NOVEMBER 1947 /OH CClaCH \OH
88!3
+ ROH
/OR
CCliCH +HzO O 'H alcoholate
and this liberated water can then react with the Karl Fischer reagent. The nonalcohoilc diluents cannot accomplish this. The methanol present in the reagent accounts for the improved results obtained when the titri-meter and nonalcoholic diluents are used. When used alone methanol reacts very vigorously with chloral and evolves sufficient heat to bring about acetal formation with liberation of a further quantity of water. Isopropanol and butanol are readily available at moderate cost and are suitable for both the direct visual and indirect titrimeter method of analysis. RECOMMENDED METHOD O F ANALYSIS
Pipet a 10-ml. sample of refined chloral into 50 ml. of anhydrous isopropanol or butanol. If the ratio of diluent to sample is less than 5 to 1, the heat developed on titration is excessive. More sample is required if the water present is in the order of 0.01%. Mix and titrate immediately with Karl Fischer reagent prepared and standardized according to the method of Smith, Bryant, and Mitchell (8). Do a blank on the alcohol used. In the visual method the reagent is used for direct titration, the yellow color turning red brown a t the end point.
% by weight of water
=
(ml. of reagent for sample) (ml. of reagent for blank) X 100 F volume of sample X 1.51
where F = grams of water equivalent to 1 ml. of reagent. The.electrometric method, in which an excess of Karl Fischer reagent is added to the diluted sample, the excess is titrated with a standard water solution, and the end point is obbained by the apparatus of Kieselbach (I), has been used recently in place of, the visual method with equally good results. The reproducibility of the visual method in the hands of a laboratory technician from a different subdepartment when iso-
propanol is used as a diluent is shown in Table 11. This method has proved entirely reliable for both research and control purposes over the past year.
Table 11. Accuracy and Reproducibility of Isopropanol Method Item
Water Calculated
% 1 2 3 4 5
0:26 0.46 0.76 1.06
Water Found
% 0.06 0.25,O 28 0.44,O 46 0.71,0.70 1.06,1.12
Remarks Blank, no added.water Measured quantities of water added
STANDARDIZATION O F KARL FISCHER REAGENT
Warren (3) recommends the use of crystallized sodium acetate, C2Ha02Na.3H~0, as a standard source of water. The authors have not found it satisfactory, since the crystals are slightly hydroscopic and samples obtained from the supply houses contained up to 103.5% of the theoretical water content. Attempts to dry a t elevated temperature or a t ordinary temperature over calcium chloride led to some loss of water of crystallization. The method of standardization given in the original paper (2) is satisfactory. ACKNOWLEDGMENT
The authors thank Shawinigan Chemicals, Ltd., for permission to publish these results. LITERATURE CITED
(1) Kieselbaoh. R.. IND.EXG.CHEM..ANAL.ED..18.727 (1946). (2) Smith, D. M., Bryant, W. M . D . , and Mitchell, J., J. Am. &em , SOC.,61,2409 (1939). (3) Warren, G. G., Can. Chem. Process Ind., 29,370 (1945). RECEIVED November 25, 1946
Determination of High Molecular Weight Quaternary . Ammonium Compounds as the Triiodides 0. B. HAGER', E. M. YOUNG, T. L. FLANAGAN, AND H. B. WALKER2 Research Laboratories, Rohm & Haas Co., Philadelphia, Pa. T w o qualitative and three quantitative methods of analytical value are described which are based on the insolubility of the triiodides of many high molecular weight quaternary ammonium compounds in water. These triiodides are rapidly precipitated from aqueous solutions by addition of a solution containing iodine and potassium iodide. They separate as
4
H
IGH molecular weight quaternary ammonium compounds have vaulted into tonnage production in the last decade. They have found uses in medicine and sanitation and in the textile, paper, and leather industries. This increasing industrial importance of the quaternaries has been accompanied by a like demand for analytical methods suitable for use as control measures in both their manufacture and the industrial processes in which they are used. Of the many restrictions on the development of analytical 1
Present address, Alco Oil and Chemical Gorp., Philadelphia, Pa. & C o . , Philadelphia, Pa.
