1906 ATRINSOX’ ASD THORPE: THE FORJIATIOS ASD By ERKEST FRAXCIS JOSEPH ATKINSOX and JOCELYN FIELD THORPE. The Condensation of Benxyl Cyanide with the Soc2izcm Derivcctive OJ Ethgl Cyanoiccetate. IN Part I. of this series it was pointed out by Baron, Remfi-y, and Thorpe (Trans., 1904, 85, 1726) that ethyl cyanoscetate readily reacted with its sodium derivative, forming ethyl p-imino-a-cyano- glutarate, and that tlhe reaction evidently proceeded according to the equation * Co2Et*~HNn + HCH(CN)-C0,Et -+ CN CO,Et*yHNa C( :NH)*CH(CN)*CO,Et* It was also pointed out that, on treatment with concentrated sulphuric acid, ethyl P-imino-a-cynnoglutarate (1) is quantitatively converted into ethyl glutazinecarboxylate (2). C:NH C:NH C:NH JYH bEt I OH (1.) (2.1 (3.) The tendency for the straight chain compound to pass into the if Subsequent experimeiits have shown that this view of the course of the conden- sation, which was based on the position of the sodium atom in the condensation product, is not quite correct, and that it is the reactive hydrogen atom of the sodium derivative which forms the imino-group.The course of the action is therefore best represented by the equation : CO,Et*CH CO,E t*CH, 1 t- HCNn(CN)CO,Et -+ I CN C(:NH)*CN~(CN)CO,E~‘ The change in the position of the sodium atom must be attribnted t o the mom strongly acidic character of the hydrogen of the methyleiie ~ T O I I ~ .11EACTIOSS OF IXISO-COMl'OUh'DS. PART 11. 1907 pyridine ring, which is indicated by this reaction, was still more strongly shown by the behaviour of ethyl hydrogen P-imino-a-cyano- glutarate (3), which was found to undergo conversion into ethyl gliitazinecarboxylate (2) on being heated a few degrees above i t s melting point (Zoc.cit., 1746). There seemed to be, therefore, a con- siderable tendency for a nitrile group in the y-position to a carboxyl group in n compound such as ethyl hydrogen p-imino-a-cyanoglutarate to pass, by intermolecular change with that group, into a six- membered ring, and that this tendency is also shown by the cor- responding p-ketonic derivatives mas demonstrated by the conversion of ethyl hydrogen cyanoacetonedicarboxylate (4) into ethyl 2 : 4 : 6- trioxypyridine-3-cnrboxylate (5), on heating a few degrees above its melting point. CO co CH,/\CH*CO,Et c Whilst desiring to continue the investigation of the conditions of formation of these irnino-compounds, we wished at the same time to ascertain whether the tendency of the nitrile group to enter into ring formation inigbt not be applied to the formation of carbon rings as well as those containing nitrogen, and with this object in view and for other reasons which will be given later, we have investigated the condensation of benzyl cyanide with the sodium com- pound of ethyl cyanoacetate and also with its own sodium derivative.It is apparent that the most probable course of the first reaction mould be represented by the equation (I), and that the product on. acidifying would be ethyl p-imino- a-cyano-y-phenyl-n-butyrate (6). (I) Ph'gt2 + HCNa(CN)*CO,Et -+ Ph*$!H, -+ C:(NH)C]H(CN)S P1i*$!H2 C0,Et C (NH) C Na( CN) CO,E t But it is also possible that the sodium derivative of ethyl cyano- acetate might react with benzyl cyanide, forming the sodium derivative of the latter, in which case the course of the condensation would be represented by equation (11), and on acidifying ethyl p-imino- yy-cyanophenyl-n-butyrate (7) would be the chief product.1908 ATKINSON AND THORPE: THE FORMATION AND CO,Et*$lH, CO,Et*$!H, C: ( N H)* CH( CN)Ph -+ C: (N Hj*CNa( CN)Ph (7.) Under conditions given in the experimental portion of this paper, a good yield of a well-defined crystalline compound (m.p. 125O) is formed on condensing benzyl cyanide with ethyl sodiocyanoacetate, and there can be no doubt that its constitution is represented by formula (6), although, under certain experimental conditions, there is evidence t h a t a compound of formula (7) also occurs as an intermediate product.I n Part I. of this series (p. 1728), it was shown that P-imino-ethyl salts are decomposed on treatment with alkaline hydrolytic agents and that the products formed, which are the same as those derived by the alkaline hydrolysis of the corresponding P-keto-ethyl salts, often afford