1556 ANALYTICAL CHEMISTRY

mixtures. The phyaical methods-e.g., thermal conductivit L , of the molecular weights of the various components is not very
refractometry, et?.-ordinarily applied to the analysis of these large.
fractions do not distinguish the possible presence of a third con- COlVCLUSION
stituent and the analytical values may contain large and com- The method that has been described brings reasonably accurate
pletely unsuspected errors when such a contaminant orcurs. analysis of binary mixtures within the compass of effusiometric
Even when it does not allon total analysis of the mixture in- work. Rather more important, it is one of very few purely physi-
volved, the effusiometric examination can still shed some light
cal methods (other than those based on spectrometry and spec-
on the degree of complexity of the sample. trophotometry) that permits of even approximate quantitative
Other Applications. If a specimen of a completely unknown
analysis of polycomponent mixtures and is also capable of fur-
g x i is a t hand, it is advisable to determine the proportion of effu-
nishing some indication of the qualitative make-up of such mix-
sion a t a minimum of two time intervals. Using Equation 2 iii tures. The proposed method of analysis should be particularly
the interpretation of these data, we secure two or more inde- valuable for the examination of any group of gases that are not
pendent cvsluations of the apparent molecular weight of the
readily separated from one another and that show suhstantial
original specimen. If the vwlues 10 obtained are all concordant, gradations in their niolecular weights. Two such groups are the
the molecular weight ot the g:ts, and thence a valuable clue to its paraffin hydrocarbons and the inert gases.
identity, is derived. (Thi?, of course, does not exclude the possi- This exploratory investigation has been conducted with rela-
hility that the sample is a mixture of two gases of nearly identical tively crude equipment, and the experimental techniques could be
molecular weight.) On the other hand, a systematic variation in cwnsiderably refined. Thus, x n increase in the size of the effusion
the value9 of the apparent molecular weight of the gas provides a vessel, the size of the sample, and the duration of the experiment
definitive indication of the complexity of the sample. The should greatly improve the accuracy of all the measurements, with
divergence of the bracketed results in the fourth column of Tiihle a corresponding improvement in the quality of the analytical re-
I1 suggests the magnitude of the corresponding variation in thc sultq. Furthermore, i t appears possible to replaw the present
apparent molecular weights calculated by the application of method of measuring the proportion of effusion, in terms of the
Equation 2 to a mixture containing molecular species of divcme residual quantity of gas in the effusion chamber, by an alternative
we i ghtm .
technique that would not require a separate trial for the deter-
When incomplete analytical data are available from some other mination of each individual proportion. Thus, one might use an
hource(s) there ensues a noteworthy simplification, and siniul- indirect hut semicontinuous method of assaying the proportion
taneous increase in power, of the effusiometric methods. of effusion, measuring the quantities of gas that pass through to
-4s a concrete example, consider the problem of identifying a the low-prcssure side of the orifice. The authors hope to in-
minor component that is not absorbable in volumetric gas ana- vestigate the effect of these improvements. However, the effusi-
lytical procedures. From such procedures the complete quantita- ometer described will do practically everything (except handle
tive make-up of the sample, and the qualitative character of all
but the unabsorbed component, would be known. The deter- condensable vapors) that can he done by any previously described
mination of the molecular weight of the unknown constituent can effusiometer, and it is capable of many other determinations that
then be based on a single measurement of the effusion rate. no previous effusiomrter has brought even remotely within the
The "identification" of the argon in air provides an exemplifi- range of possibility
cation of this possibility. Assume that it has been found that air
consists of the following components in the indicated proportions:
ovygen 21.0; nitrogen 78.0; component X 0.9. A single effusion ,iCKYOWLEI)C;MIE\IT
trial made with air that had been passed over Ascarite gave PI!Pc
a q 0.5811 after 5 minutes For a ternary mivture n e may write: The authors are indebted to E. H. deButts and It. E. Lundin
for a number of helpful suggestions.

LITEK4TURE CITED

ill1 the mole fractions, and two of the three molecular weights, (1) Benson, S. W., and C'aswell, S.,J . Phys. Colloid. <'him,, 52, 1332
are known. Solution for the third molecular weight then gives (1948).
(2) Buckingham, E., and Edwards, J. D., U. S. But. Standards. S a .
Paper 359 (1920).
(3) Edwards, J. D., U. S. Bur. Standards, Tech. Paper 94 (1917).
(4) Kemp, L. C., Collins, J. F., Jr., and Kuhn, W. E., IND.E s ~ .
CHEM., ANAL.ED.,7,338 (1935).
(5) Knudsen, M., Ann. Physik, 28,75 (1909).
from which the molecular weight of component X is calculated as ( 6 ) Nash, L. K., ANAL.CHEM.,20,258 (1948).
39.3. The closeness of this approximation to 40, the molecular (7) Rosenberg, P., Rei. S c i . Instruments, 10, 131 (1939).
weight of argon, is rather striking when it is recalled that less than
1 of the unknown constituent was present, and that the spread K E C E I V E D May 11, 1950.