* Present address, E. F. Houghton
colored solids which show a marked change in color when more than the equivalent amount of iodine is provided. The triiodides can be isolnted, dissolved in dilute alcohol, and determined colorimetrically or by titration w-ithsodium thiosulfate, but the most accurate and precise method uses potentiometric titration w-ith a solution of iodine. methods, some are intrinsic in the nature of the compounds and others are imposed by technical requirements of the process to be controlled. Most of the industrially important quaternary ammonium salts are dispersible in water to form nearly colorless solutions. The substituents about the nitrogen are frequently without good analytical points of attack or action-that is, without color, double bonds, or ionizing groups. For this reason organic chemists have used nitrogen determinations and some measure of the anion, usually a halide, to guide them in their elucidation of structure and purity. Estimations based on surface activity have been suggested by
V O L U M E 19, NO. 1 1
886 Wakeham and Skau (16) and Preston (14). DuBois (6) has described details for determination of the ionizable chlorides, bromides, or iodides, using the argentimetric methods of Mohr and Fajans. It is highly desirable in many instances to have a direct measure of the quaternary ammonium ion and precipitation methods have been suggested by various 17-orkers. DuBois (5) has reviewed most of those and described in detail a method for precipitation with iodine in potassium iodide solution. In the presence of starch he reports the color change to be “from yellow through various shades of green and quite sharply to blue, and this gradual change in color varies with each compound tested.” Flotow (8) mentions nitric acid, mercurous chloride, iodine in potassium iodide solution, picric acid, Nessler’s reagent, bromine solution, trichloroacetic acid, perchloric acid, ammonium persulfate, potassium ferricyanide, potassium ferrocyanide, silicotungstic acid, potassium dichromate, and cesium sulfate as precipitants of many high molecular weight quaternary ammonium compounds, giving detailed procedures using potassium permanganate in a qualitative test and potassium dichromate in a quantitative method. Precipitation with excess standard potassium ferricyanide solution and iodometric determination of the excess is described in detail in ( 2 ) . Dansi et al. (4)proposed a gravimetric procedure based on precipitation with Reinecke’s acid [Cr(sH,),(SCS),(H,O),1H. Amelink (1) has suggested a variety of complex salts with the noble metals for purposes of microchemical identification of certain quaternary ammonium compounds. The use of anionic dyes as precipitants has been proposed in methods given by huerbach ( S ) , DuBois (j)Hartley , and Runnicles ( I O ) , and Krog and Marshall (12). These have been reviewed by DuBois (’7) and special mention should be made of the proposal by DuBois (5) to determine the amount of quaternary ammonium compound on textiles and paper directly by exposing such treated materials to a solution of an appropriate dye and matching the color obtained on the dyeing of the unknown material with that of some of the same textile or paper Tvhich has been treated with known quantities of the same quaternary ammonium compound and dyed according to the given procedure. For this purpose he recommends using Azo Wool Violet 4B (Colour Index S o . 53) for cotton textiles and paper and Benzo Fast Scarlet 4BA (Colour Index Yo. 327), among others, for wool. The use of iodine, in a solution of potassium iodide, for precipitating high molecular weight phosphonium compounds has been proposed as the basis of quantitative methods by Rath and Schoniger (Is),llecheels ( I S ) , Heidler ( I I ) , and Gross and von Roll (9). The extensive investigations of the structure of the “perhalides” of the quaternary ammonium compounds by T. H. Reade and eo-workers in the 1920’s revealed the rather remarkable stability of the triiodides and the low order of solubility which suggested their usefulness as analytical precipitants. Experience in this laboratory since 1940 has shown this suggestion t o be fruitful as the basis for several successful methods based on the triiodide chemistry of high molecular weight quaternary ammonium compounds and this paper describes these methods. All the important chemistry of analytical value in the methods described in this paper can be summarized in the following equations:
