f o r n o m i n a t i o n s
a n a ly t ic a l c o n t r ib u t io n s in a n a r e a o f in t e r e s t to f o r t h e 1 9 9 5 a w a r d .*
AOAC IN T E R N A T IO N A L . The 1995 cash * N o m in e e s w ill c o n tin u e to b e e lig ib le f o r th r e e a d d itio n a l
a w a r d w ill b e $ 4 ,0 0 0 . y e a r s w i t h o u t r e n o m i n a t i o n a n d th e ir e lig ib ility m a y b e
e x te n d e d a n a d d it i o n a l f o u r y e a r s b y w r i t te n r e q u e s t o f
th e n o m in a t o r .
W e know t h e r e a r e s c i e n t i s t s o u t t h e r e , in a n d
o u ts id e o f A O A C , w h o d e s e r v e to F or m o re in fo rm a tio n o r a p p lic a tio n fo rm s,
b e r e c o g n iz e d f o r th e ir a c h ie v e m e n ts c o n ta c t: M a n a g e r, M e m b e rs h ip & M e m b e r
A i m Services, A O A C IN T E R N A T IO N A L , 2 2 0 0
in a n a l y t i c a l s c i e n c e . B u t l a t e l y y o u
W ilso n B o u le v a rd , S u ite 4 0 0 , A rlin g to n , VA
h a v e n o t b e e n id e n t ify in g o r n o m i­ A O A C 2 2 2 0 1 -3 3 0 1 . T e le p h o n e + 1 - 7 0 3 -5 2 2 -3 0 3 2 ,
INTERNATIO NAL
n a t in g t h e s e d e s e r v in g s c ie n tis ts ! FAX + 1 -7 0 3 -5 2 2 -5 4 6 8 .
The Scientific Association Dedicated to Analytical Excellence
 N ow A vailable from A O A C INTERNATIONAL

T h is n e w b o o k w ill p r o v id e th e u s e r w it h a

HEII f t h o r o u g h k n o w l e d g e o f h o w to te s t fo o d
p r o d u c t s fo r c o m p li a n c e w it h p r o v is io n s o f th e
N u tritio n L a b e lin g a n d E d u c a tio n A c t ( N L E A )
o f 1 9 9 0 , w h ic h m a d e m a n d a t o r y th e n u t r i­
t io n a l c o n t e n t la b e li n g o f n e a r l y a ll p r o c e s s e d
fo o d s. T h e b o o k c o n ta in s a c o n c ise a b stra c t
all AOAC® Official Methods
o n th e N L E A a n d
determined to be acceptable for use in nutrition
labeling.
T h e m e t h o d s a r e l is t e d in a l p h a b e t i c a l o r d e r
b y n u trie n t, a n d d iv id e d in to c h a p t e r s o n
c a r b o h y d r a t e s , m in e r a l s a n d p r o x i m a t e , fat
a n d fatty a c id s, a n d v ita m in s . D is c u s s io n s on
th e u s e o f th e s e m e t h o d s in th e fo o d l a b o r a ­
to ry, c u rr e n t id e a s o n m e th o d a d a p tio n , a n d
th e u se o f m a n y n e w u n o ffic ia l m e t h o d s are
in c lu d e d .
A c h a p t e r o n S ta n d a r d R e fe re n c e M a te ria l
w i l l a s s i s t th e r e a d e r in l o c a t i n g a p p r o p r i a t e
s ta n d a r d s fo r e a c h m e t h o d / m a tr ix c o m b in a ­
tio n a n d p r o v id e s r e c o m m e n d a t io n s o n h o w
to p r e p a r e y o u r o w n r e f e r e n c e m a t e r ia ls .

O c t o b e r 1 9 9 3 . S p e c ia l p r e p u b l i c a t i o n d i s c o u n t b e f o r e
D e c e m b e r 3 1 , 1 9 9 3 : $ 1 0 0 . 8 0 in N o r t h A m e r ic a (U S A ,
C a n a d a , M e x ic o ) , $ 1 1 2 . 5 0 o u t s i d e N o r t h A m e r ic a .
A fte r D e c e m b e r 3 1 , 1 9 9 3 : $ 1 3 9 . 0 0 in C o n t i n e n t a l
N o r t h A m e r ic a , $ 1 5 5 . 0 0 o u t s i d e C o n t i n e n t a l N o r t h
A m e r ic a . A O A C I N T E R N A T I O N A L m e m b e r s : s u b t r a c t
10% d i s c o u n t .

T o o r d e r : S e n d y o u r n a m e , a d d r e s s a n d p a y m e n t to
A O A C IN T E R N A T IO N A L . A O A C a c c e p t s c h e c k s (U S
f u n d s o n U S b a n k s o n l y p l e a s e ) o r V IS A , M a s t e r C a r d ,
a n d D in e r s c r e d it c a r d s. W h e n p a y in g b y c r e d it c a r d
p l e a s e i n c l u d e t h e t y p e o f c r e d it c a r d , c a r d n u m b e r ,
e x p i r a t i o n d a t e , a n d y o u r s ig n a t u r e .

E d it e d B y S e n d to: A O A C IN T E R N A T IO N A L -J , 1 9 7 0 C h a m B r id g e
R oad , D e p t. 0 7 4 2 , M cL ea n , V A 2 2 1 0 9 - 0 7 4 2 .
D A R R Y L M, S U L L I V A N
C r e d it c a r d o r d e r s m a y a l s o b e p l a c e d b y p h o n e :
D O N A L D E . C A R P E N TE R + 1 -7 0 3 -5 2 2 -3 0 3 2 , or FAX + 1 -7 0 3 -5 2 2 -5 4 6 8 .

M U
AOAC
INTERNATIONAL

------------------------------------------------ TheScientific Association Dedicated to Analytical Excellence

170A Journal O f AOAC INTERNATIONAL V o l . 76 No. 6,1993
For Your Information

N L E A is the m o st sig n ifica n t p ie c e o f fatty a cid s from m o n o -, d i-, and w h e n o n e m eth o d con tain s a referen ce
le g isla tio n regu latin g the fo o d industry. triglycerid es, and an y other sou rce. T h e to a secon d referen ce, th is seco n d
N e v e r h as so m eth in g b een s o w id e rang­ L ip id A n alysis chapter e x p la in s this n ew m eth o d is in clu d ed as a tim e -sa v in g fe a ­
in g - a ffectin g sm all, m ed iu m , and large d efin itio n , co m p a res it to other fo o d la ­ ture. A n in d e x is in c lu d e d as w e ll, listin g
b u sin e sse s - in an effort to ed u cate the b e lin g d e fin itio n s for fats and lip id s, and the A O A C O ffic ia l M eth o d n u m b er and
p u b lic. N L E A le ft th e fo o d industry in a lists th e cla ssific a tio n s and ty p es o f fatty th e p a g e o n w h ic h it is found.
h u g e v o id in term s o f h o w to co m p ly , a cid s fo u n d in fo o d s . T h e chapter a lso
and w e h o p e this b o o k w ill f ill the v o id .” d etails e a c h appropriate m e th o d o lo g y Methods Adopted First Action
M ethods o f A n alysis f o r N utrition L a­ and states w h ic h fats or lip id s are a ccu ­
beling c o n ta in s chapters o n su ch top ics rately m easured. ■ D ru g s a n d R e la te d Topics: 993.32
as international ap p roaches to fo o d la­ T h e chap ter C a rb o h yd ra te/D ieta ry M u ltip le S u lfo n a m id e R e sid u e s in B o ­
b e lin g , a c o n c is e exp lan ation o f N L E A F ib er A n a lysis ex p la in s th e term in o lo g y v in e M ilk , L iq u id C hrom atograp h ic
p ro v isio n s, and the fin al report o f the b eh in d the N L E A d e fin itio n o f sugars, M eth o d w a s ad op ted first action b y a
A O A C T ask F o rce o n M eth o d s for N u ­ in clu d in g th e tro u b leso m e term “c o m ­ b allot c o m p le te d on S ep tem b er 15,
trient L a b elin g A n a ly se s. T h e task fo rce, p le x carboh yd rates.” T h e chapter a lso 1 9 9 3 . T h is m eth o d w ill b e p u b lish ed in
form ed in early 1 9 9 2 and c o n sistin g o f groups the appropriate su gar a n a ly ses b y the fifth su p p lem en t (1 9 9 4 ) to the 15th
m em b ers o f g o v ern m en t a g e n c ie s , aca­ ty p es o f m e th o d o lo g y , in c lu d in g co lu m n ed itio n (1 9 9 0 ) o f O fficial M eth ods o f
d em ia, p rivate co m p a n ie s, and trade as­ chrom atography, colorim etric m eth od s, A n alysis.
so cia tio n s, w a s instrum ental in d efin in g and e n z y m a tic -c h e m ic a l m eth od s.
w h ic h o ffic ia l m eth od s are accep tab le O n e o f th e m o re n o v e l chapters c o v ­ Method Modification Received
for nutrition la b elin g and w h ic h areas ers th e n e e d for referen ce m aterials to
n eed further m eth o d d ev elo p m en t. B o th v erify the accu racy o f n e w or m o d ifie d ■ M ic ro b io lo g ica l M ethods: 990.13
o f th ese lists are in clu d ed in th is p u b lica ­ m eth o d s, as w e ll as other id e a s d is­ Salm on ella in F o o d s, C olorim etric D e -
tion. cu sse d b y th e task fo r c e and its su b c o m ­ o x y n u c le ic A c id H yb rid ization M eth od
A ls o in clu d ed are in d ivid u al chapters m ittees. O n e o f th e se id ea s p resen ts a (G E N E -T R A K )
ad d ressin g a n a ly ses o f lip id s, carbohy- u n iq u e w a y o f p e r c e iv in g th e collab ora­ ■ M etals a n d O th er Elem ents a t Trace
drates/dietary fiber, p roxim ate and m in ­ tiv e study p ro cess. Instead o f d e fin in g a Levels in Foods: 9 7 3 3 2 C adm ium and
erals, fa t-so lu b le v ita m in s, and w ater- range o f fo o d s o n a c o m m o d ity b a sis, L ead in Earthenware, A to m ic A bsorption
so lu b le vitam in s. A ll o f th ese chapters the ch e m ic a l variation o f fo o d s is u sed to Spectrophotom etric M eth od
detail th e appropriate A O A C O fficia l lim it th e num ber o f sa m p les that w o u ld
M eth o d s for in d ivid u al analytes and represent all fo o d s. T h is variation is d e ­ Beta-Lactam Test Kit Evaluation
state p lu se s and m in u se s o f the different fin ed b y fat, protein, and carbohydrate Completed
m e th o d o lo g ie s u sed . In all o f th ese concen tration . T h erefore, a sa m p le c o n ­
chapters, w h en m o d ific a tio n s o f A O A C taining a sim ila r ch e m ic a l m ak e-u p can T h e A O A C R esea rch Institute and the
O fficia l M eth o d s are n e e d e d to co m p ly b e u sed fo r eith er ch ed d ar c h e e s e or F D A C en ter fo r V eterinary M e d ic in e
w ith N L E A , the m o s t current adapta­ brow n m ustard, ea ch o f w h ic h con tain s c o m p le te d th eir jo in t e v a lu a tio n o f beta-
tion s are d iscu ssed . sim ilar am ounts o f fat, protein, and car­ la cta m test k its su b m itted in January
In sev era l k e y areas o f an alysis, b oh ydrates. Currently, 2 d ifferen t c o m ­ 1 9 9 3 . S e v e n b eta-lactam test k its have
N L E A requires a n o v e l approach, and m o d ity -b a sed sa m p les w o u ld n e e d to b e been r eco m m en d ed fo r A O A C R e­
M ethods o f A nalysis f o r N utrition L a b el­ u sed . T h e chapter a lso p ro v id es g u id e ­ search Institute “P erform an ce T ested”
ing p rovid es so m e m u ch n e e d e d g u id ­ lin e s fo r the preparation and u se o f in- certification and are fisted b e lo w w ith
an ce. T h e chapters on lip id an d sugar h o u se q u ality assu ran ce con trol m ateri­ their va lid a ted se n sitiv ity le v e ls . A ll
a n alyses w e r e c o m p ile d to reflect the als. b eta-lactam test k its w ere evalu ated a c­
new d efin itio n s fo r th e se analytes M eth ods o f A n a lysis f o r N utrition L a ­ cord in g to a va lid a tio n p ro to co l d e v e l­
adop ted b y the U .S . F o o d and D ru g A d ­ beling is a hardbound m an u al con tain in g o p ed jo in tly b y th e A O A C R esea rch In­
m inistration in N L E A . o v er 6 2 0 p a g e s. E a ch O fficia l M eth o d is stitute and the FDA C en ter for
T h e fin al r a le s o f N L E A d e fin e fat as referen ced b y the O ffic ia l M e th o d n u m ­ V eterinary M e d ic in e . A d d itio n a l test kits
the su m o f fatty acid s ex p ressed as ber and a sp e c ia l chapter num b er for are still u nder evalu ation . T h e F D A C en ­
triglycerid e, in clu d in g free fatty acid s, ea sy cro ss-referen cin g. In m o st c a se s, ter for V eterinary M e d ic in e and the

Jo u r n a l O f AOAC I n t e r n a t io n a l Vol. 76, No. 6,1993 171A

E s s e n t ia l F o r Y o u r Q u a l it y A s s u r a n c e P r o g r a m ...

i A O A C.

Q U A l I T Y
i
| a s s u r a n c e

' p r i n c i p l e s

A N A L Y T I C A L

LABORATORIES

/ E s s e n t ia l for any lab w a n tin g to im prove or initiate a q u ality assu ran ce (Q A ) program .
M aterial w as e xte n sive ly revised for th is n e w edition w h ic h features n e w in form ation and
concepts developed since the h an d b o o k w as last published. A n im portant ad d ition is a chap ter
on u tilizin g statistical ap p lication s and analytical control c h artin g techniques. A p p e n d ix e s, too,
have been revised and one added on laboratory accreditation criteria — crite ria w h ic h can be
used for self-evaluation o f lab Q A p rogram s and op e ration s m anagem ent procedures.
E a c h chapter offers recom m endations for developing and operating a Q A program. T h e book
also provide s solid ju stifica tio n for com m itm ent of re sou rce s to a q u ality assu ran ce program .
C o n te n ts: Q u ality A ssu ra n c e Planning. Statistical A p p lic a tio n s and C o n tro l Charts.
P e rso n n e l C onsiderations. M an age m e n t o f E q u ip m e n t an d Su pp lies. Sam p le and Record
H and ling. S a m p lin g and Sam p le A n a ly sis. P ro fic ie n cy and C h e c k Sam ples. A u d it Procedure.
D e sig n and Safety o f Facilities. La b o ra to ry Accreditation.
A p p e n d ix e s : T yp ical C on te n ts o f a Q u ality M a n u a l for T e stin g Laboratories. F o rm s U sed
by U.S. Fe d era l Agencies. In stru m e n t Perform ance Checks. F D A A u d it M e a su re Procedures.
P roficie n cy and C h e ck Sam ple Program s. A ccred itation Criteria.
192 p a g e s . 2 n d e d itio n . M a y 1991. S o ftb o u n d . IS B N 0-93 5 5 8 4 -4 6 -3 .
$ 6 3 .0 0 I n C o n tin e n ta l N o r th A m e ric a (U S A , C a n a d a , M exico)
$ 6 9 .0 0 O u ts id e C o n tin e n ta l N o r th A m e ric a

T o o rd e r:S e n d y o u r nam e and ad d ress and paym ent. A O A C In te rn a tio n al accepts ch e cks (U S
fu n d s on U S b an ks only, p le ase ) and V IS A , M a ste rC a rd or D in e rs credit cards. W h e n p a y in g by
credit card please include: type of credit card, card num ber, e xp iration date and y o u r signature.

S e n d to: A O A C In te rn a tio n al - J ____________________________________ ____________

1970 C h a in B rid ge Road C redit card orders m ay also be placed b y p hone
Dept. 0742 + 1 (7 0 3 ) 522-3032, or F A X + 1 (7 0 3 ) 522-5468.'
M cLe an , V A 22109-0742

Members: subtract 10 % discount

For Your Information

AOAC Research Institute will certify fects of contamination of milk by so­ AOAC Research Institute Perform­
these test kits after the manufacturers matic cells and bacterial cells or cross­ ance Tested certification is effective for
submit revised labeling for approval. reactivity, or any other limitations dis­ 1 year, after which the test kit is re­
The beta-lactam validation protocol covered during the independent viewed to ensure the original conditions
specifically applies to raw, commingled evaluation of the test kit. of the certification are still appropriate.
milk (bulk tank milk) and requires the Under the requirements of the valida­
collection of data on sensitivity, selec­ tion protocol, beta-lactam test kits must Correction
tivity, dose response, ruggedness, cross­
exhibit a minimum sensitivity rate of
reactivity, adverse affects of contamina­ The Jou rnal o f AOAC INTERNA­
90% at or below established tolerance
tion of milk by somatic cells and TIONAL (1993) Volume 76, Number 4,
and/or safe levels for each FDA ap­
bacterial cells, lot-to-lot consistency, “Ochratoxin A in Cow’s Milk and in Hu­
proved beta-lactam drug claimed on the
and stability. Sensitivity, selectivity, man Milk with Corresponding Human
dose-response, and cross-reactivity re­ product labeling, and approved test kit
Blood Samples,” page 842, abstract, line
sults submitted by manufacturer were labeling must reflect the specific beta- 14, ng/mL should be ng/L; page 843,
verified by an independent testing labo­ lactam drugs claimed as detectable, i.e., column 1, line 8, 10% boron should be
ratory, and the results are required to ap­ “Beta-Lactam Test for Penicillin G, 20% boron; page 843, column 2, line 5,
pear on the test kit labels. Test kit pack­ Cloxacillin, Ampicillin, and Amoxi­ 5 mL ethanol should be 5 mL methanol;
age inserts were reviewed to ensure that cillin” or “Beta-Lactam Test for Cefdo- and page 845, table 2, Ochratoxin A,
users are warned about any adverse af- fur and Cephapirin.” mg/L should be Ochratoxin A, ng/L.

AOAC Research Institute“Performance Tested”certifiedtestkitswith validated beta-lactam detection claims

Manufacturer Test kit Ampicillin Amoxicillin Ceftiofur Cloxacillin Cephapirin Penicillin

Gist-Brocades Delvo-X-Press 10 ppb 10 ppb 10 ppb no claim 10 ppb 5 ppb

Delvotest SP* 8 ppb 8 ppb 50 ppb no claim 8 ppb 3 ppb
LacTek Beta-
Lactam 8 ppb 10 ppb no claim 8 ppb 16 ppb 5 ppb
Idetek Lactek Ceftiofur no Claim no claim 50 ppb no claim no claim no claim
Idexx SNAP 10 ppb 10 ppb 50 ppb no claim 8 ppb 5 ppb
Smith-Kline Beecham Penzyme III 10 ppb 8 ppb no claim no claim 8 ppb 5 ppb
Penzyme Milk 10 ppb 8 ppb no claim no claim 8 ppb 5 ppb
FDA safe and/or tolerance level 10 ppb 10 ppb 50 ppb 10 ppb 20 ppb 5 ppb

1 The Gist-Brocades Delvotest SP was not evaluated by the FDA Center for Veterinary Medicine, because the test kit claims to detect sulfona­
mide drugs above FDA safe and/or tolerance level. However the Delvotest SP was validated for beta-lactam drugs by the AOAC Research
Institute according to beta-lactam validation protocol.

COMING IN THE NEXT ISSUE

■ KEYNOTE ADDRESS

ISO 9000.. .The Foundation for Good Business Relationships — I.R. Waples

■ PRESIDENT’S ADDRESS

The Changing Face of AOAC — H.B.S. Conacher

■ TRANSACTIONS OF THE 107TH ANNUAL AOAC INTERNATIONAL MEETING AND EXPOSITION

Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3 173A

 NEW EDITION!
This manual contains analytical methods to detect
FDA microorganisms and certain of their metabolic
products. Although the methods are primarily
for foods, the manual also contains methods for
B a c te r io lo g ic a l analysis of beverages and cosmetics. For use in
regulatory and industry laboratories. Most
chapters have been revised, expanded and
A n a ly tica l updated.
Not all procedures in this manual have achieved
official AOAC status through collaborative testing;
but all represent the methodology currently in
M an u al use in FDA laboratories.
CONTENTS
(BAM), 7th Edition Food Sampling and Preparation of Sample
Homogenate. Microscopic Examination of Foods.
Aerobic Plate Count. E sch erich ia c o li and the
Coliform Bacteria. Salm onella. Shigella. C am pylo­
ba cter. Y ersin ia e n te ro c o litic a and Y ersin ia
M e th o d s c u r r e n tly in
p seu d o tu h ercu lo sis. V. ch olerae, V. p a ra h a em o -
u s e in U S . F o o d a n d lyticu s, V. vu ln ificu s and Other V ibrio spp.
D r u g A d m in is tr a tio n h ysteria m on ocytogen es. Serodiagnosis of L isteria
m on ocytogen es. S ta ph ylococcu s au reu s.
L a b o r a to r ie s Staphylococcal Enterotoxins. B a cillu s cereu s.
Diarrheagenic Enterotoxin. Clostridium perfrin-
gens. C lo strid iu m b o tu lin u m . Yeasts, Molds and
Mycotoxins. Parasitic Animals in Foods. Inhibitory
■G&O
Substances in Milk. Rapid HPLC Determination of
Sulfamethazine in Milk. Examination of Canned
Foods. Modification of Headspace Gas Analysis
Methodology, using the SP4270 Integrator.
Examination of Containers for Integrity. Micro­
biological Methods for Cosmetics. Identification
of Foodbome Bacterial Pathogens by Gene Probes.
Investigation of Food Implicated in Illness.
Appendixes: Rapid Methods for Detecting
Foodbome Pathogens. Most Probable Number
Determination. Media and Reagents.

529 pages. Seventh Edition. 1992 publication.

Ulus. Appendixes. 3 hole d rill w ith binder.
$99-00 in Continental North America
$1 2 2 .0 0 outside Continental North America
Members: subtract 10% discount and include member number.

TO ORDER: send your name and address and payment. AOAC

International accepts checks (US binds on US banks only, please) and
credit cards: VISA, MasterCard or Diners. When paying by credit card
please include: type o f credit card, card number, expiration date and
your signature. â i m
SEND TO: AOAC International - J, 1970 Chain Bridge Road, Dept. 0742,
McLean, VA 22109 0742. A O A C
Credit card orders may also be placed by phone: + 1 (7 0 3 ) 522-3032, INTERNATIONAL
or FAX + 1 (703) 522-5468.
T h a n k s t o R e v ie w e r s

Each year, we take the opportunity to convey our gratitude to the hundreds of individuals who voluntarily set aside valuable time
to assist the Journal o f A O A C In tern ation al by reviewing submitted manuscripts. The reputation of the scientific journal rests on
the quality of the science reported therein. Those of you listed below have helped immeasurably to assure the reputation of the
Jou rnal o f A O A C In ternational by insisting on high standards of experimental design, method execution, and data interpretation.
Please accept our thanks for your contribution to fulfilling AOAC’s mission.

Ackman, R. G. Bennett, Glenn Callahan, J. Del Grosso, Alf

Agarwal, Vipin Betageri, G. V. Campbell, Harold Desjardins, Ann
Ahmad, I. Betz, Joe Caputti, Arthur Desmarchelier, James M.
Ahmad, Nazir Beveridge, Tom Cardarelli, Nate F. Deutsch, Mike J.
Akerblom, M. J. Bicking, Merlin Carlson, Michael Doerge, Daniel
Akkara, Joseph Bicsak, Ronald Carpenter, Don Dogheim, Salwa
Albaladejo, J. Bigalli, Giovanni Carter, Stephen Dombrink-Kurtzman,
Albert, Richard Bishop, J. Russ Casper, Howard Mary A.
Alexander, Thomas Bjorklund, Harry Cervera, Luisa Doner, Landis W.
Allen, J. L. Blass, Wolfgang Chafetz, Lester Donnelly, J. R.
Alliers, J. J. L. Blittman, S. I. Chamkasem, Narong Donoho, Alvin L.
Allied, Maryann Blumenthal, M. Chaput, Daniel Domer, Joe W.
Altstein, Miria Boatman, John A. Chase, G. William Dowling, Thomas
Anderson, Rollen Boese, Jack L. Clark, C. Randa Draper, William
Andrews, Wallace Bolygo, Elek Clark, Charles Dreyfuss, P.
Ard, Edwina Bos, K. D. Class, T. J. Driani, Julia A.
Asp, Nils-Georg Bousquet, William Cohen, Steven A. Dube, David G.
Asselin, Jaques Bradley, Robert Cole, Richard J. Dugar, Sumer M.
Atreya, Naresh Brandon, David Coleman, Margaret Duinker, Jan C.
Ault, James Braselton, W. E. Coleman, Mark R. Duke, Steve
Bachman, William Brasselton, W. Coleman, Pam Dyer, Randolph
Bailey, J. Stan Brause, Allan R. Collier, Richard Dyminski, John
Baker, Randall Brennecke, R. Conacher, Henry Eckert-Tilotta, Sally
Baker, Ronald L. Bright, Susan A. Crocker, Ron Edgell, Kenneth
Balasubramanian, N. Brinkman, U. A. Cunningham, J. Egelhofer, David
Barcelo, Damia Brodsky, Michael Czuczwa, Jean M. Elkins, Edgar R.
Barker, Steven Brooks, Marvin Dabbah, R. Emmons, D. B.
Barnes, Charlie Brumbaugh, William G. Dabeka, Robert Epstein, Mike
Barnes, Ramon M. Brumley, William Dacus, Steve Erickson, Dennis
Barry, Carla P. Brunner, Charlotte Daniels, Karen Espinosa-Mansilla, A.
Barry, Thomas L. Buchanan, Robert Danielson, James Falcone, F.
Bauer, John Bueno, Mark P. Danielson, Neil Farwell, Sherry
Baumann, R. A. Buglio, Benjamin Daun, Henry Favell, D. J.
Bean, G. Bushway, Rodney Davidek, J. Felice, Lawrence
Beasley, Val Buslig, Bela S. Davis, Andy Feng, Paul C. C.
Behrens, W. A. Butterwoith, Fred Davis, Raymond Fernandes, A. R.
Berne, Robert Cabanis, Marie De Vries, Ellen Fiddler, Walter
Belkind, Benjamin Cabras, Paolo DeKok, A. Finney, J. R.
Bellman, Sander Caine, Douglas DeVries, Jonathan Firestone, David
Bennett, Carl D. Cairns, Thomas Defiburgh, Phillip Fish, David L.

Journal O f AO AC International V ol . 76, N o. 6 ,1 9 9 3 175A

r
An Important Reference from
AOAC International

A n y o n e in v o lv e d in f o o d s to r a g e , p r o c e s s ­
in g , d i s t r i b u t i o n o r r e g u l a t i o n w i l l f i n d t h i s
fu lly illu s t r a t e d b o o k to b e t h e e s s e n t ia l s o u r c e
o f in f o r m a tio n o n f o o d p e s t m a n a g e m e n t .

E C O L O G Y A N D It i s a c o m p r e h e n s i v e c o m p i l a t i o n o f t h e
w o r k o f l e a d i n g s c i e n t i s t s i n t h e f i e ld , p r e p a r e d
u n d e r t h e d i r e c t i o n o f t h e U .S . F o o d a n d D r u g
M A N A G E M E N T O F
A d m i n i s t r a t i o n . S t a r t in g w i t h t h e b a s i c s , t h e
b o o k c o n t i n u e s a ll t h e w a y t h r o u g h t h e “ s t a t e
F O O D -IN D U S T R Y o f th e a r t” te c h n iq u e s b e in g e m p lo y e d to d a y .
P o te n tia l u s e r s in c lu d e f o o d in d u s tr y
P E S T S p r o f e s s io n a l s r e s p o n s ib le fo r o r i n t e r e s t e d in
f o o d s a n ita tio n , p e s t c o n t r o l a n d q u a lity
a s s u r a n c e , w o r k in g fo r f o o d p r o c e s s o r s ,
r e ta ile r s , w h o le s a l e r s , s t o r a g e fa c ilitie s ,
im p o r te r s a n d e x p o r te r s , r e s ta u r a n ts a n d o th e r
f o o d s e r v ic e s , f o o d b a n k s; e d u c a t io n a l in s t it u ­
t io n s w it h d e p a r t m e n t s o f f o o d s c ie n c e , a g r ic u l­
tu r e a n d e n to m o lo g y ; a n d r e g u la t o r y a g e n c ie s .
I n a d d i t i o n , it p r o v i d e s a v it a l r e s o u r c e f o r
t h o s e e n g a g e d in p r o a c tiv e e ffo r ts to p r e s e r v e
a n d e n s u r e c le a n a n d a d e q u a te w o r ld fo o d
s u p p lie s .
C o n t e n t s : E c o lo g y , P r e v e n t i o n , S u r v e y a n d
C o n t r o l, H e a l t h A s p e c t s , R e g u l a t i o n a n d
I n s p e c tio n a n d M a n a g e m e n t o f s u c h p e s t s a s
m ic r o o r g a n is m s /d e c o m p o s e r s , m ite s, in s e c t s
( c o c k r o a c h e s , b e e t l e s , s p r i n g t a i l s , m o t h s , f l ie s ,
a n ts ), a n d v e r t e b r a t e s ( r o d e n t s , b i r d s , a n d b a ts );
G lo s s a r y a n d T a x o n o m ic a n d S u b je c t I n d e x e s .

5 9 5 p a g e s . 1 9 9 1 . I llu s tr a te d . H a r d b o u n d .
IS B N 0 - 9 3 5 5 8 4 - 4 5 - 5 .
$ 1 5 0 .0 0 i n C o n t i n e n t a l N o r t h A m e r i c a
$ 1 6 7 .0 0 o u t s i d e C o n t i n e n t a l N o r t h A m e r i c a
M e m b e r s s u b t r a c t 10% d i s c o u n t .

To o rd e r: s e n d y o u r n a m e a n d a d d r e s s , a n d
p a y m e n t . A O A C I n t e r n a tio n a l a c c e p t s
c h e c k s (U S f u n d s o n U S b a n k s o n ly ,
p l e a s e ) a n d c r e d i t c a r d s : V ISA,
M a s te r C a r d o r D in e r s . W h e n p a y in g
b y c r e d it c a r d p le a s e in c lu d e : t y p e o f
J. R. Gorham, Editor c r e d it c a r d , c a r d n u m b e r , e x p ir a t io n
d a te a n d y o u r s ig n a tu r e .

S e n d to: A O A C I n t e r n a t i o n a l - J
1 9 7 0 C h a in B r id g e R o a d , D e p t . 0 7 4 2
M c L e a n VA 2 2 1 0 9 - 0 7 4 2

C r e d it c a r d o r d e r s m a y a ls o b e p l a c e d b y
p h o n e + 1 (703) 5 2 2 -3 0 3 2 , o r FAX + 1 (703) 5 2 2 -5 4 6 8 .
k. J
Thanks to Reviewers

Flowers, Russel Harlin, Karen S. Kelly, Clark Lucy, Charles A.

Fodor-Csorba, Katalin Harmon, Robert Kelly, J. Luke, Milton A.
Fogg, Arnold G. Harriott, M. Keukens, Henk J. MacDonald, Alex
Fong, W. George Harvey, Robert Kevra, Susana MacErlane, Keith
Ford, Joseph H. Haschek, W. M. Kierstead, Todd MacNeil, James
Forsyth, Donald Hassan, Saad S. Kim, Hie-Joon J. Mallet, Victorin
Fouda, Hassan Hasselberger, M. King, Jerry W. Mallinson, Edward
Fox, Allen Hauck, Heinz E. Klopf, Lori Manko, R. W.
Fox, Dennis R Hauptversuchsanstalt, Knowles, David Maragos, Chris
Fremy, J. Marc Bayerische Koppenaal, Gary Marderosian, Ara der
Fritz, James S. Hawthorne, Steven Korffnacher, Walter Marlett, Judith
Fung, Daniel Y. Heeschen, Walter Kovar, J. Marshall, James
Funk, Wemer Heikes, David L. Krause, Richard Marshall, Willim
Gallman, James Henry, Carolyn Lalancette, Reger A. Martin, James L.
Galoux, M. Hesselberg, Robert Lamont, William Marts, Ronald W.
Gamble, D. S. Hild, J. Lampel, Keith A. Maskarinec, Michael
Gardner, Audrey Hill, A. Lancette, Gayle Massey, Robert
Gay, Martha Hill, Walter E. Landen, William Maxey, Robert
Geerdink, R. B. Hillery, Barbara Landgraf, Wayne May, Joan C.
Gehlhausen, Jay Hilliard, J. H. Landi, Silvio Maynard, M. S.
Geokeler, Ulrich K. Ho, Chi-Tang Lane, Ralph H. Mazza, G.
George, Glenn M. Hock, Bertold Larsen, B. McClure, Fred
Giang, Benjamin Holak, Walter Laski, Ronald R. McClymont-Peace, Diane
Gilbert, John Holstege, Dirk M. Lasota, Joan McCurdy, Dennis
Gilchrist, J. Holt, Douglas L. Lavy, Terry L. McGarvey, Brian
Gilvydis, Dalia Hong, Victor Lawrance, P. McGill, A. S.
Godshall, Mary Hopper, Marvin Layman, Lawrence McGrath, S. P.
Goh, K. S. Horwitz, William LeVan, Leon McKenna, David
Goodspeed, Donald Houglum, Joel E. Lee, Healther A. McKim, James M.
Goodwin, Phillip Huber, Edward Lee, Hing-Biu McKinney, John
Gorham, Richard Hui, J. Y. Lee, S. Mark McLaughlin, Lee
Gram Jensen, Karen Hungerford, J. Lee, Sungsoo McMahon, Bernadette M.
Grant, David J. Hurlbut, Jeffrey Lentza-Rizos, C. McNair, Harold
Greenhalgh, Roy Ibe, Akihiro Leslie, Ken McNally, Mary E.
Gregory HI, J. Ihnat, Milan Levy, Joseph McNeal, Jon E.
Grey, Vivian A. Indyk, H. Lewis, Eugene L. Mennear, John
Guarino, Phil A. Isaac, Robert A. Li, Min Meyer, Keith A.
Gump, Barry H. Jain, Anant V. Liao, Wenta Miakong, Shen
Haagsma, Neil Jarvis, Bruce B. Lindsay, Robert Mijares, Rosalinda
Haddad, Paul R. Jenkins, Thomas Llames, Cynthia Miles, Carl J.
Hadfield, Stephen Jensen, Thomas Loewen, S. R. Miller, Glenn C.
Hagler, Winston Johnson, D. B. Long, Austin R. Miller-Ihli, Nancy J.
Halley, Bruce A. Johnson, Frank Lopez, L. F. Milner, Nicholas
Hamilton, J. E. Jork, Heilmut Lopez-Avila, Viorica Mindak, William
Hammock, Bruce Kadin, Harold Losada, P. Pase Minyard, James
Hangaard, J. Kandetzki, Paul Lovering, Edward Mirocha, Chester J.
Hanks, Alan Kane, Peter F. Low, Nicholas H. Mitchum, Ronald
Hanna, George M. Kantor, Edward Lowie, Dwight M. Miyahara, Makoto
Hanni, R. Katz, Stanley E. Luchtefeld, Ronald G. Moats, William

Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3 177A

Thanks to Reviewers

Moller, Gregory Newsome, W. Harvey Pagani, Silvia Plattner, Ronald

Moorman, Thomas Newton, John M. Papadopoulou-Mourkidou, Pohland, Albert
Morita, Masatoshi Ngeh-Ngwainbi, Jerry Euphemia Pollman, Roger
Mortensen, B. K. Nickerson, Steven Park, Douglas L. Poole, C. F.
Mortimer, R. D. Nierman, Paul Parks, Douglas Porter, Nathan
Motohashi, Nobom Nishizawa, Hideo Patterson, John Portoghese, Philip
Moye, H. Anson Noegrohati, Sri Patterson, Sean Prelusky, Dan B.
Mueller, Thomas Nordländer, Ingrid Pavoni, B. Rabinowitz, Michael
Munns, Robert K. Norred, William Pawliszyn, Janusz Rastogi, Suresh
Munro, Nancy B. Nott, R. Perez-Caballero, Guadalupe Ratnayaki, N.
Murphy, Patricia Nowicki, Thomas Perfetti, Gracia Rees, Naomi
Nagy, Steve Nygard, Gloria Perkins, Larry Reeves, D. W.
Nakamura, Yumiko O’Connor, B. Perry, Dana F. Reeves, Valerie
Nakazawa, Hiroyuki O’Rangers, John Pestka, James J. Reimann, Kurt
Nandrea, Gene J. Oldenburg, Lufa Petz, Michael Reimann, Lars
Nappier, John L. Öles, P. J. Pfeiffer, Sandra Reimer, Gerry J.
Navarro, Miguel Omaye, D. T. Pfender, Kathleen Reina, Robert J.
Neidert, Eh Oppenhuizen, Mark Phifer, Edwin C. Reisig, Alan
Nelsen, Terry C. Osteryoung, Janet Pignatello, Joseph J. Rice, Larry G.
Nesheim, Stanley Padmore, Joel M. Plakas, Steven Richter, Bmce

A O A C W A N T S T O P U B L IS H

D o y o u h a v e an id e a fo r a
Y O U R T h e P u b lic a tio n s D e p a r tm e n t

b o o k o n a su b ject in th e an a­ o f A O A C IN T E R N A T IO N A L

ly tic a l scien ces?

D o y o u h a v e a m a n u scrip t
B O O K S is s e e k in g p r o p o s a ls fr o m

a u th o r s fo r b o o k s to b e

b u t n o o th er p u b lish e r c o m m itte d to p u b lis h e d b y A O A C .

p u b lish in g it?
AOAC o ffers c o m p e titiv e co n tra ct term s,
A re y o u p rep a rin g a w o rk sh o p , sy m p o ­
ro y a lties, an d co m p reh en siv e m a rk etin g .
siu m , o r tr a in in g co u rse a n d w a n t to p u b ­
P r o m o tio n a l c a m p a ig n e ffo r ts a r e d e s ig n e d
lish th e p ro ceed in g s o r w ork w ith AOAC to
to p r o v id e t h e w id e s t a p p r o p r ia te e x p o s u r e
d ev elo p a m an u al?
th r o u g h th e u se o f sp a c e a d s, e x c h a n g e a d

I f y o u h a v e a n s w e r e d "Yes" t o a n y o f t h e p r o g r a m s , c o n f e r e n c e d is p la y s , a n d ta r g e te d

q u e s t io n s w e 'v e p o s e d , p le a s e c a ll o r fa x : m a ilin g s . A O A C p u b lic a t io n s r e a c h a w o r ld ­

R r y s ty n a M c lv e r , D ir e c to r o f P u b lic a t io n s , w id e a u d ie n c e o f a n a ly tic a l c h e m is ts , m ic r o ­

A O A C IN T E R N A T IO N A L , + 1 - 7 0 3 - 5 2 2 -3 0 3 2 b io lo g is t s , a n d o t h e r b io lo g is t s a n d a d m i n is ­

o r fa x + 1 -7 0 3 -5 2 2 -5 4 6 8 . tr a to r s in in d u s tr y , g o v e r n m e n t , a n d a c a d e m ia .
Thanks to Reviewers

Rizzi,George P. Silkey,James Sundaram, K.M.S. Vilim,Amost B.

Rogers,W. A. Simon,Veme A. Suzuki,Takashi Viscomi,Anthony
Rosen, Joseph D. Simpson, Kenneth Suzuki,Toshina Voyksner,Robert
Ross,StephenC. Skerritt,John Takasuki, K. Wagner, Dean
Rottinghaus,George Slahck, Stephen Takoka,T. Ware, George M.
Rowe, L.D. Slavin,Walter Tan, Shigem Warner, Charles
Rowe, Norman W. Sliva,Matthew Tang,PeterH. Weagant, StephenD.
Roybal, JoseE. Smallidge,RobertL. Tate,DonaldF. Wedzicha, B. L.
Ruch,Wallace Smedley, MichaelD. Taylor,Larry Wehling, Randy
Salisbury,Craig Smith,J.Scott Taylor,R. F. Wehner, Teresa
Sanchez, Mercedes Smith,James L. Taylor,ScottL. Wei,Lester
Samoff, Evelyn Smith,Larry Taylor,Sue G. Westerhaus,Mark 0.
Sarwar, Ghulam Smith,Randall Taylor,Wesley Westheimer,Joseph
Saschenbrecker,P.W. Smith,T.K. Tena, M. T. Weston,Leslie
Sauve, J.Paul Soderberg, David Thiex,Nancy Whitaker,Thomas
Scanlan,R.A. Sokoloff,Helen Thomas, Michael White, Jonathan
Schattenberg,HerbertJ.,Ill Soon, Y.K. Thompson, D. White, RalphT.
Schenck, Frank Spanedda,Loren Thompson, Hard Wicklow, Donald
Schmid, E.R. Sphon,James A. Thompson, Raymond Wigfield,Yuk Y.
Schrader,Doug Spittler,Terry Torma, Laszlo Wilkins,John P.
Schroeder,H. F. Stack,Michael Torreti,Luigi Wilson, ClydeR.
Schumacher,David Stafford,Steve Townshend, Alan Wilson, David M.
Schwartz,Steve Stahr,Henry M. Tracy,Mark L. Wilson, TerryD.
Scott,PeterM. Steffen,Edward Trout,Graham Windham, Wiliam
Sebranek,Joseph Steinbrecher,K. Trucksess,Mary Wolters,MechteldisG. E.
Segelman,Alvin Steindl,Eric Tscheme, Ronald Woodbury, James
Sen,Nisu P. Stephens,Marvin Tubino,Matthieu Wright,Robert
Settepani,Joseph Stephens,RobertD. Tumipseed, Sherri Wrolstadt,R. E.
Seymour, Sandra Stevens,Timothy S. Tyler,R. Wyhowski de Bukanski,
Shaikh, Badar Stevenson,Derek Uden, Peter Brigitte
Shalaby,Lamaat M. Stewart,James Ueta,Tadahiko Yeager,Russell
Shea,Patrick Stoloff,Leonard Uhler,AllenD. Yess,Norma J.
Shearer,G. Stout,StevenJ. Ullman,E.F. Ymyu, Shi
Shearer,M. J. Strom,J.Grady Venant,A. Zini,Guido
Shibamoto, Takayuki Strong,Ann B. Venkateswarlu, van Egmond, Hans
Shillington,D. Suhre,Francis Pothapragada van Ginkel,Leendert
Shipe,W. F. Sullivan,Darry Vesonder, Ronald van de Voort,Frederick
Sieh,David H. Sullivan,John Vidnes,Ame von Elbe,Joachim H.

Journal Of AOAC International V ol . 76, No. 6,1993 179A

SPECIAL REPORT

R e p o r t o f th e A O A C IN T E R N A T IO N A L T a s k F o rc e o n M e th o d s

f o r N u t r i e n t L a b e li n g A n a ly s e s

T heNutritionLabelingandEduca­ lookedprimarilytoAOAC INTERNA­ terialsaspartofthemethods validation

tion Act of 1990 (NLEA) was TIONAL forappropriatemethods. process.
signed intolaw on November 8,
1990. Even though many foods have AOAC Response Task Force Actions
been labeled voluntarily fornutritional
content since 1973, this far-reaching The analyticalcommunity was quick Identification o f Available
pieceoflegislationmade mandatory,for to recognize the immense task that lay Adequate Validated M ethods
the firsttime, the labeling ofnearly all before it.Analyticaldatawould have to
foodswithinformationabouttheirnutri­ be generatedforhundreds ofthousands The task force drew upon the re­
tionalcontent. NLEA requiredthe U.S. of foods that may not have been ana­ sources offered by the entire AOAC
FoodandDrugAdministration(FDA) to lyzedpreviously.As FDA hadrightfully membership and worked closely with
develop revised regulations that would recognized, validated methodology for other organizations. One of the first
mandate the labeling of several pre­ some of the new mandatory nutrients tasksundertakenby thetaskforcewas a
viouslyunlabeledfoodcomponents (to­ was notavailable.Without accurateand surveyofanalystsworking inanalytical
laboratories that specialized in food
taldietaryfiber,complexcarbohydrates, reliableanalyticalmethods, compliance
analysis. This was done to determine
sugars, saturated fat,and calories from withtheNLEA would be impossible,as
whatmethodswerecurrentlybeingused
fat). In November of 1991, FDA pub­ would enforcement. To address theim­
and theanalyst’sevaluation ofthe ade­
lished their response to NLEA in the mediate needsrequiredby theproposed quacyofthosemethods.The surveyalso
formofproposedregulationsintheF e d ­ regulations, AOAC [with participants askedaboutreferencestandardsand the
e r a l R e g is te r ( FR, November 27, 1991, drawn from government agencies, aca­ use ofcontrolsamples. A review ofthe
pp. 60365-60891). Although not re­ demia, privatecompanies, and tradeas­ surveyresponsesrevealedthatlaborato­
quired by the NLEA, the U.S. Depart­ sociations(s e e C o m m itte e M e m b e r s a n d rieswereusingagreatvarietyofAOAC
ment of Agriculture (USDA) also pro­ P a rtic ip a n ts )]formedtheTaskForceon methods, sometimes with modifica­
posed regulations requiring mandatory MethodsforNutrientLabelingAnalyses tions, and that a great many methods
labeling of nutrients. The USDA pro­ withthefollowingmission/responsibili- werebeingappliedsuccessfullytoprod­
posal (F e d e r a l R e g iste r, November 27, ties: uctmatrixesforwhich themethods had
1991, pp. 60301-60364) was harmo­ 1. Identify and publicize the avail­ not been explicitly validated. Assign­
nizedwiththeFDA proposal.The FDA abilityofvalidatedmethods fornutrient ments were undertaken by individual
document amounted toover 500 pages analysis currently available through members of the task force to review
and discussedeverything from method­ AOAC. AOAC methods for specific nutrients.
ology and serving sizes to exemptions 2. Identify needed revisions inex­ Using all resources available to them,
and preemptions. The proposalnotonly istingAOAC OfficialMethodsfornutri­ the individual members determined
includedthe listofmandatory nutrients tionlabeling. which methods were adequate for the
tobe labeledbutalsoexpanded thepre­ 3. Identify additional methods for purposes ofnutritionallabeling.The re­
viouslistofthosethatcouldbevoluntar­ nutritionalanalysisinneedofvalidation viewprocesswas carriedoutby nutrient
ilylabeled.Intheproposal,FDA explic­ and officialactionby AOAC. and productmatrix combinations. Even
itly acknowledged concern over the thoughthesurveydidindicatethatmany
4. Develop an approach or ap­
non-AOAC methods wereroutinelybe­
availability of analytical methodology proaches thatcan be used by AOAC to ing used and that laboratories found
forcomplex carbohydrates, sugars, and initiatenecessary studiesand recruitre­ some ofthesealternativemethods tobe
protein quality. FDA also continued to searchersand collaborators. acceptable,themissionofthetaskforce
acknowledge that, for compliance pur­ 5. Identify and publicize standard was toidentifyAOAC OfficialMethods
poses,theknown variabilityofappropri­ reference materials available for use that have been accepted and validated.
ateanalyticalmethods would be an im­ withcurrentvalidatedmethodsfornutri­ Aftertheindividualtaskforcemembers
portantaspectinevaluatingcompliance. tional analysis. Identifyneeds foraddi­ reviewed each analyteand matrix com­
With only the single exception of the tional reference materials for current bination, their recommendations were
new analytical methodology proposed validated methods. Develop an ap­ collectivelyreviewed by the entiretask
for protein quality, the FDA proposals proach forincorporating reference ma­ forcewiththeassistanceofAOAC tech-
180A S pecial R eport : Journal of A O AC International V ol . 76, N o . 6 ,1 9 9 3
meal staff.The resultofthereviewproc­ was alsoformedtospecificallyexamine S u b c o m m itte e on c a rb o h y d r a te
esswas alistingofadequateAOAC Of­ the suitability of methods for carbohy­ m e th o d s. —
On the basis of the current,
ficialMethods thatwas publishedinthe drates. most widely accepted definitionofdie­
July 1992 issue of T h e R e fe re e , the S u b c o m m itte e on to ta l f a t m e th ­ taryfiber,the Subcommittee on Carbo­
AOAC newsletter(Table 1).A comple­ o d s .—
The Subcommittee on Total Fat hydrate Methods concluded that the
mentary listwas generated, specifying Methods conduced that the lack of a AOAC OfficialMethods currentlyvali­
methods that needed development and single,clear,concisedefinitionforfatas datedareadequateforthedetermination
validation,updating,revision,orfurther ananalytewas asignificantcontributing oftotal,insoluble,andsolubledietaryfi­
work toextend theirapplicability. This factortotheissuesregarding fatanaly­ ber fornutrition labeling. The subcom­
listwas published inthe October 1992 sis.AOAC does notdefineanalytesbut mittee also determined that reference
issueofT he R e fe re e (Table2). rathervalidatesmethods foranalytesas materials are available for laboratories
defined. Therefore, the subcommittee touseforanalyticalqualitycontrol.This
Identification o f Needed information and the conclusions of the
recommended thattheregulatoryagen­
M ethodology Revisions subcommitteewerecontainedinareport
cies involved with nutrition labeling
(i.e.,USDA and FDA) adopt a single, acceptedby thetaskforceandpublished
During the review process, AOAC intheNovember 1992issueofT h e R e fe ­
members raisedseveralsubstantia]con­ clear,concise definitionforfatand that
re e (Tables 5 and 6;s e e also Lee, S.C.,
cerns about the applicability of some AOAC INTERNATIONAL then vali­
DeVries,J.W.,& Sullivan,D.M. (1993)
AOAC methods forcertainanalytes.Of date methods commensurate with those
M e th o d s o f A n a ly s is f o r N u tritio n L a b e l­
particular concern was the availability definitions. The subcommittee further
ing, D.M. Sullivan & D.E. Carpenter
ofmultipleOfficialMethods foragiven went on toevaluate each ofthe current
(Eds), AOAC INTERNATIONAL, Ar­
nutrient or food component. The task AOAC Official Methods for fat or fat lington,VA, Chapter4).When theregu­
force was aware thatin some instances fractionsand outlinedand tabulatedthe latory agencies published theirfinal la­
the analytical resultwas a reflection of key characteristics of each. The task beling regulations in early 1993, the
thebiasesofthemethod and notneces­ force adopted the subcommittee’s re­ subcommittee updated itsreportby re­
sarily a true measure of the nutrient in port,anditwaspublishedintheSeptem­ viewing methods for the analysis of
the food. The method reviewers deter­ ber 1992 issueofThe R eferee (Table3; mono- and disaccharides and determin­
mined thatthevariousOfficialMethods see also Carpenter, D.E., Ngeh- ingwheremethods areneeded.The sub­
designatedfortotalfatdo notnecessar­ Ngwainbi, J.,& Lee, S.C. (1993)M e th ­ committee found thatestablished liquid
ilymeasure thesame food components o d s o f A n a ly sis f o r N u tritio n L a b elin g , chromatographic methods were suitable
andthusmay notproduceequivalentre­ D.M. Sullivan & D.E. Carpenter(Eds), for the determination of glucose, fruc­
sults.Certain methods fortotalfatmay AOAC INTERNATIONAL, Arlington, tose,maltose, sucrose, and lactose.The
include food components that do not VA, Chapter5). report also suggested that the classical
contribute to dietary fat intake. Other S u b c o m m itte e on m o istu re m e th ­ chemical and enzymatic procedures
methodsmay underestimatethetotalfat o d s. — The Subcommittee on Moisture would have utilityincertainlimitedap­
in the food being analyzed; therefore, Methodsreachedaconclusionsimilarto plications. Even though the finallabel­
theyalsounderestimatethecontribution that of the Subcommittee on Total Fat ingregulationsdroppedtherequirement
thatthefoodbeinganalyzedmakes tofat forthemandatory declarationof“com­
Methods regardingthecauseofdiscrep­
intake. Similarly, some moisture meth­ plex carbohydrates,” the subcommittee
ancies inresultsbetween methods used
ods do notmeasure moisturepersebut recommended that,asinthecaseoftotal
formoisture analysis. As in the case of
eitherresidualsolidsortotalvolatilema­ fatand moisture,aclear,concisedefini­
fat, moisture has not been clearly de­
terials,whicharenotnecessarilyequiva­ tionofcomplexcarbohydratesshouldbe
lent. Therefore, the task force decided finedbutisafunction ofthemethodol­
ogy usedforthedetermination.Because developed. AOAC should then support
thatthemethodology forthesenutrients method development and validation to
needed further study. Two subcommit­ moisturecontentoffoods isessentialto
allcaloriecalculations,the subcommit­ meet thisdefinition.
tees were formed to carefully examine
theseissues and tomake recommenda­ tee recommended this issue be ad­
Addressing M ethodology Needs
tions to the task force. The task force dressed on a regulatory basis inthe fu­
members also feltthat, because of the ture.To assistthemembershipofAOAC The response ofAOAC members to
importantrolecarbohydrateshaveinthe and the users of O ffic ia l M e th o d s o f theeffortsofthetaskforcehas,thusfar,
new nutritionlabel(declarations forto­ A n a ly sis, the subcommittee compiled a been highlyencouraging.Notonlyhave
tal carbohydrate, complex carbohy­ listing of the AOAC Official Methods AOAC members responded with meth­
drates, total dietary fiber, and simple for moisture along with the methods’ odsforvalidationandofferedtoserveas
sugars)andbecauseofconcernsoverthe characteristics. The compiled list and collaboratorsand associatereferees,but
chemicaldefinitionofthevariouscarbo­ thesubcommitteereport,which was ac­ various trade associations and profes­
hydrate fractions, it was important to ceptedby thetaskforce,was published sional groups have recruited members
evaluate the adequacy of carbohydrate intheJanuary 1993issueofThe R efe re e fromtheirranksfortheeffortsnecessary
methods. Therefore, a subcommittee (Table4). tovalidateadditionalmethods thatmeet

S pecial Report : Journal of A O AC International V ol . 76, N o. 6 ,1 9 9 3 181A

T a b le 1 . A c c e p t a b le O fficial M e th o d s fo r n u tritio n la b e lin g ( a n a ly t e s n o t in c lu d e d in N L E A fin a l r u lin g ; s e e T a b l e 7 fo r
fin a l lis t o f m e t h o d s ) 9

AOAC No. Method Matrix

CARBOHYDRATES-TOTAL
986.250 Carbohydrate of Milk-Based Infant Formula K

CHLORIDE
971.27 NaCI in Canned Vegetables, Potentiometrie Method All
986.26 Chloride in Milk-Based Infant Formula, Potentiometrie Method All
952.24 Microchem Determination of Br, Cl, and I, Carius Combustion Method All
974.36 Microchem Determination of Br, Cl, and I, Oxygen Flask Combustion Method All
915.01 Chlorides in Plants, Volumetric Method A, C, F, J, P, T
935.05 Chlorides in Plants, Volumetric Method A, C, F, J, P, T
935.47 Salt (Chlorine as NaCI) in Meat, Volumetric Method B, L
983.14 Chloride in Cheese, Potentiometrie Method G
935.43 Total Chloride in Cheese G
976.18 Salt (Chlorine as NaCI) in Seafood, Potentiometrie Method I, Q
937.09 Salt (Chlorine as NaCI) in Seafood, Volumetric Method LQ
933.06 Chloride in Eggs, Potentiometrie Method L
941.13 Total Chlorides in Prepared Mustard, Titrimetric Method S

CHROMIUM
975.03c Metals in Plants, AAS Method All
985.01e Metals and Other Elements in Plants, ICP Method All
985.35e Minerals in Ready-To-Feed Milk-Based Infant Formulation, AAS Method
984.27e Ca, Cu, Fe, Mg, Mn, P, K, Na, and Zn in Infant Formulation, ICP Method All

FAT-SATURATED
969.33 Fatty Acids in Oils and Fats, Preparation of Methyl Esters with 963.22 Methyl Esters of Fatty
Acids in Oils and Fats, GC Method (Standardize with each fatty acid or with one and All
normalize.)

FAT-TOTAL
983.23 Fat in Foods, Chloroform Methanol Extraction Method (Extract high sugar, glycerol-containing, All
and chocolate products)
960.39 Crude Fat in Meat B, L
976.21 Crude Fat in Meat, Specific Gravity method B, L
985.15 Crude Fat in Meat and Poultry Products, Microwave-Solvent Extraction Method B, L
922.06 Fat in Flour, Acid Hydrolysis Method E, F, L, O, P, R
920.39C Crude Fat in Animal Feed F, P, R
945.18A Fat in Cereal Adjuncts F, R
925.12 Fat in Macaroni Products, Add Hydrolysis Method F
945.38F Fat in Grains F
933.05 Fat in Cheese, Roese-Gottlieb Method G
989.05 Fat in Milk, Mojonnier Ether Extraction Method H
938.06 Fat in Butter
905.02 Fat in Milk, Roese-Gottlieb Method H
920.111 Fat in Cream, Roese-Gottlieb Method H
952.06 Fat in Ice Cream and Frozen Desserts, Roese-Gottlieb Method H
945.48G Fat in Evaporated Milk, Roese-gottlieb Method H
932.06 Fat in Dried Milk, Roese-Gottlieb Method H
948.15 Crude Fat in Seafood, Add Hydrolysis Method LQ
964.12 Crude Fat in Seafood, Mod Babcock Method LQ
986.25 Fat in Milk-Based Infant Formula, Roese-Gottlieb Method K
925.32 Fat in Eggs, Add Hydrolysis Method L
948.22 Crude Fat in Nuts and Nut Products, Gravimetric Method N
950.54 Total Fat in Food Dressings O

182A S pecial R eport : Journal of A O AC International V ol . 76, N o . 6 ,1 9 9 3

T a b le 1 . ( c o n tin u e d )

AOAC No. Method Matrix

FAT-TOTAL(continued)
935.39D Fat in Baked Products, Add Hydrolysis Method R
945.44 Fat in Fig Bars and Raisin Cookies, Ether Extraction Method R

FAT-UNSATURATED
969.33 Fatty Acids in Oils and Fats, Preparation of Methyl Esters with 963.22 Methyl Esters of Fatty
Adds in Oils and Fats, GC Method (Standardize with each fatty acid or with one and All
normalize.)

FLUORIDE
944.08 Fluorine in Food, Distillation Method All
961.16 Microchemical Determination of Fluorine, Titrimetric Method All

MANGANESE
975.03 Metals in Plants, AAS Method All
980.03 Metals in Plants, Spectrographic Method All
965.09 Minor Nutrients in Fertilizers, AAS Method All
968.08 Minerals in Animal Feed, AAS Method All
953.01 Metals in Plants, Emission Spectrographic Method A, C, F, J, P, T
985.01 Metals and Other Elements in Plants, ICP Method A, C, F, J, P, T
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na, and Zn in Infant Formulation, ICP Method K
985.35 Minerals in Ready-To-Feed Milk-Based Infant Formula, AAS Method K

MOISTURE**
925.09 Total Solids and Moisture in Flour, Vacuum Oven Method All (Low sugar products)
A, C, P, T (high sugar
964.22 Solids (Total) in Canned Vegetables, Gravimetric Method products)
D, E, G, H, K, S (high
925.45A Moisture in Sugars, Vacuum Drying Method sugar products

MOLYBDENUM
980.03 Metals in Plants, Spectrographic Method All
975.03e Metals in Plants, AAS Method All
985.01e Metals and Other Elements in Plants, ICP Method All
985.35e Minerals in Ready-To-Feed Milk-Based Infant Formula, AAS Method All
984.27e Ca, Cu, Fe, Mg, Mn, P, K, Na, and Zn in Infant Formulation, ICP Method All
960.05 Molybdenum in Plants, Colorimetric Method A, C, F, J, P, T

SELENIUM
974.15 Selenium in Food, Fluorometric Method All
986.15 As, Cd, Pd, Se, and Zn in Food, Multielement Method All
969.06 Se in Plants, Fluorometric Method A, C, F, J, P, T

a This list was prepared thanks to the efforts of M. Bueno, D. Carpenter, H. Chin, M. Deutsch, J. DeVries, N. Fraley, W. Flummer, W. Ikins, W.
Landen, S. Lee, J, Morawski, P. Oles, L. Prosky, D. Soderterg, A. Soliman (deceased), D. Sullivan, J. Tanner, and W. Wolf. Matrixes were
considered representative of all food types and are as follows: A, fruit baby foods; B, meat baby foods; C, vegetable baby foods; D,
beverages and juices; E, candy; F, cereals and products; G, cheese; H, dairy products; I, fish; J, fruits; K, infant formula or medical diet; L,
meat (beef, pork, or fowl); M, mixed dinners (TV dinners); N, nuts; O, oils/fats (dressings); P, potatoes and products; Q, shellfish; R, sweet
mixes (cake, pie, etc.); S, spices; and T, vegetables.
b Calculation by difference is acceptable for all other matrixes.
c Commonly used for determining Cr but not collaboratively studied for Cr.
d Numerous other satisfactory methods in Official Methods of Analysis.
e Commonly used for determining Mo but not collaboratively studied for Mo.

S pecial Report : Journal of A O AC International V ol . 76, N o. 6 ,1 9 9 3 183A

T a b le 2 . M a tr ix e s n e e d in g m e t h o d s (n o t in c lu d e d in N LEA fin a l ru lin g; s e e T a b le 8 fo r fin a l lis t o f m e t h o d s )

Recommendations Matrixes3

CARBOHYDRATES-SUGARS All

• If FDA/USDA defines as 1 & 2 saccharide units, HPLC methods 977.20, 980.13, 982.14, 984.17 are suitable for the stated
matrixes
• Cs of HPLC method recommended

CARBOHYDRATES-TOTAL A-J, L-T

• Need method/cs

CARBOHYDRATES-SUGAR OLIGOMERS All

• Need method/cs for 5-9,1 -9, 3-9

CHLORIDE

• Cs combining 971.27, 986.26, 976.18, 983.14 for all matrixes recommended

• Cs of automated method recommended

CHROMIUM

• Need cs to add Cr as analyte in 975.03, 985.01,984.27

FAT-UNSATURATED

• Recommend extension of matrixes for 969.33

• See report of Fat Subcommittee

FLUORIDE

• Recommend standardization of titles (Fluorine vs Fluoride) and addition of applicability statements to current methods

MANGANESE

• Recommend extension of applicability statements to foods for 965.09, 985.35, 975.03, 968.08, 980.03
• Recommend extension of applicability statement to A, B, D-F, H-M, O, P, R, T for 984.27

MOISTURE B, F, I, J, L-O,
Q, R for high
sugar products

• Recommend modification of 952.08 to eliminate asbestos use

MOLYBDENUM

• Need cs to add Mo as analyte in 975.03, 985.01,984.27

SELENIUM

• Cs recommended to combine digestion from 974.15 with 986.15

VITAMIN K

• Recommend cs of method by HPLC

• Recommend cs of method for infant formula, JAOAC 68, 684(1985); for medical foods, JAOAC 68,183(1985); and for infant formula but
révisable to include other matrixes, JAOAC 66,1063(1983)

This list was prepared thanks to the efforts of M. Bueno, D. Carpenter, H. Chin, M. Deutsch, J. DeVries, N. Fraley, W. Hummer, W. Ikins, W.
Landen, S. Lee, J. Morawski, P. Oles, L. Prosky, D. Soderberg, A. Soliman (deceased), D. Sullivan, J. Tanner, and W. Wolf. Matrixes were
considered representative of all food types and are as follows: A, fruit baby foods; B, meat baby foods; C, vegetable baby foods; D,
beverages and juices; E, candy; F, cereals and products; G, cheese; H, dairy products; I, fish; J, fruits; K, infant formula or medical diet; L,
meat (beef, pork, or fowl); M, mixed dinners (TV dinners); N, nuts; O, oils/fats (dressings); P, potatoes and products; Q, shellfish; R, sweet
mixes (cake, pie, etc.); S, spices; and T, vegetables. Cs = collaborative study.

184A S pecial R eport : Journal of AOAC International V ol. 76, N o . 6 ,1 9 9 3

T a b le 3 . F at a n a l y s i s fo r n u tritio n a l la b e lin g — A O A C O fficial M e th o d s 3

Method No. Title Brief Description App ¡cable Matrixes Comments

960.39 Crude Fat in Meat Pet or diethyl ether extn Baby food-meats, meats Mono-, di-, and triglycer­
ides and most of the
sterols, glycolipids,
phospholipids, and waxes
976.21 Crude Fat in Meat Rapid Sp gr mth; Baby food-meats, meats Mono-, di-, and triglycer­
tetrachloroethylene extn ides and most of the
sterols, glycolipids,
phospholipids, and waxes
985.15 Crude Fat in Meat & Rapid microwave- Baby food-meats, meats Mono-, di-, and triglycer­
Poultry Prod methylene chloride solv. ides, glycolipids
extn phospolipids, and waxes;
sterol yield may be
reduced
920.39B Crude Fat in Animal Feed Diethyl ether extn Cereals & prod; not Mono-, di-, and triglycer­
adequate if heat ides and traces of other
processed, extruded, or lipid components
has sugar added
920.39C Crude Fat in Animal Feed Diethyl ether extn; H2O Cereals & prod, sweet Mono-, di-, and triglycer­
prewasi if high in sugar mixes (cakes & pies); not ides; may not quant,
adequate if heat extract tot lipids;
processed recommend further
review or study of mth
945.18A Fat in Cereal Adjuncts Pet ether extn Cereals & prod, sweet Mono-, di-, and triglycer­
mixes (cakes and pies); ides; may not quant,
not adequate if heat extract tot lipids;
processed recommend further
review or study of mth
945.38F Fat in Grains Refers to 920.39C Cereals & prod; adequate Mono-, di-, and triglycer­
if not heat processed ides and traces of other
lipid components
933.05 Fat in Cheese Acid hydr., pet and diethyl Cheese Mono-, di-, and triglycer­
ether extn re-extraction ides and traces of other
lipid components; sterol
yield greatly reduced

905.02 Fat in Milk Aik treatmt, pet and diethyl Dairy Mono-, di-, and triglycer­
ether extn re-extraction ides and traces of other
lipid components
989.05 Fat in Milk Aik treatmt, pet and diethyl Dairy Mono-, di-, and triglycer­
ether extn ides and traces of other
lipid components
938.06 Fat in Butter Pet or diethyl ether extn Butter Mono-, di-, and triglycer­
ides and traces of other
lipid components
920.111A Fat in Cream Aik treatmt, pet and diethyl Dairy Mono-, di-, and triglycer­
ether extn re-extraction ides and traces of other
lipid components
920.111B Fat in Cream Babcock, acid hydr., vol. Dairy Mono-, di-, and triglycer­
anal ides and phospholipids;
sterol yield reduced
952.06 Fat in Ice Cream & Frozen Aik treatmt, pet and diethyl Dairy Mono-, di-, and triglycer­
Desserts ether extn re-extraction ides and traces of other
lipid components
945.48G Fat in Evaporated Milk Aik treatmt, pet and diethyl Dairy Mono-, di-, and triglycer­
ether extn re-extraction ides and traces of other
(refers to 905.02) lipid components
932.06 Fat in Dried Milk Aik treatmt, pet and diethyl Dairy Mono-, di-, and triglycer­
ether extn ides and traces of other
lipid components

S pecial Report : Journal of AO AC International V ol . 76, N o. 6 ,1 9 9 3 185A

Table 3. (continued)
Method No Title Brief Description Applicable Matrixes Comments

948.15 Crude Fat in Seafood Acid hydr., pet and diethyl Fish, shellfish Mono-, di-, and triglycer­
ether extn ides, fatty acid portion of
phospholipids and glyco-
lipids; in some prod with
high sugar content, may
overest. fat; sterol yield
may be reduced 964.12
986.25 Fat in Milk-Based Infant Aik treatmt, pet and diethyl Infant formula/medical Mono-, di-, and triglycer­
Formula ether extn; re-extn (refers ides and traces of other
to 945.48G) lipid components
925.32 Fat in Eggs Add hydr., pet and diethyl Eggs/egg prod; May not Mono-, di-, and triglycer­
ether extn be adequate for some ides, fatty acid portion of
egg products phospholipids and
containing sugar glycolipids; sterol yield
may be reduced
948.22 Crude Fat in Nuts and Nut Diethyl ether extn Nuts; not adequate for Mono-, di-, and triglycer­
Prod nuts containing sugar ides and traces of other
lipid components
950.54 Tot Fat in Food Dressings Add hydr., pet and diethyl Oils/fats (dressings) Mono-, di-, and
ether extn triglycerides, fatty acid
portion of phospolipids
and glycolipids
945.44 Fat in Fig Bars & Raisin Add hydr., pet and diethyl Mono-, di-, and triglycer-
Filled Cookies ether extn re-extraction Sweet mixes (cakes & pies) ides and fatty acid portion
and baked cereal products of phospholipids and
glycolipids; sterol yield
may be reduced; re-extn
may not remove all sug­
ars; recommend further
review or study of mth
963.15 Fat in Cacao Prod Pet ether extn Chocolate prod Mono-, di-, and
triglycerides and traces of
other lipid components
925.07 Fat in Cacao Prod Pet and diethyl ether extn Candy Mono-, di-, and triglycer­
ides and traces of other
lipid components
920.177 Ether Extract of Pet and diethyl ether extn; Candy Mono-, di-, and
Confectionery re-extn triglycerides and traces of
other lipid components
Mono-, di-, and
920.172 Ether Extract of Prepared triglycerides and traces of
Mustard Diethyl ether extn Mustard other lipid components
963.22* Methyl Esters of Fatty GC mth following prep of All Fatty acid profile; recom­
Adds in Oils and Fats methyl esters (according mend further review or
to 969.33) study of mth to make
method adequate for
quantitative analysis of
fatty adds
979.19* c/s,c/s-Methylene Enzymatic-Spectro-photo- All
Interrupted metric mth
Polyunsaturated Fatty
Adds in Oils

a This list has been prepared thanks to the efforts of S. Bailey, D. Carpenter, H. Chin, J. DeVries, N. Fraley, W. Hummer, A. Kistler, S. Lee,
J. Ngeh-Ngwainbi, P. Oles, and D. Sullivan.
6 Fatty acid methodology and c/'s.c/s-methylene interrupted procedure.

186A S pecial R eport : Journal of A O AC International V ol. 76, N o . 6 .1 9 9 3

T a b le 4 . M o is tu r e a n a l y s i s fo r n u tritio n a l la b e lin g — A O A C O fficial M e th o d s

AOAC Nc. Title Brief Description Applicable Matrixes Comments

925.04 Moisture in Animal Feed Distillation with toluene Spices

986.21 Moisture in Spices Distillation with toluene or Spices Method measures water as
hexane moisture In samples.
972.20 Moisture in Prunes and Moisture meter method Dried fruit Measures conductivity and
Raisins equates this to moisture.
977.10 Moisture in Cacao Products Karl Fischer, Milk chocolate & Method measures water as
electrochemical method confectionery coatings moisture in samples.
984.20 Moisture in Oils and Fats Karl Fischer, Oils & fats except for alk. or Method measures water as
electrochemical method oxidized samples moisture in samples.
920.116 Moisture in Butter Pressure = atmospheric Butter Method measures all 100°C
Temp = bp (H2O) on sand volatile components.
to constant weight Residual Solids Method.
925.10 Solids (Total) and Moisture Pressure = atmospheric Flour Method measures all 100°C
in Flour Temp = 130°C-100°C Air volatile components.
oven for 3 h Residual Solids Method.
925.23 Solids (Total) in Milk Pressure = atmospheric Milk Method measures all 100°C
Temp = 98-100°C Air volatile components.
oven for 3 h Residual Solids Method.
930.15 Moisture in Animal Feed Pressure = atmospheric Animal feeds Method measures all 130°C
Temp = 135°C for 2 h volatile components.
Residual Solids Method.
931.04 Moisture in Cacao Products Pressure = atmospheric Cacao products Method measures all 100°C
Temp = 100°C air oven to volatile components.
constant weight Residual Solids Method.
935.29 Moisture in Malt Pressure = atmospheric Malt Method measures all
Temp = 103-106°C Air 103-106°C volatile
oven for 3 h components. Residual
Solids Method.
941.08 Total Solids in Ice Cream Pressure = atmospheric Ice cream and other frozen Method measures all 100°C
and Frozen Desserts Temp = 100°C Air oven for desserts volatile components.
3.5 h Residual Solids Method.
948.12 Moisture in Cheese (Rapid Pressure = atmospheric Cheese Method measures all 130°C
screening method) Temp = 130°C Air oven for volatile components.
1.25 h Residual Solids Method.
952.08 Solids (Total) in Seafood Pressure = atmospheric All marine products except Method measures all 100°C
Temp = 100°C air oven raw oysters volatile components.
4 h or Temp = lOO'C draft Residual Solids Method.
oven for 1 h
984.25 Moisture (Loss of Mass on Pressure = atmospheric Quick-frozen french-fried Method measures all 130°C
Drying) in Frozen Temp = 103°C Air oven for potatoes volatile components.
French-Fried Potatoes 16 h Residual Solids Method.
985.14 Moisture in Meat and Pressure = atmospheric Meat and poultry Method measures all 130°C
Poultry Products Temp = 130°C + volatile components.
Microwave oven Residual Solids Method.
950.46B(a) Moisture in Meat Pressure = atmospheric Meat and meat products Method measures all 100°C
Temp = 100-102°C air volatile components.
oven for 16-18 h Residual Solids Method.
950.46B(b) Moisture in Meat Pressure = atmospheric Meat and meat products Method measures all 125°C
Temp = 125°C ai' oven for volatile components.
2-4 h Residual Solids Method.
920.151 Solids (Total) in Fruits and Pressure = <100 mm Fig Fruits and fruit products Method measures moisture
Fruit Products Temp = 70°C 2 h wts to in sample.
<3 mg change
925.40 Moisture In Nuts and Nut Pressure <100 mm Fig Nuts and nut products Method measures moisture
Products Temp = 95-100°C excluding high sugars or in samples.
Constant Weight (ca 5 h) glycerols
926.08 Moisture in Cheese Pressure <100 mm Fig Cheese Method measures all 100°C
Temp = 100°C Constant volatile components.
Weight (ca 4 h) Residual Solids Method.

S pecial Report : Journal of A O A C International V ol . 76, N o. 6 ,1 9 9 3 187A

T a b le 4 . ( c o n tin u e d )

AOAC No. Title Brief Description Applicable Matrixes Comments

926.12 Moisture and Volatile Matter Pressure = <100 mm Fig Measures all volatile
in Oils and Fats Temp = bp (FI2O) + 25°C components in sample.
to constant weight Fats & Residual Solids Method.
oils
927.05 Moisture in Dried Milk Pressure = <100 mm Hg Dried Milk Method measures all 100°C
Temp = 100°C Constant volatile components.
Weight (ca 5 h) Residual Solids Method.
934.01 Moisture in Animal Feed Pressure = <100 mm Fig Animal feedsMethod
Temp = 95-100°C(P=<50 measures moisture, better
T=<70 w/i sugar) at lower T and P.
934.06 Moisture in Dried Fruits Pressure <100 mm Fig Dried fruit Method measures moisture
Temp = 70°C Constant in samples.
Weight (ca 6 h)
920.115(d) Sweetened Condensed Pressure <100 mm Fig Sweetened condensed milk Method measures all 100°C
Milk; Total Solids Temp = 100°C Constant volatile components.
Weight (ca 6 h) Residual Solids Method.
925.30(b) Solids (Total) in Eggs Pressure <25 mm Fig Temp Liquid or dried eggs Method measures all
= 98-1OOTS Constant 98-100°C volatile
Weight (ca 5 h) components. Residual
Solids Method.
925.09 Solids (Total) and Moisture Pressure <25 mm Fig Temp Flour Method measures moisture
in Flour = 98-100°C Constant in samples.
Weight (ca 5 h)
945.43 Moisture in Fig Bars and Pressure <50 mm Fig Temp Fig bars and raisin-filled Method measures moisture
Raisin-Filled Crackers. = 98-70°C 16 h with sand crackers in samples.
964.22 Solids (Total) in Canned Pressure <50 mm Hg Temp Canned vegetables Method measure moisture
Vegetables = 69-71 °C Multi-Step in samples.
steam-air oven-vacuum
oven drying, final = 2 h in
vacuum
925.45(a) Moisture in Sugars-Vacuum Pressure <50 mm Hg Temp Cane and beet, raw and Method measures moisture
Drying = 70°C (preferably 60°C) refined sugars in samples.
2 h + until weight loss <2
mg
925.45(b) Moisture in Sugar-Drying at Pressure = atmospheric Cane and beet, raw and Method measures all 100°C
atmospheric pressure Temp = 100°C 3 h + until refined sugar volatile components.
weight loss <2 mg Residual Solids Method.
925.45(d) Moisture in Sugars-Drying Pressure <50 mm Hg Temp Massecuites, molasses, Method measure moisture
on Quartz Sand = 70°C (preferably 60°C) and other liquid and in samples.
until wt loss <2 mg (ca semiliquid products
18 h)
985.26 Total Solids in Processed Pressure = atmospheric Tomato products Method measures volatile in
Tomato Products by Temp = 125°C + To samples.
Microwave Moisture constant weight
analyzer
972.16 Solids in Milk by NIR Reflectance spectroscopic Milk products Method measure moisture
method in samples.

188 A S pecial R eport : Journal of AO AC International V ol . 76, N o. 6 ,1 9 9 3

T a b le 5 . LC m e t h o d s fo r c a r b o h y d r a te a n a l y s i s

Analytical
Method Sugars Matrix Traction Clean-up column/detector Mobile phase
Analysis on amino silica columns®

982.14 glucose Presweetened 50% ethanol, Centrifuge; C18 Amino bonded CH3CN-H2O
fructose cereals 85-90°C, cartridge 0.45 silica 300 x 4 (80 + 20)
sucrose (defatted) water bath, 25 pm filter mm/RI
maltose min
980.13 fructose Milk chocolate H20, 85-90°C Centrifuge; 0.45 Amino bonded CH3CN-H2O
glucose (defatted) water bath, 25 pm filter silica 300 x 4 (80 + 20)
lactose min mm/RI
maltose
sucrose
984.17 fructose Licorice extracts H2O, 0-90°, 25 0.45 pm filter Amino bonded CH3CN-H2O
glucose min C18 cartridge silica 300 x 4 (83 + 17)
sucrose mm/RI
maltose
977.20 fructose Honey Aqueous 0.45 pm filter Amino bonded
glucose CH3CN silica 300 x 4 CH3CN-H2O
sucrose mm/RI (83 + 17)

Analysis cn ion exchange resin columns

983.22 Minor saccharides Corn sugar H2O 5% (w/w) mixed 610 x 7 mm Ion H2O
(dextrose total exchange exchange
DP2 total DP3) resin in Ca2+ 78°C
sample soln.
979.23 Major saccharides Corn syrup H2O 5% (w/w) mixed 610 x 7 mm Ion H2O
(glucose exchange exchange 50
fructose resin in W-x 4, Ca2+
maltose sample soln. 78°C
psicose)

Methods 982.14 and 980.13 should have broad applicability, because samples are defatted before extraction. May need to eliminate the
NaCI interference as indicated in 982.14. Recommend collaborative study to extend to other matrixes (i.e., dairy products).

Table 6. Other m ethods fo r sugar analysis

AOAC No. Method Brief description Applicable matrixes

Gas chromatography

971.18 Carbohydrates In Fruit Juices All

Sample clarified with Pb acetate, deriv to
TMS deriv and sep on silicone phase; det
mono and disaccharides3
973.28 Sorbitol in Foods Fruits
Acetate deriv of sorbitol are detd

Charcoal column chromatography

954.11 Separation of Sugars in Honey Charcoal Col Chrom followed by detn of All
fractions3
979.21 Separation of Sugars in Honey Charcoal Col Chrom, followed by detn of All
fractions6

Enzymatic methods

969.39 Glucose In Corn Syrups and Sugars Glucose oxidase mth (colorimetric detn Corn syrups
using o-dianisidine)
984.15 Lactose in Milk Hydr lactose with ff-galactosldase, released Milk
galactose meas by NAD red mth (UV)

S pecial Report : Journal of AO AC I nternational V ol . 76, N o. 6 ,1 9 9 3 189A

Table 6. (continued)
AOAC No. Method Brief description Applicable matrixes

Reducing sugar methods—copper reduction

906.03 Invert Sugar in Sugars and Syrups Munson-Walker mth Sugars and syrupsb
923.09 Invert Sugar in Sugars and Syrups Lane-Eynon gen vol mth using Soxhlet reag Sugars and syrups®’d
(mod Fehling soln)d
929.09 Invert Sugar in Sugars and Syrups Sugars and syrups
945.59 Invert Sugar in Sugars and Syrups® Sugars and syrups^
945.60 Invert Sugar in Sugars and Syrups Meissl-Hiller mth Sugars and syrups®
950.56 Invert Sugar in Sugars and Syrups® Quisumbing-Thomas mth Sugars and syrups
955.36 Invert Sugar in Sugars and Syrups Berlin Institute mth with Müller’s soin Sugars and syrups
970.58 Invert Sugar in Molasses Lane-Eynon const vol mth Molasses
931.07 Glucose and Sucrose in Eggs Dissol, pptn of protein, quanti by titr Eggs
936.06 Glucose in Cacao Products Defat, dissolv in H2O, prod clar with Pb Chocolate
acetate and quanti by mod
Munson-Walker mth (using Sichert-Bleyer
table for glucose)
938.18 Glucose in Cacao Products Defat, dissolv in H2O, quanti using Soxhlet Chocolate
reag, clar with Pb acetate, Zerban-Sattler
mthd
938.02 Glucose in Cacao Products Ref to 938.18 or 936.06
925.37 Glucose (Commercial) in Fruits and Fruit Ref to 930.37 Fruits and fruit prod
Products
933.02 Glucose in Plants Shaffer-Somogyi micro mth
935.62 Glucose in Sugars and Syrups Lane-Eynon and Munson-Walker mths
945.67 Glucose in Corn Syrups and Sugars® Steinhoff mth
959.11 Glucose in Sugars and Syrups Shaffer-Somogyi micro mth
935.63 Fructose in Sugars and Syrups Lane-Eynon and Munson-Walker mth
932.15 Fructose in Sugars and Syrups Jackson-Mathews, mod of Nyns Selective
mth (similar to Shaffer-Somogyi)
960.06 Fructose in Plants Somogyi-Micro or Munson-Walker mth
950.57 Arabinose, Galactose, and Xylose and other Shaffer-Somogyi Micro and other mths
Sugars in Sugars and Syrups
920.139 Sucrose in Lemon, Orange, and Lime Ref to 945.27,925.48,930.36
Extracts
902.02 Sucrose in Vanilla Extract Ref to 925.47, 925.48, 930.36
920.184 Sucrose in Honey Ref to 920.183
920.189 Sucrose in Maple Products Ref to 930.36
925.35 Sucrose in Fruits and Fruit Products Invert with HCI or invertase then
Munson-Walker mth
925.44 Sucrose in Food Dressings Munson-Walker % invert sugar bef/aft
inversion
930.07 Sucrose in Plants Invert sugar by HCI or invertase, then
Lane-Eynon mth
930.36 Sucrose in Sugars and Syrups Invert sugar with HCI or invertase,
Munson-Walker mth
950.51 Sucrose in Nuts and Nut Products Ref to 920.40®
927.07 Lactose in Meat Benedict Solution Mth
930.32 Lactose in Process Cheese Invert sugar with HCI, then Munson-Walker
mth
933.04 Lactose in Milk Chocolate
935.65 Lactose in Sugars and Syrups Munson-Walker or Lane-Eynon mth
945.48 Evaporated Milk—Lactose Ref to 896.01,930.28
952.05 Lactose in Bread Titr mth with Somogyi reag
984.15 Lactose in Milk Enzymatic mth
935.64 Maltose in Sugars and Syrups Munson-Walker or Lane Eynon mths
906.01 Sugars (Reducing) in Plants Munson-Walker mth
920.190 Sugars (Reducing) in Maple Products as
Invert Sugar

190A S pecial R eport : Journal of A O AC International V ol . 76, N o. 6 ,1 9 9 3

T a b le 6 . ( c o n tin u e d )

AOAC No. Method Brief description Applicable matrixes

Reducing sugar methods—copper reduction (continued)
920.183 Sugars (Reducing in Honey Munson-Walkerand Lane-Eynon mths
921.03 Sugars (Reducing) in Plants® Quisumbing-Thomas mth
925.15 Sugars in Roasted Coffee Munson-Walker or Lane-Eynon mths
925.36 Sugars (Reducing) in Fruits and Fruit Ref to 925.35
Products
925.42 Sugars (Reducing) Before Inversion in Food Munson-Walker mth
Dressings
925.43 Sugars (Reducing) After Inversion in Food Invert sugars by HCI then Munson-Walker
Dressings mth
925.52 Sugars in Canned Vegetables Munson-Walker mth on red sugars bef/aft
inversion with HCI
931.02 Sugars in Plants Clar with Pb acetate or ion-exchange resins
935.39G Sugars in Baked Products Ref to 975.14
945.29 Sugars (Total Reducing) in Brewing Sugars Munson-Walker or Lane-Eynon mths
and Syrups
945.66 Total Reducing Sugars Ref to 923.09 and 906.03
950.50 Sugars (Reducing) in Nuts and Nut Products Munson-Walker mth
968.28 Total Sugars in molasses as Invert Sugar Lane-Eynon Const Vol Vol mth
975.14 Sugars in Bread Munson-Walker mth for red sugars, for
sucrose ref to 925.05

Reducing sugar methods—ferricyanide reduction

939.03 Sugars (Reducing and Nonreducing) in Flour Clar with Na tungstate followed by titr with
ferricyanide and thiosulfateCereals,
potatoes, sweet mixes

Polarimetrie methods

896.01 Lactose in Milk

896.02 Sucrose in Sugars and Syrups Polarizing bef/aft inversion with HCI or Reducing Sugars
invertase
920.189 Sucrose in Maple Products Direct and invert polarizations bef/aft
inversion with HCI
920.115 Sweetened Condensed Milk-Sucrose Polarizing bef/aft inversion with HCI
925.35 Sucrose in Fruits and Fruit Products Polarizing bef/aft inversion with HCI
925.47 Sucrose in Sugars and Syrups Polarizing bef/aft inversion with invertase
925.46 Sucrose in Sugars and Syrups Saccharimeter reading, sample soln clar
with Pb acetate or alumina cream, ICUMA
mthc
925.48 Sucrose in Sugars and Syrups Polarizing bef/aft inversion with HCI

926.13 Polarizing bef/aft inversion with

Sucrose and Raffinose in Sugars and Syrups invertase-melibiase
926.14 Sucrose and Raffinose in Sugars and Syrups Polarizing bef/aft inversion with HCI
970.57 Sucrose in Molasses Polarizing bef/aft inversion with invertase
942.20 Sucrose in Sugar Beets Hot H2O digestion then polarization
945.55 Sucrose in Gelatin Polarizing bef/aft inversion with HCI
950.30 Sugars (Reducing) in Nonalcoholic Ref to 950.29
Beverages
950.29 Sucrose in Nonalcoholic Beverages Polarizing bef/aft inversion with HCI or
invertase
950.31 Glucose (Commercial) in Nonalcoholic Ref to 930.37
Beverages
930.37 Glucose (Commercial) in Sugars and Syrups

S p e c ia l R e p o r t : J o u r n a l of A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 191A
T a b le 6 . (c o n tin u e d )

AOAC No. Method Brief description Applicable matrixes

Infrared methods

972.16H„I Lactose in Milk Mid-infrared9

975.19 Lactose in Milk Ref to 972.16H„I

Gravimetric methods

930.28 Lactose in Milk Gravimetric mth Dairy prod

920.110 Lactose in Cream Ref to 930.28 Dairy prod

a Applicable to broad range of matrixes if samples are defatted and extracted before analysis.
6 International Commission for Uniform Methods of Sugar Analysis (ICUMSA) approved method.
c Lane-Eynon method may require filtration of reagents. Recommend not using asbestos as a filtering aid.
d Be cautious of potential interferences from oligosacch and lactic acid.
e Surplus method.
1 For low levels of invert in sucrose.
9 Must be calibrated against a primary method.

Table 7. Acceptable Official Methods for nutrition labeling

AOAC No. Method Matrixes3

ASH
940.26A" Ash of Fruits and Fruit Products A, J
920.153" Ash of Meat B, L, M
925.51 A6 Ash of Canned Vegetables C, P, T
950.14ä " Ash of Nonalcoholic Beverages D
920.100" Ash In Tea D
925.49" Ash of Confectionery E
972.15 Ash of Cocoa Products E
945.18 Ash of Cereal Adjuncts, Direct Method F
930.22 Ash of Bread, Direct Method F
935.42 Ash of Cheese, Gravimetric Method G
945.46 Ash of Milk, Gravimetric Method H
920.117 Ash of Butter H
930.30 Ash of Dried Milk H
920.108 Ash of Cream, Gravimetric Method H
945.48E Ash of Evaporated Milk H
920.115E" Ash of Sweetened Condensed Milk H
938.08" Ash of Seafood I, Q
986.25 Ash of Milk-Based Infant Formula,Gravimetric Method K
950.49" Ash of Nuts and Nut Products N
923.03 Ash of Flour, Direct Method R
935.39 Ash of Baked Products, Direct Method (without fruit) R
941.12 Ash of Spices, Gravimetric Method S

BETA CAROTENE
941.15 Carotene in Fresh Plant Materials, and Silages, Spectrophotometric Method (naturally A, C, J, P, T
occurring carotene only)

CALCIUM
968.08 Minerals in Animal Feed, AAS Method All
975.03 Metals in Plants, AAS Method All
980.03 Metals in Plants, Spectrographic Method All
965.09 Minor Nutrients in Fertilizers, AAS Method All

192A S p e c ia l R e p o r t : J o u r n a l o f AO AC I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3
T a b le 7. (c o n tin u e d )

AOAC No. Method Matrixes‘

CALCIUM (continued)
953.01 Metals in Plants, Emission Spectrographic Method A, C, F, J, P, T
985.01 Metals and Other Elements in Plants, ICP Method A, C, F, J, P, T
921.01 Ca in Plants, Micro Titrimetr c Method A, C, F, J, P, T
951.01 Ca in Plants, Nitrocreosol Method A, C, F, J, P, T
983.19 Ca in Mechanically Separatee Poultry and Beef Titrimetric Method B, L
968.31 Ca in Canned Vegetables, Titrimetric Method C, T
945.41 Ca in Bread, Titrimetric Method F
944.03 Ca in Flour, Titrimetric Method F
991.25 Ca, Mg, and P in Cheese, AAS and Colorimetric Methods G
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na, and Zn in Infant Formula, ICP Method K
985.35 Minerals in Ready-To-Feed Milk-Based Infant Formula, AAS Method K

CHOLESTEROL
976.26 Cholesterol in Multi-component Foods, GC Method (only 10 mg/100) Ail

COPPER
985.35 Minerals in Ready-To-Feed Milk-Based Infant Formula, AAS Method Ail
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na and Zn in Infant Formula, ICP Method Ail
960.40 Copper in Food, Colorimetric Method Ail
968.08 Minerals in Animal Feed, AAS Method Ail
975.03 Metals in Plants, AAS Method Ail
980.03 Metals in Plants, Spectrographic Method Ail
953.01 Metals in Plants, Emission Spectrographic Method A, C, F, J, P, T
985.01 Metals and Other Elements in Plants, ICP Method A, C, F, J, P, T
953.03 Cu in Plants, Colorimetric Method A, C, F, J, P, T
985.40 Cu in Liver, AAS Method L
990.05 Cu, Fe, Ni in Edible Oils and Fats, AAS Method O
CYANOCOBALAMINE
960.46 Vitamin Assays, Microbiological Method with 952.20 Cobalamin (Vitamin B12 Activity) Ail
in Vitamin Preparations, Microbiological Method
986.23 Cobalamin (Vitamin B12 Activity), in Milk-Based Infant Formula, Turbidimetric Method K

DIETARY FIBER-INSOLUBLE
991.42 Insoluble Dietary Fiber in Food and Food Products, Enzymatic-Gravimetric Method A, C-F, J, K, N, P,
R-T
991.43 Total, Insoluble and Soluble Dietary Fiber in Foods, Enzymatic-Gravimetric Method A, C-F, J, K, N, P,
R-T

DIETARY FIBER-SOLUBLE
991.42 Insoluble Dietary Fiber in Food and Food Products, Enzymatic-Gravimetric Method, A, C-F, J, K, N, P,
(by calc; direct method is under AOAC review) R-T
991.43 Total, Insoluble and Soluble Dietary Fiber in Foods, Enzymatic-Gravimetric Method A, C-F, J, K, N, P,
R-T

DIETARY FIBER-TOTAL
985.29 Total Dietary Fiber in Foods, Enzymatic-Gravimetric Method A, C-F, J, K, N, P,
R-T
991.43 Total, Insoluble and Soluble Dietary Fiber in Foods, Enzymatic-Gravimetric Method A, C-F, J, K, N, P,
R-T

S p e c ia l R e p o r t : J o u r n a l of AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 193A
T a b le 7 . (c o n tin u e d )

AOAC No. Method Matrixes'

FAT-POLYUNSATURATED
979.19 c/s„c/s-Methylene Interrupted Polyunsaturated Fatty Acids in Oils, Spectrophotometric All
Method (Suitable after isolation of fat.)
957.13 Polyunsaturated Acids in Oils and Fats, Spectrophotometric Method O

FAT-TOTAL
Total fat has been concisely defined under NLEA as the sum of lipid fatty acids expressed as triglycerides. Currently, no AOAC Official
Methods meet that definition per se. See the February 1993 issue of The Referee for the task force report on fat that contains a number of
methods that approximate the definition and may be useful on an interim basis until methods can be collaboratively studied.

FOLACIN
960.46 Vitamin Assays, Microbiological Method with 944.12 Folic Acid in Vitamin All
Preparations, Microbiol. Method (Suitable for free form only) (L. casei commonly
preferred, but not collaboratively studied.)
992.05 Folic Acid in Infant Formula, LC Method K

IRON
968.08 Minerals in Animal Feed, AAS Method All
975.03 Metals in Plants, AAS Method All
980.03 Metals in Plants, Spectrographic Method All
965.09 Minor Nutrients in Fertilizers, AAS Method All
953.01 Metals in Plants, Emission Spectrographic Method A, C, F, J, P, T
945.40 Iron in Bread, Spectrophotometric Method F
950.39 Iron in Macaroni Products, Spectrophotometric Method F
944.02B Iron in Flour, Spectrophotometric Method F
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na and Zn in Infant Formula, ICP Method K
985.35 Minerals in Ready-To-Feed Milk-Based Infant Formula, AAS Method K
990.05 Cu, Fe, Ni in Edible Oils and Fats, AAS Method O
MAGNESIUM
975.03 Metals in Plants, AAS Method Ail
980.03 Metals in Plants, Spectrographic Method Ail
965.09 Minor Nutrients in Fertilizers, AAS Method Ail
953.01 Metals in Plants, Emission Spectrographic Method A, C, F, J, P, T
985.01 Metals and Other Elements in Plants, ICP Method A, C, F, J, P, T
991.25 Ca, Mg, P in Cheese, AAS and Colorimetric Methods G
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na and Zn in Infant Formula, ICP Method K
985.35 Minerals in Ready-To- Feed Milk-Based Infant Formula, AAS Method K

MOISTURE
See the task force report on moisture in the January 1993 issue of The Referee.
NIACIN
960.46 Vitamin Assays, Microbiological Method with 944.13 Niacin and Niacinamide in
Vitamin Preparations, Microbiological Method Ail
961.14 Niacin and Niacinamide inDrugs, Foods and Feeds, Colorimetric Method Ail
981.16 Niacin and Niacinamide inFoods, Drugs and Feeds, Automated Method Ail
975.41 Niacin and Niacinamide inCereal Products, Automated Method F
985.34 Niacin and Niacinamide inReady-To-Feed Milk-Based Infant Formula, K
Microbiological-Turbidimetric Method

PANTOTHENIC ACID
960.46 Vitamin Assays, Microbiological Method with 945.74 Pantothenic Acid in Vitamin Ail
Preparations, Microbiological Method (Suitable for free form only)

194A S p e c ia l R e p o r t : J o u r n a l o f AO AC In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3
T a b le 7 . (c o n tin u e d )

AOAC No. Method Matrixes'

PHOSPHORUS
958.01 Total Phosphorus in Fertilizers, Spectrophotometric Molybdovanadophosphate Method All
957.18 Microchemical Determination of Phosphorus, Kjeldahl Digestion Method All
964.06 Phosphorus in Animal Feed, Aik Ammonium Molybdophosphate Method All
965.17 Phosphorus in Animal Feed, Photometric Method All
980.03 Metals in Plants, Spectrographic Method A,C, J, P, T
953.01 Metals in Plants, Emission Spectrographic Method A,C,:, J, P, T
985.01 Metals and Other Elements in Plants, ICP Method A,C,:, J, P, T
966.01 Phosphorus in Plants, Gravity Quinolinium Molybdophosphate Method A,C, J, P, T
931.01 Phosphorus in Plants, Micro Method A,C, J, P, T
970.39 Phosphorus in Fruits and Fn.it Products, Spectrophotometric Molybdovanadate
Method A, J
969.31 Phosphorus in Meat, Aik Ammonium Molybdophosphate Method B, L
972.22 Phosphorus in Meats, Automated Method B, L
991.27 Phosphorus in Meat and Meat Products, SpectrophotometricMethod B, L
948.09 Phosphorus in Flour F
990.24 Total Phosphorus in Cheese and Processed Cheese, Photometric Method G
991.25 Ca, Mg, P in Cheese, AAS and Colorimetric Methods G
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na, and Zn in Infant Formula, ICP Method K
986.24 Phosphorus in Milk-Based Infant Formula, Spectrophotometric Method K
935.59 Total Phosphorus in Food Dressings, Aik Ammonium Molybdophosphate Method O
930.35G, H, I Phosphorus in Vinegars, Aik Ammonium Molybdophosphate Method S

POTASSIUM
985.35 Minerals In Ready-To-Feed Milk-Based Infant Formula, AAS Method All
956.01 K/Na in Plants, Flame Photometric Method A,C,:, J, P, T
953.01 Metals in Plants, Emission Spectrographic Method A,C,:, J, P, T
985.01 Metals and Other Elements in Plants, ICP Method A,C,:, J, P, T
975.03 Metals in Plants, AAS Method A,C,:, J, P, T
980.03 Metals in Plants, Spectrographic Method A,C, J, P, T
965.30 K in Fruits and Fruit Products, Flame Photometric Method A, J
990.23 Na and K in Dried Milk, Flame Emission Spectrophotometric Method H
969.23 Na and K in Seafood, Flame Photometric Method I,Q
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na, and Zn in Infant Formula, ICP Method K

PROTEIN
981.10 Crude Protein in Meat, Block Digestion Method All
955.04 Total N in Fertilizers, Kjeldahl Method All
977.02 Total N (Crude Protein) in Plants, Automated and Semiautomated Methods A,C, J, P, T
978.04 Total N (Crude Protein) in Plants, Kjeldahl Methods A,C, J, P, T
920.152 Protein in Fruit Products, Kjeldahl Method A, J
992.15 Crude Protein in Meat and Meat Products, Combustion Method B, L
939.02 Milk Protein in Milk Chocolate, Kjelcahl Method E
990.03 Crude Protein in Animal Feed, Combustion Method F
990.02 Crude Protein in Animal Feed, Semiautomated Method F
920.87 Total Protein in Flour F
945.18B Protein in Cereal Adjuncts, Kjeldahl Method F
930.25 Protein in Macaroni Products F
950.36 Protein in Bread F
991.20 Total Nitrogen Content of Milk, Kjeldahl Method H
991.22 Protein Nitrogen Content of Milk, Kjeldahl Method (Direct) H
991.23 Protein Nitrogen Content of Milk, Kjeldahl Method (Indirect) H
930.33 Protein in Ice Cream and Frozen Desserts, Kjeldahl and DyeBinding Methods H
930.29 Protein in Dried Milk, Kjeldahl Method H
967.12 Protein in Milk, Dye Binding Method I H

S p e c ia l R e p o r t : J o u r n a l of A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 I95A
T a b le 7 . ( c o n tin u e d )

AOAC No.

PROTEIN (continued)
975.17 Protein in Milk. Dye Binding Method II H
975.18 Protein in Milk, Mid-Infrared Spectrophotometric Method H
972.16 Fat, Lactose, Protein, and Solids in Milk, Mid-Infrared Spectrophotometric Method K
986.25C Protein in Milk-Based Infant Formula, Kjeldahl Method K
928.08 Nitrogen in Meat, Kjeldahl Method L
977.14 Nitrogen in Meat, Automated Kjeldahl Method L
981.10 Nitrogen in Meat, Block Digestion Method L
950.48 Crude Protein in Nuts and Nut Products, Kjeldahl Method N
935.39C Protein in Baked Products R

PROTEIN QUALITY
991.29 True Protein Digestibility of Foods and Food Ingredients, Rat Bioassay Method with All
982.30 Protein Efficiency Ratio, Calc Method
960.48 Protein Efficiency Ratio, Rat Bioassay Method All

PYRIDOXINE
960.46 Vitamin Assays, Microbiological Method with 961.15 Vitamin B6 in Food Extracts, All
Microbiological Method
985.32 Vitamin B6 in Ready-To-Feed Milk-Based Infant Formula, Microbiological Method K

RIBOFLAVIN
960.46 Vitamin Assays, Microbiological Method with 960.33 Riboflavin (Vitamin B2) in Vitamin All
Preparations, Microbiological Method
970.65 Riboflavin (Vitamin B2) in Foods and Vitamin Preparations, Fluorometric Method All
981.15 Riboflavin in Foods and Vitamin Preparations, Automated Method All
985.31 Riboflavin in Ready-To-Feed Milk-Based Infant Formula, Fluorometric Method K

SODIUM
985.35 Minerals in Ready-To-Feed Milk-Based Infant Formula, AAS Method All
956.01 K/Na in Plants, Flame Photometric Method A, C, F, J, P, T
953.01 Metals in Plants, Emission Spectrographic Method A, C, F, J, P,T
980.03 Metals in Plants, Spectrographic Method A, C, F, J, P,T
966.16 Na in Fruits and Fruit Products, Flame Spectrophotometric Method A, J
990.23 Na and K in Dried Milk, Flame Emission Spectrophotometric Method H
969.23 Na and K in Seafood, Flame Photometric Method I,Q
976.25 Na in Foods for Special Dietary Use, ISE Method K
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na and Zn in Infant Formula, ICP Method K

THIAMINE
942.23 Thiamine (B1) in Foods, Fluorometric Method All
957.17 Thiamine (B1) in Bread, Fluorometric Method F, R
953.17 Thiamine (B1) in Grain Products, Fluorometric Method F, R
986.27 Thiamine (B1) in Milk-Based Infant Formula, Fluorometric Method K

TRYPTOPHAN
960.46 Vitamin Assays, Microbiological Method with 988.15 Tryptophan in Foods and Food All
and Feed Ingredients, Ion Exchange Chromatographic Method

VITAMIN A
974.29 Vitamin A in Mixed Feeds, Premixes, and Foods, Colorimetric Method B, G, H, I, K, L, M,
Q
992.04 Vitamin A (Retinol Isomers) in Milk and Milk-Based Infant Formula, LC Method H, K
992.06 Vitamin A (Retinol) in Milk-Based Infant Formula, LC Method K

196A S p e c ia l R e p o r t : J o u r n a l o f A O AC In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3
T a b le 7 . (c o n tin u e d )

AOAC No. Method Matrixes3

VITAMIN C
984.26 Total Vitamin C in Food, Semi-automated Fluorometric Method A, C-H, J, K, P, R, T
967.22 Vitamin C (Ascorbic Acid) in Vitamin Preparations, Microfluorometric Method A, C-H, J, K, P, R, T
967.21 Vitamin C (Ascorbic Acid) in Vitamin Preparations and Juices, Titrimetric Method D
(Colorless juices only)
985.33 Vitamin C (Reduced Ascorbic Acid) in Ready-To-Feed Milk-Based Infant Formula, K
Titrimetric Method

VITAMIN D
982.29 Vitamin D in Mixed Feeds, Premixes, and Pet Foods, LC Method All
936.14 Vitamin D in Milk, Vitamin Preparations and Feed Concentrates, Rat Bioassay Method All
981.17 Vitamin D in Fortified Milk and Milk Powder, LC Method G, H, K

VITAMIN E
971.30 a-Tocopherol and a-Tocopheryl Acetate in Foods and Feeds, Colorimetric Method All
992.03 Vitamin E Activity in Milk-Based Infant Formula, LC Method K

ZINC
968.08 Minerals in Animal Feed, AAS Method All
975.03 Metals in Plants, AAS Method All
980.03 Metals in Plants, Spectrographic Method All
965.09 Minor Nutrients in Fertilizers, AAS Method All
986.15 As, Cd, Pd, Se, and Zn in Food, Multielement Method All
944.09 Zn in Food, Colorimetric Method All
969.32 Zn in Food, AA Method All
953.01 Metals in Plants, Emission Spectrographic Method A, C, F, J, P, T
985.01 Metals and Other Elements in Plants, ICP Method A, C, F, J, P, T
941.03 Zn in Plants, Mixed Color Method A, C, F, J, P, T
953.04 Zn in Plants, Single Color Method A, C, F, J, P, T
984.27 Ca, Cu, Fe, Mg, Mn, P, K, Na and Zn in Infant Formula, ICP Method K
985.35 Minerals in Ready-To-Feed Milk-Based Infant Formula, AAS Method K

8 Matrixes were considered representative of all food types and are as follows: A, fruit baby foods; B, meat baby foods; C, vegetable baby
foods; D, beverages and juices; E, candy; F, cereals and products; G, cheese; H, dairy products; I, fish; J, fruits; K, infant formula or medical
diet; L, meat (beef, pork, or fowl); M, mixed dinners (TV dinners); N, nuts; O, oils/fats (dressings); P, potatoes and products; Q, shellfish; R,
sweet mixes (cake, pie, etc.); S, spices; and T, vegetables.
6 These methods refer to 900.02A, Ash of Sugar, in surplus status, no longer in use with sugar. Method 923.03 may be substituted.

Thanks to the efforts of M. Bueno, D. Carpenter, H. Chin, M. Deutsch, J. DeVries, N. Fraley, W. Hummer, W. Ikins, W. Landen, S. Lee,
J. Morawski, P. Oles, L. Prosky, D. Soderberg, A. Soliman (deceased), D. Sullivan, J. Tanner, W. Wolf.

the needs of the new regulations. New Whereas SRMs arereadily availablefor S u b c o m m itte e on R e feren ce M a te r i­
AOAC Associate Referees have been many inorganic compounds of environ­ a ls .
—The Task Force on Methods for
appointedinseveral areas,and collabo­ mental or manufacturing interest, food- NutrientLabeling Analyses established
rativestudiesonneededmethodsarebe­ based SRMs containing labile nutrients aSubcommitteeon ReferenceMaterials
ingpursued. are less available. Even though many toaddresstheissuesinvolvedwithiden­
laboratories have tried to overcome this tifyingandpublicizingreferencemateri­
Identification o f N e e d s fo r
deficiencyby developing in-houserefer­ alsavailableforusewithcurrentAOAC
R e fe re n c e M a te ria ls
encematerials,therehavebeenno guide­ OfficialMethods fornutrition analysis.
The task force recognized that ana­ linesastocertificationofthesematerials They also addressed the need foraddi­
lytical laboratories do not have a great or theirdevelopment and handling. Two tional reference materials for current
dealofresourcestodrawfromforstand­ subcommittees were formed to address validatedmethods anddevelopedanap­
ard reference materials (SRMs), certi­ referencematerialsandguidelinesforde­ proach for incorporating reference ma­
fied reference materials, or quality as­ velopingin-houseanalyticalqualityassur­ terialsaspartofthemethods validation
surance materials fornutrient analyses. ancecontrolmaterials,respectively. process.

S p e c ia l R e p o r t : J o u r n a l of AO AC In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 197A
T a b le 8 . M a tr ix e s n e e d in g m e t h o d s

Recommendations Matrixes3 Recommendations Matrixes3

ASH (D. Soderberg) O • Cs recommended for 960.46 + 944.12 + FDA

• Recommend replace references to 900.02 SOP 315 to use L. case/'for free form
with 923.03 • Cs recommended for 960.46 + 944.12 + FDA
BETA CAROTENE (M. Deutsch) B, D—I, K-O, SOP 320 to use S. faecalei for products
Q-S containing free form
• Need method/cs6 and source of standards • Cs recommended for 960.46 + FDA SOP 330
• Need method to separate c is and tra ns to use K. apriculata for bound form
isomers IODINE (Open0) All
• Need method for encapsulated products • Need method/cs; consideration of ceric
BIOTIN (M. Deutsch) All sulfate reduction recommended
• Need method/cs IRON (Open c)
• Modification of 960.46 + FDA SOP 305 to use • Recommend extension of applicability
L. p la n ta ru m recommended statements to foods for 965.09, 968.08,
985.35, 975.03, 980.03
CARBOHYDRATES-SUGAR ALCOHOLS (M. All
Clarke) MAGNESIUM (Open c)
• Need method/cs • Recommend extension of applicability
statements to foods for 965.09, 985.35,
CARBOHYDRATES-SUGARS (M. Clarke)
975.03, 985.35, 980.03
• Cs of HPLC method with wide matrix
• Recommend extension of applicability
applicability recommended
statement to A-C, F-H, K-M, P, Q, T for 984.27
CHOLESTEROL (M. Deutsch) All <10 mg/100
MOISTURE (Open c) B, F, I, J, L-O,
• Cs underway on direct saponification method Q, R for high
COPPER (Open c) sugar products
• Need cs for Cu as analyte in 968.08, 975.03, • Recommend modification of 952.08 to
965.09, 984.27, 985.35, 970.18, 980.03 eliminate asbestos use
• Add note to 968.08 that Pt dishes are • See Moisture Report in the January 1993
preferable, that losses may occur using issue of The Referee
porcelain dishes NIACIN (M. Deutsch)
CYANOCOBALAMINE (M. Deutsch)
• Recommend extension of applicability
• Cs recommended for 960.46 + 952.20 to use statement to B, D, F-H, J, P for 985.24
L le ic h m a n u i for all matrixes • Recommend extension of applicability
DIETARY FIBER-INSOLUBLE (L. Prosky) statement to all foods for 944.13
• Cs recommended for sample preparation of • Recommend cs on niacin in fortified foods
high-fat samples. using HPLC method in JAOAC (in press)
DIETARY FIBER-SOLUBLE (L. Prosky) • Cs recommended for 960.46 + 944.13 to use
• Cs recommended for sample preparation of L. plantarum for all matrixes
high-fat samples. PANTOTHENIC ACID (M. Deutsch) All
DIETARY FIBER-TOTAL (L. Prosky) • Cs recommended for 960.46 + 945.74 to use
• Cs recommended for sample preparation of L. plantarum for free form
high-fat samples. • Cs recommended for 960.46 + FDA SOP 350
DRAINED WEIGHT (Open c) + USDA 97 to use L plantarum for total
• Need procedure for pickled veg., fruits, and pantothenic acid for all matrixes
tuna • Cs recommended for 960.46 + FDA SOP 340
FAT-MONOUNSATURATED (D. Firestone) All to use A. suboxydaus for panthenol
• Recommend adoption of AOCS method PHOSPHORUS (Open c)
where applicable • Recommend extension of applicability
FAT-POLYUNSATURATED (D. Firestone) statements to foods for 958.01,957.18,
• Recommend revision of extraction procedure 964.06, 965.17
FAT-SATURATED (D. Firestone) A, C, J, T • Recommend extension of applicability
• Recommend revision of extraction procedure statement to A, E-M, O-R, T for 986.24
• Modification of 969.33 to extend matrixes and POTASSIUM (Open0)
quantitation of fatty acids in all foods • Recommend extension of applicability
recommended statement to include food for 985.35
• Recommend cs for samples <350 mg fat • Recommend extension of applicability
content statement to C, K, O for 984.27
• Need method for C-4 to C-6 fatty acids PROTEIN (D. Soderberg)
FAT-TOTAL (D. Firestone) All • Recommend modification of Kjeldahl method
• Need quantitation method for fatty acids to include spices
FOLACIN (M. Deutsch) • Recommend cs of Dumas for all matrixes

198A S p e c ia l R e p o r t : J o u r n a l o f AO AC I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3
Table 8. (co n tin u ed )

Recommendations Matrixes3 Recommendations Matrixes3

PYRIDOXINE (M. Deutsch) • Recommend modification of 984.26 to include

• Recommend cs based on JAOAC 39, meats, nuts, seafood, spices, and oils/fats
157(1956) • Recommend adoption of ACNA HPLC
• Cs recommended for 960.46 + 961.15 to use method, JAOAC 75, 887(1992), for vitamin C
S. u v a ru m for total Bg, all matrixes in variety of foocs
RIBOFLAVIN (M. Deutsch) VITAMIN D (M. Deutsch)
• Recommend cs for HPLC method • Recommend extension of applicability
• Recommend extension of applicability statements to foods for 982.29 936.14
statement to foods for 940.33 • Recommend cs on GPC/HPLC method,
• Recommend modification of 970.65 and NIST-Anal. Chem. 60,1929(1988)
981.15 to incorporate standard addition • Recommend cs using HPLC method,
• Cs recommended for 960.46 + 940.33 to use reversed or normal phase
L. case/'for all matrixes VITAMIN E (M. Deutsch)
SAMPLING/SAMPLE PREP (Open c) • Recommend cs of HPLC method
• Recommend completion of AOAC book on • Recommend modification of current methods
sampling with solvent substitution for chloroform
• Recommend inclusion of sampling and • Need further investigation into standards,
sample preparation in future AOAC methods methodology, and standard conversion factors
SODIUM (Open c) ZINC (Open °)
• Recommend extension of applicability • Recommend extension of applicability
statement to food for 985.35 statements to foods for 985.35, 968.08,
• Recommend extension of applicability 975.03, 980.03, 965.09
statement to C, K, O for 984.27 • Recommend extension of applicability
STEARIC ACID (D. Firestone) statement to A-F, H-M, O, P, R, T for 984.27*2
0
• Need quantitative method
a Those analytes for which no matrix is specified can be analyzed
THIAMINE (M. Deutsch) using AOAC Official Methods listed in the March 1993 issue of
• Recommend extension of applicability The Referee. The methods in that list are deemed to be adequate
statement to all food for 986.27 on the basis of collaborative studies and/or common use for the
analyte/matrix combinations listed although not all analyte/matrix
• Recommend extension of applicability
combinations listed have been fully collaboratively studied. See
statement to D, H-J, Q for 957.17 Appendix 2, Table 1, for matrix identification.
• Cs recommended for simultaneous b cs, collaborative study.
determination of B-,, B2, and Be using c No current general referee (GR) for these specific analytes/topics.
fluorescence detection, JAOAC 75, 561 (1992) Contact AOAC International, Technical Services, 2200 Wilson
• Cs recommended for 960.46 + FDA SOP 360 Blvd, Suite 4C0, Arlington, VA 22201, +1 (703) 522-3032.
to use L. virid e sce n s for all matrixes
Note: GR for analytes in infant formula/medical diets is M. Bueno,
TRYPTOPHAN (M. Deutsch) FDA HFF-266, 200 C St, SW, Washington, DC 20204,
• Cs recommended for FDA SOP 500, 501, and +1 (202) 205-4445.
502, HPLC for free form AOAC General Referees and Their Topics
• Cs recommended for 960.46 + FDA SOP 365 ■ D. Sodemerg.—Meat, Poultry, Meat and Poultry
Products—USDA/FSIS, 300 12th St, SW, Washington, DC 20205,
to use L. p la n ta ru m for bound and free forms
+1 (202) 205-0039
VITAMIN A (M. Deutsch) A, C-F, J N-P, ■ M. Deutsch.—Vitamins and Other Nutrients—FDA, HFF-264,
R-T 200 C St, SW, Washington, DC 20204, +1 (202) 205-4263
• Cs recommended for new HPLC method ■ M. Clarke.—Sugars and Sugar Products—Sugar Processing
based on 992.04, for other than infant formula Research, Inc., 1100 Robert E. Lee Blvd, New Orleans, LA 70124,
to apply to variety of matrixes. Modify sample +1 (504) 286-4542
■ L. Prosky.—Dietary Fiber—FDA, HFF-268, 8301 Muirkirk Rd,
size and replace ambient 18 h with high
Laurel, MD 20708, +1 (301) 344-6005
temp/short time saponification ■ D. Firestone—Fats & Oils—FDA, HFF-426, 200 C St, SW,
VITAMIN C (M. Deutsch) B, I, L-O, Q, S Washington, DC 20204, +1 (202) 205-4381Thanks to the efforts
• Need method/cs using HPLC to separate of M. Bueno, D. Carpenter, H. Chin, M. Deutsch, J. DeVries,
erythorbate, ascorbate, and dehydroascorbate N. Fraley, W. Hummer, W. Ikins, W. Landen, S. Lee, J. Morawski,
P. Oles, L. Prosky, D. Soderberg, A. Soliman (deceased),
D. Sullivan, J. Tanner, W. Wolf.

S p e c ia l R e p o r t : J o u r n a l of AO AC In t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3 199A
 The subcommittee surveyed the ref­ Analysis fo r Nutrition Labeling, D.M. of this more systematic approach toward
erence materials currently available to Sullivan & D.E. Carpenter (Eds), classifying foods should greatly sim­
laboratories doing nutrition analysis. As AOAC INTERNATIONAL, Arlington, plify the collaborative study process and
a result of this survey, a report on their VA, Chapter 7). the process of extending the applicabil­
availability, compiled by analyte and Definition of Food Matrices.—The ity of methods.
matrix, along with their sources, was task force selected twenty food groups Likewise, there is a great deal of con­
published in the August 1992 issue of to be used in evaluating the adequacy of cern regarding the availability of refer­
The Referee (see also Wolf, W.R. (1993) methods for nutrition labeling. The ence materials relevant to the particular
Methods o f Analysisfor Nutrition Label­ available methods for a given nutrient food under study. With this in mind, the
ing, D.M. Sullivan & D.E. Carpenter were evaluated specifically for each ma­ task force recommends the development
(Eds), AOAC INTERNATIONAL, Ar­ trix group of foods (e.g., fruits, fish and or selection of a series of reference sam­
lington, VA, Chapter 7) for those nutri­ shellfish, beverages, or meats). Whereas ples that could be used by laboratories
ents whose inclusion is mandatory on such groupings are reasonable from a during the methods validation process.
the nutrition label. dietary perspective, the groups are nei­ To encourage the incorporation of refer­
Recognizing the ongoing needs in ther inclusive nor exclusive from the ence materials into the methods valida­
the area of reference materials for perspective of the applicability of ana­ tion process, the Subcommittee on Ref­
method development, validation, and lytical methods. The twenty food groups erence Materials determined that the
quality control, and further recognizing selected by the task force as a basis for reference materials to be incorporated
the need to define reference materials judging methods currently validated by must be truly representative of food or
that will represent food and food matri­ AOAC have certain commonalities, and food matrices, readily available, and
ces during methods validation work, the in many cases, their characteristics over­ available at prices that encourage their
task force supports the formation and lap. These matrix characteristics are as use. For reference materials to be repre­
development of a Technical Division on follows: fat (high or low), moisture sentative of foods or food matrices,
Reference Materials within AOAC. (high or low), protein (high or low), and those matrices must be clearly defined,
Task force members have been suppor­ carbohydrate (high or low); or they are and the definitions must be accepted
tive of and active in initiating such a at an intermediate level in all these ma­ throughout the validation process. The
technical division. A group of scientists, jor components. The task force recog­ reference samples would have charac­
of which many task force members were nized that the ability to categorize foods teristics such that the set of samples
a part, met at AOAC headquarters on on a more fundamental chemical basis would represent all the food matrices
December 8 , 1992, to petition the would facilitate both the development of and would allow the collaborators who
AOAC Board of Directors to form the analytical methods and appropriate have completed a collaborative study
Technical Division. The Board of Direc­ standard reference materials. A subcom­ using this set to be confident of the ap­
tors approved the Technical Division on mittee was formed to develop a proposal plicability of their method. The task
Reference Materials on March 23,1993. for defining foods in terms of a matrix of force recommends that this approach
This group will be charged with pursu­ the major food components (i.e., fat, [outlined by the Subcommittee on Ref­
ing the following recommendations: the protein, and carbohydrate). erence Materials and reported in The
identifying and/or developing reference Referee (1993) 17(7), p. 1,6-7; see also
The subcommittee determined that
Wolf, W.R. (1993) Methods o f Analysis
materials for quality control of methods defining foods of a particular matrix
for Nutrition Labeling, D.M. Sullivan &
for nutrition analysis, identifying and/or group on the basis of fat, protein, and
D.E. Carpenter (Eds), AOAC INTER­
developing a set of reference materials carbohydrate content would be both
NATIONAL, Arlington, VA, Chapter 7]
representative of foods for use in valida­ possible and advantageous. The mois­
be adopted by the Official Methods
tion studies, and facilitating the avail­ ture level of a sample can readily be ad­
Board of AOAC INTERNATIONAL
ability of reference materials to labora­ justed by drying or adding water. Thus,
for future collaborative studies used for
tories in a cost effective manner. if a method is affected by the moisture
the methods validation process for
Subcommittee on in-house control level of the matrix, this can be handled.
analyses of foods for nutrition labeling
materials.—The Subcommittee on In- The remaining major components of the
purposes.
House Control Materials surveyed the samples and their various combinations
AOAC membership regarding the for­ of levels can be represented by a set of Follow-up Recom m endations
mulation, validation, and use of in- nine samples that span specific ranges of
house quality control materials. Using fat, protein and carbohydrate content. On January 6,1993, USDA and FDA
the survey results where possible, the Deliberations within the task force and issued the final regulations to be oper­
subcommittee developed a set of guide­ at other AOAC forums have indicated a able under NLEA (Federal Register,
lines to aid laboratories in the develop­ great deal of concern about the level of January 6,1993, pp. 631-691 and 2065-
ment of in-house quality control materi­ effort necessary to extend the applicabil­ 2964, respectively). The Task Force on
als. These guidelines were published in ity of Official Methods from foods spe­ Methods for Nutrient Labeling Analy­
the May 1993 issue of The Referee (see cifically studied in a collaborative study ses, after reviewing the final regulations,
also Wolf, W.R. (1993) Methods of to other foods of interest. The adoption again reviewed the status of AOAC Of­

200A S pecia l R eport : J o u r n a l o f A O A C In ternational V o l . 76, N o. 6 ,1 9 9 3

ficial Methods for nutrition labeling in its individual members and the AOAC N. Fraley, Armour Swift Eckrich,
light of changes made since the regula­ staff for their time, efforts, and devotion.
Inc.
tions were proposed. Updated reports The accomplishments of this task force W. Hummer, U.S. Food and Drug
[listing methods deemed adequate (Ta­ have demonstrated the importance and Administration
ble 7) and also methods in need of devel­ value of AOAC INTERNATIONAL W. Dans, Silliker Laboratories Inc.
opment and validation, updating, revi­ and the forum it provides for regulatory,
A. Kistler, Pet Inc.
sion, or further work to extend their academic, and industrial analytical
S. Lee, Kellogg Co.
applicability (Table 8 )] from the review chemists to meet and resolve issues of
W. Landen, U.S. Food and Drug Ad­
were published in the March and April concern. The benefits derived are of
ministration
1993 issues of The Referee. value to the scientific and regulatory
The final regulations, in response to community and to the general public. R. Lane, University of Alabama
concerns expressed regarding the meas­ J. Morawski, Waters Associates
urement and labeling of total and satu­ C om m ittee M em bers an d J. Ngeh-Ngwainbi, KeDogg Co.
rated fats, adopted a concise definition P articip an ts P. Oles, Lancaster Laboratories Inc.
for each analyte. Total fat is defined as L. Prosky, U.S. Food and Drug Ad­
the sum of the lipid fatty acids (ex­ J. DeVries, Chair, General Mills Inc. ministration
pressed as their triglycerides), and satu­ S. Bailey, Lancaster Laboratories L. Reimann, Woodson Tenent Labo­
rated fat is defined as all lipid fatty acids Inc. ratories
containing no double bonds. The Task K. Boyer, Southern Testing & Re­ A. Sheppard, U.S. Food and Drug
Force Subcommittee on Total Fat Meth­ search Laboratories Administration
ods reassessed the information reported M. Bueno, U.S. Food and Drug Ad­
D. Soderberg, U.S. Department of
in the September 1992 issue of The ministration
Agriculture
Referee on validated methods for fat in D. Carpenter, Kraft General Foods
A. Soliman (deceased)
light of the adopted definition and issued H. Chin, National Food Processors
an updated report in The Referee in Feb­ Association D. Sullivan, Hazleton Laboratories
ruary 1993. N. Craft, Southern Testing and Re­ J. Tanner, U.S. Food and Drug Ad­
The last meeting of the task force was search Laboratories ministration
held on February 8 , 1993. It has com­ M. Deutsch, U.S. Food and Drug Ad­ E. Waysek, Hoffman La Roche Inc.
pleted its mission and has put into place ministration W. Wolf, U.S. Department of Agri­
procedures that will assure the availabil­ D. Emery, Emery & Associates culture
ity of validated analytical methods for D. Firestone, U.S. Food and Drug E. Young, U.S. Food and Drug Ad­
nutrients in foods. The task force thanks Administration ministration

S pecia l R epo r t : J o u rn a l o f A O A C In ter n a tio na l V o l . 76, N o. 6 ,1 9 9 3 201A

C u m u l a t i v e A u t h o r I n d e x

Adamo, Nicholas C., 287 Cairns, Thomas, 306 Duran-Meras, Isabel, 754
Adams, Freddy C., 1262 Campbell, Harold M„ 1163 Edberg, Ulla, 53
Akasaka, Kazuaki, 1385 Cancalon, Paul E, 584 Edgell, Kenneth W., 1098,1113
Al-Hasani, Sami M., 902,1014 Cardwell, Terence J., 1389 Eerola, Susanna, 575
Al-Showiman, S.S., 601 Carignan, Germain, 325 Eichelberger, James W., 72
Al-Tamrah, S.A., 601 Carleer, Robert, 1138 Eilers, Paul P., 1344
Al-Warthan, A.A., 601 Carpenter, Mark W, 902 Eitenmiller, Ronald R., 390,1276
Alanko, Timo, 674 Carrier, Karen, 325 Elkins, Edgar R., 4
Albert, Richard, 461 Carro, A., 608 Ellis, Richard L„ 907, 1309
Ali, M. Sher, 907,1309 Carson, Mary C.. 329 Emara, Laila H., 847
Allen, Geraldine, 430 Casais, C., 608 Erb, Elizabeth J., 1098,1113
Amaguana, R. Miguel, 1240 Castle, Laurence, 760 Espinosa-Mansilla, A., 1255
Anderson, Kim A., 910 Cattrall, Robert W„ 1389 Fabbrini, Riccardo, 650
Anderson, Ellen M., 682 Cela, R., 608 Falbo-Nelson, Maria T., 694, 988
Andrews, Paul, 703, 707 Cepeda, A., 838 Fan, Titan S., 851
Andrews, Wallace H., 1240 Chakravarty, Sumon, 604 Feldsine, Phihp T., 694,988
Angyal, Gerald, 414,682 Chamkasem, Narong, 691 Felkner, I. Cecil, 682
Aoki, Kazuharu, 436 Chase, Jr, G. William, 390,1276 Fiddler, Walter, 578
Armentia-Alvarez, Arantza, 565 Chatt, A., 703 Fleming, J. Richard, 1033, 1300
Armishaw, Paul, 1317 Chen, Guo Nan, 1389 Fletouris, Dimitrios J., 1168
Atkinson, June, 620 Chen, Junshi, 1193,1206 Franco, C., 838
Atwal, AvtarS., 1010 Chiavarmi, S., 1133 Fujii, Yuko, 436
Baker, Randall J., 913 Chiba, Mikio, 1187 Fukaya, Megumu, 436
Bakowski, Ralph S., 907,1309 Chichila, Tina M.P. 1323 Fuleki, Tibor, 59,591
Balasubramanian, N., 730 Chou, Chiu L., 794 Gagnon, Jacques, 26
Bandler, Ruth, 430 Chowdhury, Bhabadeb, 1152 Gales, Peter W., 918
Baneijee, ArunB., 1152 Cleroux, Chantal, 14 Gao, Junquan, 1193,1206
BaOsman, A.A., 601 Cole, Richard J., 637,983 Garcia-Moreno, Concepcion, 565
Barbano, David M., 1033,1300 Coleman, Mark R., 945 Gilbert, John, 760
Barford, Robert A., 335 Collins, Peter G., 381 Gilvydis, DaliaM., 1323
Barker, Steven A., 67,663 Conacher, Henry B.S., 14, 26, 703 Ginn, Roy E., 297
Barnett, Stephen A., 399,1042 Conrad, Steve M., 1344 Glaze, Larry E., 44
Bauer, Karin, 864 Cox-Trout, Candace, 1174 Gliksman, Joseph E., 920
Beaulieu, Normand, 962 Cremisini, C, 1133 Gooch, Emmett G., 581
Beckert, Werner F., 555, 864 Crisippi, Teodoro, 650 Gopal, Madhuban, 283
Beljaars, Paul R., 570 Cutrufelli, Mark E., 1022 Graham, Russell A., 14
Bicchi, Carlo, 657 D’Amato, Angela, 657 Graham, Susan J., 962
Bicsak, Ronald C., 780 D’Aoust, Jean-Yves, 814 Greco, Paula, 814
Bidasee, Keshore R., 366 Dabeka, Robert W., 14 Gunderson, Ellis L., 492
Blankenship, Paul D., 637 Dacunha, Adrian R., 1230 Guy, Robert D., 794
Boisen, Flemming, 674 Daft, James L., 1083 Haggett, T.O. Richard, 1280
Botsoglou, Nikos A., 1168 Danielson, James W, 355 Hamilton, Ian C., 1389
Bowe, Susan, 14 Dauphin, C. 1295 Hammack, Thomas S., 1240
Brands, Arie, 570 Davidson, Sandra, 956 Hamon, M., 1295
Brassard, Rene, 923 de la Peña, A. Muñoz, 1255 Hanna, George M., 526
Breitholtz-Emanuelsson, Anna, 842 Deb, Manas Kanti, 604 Harnett, Michelle, 1280
Brousseau, Robert, 26 Decker, Eric Andrew, 644 Harris, Loralyn, 822
Bruelhart, Milene, 268, 275 Diachenko, Gregory W., 1213 Hashimoto, Hideki, 32
Bui, Lap V., 966 Dickerson, Richard, 1220 Hasselberger, Mary Lee, 39
Burini, Giovanni, 1017 Doerr, Robert C., 578 Hayakawa, Junko, 436
Bushway, Rodney J., 90, 851 Donnelly, Joseph R., 1092 Heame, Loretta A., 1400
Cabanis, Jean-Claude, 1262 Domer, Joe W, 637,983 Hennig, Bernhard, 644
Cabras, Paolo, 92 Duke, Paul D., 320 Hermida, M., 838

202A J o u rn a l O f A O A C I n ternational V o l . 76, N o . 6 ,1 9 9 3

C u m u la t iv e A u t h o r I n d e x

Higgins, DonL., 831 Kumar, B.S.M., 730 Mongeau, Roger, 923

Hinkkanen, Riitta, 575 Kuntom, Ainie, 371 Mooser, André E., 976
Hipp, Sharon, 394 Kurtz, David A., 735 Moran, John W, 945
Hirvi, Irnio, 575 Lacroix, Gladys M.A., 14 Mori, Brian, 26
Hlavac, Jan, 902,1014 Landen, Jr, William O., 390,1276 Mortimer, Richard D., 377
Ho, James S., 72 Lati.E., 1295 Mountford, Mardi K„ 399,1042
Holland, David C„ 720 Latimer, Jr, George W., 749 Mukherjee, Irani, 283
Hollifield, Henry C„ 1213,1268 Lau-Cam, Cesar A., 526 Mullens, Jules, 1138
Hopper, Marvin L., 857 Leadbetter, Mary G., 420 Mullin, William J.,508
Horwitz, William, 461 Lee, S. Mark, 691 Munns, Robert K., 720,1235
Hsu, Mei-Chich, 1143 Leoni, V., 1133 Muñoz De La Peña, Arsenio, 754
Huang, Weng R, 1143 Li, Maojuan, 1289 Nakamura, Yumiko, 1348
Huit, Karl, 842 Lin, Peter Y.T., 1396 Nakatani, Aya, 1374
Humber, John, 822 Lindfors, Erja, 575 Nakashima, Marvin J., 47, 50
Huntsman, Mark A., 1014 Lindsay, Chester W., 424 Nelson, Donald L., 1289
Hurlbut, Jeffrey A., 720 Lobinski, Ryszard, 1262 Nesheim, Stanley, 461
Hustead, David L., 694, 988 Long, Austin R., 720,1235 Newlon, Natalie, 1174,1182
Ibe, Akihiro, 32 Longbottom, James E., 1098,1113 Newsome, W. Harvey, 381,703,707,1225
Ichinoe, Masakatsu, 1006 Lopez-Avila, Viorica, 555, 864,1113,1329 Ng, Wing-Yan, 540
Iida, Mami, 32 Lorenzo, R.A., 608 Niemann, Richard A., 1362
Ikebuchi, Hideharu, 1006 Lott, Heidi M., 67, 663 Nishima, Taichiro, 32
Imerman, Paula Martin, 899 Louis, Judith B., 1121 Nixon, Gerald R., 26
Isaacs, Brandon, 910 Lovering, Edward G., 962 Nott, Reginald, 6 6 8
Ito, Yoshio, 1348 Low, Nicholas H., 342 Nyman, Patricia J., 1213
Izquierdo-Pulido, M.L., 1027 Luchtefeld, Ronald G., 857 O’Rangers, John, 442,443
Jackson,Jean, 543 Lynch, Joanna M., 1033, 1300 Ogawa, Masao, 83
Jain, Anant V., 948 MacDonald, Alexander, 320 Ohuri, Hiroshi, 1385
Jickells, Sue M., 760 Mageau, Richard P., 1022 Ohtsubo, Toshiro, 83
Jiusheng, Zhu, 6 6 8 Mahadevan, Subramaniam, 1010 Oles, Philip J., 615
Johnson, Robyn C., 907 Malone, William, 711 Olsen, Monica, 842
Johnston, Ralph W., 1022 Mantis, Antonios J., 1168 Orton, William L„ 857
Jones, Tammy L., 1329 Marine-Font, A., 1027 Oskarsson, Agneta, 842
Jorhem, Lars, 798 Markus, John R., 442, 443, 444, 447, 449, Packard, Vernal, S., 297
June, Geraldine A., 1240 451,459 Page, B. Denis, 26,765
Kamimura, Hisashi, 32 Martin, Alain, 632 Paisley, Steven D., 1069
Kaminski, Jerzy, 1010 Matheson, Man R., 1280 Palabay, Rodrigo B., 591
Kane, Peter, 1174 Matusik, Jean E., 420,1344 Palminger, Ira, 842
Katz, Stanley E„ 514,632,952 McKenzie, Arthur D., 14 Papageorgiou, Georgios, 1168
Kawamura, Norihisa, 436 McNeal, Timothy P„ 1213, 1268 Papthakis, Michael L., 691
Kendall, Donald C., 1146 McSheffrey, Sandra, 342 Paradis, Lance R., 851
Khunachak, Alisa, 1230 Meguro, Hiroshi, 1385 Parks, Owen W., 698
Kijima, Keiji, 292 Melis, Marinella, 92 Patil, Vitthal B„ 1394
Kim, Henry S., 414 Mendes, A. Silva, 1 Pelayo, Estela, 59, 591
Kim, Robert, 555 Merola, George V., 1057 Peña-Egido, M. Jesus, 565
King-Brink, Marcia, 787 Metzger, Dick T., 297 Perkins, Lewis B., 90, 851
King, Jerry W., 857, 8 8 8 Mikami, Eiichi, 436 Phillippo, Evan T., 907, 1309
Kirchhoefer, Ross D., 394 Milanes, June, 864 Philo, Mark R., 760
Kmostak, Svatomir, 735 Millar, Roderick G., 1317 Poucke, Lucien Van, 1138
Koch, Herbert, 976 Miller, Lisa S., 313 Prodolliet, Jacques, 268,275
Kojima, Shigeo, 292 Mishra, Rajendra Kumar, 604 Purcell, Steve, 620
Komla, K.L., 549 Mishvec, Philip B., 430 Pylypiw, Harry M., Jr, 1369
Krause, Richard T., 1146 Miwa, Harufumi, 1385 Rader, Bobby, 306
Krueger, Dana A., 418 Moats, William A , 535,549 Ramstad, Tore, 313

J o u r n a l O f A O A C I n tern a tio na l V o l . 76, N o. 6 ,1 9 9 3 203A

C u m u la t iv e A u t h o r I n d e x

Rao, R.R., 703 Sjöberg, Anna-Maija, 674 Tsumura,Yukari, 1348

Reeves IE, James B., 741 Skurikhin, Igor M., 257, 262 Tuinstra, Louis G.M.Th., 1248
Reggers, Guy, 1138 Smedley, Michael D., 725 Unruh, Joseph, 335
Renger, Bemd, 7 Smith, Erika B., 1033 Uthe, John R, 794
Riedel, Gunther, 26 Snell, Robert P.,531, 1127 van Schalkwyk, Dirk J., 361
Robison, Barbara J., 831 Snyder, Janet M.. 8 8 8 van Dijk, Remmelt, 570
Rodríguez-Otero, J.L., 838 Soliman, Abdel-Gawad M., 390 van Trijp, John M.P., 1248
Roinestad, Kurt S.,, 1121 Sovocool, G. Wayne, 1092 Verdier, Pam, 14
Rosas, Miguel López, 754 Spanedda, Lorenzo, 92 Vidal-Carou, M.C., 1027
Rosen, Joseph D., 1121 Sphon, James A.. 1344 Viggers, Elizabeth A., 620
Rowe, Loyd D., 8 8 8 Sporns, Peter, 1289 Vitali, M„ 1133
Roy, Ronald R., 492 Stahr, Henry M., 893 Waldron, Ellen M., 1344
Roybal, José E., 720 Stapleton, Nancy K., 907 Walser, Paul E„ 851
Rubí, E„ 608 Stout, Steven J., 1230 Weaver, Carol M., 682
Rupp, Heidi S., 1235 Suekane, Sachiko, 1374 Weber, Dorcas F, 377
Sack, Chris A., 1146 Sumitani, Hidenobu, 1374 Weber, John D., 725
Salinas, Francisco, 754, 1255 Sun, Thomas S.C., 893 Webster, Gregory K., 1400
Salisbury, Craig D.C., 1149 Sundaram, Kanth Maran Somu, 6 6 8 Weiss, George, 320
Salminen, John, 26 Suzuki, Tateo, 1385 Wesselman, Raymond J., 1098
Salvatore, Michael J., 514,952 Sved, Stephen, 325 Wheeler, Mark M., 941
Sanyal, Asis K., 1152 Sydenham, Eric W., 361 White, Jeny D., 907, 1309
Sapp, Robert E., 956 Szpunar-Lobinska, Joanna, 1262 White, C.R., 549
Sasaki, Kumiko, 292 Tabata, Setsuko, 32 Wicker, Alan L„ 945
Sauvé, François, 1163 Takeda, Mitsuham, 292 Williams, Kay A., 907
Sawada, Jun-ichi, 1006 Tallmadge, Daniel H., 1396 Wirtanen, Gun, 674
Scher, Alan L., 287 Tamura, Yukihiro, 32 Woemer, Julie A., 8 8 8
Schmitt, Thomas M., 387 Tan, Yew Ai, 371 Wolf, Wayne R., 682
Schwab, Bernard, 1022 Tanaka, Touichi, 1006 Wolynetz, Mark S., 508
Schwartz, Daniel R, 335 Tang, Peter H., 72 Wong, Siu-Kay, 540
Schwiesow, Mark, 918 Tanner, James T., 399, 1042 Wood, Roger, 926
Scollary, Geoffrey R. 1389 Tatsuka, Kiyoaki, 1374 Woods, Robert W„ 907
Sebranek, Joseph G., 787 Taylor, Scott L., 857 Worobey, Brian L., 881
Semeraro, Isabella, 657 Teissedre, Pierre-Louis, 1262 Worthy, Brenda E., 682
Serti, David, 711 Terao, Chikahiro 436 Wudrich, Gerald G., 342
Sewell, Anne M., 814 Terao, Tadao, 1006 Yamada, Sadaji, 436
Shaikh, Badar, 543 Teshima, Reiko, 1006 Yen, Ivan Chang, 366
Shantha, Nalur Chandrasekaran, 644 Thiel, Pieter G., 361 Yess, Norma J., 36, 492
Shephard, Gordon S., 361 Thomas, Virginia N., 313 Yeung, Jupiter M., 381,1225
Sherma, Joseph, 444,447,449, 451, 459 Thompson, Joseph J., 1378 Young, Richard, 555
Sherrod, Patricia S., 1240 Thompson Raymond H., 1057 Young, Barbara E.S., 851
Shingare, Murlidhar S., 1394 Thompson, Michael, 926 Yperman, Jan, 1138
Shoemaker, Dirk, 1156 Titus, Richard L.. 1092 Zini, Guido, 650
Siegmund, Emil G., 306 Tonogai, Yasuhide, 1348 Zygmunt, Lucian C., 1069
Silvestre, M.P.C., 1295 Trotter, William J., 1220
Sims, Art, 1156 Tsuda, Shigenori. 83
Singh, Raj P, 1187 Tsuji, Kozo, 83

204A J o u r n a l O f A O A C In ternational V o l . 76, N o. 6 ,1 9 9 3

 C a m pbell & S a u v é J o u r n a l O f A O A C I ntern a tio na l V o l . 76, N o. 6 ,1 9 9 3 1163

A G R IC U L T U R A L M A T E R IA L S

L iq u id C h r o m a t o g r a p h ic D e t e r m in a t io n o f M e le n g e s t r o l A c e t a t e

in F e e d s

H M. C a m p b e l l and F r a n ç o is S a u v é
arold

Agriculture Canada, Laboratory Services Division, Food Production and Inspection Branch, Ottawa, ON, K1A0C6, Canada

A m eth o d for m elen g estro l a c e ta te (MGA) w a s d e­ phy (LC). Roybal (3) investigated the use of separation and
v elo p ed to ac cu ra te ly d eterm in e 0.08-220 ppm cleanup steps from the GC procedure (1) combined with LC
. MGA in anim al fe ed s. MGA is e x tra c te d from an determination. Examination of this method has shown low re­
a q u e o u s slu rry of fe e d sa m p le w ith h ex a n e a n d p ar­ coveries for some feeds and the frequent occurrence of signifi­
titio n ed from h ex a n e into a q u e o u s m ethanol an d cant interferences. Weigand and Dille (4) reported a simplified
th e n into d ich lo ro m eth an e. After ev a p o ratio n of th e method using preparative LC and quantitation by reversed-
d ich lo ro m eth an e, th e dried e x tra c t is tran sferre d phase LC. The methods of Roybal and of Weigand and Dille
with chloroform into a volum etric flask. An aliquot use Soxhlet extraction with hexane; however, we found this
is p ip etted into a m ixed colum n of silica gel an d procedure to give incomplete extraction for some types of
acid alum inum oxide, a n d th e MGA is elu ted with feeds (Table 1).
h e x a n e -a c e to n e . T he re sid u e is d isso lv e d in m eth a­ The method presented here incorporates many of the con­
nol, an d th e MGA is q u an titated by re v e rse d -p h a se cepts of the GC procedure, i.e., liquid-liquid extraction, parti­
liquid c h ro m a to g rap h y (LC). T he m eth o d w a s vali­ tioning, and cleanup by a modification of Roybal’s mixed-
d a te d by u sin g com m ercial fe e d s a n d laboratory- phase column. An option is presented for a simplified
fortified fe e d s. T he m eth o d w a s a lso applied s u c ­ liquid-liquid extraction that avoids the necessity of specialized
cessfu lly to co rn s ila g e s su p p le m e n te d w ith MGA. glassware and reduces the volume of solvents required. Com­
R eco v eries [and a s s o c ia te d coefficients of vari­ pared to the GC method with either extraction, this LC method
ation (CVs)] for fe e d s fortified in th e laboratory with provides improved recoveries, lower coefficients of variation
MGA 100 a t 0.2, 0.8, a n d 5 ppm w ere 96.9 (6.28), 102 (CVs), a wider range of application, cleaner chromatograms,
(3.59), a n d 109% (2.06%), respectively. CVs for a v a­ and elimination of the need for internal standards because re­
riety of co m m ercial fe e d s varied from 0.88 to coveries are quantitative.
3.84%. An interlaboratory s tu d y w a s a lso c o n ­ A broad spectrum of samples was analyzed, including com
d u c te d to co m p a re th e LC m eth o d to th e AOAC g a s silage supplemented with MGA. The silage samples generate
ch ro m a to g rap h ic m ethod. extraneous peaks, and MGA is adequately resolved. Chroma­
tograms of other sample types generally show only a major
matrix peak after the solvent front and a well-resolved MGA
M elengestrol acetate (MGA) is a progestational steroid peak.
added to heifer feeds as an aid in suppression c f estrus
to increase weight gains and improve feed efficien­ METHOD
cies. The AOAC method (1) uses gas chromatography (GC)
with electron capture detection for MGA measurement. An in­ Reagents
ternal standard is necessary to compensate for losses in the ex­ (a) Solvents.—Distilled-in-glass hexane, methanol, dichlo­
traction and cleanup steps, and another internal standard is romethane, chloroform, acetone; LC grade methanol, dioxane,
added before GC analysis to compensate for variability in the water (Milli-Q, or equivalent); and absolute ethanol.
chromatography and detection. The collaborative study (2) (b) Silica gel 60.— For column chromatography, Kieselgel
showed other deficiencies: recoveries were low (overall recov­ 60,70-230 mesh ASTM (Merck & Co., Inc.).
ery 83.2%), within-laboratory repeatability was significant (c) Deactivated silica gel.—Combine 45.0 g silica gel and
(±9 .5 %), and the method was designed for the limited range of 5.00 g distilled water in a tightly sealed container. Mix well to
0.28-2.2 ppm. Recoveries below 0.28 ppm were low (72.5% remove any lumps and then mix occasionally for 2 h before
for a 0.14 ppm sample). use. Discard after 2 weeks.
Other published methods attempted to improve and sim­ (d) Aluminum oxide.—Woelm acid, anionotropic, activity
plify the method of analysis or tried to use liquid chromatogra- grade 1 for column chromatography (InterScience, Inc.).
(e) Deactivated alumina.—Combine 29.1 g acid aluminum
Received October 5, 1992. Accepted January 25, 1993. oxide and 0.90 g distilled water in a tightly sealed container.
1164 C a m pbell & S a uv é : J o u rn a l O f A O A C I n tern a tio na l V o l . 76, N o. 6 ,1 9 9 3

T a b le 1. C o m p a r is o n o f S o x h le t e x tra c tio n a n d A O A C (f) Chromatographic column.—Glass, 27 x 1.0 cm id with

liq u id - liq u id e x tra c tio n 100 mL reservoir and Teflon stopcock.
M G A analysis, ppm (g) LC system.—Instrument: capable of maintaining con­
stant pulseless flow and with autosampler or loop injector for
AOAC Soxhlet extraction
MGA liquid-liquid
50 pL injections. Detector: wavelength, 290 nm; able to pro­
Sam ple guarantee, extraction, Extraction vide minimum sensitivity of 0.01 AUFS. System must also
description ppm 3 h Result tim e, h consist of an integrator, recorder, or data station. Spherisorb
ODS2 column, 3 pm thickness, 10 x 0.45 cm id, or equivalent
Beef sup- 0.8 0.636 0.415 6 equipped with a Clg guard column. (In this study, we used a
plement for 0.605 16
0.6 27 40
Varian 5560 LC equipped with a 8085 autosampler and
heifers, 40%
Rheodyne 7126 injector, UV 200 detector, and Vista 402 data
S w eet cattle 0.8 0.505 0.3 56 40 system.)
ration, 32%
Extraction and Solvent Partitioning
Beef sup- 3.5 2.35 2.41 3
plement, 2.44 6 Either the A or the B procedure can be used for extraction
10% 2.45 22
and purification. Procedure A does not require purchasing the
Beef sup- liquid-liquid extractors and accessories, uses 40% less solvent
plement, 1.76 1.05 1.15 3 and is performed more quickly.
32% 1.17 6 A. Modified AOAC extraction method.—Weigh 10 g finely
ground sample into 500 mL flat-bottom boiling flask. Add 10
mL absolute ethanol and 30 mL distilled water and mix. Add
250 mL hexane. Connect to a reflux condenser, and reflux with
Mix well to remove lumps and then shake occasionally for 2 h stirring, using a heater/magnetic stirrer, for 1 h. If sample does
before use. Discard after 2 weeks. not stir well, <30 mL additional water may be added.
(f) Sodium sulfate.—Anhydrous, reagent grade. Let cool, and decant the hexane layer into another 500 mL
(g) Partitioning solvent.—Methanol-0.25% aqueous so­ evaporating flask. (If sample matrix does not settle completely,
dium sulfate (70 + 30). Dissolve 1.50 g anhydrous sodium sul­ stop decanting before solids are transferred, add 50 mL hexane,
fate in 600 mL distilled water. After complete dissolution, add mix, let solids settle, and decant again.) Add 250 mL hexane to
1400 mL methanol (distilled-in-glass). Mix well. the extraction flask, and reflux, with stirring, for 1 h. Evaporate
(h) Hexane-acetone.—(7) (95 +5). Pipet 50 mL acetone first hexane portion to dryness on rotary evaporator during re­
into a 1 L volumetric flask, and dilute to volume with hexane. fluxing, and dissolve residue in ca 10 mL hexane. Let second
Mix well. (2) Hexane-acetone (4 + 1 , v/v). extract cool and then decant hexane layer into 500 mL evapo­
(i) LC mobile phase.—Combine 350 mL methanol, 350 rating flask. Evaporate combined extracts to dryness on rotary
mL dioxane, and 300 mL water (all solvents LC grade). Mix evaporator. Rinse extraction flask twice with 50 mL hexane,
well and filter through a 0.45 or 0.2 pm filter. and decant hexane layer into 500 mL evaporating flask. Trans­
fer hexane extract in 500 mL evaporating flask to 250 mL sepa­
(j) Melengestrol acetate.—Reference standard, available
ratory funnel. (Filter through glass wool if necessary.) Rinse
from the Upjohn Co., Kalamazoo, MI 49001, and TUCO Prod­
500 mL evaporating flask with 50 mL partitioning solvent, and
ucts Co., Animal Health Division, Orangeville, ON, L9W 3T3,
transfer into same 250 mL separatory funnel. Gently shake fun­
Canada. (1) MGA stock solution, 100 \ig/mL—Accurately
nel by inverting 15 times, let solvent layers separate com­
weigh 10 mg MGA reference standard into a 100 mL volumet­
pletely, and drain lower layer into 500 mL separatory funnel
ric flask, dissolve in methanol by using an ultrasonic bath if
containing 100 mL CH2 Cl2. If precipitate or emulsion forms at
necessary, and dilute to volume with methanol. (2) MGA LC
interface, retain this with upper phase. (If lower layer is cloudy,
standard solutions; 0.5, 1, 2 , 5, and 10 \lg/mL.—Dilute MGA
filter through No. 4 filter paper into the separatory funnel.)
stock standard with LC mobile phase to obtain the respective
concentrations: 1.0 mL to 200 mL, and 1.0, 2.0, 5.0, and Repeat partitioning steps 3 more times, rinsing 500 mL
10.0 mL to 100 mL. evaporating flask with each 50 mL volume of partitioning sol
vent. If emulsion forms during partitioning, let it pass into
Apparatus CH2 C12 in fourth transfer. Vigorously shake combined extracts
in 500 mL separatory funnel 15 s, let phases separate com­
(a) Liquid-liquid extractor.—Ace Glass No. 6840-96, or pletely, and drain lower (CH2 C12) layer into another 500 mL
equivalent. round-bottom flask. Repeat partitioning 3 more times, using 30
mL CH2 C12 each time. Roto-evaporate combined CH2 C12 ex­
(b) Heating mantle.—For 500 mL flask, 250 watt (Glas-
tracts to near dryness. Remove residual HzO with 2 separate
Col Apparatus Co.).
additions of 20 mL absolute ethanol, and roto-evaporate to dry­
(c) Rotary evaporator.
ness after each addition of ethanol. Quantitatively transfer resi­
(d) Nitrogen evaporator. due to 10 mL volumetric flask with CHC13 and dilute to volume
(e) Heater/magnetic stirrer.—4-unit or equivalent. with CHCI3 . Proceed to chromatography cleanup.
 C a m pb el l & S a u v é J o u r n a l O f A O A C In tern a tio na l V o l . 76, N o. 6 ,1 9 9 3 1165

B. AOAC Extraction Method.— (As described in reference LC Determination

1, sections D and E, with omission of extraction standard.)
Weigh 10 g finely ground sample into extractor. Add 10 ml. Set following parameters and let system equilibrate: mobile
absolute ethanol and 30 mL distilled water and mix. Fill to near phase, MeOH-dioxane-H20 (35 + 35+30); flow rate, 1.0
the side arm with hexane, insert solvent return tube into posi­ mL/min; chart speed, 1.0 cm/min; detector wavelength, 290
tion, and start magnetic stirrer. Attach 500 mL receiving flask nm; and detector absorbance range, 0.01 AUFS.
containing 100 mL hexane and several glass beads to extractor. Inject 50 pL (full loop) of MGA LC standard solution, 1
Fill extractor to side arm with hexane and attach to condenser. pg/mL. Adjust amount of water in mobile phase, if necessary,
Extract sample 3 h by heating receiving flask enough to have to obtain K of ca 1.6-2.0 [K = (t1 —t0)lt0, where t0 = distance
rapid reflux (>20 mL/min) at condenser while stirring sample from injection to first perturbation of standard chromatogram,
vigorously and continuously. Control emulsions by regulating and t1 is the same distance for MGA],
stirring rate. Up to 30 mL additional distilled water may be Make 2 or more injections of 1 pg/mL standard to ensure <
added through condenser to aid stirring action, if necessary. Let 2% repeatability of peak area responses. Verify linearity of in­
apparatus cool, set receiving flask aside, and decant hexane strument by injecting different concentrations (0.5-10 pg/mL).
layer from extractor into a 2 L round-bottom flask. Rinse ex­ Filter sample solutions through 0.45 pm filter. Inject 50 pL of
tractor 2 or 3 times with ca 10 mL hexane, and decant hexane each. Bracket each 2 sample injections by injection of 1 or 2
into the 2 L round-bottom flask. pg/mL standard. If sample peak area is more than double that
Roto-evaporate contents of the 2 L round-bottom flask to for standard, make appropriate dilutions. At end of analysis,
dryness. Quantitatively transfer extract in 500 mL receiving pump MeOH through system for 30 min to remove any ad­
sorbed material.
flask into 500 mL separatory funnel. Add 100 mL partitioning
solvent into the 2 L round-bottom flask and swirl. Transfer ex­
tract in the 2 L round-bottom flask into the corresponding 500 C alculations
mL receiving flask, swirl, and then quantitatively transfer into
same separatory funnel. Gently shake separatory funnel by in­ Calculate MGA content of samples by using average re­
verting 15 times, let solvent layers separate completely, and sponse of standards bracketing each sample.
drain lower layer into 1L separatory funnel containing 200 mL
PRy 10 mL
CH2 C12. Repeat partitioning step 3 times, and rinse the 2 L ppm MGA = x Cx
pr2 2 mL
round-bottom flask and the 500 mL receiving flask with each
100 mL portion of partitioning solvent. Vigorously shake com­
bined extracts in 1 L separatory funnel 15 s, let phase separate where PR] - sample peak response; PR2 = average standard
completely, and drain lower (CH2 C12) layer into another 500 response (standards bracketing the sample); C = standard con­
mL flask. Repeat partitioning with CH 2 C12 twice, adding 40 centration, pg/mL; V - final volume of sample extract, mL; and
mL CH2 C12 each time. Roto-evaporate combined extracts to W = sample weight, g.
near dryness. Remove residual H20 with 2 separate additions
of 20 mL absolute ethanol, and roto-evaporate to dryness after R e su lts a n d D iscu ssio n
each addition of ethanol. Quantitatively transfer residue to 10
mL volumetric flask with CHC13, and dilute to volume with In establishing the LC conditions, a variety of LC columns
CHC13. Proceed to chromatography cleanup. and mobile phases were evaluated. The 3 pm C ] 8 column iden-

Chromatography Cleanup

(Column recovery should be 98-102%. Do not let columns

ran dry.)
Place glass wool plug at bottom of column and fill column
with hexane. Slowly add 3.4 g deactivated silica gel. Open
stopcock and gently tap column to pack material. Add 2.2 g
deactivated alumina oxide. Add 1 g anhydrous Na2 S 0 4 and
drain hexane to top of Na2 S 0 4 layer. Pipet 2 mL sample into
column and let drain to top of Na2 S 0 4 layer. (Place the pipet tip
below reservoir before draining.) Wash column with total of 70
mL hexane-acetone (95 + 5), using 2 or 3 ca 2 mL portions
(Pasteur pipet) to rinse column walls. Let solvent drain to
Na2 S 0 4 surface between additions of solvent. Discard eluate.
Elute MGA with 20 mL hexane-acetone (4 + 1) into centrifuge
tube. (Volume of elution solvent required should be verified.) F ig u r e 1. E x a m p le c h ro m a to g ra m s : (A ) M G A s ta n d a rd ,
Evaporate to dryness, using N-evaporator, and dissolve residue 20.2 n g; (B ) 40% b e e f s u p p le m e n t, 0 .7 8 0 p p m M G A ,
in appropriate volume of mobile phase to give 1-2 pg/mL fo rtifie d in th e la b o ra to ry w ith M G A 100; a n d (C ) c o m ­
(minimum volume, 1.0 mL). m e rc ia l fe e d , 62% b e e f s u p p le m e n t, 0 .5 9 9 p p m M G A .
1166 C a m pbell & S a u v é : J o u rn a l O f A O AC In ternational V o l . 76, N o. 6 ,1 9 9 3

Table 2. Comparison of the AOAC and modified extraction procedures for the analysis of commercial feeds
Modified A O A C liquid-liquid extraction A O A C liquid-liquid extraction

G uarantee, ppm X, ppm n CV, % X, ppm n CV, %

0.4 4 0.376 3 1.81 0.3 74 3 2 .5 4

0 .8 0 0.636 3 1.50 0.6 42 3 1.22
Unknown 0.665 3 0.3 5 0.6 55 3 2.3 7

0.4 0 0.313 3 0.49 0.300 3 1.07

0.88 0.7 34 3 1.74 0.726 3 2.1 6
2 2 1 .6 (M G A 100) 194 8 3.8 4 198 3 3.2 9
3.5 2.35 3 0.88 2.4 5 1 —
1.76 1.05 3 0.9 5 1.04 1 —

tifiedinthemethod issuperiortoother5 pm and 10 pm col­ volumes ofsolvents (2).The modified AOAC extractionisa
umns evaluated. The use ofdioxane inthe mobile phase sig­ good alternative.As demonstratedinTable2,MGA valuesob­
nificantlyaltersthecolumn selectivity.The selectedLC condi­ tainedbyeitherliquid-liquidextractionareequivalent;CVs for
tionsprovide a rapid elutiontime (approximately 3 min), and themodifiedAOAC extractionareslightlylower.Table3 pre­
witha 10min run time,subsequentchromatograms arefreeof sentsrecoverydataby themodifiedAOAC extractionforsam­
ghostpeaks. ples prepared inthe laboratory (fortifiedby standard addition
Figure 1 shows a standard chromatogram and chromato­ or formulated by using MGA 100). This table shows the
grams offeedextractspreparedby using the modified extrac­ method tobe effectiveforbothmineralandorganictypefeeds.
tion.The LC responsewas shown tobe linearover arange of The 0.8 ppm sample isused asthequalitycontrolsample for
5-200 ng (higherlevelswere notchecked). routineanalyses.
Attempts to simplify the method by using Soxhlet extrac­ A criticalstepinthemethod istheuseofCHC13todissolve
tion, solid-phase extractions, and gel permeation for sample theresidueafterroto-evaporationofCH2C12andbeforethecol­
cleanupwereunsuccessful. Soxhletextractionwithhexane,as umn chromatography step;hexane, asspecifiedintheAOAC
publishedby Roybal (3)and Weigand and Dille (4),isnotal­ method, is not always effective. Because of the significant
ways adequate. As shown inTable 1,and asreportedby Wei­ amount of coextractives occurring with the liquid-liquid ex­
gand and Dille (4)(recoveryproblems with mineral samples), traction,alumina,asused inreferences2 and 3,inthecleanup
Soxhletextractionisincompleteforsome feedtypes.The use column gaveextractsthatoftengeneratedchromatograms with
of liquid-liquidextraction (presence of water isessential for unacceptable interferences.The mixed-phase cleanup column
quantitative recoveries) affects the method and the extent of largelyeliminatesthisproblem.The columncleanupisnotnec­
extractpurificationrequiredbeforeLC analysis.The combina­ essaryforhigh-MGA-content samples, such asMGA 100 and
tionofpartitioningand column chromatography was found to mineral samples atapproximately 5 ppm orhigher. The stock
be essential and resulted in very clean extracts and problem- standard solution is stable for 1 year ifrefrigerated. Diluted
freechromatograms. standard solutions should be refrigerated and prepared fresh
Liquid-liquidextractionofM G A givesconsistentandquan­ monthly.Sample solutionsarestablefor1week ifrefrigerated.
titativeextractionbutrequiresspecializedglassware and large The throughput of samples isimproved over the AOAC GC

Table 3. Recoveries for modified AOAC liquid-liquid extraction

M G A found

Sam ple ppm X, ppm n CV, % Rec., %

Check sample
5 ppm a 4.85 5.28 3 2.0 6 109
0.8 ppm a 0.7 80 0.797 15 3.59 102
0.2 ppm a 0.1 94 0.1 88 3 6.28 96.9
Mineral, 4 ppm 6 3.96 3.7 8 3 2.68 95.4
0.5 ppm spikec 0.505 0.482 3 1.86 95.4

S am ples form ulated in laboratory using 194.4 ppm, M G A 100, added to a beef supplement.
b Mineral feed sample formulated In laboratory using M G A 100.
c S am ple recovery using standard addition to a beef supplement.
 C ampbell & S au v é J ournal O f AOAC International V ol . 76, No. 6,1993 1167

T a b le 4. C o m p a r is o n o f L C a n d A O A C G C m e th o d s lin, piperazine, poloxalene, pyrantel tartrate, ractopamine,