valuable evidence as to the formula of tho compound hydro- lysed. It is apparent, however, that this method fails when applied t o two compounds of formulze (6) and (7), since these two substances would yield the same products, namely, a mixture of malonic and phenylacetic acids, on complete hydrolysis, thus : CH2Ph*C: (NH). CH(CN)CO,Et CO,Et*CH,* C:(NH) CH( CN)Ph (6.1 (7.) CH,Ph*CO,H CH,(CO,H), CH,(CO,H), CH,Ph*CO,H.The compound melting at 125" is readily hydrolysed by means of an alcoholic solution of potash, and the product consists of about equal proportions of malonic and phenylacetic acids. Although by this means, therefore, i t is not possible to distinguish between the two formuh ( 6 ) and (7), yet the evidence is valuable, since, taken i n conjunction with the boiling point of the compound (235'/20 mm.) and its percentage composition, it shows t h a t it has been formed by the condensation of one molecule of benzyl cyanide with one molecule of ethyl cysnoacetate. It was necessary, therefore, to adopt some other means of dis- tinguishing between the two compounds of formula: (6) and (7), and since me were unable t o isolate any inkermediate products during the alkaline hydrolysis, the formation of which might throw some light on the problem, we were led to study the action of concentrated sulphuric acid on the cornpound melting a t 125'.If the two formuh are examined, it will be seen- that the action of this reagent must give rise t o different compounds.REACTIONS OF 3MINO-COMPOUNDS. PARr 11. 1909 A substance of formula (7) is closely allied to ethyl p-imino-a- qanoglutarate (see page 1906) in constitution, and therefore should readily give phenylglutazine (8) on treatment with concentrated sulphuric acid, thus : C:NH CXH /\ A YH2 YHPh -+ VHP YHPh co CN co co ' I OEt (7.) \/ NH (8.) whereas a compound of formula (6), in which the nitrile group and the carboxyl group are united with the same carbon atom, cannot yield a derivative of glutazine under the same conditions.The compound melting at 125' does not give phenylglutazine on treatment with concentrated sulphuric acid, and there can therefore be no doubt that it possesses a constitution represented by formula (6), a view which is confirmed by a close study of the compound which is actually formed with this reagent. It has already been mentioned that one of our objects in investigat- ing t h e interaction of benzyl cyanide and ethyl sodiocyanoacetate was to prepare a compound in which the nitrile group could by reason of its position pass into a six-membered carbon ring instead of the pyridine ring formed from the other compounds which have already been investigated.This object will be more clearly understood when the formula of ethyl p-imino-a-cyano-y-phenylbutyrate is written thus (9) : CH, CHo (9.) (10.1 (11.1 It was thought possible that the nitrile group might react with the ortho-hydrogen atom of the benzene ring, forming a di-imino-compound (lo), which in its amino-form would be ethyl 1-3-diaminonaphthalene- 2-carboxylate (ll), and that by this means an interesting synthesis of the naphthalene ring might be effected. Since the original imino-compound (9) was formed in the presence of sodium ethoxide, we intended to attempt to obtain the di-imino- derivative (1 0) by treating ethyl P-imino-a-cyano- y-phenylbutyrate with excess of this reagent. We mere therefore considerably astonished to find that, on treating this ethyl salt with concentrated1910 A'l'KIXSOK' AND THORTE: THE FORMATIOX ASD sulphuric acid, in the attempt already mentioned to determine its constitution, it was quantitatively converted into the sulphate of ethyl 1 : 3-diaminonaphthalene-2-cnrboxylate (1 1).This result was all the more surprising because it has been shown by Metzner (Annalen, 1901, 298, 3S6) that a similarly constituted com- pound, namely, ethyl phenylacetylmalonate (1 a), on treatment with cold, concentrated eulphuric acid, is converted with loss of alcohol into ethyl 1 : 3-dihydroxynaphthalene-2-carboxylate (13), the reaction taking place a t the ordinary temperature, CH, (13.) and i t mas therefore to be expected that ethyl P-imino-a-cyano-y- phenylbutyrate (9) would behave in the same w ~ y , yielding 3-amino-2- cyano-1 -naphthol (14) : CH, CH, ., .I " co OH (14.) The yield, however, of ethyl 1 : 3-diaminonaphtbalene-2-carboxylate (1 1) from ethyl P-imino-a-cyano-y-phenylbutyrate (9), on treatment with sulphuric acid, is quantitative, and the reaction takes place with remarkable ease. Thus it is only necessary to add n quantity of well- ground ethyl salt t o three times its weight of concentrated acid and allow the mixture t o stand for one minute in order to complete the change.On first adding the ethyl salt it dissolves, forming a yellow solution which rapidly changes in colour in the course of a few seconds to deep malachite-green, heat a t the same time being generated.REACTIONS OF ININO-COMPOUNDS. PART 11. 1911 When this green solution is poured on ice the colour is discharged and a clear yellow solution is formed from which the sulphate of ethyl 1 : 3-diaminonaphthalene-2-carboxylate separates on standing as a pale yellow, crystalline precipitate.The sulphate is readily soluble in water, and if the aqueous solution is rendered alkaline the base is precipitated as an intense yellow crystalline compound, which, when purified by recrystallisation from alcohol, is obtained in deep yellow plates. There can be no doubt that this yellow compound represents the di-imino-form of ethyl 1 : 3-diarninonaphthalene-2-carboxyl;~te (10). The salts, especially the hydrochloride, can be obtained in almost colourless crystals, and they are therefore probably the salts of the amino-form (11). It is interesting to note, however, that although the solutions of these salts are colourless when cold, they become intensely coloured on boiiing, the colour being again discharged on cooling, The green colour which is formed when ethyl p-imino-a-cyano- y-phenylbutyrate is treated with concentrated sulphuric acid is difficult to explain, since ethyl 1 : 3-diaminonaphthalene-2-carboxylate dis*olves i n the concentrated acid, forming a yellow solution which does not change on warming. That this coloration is not necessary for the production of the naphthalene ring is shown by the fact that i f the solution of ethyl P-imino-a-cyano-y-phenylbutyrate in concentrated sulphuric acid is brought about a t Oo, and the temperature is not allowed t o rise above this point, no green colour appears, the solution remaining a light brown.Whether the green colour is formed or not, the yield of the naphtlialeue is not apparently affected. The corresponding carboxylic acid (15) can be readily prepared from ethyl 1 : 3-diaminonaphthalene-2-carboxylate (I 1) on treating it with a n alc+oholic solution of potash. The free acid, which is colourless, is unstable a t temperatures above its melting point (85"), being then transformed with evolution of carbon dioxide into 1 : S-naphthylene- diamine (16). 1 : 3-Naphthylenediamine prepared in this way is identical with the compound obtained by Urban(Ber., 1887, 20, 973) by the reduction of 1 : 3-dinitronaphthalene, and by Friedlander (Ber., 1895, 28, 1953) by the action of ammonia on 4-amino-2-naphthol. It was pointed out in Part I.of this series that ethyl p-imino- a-cyanoglutnrate (1 7) reacted with sodium ethoxide, forming a sodium1912 ATKINSON AND THORPE: THE FORMATION AND derivative, and that tlie metal entered in the po'ition (l), that after the sodium atom in this had been replaced by a n alkyl radicle, the hydrogen marked (2) could be replaced by sodium, and that finally, after this sodium had been treated with a n alkyl iodide, the third hydrogen atom marked (3) could be replaced. CO,Et* CHH*C:(NH)*CH(CN)*CO,Et (17.) (1) (4 (2 ) Ph.CHH*C(:NH)*CH(CN)*C02Et (18.) It is interesting to note that, although the hydrogen atom marked (2) is attached to the same carbon atom as a nitrile- and a carbethoxyl- group, yet the first hydrogen atom replaced is that marked (l), which is attached to the same carbon atom as the carbethoxyl- and the carb- imino-group.I n the case of ethyl p-imino-a-cyano- 7-phenylbutyrate (1 S), the influence of the phenyl group is at once apparent, and we have only succeeded in introducing one alkyl group into this compound after repeatedly treating it with sodium ethoxide and the nlkyl iodide. The hydrogen atom replaced is that marked (*), and apparently the hydrogen atom of the cyanoacetic residue