Colorimetric Determination of Ethyl Alcohol

M.4X R. WILLIAMS AND H. DARWIN KEESE, Oregon State College, Corvallis, Ore.

T H E importance of determining ethyl alcohol in aqueous solu-

tions can be exemplified by the 200 or more references appear-
ing in the past 12 years concerning this analysis. Most of the
has long been realized that an accurate colorimetric method for
determining ethyl alcohol would be very desirable in order to
eliminate troublesome titrations and preparation of rather un-
methods employ potassium dichromate and sulfuric acid to oxidize stable standard solutions, some of which should be checked every
the ethyl alcohol to acetic acid and determine the excess dichro- day. Comparatively few attempts have been made, however, to
mate by a titration with standard ferrous ammonium sulfate determine this important compound colorimetrically, and almost
( 2 . 8, IO, 18, 1 9 ) or by an iodonietric titration (1:.9, 26, 86). I t every method involves the estimation of the excess potassium
V O L U M E 22, NO. 12, D E C E M B E R 1 9 5 0 1557

.Isimple, accurate spectrophotometric method is described for the determina-

tion of 0.0010 to 5.800 mg. of ethyl alcohol per nil. in aqueous solution.
The procedure is based on the quantitati\e oxidation of the ethyl alcohol by a
potassium dichromate-sulfuric acid reagent and dilution of the resulting
solution to a concentration where the chromic ions will not interfere with the
\ iolet-colored complex developed with the excess dichromate when s-diphenyl-

rarhazide is added. This complex follows Beer's law accuratel? at 540 mp.

iIi(~hroiiiatt~(4-3, 1 7 , 23-25), These colorimetric. inrthods have I alcohol pei nil. of solution. Ilenio\e ;tliquots froin thib Ytoch
it h j

