168 CIERI: J. ASSOC. OFF. ANAL. CHEM. (VOL. 62, NO.

1, 1979)

High Pressure Liquid Chromatographic Detection and Estimation

of Furazolidone and Nitrofurazone in Animal Feeds
UGO R. CIERI
Food and Drug Administration, 2nd and Chestnut Sts, Philadelphia, PA 19106

The feed sample is extracted with acetone or connected to Spectroflow Model SF770 monitor
dimethylformamide-acetone ( l + l ) and the (Schoeffel Instruments Corp., Westwood, NJ

Downloaded from https://academic.oup.com/jaoac/article/62/1/168/5702220 by guest on 31 October 2020

filtered extracts are evaporated to dryness. The 07675). If monitor is not available, adapt Waters
residue is dissolved in chloroform and trans- instrument to 365 nm detector with appropriate
ferred to a silica gel column. The nitrofurans kit.
are eluted with methanol-chloroform (50-|-50). (b) HPLC Column.—/iBondapak C 18 , 3.9 mm
A portion of the eluate is evaporated to dryness id X 30 cm (Waters Associates).
and the residue is redissolved in a small volume (c) Silica gel.—For column chromatography,
of methanol. Aliquots of the methanolic solu- 0.05-0.2 mm mesh (Brinkmann Instruments, West-
tion are injected into a liquid chromatograph bury, NY 11590). Prepare column as follows: Plug
with a /iBondapak Ci8 column, using 30% ace- constricted end of 30 X ca 2 cm id glass column
tonitrile as the eluting solvent and ultraviolet with glass wool, add ca 3 cm silica gel, tamp to
detection at 365 nm. Several samples spiked compact, and overlay with plug of glass wool.
with 0.5—50 ppm furazolidone or nitrofurazone (d) Furazolidone standard solutions.—(1) Stock
and 2 commercial samples were analyzed by solution.—Transfer 40.0 mg furazolidone (Hess and
the proposed method. Clark, Inc., Ashland, OH 44805) to 200 ml volu-
metric flask; dissolve and dilute to volume with
dimethylformamide (DMF), stopper, and mix.
The nitrofurans furazolidone and nitrofura-
Store in dark and use within 1 week. (2) Working
zone are added to animal feeds in amounts rang- solution.—Dilute 1.0 ml stock solution to 100.0 ml
ing from 0.0008 to 0.022% (8-220 ppm) to with methanol. Prepare on day of use.
stimulate growth and to prevent and control (e) Nitrofurazone standard solutions.—(1) Stock
diseases. Colorimetric (1), polarographic (2), solution.—Proceed as described under (d), using
and quantitative thin layer chromatographic (3) nitrofurazone (Eaton Laboratories, Inc., Norwich,
methods have been published for determining NY 13815). Store in dark and use within 1 week.
these 2 compounds in feeds. These methods are (2) Working solution.—Dilute 1.0 ml stock solu-
generally not suitable or not reliable for sam- tion to 200.0 ml with methanol. Prepare on day
ples containing < 3 0 ppm of the drugs. A more of use.
sensitive method is, therefore, needed to de- (f) 'HPLC eluting solvent.—30% acetonitrile
(Burdick and Jackson Laboratories, Muskegon,
termine low levels of nitrofurans in feeds and
MI 49442J in water.
also to detect them in samples that may contain
trace amounts of the compounds as a result of Determination
accidental contamination. Because of its greater (a) Samples containing >2£ ppm furazolidone.
flexibility, high pressure liquid chromatography —Transfer 1 g finely ground sample to beaker,
(HPLC) seemed to permit higher sensitivity add 10 ml acetone, and stir well with glass rod.
than other techniques currently in use and, Filter with suction through coarse porosity fritted
consequently, was chosen as the basis of this glass crucible, retaining most of solid material in
study. beaker. Repeat extraction with 9 additional 10 ml
portions of acetone and filter each portion. Trans-
fer nitrates to 250 ml beaker and rinse flask with
METHOD
small portions of acetone. Evaporate to dryness
Apparatus and Reagents on steam bath, using air current. Rinse sides of
(a) Liquid chromatograph.—Model U6K uni- beaker with 20 ml CHC1 3 ; then swirl to dissolve
versal injector and Model 6000A solvent delivery all residue in beaker. Transfer ^solution to silica
system (Waters Associates, Milford, MA 01757) gel column and let solvent elute. Rinse beaker
with 20 ml CHC13 and add to column. Continue
rinsing beaker and eluting with small portions of
This report of the Associate Referee was presented at the CHC13 until green or yellow band is eluted but
92nd Annual Meeting of the AOAC, Oct. 16-19, 1978, at
Washington, DC. do not use >120 ml CHC13. Add 25 ml methanol-

