2762 WERNER : THE ISOMERIC TRANSFORMATION OFC C t V I 1 I. - The Is0 n ~ e r ic T? -a 12 s f omnatioyb of Ammom’ 11 mMethyl Sulphate, and of Substituted Anhmo?iizcmMethyl Xulphates ; the Intewxction o f Amivzesaid Methyl Sdphate.By EMIL ALPHONSE WERNER.IN a recent communication (this vol., p. 925) containing an accountof the study of the’ decomposition by 4ieat of the methyl sulphatesof certain isocarbamides, it was pointed out that ammonium methylsulphate can evidently undergo isomeric change when heated, inaccordance with the equation :NH,*MeSO, = NH3Me*HS0,.This interesting readion, which does not appear t o have beenhitherto rccorded, was only noticed as a qualitative change inconnexion with tlie investigation referred t o above. A more inti-mate study of the change has now been made, the results of whichhave shown that an isomeric transformation on the line indicatedabove is apparently a property common to all substituted ammon-ium methyl sulphates, where such an interchange in the position ofa hydrogen atom and a methyl group is possible.Thus when methylammonium methyl sulphate, the simple pre-paration of which from ammonia and methyl sulphate is describedfurther on, is heated it gradually changes t o the isomeric dimethyl-ammoniuin hydrogen sulphate,NH,Me*MeSO, = NH,Me,*HSO,,and the further progress and limitl of the isomeric traiisformationmay be expressed by the general equations:NH2RR/*MeS0, = NHRR/Me*HSO,NHRR/R”*MeSO, = NRR/R//Me*HSO,.Whilst the alcohol radicles may be variants, tlie isoinerisationappears to be restricted to the methyl sulphates, since neitherammonium ethyl sulphate nor ammonium n-propyl sulphate wasfound to undergo any isomeric change.Decomposition with theformation of ammonium hydrogen sulphate and ethylene andpropylene respectiveIy was the only result of the action of heaton these two salts.The results of a quantitative study of the extent of isomericchange with rise of temperature in the case of ammonium andinethylammonium methyl sulphates are given in the tables below;the time of heating was fifteen minutes in all the experiments; thiscomparatively short period was chosen, as preliminary experimentsshowed that the velocity of the isomeric change was chiefly a funcAMMONIUM BIETHSL SULPHATE, ETC.2763tion of the temperat'ure. Since the methyl sulphates under investi-gation were found t o 'be neutral in reaction, the progress of thechange was readily measured by determining the acidity developedin the product ah the end of each experiment.TABLE I,*NH4*M(?S0, = NH,&lIe*€ISO,.I'empera-ture.110-120"135-145160180-185200-205Tempera-ture.100-105°135-140150- 15 517G-175190-200Per cent. ofisomeric change.7.09.411.314.325.8Tempera- Per cent. of220-230' 46.3240-250 7 3 4250-260 85.9275 97.2ture. isomeric change.TABLE II.*NH3Me*&CeS0, = NH,Me,-HS04.Per cent. ofisomeric chaaige.2.66.47.28.313.9Tempera- Per cent. of2 10- 4 20" 35.1230-240 60-3-2 50--260 81.3275 97.5ture. isomeric chmgr.* The nnmbers for the temperatures from 240" upwards are slightly liigher thanthe true values, on account of the sinall amount of secondary change, referred to inthe experimental part.It will be' seeii from the above that in both cases the amount ofisomeric change is small until a temperature of about 220° isreached; beyond this it proceecb rapidly, and is almost completea t 275O, and with the exception of a slightly greater amount ofisomerisation for the lower temperatures in the case of ammoniummethyl sulphate these is practically no difference in the generalorder of the change for these two salts.The influence of time onthe velocity of isomeric change is small even just below 220O; thusin the case of ammonium methyl sulphate after heating for onehour a t 200-205° the percentage of isomerisation was raised from25.8 (for fifte'en minutes) t o 34.2, whilst with the methylammoniuinsalt a t 210-220° there was an increase from 34.1 t o 46.9 per cent.for the same difference in period.A few different types of substituted ammonium methyl sulphateshave been examined, all of which apparently undergo an isomericchange, and in the1 case of derivatives containing more complexalkyl groups or a phenyl group this proceeds more readily thanwith the methyl sulphates recorded abo've ; for example, methyl-dipropylammonium and plienyldimethylammonium methyl sul2764 WERNER : TIIE ISOMERIC TRANSFORMATION OFphates were found t o undergo isomeric change to the extent ofabout 35 and 48 per cent.respectively a t 140°, wliilst even a t looothere was a very appreciable amount of change.It is hoped to make a systematic quantitative study of a numberof different methyl sulphates as soon as opportunity permits.As regards the mechanism of this isomeric change it appearsvery probable that dissociation, as the first step towards morestable equilibrium, precedes the formation of the isomeride ; thuswhen ammonium methyl sulphate is heated it will dissociate withthe production of methylamine and sulphuric acid rather thanammonia and the very unstable methyl hydrogen sulphate, andfrom the union of the disso_ciation products the still more stablemethylammonium hydrogen sulphate will be formed. This explana-tion is in agreement with the conditions under which the isomericchange has been shown to take place, and also with the fact thatthe methyl sulphates of the feeble (benzenoid) ammonium basesundergo isomerisat.ion readily a t correspondingly lower tempera-tures. Considered from a practical point of view, the recognitionof this general isomeric change is likely to prove of some value,since it appeazs to place in our hands a method more simple thanany of those hitherto available for the preparation of varioussubstituted methylammonium bases, more particularly on accountof the ease with which the methyl sulphates may be prepared fromthe interaction of the amines and methyl sulphate, in accordancewith the general equations :(a) NH,R + Me$O, = NH,RMe*MeSO,.( b ) NHR, + Me2S0, =NHR2Me*MeS0,.It also explains certain discrepancies in the results which have beeiipublished hitherto with regard to the above reactions.Thus, whilstClaesson and Lundvaal ( B e y . , 1880, 13, 1699) have shown thatammonia and aniline respectively react with methyl sulphateaccording to equation (a), they state that in the case of diethyl-amine and methyl sulphate the products formed are diethyl-ammonium and dimethyldiethylainmonium methyl sulphates, andmore recently Ullmann (9nnaZen, 1903, 327, 104), who hasexamined the behaviour of aniline and a number of its differenthomologues towards methyl sulphate, has arrived at the conclusionthat in the case of aromatic amines the interact.ion does not followthe course described by Claesson and Lundvaal (Zoc. cit.), but forprimary amines is to be represented by the general equation:2NH2R + Me,SO, = NH,R*MeHSO, + NHMeR ;for example, with aniline the products formed are stated to beaniline methyl sulphate, methylaniline, and a certain amount ofdimethylaniline.The experimental conditions adopted, more eepeciAMMONIUM METHVL SULPHATE, ETC. 2 7-6 5ally by Ullniann, were such that much heat was allowed to de,velopduring the interactions, with the result that more or less isomericchange must have taken place, to which cause must be attributedthe different conclusions arrived a t regarding the general order ofthe interaction. The results of experiments with several differentamines, aliphatic and aromatic, have shown that if care be takento avoid anything more than a slight development of heat, thereactions with methyl sulphate proceed in a perfectly straight-forward manner in accordance with equations ( a ) and ( b ) ; practi-cally quantitative yields oE the substituted ammonium methylsulpliates have been obtained without any trouble.EXPERIMENTAL.Preprution of, and Action of Heat on, Anlmonium MethylSulphate.The following method of preparation was found t o give a fairlygood yield of the above salt in a high degree of purity.Theproduct obtained after heating a mixture of 50 grams of puremethyl alcohol and 100 grams of pure sulphuric acid on the water-bath for half-an-hour was cooled and directly neutralised byaddition of powdered commercial ammonium carbonate until, withthe occasional addition of a small quantity of water, a pasty,faintly alkaline mass was obtained; this was extracted with about150 C.C.of boiling methyl alcohol, the cold solution was poured offfrom any ammonium sulphate which separated, and concentratedt o about half the volume by careful distillation. While hot thesolution was again poured off from any substance which hadsep:wated, and on cooling it set to a mass of thin, plate-like crystalswith a satiny lustre. After a further recrystallisation from absolutemethyl alcohol the salt was obtained quite free from even a traceof sulphate, and was almost neutral in reaction; 32 grams of pureammonium methyl sulphate were obtained from 50 grams of methylalcohol.The pure salt melts* a t 1 3 7 O , and is exceedingly hygroscopic; i tis less soluble in ethyl alcohol than in methyl alcohol, hence thepreference of the latter solvent for its purification.In order to study the progress of the isomeric change, a series oftest-tubes containing weighed quantities of the well-dried salt wereheated in a bath of glycerol for fifteen minutes to the desired tem-perature; the bulb of a thermometer, passed through a loosely-fitting cork, was kept immersed in the fused salt during each* When slowly heated the salt may be found to melt at 125', as 9 result of acertain amount of isomeric change.VOL.cv. 8 2766 WERNER : THE ISOMERIC TRANSFORMATION OFexperiment. The product was dissolved in water and titrateddirectly with I?