QN
+
+ K13 + starch +yellow t o yellow-brown colors Q Y + + KI3 +QNII + K + colorless &SI3
yellow
yellow
+ KI3 +brownish &NIX+ KI
(1) (2)
(3)
orange yellow
+ K I --+
QNI
white
+ KI3
(4)
where QN+ is a completely substituted quaternary ammonium ion and KIa is, for these purposes, the active material in the ordinary analytical potassium iodide solutions of iodine. QUALITATIVE METHODS
Starch-Iodide Method. If 5 drops of a 5% solution of sodium hypochlorite and 0.5 ml. of glacial acetic acid are added to 25 ml. of test solution and a piece of starch-iodide test paper is inserted, the paper will be made blue if the concentration of the
quaternary ammonium salt is about 0.1% (solids by weight) or less and bright yellow if 0.5y0or more. For concentrations in between, the colors vary from blues to greens through browns t o yellows. This type of procedure can be varied in an obvious way for detecting the presence of a quaternary ammonium compound on a textile. I t ran be done conveniently by immersing the test fabric for a few moments in a dilute solution of acidified hypochlorite solution, removing it, and then spotting it with a starch iodide solution. If a quaternary ammonium compound is present a yellow color will develop; if absent, a blue color. Turbidity Method. If 0.05 S pdtassium triiodide solution (the 0.1 Y iodine solution used for general oxidative analytical purposes) is added dropwise to 25 ml. of previously acidified test solution, a precipitate will develop immediately in the presence of a quaternary ammonium salt. For extremely dilute solutions only pale colored turbidities will develop; at levels of about O.Ol$& the turbidity will be great and there will be decided yellow, orange, to reddish-broivn colored precipitates. QUANTITATIVE METHODS
Determination of Quaternary Ammonium Salts on Textiles and .Paper by Formation of the Triiodide in Situ. If a 0.6-gram piece of paper or fabric which has been impregnated with from 0.25 to 2% of octadecyldimethylbenzylammonium chloride is immersed in a mixture of 25 ml. of water and 1 ml. of 0.05 N potassium triiodide solution and stirred about for 20 minutes, a polyiodide (Equations 2 and 3) will be formed in the paper or cloth. The cloth is then rinsed wit,h two to four 10-ml. portions of 5y0potassium iodide solution until the rinse solution is colorless, usually about three rinses. By this rinsing procedure the polyiodide of rather indefinite composition is reduced in halogen content to the fairly stable triiodide (Equation 3). Prolonged standing during the rinsing or the use of more portions of the potassium iodide solution will give low results because of appreciable removal of iodine of addition by the potassium iodide (Equation 4). The sample is removed from the last rinse and squeezed as free of solution as possible by hand. The triiodide left on the paper or cloth is removed by immersing and “working” t.he sample in about 20 ml. of alcohol (Formula 2B denatured, or any alcohol which can be mixed with dilute potassium iodide solution without becoming turbid is satisfactory). The extract is filtered through a dry sintered-glass filter (porosity equivalent to Jena 1-b G-3), using suction. The filtrate may then be compared for transmittancy at a w v e length of 450 millimicrons against the transmission of the alcohol. The resulting transmittancy is convert,ed to the amount of the quaternary ammonium salt on the textile by reference t o the conversion chart derived from measurements of known amounts of the quaternary triiodide in alcohol. When a photometer is not available the alcoholic filtrate can be diluted with water, acidified, and t.itrated with sodium thiosulfate, I n this case it must be remembered t’hat only two of the three iodine atoms are