June 15, 1942 ANALYTICAL EDITION
the weight azotometer, with satisfactory results. The sample
TABLE OF ACETASILIDE
I. ANALYSIS
(Calcd. X = 10.3717,)
Series I Series I1 Series I11
% 7c %b
10.27 10.25 10.33
10.2s 10.24 10.34
10.263 ... IO.32
A\.. 10.29
mined in grams of mercury and is subtracted from the weight
of displaced mercury (the loss in weight of A ) . The percent-
age of nitrogen is given by the formula:
yo s =
100 x (wt. of Hg displaced - n t . of blank) X NZreduction factor
density of Hg X wt. of sample
Results
Acetanilide is the only substance whose analysis is reported
in this paper, inasmuch as the method Only with the cO1-
lection of nitrogen produced b y a standard combustion pro-
cedure. If the combustion method is valid for anv substance
when a volumetric azotometer is used, it should- be equally
acceptable in the present scheme. Research samples of
materials of established structure have been analyzed using
Determination of Ethyl Alcohol
bv MicrodiffusionJ
THEODORE WINNICK, Department of Surgery, Wayne University College of llledicine, Detroit, %rich.
L EVISE and Bodansky (6) have pointed out that most
current methods for the determination of ethyl alcohol
in tissue and body fluids employ a solution of potassium di-
chromate in sulfuric acid for oxidation of the alcohol to acetic
acid, and that these methods differ only in the means of sepa-
rating the alcohol from the tissue or fluid, and in measuring
the partial reduction of the dichromate.
I n addition to distillation and aeration methods, several
procedures based upon desiccation of the sample have been
developed. These are modifications of Widmark’s alcohol
method (8), which employs a 50-ml. glass-stoppered flask with
a small cup suspended from the bottom of the stopper. The
alcohol passes by gaseous diffusion from the sample contained
in the cup into a solution of dichromate in sulfuric acid in the
bottom of the flask. The quantity of alcohol oxidized is de-
termined by reducing the excess dichromate with potassium
iodide, and titrating the liberated iodine with standard thio-
sulfate solution. I n Cavett’s modification of this method
( I ) , the excess dichromate is titrated with ferrous sulfate solu-
tion containing methyl orange.
The present method substitutes the Conway microdiffusion
unit ( 2 ) for the Widmark flask. The alcohol diffuses from a
blood or urine sample placed in the outer chamber of the unit
into a solution of potassium dichromate in sulfuric acid in the
central chamber. The excess dichromate is determined iodo-
metrically, as in the method of Widmark. The apparatus is
inexpensiye and simple to operate, so t h a t a large number of
determinations may be run simultaneously. I n addition, the
Conwvay unit has the advantage of being applicable to a vari-
ety of biochemical determinations ( 2 , 9).
Experimental
SIZESOF SAMPLES FOR AI~ALYSIS. Analyses are performed on
blood or urine samples containing approximately 1 to 1.5 mg. of
ethyl alcohol. Since the concentration of alcohol in blood
523
of acetanilide nas ILIallinckrodt’s U. S. P. powder recrystal-
lized twice from xater. I t s melting point was 114.20”, and a
cooling curve proved it to be homogeneous. A series of ten
Kjeldahl determinations with the acetanilide averaged 10.35
per cent, and the same number of determinations b y the
Dumas method using a n old and somewhat corroded Pregl
azotometer averaged 10.45 per cent.
All the results given in Table I are for analyses by the junior
author, and represent the valid results of three series of four
determinations each.
lclinowledgment
The authors wish to express their thanks to
the Atmater Reseaxh Fund for financial assistance
and to the Coming Glass Korks for con-
struction of the apparatus.
Literature Cited
(1) Milner and Sherman, IXD.ESG.CHEY.,ANAL.ED.,8, 331 (1936).
(2) Kiederl and Ni .derl, “Organic Quantitative Microanalysis”,
2nd ed., pp. 79-100, Kew York. John Wiley & Sons, 1943.
(3) p. ”,
(4) Pregl and Fyleman, “Quantitative Organic Microanalysis”,
pp. 79-94, Philadelphia. P. Blakiston’s Son & Co., 1924.