M G A found, ppm
robenidine, salinomycin, streptomycin, sulfamethazine, tia-
Labeled M G A mulin, tylosin, virginiamycin, and zoalene.
Sam ple content, ppm GCa LC*
Table 4 provides a comparison of results obtained by this
method and the AOAC GC method (1). Samples and GC re­
Corn silage 0.035 0 .0 4 3 c ND
sults were provided by the Upjohn Co. Although a validated
Corn silage 0.035 0 .0 4 6 c 0 .0 280
study protocol was not used (single analysis only by Upjohn for
Corn silage 0.035 0 .0 3 7 5 c 0.0 372
feed samples; sample storage and sampling procedure not
Feed 0.88 0.8 69 0.7 52
specified), the MGA results compare favorably. On the basis of
Feed 1.76 1.62 1.55
0.4 12
these data and the recovery study and repeatability data, this
Feed 0,595 0 .5 6 9
Feed 0.586 0.5 86 0.5 04
LC method provides a good alternative to the official AOAC
Feed 0.88 0.9 24 0.9 90
GC method. Using the modified AOAC liquid-liquid extrac­
Feed 0.88 0.772 0.7 56 tion, the LC method gives significant savings in time, equip­
Feed 1.1 0.9 59 0.9 08 ment, and chemicals.

a Sam ples and G C results were provided by the Upjohn Co.,

A cknow ledgm ents
Kalam azoo, M l. Results are single analysis except for silage
samples.
b Average results of duplicate analysis. We express appreciation to Domenic Brescacin and Luc
c Silage samples were analyzed by an unofficial method in
Provost for providing some of the LC data; to Trade Wolthuis,
duplicate. N D = not detected.
Sharon Barney, and Cindy Biljum of the Upjohn Co., Kalama­
zoo, MI, for providing samples and results by the AOAC GC
method ( 1 ) but is only about half the 1 2 samples per day cited method; and to E.J. Moore of this laboratory for his support and
by Weigand and Dille (4). The latter method, however, uses encouragement.
Soxhlet extraction, which is not always effective.
The following dmgs and antibiotics were checked and R eferen ces
found not to interfere in the LC analysis: amprolium, acetyl-(p-
nitrophenyl) sulfanilamide, arsanilic acid, bacitracin, car- (1 ) O fficia l M e th o d s o f A n a ly sis ( 1 9 9 0 ) 1 5 t h E d . , A O A C , A r ­
badox, chlortetracycline, clopidol, decoquinate, dibutyltin di- lin g to n , V A , s e c . 980.36
laurate, diclazuril, dimetridazole, 3,5-dinitrobenzamide, (2 ) D a v i s , R . A . , K r a t z e r , D . D . , & G e n g , S . ( 1 9 8 0 ) J. A sso c. Off.

ethopabate, flavomycin, furazolidone, lasalocid, lincomycin, A nal. Chem . 6 3 , 4 2 5 ^ 4 4 3

maduramicin, morantel tartrate, monensin, narasin, neomycin, (3 ) R o y b a l , J . E . ( 1 9 8 1 ) J. A sso c. Off. A nal. Chem . 6 4 , 6 6 1 - 6 6 4

nequinate, nicarbazin, nifursol, 3-nitro-4-hydroxyphenylar- (4 ) W e i g a n d , J . L . , & D i l l e , D . S . ( 1 9 8 8 ) J. A ssoc. Off. A nal.

sonic acid, 4-nitrophenylarsonic acid, oxytetracycline, penicil­ Chem . 71, 7 0 7 - 7 0 9
1168 F letouris E t Al .: Journal O f AOAC International V ol . 76, No. 6,1993

AGRICULTURAL MATERIALS

R a p id D e te rm in a tio n o f T ry p to p h a n in I n ta c t P ro te in s b y

D e riv a tiv e S p e c tro p h o to m e try

D e m it r io s J. F l e t o u r is

Aristotle University, School of Veterinary Medicine, Laboratory of Milk and Milk Products, 54006 Thessaloniki, Greece
N ic k o s A . B otsoglou
Aristotle University, School of Veterinary Medicine, Department of Animal Production, Ichthyology, Ecology, and
Protection of Environment, 54006 Thessaloniki, Greece
G e o r g io s E. P a p a g e o r g io u

Aristotle University, School of Medicine, Laboratory of Biological Chemistry, 54006 Thessaloniki, Greece
A n t o n io s J. M a n t is

Aristotle University, School of Veterinary Medicine, Laboratory of Milk and Milk Products, 54006 Thessaloniki, Greece

A m eth o d for rap id determ in atio n of total try p to ­ sis of the majority of the amino acids whereas alkaline hydroly­
p h an in in tact p ro tein s w a s d ev elo p ed . S am p le is sis is recommended for tryptophan assay because this amino
h o m o g en ized with 0.1 M so d iu m hydroxide, a n d th e acid is extensively degraded during acid hydrolysis and must
h o m o g en ate , if it c o n ta in s s u s p e n d e d m aterial, is be analyzed separately ( 1 ).
cen trifu g ed . T ryptophan is directly q uantified in The determination of tryptophan, however, is still a problem
sam p le e x tra ct o n th e b a s is of its fourth-derivative (2 ); losses of tryptophan can occur even under alkaline hy­
UV ab so rp tio n sp e c tru m . P rotein h y d ro ly sis an d /o r drolysis (3). Lucas and Sotelo (4), studying the effect of differ­
e x tra ct purification is not req u ired b e c a u s e th e ent alkalis, temperatures, and hydrolysis times on the determi­
fourth-derivative tran sfo rm atio n of th e c o n v e n ­ nation of tryptophan in pure proteins and foods by EEC,
tional analytical b a n d aro u n d 285.5 nm virtually reported that lithium hydroxide at 145°C gives the best results
elim in ates in terfe ren c es arisin g from both ty ro sin e provided that hydrolysis proceeds for 4 and 8 h, respectively,
a n d o th e r U V -absorbing c o m p o n e n ts. W hen p u re whereas Nielsen and Hurrell (5) found no advantage in using
p ro tein s w ere an aly zed by th e m ethod, th e v alu e s lithium rather than sodium hydroxide and recommended hy­
of try p to p h an re s id u e s found co in cid e d well with lit­ drolysis for 20 h at 110°C for all proteins. Other workers (6 ,7)
e ra tu re d a ta b a s e d on s e q u e n c e an a ly sis. T he a p ­ have proposed barium hydroxide hydrolysis at 125°C for 16 h
plicability of th e m eth o d w a s a ls o te s te d on sev eral in absence of air for the determination of tryptophan in proteins
fo o d a n d feed in g red ient s a m p le s se le c te d to in­ and feedstuffs by IEC or LC, whereas Sato et al. (8 ) have rec­
c lu d e a ra n g e of p rotein c o n te n t a n d a variety of ommended sodium instead of barium hydroxide to improve
p ro tein s o u rc e s . Owing to its sim plicity a n d reliabil­ tryptophan recovery. Jones et al. (9) have also suggested so­
ity, th e m eth o d is particularly re co m m en d ed for dium hydroxide with maltodextrin to prevent degradation of
ev ery d ay a n a ly sis of a large n u m b er of sa m p le s. tryptophan during hydrolysis, whereas Allred and MacDonald
( 1 0 ), who performed a collaborative study, have reported that
sodium hydroxide hydrolysis at 110°C for 20 h under vacuum
A mino acid analysis is of major importance in the assess­ is suitable for the determination of tryptophan in foods and
ment of protein quality, because the nutritive value of food and feed ingredients by IEC or LC.
proteinaceous products depends on their amino acid Because analysis of tryptophan in proteinaceous materials
composition. Amino acid analysis of food proteins is conven­ by hydrolysis is generally a destructive process, alternative
tionally carried out by ion-exchange (IEC) or liquid chroma­
spectrophotometric methods based on its prominent UV ab­
tography (LC) after hydrolysis into the constituent amino ac­
sorption have been developed (11-13). However, interferences
ids. The hydrolysis conditions are critical, because they
due to inherent components, impurities, or turbidity can affect
represent a compromise aimed at yielding the best estimate of
the analysis (14). To eliminate or reduce interferences, other
amino acid composition. Differences in amino acid stabilities,
workers (15-18) have proposed methods for direct determina­
variations in ease of peptide bond cleavage, and matrix effects
tion of tryptophan in pure proteins by second-derivative spec­
from nonproteinaceous components all prohibit a single set of
trophotometry. With these methods, the complex UV absorp­
hydrolysis conditions. Thus, acid hydrolysis is used for analy-
tion spectrum of protein at wavelengths above 275 nm could
be partly resolved, and quantitation of tryptophan could be per­
Received September 30, 1992. Accepted Febuary 18, 1993.
formed at neutral pH; however, serious interferences, mainly
 F letouris E t A l .: J ournal Of AOAC International V ol . 76, No. 6,1993 1169

from tyrosine, necessitate application of appropriate correction Determination

factors that complicate the analysis.
The method described in this paper is a new approach to the Run fourth-derivative UV spectra of both extracts and
analysis of tryptophan in unhydrolyzed proteinaceous samples. working solutions against solvent according to described con­
Band discrimination and measurement of tryptophan in the in­ ditions. Record depth of trough at 285.5 nm, as printed on in­
tact protein is successfully carried out at pH 13, where the ab­ strumental chart in arbitrary units, and plot against concentra­
sorbance contribution from tyrosine is almost negligible, tion of working solutions (pg/mL or pM tryptophan). Calculate
through use of fourth-derivative spectrophotometry. slope, intercept, and least-squares fit of standard curve. Use
standard curve slope and intercept to compute concentration of
E xperim ental tryptophan in sample extract. Determine tryptophan content of
samples or number of tryptophan residues in pure proteins by
the following formulas:
Apparatus
C] x V„ v
(a) Spectrophotometer.— Shimadzu Model UV-160A (Ja­ Tryptophan content, % = (- ) x (—)
pan) with 10 mm quartz absorption cells; scan speed, 480 10 Vv v
nm/min; derivative wavelength difference (AX), 3 nm.
(b) Vortex mixer.— Heidolf Reax 2 (Germany).
C2 x MW x V0 V
(c) Laboratory mill.—Retsch (Germany) with 0.5 mm Number of tryptophan residues = (------------- -— ) x (—)
screen. v Wx 106 V
(d) Blender.—Sorvall Omni-Mixer 17220 (Dupont Instru­
ments Products, Newtown, CT).
where Cl and C2 are tryptophan concentrations determined in
(e) Centrifuge.—BHG 1100 (Germany).
pg/mL and pM, respectively, according to calibration curves;
Reagents and Materials V0 is volume, mL, of sodium hydroxide solution added for ex­
traction; W is the weight of sample, mg; v is the volume of
(a) Extracting solvent.—0.1M sodium hydroxide solution. aliquot (mL) taken; V is the volume (mL) of aliquot dilution;
(b) Amino acid standards.—V-Acetyl-L-tryptophanamide, MW is the molecular weight of the protein.
V-acetyl-L-tyrosinamide, /V-acetyl-L-phenylalanine ethyl ester,
and L-cystine (Sigma Chemical Co., St. Louis, MO, USA.) R e su lts a n d D iscu ssio n
(c) N-Aceytl-L-tryptophanamide standard solutions.—
Weigh accurately ca 15 mg V-acetyl-L-tryptophanamide and The performance of second-derivative spectrophotometry
dissolve in and dilute to 25 mL with 0.1M sodium hydroxide. in the determination of tryptophan in intact proteins at neutral
Further dilute aliquots of this stock solution to prepare working pH has been long studied (15-18). The absorbance differentia­
solutions ranging from 2.5-20 |lg tryptophan/mL that corre­ tion d2A/dX2, where A is the absorbance and X is the wave­
spond to 12.2-97.9 (iM. Refrigerate standard solutions when length, transforms the normal spectrum into a series of sharp
not in use. Prepare fresh solutions weekly. peaks and troughs that are used for analytical purposes.
(d) Protein samples.—Gluten (wheat) and gliadin (wheat) In the spectral region above 275 nm, the second-derivative
obtained from BDH (Poole, England), casein (cow milk), UV spectrum of an aqueous buffer solution of V-acetyl-L-tryp-
lysozyme (chicken egg white), ovalbumin (chicken egg), bo­ tophanamide, pH 6.7, exhibits 2 troughs at 281.5 and 290 nm
vine albumin, and histone (calf thymus) purchased from and 2 peaks at 285.5 and 295 nm (Figure 1). The distance be­
Sigma, cytochrome C (horse heart) obtained from Boehringer tween the trough at 290 nm and the peak at 295 nm or the
Mannheim (Germany), colagenase purchased from Serva, baseline is strictly related to tryptophan content, and this rela­
Feinbiochemica (Heidelberg, Germany). tion constitutes the basis of several methods for quantitation of
(e) Food and feed ingredient samples.—Com, barley, tryptophan in pure proteins (15-18). A major drawback, how­
wheat, skimmed milk powder, cow milk, soybean meal, iso­ ever, in all published methods is a large spectral interference
lated soy protein, fish meal, and meat meal purchased from arising from tyrosine in the second-derivative spectra of prote­
local market. inaceous samples containing all 3 aromatic amino acids. N-
acetyl-L-tyrosinamide, whose spectrum matches that of tyro­
Sample Preparation
sine molecules in proteins (13, 19), exhibits 2 troughs above
Weigh 15-20 mg protein into 50 mL tube or an amount of 275 nm centered at 275.5 and 283 nm, and 2 peaks at 279 and
ground (through 0.5 mm screen) proteinaceous sample con­ 288.5 nm (Figure 1). Because the tyrosine peak at 288.5 nm
taining 60-80 mg protein into blender vessel. Add 25 or 100 overlaps the tryptophan trough at 290 nm, the intensity of the
mL 0.1M sodium hydroxide, respectively, and mix on vortex tryptophan trough depends not only on the tryptophan concen­
mixer or blend for 2 min at high speed. Centrifuge homogenate tration but also on the tyrosine concentration of the sample; any
for 3 min at 2 0 0 0 x g to remove suspended material, and meas­ increase in the tyrosine content reduces the intensity of trypto­
ure absorbance of supernatant at 280 nm. If absorbance value phan trough. Thus, when the tryptophan residues in a protein
exceeds 1 .0 , further dilute an aliquot of the supernatant with the molecule are fewer than those of tyrosine, analytical errors are
same solvent. introduced. Several proposals (15-18) have been made to over-
1170 F letouris Et Al .: J ournal Of AOAC International V ol . 76, No. 6,1993

W a v e le n g th , nm
F ig u r e 1. R e c o r d in g s o f s e c o n d - d e r iv a tiv e U V s p e c tra
o f 1 2.2pM N -a c e ty l-L -try p to p h a n a m id e (— ) a n d 6 0 .9 p M
/V -a ce ty l-L -ty ro s in a m id e (...) in p h o s p h a te bu ffer, p H 6.7.

come this problem, but all have been deficient in one way oi
another.
Ionization of tyrosine produces large changes in its absorp­
tion properties (11, 12, 20). Therefore, our first experiments W avelength, nm
were directed toward examining whether tyrosine ionization
could reduce or eliminate interference to tryptophan determi­ F ig u r e 3. R e c o r d in g s o f fo u rth -d e riv a tiv e U V s p e c tra :

nation by derivative spectrophotometry. Figures 2a and 3a il­ (a) 2 4 .3 pM A f-a ce ty l-L -try p to p h a n a m id e (— ) a n d 1 2 1 .8pM

lustrate second- and fourth-derivative spectra, respectively, of A /-a ce ty l-L -ty ro s in a m id e (...) in 0.1 M N a O H ; (b) m ix tu re o f
th e a b o v e a t 24.3/121,8 p M ra tio in 0.1 M N a O H .

iV-acetyl-i-tryptophanamide, and Af-acetyl-L-tyrosinamide so­

lutions in 0.1M sodium hydroxide. When these spectra are
compared with those recorded at neutral pH, it becomes evi­
dent that the /V-acetyl-i.-tryptophan amide derivative spectrum
does not change at alkaline conditions, whereas that of N-ace-
tyl-L-tyrosinamide shows a large bathochromic shift following
the formation of the phenolate ion. At alkaline pH, the finely
detailed derivative pattern of Wacetyl-L-tyrosinamide in the re­
gion between 275 and 295 nm is essentially lost possibly be­
cause of the broadness of the phenolate absorption band. This
region, however, does not show a zero value in the second-de­
rivative spectra, and therefore, reliable quantitation of trypto­
phan on the basis of the intensity of its second derivative trough
at 290 nm could not be made even under alkaline conditions.
The fourth-derivative spectra presented in Figure 3 seemed to
be more promising. The spectrum of Wacetyl-L-tyrosinamide
is almost flat between 275 and 295 nm, and therefore, interfer­
ence to tryptophan determination by fourth-derivative spectro­
photometry is not normally expected. The presence, on the
other hand, of both a major trough at 285.5 nm and a major
250 270 290 310
peak at 290 nm in this spectral region of ,/V-acetyl-L-tryptopha-
namide provided the means of further increasing the accuracy
W avelength, nm
of the determination by selecting the more appropriate refer­
F ig u r e 2. R e c o r d in g s o f s e c o n d - d e r iv a tiv e U V s p e c tra : ence point for calibration.
(a) 2 4 .3 p M W -a ce ty l-L -try p to p h a n a m id e (— ) a n d 1 2 1 .8pM To select the optimum reference point for the fourth-deriva­
W -a c e ty l-L -ty ro s in a m id e (...) in 0.1 M N a O H ; (b) m ix tu re o f tive measurement of tryptophan at alkaline pH, the effect of the
th e a b o v e a t 24.3/121.8 p M ra tio in 0.1 M N a O H . multiple amounts of Af-acetyl-L-tyrosinamide on the intensity
 F letouris Et Al .: J ournal O f AOAC International V ol . 76, No. 6,1993 1171

T a b le 1. E ffe c t o f L -c y s tin e , W -a ce ty l-L -p h e n y la Ia n in e e th y l e s te r, a n d W -a c e ty l-L -try o s in a m id e o n th e s e c o n d - a n d

fo u r th -d e riv a tiv e s p e c tr o p h o to m e tr ic d e te rm in a tio n o f try p to p h a n a t p H 13

/V-Acetyl-tryptophanamide found ± S D , p M a

Amino acid mixtures, pM Second-derivative Fourth-derivative

/V-Acetyl-L-trypto- /V-Acetyl-phenyl- /V-Acetyl-tyrosin- Trough depth at P eak height at Trough depth at

phanamlde L-Cystine alaninethyl ester am ide 2 9 0 nm 2 9 0 nm 2 8 5 .5 nm

2 4 .3 0 0 0 24 .3 ± 0.0 24 .3 ± 0.0 2 4 .3 ± 0.0

2 4 .3 300 350 0 24 .3 ± 0.5 2 4 .3 ± 0 .0 2 4 .3 ± 0.5
24 .3 300 350 121.8 30 .4 ± 0 .0 2 4 .3 ± 0 .5 2 4 .9 ± 0.8
24 .3 300 350 24 3.7 40 .0 ± 0.8 2 6 .7 ± 0 .8 2 4 .3 ± 0.0
24 .3 300 350 365.4 50 .6 ± 1 .0 28 .7 ± 0 .5 2 4 .3 ± 0.5
24 .3 30 0 350 487.3 58 .2 ± 0.8 30 .6 ± 0 .0 2 7 .8 ± 0 .5

a M ean of 3 determinations.

of either the tryptophan trough at 285.5 nm or the peak at 290 measurement of tryptophan even when these amino acids are
nm was investigated. Trough depth or peak height measure­ mixed at a molar ratio of as low as 5/1. Furthermore, they point
ments were based on values printed on instrumental chart (in out that the fourth-derivative measurement considerably re­
arbitrary units) during derivative processing, values that repre­ duces all interferences arising from /V-acetyl-L-tyrosinamide;
sented, in fact, the vertical distance between the baseline and with the 290 nm peak as reference point, a slight interference
the minimum of the trough (Figure 3b) or the maximum peal: appears at molar ratios greater than 5/1, whereas with the 285.5
height. Because interferences to tryptophan determination nm trough, no interference appears even at a molar ratio of
might also come from other amino acids that contribute to the 15/1. On the basis of these results, the trough at 285.5 nm was
absoiption spectrum of proteinaceous samples at alkaline pH, selected as the optimum reference point for measuring trypto­
(V-acetyl-L-phenylalanine ethyl ester and L-cystine, which is phan concentration.
known (21) to contain sulfhydryl groups with pH values close To further optimize the quantitation, the influence of the de­
to those of ionizing phenol groups, were also included in the rivative wavelength difference (AA) setting of the instrument
experiment. The results presented in Table 1, which also in­ on the intensity of tryptophan trough at 285.5 nm was also ex­
clude measurements on the basis of the second-derivative amined. Table 2 gives the regression equations relating the
trough at 290 nm, show that L-cystine and W-acetyl-L-pheny- concentrations of /V-acetyl-L-tryptophanamide working solu­
lalanine ethyl ester do not interfere with the derivative determi­ tions and trough depths recorded at 2 selected AX settings. The
nation of tryptophan at pH 13. They also suggest that /V-acetyl- raw data suggested that the response becomes greater with in­
L-tyrosinamide seriously affects the second-derivative creasing AA setting, but the regression equations further
showed that such an increase exerts some influence on the
T able 2. R a w d ata a n d re g re s s io n e q u a tio n s of linearity of the calibration curve. Because an increase in AA
calibration c u r v e s fo r tryp to p han determ ination b y generally reduces noise, the slight decline in linearity observed
fourth-derivative sp e c tro p h o to m e try at w avelen gth
with the 3.6 nm setting must be due to deterioration of spectral
difference (A A ) o f 3 o r 3.6 nm
resolution. Thus, the setting of 3 nm was finally adopted.
Tryptophan concentration, Trough depth, arbitrary units The accuracy and the precision of the method were studied
in standard solution at A X
by analyzing a number of pure proteins and comparing the con­
pg/m La |iM 6 3 nm 3 .6 nm centrations found with those expected. The results presented in
Table 3 demonstrate the accuracy of the method as the number
2.4 84 12.165 0.021 0.023
of tryptophan residues found coincided quite well with litera­
4.9 68 2 4 .329 0.041 0.045
ture data based on sequence analysis. The success of the accu­
7.452 3 6 .4 9 4 0.061 0.067
racy experiment shows that proteins can be effectively dena­
9.936 4 8 .659 0.081 0.C91
0.122 0.138
tured in sodium hydroxide solution to expose all buried
14.904 7 2 .988
19.872 9 7 .317 0.162 0.182 tryptophan in the solvent, a finding that lends support to pre­
vious studies (11). It also shows that any substance that pro­
a Regression equation (AT. = 3 nm): Y = (6 .274 x 1 0 " ) + (8 .122 x duces a constant background absorption or a gradual back­
10-3) X (correlation coefficient, r= 0.9 999 9); regression equation
(AX = 3 .6 nm): Y = (-5 .2 9 4 x 1 0 " ) + (9 .2 1 2 x 10“3) X (correlation ground variation can be normalized by the fourth-derivative
coefficient, r = 0 .9 999 0) function, which takes zero value. This makes the method par­
b Regression equation ( A X = 3 nm): Y = (6.271 x 10 4) + (1 .659 x
ticularly useful for analysis in the presence of turbidity or when
10"3) X (correlation coefficient, r = 0.9999 9); regression equation
( A X = 3.6 nm): Y = (-5 .2 9 8 x 1 0 " ) + (1.881 x 1 0 " ) X(correlation the background absorbance is high, as in samples of food and
coefficient, r = 0 .9 999 0). feed ingredients.
1172 F letouris E t Al.: Journal Of AOAC International V ol . 76, No. 6,1993

Table 3. Precision and accuracy data for the Table 5, Precision and accuracy data for the determ­
determination of tryptophan residues in proteins by ination of tryptophan in soybean meal by fourth-
fourth-derivative spectrophotometry derivative spectrophotometry

Num ber of tryptophan Tryptophan M ean

residues Spiking added, M ean tryptophan RSD, recovery,
level ppm 3 found (± SD), ppm % %
Determined Literature 3
Protein 3 value 0 value Recovery, % RSD,
0 ___ 2.11 ( ± 0 . 10) 4.7 —

1 2 .1 5 4.1 9 (± 0 .1 4 ) 3.3 96 .7
Lysozyme 6.0 9 + 0.0 7 6 101.5 1.1
2 4.3 0 6.35 (± 0 .1 0 ) 1.6 98 .6
Albumin bovine 2.0 5 ± 0 .0 5 2 102.5 2.4
3 6.4 5 8.42 (± 0 .1 9 ) 2.3 97 .8
Ovalbumin 2.9 9 ± 0 .0 8 3 99 .7 2.7
4 8.60 10.55 (± 0 .1 8 ) 1.7 98.1
Cytochrome C 1.00 ± 0.0 6 1 100.0 6.0
Histone 0.00 0 — —
a Five replicates.

a Protein concentrations were determ ined by using the following

absorption coefficients: lysozyme, = 3.7 8 x 1 03M 'e m “1 (22);
albumin bovine, = 4.2 x ICFM ’em "1 (23); cytochrome C,
The results from application of the method to a number of
£550 = 2 9 .5 x K f M 'e n r 1 (24); ovalbumin, = 3 .3 x 10 3M " 1cm “1 proteins and food and feed ingredients selected to include a
(25). range of protein content and a variety of protein sources are
b M ean of triplicate analysis ± standard deviation.
c Literature values are from Dayhoff (25).
summarized in Table 4. Although comparison of the found val­
ues with those reported from one sample of a given food or feed
ingredient can be approximated, a survey of the literature
shows that most of these values are identical or close to those
presented by other workers. The minor discrepancies seen for
some of the samples might be due to the quality of proteins
present in these samples or/and to the different quantitation
Table 4. Tryptophan content of proteins and food and systems applied. A recovery experiment conducted on soybean
feed ingredients determined by fourth-derivative
meal samples spiked with lysozyme at 4 levels (Table 5)
spectrophotometry
showed that the relationship between tryptophan added and
Tryptophan, g /100 g Literature data tryptophan found was adequately described by a regression
Sam ple protein® (reference)
(correlation coefficient, 0.99897) and, therefore, the slope
(0.9803) of the regression line (y —2.10 + 0.9803a) could be
Casein 1.38 1.31 (26), 1 .3 6 (4 ) used as an estimate of the overall recovery (98%) of the method
1 .0 2 -2 .2 4 (27)
in the presence of a product matrix. The recovery and relative
C olagenase 1 .5 0 ù — standard deviation values compared well with those reported in
Gliadin 0.83 — a pertinent study (7). The major discrepancies noted for fish
Gluten 0.78 — meal and meat meal samples could not be explained without
C ow milk 1.57 1 .3 8 -1 .5 9 (5) further experiments. Reanalysis of these samples with a stand­
Skim med milk 1 .4 0 (4 ), ard addition method gave substantially similar results.
powder 1.57 1 .1 8 -1 .6 0 (27) Representative fourth-derivative spectra of the analyzed
Isolated soy samples are shown in Figure 4. These spectra show the same
protein 1.60 — general features observed for mixtures of A-acetyl-L-tyrosi-
W heat 1.20 1 .1 4 (7 ,2 8 ) , 1 .1 2 (2 6 ), namide and /V-acelyI-L-tryptophananlide (Figure 3b). The
1.20 (4),
spectral match, however, is not perfect; there is a bathochromic
1 .0 1 -1 .2 0 (27)
shift of less than 1.5 nm within proteins, but this shift has no
Com 1.01 0 .8 0 (7)
influence on tryptophan quantitation.
Barley 1.27 1 .1 3 (7 ), 1 .2 7 (2 ),
The results of this study suggest that the developed method
1.49 (5)
can be a low-cost, fast, and simple alternative to existing meth­
Soybean meal 1.66 1 .3 8 -1 .7 1 (7),
ods for the determination of tryptophan in intact proteins. De­
1.32 (28),
1 .5 2 (5 ,2 9 ), 1 .1 4 (2 6 ), structive hydrolysis and/or time-consuming purification and
1 .1 5 (1 0 ) laborious quantitation is not required. Special safety precau­
Fish meal 0.55 0 .8 4 -0 .9 8 (7), tions, which are quite indispensable in other methods based on
1 .1 7 (2 8 ), 1.2 7 (29), alkaline hydrolysis of proteins in evacuated sealed tubes at high
1.04 (5) temperatures for many hours, are not needed. Further applica­
M eat meal 0.21 0 .6 4 (7) tion of the method may be possible, as indicated by the analyti­
cal results for food and feed ingredients, but each protei­
a Average of duplicate analyses.
naceous product should be evaluated individually for protein
6 Value is expressed in percentage of sam ple weight.
dissolution into extracting solvent. These advantages make the
 F letouris E t A l .: J ournal O f AO AC International V ol . 76, No. 6,1993 1173

270 280 290 300 310

W avelength, nm W avelength, nm
Figure 4. Representative fourth-derivative UV spectra of sam ple extracts: (a) soybean meal; (b) histone; (c) skimmed
milk powder; and (d) fish meal.

fourth-derivative method particularly useful for routine control (14) Nozaki, Y (1986) Arch. Biochem. Biophys. 249,437-446
of protein quality. (15) Balestrieri, C., Colonna, G., Giovane, A., Irace, G., &
Servillo, L. (1978) Fur. J. Biochem. 90, 433-440
R efe ren c es (16) Levine, R.L., & Federici, M.M. (1982) Biochemistry 21,
2600-2606
(1) Pettersson, G., & Eaker, D.L. (1968) Arch. Biochem. Bio- (17) Servillo, L., Colonna, G., Balestrieri, C., Lagone, R., & Irace,
phys. 124,154-159 G. (1982) Anal. Biochem. 126, 251-257
(2) Williams, A.P., Hewitt, D., & Buttery, P.J. (1982) J. Sci. Food (18) Nozaki, Y. (1990) Arch. Biochem. Biophys. 277, 324—333
Agric. 33, 860-865 (19) Solli, N.J.. & Herskovits, T.T. (1973) Anal. Biochem. 54,
370-378
(3) Spies, J.R., & Chambers, D.C. (1948)Anal. Chem. 20, 30-39
(20) Beaven, G.H., & Holiday, E.R. (1952) Advan. Protein Chem.
(4) Lucas, B., & Sotelo, A. (1980) Anal. Biochem. 109, 192-197
7,319-386
(5) Nielsen, H.K., & Hurrell, R.F. (1985) J. Sci. FoodAgnc. 36,
(21) Benesh, R.E., & Benesh, R. (1955) J. Am. Chem. Soc. 77,
893-907
5877-5881
(6 ) Delhaye, S., & Landry, J. (1986) Anal. Biochem. 159, 175— (22) King, T.P., & Spencer, M. (1970)7. Biol. Chem. 245, 6134—
178 6148
(7) Landry, J., Delhaye, S., & Viroben, G. (1988) J. Agric. Food (23) Van Gelter, B.F., & Slater, E.C. (1962) Biochim. Biophys.
Chem. 36, 51-52 Acta 58, 593-595
(8 ) Sato, H., Seino, T., Kobayashi, T., Murai, A., & Yugari, Y. (24) Glazer, A.N. (1959) Australian J. Chem. 12, 304—307
(1984) Agric. Biol. Chem. 48, 2961-2969 (25) Dayhoff, M.O. (1972) Atlas of Protein Sequence and Struc­
(9) Jones, A., Hitchcock, C.H.S., & Jones, G.H. (1981) Analyst ture, vol. 5 (suppl. 2, 1976) National Biological Reference
106,968-973 Foundation, Washington, D. C.
(10) Allred, M.C., & MacDonald, J.L. (1988) J. Assoc. Off. Anal. (26) Sarwar, G., Christensen, D.A., Finlayson, A.J., Friedman, M.,
Chem. 7 1 ,603-606 Hackler, L.R., Mackenzie, S.L., Pellett, P.L., & Tkachuk, R.
(11) Bencze, W.L., & Schmid, K. (1957) Anal. Chem. 29, 1193— (1983) J. Food Sci. 48, 526-531
1199 (27) Food and Agriculture Organization (1970) Amino Acid Con­
(12) Goodwin, T.W., & Morton, R.A. (1946) Biochem. J. 40, tent of Foods, FAO Nutritional Studies No. 24, FAO, Rome
628-632 (28) Wemer, G. (1986) Landwirtsch. Forsch. 36, 109-117
(13) Edelhoch, H. (1967) Biochemistry 6 , 1948-1954 (29) Delhaye, S., & Landry, J. (1986) J. Cereal Chem. 4, 117-123
1174 N ewlon E t Al.: J ournal Of AOAC International V ol . 76, No. 6,1993

AGRICULTURAL MATERIALS

E v a lu a tio n o f C u r r e n t A O A C F e rtiliz e r S a m p le P r e p a r a tio n

R e q u ire m e n ts

N atalie N ewlon , C andace C ox-T rout , a n d P eter K ane

Purdue University, Department of Biochemistry, West Lafayette, IN 47907

T he official AOAC fertilizer sa m p le p rep aratio n re­ screen. This resulted in a product of varying particle size, and
q u ire s th a t all dry m ixtures be g ro u n d to p a s s a grinding only the fraction of the sample that did not pass the
U.S. No. 40 siev e. With cu rren t fertilizers a n d m e­ No. 40 sieve, often with a mortar and pestle, was a common
ch an ical g rin d ers, th e s e criteria m ay no lo n g er be practice.
ap p ro p riate. B lended fertilizers w ere g ro u n d a n d In time, these pulverized fertilizers gave way to “granu­
siev ed , an d th e fractio n s w ere analy zed sep a rately lated” fertilizers, in which various dry materials were mixed
to sh o w potential variability in re su lts. In general, together and then treated with ammonia or combinations of
p o ta ssiu m w a s heavily c o n c e n tra te d in th e sm aller ammonia and phosphoric or sulfuric acid. The result was a
p articles of th e g ro u n d sam p le, w h e re a s p h o s p h o ­ more uniform granulated product that did not need to be
ru s te n d e d to b e c o n c e n tra te d in th e c o a rs e r parti­ crushed, and therefore, had a somewhat smaller percent of fine
c les. A re p resen ta tiv e s e t of fertilizers w a s s u b ­ materials than the typical pulverized fertilizer.
jecte d to 7 grinding tre a tm e n ts d e sig n e d to Although fertilizer production progressed through many
p ro d u c e sa m p le s with a w ide distrib u tio n of parti­ variants of granular products, the materials remained granular,
cle size. N, P, a n d K w ere d eterm in ed in th e s a m ­ not pulverized. In fact, current emphasis on bulk blending of
p les by traditional m eth o d s. A nalysis of th e re su lts granular components has made control of particle size of even
d e m o n stra te d th a t p re cise, a c c u ra te re su lts could more concern than in the past because of segregation tenden­
b e o b tain ed from sa m p le s th a t did n o t technically cies when particle sizes of components of bulk blended prod­
m eet cu rren t sa m p le p reparation req u irem en ts. ucts are not uniform.
T he relatio n sh ip b etw een fin e n e ss of grind an d The granular nature of today’s product does not lend itself
size of th e analytical sa m p le portion w a s ex am ­ to first sieving, then grinding the coarser portion to pass a No.
ined. With p ro p e r sa m p le grinding, sa m p le s iz e s of 40 sieve. Also, the mortar and pestle method is no longer the
< 2 0 m g co u ld give re p resen ta tiv e re su lts for nitro­ method of choice for the preparation of analytical samples, and
g en an aly sis, with p recisio n eq u al to o r b etter th an virtually everyone grinds the entire analytical sample with a
traditional Kjeldahl an a ly sis u sin g 1 g sa m p le s. mechanical grinder of some sort. Sieving, if done at all, is used
T his is of particular in terest, b e c a u s e v ario u s com ­ mainly as a periodic check of the grinder’s performance.
b u stio n in stru m e n ts now beco m in g po p u lar for ni­ The mechanical grinders most widely used fall into 2 main
tro g e n a n a ly sis are limited in sa m p le size. classes. The first is the traditional Wiley Mill type, which has
rotating knife blades and uses a shearing and slicing motion to
fracture the large particles in blended fertilizers. The second
T he present AOAC method for fertilizer sample prepara­ type has no blades, only paddles, and operates via high-speed
tion (1) was established in 1941, on the basis of a col­ impact to produce a finely ground sample. The fineness of the
laborative study reported that year (2). For fertilizer ma­ ground fertilizer sample may be affected by the method of
terials and moist fertilizer mixtures, the method specifies that grinding. Mechanical grinders tend to produce samples in
the sample be ground to pass a U.S. No. 20 sieve. For dry mix­ which the majority, but not all, of the particles will pass a U.S.
tures, the method specifies that the sample be ground to pass a No. 40 sieve. On the other hand, modem fertilizers ground in
U.S. No. 40 sieve. mechanical grinders produce a sample in which the overall av­
The method was suitable for the usual type of fertilizer erage particle size is much smaller than the average particle
product at that time, called “pulverized” fertilizer, in which sizes of the samples used in the 1941 collaborative study. Data
various solid materials were mechanically mixed together, al­ in Table 1 are examples of this kind of distribution.
lowed to cure and cake, and then crushed to pass a 6 -mesh In view of the changing physical nature of fertilizer prod­
ucts and the outmoded state of AOAC fertilizer sample prepa­
Received July 15, 1992. Accepted February 25, 1993. ration methodology, past assumptions and procedures are due
Published as Journal Paper No. 13452 of the Purdue University for review. This paper will attempt to document the potential
Agricultural Experiment Station. variability inherent even in fine, mechanically ground samples.
 N ewlon E t A l . : J ournal O f AOAC International V ol . 76, No. 6,1993 1175

Table 1. Distribution of particle size (% of total), pooled for each treatment

Treatment

Sieve fraction A B C D E F G

+12 68.27 27.24 0.09 0.15 0.01 0.00 0.00

-1 2 + 20 28.67 47.40 14.83 3.71 1.19 0.10 0.08
-2 0 + 40 1.48 11.69 39.27 13.25 9.32 2.29 0.14
-4 0 + 60 0.45 4.23 14.33 15.10 13.15 8.11 1.14
-6 0 + 100 0.29 2.72 8.24 14.28 13.66 12.12 6.25
-1 0 0 + 200 0.38 3.10 9.41 22.00 22.27 23.17 22.70
-2 0 0 0.29 3.25 13.20 30.71 39.67 53.46 68.56
% Lost3 0.17 0.37 0.63 0.80 0.73 0.75 1.13

a Percent lost due to sieving process.

We will measure particle size differences in samples ground by Results and discussion.—The NPK results obtained in this
different grinding techniques and examine the effect on preci­ study were highly dependent on particle size. For example,
sion of nitrogen, phosphorus, and potassium (NPK) results. Fi­ K20 values for one fertilizer ranged from 7.74 to 52.62% de­
nally, we will explore in more detail the relation between par­ pending on the sieve fraction analyzed. Because all fertilizers
ticle size, sample size, and precision in nitrogen results. behaved similarly, fertilizer 5 was selected as a representative
example (Table 3). Duplicate results represent independent
within-day analyses. Because of the limited amounts of some
S eg reg atio n P otential of G round S am p les
sieve fractions, nitrogen results were incomplete.
In general, K20 values increased with decreasing particle
This study was designed to investigate the extent to which size, indicating that potassium was concentrated in the smaller
potential segregation of particles in ground blended fertilizers particles. P2 0 5 values (total and available) tended to decrease
could affect nitrogen, K2 0 , and P2 0 5 measurements. Seven fer­ with decreasing particle size, indicating that phosphoms was
tilizers (Table 2) were selected as representative of typical concentrated in the larger particles. For nitrogen, although val­
blended products. They were ground in a Brinkmann Model ues were variable, there was no clear trend with respect to par­
ZM-1 grinder, a high speed impact type grinder using a screen ticle size.
with 5.0 mm openings. The screen size was selected to give a The potential harm of ground sample segregation is obvi­
relatively higher proportion of the larger particle sizes, but the ous, and Table 3 illustrates that segregation is more likely to
average particle size was still small compared to 1941 stand­ affect K20 and P2 0 5 results and less likely to affect nitrogen
ards. Each sample was then sieved through U.S. standard results.
sieves No. 14, 20,35,40, 60, 100, and 200 to separate the dif­
ferent particle sizes within the sample. The sieve fractions for Varying th e F in e n e ss of G rind
the 7 fertilizers were each analyzed in duplicate for nitrogen
(978.02) , K20 (983.02), and both total (958.01) and available This study was designed to determine the minimum grind­
(960.03) P2 Os (1). ing treatment necessary to achieve acceptable precision and ac-

T a b le 2. C o m p o n e n t m a k e u p o f b le n d e d fe r tiliz e r s b y p e rc e n t

Fert No. Grade N-P-K Urea3 DAP* G TSPC Potash d S-Ureae Limestone'

10-38-10 __ 55.6 27.0 16.7 — 0.70

1
2 6-15-40 — 33.3 — 66.7 — —

3 16-16-16 21.2 34.8 — 26.7 — 17.3

4 12-12-12 15.9 26.1 — 20.0 — 38.0

5 10-20-20 4.7 A3.5 — 33.3 — 18.5

6 20-10-10 35.0 21.7 — 16.7 — 26.6

7 32-3-3 46.65 — 6.5 4.9 33.0 8.95

3 Urea, 46-0-0.
b Diammonium phosphate, 18-46-0.
c Granulated triple super phosphate, 0-46-0.
d Coarse potash, 0-0-60.
e Sulfur-coated urea, 32-0-0.
' 6 x 1 6 imestone filler, 0-0-0.
1176 N ewlon Et A l .: J ournal Of AOAC International Vol. 76, No. 6,1993

T a b le 3. V a ria b ility o f n itro g e n , p o ta s s iu m , a n d p h o s p h o r u s d u p lic a te r e s u lt s fro m a r e p re s e n ta tiv e fe r tiliz e r

(F e rtiliz e r N o . 5 ,1 0 -2 0 -2 0 )

Sieve fraction Nitrogen, % Soluble K20 , % Available P 2O 5, % Total P 20 5, %

+14
__ __ a __ __ --- ------------ -------------- --- , ---

- 1 4 + 20 — , — 5.6 5, 5.7 0 2 1 .9 0 ,2 1 .1 7 24 .03, 2 5 .3 7

- 2 0 + 35 13.31, 12.27 — , 3 .5 8 29 .70, 3 2 .5 3 30 .33, 3 3 .8 9

- 3 5 + 40 1 3 .4 8 ,— 4.12, 4.3 2 28 .86, 3 0 .7 3 31 .86, 30 .70

- 4 0 + 60 1 3 .1 1 ,— 8.68 , 8.65 28 .16, 2 8 .5 7 30 .65, 2 8 .4 3

-6 0 + 100 11.81, 11.81 18.84, 18.69 22 .76, 22 .30 2 2 .7 0 , 2 2 .5 5

-10 0 + 200 10.39, 10.34 29 .22, 2 9 .0 0 14.79, 14.79 13.37, 14.71

-200 10.95, 9.12 29 .18, 28 .72 lowb, low 12.66, 12.71

a indicates no results.
b indicates result w as below instrument standard range.

curacy and to avoid segregation problems in blended fertilizers sample. Within each grinding treatment, particle size distribu­
when the “traditionally” accepted sample sizes are used for tions were similar for all fertilizers. Therefore, the results from
analysis. All K20 and P2 0 5 analyses started with 1 g samples. all fertilizers were pooled to produce an average distribution
Kjeldahl analyses used 1.2 g samples, except for the 32-3-3, for each grinding treatment.
which used a 0 . 6 g sample. Results and discussion.—For each of the 9 fertilizer sam­
Each of the 7 fertilizers used in the previous study (Table 2), ples, a mean and standard deviation (SD) of results over the 7
a granular triple super phosphate fertilizer, and a diammonium grinding treatments were calculated for nitrogen, K2 0 , total
phosphate fertilizer were sampled by using a standard probe to P2 0 5, and available P2 0 5. An approximate 95% confidence in­
obtain approximately 2 quart samples. Each sample was split terval was also calculated (mean ± 2 SDs). Data are shown in
into 8 equal portions with a gated riffle (Carpco, Jacksonville, Tables 5-8.
FL 32206), and each portion was placed in an 8 oz. jar. Examination of the tabulated results showed that treatments
The splits from each fertilizer were processed by 7 different A (unground) and B (Wiley 4.0 mm screen) produced the most
grinding treatments (Table 4). These treatments were designed variable results. Many of the values obtained from these 2 treat­
to produce samples with a wide range of average particle size, ments were more than 2 SDs away from the sample mean, and
therefore, fell outside the 95% confidence interval. This was
from unground to very fine.
true for all 4 analyses. Because of their extreme variability',
All fertilizer grinds were then analyzed in duplicate for ni­
treatments A and B were considered unacceptable, and these
trogen, K2 0 , and total and available P2 0 5 by the Official Meth­
results were eliminated from further consideration.
ods described in the segregation study.
When Youden’s test for outlying values (3) was applied to
After all analyses were complete, the remaining sample por­
the remaining 5 treatments (C-G), 24% of total P2 0 5 values
tions were sieved by using a Tyler Ro-Tap® sieve shaker and
and 17% of available P 2 0 5 values obtained from treatment C
U.S. standard sieves No. 12,20,40,60,100, and 200. The pur­
were determined to be outliers. Therefore, grinding treatment
pose of the sieving was to quantitatively define the particle size C was also eliminated from further consideration. (Nitrogen
distribution of each grinding treatment. The sieve fractions for and potassium results were not as variable for treatment C.)
each sample were weighed and compared to the total sample Seven other values were determined to be outliers by Youden’s
weight, and each fraction was expressed as a percent of the total test and were eliminated, but these were randomly distributed
among all other treatments and no pattern was evident.
T a b le 4. G r in d in g tre a tm e n ts to a c h ie v e v a ry in g Tables 9-12 summarize the means and variability for treat­
d e g re e s o f f in e n e s s
ments D-G (Brinkmann 5.0, 1.5, 0.75, and 0.75/0.50 mm
Treatment Description screens, respectively). Outlying values were excluded from
calculations.
A Portions left unground In general, results from treatments D-G were consistent and
B Portions ground on W iley Mill3 (4.0 mm) accurate, as reflected in the SD values. Further elimination of
C Portions ground on W iley Mill (2.0 mm) treatments did not significantly reduce SDs. Analyses (P2 0 5,
D Portions ground on Brinkmann 0 (5.0 mm) K2 0 , and nitrogen) were similar with respect to SD, relative
E Portions ground on Brinkmann (1.5 mm) standard deviation (RSD), and range values for all fertilizers.
F Portions ground on Brinkmann (0 .75 mm)
Analysis of variance applied to the results from these 4 treat­
G Portions double ground on Brinkmann (0 .7 5 /0 .5 0 mm)
ments indicated no significant differences between treatments
H Portions left unground as reserves
D-G.
a W iley mill has a shearing/cutting type of grinding motion. Table 1 summarizes the results obtained when all fertilizer
samples were sieved to classify the grinding treatments accord-
 N ewlon E t A l .: J ournal Of AOAC International V ol . 76, No. 6,1993 1177

T ab le 5. T otal P 2O 5 r e s u lts (% o f to ta l) fo r all tr e a tm e n ts

Fertilizer No.

Treatment 1 2 3 4 5 6 7 8 9

A 34 .14 11.11 20 .77 13.54 19.88 1 2 .2 2 a 3.0 8 4 7 .9 4 46 .62

33.86® 10.83 20.11 14.32 21 .82 8 .0 4 3 .2 7 4 7 .7 5 4 6 .5 4
B 34 .20 9.4 4 13.66 7 .2 1 a 20 .19 8.8 0 3 .7 4 4 7 .5 7 46.21
37.42 7 .1 9 a 11,74a 7 .2 5 a 19.99 7.3 0 3 .9 2 a 4 7 .2 3 a 4 6 .5 3
C 36.02 13.34 18.08® 13.83® — 7 .5 4 3 .2 7 4 7 .7 5 46 .55
36.01 13.26 18.82 11.06* 2 0 .9 1 ® 7 .3 7 3 .3 7 a’® 4 7 .2 3 46 .53
D 37 .20 13.86 19.04 11.81 22.61 7 .0 8 3.51 4 7 .6 9 4 6 .3 5
— 13.44 19.03 11.89 2 2 .2 8 7.4 6 3 .3 9 4 7 .6 5 4 6 .4 6
E 37 .10 13.42 18.94 12.23 2 2 .1 4 7 .4 3 3 .2 9 47.81 46 .79
37.01 13.89 18.74 12.09 2 2 .6 3 7 .0 9 3 .2 9 47 .92 46.86
F 36.84 14.02 19.32 11.82 22.38 7 .1 2 3.0 2 4 8 .0 7 46 .68
37 .02 14.07 19.29 11.95 2 2 .3 6 7.0 6 3.0 4 4 7 .8 9 46 .70
G 37 .08 13.69 19.47 12.23 2 2.12 7 .1 5 3 .2 2 4 7 .5 6 46 .73
37 .19 13.71 19.42 12.37 21 .95 7 .2 4 3 .3 3 4 7 .7 0 46 .84

M ean 36.31 12.52 18.32 11.69 21 .59 7 .7 8 3 .3 4 47 .70 46 .60

SD 1.16 2 .12 2.55 2.09 0.95 1.08 0.2 5 0.21 0.18
3 3 .9 9 - 8 .2 8 - 1 3 .2 2 - 7 .5 1 - 1 9 .6 9 - 5 .6 2 - 2 .8 4 - 4 7 .2 8 - 4 6 .2 4 -
95% C lc 3 8 .6 3 16.76 23 .42 15.87 23 .49 9.94 3 .8 4 4 8 .1 2 4 6 .9 6

8 Outside 95% confidence interval.

b Outlier according to Youden's test.
c Cl, confidence interval.

T a b le 6. A v a ila b le P 2O 5 r e s u lt s (% o f to ta l) f o r a ll tre a tm e n ts

Fertilizer No.

Treatment 1 2 3 4 5 6 7 8 9

A 36 .88 13.04 14 .51a 7.63 21 .36 5 .3 4 a 3 .1 9 4 6 .0 8 45.45®

34 .83 9 .4 5 18.43 9.15 19.45 6.23 3 .1 3 4 5 .6 0 45 .88
B 3 5 .3 8 7.5 4 15.97 7.08 16 .18a 6.8 7 3 .4 5 4 5 .6 5 45 .66
3 3 .5 4 a 7 .2 9 14.57® 7.58 18.53 5.33® 4.00® 45 .63 45.78
C 36.41 13.29 18.69 12.04 2 0 .0 7 7.3 3 3.66® 45 .92 45 .84
35.89® 13.04 17.96 13.37® 21 .07 7.79 3 .0 4 4 6 .0 0 45 .72
D 36 .56 7.55® 18.33 11.93 22 .08 7.2 9 2 .8 5 4 5 .9 7 45 .58
36 .54 13.68 19.03 11.83 2 2 .3 7 7.44 2 .9 5 4 5 .0 4 45 .58
E 36.83® 13.82 18.52 11.88 22.52 7.27 3 .1 8 4 6 .1 4 45 .96
36.53 13.87 18.70 11.87 22.51 6.36® 3.2 6 4 6 .1 0 45 .95
F 36.42 — 19.42 11.79 2 2 .2 9 7.02 2.9 8 4 5 .3 0 45 .68
36.51 14.04 19.33 11.96 2 2 .0 3 7.00 2 .9 5 4 5 .2 8 45 .92
G — 13.68 19.31 11.91 2 1 .9 5 7.12 3.2 2 45.55® 45 .96
36 .60 13.80 19.31 12.24 21 .96 7.12 3 .1 4 4 5 .9 4 45 .00

M ean 36.11 12.01 18.01 10.88 21 .03 6.81 3.21 4 5 .9 4 45 .78

SD 0.78 2.4 3 1.55 2.10 1.82 0.67 0.2 7 0.22 0.15
95% C lc 3 4 .5 5 - 7 .1 5 - 1 4 .9 1 - 3 .6 8 - 1 7 .3 9 - 5 .4 7 - 2 .6 7 - 4 5 .5 0 - 4 5 .4 8 -
37 .67 16.87 2 1.11 15.08 24 .67 8 .1 5 3.7 5 46 .38 46 .08

8 Outside 95% confidence interval.

b Outlier according to Youden’s test.
c Cl, confidence Interval.
1178 N ewlon Et A l .: J ournal O f AOAC International V ol . 76, No. 6,1993

T a b le 7. K 2O r e s u lts (% o f to ta l) f o r a ll tre a tm e n ts

Fertilizer No.

Treatment 1 2 3 4 5 6 7

A 12.58 3 5 .2 4 a 13.32 14.43 20.51 11.46® 4.0 0

14.82 4 7 .6 6 1 1 .2 1 a 15.44 22.25 15 .95a 4.9 3
B 14.06 48 .97 16 .62a 14 .14a 22 .16 14.39 5.15
16 .82a 5 1 ,28a 15.93 16 .61a 21.20 15 .27a 5.06
C 13.15 4 3 .0 7 13.76 15.02 17.16 16.54 4.1 4
13.70 43 .72 13.66 14.82 18.40 14.26 3.69
D 13,06 4 2 .3 5 13 .4 2 ai> 16.08 17.12 12.83 4.1 2
12.83 4 3 .1 6 13.33 14.96 17.87 13.26 4.09
E 12.66 4 2 .5 8 14.46 14.72 16.88 13.72 3.7 8
13.22 4 2 .5 8 14.17 14.67 18.07 1 2 . 2 3 a ,b 3.71
F 13.32 42 .52 13.24 14.67 17.05 13.86 3.66
13.32 4 2 .2 6 13.31 14.92 17.35 13.87 3.65
G 13.16 43 .03 13.50 14.63 17.47 13.74 4.0 8
13.20 42 .60 13.58 14.83 17.82 14.16 4.0 8

M ean 13.57 43 .65 13.83 15.00 18.67 13.76 4.1 6

SD 0.8 7 2.92 1.24 0.3 2 1.97 0.6 2 0.50
95% Cl 1 1 .83-15.3 1 3 7 .8 1 -4 9 .4 9 11.35-16.3 1 1 4 .3 6 -1 5 .6 4 1 4 .7 3 -2 2 .6 1 1 2 .5 2 -1 5 .0 0 3 .1 6 -5 .1 6

a Outside 95% confidence interval.

b Outlier according to Youden’s test.
c Cl, confidence Interval.

T a b le 8. N itro g e n r e s u lts (% o f to ta l) fo r a ll tre a tm e n ts

Fertilizer No.

Treatm ent 1 2 3 4 5 6 7 8 9

A 10.97 6.11 22.75® 10.34 8.90® 25.34® 3 7 .2 4 _

17.90
10.51 6.25 18.33 10.57 11.64 21.20 — — 17.89
B 10.92 4 .4 4 a 16.50 10 .12 13.08 19.05 2 7 .8 5 — 17.90
1 1 .1 1 ® 4.84 15.88® 10.92 11.22 20.49 2 8 .5 3 — 17.64®
C 10.34 5.34 20 .14 12.19 12.93 22 .23 31 .99 — 17.82
10 .54b 5.15 19.86 12.87 12 .12 22 .07 32 .55 — 17.90
D 10.28 5.20 18.54 12.69 12.98 21 .24 31 .75 — 17.79
10.23 5.25 18.59 12.28 12.88 20.69 31 .99 — 17.74
E 10.34 5.52 18.78 12.83 12 .12 19.98 31 .83 — 17.85
10.30 5.46 18.52 12.44 12.61 20.02 31.59 —
17.65®’6
F 10.23 5.47 18.41 11.76 12.95 20.51 32 .49 — 17.86
10.32 5.49 18.61 11.91 12.75 20.45 31 .89 — 17.89
G 10.25 5.37 18.67 12.04 12.32 19.92 31 .42 — 17.86
10.18 5.41 18.68 12.00 12.36 19.83 31 .52 — 17.70

M ean 10.47 5.38 18.73 11.79 12.21 20.93 32 .13 __ 17.82

SD 0.30 0.46 1.38 0.93 0.90 1.32 2.66 — 0.06
9 .8 7 - 4 .4 6 - 1 5 .9 7 - 9 .9 3 - 1 0 .4 1 - 1 8 .2 9 - 2 6 .8 1 - 1 7 .7 0 -
95% Cl 11.07 6.3 0 21 .49 13.65 14.01 23 .57 37.45 —
17.94

a Outside 95% confidence interval.

6 Outlier according to Youden’s test.
0 Cl, confidence Interval.
 N ewlon E t A l .: J ournal Of AOAC International V ol . 76, No. 6,1993 1179

Table 9. Total P2O5 results for treatments D-G

Fertilizer No.

Param eter 1 2 3 4 5 6 7 8 9

M ean (%) 37 .06 13.76 19.16 12.05 22.31 7.20 3.26 47 .79 46 .68
SD 0.123 0.2 44 0.2 57 0.212 0.2 38 0.159 0 .1 6 6 0.1 67 0.181
RSD 0.332 1.772 1.341 1.760 1.065 2.208 5.1 03 0.3 50 0.388
Range 3 6 .8 4 - 1 3 .4 2 - 1 8 .7 4 - 1 1 .8 1 - 2 1 .9 5 - 7 .0 6 - 3 .0 2 - 4 7 .5 6 - 4 6 .3 5 -
37 .20 14.07 19.47 12.37 22 .63 7.46 3.51 4 8 .0 7 46 .86

T a b le 10. A v a ila b le P 2O 5 r e s u lts f o r tre a tm e n ts D - G

Fertilizer No.

Param eter 1 2 3 4 5 6 7 8 9

M ean (% ) 36.51 13.82 18.99 11.93 22.21 7.18 3.07 46.11 4 5 .8 3

SD 0.078 0.1 34 0.4 22 0.138 0.2 38 0.160 0.151 0.141 0.181
RSD 0.212 0.970 2 .2 2 3 1.155 1.073 2.224 4 .9 3 7 0.306 0.3 96
Range 3 6 .4 1 - 1 3 .6 8 - 1 8 .3 3 - 1 1 .7 9 - 2 1 .9 5 - 7 .0 0 - 2 .8 5 - 4 5 .9 4 - 4 5 .5 8 -
36 .60 14.04 19.42 12.24 22.52 7.44 3.26 46 .30 46 .00

T a b le 11. T o ta l N itro g e n r e s u lts f o r tre a tm e n ts D - G

Fertilizer No.

P aram eter 1 2 3 4 5 6 7 8 9

M ean (%) 10.27 5.40 18.60 12.24 12.62 20.33 3 1 .9 4 _ 17.81

SD 0.053 0.116 0.113 0.382 0.3 23 0.4 84 0.3 84 — 0.071
RSD 0.5 20 2.150 0.6 10 3.132 2.5 60 2.3 82 1.201 — 0.3 99
Range 1 0 .1 8 - 5 .2 0 - 1 8 .4 1 - 1 1 .7 6 - 1 2 .3 2 - 1 9 .8 3 - 3 1 .4 2 - — 1 7 .7 0 -
10.34 5.52 18.78 12.83 12.98 21.24 3 2 .4 9 17,89

T a b le 12. P o ta s s iu m r e s u lt s f o r tre a tm e n ts D - G

Fertilizer No.

Parameter 1 2 3 4 5 6 7 8 9

M ean (% ) 13.10 42 .64 13.66 14.77 17.45 13.60 3.90 _ _

SD 0.237 0.3 10 0.4 73 0.132 0.431 0.4 74 0 .2 1 4 — —
RSD 1.808 0.7 28 3.4 59 0.893 2.471 3.4 85 5 .4 8 0 — —
Range 1 2 .6 6 - 4 2 .3 5 - 1 3 .2 4 - 1 4 .6 3 - 1 6 .8 8 - 1 2 .8 3 - 3 .6 5 - — —
13.32 43 .16 14.46 14.96 18.07 14.16 4.12
1180 N ewlon E t A l .: J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3

Table 13. Comparison of nitrogen results (% of total) Table 14. Variation of 0.75-0.5 mm double ground
for Kjeldahl 0.75 mm grind and combustion method with blended fertilizer compared to the variation of standard
0.75 mm grind and 0.75-0.5 mm double grind of a materials within a run (% of total)
blended fertilizer
Ground blended
Carlo Erba fertilizer,
Kjeldahl 0 .7 5 Carlo Erba 0 .7 5 0 .7 5 -0 .5 mm P aram eter 0 .7 5 -0 .5 mm (N H 4)2S 0 4 nh 4n o 3
mm grind, 1200 mm grind, 15 grind, 15 mg
Param eter mg sample mg sample sample 6.388 2 1 .3 1 9 3 5 .0 2 4
6.2 84 2 1 .3 6 7 3 4 .7 8 8
6.32 6.2 73 6 .3 8 8 6.3 18 2 1 .3 7 5 34.691
6.32 6.2 46 6.2 84 6.321 21.351 3 4 .9 7 3
6.3 3 6.3 54 6.3 18 6.2 08 2 1 .3 5 4 3 4 .8 0 5
6.3 5 6.1 80 6.321 6.3 65 2 1 .375 34.791
6.39 6.241 6.2 08 6.3 53 21.351 3 4 .6 5 3
6.4 0 6.0 30 6.3 65 6.3 15 2 1 .3 1 2 3 4 .9 0 9
6.41 6 .4 0 3 6 .3 5 3 6.2 18 2 1 .2 9 0 3 4 .802
6.36 6 .4 5 5 6 .3 1 5 6.3 47 2 1 .3 3 9 3 4 .817
6.45 6.4 96 6.2 18 6.392 — —
6.3 7 6 .2 3 0 6.3 47 6.321 — —
6.3 4 6.668 6.3 92 6.393 — —
6.33 6.3 02 6.321 6.356 — —
6.3 8 6.5 78 6.3 93 6.3 54 — —
6.41 6.5 04 6.3 56
6.39 6.2 90 6 .3 5 4 M ean 6.3 29 2 1 .343 3 4 .8 2 4

6.33 — — SD 0.056 0.028 0.1 157

RSD 0.885 0.133 0 .3 3 2
M ean 6.368 6.3 5 6.3 29
SD 0 .0 3 8 7 0.167 0.0 56
RSD 0.608 2.64 0.8 85

smaller sample sizes. Depending on the particular instrument

model, sample sizes of 2 0 mg or less might be appropriate.
ing to particle size distribution. Because results from all 9 fer­ Traditional wisdom suggests that a representative sample could
tilizers were similar, they were pooled to produce an average not be expected for sample weights in this range. This section
distribution for each grinding treatment. The distribution of attempts to determine what the lower limits of sample size
particle size varied enormously from treatment to treatment, might be, given the considerably finer grinds possible with cer­
but even the finest grinding treatment left a very small percent tain modem grinders.
of material retained on the No. 40 sieve. Technically, none of The specific combustion instrument used was a Carlo Erba
the treatments would pass the official recommendation, by Nitrogen Analyzer 1500. Instrument parameters were as fol­
which all of the sample must pass a U.S. No. 40 sieve to be lows: combustion tube, 1020°C; reduction tube, 650°C; chro­
acceptable. matography oven, 60°C.; filament, 190°C; total cycle time, 190
The critical point, however, is that the majority of the parti­ s; sample start time, 2 s; and stop time, 80 s. This sets the purg­
cle sizes in the samples from the 1941 study establishing these ing time of the sample in the autosampler. Oxygen inject time
criteria were in the range of 40 to 100 mesh. In the present was 81s, and peak enable was 0 1 .
study, for grinding treatments D-G, the majority of the particle Software parameters were set to match the instrument pa­
sizes were concentrated well below 100 mesh. Average particle rameters for furnace, oven, and filament temperatures. The he­
size is at least as much a controlling factor in sample prepara­ lium flow (measure) was 100 mL/min, and helium flow (refer­
tion as is an arbitrary minimum particle size. Therefore, AOAC ence) was 40 mL/min. Oxygen pressure was 100 Kpa, and the
sample preparation requirements should be reconsidered with oxygen flow rate was 20 mL/min. The helium flow (purge) was
average particle size in mind. 50 mL/min. The oxygen loop was 10 cc, and sample delay time
was 2 s.
N itrogen A n aly sis a n d Sm all S am p le S izes Two different grinds of a 6/24/24 fertilizer, a product
blended from multiple ingredients, were prepared on a Brink-
In Varying the Fineness o f Grind, we demonstrated that mann Mill. The first was ground once with a 0.75 mm screen,
among grinding treatments D-G, relatively little precision is and the second was double-ground with 0.75 and 0.5 mm
gained by using the finer grinds. This work used sample screens. The single grind was analyzed in replicate by the
weights of approximately 1 g. Recently, there has been much Kjeldahl method by using 1200 mg samples. Both the single
interest in combustion instrumentation as a substitute for the and the double grinds were analyzed in replicate by the com­
traditional Kjeldahl analysis, but these instruments require bustion instrument by using 15 mg samples.
 Newlon E t A l .: J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3 1181

Results and discussion.—The 0.75 mm single grind gave Sample segregation is more likely to influence K20 and P2 0 5
more variable results by nitrogen analyzer than by the Kjeldahl results than nitrogen results.
method. The SD and RSD for the nitrogen analyzer were con­ The 1941 sample preparation collaborative study is out­
siderably higher than the Kjeldahl method for the single grind. moded and needs to be reconsidered by taking average particle
However, the variation of results of the 0.75/0.5 mm double­ size into account. With modem grinding techniques, particle
ground sample by the nitrogen analyzer was very close to the sizes are much smaller than the 40-100 mesh range found in
variation of the single grind by the Kjeldahl method. The data the 1941 study. The majority of particle sizes can be smaller
for both the nitrogen analyzer and the Kjeldahl method appear than 200 mesh. Average particle size is a better measure of
in Table 13. good sample preparation than maximum particle size.
To separate method error from sample preparation enor, the With finer grinding techniques and smaller average particle
same 0.75/0.5 mm double-ground blended fertilizer sample sizes, procedures can be used that require smaller sample
and pure standard materials, ammonium sulfate and ammo­ weights than the traditional 1 g sample. The combustion instru­
nium nitrate, were analyzed by the combustion instrument (Ta­ ment in this study used a 15 mg double-ground blended fertil­
ble 14). The RSD of the double-ground blended fertilizer rea­ izer, and the variation in results was similar to the classical
Kjeldahl method, in which 1200 mg regular ground samples
sonably approached RSDs of the 2 standard materials. RSDs of
are used. Combustion instruments using these small sample
all 3 materials were <1.0%. Additional sample grinding for
sizes are able to give results with SDs on blended fertilizers that
blended fertilizers would lead to relatively little improvement
approach the SD values of the standard materials. RSDs of the
in overall variation.
blended fertilizers and of standard materials were <1%. With
The nitrogen analyzer gave results on a blended fertilizer with
the finer grinding techniques commonly used today, the small
variability comparable to the Kjeldahl method. With the finer sample size can give a representative result for nitrogen.
grinding techniques commonly used today, considerably smaller
sample sizes can give representative results for nitrogen.
R eferen ces

C o n clu sio n s (1) Official Methods of Analysis (1990) 15th Ed., AOAC,
Arlington, VA
Generally, K20 tends to be more concentrated in the finer (2) Ross, W.H., Rader, L.F., Jr, & Hardesty, J.O. (1941) J. Assoc.
particle sizes, and P2 0 5 tends to be concentrated in the larger Off. Agric. Chem. 24,253-263
particles of ground fertilizers. Nitrogen does not tend to con­ (3) Youden, W.J., & Steiner, E.H. (1975) Statistical Manual of
centrate in any one particle size and is more evenly distributed. the AOAC, AOAC, Arlington, VA
1182 N ew lon : J ournal O f AOAC International V ol . 76, No. 6 , 1993

AGRICULTURAL MATERIALS

E valu atio n o f N ew Flam e P hotom etric In stru m en tatio n to M e e t

R equirem ents o f A O A C O ffic ia l M eth o d fo r Potassium in
F ertilize rs

N a t a l ie N ew lon

Purdue University, Department of Biochemistry, West Lafayette, IN 47907-1154

AOAC Official M ethod 983.02 for p o ta ssiu m in fertil­ Also specified is the correction for the effect of phosphate on
izers c o n ta in s in stru m en t p erfo rm an ce criteria. T he potassium response in the flame. This correction is made by
B ran & L uebbe Flam e P h o to m e te r 410 with a varying the lanthanum oxide concentration in the final aspi­
TRAACS 800 w a s ev a lu a ted by th is m ethod. Noise, rated solution. All these performance criteria permit some vari­
drift, carryover, a n d p re cisio n w ere investigated. ation among instrument types and setups as long as final per­
Drift a n d n o ise w ere well below th e sp ecificatio n s. formance is maintained and chemistry is not changed.
C arryover w a s e sse n tia lly zero. T he overall p er­ This paper describes a new instrument configuration that
fo rm a n ce of th e s y ste m b a se d on National Institute involves no change in the basic chemistry of the current official
of S ta n d a rd s a n d T echnology s ta n d a rd s a n d q ual­ method, meets all of the performance criteria of the method,
ity con tro l s a m p le s sh o w e d an a v e ra g e b ias of and represents a significant upgrade in automated technology.
0 .10%, w hich is within th e m eth o d sp ecificatio n s;
th e a b s o lu te v alu e s of d ifferen ces w ere 0.166% for E xperim ental
th e citrate ex tractio n a n d 0.063 a n d 0.171%, re s p e c ­
tively, fo r th e o x alate ex tra ctio n s. T he m eth o d Apparatus
s ta te s th a t th e av e ra g e b ias sh o u ld b e <0.1, a n d th e
a b s o lu te v alu e of d ifferen ces sh o u ld b e <0.4. O ver­ The instrument used was one channel of a Bran & Luebbe
all, th e s y s te m p a s s e d all th e p erfo rm an c e criteria TRAACS 800 2-channel instrument. An Alternate Detector In­
sp ecified in th e m eth od. terface (ADI) box, supplied by the instrument manufacturer,
connected the instrument’s software with a Bran & Luebbe
(Coming) 410 flame photometer detector. The instrument
T he original flame photometric method for potassium in manifold supplied by Bran & Luebbe was designed for soil
fertilizers ( 1 ) was a labor-intensive and time-consuming analysis and was modified as illustrated in Figure 1. The fol­
manual method. After extraction, a resin bed had to be lowing parameters are used with the instrument: 90 samples/h;
prepared and an aliquot of the neutralized fertilizer sample 4:1 sample/wash ratio; no pecking; quadratic curve fit; base not
rinsed through the resin bed. The sample was then diluted and in calibration; baseline drift correction; ran protocol P, 2N, 6 C,
atomized through the flame. H, 3L, (10S,I)x3, H, 3L, 2N, G, 4N, E; null peaks are lower
concentrated standard; carryover lows are No. 1 standard; gain
Automating this method (2) made a significant improve­
references to higher concentrated standard in calibration se­
ment over the manual method, because, in addition to the bene­
quence.
fits of automation, the extracted and diluted sample could be
directly analyzed on the instrument. Unfortunately, the auto­ If standard concentrations are input in mg/mL:
mated method was written for a specific instrument and spe­ q (calc, sample concn) x (5000000)
cific modules and parameters and did not allow for variation 2 (aliquot) x (sample weight in mg)
among instruments or manifolds.
The current manual/automated method for potassium in fer­
tilizers (3) includes the same extraction as the original auto­ Or for spiked samples:
mated method. The method is more flexible, however; it in­
„ „ „ [(calc, sample concn) x (100) - 30] x [1000]
cludes a set of performance criteria that every instrument must /OJVjU — .
(sample weight in mg)
meet and allows for instrument variation. The criteria specify
the permissible amount of noise, carryover, drift, and precision.
(Calculation assumes spike =10 mL of 3 mg K2 0/mL solution.)
Received December 19, 1992. Accepted January 24, 1993. Although not described in this paper, the Alternate Detector
Published as Journal Paper No. 13647 o f the Purdue University Interface can also be used to interconnect the TRAACS soft­
Agricultural Experiment Station. ware and an AAII autoanalyzer witn the older flame IV detec-
 N ewlon : J ournal Of AOAC International V ol . 76, No. 6,1993 1183

K2Q in F e r tiliz e r
Rangs 0 .1 - 0.B Mg K20/ml Sampler
9 0 /H r , 4 :1 S/W
M a n ifo ld # lfi5 -D 0 3 5 -fll, m o d ifie d

5 Turn A ir
B l/ B l LB2O3
___________» ____ 0QDD_
| ¡S ' W t/U t S aspla
____________________B l/B l D ilu a n t
„ G ry/G ry Sampler Wash

T° UMhlel" *— Sa"PlBr Uash

z . B l / V l D s ta c to r Waste
---------- — ----------------

20 Turn 20 Turn

_ u u u u ______ u u u u X X II

N o ta: Usa 0 .0 4 "

410
Tygon f o r a l l
Flans
* 11B-G001-01 In je c t io n F i t t i n g Photometer conn ections ex cep t

* * 178-G 215-01 G lass “Tee" 768™ F i l t e r use 20mn o f 0 .0 2 5 "

* * * 11B-B332-01 S trean S p l i t t e r Tygon from dabubblar

**h G lass Dabubblar F i t t i n g to Flame Photometer

F ig u re 1. F lo w d ia g ra m o f in s tru m e n t m a n ifo ld .

tor. This connection is straightforward and provides another Because the fertilizer methodology required La2 0 3 to damp
route to an official instrument configuration and retains some the phosphate effect in the flame detector, an La2 0 3 reagent
of the benefits of the newer instrument software technology. fine was substituted for one of the diluent fines. The concentra­
tion of that reagent was changed (see Performance Parame­
Instrument Manifold Modifications ters).
The sample: wash ratio was changed from 6 :1 to 4:1, and the
The instrument manufacturer supplied a method (4) for po­
sampling rate was slowed from 120 to 90 samples/h to improve
tassium in soil extracts. Several modifications were made so
peak separation. The ADI box gain select control was set at 10
that the system would be applicable to fertilizer extracts. The
instead of the 1 0 0 setting suggested by the instrument manu­
dialysis membrane was omitted from the manifold, because its
facturer.
function was to dilute solutions in the concentration range of
1000 mg K2 0/L. This laboratory did not plan to analyze solu­ The wetting agent was changed from Renex 30 to Brij 35.
tions more concentrated than 600 mg K2 0/L. This simplified Brij 35 was chosen, because it was used with the AAII systems
the manifold. and is much more stable than the Renex 30, which tends to
An extra 20-turn coil was added to the manifold for addi­ deteriorate after dilution. The amount of wetting agent in the
tional mixing before the solution enters the nebulizer. This coil diluent reagent was increased to 2 mL/L.
reduced the noise considerably. To minimize carryover, the The Official Method suggests that chloroform be added to
tubing line from the glass debubbler fitting to the nebulizer was the standards as a preservative. Chloroform biased the standard
shortened as much as possible and the inlet tube remained curve slightly, and it is no longer used in this laboratory. In­
pointing down. stead, the standards are stored in total darkness, which pre­
Pump tube sizes for the sample and waste fine were ad­ serves them for ca 1 month.
justed. The sample pump tube size was increased to get the The material used to prepare the standard curve was
least possible dilution ratio on line without flooding the detec­ changed from National Institute of Standards and Technology
tor. The waste fine size was increased but not to the point that (NIST) KH2 P 0 4 to NIST K N 03. The phosphate ion concentra­
bubbles were aspirated into the nebulizer. This reduced the tion from the KH2 P 0 4 material was high enough to cause re­
noise level of the detector. sidual phosphate effect and bias the standard curve. The low
Because the dialysis membrane was eliminated, HC1 was no end of the standard range was changed from 0.0 to 0.1 g K2 0/L.
longer needed to provide ionic strength across the membrane. Because of software considerations in the new instrument, a
Therefore, that fine was also removed. The sample line, 2 dilu­ spike is now added to bring low-level samples into the concen­
ent fines, and the waste fine were left. tration range of the standards.
1184 N ew lon : J ournal O f AOAC International V ol . 76, No. 6,1993

Reagents the official method, 983.02. Each performance test is discussed

in turn.
(a) Sam ple w ash solution. —For ammonium oxalate-ex­ The specification of the LiN0 3 concentration in the final
tracted samples. Use degassed, deionized H2 0 . solution aspirated into the flame was written for a photometer
(b) Sam ple w ash solution. —For ammonium citrate-ex­ equipped with a second (lithium) detector used for an internal
tracted samples. Dilute 200 mL ammonium citrate reagent standard. The LiN 0 3 concentration was adjusted so that detec­
[963.03A(a)] to 1000 mL with degassed, deionized H2 0 . tor responses for Li and K were approximately equal. The new
(c) D ilution w a te r solution. —Add 2 mL Brij 35 wetting flame photometer has only a potassium detector. Because the
agent to 1 L degassed H20 and mix. potassium channel is stable by itself, it has no need for the Li
(d) L a20 3 solution. —Dry La2 0 3 for 2 h at 105°C. Weigh internal standard. Therefore, LiN0 3 is not added to the La2 0 3
3.5214 g into 1 L volumetric flask, add 90 mL H N 03, and dis­ reagent.
solve. Dilute to volume with degassed HzO, add 1 mL Brij 35 The phosphate effect specification was checked first so that
wetting agent, and mix. reagents could be adjusted to the final concentrations before
(e) P otassium stan da rd solutions. —Dry NIST K N 0 3 for 2 specifications for noise, drift, carryover, and precision were
h at 105°C. (7) Prepare working standards of 0.1, 0.2,0.3,0.4, checked. Phosphate effect, i.e., the depression of the potassium
0.5, and 0.6 g K2 0/L for ammonium citrate-extracted samples response in the flame because of the presence of phosphate, is
by weighing 0.2148 g K N 0 3 into one 1000 mL volumetric determined by preparing 2 K N 0 3 solutions with the potassium
flask; 0.2148, 0.3222, 0.4296, and 0.5370 g samples into 500 concentrations equal to the highest standard concentration.
mL volumetric flasks; and 1.2886 g into another 1000 mL volu­ Enough NH4 H2 P0 4 is added to one flask to equal the highest
metric flask, respectively. Add 100 mL ammonium citrate re­ concentration of POj 3 expected in any real sample. These 2
agent per 500 mL final volume, dilute to volume with deion­ solutions are analyzed alternately, and the average is calculated
ized H 2 0 , and mix. Standards are stable until they begin to turn for each solution. The averaged responses should not differ by
cloudy or for maximum of 1 month. Store in total darkness. more than 1 %.
(2) Prepare working standards of 0.1,0.2,0.3, 0.4,0.5, and The concentration of the La2 0 3 reagent used for the
0.6 g K2 0/L for ammonium oxalate-extracted samples by AutoAnalyzer II was adjusted for the TRAACS instrument,
weighing 0.2148 g K N 0 3 into one 1000 mL volumetric flask; and the differences in pump tube sizes and amounts of reagents
0.2148, 0.3222, 0.4296, and 0.5370 g samples into 500 mL used were taken into account. To eliminate the phosphate effect
volumetric flasks; and 1.2886 g into another 1000 mL volumet­ for citrate and oxalate extractions in the TRAACS instrument,
ric flask, respectively. Add 1.0 g (N H ^Q tL t to the 500 mL La2 0 3 and U N 0 3 concentrations had to be increased 3-fold
flasks and 2.0 g to the 1000 mL flasks. Dilute to volume with over those used with the AAII (Table 1). The amount of H N 0 3
deionized H20 and mix. Standards are stable for maximum of used was the minimum amount that still prevented the precipi­
1 month. tation of La2 0 3 in the coils of the TRAACS manifold.
(3) Prepare a spike containing 3 mg K2 0/m L for low guar­ The performance specification for noise requires that noise
antee samples by weighing 6.4433 g K N 03, diluting to 1000 should be <2% full scale. To determine the noise, the highest
mL with deionized H2 0 , and mixing. Use 10 mL of the spike concentrated standard was aspirated continuously through the
solution with each sample >5% K2 0 . instrument. The instrument output was monitored for 5 min,
and the highest and lowest readings were recorded. Both the
Determination oxalate and the citrate standards passed the criteria.
Carryover is determined by analyzing 3 higher concentrated
Extract samples with either ammonium citrate or ammo­ standards followed by 5 lower concentrated standards. The car­
nium oxalate (3). With the citrate extraction, if the sample has ryover on the first lower standard must be < 1 % full scale when
a guarantee <5%, add 10 mL of the K20 spike standard. If the compared to the average of the following 4 lower standards.
guarantee is >30%, take a 50 mL aliquot and dilute it to 100 Carryovers on the TRAACS instrument were essentially zero
mL. Add 10 mL ammonium citrate extractant before diluting for both the oxalate and citrate standards.
to volume to keep the citrate concentration the same in samples
The drift criteria require that detector output not vary by
and standards.
more than 1 % full scale for length of time required to analyze
The oxalate extraction is used for samples with >30% guar­ 10 samples. There was essentially no drift for either the oxalate
antees if no phosphorus determination is required. Therefore, and citrate standards.
the oxalate extraction requires that a 50 mL aliquot of all sam­
The precision criteria require that when 30 mid-range stand­
ples must be taken and diluted to 100 mL. The oxalate stand­
ards are sampled, the range of the instrument response not vary
ards were prepared with the amount of ammonium oxalate to
by more than 2% (optimum is 0.7%). The range of the response
reflect the amount of oxalate in the diluted samples.
was 0.6% for the citrate extraction and 0.59% for the oxalate
extraction.
P erfo rm an ce P a ra m e te rs The criteria for the standard curve require that for each of
the 6 standards the difference between calculated and known
The instrument as configured in the A p p a r a tu s section was concentration not be more than 2%. The average of the absolute
tested to see if it met the performance specifications given in values of the percent differences may not be more than 1 %.
 N e w l o n : J o u r n a l Of AOAC I n t e r n a t io n a l V o l . 76, No. 6,1993 1185

Table 1. Elimination ofphosphate effectby use ofLa203 (g)forcitrateand oxalate

No [La20 3 ] AAII [La20 3] AAII [La20 3] x2 AAII [La20 3] x3

Parameter Aa Bb Aa Bb Aa B6 Aa Bb
Citrate reagent

0.2491 0.2548 0.2960 0.3021 0.2994 0.3024 0.3023 0.3041

0.2514 0.2542 0.2954 0.2995 0.2965 0.2993 0.3012 0.3013
0.2515 0.2531 0.2992 0.2993 0.2963 0.3002 0.3012 0.3028
0.2490 0.2521 0.2984 0.3011 0.2973 0.3013 0.3021 0.3046
0.2485 0.2515 0.2961 0.3002 0.2966 0.2996 0.3021 0.3013
0.2490 0.2548 0.2966 0.3001 0.2995 0.3002 0.3029 0.3033
0.2478 0.2525 0.2971 0.2989 0.2991 0.3030 0.3032 0.3057
0.2487 0.2523 0.2934 0.3001 0.2976 0.3001 0.3049 0.3044
0.2477 0.2534 0.2934 0.2992 0.2933 0.3028 0.3041 0.3057
0.2502 0.2519 0.2969 0.2963 0.2996 0.3012 0.3054 0.3062

Mean 0.24929 0.25306 0.29626 0.29968 0.2980 0.3010 0.3029 0.3039

SD 0.0013 0.0020 0.0019 0.0015 0.0013 0.0013 0.0015 0.0018

Ratios 0.9851 0.9886 0.9900 0.9967

Oxalate reagent

0.2542 0.2579 0.2928 0.2983 0.3012 0.3028 0.3036 0.3050

0.2556 0.2586 0.2915 0.2973 0.3001 0.3027 0.3053 0.3043
0.2544 0.2610 0.2948 0.2950 0.3031 0.3050 0.3031 0.3057
0.2547 0.2607 0.2961 0.2974 0.3011 0.3041 0.3045 0.3048
0.2529 0.2613 0.2932 0.2976 0.3032 0.3034 0.3070 0.3050
0.2542 0.2594 0.2965 0.2987 0.3004 0.3044 0.3058 0.3065
0.2534 0.2575 — — 0.3028 0.3013 0.3040 0.3041
0.2535 0.2589 — — 0.3023 0.3043 0.3038 0.3041
0.2558 0.2601 — — 0.3044 0.3025 0.3038 0.3025
0.2555 0.2608 — — 0.3024 0.3030 0.3031 0.3041

Mean 0.2544 0.2596 0.29412 0.29738 0.3021 0.3034 0.3044 0.3046

SD 0.0010 0.0014 0.0020 0.0013 0.0014 0.0011 0.0013 0.0011
Ratios 0.9800 0.9890 0.9957 0.9993

a NH4H2P04added to solution.
b NH4H2P04not added to solution.

Table2. Check ofstandard curve

Citrate Oxalate

Theoretical, g Actual, g Difference, % Actual, g Difference, %

0.1000 0.1003 +0.3 0.09972 -0.28

0.2000 0.1997 -0.15 0.2007 +0.35
0.3000 0.3030 +1.0 0.3016 +0.53
0.4000 0.4049 +1.22 0.3998 -0.05
0.5000 0.5041 +0.82 0.4967 -0.66
0.6000 0.6010 +0.17 0.6020 +0.33

Mean of differences 0.61 0.37

1186 N e w l o n : Jo u rn a l Of AOAC I n t e r n a t io n a l Vol. 76, No. 6,1993

Table3. Performance check and qualitycontrol Table4. Performance check and qualitycontrol
samples forthecitrateextraction samples fortheoxalate extraction:sample, KNO3;
theoretical % K20 ,46.56
Difference,
% K20, % K20, actual - Difference,
Sample theoretical actual theoretical, % % K20, actual actual - theoretical, %

kh 2po 4 34.64 35.068 0.428 46.92 0.36

35,198 0.558 46.57 0.01
34.958 0.318 46.62 0.06
34.583 -0.057 46.64 0.08
34.512 -0.128 46.65 0.09
46.77 0.21
KN03 46.56 46.645 0.085
46.75 0.19
46.496 -0.064
46.71 0.15
46.687 0.127
46.63 0.07
QC 12 24.85 24.843 -0.007 46.36 -0.20
25.000 0.15 46.82 0.26
25.064 0.214 46.63 0.07
25.092 0.242 46.90 0.34
46.62 0.06
QC 13 12.07 11.883 -0.187 46.56 0.00
12.083 0.013 46.98 0.42
12.087 0.017 46.71 0.15
12.004 -0.066 46.61 0.05
46.53 -0.03
Mean 0.10
46.32 -0.24
Absolute mean 0.17
46.13 -0.43
46.27 -0.29

Mean 0.063
Absolute mean 0.171

Originally, the standard solutions were made from KH2P04. Conclusions

The standard material was changed to KN03when Idiscov­
eredthatthecurvewas biasedbecauseoftheamount ofphos­ With thechangeofthematerialsusedforthestandardsolu­
phateinthestandardmaterial.With KN03asthestandardma­ tionsand adjustment oftheLa203concentration,resultsfrom
terial,thebiasdisappeared. A quadraticcurvefitwas usedfor performancecheck samplesagreedwiththetheoreticalvalues.
bothcitrateandoxalatestandards. Performance criteria were met. The new flame photometer
The largestdifferencebetweenknown andcalculatedstand­ passed all the performance criteria specified in the AOAC
ardconcentrationswas 1.22%.The averageoftheabsoluteval­ method.
ues(theoreticalvsactual)was 0.61% forcitrateand0.37% for
oxalate(Table2).Both ofthesenumbers were wellwithinthe References
parametersoftheperformancecriteria.
The final performance criteriainvolved running standard (1) O ffic ia l M ethods o f A nalysis (1984) 14th Ed., A O A C , A r­
materials and qualitycontrolsamples. The citrateand oxalate lington, VA, secs 2.108-2.113
methods both gave good agreement between the actual and (2) O ffic ia l M ethods o f A nalysis (1984) 14th Ed., A O A C , A r­
theoreticalresults(Tables 3 and 4).The performance criteria lington, VA, secs 2.114-2.118
fortheaverage biasfrom themethod is0.10, and theaverage (3) O ffic ia l M ethods o f A nalysis (1990) 15th Ed., A O A C , A r­
oftheabsolutedifferencesis0.4.Resultsforcitratewere 0.10 lington, V A sec. 983.02
fortheaveragebiasand0.166forthe average oftheabsolute (4) Sim ultaneous D ete rm in a tio n o f Sodium a n d Potassium in
differences.The oxalateextractionsresultedinaveragebiasof S o il Extracts (1990) M ethod 894-90T, B ran & L uebbe A na­
0.063 and averageoftheabsolutedifferencesof0.171. lyzing Technologies, Inc., E lm sford, N Y
 S in g h & C h ib a : J o u r n a l O f AOAC I n t e r n a t io n a l V o l . 76, No. 6,1993 1187

AGRICULTURAL MATERIALS

R e v e rs e d -P h a s e L iq u id C h r o m a t o g r a p h ic M e th o d f o r th e

S im u lta n e o u s D e t e r m in a t io n o f B e n o m y l a n d M e t h y l

1/ / - B e n z i m i d a z o l- 2 - y lc a r h a m a t e i n W e t t a b le P o w d e r

F o r m u la t io n s

R ajP.S ingh
Brock University,Chemistry Department, St.Catharines,ON, L2S 3A 1,Canada
M ikio C hiba *1
AgricultureCanada, Research Station,Vineland Station,ON, LOR 2E0, Canada

A liquid chromatographic (LC) method isreported pound or as an impurity. The concentration of MBC in WP
forthe simultaneous determination of benomyl, formulations may be high ifthe formulatedproducts arekept
methyl {1-[(butylamino)carbonyl]-1 H-benzimidazol- under moist conditions or at elevated temperatures (5, 6).
2-yl}carbamate and carbendazim, methyl 1H- Therefore, themethod reportedby Teubertand Stringham (1,
benzimidazol-2-ylcarbamate (MBC), inwettable 2) may overestimate theactualbenomyl contentofaWP for­
powder (WP) formulations. Inthis method, mulation.Thus,thereisaneed foran accuratemethod thatsi­
benomyl isconverted to 3-butyl-2,4-dioxo-s-triaz- multaneouslydeterminesintactconcentrationsofbenomyl and
ino[1,2-a]benzimidazole (STB) in 0.125M sodium hy­ MBC inWP formulations.Thisneedhasbecome more appar­
droxide. After conversion (atroom temperature in entinlightoftherecentcontroversyoverbenomylphytotoxic­
20 min), STB isdetermined simultaneously with ityproblems observedinsome partsoftheUnitedStates(7).
Previously,we reportedasimpleRP-LC method forsimul­
MBC (which remains intactwith the sodium hydrox­
taneousdeterminationofbenomyl andMBC inwatersamples
ide treatment) at286 nm after LC separation on a
(8, 9). In thispaper we report a modified version of RP-LC
Cis column. The method was accurate and showed method for the simultaneous determination of benomyl and
good reproducibility, with relativestandard devia­ MBC inWP formulations.We alsocarriedoutadditionalstud­
tions of 0.7 and 2.2% for benomyl and MBC determi­ iestoinvestigatethereliabilityofthismethod.
nations, respectively.
Experimental
D eterminationofbenomyl,methyl {l-[(butylamino)car- R e a g e n ts
bonyl]-l//-benzimidazol-2-yl)carbamate, in wettable
powder (WP) formulations by reversed-phase liquid (a) B e n o m y l a n d M B C . — Analytical standards were ob­
chromatography (RP-LC) has been described intheliterature tainedfromE.I.duPontdeNemours andCo.,Inc.Benomyl and
(1,2).In this method (1, 2) WP formulations were extracted MBC wereassumed tobe 100% pure.
with acetonitrilecontaining 3% «-butyl isocyanate (BIC) and (b) S T B (3 -b u ty l-2 ,4 -d io x o -s -tr ia z in o [l,2 -a ]b e n z im id a -
chromatographedon areversed-phaseC18column. z o le ) a n d B B U [l-(2 -b e n z im id a z o ly l)-3 -n -b u ty lu r e a ]. — STB

The reported method (1, 2) was based on the finding by and BBU werepreparedfrombenomyl asreportedearlier(8).
Chiba and Cherniak (3) that the degradation of benomyl to The purityofSTB was testedby LC and assumedtobe 100%
MBC (methyl lH-benzimidazol-2-ylcarbamate) in most or­ inthisstudy.The structuresofbenomyl,MBC, STB,BBU, and
2-AB (2-aminobenzimidazole)areshown inFigure 1.
ganic solvents isreversible,thatis,benomyl (A2i) MBC +
(c) S o lv e n ts. — Acetonitrile and methanol were LC-grade
BIC (C|).Thereversibilityofthisreactionwas documentedby
reagents from Caledon Laboratories Ltd., Georgetown, On­
stabilizingbenomyl inorganic solventsinthepresence ofex­
cess BIC (4). However, excess BIC would also produce tario,Canada.
(d) O th e r re a g e n ts. — Sodium hydroxide, nitric acid, and
benomyl from MBC. MBC, which isone ofthemajor degra­
disodiumhydrogen phosphate were AnalaR grade from BDH
dation compounds ofbenomyl, isalways present in variable
Chemicals Canada Ltd.,Toronto,Ontario,Canada. Potassium
quantities in WP formulations either as a degradation com­ dihydrogen phosphate (FishercertifiedACS grade)was from
FisherScientificLtd.,Toronto, Ontario, Canada. Buffersolu­
Received November 4, 1992. Accepted March 8, 1993. tions (pH7) were prepared by mixing 0.067M solutions (v/v)
1 To whom correspondance should be addressed.
ofNa2HP04andKH2P04ina3:2ratioasreportedearlier(9).
1188 S in g h & C h ib a : J o u r n a l O f AO AC I n tern a tio n a l V o l . 76, N o . 6 ,1 9 9 3

nitricacid,5 mL methanol,and 5 inLacetonitrileand made up

tovolume withpH 7 buffer.
0= C— N— C4H9 C,H„
A p p a r a tu s
( .N c N C -OCH, c— N-
II A Perkin-Elmerseries3LC equippedwithaRheodynesyr­
H O a .
H inge loop-typeinjectorand a Perkin-Elmer LC-55-S detector
BENOMYL STB was usedatroom temperature25-27°C. The wavelengthused
I
N\ was 280 nm, and injectionvolumes were 10-50 p.L.The col­
C— N— C— OCHs
^ I II umn usedwas anODS (C18),15cm x 4.6m m RegisHiChrom
H O reversiblereversed-phase,packed with 5-|im Spherisorb.A 5
MBC cm x 4.6mm precolumndrypacked withCo:PellODS 25-37
N\ I pm (Whatman) was used in series before the analytical col­
C N— C— N— C,H8 nn
I II I umn.
H O H
M o b ile P h a s e
BBU 2-AB
A mixture of40% acetonitrile,45% water, and 15% pH 7
Figure 1. Molecular structuresofbenomyl, MBC, STB, phosphate buffer was run isocratically at a flow rate of 0.8
BBU, and 2-AB. mL/min.

Calculations
(e) S a m p le s .
— Benlate 50% WP (50% benomyl) (DuPont
ofCanadaLtd.,Montreal, Quebec, Canada) and WP formula­ P e r c e n ta g e c o n c e n tr a tio n o f a n a ly z e d b e n o m y l a n d
tions consisting of 10% benomyl and 50% captan (Chipman M B C .— The percentages of analyzed benomyl and MBC in
Chemicals Ltd., Stony Creek, Ontario, Canada) were used. 50% WP (equationscanbe easilymodifiedfor 10% formula­
These samples (some of which were 4 to 7 years old) were tions)were calculatedasfollowsfrom LC results:
collectedfrom growers.
Ps
(p ) * Cs
S a f e t y C o n s id e r a tio n s
% Benomyl = -- xl.124x100 (1)
w
The (LD50)values (orally in rats)ofbenomyl, MBC, and
captan, are 10000 mg/kg, >15 000 mg/kg, and 9000 mg/kg,
respectively(10).Normal laboratoryproceduresshouldbefol­
lowed when handling thesepesticidesand sodium hydroxide
(astrongbase). %MBC =— -- x 100 (2)
Vv
S t a n d a r d S o lu t io n s
where ( P s / P s -)x Cs istheweightofSTB inWP analyzed;P s
A standard solution of MBC was prepared inmethanol at and P s'arethepeak heightsofSTB insample and standard,
100mg/L (weighedtothenearest0.1mg).A standardsolution respectively; C s = (5 mg/1000 mL) x 250 mL x (100 mL/5
ofSTB was preparedinacetonitrileat100mg/mL (weighedto mL); W isthe weight (mg) of WP; 1.124 isthe ratio of the
thenearest0.1mg).A mixed standardsolutioncontainingSTB molecular weight of benomyl to that of STB (290/258);
and MBC was prepared by appropriate dilutionofthe MBC ( P m / P m - ) x C m istheweightofMBC inWP analyzed;P m and
and STB standardsolutionsasfollows: 5 mL STB standard,5 P m -arethepeak heightsofMBC insample and standard,re­
mL MBC standard,5mL 0.125M sodiumhydroxide,and5mL spectively; and Cm = (5 mg/1000 mL) x 250 mL x
0.125M nitricacid were mixed ina 100 mL volumetric flask (100mL/5 mL).
andmade uptovolume withpH 7buffer.The peakheight-ver-
sus-concentrationresponsesforSTB and MBC were linearin Results and Discussion
the0 to10mg/L and0 to7 mg/Lconcentrationranges,respec­
tively.The linearityshouldbechecked iftheworkingrangeis L C S e p a r a tio n o f S T B a n d M B C
substantiallyhigherthantheseconcentrationranges.
The factorsthataffecttheLC separationofSTB andMBC
S a m p l e P r e p a r a tio n havebeendiscussedinmoredetailinourpreviouspublications
(8,9).As mentionedinthesepapers,theseparationofSTB and
FiftymilligramsofBenlate50% WP or250mg (weighedto MBC isinfluenced by the composition (i.e.,ratioofacetoni­
thenearest0.1mg) ofBenlate-captan(10% benomyl and50% trile,water,andbufferinthemobilephase)andpH ofthemo­
captan)was quicklydispersedin250 mL 0.125M sodium hy­ bilephase. Because thepeak height and theretentiontime of
droxide(pH 13)solution.After20min,5mL clearsolutionwas STB arealsoinfluencedby changesinsamplecomposition,for
put into a 100 mL volumetric flaskcontaining 5 mL 0.125M accurate determinations the composition of sample solutions
 S in g h & C h ib a : J o u r n a l O f AO A C In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1189

andthestandardsshouldbekeptthesame. Inthepresentwork
we usedamixtureof40% acetonitrile,45% water,and 15% pH
7phosphatebufferasmobilephase.Thismobilephasegavethe
shortestretentiontimes ofSTB and MBC with theirbaseline
resolution on an ODS (C18) column at a flow rate of
0.8mL/min. Linearityoftheresponse(peakheightversuscon­
centration) was checked in the concentration range 0.35 to
7.0 mg/L forMBC and 0.5 to 10.0mg/L forSTB. The slopes
oftheplotsofpeakheightversusconcentrationremainedcon­
stantthroughoutthe analysisforallcompounds, with a linear
correlationcoefficient of 0.9999 or better. With the detector
sensitivity at0.02 absorbance units and 50 jiLinjection vol­
umes,thedetectionlimitsforMBC andSTB wereestimatedat
0.035mg/L.
E f f e c t o f S o d i u m H y d r o x id e C o n c e n tr a t io n o n
Q u a n tita tiv e C o n v e r s i o n o f B e n o m y l to S T B

The quantitativeconversionofbenomyl toSTB inWP for­

mulations involves 2 steps: (/) thedissolution ofbenomyl in
theWP formulationsinsodiumhydroxidesolutionand(2)the
conversionofdissolvedbenomyl toSTB. A delay,which may
beafunctionofsodiumhydroxideconcentration,intheseproc­
essesmay leadtothepartialconversion ofbenomyl toMBC.
As evidentfromourpreviousresults(9),dissolutionofWP and
conversionofdissolvedbenomyltoSTB werefasterinsodium
hydroxide solutions of pHs greater than 12. Although ithas
beenreported(11)thatbenomyl selectivelyconvertstoSTB in
alkalinesolutionsofpHs greaterthan 10.5,on thebasisofour
previouswoik(8),we suspectthatapartialconversionofbenomyl MIN
toMBC alsotakesplaceinthepH rangefrom 10.5to12.6.
Figure2. Chromatograms ofbenomyl solutionsafter
Inourpreviouspaper(9)we reportedthatapH 13 sodium approxiately40 min in(a)0.125,(b)0.2,and (c)0.5mol/L
hydroxide solution(0.125M) was appropriateforquantitative NaOH.
conversion of benomyl to STB in water samples containing
relativelylowconcentrationsofbenomyl(<25mg/L).Because
the concentration of benomyl in WP formulations ishigher ofbenomyltoSTB arethatSTB startsconvertingtoBBU with
thaninwatersamples,an optimizationofalkaliconcentration increasedsodiumhydroxideconcentrations(Figure2)andthat
toyieldselectiveconversionofbenomyltoSTB wasdesirable. a slow conversion ofMBC to2-aminobenzimidazole (2-AB)
To accomplishthisoptimization,LC analysisofaWP sample isobservedinstrongeralkalisolutions.
dissolvedindifferentconcentrationsofsodiumhydroxidewas
carriedout.TheseresultsarereportedinFigure2.The chroma­ K in e t ic s o f C o n v e r s i o n o f B e n o m y l to S T B in p H 1 3
togramspresented inFigure2 revealthattheheightsofMBC N aO H
peaksareapproximatelyequalinsolutionsofpH 13andhigher.
Thisfactmay beindicativeofcompleteconversionofbenomyl To determine the minimum time required forquantitative
toSTB inthesesolutions(i.e.,sodium hydroxide solutionsof conversionofbenomyl toSTB, a kineticstudyofthereaction
pHs greaterthanorequalto13)duringdissolutionandconver­ was undertaken. The results indicate fast conversion of
sion steps, because an alkali concentration as high as 0.5M benomyltoSTB atroomtemperature(24°C),withafirst-order
mustassuretheselectiveconversionofbenomyl toSTB (11). rateconstant(k)ofapproximately0.01 sec'1andareactionhalf
The chromatogram (notshown here)resultingfrom WP solu­ time (t0 5) of 1.15 min. On thebasis ofkineticsresults,more
tion in pH 12.6 (0.1M) sodium hydroxide showed a slightly than 99.9% benomyl should be converted to STB in 12 min.
largerMBC peak thanthatresultingfrom pH 13 sodium hy­ BecausethedissolutionofWP isnotinstantaneous,areaction
droxidesolution.Thisdifferencemay beduetopartialconver­ timeof20min was usedtoensurecompleteconversionofdis­
solvedbenomyl toSTB.
sionofbenomyltoMBC duringthedissolutionstepinpH 12.6
alkalisolution.From thesestudieswe concludethatpH 13so­ E f f e c t o f N a tu r e o f H y d r o x id e C o m p o u n d o n
dium hydroxide solutionisthe most appropriatemedium for Q u a n tita tiv e C o n v e r s i o n o f B e n o m y l t o S T B
quantitative conversion of benomyl to STB in WP formula­
tions.Thedisadvantagesofusinghigherconcentrations(higher The conversionofbenomyl toSTB was alsocarriedoutin
than0.125M) ofsodiumhydroxideforquantitativeconversion tetramethylammoniurnhydroxide(TMAH) tostudytheeffect
1190 S in g h & C h ib a : J o u r n a l Of A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

Table 1. Concentrations ofbenomyl and MBC inWP formulations as determined by STB and spectrophotometric
(SP) methods3
Benomyl, % MBC, %

Sam ple No. S P (X ) S TB (V) Bias, % S P(X) S TB (Y ) Bias, %

1 32 .4 29 .6 +8.6 15.8 16.8 - 6 .3

2 28 .9 32 .0 -1 0 .6 15.2 15.3 - 0 .7

3 13.9 13.8 +0.7 29.1 29 .4 - 1 .0

4 37 .4 32 .0 + 14 .4 13.4 15.2 - 1 3 .4

Bias (%) = [ ( X - V)/X] x 100. Statistical param eters for bias for benomyl are as follows: m ean, +3.3; standard deviation, 10.9; Student’s f-test
0 .5 2. For M B C , statistical param eters are as follows: m ean, -5 .4 ; standard deviation, 6.9; Student’s f-test, -1 .5 6 . Th e value of Student’s
f-test at 95% confidence interval is 3.18.

ofthecationassociatedwithhydroxide. Quantitativeanalysis A p p lic a tio n o f M e t h o d f o r A n a l y s i s o f s o m e

of a WP sample carried out inpH 13 TMAH solutions pro­ C o m m e r c i a l W P F o r m u la tio n s
duced results similarto thatof analysis carried out in pH 13 The method was used toanalyze the intactconcentrations
sodiumhydroxidesolution.Therefore,theseexperimentsindi­ ofbenomyl andMBC inanumberofcommercialWP samples
catethatthekineticsoftheconversionofbenomyl toSTB de­ containing 50 and 10% benomyl (the 10% formulations also
pends only on the hydroxide ion concentration of alkali hy­ contained50% captan).These resultsarereportedinTable 2.
droxideratherthantheparticularcation.

E f f e c t o f S a m p l e S i z e - t o - S o d i u m H y d r o x id e S o lu tio n
R a tio o n A n a ly tic a l R e s u l t s
Table 2. Analysis ofbenomyl and MBC inWP
formulationscollectedfrom growers3
The effectofdifferentsolid (WP)-to-solution (sodium hy­ Benomyl +
Location in M B C (as
droxide)ratiosonthequantitativeanalysisofWP formulations S am ple No. the bag Benomyl, % MBC, % benomyl), %
was alsostudied.These studieswerecarriedoutby dissolving
50mg ofaWP samplein50,100,and250mL ofpH 13sodium 1 T 0.2 7.6 11.7
hydroxidesolution.Althoughnodifferenceinanalyticalresults D 0.3 7.4 11.5
was observed, becauseofthecomparativelyfasterdissolution 2 T 0.3 8.4 13.1
ofWP, a solid-to-solutionratioof50 mg/250 mL was used in D 2.0 7.8 13.8

thepresentmethod. 3 T 0 .3 8.4 13.1

D 0.3 8.4 13.1

R e p r o d u c ib ility 4 T 37.4 11.9 5 5 .5

D 40.6 9.2 54 .6

The reproducibilityofthemethod was determinedby ana­ 5 T 37.4 12.2 55 .9

D 40.2 10.2 55 .7
lyzinga50% WP samplein6duplicatesof50mg sampleeach.
6 T 40 .7 8.9 54 .2
The percent relative standard deviations (%RSDs) on an in­
D 43 .2 7.4 54 .4
trarunbasisfortheoverallprocedure(i.e.,weighinganddisso­
7 T 29.6 16.8 55.1
lutionofthesamplein0.125M sodiumhydroxide,dilution,and D 32.0 15.3 5 5 .2
LC) were0.7and2.2% forbenomyl andMBC determinations, 8 T 36.0 11.8 5 3 .9
respectively. D 36.5 11.0 5 3 .2
9 T 13.8 29.4 5 8 .4
A ccu racy D 32 .0 15.2 54 .9
10 T 48 .7 3.5 54 .0
The accuracyofthemethod was assessedby comparingthe D 49 .9 3.0 54 .4
results ofthe analysis ofsome commercial WP formulations
a Most samples at the tim e of collection had been stored from 4 to 7
with thoseobtainedby spectrophotometric method (5).These
years. S am ples were analyzed approximately 1 year after
resultsarereportedinTable 1.Becausethecalculatedvalueof collection. After collection, sam ples w ere kept in polyethylene
Studentt formean biasbetween the2 methods islessthanthe bags without proper seals at tem peratures ranging from 22 to
28°C . Sam ple 10 w as relatively new, only 1 year old at the tim e of
criticalvalueoft atthe95% confidence interval,thenullhy­ analysis. Sam ples 1 , 2 , and 3 contained 10% benomyl and 50%
pothesisisretained.Therefore,thebiasbetweenthe2 methods captan. Sam ples 4 to 10 were 50% benomyl formulations. T,
sam ple collected from the top layer of the formulation in the bag;
forthedetermination ofintactconcentrationsofbenomyl and D, sample collected 2 to 3 in. down from the top layer of the
MBC in WP formulations does not differ significantly from formulation in the bag.
zero.
 S in g h & C h ib a : J o u r n a l O f AO A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1191

o
CD

STB
2

UJ
if)
HI
if)
z Z
o
Q_ o
CL
if) if)
HI HI
CC CC
t
or CC
O

M BC

p
if)

JL — t— r
4
r
L
0 3 6 9 12 0369
MIN MIN

Figure3. Representative chromatogram ofa 50% WP Figure4. Representative chromatogram ofa 10% WP
formulation. formulation.

Representative chromatograms of the 50 and 10% formula­ WP formulationsthaninthe50% WP formulations.We found

tionsareshown inFigures3 and4. thatduringa6 month storageperiodundersimilarconditions,
As canbe seenfromTable2,allthesamplescontainedvari­ 63% oftheactivebenomylpresentinthe10% WP formulations
ablelevelsofMBC inadditiontobenomyl.Thisfindingshows was convertedtoMBC versusonly 17% oftheactivebenomyl
thatbenomyl isconvertedtoMBC duringstorage.The conver­ in the 50% formulations. Faster conversion of benomyl to
sionofbenomyl toMBC seems tobe influencedby exposure MBC inthe 10% WP formulations than inthe 50% formula­
tomoistureand air,becausethesamples takenfrom thetopof tionswas alsoreportedby Chiba (5).The samples reportedin
thebag showed more MBC thandidthesamples takenfrom2 Table 2 were 4 to7 years old,and the age ofsamples may be
to3 inches down from thetop.Analyticalresultsforsamples the main reason for the decomposition of up to 98% of the
9T and9D providefurtherevidenceofthiseffect.Thebagfrom activebenomyl toMBC inthe 10% WP formulations.
which these samples were taken was keptopen atthetime of The resultsreportedinthelastcolumnofTable2revealthat
samplecollection,andbecauseofexposuretomoistureandair, the50and 10% WP formulationscontained55.0± 1.2and 12.7
more benomyl inthetoplayerwas converted toMBC. These ±0.9% benomyl,respectively.Theseresultsimplythatboththe
resultssuggestthatafteruse any remainingproduct shouldbe formulations were overformulated. Although overestimation
storedinair-tightbags and incoldplacestokeep benomyl in­ ofbenomylinWP formulationscanbecausedbyexperimental
tact. error,inview of similarconclusions drawn by Chiba (5)and
To calculate the total content of benomyl and MBC as TeubertandStringham(1,2)theseformulationsmay wellhave
benomyl in a WP formulation, the MBC concentrations re­ been overformulated. We would like to emphasize here that
portedincolumn 4 ofTable2 wereconvertedtobenomyl con­ determination of individual concentrations of benomyl and
centrationsby multiplyingwith thefactor 1.52,and adding to MBC, reported in thepresent work, was not possible by the
benomyl concentrationsreportedincolumn 3 (Table2).These
method ofTeubertand Stringham(1,2).
resultsarereportedinthelastcolumn inTable2.The valuesof
totalconcentrationsofbenomyl thuscalculatedforsamples T
(takenfrom thetop)andD (taken2 to3 inchesdown fromthe Conclusion
top)from thesame bag matched closely(thebiasbetween the
2 values was lessthan 2%, exceptforsamples 2 and 9).This We developedasimpleRP-LC methodforthesimultaneous
resultindicatesthattheinitialdistributionofbenomyl withina determinationofbenomyl andMBC inWP formulations. The
bag isquiteuniform. However, ifasample iskeptopen foran method has theabilitytoselectivelyconvertbenomyl toSTB
extendedperiod,itisnecessarytomix wellthecontentsofthe while keeping intact the MBC present in the sample. The
bagtoobtainarepresentativevalue. method isreproducible, accurate, and suitableforroutinede­
The analyticalresultsforthe 10% WP formulations reveal terminations. Further evaluation toward the establishment of
thatup to98% ofthebenomyl intheformulationdecomposed anofficialAO AC methodmay becarriedoutthroughcollabo­
toMBC. ConversionofbenomyltoMBC wasfasterinthe10% rativestudies.
1192 S in g h & C h ib a : J o u r n a l Of AO AC I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

References (7) Hanson,D.J.(1992)Chem. Eng. N ew s April6,15-17

(8) Singh,R.P.,& Chiba,M. (1985)J. Agric. F ood Chem. 33,
(1) Stringham,R.W.,& Teubert,W.E.(1984)J. Assoc. Off. Anal. 63-67
Chem. 67,302-303
(2) Teubert,W.E.,& Stringham,R.W. (1984)J. A ssoc. Off. Anal. (9) Chiba,M.,& Singh,R.R (1986)J. Agric. F ood Chem. 34,
Chem. 67,303-305 108-112
(3) Chiba,M.,& Cherniak,E.A.(1978)J. Agric. F ood Chem. (10) Spencer,E.Y.(1981)Guide to the Chem icals U sed in Crop
26,573-576 Protection, Publication 1093,7thEd.,ResearchBranch,Agri­
(4) Chiba,M. (1977)J. Agric. F ood Chem. 25, 368-373 cultureCanada
(5) Chiba,M. (1979)J. A ssoc. Off. Anal. Chem. 62,488-493 (11) Calmon,J.P.,& Sayag,D.R.(1976)J. Agric. F ood Chem. 24,
(6) Hall,D.J.(1980)Proc. FI. State H orde. Soc. 93,341-344 314-317
 Please check one from each category below:
JOURNAL OF AOAC INTERNATIONAL
NOVEMBER 1993
ThisinquirycardvaliduntilMAY 1994 PRIMARY FIELD OF WORK: □D Sales/Service/Mkt.
ADVERTISEDPRODUCTS: □ A. Food □E Other_____ ,____
1 2 3 4 5 6
□B Feeds

To 7

18

29
8

19
30
9

20

31
10
21

32
11

22

33
12

23
34
13

24

35
14

25
36
15

26

37
16

27

38
17

28

39
□ C. Fertilizers
□ D. Pharmaceuticals
□ E. Cosmetics
PRIMARY AREA OF EMPLOYMEN
INDUSTRIAL
□ A. Research/Development
40 41 42 43 44 45 46 47 48 49 50
□ F. □ B. Quality Control

R e c e iv e
Environment
51 52 53 54 55 56 57 58 59 60 61
□ G. Veterinary and/or animal science GOVERNMENT
62 63 64 65 66 67 68 69 70 71 72
□ H. Forensics □ C. Research/Development
73 74 75 76 77 78 79 80 81 82
84 85 86 87 88 89 90 91 92 93
83
94
□ I. Toxicology LI D Regulate/lnvestigate
□ J. General Analytical Chemistry COLLEGE/UNIVERSITY
95 96 97 98 99 100 101 102 103 104 105
□K General Microbiology □ E. Research/Development
□ L. Medical □ F. Teaching
NEWPRODUCTS:

FREE ; 301 302 303 304 305 306 307 308 309 310 □ M. Other INDEPENDENT/CONSULTING
311 312 313 314 315 316 317 318 319 320 □ G Research/Development
321 322 323 324 325 326 327 328 329 330 PRIMARY JOB FUNCTION: LJ H. Analysis/Testing
331 332 333 334 335 336 337 338 339 340 □ A. Laboratory Director/Manager MEDICAL
341 342 343 344 345 346 347 348 349 350 □ B. Laboratory Scientist □ I Research/Development
351 352 353 354 355 356 357 358 359 360 □ C. Purchasing □ J Climcal/Diagnostic
361 362 363 364 365 366 367 368 369 370
371 372 373 374 375 376 377 378 379 380
381 382 383 384 385 386 387 388 389 390 NAME: .
391 392 393 394 395 396 397 398 399 400
401 402 403 404 405 406 407 408 409 410 TITLE: .
â i m 411 412 413 414 415 416 417 418 419 420
ORGANIZATION:.
421 422 423 424 425 426 427 428 429 430

AOAC
432 433 434 435 436 437 438 439 440
431 STREET:_____
441 442 443 444 445 446 447 448 449 450
451 452 453 454 455 456 457 458 459 460 CPF:____
INTERNATIONAL 461 462 463 464 465 466 467 468 469 470
471 472 473 474 475 476 477 478 479 480 STATE or PROVINCE. .POSTAL CODE.
481 482 483 484 485 486 487 488 489 490
COUNTRY_________

’ublications
491 492 493 494 495 496 497 498 499 500

PHONE:__________

Catalog
F o r F re e in fo r m a tio n

C ir c le O n P ro d u c ts A d v e r t is e d in T h is Is s u e

U se T h e se R e a d e r S e r v ic e C a rd s
500 Circle corresponding number
Fill out and send card...
on And you’ll soon receive the data requested.

C a rd

!
Please check one from each category below:
JOURNAL OF AOAC INTERNATIONAL
NOVEMBER 1993 PRIMARY FIELD OF WORK: [ J D. Sales/Service/Mkt.
ThisinquirycardvaliduntilMAY 1994 □ A Food n e . Other__________
ADVERTISEDPRODUCTS: 1 2 3 4 5 6 □ B Feeds
PRIMARY AREA OF EMPLOYMENT
7 8 9 10 11 12 13 14 15 16 17 n c Fertilizers
Pharmaceuticals INDUSTRIAL
18 19 20 21 22 23 24 25 26 27 28 □ D
□ A Research/Development
29 30 31 32 33 34 35 36 37 38 39 □ E Cosmetics
H B. Quality Control
40 41 42 43 44 45 46 47 48 49 50 □ F Environment
Veterinary and/or animal science GOVERNMENT
51 52 53 54 55 56 57 58 59 60 61 n g .

72 Forensics □ C. Research/Development
62 63 64 65 66 67 68 69 70 71 □ H
81 82 Toxicology □ D. Regulate/lnvestigate
73 74 75 76 77 78 79 80 83 a I.
COLLEGE/UNIVERSITY
84 85 86 87 88 89 90 91 92 93 94 □ j . General Analytical Chemistry
□ E. Research/Development
95 96 97 98 99 100 101 102 103 104 105 n k . General Microbiology
□ F. Teaching
□ L. Medical
INDEPENDENT/CONSULTING
NEW PRODUCTS: □ M. Other
□ G. Research/Development
308 309 310
301 302 303 304 305 306 307
PRIMARY JOB FUNCTION: □ H Analysis/Testing
314 315 316 317 318 319 320
311 312 313
330
□ A. Laboratory Director/Manager MEDICAL
323 324 326 326 327 328 329
321 322
338 339 340 □ B. Laboratory Scientist □ I. Research/Development
332 333 334 335 336 337
331
344 345 346 347 348 349 350 □ C. Purchasing □ J. Clinical/Diagnostic
341 342 343
354 355 356 357 358 359 360
351 352 353
364 365 366 367 368 369 370
361 362 363
373 374 375 376 377 378 379 380 NAME:
371 372
384 385 386 387 388 389 390
383
381 382
400
TITLE:
392 393 394 395 396 397 398 399
391
402 403 404 405 406 407 408 409 410 ORGANIZATION:-.
401
414 415 416 417 418 419 420
411 412 413
421 422 423 424 425 426 427 428 429 430 STREET: ______
434 435 436 437 438 439 440
431 432 433
447 448 449 450 CITY:_________
441 442 443 444 445 446

451 452 453 454 455 456 457 458 459 460
STATE or PROVINCE. .POSTAL CODE.
464 465 466 467 468 469 470
461 462 463
471 472 473 474 475 476 477 478 479 480 COUNTRY_________
482 483 484 485 486 487 488 489 490
481
493 494 495 496 497 498 499 500 PHONE:__________
491 492
 NO POSTAGE
NECESSARY
IF MAILED
IN THE
UNITED STATES

B U S IN E S S R E P L Y M A IL
FIRST CLASS • PERMIT NO. 2758 • ARLINGTON, VA

POSTAGE WILL BE PAID BY ADDRESSEE

MARKETING COORDINATOR
JOURNAL OF AOAC INTERNATIONAL
2200 WILSON BLVD SUITE 400
ARLINGTON VA 22201-9907 USA

L I.U .I..I,III..... Ill,LI.M l...I.

F o r F re e I n fo r m a t io n

O n P ro d u c ts A d v e r tis e d in T h is Is s u e

U se T h e se R e a d e r S e r v ic e C a rd s
Circle corresponding n u m b e r
Fill out and send card...
A n d y o u ’ll soon receive the data requested.

NO POSTAGE
NECESSARY
IF MAILED
IN THE
UNITED STATES

B U S IN E S S R E P L Y M A IL
FIRST CLASS • PERMIT NO. 2758 • ARLINGTON, VA

POSTAGE WILL BE PAID BY ADDRESSEE

MARKETING COORDINATOR
JOURNAL OF AOAC INTERNATIONAL
2200 WILSON BLVD SUITE 400
ARLINGTON VA 22201-9907 USA
 C hen & G a o : Journal O f AO A C International V ol . 76, N o. 6 ,1 9 9 3 1193

CHEMICAL C O N T A M I N A N T S MONITORING

T h e C h in e s e T o t a l D i e t S t u d y i n 1 9 9 0 . P a r t I . C h e m ic a l
C o n ta m in a n ts

JuN sm C h en a n d J u nq uan G ao
InstituteofNutritionand Food Hygiene, Chinese Academy ofPreventiveMedicine, 29 Nan Wei Rd, Beijing 100050, China

T h e C h i n e s e to ta l d ie t s t u d y in 1 9 9 0 e s t im a t e d th e inthegeneralpopulation.Thispaperwillmainlycoverthegen­
d ie t a r y in t a k e o f 2 4 c h e m ic a l c o n t a m i n a n t s a n d 7 2 eralmethodologyofthestudyandtheresultson chemicalcon-
n u t r ie n t s fr o m 4 m a r k e t b a s k e t s c o lle c t e d a n d p r e ­ taminants.Theresultson nutrientsarereportedinacompanion
p a r e d in 12 p r o v in c e s . T w e lv e f o o d g r o u p c o m p o s ­ paper(2).
ite s w e r e m a d e fo r e a c h r e g io n a l m a r k e t b a s k e t.
T h e o v e r a ll d ie ta r y P b , C d , H g , h e x a c h lo r o c y c lo h e x - E x p e r im e n t a l
a n e H C H , a n d D D T in t a k e s w e r e w e ll b e lo w th e ir
c o r r e s p o n d i n g a c c e p t a b le d a ily in t a k e s . H o w e v e r,
The food composite approach was used to study the total
th e P b c o n t e n t o f e g g s fr o m t h e 2 s o u t h e r n r e g io n s
dietin4 regionalmarketbasketsrepresentingtheaveragedie­
e x c e e d e d th e t o le r a n c e lim it. T h e H g c o n t e n t o f le g ­
tarypatternsofdifferentgeographicalareasinmainlandChina.
u m e s j u s t r e a c h e d th e t o le r a n c e lim it, a n d H g in
As shown inFigure 1,eachregioncomprised3provinces.The
e g g s fr o m th e N o r t h 1 r e g io n e x c e e d e d th e to le r ­
followingunifiedprocedureswerefollowedthroughouttheen­
a n c e lim it. T h e d ie t a r y H C H in t a k e h a s d e c r e a s e d tirestudy (Figure2).To minimize theeffectsofseasonalvari­
s ig n if ic a n t ly s i n c e th e 1 9 8 0 s, b u t d ie ta r y D D T in ­ ationinthefoodsupply,studiesinthe2 southernregionswere
ta k e h a s d e c r e a s e d ra th e r s lo w ly . F iv e o r g a n o p h o s - carried out inthe spring (May and June) and studies in the 2
p h o r u s p e s t ic i d e s w e r e d e te c t e d o u t o f a to ta l o f 12 northernregionswere carriedoutinfall(Septemberand Octo­
o r g a n o p h o s p h o r u s p e s t ic i d e s a n a ly z e d . A m o n g ber).
th e m , m e t h a m i d o p h o s w a s t h e m o s t o u t s t a n d in g .
T h e in ta k e o f to ta l c o m m it t e d d o s e e q u iv a le n t s Food Consum ption Survey
( C D E s ) o f th e 6 r a d i o n u c li d e s w a s 0 .2 4 m S v / a ; o n ly
1 .5 % w a s a c c o u n t e d f o r b y ^ S r a n d 137C s . 210P b , The food consumption pattern ineach ofthe 12 provinces
210P o , ^ R a , a n d 228R a a c c o u n t e d f o r 9 8 .5 % o f th e wasdeterminedby a3-dayhouseholddietarysurveythatdocu­
to ta l C D E s . T h e m a in f o o d s o u r c e s o f t h o s e ra ­ mented allthefoodconsumed inthehouseholdsin3daysby a
d i o n u c li d e s w e r e c e r e a ls , v e g e t a b le s , a n d a q u a t ic weighingandrecordingmethod (3).Three surveysites(2rural
f o o d s . A f la t o x in B i w a s d e te c t e d a t v e r y lo w le v e ls and 1urban) atmedium economic levelswere selectedtorep­
o n ly in th e c e r e a l c o m p o s it e o f th e N o r t h 1 r e g io n . resent the dietary pattern ofeach province, because approxi­
A f la t o x in M i w a s n o t d e te c t e d in a n y o f th e m ilk mately two-thirdsoftheChinese populationresideinruralar­
a n d m ilk p r o d u c t s . T h e o v e r a ll r e s u lt s s h o w th a t eas. Ineachsurveysite,30householdswererandomlyselected
th e re i s n o s i g n i f i c a n t e n v ir o n m e n t a l c o n t a m i n a ­ and the average food consumption ofa standard man (18-45
tio n o f th e a v e r a g e C h i n e s e diet. years old,60-kg body weight, tightphysicalactivity)foreach
household was calculated from thetotalhousehold food con­
sumption.The averagefoodconsumptionofthe90households
T he totaldietstudyisrecommendedby theWorld Health was thenusedasthestandardfoodconsumptionpatternforthe
Organizationtoestimatedietaryintakeofcontaminants province.
and nutrients by analyzing a standard set of prepared
Food Aggregation
foodsorfood groups toreflecttheaverage dietaryhabitofthe
generalpopulationoraparticularsubsetofthepopulation (1).
Allthefoods(includingwaterandbeverages)consumed by
The firsttotaldietstudyinChina, which was carriedoutin
the standard man in each province were aggregated into 13
1990,estimatedthedietaryintakeof24chemicalcontaminants
groups (Table 1). The food items consumed in significant
and 72 nutrientsfrom food samples collectedand preparedin
amounts arelistedinAppendix la.Any fooditemconsumed at
12provinces.Thepurposeofthestudywastoassessandmoni­ lessthan 1% ofthetotalfoodconsumption (by weight)ofthis
torthesafetyandnutritionalqualityoftheaverageChinesediet foodgroupwas combined intoasimilarorcloselyrelatedfood
item.A foodlistwas formulatedforthefoodsamplecollection
Received November 24, 1992. Accepted M arch 4, 1993. ineach province.
1194 Chen & G a o : Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3

diments were usedduringpreparationand cooking offoodsin

othergroups.The preparedfoodswerethenblendedtoformthe
respective group composites (Table 1) with weights propor­
tional to the average daily consumption for the province.
Therefore, 12 food group composites were availableforeach
province.
Sample Shipm ent

These provincial composites were frozen at -30°C and

shippedtothecentrallaboratoryattheInstituteofNutritionand
Food Hygiene inBeijing.There, thecomposites were further
mixed toformthe4 regionalmarketbaskets(Figure2)accord­
ingtotheircorrespondingweightproportioninfoodconsump­
tion.
Sample Preparation

lj§ l N o r th 1 H e ilo n g jia n g L ia o n in g H e ib e i Dietary survey and food sampling and preparation (cook­
■ H N o r th 2 H enan S haanxi N in g x ia ing) were carried out at the local level by provincial health
H I S o u th 1 S hanghai J ia n g x i F u jia n
teamsaccordingtounifiedprotocolsandprocedures.The final
W M iM i S o u th 2 H ubei S ic h u a n G uangxi
food composite preparation was carriedout atthe Instituteof
Nutrition and Food Hygiene. Samples were stored in 1 kg,
F igu re 1. S tu d y s ite s o f 1990 total diet s tu d y in C hin a. high-pressure,polyethylenecontainers(washedwithacidsand
nonionizedwater)at-30°Cbeforeanalysis.Samplesforheavy
metalsanalysiswere freeze-dried(—403C,72 h,760 mm Hg).
Food Sam pling and Preparation
Sample Analysis and Q uality Assurance
Food samples were collected from 4 local food markets, LaboratoryanalyseswerecarriedoutattheInstituteofNu­
grocery stores, and rural households in each of the 3 survey tritionandFood Hygiene,excepttheanalysisofradionuclides,
sitesineachprovinceandthencookedandpreparedaccording whichwas carriedoutattheInstituteofRadiationMedicine of
tothelocalfoodhabitsofeachprovince.Cookingoilandcon- theChinese Academy ofMedical Sciences. The final48 food

F igu re 2. Unified p ro ce d u re fo r total diet s tu d y in C h in a (with N orth I m arket b a sk e t a s an exam ple).

 C hen & G a o : Journal O f AO A C International V ol . 76, N o. 6 ,1 9 9 3 1195

T able 1. F o o d c o n s u m p tio n (g/d ay)a in the 4 re g io n s in the C h in e s e total diet s tu d y

Food group North 1 North 2 South 1 South 2 Average %b

Cereals and cereal products 424.6 526.1 440.7 454.3 461.4 26.5
Legumes, nuts, and their products 36.1 41.8 38.8 41.3 39.5 2.3
Potatoes and potato products 79.7 151.1 16.0 157.3 101.0 5.8
Meats and meat products 36.1 36.4 59.3 63.6 48.9 2.8
Eggs and egg products 20.8 14.1 17.4 16.0 17.1 1.0
Aquatic foods and aquatic food
products'7 42.2 9.3 27.0 12.9 22.9 1.3
Milk and milk products 14.6 19.1 6.1 4.2 11.0 0.6
Vegetables and vegetable products 348.2 424.9 234.4 288.3 323.9 18.6
Fruits and fruit products 205.5 189.1 8.9 1.0 101.2 5.8
Sugar 2.3 3.2 5.2 3.0 3.4 0.2
Beverages and water 626.1 464.5 311.6 646.1 512.1 29.4
Alcohol beverages 16.9 3.6 16.5 19.1 14.0 0.8
Condiments and cooking oils 104.5 82.1 77.0 71.6 83.9 4.9
Total 1957.7 J965.2 1259.2 1778.5 1740.2 (100.0)

a Amount of edible part before cooking and preparation.

b Percentage of total amount consumed.
c Including fish, shrimp, and oyster.

composite samples (12 composites x 4 market baskets) were Reporting o f Results

analyzed for a wide variety ofchemical contaminants by the
methodsdescribedinTable2.Dataofanalyticalmethodevalu­ This study defined the limitofdetectionas 3 standard de­
ationarelistedinAppendix lb.Referencematerialswereused viations(SDs) oftheblank and thelimitofquantitationas 10
foranalyticalqualityassuranceofheavymetals(AppendixIc). SDs oftheblank(1,4).The levelsofintakeofallthechemical
contaminants and nutrientsofa standardman foreach region
were estimatedaccordingtotheconcentrationsoftheanalytes
inthefood composites and thecorrespondingfood consump­
tion.Theseresultswerecompared withtheacceptabedailyin­
T able 2. A n a ly te s a n d an alytical m e th o d s for takes (ADIs) of chemical contaminants (5-8) and recom­
d eterm in in g ch e m ic al c o n ta m in a n ts in C h in e s e total diet
mended dailyintakes(RDAs) ofnutrients(9).
s tu d y in 1990

Analytical Results and Discussion

Analyte method3 Ref.

Food Consumption Patterns

Heavy metals
Lead, cadmium GFAAS 19 The food consumption patternsinthe4 regions areshown
Mercury CVAAS 19 in Table 1.In general, the average Chinese dietin 1990 was
Pesticides mainly composed ofplantfood (81.5%), such ascereals,leg­
Organochlorine pesticides umes, potatoes,and vegetables,ifwaterandbeverages areex­
HCH, DDT GC 20
cluded.Animal food(meat,eggs,fish,andmilk)accountedfor
Organophosphorus pesticides
Methamidophos, dichlorvos,
only asmall proportion(8.0%) ofthetotalfood consumption.
parathion, parathion-methyl, Consistent with theregional levelofeconomic development,
trichlofon, dimethoate, totalfoodconsumptionwas higherinthenorthernregionsthan
acephate, disulfoton, in the southern regions, and animal food consumption was
fenitrothion, malathion, higherinthesouthernregionsthaninthenorthernregions.
fenthion, phosmet Capillary GC 21
Aflatox ns (B-j, M-,) LC 22 Heavy M etals
Radionuclides
^Sr, 137Cs, 210Pb, ^ P o , 210Ra, BC 23 Three heavy metals, lead, cadmium, and mercury, were
228Ra AS 24 measured. The overall dietary intakes of these heavy metals
were wellbelow theircorresponding ADIs orprovisional tol­
a GFAAS, graphite furnace atomic absorption spectrometry:
CVAAS, cold vapor atomic absorption spectrometry; GC, gas
erable weekly intakes (PTWIs) recommended by the World
chromatography; LC, liquid chromatography; BC, beta cojnting Health Organization (Table 3). Some geographical variations
apparatus; AS, alpha spectroscopy. inPb intakewerenoted;thatis,North2regionhadarelatively
1196 C hen & G a o : Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3

T a b le 3 . H e a v y m e ta l in ta k e (p g p e r p e r s o n p e r d a y ) in C h in e s e to ta l d ie t s t u d y

Acceptable Dietary intake (% ADI)

Population daily -------------------------------------------------------------------------------------------------------
Heavy metal sampled intake3 North 1 North 2 South 1 South 2 Mean

Pb Adults 429 70.7 (16.5) 166.2 (38.7) 72.7 (16.9) 35.5 (8.3) 86.3 (20.1)
Children 62b 47.1 (75.9) 110.8 (178.3) 48.5 (78.0) 23.7 (38.2) 57.5 (92.6)
Cd Adults 60 10.2 (17.0) 9.3 (15.4) 19.1 (31.8) 16.4 (27.4) 13.8 (22.9)
Hg Adults 43 12.3 (28.5) 9.3 (21.6) 8.6 (20.1) 10.9 (25.5) 10.3 (23.9)

a Acceptable daily intake (pg per person per day) was calculated from the provisional tolerable weekly intake (PTWI) proposed by FAO/WHO
(50 pg/kg body weight per week for Pb, 7 pg/kg body weight per week for Cd, and 5 pg/kg body weight per week for total Hg) (3, 4) and
based on 60 kg body weight for male adult.
b Based on the proportion of energy intake of 5-year-old children to male adult. ADI (pg per person per day) = 25 pg/kg body weight per week
for child/7 days x 17.4 kg body weight (4).

T able 4. C o m p a r is o n o f dietary h e avy m etal intake (pg high intakeofPb and South 2 region had a low intakeofPb.
per p e rso n per d a y ) fro m total diet stu d y in v a rio u s However, ifthePb intakeofpreschoolchildren(<5 yearsold)
c o u n trie s (adult)3 is estimated on the basis of the ratioof theirdietary energy
Country (year of intaketothatofadults,theaverage intakeinthe4 regionswas
study) Lead Cadmium Mercury ashighas96% oftheADI (Table3),and thehighestintakein
theNorth 2 regionwas outstanding.Because ofthesensitivity
China (1990) 86.3 13.8 10.3 ofyoung children’scentralnervous systemtoPb toxicity,fur­
Japan (1988) 85.0 29.0 —
ther investigation and monitoring are necessary to lower the
United States
13.0 3.2
exposureofchildrentoPb.The intakesofheavy metalsinthe
(1987-1988) 9.8
United Kingdom
Chinese diet are still much higher than those of developed
(1988) 60.5 18.9 — countries but lower than estimated global intakes (Table 4)
Sweden (1988) 17.0 12.0 1.8 (10-11).
Netherlands Consistentwiththeaveragefoodconsumptionpattern,most
(1986-1988) 47.0 22.9 9.0 oftheintakesofPb and Cd were from cerealsand vegetables
Finland (1986) 15.8 11.3 1.9 (Table5).However, significantproportionsofHg intakewere
Australia (1987) 182.9 28.6 —
fromcereals(54.3%),aquaticfoods (9.7%;fish,shrimp,etc.),
Guatemala
(1986) 254.0 29.0 10.8
andvegetables(8.7%).The foodsourceofHg isunknown and
Global (1991)" 153.0 25.0 — needs furtherinvestigation.No significantgeographical vari­
ationswere notedon thefoodsourcesofthese3heavy metals.
a From reference 10. When theconcentrations(mg/kg) ofPb,Cd, and Hg inthe
b From reference 11. 12foodcomposites(asraw materials)werecompared withthe
tolerance limits in the Chinese National Hygienic Standards

T able 5. C o n trib u tio n (% o f d aily intake) o f v a rio u s fo o d g r o u p s to dietary in tak e s o f Pb, C d , a n d H g

Heavy Legumes Aquatic Beverages Alcoholic

metal Cereals and nuts Potatoes Meats Eggs foods Milk Vegetables Fruits Sugar and water beverages

Pb 39.6 4.4 7.8 5.6 5.1 2.8 0.3 27.5 5.4 0 1.3 0.1
Cd 54.3 4.3 6.5 4.3 0.7 3.6 0.7 23.9 2.9 0 0 0
Hg 54.4 3.9 6.8 5.8 5.8 9.7 0 8.7 5.8 0 0 0.3

T able 6. C o m p a r is o n o f dietary o rg a n o c h lo rin e p e sticid e in tak es with A D Ia

Dietary intake (% ADI)

Pesticide intake North 1 North 2 South 1 South 2 Mean

DDT 1200 22.8 (1.9) 4.2 (0.4) 8.8 (0.7) 46.1 (3.8) 20.5(1.7)
HCH _b 5.9 6.2 3.7 4.4 5.0

a ADIs are expressed as pg per person per day based on FAO/WHO ADI(5) 20 pg/kg body weight x 60 kg body weight.
b No ADI is available for total HCH.
 C hen & G ao : Journal O f AO AC International V ol . 76, N o. 6 ,1 9 9 3 1197

T able 7. C o n trib u tio n ( % o f d a ily intake) o f v a rio u s fo o d g r o u p s to dietary in tak es of H C H a n d D D T

Legumes
Pesticide3 Cereals and nuts Potatoes Meats Eggs Aquatic foods Milk Vegetables Fruits

HCH 40.9 3.0 3.0 20.5 6.2 4.7 0.8 15.2 5.8
DDT 2.6 0.4 0.6 56.9 3.4 26.2 0.1 8.5 1.3

3 Total HCH and total DDT.

Pesticide Residues
Table 8. C o m p a r is o n o f dietary total H C H an d total D D T
intake (pg per p e rso n per d a y ) in v a rio u s c o u n trie s3
Oiganochlorine pesticides, particularly DDT (dichlo-
Country (year of study) Total HCH Total DDT rodiphenyltrichloroethane) and HCH (hexachlorocyclohex-
ane),werethemajorpesticidesusedinChinabeforetheywere
China (1990) 5.04 20A7 banned in 1983. Although thedietary intakeofHCH has de­
Japan (1988) 0.88 1.30
creased significantly from 150 pg per person per day from
United States (1988) 0.16 1.30
Australia (1987)
1973 through 1978 to 5 pg per person per day in 1990, the
1.95
dietary intakeof DDT has decreased ratherslowly from 41.4
3 From reference 10. pg perpersonperday from 1973 through 1978 to20.5 pg per
personperdayin1990(15).The South2regionhadthehighest
DDT intake (Table 6).The main food sources of DDT were
(Appendixlia;12-14),thePbcontentofeggsandeggproducts meat(andmeatproducts)andaquaticfoods,andthoseofHCH
from the2 southernregionsboth exceeded thetolerancelimit were cereals,meats, and vegetables (Table 7).The DDT and
(0.2mg/kg) by 207.5 and 3.7% respectively;thus,theaverage HCH intakeintheChinesedietarestillmuch higherthanthose
national content exceeded the tolerance limitby 29.1%. The in developed countries (Table 8) (10). The concentrations
high Pb intake may be caused by consumption ofpreserved (pg/kg) of DDT and HCH in the 9 food composites (asraw
eggs,whichhadahighPb content.Hg concentrationsinallthe materials)arelistedinAppendix lib.The isomersandmetabo­
compositesofthe4 regionswere wellbelowthetolerancelim­ litesofHCH and DDT are listedinTables 9 and 10.In plant
its,exceptthatlegumes(andlegumeproducts)reachedthetol­ foods,theorderofoccurrenceofHCH isomerswas a>y>P>8,
erance limit (0.1 mg/kg) and eggs (and egg products) in the whichwasconsistentwiththeconrespondingproportionsinthe
North 1 region exceeded the tolerance limit (0.05 mg/kg) by commercial products of HCH. On the other hand, in animal
88.9%. The sources of these contaminations are unknown. foods,theorderofoccurrenceofHCH isomerswas P>a>y>5.
However, because eggs and legumes accounted for only a p , p -DDE isthemain form ofDDT inallthefood groups,but
small proportion ofthe diet, theircontribution to the overall p.p'-DDT and p , p '-DDE are the main forms of DDT in un-
intakeofPb andHg isoflittlesignificance.
T able 10. R a tio s (% ) o f D D T is o m e r s an d m etabo lites
T able 9. R a tio s (% ) o f H C H is o m e r s a n d m etabo lites b y b y fo o d g r o u p s 3
fo o d g r o u p s 3
Food group p,p '-DDE o,p '-DDT p,p '-DDD p,p '-DDT
Food group ct-HCH ß-HCH y-HCH 5-HCH
Plant foods
Plant foods Cereals 100.0 0 T T
Cereals 64.2 13.5 22.3 0 Legumes
Legumes and nuts 60.9 T 7.3 31.7
and nuts 64.5 7.6 26.2 1.7 Potatoes 40.1 0 13.9 46.0
Potatoes 61.6 11.4 27.0 0 Vegetables 30.4 17.1 12.7 39.8
Vegetables 63.2 16.7 20.1 0 Fruits 37.0 46.6 2.9 13.5
Fruits 65.6 15.2 19.2 0 Combined 39.8 19.9 9.1 31.2
Combined 64.1 12.8 22.7 0.4 Animal foods
Animal foods Meats 41.5 0 27.2 31.3
Meats 49.8 44.1 6.1 0 Eggs 55.8 0 14.2 30.0
Eggs 29.1 62.1 7.5 1.3 Aquatic
Aquatic foods 39.5 0 54.4 6.1
foods 52.0 35.4 8.7 3.9 Milk 73.9 0 21.6 4.4
Milk 33.5 59.7 6.8 0 Combined 42.3 0 36.1 21.6
Combined 42.4 49.2 7.2 1.3
3 Values below the limit of detection are reported as zero(0). Values
3 Values below the limit of detection are reported as zero (0). below the I mit of quantitation are reported as trace (T).
1198 C hen & G ao : Journal O f AOAC International V ol 76, N o. 6 ,1 9 9 3

T a b le 11. C o m p a r is o n o f d ie ta r y o r g a n o p h o s p h o r u s p e s t i c id e in ta k e (u g p e r p e r s o n p e r d a y ) w ith A D Isa

Acceptable Dietary Intake (% ADI)

daily
Pesticide intake North 1 North 2 South 1 South 2 Mean

Dichlorvos 240 7.8 (3.3) 2.3 (1.0) 6.8 (2.8) 6.4 (2.7) 5.8 (2.4)
Methamidophos 240 26.7(11.1) 10.3 (4.3) 35.6 (14.8) 22.9 (9.5) 23.9 (10.0)
Parathion 300 1.2 (0.4) 0 0.02 (0.01) 0 0.3 (0.1)
Trichlorfon 600 0 0 0 11.5 (1.9) 2.9 (0.5)
Dimethoate 600 0 0 2.5 (0.4) 0 0.6 (0.1)

ADIs are expressed as gg per person per day [FAO/WHO ADIs (5) gg/kg body weight x 60kg body weight]. Values below the limit of
quantitation but above the limit of detection are reported as half the limit of quantitation, and values below the limit of detection are reported
as zero (0).

cookedfood(16).Inthisstudy,asignificantproportionofp , p - milk productcomposites from the4 regions.These resultsin­

DDD was found, formed fromp,//-DDT duringcooking (17). dicatethataflatoxincontaminationisnotamajor healthprob­
Five organophosphorus pesticides were detected out of a lem intheaverageChinese diet,althoughtheproblem ofafla­
total of 12 organophosphorus pesticides analyzed. Among toxincontaminationincom,peanuts,andpeanutoilshouldnot
them, methamidophos was widely found inall4 regions (Ap­ be overlookedinsome ofthesouthernprovinces.
pendix lie),especially in cereals, vegetables, and fruits (Ta­
ble 11).AccordingtotheNationalRegulationsontheSafeUse Radionuclides
ofPesticides(18),thispesticideisnotpermittedforuseonfood
The concentrations (Bq/kg) (asraw materials) ofradionu­
crops. In addition, dimethoate, dichlorvos, trichlorfon, and
clides in the 12 food composites are listed in Appendix Be.
parathion were detected incereals, vegetables, and fmits,al­
though theirintakeswere well below thecorrespondingADIs Among the 6 radionuclides studied, 210Pb, 210Po, 226Ra, and
(Table 12). 228Ra areindicatorsofnaturalradionuclide.The overallintake
The concentrations Qig/kg) of organophosphorus in the 3 of these radionuclides was quite low, although some geo­
mainfoodcomposites(asrawmaterials)arelistedinAppendix graphicaldifferences were noticeable (Table 14).The intakes
He. Acephate, disulfoton, fenitrothion, fenthion, malathion, inthesouthernregions were higherthanthose inthenorthern
parathion-methyl, and phosmet were not detected in samples regions,eventhoughthenorthernregionshadhighertotalfood
from allthe 4 regions. These results show thatthecontrol of consumption than thesouthernregions.90Srand 137Cs arein­
pesticide use and pesticide residue monitoring should be dicatorsofman-maderadionuclide;nosignificantradionuclide
strengthened. contaminationwas foundinallthefoodcompositesfromthe4
Details of the pesticide residues in thisChinese total diet regions. The totalcommitted dose equivalent (CDE) ofthe 6
studywere reportedby Wang etal(15). radionuclidesbyingestionwas0.24mSv/a,inwhichonly 1.5%
was due to90Sr and 137Cs. 210Pb, 210Po, 226Ra, and 228Ra ac­
Aflatoxins countedfor98.5% ofthetotalCDEs (Table15).The main food
The concentrations (pg/kg) (asraw materials) ofaflatoxin sources of those radionuclides were cereals, vegetables, and
B |andM ,inthe3mainfoodcompositesarelistedinAppendix aquatic foods (Table 16). Although the average totalCDE is
Ed. AflatoxinB, was detectedatvery low levels(1.41 pg/kg) very low, remember that certain people consuming greater
only inthecereal composite ofthe North 1region,which re­ quantities of food (especially those foods high in radionu­
sultedinan aflatoxinB1intakeforthisregion of0.60 pg/day clides)willhavehigherintakesoftheseradionuclides.
(Table 13).AflatoxinMj was notdetected in allthe milk and In comparison with the global exposure levelsreported in
1988, theChinese CDEs of210Pb, 210Po, 226Ra, and 228Ra are
T able 12. C o n trib u tio n (% of d aily intake) of v a rio u s somewhat higher (Table 15), perhaps because of the higher
fo o d g r o u p s to dietary intake o f o r g a n o p h o s p h o r u s concentrationsofthoriumradionuclidesinChinese waterand
p e s tic id e s 3 soil.
Pesticide Cereals Vegetables Fruits
Table 13. Dietary aflatoxin intake (p g per p e rso n per
Methamidophos 72.3 18.7 9.0 d a y ) in the 1990 C h in e s e total diet stu d y 3
Dichlorvos 60.0 36.3 3.7 Aflatoxin North 1 North 2 South 1 South 2 Mean
Dimethoate 0 95.7 4.3
Parathion 0 0 100.0 Aflatoxin B-, 0.60 0 0 0 0.15
Trichlorfon 0 100.0 0 Aflatoxin M1 0 0 0 0 0
3 Values below the limit of detection are reported as zero (0). 3 Values below the limit of detection are reported as zero (0).
 C h e n & G a o : J o u r n a l O f AO A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1199

T a b le 1 4 . C o m p a r is o n o f d ie ta r y r a d io n u c lid e in ta k e (B q p e r p e r s o n p e r y e a r ) w ith a n n u a l lim it in ta k e 3

Dietary intake (% ALI)

Radionuclide Annual limit intake North 1 North 2 South 1 South 2 Mean

90Sr 2.4 x104 47.1 (0.2) 33.4 (0.1) 157.1 (0.7) 132.2 (0.6) 92.4 (0.4)
137Cs 7.4 x104 5.6 (0.01) 11.4 (0.02) 43.3 (0.06) 29.5 (0.04) 22.5 (0.03)
210Pb 4.6 x102 87.0(18.9) 82.4(17.9) 130.4 (28.4) 136.5 (29.7) 109.1 (23.7)
210Po 2.2 x103 71.0 (3.2) 113.5 (5.2) 160.4 (7.3) 157.2 (7.1) 125.5 (5.7)
226Ra 1.5 x103 20.3 (1.4) 27.6 (1.8) 27.2 (1.8) 33.4 (2.2) 27.1 (1.8)
228Ra 1.7 x103 21.6 (1.3) 30.0 (1.8) 133.6 (7.9) 85.8 (5.1) 67.7 (4.0)

a Annual limit intake is expressed as Bq per person per year.

T able 15. C o m p a r is o n of dietary ra dio n u clide intake (pSv/year) o f total com m itted d o s e e q u iv a le n ts3 in 4 re g io n s an d
g lo b a l m e a n 6

% of daily mean
Radionuclide North 1 North 2 South 1 South 2 Mean intake Global mean

90Sr 1.7 1.2 5.7 4.8 3.4 1.4 _

137Cs 0.1 0.2 0.6 0.4 0.3 0.1 —
210Pb 121.8 115.3 182.6 191.1 152.7 63.0 56.0
210Po 31.2 50.0 70.6 69.2 55.3 22.8 17.6
226Ra 6.3 8.6 8.4 10.4 8.4 3.5 13.0
228Ra 7.1 9.9 44.1 28.3 22.4 9.2 7.0
Total 168.2 185.1 312.0 304.1 242.4 100.0 93.6

a Based on weighting total committed dose equivalent of radionuclides (25).

b From reference 25.

T able 16. C o n trib u tio n (% of d aily intake) of v a rio u s fo o d g r o u p s to dietary in tak es o f ra d io n u c lid e s

Legumes Aquatic Beverages Alcoholic

Radionuclide Cereals and nuts Potatoes Meats Eggs foods Milk Vegetables Fruits Sugar and water beverages

90Sr 39.4 5.3 11.3 2.0 0.9 3.9 0.1 34.8 1.2 0.02 1.0 0.03
137Cs 20.6 5.1 11.2 10.6 1.6 4.6 0.1 30.5 1.1 0.1 14.2 0.2
210Pb 15.3 1.5 0.5 4.0 2.4 35.7 0.1 39.2 0.4 0.01 0.9 0.02
210Po 18.1 1.1 0.3 2.0 1.7 35.7 0.04 40.1 0.3 0.004 0.8 0.01
226Ra 29.4 5.4 17.3 2.8 4.5 2.0 0.2 34.6 2.0 0.1 1.7 0.1
228Ra 27.7 5.9 10.9 3.9 2.0 5.4 0.1 39.7 1.0 0.03 3.3 0.04

Conclusions Thistotaldietstudybyitsnatureprovidesinformationonly
on the safety ofthe average dietofthe generalpopulation in
Results on the food content and dietary intakes of heavy China.Itshouldbefullyacknowledgedthattherearehugegeo­
metals,pesticides,radionuclides,andaflatoxinsobtainedfrom graphicaldifferencesinthechemicalcontaminationoffoodsin
thefirstChinesetotaldietstudycarriedoutin 1990 show that China.Therefore,thenegativeresultsobtainedfromthisstudy
the average Chinese dietwas not significantly contaminated do notexclude thepossibilityoflocalizedfoodcontamination
withtheseenvironmentalcontaminants.However, some prob­ problems.Thetotaldietstudyshouldbecomplementedby spe­
lems with considerable health significance were found. The cificsurveyscarriedoutinspecificregions.
majorfindingsaretherelativelyhighintakeofPb by children,
which shouldbe causeforalarmandpromptfurtherinvestiga­ Acknowledgments
tion,andthedetectionof5organophosphoruspesticidesinsev­
eral major food composites, that is,cereals, vegetables, and The authors express theirappreciationtothe 12provincial
fruits.Methamidophos should be considered as a major con­ healthteams who carriedoutallthefieldwork, aswell asthe
cern inthe safecontrol ofpesticideuse and pesticideresidue following scientistsintheInstituteofNutritionand Food Hy­
monitoring. giene ofthe Chinese Academy ofPreventive Medicine, who
1200 C h e n & G a o : J o u r n a l O f AO A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

were responsible for laboratory analysis of the various con­ (13) N ational Standard o f P eople’s Republic o f C hina (1984) G B n
taminants: Xuqing Wang (pesticides),Yan Su (heavy metals) 238-84, P e r m i t t e d L e v e l o f C a d m iu m in F o o d s , Standard
Nan Wu (aflatoxins),andHongdaZhu and ShouliangWang of Press, Beijing, C hina (in Chinese)
the Institute of Radiation Medicine, Chinese Academy of (14) N ational Standard o f People ’s Republic o f C hina (1981)
MedicalSciences(radionuclides).The authorsalsothankJohn G B 2762-81, P e r m i t t e d L e v e l o f M e r c u r y in F o o d s , Standard
W. Jones of the U.S. Food and Drug Administration forhis Press, B eijing, C hina (in Chinese)
adviceon thestudydesignandhisreviewofandcomments on (15) W ang, X., Lin, Y., Chen, H., & F eng, Y , (1993) J. H y g . R e s .
thispaper. 22, Supplem ent 1 on The C hinese Total D iet S tudy in 1990
(in Chinese)
R e fe re n c e s
(16) Institute o f H ealth, C hinese A cadem y o f M edical Sciences
(1981) C o m p ila tio n s o f S tu d ie s o n T o x ic ity a n d R e s i d u e s o f
(1) World Health Organization, Offset Publication (1985) No. P e s tic id e s in 1 9 7 8 - 1 9 8 0 , Institute o f H ealth, C hinese A cad­
87, Geneva, Switzerland em y o f M edical Sciences, B eijing, C hina (in C hinese)
(2) Chen, J„ & Gao, J. (1993) J. A O A C I n t. 76, 1206-1213
(17) D epartm ent o f Food H ygiene, Institute o f H ealth, Chinese
(3) Chen, J., Campbell, T.C., Li, J., & Peto, R. (1990) D ie t, L if e ­
A cadem y o f M edical Sciences (1978) C h in e s e J . P re v . M e d .
s ty le a n d D is e a s e M o r ta lity : A S u r v e y o f th e C h a r a c te r is tic s
2, 77 -8 1 (in C hinese)
o f 6 5 C h in e s e C o u n tie s , Oxford University Press, Oxford,
England, pp. 14-16 (18) T h e N a t io n a l R e g u l a ti o n s o n th e S a f e U s e o f P e s tic id e s
(4) Dokkum, W.V., & Vos, R.H.D. (1987) T o ta l D i e t S tu d y in (1982) , M inistry o f A griculture, A nim al H usbandry and
E u r o p e , EURO NUT Report 10, A C o n c e r te d A c tio n o n N u ­ A quaculture and M inistry o f Public H ealth, Beijing, China
tr itio n a n d H e a lth in th e E u r o p e a n C o m m u n ity , The (in Chinese)
Netherlands, pp. 65-80 (19) Su, Y., G ao, J., & W ang, H. (1993) J. H y g . R e s . 22 Supple­
(5) WHO Technical Report Series (1972) No. 505, Geneva, Swit­ m ent 1 on T he C hinese Total D iet Study in 1990 (in Chinese)
zerland
(20) Lin, Y., C hen, H., Feng, Y , & W ang, X . (1991) J. H y g . R e s .
(6) WHO Technical Report Series (1987) No.751, Geneva, Swit­
20, 3 4 -3 8 (in Chinese)
zerland
(7) FAOAVHO (1986) Codex Alimentarius Commission, Vol. (21) Li, H., & W ang, X. (1993) C h in e s e J . P re v . M e d . 27 (in C hi­
Xm-Ed. 2, Rome, Italy nese) (in press)
(8) WHO/IPCS/91.47 Pesticide Residues in Food— 1990, (22) Park, D., N esheim S., Trucksess, M .W ., Stack, M .E., & N ew ­
UPCS, Geneva, Switzerland ell, R. (1 9 9 0 ) ./. A s s o c . O ff. A n a l. C h e m . 73, 2 6 0 -2 6 6
(9) Chinese Nutrition Society (1989) A c ta N u tri. S in . 11,93-96
(23) N ational Standard o f P eople’s R epublic o f C hina (1991)
(in Chinese)
S t a n d a r d M e th o d s o f R a d io n u c lid e s in F o o d s , Standard
(10) FAO/UNEPAVHO (1991) GEMS: Food Contamination Press, B eijing, C hina (in Chinese)
Monitoring Programme, Summary of 1986-1988 Monitoring
Data, WHO, Geneva, Switzerland (24) N ational Standard o f P eople’s R epublic o f C hina (1985)
G B 4792-84, B a s ic H y g ie n ic S t a n d a r d f o r R a d io l o g i c a l P r o ­
(11) Gorchev, H.G. (1991) F o o d A d d itiv e s a n d C o n ta m in a n ts 8,
te c tio n , Standard Press, Beijing, C hina (in C hinese)
793-806
(12) National Standard of People’s Republic of China (1992) T o l­ (25) Z hang, J., & Zhu, H. (1989) R a d io n u c lid e s a n d T h e ir
e r a n c e L im it o f L e a d in F o o d , Standard Press, Beijing, China C a u s e d i n t e r n a l D o s e in C h in e s e F o o d s , C hinese E nviron­
(in Chinese) m ental Science P ress, B eijing, C hina, pp. 240 (in Chinese)
 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1201

A p p e n d ix I

Table la. M a in fo o d ite m s stu d ie d in 1990 C h in e s e total diet stu d y

Food group Food items

Cereals and cereal

products W h ea t flour, fine dried noodles, steam ed bud, bread, biscuits, moon cake, deep-fried twisted dough sticks, rice, rice
noodles, glutinous rice, com m eal, corn grit, millet, glutinous millet, Chinese sorghum
Legumes, nuts, and
their products Soybean curd, soybean milk film, soybean flour, soybean curd strip, dried bean curd, deep-fried bean curd,
soybean, soybean cheese (red), soybean cheese (pungent), mung bean, adzuki bean, dried kidney bean, dried
bean, mung bean sprout, soybean sprout, vermicelli, peanut kernel, sunflower seed, walnut kernel
Potatoes and potato
products Potato, sweet potato, taro, sw eet potato powder, potato powder noodles
M eats and m eat
products Pork, mutton, beef, grilled chicken, chicken, sausage, pork luncheon m eat, ham sausage, trotters, pork liver, pork
kidney, pork blood, beef liver, pork casing, bacon, stew pork, cooked pork, duck m eat
Eggs and egg products Chicken egg, duck egg, salted duck egg, limed duck egg
Aquatic foods and
aquatic food products C arp silver, small carp silver, carp, carp crucian, carp grass, hairtail, coilia, anchovy, yellow croaker, pomfret,
C hinese herring, eel, sardine, Spanish m ackerel, cod, salted fish, spitchcock, clams m eat, sea snail meat,
shrimp, dried shrimps, loach, frog m eat
Milk and milk products C ow milk powder, cow milk, sheep milk
Vegetables and
vegetable products C hinese cabbage, immature C hinese cabbage, cabbage, cucumber, pumpkin, serpent melon, Chinese w ax gourd,
sponge gourd, C hinese squash, rutabaga, C hinese chive, celery, spinach, am aranth, w ater spinach, greens,
sword bean, lentil, string bean, kidney bean, eggplant, tomato, green pepper, green chilli, mushroom (fresh),
forest mushroom (fresh), radish, sauerkraut, broad bean (fresh), pea, cauliflower, kelp, Chinese lettuce, garlic
sprouts, onion, leek, wild rice stem, pickles, lotus root, bam boo shoots, pumpkin young stem, sw eet potato leaf
bud, salted vegetables, salted cabbage, salted radish dried, edible fungus
Fruits and fruit products Apple, flowering crabapple, pear, peach, grape, banana, watermelon, persimmon, orange, mandarin orange,
pineapple, plum, Chinese date, w ater chestnut, strawberry, hawthom e
Sugar W hite sugar, brown sugar
Beverages and water Boiled water, running water, tea, ice sucker, soda
Alcoholic beverages Spirits, beer, wine, yellow wine, rice wine
Condiments and
cooking oils Peanut oil, soybean oil, cotton seed oil, sesam e oil, rape seed oil, lard, soy sauce, salt, vinegar, green onion, garlic,
ginger, sesam e butter, sw eet brown sauce, soybean paste, chilli paste, aniseed, wild pepper, gourmet powder,
starch, baking yeast
1202 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

T able lb. E va lu a tio n data of an alytical m e th o d s for ch e m ic al c o n ta m in a n ts in C h in e s e total diet stu d y

Recovery

Chemical Limit of detection, pg/kg M ean, % Range, % CV, %

H eavy metals
Lead 1.56 104 8 5 -1 1 2 11.5
Cadmium 0.0 3 97 9 2 -1 0 3 4.3
Mercury 0.11 96 8 5 -1 2 1 14.0
Organochlorides
DDTa 0 .1 3 -1 .0 0 — 8 2 -1 0 6 < 20.8
HCHb 0 .0 5 -0 .1 0 — 9 6 -1 0 3 <24.2
Organophosphorus
Dichlorvos 0.2 5 — 5 8 -7 5 <9.3
Trichlofon 0.25 — 4 5 -8 7 < 16.8
M etham idophos 0.5 0 — 6 1 -7 1 <9.8
Parathion 0.125 — 7 6 -1 0 9 <9.8
Dimethoate 0.125 — 8 2 -1 1 0 <7.9
Acephate 0.5 — 5 1 -6 5 < 11.0
Parathion-methyl 0.125 — 8 0 -1 0 8 <7.3
Disulfoton 0.125 — 5 1 -9 6 < 10.8
Fenitrothion 0.125 — 8 2 -1 0 7 <7.8
Fenthion 0.125 — 8 0 -1 0 5 <7.1
Malathion 0.125 — 6 0 -1 0 6 <10.7
Phosmet 0.2 5 — 9 1 -1 0 5 <9.8
Aflatoxins
Bi 0.5 81 .6 7 3 -8 8 5.1
Mi 0.5 92 .3 8 1 -9 9 7.6
Radionuclides
90Sr 16.0 m Bq/g ash 75.1 6 8 -8 5 4 1 .9
137Cs 13.0 m Bq/g ash 78 .0 6 1 -8 1 3 9 .8
21° p b 8 9 .0 pBq/kg 91 .0 8 0 -9 8 2 6 .9 -2 1 4 .9
210Po 4 1 .0 pBq/kg 91 .7 8 1 -1 0 1 9 .1 -3 7 1 .9
226Ra 2.2 mBq/g ash 85 .6 7 4 -9 3 19.7
228Ra 3 0 .0 mBq/g ash 85 .7 7 4 -9 1 4 5 .6

a p ,p '-D D T , o,p '-DDT, p ,p '-D D D , p ,p '-D D E .

b a -H C H , P-HCH, y H C H , 8-H C H .

