Table IV. Recovery of Tween 20 from Mixtures of Surface-Active Agents illicrogran~s/Milliliter Tween .... ~
Tween 20
added 7.25 7.25 7.25
IMax. concn. of additive Span 20, 0 . 1 2 Cetylpyridium chloride, 0 . 6 7 Sodium laurvl sulfate, 0 .“07
20
recov- Recovered ery, 70 7 . 3 5 101.4 7 . 4 0 102.0 7 . 2 8 100.40
Data recorded in Table IV indicate that good recoveries of Tween 20 are possible in the presence of low concentrations of Span 20 (nonionic), cetylpyridium chloride (cationic), and sodium lauryl sulfate (anionic). For successful assay, an anhydrous medium must be used and, as summarized in Table I, several variables affect the reaction. The time of heating is rather critical, 10 minutes or more being required for the completion of the reaction. For optimum results, the 2,4DNPH reagent must contain a t least 0.5 mg. of 2,4DNPH/ml. and the acid normality must be a t least 0.05. Although carbonyl-free alcohol or alcoholic KOH can be used as a diluent for the color developed, high final concentrations will lead to salt formation, hence distilled water is the preferred diluent. If required, the proposed method can be scaled down to meet any obtaining conditions. The reaction apparently depends on the presence of a free hydroxyl group since the glycols and monoesters of polyoxyethylene glycols responded, but the diester fraction as determined by the method of Papariello et al. (18) did not. This is indicated by the results which are summarized in Table 111. The negative results obtained with the fractions listed may be due to the total absence of any of the various forms of the compounds or, alternately, to the
presence of glycol diesters which do not respond. When the diester fraction was saponified according to Birkmeier and Brandner (1) and then tested, a positive test was obtained. Since ethylene glycol does not respond while di- and tri-ethylene glycols as well as propylene glycol do, it seems that color development depends on the molecular weight of the compound and that a minimum chain length of the order of propylene glycol or diethylene glycol is necessary. Hydrazones of Tween 20, Altox 1045, Ethomid HT/60, Igepal, Brij 30, blyrj 45, and Pluracol P-2010 were recrystallized from alcohol three times, air-dried, and their melting points determined. All melting points fell within the range of 320“ to 325’ C. This similarity indicates that a common mechanism of action is involved. The hydrazone formed is not that of formaldehyde nor acetaldehyde, since these compounds give red or wine colors with absorption maxima a t 440 and 425 mp, respectively (4, IS, 16). Possibly, in the process, bisphenylhydrazones are formed. iiS suggested by the work of Khym and Cohn (8) these substances can then further react with phenylhydrazine to foim two- and threecarbon fragments. Since ethylene glycol does not respond, it is suggestive that very little or no two-carbon fragment results. The positive results obtained with propylene glycol do suggest the production of a three-carbon fragment. The exact mechanism, however, awaits further elucidation. Commerical samples of diesters may give positive tests because of the presence of monoesters and glycols (1). Furthermore, during the heating processes, some hydrolysis may occur resulting in reactive forms. , Addition of the alcoholic potassium hydroxide to the 2,4-dinitrophenylhydrazone produces a very intense purple color because of, presumably, the formation of a resonating quinoidal
ion (10). A similar quinoidal ion has been suggested for the colored solution formed when alcoholic potassium hydroxide is added to the phenylhydrazone of a nitroaromatic aldehyde (8). ACKNOWLEDGMENT
The authors thank the several manufacturers who so kindly donated samples of the various surface-active agents. A portion of this research was financed by a grant from the Petroleum Research Fund. LITERATURE CITED
(1) Birkmeier, R. L., Brandner, J. D.,
J.Agr. Food Chem. 6,471 (1958). (2) Brokke, ill. E., Kiigemagi, U., Terriere, L. C., Zbid., 6,26 (1958). (3) Chataway, F., Ireland, S., Walker, A., J . Chem. SOC.1925,1851. (4) Dunn, C. L., J . -4gr. Food Chem. 6, 203 (1958). (5) Ginn, M. E., Church, C. L., Jr., Harris, J. C., ANAL. CHEM.33, 143 (1961). (6) Gunther. P. A,, Glinn, R. C., Kolbezen, J., Barkley, J. H., Harris, W. D., Simon, H. S.,Ibid., 23, 1835 (1951). ( 7 ) Hobson, B. C., Hartley, R. S., Bnalyst 85, 193 (1960). (8) Khym, J. X., Cohn, W. E., J . Am. Chem. SOC.84,6380 (1960). (9) Kilheffer, J. V., Jr., Jungermann, E., ANAL.CHEM. 32, 1178 (1960). (10) Lappin, G. R.. Clark, L. C., Ibid., 23,541 (1951). (11) McCutcheon, J. W., “Detergents and Emulsifiers Up to Date,” J. W. McCutcheon, Inc., 475 Fifth Ave., New York 17, N. Y., 1960. (12). Papariello, G. J., Chulkaratana, S., Higuchi, J., Martin, J. E., Kuceski, V. P., J . Ani. Oil Chem. SOC.37, 396 (1960). (13) Roberts, J. D., J. Am. Chem. SOC. 68,214 (1946). (14) Rosen, M. J., ANAL.CHEM.27, 787 (1955). (15) Ibid., 29, 1675 (1957). (16) Sharefkin, J. G., Sulzberg, T., Ibid., 32,993 (1960). RECEIVED for review August 15, 1960. Acce ted June 23, 1961. Data taken from M. S? thesis submitted by Lacey Gatewood, Jr., to Tuskegee Institute, August 1960.