available for oxidation of the thiosulfate. This method has been used for several years for studying the exhaustion of quaternaries from aqueous solution by cellulose, alt,hough it requires acquired skill on t,he part of the analyst to wash the quaternary polyiodide enough to convert it to the triiodide but not also convert some of the triiodide to t,he simple iodide (Equation 4). The fabric, or paper, should contain very little if any starch. Approximat,ely, 2% tapioca starch is equivalent in consumption of iodine to about 0.8% octadecyldirnethylbenzylammonium chloride. .4number of samples of cloth or paper may be “halogenated” together but each sample must be rinsed individually in the potassium iodide solutions. Once in alcohol the solutions are stable; hence the photometric measurements can be made leisurely. The average of the deviations observed between days is -0.09 and since the average of the deviations for a very large number of observations under statistical control should be zero, we find the probability of difference of average deviations from zero as great or greater than 0.09 to be 18%; hence, we may conclude there is no statistically significant, difference between the two sets of analyses. Rapid Assay Method for Dilute Solutions. An assay method for rapidly estimating the strength of solutions of Hyamine 1622
~
N O V E M B E R 1947
887 PROCEDURE
Table I. Reproducibility of Method by F o r m a t i o n in S i t u
Weigh 0.05 to 0.20 gram of the quaternary ammonium compound to the nearest 0.1 mg. and place the weighed sample in a 250-ml. beaker. [If the quaternary ammonium compound is on a textile material, such an amount of the textile material should be taken as will contain the prescribed quantity of the quaternary and a separation should be made by extracting the quaternary compound with hot ethyl alcohol (denatured Formula 2B is -sat,isfact,ory). The ethyl alcohol extract must' be evaporated t o less than 3-ml. volume, although it is better to evap0rat.e this extract to dryness on a steam bath, providing t>hatthis latter procedure does not cause decomposition of the quaternary compound. The ext,rected solids are then taken up with the prescribed 100 ml. of viater and the determination is carried out from t,his point as described in the procedure.] Add approximately 100 ml. of distilled water to the beaker and heat until the solution becomes homogeneous. Adjust the pH to approximately Table 11. Composition of P e r m a n e n t Color S t a n d a r d s 3.0 by adding a small amount of concentrated hydrochloric acid D y e Blend Dilution of Hyamine 1622 (0.5 to 0.1 ml.). Cool t,he test solution t o 20" C. and by use of a 15 nil. Celliton Yellow 50.4 1:9000 const,ant-temperat,ure bath hold the temperature of this solution 3 drops Resorcin Brown 3R a t 20" =t2 O C. throughout the titration. Place in the test solu10 ml. Celliton Yellow 5G.A 1:8000 tion a platinum and a calomel elect'rode, the lead wires of Ivhich 10 ml. Distilled water 2 ml. Titanox A suswensiona are attached to a potentiometer. Place a small electric or air1 drop %thraquinonk Green G driven mixer in the solution, and drive the mixer rapidly enough 7 . 5 ml. Celliton Yellow 5G.A 1 : 6600 t o cause a turbulent mixing of the solution in t'he beaker. 1 2 . 5 ml. Distilled water 2 ml. Titanox 4 suspensiona Start, the mixer, note the initial potential reading of the solu1 drop Resorcin Brown 3R tion, then add 0.05 S potassium triiodide solution from a buret in 5 ml. Celliton Yellow 5G.A 1 : 5000 increments of about 0.25 ml., recording the potential after each 2 rnl. Titanox A suspensiona 10 ml. Distilled water addition of the solution in the regular manner. Alloiv from 30 to 0 . 