CITRIC ACID DETERMINATION BY AARON S. GOLDBERG AND ALICE R. BERNHEIM (From the Laboratory for the Study of Peripher al Vascular Diseases, Department o f Surgery, the New York Ho spital, and Corne ll University Med ical Colleg e, New York) (Rece ived fo r publication, April 19, 1944) Three types of methods hav e been proposed for the determination of citric acid in urine. Salan t and Wise (1) used the unspecific mercuric sulfate reagent of Deniges (2) in a study of the toxicity of citrates. Thun- berg (3) developed an enzy mati c method employing a dehydrogenase said to be specific for citrate. The remaining methods depend on the so called pentabromoacetone method of Stahre (4), a qualitative test publi shed in 1897 which Ku nz (5), in 1914, used as the basis of the first quan titative pentabromoacetone method. It has been generally overlooked t hat i n 1847 Cahours (6) prepared a bromination product from several citrates which he called “bromox aforme” which according to his melting point and solubility data is withou t doubt identical with pentabro moaceton e. Cahours was aware of the specificity and analytical applicability of the reaction. (“Le brome peut done servir a reconnaftre de petites quantit& d’acide citrique mhlangbes a l’acide tartri que.“) Kunz’ s techn ique was adapted for use with urine by Amberg and McClure (7). Fasol d (8) with this method was the first to isolat e citric acid from normal ur ine. Kometiani’s (9) modification of this method was applied to urine b y Stillmann and Schaerer (10). Pucher, Sherman, and Vi ck er y (11) de- velo ped a spectroph otometric method based on the color form ed when pentabromoacetone is treat ed with sodi um sulfide. Ou r pr ocedure is another modification of this method. It is precise, accu rate , and time- saving. The “pentabromoacet one reaction” has been clarified and new reagents with certain advantages hav e been introduced. While the procedure was designed for 15 cc. sample s of urine, it is also applicable to 2 cc. samples, as well as to other fluids of comparable citrate cont ent. The range is from 1 to 40 mg., although with care , useful results are ob- tainable down to 0.2 mg. The method has al ready been used in a stud y of the variation of urinary citrate with the menstr ual cyc le (12). Principle The citric acid is oxidi zed by manganese dioxide, in the presence of bromine, to acetonedica rboxylic acid, which is then rapidly brominated with simultaneous decarbo xylatio n, yielding pentabr omoacetone. Af te r reduction of the excess manganese dioxide and bromine with hydr azin e,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
showed a violet tinge. This was removed by adding a few mg. of sodium
arsenite to the wash water.
Glucose-As noted by Reichard (13), low results are ordinarily obtained
in the presence of glucose. If, before analysis, the specimen is not known
to contain sugar, its presence will be indicated by the unusually high per-
manganate consumption. The following procedure yields quantitative
results even in the presence of 100 times as much glucose as citric acid.
A small sample is taken, 1 to 2 mg. instead of 5 to 15. Bromination is
carried out at 0” instead of at room temperature, with a corresponding
increase in reaction time. ,4t 24” the reaction time is about 20 minutes,
whereas at 0” it is 4 to 5 hours. Manganese dioxide suspension is sub-
stituted for permanganate and the bromide-bromate is increased to 2 cc.
The following procedure is applicable to samples containing up to
400 mg. of glucose, although a smaller sample should be t,aken if the
citrate concentration permits. The neutral sample is placed in a 60 cc.
separatory funnel, the volume is brought to 15 cc., and 4.3 cc. of 27 N
sulfuric acid are added; or if a preliminary bromination has been made
15 or 20 cc. of the filtrate are used.2 cc. of bromide-bromate and 1 cc. of manganese sulfate are added, and
the funnel is placed in an ice water bath. The manganese dioxide sus-
pension is prepared as follows: 7.5 cc. of 2 M manganese sulfate are placed
in a small beaker immersed in an ice bath, or, better, a freezing mixture,
and 4 cc. of 27 N sulfuric acid are added. When the solution is below 5”,
7.5 cc. of cold 26.1 per cent NaMnOd.3Hz0 (1.33 M) are added with stirring.
The solution should be allowed to stand in the bath for some minutes
before use. The requisite (not total) amount of suspension is added tothe funnel and allowed to stand for at least 3 hours at 0”. The funnel is
inspected occasionally to see that the manganese dioxide is present in
excess and that it has not settled out. At the end of this time the contents
of the funnel are allowed to come to 20-25” and to stand for another hour.
Then hydrazine is added and the analysis completed as usual.2
In the presence of smaller amounts of glucose (e.g., 50 mg.) a suspension
may be made from the 0.2 M potassium permanganate with 1 cc. of manga-
nese sulfate, 6.6 cc. of 0.2 M KMn04, and 2 cc. of 27 N sulfuric acid. This
suspension may be used routinely for milk, animal urines, and in other
cases in which the consumption of 0.2 M KMnOa tends to exceed about
3 cc., whether because of glucose or other reducing substances.
