A'VALYTICAL EDITION
302
inents-namely, the alkaline lead a c e t a t e 1 0 cc. of air containing one part in 10,650,000 of hydrogen sulfide gave a reproducible test. Assuming that all the gas reacted to form cc. or 1.5 X 10-9 gram lead sulfide, there was used 10 X of hydrogen sulfide. At this same dilution, by using a liquid air trap just ahead of the testing device i t was possible to condense the hydrogen sulfide present in the gas filling the t r a p and connecting tube (4or 5 cc.), and then upon removing the liquid air to pass pure hydrogen through the trap and drive the hydrogen sulfide onto the test paper, thus obtaining a faint spot. The tests here reported were carried out under conditions which would exist in the radioactivity experiments for which they were necessary, and hence probably do not
Vol. 2, s o . 3
represent the final limit of sensitivity. It is probable that by using much larger volumes of gas somewhat smaller concentrations of hydrogen sulfide could be detected. Literature Cited (1) Almy, J . Am. Chem. Soc., 47, 1381 (1925). (2) Fischer, Ber., 16, 2234 (1883). (3) Gmelin-Kraut, Handbuch der anorganischen Chemie, vol. I V , P t . 11, p. 564 (1924). (4) International Critical Tables, Vol. 111, p . 207, ?vfcCraw-Hill, 1927. (5) Latshaw and Reyerson, J . A m . Chem. SOL.,47, 610 (1925). (6) McBride and Edwards, Bur. Standards, Tech. P a p e r 4 1 (1914). (7) hIcBride and Edwards, I b i d . , pp. 5 and 42 and references there given.
Determination of Organic Acids in Mixtures I-Determination
of Fatty Acids in Mixtures by Partition between Isopropyl Ether and Water' C. H. Werkman BACTERIOLOGY SECTION, IOWA AGRICULTURAL EXPERIMEXT ST.4TIOS,
I
K THE course of investigations on the products of fer-
mentation it became imperative to have a method for the determination of saturated fatty acids which would give results of a higher degree of accuracy and require less time than that of Duclaux ( 2 ) . Much effort in the past has been directed toward a n improvement of the Duclaux rather than toward the development of a method which would avoid the unreliability and time-consumption of distillation methods. The method to be described is based upon the partition of t h e fatty acids between isopropyl ether and water. It accomplishes all that the Duclaux does and has fewer of its limitations. Among the advantages may be enumerated: (1) It is much more accurate and reliable; (2) requires only a fraction of the time of the Duclaux; (3) less apparatus is required; (4) it is more economical; (5) determinations may be made on smaller quantities; (6) no tedious calculations are required; (7) no standardization of equipment is required; (8) it is simpler than the Duclaux; (9) it may be extended to more than two acids; (10) it may be used to determine non-volatile acids or mixtures of volatile and non-volatile acids. The greater accuracy of the partition method is apparent when a comparison is made of the ratios of distillation constants and partition coefficients. Wiegner and Magasanik ( 4 ) found that the relative proportions of various acids distilling over when the volume of dilute solutions was reduced by one-half were: acetic 36.59, propionic 58.48, and butyric 72.77 per cent. Dyer (3) gives t h e calculated distillation constants in his modification of the Duclaux for the first 100 cc. as: acetic 30.27, propionic 51.76, butyric 70.43 per cent. Dyer's constants are in the ratio 1: 1.7:2.4 and Wiegner and hlagasanik's values, 1: 1.6:1.9. The ratios are narrow and do not allow of any considerable error in the distillation. The partition coefficients of acetic, propionic, and butyric acids a t 25" C. between isopropyl ether and water were found t o be: 0.168, 0.809, and 1.717, respectively, in the ratio of 1:4.9:10.2. That is, the range, of the partition coefficients between acetic and butyric acids is approximately five times as great as the range shown by the distillation constants of the same acids. Furthermore, the accuracy of the deter1
Received April 20, 1030.