the common fault that Beer's law is not oheyeti because one of >ample and dilute to make up the various standard ethyl alcohol
Yolutions with concentrations from 0.0010 to 5.800 mg. pel nil
the osidation products, chromic ion, has Y I I intense green color
xhich is mixed in varying concentrations with the exccw yellox The large range of ethyl alcohol conc~ntrationsdetermined by
cwlored dichromate ions that are being determined. this new method necessitates various dichromate reagents. The
LIost colorimetric methods attempt to overcome this dificwlty (*oncentrationsof potassium dichromate in the reagents, the COII-
I)y visual comparison of the urikiiowii ethyl alcohol sample with :i iw~tratioriranges of ethyl alcohol for which these reagents were
series of tAthyl :tl(:ohol standards aft,rr oxidation of both with PS- vmployed, and the theoretical upper limit of ethyl alcohol for ear11
dichromate reagent (26). The :trcuracy of this method. reiigent are given in Table I
however, is approximately 10% with an ethyl alcohol range of
0.1 t o 2.0 mg. per ml. Other methods use various filters in at- _____ __ - __
tempts to absorb most of one color ( 6 ) ,and one method involves
'l'ahle 1. Ethyl llcohol Concentration Ranges for Various
multiple wave-length analysis (4). These latter methods give Dichromate Reagents
only fair results, are usually too complicated for practical and I lichromatr Potassium Ethyl Alcohol
rapid analyses, and, moreover, are not applkahle for ethyl alcohol Reagent Dichromate Ethyl A l r o h i i i Theoretical
YO. Solution Range r p p e r Limit
mncent,rations much below I mg. per ml. Grams/100 ml. Mg./ml .Ug./ml
The following method utilizes a n e w technique to iiullify the 1 5.0000 1.000-5.800 5,880
interference of the chromic ions completely, so that an accurate 2 1.0000 0.2t50-1.150 1,170
2 0.1000 0.050-0.100 0.117
colorimetric determination of the ewes8 dichromate ions is 4 0.0200 0.0010-0.0100 0.023
feasihle. The sensitivity of this method makes it possihle to
:irialyze 11 wide ra.nge of ethyl alcohol concentrations.
'The highest concentration of ethyl :ilcohol which can be deter-
PRINCIPLE niined hy reagent 1 is limited only hy the solubility of potassiuiii
Beer's law is not obeyed for the variatilc colored system re- Ilichromate in concentrated sulfuric acid at, 20' C. The masi-
sulting from the oxidation of ethyl alcohol by potassium dichro- mum concentration which can be determined by any dichromate
mate in sulfuric acid. However, when this colored solution is reagent is computed from t,he theoretical reaction, Equation 1 .
tiilut,ed to a chroniiuni concentration of less than 1 mg. per liter,
2K,Cr2O7 + 8HzSOa+ 3CH8CHp0H +
2KzS04 + 2Crr(S04)s+ 3CH&OOH + llHOH
neither the chromic nor the dichromate ions show appreciable
i1)
absorption for wave lengths in the visible region. s-Diphenyl-
cmbazide, when added to this dilute solution, forms a deep violet- The lowest ethyl alcohol concentration which can be deter-
colored complex with the dichromate ions, but no color with th(3 mined by any reagent is that which reduced so little dichromat(>
vhromic ions. This violet-colored complex accurately follows t.hat thc spectrophotometric reitdings of the violebcolored com-
' s :md a precise detwmination o f ethyl alrohol is possihh,.
R ~ ~ e rlaw plex closely approach those of the blank which cont'ains no ethyl
:tlcohol.
REAGENTS Other coiicentrations of ethyl alcohol which will fall het,weeti
t Iw ranges given can be determined by preparing appropriate
s-Diphenylcarbazide. The reageiit was obtained from thiz
Paragon Testing Laboratories, Orange, N. J. (now the Matheson ilic,hroniat,ereagents which will leave a slight excess of dirhromate
Co., Inc., Paragon Division, East Rutherford, K. J.), and is :ifttJr the oxidation reartion is complete.
suitable for use wit>hout furthcr purification. This reagent is
xvailable elseyhcre. iPP.4R 4TU S
Solution of s-Diphenylcarbazide. Saturate 95% ethyl alcohol
with s-diphenylcarbazide and allow to st:tnd until the solid has All ineasurements of absorptioIi were made with a Beckri1a11
settled, leaving a clear slightly yellow-colored solution ready for LIodel DU quart,z spectrophot'ometer using I-cm. Corc:x cells.
use. A fresh solution should be prepared every week, or morc (?rdinary good grade volumetric flasks, pipets, burets, and Imo-
often should a pink color develop in the reagent. .ilicate glass microburets were cinployed.
Potassium dichromate, analytical reagent grade.
Sulfuric acid, analytical reagent grade. GENERAL PROCEDURE
Potassium Dichromate-Sulfuric Acid Reagent. Weigh out thc
wquisite amount, of dried potassium dichromate, transfer to a .Igc~iic~t~alprocedure is applicahle to ariy one of the dichromat(.
glass-stoppered volumetric flask, niid add concentrated sulfuric rthagents given in Table I. 13ichrotnate reagent 2 is chosen as an
:wid nearly to t,he mark. JT7~irmthe solution slightly, and shake
until the crystals dissolve. Bring the contents of the flask to i 1lust rat ion,
20" C., add sulfuric acid to the mark, mix t,horoughly, and store
in t,he tightly stoppered volumetric flask. This reagent is stahle lhactly 10.00 nil. of ethyl alcohol solution are pipetted into a11
indefinitely. 8-inch borosilicate glass test tube and 5.00 ml. of dichromate
Absolute ethyl alcohol, obtained from Commercial Solvents reagent 2 are added slowly from a 10-ml. microburet as the mix-
Corporation, 110 Sutter St., San Francisco 4, Calif., or elsewhere. t,ure is constantly being shaken. The tube is stoppered with a
Ethyl Alcohol. Prepare a stock samplr of ethyl alcohol solu- one-hole rubber stopper containing a short length of capillary
tion by measuring out accurately at 20' C. with a 10-ml. micro- tubing, immersed in a boiling water bat.h, heated for 2 minutes:
buret the proper volume of absolute ethyl alcohol, into a I-liter :ind cooled in cold wat,er. The reaction mixture is carefully trans-
volumetric flask, which volume, when multiplied by the den- i'cwed to a 2-liter volumetric flask, tho test tube is rinsed several
sity of nhsolute ethyl alcohol at 20" C., \\ill represent 10 nig of times with distilled water, the solution is made to the m:irk a t
1558

20" C. with water, and the contents of the flask are thoroughly
shaken. An aliquot of 10.00 ml. is transferred with a pipet into a
100-ml. volumetric flask, 3.3 ml. of 6 -V sulfuric acid and 1.0 ml. of
s-diphenylcarbazide reagent are added, and the solution is
brought to the mark a t 20" C. with distilled water. This solution,
when thoroughly shaken, develops a pronounced violet color, the
optical density of which is determined a t 540 mp within 10 minutes
100 , ANALYTICAL CHEMISTRY

against an s-diphenylcarbazide-sulfuric acid solution as :t blank.

