February, 19581 POLY (ETHYL ESTERS) IN METHYL METHACRYLATE COPOLYMERS 71 The Separation of Quaternary Halides by Paper Chromatography BY H. HOLNESS* AND W. R. STONE (Chemistry Department, South West Essex Technical College, Walthamstow, London, E . 17) Separations within the homologous series of certain surface-active cationic germicides by means of paper chromatography are described. By using a new spray reagent, the detection on the paper of as little as 0.6p.g of the quaternary base is possible. WITHIN recent years increasing use has been made in both industry and medicine of germicides whose activity is derived from the presence of surface-active quaternary ammonium or pyridinium salts, and this entails a need for their recognition. Since the publication of our Notel on the paper chromatography of cationic surface-active agents, three papers2JA have been published on this subject.An electrochromatographic method used by Fumasoni, Mariani and Torraca2 for the separation of the homologous series of n-alkylpyridinium halides requires amounts of material greater than 50 pg and the spots formed exhibit “tailing.” An ascending chromatogram described by Garcia and Couerbe3 separates commercial products having mixtures of chain lengths in the n-alkyl radical, but the minimum amount detectable is 5 pg. Negoro and Sen04 also used electrochromatography to separate quaternary bases from non-ionic surface- active agents when present in admixture, but they again experienced considerable “tailing.” The method described in this paper allows separations within the homologous series of saturated ut-alkyltrimethylammonium halides, n-alkylbenzyldimethylammonium halides and * Present address : Chemical Engineering Department, College of Science and Technology, Manchester, 1.72 HOLNESS AND STONE: TH:E SEPARATION OF [Vol.83 n-alkylpyridinium halides when the alkyl chain possesses from twelve to eighteen carbon atoms. The limit of detection is of the order of 0.6pg. EXPERIMENTAL APPARATUS AND TECHNIQUE- All the chromatograms were prepared by using the ascending-solvent technique and Whatman No. 1 paper (for chromatography). Two types of container were used, the first a rectangular tank, which allowed two sheets, 15.75 cm x 33.5 cm, to be run simultaneously. The second container was a large cylinder with vertical sides 30 cm high and with a diameter of 15 cm.This allowed one sheet, 33.5 cm x 31.5 cm, to be used, this sheet being rolled into a tube as described by Wolfson, Cohn and De~aney.~ In this manner a dozen spots could be run on the one sheet. The containers were closed and sealed by covering the top with a sheet of polythene film and securing it round the sides with a strong rubber band. Both the containers were placed in a thermostatically controlled oven adjusted to a temperature of 30" C. All the quaternary salts used in this work were white crystalline products and were supplied as single substances containing only the walkyl chain length indicated by its name. For application to the chromatogram, the salts were dissolved in 50 per cent. aqueous ethanol to concentrations that allowed volumes not greater than 0.003 ml to be spotted on the paper. These spots were applied along a line 3 cm from the base by means of a capillary pipette graduated in 0401 ml.When the chroniatograms had run the required distance, they were removed from the container and allowed to dry by exposure to air at room tem- perature. In this connection it is of interest to note that at a certain time in the process of drying, the separated spots of the quaternary base became faintly visible by reflected light as lighter and more opaque areas on the still damp paper. I t is thought that this might be due to a decreased rate of evaporation of the solvent from the paper in the presence of the surface-active agent. It is, however, unlikely that this could be of general use for determining the position of the spots, since the phenomenon was only observed when the material was present in relatively high concentr, t' ions. When the papers were dry, they were sprayed with the reagent and examined under ultra-violet light while damp.The spots of the quaternary base fluoresced a bright vermilion on a bright white background only faintly tinged with pink; their positions were marked by means of a soft pencil. REAGENTS- The solvent mixtures used in developing the chromatograms, 35 per cent. and 40 per cent. aqueous ethanol, were prepared by mixing 35 and 40 ml, respectively, of 