/’-sodium hydroxide, methyl-orangel being used asindicator. On account of the very hygroscopic nature) of the salti t was not advisable to attempt to weigh out a similar quantity foreach experiment; from the equation :NH,*MeSO, = NH,Me*HSO, (M.W.= 129),i t will be readily seen that 1-29 grams of ammonium niethylsulphate would require, after complete isomerisation, 10 C.C. ofW-sodium hydroxide for neutralisation, hence the values for amolecular proportion were calculated from the titration resultsobtained ill the different experiments, and as these numbersexpressed in terrns of percentage of isomeric change are givenunder table I, the full details would be of no particular interest,an1 have therefore been omitted.When ammonium methyl sulphate was heated to about 240°and upwards, a slight evolution of gas commenced after fiveminutes’ heating, and continued very slowly to the end of theexperiment; this was found to be ethylene, the result of asecondary decomposition, namely,2NH4-MeS0, = C,H, + 2NH,HSO,,which to a very slight extent accompanies the main isomeric change.From 20 grams of ammonium methyl sulphatel, after heating underthe most favourable conditions, 8.9 grams of methylamine hydro-chloride were ultimately obtained, corresponding with 85.6 percent.of the theoretical yield for complete isomerisation ; hence thesecondary decomposition referred to above has no serious influenceon the main change.Preparation of, and Action of Heat on, MethylammoniumMethyl Sulpha t e.This salt was readily prepared, in quantitative yield, by passinga current of dry ammonia ‘into a solution of pure methyl sulphatein about fifteen times its volume of benzene until the product, afterkeeping f o r one hour, had a slight odour of ammonia, the flaskcontaining the solution being immersed in ice-cold water.The saltgradually separated in micro-crystalline form ; it was quite colour-less, neutral in reaction, and gave an absolutely negative resultwhen tested with barium chloride solution. The dry salt melts a t5Z0, and is very hygroscopic; its purity was confirmed by analysis,and hence the reaction is correctly represented by the equation:NH, + Me2S0, = NH,Me-MeSO, ;the formation of some ammonium methyl sulphate, recorded bAMMONIUM METHYL SULPHATE, ETC.2767Claessoii nntl T,unclvaal (Lor. ciL.), was probably due to the presemeof some methyl liytlrogen siilphate in the ester nsed by tliein.The esaniiiiation of the progress of the isomeric cliange whenmethylammonium methyl sulphate is heated was carried out exactlyas in the case of the amnioiiiuir. salt; the results so obtained aregiven under table 11.liiferirctioit of -4 nziiies nnd dfutlhyl SiJphnie.The ester used in these experiments was purified from any acidiiiipurities by shaking it with a solution of sodium hydroxide untilneutral; the separated ester was then dried over anhydrous sodiumsulphste, and used directly bithout redistillation.E'rpt. T. A d i n e niid Jfethyl Sulphate (equal molecular propor-tions).-To a solution of 9.3 grams of freshly distilled aniline in75 C.C.of pure benzene, 12.6 grams of methyl sulphate dissolved in25 C.C. of benzene were added; there was very slight developmentof heat, and after a short time, minute, thin, glistening plates begant o separate. After twenty-four hours these were collected, washedwith benzene, and dried over sulphuric acid and afterwards for ashort time in a vacuum. The weight obtained was 21.1 grams,whilst theory requires 21.9 grams if the reaction proceeded inaccordance with the equation :C,H,*NH, + Me,SO, = C,H,*NH2Me*CH3S04.An aqueous solution of the product gave no precipitate withbarium chloride solutlion, and on the addition of bleaching-powdersolution and a few drops of dilute sulphuric szid, a deep indigo-bluecolour was developed,* a reaction characteristic of methylaniline(Found, SO, =44*02.C8H,,04NS requires SO, = 43.83 per cent.).Phenylmethylammonium methyl sulphate melts at 159O, andundergoes isomeric change very readily ; this was strikingly demon-strated in the benzene filtrate from the preparation, which retaineda small quantity of the salt in solution, in the following manner: aportion of the benzene solution was evaporated to dryness on thewater-bath, and the residue gave a copious precipitate with bariumchloride solution, and, on addition of bleaching-powder solution, adeep orange-red colour was a t once developed, a reaction character-istic of dimethylaniline. The original benzene solution when shakenwith dist?ilied water gave no reaction with barium chloride, butwhen previously heated for a few minutes on the water-bath andagain shaken with water, on applying the test reagent a precipitate* It i q generally stated that methylaiiiline gives no colour with bleaching-powder solution, but this is only trne of the free base ; on acidification an intenseindigo-hlue colour is quickly developed.8 s 2768 WERNER : THE ISOMERIC TRSNSFORMATION OFof harimti snlplinte was imine~lintely fornierl as R, resalt; of theisomeric change.The experiment was repeated, using an excess of aniline, andthe weight of crystalline precipitate obtained was practically thesame as before, the excess of aniline being found unchanged in thebenzene filtrate, thus proving that the interaction takes placestrictly in accordance with the equation given above, and that thefailure t o recognise the isomerisation which takes place so readilywas responsible €or t4he erroneous explanation of the change givenby Ullmann (Zoc.cit.).E x p t . ZI. Methylaniline a d Methyl Sulphate (equal molecularproportions).