It is recommended, but not necessary, that a blank be run for

each series of experiments in which distilled water can be sub-
stituted for the ethyl alcohol solution. If a blank is run, it is not
necessary to make up the dichromate reagent of exact weight.
The reaction mixture is always diluted with distilled water
according to the original chromium concentration in the dichro-
mate reagent used, so that the final solution will contain less than
1 nig. per liter of chromium.
Slthough the intensity of the s-diphenylcarbazide-dichromate
color is influenced very little by moderate pH changes, the sulfuric
acid concentration in the final solution should be approximately
0.2 S; sulfuric acid must be added to the solution when making
the final dilution in order to attain this final acid concentration. 50 I ' '
400
' I ' "
500
' I ' ' I
600
' ' '
Much higher acid concentrations result in more rapid fading of WAVE LENGTH, mp
the violet color, whereas lower acid concentrations cause the color
Figure 1. Spectral Transmittance Curves
to be developed slowly (21).
1. 0.4 mg. C r + + +per liter of solution in 0.2 ,V H I S O , and excess
The authors consider the measurement of the optical density s-dipheuylcarbazide
2. 0.4 mg. chromium (as K*Cr?Oi) per liter of solution in 0.2 %
within 10 minutes after the addition of the s-diphenylcarbazide HISO, and excess s-diphenylcarbazide
necessary for accurate determinations, because a very slight but
definite fading of the color always takes place on standing.
The amounts of s-diphenylcarbazide reagent used, 1 ml. per The concentration of ethyl alcohol in an unknown alcohol solu-
100 ml. of solution, are such that a large excess of this compound tion ni:ty then be calculated from the expression
is always present in the final solution.
The dilutions and computed dilution numbers required for each A = - -d-( B - c )
of the four dichromate reagents employed for determining the aF
large range of ethyl alcohol, and the amounts of dichromate re- where .4 = concentration of ethyl alcohol, mg. per ml.
agents, 6 N sulfuric acid, and s-diphenylcarbazide are given in B = concentration of chromium in diluted blank, mg. per
T:thle IT. liter
c = concentration of chromium in diluted sample, mg. per
liter
d = dilution number (Table 11),computed by multiplying
total volume in liters of first dilution by subsequent
Table 11. Amounts of Reagents Used in Ethyl Alcohol dilution factor of aliquot
Determinations F = theoretical factor, 1.5053, computed from Equation 1,
Dichromate 6 A' S- relating unknown ethyl alcohol concentration to
Reagents Dilution of Dilution Sulfuric Diphenyl- mg. of Cr reduced
so. Reaction Mivture No. ( d ) Acid carbazide (I = alcohol solution used, nil.
Ml. MI. M1.
1 5.00 (1) Reaction rnixtiire 100 16.7 ;1.0
t o 2 liters DISCUSSION AND RESULTS
(2) 10 mi. t o SO0 1111.