1 ml. Kiton-Fast Orange G 60 seconds after each addition of the potassium t'riiodide solution 5 rnl. Celliton Yellow 5G.A 1 : 3000 for the potential reading to become constant. [Potassium tri2 ml. Titanox A suspensiona 15 ml. Distilled water iodide solution is prepared in 0.05 concentration by dissolving 1 drop Kiton Fast Orange G 12.7 grams of iodine cryst,als and 25 grams of potassium iodide 1 drop Anthraquinone Green G crystals in 25 ml. of distilled water and then diluting the resulting 1 drop Quinoline Tellou Concentrated solution to 1 liter. This iodine solution is standardized against sodium thiosulfate in the regular manner. I t must not be overlooked that potassium triiodide which is 0.05 S to Q S + is 0.10 N t o sodium thiosulfat,e. Lower concentrations of potassium triiodide solut,ion (0.025 S , 0.01 S , etc.) can be used to good adblends. and the sludge a t the bottom of the graduate was discarded vantage \Then t,he concentrations of quaternary are unavoidably small. Khen these lower concentrations of triiodide solutions are used in t,he analysis, the same general procedure is followed as outlined. h s the potassium triiodide solution is added to the test [ (1,1,3,3 - tetranietliylbutylphenoxyethoxyethyl~dimethy~benzyl- solut,ion, the potential of the latter solution n-ill decrease sharply ammonium chloride] is based on Equation 3. at, first,, t,hen slon-l!- turn and increase, and finally reach a nearly constant value. The end point is taken at' the inflection point on the rising part of curve. Sear the break point, the triiodide To 25 ml. of a solution of the quaternary salt, about 0.015?, in solution can be added advant,ageously in smaller increments than strength, contained in a 16 X 180 mm. test, tube is added 1 ml. of 0.25 ml. for a more accurate analysis.] 0.005 S potassium triiodide solution. h yellow colored turbidity .liter the titration is completed, wash the electrodes and stirrer indicat'es concentrations equal t.o or greater t,han 0.01570 and a well wit,h ethanol, not with distilled lvater, as the latter \vi11 not brownish orange turbidity indicates concentrations lower t,han remove t8hequaternary t,riiodide precipit,ate. 0.01570. The zone of uncertainty, which includes the errors of Plot the volume of triiodide solution consumed against, the judgment of a number of observers and the error of producing the corresponding observed potentials and determine the end point ident,ical color, amounts to *0.005% of the quaternary. This from the graph in the usual manner. The concent,ration of the test, can differentiate with reasonable accuracy between 1 part of quaternary sample is calculat,ed as follows: the quaternary in 1000 parts of water over the range from 1 to 3000 t o 1 t o 9000. MI. of KI, used X of KI, X E.T. of QX compound X 100 Color standards made from the quaternary and potassium tri1000 X grams of sample iodide do not keep well for much longer than one hour. Perma% compound nent color standards can be made by blending 0.1% solutions of dyes with a suspension of Titanox h in the proportions given in E.W. = 1I.W. only if the QY ion is univalent. Table 11. (Cotton cloth soaked for various lengths of time in various concentrations of octadecyldimetbylbenrylammonium chloride) % ' Quaternary Content Analysis 1 Analysis 2 Deviations, d Sample (1/26-28/42) (2/4-10/42) +o, 08 0.54 0 46 -0.12 0 88 1 0 -0.2 1 3 1 5 -0.1 2 1 2 2 -0 3 2 1 2 4 3 0 +o 1 2 9 f0.2 2 8 2 6 -0.3 2 7 3 ~ n. . -0 2 3.6 9 3.8 d = -0 09
~
~~
~
Potentiometric Determination of Quaternary Ammonium Compounds. The most versatile and precise quantitative procedure is based on a potentiometric titration of the solution of a quaternary salt with iodine solution. The method makes use of the fact that the precipitating agent, potassium triiodide, is also an oxidizing agent. When the quaternary salt is in excess a low oxidation potential is shown by a platinum us. calomel cell electrode pair. A slight excess of potassium triiodide over the end point (Equation 2) results in a much higher oxidation potential and this change is used to indicate the end of the reaction. The formation of the insoluble Qh-1, will be visually evident as a yellow turbidity during the early stages of the titration with a tendency to become rapidly agglomerated as the end point is approached. Obviously, reducing agents and organic compounds rvhich easily form iodine derivatives must be absent, as well as heavy metals which form insoluble iodides and materials which can act as solvents for the otherwise insoluble &IK13.