0004-5756/79/6201-0168/08$01.00
© Association of Official Analytical Chemists, Inc.
 CIERI: J. ASSOC. OFF. ANAL. CHEM. (VOL. 62, NO. 1, 1979) 169

CHC13 (50+50) to column and collect eluate in

>25 ml volumetric flask, dilute to volume with
CHCI3, and mix. Transfer aliquot containing ca
2.5 ng furazolidone to 25 ml volumetric flask, add
CHCI3 to volume, and mix. Pipet 20.0 ml into
25 ml glass-stoppered Erlenmeyer flask (identify
as A) and 2.0 ml to another flask (identify as
B). Evaporate solvent to dryness on steam bath,
using air current. Add 1.0 ml methanol to flask
A and 1.0 ml furazolidone working solution to

Downloaded from https://academic.oup.com/jaoac/article/62/1/168/5702220 by guest on 31 October 2020

flask B, swirl to dissolve residue, and stopper. Set
wavelength of monitor at 365 nm, range at 0.01,
and flow rate at 3.0 ml/min. Inject 50 fil solution
A and, after furazolidone peak has appeared, let
chromatogram run until horizontal or nearly hori-
zontal line is obtained. Inject 2 more 50 nl ali-
quots of solution A and three 50 (A aliquots of
solution B, waiting in each case until horizontal
line is obtained.
(b) Samples containing <2.6 ppm furazolidone.
—Transfer sample containing 2.5 /ig furazolidone
to beaker; extract with 10 successive portions
of acetone, using > 5 ml acetone/g sample. Filter
each portion and continue as described in (a),
collecting column eluate in 25 ml volumetric flask.
(c) Samples containing >2.5 ppm nitrofurazone.
—Transfer 1 g sample to beaker, add 5 ml DMF, MIN
stir well with glass rod, add 5 ml acetone, and mix FIG. 1—Chromatograms of samples containing 2 ppm
well. Filter with suction through coarse porosity furazolidone (I) or nitrofurazone (II): B, standard solu-
fritted glass crucible, retaining most of solid ma- tion; A, sample. See text for HPLC conditions.
terial in beaker. Repeat extraction with 9 addi-
tional 10 ml portions of acetone, filtering each
portion. Evaporate combined filtrates to dryness Results and Recommendations
and continue as described under (a). Add 2.0 ml Separate chromatograms resulting from the
methanol to flask A and 2.0 ml nitrofurazone analysis of feeds containing 2 ppm furazoli-
working solution to flask B, and stir well. Inject done and nitrofurazone are shown in Fig. 1. The
three 50 fd aliquots each of solutions A and B, 2 compounds appear to have similar retention
using the operating conditions described above times and peak heights but these chromato-
but with a flow rate of 2.0 ml/min. grams were obtained by changing the flow
(d) Samples containing <2.5 ppm nitrofura-
rate of the eluting solvent. If the chromato-
zone.—Proceed as described under (b), using only
acetone to extract sample and collect eluate in graphic conditions are the same, nitrofurazone
25 ml volumetric flask. Add 2.0 ml methanol to elutes first, followed soon by furazolidone. The
flask A and 2.0 ml nitrofurazone working solution sample chromatogram contains some extraneous
to flask B, and stir well. peaks; these peaks are almost completely ab-
Measure height (cm) of each peak by extending sent if the sample examined contains > 5 ppm
horizontal line of later section of the chromato- nitrofuran whereas they become considerably
gram; see Fig. 1. larger at levels < 2 ppm. For this reason, 0.5
Average peak heights of 3 chromatograms of ppm is the lowest level that can be accurately
solution A (HA) and also 3 peak heights of solu- quantitated. However, even below this level, the
tion B (HB). Calculate ppm furazolidone (or nitrofurans can be detected. The monitor was
nitrofurazone) in sample as follows:
set at 365 nm because, at this wavelength, many
ppm =(HAXDX 2.5)/
[(HB — 0.1 HA) X g sample] The recommendations were approved by the General
Referee and by Subcommittee G and were accepted by the
where D = dilution of eluate, if any, after it was Association. Their reports will appear in the March 1979
made to volume. issue of / . Assoc. Off. Anal. Chem.
170 CIERI: J. ASSOC. OFF. ANAL. CHEM. (VOL. 62, NO. 1, 1979)