2 5.00 (1) Reaction mixture 20 3.3 1 0 -4typical absorption curve of potassium dichromate in 0.2 .V
t o 2 liters sulfuric acid and excess s-diphenylcarbazide is given in Figure 1.
(2) 10 ml. t o 100 i d .
Also is shown a curve for approximately the same concentration of
3 5,OO (1) Reaction mixture 5 3 3 1 0
t o 500 ml. chromic ion in 0.2 iV sulfuric acid and excess s-diphenylcarbazide.
(2) 10 ml. t o 100 inl. I t is clearly seen that only the dichromate ions involve an absorp-
4'' 2.50 (1) Reaction mixture 0.5 1.7 5.0 tion in the visible region and the chromic ions give absolutely no
t o 500 ml.
interference. The slit width used ranges from 0.1 mm. a t 400
5.00 ml. of ethyl alcohol solution used. nip*,and 0.04 mm. a t 540 mp to 0.15 mm. a t 620 mp.
The violet-colored complex follows Beer's law strictly for a
chromium concentration below 0.8 mg. per liter of solution :It
CALCU LATIOSS 540 nip, the wave length of maximum absorption.
A series of experiments was made to determine whether or not
The concentration of chromium i n a diluted solution is citlcu- the potassium dichromate reagents used in the ethyl alcohol
:tted by means of the Bouguer-Beer law, analysis could be diluted to less than 1 mg. of chromium per liter
with sufficient accuracy for reproducible results. Previous
c = LYE (2)
studies by Mehlig ( l l - l d ) , Mellon (15),Murray and Ashley (16).
where c = concentration of chromium, mg. per liter, in the Drabkin and Austin (S), and others on the determination of
diluted solution macro amounts of materials using dilution methods indicate that
E = memured optical density a t 540 mp no difficulty should be experienced in the present problem.
K = -'O0O M , a constant determined on our instrument
€1 Exactly 5.00 ml. of the dichromate reagent were added from a
for known dichromate solutions 10-ml. microburet to 10 ml. of water, the mixture being con-
M = atomic weight of chromium stantly shaken. The resulting solution was heated for 2 minutes
1 = light path of absorption cell, cm. and diluted, and the analysis was completed according to the stand-
e = molecular extinction coefficient ard procedure given above.
V O L U M E 2 2 , NO. 12, D E C E M B E R 1 9 5 0 1559

rounded by an ice-water bath, The typical result for the longest

Table 111. Effect of Heating Time heating time is shown and compared in Table 111. This indi-
Ethyl Ethyl cates that an ice bath is not necessary and that ethyl alcohol is
Alcohol Alcohol
Experimental Conditions Present Found Difference not volatilized during either the addition of the dichromate re-
Mg./ml. Mg./ml. il.lg./rnl. agent or the subsequent steps in the standard procedure. Simi-
No heat 1.000 0,990 -0.010 lar results were obtained with other concentrations of ethyl
Heat 0.5 minute ... 0.998 -0,002
Heat 2 minutes ... 1.000 0.000 alcohol,
Heat 10 minutes ... 1 ,000 0.000 Standard ethyl alcohol solutions were made to contain various
Ice bath, heat 10 minutes ... 1.000 0.000
concentrations for the particular ranges and various dichromate
reagents listed in Table I. These solutions were analyzed by the
standard procedure and the results are given in Table IV. They
Thr theoretical chroniiuni concentration expected from reagvnt indicate that the reaction between the dichromate reagent and
2 was 0.883 mg. per liter. During several months' study of this the ethyl alcohol in the solutions is quantitative, under the condi-
problem the over-all average found experimentally was 0.881 mg. tions of the experimental procedures, according to Equation 1.
p"r liter. The maximum deviation from the theoretical concen-
trat,ion was +0.009 mg. per liter.
These results clearly illustrate that accurate dilutions can Table IV. Calculated and Experimental Chromiurn
easily be made using only ordinary volumetric equipment and Reduced
good quantitative technique. In a standard ethyl alcohol deter- Chromium Reduced b y
mination, the authors have found that the use of a blank tends to 10.00 M I . E t h y l Av., *
Ethyl Dichromate *lcohol Solution Chromiuiii
diminish the experimental errors. C u e should be taken, how- hlcohol Reagent Experi- Reduced,
ever, that the pipets and burets are thoroughly clean and that all Concn. Used Calcd. mental Experimental Difference
Mg./ml. Mg. JIQ. M Q. Vg.
work is done with the solutions near 20" C.
3,800 1 $!.31 86.9 86.9 -0.4
The effect, on the ethyl alcohol determination of various heat- 5,000 1 (3.27 76 4 75.5 +0.2
ing times for the reaction mixture was determined by making 1 74 7
4.000 1 60.21 61.2 61.2 $1.0
several runs on a standard ethyl alcohol solution containing 1 mg. 3.000 1 43.16 45.2 45.5 +0.:1
1 4.5 ,
per nil. These results, t,abulated in Table 111, indicate that 2,000 1 30.11 30.1 30.1 0.0
heating time is of little consequence, although heating in a water 1,150 2 17.31 17.2 li.2 -0.1
1 000 1 15.05 13.00 15.00 -0.05
bath for at least 2 minutes gave slightly better results than no 2 15.00
0.750 2 11.29 11.16 11.27 -0.02
heat. 2 11.37
Because the ethyl alcohol solutions become hot when the di- 0 , 3 0 0 2 i , 5 3 7 . 4 0 7.48 -0.05
2 7.56
chromate reagent is added, the possibility of the unreacted ethyl n 250 2 3.76 3 82 3.79 +0.03
2 3 78
alcohol being volatilized indicated a likely source of error. -IC- 2 3.78
cordingly, several runs were made in which the dichromate reagent 0 100 3 1 . 3 0 1.48 1.33 +o 03
3 1.56
w:is added to the thoroughly chilled tsthyl alcohol solution sur- 3 1 . ,56
0 030 3 0 73 0.76 0.76 +0.01
0.0100 4Q 0 150 0.139 0.133 -0.017
4 0.127
4 0.138
4 0.129
0 ,0030 4 O.Oi3 0.055 0,057 -0,018
4 0.058
0,00"0 4 0,030 0,029 0.026 -0.004
4 0.030
4 0.023
4 0.022
0.0010 4 0.015 0,006 0.011 -0.004
4 0.006
4 0,003
4 0.021
5.00 ml. of ethyl alcohol ured with dichromate reagent 4.