.%XES, IOWA
minations of the partition constants is much greater than that of the distillation constants. The method as here developed is based upon the practicable utilization of the distribution of the organic acids between isopropyl ether and water in relatively dilute conrentrations. Furthermore, the total acidity of the two phases is kept constant, although t'he relative proportions of acids present vary. Behrens (1) has described a method of fatty acid analysis making use of the partition coefficients between ethyl ether and water. His method requires the titration values for both the ether and aqueous solutions. These values are then s u b stituted in simultaneous equations. Behrens did not extend and standardize his method to provide solution by reference to graphs. Any method based on the use of ethyl ether offers a number of disadvantages when compared with the use of isopropyl ether: (1) higher vapor pressure of ethyl ether; ( 2 ) lower boiling point; (3) greater expansion with rise of temperature; (4) necessity of adding sodium to remove alcohol and water; ( 5 ) greater solubility of water in ethyl ether; (6) greater solubility of ethyl ether in water; ( 7 ) higher relative cost. T a b l e I-Experimental Values of P a r t i t i o n C o n s t a n t s of Two Acids between I s o p r o p y l E t h e r a n d W a t e r . 25 P h a s e a t 25' C. BCTY- PRO- P R O PRO0.1 .\- BUTY- BUTY- RIC, PIPIPIFORRIC, PRO- OXIC, OXIC, 0x1'2, M I C , ACID RIC, (FIRST LAC- ACEPILAC- FOR- .4CE- ACENAMED) T I C
TIC
ONIC
TIC
11.05 11.6 12.1 12.6 13 1 13.7 14.1 14.6 l5,Z 15.65 16.2
16.2 17.0 17.8 18.6 19 5 20.4 21 2 22.0 22.8 23.6 24.4
MIC
TIC
TIC
16.2 16.6 17.4 18.1 18 7 19.3 19.9 20.5 21.1 21.7 22.3
22.8
LACTIC,
ACETIC
0.1 s ACID (SSCOND
NAMED)
R
70 100 90 80 70 60 50 40 30 20 10 0
for M i x t u r e s cc. A q u e o u s
11,05 12.35 13.7 15.1 16.4 17.7 19.1 20.4 21.7 23.1 24.4
11.05 12.2 13.3 14.4 15.6
16.75 17.8 19.0 20.1 21.2 22.3
16.2 16.9 17.6 16.25 18.9 19.55 20.2 20.9 21.5 22.2 22.8
22'7
...
22:&
...
2Y.i
22:i
24 4 0 24.3 10 24.0 20 23.65 30 23 6 40 23.4 50 23.2 60 23.0 70 22.8 80 22.6 90 22.35 100
Isopropyl ether boils at 67.5' C., has a vapor pressure of 158 mm. a t 20" C., and may be shaken vigorously in a separatory funnel a t room temperature with no danger of blowing the stopper and loss due to evaporation. It is soluble in water to the extent of onlv 2.7 Der cent bv weight a t 23" C.. whereas ethyl ether is 6.0iper cknt solubli in w k e r a t 25" C.
I.liDCSTRIA L A.VD ENGILVEERISG CHEMISTRY
July 15, 1930
Water is soluble in isopropyl ether to the extent of 0.4 per cent by weight a t 23" C., and 1.3 per cent in ethyl ether (25" C'.). Technically pure and commercial isopropyl ether gave the same partition constants by experiment. The addition of metallic sodium did not change these readings. The commercial product i j somewhat less soluble in mater (1.71 per cent by weight a t 19" C.) than the technical grade. Water is 6.9 per cent by weight soluble in the commercial product a t 19" C. The organic acids used in this work were the purest obtainable in the market. Further investigations may reveal slight corrections of the partition constants shown in Table I. %
25 cc. of the aqueous phase require 17.4 cc. of 0.1 N alkali t o neutralize. This latter figure is the partition constant, and from the nomogram in Figure 1 i t is seen that the mixture is 80 per cent 0.1 normal propionic and 20 per cent 0 1 normal acetic acid.
Since graphs of the partition constants as abscissas plotted against relative percentages of the acids as ordinates (Figure 1) are straight lines, there is no difficulty in interpolating. I n Figure 1 the percentage of the acid first named on the graph is given on the left; on the right is given the percentage of the second named acid. These percentages are calculated as cubic centimeters 0.1 -Ir acid.
%
//
12
N
14
15
/6
/7
18
19
20
Z/
2 2 23
24
25 26
CC.
