AUGUST, 1912, Vol. XXXVII., NO. 437. THE ANALYST. PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS. ON THE APPLICATION OF ADSORPTION TO THE DETECTION AND SEPARATION OF CERTAIN DYES. BY A. CHASTON CHAPNAN AND ALFRED SIEBOLD. (Read a t the Meeting, June 5, 1912.) DURING recent years a very considerable amount of attention has been devoted to the study of certain phenomena to which the general term ‘( adsorption ” has been applied.Whatever may be the precise explanation of these phenomena, the facts are comparatively simple, and many instances-e.g., the decolorisation of solutions by means of animal charcoal-have been known and utilised for a very long time. Among adsorbent substances, kaolin occupies an important position, and a number of references to its power of removing dyestuffs from solution are to be found in chemical literature.These, as a rule, occur in communications dealing with the theory or practice of dyeing, and, with the exception of the valuable work of Gop- pelsroeder on the separation and identification of certain colouring matters by capillary methods, no systematic study of this property, from the analytical stand- point, has, so far as we are aware, been made.It is true that H. Wislicenus (Zeitsch. Chem. Id. KoZZoide., 1908, 2, Supt., II., 11-20) has suggested the application of the so-called (( fibre-alumina ” to the examination of various kinds of extracts, including dyes, and that LaWall has employed kaolin as one of the reagents in a scheme for the detection of certain natural colouring matters (“ Select Methods of Food Analysis,” Leffmann and Beam, 2nd Edition, pp.74-75), but no very definite results appear to have been obtained, and no mention is made of the separation and identification of artificial dyes. Among the more important papers dealing with the question of adsorption of coal-tar dyestuffs, attention may be specially directed to a paper by Suida (lkfonatsh. Chem., 1904, 25, 1107), in which the adsorbing properties of kaolin and other silicates and silicious materials have been studied.Few problems are more difficult than the identification of the dyes used in the colouring of foodstuffs, particularly when two or more are present together, and no apology is needed for bringing to the notice of analysts any method which is capable of facilitating such investigation, even if only to a small extent.The experiments described below had their origin in some observations made in the laboratory of one340 CHAPMAN AND SIEBOLD : APPLICATION OF ADSORPTIOK of US, that certain artificial dyestuffs could be readily separated by the use of kaolin, and that this property could in some cases be usefully applied to the examina- tion of artificially coloured food-products.The method of experimenting was very simple, and consisted in stirring in R small glass mortar the solution of the dye with a weighed quantity of kaolin, pre- viously ground with a little water. Alternatively the dry kaolin was added to the dye solution, and the mixture thoroughly stirred, or in some cases the solution of the dye was merely filtered through B layer of kaolin.The dye solutions contained in all cases 1 grm. of the dyestuff per litre. For the purposes of the experiments, 10 C.C. of the dye solution were stirred for several minutes with 5 grms. of kaolin, previously mixed with 10 C.C. of water, and the mixture filtered through a Buchner funnel with the aid of a, pump.The above amount of kaolin was always employed, but in the case of those colours which are very rapidly adsorbed, it was found that the kaolin was capable of taking up in some cases more than 1 per cent. of its weight of the dye. When tested in this manner, it was found that the various dyes experi- mented with could be divided into three classes: (a) those which are entirely adsorbed under the above-mentioned conditions ; ( b ) those which are partially adsorbed, giving coloured kaolin from which the bulk of the dye can be extracted by washing with water ; and ( c ) those which are not appreciably adsorbed.The following table contains a list of the dyes experimented with, and shows their behwiour when treated in the manner above described : TABLE I.Adsorbed. Congo red. Safranine. Magenta. Neutral red, Malachite green. Brilliant green. Methylene blue. Bismarck brown. Crystal violet. Methyl violet B extra. Auramine. Partiall Adsorbed, but Adsorbed CIo%nw partly extracted by Washing with Water. Acid green. Patent blue. Soluble blue. Acid green. (Practically all extracted by boiling water.) Not Adsorbed. Acid magenta. Eosin.Erythrosin. Fluorescein. Methyl orange. Tr opaoli n. Orange IV. Ponceau 4R. Bordeaux R. Sudan red. Tartrazin. Naphthol yellow. Picric acid. Naphthol green. Indigo carmine. Cochineal (1 per cent. solution).TO THE DETECTION AND SEPARATION OF CERTAIN DYES 341 Although three entirely different lots of kaolin which were experimented with behaved very similarly so far as their powers of adsorption were concerned, yet slight differences were observed, particularly in the case of certain of the partially adsorbable dyes.A convenient method of ascertaining whether any particular sample of kaolin is suitable for the examination of dyes by this method, is to mix 10 C.C. of a 0.1 per cent. aqueous solution of naphthol yellow, and 5 C.C. of a 0.1 per cent. aqueous solution of soluble blue.I n the case of a really good kaolin, 5 grms. should be sufficient for the coqplete