( 5 ) Trautr, o., Mllikrochemie, 9,300 (1931).
and urine is roughly proportional to the degree of intoxication
(4),the size of the sample taken for analysis is varied accord-
ing to the apparent condition of the subject as follows: 1 ml.
for mild or doubtful, 0.5 ml. for moderate, and 0.25 ml. for severe
intoxication.
COLLECTIOX A S D hTEASUREMENT O F SAMPLES FOR ANALYSIS.
Blood and freshly voided urine specimens are collected in tubes
which are t,ightly fitted xith rubber stoppers. The tubes for
blood samples contain oxalate or citrate. The author has con-
firmed Cavett’s observation ( 1 ) that the blood or urine can stand
for 24 hours in the ice box without appreciable change in alcohol
concentration.
The 0.25 ml. pipet is made from capillary tubing of about 1-mm.
bore, and is calibrated for content. In measuring samples with,
this pipet, about 0.1 ml. of water is drawn up first, leaving a small
air space at the tip, and then the pipet is filled to the calibration
mark with blood or urine. The air space between the two columns
prevents mixing of the fluids. When the pipet is emptied, the
water washes out the film of blood or urine left along the inner
wall of the pipet.
MICRODIFFUSION PROCEDURE.The glass cover plate is suit-
ably smeared with “cello-seal” (Fisher Scientific Co.), a lubri-
cant which does not liquefy at the temperatures employcd (25’ to
50”).
Approximately 0.5 ml. of 0.4 N potassium dichromate in 10 W
sulfuric acid is delivered with a volumetric pipet into the central
chamber of the unit. A constant time of drainage is employed.
The exact volume or concentration of the dichromate solution
need not be known, since blank analyses are run with each series.
The blood or urine sample is pipetted quickly into the outer
chamber, and the vessel is sealed immediately with the greased
glass cover plate. The unit is rotated to spread the fluids over
the floors of the chambers. After the unit has stood for 2 hours
a t 50”, 6 hours a t 37”, or 10 hours at 25” (or for longer times), the
cover plate is removed and the dichromate solution is diluted
with approximately 1 ml. of water. Then about 0.5 ml. of 3 M
potassium iodide solution is added with stirring to the centrak
solution, and the liberated iodine is titrated a t once with stand-
ard 0.1 N thiosulfate solution. (The large excess of potassium
iodide reduces the loss of iodine vapor from the solution before
addition of thiosulfate. Tests showed that 0.4 to 0.5 per crnt
of the iodine is lost per minute if the central solution is allowed
to stand after addition of the potassium iodide. Accordingly,
524 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 14, No. 6

in the blood and urine of persons

TABLE I. RECOVERY
O F ETHYL
ALCOHOL ADDEDTO BLOOD
AND URINE following anesthesia, were also tested.
7 .Alcohol in Blood 7 --.llcohol in Urine-- Analyses were performed at 37' on
1-ml. portions of solutions (or disper-
sions) containing 50 to 100 mg. of the
organic compound to be tested per 100
0.5 26.5 ... 30 47 41 ... 37.5 ... ml. of water.
1.0 57 27.5 60.5 79.5 70 44.5 72
2 0 80 48 84 9s 103 48 82 99 23
I 1
No appreciable quantity of dichromate
4.0 92 79.5 96 .. ... 66 5 92.5 .. 81 in the central chamber was reduced when
6.0 97.5 86.5 99 ... 94 5 100,s .. 103
8.0 93 .. ... 97 5 . . ... solutions of the following substances
10.0 102 .. ... 101 ... ... were placed in the outer chamber:
benzene, chloroform, carbon tetrachlo-
ride, naphtha, and trichloroethylene.
if most of the thiosulfate is added less .than a minute after the Ten to 20 per cent of the dichromate was reduced -when
liberation of the iodine, the amount of iodine. lost is, negligible,) methyl alcohol, isopropyl alcohol, amyl alcohol, octyl alcohol,
The thiosulfate is added rapidly with efficient stirring, until
nearly all of the iodine is reduced, and then dropwise until the isoamyl acetate, acetaldehyde, and propionaldehyde were
color of the solution is light yellow-green. A drop of 1 per cent tested, indicating that the vapors of these substances were
starch solution is added, and the titration is completed to the readily oxidized by the dichromate solution. Accordingly,
point a t which the deep blue starch-iodine color is replaced by the drunkenness caused by these compounds cannot be differ-
the light blue-green color of the chromic ion Owing to the
small volume of the solution being titrated, this color change is entiated readily from ethyl alcohol intoxication by the present
very abrupt. method, as is possible in the case of the preceding group.