T a b le Ic . A n a ly t ic a l q u a lit y a s s u r a n c e o f h e a v y m e t a l s in C h in e s e t o t a l d i e t s t u d y 9

N um ber of
Analyte R eference material determinations M ean ±S D , pg/kg Range, pg/kg Certified ± S D , pg/kg

Lead Spinach N B S 1570 5 1.05 ± 0 .0 6 1 .0 1 -1 .0 5 1.2 ± 0.2

Oyster N B S 156 6 5 0.4 5 ± 0 .0 3 0 .4 3 -0 .4 7 0 .4 8 ± 0.0 4
Cadmium Spinach N B S 1570 4 1.17 ± 0.0 8 1 .0 9 -1 .2 7 1.5
Mercury Spinach N B S 1570 5 0 .0 3 3 ± 0.0 04 0 .0 2 8 -0 .0 3 6 0.0 3 ± 0 .0 0 5
Oyster N B S 156 6 5 0 .0 5 5 ± 0.002 0 .0 5 4 -0 .0 5 8 0.0 57 ± 0 .0 1 5

3 NBS reference material w ere gifts from John Jones of the U .S. Food and Drug Administration.
 >
►d
a>
d
Q*•
X
HH

T able H a . C o n c e n tra tio n s (m g /k g ) of h e a v y m etal b y fo o d g r o u p s 3

H eavy Legum es and B everages Alcoholic

metal Cereals nuts Potatoes M eats Eggs Aquatic foods Milk V egetables Fruits Sugar and w ater beverag es

Lead

C
Average 0.0 72 0.0 97 0 .0 6 9 0.115 0 .2 5 8 b 0.0 98 0 .0 2 6 0.0 68 0.0 39 0 0.002 0.0 09

h e n
Range 0 0 1 7 -0 .1 3 9 0 .0 6 7 -0 .1 5 6 0 .0 5 2 -0 .0 9 0 0 .0 6 7 -0 .1 9 4 0 .0 8 1 -0 .6 1 5 0 .0 7 5 -0 .1 4 4 0 .0 2 -0 .0 3 3 0 .0 1 3 -0 .1 5 2 0 .0 2 4 -0 .0 6 8 0 .0 0 0 -0 .0 0 7 0 .0 0 0 -0 .0 1 9

&
C adm ium

G
Average 0 .0 1 7 0.015 0.011 0 .0 1 3 0 .0 0 6 0 .0 1 8 0.0 04 0.022 0.011 0 0 0

a o
Range 0 .0 0 9 -0 .0 3 0 0 .0 1 0 -0 .0 2 3 0 .0 0 7 -0 .0 1 3 0 .0 0 6 -0 .0 1 8 0 .0 0 0 -0 .0 1 5 0 .0 0 4 -0 .0 3 9 0 .0 0 1 -0 .0 0 9 0 .0 0 3 -0 .0 2 0 0 .0 0 2 -0 .0 0 7

: Jo
Mercury

u r n a l
Average 0 .0 1 3 0.010° 0 .0 0 7 0 .0 1 3 0 .0 2 9 d 0.0 40 0 0 .0 0 3 0.0 04 0 0 0.002
Range 0 .0 0 8 -0 .0 1 9 0 .0 0 7 -0 .0 1 3 0.014—0 .0 1 9 0 .0 0 8 -0 .0 1 9 0 .0 0 5 -0 .0 9 4 0 .0 2 3 -0 .0 8 3 0 .0 0 1 -0 .0 0 5 0 .0 0 2 -0 .0 0 8 0 .0 0 1 -0 .0 0 3

Of AOAC I n t e r n a t i o n a l
a Th e average and range of concentrations are based on data from 12 food groups in 4 regions and are expressed as m g/kg of raw food. Values below the limit of detection are reported as zero (0).
6 The Pb content of eggs in the 2 southern regions both exceed ed the national tolerance limit (0.2 m g/kg) (12).
c The Hg content of legum es and nuts in North 1 region exceed ed the national tolerance limit (0.1 m g/kg) (14).
d The Hg content of eggs in North 1 region exceeded the national tolerance limit (0 .05 m g/kq)(14).

V
.o l
76, No. 6,1993 1203
1204 C h e n & G a o : J o u r n a l O f A O A C In t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

T able lib . C o n c e n tra tio n s (p g/kg) o f o rg a n o c h lo rin e p e stic id e s in fo o d g r o u p s 3

Organochlorine Legumes Aquatic

pesticides Cereals and nuts Potatoes Meats Eggs foods Milk Vegetables Fruits

HCH
Average 4.4 3.9 1.8 23 .7 18.8 9.1 2.9 2 .3 2.5

to
CO
Range 3 .0 -5 .8 1 .7 -7 .2 1 .6 -1 .9 1 4 .1 -3 1 .2 6 .7 -2 4 .1 4 .9 -1 3 .5 1 .7 -2 .7 1 .9 -3 .2

I
DDT
Average 1.2 2.1 1.6 238.4 40.6 140.5 2.8 0.9 3.0
Range 0 .9 -1 .4 1.5 -2 .8 0.4—3.3 2 4 .4 -8 2 1 .3 1 2 .7 -7 1 .5 9 .7 -4 4 3 .7 0 .5 -6 .3 3 .2 -8 .0 0 .7 -6 .0

Th e average and range of concentrations are based on data from 9 food groups in 4 regions and are expressed as pg/kg of raw food.

T able lie . C o n c e n tra tio n s (p g/kg) o f T able lid . C o n c e n tra tio n s (g g /k g ) o f aflato x in s in 3 fo o d
o r g a n o p h o s p h o r u s p e stic id e s in 3 fo o d g r o u p s 3 gro u p s3

Organophosphorus Legum es and

pesticides Cereals Vegetables Fruits Aflatoxin Cereals nuts Milk

Dichlorvos Aflatoxin B-,

Average 7.9 6.8 4.1 Average 0.35 0 NA
R ange T - 1 2 .0 3 .0 -9 .1 0 .0- 6 .4 Range 0 .0 0 -1 .4 1
Methamidophos Aflatoxin M-,
Average 39.2 14.5 17.7 Average NA NA 0
R ange T -6 9 .1 T - 2 1 .4 1 4 .3 -2 7 .3 Range
Parathion
8 The average and range of concentrations are based on data from
Average 0 0 2.0
3 food groups in 4 regions and are expressed as pg/kg of raw
Range 0 .0 -5 .7 food, values below the limit of detection are reported as zero (0).
Trichlorfon NA indicates that aflatoxin was not analyzed in that food group.
Average 0 10.0 0
Range 0 .0 -3 9 .8
Dimethoate
Average 0 2.6 3.1
Range 0 .0 -1 0 .3 0 .0 -1 2 .3

a Th e average and range of concentrations are based on data from

3 related food groups in 4 regions and are expressed as pg/kg of
raw food. Values below the limit of quantitation are reported as
trace (T ) ; values below the limit of detection are reported as zero
(0). Averages for the following pesticides in cereals, vegetables,
and fruits are zero: parathion-methyl, acephate, disulfoton.
fenitrothion, fenithion, malathion, and phosmet.
T able Ile . C o n c e n tra tio n s (B q /k g ) of ra d io n u c lid e s in fo o d g r o u p s 9

Legum es and B everages Alcoholic

Radionuclide Cereals nuts Potatoes M eats Eggs A quatic foods Milk Vegetables Fruits Sugar and w ater beverages

90Sr
Average 0.1 9 0.3 3 0 .3 2 0.10 0.1 3 0 .3 2 0.02 0.2 9 0.0 4 0.02 0.01 0.01
R ange 0 .0 6 -0 .5 1 0 .1 2 -0 .7 4 0 .0 4 -0 .8 7 0 .0 2 -0 .1 9 0 .0 0 -0 .4 2 0 .0 1 -0 .7 9 0 .0 0 -0 .5 4 0 .0 8 -0 .5 6 0 .0 1 -0 .0 8 0 .0 0 -0 .0 4 0 .0 0 -0 .0 3 0 .00- 0.01
137Cs
Average 0 .0 3 0.0 8 10.21 0.0 9 013 0.10 0.01 0.05 0.01 0.01 0.02 0.01

C
Range 0 .0 0 -0 .0 9 0 .0 2 -0 .1 3 0 .01- 0.21 0 .0 0 -0 .3 0 0 .02- 0.12 0.00- 0.20 0 .00- 0.01 0.00- 0.11 0 .0 0 -0 .0 3 0 .0 0 -0 .0 3 0 .0 0 -0 .0 7 0 .00- 0.02

h e n
ZtOpb

&
Average 0.0 7 0.11 0 .0 3 0.20 0.41 3 .0 9 0.02 0.39 0.0 3 0 .0 8 0 .0 0 4 0 .0 0 4

G
R ange 0 .0 0 -0 .1 8 0 .0 0 -0 .2 7 0 .0 0 -0 .0 9 0 .0 9 -0 .3 1 0 .1 3 -0 .6 4 0 .7 2 -5 .1 6 0 .0 0 -0 .0 4 0 .1 8 -0 .7 1 0 .0 0 4 -0 .1 4 0 .0 2 -0 .1 9 0 .00- 0.01 0 .00- 0.01

a o
210Po

: Jo
Average 0.0 9 0.10 0 .0 3 0.13 0.32 3 .7 6 0.02 0.46 0.0 3 0 .0 0 3 0 .0 0 4 0 .0 0 3

u r n a l
Ranqe 0 .0 0 -0 .3 1 0 .0 0 -0 .2 3 0 .00- 0.11 0 .0 5 -0 .2 8 0 .0 0 -0 .6 7 0 .4 0 -3 3 .8 7 0 .0 0 -0 .0 4 0 .0 6 -0 .9 1 0 .0 0 -0 .0 9 0.00- 0.01 0 .00- 0.02 0 .00- 0.01
226Ra

Of AOAC I n t e r n a t i o n a l
Average 0.0 4 0.0 9 0.12 0.0 4 0.1 9 0 .0 5 0.02 0.09 0.0 3 0.02 0 .0 0 3 0.01
R ange 0 .0 3 -0 .0 5 0 .00- 0.22 0 .0 1 -0 .3 4 0 .0 3 -0 .0 6 0 .0 8 -0 .3 2 0 .0 0 -0 .1 3 0 .0 0 -0 .0 5 0 .0 4 -0 .1 6 0 .0 1 -0 .0 6 0 .0 0 -0 .0 3 0 .0 0 -0 .0 0 4 0 .0 0 -0 .0 3
228Ra
Average 0.10 0.26 0 .1 9 0.1 4 0.22 0 .3 4 0.02 0.2 6 0.0 6 0.02 0.01 0.01
R ange 0 .0 0 -0 .3 1 0 .0 9 -0 .4 3 0 .0 2 -0 .5 6 0 .0 4 -0 .2 4 0 .1 0 -0 .3 4 0 .0 5 -0 .9 9 0 .0 0 -0 .0 4 0 .0 5 -0 .5 9 0 .0 1 -0 .1 5 0 .0 0 -0 .0 4 0 .01- 0.12 0 .00- 0.01

a The average and range of concentrations are based on data from 12 food groups in 4 regions and are expressed as Bg/kg of raw food. Values below the limit of detection are reported as zero (0).

V
.o l
76, No. 6,1993 1205
1206 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

CHEMICAL CONTAMINANTS MONITORING

T h e C h in e s e T o ta l D ie t S tu d y in 1 9 9 0 . P a r t II. N u trie n ts

J u n sh i C hen and J unquan G ao

Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Medicine, 29 Nan Wei Rd, Beijing 100050, China

T h is p a p e r r e p o r t s t h e in ta k e s o f 7 2 n u tr ie n ts a n d sition pattern of a standard man (18^15 years old, 60 kg body

th e ir d ie ta r y s o u r c e s o b ta in e d fr o m th e C h in e s e to ­ weight, light physical activity) in each of the 1 2 provinces was
t a l d i e t s t u d y in 1 9 9 0 . M o s t o f t h e n u t r i e n t i n t a k e s determined by a 3-day household dietary survey. All the foods,
a r e c lo s e o r e q u a l to th e ir c o r r e s p o n d in g r e c o m ­ beverages, and water consumed by the standard man were in­
m e n d e d d a ily a llo w a n c e s (R D A s ) . B o th t h e to ta l e n ­ tegrated into 1 2 food groups: cereals, legumes, tubers, meats,
e r g y in ta k e (2 2 0 3 k c a l) a n d t h e p r o p o r t io n s c o n t r ib ­ eggs, aquatic foods, milk, vegetables, fruits, sugar, beverages
u t e d b y p r o te in , fa t, a n d c a r b o h y d r a t e m e e t th e and water, and alcoholic beverages. Food samples were col­
c u r r e n t C h in e s e R D A s a n d th e W o r ld H e a lth O r g a n i­
lected and prepared (cooked) in each province according to the
z a t i o n ( W H O ) n u t r i e n t g o a l s . T h e a v e r a g e p r o t e i n in ­
local food habits. Twelve food group composites were made
t a k e w a s 6 4 g /d a y . T h e in ta k e o f e s s e n tia l a m in o a c ­
for each province. Finally, the same food composites from each
id s a ll e x c e e d e d t h e C h in e s e R D A , a n d t h e ir
of the 3 provinces were combined to formulate regional market
p r o p o r tio n s w e r e g e n e r a lly c o n s is t e n t w it h t h e
baskets. A total of 48 food composites (12 composites x 4 mar­
W H O r e c o m m e n d e d p a tte r n . T h e a v e r a g e fa t in ta k e
ket baskets) were subjected to nutrient analysis. The nutrients
w a s 5 1 .2 g /d a y (2 1 .2 % o f t h e to ta l e n e r g y in ta k e ).
analyzed and the methods of analysis are shown in Table 1.
H o w e v e r , t h e d ie ta r y f a t in ta k e h a s b e e n in c r e a s in g
Data of analytical method evaluation are listed in Appendix a.
s i g n i f i c a n t l y in t h e C h i n e s e d i e t a n d t h e p r o p o r t i o n
Reference materials were used to assure analytical quality for
o f a n im a l f a t h a s r e a c h e d 5 3 % o f t h e t o ta l f a t in ­
ta k e . T h e to ta l s a tu ra te d :to ta l m o n o u n s a tu ra te d :to -
minerals and vitamins (Appendix b).
ta l p o ly u n s a tu r a t e d f a t t y a c id r a tio w a s 1 .0 :1 .5 :1 .0 .
A lth o u g h t h e a v e r a g e c h o le s te r o l in ta k e w a s o n ly R e s u lts a n d D is c u s s io n

1 7 9 m g / d a y , i t w a s 2 4 8 m g / d a y in t h e S o u t h 1 r e ­
g io n . T h e in ta k e s o f t h ia m in e a n d r ib o fla v in w e r e The intake levels of energy, protein, fat, carbohydrate, alco­
b e l o w t h e R D A . R e t i n o l i n t a k e s in a l l t h e 4 r e g i o n s hol, dietary fiber, vitamins, and minerals and the comparison
w e r e l o w . M o s t ( 8 0 % ) o f t h e r e t i n o l ( e q u i v a l e n t ) in ­ with the corresponding Chinese recommended dietary allow­
ta k e s w e r e fr o m c a r o te n o id s . T h e a v e r a g e in ta k e o f ance (RDA) and estimated safe and adequate daily intake
to ta l to c o p h e ro l w a s 8 9 % o f th e R D A , a n d a m o n g (ESADDI) (2) are listed in Tables 2-5. Most of the nutrient
t h e 4 r e g io n s , o n ly t h e S o u th 2 r e g io n h a s r e la t iv e ly intakes were close or equal to their corresponding RDAs and
lo w in ta k e . T h e in t a k e s o f ir o n , c o p p e r , m a n g a n e s e , ESADDIs, a fact that indicates that the composition of the diet
s o d iu m , a n d p h o s p h o r u s w e r e a d e q u a t e . T h e in ­ of the average Chinese male is adequate relative to nutrient
ta k e s o f c a lc iu m , z in c , a n d p o ta s s iu m w e r e in s u ffi­
content.
c ie n t, a n d in ta k e s o f s e le n iu m a n d m a g n e s iu m
w e r e a lit tle lo w . H ig h s o d iu m a n d lo w p o t a s s iu m in ­ Energy
t a k e i s a t r a d i t i o n a l p r o b l e m in t h e C h i n e s e d i e t .
Both the total energy intake (2203 kcal) and the proportions
contributed by protein, fat, and carbohydrate meet the Chinese
T he first Chinese total diet study was carried out in 1990. RDAs and the nutrient goals of the World Health Organization
The study design, experimental methods, and results on (WHO) (3) (Table 2). Among the 4 regions, North 2 had the
chemical contaminants are reported in the companion highest energy intake. Because plant foods and carbohydrate
article ( 1). The present paper focuses on the results on nutrients. provided 84.9 and 66.1%, respectively, of the total energy in­
take, the main source of energy in this region was cereals. On
E x p e r im e n ta l the other hand, the contribution of fat to the energy intake was
lowest in the North 2 region. These characteristics show that
this region consumed more cereals and less meat and fat (oil)
The overall study design and experimental methods were
than the other regions, a fact indicating a lower living standard.
described in a separate paper (1). In brief, the food composite
approach was used to study the total diet in 4 regional market
baskets; each region comprised 3 provinces. The food compo- Received November 24, 1992.Accepted M arch 4, 1993.
 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1207

T a b le 1 . A n a ly t e s a n d a n a ly t ic a l m e t h o d s f o r Protein
d e t e r m i n in g n u t r ie n t s in C h in e s e t o t a l d i e t s t u d y in 1 9 9 0
The average protein intake was 64 g/day, which is 91.4% of
Nutrient Method (Ref.)
the Chinese RDA (Table 3) and exceeds the protein require­
ment (0.75 g/kg body wt) recommended by WHO (4). Even
Macronutrients
considering the possible lower digestibility and bioavailability
Protein Kjeldal method (12)
Fat Soxhlet extraction (12)
of plant protein (the actual protein requirement of Chinese peo­
Dietary fiber Neutral detergent m ethod(11)
ple may be a little higher, 1.0 g/kg body wt), the current protein
Ash and moisture Weighing m ethod (12)
intake is still adequate. In addition, the combined proportion of
animal protein (21.8%) and legume protein (8.3%) was 30.1 %,
Minerals and trace elements which indicates that the overall quality of protein is fine.
Sodium, potassium Emission spectroscopy (11) Among the 4 study regions, the 2 northern regions had higher
Phosohorus, calcium, protein intakes than the 2 southern regions. This is a result of
magnesium, m anganese, zinc,
the higher total food consumption and cereal consumption in
copper, iron, soluble iron Flam e AAS (11)
the northern regions. However, the proportion of animal pro­
Total ion iron, ferrous ion Spectrophotometry (15)
N onhem e iron and its
tein consumed in the southern regions was higher than in the
bioavailability Spectrophotom etry (16) northern regions.
Selenium Spectrophotofluorometry (11)
Amino Acids
Vitamins
Consistently, the intakes of the 8 essential amino acids (Ta­
Retinol, tocopherols LC (13)
Thiam ine Spectrophotofluorometry (13)
ble 6) all exceeded the Chinese RDAs. The relative proportions
Riboflavin Microbiological method (13)
of these essential amino acids (Table 7) were also close to the
Carotenoids P aper LC (13)
WHO recommended pattern, but only for the adult diet, which
may indicate problems in the diets of children. Details will be
Fatty acids G C (13) reported in another paper (5).
Palmitic, stearic, oleic, linoleic,
linolenic, etc. (2 5 fatty acids) Fat

Cholesterol
Spectrophotom etry (13)
The average fat intake was 51.2 g/day (Table 3), which ac­
counted for 21.2% of the total energy intake (Table 2). Al­
Amino acids though this level of dietary fat intake is considered ideal and is
Tryptophan Spectrophotofluorometry (13) far below the upper limit of the WHO nutrient goal (30%) (3),
Histidine, isoleucine, leucine, dietary fat has been increasing significantly in the Chinese diet
lysine, methionine, cystine,
(Table 8), and the proportion of animal fat consumed has
tyrosine, phenylalanine,
threonine, valine, glycine,
reached 53% of the total fat intake (Table 3).
alanine, serine, proline,
Fatty Acids
cysteine, arginine, glutamic acid Amino acid analyzer (13)

It has been suggested that the appropriate proportion of the

total saturated, total monounsaturated, and total polyunsatu­
rated fatty acids (S:M:P) is 1:1:1 (6). The S:M:P ratio obtained
in this study was 1.0:1.5:1.0; the P:S ratio (1.0) is appropriate,

T a b le 2 . E n e r g y in t a k e ( k c a l) a n d c o n t r ib u t io n (% ) t o d a ily in t a k e b y m a in n u t r ie n t s a n d a lc o h o l a n d c o m p a r is o n w it h
R D A sa

Total energy Protein Fat Carbohydrates Alcohol

% from % from
animal plant % from % total % total % total % total
Region Intake % RDA food food alcohol Intake energy Intake energy Intake energy Intake energy

North 1 2159 9 0 .0 15.2 83.4 1.4 248 11.5 481 22 .3 1397 64.8 32 1.4
North 2 24 44 101.8 7.9 91.9 0.2 294 12.0 361 14.8 1784 73.0 4 0.2
South 1 2121 8 8 .4 18.5 80.7 0.8 245 11.6 546 25 .8 1311 61.8 18 0.8
South 2 20 90 87.1 15.4 82.6 2.0 236 11.3 455 21.8 1357 64.9 42 2.0
M ean 2203 91 .8 14.0 84.9 1.1 256 11.6 461 21.2 1462 66.1 24 1.1

a C hinese R D A for energy is 2 4 0 0 kcal for an adult m ale with light physical activity (2). The Chinese National Nutrition Society also
recomm ends ( 2) that the proportion of energy contribution by protein, fat, and carbohydrates to the total energy intake should be 1 0 -1 5 ,
2 0 -2 5 , and 6 0 -7 0 % , respectively.
1208 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

T able 3. Protein, fat, carb o h yd rate s, dietary fiber, a n d a lc o h o lic intake (g per p e rso n per day)

Protein Fat
C arbohy­ Dietary
% from % from % from % from % from drate Alcohol fiber
Region Intake % RDAa animal legum es plant Intake animal plant intake intake intake

North 1 62.1 88 .7 23 .3 11.6 76.7 53.5 51.2 48 .8 349 4.6 2 6 .3

North 2 73.4 104.9 11.5 6.5 88.5 40.2 42 .3 57 .7 446 0.6 26.4
South 1 61.3 87 .5 32.1 6.9 67.9 60.7 56.8 43 .2 328 2.6 26 .7
South 2 59.1 84 .4 22.2 8.6 77.8 50.5 59.0 41 .0 33 9 6.0 26 .9
M ean 64.0 91.4 21.8 8.3 78.2 51.2 53 .0 47 .0 36 6 3.4 2 6 .6

“ Chinese RDA for protein is 7 0 g for an adult m ale with light physical activity (2).

T able 4. Dietary vitam in intake (m g per p e rso n per d ay )

Retinol Tocopherols Thiam ine Riboflavin

Total retinol % from % from

Region equivalent, pg % R D A a vitamin A carotenoids Intake % RDAb Intake % RDAC Intake % BDAd

North 1 326 40 .8 14.1 85.9 11.3 113.0 1.0 86 .4 0.7 60 .8

North 2 558 69.8 9.0 91.0 11.0 110.0 1.3 107 0.9 73 .3
South 1 290 36.3 4 1 .0 59.0 8.6 86.0 0 .7 58.2 0.8 65 .0
South 2 175 21 .9 3 0 .3 69.7 4.5 45.0 0.4 37.0 0.7 57.5
Mean 337 42 .2 19.9 80.1 8.9 89.0 0.9 72.2 0.8 64 .2

a RDA for retinol = 80 0 mg per adult m ale per day (2).

6 R D A for tocophols = 10.0 mg per adult male per day (2).
c RDA for thiamine = 1.2 mg per adult m ale per d ay (2).
d RDA for riboflavin = 1.2 mg per adult male per day (2).

Table 5. C o m p a r is o n o f m ineral in takes (m g per p e rso n per day) with R D A a an d E S A D D I 6

Ca Mg Iron Selenium 0 Zinc

(R D A = 800) (E S A D D I = 350) (R D A = 12) (R D A = 5 0 ) (R D A = 15)

Region Intake % RDA Intake % ESA D D I Intake % RDA Intake % RDA Intake % ESA D D I

North 1 594.0 74 .2 34 1.8 97.7 25 .7 21 4.3 4 3 .4 97.7 84 56 .0

North 2 774.8 96.9 38 8.0 110.8 29.7 24 7.9 32 .8 65.6 12.6 84.1
South 1 48 2.3 60.3 20 7.2 59.2 14.9 124.1 53 .8 107.6 9.3 62 .0
South 2 47 6.9 59 .6 2 0 3 .6 58.2 20 .5 170.5 39.1 78.1 8.9 59 .2
Mean 58 2.0 72 .8 28 5.2 81.5 22 .7 189.2 4 2 .3 84.6 9.8 65.2

Copper M anganese Sodium Potassium

(% ED A D D I = 2) (% E SA D D I = 2.5) Phos- (% E SA D D I = 3 3 00) (% ESA D D I = 1875)
Na:K
Region Intake % ESA D D I Intake % ESA D D I intake Intake % E SA D D I Intake % ESADDI ratio

North 1 3.0 150.0 5.8 23 2.0 95 3.4 41 23 124.9 1231 65 .7 3.4

North 2 11.0 547.8 6.7 26 8.0 989.2 4442 134.6 1237 66.0 3.6
South 1 2.6 127.8 5.6 22 4.0 76 3.4 3338 101.1 935 49.9 3.6
South 2 2.5 125.4 5.4 21 6.0 76 0.2 30 87 93.5 1075 57 .3 2.9
Mean 4.8 237.7 5.9 23 6.0 866.5 37 47 113.6 1119 59.7 3.4

a R D A values are from reference 2.

b ESA D D I = estimated safe and adequate daily dietary intake; values are from reference 2.
c Intake and R D A are in gg per person per day.
 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1209

Table 6. C o m p a r is o n o f e sse n tia l a m in o a c id intake (m g per p e rso n per d ay ) with R D A a

Methionine Phenylalanire
Isoleucine Leucine Lysine and cystine and tyrosine Threonine Tryptophan Valine
(R D A = 6 0 0 ) (R D A = 84 0) (R D A = 720) (R D A 780) (R D A = 840) (R D A = 4 2 0 ) (R D A = 210) (R D A = 6 0 0 )

% % % % % % % %
Region Intake RDA Intake RDA Intake RDA Intake RDA Intake RDA Intake RDA Intake RDA Intake RDA

North 1 23 73 39 6 48 25 574 29 08 404 20 18 259 49 14 585 2023 482 735 35 0 29 28 488

North 2 26 78 446 55 66 663 27 28 379 2226 285 64 99 774 2317 552 834 397 2946 491
South 1 24 07 401 46 10 549 31 84 442 2234 286 5041 600 2151 512 774 369 31 78 530
South 2 22 52 375 44 95 535 30 84 423 2111 271 4974 592 2129 507 744 354 2965 494
M ean 24 28 404 4874 580 29 67 412 2147 275 53 57 638 2155 513 774 367 30 04 501

RDA values are in mg per person per day; based on RDA recom m ended by the C hinese Nutrient Society (2) for 60 kg body mass.

T able 7. E sse n tia l am in o a c id intake pattern co m p a re d with the W H O re c o m m e n d e d intake pattern (m g /g protein)

Amino acid N o rth 1 a North 2 a South 1a South 2 a C hinese pattern 3 W H O pattern

Isoleuclne 16 16 16 15 16 13
Leucine 33 33 30 30 31 19
Lysine 20 16 21 20 19 16
Methionine and cystine 14 13 14 14 14 17
Phenylalanine and tyrosine 33 39 33 33 35 19
Threonine 14 14 14 14 14 9
Tryptophan 5 5 5 5 5 5
Valine 20 18 21 20 19 13
Total 155 154 153 152 153 111

To compare with the W H O recom m ended intake pattern(4), the intake of tryptophan in the total study w as set as 5.

but the M:S ratio (1.5) is 50% higher than the suggested ratio
of 1.0 (Table 9). However, because the major part of the
monounsaturated fatty acid was oleic acid, this high M:S ratio
may indicate a protective effect against coronary heart disease
(7).
Table 8. M acron utrient an d a lc o h o l intake in 1990 total
diet s tu d y an d in 1982 nation al nutrition s u r v e y 3 Cholesterol

Nutrient 1982 survey 1990 Survey % difference

Although the average cholesterol intake was only
179 mg/day (Table 10), regional variation was significant. The
Energy Intake, kcal 2498 22 03 - 11.8
cholesterol intake in the South 1 region was 248 mg/day, which
% from animal food 5.4 14.0 + 15 9.3
is 83% of the WHO upper limit (300 mg/day) (3). The main
% from plant food 94.0 84.9 -9 .7
source of dietary cholesterol is egg, which accounts for more
% from alcohol 0.6 1.1 + 33 .3
than 50% of the total cholesterol intake. The second main
Protein intake, g 66.0 64.0 - 3 .0 source is meat, which accounts for 30% of the total intake (Ta­
% from animal food 8.1 21.8 +169.1 ble 11).
% from legum es 9.0 8.3 - 7 .8
% from other plant Vitamins
food 82.9 69 .9 -1 5 .7
Table 4 shows the intake of vitamins. Consistent with pre­
Fat intake, g 44.1 51.2 +16.1 vious surveys (8, 9), intakes of thiamine and riboflavin were
% from animal food 36.3 53.0 + 46.0 still below the RDA. The thiamine intakes in the northern re­
% from plant food 63.7 47 .0 -2 6 .2
gions were sufficient, but the thiamine intakes in the southern
a Data of 1982 and 19 90 surveys were from the sam e 12 provinces.
regions were insufficient, especially in the South 2 region,
Th e 1982 data were calculated from the Chinese Food where it was only 37% of the RDA. The most likely cause of
Composition Table, and the 1990 data were the directly m easured this low intake of thiamine is the consumption of rice (the main
results.
food in southern regions) with a high milling rate. In contrast,
1210 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

T able 9. Fatty a c id intake (m g per p e rso n per d a y ) a n d S : M : P ratio3

Monounsatu rated
Saturated fatty acid fatty acid Polyunsaturated fatty acid

Myristic Palmitic Stearic Oleic Linoleic Linolenic Total fatty

Region Total (14:0) (16:0) (18:0) Total (18:1) Total (18:2) (18:3) acids S:M:P

North 1 14.9 0.5 10.4 3.4 20.0 17.5 16.6 14.7 1.6 53.2 1.0:1.5:1.3
North 2 11.8 0.4 7.3 3.5 14.9 13.0 13.4 12.3 0.8 39.8 1.0:1.3:1.2
South 1 16.7 0.5 11.2 4.2 29.8 24.8 13.7 12.5 1.1 60.5 1.0:1.8:0.9
South 2 15.9 0.5 10.6 4.4 23.4 18.4 10.9 9.6 1.3 50.2 1.0:1.5:0.7
Mean 14.5 0.5 9.8 3.8 22.2 18.6 14.3 12.7 1.5 51.0 1.0:1.5:1.0

a Suggested S:M:P ratio is 1:1:1 (5).

T able 10. C h o le ste ro l intake (m g per p e rso n per day) in regional differences in riboflavin intake were not significant,
total diet stu d y and riboflavin intake in all the 4 regions was less than the RD A.
Region Cholesterol intake % of WHO goal3 The highest intake was in the North 2 region, which was 73%
of the RDA. However, some nutritionists have questioned the
North 1 216 72.1 riboflavin RDA and believe that it is too high (10). Therefore,
North 2 135 44.9 the amount of riboflavin in the average Chinese diet may not
South 1 248 82.6 really be insufficient.
South 2 118 39.2 Retinol intakes in all the 4 regions were low, with intake in
Mean 179 59.7 the North 2 region as the highest (70% of RDA). Most (80%)
of the retinol (equivalent) intake was from carotenoids, and
a WHO goal, 300 mg per person per day (upper limit) (3).
only 20% was from vitamin A. The major food source of ca­
rotenoids was vegetables, which accounted for 8 6 % of the total
carotenoid intake. The best result was from the North 2 region:
Table 11. C on tribu tio n (% of daily intake) of v a rio u s The retinol (equivalent) intake was 70% of the RDA and 91%
fo o d g r o u p s to dietary intake o f ch olestero l of the intake was from carotenoids, 85% of which was from
Region Meat Eggs Aquatic foods Milk vegetables. The average intake of total tocopherols was 89% of
the RDA. Among the 4 regions, the northern regions had higher
North 1 18.2 58.0 23.2 0.6 intakes than the southern regions and were equal to RDA; how­
North 2 29.5 63.3 5.9 1.3 ever, intake of tocopherols in the South 2 region was only 45%
South 1 28.6 52.4 18.8 0.2 of the RDA.
South 2 48.6 41.9 9.2 0.3
Mean 28.9 54.4 16.1 0.6 Minerals and Trace Elements

Table 5 lists the intakes of minerals and trace elements. In­

takes of iron, copper, manganese, sodium, and phosphorus
were adequate in the average Chinese diet. Consistent with pre­
vious reports (8 , 9), intakes of calcium, zinc, and potassium
were insufficient and intakes of selenium and magnesium were

T able 12. Intake3 (m g per p e rso n per d ay ) an d bioavailability (A % ) of different c h e m ical fo rm s o f dietary iron

Heme Fe Nonheme Fe Total bioavailable Fe

Composite Total Fe
diet intake Intake A % b-C IBId Intake A% c IBIe Intake^ A% b

Northern 27.67 8.94 20.0 1.79 18.73 5.70 1.07 2.86 10.3
Southern 17.53 4.58 20.0 0.92 12.95 6.03 0.78 1.70 9.7
Mean 22.60 6.76 20.0 1.35 15.84 5.87 0.93 2.28 10.0

b Bioavailability in %.
Quoted from reference 14.
d Intake of bioavailable iron = heme iron intake xA%.
° Percentage of bioavailable nonheme iron was actual determined value.
1 Intake of total bioavailable iron=IBI of heme iron plus IBI of nonheme iron.
 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1211

a little low. High sodium and low potassium intake is a tradi­ fatty acids, and cholesterol), Guangya Wang (minerals, trace
tional problem in the Chinese diet due to the high intake of salt. elements, vitamins, dietary fiber, ash, and water), Shengjie Liu
All 4 regions showed high Na:K ratios, with the South 2 region (chemical forms of iron), and Ruihua Zhou (selenium). The
showing the lowest ratio. The obvious measure to correct this authors also wish to thank John W. Jones of the U.S. Food and
problem is to reduce salt intake (13.9 g/day in this study). Al­ Dmg Administration for advice on the study design and his
though iron intake (22.7 mg/person/day) was much higher than review of and comments on this paper.
the RDA and the reported value for the western population, the
bioavailability of iron in the Chinese diet is low. As shown in References
Table 12, the total bioavailable iron accounts for only 10% of
the total iron intake. However, because of the high total iron (1) Chen, J„ & Gao, J. (1993) J. AOACInt. 76,1193-1205
intake, the total available iron from heme and nonheme iron (2) Chinese Nutrition Society (1989) Acta Nutr. Sin. 11, 93-96
seems sufficient for adults. On the other hand, for children and (in Chinese)
pregnant women, more available iron from animal foods might (3) WHO Technical Report Series No.797 (1990) Geneva, Swit­
be needed. More than 80% of the intakes of the above-men­ zerland
tioned minerals and trace elements (except selenium) were (4) World Health Organization (1985) Energy and Protein Re­
from plant foods, mainly cereals, legumes, tubers, and vegeta­ quirements, Report of a Joint FAO/WHO/UNU Expert
bles. Therefore, how to improve the mineral and trace element Consultation, Technical Report Series No. 724, Geneva, Swit­
status of the Chinese population whose diet is mainly com­ zerland, p. 206
posed of plant foods is an important nutritional problem in (5) Jia, J., Zhao, X., Lin, Y., Xu, Z., Zhang, Q., & Gao, J. (1993)
J. Hyg. Res. 22
China. The current practical measure will be to increase food
varieties and consumption of fresh green leafy vegetables as (6) Dietary Fats and Oils in Human Nutrition (1980) FAO Food
and Nutrition Series, No. 20, p. 38
well as milk and milk products.
(7) Fan, W, Parker, R., Parpia, B., Qu, Y., Cassano, R, Craw­
ford, M., Leyton, J., Tian, J., Li, J., Chen, J., & Campbell, T.
Conclusions
C. (1990) Am. J. Clin. Nutr. 5 2 , 1027-1036
(8) Jin, D. (1986) Med. China 2, 66-67
Results of the first Chinese total diet study show that the
(9) Chen, J., Campbell, T.C., Li, J., & Peto, R. (1990) Diet, Life­
average Chinese diet provides sufficient energy, protein, fat, style and Disease Mortality: A Survey of the Characteristics
carbohydrate, and fiber, as well as most vitamins and minerals. of 65 Chinese Counties, Oxford University Press, Oxford,
However, retinol, riboflavin, calcium, zinc, and potassium are UK
not present in sufficient amounts. Children and women (includ­ (10) Campbell, T. C., Bran, T., Chen, J., Feng, Z., & Parpia, B.
ing pregnant and lactating women) who have different dietary (1990) Am. J. Clin. Nutr. 51, 436-445
requirements may have problems different from those of a (11) National Standard of People’s Republic of China (1990)
standard Chinese man. GB12388-12399-90, Methods for Determination of Nutrient
This total diet study studied only the average diet of the gen­ Composition in Foods, Standard Press, Beijing, China (in
eral Chinese population. Considering the significant geo­ Chinese)
graphical variation of the Chinese diet (dietary pattern as well (12) National Standard of People’s Republic of China (1985)
as geochemical environment of food crop production), the total GB5009.3-6-85, Methodfor Food Hygiene Analysis, Physi­
diet study cannot reveal specific nutritional problems such as cal-Chemical Section, Standard Press, Beijing, China (in
Chinese)
selenium deficiency in individual regions. Specific studies
should be designed to study these particular regional problems. (13) Institute of Nutrition and Food Hygiene, Chinese Academy
of Preventive Medicine (1990) Methods for Determination of
Nutrient Composition in Foods, People’s Medical Publishing
Acknowledgments House, Beijing, China (in Chinese)
(14) USA Recommended Dietary Allowances (1989) lOthEd.,Na­
The authors would like to express their appreciation to the tional Academy Press, Washington, DC, p. 198
12 provincial health teams who carried out the field work, as (15) Ma, Y, & Liu, S. (1993) J. Hyg. Res. 22, Supplement I on
well as the following scientists in the Institute of Nutrition and The Chinese Total Diet Study in 1990 (in Chinese)
Food Hygiene of the Chinese Academy of Preventive Medi­ (16) Liu, S., Liu, H„ Ma, Y, Li, W„ & Xu, J. (1993) J. Hyg. Res.
cine, who were responsible for laboratory analysis of :he nutri­ 22, Supplement I on The Chinese Total Diet Study in 1990
ents: Xihe Zhao (protein and amino acids), Wenxun Fan (fat, (in Chinese)
1212 C h e n & G a o : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

Appendix

T able a. E va lu a tio n data of an alytical m e th o d s for ch e m ic al c o n ta m in a n ts in C h in e s e total diet stu d y

Chemical Limit of detection Recovery, % CV, %

Minerals and trace elements

Potassium 0.05 mg/L 101.2-103.1 <2.6
Sodium 0.3 mg/L 100.2-101.3 <5.1
Calcium 0.1 mg/L 99.4-100.5 <5.7
Magnesium 0.05 mg/L 99.2-100.0 <5.9
Phosphorus 1.49 mg/L 98.1-101.0 <6.4
Copper 0.5 mg/L 98.5 <2.8
Manganese 0.1 mg/L 100.5-101.2 <9.6
Zinc 0.5 mg/L 95.0 <3.4
Selenium 3.0 pg/kg 92.6-101.6 8.2
Total iron 0.2 mg/L 101.5-102.3 <7.2
Soluble iron 0.2 mg/L 98.3 <5.0
Total ion iron 0.025 mg/L 96.4 (92-101) <5.1
Ferrous ion 0.025 mg/L 97.1 (96-98) <7.3
Nonheme 0.1 mg/L 93.3-105.3 <3.5

Vitamins
Retinol 0.04 mg/L 97.4 <2.4
Carotenoids 0.11 pg 93.7 <1.7
Tocopherols
a 4.59 mg/L 97.3-98.0 <9.0
ß+Y 1.83 mg/L 98.2-101.2 <9.0
Ô 1.03 mg/L 96.6 (95-98) <9.0
Thiamine 0.05 pg 90.0-99.0 <8.5
Riboflavin 0.05 pg 100.1-101.7 <5.2

Fatty acids
Palmitic, stearic, oleic, linoleic, linolenic, etc. (25 fatty
acids) 0.1%

Cholesterol 30.0 mg/L 95.2-106.5 <5.4

Amino acids
Cystine 100.0 mg/L 95.5-98.9 <4.7
Tryptophan 5.0 mg/L 87.4-98.9 <4.0
Histidine, isoleucine, leucine, lysine, alanine,
methionine, tyrosine, phenylalanine, threonine, valine, <6.0
glycine, serine, proline, cysteine, arginine, glutamic
acid 100 mg/L 80.3-109.2

Macronutrients
Protein 50 mg/L 96.6-100.0 <2.0
Fat 200 mg/kg 96.1-99.9 <4.4
Dietary fiber 0.1 g/kg <7.7
 M c N e a l E t A l .: J o u r n a l O f A D A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1213

T able b. A n aly tic al quality a s s u r a n c e o f m in e ra ls an d vita m in s in C h in e se total diet stu d y

Number of Certified value ±SD,

Analyte Reference material3 determinations Mean ± SD, mg/kg Range, mg/kg mg/kg

Potassium Bread crumbs 41B 3 903.5 890.9-912.1 910.0176.0

Sodium Bread crumbs 41B 3 755.7 739.7-765.4 743.0 1 54.0
Calcium Bread crumbs 41B 3 256.6 253.8-260.6 257.0 + 13.0
Selenium Beef liver NBS1577 2 — 0.686-0.702 0.71 10.07
Iron Bread crumbs 41B 3 8.63 8.43-8.92 8.2810.85
Carotenoids Bread crumbs 41B 3 1.03 0.96-1.08 1.42 + 0.31
Thiamine Bread crumbs 41B 6 0.46 0.45-0.47 0.6510.12
Riboflavin Bread crumbs 41B 2 0.70 — 0.7510.08

USA Campbell Soup Co. reference materials, except beef powder from NBS.

CHEMICAL CONTAMINANTS MONITORING

Survey of Benzene in Foods by Using Headspace Concentration

Techniques and Capillary Gas Chromatography

T im o t h y P. M cN e a l, P a t r ic ia J . N y m a n , G reg o ry W. D ia c h e n k o , and H en ry C. H o l l if ie l d
U.S. Food and Drug Administration, Division of Food and Chemical Technology, Washington, DC 20204

Recently, the com bination of sodium or potassium without added benzoates (including eggs) con­
benzoate with ascorbic acid was shown to produce tained benzene at levels equal to or less than 2
low levels (ng/g) of benzene in fruit-flavored soft ng/g. Slightly higher levels were present in some
drinks. The presence of benzene also was reported foods and beverages containing both ascorbic acid
in butter, eggs, meat, and certain fruits; levels of and sodium benzoate.
these findings ranged from 0.5 ng/g in butter to
500-1900 ng/g in eggs. Because benzoates are
widely used as food preservatives, a lim ited survey S ince 1990, 3 reports of benzene in beverages have oc­
of other foods containing added benzoate salts curred (1-3). In the first case, low levels (ng/g) were
was conducted to investigate the potential for ben­ found in Perrier mineral water. In the other cases, 2 bev­
zene form ation. Selected foods that did not contain erage manufacturers independently discovered low levels
added benzoates but were previously reported to (ng/g) in a limited number of their products. Even though the
contain benzene were analyzed for comparison. presence of such low levels of benzene was not considered to
More than 50 foods were analyzed by purge-and- be a significant short-term health hazard, the companies in­
trap or static headspace concentration and capil­ volved voluntarily removed the contaminated products from
lary gas chromatography. Benzene was quantitated the market.
by using the method of standard additions, and its Each company began investigations to find the causes for
identity was confirmed by mass selective detec­ the presence of benzene in their products. The makers of Per­
tion. Results of this lim ited survey show that foods rier mineral water identified the source of their contamination
as the inadequate maintenance of a carbon filter used to purify
the carbon dioxide derived from an underground gas source;
Received July 6, 1992. Accepted January 29, 1993
Part o f this work was presented at the 105 th AO AC Annual
once identified, the problem was corrected. The soft drink
International Meeting, August 1 2 -1 5 , 1991, at Phoenix, AZ. manufacturers attributed the benzene in their beverages to the
1214 M c N e a l E t A l .: J o u r n a l O f A O A C In t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

A B

F igu re 1. F ID c h ro m a to g ra m s from pu rge-an d-trap G C a n a ly s is o f cran berry juice cocktail: A, a n a ly s is o f 10 g

unfortified juice c o n ta in in g <0.5 n g benzene/g; B, a n a ly s is o f 10 g juice fortified with benzene at 0.5 ng/g.

presence of ascorbic acid and sodium benzoate in their drink Experimental

formulations (3). Removing one or the other of these com­
pounds eliminated the benzene. Apparatus
Reports of naturally occurring benzene in foods are rela­
tively common in the literature (4-9), but these reports are the (a) Headspace vials.—22 mL, with Teflon-faced silicon
first implication of “generally recognized as safe” (GRAS) disks and aluminum crimp seals (Shamrock Glass Co., Inc., PO
Box 6 8 6 , Seaford, DE 19973). Store vials in 90°C forced-air
substances in the formation of benzene. Subsequently, several
oven until ready for use, and keep disks at 90°C under vacuum
topics needed to be investigated. Benzene levels in certain bev­
until ready for use. Demonstrate absence of coeluting materials
erages needed to be compared with levels of naturally occur­
under GC/MSD conditions.
ring benzene in other foods. For regulatory purposes, possible
(b) Syringes.—Hamilton 700 series; 5 mL Pressure-Lok®
reactions between foods and added substances in human diets
gas syringe (Precision Sampling Corp., Baton Rouge, LA
needed to be investigated, as well as the concentration of ben­ 70815); 89 mm, 13-gauge hypodermic needle.
zene resulting from such reactions. (c) Purge-and-trap concentrator/gas chromatograph.—
To address these issues, we examined more than 50 foods Tekmar Model 4000 dynamic headspace concentrator coupled
for benzene, including a number that were previously reported to Perkin-Elmer Sigma 2000 gas chromatograph with flame
to contain naturally occurring benzene. Included in the survey ionization detector. Remove injector module A; pass heated
were foods containing benzoates and ascorbates, which either transfer line of concentrator through opening into GC oven,
were added as preservatives or occurred naturally. and connect to manual splitter. Lit the purge line and the trans­
The foods were collected and analyzed from 1991 to early fer line to the trap with standard 1. 6 mm od x 0.5 mm id nickel
1992. The surveyed foods encompassed a wide variety of ma­ tubing for sampling. Use traps filled with Tenax adsorbent.
trixes. To determine benzene concentrations in matrixes rang­ Capillary column: 50 m x 0.32 mm id, 1 |im film, HP-5
(Hewlett-Packard Co., Rockville, MD 20850). Operating con­
ing from water to egg, jam, and other heterogeneous food
ditions: carrier gas, ultra high purity (UHP) helium; purge flow
types, purge-and-trap headspace concentration/capillary gas
rate, 30 mL/min; desorption flow rate, 30 mL/min; split injec­
chromatography with flame ionization detection (GC/FID)
tion; split ratio, 15:1; column flow, 1.3 mL/min; hydrogen pres­
was selected as the method of choice. Findings were confirmed sure for LID, 19 psi; air pressure for LID, 20 psi. Purge Tekmar
by static headspace concentration/capillary gas chromatogra­ 4000 for 9 min at room temperature, desorb 3 min at 180X1,
phy with mass selective detection (GC/MSD). The analytical and heat trap 10 min at 225X3. Heat GC oven 4 min at 70X1,
techniques used, the findings, and some of their implications increase temperature 107min to 200X1, and hold 8 min. Ben­
are discussed below. zene retention time, ca 11 min. Data handling, WINner® data
 M c N e a l E t A l .: J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1215

A B
+ 10 ppb Benzene

F igu re 2. F ID c h ro m a to g ra m s fro m pu rge-an d-trap G C a n a ly s is o f ta c o sa u c e : A, a n a ly s is o f 5 g unfortified taco

s a u c e c o n ta in in g 22 n g benzene/g; B, a n a ly s is of 5 g ta c o s a u c e fortified with benzene at 10 ng/g.

system with SP 4270 computing integrator (Spectra-Physics (c) Standard solutions.—(7) Stock solution.—Transfer 50
Corp., Piscataway, NJ 08854). pL benzene standard to 20 mL methanol in a tared and sealed
(d) Gas chromatograph/mass spectrometer.—Hewlett- headspace vial, weigh, and calculate the exact concentration of
Packard 5890 gas chromatograph with capillary direct inter­ benzene (ca 2400 pg benzene/mL solution). Replace Teflon
face to Hewlett-Packard 5970B mass selective detector. Gas disk after each day of use, and store vial in freezer until needed.
chromatograph equipped with subambient oven accessory. GC With new septa replacements, stock solution is stable for 2
conditions: capillary column, 50 m x 0.32 mm id, 0.5 pm film, weeks. (2) Working standards.—Prepare daily. Transfer
HP-5; carrier gas, UHP helium; column head pressure, 8 psi; aliquots of stock solution to 20 mL portions of water in sealed
linear velocity, ca 35 cm/s; split/splitless injector, vent opened headspace vials to obtain working concentrations of 0 .1 - 2 0
pg/mL.
1.5 min; oven program, 4 min at 20°C, increase temperature
207min to 50°C, hold 0.5 min, increase temperature 57min to
Purge-and-Trap Headspace Concentration/Capillary
10073, hold 0.5 min, increase temperature 257min to 20073.
GC/FID
Benzene retention time, ca 12.6 min. MSD parameters: Tune
manually with perfluorotributylamine; optimize lenses and
Control solutions.—For external calibration, transfer 5 g
voltages for mJz 414; acquisition mode, selected ion monitor­
water to headspace vials, fortify with appropriate benzene
ing; benzene ions scanned, mJz 52, 77, and 78; scan time, working standard to yield a benzene concentration between 1
10.24—13.0 min; dwell time, 50 ms with low mass resolution, and 200 ng/g, and seal with a Teflon-faced silicon disk. For the
ca 5 scans/s. benzene formation studies, transfer 5 g of test solution to a
headspace vial and seal. To analyze the control solution, pierce
Reagents the Teflon disk twice with a 13-gauge needle, and fit purge-
and-trap transfer lines through holes in disk. Push the purge fine
Caution! Benzene is a human carcinogen and is highly flam­
to the bottom of vial, let the transfer fine to the trap remain in
mable. Use appropriate safety precautions when handling.
upper portion of the vial well above the level of the solution,
(a) Benzene standard.—Certified ACS (Fisher Scientific, and start the analysis by using purge-and-trap GC conditions
Fair Lawn, NJ 07410). described in Apparatus (c). Report the integrated area for the
(b) Methanol and deionized water.—Absolute methanol peak at the benzene retention time, and repeat the analysis until
suitable for purge-and-trap analysis. Distilled and deionized all control solutions are analyzed. For external calibration, sub­
water obtained from Milli-Q® water purification system (Mil- ject all data to linear regression analysis. Plot data from the
lipore Corp., Bedford, MA 01730). Demonstrate absence of analysis of the standard curve series. Determine the benzene
interferences under the conditions used for GC. concentration in the test solution from the standard curve.
1216 M c N e a l E t A l .: J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

T able 1. S u m m a r y o f b e nzene fin d in g s in f o o d s

Food Benzene, ng/g Food Benzene, ng/g Food Benzene, ng/g Food Benzene, ng/g

Foods with no added benzoates Foods containing added benzoates

Brewed instant Bloody Mary Margarita mix,

Apple juice <1 coffee 0 mix, liquid 3 liquid 2
Strained apple Red raspberry Fruit punch <1 Iced tea 0
juice 0 preserves 1 Diet raspberry
Strained Diet cola <1 soda <1
apple-cherry Black currant Diet orange Diet cherry
juice <1 preserves 2 soda A <1 berry soda 2
Cranberry juice Strawberry Diet orange Diet grapefruit
cocktail preserves 0 soda B 0 soda 1
Brand A 1 Grape jelly 0 Diet white Lite strawberry
Liquid smoke, grape soda <1 preserves <1
Brand B 1 Brand A 121 Imitation
Liquid smoke, Imitation grape strawberry
Brand C <1 Brand B 21 jelly 5 preserves 38
Liquid beef Lite orange
Brand D <1 bouillon 0 Duck sauce 7 marmalade 1
Cranberry juice All purpose
concentrate 0 Raw potato3 0 sauce 1 Lite grape jelly 1
Fresh Taco sauce,
cranberries <1 Baked potato3 0 BBQ sauce 5 Brand A 9
Ground Soy sauce, Taco sauce,
Raspberry drink 0 nutmeg3 0 Brand A 0 Brand B 22
Grape drink <1 Fried egg3 0 Soy sauce,
Hard-boiled Brand B 0 Sweet relish <1
3
Fruit punch <1 egg <1 Pickled
Orange soda <1 Fresh tomato3 0 vegetables <1 Salad peppers <1
Cola soda, Smoked fish Lite syrup
Brand A 0 (chub)3 <1 Citrus salad <1 product <1
Cola soda, Roasted
Brand B 0 peanuts3 <1 8 Foods previously reported to contain naturally occurring benzene
(4-9).

Under the conditions described, analyses of water fortified foods as for the control solutions, and quantitate benzene resi­
with 1 - 2 0 0 ng benzene standard per g resulted in a greater than dues by external calibration. For foods that foam when purged,
95% transfer of benzene from the water to the trap after 1 use an empty, sealed headspace vial as the foam trap; run an
purge-and-trap cycle. Also, the data were linear over the same exit line through the empty vial and another tube out of the
concentration range. A signal-to-noise ratio of ca 10:1 was ob­ empty vial into the transfer line to the Tenax trap. For
served for 5 ng benzene in 5 g water (a concentration of 1 ng/g). nonaqueous foods, stir the food or sweep the surface with the
A concentration of 0.5 ng benzene/g water provided 400 area gas used to purge, and if needed, apply heat to the test vial to
counts, a response that is substantially above that of the blank force more benzene into the vapor phase. Fortify at least 3 rep­
(ca 1 0 0 area counts). licate aliquots in test vials with the benzene standard at ca 0.5x,
Preparation o f foods.—To minimize losses due to volatili­ lx, and 2 x the benzene concentration suspected in the food.
zation, quickly prepare all foods for analysis, and when possi­ Cap, seal, and shake each vial well. Analyze the fortified repli­
ble, take the portion to be tested directly from container. Trans­ cate aliquots by the method of standard additions, subject all
fer 5 or 10 g aqueous food to a tared headspace vial by using a data to linear regression analysis, and determine the benzene
10 mLLuer-Lok® syringe with a 89 mm, 13-gauge hypodermic concentration in each food.
needle, and cap and seal the vial. Transfer 5 g viscous food with
Food Preparation and Analysis by Static Headspace
syringe or spatula, add 5 mL water, and then cap, seal, and
Concentration/Capillary GC/MSD
shake the vial well. Finely divide solid foods with a knife or
with a mortar and pestle, transfer 5 g to a headspace vial, add 5 Finely divide the solids; transfer 10 g of each solid to a head-
mL water, and cap, seal, and shake the vial well. space vial, add 10 mL water, and seal and shake the vial well.
Determination and quantitation by the method o f standard Analyze semisolids and liquids without diluting. For quantita­
additions.—Except for certain foods listed below, analyze tion, fortify a replicate aliquot of food in the test vial with ben-
 M c N e a l E t A l .: J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1217

A B

F igu re 3. S e le c te d ion c h ro m a t o g ra m s fro m h e a d s p a c e a n a ly s is o f a m a n g o : A, a n a ly s is o f 10 g unfortified fruit; B,

a n a ly s is o f fruit fortified with b e n ze n e at 2 ng/g.

zene at the level suspected in the food. Equilibrate each vial 30 eled the procedure after the Environmental Protection Agency
min in a 90°C oven. In same oven, preheat 5 mL Pressure-Lok method 524.2 (EPA 524.2) (10), except that benzene was the
gas syringe. Use gloves to protect hands from hot syringe; then only analyte determined, a flame ionization detector served as
pierce the vial septum and draw up 2 mL of the headspace a general detector, and the method of standard additions was
gases. Leave vial and syringe (with the needle remaining in the used for quantitation to compensate for the many food ma­
vial headspace) in the oven for 3 min. Quickly lock the syringe trixes. The presence of benzene in foods was determined by
valve, remove the syringe from the vial, inject the headspace purge-and-trap headspace concentration/capillary GC/FID.
gas into the gas chromatograph, and begin the data acquisition. Confirmation by static headspace concentration/capillary
GC/MSD was performed, because no purge-and-trap interface
Post GC/MSD Data Analysis
to the gas chromatograph/mass spectrometer was available.
Obtain ion chromatograms for m/z, 52,77, and 78. Integrate Figures 1 and 2 are representative chromatograms of foods
responses, and calculate area ratios normalized to the m/z 78 analyzed by purge-and-trap headspace concentration/capillary
ion (ca 0.20:0.30:1.0). If the area ratios are within ±20% of the GC/FID. The chromatograms in Figure 1 illustrate the sensitiv­
ion ratios from the benzene standard, the presence of benzene ity of the method. Chromatogram A shows the analysis of 10 g
is confirmed. Quantitate benzene by using integrated responses cranberry juice cocktail. Chromatogram B represents the same
for the m/z 78 ion in fortified and unfortified food. Determine juice fortified with 5 ng benzene (0.5 ng/g). By the method of
area response for the amount of benzene added, and calculate standard additions, less than 0.5 ng benzene/g juice was found
the benzene concentration in the food by using a single-point in the cranberry juice cocktail. The peak to the right of benzene,
calibration. at approximately 1 1 min, is fluorobenzene, a component of the
working standard used as an internal standard in related experi­
Results and Discussion ments. The chromatograms in Figure 2 represent the analysis
of a more complex food matrix. Chromatogram B represents
An analytical technique for determining benzene in simple headspace measured over a taco sauce fortified with 50 ng ben­
and complex matrixes at nanogram levels of benzene per gram zene and 50 ng fluorobenzene. We found 22 ng benzene/g un­
of food was needed. The purge-and-trap headspace technique fortified taco sauce (chromatogram A). Data in Table 1 sum­
with capillary GC seemed to offer the most promise. We mod­ marize the levels of benzene found in foods by using
1218 M c N e a l E t A l .: J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

A B
Io n 5 2 . 0 0 amu . -fro m D RTR5 : T flC O R . D Io n 5 2 . 0 0 am u. f r o m D R T R 5 : TR C O B . D

Io n 7 7 . 0 0 am u. -fro m D RTR5 : TR C O R . D Io n 7 7 . 0 0 am u. f r o m D R T R 5 : T R C O B . D

Io n 7 8 . 0 0 am u. -fro m DRTR5 : TR C O R . D Io n 7 8 . 0 0 am u. f r o m D R T R 5 : T R C O B . D

F igu re 4. S e le cte d ion c h ro m a to g ra m s fro m h e a d s p a c e a n a ly s is o f ta c o sa u c e : A, a n a ly s is o f 10 g unfortified ta c o

s a u c e c o n ta in in g 22 n g benzene/g; B, a n a ly s is of ta c o s a u c e fortified with b e nzene at 11 ng/g.

purge-and-trap headspace concentration/capillary GC/FID. naturally occurring benzene levels. To investigate benzene for­
Foods previously reported to contain benzene were analyzed in mation from precursor compounds, we prepared aqueous solu­
duplicate. Analyses were repeated for all foods that showed an tions containing 0.04% sodium or potassium benzoate and
initial result > 1 ng/g. 0.025% ascorbic acid. These concentrations are typical of those
With the mass selective detector operating in the selected used in beverage formulations. Aliquots of the solutions were
ion monitoring mode, a benzene concentration of 1 ng/g could then exposed to strong UV light, 45 °C in an oven, or storage at
easily be confirmed in many foods by the static headspace tech­ room temperature in a dark cabinet. After 20 h either in strong
nique and capillary GC. The same technique was suitable for UV light or at 45°C, about 300 ng benzene/g had formed in
verifying the quantitative results. Figures 3 and 4 are ion chro­ these solutions. Only 4 ng benzene/g formed in those solutions
matograms representing determination of benzene in different stored in the dark at room temperature. After 8 days, the ben­
foods by the static headspace technique. They demonstrate the zene yield from the solutions stored at room temperature in­
selectivity and sensitivity of the MSD procedure. Figure 3 rep­ creased to 266 ng/g, whereas levels remained constant in the
resents the analysis of a mango. In the unfortified fruit (Fig­ heated and UV-treated solutions. Benzene was not found in
ure 3A), benzene ions are absent, but when the fruit was forti­ identically treated control solutions containing benzoate alone
fied at 2 ng/g (Figure 3B), all 3 ions gave responses. The or ascorbic acid alone.
chromatograms of Figure 4 are the results of static headspace Benzaldehyde, benzoic anhydride, and acetophenone at
concentration/capillary GC/MSD analyses of the same taco 0.04% (w/w) in water, when mixed with 0.025% ascorbic acid
sauce represented in Figure 2. Ion chromatograms A represent and subjected to similar test conditions, also yielded trace
the analysis of the unfortified taco sauce. The 3 benzene ions amounts of benzene. For example, we found 74 ng benzene/g
are present in the proper ratios; 2 2 ng benzene/g was found, in the test solution when we substituted benzaldehyde for ben­
which is the same concentration found by using purge-and-trap zoate salts. We found benzene yields of 4 and 5 ng/g of test
headspace concentration/capillary GC/FID. Chromatograms B solution when we tested benzoic anhydride and acetophenone,
represent analysis of the same sauce after fortification with respectively. Toluene, benzyl alcohol, butylated hydroxytolu-
benzene at the 1 1 ng/g level. ene, and butylated hydroxyanisole did not generate benzene in
T h e s e a n a ly tic a l p ro c e d u re s w e re u s e d in s tu d ie s o f th e fo r­ similar tests. When benzoate salts were analyzed for residual
m a tio n o f b e n z e n e fro m p re c u rs o rs a n d in su rv e y s o f fo o d s fo r benzene by the procedure described, no benzene was found.
 M c N eal Et Al .: J ournal O f AOAC International V ol . 76, No. 6,1993 1219

Following the experiments on prepared laboratory stand­ interaction of these 2 compounds. In these cases, removal of
ards, we conducted a limited survey of selected foods. No dif­ one of the compounds may mitigate benzene formation.
ficulties were encountered in the analysis of juices or soft Although the benzene levels found in the foods we analyzed
drinks. However, the analysis of certain foods proved challeng­ were quite low, some soft drink products have already been
ing and required changes in food preparation and purging tech­ reformulated to avoid the formation of benzene from additives.
niques. Solids and viscous foods required dilution with water Other soft drink manufacturers are re-evaluating both their for­
to improve the efficiency of the transfer of benzene from the mulations and their processes to minimize the possibility of
food to the trap, i.e., the percent transfer from foods when com­ benzene formation (3). The limited data developed from this
pared with that from water standards. In addition, 70°C heat study show that, for the most part, the foods containing both
was applied to both fried and boiled eggs to drive the benzene benzoate and ascorbic acid form at most very small concentra­
into the vapor phase. Percent transfer for diluted viscous foods tions of benzene.
ranged from 95% in sauces containing salt and tomatoes to
about 75% in some preserves. The lowest percent transfer, References
about 25%, was obtained with cooked eggs, smoked fish, and
peanuts. (1) Food Chemical News (1990) Feb. 19, pp. 27-28
More than 50 foods were analyzed for benzene by the pro­ (2) Florida Department of Health Bulletin, November 1990
cedure described. Except for the liquid smoke products, in (3) Memorandum of meeting, National Association of Soft
which benzene levels may be source-related, < 2 ng/g was Drink Manufacturers, December 17, 1990
found in certain foods reported to contain naturally occurring (4) Maarse, H., & Visscher, C.A. (1989) Volatile Compounds in
benzene. These data contradict many earlier reports (4—8) of Foods, Vol. II, 6th Ed., TNO-CIVO, Zeist, The Netherlands
high benzene levels in foods such as eggs, mangos, and roasted (5) MacLeod, A.J., & Cave, S.J. (1975) J. Sci. FoodAgric. 26,
peanuts. Although this study represents a limited selectve sam­ 351-360
pling, it appears that some of the earlier reports of benzene in (6) MacLeod, A.J., & Cave, S.J. (1976) J. Sci. FoodAgric. 27,
foods may have been the result of laboratory contamination. 799-806
Little or no benzene (about <1 ng/g) was found :n foods (7) Lovegren, N.V., Fisher, G.S., Legendre, M.G., & Schuller,
containing naturally occurring benzoates and ascorbic acid, W.H. (1979) J. Agric. Food Chem. 27, 851-853
and no correlation could be made. However, the level of ben­ (8) Engel, K.H., & Tresal, R. (1983) J. Agric. Food Chem. 31,
zene found in foods containing added benzoates in addition to 796-801
ascorbates (e.g., imitation strawberry preserves, taco sauce, (9) Stein, V.B., & Narang, R.S. (1990) Arch. Environ. Contam.
and duck sauce) ranged from <1 to 38 ng/g. These findings are Toxicol. 19, 593-596
similar to the survey results reported by Page et al. (11) for (10) U.S. EPA Method 524.2, Revision 3, EPA/600/4-88/039
benzene residues in beverages. Both our survey data and the (1988) U.S. Environmental Protection Agency, Cincinnati,
data we obtained from analyses of prepared laboratory stand­ OH
ards suggest that when benzoates and ascorbates are present (11) Page, B.D., Conacher, H.B.S., Weber, D., & Lacroix, G.
together in foods, the benzene formed is associated with the (1992) J. AOAC Int. 75,334-340
1220 T rotter & D ickerson : J ournal Of AOAC International V o l 76, No. 6,1993

CHEMICAL CONTAMINANTS MONITORING

P e s tic id e R e s id u e s in C o m p o s ite d M ilk C o lle c te d T h r o u g h th e

U .S . P a s t e u r i z e d M ilk N e tw o r k

W i l l ia m J. T r o t t e r

U.S. Food and Drug Administration, Division of Contaminants Chemistry, Washington, DC 20204
R ic h a r d D ic k e r s o n

Woodson-Tenent Laboratories, Inc., Memphis, TN 38101

The U.S. Food and Drug Administration (FDA) has Panama Canal Zone. Milk collected by the PMN through the
implemented a comprehensive monitoring pro­ EPA Environmental Radiation Ambient Monitoring System
gram to determine the incidence and levels of or- (ERAMS) program for radionuclide monitoring represents an
ganohalogen pesticide residues in milk repre­ estimated 80% of the milk delivered to the population centers
senting most of the U.S. supply consumed in and at least 40% of that consumed by the U.S. population (C.
metropolitan areas. Residue findings for 806 com­ Porter, EPA, 1989, personal communication). Radionuclide
posite milks collected through the Pasteurized Milk determinations are conducted at the EPA National Air and Radia­
Program by the U.S. Environmental Protection tion Environmental Laboratory (NAREL) in Montgomeiy, AL.
Agency (EPA) in 1990-1991 are reported. Milk was In 1989, EPA and FDA signed an interagency agreement
collected on a monthly basis from 63 stations se­ that allowed a representative portion of all milk received by
lected by EPA for radionuclide monitoring. These NAREL to be transferred to an FDA contract laboratory
stations provide an estimated 80% of the milk deliv­ (Woodson-Tenent Laboratories, Inc., Memphis, TN) for pesti­
ered to U.S. population centers. At each station, cide and PCB residue analysis. This report describes the estab­
milk from selected sources had been composited lished sampling mechanism, the analytical procedures, and the
to represent the milk routinely consumed in its met­ controls and summarizes the findings from this monitoring in­
ropolitan area. Portions of these composites were itiative for milk collected from May 1990 through July 1991.
forwarded to an FDA contract laboratory for pesti­
cide residue analysis. Pesticide residues were Experimental
found in 398 (49.4%) of 806 test samples, on the ba­
sis of a 0.0005 ppm limit of detection for each resi­ Milk Collection
due on a whole-product basis. A total of 455 occur­
Nalgene bottles (250 mL), each containing 3 mL formalin
rences of pesticide residues were found; p,pf-DDE
added for microbial inhibition, are sent by NAREL to the PMN
and dieldrin accounted for 384 (84.4%) of these oc­
collection sites monthly. (The actual number of collections
currences. The highest level was 0.019 ppm p,pf-
may vary occasionally because of nonparticipation of some
DDE.
sites.) The bottles are filled at the collection sites, mostly by
local Department of Health personnel, with pasteurized whole
milk composited to represent milk consumed in the metropoli­
T he U.S. Food and Drug Administration (FDA) has regu­
tan center. Each composite, along with a separate 1 L portion
latory responsibility for enforcement of pesticide and
for radionuclide analysis, is sent to NAREL, which then for­
polychlorinated biphenyl (PCB) tolerances in foods and
wards the 250 mL test sample to the FDA contract laboratory.
routinely monitors the food supply for residues of these chemi­
Upon arrival at the contract laboratory, test samples are stored
cals (1^4). Because of the importance of milk and milk prod­
at <-10°C. Milk shipments to NAREL and from NAREL to
ucts in the U.S. diet, the FDA, in cooperation with the Environ­
FDA’s contract laboratory are sent by express mail at ambient
mental Protection Agency (EPA), developed this monitoring temperature.
initiative to supplement pesticide surveillance that is ongoing
in other programs. R eagents
Every month the EPA Office of Radiation Programs collects
(a) Solvents.—Ethyl ether, «-hexane, isooctane, methanol,
milk from the Pasteurized Milk Network (PMN), which con­
methylene chloride, petroleum ether; pesticide residue analysis
sists of approximately 63 stations located in large metropolitan
grade (Burdick & Jackson Laboratories, Inc., Muskegon, MI
population centers in the United States, Puerto Rico, and the
49442).
(b) Alumina.—Neutral, activity grade I, 80-200 mesh
Received September 28, 1992. Accepted February 2, 1993.
(ICN Biomedicals, Costa Mesa, CA 92626); heat for 16 h at
 Trotter & Dickerson: Journal Of AOAC International Vol. 76, No. 6,1993 1221

T a b le 1. P e s t ic id e r e s id u e f in d in g s 3 fo r c o m p o s it e m ilk s fro m 6 3 m e tro p o lita n a re a s

No. of milks

Positive for No. of

Metropolitan area Analyzed pesticides findings Residue (m ean ppm /frequency of occurrence)

Albuquerque, NM 14 13 13 p ,p '-D D E [Tr.b/9]; dieldrin [Tr./4]

Atlanta, G A 15 13 22 p p '- D D E [Tr./13]; dieldnn [Tr./4]; hept.ep.c [Tr./2]; chlorpyrifos [Tr./2]; endrin
[Tr./1]
Austin, T X 14 14 14 p p '- D D E [0.003/14]
Baltimore, M D 15 0 0
Boston, MA 15 4 4 p p '-D D E [Tr./1 ]; dieldrin [Tr./3]
Buffalo, N Y 14 4 5 p p '- D D E [Tr./2]; dieldrin [Tr./3]
Burlington, V T 10 9 9 p p '- D D E [0.002/2]; die drin [0.001/6]; lindane [0.001/1]
Charleston, SC 15 2 5 p p '-D D E [Tr./2]; dieldrin [0.001/1]; hept.ep. [0.001/1]; p,p '-D D D [0.001/1]
Charleston, W V 14 10 11 p p '- D D E [0.001/2]; die drin [0.001/8]; hept.d [Tr./1]
Charlotte, NC 14 2 2 p p '- D D E [Tr./1 ]; hept. [Tr./1]
Chattanooga, TN 13 2 5 p ,p '-D D E [Tr./1]; dieldrin [Tr./2]; chlorpyrifos [0.001/1]; hept.ep. [Tr./1]
Chicago, IL 15 14 14 Dieldrin [Tr./14]
Cincinnati, O H 13 4 4 p ,p '-D D E [Tr./1 ]; dieldrin [Tr./3]
Cleveland, OH 13 2 2 Chlorpyrifos [Tr/1]; hept. [Tr./1]
Cristobal, Panam a 13 2 4 p ,p '-D D E [Tr./1]; chlorpyrifos [Tr./1]; H C B [Tr./1]; mirex [Tr./1]
Denver, C O 9 8 a p ,p '-D D E [Tr./7]; dieldrin [0.002/1]
D es Moines, IA 15 12 14 Dieldrin [Tr./11]; chlorpyrifos [0.001/1]; hept.ep. [Tr./2]
Detroit, Ml 15 0 0
Dover, DE 2 1 1 Endrin [Tr./1]
Fort W orth, T X 14 0 0
Grand Rapids, Ml 14 12 12 p ,p '-D D E [Tr./2]; dieldrin [Tr710]
Hartford, C T 10 2 2 p ,p '-D D E [Tr./1]; hept.ep. [Tr71]
Helena, M T 14 1 1 Chlorpyrifos [Tr./1]
Honolulu, HI 14 14 14 p ,p '-D D E [Tr./14]
Idaho Falls, ID 15 15 10 p ,p '-D D E [0.002/15]; hept. [T r/1]
Indianapolis, IN 15 15 19 p ,p '-D D E [0.001/1]; dieldrin [Tr./14]; chlorpyrifos [0.001/2]; hept. [Tr./1];
hept.ep. [Tr./1]
Jackson, M S 15 5 3 p ,p '-D D E [Tr./5]; dieldrin [Tr./2]; chlorpyrifos [Tr./1]
Kansas City, MO 12 2 4 Dieldrin [Tr./3]; H C B [T r./1]
Knoxville, TN 14 13 14 Dieldrin [Tr./12]; chlorpyrifos [0.001/1]; lindane [Tr./1]
Las Vegas, NV 13 13 14 p ,p '-D D E [0.003/13]; hept. [Tr./1]
Little Rock, AR 15 2 2 Dieldrin [Tr./1]; chlorpyr fos [Tr./1]
Los Angeles, C A 15 15 16 p ,p '-D D E [0.003/15]; HCB [Tr./1]
Louisville, KY 14 4 4 Dieldrin [Tr./2]; hept.ep. [0.001/1]; p .p '-D D T [Tr./1]
Manchester, NH 7 3 3 p ,p '-D D E [Tr./1 ]; dieldrin [Tr./2]
Memphis, TN 14 3 4 Dieldrin [Tr./2]; a -B H C [0.002/1]; H C B [Tr./1]
Minneapolis, M N 2 2 3 p ,P '-D D E [Tr./2]; chlorpyrifos [Tr./1]

Minot, N D 15 0 0
Montgomery, AL 14 2 2 p ,p '-D D E [Tr./1 ]; chlorpyrifos [Tr./1]
Montpelier, V T 5 4 4 p ,p '-D D E [Tr./1 ]; dieldrin [0.001/3]
N ew Orleans, LA 5 1 1 Chlorpyrifos [Tr./1 ]
N ew York, N Y 13 10 11 p,p '-D D E [Tr./1]; dieldrin [0.001/9]; chlorpyrifos [Tr./1]

Norfolk, VA 13 2 2 Dieldrin [Tr./2]

O klahom a City, O K 11 2 2 p ,p '-D D E [0.001/1]; a -B H C [Tr./1]

O m aha, NE 14 8 10 Dieldrin [Tr./6]; chlorpyrifos [Tr71]; hept.ep. [Tr./3]

Philadelphia, PA 15 13 14 p ,p '-D D E [Tr./4]; dieldnn [0.001/9]; chlorpyrifos [Tr./1]

Phoenix, A Z 14 14 14 p ,p '-D D E [0.012/14]

Pittsburgh, PA 14 14 15 p ,p '-D D E [Tr./2]; dieldrin [Tr./11]; a -B H C [Tr /1 ]; p ,p '-D D T [Tr./1]

P o ila n d , M E 14 0 0
Portland, OR 15 13 19 p ,p '-D D E [Tr./12]; dieldrin [Tr./5]; chlorpyrifos [0.002/1]; H C B [Tr./1]

Rapid City, SD 15 1 2 Chlorpyrifos [0.002/1]; HCB [0.001/1]

Riverton, W Y 2 0 0
Sacram ento, CA 14 14 14 p ,p '-D D E [Tr./14]
1222 T rotter & D ickerson : Journal Of AOAC International V ol . 76, No. 6,1993

T a b le 1. C o n tin u e d

No. of milks

Positive for No. of

Metropolitan area Analyzed pesticides findings Residue (m ean ppm /frequency of occurrence)

San Francisco, CA 15 15 15 p,p '-D D E [0.002/14]; dieldrin [0.002/1]

San Juan, PR 15 0 0
Seattle, W A 12 11 13 p ,p '-D D E [Tr./11 ]; dieldrin [Tr./1 ]; hept.ep. [Tr./1]
Spokane, W A 13 4 4 p,p '-D D E [Tr./2]; chlorpyrifos [Tr./1]; H C B [Tr./1]
St. Louis, M O 14 12 15 Dieldrin [Tr./11]; hept.ep. [Tr./4]
St. Paul, M N 13 0 0
Syracuse, NY 15 6 6 p ,p '-D D E [0.001/1]; dieldrin [Tr./3]; chlorpyrifos [Tr./1 ]; p,p '-D D T [Tr./1]
Tam pa, FL 14 7 10 p ,p '-D D E [Tr./6]; chlorpyrifos [Tr./1]; hept. [Tr./T]; m ethoxychlor [0.002/2]
Trenton, NJ 13 2 2 p,p '-D D E [Tr./2]
Wichita, KS 13 2 2 Dieldrin [0.002/1]; H C B [T r./1]
Wilmington, DE 11 4 5 p ,p '-D D E [Tr./1 ]; dieldrin [Tr./2]; hept. [Tr./1]; lindane [Tr./1]
Total 806 398 455

a No PCB residues w ere found at or below the PCB limit of quantitation of 0.0 2 ppm, based on G C response to A rodor 1254.
b Tr. = trace value, i.e., below the pesticide limit of quantitation of 0.001 ppm and above or equal to the pesticide limit of detection of 0 .0 0 0 5
ppm.
c hept.ep. = heptachlor epoxide.
d hept.= heptachlor.

260°C, cool to room temperature, and add sufficient water to program: 160X1 for 3.5 min, 7°C/min to 235X1, and 235X1 for
produce alumina-water (85 + 15 w/w). 4 min. Column 1 is used for quantitation; column 2 is used as
(c) Florisil.—60-200 mesh (Curtin-Matheson Scientific, the primary confirmatory column; and column 3 is used for
Marietta, GA 30067); heat for 16 h at 260°C before use. additional confirmation if needed.
(d) Potassium oxalate.—Certified ACS grade (Fisher Sci­ (f) Milk shipping containers.—250 mL Nalgene bottles
entific, Pittsburgh, PA 15219). (Northwest Bottle, Memphis, TN 38118).
(e) Sodium sulfate.—Anhydrous granular (12-60 mesh) (g) Centrifuge.—Model CRU-5000 (Damon/IEC, Need­
for pesticide residue analysis (J.T. Baker, Phillipsburg, PA ham, MA 02194).
08865).
Extraction and Cleanup
Apparatus
Weigh 50 g milk into 250 mL centrifuge bottle. Add 0.25 g
(a) Gelpermeation chromatograph.—Model 1002A(Ana­ potassium oxalate and 50 mL methanol. Stopper and then
lytical Bio-Chemistry Laboratories, Inc., Columbia, MO shake centrifuge bottle until potassium oxalate dissolves. Add
65205). Calibrate according to the method described by Hop­ 100 mL ethyl ether-petroleum ether (1 + 1), and shake bottle
per (5). vigorously 1 min. Centrifuge at ca 1500 rpm for 10 min.
(b) Gel permeation column.—30 x 2.5 cm id, packed with Pipet top layer (extract) into 500 mL separatory funnel con­
a slurry of 33 g Bio-Beads SX-3 resin, 200-400 mesh (Analyti­ taining 250 mL distilled or deionized water and 0.5 g potassium
cal Bio-Chemistry Laboratories, Inc.), and compressed to a bed oxalate; allow organic layer to settle. Add 2 more 50 mL por­
length of 2 0 cm. tions of ethyl ether-petroleum ether ( 1 + 1 ) to centrifuge bottle;
(c) Eluting solvent.—Methylene chloride-hexane (1 + 1) after each addition, shake bottle vigorously 1 min, centrifuge
pumped at a rate of 5.0 mL/min with operating pressure of 7 - 10 min, and pipet top layer into 500 mL separatory funnel; al­
15 psi. low organic layer to settle. Gently mix combined extract in
(d) Gas chromatograph.—Model 5840 (Hewlett-Packard, separatory funnel. (Caution: Vigorous shaking may result in
Avondale, PA 19311) equipped with 63Ni electron-capture de­ unbreakable emulsion.)
tector such that 0.1 ng chlorpyrifos produces 50% full-scale Drain and discard lower (aqueous) layer. Wash extract with
deflection. 200 mL water by gently shaking funnel; let layers separate and
(e) Gas chromatographic columns.—Supelco, Inc. (Belle- discard aqueous portion.
fonte, PA 16823). (7) 8 ft x 4 mm id glass, packed with 10% Drain organic layer through funnel containing glass wool
OV-101 on 80-100 mesh Supelcoport and operated isother- plug and 4 g Na2 S 0 4 into beaker. Rinse Na2 S 0 4 with 10 mL
mally at 230°C; (2) 6 ft x 2 mm id glass, packed with 1.95% ethyl ether-petroleum ether (1 + 1). Gently concentrate extract
SP-2250/1.95% SP-2401 on 100-120 mesh Supelcoport and to ca 0.5 mLby using 40X1 water bath and stream of nitrogen.
operated isothermally at 215°C; and (3) 15 m x 0.53 mm SPB- Add 3 mL n-hexane-methylene chloride (1 + 1) to beaker.
608 fused silica with 0.5 pm film thickness and temperature Mix and transfer to 15 mL centrifuge tube. Add 2 more 3 mL
 Trotter & D ickerson : J ournal O f AOAC International V ol . 76, No. 6,1993 1223

T a b le 2. P e s t ic id e r e s id u e s fo u n d in 8 0 6 m ilk pretested to demonstrate that plasticizers and other potential

c o m p o s it e s interfering materials would not be introduced into the milk dur­
Residue Frequency of occurrence ing extended storage periods. Before use in this survey, 8 ran­
domly selected bottles containing preanalyzed milk and an
p,p '-D D E 212 added preservative were allowed to stand at room temperature
Dieldrin 172 for 2 to 10 days. Reanalysis of the milk after storage demon­
Chlorpyrifos 23 strated that nc detectable interfering materials were introduced.
Heptachlor epoxide 17 Control (blank) and fortified milks were analyzed concur­
Heptachlor 8 rently with each batch of composite milks. Fortification solu­
HCB 8 tions, containing chlorinated pesticide mixtures or Aroclor
a -B H C 3 1254, were added to duplicate portions of control milks before
Lindane 3 the extraction step. A total of 17 pesticides and Aroclor 1254,
p,p '-D D T 3 each at 3 concentrations, were included in the fortification so­
Endrin 2 lutions. All 13 pesticides that were eventually identified in the
Methoxychlor 2
composite milks were included in the fortification solutions.
p,p '-D D D 1
Control limits of 80-110% recovery were established for each
Mirex 1
pesticide and Aroclor 1254. Reanalysis of the entire batch of
Total 455 milks was required if a concurrent recovery was not within the
control limits or if excessive interferences were encountered in
the control.

portions of n-hexane-methylene chloride ( 1 + 1 ) to beaker; af­ Results and Discussion

ter each addition, mix and transfer to 15 mL centrifuge tube to
ensure quantitative transfer. Adjust volume to 10 mL, mix, cap Table 1 lists analytical findings for composite milks col­
centrifuge tube, and centrifuge at ca 1500 rpm for 5 min. lected from 63 major U.S. metropolitan areas from May 1990
Transfer 5 mL aliquot of extract from centrifuge tube (use through July 1991. There were 455 occurrences of pesticide
Luer-Lok syringe equipped with 0.45 |im filter if solution ap­ residues in 398 (49.4%) of the 806 milks tested; no PCB resi­
pears cloudy) to gel permeation chromatograph (GPC) 5 mL dues were found. Of the 13 pesticide residues that were identi­
sampling loop. Take care not to disturb sediment at bottom of fied, p,p'-DDE (212 findings) and dieldrin (172 findings) rep­
culture tube. The GPC dump, collect, and wash timing cycles resented 84.4% of the total.
are 12,23, and 0 min, respectively. Collect 135 mL extract from Table 2 summarizes the frequency of occurrence of each
GPC in 250 mL beaker. pesticide found. For this survey, limits of detection (LD) and
Gently concentrate extract to ca 20 mL and quantitatively quantitation (LQ) for all pesticide residues were defined as
transfer to 50 mL graduated centrifuge tube with additional 0.0005 and 0.001 ppm, respectively, on a whole-milk basis.
portions of petroleum ether to final volume of 40 mL. Concen­ The LD was based on a detector sensitivity set to produce a
trate to 1-2 mL, using stream of nitrogen and 40°C water bath. 50% full-scale response for 0.1 ng chlorpyrifos injected in a
Add petroleum ether to final volume of 40 mL and again con­ volume equivalent to 25 mg milk. At these settings, a detect­
centrate to 1-2 mL. able response (5% full-scale response) was obtained for the
If gas chromatographic (GC) interferences are present, elute later-eluting chemicals of interest (e.g., p,p'-DDT and mirex) if
concentrate with 35^40 mL petroleum ether through 10 mm id present at a level of 0.0005 ppm. Detector responses for the
column containing 5 g alumina. (In our study only ca 20% of early-eluting chemicals (e.g., HCB and a-BHC) produced
the extracts required alumina cleanup.) If interference persists, theoretical LDs of <0.0005 ppm; however, the generalized LD
remove by Florisil chromatography (6 , 7). (About 5% of the and LQ values were used to provide uniformity. A trace was
extracts required Florisil cleanup.) defined as a value >LD but <LQ. The PCB LD was established
as 0.01 ppm, on the basis of the response to Aroclor 1254. In
Concentrate or dilute extracts to appropriate concentration
Table 1, trace values were assigned the numerical value of the
for GC analysis. Inject ca 5 pL test solution, which is equiva­
LD (0.0005 ppm) for determining averages. If the calculated
lent to ca 25 mg whole milk when extract is diluted to 5 mL.
average was >LQ, a numerical average was assigned; if the
Identify responses from gas chromatograms by matching re­
resultant average was <LQ but >LD, the average was consid­
sponse times with those of appropriate reference standards; use
ered a trace.
multiple columns as necessary. Compare chromatographic re­
sponse with that of appropriate reference standard to quantitate Of the 212 p,p'-DDE and the 172 dieldrin findings, the high­
est levels found were 0.019 ppm (Phoenix, AZ) and 0.003 ppm
residue level.
(Indianapolis, IN), respectively. None of the other pesticides
Quality A ssurance exceeded 0 . 0 0 2 ppm.
All 14 milks from Phoenix, AZ, contained p ,p -DDE (mean
For ease of handling, plastic containers were chosen for = 0.012 ppm). All 15 milks from Los Angeles, CA, 14 from
storing and shipping milk collections. These containers were Austin, TX, and 13 from Las Vegas, NV, contained p,p'-DDE
1224 T rotter & D ickerson : J ournal O f AOAC International V ol . 76, No. 6,1993

T a b le 3. M e th o d p e rfo r m a n c e d a ta f o r te s t s a m p le s a n a ly z e d c o n c u r r e n t ly w ith m ilk c o m p o s it e s 3

Amount Amount
Chemical added, ppm Av. rec., % n RSD, % Chemical added, ppm Av. rec., % n RSD, %

a-B H C 0 .0 007 98 2 13.0 Endrln 0.0011 89 2 5.5

0 .0 013 86 12 7.6 0 .0 0 2 3 91 12 6.7
0 .0 033 92 20 5.8 0 .0 0 5 7 99 20 5.9
HCB 0.0 005 91 2 2.3 p,p '-D D D 0.0020 100 2 6.4
0.0010 88 12 7.1 0 .0 040 95 12 7 .0
0 .0 026 89 20 7.6 0 .0 0 9 9 102 20 4 .7
Lindane 0.0010 96 2 3.7 p ,p '-D D T 0 .0 0 4 2 103 2 2.0
0.0020 87 12 7.1 0 .0 083 97 12 7 .6
0.0051 94 20 4.3 0 .0 208 105 20 5 .6
Heptachlor 0.0011 95 2 4.5 Methoxychlor 0 .0 0 6 3 103 2 8.2
0.0022 89 12 5.5 0 .0 1 2 7 92 12 6 .3
0 .0 055 96 20 5.6 0 .0 3 1 7 103 20 7 .6
Chlorpyrifos 0.0010 94 2 8.3 Mi rex 0 .0 057 96 2 8.1
0.0020 90 12 6.7 0.0011 93 12 6 .5
0 .0 050 100 20 5.1 0 .0 284 99 20 5.9
Aldrin 0.0010 103 2 6.9 a-Chlordane 0 .0 0 2 4 91 15 9.0
0.0020 88 12 8.5 0 .0 0 3 6 93 3 5 .3
0 .0 052 96 20 6.3 0 .0 060 92 17 4 .9
Heptachlor epoxide 0.0012 104 2 2.0 y-Chlordane 0.0020 90 15 9.5
0 .0 023 93 12 8.3 0 .0 0 3 0 95 3 6.7
0 .0 059 99 20 4.7 0 .0 0 5 0 92 17 5 .7
p,p '-D D E 0.0010 110 2 0.0 Octachlor epoxide 0.0022 85 15 8 .3
0.0021 100 12 6.8 0 .0 0 3 3 96 3 4 .2
0 .0 052 101 20 3.9 0 .0 056 95 17 6.2
Dieldrin 0.0011 107 2 1.3 Aroclor 1254 0 .0 2 1 6 92 4 2 .7
0.0022 94 12 9.7 0 .0 4 3 2 97 10 9.5
0.0 055 100 20 7.2

Control milks (n = 7 2 ) were analyzed concurrently with test samples; 6 8 controls dem onstrated no interferences, but G C chrom atogram s of 4
showed trace levels of interference at the retention tim e of lindane.

at identical mean values o f0.003 ppm. No residues were found with the GC column used for quantitation, but this GC response
in 7 metropolitan areas (Baltimore, MD; Detroit, MI; Fort did not match the response of lindane on the confirmatoiy col­
Worth, TX; Minot, ND; Portland, ME; San Juan, PR; and St. umns. Responses of coeluting components obtained from milk
Paul, MN) over the 15-month testing period. and fortified-milk extracts that were analyzed concurrently
Table 3 summarizes the concurrent quality assurance data with these 4 controls were corrected for these small interfer­
generated during the course of these analyses. All 13 analytes ences; these extracts were not reanalyzed. All analytes identi­
that were found in the survey were included in the recovery fied for the first time and all residues determined at >0.003 ppm
studies along with Arcolor 1254 and 4 other pesticides that had were confirmed qualitatively and quantitatively by comparing
a high likelihood of occurrence. Recoveries of each of the 13 GC retention time and residue response with those of the cor­
analytes found in the survey were determined 34 times, and all responding reference standard on either or both of the 2 confir­
results were within the predetermined control limits of 80- matory columns.
110% of the fortified amounts. Two recoveries each of a-BHC None of the 806 composite milks contained pesticide or
and HCB were determined at the true LQ (0.0007 and 0.0005 PCB levels that approached established EPA or FDA regula­
ppm, respectively) in addition to levels equal to or above the tory limits. The residues that were found reflect the persistence
generalized LQ of 0.001 ppm. These lower LQs were readily of the chlorinated hydrocarbon pesticides that are currently
obtained because of the relatively high detector response to prohibited for use in the United States. Because PMN compos­
these early-eluting chemicals. Control milks were analyzed 72 ite milks substantially represent the U.S. milk supply, this study
times, and only 4 contained a potential interference. For each is a valuable supplement to FDA’s regulatory monitoring and
of these 4 milks, a small (equivalent to trace) peak appeared at Total Diet programs (1—4), which have reported similar low-
the retention time of lindane in the chromatogram obtained level findings.
 Y eung & N ewsome : J ournal Of AOAC International V ol . 76, No. 6,1993 1225

References (4) Food and Drug Administration Pesticide Program (1991) J.

Assoc. Off. Anal. Chem. 74, 121A-140A
(1) Food and Drug Administration Pesticide Program (1988) J. (5) Hopper, M.L. (1982) J. Agric. Food Chem. 30, 1038-1041
Assoc. Off. Anal. Chem. 71, 156A-174A (6) Pesticide Analytical Manual, 2nd Ed., Vol. I (revised 1968
(2) Food and Drug Administration Pesticide Program (1989) J. and subsequent revisions), U.S. Food and Dmg Administra­
Assoc. Off. Anal. Chem. 72, 133A-152A tion, Washington, DC
(3) Food and Dmg Administration Pesticide Program (1990) J. (7) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
Assoc. Off. Anal. Chem. 73, 127A-146A lington, VA, sec. 983.21

CHEMICAL CONTAMINANTS MONITORING

S u rv e y o f T o ta l T e tr a h y d r o p h th a lim id e in B a b y F o o d s U s in g

B o th E n z y m e -L in k e d Im m u n o s o rb e n t A s s a y a n d G a s

C h ro m a to g ra p h y /M a s s S p e c tro m e try : A C o m p a ra tiv e S tu d y

J u p it e r M. Y e u n g and W. H a r v e y N ew so m e
Health and Welfare Canada, Health Protection Branch, Food Directorate, Bureau of Chemical Safety, Food Research
Division, Ottawa, ON, Kl A 0L2, Canada

An enzyme-linked immunosorbent assay (ELISA) tration of THPI in these samples can then be veri­
method was compared with a gas chroma- fied using the GC/MS method.
tographic/mass spectrometric (GC/MS) method for
determining the concentration (in parts per million)
of the combination of captan and its degradation C aptan (A-trichloromethylthio-4-cyclohexene-l,2-dicar-
product tetrahydrophthalimide (THPI) in 13 fruit bo xiiride) is a broad spectrum, non-systemic fungicide
samples and in a survey of baby foods. Ninety widely used for the protection of many fruit and vege­
table crops against fungal pathogens. In Canada, the maximum
baby foods (49 fruits, 28 juices, and 13 vegetables)
residue limit for captan is 5.0 ppm in apples, apricots, blueber­
from 2 different suppliers were sampled. All captan
ries, cranberries, cherries, grapes, peaches, pears, plums, rasp­
in the samples was converted to THPI before each
berries, strawberries and tomatoes (1). Gilvydis et al. (2) re­
analysis. None of the samples contained a concen­
ported residues of captan in a survey of treated strawberry and
tration of combined captan and THPI that violated grape crops and in all strawberry samples. In addition, they
the maximum residue limit of 5.0 ppm. Eight sam­ found captan residues to be stable on samples that had been
ples of baby food tested positive for THPI at levels stored frozen. In a field trial, the kinetics of captan and tetrahy­
ranging from 0.019-0.041 ppm by the GC/MS drophthalimide (THPI) in strawberries were studied. Captan
method, whereas 20 samples tested positive in the was found to be stable on the fruit, and the ratio of THPI to
ELISA assay. All samples that tested positive with captan was consistently 0.036 (3). However, captan has been
the GC/MS method also tested positive with the shown to undergo extensive degradation, presumably to THPI,
ELISA method. Thirteen percent of the baby food when apples (4), strawberries (5), or tomatoes (6 ) containing
samples tested false positive with the ELISA residues were subjected to cooking. The present study was con­
ducted to determine the concentration of combined captan and
method. The ELISA assay also gave higher values
THPI in infant foods that had been subjected to heat during
than the GC/MS method. The ELISA method can be
processing.
effectively used as a primary screening tool to se­
Enzyme-linked immunosorbent assay (ELISA) methods
lect samples testing positive for THPI. The concen-
are well suited for screening purposes, such as in surveying a
large number of samples for analytes that are difficult to deter­
Received November 9, 1992. Accepted January 27, 1993 mine by conventional chemical procedures (7,8). The growth
1226 Y eung & N ewsome : J ournal O f AOAC International V ol . 76, No. 6,1993

F ig u re 1. G a s c h ro m a to g ra m s o f (A ) s ta n d a rd T H P I in e th y l a c e ta te a t a c o n c e n tra tio n o f 0.030 |ig/m L, (B ) b a n a n a

s a m p le w ith n o T H P I d e te c te d , a n d (C ) p e a c h s a m p le w ith 0 .0 2 6 p p m T H P I d e te c te d . A r r o w in d ic a te s th e re te n tio n tim e
f o r T H P I a t 9.38 m in . A b u n d a n c e s c a le s fo r s ta n d a r d a n d s a m p le s w e re d iffe re n t. A 10 m g p o rtio n o f b a b y fo o d
e q u iv a le n t w a s in je c te d fo r s a m p le s (B ) a n d (C).

in the use of the ELISA method has been rapid in recent years foods, all with different lot numbers but obtained from 2 sup­
due to its simplicity, sensitivity, and selectivity. We have pre­ pliers (Heinz and Gerber), and 13 samples of fruits were pur­
viously reported the production of polyclonal antibodies spe­ chased from grocery stores in the Ottawa area. All samples
cific for THPI and described a validated ELISA procedure for were randomly assigned a number. Both ELISA and GC/MS
fruits using these antisera (9). In the present study, a gas chro- analyses were performed on the same methanol extract to
matographic/mass spectrometric method (GC/MS) for quanti­ eliminate sampling variation.
fying THPI is characterized, and the results obtained by it are
compared to those obtained by the ELISA method in a survey ELISA
of foods.
The sample preparation and ELISA protocol have been re­
METHOD ported previously (9). Briefly, 10 g fruit or baby food or 20 mL
baby food juice were refluxed 20 min in 60 mL methanol to
Materials convert any captan present to THPI. After the solids were re­
moved by filtration, the filtrate volume was increased to 1 0 0
Captan standard was obtained from the pesticide repository mL with methanol.
of the Food Research Division, and THPI was purchased from Five milliliters of this crude extract was partitioned with an
Aldrich Chemical Co. (Milwaukee, WI 53201). Both were equal volume of 1 0 % ethyl ether in hexane prior to use in the
stated to be at least 99% pure by the respective manufacturers. ELISA. A 100 (iL aliquot of the final purified methanol extract
All reagents, buffers, and instruments that were used were ob­ was dried overnight, and the residue was taken up in 25 fiL
tained as previously reported (9, 10). Ninety samples of baby PBS. One milliliter of the diluted antiserum was added to the
 Y eung & N ewsome : J ournal Of AOAC International V ol . 76, No. 6,1993 1227

T a b le 1. R e c o v e r y o f T H P I fro m 5 s a m p le s o f T a b le 3. C o m p a r is o n o f th e E L I S A a n d G C / M S m e th o d s
s tra w b e rrie s b y th e G C / M S m e th o d f o r d e te rm in in g th e c o n c e n tra tio n o f T H P I in 13 fru it
s a m p le s
T H P I found, ppm®
Concentration of T H P I, ppm
Sam ple Blank 0.5 1.5 3.5
S am ples3 ELISA G C /M S
Strawberries A 0.6 07 0.4 10 1.372 3.2 54
Strawberries B 0.2 43 0.431 1.225 3.3 10 Strawberries A 0.8 37 0.6 07
Strawberries C 0.2 77 0.4 24 1.228 3.1 47 Strawberries B 0.241 0.2 43
Strawberries D 0.2 66 0.415 1.320 3 .3 0 3 Strawberries C 0.1 67 0.2 77
Strawberries E 0 .2 7 4 0.4 00 1.233 3.4 48 Strawberries D 0.2 24 0.265
Recovery (%) 8 3 .2 ±2 .2 8 5 .0 +4.0 94.1 ±2.8 Strawberries E 0.181 0.2 39
Raspberry A 0.467 0.5 86
a Values for the fortified sam ples are corrected for blanks. Recovery
Raspberry B nd 0.0 43
data are the m ean percent recoveries ± s.d. of 5 sam ples from
different suppliers spiked at 0.5 , 1.5, and 3 .5 ppm of TH P I. Raspberry C nd nd
Raspberry D nd nd
Apple 0.175 0.135
Grape nd 0.033
test sample, which was incubated and then transferred in trip­
Apricot nd 0.071
licate to the coated plate. After further incubation and v/ashing,
Nectarine nd nd
a secondary antibody, goat anti-rabbit IgG peroxidase conju­
gate, was added. Following further incubation and washing, the a Strawberries and raspberries were purchased from different
substrate, o-phenylenediamine dihydrochloride, and H 2 0 2 suppliers. Va ues are the m ean of 2 repeat determinations,
expressed in ppm. Minimum detection limits for the E LISA and
were added. Thirty minutes later, the color reaction was G C /M S methods are 0 .1 5 0 and 0 .0 1 5 ppm, respectively: nd =
stopped and the optical densities were read at 492 nm. below the detection limit, r2 = 0 .8 8, slope = 0.80.

GC/MS
Sample preparation for the GC/MS procedure was similar
A Hewlett Packard 5890 gas chromatograph equipped with
to that for the ELISA assay, except that a more vigorous
an autosampler and on-column injector was coupled to a
cleanup was necessary for the GC/MS method. Thus, a 25 mL
Hewlett Packard 5970 mass selective detection (MSD) system.
aliquot of the crude methanol extract was partitioned with
Separations were carried out using a i m deactivated, uncoated,
25 mL hexane. After discarding the hexane layer, 25 mL water
fused silica retention gap coupled to a 30 m x 0.25 mm id DB-5
and 5 mL saturated NaCl were added. This solution was ex­
column (J & W Scientific, Folsom, CA 95630) with a 0.25 pm
tracted 3 times with 25 mL CH2 C12. The final CH 2 C12 extract
film thickness. Operating conditions: injection port tempera­
ture, 80X1 for 0.5 min then programmed to 280°C at 100°C/min was filtered through anhydrous Na2 S 0 4 and evaporated at
and held at 280°C for 5 min; oven temperature, 80°C for 6 min 35 °C in vacuo to dryness. The sample was then transferred first
then programmed to 230°C at 50X/min and held at 230°C for with 1 mL ethyl acetate and then with 1 mL hexane to an 8 x
10 min; inlet pressure of 15 psi; electron impact at 70 eV; dwell 160 mm column packed with 2 g silica, and eluted from the
time at 50 msec, SIMS at m/z 79 and 151, and m/z 151 for the column with 40 mL methanol:ethyl acetate:hexane solution at
M+’ of THPI was used as the quantitating ion. The THPI reten­ a ratio of 3:25:72. The sample was taken to dryness and then
tion time was 9.38 min. reconstituted in 2.5 mL ethyl acetate. One microliter of the re­
constituted sample was injected into the gas chromatograph. A
T a b le 2. W ith in - d a y a s s a y a n d b e tw e e n -d a y a s s a y 7-point standard curve of THPI in ethyl acetate, ranging in con-
c o e ff ic ie n t s o f v a ria tio n w ith th e G C / M S m e th o d f o r 4
fru its T a b le 4. P e rc e n t r e c o v e r y o f T H P I fr o m s e le c t e d b a b y
f o o d s b y th e E L I S A a n d G C / M S m e th o d s 9
Betw een-day
W ithin-day assay® assay* T H P I recovered, %
T H P I spiked,
T H P I found, ELISA G C /M S
Sam ples ppm
Fruit ppm c CV, % CM, %

Apple sauce 0.25 89.1 85.6

Raspberry 0.042 3.5 7.6
0.50 90 .3 89.6
Apple 0.135 1.9 7.4
1.00 96.1 88.3
Grape 0.0 33 4.3 6.1
Apple juice 0.25 95 .3 94.2
Apricot 0.071 5.5 9.1
0.50 95.7 93.9

a n= 6 1.00 98.2 95.3

b n= 3
c The average incurred concentration of T H P I residue found in each a No incurred TH P I w as detected in blank samples by either the
fruit. E LISA or G C /M S methods.
1228 Y e u n g & N e w s o m e : Jo u r n a l O f A O AC I n t e r n a t io n a l V ol . 76, N o . 6 ,1 9 9 3

Table 5. ELISA and GC/MS quantification of T H P I in baby foods

Solid Juice

Sam ples Positive3 E LISA 0 G C /M S 6 Positive3 E LISA 6 G C /M S 6

Banana 1/4 0 .1 5 7 (nd) nd

Apricot 1/3 0.171 (nd) nd
Pears 1/6 0 .1 7 7 (nd) nd 0/4 nd nd

Peaches 2/7 0 .1 9 9 (0.189) 0.0 26 0/2 nd nd

0 .3 4 8 (nd) nd
Apple 2/3 0 .2 3 6 (0.183) 0.032 0/3 nd nd
0 .2 1 4 (nd) nd
Mixed fruits 1/10 0 .2 3 4 (nd) nd 1/2 0.171 (nd) nd

G reen beans 4/4 1.0 88 (0.160) 0.0 36

1.049 (0.199) 0.0 34
1.755 (0.191) 0.0 27
1.7 66 (0.155) 0.0 26
S w eet potatoes 1/1 0 .2 7 3 (0.218) 0.0 19
Mixed vegetables 1/2 0 .2 7 7 (nd) nd
Apple raspberry 1/2 0 .1 8 2 (0.181) 0.041 1/2 0.5 86 (0.470) nd

Apple grape 1/3 0.1 89 (nd) nd

Apple orange 1/1 0.2 38 (nd) nd
Apple pineapple 1/2 0.2 47 (nd) nd
Apple apricot 0/1 nd nd
Apple banana 0/2 nd nd
Apple cherry 0/3 nd nd
Apple prune 0/1 nd nd
Apple peach 0/1 nd nd
Apple raisin 0/2 nd nd
Pineapple pear 0/1 nd nd 0/1 nd nd
Pears carrots 0/1 nd nd
O range 0/3 nd nd
Prunes 0/3 nd nd
Blueberries 0/3 nd nd
Strawberries 0/4 nd nd
Corn 0/2 nd nd
Carrots 0/2 nd nd
Squash 0/1 nd nd

a Values are the number that tested positive by the E LISA method over the total num ber of baby food samples of each type tested.
b Average values were expressed in ppm, and the values in parenthesis were obtained from the split samples that underwent cleanup steps
normally used for the G C /M S method. Ninety baby foods (28 juices, 13 vegetables, and 49 fruits) from 2 different suppliers w ere sampled.
Eight samples tested positive with the G C /M S method, w hereas 20 sam ples tested positive with th e E LISA method, nd = below detection
limit. Minimum detection limits for the ELISA and G C /M S methods w ere 0 .1 5 0 and 0 .0 1 5 ppm, respectively.

centration from 0.015 to 1 jig/mL, was run with every 10 sam­ comparative study of this ELISA method with a GC/MS
ples. The concentration of THPI in each sample was calculated method on a survey of combined captan and THPI in commer­
from the linear regression of the standard curve. cial fruit and baby food preparations.
To monitor the recovery by GC/MS, THPI was spiked at 3 The GC method for the detection of THPI was similar to that
levels into homogenized strawberry and baby food samples. reported by Schoen and Winterlin (11), except we used a cap­
These fortified samples were incubated 30 min at room tem­ illary column and an MSD system instead of a packed column
perature prior to extraction.
and a nitrogen-phosphorous detector. The modified liquid-liq­
uid extractions and silica column cleanup steps were necessary
R esu lts a n d D iscu ssio n to obtain good chromatography (clean separation without in­
terference) in the food matrixes (Figure 1). We have demon­
We have recently reported a validated ELISA procedure for strated the complete conversion of captan to THPI when captan
the determination of the concentration of captan and THPI as was refluxed in methanol in our experimental conditions (9 ).
THPI in various commodities (9). The present paper reports a Therefore, only THPI was used to spike the samples in our
 Y e u n g & N e w s o m e : Jo u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3 1229

recovery study. The mean recovery of THPI with the GC/MS more false positives were obtained (data not shown), indicating
method at 3 levels of fortification in strawberry samples was a matrix effect caused by non-specific binding. This phenome­
87.4% (Table 1), whereas 99.5% mean recovery was pre­ non (false positive results) was not specific to the fruit or to the
viously obtained with the ELISA method (9). However, when physical state, whether it was solid or liquid. However, when
the standards were also subjected to the same cleanup proce­ the false positive samples were re-analyzed with the ELISA
dure, the recovery was 99.7%. The within-day assay coeffi­ method, using the extract that had undergone the vigorous
cients of variation (CV) were 1.9-5.5%, whereas the between- cleanup steps normally used for GC/MS, all of the false posi­
day assay CV were 6 .1-9.1% on 4 fruit commodities with tive results, except one, were eliminated (Table 5). However,
incurred residues (Table 2). The minimum detection limit for some matrix effect was still evident because the results ob­
the GC/MS method, defined as 5 times the signal-to-noise ra­ tained in this manner still remained higher than those obtained
tio, was 0.015 ppm. Similar results were obtained by Schoen with the GC/MS method. Because the incurred THPI residues
and Winterlin ( 1 1 ). found in the baby foods were so low, we did not attempt to
When the results for 13 fruits obtained through the ELISA determine the possible precursors of THPI, such as captan or
and GC/MS methods were compared, the correlation was captafol.
good. The correlation coefficient between the ELISA results (x In conclusion, none of the baby food samples contained a
axis) and the GC/MS results (y axis) was 0.88 and the slope was concentration of combined captan and THPI that exceeded the
0.80 (Table 3). All strawberry samples contained THPI and/or maximum residue limit in Canada of 5.0 ppm. A sensitive
captan. No false positives were observed with the ELISA GC/MS method was developed for the confirmative determi­
method. However, 3 samples were detected with the GC/MS nation of THPI levels in foods. The described ELISA method
method that were below the detection limit of the ELISA for determining THPI concentration was less accurate and pre­
method. The detection limits of the ELISA and the GC/MS cise in complex matrixes, such as baby foods, than GC/MS. On
methods were 0.150 and 0.015 ppm, respectively. If the mini­ the other hand, it did not give any false negative data. Our re­
mum detection limit for the ELISA method is considered the sults demonstrate that the ELISA method can serve effectively
predetermined cutoff point for both systems, then no false as a primary screening method in a complex food matrix. It has
negatives were found with the ELISA method. the advantage of being both time-saving and cost-effective.
The mean recovery data for apple sauce and apple juice Many samples can be screened for THPI in a short time with
baby foods at 3 levels of fortification were 91.8 and 96.4%, the ELISA method, and any negative results can be eliminated
respectively, for the ELISA method and 87.8 and 94.5%, re­ from further analysis by the GC/MS method.
spectively, for the GC/MS method (Table 4). These figures
were similar to the results obtained with the GC/MS method R eferen ces
for strawberry samples. Excellent correlation was obtained
when the actual quantification values were correlated, with r = (1) Canadian Food & Drugs Act and Regulation (1990) 503
0.99 and slope = 0.93. (2) G ilvyd is, D .M ., W alters, S.M ., Spivak, E.S., & Hedblad,
In the survey of combined captan and THPI in baby foods, R. K . (1986) J. Assoc. Off. Anal. Chem. 69, 803-806
90 products were sampled. With the GC/MS method, 8 sam­ (3) W interlin W .L., K ilgore, W .W ., M ourer, C.R., & Schoen,
ples were found to contain THPI at levels ranging from 0.019 S. R. (1984) J.Agric. Food Chem. 32, 664-672
to 0.041 ppm. With the ELISA method, 20 samples tested posi­ (4) Frank, R., Braun, H.E., & Stanek, J. (1983) Arch. Environ.
tive for THPI with levels ranging from 0.157 to 1.776 ppm Contam. Toxicol. 12, 265-269
(Table 5). Twelve samples showed false positive results with (5) El-Zem aity, M .S. (1988) Bull. Environ. Contam. Toxicol. 40,
the ELISA method, but no false negative result was observed. 75-79
All of the green bean samples were positive with both methods, (6) Ritcey, G ., Frank, R., M cEwen, F.L., & Braun, H .E. (1987)
but the values were significantly higher with the ELISA assay. Bull. Environ. Contam. Toxicol. 38, 840-846
This result suggested that a constituent or an additive of the (7) Kaufm an, B .M ., & C low er Jr, M . (1991) J. Assoc. Off. Anal.
green beans produced interference in the ELISA. Uniformly, Chem. 74, 239-247
the ELISA method gave higher estimates of THPI than the (8) Jung, F. Gee, S.J., Harrison, R.O ., Goodrow, M .H ., Kara,
GC/MS method in baby foods other than green beans. Presum­ A .E ., Braun, A .L ., L i, Q.X., & Hammock, B .D . (1989) Pes-
ably, the antibodies cross-reacted with some of the ingredients tic. Sci. 26, 303-317
in baby foods because this phenomenon was not observed with (9) Newsome, W .H., Yeung, J.M ., & C ollins, P.G. (1993) J.
the raw fruit. It should be pointed out that this matrix effect was AOAC Int. 76, 381-386
not evident in the spiked samples of one of the solid (apple (10) Newsome, W .H., & C ollins, P.G. (1990) Food & Agriculture
sauce) and one of the liquid (apple juice) baby foods (Table 4). Immunology 2, 75-84
When 200 pL instead of 100 pL of each of the extracts was (11) Schoen, S.R., & W in te rlin W .L. (1982) J. Assoc. Off. Anal.
analyzed, similar results for THPI were obtained, except that Chem. 65, 1382-1384
1230 K h u n a c h a k Et A l .: Jo u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

D R U G S, C O S M E T IC S , F O R E N S IC M A T ER IA L S

L iq u id C h ro m a to g ra p h ic D e te rm in a tio n o f M o x id e c tin R e s id u e s

in C a ttle T is s u e s a n d C o n firm a tio n in C a ttle F a t b y L iq u id

C h ro m a to g ra p h y /M a s s S p e c tro m e try

A lisa K hunachak, A drian R. D acunha, and Steven J. Stout 1

American Cyanamid Company, Agricultural Research Division, PO Box 400, Princeton, NJ 08543-0400

M oxidectin, a p o te n t new en d o - a n d e c to p a ra sitic nation of the structurally similar ivermectin in cattle and sheep
ag e n t, is d eterm in e d in cattle tis s u e s by liquid ch ro ­ tissues, our first objective was to ascertain the method’s appli­
m ato g rap h y (LC) with flu o re sc e n c e d etectio n . Mox­ cability to the determination of moxidectin residues in cattle
idectin re s id u e s in cattle fat are confirm ed with fat, muscle, liver, and kidney. Our second objective was to
th e rm o sp ra y L C /m ass sp e c tro m e try (MS). Mox­ evaluate the suitability of thermospray LC/mass spectrometry
idectin is ex tra cted from th e tis s u e with a c e to n i­ (MS) as a confirmatory technique for the direct analysis of cat­
trile; th e ex tra ct is partitioned with h exane, c o n c e n ­ tle fat extracts from the LC/fluorescence method. Direct
trated , a n d re a c te d with a c e tic anhydride, LC/MS of these extracts for confirmation would circumvent
1 -m ethylim idazole, a n d dim ethylform am ide to pro­ the lengthy and involved confirmatory method developed for
d u c e a flu o re sc e n t p ro d u ct. T he validated sen sitiv ­ ivermectin, which requires a tedious cleanup before determina­
ity of th e L C /fluorescence m ethod w a s 10 ppb, with tion by tandem MS (9).
a limit of d etec tio n typically betw een 1 an d 2 ppb.
A verage re co v erie s from cattle fat, m uscle, liver, METHOD
a n d kidney w ere 99, 95, 89, a n d 92%, respectively.
LC/MS confirm atory m ethod d eterm in ed th e underi- Special N otes
vatized p aren t c o m p o u n d follow ing th e a c e to n itrile -
h ex a n e partitioning ste p , with an a v e rag e recovery Thoroughly rinse all clean glassware with water followed
of 108% at th e 250 p p b level in cattle fat. by acetone, and let dry before use. The presence of moisture
will interfere with the derivatization reaction. All solvents
should be distilled in glass and suitable for pesticide analyses
M oxidectin (Figure 1) is a potent new antiparasitic (Burdick & Jackson Laboratories, Inc., or equivalent). Water
agent. Its structure is related to the milbemycins (1, should be purified by Milli-Q water purification system (Mil-
lipore Corp.), or equivalent. To avoid contamination, the reac­
2) and avermectins (3), which have a novel mode
tion of
vials should be cleaned after each use as follows: Rinse
action against a broad spectrum of nematode and arthropod with acetone and soak in detergent and water for several hours
parasites of animals (4). Common features of these compounds or overnight. Rinse well with distilled water and then deionized
are a fused cyclohexene-tetrahydrofuran ring system, a bicy- water. Rinse with acetone again and let vials dry completely.
clic 6 ,6 -membered spiroketal, and a cyclohexene ring fused to
the 16-membered macrocyclic lactone ring. R eagents and Apparatus for LC/Fluorescence
Studies of the disposition, excretion, and metabolism of ra­
diolabeled moxidectin in cattle, sheep, and rats were reported (a) Moxidectin standard solutions.—Prepare standard so­
by our laboratories (5-7). Only part-per-billion levels of drug- lutions containing 10, 1.0, and 0.1 pg moxidectin/mL metha­
related radioactivity were detected in various organs and tis­ nol. Standard solutions are stable for 1 month when stored in a
sues. All species showed very similar metabolic patterns, and refrigerator at 4°C. Moxidectin is available from American Cy­
the unaltered drug is the major residue at all time points studied. anamid Co., Agricultural Research Div., Princeton, NJ.
Considering the lipophilic nature of moxidectin, its greatest (b) LC standard solutions.—Pipet 1.0 mL of the 0.1 pg/mL
persistence in fat was entirely reasonable; therefore, fat was the standard solution and 0.2, 0.5, and 1.0 mL of the 1.0 pg/mL
target cattle tissue for regulatory purposes. standard solution into separate reaction vessels. Evaporate to
Because the liquid chromatographic (LC)/fluorescence dryness and derivatize as described in Determination by
method of Tway et al. (8 ) proved satisfactory for the determi- LC/Fluorescence. Dilute each with 10 mL methanol to give
derivatized standard solutions containing 0.01, 0.02,0.05, and
Received July 24, 1992. Accepted February 9, 1993.
0.1 pg/mL, respectively. The 0.05 pg/mL derivatized standard
Author to whom correspondence should be addressed.
is prepared with each set of samples analyzed and is the work-
 K h u n a c h a k E t A l .: Jo u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1231

R eagents and Apparatus for LC/MS

C H j-Q ^
N (a) LC/MS standard solution.—Pipet 5.0 mL of the
1.0 pg/mL standard solution of moxidectin into a 50 mL round-
bottom flask, evaporate to dryness, and reconstitute in 20 mL
LC/MS mobile phase. The LC/MS standard solution contains
0.25 pg moxidectin/mL.
(b) LC/MS mobile phase.— Weigh 7.7 g ammonium ace­
tate (analytical reagent, Mallinckrodt), transfer to a 1 L high
density polyethylene bottle, dissolve in 170 mL water, and di­
lute with 830 mL methanol.
(c) Liquid chromatograph/mass spectrometer.—An ABI
Kratos Spectroflow Model 400 pump delivered the LC/MS
mobile phase at a flow rate of 1.5 mL/min to a Rheodyne
Model 7125 injector equipped with a 100 pL loop. The LC col­
umn was a Whatman RAC II Partisil 5-C8 (10 cm X 4.6 mm)
maintained at ambient temperature. UV absorbance at 245 nm
was monitored by an ABI Kratos Spectroflow 783 programma­
ble absorbance detector and recorded with a Spectra-Physics
Model 4290 recording integrator. The mass spectrometer was
a Finnigan-MAT TSQ70 triple stage quadrupole equipped with
a Finnigan-MAT thermospray LC/MS accessory. Operational
parameters specific to the thermospray interface were as fol­
lows: aerosol temperature, 230°C; vaporizer temperature,
75°C; repeller voltage, 150 V. For LC/MS analyses, all ions
were allowed to pass through the first 2 quadrupoles; the third
quadrupole monitored ions indicative of moxidectin at m/z 640,
622, and 528, with a dwell time of 1.0 s/ion. Mass spectromet-
Figure 1. Chemical structures of (A) moxidectin and (B) ric operating parameters included the following: conversion
the fluorescent derivative, didehydromoxidectin. dynode voltage, -15 kV; electron multiplier voltage, -1250 V;
preamplifier gain 10' 9 A/V. Retention time of moxidectin was
ca 5 min.
ing standard for quantitation of samples. The other derivatized
standard solutions are used for the linearity check. Determination b y LC/Fluorescence
(c) Reaction solution.—Mix 9 mL dimethylformamide (re­
Freeze quantity of commodity sufficient to provide repre­
agent grade, J. T. Baker), 3 mL acetic anhydride (reagent grade,
sentative sampling, mix with powdered dry ice, and chop with
J. T. Baker), and 2 mL 1-methylimidazole (99%, Aldrich). This
prechilled food chopper. Mix chopped sample well, and let
solution must be prepared just before derivatization and used
stand in freezer until dry ice has completely dissipated.
immediately.
Accurately weigh 10 g cattle tissue into a 200 mL stainless
(d) LC mobile phase.—Mix 20 mL purified water and
steel chamber of an Omni-Mixer. Add 100 mL acetonitrile and
980 mL methanol. Filter through a Rainin Nylon- 6 6 (0.45 pm)
blend at low speed for 5 min. Filter through a glass microfiber
filter, or equivalent.
filter paper by using a 150 mL sintered glass funnel (medium
(e) Heating block.—Reacti-Therm heating module with porosity) under vacuum, and leave the tissue in the chamber.
Reacti-Block B-l and Reacti-Vap evaporator (Pierce). Add 100 mL of additional acetonitrile and extract for another
(f) Micro reaction vessels.—5.0 mL vessel, 20 mm cap di­ 5 min at low speed. Filter through the same funnel. Transfer the
ameter with Teflon-lined, solid plastic caps (Supelco). combined filtrates to a 500 mL separatory funnel. Rinse the
(g) Liquid chromatograph.— An ABI Kratos Spectroflow filtering flask with 100 mL hexane, add the rinse to the separa­
Model 400 pump delivered the LC mobile phase at a flow rate tory funnel, and shake for 1 min. Collect the lower acetonitrile
of 1.8 mL/min to a Rheodyne Model 7125 injector equipped phase in a 500 mL round-bottom flask, and evaporate to dry­
with a 100 pLloop. The column was a 25 cm x 4.6 mm id Zor- ness by using a rotary evaporator. Discard the hexane in the
bax ODS (5 pm) maintained at ambient temperature. Detection separatory funnel.
was provided by an ABI Kratos Model FS970 Spectrofluoro Dissolve the residue in 20 mL acetonitrile and then add
Monitor set for excitation at 364 nm and for detection of emis­ 80 mL deionized water. Swirl for 10 s and transfer to a 500 mL
sion above 470 nm by use of a 470 nm filter on the phototube. separatory funnel. Partition with three 100 mL portions of hex­
Chromatograms were recorded with a Spectra-Physics Model ane in the following manner: Rinse the flask with 100 mL hex­
4270 recording integrator. Retention time of derivatized mox­ ane and add to the separatory funnel. Shake for 1 min, releasing
idectin was ca 9 min. the pressure formed as necessary. Let the phases completely
1232 K h u n a c h a k E t A l .: Jo u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

separate, and draw the lower aqueous phase into a 150 mL in ng/mL of the LC standard solution (usually 50 ng/mL), and
beaker. Slowly decant the upper hexane phase into a new W = g sample used in the analysis (usually 10 g).
500 mL round-bottom flask, but leave the last few milliliters of
Determination b y LC/MS
the aqueous phase in the separatory funnel. Pour the aqueous
phase in the beaker back into the separatory funnel. Rinse the Follow Determination by LC/Fluorescence for sample ex­
beaker with 100 mL hexane, add to the funnel, and repeat the traction and the partitioning cleanup. Do not use the derivati­
partitioning. Evaporate the combined hexane to dryness. zation procedure and subsequent cleanup. After evaporating
Transfer the residue in the 500 mL round-bottom flask to a the combined hexane extracts from the partitioning step, recon­
100 mL round-bottom flask by using two 10 mL portions of stitute the dry residue in 10 mL LC/MS mobile phase. Inject
methanol. Evaporate to dryness or as dry as possible. Dissolve 100 |iL aliquots of the LC/MS standard solution (250 ng/mL)
the residue with 2 mL methanol, and transfer to a micro reac­ and the sample extracts. Measure the retention time of the ana­
tion vessel by using a disposable Pasteur pipet. Repeat with lyte peak in the m/z 640 mass chromatogram, and obtain the
another 2 mL methanol and transfer to the same vial. Evaporate background subtracted intensities of the m/z 640,622, and 528
the combined methanol in the vial to dryness or as dry as pos­ ions. Calculate the residue in parts per billion by using the
sible by using the Reacti-Vap evaporator with a stream of air. equation given in Determination by LC/Fluorescence.
Add 0.5 mL of the reaction solution to the vial and cap tightly.
Heat the vial, using the Reacti-Block and the heating module, R esu lts a n d D iscu ssio n
at 95°C for 1 h. Let cool for at least 30 min.
Add 1 mL hexane to the cooled solution in the vial, cap The basic approach of Tway et al. (8 ) for determining iver­
tightly, and shake for 1 min. Let sit for a few minutes, and de­ mectin in animal tissues was found to be directly applicable to
cant the solution into a 10 mL disposable syringe connected to
determining moxidectin residues in cattle tissues. The analyti­
a Florisil Sep-Pak. Let the solution move into the Florisil by
cal scheme used extraction, partition, derivatization, final
gravity for ca 1 min and then use the plunger and apply slight
cleanup, and determination by LC with fluorescence detection.
pressure to push the clean hexane solution into the Florisil Sep-
Several modifications to the procedure of Tway et al. proved
Pak. Collect the eluant in a 100 mL round-bottom flask. Pull
particularly beneficial for the determination of moxidectin in
the plunger out of the syringe. Connect a Millex-HV micro fil­
cattle tissues. First, acetonitrile was substituted for acetone-
ter to the other end of the Sep-Pak. Add 3 mL hexane to the vial,
water as the initial extractant and was found applicable for all
cap, and shake for 10 s. Add the hexane rinse to the syringe, use
tissues. For ivermectin, isooctane had to be added to the ace­
the plunger to apply pressure to push the hexane through the
tone-water to extract fat samples. Second, the precipitation of
Sep-Pak, and collect this rinse in the same 100 mL round-bot­
extraneous tissue materials in the initial ivermectin extracts
tom flask. Disconnect the syringe from the Sep-Pak and pull
was not required in the determination of moxidectin. Sample
the plunger out. Reconnect the syringe to the Sep-Pak, add
preparation of moxidectin extracts did not require centrifuga­
5 mL hexane to the syringe, and use the plunger to push the
tion. The final and most analytically useful modification was
hexane though the Sep-Pak into the same round-bottom flask.
the substitution of hexane for chloroform as the eluant in the
Evaporate the combined hexane to dryness by using a rotary
Sep-Pak cleanup step before LC/fluorescence determination.
evaporator. Add 5 mL methanol to the flask, swirl, and transfer
Whereas hexane elution left essentially all of the color of the
this solution by using a disposable pipet into a 10 mL volumet­
final derivatized extract on the cartridge, elution with chloro­
ric flask. Rinse the flask with 4 mL of additional methanol, and
form did not. The net result was a much cleaner chromatogram
transfer into the same volumetric flask. Add more methanol to
(Figure 2) without the intense, prolonged solvent tail present in
the volumetric flask to dilute to the mark and mix well. Inject
chromatograms of ivermectin extracts.
100 pL aliquots onto the LC system. Dilute samples containing
To validate the LC/fluorescence method, nonmedicated cat­
high drug residues to fit on the standard curve.
tle tissues were spiked with methanolic solutions of moxidectin
Quantitate samples and standards by peak height. Check the
and immediately carried through the assay. Recoveries were
linearity of the standard curve by using the LC standard solu­
ran in the fortification ranges of 1 0 to 1 0 0 0 ppb for fat samples
tions (10-100 ng/mL) prepared from moxidectin standard so­
and 10 to 500 ppb for the other tissues (Table 1). Recoveries
lutions carried through the derivatization and Sep-Pak purifi­
expressed as the average ±1 standard deviation (SD) were 99
cation procedures. Because the LC standard solutions are not
± 9% for fat, 95 ± 15% for muscle, 89 ± 18% for liver, and 92
stable for long-term storage, prepare a fresh 50 ng/mL standard
± 12% for kidney. Control tissues showed apparent moxidectin
for use as the working standard for each set of samples ana­
residues of < 1 - 2 ppb, which is the limit of detection of the
lyzed. Calculate the residue in ppb as follows:
method. The specificity of the method was checked against the
Residue (ppb) = R(samp) x V x C(std) x DF/[R(std) x W\ following animal health products, none of which were found to
interfere: ivermectin, fenbendazole, levamisole, monensin,
oxytetracycline, procaine penicillin G, rafoxanide, sul­
where //(samp) = response of the sample, R(std) = response of famethazine, and trenbolone.
the standard, V = volume in mL of the final solution for LC LC/fluorescence determination was applied to a residue de­
(usually 10 mL), DF - dilution of the sample if V is too con­ pletion study of cattle dosed at 0 . 2 mg/kg moxidectin by sub­
centrated (DF = 1 if no dilution made), C(std) = concentration cutaneous injection. For this study, liver, muscle, kidney, and
 K h u n a c h a k E t A l .: Jo u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1233

Table 2. Average moxidectin residue in cattle tissues

Residue, ppb
Days
postdose Muscle Liver Kidney Fat

Control <10 <10 <10 <10

14 <10 14 27 27 5
21 <10 15 29 24 3
28 <10 <10 22 225
35 <10 <10 19 153
42 <10 <10 <10 77

for fluorescence detection generated a satisfactory mass spec­

trum for confirmatory purposes (Figure 3), its response in ther­
T I M E ( mi n ) mospray LC/MS was 100-fold lower than that of underivatized
moxidectin. Such a dramatic decrease in response was not en­
Figure 2. Liquid chrom atogram s of (A) control cattle
tirely unreasonable, because elimination of 2 hydroxyl groups
m uscle and (B) cattle m uscle fortified at 10 ppb.
from the parent could dramatically affect its proton affinity in
thermospray LC/MS. In view of the decreased polarity of dide­
hydromoxidectin, an attempt was also made to determine it by
fat from 6 groups of animals were assayed. Three animals in
GC/MS. Didehydromoxidectin could be eluted from a GC col­
one group served to furnish nonmedicated control tissues. The
umn by using high column temperatures (injector, 325°C; col­
5 treated groups ( 6 animals each) were sacrificed at 14,21,28,
umn, 200-325°C at 30°C/min) in conjunction with an alumi­
35, and 42 days postdose, respectively. Table 2 summarizes the num-clad capillary GC column (4 m x 220 pm id, 0.1 pm HT-5
results of this study. The apparent moxidectin residues in all from SGE, Inc., Austin, TX). However, the broad chroma­
control cattle tissue samples were less than the 1 0 ppb vali­ tographic peak that was obtained, coupled with coeluting inter­
dated sensitivity of the method. Residues of moxidectin were ferences at the monitored ions from fatty acid coextractives,
< 1 0 ppb in the muscle of all treated cattle and rapidly dissi­ destined this approach to failure. Consequently, thermospray
pated to the validated sensitivity of the method in liver and LC/MS of the parent moxidectin was the most feasible ap­
kidney. In agreement with the cattle metabolism study (5), proach for a confirmatory method.
moxidectin exhibited its greatest persistence in cattle fat, which
made fat the target tissue for regulatory purposes.
Because confirmatory methods of analysis for regulatory
agencies require structural specificity, MS has gained wide­
spread acceptance as the preferred analytical technique (10). In
the case of moxidectin, attempts to perform MS confirmation
directly on the same extracts used in the LC/fluorescence de­
termination were unsuccessful. The first attempt at a confirma­
tory procedure was the obvious substitution of a mass spec­
trometer for the fluorescence detector. Although
didehydromoxidectin produced in the derivatization procedure

Table 1. Recovery3 (%) of moxidectin from cattle

tissu es determined by LC/fluorescence
Fortification level, added, ppb
Control,
Com modity ppb 10 20 100 500 1000

Fat <1.1 106.7 107.9 92.2 101.4 88.8

__b
Muscle <0.9 109.0 100.6 86 .9 8 3 .3
Liver <1.0 82.3 100.1 90.5 84 .2 —
Kidney <1.6 98 .4 97.9 82.6 96 .8 —

a Th e average of duplicate sam ples processed through the

procedure. Figure 3. Mass spectra from therm ospray LC/MS of (A)
b Recovery experim ent not performed at this fortification level. moxidectin and (B) didehydromoxidectin.
1234 Khunachak Et A l .: Journal Of AOAC International V ol . 76, N o. 6 ,1 9 9 3

Table 3. Thermospray LC/MS data from cattle fat

528 3 622
Sam ple 640 640 ppb

Standard (n = 11) 63.2 (4 .5 )b 58 .7 (5.4) —

Fat fortified at 25 0 62.9 (4.8) 55.1 (7.5) 2 7 0 (58)

ppb ( n = 5)
Incurred fat ( n = 5) 55.0 (2.3) 58.1 (6.1) 251 (21)

a Ion ratios monitored, %.

b Standard deviation in parentheses.

though quantitative data are not as important to a confirmatory

method as the structural information provided, the average re­
covery from the fortified tissues was 108% (SD, 23%) with a
range of 70.9-126%. Analyses of the 5 biologically incurred
samples gave an average residue level of 251 ppb (SD, 21 ppb),
which is in excellent agreement with the 250 ppb determined
by LC/fluorescence for this specific cattle fat sample. These
quantitative results are within guidelines for regulatory meth­
ods (10) and further support the ability of the LC/MS method
to reliably confirm moxidectin residues in cattle fat at the
250 ppb level.
An LC/fluorescence method was developed for the determi­
nation of moxidectin residues in cattle fat, muscle, liver, and
T IM E (m in :s e c ) kidney. With a few modifications, the method of Tway et al. (8 )
developed for ivermectin in cattle and sheep tissues was found
Figure 4. Chromatograms from thermospray LC/MS of applicable for determining moxidectin in cattle tissues. Ex­
(A) control cattle fat and (B) cattle fat fortified at 250 ppb
tracts analyzed by LC/fluorescence were not directly amenable
to MS confirmation. Consequently, confirmation required the
determination of the underivatized parent by thermospray
Validation of the LC/MS confirmatory method used control LC/MS. The LC/MS confirmatory method required processing
cattle fat, cattle fat fortified at 250 ppb, and incurred cattle fat another sample through the extraction and partition steps but
containing approximately 250 ppb from the previously dis­ avoided the subsequent derivatization and cleanup. As an alter­
cussed residue depletion study (day 28 in Table 2). Figure 4 native, the extract could be split after the partitioning step, and
shows the results obtained from a control and a fortified sam­ half could be processed for LC/fluorescence determination and
ple. No detectable peaks (apparent residue <25 ppb) were the other half could be retained for potential LC/MS confirma­
found at any of the monitored ions at the retention time of mox- tion. The cleanup provided by coupling LC with MS elimi­
idectin in any of the control extracts. Table 3 summarizes the nated the extensive sample cleanup and tandem MS required
qualitative confirmatory data obtained on the fortified and in­ for the confirmation of ivermectin (9).
curred samples over the course of the validation. For each se­
lected sample type (standard, fortified, and incurred), the abso­ R efe ren c es
lute SDs of the relative intensities of the confirmatory ions
(1) M ishim a, H „ Ide, J., Muram atsu, S., & Ono, M . (1983)7. A n-
correspond, predominantly, to better than 10% relative SDs.
tibiot. 36, 980-989
The average relative intensities of the confirmatory ions for the
(2) Takiguchi, Y., M ishim a, H ., Okuda, M ., Terao, M ., A o k i, A .,
fortified and incurred extracts are generally within 1 SD of the
& Fukuda, R. (1980) J. A ntibiot. 33, 1120-1127
mean standard intensities. These values actually understate the
(3) Albers-Schonberg, G., A rison, B .H ., Chabala, J.C., Douglas,
qualitative confirmatory capabilities of the method, because
A.W ., Eskola, R, Fischer, M .H ., Lusi, A ., M rozik, H ., Sm ith,
the results are calculated over the entire time frame of the study.
J.L., & Tolman, R .L. (1981) J. Am. Chem. Soc. 103, 42 16 -
In practice, results are usually confirmed by bracketing the 4221
sample injection with a pair of standard injections to reduce the (4) Putter, I., M acC onnell, J.G., Preiser, F.A., H aidri, A .A ., Ris-
impact of drift and instability on the ion ratios. Reviewing the tich, S.S., & Dybas, R .A . (1981) E xperientia 37, 963-964
data in this manner showed that the ion ratios from each forti­ (5) Z ulalian, J., Stout, S.J., daCunha, A .R ., Garces, T., & M ille r,
fied and incurred extract were within 1 0 % of the average ion P. (1993) J. A g ric. F o o d Chem ., in press
ratios of the bracketing standards. This performance is within (6) Stout, S J., daCunha, A .R ., W u, S.-S., Z ulalian, J., & A fza l, J.
established guidelines for confirmatory methods (10, 11). Al- (1993) J. A gric. F o o d C hem ., in press
 R upp E t A l .: Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3 1235

(7) W u, S.-S., Stout, S.J., daCunha, A .R ., B iro c, S., Singh, S., (9) Tway, PC., D ow ning, G.V., Slayback, J.R.B., Rahn, G.S., &
W ashington, J., & Rajan, S. (1993) “ A bsorption, D istrib u ­ Isensee, R .K . (1984) Biomed. Mass. Spectrom. 11, 172-176
tion , E xcretion, and M etabolism o f M oxidectin in the Rat,” (10) Sphon, J.A. (1978) J. Assoc. Off. Anal. Chem. 61, 1247-1252
Presented at the 5th N orth Am erican ISSX M eeting, Tucson, (11) DeRuig, W .G., Stephany, R.W ., & D ijkstra, G. (1989) J . Ay-
A Z , paper N o. 190 soc. O ff Anal. Chem. 72, 487^190
(8) Tway, P.C., W ood, J.S., & D ow ning, G.V. (1981) J. Agric.
F o o d Chem. 29, 1059-1063

DRUGS, COSMETICS, FORENSIC MATERIALS

S im u lta n e o u s D e te rm in a tio n o f N itro fu r a z o n e a n d F u ra z o lid o n e

in S h rim p (P e n a e u s v a n n a m e i) M u s c le T is s u e b y L iq u id

C h ro m a to g ra p h y w ith U V D e te c tio n

H eidi S. R upp, R obert K. M unns, and Austin R. L ong

U.S. Food and Drag Administration, Animal Drags Research Center, PO Box 25087, Denver Federal Center, Denver, CO
80225-0087

A liquid ch ro m a to g ra p h ic (LC) m eth o d w a s devel­ N itrofurazone (NFZ) and furazolidone (FZD) belong to
o p e d for th e sim u lta n e o u s determ in atio n of nitro­ a class of highly effective synthetic antimicrobial com­
fu razo n e (NFZ) a n d fu razo lid o n e (FZD) in sh rim p pounds called nitrofurans. A furan ring bearing a nitro
m u sc le tis su e . T h e d ru g s a re e x tra c te d from th e tis­ group at position 5 is a characteristic structural component of
s u e with acetonitrile, a n d th e lipids a n d lipophilic these compounds (Figure 1).
p ig m en ts a re rem ov ed from th e e x tra c t with hex­ Nitrofurans have a history of common use in the poultry and
an e. The rem aining aceto n itrile ex tra ct is ev a p o ­ swine industries as both therapeutics and feed additives. How­
rated by ro tary ev a p o ratio n , a n d th e re su lta n t ever, nitrofuran residues in food are now a major concern, be­
re s id u e s a re d isso lv e d with LC -grade w ater, a p ­ cause they have been implicated as carcinogens and mutagens
plied to a p re co n d itio n e d Cis so lid -p h a se ex tra c­ (1) and can cause allergic reactions in sensitive individuals. Re­
tion colu m n , a n d elu ted with acetonitrile. T he cently, the U.S. Food and Drag Administration (FDA) with­
aceto n itrile e lu an t is th e n d ried u n d e r nitrogen, an d drew approval of nitrofurans and moved to ban them for oral
th e re su lta n t d ru g re s id u e s a re d isso lv e d with m o­ or parenteral use in all food-producing animals (2). FDA also
bile p h a s e a n d filtered. T he d ru g s a re d eterm in ed placed the highest priority on preventing nonpermitted extra­
by LC by u sin g a Cis re v e rse d -p h a se (octyldecyl- label use of still-approved topical nitrofuran preparations (2 ).
silyl Hypersil) colu m n, a m obile p h a s e of acetoni- Although successful aquaculture requires the use of thera­
trile-1 % a q u e o u s a c e tic acid (25 + 75, v/v), a n d a peutic treatments, few therapeutic drugs are registered for use
p h o to d io d e array UV d e te c to r a t 375 nm . NFZ an d in the United States, and only slightly more are permitted in
FZD w ere d eterm in e d in sh rim p tis s u e at e a c h of 5 European and Asian countries. Regulations governing the use
sp ik in g lev els (64, 3 2 ,1 6 , 8, a n d 4 ng drug/g tissu e). of nitrofurans for this purpose vary from country to country.
A b so lu te re c o v e rie s ra n g e d from 70.6 to 78.4%, FDA does not allow the registration of any nitrofuran for
a n d relative s ta n d a rd d ev iatio n s ra n g ed from 4.0 to aquaculture and has set a zero tolerance limit for these com­
13.6%. T he limit of d e te c tio n of p u re sta n d a rd of pounds (3). Concern arises from the potential extra-label use of
e a c h d ru g w a s app roxim ately th e eq u iv ale n t of 1 any therapeutic compounds (especially nitrofurans) in
ng drug /g tis s u e , a n d th e limit of determ in atio n in a aquaculture species such as shrimp.
sa m p le w a s 4 ng drug/g tissu e . The detection of parent nitrofurans in animal tissue after
slaughter is difficult at best, because nitrofuran drags metabo­
lize quickly, forming multiple metabolites. On the basis of re­
Received Novem ber 10, 1992. Accepted February 24, 1993. cent research, this is apparently true of catfish (Ictalurus punc-
1236 Rupp E t A l .: Journal O f AO AC Internation a i . V ol . 76, N o. 6 ,1 9 9 3

of mixed working solution will contain the same appropriate

quantity of each of the 2 nitrofurans. Use a 100 pL aliquot of
mixed working solution to spike a 5 g sample of tissue. Because
nitrofurans are light-sensitive, work under subdued lighting.
Store all solutions in a refrigerator and make fresh each week.
(b) Solvents.—Acetonitrile (UV grade), water [liquid chro­
matographic (LC) grade], hexane, and dimethylformamide
(Burdick & Jackson, Muskegon, MI 49442).
(c) Ethyl alcohol.—200 proof dehydrated alcohol, USP
(Quantum Chemical Corporation, USI Division, Tuscola, EL
61953).
(d) Acetic acid.—Glacial acetic acid, 99.9% pure (J.T.
Baker Inc., Phillipsburg, NJ 08865). Prepare 1% aqueous ace­
Figure 1. Chemical structures of nitrofurazone and tic acid by adding 10 mL glacial acetic acid to ca 900 mL LC
furazolidone. water in a 1.0 L volumetric flask and adding LC water to the
mark; filter through a glass Fluoropore-type FH membrane fil­
ter, 0.5 pm pore size (Cat. No. FHLP 047 00, Millipore, Bed­
ford, MAO 1730).
tatas) (personal communication, Steve Plakas, Division of
(e) Saturated hexane.—Prepare by adding acetonitrile to
Seafood Research, FDA) and is well known for other animal
hexane ( 1 + 1 0 , v/v) in a separatory funnel, stoppering, and
species (4, 5). The toxicological significance of these metabo­
shaking well. Drain off the lower acetonitrile layer and save for
lic products is unknown, however, because the metabolism,
later use. Use upper hexane layer for lipid extraction.
disposition, and elimination of nitrofurans in aquaculture spe­
cies including shrimp have not been thoroughly studied to date. Apparatus
In the present study, a method was developed for the simul­
taneous extraction, separation, and detection of nitrofurazone (a) Homogenizer.—Tissuemizer SDT1810, Model TR-10
and furazolidone in shrimp (Penaeus vannamei) muscle. The power control, SD T182 EN shaft/blades (Tekmar Co., Cincin­
method is based on our previously developed method for de­ nati, OH 45222-1856).
termination of 3 nitrofurans in catfish muscle (unpublished (b) Centrifuge.—Model DPR-6000, rotor Model 823A (In­
data), which was developed in response to the need for a ternational Equipment Co., Needham Heights, MA 02194).
method for nitrofurans in aquaculture that was efficient analyti­ (c) Rotary vacuum evaporator.—Buchi Rotavapor No.
cally and in terms of labor, time, and materials. Previous nitro- R110 with ice trap (Brinkmann Instruments, Inc., Westbury,
furan methods found in the literature were intended for ma­ NY 11590). Set water bath temperature at 45X1
trixes other than shrimp (6-11). The present method for shrimp (d) Sonicator.—Branson 2200 (Branson Ultrasonics
is a refinement of the catfish method, because it provides for an Corp., Danbury, CT 06813-1961).
increased cleanup of matrix components, including orange pig­ (e) Liquid chromatograph.—HP 1090 Series II/M LC sys­
ments, that are found in appreciable amounts in shrimp but not tem with Pascal ChemStation and autosampler (Hewlett-
in catfish. In this study, incurred shrimp tissue was not analyzed Packard Co., Avondale, PA 19311-0900): 50 pL injection at
because of the difficulty of obtaining shrimp out of season for 40X1 mobile phase, acetonitrile-1% aqueous acetic acid (25 +
incursion. 75, v/v), run at 1.0 mL/min isocratic; photodiode array detec­
tor, set at 375 nm (10 nm bandwidth), referenced against 450
Experim ental nm (80 nm bandwidth); threshold, 0.1 mAU; integrator, on at
3.0 min, with initial threshold at - 6 (this threshold for levels <5
R eagents and Chemicals ppb).
(f) Chromatographic column.—ODS Hypersil C ) 8 (oc-
(a) Nitrofurazone and furazolidone.—U.S. Pharmacopeia tyldecylsilyl) column, 5.0 pm pore size, 200 x 4.6 mm (Cat.
(USP) Reference Standards (U.S. Pharmacopeia, Rockville. No. 799160D-574, Hewlett-Packard Co.).
MD 20852). Stock solutions.—Weigh 5.00 mg each of NFZ (g) Guard column and holder.—Guard column cartridges,
and FZD, and place in separate 50 mL volumetric flasks. Add ODS Hypersil (C18), 5 pm, 20 X 4.0 mm (Cat. No. 79916KT-
5 mL dimethylformamide to each flask, swirl until solids are 120); pre-column cartridge holder (Cat. No. 79900CH-010)
dissolved, and dilute to volume with acetonitrile (concentra­ (Hewlett-Packard Co.).
tion, 100 pg/mL for each solution). Mixed working solu­ (h) Solid-phase extraction (SPE) columns.—Analytichem
tions.—From the stock solutions, make a dilution series of 5 Bond Elut, C 18 6 cc (Cat. No. 1210-2052, Varían Sample
mixed working solutions used for spiking at concentrations of Preparation Products, Harbor City, CA 90710).
3200, 1600, 800,400, and 200 ng drug/mL solution. To make (i) SPE elution vacuum.—Vac-Elut SPS 24 (Analytichem
each mixed working solution, transfer an appropriate aliquot of International, subsidiary of Varían).
each drug’s stock solution into the same 50 mL volumetric (j) Flasks.—Pear-shaped, 100 mL, 24/40 standard taper
flask, and dilute to volume with acetonitrile. Each dilution level (Cat. No. 608700-0224, Kontes Co., Vineland, NJ 08360).
 R u p p E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1237

(k) Tubes.— Falcon graduated 50 mL polypropylene cen­ upper hexane layerwith an aspirator,taking carenottocarry
trifugetubes(Cat.No. 2098,Becton Dickinson,Lincoln Park, overany acetonitrile.Add 10 mL absoluteethanol totheace­
NJ 07035). Eppendorf 1.5 mL polypropylene standard micro tonitrile,and evaporate on arotaryevaporatorunder medium
testtubes(Cat.No. 22 36 380-8,Brinkmann InstrumentsCo.). vacuum at45°toavolume of2-5 mL (orwhen theliquidjust
(l) S y rin g e b a rr e ls. — Luer-Lok 3 cc disposable poly­ startstolookmilky).Add anadditional2 mL absoluteethanol,
propylenesyringes(Cat.No. 9585, Becton Dickinson & Co.). andcontinuerotaryevaporationuntilthereisathickliquidresi­
(m) S y rin g e f ilte r s . — 13 mm Puradisc disposable due (ca2 mL). At thispoint,add an additional2 mL absolute
polypropylene filters(Cat.No. 6788-1304, Whatman Labora­ ethanol toevaporate todryness. Remove promptly when dry.
toryDivision,Clifton,NJ 07014). (Thisisagood shortbreakingpoint.)
(n) A s p ir a to r . — 500mL vacuumflaskpluggedwithTeflon Add 2 mL LC-grade watertotheflask,stopper,swirlwell
tube20x 1/8in.od, 1/16in.id,fittedthrougharubberstopper aroundthewallsoftheflasktodissolveanysolids,ventbriefly,
with side arm connected towater aspirator vacuum. [Tubing andsonicatefor5min.PrimeaC18SPE column (6cccapacity)
functionsastheworkingendofwateraspirator.Used withcare by rinsingwith5mL methanolfollowedby5 mL water.Trans­
and placed atthetop edge ofthe hexane againsttheinsideof fer the analyte-containing water to the primed SPE column
theglass,thetube can suctionoffalmost allthehexane layer. withaPasteurpipet.Rinsethecolumn with4 mL water.Elute
Care must be taken so asnottoremove theloweracetonitrile theanalytesslowly(<3 mL/min) with 5 mL acetonitrileintoa
layer(ifthishappens,theinterfaceoflayersatthesuctionarea smallglasstesttubeuntiltheSPE columnisdry.Dry theeluate
appearsdistorted).The hexaneiscollectedinthevacuum flask under nitrogen at <45°C. Remove promptly when dry. Add
forlaterrecycling.] 1.00 mL mobilephasetothetesttube,mixbyvortexingbriefly,
(o) L C v ia ls a n d se p tu m c a p s. — Clearglassvials,32 X 11 and then transferthe liquidby using a disposable glass pipet
mm id(partNo.5180-4197);aluminumcapswithseptum(part intoa 3 cc syringebarrelfittedwith a0.45 pm polypropylene
No. 5061-3370) (Hewlett-PackardCo.). filter(13 mm diameter).With theplunger,forcetheliquidout
(p) P a s te u r p ip e ts . — Disposableglass,5.75in.(14.6mm). ofthesyringebarreland throughthefilterintoanLC vial.Cap
(q ) P ip e tto r s . — Eppendorf micro adjustable KMOOpL thevialwith a septum cap.Injecta 50 pL portionintotheLC
and 100-1000pL pipettors(Brinkmann Instruments Co.); ad­ system foranalysis.
justable 5 mL pipettor(Cat. No. 851350, Wheaton Manufac­
turers,Millville,NJ 08332). Determ ination
(r) P ip e t tip s. — 200-1000 pL Veri-tips (Cat. No. P5054-
24);1-250pL Veri-tips(Cat.No. P5054-23)(UlsterScientific, Measure fiedailychromatographicpeakareasobtainedfor
Div. of Baxter Diagnostics Inc., McGraw Park, EL 60085- dailystandards,equivalenttoatleast0.5a, and 2 x ng drug/g x ,

6787); 1-5 mL macro pipettips (Cat. No. 53503-826, VWR tissue, where x istheconcentration under study. These daily
Scientific,nationwide). standards are prepared by taking an appropriate aliquot of a
mixed working solutionanddilutingto 1mL withappropriate
Tissue Preparation microliteramounts ofacetonitrileand aqueous 1% aceticacid
togive the same composition as the mobile phase. Calculate
Remove headsandchitinfromfrozenshrimp. Keep tailtis­ regressiondataforeach analytefrom themeasuredpeakareas
suecoldand semiffozenby keeping inabeakerovericeuntil and correspondingconcentrations.
ready toprocess.Let shelledshrimp tailsp a r tia lly th a w (e.g., Injecta 50 pL aliquotofsample from theLC vialintothe
cold and firm to the touch but can be cut by knife). Using a liquidchromatograph, and measure thechromatographicpeak
stainlesssteelchef’sknifeandanonporous,high-densitypoly­ areas.Calculate theconcentrationsofNFZ and FZD from re­
ethylene cutting board (Cat. No. 7339-00-900-9861, Federal gressionequationsofdailystandardcurves.
Supply Service, General ServiceAdministration, Washington
DC 20406, or equivalent), coarsely mince the tissue with a Recovery Study
chopping motion. (Thiscan be easilyaccomplished ifdone in
a semi-frozen state as specified. This hand-chopped method, Five 5 g portions of control shrimp were weighed into
compared tousing a blender, does not mash thetissue,which 50 mL polypropylene centrifugetubes.Contents ofeach tube
can promote degradation, and italso allows easy recovery of were fortifiedat64.00ng drug/gtissuewith 100 pL ofa3200
allbitsoftissueofftheboard.) ng/mL mixed working solution. Each of a setof5 replicates
was alsofortifiedatlevelsof32, 16,8,and4 ng drug/g tissue
Extraction and Cleanup
with 100 pL aliquotsfrom mixed working solutions of 1600,
Accurately weigh 5 ± 0.05 g finelychopped shrimp intoa 800, 400, ar.d200 ng/mL, respectively. The method was also
50 mL polypropylene centrifuge tube. Add 100 pL mixed usedtostudya setof5 replicatesofshrimptissueblanks.
working solution to tissue. Add 20 mL acetonitrile, and ho­ Each day that a setof replicates was chromatographed, a
mogenizewithaTissuemizeratmedium speedforca45s.Cen­ dailystandardcurvewas constructedforeachdrugat0.5x, lx,
trifuge at3000 rpm for5 min, and decant supernatant into a and 2x the concentration under study. The equations of the
100 mL pear-shaped flask.Add 30 mL hexane (saturatedwith daily NFZ and FZD standard curves were used to determine
acetonitrile), stopper, and shake well forca 30 s.Let hexane drug levels (ng drug/g tissue) on the basis of peak areas ob­
and acetonitrile layers fully separate (2-3 min). Remove the servedinchromatograms oftheseportions.
1238 R u p p E t A l .: J o u r n a l O f A O AC In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

Table 1. Average recovery of nitrofurazone and tionofnitrofurazoneis 10ppb; recoveriesat4 ppb and above
furazolidone from fortified shrimp muscle tissue by LC were over70% and fellwellwithinCVM guidelines.
determination NFZ and FZD have very similarUV spectra,withmaxima
Nitrofurazone Furazolidone atabout 375 nm. At thiswavelength, there are practically no
analyticalinterferencesfromthetissuematrix(Figure2a),and
Added, Found, Absolute Found, Absolute
ng/g ng/ga rec., %b RSD, %c ng/ga rec., %b RSD, %c therefore, a selective chromatogram can be obtained (Fig­
ure2b).
4.00d 2.9 73.0 11.5 2.8 70.6 12.6 The limitofdetection(signal:noise,3:1)ofapure standard
8.00 6.1 76.4 11.2 5.8 72.2 5.5 ofNFZ andFZD isapproximatelytheequivalentof1ngdrug/g
16.00 12.3 77.2 10.4 11.7 73.1 13.6 tissue. Intissue,4 ng drug/gtissueisthelimitofquantitation
32.00 24.5 76.6 4.2 23.9 74.7 4.9 (signaknoise, 10:1).
64.00 48.8 76.2 5.2 50.1 78.4 4.0
Extraction and Cleanup
a Average peak area of 5 replicates vs daily standard curve of Vfe x,
1x, and 2 x level added.
b Absolute rec. = (Found + Added) x 100. One extractionoftheshrimpmuscletissuewithacetonitrile
c RSD, relative standard deviation am ong peak areas of 5 replicates. gave good recoveriesoftheanalytes (Table 1) and prevented
d Four replicates. carryover ofunwanted particulatematter. On thebasis ofour
previous research with catfishtissue, a second extraction did
notsignificantlyimproverecovery.Additionally,a secondho­
Atthebeginningofeachsetofspikedshrimpreplicates,the
mogenizationwas made difficultasaresultofhardeningofthe
3 daily standards were injectedonto theLC column todeter­
tissuepelletfromexposuretoacetonitrile.Thehexanewashfor
mine peak areas foruse inthecalculation ofregression data.
lipidremoval followed by SPE cleanup and filtrationresulted
FDA Center forVeterinary Medicine (CVM) guidelines (12)
in a solution thatchromatographed well, with no significant
setaminimum analyterecoveryat60% forsamplescontaining
inherentinterferingpeaks (Figures2aand2b).The initialpeak
<100 ng drug/g tissue, and a relative standard deviation of
atabout0.9 min inthefortifiedtissueextractisresidualfrom
<20%. the previous run; the retention time isconsistent because of
injectionontotheLC by automaticsequencing.Thispeak was
Results and Discussion not evident when single injections were run manually during
development, and therefore, could be minimized by a se­
M ethod Performance quenced rantimelongerthan6 min asperformedon therepli­
cates.
Average absoluterecoveries ofdrug residuesfrom 5 repli­ When both nitrogen and rotary vacuum evaporation are
cates offortified shrimp tissue were calculated for NFZ and used,itisimportanttouseawaterbathtemperature<45°C and
FZD ateach of 5 fortificationlevels. These average absolute toremovesamplespromptlywhen drytoavoidlossofanalytes.
recoveries, one foreach level and drug, ranged from 70.6 to Absoluteethanolwas usedduringtherotaryvacuum evapora­
78.4%, and the corresponding relative standard deviations tionsteptofacilitatetheremovalofwater.The ethanolmustbe
rangedfrom4.0to 13.6% (Table 1).The targetlevelfordetec­ added before rotaryvacuum evaporation toadjustthesurface

Figure 2. Liquid chromatograms of (a) shrimp control and (b) shrimp fortified with nitrofurazone (NFZ) and
furazolidone (FZD): concentration, 16 ng drug/g tissue.
 R u p p E t A l .: J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1239

tensionandpreventsudden,strongbubblingupintotheinstru­ Acknowledgments
mentthatcanresultinalossofanalyte-containingsolution.No
cleartrendwas obviousintheexactamount ofethanolneeded We thankthosechemists who contributedtime and insight
tocompletelydryasample,butstepwiseadditionsof10,2,and intothedevelopmentofthismethod:JoseE.RoybalandDavid
2 mL gave good results.The number ofethanol additions ap­ C. Holland, Animal Drugs Research Center, FDA; Steve M.
pearstobe more criticaltotheevaporationthanthevolume of Plakas,DivisionofSeafoodResearch(DauphinIsland),FDA;
ethanoladded.Thepurposeoftakingtheextracttodrynesswas andJeffreyA.Hurlbut,ScienceAdvisor,DenverDistrict,FDA.
toleavebehindno organic solventcomponent thatmightpre­
vent some of the analytes from fully adhering when sub­ References
sequently applied in 100% water to a C18SPE column. The
aliquotofwateraddedtotheresidueintheflaskcontainingthe (1) Yndestad,M. (1990)“PublicHealthAspectsofResiduesin
analytes must be swirled well and sonicated to solubilize as AnimalProducts:FundamentalConsiderations,”paperfrom
much analyte as possible. Note thatthe water istoo polar to DepartmentofFoodHygiene,NorwegianCollegeofVeteri­
dissolve allthe less-polarresidues (mainly pigments) on the naryMedicine,Oslo,Norway
flaskwalls. (2) Anonymous (1992)F ood Chem. N ew s 35(4),12
Throughout theprocedure, portions oforange shrimp pig­ (3) Schnick,R.A.(1990)“ChemicalsforWorldwideAquacul­
mentsareremoved witheachsucceedingstep.Roughlyhalfof ture,”presentedattheWorkshoponFishDiseaseandFish
thepigmentseems tobe lipophilicand isremoved alongwith HealthManagement,Pusan,Korea,October8-15
thehexane wash. Some pigmentisnotsolubilizedinthewater (4) Vroomen,L.H.M.,Beighmans,M.CJ.,Groten,J.P.,Koe-
with the analytes; itadheres to the sides of the pear-shaped man,J.H.,& VanBladeren,P.J.(1988)Toxicol. Appl.
Pharmacol. 95,53-60
flask.More pigmentstaysontheSPE column.Duringthefinal
nitrogen evaporation, the remaining pigment tends to clump (5) Vroomen,L.H.M.,Berghmans,M.C.J.,Hekman, P.,Hoogen-
boom,L.A.P.,& Kuiper,H.A.(1988)Xenobiotica 17(12),
togetherand sticktotheglasstesttubewallsortoitself.Even 1427-1435
afteritismixed by vortexingwith mobile phase, thepigment
(6) Laurensen,J.J.,& Nouws,J.F.M.(1989)J. Chromatogr. 472,
remains undissolved and is easily removed by filtering (a 321-326
polypropylene filter,0.45 pm X 13 mm diameter,works best)
(7) Suortti,T.,& Heinonen,K.(1987)Chrom atographia 24,
togiveaclearsolution. 344-346
(8) Burkepile,R.G.,& Torda,S.B.(1980)FDA Lab. Information
Conclusion Bull., No. 2449,Rockville,MD
(9) Nagata,T.,& Saeki,M. (1991)J. Liq. Chromatogr. 14,2551-
Low nanogram quantitiesofNFZ and FZD in shrimp can 2561
be determinedby thismethod.The methodrequiresonly 1day (10) Nose,N.,Hoshino,Y.,Kikuchi,Y.,Horie,M.,Saitoh,K.,
fortheanalysisof8samples,from thetissuepreparationtothe Kawachi,T.,& Nakazawa,H.(1987)J. Assoc. Off. Anal.
chromatography. The results can be expected to fall within Chem. 70,714-717
CVM guidelinesforfishtissuecontaining<100 ng drug/gtis­ (11) Samuelsen,O.(1990)J. Chromatogr. 528,495-500
sue.The use ofthismethod on othershrimp and aquaculture (12) Anonymous (1989)A n alytical M eth od Trial P rocedures ,
speciesisbelieved possible. Incurred shrimp samples willbe CenterforVeterinaryMedicine,FoodandDmg Administra­
analyzeddependingon availability. tion,Rockville,MD
1240 J u n e E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

FOOD BIOLOGICAL CONTAMINANTS

E ffe c tiv e n e s s o f th e Bacteriological Analytical Manual C u l t u r e

M e th o d f o r th e R e c o v e r y o f Shigella sonnei f r o m S e l e c t e d F o o d s

G eraldine A.June,Patricia S.Sherrod,R. M iguel Amaguana,W allace H . A ndrews,and Thomas S.H ammack
U.S. Food and Drag Administration,Division ofMicrobiological Studies,200 C St,SW, Washington, DC 20204

The relative retention of the indigenous morpho­ S h ig e lla (2,3),shigellosisoutbreakshave been causedby raw
logical, biochemical, and serological charac­ oysters,fish,andgroundmeats(2,5,6).Thesefoodshadeither
teristics by S h ig e lla s o n n e i was tested under vari­ beentakenfromfecallycontaminatedwatersorwerecontami­
ous storage conditions (room temperature, natedby infectedfoodhandlers.However, thenumber ofcases
refrigeration, freezing at -20°C and at -70°C, and of foodbome shigellosis may actuallybe higher than thatre­
lyophilization). The use of a selective (desoxycho- ported, because traditional culture methods for isolating
late citrate) agar rather than a nonselective (brain S h ig e lla fromfoods arerelativelyinsensitive(2,3,7).The cul­
heart infusion) agar gave a lower conversion rate turingofS h ig e lla fromfoodsisdifficult,becausetheorganisms
of smooth to rough colonies, and the percentage of
are oftenfound in very low numbers (7), and theirgrowth is
rough colonies derived from cultures stored for
inhibitedby acidsproducedby competitivemicroflora(8-11).
prolonged periods increased under all conditions.
With respect to biochemical characteristics, there
Even withanenrichmentmedium containinglittleornocarbo­
were no major differences in the reactions of
hydrate,S h ig e lla cannotbe recoveredfromfoodsunlessitsin­
smooth vs rough variants. For serological charac­ itiallevelofcontamination isatleast 106organismsper gram
teristics, smooth variants agglutinated more read­ (7).
ily in homologous antisera than did rough variants. The difficultyinisolatingS. so n n e i isfurtherexacerbatedby
S. s o n n e i populations maintained at -70°C with thefactthatS. so n n e i producesbothsmoothandroughcolonies
glycerol remained reasonably stable and were (3,12).CellsofS. so n n e i thatproducesmoothcoloniesaredes­
used in recovery studies. Up to six foods (potato ignatedForm Icoloniesandarevirulent.Form IIcoloniesgive
salad, chicken salad, cooked salad shrimp, lettuce, arough appearance and areaviralent.Loss ofa 120-megadal-
raw ground beef, and raw oysters) were inoculated tonplasmidwasfoundtocorrelatewithalossofvirulence.This
with unstressed, chill-stressed, or freeze-stressed plasmid codes for the synthesis of specific O antigen side
S. s o n n e i cells. Test portions (25 g) were inocu­ chains of the lipopolysaccharide thatgives the colonies their
lated with serial 10-fold dilutions of culture and sub­ smooth appearance(13).Itwas reportedthatthecellslosethis
sequently analyzed by the culture method de­
plasmid at a rate of 1-50%, depending on the strain (13).
scribed in the U.S. Food and Drug Administration’s
RoughcoloniesofS. so n n e i may appearatypicalandtheorgan­
B a c te r io lo g ic a l A n a ly t ic a l M a n u a l. It was found that
ismsmightnotberecognizedonplatingagars.Ourpreliminary
the method was relatively ineffective for the recov­
ery of S. s o n n e i from raw ground beef and raw oys­
studies showed that this dimorphism was especially pro­
ters. nounced when thecultureswere stored forprolonged periods
atroom temperature.
The currentculturemethod describedintheU.S. Food and
I n recentyears,thenumber ofreportedcasesofshigellosis DragAdministration’s(FDA)B a c te r io lo g ic a l A n a ly tic a l M a n ­
has increased. Classically known as a waterborne patho­ u a l (BAM) (14) selectsforS h ig e lla by enrichinginalow car­
gen, S h ig e lla has been found with greater frequency in bohydratemedium, andincubatinginananaerobicatmosphere
foodbome outbreaks. From 1961 to 1975, there were 72 out­ atanelevatedtemperatureinthepresenceoftheantibioticno­
breaks (10 648 cases) of foodbome shigellosis and 38 out­ vobiocin.Before attemptingtoimprove thecurrentmethodol­
breaks (5893 cases) ofwaterborne shigellosis (1).The foods ogy,we decided todetermineitseffectivenessinrecoveringS.
most oftenimplicatedascauses ofshigellosisarepotato salad so n n e i from variousfoods.The objectivesofthisstudy,there­
and salads containing chicken or fish(1, 2). Raw vegetables fore,weretodetermineoptimalstorageconditionsforretaining
have been implicated toa lesserextent(1,3,4).Even though the organism’s indigenous morphological, biochemical, and
humans and higher primates are the only natural host of
serologicalcharacteristicsandtodocumenttheeffectivenessof
recovery ofS. so n n e i from variousfoods by thecurrentBAM
Received February 13, 1993. Accepted May 4, 1993.
culturemethod.
 J u n e E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V ol . 76, N o. 6 ,1 9 9 3 1241

Experimental CellsfromtheBHI brothculturewere washed twicewithBut­

terfield’sphosphate bufferand aliquotswere frozen (withand
R e a g e n ts a n d M e d ia without20% glycerol)and storedat-20 and-70°C. Aliquots
ofthewashed cellswere alsoadded toequalvolumes of20%
Allmediaandreagentswerepreparedaccordingtomethods nonfatdrymilk inserum vials.The contentsofthevialswere
recommendedinBAM (14)andinE d w a r d s a n d E w in g ’s Id e n ­ shell-frozeninan ethanol-dryicebath. The vialswere placed
tific a tio n o f E n te r o b a c te r ia c e a e (15).When conventionalbio­ in a freeze-dryer (Virtis Co., Gardiner, NY) and the contents
chemicaltestswerenotperformed,S h ig e lla culturesw erecon­ werelyophilized24± 2h.The vialswerestopperedundervac­
firmedbiochemicallywithMICRO-ID kits(OrganonTeknika, uum andstoredintherefrigerator.Thus, subculturesofS. so n ­
Durham, NC). Cultures were confirmed serologically by ag­ n e i were storedunder7 conditions: (7)room temperature; (2)
glutinationinBacto S h ig e lla poly Group D antiserum (Difco 4°C;(3 ) -20°C withoutglycerol;(4 )-20°C with20% glycerol;
Laboratories,Detroit,MI). (5) -70°C withoutglycerol;(6)-70°C with20% glycerol;and
C u ltures
(7) lyophilizedand storedintherefrigerator.
At approximate 1-week intervals, BHI broth subcultures
S. so n n e i strains 9290, 11060, and 25931 were obtained weremade fromculturesstoredundereachofthe7 conditions,
from the American Type Culture Collection (ATCC), and the subcultures were incubated 18-24 h at 35°C. From
Rockville,MD. these overnight BHI broth cultures, serial 10-fold dilutions
were made, and 0.1 mL aliquots were spread onto BHI agar
O p tim a l G ro w th p H o f S . s o n n e i
plates.Counts were made and isolatesofeach morphological
Novobiocin was added toflasksofS h ig e lla brothtoafinal typewere transferredtoBHI agarslants.Serological and bio­
concentrationof0.5 pg/mL. The pH ofthecontentsofdupli­ chemicaltestswereperformedasdescribedforthecontrolcul­
catesetsofflaskswas adjustedto6.2,6.6,7.0,7.4,7.8,and 8.2 ture.
by adding IN NaOH or IN HC1. A 3 mm loopfulofan 18-24 E ffe c t o f D e s o x y c h o la te C itra te A g a r on C o n versio n
h brainheartinfusion(BHI) brothcultureofS. so n n e i was in­ o f S m o o th -to -R o u g h C o lo n ie s o f S. s o n n e i
oculatedintoeachoftheflasks.One setofflaskswasincubated
aerobically20 h at44°C inacirculatingwaterbath.The other A lyophilized strain of.
S’.s o n n e i was rehydrated and then
wasplacedinanaerobicjarswithfreshcatalysts;GasPakenve­ grown anaerobically20h inS h ig e lla broth(14)at44°C. From
lopeswereinsertedandactivatedbyaddingwater.Theseflasks thiscultureserial 10-folddilutionswere made, and 0.1 mL of
were alsoincubated 20 h at44°C ina circulating waterbath. each was spread-plated onto duplicate desoxycholate citrate
IncubatedS h ig e lla brothcultureswerewashed twicewithBut­ (DC) agarplatesandincubated 18-24h at35°C. Two colonies
terfield’s phosphate buffer and serial 10-fold dilutions were ofeachmoiphological typewere transferredtoDC agarslants
made. Viablecellsinovernightcultureswerecountedtodeter­ and incubated 18-24 h at 35°C. The somatic group of each
mine thenumber ofcellspermilliliterofinoculum. isolatewas confirmed serologically,and thebiochemicaltests
listedabove were performed on eachisolate.From thecontrol
D e term in a tio n o f the O p tim a l S to ra g e C ondition fo r
culture,subcultureswere made and storedunderthe7 storage
S. s o n n e i
conditionsfistedabove.At ca 1week intervals,S h ig e lla broth
A lyophilized strainofS. so n n e i was rehydrated and then subcultureswere made from cultures storedundereach ofthe
grown 18-24h inBHI brothat35°C.From thisculture,serial 7 storageconditionsandthesubculturesweregrownanaerobi­
10-folddilutionswere made, and 0.1mL ofeach dilutionwas cally20hat44°C.From theseovernightcultures,serial10-fold
spread-platedontoduplicateBHI agarplates.The plateswere dilutionsweremade, and0.1mL aliquotswere spreadontoDC
incubated 18-24h at35°C.Two coloniesofeachmorphologi­ agar plates.Counts were made and isolates ofeach morpho­
caltype(smoothandrough)weretransferredtoBHI agarslants logicaltypeweretransferredtoDC agarslants.Serologicaland
and incubated 18-24 h at 35°C. The somatic group of each biochemicaltestswere performed asdescribedforthecontrol
isolate was determined by group specific antisera. The bio­ culture.
chemical profile of each isolate was then determined by the D e te rm in a tio n o f th e E ffe c tiv e n e s s o f th e B A M
following tests: glucose (acid and gas production); hydrogen C u ltu re M e th o d fo r the R e c o v e ry o f S . s o n n e i from
sulfide production; motility; urease production; growth in S e le c te d F o o d s
KCN medium andmalonatebroth;lysineandornithinedecar­
boxylase production; indole production; methyl red reaction; The effectivenessoftheBAM (14)culturemethod forthe
citrateutilization;and acidproductionfromthefollowingcar­ recovery of.S’.so n n e i was determined fortherecovery ofun­
bohydratesinabromcresolpurplebase:lactose,sucrose,inosi­ stressed, chill-stressed and freeze-stressed cellsfrom up to 6
tol, salicin, glycerol, raffinose, rhamnose, xylose, dulcitol, foods (potato salad, chicken salad, cooked salad shrimp, let­
mannitol,adonitol,maltose,arabinose,andtrehalose. tuce,raw ground beef, and raw oysters).The foods were pur­
The reactionswiththisprimarycultureservedasthecontrol chased at retail outlets. Portions (25 g) of the potato salad,
to which the experimental results were compared. From the chicken salad, cooked salad shrimp, lettuce, and raw ground
primary BHI broth culture above, subcultures were made to beefwere placedinsterile500 mL Erlenmeyerflasks.Forthe
BHI agar slants and stored atroom temperature and at4°C. raw oysters, a slurry was made with a 1:1 ratioof oysters to
1242 J u n e E t A l .: J o u r n a l O f AO AC I n t e r n a t io n a l V o l 76, N o. 6 ,1 9 9 3

Figure 1. Effect of plating agars BHI and DCA on the smooth-to-rough conversion of Shigella sonnetstrain 9290
stored under the following conditions: A, at room temperature and at 4°C; B, at -20°C; C, at -70°C; D, lyophilized and
stored in the refrigerator.