2 Since the water content of commercial sodium permanganate is variable, it is
desirable to test how well the solution matches the manganese sulfate solution bycentrifuging suspensions made with volumes varying slightly from the theoreticalone specified and find.ing by inspection of the supernatant the proper volume to beused in subsequent determinations.
to the nai l on the motor. Alternate stretching and releasing of the rubber
band agitates the suspension in the separatory funnel. The stems of the
funnels are cut off and ground down to an acute angle. The over-all
length of all funnels of each size should be the same to avoid shifting the
clamp holding the rubber stopper. No lubricant is used on the stopper
of the 125 cc. funnel. It is wetted with a drop of the filtrate and inserted
firmly with a slight, twisting motion. A drop of water is applied to the
stopper of the smaller funnel before it is shaken. Stop-cocks should
have the thinnest possible film of a heavy rubber paraffin grease. When
the funnel is removed from the shaker, the stop-cock is always openedbefore the stopper to equalize the pressure.
Many workers have used permanganate in large excess. However,
especially with small amounts of citric acid (less than 3 mg.), it is im-
portant to add permanganate only as rapidly as it is used up. Use of the
bromide-bromate mixture, instead of bromide alone, reduces the amount
of permanganate needed and makes for greater convenience and accuracy.
Since permanganate vigorously decomposes the intermediate acetone-
dicarboxylic acid, its rapid reduction to manganese dioxide (hastened bythe large excess of manganese sulfate) is obviously advantageous.
Sufficient hydrazine must be added to remove the last trace of bromine,
to avoid its extraction by the petroleum ether with which it would subse-
quently react. Some of the bromine might later be reduced by the sulfite,
giving rise to a high result. We have avoided the use of ferrous sulfate
for decolorizing because the color of the resulting ferric salt is indistinguish-
able from that of the free bromine. Methods in which pentabromoacetone
is filtered off result in a loss equivalent to about 3 mg. of citric acid per100 cc. of reaction mixture, or an error of perhaps 50 per cent in urine of
low citrate content. In the petroleum ether procedure, the loss of penta-
bromoacetone has been shown to be negligible (about 0.001).
During the treatment with sulfite a trace of thiosulfate is formed, which,
in the presence of silver halide, does not give the usual white precipitate
turning to black with silver nitrate, but behaves like a halide and uses up
an exactly equivalent amount of silver solution. The thiosulfate formed is
equivalent to about 0.03 cc. of 0.4 N AgN03. The treatment with hot
ferric alum suffices to oxidize most of it, but for accurate results with small
amounts it is desirable to run a blank, starting at the point in the proce-
dure where the sulfite is added to the alcohol. In our experience this
amounts to only 0.01 cc.
The principle of the titration is due to Kolthoff (15), but we have modi-
fied the details to our needs. The thiocyanate solution, unlike that in
the usual Volhard titration, need not be accurately prepared and need
be only roughly pipetted. The addition of a measured number of drops
induce scission of the carbon chain. At 15 N, when pure hexabromoacetone
is obtained, the apparent yield is only 72 per cent, or 60 per cent in termsof moles of hexabromoacetone. This indicates destruction of 40 per
cent of the citrate.
Temperature-Kunz, followed by others, prescribed a temperature of
50” for the bromination, while some advised room temperature. Fasold
stated that at 55” results were 30 per cent too low. Reichard advised
cooling to 5” or less. Our work shows a slight decrease in yield (1 per
cent) from 24” to 48” and a 1 per cent increase at 0”. Qualitatively this
is to be expected in the exothermic bromination reaction, but the quan-titative effect of temperature variation can be calculated from van%
Normality of sulfuric acid
FIG. 2. Recovery with varying acidity
Hoff’s equation. These calculations are in good agreement with our
findings where they have been checked.3
3 We select, for example, the point at which 101 per cent recovery is obtained
(about 10 mg.) and find that this corresponds to 1 mole of hexabromoacetone to
19 moles of pentabromoacetone. The equilibrium constant for the reaction penta-
bromoacetone + BrS + hexabromoacetone + HBr may then be written K24= (hexa-
bromoacetone)/(pentabromoacetone) = l/19. Since we solve for the ratio of the
constants at two different temperatures, and since practical constancy of the Br andHBr concentrations is assured by the experimental conditions, these concentrations
need not be expressed. From the data of Magee and Daniels (17) we find values of
AH = 6.5, 15,12, and 8 kilocalories for the bromination of an aliphatic hydrogen in
methane, and mono-, di-, and triphenylmethane respectively. We can assume a
rounded average of 10 kilocalories for the bromination of the residual hydrogen of
pentabromoacetone. This may seem rash, but calculation shows that either doubling
or halving the value does not alter the outcome materially. Substituting in the
integrated form of the van’t Hof f relation, we have for 48” and 24”,log KC, log Kzt =-10,000/4.57 (l/297 - l/321), or K~~/KH = 0.288,a yield of 100.3 per cent; i.e., a
Citrate Concentration-Fig. 3 shows the effect of the citrate sample on
the yield. The yield is seen o be strictly stoichiometric only with a 3 mg.
sample. With another acidity or temperature the value would be different.