The first column in Table 11- gives the wide rnngc of ethyl
alcohol solutions used, while column two shows the particular
dichromate reagent used in each experimental determination.
Chluinn three is the theoretical amount of chromium reduced by
rach solution computed according to Equation 1. Column four
s h o w typical experimental results of the amount of chromium
reduced as determined by the procedure set forth in t'his paper.
Plots of the ethyl alcohol concentrations against the milligrams of
chromium reduced are shown in Figures 2 and 3.
Figure 2 gives an experimental curve for determining ethyl
alcohol concentrations of solutions between the range of 0.050 to
5.800 mg. per ml., and Figure 3 shows a continuat,ion of the curve
for concentrations below 0.050 mg. per ml. with the coordinate
axes increased over one hundredfold for clarity. The straight
line in Figure 3 is the theoretical curve as computed from Equa-
tion l for this range, and the small circles indicate the experi-
mental data.
In Table V are given the experimentally determined ethyl
alcohol concentrations computed from Equation 3 as compared
Figure 2. Potassium Dichromate Reduced by Various with the standard solutions.
Ethyl Alcohol Concentrations -4study of the plots and Tables I V and V indicates that the
 1560 A N A L Y T I C A L CHEMISTRY

accuracy of determining this important compound is very satis-

factory, especially in view of the errors existing in the present Table V. Deterniinatioii of Ethyl Alcohol
methods ( 7 ) . However, the analyses become somewhat erratic Ethyl Alcohol Ethyl Alcohol
Concentration Found Difference
brloiv concentrations of 0.0100 mg. per ml. lfg./mL. Jfg./ml. Mu.
.5 800 5 77 - n . nA
_.
.5 ,000 3.02 +o.oz
4,000 4.06 +0.06
3.000 3.02 f0.02
2,000 2.00 0.00
1,150 1.14 -0.01
1.000 0.997 -0.003
0.750 0,749 -0.001
0.500 0.497 -0.003
0.250 0,252 +0.002
0.100 0.102 +0.002
0,050 0,050 0.000
0.0100 0,0088 - 0.0012
0.0050 0,0039 - 0.0012
0.0020 0.0017 - 0.0003
0.0010 0.0007 - 0.0003

Interf~rc~nws with s-dipheii?'lcarbazide are very fev . 1,:ti g ~ >

quantitirs of molybdenum (VI), mercury (I and 11),iroii (111).
:ind wnadiuni (V) give some color with this reagent, but, rsrrpt
at low :widities, the reactions are not sensitive. Under the c w i t i i -
tions described in this paper small quantities of these metah \vi11
not interfere; moreover, their prrsence in an ethyl alcohol deter-
inination is not likely. Reasonable amounts of neutral 5 i t l t h
(a%) :trv without effect on the c~ilorof the complex (20, 2 1 ) .
COLOHIV ETHI