Butterfield’sphosphate buffer, and 50 mL of this slurry was S h ig e lla brothwithnovobiocin(0.5pg/mL) was addedtoeach
added to the flask. From a stock culture, a subculture was foodexceptraw oysters.A 200mL portionofaspecialstrength
placedinBHI brothand incubated 18-24 h at35°C.The cells of S h ig e lla broth was added to 50 mL aliquotsofinoculated
werewashedtwicewithButterfield’sphosphatebuffer.Forthe rawoysters.The specialstrengthofS h ig e lla brothwas adjusted
chill-stressed cellinjury,the cellswere stored 7 days at4°C. to account forthe buffer in the raw oyster slurry. The flasks
The method of Ray et al. (16) was used to lfeeze-stress the were held atroom temperature for 10 min and theircontents
cells. Specifically, cells from overnight trypticase soy (with were swirled at5 min intervals. After 10 min, thepH of the
yeast extract) broth cultures were washed twice with sterile suspension was adjustedto7.0± 0.2 ifnecessary. The flasks
waterandresuspended. Aliquots (10 mL) were frozenrapidly wereincubatedanaerobically20 h incirculatingwaterbathsat
in an acetone-dry ice bath, and then thawed 75 min at4°C. 44°C. Each flask was mixed well, and the contents were
From theunstressed,chill-stressed,orfreeze-stressedcellsus­ streakedonto MacConkey agar plates.The plateswere incu­
pensions,serial 10-folddilutionswere made. From thesedilu­ batedaerobically20hat35°C. Suspiciouscoloniesweretrans­
tions, 1mL aliquotswere inoculatedintoduplicatesetsofthe ferredtotriplesugariron(TSI)agar.TSI slantsthatgave typi­
25 g foodportionsinErlenmeyerflasksandmixed well.Aero­ calS h ig e lla reactions were confirmed biochemically with the
bicplatecountswere performed on thedilutionstodetermine MICRO-ID system and serologicallywith group-specific so­
the viable counts of the inocula. After inoculation, 225 mL maticantiserum.
 J u n e E t A l .: J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1243

Results and Discussion Table 1. Serological reactions of smooth9 and rough

variants of Shigellasonnei (strain 9290) stored under
Becausethemethod forS h ig e lla inthe6theditionofBAM various conditions and subsequently plated on BHI agar
(17) didnotrequiredetermination oradjustmentofthepH of Rough colonies6
Maintenance Length of
theS h ig e lla broth culture, the optimal pH forthe organism’s condition storage, days Trial 1 Trial 2
growthwasdeterminedinS h ig e lla broth.No significantdiffer­
ences were observed intheviablecountsofthe2 strainsofS. Control NAtf 2+ 2+
so n n e i among pH valuesof6.2to7.8.Moreover,theincubation Room temperature 0 1+ 1+
atmospheredidnotaffecttherecoveryofS. so n n e i atthesepH 4 1+ 4+
values. At pH 8.2, however, the number of S. so n n e i strain 11 1+ 1+
11060 cellsgrown aerobicallydecreased5 logs. 17 1+ 1+
Because ofthedimorphism ofS. so n n e i, itwas questioned 25 1+ 1+
whetherthebiochemicalcharacteristicsofthe2 typesofcolo­ 32 2+ 2+
nies would differ with prolonged storage. Subcultures of S. Refrigeration 4 1+ 2+
so n n e i were stored under 7 conditions to determine which 11 1+ 3+
would retainmost oftheculture’sindigenous morphological, 17 1+ 2+
biochemical,andserologicalcharacteristics.Results(Figure 1) 25 1+ 3+
showedanincreaseinthepercentageofroughcoloniesderived 32 2+ 3+
from culturesthatwere grown inBHI broth,platedonto BHI -20 °C w/o glycerol 6 1+ 1+
agar,andstoredforprolongedperiods. 13 1+ 4+
19 1+ 2+
Wheeler and Mickle (12) reported thatthetype ofplating
27 1+ 2+
medium determined whetherthere would be 2 forms ofcolo­
34 1+ 2+
nies,and thatPhase II(i.e.,rough) colonieswere more inhib­ -20°C w/ glycerol 6 1+ 1+
itedbyDC agarthanbytryptose,S a lm o n e lla -S h ig e lla , desoxy- 13 1+ 1+
cholate, eosin methylene blue, and MacConkey agars. We 19 1+ 1+
decided todetermine ifthesmooth-to-roughvariationand the 27 1+ 1+
retention of biochemical and serological characteristics de­ 34 1+ 3+
pendedonthetypeofmedium usedand/orthelengthofculture -70°C w/o glycerol 6 1+ 4+
storage.Cultures grown inS h ig e lla broth andplatedontoDC 13 1+ 1+
agarshowed littleincreaseinthepercentageofrough colonies 19 1+ 1+
(Figure 1).They produced lessthan 9% rough coloniesat30 27 1+ 1+
daysofstorageunderthe7 conditions. 34 1+ 2+
In group-specific antiserum, smooth variants agglutinated -70°C w/ glycerol 6 1+ 1+
more readilythandidroughvariantswhen platedon BHI agar 13 1+ 4+
(Table 1)or DC agar (Table 2).The smooth colonies consis­ 19 1+ 2+
tently gave 4+ (highest) agglutination, whereas most rough 27 1+ 2+
coloniesgave 1+or2+agglutinationreactions.Theseserologi­ 34 1+ 1+
Lyophilizatlon 5 1+ 1+
calreactionswerenotaffectedby thelengthofstorage.
12 1+ 1+
There were no major differencesinthe biochemical reac­
18 1+ 1+
tionsoftheculturesstoredunderthe7 conditionson eitherof 26 2+ 2+
the 2 media. Of the 25 tests performed, discrepant reactions 37 1+ 2+
occurredonlywithlactoseutilizationandgrowth inKCN me­
dium (Table3),andmostofthesediscrepantreactionsoccurred a Reactions ot smooth colonies were 4+ in duplicate determinations,
with prolonged (>18 days) storage ofthecultures.There was including controls, under all conditions.
nocorrelationbetweenthediscrepantbiochemicalreactionand
b Amount of agglutination graded from 1+ to 4+ (highest); duplicate
determinations.
themorphologicaltypeofthecolony (smoothorrough)orthe c The control culture was not subjected to storage; it was rehydrated
typeofmedium (selectiveornonselective). and grown f'om the ATCC stock culture.
S. so n n e i maintained at-70°C withglycerolyieldedarea­
d Not applicable.
sonably stablepopulationofcellsthatcouldbe usedinrecov­
ery studies.The effectiveness oftheBAM method forthere­
covery ofS. so n n e i from selectedfoods (potatosalad,chicken and raw oysters (8.5 x 102); for strain 25931: potato salad
salad,cooked saladshrimp,lettuce,raw ground beef, and raw (5.3x 10“1),chicken salad (6.6 x 10-1),cooked saladshrimp
oysters)was thendetermined. (1.6x 100),lettuce(4.2x 10_1),raw ground beef(6.9X 104),
The lowest number of unstressedS. so n n e i cells 'Table4) andraw oysters(5.4x 102).
recoveredper 25 g offood were forstrain9290: potato salad Because theBAM culturemethod was abletorecover 1S.
(1.0X 10°),chicken salad (8.5 X 10_1),cooked salad shrimp so n n e i cellper25 g from 4 ofthe6 foods, we testeditseffec­
(8.8 X 10_1),lettuce(7.6X 1CT1),raw ground beef(8.4X 10), tiveness with injured cells. Chill-stress and freeze-stress cell
1244 J u n e Et A l .: J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

Table 2. Serological reactions of smooth3 and rough Table 3. Discrepant biochemical reactions of smooth
variants of Shigella sonnei (strain 9290) stored under and rough variants of Shigella sonnei (strain 9290)
various conditions and subsequently plated on stored under various conditions
desoxychoiate citrate agar Discrepant
Length biochemical
Rough colonies'*
Maintenance Length of of reactions3
condition storage, days Trial 1 Trial 2 Maintenance storage, Morphc logical
condition days appearance KCNb Lactose0

Control Nonselectlve medium0'

NAd 0 0
Room temperature 0 0 0
Room temperature 25 Smooth +
4 0 0
11 0 0 Refrigeration 4 Rough +
19 2+ 3+ 25 Rough +
25 0 0
32 2+ 4+ -20°C w/o glycerol 19 Smooth +
Refrigeration 4 0 0
11 0 0
-20°C w/ glycerol 19 Smooth (1)e +
19 0 0
(2) +
25 0 0
27 Rough +
32 1+ 3+
- 20C
C w/o glycerol 6 0 0 -70°C w/o glycerol 19 Smooth (1)e +
13 0 0 (2) +
21 0 0 Rough +
27 4+ 0 27 Rough (1) +
34 4+ 4+ (2) +
-20°C w/ glycerol 6 2+ 2+ 34 Smooth +
13 0 0
21 0 0 -70°C w/ glycerol 27 Smooth +
27 0 0 Rough +
34 1+ 0 34 Smooth +
-70°C w/o glycerol 6 0 0
Lyophilization 18 Rough +
13 2+ 2+
26 Rough +
21 0 0
27 0 0
Selective medium'
34 4+ 4+
-70°C w/ glycerol 6 1+ 0
Room temperature 19 Rough +
13 0 0
32 Rough +
21 0 0
27 0 0 -20°C w/ glycerol 13 Smooth +
34 0 0 34 Rough +
Lyophilization 5 0 0
12 2+ 2+ -70°C w/o glycerol 6 Smooth +
20 0 0 34 Smooth +
26 2+ 0 Rough +
33 4+ 0
-70°C w/ glycerol 6 Rough +
3 Reactions of smooth colonies were 4+ in duplicate determinations,
including controls, under all conditions. 3 Different from those of the control.
6 Amount of agglutination graded from 1+ to 4+ (highest); duplicate b KCN = growth in potassium cyanide medium.
determinations; zero (0) indicates no rough colonies. c Lactose = lactose utilization.
c The control culture was not subjected to storage; it was rehydrated d Cultures were grown in BHI broth and spread-plated onto BHI
and grown from the ATCC stock culture. agar.
d Not applicable. e Particular Isolate number giving the reaction. If no number is
designated, only one isolate gave the discrepant reaction.
' Cultures were grown in Shigella broth and spread-plated on
desoxychoiate citrate agar.
 J u n e E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1245

Table 4. Effectiveness of BAM culture method for recovery of unstressed Shigella sonnei cells from selected foods

Strain number Inoculum viable count3 Recovery41 Strain number Inoculum viable count3 Recovery6

POTATO SALAD LETTUCE

9290 1.0x105 + + 9290 7.6x10° + +
1.0x104 + + 7.6x102 + +
1.0x103 + + 7.6x10 + +
1.0x102 + + 7.6x10° + +
1.0x10 + + 7.6x10” 1 + -
1.0x10° + + 7.6x10"2 - -
1.0x10“1 - - Negative control - -
Negative control - -
25931 4.2x10° + +
25931 5.3x103 + + 4.2x102 + +
5.3x102 + + 4.2x10 + +
5.3x10 + + 4.2x10° + +
5.3x10° + + 4.2x10-1 + +
5.3x10“1 + + 4.2x10”2 - -
5.3x10“2 - - Negative control - -
Negative control - -
RAW GROUND BEEF
CHICKEN SALAD
9290 8.4x104 + +
9290 8.5x104 + + 8.4x10° + +
8.5x103 + + 8.4x102 + +
8.5x102 + + 8.4x10 + +
8.5x10 + + 8.4x10° - -
8.5x10° + + Negative control - -
8.5x10 1 _ +
8.5x10-2 _ _ 25931 6.9x10® + +
Negative control _ _ 6.9x10s + +
6.9x104 + -
25931 6.6x103 + + 6.9x10° - -
6.6x102 + + 6.9x102 - -
6.6x10 + + 6.9x10 - -
6.6x10° + + 6.9x10° - -
6.6x10“1 + + 6.9x10“1 -
6.6x10“2 - Negative control - -
Negative control - -
RAW OYSTERS
COOKED SALAD SHRIMP 9290 8.5x10° + +
9290 8.8x104 + + 8.5x102 - +
8.8x103 + + 8.5x10 - -
8.8x102 + + 8.5x10° - -
8.8x10 + + 8.5x10‘ 1 - -
8.8x10° + + Negative control - -
8.8x10 1 _ +
25931 5.4x104 + +
8.8x10“2 _ _

5 4x10° + _
Negative control - - O
5.4x10 — +
25931 1.6x104 + + 5.4x10 - -
1.6x103 + + 5.4x10° - -
1.6x102 + + 5.4x10"1 - -
1.6x10 + + Negative control - -
1.6x10° + +
1.6x10“1 - -
Negative control - -

a Approximate number of cells inoculated into the enrichment broth.

b Duplicate determinations.
1246 June Et Al.: Journal Of AOAC International V ol . 76, No. 6,1993

Table 5. Effectiveness of BAM culture method for recovery of chill-stressed3 S h i g e l l a sonnet cells from selected foods

Strain number Inoculum viable countb Recoveryc Strain number Inoculum viable countb Recoveryc

POTATO SALAD LETTUCE

9290 8.8x10s + + 9290 6.8x10s + +
8.8x104 + + 6.8x102 + +
8.8x103 + + 6.8x10 + +
8.8x102 + + 6.8x10° + +
8.8x10 + + 6.8x10_1 -
8.8x10° + + Negative control -
8.8x10 '1 — -

25931 4.8x104 + +
Negative control _ -
q
4.8x10s + +
25931 5.3x10s + + 4.8x102 + +
5.3x102 + + 4.8x10 + +
5.3x10 + + 4.8x10° + +
5.3x10° + + 4.8x10"’ -
5.3x10"’ + - Negative control -
5.3x10"2 - -

Negative control - - RAW GROUND BEEF

9290 8.3x104 + +
CHICKEN SALAD 8.3x10s + -

9290 5.9x10s + + 8.3x102 -

5.9x104 + + 8.3x10 -
5.9x103 + + 8.3x10° -
5.9x102 + + 8.3x10"’
5.9x10 + + Negative control
5.9x10° + +
25931 5.9x106 + +
5.9x10"’ _ _

5.9x10s + +
Negative control — —

5.9x104 + +
25931 5.8x104 + + 5.9x10s + -
5.8x103 + + 5.9x102 - -
5.8x102 + + 5.9x10 - -
5.8x10 + + 5.9x10° - -
5.8x10° - + Negative control - -
5.8x10"’ - -

5.8x10"2 - -
RAW OYSTERS
Negative control - - 9290 8.3x104 + +
8.3x10s + +
COOKED SALAD SHRIMP 8.3x102 +
+
9290 7.2x104 + + 8.3x10 + -

7.2x10s + + 8.3x10° - -

7.2x102 + + 8.3x10"’ - -

7.2x10 + + Negative control - -

7.2x10"' + +
7.2x1 O'2 _ _ 25931 3.3x106 + +
Negative control — _
3.3x10s + +
3.3x104 + +
25931 6.0x103 + + 3.3x10s + +
6.0x102 + + 3.3x102 - -

6.0x10 + + 3.3x10 - -

6.0x10° + + 3.3x10° - -

6.0x10"’ - - 3.3x10"’ - -

Negative control - - Negative control - -

a Cells were stressed by storage In the refrigerator for 7 days.

b Approximate number of cells inoculated Into the enrichment broth.
c Duplicate determinations.
 Juke E t A l .: Journal O? AOAC International V ol . 76, No. 6,1993 1247

Table 6. Effectiveness of BAM culture method for injurywere chosen,because S h ig e lla contamination occursin
recovery of freeze-stressed3 S h i g e l l a s o n n e i cells from foodssuchasdelisalads,lettuce,andchoppedmeats,whichare
selected foods normally stored in the refrigerator or the freezer. The lowest
Strain number Inoculum viable count* Recovery17 number ofchill-stressedcells(Table5)recoveredper25 g of
food were for strain 9290: potato salad (8.8 x 10°), chicken
COOKED SALAD SHRIMP salad (5.9 X 10°), cooked salad shrimp (7.2 X 1CT1)’lettuce
(6.8xioo}igj.ouncj(3CCf(g3 x 103),and raw oysters (8.3X 10);
9290 1.4x10® + +
forstrain25931: potato salad (5.3X KT»-chickensalad(5.8xioo^
1.4x104 + +
1.4x103 + + cookedsaladshrimp(6.0X 10°),lettuce(4.8X 10°),rawground
1.4x102 + + beef(5.9X 103),andraw oysters(3.3x 103).
1.4x10 + + Raw oysters,cooked saladshrimp, and ground beef,which
1.4x10° + + arelikelytobestoredfrozen,wereusedtoevaluatetherecov­
1.4x10-1 - - eryoffreeze-stressedS. so n n e i cellsfromfoods (Table6).The
Negative control - - lowest number offreeze-stressed cellsrecovered per25 g of
foodwereforstrain9290:cookedsaladshrimp(1.4x10°),raw
25931 5.2x103 + + ground beef(1.1X 104),andraw oysters(7.6X 102);forstrain
5.2x102 + +
25931: cooked salad shrimp (5.2 x 10°), raw ground beef
5.2x10 + +
(4.4X 103),and raw oysters(2.1 X 103).
5.2x10° + +
EventhoughthecurrentBAM (14)methodrecoveredfewer
5.2x10“1 - -
cellsthanpreviouslyreported(7),itwas relativelyineffective
Negative control - -
forthe recovery ofS. so n n e i from raw ground beefand raw
RAW GROUND BEEF oysters;itseffectivenessforS. so n n e i decreased asthedegree
9290 1.1x10° +
ofinjurysustainedby thecellinoculumincreased.Becausethe
+
1.1x105 + +
infectivedose ofS h ig e lla canbe aslow as 10cellsperperson
1.1x104 + - (2,3),culturemethodsmustbe abletodeterminethislowlevel
1.1x103 - - ofcells.
1.1x102 - - Otherdisadvantagesofthemethod aretheneed touseele­
1.1x10 - - vated temperatures and to use large anaerobic jars to obtain
1.1x10° - - anaerobicatmospheres.Theseconditionsmay limitthenumber
Negative control - - oftestportionsthatcanbeanalyzed,especiallyinsmallerlabo­
ratories.Futureplansaretodevelopmore effectiveandpracti­
25931 4.4x104 + + calmethods fortherecoveryofS h ig ella .
4.4x10s + +
4.4x10s - -
References
4.4x10 - -
4.4x10° - -
(1) Black, R.E., C raun, G.F., & B lake, P.A. (1978) A™. J.
4.4x10“1 - -
108,4 7 -5 2
E pidem iol.
Negative control - -
(2) Sm ith, J.L., & B uchanan, R .L. (1992) in Com pendium o f
M ethods f o r the M ic ro b io lo g ic a l E xam ination o f Foods, 3rd
RAW OYSTERS
Ed., C. Vanderzant & D.F. S plittstoesser (Eds), A m erican
9290 7.6x104 + + Fhiblic H ealth A ssociation, W ashington, D C, pp. 423-431
7.6x10s + + (3) Sm ith, J.L. (1987) J. F o o d Prot. 50, 788-801
7.6x10s + +
(4) D avis, H., Taylor, J.P., Perdue, J.N ., Stelm a, Jr., G .N ., H um ­
7.6x10 - -
phreys, J.M . Jr, Row ntree, R. Ill, & G reene, K .D . (1988 )A m .
7.6x10° - -
J. Epidem iol. 128, 1312-1321
7.6x10“1 - -
(5) Saddik, M.F., E l-Sherbeeny, M .R ., M ousa, B .M ., El-A kkad,
Negative control - - A., & Bryan, F.L. (1985) J. F o o d Prot. 48, 403 4 0 6

25931 2.1x104 + + (6) Reeve, G., M artin, D.L., Pappas, J., T hom pson, R.E., &
2.1x10s + - G reene, K.D. (1989) N. Engl. J. Med. 321, 2 2 4 -2 2 7
2.1x10s - - (7) M ehlm an, I.J., R om ero, A., & W entz, B.A . (1985) J. Assoc.
2.1x10 - - Off. A nal. Chem. 68, 552 -5 5 5
2.1x10° - - (8) Fishbein, M ., M ehlm an, I.J., & W entz, B. (1972) J. Assoc.
2.1x10“1 - - Off. A n a l. Chem. 55, 1323-1327
Negative control - - (9) H entges, D.J. (1967) J. B a cte rio l. 9 3 ,1 3 6 9 -1 3 7 3
(10) H entges, D.J. (1967) J. B a cte rio l. 93, 2 0 2 9 -2 0 3 0
a Cells were stressed by the method of Ray et al. (16). (11) H entges, D.J. (1969) J. B a cte rio l. 97, 5 1 3 -5 1 7
6 Approximate number of cells inoculated into the enrichment brcth.
c Duplicate determinations. (12) W heeler, K .M ., & M ickle, F.L. (1945) J. Im m unol. 51, 2 5 7 -
267
1248 T uinstra Et A l . : Journal Of AOAC International V ol 76, No. 6,1993

(13) Sansonetti, P.J., K opecko, D .J., & Form al, S.B. (1981) Infect. (15) E w ing, W .H. (1986) in Edw ards and E w ing ’s Id e n tifica tio n
Immun. 34, 7 5 -8 3 ofE nterobacteriaceae, 4th E d., E lsevier Science Publishing
(14) U .S . F ood and D m g A dm inistration, B a c te rio lo g ic a l A n a ly ti­ Co., N ew York, NY, pp. 5 0 9 -5 2 9
c a l M a n u a l (1992) 7th Ed., A O A C IN T E R N A T IO N A L, (16) Ray, B., Janssen, D.W., & B usta, E F. (1972) A ppi. M ic ro b io l.
A rlington, VA 2, 803 -8 0 9
(17) U .S. F ood and D rug A dm inistration, B a c te rio lo g ic a l A n a ly ti­
ca l M a n u a l (1984) 6th Ed., A O A C , A rlington, VA

FOOD BIOLOGICAL CONTAMINANTS

L i q u i d C h r o m a t o g r a p h ic D e t e r m in a t io n o f A f la t o x in M l in M ilk

P o w d e r U s i n g I m m u n o a f f i n i t y C o lu m n s f o r C l e a n u p :

I n t e r la b o r a t o r y S tu d y

Louis GJVLTh. Tuinstra, Aril H. Roos, and John M.P. van Trijp
StateInstituteforQualityControlofAgriculturalProducts (RIKILT-DLO), Postbus 230, 6700 AE Wageningen, The
Netherlands

Collaborators: P.A.Burdaspal;B. Declercq; M.F. Dutton;K.F.Donahue; J.M.Fremy; A. Pittet;G. Rasmussen; M.A.

Rutschmann;M. Sharman; R.D. Stubblefield;M. Toyoda; H.P.van Egmond; J.van den Bedem; R.vanRenterghem;Yau-Huei
Wei

A liquid chromatographic method for determining I n March 1989, the Working Group on Mycotoxins ofthe
low aflatoxin Mi concentrations in milk was evalu­ International Dairy Federation planned an interlaboratory
ated in an International Dairy Federation (IDF) inter­ studytotestamethod applicableforlow aflatoxinMi con­
laboratory study. The study involved 16 centrations inmilk (<50ng/kg milk).The method isbased on
participants from 11 countries. The method, cho­ immunoaffinity chromatography cleanupwith liquidchroma­
sen after a comparison of several methods by a
tographic (LC) determination. A literaturereviewofmethods
fordeterminingaflatoxininmilkancmilkproductsrevealed5
preparatory group, uses an immunoaffinity column
collaborativelystudiedmethods usingthin-layerchromatogra­
for cleanup. As the sample passes through the col­
phy (TLC) orLC, andmorethan20othermethodsusingTLC,
umn, antibodies selectively bind with aflatoxin Mi
LC, radioimmunoassays, and enzyme-linked immunosorbent
(antigen) present and form an antibody-antigen
assaysasdetectiontechniques(1).None ofthesemethods,with
complex. All other components of the sample ma­
a claimed detection limitbelow 50 ng/L milk,have been col­
trix are washed off the column with water. Then,
laboratively tested. Most methods used C18 or silica for
aflatoxin Mi is eluted from the column with acetoni­ cleanup. Only 1 method used affinitycolumns (2).Recently,
trile, which is collected. Final determination is car­ other methods have been published thatuse affinitycolumns
ried out by reversed-phase liquid chromatography (3),aswellasseveralnew techniquesforhighsamplethrough­
with fluorescence detection. Over the tested range put.Antibodiescoatedonpolystyrenebeadswereusedforiso­
(80-600 ng aflatoxin Mi/kg milk powder), an RSDr lating aflatoxinM tfrom milk (4).The methods were further
ranging from 11 to 23% was obtained by analyzing evaluatedforuseunderpracticalconditions(5).On-lineproce­
24 samples (blind duplicates), 2 samples of which duresarepossibleby usingroboticanalysis(6)ordialysis(7).
were blanks. A preparatory group, consisting of the organizer and 3
members, examined andtestedseveralmethods thatuseaffin­
ity columns and proposed a subjoined method, after a pilot
R e c e iv e d S e p te m b e r 16, 1 9 9 2 . A c c e p te d J a n u a r y 2 5 , 1 9 93.
studywithapracticesample was performed.
 T uinstra E t A l .: Journal Of AOAC International V ol . 76, No. 6,1993 1249

Interlaboratory Study (d) A fla to x in M t sta n d a rd . — Sigma Chemical Co., St.

Louis, MO, orMakor Chemicals, Jerusalem, Israel.(7)S to ck
The 16participatinglaboratorieseachreceived6 randomly so lu tio n . 0.1 pg aflatoxinM|/mL chloroform. Stoppertightly,
codedsamplesofmilkpowder (A,B,andC) withlow fatcon­ wrapinaluminumfoiltoexcludelight,andstoreinarefrigera­
tent(about1%) and6randomlycodedsamplesofmilkpowder torat<5°C inthedark.(2)W orking d ilu tio n s. Beforepreparing
(D,E,andF)withhighfatcontent(about28%). Theselatter6 workingdilutionsoftheaflatoxinM, standard,the0.1pg/mL
samples were remainders ofbatches ofmilk powders used to solution,previously storedinarefrigerator,must attainambi­
prepare certified reference materials; therefore, the aflatoxin enttemperature.
Mj contentofthesesamples was known totheorganizer.The Transfer 1.0 mL of the 0.1 pg/mL solution by pipet to a
contentofsampleA-F wereunknown totheparticipants.Sam­ 20 mL conical tube.Evaporate the solutiontodryness with a
plesB andC withlow fatcontentoriginatedfromThe Nether­ constantstreamofinertgasanddissolvetheresiduein20.0mL
landsInstituteforDairyResearch (NIZO, Ede) and were used 10% acetonitrile.Shake occasionally over 30 min. This solu­
asinternalcheck samples. tioncontains0.005pg aflatoxinM|/mL. Use thisdilutedsolu­
All samples were present as blind duplicates, insufficient tion toprepare a series of dilutions of aflatoxin M! standard
quantities for a single extraction procedure according to the solution,containing0.1,0.2,0.4,and0.8ng aflatoxinM,/mL,
method protocol,i.e.,10g milkpowder. Two ofthe 6 low fat respectively,in 10% acetonitrile.
sampleswere blanks.The contaminationlevelofaflatoxinM!
E q u ip m e n t
in the other 10 samples varied from 0.08 to 0.6 pg/kg milk
powder, i.e.,8-60 ng/kgreconstitutedmilk. (a) D is p o s a b le s y r in g e s. —Capacity 10and 50 mL.
Participantsalsoreceivedapracticesample(20g,sufficient (b) Vacuum sy s te m . — Büchner flask, Vac-Elut system, or
for2analyses)withaknown aflatoxinM] content,anunknown peristalticpump.
practice sample (10 g), 15 immunoaffinity columns, and a (c) C e n trifu g e . — Capableofgenerating lOOOxg , fordefat­
standard solution of aflatoxin Mj inchloroform (0.1 pg/pL). tingthemilk.
The resultobtainedwith theunknown practicesample was to (d) P ip e ts. — Capacity 1.0,2.0,and50.0mL.
be returnedby FAX totheorganizer,beforetheparticipantre­ (e) W a ter b a th . — Capable ofoperatingat50°C.
ceivedpermissiontoproceedwiththeotherunknown samples. (f) G r a d u a te d c o n ic a l g la s s tu b e s. — With ground glass
Participantsreceiveddetailedinstructionsaboutthemethod neckand stopper,capacity 10 and 20 mL.
tofollow,includingatestforthelinearityoftheLC system. (g) L C a p p a ra tu s. — Consisting of injector fitted with
500 pL loop,solventdeliverysystem,andfluorescencedetec­
METHOD torwith X ex of360 nm and X em of435 nm, reversed-phase
analyticalcolumn, and guardcolumn.
Principle N o te: ASpherisorbS5 ODS2 (12% C loading)(PhaseSepa­
rationsInc.,Norwalk, CT), 25 x 4.6m m id,and aChrompack
AflatoxinM, isextractedfrom themilk sample by passing LC cartridge(Chrompack InternationalB.V.,Middelburg,The
itthrough an immunoaffinity column. The column contains Netherlands), 5 pm Chromspher-Ci8,20 ± 1 cm (including
specificmonoclonal antibodiesbound toasolidsupportmate­ guard column) x 3 mm idaresuitable.A guard column filled
rial.As thesample passes through thecolumn, the antibodies withreversed-phasematerialmay be used.
selectivelybind with aflatoxinM t(antigen)presentand form
an antibody-antigen complex. All other components of the D e term in a tio n
sample matrix are washed offthe column with water. Then, See Remarks.
aflatoxinM| iselutedfromthecolumn withacetonitrile,which (a) Im m u n oaffin ity c o lu m n p r e p a r a tio n . —
Attach the bar­
iscollected.The amount ofaflatoxinM! presentinthiseluate relofa50 mL disposable syringetothetop ofan immunoaf­
is determined by liquid chromatography coupled with finitycolumn.Connecttheimmunoaffinitycolumn tothevac­
fluorimetricdetection. uum system.
R e a g e n ts (b) E x tra c tio n . — (7) Milk.— Warm the milk to 35-37X3
andeitherfilterenoughmilkthroughWhatman No.4 filterpa-
Use analyticalgradereagentsanddistilledwaterorwaterof peps) orcentrifuge 15 min at1000 X g . Collectatleast50 mL
atleastequivalentpurity. milk (ifnecessary, use several filters).Pipet50 mL miff into
(a) Im m u n o a ffin ity c o lu m n .— Containingmonoclonalanti­ the syringe barrel and letitpass through the immunoaffinity
bodies againstaflatoxinM, (Rhone Poulenc Diagnostics Ltd., column ata slow, steady flow rateof 2-3 mL/min, using the
Glasgow, Scotland). vacuum system to control the flow rate. Continue as in (2),
(b) A c e to n itr ile so lu tio n s. — (7) 25% (v/v) acetonitrile in beginning“Remove the50 mL syringebarrel....”
water:Dilute250 mL acetonitrileto 1000 mL withwater.De­ (2) Milkpowder.— Weigh 10g milkpowder,tothenearest
gas before use. (2) 10% (v/v) acetonitrile in water: Dilute 0.1g,intoa250mL beaker.Take50mL waterwarmed to50X1
100mL acetonitrileto 1000 mL withwater. and add thisinsmall amounts tothe milkpowder. Mix, using
(c) C h lo ro fo rm .— Stabilized with 0.5-1.0% ethanol by a stirringrod, until a homogeneous mixture isobtained (see
mass. Remarks). Let the solutionofmilkpowder cool to20X3 and
1250 T uinstra E t A l .: Journal O f AO AC International V ol . 76, N o. 6 ,1 9 9 3

thenquantitativelytransferittoa 100mL volumetricflask,us­

ingsmallamountsofwater.Dilutethemilkpowder solutionto
the mark. Filterenough reconstitutedmilk through Whatman
No. 4 filterpaper(s) orcentrifuge 15 min at lOOOx g , transfer
50 mL solutionintothesyringebarrel,and letitpass through
the immunoaffmity column ata slow, steady flow rateof 2-
3 mL/min, usingthevacuum systemtocontroltheflow rate.
Remove the50 mL syringebarreland replacewitha clean
10mL syringebarrel.Wash thecolumnwith 10mL water.Pass
the water through the column at a steady rate, and afterthe
washing,blow thecolumn completelydry.Disconnectthecol­
umn fromthevacuum system.
Slowly eluteany aflatoxinMj from thecolumn by passing
4 mL acetonitrilethroughthecolumn, using a 10 mL syringe.
The acetonitrileshouldtakeca60stopassthroughthecolumn;
controltheflowratewiththesyringeplunger.Collecttheeluate
in a 10 mL conical tube. Reduce the volume of the eluate to
300-500 pL at30°C,usinga streamofnitrogen. r e t e n t io n t im e ( m in )
N o te: Lossesmay occurifevaporatedtocomplete dryness.
F igu re 1. C h ro m a to g ra m o f a m ilk p o w d e r extract
Diluteto5.0mL withwater(seeRemarks).
(labo ratory 3) c o n ta in in g aflatoxin M i (190 n g /k g m ilk
(c) L iq u id c h ro m a to g ra p h y . — (7) Pump the eluant at a pow der). L C c o n d itio n s a s in text.
constantflow ratethrough theLC column. Injectinsequence
500 pL aflatoxin Mj standard solutions containing 0.05, 0.1,
0.2, and 0.4 ng aflatoxin M,. Prepare a calibration graph by
plottingthepeakareaorpeakheightforeachstandardagainst (b) M ilk p o w d e r .
— Calculate the aflatoxin Mi content of
thequantityofaflatoxinM, injected. thesampleinpg/kgby usingthefollowingformula:
N o te: Ifnecessary(depending on thetypeofcolumn used),
theacetonitrile:waterratiooftheLC eluantmay be adjustedto J x = pg/kg aflatoxinMi
ensureanoptimalseparationoftheaflatoxinMj fromanyother
extractcomponents. The flow rateoftheeluantalso depends
on the column. As a guideline forconventional columns (ca
25 cm x 4.6mm id),aflowrateofca 1mL/min givesoptimal where a = ng aflatoxin Mi corresponding to the area of the
results;forLC columns (3mm id),aflowrateof0.5 mL/min aflatoxinMi peakofthesampleextract;Vm = volume ofsam­
isoptimal. The bestconditionscan be determined by using a p leextractinjected(500pL); Vext = volumeinwhich thesam­
sample extract (preferably free from aflatoxin Mi), which is p le extract is dissolved (5000 pL); M = volume of milk
injectedbothincombinationwithanaflatoxinM! standardand ( 50mL) ormass ofmilkpowder(5g)passingthroughthecol­
alone.SeeFigure 1. umn.
This formulaisapplicableonlywhen no dilutionhas been
Inject500 pL sample extractintotheLC apparatus viathe
carriedout.Otherwise,dilutionshouldbe takenintoaccount.
injectionloop,and separateany aflatoxinM, presentby using
the same conditions as forthe standard solutions. Determine
Rem arks
theareaoftheaflatoxinM! peakofthesampleextract.
From thecalibrationgraph,determinengaflatoxinM tinthe
sampleextract.IftheareaorheightoftheaflatoxinM! peakin The method described in thisprotocol requires the use of
the sample extractisgreaterthan thatofthehigheststandard solutionsofaflatoxin M|. This compound isa carcinogen for
solution,dilutetheextractquantitativelywithwaterandre-in­ some rodentstrainsand shouldbe handledasacarcinogen.
jectintotheLC apparatus. The laboratory where the analyses are performed must be
N o te: Ifaseriesofsampleextractsisbeinginjectedoneafter
a deq ua telyprotectedfromdaylight(darkenedlaboratory,gold
the other, an aflatoxin Mi standard should be injected after l ight), and aflatoxinstandardsolutionsmust be keptprotected
every5 injectionsofsampleextracts. from li ghtby usingaluminum foil.
The use of nonacid-washed glassware (e.g., tubes, vials,
C a lc u la t io n
flasks,beakers,syringes)foraqueous aflatoxinsolutionsmay
cause aflatoxinloss.Particularcareshouldbe takenwith new
(a) M ilk .— Calculate theaflatoxinM[ contentofthesam­ glasswareanddisposableglassware,suchasautosamplervials
pleby usingthefollowingformula: and Pasteurpipets. Therefore, laboratory glassware thatcon­
tactsaqueous solutionsofaflatoxinshouldbe soaked indilute
acid (e.g., sulfuric acid, 2 mol/L) for several hours and then
x = pg/L aflatoxinMi rinsed well with distilled water to remove all traces of acid
Vm M
 T uinstra E t A l .: Journal O f AOAC International V ol . 76, N o. 6 ,1 9 9 3 1251

(checkwithpH paper) (H.P.vanEgmond, personalcommuni­ particulatesfrom thedrainedmilk. Furthermore, theenergyof

cation). the fluorescence detectorlamp significantlydecreased during
Ifthe milk powder isnotcompletely dissolved, thebeaker theanalysis.
should be placed ina 50^0 waterbathforatleast30 min, and Laboratory 6 reportedno orlow recoveryforaflatoxin
thesolutionshouldbe mixed regularly. inthepracticesamples.ExperimentswithaflatoxinM, inwater
When the acetonitrile content ofthe injected extractcon­ alsogave no recovery. Therefore, new practice samples were
taining aflatoxinM! exceeds the 10% (v/v)limit,peakbroad­ distributed,togetherwith new affinitycolumns. Resultswere
eningwilloccur.However, awatercontentof>90% (v/v)has thesame; no aflatoxinM, was recoveredandno datawerere­
no influenceon thepeak shape. ported.
Laboratory 7 reported an increase of retention time (10-
Results and Discussion 20%) duringtheLC measuring period on 1day.The tempera­
tureinthelaboratoryincreasedto33°C.Laboratory8requested
Resultsnotcorrectedforrecoveryasreportedbyeachlabo­ the elutionofthecolumn by gravity only insteadofvacuum.
ratory arepresentedinTable 1forthepracticesamples and in Laboratory 9 used a syringepump forelutingthecolumn. As
Table2 forthefinalstudysamples. the reconstituted milk passed easily through the column, no
Laboratory 1reportedproblems with the sensitivityofthe filtrationorcentrifugationwasused.Laboratory 14usedacali­
detectionsystemand,therefore,carriedoutaprecolumn deri- bration curve foronly 3 datapoints insteadofatleast4.The
vatization with trifluoroacetic acid (TFA). Results for the lowestquantityproposed inthemethod forinjection(0.05ng)
known practicesample were much toolow (approximate fac­ was not injected. This quantity, according to the method, is
tor, 10).The recovery experiments with thelaboratory’sown equivalentto 100 ng/kgmilkand appearedlatertobethelimit
milkpowderwerelow.Attheend,theimmunoaffinitycolumn ofdetectionforthislaboratory.
forthe unknown samples appeared tobe elutedwith acetone Laboratory 15 concentrated the acetonitrile fraction from
instead ofacetonitrile,because precolumn derivatizationwas the column to 300 pL and diluted to 2000 pL, i.e.,the ratio
required. Therefore, the reported data was not included (Ta­ CH3CN:H2wasabout1:5.7.Accordingtothemethodprotocol,
ble2)inthestatisticalcalculations. theCH3CN contentshouldbe<10% toavoidpeakbroadening.
Laboratory 2 reported slow filtration of the reconstituted Thisindeedshowedup inthechromatogramsoflaboratory 15.
milk made from high fatmilk powder. Laboratory 4 had the See alsoFigure2.
same problem even after centrifuging. Laboratory 5 centri­
fuged allsamples and then used filtrationtoremove any large

T able 1. R e p o rte d aflatoxin M i con ten t in the practice

s a m p le s of high -fat m ilk p o w d e r (n g /k g m ilk po w der)

Known sample (580 ±

170 ng aflatoxin
Laboratory M,/kg)a Unknown sample a
b

b b
1
2 646 184
3 593 200
4 570 190 y_ -A-

5 505 315
b b
6
7 645 211

8 686 246
9 530 207
10 716 182
11 588 185
12 586 204
13 608 190
14 500 223
15 485 190
16c
17 593 244 F igu re 2. A flato xin Mi sta n d a rd (each injection 4 ng),
injected in 500 pL so lv e n t mixture, with c h a n g in g
a Mean value, found by participants of the preparatory group during
the pilot study. C H 3 C N /H 2 O ratio (a = 5:95; b = 10:90; c = 15:85; d =
b Results not included, see text. 20:80). P e a k b ro a d e n in g d u e to in c re a sin g acetonitrile
c Participation cancelled. content. P e ak width: a, 2.5; b, 2.7; c, 3.0; d, 5.1 mm.
1252 T uinstra E t A l .: Journal O f AO AC International V ol . 76, N o. 6 ,1 9 9 3

T able 2. R e p o rte d aflatoxin M i con ten t in the final s tu d y s a m p le s (n g /k g m ilk po w der) an d limit o f detection

Sample
A B C D E F
LOD (ng/kg
Laboratory LFa LF LF HFb HF HF milk powder)

1c NDrf 144 156 34 84 352 _

ND 100 ND 97 176 520
2 ND 65.9 132 88.4 220 602 25
ND 80.8 134 88.8 205 608
3 ND 77.2 153 84.3 190 571 10
ND 78.1 153 82.1 189 592
4 ND 67.0 155 85.3 191 572 40
16.9 72.5 148 72.6 204 607
5 46 112 69 54 167 427 50-60
ND 61 145 48 253 339
6e — — — — — — —
7 ND 89 155 93 200 630 10
ND 85 156 91 214 586
8 ND 91 168 106 220 597 10
ND 85 168 94 210 698
9 ND 69 141 75 172 569 30
ND 77 128 71 183 649
10 ND 6.6 126 66.9 166 536 10
ND 73.0 136 74.0 190 558
11 ND 70.7 141 70.7 186 539 30
ND 79.6 150 79.6 190 575
12 ND 102 172 77.6 250 809 80
ND 124 179 89.4 253 740
13 ND 71.7 163 77.2 204 650 40
ND 85.2 151 82.6 202 654
14 ND 125 215 100 210 470 100
ND ND 180 294 186 690
15 279 62 ND 64 322 400 50
ND 45.6 105 58.8 128 390
16f — — — — — — -- ---
17 ND 104 192 104 800 454 150
ND ND 260 96 276 708

a LF, milk powder with low fat content.

b HF, milk powder with high fat content.
c All data eliminated for statistical analysis.
d ND, not detectable (below limit of detection), according to participant.
e No results received, see text.
' Participation cancelled.

Participants were asked to provide a chromatogram of a sitivityofthedetectorforthegivencompoundbutisalsoinflu­

relevantsample and achromatogram ofan M, standardsolu­ enced by thenoise (chemical and electronic).Ifthechemical
tion.Most participantssentthechromatogram ofsample E or noise isnegligible, LOD can be improved by injectingmore
F. The authors derived the limit of detection (LOD) (sig- sample. In thisexercise, a few laboratories appeared to have
nalinoise ratio is 3:1) from these chromatograms. At a later injectedasmallerquantityofsamplethandescribedinthepro­
stage,theauthorsrequested5 oftheremaining 14participants tocol (an amount equivalent to 0.5 g milk powder was re­
tofurnishchromatograms ofsamples B orD forbettercom­ quested).Nevertheless,thederivedLOD wasclearlybelowthe
parisonofLOD obtainedfromtheextrapolationwithLOD ob­ reported resultsofthesefew laboratories. Laboratories 5, 12,
tainedfrom sample B orD. 14, 15, and 17 injected the requested quantity equivalent to
The resultingLODs arereportedinthe lastcolumn ofTa­ 0.5g milkpowderbutreportedresultsaroundorbelow thede­
ble2.ItissurprisingthatdifferencesinLOD rangeoveratleast rivedLOD forsome samples.With 1exception(laboratory12),
1orderofmagnitude. LOD isnotonlyinfluencedby thesen­ alloftheselaboratoriesusedfilterdetectors.From thedataof
 T uinstra E t A l .: Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3 1253

T able 3. S u m m a ry o f sta tistic al a n a ly s is o f co llab o rativ e data fo r the L C determ ination of aflatoxin M i in m ilk powder,
u s in g im m u n oaffin ity c o lu m n cle a n u p

Parameter3 Dim Known Unknown B C D E F

n 14 13b 12c 14 13d 11® 14

m ng/kg 589.36 204.31 81 150 80 202 580
r ng/kg 23 60.1 15 27 203
R ng/kg 52 98 41 61 310
RSDr % 9.9 14.0 6.8 4.7 12.5
RSDr % 11.7 10.6 23.0 22.7 18.3 10.8 19.1
Sr ng/kg 8.05 21.68 5.39 9.52 72.59
SR ng/kg 69.04 21.66 18.63 35.08 14.62 21.84 110.53

8 n = number of participants; m = arithmetic mean; r = repeatability; R = reproducibility.

b Laboratory 5, Grubbs test outlier.
c Laboratory 5, Cochran outlier.
d Laboratory 14, Cochran and Grubbs test outlier.
8 Laboratory 5,15, Cochran outlier; Laboratory 17, Chochran and Grubbs test outlier.

theaflatoxinM, standard,thederivedLOD oftheselaborato­ study, whose data are not corrected forrecovery, obtained a
ries appeared tobe poorer than thatof the other laboratories good result.
usingdetectorswithmonochromators. TheaflatoxinM! contentinsamplesD,E,andF,made from
Theextenttowhichparticipantsusedelectronicfiltersisnot remaindersofbatchesofmilkpowder,usedtopreparecertified
known. Obviously, when a high LOD isneeded, preference reference materials, and corrected for recovery, are, respec­
shouldbe giventodetectorswithmonochromators, insteadof tively: 90 +40/-20,310 ± 60,and760± 50 ng/kg (9,10).
filters.Sample A was ablanksample;nevertheless,laboratory The indicated 95% confidence interval covers the true
5,and especiallylaboratory 15,reportedthepresenceofafla­ value.Comparedtotheresultsoftheparticipants(notcorrected
toxinMj in 1blindduplicate.Inview ofthelevelreportedby forrecovery) inthisstudy,resultsseem todiffergreatly,espe­
laboratory 15, an interchange of samples seems likely, but ciallyforsamplesE andF.The recoverycorrectionusedinthe
couldnotbeexplained. BCR study (9, 10), however, can account forup to 20% for
StatisticalanalysiswascarriedoutaccordingtoISO 5725 to materialF.
determineoutliersandtocomputerepeatabilityandreproduci­ Therefore, based on a general RSDr ofabout 20% inthis
bilityvaluesfrom theraw datagiveninTables 1and 2.How­ study (Table 3), there isa considerable overlap between the
ever,datafromlaboratory1arecompletelyeliminatedfromthe mean contentofsamples D, E, and F found inthisstudy and
those oftheBCR exercise. Nevertheless, the values forsam­
statisticalanalysis.In Table 3,summarizing statisticalresults
ples E and F obtained inthe BCR exercise tend tobe a little
fromTables 1and2 aregivenforthepracticesamplesand un­ higher.Separatelyperformedrecoveryexperimentscarriedout
known milkpowder samples aftereliminationoftheCochran atRIKILT-DLO attheaflatoxinMj levelsofsamplesE and F
andthe(double)Gmbbs testoutliers. didnotgiveany indicationthatthecapacityoftheimmunoas­
Except for sample E, RSDr isabout 20% over the tested saycolumn was thereason.The recoveryof1|igaflatoxinM,
rangeforthelowfatandhighfatmilkpowders.SampleE gives appliedtothecolumn was 90%.
most outliers(3laboratoriesby theCochrantestand 1labora­ The extraction procedure forfluidmilk isincluded in the
tory by the Gmbbs test).RSDrisabout 10-15%, except for method description,even when nottestedinthisstudy.From
samplesD and E,forwhich lowervalueswerefound. Inview experiments during thepreparation ofthisexerciseinthe or­
ofthelow aflatoxinM, levels,theseresultsshouldbe consid­ ganizer’slaboratoryand inthelaboratory ofanothermember
eredasverygood.ResultsobtainedwithotherLC methodsand of the preparatory group (J.M. Fremy, personal communica­
higheraflatoxinMj levelsreportRSDr valuesof43 and27.9% tion),resultswere good, and no practicalproblems were met.
(8),butinthelatterexerciseonly spikedsamples were used. Currently, thisprocedure forfluidmilk isbeingring-testedin
Not known totheparticipants,ofcourse,was thefactthat France.The fluidmilk shouldapparentlybecentrifugedatthe
theknown practice sample was identicalto sample F and the highestpossiblespeed,even >1000 x g .
unknown practice sample was identicalto sample E. Partici­
pantsfoundessentiallythesame values.Furthermore,forsam­ C onclusions
ple F,the same contentwas obtained 6 months earlierby the
members ofthepreparatorygroup (seeTable 1).The aflatoxin Sixteen laboratories from 11 countries participated in a
M| contentofsamplesB andC correctedforrecoveryare,ac­ studyofthedeterminationofaflatoxinM, inmilkpowder.Five
cordingtoNIZO, 80and 150ng/kg,respectively.As therecov­ samples, containing aflatoxin M, levels in the range of 80-
ery isalways between 90 and 100%, the participants in this 600 ng/kg milk powder, were analyzed as blind duplicates.
1254 T uinstra Et A l .: Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3

Fourteen laboratoriesproducedacceptableresults.The overall J.van den Bedem,* Netherlands controlling authority for
RSDr rangedfrom 11to23%, averyacceptableresult.Hence, milkand milkproducts,Leusden,The Netherlands
immunoaffinitycolumns shouldbe consideredasan appropri­ R. van Renterghem, Rijkszuivelstation,Melle, Belgium
ate tool for aflatoxin M, determination atlow levels of con­ J.M.P.vanTrijp,StateInstituteforQualityControlofAgri­
tamination. cultural Products (RIKJLT-DLO). Wageningen, The Nether­
lands
A c k n o w le d g m e n t s
Yau-Huei Wei, NationalYang-Ming Medical College,De­
partmentofBiochemistry,Taipei,Taiwan
We thanktheparticipantsinthisstudy and especiallythose
ofthepreparatorygroup(marked with *).
R e fe re n c e s
P.A.Burdaspal,CentroNacionaldeAlimentación,Madrid,
Spain
B. Declercq,LaboratoireInterregionaldelaRepressiondes (1) Scott, P.M. (1989) F o o d A d d i t . C o n ta m . 8, 283-305
Fraudes de Paris-Massay, France (2) M ortim er, D .N ., G ilbert, J., & Shepherd, M .J. (1987) J. C h r o -
M.F. Dutton, University ofNatal, Pietermaritzburg, South m a to g r. 407, 393-398

Africa (3) Hansen, T.J. (1990) J. F o o d P r o t. 53, 75-77

K.F.Donahue, Vicam, Sommerville,MA, USA (4) Steimer, J., Hahn, G ., Heeschen. W ., & Bluthgen, A . (1988)
J.M. Fremy,* Laboratoire Central d’Hygiène Alimentaire, M ilc h w iss e .m c h a .ft 43, 772-776
Paris,France (5) N ieuw enhof, F.F.J., H oolw erf, J.D., & Bedem, J.W. van den
A. Pittet,Nestlé,Vevey,Switzerland (1990) M ilc h w is s e n s c h a f t 45, 5 8 ^ 5 8 8
G.Rasmussen,NationalFoodAgencyofDenmark,Soborg,
(6) G iffo rd , L .A ., W right, C „ & G ilbert, J. (1990) F o o d A d d it.
Denmark
C o n ta m . 7, 829-836
MA. Rutschmann, Migros Central Laboratoire, Zurich,
Switzerland (7) Tuinstra, L.G .M .T h., Kienhuis, P.G.M ., & D ols, P. (1990) J.
A s s o c . O ff. A n a l. C h e m . 73, 969-973
M. Sharman, Food Science Laboratory, MAFF, Norwich
UK (8) S tubblefield, R.D., & K w olek, W F. (1986) J. A s s o c . Off.
R.D. Stubblefield,U.S. DepartmentofAgriculture,Peoria, A n a l. C h e m . 69, 880-885

Illinois,USA (9) Com m ission o f the European Com m unities, C om m unity B u­

M. Toyoda,NationalInstituteofHygienicSciences,Tokyo, reau o f References, BCR. The ce rtifica tio n o f aflatoxin M , in
Japan three m ilk powder samples, CRM No. 282, 284, 285. H.P.
H. P.vanEgmond,* RijksinstituutvoorVolksgezondheiden van Egmond, P.J. W agstaffe, (1986) EUR 10412.
Milieuhygiëne,Bilthoven,The Netherlands (10) Addendum to report EUR 10412
 Espinosa -M ansilla E t Al.: J ournal Of AOAC International V ol . 76, No. 6,1993 1255

FO O D C H E M IC A L C O N T A M IN A N T S

Sem iautom atic D eterm in atio n o f F u ra n ic Aldehydes in Food and

P harm aceutical Samples by a Stopped-Flow In jec tio n Analysis
M eth o d

A. E s p in o s a - M a n s il l a , A. M u n o z d e l a P ena, and F. S a l in a s
University of Extremadura, Department of Analytical Chemistry, 06071 Badaoz, Spain

A kinetic stu d y of th e re a c tio n s of 5-hydroxym ethyl- specificity because they determine total furanic aldehydes. De­
2 -furfuraldehyde a n d furfural with 2-thiobarbituric rivative spectrophotometiy has been used in our laboratory to
acid (TBA) by a sto p p ed -flo w flow injection a n a ly ­ resolve this problem, and simultaneous determination of HMF
s is te c h n iq u e h a s b ee n u n d erta k en . A se m ia u to ­ and furfural has been proposed on the basis of the Winkler re­
m atic m eth o d fo r th e analytical determ ination of action (6 ).
th e s e fu ran ic a ld e h y d e s is p ro p o se d o n th e b a s is On the other hand, flow injection analysis (FIA) techniques
of reactio n with TBA. T he p ro p o se d stopped-flo w have been applied to the differential kinetic determination of
m eth o d w a s su c c e ss fu lly applied to sev e ral co m ­ furfural and vanillin (22) in artificial samples, and an FIA semi­
m ercial p h arm aceu tical p re p a ra tio n s a n d food s a m ­ automatic method based on the Winkler reaction has been pro­
p les. T he p ro c e d u re is fa ste r th a n th e earlier p ro c e­ posed for the determination of HMF in honey (17).
d u re for d eterm in atio n of th e s e c o m p o u n d s in The determination of furfuraldehyde by a continuous FIA
fo o d s a n d p h arm a ceu tica ls. method based on the classical reaction with aniline has been
described (23). This method avoids the instability problems of
the aniline reaction products and has been used to determine
5 -Hydroxymethyl-2-furfuraldehyde (HMF) is the princi­ furfuraldehyde in alcoholic beverages that are only slightly ab­
pal decomposition product of the acid-catalyzed hydroly­ sorbing at the measurement wavelength, such as cognac, gin,
sis of glucose and fmctose, and furfural is the principal whiskey, and mm (24). However, the continuous FIA approach
product of the hydrolysis of pentoses (1). Both seem to be re­ cannot be applied directly to highly colored or turbid samples
lated to the browning of diverse foods during storage and are (e.g., honey, juices, and red wines).
also useful as indicators of temperature abuse (2 ). In this paper, a kinetic study of the reactions of HMF and
The analysis of these compounds is very important in the furfural with 2-thiobarbituric acid (TBA) based on a stopped-
food industry. For example, to meet government requirements flow technique is described. The influence of temperature and
pH on the slow reaction between TBA and furanic aldehydes
for cognac, the furfural formed from pentoses during wine dis­
was investigated. The purpose of this work was to develope an
tillation must not be less than 1.0 g/100 L ethanol. For many
FIA system for the semiautomatic determination of furanic al­
years, honey has been known to contain HMF arising from the
dehydes in diversely colored and turbid food samples to avoid
action of normal honey acidity on levulose at ambient tempera­
the problems encountered with the continuous FIA methods.
tures and during heat processing or storage at elevated tem­
The stopped-flow process makes pretreatment of the samples
perature. A maximum content of 4 mg HMF/100 g of honey is
unnecessary and allows direct determination.
allowed by the Dutch Food Law to ensure that table honey has
not been denatured by heat (3).
Experim ental
Recently, several authors have determined these com­
pounds in citrus juice (4-7), grape juice (2,7, 8 ), wine (9, 10),
plant extracts (11), pharmaceutical syrups (12-15), honey (6 , Apparatus
16,17), milk (18), and spirits (19).
A home-made stopped-flow FIA system (Figure 1) was fit­
Spectrophotometric techniques have been used to deter­
ted to a Beckman DU-50 spectrophotometer equipped with a
mine HMF and furfural in foods. Dinsmore and Nagy (20) de­
10-mm microflow cell (18 pL) connected via RS-232 to an
scribed a colorimetric method for determining furfural in citrus
IBM PC-XT 286 microcomputer. The Beckman Data Leader
juice on the basis of the reaction with aniline. HMF is often software was used for data acquisition and measurement and
determined by using the Winkler reaction or the reaction with analysis of the kinetic data. A thermostatic bath (Selecta) and a
2-thiobarbituric acid (TBA) (21); however, these methods lack thermostated cell holder were used for temperature control.
A precisian peristaltic pump (Gilson Minipuls-2) and a 6 -
R e c e iv e d F e b u a ry 28 , 199 2 . A c c e p te d M a r c h 12, 1993 way injection valve with a variable loop (Omnifit) were used.
1256 E spinosa -M ansilla E t A l .: J ournal Of AOAC International V ol . 76, No. 6,1993

Analytical signal
N Td Ts
j
E
« 7 ------ ♦ «---- --------- e------------ — ►
c
T -----
4------- 40 «----------- 6 0 s —
I S —

O
N

(b)
F ig u r e 1. (a) S c h e m a tic d ia g ra m o f s to p p e d - flo w F IA s y s te m fo r th e d e te rm in a tio n o f f u r a n ic a ld e h y d e s ; (b) T im e
p ro g ra m . T h e in s e r t s h o w s V, in je c tio n w a v e ; W, w a ste ; T d , d e la y tim e ; T s, s t o p tim e ; a n d t, tim e .

F ig u r e 2. (a) A b s o r p tio n s p e c tra o f T B A - f u r f u r a l (4.6 p g /m L ) a n d T B A - H M F (5.3 p g/m L). R e a c tio n tim e , 4 m in . (b)
A b s o r b a n c e - t im e c u r v e s o b ta in e d fo r th e T B A - f u r f u r a l a n d T B A - H M F re a c tio n s .
 E spinosa -M ansilla E t Al.: J ournal O f AOAC International V ol . 76, No. 6,1993 1257

POKE instruction in the 56579 memory address, the corre­

sponding data lines are activated. By activating (POKE 56579,
8 ) or deactivating (POKE 56579, 2) a relay, the pump is turned

in a direction or is stopped. By activating or deactivating a sec­

ond relay, the pump is turned in the reverse direction or is
stopped.
The time program used for the stopped-flow method is de­
scribed in Figure lb. The delay time (Td) and the stop time (Ts)
are controlled by input instructions through the internal clock
of the Commodore 64 microcomputer.

Reagents

(a) 2-Thiobarbituric acid.—A 3 x 10- 2 M solution was pre­

pared by dissolving 1.14 g reagent (Sigma) in 250 mL of water.
(b) Hydroxymethyljurfural stock solution.—A 0.01% (w/v)
solution was prepared by dissolving 0.01 g reagent (Carlo
Erba) in 100 mL of water.
F ig u re 3. P lo t o f th e lo g a rith m o f th e re a c tio n ra te v s (c) Furfural stock solution.—A 0.01% (w/v) solution was
th e n e g a tiv e lo g a rith m o f th e h y d ro g e n io n
prepared by dissolving 0.01 g reagent (Merck) in 100 mL of
c o n c e n tra tio n .
water.
All chemicals and solvents used were analytical reagent
grade.
The starting, stopping, and turning direction of the pump were
manually controlled by a switch located on the front panel of Determination of HMF and Furfural by Stopped-Flow
the pump. An interface was built to allow the starting, stopping, FIA System
and control of turning direction of the pump. The interface sub­
A 100 pL test solution containing up to 1.5 pg of HMF or
stitutes for the manual functions of the pump switch, allowing
furfural was injected through the injection valve into the carrier
the automation and control of the peristaltic pump through a
stream of water by using the manifold shown in Figure la. The
microcomputer. The interface was connected to the user port of computerized timer was used to control the peristaltic pump
a Commodore 64 microcomputer in which data lines were pre­ and the injection valve. The time program is shown in Figure
sent. lb. The kinetic curve and initial rate measurements were auto­
A program written in BASIC was used to activate ar.d deac­ matically obtained by the computer system interfaced to the
tivate a relay by means of the POKE instruction. By writing a spectrophotometer and by using the Data Leader software. The

F ig u r e 4. (a) E ffe c t o f H M F c o n c e n tr a tio n o n th e re a c tio n ra te a t s e v e r a l te m p e ra tu re s , (b) P lo t o f th e lo g a rith m o f th e

ra te c o n s ta n t (K ) v s th e in v e r s e o f th e a b s o lu te te m p e ra tu re .
1258 E spinosa -M ansilla E t Al.: J ournal Of AOAC International V ol . 76, No. 6,1993

F ig u r e 5. T h re e - d im e n s io n a l is o m e tr ic p ro je c tio n o f th e v a ria tio n o f th e re a c tio n ra te w ith te m p e ra tu re a n d H M F

c o n c e n tra tio n .

reaction was monitored at 425 nm, and the temperature of the Syrup oflevulose and dextrose.—For each sample, 5 mL of
flow cell was kept constant at 40 ±0.1 °C. a sugar syrup (Apiroserum, YBYS, intravenous; dextrose, B.
Martin, oral, pediatric solution) is placed in a 25 mL volumetric
Determination of Furanic Aldehydes in Food flask and diluted to the mark with distilled water. The samples
Samples and Pharmaceutical Formulations are analyzed as described above.

Honey.—About 10 g honey is dissolved in 25 mL distilled

R esu lts a n d D iscu ssio n
water, and an ultrasonic bath is used to dissolve the sample.
Aliquots (5 mL) of this solution are placed in 10 mL volumet­
Furanic aldehydes react with 2-thiobarbituric acid in acid
ric flasks and diluted to the mark with distilled water. The sam­
medium to form a yellow product according to the reaction
ples are then analyzed by the stopped-flow method. Conven­ shown in Scheme 1.
tional calibration graphs and the method of standard additions The spectral features of the reaction products formed are
are used. shown in the Figure 2a. Absorbance-time curves are shown in
Orange and red grape juices.—For each sample, 10 mL of Figure 2b for HMF and furfural concentrations of 5.3 and
commercial juice is placed in a 25 mL volumetric flask and 4.6 |0 .g/mL, respectively.
diluted to the mark with distilled water. Aliquots (5 mL) of the
solution are placed in 10 mL volumetric flasks, diluted to the Kinetic Behavior
mark with distilled water, and analyzed by the stopped-flow
A kinetic study of the effect of concentration of TBA, pH,
technique as described above. Conventional calibration graphs
and temperature on the rate of the condensation reaction was
and the method of standard additions are used.
made with HMF samples, and later, the calibration curves were
Red wine.—For each sample, 5 mL of commercial red wine tested on furfural samples.
is placed in a 25 mL volumetric flask and diluted to the mark The kinetic data were obtained from the initial plots of rate
with distilled water. The samples are analyzed as described versus concentration by using the manifold and time program
above. shown in Figures la and lb, respectively. The partial orders for
Sweet red wine.—For each sample, 0.1 mL of sweet red each variable were calculated from the resulting log-log plots.
commercial wine (Malaga type) is placed in a 25 mL volumet­ Samples containing HMF were injected, and the temperature
ric flask and diluted to the mark with distilled water. The sam­ was maintained constant at 40 ± 0.1 °C by using a thermostated
ples are analyzed as described above. flow cell.
 E spinosa -M ansilla E t A l .: J ournal Of AOAC International V ol . 76, No. 6,1993 1259

OH variation of reaction rate with temperature and concentration is

plotted in Figure 5. The response surface clearly indicates the
exponential relationship between reaction rate and temperature
and the linear relationship between reaction rate and HMF con­
centration.
R = - C H 2O H , 41

S c h e m e 1. R e a c tio n o f fu r a n ic a ld e h y d e s w ith Calibration Curves

2 -th io b a rb itu ric a c id .
Absorbance-time stopped-flow signals were recorded for
solutions containing different amounts of HMF at the experi­
Effect ofTBA concentration.—The influence of TBA con­ mental conditions indicated above (Figure 6 a), and a calibra­
centration from 0.15 x 10 2 to 3 x 1(T2 M) was studied. The tion curve was obtained by use of the reaction rate versus con­
reaction rate increased with the concentration of TBA. A TBA centration plot (Figure 6 b). The reaction was first order with
concentration of 3 x 10_2M was selected as optimum. A partial respect to HMF in the range 1.1-14.6 pg/mL. The correlation
order of 1 was found for the concentration range studied.
coefficient (r) was 0.9996. The slope of the calibration plot was
Effect o f pH .—The influence of pH was studied by using
0.025 mL/pg/min. The relative standard deviation (RSD) (con­
mixtures of hydrochloric acid and sodium chloride solutions. A
fidence level, 95%; n = 10) for 5.6 pg/mL was 2.4%. The de­
6 M chloride concentration was maintained constant, and the
tection limit was defined as the concentration of HMF giving
hydrogen ion concentration was varied from 0.5 to 6 M. The
rise to an initial rate equal to 3 times the standard deviation of
initial reaction rate increased with an increase in hydrogen ion
the initial rate (n = 11) on a sample containing 5.6 pg/mL of
concentration up to 1.5M (a Vi partial order with respect to
HMF (25). A value of 0.47 pg/mL was found. The sampling
[H+] was obtained) and remained practically constant above
rate was about 40/h. The reaction between furfuraldehyde and
1.5M (a 0 partial order was found). The results are plotted in
Figure 3. A 5M hydrochloric acid concentration was selected TBA follows the same kinetic behavior.
as optimum.
Kinetic Equation
Effect o f temperature.—The effect of temperature on the re­
action rate was examined in the 20-42^0 range for samples
The simplified kinetic equations can be applied on the fol­
containing between 1 to 14 pg/mL of HMF (Figure 4a). A tem­
lowing conditions:
perature of 40°C was chosen as sufficient for proper develop­
ment of the reaction. A plot of the logarithm of the rate constant
For [H+] > 1.5M:
against the inverse of the absolute temperature gives the value
of the activation energy (Figure 4b): 5.9 kcal/mol/K
¿[ h m f - t b a ] = r [HMF| [TBA]
(24.75 kJ/mol/K). A 3-dimensional isometric projection of the
at

F ig u r e 6. (a) S to p p e d -flo w s ig n a l f o r d iffe re n t c o n c e n tr a tio n s (p p m ) o f H M F. (b) C a lib r a tio n c u r v e f o r H M F

d e te rm in a tio n .
1260 E spinosa -M ansilla E t Al .: J ournal Of ADAC International V ol . 76, No. 6,1993

T a b le 1. R e s u lt s o f p r o p o s e d m e th o d f o r d e te rm in a tio n o f f u r a n ic a ld e h y d e s in f o o d a n d p h a rm a c e u tic a l s a m p le s b y
u s in g th e m e th o d o f s ta n d a rd a d d itio n s

Slope of
calibration curve, Regression
Sam ple Added3 Found3 Ree., % pg/mL/min coefficient

H oney (commercial) 0 4 6 .0
15.8 60 .5 92
31.6 78 .4 102 0.0 26 0 .9 9 8 0

O range juice (commercial) 0.0 4.2

14.1 17.3 93
28.2 31 .5 97 0.0 25 0 .9 9 9 0

G rape juice (commercial) 0.0 2.7 —

5.6 9.0 112
11.9 16.8 118 0 .0 3 0 0 .9 9 8 0

Sw eet red wine (M alaga) 0.0 875.0

595.0 1519.0 108
1192.0 2 1 2 5 .0 105 0.0 26 0 .9 9 9 7

Red wine (commercial) 0.0 2.0 —

5.9 8.5 110
11.9 14.7 106 0.0 28 0 .9 9 9 4

Apiroserum levulose (YBYS, intravenous) syrup 0.0 77 .2

11.9 88.0 91
23.8 105.0 101 0.0 30 0 .9 9 9 3

Dextrose syrup (B. Martin, oral, pediatric) 0.0 41 .4

28.2 66 .3 88
56.2 97.1 99 0.0 25 0 .9 9 9 5

Th e total content of furanic aldehydes (H M F plus furfuraldéhyde) is expressed as H M F concentration In pg/mL.

For 1.5M > [H+] > 0.5M: The proposed stopped-flow method for HMF and furfural­
déhyde was tested on various commercial food and pharma­
ceutical samples by using the external standard and standard
= K" [HMF] [TBA] [H+ ] ' /2
additions procedures for calibration. The total content of fura-

where K ' and K" are the conditional rate constants of the con­
T a b le 2. C o m p a r is o n o f n o rm a l a n d s ta n d a r d a d d it io n s
densation reaction.
c a lib r a tio n p r o c e d u r e s

Influence of Foreign Species Furanic

aldehydes found

Several compounds were tested as possible interferents. External Standard

Glucose, levulose, sorbitol, citric acid, ascorbic acid, oxalic Sam ple standard additions
acid, and tartaric acid were tolerated at an interferent:analyte
ratio of at least 1000:1. The study was performed with samples Honey (commercial) 4 6 .0 pg/g 4 5 .0 pg/g
containing HMF at 5.6 fig/mL. G rape juice (commercial) 2.7 pg/m L 2.6 pg/m L
O range juice (commercial) 4.2 pg/m L 3 .7 pg/m L
S w eet red wine (M álaga) 6 7 5 .0 pg/m L 8 5 0 .0 pg/m L
A pplications
Red wine (commercial) 2.0 pg/m L 2.0 pg/m L
Apiroserum levulose (YBYS,
The continuous FIA method can be applied when the sam­ intravenous) syrup 77.2 pg/m L 7 5 .0 pg/m L
ple background does not absorb, whereas the stopped-flow Dextrose syrup (B. Martin, oral,
method can be used also in the presence of colored and turbid pediatric) 4 1 .4 pg/m L 4 6 .0 pg/m L

background matrixes.
 Espinosa -M ansilla E t Al .: J ournal Of AOAC International V ol . 76, No. 6,1993 1261

J. Morales Bruque, Department of Physics of the University of

Extremadura, for the design and construction of the interface
used in this work.

R efe ren c es

(1 ) H o d g e , J .E . ( 1 9 5 3 ) 7 Agrie. Food. Chem. 1, 9 28

(2 ) L e e , H .S ., & N a g y , S . ( 1 9 8 8 ) J. Food Sci. 5 3 , 168

(3 ) C o d e x A lim e n ta riu s C o m m is s io n ( 1 9 6 9 ) Recommended

European Standard f o r Honey, C A C /R S -1 2 -1 9 6 9 J o in t
F A O A V H O F o o d S ta n d a r d P r o g r a m , R o m e , Ita ly

(4 ) M ija r e s , R .M ., P a r k , G .L ., N e ls e n , D .B ., & M c lv e r , R .C .
(1 9 8 6 )7 . Food Sci., 51, 843

(5 ) l e e , H .S ., R o u s e f f , R .L ., & N a g y , S . ( 1 9 8 6 ) J. Food Sci. 51,

1076

(6 ) E s p in o s a - M a n s illa , A ., S a lin a s , E, & B e r z a s N e v a d o , J .J .

(1 9 9 2 ) J. AOAC Int. 7 5 ,6 7 8 -6 8 4
F ig u re 7. S to p p e d -flo w s ig n a ls o b ta in e d f o r a p p li­
c a t io n s in re d s w e e t w in e .
(7 ) B e e m a n , C .P . ( 1 9 8 7 ) J.Assoc. Off. Anal. Chem. 70, 6 0 1 -6 0 2

(8 ) P o m p e i, C ., R o s s i, M ., & B a r o z z i, E . ( 1 9 8 6 ) J. Food Sci. 51,

8

nic aldehydes found (HMF plus furfuraldéhyde) was expressed (9 ) S a n d o v a l, J. A ., & H id a lg o , T. ( 1 9 6 7 ) Bol. INIA, 1 6 .4 , 4 5 4

as HMF concentration in pg/mL. In Table 1, the results of the (1 0 ) M o n t i l l a G ó m e z , J ., O l e a , F ., & G a r c í a V i l l a n o v a , R . ( 1 9 8 8 )

proposed method combined with the method of standard addi­ Quim. Anal. 7, 77
tions are summarized. The slopes of the different calibration (1 1 ) L y d o n , J ., D u k e , S . O . , & H e d i n , P .A . ( 1 9 8 7 ) 7 . Chromatogr.
curves are similar in all cases and also similar to the slopes 410, 4 7 ¿
obtained for the determination in water. This fact shows that (1 2 ) D a v id s o n , A .G ., & D a w o d u , T .O . ( 1 9 8 7 ) 7 Pharm. BiomedL
sample matrixes do not interfere in the kinetic proposed Anal. 5, 213
method. The results of using the external standard and the (1 3 ) D a v id s o n , A .G . ( 1 9 8 7 ) 7 Clin. Pharm. Therap. 1 2 , 11
standard additions procedures are similar, as shown in Table 2.
(1 4 ) D a v id s o n , A .G ., & D a w o d u , T .A . ( 1 9 8 8 ) 7 Pharm. Biomed
Figure 7 shows the stopped-flow signals obtained for different
Anal. 6 , 61
dilutions of a red sweet wine sample. Two injections were
(1 5 ) R o c h a , M . I . , H a c k m a n n , E . R . , M a g a l l a n e s , J .F . , & V e m e n g o ,
made for each diluted sample.
M .J . ( 1 9 8 6 ) Rev. Farm. Bioquim. Univ. A. Paulo 22, 77

(1 6 ) M a c h e r a s , R , K o p p a r is , M ., & T s a p r o u n is , C . ( 1 9 8 6 ) Int. J.
C o n clu sio n s
Pharm. 33, 125
(1 7 ) S a lin a s , F, E s p i n o s a - M a n s i l l a , A ., & B e r z a s N e v a d o , J .J .
The proposed stopped-flow method based on the reaction
(1 9 9 1 ) Fresenius Z Anal. Chem. 340, 250
with TBA allows the determination of furanic aldehydes in
(1 8 ) D e l l a M o n i c a , E . S . , G r a i g , J . C . , & C a l h o u n , M . J . ( 1 9 6 8 ) 7.
various commercial samples, without previous separation pro­
D a iry Sci. 51, 353
cedures, in the presence of colored (red wine, orange juice) and
turbid (honey, citric juice) background matrixes. The proposed (1 9 ) J e u r in g , H .J ., & K u p p e r s , F .J .E .M . ( 1 9 8 0 ) 7. Assoc. Off.
Anal. Chem. 6 3 , 1 2 1 5 -1 2 1 8
method is fast and simple.
Use of the proposed stopped-flow method is more advanta­ (2 0 ) D in s m o r e , H .L ., & N a g y , S . ( 1 9 7 4 ) 7 Assoc. Off. Anal.
Chem. 57, 3 3 2 -3 3 5
geous than use of the conventional continuous FIA system, be­
cause the turbidity or the colored background signal does not (2 1 ) W in k le r. O . ( 1 9 5 5 ) Z Lebensmitt-Untersuch. 102, 161

affect the analytical signal, which in this case is the initial rate (2 2 ) L in a r e s , P , L u q u e d e C a s tr o , M .D ., & V a lc á rc e l, M . (1 9 8 7 )

of the kinetic reaction and is given by the slope of the absor­ Microchem. 7 35, 120

bance-time curve (stopped-flow FIA signal). (2 3 ) S tillin g , R .A ., & B r o w n in g , B .L . ( 1 9 4 0 ) Ind. Eng. Chem. 12,
499

A ck n o w led g m en ts (2 4 ) A lo n s o , J ., B a r to li, J ., & B la n c o , M . ( 1 9 8 3 ) Quim. Anal. 1,

261

Financial support for this work from the DGICYT of Spain (2 5 ) G la s e r , J .A ., F o e r s t, D .L ., M c k e e , G .D ., Q u a v e , S .A ., &
(Project PB91-0856) is acknowledged. We are also grateful to B u d d e , W .L . ( 1 9 8 1 ) Environ. Sci. Technol. 15, 1426
1262 Lobinski Et Al .: J ournal O f AOAC International V ol . 76, No. 6,1993

FOOD CHEMICAL CONTAMINANTS

Spéciation Analysis o f O rganolead Com pounds in W in e by

C a p illa ry Gas C hrom atography/M icrow ave-Induced-P lasm a
A tom ic Em ission S pectrom etry

R yszard L obinski, J oanna Szpunar-L obinska, and F reddy C. A dams

University of Antwerp (U.I.A.), Universiteitsplein 1, Department of Chemistry, 2610-Wilrijk (Antwerpen), Belgium
P ierre-L ouis T eissedre *1 and J ean-C laude C abanis
Université de Montpellier I, Centre de Formation et de Recherche en Œnologie, 15, av. Charles Flahaut, 34060 Montpellier,
France

A m eth o d w a s d ev e lo p e d for th e s p é c ia tio n analy­ process and result from the release of lead from capsules used
s is of ionic o rg a n o le ad c o m p o u n d s in w ine. The during wine storage, from bronze faucets, from pumps and tub­
a n a ly te s w ere ex tra cted a s dieth y ld ith io carb am ate ing, and from enamelled containers or concrete (1,4).
c o m p le x e s into h e x a n e a n d p ropylated with a Grig­ Until now, the research has focused exclusively on the de­
n ard reag en t. T he derivatized e x tra c t w a s analyzed termination of the total amount of lead present in wine regard­
by capillary g a s chro m atography/m icrow ave-in- less of what species were present. The latter question, however,
d u c e d -p la sm a ato m ic e m issio n sp ectro m etry . T he may be of vital importance, because the harmful effects of or­
m eth o d of sta n d a rd a d d itio n s w a s u s e d for calibra­ ganolead compounds are considered much larger than those of
tion to c o rre c t for variable re co v erie s a n d signal e n ­ inorganic lead (5). Unlike Pb2+, organolead compounds are
h a n c e m e n ts. Red, ro sé , a n d w hite w in es from liposoluble and, therefore, more readily absorbed by humans
so u th e rn F ran ce w ere analyzed. Trim ethyllead w as than inorganic lead. Oiganolead compounds tend to concen­
th e u b iq u ito u s s p e c ie s. T he w in es m ad e from trate in the human brain (6 ) and are responsible for metabolic
g ra p e s grow n c lo s e to industrial z o n e s sh o w e d ele­ and neurophysical deficiencies in subjects exposed to these
v a te d c o n c e n tra tio n s of ethyllead sp e c ie s . T he c o n ­ compounds over a long period of time (7).
c e n tra tio n s of m ethyllead a n d ethyllead fo u n d in Tetraalkyllead compounds used as antiknock additives to
w in es w ere co m p ared with th e c o n c e n tra tio n s of or­ gasoline and their ionic degradation products are common pol­
g an o lea d fo u n d in rain w ater a n d plant s a p col­
lutants found in atmospheric dust and deposits. The pollution
lected in th e viticultural reg io n s. T he ratio of
has a global character ( 8 ) and is particularly intense in urban
m ethyllead to ethyllead in w in es greatly e x c e e d e d
environments and at locations close to highways and gas sta­
th e s a m e ratio fo u n d in a tm o sp h e ric d e p o sits.
tions. Organolead compounds may arise in wine due to the
natural formation of methyllead from inorganic lead present
during the fermentation process. Biomethylation of lead spe­
T he growing concern about human exposure to lead pro­
cies was tentatively identified in media containing high micro­
vided the impetus for studying this element in wine ( 1 ).
bial activity (9) and the biomethylation of other metals (Hg, Sn)
Both the quantity of lead and the species of lead present
and of metalloids (As, Ge, Se) is well-established, but the
in wine are of interest. An excess of ingested lead may present
a real health hazard affecting both the nervous system and the biomethylation of inorganic lead remains controversial ( 1 0 ,
biosynthesis of hemoglobin (2). In response to this health con­ 11). The long-term exposure of grapes to the atmosphere and
cern, regulations on the maximum permissible lead level in the large wine consumption in some countries make the pres­
wine have been issued. The value of 300 pg/L, recommended ence of organolead compounds in wine worthy of investiga­
by the Wine and Vines International Office, was accepted by tion.
the European Community (3). Speciation analysis of organolead is performed by using hy­
The lead in wine may be of atmospheric origin, resulting phenated techniques combining chromatographic separation
from the deposition of airborne particulate matter on grapes with element-selective and element-sensitive detection ( 1 2 ).
and the uptake by the grapevine of lead present in ground- Most workers have used gas chromatography/atomic absorp­
waters and in soil. Other sources are related to the production tion spectrometry (5, 12), but capillary gas chromatogra-
phy/microwave-induced-plasma atomic emission spectrome­
Received December 4, 1992. Accepted March 12, 1993. try (GC/MIP-AES) is more suitable due to increased
1 Com ité Interprofessionnel des Vins d’A.O.C. Côtes du Rhône et delà sensitivity, reduced sample size, and simplified sample prepa­
Valleé du Rhône, 2A, rue Henri Fabre, 84000 Avignon, France. ration (13, 14).
 L obinski Et A l .: J ournal O f AOAC I nternational V ol . 76, No. 6,1993 1263

Table 1. GC/AES operating param eters

Analysis of plant juice
Param eter Analysis of wine and rain water

IN JE C TO R P A R A M E T E R S

Injection volum e, p.L 1 25

Split ratio splitless splitless
Injection tem p., °C 40 5
1st heat-up rate, °C/s 12 2
1st retention tem p., °C 260 10
1st retention period , s 60 60
2nd heat-up rate, °C/s — 12
2 nd retention temp.,
°C — 260
2 nd retention period, s — 60

G C PARAM ETERS
Figure 1. Effect of ethanol concentration on the extrac­
Column head pressure 130 k P a o f He 130 kPa of He
tion recovery of organolead com pounds. Recovery
Oven program:
values are relative to the respective extraction yields
Initial temp., °C 45 45
from the Milli-Q purified water.
Tim e, min 1 2
Ram p rate, °C/min 20 20
Final tem p., °C 280 280 direct extraction or leaching are often affected by poor recov­
Purge valve program: ery and co-preconcentration of organic compounds.
O N , min — <0.9 The aim of this work was to develop a method for the spe-
OFF, min 0 -0 .9 0 .9 -2 .5 ciation analysis of organolead compounds in wine and to inves­
O N , min >0.9 >2.5 tigate the extent of contamination of this beverage by or­
ganolead compounds. The organolead species patterns and
IN TE R FA C E P A R A M E T E R S
organolead concentrations in various wines are compared with
Transfer line HP-1 column HP-1 column those in rainwater and plant sap.
Transfer line tem p., °C 280 280
Solvent vent-off program:
Experimental
O N , min <2 <3
OFF, min 2 -5 .8 3 -6 .8
ON, min >5.8
Apparatus
>6.8
Column-detector Propylated ionic organolead compounds were separated on
coupling column to cavity column to cavity
an HP-1 capillaiy column (25 m x 0.32 mm x 0.17 pm) using
AED PARAM ETERS
an HP Model 5890 Series II gas chromatograph (Hewlett-
Packard, Avondale, PA) equipped with a programmable tem­
Wavelength, nm 4 0 5 .7 8 3 40 5.78 3
perature injector (Gerstel, Miilheim a.d. Ruhr, Germany). A
H e m ake-up flow,
mL/min
Hewlett-Packard 5921A atomic emission detector was used.
300 300
S cavenger gases:
Injections were made either by means of an HP Model 7673A
H2 pressure, psi 90 90
autosampler (1 pL) or manually (25 pL) using a Hamilton
0 2 pressure, psi 20 20 701RN syringe equipped with a needle (0.64 mm o.d.) with a
Spectrom eter purge polished finish (Gerstel) to fit the septumless injection head.
flow, L/min of N 2 2 2 Smooth-finished glass vaporization tubes packed with a 2 cm
Cavity temp., °C 280 280 plug of Tenax (80-100 mesh) (Hewlett-Packard) were used.

Reagents and Materials

(a) Solvents and reagents.—Analytical grade (E. Merck,
Darmstadt, Germany), unless otherwise stated.
(b) n-Propylmagnesium chloride (PrMgCl).—2 M solu­
Relatively little work has been devoted to spéciation analy­ tion in diethyl ether (Aldrich Chemical Co., Milwaukee, WI).
sis of organolead in foodstuffs; studies until now have been (c) Trimethyl- and triethyllead chlorides.— (Alfa Ventrón,
restricted to seafood (15). The most probable reason for this Karlsruhe, Germany).
fact is the complexity of the matrix and insufficient sensitivity (d) Standard solutions.—Tri alkyl lead standard solutions
and selectivity of the analytical techniques used. Methods for (100 pg/mL) were prepared in ethanol. Dialkyllead stock solu­
digestion of food risk destruction of the original analytes, while tions were prepared by reduction of the trialkyllead solutions
1264 L o b in s k i E t A l .: Jo u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

(i) Glassware.—Cleaned with a common detergent, thor­

oughly rinsed with distilled water, soaked 12 h in 10% H N 0 3
and finally rinsed with deionized water just before use.
(j) Tenax (80-100 mesh).—Hewlett-Packard.

P ro ce d u re s

Analysis of Wine

A 60-80 mL sample of wine was placed in an extraction

vessel and diluted with water to a total volume of ca 120 mL.
Figure 2. Chromatogram of a white wine sample EDTA solution (2 mL) and 3 mL buffer were added to the ves­
recorded on a carbon-193 channel. sel. Thereafter, the pH of the solution was adjusted to 8.0-8.5
with concentrated ammonia. NaDDTC solution (1 mL) was
added and the solution was shaken 2 min with 2 mL hexane.
with iodine monochloride (16). Working solutions were pre­ Solid NaCl (ca 0.5 g) was added to facilitate phase separation.
pared daily by dilution of the stock solutions with 50% ethanol. After phase separation, 500 pL of the extract was transferred to
The standards were quantified and their purity was checked as a derivatization vessel, followed by the addition of 50 pL
described previously (13). PrMgCl solution. After 10 min, 10 mL 0.1 M H 2 S 0 4 was
(e) Sodium diethyldithiocarbamate (NaDDTC).—0.25 M added to the vessel, and the vessel was shaken 1 min. After
solution (1.406 g NaDDTC dissolved in 25 mL water) was pre­ separation of the phases, the hexane layer was pipetted off and
pared daily and purified by shaking with hexane. dried over ca 0.1 g anhydrous Na2 S 0 4, and a 1 pL aliquot of
(f) Disodium saltofEDTA.—0.1 M solution in water (h). dried organic phase was analyzed by GC/AES. The method of
(g) Citric acid/ammonia buffer (pH 8.5).—Prepared by standard additions was used for calibration. An amount of a
dissolving 20.6 g citric acid in 1 L water (h) and adjusting pH mixed standard solution of ionic organolead compounds corre­
with concentrated ammonia solution. sponding approximately to the content of the organolead com­
(h) Water.—Deionized and further purified in a Milli-Q ap­ pounds in the analyzed sample was added to another sample,
paratus (Millipore Corp., Bedford, MA). which was analyzed in parallel according to the same proce­
dure.

Analysis o f Rain Water and Plant Sap

A 100 mL sample was analyzed as described above, except

that 1 mL hexane was used for the extraction and no NaCl was
added. A 25 pL aliquot of the extract was analyzed. The cali­
bration was done using linear regression. The standards used
contained 0 ,1 ,2 , or 5 ng/L Pb as organolead compounds.

GC/AES Conditions

The operating parameters for the injector, GC oven, and

atomic emission detector are summarized in Table 1.

R e s u lt s a n d D is c u s s io n

Wine is a complex sample containing a variety of inorganic

and organic substances in an ethanol-aqueous solution. The
principal dissolved species are inorganic ions (ca 2 g/L), such
as K+, Na+, Ca2+; organic acids, such as tartaric, lactic, and cit­
ric; polyphenols, such as anthrocyanines and tanins; polyhy-
droxyalcohols, such as glycerol (ca 7 g/L); and various pro­
teins, aminoacids, and polysaccharides. White wines are less
abundant in dissolved substances than red wines, especially
with respect to polyphenols. The wines analyzed in this work
contained only small concentrations of sugars (less than 2 g/L)
Figure 3. (A) Chromatogram of procedure blank; (B) and 50-200 pg/L of inorganic lead. The grapes from which the
chromatogram of a white wine sample (80 mL; Me3Pb+ wines were produced had not been washed prior to processing.
48.0 ng/L, Et3Pb+ 3.6 ng/L, Et2Pb2+ 6.3 ng/L). Peaks: 1 = The analytical procedure developed earlier for the specia-
Me3Pb+, 2 = Et3Pb+, 3 = Et2Pb2+. tion analysis of organolead compounds in water (13) was
 L o b in s k i E t A l .: J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3 1265

nol concentration. Because the polarity of all organolead

chelates is similar, this decrease is attributed to a hampering
effect of ethanol on the formation of the Me3 Pb+-DDTC com­
plex.
Effect o f foreign ions.—Tartrate, lactate, and citrate at con­
centrations not exceeding 0.1 M were found not to hamper the
extraction of ionic organolead compounds, while the transfer
of inorganic lead into hexane was diminished. This last phen-
momenon is desirable, because very large quantities of coex­
tracted Pb2+ might negatively affect the derivatization and
might cause carry-over effects in the injector and the formation
of deposits in the discharge tube. EDTA was used to increase
the retention of Pb2+ in the aqueous phase. It was observed,
however, that its effect was limited on account of the presence
of other metal ions, such as Ca2+binding to EDTA. At an EDTA
concentration of 0.005 M, the inorganic lead signal from a
wine sample was identical to that from a Milli-Q water sample
analyzed according to the procedure.
Co extraction o f organic substances.—The degree of coex­
traction of organic substances is dependent on the pH of the
sample and the polarity of the extraction solvent. The least po­
lar solvent enabling quantitative extraction of organolead com­
pounds, hexane, was used. The organic phase after extraction
turns yellow due to the presence of copper. The alkaline extrac­
tion pH favors the dissociation of acids and phenols in the
aqueous phase; these are then not coextracted. Nevertheless,
some organic compounds pass into the organic phase. There­
fore, further preconcentration of a wine extract by evaporation
F igu re 4. (A ) C h ro m a to g ra m o f a rainw ater s a m p le (R W affects the baseline stability and is inadvisable unless the inter­
III, 50 m L) from a viticultural area; (B ) c h ro m a to g ra m o f a fering substances are removed by a cleanup procedure. The co­
gra p e vin e s a p s a m p le ( G S 1,100 m L). P e a k s: 1 = extraction of organic substances makes the use of a lead-selec­
M e 3 P b +, 2 = M e 2 P b 2+, 3 = E t 3 P b +, 4 = Et2P b 2+. tive detector for analysis mandatory. Otherwise, the carbon
background prevents the detection of the analytes (Figure 2).
adapted for wine analysis. Recovery of the analytes, however,
required a more detailed investigation. The recovery studies Characteristics o f the Method
included verification of the extraction yield from ethanolic me­
dia; examination of the effect of tartrate, lactate, and citrate on GC/AES conditions. The operating conditions for the detec­
tor were set as optimized earlier (13,14). The operating condi­
the efficiency of extraction of organolead; and consideration of
tions for the injector and GC were adapted from earlier works
coextraction of matrix organic compounds and Pb2+.
(13, 14) for wine and sap, respectively. Some modifications
Recovery Studies with respect to the temperature and valve switching time pro-

Effect o f ethanol.—The introduction of ethanol, miscible T able 2. D etection lim its in the a n a ly s e s o f different
both with water and with non-polar solvents, into the extraction s a m p le s
system may affect the recovery of organolead in 2 ways: (7) by
White Red Plant Rain MQ- Absolute
affecting the distribution coefficient of organolead chelates be­ Com ­ w n e, wine, sap, water, water, detection
tween the aqueous and the organic phase or (2 ) by changing the pound ng/La ng/Lb ng/Lc ng/Lc ng/Lc limit, pgd

initial sample-to-hexane volume ratio. The effect was studied

by extracting organolead compounds from aqueous-ethanol M e 3Pb+ -.2 1.8 0.03 0.03 0.0 25 0.06
M e 2Pb2+ -.6 2.4 0.0 4 0.04 0.0 3 0.07
solutions containing the same initial amounts of ionic or­
EtgPb+ ‘ .8 2.7 0.0 4 0.04 0 .0 3 0.07
ganolead compounds with 1 mL hexane. The solutions were
Et2Pb2+ 2.1 3.2 0.05 0.05 0.0 35 0.09
100 mL mixtures of water and ethanol with varying propor­
tions of ethanol. The volume of the organic phase remained 8 Based on an 80 m L sample, 2 mL extraction.
unchanged. The influence of ethanol concentration on the re­ b Based on an 80 m L sample, 3 mL extraction.
c Based on a 100 mL sample, 1 m L extraction, after 25-fold In-llner
covery of analytes is shown in Figure 1. No effect is noticed for pre concentration.
organolead species except for trimethyllead, for which a de­ d For injection of pure standards, in in-liner venting mode.
crease in extraction efficiency is observed with increasing etha­
1266 L o b in s k i E t A l .: Jo u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

Table 3. Recovery and precision of the method developed (5 determinations)

Found, ng/L Recovery, % RSD, %

Com pound Added, ng/L W hite wine Red wine W hite wine Red wine W hite wine Red wine

M e 3Pb+ 39 .9 30.2 __ __ 7.5 9.8

25 61.1 49.8 84.8 78.4 4 .4 8.3
50 84.1 72.8 88.4 85.2 3.9 6.7

M e2Pb2+ — — — — — — —
5 4.82 4.14 96.4 82.8 15.1 19.0
10 9.13 9.18 91.3 91.8 11.6 14.8

Et3Pb+ — 10.4 8.05 — — 13.6 17.5

5 15.4 12.7 100 93.0 7.5 11.0
10 19.9 16.7 95.0 86.5 7.4 9.1

Et2Pb2+ — 5.43 5.1 4 - — 25.5 28 .0

5 9.42 9.41 79.8 85.4 12.4 15.2

10 14.6 14.4 91.7 92.6 9.9 11.2

grams were necessary due to the significant complexity of the is not seen on the chromatograms, because the derivatized Pb2+
matrix. In the wine analyses, the Tenax packing was used to is vented off the discharge tube. The baseline is generally more
retain coextractives with high boiling points to reduce contami­ stable for the white wine extracts, which are less abundant in
nation of the capillary column, while in the analyses of plant organic matter than the red wine extracts. For samples of rain
sap and rain water its role was different. Tenax served then to water and grapevine sap, a stable baseline was obtained even
retain the analytes in the liner, while the extraction solvent was after the 25-fold preconcentration of the derivatized extract in
vented off (14). the injection liner. Typical chromatograms obtained for rain
Chromatograms.—Figure 3 shows typical chromatograms water and plant sap samples are shown in Figure 4. The unla­
obtained for a blank and a wine sample. The procedure is belled peaks are due either to some impurities of the Grignard
blank-free with respect to organolead compounds but a contri­ reagent or to artifacts aroused in the course of the derivatization
bution of inorganic lead is present in the blank. Inorganic lead
process (14).
Detection limits.—The detection limits, calculated as 3
times the standard deviation due to noise, that were obtained in
Table 4. Concentrations of lead in organolead this work for different compounds in different matrixes are
compounds from wine samples (individual results are summarized in Table 2. Poorer detection limits obtained for
given)
ethyl than for methyl species result from signal discrimination,
M e3Pb+, ng/L Et3Pb+, ng/L Et2Pb2+, ng/L which is more pronounced in splitless injections than in large
Sam ple No. (of Pb) (of Pb) (of Pb)
volume injections (14). Phase separation difficulties after ex­
traction of red wine samples made it necessary to use a larger
W hite wines
volume of hexane. This modification of the procedure resulted
W1 47.75; 46.0; 43 .7 <1.8 <2.1
in poorer detection characteristics of the method.
W2 14.4; 15.6 <1.8 <2.1
Precision and accuracy.—A white wine sample and a red
W3 24.6; 26.0 9.0; 10.1 <2.1
wine sample were analyzed (in qumtuplicate) unspiked and
W4 42.6; 27 .8 <2 <2.1
W5 39.8; 42.1 10.0; 10.8 5.0; 5.6
spiked with ionic organolead at 2 different concentration lev­
els. The results are given in Table 3. Other samples (Table 4)
Rosé wines

R 01 46.9; 47.2 5.2; 5.4 <3.2 Table 5. Concentrations of organolead in rain water
R 02 26.4; 26 .7 <2.7 <3.2 (RW) and grapevine sap (GS)
R 03 31.5; 29.5 <2 <3.2
M e3Pb+, M e2Pb2+, Et3Pb+, ng/L Et2Pb2+,
Sam ple ng/L (as Pb) ng/L (as Pb) (a s P b ) ng/L (as Pb)
Red wines

R1 34.4; 35.8 <2.7 <3.2 RW I 2.1 0.1 6 1.8 0.9 6

R2 66.5; 68.1 11.7; 10.2 <3.2 R W II 3.4 <0.04 0.82 0.3 3
R3 65.4 24.3 <3.2 R W III 2.3 1.2 2.7 3.1
R4 112; 102 20 .0; 22.1 11.2; 12.0 RW IV 1.6 <0.04 0.14 0.2 8
R5 8.1; 13.7 <2.7 <3.2 GS I 1.4 0.0 8 2.2 0.81
R6 13.0; 22.1 7.5; 9.1 <3.2 G S II 1.3 0.11 1.8 0.76
 L o b in s k i E t A l .: Jo u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3 1267

were analyzed in duplicate by the method of standard addition Acknowledgments

(spike at one level only). Variable recoveries from 70 to 130%
were observed. The signal depression was the largest for We thank S. Bmn and M.T. Cabanis for the comments on
Me3 Pb+ due mostly to the effect of ethanol (Figure 1). Signal the manuscript. Joanna Szpunar-Lobinska acknowledges a re­
enhancement may result from evaporation of the extract during search grant by the Belgian Science Policy Office. The work
lengthy phase separation or the presence in the extract of un­ was financially supported by the EC program “Lead in wine”
identified compounds facilitating the release of the species (P.V1NS/FR/1512) and the IGBP “Global Change”
from the injector. No organolead compounds were found in the (GC/06/006) program.
second extraction. Recoveries in the analysis of sap and rain
water were comparable with those from MQ-water and ex­ References
ceeded 90%. Because wines may have different compositions,
the method of standard additions was applied to all wine sam­ (1 ) L e g r a n d , V , M e d i n a , B ., G r e n o n , J .P ., & P a u c h e t , M . ( 1 9 9 1 )
ples, whereas the analyses of sap and rain water calibration was Analusis, 19, M 3 9 -M 4 3
accomplished with linear regression. (2 ) B o u d e n e , C . Toxicité des M é ta u x (1 9 8 6 ) i n Toxicologie et
Sécurité des alim ents R . D e r a c h e (E id .), L a v o i s i e r , F r a n c e
A n a ly t ic a l R e s u l t s (3 ) In te rn a tio n a l M ethods o f Wine Analysis, Annexe C (M a x i­
mum Levels o f D iffere n t Elem ents in W ine) ( 1 9 9 0 ) P a r i s ,
F ran ce
The results of the analysis of wine samples are summarized
in Table 4. Trimethyllead was found in all the wine samples (4 ) T e i s s e d r e , P L . , C a b a n i s , M . T . , D a u m a s , F ., & C a b a n i s , J . C .
(1 9 9 3 ) Revue F ran çais e d ’Œ no log ie, i n p r e s s
analyzed, at levels dependent on the origin of the grapes used
in the wine and varying from 8 .1 to 112 ng/L. Ethyllead com­ (5 ) V a n C l e u v e n b e r g e n , R ., & A d a m s , F .C . ( 1 9 9 0 ) i n H an dbo ok
o f E nvironm ental Chem istry, V o l. 3 E , O . H u t z i n g e r ( E d .) ,
pounds were generally not found in wines produced in rural
S p rin g e r, B e r lin , G e rm a n y
areas away from the highway nor in vintages of the prior year
(6 ) N i e l s e n . T ., J e n s e n , K . A . , & G r a n d j e a n , P . ( 1 9 7 8 ) N atu re
from industrial zones. In older vintages, however, ethyllead
(London), 274, 6 0 2 - 6 0 3
concentrations of approximately 10-20 ng/L were regularly
found. Wines made from grapes grown close to large industri­ (7 ) B io lo g ic al Effects o f O rg an o le ad Compounds ( 1 9 8 4 ) P.
G ra n d je a n & E .C . G ra n d je a n (E d s ), C R C P re s s , B o c a R a to n ,
alized centers showed an elevated content of triethyllead.
FL
Me2 Pb2+ was not detected in any of the wine samples analyzed.
(8 ) L o b i n s k i , R „ & A d a m s , F .C . (m i) Analusis, 2 0 , M 2 8 - M 3 1
Table 5 shows results of the analysis of rain water and
(9 ) R i d l e y , W .P ., D i z i k e s , L . J . , & W o o d , J . M . ( 1 9 7 7 ) Science,
grapevine sap samples from the viticultural zones. No clear
197, 3 2 9 - 3 3 2
dominance of trimethyllead is seen. Possible explanations for
(1 0 ) S c h m i d t , U ., & H u b e r , E . ( 1 9 7 6 ) N atu re (London), 2 5 9 , 1 5 7 —
the different organolead species patterns of wines are either the
158
relative increase in Me3 Pb+ (e.g., via biomethylation) or the
(1 1 ) R e i s i n g e r , K ., S t o e p p l e r , M . , & N ü r n b e r g , H .W . ( 1 9 8 1 ) N a ­
relative decrease of the ethyllead species concentration as a
ture (London), 291, 2 2 8 - 2 3 0
result, perhaps of faster degradation.
(1 2 ) R a d o j e v i c , M . ( 1 9 8 9 ) i n E nviro nm ental A nalysis Using C h ro ­
m atography In terfac ed w ith A to m ic Spectroscopy, R .M .
C o n c lu s io n H a rris o n & S . R a p s o m a n ik is (E d s), H o rw o o d , C h ic h e s te r,
UK

This study developed a method for the speciation analysis (1 3 ) L o b i n s k i , R ., & A d a m s , F . C . ( 1 9 9 2 ) A n a l. Chim . A cta 2 6 2 ,

of organolead in wine and indicated the extent of contamina­ 2 8 5 -2 9 7

tion of this beverage by organic forms of lead. Although the (1 4 ) L o b i n s k i , R . , & A d a m s , F .C . ( 1 9 9 2 ) 7 . A n a l. At. Spectrom. 7,
source of this contamination is probably airborne, the greater 9 8 7 -9 9 2

amount of trimethyllead compared with ethyllead species in (1 5 ) L o b i n s k i , R . , & A d a m s , F .C . ( 1 9 9 3 ) i n A nalysis o f C o n tam i­

wine may be indicative of biomethylation. Further investiga­ nants in E dib le A qu atic Resources, V .I I , S . R a y & J .W .

tion of the organolead species content in soils, grapes, sap, K ic e n iu k (E d s ), V C H P u b lis h e rs , W e in h e im , G e rm a n y

must, and wine is necessary. (1 6 ) H a n c o c k , S „ & S l a t e r , A . ( 1 9 7 5 ) A nalyst 100, 4 2 2 - 4 2 9

1268 M c N e a l & H o l l m e l d : Jo u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

FOOD CHEMICAL CONTAMINANTS

D e t e r m i n a t i o n o f V o l a t i l e C h e m ic a ls R e le a s e d f r o m

M ic r o w a v e - H e a t - S u s c e p t o r F o o d P a c k a g in g

T imothy P. M cNeal and H enry C. H ollifield

U.S. Food and Drug Administration, Division of Food and Chemical Technology, Washington, DC 20204

M i c r o w a v e h e a t s u s c e p t o r s th a t c o n v e r t e le c tr o ­ F oods such as pizza, waffles, meat sandwiches, and french

m a g n e t i c e n e r g y t o h e a t a tta in h ig h t e m p e r a t u r e s fries are now marketed as convenience foods that can be
th a t m a k e it p o s s i b l e to c o o k s o m e f o o d s to g o ld e n prepared quickly and easily in microwave ovens by using
c r i s p n e s s in a m ic r o w a v e o v e n . S u s c e p t o r s a re
susceptor packaging. When placed in a microwave field, the
t y p ic a lly p a c k a g e d w ith f o o d s in t e n d e d f o r m ic r o -
susceptor acts much like a frying pan, instantly achieving very
w a v e u s e , e.g., w a ffle s , p iz z a s , a n d f r e n c h frie s.
high temperatures that enable it to brown and crisp foods in
moments. Most of the microwave susceptors in use today are
T h e h ig h t e m p e r a t u r e s >3 02 ° F u s e d t o c o o k s o m e
laminated structures of paper, adhesives, and a metalized poly­
f o o d s r e le a s e t r a c e le v e ls o f v o la t ile c h e m i c a ls
ethylene terephthalate (PET) film. Although it may seem sur­
f r o m m e t a liz e d p o ly e s t e r film , a d h e s iv e , a n d p a p e r
prising, susceptors made of these simple materials are capable
p a c k a g i n g m a te r ia ls ; t h e s e v o la t ile c h e m i c a l s m a y
of achieving surface temperatures greater than 500T when
b e a b s o r b e d b y th e fo o d . W e s im u la t e d m ic r o w a v e
cooking foods that have a low water content ( 1 ).
s u s c e p t o r c o o k i n g c o n d i t i o n s a n d d e v e lo p e d p r o t o ­
Two main susceptor constructions are currently in use. The
c o l s b y u s i n g h e a d s p a c e c o n c e n t r a t io n c a p illa r y
first is a bilaminate, in which the food contact surface is a
g a s c h r o m a t o g r a p h y a n d m a s s s p e c t r o m e t r y to
metalized PET film glued directly to a paperboard support.
id e n tify v o la t ile c h e m i c a l s r e le a s e d f r o m h e a te d This type is generally used with foods such as fish and pizza.
s u s c e p t o r s . W e p u r c h a s e d a lim ite d , c r o s s - s e c ­ The second type is a trilaminate used primarily for popcorn. In
t io n a l s a m p l e o f lo c a l re tail m ic r o w a v e f o o d p r o d ­ this type, the metalized PET film is in the wall of a bag between
u c t s p a c k a g e d w ith s u s c e p t o r s a n d u s e d o u r 2 layers of paper—an inner layer of greaseproof paper serving

p r o t o c o l t o a n a ly z e 1 0 d iffe re n t s u s c e p t o r p r o d ­ as the food contact surface and an outside printed layer.

u c t s . A l t h o u g h m o r e t h a n 1 4 0 u n iq u e c h r o m a ­ Previous studies showed that both constructions are capable
t o g r a p h i c p e a k s w e r e ta b u la te d , o n l y 4 4 v o la t ile of reaching cooking temperatures that cause their constituent
c h e m i c a ls w e r e id e n tifie d , i n c lu d in g 1 ,1 ,1 -tric h lo - parts to physically break down and to release numerous com­
r o e th a n e , b e n z e n e , a n d 2 -(2 -b u t o x y e th o x y )e t h a n o l, ponents and degradation products (2-6). The PET film con­
w h ic h w e r e d e r iv e d p r im a r ily fr o m th e p a p e r a n d tains oligomers, residual starting materials, and reaction by­
a d h e s iv e s u s c e p t o r c o m p o n e n ts. N o o n e s u s c e p ­ products. Many adhesive products, each containing a unique
to r c o n t a in e d a il th e id e n tifie d s u b s t a n c e s . T h e combination of chemicals, can be used to hold the PET film in
s t a n d a r d a d d i t i o n s t e c h n iq u e w a s th e p re fe rre d place. Finally, the paper or paperboard is a complex mixture of
m e t h o d f o r q u a n tita tio n . T r ic h lo r o e t h a n e a n d 2 -(2 - cellulose, wood-based residuals, and by-products of the paper­
b u t o x y e t h o x y ) e t h a n o l w e r e p r e s e n t in s e v e r a l p r o d ­ making process. Added to the raw paper are chemicals and ma­
u c t s a t 7 5 -1 2 2 p g /in .2 o f s u s c e p t o r s u r f a c e a re a . terials that promote high temperature cohesion, reduce water
B e n z e n e w a s f o u n d in 3 s u s c e p t o r s a t < 0 .2 2 |ig/in.2
and grease absorption, improve appearance, and even enhance
le v e ls . E x a m in a t io n in d ic a t e s th a t a d h e s i v e s u s e d
paper printability. When heated above 300°F, many of the
in m o r e re c e n t s u s c e p t o r p r o d u c t s w e r e r e fo r m u ­
chemical entities in the PET film, adhesive, or paper may de­
grade to other chemical forms. At these high temperatures, nei­
la te d to r e m o v e e v e n t h i s t r a c e le v e l o f b e n z e n e .
ther the softened PET film nor the greaseproof paper is an ef­
fective barrier against migration of most of the chemicals that
may be present in the hot microwave susceptor. Food in contact
with or surrounded by these hot susceptor surfaces may adsorb
Received July 10, 1992. Accepted January 26, 1993. many of these chemicals.
Part of this work was presented at the 104th AOAC Annual
To obtain information about the chemical makeup of micro-
International Meeting, September 9-13, 1990, at New Orleans, LA.
Throughout this paper, concentration units of pg/in.2 and temperatures
wave susceptors, the potential thermal decomposition prod­
in °F are used to be consistent with units used in food packaging regulations ucts, and data on their migration to food products, the U.S.
in the C o d e o f F e d e r a l R e g u la tio n s . Food and Drug Administration (FDA) published an Advance
 M c N e a l & H o l u f ie l d : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3 1269

Notice of Proposed Rulemaking (ANPR) (7). In response to the (e) Automated headspace gas chromatograph.—Perkin-
ANPR, the American Society for Testing and Materials Elmer HS-100 automated headspace sampler coupled to Per-
(ASTM), the National Food Processors Association in con­ kin-Elmer Sigma 2000 gas chromatograph with flame ioniza­
junction with the Society of Plastics Industries (NFPA-SPI), tion detection (FID) system. HS-100 operating conditions:
and individual manufacturers developed test protocols for the headspace program, constant mode. Times (min): vial equili­
determination of volatiles released from microwave suscep­ bration, 30; injection, 0.05; cycle, 60. Temperatures (°C):
tors, and each submitted data to FDA (2-4,8-10). equilibration, 90; transfer line, 120. Vial head pressure: 45 psi
Although these efforts provided some useful information to (helium). GC parameters: capdlary column, 30 m x 0.25 mm
FDA, the test conditions were not consistent, and not enough x 1.05 (im film of bonded Restek Rtx-5 (5% diphenyl-95%di-
replicate analyses were performed. Lack of a consistent test methyl polysiloxane) with open split injection; ca 15:1 split
protocol by the submitters made it difficult to draw meaningful ratio. Helium carrier gas: column head pressure 23 psi, linear
conclusions about different susceptor products. velocity 30 cm/s at 4073. FID gases (mL/min): hydrogen 20,
Meanwhile, FDA’s Indirect Additives Laboratory devel­ ultrahigh purity air 200. Detector temperature, 22573. Oven
oped general test protocols for both volatile and nonvolatile temperature program: 8 min at 073,47min to 16073, hold 8.5
chemicals produced during simulated use of the typical suscep­ min. Total chromatographic mn time, 37 min.
tor product. The protocols are appropriate for comparing simi­ (f) Data acquisition.—Spectra-Physics LABNET chroma­
lar susceptor products under standardized conditions, inde­ tographic data system; Model 4270 computing integrator with
pendent of potential food applications. Protocols for data capture feature and Epson Equity I+ computer.
nonvolatile migrants from susceptors were discussed in pre­ (g) Microwave oven.—Microwave oven rated at 700 watts
vious articles (11-19). This paper describes the protocol used total power output. Calibrate weekly, using ASTM procedure
by FDA for the determination of volatile chemicals generated F-1317-90 (20). Repeat calibration weekly to monitor any
from susceptor products under simulated microwave use con­ change in total output. When variability of the output wattage
ditions. exceeds a relative standard deviation (RSD) of 10%, use an­
other oven that meets the calibration requirement. Reproduci­
E x p e r im e n t a l bility observed with ASTM procedure F-1317-90 was 7.5% for
7 different microwave ovens, all rated at 700 watts. Precondi­
tion magnetron by heating 2 L water in a 3 L Pyrex beaker on
Note: As a matter of good laboratory practice, the analyst
high power for 1 0 min and allow to cool 1 0 min before testing.
should review the toxicity of all chemicals used in this method
and exercise necessary safety measures when handling them. Reagents

Apparatus (a) Solvents.—Deionized water. Demonstrate absence of

materials coeluting at retention times of analytes under condi­
(a) Headspace vials.—No. 899701, Teflon-faced silicon tions of headspace GC/MSD analysis.
septa, aluminum caps, and spacer stars No. 887803 (Shamrock (b) Analytes.—ACS grade or better.
Glass Co., PO Box 6 8 6 , Seaford, DE 19973). (c) Standard solutions.—All analytes were liquids at room
(b) Syringes.— 10, 50, and 100 jiL Hamilton 700 series; temperature. Concentrations are expressed as weight/volume.
500 pL Hamilton 1000 series; and 5 mL Pressure-LOK series Prepare solutions daily. (1) Internal standard solution.—245
A-2 gas syringes (Precision Sampling Corp., Baton Rouge, LA pg/mL. Transfer 6 pL 4-heptanone into taied headspace vial
70815). containing 20 mL water. Quickly cap, seal, and reweigh vial.
(c) Forced air oven.— Stable-Therm constant temperature Calculate concentration of internal standard. (2) Mixed fortifi­
cabinet (Blue M Electric, Blue Island, EL 60406). Oven stabi­ cation standard.—Transfer 100 mL water to 125 mL HYPO-
lized at 90°C. VIAL™. Cap, seal, and weigh vial. With microliter syringe,
(d) Gas chromatograph/mass spectrometer.— Hewlett- transfer selected analyte(s) by piercing vial septum and intro­
Packard 5890A gas chromatograph with subambient oven ac­ ducing analyte to water (e.g., 250 pL2-(2-butoxyethoxy)etha-
cessory and split/splitless injector connected (capillary direct nol, ca 2400 pg/mL, or 60 pL 2-methyl-1-propanol, ca 400
interface) to Hewlett-Packard 5970B mass selective detection pg/mL) and reweigh vial after each fortification. Repeat for all
(MSD) system. Gas chromatographic (GC) parameters: capil­ analytes and include 30 pL 4-heptanone as an internal standard.
lary column, 50 m x 0.32 mm x 0.5 pm film of bonded HP-5 Calculate individual analyte concentrations.
(5% diphenyl-95 % dimethyl polysiloxane), operated in the Preparation o f Susceptor Products for Analysis
splitless mode; split vent, open 0.75 min; carrier flow, ca 1.0
mL helium/min. Column head pressure 10 psi, linear velocity Preparation o f susceptor test strips.—Using a paper cutter,
35 cm/s at 50°C. Temperatures (°C): injector, 225; interface, cut and discard susceptor edges. If the susceptor was packaged
280. Oven temperature program: 5 min at 0°C, 87min to in contact with food, remove the larger portion of the edge. Cut
15073, hold 4.25 min. Total mn time, 28 min. MSD parame­ rectangular 1 in. 2 (ca 1.05 x 6.15 cm) susceptor pieces for
ters: calibration, autotuned with perfluorotributylamine, opti­ analysis, preferably by using a pattern. Buff surfaces of strips
mized at mJz 502. Scan time, 4—28 min; scan range, m/z 35- with laboratory tissues and store in fod-covered glass beaker
2 0 0 ; scan rate, 2 . 6 scans/s. until ready for analysis.
1270 M c N e a l & H o l l if ie l d : Jo u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

Fortification o f susceptor test strips fo r qualitative and fortified and unfortified susceptor test strips to determine
quantitative analysis.—For quantitative analysis using the amount of each volatile chemical released from susceptor.
method of standard additions, place 1 susceptor test strip into When the standard for a given analyte is not available, estimate
each of 5 headspace vials. Transfer 10 (iL of the internal stand­ concentration based on the use of 4-heptanone internal stand­
ard solution (c)(7) to the paper surface of each strip in 2 of the ard. Estimate analyte levels by multiplying the amount of
vials (vials 1 and 2). Cap and seal the vials. Transfer 2, 5, and added 4-heptanone (2.45 jig) by the ratio of the peak area re­
10 pL of mixed fortification standard to vials 3, 4, and 5, re­ sponses of the unknown and the internal standard. Report all
spectively; add internal standard solution to vials 3 and 4 to data in micrograms of chemical found in 1 in . 2 of susceptor
make a total volume of 10 uL in each vial. Cap and seal the surface (pg/in.2).
vials. The final concentration of the internal standard will be
2.45 pg/in . 2 in all vials and vials 1 through 5 will contain 0,0, Results and Discussion
2X, 5X, and 10X of the added analytes. For qualitative
GC/MSD analysis, prepare a susceptor strip in a vial with 10
This analytical protocol was developed to characterize the
pL of the internal standard solution.
gaseous components released by microwave susceptors during
Microwave heating o f susceptor strips.—Precondition mi­
heating. Our objective throughout has been to evaluate the per­
crowave oven as indicated above (.Apparatus (g)). Place 1 of
formance of microwave susceptor products under simulated
the test vials in the center of microwave oven floor with the
consumer use conditions. For this reason, microwave ovens in­
active surface of the susceptor facing up. Place a 400 mL Pyrex
tended for home use have been used when appropriate, instead
beaker with 250 mL water and boiling chips next to the vial.
of microwave guides or other specialized ovens. Development
Microwave on high for 5 min. For quantitative analyses, re­
of the protocol involved 2 principal tasks. The first was to de­
move the vial from the oven, let oven cool for 1 0 min, and
vise a microwave heating test that would realistically simulate
microwave the next test vial. Always position the vial and the
susceptor cooking conditions and contain any escaping gases
beaker the same way for all microwave heating tests. For quali­
for analysis. This task involved trying to achieve susceptor sur­
tative analysis, always precondition oven before microwaving
face temperatures in a microwave environment appropriate for
the test vial.
the cooking of foods but not overly abusive to the susceptor
Qualitative Analysis material. The second task was to select or develop an analytical
method to separate susceptor-generated gases into individual
Manual headspace sampling and GC/MSD analysis.— Use chemicals for identification and measurement.
gloves when handling hot syringes and vials. After microwave In the microwave oven, food/susceptor interface surface
heating of the test vial, place vial in forced-air oven set at 90°C. temperatures rely upon the interdependent parameters of the
With preheated 5 mL Pressure-LOK gas syringe, pierce the vial food, oven, susceptor, and heating times in ways that hamper
septum, draw up a full syringe of headspace vapor, return vol­ simple analysis. Lentz and Crossett (1) provide an excellent
ume into vial, and redraw 2 mL. Allow headspace vapors to discussion of these interdependencies in their paper,
equilibrate in vial and in syringe in a closed oven for at least 2 “Food/Susceptor Interface Temperature During Microwave
min. Close syringe valve, remove syringe, quickly inject 2 mL Heating.” They measured the maximum temperatures
of the headspace gases onto the GC column, and initiate data achieved by a specific susceptor for several foods and cooking
acquisition. times, and compared them to the maximum temperature
Identification o f susceptor-generated volatiles.—Evaluate achieved by the susceptor without a food load. For example,
individual scans of peaks of the total ion chromatogram. Iden­ the authors observed a surface temperature of 600°F during a
tify any volatile chemicals. When possible, use retention data cooking time of 100 s when there was no food load. In this case,
and mass spectra of standards to identify the peaks. When no the susceptor was severely blackened over 25 % of its area, and
standard is available, identify the unknowns by using prob­ the remainder of the surface was far darker than susceptors
ability-based matching with spectra from a mass spectral refer­ used under normal cooking conditions. The temperature of an­
ence library. other susceptor reached 468°F when popcorn was cooked for
Quantitative Analysis 150 s and increased to 536°F when the cooking time was ex­
tended to 220 s. The temperatures reached 432°F and 433°F
Automated headspace sampling and GC/FID analysis.— when fish filets and pizza, respectively, were cooked for 290 s.
Load vial rack of the HS-100 headspace sampler with vials for None of the foods burned under these conditions. These find­
each set of fortified and unfortified susceptor sheets being ana­ ings by Lentz and Crossett are in agreement with those of Cas­
lyzed. When necessary, to prevent carryover from previous in­ tle et al. (12) as well as our own results. Collectively, these data
jections, place a headspace vial containing 10 pL of the mixed reflect the difficulty of prescribing “standardized” test proce­
fortification standard without susceptor in the sampler between dures for evaluating every susceptor and its various uses. Not
each set of vials analyzed. Record analysis sequence and initi­ only do cooking conditions vary for different foods, they may
ate automated data acquisition. vary for different amounts of the same foods.
Quantitation o f susceptor volatiles.—Record the chromato­ After numerous experiments, the test protocol described in
grams and the corresponding peak areas for each headspace the Experimental section was developed; it gives reproducible
analysis. Use the linear regression plot of analyte response for results when replicate portions of the same susceptor material
 M c N e a l & H o l l ih e l d : Jo u r n a l O? A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3 1271

are heated. To ensure that each test portion yields the same mode, a gain in column capacity and chromatographic resolu­
volatile components in the same relative amounts, strips of the tion was obtained.
susceptor are placed in septum-sealed vials and microwaved When possible, we used the method of standard additions
for 5 min in a 700 watt microwave oven that also contains for quantitation. When analytical standards were not available,
250 mL water. The temperature achieved by a typical pizza- we obtained quantitative estimates based on the response of
type susceptor under these conditions is about 347°F. The ap­ 4-heptanone. The method of standard additions is the best
pearance of the susceptor after heating is typical of other means to overcome matrix interactions caused by the presence
susceptors used to cook pizza and fish filets; the paperboard has of water and cellulose in the susceptor.
a slight beige color and possesses no excessively darkened or The FID chromatograms in Figure 1 illustrate these matrix
black areas. Both bilaminate and trilaminate susceptors are interactions and their effect on detector response when the
subjected to the same test conditions. Certainly the test protocol headspace sampling technique is used. The same segments of
temperatures are milder than the cooking temperatures re­ 3 headspace chromatograms are shown. Chromatogram A is
ported by Lentz and Crossett. However, when higher tempera­ the headspace chromatogram representing the analysis of a 7-
tures were attempted, controlled heating of the susceptors be­ component standard in 10 |iL water. All 7 analytes can be seen
came more difficult and reproducible results were not obtained, (see caption to Figure 1, chromatogram A, for fist of analytes).
often because the susceptor material was partially destroyed. Chromatogram B represents the analysis of a susceptor test
Not only is careful control of both the temperature and the strip that was oven-dried overnight at 212°F prior to fortifica­
heating conditions important for achieving reproducible re­ tion. Typically this action removes the volatiles, thus creating
sults, but preconditioning the magnetron tube or microwave- a blank matrix for external calibration. After oven-drying, the
producing device is also important. Microwave ovens are well test strip was fortified with the same 7 analytes at the same
known for uneven heating within the oven cavity. Much of this levels used in A, and then was analyzed. Only 3 compounds
has to do with the performance of the magnetron tube. By pre­ responded: furfural (a), 4-heptanone (c), and styrene (d). No
heating the tube and using controlled cooling between tests, we responses were obtained for the 4 remaining analytes. Com­
improved the reproducibility of our results, as indicated by the parison of chromatograms A and B shows that the susceptor has
appearance of relatively reproducible amounts of furfural, cro- a strong affinity for certain chemicals, namely, polar oxygen­
tonaldehyde, and other paper degradation products. ated compounds, and little affinity for nonpolar compounds
such as styrene.
The selection of appropriate microwave test conditions for
Chromatogram C in Figure 1 represents the headspace sam­
heating specimens and the choice of analytical methodology
pled over another test strip of the susceptor shown in chroma­
were not mutually exclusive tasks and proceeded simultane­
togram B. However, this time the strip was first fortified with
ously. When it became apparent that many volatile chemicals
only 1 analyte, the internal standard 4-heptanone (c) (at the
were being formed during microwave heating of susceptors, it
same level as in chromatogram B), then microwaved under the
was obvious that capillary GC would be the method of choice
previously described heating test conditions, and analyzed. All
for optimum resolution of this mixture. A combination of cap­
of the analytes responded except 2 -ethylhexanol (f), which was
illary GC and static headspace sampling provided great sensi­
not included in the susceptor formulation. All of the chroma­
tivity for identifying and quantifying minor components that
tographic peaks represent compounds released from the
may be present at only trace (ng/in.2) levels. Moreover, this susceptor as the result of microwave heating. The residual
methodology has been used previously in multiresidue deter­ water present in the susceptor paperboard plus the 10 |iL water
minations, and when automated, has excellent reproduciblity added (containing the internal standard, 4-heptanone) was ade­
(21). quate to reduce the strong affinity of the paper for certain polar
Volatiles released from susceptors were identified by analytes in this susceptor. More or less added water may be
GC/mass spectrometry (MS) using manual headspace sam­ required for the determination of other chemicals. However,
pling and capillary GC/MSD. We identified a large number of the addition of too much water may reduce the amount of polar
volatile chemicals, including many polar compounds. Because analytes that can partition into the vapor phase and raise their
of the broad range of boiling points represented by these com­ detection limits.
pounds, we chose a 5% diphenyl-95% dimethyl polysiloxane How this variable (the total of residual plus added water)
capillary column. Because of the limited sensitivity of the mass affects quantitative results is illustrated by the variation in the
selective detector for trace volatiles, 2 mL headspace injections responses for 4-heptanone in the 3 chromatograms (Figure 1).
were made in the splitless mode. This resulted in some loss of Although the concentration of 4-heptanone, peak c, was the
chromatographic resolution, especially for the very early­ same for all 3 analyses, its response differed by as much as a
eluting volatiles. Also, the presence of significant amounts of factor of 2. The 4-heptanone response in chromatogram A was
water from the susceptor paper interfered with chroma­ twice the response in chromatogram B and slightly greater than
tographic resolution and the identification of early-eluting that in chromatogram C. Therefore, when this chemical is used
compounds. as the internal standard, the less polar substances will appear to
To achieve greater sensitivity for quantitation, automated be present in greater amounts because of their low affinity for
headspace GC/FID was used instead of GC/MS. By using a cellulose and water. On the other hand, chemicals more polar
capillary column with a 1 . 0 p,m film operating in the split than 4-heptanone will appear to be present in lesser amounts
1272 M c N e a l & H o l l if ie l d : J o u r n a l O f A O A C I n t e r n a t io n a l Y o l . 76, N o . 6 ,1 9 9 3

A B C

MU lL
i—.—i—i—i—i—i i i
18 24 30 36

F ig u r e 1. C h ro m a to g r a m A : A u to m a te d h e a d s p a c e a n a ly s is o f 10 p L m ix e d a q u e o u s s ta n d a rd . F o rtific a tio n le v e ls o f
c o m p o u n d s a - g in o rd e r o f e lu tio n : (a) 5.7 |ig fu rfu ra l, (b) 5 .6 p g fu ra n m e th a n o l, (c) 2 .45 p g 4 -h e p ta n o n e , (d) 2 .4 p g
s ty re n e , (e) 2 .3 p g 2 -b u to x y e th a n o l, (f) 2.0 p g 2 -e th y lh e x a n o l, a n d (g) 2 3 .7 p g 2 -(2 -b u to x y e th o x y )e th a n o l.
C h ro m a to g r a m B : A u to m a te d h e a d s p a c e a n a ly s is o f o v e n -d rie d s u s c e p t o r te s t s t r ip fo rtifie d w ith m ix e d s ta n d a rd
u s e d in A . C h ro m a to g r a m C : A u to m a te d h e a d s p a c e a n a ly s is o f m ic r o w a v e d s u s c e p t o r te s t s t r ip fo rtifie d w ith 2.45 p g
4 -h e p ta n o n e , th e in te rn a l s ta n d a rd , in 10 p L w ater.

because of their greater affinity for water and cellulose. This stance migrated into the food and remained there, its concen­
suggests that any quantitative data developed by using a single tration in the food would therefore range from ca 0 . 2 2 pg/g for
internal standard and a standard technique for quantitation waffles and meat pies to ca 0.081 pg/g for the pot pie. Large
based on the formation of ratios are at best rough estimates of numbers of chromatographic peaks, many representing chemi­
the probable amounts of substances present. cals below the estimated 0.5 pg/in . 2 level, were encountered in
the susceptor survey. We arbitrarily selected this value, i.e., 0.5
Survey o f Susceptors in the Local Market pg chemical/ 1 in .2 susceptor surface as the lowest level for de­
termining unidentified chemicals. With this criterion, we were
Using the above procedures, a limited survey was con­ able to compare the different susceptors.
ducted that included analysis of 1 1 susceptors packaged with A test strip representing each susceptor was microwaved ac­
1 0 different food products purchased from grocery stores in the
cording to the conditions described above and analyzed by
Washington, DC area during September 1989. All of the
manual headspace sampling and capillary GC/MSD. Total ion
susceptors were bilaminate constructions. Susceptor designs
chromatograms were recorded, mass spectra were obtained,
included flat sheets and trays, sheets glued to corrugated paper,
and peaks were identified. Table 1 lists 44 volatile chemicals
sleeves with overlapping seams, a paper pie pan, and some
identified from the 11 susceptors analyzed. Ten peaks having
susceptors with perforations penetrating through the food con­
tact surface to ease creasing or to serve some other function. responses greater than the 0.5 pg/in . 2 response of the internal
One package, for heating fish sticks, incorporated 2 different standard remain unidentified. Many more compounds having
susceptor sheets that sandwiched the fish between them. Seven responses less than or equal to 0.5 pg/in .2 could not be identi­
of the susceptors were packaged in contact with the foods, and fied; most of them appear to be aldehydes, ketones, furan de­
may have contained some volatiles from the foods. rivatives, and aliphatic hydrocarbons, and also appear to be
Because foods varied in mass and shape, different amounts common to all the susceptors tested.
of food came in contact with the susceptor surface area. In the All the susceptors were screened for the suspected carcino­
survey, the weight of the food in contact with 1 in . 2 of susceptor gens acrylonitrile, methylene chloride, chloroform, benzene,
surface ranged from ca 2.3 g/in. 2 for waffles and meat pies to and ethyl acrylate. Only benzene was found. It was present in
ca 6.2 g/in. 2 for the pot pie. If 0.5 pg/in . 2 of a chemical sub­ 8 of the susceptors tested at a <0.01-0.22 pg/in . 2 level. Accord-
 M c N eal & H o l l if ie l d : Jo urnal O f A O AC I n t e r n a t i o n a l V o l . 76, N o. 6 ,1 9 9 3 1273

T a b le 1. Id e n titie s a n d c o n c e n tr a tio n s o f 4 4 v o la tile c o m p o u n d s re le a s e d fro m 11 d iffe re n t s u s c e p t o r s d u rin g

m ic r o w a v e c o o k in g

MSD No. Concn MSD No. Concn

Compound response3 occurrences range, pg/in2 Compound response3 occurrences range, gg/in2

Standard additions m ethod6 Identified by using M S library3

Acetone/furan (coelutors) X 11 0 .0 6 -3 6 Butanal 2

Isopropanol 1 Methyl vinyl ketone X 8
Vinyl acetate X 6 0 .0 1 -0 .8 8 Pentanal X 9 0 .3 5 -3 .1
Methyl ethyl ketone 1 Methyl furan 2
Formic acid 2 Ethyl furan 1
2-Methyl-1-propanol X 7 1 6 .4 1 -1 0 5 Butanoic acid 1
1,1,1-Trichloroethane X 4 1 .0 5 -7 5 3-M ethyleneheptane X 2 0 .2 4 -1 .8
Acetic acid X 8 2-(Propoxy)ethanol X 1 0 .8 1 -1 .3
Crotonaldehyde X 7 0.1 0—4.5 Heptanal X 5
Benzene X 8 < 0 .0 1 -0 .2 2 2-Heptanone 2
Butanol X 8 2 .0 4 -7 2 3-Heptanone 1
Methylmethacrylate X 7 0 .0 2 -4 .2 Octanal X 5
Toluene X 5 0 .0 6 -1 .1 1 -Phenylpropanedione X 2
Hexanal X 11 0 .2 2 -8 .5 Nonanal X 4
Furfural X 10 0 .5 2 -3 2 Octyl acetate 1
Furan methanol 7 0 .0 7 -1 8 Benzoic acid 1
Butyl ether 1 0 .3 5 -0 .3 9 Decanal 1
Styrene X 7 0 .0 4 -2 .0 5-Hydroxymethylfurfural 1
Butyl acrylate 4 0 .0 5 -0 .6 8
2-(Butoxy)ethanol 6 0 .0 2 -5 .8
2-Ethyl-1-hexanal X 1
Benzaldehyde X 9 0 .0 2 -7 .0 4
2-Ethyl-1-hexanol X 3 0 .2 6 -6 .8
o-Dichlorobenzene 1
2-(2-Butoxyethoxy)ethanol X 4 2 1 .4 4 -1 2 0

a Comoound occurring at least once having M S D response ca >20% of that of the internal standard.
b Identification is based on coincident retention times of autnentic standards plus agreem ent within 5% intensity of the 10 most abundant ions.
c Identification of a susceptor volatile is based on a >90% match with the NB S (formerly National Bureau of Standards; now National Institute
of Standards and Technology (N IS T)) library reference spectrum.

ing to NFPA-SPI, the plastics industry has changed formula­ min) are common to both susceptors and originate from paper
tions, thus eliminating benzene from microwave susceptor degradation. In chromatogram A, many of the major peaks are
products. None of the other 4 compounds mentioned above adhesive-related. They are solvents, terminal units of the
was found at >0 . 1 (ig/in. 2 levels. acrylic polymer adhesive backbone such as 2 -methyl-1 -pro­
Also in Table 1 are the results of quantitative analyses of the panol (6.7 mm), 1,1,1-trichloroethane (7.4 min), n-butanol (8 . 6
survey susceptors. The data developed by the method of stand­ min), and 2-(2-butoxyethoxy)ethanol (33.0 min), or monomers
ard additions are indicated. The amounts of the other chemicals such as styrene and butyl acrylate (22.4 and 22.5 min, respec­
listed are only estimates derived from forming a ratio between tively). In chromatogram B, except for 2-ethylhexanol (at 28.0
the peak area of the chromatographic response for each chemi­ min), no major adhesive-related peaks are observed.
cal and the peak area response for 2.45 (ig/in. 2 of 4-heptanone, It has been reported that the major thermal decomposition
the internal standard. product of vinyl acetate-based adhesives is acetic acid (22). Its
All the survey susceptors were screened by automated characteristic ion fragmentation pattern is easily identified by
headspace sampling and capillary GC/FTD, and their chroma­ MSD but not by FID. It is poorly resolved on the 5% diphenyl-
tograms were compared. Most of the GC profiles (Figure 2) 95 %dimethyl polysiloxane column and elutes as a shallow
were of 2 types: those containing acrylic adhesives (chromato­ broad band at about 8 min (Figure 2B).
gram A) or those containing vinyl acetate-based adhesives Also, the chromatograms shown in Figure 2 represent 2
(chromatogram B). In both chromatograms, the response at susceptors that were not packaged with the food product. More
about 21.5 min is due to the internal standard, 4-heptanone. than 140 unique peaks were observed in these 2 chromato­
Many peaks including crotonaldehyde (at 7.8 min), hexanal (at grams, of which 41 were estimated to represent volatile chemi­
17.1 min), furfural (at 19.2 min), and furan methanol (at 21.4 cals at >0.5 (ig/in. 2 levels. When the total area for each chro-
1274 M cN e a l & H o l l if ie l d : Jo urnal O f AO AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3

F ig u r e 2. C h ro m a to g r a m A : FID c h ro m a to g ra m o f b ila m in a te -ty p e s u s c e p t o r c o n ta in in g a n a c r y lic - b a s e d a d h e s iv e .

C h ro m a to g r a m B : FID c h ro m a to g ra m o f b ila m in a te -ty p e s u s c e p t o r c o n ta in in g a v in y l a c e ta te -b a s e d a d h e s iv e .

matogram was divided by the area of the internal standard, the typical of susceptor materials following actual cooking appli­
estimated concentration of total volatiles released by the cations. Although this criterion is somewhat arbitrary, we feel
susceptor in chromatogram A was about 5 times greater than that it results in a reasonable test protocol that is neither too
that released by the susceptor in chromatogram B. These very mild nor too abusive. These uniform conditions permit the
limited data illustrate the variability in the number and amounts comparison of results from one susceptor product with those
of chemicals generated from product to product. Within-labo- from another.
ratory repeatability of responses generated by the internal The ASTM protocol differs from FDA’s method by allow­
standard is illustrated by replication of the peak area response ing the analyst to choose different microwaving conditions for
for 2.45 |ig 4-heptanone and is reflected in the following data: each test to mimic the time-temperature profile that would be
n = 55, mean - 7308, standard deviation = 546, and relative obtained by monitoring food-susceptor interface temperatures
standard deviation = 7.5%. Recoveries for representative com­ when cooking the corresponding food (8 ). These profiles are
pounds are shown in Table 2. They were calculated for the low­ plots of the temperature of the food-susceptor surface interface
est fortification level of each compound from the method of at a given location as a function of microwave heating time.
standard additions. Recoveries for toluene were high because The measurements are spatially dependent determinations usu­
the toluene peak coeluted with an unidentified component. ally made with fluoroptic probes such as those used with the
Luxtron™ thermometer. Such profiles, based on interface
Comparison o f FDA Protocol with A ST M Standard
measurements, often represent cooking temperatures that have
M ethod for Microwave Susceptor Volatiles
been greatly moderated by many factors, such as the food’s
FDA’s protocol is very similar to ASTM Method F-1308- initial temperature, mass, and water and oil content.
90, for which the agency collaborated with industry repre­ Furthermore, Lentz and Crossett (1) showed that the
sentatives (10). However, FDA’s procedure uses constant mi­ fluoroptic probes of the Luxtron™ thermometer cannot follow
crowave heating parameters. Susceptor test strips are heated in rapidly rising susceptor temperatures, which are more accu­
a 700 watt microwave oven for 5 min with 250 mL water that rately observed with an infrared camera. For example, when
acts as a food load. Volatiles from the susceptor test strips are popcorn was microwaved, after 2 0 s of microwave heating the
determined by static headspace sampling and capillary GC. Al­ infrared camera showed interface temperatures as high as ca
though not routinely monitored, temperatures of our test strips, 482°F, compared with temperatures of only ca 266T obtained
as determined by fluoroptic probes, generally reach ca 370°F. with the Luxtron™ thermometer. After 3 min, the recom­
More important, the appearance of the test strips afterwards is mended cooking time for the popcorn, temperatures measured
 M cN e a l & H o l u f ie l d : Jo urnal O f A O AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3 1275

T a b le 2. R e c o v e r y o f s e le c t e d c o m p o u n d s a t lo w e s t le v e l o f fo r tific a tio n 9

Fortification level,
Compound pg/in.2 n M ean ree., % SD RSD, %

2-M ethyl-1-propanol 0.98 3 86 2.5 2.8

1 ,1 ,1-Trichloroethane 0.8 4 104 2.0 1.9
Toluene 0.05 3 152 18.8 12.4
2-Butoxy-1 -ethanol 2.69 4 87 15.1 17.5
2-Ethyl-1-hexanol 0.83 4 69 15.0 21.8

a Derived by the method of standard additions.

were ca 464°F with the infrared camera but only ca 374°F with (3) TAPPI (1991) Proceedings, Polymers, Laminations & Coat­
the Luxtron™ thermometer. Thus, the difficulty of obtaining ings Conference 2, 801-825
reliable interface temperatures, documented by Lentz and (4) TAPPI (1991) Proceedings, Polymers, Laminations & Coat­
Crossett, makes reliance on time-temperature profiles as a ings Conference 2, 767-772
measure of susceptor performance a questionable tactic. (5) TAPPI (1991) Proceedings, Polymers, Laminations & Coat­
ASTM Method F-1308-90 no doubt provides some data that ings Conference 2, 773-779
are appropriate for evaluating specific food applications. How­ (6 ) Booker, J.L., & Friese, M.A. (May 1989) Food Technol. 110—
ever, because of the many parameters that influence tempera­ 117
tures at the susceptor-food interface as well as the difficulty of (7) Advance Notice of Proposed Rule-Making; Materials for
accurately obtaining reliable interface temperature measure­ Use Under High Temperature Conditions in Microwave Ov­
ens. Fed. Regist. Sept 8 , 1989, 54, 37340-37343
ments, we felt that this method was not appropriate for compar­
(8 ) Kashtock, M.E., Wurtz, C.B., & Hamlin, R.N. (1990) J.
ing microwave susceptor products.
Packaging Technol. 4(2), 14-19
(9) ASTM Standard F-874-91 (1991) American Society for Test­
Summary ing and Materials, Philadelphia, PA
(10) ASTM Standard F-1308-90 (1990) American Society for
A general protocol has been described for using simulated Testing and Materials, Philadelphia, PA
cooking tests to determine the volatile chemicals generated (11) Castle, L., Mayo, A., Crews, C., & Gilbert, J. (1989)7. Food
when microwave susceptors are heated. The difficulties of ac­ Prot. 52, 337-342
curately quantifying the amounts of specific chemicals have (12) Castle, L„ JickeUs, S.M., Gilbert, J., & Harrison, N. (1990)
been discussed when evaluating the method of standard addi­ FoodAddit. Contam. 7, 779-796
tions as opposed to internal standard ratioing based on 4-hep- (13) Begley. T.H., & Hollifield, H.C. (1989) J. Agric. Food Chem.
tanone. When possible, the method of standard additions is pre­ 3 8 , 145-148
ferred because of strong matrix effects. (14) Begley, T.H., Dennison, J.L., & Hollifield, H.C. (1990) Food
Results of a limited survey illustrate the type of data ob­ Addit. Contam. 7, 797-803
tained when this protocol is applied to susceptor products from (15) Begley, T.H., & Hollifield, H.C. (1990) J. Food Prot. 53,
different manufacturers. In this study, most of the volatile 1062-1066
chemicals occurred in very small amounts, with fewer than 50 (16) American Chemical Society (1991) Food and packaging in­
teractions II, ACS Symposium Series 473, FDA Studies of
reaching levels as high as the fig/in. 2 level. Many peaks remain
High-Temperature Food Packaging, Chapter 3, pp. 22-36
unidentified. Future investigations will focus on use of similar
(17) American Chemical Society (1991) Food and packaging in­
headspace techniques to study the effect of high temperatures
teractions II, ACS Symposium Series 473, Application of a
on the paper, adhesive, and polymer components of microwave Teflon Single-Sided Migration Cell for Measuring Migration
susceptors and similar food packages. Through Microwave Susceptor Films, Chapter 5, pp. 53-66
(18) TAPPI (1991) Proceedings, Polymers, Laminations & Coat­
References ings Conference 2, 789-799
(19) ASTM Standard F-1449-91 (1991) American Society for
(1 ) L e n t z , R .R . , & C r o s s e t t , T .M . ( 1 9 8 8 ) Microwave World 9 ( 5 ) , Testing and Materials, Philadelphia, PA
1 1 -1 6 (20) ASTM Standard F-1319-90 (1990) American Society for
(2 ) A m e ric a n C h e m ic a l S o c ie ty (1 9 9 1 ) F o o d a n d p a c k a g in g in ­
Testing and Materials, Philadelphia, PA
te ra c tio n s II, ACS Symposium Series 473, D e t e r m i n i n g (21) McNeal, T.P., & Hollifield, H.C. (1990) J. Assoc. Off. Anal.
V o la tile E x t r a c tiv e s f r o m M ic r o w a v e S u s c e p to r F c o d P a c k a g ­ Chem. 7 3 ,328-331
in g , C h a p t e r 6 , p p . 6 7 - 7 8 (22) Grassie, N. (1951) Trans. Faraday Soc. 48, 379-387
1276 C h a s e Et A l .: J o urnal O f A O AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3

F O O D C O M P O S IT IO N A N D A D D IT IV E S

M e th o d M o d ific a tio n fo r L iq u id C h ro m a to g ra p h ic

D e te rm in a tio n o f T h ia m in e , R ib o fla v in , a n d P y rid o x in e in

M e d ic a l F o o d s

G. W illiam C hase, W illiam O. L anden, Jr, and Abdel-G awad M. Solevian

U.S. Food and Drag Administration, Atlanta Center for Nutrient Analysis, 60 8 th St, NE, Atlanta, GA 30309
R onald R. E ttenmiller
University of Georgia, Department of Food Science and Technology, Athens, GA 30602

A r e v e r s e d - p h a s e io n p a ir li q u id c h r o m a t o g r a p h ic frequently for a wider variety of sample matrixes, we observed

m e t h o d d e v e lo p e d f o r t h e s im u lt a n e o u s d e t e r m in a ­ that certain soy-based formula extracts contained components
t io n o f t h ia m in e ( B i), r ib o f la v in ( B 2 ), a n d p y r id o x in e that interfered with the B , and B6 peaks and made quantitation
(B e ) in p e r c h lo r ic a c id e x t r a c t s o f in f a n t f o r m u la s difficult. In addition, when the LC method was applied to
w a s m o d if ie d t o in c lu d e m e d ic a l f o o d s . U V d e t e c ­ medical foods, interferences were observed in the chromato­
t io n o f B 1 a n d B 2 w a s r e p la c e d b y f lu o r e s c e n c e d e ­ gram that made quantitation impossible for B| and B6.
t e c t io n , w h ic h r e s u lt e d in im p r o v e d s e n s i t iv it y a n d The objectives of this study were 3-fold. The original LC
s p e c if ic it y . B 1 w a s d e t e c t e d b y f lu o r e s c e n c e a ft e r method needed improvement for reasons previously men­
c o n v e r s io n t o t h io c h r o m e b y a p o s t c o lu m n r e a c ­ tioned, and a multivitamin LC method was needed for analysis
t io n w it h s o d i u m h y d r o x id e a n d p o t a s s iu m f e r r ic y -
of medical foods. A programmable LC fluorometric detector
was used to detect B2 and B6 simultaneously. Additionally, B,
a n id e . T h e m e t h o d u s e s a m o b ile p h a s e o f w a te r ,
was converted to thiochrome so that B! could be detected
a c e t o n it r ile , h e x a n e s u lf o n ic a c id s o d i u m s a lt , a m ­
fluorometrically, which eliminated the use of UV detection and
m o n iu m h y d r o x id e , a n d p h o s p h o r ic a c id a d j u s t e d
increased the sensitivity and specificity of the method.
t o p H 3 .6 . T h e c o lu m n i s a 3 0 0 x 3 .9 m m N o v a P a k
C 18- L im it s o f d e t e c t io n w e r e 0 .0 5 p g / m L f o r B 1 a n d
M ETHOD
B 2 a n d 0 .0 1 p g / m L f o r B 6 b y f lu o r e s c e n c e d e t e c ­
t io n . T h e s y s t e m r e p r o d u c ib i lit y w a s e v a lu a t e d b y
c o m p le t in g 1 0 r e p e t it iv e d e t e r m in a t io n s o n a m e d i­ A p p a r a tu s
c a l f o o d t h a t g a v e a c o e f f ic ie n t o f v a r ia t io n o f 5 .9 ,
6 .0 , a n d 1 0 .7 % f o r B 1 , B 2 , a n d B6, r e s p e c t iv e ly . (a) Liquid chromatograph.—Model 6000 A solvent deliv­
M e a n r e c o v e r i e s ( n = 1 0 ) w e r e 1 1 1 ,9 6 .3 , a n d 1 13% ery system, and Model 712 WISP autoinjection or U6 K injec­
f o r B i , B 2, a n d B6, r e s p e c t iv e ly . T h e r e s u lt s c o m ­
tor for manual injection (Waters Associates, Milford, MA
p a r e d f a v o r a b ly w it h t h o s e b y A O A C O f f ic ia l M e t h ­
01757).
o d s 9 4 2 .2 3 ,9 4 0 .3 3 , a n d 9 6 1 .1 5 f o r B i , B 2, a n d B 6, re ­
(b) Column.—Pico Tag (C 18 Nova Pak) stainless steel, 300
s p e c t iv e ly .
x 3.9 mm, No. 10950, with in-line precolumn filter No. 84560
(Waters Associates).
(c) Integrator.— Model 3396 or equivalent (Hewlett-
R ecently, we developed a liquid chromatographic (LC) Packard, Atlanta, GA 30339).
method for the simultaneous assay of thiamine (BL), ri­ (d) Fluorescence detector.—Model 1046A programmable
boflavin (B2), and pyridoxine (B6) in infant formula (1). fluorescence detector (Hewlett-Packard).
The LC method uses a perchloric acid extraction followed by (e) Column heater.—Model CH30 with Model TC-50 con­
reversed-phase ion pair chromatography. It provides a fast, ac­ troller (Eppendorf, Madison, W I53711).
curate, and reliable means to determine B b B2, and B 6 simul­ (f) Postcolumn reaction pump.—Model A-30-SW (Eldex
taneously and avoids the use of 3 time-consuming AOAC as­ Laboratories, San Carlos, CA 94070).
says, Official Methods 942.23, 94033, and 961.15 for Bb B2, (g) Reaction coil.—Constructed of 18 ft, 0.010 in. id stain­
and B6, respectively. Because we used the LC method more less steel tubing coiled and placed inside the column heater.
(h) pH meter.—Model 125 (Coming Science Products,
Received August 8, 1992. Accepted January 21, 1993. Medfield, MA 02062).
Presented at the 106th AOAC Annual International meeting, August (i) Nylon filter.—0.45 pm (Universal Scientific, Inc., At­
31-September 3, 1992, at Cincinnati, OH. lanta, GA 30360).
 Chase Et A l .: J o urnal O f A O AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3 1277

R eagents Loop

(a) Perchloric acid.—69-72% (J.T. Baker Inc., Phil-

lipsburg, NJ 08865).
(b) Acetonitrile.—Disti Iled-in-glass, UV grade (Burdick
and Jackson Laboratories, Inc., Muskegon, MI 49442).
(c) Ion-pairing reagent.— 1-Hexanesulfonic acid sodium
salt (Eastman Kodak Co., Rochester, NY 14650).
(d) Potassium ferricyanide.—Crystals (J.T. Baker).
(e) Stock standard solutions.—Individual stock standards
of thiamine-HCl (0.1 mg/mL), riboflavin (0.02 mg/’mL), and
F ig u re 1. S c h e m a tic o f th e L C s y s t e m f o r th e a s s a y o f
pyridoxine-HCl (0.1 mg/mL or 0.08 mg/mL as free base) were
th ia m in e , rib o fla v in , a n d p y rid o x in e . T o a s s a y rib o fla v in
prepared from U.S. Pharmacopeial (USP) Standards
a n d p y rid o x in e , t u b in g s m a rk e d “ A ” a n d “ C ” a re
(Rockville, MD 20852) in 0.1N HC1 after being dried as di­
c o n n e c te d . T o a s s a y th ia m in e , tu b in g m a rk e d “ A ” is
rected by USP. The stock standards were stable for several
c o n n e c te d to tu b in g m a rk e d “ B ,” a n d tu b in g m a rk e d “ C ”
months under refrigeration.
is c o n n e c te d to tu b in g m a rk e d “ D .”
(f) Mobile phase.— 1-Hexanesulfonic acid (0.95 g) was
dissolved in 1 L water containing 9.5% acetonitrile and 0.5 mL
ammonium hydroxide. The entire solution was adjusted to pH mulations represented different formulation bases that con­
3.60 with phosphoric acid. A 930 mL portion of this solution sisted of a protein source derived from egg white solids, con­
was diluted to 1 L with water. The mobile phase was then fil­ centrated sweet skim milk, 3 formulations of sodium and cal­
tered through a nylon filter. cium caseinates and soy protein isolates, calcium and sodium
caseinates, and protein components of partially hydrolyzed
(g) Derivatizing solution.—A 1 g portion of potassium fer-
whey, meat, and soy with free essential amino acids. Sample
ricyanide, K3 Fe(CN)6, was dissolved in 100 mL water. A
composites were prepared under subdued light according to
6.0 mL aliquot of the K3 Fe(CN ) 6 solution was added to 100 mL
AOAC instructions (2) for proper warming, opening, mixing,
15% NaOH. The solution was filtered through a nylon filter.
and storage under nitrogen atmosphere and refrigeration.
(h) Other reagents.—Hydrochloric acid, phosphoric acid,
ammonium hydroxide, potassium hydroxide, sodium hydrox­ Sam ple Extraction
ide; all reagent grade (Fisher Scientific, Fair Lawn, NJ 07410).
Accurately weigh an analytical portion of sample to contain
Chromatographic Conditions ca 120-180 pg B1; B2, and B6 based on 150% of the declared
amounts into a 500 mL Phillips beaker. Mix with enough water
(a) Instrument parameters.—Injection volume, 125 jlL;
to bring the total volume to ca 60 mL.
flow rate, 1.0 mL/min. Fluorescence detection parameters for
Add 2 mL perchloric acid to each solution while stirring,
B6 and B2: gain 8 , time 0, excitation wavelength (A.) 295 nm,
and continue stirring for 0.5 h. Carefully add 6 M potassium
emission A 395 nm; at time 14 min, excitation A 440 nm, emis­
hydroxide dropwise with constant stirring until the pH is
sion A 565 nm; at time 27 min, excitation A 295 nm; emission
3.2 ± 0.4 (pH meter). Transfer each solution quantitatively to a
A 395 nm. Fluorescence detection parameters for B,: gain 8 ,
200 mL volumetric flask and dilute to volume with mobile
excitation A 360 nm, emission A 435 nm. Reaction coil heater,
phase. Refrigerate the solutions overnight to allow complete
55°C; flow rate of postcolumn reaction pump, 0.30 mL/min.
precipitation of the protein and perchlorate; then filter through
(b) LC system configuration.—With an autoinjector, the
a 32 cm grade 588 pre-pleated filter paper (Schleicher and
injection sequences for the standards and samples in the carou­
Schuell, Keene, NH 03431). Refilter the solutions through a
sel are repeated twice. A specially designed flow system en­
nylon filter.
abled the same fluorescence detector to be used for B2 and B6
as well as B , (Figure 1). During the first injection sequence, the R ecovery Studies
B2 and B 6 are quantitated with the same detector by using a
time program for wavelength changes and connecting tube Perform a recovery experiment by spiking the sample with
ends “A” and “C” together with a low dead volume female-fe­ each standard of B,, B2, and B6 at 200% of the declared values
male connector (Waters Associates P/N 97332) (Figure 1). Af­ for each vitamin. Use the same sample weight as for the sample
ter completion of the first sequence, the “loop” (Figure 1) was analysis. Treat the recovery sample as in Sample Extraction,
installed by connecting end “A” to end “B” and end “C” to end beginning with “Mix with enough w ater....” However, make
“D” with 2 female-female low dead volume connectors. The the appropriate dilutions to keep the peak responses of the re­
second injection sequence was then begun to quantitate the covery within the standard curve.
thiochrome by using a second set of instrument parameters.
Preparation o f Standards for Calibration Curve
Sam ple Description and Preparation
Prepare the working standard by combining 2.0, 10.0, and
For this study, 7 medical food samples from 3 different 2.0 mL aliquots of stock standard solutions of B,, B2, and B6,
manufacturers were used: 3 powders and 4 ready-to-feed for- respectively, in a 250 mL Phillips beaker containing ca 50 mL
1278 Chase Et A l .: J o urnal O f A O AC I n t e r n a t i o n a l V o l. 76, N o, 6 ,1 9 9 3

water. From this point on, treat the standard and the sample
portions the same, beginning as in Sample Extraction, “Add
2 mL perchloric acid... Once the standard mixture has been
filtered through a nylon filter, make 3 additional working
standards for the standard curve by diluting 2.0, 4.0, 8.0, and
10.0 mL aliquots to 10.0 mL with mobile phase to give final
concentrations of 0.2, 0.4, 0.8, and 1.0 |ig Bb B2, and B 6 per
mL, respectively.