96 per cent. V/V industrial methylated spirit with 5 ml of concentrated. hydrochloric acid and diluting each mixture to 100ml with water.The indicator spray reagent was freshly prepared immediately before use by mixing 5 ml of a 0.2 per cent. w/v aqueous solution of rhodamine BS with 10 ml of a 0.2 per cent. w/v aqueous solution of Tinopal WG, adding 40 ml of ammonia solution, sp.gr. 0.880, and diluting with water to 100ml. RESULTS With both the developing solvents clear separations were obtained within each homo- logous series of the lauryl, myristyl, cetyl and stearyl derivatives. The RF values quoted in Table I represent the average in each case of at least thirty determinations carried out with a solution containing equal weights of each member in the series. It was unfortunate that no pure sample of benzyldimethylmyristylalnmonium chloride was available and the R, value quoted in this instance is that found when a commercial product was examined.It will be seen that the benzyldimethyl-la.urylammonium, benzyldimethylmyristyl- ammonium and benzylcetyldimethylammonium halides can be differentiated from the corres- ponding n-alkyltrimethylammonium and n-alkylp~7ridinium halides, but only poor separation of the stearyl derivatives is shown. The R, values of the n-alkyltrimethylammonium and n-alkylpyridinium salts in both solvents are practically identical and separations could not be achieved, although the pyridinium salts generally showed slightly lower R, values. Better separations might well be achieved by allowing the solvent to run a greater distance, but limitation of the size of the containers available did not allow the examination of this possibility,February, 19581 QUATERNARY HALIDES BY PAPER CHROMATOGRAPHY TABLE I R, VALUES FOR PUKE COMPOUNDS Paper, Whatman No.1 Temperature] 30' C n-A lkyltrimethylammonium salts- Lauryl .. .. .. Myristyl . . .. .. Cetyl . . .. .. Stearyl . . .. .. Lauryl . . .. .. Myristyl* .. .. Cetyl . . .. .. Stearyl . . .. .. n-Alkylpyridinium salts- Lauryl . , 6 . .. Myristyl . . .. * . Cetyl . . .. .. Stearyl . . .. .. n-A lkylbenzyldimethyl- ammonium salts- RF with 35 per cent. ethanol 0.86 0.65 0.36 0.10 0.7 1 0.39 0.20 0-06 0.81 0.58 0.3 1 0.08 Time of run, 7 hours approximately Length of run, 27 cm approximately Standard deviation 0-03 0.03 0.02 0.01 0.04 0.02 0.04 0.01 0.03 0.05 0.04 0.02 RF with 40 per cent. ethanol 0.92 0.79 0-65 0-23 0.85 0.61 0.40 0.14 0.90 0-76 0.64 0-20 Standard deviation 0.02 0.02 0.04 0.04 0.02 0.01 0.03 0.03 0.01 0.02 0.03 0.03 73 * These values were found by using a commercial material of mixed chain lengths.In Table I the figures for the standard deviations are derived from the values for several separate runs. Some of these values are rather large, but this is due to variation from one sheet of paper to another. When several spots of identical material are run simultaneously on the same sheet of paper, the spread of R, values for each component is seldom greater than +0*01. It is therefore recommended that, for purposes of identification, control spots of Known composition should be run on the same sheet as spots of an unknown mixture. Number of carbon atoms in n-alkyl chain Graph of RM values plotted against n-alkyl chain length, with the 35 per cent.ethanol mixture as developing solvent: curve A, n-alkylbenzyldimethy1ammonium homologues; curve B, n-alkylpyridinium homologues; curve C, n-alkyltrimethylam- monium homologues Fig. 1. The R, value, log (l/RF - l), proposed by Bate-Smith and WestallJ6 was evaluated from the R, values shown in Table I and the values obtained within each homologous series were plotted against the corresponding number of carbon atoms in the 12-alkyl chain. The points74 HOLNESS AND STONE: THE SEPARATION OF [Vol. 83 on each graph fell close to the straight line drawn through points calculated by the method of least squares to give the best fit to the experimental data. It is of interest to note that the R, value obtained from the graph for the n-octyl, n-decyl and n-tetradecyl derivatives of the homologous series of n-alkylbenzyldimethylammonium salts were in close agreement with those found experimentally from the chromatogram of the commercial alkylbenzyl- dimethylammonium halide by using the 35 per cent. ethanol solvent.With 40 per cent. ethanol as solvent (see Fig. 1), the rt-octyl and rt-decyl derivatives formed only one spot near the solvent front. LIMIT OF DETECTION- but the cetyl and stearyl derivatives required 