$UOH
Figure 1-Partition
of a Mixture of Two Acids between Isopropyl Ether and Water
303
.4pplication to Three Acids The partition method in the form here presented cannot be used to determine the proportion of each of three acids in a mixture. The partition constant for such a mixture will indicate only the limits within which the three acids must be present in the mixture. The partition method when applied to mixtures of three acids has in its present form one inherent defect of the Duclaux method-that is, the same distillation or partition constant may be given by various mixtures of the same acids. Thus in a mixture of butyric, propionic, and acetic acids a partition constant of 18.4 (see nomogram Figure 2 ) indicates that the limits of the proportions are approximately 34: 1: 65 to 1: 69: 30. Considerable information is made available regarding the composition of the mixture, but no definite proportion is determinable as is the case in a mixture of two acids.
The method here developed is used for the determination of fatty acids present in mixtures. The present paper is limited to acetic, propionic, and butyric acids. It will discuss the determination of any two of these acids in a mixture. Later papers will extend the method to include additional acids, mixtures of three acids, and identification of acids in mixtures of two and three acids. Experimental Procedure The mixture of acids in aqueous solution is titrated with 0.1 N patassium hydroxide and adjusted to a definite normality. The appropriate normality is determined by the strength of the solution and will vary with the type of investigation. For purposes of illustration in the present paper we shall assume adjustment to 0.1 normal. After adjustment, 30 cc. of the acid mixture are placed in a 100-cc. separatory funnel, 20 cc. of isopropyl ether added, and the mixture is shaken vigorously for 1 minute. Three minutes are allowed for separation of the phases. For very careful work the mixture may be placed in a constant-temperature compartment, but. this has not been found necessary for routine work. Enough of the aqueous phase is withdrawn into a buret to permit measurement of exactly 25 cc. into a beaker. The use of 25 cc. is convenient, whereas the recovery of 30 cc. of the aqueous layer is impracticable. This is titrated with 0.1 N potassium hydroxide using phenolphthalein as an indicator. The nuniber expressed in cubic centimeters of potassium hydroxide required to neutralize 25 cc. of the aqueous solution may he termed the "partition constant." Table I gives the experimental values for partition constants for various combinations of two acids. These values were obtained a t 25" C. by use of commercial isopropyl ether. The percentage composition may then be determined by reference to the nomogram in Figure 1. The isopropyl layer is discarded. Example. It is desired to determine the percentage composition of a mixture of propionic and acetic acids. It is found that
idox0 %
M%
20%
30%
50%
40%
60%
70%
80%
90%
(0%
. Figure 2-Determination of Organic Acids i n a Mixture Partition between Isopropyl Ether and Water
Calculations Using Partition Coefficients Graphical solution offers a number of advantages and is the method of choice for routine work. However, quantitative determination of two acids may be calculated from a knowledge of the partition coefficients of the acids present and the number of cubic centimeters of 0.1 N potassium hydroxide required to neutralize (1) the total sample partitioned and ( 2 ) the aqueous phase. The general equations are as follows: It is assumed for purpose of illustration that the acids are partitioned between equal volumes of isopropyl ether and water; also that the original solution is adjusted to 0.1 normal. A I = cc. 0.1 N first acid present in x cc. of 0.1 N mixture A2 = cc. of 0.1 N second acid present in x cc. of 0.1 N mixture PI = partition coefficient of first acid
A-VALYTZCAL EDITIOS
304
P B = partition coeffioient of second acid N = cc. 0.1 N alkali required to neutralize aliquot of original
x' =
mixture = cc. of original solution employed after adjustment to 0.1 N . 30 cc. have been used in the graphical solution cc. 0.1 N alkali required to neutralize aqueous phase when partition is made between equal volumes of isopropyl ether and water
+ AB= I 1 1 AI + -A? 1 + PI 1i Pn Au
(-
-)
1 1 1+PB 1+p,
Al =
(1) = w
added to raise the boiling point. Table I1 gives a series of results which were obtained by preparing a nutrient medium suitable for bacterial fermentation, composed of peptone 10 grams, dextrose 5 grams, acid potassium phosphate 2 grams, water 1 liter. Known quantities of the organic acids were added to the medium. T a b l e 11-Determination of Volatile Acids in F e r m e n t a t i o n S t u d i e s SAMPLE ACIDSADDED RATIO RATIOFOUND