AOAC Method
|------------26 min------------1 |------------26 min------------ 1
The following AOAC methods were used for the analysis of
F ig u r e 2. C h r o m a to g r a m s o f th e s ta n d a rd m ix tu re (A)
the medical foods: B, manual thiochrome method (3); B 2 mi­
b y f lu o r e s c e n c e d e te c tio n a n d a m e d ic a l fo o d e x tra c t
crobiological method (4); B6 microbiological method, except
(B), w h e re (1) is p y r id o x in e a n d (2) is rib o fla v in (fo r 14
that column chromatography to separate the B6 vitamers was m in : e x c ita tio n X, 2 9 5 nm ; e m is s io n X, 395 n m ; a fte r 14
not used (5). m in : e x c ita tio n X, 4 4 0 nm ; e m is s io n X, 565 nm ).

Results and Discussion

sible. For these reasons, the use of the internal standard was
This study, unlike our earlier work (1), used a programma­ abandoned and a standard curve was used.
ble fluorescence detector for Bh B2, and B 6 determination. The standard curve was linear from 0.2 to 1.0 pg/mL for B x,
Fluorescence detection enabled the limit of detection to be low­ B2, and B6. The system reproducibility was evaluated by com­
ered from 0.15 to 0.05 pg/mL for Bj and from 0.09 to
pleting 1 0 replicate analyses of a powdered medical food with
0.05 pg/mL for B2. Because the detection was more sensitive,
a major protein source derived from egg white solids and la­
the sample size needed for the assay could be decreased or the
beled to contain 9.60, 8.67, and 12.4 pg B h B2, and B 6 per
final dilution increased.
gram, respectively. The mean values in pg/g ± the standard
The column was also modified from our earlier work (1). A
deviations [coefficient of variation (CV)] were 15.0 ± 0.9 (CV,
30 cm Pico-Tag column replaced the 15 cm column. Although
5.9%), 9.48 ± 0.6 (CV, 6.0%), and 16.0 ± 1.7 (CV, 10.7%) for
the 30 cm column is described as an amino acid analysis col­
Bb B2, and B6, respectively. Recovery studies were evaluated
umn, it contains the same packing material as the previously
with each sample assayed. The mean percent recoveries (n =
used 15 cm C 18 Nova Pak column. The difference is that the
10) and standard deviations were 111 ± 7.9, 96.3 ± 7.2, and
30 cm column contains a higher quality packing material that
is subjected to more rigid quality control testing. Despite longer 113 ± 10.0 for B |, B2, and B6, respectively.
retention times with the 30 cm column, the chromatography is Table 1 presents the declared amounts of B , and the values
excellent, with superb resolution between B2 and B 6 (Figures obtained by the LC and the AOAC methods (3). The values
2A and 2B). The capacity factors (K), where K = (time spent in declared on the label are expressed as mg/serving. For this
the stationary phase/time spent in the mobile phase) - I, for B,, study the declared units were converted to pg/g. The LC results
B2, and B6 were 7.04, 11.4, and 2.71, respectively. The chro­
matograms were free of chromatographic interferences and did
not require extra time between injections to allow for late-
eluting peaks before returning to baseline as our earlier method
( 1 ) did for soy-based infant formulas.
Figures 3A and B depict the elution profile of the thio­
chrome derivative. The conversion of thiamine to thiochrome
coupled with fluorescence detection and separation on a 30 cm
column made the system highly specific for thiamine, and no
interference peaks were observed for any of the medical foods
studied.
The use of the internal standard (1), m-hydroxybenzoic
acid, was eliminated in this study. Even though the autoinjector
was programmed to add the internal standard and therefore
eliminate human error, fluctuations were observed in peak re­
sponse that made quantitation difficult. Additionally, medical F ig u r e 3. C h r o m a to g r a m s o f th e th ia m in e s ta n d a rd (A )
foods containing free amino acids also posed problems with the b y flu o r e s c e n c e d e te c tio n a n d a m e d ic a l f o o d e x tra c t
internal standard. Tryptophan co-eluted with the internal stand­ (B), w h e re (1) is th io c h r o m e (e x c ita tio n X, 3 6 0 nm ;
ard in the B 2 and B6 assays; therefore, quantitation was impos- e m is s io n X, 4 3 5 nm ).
 C h a s e E t A l .: J o urnal O f AO AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3 1279

T a b le 1. C o m p a r is o n o f v a lu e s o b ta in e d f o r th ia m in e T a b le 3. C o m p a r is o n o f v a lu e s o b ta in e d f o r v ita m in Be
(a s th ia m in e H C I) in m e d ic a l f o o d b y L C a a n d A O A C * in m e d ic a l fo o d b y L C a a n d A O A C * M e th o d 961.15
M e th o d 942.23
Vitamin B6, gg/g
Thiamine, gg/g
Medical food Declared LC method AOAC method
Medical food Declared LC method AOAC method
a 10.1 17.4 + 0.8 20.7 + 0.4
a 7.59 21.8 + 0.2 20.9 + 0.2 b 2.49 3.55 ± 0 3.63 + 0.1
b 1.86 2.71 ±0.1 2.46 ± 0 c 12.4 16.0 + 1.7 22.7 + 0.8
c 9.60 15.0 + 0.9 9.27 + 0.2 d 2.49 3.32 + 0.1 3.50 ± 0
d 1.93 2.76 + 0.1 1.83 + 0 e 2.00 3.92 ± 0.5 4.00 + 0.2
e 1.51 3.34 ± 0 2.56 + 0.4 t 2.34 3.13 + 0 3.51 ±0.2
t 1.76 1.69 + 0 1.29 + 0.1 g 24.7 29.5 + 1.0 34.0 ± 0.5
g 18.0 17.9 + 0.7 17.7 + 0.2
3 LC data are the mean of triplicate injections for duplicate
3 LC data are the mean of triplicate injections for duplicate extractions, except for “C,” which is the mean of duplicate
extractions, except for “C,” which is the mean of duplicate injections for 10 extractions.
injections for 10 extractions. 6 AOAC data are the average of duplicate extractions.
6 AOAC data are the average of duplicate extractions.

P eak Purity Evaluations

were at least 96% of the declared values for each sample. The
LC results exceeded the AOAC results in every case and aver­ The purities of the Bb B2, and B6 peaks were confirmed by
aged 19% higher than the AOAC results. A linear regression a procedure described by Haroon et al. (6 ), which we also used
analysis comparing the LC and AOAC method gave a correla­ in an earlier study (1). The emission wavelength was kept con­
tion coefficient of r = 0.973. stant for each nutrient; the fluorescence was measured at 3 spe­
Table 2 illustrates the comparison of results by the LC and cific excitation wavelengths. For example, the emission of B6
the AOAC methods (4) to the declared values of B2. The LC at 395 nm was determined at excitation of 275, 295, and
results exceeded the declared values in every case. Excellent 305 nm. Emission ratios were calculated for 275/295,305/295,
agreement was obtained between the LC and the AOAC meth­ and 275/305. These ratios were compared with the same re­
ods (r = 0.988). The LC results averaged 3.7% lower than the sponse ratios of the standard. Table 4 illustrates the good agree­
AOAC microbiological results. ment between response ratios obtained with the standard and
Table 3 presents the declared values of B6 and the values sample for B,, B2, and B6 and indicates the purity of the peaks.
obtained by the LC and the AOAC methods (5). In every case
the LC results exceeded the declared values. The LC results Conclusions
averaged 14% lower than the AOAC microbiological assay
(r = 0.992).
This multivitamin procedure greatly improves our earlier
method (1) for determining B l5 B2, and B6. Although the thrust
of this study was aimed at medical foods, infant formulas can
be analyzed with this method. Individual AOAC methods with
time-consuming techniques can be virtually eliminated.
T a b le 2. C o m p a r is o n o f v a lu e s o b ta in e d f o r rib o fla v in
in m e d ic a l f o o d b y L C a a n d A O A C * M e th o d 940.33 T a b le 4. E v a lu a tio n o f p e a k p u rity 3

Riboflavin, gg/g Peak response ratio

Medical food Declared LC method AOAC method Nutrient wavelengths Standard Sample

a 8.60 17.1 ±0 15.4 + 1.0 Bi 340/360 0.42 0.41

b 2.12 3.33 + 0.1 3.41 ±0.2 370/360 0.93 0.97
c 8.67 9.48 + 0.6 9.67 + 0.5 340/370 0.46 0.42
d 2.17 2.63 ± 0 2.86 + 0.1 b2 420/440 0.67 0.67
e 1.71 3.08 ± 0 3.00 + 0.2 460/440 0.91 0.91
t 1.99 2.53 ± 0 2.86 + 0.2 420/460 0.74 0.74
g 20.4 20.5 ± 0.6 23.1 ±0.4 Be 275/295 0.39 0.43
305/295 0.45 0.46
a LC data are the mean of triplicate injections for duplicate 275/305 0.86 0.92
extractions, except for “C,” which is the mean of duplicate
injections for 10 extractions. 3 Emission wavelength constant for each nutrient: B„ 435; B2, 565;
b AOAC data are the average of duplicate extractions. B6, 395.
1280 H agg ett Et A l .: J o urnal O f AO A C I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3

Values obtained for B b B2, and B6 agreed closely with the (2 ) Official Methods of Analysis ( 1 9 9 0 ) 15th E d ., A O A C , A r­

results of AOAC methods. Again, as stated in our earlier work lin g to n , V A , se c . 9 8 5 . 3 0

(1), for B6 the method is limited to the determination of pyri- (3 ) Official Methods of Analysis ( 1 9 9 0 ) 15th E d ., A O A C , A r­

doxine and does not measure naturally occurring vitamers of lin g to n , V A , se c . 9 4 2 .2 3

B6. (4 ) Official Methods o f Analysis ( 1 9 9 0 ) 15th E d ., A O A C , A r­

lin g to n , V A , se c . 9 4 0 .3 3

R e fe re n c e s (5 ) Official Methods o f Analysis ( 1 9 9 0 ) 15th E d ., A O A C , A r­

lin g to n , V A , se c . 9 6 1 . 1 5

(1 ) C h a se , G .W ., L a n d e n , W .O ., E ite n m ille r, R .R ., & S o lim a n , (6 ) H aroo n , Y ., B a c o n , D .S ., & S ad o w sk i, J .A . ( 1 9 8 6 ) Clin.

A .M . ( 1 9 9 2 ) J. AOACInt. 7 5 , 5 6 1 -5 6 5 Chem. 3 2 , 1 9 2 5 -1 9 2 9

FOOD COMPOSITION AND ADDITIVES

P a r tia l A u to m a tio n o f M ic ro b io lo g ic a l A s s a y s f o r V ita m in s

T.O . R ichard H aggett, A lan R. M atheson, and M ichelle H arnett

New Zealand Dairy Research Institute, Private Bag 11 029, Palmerston North, New Zealand

T h e d i s p e n s in g a n d t u r b id it y m e a s u r e m e n t s t e p s variable nutritional efficiency. Animal growth has been re­

u s e d in m ic r o b io lo g i c a l a s s a y s o f B g r o u p v it a m in s placed by microbiological growth for this type of assay for rea­
w e re a u to m a te d b y th e u s e o f a G ils o n h a rd w a re sons of cost, precision, speed, and concern for animal welfare,
p a c k a g e , a s p e c i a l l y c o n s t r u c t e d t u r b id im e t e r , a n even though chemical conversions of complex vitamers are
I B M P C - A T o r c o m p a t i b le c o m p u t e r , a n d s o f t w a r e usually necessary to suit the narrow specificity of the test or­
w r it t e n in - h o u s e t o f a c ili t a t e d a t a e n t r y , c o n t r o l t h e ganism. Although microbiological assays (MBA) are complex,
h a r d w a r e , a n d c a lc u la t e t h e r e s u lt s . C o n s id e r a b le the full range of B group vitamins can be determined in labo­
im p r o v e m e n t in t h e s p e e d a n d p r e c is i o n o f t h e t u r ­
ratories in which MBAs are well established. For these reasons,
b id it y m e a s u r e m e n t w a s a c h ie v e d (1 % c o m p a r e d
MBA has survived for over 50 years, since the research of Snell
t o 3 -1 2 % ) . T h e p r e c is i o n o f t h e d is p e n s i n g o p e r a ­
and Strong (1). MBA is the only technique recognized by
t io n w a s b e t t e r t h a n t h a t o f m a n u a l d is p e n s i n g
AOAC for determining some B group vitamins in foods (2).
(0 .0 2 - 0 .8 % c o m p a r e d t o 0 .4 - 2 % ), a lt h o u g h t h e
The assay is labor-intensive, however, which increases the vul­
s p e e d w a s n o t im p r o v e d . H o w e v e r , r e s u lt s a n d e r ­
nerability of the assay to nonrandom human error that can be
r o r s o b t a in e d f r o m t h e a u t o m a t e d s y s t e m d id n o t
s ig n if ic a n t ly d if f e r f r o m t h o s e o b t a in e d b y f o r m e r
difficult to trace. Furthermore, hidden systematic errors relat­
p r o c e d u r e s . T h i s w a s a t t r ib u t e d t o g r o w t h r e ­ ing to the specificity of the microorganism and the nutritional
s p o n s e v a r i a t io n s n o t r e la t e d t o t h e a u t o m a t e d p r o ­ quality of the assay medium are of constant concern.
c e d u r e . T h e a u t o m a t io n p r o c e s s f a c ilit a t e d d e t e c ­ These difficulties provided strong incentives over the years
t io n o f e r r o r s a n d h a lv e d t h e o p e r a t o r t im e p e r for more reliable, simpler methods. Consequently, a variety of
assay. techniques are now successfully and routinely used to deter­
mine specific vitamins. One technique is radioimmunoassay;
however, this method may be as much as 3 times as expensive
G rowth response assays for essential nutrients in food as the corresponding MBA (3). Another new technique, en­
and food ingredients are useful techniques when a spe­ zyme-linked immunosorbent assay, shows promise; assays
cific growth requirement for a nutrient can be exploited were developed for vitamins such as pantothenic acid (3), and
for a particular animal or microorganism. The nutrient is re­ work continues on other vitamins. This technique still needs
ported as the total concentration of nutritionally available ana­ wider evaluation before it can be accepted as a versatile vita­
lyte instead of as multiple forms (vitamers), which may have
min assay methodology. Liquid chromatographic (LC) assays
are widely used for some vitamins for which adequate sensitiv­
R e c e iv e d O c to b e r 2 8 , 1 9 9 2 . A c c e p te d M a rc h 6, 1993.
ity can be achieved, the range of vitamers is narrow, and co-elu­
 H agg ett E t A l .: J o urnal O f AO AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3 1281

Gilson GSIOC link

tion problems requiring specialist attention are not a significant
risk.
Recently, several authors claimed that MBA is the method
of choice for determining natural levels of B group vitamins in
foods (4, 5). Favell (4) suggested that automation can reduce
the tedium of dispensing and of measuring turbidity. Lumley
and Lawrance (5) used a semiautomated technique; they stated
that batching of samples often made these assays more cost-ef­
fective than LC assays.
The above arguments demonstrate that the MBA is still a
necessary tool for a fully versatile vitamin assay laboratory.
The assay would be more acceptable to analysts if nonrandom
human error arising from the high labor requirement could be
more readily controlled. Automation is a way of reducing this
error, because machine error is usually readily traced. A further
benefit of automation is the increase in laboratory efficiency,
measured in terms of labor/capital cost, laboratory space, and
need for repeated assays caused by untraceable major error. F low -through p h o to m e te r
This paper describes the following: use of an automated dis-
pensing/sampling system to dispense the sample, buffer, and Figure 1. Inter-connection of modules.
assay medium into the tubes for microbiological assay; auto­
matic measurement of growth response by turbidimetry after
incubation; software required to control the above equipment dispensing of reagents in Figure 2 and for turbidity measure­
and to process the analytical data; and comparison of results ments in Figure 3.
and associated errors obtained by using the system when cal­ (b) CE373 linear readout grating spectrophotometer.—
culation, turbidity measurement, and reagent dispensing were The spectrophotometer (Cecil Instruments, Cambridge, Eng­
independently examined. land) was modified to read the turbidity directly in thin-wall,
borosilicate glass culture tubes, 150 x 20 mm, by inserting a
Experimental vertical.slit to create an acceptance angle of 1° in front of the
photocell to facilitate turbidity measurements.
Apparatus (c) Commodore computer system.—A Commodore 4032
(32K) computer plus a Commodore 4040 dual disk drive and
(a) Gilson automatic dispensing/sampling system.—A 222 an accompanying Commodore 3023 printer were all manufac­
sample changer with an adaptor kit for long tubes (No. B tured by Commodore Business Machines, Santa Clara, CA.
49376), a 401 diluter fitted with a syringe of 5 mL capacity, a This computer system was used in our laboratory for microbio­
Minipuls 3 MP1 peristaltic pump and the appropriate GSIOC logical assay calculations before the automated procedure was
(Gilson Serial Input/Output Channel) cables, GSIOC 506 mo­ developed.
dem, and 706 GSIOC driver software for MS/DOS computers
were purchased from Gilson Medical Electronics (France)
S.A., Villiers-le-Bel, France.
The computer hardware (Taiwan Redstone Industries Ltd,
Taiwan) consisted of a Redstone IBM AT compatible personal
computer operating in turbo mode (12 MHz) with EGA card
and color screen, 12 bit A/D plug-in board (Metrabyte DAS-
CON-1, Metrabyte Corporation, Taunton, MA).
A regulated power supply (12 V, 5 A continuous) was used
for the turbidimeter lamp supply (IE 12/05 Innovative Energies
Ltd, Auckland, New Zealand). The flow-cell (20 o m path-
length x 7 mm id circular cross-section), flow-cell housing,
photocell, and lamp were obtained from an obsolete Locarte
Mark 4 amino acid analyzer (Locarte Co., London, UK). A
long-pass filter (R60 Polo, Japan) possessing optical properties
similar to a Wratten 25 filter (>10% transmission above
590 nm) was used.
The cable connections between different modules are
shown in Figure 1. The fluid connections for the sample Figure 2. Fluid path for dispensing operation.
changer, 401 diluter, and Minipuls 3 pump are shown for the
1282 H a g g e t t E t A l .: J o u r n a l O f AOAC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

Gilson sample changer model 222

Figure 5. Modified needle for dispensing.
Figure 3. Fluid path for turbidity measurement.

coilwas suspendedabovethesamplechangerandwas allowed

(d) G ilso n h a rd w a re m o d ific a tio n s .— The sample changer tohang freesothatthesamplechangerarm couldmove freely
was modified inthefollowingmanner: (i)The centerrackin through allpositionswithoutsignificantlyaffectingtheaspira­
the tray was replaced with an “extract rack assembly” (Fig­ tioncoilposition.
ure4),made attheNew ZealandDairyResearchInstitute.This A dualneedleassembly (Figure6)was manufacturedatthe
consistedofthesample extractrackand aplastic,square-sec­ New Zealand Dairy Research Institutefrom Vi6 in. od stock
tioncontainer(80x 80x 120mm) toactasareservoirforbasal hypodermic tube(AirDriven Pump Industries,NZ Ltd,Auck­
medium orwater,asrequired.Thesewereheldon aframethat land,New Zealand).
conformedtothebaseoftheGilsontype22testtuberack.The
A u to m a te d S o ftw a re S y s te m
sample extractrack(80mm wide x 150 mm long)was manu­
facturedfrom316 stainlesssteelsheet.Itheld4x7 tubes,com­ Automationsoftwarewas writtenattheNew ZealandDairy
pared with 4x11 tubes forthe Gilson type 22 rack. (2)The ResearchInstitutetoperformthefollowingoperationsby using
needle guide was removed from the sample changer and the theGilsonhardware andRedstonecomputer: enterdataforthe
needleusedforsample dispensingwas stiffenedby enclosing assaystobeperformed; controlthehardwaretodispensesolu­
itin a bushed stainless steeltube as shown inFigure 5. This tionsintothetubes;controlthehardwaretomeasuretheturbid­
tubeended well clearofthewettedsurface.(J)The aspiration ityoftheculturesuspensions;calculatethevitaminconcentra­
tioninthesamplesfromtheraw data;andpasstheresultsinto
aresultsdatabase and linkthisdatabase toasample reception
database.

OOÜOOOO D

B a sa l ooooooo c
m ed iu m
ooooooo B

E x tra ct n u m b e r
ooooooo
1 2 3 4 5 6 7
A

Figure 4. Extract rack layout. Figure 6. Modified needle for turbidity measurement.
 H a g g e t t E t A l . : J o u r n a l O? AO A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1283

R e fe re n c e S o ftw a re fo r C a lc u la tin g R e s u lts L iq u id T ran sfer

The reference method forcalculation used a BASIC pro­ The 3 modes ofliquidtransferbetween tubes(airdisplace­
gram, written in-house, for the Commodore computer. This ment, liquiddisplacement, and liquiddisplacement withbub­
software was used for 10 years (reference software) and was bleseparator)were evaluatedforreproducibilityandaccuracy
extensivelycheckedagainstmanualcalculations.The standard (Table 1).All3modes usedtheaspirationcoilandsyringe.Al­
curve generated by the BASIC program was composed of a though liquid displacement predictably gave the bestresults,
composite offirstand second orderregression lines,whereas themethodwasrejectedonthegroundsthatexcessreagenthad
theautomation softwareusedasinglepolynomialoforderca­ to be aspirated. Such excess volumes could come in contact
pable ofbeing selected up to 6 terms. The Commodore soft­ with the syringe and valve, and therefore, cleaning between
ware allowed only 1 dispensing format, and thisformat was assaysfordifferentvitaminswouldbemoredifficult.Displace­
duplicatedintheautomatedsystemwhen the2calculationpro­ ment withwaterorbufferusinga 20 LlLbubble separatorwas
cedureswere beingcompared. selected.Smaller bubbles tended tobreakup, and largerones
R e a g e n ts a n d M icro b io lo g ical A s s a y
createdmore errorbecause ofairelasticity.Even so,theaspi­
ration coil speed had to be reduced from the maximum of
MBA andreagentsusedinthisworkhavebeendescribedin 1600 pL/s to 1000 pL/s, and the bubble separator had to be
detailby a number ofauthors (1,4, 6, 7). For some turbidity renewed forevery sequence oftransferoperations. Basal me­
measurementwork,asilicasuspensionwas preparedbyadding dium was dispensed by liquid transferfrom the reservoirlo­
0.25± 0.05g Spherisorb5u sphericalsilicaparticlesmade for cated inthecentralrack.The displacement liquid,which was
LC (HPLC Technology, Macclesfield, Cheshire, UK) to also used as the diluent for the assay, was obtained from an
100mL water. externalreservoirbyaspiratingitintothesyringe/valveassem­
blyand dispensingviatheaspirationcoil.The precisionofthe
S tatistical Testing
selectedmethod ofliquidtransferwas evaluatedfortheopera­
Resultsandstandarderrorsgeneratedbytheautomatedsys­ tionsrequiredforan assay(Table2).
tem and reference procedures were compared by using the
P re ve n tio n o f C ro s s C o n tam in atio n
pairedr-testandrelyingonthecentrallimittheoremtonormal­
izedata(8). Crosscontaminationwas consideredtobe ariskduringliq­
uid transfer,especially when shiftingtoassays fora different
Results vitamin. Cross contamination between liquid transfers was
eliminatedby thefollowingprocedures.To preventtheneedle
A u to m atio n S o ftw a re beingwettedtoheightsunreachableby therinsingstation,the
needletipwas programmedtodipintotheliquidtoaminimum
A total of 30 000 lines ofexecutable code occupying 1.2
depth and tofollow the fallingliquidlevel during aspiration.
Mbytes ofmemory, heldinadedicatedsubdirectory,was writ­
The computerkepttrackoftheliquidheightineachtubefrom
ten to support assay and sample data entry, Gilson hardware
thetubedimensions and content volumes. At theend ofeach
control,and record generationfordownloading tootherdata­
aspirating/dispensingsequence,theneedlewas returnedtothe
bases. Operator interaction was by menu selection; 400 kilo­
rinsing station,lowered tothemaximum depth ofthe rinsing
bytesofon-linehelpand extensiveerrorcheckingduringdata
station(adepthmore thanadequate tocoverthecontaminated
entrywere includedtofacilitateuse ofthissoftware. TURBO
externalsurface),andrinsed.Becausetheneedleguidetrapped
PROLOG (Borland) was used for database operation and fluid,itwasremoved,andtheneedlewas stiffenedwithastain­
TURBO PASCAL (Borland) was used forhardware control.
less steelsheath.The needle tipwas always kept clearofthe
dBASE 3+ (AshtonTate)was used asthedownload database
liquidsurfaceofthedestinationtube.
softwarepackage.To obtainmaximum softwareflexibility,all
parameters,includingtubecountsforsamplesetsand standard
Table 1. Dispensing (1000 pL) accuracy of the
curvesaswellastubecontents,wereincludedinthemenu-op­
automated system for different displacement modes3
eratedsoftware. Otherlow-levelmachine operatingvariables,
such as rack locations, dimensions and layout, tube dimen­ Displacement Mean volume,
mode pLb SD, pL Diff., %
sions, dispensing and pumping speeds and times, aspiration
technique,bubbleseparatorvolume,andsyringevolume,were
Air 871.4 3.84 -12.9
allocated to set up fileswritten in ASCII textformat. These
Liquid, 100 pL
variableswere infrequentlyaccessedand were accessibleonly bubble 986.6 1.30 -1.3
to the laboratory manager. Software restrictedthe assay me­ Liquid, 20 pL
dium todouble strength,thenumber ofstandardcurvesto1or bubble 1003.3 0.86 +0.33
2 perrack(notzero),and standardand extractsource tubes to Liquid, no bubble 1001.3 0.13 +0.1
specifiedlocationswithintheappropriaterack.
Datafilesweregeneratedasrequiredduringtheassay,were a Dispensing controlled by Gilson 222 keyboard. Drop on needle tip
removed by either bubble blow out or Immersion.
heldindedicated subdirectories,and were availableforaudit­ b Four tubes dispensed for each mode.
ing.
1284 H a g g e t t E t A l .: J o u r n a l O f A O A C T n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

Table 2. Dispensing precision of the automated system using bubble separator

Values obtained, \AJ

Set values, |xL Standard Water Total6 Dilution factor6

Standard Water Diln factor X %CV X % CV X %CV X %CV

0 5000 0 0 _ 5000 0.16 5000 0.16 _ _

250 4750 0.05 248 2.0 4749 0.18 4997 0.17 0.0497 2.06
500 4500 0.10 496 1.1 4496 0.13 4992 0.15 0.0998 0.65
750 4250 0.15 745 1.4 4246 0.13 4991 0.18 0.1490 0.34
1000 4000 0.20 997 0.30 3996 0.12 4993 0.13 0.1997 0.20
1250 3750 0.25 1248 0.30 3745 0.11 4993 0.12 0.2500 0.22
1500 3500 0.30 1498 0.07 3496 0.10 4994 0.07 0.3000 0.08
1750 3250 0.35 1745 0.20 3246 0.13 4991 0.13 0.3497 0.17
2000 3000 0.40 2001 0.30 2998 0.10 4999 0.14 0.4003 0.17
2250 2750 0.45 2247 0.10 2749 0.12 4996 0.11 0.4497 0.07

a Six independent experiments. Volumes obtained were determined by mass measurement and density corrector. Means (X) and percent
coefficients of variation (% CV) for the 6 experiments are shown for each parameter. Means for total volume anc dilution factor were,
therefore, not directly calculated from the means of standard volume and water volume.
6 The total volume for each experiment was determined as the sum of the volumes of standard and water.
c The dilution factor for each experiment was determined as the ratio of standard volume to total volume.

Crosscontaminationbetweenassayswas eliminatedby the thesestepswithamanual mixingoperationbeforetheturbidity

followingprocedures.As describedpreviously,onlythediluent measurement by the automated system. The Gilson hardware
was allowed tocontact the syringe and valve assembly, thus then only had tore-mix thecontentsby simple bottom-to-top
restrictingthesurfacestobecleaned.The automationsoftware liquidtransfersimmediatelybeforemeasurement(15 sinstead
mainmenu includedasectiondevotedto2cleaningoperations of45 s).
tobeperformedafteranassayandalsoimmediatelybeforethe
startof the next assay. One of these operations continuously Turbidity M easurements
aspiratedcleaning reagent from thereservoirintothe syringe
and dispensedthesame viathecoiland needle.The otherop­ A singleneedleassemblycouldnotbeusedforbothsimple
eration continuously aspirated cleaning reagent through the mixing and deliveryofthetubecontentstotheflow-cellwith­
turbiditymeasurement liquidpath. outtheinstallationofanadditional2-wayvalve.Consequently,
adualneedleassembly(Figure6)was usedforturbiditymeas-
Dispensing a Microbiological A ssay

The arrangement of tubes used for the Gilson rack 22 is Column

showninFigure7.Theextractstobedispensedintotheseracks 1 4
werelocatedinthesampleextractrackshown inFigure4.The
sequenceofoperationsperformedby theGilsonhardwaredur­
ingtubedispensingwas asfollows:( I ) Add watertotheblank; Uninoculated tubes
(2) addstandardtothetubes;(3 ) addwaterto(2);(4 ) add sam­
a S ta n d a rd fo r curve /
pleextracttothetubes;(5)add waterto(4);(6)addbasal me­
b S ta n d a rd fo r curve 2
dium toalltubes.
c Blank, curve 1
d Blank, curve 2
R esuspension for Turbidity M easurement
e onwards
Beforetheturbiditymeasurementswereperformed,thepel­ C olour correction tubes
letoforganisms inthe bottom ofeach culture tube had tobe fo r each sa m p le s e t
resuspended. Experiments demonstrated thattheGilson hard­ as requ ired
ware couldbeusedtoresuspendapelletofmicroorganismsby
aseriesofaspirations(800pL) and rapidejectionstobreakup
the pellet,followed by liquidtransfersfrom thebottom tothe
top ofthe tube to mix the contents. However, the time taken
(45s)was consideredtobeexcessive(40tubesperrackwould Tubes a re p ro ces se d s ta rtin g a t colum n 1, row 1
require 30 min formixing alone, not including the measure­ a n d then down each colum n
ment step). Because “sterilization,” inoculation, and incuba­
Figure 7. Assay rack layout.
tion were already manual operations, we decided to follow
 H a g g e t t E t A l .: J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1285

urements. One needle of the dual needle assembly was con­

nected viathe aspirationcoiltothe401 diluterstainless steel
3-wayvalveandusedforthemixingoperationbybubblesepa­
ratorliquiddisplacement.The secondneedlewas connectedto
thephotocell,andtherefore,totheperistalticpump andusedto
draw the mixed suspension through the photocell, where the
transmittanceofthesuspensionwas measured.
During turbidity measurement for each tube, 50 readings
were taken overthereadingcycle.Repeated measurements of
thephotocelloutputvoltageweretakenby softwarecontrolof
the A/D interface. The mean resultwas then determined. To
performthereading,thetubecontentsweredrawn throughthe
flow-cellfor 10stoflushthepreviousfluid.The flowwas then
continuedatthesame rateforafurther15 swhilethe50 read­
ingsweretaken.Ataflowrateofabout360pL/s,about90% of
thetubecontentswere used.
After the transmittance voltages ofeach setoftubes were Silica Suspension, ml
measured,thephotocellwas flushedfor30 swithwaterdrawn
fromasupplyinthereservoirintheextractrack.Neartheend • Specrophotometer measurements,
oftheflushoperation, thetransmittancewas measured (blank o Gilson measurements.
reading)inthesameway asthesuspensions.Any changeinthis Figure 8. Calibration of automated turbidity
readingfrom thepreviousblankreadingwas canceledby cor­ measurement.
recting allintermediate readings. Inthisway, photocell/lamp
drift was eliminated. The transmittance voltages were then
converted to a suitabledesired absorbance scale by software When various culture suspensions derived from standard
usingthefollowingformula: curve tubeswere agitatedby vortex mixing and theturbidities
were measuredby spectrophotometer,an initial,relativelysta­
Vs
Turbidity/absorbance= -In( ~ )x F ble reading was obtained thatthen driftedcontinually untila
vb new stablereadingwas obtained(Table3).
Because agitationwould altertherelationshipbetween cell
where Vs = sampletransmittancevoltage; Vb = blanktransmit­ mass concentration and apparent turbidity, the effectof flow
tance voltage; and F = correction factor (dependent on cell was examined closely for the automated turbidity measure­
length). ment. Standardcurveswereplottedusingstationaryandflow­
ingsuspensionsofL a c to b a c illu s leich m a n n ii, L. c a se i, L p la n -
Turbidity M e a s u re m e n ts o f S ta tio n a ry F lo w in g S ta te s
ta ru m , and S a c c h a ro m y c e s c a rlsb e rg e n s is. Inno casewas the
Turbidity measurements were usually taken while thecul­ curve significantlyaffectedby flow. A period exceeding 15 s
ture was flowing parallel to the light path in the horizontal (usually 45 s) was considered necessary to achieve a stable
plane.Bubblesaccumulatedwhentheflow-cellwaspositioned reading in stationary mode. The variability of the readings
atan angle tothehorizontal.Verticalinstallationwas notsuit­ withineachsetof50voltagereadingswas typically± 0.5%for
ablebecause ofdifficultiesinclearingthecelloftheprevious
sample. Table 3. Typical turbidity measurements of agitated3
The calibration and linearity ofthe automated dispensing culture suspensions using the spectrophotometer
operation and turbiditymeasurement were routinely checked
Initial reading,6 Final reading,6 Duration,
by dispensing silica suspension instead of working standard Organism optical density optical density sc
solution and measuring the turbiditiesby both the automated
system and the spectrophotometer (Figure 8). The turbidity L. leichmannii 0.92 ± 0.02 0.98 ±0.04 120
wasproportionaltosilicaconcentrationuptoaturbidityof1.0, L. casei 0.83 ±0.01 0.93 + 0.01 55
which isbeyond therange ofatypicalmicrobiologicalassay. L. plantamm 0.495 ± 0.005 0.510 ±.0075 30
No effectofagitationorflow on the turbidityofthe silica S. carlsbergensis 0.485 ± 0.005 0.485 + 0.005 0
suspensionmeasured manually by spectrophotometerwas ob­
a Cultures were mixed by vortex mixer, and the tube was
served.Consequently,thedifferenceinlineslopebetween the immediately inserted into the cell holders.
2 measurement modes was attributedtothedifferencesincali­ b Both initial and final readings, although relatively constant,
bration of the 2 instruments used. The automated turbidity typically varied to the limits shown.
measurement was calibrated by the correction factorin soft­ c The times taken for the tubidity readings to transition from the
initial agitated state to the final quiet state are shown. For L.
ware, whereas the absorbance calibration of the spectro­ leichmannii, these times varied from 30 to 120 s. Visual
photometerwasrenderedinapplicablebythecellcompartment observation showed that suspension uniformity was not achieved
modification. until 7-10 min after mixing.
1286 H a g g e t t E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

both modes, butvariabilityroseto± 2% forL le ich m a n n ii in Table 4. Comparison of calculation methods using
flowing mode. The variabilityofthereadings, thetime taken reference and automated methods for vitamin B12
for the readings to stabilize afterflow was stopped, and the (pg/kg)a
differencebetween flowing and stationarymodes ofmeasure­ Reference method Automated method
ment were all less forthe automated turbidity measurement
thanforthespectrophotometerturbiditymeasurement.
Sample No. Result Std error, % Result Std error, %

C a lc u la tio n o f R e su lts 1 27.2 4.1 27.0 4.1

1 26.3 3.3 24.3 3.6
The automation software calculated results as follows: A 2 25.5 4.9 26.0 5.2
least-squares polynomial of selected degree was fittedto the 2 23.3 7.6 24.4 6.7
standard curve resultsto generate an equation ofturbidity or 3 31.7 2.3 27.1 5.2
opticaldensityversusvolume ofworking standard.Inturn,the 3 30.9 5.3 28.6 3.2
vitamin concentrationcorresponding toeach extractturbidity 4 27.1 7.1 27.9 5.0
was interpolatedfrom thecurve. The estimated concentration 5 22.3 5.7 23.5 5.4
ofthevitaminintheextractinthetubeand theestimatedcon­ 5 20.9 6.9 19.6 6.1
centrationintheoriginal sample couldthenbe calculated.Fi­ 6 37.8 5.4 38.3 7.5
nally,themean sample concentration and the standard devia­ 6 40.0 2.5 41.3 3.4
tionwere calculatedfrom allofthetubesinthesample set. 7 31.8 2.9 29.6 7.1
Some extractionproceduresrequiredthatthesample bein­ 8 24.9 8.2 23.4 7.3
cubated with enzymes to convert various vitamers to a form 9 26.7 4.0 24.7 2.4
suitableformicrobiological assay. These enzymes were usu­ 9 22.5 6.1 23.8 5.1
allycontaminatedwithlow levelsoftheassayvitamin.Conse­ 10 28.8 10.6 26.7 8.7
quently,thiscontaminationhad tobe correctedby thefollow­ 11 25.4 3.6 23.8 0.7
ingformula: 12 22.2 6.5 23.6 5.5

£>, a Samples assayed over a 3-day period.

Correction= Ex —
De

sample setthatcould notbe accounted forby thevarying ex­

where E = vitaminconcentration intheenzyme, D, = product tractconcentrationwould appearasdrift.
ofthe initialsequence ofdilution factors up toand including
C o m p a ris o n o f A u to m a te d a n d R e fe re n c e
the dilutionto which the enzymes were added, and D e = en­
P ro c e d u re s
zyme dilutionfactor.
When colorcorrectiontubeswere includedforthesample, C o m p a riso n o f m e th o d s u se d to c a lc u la te resu lts. — The tur­
thefollowingcorrectionwas appliedtoeachturbidityreading: biditiesofhand-dispensed assays (standardcurve and 7 sam­
ples)weremeasuredby usingtheCecilspectrophotometer.Re­
sultscalculatedfrom theseturbiditiesusingboththereference
software and the automation software are shown in Table 4.
The differencesinresultsand standarderrorswere notsignifi­
where C = absorbanceofthecolorcorrectiontube,B = absor­ cantatthe95% confidencelevel.
bance of the blank reading, V = volume ofthe extractin the C o m p a riso n o f m e th o d s u s e d to m e a su re tu r b id ity . — The
tube,and Vb - volume ofthebasalmedium. Such corrections turbiditiesofhand-dispensedassayswerefirstmeasuredby us­
were approximate,because samplecolorisusuallylostduring ing the Cecil spectrophotometer. The turbidities of the same
thegrowth cycleoftheassayorganism. assaytubeswere thenmeasuredby theautomatedsystem.Re­
Driftwascalculatedbytheautomationsoftwareastheslope sultsforboth setsofturbiditiescalculatedby using the auto­
ofthe firstorderregressionplot(straightline)ofthe relative mation software are shown inTable 5.The differences in re­
sample concentration ( C r) (dependent) againstsample extract sults and standard errors were not significant at the 95%
volume inmilliliters(independent),where confidencelevel.
C o m p a riso n o f m e th o d s u s e d to d is p e n s e re a g e n ts. — Hand-
^ sample concentrationestimate x 100
dispensed assays and assays dispensedby the automated sys­
mean sample concentration tem werepreparedfromthesame sampleextracts.The turbidi­
ties of both assays were measured, and the results were
Thiswas interpretedastheaveragepercentdriftpermillili­ calculatedbyusingtheautomatedsystem(Table6).The differ­
terofsample. ences inresultsand standarderrorswere notsignificantatthe
Driftwas usedtodeterminewhetherthegrowthresponseof 95% confidencelevel.
the organism tosample extractconcentrationwas ofasimilar
naturetogrowth response toworking standard concentration.
Consistentvariationofgrowth response indifferenttubes ina
 H a g g e t t E t A l .: J o u r n a l O? A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1287

Table 5. Comparison of the methods of turbidity measurement using reference and automated methods3

Reference6 Autorratedc Mean difference6

Assay (organism) X % SE X % SE X % SE
Folate ;L casei), pg/kg 544-876 0.9^t.9 575-873 1.0-8.0 +11 +0.1
B12 (L leichmannii), pg/kg 12.9-39.5 2.8-10.7 13.8-44.4 2.6-9.3 +1.1 -0.1
Panto. (L. plantarum), mg/kg 32.5-93.1 2.2-14.9 34.0-98.9 2.4-12.4 +1.2 -0.4

3 Results obtained for 15 infant formulas for each assay repeated on different days; SE = standard error.
b Turbidities measured with the Cecil spectrophotometer.
c Turbidities measured with the automated system.
d Automated result minus reference result.

Discussion The difference in turbiditiesofL leic h m a n n ii and L

g e n sis.
ca seiwas largeenough for2 stablestatestobeobserved: fully
Full automation of MBA for vitamins, although an obvi­ aligned organisms flowing in a horizontal circular path, and
ouslydesirableobjective,requiresacomplex,versatilerobotic fullyrandomized organisms. Because theformer statecannot
systemthatcancopewiththevariationsinextractionmethods be maintainedlongenough forusefulreadingstobetaken,the
thatare required. Steps such as carefulpH adjustment, filtra­ analyst isforced to wait for the latter state to be stabilized.
tion,volumetricdilution,andautoclavingarebeyondthecapa­ Therefore, cultures must be resuspended by vortex mixing
bility of a diluting, dispensing, sampling system such as the some timebeforemeasurement.When thetubeisplacedinthe
Gilsonpackage.Nevertheless,theuseofthishardwaretogether spectrophotometer,randomizationmay takeabout30-120 sto
withthecomprehensiveautomation softwaredoes achievethe be re-established (L. le ic h m a n n ii willtake longerbefore uni­
desired objectives of eliminating the labor-intensive, error- formturbiditydevelopsthroughoutthesuspension).Also,ther­
prone portions ofthe assay atmoderate cost and without the mal effectsfromthelightbeam canmean thatareadingcanbe
need to reorganize the laboratory to accommodate the larger taken of a convection stream of partially aligned organisms
circularfootprintofafullroboticssystem.However,theGilson withouttheoperatorbeingaware ofit.
packagecouldbe usedasaunitwithinalarger,more versatile Incontrast,when thefullyalignedflowing statewas estab­
roboticssystem, ifsodesired. lishedintheautomatedsystemflow-cellandthelightwas par­
A very satisfactory improvement in both precision and alleltotheorganism axis,no significantdifferences inmeans
speed was achieved with turbidity measurement. When rod­ or errors between flowing and randomized states were ob­
shaped organisms are agitated, flow birefringence is set up, served.The reasonsforthislackofdifferenceareprobablyre­
which causes drifting turbidity readings (9). Our work has latedtoflow-cellsizeand shape aswell asthelightbeam ori­
shown that, as expected, the effect is absent with both S. entation.The factthatthedifferenceissmallfortheautomated
c a rlsb e rg e n s is and spherical silica particles, presumably be­ systemmeans thattherelationshipbetween cellmass concen­
cause ofthe spherical-to-ovoidshapes. With theL a c to b a c illi, trationandturbidityisnotstronglyaffectedby flow,regardless
all of which are rod-shaped, the effect was particularly pro­ ofthemorphology oftheorganism. Inturn,thismeans thatthe
nounced when thesuspensions were agitatedby vortexmixer cell morphology can be differentbetween cultures grown in
andmeasuredby thespectrophotometer.The lightbeam was at standardand sampleextractswithouttheturbidityreadingbe­
rightanglestothelongitudinal axisoftheorganism when the ing significantly affected by the use offlowing mode. Com­
suspensionwas flowinginthevortexpattern.Turbiditymeas­ parisonsoftheresultscalculatedfrom turbiditiesmeasured by
urementvariabilitydiminishedfortheorganismsinthefollow­ bothspectrophotometer(stationary)and theautomatedsystem
ing order: L. leic h m a n n ii, L c a sei, L p la n ta ru m , S. c a r ls b e r ­ (flowing)confirmedthattherewasnosignificanteffectofflow.

Table 6. Comparison of methods of dispensing reagents using using reference and automated methods3

Reference6 Automated6 Mean difference6

Assay (organism) X % SE X % SE X % SE
Folate (L. casei), pg/kg 544-720 0.2-11.5 500-710 0.2-9.6 -5 -0.2
B,2 (L. leichmannii), pg/kg 16.5-26.5 1.4-5.5 15.8-29.0 1.6-8.6 -0.3 +1.4
Panto. (L. plantanjm), mg/kg 34.0-98.8 2.4-12.4 36.2-104.6 0.0-11.4 +1.4 -1.6

3 Results obtained for 13-15 infant formulas for each assay repeated on different days.
b Assays dispensed by hand.
c Assays dispensed by the automated system.
d Automated result minus reference result.
1288 H a g g e t t E t A l . : J o u r n a l O f AO A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

The precisionoftheflowingmode isconsiderablybetter(typi­ jectcode areavailablefrom thesame source.The objectcode

cally 1%) than the manual randomized mode measurements sizeis1.2Mbyte heldinone subdirectoryandrunsunderDOS
(typically3-12%). Thisprecision,coupledwiththemorerapid 3.2 on anIBM orIBM clonecomputer. A 20Mbyte harddrive
readingtime,means thattheautomatedturbiditymeasurement isneededforsoftwareand data-filestorage.
isfarsuperior.
Even thoughautomateddispensingisnotquickerthanman­ Acknowledgments
ual dispensing, itdoes allow the operator to perform other
tasks. In addition, automated dispensing ismore precise and We thanktheworkshop, laboratory,and supportstaffofthe
alsoallowsmore flexibilityintheallocationoftubesforstand­ New Zealand Dairy Research Instituteforalltheirassistance,
ardsandextracts. andJohnMorrisScientific,Auckland,New Zealand,forinfor­
Inherently, a new softwarepackage may contain well-hid­ mationon theGilsonhardware.
den errors,and theseerrorsmay create occasionalbias inthe
results.However, by comparing oldandnew calculationmeth­ References
odologies,we have notfound any such errorsinthe software.
These same comparisons have alsoshown that,despitethere­
(1) Snell, E .E ., & Strong, F.M. (1939) Ind. Eng. Chem. Anal. Ed.
ductionoferrorinthemeasurementfunctions,theoverallassay
11, 3 4 6 -3 5 0
error isunaltered. This finding clearly demonstrates thatthe
major source oferrorinthe MBA isthe day-to-day variation (2) Official Methods of Analysis (1990) 15th Ed., A O A C , A r­
lington, VA
relatedtogrowthresponseandnotmeasurementerror.
However, the precisely repeatable nature of the machine (3) Finglas, P.M., Faulks, R .M ., M om s, H.C., Scott, K.J., & M o r­

functionsdoes make iteasyfortheoperatortospota system- gan, M .R .A . (1988) J. Micronutr. Anal. 4, 4 7 -5 9

relatedproblemandalsoallowsthelaboratorystafftofocuson (4) Favell, D.J. (1990) J. Micronutr. Anal. 7, 2 3 7 -2 4 6
more majorsourcesoferror.Checkingforerroneousdataentry (5) Lum ley, I.D ., & L aw rance, P.R. (1990) J. Micronutr. Anal. 7,
is faciiitated by an audit trail, which is updated every time 301 -3 1 3
changes aremade, and adatabase,which can facilitatethede­ (6) Bell, J.G . (1974) Lab. Practice, 23, 2 3 5 -2 5 2
tectionofanomalous results. (7) Methods of Vitamin Assay (1985) A ugustin, J., K lein, B.P.,
The system has been operationalforover 3 years and has Becker, D., & V enugopal, P.B. (Eds), 4th Ed., W iley-Inter-
proved reliable; only minor maintenance to tubing and me­ science, N ew York, N Y
chanicalpartsisrequired.The speedofthesystemwhen proc­ (8) B hattacharyya, G .K ., & Johnson, R.A. (1977) in Statistical
essing pre-extracted samples issimilartothatofa competent Concepts and Methods, John W iley & Sons, N ew York, NY,
technician. p. 2 1 3 -2 1 5
The plans for modification of the Gilson hardware and a (9) Analytical Microbiology (1972) K avanagh, F. (Ed.), 2, A ca­
demonstrationdiskforthesoftwareareavailablefromtheNew dem ic Press, N ew York, NY, and L ondon, E ngland, pp.
ZealandDairyResearchInstitute.The fullsourcecode and ob­ 4 9 -5 2
 Li E t A l .: J o u r n a l O f A O AC In t e r n a t io n a l V ol . 76, N o. 6 ,1 9 9 3 1289

FOOD COM POSITION AND ADDITIVES

D e t e r m in a t io n o f M e n t h o l i n H o n e y b y G a s C h r o m a t o g r a p h y

M aojuan Li,D onald L.N elson,1

*and Peter Sporns 2
University ofAlberta,Food Science Department, Edmonton, AB, T6G 2P5, Canada

A gas chromatographic method was developed for Despite the growing use ofL-menthol by North American
the determination of L-menthol in honey at levels beekeepers, thereisvirtuallyno published scientificliterature
as low as 0.1 ppm. The method includes steam dis­ on analysismethods, stability,orthetastethresholdofL-men­
tillationand hexane extraction with an internal thol inhoney. The goal ofthis work was toinvestigatethese
standard (2,6-dimethylphenol). Beehives treated to issues.
control A c a r a p i s w o o d i over 21 days with 30-60 g
L-menthol contained L-menthol residues in honey Experimental
and beeswax. L-Menthol was found only intreated
portions ofthe hive and not latertransferred to H o n e y S a m p le s
added honey supers. The highest levels of resi­ GB honeywas obtainedfrom an Edmonton beekeeperwho
dues in honey (18 ppm) and beeswax (2790 ppm) used no chemical compounds in any part of his beekeeping
were found when pure L-menthol was adsorbed operation.The othertestsampleswere obtainedfrom avariety
into foam strips placed on top of the hives. L-Men- ofmenthol hivetreatmentssummarized inTable 1.
thol residues in honey were not reduced by storage
in open containers at room temperature for up to R e a g e n ts

55 days. Untrained panelists could not detect l - Allwaterwas purifiedby usingaMilliporeMilli-Qsystem

menthol in honey untilthe levels approached (Millipore Corp., Bedford, MA). Unless otherwise noted, all
36 ppm. otherreagentsused were reagentgradeorbetter.
(a) S ta n d a r d 2 ,6 -d im e th y lp h e n o l (2 ,6 -D M P ) so lu tio n (1 0
p p m ). — Dissolve 0.0020 g 2,6-DMP (Aldrich Chemical Co.,
I n 1984, a new mite,A c a r a p is w o o d i, widely known asthe Milwaukee, WI) inhexane ina 200 mL volumetric flaskand
Tracheal mite, appeared inAmerican beehives(1).Before make to volume. Protect the solution from lightand prepare
thisdate,themitehadnotbeenfoundinNorthAmerica,but freshevery2 days.
Europeans had longexperiencewiththispest.The miteenters (b) i-M e n th o l sta n d a r d so lu tio n (2 0 p p m ) . — Dissolve
the breathing tubes or trachea of bees, where it sucks the 0.0010 g L-menthol (Aldrich Chemical Co.) ina50 mL volu­
hemolymph, or“bee fluid,” layseggs,and multipliesuntilthe metric flaskwithstandard2,6-DMP solutionand make tovol­
beeissoweakened thatitcandie.Although problematicinall ume.
seasons,themiteisofparticularconcerninover-winteredbee
S ta n d a rd C u rv e
colonies.Inwinter,bees clusterforwarmth, allowingthemite
to easily move from bee to bee. Along with the other stress Standardsrepresenting0.1,0.5, 1.0,2.5,5.0,and 10.0ppm
factorsofwinter,theresultingwinterkillinTracheal mite-in- L-menthol were preparedby adding 0.125,0.625, 1.25,3.125,
fectedhivescanbe disastrous(2,3). 6.25, and 12.5mL L-menthol standardsolutionto25 mL volu­
Soon afteritbecame apparent that the mite could not be metricflasks.The flasksweremade uptovolumewithstandard
eradicated,scientistsintheUnitedStatesandCanada began to 2,6-DMP solution.
searchformethodsoftreatmenttocontrolTrachealmiteinfec­ To 10mL ofeach oftheabove standards,3.5g sodium sul­
tionsinbees.One compound thatexhibitedgreatpromisewas fate(Anachemia, Montreal, PQ) was added. The mixture was
L-menthol (1,4-6), which occurs naturally in honey derived shaken vigorously and filtered.The remaining sodium sulfate
from mint plants.Because use ofL-menthol inhivesproduces was washed with 5 mL hexane. The hexane wash and thefil­
residuesinhoney,thecompound isconsideredafoodadditive tratewere combined and evaporated under a gentle stream of
(7).The United Statesand Canada recentlypermitted theuse nitrogen (Linde, Edmonton, AB) untilthe solutionreached a
ofL-mentholby beekeeperstocontroltheTrachealmite. volumeofca1mL. Approximately 1.0|iLofeachconcentrated
solutionwas injectedintoagaschromatograph (GC).
R e c e iv e d J u ly 2 7 , 1 9 9 2 . A c c e p te d O c to b e r 15, 1 9 9 2 . G a s C h ro m a to g ra p h ic (G C ) S y s te m
1 A g ric u ltu r e C a n a d a R e s e a rc h S ta tio n , P O B o x 2 9 , B e a v e r lo d g e , A B ,
TOH 0C 0, C anada. The Varian Model 3700 gas chromatograph (Georgetown,
2 T o w h o m c o rre s p o n d e n c e s h o u ld b e a d d re s s e d .
Ontario, QC) was equipped with a flame ionization detector.
1290 Li E t A l .: J o u r n a l Of ADAC I n t e r n a t io n a l V o l . 76, No. 6,1993

Table 1. Honey and hivetreatments

Honey
Code Treatment3
2.1 cm

A Control beehives, no menthol applications

B 60 g L-menthol in a 50:50 paste with vegetable shortening
(60 g) applied on top of notched (v-grooved) cardbord and
placed on the bottom of the hive for a total of 21 days
C 6 foam strips each containing 10 g pure L-menthol applied
on the top of the hive for 21 days
D 3 foam strips each containing 10 g pure L-menthol applied
on the top of the hive for 21 days
E 30 g L-menthol in a 50:50 paste with vegetable shortening
(30 g) applied on top of notched cardboard and placed on
the bottom of the hive for a total of 21 days
F 30 g L-menthol in a 50:50 paste with vegetable shortening
(30 g) absorbed in liquid form into cardboard placed on the
bottom of the hive for a total of 21 days. These “abusive
treatment” hives had a second similar menthol treatment
after 31 days. This second application was left from day 31
to day 55 and followed by a third similar treatment from day
55 to day 89 (August 14,1991) o
vi
All treatments began on May 17,1991, and all but F hives had the
treatment removed after21 days (June 7, 1991).

The majorcolumnusedwas a30m x 0.25m m DB-5 capillary

column (J&W Scientific,Folsom, CA) protectedwith a0.4m Figure 1. Steam distillation/extractionapparatus
x 0.25 mm deactivated silicacapillary column (J&W Scien­ adapted from Veith and Kiwus (8).
tific).Some analyseswerealsoperformedusinga30m x 0.25
mm Cyclodex B capillarycolumn (J&W Scientific).The car­
riergaswas UHP helium(Linde)passedthroughoxygentraps
flask with 100 mL water. Standard 2,6-DMP solution (10.0
(Chromatographic Specialties,Inc.,Brockville,ON). The flow
mL) was added tothe top ofthe steam distillation/extraction
rate was 26.6 cm/s. Extracts were introduced with a splitter
apparatus.The apparatus was placed on topoftheround-bot­
ratio of 42:1. Temperature program was as follows: initial,
tom flask,andtheflaskandcontentswereheated.The solution
80°C, hold 1min; programming rate,5°C/min; final, 120°C,
was distilledfor20min fromthetimeitfirstbegantoboil.The
hold 10min. Injectorand detectortemperatureswere 140 and
hexane solutionwas removedfromtheapparatus,andanaddi­
280°C, respectively.Chromatographic datawerecollectedus­
tional 10mL hexanewas usedtothoroughlywash theappara­
inga 3388Aintegrator(Hewlett-Packard,Avondale, PA).
tus.The hexane portionswere combined and driedover 3.5 g
S te a m D istillation/E xtraction A p p a ra tu s sodium sulfate.The textextractwas concentratedtoca 1mL
undernitrogen.An aliquotoftheconcentratedextractwas in­
The steam distillation/extraction apparatus was firstintro­ jectedintothegaschromatographforanalysis.
ducedbyVeithandKiwus (8),andourvariation(Figure 1)was Residual honey was removed from the wax with a rapid
producedby theGlassblowingShop oftheChemistryDepart­ washusing50mL water.The wax wasdriedwithpapertowels.
mentoftheUniversityofAlberta.A similarapparatuscanalso To extractthementhol, theprocedure used forhoney was re­
be purchased (Kontes, Vineland, NJ; Part No. K-523026- peatedfor0.4-2 g wax in200 mL water,themixture was dis­
0000). tilledfor30 min.
S a m p le P re p a ra tio n
S e n s o ry E valuatio n
Honey frames were processedimmediately astheyarrived
in the laboratory. Honey and surrounding wax (ca 10 square Honey was prepared forevaluation 2 days before use for
cm) were removed and placedin4 layersofcheesecloth.The sensorytesting.GB honey (300g)was addedtoa 1L beaker,
cheesecloth and contents were bundled and secured with a and the top of the beaker was covered with parafilm. The
clamp overalargefunnel.Honey andwax wereemptiedintoa beaker was kept in a 50°C water bath for 1 h to liquefy the
250 mL screw-topjar.Afterallofthehoney possiblewas col­ honey.The honey was thenweighedinto2jarsas 100and200
lected(ca2h),honey andwax were removedfromthecheese­ gportions.A known amountoffoodgradeL-menthol(supplied
clothandstoredseparatelyinsealedjarsat-20°Cuntilanalysis. courtesyofH&R, Springfield,NJ) was added to 1jar(usually
ForL-mentholanalysisofhoney, 10.0ghoneywas weighed thejarcontaining 100ghoney).Thejarwas thensealedimme­
intoa500mL round-bottomflask.Honey was washed intothe diately.Bothjarswereplacedbackintothe50°C waterbathfor
 Li E t A l .: J o u r n a l O f A O A C In t er n a tio n a l V o l . 76, N o. 6 ,1 9 9 3 1291

1 h and then shaken vigorously for 10 min. The honeys were for treating the bee colonies and for the analysis of a few initial
kept at room temperature overnight and analyzed for L-menthol honey samples.
the next day. To overcome some of the analytical problems noted above,
Honey used for individual sensory evaluations was pre­ a completely new procedure for menthol determination in
sented in ca 3 g portions in 7 mL vials and served with coffee honey was developed. This new procedure involved the use of
stirrers. Lemon juice (1%) in plastic cups was supplied for use steam distillation and extraction using the apparatus designed
as a rinse. All sensory evaluations were conducted in individual originally by Veith and Kimus (8 ). The solvent, which now
booths equipped with a sink, from noon to 12:30 p.m. or from served to extract the steam volatile materials in honey, was
1:30 to 3:30 p.m., under a yellow light. All apparatus was dis­
changed to hexane, and the GC capillary column was changed
posed of between analysis sessions. To avoid interference, pan­
to the chemically bonded DB-5 (J&W Scientific). With these
elists were not to chew gum 1 h before testing. All triangle test
modifications, the methodology was thoroughly investigated
protocol was as described in Poste et al. (9).
(Figure 3).
We were concerned that the sodium sulfate used for drying
R e s u lt s a n d D i s c u s s i o n
the hexane may retain menthol or 2,6-DMP; therefore, 3 sepa­
rate hexane solutions containing 5.0 ppm L-menthol and 10.0
Because government regulations (10,11) often use 0.1 ppm ppm 2,6-DMP were examined. Portions of each solution (10
as a tolerance level for unregistered agricultural chemicals, it
mL) were treated with 3.5 g sodium sulfate and the remaining
was desirable to have a honey analysis method for L-menthol
hexane solutions were retained without a drying treatment. In­
with a similar detection limit. To achieve this detection limit,
jection of each solution gave a peak height ratio (PHR, L-men­
some extraction and concentration steps were necessary even
thol peak height divided by 2,6-DMP peak height) of 0.599 ±
though sample loss was possible at each step. An internal stand­
ard was used to allow for such losses and to eliminate the need 0.003 (standard deviation) for the dried solutions and 0.594 ±
for volumetric control throughout the isolation procedure. We
experienced difficulty in choosing an appropriate internal
standard for honey analysis, because the internal standard had
to be similar to menthol but not occur naturally in honey. A
remarkable diversity of compounds are found in honeys ( 1 2 ).
The compound chosen was 2,6-DMP. To our knowledge, this
compound has never been reported in any honeys. 2,6-DMP is
a stable solid with physical properties and solubility (molecular
weight 122, mp 46-48°C, bp 203°C) that are very similar to
those of L-menthol (molecular weight 156, mp 41^f3°C, bp
212°C), and it has been used before as an internal standard for
quantitating volatile compounds in honey (13).
Because researchers had reported successful extraction of
menthol from honey using acetone (14), this solvent was inves­
tigated first. Also, because only L-menthol (15), in contrast to
its enantiomer D-menthol, was known to be effective for mite
control, the initial GC column chosen for separation and quan­
titation of menthol was the Cyclodex B capillary column (J&W
Scientific). This column contained methylated B-cyclodextrin,
which was known to separate terpene enantiometers (16).
However, there were several problems with this combination
of conditions. Although the GC separation of menthol and the
internal standard was obtained (Figure 2), the acetone was dif­
ficult to successfully dry after extraction and difficult to keep
dry as it was concentrated, even with the liberal use of sodium
sulfate. The best detection limit (at least 3 times background
noise level) that we were able to obtain for L-menthol in honey Figure 2. GC separation of a standard solution of
was 0.5 ppm. Although standards (Figure 2) allowed for a menthol enantiomers in acetone on a Cyclodex B
lower detection limit, other materials extracted from honey re­ capillary column. Axis is time in min. GC conditions are
sulted in considerable background noise. Most problematic, as stated in the experimental except that temperature
however, was the rapid deterioration of the expensive Cy­ programming was as follows: 100°, hold for 1 min;
clodex B capillary column with injection of honey samples. increase 1°C/min to 120°C, hold 10 min. L = L-menthol
The deterioration was noted by a loss of resolution of the men­ (5.0 ppm); D = D-menthol (5.0 ppm); s = internal
thol enantiomers and an increase in tailing. This analysis pro­ standard, 2,6-dimethlyphenol (5.0 ppm); x = acetone
cedure was used to confirm the authenticity of L-menthol used impurities; a = 2-fold increase in attenuation.
1292 Li E t A l .: J o u r n a l O f A O A C I n t er n a tio n a l V o l . 76, N o, 6 ,1 9 9 3

Table 2. Peak height and peak area ratios for L-menthol

and 2,6-DMP standards
L-Menthol concn, ppm P eak height ratio3 P eak area ratio3

0.1 0 .0 1 4 (0 .1 4 0 ) 0 .0 1 3 (0 .1 3 0 )
0.5 0.0 96 (0.192) 0 .0 7 3 (0 .1 4 6 )
1.0 0 .1 6 3 (0 .1 6 3 ) 0 .1 3 4 (0 .1 3 4 )
2.5 0 .3 5 2 (0.141) 0.281 (0.112)
5.0 0 .6 3 5 (0.127) 0 .5 4 0 (0 .1 0 8 )
10.0 1 .1 1 6 (0 .1 1 2 ) 0 .9 5 5 (0 .096 )

2 ,6 -D M P constant at 10.0 ppm. The value in parentheses is the

slope (ratio/L-menthol concentration) for individual L-menthol
concentrations.

shows that there was a slight curvature to the peak ratios rela­
tionships. However, for simplicity, all quantitations were deter­
mined using the above relationships. Because numerous cor­
rections for varying peak heights and areas also require an
accurate accounting for volumes throughout the procedure,
making these corrections would defeat many of the advantages
of using an internal standard. Systematic errors introduced by
using a linear relationship were small. In the following L-men­
thol determinations, any extract containing more than 1 0 ppm
menthol was volumetrically diluted with the appropriate
amount of 10.0 ppm 2,6-DMP hexane solution to ensure that
the peak ratio was within the 0.1-10 ppm L-menthol range.
To use peak height measurements for quantitation of L-men­
thol, the variation in retention times should be minimal (longer
retention times result in smaller and broader peaks). With this
caveat in mind, the retention times of both L-menthol and 2,6-
Figure 3. GC separation of menthol found in honey
from a honey super (hive F) after 89 days. Axis is time in
DMP were continually monitored. In one set of experiments,
min. A DB-5 capillary column and conditions as stated in using 5 injections of both standard solutions and spiked honey
Experimental were used, m = menthol (0.4 ppm); s = extracts, the retention times for L-menthol and 2,6-DMP were
internal standard, 2,6-dimethlyphenol (10.0 ppm). 9.00 ± 0.10 and 7.35 ± 0.10 min, respectively.
Steam distillation of GB honey obtained from an Edmonton
beekeeper who used no chemical compounds in any part of his
0.012 for the undried solutions. Therefore, there was no evi­
beekeeping operation gave no interfering peaks for either l -
dence of loss of soluble material due to drying of the hexane
menthol or 2,6-DMP. To investigate how much steam distilla­
solutions.
tion was required to remove all menthol in a honey test solu­
After extraction, the hexane solution was dried with sodium
tion, GB honey was spiked with 100 ppm L-menthol. In
sulfate and then concentrated under a stream of nitrogen. Dur­
separate experiments, after 10 and 15 min of distillation, 93.9
ing concentration, either L-menthol or 2,6-DMP may be lost.
and 102.2% menthol was recovered, respectively. All honey
Also, the detector response may not be linear with the variation
steam distillations were carried out for at least 2 0 min to ensure
in peak height for different L-menthol concentrations. Both of
complete recovery of L-menthol, because, in some cases, honey
these possibilities were investigated by preparing solutions is known to contain small amounts of crystalline L-menthol, as
containing 10.0 ppm 2,6-DMP and from 0.1 to 10.0 ppm l - well as soluble L-menthol.
menthol in hexane. A 10 mL aliquot of each of these solutions Repeatability of the analysis was investigated with 3 sepa­
was dried by using sodium sulfate and concentrated to about 1 rate honey test samples spiked at 3.3 ppm L-menthol. By using
mL before injection into the GC system (Table 2). The PHR (y) the above calibration curve for PHR and PAR, the L-menthol
vs the L-menthol concentration in ppm (x) was fitted to the content was 2.93 ± 0.07 and 3.09 ± 0.20 ppm, respectively.
equation y = mx, and the slope was calculated using least- The method was then used to evaluate the menthol levels of
squares estimate. The slope (m) was 0.12 with a coefficient of honey from hives treated in different ways. Honey from all
determination (r2) of 0.994. Relative areas of the 2 peaks were hives was initially analyzed for menthol (background levels),
also compared (i.e., peak area ratio, or PAR) and found to have and no menthol (0.5 ppm detection limit) was found in these
a slope of 0.10 and a coefficient of determination of 0.994. Al­ samples. After 21 days (end of the menthol treatment for all
though both sets of data fit a straight line relationship, Table 2 except “abusive treatment” F hives) more boxes, which are fre­
 Li E t A l .: J o u r n a l O f A O A C Intern a tio n a l V o l . 76, N o. 6 ,1 9 9 3 1293

quently called honey supers, were added to each hive for honey hives. Also, higher menthol levels in honey in the brood cham­
production. Honey from the original box (brood super) and the ber resulted from foam strips being placed on top of the brood
added honey supers (Figure 3) were analyzed. The results of sections for menthol treatment rather than from menthol paste
this testing are presented in Table 3. The efficacy and effect of being placed at the bottom of the hive.
the menthol treatments on the bees in the hives will be reported Because beekeepers reuse the wax frames obtained from
elsewhere, but it was interesting that all menthol treatments hives after the honey is removed, and because hydrophobic
resulted in dramatic reductions of Tracheal mite infestation menthol is more soluble in wax than honey, some wax analyses
compared with that o: the control (A hives). were also performed. For analysis of beeswax, the steam distil­
The only unusual honey was 1 sample from the untreated lation procedure was carried out for 30 min, rather than only
2 0 min.
control hives A at the end of the menthol treatment (day 21).
This honey contained 0.3 ppm menthol. Because no menthol The results of the beeswax analysis and a comparison with
was used with this hive, we assume that this single anomaly that of accompanying honey are presented in Table 4. Again,
was due to contamination of the honey in collection, transpor­ an anomaly was apparent; a small amount of menthol (0.5
ppm) was detected in wax from hive A (no menthol detected in
tation, sample preparation, or analysis. Generally, honey from
the corresponding honey). We believe that this further con­
the sections of the hive where honey is collected in commercial
firmed contamination of the samples collected from hive A on
honey operations (the honey supers) are free of detectable men­
day 21. It was interesting that with all of the errors in analysis,
thol, with the exception of small amounts of menthol (averag­
especially for honey low in menthol, that a fairly consistent
ing only 0.2 ppm, Figure 3) found in the abusively treated F
ratio (about 70 to 1) of wax menthol compared with honey
menthol was noted. This ratio doubled or even tripled for the
Table 3. Menthol levels in honey from treated hives
last 3 samples from hives C and D, which contained the men­
D ays after thol in foam strips placed on top of the brood nest. The higher
initial Honey No. Av. R ange of
menthol ratio for these last 3 samples probably occurred be­
menthol ocatlon in samples menthol menthol
Hive treatment hive analyzed level, ppm evel, ppm cause the foam strips are handled by the bees as they attempt
to wall off or remove the source of the menthol. The bees did
A 21 brood 3 0.1 N D -0.3® not approach the menthol paste placed on the bottom of hives
B 21 brood 2 1.8 1 .7 -1 .8 B, E, and F. In handling the foam strips, the bees may have
C 21 brood 4 6.2 3 .0 -9 .0 distributed menthol around the brood frames. In one case, we
D 21 brood 8 4.5 1 .4 -1 8 .0 could actually detect the white menthol crystals in a sample.
E 21 brood 2 1.2 0 .8 -1 .6 The menthol crystals probably stick to wax better than to
F 21 brood 5 0.8 0 .4 -1 .4
honey, elevating the apparent menthol in these samples.
A 55 brood 2 ND ND Because menthol treatment of hives obviously results in
B 55 brood 2 0.3 0.3 some residue in honey, it was useful to determine if menthol
C 55 brood 2 0.6 0 .5 -0 .7 levels in honey decreased with time. Rivera and Wilson (17)
D 55 brood 2 0.5 < 0 .1 -0 .9 *
E 55 brood 2 <0.1 <0.1 Table 4. Menthol levels in individual beeswax
F 55 brood 2 0.6 0 .5 -0 .7 compared with that in honey from same hive and
treatment
A 55 honey 2 ND ND
B 55 honey 2 ND ND D ays after Menthol concn, Ratio of
initial Honey ppm menthol,
C 55 honey 2 ND ND
menthol concn
D 55 honey 2 ND ND Hive treatment hive Honey W ax
wax/honey
E 55 honey 2 ND ND
F 55 honey 2 ND ND A 21 brood ND® 0.5 __

E 55 brood < 0 .1 * 6.9 —

A 89 honey 2 ND ND
E 55 honey ND ND —
B 89 honey 2 ND ND
B 89 honey ND ND —
C 89 honey 2 ND ND
B 21 brood 1.7 106 62
D 89 honey 2 ND ND
F 89 honey 0.4 30 .7 77
E 89 honey 2 ND ND
F 55 brood 0.7 60 .4 86
F 89 honey 2 0.2 < 0 .1 -0 .4
C 21 brood 9.0 1790 199

A 89 brood 2 ND ND C 21 brood 8.8 1550 176

F 89 brood 2 0.6 0 .4 -0 .8 D 21 brood 18.0 2790 155

3 ND = none detected. 3 ND = none detected.

b <0.1 ppm m eans that menthol seem s to be present but below any b <0.1 ppm m eans that menthol seem s to be present but below any
level for which quantitation should be used. level for which quantitation should be used.
1294 Li E t A l .: J o u r n a l O f A O A C In tern a tio n a l V o l . 76, N o. 6 ,1 9 9 3

T a b le 5. M e n th o l le v e ls o f 100 p p m in h o n e y s to re d a t Table 6. Triangle tests to determine L-menthol taste

ro o m te m p e ra tu re threshold levels

Menthol, ppm No. of

L-Menthol panelists
D ay Closed container O pen container in 2 i-M enthol correctly Probability of
samples, in single No. of identifying selection of
ppm sample, ppm panelists odd sample odd sample
0 98 100
7 98 97
96 93
0 20 .7 13 3 0.861
16
100 100 0 36.2 15 9 0.031
36
0 54 .5 14 10 0.0 04
55 90 97

38 .3 0 9 3 0.6 23

reported that levels of up to 5 ppm menthol in honey were re­

moved by heating honey (presumably open to the air) for 2 Overall, however, it seemed as though even the highest l -
days at 65°C. These high temperature conditions are unlikely menthol containing honey that we found in our experimental
to be used for commercial honey, because extensive heating is hives (18 ppm) would not be detected by the average consumer.
expensive and destructive to honey (12). To investigate if l - When panelists are trained to detect L-menthol in a 5% sucrose
menthol may be lost or changed chemically in the relatively solution, they can detect it at levels as low as 0.5 ppm (18). The
acidic (average pH 3.9 [12]) honey at more moderate tempera­ sweetness of honey (approximately 80% sugar content) does
tures, the GB honey was spiked with 100 ppm L-menthol and seem to mask the flavor of L-menthol (noted before for other
left for 55 days at room temperature in both open and closed volatiles; 19), but if panelists were trained to detect L-menthol
containers (Table 5). Over 55 days, variation in menthol levels in honey, they may decrease the detection lim it somewhat from
in both containers seemed to be attributable solely to sample the 36 ppm level that we have observed.
and analysis variation. There was no evidence that menthol va­
porized or changed chemically to any large extent. A cknow ledgm ents
We were also concerned that after a certain threshold level
of L-menthol in honey, consumers would detect the difference This work was supported by the Alberta Farming for the
in flavor. To determine the threshold level of L-menthol in Future program. Assistance was provided by Preben Kristian­
honey, tests were carried out with volunteer, untrained panel­ sen and Denis McKenna (Fairview College, Alberta) in over­
ists. The triangle test was chosen because fewer correct choices seeing the beehive experiments. All L-menthol hive treatments
and, therefore, fewer panelists were required for a significant were prepared by Willy Baumgartener (Medivet Pharmaceuti­
difference between samples; there is only a Vi probability (vs cals Ltd, High River, Alberta).
Vi in paired comparison tests) of choosing the correct sample
by chance (17). The results of these experiments are given in R eferen ces
Table 6 .
(1) H erbert, E.W ., Jr, Shim anuki, H.. & M attenius, J.C., Jr
The level of L-menthol spiking of all honey was confirmed
(1987) Amer. Bee J. 127, 7 7 6 -7 7 8
by analysis. At the 20.7 ppm L-menthol level, the number of
(2) O tis, G.W. (1990) Amer.
Bee J. 130, 28-31
panelists correctly identifying the single L-menthol containing
(3) Lord, W. (1990) Amer. Bee J. 130, 456
honey was about what one would predict by chance. However,
(4) Cox, R.L., M offett, J.O ., W ilson, W.T., & EUis, M . (1989)
at the 36.2 ppm L-menthol level, a significant number (prob­
Amer. Bee J. 129, 547 -5 4 9
ability of 0.031, Table 6 ) of panelists correctly chose the L-men-
(5) W ilson, W.T., C ox, R.L., & M offett, J.O . (1990) Amer. Bee J.
thol containing honey. Two honey samples, rather than only 1
130, 4 0 9 -4 1 2
sample, were then spiked with L-menthol at very similar levels
(6 ) W i l s o n , W .T ., C o x , R . L . , M o f f e t t , J . O . , & E l l i s , M . ( 1 9 9 0 )
(38.3 ppm), and panelists were asked to identify the odd sam­
Bee Sci 1 , 4 8 -5 4
ple, which now contained no L-menthol. It is interesting that the
panelists were unable to identify it at the same level of signifi­
(7) Rivera, R., & Wilson, W.T. (1989) Amer. Bee J. 129, 821
(8) Veith, G.D ., & K iw us, L.M . (1977) Bull. Environ. Contam.
cance. This finding may be due to a considerable aftertaste of
Toxicol. 17, 6 3 1 -6 3 6
L-menthol, which becomes more of a problem when 2 of 3 sam­
(9) Poste, L.M ., M ackie, D .A ., Butler, G., & Larm ond, E. (1991)
ples contain menthol, than when only 1 of 3 samples contains
in Laboratory Methods for Sensory Analysis of Food, A gri­
menthol. culture C anada R esearch B ranch Publication 1864/E
We tried to overcome the problem with the L-menthol after­ (10) Section B .15.002(1), D ivision 15. Canadian Food and Drug
taste by encouraging panelists to rinse their mouths with 1 % A ct and R egulations. Issued by the D epartm ent o f N ational
lemon juice between tasting samples. Obviously, this precau­ H ealth and W elfare. Available from the M inister o f Supply
tion was not sufficient to overcome the aftertaste with 2 L-men­ and Services, C anada
thol containing samples. (11) Fed. Regist. (1988)53,47810-47811
 S ilv estre E t A l .: J o u r n a l O f A O A C I n t er n a tio n a l V o l . 76, N o. 6 ,1 9 9 3 1295

(12) White, J.W., Jr (1979) in Honey: A Comprehensive Survey, (16) Schurig, V., & Nowotny, H.P. (1988) J. Chromatogr. 441,
E. Crane (Ed.), Heinemann, London, pp. 157-206 155-163
(13) Daharu, P.A., & Spoms, P. (1984) J. Agric. Food Chem. 32, (17) Roessler, E.B., Pangbom, R.M., Sidel, J.L., & Stone, H.
108-111 (1978) J. Food Sei. 43, 940-943
(14) Rivera, R., Cox, R.L., & Moffett, J.O. (1988) Amer. Bee J. (18) Emberger, R., & Hopp, R. (1985) Proceedings of the Interna­
128, 807 tional Conference of Topics in Flavor Research pp. 201-218
(15) Wilson, W.T., Pettis, J.S., & Collins, A.M. (1989) Amer. Bee (19) Daharu, R, & Spoms, P. (1985) Can. Inst. Food Sei. Technol.
J. 129, 826 J. 18, 63-66

FOOD COMPOSITION AND ADDITIVES

C u p r i m e t r i c A s s a y o f C a s e in H y d r o l y s a t e s

M.P.C. S il v e s t r e
Faculdade de Farmacia da UFMG, Departamento de Alimentos, Av. Olegario Maciel, 2360-30 180 Belo Horizonte, Brasil
E. L a t í , C. D a u p h in , and M. H a m o n
Faculté de Pharmacie, Laboratoire de Chimie Analytique, 5 me Jean Baptiste Clément, 92290 Chatenay-Malabry, France

A sim ple, rep ro d ucible, a n d p re cise potentiom etric On the basis of the well-known complex formation between
a s s a y (c o p p er in d icator electro d e) h a s b ee n d ev el­ copper (II) and both amino acids and peptides (12-16), we
o p e d for alkalinized so lu tio n s of p u re c a se in h ydro­ have studied the analytical behavior of such complexes with
ly sates, with o r w ith out additional am ino ac id s, c a r­ regard to copper ions in media alkalinized by sodium borate
b o h y d ra te s, o r lipids (sh o rt ch ain tryglicerides; (pH 10.0), by measuring electric potential with a copper elec­
SCT) u sin g a titrated c o p p e r (II) su lfa te solution. trode (17-20). We showed (18) that the average copper con­
sumption for a mixture of amino acids is 0.65 mol copper
T he c o p p e r c o n su m p tio n is related to both free
(II)/mol residue. For a di-, tri-, and tetrapeptide, the copper con­
am in o acid a n d p ep tid e c o n te n ts. T hus, in com pari­
sumption per mol of peptide is the same (about 1.30 mol).
s o n with th e a-am in o nitrogen determ ination, th e
Thus, we concluded that the copper consumption per mol of
m eth o d d e sc rib e d h e re giv es b etter inform ation amino acid decreases with the degree of polymerization
a b o u t th e co m p o sitio n of p rotein h y d ro ly sates. (0.44 mol and 0.33 mol for the tri-and tetrapeptide, respec­
tively). These observations led us to extend this method to the
cuprimetric assay of protein hydrolysates. In addition, because
T he nutritional value of protein hydrolysates is directly of the existence of commercial mixtures of protein hydro­
related to their composition. Over the last 2 decades, it lysates, some of which also contain carbohydrates and lipids,
has been shown that small peptides (fewer than 4 amino we extended our study to these products.
acids) produced by enzymatic proteolysis are better absorbed The results were compared with those obtained by classical
than free amino acids (1-9). Indeed, Grimble et al. (10) found methods (total and a-amino nitrogen determination).
that by increasing the dipeptide and tripeptide content, the nu­
tritional efficiency of such hydrolysates was improved. METHOD
Official analytical methods are unsuited to determining
these parameters. Gel permeation chromatography (11) can be
used but is not well adapted for distinguishing between large Apparatus
amino acides (e.g., tryptophan) and small peptides (e.g., digly­
Potentiometer.—Tacussel TA T 6 -EPL 3 coupled to an auto­
cine and even triglycine), which have lower molecular weights.
mated titrating reagent buret with a variable capacity ( 1 -
The search for simple and effective methods for determining
50 mL) and a recorder adjusted to a speed of 1 cm/mL. The
the composition of these products is, therefore, important.
apparatus was connected to a derivation Derivol curve system
giving the potential drift relative to the volume of reagent
Received June 2 2, 1992. Accepted Decem ber 4, 1992. added. The measurement electrode consisted of a copper wire
1296 S ilv estre E t A l .: J o u r n a l O f A O A C I n t er n a tio n a l V o l . 76, N o. 6 ,1 9 9 3

3 mm in diameter; the reference electrode was mercury chlo­ with complex formation (14, 20, 22). We have previously
ride (calomel). shown that the most suitable neutralizing agent was a solution
of sodium borate-boric acid (17).
Reagents
The optimal experimental conditions were determined by
(a) Potassium caseinate.— Nutripharm-Gallia. using casein hydrolysate 4 in the presence of increasing vol­
(b) Casein hydrolysates.—Differing in degree of hydroly­ umes of borate solution at different pH values. The assay was
sis (numbered l-5)(Nutripharm-Gallia). optimal with 40-80 mL alkalinizing solution at an initial pH of
(c) Glucose, maltose, and lactose.—Analytical grade (E. 9.5-10.0. When the final pH was maintained above 7.0, reagent
Merck). consumption was stable and the potential variation was maxi­
(d) Medium-chain triglycerides.—MCT (Société Indus­ mal. These conditions were identical to those determined for
triel des Oléagineus). the assay of amino acids and small peptides alone and in com­
bination (17-20), so we chose 40 mL borate solution at
(e) Maltodextrins.—(Roquette).
pH 10.0.
(f) Copper sulfate solution.—Titrated at 0.1 mol/L.
(g) Alkalinizing solution.—0. IM boric acid-sodium borate, Analytical Parameters
pH 10.
Repeatability.—The method was used to perform 10 assays
Nitrogen Operating Conditions
of hydrolysate 4. The mean value was 173.0 mmol copper
Total nitrogen (TN) and a-amino nitrogen (N a -Am) were (II)/100 g product, with a standard deviation (SD) of 0.7 and a
determined according to AOAC methods (21). coefficient of variation (CV) of 0.4%. Therefore, the cuprimet­
ric assay showed very good repeatability under these operating
Preparation o f Sample conditions.
(a) Soluble proteins, casein hydrolysates, and carbohy­ Reproducibility.—Ten solutions of hydrolysate 4 at 1.6 g%
drates, alone or in combination.— A precise amount of product were prepared in distilled water, and 25 mL of each was as­
(4.0 or 8.0 g, according to expected reagent consumption) was sayed by the cuprimetric method. The mean value was
placed in a 500 mL volumetric flask and diluted to the required 172.3 mmol of copper (II)/100 g product (SD 2.3; CV 1.4%).
volume with distilled water. A 25 mL aliquot of this solution Therefore, the cuprimetric assay showed very good reproduci­
was placed in a 250 mL flask containing 40 mL alkalinizing bility under these operating conditions.
solution, and the potentiometric assay was then carried out with Linearity.—Linearity was also studied using hydrolysate 4.
the apparatus described above. Good linearity was demonstrated between 80 and 960 mg hy­
(b) Pure lipids or lipids mixed with casein hydrolysates drolysate, yielding a conrelation coefficient (r) of 0.9995. The
and/or carbohydrates.—A precise amount of product (4.0 or regression equation was as follows: y = 1.629a + 0.008, where
8 .0 g according to expected reagent consumption) was placed y is the copper consumption in mmol copper (II), and x is the
in a 500 mL volumetric flask containing 10 or 30 mL 95% quantity of hydrolysate in grams.
ethanol, and diluted to the required volume with distilled water.
A 25 mL portion of the formed stable emulsion was then placed R esu lts a n d D iscu ssio n
in a 250 mL flask containing 40 mL alkalinizing solution.
Determination o f total nitrogen (NT), a-amino nitrogen (N
Cuprimetric Assay Operating Conditions
a-Am), and N a-Am/NT ratio.—Table 1 shows the results for
Influence o f pH and the neutralizing agent.—During copper total nitrogen, a-amino nitrogen, and the N a-Am/NT ratio for
(Il)-induced amino acid complex formation, protons are re­ potassium caseinate and the 5 casein hydrolysates. The hydro­
leased in the medium. For the reaction to be complete, the acid­ lysates are listed in order of increasing a-amino nitrogen con­
ity must be neutralized by adding a base that does not interfere tent.

Table 1. Total nitrogen and a-amino nitrogen concentrations and a-amino nitrogen/total nitrogen ratios of potassium
caseinate and casein hydrolysates*
Sam ple Total nitrogen, g % a-Am lno nitrogen, g % a-Am ino nitrogen/total nitrogen ratio

Potassium caseinate 15.72 + 0 .4 3 a 1.05 ± 0.0 2a 0.0 7 ± 0.0 0a

Hydrolysate 1 13.15 + 0.75b 1.83 + 0.03b 0.1 4 + 0.01b
Hydrolysate 2 13.19 + 0.7 4b 2.1 4 + 0.03c 0.1 6 ± 0.01b
Hydrolysate 3 14 .36 + 0.3 9c 3.8 4 ± 0.23d 0.2 7 + 0.01c
Hydrolysate 4 13.92 + 0.82b,c 4.4 4 ± 0.0 2e 0.3 2 ± 0.02d
Hydrolysate 5 13.84 + 0.26b,c 5 .8 2 + 0 .1 9f 0.4 2 ± 0.0 2e

" Th ese values represent the m ean and the standard deviation of 5 m easures. Statistical m easures by Student’s f-test. Within a column, the
m eans followed by the sam e letter w ere not significantly different at the 5% level.
 S:lv estr e E t A l .: J o u r n a l O f A O A C In t ern a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1297

T able 2. C o p p e r c o n su m p tio n , copper/total n itro gen ratio ( R i) an d c o p p e r/a-am in o nitro gen ratio (R 2 ) o f p o ta s s iu m
ca se in a te a n d c a s e in h y d ro ly sa te s

Sam ple C opper consumption, mmols copper(ll) %a Rf R2

Potassium caseinate 52 .72 1 1 ,23a 3 .3 6 1 0 .1 6 a 5 0 .3 2 1 1 .7 2 a

Hydrolysate 1 9 4 .3 4 1 2 .2 2b 7 .1 9 1 0 .4 1 b 5 1 . 5 8 1 0.9 4a
Hydrolysate 2 9 3 . 1 7 1 0.9 9b 7.0 8 1 0.40b 4 3 .2 7 1 0.25b
Hydrolysate 3 1 6 0 .0 0 1 1 .5 8 c 1 1 .1 5 1 0 .3 7 c 4 1 .7 9 1 2.56b
Hydrolysate 4 173.01 1 0.7 4d 12.47 ! 0 .7 4 d 3 8 .9 5 1 0 .1 4 c
Hydrolysate 5 2 1 3 .0 0 l2 .5 5 e 1 5 .3 9 1 0 .1 1e 3 6 . 6 3 1 1.34d

a Th ese values represent the m ean and the standard deviation of 10 m easures. Statistical m easures by Student’s f-test. Within a column, the
m eans followed by the sam e letter w ere not significantly different at the 5% level.

As expected, the total nitrogen content of the 5 hydrolysates complementary to that given by the classical methods with re­
was very similar (14.36-13.15 g%), because all were derived gard to free amino acid and small peptide contents and, as a
from the same protein. However, the slightly higher value ob­ result, the degree of hydrolysis.
served for potassium caseinate suggests a decrease in nitrogen Comparison o f copper/total nitrogen (Rl) and copper/a-
content during the preparation of the hydrolysates. This could amino nitrogen (R2) ratios fo r casein hydrolysates and potas­
be accounted for by a loss of tyrosine (23), as well as by partial sium caseinate.—We calculated these ratios to improve the in­
hydrolysis of glutamine and asparagine residues due to the terpretation of the results of the cuprimetric assay and to gain
amidase activity of the pancreatic enzymes used. precision in the analytical study of protein hydrolysates (Ta­
a-Amino nitrogen content varied between 1.83 and ble 2 ).
5.82 g/100 g of product for the different hydrolysates. Again, The a-amino nitrogen/total nitrogen and copper/total nitro­
as expected, the results for potassium caseinate were lower gen ratios (Tables 1 and 2) of the hydrolysates showed similar
than for the hydrolysates. Given that a-amino nitrogen reflects
free amino acid content, and that the N a-Am/TN ratio indi­
cates the proportion of the total nitrogen represented by a-
R1
amino nitrogen, these results confirm that the 5 hydrolysates
differed in degree of hydrolysis. Products 1 and 2 were weak
hydrolysates, products 3 and 4 were intermediate, and product
5 was strongly hydrolyzed. However, on the basis of these re­
sults, it was not possible to distinguish clearly between prod­
ucts 1 and 2 .
Cuprimetric assay o f casein hydrolysates and potassium
caseinate.—The results of the cuprimetric assay of the 5 casein
hydrolysates and potassium caseinate, expressed as millimoles
of copper (II) ions per 100 g of product, are shown in Table 2.
Values for a-amino nitrogen content (Table 1) and copper con­
sumption were roughly parallel; the latter parameter increased,
as expected, with the degree of hydrolysis (20). Copper (II)
consumption by the hydrolysates varied from 93.2 to 213.0
mmol%; the value for potassium caseinate was approximately
half of the weakest hydrolysate.
Hydrolysates 1 and 2 showed similar copper consumption,
although the N a-AM content of hydrolysate 2 was higher than
that of hydrolysate 1. This suggests that the latter contained a
smaller amount of free amino acids and a larger quantity of
small peptides, because a-amino nitrogen corresponds only to
free amino acid content, whereas copper consumption also re­
flects peptide content. In fact, we confirmed these results by
developing a chromatographic method for separating and
measuring the small peptide and amino acid content in casein
F igu re 1. R e la tio n sh ip betw een copper/total nitrogen
hydrolysates (these results will be published shortly). (R i) an d a -a m in o nitrogen/total nitro gen ra tio s o f c a se in
These observations are interesting, because they show that h y d ro ly sate s.
cuprimetric assay of protein hydrolysates provides information
1298 S ilv estre E t A l .: J o u r n a l O f A O A C In t er n a tio n a l V o l . 76, N o. 6 ,1 9 9 3

T able 3. Total nitrogen, a -a m in o nitrogen, a n d c o p p e r c o n su m p tio n of s o m e protein p ro d u c ts a n d m ixtures

Total nitrogen, a-Am ino nitroqen, C opper consumption,

Product g % g % mmol copper(ll) %

Potassium caseinate 15.7 1.1 52

M oderate casein hydrolysate 14.3 3.9 160
Strong casein hydrolysate 14.0 5.8 215
Amino acid mixture 12.8 10.5 48 0
45% amino acids + 55% m oderate casein hydrolysate 13.2 6.3 311
90% potassium caseinate + 10% am ino acids 15.4 2.5 84

trends, confirming their classification on the basis of the degree commonly added to hydrolysates. The lipids chosen were
of hydrolysis. There was a strong correlation between the 2 short-chain triglycerides (SCT), which are commonly used in
ratios (r = 0.993, Figure 1; Student’s t-test result, t = 0.028 at enteral nutrition.
the 5% level). We first determined whether these components reacted with
As expected, the order of the R2 ratios was reversed, i.e., the copper under the conditions used for assaying the hydrolysates
weakest hydrolysate had the highest value. A less strongly de­ by determining the potential of each component, separately and
graded hydrolysate would contain a higher proportion of small in a mixture. As SCTs are insoluble in water, it was necessary
peptides than of free amino acids, and this would tend to in­ to add 95% ethanol. However, because the cuprimetric assay
crease R2. The values of R2 for hydrolysates 1 and 2 confirmed of hydrolysates is carried out in aqueous medium, we deter­
the higher small-peptide content in the former. However, it is mined whether the presence of alcohol interfered with copper
important to note that although the R2 ratios for hydrolysates 2 consumption by performing a potentiometric assay with the
and 3 were similar, higher values were obtained with hydro­ same volume of alcohol as that used to dissolve the SCT, under
lysate 3 for the other parameters, particularly a-amino nitrogen the operating conditions described above. It was noted that in
and copper consumption. This indicates that product 3 was all cases (carbohydrates, lipids, and alcohol), there was no cop­
richer in both free amino acids and small peptides. per consumption.
To check whether copper consumption by a given protein
Cuprimetric assay o f mixtures of proteins (hydrolyzed and
hydrolysate was affected by these carbohydrates and lipids, we
nonhydrolyzed) and amino acids.—The cuprimetric assay was
assayed a casein hydrolysate (number 1 ) to which these com­
applied to a mixture of free amino acids and the most weakly
ponents had been added, separately or mixed. Table 4 gives
hydrolyzed casein preparation. The mixture was prepared in
comparative results for the pure hydrolysate (no added carbo­
such a way that the proportion of a-amino nitrogen was similar
hydrates or SCT); no interference with the protein hydrolysate
to that contained in the preparation with the greatest degree of
assay was shown. Thus, the cuprimetric assay can be used to
hydrolysis. Indeed, this is often the case in commercial prepa­
analyze protein hydrolysates containing carbohydrates and lip­
rations. Some international manufacturers of dietary products
ids.
modified the composition of their products a number of years
ago (without informing the consumer) by using either a rela­
tively homogeneous hydrolysate (in terms of molecular
weight), or a mixture of weakly degraded proteins and
amino acids. T able 4. C o p p e r c o n s u m p tio n b y c a s e in h ydrolysate,
A comparison of the results obtained by the 3 methods de­ alo n e a n d with c a rb o h y d ra te s and/or lip id s
scribed above and those obtained for the strong hydrolysate
C opper consumption, mmols
showed that the difference in total nitrogen and a-amino nitro­ Sam ple coDper(ll)/100 g hydrolysate3
gen was relatively small, whereas copper (II) consumption by
the mixture was almost 50% higher than that by the strong hy­ Hydrolysate 1 9 4 .3 4 ± 2.22
drolysate (Table 3). The difference in copper (II) consumption Hydrolysate 1 + glucose 9 5 .3 7 ± 2.0 0
by the mixture of amino acids with casein (no peptides), re­ Hydrolysate 1 + maltose 93 .93 ± 3.55
mained significant but was less marked than that of a-amino Hydrolysate 1 + lactose 9 2 .4 9 ± 1.31
nitrogen. Hydrolysate 1 + maltodextrin 9 2 .4 9 ± 2.7 3
Hydrolysate 1 + S C T 6 9 5 .3 7 + 2.0 0
Lack o f effect o f carbohydrates and lipids.—Given the ex­
Hydrolysate 1 + mixture6 9 2 .7 6 ± 3.31
istence of various commercial protein hydrolysates containing
different quantities of carbohydrates and lipids, we conducted a Th ese values represent the m ean and the standard deviation of
studies to determine the possible influence of these compo­ 10 m easures. Statistical m easures by Student’s f-test. Th ere is no
nents in the cuprimetric assay. The carbohydrates studied were significant differences between these values at the 5% level.
b SCT, short-chain triglycerides.
glucose, maltose, lactose, and a maltodextrin. Lactose is a nor­ c Mixture, glucose + m altose + lactose + maltodextrin + SCT.
mal component of milk; the others are the carbohydrates most
 S ilv estre E t A l .: J o u r n a l O f A O A C Intern a tio n a l V o l . 76, N o. 6 ,1 9 9 3 1299

C o n clu sio n (6) Hara, H., Funabiki, R., Iwata, M., & Yamazaki, K.I. (1984) J.
Nutr. 114, 1122-1129
These results show that copper (II) ion consumption by ca­ (7) Webb, K.E., Jr (1990) J. Anim. Sci. 68, 3011-3022
sein hydrolysate solutions increases with the degree of hy­ (8) Cosnes, J. (1991) Rev. Prot. (Paris) 41, 699-702
drolysis. This opens the way to a new assay method for protein (9) Debry, G. (1991) Rev. Prot. (Paris) 41, 715-717
hydrolysates that complements existing methods. The pro­ (10) Grimble, G.K., Keohane, P.P., Higgins, B.E., Kaminski,
posed assay technique is better suited than the classical M.V., & Silk, D.B.A. (1986) Clin. Sci. 71, 65-69
a-amino nitrogen assay for differentiating between homoge­ (11) Pellerin, E, Baylocq, D., Majcherzyk, C., & Hamon, M.
nous hydrolysates and mixtures consisting of amino acids and (1985) Rev. Ind. Agro-aliment. Tech. Lait 356, 9-13
poorly hydrolyzed proteins. It can be applied both to pure pep­ (12) Amico, R, & Danielle, P.G. (1979) Inorg. Chim. Acta 35,
tide preparations and to mixtures containing lipids and carbo­ 383-386
hydrates. (13) Hay, R.W., & Williams, D.R. (1982) Spec. Period. Rep.
Amino Acids Pept. Proteins 13, 408^140
A ck n o w led g m en ts ( 14) Margerum, D.W. (1982) Pure Appi. Chem. 55, 23-34
(15) Mori, W, & Nakahara, A. (1979) Inorg. Chim. Acta 37, 507-
508
We are grateful to Nutripharm-Gallia for providing samples
of potassium caseinate and the various casein hydrolysates. (16) Norman, C.L., & Doody, E. (1952) Inorg. Chim. Acta 20,
4184-4188
(17) Lati, E., Dauphin, C, & Hamon, M. (1989) Ann. Pharm. Fr.
R efe ren c es 47, 344-352
(18) Lati, E., Dauphin, C, & Hamon, M. (1991) Ann. Pharm. Fr.
(1) Matthews, D.M. (1975) Physiol. Rev. 55, 537-608 49, 24-30
(2) Silk, D.B.A., Marrs, T.C., Addisson, J.M., Burston, D., (19) Lati, E., Dauphin, C, & Hamon, M. (1991) Anal. Chim. Acta
Clark, M.L., & Matthews, D.M. (1973) Clin. Sci. Mol. Med. 254, 89-94
45,715-719 (20) Lati, E., Dauphin, C, Silvestre, M., & Hamon, M. (1992)
(3) Fairclough, P.D., Hegarty, J.E., Silk, D.B.A., & Clark, M.L. Anal. Chim. Acta 268, 163-169
(1980) Gut 21,829-834 (21) Official Methods of Analysis (1984) 14th Ed., AOAC, Ar­
(4) Rees, R.G., Grimble, G., Keohane, P.P., Higgins, B.E., West, lington, VA
M„ Spiller, R.C., & Silk, D.B.A. (1984) Gut 25, A547 (22) Hamon, M., Hoppenot, A., & Guemet, M. (1972) Ann.
(5) Keohane, P.P., Grimble, G.K., Brown, B., Spiller, R.C., & Pharm. Fr. 9, 595-600
Silk, D.B.A. (1985) Gut 26,907-913 (23) Knights, R.J., & Manes, J.D. (1987) Food Allerg. 273-285
1300 Smith E t A l .: J ournal O f AO AC International V ol . 76, No. 6,1993

FOOD COMPOSITION AND ADDITIVES

A Q u an titative L in e a rity E valu atio n M eth o d fo r In fra re d M ilk

A nalyzers

E rika B. Smith, D avid M. Barbano, and J oanna M. L ynch

Northeast Dairy Foods Research Center, Cornell University, Ithaca, NY 14853
J. Richard F leming
U.S. Department of Agriculture, Agricultural Marketing Service, Texas Milk Market, Carrollton, TX 75006

N o n lin e a r it y in th e u n c o r r e c t e d s i g n a l fo r a n y c h a n ­ tor periodically and adjust appropriately the uncorrected signal

n e l o f th e in fr a re d m ilk a n a ly z e r c a n c a u s e d e ­ linearity for each channel. Atypical IR has 4 channels that use
c r e a s e d a c c u r a c y o f th e f in a l c o r r e c t e d re su lt. A optical filters (fat A, 5.723 pm; fat B, 3.48 pm; protein,
p r o c e d u r e q u a n t if y in g th e a m o u n t o f r e s id u a l n o n ­ 6.465 pm; and lactose, 9.610 pm) in combination with specific
lin e a r ity in th e u n c o r r e c t e d s i g n a l f o r e a c h c h a n n e l reference wavelengths (2). Linearity adjustment is necessary
w a s d e v e lo p e d fo r u s e in th e r a n g e o f m ilk c o m p o ­ because the Beer-Lambert law is not always valid for low-reso­
n e n t c o n c e n t r a t io n s a p p r o p r ia t e f o r p r o d u c e r p a y ­ lution spectroscopy (3). Nonlinearity is generally caused by
m e n t te s t in g . F o r th e fa t c h a n n e ls , a h o m o g e n i z e d problems or changes associated with the servocomb, servopo-
6 % fa t m ilk i s d ilu t e d w ith w a te r t o o b t a in a s e r i e s tentiometer, or optical filters (3). Biggs et al. (3) recommended
o f s o l u t i o n s f r o m 0 t o 6 % fat. F o r th e p r o t e in a n d that signal linearity should be checked after servicing or re­
la c t o s e c h a n n e ls , e v a p o r a t e d s k i m m ilk i s d ilu t e d placement of the sample cell, optical filters, or homogenizer.
w ith w a te r t o o b t a in lin e a rity s o l u t i o n s fr o m 2.8 to According to the instrument manufacturer, sample cell or ho­
3 . 9 % tr u e p r o t e in a n d 4.3 t o 6 . 0 % a n h y d r o u s la c ­ mogenizer replacement should not influence linearity (per­
t o s e . T h e s e le c t io n o f d ilu t io n m e t h o d (i.e., sonal communication, Gert Hoy Jakobsen, A/S Foss Electric,
w e ig h t / w e ig h t o r v o lu m e / v o lu m e ) s h o u l d b e b a s e d Hillerod, Denmark).
o n h o w th e fin a l c a lib r a t e d in s t r u m e n t v a l u e s a re to The most important consequence of signal nonlinearity is
b e e x p r e s s e d (i.e., w e ig h t / w e ig h t o r w e ig h t / v o l- decreased accuracy of the final estimates of the fat, protein, and
u m e ). A n o n - z e r o f o r c i n g lin e a r r e g r e s s i o n i s c o n ­ lactose content of milk (3). Nonlinearity increases the magni­
d u c t e d w it h in a c h e m ic a l r a n g e b y u s i n g k n o w n tude of the standard deviations of the difference between in­
c o m p o n e n t c o n c e n t r a t io n s a s th e d e p e n d e n t v a r i­ strument and chemical reference tests and influences the mag­
a b le (V ) a n d u n c o r r e c t e d in fr a re d s i g n a l s a s th e in ­ nitude of intercorrection coefficients (4).
d e p e n d e n t v a r ia b le (X ). R e s id u a l n o n lin e a r it y v a l­ The purpose of a linearity evaluation is to determine the
u e s a r e c a lc u la t e d . A c h a n n e l p o s s e s s e s extent to which uncorrected signals change linearly with linear
u n a c c e p t a b le n o n lin e a r it y if r e s id u a l v a l u e s a r e changes in fat, protein, and lactose concentrations. Linearity
>fc0 .0 2 % a t a n y p o in t w it h in th e c o m p o n e n t r a n g e evaluation procedures for each IR channel have been published
n o r m a lly e n c o u n t e r e d in p r o d u c e r p a y m e n t t e s t in g . by the AOAC (2) and the International Dairy Federation (IDF)
T h e lin e a rity s t a t u s o f 1 8 in s t r u m e n t s w a s d e te r­
(5). The general procedure involves (/) preparing a series of
m in e d .
dilutions of the major chemical component absorbing fight at a
particular wavelength (i.e., fat for the fat channels), (2 ) testing
each of the linearity solutions with the instrument, and (3)
I nfrared milk analyzers (IRs) are mid-infrared transmittance evaluating signal linearity for each channel by plotting instru­
instruments that use optical filters to select mean wave­ ment values against hue chemical values for each solution. The
lengths for measurement of fat, protein, and lactose content graphs are then inspected visually; if a plot is judged not linear,
of milk. Detector voltage is converted to a partially linearized
both the AOAC and the IDF methods instruct that the linearity
uncorrected signal within the instrument by a log,, amplifier. IR
for that channel should be adjusted as indicated in the instru­
uncorrected signals are linear when they change linearly with
ment’s operating manual.
changes in fat, protein, and lactose concentration, as defined by
Both the AOAC and the IDF linearity evaluation procedures
the Beer-Lambert law for monochromatic light (1). Although
have limitations (2, 5). First, neither procedure provides a
the instrument manufacturer initially sets the linearity for each
method to quantify the residual nonlinearity of the signal or
instrument channel, it is the operator’s responsibility to moni-
specifies how much nonlinearity is acceptable. Second, the
IDF method suggests that primary (uncorrected) signals be
R e c e iv e d J u ly 2 7 , 1992. A c c e p te d F e b u a ry 15, 1993.
used whenever possible, whereas the AOAC method does not
 Sm it h E t A l .: J ournal O f AO AC International V ol .7 6 ,N o .6 , 1993 1301

specify if uncorrected or corrected signals (6 ) should be used als were used to set limits on the amount of nonlinearity that
for the evaluation. Third, the choice of the specific range of can be accepted before it is necessary to readjust linearity.
component concentrations over which each channel should be It was determined that uncorrected instrument readings are
linear is not specified clearly for all components. more appropriate for linearity evaluation of residual nonlinear­
Fourth, the procedures differ with respect to the chemical ity. Uncorrected signals are the primary signals from each
composition of the stock materials used for evaluation of the channel prior to subtraction or addition of the intercorrection
fat and protein channels (Table 1). In particular, the AOAC values from the other channels and prior to the final slope and
method uses a dilution series of homogenized and/or unhomo­ bias adjustments of the intercorrected signal. The corrected sig­
genized cream to evaluate the fat channels, whereas the IDF nals are the final predictions of the chemical values for each
procedure recommends using a dilution series of unhomo­ component by the instrument. Uncorrected signals should be
genized cream. Furthermore, the AOAC method instructs that used for linearity evaluation, if available. On some older mod­
dilutions be made with water, whereas the IDF procedure calls els of IRs (e.g., Milkoscan 300,203, and 104), uncorrected sig­
for dilution with skim milk. It is not clear if these different di­ nals may not be easily accessible. On new models, uncorrected
luents will yield different linearity evaluation results. For the signals are obtained by setting the intercorrection coefficients
protein channel linearity evaluation, the AOAC method recom­ to zero, the slope of the component being measured to one, and
mends water dilutions of a 1.294% calcium propionate solu­ the intercept to zero.
tion. The IDF procedure recommends an ultrafiltration (UF) Some calibration methods may mask uncorrected signal
skim milk retentate stock material diluted with ultrafiltrate (i.e., nonlinearity. Although these calibration methods may improve
permeate) to make a series of protein concentrations. the final mean corrected signal, the overall standard deviation
Finally, the methods differ with respect to the recommended of the differences between instrument and chemical values
method of dilution. The AOAC specifies volume/volume dilu­ may be increased because of poor linearity. The use of uncor­
tions (no temperature of the solutions is specified), whereas the rected signals for linearity evaluation provides a better assess­
IDF procedure specifies weight/weight or volume/volume di­ ment of the true linearity behavior of the log,, amplified signal
lutions. for each channel.
The objectives of this study were to develop a residual non­ The linearity for each channel should be evaluated over the
linearity procedure that quantifies the degree of nonlinearity range of component concentrations for which the final calibra­
remaining in the uncorrected IR signal for each channel and to tion will be used. New instrument models can separate linearity
use the residual nonlinearity procedure to assess the linearity adjustment for each set of calibration equations used for differ­
status of a group of IRs currently used in the dairy industry. ent types of products. For producer payment testing in the
United States, the most critical ranges in which residual non­
E x p e r im e n t a l linearity should be minimized are from 2.5 to 5.8% fat, 2.5 to
3.9% true protein, and 4.3 to 4.9% anhydrous lactose. These
Development of Linearity Evaluation Procedure ranges were selected on the basis of a survey of milk composi­
tion for 300 farms tested during January and again in June in a
Method for quantitation of nonlinearity.—The limitations federal milk-marketing area. Although these ranges may not
of the AOAC and the IDF linearity evaluation procedures were include all concentrations that may be encountered in producer
addressed in preliminary work. A residual plot of differences payment testing, the majority of producer payment samples
between chemical values and instrument readings was selected would fall within these ranges. In other areas of the world, ap­
as a quantitative method of expressing the amount of residual propriate ranges should be selected on the basis of range of
nonlinearity. In addition, absolute values for maximum residu­ composition encountered in the population.

Table 1. Stock materials for evaluation of signal linearity according to AOAC and IDF procedures
Linearity Solvent for
Channel Procedure Dilution method stock material dilution series

Fat A and B AOAC vol/vol Homogenized cream, unhomogenized cream Water

IDF vol/vol Unhomogenized cream Skim milk
wt/wt

Protein AOAC vol/vol 1.294% Anhydrous calcium propionate solution Water

IDF vol/vol UF Skim milk retentate ultrafiltrate (i.e., permeate)
wt/wt

Lactose AOAC vol/vol 4.75% Anhydrous lactose solution with 0.25 g/L HgCI Water
IDF vol/vol 4.75% Anhydrous lactose solution Water
wt/wt
1302 S mith E t Al .: J ournal O f AO AC International V ol . 7 6 ,N o . 6,1993

Weight/weight vs volume/volume dilutions.—Ideally for to equal the protein and lactose concentrations in the 6 % fat
volume/volume dilutions, the solution volumes should be stock solution.
measured at the temperature at which they will be tested on the Dilution of the homogenized stock with a skim milk that has
instrument. One rationale for the volume/volume dilution ap­ been adjusted to match the protein and lactose concentrations
proach is that the instrument’s cuvette holds a constant volume, in the 6 % fat stock solution would also be appropriate for
and therefore, the instrument signal changes linearly with vol­ linearity evaluation. This would result in a series of solutions
ume/volume changes in concentration of a component. How­ with a decreasing fat content but with a constant concentration
ever, the instrument can be adjusted to make the signals change of protein and lactose. The errors (discussed above) caused by
linearly with weight/weight changes in component concentra­ use of unadjusted skim milk would be avoided. To prepare so­
tion. In some countries instruments are calibrated with chemi­ lutions for this type of linearity evaluation, the analyst would
cal test results expressed on a weight/weight basis, whereas in need to measure by chemical methods the fat, protein, and lac­
other countries chemical test results for producer payment are tose contents of both the skim milk and the high-fat stock solu­
expressed on a weight/volume basis. When chemical test re­ tion and adjust (by water dilution of the skim milk) the protein
sults are reported on a weight/volume basis, the temperature at and lactose concentrations of the skim to match the concentra­
which the test applies must be specified. In the case of calibra­ tions in the high-fat stock solution. This approach appears to be
tion of an IR, the weight/volume chemical test used as the basis the most scientifically correct for preparing the evaluation so­
for calibration should be reported at the same temperature used lutions. However, in practice it would be very difficult to exe­
cute correctly in most laboratories. A direct comparison of the
for the infrared analysis (ca 40°C). This should be considered
2 approaches (water diluent vs composition-adjusted skim
when setting and evaluating IR linearity.
milk) for preparation of linearity solutions and evaluation of
If the final instrument results need to be expressed on a
instrument linearity was conducted. Given the linearity criteria
weight/weight basis, then the solutions used to set and check
described earlier in this paper, the results of the residual non­
linearity should be made on a weight/weight basis. However, if
linearity evaluation (i.e., pass/fail) were not different with 2
the final results are expressed on a weight/volume basis, then
different sets of solutions (data not shown). Therefore, water
the solutions used to set and check linearity should be made on
dilution of the homogenized stock is a more practical approach
a volume/volume basis. Neither the AOAC nor the IDF method
that is more likely to be executed correctly in all laboratories.
addresses this important issue.
The choice of materials recommended by AOAC and IDF
Selection o f stock materials and diluents.—The high-fat
for evaluation of residual nonlinearity of the signal at the pro­
stock material used for the linearity evaluation of the fat A and
tein wavelength was addressed. UF skim milk retentate diluted
fat B channels should be homogenized. The use of homoge­
with permeate (solution of lactose and minerals) may not be
nized milk for the preparation of linearity solutions will pro­
adequate to evaluate protein uncorrected signal linearity as
vide more homogeneous samples that will reduce mixing and
suggested by the IDF method. When permeate is used as a di­
sampling errors.
luent, the concentration of true protein decreases linearly while
There are advantages to using water instead of skim milk as the lactose concentration varies because the lactose concentra­
the diluent for the fat channel linearity solutions. Experiments tion in the permeate is not the same as the lactose concentration
were conducted to compare use of water and skim milk to make in the retentate. The result is apparent nonlinearity of the pro­
dilutions of a homogenized milk containing 6 % fat. The con­ tein signal because of the systematic change in background lac­
centrations of all the milk components (i.e., fat, protein, and tose concentration. In addition, UF skim milk retentate and per­
lactose) decreased linearly with a water diluent. Skim milk meate cannot be obtained easily by the average IR operator in
contains a small amount of fat, which will vary from one linear­ most laboratories.
ity evaluation to the next. This makes it necessary to accurately Water dilutions of a calcium propionate solution, as recom­
measure the fat content of the skim milk by ether extraction (7). mended by AOAC, are not ideal for protein linearity evalu­
No fat measurement is necessary when water is the diluent. ation, because calcium propionate absorbs light at a slightly
When milk is separated into skim and cream, the protein and different wavelength (6.21-6.45 |um) and has an absorption
lactose concentrations in the skim milk are higher than in the band shape different from that of the peptide bond in protein
cream or a portion of cream diluted to a 6 % fat content with (6.465 |im) (1, 8 ). Several alternatives were evaluated; com­
skim. Thus, when skim milk is used as the diluent for a milk mercially canned evaporated skim milk (ca 7.1% true protein
containing 6 % fat, the concentrations of protein and lactose in and 1 1 .0 % anhydrous lactose) was selected to serve as stock
the linearity solutions will increase as fat concentration de­ material for preparation (i.e., water dilution) of linearity solu­
creases. The increase in background protein and lactose con­ tions for simultaneous assessment of protein and lactose uncor­
centrations as the fat concentration decreases causes an incor­ rected signal linearity. This material is available as a canned,
rect apparent nonlinearity in the signal of the fat A and fat B shelf-stable product in most grocery stores in North America.
channels and may cause the operator to make an incorrect The linearity evaluation procedure described below was de­
linearity adjustment. This effect is due to the change in back­ veloped after the above issues had been considered. The result­
ground light absorbance by protein, lactose, and water at the fat ing quantitative linearity evaluation procedure represents a
wavelengths. Thus, water is a better diluent than a skim milk genera] method designed for instruments that are used to test
that has not had its protein and lactose concentrations adjusted raw milk for producer payment.
 S mith E t A l .: J ournal O f AOAC International V ol . 76, No. 6,1993 1303

Quantitative Linearity Evaluation Procedure should be poured into 1 of 2 dry containers. It is not unusual to
observe a small amount of milk mineral residue remaining on
Preparation o f linearity solutions.—Two sets of linearity the bottom of the can. This residue is not a problem and does
evaluation solutions are prepared: one for both fat channels not need to be transferred to or mixed with the sample. The
and one for the protein and lactose channels. The solutions evaporated skim milk is mixed by transferring from one con­
should be prepared on a volume/volume basis if the final in­ tainer to the other to ensure a homogeneous solution. Water
strument-predicted chemistry values are expressed on a dilutions of the evaporated skim milk are made such that the
weight/volume basis. However, if the final values are ex­ true protein content of the linearity solutions ranges from 3.9 to
pressed on a weight/weight basis, then the solutions should be 2 .8 %; as a result, the anhydrous lactose content of these solu­
prepared on a weight/weight basis. tions will range from about 6.0 to 4.3% anhydrous lactose (Ta­
The following method describes preparation of linearity ble 3). The true protein content of the commercial evaporated
evaluation solutions made on a weight/weight basis. If vol­ skim milk should be determined by the Kjeldahl method for
ume/volume solutions are prepared, glassware of appropriate true protein (9). To determine the anhydrous lactose content of
volumetric calibration accuracy and temperatures at which so­ the evaporated skim milk stock material indirectly, the total
lutions are prepared and tested need to be defined clearly. solids ( 1 0 ), ash ( 1 1 ), total nitrogen-based protein ( 1 2 ), and the
For the fat A and fat B channels, a 6 % fat (approximate) fat (7) content of the evaporated skim milk stock must be de­
stock material is prepared by diluting homogenized half-and- termined. The anhydrous lactose concentration can be calcu­
half (ca 1 0 % fat) milk with an appropriate amount of pasteur­ lated by subtracting values determined for total nitrogen-based
ized skim milk. Homogenized light cream (ca 18-20% fat) di­ protein, fat, and ash from the total solids value. The lactose
luted to 6 % fat with skim milk would also be appropriate as a content also can be measured directly (13). Again, an assump­
stock material. A series of weight/weight dilutions of the 6 % fat tion about the true protein and anhydrous lactose content of the
stock material are made with distilled water to obtain 5,4,3,2, evaporated skim milk can be made at this point to avoid doing
and 1% fat concentrations (Table 2). Distilled water is used for any chemical analysis of the evaporated skim milk. The limi­
the solution with 0% fat. Weights of water and stock material tation of this assumption becomes important when comparing
are determined to 4 decimal places. Fat determination by ether linearity performance across time or from instrument to instru­
extraction can be performed on the 6 % fat stock material to ment when the composition of the evaporated skim milk stock
establish the fat content (7); the fat chemistry of each dilution varies from one set of linearity evaluation solutions to another.
is calculated from the actual weights of the stock material and Evaluation o f linearity status o f fat channels.— Prior to
water used to make each solution. It is not absolutely necessary evaluation of linearity, the instrument should be evaluated for
to measure the fat content of the 6 % fat stock material by a repeatability on all channels. If the instrument cannot satisfy
chemical testing method if an assumption of a 6 % fat concen­ basic repeatability performance requirements, then the instru­
tration is made. However, chemical analysis of the stock mate­ ment requires servicing prior to linearity evaluation. To evalu­
rial is recommended if linearity performance of an instrument ate linearity the flow system of the instrument is cleaned and
or a group of instruments is to be monitored over time at spe­ all channels are set to zero. Ten measure cycles of 40°C pas­
cific fat concentrations with different sets of linearity solutions. teurized whole milk should be run through the instrument
For protein and lactose channels, a series of 7 solutions with (without collecting data) prior to conducting the linearity
a known linear change in true protein and anhydrous lactose evaluation, to coat the cell and provide for stable instrument
content is prepared by diluting a canned evaporated skim milk readings. Next, the flow system of the instrument should be
with distilled water on a weight/weight basis. Prior to dilution flushed with water and zeros should be checked and reset, if
of the evaporated skim milk, the can should be tempered in a necessary, for all channels. Large shifts in zero are not normal,
40X1 water bath for 1 0 min, and then opened, and the milk and if they occur, the cuvette probably needs to be replaced.

Table 2. Sample data set for linearity solutions to evaluate fat A and fat B uncorrected signal linearity
Stock material Fat
Solution Water, g (6% fat), g concn, %a

F0 100 0 0.0000
F1 83.3342 16.6658 0.9969
F2 66.6667 33.3333 1.9939
F3 50.0008 49.9992 2.9908
F4 33.3333 66.6667 3.9878
F5 16.6675 83.3325 4.9847
F6 0 100 5.9817to

a Fat concentration calculated based on fat content of solution F6 and actual weights of water and stock material.
b Fat concentration determined by Mojonnier method or assumed to be exactly 6.0000% fat.
1304 S mith E t A l .: J ournal O f AOAC International V ol . 76, No. 6,1993

Table 3. Sample data set for linearity solutions to evaluate protein and lactose uncorrected signal linearity
True
Stock material protein Anhydrous lactose
Solution Water, g (evaporated skim milk), g concn, %a concn, %a

P0 100 0 0.0000 0.0000

P1 60.7112 39.2564 2.7999 4.3157
P2 57.9451 42.1722 3.0034 4.6293
P3 56.6008 43.5582 3.1008 4.7794
P4 55.1182 44.8884 3.2003 4.9329
P5 50.7125 49.1229 3.5082 5.4075
P6 45.4327 54.8342 3.8993 6.0102

a True protein and anhydrous lactose concentrations calculated on the basis of chemical content of evaporated skim milk stock material and
actual weights of water and stock material. Analyst can assume evaporated skim milk stock contains 7.13% true protein and 10.99%
anhydrous lactose, or analyst can measure the protein and lactose content of the evaporated skim milk stock material by chemical analysis.

Prior to running each set of linearity samples, 3 initial water

readings are recorded for each channel (Table 4).
A set of 7 linearity evaluation solutions (F0 to F 6 ) is tem­ Table 4. Linearity evaluation data set for fat A and fat B
pered to 40°C, mixed, and mn through the IR in order of in­ channels of instrument 9, Multispec M-1
creasing fat concentration. An automatic rack-sampling system
Solutions® Reading No. Fat Afc Fat Bb
should not be used during this procedure unless 3 consecutive
readings can be taken from each linearity solution before test­
Water, Initial 1 0.01 0.00
ing the next linearity solution. Otherwise all samples should be
2 0.00 0.01
presented manually to the instrument so that 3 consecutive 3 0.01 0.00
readings may be taken from each sample vial. Sample tempera­
ture should be kept as uniform as possible. Three uncorrected F0 1 0.01 0.01
signals are collected for each linearity sample from the fat A 2 0.01 0.00
and fat B channels. A linearity evaluation data set for the fat A 3 0.01 0.00
and fat B channels of an instrument (instrument 9 Multispec F1 1 0.94 1.32
M -l) used in the linearity evaluation survey is given in Table 4. 2 0.94 1.33
After running the linearity sample with the highest fat concen­ 3 0.94 1.33
tration in the series, the instrument’s flow system is flushed
with water several times and then 3 consecutive ending zeros F2 1 1.95 2.69
2 1.96 2.68
(i.e., final water) are recorded (Table 4). The difference be­
3 1.96 2.72
tween the means of the initial and final water signals should be
<0.04%. If the zero shift is >0.04%, the linearity evaluation F3 1 3.01 4.13
should terminate at this point because the zero shift problem is 2 3.01 4.14
serious and needs to be rectified before linearity can be evalu­ 3 3.02 4.13
ated properly. The first reading for each linearity sample is not
F4 1 4.07 5.55
used (Table 4) to avoid possible distortions of the linearity
2 4.08 5.53
evaluation due to poor purging efficiency. The mean of the sec­
3 4.07 5.56
ond and third uncorrected readings for each linearity sample is
calculated and rounded to the third decimal place (Table 5). F5 1 5.23 6.94
A non-zero forcing linear regression is conducted with fat 2 5.23 6.97
chemistry values as the dependent variable (Y) and the uncor­ 3 5.24 6.97
rected infrared signals as the independent variable (X) in the
F6 1 6.41 8.42
chemical target range for a particular channel (Table 5). Con­ 2 6.41 8.44
ducting the regression in this manner (X = uncorrected signals, 3 6.42 8.43
Y = fat chemistry) ensures that the results of the linearity evalu­
ation (i.e., the magnitude of residual values) will not be influ­ Water, final 1 0.00 0.02
enced by the primary slope setting of the fat channels for the 2 0.01 0.00
instrument. For the fat A and fat B channels the regression 3 0.02 0.00
analysis was based on the set of data points between 2 and 5 % a Solutions F0 to F6 contain ca 0-6% fat. Actual fat concentrations
fat (i.e., solutions F2 through F5), which is the target range are given in Table 2.
selected for producer payment testing (Table 5). Selection of b Values are uncorrected signals.
 S mith E t A l .: J ournal O f AOAC International V ol . 76, No. 6 ,1993 1305

Table 5. Example of linearity evaluation calculations for fat A and fat B channels using data set for instrument 9,
Multispec M-1
Linearity Fat Uncorrected Predicted Residual
Channel solution concn, % signal, % concn, % value, %a

Fat Aft F0 0.0000 0.010 0.2294 0.229

F1 0.9969 0.940 1.0807 0.084
F2 1.9939 1.960 2.0144 0.021
F3 2.9908 3.015 2.9801 -0.011
F4 3.9878 4.075 3.9505 -0.037
F5 4.9847 5.235 5.0123 0.028
F6 5.9817 6.415 6.0925 0.111

Fat Bc F0 0.0000 0.000 0.0978 0.098

F1 0.9969 1.330 1.0303 0.033
F2 1.9939 2.700 1.9908 -0.003
F3 2.9908 4.135 2.9968 0.006
F4 3.9878 5.545 3.9854 -0.002
F5 4.9847 6.970 4.9845 0.000
F6 5.9817 8.435 6.0116 0.030

a Residual value = (predicted concentration - fat concentration).

b Fat A regression equation: predicted concentration = [0.9154 x (uncorrected signal)] + 0.2202, r2 = 0.99946.
c Fat B regression equation: predicted concentration = [0.7011 x (uncorrected signal)] + 0.0978, r2= 0.99999.

the appropriate number of solutions to be included in the re­ range (i.e., 2 to 5% fat) over which the regression analysis has
gression should be based on the concentration range over been based are <±0 .0 2 % residual.
which the final calibration will be used. Predicted chemical Channels that exceed the criteria by having residuals
values are calculated for each channel by using the regression >±0.02% within the chemical target range (i.e., 2 to 5% fat)
equation (Table 5). Residual nonlinearity values are calculated demonstrate unacceptable residual signal nonlinearity for use
for each channel by subtracting actual fat chemistry values in payment testing, where testing accuracy is very important.
from predicted chemical values. Finally, residual values are Such instruments require signal linearity adjustment. The mag­
plotted against actual fat chemistry values in a residual non­ nitude of residual nonlinearity in the uncorrected signal will be
linearity plot for each channel. An example residual nonlinear­ reflected directly in the corrected signal. Therefore, the
ity plot for the fat A and fat B channels for the data set from
<±0 .0 2 % criterion for residual nonlinearity results in a limit for
Table 5 is given in Figure 1. The signal response for a channel
maximum systematic error in corrected values, due to non­
is linear if the values of the residuals within the chemical target
linearity, of no greater than 0 .0 2 % difference from chemistry at
any fat concentration within the target chemical range. The re­
sidual nonlinearity plot for instrument 9 (Figure 1) indicates
that this instrument exceeds the residual nonlinearity criteria at
2, 4, and 5% fat on the fat A channel and, therefore, fails the
linearity evaluation for the fat A channel. Thus, the uncorrected
signal linearity for the fat A channel of instrument 9 must be
adjusted and then reevaluated for residual nonlinearity. The re­
sidual nonlinearity is acceptable within the target range (i.e., 2
to 5% fat) for the fat B channel (Figure 1).
Evaluation o f linearity status o f protein and lactose chan­
nels.—The instrument should be prepared for linearity evalu­
ation as described above. In addition, the sequence of calcula­
tions for fat channel residual nonlinearity should be followed
for evaluation of the protein and lactose channels. However,
the regression analysis for the protein channel was based on the
set of data points between 2.8 and 3.9% true protein (i.e., sam­
ples PI through P6 ), whereas the regression analysis for the
lactose channel was based on the set of data points (Table 3)
Figure 1. Residual nonlinearity plot for fat A ( ■ ) and
between 4.3 and 5.4% anhydrous lactose (i.e., samples PI
fat B ( A ) channels of instrument 9, Multispec M-1.
through P5). As described earlier, the signal response for a
1306 S mith E t A l .: J ournal O f AOAC International V ol . 76, No. 6,1993

T able 6. In stru m e n ts u se d in linearity evalu ation su rv e y T able 7. Initial n u m b e r of in stru m e n ts u s e d in survey,

n u m b er o f in stru m e n ts evaluated for nonlinearity, an d
Instrument Abbreviation No. used
n u m b er o f in stru m e n ts that p a s s e d linearity e valu atio n

Milkoscan 104 MS 104 7 Channel Initial Evaluated Passed

Mllkoscan 134 MS 134 1
Multispec Dairy Lab 1 DL-1 1 Fat A 18 14a 12
Multispec Dairy Lab 2 DL-2 1 Fat B 18 14a 14
Multispec M1 M-1 2 Protein 18 17b 17
Multispec M1 : Bentley3 M-1 B 2 Lactose 18 17b 17
Multispec M2 M-2 4
a Results of 3 instalments discarded for falling repeatability
a The Bentley modification of the Multispec M-1 uses the optical evaluation; 1 instrument discarded for excessive zero difference.
system of the original instrument and then transfers total control of b Results of 1 instrument discarded for falling repeatability
the signal to an external computer using software developed by evaluation.
Bentley Instruments (Chaska, MN).

The fat concentration of the 6 % fat stock material was de­

termined by ether extraction (7). The true protein content of the
channel is linear if the values of the residuals within the chemi­ commercial evaporated skim milk stock was determined by the
cal target range over which the regression analysis is based are indirect Kjeldahl method for true protein (9). The total solids
<±0 .0 2 % residual. (10), ash (11), and fat (7) contents of the evaporated skim milk
stock material were determined so that the anhydrous lactose
Linearity Evaluation Survey
concentration could be calculated.
One set of milk solutions for evaluation of fat A and fat B Eighteen instruments were included in the linearity evalu­
channel signal linearity was prepared through dilution ation survey (Table 6 ). All instruments were used routinely for
(weight/weight) of a homogenized 6 % fat stock material with raw milk testing, and no special adjustments of linearity were
water. A separate set of solutions for evaluation of protein and made to the instruments prior to the survey. All channels for
lactose uncorrected signal linearity was prepared through dilu­ each instrument used in the survey were evaluated first for the
tion (weight/weight) of a commercial evaporated skim milk performance characteristic of signal repeatability before linear­
stock material with water. Both sets of linearity solutions were
ity evaluations were performed.
subdivided into 18 sets of test samples. One set of test samples
A channel was considered to have a repeatability problem if
for evaluation of fat channel linearity and 1 set of test samples
the range for 2 0 consecutive readings of a homogenized or un­
for evaluation of protein and lactose linearity was sent by over­
night delivery to each of 18 laboratories. Distilled water for the homogenized milk was >0.04% fat. Only data from the chan­
purposes of instrument zeroing and flushing was also included nels of instruments that demonstrated satisfactory repeatability
with the 2 sets of test samples sent to each laboratory. In addi­ prior to linearity evaluation were included in the linearity sur­
tion, all analysts were provided with detailed instructions de­ vey results. In addition, zero stability during testing each set of
scribing sample handling and testing procedures. linearity samples was monitored.

T able 8. Linearity evalu ation su rv e y o f fat A a n d fat B c h a n n e ls

Fat linearity solution3

F2b
-Q

-Q

F0 F1
2

F4b F6
LL
LO

Fat channel Residual value, % No. of instruments

Fat A >0.07 9 4 0 0 0 0 1
<0.07 to > 0.02 4 6 1 0 0 1 2
<0.02 to > -0.02 1 4 13 13 13 13 6
<-0.02 to >-0.07 0 0 0 1 1 0 5
<-0.07 0 0 0 0 0 0 0

Fat B >0.07 5 0 0 0 0 0 0
<0.07 to > 0.02 6 11 0 0 0 0 5
<0.02 to > -0.02 2 2 14 14 14 14 7
<-0.02 to > -0.07 0 0 0 0 0 0 2
<-0.07 1 1 0 0 0 0 0

a Solutions F0, F1, F2, F3, F4, F5, and F6 contain ca 0,1,2, 3, 4, 5, and 6% fat, respectively. See Table 2 for actual values.
6 The solution was used for regression analysis in the linearity evaluation procedure.
 Sm it h E t A l .: J ournal O f AO AC International V ol . 76, No. 6 ,1 9 93 1307

T able 9. Linearity evalu atio n su rv e y o f protein ch an n e l

Protein linearity solution3

P0 P1b P2b P3b P4b P5b P6
Residual value, % No. of instruments

>0.07 15 0 0 0 0 0 0
<0.07 to > 0.02 1 0 0 0 0 0 2
<0.02 to >-0.02 0 17 17 17 17 17 15
<-0.02 to >-0.07 0 0 0 0 0 0 0
<-0.07 1 0 0 0 0 0 0

a Solutions P0, P1, P2, P3, P4, P5, and P6 contain ca 0, 2.7, 3.0, 3.4, 3.7, 4.1, and 4.4% true protein, respectively.
b The solution was used for regression analysis in the linearity evaluation procedure.

S u rv ey R esu lts results for ranges of residual values for the 7 linearity solutions
used to evaluate the fat, protein, and lactose channels.
For the fat A and fat B channels, 3 instruments failed the
repeatability criteria and were not included in the linearity D iscu ssio n a n d C o n clu sio n s
evaluation for these channels. Linearity results for 1 additional
instrument were discarded for the fat A and fat B channels, A quantitative method for evaluation of residual nonlinear­
because the difference between final and initial zeros for each ity of uncorrected signals was developed and tested by a group
of these channels exceeded the zero drift criteria during the of laboratories. The selection of the method for preparation of
actual linearity evaluation. Therefore, 14 instruments out of an solutions for setting and evaluating linearity (i.e., weight/
initial field of 18 were included in the final linearity evaluation weight or volume/volume) is an important decision. This deci­
for the fat A and fat B channels (Table 7). sion must be based on the form (i.e., weight/weight or
Results for the protein and lactose channel for 1 instrument weight/volume) in which the final instrument predictions of
were eliminated because of poor lactose and protein signal re­ chemical values will be expressed.
peatability. Thus, 17 instruments out of an initial field of 18 Evaluation of signal repeatability (in addition to other in­
were included in the final linearity results for the protein and strument performance characteristics) for each channel is es­
lactose channels (Table 7). sential prior to evaluation of linearity, because other mechani­
cal or electrical performance problems of the instrument may
Results of the linearity evaluation survey are summarized in
cause false conclusions about the acceptability of linearity. In
Tables 7-10. Overall, the IRs that demonstrated good repeat­
the survey, 4 of 18 instruments could not be evaluated on the
ability and stable zeros had excellent linearity in the target
fat A and B channels because of poor repeatability or zero in­
chemical ranges used for payment testing and, in almost all
stability. This result suggests that repeatability and zero stabil­
cases, had no difficulty achieving the <±0 .0 2 % residual non­ ity are more of a problem than uncorrected signal linearity. The
linearity criterion. Two instruments failed the linearity evalu­ fat, protein, and lactose channels gave acceptable linearity over
ation (i.e., >±0.02% residual nonlinearity) for the fat A channel the selected ranges of concentrations appropriate for producer
(Table 7). All instruments evaluated satisfied the linearity payment samples. Linearity may have to be checked after any
evaluation criteria for the fat B, protein, and lactose channels changes or adjustments are made to the optical filters or peri­
(Table 7). Tables 8-10 contain a summary of the distribution of odically during troubleshooting procedures.

Table 10. Linearity e valu atio n su rv e y of la c to se ch an n e l

Lactose linearity solution3

P2b
-O

P4b P5 P6
CL

P0 P1b
CO

Residual value, % No. of instruments

>0.07 10 0 0 0 0 0 0
<0.07 to >0.02 2 0 0 0 0 3 3
<0.02 to >-0.02 2 17 17 17 17 11 9
<-0.02 to >-0.07 0 0 0 0 0 3 5
<-0.07 3 0 0 0 0 0 0

a Solutions P0, P1, P2, P3, P4, P5, and P6 contain ca 0, 4.2, 4.7, 5.2, 5.7, 6.3, and 6.8% anhydrous lactose, respectively.
b The solution was used for regression analysis in the linearity evaluation procedure.
1308 S mith E t A l .: J ournal O f AOAC International V ol . 76, No. 6,1993

A c k n o w le d g m e n t s (4) Biggs, D.A. (1979) J. Assoc. Off. Anal. Chem. 62,1211-1214

(5) International Dairy Federation (1990) Provisional IDF Stand­
The authors acknowledge the technical assistance of the ard 141A
staff of all the Federal Milk Market Laboratories, Northeast (6) Barbano, D.M., & Clark, J.L. (1989) J. Dairy Sci. 72, 1627-
Dairy Herd Improvement Association, and Dairy Quality Con­ 1636
trol Institute. We are grateful for the assistance of Maureen (7) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
Chapman, Pat Nelson, and Enda Hu in this work. Assistance lington, VA sec. 989.05
from the staff of Foss Electric and Foss Food Technology was (8) Brugel, W. (1962) Introduction to Infrared Spectroscopy,
very helpful. John Wiley & Sons, Inc., New York, NY
(9) Barbano, D.M., Lynch, J.M., & Fleming, J.R. (1991) J. As­
R e fe re n c e s soc. Off. Anal. Chem. 74, 281-288
(10) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
(1) Dyer, J.R. (1965) Applications of Absorption Spectroscopy of lington, VA, sec. 990.20
Organic Compounds, Prentice-Hall, Inc.. Englewood Cliffs, (11) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
NJ lington, VA sec. 945.46
(2) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­ (12) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
lington, VA, sec. 972.16 lington, VA sec. 991.20
(3) Biggs, D.A., Johnsson, G., & Sjaunja, L.O. (1987) IDF Bulle­ (13) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
tin 208, International Dairy Federation lington, VA, sec. 896.01,984.15
 A li E t A l .: J ournal O f AOAC International V ol . 76, No. 6,1993 1309

R E S ID U E S A N D T R A C E E L E M E N T S

A n alyte S ta b ility Study o f N - IVIethylcar ham ate Pesticides in B eef

and P o u ltry L iv e r Tissues by L iq u id C hrom atography

M . Sher Ali, J erry D . W hite, R alph S. Bakowski, and E van T. P hillippo

U.S. Department of Agriculture, Food Safety and Inspection Service, Eastern Laboratory, PO Box 6085, Athens, GA 30604
R ic h a r d L. E l l is
U.S. Department of Agriculture, Food Safety and Inspection Service, Science and Technology, Chemistry Division, 300
12th St, SW, Washington, DC 20250

To optimize conditions for sample collection, prepa­ T he mission of the Technical Support Services Labora­
ration, storage, and analysis and to assure the va­ tory, U.S. Department of Agriculture, is to provide tech­
lidity of our previously published liquid nical support for inspection operations that ensure safe,
chromatographic (LC) method for carbamate analy­ wholesome, and accurately labeled meat and poultry products
sis in tissue, stabilities of 16 W-methylcarbamates for the American consumer. As a part of this mission under the
in beef, duck, and chicken liver tissues were stud­ auspices of the National Residue Plan, Food Safety and Inspec­
ied by using 2 sampling protocols. Tissue samples tion Service (FSIS), we have analyzed carbamate residues in
were fortified at room temperature to a concentra­ liver tissues from beef, swine, and poultry since 1988.
tion 5 to 10 times greater than either the Environ­ For any method of residue determination, information about
mental Protection Agency tolerance level for each the stability over time of an analyte in tissues or in environ­
compound (if established) or the concentration mental matrixes is useful for the following reasons: it enables
used in the previously published method. Thereaf­ the agency or program to establish acceptable conditions for
ter, samples were continuously frozen at -4°C for collection and laboratory storage of samples; it enables optimi­
varying time intervals. In the first study, samples zation of analytical conditions that minimize loss during ex­
were analyzed one day (initial) and 0 .5 ,1 ,1 .5 ,2 ,3 , traction, purification, and analysis; it can be used to assess po­
4,5, and 6 months after fortification. In the second tential interconversion of an analyte in a multiresidue method;
study, samples were analyzed one day (initial) and and, above all, it can be used to develop a quality assurance
0.5,1,2, 3, and 6 months after fortification. For plan.
each residue and species, a minimum of 4 samples In general, carbamates are thermally unstable compounds.
were analyzed by LC at each point in time, and the The effects of pH on the stabilities of some N-methylcar-
mean represented analyte concentration at the end bamates in aqueous solutions were reported (1). Recent studies
of each time interval. Rates of residue depletion found that in cool, anaerobic conditions (as in northern aqui­
varied among analytes and among species. Deple­ fers) and in limestone water (as in Florida), the breakdown of
tion rates were greater in duck livers than in beef aldicarb takes much longer than in warmer, aerobic, or more
livers. Methomyl and oxamyl were depleted com­ acidic conditions (2). The in vivo and in vitro metabolism of
pletely within 2 weeks. Between 2 and 6 months af­ the parent compound, N-methylcarbamate, in animal, plant,
ter sample fortification, residue depletions to levels and mmen microbes was reported (3,4). The stability overtime
below detection limits were observed for aldicarb,
of A-methylcarbamates in liver tissues from cattle, swine, and
aldicarb sulfoxide, aldicarb sulfone, dioxacarb, p ro
poultry has not been reported.
mecarb, propoxur, and bendiocarb. The initial loss
Stabilities of 16 /V-methylcarbamates were studied in beef,
of certain carbamates during sample preparation in
duck, and chicken livers by liquid chromatography (LC). Ten
tissues exposed to room temperature for up to 8 h
carbamates cited in a previous method (5) were analyzed one
was greater than the subsequent rate of loss. Results
day and 0.5, 1, 1.5,2, 3,4, 5, and 6 months after sample forti­
indicate that cryogenic conditions are required for
fication. The remaining 6 carbamates, which were cited in the
sample preparation and storage. Results also pro­
method extension (6 ), as well as 2 of the original 1 0 car­
vide information on how long a violative eviden­
bamates, were analyzed one day and 0.5,1,2,3, and 6 months
tiary sample can be stored for potential litigation.
after sample fortification.
In the original method (5), 5, 10, and 20 ppb levels of each
of the carbamate residues were used to generate standard
Received October 28, 1992. Accepted January 27, 1993.
Presented at the 105th AOAC INTERNATIONAL Annual Meeting, curves and test recovery in a repeatability study. These concen­
August 12-15, 1991, at Phoenix, AZ. trations represented, respectively, 0.5, 1, and 2x the tolerance
1310 Ali E t A l .: J ournal Of AOAC International V ol . 76, No. 6,1993

T a b le 1. C a rb a m a te a n a ly te s a n d in itia l fo rtific a tio n

le v e ls p e r s p e c ie s o f liv e r

C arbam ates Concentration, ppb Species

Group A
Aldicarb 50 beef and duck
Aldicarb sulfoxide 50 beef and duck
Aldicarb sulfone 50 beef and duck

Group B
Carbofuran 100 beef
3-Hydroxy carbofuran 100 beef
Methomyl 200 beef
Methiocarb sulfoxide 200 beef
Methiocarb 200 beef
Bufencarb 200 beef
Carbaryl 1000 beef

Group C
Carbaryl 50 00 poultry

Group D
Oxamyl 200 beef
Methomyl 200 beef
TIME (MONTHS AFTER FORTIFICATION)
Methiocarb sulfoxide 200 beef
□ Sulfoxide/Beef + Sulfoxide/Duck 0 Sulfone/Beef A Sulfone/Duck
Dioxacarb 200 beef
X Aldlcarb/beef M Aldicarb/Duck
Propoxur 200 beef
Bendiocarb 200 beef
Isoprocarb 200 beef F ig u re 1. A n a ly te d e p le tio n c u r v e s o f a ld ic a rb , a ld ic a r b
Promecarb 200 beef s u lfo x id e , a n d a ld ic a r b s u lfo n e , f o rtifie d a t 5 0 p p b in
b e e f a n d d u c k liver.

level (10 ppb) established by the Environmental Protection R eagents

Agency (EPA) for aldicarb in animal tissues. In the present sta­
bility experiment, we chose a concentration of each compound Carbamate standards.—Aldicarb sulfoxide, oxamyl, aldi­
for tissue fortification that was 5 to 1Ox greater than either the carb sulfone, methomyl, methiocarb sulfoxide, 3-hydroxycar-
EPA tolerance level of that compound (if established) or the bofuran, dioxacarb, aldicarb, propoxur, carbofuran, bendio-
highest concentration used in the original method (5) (if the carb, carbaryl, isoprocarb, methiocarb, promecarb, and
bufencarb standards were EPA/U.S. Food and Drug Admini­
EPA tolerance level was not established). Our aim was to detect
stration Reference Standards (EPA Pesticides and Industrial
any measurable loss over the 6 -month period of the study.
Chemical Repository, MD-8 , Research Triangle Park, NC
The fortified samples for the 3 residue fortification groups
27711). Reagents were used as previously described (5), except
(A, B, and C) were prepared by adding the appropriate fortifi­ for preparations of the LC working standards and the fortifica­
cation standard (A, B, or C). For each residue group and animal tion standards from the stock solutions. These preparation are
species, a minimum of 4 fortified samples were analyzed at detailed below.
each point in time to compensate for variability in the analytical
method. Preparation of LC Working Standards

Experimental (a) Stock standards (0.5 mg/mL o f each compound).—Pre­

pared as described previously (5).
(b) Working standard 1 (2.0 ng/\iL).—Transfer via syringe
Apparatus 100 pL of each stock standard (equivalent to 50 jig) of aldi­
carb, aldicarb sulfone, aldicarb sulfoxide, bufencarb, carbaryl,
Apparatus were the same as published in the previous carbofuran, 3-hydroxycarbofuran, methiocarb, methiocarb sul­
method (5), except for one additional item, a 50 mL Blue Max foxide, and methomyl into a single, 25 mL actinic volumetric
Falcon polypropylene tube (Becton Dickinson and Company, flask, and dilute to volume with degassed methanol. Swirl gen­
Lincoln Park, NJ 07035), which was used for storing fortified tly several times to mix. Store working standard in a freezer at
samples for analyte stability tests. -4°C. This mixed working standard was used to generate a lin-
 Ali E t Al .: J ournal Of AOAC International V ol . 76, No. 6,1993 1311

Table 2. Results of carbamate stability study for 10 carbamates in beef, duck, and poultry livers
Concentration (ppb) by month

Sam ple Initial

Analyte No. (D ay 1) 0.5 1 1.5 2 3 4 5 6

Aldicarb sulfoxide, 50 1 31.3 28 .9 23.0 15.2 19.3 2 0 .7 16.8 15.8 19.7

ppb in beef liver 2 31.5 23 .4 22.0 17.6 21.6 2 2 .0 19.9 17.3 16.9
3 30.6 25 .2 24.5 18.3 21.6 2 3 .5 2 0 .3 17.8 16.7
4 35.9 23.1 23 .8 17.3 23.9 22 .6 2 6 .6 21.2 18.0
M ean 32.3 25.1 23.3 17.1 21.6 2 2 .2 20 .9 18.0 17.8

Aldicarb sulfoxide, 50 1 21 .5 12.1 13.5 6.0 7.1 0.0 0.0 0.0 0.0
ppb in duck liver 2 22 .6 13.6 12.1 7.0 4.8 0.0 0.0 0.0 0.0
3 20.2 11.6 15.6 3.8 6.4 0.0 0.0 0.0 0.0
4 21 .0 11.1 9.4 6.0 4.7 0.0 0.0 0.0 0.0
M ean 21 .3 12.1 12.7 5.7 5.8 0.0 0.0 0.0 0.0

Aldicarb sulfone, 50 1 30.9 27 .4 19.1 12.5 16.0 11.2 9.1 4.7 7.8
ppb in beef liver 2 29.1 24 .8 19.4 15.4 17.5 13.4 10.6 7.9 4 .3
3 29 .2 2 5 .0 20 .8 18.4 17.8 18.0 9.3 6.0 5.6
4 33 .7 2 7 .9 19.9 14.7 24.3 15.3 18.9 8.3 6.6
M ean 30 .7 2 6 .3 19.8 15.3 18.9 14.5 12.0 6.7 6.1

Aldicarb sulfone, 50 1 37 .9 3 2 .7 33 .7 2 0 .0 26 .9 11.0 7.0 7.2 0.0

ppb in duck liver 2 39.2 31 .4 31.3 27.1 24 .0 7.4 4.4 5.4 0.0
3 37.5 30.1 33.1 19.6 27 .7 12.5 8.6 7.9 0 .0
4 38.8 32 .0 32.4 26 .3 21.7 6.6 5.2 10.4 0.0
M ean 38.3 31 .5 32.6 23 .3 25.1 9.4 6.3 7.7 0 .0

Methomyl, 2 0 0 ppb in 1 10.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
beef liver 2 15.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 .0
3 12.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4 12.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
M ean 12.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Methiocarb sulfoxide, 1 19.6 43 .0 76.7 24 .6 35.0 71.1 14.2 24.4 63.2

2 0 0 ppb in beef liver 2 24.1 55 .2 65 .4 4 9 .7 43 .4 59 .4 4 7 .0 22 .7 31.6
3 21.5 45 .5 55 .7 58.1 68.9 79 .4 31 .2 21.2 47.4
4 34.1 28 .5 84.1 21 .4 65.9 32 .4 4 0 .8 79.8 32.6
M ean 24.8 43 .0 70.5 38 .4 53.3 60 .6 3 3 .3 37.0 43.7

3-Hyd roxycarbofu ran, 1 83.7 76 .9 86.7 78 .8 81.4 85.2 90 .6 86 .3 86.2

100 ppb in beef liver 2 75.9 71 .3 83.0 77 .4 92.6 84 .6 100.0 91.6 87.7
3 85.8 80 .8 88.7 83 .0 97.5 98 .2 9 4 .9 90.5 87 .0
4 88 .3 82 .7 83.2 83 .2 98.0 97 .5 100.5 95.9 98.8
M ean 83 .4 77 .9 85.4 80 .6 92.4 91 .4 96 .5 91.1 89.9

Aldicarb, 50 ppb in beef 1 41.7 42 .8 35.5 20 .3 21 .6 13.5 5.9 3.5 0.0

liver 2 41.1 39 .7 31.5 20 .8 24.2 16.2 8.8 6.1 0.0
3 42.0 41 .9 36.3 23 .5 24.6 2 1 .0 9.3 4.8 0.0
4 41.6 41 .4 33.3 22 .3 27.5 15.1 13.1 5.0 0.0
M ean 41 .6 41 .4 34.1 21 .7 24.5 16.5 9.3 4.9 0.0

1 42 .7 33 .6 33.0 20 .9 8.0 3.7 3.6 0.0 0.0

Aldicarb, 50 ppb in
duck liver 2 46 .0 30 .7 30.5 19.7 9.0 4.3 10.5 0.0 0.0

3 38.9 32 .0 21.3 15.5 5.9 5.2 3.1 0.0 0.0

4 43 .9 35.1 29.0 7.3 5.3 10.8 4.0 0.0 0.0

M ean 42.9 32 .9 28.5 15.9 7.1 6.0 5.3 0.0 0.0

1312 Ali E t A l .: J ournal O f AOAC International V o l . 76, No. 6 , 1993

Table 2. (Continued)
Concentration (ppb) by month

Sam ple Initial

Analyte No. ( D a y 1) 0.5 1 1.5 2 3 4 5 6

Carbofuran, 100 ppb in 1 80.2 80 .9 82.2 68.8 71.7 72 .3 64.1 62.5 59 .4

beef liver 2 79.1 76 .8 78.5 69 .5 80.0 72.0 70.3 66 .8 53 .7

3 82.6 80 .9 81.6 73.6 86.3 83.1 68.1 65 .8 52.1

4 90.2 86 .2 76.8 72 .0 87.9 82.2 83.8 69.1 62 .3

M ean 83.0 81 .2 79.8 71 .0 81.5 77 .4 71.6 66.1 56.9

1 577.2 64 9.5 68 5.9 55 5.2 59 3.0 62 2.8 529.2 417.9 45 6.2

Carbaryl, 1000 ppb in
beef liver 2 55 3.2 634.1 661.1 59 6.4 670.7 611.5 546.2 445.1 40 6.8

3 548.6 63 6.3 68 8.4 642.8 730.1 694.0 530.2 460.0 38 9.8

4 63 8.3 66 1.6 65 9.8 62 7.6 749.9 66 0.0 690.5 506.0 4 9 2 .2

M ean 57 9.3 6 4 5 .4 67 3.8 60 5.5 68 5.9 647.1 57 4.0 457.2 43 6.2

Carbaryl, 5 0 0 0 ppb in 1 3786 45 02 48 96 41 52 46 45 3442 3749 2213 1897

chicken liver 2 39 63 4875 4723 41 49 4537 43 14 4305 24 05 2148

3 40 08 44 63 4741 37 28 48 36 3951 4130 24 86 23 26

4 40 58 40 65 4785 39 27 4747 41 20 4127 2451 2088

M ean 3954 44 76 4786 39 89 4692 3957 40 78 23 89 2115

Methiocarb, 2 0 0 ppb in 1 303.1 24 7.9 2 7 4 .3 28 2.4 28 7.3 27 8.6 348.8 336.4 29 0.3
beef liver3 2 26 6.2 22 0.8 26 6.3 25 5.5 33 7.4 282.7 362.2 358.3 30 0.6

3 310.8 27 1.2 30 0.5 26 6.2 322.5 335.4 35 2.9 356.9 2 7 0 .7

4 29 5.5 30 8.5 251.1 31 0.9 33 1.7 372.9 423.9 353.8 34 2.2

M ean 29 3.9 262.1 273.1 27 8.7 31 9.7 31 7.4 372.0 351.4 30 1.0

Bufencarb, 2 0 0 ppb in 1 156.3 161.7 155.3 129.1 134.4 127.1 113.3 107.9 103.7
beef liver 2 159.1 154.2 152.4 130.7 150.4 124.1 126.8 114.8 88 .0
3 160.0 161.7 157.0 132.3 165.1 151.4 123.3 115.4 83 .2
4 182.4 173.1 149.0 132.6 163.4 145.6 154.2 126.9 102.5
M ean 164.5 162.7 153.4 131.2 153.3 137.0 129.4 116.3 94 .3

a Methiocarb levels rose to approximately 1.5 tim es the Initial fortification level and remained at that level for the duration of the study.

ear standard curve for each of the 1 0 carbamates that were used methiocarb, methiocarb sulfoxide, and bufencarb), and 10 mL
in the original LC method (5). of the carbaryl stock standard solution into a single, 50 mL
(c) Working standard 2 (2.0 ng/\xL).—A second working actinic volumetric flask. Dilute to volume with ethyl acetate. A
standard was prepared in the same manner as working standard 21 g beef liver sample fortified with 105 pL of this solution
1 for oxamyl, dioxacarb, propoxur, bendiocarb, isoprocarb, results in concentrations of 100 ppb for carbofuran and 3-hy-
promecarb, methomyl, and methiocarb sulfoxide. Working droxycarbofuran; 2 0 0 ppb for methomyl, methiocarb, methio­
standard 2 was used to generate linear standard curves for each carb sulfoxide, and bufencarb; and 1 0 0 0 ppb for carbaryl.
of these 8 residues.
(c) Fortification standard C (1 \ig/\iL).—Transfer 10 mL
carbaryl stock standard into a 15 mL graduated test tube,
Preparation o f Fortification Standards
evaporate to ca 1 mL on N-Evap at 30°C, and reconstitute to
5 mL with ethylacetate. A 21 g poultry liver sample fortified
(a) Fortification standard A ( 10.0 [ig/mL).—Transfer 1 mL
each of 3 stock standard solutions (0.5 mg/mL) of aldicarb, with 105 ¡j L of this solution results in a carbaryl concentration
aldicarb sulfoxide, and aldicarb sulfone into a single, 50 mL of 5000 ppb.
actinic volumetric flask, and dilute to volume with ethyl ace­ (d) Fortification standard D (40 \lg/mL).—Transfer 4 mL
tate. A 21 g beef or duck liver sample fortified with 105 pL of each of stock standards solutions of oxamyl, methomyl,
this solution results in concentrations of 50 ppb of each of these methiocarb sulfoxide, dioxacarb, propoxur, bendiocarb, iso­
3 residues. procarb, and promecarb into a single, 50 mL actinic volumetric
(b) Fortification standard B.—Transfer 2 mL each of 2 flask. Dilute to volume with ethylacetate. A21 g beef liver sam­
stock standard solutions (carbofuran and 3-hydroxy carbo- ple fortified with 105 pL of this solution results in a concentra­
furan), 4 mL each of 4 stock standard solutions (methomyl, tion of 2 0 0 ppb for each of these 8 residues.
 Ali E t A l .: J ournal O f AOAC International V ol . 76, No. 6,1993 1313

Table 3. Results of carbamate stability study for 8 carbamates in beef liver

Concentration (ppb) by month

Sam ple Initial

Analyte No. (D ay 1) 0.5 1 2 3 6

Oxam yl, 20 0 ppb 1 16.9 0.0 0.0 0.0 0.0 0.0

2 13.0 0.0 0.0 0.0 0.0 0.0
3 0.0 0.0 0.0 0.0 0.0 0.0
4 0.0 0.0 0.0 0.0 0.0 0.0
M ean 7.5 0.0 0.0 0.0 0.0 0.0

Methomyl, 2 0 0 ppb 1 87.4 0.0 0.0 0.0 0.0 0.0

2 91.4 0.0 0.0 0.0 0.0 0.0
3 92.8 0.0 0.0 0.0 0.0 0.0
4 78.1 0.0 0.0 0.0 0.0 0.0
M ean 87.4 0.0 0.0 0.0 0.0 0.0

Methiocarb sulfoxide, 2 0 0 ppb 1 90.1 158.3 141.5 95.7 85.4 67.3

2 132.8 145.5 78.8 138.5 89.6 69.9
3 144.2 149.2 129.4 139.8 70.1 112.9
4 139.7 135.9 123.9 105.0 83.3 102.0
M ean 126.7 147.2 118.4 119.8 82.1 88.0

Dioxacarb, 2 0 0 ppb 1 162.2 105.1 58.1 31 .7 15.8 0.0

2 214.1 96.5 46.5 19.6 15.8 0.0
3 198.1 114.8 56.8 19.6 14.4 0.0
4 187.0 116.0 48.9 29.2 19.8 0.0
M ean 190.4 108.1 52.6 25.0 16.5 0.0

Propoxur, 20 0 ppb 1 163.7 110.6 58.3 40.7 12.3 20.7

2 2 1 8 .4 98.2 50.4 25.4 12.8 0.0
3 198.9 122.8 66.4 25.4 12.8 0.0
4 183.7 116.7 51.3 37.4 18.2 0.0
M ean 191.2 112.1 56.6 32.2 14.0 5.2

Bendiocarb, 2 0 0 ppb 1 151.8 114.7 105.9 64.0 30.2 24.0

2 20 1.2 102.9 60.0 36 .7 47.0 23.0
3 178.6 125.0 77.0 39.8 16.2 20.3
4 163.1 118.6 72.9 52.6 33.5 0.0
M ean 173.7 115.3 79.0 48.3 31.7 16.8

Isoprocarb, 2 0 0 ppb 1 174.6 187.8 155.8 162.5 76.0 66.2

2 229.1 151.4 121.3 117.2 113.0 78.3
3 211.2 186.8 150.1 117.8 65.8 72.4
4 197.2 171.1 141.2 142.9 92.4 69.2
M ean 20 3.0 174.3 142.1 135.1 86.8 71.5

Promecarb, 2 0 0 ppb 1 150.6 103.4 89.1 48.8 14.4 0.0

2 196.7 96.6 46 .7 29.6 36.0 0.0
3 171.3 111.0 65.4 32.4 9.0 0.0
4 156.9 106.2 55.4 38.3 22.2 0.0
M ean 168.9 104.3 64.2 37.3 20.4 0.0
1314 A li E t A l .: J ournal Of AOAC International V ol . 76, No. 6,1993
( p p b )
C O N C E N T R A T IO N

TIME (MONTHS AFTER FORTITICATION) TIME (MONTHS AFTER FORTIFICATION)

0 Corbofuran + 3-Hydroxycarbofuran D Methomyl + Methiocarb Sulfox. 0 Methiocorb A Bufencarb

F ig u re 2. A n a ly te d e p le tio n c u r v e s o f c a rb o fu ra n a n d F ig u r e 3. A n a ly te d e p le tio n c u r v e s o f m e th io c a rb ,
3 -h y d ro x y c a rb o fu rn , f o rtifie d a t 100 p p b in b e e f live r. m e th io c a rb s u lfo x id e , m e th o m y l, a n d b u fe n c a rb ,
fo rtifie d at 2 0 0 p p b in b e e f liver.