About 2 ml. of thiosulfate solution are required in the blank Diethyl ether was oxidized slowly by the dichromate. This
analyses, so that the final volume of fluid in the central chamber compound is not likely to interfere seriously \vith the ethyl
is never greater than 4 ml. The central chambers of the units
used in this study have capacities of 4.5 to 5 ml alcohol determinations, since ethyl alcohol is present a t a
BLANKANALYSIS. An analysis is performed nith water in relatively much higher level during intoxication and is more
place of blood or urine in the outer chamber. In the author's rapidly oxidized by the dichromate.
experience, blank analyses usually agreed to within 0.01 ml. (0.5
per cent).
CALCULATION. The volume of thiosulfate solution which
corresponds to the quantity of dichromate reduced, and to the TABLE11. CONCENTRATION
OF ETHYLALCOHOL
IN BLOOD
AND
quantity of alcohol oxidized to acetic acid, is given by the differ- URINEOF INTOXICATED
PERSONS
ence between the volumes of thiosulfate required in the blank Alcohol Found
analysis, VB, and in the alcohol determination, VA. Since 1 ml. Apparent Degree of Intoxication Blood Urine
of 0.1 N thiosulfate is equivalent to 1.152 mg. of alcohol, the Mg./100 ml. M g . / l O O ml
milligrams of alcohol per 100 ml. of blood or urine equal Mild t o moderate 132, 135 185, 188
159, 163 260, 261
( V B - Va) X 1,152 X 100 219, 224 276, 289
Volume of sample analyzed Moderate t o severe 162, 170 170, 176
188, 190 269, 276
236, 254 267, 278
RECOVERY OF ETHYL ALCOHOL ADDEDTO BLOODAND URINE 249, 260 313, 316
Standard solutions were prepared by adding weighed quantities 304, 310 442, 452
of 97.5 per cent ethyl alcohol (percentage determined by specific 380. 387 468, 480
gravity measurement) to urine and to a 1 per cent sodium chloride Very severe, "dead drunk" 520, 547 .....
solution. The alcohol was weighed in sealed glass bulbs, which
were broken beneath the surfaces of the solutions, so that no
alcohol vapor could escape. Three standard solutions, contain-
ing 65.5, 155, and 341 mg. of alcohol per 100 ml. of blood, were Blood and Urine L e v e l s
prepared by diluting blood (found previously to contain no
alcohol) nith suitable volumes of the sodium chloride solutions, SORMAL. The blood and urine of six individuals who had
which contained 1705 mg. of alcohol per 100 ml. taken no alcoholic beverages for several days contained no
Series of determinations with varying reaction times were measurable amounts of alcohol. When 1-mi. samples were
made a t three different temperatures on aliquots of the blood analyzed, the thiosulfate titrations ranged from 1.905 to 1.94
and urine solutions.
ml., as ccmpared with 1.925 ml. for the blank titration
Table I s h o w that 97.5 to 103 per cent of the added alcohol Cavett ( I ) and Gibson and Blotner (3)found traces of alcohol.
was recovered from the blood or urine after reaction periods 0 to 6 mg. per 100 ml. of blood, and 0 to 2 mg. per 100 ml. of
of 2 hours at 50°, 6 hours at 37", or 10 hours at 24-25'. The urine, in persons who had taken no alcohol. These levels are
analyses were performed on 1-ml. portions of the solutions, probably close to the limits of sensitivity of these methods,
except for the blood containing 341 mg. and the urine con- and insignificant compared to the values encountered in
taining 374 mg. of alcohol per 100 ml.; 0.25-ml. samples were intoxication.