(2)
(-* 1 ) -
1 2 3
w
Propionic, acetic Butyric, propionic Butyric, acetic Propionic, acetic Propionic, acetic
4
5
1 f P . "
A B = x - A1
VOl. 2, s o . 3
(4)
The partition coefficient is defined as the ratio of distributions of acid between equal volumes of ether and water, the latter phase being unity-i. e., 0.168 : 1 for acetic; 0.809 : 1 for propionic; and 1.72 : 1 for butyric. It is not necessary to adjust the original solution to known normality. Determinations may be made by titrating both (ether and aqueous phases and substituting for Equation 1 :an equation obtained similarly to Equation 2 but applicable t o the ether phase:
=
cc. 0.1 N alkali required to neutralize isopropyl ether phase Application t o Fermentation Studies
The partition method will find its greatest use in the same kinds of analytical work as that in which the Duclaux method has been used. This applies especially to the determination of volatile acids in fermentation studies. I n this case a convenient quantity of the fermented medium is distilled to obtain the volatile acids in the distillate. We have generally used 200 cc. adjusted to pH 3.0 by the addition of sulfuric acid to liberate the organic acids. Sodium chloride may be
Per cent
60: 40 50 : 50
59:41 50 31:69
30:70
so: zn
5 0 : 50
a:
80:20
48:52
The distillations were continued to a minimum volume of residue in order to drive over the last of the less volatile acids. Percentages of acids were obtained by interpolation where necessary. Some investigators may prefer to distil the dilute solutions of acids to one-half volume and calculate the percentages of the various acids which have distilled over according to figures given in the literature. The acidity of fermented media is frequently due to one or two acids with only traces of other acids which do not interfere significantly with the determinations. Conclusion
(5)
U
Per cent
The partition method for the determination of fatty acids offers many advantages over the Duclaux. As outlined, the quantities of two fatty acids in a mixture may be quickly and accurately determined. I n its present form the method will serve to indicate the limits of the proportions of three acids in a mixture. The method will be extended in further communications. Literature Cited (1) Behrens, Z.anal. Chem., 69, 97 (1996). (2) Duclaux, Ibid., 39, 376 (1900); 47, 615 (IQOS). (3) Dyer, J. B i d . Chem., 28, 446 (1917). (4) Wegner and Magasanik, .I.lift.Lebensm. H y g . , 10, 156 (19191
Determination of Lithium' M. H. Brown a n d J. H. Reedy CHEMISTRY
DEPARTMENT, VNIVERSITY
OR the determination of lithiuiii in the presence of
F
sodium and potassium, preference is usually given to the Gooch method (1). I n this procedure the lithium chloride is extracted from the combined alkali chlorides with amyl alcohol and, after heating with sulfuric acid, weighed as the sulfate. Owing t o the slight solubility of sodium and potassium chlorides in amyl alcohol a correction factor must be applied. For every 10 cc. of amy1 alcohol used a deduction of 0.00041 gram must be made when sodium is present, of 0.00051 gram when potassium is present. or of 0.00092 gram if both sodiuni and potassium are encountered in the sample. Kahlenburg and Krauskopf (3) attempted to improve this method by using pyridine as the extractant. S o corrections are necessary in this case, as sodium and potassium chlorides are not soluble, even in traces, in the solvent. However, the odor of pyridine is so highly objectionable that the method has been avoided. I n the Rammelsberg method ( 5 ) the separation is effected 1
Received March 27. 1930
O F ILLIXOIS, U R B A N A ,
ILL.
by using a mixture of equal parts of alcohol and ether, which is kept saturated with hydrogen chloride gas. The manipulation is difficult, and the method is seldom used. Use of Acetone as a n Extractant
The fact that sodium and potassium chlorides are entirely insoluble in acetone, while lithium chloride is very soluble, seems to have been overlooked. This makes available a method that is free from all the disadx-antages of previous procedures. The solubilities of the various chlorides in acetone were determined (1) by the "spot method'' and (2) b y weighing the residue obtained by evaporating a known amount of saturated solution. In the first method, measured quantities of standardized solutions corresponding to 0.05, 0.1, 0.2, and 0.5 mg. of the chlorides were introduced into small flasks and gently evaporated to dryness, leaving the salts as a residue in the form of a spot. Known volumes of dry acetone were then introduced,