Preparation o f Fortified Sam ples included in the stability test for 6 additional carbamates (6 ).
This second study was conducted with an abbreviated version
(a) Fortification group A .— Blank beef and duck liver sam­ of the earlier study (see fortified samples A-C above) and used
ples (21 g each in a 50 mL screw-capped polypropylene tube) only blank beef fiver samples. Each sample was fortified with
were fortified with 105 |xL fortification standard A (50 ppb 105 (iL (equivalent to 200 ppb) fortification standard D
aldicarb, aldicarb sulfoxide, and aldicarb sulfone). For the first ( 2 0 0 ppb oxamyl, methomyl, methiocarb sulfoxide, dioxacarb,
study, which involved 9 points in time, 36 fortified samples for propoxur, bendiocarb, isoprocarb, and promecarb) and ana­
each species were prepared in addition to 18 blank samples. lyzed at amended time intervals of 1 day and 0.5,1,2,3, and 6
The fortified samples and the blank samples were stored con­ months. For the second study, a total of 20 fortified samples and
tinuously at -4°C until each was analyzed. 10 blank samples were prepared. The fortified samples and the
(b) Fortification group B.—Blank beef liver samples (21 g blank samples were stored at -4°C until each was analyzed.
each in a 50 mL screw-capped polypropylene tube) were forti­
fied with 105 (xL fortification standard B (100 ppb carbofuran Sam ple Extraction
and 3-hydroxy carbofuran; 200 ppb methomyl, methiocarb,
methiocarb sulfoxide, and bufencarb; and 1 0 0 0 ppb carbaryl). At the end of each time interval, 4 fortified fiver samples
A total of 36 fortified samples and 18 blank samples were pre­ from each species and fortification group and 2 blank fiver
pared and stored at -4°C. until each was analyzed. samples were transferred quantitatively to homogenizer flasks
(c) Fortification group C.—Blank poultry liver samples with 60 g anhydrous Na2 S 0 4 and 200 mL methylene chloride
(21 g each in a 50 mL polypropylene tube) were fortified with in the following way: ca 20 g Na2 S 0 4 was added to a sample in
105 |iL fortification standard C (5000 ppb carbaryl). A total of a polypropylene tube, and the contents of the tube were mixed
36 fortified samples and 18 blank samples were prepared and with a spatula and transferred to a homogenizer flask until all
stored at -4°C until each was analyzed. 60 g anhydrous Na2 S 0 4 was used. The sample tube was then
(d) Fortification group D.—Stability results for methomyl washed repeatedly with methylene chloride until all 200 mL
and methiocarb sulfoxide in fortified sample B were not con­ was used. The washings were added to the homogenizer flask
clusive (see Results and Discussion). Consequently, stability containing the sample mixed with Na2 S 0 4. For each set of
tests on methomyl and methiocarb sulfoxide were repeated and samples, 1 of the blank samples was fortified at room tempera-
 Ali E t A l .: J ournal O f AOAC International V ol . 76, No. 6,1993 1315

TIME (MONTHS AFTER FORTIFICATION)

TIME (MONTHS AFTER FORTIFICATION) 0 Oxamyl + DIoxacarb A Propoxur A Bendiocarb X Isoprocarb
□ Carbaryl/Beef + Carbaryl/Chicken V Promecarb

F ig u r e 4. A n a ly te d e p le tio n c u r v e s o f c a r b a r y l in b e e f F ig u re 5. A n a ly te d e p le tio n c u r v e s o f o x a m y l,
a n d c h ic k e n liv e r, fo rtifie d a t 5 0 0 p p b in c h ic k e n a n d d io x a c a r b , p r o p o x u r b e n d io c a r b , is o p r o c a r b , a n d
1000 p p b in beef. p ro m e c a rb , fo rtifie d a t 2 0 0 p p b in b e e f liver.

ture with 105 (iL (equivalent to 10 ppb) of the corresponding compounds cited in the original method (5) (Table 2). The resi­
LC working standard immediately before analysis to act as a due values represent the mean analyte concentration at the end
calibration standard for correcting sample recovery. Samples of each time interval. Anomalies in the results for methomyl
were extracted and purified following the LC method (5). and methiocarb were found. Methomyl was depleted quickly,
and methiocarb increased, possibly because methiocarb sul­
Determination o f R esidue Concentrations
foxide was reduced to methiocarb. To verify this finding,
methomyl and methiocarb sulfoxide were included in the sub­
At the end of each time interval, 4 fortified sample extracts
sequent analyte stability study, along with 6 additional car­
from each residue group and species, a blank liver sample pre­
viously analyzed and found to contain no carbamates (to act as bamates ( 6 ) (see group D in Table 1). The fortified samples for
a negative control), and another blank liver sample fortified at the second study were prepared by adding 2 0 0 ppb fortification
the time of analysis with 1 0 ppb of the corresponding working standard D to beef liver only, and an amended sampling proto­
standard (to act as a calibration standard for correcting sample col was used. Table 3 shows the results of this second experiment.
recovery) were analyzed. A 3-point regression curve was gen­ The preparation of each batch of fortified samples, both for
erated for each compound in each group by analyzing the ap­ the initial carbamate analyte stability study on 1 0 compounds
propriate working standard with the LC method (5). When nec­ (Table 2) and for the subsequent study on 6 additional car­
essary, the sample extracts were diluted to accommodate the bamates (Table 3), took almost an entire day. Therefore, forti­
peak heights in the chromatogram. The concentration of each fied samples were exposed to room temperature for approxi­
compound in parts per billion was calculated by using the mately 8 h before they were transferred to the freezer for
1 0 ppb calibration standard and the appropriate dilution factor. storage. The baseline samples for initial analysis were analyzed
A mean value was calculated from the 4 determinations. the next day, one day following sample fortification (day 1). A
significant depletion of certain carbamates was observed on
Results and Discussion day 1 compared with the recoveries obtained immediately after
fortification, as reported earlier (5). The average recovery of
The first stability study analyzed 3 fortification samples (A- carbamates immediately after fortification was greater than
C) at 9 time intervals (Table 1) and was conducted on the 10 80%, with a coefficient of variation less than 17%. Compared
1316 Ali E t Al .: J ournal Of AOAC International V ol . 76, No. 6,1993

with the initial loss, the subsequent rate of depletion was This study provided valuable information regarding the sta­
smaller. The initial residue depletion may result from enzy­ bility of 16 N-methylcarbamates in liver tissues of beef, swine,
matic activity within the fortified tissues at room temperature. and poultry. The tissues are analyzed as part of the FSIS Na­
The results of the 2 experiments are shown in the analyte tional Residue Plan. The significant initial loss of certain car­
depletion curves of Figures 1-5. The rate of depletion varied bamates in tissues at room temperature due to enzymatic activ­
among carbamates. The depletion rate was greater in duck than ity indicates that cryogenic conditions are required for sample
in beef tissue for a given analyte. Methomyl and oxamyl were preparation and storage. The results of this study also provided
depleted completely within day 1 and 0.5 month. Two to 6 information on how long an evidentiary sample containing viola­
months after sample fortification, 1 0 0 % depletion of residues tive carbamate residue levels can be stored for potential litigation.
in fortified tissue was observed for 7 other carbamates (aldi-
carb, aldicarb sulfoxide, aldicarb sulfone, dioxacarb, prome- References
carb, propoxur, and bendiocarb). Increases in methiocarb con­
centration to a level 1.5 times that of the fortified sample were (1) Hill, M.K., & HoUowell, R. (1984) Anal Chem. 56, 2465-
observed in the original experiment. In the latter experiment, 2466
with the decrease of methiocarb sulfoxide concentration, a (2) Marshall, E. (1985) Science 229,1369-1371
peak was observed at the retention time of methiocarb in the (3) Belasco, J.I., & Harvey, J., Jr (1980) Am. Chem. Soc. 28,
chromatogram. A possible explanation may be that enzyme ac­ 689-692
tion caused a reversed equilibrium shift that produced free (4) Kearney, C.P. (1976) J. Assoc. Off. Anal. Chem. 59, 866-893
methiocarb from methiocarb sulfoxide. Either chemical or en­ (5) Ali, M.S. (1989) J. Assoc. Off. Anal. Chem. 72, 586-592
zymatic reduction appears to have occurred. (6 ) Ali, M.S., White, J., Bakowski, R., Stapleton, N., Williams,
K., Johnson, R., Phillippo, E., Woods, R., & Ellis, L.R.
(1993) J. AOACInt. 76,907-910
 Armishaw & M illar : J ournal O f AOAC International V ol . 76, No. 6,1993 1317

RESIDUES AND TRACE ELEMENTS

C om parison o f G el P erm eation C hrom atography, Sweep

C odistillation, and F lo ris il C olum n A dsorption C hrom atography
as Sam ple C leanup Techniques fo r the D eterm in atio n o f
O rganochlorine Pesticide Residues in A n im a l Fats

P aul Armishaw and R oderick G. M illar

Australian Government Analytical Laboratories, Department of Administrative Services, PO Box 83, Cottesloe 6011,
Western Australia

Methods using a commercial sweep codistillation ment of Primary Industries and Energy (DPIE). This program
apparatus, gel permeation chromatography (GPC), aims to ensure compliance with maximum residue levels set by
and Florisil column adsorption chromatography the Australian National Health and Medical Research Council
were compared as cleanup techniques for the de­ (1).
termination of organochlorine pesticide residues in The agricultural use of organochlorine pesticides was
animal fats by Megabore capillary gas chromatogra­ banned in Australia in 1987. However, animals may still incur
phy with electron capture detection. Animal fat that significant residues from land contaminated by previous use,
had been previously found to contain no detectable from contaminated dipping or jetting equipment, or from inad­
organochlorine residues was spiked with 17 or­ vertent exposure to organochlorine pesticides used for termite
ganochlorine pesticides at levels ranging from 0.1 control.
to 0.4 mg/kg and cleaned up by each of the 3 tech­ Operating conditions for a commercial sweep codistillation
niques. Recoveries obtained for all 3 methods were
apparatus for the cleanup of animal fats prior to the determina­
in the range 73-113%, with coefficients of variation
tion of organochlorine and organophosphorus pesticide resi­
between 1.1 and 11.2%. Equivalence of method per­
dues have been described previously (2-4). This apparatus is
formance was further demonstrated by performing
used in many Australian residue testing laboratories and is
replicate analyses of beef and sheep fat containing
gaining increasing acceptance in other countries. It is used rou­
naturally incurred residues of heptachlor epoxide,
tinely by AGAL for the determination of organochlorine and
dieldrin, p,pf- DDE, p,pf-DDD and p.p'-DDT. All 3
organophosphorus pesticides in thousands of animal fat, butter,
methods offer effective cleanup and acceptable re­
and cheese samples annually.
covery of organochlorine pesticides in animal fat.
Gel permeation chromatography (GPC) is a cleanup tech­
The sweep codistillation method has the advan­
nique recommended by Codex (5) and the AOAC (6 ) for the
tages of low solvent and reagent use, simultane­
ous cleanup of 10 samples, and rapid turnaround, cleanup of organochlorine pesticides in animal fat and is the
although thermal degradation of pp'-DDT requires subject of several recent publications (7-10).
monitoring and control. GPC offers a high degree Florisil column chromatography is a recognized cleanup
of automation but is a relatively slow sequential technique for fatty foods. Procedures are described in detail in
sample cleanup with high solvent use. Florisil col­ AOAC Official Methods o f Analysis (11) and the U.S. Food and
umn adsorption chromatography is a simple, Drug Administration Pesticide Analytical Manual (12). This
proven technique but requires considerable sol­ technique is used by AGAL as an alternative cleanup for the
vent and reagent and has a low potential for auto­ confirmation of organochlorine residues in foodstuffs.
mation. The performance of the commercial sweep codistillation
apparatus used in this study has been evaluated in interlabora­
tory proficiency studies (13) conducted by AGAL on behalf of
T he Australian Government Analytical Laboratories the DPIE. However, no study comparing its performance with
(AGAL) routinely analyze fat samples for organo­ that of other recognized cleanup techniques for animal fats has
chlorine pesticide residues as part of the National Resi­ been published.
due Survey program administered by the Australian Depart- The present study was conducted to provide comparative
performance data for the 3 cleanup techniques (GPC, sweep
Received September 10, 1992. Accepted February 11, 1993 codistillation, and Florisil column chromatography).
1318 A r m is h a w & M il l a r : J o u r n a l O f AO AC I n t e r n a t i o n a l V o l . 76, N o. 6 ,1 9 9 3

Comparative data are presented for 17 organochlorine pes­ Prepare individual stock solutions of 0.2 mg/mL in light petro­
ticides at internationally significant residue levels from spiked leum.
animal fat and from fat containing naturally incurred residues. (1) Organochlorine pesticide standard solution.—Dilute in­
dividual stock solutions to obtain the following concentrations
METHOD in light petroleum; hexachlorobenzene (HCB), lindane,
a-BHC, heptachlor, 8 -BHC, aldrin, oxychlordane, heptachlor
epoxide, /ran.v-chlordane, d.v-chlordane, /?,//-DDE, dieldrin,
Apparatus and Reagents endrin, p,p'-DDD, and /?,//-DDT, all 0.01 |ig/mL; P-BHC,
0.02 pg/mL; p,/?'-methoxychlor, 0.04 jig/mL.
(a) Sweep codistillation apparatus.— Unitrex (Universal
Trace Residue Extractor) supplied by Scientific Glass Engi­
Procedure
neering (SGE) Pty Ltd., Melbourne, Australia. This unit has
been described previously (2). Remove the glass beads from
the distillation tubes before use.
Sweep Codistillation Method
(b) Gel permeation chromatography system.—LC pump Operate the sweep codistillation apparatus as described by
(ISCO Model 2350), autoinjector (ISCO Model ISIS), fraction Luke et al. (2) except for the following modifications: (7) Use
collector (ISCO Model Foxy 200), electrically actuated 6 -port beadless distillation tubes, (2) increase sweep time to 45 min,
injection valve, 5 mL sample loop, peristaltic transfer pump (3) use 10 mL diethyl ether-light petroleum (22 + 78) to elute
(ISCO Inc., Lincoln, NE). traps.
(c) Gel column.—Glass, 50 cm x 2.5 cm id, slurry packed Collect the eluate from the traps in a 10 mL volumetric flask
with 70 g Bio Beads SX-3 (Bio-Rad Laboratories, Richmond, and adjust to volume with light petroleum. Mix thoroughly and
CA) and compressed to a bed length of 48 cm. Operate at a flow inject into the gas chromatograph.
rate of 3 mL/min. These variations from the sweep codistillation operating
(d) Gas chromatograph.—Varian 3400 fitted with 63Ni conditions described by Luke et al. (2) are not necessary for the
electron capture detector, Varian packed column injector fitted determination of organochlorine residues alone. However, they
with Megabore column installation kit (J&W Scientific, Fol­ do allow the simultaneous determination of certain organo-
som, CA, part no. 125-1100); fused silica, 30 m x 0.53 mm id phosphorus pesticide residues, including diazinon, chlorpyri-
column with 1 pm DB-1701 stationary phase (J&W Scientific, fos, bromophos ethyl, and ethion, and are used routinely in our
part No. 93408). Operating conditions: temperatures, column laboratory.
oven 170°C, temperature program at 40°C/min to 230°C, hold
at 230°C for 4 min, then temperature program at 40°C/min to GPC Method
270°C, hold at 270°C for 3 min, injector 210°C, detector 3(X)°C; (a) Calibration.—Inject 5 mL organochlorine pesticide
He carrier gas at 6 mL/min, N 2 makeup gas at 30 mL/min. standard prepared in GPC eluting solvent. Use transfer pump
(e) Solvents.—Light petroleum (Shell X4, redistilled in to load sample loop as described by Daft et al. (9). Collect six­
glass); dichloromethane, LC grade (Ajax Chemical Co., Syd­ teen 10 mL fractions from 150 to 310 mL. Transfer collected
ney, Australia); diethyl ether, nanograde (Mallinckrodt Austra­ fractions to a Kudema-Danish concentrator fitted with a 5 mL
lia Pty Ltd., Melbourne, Australia). calibrated tube. Evaporate the collected fractions to a small
(f) Sweep codistillation trap eluting solvent.—Diethyl volume; add 10 mL light petroleum during the final stages of
ether-light petroleum (22 + 78). Add 44 mL diethyl ether to a evaporation. Adjust to a final volume of 5 mL with light petro­
200 mL volumetric flask, adjust to volume with light petro­ leum, mix thoroughly, and inject into the gas chromatograph.
leum, and mix thoroughly. Determine the elution profiles of pesticides.
(g ) GPC eluting solvent.—Dichloromethane-light petro­ (b) Cleanup.—Weigh 1.0 g molten fat into a 10 mL volu­
leum (1 + 1). Add 1 L of dichloromethane to 1 L of light petro­ metric flask, adjust to volume with GPC eluting solvent, and
leum in a glass Winchester bottle and mix thoroughly. Vacuum mix thoroughly. Filter the fat solution through a 0.45 pm mem­
filter solvent through a 0.45 ¡am membrane filter prior to use. brane filter into a GPC autosampler vial. Inject 5.0 mL fat so­
(h) Florisil column eluting solvent.—Dichloromethane- lution (equivalent to 0.5 g sample) onto the GPC column and
light petroleum (20 + 80). Add 400 mL dichloromethane to collect the fraction determined from the calibration procedure.
1600 mL light petroleum in a glass Winchester bottle and mix Transfer the collected fraction to a Kudema-Danish concentra­
thoroughly. tor fitted with a 5 mL calibrated tube. Evaporate to a small
(i) Florisil.—60-100 mesh (Floridin Co., Pittsburgh, PA). volume; add 10 mL light petroleum during the final stages of
Prepared as described by Luke et al. (2). (Deactivated Florisil evaporation. Adjust to a final volume of 5 mL with light petro­
2 % was used for sweep codistillation traps and adsorption col­ leum, mix thoroughly, and inject into the gas chromatograph.
umns.)
(j) Sodium sulfate.—Granular, anhydrous. AR grade. Heat
Florisil Column Adsorption Chromatography
at 160°C for a minimum of 16 h before use. Weigh 0.5 g molten fat into a 25 mL beaker and dissolve in
(k) Reference standards.—Curator of Standards, Australian 5 mL Florisil column eluting solvent. Weigh 16.0 g Florisil into
Government Analytical Laboratories, Melbourne, Australia. a glass chromatography column and top with 1 cm of sodium
 A r m is h a w & M il l a r : Jo urnal O f AO AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3 1319

sulfate. Rinse column with 75 mL eluting solvent. Transfer the For the GPC cleanup the amount of fat carried over into the
fat solution to Florisil column; rinse the beaker with small por­ final extract was measured by collecting the pesticide-contain­
tions of eluting solvent. Elute pesticides from the column with ing fraction in an oven dried (70°C, 45 min), tared beaker. The
250 mL eluting solvent. Collect the eluate in a Kudema-Danish fraction was then evaporated to dryness on a steam bath, and
concentrator fitted with a 5 mL calibrated tube and concentrate the beaker was returned to the drying oven (70°C, 45 min). The
to 5 mL for injection into the gas chromatograph. beaker was then reweighed.
Linearity of electron capture detector response was estab­
Gas Chromatographic Determination lished (R 2 > 0.99) within the range equivalent to 0.01 to 0.2
mg/kg of residue in the sample for 15 of the 17 organochlorine
Inject 2 pL of sample extract and calibration standard into
pesticides tested and within the range 0.02 to 0.4 mg/kg for
the gas chromatograph. Identify peaks by comparison of reten­
(3-BHC, and 0.04 to 0.8 mg/kg for p,p'-methoxychlor. When
tion times and quantitate by peak height.
necessary, sample extracts were diluted to maintain detector
response within this range.
Experimental
Results and Discussion
A composite animal fat was prepared by mixing rendered
beef, sheep, and swine fat that had been analyzed previously
Recoveries fromSpiked Fat
and found to be free from detectable residues of organochlorine
pesticides. The recoveries from the 3 cleanup techniques are Usted in
A portion of this fat was spiked with 17 organochlorine pesti­ Table 1. The 17 organochlorine pesticides tested were recov­
cides at residue levels (Table 1). Portions of the spiked and un­ ered in the range 73-113% for residue levels between 0.1 and
spiked fats were then analyzed by the 3 techniques specified. 0.4 mg/kg.
The 3 cleanup techniques were also compared by analyzing Because each of the rephcate spiking trials was ran as a
samples of beef and sheep fat containing naturally incurred or­ single batch, the results in Table 1 indicate the repeatabihty of
ganochlorine pesticide residues. The beef fat sample was ana­ each cleanup method and demonstrate a comparable level of
lyzed in quadruplicate using each of the 3 cleanup techniques. precision for each method. Our laboratory has considerably
The sheep fat sample was analyzed in quadruplicate by the less experience with the GPC technique than with either sweep
sweep codistillation and GPC cleanups but not by the Llorisil codistillation or Florisil column. Experience within our labora­
column adsorption cleanup. A spiked fat was analyzed along tory over many years has shown the long-term intralaboratory
with each batch of samples, and the raw results were corrected reproducibility for both sweep codistillation and Florisil col­
for recovery (Tables 2 and 3). umn cleanups to be between 5 and 10% at the spiking levels

Table 1. Results of replicate analyses of spiked animal fat cleaned up by sweep codistillation, GPC, and Florisil
column adsorption chromatography
Sweep3 GPC3 Florisil0

Spike, Mean ree., cv, Mean ree., cv, Mean ree., CV,
Pesticide mg/kg % % % % % %
HCB 0.1 101 3.6 105 8.7 93 3.5
a-BHC 0.1 103 2.6 97 4.5 89 4.0
Lindane 0.1 102 1.7 95 3.4 93 3.1
ß-BHC 0.2 109 3.1 98 2.2 94 4.2
Heptachlor 0.1 103 5.2 98 5.2 90 5.6
8-BHC 0.1 95 3.3 100 1.9 91 3.8
Aldrin 0.1 102 2.7 84 7.2 91 4.8
Oxychlordane 0.1 105 3.6 73 11.2 93 4.6
Heptachlor epoxide 0.1 102 2.2 96 3.2 91 4.3
frans-Chlordane 0.1 112 10.2 100 2.2 93 3.6
c/s-Chlordane 0.1 108 2.7 100 3.2 93 3.7
p,p'-DDE 0.1 113 1.6 104 4.1 93 4.6
Dieldrin 0.1 107 1.1 102 3.2 97 5.4
Endrin 0.1 103 2.6 101 2.6 92 4.9
p,p'-DDD 0.1 111 2.8 103 3.4 97 6.5
p,p'-DDT 0.1 94 3.5 101 4.3 97 2.9
p.p'-Methoxychlor 0.4 96 4.2 102 6.5 83 4.4

a Six replicate analyses.

6 Five replicate analyses.
1320 A r m is h a w & M il l a r : J o u r n a l O f A O AC I n t e r n a t i o n a l V o l . 76, N o. 6 ,1 9 9 3

used here. On the basis of the repeatability of results obtained in our laboratory has shown that by removing the glass beads
in this study, we would expect a similar level of reproducibility from the distillation tubes, it is possible to extend to over 1 0 0
for the GPC cleanup. the number of analyses performed with each tube before acid
For the GPC cleanup, the observed recoveries were consis­ washing and silanizing are required.
tent with those obtained by Ault et al. (8 ) in their validation
studies of the official AOAC method (7). The somewhat lower
Cleanup Efficiency
mean recoveries observed for aldrin (84%) and oxychlordane Efficiency of sample cleanup, measured in terms of fat car­
(73%) in our GPC cleanup are consistent with the use of a ryover and “cleanliness” of gas chromatograms is excellent for
slightly long dump time, because these are the first pesticides all 3 methods. Figure 1 illustrates electron capture detector gas
eluted from the GPC column. When the dump time was short­ chromatograms of composite, residue-free animal fat cleaned
ened to improve recovery of aldrin and oxychlordane, an in­ up by each of the 3 techniques. Sweep codistillation and
crease in interferences in the gas chromatogram was observed. Florisil column cleanups provide a slightly cleaner chromato­
These interferences were presumably caused by increased gram, particularly in the early part of the trace. Figure 2 reveals
overlap of the fat elution band into the collected fraction. No some minor interferences (particularly for HCB) in the chro­
correction for interferences was made when calculating recov­ matogram of the GPC extract.
eries. Luke et al. (2) reported less than 2 mg fat carryover into the
For sweep codistillation cleanup, the slightly high mean re­ final sample extract with the sweep codistillation method. We
coveries observed forp,//-DDE (113%) and p,p'-DDD (111%) measured fat carryover for beef, sheep, swine, goat, and poultry
coupled with a correspondingly low mean recovery of p ,p - fat with the GPC technique to be less than 2 mg. No deteriora­
DDT (94%) is consistent with minor breakdown of p,//-DDT tion of gas chromatograph performance, caused by lipid con­
to p ,p -DDE and /?,//-DDD in the hot distillation tubes. This tamination of the column or injection port, was observed dur­
degradation effect is statistically significant (Student’s two- ing this work.
tailed t test p < 0.05) when compared with the ambient tem­
Comparison of Method Performance for Samples
perature GPC and Florisil column techniques. However, the
Containing Naturally Incurred Residues
results for p ,p -DDT and its breakdown products are generally
summed and reported as total DDT analogues. Thus minor The beef fat sample selected for this study contained repre­
breakdown of DDT is of little practical significance. The de­ sentative levels of p,//-DDE, dieldrin, and heptachlor epoxide,
gree of thermal breakdown is controlled by regular monitoring organochlorine pesticides commonly found in animal fat. Raw
of p,p -DDT recoveries and by silanizing the distillation tubes results for the beef fat sample (Table 2) are in excellent agree­
when significant p ,p -DDT breakdown is observed. Experience ment for the 3 cleanups. Corrected recovery results show a

----------------------------------- 1----------------------------------------1------------------ ------------------- i

1

0 3 6 9
Time (min)
Figure 1. ECD gas chromatograms of residue-free fat cleaned up by GPC (A), sweep codistillation (B), and Florisil
column adsorption chromatography (C).
 A r m is h a w & M il l a r : Jo urnal O f AO A C I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3 1321

1----------------------------------------- ------------------------------ ------------ 1--------------------------------------- _ j

0 3 6 9
Time (min)
Figure 2. ECD gas chromatograms of residue-free fat cleaned up by GPC (A), and organochlorine pesticide standard
(B). Peak identification: 1, HCB; 2, a-BHC; 3, lindane; 4, heptachlor; 5, aldrin; 6, (3-BHC; 7, 5-BHC; 8, oxychlordane; 9,
heptachlor epoxide; 10, frans-chlordane; 11, cts-chlordane; 12, p , f i - DDE; 13, dieldrin; 14, endrin; 15, p ,p f- DDD; 16,
p,/f-DDT; 17, p,p'-methoxychlor. All 20 pg, except for p-BHC (40 pg) and pp'-methoxychlor (80 pg).

greater diversity consistent with the practice of batch recovery between the sweep codistillation and GPC cleanups for p,p'-
correction based on the result from a single spike. Batch recov­ DDE and p,//-DDD, but a lower result for p,p'-DDT by sweep
ery correction is specified by DPIE for testing the fat in Austra­ codistillation. This result is consistent with the typically lower
lian export meat. recovery obtained for p,p'-DDT by sweep codistillation as pre­
The sheep fat selected for this study contained repre­ viously discussed. When the raw results for individual DDT
sentative levels of p.p'-DDT and its metabolites. This enabled analogues are summed, no statistically significant difference
a direct comparison of the performance of the GPC and sweep between the sweep codistillation and GPC results is observed
codistillation cleanup techniques for these pesticides. Raw re­ (Student’s two-tailed t test; p < 0.01). The recovery corrected
sults for the sheep fat sample (Table 3) show good agreement results are in close agreement with the corresponding GPC results.

Table 2. Results of replicate analyses of a beef fat containing naturally incurred organochlorine pesticide residues
after cleanup by codistillation, GPC, and Florisil column adsorption chromatography
Corrected result,
Cleanup technique3 Pesticides Raw result, mg/kg CV, % mg/kg Recovery,

Sweep Fleptachlor epoxide 0.07 5.8 0.06 108

p.p'-D D E 0.57 2.6 0.49 116
Dieldrin 0.21 2.5 0.18 113

GPC Heptachlor epoxide 0.06 8.1 0.0 7 92

p,p'-D D E 0.60 5.5 0.57 104
Dieldrin 0.21 6.4 0.21 97

Florisil Heptachlor epoxide 0.06 7.9 0.06 91

p.p'-D D E 0.58 5.5 0.63 93
Dieldrin 0.20 7.0 0.21 97

a For each techique, 4 replicate analyses were performed.

b Corrected for batch recovery based on a single spike.
1322 A r m is h a w & M il l a r : Jo urnal O f A O AC I n t e r n a t i o n a l V o l . 76, N o. 6 ,1 9 9 3

Table 3. Results of replicate analyses of a sheep fat containing naturally incurred organochiorine pesticide residues
after cleanup by codistillation and GPC
Corrected result,
Cleanup technique3 Pesticides Raw result, mg/kg CV, % mg/kgb Recovery, %
Sweep p.p'-DDE 0.81 2.4 0.84 97
p.p'-DDD 0.23 5.2 0.24 97
p.p'-DDT 0.60 1.8 0.65 93
Total DDT 1.66 2.1 1.73 NAC

GPC p.p'-DDE 0.87 7.5 0.84 104

p.p'-DDD 0.22 3.0 0.22 100
p.p'-DDT 0.71 7.5 0.66 107
Total DDT 1.79 7.0 1.72 NAC

a For each techique, 4 replicate analyses were performed.

b Corrected for batch recovery based on a single spike.
c Not applicable.

Comparison of Operational Features low capital cost with the important advantage of greatly re­
duced solvent use.
The 3 cleanup techniques give equivalent results for the de­
termination of organochiorine pesticide residues in animal fats. Acknowledgments
However, for the regulatory laboratory chemist, each technique
has advantages and disadvantages. The authors wish to thank Mr. A. Hirst for performing some
Advantages of the sweep codistillation method include low of the sweep codistillation analyses and Mr. R. Hogg for help­
solvent use, no use of chlorinated solvents, no evaporation of ful assistance with proofreading the manuscript.
solvents, and relatively low capital equipment costs. The tech­
nique is simple in operation and offers rapid sample turnaround References
and high sample throughput in comparison with the other
methods. A batch of 10 samples can be cleaned up in 1 h. How­ (1) National Food Authority (1992) Australian Food Standards Code,
ever, the method is not automated and requires the full-time Australian Government Publishing Service, Canberra, Australia
attention of the analyst. AGAL has routinely used sweep codis­ (2) Luke, B.G., Richards, J.C., & Dawes, E.F. (1984) J. Assoc.
tillation cleanup to analyze fat samples taken from animals bound Off. Anal. Chem. 67, 295-298
for export when a turnaround time of less than 24 h is required. (3) Luke, B.G., & Richards, J.C. (1984) J. Assoc. Off. Anal.
The GPC method has been highly automated, with commer­ Chem. 67, 902-904
cial systems available. However, these systems are expensive. (4) Brown, R.L., Fanner, C.N., & Millar, R.G. (1987) J. Assoc.
The fact that GPC is a sequential cleanup procedure places Off. Anal. Chem. 70, 442—445
limitations on both the number of samples that can be proc­ (5) Codex Alimentarius Commission (1989) Recommendations
essed in a given period, and on the minimum achievable turn­ for Methods of Analysis of Pesticide Residues, Document
around time. The high solvent use, with attendant disposal CAC/PR8-1989, Food and Agriculture Organization, Rome,
problems, is also a disadvantage, although a recent study ( 1 0 ) Italy
reports a GPC cleanup that reduces solvent use by 85%. The (6 ) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
automated nature of GPC systems makes them well suited to lington, VA secs. 984.21A-F
laboratories with a constant load of samples with a relatively (7) Ault, J.A., & Spurgeon, T.E. (1984) J. Assoc. Off. Anal.
long turnaround time and when the saving in labor can com­ Chem. 67, 284-286
pensate for the capital investment and the solvent purchase and (8 ) Ault, J.A., Spurgeon, T.E., Gillard, S., & Mallinson, E.T.
(1985) J. Assoc. Off. Anal. Chem. 6 8 , 941-944
disposal costs incurred.
(9) Daft, J., Hopper, M., Hensly, D., & Sisk, R. (1990) J. Assoc.
Florisil column adsorption chromatography combines high
Off. Anal. Chem. 73, 992-994
solvent and reagent use and a low potential for automation. It
(10) Patterson, J.R. (1991) J. Assoc. Off. Anal. Chem. 74, 1016-
is, however, a rapid, proven, and reliable cleanup technique,
1018
which is well suited for confirmatory analyses.
(11) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
lington, VA, sec. 970.520
Conclusion (12) Pesticide Analytical Manual (1968) 2nd Ed., Vol. 1, U.S.
Food and Dmg Administration, Washington, DC
Results of the sweep codistillation cleanup are equivalent to (13) Hotchin, M. (1987-1991) National Pesticide Proficiency
those of the GPC and Florisil column adsorption cleanups. The Testing Program, Reports of Studies No. 1 to 7, Australian
sweep codistillation technique combines rapid analysis and Government Analytical Laboratories, Sydney, Australia
 C h ic h il a & G il v y d is : Jo urnal O f AO AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3 1323

RESIDUES AND TRACE ELEMENTS

D e te rm in a tio n o f P a r a q u a t a n d D iq u a t in L o w -M o is tu re F o o d

C r o p s U s in g S ilic a C o lu m n C le a n u p a n d L iq u id

C h ro m a to g ra p h y w ith U V D e te c tio n

T ina M.P. Chichila and D alia M. G ilvydis

U.S. Food and Drag Administration, Pesticides and Industrial Chemicals Research Center, Detroit, MI 48207

A sample cleanup method was developed for the METHOD

determination of paraquat (PQ) and diquat (DQ) in
low-moisture food crops. Low-moisture commodi­ Reagents
ties, such as milled dry navy beans, are digested in
(a) Solvents.—Methanol and acetonitrile (Omnisolv, EM
acid. PQ and DQ are isolated from the digestates
Science, Gibbstown, NJ 08027).
by using a 4 g column of preconditioned silica gel.
(b) LC grade water.—Obtained from Milli-Q water purifi­
The analytes are concentrated and then deter­ cation system (resistivity, 15-18 megohms/cm) (Millipore
mined by liquid chromatography with a silica ana­ Corp., Bedford, MA 01730).
lytical column, sodium chloride as an ion-pairing re­ (c) Column chromatography solutions.—Prepare the fol­
agent, and acetonitrile as an organic modifier. PD lowing aqueous solutions with LC grade water: Solution A,
and DQ are determined simultaneously with a di­ 0.1M NaOH; Solution B, 0.1M HC1 in methanol; Solution C,
ode array UV absorbance detector. Recoveries for 7 parts methanol and 3 parts 6.5M HC1.
PQ and DQ were determined on 3 different fortified (d) LC mobile phase.—Dissolve 5.0 g NaCl in 600 mL LC
low-moisture crops. Fortification levels ranged grade water previously adjusted to pH 2.2 with 1M HC1. Add
from 0.01 to 0.30 ppm; average recoveries ranged 400 mL acetonitrile, mix, and filter through Millipore Type DV,
from 47.5% (DQ) to 95.3% (PQ). 0.65 pm nylon filters. Degas before use by sparging with he­
lium.
(e) Reference standard solutions.—Paraquat dichloride
and diquat dibromide were obtained from U.S. Environmental
P araquat (PQ) and diquat (DQ) are quaternary ammonium
Protection Agency (EPA), Pesticides and Industrial Chemicals
compounds that are registered in the United States for
Repository, Research Triangle Park, NC 27711. Use plas-
many agricultural applications. For instance, they are
ticware or silanized glassware for preparation of standard solu­
commonly used as general herbicides and preharvest desic­
tions. Use methanol as the diluent throughout. Use the follow­
cants on many food crops (1, 2). We proposed a method for
ing equations to calculate the amount of PQ and DQ salts
their determination in high moisture food commodities, such as
necessary for the appropriate concentration of cations (C) in
potatoes and com (3). This method used a pH-controlled silica solution.
solid-phase extraction (SPE) for sample cleanup and isolation For DQ dibromide (MW, 344.07):
of PQ and DQ from the sample matrix. Commercially available
344 07
SPE cartridges containing 400 mg silica gel were used. After DQ dibromide salt, mg = [DQ (C) needed, mg] x —
1o 4 . Z j
the SPE procedure, the analytes were determined by liquid
For PQ dichloride (MW, 257.16):
chromatography (LC) with UV detection. However, the SPE
25 V 16
procedure was not adequate for low-moisture commodities, PQ chloride salt, mg = [PQ (C) needed, mg] x '
186.25
such as dry beans and whole wheat flour. Recoveries were low
and irreproducible, particularly for DQ. The cartridges were Prepare a stock solution of each standard at a concentration
of 100 ng/pL. Prepare mixed standard solutions by serial dilu­
also easily plugged by sample fines. As a result, the SPE pro­
tion of combined aliquots of stock solutions. Store all standard
cedure was modified for low-moisture commodities. The silica
solutions under refrigeration, and equilibrate at room tempera­
cartridges were replaced by 4 g columns of preconditioned,
ture before use. Prepare a working reference standard solution
bulk silica gel. PQ and DQ were determined by the same LC
by evaporating a 1 mL aliquot of a mixed standard solution to
system described in the original method for high-moisture
dryness at 50X1 under a stream of nitrogen. Dissolve the resi­
commodities.
due in 1 mL LC mobile phase.
(f) Cleaner.— Micro Laboratory Cleaner (International
Received October 29, 1992. Accepted January 26, 1993. Products Corp., Trenton, NJ 08601).
1324 C hichila & G ilvydis : Journal O f A O AC International V ol . 76, N o . 6 ,1 9 9 3

GeneralApparatus LC Operation
(a) Vortex mixer.—Scientific Products, McGraw Park, IL Set column oven temperature to 40°C. Equilibrate system
60085. with helium-sparged mobile phase at 1.0 mL/min for 15 min or
(b) Centrifuge.— Must accommodate Coming 50 mL cen­ until retention times for PQ and DQ are reproducible (ca 5 and
trifuge tubes; 2 0 0 0 rpm minimum, with swing-out horizontal 5.6 min for DQ and PQ, respectively). Thoroughly rinse system
with 20% acetonitrile in water before shutting down systemfo r
rotors.
extended periods, e.g., overnight. Maintain slow flow (0.1-0.2
(c) Food mill.—Retsch KG, 5657 Hahn, Germany; process
mL/min) when system is not in use for shorter periods to pre­
dry food crops to pass through 20-40 mesh sieve.
vent salt deposition. Program diode array detector for the quan­
(d) pH measurement.—pH meter (Fisher Scientific,
titation of PQ and DQ at maximum absorbance wavelengths of
Springfield, NJ 07081).
257 and 310 nm, respectively. Adjust sensitivity of detector so
(e) Filter paper.—Sharkskin analytical paper, 5.5 cm that 1 0 ng of each analyte on-column produces 2 0 % full-scale
(Schleicher & Schuell, Keene, NH 03431). response. (This was achieved at a setting of 0.01 AUFS by us­
(f) Polypropylene containers.—Coming 250 mL beakers ing the specified system.)
and 50 mL centrifuge tubes with conical bottoms (Fisher Sci­
entific). Preparation of Silica Gel for Column
(g) Pyrex test tubes.— 100 mL, 159 x 33 mm (Fisher Sci­ Chromatography
entific).
Weigh 100 g EM Science bulk silica gel into 1L beaker. Add
(h) Filters.—For column chromatography, use 20 ¡urn po­ 1 L LC-grade water. Mix with glass stirring rod. Pour silica
rous polyethylene frits (1 V\ 6 in. diameter; Varian Associates, slurry into 500 mL (or larger) coarse sintered glass filter. Apply
Inc., Harbor City, CA 90710). For sample filtration before LC suction and elute water. Wash silica with 500 mL 1M HC1.
analysis, use Nylon Acrodisc 13 mm Disposable Filter Unit, Rinse silica with LC-grade water, ca 2-3 L, until pH of eluate
0.45 pm pore size (Gelman Sciences, Ann Arbor, MI 48106). is ca 6-7. Transfer silica from funnel to 500 mL beaker by using
(i) SPE apparatus and silica gel for sample cleanup.— mbber policeman and rinsing with methanol. Dry silica in oven
Visiprep Solid Phase Extraction Vacuum Manifold with Teflon at 130°C for at least 3-4 h. Silica may be stored in oven or at
solvent guides (Supelco, Bellefonte, PA 16823); 75 mL poly­ room temperature in covered container. For convenience, 4 g
ethylene reservoirs and bulk silica gel (Silica Gel 60, particle portions of silica may be weighed into 50 mL polypropylene
size 0.040-0.063 mm, 230-400 mesh ASTM, part number centrifuge tubes, capped, and stored at room temperature for
9385-1), no substitutions (EM Science). subsequent use.

LCSystem Acid Digestion of Crop Samples

(a) Pump and autosampler.—SP8800 ternary LC pump Mill barley or dry navy beans to pass through 20-40 mesh
with backflush seals (part No. A3472-010 for backup seals kit sieve. Weigh 10.0 g milled barley, dry navy beans, or whole
or part No. A2963-010 for replacement backup seals only) and wheat flour into 100 mL Pyrex test tube and add 25 mL 6 M
SP8880 autosampler (Spectra-Physics, San Jose, CA 95134). HC1. Digest sample in water bath at 100CC for 1 h. Cool, add
(b) Column.—Du Pont Zorbax silica analytical column (25 ca 10 mL water, and carefully mix by vortexing. Filter acid
cm x 4.6 mm id) with Du Pont Reliance Guard Column Hard­ digest with suction through sharkskin paper in 6.3 cm Buchner
ware Kit and Zorbax silica guard cartridge (1.25 cm x 4 mm funnel into 250 mL vacuum flask. Rinse test tube and funnel 3
id), no substitutions (Mac-Mod Analytical, Chadds Ford, PA times with total of ca 50 mL water. Transfer sample to 250 mL
19317). polypropylene beaker, and rinse flask with ca 25 mL water.
(c) Column oven.—Model LC-100 (Perkin-Elmer Corp., Acid digests may be stored under refrigeration.
Norwalk, CT 06856). Preparation of Digestates for Column Cleanup
(d) Detector.—HP 1040M diode-array UV/vis absorbance
detection system (Hewlett-Packard Corp., Palo Alto, CA Immediately before silica column cleanup, add 10 mL 50%
94303). NaOH (w/v) to acid digestates with swirling. By dropwise ad­
dition, adjust filtrate to pH 9 with 50 and 10% NaOH. Lower
Safety Notes the pH, if necessary, by using solutions of HC1 prepared in LC-
grade water. As soon as pH is adjusted, process samples to
Silica dust should not be inhaled. Dry silica gel should be minimize possible degradation of DQ.
managed under a fume hood. Silica dust is also irritating to the Transfer ca 50 mL pH-adjusted filtrate to 50 mL
skin, eyes, and mucous membranes. Direct exposure should be polypropylene centrifuge tube with conical bottom. Centrifuge
avoided. sample at 2000 rpm for 15-20 min to remove solids. After de­
Hydrochloric acid is used extensively throughout this meth­ canting supernatant for column cleanup (see Silica Gel Column
odology. Concentrated acid or acid solutions should be dis­ Cleanup), transfer another aliquot of filtrate to tube, centrifuge,
pensed, stored, and used under a fume hood. Consult material and decant. Repeat this process until entire sample has been
safety data sheets for further information. centrifuged. Rinse beaker with LC-grade water previously ad­
 C h ic h il a & G il v y d is : Jo urnal O f AO AC I n t e r n a t i o n a l V o l. 76, N o. 6 ,1 9 9 3 1325

justed to pH 9 with NaOH. (Several tubes may be used simul­ Silica was originally used to isolate PQ and DQ from high-
taneously to accommodate total volume of filtrate.) moisture crop sample extracts, because it exhibits primarily
cation exchange selectivity under aqueous conditions (3, 4).
Silica Gel Column Cleanup The capacity of silica to act as a cation exchanger depends on
several factors, particularly pH (4). In the original methodol­
Insert 20 pm frit into 75 mL polyethylene sample reservoir,
ogy, recovery of PQ and DQ from potato extracts appeared to
and place onto vacuum manifold. Add 25 mL Solution A to 4 g
be optimized at pH 9. With low-moisture commodities, the ca­
silica gel in 50 mL polypropylene centrifuge tube. Shake vig­
pacity of the cartridges was exceeded because of excessive
orously and pour slurry into sample reservoir. Let silica settle
sample matrix. Increasing the pH would theoretically increase
(ca 2 min). Open valve and elute basic solution, but do not let
capacity, but at pH >9, degradation of DQ (5) and “irreversible”
gel mn dry at this time. After this elution, transfer 50 mL pH-
retention of PQ were probable. Another way to increase capac­
adjusted sample supernatant to silica gel column. Let gel settle
ity was to use a greater mass of silica gel. A column of 4 g silica
again and then elute sample. Continue adding sample super­
gel was found to provide sufficient capacity for PQ and al­
natant to silica column until entire sample is transferred. Sam­
lowed for the low but reproducible recovery of DQ from ex­
ple solutions should elute freely; if not, apply vacuum to
tracts prepared from low-moisture crop extracts.
achieve flow of ca 1 drop/s, but do not allow silica to run dry.
An inherent advantage in using silica gel columns was the
Using vortex mixer, thoroughly mix insoluble residue remain­
elimination of plugging resulting from sample fines. The top of
ing in centrifuge tube(s) with ca 10 mL LC-grade water pre­
the silica bed had 3 times as much sampling area as the SPE
viously adjusted to pH 9, and centrifuge for 15 min. Apply this
cartridges (27 mm diameter vs 9 mm), which resulted in a less
supernatant to silica column. After sample elution, wash silica
dense accumulation of sample fines. Sample solutions could
column with 100 mL LC-grade water, then 50 mL methanol,
freely permeate the silica column bed.
and then 40 mL solution B. Dry silica by applying vacuum (ca
The final sample volume before the column cleanup was
5 in. Hg) for ca 5-10 s after eluting each of these washes. Dis­
about 100 mL. This volume was exceeded on occasion, how­
card all sample eluate and waste. (Flow may be stopped at any
ever (by as much as 200 mL), with no apparent effect on sample
time to empty waste beakers.) Elute analytes with 100 mL so­
cleanup or analyte recoveries. Therefore, additional aqueous
lution C into clean, HCl-rinsed, unsilanized 250 mL round-bot­
washings in the sample preparation steps before the silica col­
tom flasks without vacuum unless necessary. Carefully evapo­
umn cleanup are permissible.
rate to ca 1 mL by using vacuum rotary evaporator at 90°C.
With vacuum on, vent evaporator and allow stream of air to dry Because of the increased mass of silica, greater wash and
eluate volumes were required to achieve sample cleanup and
remaining 1 mL eluate. Final 1 mLmay also be taken to dryness
in 100°C water bath under gentle stream of nitrogen. Dissolve analyte elution than in the original methodology. The volumes
residue in 1 mL LC mobile phase with swirling. Filter through described in this modified method were adjusted for efficiency
0.45 pm porosity membrane. and practicality. In one instance, an analyst using the method to
determine PQ in wheat samples increased the volume of Solu­
LCDetermination tion B (0.1M HC1 in methanol) from 40 to 60 mL. The loss of
PQ was negligible and the final extracts were cleaner (personal
Inject 100 pL sample solution into LC system. Compare communication, U.S. Food and Drug Administration, Dallas
chromatographic response (peak retention times, heights, District Office). In general, because PQ is highly retained on
and/or areas) with that of standard solution, and calculate resi­ silica, increasing the wash volumes should be permissible. The
due amount. Standard peak responses established for screening recoveries of PQ and DQ reported in this paper were, however,
analysis are equivalent to ca 0.01 ppm PQ and DQ. If levels established by using the method as described earlier.
found in sample significantly exceed screening level, prepare PQ and DQ can be readily adsorbed onto surfaces such as
standard solutions with higher concentration giving peak re­ glass, particularly under alkaline conditions. Because this is a
sponses within 25% of those of sample. Diode array scans of troublesome source of contamination, polypropylene contain­
respective peaks provide qualitative information if desired. ers were specified wherever possible. The sample residues
from low-moisture crops are particularly difficult to remove
R e s u lt s a n d D is c u s s io n from the round-bottom glass sample flasks. Soaking the flasks
overnight in undiluted Micro detergent and rinsing with water,
2M HC1, water, and methanol were usually sufficient to re­
Silica Column Chromatography move contaminants. (The Micro can be reused.)
The original method for sample cleanup and the isolation of Liquid Chromatography
PQ and DQ from high-moisture food crops (3) was not success­
fully applied to low-moisture commodities for at least 2 rea­ Sodium chloride is generally not used as an additive in LC
sons. First, the capacity of the SPE cartridges was apparently mobile phases, because it can cause corrosion of stainless steel.
exceeded; therefore, recoveries of PQ and DQ from low-mois­ The backflush seal arrangement used with the Spectra Physics
ture crops were low and irreproducible. Second, sample fines Ternary Solvent System apparently protected the pump heads
that were not removed despite filtration and centrifugation from seepage. Nevertheless, the pump heads were dismantled
plugged the SPE cartridges. and checked for signs of corrosion periodically, depending on
1326 C hichila & G ilvydis : Journal O f AO AC International V ol . 76, N o. 6 ,1 9 9 3

T a b le 1. R e c o v e r ie s o f p a ra q u a t fro m fo rtifie d c r o p s T a b le 2. R e c o v e r ie s o f d iq u a t fro m fo rtifie d c r o p s

Crop Added, ppm Av. rec., % CV, % Crop Added, ppm Av. rec., % CV, %

Dry navy beans 0.01 86 .0 (5 )a 15.2 Dry navy beans 0.01 4 7 .5 (5 )a 16.6
0.05 7 9 .3 (9) 7.4 0 .0 2 b 52.1 (9) 14.0
0 .3 0 b 87.1 (10) 5.6 0.50 6 5 .8 (3) 5.9
0.50 8 4 .0 (3) 3.1 1.00 74 .5 (3) 1.1
1.00 8 8 .3 (3) 1.5 W hole wheat
W hole wheat flour 0 .026 68 .5 (9) 13.1
flour 0 .0 5 * 9 5 .3 (9) 17.3 Milled barley 0 .0 2 b 59.1 (11) 13.4
Milled barley 0 .0 5 b 9 4 .7 (1 1 ) 12.5
a Num ber of samples In parentheses.
a N um ber of samples in parentheses. b EPA tolerance level.
b EPA tolerance level.

levels of 0.02 and 0.05 ppm for DQ and PQ, respectively, in 10

the amount of use. Autosamplers and other sample injection g crop sample. The average recoveries ranged from 86.4% for
systems must be maintained in a similar fashion. DQ [n = 12, coefficient of variation (CV) = 6.4%] to 90.9% for
An attempt was made to replace NaCl with sodium per­ PQ (n = 12, CV -1.0% ).
chlorate (NaC104) as the ion-pairing reagent. Adequate reten­
tion of PQ and DQ standards with NaC104 (98mM) was Recovery fromFortified Crops
achieved only by using a ternary mobile phase consisting of
acetonitrile, methanol, and water adjusted to pH 2.2 (40 + 30 + Average recoveries of PQ and DQ in 3 low-moisture com­
30). Unfortunately, PQ and DQ in dissolved sample residues modities and potatoes are shown in Tables 1 and 2, respec­
had longer retention times (about 0.05 min longer) and chro­ tively. Crop samples (10 g) were fortified with 1 mL aliquots
matographed with 30% narrower peak widths than the refer­ of diluted standards in methanol. Fortified samples were then
ence standards in mobile phase alone. As previously discussed equilibrated overnight at room temperature. The recoveries for
(3), residual chloride ions that may be present in the sample PQ ranged from 79.3 to 95.3% in all crops at all fortification
residues probably compete or otherwise affect the “ion-pair­
levels. Recoveries for DQ ranged from 47.5 to 74.5% in dry
ing” process of perchlorate ions in the mobile phase. The pres­
navy beans at fortification levels ranging from 0 . 0 1 to 1 . 0 ppm.
ence of chloride ions in the mobile phase appears to be neces­
Recoveries of DQ are low at levels below 1.0 ppm, but their
sary for representative and reproducible chromatographic
separation of PQ and DQ in sample residues after the column CVs are within acceptable limits at these levels. Increasing the
cleanup described in this method.
A Zorbax silica column with a new guard cartridge pro­
duces >5000 theoretical plates per column for both PQ and DQ
according to the following formula:

TV= 5.54 x (retention time/peak width at half height) 2

A simultaneous loss in efficiency for both analytes is indica­

tive of a spent guard cartridge. The Du Pont Zorbax analytical
silica columns were stable with repeated use (e.g., 1 0 0 0 injec­
tions of sample and standard solutions) before showing a sig­
nificant loss in efficiency (3).
[Note: Tailing, peak splitting, enhanced or absent signal
(particularly for DQ), or the appearance of extraneous peaks at
310 nm were associated with the LC-grade water used. These
inconsistencies were encountered in intralaboratory and inter­
laboratory situations (6 ) with LC-grade water that was appar­
ently suitable for other types of analyses. Purified water, as de­
scribed in this method, or bottled water, if necessary, should be
F ig u re 1. L C - d io d e a rra y U V c h ro m a to g ra m s
used.]
d is p la y e d a t 2 5 7 n m (P Q d e te rm in a tio n ) f o r th e a n a ly s is
Recovery of Standards o f fo rtifie d d r y n a v y b e a n s : A , 0.3 p g s ta n d a rd o f P Q ; B,
n a v y b e a n c o n tro l; C , fo rtifie d n a v y b e a n s a t th e
Recoveries of standards in the absence of crop matrix were t o le r a n c e le v e l, 0.30 p p m . R e c o v e r y fo P Q f o r t h is
determined. Fortifications of 0.2 and 0.5 pg were equivalent to s a m p le w a s 89.7%.
 C hichila & G ilvydis: J ournal O f A O A C International V ol . 76, N o. 6 ,1 9 9 3 1327

F ig u r e 2. L C - d io d e a rra y U V c h ro m a to g ra m s F ig u re 4. L C - d io d e a rra y U V c h ro m a to g ra m s
d is p la y e d a t 2 5 7 n m ( P Q d e te rm in a tio n ) f o r th e a n a ly s is d is p la y e d a t 3 1 0 n m (D Q d e te rm in a tio n ) f o r th e a n a ly s is
o f fo rtifie d w h o le w h e a t flo u r: A , 50 n g s ta n d a rd o f P Q ; o f fo rtifie d w h o le w h e a t flo u r: A , 2 0 n g s ta n d a r d o f D Q ;
B , w h o le w h e a t f lo u r c o n tr o l; C , fo rtifie d w h o le w h e a t B , w h o le w h e a t flo u r c o n tr o l; C , fo rtifie d w h o le w h e a t
flo u r a t th e to le r a n c e le v e l, 0.05 p p m . R e c o v e r y o f P Q f o r flo u r a t th e t o le r a n c e le v e l, 0.02 p p m . R e c o v e r y o f D Q f o r
t h is s a m p le w a s 95.2%. t h is s a m p le w a s 60.7%.

mass of silica gel (e.g., from 4 to 5 g) did not increase the re­ Dry navy beans, in particular, were the most difficult to work
covery ofDQ. with; peak heights for DQ were measured manually by using
Typical UV diode array chromatograms of standards and baseline correction because of the presence of an interférant
sample solutions are shown in Figures 1-4. peak (Figure 3). Blank correction was also usually required for
The determination of PQ and DQ in low-moisture com­ calculating PQ levels in dry navy bean extracts.
modities was often complicated by the presence of interférants.
High-Moisture Crop Cleanup and Recoveries

The method for low-moisture commodities can be used for

high-moisture crops. The recoveries of PQ at fortification lev­
els of 0.05 and 0.50 ppm in potatoes were 90.2% (n = 4, CV =
7.2%) and 93.8% (n = 5, CV = 6.4%), respectively. The recov­
eries of DQ at fortification levels of 0.02 and 0.10 ppm were
78.4% (n = 4, CV = 6.3%) and 85.7% (n = 5, CV = 5.5%).
These recoveries are comparable to those obtained with the
original SPE method (3). However, the original method is rec­
ommended for high-moisture commodities, because the col­
umn cleanup technique was not as convenient as the cartridge
method. The preparation of bulk silica was time-consuming,
and the solvent requirements were about 1 0 times those of the
original method. The column cleanup technique also resulted
in sample residues that were particularly difficult to remove
from glassware.

C o n clu sio n s

F ig u r e 3. L C - d io d e a rra y U V c h r o m a to g r a m s A method providing selective, multiresidue determination

d is p la y e d a t 3 1 0 n m (D Q d e te rm in a tio n ) f o r th e a n a ly s is of PQ and DQ in high-moisture food crops was modified for
o f fo rtifie d d r y n a v y b e a n s : A , 20 n g s ta n d a r d o f D Q ; B , low-moisture commodities. Recoveries for PQ and DQ were
n a v y b e a n c o n tro l; C , fo rtifie d n a v y b e a n s a t th e reproducible with this method but were generally low for DQ.
to le r a n c e le v e l, 0.02 p p m . R e c o v e r y o f D Q f o r th is Sample cleanup allowed the determination of PQ and DQ at
s a m p le w a s 56.7%. fortification levels ranging from 0.01 ppm to EPA tolerance levels.
1328 C hichila & G ilvydis: J ournal O f A O A C International V ol . 76, N o. 6 , 1993

R e fe re n c e s (3 ) C h ic h ila , T .M ., & W alters, S .M . ( 1 9 9 1 ) 7. A ss o c . Off. A nal.

Chem . 74, 9 6 1 - 9 6 7
(1 ) F D A S u rveilla n ce In dex f o r P e s tic id e s ( 1 9 8 1 ) P B 8 2 - 9 1 3 2 9 9 , (4 ) B ro w n , D .M ., & P eitrzy k , D J . ( 1 9 8 9 ) 7 . C hrom atogr. 466,

D .V . R e e d (E d .), N a tio n a l T e c h n ic a l In fo rm a tio n S e rv ic e , 2 9 1 -3 0 0

S p rin g field , V A (5 ) W orob ey , B .L . ( 1 9 8 7 ) P estic. ScL 1 8 , 2 4 5 - 2 5 7

(2 ) F u n d erb u rk , H .H . ( 1 9 6 9 ) in D e g ra d a tio n o f H erb ic id es, R C . (6 ) C h ic h ila , T .M ., & G ilv y d is, D .M . ( 1 9 9 1 ) L a b o ra to ry In for­

K e a rn e y and D .D . K a u fm a n (E d s), M a r c e l D e k k e r, In c .. N ew m ation B u lletin 3 6 2 0 , U .S . F o o d and D rug A d m in istra tion ,
Y o rk , N Y , pp. 2 8 3 - 2 9 8 W ash in g ton , D C
 Lopez -A vila & J ones : J ournal O f AOAC International V ol . 76, N o. 6 ,1 9 9 3 1329

RESIDUES AND TRACE ELEMENTS

I n te r la b o r a to r y S tu d y o f a T h e rm o s p ra y -L iq u id

C h r o m a to g r a p h ic /M a s s S p e c tr o m e tr ic M e th o d f o r S e le c te d

A f-M e th y l C a r b a m a t e s , /V -M e th y l C a r b a m o y lo x im e s , a n d

S u b s titu te d U r e a P e s tic id e s

V io r ic a L o p e z - A v il a

Midwest Research Institute, California Operations, 625-B Clyde Ave, Mountain View, CA 94043
T ammy L . J ones
U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory, Las Vegas, NV 89119

Collaborators: B. Anderson; T. Behymer; R. Christensen; T. Dougherty; T. Getek; L. Dnicki; D. Richards; L. Shalaby; C. Vestal

A th erm o sp ray -liq u id c h ro m a to g ra p h ic /m a ss s p e c ­ deviation o v er th e co n c en tratio n ra n g e te ste d .

tro m etric (TS-LC/MS) m eth o d w a s e v a lu a te d in a n A nalysis of v aria n ce in d icates th a t for all com ­
in terlab o rato ry stu d y fo r d eterm in in g 3 /V-methyl p o u n d s te s te d , th e variation from laboratory to
c a rb a m a te s (b en d io carb , carbaryl, a n d carbofuran), laboratory is g re a te r th a n th a t attrib u ted to analyti­
3 /V-methyl ca rb am o y lo x im es (aldicarb, m ethom yl, cal erro r d isp lay ed within lab o rato ries. T here are
a n d oxam yi), 2 su b stitu te d u rea p e stic id e s (diuron m any o p eratio n al p a ra m e te rs th a t co u ld co n trib u te
a n d linuron), a n d 1 e s te r of a s u b s titu te d ca rb am ic to interlaboratory variability; o n e of th em , th e th e r­
acid (carbendazim ). T he p u rp o s e of th is s tu d y w a s m o sp ra y tip te m p eratu re, c a n play a m ajor role in
to e sta b lish w h e th e r th e s e 9 c o m p o u n d s c a n b e re­ a d d u c t form ation a n d ion frag m en tatio n in th e c a s e
liably d e te c te d a n d q u a n tita te d with th is m ethod, of therm ally labile c a rb a m a te p e stic id e s and, th e re ­
a n d to e s ta b lis h th e in terlab o rato ry p re cisio n a n d fore, n e e d s to b e m onitored a n d co n tro lled c a re ­
a c c u ra c y of th e m eth o d w ith cu rren tly available in­ fully. T he correlation of m a s s sp e c tra l d a ta with th e
stru m en tatio n . T he s tu d y d e sig n w a s b a se d o n th e rm o sp ra y tip te m p e ra tu re w a s atte m p te d u sin g
AOAC INTERNATIONAL’S blind replicate d e sig n principal co m p o n e n t an a ly sis.
with b alan c ed re p lic ates. T he sa m p le s c o n s is te d of
so lu tio n s of th e 9 te s t c o m p o u n d s in m eth an o l a t 3
c o n c e n tra tio n s th a t w e re u nknow n to th e p articip at­ T he Methods Research Branch, Quality Assurance and
ing lab o ra to ries a n d th a t co v e red th e linear ran g e Methods Development Division, U.S. Environmental
of th e m eth o d . Nine v o lu n teer la b o ra to ries partici­ Protection Agency (EPA), Environmental Monitoring
p ated in th e study. L inear re g re ssio n e q u a tio n s are Systems Laboratory, Las Vegas, NV, develops analytical meth­
p re s e n te d th a t c a lc u late th e a c c u ra c y of th e ods for the determination of toxic compounds in environmental
m eth o d , i.e., th e p e rc e n t re co v ery of e a c h of th e 9 samples. Once a new method has been developed in a single
c o m p o u n d s a t a n y co n c e n tra tio n within th e ra n g e laboratory, an important step in further evaluating the method
of c o n c e n tra tio n s te s te d (5-90 pg/m L for e a c h c o m ­ is the completion of an interlaboratory study. This paper con­
p o u n d , e x c e p t carb en d azim , for w hich th e ra n g e tains the results of an interlaboratory study designed as part of
w a s 1.25-22.5 pg/mL). T h e in tralab o rato ry preci­ the method development program for the analysis of selected
s io n s of th e TS-LC/MS m eth o d ra n g e d from 6.5 to pesticides by thermospray-liquid chromatography/mass spec­
33.1% relative s ta n d a rd deviation, d ep e n d in g on trometry (TS-LC/MS). The study was conducted under the
th e c o m p o u n d . T he in terlab o rato ry m eth o d preci­ auspices of the Methods Research Branch, Environmental
s io n s ra n g e d from 29.8 to 98.2% relative sta n d a rd Monitoring Systems Laboratory. Nine compounds, including 3
V-methyl carbamates (bendiocarb, carbaryl, and carbofuran), 3
carbamoyloximes (aldicarb, methomyl, and oxamyi), 2 substi­
Received D ecember 10, 1992. Accepted M arch 12, 1993.
Although the research described in this article has been supported by tuted urea pesticides (diuron and linuron), and 1 ester of a sub­
the U.S. Environmental Protection Agency under Contract No. 6 8 -0 -0 0 2 9 stituted carbamic acid (carbendazim) were selected for evalu­
to M idwest Research Institute, it has not been subjected to Agency review. ation in this study. Their CAS Registry numbers, chemical
Therefore, it does not necessarily reflect the views o f the Agency. Mention
formulae, chemical structures, and chemical names are given
o f trade names or commercial products does not constitute endorsem ent or
recommendation for use. in Table 1.
1330 L opez -A vila & J ones : J ournal O f AOAC International V ol . 76, No. 6,1993

T a b le 1. T a rg e t c o m p o u n d s , th e ir C A S r e g is try n u m b e rs , c h e m ic a l fo rm u la e , c h e m ic a l s tru c tu re s , a n d c h e m ic a l
nam es

Com pound nam e C A S registry No. Chemical form ula Chem ical structure Chem ical nam e

Aldicarb 116-06-3 C 7H 14N20 2S 2-Methyl-2-(m ethylthio)propionaldehyde

0 O-m ethyl carbamoyloxime
C H ,S - C - C H = N - 0 - C - N - C H ,

CH, H

Bendiocarb 2 2 7 8 1 -2 3 -3 C 11H 13N 0 4 1,3-Benzodioxol-4-ol

2,2-dim ethyl-N-m ethylcarbam ate

O -C -N H -C H .
ii
0

Carbaryl 6 3 -2 5 -2 C^H^NCT, OH 1-Naphthyl N -m ethylcarbam ate

ii i
O -C -N -C H ,

C arbendazim 1 0 605-21-7 C 9H 9N30 2 C arbam icacid 1-H benzim idazole-2-yl

methyl ester

Qc>r°“' H

Carbofuran 1563-6 6-2 C 12H 15N 0 3 2,3-Dihydro-2,2-dim ethylbenzofuran-7-yl

N-m ethylcarbam ate

Diuron 330-54-1 C 9H 10CI2N 2O 3-(3,4-D ichlorophenyl)-1,1-dimethyl urea

Linuron 33 0 -5 5 -2 C 9H 10CI2N 2O 2 3-(3,4-Dichlorophenyl)-1 -

methoxy-1 -methyl urea

Methomyl 16752-77-5 C 5H 10N2O 2S 1 -(M ethylthio)acetaldehyde

o O-methylcarbam oyloxime
C H ,-C = N -0 -C -N -C H ,
1 I l 1
S -C H , H

Oxamyl 2 3 1 3 5 -2 2 -0 C7 H1 3 N3 O3 S 2-Dimethylamino-1-(methylthio)glyoxal
O O-m ethylcarbam oyl-m onooxime
11
(C H .l.N -C -O N -O -C -N -C H ,
SCH,

These pesticides are typical examples of polar, nonvolatile, pounds are not amenable to gas chromatography, because they
and thermally labile compounds. The need for sensitive and are thermally labile. Several investigators reported methods
specific analytical methods for such compounds is increasing that involve the derivatization of these compounds or of the
as they become more widely used in agricultural pest-control phenols or methylamines that result from their basic or acidic
programs. Reports in the literature indicate that these com­ hydrolysis, respectively. For example, Coburn et al. (1) used
 L opez -A vila & J ones : J ournal O f A O A C International V ol . 76, N o. 6 ,1 9 9 3 1331

pentafluorobenzyl (PFB) bromide to derivatize the phenols were advised to follow the instrument manufacturer’s specifi­
that resulted from the hydrolysis of 2 /V-methyl carbamates, cations for optimal interface performance.
carbaryl and carbofuran, with methanolic potassium hydrox­
ide, and analyzed the derivatives by gas chromatography with E xperim ental
electron capture detection.
Other very sensitive analytical methods for /V-methyl car­
Study Design
bamates make use of liquid chromatography with post-column
derivatization and fluorescence detection (2,3), or ultraviolet The study design was based on the AOAC INTERNA­
detection without post-column derivatization (4). In the former TIONAL’S blind replicate design with balanced replicates for
case, the aqueous sample is either filtered and directly injected the collaborative evaluation of precision and accuracy for ana­
(2) or it is extracted with methylene chloride (3) and the extract lytical methods (12). In this design, solutions in methanol of the
is then injected into the LC column. After elution from the LC 9 target compounds at 3 concentrations were analyzed to pro­
column, the A-methyl carbamates are hydrolyzed with sodium vide data for estimating the intralaboratory and the interlabora-
hydroxide at 95°C; the methylamine that is formed then reacts tory precision of the method. Three replicate samples at each
immediately with added o-phthalaldehyde and 2 -mercap- concentration were analyzed at the high concentration (identi­
toethanol to give a highly fluorescent derivative. Accurate de­ fied as H-l, H-2, H-3) and at the low concentration (identified
termination of /V-methyl carbamates by LC with fluorescence as L-1, L-2, L-3) while 2 replicate samples were analyzed at the
or ultraviolet detection is highly dependent upon the sample medium concentration (identified as M-2, M-3). In addition, 1
matrix. Since compound identification is based solely on additional sample (identified as M -l) with lA lower concentra­
matching chromatographic retention times, co-elution of inter­ tion than M2 and M3 was analyzed in the medium-concentra­
fering compounds and shifts in retention times can have a great tion range. The true concentrations, unknown to the collabora­
impact on data interpretation. tors, of the 9 target compounds (excepting carbendazim) were
LC with mass spectrometric detection has also been used for 90 pg/mL of each compound in the H -l, H-2, and H-3 samples;
the determination of pesticides (5, 6 ); this technique provides 35 pg/mL of each compound in the M -l sample; 40 pg/mL of
both qualitative and quantitative information. TS-LC/MS in each compound in the M-2 and M-3 samples; and 5 pg/mL of
the positive- and negative-ion modes was used for the charac­ each compound in the L-1, L-2, and L-3 samples. The true con­
terization of aldicarb, carbaryl, and linuron (7). Using the posi­ centrations of carbendazim were 22.5 pg/mL in the H -l, H-2,
tive-ion mode in combination with reversed-phase eluents, the and H-3 samples; 8.75 pg/mL in the M-l sample; 10 pg/mL in
base peaks in the mass spectra of carbamates corresponded to the M-2 and M-3 samples; and 1.25 pg/mL in the L-1, L-2, and
(M + NH4)+ ions (8 ). In the negative-ion mode, the formation L-3 samples. Carbendazim was not soluble in methanol at a
of (M-CONHCH3)~ ions was reported for carbamates. De­ concentration of 100 pg/mL; therefore, it had to be spiked at
lower concentrations.
pending upon which mobile phase was used, other adduct ions
were reported (9). For example, the formation of (M + CH3CN The 9 volunteer laboratories participating in this study,
+ H)+ ions was reported for acetonitrile-water mixtures (9). identified as laboratories 01, 03,05,06,07,08,10,11, and 12,
were instructed to analyze the 9 samples and a sample blank
The variation in the composition of the mobile phase provided
(i.e., a methanol blank identified as B), and to submit the data
additional structural information, with no appreciable loss in
on a 51/4-in. diskette in the format specified by EPA-LV. In ad­
sensitivity.
dition to the samples, each laboratory received 2 ampules of a
TS-LC/MS has been done primarily with single-stage quad-
stock standard solution containing all 9 target compounds, each
rupole mass spectrometers. Because thermospray ionization is
at a concentration of 1 mg/mL, in methanol except for carben­
a relatively “soft” technique resulting in very little fragmenta­ dazim at a concentration of 0.25 mg/mL. The laboratory staff
tion, it may be difficult to identify the compounds when then- were instructed to prepare calibration standards at 10, 30, 60,
mass spectra exhibit mainly the adduct ions. To maximize the and 100 pg/mL by diluting these stock standard solutions
identification power of the mass spectrometer, Betowski and 1:100, 1:33, 1:17, and 1:10 (vol. to total vol.) with methanol,
Jones (10) and Chiu et al. (11) used mass spectrometry/mass respectively. Prior to analyzing the samples, the staff of each
spectrometry (MS/MS) in conjunction with thermospray ioni­ laboratory performed a 4-point calibration curve using the
zation. Using low-energy collision with an inert gas, the daugh­ working standards at concentrations of 10, 30, 60, and
ter ions resulting from the collision-activated dissociation re­ 100 pg/mL and calculated the slope, intercept, and correlation
sembled those obtained when electron impact or chemical coefficient for each target compound using the linear regres­
ionization modes are used. sion technique. Quantitation of the test compounds in the sam­
The main purpose of this study was to evaluate whether TS- ples was done by external standard calibration.
LC/MS can reliably detect and quantity the 9 compounds, and The TS-LC/MS conditions recommended by EMSL-LV are
to establish the interlaboratory precision and accuracy of the summarized in Table 2. These conditions were developed from
method with currently available instrumentation. Because of those reported by Anderson (12) and optimized for this set of 9
the complex nature of the TS-LC/MS operation, specific oper­ compounds using Dry Lab software (LC Resources, Inc., Wal­
ating parameters, such as instrument tuning and calibration, nut Creek, C A). The laboratories were asked to use the oositi ve
1332 L o p e z -A v ila & J o n e s : J o u r n a l O f A O A C I n tern a tio n a l V o l . 76, N o. 6 ,1 9 9 3

filament turned on whenever possible. The ions recommended Preparation of the Sam ples
for quantitation are listed in Table 2; however, laboratory 08
used the ion at mass-to-charge ratio (m/z) 191 to quantitate aldi- The 9 target compounds used in this study were purchased
carb, laboratories 07 and 08 used the ion at m/z 202 to quanti­ from Crescent Chemical (Hauppauge, NY) and were used
tate carbary 1 , and laboratory 08 used the ion at m/z 2 2 0 to quan­ without further purification. Their purities were stated to be at
titate oxamyl. least 96%. One concentrated stock solution containing carben-
The tip temperatures used by the 9 laboratories were as fol­ dazim at a concentration of 0.25 mg/mL and each of the other
lows: laboratory 01, gradient from 114° to 90°C; laboratory 03, 8 target compounds at a concentration of 1 mg/mL was pre­

99°C; laboratory 05, gradient from 240° to 180°C; laboratory pared in methanol. Sample dilutions were prepared as follows:
06, I05°C; laboratory 07, approximately 240X; laboratory 08, for the L-1, L-2, and L-3 samples, add 1 mL concentrated stock
the heated vaporizer was set at 200°C (this corresponds to an solution to 200 mL volumetric flask and increase the volume to
exit gas temperature of 125X1); laboratory 10, gradient from 200 mL with methanol; for the M-l sample, dilute 7 mL stock
185° to 145X; laboratory 11,200X; and laboratory 12, gradi­ solution into total volume of 200 mL with methanol; for the
ent from 205° to 143X. M-2 and M-3 samples, dilute 8 mL stock solution into total
volume of 200 mL with methanol; and for the H -l, H-2, H-3
samples, dilute 18 mL stock solution into total volume of
T a b le 2. T S - L C / M S o p e ra tin g c o n d it io n s re c o m m e n d e d 200 mL with methanol. The samples were kept refrigerated at
f o r th e in te rla b o r a to r y s tu d y 4°C in the dark until their shipment to the laboratories partici­
Value
pating in the collaborative study. The concentrations of the
P aram eter
samples, as well as of the 7 replicates of stock solution, were
Ionization mode Positive verified prior to the study at the EMSL-LV by LC/UV.
Filament, on/off On
Flow rate, mL/min 1.0
Outlier Testing
Column type C 18-bonded silica
The outlier testing was done using both the Cochran test and
Column dimensions 150 mm length x 4 .6 mm id
the Grubbs test (13). The aim of the former test is to remove the
Particle size, pm 5
Injection volume, pL 20
results obtained from the laboratories showing significantly
Tip tem perature, °C 190 to 2 1 0 greater variability among replicate intralaboratory analyses
Source temperature, °C 2 6 0 to 27 0 than the other laboratories for a given concentration. The pur­
Mass range, amu 140 to 4 0 0 pose of the latter test is to remove the results obtained from
Scan time, s/scan 1.5 laboratories with extreme averages. When a laboratory was re­
jected on the basis of these tests, its results were removed from
G R A D IE N T E LU TIO N P R O G R A M
the set of data for a particular compound, and the test was then
Tim e, min Mobile phase A, % a Mobile phase B, % b repeated 1 more time using the remaining data in the subset.

0 5 95 Statistical Analysis
30 100 0
35 100 0 Statistical analysis was performed using the AOAC Lotus
40 5 95 spreadsheet developed for the analysis of data from interlabo­
45 5 95 ratory studies (14). The spreadsheet program calculates per­
formance parameters of the method according to the Harmoni­
R E C O M M E N D E D ION FO R Q U A N TITA TIO N
zation Guidelines in Ref. 13. Summary statistics (sr, sR, RSDr,
Com pound Ion, m /z RSDr ) were calculated for the average concentrations and for
the overall method precisions at each of the 3 concentration
Aldicarb 2 0 8 (M + N H 4) levels. The single-analyst standard deviation (sr, repeatability)
Bendiocarb 2 2 4 (M + H) or 241 (M + N H 4)
is the precision associated with the performance of an individ­
Carbaryl 2 1 9 (M + N H 4)
ual laboratory, and the overall standard deviation (sR, repro­
Carbendazim 192 (M + H)
ducibility) is the precision associated with measurements gen­
Carbofuran 2 2 2 (M + H) or 2 3 9 (M + N H 4)
Diuron 2 3 3 (M + H)
erated by a group of laboratories. The repeatability relative
Linuron 2 4 9 (M + H) or 2 6 6 (M + N H 4) standard deviation (RSDr), which was determined from the re­
Methomyl 163 (M + H) or 180 (M + N H 4) peatability standard deviation (sr) and the average concentra­
Oxamyl 163 (M + H -C H 3N C O )c or 2 3 7 (M + N H 4) tion at a particular concentration level, is an indication of the
intralaboratory precision. The reproducibility relative standard
a Mobile phase A = 0 .1 M ammonium acetate and 1 % acetic acid in
deviation (RSDr ), which was determined from the reproduci­
methanol.
b Mobile phase B = 0.1 M ammonium acetate and 1 % acetic acid in bility standard deviation (sR) and the average concentration at
water. a particular concentration level, is an indication of the inter­
c As reported in Ref. 18.
laboratory method precision.
 Lopez -A vdla & J ones : J ournal O f AOAC International V ol . 76, No. 6 ,1993 1333

C
C3

F igu re 1. T S -L C / M S c h ro m a to g ra m reported b y labo rato ry 03 fo r the c o m p o s ite sta n d a rd c o n ta in in g the 9 c o m p o u n d s

a t 100 p g /m L e x cept carb e n d az im at 25 p g /m L (injection v o lu m e is 20 |j.L).

Quality A ssurance R esu lts a n d D iscu ssio n

Each laboratory implemented the following quality control

Chromatography
procedures:
(7) The LC/MS system was tuned to meet the PEG 400 cri­ The average retention times (in min) of the 9 target com­
teria for accurate mass assignment, sensitivity, and resolution. pounds reported by the 9 laboratories are reported elsewhere
(2 ) A 4-level calibration (10,30, 60, and 100 pg/mL for all (15). Laboratory 06 did not submit hard copies of the TS-
compounds, except for carbendazim at 2.5, 7.5, 15, and LC/MS data. Although the laboratories were instructed to fol­
low the recommended conditions fisted in Table 2, we found
25 pg/mL) was performed before the samples were analyzed.
large variations in compound retention times; we attribute
The linear regression correlation coefficient had to be greater
these variations to the fact that only 2 of the laboratories (labo­
than 0.985 for the multilevel calibration to be acceptable. The
ratory 07 and 10) followed the instructions closely, while the
multilevel calibration was verified daily with a mid-level other laboratories deviated from the recommended instructions
standard (60 pg/mL) whenever all study samples could not be by either using a longer column, a different particle size, a dif­
analyzed in 1 day. If the percent difference between the calibra­ ferent eluting solvent, or a different gradient. For example,
tion factor (total area of the quantitation ion divided by the laboratories 0 1 and 1 2 used the gradient elution program speci­
amount injected) for the mid-level standard and the average fied by EMSL-LV; however, their column lengths were
calibration factor from the initial calibration was more than 250 mm instead of 150 mm. Laboratory 01 used a packing ma­
±30%, the laboratories were required to perform a new multi­ terial composed of 7 pm particles instead of the 5 pm particles
level calibration. specified in the instructions package. Laboratory 06 used ace­
tonitrile instead of methanol as the eluant, and laboratory 08
(3) An instrument blank was performed after the last stand­
used a gradient starting at 20% methanol instead of 5% metha­
ard in the multilevel calibration sequence or after the mid-level
nol in buffer. The shorter retention times reported by laboratory
standard to ensure absence of carryover contamination from
08 may be explained by the fact that this laboratory started with
the calibration standards. more methanol initially. These deviations severely limit our
(4) A methanol blank was included with the study samples ability to compare the data generated. Despite these deviations,
and was analyzed by each of the laboratories. The laboratories the elution order of the compounds from the LC column was
were instructed not to subtract the concentration results for the the same, except that laboratory 03 reported the elution of car­
blank from the concentration results for the sample. bendazim between carbaryl and diuron. Figures 1 through 3
1334 L opez -A vila & J ones : J ournal O f A O A C International V ol . 76, N o. 6 ,1 9 9 3

F igu re 2. T S -L C / M S c h ro m a to g ra m reported by labo rato ry 07 fo r the c o m p o site sta n d a rd c o n ta in in g the

9 c o m p o u n d s at 100 p g /m L e x cept c arb en d azim at 25 p g /m L (injection v o lu m e is 20 pL).

show TS-LC/MS chromatograms submitted by laboratories 03, Rejection o f Outliers

07, and 10, respectively. The best separation was achieved by
From the entire study data (648 data points, not including
laboratory 03 by using a gradient starting at 5% methanol (sol­
the M-l values) the program rejected 46 data points (7.1%) on
vent B)/95% water with 5 mM ammonium acetate and 0.1 M
the basis of the Cochran test. All of the determined outliers
acetic acid (solvent A) to 80% solvent B/20% solvent A and
were observed in the data set from laboratory 05 as follows:
then to 100% solvent B. Under these conditions, carbendazim
aldicarb (H samples), bendiocarb (L samples), carbaryl (all
and aldicarb are completely separated, with carbendazim
samples), carbendazim (FI samples), diuron (all samples),
eluting between carbaryl and diuron. Also, under these condi­
linuron (all samples), methomyl (all samples), and oxamyl (H
tions, oxamyl is partially resolved from methomyl, and bendio- and M samples). Laboratory 05 data were all biased low; this
carb is partially resolved from carbofuran. To achieve the bias may have been due to the fact that this laboratory operated
higher concentration of the buffer needed for TS-LC/MS, labo­ the mass spectrometer with the filament off.
ratory 03 used postcolumn addition of 0.5 M ammonium ace­
tate at 0.2 mL/min. M ethod Precision
Although laboratory 03 achieved the best separation, the
The overall standard deviation (sR, reproducibility) is the
method accuracy (i.e., defined as the average recovery) re­
precision associated with measurements generated by a group
ported by this laboratory for the high-concentration samples
of laboratories; the single-analyst standard deviation (Sr, re­
was not significantly different at the 95% confidence level
peatability) is the precision associated with the measurement
from that reported by laboratories 07 and 10, which used the
performance of an individual laboratory. The values for sr and
gradient elution program that was specified in our instructions
sR, in Tables 3 and 4, were used to calculate method precision,
letter. For example, the average recovery for the high-concen- which is presented here as the relative standard deviation for
tration samples ranged from 81.9 to 112 (average 100, standard repeatability (RSDr) and the relative standard deviation for re­
deviation 11.5) for laboratory 03, from 40.3 to 121 (average producibility (RSDr ). When data from all laboratories were
82.4, standard deviation 24.3) for laboratory 07, and from 74.8 pooled, the RSDr values before outlier removal ranged from
to 141 (average 110, standard deviation 20.7) for laboratory 10. 21.2 to 70.7% relative standard deviation for the high-concen­
Therefore, we concluded that the conditions used by laboratory tration samples, from 10.4 to 40.6% relative standard deviation
03 would not be a viable alternative to the recommended TS- for the medium-concentration samples, and from 9.7 to 93.2%
LC/MS operating (Table 2) conditions. relative standard deviation for the low-concentration samples.
 Lopez -avila & J ones : J ournal O f A O A C International V ol . 76, N o. 6 ,1 9 9 3 1335

TIC
4SI95

c
<
un
a
o
xu
C3
u
35088.
o03
.2
■0
c
38888. o
«

ac nnr.
25bbh.

28888. u.
C3
Xu-I
E
o
-C u
15886.

es
o
10888.
E 12
N
->»J cs
-g ii
E
5880. x
a o
X
Oll
ÜI
J
8 .r.r.r ..,.rv..rrT .T.T , T.. ( T ,T r .T v r -p-r -T .p. r , r.,........ r ,T , r , r ,T , .
8 5 10 15 29 25
TIME (min)
F igu re 3. T S - L C / M S c h ro m a to g ra m reported b y labo rato ry 10 for the c o m p o site sta n d a rd c o n ta in in g the
9 c o m p o u n d s at 100 p g /m L e x cept c arb en d azim at 25 p g /m L (injection v o lu m e is 20 pl_).

After outlier removal the RSDr values were lower, ranging following facts (7) that the TS-LC/MS technology is not nearly
from 10.9 to 21.9% relative standard deviation for the high- as mature as the LC/UV technology, (2) that the laboratories
concentration samples, from 6.5 to 17.5% relative standard de­ were allowed to deviate from the recommended conditions,
viation for the medium-concentration samples, and from 9.7 to and (3) that the technique requires highly skilled analysts, this
33.1% relative standard deviation for the low-concentration wide range in percent RSD r is not unexpected. Because only
samples. Before outlier removal the RSDr values were signifi­ one other collaborative study dealing with the TS-LC/MS tech­
cantly higher than the RSDr values, ranging from 39.0 to 98.5% nique has been done, we are not making any claims as to what
relative standard deviation for the high-concentration samples, is acceptable for method precision at the present time.
from 29.4 to 68.2% relative standard deviation for the medium-
concentration samples, and from 68.2 to 135% relative stand­ M ethod Accuracy
ard deviation for the low-concentration samples. After outlier
removal, the RSDr values ranged from 35.0 to 54.4% relative Data on method accuracy given here as the interlaboratoiy
standard deviation for the high-concentration samples, from average percent recovery of each of the target compounds (cal­
29.8 to 64.2% relative standard deviation for the medium-con­ culated from the average concentrations in Tables 3 and 4) at
centration samples, and from 51.4 to 98.2% relative standard each of the 3 concentration levels (5,40, and 90 pg/mL for all
deviation for the low-concentration samples. compounds except carbendazim, for which the 3 concentra­
The results of this interlaboratoiy study, although poor in tions were 1.25,10, and 22.5 pg/mL) are presented in Tables 5
terms of method precision in comparison with those reported and 6 . The method accuracy data for the other medium-concen­
by Edgell et al. in Ref. 2, are comparable to those reported by tration samples M l samples, concentration 35 pg/mL, except
Jones et al. (5) for 11 chlorophenoxy acid herbicides. Given the for carbendazim at 8.75 pg/mL, are not included, because we
1336 Lopez -A vila & Jones : Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3

T able 3. Interlaboratory m eth od p re c isio n (before outlier re m o v al)3

High concentration Medium concentration Low concentration

Compound Av. sr sR RSDr, % RSDr , % Av. sr sR !RSDr, % RSDp, % Av. Sr SR RSDr, % RSDr , %

Aldicarb 89.9 19.6 35.0 21.8 39.0 44.1 7.7 17.0 17.5 38.5 2.6 0.9 2.6 33.1 98.2
Bendiocarb 73.3 16.1 39.3 21.9 53.6 38.0 6.6 16.6 17.3 43.7 4.0 3.8 5.4 93.2 135
Carbaryl 93.6 66.2 70.0 70.7 74.8 45.4 9.2 17.3 20.2 38.1 3.7 1.6 3.2 43.0 85.1
Carbendazim 30.2 13.0 18.4 43.1 60.9 13.8 1.4 8.9 10.4 64.2 1.6 0.4 1.1 26.1 68.2
Carbofuran 79.0 16.7 35.2 21.2 44.5 36.9 5.0 16.3 13.6 44.3 3.6 0.9 3.3 25.2 91.6
Diuron 79.2 44.2 47.8 55.8 60.3 40.6 5.1 11.9 12.5 29.4 4.9 2.7 5.7 56.2 117
Linuron 80.6 23.0 37.5 28.5 46.5 40.2 10.3 16.7 25.7 41.7 5.4 4.6 5.8 86.0 108
Methomyl 119 53.0 118 44.3 98.5 44.1 17.9 30.1 40.6 68.2 5.6 3.1 5.7 55.3 102
Oxamyl 82.3 31.8 47.0 38.6 57.1 39.4 9.2 24.0 23.2 60.8 4.9 0.5 4.6 9.7 93.6

a sr and sRare the standard deviations for repeatability and reproducibility, respectively. RSDr and RSDr are the corresponding relative
standard deviations for repeatability and reproducibility, respectively. The units for average, sr, and sRare pg/mL. The number of replicates
was 3 for the high-concentration and the low-concentration samples and 2 for the medium-concentration samples.

had only 1 replicate per laboratory; however, they were in­ intoan intralaboratoryportion and an interlaboratoryportion,
cludedinTable7. with a corresponding splitof the total number ofdegrees of
To determinetheoverallmethod recovery,we performeda freedom. For all9 compounds and inboth high-concentration
linearregression analysis(Table7).The measured concentra­ and low-concentration samples, we found that the variation
tionswereplottedon they-axis,andthetrueconcentrationsof fromlaboratorytolaboratorywas greaterthanthatattributedto
thetargetcompounds were plottedon thex-axis.The correla­ analyticalerrordisplayedwithinlaboratories.There aremany
tioncoefficientscalculatedforthelinearregressionequations reasons forthe greatervariation; however, because the study
were allabove 0.98 (Table 7), which confirms the utility of samples didnotcontain any ofthecontaminants thatareusu­
theseequationsforestimatingthemethod accuracyatanycon­ allyfound inenvironmental samples, we attributedthisinter­
centration levelwithin therange tested.The slopes of there­ laboratory variation tooperationalparameters such as mobile
gressionequationsvariedfrom0.785fordiuronto1.24forcar- phase flow rate,mobile phase and buffercomposition, vapor­
bendazim (Table 7). These data indicate that the overall izer temperature, tip temperature, and source temperature.
method recoveries,for8ofthetargetcompounds atconcentra­ Among these operational parameters, the tiptemperature was
tionsrangingfrom 5 to90 pg/mL and forcarbendazimatcon­ thoughttobetheparametermostlikelytoaffectthemass spec­
centrationsrangingfrom 1.25to22.5pg/mL,rangedfrom78.5 tralfragmentationofthesecompounds.
to 124%.
Mass Spectral Analysis
Analysis of Variance
We analyzed the mass spectraldatareported by 7 ofthe 9
Analysisofvariance(ANOVA) wasusedtomathematically laboratories, excepting laboratories 06 and 08 which did not
separatethetotalvariation oftheexperimental measurements submit mass spectra, and looked forpossible correlationsbe-

Table 4. Interlaboratory m ethod p re c isio n (after outlier re m oval)3

High concentration Medium concentration Low concentration

Compound Av. Sr SR RSDr, % RSDr , % Av. Sr sr IRSDr, % RSDr, % Av. sr SR RSDr, % RSDr , %

Aldicarb 88.8 11.4 34.4 12.9 38.8 44.1 7.7 17.0 17.5 38.5 2.6 0.9 2.6 33.1 98.2
Bendiocarb 73.3 16.1 39.3 21.9 53.6 38.0 6.6 16.6 17.3 43.7 2.6 0.6 1.6 21.3 61.9
Carbaryl 82.8 11.7 34.0 14.2 41.1 43.1 3.0 15.7 7.0 36.4 3.1 0.7 2.3 23.3 75.8
Carbendazim 28.1 5.6 15.3 19.9 54.4 13.8 1.4 8.9 10.4 64.2 1.6 0.4 1.1 26.1 68.2
Carbofuran 79.0 16.7 35.2 21.2 44.5 36.9 5.0 16.3 13.6 44.3 3.6 0.9 3.3 25.2 91.6
Diuron 71.9 13.1 26.1 18.2 36.3 39.5 2.6 11.8 6.5 29.8 3.3 0.5 2.6 16.2 77.9
Linuron 76.3 8.3 32.5 10.9 42.6 37.2 3.9 13.4 10.5 35.9 4.1 0.6 2.1 15.7 51.4
Methomyl 84.0 10.8 29.4 12.9 35.0 36.3 2.8 15.0 7.8 41.2 4.5 0.7 4.1 15.3 92.9
Oxamyl 75.5 12.4 37.0 16.4 49.1 35.2 3.7 20.8 10.4 59.1 4.9 0.5 4.6 9.7 93.6

sr and sRare the standard deviations for repeatability and reproducibility, respectively. RSDr and RSDr are the corresponding relative
standard deviations for repeatability and reproducibility, respectively. The units for average, sr, and sRare pg/rnL. The number of replicates
was 3 for the high-concentration and the low-concentration samples and 2 for the medium-concentration samples.
 Lopez -A vila & Jones : Journal O f A O A C International V ol . 76, N o. 6 ,1 9 9 3 1337

T able 5. M e tho d a c c u ra c y (before outlier rem oval) using unsealed data) was generated to get uncorrelated new
Rec., % variablescalledprincipalcomponents (PCs).
The PCA program gives 2 useful pieces of information:
Compound High-concn Medium-concn Low-concn
name samples3 samples* samples*
scoresand loadingplots.The scoresforeachlaboratoryresult
from the laboratory data vectorbeing projectedonto thePCs
Aldicarb 99.9 110 52.0 (eigenvectorsthatcontaininformationversusthosethatrepre­
Bendiocarb 81.4 95.0 80.0 sentnoiseinthedatamatrix).The loadingsarethecoefficients
Carbaryl 104 114 74.0 of the linear equation defining the PCs; they describe how
Carbendazim 134 138 128 much each variable contributes to the PC. The distance of a
Carbofuran 87.8 92.3 72.0 variableto the origin along a PC isa quantitativemeasure of
Diuron 88.0 102 98.0 theimportanceofthatvariableinexplainingthevariationofthe
Linuron 89.6 101 108 data.Availableneartheorigincarrieslittleinformation,while
Methomyl 132 110 112 alargedistancefrom theoriginmeans thatthevariableisvery
Oxamyl 91.4 98.5 98.0 important.Inaddition,variablesgroupedtogetherholdsimilar
information, and they could be either positively correlated
a Three replicates per laboratory; 9 laboratories. The true
concentration is 90 pg/mL per compound, except carbendazim at (when located on the same side of the origin) or negatively
22.5 pg/mL. correlated(when locatedon opposite sidesoftheorigin).
b Two replicates per laboratory; 9 laboratories. The true Table 8 shows thevariance inourdatathatisexplainedby
concentration is 40 pg/mL per compound, except carbendazim at
10 pg/mL. the5 PCs.As shown, thefirst2 PCs,PC] andPC2,accountfor
c Three replicates per laboratory; 9 laboratories. The true over 70% of the total variance. What we have achieved isa
concentration is 5 pg/mL per compound, except carbendazim at reductionofa5 dimensionalproblemtoa2 dimensionalprob­
1.25 pg/mL.
lem, yeta preservation ofapproximately 70% oftheinforma­
tion.
tween the instrument operating parameters, specifically ther­ Figures4through 12showboththeloadingandscoresplots
mospray tiptemperature,and themass spectralfragmentation forPC, vs PC2forthe9 compounds arrangedinalphabetical
patterns. Principal component analysis (PCA) was employed order.Each ofthesefiguresisdescribedbelow.
forthispurpose (16).The E in * S ig h t™ software(17)was used Figure4 displaystheloadingsand scoresplotsforaldicarb.
toperformPCA. The loadings foreach variable were ordered from thehighest
To perform PCA, we setup a data matrix for each com­ contributiontothelowestcontribution.Inthecaseofaldicarb,
the ions atm /z 191 and 208 are the strongestcontributors to
pound inwhich the normalized intensities (i.e.,those greater
PC],andtheionsatm /z 148and 192arethestrongestcontribu­
than 10% relativeintensity)ofthevariousfragmentionsatspe­ torstoPC2.Because PC] accountsforalmost62% ofthetotal
cificin/z. ratioswere listedforeach laboratory.Next, a covari­ variance (Table 8),we willonly focus on theions atm /z 208
ancematrix(i.e.,atableofthevariancesbetweenthevariables and 191 when analyzingthescoresplot.We seethatlaborato­
riesL10, L03, L05, and L01 clusterinto 1group, while L12,
Table 6. M e tho d a c c u ra c y (after outlier rem oval) Lll, and L07 each constitute separate groups. Although L12
Rec., % andLI1appeartobefarfromtheclusterofL10,L05,L03,and
L01, they liewithinthesame planeastheLI0,L05, L03, and
Compound H gh-concn Medium-concn Low-concn L01 cluster,and can,therefore,be consideredpartofthatclus­
name samples3 samples* samples*
ter.L07, however, standsby itself.Inanalyzingtheindividual
Aldicarb 98.7 110 52.0
mass spectrafromtheselaboratories,we findthatthelaborato­
Bendiocarb 81.4 95.0 52.0
riesbelongingtothebigclusterexhibitabasepeakatm /z 208,
Carbaryl 92.0 108 62.0
correspondingtothe(M + NH 4)+ion.L07, whichwas notpart
Carbendazim 125 138 ^28 of this cluster, reported a mass spectrum for aldicarb with a
Carbofuran 87.8 92.3 72.0 basepeakatm /z 191 (correspondingtoM + H)+.Itappearsthat
Diuron 79.9 98.8 66.0 thehighthermospraytiptemperature(240°C)usedby L07 has
Linuron 84.8 93.0 82.0 resultedinthedecompositionofthe(M + NH 4)+adduction.
Methomyl 93.3 90.8 90.0 Figures 5 and 8 display the loadings and scores plots for
Oxamyl 83.8 88.0 98.0 both bendiocarb and carbofuran. In this case, because the 2
compounds coelutedandtheirmass spectrawereunresolvable,
a Three replicates per laboratory; 8 or 9 laboratories. The true
concentration is 90 pg/mL per compound, except carbendazim at
eachlaboratorystandsoutseparately,indicatingnocorrelation
22.5 pg/mL among thethermospraytiptemperatures.
b Two replicates per laboratory; 8 or 9 laboratories. The true The loadingsplotforcarbaryl(Figure6)shows thattheions
concentration is 40 pg/mL per compound, except carbendazim at
10 pg/mL. atm /z 202 and219 arethemostimportantcontributorstoPCj,
0 Three replicates per laboratory; 8 or 9 laboratories. The true whichaccountsfor75% ofthevarianceinthedata,andthatthe
concentration is 5 pg/mL per compound, except carbencazim at ionsatm /z 145 and 201 arethemostimportantcontributorsto
1.25 pg/mL.
PC2.The scores plot shows thatlaboratories L05, Lll, L03,
1338 L o p e z -A v il a & J o n e s : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

T able 7. S u m m a ry sta tistic s a n d re g re ssio n e q u a tio n s

Compound name Ca Xb SRC

C Regression equation Correlation coefficient, r

Aldicarb 90 88.8 34.4 X = 1.004C + 0.1674 0.9968

40 44.1 17.0 sr = 0.374X + 0.5437 0.9956
35 35.9 12.2
5 2.6 2.6

Bendlocarb 90 73.3 39.3 X = 0.817C + 1.886 0.9936

40 38.0 16.6 sr = 0.529X - 0.2763 0.9895
35 32.6 18.9
5 2.6 1.6

Carbaryl 90 82.8 34.0 X = 0.925C + 1.711 0.9945

40 43.1 15.7 sr = 0.399X + 0.0410 0.9957
35 35.1 13.6
5 3.1 2.3

Carbendazim 22.5 28.1 15.3 X = 1.24C + 0.5500 0.9985

10 13.8 8.9 sr = 0.535X + 0.6051 0.9946
8.75 11.4 6.5
1.25 1.6 1.1

Carbofuran 90 79.0 35.2 X = 0.879C + 0.8621 0.9987

40 36.9 16.3 sr = 0.422X+ 1.829 0.9977
35 33.4 17.0
5 3.6 3.3

Diuron 90 71.9 26.1 X = 0.785C + 4.328 0.9859

40 39.5 11.8 sr = 0.343X-0.1789 0.9808
35 36.0 10.4
5 3.3 2.6

Linuron 90 76.3 32.5 X = 0.840C + 2.006 0.9977

40 37.2 13.4 sr = 0.423X - 0.5095 0.9951
35 33.2 13.7
5 4.1 2.1

Methomyl 90 84.0 29.4 X = 0.934C - 0.0859 0.9997

40 36.3 15.0 sr = 0.315X + 3.359 0.9978
35 33.6 14.8
5 4.5 4.1

Oxamyl 90 75.5 37.0 X = 0.823C + 2.292 0.9984

40 35.2 20.8 sr = 0.456X + 3.215 0.9967
35 33.4 18.4
5 4.9 4.6

a True concentration, pg/mL.

b Average measured concentration, pg/mL.
c Reproducibility standard deviation, pg/mL.

LI2,andL01 allcluster,butL10 andL07 aresingulargroups. laboratory 10 data, theresults indicate thatcarbarylbehaved

L07 reportedabasepeak atm /z 202,correspondingtothe(M similarlytoaldicarbunderthesethermosprayconditions.
+ H)+ion,whereastheotherlaboratories,exceptL10,reported Inthecaseofcarbendazim(Figure7),90% oftheinforma­
a base peak atm /z 219, corresponding tothe(M + NH4)+ion tioniscontained inPC] and only 6.4% inPC2(Table 8).Be­
and afragment ion atm /z 202 ofapproximately 30% relative cause PC2contains only 6.4% ofthevariance inthedata, we
intensity.L10 reportedafragmentionatm /z 202 of99% rela­ willnotdiscusstheionsatm /z 134 and 160,which seem tobe
tiveintensity.Although we do nothave an explanationforthe importantcontributorstoPC2.From theionsthatcontributeto
 L o p e z -A v il a & J o n e s : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1339

T able 8. V a rian ce in the m a s s sp e c tral data c o r re s p o n d in g to e a c h o f the 5 P C s

Compound PC,

o
Q_

CL
o
PC4 PC5

CO
VJ
Aldicarb 61.9 24.2 9.7 3.4 0.7
Bendiocarb 44.2 37.4 10.6 5.5 2.2
Carbaryl 75.0 15.9 9.0 0 0
Carbendazim 89.9 6.4 3.0 0.8 0
Carbofuran 60.4 24.9 10.9 3.8 0
Diuron 51.7 23.6 12.5 10.7 1.5
Linuron 69.2 21.9 4.3 2.6 1.6
Methomyl 95.5 3.0 1.5 0 0
Oxamyl 42.7 29.0 17.6 7.7 2.6

PC1;onlytheionatm /z 192isimportantbecausetheotherion mospray tiptemperature (90°to 114°C) thantheotherlabora­

atm /z 208 doesnotbelongtocarbendazimbuttoaldicarb.This tories.Thus,itappearsagainthatthelow thermospraytiptem­
explainswhy laboratoriesL01, L03, L05, L07, L10, and L12 peraturefavorstheformationoftheammonium adductions.
aregroupedtogether,and LI1standsseparately. Figure 10 shows the loadings and the scores plots for
The loadingsplotofdiuron(Figure9)showsthattheionsat linuron.Inthecaseoflinuron,theionsatm /z 249 and266 are
m /z 250and252 (ionatm /z 250cannotbeseenbecauseitover­ the strongestcontributors toPCb and theions atm /z 148 and
lapswithionatm /z 252)arethemostimportantcontributorsto 233 the strongestcontributors to PC2.Because PC, accounts
PC |,whichaccountsforalmost52% ofthevarianceinthedata, for69% ofthetotalvariance (Table 8),we willonly focus on
andthattheionatm /z 234 isthemostimportantcontributorto the ions atm /z 249 and 266 when analyzing the scores plot.
PC2,whichaccountsforalmost24% ofthevarianceinthedata. Thereare4distinctgroupsinthescoresplot.LaboratoriesL03,
When analyzingthescoresplotinFigure9,we seethatlabora­ L07, and L10 make up 1 clusterand L05 and L12 make up
toriesL10 and LI1make up 1cluster,L03,L05,L07,andL12 another, whereas L01 and Lll each stand separately. When
anothercluster,and L01 standsseparately.L10 and LI1arein comparingthemass spectraoflinuronreportedbylaboratories
the same planeasthelargeclusterand areseparatedby small L05 andL12 withthosereportedbyL03,L07,andL10,we find
differences.They canbe consideredpartsofthelargercluster, only minor differences ineitherthefragment ions ortheirin­
becausetheyweretheonly2 laboratoriestopickup thecontri­ tensities.Therefore,L03, L05, L07, L10, and L12 can be con­
bution from thecarbon isotope, m /z 234. When analyzingthe sideredpartsofthesame largecluster.LI1reportedafragment
mass spectrumfromL01,we findnotonlythe(M + Fi)+ionat ion atm /z 148 thatno otherlaboratory reported, and L01 re­
m /z 233,butalsothe(M + NH4)+ionatm /z 250, aswellasthe portedthe(M + NfL()+ionatm /z 266 asthebasepeak.Because
contributionfrom thechlorine-37isotopeatm /z 252 (datanot L01 used a lower thermospray tiptemperature, theformation
shown). The otherlaboratorieseitherdidnotreportionsat(M oftheammonium adductionswas alsofavored.
+ NH 4)+ortheintensitiesofsuchionswerelow(approximately In the case of methomyl, 95% of the information iscon­
10% for Lll and LI2). Laboratory L01 used a lower ther­ tained inPCj. This implies that,intheloadingsplot,only the

L oadi ngs Sc o r e s

F ig u re 4 . L o a d in g s a n d s c o r e s p lo ts fo r a ld ic a rb .
1340 L o p e z -A v il a & J o n e s : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

F igu re 5. L o a d in g s an d s c o r e s p lo ts for b endio carb.

F igu re 6. L o a d in g s a n d s c o r e s p lo ts for carb aryl (L03, L11, L12 are not s h o w n b e c a u s e th e y o ve rla p with L01).

Load? n g s S cot as

F ig u re 7 . L o a d i n g s a n d s c o r e s p l o t s f o r c a r b e n d a z i m ( L 0 3 a n d L 0 7 a r e n o t s h o w n b e c a u s e t h e y o v e r l a p w i t h L 0 1 ).
 L o p e z -A v il a & J o n e s : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1341

F igu re 8. L o a d in g s a n d s c o r e s p lo ts for carbofuran.

Loadi n g e Scorse

F igu re 9. L o a d in g s a n d s c o r e s p lo ts for d iuron (L05 is not s h o w n b e c a u s e it o v e rla p s with L12).

Load!n ge Scorse

F ig u re 10. L o a d in g s a n d s c o r e s p lo ts f o r lin u ro n .
1342 L o p e z -A v il a & J o n e s : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

Loa dî ngs Sc o r 9 s

F igu re 11. L o a d in g s an d s c o r e s p lo ts fo r m eth om yl (L10 is not s h o w n b e c a u se it o v e rla p s with L07).

Load? nçs Scorsa

F igu re 12. L o a d in g s an d s c o r e s p lo ts fo r o xam yI.

masses m /z 163, (M + H)+,and m /z 180, (M + NH4)+,areim­ methomyl (L01 couldonlypartiallyresolvemethomyl andox­

portant.Inanalyzingthescoresplotofmethomyl (Figure 11), amyl).Lll andL12 reportedthe(M + NH4)+ionatm /z 237 as
we seethatalllaboratories,exceptLOI, aregrouped together. thebase peak. Laboratories L03 and L10 reportedafragment
The mass spectrareportedformethomyl containabasepeakat ionatm /z 163 asthebasepeak and alsoan ionatm /z 237,the
m /z 163 (M + H)+,exceptforLOI where thebasepeak was at (M + NFI4)+ion,thathad almostthesame intensityasthefrag­
m /z 180 (M + NH4)+.Again, theformation oftheammonium ment ion atm /z 163. Because of the different fragmentation
adduct ions isdue to the lower thermospray tiptemperature patternsunderthermosprayconditions,we couldnotfindacor­
usedby LOI. relationbetweenthethermospraytiptemperatureandthemass
Table8indicatesthatforoxamyl mostoftheinformationis spectraofoxamyl.
containedinPQ, PC2,and PC3.PC, and PC2contain almost To summarize, the thermospray dp temperature seems to
72% oftheinformation;therefore,we willconcentrateonlyon playamajorroleinadductformationandionfragmentationof
theionsthatcontributetoPQ andPC2.The ionsatm /z 163 and thermally labilecompounds such asthecarbamate pesticides.
180areimportantcontributorstoPC1;andtheionatm /z 237 is Therefore, itneeds tobe monitored and carefully controlled.
an important contributortoPC2.Inthe scores plotofoxamyl Otherresearchershavenotedasimilarcorrelationbetweendif­
(Figure 12),we find4 distinctgroups:LOI; Lll andL12; L03 fering spectral patterns and source block temperatures (18),
and L10; and L05 and L07. LOI stands alone because of the and, therefore, the source block temperature should also be
presenceoffragmentionsatm /z 169and 180,whichbelongto closelymonitoredand controlled.
 L opez -A vila & Jones : Journal O f AOAC International V ol . 76, No. 6 ,1 9 9 3 1343

Conclusions References

The TS-LC/MS method used in this study uses sophisti­ (1) Cobum, J.A ., R ipley, B .D ., & Chau, A.S.Y. (1976)7. Assoc.
cated instrumentation, and, thus,itsperformance depends on O ff. Anal. Chem. 59, 188-196

numerous instrumental parameters (e.g., mobile phase flow (2) Edgell, K.W ., Biederman, L .A ., & Longbottom , J.E. (1991)
7. Assoc. O ff. Anal. Chem. 74, 309-317
rate,mobilephaseandbuffercomposition, vaporizertempera­
ture,thermospray tiptemperature, and source block tempera­ (3) Proposed EPA M ethod 8318, A-M ethylcarbam ates by H igh
Performance L iq u id Chrom atography (H PLC )
ture). The results of this study indicate that the TS-LC/MS
(4) Pressley, T .A ., & Longbottom , J.E. (1982) “ The Determ ina­
method may be aviableanalyticalmethod forthermallylabile
tion o f Carbamate and Urea Pesticides in Industrial and
compounds. However, the control of certain operational pa­ M un icip al Wastewater,” EPA 600/4-82-014
rameters iscritical.For example, thetiptemperature needs to
(5) Jones, T .L., Betow ski, L .D ., Lesnik, B ., Chiang, T.C., & Te-
bemonitoredandcarefullycontrolled,becauseitplaysamajor berg, J.E. (1991) Environ. Sci. Technol. 25, 1880-1884
roleinadductformationandionfragmentationforcompounds (6) Behymer, T.D ., B ellar, T .A ., & Budde, W .L. (1990) Anal.
such asthecarbamate pesticides.Although itisnotimportant Chem. 62, 1686-1690
whethertheammonium adductionortheprotonatedionisthe (7) Durand, G., DeBertrand, N ., & Barcelo, D. (1991) 7. Chro-
dominant ion, itisimportant thatthe thermospray spectra be matogr. 554, 233-250
consistentbetween thestandardandthesample andfromday- (8) Barcelo, D ., Durand, G., Vreeken, R.J., DeJong, G.J., Linge-
to-day. Because the TS-LC/MS data reported here were ob­ man, H ., & Brinkm an, U .A .Th. (1991) 7. Chromatogr. 553,
tained with standards and not with extracts ofenvironmental 311-328
samples,method performancewithactualsamples needstobe (9) Durand, G., DeBertrand, N ., & Barcelo, D. (1991) 7. Chro­
addressedinfuturestudies. matogr. 5 6 2 ,507-523
(10) Betow ski, L .D ., & Jones, T.L. (1988) Environ. Sci. Technol.
Acknowledgment 22,1430-1434
(11) C hiu, B.S., Van Langenhove, A ., & Tanaka, C. (1989)
Biomed. Environ. Mass Spectrom. 18, 200-206
The authorsthankthevolunteerparticipantsinthecollabo­
(12) Anderson, B. (1990) “ Evaluation o f TS-LC /M S M ethod fo r
rativestudyfortheirwillingnesstoundertaketheeffort:
Analysis o f Carbamates and Urea Pesticides in S oil and
BradAnderson, APPL, Inc.,Fresno,CA Water,” Interim Report to Lockheed Engineering and Serv­
Thomas Behymer, U.S. Environmental ProtectionAgency, ices Co.
Cincinnati,OH (13) G uidelines fo r C ollaborative Study Procedure to Validate
RichardChristensen,NIST, Gaithersburg,M D Characteristics o f a M ethod o f A nalysis (1989) 7. Assoc. Off.
TerryDougherty,TennesseeEastman, Kingsport,TN Anal. Chem. 72, 694—704
Tim Getek,Battelle-Columbus, Columbus, OH (14) “ Lotus Spreadsheet Program fo r the C alculation o f Perform ­
Leon Bnicki,USDA/FSIS, Alameda, CA ance Parameters from C ollaborative Study Data Including
O utlier Analyses,” revision 3/5/91, received from John G.
Don Richards,VG/Fisons, Manchester, United Kingdom
P hillips, Chairm an, A O A C Statistics Com m ittee
Lamaat Shalaby, DuPont Agricultural Products, Wilming­
(15) Lopez-A vila, V. (1992) “ Interlaboratory Study o f a TS-
ton,DE LC /M S M ethod fo r Selected A -M e th yl Carbamates,
ChristineVestal,VestecCorp.,Houston,TX A -M e th yl Carbamoyl O xim es, and Substituted Urea Pesti­
The authorsaregratefultoLockheed Environmental Serv­ cides,” EPA 600/X-92/102
icesCorp.(LockheedEnvironmentalSystems andTechnology (16) W old, S., Esbensen, K ., & G eladi, P. (1987) Chemometrics
Co.)forpreparingthesamplesand sendingthem tothepartici­ and Intell. Lab. Syst. 2, 37-52
patinglaboratories, and toJohn Dolan ofLC Resources, Inc., (17) Ein*Sight, R evision 2.5 (1989) Infom e trix, Inc., Seattle, W A
forassistancewith optimizing theHPLC conditions forthe 9 (18) Pleasance, S., Anacleto, J.F., Bailey, M .R ., & N orth, D .H .
carbamatepesticides. (1992) 7. Am. Soc. Mass. Spectrom. 3, 378-397
1344 M a t u s ik Et Al .: J o u r n a l Of AOAC In t e r n a t io n a l V o l . 76, No. 6,1993

RESIDUES AND TRACE ELEMENTS

C o n f i r m a t i o n o f I d e n t i t i e s o f P r o p y l e n e a n d E t h y l e n e G ly c o l s i n

A n c h o v ie s b y T a n d e m M a s s S p e c tr o m e tr y

Jean E. M atusik
U.S. Food and Drug Administration,CenterforFood Safety and Applied Nutrition,DivisionofContaminants Chemistry,
Washington, DC 20204
Paul P.Eilers,Ellen M. W aldron,and Steve M. Conrad
U.S. Food and Drug Administration,CenterforFood Safetyand AppliedNutrition,Division ofSeafood Research,
Washington, DC 20204
James A. Sphon
U.S. Food and Drug Administration,CenterforFood Safety and Applied Nutrition,DivisionofContaminants Chemistry,
Washington, DC 20204

A gas chromatographic/tandem mass spectromét­ toinsoluble calcium oxalate. The crystals ofcalcium oxalate
rie (GC/MS/MS) method has been developed for aredepositedintherenaltubules,and cause eitheroliguriaor
confirming the identity of propylene and ethylene anuriaand eventuallynecrosis.These human healthand regu­
glycols added to bait fish for preservation. Bait fish latoryconcernsprompted ourinterestindeveloping a method
are occasionally illegally diverted to human food fortheidentificationofthesecomponents (1).
use. The bait fish were extracted with methanol, the Anchovies destined foruse as “baitfish” were embargoed
extract was centrifuged and filtered, and the filtrate by theStateofCaliforniainOctober 1991.Therewasevidence
was concentrated 10-fold and then analyzed by thatthefishhad beenredirectedforhuman foodconsumption.
GC/MS/MS. The glycols were separated chroma- The anchovies were adulterated with propylene and ethylene
tographically without derivatization or preliminary glycols, and some ofthefishwere alsocontaminated withthe
cleanup. Isobutane positive ion chemical ionization seafood toxin domoic acid, which occurs naturally in certain
was used to generate the protonated molecular ion algaeeatenby fish.
species of each glycol. Product-ion MS/MS experi­ Selected lotsofthe anchovies were extractedwith metha­
ments were performed to obtain spectra to confirm nol,and theconcentrated extractswere analyzedby gas chro-
the identities of propylene and ethylene glycols. matography/tandem mass spectrometry (GC/MS/MS) using
The identities of these 2 compounds in anchovy ex­ positiveionchemical ionization (PICI).The identitiesofpro­
tracts were successfully confirmed by this ap­ pylene and ethylene glycols were successfully confirmed by
proach. theuniquemethod reportedbelow.

Experimental
F ishused forbaitareno longerconsideredacceptablefor
human consumption becauseantifreeze,which contains M a te r ia ls

propylene and ethylene glycols, issometimes added to

(a) M e th a n o l. — Distilled-in-glass and pesticide quality
these“baitfish” topreservethem. To identifytheuseofthese
(Burdick and Jackson Laboratories, Inc., Muskegon, MI
adulteratedbaitfishforhuman foodconsumption,we havede­ 49442);used asreceived.
veloped a method for confirming the identities of propylene
(b) E th y le n e (1 ,2 -e th a n e d io l) a n d p r o p y le n e (1 ,2 -
and ethyleneglycolsisolatedfrom fish.
p r o p a n e d io l a n d 1 ,3 -p r o p a n e d io l) g ly c o l s ta n d a r d s .
—
Propylene and ethylene glycols are classed chemically as (Aldrich Chemical Co.,Inc.,Milwaukee, WI 53233); used as
simpleglycols.Althoughpropyleneglycolisnontoxiciftaken received.
internally,ethyleneglycolisahealthhazardtohumans.Neither
compound isapprovedasafoodadditiveforthisreason.When T is s u e E x tra ctio n
takeninternally,propyleneglycolismetabolizedby oxidation
to pymvic and acetic acids. Ethylene glycol is generally not Anchovy testportions(5g)werehomogenizedinmethanol
lethaltohumans when takeninternally;however, itismetabo­ (5 mL). The homogenate was centrifugedfor 15 min at 1200
lizedinthehuman body tooxalicacid,which isthenconverted x g . The alcoholicsupernatantwas passed through a0.45 pm
Nylon66filter.Approximately4.5mL ofthefiltratewasrecov­
ered, and 2 mL was evaporated under nitrogen to0.2 mL for
R e c e iv e d O c to b e r 2, 1 9 9 2 . A c c e p te d F e b ru a ry 5, 199 3 .
GC/MS/MS analysis.
 M atusdc E t A l . : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 , 1993 1345

Instrumentation w ere de te rm in e d b y M S /M S to fo rm the p ro d u c t io ns, w h ic h

w ere the n reco rde d. Isobutane w as chosen o v e r m ethane as the
A F in n ig a n M o d e l T S Q 46 gas chrom atog raph /ta nd em c h e m ica l io n iz a tio n (C l) reagent gas, because it fo rm e d the
m ass spectrom eter was fitte d w ith a 15 m N u k o l fuse d s ilic a m ost intense M H + w ith these com pounds. M ethane C l p ro ­
c a p illa ry c o lu m n (0 .3 2 m m id ; 0.2 5 p m film th ickn e ss). T he duced a lo w percent re la tiv e abundance (% R A ) o f M H + , w ith
m ass spectrom eter was c o n fig u re d to p e rfo rm P IC I M S /M S the p re d o m in a n t io n b e in g due to the loss o f w ater. Isobutane
p ro d u c t-io n exp erim e nts. F o r G C separation, the c a p illa ry c o l­
C l dem onstrated 1 0 0 % R A o f the M H + io n and little fragm e n­
um n was tem pe rature -prog ram m e d. T he in je c to r tem perature
ta tio n , th e re b y im p ro v in g s e n s itiv ity fo r the p ro d u c t-io n ex­
was set at 2 2 0 X ; the in itia l c o lu m n tem perature was set to
pe rim en ts. A m m o n ia C l pro du ced a lm ost a ll g ly c o l am m onia
50°C. and h e ld fo r 1 m in , the n pro gra m m e d at 5X1/11110 to a fin a l
adduct io n s (M + 18) and little M H + fo r these g ly c o ls , as dem ­
tem perature o f 1 5 0 X and h e ld fo r 1 m in . U n d e r these c o n d i­
onstrated in T able 1. T he a b ility o f C l to fo rm 1 species o f io n ,
tio n s , p ro p yle n e g ly c o l had a re te n tio n tim e o f ca 8.52 m in , and
th e re b y in cre a sin g s e n s itiv ity , is a u se fu l c h a ra cte ristic fo r
e th ylen e g ly c o l had a re te n tio n tim e o f ca 9.4 7 m in . T he h ig h e r
M S /M S analysis. T he M H + is p re fe rre d in th is case because it
fin a l co lu m n tem pe rature was used to e lu te o th e r less v o la tile
g ive s m o le c u la r w e ig h t in fo rm a tio n in c o n ju n c tio n w ith
com ponents fro m the tissue e xtra ct. T he M S /M S o p e ra tin g
c o n d itio n s w ere as fo llo w s : e le ctro n energy 70 eY, e le ctro n M S /M S , w h ic h fu rth e r characterizes these com pounds.

m u ltip lie r 1300 V, co n ve rsio n dynode 5 keV , e m issio n cu rre n t P rop yle ne g ly c o l (1 ,2 -p ro p a n e d io l), w h ic h has a m o le cu la r

0.35 m A , io n source tem pe rature 100"C. Isob utane w as used w e ig h t o f 76.09 and an e le m e n ta l c o m p o s itio n o f C 3H 80 2, was
fo r P IC I. P rod uct ions w ere generated w ith argon as the c o lli­ the firs t com pound to e lu te fro m the G C co lu m n . T he p ro d u ct­
sio n gas set to 1.8 m to rr and the c o llis io n energy set to -2 8 eV. io n spectrum o f the M H + a t m/z 77 can be seen in F ig u re 2. The
T he p ro d u c t io n s fo rm e d b y c o llis io n a lly a ctiva te d de com posi­ io n at m/z 76 fo rm s as a re s u lt o f the loss o f a p ro to n , and the
tio n s o f the p ro to n a te d m o le c u la r io n s o f p ro p yle n e g ly c o l, m/z io n at m/z 59 is due to the loss o f H 20 to fo rm C 3H 70 +. T he loss
77 , and eth yle n e g ly c o l, m/z 63, w ere scanned fro m 20 to 81 o f 15 da lton s, o r C H 3, fro m the io n at m/z 59 gives ris e to the
daltons. A F in n ig a n Inco s data system , M o d e l N o v a 4 X , w ith io n at m/z 44 . T he io n at m/z 41 is due to the loss o f w a te r fro m
T S Q so ftw a re re v is io n 6.1 w as used to acq uire and process the the m/z 59 io n . T he io n at m/z 31 resu lts fro m the loss o f C H
data. fro m the io n at m/z 44, and the io n a t m/z 30 re su lts fro m the
loss o f an a d d itio n a l p ro to n . T h is fra g m e n ta tio n p a th w a y c o r­
Results and Discussion relates w e ll w ith th a t rep orted b y L o n g and P rich a rd (5 ), u sin g
E I/M S .

P rop yle ne and eth yle n e g ly c o ls w ere separated b y a N u k o l E th yle n e g ly c o l (1 ,2 -e th a n e d io l), w h ic h has a m o le cu la r
fused s ilic a glass c a p illa ry c o lu m n w ith baseline re s o lu tio n , w e ig h t o f 62.07 and an elem e ntal c o m p o s itio n o f C 2H 60 2,
good peak sym m etry, and a separation o f a p p ro x im a te ly 1 m in e lu te d fro m the chro m a to g ra p h ic c o lu m n at 9.5 m in . T he p ro d ­
w ith o u t d e riv a tiz a tio n . O ne o f the s p e c ifie d a p p lic a tio n s o f the u c t-io n spectrum o f the M H + at m/z 63 can be seen in F igu re 3.
p o la r N u k o l co lu m n is the separation o f un alte red g ly c o ls . T he T he M H + loses a p ro to n to fo rm the m o le c u la r io n at m/z 62.
separation at lo w tem peratures e lim in a te s the th e rm a l decom ­ T he loss o f H 20 fro m the M H + fo rm s the io n a t m/z 4 5 . T he io n
p o s itio n o f g ly c o l. F ig u re 1 is a re co nstructed io n ch ro m a to ­ at m/z 2 1 fo rm s as a re s u lt o f a d d itio n a l loss o f a w a te r m olecu le
gra m o f an in je c tio n o f a co m b in e d s o lu tio n c o n ta in in g p ro p y l­ and has an elem ental c o m p o s itio n o f C 2H 3.
ene and eth yle n e g ly c o l standards, each at a co n ce n tra tio n o f
25 ng. G ly c o l m ix tu re s have been p re v io u s ly separated b y o th ­
ers (2 ) on a S E -52 coated co lu m n , b u t o n ly a fte r fo rm a tio n o f
the trim e th y l s ily l eth er (T M S ) d e riv a tiv e s . K a s h to k and B re -
de r (3 ) attem pted ch ro m a to g ra p h y o f u n a lte re d e th yle n e g ly c o l
b y packed c o lu m n G C u sin g C a rb o w a x 20 M , b u t the re su lta n t
peak shape was po or, p o s s ib ly because o f th e rm a l bre akd ow n ,
o r because the packed c o lu m n d id n o t p ro v id e the necessary
re so lu tio n . I t is p re fe ra b le to c o n firm th e id e n tity o f g ly c o l
w ith o u t ch e m ica l a lte ra tio n s because the d e riv a tiz a tio n reac­
tio n m ay n o t go to c o m p le tio n , and th is fa ilu re in tro d u ce s o th e r
products in to the s o lu tio n used fo r an alysis. T h e re fo re , because
o f the ch ro m a to g ra p h ic peak shape and separation p ro p e rtie s
o f the g ly c o ls , the use o f the N u k o l c o a tin g fo r the c a p illa ry
c o lu m n appears to be the b e tte r ch o ice fo r these com pounds.
These 2 com pounds do n o t fo rm m o le c u la r io n s b y e le c tro n
io n iz a tio n (E l) m ass spe ctro m e try (4 ) n o r do th e ir T M S d e riv a ­
Figure 1. Reconstructed ion chromatogram of
tiv e s . T h e re fo re , P IC I w as in ve stig a te d because it can fo rm standards of propylene and ethylene glycol at 25 ng
pro to na te d m o le c u la r io ns (M H + ), w h ic h are in d ic a tiv e o f a each.
com p ou nd’s m o le c u la r w e ig h t. M H + io ns o f these com pounds
1346 M a t u s ik E t A l .: J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

Table 1. Ions produced by PICI mass spectrometry of

propylene and ethylene glycols with 3 reagent gases

% Relative abundance*30
Propylene glycol Ethylene glycol
Reagent ----------------------------------------------------------
gas (M -18) (MH+) (M + 18) (M -18) (MH+) (M + 18)

Methane 100 10 0 100 60 0

Isobutane 0 100 0 0 100 0
Ammonia 0 10 100 0 15 100

T he m etha no l reagent b la n k gave no ch ro m a to g ra p h ic re ­ 30 40 50 60 70 80 90

sponse, de m on stra ting th a t th e re was no system c o n ta m in a tio n M/Z

o r e x tra c t in je c tio n carryo ver. E xtra cts o f the te st sam ples w ere
Figure 2. Background-subtracted product-ion
concentrated 1 0 -fo ld , and 2 | iL p o rtio n s o f these concentrates
spectrum of propylene glycol.
w ere in je c te d in to the gas chro m atog raph /ta nd em mass spec­
trom eter.
T h e id e n titie s o f p ro p yle n e and e th ylen e g ly c o ls w ere con­ tita tiv e procedure. F o r q u a n tita tiv e M S , is o to p ic a lly labe led in ­
firm e d in 9 test sam ple extracts. O ne o f these 9 extracts gave 2 te rn a l standards sho uld be used.
peak responses fo r the m/z 77 p ro d u c t ions (F ig u re 4 ). These 2 C h e m ica l d e riv a tiz a tio n is o fte n used to enhance e ith e r the
responses w ere separated c h ro m a to g ra p h ic a lly b y m o re than chro m atog raph y o r the de tection . O u r d e riv a tiz a tio n studies o f
30 s. B o th spectra show the same io ns, b u t w ith d iffe re n t % R A these g ly c o ls show ed th a t even the T M S d e riv a tiv e d id n o t
values. T he e a rlie r-e lu tin g peak was id e n tifie d as 1,3- fo rm m o le c u la r io ns b y G C /M S /E I. O th e r researchers have ob­
p ro p a n e d io l, an iso m e r o f 1 ,2 -p ro p a n e d io l; the second peak served th is phenom enon (2 ). T h e re fo re , P IC I w as used to ob­
was 1,2 -p ro p a n e d io l at the co rre ct re te n tio n tim e . ta in m o le c u la r io n in fo rm a tio n . Because iso bu ta ne P IC I gave
T hus, the id e n tific a tio n o f 1,3 -p ro p a n e d io l, in a d d itio n to the m ost intense M H + io n , it w as the best ch o ice to enhance the
the o th e r g ly c o ls , was c o n firm e d in 1 o f the 9 test sam ple ex­ o v e ra ll sig n a l and, th e re fo re , u ltim a te d e te c tio n s e n s itiv ity .
tracts. V ariou s a n tifre e ze m anufacturers use d iffe re n t fo rm u la ­ M S /M S w as also the m ethod o f cho ice because the isobutane
tio n s . In th is p a rtic u la r test sam ple, the ad ulte rated anchovies reagent gas io n at m/z 57 w o u ld do m in ate any spectrum ob ­
co n ta in e d 1 ,3 -p ro p a n e d io l, as w e ll as pro p yle n e and ethylene ta in e d at lo w masses b y c o n ve n tio n a l sin gle -stage M S . T he
g ly c o ls . N o attem pt was m ade to id e n tify the o rig in o f the an­ p ro d u c t-io n spectra are cle an er than those ob ta in e d b y con ven­
tifre e z e . tio n a l M S because the fir s t qu ad rupo le acts as a m ass filte r and
C o n firm a tio n o f the id e n titie s o f p ro p yle n e and eth ylen e a llo w s o n ly a 1.5 am u w in d o w fo r the selected io n s, in th is case
g ly c o ls was based on agreem ent o f chro m a to g ra p h ic re te n tio n m/z 77 o r m/z 63 , to pass in to the c o llis io n c e ll fo r p ro d u c t-io n
tim e s and m ass spectra. M ass spectra o f the test sam ple extracts fo rm a tio n . T he re la tiv e abundances o f the io ns in the spectrum
w ere com pared w ith those o f au then tic standards. T he se le c tiv ­ are n o t in flu e n c e d b y those produced b y ba ckgro und . T he
ity o f M S /M S e lim in a te d responses fo r a ll io ns exce pt the 2
b e in g m o n ito re d (m/z 77 and m/z 63 ) and a llo w e d id e n tific a tio n
o f lo w m o le c u la r w e ig h t com ponents n o t e a s ily m o n ito re d b y R ch2-ch2
c o n v e n tio n a l sin gle -stage quad rupo le mass spectrom etry, E OH OH
L
w h ic h is m ore a ffected in the lo w scanning re g io n (un de r 100 A
T
am u) b y m a trix co e xtra ctive s. I
V
T he in te n s itie s o f the sig n a l responses fo r p ro p yle n e and E
eth yle n e g ly c o l standards w e re 1.6 m illio n counts on each A 50.0 -

spectrum (F ig u re s 2 and 3 ) fo r a 25 ng in je c tio n . T h is suggests

B
u
th a t at le ast a 10-fo ld decrease in am ount (25 0 p g ) c o u ld be N CH2CH2=OH
D
e a s ily m easured b y u sin g th is fu ll-s c a n approach. I f greater A 45
N
s e n s itiv ity w ere re q u ire d , a selected io n m o n ito rin g (S IM ) C M t
E 27
m ode w o u ld need to be used. G en erally, S IM an alysis w ill in ­
29
crease d e te ctio n lim its b y 5 - to 1 0 -fo ld ; how ever, io n in fo rm a ­ ill
30 40 50 60
tio n is lim ite d fo r id e n tity c o n firm a tio n . O f the test sam ple ex­ M/Z
tracts c o n firm e d , a ll w ere suggestive o f b e in g in the lo w
F ig u r e 3 . B a c k g r o u n d -s u b tr a c te d p r o d u c t-io n
nanogram range, i.e ., 1 -1 0 ng in je c te d . H o w eve r, th is w o rk
s p e c t r u m o f e t h y le n e g ly c o l.
w as in te n d e d o n ly as a c o n firm a tio n o f id e n tity , n o t as a quan-
 M a t u sik E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1347

Conclusion

T he use o f the N u k o l c a p illa ry c o lu m n G C a ffo rd s good

separation o f e th ylen e and p ro p yle n e g ly c o ls w ith o u t d e riv a ti-
z a tio n o r p re lim in a ry cleanup. T he use o f C l p ro vid e s m o le cu ­
la r w e ig h t in fo rm a tio n fo r these com pounds. T he p ro d u c t-io n
spectra p ro v id e fra g m e n t io n s th a t u n e q u iv o c a lly characterize
these g ly c o ls . T he a n alysis b y G C /C I/M S /M S o f an cho vy ex­
tracts co n ta in in g un alte red p ro p y le n e and e th yle n e g ly c o ls is a
u n iq u e m ethod fo r c o n firm in g the id e n tity o f these com pounds.

References

(1) Clarke, E.C.G . (Ed.) (1969) T h e I s o la tio n a n d I d e n tific a tio n

o f D r u g s , T he Pharm aceutical Press, London, p. 339
(2) Van W illaer, W .G., & B eem aert, H. (1983) Z L e b e n s m . U n -
Figure 4. Reconstructed ion chromatogram of the test
te r s. F o r s c h . 1 7 7 ,1 9 6 -1 9 9
extract, showing 1,3-propanediol, an isomer of both
propylene and ethylene glycols. (3) Kashtok, M „ & Breder, C.V. (1980) J. A s s o c . O ff. A n a l.
C h e m . 63, 168-172
(4) Heller, S.R., & M ilne, G.W .A. (1978) E PA /N IH M ass Spec­
tral D ata B ase, pp. 10,2 5
p ro d u c t-io n spectra are a m ore accurate d e p ic tio n o f the fra g ­
(5) Long, F.A., & Prichard, J.G . (1956) J. A m . C h e m . S o c . 78,
m e n ta tio n p a tte rn o f a com p ou nd because the y are n o rm a lly
266 3 -2 6 6 7
n o t in flu e n c e d b y o th e r io ns generated fro m the m a trix .
T he id e n tity o f d o m o ic a cid was also c o n firm e d in som e o f
the an cho vy e xtra cts; the liq u id ch ro m a to g ra p h ic io n sp ra y M S
tech niq ue used is described in a separate paper.
1348 N a k a m u r a E t A l .: J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

RESIDUES AND TRACE ELEMENTS

D e t e r m i n a t i o n o f P y r e t h r o i d R e s id u e s i n V e g e t a b le s , F r u i t s ,

G r a in s , B e a n s , a n d G r e e n T e a L e a v e s : A p p lic a t io n s to P y r e t h r o id

R e s id u e M o n i t o r i n g S t u d ie s

Y iim ik o N a k a m u r a , Y a s u h id e T o n o g a i, Y u k a r i T su m u r a , a n d Y o s ra o I to
N a tio n a l In s titu te o f H y g ie n ic Sciences, D iv is io n o f F ood C hem istry, Osaka B ranch, 1-1-43, Hoenzaka, C huo-ku, Osaka, 540, Japan

Determination of natural pyrethrins and 12 syn­ o f H e a lth and W e lfa re in Japan plans to le g is la te lim its fo r 50
thetic pyrethroids in agricultural products was in­ p e sticid e residues, in c lu d in g 6 p y re th ro id s (c y h a lo th rin , cype r­
vestigated. Vegetables and fruits were extracted m e th rin , d e lta m e th rin , flu c y th rin a te , p e rm e th rin , and flu v a li­
with acetone, filtered after addition of coagulating nate) and n a tu ra l p y re th rin s d e rive d fro m p yre th ru m .
solution, partitioned into n - hexane, and cleaned up V ariou s m ethods fo r e x tra c tio n and cle an up o f in d iv id u a l
on a Florisil column (as necessary). Pyrethroids p y re th ro id s are described in the lite ra tu re . P yre th ro id s in foo ds
were determined by gas chromatography/electron w e re de term in ed b y gas chro m atog raph y w ith e le c tro n capture
capture detection (GC/ECD) by using a methyl sili­ d e te ctio n (G C /E C D ) (3 , 4 ), c a p illa ry G C /E C D (5 -7 ), liq u id
cone-coated fused-silica capillary column, and re­ chro m atog raph y (4 , 5 ), and im m unoassay (8 ,9 ).
coveries were calculated by summing peak areas A co a g u la tin g reagent co n sistin g o f am m o niu m c h lo rid e
of the components. GC/mass spectrometry was and ph osp ho ric a c id was show n to be e ffic ie n t fo r the cleanup
used to identify the pyrethroids detected by o f c ro p extracts in carbam ate (1 0 , 11) and organophosphorus
GC/ECD monitoring. Grains and beans containing (1 2 ) pesticides.

lipids were analyzed by extraction with acetonitrile, W e developed an a n a ly tic a l m ethod fo r n a tu ra l p y re th rin s
partition into n - hexane, Florisil column cleanup, and 12 syn th e tic p y re th ro id s (flu c y th rin a te , flu v a lin a te , pe r­
and capillary GC/ECD. The coagulation method m e th rin , cy p e rm e th rin , fe n va le ra te , tra lo m e th rin , a lle th rin ,
was suitable for nonfatty crops such as vegetables, te tra m e th rin , fe n p ro p a th rin , c y h a lo th rin , c y flu th rin , and d e l­
fruits, and green tea leaves, because recoveries ta m e th rin ) in a g ric u ltu ra l pro du cts (F ig u re 1). T he m etho d w as

were good and Florisil column cleanup was not a p p lie d to p y re th ro id residue m o n ito rin g studies.

needed in most cases. The coagulation method

was not applicable to lipid-containing crops such METHOD
as grains and beans because of low recoveries. Re­
coveries for 18 crops at fortification levels of 0.25- Pesticide Standards
1.0 ppm were 60.0-103.5%. No pyrethroids were de­
tected from the nonfortified crops tested except P esticide standard p u ritie s w ere as fo llo w s : p y re th rin I,
green tea leaves, in which fluvalinate was detected 13.21% ; p y re th rin II, 12.00% ; p e rm e th rin , 9 1 .6 % ; cype r­
at 0.89 ppm. m e th rin , 95 .4% ; flu c y th rin a te , 93% ; fe n va le ra te , 9 5 .4 % ; flu ­
v a lin a te , 91 .0% ; tra lo m e th rin , 9 8 .7 % ). Standard solutio ns (1
m g /m L ) w ere prepared b y d isso lvin g each standard in «-hexane
P y re th ro id in se cticid e s are w id e ly used fo r the c o n tro l o f (tra lo m e th rin ) o r acetone (pyre th rin s, perm ethrin, cyperm ethrin,
insects in a g ric u ltu re , e lim in a tio n o f ho use ho ld insects in flu c y th rin a te , fenvalerate, and flu v a lin a te ). F igu re 1 shows the
Japan, and the p re v e n tio n o f in fe s ta tio n d u rin g storage in chem ical structures o f the 13 pyrethroids used in th is study.
the U n ite d States. T h e y are v e ry e ffe c tiv e , have sh o rt life tim e s
in the fie ld , and have re la tiv e ly lo w m am m a lia n to x ic ity (1 ,2 ). Pesticide Working Solutions
Ten p y re th ro id s (fe n va le ra te , p e rm e th rin , flu c y th rin a te ,
D ilu te standard s o lu tio n s w ith acetone, and prepare th e fo l­
c y p e rm e th rin , flu v a lin a te , c y c lo p ro th rin , tra lo m e th rin , cyha -
lo w in g w o rk in g s o lu tio n s: A , m ix tu re o f 20 p g p y re th rin s /m L ,
lo th rin , c y flu th rin , and fe n p ro p a th rin ) are c u rre n tly registered
10 p g flu c y th rin a te /m L , and 10 |ig T u v a lin a te /m L ; B , m ix tu re
b y the M in is try o f A g ric u ltu re and F ish e ry in Japan. N o p yre ­
o f 20 p g p e m ie th rin /m L , 10 p g c y p e rm e th rin /m L , 10 p g fe n -
th ro id s have been le g is la te d b y the L a w o f S a n ita tio n , and p y ­
va le ra te /m L , and 20 p g tra lo m e th rin /m L ; C : m ix tu re o f 10 p g
re th ro id residues sho uld n o t be detected in crops. T he M in is try
a lle th rin /m L , 20 p g te tra m e th rin /m L , 20 p g fe n p ro p a th rin /m L ,
10 p g c y h a lo th rin /m L , 20 p g c y flu th rin /m L , and 20 p g d e l-
R e c e iv e d O c to b e r 2 1 , 1 99 2. A c c e p te d F e b ru a ry 1 5 ,1 9 9 3 .
ta m e th rin /m L .
 N a k a m u r a E t A l .: J o u r n a l O f AO A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1349

H CN
C H ,V F ,C -< f \ n H -C -C O O C H
C H ,X
q C H (C H
fluvalinate
:a o
allethrin
CK CH,
^ 3 = C H —t I t -C O -O C H ;
C l\
,C = CH
C V CH,
permethrin
cyfluthrin
CH,
CH,
t / COO H ,C H = C H C H ,
CH,
> C = c h \ f H ^O
CH,
cin e rin I : R = — C H 3; c in e rin II: R = - C O O C H 3
( Z H 1 R ) - c is -
CH,
H CH,

CH, CCY V iCH=CHC,H,

,1
>C=CH

CH,
ja s m o lin I : R = - C H , ja s m o lin II : R = COOCH,
C II, CH,
H
Hv /C O O ,.. _ /
C H ,\( T T IL -C
C1\ CH, __ |
: = c i i y 'H „ / 7 c - ii
C = CH —^ • O CH R
CH, CH =CH,
c/ I
CH, R = - C H , ( p y r e t h r i n I ) ; C O O C H , ( p y r e th r in II)
cypermethrin pyrethrins

B\
c = c h c h 3c o o . . cn
>C = C H ^ ^ C O • O C H ,N ^ j ] ^ ^ j
Br H' CH,
ch 3 CH,
deltamethrin tetramethrin

CH, CH, COO^ /C N Br

T il 0. i CHl
Br,C TH11 ir COO, ,CN
)4 K
CH, I H
CH, hY h h /C V Y
CH,
fenpropathrin
tralomethrin

C H (C H ,) ,
flucythrinate

Figure 1. Structures of natural pyrethrins and 12 synthetic pyrethroids.

 1350 N a k a m u r a E t A l .: J o u r n a l Of AO AC In t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

F igu re 2. G a s c h ro m a to g ra m s of c o m p o n e n ts or iso m e rs of natural pyreth rin s an d 12 syn th e tic p y re th ro id s in

sta n d a rd w o rk in g s o lu t io n s (10 o r 20 pg/m L).
 N a k a m u r a E t A l .: J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3 1351

Peak ® C in e r in I Peak ® C in e rin II

Peak ® J a s m o lin I Peak © J a s m o lin II

Peak © P y r e th r in I Peak ® P y r e th rin II

F igu re 3A. M a s s sp e c tra of natural pyrethrins.

R e a g e n ts a n d A p p a ra tu s 1 3 0 ^ o v e rn ig h t b e fo re use. N e ith e r C e lite 545 n o r F lo ris il PR

was prew ashed b e fo re use.
(a ) C o a g u la tin g so lu tio n .— D isso lve 10 g am m onium c h lo ­
(c ) O th er rea g en ts.— A ce to n e , n-hexane, e th y l acetate, and
rid e and 20 m L ph osp ho ric a cid in 800 m L d is tille d w a te r (4 ).
a c e to n itrile w ere special grade fo r d e te rm in a tio n o f pesticides.
(b ) D ia to m a ceo u s ea rth a n d syn th etic m agn esiu m s ili­
S od iu m c h lo rid e (N a C l), anhydrous so d iu m su lfa te (N a 2S 0 4),
c a te .— C e lite N o. 545 (J o h n s -M a n v ille P roducts C o rp .);
p h o sp h o ric a cid (H 3P 0 4), and a m m o n iu m c h lo rid e (N H 4C1)
F lo ris il P R (6 0 -1 0 0 m esh, F lo rid in C o .). H e at F lo ris il PR at
w ere a n a ly tic a l grade.
1352 N a k a m u r a E t A l .: J o u r n a l O f AOAC I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

Peak ® P eak (D

Peak (2)

P e ak ( D

F igu re 3B. M a s s sp e c tra of flucythrinate.

Peak ©
F igu re 3D. M a s s sp e c tra of perm ethrin.

(d) F ilt e r p a p e r .— Ashless filte r paper N o . 5 A and 5C (A d -

van tec T o yo C o ., L td ., T o k y o , Japan).
(e) C h ro m a to g ra p h ic g la ss tu b e.— 30 cm x 15 m m id ; fo r
co lu m n chro m atog raph y.
(f) G C co lu m n .— F u s e d -s ilic a c a p illa ry c o lu m n C B P 1, 25
m x 0 .2 m m id , film th ickn ess 0.25 p m , m e th y l p o ly s ilo x a n e
liq u id phase, che m ica l-b o n d e d (S him ad zu C o rp .).
Peak ©
(g) H o m o g en iz er.— E xce l A u to H o m o g e n ize r (N ih o n S eki
K a isha L td .).
(h) R o ta ry e v a p o ra to r.— R E 111 R o ta va p o r (B u c h i,
S hiba ta K ag a ku K ik a i K o g y o ) eq uippe d w ith w a te r ba th and
vacu um pum p (M o d e l JS -7 5 A , A d v a n te c ). Set w a te r b a th to
3 5 ^)0 °C .
(i) G a s ch ro m a to g ra p h .— G -2 8 0 0 (Y anaco) eq uippe d w ith
e le c tro n cap ture d e te cto r (63N i). O p e ra tin g c o n d itio n s : co lu m n ,
C B P 1 ; c o lu m n tem perature, 250X 1; in le t and d e te cto r tem pera­
tu re , 270°C ; c a rrie r and m ake-up gas, N 2 (p u rity 99 .9 9 9 % );
flo w rate, 1.0 m L /m in ; in je c tio n m etho d, s p lit, s p lit ra tio 1:10;
F igu re 3C. M a s s sp e c tra of fluvalinate.
in je c tio n vo lu m e , 2 p L .
 N a k a m u r a E t A l .: J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1353

Peak (D
Peak ®

Peak (D

P eak ©

Figure 3F. Mass spectra of fenvalerate.

S a m p le E xtractio n

(a ) V egeta bles, fru its , a n d g reen tea le a v e s .— Place 20 g

m in ce d vegetables o r fru its o r 10 g m in ce d na tsum ika n peels
o r green tea leaves in to a b le n d e r cup. A d d 100 m L acetone, and
hom ogenize at 10 00 0 rp m fo r 2 m in . F ilte r hom ogenate
th ro u g h N o . 5 A filte r paper, reh om o gen ize re sid u e w ith 100
m L acetone, and filte r again. C o nce ntrate filtra te to 50 m L b y
u sin g a ro ta ry evaporator.
Figure 3E. Mass spectra of cypermethrin. A d d 5 0 m L c o a g u la tin g rea ge nt and 5 g C e lite 54 5, and le t
stand 30 m in w ith o cca sio n a l sha kin g. F ilte r m ix tu re un de r
vacu um , and w ash filte r cake w ith 50 m L a ce to n e -co a g u la tin g
reagent (1 + 1, v /v ).
(b) G ra in s a n d b ea n s.— S oak 10 g g ro u n d gra in s o r beans
in 20 m L w a te r fo r 2 h, tra n s fe r to b le n d e r cup, add 100 m L
( j) G a s ch ro m atograph /m ass sp ectro m eter.— D o u b le fo ­ a c e to n itrile , and ho m og enize a t 10 0 0 0 rp m fo r 2 m in . F ilte r
cu sin g m ass sp e ctro m e tric (M S ) system (J E O L J M S -D X 302) hom ogenate th ro u g h N o . 5 A filte r paper, rehom o gen ize residue
w ith gas chro m atog raph (JE O L M S G C G 0 6 ) and h ig h speed w ith 100 m L a c e to n itrile , and filte r ag ain. W ash filtra te w ith 30
data a c q u is itio n system (JE O L J M A -D A 5 0 0 0 ). m L a c e to n itrile -s a tu ra te d n-hexane, and con cen tra te ace toni­
trile -w a te r fra c tio n to < 30 m L .
S a m p le s
L iq u id -L iq u id Partitio n
C abbage, Chinese cabbage, cucum ber, japanese rad ish , to ­
m ato, eggp lan t, p u m p k in , spina ch, o n io n , p o ta to , apple, straw ­ A d d 100 m L 10% so d iu m c h lo rid e aqueous s o lu tio n and
b e rry, japanese sum m er orange, green tea leaves, m aize, w heat, 100 m L n-hexane to S a m p le E x tra ctio n (a ) o r (b ). Shake v ig ­
b ro w n ric e , and soybeans w ere purchased fro m re ta il sources. o ro u s ly 5 m in . A d d 100 m L n-hexane to aqueous la y e r and
1354 N akamura E t A l .: J ournal O f A O A C International V o l . 76, N o . 6 ,1 9 9 3

F ig u re 3 G . M a s s s p e c tr a o f tra lo m e th rin .

F ig u r e 3H . M a s s s p e c tr a o f a lle th rin .

repeat extraction. Partition pesticides into «-hexane. Collect or­ GC/ECD. GC conditions: apparatus, MS-GCG 06; column,
ganic layers, dehydrate with ca 2 0 g anhydrous sodium sulfate, CBP1; inlet temperature, 270X1; column temperature, 60°C
and concentrate to 5 mL for vegetables, fruits, and green tea (hold 2 min), increase 32°C/min to 250°C; carrier gas, He (pu­
leaves or 1 mL for grains and beans. rity 99.999%); flow rate, 1.0 mL/min; injection method, split­
less; injection volume, 2 pL. MS conditions: apparatus, JMS-
Florisil Column Cleanup
DX 302; data system, JMA-DA 5000; separator temperature,
Place 10 g Florisil PR suspended in n-hexane into a chro­ 255°C; ion source temperature, 250°C; acceleration voltage, 3
matographic tube plugged with cotton wool, and add 1 0 g an­ kV; ionization voltage, 70 eV; ionization mode, electron im­
hydrous sodium sulfate continuously. Transfer «-hexane ex­ pact; trap current, 300 pA; scan time, 1.0 s; scan range, m/z
tract quantitatively to Florisil column. Wash column with 50 50-700.
mL «-hexane and discard this fraction. Next, elute with 50 mL
«-hexane-ethyl acetate (70 + 30, v/v). Concentrate eluate to 5 Recovery
mL (the test solution).
Recoveries were determined in triplicate for nonfortified
Determination
crops purchased from retail sources at fortification levels of
Gas chromatography.—Apply test solution to gas chroma­ 0.25-1.0 ppm for each pyrethroid. Working solution was added
tograph and determine pyrethroids. Sum peak areas of all com­ to each test portion 2 h before blending, and fortified samples
ponents or isomers for pyrethroids. were analyzed by the method outlined above. Recoveries were
Gas chromatography/mass spectrometry.—Evaporate test calculated by a standard curve prepared by diluting pesticide
solution and dissolve the residue in 100 pL «-hexane (50-fold working solutions with «-hexane (from 0.1 or 0.2 pg/mL to 1
concentration). Identify pyrethroids by GC/MS if detected by or 2 pg/mL).
 N akamura E t A l .: J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3 1355

Peak ®

Peak (2)

F ig u r e 31. M a s s s p e c tr a o f te tra m e th rin .

F ig u r e 3 J . M a s s s p e c tra o f fe n p ro p a th rin .

i
1356 N akamura E t Al .: Journal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3

Peak (D

Peak ©

F ig u r e 3 K . M a s s s p e c tr a o f c y h a lo th rin .

R e su lts a n d D iscu ssio n time-consuming, so the split injection mode was adopted for
GC/ECD.
This method is the first application of the coagulation Pyrethroid isomers or components were identified by
method and capillary GC/ECD for determination of natural py- GC/MS by using standard solutions (200-1000 (Xg/mL). Mass
rethrins and 1 2 synthetic pyrethroids. spectra of these isomers or components in 13 pesticides are
The gas chromatograms of natural pyrethrins and 12 syn­ shown in Figures 3A-3M.
thetic pyrethroids are shown in Figure 2. Six components were The relative retention time of each component of those 13
detected in natural pyrethrins (cinerin I and II, jasmolin I and pesticide working solutions was investigated in triplicate. Re­
H, and pyrethrin I and H). Several peaks were detected in syn­ sults are shown in Table 1. Coefficients of variation (CVs) of
thetic pyrethroids: 2 from flucythrinate, fluvalinate, per- relative retention times for 3 experiments were within 0.1%.
methrin, fenvalerate, tetramethrin, and cyhalothrin; 4 from Complete separation of all components of the 13 pesticides by
cypermethrin and cyfluthrin; and 1 from tralomethrin, al- the CBP1 column was not possible. Fortification for recovery
lethrin, fenpropathrin, and deltamethrin. Pyrethrin I and II was divided into 3 groups for complete separation of the com­
peaks were broader than the peaks of other components or iso­ ponents: Group A contained pyrethrins, flucythrinate, and flu­
mers of 13 pesticides, because pyrethrin I and II are heat-unsta­ valinate; Group B contained permethrin, cypermethrin, fen­
ble and easily degraded by thermal conditions. The determina­ valerate, and tralomethrin; and Group C contained allethrin,
tion of pyrethroids such as tralomethrin and deltamethrin are tetramethrin, fenpropathrin, cyhalothrin, cyfluthrin, and del-
 N akamura E t A l .: J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3 1357

Peak ©

Peak ©

Peak ©

Peak @

F ig u r e 3 L . M a s s s p e c tr a o f c y flu th rin .

tamethrin. Minimum detection limits were 0.002-0.05 ppm for Coagulating solution converts indigenous compounds co­
vegetables and fruits and 0.005-0.1 ppm for green tea leaves, extracted from plant tissue samples into a precipitate that can
citrus peels, grains, and soybeans. be removed by filtration (12). The coagulation method was suit­
Florisil column cleanup was not necessary for most vegeta­ able for analysis of natural pyrethrins and 1 2 synthetic pyiethoids
bles and fruits but was necessary for onion, spinach, peels of
in nonfatty crops such as vegetables and fruits because of good
japanese summer orange, green tea leaves, grains, and soy­
recoveries without the need for column cleanup in most cases.
beans. Interfering compounds could not be excluded from the
«-hexane extracts of onion, peels of japanese summer orange, Pyrethroids are nonpolar and have a high affinity for «-hex­
and green tea leaves, despite Florisil column cleanup. ane because it is also nonpolar. Recovery by the coagulation
method was low (nearly less than 2 0 %; data not shown) for
Î358 N akamura E t Al .: J ournal O f A O A C International V ol . 76, N o . 6 ,1 9 9 3

F ig u r e 3M . M a s s s p e c tr a o f d e lta m e th rin

grains and soybeans. Thus, the coagulation method is not suit­ No pyrethroids were detected in nonfortified crops tested in
able for the analysis of these commodities. This fact suggests this study except green tea leaves. Fluvalinate residue (0.89
the possibility that lipids and pyrethroids coprecipitate. The ppm) was only detected in nonfortified green tea leaves. An
acetonitrile-hexane partition step was performed to remove ECD/gas chromatogram of nonfortified green tea leaves is
some of the lipids. If this partition step was completed more shown in Figure 4. Fluvalinate was identified by GC/MS. Mass
than twice, recoveries of pyrethroids were low (nearly less than spectra of fluvalinate in a test solution of green tea leaves and
40%; data not shown). Therefore, the acetonitrile-hexane par­ standard solution are shown in Figure 5.
tition step for removal of lipids was performed only once. Re­ Recoveries of natural pyrethrins and 12 synthetic pyre­
coveries ranging from 60.0 to 103.5% were obtained for lipid- throids in 18 crops fortified at 0.25-1.0 ppm are shown in Ta­
containing crops such as grains and beans by acetonitrile ble 2. Values are the means for triplicate determinations. CV of
extraction followed by partitioning into «-hexane and Florisil the recovery for each pesticide was within 1 0 % (data not
column cleanup. shown). Recoveries of 13 pyrethroids were 60.3-102.7% for

F ig u r e 4. G a s c h ro m a to g ra m o f te s t s o lu tio n o f n o n fo rtifie d g re e n te a le a v e s .
 N akamura E t Al .: J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3 1359

T a b le 1. R e la tiv e re te n tio n tim e s a n d m in im u m d e te c tio n lim its o f n a tu ra l a n d 12 s y n th e tic p y r e th r o id s 3

Minimum detection limits, ppm

G reen tea leaves, citrus peels,

Pesticides Relative retention tim es0 Vegetables and fruits grains, an d soybeans

Pyrethrins
P eak ® Cinerin 1 0.5 75 0.025 0 .0 5
P eak ® Jasmolin 1 0.6 70 — —

P eak ® Pyrethrin 1 0.8 90 — —

P eak © Cinerin II 1.141 — —

P eak ® Jasmolin II 1.373 — —

P eak © Pyrethrin II 1.929 — —

Flucythrinate
P eak ffi 2.220 0.025 0 .0 5
Peak© 2.3 72 — —

Fluvalinate
P eak ® 3.1 09 0.025 0 .0 5
P eak ® 3.1 85 — —

Permethrin
P eak ® 1.579 0.05 0.1
Peak® 1.639 — —

Cypermethrin
P eak ® 2.081 0.05 0.1
Peak® 2.1 48 — —

P eak ® 2.2 08 — —

Peak® 2.2 38 — —

Fenvalerate
P eak ® 2.7 90 0.025 0 .0 5
P eak ® 2.9 96 — —

Tralomethrin 3.5 83 0.025 0 .0 5

Allethrin 0.4 60 0.002 0 .0 0 5
Tetramethrin
Peak® 0.901 0.01 0 .0 2 5
P eak © 0.9 29 — —

Fenpropathrin 1.000 0.005 0.01

Cyhalothrin
Peak ® 1.213 0.002 0.0 05
P eak © 1.285 — —

Cyfluthrin
P eak ® 1.904 0.01 0 .0 2 5
Peak ® 1.965 — —

P eak ® 2.0 15 — —

Peak® 2.0 46 — —

Deltamethrln 3.591 0.002 0 .0 0 5

a Pyrethroids determined by ECD/GC. Operating conditions: apparatus, G2800; column, CBP1; column, 250°C; injection port, 270X1; detector,
270°C; detector, ECD; carrier gas flow, N2, 1.0 mL/min; injection method, split; split ratio, 1:10; injection volume, 2 pL.
b Fenpropathrin (12.52 min) calculated as 1.000.

vegetables, 68.0-102.4% for fruits, 77.0-92.9% for green tea tamethrin (500 pg/mL; n-hexane solution for tetramethrin and
leaves, 64.5-101.3% for grains, and 60.0-103.5% for soybeans. acetone solutions for other pyrethroids) was kindly supplied by
Seiya Kato, Japan Food Research Laboratories.
A ck n o w led g m en ts
R eferen ces
We are greatly indebted to the Ministry of Health and Wel­
fare and Japan Food Research Laboratories for supplying the (1) Casida, J.E., & Ruzo, L.O. (1980) Pestic. Sci. 11,257-269
pesticide standards. Each standard solution of allethrin, tetra- (2) Demoute, J.-P. (1989) Pestic. Sci. 27, 375-385
methrin, fenpropathrin, cyhalothrin, cyfluthrin, and del- (3) Braun, H.E., & Stanek, J. (1982) J. Assoc. Off. Anal. Chem.
65, 685-689
 1360
T a b le 2a. R e c o v e r ie s o f 13 p y r e th r o id s in a g r ic u ltu r a l p r o d u c t s 9 a t f o r tific a tio n le v e ls o f 0.25 a n d 0.5 p p m

Rec., %

Na k a m u r a
Japan ese
Fortification Chinese Japan ese sum m er orange
Pesticides level, ppm C abbage cabbage Cucum ber radish Tomato Eggplant Pumpkin Spinach Onion Potato Apple Strawberry (Fruit)

Et A l . :
__b
Pyrethrins 0.5 75.3 86.4 81.6 78 .0 92 .0 83.2 90 .0 82 .5 88 .4 78 .6 68.0 83 .2
Flucythrinate 0.25 81 .3 86.2 81.0 83 .8 75 .0 80.0 94 .0 81 .9 90 .9 88.0 79 .3 80.4 78.0

Jo
Fluvalinate 0.2 5 79.6 81.1 88.0 90 .0 70 .3 75.8 85.1 78 .0 89 .8 82 .0 80 .4 77.1 81.0

u r n a l
Permethrin 0.5 87 .0 85.3 84.5 9 6 .7 80 .4 92 .0 80 .0 93.1 77.1 92 .2 88 .9 90 .3 89 .2
Cypermethrin 0.25 81.3 85.4 90.5 8 0 .0 80 .5 94 .0 70 .7 93.1 9 5 .0 93 .4 86.0 82 .0 93.2

Of AOAC I n t e r n a t i o n a l
Fenvalerate 0.2 5 81 .6 83.5 95.0 82 .5 95.0 96 .6 65 .8 97.2 7 8 .9 88.8 88.6 95 .5 90 .0
Tralomethrin 0.5 85 .0 87.0 99.3 86.6 102.7 88.0 67 .0 88.0 7 8 .4 83.1 7 9 .9 90.1 85 .5
Allethrin 0.25 84.1 84.3 80.6 92 .8 100.0 77.3 66 .9 75.1 9 4 .2 79 .9 77 .8 85 .4 94.6
Tetramethrin 0.5 86.1 84.3 84.6 89 .7 102.5 83.3 6 5 .8 73.2 80 .0 80 .7 7 5 .4 81 .4 101.3
Fenpropathrin 0.5 90 .8 85.0 88.4 8 4 .7 100.0 77.9 60 .3 75.2 8 2 .7 87 .5 8 1 .0 85 .5 99 .7
Cyhaiothrin 0.2 5 97 .6 78.5 84.4 8 2 .8 93 .2 77 .8 6 2 .0 80 .0 84 .8 84 .8 7 8 .4 95.4 94 .2
Cyfluthrin 0.5 95 .3 78.5 85.7 8 2 .6 95 .0 82.1 65 .6 82.1 8 5 .5 85 .6 7 5 .5 95 .7 102.7
Deltamethrin 0.5 94 .8 80 .2 82.5 84 .6 101.5 80.0 68.2 91.0 7 7 .7 91 .2 7 4 .8 90 .3 101.3

V
.o l
76, No. 6,1993
T a b le 2 b. R e c o v e r ie s o f 13 p y r e th r o id s in a g r ic u ltu r a l p r o d u c t s 9 a t f o r tific a tio n le v e ls o f 0.5 a n d 1.0 p p m

Rec., %

Fortification level, Japanese sum m er

Pesticides ppm orange (peel) G reen te a leaves M aize W heat Brown rice Soybean

__b __b
Pyrethrins 1.0 81.6 8 4 .0 81.1 101.5
__b 99 .7
Flucythrinate 0.5 77 .0 88.7 90.1 10 1.7
Fluvalinate 0.5 83.8 85.1 90.3 81.1 9 9 .6 92.8
__b
Permethrin 1.0 78 .8 78 .7 9 4 .0 103.5 60.6
Cypermethrin 0.5 83 .3 82 .8 85 .8 8 8 .5 103.3 72 .4
Fenvalerate 0.5 86.0 92 .9 93 .4 101.5 100.0 73.0
Tralomethrin 1.0 77 .4 80.1 102.1 9 0 .4 101.1 60.8
__b
Allethrin 0.5 80 .7 98.1 8 9 .9 9 7 .4 67.2
Tetramethrin 1.0 76 .9 8 3 .4 89.5 93.1 96 .0 60.0
Fenpropathrin 1.0 87 .4 8 2 .0 100.0 6 4 .5 10 0.5 61.6
__b
Cyhaiothrin 0.5 8 0 .3 93.5 88.1 101.2 60.7
Cyfluthrin 1.0 72 .8 8 1 .5 92.0 86.6 102.5 60.6
Deltamethrin 1.0 66.0 78.1 91.5 8 5 .0 9 6 .4 60.0

3 Pyrethroids analyzed using CBP1 column. S ee Table 1 for G C /E C D conditions.

b Determ inations w ere impossible because of the interfering peak.
 N akamura E t Al .: J ournal O f A O A C I nternational V ol . 7 6 ,N o . 6,1993 1361

F ig u r e 5. M a s s s p e c t r a o f flu v a lin a te o b ta in e d fro m s ta n d a rd s o lu t io n a n d te s t s o lu tio n o f n o n fo rtifie d g re e n te a

le a v e s : A , T IC o f s ta n d a r d s o lu t io n o f f lu v a lin a te (200 |ig /m L in n-h exane); B , m a s s s p e c tr a o f flu v a lin a te fro m s ta n d a rd
s o lu tio n o f flu v a lin a te ; C , T IC o f t e s t s o lu t io n o f g re e n te a le a v e s (10 g/0.01 m L in n -h exane); a n d D, m a s s s p e c tr a o f
flu v a lin a te o b ta in e d fr o m t e s t s o lu t io n o f n o n fo rtifie d g re e n te a le a v e s .