PRECISION AND ACCURACY
As an example of the precision obtainable by the procedure described, Table I11 gives the results from twelve determinat,ions made of the concentration of a nominally 0.57, solution of octa- ' decyltrimethylammonium salicylate, using 25 ml. aliquots diluted to 100 ml.' Aisevidence for accuracy, Table I11 also gives the results of seven determinations of the concentration of the same solution calculated from the nitrogen contents obtained by the Kjeldahl method. On the basis of the data in Table 111, 99yo of the individual values obtained from the titration of 26-ml. aliquots of the 0.5y0 solution of quaternary salt would be expected to lie within the limits 0.491 to 0.5317, if titrated with 0.05S potassuim triiodide and within the limits 0.465 to 0.531 if determined by Kjeldahl nitrogen. Some evidence of freedom from any gross systematic error was sought by making known dilutions of the same nominally 0.5% solution of octadecyltrimethylammonium salicylate nien-
V O L U M E 19, NO. 1 1
888 Table 111. Reproducibility and Accuracy of Potentiometric Titration of an Octadecyltrimethylammoniurn Salicylate Solution % Solution Concentration Indicated by Titration
-
% Solution Concentration Indicated by Kjeldahl K t r o g e n
0,502 0.507 0.513 0,512 0.516 0.514 0.514 0.520 0.508 0.497 0.515 0.516 Av. 0.511 8 = 6.27 X 10-8 = 6.78 X 10-3
0.487 0.497 0.570 0.497 0.503 0.503 0,495
....
.... .... t...
....
0,498 9.65 x 10-1 11 x lo-’
a = sample standard deviation.
E
estimated universe standard deviation.
Table IV. Systematic Dilution of a 0.05% Solution of Octadecyltrimethylammonium Salicylate
-
(Total volume a t beginning of titration 100 ml.) Relative Normality % QN Salicylate Potential Concentration of KIs Expected Found Rise, Mv.
1. on’ 0.50 0.40 0.20 0.10 0.05
0
0.258 70 100 0,128 0.104 100 0..0537 100 0,0266 100 0.0132 80 0.02 0.0058 80 0,0029 60 0.01 0.0014 0.005 14 0,00057 0,002 6 Relative concentration, with actual concentration of 25 ml /IO0 ml 0.05 0.05 0.05 0.05 0.005 0.005 0.005 0.005 0.0005
0.255 0.127 0.102 0.0512 0.0255 n ,0128 o.on5i 0.0026 0.0013 0.00051
Octadecyltrimethylammoniurn salicylate Dimethyldibenz lammonium methyl sulfate Octadecyldimet iylbenz ylammonium chloride Hexadecyltrimethylammonium bromide
(1,1,3,3-Tetramethylbutylphenoxyethoxyethyl) trimethylammonium iodide (1,1,3,3-Tetramethylbutylphenoxyethoxyethyl) dimethylbenzylammonium chloride Somewhat low results were obtained on technical grades of the following:
Dodecyldimethylbenzylammonium chloride Octyloxymethyldimethylbenzylammonium chloride Octadecylpyridinium bromide No potential break could be obtained with the following: Trimethylbensylammonium phosphate Trimethylbensylammonium iodide Trimethylbensylammonium sulfate Trimethylbensylammonium methyloxalate (Phenylmercuri)trihydroxyethylammonium cu-hydroxypropionate Tribenzylmethylammonium iodide Octadecylm~thallyldimethylammoniumchloride
Table V. Effect of Ethyl Alcohol and Acetic Acid on KII Titration of Octadecyltrimethylammonium Salicylate ’ C o n c e n t r 3 o n Added Acetic acid Ethyl alcohol 0 0 2 0 3 0
5 10
0 0
0 tioned previously For this purpose carefully measured aliquots were taken from the stock solution and in all cases made to a total of 100 ml. for titration. As the concentration of quaternary about the electrodes decreased, it became necessary also to decrease the concentration of the potassium triiodide solution from time to time. As may be expected, the voltage change a t the end point became less as the concentration of quarternary decreased. These observations are evident from the results of the experiments reported in Table IV. INTERFERENCES