Prepa ration of Carbonyl- Free Solvents DANIEL P. SCHWARTZ and OWEN W. PARKS Dairy Products laboratory, U. S. Department o f Agriculture, Washington 25, D. C.
b A
method is described for the Preparation of CarbOnYl-free Solvents. A Celite column impregnated with 2,4dinitrophenylhydrazine, phosphoric acid, and water is used to effect a rapid quantitative reaction with the carbonyls in the solvent. Partially deactivated alumina is employed to
1396
ANALYTICAL CHEMISTRY
isolate the monocarbonyl fraction which is subsequently estimated. The procedure is theoretically applicable to all nonoxygenated, water-immiscible sol-
vents On a continuous basis. The aliphatic monocarbonyl content of 1 3 solvents is presented.
0
for use in the microanalysis of carbonyls are usually purified by refluxing a derivative of hydrazine in the solvent in the presence of an acid catalyst. The most common derivative employed is 2,4dinitrophenylhydrazine (DNPH) since it is highly nonspecific, reacts rapidly, RGANIC SOLVENTS
and yields hydrazones which are relatively acid stable. I n the instance where the reagent is insoluble or only very slightly soluble in the solvent, other methods must be resorted to. These, for the most part, consist essentially of reaction with an oxidizing agent or reaction with DNPH a t the interface of a two-phase system. Begemann and DeJong (1) evaluated the more popular methods of forming 2,4-dinitrophenylhydrazonesand found them unsuitable for quantitative microanalysis. They described a two-phase quantitative method in which the aqueous phase (containing DKPH and HCl) was supported on Celite, and the nonpolar phase (petroleum ether containing the carbonyl) was permitted to flow over the support. In this manner they purified petroleum ether, but indicated that small amounts of acetone still remained unreacted. Their technique, although giving excellent yields was rather slow and ostensibly practical (as far as solvent purification is concerned) where only small volumes of solvents are employed. However, the method appeared to have excellent potentialities and considerable time was spent in this laboratory in a study of improvising modifications without impairing its sensitivity and quantitative aspect. REAGENTS A N D APPARATUS
.-llumina, Grade F-20, Iluminum Co. of America. Activated by heating 24 hours at 150" C., then partially deactivated by the addition of 6% (w./w.) distilled water. The wet alumina was shaken until all lumps were broken and allowed to equilibrate overnight. Chromatographic tubes, Type 1, approximately 2.5 cm. i.d. x 30-cm. length, with or without coarse sinteredglass disk; Type 2, approximately 12 mm. i.d. X 30-cm. length, with or without coarse sintered-glass disk. Solvents Investigated. Table I lists the solvents which were investigated. All solvents except chloroform were analyzed directly from the container. Chloroform was washed six times with half its volume of distilled water, then dried 24 hours over CaClz before purification. Unwashed chloroform could not be purifipd (see discussion). METHODS
Preparation of Reaction Column. Five tenths gram of D N P H is dissolved in 6 ml. of 85% H J P O ~by grinding in a 6-inch mortar. Four milliliters of distilled water is added to the clear yellow solution and the precipitated D N P H is redissolved by grinding. Ten grams of Celite (JohnsManville, analytical grade) is then ground with the solution until a homogeneous damp preparation is obtained. This is best attained by grinding, scraping the sides with a spoon
Solvent Benzene Carbon tetrachloride Chloroform Cyclohexane Ethylene chloride n-Heptane n-Hexane n-Hexane Methyl cyclohexane Methylene chloride n-Pentane Petroleum ether (35.3-52.8)
Toluene
Table 1. Sohtents Investigated Grade Supplier J. T. Baker ACS Fisher ACS Fisher ACS Fisher Reagent Fisher High purity Phillips Fisher Phillips High purity Yellow Label Eastman Kodak Cert. reagent Fisher Blue Label Eastman Kodak Fisher ACS ACS