(4) Goto, M., & Kato, S. (1987) Analytical Methodsfor Residual (8) Stanker, L.H., Bigbee, C., Emon, J.V., Watkins, B., Jensen,
Pesticides, an Enlarged Edition, Soft Science Co., Tokyo, Ja­ R.H., Morris, C., & Vanderlaan, M. (1989) J. Agric. Food
pan Chem. 37, 834-839
(5) Bottomley, P„ & Baker, P.G. (1984) Analyst 109, 85-90 (9) Skerritt, J.H., Hill, A.S., McAdam, D.P., & Stanker, L.H.
(6) Dicke, W., Ocker, H.-D., & Thier, H.-P. (1988) Z. Lebensm. (1992) J. Ague. Food Chem. 4 0 , 1287-1292
Unters. Forsch. 186, 125-129 (10) Holden, E.R. (1973) J. Assoc. Off- Anal. Chem. 56, 713-717
(7) Nakamura, Y., Hasegawa, Y., Tonogai, Y., & Ito, Y. (1990) (11) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
Eisei Kagaku 36, 525-537 lington, VA, sec. 975.40
(12) Sasaki, K., Suzuki, T, & Saito, Y. (1987) J. Assoc. Off. Anal.
Chem. 7 0 , 460-464
1362 N iemann : J ournal O f A O A C International V ol . 76 , N o . 6 ,1 9 9 3

RESIDUES AND TRACE ELEMENTS

D eterm in atio n o f Form etanate H ydrochloride in Selected F ru its

by C oupled-C olum n C ation Exchange L iq u id C hrom atography

R ic h a r d A . N ie m a n n
U.S. Food and Drug Administration, Division of Contaminants Chemistry, Washington, DC 20204

A s t r o n g c a t io n e x c h a n g e ( S C X ) li q u id c h r o m a ­ the methodology is time consuming. These tedious single resi­

t o g r a p h ic ( L C ) m e t h o d i s d e s c r ib e d f o r d e t e r m in a ­ due methods (3-5) used formetanate’s weak basicity to isolate
t io n o f f o r m e t a n a t e h y d r o c h lo r i d e r e s id u e in p o m e , it from coextractives through a series of multiple liquid-liquid
c it r u s , a n d s t o n e f r u it s . A t e s t p o r t io n o f f r u it , h o ­ partitionings; pH was adjusted to cycle the analyte between its
m o g e n iz e d w it h t h e p e e l le f t o n , w a s b le n d e d w it h free-base and cationic conjugate-acid forms, as many times as
a c id if ie d a c e t o n it r ile a n d f ilt e r e d . A p o r t io n o f e x ­ necessary, to enhance cleanup. Special precautions were nec­
t r a c t w a s f in e ly f ilt e r e d , a n d a 5 0 0 | iL a liq u o t ( c a 0 .2 essary to minimize losses due to hydrolysis, the rate of which
g t e s t s a m p le e q u iv a le n t ) w a s lo a d e d o n t o a n S C X accelerated while formetanate was in the free-base form. After
s o l id - p h a s e e x t r a c t io n ( S P E ) L C c o lu m n , w h ic h re ­ residue isolation, electron capture gas chromatographic (GC)
p la c e d t h e in j e c t io n lo o p o f t h e L C in j e c t io n v a lv e . methods (3, 4) specified lengthy processes of hydrolysis and
C a t io n s w e r e s e le c t iv e l y e n r ic h e d ; n o n c a t io n s
derivatization of the hydrolysis product to form a detectable
w e r e e lu t e d b y a c e t o n it r ile in a p r e - s e p a r a t io n
surrogate. Replacing GC with reversed-phase liquid chroma­
c le a n u p . T u r n in g t h e v a lv e t o t h e in j e c t p o s i t io n
tography (LC) with UV detection (5) eliminated these steps,
but sample throughput was still limited by continued depend­
c o u p le d t h e S P E c o lu m n t o a n S C X a n a ly t ic a l c o l ­
ence on liquid-liquid partitionings for cleanup. In addition, for­
u m n f o r s e p a r a t io n a n d d e t e c t io n a t 2 5 0 n m . T h e
metanate was chromatographed as the hydrolytically suscepti­
m o b i le p h a s e w a s 0 .4 M p H 3 .0 a m m o n iu m p h o s ­
ble free base at pH 8 , very near the upper pH limit of stability
p h a t e b u f f e r - w a t e r - a c e t o n i t r il e (5 0 + 2 5 + 2 5 ). F o r ­
of bonded-phase silica.
m e t a n a t e c a t io n w a s q u a n t it a t e d b y p e a k a r e a a n d
Our research goal was to develop faster and simpler meth­
r e g r e s s io n c o e f f ic ie n t s f r o m a 5 - p o in t lin e a r c a li b r a ­
odology for use in the U.S. Food and Drug Aministration’s
t io n c o v e r i n g a 1 0 0 - f o ld r a n g e . R e c o v e r y o f d u p l i ­
(FDA’s) program of residue tolerance enforcement. Because
c a t e f o r t i f i c a t io n s o f a p p le , p e a r, o r a n g e , a n d p e a c h
the legal tolerance was established for residues of the parent
a v e r a g e d 8 9 - 0 9 % a t t h e r e s p e c t iv e U . S . t o le r a n c e s
formetanate hydrochloride only, determination of various for­
o f 3 , 3, 4 , o r 5 p p m a n d a v e r a g e d 9 3 - 9 9 % a t o n e -
metanate metabolites, except for the free base formetanate, was
t e n t h o f t h e r e s p e c t iv e t o le r a n c e le v e l. P e e l p ig ­
not a method objective. The analytical approach kept for­
m e n t s o r v a r ia b le p e e l b u lk o f c r o p v a r ie t ie s t e s t e d ,
metanate in the more stable conjugate acid form by specifying
a s w e ll a s o t h e r e n d o g e n o u s f r u it m a t e r ia l, c o n t r ib ­
extraction and chromatography at low pH. All separatory fun­
u t e d in t e r f e r e n c e t h a t w a s b e lo w t h e 0 .0 2 p p m lim it nel extractions were replaced by a single solid-phase extraction
o f d e t e c t io n . In a 1 9 9 1 lim it e d s u r v e y c o m p r is in g 15 (SPE) and enrichment by strong cation exchange (SCX).
s a m p le s , n o n e w e r e f o u n d v io la t iv e . R e s id u e s w e r e Placement of the SPE column in the flow path normally occu­
f o u n d in 2 s a m p le s , b u t o n ly 1 m e a s u r e m e n t w a s pied by the injection loop of the LC injection valve provided
q u a n t if ia b le , n e a r t h e 0 .0 6 p p m lim it o f q u a n t it a t io n . on-line coupling and elution to an SCX analytical column for
separation and UV quantitation of the formetanate cation. A
pre-separation cleanup of the SPE column was performed
F ormetanate hydrochloride (Figure 1) is a miticide/acari- while the 2 columns remained uncoupled. The present paper
cide with U.S. tolerances (1) on various fruits. The U.S. describes the method based on the coupled-column SCX
Environmental Protection Agency has estimated (2) that LC/UV approach and presents performance data and results
about 150 tons were used in the United States during 1989, from testing with apples, pears, oranges, and peaches.
primarily on apples and oranges (36% and 33% of total usage,
respectively), and that the trend in overall usage has been in­ E x p e r im e n t a l
creasing. However, food monitoring data are lacking because
Reagents
Received September 14, 1992. Accepted February 3, 1993
This paper was presented at the Sixteenth International Symposium on (a) Solvents.—Acetonitrile, UV; water and methanol, high
Colum n Liquid Chromatography, June 14-19, 1992, at Baltimore, MD. purity (Baxter, Burdick & Jackson, Muskegon, MI).
 N e m a n n : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1363

port injection valve and electric actuator (Models C6 U, or

o equivalent, and E60, Valeo Instrument Co., Inc., Houston, TX);
II fill port syringe adapter (for 0.75 in. needle, P/N VISF-1), in­
o - c — nhch 3 jection loop not used; column compartment convection oven
(DuPont Instruments; LC line has since been sold to The An-
spec Co., Ann Arbor, MI); variable wavelength UV detector
with 8 pL flow cell (DuPont Instruments); integrator equipped
with timed-events (I/O) and data capture/communications ca­
ch3 pabilities (Model 4270, Spectra Physics Analytical); PC data
I system (Model Chromstation-AT, Spectra Physics Analytical).
N*HCI Operating conditions: pump at 1.0 mL/min; columns at 35 3C;
I detector at 250 nm and 0.020 AUFS range. Keep space behind
ch3 pump’s piston seals filled with water-methanol (90 +10) solu­
tion delivered from attached syringe for batchwise flush-out
F ig u r e 1. S tru c tu r e o f fo rm e ta n a te h y d ro c h lo rid e . and renewal.
(b) LC columns.—3 cm x 4.6 mm id Whatman Partisil
SCX, 10 pm, SPE column (P/N 635-030-26, Keystone Scien­
tific, Bellefonte, PA); 25 cm x 4.6 mm id ZORBAX 300-SCX,
7 pm, analytical column (MAC-MOD, Chadds Ford, PA).
(b) Acids and bases.—Phosphoric acid (85%), LC grade
(c) LC syringes.—Gas-tight with removable needle, 2 each
(Fisher, Fair Lawn, NJ); hydrochloric acid (36.5-38%, 12M),
of 100, 250, and 500 pL sizes (Models 1710, 1725, and 1750,
ACS reagent grade (Fisher); ammonium hydroxide (30%), re­
respectively , Hamilton, Reno, NV); and 2 gas-tight 1.0 mL size
agent grade (Baker, Phillipsburg, NJ). Prepare 50 mL aqueous
with Teflon Luer-Lok (TLL; Model 1001, Hamilton). Injection
solution of 1.2M HCl from 10-fold dilution of concentrated
valve and syringe adapter specified above require 0.75 in.,
acid.
blunt end point, style No. 3 needles (removable type, listed in
(c) Buffers.—Standard solutions of pH 4 and pH 7 (Baker).
same order as syringe sizes above, P/Ns 76517, 76518, and
Prepare 0.4M ammonium phosphate solution at pH 3.0 by add­
76515; Kel-F Luer hub to fit TLL type, P/N 90172; Hamilton).
ing 46.2 g phosphoric acid and 24 mL ammonium hydroxide
Other valves or adapters may have different needle require­
to 950 mL water stirred magnetically. Add additional base until
ments.
pH measures 2.98-3.00 by combination glass electrode.
(d) Other syringes.—Several 10 mL size glass with metal
(d) LC mobile phase.—Dynamically proportion (or alter­ Luer-Lok (Becton Dickinson M U LilEil, VWR Scientific,
natively premix) A -B -C (50 + 25 + 25), where solvent “A” is
Philadelphia, PA; or equivalent).
0.4M ammonium phosphate buffer, pH 3.0; solvent “B” is
(e ) pH measurement.—Digital pH meter with ATC probe
water; and solvent “C” is acetonitrile. Degas before pumping.
(Model 811, Orion Research, Boston, MA) and combination
(e) Analytical standard.—Obtain formetanate hydrochlo­
glass electrode with epoxy body (Model 91-55, Orion). Per­
ride (Std. No. CR-3680) from Ultra Scientific, North Kings­ form a 2-point standardization in pH 4 and 7 standard buffer
town, RI.
solutions.
(f) Formetanate hydrochloride standard solutions.—Pre­
(f) Food processor.— 1 3/4 qt capacity (Classic, Model
pare 100 pg/mL stock solution in acetonitrile plus 1-2% water DLC-10C, Cuisinarts, Greenwich, CT).
by accurately weighing 10.0 mg standard into 10 mL beaker, (g) Blender.—Explosion proof, variable speed (Waring,
transferring it by acetonitrile rinses to a partially filled 100 mL New Hartford, CT) with 1 L glass jar and A1 foil-wrapped
volumetric flask, dissolving it with the aid of 0.5 mL additions cover.
of water as needed, and diluting to volume. Make quantitative
(h ) Filters.— 11 cm sharkskin (Schleicher & Schuell,
1 0 -fold serial dilutions in acetonitrile to prepare 1 0 .0 , 1 .0 0 , and
Keene, NH) and 13 mm Nylon 6 6 disposable syringe filter,
0.10 pg/mL LC calibration solutions in separate 10 mL volu­
0.45 pm (Rainin, Woburn, MA).
metric flasks. Prepare 10.0 pg/mL acetonitrile solution by di­
luting 5.00 mL stock solution in 50 mL volumetric flask. Use LC Coupled-Column Configuration and Conditioning
this solution to fortify at low levels ( 1 0 % of tolerance), and use
stock solution for high level (tolerance) fortifications. Store all Refer to Figure 2 for details and component location. Install
these solutions in freezer, where they remain stable for months. 3 cm SPE column in flow path of injection valve normally oc­
(g) Extraction solution.—Acetonitrile-1.2M HQ (2000 + 2, v/v). cupied by injection loop, such that flow direction through col­
umn reverses when valve’s position is changed. Using short
(h) Conditioning solution.—Water-acetonitrile (70 + 30, v/v).
lengths of 0.010 in. id tubing, install 0.5 pm precolumn filter
Apparatus (P/N A-316, Upchurch, Oak Harbor, WA) ahead of each LC
column and have available 10-20 replacement filters (P/N A-
(a) Liquid chromatograph.—Ternary solvent proportion­ 103X, 10/pkg, Upchurch). Between pump’s drain valve (not
ing pump equipped with sapphire pistons and back-up seals shown) and injection valve, install 0.5 pm in-line filter (P/N
(Model 8800, Spectra Physics Analytical, San Jose, CA); 6 - 05-105, SSI, State College, PA).
1364 N ie m a n n : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

Clean and condition columns (valve in the inject position), Integrator M ethod File
using the following recommended mobile phases at 2 mL/min:
0-50-50 (hold 10 min) to 0-100-0 in 5 min; 0-100-0 (hold 5 Use external standard method and peak area, set up the com­
min) to 100-0-0 in 5 min; 100-0-0 (hold 10 min) to 50-25-25 ponent table, and establish linear multilevel calibration at 5 lev­
in 5 min; hold at 50-25-25 final mobile phase for LC analyses. els in microgram quantities entered to 3 significant figures.
After 20 min pumping at final mobile phase, reduce flow Disable peak integration until the baseline has recovered from
rate to 1 mL/min and let system equilibrate. Perform this pro­ negative excursions caused by void volume elution; activate
cedure once daily before starting a sequence of chroma­ peak integration typically at 4—5 min into the ran. Enter 3.5%
tographic analyses. At the end of a day’s work, prepare the col­ for the range (component window) in which analysis retention
umns for overnight storage in buffer-free mobile phase as times are compared against formetanate cation retention time
in the component table. When LC system has equilibrated,
follows: Increase flow rate to 2 mL/min; run gradient 50-25-
have the integrator evaluate signal noise.
25 to 0-70-30 in 5 min; 0-70-30 (hold 5 min) to 0-30-70 in 2
min; hold 0-30-70 for 5-10 min before stopping the pump. Multilevel Calibration
CAUTION: Do not let pump components and solvent lines
contact buffers overnight. After shut-down, stop pump, put Calibrate at 0.02,0.06,0.2,0.6, and 2.0 pg formetanate hy­
Solvent “A” (buffer) inlet line into water, open pump drain drochloride. Use dedicated syringes. Begin with 200 pL most
valve, purge pump several minutes, stop pump, change “A” dilute LC calibration standard solution. Rinse outside of needle
inlet to methanol, repeat purging, stop pump, close drain valve, with solvent, and then insert into syringe adapter to load aliquot
and leave stored in methanol. Flush fresh solution behind pis­ onto SPE column (details below). After highest calibration
ton seals. level has been injected, flush inlet pathway to injection valve
(bypassing SPE column) with 0.5 mL acetonitrile followed by
1 mL water-acetonitrile conditioning solution, to remove ana­
lyte traces adsorbed onto tubing walls. When sequence has
been completed, integrator computes and prints linear regres­
sion coefficients (intercept and slope), which are used inter­
nally in subsequent analyses to quantitate peak area response
at analyte retention time.

Test Sam ple Preparation

L o a d
Remove apple and pear stems and peach pits. Section ca 3
lb of each fruit, leaving on peel. Add fruit to food processor
P o s itio n
while chopping; chop until consistency is homogeneous.
Weigh 100 g portions into cleaned 4 oz jars, cover with alumi­
num foil (acetone-rinsed, dull side down), cap, and store in
freezer until analysis.

Extraction

Rinse all glassware and filter paper on Buchner funnel with

10 mL extraction solution delivered from tilting dispenser
head. Measure 200 mL extraction solution in 250 mL glass-
stopper graduated cylinder. For recovery study, pipet 3,4, or 5
mL of desired fortifying solution into cylinder and mix. Pour
extraction solution (with or without analyte added) onto
In je c t thawed test portion of fruit placed in bottom of blender jar,
blend contents at high speed for 5-6 min, and filter through
shark skin paper, using vacuum. Rinse blender jar with two 10
P o s itio n
mL portions of fresh extraction solution, each time pouring
rinse onto filter cake and combining it with main extract. When
no drop of liquid appears at end of funnel stem, cease filtration
and pour filtrate back into cylinder. Record extract volume
(convert height above 250 mL mark to mL, usually 19 mm =
20 mL; 290 mL maximum capacity is rarely exceeded). Invert
cylinder 1 0 times with vigorous agitation to thoroughly mix
F ig u re 2. C o n n e c t io n s to a V a le o 6 -p o rt L C in je c tio n extract (lower immiscible layer <5-10% total volume). Wait 30
v a lv e th a t r e v e rs e th e d ir e c t io n o f f lo w b e tw e e n lo a d a n d s (ca 90 s for apple) for some settling of particulates, which
in je c t p o s itio n s . A : s y r in g e a d a p te r a n d u n io n , B: 0.5 p m would otherwise completely block syringe filter. Draw up
filte r, C : L C p u m p , D: S P E c o lu m n , E: a n a ly tic a l c o lu m n . 11-12 mL off the top with a syringe, attach a disposable syringe
 N ie m a n n : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1365

APPLE PEAR

ORANGE PEACH

F ig u r e 3. R e c o v e r y s tu d y c h r o m a to g r a m s o f 4 f r u its fo rtifie d a t to le r a n c e le v e l: a p p le a n d p e a r a t 3.0 p p m , o ra n g e at

4.0 p p m , a n d p e a c h a t 5.0 p p m ( L C a n a ly s is o f c a 0 .2 g e q u iv a le n t). A r r o w in d ic a te s th e fo rm e ta n a te c a tio n p e a k .
A b s o r b a n c e w a s m e a s u re d a t 2 5 0 n m a n d s c a le d t o 0 .0 2 0 A U F S , th e s ig n a l w a s d ig it iz e d a n d c o n v e rte d to A S C II
fo rm a t f o r c o m p u te r g r a p h ic s s o ftw a re m a n ip u la tio n , a n d th e d is p la y e d o u tp u t w a s e q u iv a le n t t o a s t r ip c h a rt re c o rd
a tte n u a te d 8 -fo ld .

filter, and discharge 1 mL to waste before collecting 10 mL in where V is total extract volume (mL), Q is quantitation from
vial. peak area (pg), W is test sample weight (g), and A is aliquot
(mL) of extract loaded onto SPE column. Typically, VP = 100
Liquid Chromatography and A = 0.500.

Uncouple 2 columns by actuating valve to load position. R e su lts a n d D iscu ssio n

Perform following sequence for loading SPE column: (7) 0.5
mL conditioning solution to flush out buffer salts; (2) 500 pL
filtered extract, or 60 or 200 pL desired calibration solution; (5) On-line Enrichment
0.5 mL acetonitrile; and (4) 0.5 mL conditioning solution to
The short SCX column selectively extracted cations from
avoid buffer salt precipitation when coupled to analytical col­
the test solutions passed through it during the loading se­
umn.
quence. Acidification of the extraction solution kept for­
Switch valve to inject position, which couples columns, be­ metanate and other free-base nitrogen residues in a cationic
gins chromatography, and starts integrator. Formetanate cation recoverable form. Any neutral or anionic compounds in the test
elutes between 7.7 and 8.1 min. solution that were retained by secondary separation mecha­
nisms, such as partitioning or adsorption, were eluted by ace­
C alculation tonitrile in a pre-separation cleanup of this SPE column. For­
metanate cation remained virtually immobilized by strong
Calculate residue concentration in test sample as follows: interaction with not only the fixed charge sulfonate groups but
also the aromatic backbone of the bonded phase (analytical col­
Formetanate HC1, ppm = (V x Q)/(W x A) umn phase was structurally similar). Eluting with only a high
1366 N ie m a n n : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

APPLE PEAR

( a ) A/>------ J l XV'
(b )/ A ^ ^ / “v f B a r tle tt

CD

ORANGE PEACH

F igu re 4. Partial c h ro m a to g ra m s of reco very s tu d y te st p o rtio n s (a) fortified at 1 0 % o f to lerance (apple a n d p e ar at

0.3 ppm , o ra n g e at 0.4 ppm , an d pe ach at 0.5 p p m ) an d (b) c ro p b la n k s of se le cte d varieties. A rro w in d ic ate s the
retention tim e o f form etanate cation. A n a ly s is an d m e a su re m e n t c o n d itio n s w ere the s a m e a s th o s e fo r F igu re 3
e x ce p t y -a x is w a s attenuated 8 tim es.

ionic strength mobile phase was unsuccessful until acetonitrile umn performance after many switching operations appeared
was blended in. During optimization of the mobile phase com­ unaffected by the small 5-8 bar change in system pressure con­
position, analyte retention by the coupled columns was more tributed by the SPE column.
profoundly affected by small changes in the acetonitrile con­ A study of extraction efficiency showed that formetanate
tent than by changes in the ionic strength. Premixing the mobile cation was quantitatively recovered from 25 to 500 pL loadings
phase was advisable for pumps with poor proportioning repeat­ of standard test solutions. For example, recovery was complete
ability. for 0.5 (_ig delivered in 50 and 500 pL loadings and for 2.5 pg
The relatively larger particle size of the packing material delivered in 25 and 250 pL loadings.
chosen for the SPE column made the manual loading operation
easier but tended to degrade coupled-column chromatography. Chromatography
Therefore, opposite flow directions were used for the loading
Analysis of test portions of recovery study test samples gave
and injection steps to prevent chromatography in the lead col­
the representative chromatograms in Figures 3 and 4, which
umn. The SPE column functioned solely as a chemical filter for
compare the relative magnitudes of and interferences from
cations now held in a tight band at the column end closer to the
food matrix material. Figure 3 shows that formetanate cation,
analytical column. As the lead column, its lifetime was most which was fortified at the tolerance level for the respective fruit
affected by the step pressure rise and fall from the atmosphere, and eluted in about 8 min, was well separated from the earlier,
resulting from valve switching. The use of wide-pore (300 A major crop coextractives. Orange provided the most extrane­
[30 nmj) packing material in the analytical column maintained ous material absorbing at 250 nm. In Figure 4 are displayed
the system head pressure (55 ± 5 bar) low enough to protect the partial chromatograms of fortifications 1 0 -fold below toler­
SPE column from pressure-shock damage. Also, analytical col­ ance and of crop blanks, all with signal expansion to elucidate
 N ie m a n n : J o u r n a l O f AO A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1367

whether varietal differences in peel color or peel thickness in­ T able 1. R e c o v e ry of form etanate h yd ro c h lo rid e fortified
fluenced the detector response equivalent of formetanate. in dup licate at 2 levels
Among apple blanks of red and golden peels, virtually none Ree., %
gave responses above baseline where the analyte eluted except
10% of tolerance level3 Tolerance level3
for Red Delicious, the peak of which, although discernible, was
still below the detection limit of 0.02 ppm. Oranges of thick Fruit, variety
Trial 1 Trial 2 Trial 1 Trial 2
(navel) and thin (Valencia) peel gave surprisingly similar and
undetectable crop blank responses. Despite the percentage Apple, Red

change that different orange peels made to the test sample com­ Delicious 92.1 94 .6 89.0 89.6
Apple, Golden
position, the inedible peel bulk was not a source of analytical
Delicious 97.8 99 .4 100.2 97 .5
bias. In summary, none of the 4 fruits tested interfered with
Apple, Rome 99.1 96 .5 96.2 95.6
formetanate cation measurement above the limit of detection
Pears, Bartlett 94.1 93 .4 93.1 94.2
(LOD). Chromatographic mn times of 12 min for apple, pear,
Orange, Navel 97 .3 9 7 .0 90.4 89.2
and peach extracts and 30 min for orange extract left the col­
Orange, Valencia 95 .7 93 .8 91.4 92.5
umn clean for the next analysis. Peaches,
Freestone 94.0 91 .9 94.6 94 .7
Calibration
Av. (n = 14) 95 .5 93.4

The 2-5 ppm range of established fruit tolerances was ade­ % RSD 2.6 3.6

quately encompassed by the 5-point calibration spanning 100-

U.S. tolerance (ppm): apple and pear, 3.0; orange, 4.0; peach, 5.0.
fold from approximately 0.1 to 10 ppm. Assuming a non-zero
intercept model, the set of 13 daily calibrations from the recov­
ery study had coefficient of determination R2 values >0.99992, tion level. At the 95% confidence level, the F-test statistic was
an average slope of 3.56 x 105 (scaled peak area counts/|j.g) consistent with the assumption of no difference in variance be­
with 0.9% relative standard deviation (RSD), and an average tween the lower and higher recovery data sets (calculated F =
intercept of 97 (scaled peak area counts) with 447% RSD. 1.84 <4.28 critical F for 6 , 6 degrees of freedom and 95% con­
Measurement bias was controlled more by the large spread in fidence). Average analyte recovery of duplicate determinations
intercept, ranging ±700 about zero, than by the weaker or non­ was 89-99% for additions of 3.0-5.0 ppm and was 93-99% for
existent responses observed with crop blanks. Accordingly,
calibration data and statistics were used to compute the LOD
and the limit of quantitation (LOQ) in terms of the standard
deviation of calibration blanks (SDb): LOD = 3 x SDb and LOQ
= 10 x SDb. The calibration blank (|ig) at zero response was
calculated from (-intercept/slope) or -b/m. For n = 13, SDb was
0.0012, which gave an LOD o f0.0036 (lg and an LOQ of 0.012
(J.g. If A is peak area response, a peak is detected if [(A-b)/m]
>LOD and is quantitated if [(A-b)/m] >LOQ. Assuming a 270
mLextract (the middle value of the 250-290 mL typical range),
then LOD and LOQ become, respectively, 0.02 and 0.06 ppm.
Within-day retention time (RT) repeatability was usually
0.2-0.3% RSD from calibration. Despite the excellent preci­
sion, the actual RT for analysis of fruit extracts was always
longer. The crop extracts, unlike the calibration solutions, have
ionic strengths that, together with the organic solvent compo­
nent, met mobile phase conditions for organocation elution in
SCX columns. Formetanate cation, carried further into the SPE
column during loading, required more time to exit during in­
jection. On average, the extra retention time was small: less
than 0.1 min for pome fruit, less than 0.15 min for orange, and
less than 0.2 min for peach. Keeping the extract loading vol­ TIME (min)
ume constant at 500 pL stabilized this increase in retention F igu re 5. Lim ited su rv e y c h ro m a to g ra m s o f (a) im port­
time. If desired, a portion of extract from each crop of interest ed A s ia n pear with 0.07 p p m incurred re sid u e at the
could be fortified and chromatographed to specify analyte RT LO Q , (b) P au la red a p p le with 0.03 p p m incurred residue
for subsequent analyses of that crop. a b o v e the L O D bu t b elo w the LO Q , an d (c) Bartlett pear
with n o detectable residue. A rro w in d ic ate s the retention
Recovery Study
tim e of fo rm etanate cation. M e a su re m e n t c o n d itio n s
Results presented in Table 1 verify that formetanate recov­ w ere the s a m e a s th o se for F igu re 4 e x ce p t the early half
ery was complete and independent of commodity and fortifica­ o f the c h ro m a to g ra m s w a s attenuated 64 tim es.
1368 N ie m a n n : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

additions of 0.3-0.5 ppm. Overall, the average recovery for 28 the methodology is being successfully extended to other resi­
trials was 94.5% with 3.2% RSD. dues amenable to cation exchange with the goal of a mul­
tiresidue method for later publication.
Limited Survey

The method was applied to 15 fmits purchased locally at A ck n o w led g m en ts

chain food stores, farmers’ markets, or orchards during the
summer of 1991. Most of them, about equally distributed I thank Gary Schieffer, Procter and Gamble, for helpful dis­
among apples, pears, peaches, and oranges, had no formetanate cussions on the analytical approach: Terry Yates, Analytical
residues above the LOD. The 2 exceptions were one imported Biosystems Inc., for a generous gift of a cartridge SCX column
Asian pear sample, which contained 0.07 ppm (the LOQ), and useful in exploratory investigations; Richard Albert, FDA, for
one Paula red apple, which measured 0.03 ppm (between the statistical analysis, and Donald Reec, FDA, for insightful in­
LOD and the LOQ). Their chromatograms, plus one from the formation about pesticides and their regulation.
set of blank residue analyses, are compared in Figure 5 to show
the relative peak height response of formetanate incurred R efe ren c es
slightly above the LOD and at the LOQ.
In summary, a single, selective cleanup by solid-phase ex­ (1 ) Code o f Federal Regulations ( 1 9 9 2 ) T itle 4 0 , U .S . G o v e r n ­
traction and cation enrichment coupled on-line to LC separa­ m e n t P r i n t i n g O I F ic e , W a s h i n g t o n , D C , s e c . 1 8 0 . 2 7 6
tion and measurement made this method faster, simpler, and (2 ) B r u c e , C .M . ( J u ly 1 9 8 9 ) U .S . E n v i r o n m e n ta l P r o te c tio n
easier to use than existing methodology for determination of A g e n c y R e p o r t, O ffic e o f P e s tic id e P r o g ra m s , E c o n o m ic
parent residues. Analysis time from blending to LC determina­ A n a ly s is B ra n c h , W a s h in g to n , D C

tion was less than 30 min. Recoveries and measurements were (3 ) Pesticide Analytical Manual, Volume 7 / ( 1 9 6 8 a n d re v is io n s )
quantitative and precise. The method was not tested with sam­ U .S . F o o d a n d D r u g A d m in is tr a tio n , W a s h in g to n , D C , s e c .
ples known to have field-incurred formetanate residues, as 1 8 0 .2 7 6

these were unavailable. Significant findings should be con­ (4 ) J e n n y , N .A ., & K o s s m a n , K . ( 1 9 7 4 ) Anal. Methods Pestic.
firmed by independent analysis using a different method. Once Plant Growth Regul. 7, 2 7 9 -2 9 6

samples with incurred residues are found, validation of this (5 ) L a w r e n c e , J .F . , P a n o p i o , L . G . , L e w i s , D . A . , & M c L e o d ,

method against a referenced method can proceed. Currently, H .A . ( 1 9 8 1 ) J. Agric. Food Chem. 29, 7 2 2 -7 2 4
 P y l y p iw : J o u r n a l O f A O A C In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1369

R E S ID U E S A N D T R A C E E L E M E N T S

Rapid Gas Chrom atographic M ethod fo r the M ultiresidue

Screening of Fruits and Vegetables fo r Organochlorine and
Organophosphate Pesticides

H a r r y M. P y l y p iw , J r
The Connecticut Agricultural Experiment Station, Department of Analytical Chemistry, 123 Huntington St, New Haven, CT
06504-1106

A rapid a n d reliable m ethod w a s dev elo p ed for th e based on the sample matrix. The method provides accurate re­
d eterm in atio n of p e stic id e s in fru its a n d v e g e ta ­ sults for a variety of organochlorine and organophosphorus
b les. A 100 g sa m p le is ex tra cted with a m ixture of pesticides that are present in fruit and vegetable samples. The
200 mL petroleum e th e r a n d 100 mL 2-propanol. method uses a simple extraction and cleanup; pesticides are
T he e x tra ct is b ac k w ash ed 4 tim es, tw ice with a q u e ­ determined by capillary GC coupled with highly specific de­
o u s so d iu m su lfate a n d tw ice with 350 mL distilled tectors for organochlorine and organophosphate pesticides.
w ater, a n d th e n dried o v er 15 g so d iu m su lfate. T he
dried ex tract is an alyzed by capillary g a s c h ro m a ­ Experim ental
to g rap h y with selec tiv e o rg a n o ch lo rin e a n d o rg an o -
p h o s p h o ru s d etectio n . T he m eth o d d e te rm in e s pri­
m arily n o n p o lar p esticid es, with re co v erie s ranging Apparatus
from 81 to 114%, a n d h a s an av e ra g e limit of d e te c ­
tion of 10 pp b for both d e te c to rs. (a) Gas chromatograph.— Model 5890 (Hewlett-Packard
Co., Avondale, PA), equipped with the following detectors: (7)
63Ni electron capture detection (ECD) system; (2) flame
P esticides are applied to a broad variety of crops to reduce photometric detection (FPD) system operating in P mode (in S
losses from weeds, insects, and diseases. The determina­ mode for confirmation); (3) electrolytic conductivity detection
tion of pesticide residues in fruits and vegetables is a for­ (ELCD) system operating in the halogen mode. One GC was
midable task, because the actual chemicals used on a crop are configured with dual detectors (for ELCD/FPD), and another
seldom known. In addition, residue testing of fmits and vege­ was configured with a single detector (for ECD). General op­
tables that are sold to the public must be rapid, because state erating conditions were as follows: initial temperature, 175°C;
and federal regulatory agencies must know if the pesticide con­ no initial hold time; ramp rate, l°C/min; final temperature,
tamination on a crop exceeds its allowable tolerance or if a 250°C; final hold time, 10 min; total run time, 85 min; carrier
pesticide is present on a crop where it is not allowed ( 1 ). gas, He; make-up gas for ECD, 5% C’H4 /Ar, with flow rate of
To analyze fruits and vegetables for pesticides, many labo­ 20 mL/min; make-up gas for all other detectors, He, with flow
ratories use multiresidue screening methods that are outlined in rate of 20 mL/min. Gas purifier (all gasses except air), OMI-1
the U.S. Food and Drug Administration’s Pesticide Analytical (Supelco Inc.) indicating purifier. Injector, HP-19251A; tem­
Manual (2). Most of these methods use solvent extraction cou­ perature, 225°C; operated in the splitless mode; purge off time,
pled with preparatory chromatography or solvent partitioning, 0.50 min. Auto injector, HP-7673; 2-4 |iL injection volume.
followed by gas chromatography (GC) with various detectors. Detector conditions: ECD, 325°C; FPD, 265°C, flame mixture,
Each method, however, is limited by the sample matrixes that hydrogen-air; ELCD, Model 4420 (OI Analytical Corp., Col­
can be processed, the extraction and cleanup steps, and the lege Station, TX); reactor temperature, 900°C; vent time,
sample preparation time. For instance, several of the methods 3.5 min.
require additional steps depending on the sugar content of the (b) Chromatographic column.—Capillary, 30 m x
sample; other methods have lengthy cleanup steps to remove 0.53 mm, 0.5 pm film, SPB-1 (Supelco Inc., Bellefonte, PA).
coextracted plant pigments. Alternative columns, 30 m x 0.53 mm, or 15 m x 0.53 mm,
In this study, a multiresidue method was developed that is 0.5 pm film, SPB-5, SPB-608, and SPB-20.
quick, simple, and needs no modifications of the procedure (c) Data collection.—All GC data were collected on a PE-
Nelson 1020X, dual channel personal integrator (Perkin-Elmer
Corp., Norwalk, CT).
Received November 16, 1992. Accepted M arch 4, 1993.
This paper was presented at the 29th Annual Pesticide Residue (d) Glassware.—Separatory funnel, 1000 mL, with Teflon
Workshop for the Florida Department of Agricultural Services, July 1992. stopcock and glass stopper; filtering funnel, 150 x 150 mm;
1370 P y l y p iw : J o u r n a l O f A O A C In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

glass wool; cylinder, graduated with stopper, 100 mL (Fisher decanted into a separatory funnel from the blender container
Scientific Co., Pittsburgh, PA). through a glass filtering funnel containing a small plug of glass
(e) Food cutter.—Model 84145 (Hobart Corp., Troy, OH). wool in the neck. Approximately 130-150 mL of the extract
(f) Blender.—Explosion-resistant (Waring), Cat. No. 14- was collected from the blended sample.
509-53; blender containers, 1 qt, Cat. No. 14-509-11A (Fisher To the separatory funnel, 350 mL distilled water and 20 mL
Scientific). saturated sodium sulfate solution were added. The separatory
funnel was capped, vented, gently swirled for 1 min, and then
R eagents allowed to stand for 15-20 min to let the water and petroleum
(a) Solvents.— Petroleum ether (30-60°C), Cat. No. 9265- ether layers separate. After separation, the water layer was
03; 2-propanol, Cat. No. 9334-03; toluene, Cat. No. 9336-03; drained off and discarded. With some crops, such as tomatoes,
2,2,4-trimethylpentane, Cat. No. 9355-03 (J. T. Baker Chemi­ apples, and grapes, a small emulsion forms at the solvent-water
cal Inc., Phillipsburg, NJ). interface. That emulsion should not be discarded until the final
(b) Distilled water.—Water was used directly from the still wash.
(Bamstead/Thermolyne, Model A1013-B) without further pu­ The petroleum ether layer was washed by gently agitating
the separatory funnel with 2 more 350 mL portions of distilled
rification.
water without added sodium sulfate. After the third wash was
(c) Saturated sodium sulfate.—Approximately 250 g an­
drained, 100 mL distilled water and 10 mL saturated sodium
hydrous granular sodium sulfate (Cat. No. 3375-05, Resi-Ana­
sulfate solution were added to the separatory funnel, and the
lyzed grade, Baker) was added to 800 mL distilled water and
funnel was swirled for 20-30 s. This final wash served to break
warmed on a steam bath until the sodium sulfate crystals dis­
solved. The solution was cooled overnight at room temperature any emulsions formed from the previous water washes.
to allow the excess sodium sulfate crystals to precipitate. The final wash was discarded, and the petroleum ether ex­
tract was transferred to a graduated cylinder containing 15 g
(d) Pesticide standards.—All pesticides were obtained
anhydrous sodium sulfate. The cylinder was stoppered and
from the U.S. Environmental Protection Agency, Pesticides &
Industrial Chemicals Repository, Research Triangle Park, NC. shaken for 1 min, and the contents were allowed to settle for 1 0
All standards were prepared from the compound as received. min. This step was repeated 2 more times. Before the GC de­
An approximately 10-15 mg sample of each pesticide was termination, the extract was allowed to stand over the sodium
weighed accurately on an analytical balance and dissolved in sulfate >1 h to let any particulates settle. (At this point, the pro­
25 mL toluene, and the concentration was calculated. This was cedure could be stopped and the extract held for GC determi­
the individual pesticide stock standard. The stock standard was nation at a later date.) The extraction procedure is summarized
diluted with 2,2,4-trimethylpentane to give a 10 ppm interme­ in Figure 1.
diate standard, from which individual and mixed standard so­ Sam ple Spike Preparation
lutions were prepared. Individual standards were prepared at
0 . 2 and 1 ppm; mixed standard solutions were prepared with A separate 100 g sample was taken after the chopping or
individual analyte concentrations ranging from 0.1 to 3 ppm. blending step, and a known quantity of pesticide standard(s)
was added to this sample. No attempt was made to determine
Sam ple Preparation if the sample to be spiked contained any pesticide residues be­
Fruits and vegetables were processed as received in their fore spiking. The pesticide concentrations in spiked samples
raw, unwashed, and unpeeled form, unless otherwise specified were usually 0.1, 0.2, and 0.5 ppm. The spiked samples were
in the protocol given in the Code o f Federal Regulations, then prepared in the manner described in Sample Preparation.
§180.1 (1). Most fmit and vegetable samples, such as apples, GC Calibration
squash, bell peppers, cucumbers, cabbage, and carrots, were
placed in a food cutter and chopped for 2-3 min. Other sam­ A single-point GC calibration using pesticide standard mix­
ples, such as tomatoes, peaches (pitted), plums (pitted), straw­ tures was performed daily before and after each sample set. A
berries (hulled), and grapes, were placed in a blender container 3- point standard curve using individual pesticides was con­
and blended at high speed for 1-2 min. Samples of apple cider structed over 2 orders of magnitude, 0.05, 0.5, and 5 ppm, at
or orange juice did not require any pre-extraction processing. 4- month intervals. Response to the standards was linear for all
A 100 g portion of the fruit or vegetable sample was placed 3 GC detectors.
into a 1 qt blender container, and 100 mL 2-propanol was
Calculation of Sam ples and Spiked Sam ples
added. The container was covered, and the sample was blended
at high speed for 20-30 s. The blender speed was reduced, and The weight: volume ratio of the final extract for GC injec­
200 mL petroleum ether was added to the jar. The sample was tion was based on the initial sample weight vs the initial vol­
then blended at reduced speed (ca 40% full speed) for 3-4 min. ume of petroleum ether, i.e., 1 g sample equaled 2 mL petro­
(The 200 mL aliquot of petroleum ether must be measured ac­ leum ether. All pesticide concentrations were calculated by
curately, because sample concentration calculations are based using an external standard calculation based on peak height.
on that volume.) Spiked samples were calculated in the same manner as regular
After blending, the sample was allowed to settle in the samples. Recovery of the spikes was calculated as amount
blending container for 2-3 min. The extract was then slowly found divided by the amount spiked plus the amount of pesti-
 P y l y p iw : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1371

T able 1. P e s tic id e s reco vered fro m sp ik e d s a m p le s

with 3 or m ore replicate determ ination s

No. LOD LOD LO D

replicates ECD, ELCD, FPD , Rec.
Pesticide ppm ppm ppm range, %

Azinphos-methyl 15 __ __ 0.0 5 8 9 -1 0 6
Captan 8 0.1 0.0 2 — 8 2 -1 1 4
Chlordane 5 0.0 02 0.01 — 9 3 -1 0 2
Chlorothalonil 6 0.0 02 0.01 — 9 0 -1 0 3
Chlorpyrifos 11 0.0 02 0.01 0.0 2 8 7 -1 0 6
DC PA (Dacthal) 7 0.0 05 0.0 2 — 8 2 -1 1 2
DDE 17 0.0 02 0.01 — 8 3 -1 0 8
DDT 4 0.01 0.0 5 — 8 8 -9 4
Diazinon 7 — — 0.0 2 8 4 -1 1 3
Didoran 8 0.0 02 0.0 2 — 9 7 -1 0 9
Dicofol 7 0 .0 0 5 0.1 — 8 6 -1 0 3
Dimethoate 3 — — 0 .0 5 8 7 -1 0 1
Endosulfan I 11 0.0 02 0.01 — 8 4 -1 0 5
Endosulfan II 10 0.0 02 0.01 — 8 6 -1 0 8
Endosulfan sulfate 5 0.0 02 0.01 — 8 2 -9 8
Ethion 9 — — 0.01 9 1 -1 0 8
Heptachlor epoxide 3 0.0 02 0.01 — 8 5 -9 7
Iprodione 6 0.01 0.2 — 97-111
lsofenphosa 3 — — 0.0 2 8 1 -8 7
Malathion 3 — — 0 .0 2 8 7 -1 0 3

F igu re 1. M etho d outline. Parathion-methyl 4 — — 0 .0 2 8 9 -1 0 3

Permethrin 4 0.01 0.5 — 8 5 -1 0 8
Phosmet 8 — — 0.0 5 84-111
Procymidone 5 0.0 02 0 .0 5 — 8 7 -1 1 3
cide in the sample, if any. All samples containing residues
Vinclozolin 11 0.0 02 0.01 — 8 7 -1 0 6
> 1 0 times the calibration standard amount were diluted to be

within the calibration range, reinjected, and recalculated. a This pesticide not found in any sam ple from 1 9 8 8 -1 9 9 2 .

R e su lts a n d D iscu ssio n

that usually lowers pesticide recovery. For these reasons, ace­

This multiresidue method was developed in 3 steps. Step 1
tone or acetonitrile extracts are unsuitable for direct determina­
was the selection of appropriate pesticides to be included in the
tion by GC/ECD (4,5).
method, step 2 was the development of an extraction procedure
for a variety of fruit and vegetable samples, and step 3 was the
analysis of the extracts by capillary GC with 3 GC detectors.
T able 2. P e s tic id e s reco vered fro m sp ik e d s a m p le s
Initially, the following 6 pesticides were selected on the ba­ with few er than 3 rep licates
sis of pesticide residues found in produce in Connecticut before
No. LO D Minimum
1988 (3): chlorothalonil, chlorpyrifos, endosulfan, dicofol, Pesticide replicates ECD, ppm rec., %
DDE, and vinclozolin. These compounds could easily be de­
tected and quantitated by GC/ECD. As development of the Aldrin3 1 0.0 02 88
method progressed, additional pesticides were found in the ß -B H C 3 1 0.0 02 92
samples analyzed by using GC/ELCD and GC/FPD. Tables 1 Dieldrin 2 0.0 02 83
and 2 list the pesticides that were recovered over the past 5 Endrin 2 0.0 02 83
years with this method, the limits of detection (LOD) for each Heptachlor3 1 0.002 89
GC detector, and the recoveries from spiked samples. HCB 1 0.0 02 93

Extraction of pesticides from fruits and vegetables by using Lindane3 2 0.002 91

a solvent such as acetone or acetonitrile, which is miscible in Methoxychlor3 1 0.01 94

plant materials, is the usual approach of most multiresidue Mirex3 1 0.0 02 82

O xadiazon3 2 0.0 05 83
methods (2,4,5). However, acetone and acetonitrile will also
TDE 2 0.002 89
extract large amounts of undesirable polar coextractable plant
Toxaphene3 1 0.0 2 89
compounds, especially from matrixes such as com, celery, and
spinach. The coextraction of various plant pigments often re­ This pesticide not found in any sam ple from 1 9 8 8 -1 9 9 2 .
quires the use of an additional, time-consuming cleanup step
1372 P y l y p iw : J o u r n a l O f A O AC In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

2 4

F ig u r e 2. C h r o m a to g r a m s o f a p e s tic id e s ta n d a rd
m ix tu re (A), s tra w b e rry s a m p le (B), a n d c h e r r y s a m p le
(C ) a n a ly z e d b y G C / E L C D , h a lo g e n m o d e ; G C c o lu m n ,
3 0 m x 0.53 m m , 0.5 p m film , S P B -1 . S ta n d a rd c h ro m a to ­
g ra m A : (1) d ic lo ra n , 0.50 p p m ; (2) c h lo ro th a lo n il,
0.50 p p m ; (3) v in c lo z o lin , 0.5 p p m ; (4) c h lo r p y r ifo s ,
0.50 p p m ; (5) c a p ta n , 1.00 p p m ; (6) e n d o s u lfa n I, F ig u re 3. C h r o m a to g r a m s o f a p e s t ic id e s ta n d a r d
0.25 p p m ; (7) p , p i - D D E , 0.25 p p m ; (8) e n d o s u lfa n II, m ix tu re (A), s tra w b e rry s a m p le (B), a n d c h e r r y s a m p le
0.25 p p m ; (9) ip ro d io n e , 3.00 p p m ; (10) p e rm e th rin , (a) (C ) a n a ly z e d b y G C / E C D ; G C c o lu m n , 3 0 m x 0.53 m m ,
c/s-, (b) t r a n s - , 2.00 p p m to ta l. S tra w b e rry s a m p le B : (3) 0.5 p m film , S P B -6 0 8 . S ta n d a rd c h ro m a to g ra m A : (1)
v in c lo z o lin , 0.05 p p m ; (5) c a p ta n , 10 p p m . C h e r r y s a m p le d ic lo r a n , 0.10 p p m ; (2) c h lo r o th a lo n il, 0.40 p p m ; (3)
C : (1) d ic lo ra n , 0.45 p p m ; (9) ip ro d io n e , 0.38 p p m . v in c lo z o lin , 0.2 p p m ; (4) c h lo r p y r ifo s , 0.20 p p m ; (5)
c a p ta n , 1.00 p p m ; (6) e n d o s u lfa n I, 0.20 p p m ; (7)
p , p i - D D E , 0.20 p p m ; (8) e n d o s u lfa n II, 0.20 p p m ; (9)
To achieve a cleaner extract, 2 key points had to be ad­ ip ro d io n e , 1.00 p p m ; (10) p e rm e th rin , (a) c/s-, (b) t r a n s - ,
dressed. First, a solvent mixture had to be chosen that would 1.00 p p m to ta l. S tra w b e rry s a m p le B : (3) v in c lo z o lin ,
extract pesticides quantitatively from the sample matrixes. 0.05 p p m ; (5) c a p ta n , 10 p p m . C h e r r y s a m p le C : (1)
Second, interfering water-soluble compounds that were coex­ d ic lo r a n , 0.45 p p m ; (9) ip ro d io n e , 0.38 p p m .
tracted by the solvent mixture had to be removed. To accom­
plish these objectives, a 2 + 1 mixture of petroleum ether and
2 -propanol was used to extract the samples, followed by a Method blanks, spiked samples, and duplicate samples were
backwash with distilled water to remove the coextracted inter­ tested on a weekly basis. The method blank showed no inter­
ferences. This wash step removed hydrophilic compounds fering peaks at the retention times of the pesticides of interest.
from the extract and produced a finished extract that was dried Recoveries of spiked samples ranged from 81 to 114%, and the
over sodium sulfate. The procedure is summarized in Figure 1. relative standard deviations of both spiked and duplicate sam­
This method, based on a mixed solvent extraction followed ples within a group of 6 - 8 samples and their respective spiked
by a distilled water backwash, was successful in eliminating samples prepared at the same time ranged from 2 to 5%.
many of the water-soluble coextracted interferences from Chromatograms of authentic sample extracts are shown in
vegetables such as tomatoes, green beans, and squash, as well Figures 2-4. Because the final extract contains very few inter­
as com, celery, and spinach. The extract is suitable for direct fering peaks, as is shown in the figures, an additional benefit is
analysis by GC/ELCD, GC/FPD, and GC/ECD. Spiked sam­ lower LODs, especially when ECD is used (Figure 3). LODs
ples processed by the method produced excellent recoveries. for most pesticides in samples prepared by this method and
These data are given in Table 1. Table 2 lists additional pesti­ analyzed by GC/ECD are 2 ppb. For GC/ELCD and GC/FPD,
cide compounds that were recovered by the method but were LODs are in the range of 10 ppb, on the basis of a signaknoise
not completely validated by the criteria given by Conacher (6 ). ratio of 3:1.
 P y l y p iw : J o u r n a l O f AO A C I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1373

In summary, the method provides fast and convenient sam­

ple preparation with clean extracts that can be analyzed by a
wide variety of GC detectors, including an ECD system. Re­
sults of actual sample testing by this method from 1988 to 1992
were summarized in bulletins of The Connecticut Agricultural
Experiment Station (7-11). Further studies with the method
will validate its use for more pesticides and will emphasize
GC/MS for confirmation of pesticide findings.

A cknow ledgm ents

I thank Maty Jane Incorvia Mattina (Connecticut Agricul­

tural Experiment Station) and Michael W. Dong (Perkin-Elmer
Corp.) for helpful discussions and guidance with this work. I
thank Terri Misenti, Chris Baldoni, and Serge Golden for their
technical assistance. I also thank John Wadhams and Ellen
Sloan (Connecticut Department of Consumer Protection) for
providing the many samples that were necessary to accomplish
this work.

R efe ren c es

(1) Code of Federal Regulations (1990) Title 40, Parts 180.1-

180.1085
0 10 20 30
(2) Pesticide Analytical Manual (1968 Rev.) Vol. I, U.S. Food &
Time (min) Drug Administration, Washington, DC
F ig u r e 4. C h r o m a to g r a m s o f a p e s tic id e s ta n d a rd
(3) Hankin, L. (1988) Pesticide Residues in Produce Sold in
Connecticut, Bull. 863, The Connecticut Agricultural Experi­
m ix tu re (A ) a n d a lim e s a m p le (B ) a n a ly z e d b y G C / F P D ;
ment Station, New Haven, CT
G C c o lu m n , 15 m x 0.53 m m , 0.5 p m film , S P B - 1 .
(4) Luke, M.A., Froberg, J.E., Doose, G.M., & Masumoto, H.T.
S ta n d a rd c h ro m a to g r a m A : (1) d im e th o a te , 0.50 p p m ; (2)
(1981) J. Assoc. Off. Anal. Chem. 64, 1187-1195
d ia z in o n , 0 .5 0 p p m ; (3) p a ra th io n -m e th y l, 0.5 p p m ; (4)
(5) Okumura, D., Melnicoe, R., Jackson, T., Drefs, C., Maddy,
c h lo r p y r ifo s , 0 .5 0 p p m ; (5) is o fe n p h o s , 0.50 p p m ; (6)
K., & Wells, J. (1991) Rev. Envir. Contam. Toxicol. 118, 87-
e th io n , 0 .5 0 p p m ; (7) p h o s m e t, 0.50 p p m ; (8) a z in p h o s -
151
m e th y l, 0 .5 0 p p m . L im e s a m p le B : (6) e th io n , 0.44 p p m .
(6 ) Conacher, H.B.S. (1990) J. Assoc. Off. Anal. Chem. 73, 332-
334
Also of note is the difference in response to the pesticides (7) Hankin, L., & Pylypiw, H.M. (1989) Pesticide Residues in
among the 3 detectors. In general, ECD is about 10 times as Produce Sold in Connecticut—1988, Bull. 873, The Con­
sensitive as ELCD for most organochlorine pesticides. ECD necticut Agricultural Experiment Station, New Haven, CT
will also respond, although not in a linear manner, to organo- (8 ) Pylypiw, H.M., Jr, & Hankin, L. (1990) Pesticide Residues in
phosphate and organosulfur pesticides. For example, if diazi­ Produce Sold in Connecticut—1989, Bull. 881, The Con­
necticut Agricultural Experiment Station, New Haven, CT
non is found in an extract analyzed by GC/FPD, it will also give
a small peak at the correct retention time for diazinon when the (9) Pylypiw, H.M., Jr, & Hankin, L. (1991) Pesticide Residues in
Produce Sold in Connecticut—1990, Bull. 8 8 6 , The Con­
extract is analyzed by GC/ECD.
necticut Agricultural Experiment Station, New Haven, CT
Pesticides can also be identified by using different GC cap­ (10) Pylypiw, H.M., Jr, & Hankin, L. (1992) Pesticide Residues in
illary columns. For example, when an SPB-20 column is used, Produce Sold in Connecticut—1991, Bull. 900, The Con­
retention times for most pesticides will be longer than with an necticut Agricultural Experiment Station, New Haven, CT
SPB-1 column; however, an SPB-608 column will only delay (11) Pylypiw, H.M., Jr, & Mattina, M.J.I. (1993) Pesticide Resi­
the retention times of earlier eluting pesticides such as dicloran dues in Produce Sold in Connecticut—1992, Bull. 908, The
and diazinon. For absolute confirmation of the identity of any Connecticut Agricultural Experiment Station, New Haven,
pesticides, mass spectrometry (GC/MS) must be used. CT
1374 SuMiTANi E t A l . : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

RESIDUES AND TRACE ELEMENTS

In d u c t iv e ly C o u p le d P la s m a A to m ic E m is s io n S p e c t r o m e tr ic

D e t e r m in a t io n o f T in in C a n n e d F o o d

H idenobu S umitani, Sachiko Suekane, and Aya Nakatani

Toyo Institute of Food Technology, 4-23-2, Minamihanayashiki, Kawanishi-shi, Hyogo 6 6 6 , Japan
K iyoaki T atsuka
Toyo Junior College of Food Technology, 4-23-2, Minamihanayashiki, Kawanishi-shi, Hyogo 6 6 6 , Japan

V arious c a n n e d fo o d s w ere d ig e ste d seq u en tially atomic absorption spectrometry (AAS) have been investigated
with HNO3 a n d HCI, diluted to 100 mL, a n d filtered, very actively (4, 5). Inductively coupled plasma atomic emis­
a n d th e n tin w a s d eterm in ed by inductively c o u ­ sion spectrometry (ICP/AES), however, offers wide dynamic
pled p lasm a ato m ic em issio n sp ec tro m e try range and relative freedom from interferences compared with
(ICP/AES). S a m p le s of c a n n e d S a tsu m a m andarin, AAS.
p each , apricot, pineapple, a p p le juice, m u sh ro o m , The objective of the present study was to establish a method
a s p a ra g u s , ev a p o rated milk, sh o rt-n e c k e d clam , for the rapid determination of tin in canned foods by ICP/AES.
sp in ach , w hole to m ato , m eat, a n d salm o n w ere Wet digestion with a combination of concentrated nitric acid
ev alu ated . S am p le p re p a ra tio n s did not require (HN03)and concentrated hydrochloric acid (HCI) was used for
tim e-co n su m in g dilutions, b e c a u s e ICP/AES h a s sample preparation.
w ide d y n am ic ran g e. T he sta n d a rd addition m ethod
w a s u se d to d eterm in e tin co n c en tratio n . A ccuracy Experim ental
of th e m eth o d w a s te s te d by analyzing analytical
s ta n d a rd s co n tain in g tin a t 2 levels (50 a n d 250 Reagents
|ig/g). T he a m o u n ts of tin fo u n d for th e 50 a n d 250
(a) Concentrated HCI (35% v/v), concentrated HN03 (61%
ug/g levels w ere 50.5 a n d 256 pg/g, respectively,
v/v), and acetone.—Analytical grade iWako Pure Chemical In­
a n d th e repeatability coefficients of variation w ere
dustries Ltd., Osaka, Japan).
4.0 a n d 3.8%, respectively. R ecovery of tin from 13
(b) Tin stock solution.—Dissolve 1.000 g tin (> 99.999%
c a n n e d fo o d s sp ik ed at 2 levels (50 a n d 250 pig/g)
pure, Wako Pure Chemical Industries Ltd.) in 50 mL concen­
ra n g ed from 93.9 to 109.4%, with a m ean of 99.2%.
trated HCI. Dilute to 1 L with 3N HCI.
T he q u an titatio n limit for tin sta n d a rd so lu tio n w as
(c) Tin metal plate.—A thickness of 0.2 mm, purity of
ab o u t 0.5 ¡ig/g. 99.99% (Rare Metallic Co. Ltd., Tokyo, Japan). All canned
foods were purchased from local markets.

B oth tin plate and lacquered cans are used for food stor­ Apparatus
age. Tin plate cans are used when the can-food interac­
(a) Inductively coupled plasma atomic emission spectrome­
tions are not significant or when the quality of the food
ter.— Model ICPS-1000HI (Shimadzu, Kyoto, Japan) inter­
is better in tin plate can. In canned foods packed in tin plate can,
faced under standard operating system.
tin dissolves into the food because of the interaction between
(b) Water purification system.—Elgastat UHQ (Elga Ltd.,
container and contents. The tin content indicates the extent of
Lane End, United Kingdom).
corrosion of the container and the acceptability of the contents.
Estimation of tin in various canned foods is important in assess­ Preparation o f A nalytical Standards
ing food quality. The Japanese regulatory limit for tin in canned
beverages is 150 |Lg/g. The Canadian regulatory limit for To check the accuracy of the method, analytical standards
were prepared as follows. An accurately weighed tin metal
canned food is 250 (ig/g. An excellent review of commonly
used methods of determination of tin in foods has been pre­ plate (50 x 260 x 0.2 mm) was dipped in 300 g fresh Satsuma
mandarin juice in a bottle. The bottle was tightly capped with
sented by Horwitz (1). The photometric determination of tin
a screw cap and placed in a constant-temperature oven at 90TT
using salitylidenamino-2 -thiophenol as a reagent (2 ), not cited
for 41.5 h. The tin plate was then removed from the juice,
in the review, has been authorized by the pharmaceutical soci­
washed with acetone and then water, dried at 60°C, and cooled
ety of Japan (3). Determinations of tin in canned foods by
to room temperature in a desiccator. The weight of the tin plate
was taken. The amount of dissolved tin (98.8 mg) was esti­
Received September 8, 1992. Accepted Febuary 23, 1993.
mated from the difference in the weight of tin before and after
 S u m it a n i E t A l .: J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1375

it was placed in mandarin juice. Two levels of analytical stand­ cept the negative concentration axis. The concentration of tin
ard (50 and 250 pg/g) were prepared from the remaining man­ in sample was calculated by using the extrapolated value, and
darin juice by quantitative dilution with fresh mandarin juice. allowing for sample weight dilution.
(c) Calculations.
Preparation o f Samples
Tin, pg/g sample = pg tin/mL x (50/25) x (100/m)
Homogenize canned food with a Waring blender. Accu­
rately weigh 40 g of homogenate or analytical standards into
300 mL Erlenmeyer flask. Dry homogenate in oven at 110°C.
where 25,50, and 100 are dilution factors and m is the sample
In a hood, add 30 mL HN0 3 to flask and heat mixture gently to weight in grams.
initiate digestion; avoid excessive frothing. When frothing has
subsided, gently boil mixture at moderate temperature for 2 h
R esu lts a n d D iscu ssio n
or until sample begins to dry at the bottom of the flask. Do not
let sample char. Add 20 mL HC1 to flask and heat mixture gen­
tly ca 15 min. Increase heat and boil 1.5 h. Remove flask from Immediately after a can is opened, the contents should be
heat. Transfer digest to 100 mL flask and then thoroughly rinse transferred to a glass vessel, because tin plated on the internal
Erlenmeyer flask twice with water. Let the solution cool to surface of the can rapidly dissolves into the contents in the pres­
room temperature and dilute to volume with water. Fat floating ence of oxygen.
on top should not be considered part of volume. Mix well and Most methods now rely on wet digestion to destroy organic
filter under reduced pressure through dry Whatman No. 1 paper matter, because dry ashing leaves the tin as insoluble oxides.
into a clean, dry bottle. The wet digestion used in this study was a modification of the
AOAC official method for tin determination by AAS (6 ). It was
Determ ination difficult to digest fat and oil in fatty samples such as meat,
salmon, and evaporated milk. In this method, the limit of fat
The instrumental conditions are summarized in Table 1.
and oil in sample is about 10-20%. We did not examine how to
(a) Preparation o f standard addition sample.—Pipet 25 mL
digest fat and oil.
digested sample solution into three 50 mL volumetric flasks,
The analytical line of sulfur (190.027 nm) is adjacent to that
and then add 0,1.0, and 2.0 mL tin stock solution (1000 pg/mL)
of tin (189.989 nm) (7). If the sample contains large amounts
into each flask and dilute to volume with water.
of sulfur compared with tin, the method is inaccurate because
(b) Standard addition method.—Three standard addition
the instrument catches the sulfur line instead of the tin line. In
samples were measured by ICP/AES after background correc­
this case, a fixed-wavelength method should be used.
tion. Background subtraction was done in the wavelength scan
In the initial stage of this investigation, we attempted to de­
profile on a display by marking 2 suitable positions on the base­
termine tin by the internal standard method using yttrium as
line of both sides of the peak. The results were plotted against
internal standard. However, we could not obtain correct results,
the added concentrations. The plot was extrapolated to inter­
because we could not match the matrix of sample solution with
that of tin standard solution. The pronounced matrix effect may
T a b le 1. In s tru m e n ta l c o n d it io n s f o r tin d e te rm in a tio n
have been due to spectral or physical interference from organic
b y I C P /A E S
materials and acids remaining in sample solution after diges­
Parameter Setting tion. Results of the standard addition method indicated that the

Rf (radio frequency) generator frequency 27.120 MHz T a b le 2. R e s u lt s o f a c c u r a c y te s t o b ta in e d b y a n a ly z in g

Vacuum spectrometer a n a ly tic a l s t a n d a r d s 3 c o n ta in in g tin a t 2 le v e ls (50 a n d
Grating 2 5 0 pg/g)_________________________________________________
Radius of curvature 1m
Level of tin in Amount of tin
Number of grooves 3600/mm
standard, pg/g found, pg/g Average, pg/g RCV, % b*250
Wavelength range 160-458 nm
Reciprocal linear dispersion 0.22 nm/mm
50 52.2 50.5 4.00
Entrance slit width 20 pm
52.2
Exit slit width 30 pm
48.1
Plasma output power 1.2 kW
51.6
Argon flow rate
48.6
Coolant 14 L/min
250 264.6 256.0 3.73
Plasma 1.2 L/min
258.9
Carrier 1.0 L/min
243.1
Purge 3.5 L/nin
249.2
Observation height in plasma 15 mm
264.4
Integration time 5s
Tin analytical line 189.989 nm a Prepared as described in text.
b RCV, repeatability coefficient of variation.
1376 S u m it a n i E t A l .: J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

T a b le 3. R e c o v e r y o f tin fro m c a n n e d f o o d s b y s ta n d a r d a d d itio n m e th o d 3

Found Recovery

Sample 49/9 RCV, % Spike level, (ig/g % RCV, %

Satsuma mandarin6 57.7 3.07

50 102.4 5.53
250 94.1 2.44
Peach6 50.0 4.78
50 100.8 3.95
250 96.4 3.16
Apricot6 87.7 3.27
50 94.1 5.70
250 99.8 2.95
Pineapple6 56.5 1.76
50 105.2 9.44
250 98.6 4.48
Apple juice6 33.4 0.91
50 102.2 8.40
250 98.2 1.77
Mushroom6 19.8 10.3
50 109.4 3.99
250 95.8 8.54
Asparagus0 81.2 3.25
50 101.4 3.56
250 97.1 4.20
Evaporated milk6 19.3 7.48
50 93.9 4.88
250 101.1 4.54
Short-necked dam® 8.9 4.59
50 95.3 8.14
250 100.0 4.44
Spinach® NDf
50 98.4 4.27
250 101.8 3.69
Whole tomato® ND
50 100.3 6.15
250 95.4 5.52
Meat® ND
50 95.3 5.04
250 94.4 8.63
Salmon® ND
50 107.1 4.32
250 102.3 1.66

a Five replicates per food. RCV, repeatability coefficient of variation. Mean recovery, 99.2%.
b Food packed in tin plate can.
c Food packed in high-tin fillet can.
d Food packed in tin plate can with inside lacquered ends.
e Food packed in inside lacquered can.
' ND, not detected.

most favorable method of determining tin in canned food is tively. Results for 13 canned foods are shown in Table 3. RCVs
ICP/AES. for unspiked samples varied from 0.91 (apple juice) to 10.3%
The accuracy of the method was tested by analyzing ana­ (mushroom). Recovery of tin from 13 canned foods spiked at
lytical standards containing tin at 2 levels (50 and 250 |lg/g). 2 levels (50 and 250 |ig/g) ranged from 93.9 to 109.4%, with a
Results are shown in Table 2. Amounts of tin found and repeat­ mean of 99.2%. Recovery at the lower spiking level (50 pg/g)
ability coefficients of variation (RCV) for the 50 and 250 (ig/g was slightly erratic compared with the higher spiking level
standards were 50.5 and 256.0 |ig/g and 4.0 and 3.8%, respec­ (250 frg/g).
 S u m it a n i E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1377

The quantitation limit was defined as the concentration at Sample preparation was easy; however, it was difficult to di­
which RCV was less than 10% (n = 5). By measuring 5 repli­ gest fat and oil. Nevertheless, it was possible to detennine tin
cates of tin standard solutions ranging from 0.1 to 3 pg/g, we in fatty foods. The standard addition method is the most pref­
found that 0.5 pg/g tin was the minimum concentration at erable method for tin determination by ICP/AES.
which RCV did not exceed 10%. The quantitation limit of the
method was estimated to be about 0.5 pg/g. R efe ren c es
Tin was not detected in foods packed in lacquered can, such
as spinach, whole tomato, meat, and salmon, except for short­
(1) Horwitz, W. (1979) J. Assoc. Off. Anal. Chem. 62, 1251-1264
necked clam. Tm concentration detected in canned short­
necked clam was 8.9 pg/g. Internal corrosion was found at the (2) Gregory, G.R.E.C., & Jeffery, P.G. (1967) Analyst 92, 293-
299
side seam of the lacquered can. Tin detected in the sample is
assumed to originate from the tin plate exposed by corrosion. (3) Standard Methods of Analysisfor Hygienic Chemists With
The large dynamic range of ICP/AES, compared with AAS, Commentary (1990), authorized by the Phannaceutical Soci­
generally permits calibration over a very wide concentration ety of Japan, Kanahara Publishing Co., Ltd., Tokyo, pp.
58-60, 574-577
range. The time-consuming dilutions are, therefore, not re­
quired. (4) Dabeka, R.W., & McKenzie, A.D. (1981) J. Assoc. Off. Anal.
Chem. 64, 1297-1300
C o n clu sio n (5) Dabeka, R.W., McKenzie, A.D., & Albert, R.H. (1985) J. As­
soc. Off. Anal. Chem. 6 8 , 209-213
The purpose of the study was to establish a rapid and accu­ (6 ) Official Methods of Analysis (1990) 15th Ed., AOAC, Ar­
rate method of quantifying tin in canned foods by ICP/AES. lington, VA, sec. 985.16
Results from wet digestion by H N 0 3 and HC1 followed by (7) Hayakawa, T., Kikui, F., & Ikeda, S. (1982) Spectrochem.
ICP/AES were obtained rapidly and were moderately accurate. Acta 37B, 1069-1073
1378 T h o m p s o n : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

RESIDUES AND TRACE ELEMENTS

D e t e r m in a t io n o f U lt r a t r a c e L e v e ls o f L e a d in In f a n t F o r m u la b y

Is o t o p e D ilu t io n In d u c t iv e ly C o u p le d P la s m a M a s s S p e c tr o m e tr y

J o seph J. T h o m pso n

Abbott, Ross Products Division, Analytical Research and Development, 625 Cleveland Ave, Columbus, OH 43215

A sim p le m eth o d w a s d ev e lo p e d for th e a c c u ra te levels in infant formula are likely to be compared on this basis.
a n d p re c ise d eterm in atio n of low- a n d su b -p p b Although current methods based on GFAAS can be modified
(ng/g) c o n c e n tra tio n s of lead in infant form ula by to have a quantitation limit of less than 5 ppb (ng/g) lead, the
iso to p e dilution inductively c o u p le d p la sm a m a s s moderate precision of the results and the lack of suitable stand­
sp e c tro m e try u sin g u ltraso n ic nebulization. After ard reference materials (SRMs) certified for lead at this level
ad d ition of a know n am o u n t of 207Pb, sa m p le s w ere make it desirable to ascertain accuracy by comparison with re­
m icrow ave d ig e ste d a n d th e r a tio 201 Pb/20®Pb w as sults of an independent method, preferably one with greater
m e a su re d in th e d ig e sts. A g reem en t w ith certified precision.
v alu e s for lead in milk po w d er sta n d a rd referen ce For several years, isotope dilution inductively coupled
m aterials w a s g o od , a n d iso to p e dilution an a ly sis plasma mass spectrometry (ID/ICP/MS) has been used for the
u sin g ^ P b y ielded identical re su lts for th e sta n d ­ accurate and precise determination of elemental concentrations
ard re fere n ce m aterials. Lead c o n c e n tra tio n s d e te r­ in the ppm to upper ppb range (7-10). However, only a few
m ined for sev eral infant nutritional p ro d u c ts w ere reports deal with lead determinations in foods at low-ppb levels
verified by a n in d e p e n d e n t m ethod. Typically, rela­ by ID/ICP/MS (11-12) or evaluate the accuracy of results at
tive sta n d a rd d ev iatio n s of <4% w ere o b ta in e d with this level. The problem is complicated by a lack of suitable
th is m eth o d for lead c o n c e n tra tio n s a b o v e 2 ppb. SRMs and by contamination. Despite the impressive precision
T he reco v ery of 2 ng of lead from an a q u e o u s of isotope dilution methods, the key issue in this application is
sta n d a rd carried th ro u g h th e m icrow ave d ig estio n accuracy in the 1-5 ppb range typical of lead concentrations in
w a s 104 + 4%. Infant form ula (containing 0.6 p p b infant formula. In this paper, an ID/ICP/MS method is de­
lead) to w hich 0.4 ng of n atu ra l-ab u n d an c e lead scribed for the determination of low- and sub-ppb concentra­
h ad b e e n a d d e d , to sim u la te a form ula containin g tions of lead in infant formula, with special attention given to
0.9 p p b lead, w a s an alyzed by iso to p e dilution, an d ensuring accuracy. This approach is intended to serve as a ref­
th e re su lt w a s 96 ± 18% of th e th eo retical value. erence method forjudging the accuracy of auxiliary methods
T h u s, d ifferen ces of 0.3 p p b lead could b e clearly developed for quality control and for resolving disparities be­
d istin g u ish ed , a n d th e d etec tio n limit w a s e sti­ tween lead concentrations measured by different laboratories.
m ated to b e 0.1 ng lead p er gram of infant form ula.
T he k ey s to a c c u ra c y for th is m eth o d a re m inim iz­ Experim ental
ing co n tam in atio n a n d accu rately determ ining th e
co n c en tratio n of lead in th e isotopically e n ric h ed Reagents and Standards
sta n d a rd .
Standard reference materials (nonfat milk powder 1549,
trace elements in water 1643c, and isotopic SRM 981) were
obtained from the National Institute of Standards and Technol­
R ecently, there has been increased concern over the ex­
ogy (NIST, Gaithersburg, MD); another milk powder reference
posure of infants to ultratrace levels of lead (1-3). In
material, BCR 63, was obtained from the Community Bureau
any efforts to assess the risk to infants, background lev­
of Reference (Brussels, Belgium). Enriched isotopic materials
els of lead in infant formulas must be measured accurately at
(2 °6 pb, 9 9 9 9 %^ anci 2 0 7 Pb, 91.62%, both supplied as lead car­
very low concentrations. Methods involving differential pulse
anodic stripping voltammetry (4) and graphite furnace atomic bonate) were obtained from Oak Ridge National Laboratories
(Oak Ridge, TN). The 207Pb material was stored overnight in a
absorption spectrophotometry (GFAAS) (5-6) have been de­
vacuum desiccator before weighing on a microgram balance.
veloped to measure ultratrace concentrations of lead in foods.
The 206Pb material was weighed without drying. Stock solu­
Because most laboratories are equipped with GFAAS instru­
tions were prepared and serially diluted to reach working con­
mentation, nutrition research and routine measurements of lead
centrations of 30.0 and 38.2 ppb total lead for the 206Pb and
207Pb materials, respectively. Other solutions prepared were
Received Novmber 1 6 ,1 9 9 2 . Accepted January 30, 1993.
2 0 0 , 2 0 .0 , and 0 . 8 ppb of natural-abundance lead (prepared
 T h o m p s o n : J o u r n a l O f AO AC In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1379

from NIST 1000 mg/L lead stock solution) and (optionally) cap were rinsed with water, and, optionally, 1.00 mL of a 100
100 ppb of Th or Bi internal standard. All standards were care­ ppb Th or Bi internal standard was added. The digests were
fully prepared in 2 or 3% nitric acid with calibrated, acid- diluted to a reference mark on each vessel (volume ~ 70 mL)
leached glass volumetric ware and Millipore water. Solutions and transported directly to the ICP/MS for measurement of the
were stored in acid-leached poly(tetrafluoroethylene) (PTFE) 2 0 7 Pb/208Pb ratio. The entire procedure, including data reduc­

or poly(methylpentene) bottles. High-purity nitric acid and hy­ tion, took about 5 h. This time probably could be reduced to 2
drogen peroxide (Ultrex, J.T. Baker, Inc.) were used. to 3 h with newer microwave technology.
Because of uncertainty about the moisture content of the For a smaller number of samples, 206Pb was added as the
original isotopic materials and error propagation from the dilu­ spike isotope and the ratio 2 0 6 Pb/208Pb in the digests was meas­
tions, the concentration of lead in the working standards, ured.
[Pb]spike, was checked by reverse isotope dilution. A known
amount of spike solution was mixed with a known amount of Instrum entation
the 20 ppb natural-abundance lead standard and the 2 0 7 Pb/208Pb The ICP/MS instrument (PQ II Plus, Fisons) was operated
of the mixture ratio was determined. Both lead standards were with an ICP forward power of 1.4 kW and typical ion lens set­
remade from their parent standards every month whenever tings (dial readings) as follows: extractor, 100; collector, 470;
their concentrations changed by more than 5%. Initially, LI, 800; L2, 440; L3, 580; and L4, 490. The pulse-counting
fPb]spike must be measured for various mixtures differing in the multiplier was set to a dial reading of 500-550, and the resolu­
ratio of spike to 20 ppb standard, as described in Discussion. tion was lowered somewhat from factory settings to maximize
From experience, a best value for [Pb]spike was chosen that sensitivity. The quadrupole was operated in a peak-jumping
agreed closely with the nominal concentration calculated from mode. The ion intensity was measured repetitively for 160 ps
the mass and volume used in preparation of the spike solution. (dwell time) at each of 3 points per peak until a total of 10 was
In addition, an accurate determination of lead in NIST 1643c spent on all of the selected isotopes (2 0 7 Pb, 2 0 9 Bi, 2 0 6 Pb, 2 0 8 Pb,
could be made by using the best value for [Pb]spike. If internal 2 3 2 Th). Ten integrated signal counts were acquired and aver­
consistency was not obtained with these 3 solutions, all master aged for each sample acquisition. The ICP/MS was equipped
standards were remade, and a fresh bottle of the SRM was ob­ with an ultrasonic nebulizer (Cetac Technologies, Inc., Model
tained. The concentration of lead in the spike typically was U-5000) operated with an aerosol flow rate of 1.3 L/min.
quite stable over a month or more, and subsequent determina­
tions of [Pb]spike could be made rapidly, without checking Data Handling
against NIST 1643c.
A commercial spreadsheet program facilitated the examina­
Preparation o f Sam ples tion and visualization of the effects of isotope dilution parame­
ters on method accuracy. Data reduction was performed in the
Twelve PTFE microwave vessels were cleaned by heating following manner. Because the natural abundance of the lead
(with vessel caps on and tightly sealed), first with Ultrex nitric isotopes varies, the spreadsheet calculated the abundances of
acid for 1 0 min and then with fresh nitric acid and hydrogen the 4 lead isotopes from the 2 0 4 Pb/2 0 8 Pb, 2 0 6 Pb/2 0 8 Pb, and
peroxide for another 10 min. Then, 0.20 g powder, 0.6 g con­ 2 0 7 Pb/2°Spb ratios measured for the 0.8 ppb standard. All 3 ra­

centrated liquid, or 1.2 g ready-to-feed (RTF) product was tios were determined at the start of each day and did not vary
weighed into 8 vessels, followed by the addition of a known by more than 2 to 3% over the course of these experiments (a
amount of 207Pb (spike isotope). The spike/sample mass ratio few months). In addition, the ratios from NIST 981 were moni­
was adjusted to yield a 2 0 7 Pb/208Pb ratio in the range of 2-5. tored every few weeks to see if there was inordinate mass dis­
Thus, 100 pL of the 38 ppb spike (~4 ng) was added to the RTF crimination in the lead isotope measurements and to check for
products and 500 pL of spike solution was added to the NIST the presence of spectral interferences. Certified lead isotope
1549 powder. Vessels were always weighed to verify the accu­ ratios in NIST 981 and in the 2 0 7 Pb-enriched standard always
racy of the volumes delivered by micropipets. The 4 vessels agreed with experimental values within 1 to 3%. The natural-
used for blanks were treated differently depending on the abundance ratios in samples and milkpowder SRMs were
method of blank subtraction. For blank subtraction by isotope checked initially and found to be essentially equal to the ratios
dilution (ID) 100 pL of the 38 ppb spike standard was added to obtained from the 0.8 ppb standard. Typical natural ratios were
each blank vessel; otherwise, nothing was added to the vessels as follows: 2 0 4 Pb/208Pb = 0.0261, 2 0 6 Pb/208Pb = 0.513, and
prior to nitric acid. 2 0 7 Pb/208Pb = 0.410 (abundances: 2 0 4 Pb, 1.3%; 2 0 6 Pb, 26.3;

RTF products first were dried by heating in a 650 W micro- 2 0 7 Pb, 21.0%; and 2 0 8 Pb, 51.3%). Thereafter, samples were as­

wave oven (Model 81D, CEM Corp.) at 45% power for 4 min sumed to have the same natural abundances as the standard.
in open vessels. Then, 5 mLconcentrated nitric acid was added, From each ran of 12 vessels, the blank count rates from the
the vessels were sealed, and the samples were heated for 15 min 4 unspiked vessels were normalized to the samples by using the

at 80% power to a pressure of 80 psi. After manual venting, the internal standard, averaged, and subtracted from the count rate
vessels were heated for 2 0 min at 80% power to a pressure of of the sample at the mass 207 and mass 208 channels. A cor­
25 psi. Next, 2 mU hydrogen peroxide was added, followed by rected ratio was then calculated by dividing the net mass 207
heating for 15 min at 80% power. After cooling to room tem­ counts per second (cps) by the net mass 208 cps. This ratio was
perature, the vessels were opened, the walls and inside of the inserted into the normal isotope dilution equation to calculate
1380 T h o m p s o n : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

the lead concentration. Alternatively, the quantity of Pb in the pression due to the matrix was observed, usually 30-50%, re­
blank could be determined by isotope dilution. gardless of the microwave digestion procedure. For products,
isotope ratios changed significantly upon dilution of a digested
Recovery Studies sample. As noted previously, high concentrations of concomi­
Lead loss from microwave digestion was measured as fol­ tant elements can change isotope ratios significantly (13), and
lows: 100 pL of the 20 ppb natural-abundance lead standard (2 so sample dilutions were optimized individually, by choosing
ng) was placed into each of 6 cleaned, PTFE microwave ves­ a sample weight below which the ID result did not change.
sels (A-F) containing 5 mL nitric acid. Six other vessels (G-L), Some ID/ICP/MS results for lead in the 2 milk powder
containing only nitric acid, were put in the oven. After the di­ SRMs are shown in Table 1. Although the NIST powder is cer­
gestion procedure, the same amount of lead was placed into 3 tified for a lead concentration of 19 ppb, it really represents the
empty vessels (G-I). The other 3 vessels served as blanks (J- lower levels of lead found in RTF infant formula, because com­
L). After all the vessels were filled with water, the blank-sub­ ponents in a powder are concentrated 6 to 7 times relative to an
tracted counts per second at 208Pb for vessels A -F were com­ RTF liquid. The excellent within-run precision for the SRMs
pared with the average blank-subtracted counts for vessels G-I. indicates that they were homogeneous at the 0 . 2 0 g sample size
The isotope dilution recovery was determined for samples taken (NIST certifies homogeneity at a 0.5 g level). Day-to-day
spiked with natural-abundance lead. Samples of BCR 63 (0.35 repeatability also was excellent. No difference was observed in
g) were weighed into 3 vessels (A-C) along with 1 mL of the the results whether 206Pb or 207Pb was used as the spike isotope.
38 ppb 207Pb spike standard. Six other vessels (D—I) contained Lead concentrations determined by this method generally
sample, 2 mL of spike, and 200 jiLof a 200 ppb natural-abun­ agreed with the certified concentrations for lead in the SRMs,
dance lead standard. Three vessels (J-L) served as blanks. The which is indicative of the accuracy of the method. The
usual microwave digestion was performed. From vessels A-C, ID/ICP/MS results were slightly below the certified mean val­
the average lead concentration (ng/g) in BCR 63 was deter­ ues, even though the values in Table 1 were corrected for mois­
mined by ID. With this value and the individual sample ture in the powders (~6 % in the BCR materials and ~1% in
weights, an expected lead concentration and recovery could be NIST 1549). A typical value obtained for lead in NIST 1643c
calculated (also by ID) for the adulterated samples in vessels (water) was 34.5 ±0.3 ppb, also below the mean certified value
D-I. This experiment was repeated twice, first for 3 g of infant of 35.3 ± 0.9 ppb. The reason for this apparent bias is unknown,
formula plus 5 ng of additional lead and then for 1.2 g of the but the difficulty in accurately determining the lead concentra­
same product spiked with 0.4 ng of natural-abundance lead. tion of the spike solution could play a role. Small positive or
negative biases of ID/ICP/MS results have been noted pre­
R e su lts a n d D iscu ssio n viously (13). However, a ~5% bias at the 5 ppb level would
certainly be deemed inconsequential for the purpose of this
The method is feasible only with the sensitivity afforded by method.
the ultrasonic nebulizer. For example, after digesting 0.20 g of Lead concentrations determined by the ID/ICP/MS method
NIST 1549 (19 ng Pb/g) and diluting to 70 mL, the concentra­ in milk and soy-based infant formulas from different manufac­
tion of the solution taken to the ICP/MS was about 50 ppt turers are shown in Table 2. All formulas tested contained less
(pg/g). With conventional pneumatic nebulization, the limit of than 2 ppb lead on an as-fed basis. The soy-based formulas
detection was 30 ppt, and the isotope ratio 2 0 7 Pb/208Pb could be contained a higher concentration of lead than the milk-based
determined with a precision of only 5% RSD on the 0.8 ppb formulas. Otherwise, the concentration of lead did not differ
standard. Digested RTF samples also had lead concentrations greatly among manufacturers or among the 3 forms of a prod­
that were too close to conventional detection limits with the uct (powder, concentrated liquid, or RTF). From the data in
instrument. With ultrasonic nebulization the sensitivity was Tables 1 and 2, one can see that the RSDs obtained with the
typically 40 000 cps/ppb lead, with a detection limit below 1 method for 3 to 4 replicates are generally 10% for samples with
ppt (pg/g). Ratios were routinely measured with 0.5% RSD for 0.3 to 2 ppb lead and <4% for samples with >2 ppb lead.
the 0.8 ppb standard, and were typically 1-3% for digested Lead concentrations determined in 6 Ross pediatric nutri­
samples containing about 10 times less lead. Some signal sup­ tional products by the ID/ICP/MS method were compared with

T a b le 1. L e a d c o n c e n tr a tio n s (p p b ) d e te rm in e d f o r m ilk p o w d e r S R M s b y ID /ICP/M S

NIST 1549 207Pb NIST 1549 ^ P b BCR 63 ^ P b

Day Mean ± SD RSD, % N Mean ± SD RSD, % N Mean ± SD RSD, % N

1 17.0 + 0.7 4 6 17.4 + 0.4 2 4 100 + 2 2 3

2 17.3 + 0.7 4 4 16.5 + 0.2 1 4 98 + 2 2 4
3 17.0 + 0.8 4 3 17.7 + 1.1 6 4 100 + 1 1 4
Total3 17.1 ±0.7 4 13 17.1 ±0.8 5 12 99 + 2 2 11

a Pooled data. Certified lead concentration: 19 ± 3 ppb for NIST 1549 and 104 ± 3 ppb for BCR 63.
 T h o m p s o n : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1381

Table 2. Lead concentrations (ppb) in infant formulas niques were briefly investigated, but each technique had limi­
from different manufacturers, determined by ID/lCP/MSa tations for measurements at the 1 ppb level. Possible problems
Powder CLb RTFC with these methods of sample preparation include selective
volatilization of lead species from the spike/sample mixture,
Milk-based formula
variable blank contamination or adsorption from quartz or
0.24 ± 0.01 0.48 ± 0.08 0.61 ± 0.06 platinum surfaces, or matrix effects from undigested material.
0.90 ± 0.06 0.24 ± 0.06 0.42 ± 0.05 With microwave digestion in PTFE vessels, the blanks were
0.92 ± 0.01 0.30 ± 0.03 1.1 ±0.1 very low (<0.005 ppb or <0.3 ng total lead) and reproducible
(within-run precision typically 10-15% RSD) provided that
Soy-based formula the vessels were cleaned as described previously. It could be
argued that the isotopes have an excellent chance of equilibrat­
1.81 ±0.08 1.90 ± 0.07 1.78 ±0.03
ing in the closed, high-temperature environment before any se­
1.20 ±0.02 1.06 ±0.07 1.48 ±0.05
_d _d lective isotope losses can occur. This hypothesis was tested by
1.2 ±0.2
performing the spike recovery studies (Table 4, and described
3 Each entry is the mean ± SD of 3 or 4 determinations of a single in Experimental), which are commonly used to evaluate the
product expressed on an as-fed (RTF) basis: powder + 6.8 = RTF;
CL + 2 = RTF. Entries are randomized so that values from a single accuracy of methods that rely on external calibration. The re­
row do not represent one manufacturer’s line of products. sult of the first experiment listed in Table 4 indicates that the
b Concentrated liquid. recovery of lead after microwave digestion in a closed vessel
c Ready-to-feed.
d Not determined. is nearly complete. This result is significant, because experi­
ence in developing the GFAAS method has shown that, at low
ppb levels, lead from an aqueous standard is lost preferentially
to lead from the sample when the sample is heated in an open
those from a newly developed method with GFAAS. The re­
container. Therefore, open heating of a sample with an aqueous
sults are shown on Table 3. Given the very low levels of lead
isotopic spike standard may result in loss o f the spike isotope
in the product, the agreement between the 2 independent meth­
ods is remarkable and is a further indication of accuracy. Also, before isotopic equilibration is achieved and in inaccurate iso­
the sample preparation for the GFAAS method was substan­ tope dilution analysis. As further proof that isotopic equilibra­
tially different:dry ashing of a 20 g sample of RTF (17 times tion occurred with the present method, equilibrating the sam­
more than for ICP/MS) followed by extraction with sodium ple, acid, and enriched isotope solution for longer periods ( 2 h)
diethyldithiocarbamate in diisobutylketone and back-extrac­ before microwave heating did not significantly change the ID
tion into an aqueous layer containing palladium chloride. Prob­ result. Furthermore, when natural-abundance lead was added
ably because of this extensive sample handling, the precision to an SRM or an infant formula, the ID result agreed very
of the GFAAS method was generally inferior to that of the closely with the expected result (Table 4, last 3 experiments).
ID/ICP/MS method. The precisions also indicate that the Lead concentrations from a 0.4 ng spike added to product A
ID/ICP/MS method has a lower quantitation limit than the (Table 3) could be clearly differentiated from that from the un­
GFAAS method. adulterated product. Product A contains about 0.6 ppb lead, and
The mode of sample preparation can affect accuracy. Open- the addition o f 0.4 ng lead simulated a product containing about
vessel wet digestion, dry ashing, and simple dilution tech­ 0.9 ppb lead, a difference of 0.3 ppb. Therefore, the detection
limit by ID is estimated to be 0.1 ng/g in product.

Table 3. Precision, number of trials ( N ), and relative agreement between lead concentrations determined by GFAAS
and ID/ICP/MS in various infant formulas and in NIST 1549
GFAAS ID/ICP/MS

Product, type3 RSD, % N (days) RSD, % N (days) Ratio6

NIST 1549 15.0 83 (29) 3.9 13(3) 1.05

A, RTFC 50.0 42 (12) 26.0 6(2) 0.80
A, RTFC 22.0 33(7) 9.5 7(2) 0.82
B, RTF 25.0 17(6) 5.6 7(2) 0.87
C, RTF 18.0 6(2) 8.1 5(1) 1.03
D, RTF 6.5 4(1) 0.7 3(1) 0.95
E, CL 6.4 6(2) 2.6 5(1) 1.05
F, powder 9.4 6(2) 4.3 4(1) 0.99

a Abbreviations: RTF= ready-to-feed; CL= concentrated liquid.

b Mean GFAAS result/mean ICP/MS result. Specific lead concentrations for these products are not listed but are all < 5 ppb.
c Different batches.
1382 T h o m p s o n : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

Table 4. Recoveries of lead and lead spikes added to samples

Mean recovery
Experiment +SD, %
N Spike level, ng

Total lead from aqeous standard 104 + 4 6 2

ID recovery from BCR 63 101 ±1 6 40
ID recovery from product A (Table 3) 102 + 4 5 5
ID recovery from product A (Table 3) 96 + 18 4 0.4

The relatively simple procedures used in this method re­ signal suppression from the matrix) count rates of the blank
duced the chance for sample contamination or high blanks. All were subtracted from the lead count rates for the samples. The
dilutions, weighings, and digestions were performed in a Class net count rates were then used to calculate a corrected
1 0 0 clean room, and a set of glassware and plasticware was set 207Pb/208Pb ratio. The choice of Th as internal standard was

aside exclusively for this project. Despite all precautions, how­ arbitrary; subsequent experiments in which Bi was substituted
ever, contamination of blanks and samples was relatively com­ for Th proved that the choice of internal standard did not affect
mon when working at this level; typically, one blank or sample method accuracy. Eventually, the z07Pb -enriched standard
was rejected out of each mn of 12 vessels. For the data col­ (which was 6 % 208Pb) was added to the 4 blank vessels before
lected for Tables 1 and 2, for instance, the data for 9 of 61 digestion, and the blank was determined by ID. The results
blanks and 8 of 106 samples had to be discarded. showed that both methods of blank subtraction were equiva­
In most ID/ICP/MS applications, the blank is determined by lent: they yielded the same average result for lead in NIST 1549
ID. However, in the present method the total amount of 208Pb and exactly the same value for lead in the blank (0.26 ng).
in the spiked blank is extremely small and whether isotope ra­ Therefore, when the spike solution contained a significant
tios could be measured with sufficient accuracy in these solu­ amount of the reference isotope, the consistency of the lead
tions was doubtful. At first, normalized (to account for lead levels found in the blank was more important than the method

Pb-2 0 6
.Q
Q.

(/)
<0 0 .8 1.1 1 .4 1 .7 2 .0 2 .3 2 .8 3 .1 3 .5 3 .7 4 .2 5 .3 6 .3 7 .6 9 .8 1 4 26
<D
^ 2 0 6 /2 0 8 R atio in M ixture
Figure 1. Total lead concentration of spike determined for various mixtures of natural-abundance standard and 206Pb
spike standard. The concentration derived from careful serial dilution of the original isotopic material was 30.0 ppb
(horizontal line). The precision of individual measurements at a given ratio is about 0.3%, indicated by the error bars
on top of each bar.
 T h o m p s o n : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1383

Pb-2 0 7
-Q
CL

CO
5 1 .6 1 .8 2 .2 2 .7 2 .9 3 .1 3 .4 3 .7 4 .5 5 .0 5 .7 6 .6 8 .3
<D
s 2 0 7 /2 0 8 R atio in M ixture
Figure 2. Total lead concentration of spike determined for various mixtures of natural-abundance standard and 207Pb
spike standard. The concentration derived from careful serial dilution of the original isotopic material was 38.2 ppb
(horizontal line). The precision of individual measurements at a given ratio is about 0.3%, indicated by the error bars
on top of each bar.

of blank subtraction. However, blank subtraction using 206Pb The proportions of the 2 solutions (spike and natural stand­
ID (0.05% 208Pb) was not attempted, because the small amount ards) determines the 207Pb/208Pb or 206Pb/208Pb ratio in the re­
of reference isotope in the mixture would likely cause large sultant mixture (Rm). For mass spectrometric measurement, the
uncertainty in the isotope ratio. best precision and accuracy is obtained if Rm is near unity.
A final, rather insidious source of error in these experiments However, for the ID method as a whole, error propagation is
appeared in the determination of the total lead concentration of often minimized by adjusting the spike/sample ratio to obtain
the spike by reverse ID. The calculated value for [Pb]spike de­ Rm substantially different from unity (14). The optimal Rm can
pended on the aliquots of spike and natural-abundance stand­ be predicted from theory, but in practice, it is the ratio that gives
ards taken (Figures 1 and 2), a phenomenon that has been noted the best value for the concentration of lead in the spike stand­
previously in work with low levels of lead by ICP/MS (11). A ard, as discussed in Experimental. Luckily, there is a fairly
range of as much as 2 ppb in the 207Pb spike concentration and broad range of spike/sample ratios (Rms) that yields an accurate
> 4 ppb in the 206Pb spike concentration was observed, whereas [Pb]spike value; for example, when Rmis 2 - 6 , the concentrations
repeated determinations using the same aliquots showed differ­ of lead determined in the 206Pb solution are accurate and not
ences of no more than 0.5 ppb in the spike concentrations. The significantly different from each other. However, some of the
horizontal line through the bars in Figures 1 and 2 indicates the small apparent bias noted earlier in the ID/ICP/MS results may
nominal concentration of the working isotopic spike solutions, be due to these errors in the determination of [Pb]spike.
obtained from a knowledge o f the mass and volumes used in its
preparation. The choice of whether to use this nominal concen­ Conclusion
tration or one of the other measurements of [Pb]spike shown in
the figures for ED calculations represents a dilemma for the A relatively rapid and rugged method has been devised for
analy st. Any error in the spike concentration would translate the accurate and precise determination of lead in infant for­
into a similar relative error in the lead concentration deter­ mula. Accuracy was confirmed by comparison with certified
mined for a sample. values in reference materials, by spike recoveries, and by com­
1384 T h o m p s o n : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

parison with concentrations determined by an independent (4) Capar, S.G., G ajan, R.J., Subjoc, C .A ., & Sanders, M . (1982)
method. The method was reliable during several months of J. Assoc. Off. Anal. Chem. 65, 9 7 0 -9 7 7
testing of over 2 0 0 individual samples of pediatric nutritional (5) D abeka, R.W ., & M cK enzie, A .D . (1986) C a n . J. S p e c tr o s c .
products. The keys to accuracy for this method are controlling 3 1 ,4 4 -5 2
the concentration and variability of lead in the blank and accu­ (6) D abeka, R.W. & M cK enzie, A .D . (1988) Food Addit. Con-
rately determining the concentration of lead in the isotopically tam. 5, 3 3 3 -3 4 2
enriched standard.
(7) G arbino, J.R. & Taylor, H .E. (1987) Anal. Chem. 5 9 ,1 5 6 8 —
1575
Acknowledgment
(8) B eauchem in, D ., M cL aren, J.W ., M ykytiuk, A.P., & B erm an,
5.5 . (1987) Anal. Chem. 59, 7 7 8 -7 8 3
Andre Szabo and Donald Sgontz, Jr., developed the GFAAS
(9) B eauchem in, D., M cL aren, J.W ., & Berm an, S.S. (1988) J.
method for lead and graciously supplied the data for the Anal. At. Spectrom. 3, 7 7 5 -7 8 0
method comparison studies.
(10) B eauchem in, D., M cL aren J.W ., W illie, S.N., & B erm an,
5.5 . (1988) Anal. Chem. 60, 687 -6 9 1
References
(11) V iczian, M ., L asztity, A., W ang, X ., & B arnes, R. M . (1990)
J. Anal. At. Spectrom. 5, 125-133
(1) N eedlem an, H.L. (1984) in A d v a n c e s in C lin ic a l C h ild P s y ­
c h o lo g y , B. L ahey & A. K azdin (Eds), N ew York, NY, pp. (12) D ean, J. R., E bdon, L., & M assey, R. (1987) J. Anal. At.
195-220 Spectrom. 2, 3 6 9 -3 7 4
(2) Ryu, RE., Ziegler, E.E., N elson, S.E., & Fom on, S.J. (1985) (13) H ouk, R.S., & T hom pson, J.J. (1938) Mass Spectrom. Rev. 7,
in D ie ta r y a n d E n v ir o n m e n ta l L e a d : H u m a n H e a l t h E ffe c ts , 425^461
K.R. M ahaffey (Ed), Elsevier, A m sterdam , pp. 187-209 (14) D e B ievre, P.J., & D ebus, G.H. (1965) Nucl. Instr. Meth. 32,
(3) Shannon, M . (1989) C lin . P e d ia tr . 28, 3 8 0 -3 8 2 2 2 4 -2 2 8
 A k a s a k a E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1385

TECHNICAL COMMUNICATIONS

F lu o r o m e t r ic D e t e r m in a t io n o f T o t a l a n d B o u n d S u lf it e in W in e

b y A f - ( 9 - A c r id in y l) m a le im id e

K azuaki A kasaka, H iroshi O hrui, and H iroshi M eguro

Tohoku University, Faculty of Agriculture, 1-1 Tsutsumidori-Amamiyamachi, Aoba, Sendai 981, Japan
T ateo S uzuki
National Food Research Institute, 2-1-2 Kannondai, Tsukuba 305, Japan
H arufumi M iwa
Ajinomoto Technical Center, 1-1 Suzukicho, Kawasaki, Kawasaki 210, Japan

A/-(9-Acridinyl)maleimide (NAM) reacts with sulfite w h ite w in e (1 1 , 12). Ion e x c lu s io n ch rom atograp h y (1 3 ) a lso
in wine and gives strong fluorescent derivatives m ad e it p o ss ib le to d eterm in e both total and free su lfite in w in e.
that lead to highly sensitive fluorometry of both to­ R ecen tly, w e reported a h ig h ly se n sitiv e flu orom etric
tal and bound sulfite in wine. Values of free and m eth od for total su lfite in w in e that u se s A -(9 -a crid in y l)m a le-
bound sulfite in wine determined by the NAM im id e (1 4 ). In the study reported here, w e ap p lied the

method and the modified Rankine method agreed. fluorom etric m eth od to th e d eterm in ation o f both b ou n d and
total su lfite in sim p le p rocedures.
Sulfite was determined in <200 pL wine within 2 h.

METHOD
A relatively large amount of sulfite is added to wine as a
Reagents
food additive to control fermentation. Sulfite is used
because of its sterilizing and antibacterial properties. It U s e a n alytical or superpure grade c h e m ic a ls and d istilled
is also useful as an antioxidant and for ripening (1). In wine, water.
sulfite exists in 3 chemical forms that can be distinguished by (a) Ethylenediaminetetmacetic acid, disodium salt
analytical methods (2 ): free, reversibly bound (called bound (Na2EDTA), 0.02M.— D is s o lv e 7 .4 g N a 2E D T A in 1 L water.
sulfite hereafter), and irreversibly bound. The bound sulfites (b) N-(9-Acridinyl)maleimide (NAM), 0.9ImM.— D is ­
are chemically bonded with organic compounds such as alde­ s o lv e 1 m g N A M in 4 m L a ceto n e b efo re use.
hydes and a-keto acids (3), which are released as free sulfite (c) Reaction buffer, p H 9.2.— M ix 6 .1 8 g b oric acid, 7 .4 6 g
under alkaline conditions. The current legislation requires de­ p o ta ssiu m ch lorid e, and 7 .4 g N a 2E D T A in 1 L w ater (so lu tio n
termination of the free and the reversibly bound sulfite (2 ). I). M ix 10.6 g so d iu m carbonate and 7 .4 g N a 2E D T A in 1 L
The method of Rankine (4) and modifications of it have w ater (so lu tio n II). A d ju st p H to 9 .2 b y ad d in g so lu tio n II to
been widely used for the determination of sulfite in wine and so lu tio n I.

other foods (5). In these methods, sulfite is distilled under (d) Phosphoric acid solution, 25% . — A d d 2 4 0 m L w ater to
strong acidic conditions with bubbling nitrogen gas. The dis­ 100 m L p h osp h oric acid (85% so lu tio n ).

tilled sulfur dioxide is trapped and oxidized to sulfuric acid in (e) Sodium carbonate solution, 0.3M.— M ix 1 5.9 g sod iu m
carbonate and 4.1 g E D T A trisod iu m salt in 0 .5 L water.
hydrogen peroxide solution, and the amount of the sulfuric acid
is determined by titration. To improve the sensitivity, Ogawa et Apparatus
al. ( 6 ) modified the titration methods to use colorimetry. The
(a) Spectrofluorometer.— M o d e l F P -5 5 0 A (Japan S p ectro­
sulfite distilled at O^C in an ice bath is defined as free sulfite;
sc o p ic C o., Ltd., T ok yo, Japan) eq u ip p ed w ith F P -2 0 5 0 q u ick
the sulfite obtained by heating is defined as total sulfite, which
flo w sam pler; F L U O R O R E A D M o d e l 2 0 0 (A jin o m o to C o.,
is the sum of free and bound sulfites. However, these methods
Inc., T o k y o , Japan).
are not satisfactory with regard to simplicity or rapidity. In ad­
(b) Isothermal incubator.— M od el T C -1 (T o y o S ci. Ind.
dition, no special apparatus is required.
C o., T o k y o , Japan).
An enzymatic method was used to determine only total sul­
(c) p H meter.— M o d el F -1 3 (H orib a Instrum ents, Inc.,
fite (7, 8 ). A pulsed polarographic method (9) and a modified
K y o to , Japan).
Monier-Williams method (10) were reported. Flow injection
analysis was used to determine both total and free sulfite in Sulfite Standard Preparation

D is s o lv e 0 .4 g so d iu m b isu lfite (N a 2S 20 5) in 100 m L d e­

R e c e iv e d O c to b e r 6 . 1 9 9 2 . A c c e p te d J a n u a ry 18, 1 9 9 3 . g a sse d 0 .0 2 M N a 2E D T A . P ip et e x a c tly 1.0 m L o f th e solu tion
1386 A k a s a k a E t A l .: J o u r n a l O f AOAC In t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

and 1.5 mL 0.1N iodine solution to 10 mL water. Add 2 mL

hydrochloric acid to the solution, and titrate with 0.01N sodium
thiosulfate, using starch solution as an indicator (x mL). Run
the blank test at the same time (y mL). Concentration of the
standard solution is [F(x -y)/200]mM, where F = titer of 0.01N
thiosulfate solution. Store this stock solution in the refrigerator
at 5°C, and use within 3 days of preparation. Dilute the stock
solution to 1:2000-l:100 with 0.02M Na2EDTA before use.
S a m p le P re p a ra tio n

(a) Total sulfite .—Mix 20 |iL wine and 180 (iL reaction
buffer (pH 9.2). Use as the sample.
(b) Bound sulfite .—To a 75 x 12 mm id test tube, add 20
(J.Lwine and 10 |J,L25 % phosphoric acid solution with cooling
in an ice bath. Mix vigorously. Let stand for at least 40 min in
the ice bath and then mix with 170 \iL sodium carbonate solu­
tion.
F igu re 1. F lu o re sc e n c e d e ve lo p m e n t o f the reaction of
D e term in a tio n N A M with sulfite at 5(FC. S a m p le s : sulfite sta n d a rd
so lu tio n (262 ppm ; • ), ro s é w ine (1/10; ■ ), w hite w ine
Add 25 pL sample solution to 3.0 mL reaction buffer, pH (1/10; O ), an d red w ine (1/10; □ ).
9.2. Add 50 pL NAM acetone solution and then mix immedi­
ately. Seal the tube (75x10 mm id) with Parafilm, and let stand
for >5 min at room temperature. Incubate at 50°C for >20 min
Both free and bound sulfite can react with NAM under these
and then cool to room temperature by immersion in a water
conditions within 15 min because of the rapid conversion of the
bath for >5 min. Measure the fluorescence intensity (FT) at 435
bound sulfite to free sulfite by equilibrium. This is supported
nm (excitation wavelength, 360 nm). Correct FI of sample by
by the fact that the total sulfite is determined by the enzymatic
that of the stock solution that was diluted 1:100. Determine
sulfite from a calibration curve prepared from the stock solu­ method conducted at pH 8.
tion that was diluted from 1:2000 to 1:100. The calibration curve of sulfite shows a linear relation in the
range of 1.31-26.2 ppm (correlation coefficient, r = 0.999).
Results and Discussion The coefficients of variation (CVs) of the fluorescence intensi­
ties for 6 runs were 0.83% (26.2 ppm) and 0.91% (5.25 ppm).
The reaction of sulfite with NAM was discussed in previous The recoveries from the 3 wine samples are shown in Table 1.
reports (15, 16). The reaction proceeds in 2 steps: a Michel- These results show that the present method can rapidly deter­
type addition to form a succimide derivative and then its hy­ mine sulfite in wine with good reproducibility and high sensi­
drolysis to 2 succinamide derivatives. The first step proceeds at tivity.
20°C in a few minutes. Higher pH and temperature and an in­ The determination of free sulfite is important, because wine
crease of background FI accelerate the second step. The best production is controlled only by the free sulfite, not by the
condition for minimum variation and low background FI was bound sulfite. In our present method, the procedure is designed
to conduct the reaction at pH 9.2 and at room temperature for to calculate the free sulfite from the difference between total
>5 min and then at 50°C for >20 min. and bound sulfite. The bound sulfite was determined after the
Figure 1 shows the time course of fluorescence develop­ free sulfite was eliminated under acidic conditions at low tem­
ment during incubation at 50°C. The gradual increase of FI perature, which is similar to the distillation of free sulfite in the
shows the progress of the second step; the FIs of the succi- modified Rankine method.
namides are slightly stronger than that of the succimide. With Figure 2 shows the decrease of sulfite in the standard solu­
wine samples, the FI reached a constant level within 15 min tions to 20 pL sulfite solutions (26.2 and 2.62 ppm). More than
(Figure 1). 93% of sulfite disappeared upon standing for 40 min in an ice

T able 1. R e c o v e ry o f sulfite from w ine

65.3 ppm added 130 ppm added

Wine Found, ppm Found, ppm Flea, ppm Rec., % Found, ppm Rea, ppm Rec., %
White 127.0 193 66 101 255 128 98.5
Rosé 99.6 167 67 103 224 124 95.4
Red 56.9 126 69 106 186 129 99.2
 A k a s a k a E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1387

F igu re 2. D e c re a se o f sulfite u nder a c id ic co n d itio n s.

T h e 262 p p m ( • ) an d 26.2 p p m ( O ) sulfite sta n d ard
s o lu tio n s were used.
by Modi f i ed Ranki ne Method

F igu re 4. R e la tio n sh ip betw een total sulfite

bath. Part of the sulfite might be diffused; the remainder might determ ined b y m odified R a n k in e m ethod an d b y
be oxidized by the oxygen in the air in this method. Figure 3 fluorom etry. S a m p le s : ro sé ( ▲ ) , w hite (• ), a n d red ( ■ )
shows the decrease of sulfite in the wine by the same procedure wine. W in e s to w h ich sulfite w a s ad d e d : ro sé ( A ),
as described above. Sulfite decreased for 40 min and then re­ white ( o ) , an d red ( □ ).
mained constant. This means that free sulfite disappeared
within 40 min, whereas the bound sulfite is not affected under
these conditions. The remaining bound sulfite was determined method as modified by Fujita et al. (5), which is one of the
official methods in Japan. The free sulfite is calculated from the
after neutralization with sodium carbonate solution. The CVs
difference between the total and bound sulfite. Figures 4 and 5
of the bound sulfite were 1.7-3.2% (3 wine samples; 6 mns for
show the relationships between total and bound sulfite by the 2
each sample).
methods, respectively. Good agreement is obtained between
Total and bound sulfites were determined in 18 wine sam­
the 2 methods. Their correlation coefficients were 0.990 for
ples by the present method. Three samples of wine were spiked
with known amounts of sulfite (100-150 ppm) as the sodium
bisulfite salt and stored for >3 days at 5°C. For comparison, the
total and the free sulfite were also determined by the Rankine

by Modi f i ed Ranki ne Method

F igu re 5. R e la tio n sh ip betw een b o u n d sulfite deter­

F igu re 3. D e c re a se o f sulfite in w ine u nder an acid ic m ined by m odified R an k in e m ethod a n d b y fluorom etry.
c o n d itio n at 0°C. S am p le : red w in e s; total an d free S a m p le s : ro sé ( ▲ ) , w hite (• ), an d red ( ■ ) wine. W in e s
sulfite, 141 an d 37.2 ppm , respectively, determ ined by to w h ich sulfite w a s added: ro sé ( A ), w hite ( o ) , an d
m od ifie d R an k in e m ethod. red ( □ ).
1388 A k a s a k a E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

total and 0.987 for bound sulfite. The free sulfite by the present (5) Fujita, K., Ikuzaw a, M ., Izum i, T., H am ano, T., M ituhashi,
method also agrees well with the values for free sulfite deter­ Y., M atsuki, Y , A dachi, T., N onogi, H., & Fuke, T. (1979) Z
mined by the modified Rankine method (r = 0.983). Fig 4, 5 L e b e n sm . U n te rs . F o r sc h . 168, 206—211

The present method requires <200 pL of each sample to (6) O gaw a, S., Suzuki, H., Toyoda, M ., Ito, Y , & Iw aida, M.
obtain the average values of 5 runs for both the total and the (1 9 7 9 )Z L e b e n s m . U n te rs . F o r sc h . 168, 2 9 3 -2 9 8
bound sulfites. A single analyst can determine several samples (7) Beutler, H.O ., & Sehüte, I. (1983) D ts c h . L e b e n s m . R u n d s c h .
concurrently in 2 h by using a simple method. Also, no special 79, 3 2 3 -3 3 0
apparatus is required. Substances in wine such as dye, other (8) Beutler, H.O. (1984) Z A n a l C h e m . 317, 6 6 4 -6 6 5
anions, and cations do not interfere in the present method. (9) H olak, W„ & Patel, B. (1987) 7. A s s o c . O ff. A n a l. C h e m . 70,
5 7 2 -5 7 8
Acknowledgments (10) Hillery, B.R., Elkins, E.R., W arner, C.R., D aniels, D., &
Fazio, T. (1 9 89)7. A s s o c . O ff. A n a l. C h e m . 72, 4 7 0 -4 7 5
This work was supported in part by a grant-in-aid (Bio Me­
(11) Sullivan, J.J., H ollingw orth, T.A., W ekell, M .M ., M eo, V.A.,
dia Program) from the Ministry of Agriculture, Forestry, and Saba, H.H., M oghadam , A., E klund, C., Phillips, J.G ., &
Fisheries, Japan. G um p, B.H . (1990)7. A s s o c . O ff. A n a l. C h e m . 73, 3 5 -4 2
(12) Sullivan, J.J., H ollingw orth, T.A., W ekell, M .M ., M eo, V.A.,
References Saba, H.H., M oghadam , A., Phillips, J.G ., & G um p, B.H.
(1990) 7. A s s o c . O ff. A n a l. C h e m . 73, 2 2 3 -2 2 6
(1) O tsuka, K , Totsuka, A., Ito, M ., M iyazaki, K., K aw am atsu,
(13) Kim , H.J., Park, G.Y., & K im , Y.K. (1987) F o o d T ech n ol.
M ., Sim izu, S., Shiragam i, K., & Sano, E. (1973) 7. S o c .
41(1), 85-91
B rew . J a p a n 68, 4 4 5 -4 5 9
(14) A kasaka, K., M atsuda, H., O hrui, H., M eguro, H., & Suzuki,
(2) W edzicha, B.L. (1984) C h e m is tr y o f S u lp h u r D io x id e in
T. (1990) A g r ic . B io l. C h e m . 54, 5 0 1 -5 0 4
F o o d s , Elsevier A pplied Science Publishers LTD, E ssex, E ng­
land (15) M eguro, H., Takahashi, C., M atsui. S., & O hrui, H . (1983)
(3) Yoshizawa, S., & O tsuka, K. (1971)7. S o c . B rew . J a p a n 66, A n a l. L e tt. 16, 1625-1632

6 1 6 -6 1 9 (16) A kasaka, K., Suzuki, T., O hrui, H , & M eguro, H. (1986)

(4) R ankine, B.C. (1962) A u str. W ine B rew . S p ir it R ev. 80, 14—15 A g r ic . B io l. C h e m . 50, 1139-1144
 C a r d w e l l E t A l .: J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1389

TECHNICAL COMMUNICATIONS

D e t e r m i n a t i o n o f S u l f u r D i o x i d e i n W in e s a n d B e v e r a g e s b y F l o w

I n j e c t i o n A n a l y s i s w i t h R e d u c t iv e A m p e r o m e t r i c D e t e c t io n a n d

E le c t r o ly t ic C le a n u p

T erence J. C ardwell, R obert W. C attrall, and G uo N an C hen

La Trobe University, Centre for Scientific Instrumentation, Chemistry Department, Melbourne, Victoria 3083, Australia
G eoffrey R. Scollary
University of Melbourne, School of Chemistry, Parkville, Victoria 3052, Australia
I an C . H amilton
Newcastle Laboratories, BHP Research, Wallsend, New South Wales 2287, Australia

A new flow injection method is described for the de­ On-line electrolytic cleanup has rarely been used in FIA, yet
term ination of sulfur dioxide in red and white wines this methodology seems to have significant potential for the
and other beverages. A dual-electrode electro­ electrochemical removal of interferences in the analysis of
chemical detector eliminates interferences by re­ complex mixtures without the need for chemical pretreatment
duction at an upstream coulom etric electrode be­ or procedures for separation of the samples. In a recent paper
fore reductive detection of sulfur dioxide at the (11), we described an FIA system that consisted of a coulomet­
amperometric electrode. The data for free and total ric electrode to eliminate interferences oxidatively in the deter­
sulfur dioxide in wines and other beverages agree mination of sulfur dioxide in wines by amperometric detection.
well with those obtained by the standard aspiration- The results obtained for free and total sulfur dioxide in white
oxidation method. wines agreed well with those determined by the standard aspi­
ration-oxidation method, but the method gave unreliable re­
sults for the analysis of red wines.
T he determination of sulfur dioxide in wines and other This paper describes an improved FIA method that is based
beverages by flow injection analysis (FIA) is difficult on the reductive elimination of interfering components com­
because of the complex nature of the sample matrix. bined with reductive amperometric detection of the analyte.
Measurements usually require some form of sample cleanup or The method is suitable for the analysis of red and white wines
pretreatment to remove interferences, and most of the reported and some other beverages.
methods incorporate a gas diffusion membrane in the FIA sys­
tem to separate the gaseous analyte. Ruzicka and Hansen (1) Experimental
described a stopped-flow procedure based on the reaction of
sulfur dioxide with p-rosaniline; this method was improved Flow-injection system .—The flow injection system consists
later by introduction of a gas-diffusion unit into the FIA mani­ of an Altex Model 110A reciprocating pump, a pulse damper,
fold (2). Alternative methods that achieve separation by a gas- a Rheodyne 7125 injection valve fitted with a 10 pL sample
diffusion membrane combined with spectrophotometric detec­ loop, and an ESA 5100A Coulochem electrochemical detector
tion include the decolorization of malachite green (3,4) and the with a 5011A analytical cell (Figure 1). An in-line filter is
reaction with /j-aminoazobenzene (5). Other methods use am­ placed between the injector and the detector to protect the ana­
perometric detection (6) and suppression of luminol chemilu­ lytical flow cell from particulate matter.
minescence (7) after on-line sample treatment by gas diffusion. The dual-electrode Model 5011A analytical cell contains a
Dual electrode amperometric detectors consisting of 2 coulometric electrode of large surface area upstream from and
working electrodes in a series or parallel configuration have in series with a high-efficiency amperometric electrode. The
been used widely in liquid chromatography, mainly to improve working electrodes are constructed from porous graphite, and
selectivity (8). Matson and coworkers (9, 10) described a the ratio of the surface area of the coulometric electrode to the
unique cell consisting of 2 porous graphite coulometric elec­ amperometric electrode is 8.3 (12). The controlling couple for
trodes in series. This detector improved the ability to eliminate the reference electrode is the H+/H2 ion couple (13). Through­
unwanted signals from the chromatographic eluant or to re­ out this work, the Model 5011A analytical cell is operated in
move interfering components from the sample. the SCREEN mode, where the upstream coulometric electrode
(DET 1) eliminates interferences before detection of the ana­
R e c e iv e d S e p te m b e r 4 , 1 9 9 2 . A c c e p te d F e b ru a ry 2 3 , 1993. lyte at the amperometric electrode (DET 2).
1390 C ardwell E t Al .: J ournal Of A O A C International V ol . 76, N o . 6 ,1 9 9 3

hyde; 200 mg/Lchloride; 100 mg/L quinone, gallic, and ascor­

bic acids; and 10 mg/L bromide.

FIA for Wine and Beverage Samples

For free S 0 2, pipet 2 mL wine (or beverage) into a 10 mL

F ig u r e 1. B lo c k d ia g ra m o f F IA s y s te m : C , c a rrie r; P,
volumetric flask, add 1 mLO.OlM H2 S 0 4, and dilute to volume
p u m p ; A , p u ls e d a m p e r; V, in je c tio n v a lv e ; D1/D2, E S A
with H20 before injection under the selected FIA conditions.
5011A D u a l E le c tr o d e A n a ly tic a l C e ll; M, E S A M o d e l
For total S 0 2, pipet 1 mL (or less, depending on the concen­
5 1 0 0 A C o u lo c h e m C o n tr o l M o d u le ; R, r e c o r d e r o r
tration of S 0 2 in the sample) of a wine or beverage sample into
e le c t r o n ic in te g ra to r; W, w a s te .
a 10 mL volumetric flask, add 0.5 mL 0.5M NaOH solution,
and dilute to volume with H20 before injection under the se­
Reagent lected FIA conditions. Calculate the concentrations of free and
total S 0 2 from the respective calibration plots.
Sulfur dioxide standards.—Standard solutions were pre­
pared from a stock solution containing 1 0 0 mg sulfur diox-
ide/L. The stock solution was standardized daily by iodimetric R esu lts a n d D iscu ssio n
titration.
Hydrodynamic voltammetry o f sulfur dioxide in the dual­
Simplex Optimization of the FIA System electrodeflow cell.—Hydrodynamic voltammograms of sulfur
dioxide generated by stepwise variation of the potential of the
The Modified Simplex Method (14) was used to optimize
amperometric electrode (DET 2) in the ESA 5011A flow-cell
the performance of the FIA system for the determination of
are shown in Figure 2. Plots give a plateau beginning at about
sulfur dioxide in the presence of other electroactive species.
-0.40 V (vs H+/H2 couple), which is in good agreement with a
Quinone was selected as a model interfering component, and
voltammetric wave at -0.60 V (vs standard calomel electrode)
the system was optimized by determining the conditions re­
on a glassy carbon electrode reported previously (17). The
quired for complete elimination of quinone interference by the
electrochemical reduction of sulfur dioxide at the carbon elec­
upstream electrode (DET 1). At DET 2, the response of an in­
trode is dependent on the pH of the carrier, i.e., the lower the
jected solution of 10 mg sulfur dioxide/L was compared under
pH the higher the sensitivity. Although the responses in 0.1 M
identical conditions with the response from a solution of 1 0 0
mg/Lin quinone and 10 mg/L in sulfur dioxide. In the simplex-
ing experiments, the parameters that were varied included the
potentials of DET 1 and DET 2 (range, 0.0 to -0.7 V), carrier
flow rate (0.2-4.0 mL/min), line length from injector to detec­
tor cell (10-100 cm), carrier pH (1.0-5.0), and sample volume
(3-100 (XL).

Aspiration-Oxidation for Wines and Beverages

Free sulfur dioxide was determined by aspirating 20 mL of

an acidified wine sample by air flowing at 1 L/min for 20 min,
trapping the evolved gas in 10 mL 0.3% hydrogen peroxide
solution, and titrating the resulting sulfuric acid with a standard
sodium hydroxide solution (15, 16). Total sulfur dioxide was
determined by the same procedure, but the acidified sample
was boiled gently during the aspiration step.

Interference Studies

A 10 mg/L solution of sulfur dioxide was injected into the

FIA system under the optimum conditions (DET 1 at -0.26 V
and DET 2 at -0.41 V), and the signal of the amperometric cell
(DET 2) was recorded. The desired amount of interfering com­
pound plus 10 mg/L sulfur dioxide solution was then injected
under the same conditions, and the signal heights of the 2 meas­
urements were compared. The maximum level of each interfer­ F ig u re 2. H y d r o d y n a m ic v o lt a m m o g ra m s o f S O 2 at
ing compound was as follows: 15% ethanol; 5 g/L tartaric, cit­ D E T 2 : D ET1 p o te n tia l, 0.0 V; S O 2 c o n c e n tra tio n , 10
ric, malic, succinic, acetic, and benzoic acids, glucose, m g/L; c a rrie r 1 ,0 .0 0 1 0M H 2S O 4; c a rrie r 2 ,0 .0 1 M H 2S O 4;
fructose, carbonate, and phosphate; 1 g/L pyruvic acid, 2-oxo- c a r r ie r 3, 0.1 M H 2S O 4; f lo w rate, 1 m L /m in ; In je ctio n
glutaric acid, arabinose, proline, ethyl acetate, and acetalde­ v o lu m e , 10 p L .
 C ardwell E t A l .: J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3 1391

and 0.01M H2 S 0 4 carriers are similar, 0.1M H 2 S 0 4 is too

acidic and leads to corrosion in the FTA system. Therefore,
0.01M H 2 S 0 4 was chosen as the carrier for the remaining ex­
periments. The hydrodynamic voltammogram of sulfur diox­
ide at DET 1 was found to be identical to that on DET 2.
Hydrodynamic voltammetry o f ascorbic acid, gallic acid,
and quinol.—Ascorbic and gallic acids and quinol were se­
lected as typical organic acids and model phenolics that may be
present in wines, and their electrochemistry was investigated to
determine whether these or similar compounds would interfere
in the analysis of sulfur dioxide at a detection potential of about
-0.4 V. In those cases in which interference occurs, elimination
of the interference may be possible by electrolysis at a carefully
chosen potential at the upstream electrode. The oxidation of F ig u re 3. E le c tr o c h e m ic a l b e h a v io r o f a s c o r b ic a c id :
these compounds was also examined, because they are easily c u r v e 1, h y d r o d y n a m ic v o lta m m o g ra m o f a s c o r b ic a c id
oxidized in solution by oxygen. Therefore, the reduction of a t D ET1 ; c u r v e 2, D E T 2 r e s p o n s e w h e n D E T 1 is s e t a t
their oxidized forms must be considered, because they are +0.2 V; a s c o r b ic a c id c o n c e n tra tio n , 10 m g/L; ca rrie r,
likely to be present in significant proportions in wine samples 0.01 M H 2S O 4; f lo w ra te a n d in je c tio n v o lu m e , s e e
(18,19). F ig u re 2.
The hydrodynamic voltammogram of ascorbic acid is
shown in Figure 3. It indicates that ascorbic acid is not reduced gives a hydrodynamic voltammogram at DET 2 that shows a
in the range -0.2 to -0.5 V and is easily oxidized at positive reduction wave in the region of -0.2 to -0.5 V (curve 2, Figure
potentials, giving a plateau starting at +0.05 V. If DET 1 is set 5). Therefore, if DET 1 is setat—0.2 V, quinone will be reduced
at -0.2 V to let all ascorbic acid pass through to DET 2, only an and will not interfere in the detection of sulfur dioxide at higher
oxidation wave is observed for ascorbic acid at DET 2. Because negative potentials on DET 2.
dehydroascorbic acid is the main form of ascorbic acid in wine Optimization o f the FIA systemfo r sulfur dioxide.—A series
and other foods (18, 19), it is easily generated in a flowing of 25 experiments performed by the “variable-size” Modified
stream by setting the potential of DET 1 to +0.20 V to oxidize Simplex Method (14) led to the selection of the following op­
all the ascorbic acid passing through DET 1. When the poten­ timized conditions for the determination of sulfur dioxide by
tial of DET 2 is varied from 0 to -0.50 V, there is no indication the dual-electrode FIA system: DET1 potential, (vs H+/H 2 cou­
that dehydroascorbic acid is reduced at DET 2 (curve 2, Figure ple), -0.26 V; DET2 potential (vs H+/H2 couple), -0.41 V; flow
3). Therefore, ascorbic acid and dehydroascorbic acid should rate, 0.8 mL/min; line length from injector to analysis cell, 65
not interfere in the determination of sulfur dioxide at a reducing cm; carrier pH, 1.47; injected volume, 10 pL. By using these
potential.
The hydrodynamic voltammogram of gallic acid (curve 1,
Figure 4) indicates that it is not reduced in the region of -0.1 to
-0.5 V but is oxidized in 2 steps with a plateau appearing at
about -0.05 V and +0.04 V. If the DET 1 potential is set at +0.2
V on the plateau of the first oxidation step, the response of DET
2 at different potentials shows that the first oxidized form of

gallic acid is easily reduced at potentials more negative than

-0.1 V (curve 2, Figure 4). When the DET 1 potential is set at
+0.5 V on the second oxidation wave, a potential sweep of DET
2 in the region of 0 to -0.50 V gives negligible current re­
sponse, indicating that the fully oxidized form of gallic acid is
not reduced in this potential region. These results suggest that
the oxidized forms of phenolic components similar in structure
to gallic acid are unlikely to interfere in the determination of
sulfur dioxide or are easily reduced at potentials more positive
than those required for detection of S 0 2, thereby eliminating
their interference.
The quinol/quinone couple is a well-known, thermody­
namically reversible redox couple. The hydrodynamic voltam­ F ig u r e 4. E le c tr o c h e m ic a l b e h a v io r o f g a llic a c id : c u r v e
mogram of quinol (curve 1, Figure 5) shows that it is not re­ 1, h y d r o d y n a m ic v o lta m m o g ra m o f g a llic a c id a t D ET1 ;
duced in the potential range of -0.2 to -0.5 V but is easily c u r v e 2, D E T 2 r e s p o n s e w h e n D ET1 is s e t a t +0.2 V;
oxidized to quinone to give a plateau beginning at about 0.0 V. g a llic a c id c o n c e n tra tio n , 10 m g /L ; o th e r c o n d it io n s , s e e
If the potential of DET 1 is set at +0.10 V, the resulting quinone F ig u r e 3.
1392 C ardw eix Et Al .: J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3

2 0 -

<
0 -
c
3
U
-2 0 -

-(X6 -0 4 -0*2 0 02
Potential (V )

F ig u r e 5. E le c tr o c h e m ic a l b e h a v io r o f q u in o l/ q u in o n e :
c u r v e 1, h y d r o d y n a m ic v o lta m m o g ra m o f q u in o l a t
D E T 1 ; c u r v e 2, D E T 2 r e s p o n s e w h e n D ET1 is s e t a t +0.1
V; q u in o l c o n c e n tr a tio n , 10 m g/L; o th e r c o n d it io n s , s e e
a b
F ig u r e 3. F ig u r e 6. T y p ic a l F IA r e s p o n s e s o f a w h ite w in e s a m p le
a t b o th d e te c to rs : tr a c e 1, D ET1 s ig n a l; tr a c e 2, D E T 2
s ig n a l; f lo w ra te , 0 .8 m L /m in ; o th e r c o n d it io n s , s e e
conditions, the coulometric efficiency of DET 1 is 100% for
F ig u r e 3; a: D ET1 p o te n tia l, -0 .0 5 V (n o r e d u c tio n o f
levels of the interferant quinone up to at least 100 mg/L.
w in e c o m p o n e n ts ) ; D E T 2 p o te n tia l, -0 .4 1 V; s ig n a l a t
The effectiveness of these conditions in removing interfer­
D E T 2 c o n v e r ts to 3 9 m g /L fre e s u lf u r d io x id e ; b: D ET1
ences in the determination of sulfur dioxide in one of the white
p o te n tia l, -0 .2 6 V ( re d u c tio n o f in te rfe re n c e s ); D E T 2
wine samples is demonstrated in Figure 6 . In Figure 6 (a), a
p o te n tia l, -0 .4 1 V; s ig n a l a t D E T 2 c o n v e r t s to 20 m g /L
potential of -0.05 V at the upstream electrode is ineffective in
fre e s u lf u r d io x id e .
reducing any of the components in the wine sample (i.e., this is
identical to DET 1 being switched off), whereas a potential of
-0.41 V at DET 2 gives a response that corresponds to 39 mg dium hydroxide. All of the remaining tested components do not
free sulfur dioxide/L. When the potential of DET 1 is adjusted interfere significantly in the determination of sulfur dioxide by
to the optimized potential of -0.26 V and the potential of DET this method.
2 is maintained at -0.41 V [Figure 6 (b)], the response at DET
2 decreases to a level corresponding to 20 mg sulfur dioxide/L.
Determination of Sulfur Dioxide in Wines and
Beverages
This level is in good agreement with the value of 19 mg/L ob­
tained for the sample by the aspiration-oxidation procedure.
Under the optimized conditions, the calibration plots for
These results suggest that the inflated DET 2 signal in Figure
free and total sulfur dioxide are linear up to 20 mg/L, with cor­
6 (a) arises from reduction of wine components and sulfur di­
relation coefficients of 0.998 and 0.999, respectively. As many
oxide, and the decrease in this detector signal in Figure 6 (b)
as 30 determinations per hour can be made for each analysis.
results from reduction of these components at the upstream
electrode, which leads to effective removal of the interferences. The new FIA method was applied to 27, mainly dry, wines
(16 whites, 1 1 reds) and 1 1 other beverages (fruit juices, wine
To further demonstrate the success of these conditions, the
coolers, and ciders), and the data were compared to those ob­
effects were studied, under the optimum FIA conditions, of
tained by the standard aspiration-oxidation method. The corre­
more than 2 0 frequently occurring compounds in wine, includ­
lation of these data for free and total sulfur dioxide is shown in
ing sugars (glucose, fructose, arabinose), organic acids (tar­
Figure 7. The free sulfur dioxide content ranged from 3 to 40
taric, citric, malic, succinic, benzoic, pyruvic, gallic, ascorbic,
mg/L, and excellent linearity was obtained with flow injection
acetic, 2 -oxoglutaric), inorganic ions (phosphate, chloride, bro­
giving a slightly higher concentration than that by aspiration-
mide, carbonate), ethanol, glycerol, proline, ethyl acetate, ac­
oxidation (slope = 1.058). The total sulfur dioxide content cov­
etaldehyde, and quinone. Each test solution contained sulfur ered the range from 12 to 160 mg/L, with excellent agreement
dioxide at 10 mg/L in the presence of an added constituent at between the 2 methods (slope = 0.97). Precisions of the 2 meth­
concentrations in excess of those normally found in wines (2 0 ). ods are almost identical with flow injection; relative standard
The carbonyl compounds acetaldehyde, pyruvic acid, and 2- deviation (RSD) values are about 3% for free and 4% for total
oxoglutaric acid interfere by giving low recoveries of sulfur sulfur dioxide, whereas the aspiration-oxidation method gives
dioxide due to the formation of bisulfite addition products in RSD values of about 4%. These measurements confirm that the
solution. Addition reactions occur in wine and diminish the FIA method is sufficiently accurate and precise to replace the
availability of free sulfur dioxide, but the bound sulfur dioxide aspiration-oxidation procedure for routine monitoring of sulfur
is readily released by treatment of the sample with dilute so- dioxide in wines and beverages.
 C ardwell E t A l .: J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3 1393

F ig u r e 7. C o rre la tio n b e tw e e n f lo w in je c tio n a n d a s p ira tio n - o x id a tio n m e th o d s in th e d e te rm in a tio n o f s u lf u r d io x id e

in w in e s a n d b e v e ra g e s : (a), fre e s u lf u r d io x id e in 3 8 s a m p le s ; [FIA] = 1 .0 5 8 (A s - o x ) - 0.064; r = 0.982; (b), to ta l s u lfu r
d io x id e in 34 s a m p le s ; [FIA] = 0.97 ( A s - o x ) + 0.78; r = 0.984.

The method described in this papier offers significant advan­ (5 ) B a r t r o l l i , J ., E s c a l a d a , M . , J o r q u e r a , C . J . , & A l o n s o , J . ( 1 9 9 1 )

tages over existing flow injection procedures. First, the single Anal. Chem. 63, 2 5 3 2 - 2 5 3 5
channel FIA system is easier to assemble and operate than the (6 ) G r a n a d o s , M . , M a s p o c h , S ., & B l a n c o , M . ( 1 9 8 6 ) Anal.
more complex manifolds containing a gas diffusion module; Chim. Acta 179,4 4 5 ^ 1 5 1
second, the FIA system does not require modification for the (7 ) H u a n g , Y .L ., K i m , J . M . , & S c h m i d , R . D . ( 1 9 9 2 ) Anal. Chim.
determination of free and total sulfur dioxide in wines and bev­ Acta 266, 3 1 7 -3 2 3

erages; third, the use of chemical reagents is minimal, because (8 ) K i s s i n g e r , P .T . ( 1 9 8 4 ) in Laboratory Techniques in Elec-
troanalytical Chemistry , P .T . K i s s i n g e r & W .R . H e i n e m a n
the entire system is electrochemically controlled. The main ad­
(E d s ), M a r c e l D e k k e r, N e w Y o rk , N Y , C h a p te r 2 2
vantage of this new method over the FIA oxidative electro­
(9 ) M a t s o n , W .R ., A n d r e w s , R .W ., B a l l , J .. S k i n n e r , D ., V i-
chemical method in our previous report ( 1 1 ) is that it is appli­
t u k e v i t c h , R ., & Z i n k , E .W . ( 1 9 8 1 ) “ A N e w E l e c t r o c h e m i c a l
cable to red wines and beverages and still maintains the high
H P L C D e te c to r,” P a p e r 5 6 5 , P itts b u r g h C o n fe re n c e o n A n a ­
sample throughput of 30/h with good precision. ly tic a l C h e m is tr y a n d A p p lie d S p e c tr o s c o p y , A tla n tic C ity , N J

(1 0 ) A n d r e w s , R .W ., S c h u b e r t , C . , M o r r i s o n , J ., Z i n k , E . W ., &
A ck n o w led g m en ts M a t s o n , W .R . ( 1 9 8 2 ) A m . Lab. 14, 1 4 0 -1 5 5

(1 1 ) C a r d w e l l , T . J ., C a t t r a l l , R .W ., C h e n , G . N . , l i e s , P .J ., H a m i l ­

We thank the Australian Research Council for financial sup­ t o n , I . C ., & S c o l l a r y , G . R . ( 1 9 9 1 ) Electroanalysis 3, 8 5 9 -8 6 3

port, and G.N. Chen thanks La Trobe University for the post­ (1 2 ) H a r o o n , Y ., S c h u b e r t , C . A . W . , & H a u s c h k a , P .V . ( 1 9 8 4 ) J.
graduate scholarship award. Chromatogr. Sci. 22, 8 9 -9 3
(1 3 ) In s tru c tio n M a n u a l f o r th e M o d e l 5 1 0 0 a C o u lo c h e m D e te c ­
to r , E S A , I n c . , B e d f o r d , M A
R efe ren c es
(1 4 ) D e m in g , S .N ., & M o r g a n , S .L . ( 1 9 7 3 ) Anal. Chem. 45,
2 7 8 A -2 8 3 A
(1) Ruzicka, J., & Hansen, E.H. (1980) Anal. Chim. Acta 114, (1 5 ) R a n k in e , B .C . ( 1 9 6 2 ) Aust. Wine Brew. S pirit Rev. 8 0 , 14—1 6
19-44 (1 6 ) R a n k i n e , B . C . , & P o c o c k , K .F . ( 1 9 7 0 ) Aust. Wine Brew.
(2) Moller, J., & Winter, B. (1985) Fresenius Z. Anal. Chem. Spirit Rev. 88, 4 0 -4 4
320,451^456 (1 7 ) C a r d w e l l , T . J ., C a t t r a l l , R .W ., C h e n , G . N . , S c o l l a r y , G .R . , &

(3) Sullivan, J.J., Hollingworth, T.A., Wekell, M.M., Nleo, V.A., H a m ilto n , I.C . ( 1 9 9 1 ) Analyst 116, 2 5 3 -2 5 6

Saba, H.H., Etemad-Moghadam, A., Eklund, C., Phillips, (1 8 ) O u g h , C .S . ( 1 9 8 8 ) in Wine Analysis, H .F . L i n s k e n s & J .F .
J.G., & Gump, B.H. (1990) J. Assoc. Off. Anal. Chem. 73. J a c k s o n (E d s ), S p rin g e r-V e rla g , H e id e lb e rg , G e rm a n y , p 111
35-42 (1 9 ) S i n g l e t o n , V .L . ( 1 9 8 7 ) Am, J. Enol. Vine. 38, 6 9 -7 7

(4) Sullivan, J.J., Hollingworth, T.A., Wekell, M.M., Meo, V.A., (2 0 ) O u g h , C .S ., & A m e r in e , M .A . ( 1 9 8 8 ) Methods f o r Analysis
Etemad-Moghadam, A., Phillips, J.G., & Gump. B.H. (1990) o f Musts and Wines, 2 n d E d ., J o h n W ile y & S o n s , I n c ., N e w

J. Assoc. Off. Anal. Chem. 73, 223-226 Y o rk , N Y

1394 P aul & Shingare : J ournal O f A O A C International V ol . 7 6 , N o . 6 ,1 9 9 3

TECHNICAL COMMUNICATIONS

T h in - L a y e r C h r o m a to g r a p h ic D e te c tio n o f O r g a n o p h o s p h o r u s

In s e c tic id e s C o n ta in in g a N itro p h e n y l G r o u p

Vit t h a l B. P a u l 1 and M u r l id h a r S . S h in g a r e

Marathwada University, Department of Chemistry, Aurangabad, 431 004, India

A G rie ss reactio n th a t h a s b ee n u se d to d e te c t o r­ Experim ental

g an ic c o m p o u n d s containing an aro m atic am ino
g ro u p w as u s e d to d e te c t o rg a n o p h o sp h o ru s co m ­ Materials
p o u n d s co n tain in g a nitrophenyl group, s u c h a s
All reagents were analytical reagent grade. Distilled water
ethyl parath io n , m ethyl parathion, a n d fenitrothion.
was used throughout.
On red u ctio n with s ta n n o u s chloride in H C I-w ater
(a) Silica gel.—Silica gel G; particle size, 15 pg (E. Merck,
( 1 + 1 ), th e s e c o m p o u n d s give re sp ectiv e am ino d e ­
Darmstadt, Germany).
rivatives, w hich are fu rth er diazotized a n d co u p led (b) Insecticide standard solutions.—Technical-grade ethyl
with 1 -n ap h th y lam in e to give in te n se p in k -o ra n g e parathion, methyl parathion, and fenitrothion were used as
s p o ts . T his re ag en t also g iv es a bluish to violet standards. Prepare a 1 mg/mL solution in ethanol of each insec­
s p o t with p-nitrophenol, a m etabolite of feni­ ticide.
tro th io n . T he p ro c e d u re c a n b e u se d to d e te c t (c) Stannous chloride solution (5%).—Dissolve 5 g stan­
th e s e in secticid es in biological m aterials in fo ren ­ nous chloride in 100 mL 50% (v/v) HC1 (32%) in distilled
s ic toxicology. water.
(d) Sodium nitrite solution (5%).—Dissolve 5 g sodium ni­
trite in 100 mL 10% (v/v) acetic acid solution.
O rganophosphorus insecticides containing an aromatic (e) 1-Napthylamine reagent (0.1%).—Dissolve 0.1 g 1-
nitro group (-N 02) in their molecules, such as ethyl naphthylamine in 10 mL glacial acetic acid and dilute to 100
parathion, methyl parathion, and fenitrothion, are mL with distilled water.
widely used in agriculture. They are frequently misused in
homicidal and suicidal poisoning cases, and their charac­ Extraction of Insecticides from Biological Material
terization is, therefore, necessary. Thin-layer chromatography Approximately 50 g portions each of various visceral tissue
(TLC) is the method of choice for the identification of these types (stomach, intestine, liver, spleen, and kidney) containing
insecticides in biological material. insecticide were minced in aqueous solution. The insecticide
Reagents that can be used to detect organophosphorus in­ was extracted with diethyl ether, and the solvent was evapo­
secticides by TLC include palladium(II) chloride (1,2), 4-(p- rated at room temperature. The residue was dissolved in 1-
nitrobenzyl)pyridine tetraethylene pentamine (3), bromine 2 mL ethanol. A known volume of the solution was spotted on
fluorescein silver nitrate (4, 5), congo red (6 ), mercury® ni­ an activated TLC plate together with the standard solutions of
trate (7), mercury(II) nitrate-potassium hexacyanoferrate(II) the insecticides.
(8 ), and acidified potassium iodate-starch (9). Hamada (10)
reported the detection of parathion from organs of cadaver, and TLC Procedure
Gage (11) used KOH solution to detect parathion and related A standard glass TLC plate was coated with a slurry of silica
nitrophenyl derivatives by paper chromatography. gel G in water (1 + 2) to a thickness of 0.25 mm. The plate was
In this paper, we describe the use of the Griess reaction to activated by heating at 110°C for ca 1 h. Standard solutions
detect organophosphorus insecticides containing a nitrophenyl (10 (iL) of ethyl parathion, methyl parathion, and fenitrothion
group by TLC. After reduction with acidic stannous chloride, (Sumithion) in ethanol (each at a concentration of 1 mg/mL)
these insecticides give respective amino derivatives, which are were spotted on the plate, which was then developed in a pre­
further diazotized and coupled with 1-naphthylamine. This re­ viously saturated TLC chamber with hexane-acetone ( 9 + 1 )
agent gives an intense pink-orange spot with organophospho­ as solvent. After the solvent had traveled 10 cm up the plate,
rus insecticides containing nitrophenyl groups. the plate was removed from the chamber, dried in air, and
sprayed with 5% stannous chloride solution. The plate was then
Received November 30, 1992. Accepted March 2, 1993 heated for 10 min at 100°C. It was cooled and sprayed with
'Present Address: Assistant Director, Regional Forensic Science freshly prepared 5% sodium nitrite solution in 10% (v/v) acetic
Laboratory, State of Maharashtra, Cantonment, Aurangabad, 431 002, India. acid, followed by 1-napthylamine reagent. Ethyl parathion,
 P atil & S hingare : J ournal O f A O A C International V ol . 76, N o . 6 ,1 9 9 3 1395

T a b le 1. R F v a lu e s f o r 3 o r g a n o p h o s p h o r u s occur in blanks from untreated individuals and hence do not

in s e c t ic id e s c o n ta in in g a n itro p h e n y l g r o u p 3 interfere.
Insecticide RF The reagent described here is sensitive and selective for or­
ganophosphorus insecticides containing a nitrophenyl group
Ethyl parathion (Ekatox, Follidon) 0.60 and their active metabolite p-nitrophenol and can be used rou­
Methyl parathion (M e ta c id e -5 0 ) 0.35 tinely for the detection and semiquantitative determination of
Fenitrothion (Folithion, Sumithlon) 0.50 these insecticides in biological materials in forensic toxicology.

Solvent system, he xan e-a ceto n e (9 + 1).

A cknow ledgm ent

The authors are grateful to Prof. D. B. Ingle, Head, Depart­

methyl parathion, and fenitrothion gave dark pink-orange
ment of Chemistry, Marathwada University, Aurangabad, and
spots. The RF values are listed in Table 1.
to the Director, Forensic Science Laboratories, State of Ma­
harashtra, Bombay, for valuable advice and encouragement.
R e su lts a n d D iscu ssio n
R eferen ces
Stannous chloride in 50% HC1 (v/v) reduces ethyl para­
thion, methyl parathion, and fenitrothion to their respective (1 ) B a u m l e r , J ., & R i p p s t e i n , U . S . ( 1 9 6 1 ) Helv. Chim. Acta 44,

amino derivatives (12) on heating at 100°C. The amino deriva­ 1162

tive so formed is further diazotized by sodium nitrite in acetic (2 ) R a n d e ra th , K . (1 9 6 5 ) Thin Layer Chromatography, A c a ­
acid and coupled with 1 -napthylamine to give a dark pink-or­ d e m ic P r e s s , N e w Y o rk a n d L o n d o n , p . 2 0 5

ange spot. The pink-orange spot remains stable for a couple of (3 ) Analytical Methods f o r Pesticides, Plant
Z w e ig , G . (1 9 6 7 )

days. Other organophosphorus insecticides (malathion, di- Growth Regulators and Food Additives, V o l. V , A c a d e m i c
P re ss, N e w Y o rk a n d L o n d o n , p. 109
methoate, fenthion, phorate, disulfoton, oxydemeton methyl,
(4 ) W a lk a r , K .G ., & B e r o x a , M .J . ( 1 9 6 3 ) J. Assoc. O ff. Agric.
quinalphose, phosphamidon, dichlorvos, and trichlorphon) do
Chem. 46, 250
not give a colored spot. Moreover, organochlorine and car­
(5 ) D a v id e k , J .U .J ., & P o k o m y , J . ( 1 9 6 1 ) Z . Lebensm. Unters.
bamate insecticides and various constituents of visceral ex­
Forsch. 115, 113
tracts (peptides, proteins, etc.) do not interfere. The limit of
(6 ) I r u d a y a s w a m y , A ., & N a ta r a j a n , A .R . ( 1 9 6 5 ) Analyst (Lon­
detection of the reagent is about 1 |lg.
don) 90, 5 0 3
This reagent also gives a bluish to violet spot with p-ni- (7 ) K a w a l e , G . B . , J o g a l e k a r , V .D ., B a r v e , V .P ., & M a h a l , H .S .
trophenol (13; RF value, 0.30), a metabolite of ethyl parathion (1 9 7 2 ) Set Cult. 30, 373
and methyl parathion and with 3-methyl-4-nitrophenol (RF (8 ) K a t k a r , H . N . , & B a r v e , V .P . ( 1 9 7 6 ) Curr. Sci. 45, 662
value, 0.25), a metabolite of fenitrothion that is formed in living (9 ) P a t i l , V .B ., P a d a l k a r , S .V ., & K a w a l e , G . B . ( 1 9 8 7 ) Analyst
organisms through biotransformation and also found in com­ (London) 112, 1765
mercial formulations of these insecticides as a degradation (1 0 ) H a m ad a , H . (1 9 5 5 ) Shikoku Igaku Zasshi 7 , 389
product. The limit of detection of p-nitrophenol is about 1 Jig. (1 1 ) G a g e , J .C . ( 1 9 5 3 ) Biochem. J. 54, 4 2 6
Thus, with the help of this reagent, an unchanged insecticide (1 2 ) S h r e e R e m u lu , U .S . ( 1 9 8 5 ) Chemistry o f Insecticides and
and its active metabolite can be detected and distinguished si­ Fungicides, O x f o r d a n d I B H P u b li s h in g C o ., N e w D e lh i, p .
multaneously from others present in biological material. Other 139

organic compounds containing aromatic nitro groups and aro­ (1 3 ) C la r k e , E .G .C . ( 1 9 7 8 ) Isolation and Identification o f Drugs,
matic amino groups may interfere, but such substances do not T h e P h a rm a c e u tic a l P re s s, L o n d o n , p . 4 6 8
1396 T a l l m a d g e & L i n : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

TECHNICAL COMMUNICATIONS

L iq u id C h ro m a to g ra p h ic M e th o d fo r D e te rm in in g th e P e rc e n t o f

O le s tr a in L ip id S a m p le s

Da n ie l H. T a l l m a d g e and P e t e r Y.T. L in

The Procter & Gamble Co., Food Product Development, Winton Hill Technical Center, 6071 Center Hill Ave, Cincinnati,
OH 45224

A liquid ch ro m a to g rap h ic (LC) m ethod h a s b een d e ­ dietary oil from which it was derived and performs in the same
v elo p ed to d eterm in e th e p ercen t of o le stra in lipid manner as that oil during cooking or frying. However, olestra
sa m p le s . To ach iev e th e h ig h e st d e g re e of a c c u ­ is not digested by pancreatic enzymes; therefore, it is not ab­
racy, th is m eth o d re q u ires th e u s e of a n o le stra sorbed from the human gastrointestinal tract and contributes no
s ta n d a rd with th e s a m e m o lecular co m p o sitio n a s calories upon ingestion (1,2). Olestra is under review by the
th e o le stra in th e lipid sa m p le to b e analyzed. S am ­ U.S. Food and Drug Administration for use as a fat substitute
p les w ere an aly zed by re v e rse d -p h a se LC u sin g an in the preparation of specific foods.
ev a p o rativ e lig h t-scattering d etecto r. C h ro m ato g ra­ The present study describes a liquid chromatographic (LC)
ph y w a s p erfo rm ed with a 5 pm o ctad ecy lsilan e- method that uses a reversed-phase column with nonaqueous
Z orbax co lu m n th a t s e p a ra te s o le stra from o th e r reversed-phase (NARP) elution for determining the percent of
lipophilic c o m p o n e n ts. T hree ty p e s of o le stra olestra in olestra-lipid blends. NARP/LC has been used exten­
s ta n d a r d s (soybean-oil o lestra, u n h e a te d c o tto n ­ sively for the separation of triglycerides and other lipids and is
seed -o il o lestra, a n d h e a te d co tto n seed -o il olestra), reviewed elsewhere (3). No other methods for direct quantita­
e a c h an aly zed in s o y b e a n oil, sh o w e d linearity tion of olestra in olestra-lipid blends are known to exist. For
w h en th e am o u n t of o le stra injected ra n g e d from accurate results, this LC analysis requires the generation of an
20 to 160 pg (r= 0.9996). T he a re a u n d e r th e o lestra external standard curve using an olestra with the same molecu­
p eak (retention tim e 3.5 to 4.9 min) w a s u s e d to lar composition as the olestra in the sample to be analyzed. The
quantify th e am o u n t of o le stra in olestra-lipid s a m ­ “same molecular composition” implies identical fatty acid
p les, by co m p arin g th e o le stra a re a for th e sa m p le composition (chain length and degree of unsaturation) and de­
with th a t of th e s ta n d a rd u sin g a cu rv e derived by gree of esterification. This method has been validated and used
linear re g re ssio n . T he m ethod w a s ev a lu a ted u sin g to determine the percent of olestra in preformulated olestra-
3 ty p e s of o le s tra b len d ed with so y b e a n oil an d lipid blends. In addition, this method may be used to determine
varying th e p e rc e n t of o le stra in th e olestra-lipid the percent of olestra in olestra-lipid blends resulting from the
b lend from 5 to 90%. R ecovery of o le stra from solvent extraction of a variety of food and other sample ma­
th e s e olestra-lipid b le n d s varied from 99.2 to trixes.
106.0%, d e m o n stra tin g ex cellen t a ccu racy , with
m eth o d p recisio n e x p re s s e d a s th e coefficient of METHOD
variation, 0.9%. E ach erro r e stim ate w a s derived
from 5 parallel d eterm in a tio n s. With p ro p e r valida­
tion (e.g., ru n n in g an o lestra-free blank for e a c h
Apparatus
lipid m atrix), th is m eth o d p ro v id es a rapid, a c c u ­
(a) Liquid chromatographic system.—Model HP1090 with
rate, a n d p re c ise te c h n iq u e for m e asu rin g th e per­
binary or ternary DR5 delivery system, autosampler, and
c e n t of o le stra in lipids e x tra cted from olestra-for-
heated column compartment (Hewlett-Packard, Avondale,
m ulated fo o d s a n d in olestra-lipid blends.
PA); Rheodyne inline filter (0.5 pm), Zorbax ODS analytical
column (MAC-MOD Analytical, Inc., Chadds Ford, PA), 5
pm, 8 cm x 4 mm. Operating conditions: Flow rate, 2 mL/min;
O lestra, manufactured by the Procter and Gamble Com­
injection volume, 10 pL; column temperature, 40°C; stop time,
pany, is the common name for the mixture of hexa-,
8.01 min; post time, 4.0 min; solvent program, (A = acetoni­
hepta-, and octaesters formed by the chemical reaction
trile, B = methylene chloride): 0 min, 30% B; 4.0 min, 70% B;
of sucrose and long-chain fatty acid methyl esters from edible
6.0 min, 70% B; 6.01 min, 100% B; 8.0 min, 100% B; 8.01
oils. Olestra has the physical and organoleptic properties of the
min, 30% B.
(b) Detector.—Applied Chromatography Systems Model
Recieved October 16, 1992. Accepted January 18, 1993.
#750/14 evaporative light-scattering detector. Operating con­
 T a l l m a d g e & L in : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1397

ditions: Nebulizer gas pressure, 12 psi (nitrogen); evaporator Table 1. Standard curve preparation (regression from
setting, 50; attenuation, 2; photomultiplier, 2; time constant, 5. LOTUS)
(c) Data acquisitionsystem.—Analog data output from the Injected, Av. peak
LC detector was transmitted through a Hewlett-Packard A/3 L9 a rea/1000 SD CV, %
18542AAnalog-to-Digital Convertor to a Hewlett-Packard 1000 Soybean-oil olestra standards
computer, where data were integrated using Hewlett-Packard
Laboratory Automation System (LAS rev.D.01) software. 5 20.00 587 7.3 1.2
Reagents 5 40 .00 1553 15.5 1.0
5 60 .00 25 38 7.1 0 .3
(a) Solvents.—Acetonitrile (ACN) and methylene chlo­ 5 100.0 46 84 3.3 0.1
ride, all UV grade. 5 120.0 57 49 20.2 0.4

(b) Olestra.—Olestra was synthesized from sucrose and 5 140.0 6791 17.1 0.3

fatty acid methyl esters derived from vegetable oil stocks as 5 160.0 7922 19.0 0.2
Constant -5 3 8 .9 5 8
described by Vilpenhein (4). The vegetable oil stocks we used
I2 0.9 996 48
in this study were either hydrogenated soybean oil (iodine
X Coefficient 52 .494 8
value = 48.5) or cottonseed oil (iodine value = 89.1). Three
olestra standards were used: soybean-oil, cottonseed-oil, and Unheated cottonseed-oil olestra standards
heated cottonseed-oil. The heated cottonseed-oil olestra stand­
ard was prepared by heating cottonseed-oil-derived olestra fol­ 5 20.00 591 2.7 0.5
lowing synthesis of the olestra, using typical conditions for fry­ 5 40 .00 1524 3.7 0.2
ing potato chips (kettle-fry at 365°F for 12 h) (5). Heating 5 60.00 25 58 2.6 0.1
results in some degree of olestra polymer formation; therefore, 5 100.0 47 13 27.3 0.6
the molecular composition of the heated cottonseed-oil olestra 5 120.0 5835 23.1 0 .4
is different from the molecular composition of the unheated 5 140.0 6810 22.6 0.3
cottonseed-oil olestra (5). 5 160.0 78 46 11.5 0.2
(c) Lipid.—Refined, bleached, and deodorized (RBD) soy­ Constant -5 2 5 .8 7 6
bean oil was used for all the formulated olestra-lipid blends. i2 0.9 996 54
X Coefficient 5 2 .434 58
Preparation of Olestra Standard Solutions
H eated cottonseed-oil olestra standards
For each standard, accurately weigh ca 1.0 g (standard error
= ±0 . 0 0 1 g) of olestra of the same molecular composition as 5 20.02 517 5.8 1.2
that contained in the samples to be analyzed into a 50 mL volu­ 5 40 .04 1397 6.0 0.4
metric flask. Dilute to volume with methylene chloride. Pre­ 5 60.05 2364 11.0 0.5
pare this stock solution daily. Prepare 7 standard solutions, A - 5 100.1 4374 9.6 0.2
G, as follows: pipet 1.0,2.0,3.0,5.0,6.0,7.0, or 8.0 mLof stock 5 120.1 53 99 18.2 0.3
solution into a 10 mL volumetric flask and dilute to volume 5 140.1 6391 18.7 0.3
with methylene chloride. Filter a portion of each standard so­ 5 160.1 74 24 20.8 0.3
lution through a 0 . 2 pm syringe filter into an autosampler vial Constant -5 5 7 .8 5 3
and cap the vial. Prepare standards daily. i2 0 .9 996 68
X Coefficient 4 9 .6 0 2 6 2
Preparation of Olestra Samples
a N = N um ber of parallel determinations for each prepared sample;
(a) Sample preparation.—Prepare known olestra- each injection was a 10 pL injection.
triglyceride blends by weighing both components and mixing
common lipids (e.g., triglyceride, diglyceride) with a specific
olestra standard, as mentioned above. In this study, we used 1 lipid, amount of lipid required: 0-5%, 2 g; 5.1-10%, 1 g; 10.1-
triglyceride—RBD soybean oil—to prepare all of the olestra- 20%, 0.5 g: 20.1-50%, 0.2 g; 50.1-100%, 0.1 g. Weigh the
lipid samples and the olestra-free lipid blank. Lipid extracts,
calculated amount of lipid (±0.001 g) into a 10 mL volumetric
resulting from the solvent extraction with hot ethylene dichlo­
flask. Dilute to volume with methylene chloride. Filter a por­
ride of food products formulated with blends of olestra and
tion of the sample solution through a 0 . 2 pm syringe filter into
vegetable oils, may also be used as samples.
an autosampler vial and cap the vial.
(b) Estimation of olestra concentration.—To insure that
(c) Check-sample preparation.—All check samples con­
the area under the olestra peak for a sample is contained within
the range of the calibration standards (achievable with an oles­ tained 35% olestra. Accurately weigh ca 0.175 g (±0.001 g) of
tra concentration of 5-10 mg olestra/mL), estimate the percent each olestra standard and ca 0.325 g (±0.001 g) of a typical
of olestra in total lipid (including olestra) for each olestra-lipid vegetable oil (RBD soybean oil in our study) into a 25 mL volu­
blend or sample. This estimate will determine the amount of metric flask. Dilute to volume with methylene chloride. Filter
total lipid required in 10 mL, as follows. Percent olestra in total a portion of the check sample solution through a 0 . 2 pm filter
1398 T a l l m a d g e & L i n : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

Table 2. Recovery of olestra from lipid blends

Recovered,
N8 Added, %b Found, %b %c CV, %

Soybean-oil olestra blends

5 5.1 5.3 104 0.1

5 25 .0 24.8 99.2 0.2
5 50 .0 50.4 101 0.2
5 75 .0 75.4 101 0.4
5 90.0 89.8 99.8 0.2

Unheated cotton seed-oil olestra blends

5 5.1 5.2 102 0.8

5 25 .0 25.2 101 0.3
5 5 0 .0 50.6 101 0.2
5 74 .9 75.0 100 0.2
5 90 .0 90.2 100 0.3

H eated cottonseed-oil olestra blends

5 5.1 5.4 106 0.9

5 25 .0 25.2 102 0.5
5 50.1 50.5 101 0.2
5 74 .7 74.7 99.6 0.1
5 90.4 89.9 99.5 0.2

a N = Num ber of parallel determinations for each prepared sample.