separation of these two colours-that is to say, the filtrate should be yellow and not tinged with blue. Satisfactory separations of the following dyes were effected : Methyl violet B extra and naphthol yellow, crystal violet and naphthol yellow, crystal violet and eosin, auramine and naphthol green, auramine and indigo carmine, auramine and acid magenta, magenta and tropaeolin, methylene blue and cochineal (1 per cent.), malachite green and tropaeolin, brilliant green and tropaeolin, methylene blue and tartrazin, crystal violet and methyl orange, soluble blue (5 C.C.only) and naphthol yellow. I n some cases it happens that, although a dye is itself readily soluble in alcohol, if is yet incapable of being dissolved from the coloured kaolin by means of that solvent.It is clear, therefore, that this property might in some instances be made use of in order to effect the approximate separation of three dyes when present together. Thus, the 5 grms. of kaolin, having been shaken or stirred with the dye solution in the manner indicated above, and filtered, the coloured kaolin remaining on the filter might, after one or two washings with hot water, be treated with hot alcohol.The following are examples of separations effected in this manner : TABLE 11. Ponceau 4 R : Aqueous filtrate. Red. Crystal violet : Alcoholic filtrate. Purple. Safranine : Kaolin. Pink. Naphthol yellow : Aqueous filtrate. Yellow.Magenta : Alcoholic filtrate. Red. Methylene blue : Kaolin. Purplish-blue. Fluorescein : Aqueous filtrate. Dirty pink and fluorescent. Brilliant green : Alcoholic filtrate. Green. Congo red : Kaolin. Pink. That the coloration of the kaolin in the above cases was, in fact, due to the safranine, methylene blue, and congo red respectively, was proved by the application of well- known tests.Experiments were next made for the purpose of ascertaining to what extent the adsorption results given above apply to saccharine solutions. I n the case of magenta, methyl violet, safranine, Bismarck brown, brilliant green, auramine, and soluble blue, the results were substantially the same as when sugar was absent. The above results apply entirely to the removal of dyes from aqueous soZutions by means of the kaolin, but in many cases the removal of the dye from alcoholic solutions can heTABLE 111.Kaolin Dyed by Adsorption of- Congo red. Safranine. Magenta. Neutral red. Methyl violet B extra. Crystal violet. Malac hit e green. Brilliant green. Meth ylene blue. Bismarck brown, or Vesuvin. Auramine. H,SO,. Deep blue. Green, changing to dull violet on dilution.Brownish- yellow. Green. Yellowish- brown. Brownish- yellow: Yellow. Orange. Yellowish- green, becoming blue on dilution. Puce. Fawn. H2S04, 10 per :ent. by weight, Slate blue. Purple. Brownish. Violet-blue. Fawn. Greenish. Fawn. Yellowish- brown. No change. Darker. Much lighter HCI. Deep blue. Greenish- blue. Yellow. Blue. Brownish- yellow. Orange- yellow. Bright yellow.Yellow. No change. Puce. Deoolorisod. HNO,. Bluish-sla te. Bluish-green Yellowish- brown. Bluish-slate. Greenish. Dull orange. Yellow -red. Yellowish- red. Bluish-green becoming blue on dilution. Puce. Decolorised. NaOH, 10 per cent. No change. 9 9 1, Yellow- brown. Lighter and duller. No change. Nearly decolorised. $ 9 No change. Browner. Decolorised, NH4OH. ~ No change.?9 Slowly becomes lighter. Yellow- brown. No change. $ 9 D ecolor ise d . 9 9 No change. Yellower. Lighter. Alcohol. Alcohol, tinted light pink. ilcohol, coloured light pink. Deep pink solution. Little change ; reddish- yellow filtrate. Violet solution. Greenish solu- tion; partly extracted. Deep green solu- tion; easily extrac tea. No action. Yellow on boiling. Yellow solution.TO THE DETECTION AND SEPARATION OF CERTAIN DYES 343 effected by the same method, although not in the same degree.Thus, working under the above conditions, safranine which is completely removed from water, was only partly adsorbed from a solution in 90 per cent, alcohol. The kaolin containing the adsorbed dyes can be tested by the application of reagents similar to those which have been employed for the detection of dyes on wool and other fibres.As we have verified these reactions in the case of the colours experimented with, it may be useful to give the table (p. 342), although the reactions are in many cases to be found in various textbooks dealing with the subject : The coloured kaolins were, in all cases, dried before the application of the above-mentioned reagents.The dilute sulphuric acid referred to was prepared by mixing one part by weight of the acid with nine parts of water ; the sulphuric acid, hydrochloric acid, and nitric acid were the ordinary pure laboratory reagents. The caustic soda solution was a 10 per cent. solution, and the ammonia was prepared by mixing one volume of the strong ammonia solution with two volumes of water. The alcohol referred to in the last column was 90 per cent.strength. I t will be seen that we have only dealt with a few of the more commonly occurring dyestuffs, but doubtless the method would prove applicable in the case of a great many other colouring matters with which we have not experimented. I t may be said that no difference could be observed in the adsorbing properties of kaolin when this was extracted with dilute hydrochloric acid and thoroughly washed. On the other hand, the ignition of the kaolin seemed to weaken its adsorbing properties to some extent.Precipitated