-To 10.7 grams of pure methylaniline dissolved in75 C.C. of benzenel, 12.6 grams of methyl sulphate dissolved in twiceits volume of benzene were added; a clear, pale yellow oil gradu-ally separated, witlhout any appreciable evolution of heat, andafter twenty-four hours this was collected. The weight obtainedwas 32 grams ; the theoretical yield of plienyldimethylammoniummethyl sulphate formd in accordance with the equation :C,H,-NHMe + Me2S0, = C,H,*NHMe2-MeS0,,would be 23.3 grams.The oil when added to water rapidly dis-solved with the separation of some benzene, and any attempt toexpel the latter from the oil by heat was accompanied by evidenceof isomeric change having taken place during the process. Byleaving the oil for a week in a vacuum over sulphuric acid aproduct quite free from benzene was finally obtained. The viscousresidue, which showed no signs of crystallisation, was not analysed,but gave all the reactions of a salt of dimethylaniline, and no otherproduct could be detected in the original benzene separated fromthe oil, the yield of which corresponded very closely with a com-pound of the salt with a molecular proportion of benzene whichwould require 31.1 grams.This property of forming a feeble com-pound with benzene has already been noticed by Claesson andLundvaal (Zoc. cit.) in the case of ethylaniline ethyl sulphate.Expt. IZZ. Dipropylamiue und Methyl Sulphate (equal molecularproportions).-This experiment was carried out as in the previouscase, with 10.1 grams of dipropylamine and 12.6 grams of theester, a clear, almost colourless oil quickly separating with Blightevolution of heat. The yield of oil was 30.8 grams, whilst the theo-retical yield of m&hyldipropylammonium methyl sulphate wouldbe 22.7 grams, and for a compound with one molecular proportionof benzene 30.5 grams. I n this case also an attempt to expelbenzene from the oil by heat was accompanied by isomeric change.The product, freed from benzene by the means mentioned underExpt.11, was a viscous liquid, the aqueous solution of which gavAMMONIUM METHY [, SULPHATE, ETC. 2769no precipitate with barium chloride solution, and on distillationwith pot'assium hydroxide gave methyldipropylamine, which wasidentified by the analysis of its hydrochloride (Found, GI = 23.44.Calc., C1= 23.43 per cent.).Experiments I1 and I11 were repeated, using an excess of therespective amine; in each case this excess was found unchangedin the benzene separated from the precipitated oil, proving thatwith secondary amiiles also the reaction with methyl sulphate isbetween equal molecular proportions and in accordance with thegeneral equations already given.Note 01% ,4mmonium Ethyl Sulphnte and Ammoniumn-Propyl Sulphate.Pure ammonium et'hyl sulphate was readily prepared in verygood yield by directly neutralising crude ethyl sulphuric acid(prepared in the usual manner) with commercial ammonium car-bonate, absolute alcohol being used for the extraction and recrystal-lisation of the ammonium salt.The yield was 58 gram& of thepure compound from 50 grams of ethyl alcohol, no attempt beingmade to recover a further quantity from the mother liquor. Thepure salt crystallises from alcohol in thin, flat, rhomboidal prisms,which have been easily obtained in a length of 4-5 cm., andpossessing a brillia'nt, satin-like lustre.Ammonium ethyl sulphate melts a t 9 7 O (not 6Z0, as stated invarious works of reference), and is not notably hygroscopic, whichis also contrary to the published statements.When heated the salt decomposed rapidly a t about 220° withthe evolution of ethylene; after fifteen minutes a t this temperatarethe amount of decomposition was 55 per cent., in accordance withthe equation :NH4*EtS04= C2H4 + NH4aHS04.No ethylamine could be detected in the residue, thus proving theabsence of ar,y isomeric change.Ammonium iz-propyl sulphate (m. p. 132O) was obtained in gootlyield from crude n-propylsulphuric acid in a manner similar t othat stated above. The pure salt, which is very hygroscopic, decom-posed rapidly a t 150-160° with the evolution of propylene andwithout any evidence of isomeric change.UXIVERSITY CHEMICAL LABORATORY,TRINITY COLLEGR, DUBLIN