Ilthougli the present study was done \rith a Bcckman spectro-

photometer a t 540 mp, it ha%been th(, authors' experience that
nearlv romparable results may htt obtained using a good color-
imeter. The Klett-Summerson colorimeter, for example, is par-
ticularly common and may he used very \+ell,for one of the three
filters supplied with the instrunwnt hy the manufacturer (No. 54
filter) transmits a band of light centtlring at 540 mp. The authors
have PV(W usrd a Duboscq coloriinrtc~r. ~ \ i t hfair results.
oov ' 0,:o 20 .do I io ' .do ' .do I o;. ' .io
MILLIGRAMS CHROMIUM REDUCE0
CONCLUSIOY s
Figure 3 . Potassium Dichromate Reduced by L o w
Ethyl Alcohol Concentrations Thr method described is n new c'oloriinr~tric~ method frir I I I V
Line represents theoretical values a8 computed from Fguation 1. ~lt,termiii:rtioiiof ethyl alcohol in :Iqueous solutions by oxidation
Circles represent experimental data \vith potassiuni dichromate in th(, prwencr of concentrated SUI-
furic acid, eliminating thr interfererice of chromic ions by dilutioii,
One difficulty appearing here is that the re:tgeiit grade sulfuric :ind drtrwnining accurately the excess dichromate ion concentr:t-
:wid has enough oxidizable substances in it to cause blanks run tion by adding s-diphenylcarbazide and measuring the ahsorptioii
with dichromate reagent 4 to show progressive decrease in chro- 111the rrsulting violet-colored complex a t 540 mp,
iiiiurn concent,ration as heating times are increased. It is impor- This niethod employs the use of simple volumetric equipment
tant, therefore, in analyzing these very dilute ethyl alcohol solu- ;tv;tilable in most chemical laboratories and a spectrophotometer
tions that blanks be run and heated for t,he same length of time or photoelectric colorimeter. The nTide range of concentrations
:is the ethyl alcohol solutions. This difficulty is not, encountered that can be analyzed makes this method easily adaptable to the
with reagents 1, 2, and 3 hrcausr the osidizahle impurities are so tietermination of ethyl alcohol in wines, fermentation mashes,
minute in comparison with the higher chromium concentrations biological fluids, plastics, fuels, esplosives, pharmaceuticals, fruit
t,hat the resulting rrduction of the chromium is small and does not products, lacquer solvents, and many other products.
:Lffert the original chromium cwiicentr:ttion xithin the limits of Because this method is of such general application, the authors
the experimental twor. :irv at present investigating the determination of glycerol :tnd
Even more dilute solutions could probably br aiialyzed success- othrir ra.dy Irducihle compounds by this procxdure.
fully, if precautions are taken to remove the t,races of oxidizable
substances present in tht. sulfuric : i d beforr prrparing the di- LITER4TUHE CITEI)
chromate reagents. : 1 ) Bonner, T. G., Snalysf, 72, 4 i (1947).
(2) Chaikelis, -4.S., and Floersheim, R . I]., A m . J. CEin. Path. Tech.
INTERFERENCES Sect., 10, 180 (1946).
( 3 ) Drabkin, D. L., and Akustiii..I. H.. ,J. Bid. Chem., 112, 105 (1935).
The method of determining rthyl alcohol presented in this (4) Fonteyne, R., and d e h e t . P., Mikrochemie, 13, 289 (1933).
paper suffers from the general drnn-hack that other alcohols and (5) Gaudry, R., R e t . cutz. [)id.,3, 328 (1944).
(6) Gibson, J. G., and Blot,iier, H., J . B i d . Chem., 126, 551 (1938).
other easily oxidizablt: compounds must, be ahsent or removed. (7) Henrich, W., K h . Toclischr.. 20, 148 (1941).
This difficulty, of course, is present i t l w in all the oxidation meth- 18) Johnson, G. W , , and Gibson. R. E . , J . Lab. Clin. Med., 26, 399
ods described in the literature for thc, determination of ethyl (1940).
alcohol in blood, biological fluids, wines, mashes, fuels, explosives, (9) Koselka, F. L., and Hine, C. H., IND.ENG.CHEM.,ANAL.ED.,
13, 905 (1941).
etc. However, separation from these interfering suhstnnces is 110) Labourasse, G., Bull. assoc. c h h . , 58, 161 (1941).
not difficult except in the presence of alcohols o f low nirilr~rul:i~~ I I ) Jlehlig. J. P., IND.ESG.C H E MAN.IL
1 ., ED.,7, 387 (1935).
weight (9). $ 1 2 ) Ibid., 9, 162 (1937)
V O L U M E 2 2 , NO. 1 2 , D E C E M B E R 1 9 5 0 1561