6 Percent of olestra in olestra-triglycerlde blend, calculated as
micrograms of olestra added or found divided by microgram s of
M in utes total sample injected.
c Percent of olestra found divided by percent of olestra ad ded to
Figure 1. Chromatograms of check samples of lipid injected sample.
alone or of olestra-lipid blends: (A) olestra-free lipid
blank (RBD soybean oil), (B) soybean-oil olestra blended
with RBD soybean oil (35%, w/w), (C) unheated a sample by the total micrograms of sample injected and mul­
cottonseed-oil olestra blended with RBD soybean oil tiply by 1 0 0 .
(35%, w/w), (D) heated cottonseed-oil olestra blended
with RBD soybean oil (35%, w/w).
R e su lts a n d D iscu ssio n

into an autosampler vial and cap the vial. Prepare and mn check Method selectivity, linearity, accuracy, and precision were
samples daily. evaluated for lipid blends containing olestra.
(d) Preparation o f recovery samples.—Prepare as in (c) 5 Method selectivity.—Method selectivity was established by
olestra-lipid samples ranging from 5 to 90% olestra (w/w). Cal­ analyzing olestra-free lipid samples. Figure 1 shows chromato­
culate recovery of olestra as the percent of olestra detected di­ grams for the olestra-free lipid blank, RBD soybean oil (Fig­
vided by the percent of olestra injected. ure 1, A) and 3 olestra-containing check samples (Fig­
ures 1, B-D). Under the solvent elution conditions we used,
Calculations to Determine the Percent of Olestra in triglycerides, diglycerides, and monoglycerides elute together
Samples within 3.25 min from the starting time. Olestra elutes between
3.5 and 4.9 min. The chromatogram for the olestra-free lipid
First, generate a standard curve for each olestra standard blank (Figure 1, A) shows no interference in the region of oles­
used by graphing the area under the olestra peak (eluting be­ tra (retention time 3.5-4.9 min) and demonstrates method se­
tween 3.5 and 4.9 min after solvent is applied to the column) lectivity for blends composed of these 3 olestra standards in a
for each quantity of olestra injected (20-160 |ig, see Table 1). soybean oil matrix.
Second, perform a linear regression as in Table 1. Finally, de­ Linearity o f response.—Linearity of the evaporative light­
termine the micrograms of olestra in an actual sample by com­ scattering detector (ELSD) in generating the olestra peak area
paring the peak area for olestra in the sample with the curve was determined for each of the 3 olestra standards (Table 1).
generated by the olestra standard. To calculate the percent of Linearity was evaluated for each standard over the olestra con­
olestra in the sample, divide the micrograms of olestra found in centration range of 2-16 mg/mL. (Because the injection vol-
 T a l l m a d g e & L in : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6, N o . 6 ,1 9 9 3 1399

T able 3. C h e c k s a m p le variation Discussion o f sample analysis and method scope.—For ac­

Added, Found, Recovered, curate application of this method, the composition of the oles­
Day N 3 % b % b % c CV, % tra used to generate the external standard curve must be the
Soybean-oil olestra
same as the composition of olestra found in the lipid blend to
be analyzed. For example, the day-one heated cottonseed-oil
1 5 35.1 34.8 99 .2 0.2
olestra check sample (Table 3) was formulated with 35.3%
2 5 34.8 35.1 101 0.3
olestra. When the chromatographic results for this sample were
3 5 35 .0 35.3 101 0.6 analyzed using a heated cottonseed-oil standard, 35.4% olestra
4 5 34 .9 35.2 101 0.9 was found, yielding 100% recovery. However, if this same
sample peak area were analyzed using unheated cottonseed-oil
Unheated cottonseed-oil olestra olestra standard, then only 33.3% olestra would be detected,
resulting in only a 94.3% recovery (Table 1):
1 5 35.1 35.5 101 0.1
2 5 34.1 34.2 100 0.1 i.
( -) - constant
3 5 35.3 35.8 101 0.3 1 0 0 0
jig Olestra found = -
4 5 35.1 35.1 100 0.6 slope

H eated cottonseed-oil olestra

(3005 + 525.876)
1 5 35 .3 35.4 100 0.6 52.43458
2 5 34 .9 34.4 98.6 0.7
3 5 35 .0 34.2 97.7 0.1
4 5 34.8 34.2 98.3 0.5 = 67.3 jig

a N = N um ber of parallel determinations for each prepared sample.

b Percent of olestra in olestra-triglyceride blend, calculated as ... ug olestra found
micrograms of olestra added or found divided by micrograms of % Olestra = ------- .—------- x 100
total sample injected.
jig sample injected
c Percent of olestra found divided by percent of olestra added to
injected sample.
67.3
x 100
202.1
ume was 10 pL, the amount of olestra injected into the LC
sampler ranged from 20 to 160 jig; see Table 1). Each olestra
= 33.3%
standard tested demonstrated excellent linearity in the olestra
peak area; the correlation coefficients in all 3 cases were at least
0.9996. For the purpose of this study, predetermined blends of oles­
Method accuracy.—Method accuracy was demonstrated by tra and vegetable oil were prepared to demonstrate method ac­
olestra recovery in olestra-lipid blends varying in percent of curacy and precision. Proper validation of method selectivity
olestra from 5 to 90% olestra. Table 2 presents the method re­ must be performed with all olestra-free lipid matrixes tested by
covery data for each of the 3 olestra-lipid blends. Excellent running a chromatogram of each olestra-free lipid blank before
recoveries of olestra were observed, with an average recovery running the samples. The method we describe has been vali­
of 101%. For the purpose of this study, method accuracy was dated and routinely used to determine the percent of olestra in
evaluated over a wide range of percentages of olestra in lipid. a wide range of lipid sources. This range includes preformu­
However, for the day-to-day operation of this method, a check lated lipid blends and lipids obtained through solvent extrac­
tion of food.
sample should be mn with each day’s sample set. The check
However, two limitations of this method should be noted.
sample provides an independent evaluation of the accuracy of
First, blends of an unheated olestra with a previously heated
the external standard curve. Table 3 shows typical results for
triglyceride have exhibited less than optimum resolution be­
check samples mn in this study. Check samples with recoveries
tween the olestra and the triglyceride, because of the high lev­
outside 100 ± 3% suggest errors in the preparation or evalu­
els of triglyceride polymers that may co-elute with the olestra.
ation of the external standard curve and warrant re-evaluation. The percent of olestra in this type of sample has been success­
Precision.—System precision was evaluated by analysis of fully measured by generating the olestra standard curve in the
5 replicate injections of each olestra standard (Table 1) and blank heated triglyceride matrix instead of in the solvent, al­
each prepared recovery sample (Table 2). The coefficient of though this procedure results in some loss of method sensitivity
variation (CV) for system precision was less than 1.2%. and accuracy. Second, this method is not applicable to the
Method precision was evaluated by injecting 12 separately pre­ analysis of olestra in olestra-triglyceride blends that have been
pared check sample solutions (Table 3) which were run over 4 co-heated under typical frying conditions. Again, the chemical
different days and included all 3 types of olestra. Method pre­ formation of olestra-triglyceride polymers invalidates this ap­
cision, expressed as CV, was less than 1.2%. proach to the analysis.
1400 W e b s t e r & H e a r n e : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

With proper validation, this method provides a rapid, accu­ R e fe re n c e s

rate, and precise technique for determining the percent of oles-
tra in a wide range of lipid matrixes. This method has been (1 ) M a t t s o n , F .H ., & V o l p e n h e i n , R . A . ( 1 9 7 2 ) J. Lipid Res. 13,

successfully applied to the analysis of approximately 7000 3 2 5 -3 2 8

olestra-lipid blends over the past 5 years. (2 ) M a t t s o n , F . H ., & V o l p e n h e i n , R . A . ( 1 9 7 2 ) J. Nutr. 102, 1177—
1180

A c k n o w le d g m e n t s (3 ) S h u k l a , V .K . S . ( 1 9 8 8 ) Prog. Lipid Res. 27, 5 -3 8

(4 ) V o lp e n h e in , R .A . ( 1 9 8 5 ) S y n th e s is o f h ig h e r p o ly o l f a tty a c id
The authors thank the following contributors for method de­ p o ly e s te r s u s in g c a r b o n a te c a ta ly s ts . U S P a t e n t N o . 4 ,5 1 7 ,3 6 0

velopment and technical assistance: M. Kling and P. Hudson for (5 ) G a rd n e r, D .R ., S a n d e r s , R .A ., H e n ry , D .E ., T a llm a d g e , D .H .,

early investigations and D. Ewald for contemporary evaluations. & W h a r t o n , H .W . ( 1 9 9 2 ) J. Am. Oil Chem. Soc. 69, 4 9 9 -5 0 8

TECHNICAL COMMUNICATIONS

U se o f S ta n d a r d A d d itio n s T o D ia g n o s e M a tr ix E ffe c ts in th e

A to m ic A b s o r p tio n S p e c tr o p h o to m e tr ic A n a ly s is o f F e e d B le n d e d

f o r a R o x a rs o n e C o m b in a tio n S tu d y

G K. W
r e g o r y and L
e b s t e rA . H o r e t t a e a r n e

A.L. Laboratories, Inc., Animal Health Division, 400 State Street, Chicago Heights, IL 60411

R o x a r s o n e ( 3 -n it r o - 4 - h y d r o x y p h e n y la r s o n ic a c id ) is [Arsenic (Total) in Feeds: Colorimetric Method] (1). Often, the

a fe e d a d d it iv e c o m m o n l y u s e d in th e p o u lt r y i n d u s ­ choice of method is based more on economic rather than scien­
try. It s u s e i s a p p r o v e d a t le v e ls r a n g i n g fr o m tific considerations. The required initial investment for equip­
0 .00 2 5 to 0 .0 0 5 % in fe e d . T h i s a d d it iv e i s ro u t in e ly ment to run 971.47 and 957.22 is smaller than the purchase
a s s a y e d in i n d u s t r y b y A O A C O ffic ia l M e t h o d s price of the furnace atomic absorption spectrophotometer nec­
9 7 1 .4 7 ( R o x a r s o n e in F e e d s a n d P r e m ix e s : S p e c t r o ­ essary for 9 8 6 3 9 . Yet, the economics of performing the analy­
p h o t o m e t r ic M e t h o d ) a n d 9 8 6 .3 9 ( R o x a r s o n e in sis by 9 8 6 3 9 greatly improves with the increase in sample vol­
F e e d s : A t o m i c A b s o r p t i o n S p e c t r o p h o t o m e t r ic ume, because the lower cost and analysis time per sample offset
M e t h o d ). T h e a t o m ic a b s o r p t io n s p e c t r o p h o t o m e t ­
the higher start-up costs.
ric m e t h o d i s q u ite a d v a n t a g e o u s , b e c a u s e it is a c ­
The determination of roxarsone in feeds by atomic absorp­
c u r a t e a n d e ffic ie n t. O c c a s io n a lly , a n a l y s i s o f f e e d s
tion spectrophotometry (AAS) was first reported in 1982 (2)
and updated in 1986 (3) by George et al. The original graphite
b y t h i s m e t h o d y i e ld s q u e s t io n a b le r e s u lt s . T h i s
furnace method was revised to include a tantalum-coated cup
c o m m u n ic a t io n d e t a ils th e i n v e s t ig a t io n o f s u c h a n
to improve sensitivity, a common practice in AAS. The valida­
o c c u r r e n c e a n d th e a p p lic a t io n o f s t a n d a r d a d d i ­
tion of the roxarsone assay by AAS showed its accuracy to be
t i o n s t o d e te c t t h is p r o b le m in fu tu re a n a ly s e s .
superior to the colorimetric methods, with added bonuses of
reduced analysis time and improved efficiency.
Because of the advantages of furnace AAS, our laboratory
F eed containing roxarsone (3-nitro-4-hydroxyphenylar-
uses this technique as the first choice for the analysis of roxar­
sonic acid) can currently be analyzed by 3 AOAC Offi­
sone in feed. Until recently, we have found few matrix effects
cial Methods: 971.47 (Roxarsone in Feeds and Premixes:
in the analysis of feed. The presence of hemicellulose com­
Spectrophotometric Method), 9 8 6 3 9 (Roxarsone in Feeds:
pounds cited as a problem in 971.47 should present little prob­
Atomic Absorption Spectrophotometric Method), and 957.22
lem for the furnace AAS procedure. These organic materials
should bum off during the ashing stages of the temperature pro­
Received November 19, 1992. Accepted March S, 1993. gram.
 W e b s t e r & H e a r n e : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1401

In the analysis of feeds from a recent combination study, T able 1. A A S c o n d itio n s

however, the furnace AAS procedure repeatedly and reproduc- Param eter Setting
ibly yielded low results. A review of blending records and re­
analysis of premixes used showed no traceable error in feed Lam p current 10.0 mA
formulation. The UV-visible method of 971.47 was used as an M easurem ent mode Working curve
additional check on these analyses. This procedure inexplica­ W avelength 193.7 nm
bly yielded passing results from 3 independent laboratories. Signal mode BKG corrected
Because of this discrepancy, an in-house investigation into pos­ Slit 1.3 nm
sible problems with the furnace AAS procedure was initiated. Replicates Standards = 3, S am ples = 3
This report details our investigation of this discrepancy, as Atomizer Standard burner

well as our introduction of an approach to detect this problem S am ple blank Not run

in future analy ses. Cuvette Pyro

Calculation P eak height
Temp, control Optical tem p.
Experimental
Slicing height 10% (P.W. only)
Firings 0
Apparatus Cone, unit ppm
Injection volume 20 pL
___a
(a) Furnace AAS.—Z-8100 atomic absorption spectro­ Standards
photometer (Hitachi Instruments, Inc., Tokyo, Japan) equipped
Standards 1 - 7 were set as follows: 1 , 5 , 6 , and 7, 0.00; 2, 25.00;
with an Hitachi SSC-200 autosampler. 3 ,5 0 .0 0 ; and 4, 100.00.
(b) Arsenic lamp.—Arsenic hollow cathode lamp operated
at 10 mA (Hitachi, Part No. 001-6103).
(c) Graphite tube.— Pyro-coated platform furnace (Hita­ Instrument Conditions
chi, Part No. ANO-0028).
The manufacturer-recommended conditions for arsenic de­
terminations were used and are listed in Table 1.
Reagents
Temperature Program
(a) Romrsone.—Reference grade (A.L. Laboratories, Chi­
cago Heights, IL). The temperature program used was optimized for the Hita­
(b) Water.—Milli-Q (Millipore Corp., Milford, MA). chi Z-8100 and the use of the L’Vov platform. The program is
(c) Methanol.—HPLC grade (Mallinckrodt, Paris, KY). shown in Table 2.
(d) (NH4)2C 03.—ACS grade (Sigma Chemical Co., St.
Louis, MO). Results and Discussion
(e) H N 03.—ACS Grade (Curtin Matheson Scientific, Inc.,
Houston, TX). Analysis of Study Feed

Procedure The study feed samples were received in good condition

and promptly assayed. On 2 separate occasions the results of
Method 986.39 is used in our laboratory with some vali­ the analysis of groups of samples prepared at 45.4 g/50 ppm
dated optimizations. Because of the effectiveness of the Zee- roxarsone in feed did not fall within the AOAC limit (85—
man background correction available with the Hitachi Z-8100, 120%) of the claimed potency. Each time, the premix used was
the use of a control feed extract is not necessary. The use of a verified as being within its specified range. A random pick of 3
control feed extract was also abandoned because of the lack of samples at 45.4 g/50 ppm roxarsone in feed was made to inves­
an extract that can be used as a universal control to the variety tigate these apparent low results. These samples were assayed
of feed samples that arrive for testing. The coating procedure is an additional time, as well as sent to 3 independent laboratories
not used because the furnace was purchased already coated by
the manufacturer. The use of a L’Vov platform is added to en­ T able 2. Tem perature p ro gram
hance arsenic sensitivity. Start End G as flow,
The sample and standard concentrations are not changed, Step Stage tem p., °C tem p., °C Tim e, s mL/min
although the ammonium carbonate extract is gravity filtered
before dilution. This step is done to ensure that particulate mat­ 1 Dry 50 120 60 200

ter will not hinder the use of a volumetric pipet. Finally, to en­ 2 Dry 120 220 20 20 0

hance analytical accuracy, a full calibration curve from 0 to 1 0 0 3 Ash 220 1200 10 20 0

ppm in feed (0-57 ppb As) is used instead of the single point 4 Ash 1200 1200 30 20 0
5 Atomization 3000 3000 10 0
calibration cited in 986.39. It is common practice in analytical
6 Clean 3000 3000 4 20 0
spectroscopy to run a full calibration curve with the minimum
of a blank and 3 standards.
1402 W e b s t e r & H e a r n e : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

Table 3. Initial feed results, g/50 ppm roxarsone in feed3 Method of Standard Additions
Feed No. AA1 AA2 AA3 LAB1 LAB2 LAB3
In spectroscopy, the classical way to identify and compen­
sate for a matrix effect is by the method of standard additions
2 36.4 27 .6 35.3 43 .3 43 .9 43.0
(4). The Hitachi Z-8100 was configured to run the standard
7 35.6 36 .6 33.9 39.2 43 .9 40.3
additions procedure in-cup (5-8).
10 37 .4 36 .9 35.1 44.1 45 .9 42.2
The standard additions scheme was run against a single 50
a A A 1, AA2, and AA3 are the furnace AAS results. Lab 1, Lab 2, and ppm in-feed standard and not a full calibration (0 - 1 0 0 ppm in
Lab 3 are the independent laboratory results using 971.47. feed) to make efficient use of the instrument. Standard addi­
tions is quite time consuming, because it requires a full calibra­
for testing by 971.47. The laboratories used were all approved tion to be run with each of the samples. For this reason, only
contract laboratories with excellent reputations in the field. The the 50 ppm standard was used to expedite analysis. In essence,
results are fisted in Table 3. the standard additions method is using the same calibration in
Table 3 reveals a troubling discrepancy. With 971.47, based 986.39. A standard and house blend were also used for confir­
on UV-visible spectrophotometry, all 3 feed samples were mation of the standard additions analysis.
within the AOAC limit of the 45.4 g/50 ppm roxarsone in feed Standard additions (Table 5) confirmed that an atypical ma­
claim. The results of all 3 independent laboratories correlated trix effect was affecting the AAS analysis of these feed samples
well with each other. The in-house furnace AAS procedure, and yielded a result within the AOAC guidelines (85-120%),
based on 986.39, yielded results below 85% of the claim for as did the UV-visible assay by 971.47. The correct amount of
each sample. roxarsone was present in the feed extract. Atomization in the
During this analysis, the furnaces used with these study traditional AAS technique was hindered by an unknown con­
samples had to be replaced more frequently. This frequent re­ stituent in this feed. Thus, results found by 986.39 seem to be
placement is seemingly an additional indication that this feed low because of a matrix effect.
matrix was unusual. These furnaces had to be replaced at nearly
Simple Standard Additions Method
50% fewer bums per furnace than those used during the pre­
ceding calendar year. If the feed constituent that is causing the matrix effect is well
mixed throughout the feed lot, a series of feed samples blended
Method Recovery with analogous raw materials should be affected in the same
manner. If this is the case, simple standard additions can be
Although our furnace AAS procedure correlated well with applied to shorten the analysis times incurred with the standard
the UV-visible procedure for nearly a year, further investiga­ additions technique (5).
tion was warranted. A spike recovery study was done as an Simple standard additions involves applying the traditional
additional check on our procedure. A control feed sample was standard additions method to a selected member of a series.
spiked at 45.4 g/50 ppm roxarsone in feed with the roxarsone The calibration curve prepared with the standard additions run
standard. Arecovery of 99.86% was found. of this sample is then used for the remaining samples.
The use of simple standard additions, however, applies only
Arsenic Extracts for a series of samples that are of the same type and contain the
object element over a narrow concentration range (5). Analysis
To investigate if the feed samples were affected by ammo­ by the UV-visible assay confirmed that this group of samples
nium carbonate extraction, Feed 7 was ran again as before, as
met this criteria.
well as with a nitric acid digestion. The wet-acid digestion
The feed samples previously assayed by the traditional
would free the arsenic present in the feed sample for the fur­
standard additions were tested again by the simple standard
nace AAS analysis. The results from this analysis are shown in
additions procedure; results are fisted in Table 6 .
Table 4.
The assay yielded comparable results for both sample
Table 5. Results of method of standard additions for
preparations. In other words, the total amount of arsenic in both
representative feeds, g/50 ppm roxarsone in feed
preparations was equivalent. This result indicates that the ex­
traction of roxarsone from the feed was not the problem. Standard
Sam ple Claim Original AAS 971.47a additions

Table 4. Wet ash comparison, g/50 ppm roxarsone in

50 ppm
feed3 standard 45 .4 47.1

Method Result House blend 45 .4 — — 44 .4

Feed 2 4 5 .4 36.4 43.9 48 .3

(N H 4)3C 0 3 extraction 30.6 Feed 7 45 .4 35.6 43,9 4 0 .4

W et acid digestion 32.5 Feed 10 4 5 .4 37.4 45.9 5 2 .9

a Results for Feed 7. From Independent laboratory 2.

 W e b s t e r & H e a r n e : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3 1403

Table 6. Results of simple standard additions analysis control feed extract in 986.39 serves the same purpose as that
of representative feeds, g/50 ppm roxarsone in feed of standard additions. As stated earlier, such a control extract is
Simple seldom universally applicable to the various feed samples re­
standard ceived for analysis. However, standard additions can be used
Feed No. Claim Original AAS 9 7 1 .4 7 a additions
on any sample, and simple standard additions can be used on a
homologous group of samples.
2 45 .4 36.4 43.9 48 .7
Even more important is the fact that a control feed extract’s
7 4 5 .4 35.6 43.9 42 .2
useful contribution to background correction of today’s instru­
10 4 5 .4 37.4 45.9 45 .5
mentation is negligible. As an example of the inconsequential
3 From independent laboratory 2. effect a limited control feed extract has on background correc­
tion, it has been found that high alfalfa content in feed has af­
fected the performance of 986.39. A suitable control feed ex­
Simple standard additions not only confirmed that a matrix tract should have accounted for this effect. On the other hand,
effect is present but also revealed that the matrix effect is dis­ analysis of some dark-colored feed samples has been shown to
tributed throughout the series. This technique yielded results mn high with 971.47 while the AAS mn is not affected. Hence,
that correlated well with those of the UV-visible method of neither assay is totally free of matrix interferences.
971.47, with an analysis time shorter than that of conv entional Investigation of this matrix effect is continuing. A compari­
standard additions. son of the feed matrix from this series of samples to feed ma­
Accordingly, simple standard additions is now included as trixes of feed samples that have not been a problem is under­
a check on each series of feeds that yields low results across the way. Preliminary results indicate that neither phosphate,
ran when analyzed by traditional AAS, as was done in a sub­ sulfate, protein, nor fat content is the basis for this matrix effect.
sequent investigation with the study feeds. Table 7 lists the re­ Even though scientists try to develop procedures that uni­
sults for a series of feed samples received and thought to con­ versally address each sample investigated, few methods are to­
tain a matrix affecting constituent. Because of the standard tally free of interferences. It is important, however, to be able
additions results, the original AAS results are rejected in favor to determine when an assay is subject to matrix interferences.
of the UV-visible results of 971.47. The simple standard addi­ This problem has been resolved with 986.39 through standard
tions results confirmed the presence of a matrix effect. additions. With standard additions, the analyst can empirically
challenge questionable results.
The original AAS procedure for roxarsone in feed. 986.39,
calls for the use of a control feed extract, which is added to
standard solutions for the calibration mn. Actually the use of a Conclusion

Table 7. Application of simple standard additions The feed blended for this roxarsone combination study
method, g/50 ppm roxarsone in feed yielded low results when analyzed by AAS because of a matrix
effect. Standard additions was used not only to confirm the ma­
Standard
Feed Claim Original AAS additions 9 7 1 .4 7 a
trix effect but also to confirm that the feed blend was within the
AO AC limits for claimed potency.
Flouse blend 4 5 .4 47.1 NAb 48 .5 On the basis of this investigation, the AAS procedure used
37 45 .4 36.6 48.3 44 .2 in our laboratory has been enhanced by inclusion of a standard
38 45 .4 37.4 47.9 41 .4 additions procedure. With this ability to diagnose matrix inter­
39 4 5 .4 38.1 48.4 43.0 ferences, alternative procedures, such as 971.47 and 957.22,
40 45 .4 31.5 44.3 42 .4 can be chosen when a matrix effect is confirmed, and analysts
41 45 .4 37.2 47.6 45.8 are assured that the method with the highest analytical accu­
42 45 .4 36.4 45.4 41 .7 racy for the chosen feed sample will be used.
43 45 .4 39.4 50.4 44.5
44 45 .4 40.0 48.9 39.9
Acknowledgment
45 45.4 39.2 47.2 45.8
46 45 .4 41.4 52.2 43.3
47 45 .4 45.5 51.2 43.3
We thank the management and staff of A.L. Laboratories for
48 45 .4 38.6 47.7 43.9 their support, as well as J. Jefferson and K. Seace of Hitachi
49 0 <5 <5 <5 instruments for their technical assistance.
50 45 .4 41.5 45.2 44.9
51 45 .4 40.3 52.5 42 .7 References
52 45 .4 38.5 44.4 42 .4
53 45 .4 40.6 50.0 41 .4
(1) O f f ic ia l M e th o d s o f A n a ly s is (1990) 15th Ed., AOAC, Ar­
54 45 .4 33.7 42.5 43.0
lington, VA
3 From independent laboratory 2. (2) George, G.M., & Frahm, L.J. (1986) J . A s s o c . O ff. A n a l.

b NA = not applicable, different matrix. C hem . 69, 8 3 8 - 8 4 3

1404 W e b s t e r & H e a r n e : J o u r n a l O f A O A C In t e r n a t io n a l V o l . 7 6 , N o . 6 ,1 9 9 3

(3) George, G.M., Frahm, L.J., & McDonnell, J.P. (1982) J. As­ Hitachi Technical Data Note TDN AA-2, Hitachi Instru­
soc. Off. Anal. Chem. 65, 711-719 ments, Inc., Danbuty, CT
(4) Ingle, J.D., & Crouch, S.R. (1988) Spectrochemical Analysis (7) “Simultaneous Quantitative Analysis of Aluminum, Iron, Cop­
Prentice Hall, Englewood, NJ per, and Manganese in Photoresist,” Hitachi Technical Data
(5) Graphite Atomization Analysis Guidefor Polarized Zeeman Note TDN AA-3, Hitachi Instruments, Inc., Danbury, CT
Atomic Absorption Spectrophotometry (1988) Hitachi Ltd., (8 ) “Simultaneous Determination of Trace Quantities of Fe, Cu,
Tokyo, Japan 1988 Zn, and Cd in Sea Water,” Hitachi Technical Data Note TDN
(6 ) “Simultaneous Determination of Copper, Chromium, Manga­ AAA, Hitachi Instruments, Inc., Danbury, CT
nese, and Cadmium in Bovine Liver Using a Model Z-9000
Graphite Furnace Atomic Absorption Spectrophotometer,”
 A u t h o r In d e x : J o u r n a l O f A O A C In t e r n a t io n a l V o l . 7 6 N o . 6 ,1 9 9 3 1405

AUTHOR INDEX
A dam o, N .C ., see Scher, A. termining water-soluble vitamins Barnes, C J .,
A dam s, F.C-, see Lobinski, R. in foods, 682 report on drug residues in animal tis­
A kasaka, K., Ohrui, H., Meguro, H., A nderson, K .A ., & Isaacs, B. sues, 1 0 2
Suzuki, T., & Miwa, H. determination of selenium in feeds, Barnett, S.A., see Tanner, J.T.
fluorometric determination of total and premixes, supplements, and in­ Bartz, J.K., see Strong, A.B.
bond sulfite in wine by N-{9-ac- jectable solutions by hydride-gen­ Bauer, K ., see Lopez-Avila, V.
ridinyl)maleimide, 1385 erated inductively coupled
Beaulieu, N., Graham, S.J., & Lovering,
Al-H asani, S.M., Hlavac, J., & Carpen­ plasma atomic emission spec­
E.G.
ter, M.W. trometry, 910
LC determination of glyburide
rapid determination of cholesteral in A ndrew s, P., see Newsome, W.H.
(glibenclamide) and its related
single and multicomponent pre­ A ndrew s, W.H.,
compounds in raw materials, 962
pared foods, 902 report on food microbiology—non­
Beckert, W.F., see Lopez-Avila, V.
AI-Hasani, S.M ., Hlavac, J., & Hunts­
dairy, 154
Beljaars, P.R.,
man, M.A. see also June, G.A.
A ngyal, G .N ., see Anderson, E.M.;
report of international committee, 214
simple method for determination of Beljaars, P.R., van Dijk, R., & Brands,
Kim, H.S.
dietary fiber in frozen foods, 1014 A.
Anthony, G., see Montgomery, R.M.
A l-Show im an, S.S., see Al-Warthan,
Aoki, K., see Yamada, S. continuous flow and LC determination
AA.
Arm entia-A lvarez, A , Pena-Egido, of p-toluenesulfonamide in ice
Al-Tamrah, S .A , see Al-Warthan, A.A. cream: interlaboratory study, 570
M.J., & Garcia-Moreno, C.
Al-W arthan, A A ., Al-Showiman, S.S., Beljarrs, P., see DeVries, J.
improved method for determination of
Al-Tamrah, S A ., & BaOsman, A.A. Betker, W.R.,
sulfites in shrimp, 565
speetrophotometric determination of
Arm ishaw, P., & Millar, R.G. report on pesticide formulations: or-
boron in dates of some cultivars
comparison of gel permeation chroma­ ganophosphorus insecticides, 99
grown in Saudi Arabia, 601
tography, sweep codistillation, Bicsak, R.C.,
A lanko, T., see Wirtanen, G.
and florisil column adsorption comparison of Kjeldahl method for de­
Albert, R., see Horwitz, W. chromatography as sample termination of cmde protein in ce­
Albert, R.H., see Comeliussen, RE. cleanup techniques for the deter­ real grains and oilseeds with ge­
Alexander, T.G., mination of organochlorine resi­ neric combustion method:
report on drugs V, 104 dues in animal fats, 1317 collaborative study, 780
Ali, M .S., White, J.D., Bakowski, R.S., A tkinson, .1-, see Viggers, E.A Bidasee, K.R., see Yen, I.C.
Phillippo, E.T., & Ellis, R.L. Atwal, A , see Kaminski, J. Bishop, J.R.,
LC stability study of N-methylcar- report on food microbiology—dairy,
bamate pesticides in beef and 153
poultry liver tissues, 1309 Bailey, J.S., see Lancette, G.A. Bland, P., see Collier, R.H.
Ali, M .S., White, J.D., Bakowski, R.S., Baker, R., Blankenship, P.D., see Domer, J.W.
Stapleton, N.K., Williams, K.A., fast atom bombardment MS analysis Boese, J.L.,
Johnson, R.C., Phillippo, E.T., of base hydrolyzed y-lactone, 913 report on filth and extraneous materi­
Woods, R.W.,& Ellis, R.L. Bakow ski, R.S., see Ali, M.S. als in foods and drugs, 152
extension of an LC method for N- Balasubram anian, N ., see Kumar,
see also Lancette, G A.
methylcarbamate pesticides in B.S.M.
Boisen, F., see Wirtanen, G.
cattle, swine, and poultry liver, Bandler, R ., see Mislivec, P.B.
Boland, F.E.,
907 Banerjee, A .B., see Sanyal, A.K.
Allen, G., see Mislivec, P.B.
report on fruits and fruit products, 136
BaO sm an, A .A , see Al-Warthan, A.A.
Bontoyan, W., see Collier, R.H.
A m aguana, RJVL, see June, G.A. Baratta, E J .,
report on radioactivity, 149 Botsoglou, N.A , see Fletouris, D J.
Anderson, E.M ., Angyal, G.N.,
Weaver, C.M., Felkner, I.C., Wolf, Barbano, D .M ., see Smith, E.B. Bow e, S., see Dabeka, R.W.
W.R.,& Worthy, B.E. Barford, R.A., see Unruh, J. Boyer, K.W., see Comeliussen, PE.
potential application of LASER/mi- Bark, D J ., see DeVries, J. Bradley, R.L., Jr,
crobe bioassay technology for de­ Barker, S.A, see Lott, H.M. report on dairy chemistry, 106
1406 A u t h o r I n d e x : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 N o . 6 ,1 9 9 3

AUTHOR INDEX

Brady, M S ., Cancalon, P.F., method modification for LC determi­

report of Harvey W. Wiley scholarship oligosaccharide generation in acidic nation of thiamine, riboflavin, and
award committee, 2 2 0 sugar media, 584 pyridoxme in medical foods, 1276
Brands, A ., see Beljaars, P.R. Capar, S.G., Chatt, A., see Newsome, W.H.
Braselton, W.E., Jr, see Thompson, report on metals and other elements, C hen, G .N., see Cardwell, TJ.
H.C., Jr 142 Chen, J., &Gao, J.
Breitholtz-Em anuelsson, A., Olsen, Cardw ell, T J ., Cattrall, R.W., Chen, the Chinese Total Diet Study in 1990,
M., Oskarsson, A., Palminger, L, & G.N., Scollary, G.R., & Hamilton, part I: chemical contaminants,
Hull, K. I.C. 1193
ochratoxin A in cow’s milk and in hu­ determination of sulfur dioxide in the Chinese Total Diet Study in 1990,
man milk with corresponding hu­ wines and beverages by flow in­ part II: nutrients, 1206
man blood samples, 842 jection analysis with reductive Chiavarini, S., see Vitali, M.
Britton, P., see Strong, A.B. amperometric detection and elec­ C hiba, M ., see Singh, R.P.
Brodsky, M .H ., trolytic cleanup, 1389 Chichila, T .M P., & Gilvydis, D.M.
report of official methods board, 2 0 2 C arignan, G., Carrier, K., & Sved, S. determination of paraquat and diquat
Brousseau, R., see Page, B.D. assay of oxytetracycline residues in in low-moisture food crops using
B rucciani, J.C., see DeVries, J. salmon muscle by LC with UV silica column cleanup and LC
Bruelhart, M ., see Prodolliet, J. detection, 325 with UV detection, 1323
Brunner, C.A., Carleen, R., see Yperman, J. Cichow icz, S.M .,
report of archives committee, 204 Carpenter, D.E., see Lawrence, J.F. report on forensic sciences, 105
Bui, L.V., Carpenter, M.W ., see Al-Hasani, S.M. C hin, H .B., see Comeliussen, P.E.
LC determination of 6 sulfonamide Carrier, K., see Carignan, G. C hou, C .L., Uthe, J.F., & Guy, R.D.
residues in animal tissues using Carro, A., see Lorenzo, R.A. determination of free and bound Cd,
postcolumn derivatization, 966 C arson, M.C., Zn, Cu, Ag ions in lobster
Burini, G., simultaneous determination of multi­ (Homarus americanus) digestive
indirect flurometric determination of ple tetracycline residues in milk gland extracts by gel chromatog­
lactic acid in wine, 1017 using metal chelate affinity chro­ raphy followed by AAS and po-
Bushway, R J ., & Perkins, L.B. matography, 329 larography, 794
determination of oxyfluorfen in pesti­ Casais, C., see Lorenzo, R.A. Chowdhury, B., see Sanyal, A.K.
cide formulations by LC, 90 C astle, L., see Jickells, S.M. Christensen, R.R.,
Bushway, R J ., Paradis, L.R., Perkins, Cattrall, R.W., see Cardwell, TJ. report of executive director, 197
L.B., Fan, T.S., Young, B.E.S., & C ela, R ., see Lorenzo, R.A. C larke, M .A.,
Walser, P.E. report on sugars and sugar products,
Cepeda, A., see Rodríguez-Otero, J.L.
determination of methyl 2 -benzimida- 140
Chakravarty, S., Deb, M.K., & Mishra,
zolecarbamate in wine by com­ C leroux, C., see Dabeka, R.W.
R.K.
petitive inhibition enzyme immu­
hydroxyamidines as new extracting re­ C oates, S.,
noassay, 851
agents for spectrophotometric de­ report of AOAC Research Institute,
termination of cadmium with 4- 22 1

(2 -pyridylazo)naphthol in C ole, R J ., see Domer, J.W.

Cabanis, J.C ., see Lobinski, R. industrial effluents, coals, and fly C olem an, M R ., Wicker, A.L., & Mo­
C abras, R , Melis, M., & Spanedda, L. ash, 604 ran, J.W.
determination of cymiazole residues in C ham kasem , N ., Papathakis, M.L., & need for larger analytical samples with
honey by LC, 92 Lee, S.M. granulated feed additives, 945
C airns, T., Siegmund, E.G., & Rader, B. LC determination of abamectin in Collier, R.H ., Bland, R, Bontoyan, W.,
analysis of testosterone esters by tan­ fruits and vegetables, 691 Czerkowicz, T.J., Hofberg, A.H., Ne­
dem MS, 306 C hase, G.W., Jr, Landen, W.O., Jr, Soli­ meth, M.A., Ostapenko, H., Porter,
Cam pbell, H .M ., & Sauvé, F. mán, A.-G.M., & Eitenmiller, R.R. F., & Slahck, S.C.
LC determination of melengestrol ace­ LC analysis of niacin in fortified food report of committee on pesticide for­
tate in feeds, 1163 products, 390 mulations and disinfectants on
 A u t h o r I n d e x : J o u r n a l O f A O A C I n t e r n a t io n a l V o l . 7 6 N o . 6 ,1 9 9 3 1407

AUTHOR INDEX

recommendations for official Daft, J.L., D urán-M erás, I., Salinas, E, Muñoz
methods, 172 methyl bromide determination in se­ De La Peña, A., & Rosas, M.L.
Collins, P.G., see Newsome, W.H. lected foods by headspace tech­ simultaneous determination of flavor
Conacher, see Dabeka, R.W.; nique, 1083 enhancers inosine 5'-monophos-
Newsome, W.H.; Page, B.D. Danielson, J.W., phate and guanosine 5'-mono-
C onrad, S.M ., see Matusik, J.E. evaluation of microbial loads of Bacillus phosphate in food preparations by
C ook, R.F., see Comeliussen, P.E. subtilis spores on penicylinders, 355 derivative spectrophotometry,
C om eliussen, P.E., Boyer, K.W., Chin, D auphin, C., see Silvestre, M.P.C.
754
H.B., Cook, R.F., Fong, W.G., Hun­ Dyer, R.H.,
Deb, M .K ., see Chakravarty, S.
dley, H., Ripley, B.D., McCully, report on alcoholic beverages, 131
Decker, E.A., see Shantha, N.C.
K.A., & Albert, R.H.
Deutsch, M .J.,
report of committee on residues on rec­
report on vitamins and other nutrients,
ommendations for official meth­ Edberg, U.,
141
ods, 186 enzymatic determination of sulfite in
DeVries, J., Bark, D.J., Wood, R.,
Cox-Trout, C., see Newlon, N. foods: NMKL interlaboratoiy
Peake, A.A., Brucciani, J.C., Krinitz, study, 53
Crem isini, C., see Vitali, M.
B., Smith, R., Hargraves, W., Edgell, K.W., Erb, E.J., Wesselman,
Crisippi, T., Zini, G., & Fabbrini, R.
Beljarrs, P, Egelhofer, D., & Phillips, R.J., & Longbottom, J.E.
GC determination of benalaxyl resi­
J.G.
dues in different crops and water, GC/electron capture detection method
report of committee on foods II on rec­ for determination of chlorinated
650
ommendations for official meth­ acids in water: collaborative
C utrufelli, M .E., Mageau, R.P.,
ods, 182 study, 1098
Schwab, B., & Johnston, R.W.
Diachenko, G.W., see McNeal, T.P. see also Longbottom, J.E.
development of a multispecies identi­
fication field test by modified D ickerson, R., see Trotter, W.J. Eerola, S., Hinkkanen, R., Lindfors, E.,
agar-gel immunodiffusion, 1 0 2 2 Doerr, R.C., see Fiddler, W. & Hirvi, T.
Czerkowicz, T J ., Donnelly, J.R., LC determination of biogenic amines
report on disinfectants, 97 report on cooperative studies, 168 in dry sausages, 575
Egelhofer, D., see DeVries, J.
see also Collier, R.H. Donnelly, J.R ., Sovocool, G.W., & Ti­
tus, R.L. Eichelberger, J.W., see Tang, P.H.
Eilers, P.P., see Matusik, J.E.
structures and environmental signifi­
Eitenmiller, R .R ., see Chase, G.W., Jr
cance of heptachlor epoxide iso­
D ’A m ato, A., Semeraro, I., & Biechi, C. Elkins, E.R.,
mers, 1092
simultaneous determination of linuron AOAC International— 1991-1992:
D om er, J.W ., Blankenship, P.D., &
and trifluralin residues in carrots the year in review, 4
Cole, R.J.
and their pulp by LC and GC, 657 report of Harvey W. Wiley Committee,
performance of 2 immunochemical as­ 219
D ’Aoust, J.-Y., Sewell, A.M., & Greco,
says in the analysis of peanuts for
P. see also Ali, M.S.
aflatoxin at 37 field laboratories,
detection of Salmonella in dry foods Ellis, R .L., see Ali, M.S.
637
using refrigerated pre-enrichment Em ara, L.H.,
D om er, J.W., & Cole, R.J. rapid and accurate method for determi­
and enrichment broth cultures: in­
terlaboratory study, 814 variability amoung peanut subsamples nation of niclosamide released
D abeka, R.W., McKenzie, A.D.,
prepared for aflatoxin analysis from molluscicidal formulations,
Lacroix, G.M.A., Cleroux, C , Bowe, with 4 mills, 983 847
S., Graham, R.A., Conacher, H.B.S., Dufour, A.P., Engebretsen, D .R., see Lawrence, J.F.
& Verdier, P. report on water microbiology, 160 Erb, E J ., see Edgell, K.W.; Longbot­
survey of arsenic in total diet food D uke, P.D., Weiss, G., & MacDonald, tom, J.E.
composites and estimation of die­ A.
tary intake of arsenic by Canadian LC method for the simultaneous
adults and children, 14 analysis of sulfadimethoxine and Fabbrini, R., see Crisippi, T.
D acunha, A.R., see Khunachak, A. ormetroprim in animal feeds, 320 Falbo-Nelson, M.T., see Feldsine, P.T.
1408 A uthor In d e x : Journal Of A O AC International V ol . 76 N o. 6 ,1 9 9 3

AUTHOR INDEX

Fan, T.S., see Bushway, R.J. Gales, P.W., & Schwiesow, M. Hanna, G.M., & Lau-Cam, C.A.
Fazio, T., automated beer analyzer method for stability-indicating proton nuclear
report on food additives, 108 determination of alcohol conten- magnetic resonance spectro­
Feldsine, P.T., Falbo-Nelson, M., tand original gravity of beer: sum­ scopic assay method for
Hustead, D.L. mary of cobaoborative study, 918 furoseiride in tablets and injec­
substrate supporting disc method for Gao, J., see Chen, J. tions, 526
confirmed detection of total cob- Garcia-Moreno, C., see Armentia-Al- Hargraves, W., see DeVries, J.
forms and E. coli in all foods: col­ varez, A. Harnett, M,, see Haggett, T.O.R.
laborative study, 988 Gardner, A., Harris, L,, & Humber, J.
Feldsine, P.T., Falbo-Nelson, M.T., & report of regional sections committee, autoMicrobic system for biochemical
Hustead, D.L. 219 identification of Listeria species
polyclonal enzyme immunoassay Geisler, C A ., see Thompson, H.C., Jr isolated from foods: collaborative
method for detection of motile Gilbert, J., see Jickells, S.M.; study, 822
and non-motile Salmonella in Lawrence, J.F. Hashimoto, H., see Tabata, S.
foods: comparative study, 694 Gilbertson, T.J., see Montgomery, Hasselberger, M.L.,
Felkner, I.C., see Anderson, E.M. R.M. assay of oxytetracycline in animal feed
by LC and microbiological plate
Fiddler, W., & Doerr, R.C. Gilvydis, D.M., see Chichila, T.M.P.
assay, 39
GC/chemiluminescence detection Ginn, R.E., see Packard, V.S.
(thermal energy analyzer/nitrogen
Hayakawa, J., see Yamada, S.
Glaze, L.E.,
mode) method for the determina­
Heame, LA., see Webster, G.K.
extraction of light filth from oriental
tion of dibutylamine in hams, 578 Hennig, B., see Shantha, N.C.
fish products containing spice:
Hermida, M., see Rodríguez-Otero,
Firestone, D., collaborative study, 44
J.L.
report on fats and oils, 133 Gliksman, J.E.,
Higgins, D.L., & Robison, B J.
Fleming, J.R., see Smith, E.B. citrate insoluble P205: problems with
comparison of MICRO-ID Listeria
Fletouris, DJ., Mantis, A.J., Botsoglou, ammoniated fertilizers, 920
method with conventional bio­
N.A., & Papageorgious, G.E. Gooch, E.G., chemical methods for identifica­
rapid determination of tryptophan in determination of traces of silicone de- tion of Listeria isolated from food
intact proteins by derivative spec­ foamer in fruit juices by solvent and environmental samples: col­
trophotometry, 1168 extraction/atomic absorption laborative study, 831
Flowers, R.S., see Lancette, G.A. spectroscopy, 581 Hinkkanen, R., see Eerola, S.
Fong, W.G., see Comeliussen, P.E. Gopal, M., see Mukheijee, I. Hipp, S., see Kirchoefer, R.D.
Foster, M.L., see Lawrence, J.F. Graham, C.R., see Lancette, G.A. Hirsch, J.,
Franco, C., see Rodríguez-Otero, J.L. Graham, R A ., see Dabeka, R.W. report of quality assurance committee,
Franklin, C., Graham, SJ., see Beaulieu, N. 215
report of editorial board, 202 Greco, P., see D’Aoust, J.-Y. Hirvi, T., see Eerola, S.
Fujii, Y., see Yamada, S. Gunderson, E.L., see Yess, N.J. Hitchins, A.D.,
Fukaya, M., see Yamada, S. Guy, R.D., see Chou, C.L. report on cosmetic microbiology, 150
Fuleki, T., & Pelayo, E. Hlavac, J., see Al-Hasani, S.M.
sugars, alcohols, hydroxymethylfur- Ho, J.S., see Tang, PH.
fural in authentic varietal and Hofberg, A.H., see Collier, R.H.
commercial grape juices, 59 Haggett, T.O.R., Matheson, A.R., & Holland, D.C., Munns, R.K., Roybal,
Fuleki, T., Pelayo, E., & Palabay, R.B. Harnett, M. J.E., Hurlbut, J.A., & Long, A.R.
carboxylic acid composistion of partial automation of microbiological simultaneous determination of xy-
authentic varietal and commercial assays for vitamins, 1280 lazine and its major metabolite,
grape juices, 591 Hamilton, I.C., see Cardwell, T.J. 2 ,6-dimethylaniline, in bovine
Hammack, T.S., see June, G.A. and swine kidney by LC, 720
Hamon, M., see Silvestre, M.P.C. Hollifield, H.C., see McNeal, T.P.
Hanks, A.R., Hopes, T.M., see Montgomery, R.M.
Gagnon, J., see Page, B.D. report on CIPAC studies, 95 Hopper, M.L., see King, J.W.
 A uthor In d e x : Journal O? AOAC International V ol . 76 No. 6,1993 1409

AUTHOR INDEX

Horwitz, W., Albert, R., & Nesheim, S. wave susceptors and its migration King, J.W., Hopper, M.L., Luchtefeld,
reliability of mycotoxin assays—an into foods on cooking, 760 R.G., Taylor, S.L., & Orton, W.L.
update, 461 Jiusheng, Z., see Sundaram, K.M.S. optimization of experimental condi­
Hsu, M.C., & Huang, W.F. Johnson, F.J., tions for the supercritical carbon
LC determination of dicloxacillin report of nominating committee, 218 dioxide extraction of pesticide
preparations: interlaboratory Johnson, R.C., see Ali, M.S. residues from grains, 857
study, 1143 Johnston, R.W., see Cutrufelli, M.E. see also Snyder, J.M.
Huang, W.F., see Hsu, M.C. Jones, T.L., see Lopez-Avila, V. King-Brink, M., & Sebranek, J.G.
Hult, K,, see Breitholtz-Emanuelsson, Jorhem, L., combustion method for determination
A. determination of metals in foodstuffs of crude protein in meat and meat
Humber, J., see Harris, L. by AAS after dry ashing: NMKL byproducts: collaborative study,
Hundley, H., see Comeliussen, RE. interlaboratory study of lead cad­ 787
Hungerford,, J.M., mium, zinc, copper, iron, chro­ Kirchoefer, R.D., & Hipp, S.
report on seafood toxins and seafood mium, and nickel, 798 niacin I: dissolution profiles of sus­
products, 120 June, G.A., Sherrod, P.S., Amaguana, tained-release niacin products by
Huntsman, M.A., see Al-Hasani, S.M. R.M., Andrews, W.H., & Hammack, automated and manual proce­
Hurlbut, J.A., see Holland, D.C. T.S. dures, 394
Hustead, D.L., see Feldsine, P.T. effectiveness of the Bacteriological Kmostak, S., & Kurtz, D.A.
Analytical Manual culture rapid determination of supplemental
method for the recovery of vitamin E acetate in feed pre­
Ichinoe, M., see Tanaka, T. Shigella sonnei from selected mixes by capillary GC, 735
Ibe, A., see Tabata, S. foods, 1240 Koch, H., see Mooser, A.E.
Iida, M., see Tabata, S. Kojima, S., see Sasaki, K.
Ikebuchi, H., see Tanaka, T. Kotula, K.L., see White, C.R.
Imerman, P.M., Kamimura, H., see Tabata, S. Krause, R.T., see Sack, C.A.
pH method for determination of choli­ Kaminski, J., Atwal, A.S., & Ma- Krinitz, B., see DeVries, J.
nesterase in whole blood: collabo­ hadevan, S. Krueger, D.A.,
rative study, 899 determination of formaldehyde in report on flavors, 107
Isaacs, B., see Anderson, K.A. fresh and retail milk by LC, 1010 Sample preparation bias in carbon sta­
Ito, Y., see Nakamura, Y. Kane, P.F., ble isotope ratio analysis of fruit
Izquierdo-Pulido, M.L., Vidal-Carou, report on fertilizers and agricultural juices and sweeteners, 418
M.C., & Marine-Font, A. liming materials, 163 Kumar, B.SJVL, & Balasubramanian,
termination of biogenic amines in see also Newlon, N. N.
beers and their raw materials by Katz, S.E., see Martin, A.; Thompson, pararosaniline as a new chromogen for
ion-pair LC with postcolumn H.C., Jr extractive spectrophotmetric de­
derivatization, 1027 Kawamura, N., see Yamada, S. termination of trace amounts of
Kendall, D.C., see Sack, C.A. hydrogen sulfide in air, 730
Khunachak, A., Dacunha, A.R., & Kuntom, A., see Tan, Y. A.
Jackson, J., see Shaikh, B. Stout, S.J. Kurtz, D.A., see Kmostak, S.
Jain, A.V., LC determination of moxidectin resi­
rapid test for semiquantitative determi­ dues in cattle tissues and confir­
nation of nitrate in forages: inter­ mation in cattle fat by LC/MS, Lacroix, G.M.A., see Dabeka, R.W.
laboratory study, 948 1230 Lancette, G.A., Bailey, J.S., Boese,
Jenkins, T.F., see Strong, A.B. Kijima, K., see Sasaki, K. J.L., Flowers, R.S., Sharpe, A.N.,
Jensen, T.L., Kim, H.S., & Angyal, G. Scott, V.N., Graham, C.R.,
report on nutrients in the soils, 164 comparison of LC method to AOAC Thompson, S.S., McClure, F.D.
Jickells, S.M., Philo, M.R., Gilbert, J., microbiological method for deter­ report of committee on microbiology
& Castle, L. mination of L-tryptophan in tab­ and extraneous materials on rec­
GC/MS determination of benzene in lets and capsules, 414 ommendations for official meth­
nonstick cookware and micro­ Kim, R., see Lopez-Avila, V. ods, 190
1410 A uthor In dex : Journal Of AO AC International V ol . 76 N o . 6 ,1 9 9 3

AUTHOR INDEX

Landen, W.O., Jr, see Chase, G.W., Jr ylene thiourea in finished drink­ MacDonald, A., see Duke, RD.
Lane, L.G., see Strong, A.B. ing waters: collaborative study, Mageau, R.P., see Cutrufelli, M.E.
Lane, R.H., 1113 Mahadevan, S., see Kaminski, J.
report on cereals and cereal products, see also Edgell, K.W.; Strong, A.B. Malone, W., see Sertl, D.
131 Lopez-Avila, V., Mansilla, A.E., Muñoz De La Peña,
Lati, E., see Silvestre, M.P.C. report on organics in water, 169 A.M., & Salinas, F.
Latimer, G.W., Jr, see also Longbottom, J.E. semiautomatic determination of fura-
comparison of Texas liquid sampler nic aldehydes in food and phar­
Lopez-Avila, V., Bauer, K., Milanes, J.,
with Missouri bottle for sampling maceutical samples by a stopped-
& Beckert, W.F.
liquid fertilizers: collaborative flow flow injection analysis
study, 749 evaluation of Soxtec extraction proce­
dure for extracting organic com­ method, 1255
see also Thompson, H.C., Jr
pounds from soils and sediments, Mantis, A.J., see Fletouris, D.J.
Lau-Cam, C.A., see Hanna, G.M.
864 Marine-Font, A., see Izquierdo-Pulido,
Lawrence, J.F., Park, D.L., Carpenter,
M.L.
D. E., Engebretsen, D.R., Foster, Lopez-Avila, V., & Jones, T.L.
M.L., Gilbert, J., McNeal, J.E., interlaboratory study of a ther- Markus, J.R., & Sherma, J.
Wallin, H., Wekell, M.M., Nelsen, mospray-LC/MS method for se­ method I: LC determination of tia-
T.C., & Pfeiffer, S. lected V-methyl carbamites, N- mulin hydrogen fumarate in feed
report of committee on Foods I on rec­ premixes, 444
methyl carbamoyloximes, and
ommendations for official meth­ substituted urea pesticides, 1329 method A: LC determination of tia-
ods, 179 mulin hydrogen fumarate in tia-
Lopez-Avila, V., Young, R., Kim, R., &
Leadbetter, M.G., & Matusik, J.E. mulin-poly(vinyl-chloride) for­
Beckert, W.F.
LC determination and LC-ther- mulations, 447
interlaboratory evaluation of an off­
mospray MS confirmation of ni- method III: LC determinations of tia-
line supercritical fluid extrac-
carbazin in chicken tissues: inter­ mulin hydrogen fumarate in com­
tion/infrared spectrometric
laboratory study, 420 plete swine meal feeds, 449
method for determination of pe­
Lee, S.M., see Chamkasem, N. method IV: GC determinations of tia-
troleum hydrocarbons in solid
Leoni, V., see Vitali, M. mulin residues in swine liver, 451
matrixes, 555
Leyes, G., see Strong, A.B. method V: GC/mass spectrometric
Li, M., Nelson, D.L., & Spoms, P. Lorenzo, R.A., Carro, A., Rubi, E., confirmation of 8-hydroxymu-
determination of menthol in honey by Casais, C., & Cela, R. tilin, a tiamulin metabolite, in
GC, 1289 selective determination of methyl mer­ swine liver extracts, 459
Lin, P.Y.T., see Tallmadge, D.H. cury in biological samples by Markus, J.R., O’Rangers, J.
Lindfors, E., see Eerola, S. means of programmed tempera­ animal drugs, 443
Lindsay, C.W., ture GC, 608 Martin, A., & Katz, S.E.
calculation of juice content in diluted Lott, H.M., & Barker, S.A. rapid determination of Listeria mono­
fruit juice beverage, 424 extraction and GC screening of 14 cytogenes in foods using a resusi-
Lobinski, R., Szpunar-Obiska, J., chlorinated pesticides in crayfish tation/selection/kit system detec­
Adams, F.C., Teissedre, P.-L., & (Procambarus clarkii) hepa- tion, 632
Cabanis, J.C. topancreas, 663 Matheson, A.R., see Haggett, T.O.R.
speciation analysis of organolead com­
matrix solid-phase dispersion extrac­ Matusik, J.E., Eilers, P.P., Waldron,
pounds in wine by capillary
tion and GC screening of 14 chlo­ E.M., Conrad, S.M., & Sphon, J.A.
GC/microwave-induced-plasma
rinated pesticides in oysters confirmation of identities of propylene
emission spectrometry, 1262
(Crassostrea virginica), 67 and ethylene glycols in anchovies
Long, A.R., see Holland, D.C.; Rupp,
H.S. Louis, J.B., see Roinestad, K.S. by tandem MS, 1344
Longbottom, J.E., Edgell, K.W., Erb, Lovering, E.G., see Beaulieu, N. see also Leadbetter, M.G.
E. J., & Lopez-Avila, V. Low, N.H., see Wudrich, G.G. McClure, ED., see Lancette, G.A.
GC/nitrogen-phosphorus detection Luchtefeld, R.G., see King, J.W. McCully, K.A., see Comeliussen, P.E.
method for determination of eth- Lynch, J.M., see Smith, E.B. McKenzie, A.D., see Dabeka, R. W.
 A uthor In de x : Journal O f A O AC International V ol . 76 N o. 6 ,1 9 9 3 1411

AUTHOR INDEX

McMahon, B.M., T.M., Torchia, M.G., Presser, M., & extraction of light filth from tofu: col­
report on organohalogen pesticides, Wright, W.W. laborative study, 50
147 report of committee on drugs and re­ Nakatani, A., se e Sumitani, H.
McNeal, J.E., see Lawrence, J.F. lated topics on recommendations Nelsen, T.C., see Lawrence, J.F.
McNeal, T.P., & Hollifield, H.C. for official methods, 175 Nelson, D.L., see Li, M.
determination of volatile chemicals re­ Mooser, A.E., & Koch, H. Nemeth, M.A., see Collier, R.H.
leased from microwave-heat- confirmatory method for sulfonamide Nesheim, S., se e Horwitz, W.
suscept food packaging, 1268 residues in animal tissues by GC Newlon, N.,
McNeal, T.P., Nyman, P.J., Diachenko, and pulsed positive ion-negative evaluation of new flame photometric
G.W., & Hollifield, H.C. ion-chemical ionization mass instrumentation to meet require­
survey of benzene in foods by using spectrometry, 976 ments of AOAC official method
headspace concentration tech­ Moran, J.W., see Coleman, M.R. for potassium in fertilizers, 1182
niques and capillary GC, 1213 Mori, B., see Page, B.D. Newlon, N., Cox-Trout, C., Kane, P.
McSheffirey, S., see Wudrich, G.G. Mortimer, R.D., & Weber, D.F. evaluation of current AOAC fertilizer
Meguro, H., se e Akasaka, K. determination of residual imazethapyr sample preparation requirements,
Melis, M., see Cabras, P. in soybeans by GC/nitrogen- 1174
Mendes, A.S., phosphoms detection, 377 Newsome, W.H.,
international perspectives in certifica­ Mountford, M.K., se e Tanner, J.T. report on organonitrogen pesticides,
tion and accreditation, 1 Mowrey, D.H., se e Thompson, H.C., Jr 148
Merola, G.V., se e Thompson R.H. Mukherjee, I., & Gopal, M. see a lso Yeung, J.M.
Metzger, D.T., see Packard, V.S. organochlorine pesticide residues in Newsome, W.H., & Andrews, P.
Mikami, E., se e Yamada, S. dairy milk in and around Delhi, organochlorine pesticides and PCB
Milanes, J., se e Lopez-Avila, V. 283 congeners in commercial fish
Millar. R.G., see Armishaw, P. Mullens, J., se e Yperman, J. from the Great Lakes, 707
Miller, L.S., see Ramstad, T. Mullin, WJ., see Wolynetz, M.S. Newsome, W.H., Andrews, P., Con-
Mishra, R.K., see Chakravarty, S. Mulvaney, T.R., acher, H.B.S., Rao, R.R., & Chatt, A.
Mislivec, P.B., Bandler, R., & Allen, G. report on processed vegetable prod­ total organochlorine content of fish
incidence of fungi in shared-use cos­ ucts, 138 from the Great Lakes, 703
metics available to the public, 430 Munns, R.K., see Holland, D.C.; Rupp, Newsome, W.H., Yeung, J.M., & Col­
Miwa, H., se e Akasaka, K. H.S. lins, P.G.
Moats, W.A., Muñoz De La Peña, A.M., se e Durán- development of enzyme immunoassay
determination of cephapirin and de- Merás, I.; Mansilla, A.E. for captan and its degradation
sacetylcephapirin in milk using Munson, A., product tetrahydrophthalimide in
automated LC cleanup and ion­ report of secretary/treasurer and the fi­ foods, 381
pairing LC, 535 nance committee, 199 Newton, J.M.,
Moats, W.A., see White, C.R. report on nonalcoholic beverages, 138
Mongeau, R., & Brassard, R. Ng, L.L.,
enzymatic-gravimetric determination Nakamura, Y., Tonogai, Y, Tsumura, report on drags IV, 104
in food of dietary fiber as sum of Y, & Ito, Y. Ng, W.-Y., & Wong, S.-K.
insoluble and soluble fiber frac­ analysis of pyrethroid residues in adsorptive stripping determination of
tions: summary of collaborative vegetables, fruits, grains, beans, sulfamethazine in milk, 540
study, 923 and green tea leaves; applications Niemann, R.A.,
Montgomery R.M., to pyrethroid residue monitoring determination of formetanate HC1 in
report on cosmetics, 101 studies, 1348 selected fruits by coupled-column
report on diagnostics and test kits, 101 Nakashima, MJ., cation exchange LC, 1362
report on drugs 1 ,102 extraction of light filth from oriental Nishima, T., see Tabata, S.
report on drugs III, 103 sauces containing soy sauce, Nixon, G., see Page, B.D.
Montgomery, R.M., O’Rangers, J., An­ thickeners, and spices: collabora­ Nott, R., se e Sundaram, K.M.S.
thony, G., Gilbertson, T.J., Hopes, tive study, 47 Nyman, PJ., see McNeal, T.P.
1412 A uthor In d e x : Journal O f A O AC International V ol . 76 N o . 6 ,1 9 9 3

AUTHOR INDEX

O’Rangers, J., see Markus, J.R.; Paradis, L.R., see Bushway, R.J. drug by dynamic headspace GC,
Montgomery, R.M. Parham, T.M., Jr, see Thompson, H.C 313
Ogawa, M., Ohtsubo, T., Tsuda, S., & Jr Rao, R.R., see Newsome, W.H.
Tsuji, K. Park, DA., see Lawrence, J.F. Reeves, J.B. Ill,
simplified method to measure suspen- Parks, O.W., influence of water on the near infrared
sibility of water-dispersible pow­ depletion of the monoamino metabo­ soectra of model compounds, 741
der: use of microanalytical tech­ lites of zoalene during frozen stor­ Reggers, G., see Yperman, J.
niques to reduce wastewater, 83 age, 698 Reimann, L.M., see Thompson, H.C.,
Ohrui, H., see Akasaka, K. Patil, V.B., & Shingare, M.S. Jr
Ohtsubo, T., see Ogawa, M. TLC detection of oiganophosphoms Renger, B.,
Oles, RJ., insecticides containing a ni- quantitative planar chromatography as a
fractional factorial design approach for trophenyl group, 1394 tool in pharmaceutical analysis, 7
optimizing analytical methods, Pena-Egido, MA, see Armentia-Al- Rhodig, L.L., see Thompson, H.C., Jr
615 varez, A. Riedel, G.R., see Page, B D .
Olsen, ML, see Breitholtz-Emanuelsson, Peake, A.A., see DeVries, J. Ripley, B.D., see Comeliussen, P.E.
A. Pelayo, E., see Fuleki, T. Robison, BA, see Higgins, DA.
Orton, W.L., see King, J.W. Pelayo, E., see Fuleki, T. Rodríguez-Otero, J.L., Hermida, M.,
Oskarsson, A , see Breitholtz- Perkins, L.B., see Bushway, R.J. Cepeda, A., & Franco, C.
Emanuelsson, A. Pfeiffer, S., see Lawrence, J.F. total bacteria count in raw milk using
Ostapenko, H., see Collier, R.H. Phillippo, E.T., see Ali, M.S. BactoScan 8000, 838
Phillips, J.G., see DeVries, J. Roinestad, K.S., Louis, J.B., & Rosen,
Philo, M.R., see Jickells, S.M. J.D.
Packard, V.S., Ginn, R.E., & Metzger, Placentia, A M , determination of pesticides in indoor
D.T. report on drug- and device-related mi­ air and dust, 1121
automated technique for sampling crobiology, 151 Ronk, RJ.,
milk from farm bulk tanks: col­ Porter, F., see Collier, R.H., report of bylaws committee, 205
laborative study, 297 Presser, M., see Montgomery, R.M. Roos, AH., see Tuinstra, L.G.M.Th.
Page, B.D., Prodolliet, J., & Bruelhart, M. Rosas, M.L., see Durán-Merás, I.
LC method for determination of 9 phe­ determination of acesulfam-K in Rosen, J J)., see Roinestad, K.S.
nolic antioxidants in butter oil: foods, 268 Ross, P.F.,
collaborative study, 765 determination of aspartame and its ma­ report on veterinary analytical toxicol­
Page, B.D., Conacher, H.B.S., Salmi- jor decomposition products in ogy, 165
nen, J., Nixon, G.R., Riedel, G., foods, 275 Rowe, L.D., see Snyder, J.M.
Mori, B., Gagnon, J., & Brousseau, Prosky, L., Roy, R.R., see Yess, N.J.
R. report on dietary fiber, 132 Roybal, J.E., see Holland, D.C.
survey of bottled drinking water sold Purcell, S., see Viggers, E.A Rubí, E., see Lorenzo, R.A.
in Canada, part 2: selected volatile Pylipiw, H.M., Jr, Rupp, H.S., Munns, R.K., & Long, A.R.
organic compounds, 26 rapid GC method for multiresidue simultaneous determination of nitro-
Page, S.W., screening of fruits and vegetables furazone and furazolidone in
report on meetings, symposia, and for organochlorine and organo- shrimp (Penaeus vannamei) mus­
educational programs committee, phosphate pesticides, 1369 cle tissue by LC with UV detec­
216 tion, 1235
report on plant toxins, 119
Paisley S.D., see Zygmunt, L.C. Rader, B., see Cairns, T.
Palabay, R.B., see Fuleki, T. Ragheb, H.S„ Sack, C .A , Kendall, D.C., & Krause,
Palminger, L, see Breitholtz- report on antibiotics in feeds, 161 R.T.
Emanuelsson, A. Ramstad, T., Miller, L.S., & Thomas, determination of methylene chloride
Papageorgious, G.E., see Fletouris, V.N. acceptability and its “purifica­
D J. determination of residual ethylene ox­ tion” for ethylenethiourea meth­
Papathakis, M.L., see Chamkasem, N. ide in spectinomycin HC1 bulk odology, 1146
 A uthor In de x : Journal O f A O AC International V ol . 76 N o. 6 ,1 9 9 3 1413

AUTHOR INDEX

Salinas, E , see Durân-Merâs, L; Man- Scott, V.N., see Lancette, G.A. Smallidge, R.L.,
silla, A.E. Sellers, C., see Strong, A.B. report on chugs in feeds, 161
Salisbury, C.D.C., Sertl, I)., & Malone, W. Smedley, M.D., see Weber, J.D.
modified method for determination of LC method for determination of iodine Smith, E.,
ivermectin residues in animal tis­ in milk: collaborative study, 711 report on drugs II, 102
sues, 1149 Sewell, AJVL, see D’Aoust, J.-Y. Smith, E.B., Barbano, D.M., & Lynch,
Salminen, J., see Page, B.D. Shaikh, B., & Jackson, J. J.M.
Salvatore, MJ., & Katz, S.E. improved LC determination of neomy­ a quantitative linearity evaluation
solubility of antibiotics used in animal cin B in bovine kidney, 543 method for infrared milk analyz­
feeds in selected ers, 1300
Shantha, N.C., Decker, E.A., & Hen-
solvents, 952 nig, B. Smith, E.B., Barbano, D.M., Lynch,
unified procedure for determination of J.M., & Fleming, J.R.
comparison of methylation methods
antibiotics in animal feeds, 514
for the quantitations of conjugated performance of homogenizers in infra­
Sanyal, A.K., Chowdhury, B & Baner- linoleic acid isomers, 644 red milk analyzers, a survey, 1033
jee, A.B.
Sharpe, A.N., see Lancette, G.A. Smith, R., see DeVries, J.
rapid TLA microestimation of elemen­
Shephard, G.S., see Thiel, P.G. Snell, RJ*.,
tal sulfur: apphcation to complex
Sherrod, P.S., see June, G.A. GC determination of cyclohexanone
sulpher ointments, 1152
Shingare, M.S., see Patil, V.B. leached from hemodialysis tub­
Sapp, R.E., & Davidson, S.
Shoemaker, D., see Strong, A.B.; Sims, ing, 1127
determination of roxarsone in feeds us­
A. Snell, R.P.,
ing solid-phase extraction and LC
Siegmund, E.G., see Cairns, T. solid-phase extraction and LC deter­
with ultraviolet detection, 956
Silvestre, M.P.C., Latí, E., Dauphin, C., mination of monophthalates and
Sasaki, K., Kijima, K., Takeda, M., &
& Hamon, M. phthalide extracted from solution
Kojima, S.
cuprimetric assay of casein hydro­ administration sets, 531
specific determination of ethylene ox­
lysates, 1295 Snyder, JJVL, King, J.W., Rowe, L.D.,
ide and ethylene chlorohydrin in
Sims, A., & Shoemaker, D. & Woemer, J.A.
cosmetics and polyoxyethlated
surfactants by GC with electron simultaneous LC determination of supercritical fluid extraction of poultry
capture detection, 292 thiamine and riboflavin in se­ tissues containing incunred pesti­
lected foods, 1156 cide residues, 893
Sauvé, E, see Campbell, H.M.
Sawada, J., see Tanaka, T. Singh, R.P., & Chiba, M. Soderberg, I).,
Sawyer, Leon D., reversed-phase LC method for simul­ report on meat, poultry, and meat and
report on multiresidue methods, 144 taneous determination of poultry products, 111
Scher, A., & Adamo, N.C. methyl[ 1-[(butylamino)carbonyl Soliman, A.-G.M., see Chase, G.W., Jr
LC determination of 2-chloro-4-ni- ]-1 H-benzimidazol-2-yl]carbam Sovocool, G.W., see Donnelly, J.R.
troaniline, 2-naphtol, and 2,4- ate (Benomyl) and methyl 1H- Spanedda, L., see Cabras, P.
dinitroaniline in D&C Red No. benzimidazol-2-ylcarbamate Sphon, J A ., see Matusik, J.E.
36,287 (MBC) in wettable powder (WP) Spoms, P., see Li, M.
Schmitt, TJVL, formulations, 1187 Stahr, HJVL, see Sun, T.S.C.
LC assay of bentazon formulations, Sjoberg, A-M., see Wirtanen, G. Stapleton, N.IC, see Ali, M.S.
387 Skurikhin, I.M., Stephens, 1VL, see Strong, A.B.
Schwab, B., see Cutrufelli, M.E. methods of analysis for toxic elements Stout, SJ ., see Khunachak, A.
Schwartz, D.P., see Unruh, J. in foods, part IV: general method Strong, AB., Jenkins, T.F., Lane, L.G.,
Schwiesow, M., see Gales, P.W. of ashing for determination of Leyes, G., Longbottom, J., Sellers,
Scollary, G.R., see Cardwell, T.J. toxic elements, 257 C., Stephens, M., Bartz, J.K., Britton,
Scott, P.M., new metrological characteristics of P., & Shoemaker, D.
report of joint mycotoxin committee analytical methods of analysis report of committee on environmental
on recommendations for official used for safety control of food and quality on recommendations for
methods, 220 environmental materials, 262 official methods, 195
report on mycotoxins, 112 Slahck, S.C., see Collier, R.H. Suekane, S., see Sumitani, H.
1414 A uthor In d e x : Journal O f AO AC International V ol . 76 N o. 6 ,1 9 9 3

AUTHOR INDEX

Sumitani, H., Suekane, S., Nakatani, and vitamins D and K: U.S. FDA- Tomkins, D.F.,
A., & Tatsuka, K. Infant Formula Council: collabo­ report on pesticide formulations: her­
inductively coupled plasma atomic rative study bicides, 98
emission spectrometric determi­ Tanner, J.T., Barnett, S.A., & Mount- report on pesticide formulations: or-
nation of tin in canned food, 1374 ford, M.K. ganohdogen insecticides; other
Sun, T.S.C., & Stahr, H.M. analysis of milk-based infant formula, insecticides, synergists, and repel­
evaluation and application of a biolu- phase V: vitamins A and E, folic lents, 98
minescent bacterial genotoxicity acid, and pantothenic acid, Food Tonogai, Y., see Nakamura, Y.
test, 893 and Drag Administration-Infant Torchia, M.G., see Montgomery, R.M.
Sundaram, K.M.S., Jiusheng, Z., & Formula Council: collaborative Trotter, W J., & Dickerson, R.
Nott, R. study, 399 pesticide residues in composited milk
LC determination of RH-5992, an ec- Tatsuka, K., see Sumitani, H. collected through the U.S. Pas­
dysone agonist, in some forestry Taylor, S.L., see King, J.W. teurized Milk Network, 1220
matrixes, 668 Tsuda, S., see Ogawa, M.
Teissedre, P.-L., see Lobinski, R.
Suzuki, T., see Akasaka, K. Tsuji, K., see Ogawa, M.
Sved, S., see Carignan, G. Terao, T., see Tanaka, T.
Terao, C., see Yamada, S. Tsumura, Y., see Nakamura, Y.
Sydenham, E.W., see Thiel, RG. Tuinstra, L.G.M.Th., Roos, A.H., &
Szpunar-Obiska, J., see Lobinski. R. Teshima, R., see Tanaka, T.
van Trijp, J.M.P.
Thiel, P.G., Sydenham, E.W., Shephard,
LC determination of aflatoxin M l in
G.S., & van Schalkwyk, D.J.
milk powder using immunoaf-
Tabata, S., Kamimura, H., Ibe, A., reproducibility characteristics of a LC fmity columns for cleanup: inter­
Hashimoto, H., Bda, M., Tamura, Y., method for the determination of laboratory study, 1248
& Nishima, T. fumonisins B1 and B2 in com:
aflatoxin contamination in foods and IUPAC collaborative study, 361
foodstuffs in Tokyo: 1986-1990, Thomas, V.N., see Ramstad, T.
32 Thompson, H.C., Jr, Braselton, W.E., Unruh, J., Schwartz, D.P., & Barford,
Takeda, ML, see Sasaki, K. Jr, Mowrey, D.H., Walsh, M.C., Rei- R.A.
Tallmadge, D.H., & Lin, P.Y.T. mann, L.M., Parham, T.M., Jr, quantitation of sulfamethazine in pork
LC method for determining the per­ Rhodig, L.L., Geisler, C.A., Latimer, tissue by TLC, 335
cent of olestra in lipid samples, G.W., & Katz, S.E. Uthe, J.F., see Chou, C.L.
1396 report of committee on feeds, fertiliz­
Tamura, Y., see Tabata, S. ers, and related topics on recom­
Tan, Y.A., & Kuntom, A. mendations for official methods,
GC determination of hydrocarbons in 193
van Dijk, R., see Beljaars, PR.
crude palm kernel oil, 371 Van Poucke, L., see Yperman, J.
Thompson, JJ.,
Tanaka, T., Teshima, R., Ikebuchi, H., van Schalkwyk, DJ., see Thiel, P.G.
determination of ultratrace levels of
Sawada, J., Terao, T., & Ichinoe, M. van Trijp, see Tuinstra,
lead in infant formula by isotope
sensitive determination of zearalenone L.G.M.Th.
dilution inductively coupled
and a-zearalenol in barley and Verdier, P., see Dabeka, R.W.
plasma MS, 1378
Job’s tears by LC with floures- Vidal-Carou, M.C., see Izquierdo-
Thompson, M., & Wood, R.
cence detection, 1006 Pulido, M.L.
international harmonized protocol for
Tang, P.H., Ho, J.S., Eichelberger, J.W. Viggers, E.A., Atkinson, J., & Purcell,
proficiency testing of (chemical)
determination of organic pollutants inn S.
analytical laboratories, 926
reagent water by liquid—solid ex­ reporting ongoing results of interlabo­
traction followed by supercritical Thompson, R.H., & Merola, G.V. ratory comparison programs, 620
fluid elution, 72 a simplified alternative to the AOAC Vitali, M., Leoni, V, Chiavarini, S., &
Tanner, J.T., Barnett, S.A., & Mount- official method for cholesterol in Cremisini, C.
ford, M.K. multicomponent foods, 1057 determination of 2-ethyl-1-hexanol as
analysis of milk-based infant formula, Thompson, S.S., see Lancette, G.A. contaminant in drinking water,
phase IV: iodine, linoleic acid, Titus, R.L., see Donnelly, J.R. 1133
 A uthor In de x : Journal O f AOAC International V ol . 76 N o. 6 ,1 9 9 3 1415

AUTHOR INDEX

Waldron, E.M., see Matusik, J.E. and herbs: a BCR collaborative Yess, NJ.,
Wallin, H., se e Lawrence, J.F. study, 674 U.S. Food and Drug Administration
Walser, P.E., see Bushway, R.J. Woerner, J.A., see Snyder, J.M. survey of methyl mercury in
Walsh, M.C., see Thompson, H.C., Jr Wolf, W.R., see Anderson, E.M. canned tuna, 36
Weaver, CM., se e Anderson, E.M. Wolynetz, M.S., & Mullin, W.J. Yess, NJ., Gunderson, E.L., & Roy,
Weber, D.E, see Mortimer, R.D. factors affecting the precision of die­ R.R.
Weber, J.D., & Smedley, M.D. tary fiber measurements in pota­
U.S. Food and Drug Administration
LC method for determination of sul­ toes, 508
monitoring of pesticide residues
famethazine residues in milk: col­ Wong, S.-K., see Ng, W.-Y.
laborative study, 725 in infant foods and adult foods
Wood, R., see DeVries, J.
Webster, G.K., & Heame, L.A. eaten by infants/children, 492
Woodbury, J.E.,
use of standard addition to diagnose Yeung, JJVf., & Newsome, W.H.
report on spices and other condiments,
matrix effects in the use of AAS to 139 survey of total tetrahydrophthalimide
analyze feed blended for a roxar- Woods, R.W., se e Ali, M.S. in baby foods using both ELISA
sone combination study, 1400 Worobey, B.L. and GC/MS: a comparative study,
Weeks, W.W., LC method for determination of diquat 1225
report on tobacco, 165 and paraquat herbicides in pota­ see a lso Newsome, W.H.
Weiss, G., see Duke, RD. toes: collaborative study, 881 Young, B.E.S., se e Bushway, R.J.
Wekell, M.M., see Lawrence, J.F. Worthy, B.E., see Anderson, E.M. Young, R., se e Lopez-Avila, V.
Wesselman, RJ., see Edgell, K.W. Wright, W.W., see Montgomery, R.H. Yperman, J., Carleer, R., Reggers, G.,
Wheeler, M.M.,
Wudrich, G.G., McSheffrey, S., & Mullens, J., & Van Poucke, L.
discriminating between adult mandi­
Low, N.H. automation of potentiometric meas­
bles of O ryzaephilu s surinam en-
LC detection of a variety of inexpen­ urements: determination of water-
sis and O. m erca to r using setal
sive sweeteners added to pure or­ extractable sodium in bread using
brush length, 941
ange juice, 342
White, C.R., Moats, W.A., & Kotula, a sodium ion selective electrode
K.L. with minimum sample prepara­
optimization of LC method for deter­ tion, 1138
mination of oxytetracycline, tetra­ Yamada, S., Fujii, Y., Mikami, E.,
cycline, and chlortetracycline in Kawamura, N., Hayakawa, J., Aoki,
milk, 549 K., Fukaya, M., & Terao, C.
White, JJ)., se e Ali, M.S. small-scale survey of organotin com­
Wicker, A.L., se e Coleman, M.R. pounds in household commodi­ Zini, G., see Crisippi, T.
Williams, K.A., see Ah, M.S. ties, 436 Zygmunt, L.C., & Paisley S.D.
Wirtanen, G., Sjoberg, A. M., Boisen, Yen, LC., & Bidasee, K.R. enzymatic method for determination
F., & Alanko, T. LC determinations of aflatoxins in ani­ of (1 >3)(l->,4)-beta-D-glucans in
microbiological screening method for mal feeds and feed components, grains and cereals: collaborative
indication of irradiation of spices 366 study, 1069
1416 S ubject In d e x : Journal O f AOAC International V ol . 76, N o. 6 ,1 9 9 3

S U B J E C T IN D E X

Abamectin in fruits and vegetables, AOAC silicone defoamers in fruit juices, 581
LC method, 691 archives committee report, 204 feed blended for a roxarsone combina­
Acesulfam-K in foods, LC method, 268 bylaws committee report, 205 tion study, standard additions
Aflatoxins Bylaws, 206 method to diagnose matrix ef­
contamination in foods and foodstuffs editorial board report, 202 fects, '400
in Tokyo, 32 Executive Director’s report, 197 Automated methods
in feeds and feed components, LC finance committee report, 199 AutoMicrobic system for biochemical
method, 366 general referee reports, 95 identification of Listeria species
M, in milk powder, LC method with Harvey W. Wiley committee report, isolated from foods: collaborative
immunoaffinity column cleanup, 219 study, 822
interlaboratory study, 1248 Harvey W. Wiley scholarship award Bactoscan 8000, bacterial count (total)
in peanuts, performance of 2 immuno­ committee report, 220 in raw milk using, 838
chemical assays at 37 field labora­ international committee report, 214 automated beer analyzer method, alco­
tories, 637 joint mycotoxin committee report, 220 hol content and original gravity of
in peanuts, variability among subsam­ keynote address, 1 beer: summary of collaborative
ples prepared with four mills, 983 laboratory quality assurance commit­ study, 918
Agricultural liming materials, see tee report, 215 dissolution profiles of sustained-re-
Fertilizers and agricultural liming meetings, symposia, and educational lease niacin products, 394
materials programs committee report, 216 microbiological assays for vitamins,
Air membership committee report, 217 1280
hydrogen sulfide (trace) in, nominating committee report, 218 potentiometric measurements, water-
pararosaniline for extractive spec- officers and committees, 1993, 223 extractable sodium in bread, so­
trophotometric determination, official methods board report, 202 dium ion selective electrode with
730 president’s address, 4 minimum sample preparation,
recommendations for official methods, 1138
indoor, pesticide residues in, 1121
committee reports, 172 sampling milk from farm bulk tanks:
Alcoholic beverages
regional section committee report, 219 collaborative study, 297
Methods Committee report, 182
Research Institute report, 221
referee report, 131
Secretaiy/Treasurer’s report, 199
see also Distilled liquors; Wines
Wiley Award address, 1993,
Alcohols
Arsenic, in total diet food composites Bacillus subtilis spores on penicylin-
in authentic varietal and commercial
and dietary intakes by Canadian ders, microbial loads, 355
grape juice, 59
adults and children, 14 Bacteria
in beer, automated analyzer method:
Artificial sweeteners count (total) in raw milk using Bactos­
summary of collaborative study,
acesulfam-K in foods, LC method, 268 can 8000, 838
918
aspartame and major decomposition microbial loads of Bacillus subtilis
Analytical mycology and microscopy, products in food, LC method, 275 spores on penicylinders, 355
Methods Committee report, 190
Ashing, toxic elements in foods, ana­ Shigella sonnei recovery from selected
Antibiotics in feeds lytical methods, part IV, 257 foods, effectiveness of the Bacte­
Methods Committee report, 193 Aspartame and major decomposition riological Analytical Manual cul­
referee report, 161 products in food, LC method, 275 ture method for, 1240
oxytetracycline assay by LC and mi­ Atomic absorption spectroscopy Baking powders and baking chemi­
crobiological plate assay, 39 cadmium, zinc, copper, and silver ions cals, Changes in Methods, no
solubility in selected solvents, 952 in lobster (Homarus americanus) changes
unified procedure for determination of, digestive gland extracts, 794 Barley, zearalenone and a-zearalenone
514 lead, cadmium, zinc, copper, iron, in, LC method, 1006
Antioxidants, phenolic (9), in butter chromium, and nickel in food­ Beans, pyrethroids residues in, applica­
oil, LC method: collaborative study, stuffs after dry ashing: NMKL in­ tions to pyrethroid residue monitor­
765 terlaboratory study, 798 ing studies, 1348
 S u b je c t I n d e x : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1417

SUBJECT INDEX

Benalaxyl residues in crops and mation of potentiometric measure­ protein (crude) in grains, Kjeldahl
water, GC method, 650 ments, 1138 method vs generic combustion
Bentazon, LC assay of formulations, method: collaborativestudy,780
387 Chemical contaminants monitoring
Benzene aflatoxins in foods and foodstuffs in
in foods, survey, headspace GC Tokyo, 32
method. 1213 Cacao bean and its products, Changes arsenic in total diet food composites
in nonstick cookware and microwave inMethods, no changes and intakes by Canadian adults
susceptors and itsmigration into Cadmium and children, 14
foods on cooking, GC/MS in foodstuffs after dry ashing, A AS benzene in foods, survey, headspace
method, 760 method: NMKL interlaboratory GC method, 1213
(l->3)(l->4)-beta-D-glucansingrains study,798 bottleddrinkingwatersoldinCanada,
and cereals: collaborative study, in lobster (H o m a ru s a m e ric a n u s) di­ volatileorganiccompounds in,26
1069 gestive gland extracts, gel chro­ ChineseTotalDietStudyin 1990,part
Beverages, sulfurdioxidein,flowinjec­ matography followed by AAS 1,1193
tion analysis with reductive am- and polarography,794 methyl mercury in canned tuna,FDA
perometricdetectionand electrolytic with 4-(2-pyridylazo)naphthol in in­ survey,36
cleanup, 1389 dustrial effluents, coal, and fly organochlorine pesticide residues in
Bioassay, LASER/Microbe technology, ash, hydroxyamidines as extract­ dairymilk, Delhi,India,283
water-solublevitaminsinfoods,682 ingreagentsforspectrophotomet- organochlorine pesticides and PCB
Biogenic amines ricdetermination,604 congeners in commercial fish
in beers and theirraw materials, ion- Canada from theGreatLakes,707
pairLC withpostcolumnderivati- arsenic in total diet food composites organochlorine (total) content of fish
zation, 1027 and intakes by adults and chil­ from theGreatLakes,703
indry sausages,LC method, 575 dren, 14 pesticideresidues incomposited milk
Biological samples bottleddrinkingwatersoldin,volatile collected through the U.S. Pas­
blood (whole), cholinesterase in, pH organiccompounds in,26 teurizedMilk Network, 1220
method: collaborativestudy,899 Captan and its degradation product pesticide residues in infant foods and
bovineandswinekidney,xylazineand tetrahydrothalimide in foods, adult foods eaten by infants/chil-
metabolite2 ,6-dimethylanilinein, ELISA, 381 dren,FDA monitoringof,492
simultaneous determination by Carbon isotope analysis, sample tetrahydrophthalimide in baby foods,
LC, 720 preparation bias in analysis of fruit ELISA vsGC/MS, 1225
cattle, swine, and poultry liver, N - juicesand sweeteners,418 Chinese Total Diet Study in 1990
methylcarbamate in,LC method, Carboxylic acid composition of grape partI,chemicalcontaminants, 1193
907 juice, 591 partII,nutrients, 1206
cow’smilk and human milk with cor­ Carrots and their pulp, linurinand tri- Chlorinated acids in water, GC/elec-
responding human blood sam­ fluralinresidues,simultaneousdeter­ tron capture detection method: col­
ples,ochratoxinA in,842 minationby LC and GC, 657 laborativestudy, 1098
methyl mercury in,programmed tem­ Casein hydrolysates, cuprimetric as­ 2-Chloro-4-nitroaniline in D&C Red
peratureGC method, 608 say, 1295 No. 36, LC method, 287
swine liver, tiamulin residues in, GC Cephapirin and desacetylcephapirin Chlortetracycline in milk, LC method,
method,451 in milk, automated LC cleanup and 549
Books in Brief, 19A, 37A, 55A, 105A, ion-pairingLC, 535 Chocolate and cacao products, Meth­
161A Cereals and cereal products odsCommittee report, 183
Boron in dates of Saudi Arabian culti- Changes inMethods, 253 Cholesterol
vars, 601 Methods Committee report, 182 in multicomponent foods, alternative
Bread, water-extractablesodium in,so­ refereereport, 131 toAOAC officialmethod, 1057
dium ion selective electrode with (1—>3X1—>4)-beta-D-glucans in: col­ inpreparedfoods,rapiddetermination,
minimum sample preparation, auto­ laborativestudy, 1069 902
1418 S u b je c t I n d e x : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 , 1993

SUBJECT INDEX

Cholinesterase in whole blood, pH cholinesterase in whole blood, pH summary, dietary fiber in foods, as
method: collaborativestudy,899 method, 899 sum ofinsolubleandsolublefrac­
Chromatography diquatandparaquatherbicidesinpota­ tions, enzymatic-gravimetric
Florisilcolumn adsorption,cleanupof toes,LC method, 881 method,923
organochlorines in animal fats, enzymatic method for (1 >-3)(l >4)- Color additives
1317 beta-D-glucansingrainsandcere­ Changes inMethods, no changes
gel, followed by AAS and polarogra- als,1069 Methods Committee report, 179
phy, cadmium, zinc, copper, and ethylenethiourea in finished drinking 2-chloro-4-nitroaniline, 2-naphthol,
silver ions in lobster (H om arus water, GC/nitrogen-phosphorus and 2,4-dinitroanifine in D&C
am ericanus) digestive gland ex­ detectionmethod, 1113 Red No. 36,LC method, 287
tracts,794 extraction of light filth from oriental Combustion method
gel permeation, cleanup of organo­ fishproductscontainingspice,44
chlorinesinanimal fats,1317 protein(crude)inmeatandmeatprod­
extraction of tight filth from oriental ucts:collaborativestudy,780
quantitativeplanar, inpharmaceutical sauces containing soy sauce,
analysis,7 vsKjeldahlmethod, protein(crude)in
thickeners,and spices,47 cerealgrainsandoilseed:collabo­
metal chelate affinity, simultaneous
extractionoftightfilthfrom tofu,50 rativestudy,780
determinationofmultipletetracy­
clineresidues,329 FDA-Infant Formula Council, vita­ Cooperative studies
mins A andE,folicacid,andpan­ Methods Committee report, 195
Chromium, infoodstuffsafterdry ash­
ing,AAS method: NMKL interlabo­ tothenicacidinmilk-basedinfant
refereereport, 168
formula,399
ratorystudy,798 Copper
CIPAC studies, refereereport,95 FDA-Infant Formula Council, iodide,
tinoleicacid,and vitamins D and in foodstuffs after dry ashing, AAS
Cleanup procedures method: NMKL interlaboratory
K in milk-based infant formulas,
gel permeation chromatography vs study,798
1042
sweep codistillation vs Florisil in lobster (H om arus am ericanus) di­
column adsorption chromatogra­ iodineinmilk,LC method, 711
IUPAC, fumonisinsB| andB2 incom, gestive gland extracts, gel chro­
phy, organochlorines in animal matography followed by AAS
fats,1317 reproducibility of LC method,
361 and polarography, 794
Coffee and tea
MICRO-ED L isteria Method vs con­ Corn, fumonisins B[ and B2 in,repro­
Changes inMethods, no changes ducibilityofLC method:IUPAC col­
green tea leaves, pyrethroids residues ventional biochemical methods
for identification of L isteria iso­ laborativestudy,361
in;applicationstopyrethroidresi­ Corn syrup, high fmctose com syrup
duemonitoringstudies, 1348 lated from food and environ­
mental samples,831 addedtopureorangejuice,LC detec­
Coliform organisms tion,342
in foods, substrate supporting disc phenolic antioxidants(9)inbutteroil,
LC method, 765 Cosmetic microbiology, refereereport,
method for confirmed detection:
protein(crude)incerealgrainsandoil­ 150
collaborativestudy,988
Collaborative studies seeds,Kjeldahlmethodvsgeneric Cosmetics
automated technique for sampling combustionmethod,780 Changes inMethods, 252
milkfromfarmbulktanks,297 protein(crude)inmeatand meatprod­ Methods Committee report, 177
AutoMicrobic systemforbiochemical ucts,combustion method, 780 refereereport, 101
identification of L isteria species substrate supporting disc method for ethylene oxide and ethylene chlorhy-
isolatedfromfoods,822 confirmed detection ofcotiforms drin in polyoxyethylated surfac­
BCR, microbiological method forin­ (total) and E sch erichia co li in tants and, GC method with elec­
dication of irradiation of spices foods,988 troncapturedetection,forspecific
and herbs: BCR collaborative sulfamethazine in milk, LC method, determination,292
study,674 725 shared-use, available to the public,
chlorinated acids in water, GC/elec- summary, alcoholcontentandoriginal fungiin,430
tron capture detection method, gravity of beer, automated ana­ Cuprimetric assay, casein hydro­
1098 lyzermethod,918 lysates, 1295
 S u b je c t In d e x : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1419

SUBJECT INDEX

Cured meats, hams, dibutylamine in, 2,6-Dimethylaniline and xylazine in sulfur ointments, application of mi­
GC/chemiluminescence detection bovine and swine kidney, simulta­ croestimation TLC method for
method,578 neous determinationby LC, 720 sulfur(elemental)to, 1152
Cyclohexanone leached from Dimetridazole testosteroneesters,tandem MS meth­
hemodialysis tubing, GC method, 2,4-Dinitroaniline in D&C Red No. ods,306
1127 36,LC method, 287 Drug packaging
Cymiazole in honey, LC method, 92 Diquat solution administration sets, mono-
inlow-moisturefood crops,silicacol­ phthalatesandphthalideextracted
umn cleanupandLC withUV de­ from, solid-phase extraction and
tection, 1323 LC determination,531
inpotatoes,LC method: collaborative Drugs and feed additives in animal
Dairy chemistry tissues
study,881
Methods Committeereport, 179 Changes inMethods, 253
Disc method
refereereport, 106 Methods Committee report, 177
substratesupporting,forconfirmedde­
Dairy microbiology, MethodsCommit­ tectionofcoliforms(total)andE s­ refereereport, 102
teereport, 191 c h e ric h ia c o li infoods:collabora­ cephapirin and desacetylcephapirin in
Dairy products tivestudy,988 milk, automated LC cleanup and
Changes inMethods, 253 Disinfectants ion-pairingLC, 535
ice cream, p-toluenesulfonamide in, Changes inMethods, no changes 8-hydroxymutilin in swine fiver ex­
continuous flow and LC determi­ Methods Committee report, 174 tracts,GC/MS confirmation,459
nation,interlaboratorystudy,570 refereereport,97 ivermectin,modified method, 1149
s e e a ls o Milks and milkproducts
Distilled liquors, Changes inMethods,
moxidectin residues in cattle tissues
Desacetylcephapirin in milk, auto­ no changes and cattle fat, LC/MS method,
mated LC cleanup and ion-pairing 1230
Drinking water
LC, 535 neomycin B in bovine kidney, LC
bottled, sold in Canada, volatile or­
Diagnostics and test kits method, 543
ganiccompounds in,26
Methods Committeereport, 177 nicarbazin in chicken tissues,LC de­
ethyl-1-hexanol as contaminant in, termination with LC-ther-
refereereport, 101 1133
Dibuty lamine in hams, GC/chemilu­ mospray M C confirmation, inter­
finished, ethylenethiourea in, GC/ni- laboratorystudy,420
minescence detectionmethod,578 trogen-phosphorus detection nitrofurazole and furazolidone in
Dicloxacillin preparations, LC method: collaborativestudy, 1113 shrimp (P e n a e u s v a n n a m e i) mus­
method: interlaboratorystudy, 1143 D ru g - and device-related m icrobiol­ cle tissue, simultaneous determi­
Dietary fiber ogy nation by LC with UV detection,
Methods Committee report, 183 Methods Committee report, 191 1235
refereereport, 132 refereereport, 151 oxytetracycline, tetracycline, and
infoods, as sum ofinsoluble and sol­ D ru g form ulations chlortetracycline in milk, LC
uble fractions, enzymatic-gravi­ dicloxacillin,LC method: interlabora­ method, 549
metric method: summary of col­ torystudy, 1143 oxytetracycline in salmon muscle, as­
laborativestudy,923 ethylene oxide (residual) in specti- sayby LC withUV detection,325
infrozenfoods, simplemethod, 1014 nomycin HC1 bulkdrug,dynamic sulfamethazine in milk, adsorptive-
in potatoes, precision of measure­ headspace GC method, 313 strippingdetermination,540
ments,509 furosemide in tablets and injections, sulfamethazine in milk, LC method:
Diets stability-indicating proton NMR collaborativestudy,725
arsenicintakesbyCanadianadultsand spectroscopicassay method, 526 sulfamethazine inpork tissue,quanti­
children, 14 glyburide (Glibenclamide)and related tationby TLC, 335
ChineseTotalDietStudy in1990,part compounds in raw materials, LC sulfonamide residues (6) in, LC
I,chemical contaminants, 1193 method, 962 method, 966
ChineseTotalDietStudy in 1990,part quantitative planar chromatography sulfonamide residues (6) in, GC/MS
13,nutrients, 1206 foranalysisof,7 confirmation,976
1420 S u b je c t I n d e x : J o u r n a l O f A O AC In t e r n a t io n a l V o l . 76, N o . 6 ,1 9 9 3

SUBJECT INDEX

tetracyclines in milk, multiresidue Ecdysone agonist RH-5992 in forestry Extraction

metal chelateaffinitychromatog­ matrixes, LC method, 668 hydroxyamidines as extracting re­
raphy method, 329 Eggs and egg products, Changes in agentsforspectrophotometricde­
tiamulin hydrogen fumarate in com­ Methods, no changes termination, cadmium with 4-(2-
pleteswinemeal feeds,449 Environmental analyses pyridylazo)naphthol in industrial
tiamulin hydrogen fumarate in tia­ pararosaniline for extractive spectro- effluents,coal,andflyash,604
mulin—poly(vinyl chloride) for­ photometric determination of oflightfilthfromorientalfishproducts
mulations,LC method,447 tracehydrogen sulfideinair,730 containing spice: collaborative
petroleum hydrocarbons in solid ma­ study,44
tiamulin residues in swine liver, GC
method,451 trixes, interlaboratory evaluation offightfilthfrom orientalsaucescon­
of off-line supercritical fluid ex- tainingsoy sauce,thickeners,and
xylazine and metabolite 2 ,6-dimethy- traction/infrared spectrometric spices:collaborativestudy,47
lanilineinbovine and swine kid­ method, 555 of fight filth from tofu: collaborative
ney, simultaneous determination
Environmental sanitation microbiol­ study,50
by LC, 720
ogy, Methods Committee report,191 liquid-solid,followedby supercritical
Drugs I Enzymatic method fluidelution,organicpollutantsin
Changes inMethods, 252 (1>3)(1 >4)-beta-D-glucansingrains reagentwater,72
Methods Committee report, 176 and cereals: collaborative study, matrix solid-phase, and GC screening
refereereport, 102 1069 of 14 chlorinated pesticides in
Drugs II sulfiteinfoods,NMKL interlaboratory crayfish (P r o c a m b a r u s c la r k ii)
Changes inMethods, 253 study,53 hepatopancreas,663
Methods Committee report, 176 Enzymatic-gravimetric methods matrix solid-phase, and GC screening
dietaryfiberinfoods,assum ofinsol­ of 14 chlorinated pesticides in
refereereport,102
uble and soluble fractions: sum­ oysters ( C ra ss o str e a v irg in ic a ),
Drugs III mary ofcollaborativestudy,923 67
Changes inMethods, 253 E s c h e r ic h ia c o li pararosaniline for spectrophotometric
Methods Committeereport, 177 in foods, substrate supporting disc determination of trace hydrogen
refereereport, 103 method for confirmed detection: sulfideinair,730
Drugs in feeds collaborativestudy,988 solid-phase, monophthalates and
Changes inMethods, 251 Ethyl-l-hexanol as contaminant in phthalide from drug solution ad­
Methods Committeereport, 193 drinking water, 1133 ministrationsets,531
Ethylene chlorhydrin in polyoxyethy- solid-phase,roxarsoneinfeeds,962
refereereport, 161
lated surfactantsand,LC method for Soxtec procedure, organiccompounds
melengestrol acetate in feeds, LC specificdetermination,292
method, 1163 insoilsand sediments,864
Ethylene glycols in anchovies, confir­ supercritical carbon dioxide method,
sulfamethoxine and ormetoprim, LC mation by tandem MS, 1344
methodforsimultaneousanalysis, optimizationofexperimentalcon­
Ethylene oxide ditions, pesticide residues in
320 in polyoxyethylated surfactants and, grains,857
tiamulin hydrogen fumarate in pre­ LC method for specific determi­ Extraneous materials in foods and
mixes,LC method, 444 nation,292 drugs
s e e a ls o Antibioticsinfeeds residual, in spectinomycin HC1 bulk Methods Committee report, 190
Drugs IV drug, dynamic headspace GC refereereport, 152
Changes inMethods, 253 method, 313
s e e a ls o Filth
Methods Committeereport, 177 Ethylenethiourea
Extraneous materials: isolation,
refereereport,104 in finished drinking water, GC/nitro-
Changes inMethods, 252
gen-phosphorus detection
Drugs V
method: collaborativestudy, 1113
Changes inMethods, 253 methodology, methylene chloride ac­
Methods Committee report, 177 ceptabilityand “purification” for,
refereereport, 104 1146 Fat, s e e Oilsandfats
 S u b je c t I n d e x : J o u r n a l O f A O AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1421

SUBJECT INDEX

Feeds se e a lsoExtraneousmaterialsinfoods Fluorometric methods

Changes inMethods, no changes anddrugs lacticacidinwines, 1017
Methods Committeereport, 193 Fish and other marine products sulfite in wine, by V-(9-acrid-
additives, granulated, need for laiger Changes inMethods, no changes inyl)maleimide, 1385
analyticalsamples,945 anchovies, propylene and ethylene Folic acid
aflatoxinsin,LC method, 366 glycols in, confirmation by tan­ in milk-based infant formula, Food
feeds,premixes, supplements, and in­ dem MS, 1344 and Drug Administration-Infant
jectable solutions, selenium in, crayfish (P rocam baru s clarkii )hepa- Formula Council: collaborative
hydride-generated inductively topancreas, 14 chlorinated pesti­ study,399
coupled plasma atomic emission cides in, matrix solid-phase ex­ Food additives
spectrometry,910 tractionand GC screening,663 Changes inMethods, 255
forages, nitrate content, rapid testfor Great Lakes commercial fish,organo- Methods Committee report, 180
semiquantitative determination, chlorinepesticidesand PCB con­ refereereport, 108
948 genersin,707 benzene innonstickcookware andmi­
roxarsone in, solid-phase extraction Great Lakes fish,organochlorine (to­ crowavesusceptorsanditsmigra­
andLC withUV detection,962 tal)contentof,703 tion into foods on cooking,
roxarsone in, standard additions lobster (H om arus am ericanus) diges­ GC/MS method,760
methodtodiagnosematrixeffects tive gland extracts, cadmium, biogenicaminesinbeersandtheirraw
ofuseofAAS foranalysis,1400 zinc,copper,andsilverionsin,gel materials, ion-pairLC with post­
vitamin E acetate in premixes, capil­ chromatography followed by column derivatization, 1027
laryGC method, 735 AAS andpolarography,794 biogenic amines in dry sausages, LC
water in, influence on near infrared orientalfishproductscontainingspice, method, 575
spectraofmodel compounds,741 extractionoflightfilthfrorr:col­ dibutylamineinhams,GC/chemilumi-
see a lso Drugs infeeds laborativestudy,44 nescencedetectionmethod, 578
Fertilizers and agricultural liming oysters (C rassostrea virg in ica ), 14
materials
flavor enhancers inosine 5'-mono-
chlorinated pesticides, matrix phosphate and guanosine 5'-mo-
Changes inMethods, 251 solid-phase extraction and GC nophosphateinfoodpreparations,
Methods Committee report, 194 screeningof,67 simultaneous determination by
refereereport, 163 salmon muscle,oxytetracyclinein,as­ derivative spectrophotometry,
ammoniated, citrate insoluble P2O 5, saybyLC withUV detection,325 754
920 shrimp, sulfites in, method improve­ lactic acid in wines, indirect
fertilizer sample preparation, evalu­ ment,565 fluorometricdetermination, 1017
ation of current AOAC require­
shrimp (P en aeu s vannam ei) muscle phenolic antioxidants(9)inbutteroil,
ments, 1174 tissue,nitrofurazoleand furazoli­ LC method: collaborative study,
liquidfertilizers,Texas liquidsampler done in,simultaneous determina­ 765
vs Missouri bottle for sampling, tion by LC with UV detection, siliconedefoamers infruitjuices, sol­
749 1235 ventextractionandAAS, 581
potassiumin,evaluationofnew flame see a lso Seafoodtoxins sulfite, enzymatic determination,
photometric instrumentation to
Flavors NMKL interlaboratorystudy,53
meetrequirementsofAOAC offi­
cialmethod, 1182 Changes inMethods, 254 sulfites in shrimp, method improve­
Filth Methods Committee report, 180 ment,565
light, extraction from oriental fish refereereport, 107 p-toluenesulfonamide in ice cream,
products containing spice: col­ Flow injection analysis continuous flow and LC determi­
laborativestudy,44 furanicaldehydesinfoodandpharma­ nation,interlaboratorystudy,570
light, extraction from oriental sauces ceuticalsamples, 1255 Food adulteration
containing soy sauce, thickeners, withreductiveamperometricdetection oligosaccharide generation in acidic
andspices:collaborativestudy,47 andelectrolyticcleanup,sulfurdi­ sugarmedia,584
light,extraction from tofu: collabora­ oxide in wines and beverages, sweetenersaddedtopureorangejuice,
tivestudy,50 1389 LC detection,342
1422 S u b je c t In d e x : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

SUBJECT INDEX

Food and Drug Administration Frozen storage, depletion of Gas chromatography

-InfantFormulaCouncilcollaborative monoamino metabolites of Zoalene benalaxylincropsand water,650
study, vitamins A and E, folic during,698 benzene infoods,survey, 1213
acid, and pantothenic acid in Fruits and fruit products chlorinatedpesticides (14) in crayfish
milk-basedinfantformula, 399 Changes inMethods, 254 (P ro c a m b a r u s c la r k ii ) hepa-
methyl mercury in canned tuna, sur­ Methods Committeereport, 183 topancreas,663
vey,36 refereereport, 136 chlorinated acids in water: collabora­
monitoring ofpesticideresiduesinin­ tivestudy, 1098
datesofSaudiArabiancultivars,boron
fant foods and adult foods eaten chlorinated pesticides (14) in oysters
by infants/children,492 in,601
{C r a sso stre a v irg in ic a ), 67
Food composition dilutedjuice beverage, calculation of
cyclohexanone leached from
caseinhydrolysates,cuprimetricassay, juicecontent,424
hemodialysistubing,GC method,
1295 fruits, formetanate HC1 in, coupled- 1127
dietary fiber in frozen foods, simple column cation exchange LC, dibutylamine inhams, 578
method, 1014 1362
ethylene oxide and ethylene chlorhy-
mentholinhoney,GC method, 1289 fruits,abamectin in,LC method, 691 drin in cosmetics and
protein(crude)incerealgrainsandoil­ grape juice, authentic varietal and polyoxyethylated surfactants,
seeds,Kjeldahlmethodvsgeneric commercial, sugars,alcohols,and specificdetermination,292
combustion method: collabora­ hydroxymethylfurfuralin,59 ethylene oxide (residual) in specti-
tivestudy,780 grape juice, authentic varietal and nomycin HC1 bulkdrug,313
protein(cmde) inmeatand meatprod­ commercial, carboxylic acid ethylenethiourea in finished drinking
ucts, combustion method: col­ composition,591 water:collaborativestudy, 1113
laborativestudy,780
juicesandsweeteners,sampleprepara­ hydrocarbonsincmde palm kerneloil,
Food microbiology—dairy
tionbias incarbon stable isotope 371
refereereport, 153
s e e a ls o Microbiological methods
ratioanalysisof,418 imazethapyrresiduesinsoybeans,377
Food microbiology—nondairy juices, silicone defoamers in, solvent mentholinhoney, 1289
Methods Committee report, 191 extractionand AAS, 581 methyl bromide in selected foods,
refereereport, 154 orange juice, pure, sweeteners added 1083
Food packaging to,LC detection,342 methyl mercury inbiologicalsamples,
microwave heat-susceptor food pack­ pyrethroidsresiduesin,applicationsto 608
aging,volatilechemicalsreleased pyrethroid residue monitoring organochlorine and organophosphate
from, 1268 studies, 1348 pesticidesinfruitsandvegetables,
For Your Information, 27A, 49A, Fumonisins multiresiduemethod, 1369
101A, 117A, 168A Bi and B2 in com, reproducibility of tiamulinresiduesinswine liver,451
Forensic sciences LC method: IUPAC collaborative Gas chromatography/mass spec­
Changes inMethods, 253 trometry
study,361
Methods Committee report, 177 benzeneinnonstickcookwareandmi­
Fungi in shared-use cosmetics avail­
refereereport, 105 crowave susceptorsanditsmigra­
able to the public, 430
Forestry matrixes, ecdysone agonist tionintofoodson cooking, 760
Furanic aldehydes in food and phar­
RH-5992 in,LC method, 668 ELISA vs, tetrahydrophthalimide in
maceutical samples, stopped-flow baby foods, 1225
Formaldehyde in fresh and retail
milk, LC method, 1010 injectionanalysismethod, 1255 8-hydroxymutilin in swine liver ex­
Formetanate HC1 in selected fruits, Furazolidone and nitrofurazone, in tracts,confirmatoryanalysis,459
coupled-column cation exchange shrimp (P e n a e u s va tm a m e i) muscle sulfonamideresiduesinanimaltissues,
LC, 1362 tissue, simultaneous determination confirmatorymethod, 976
Fourier transform infrared spec­ by LC withUV detection, 1235 Gas chromatography/plasma atomic
trometry Furosemide in tablets and injections, emission spectrometry, organolead
Frozen foods, dietary fiber in, simple stability-indicating proton NMR compounds in wine, speciation
method, 1014 spectroscopicassay method, 526 analysis, 1262
 S u b je c t I n d e x : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1423

SUBJECT INDEX

Gelatin, dessert preparations, and percritical fluid extraction/infra- formaldehyde infreshand retailmilk,
mixes, Changes in Methods, no redspectrometricmethod,555 LC method, 1010
changes Hydrogen sulfide (trace) in air, organic compounds in soilsand sedi­
Glyburide (Glibenclamide) and re­ pararosanilineforextractivespectro- ments, Soxtec extraction proce­
lated compounds in raw materials, photometricdetermination,730 dure, 864
962 Hydroxymethylfurfural, in authentic organicpollutantsinreagentwater,liq­
Grains varietaland commercial grapejuice, uid-solid extraction followed by
(l“>3)(l-+4)-beta-D-glucans in: col­ 59 supercriticalfluidelution,72
laborativestudy, 1069 8-Hydroxymutilin in swine liver ex­ propylene and ethylene glycolsinan­
pyrethroidsresiduesin,applicationsto tracts, GC/MS confirmation,459 chovies, confirmation by tandem
pyrethroid residue monitoring MS, 1344
studies, 1348 Infant Formula Council, EDA-, col­
Grains, pesticideresiduesin,supercriti­ laborative study, vitamins A and E,
calcarbondioxideextractionof,857 folic acid, and pantothenic acid in
Identification methods
Gravimetric methods, dietary fiberin milk-basedinfantformula,399
multispecies, field test by modified Infant foods
potatoes,precisionofmeasurements, agargelimmunodiflusion, 1022
509 pesticideresiduesin,FDA monitoring
setalbrush length, todiscriminatebe­ of,492
Great Lakes
tween adult mandibles of tetrahydrophthalimide in, ELISA vs
commercialfish,organochlorinepesti­ O ry za e p h ilu s su rin a m e n sis and
cidesandPCB congenersin,707 GC/MS, 1225
O. m ercator, 941
fish,organochlorine(total)contentof, Infant formulas
Imazethapyr in soybeans, GC withni­ leadin,ultratracelevels,isotopedilu­
703 trogen-phosphorusdetection,377
Guanosine 5-monophosphate and tion inductively coupled plasma
Immunoaffinity chromatography, MS, 1378
inosine 5'-monophosphate in food cleanup for aflatoxin M! in milk
preparations, simultaneous deter­ milk-based, iodide, linoleic acid, and
powder,interlaboratorystudy, 1248 vitaminsD and K in,FDA-Infant
mination by derivative spectro­
Immunoassays Formula Council: collaborative
photometry,754
competitiveinhibitionenzyme,methyl study, 1042
2-benzimidazolecarbamate in milk-based, vitamins A and E, folic
wine, 851 acid, and pantothenic acid in,
ELISA, captan and its degradation FDA-Infant Formula Council:
Hazardous substances product tetrahydrothalimide in collaborativestudy,399
Changes in Methods, 252 foods, 381 Infrared milk analyzers
toxic elements in foods, analytical polyclonal enzyme method, S a lm o ­ homogenizerperformance in,1033
methods, partIV:ashing,257 n e lla infoods,comparativestudy, quantitative linearity evaluation
Hemodialysis tubing, cyclohexanone 694 method for,1300
leachedfrom,GC method, 1127 ELISA vs GC/MS, tetrahydro- Infrared spectroscopy
Heptachlor epoxide isomers, struc­ phthalimideinbaby foods, 1225 near infrared spectra of model feed
tures and environmental signifi­ Immunodiffusion, modified agar gel, compounds, influence of water
cance, 1092 multispecies identificationfieldtest, on,741
Homogenizers in infrared milk ana­ 1027 petroleum hydrocarbons in solid ma­
lyzers, performance, 1033 India, Delhi, organochlorine pesticide trixes, interlaboratory evaluation
Honey residuesindairymilk,283 of off-line supercritical fluid ex-
cymiazoleresiduesin,LC method,92 Industrial chemical residues traction/infrared spectrometric
mentholin,GC method, 1289 Changes inMethods, 252 method, 555
Hydrocarbons chlorinated acids in water, GC/elec- Ion selective electrode, water-ex­
inpalmkerneloil(crude),GC method, tron capture detection method: tractablesodium inbread, 1138
371 collaborativestudy, 1098 Inorganics in water
petroleum,insolidmatrixes,interlabo­ methyl bromide in selected foods, Methods Committee report,68
ratory evaluation of off-line su- headspaceGC, 1083 refereereport,44
1424 S u b je c t I n d e x : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3

SUBJECT INDEX

Inorganic methods, Methods Commit­ Irradiation, ofspicesandherbs,micro­ aflatoxins in feeds and feed compo­
teereport, 195 biologicalscreeningmethodforindi­ nents,366
Iosine 5'-monophosphate and guanos- cation of: BCR collaborative study, aflatoxin M t in milk powder, inter­
ine 5'-monophosphate in food 674 laboratorystudy, 1248
preparations, simultaneous deter­ Iron, infoodstuffsafterdryashing,AAS aspartame and major decomposition
mination by derivative spectro­ method: NMKL interlaboratory products infood,275
photometry,754 study,798 automated cleanup and ion-pairing,
Instructions to Authors, 40A, 256,702 IUPAC, collaborativestudy,fumonisins cephapirin and desacetyl-
Interlaboratory studies B| and B2 incom, reproducibilityof cephapirininmilk,535
aflatoxin M| in milk powder, LC LC method, 361 bentazonformulations,387
method with immunoaffinity col­ Ivermectin, inanimal tissues,modified biogenicamines inbeersandtheirraw
umn cleanup, 1248 method, 1149 materials, 1027
comparisonprograms,reportingongo­ biogenic aminesindrysausages,575
ingresultsof,620 2-chloro-4-nitroaniline, 2-naphthol,
dicloxacillinpreparations,LC method, and 2,4-dinitroaniline in D&C
1143 Japan, aflatoxincontaminationinfoods Red No. 36,287
andfoodstuffsinTokyo, 32 cymiazoleresiduesinhoney, 92
evaluation of off-line supercritical
Job’s tears, zearalenone and a- dicloxacillinpreparations:interlabora­
fluid extraction/infrared spectro-
zearalenonein,LC method, 1006 torystudy, 1143
metric method forpetroleum hy­
drocarbons insolidmatrixes,555 diquatandparaquatherbicidesinpota­
toes:collaborativestudy,881
nicarbazin inchicken tissues, LC de­
termination with LC-ther- diquat and paraquat in low-moisture
Kjeldahl method vs generic combus­ foodcrops, 1323
mospray MC confirmation,420
tion method, protein (crude) in ce­ ecdysone agonistRH-5992 inforestry
nitrate in forages, rapid test for real grains and oilseeds: collabora­ matrixes,668
semiquantitative determination,
tive study, 780 formaldehyde infreshand retailmilk,
948
1010
NMKL, enzymatic determination of formetanate HC1 in selected fruits,
sulfiteinfoods,53 1362
NMKL, lead, cadmium, zinc, copper, Lactic acid in wines, indirect fumonisins B! and B2 in com, repro­
iron, chromium, and nickel in fluorometricdetermination, 1017 ducibilityofmethod: IUPAC col­
foodstuffs afterdry ashing, AAS y-Lactone, base hydrolyzed, fast atom laborativestudy,361
method,798 bombardment MS, 913 glyburide(Glibenclamide)and related
p-toluenesulfonamide in ice cream, Lead compounds inraw materials,962
continuousflow and LC determi­ in foodstuffs after dry ashing, AAS iodineinmilk:collaborativestudy,711
nation,570 method: NMKL interlaboratory melengestrolacetateinfeeds, 1163
S a lm o n e lla in dry foods, refrigerated study,798 methyl[l-[(butylamino)carbonyl]-
pre-enrichment and enrichment ininfantformula,ultratracelevels,iso­ lH-benzimidazol-2-yl]carbamate
brothcultures,814 tope dilution inductively coupled and methyl lH-benzimidazol-2-
thermospray-LC/MS method for se­ plasmaMS, 1378 ylcarbamate in wettable powder,
lected <V-methyl carbamites, N - Linoleic acid in milk-based infant for­ 1187
methyl carbamoyloximes, and mulas, FDA-Infant Formula Coun­ N -methylcarbamate inbeef and poul­
substitutedureapesticides, 1329 cil:collaborativestudy, 1042 trylivertissues, 1309
International Organization for Linurin and trifluralin residues in N - methylcarbamate pesticideresidues
Standardization, reportof carrots and their pulp, simultane­ incattle,swine, and poultryliver,
Iodide in milk-based infant formulas, ous determination by LC and GC, 907
FDA-Infant Formula Council: col­ 657 monophthalates and phthalide ex­
laborativestudy, 1042 Liquid chromatography tracted from drug solution ad­
Iodine, inmilk, LC method: collabora­ abamectininfruitsandvegetables,691 ministrationsets,531
tivestudy,711 acesulfam-Kinfoods,268 neomycin B inbovinekidney,543
 S u b je c t I n d e x : J o u r n a l O f AO AC I n t e r n a t io n a l V o l . 76, N o. 6 ,1 9 9 3 1425

SUBJECT INDEX

niacininfortifiedfoodproducts,390 residues incarrotsand theirpulp,si­ chicken tissues,nicarbazin in,LC de­

nicarbazin in chicken tissues, inter­ multaneous determination,657 termination with LC-ther-
laboratorystudy,420 Liquid chromatography-mass spec­ mospray MC confirmation, inter­
nitrofurazole and furazolidone in trometry laboratorystudy,420
shrimp(P e n a e u s v a n m m e i) mus­ moxidectin residues in cattle tissues poultrytissues,incurredpesticideresi­
cle tissue, simultaneous determi­ and cattlefat,1230 dues in,supercriticalfluidextrac­
nation, 1235 nicarbazinconfirmationinchickentis­ tionof,888
olestrainlipidsamples, 1396 sues,interlaboratorystudy,420
porktissue,sulfamethazinein,quanti­
oxyfluorfeninformulations,90 A-methyl carbamites, A-methyl car-
tationby ILC, 335
oxytetracycline residues in salmon bamoyloximes, and substituted
muscle,325 urea pesticides, interlaboratory meat and meat products, protein
oxytetracycline, tetracycline, and study, 1329 (crude) in, combustion method:
chlortetracyclineinmilk,549 L is te r ia s p e c ie s collaborativestudy,780
phenolic antioxidants(9)inbutteroil: isolated from foods, AutoMicrobic sausages(dry),biogenicamines in,LC
collaborativestudy,765 system for biochemical identifi­ method, 575
roxarsoneinfeeds,962 cationof:collaborativestudy.822 Melengestrol acetate in feeds, LC
sulfamethoxine and ormetoprim in isolatedfrom food and environmental method, 1163
feeds,simultaneousanalysis,320 samples, MICRO-ID Listeria Menthol in honey, GC method, 1289
sulfamethazine in milk: collaborative Method vs conventional bio­ Metals and other elements
study,725 chemical methods for identifica­ Changes inMethods, 252
sulfonamideresidues(6)inanimal tis­ tionof:collaborativestudy,831
Methods Committee report, 186
sues,966 L m o n o c y to g e n e s, resuscitation/selec-
refereereport, 142
sweetenersaddedtopureorangejuice, tion/kitsystemdetectioninfoods,
342 632 boron indatesofSaudi Arabian culti-
vars,601
thiamine and riboflavin in selected
foods, 1156 cadmium with 4-(2-pyridy-
thiamine,riboflavin,andpyridoxinein lazo)naphthol in industrial efflu­
medicalfoods, 1276 Malt beverages and brewing materials ents, coal, and fly ash, hy-
tiamulin hydrogen fumarate in com­ Changes inMethods, 253 droxyamidines as extracting
pleteswinemeal feeds,449 beer, alcohol content and original reagents for spectrophotometric
tiamulin hydrogen fumarate in feed gravity, automated analyzer determination,604
premixes,444 method: summary of collabora­ cadmium, zinc,copper,and silverions
tiamulin hydrogen fumarate in tia- tivestudy,918 inlobster(H o m a ru s a m e ric a n u s)
mulin-poly(vinyl chloride) for­ beersandtheirraw materials,biogenic digestiveglandextracts,gelchro­
mulations,447 amines in,ion-pairLC withpost­ matography followed by AAS
p-toluenesulfonamideinicecream,in­ column derivatization, 1027 and polarography,794
terlaboratorystudy,570 Mass spectrometry lead, cadmium, zinc, copper, iron,
vs AOAC microbiological method, - l fast atom bombardment, base hydro­ chromium, and nickel in food­
tryptophan in tablets and cap­ lyzedy-lactone,913 stuffs after dry ashing, AAS
sules,414 isotope dilution inductively coupled method: NMKL interlaboratory
vs microbiological plate assay, plasma, lead in infant formulas,
study,798
oxytetracyclineinanimalfeed,39 ultratracelevels, 1378
tandem, propylene and ethylene gly­ methylmercury inbiologicalsamples,
xylazine and metabolite 2,6-dimethy-
cols in anchovies, confirmation programmed temperature GC
lanilineinbovine and swine kid­
by, 1344 method, 608
ney, simultaneous determination
by,720 tandem,Testosteroneesters,306 methyl mercury incanned tuna, FDA
zearalenoneanda-zearalenoneinbar­ Meat, poultry, and meat and poultry survey,36
leyandJob’stears, 1006 products tinincannedfood,inductivelycoupled
Liquid chromatography-gas chro­ Changes inMethods, 255 plasma atomic emission spectro-
matography, linurin and trifluralin refereereport, 111 metricmethod, 1374
1426 S ubject In d e x : Journal O f AOAC International V ol . 76, N o . 6 ,1 9 9 3

SUBJECT INDEX

M ethod perform ance in cattle, swine, and poultry liver, LC cephapirin and desacetylcephapirin in
fractional factorial design approach for method, 907 milk, automated LC cleanup and
optimizing analytical methods, interlaboratory study of a ther- ion-pairing LC, 535
615 mospray-LC/MS method, 1329 composited, collected through the U.S.
immunochemical assays (2) at 37 field /V-Methyl carbam oyloxim es, inter­ Pasteurized Milk Network, pesti­
laboratories, for aflatoxins in pea­ laboratory study of a thermospray- cide residues in, 1220
nuts, 637 LC/MS method, 1329 cow’s milk and human milk with cor­
LC methodvs AOAC microbiological M ethylene chloride, determination of responding human blood samples,
method, L-tryptophan in tablets acceptability and “purification” for ochratoxin A in, 842
andcapsules,414 ethylenethiourea methodology, 1146 dairy milk, Delhi, India, organo-
methylation methods for quantitation S-M ethylm ethioninesulfonium chlorine pesticide residues in, 283
of conjugated linoleic acid iso­ M icrobiological methods fresh and retail, formaldehyde in, LC
mers, 644 Changes in Methods, 252 method, 1010
metrological characteristics used for AOAC method vs LC method, L-tryp- infrared analyzers, homogenizer per­
safety control of food and envi­ tophan in tablets and capsules, formance in, 1033
ronmental materials, 262 414 infrared analyzers, quantitative linear­
off-line supercritical fluid extrac- automation, partial, vitamin assays, ity evaluation method for, 1300
tion/infrared spectrometric 1280 iodine in milk, LC method: collabora­
method, petroleum hydrocarbons AutoMicrobic system for biochemical tive study, 711
in solid matrixes, interlaboratory identification of L is te r ia species oxytetracycline, tetracycline, and
evaluation, 555 isolated from foods: collaborative chlortetracycline in, LC method,
reliability of mycotoxin assays, up­ study, 822 549
date, 461 bacterial count (total) in raw milk us­ powder, aflatoxin M, in, LC method
reporting ongoing results of interlabo­ ing Bactoscan 8000, 838 with immunoaffinity column
ratory comparison programs, 620 microbial loads of B a c illu s su b tilis cleanup, interlaboratory study,
S h igella so n n ei recovery from se­ spores on penicylinders, 355 1248
lected foods, effectiveness of the MICRO-ID L is te r ia Method vs con­ raw milk, bacterial count (total) in, us­
B a c te r io lo g ic a l A n a ly tic a l M a n ­ ventional biochemical methods ing Bactoscan 8000, 838
ual culture method for, 1240 for identification of L is te r ia iso­ sulfamethazine in, adsorptive-strip­
M eth yl 2-benzim idazolecarbam ate in lated from food and environ­ ping determination, 540
wine, competitive inhibition enzyme mental samples: collaborative sulfamethazine in, LC method: col­
immunoassay, 851 study, 831 laborative study, 725
M ethyl[l-[(butylam ino)carbonyl]-l plate assay vs LC, oxytetracycline as­ tetracyclines in, multiresidue metal
H-benzim idazol-2-yl]carbam ate in say in animal feed, 39 chelate affinity chromatography
wettable powder, LC method, 1187 resuscitation/selcction/kit system de­ method, 329
M eth yl lH -benzim idazol-2-ylcar- tection, L iste ria m o n o c y to g e n e s M onophthalates extracted from solu­
bamate in wettable powder, LC determination in foods, 632 tion adm inistration sets, solid-
method, 1187 refrigerated pre-enrichment and en­ phase extraction and LC determina­
M eth yl brom ide in selected foods, richment broth cultures, S a lm o ­ tion, 531
headspace GC, 1083 n e lla in dry foods, interlaboratory M oxidectin residues in cattle tissues
M eth yl m ercury study, 814 and cattle fat, LC/MS method, 1230
in biological samples, programmed for screening of spices and herbs for ir­ M ultiresidue methods
temperature GC method, 608 radiation: BCR collaborative Methods Committee report, 187
in canned tuna, FDA survey, 36 study, 674 referee report, 144
M ethylation methods, comparison for M icrochem ical methods, Changes in organochlorine and organophosphate
quantitation of linoleic acid isomers Methods, no changes residues in fruits and vegetables,
(conjugated), 644 M ilk s and m ilk products GC method, 1369
A'-M ethylcarbam ate pesticides automated technique for sampling tetracyclines in milk, metal chelate af­
in beef and poultry liver tissues, LC from farm bulk tanks: collabora­ finity chromatography method,
method, 1309 tive study, 297 329
 S ubject In de x : Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3 1427

SUBJECT INDEX

M ycotoxins referee report, 138 O rganolead com pounds in wine, spe-

Methods Committee report, 180 Nutrients in soils ciation analysis, capillary GC/micro-
referee report, 112 Methods Committee report, 194 wave-induced plasma atomic emis­
assays, reliability of, update, 461 referee report, 164 sion spectrometry, 1262
fumonisins B, and B2 in com, repro­ N uts and nut products, Changes in O rganonitrogen pesticides
ducibility of LC method: IUPAC Methods, no changes Methods Committee report, 188
collaborative study, 361 referee report, 148
ochratoxin A in cow’s milk and human
O rganophosphorus pesticides
milk with corresponding human
Methods Committee report, 188
blood samples, 842
O chratoxin A in cow ’s m ilk and hu­ containing a nitrophenyl group, TLC
zearalenone and a-zearalenone in bar­
m an m ilk w ith corresponding hu­ detection of, 1394
ley and Job’s tears, LC method,
m an blood samples, 842 in fruits and vegetables, multiresidue
1006
O ils and fats GC method, 1369
see a lso Aflatoxins
Changes in Methods, 255 O rganotin com pounds in household
Methods Committee report, 183 commodities, small-scale survey,
referee report, 133 436
2-Naphthol in D & C R e d No. 36, L C
in butter oil, phenolic antioxidants (9) O rm etoprim and sulfam ethoxine in
method, 287 in, LC method: collaborative feeds, simultaneous L C analysis, 320
study, 765
N atural poisons, Changes in Methods, O ryza ep h ilu s species, setal brush length
255 conjugated linoleic acid isomers, com­ to discriminate between adult mandi­
Neom ycin B in bovine kidney, LC
parison of methylation methods bles of O. su rinam ensis and O. m er-
method, 543 for quantitation of, 644 cator, 941
New Products, 34A, 57A, 95A, 114A, olestra in lipid samples, LC method, O xyfluorfen in form ulations, LC
163A 1396 method, 90
N iacin palm kernel oil (crude), hydrocarbons
Oxytetracycline
in, GC method, 371
dissolution profiles of sustained-re­ in milk, LC method, 549
Oilseeds, protein (crude) in, Kjeldahl
lease niacin products, automated in salmon muscle, assay by LC with
and manual procedures, 394 method vs generic combustion
method: collaborative study, 780 UV detection, 325
in fortified food products, LC analysis,
O lestra in lipid samples, LC method,
390
1396
N icarbazin
Oligosaccharides, generation in acidic
in chicken tissues, LC determination
sugar media, 584
withLC-thermospray MC confir­
O rgan ic methods, Methods Committee Packaging materials, se e Drug packag­
mation, interlaboratory study, 420
report, 196 ing; Food Packaging
N ickel in foodstuffs after d ry ashing,
O rgan ics in water, referee report, 169 Panthenol and pantothenic acid in
AAS method: NMKL interlabora­
O rganochlorine pesticides m ilk-based infant form ula, Food
tory study, 798
in dairy milk, Delhi, India, 283 and Drug Administration-Infant For­
Niclosam ide released from m ollus-
in commercial fish from the Great mula Council: collaborative study,
cicidal form ulations, rapid accurate
Lakes, 707 399
method, 847
in fruits and vegetables, multiresidue Pararosaniline fo r extractive spectro-
Nitrate in forages, rapid test for
GC method, 1369 photom etric determ ination of
semiquantitative determination, 948
organochlorine and organophosphate trace hydrogen sulfide in air, 730
Nitrofurazone, and furazolidone in
shrimp (P enaeus vannam ei) muscle residues in fruits and vegetables, Paraquat
tissue, simultaneous determination multiresidue GC method, 1369 in low-moisture food crops, silica col­
by LC with UV detection, 1235 total, in fish from the Great Lakes, 703 umn cleanup and LC with UV de­
Nonalcoholic beverages, Changes in O rganohalogen pesticides tection, 1323
Methods, no changes Methods Committee report, 187 in potatoes, LC method: collaborative
Methods Committee report, 184 referee report, 147 study, 881
1428 S ubject In d e x : Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3

SUBJECT INDEX

Peanuts abamectin in fruits and vegetables, LC W-methylcarbamate in cattle, swine,

aflatoxin analysis, variability among method, 691 and poultry liver, LC method, 907
subsamples prepared with four benalaxyl in crops and water, GC V-methyl carbamites, V-methyl car-
mills, 983 method, 650 bamoyloximes, and substituted
aflatoxins in, performance of 2 immu­ captan and its degradation product urea pesticides, thermospray-
nochemical assays at 37 field tetrahydrothalimide in foods, LC/MS method, interlaboratory
laboratories, 637 ELISA, 381 study, 1329
Penicylinders, microbial loads of B acil­ chlorinated (14), in crayfish (P rocam - in milk (composited) collected through
lus su btilis spores on, 355 barus clarkii) hepatopancreas, the U.S. Pasteurized Milk Net­
Pesticide form ulations matrix solid-phase extraction and work, 1220
Changes in Methods, 251 GC screening, 663 organic compounds in soils and sedi­
bentazon, LC assay, 387 chlorinated (14), in oysters (C ra sso s- ments, Soxtec extraction proce­
trea virginica), matrix solid-phase dure, 864
methyl[l-[(butylamino)carbonyl]-
1H-benzimidazol-2-yl]carbamate extraction and GC screening of, organochlorine and organophosphate
and methyl lH-benzimidazol-2- 67 residues in fruits and vegetables,
ylcarbamate in wettable powder, cymiazole in honey, LC method, 92 multiresidue GC method, 1369
LC method, 1187 diquat and paraquat in low-moisture organochlorines in animal fats, gel per­
molluscicidal formulations, ni­ food crops, silica column cleanup meation chromatography vs
closamide released from, rapid and LC with UV detection, 1323 sweep codistillation vs Florisil
accurate method, 847 diquat and paraquat herbicides in pota­ column adsorption chromatogra­
oxyfluorfen in formulations, LC toes, LC method: collaborative phy for sample cleanup, 1317
method, 90 study, 881 organochlorines in dairy milk, Delhi,
ecdysone agonist RH-5992 in forestry India, 283
water-dispersible powder, suspensibil-
matrixes, LC method, 668 in poultry tissues (incurred), super­
ity measurement, and comparison
of microanalytical techniques for ethylenethiourea in finished drinking critical fluid extraction of, 888
wastewater reduction, 83 water, GC/nitrogen-phosphorus pyrethroid residues in vegetables,
detection method: collaborative
Pesticide form ulations: carbam ate in­ fruits, grains, beans, and green tea
study, 1113
secticides and substituted urea in­ leaves; applications to pyrethroid
secticides, Methods Committee re­ formetanate HC1 in selected fruits, residue monitoring studies, 1348
port 173 coupled-column cation exchange
roxarsone in feeds, solid-phase extrac­
LC, 1362
Pesticide form ulations: C IP A C stud­ tion and LC with UV detection,
ies, Methods Committee report, 173
in grains, supercritical carbon dioxide
962
extraction of, 857
Pesticide form ulations: fungicides, p H method, cholinesterase in whole
heptachlor epoxide isomers, structures
Methods Committee report, 173 blood: collaborative study, 899
and environmental significance,
Pesticide form ulations: herbicides, Phenolic antioxidants in butter oil, LC
1092
Methods Committee report, 173 method: collaborative study, 765
imazethapyr in soybeans, GC with ni­
referee report, 98 Phthalide extracted from solution ad­
trogen-phosphorus detection, 377
Pesticide form ulations: organohalo- in indoor air and dust, 1121 m inistration sets, solid-phase ex­
gen insecticides, referee report, 98 traction and LC determination, 531
in infant foods and adult foods eaten by
Pesticide form ulations: other insecti­ infants/children, FDA monitoring Plant toxins
cides, synergists, and insect repel­ of, 492 Methods Committee report, 181
lents, referee report, 98 linurin and trifluralin residues in car­ referee report, 119
Pesticide form ulations: other organo- rots and their pulp, LC and GC Plants, Changes in Methods, no
phosphorus insecticides, referee re­ method, 657 changes
port, 99 methyl 2-benzimidazolecarbamate in Polychlorinated biphenyls, congeners
Pesticide form ulations: rodenticides wine, competitive inhibition en­ in commercial Fish from the Great
and m iscellaneous pesticides zyme immunoassay, 851 Lakes, 707
Pesticide residues /V-methylcarbamate in beef and poul­ Potassium in fertilizers, evaluation of
Changes in Methods, 252 try liver tissues, LC method, 1309 new flame photometric instrumenta-
 S ubject In de x : Journal O f AO A C International V ol . 76, N o. 6 ,1 9 9 3 1429

SUBJECT INDEX

tion to meet requirements of AOAC tiamulin hydrogen fumarate in feed Selenium in feeds, premixes, supple­
official method, 1182 premixes, LC method, 444 ments, and injectable solutions, hy­
Potatoes tiamulin hydrogen fumarate in tia- dride-generated inductively coupled
dietary fiber in, precision of measure­ mulin-poly(vinyl chloride) for­ plasma atomic emission spectrome­
ments of, 509 mulations, LC method, 447 try, 910
diquat and paraquat herbicides in, LC tiamulin residues in swine liver, GC S h ig e lla so n n e i recovery from selected
method: collaborative study, 881 method, 451 foods, effectiveness of the B acterio­
Preservatives and artificial sweeten­ R iboflavin lo g ica l A n a lytica l M an u al culture
ers, Methods Committee report, 156 in medical foods, LC method modifi­ method for, 1240
Propylene glycols in anchovies, confir­ cation, 1276 Silve r in lobster (H om aru s am eri-
mation by tandem MS, 1344 in selected foods, LC method, 1156 ca n u s) digestive gland extracts, gel
Protein chromatography followed by AAS
crade, in cereal grains and oilseeds, and polarography, 794
Kjeldahl method vs generic com­ Sodium water-extractable, in bread,
bustion method: collaborative sodium ion selective electrode with
Safety
study, 780 minimum sample preparation, auto­
control of food and environmental ma­ mation of potentiometric measure­
crude, in meat and meat products,
terials, metrological charac­ ments, 1138
combustion method: collabora­
teristics of analytical methods, Soils and sediments
tive study, 780
262
Proteins (intact), tryptophan in, deriva­ indoor dust, pesticide residues in, 1121
Salmonella in foods
tive spectrophotometry, 1168 organic compounds in soils and sedi­
Pyrethroids, residues in vegetables,
diy foods, refrigerated pre-enrichment ments, Soxtec extraction proce­
fruits, grains, beans, and greer tea and enrichment broth cultures, in­ dure, 864
terlaboratory study, 814
leaves; applications to pyretkroid Soxtec analysis, organic compounds in
residue monitoring studies, 1343 polyclonal enzyme method, compara­ soils and sediments, 864
Pyridoxine, in medical foods. LC
tive study, 694 Soybeans and soybean products
method modification, 1276 Sam ples, larger, needed for granulated
imazethapyr residues in soybeans, GC
feed additives, 945 with nitrogen-phosphorus detec­
Sam ple handling tion, 377
fertilizer sample preparation, evalu­ soy sauce, extraction of fight filth from
Q uality assurance ation of current AOAC require­ oriental sauces containing: col­
international harmonized protocol for ments, 1174 laborative study, 47
proficiency testing of (chemical) sample preparation bias in carbon sta­ tofu, extraction of light filth from: col­
analytical laboratories, 926 ble isotope ratio analysis of fruit laborative study, 50
international perspectives in certifica­ juices and sweeteners, 418 Speciation analysis, organolead com­
tion and accreditation, 1 peanut subsamples prepared for afla- pounds in wine, capillary GC/micro-
toxin analysis, variability among wave-induced plasma atomic emis­
four mills, 983 sion spectrometry, 1262
Sam pling Spectinom ycin HC1, ethylene oxide
Radioactivity automated technique for milk from (residual) in bulk drag, dynamic
Changes in Methods, no changes farm bulk tanks: collaborative headspace GC method, 313
Methods Committee report 188 study, 297 Spectrom etry
referee report, 149 liquid fertilizers, Texas liquid sampler plasma atomic emission, selenium i
Regulatory analytical methods vs Missouri bottie for, 749 feeds, premixes, supplement
animal drags, 443 Seafood products and injectable solutions, 910
8-hydroxymutilin in swine liver ex­ Methods Committee report, 181 plasma atomic emission, tin in cann
tracts, GC/MS confirmation, 459 referee report, 120 food, 1374
tiamulin hydrogen fumarate in com­ s e e a ls o Fish and other marine prod­ Spectrophotom etry
plete swine meal feeds, LC ucts derivative, inosine 5'-monophospl
method, 449 Seafood toxins, referee report, 120 and guanosine 5'-monophospl
1430 S ubject In d e x : Journal O f AO A C International V ol . 76, N o . 6 ,1 9 9 3

SUBJECT INDEX

in food preparations, simultane­ in milk, LC method: collaborative sulfur (elemental), application to com­
ous determination, 754 study, 725 plex sulfur ointments, 1152
derivative, tryptophan in proteins (in­ in pork tissue, quantitation by ILC, sulfamethazine quantitation in pork
tact), 1168 335 tissue, 335
hydroxyamidines as extracting re­ S u lfu r dioxide in w ines and bever­ T iam ulin residues in swine liver, GC
agents for determination of, cad­ ages, flow injection analysis with re­ method, 451
mium with 4-(2-pyridy- ductive amperometric detection and T iam ulin hydrogen fum arate
lazo)naphthol in industrial electrolytic cleanup, 1389 in complete swine meal feeds, LC
effluents, coal, and fly ash, 604 Sulfites method, 449
pararosaniline for extractive determi­ enzymatic determination, NMKL in­ in feed premixes, LC method, 444
nation of trace hydrogen sulfide in terlaboratory study, 53 in tiamulin-polytvinyl chloride) for­
air, 730 in shrimp, method improvement, 565 mulations, LC method, 447
Spectroscopy, stability-indicating pro­ in wine, fluorometric determination by T in in canned food, inductively cou­
ton NMR, furosemide in tablets and /V-(9-acridinyl)maleimide, 1385 pled plasma atomic emission spec­
injections, 526 Sulfonam ides trométrie method. 1374
Spices and other condim ents in animal tissues, GC/MS confirma­ Tobacco
Changes in Methods, 244 tion, 976 Methods Committee report, 194
Methods Committee report, 184 in animal tissues, LC with post column referee report, 165
referee report, 139 derivatization, 966 p-Toluenesulfonam ide in ice cream,
extraction of light filth from oriental Supercritical fluid extraction continuous flow and LC determina­
fish products containing: collabo­ organic pollutants in reagent water, 72 tion, interlaboratory study, 570
rative study, 44
off-line, petroleum hydrocarbons in Toxicity testing, bioluminescent bacte­
extraction of light filth from oriental solid matrixes, interlaboratory rial genotoxicity test, evaluation and
sauces containing soy sauce, method evaluation, 555 application, 893
thickeners, and, 47
pesticide residues in poultry tissues T riflu ralin and linuron residues in
irradiated spices and herbs, microbio­
(incurred), 888 carrots and their pulp, simultane­
logical screening method for indi­
Sweep codistillation, cleanup of or- ous determination by LC and GC,
cation of: BCR collaborative
ganochlorines in animal fats, 1317 657
study, 674
Tryptophan
Standard additions method to diag­
in proteins (intact), derivative spectro­
nose m atrix effects in use of A A S to
photometry, 1168
analyze feed blended for a roxar-
Testosterone esters, tandem MS meth­ in tablets and capsules, LC vs AOAC
sone com bination study, 1400
ods, 306 microbiological method, 414
Su gars and su gar products
Changes in Methods, no changes Tetracyclines
Methods Committee report, 184 in milk, LC method, 549
referee report, 140 in milk, multiresidue metal chelate af­
acidic sugar media, oligosaccharides finity chromatography method, Vegetable products: processed
generation in, 584 329 Changes in Methods, no changes
cane sugar and beet sugar hydrolysates Tetrahydrothalim ide Methods Committee report, 184
added to pure orange juice, LC in baby foods, ELISA vs GC/MS, 1225 referee report, 138
detection, 342 in foods, ELISA, 381 Vegetables
sugars in authentic varietal and com­ Thiam ine abamectin in, LC method, 691
mercial grape juice, 59 in medical foods, LC method modifi­ pyrethroids residues in, applications to
see a lso Honey cation, 1276 pyrethroid residue monitoring
ilfamethoxine and orm etoprim in in selected foods, LC method, 1156 studies, 1348
feeds, simultaneous LC analysis, 320 T hin-layer chrom atography Veterinary analytical toxicology
lfamethazine organophosphorus insecticides con­ Changes in Methods, 252
i milk, adsorptive-stripping determi­ taining a nitrophenyl group detec­ Methods Committee report, 194
nation, 540 tion of, 1394 referee report, 165
 S ubject In dex : Journal O f A O AC International V ol . 76, N o. 6 ,1 9 9 3 1431

SUBJECT INDEX

cholinesterase in whole blood, pH L-tryptophan in tablets and capsules, W ines

method: collaborative study, 899 LC vs AOAC microbiological Changes in Methods, no changes
Vitam ins A in m ilk-based infant for­ method,414 sulfite in, fluorometric determination
m ulas, FDA-Infant Formula Coun­ vitamins A and E, folic acid, and pan­ by V-(9-acridinyl)maleimide,
cil: collaborative study, 399 tothenic acid in milk-based infant 1385
V itam in D in m ilk-based infant for­ formula, Food and Drug Admini­ lactic acid in, indirect fluorometric de­
m ulas, FDA-Infant Formula Coun­ stration-Infant Formula Council: termination, 1017
cil: collaborative study, 1042 collaborative study, 399 methyl 2-benzimidazolecarbamate in,
V itam in E, in m ilk-based infant for­ water-soluble, in foods, LASER/Mi- competitive inhibition enzyme
m ula, Food and Drug Administra­ crobe bioassay technology ap­ immunoassay, 851
tion-Infant Formula Council: col­ plied to, 682 organolead compounds, speciation
laborative study, 399 Volatile organic com pounds analysis, capillary GC/micro-
V itam in K , in milk-based infant formu­
in bottled drinking water sold in Can­ wave-induced plasma atomic
las, FDA-Infant Formula Council, ada, 26 emission spectrometry, 1262
part IV: collaborative study, 1042 from microwave heat-susceptor food sulfur dioxide in, flow injection analy­
packaging, 1268 sis with reductive amperometric
Vitam ins and other nutrients
Changes in Methods, 255 detection and electrolytic
cleanup, 1389
Methods Committee report, 184
referee report, 141 W astew ater
Chinese Total Diet Study in 1990, part reduction, suspensibility measurement
n, nutrients, 1206 of water-dispersible powder pesti­
cholesterol in multicomponent foods, X ylazine and metabolite 2,6-dimethy-
cide formulations, and compari­
alternative to AOAC official laniline in bovine and swine kid ­
son of microanalytical techniques
method, 1057 ney, simultaneous determination by
for, 83
cholesterol in prepared foods, rapid LC, 720
W ater
determination, 902 benalaxyl residues in water, GC
iodide, linoleic acid, and vitamins D method, 650
and K in milk-based infant formu­ chlorinated acids in, GC/electron cap­
las, FDA-Infant Formula Coun­ ture detection method: collabora­ Zearalenone and a-zearalenone in
cil: collaborative study, 1042 tive study, 1098 barley and Job ’s tears, LC method,
niacin in fortified food products, LC in feeds, influence on near infrared 1006
analysis, 390 spectra of model compounds, 741 Zinc
niacin dissolution profiles of sus­ reagent, organic pollutants in, liquid- in foodstuffs after dry ashing, AAS
tained-release niacin products, solid extraction followed by su­ method: NMKL interlaboratory
automated and manual proce­ percritical fluid elution, 72 study, 798
dures, 394 see a lso Drinking water in lobster (H om arus am ericanus) di­
thiamine and riboflavin in selected W ater m icrobiology gestive gland extracts, gel chro­
foods, LC method, 1156 Methods Committee report, 192 matography followed by AAS
thiamine, riboflavin, and pyridoxine in referee report, 160 and polarography, 794
medical foods, LC method modi­ W aters and salt, Changes in Methods, Zoalene, monoamino metabolites, de­
fication, 1276 no changes pletion during frozen storage, 698
Journal Information

The Association development of precise, accurate, and sensi­ M ic ro film . Volumes on microfilm are
tive methods for the analysis of foods, food available from Princeton Microfilm Corp.,
The primary objective of AOAC Interna­ additives and supplements, contaminants, PO Box 2073, Princeton, NJ 08540.
tional is to obtain, improve, develop, test, cosmetics, drugs, toxins, hazardous sub­ Single Issues. Information on single is­
and adopt precise, accurate, and sensitive stances, pesticides, feeds, fertilizers, and the sues is available from AOAC.
methods for analysis of foods, vitamins, environment.
food additives, pesticides, drugs, cosmetics, Methods. The scientific validity of pub­ Copying
plants, feeds, fertilizers, hazardous sub­ lished methods is, of course, evaluated as
stances, air, water, and any other products, part of the peer-review process. However, C op yrigh t. All articles published in the J o u r­
substances, or phenomena affecting the pub­ unless otherwise stated, methods published n a l are protected by copyright. Persons re­
lic health and safety, the economic protec­ in contributed papers in the J o u rn a l have not quiring copies of J o u rn a l articles beyond the
tion of the consumer, or the protection of the been adopted by AOAC and are not AOAC number allowed by the fair use provisions of
quality of the environment; to promote uni­ Official Methods. the 1978 U.S. copyright law may request
formity and reliability in the statement of ana­ M anuscripts. Prepare manuscripts dou­ permission to copy directly from AOAC or
lytical results; to promote, conduct, and ble spaced throughout and carefully revise may make the required copies and pay $ 1.00
encourage research in the analytical sciences before submission; 4 copies should be sub­ per copy through the Copyright Clearance
related to foods, drugs, agriculture, the envi­ mitted to Managing Editor, AOAC Interna­ Center, Inc., 222 Rosewood Dr, Danvers,
ronment, and regulatory control of commodi­ tio n al, Suite 400, 2200 W ilson B lvd, MA 01923. Articles which are copied and
ties in these fields; and to afford opportunity for Arlington, VA 22201-3301 USA. AOAC re­ royalties paid through the Copyright Clear­
the discussion of matters of interest to scientists serves the right to refuse manuscripts pro­ ance Center must be identified by the fol­
engaged in relevant pursuits. duced on dot matrix printers of less than lowing code: 0004-5756/87$1.00, indicat­
AOAC O ffic ia l M ethods are methods letter quality. ing the International Standard Serial Number
that have been validated by an AOAC-ap- Reprints. Authors may order reprints of assigned to J. A O A C I n t . , the year, and the
proved collaborative study, recommended articles when they return desktop published copying fee per copy. Information on the
by the appropriate AOAC General Referee, proofs. An order form and schedule of rates use of the Copyright Clearance Center is
Methods Committee, and the Official Meth­ is included with each author proof. Readers available from the Center.
ods Board, and adopted and published ac­ who wish to obtain individual reprints Please Note. Except for photocopies pre­
cording to the Bylaws of the Association. should contact authors directly. pared as described above, no part of this Jo u r­
Published papers that include such methods Responsibility. AOAC and the Editors n a l may be reproduced in any form or stored in
are distinguished by the words Collabora­ assume no responsibility for statements and any microform or electronic form without the
tive Study in the title and by footnotes that opinions expressed by authors of manu­ written permission of the Association.
indicate Association actions. scripts published in the Jo u rn a l.
M em bership in AOAC is open to all in­
Advertising
terested persons worldwide. Sustaining Subscriptions
memberships are available to any govern­ Placement. All space reservations and all ad­
ment agency, private company, or associa­ Jo urnal subscriptions are sold by the volume vertising copy are due 1V i months in ad­
tion interested in supporting an independent (one calendar year). Rates for 1992 are as fol­ vance of publication at AOAC International,
methods validation program. lows; Members: $120.00 in U.S.; $158.00 out­ Suite 400,2200 Wilson Blvd, Arlington, VA
European Representatives. For informa­ side U.S. Nonmembers: $175.00 in U.S.; 22201-3301 USA. Advertising inserted 3,6,
tion about AOAC and its publication, per­ $213.00 outside U.S. Two year subscriptions or 12 times within 1 year of first insertion
sons outside the U.S. may also contact the are also available. Airmail $98.00 additional earns frequency discount rates. Contact
following; Margreet Lauwaars, PO Box per year. Claim for copies lost in the mail will Marilyn Blakely, AOAC, for size require­
153,6720 AD Bennekom, The Netherlands, not be allowed unless received within 30 days ments and schedule of rates.
telephone 31-8389-18725; Derek C. Abbott, of the date of issue for U.S. subscribers or 90 Policy. Advertising is prepared by the ad­
33 Agates Lane, Ashtead, Surrey, KT21 days for all others. Claimants must state that vertisers and has been reviewed for compli­
2ND, UK, telephone 44-3722-74856. the publication was not received at their re­ ance with the AOAC Policy on Advertising
corded address. Address requests for replace­ in Association Publications. AOAC publica­
The Journal ment copies to AOAC International, Suite400, tion of an advertisement is not an endorse­
2200 Wilson Blvd, Arlington, VA 22201-3301 ment or approval, expressed or implied, by
The J o u rn a l o f A O A C In te rn a tio n a l (ISSN USA. For subscribers outside the U.S., copies the Association or the Editors of any service,
1060-3271) is published bimonthly by of the Jo u rn a l lost in transit cannot be replaced product, or claim made by the manufacturer.
AOAC International, Suite 400, 2200 Wil­ without charge because of uncertain mailing N ew Products. Notices of new products are
son Blvd, Arlington, VA 22201-3301 USA. conditions. prepared from literature received from man­
Each volume (one calendar year) will con­ Change o f Address. Notification should ufacturers and other companies and are pub­
tain about 1200 pages. The scope of the include both old and new addresses, with lished only as a service to readers.
J o u rn a l encompasses the development and mail code, and be accompanied by a mailing Publication of notices of new products is not
validation of analytical procedures pertain­ label from a recent issue. Allow 4 weeks for intended as an AOAC endorsement or ap­
ing to die physical and biological sciences change to become effective. Subscribers proval of any service, product, or claim
related to foods, drugs, agriculture, and the outside the U.S. should use airmail for noti­ made by the manufacturers.
environment. Emphasis is on research and fication. 11/92
JOIN AN AOAC SECTION
OR START ONE IN YOUR AREA!!
CURRENT AOAC SECTIONS

C e n tra l Section M id -C a n a d a R e g io n a l Section P a c ific N o rth w e st R e g io n a l

Section
President: H u sse in S. Ragheb President: M a rk T orchia
O ffice o f the Indiana St. Boniface General H ospital Chairm an: Isabel C ham berlain
State C hem ist Departm ent o f Su rge ry E P A R e gio n 10
Purdue U n ive rsity 409 Tache Avenue 7411 Beach D riv e East
Departm ent o f Biochem istry W innipeg, M B R 2 H 2 A 6 Port O rchard, W A 98366
W . Lafayette, IN 47907 Canada U SA
U SA (204) 237-2568 (206) 871-1882
(317) 494-1572

E a ste rn O n ta rio -Q u e b e c M id w e st R e g io n a l Section P a c ific Sou th w est R e g io n a l

R e g io n a l Section Section
Contact: G eorge Rottinghaus
Chairm an: G ille s Paillard U n iv. o f M isso u ri C olum b ia President: T e rry Jackson
Quebec Departm ent o f Veterinary M ed ical C alifo rn ia Departm ent o f
A griculture D iagn ostic Laboratory Foo d & A gricu ltu re
2700 R ue Einstein C 2 105 P O B o x 6023 3292 M eadow vie w R oad
Sainte Fo y, Q C G 1 P 3W 8 Colum bia, M O 65211 Sacram ento, C A 95832
Canada U SA U SA
(418) 643-2561 (314) 882-5994 (916) 2 6 2 -2070

E u ro p e Section N ew Y o rk -N e w Je rsey S o u th e a st-U S A R e g io n a l

Section Section
President: Juan Sabater
Laboratory D r. Sabater S A President: A le x M a cD o n a ld President: D o u g H ite
C alle de Lon d re s 6 H offm an-LaR oche Inc. Tennessee Dept, o f A gricu ltu re
08029 Barcelona 340 K in g sla n d Street Technical Services
Spain Nutley, N J 07110-1199 P O B o x 40527, M e lro se Station
+ 34 (3) 3228806 U SA N ash ville, T N 37204
(201) 235-4641 U SA
(615) 360-0314

M id A tla n tic U S A Section N o rth e ast R e g io n a l Section S o u th w e st-U S A R e g io n a l

Section
President: C harles Lattuada President: C athy Burchat
U SD A O ntario M in istry o f Agriculture President: A lle n C arm an
B u ild in g 322 B A R C -E a s t and Foods FDA
10300 Baltim ore Avenue 95 Stone R oad W est, Z one 2 4298 E ly sia n Fields
Beltsville, M D 20733 Guelph, O N N 1 H 8J7 N ew O rleans, L A 70122
U SA Canada U SA
(301) 504-8514 (519) 767-6214 (504) 589-2471

F o r in fo rm a tio n on sta r tin g a se ctio n , co n ta ct:

S ectio n s M a n a g er, A O A C IN TERNATION AL, 2 2 0 0 W ilson B lv d ., A rlin g to n , VA 2 2 2 0 1 -3 3 0 1 , USA,
+ 1 -7 0 3 -5 2 2 -3 0 3 2 , f a x + 1 -7 0 3 -5 2 2 -5 4 6 8 .
On why we do what we do. My family used to take long electrochemical immunoassay. I have also been very involved
car trips. We traveled Route 66 and learned about small in nuclear medicine and radiopharmaceutical development
town America before it was by­ with Ed Deutsch using electro­
passed by the interstates. I still chemistry7and chromatography.
love cars and travel. Later, a More recently, we’ve been working
summer job in Idaho convinced with polymer-modified electrodes
me that a chemist with a Ph.D. in the development of sensors and
has more opportunities. So I in EXAFS spectroelectrochemistry
veered at the next fork in (extended x-ray absorption fine
the road, went back to school, structure), in a collaboration
and proceeded toward today’s with Dick Elder.
destination.
On stating a p o in t o f view.
On electrochemistry as an Electrochemistry is becoming
analytical technique. more important because of
My next stop was a NASAfellow- the development of materials

Touring the two-lane blacktop

and research as a two-way street.
ship at the University of North Carolina where I had the and fabrication techniques that allow mass production
good fortune to study under Royce Murray who taught of disposable EC cells. We’ve worked on biosensors for
electrochemistry; which I found detecting glucose in diabetics and solid state electro­
to be an intellectually satisfying chemical cells that we hope to use to determine the oxy­
combination of math, physical gen status in premature infants.
phenomena and fairly complex One of our electrochemical
processes. immunoassay techniques has
reached a detection limit in
Now 1had my vehicle! But the the zeptomole range.
road called again and, after
obtaining my Ph.D., I went into I see fundamental and applied
William Heineman industry7where I worked on ion research as a two-way street. 450
W AVELENG TH, nm
selective electrodes and coulometry. Then, in completely Both are important and invigor­
unrelated developments I got married and left industry7for ating. The instrument companies BAS among them—
postdoc work with Ted Kuwana at Case Western Reserve supply the tools that allow us to focus on results, not the
in spectroelectrochemistry. After a year, Kuwana moved to complexity of the techniques.
OSU and 1kept going south on 1-71 to the University of
Cincinnati which has been home for more than 20 years. Southern Ohio still has intriguing two-lane roads where
it’s fun to take a swing in my old Triumph. But modern,
On where electrochemistry f its in. Though I’ve main­ computerized instrumentation is what makes work in
tained an interest in spectroelectrochemistry, for more than electrochemistry7the stimulating and productive field
a decade Brian Halsall and I have collaborated on studies of it has become. Serious sciencefor serious scientists.

B io an aly ti cal
Systems, Inc.

BASjapan •36-4 1-Chome•Oshiage,Sumida-Ku•Tokyo 131,Japan•PH (81)33624 0367•FAX (81)3 3624 0940

BAS Europe •Dellingstraat,34-1,B-2800 MechelenBelgium •PH (32) 1543 22 31•FAX(32) 1543 2385
BioanalyticalSystems, Inc. *2701 KentAve•WestLafayette,INUSA 47906•PH (317)46.3-4527•FAX(317)497-1102
CIRCLE 87 ON READER SERVICE CARD (-