aluminium hydroxide ground up with water (the so-called alumina, cream) was found to be very inferior to kaolin as an adsorbent. DISCUSSION. Mr. BEVAN asked whether the respective proportions in which the dyes were present would make any difference in the accuracy of the test, and also whether there appeared to be any connection between the chemical composition of the various dyes and their degree of adsorption.Mr. RAYMOND Ross asked whether kaolin had any action upon natural vegetable colouring matters, such as occurred in fruits and in jam.Mr. J. CONNAH asked whether this method had been applied to the identification of vegetable colours (e.g., fruit juices) as well as to the coal-tar colours, and remarked that the method would doubtless be useful for the detection of artificial colouring matters in wines. Mr. E. R. BOLTON asked whether the method could be used to separate colours from oils, and also whether the authors had tried the use of Fuller’s earth. He suggested that possibly the use of kaolin and Fuller’s earth consecutively might be found advantageous.Fuller’s earth, of course, varied a good deal in its action on oils, some samples being capable of bleaching oils that would not be affected by other samples. Mr. H. E. BURGESS said that he had recently found that when a solution con- taining three or four of these dyes was applied to ordinary blotting-paper the colours were separated, and could be fairly accurately identified by cutting up the blotting-344 THE DETECTION AND SEPARATION OF CERTAIN DYES paper and dissolving out the separate colours.The colours in that case were dis- solved in glycerol before being applied to the blotting-paper ; when water was used, the same effect was not obtained.Mr. EVERSHED said that those colours which were adsorbed by kaolin were all, or nearly all, basic colours, whereas those which were not adsorbed were “acid ” dyes. The action of the kaolin was therefore analogous to that of animal fibres. A skein of wool would take up the basic colours from a neutral bath, while the acid colours would only dye wool in the presence of a little mineral acid.He should like to ask whether the authors had tried the effect of mineral acids upon the adsorption of any of these dyes by kaolin. Had any quantitative tests been made with mixtures of acid and basic colours to see whether the former passed through the kaolin com- pletely? It was well known that if fairly strong solutions of acid and basic colours were mixed, a precipitate was generally formed. If this precipitation occurred with weak solutions in passing through the kaolin, the separation must be incomplete.Had the kaolin method any superiority over the well-known methods of testing by means of wool and cotton? Mr. L. M. NASR asked whether this method applied to spirit-soluble colours. Mr. RICHMOND said that a good many years ago a method was published for detecting artificial colours in butter and margarine by shaking up with ‘‘ Spanish earth,” which, while not affecting the natural colouring matters, removed a good many artificial colours.The PRESIDENT remarked that the fact that it was possible, when two out of a mixture of three dyes had been adsorbed by kaolin, to remove one of these by treat- ment with alcohol, showed how great an influence was exercised by the liquid in which the colours were dissolved.No doubt, if alcohol had been used in the first instance, different effects as regards adsorption would have been produced, and it might perhaps be possible, by varying the solvent, to extend the application of the method. Mr. CHAPMAN, in reply, said that the classification of the various dyes into those which were adsorbable by kaolin and those which were non-adsorbable was not, of course, an absolutely sharp one, but the method was at least capable of permitting of the separation of a number of dyes from one another, and of rendering their identifica- tion possible.It was true that, generally speaking, the dyes which were adsorbable by the kaolin were those which had a basic character, whilst those of an acid character remained, as a rule, unadsorbed.This had been pointed out some years ago by Suida and others. I t was, of course, also true that some of the dye stuffs, to which he had referred, interacted and produced some precipitation. One advantage of the kaolin method over the ordinary one of dyeing wool, and also over such capillary methods as those mentioned by Mr. Burgess, was that larger quantities of the dyes could be separated, and that the dyed kaolin could be tested by means of various reagents, which could not well be applied to wool or paper. He and Mr. Siebold had not devoted much attention to natural colouring matters, but a good deal of work had already been done on that part of the subject, and had been referred to at the begin- ning of the paper. They had had no experience of the application of this method to the testing of wine, but it seemed quite possible that it would enable certain artificialDETECTION AND ESTIMATION OF SMALL QUANTITIES OF NITROUS AClD 345 dyes to be separated from the natural colouring matters of the wine. I n the case of alcoholic solutions, separation c-ould in many cases be quite readily effected, although they were of a different character from those which had been described in the case of aqueous solutions. They had not made any experiments with oils, and could not say what results would be obtained.