(13) hlehlig, J. P.. and Hulett, H. It., Ibid., 14, 869 (1942). (22) Schulek, E., and It62sa, l’.,Z. nrtnl. Chent., 117, 400 (1939).
(14) Mehlig, J. P., and Shepherd, hl. *J., Jr.. ANAL. CHEM..21, 641 123) Shapiro. .J,, .4m.J. Clin. Path. Tech. Sect., 12, 66 (1942).
(1949). (24) Smith, L., Sceiisk Kem. Tid., 56, 153 (1944).
115) Mellon, M. G., IND.ENG.CHEM.,*\SAL. ED., 17, 81 (1945). (25) Snell. F. D.. and Snell, C. T., “Colorimetric Methods of .-\nalp-
(16) Murray, W. M., and .ishley, S.E. Q., 10, 1 (1938). sis.” Vol. 11, K e w Tork, D. Van Sostrand Co., 1937.
(17) Newman, H. W., and Abramson, h l . , J . Pharmncol. Ezptl. (26) Thouvrnin. R., Ann. chim. nnnl., 26, 72 (1944).
Therap., 74, 369 (1942).
(18) Sicloux, M., Ann. fermentalions, 1, 513 (1936).
(19) Pudovik, A. N., and sinaiskii, G. M.,J , ~ ~ ~ l i ~ RECEIVED
d April 17. 1950. Presented beforr t h e Section of Analytical
(U.S.S.R.), 21, 862 (1948). Chemistry a t the Pacific Northwest Regional Meeting of t h e AMERICAN
(20) Rowland, G. P., IND. E r o . CHEW,ANAL.ED.,11, 442 (1939). Richland, Wash.. J u n e <I and 10, 1950. Published
C H E r i c a L SOCIETY,
(21) Sandell, E. B.. “Colorimetric Determinations of Traces of with the approval of the Oregon State College Monographs Committee.
Metals,” pp. 190-4. Sew York. Intersrienre Publishel S, 1944. Research Paprzr lfi2, Devartrnent of Chemistry. School of Science.

Folin Colorimetric Procedure for Amino Acids

N. tIOWEl,l, FUR\lAN, GEORGE H. MORRISON’, AND ARTHUR F. WkGNER
Princeton l‘nirersity, Princeton, N . J .

\ careful stud! indicates that the Englis and Fiess modification of the Folin
colorimetric method for the estimation of amino acids is not adequate f o r
mixtures of \arious types of these acids. The disparitj of the extinction-con-
ceritration curies for amino acids is attrihuted to inductive effects of the various
radicals of the different amino acids, and in case of the basic amino acids, to
the participation of groups other than the alpha-amino group in the formation
of colored complexes. The method is accurate as applied to many indiiidual
amino arids and should he useful H ith effectii c separation procedures.

T HIS study was made in connection u it11

a survey of methods that might be
u>ed for the estimation of individual amino
:icids or their mixtures in chromatographic,
Ytudies. The original procedure of Folin ( 5 )
i n d the subsequently modified form of the
procedure by Englis and Fieis ( 4 ) were the Mqo( N
hasis of the present worh. This color]- E-0042
metric process offers advantages for routine D-0.028
k\ork and i t is based upon the reaction be- c -0.02 I
tween an amino w i d and sodium-2-naphtho-
quinone-4-sulfonate, the product being in
w c h case a highly c-oloiet1 cwniporintl.

ri,
0

-OH
w\/
SI’ Wavelength, m p
Figure I. Extinction-Concentration Study for Gl? cine from $00 to
R--VH (’OOIl 600 nip
Volurnea in all raws 50 rnl.

Seither of the papers referred to gave a comparative survey of amino avida was again noted. This fact \vould not militate
the behavior of the individual amino arids through the clesiretl against the use of the method in connection with efficient chro-
c~trnc.cntrationrange, 0 to 0.05 mg. of amino nitrogen in ;iO-nil. matographic separations, but it does preclude its use for an accu-
final volume of the solution after color development. Tlif rate estimation of a-amino nitrogen in mixtures of several amino
method was further studied by Frame et al. (6)withafilterphotom- acids. The wide variations in color are apparently due to the
(Iter with special referrnc,c to the cletermination of amino nitro- varying inductive effect of the group R of the amino acid. This
~ c nin blood. A comparative study of the method with thc hypothesis is considered further under discussion of results.
ninhydrin manometric. method has been made by Chinard and Block ( I ) , Bull et nl. (Z), and Moore and Stein ( 7 ) found that
\.a11 Slyke ( 3 ) . various amino acids gave different extinction values a t the same
In the course of the studies here reported the wide variation in molar concentration by the photometric ninhydrin procedure
the slopes of the extinction-concentration curves of the various according t o their iniproved technique. In t.his case the same
’ Present address. U. S. Atomic Energy Commission, New Brunswick, N. J. colored product. t l i k e t o h \ d r i n t ~ ~ l i d e n e ~ i k e t o h y d r i n d a misi