or spatula, and regrinding. The bright yellow impregnated Celite is then transferred to a Type 1 chromatographic tube which is clamped a t the outlet and contains about 25 ml. of carbonyl-free hexane. The impregnated Celite is transferred in about four equal portions. Each portion is tamped tightly before addition of the next. In this laboratory an all-stainless steel tamping rod similar to that described by Corbin, Schwartz, and Keeney (3) is used, except that the disk clears the sides of the chromatographic tube by about 2 nim. The flow rate of the column can be varied by the tightness of the packing. Although this is not reproducible exactly from column to column, and, of course, varies with the physical characteristics of the solvent to be purified, the desired flow rate can be approximated by varying the head pressure or by incorporating an inert valve assembly on the outlet of the tube. With a little experience, columns can be prepared in 5 to 10 minutes with flow rates close to those desired. Removal of Impurities from Column. The 2,4-dinitrophenylhydrazine reagent contains a small amount of colored impurities, some of which behave like aliphatic inonocarbonyl dinitrophenylhydrazones on alumina (see below). These impurities are removed before addition of the solvent by flushing the column with 50 ml. of carbonyl-free benzene using air pressure. Purification of Solvents. The solvent to be purified is stirred with excess D N P H for 15 minutes, and an aliquot is removed to serve as a blank. The solvent containing D N P H is then added to the column, and the first 50 ml. of effluent is discarded. The remainder is collected, and an aliquot taken for analysis of its aliphatic monocarbonyl content. Isolation and Estimation of Aliphatic Monocarbonyls. To ascertain the degree of purification of the solvent, the aliphatic monocarbonyl content of the effluent is determined as follows: T o about 10 ml. of hexane in a chromatographic tube (Type 2) is added 5 grams of alumina, a little a t a time, with shaking and the column packed by using light air pressure, leaving about 2 cm. of hexane above the
Fisher
Lot 9299 794616 705603 7oooo4 705088 828 54547
794575 705514
top of the bed. An aliquot of the solvent containing the dinitrophenylhydrazones is then added. Aliquot8 of hexane, heptane, carbon tetrachloride, pentane, cyclohexane, and methyl cyclohexane may be added directly. However, solvents of greater polarity (e.g., benzene, toluene, chloroform, methylene and ethylene chloride) are first taken to dryness and the residue wetted with 2 ml. of benzene (per 25-ml. aliquot evaporated) and then diluted with 20 ml. of hexane and transferred to the alumina along with hexane rinsings. Following adsorption of the hydrazones, the aliphatic monocarbonyls are eluted with 50 ml. of a 1:l benzene-hexane solution. This volume of benzenehexane solution will elute 8 pmoles or less of aliphatic monocarbonyl hydrazones from the alumina. Studies on 40 model compounds, all of which contained no other functional group, revealed that quantitative recovery was achieved, and that no artifacts were produced. I n the analysis of the solvents in this study no band was detected which moved more than 2 cm. from the top of the column after elution of the monocarbonyl fraction. The reagent (DNPH) will not move off of the column even after 100 ml. of the benzene-hexane solution has been put through. The absorbance of the monocarbonyl fraction was read in chloroform after removal of the eluting reagent on the steam bath. The blank (solvent saturated with DNPH) was carried through all steps except the reaction column. The purpose of the blank is to correct for any decomposition of D N P H which might occur on standing in solution or on contact with the alumina to give products which emerge with the monocarbonyl fraction. Also any optical absorption of the impurities in the residue of the solvents is accounted for when the absorbance of the monocarbony1 fraction is determined. The remainder of the solvent was distilled a t atmospheric pressure, and subjected, if necessary, to the exact same procedure again. VOL. 33, NO. 10, SEPTEMBER 1961
1397
~~~
Table II.