0.6 pg for positive identification. The lower members of the series gave definite spots when present to the extent of 0-2 pg, DISCUSSION In the initial stages of the work the chromatograms were run at room temperature, but it soon became evident that the variation in temperature was having an effect on the R, values.It was therefore decided to run the chromatograms at a temperature sufficiently above average room temperature to allow thermostatic control. The temperature selected was 30” C and was obtained by using an electric oven thermostatically controlled to +lo C. In a search for a sensitive means of revealing the position of the spots of the quaternary base on the developed chromatogram, a similar line of reasoning was followed to that which led to the formulation of the mixed indicator used in the scheme of semi-micro qualitative analysis of surface-active agents recently p~blished.~ Initially, a series of anionic dye-stuff s that had the property of fluorescing in ultra-violet light was tested, and with many compounds it was found that the presence of a quaternary base quenched or modified the colour of the fluorescence.Anionic optical bleaches were also found to have their fluorescence quenched by quaternary bases, but the background was then too bright to give a sensitive contrast. In order to increase this contrast, fluorescent ca.tionic dyestuffs were added to the optical bleach and this improved the sensitivity. The most sensitive reagent found was the optical bleach Tinopal WG mixed with the dye-stuff rhodamine BS or rhodamine B5OO. Previously,l we developed the chromatogram by spraying with a solution of these two materials and then exposing to the vapours of ammonia to neutralise any mineral acid remaining on the paper. Here the spray reagent contains ammonia in solution and the additional stage of exposing to ammonia vapours is unnecessary. The position of the spots must be marked while the paper is still damp, since the change in fluorescence in ultra-violet light caused by the presence of the quaternary base is substantially eliminated when the paper is dry.Several solvent systems were originally tried and were composed of various concen- trations of aqueous ethanol. In an effort to reduce “tailing,” 5 per cent. v/v of ammonium hydroxide, of glacial acetic acid and of concentrated hydrochloric acid were added in turn to the solvent, replacing an equivalent volume of water. Of these additions only concentrated hydrochloric acid proved satisfactory, as the lauryl and myristyl derivatives gave good circular spots, the cetyl derivatives gave rather elongated ovals and only the stearyl derivatives showed signs of “tailing.” It was found that, when a mixture of equal amounts of all four members of a given homologous series was chromatographed and allowed to run approximately 27 cm, the spots of the lower members became too large to allclw satisfactory separation when each spot contained more than 6 pg of material. This fact combined with the limit of detection of the spray reagent would suggest that approximately 10 per cent.of any particular chain length within a mixture of members of an homologous series could be detected when run from a single spot under the experimental conditions described. It would seem possible that a smaller percentage of any given chain length in a mixture of chain lengths might be detected if the solvent were allowed a longer run. This .would allow a greater separation between adjacent spots with the consequence that each could contain more than 6p.g of material. We should like to place on record our thanks to Mr. P. A. Lincoln, M.Sc., of Milton Industrial Chemicals Ltd., for the supply of the pure salts used in this work. We also thank Messrs. Leda Chemicals Ltd., for the commercial sample of “benzalconium” chloride, Messrs.February, 19581 QUATERNARY HALIDES BY PAPER CHROMATOGRAPHY 76 Geigy & Co. Ltd. for the Tinopal WG and Imperial Chemical Industries for the rhodamine dye-stuff. REFERENCE s 1. 2. 3. 4. 5. 6. 7. Holness, H., and Stone, W. R., Nature, 1955, 176, 604. Fumasoni, S., Mariani, E., and Torraca, G., Cham. & Ind., 1956, 69. Garcia, I., and Couerbe, J., Chim. Anal., 1956,38, 432. Negoro, H., and Seno, S., Ann. Rep. Takamine Lab., 1957, 8, 119. Wolfson, W. Q., Cohn, C., and Devaney, W. A., Science, 1949, 109, 541, Bate-Smith, E. C., and Westall, R. G., Biockim. Biophys. Acta, 1950, 4, 527. Holness, H., and Stone, W. R., Analyst, 1957, 82, 166. Received August 19th, 1967