analyzed in these two series. ALCOHOL LEVELSIN BLOODAND URINEDURING INTOXICA-
Duplicate determinations usually agreed to within about 2 TION. The results in Table I1 conform fairly well to the
per cent. Approximately two-thirds of the dichromate solu- general classification of intoxication in terms of ethyl alcohol
tion was reduced by the alcohol in most cases. concentrations (4). The subjects were persons brought into
INTERFERING SUBSTANCES.Of the volatile compounds the Emergency Service of the Detroit Receiving Hospital.
which may be encountered in body fluids, acetone need not be I n most cases the duplicate values agree to within about 3 per
considered, since it is stable even in boiling dichromate cent. The urinary levels of alcohol are considerably higher
solution. than the corresponding blood concentrations in the cases
A number of organic compounds, important in industry, studied. Jetter (5) has made this same observation, and
are known to produce conditions resembling ethyl alcohol states that this is particularly true for cases in which the
intoxication when sufficient amounts of their vapors are in- bladder was previously emptied.
haled. Tests were made with some of these substances to RATE O F DISAPPEARANCE O F -4LCOHOL FROM BLOOD.
determine whether they could be distinguished from ethyl Sewman, Lehman, and Cutting ( 7 ) showed that the alcohol
alcohol in the present method. Chloroform and diethyl ether, concentration in blood falls in a linear fashion with increasing
two common anesthetics which may linger in small amounts time, following the intravenous administration of ethyl alco-
June 15, 1942 ANALYTICAL EDITION
TABLE
111. RATEOF DISAPPEARANCE OF ETHYL ALCOHOLFROM
BLOODO F INTOXICATED P E R S O N
Time after
Taking First .%lcohol in
B:ood Sample Blood Condition of Subject
Hours . V ~ . / 1 0 0ml
0 304, 310 Moderate t o severe intoxication. speaks
with difficulty
3 214. 256 UncooDerative. fiehtine restraints
6.5 164, 168 Quiet i n d sleepy-
9 9.5, 97 Fairly rational
12 32, 38 Rational
hol to dogs. It was thought of interest to make a series of
similar measurements with a human subject, t o illustrate
further the use of the present method over a wide range of
alcohol concentration. Table I11 gives the results obtained
for a 12-hour period with a patient mho is classified in the
“moderate to severe” range of intoxication in Table 11. If
the alcohol values are plotted against the corresponding times,
the points fall along a straight line.
Summary
The Conway microdiffusion unit has been adapted to the
determination of ethyl alcohol in blood and urine. The
alcohol diffuses from the sample in the outer chamber of the
unit into the central chamber, where it is oxidized b y a solu-
tion of potassium dichromate. The excess dichromate is
determined iodometrically
Semimicrodetermination of Carbon
Using the Van Slyke-Folch Oxidation Mixture
R . \I. MCCHEADY AND W. Z. HASSID, Division of Plant Nutrition, b-niversity of California, Berkeley, Calif.
V AS SLYKE and Folch (4) pointed out that all the wet
carbon combustion methods hitherto employed were un-
satisfactory because the oxidizing mixtures used did not give
quantitative results with the more difficultly combustible com-
pounds. For this reason these methods did not find a place
in the organic laboratory.
X e t combustion methods were tried in this laboratory em-
ploying either a n iodic acid (3) or chromic acid (2) oxidizing
mixture, but the results were unsatisfactory. While theo-
retical results could be obtained with many organic com-
pounds, certain substances such as sugar acetates and some
organic acids invariably gave low results. The Van Slyke-
Folch (4) manometric method, in which a n oxidizing mixture,
consisting of fuming sulfuric, phosphoric, chromic, and iodic
acids was used, gave excellent results with all the compounds
tried. Equally satisfactory results were obtained using the
Van Slyke-Folch oxidizing mixture, and a simple apparatus,
whereby the carbon dioxide evolved was absorbed in alkali
and weighed. Inasmuch as the Van Slyke manometric ap-
paratus is not available in many laboratories, the following
procedure, requiring simple manipulation and inexpensive
equipment, is described.