of Various Organic Solvents ~l~~ Aliphatic 2,4-Dinitrophenylhydazones Found. Rate 1st pass Max.b 2nd ass Max.* 3rd ass Solvent (Ml./Hr.) (pmole/l.) (mp) (pmore/l.) (mp) (pmope/l.) 50 8 362 0 ... Benzene 58 0 .. ... Carbon tetrachloride 47 7 362 0 ... Chloroform Cyclohexane Ethylene chloride n-Heptane n-Hexane (Fisher) 47 11 n-Hexane (Phillips) 362 1 362 37 333 362 14 362 0 Methyl cyclohexane Methylene chloride n-Pentane Petroleum ether Toluene a Calculations based on readings at maxima using a molar absorptivity of 22,500 for saturated aldehydes and ketones. b Maxima in CHCll on total derivatives. Aliphatic Monocarbonyl Content
RESULTS A N D DISCUSSION
Results of the study are presented in Table 11. The efficiency of the reaction column in removing carbonyl contaminants from the various solvents studied is readily evident from the data in Table 11. Although it is to be expected that different lots of a given solvent will vary in their degree of contamination with carbonyls, relative contamination for the different solvents of a given grade should be fairly constant. The data in Table I1 could, therefore, ostensibly be used as an index for selection of a solvent. Although the work reported here was pertinent to the aliphatic monocarbonyl contaminants in the various solvents, limited experiments with aromatic carbonyls dissolved in hexane indicate that these react instantaneously on contact with the reaction column. This is manifested by the formation of a
red band a t the top of the column. This phenomenon takes place only when the aromatic carbonyl is put on the column in a solvent in which the resulting hydrazone is insoluble or feebly soluble. The data of Cheronis and Levey (2) suggest that aromatic carbonyls react more rapidly than aliphatics with DNPH. Vicinal dicarbonyls which form bis(hydrazones) also precipitate as a red zone when the parent carbonyl is put on the column in an aliphatic hydrocarbon solvent. As mentioned earlier, unwashed chloroform could not be purified on the reaction column. This is attributed to the formation of acetaldehyde via the dehydrogenation of DNPH of the ethyl alcohol used as a preservative (4). Theoretically, doubling the constituents making up the column while keeping the flow rate the same as reported in Table 11, should be as efficient as making two passes of the
solvent over a standard column. This was not attempted, since, as the data show, one pass over a standard column will purify most solvents to a satisfactory degree. The data in Table I1 were obtained on two reaction columns. In theory a reaction column can be used indefinitely, a t least for the purification of some solvents, since DKPH from the DNPHsaturated solvent displaces any DNPH lost from the column through reaction. In this laboratory a single column has been used over 6 months for the purification of benzene with no loss in efficiency. However, columns used for the purification of hexane become dark with time. This is attributed to the accumulation of decomposition products of DNPH, these being virtually insoluble in hexane. As a consequence, new columns are constructed periodically for purification of this solvent. Purification of solvents can be put on a continuous basis. At the flow rates given in Table 11, approximately 1 liter per 24 hours per column can be obtained. LITERATURE CITED
(1) Begemann, P. H., De Jong, trav. chim; 78,275 (1959). (2) Cheronis, K. D., Levey, Microchem. J. 1, 223 (1957). (3) Corbin, E. A,, Schwarts, Keenev. hl.. J . Chromatog.
(1960): '
K., Rec. V. M.,
D. P., 3, 322
(4) Caddis, -4. SI., Ellis, R., Currie,
G. T., unpublished data.
RECEIVED for review April 25 1961. Accepted June 28, 1961. Work done at a laboratory of the Eastern Utilization Research and Development Division, Agricultural Research Service, U. S.D. A., Washington 25,,D. C. The use of trade names does not imply endorsement of the product or its manufacturer by the U. S. Department of hgriculture.
An Acid Method for the Volumetric Estimation of Water-Soluble Dithiocarbamates M. L. SHANKARANARAYANA and C. C. PATEL Department o f Inorganic and Physical Chemistry, Indian Institute of Science, Bangalore- 1 2, India Dithiocarbamates have been estimated previously b y reaction with a strong acid, the carbon disulfide evolved being converted into a xanthate and the latter estimated iodimetrically. In the present method, a water-soluble dithiocarbamate is reacted with a decinormal mineral acid and the excess acid is determined to compute the amount of dithiocarbamate present. This method is applicable
1398
ANALYTICAL CHEMISTRY
for the determination of a dithiocarbamate in a thiuram disulfide.
D
mixture containing
find a variety of applications as plant fungicides, anal! itical reagents, accelerators in the rubder industry, and collectors in the flotation of minerals. Hence, methods for their estimation are becoming more ITHIOCARBAMATES
important. I n the acid method for the determination of a dithiocarbamate, employed by Callan and Strafford (d), the dithiocarbamate is decomposed by gently boiling with 30% sulfuric acid, the carbon disulfide evolved is absorbed in alcoholic potassium hydroxide to form an alkali metal xanthate, which is estimated iodimetrically, and the amount of dithiocarbamate is computed from the titer value. In Roth