Nevertheless, exact results may be obtained by employing a curve con-
1 96 -.
s 94..
92
90 -
86 I / IMs. 6 10 20
Citeicacid
44
FIG. 3. Recovery with different concentrations of citric acid
strutted from the empirical factors given in Table I. For most work the
factor 1.53 may be used routinely without serious error.4
diminution of 0.7 per cent as a result of a 24” rise in temperature. Interestingly this
value of 0.7 per cen t is not constant, but would vary with the initi al deviation from
100 per cen t. From 102 per cen t the value wou ld fall to 100.6 per cen t (1.4 per cent).Similar compu tation in the 1 and 2 mg. region (about 98 and 99 per cent recovery)
shows a larger effect. Assu ming the same heat of bromination for tetrabromoacetone
+ pentabromoacetone as for pentabromoacetone + hexabromoacetone, we find de-
creases of 5.7 and 2.2 per cen t respectively, at 48” compared to 24”. One important
factor is not taken into accou nt in this theoretical analysis; namely, the effect of an
elevated temperature on the elas tic action of the permanganate. With pure citrate
solutions , and provided the addition of permanganate is carried out as advised, loss
from increased elas tic action may be considered neg ligible, but in the presence of
considerable amoun ts of gluco se even room temperature is too high and it is neces-
sary to work at 0”. Glucose, if present in large amoun ts, must be burned completelyto CO2 and water, which increases considerably the amount of oxidant used. In this
case, as described before, permanganate shou ld be replaced by mangane se dioxide.
4 The rapid change of slope of the curve (Fig. 3) at the lower values is asso ciated
with the threshold of solubility of the pentabromoacetone. All of the bromoacetone
from 0.6 mg. of citric acid is in solution, while that in excess exists in a secon d liquid
phase con sisting of pentabromoacetone with a sm all proportion of hexabromoacetone
and whatever bromine is dissolved. In the aqueous phase, however, tetrabromo-
acetone and pentabromoacetone coexist in a ratio to give an apparent yield of about
96 per cent. Since the proportion of hexabromoacetone in the emu lsoid phase would
mg., with a precision of 0.5 per cent. With suitable care even 0.2 per cent
is attainable above 5 mg.
2. The “pentabromoacetone reaction” has been studied and it is shown
that tetrabromoacetone and hexabromoacetone are also found. The
conditions governing the relative amounts of each bromoacetone formed
are discussed. The main factors are acidity, citrate concentration, and
temperature.
3. The reaction mechanism has been clarified, leading to the introduc-
tion of new reagents with certain advantages. The decomposition of
pentabromoacetone is affected by a new, mild reagent which permits a
precise, direct titration with silver nitrate.
We wish to express our thanks to Hertha H. Taussky for her help and
many courtesies and also to Dr. Harry Sobotka for his valuable advice.
BIBLIOGRAPHY
1. Salant, W., and Wise , L. E., J. Biol. Chem., 28, 27 (1916-17).
2. Den iges, G., 2. an. Chem., 38, 718 (1899).3. Thunberg, T., Biochem . Z., 206, 109 (1929).
4. Stahre, L., 2. an. Chem ., 36, 195 (1897).
5. Kunz, R., Arch. Chem. u. Mike., 7, 285 (1914).
6. Cahours, A., Ann. chim., 19, 484 (1847).
7. Amberg, S., and McClure, W., Am. J. Physiol., 44, 453 (1917).
8. Faso ld, H., 2. Bio l., 90, 192 (1930).
9. Kom etiani, P., 2. an. Chem., 86, 359 (1931).
10. Stillmann, H., and Schaerer, E., Schweiz. med. Woc hschr., 62, 619 (1932).
11. Pucher, G. W., Sherman, C. C., and Vickery, H. B., J. Bio l. C&m., 113,235 (1936).
12. Shorr, E., Bernheim, A., and Taussky, H., Xcienc e, 96, 606 (1942).13. Reicha rd, O., 2. Untersuch. Leb ensmittel, 68,138 (1934) ; 72,50 (1936) ; Z. an. Chem.,
99, 81, 161 (1934).
14. Krebs, H., and Johnso n, W., Biochem . J., 31, 645 (1937).
15. Koltho ff, I., Pharm. Weekbl., 64, 761 (1917).
16. Weide l, H., and Gruber, M., Ber. them. Ges., 10, 1145 (1877).
17. Magee, J. L., and Dan iels, F., J. Am. Chem. Xoc., 62, 2825 (1940).