h1:hough experiments have shown that large amounts of cetyl alcohol, ethyl alcohol, and acetic acid frequently tend to give low results, the magnitude of the effect depends very markedly on the exact structure of the quaternary salt being titrated. In the case of the solution of octadecyltrimethylammonium salicylate used for the previously described results, relatively large amounts of ethyl alcohol and acetic acid could be tolerated, as shown in Table V. For these experiments 25-ml. portions of the 0.5% solution were diluted to 100-ml. total volume with water and the various amounts of alcohol or acetic acid as noted in the table. APPLICABILITY
Basically the potassuim triiodide titration of quaternary compounds is a precipitation reaction and, consequently, only quaternaries forming triiodides which are quantitatively insoluble can be determined by this procedure. Such quaternary must not have in it groups which can function as reducing groups for the iodine. The determination of purity of technical grade quaternaries may find some interference from relatively large quantities of high molecular weight tertiary amines present because of incompleteness of the quaternizing reaction. The potassium triiodide method has proved applicable to the following compounds:
Octadecyltrimethylammonium trichlorophenate Octadecyltrimethylammonium pentachlorophenste Octadecyltrimethylammonium chloride
%
QN Concentration Found by KI: Titration
.
0.491-0.531 0.527 0.523 0.532 0,532 0.532 0.514 0.502 0.510 0.503 0.518
SUMMARY
The reaction of potassium triiodide with many high molecular weight quaternary ammonium compounds to form insoluble triiodides forms the basis for two qualitative and three quantitative methods. One of the quantitative methods is simple enough for health inspectors’ field use in testing sanitizing rinse solutions for table ware and the other is a versatile laboratory potentiometric method suitable for rapid determination of moderate amounts of compounds in solution and on paper and textiles. LITERATURE CITED
(1) Amelink, F.,Pharm. Weekbkzd, 69,1221-3 (1932). (2) American Medical Assoc., Council on Pharmacy and Chemistry, “New and Non-Official Remedies,” 1944. (3) Auerbach, M. E., IND. ENG.CHEM.,~ N A L .ED., 15, 492-3 (1943); 16,739 (1944). (4) Dansi, A. O.,Mamoli, L., and Ciocca, B., Ann. chim. applicata, 22,561-5 (1932). (5) DuBois, A. S.,Am. DyestufReptr., 34,245-6 (1945). (6) DuBois, A. S.,IND.ENG.CHEM., ANAL.ED.,17,744-5 (1945). (7) DuBois, A. S., S o a p Sunit. Chemicals, 22,125,127,129,131,139. 141,143 (Nov. 1946). (8) Flotow, E.,Pharm. Zentralhulle, 83, 181-5 (1942). (9) Gross, J. W.,and von Roll, E., Melliand Teztilber., 21, 52543 (1940). (10) Hartley, G.S., and Runnicles, D. F., Proc. Roy. Soc. London, Series A, 168,420-40 (1938). (11) Heidler, K., Melliund Teztilber., 21,407-9 (1940). (12) Krog, A. J., and Marshall, C. G., Am. J . Pub. Health, 30,341-8 (1940). (13) Mecheels, O.,et al., Melliund Teztilber., 16,42-5 (1935). (14) Preston, J. M.,J . Soc. Dyers Colourists, 61,165-6 (1945). (15) Rath, H., and Schoniger, G., Klepzig’s Teztil-Z., 41,129 (1938). (16) Wakeham, N., and Skau, E., J . Am. Chem. Soc., 67, 268-72 (1945). RECEIVED August 30, 1945. Presented before the Division of Analytical and Micro Chemistry a t the Meeting-&Print of the AMERICAN CHEMICAL
SOCIETY, 1945.