Reagents
Van Slyke-Folch combustion mixture: 25 grams of chromium
trioxide, 5 grams of potassium iodate, 167 ml. of sirupy phosphoric
acid (specific gravity 1.7); and 333 ml. of fuming sulfuric acid
(20 per cent free sulfur trioxide) are placed in a 1-liter Pyrex
Erlerimever flask provided with a ground-glass stopper. The
525
Some illustrative alcohol concentrations in blood and urine
during intcxication are re,ported.
A number of industrially important organic compounds
whose vapors can cause drunkenness resembling ethyl alcohol
int’oxication were tested for possible interference with the
method.
Acknowledgment
The writer appreciates the advice and ass.oLance given bin1
by C. G. Johnston and I. B. Taylor, relative to the clinical
phase of this study. He wishes also to thank C. P. McCord
and G. C. Harrold, of the Industrial Hygiene Department,
Chrysler Corporation, Detroit, for supplying some of the or-
ganic compounds tested, and for helpful information.
Literature Cited
(1) Cavett, J. W., J . Lah. Clin. M e d . , 23, 543-6 (1938).
(2) Conway, E. J., “Microdiffusion Analysis and Volumetric Error”,
London, Crosby Lockwood and Son, 1939; Biochem. J . , 27,
419-34 (1933).
(3) Gibson, J. G., and Blotner, H., J . Biol. Chem., 126, 551-9 (1938).
(4) Goodman, L.. and Gilman, A., “Pharmacological Basis of Thera-
peutics”, New York, Macmillan Co., 1941.
(5) Jetter, W.W., Quart. J . Alcohol, 2, 512-43 (1941).
(6) Levine, H., and Bodansky, M., Am. J . Clin. Path., Tech. Section
3, 159-73 (1939).
(7) Nemman, H. W., Lehman, A . J., and Cutting, W.C., J . Pharma-
C O ~ . .61, 58-61 (1937).
(8) Widmark, E. M.P., Biochem. Z . , 131, 473-84 (1922).
(9) Winnick, T., J . B i d . Chem., 141, 115-20 (1941): 142, 461-66
(1942).
AIDED by a grant from the 3IcGregor Fund.
open flask containing the mixture is heated over a wire gauze
with a flame, until the temperature reaches 140” to 150” c‘. Dur-
ing heating, the flask is occasionally rotated to assist in the solu-
tion of the chromic anhydride. When a temperature of 150” c‘.
has been reached, the flame is removed, the flask covered with an
inverted beaker, and the mixture allowed to cool to room teni-
perature. The glass stopper is then inserted, and an inverted
beaker is kept over the stopper to prevent the solution from being
contaminated with dust.
Potassium iodate, reagent grade, pulverized.
Dehydrated phosphoric acid, prepared from 85 per cent acid
by boiling.
Apparatus
The apparatus is shown in Figure 1. Tubes A A are filled with
soda lime and serve to obtain carbon dioxide-free air. The reac-
tion flask, B , has a 15-1111. capacity and is connected to a reflux
condenser, C, with a 14/35 standard taper glass joint. The con-
struction of an Allihn condenser is modified (Figure 2) to con-
tain a cold-water column inserted into its center. This modifica-
tion greatly increases the efficiency of the condenser. A 5-ml.
capacity cup, D, with a stopcock, b, forming part of the con-
denser, serves to introduce the oxidizing mixture into the reaction
flask. Stopcock b is greased with viscous dehydrated phosphoric
acid. The rest of the stopcocks are greased with ordinary stop-
cock grease.
E is a bubble counter containing 0.5 ml. of concentrated sul-
furic acid. A small glass-wool plug is inserted into each of its
arms to trap any sulfuric acid which might splash over to tube
F or condenser C. F is filled with granular zinc (30-mesh) to
trap any volatile acids which form in the determination. The
moisture-absorption tube, G, is filled with Anhydrone (rnagne-
sium perchlorate) leaving 1-em. space at each end for a glsss-
wool plug. Previous to use in the apparatus, a slow stream of