Table 1. Recovery of furazolidone or nitrofurazone Table 2. Analysis of commercial

from spiked feed samples samples for furazolidone
Furazolidone Nitrofurazone 55 ppm labeled 165 ppm labeled
content content
Found, ppm R e c , % Found, ppm Rec, %
Found, %of Found, %of
0.5 ppm Added PPm label ppm label

0.672 134 0.566 113 63.4 115 131.4 79.6

0.591 118 0.600 120 72.4 132 122.2 74.1
0.773 155 0.456 91.2 37.1 67.5 116.9 70.8

Downloaded from https://academic.oup.com/jaoac/article/62/1/168/5702220 by guest on 31 October 2020

0.515 103 0.551 110 50.1 91.1 131.1 79.5
Av. 0.638 128 0.543 109 Av. 55.8 101 125.4 76.0
Std dev. 0.111 22.3 0.0617 12.3 Std dev. 15.4 28.1 7.09 4.32

1.0 ppm Added

1.04 104 1.27 127 other compounds often present in feeds are not
1.08 108 0.894 .89.4
1.20
detected; peaks from inert constituents are also
1.22 122 120
1.07 107 1.19 119 considerably less than at lower wavelengths.
Av. 1.10 110 1.14 114 Samples were spiked to contain 0.5-50 ppm
Std dev. 0.0802 8.02 0.167 16.7 of either furazolidone or nitrofurazone. The 2
2.0 ppm Added compounds were not determined simultaneously
because feed generally contains only one nitro-
1.83 91.5 2.23 111
2.11 3.07 153
furan. Even if both compounds were present, it
105
1.93 96.5 2.11 106 may still be advisable to perform separate de-
1.29 64.5 1.43 71.5 terminations; simultaneous analysis of the 2
Av. 1.79 89.5 2.21 110 nitrofurans by this method is feasible only when
Std dev. 0.353 17.5 0.673 33.4 nearly equal quantities are present. Recoveries
5.0 ppm Added of the compounds from spiked samples ranged
from 42.8 to 155% and standard deviations
5.38 108 3.61 72.2
4.60 92.0 4.87 97.4 ranged from 8.02 to 38.7%; see Table 1. The
5.03 101 6.34 127 accuracy of the results apparently is not
5.92 118 6.83 137
affected by the level of nitrofurans in the feed,
Av. 5.23 105 5.41 108 as indicated by the fact that, even at the 25
Std dev. 0.558 11.0 1.46 29.4
and the 50 ppm levels, the standard deviation
10.0 ppm Added is high. Two commercial samples containing
4.87 48.7 5.75 57.5 furazolidone were also examined with approxi-
13.2 132 8.13 81.3 mately the same degree of accuracy; the results
7.62 76.2 9.00 90.0 are given in Table 2. This method is thus con-
12.0 120 7.12 71.2
siderably more sensitive and specific but not
Av. 9.42 94.2 7.50
Std dev. 3.87 38.7 1.40
more accurate than the published methods ( 1 -
3 ) ; the published methods should be used for
25.0 ppm Added samples containing high levels of nitrofurans.
20.1 80.4 26.6 106 The Associate Referee recommends that study
16.4 65.6 21.9 87.6
83.2 27.7
be continued to improve the accuracy of the re-
20.8 111
32.6 130 35.5 142 sults and that, once an accurate method is de-
Av. 22.5 89.8 27.9 112 veloped, it be subjected to collaborative study.
Std dev. 7.02 27.9 5.64 22.6
REFERENCES
50.0 ppm Added
(1) Official Methods of Analysis (1975) 12th Ed.,
49.2 98.4 21.4 42.8 AOAC, Washington, DC, seca 42.068-42.069
37.1 74.2 49.3 98.6 (2) Moore, H. P., & Guertal, C. R. (1960) J.
71.3 143 59.5 119 Assoc. Off. Anal. Chem. 43, 308-309
62.8 126 37.5 75.0 (3) Cieri, U. R. (1978) / . Assoc. Off. Anal. Chem.
61, 92-95
Av. 55.1 110 41.9 83.8
Std dev. 15.1 30.3 16.4 32.7
Received July 11, 1978. Accepted September 27, 1978.