is not replaceable under ordinary conditions. We have carefully studied both the methyl and ethyl derivatives of ethyl P-imino-a-cyano-y-phenylbutyrate, and have found that it is necessary to treat the ethyl salt four successive times with the calculated quantity of sodium ethoxide and excess of methyl (or ethyl) iodide in order to convert the whole of i t into its mono-alkyl derivative.The positions of the alkyl groups in the alkylated compounds can be readily shown by the products they give on complete hydrolysis with alcoholic potash. Thus the ethyl salt which is prepared by the action of sodium ethoxide and methyl iodide on ethyl P-imino-a-cyano- 7-phenylbutyrate gives phenylscetic and malonic acids under these uonditiony, thus : Ph*CHR.l[e* C: (NH). CH (CN)*CO, E t +% \ 4 CH,: (CO,H), i' Ph*UHILIe CO,H It is therefore ethyl P-imino-a-cyano-y-phenyl-n-valerate, Ph*CHMe-C:(NH)*CH(CN)*CO,Et. In the same way, the corresponding ethyl derivative gives a mixture of a-phenyl-n-butyric and malonic acids on complete alkaline hydrolysis. Its formula is therefore Ph*CHEt*C(:NH)*CH(CN)*CO,Et.REACTIONS OF IMINO-COMPOUNDS.PART 11. 1913 Both these ethyl salts behave in the same way as ethyl /3-imino- a-cyano-y phenylbutyrate on treatment with concentrated sulphuric acid. I n each case a similar green to blue coloration is produced if the solution is allowed t o become warm, but is not formed if the temperature is kept below O", and in each case, also, on pouring the mixture on ice a clear solution is obtained from which the sulphats of the base separates on standing. Ethyl 2 : 4-diamino-1-methyl- naphthalene-3-carboxylate (1 9) and ethyl 2 : 4-diamino-1-ethyl- naphthalene-3-carboxylate (22) are in each case precipitated from the solutions of their sulphates on the addition of ammonia as yellow oils which on extraction by means of ether are obtained as syrups rapidly solidifying to yellow, crystalline solids.Me Me 31 e Et (21.1 (22.1 Ethj.1 2 : 4-diamino-1-methylnaphthalene-3-carboxylate passes on hydrolysis with alcoholic potash into the corresponding carboxylic acid (20), which, like the unalkylated derivative, is a colourless, crystalline compound, decomposing with evolution of carbon dioxide on being heated above its melting point and passing into 2 : 4-diamino- 1-methylnaphthalene (21). Et Et (23.1 (24.) I n the same way, ethyl 2 : 4-diamino-1-ethylrraphthalene-3-carboxylate (22) yields 2 : 4-diamino-1-ethylnaphthalene-3-carboxylic acid ( 2 3 ) and 2 : 4-diamino-1-ethylnaphthalene (24). I%e Condensation of Benxyl Cyanide wit?^ its Sodizcm Devivative. The interaction of benzyl cyanide with its sodium derivative was first studied by Wache (J.pr. Chem., 1889, ii, 39, 251), who investi-gated the action of finely divided sodium on an ethereal solution of the nitrile. Although this chemist was unable to isolate the bimolecular nitrile, CH,Ph*C(:NH)*CH(CN)Ph, he found that, on heating the precipitate formed in the above reaction with excess of benzyl cyanide i n a sealed tube a t 180' for several hours, i t was transformed into a white, crystalline, basic substunce of the empirical formula C,,H,,N,, t o which he gave the name of cyanbenzyline (G-amino-5-phenyl-2 : 4- N---C(NH,)' ' dibenzylpyrimidine), and the formula CH,Ph*C< on account of the similarity between its mode of formation and properties and those of cyanethine and cyanmethine, which E.v. Meyer (J. pr. C~WH,., lS89, ii, 39, 156) had previously shown to be derivatives of pyrimidine. Wache subsequently found that the best way t o prepare cyanbenzyline was t o heat a mixture of the nitrile and dry sodium ethoxide in a sealed tube a t 170-180° for many hours. E. v. Meyer (J. pr. Chenz., 1895, ii, 52, 114), in conjunction with 0. Probst (Inaug. Diss. Leipzig, 1894), repeated the work of Wache, and found that if the precipitate which was formed by the action of finely divided sodium on a n ethereal solution of benzyl cyanide was washed with light petroleum and then added to dilute acetic acid, a viscid oil was obtained which could not be distilled without under- going decomposition, but which evidently consisted of @-imino-a-cyano- ay-ciiphenylpropane, CH,€'h*C( :NH)*CH(CN)Ph, since on treatment with hydroxylaruine i t gave the same oxime as that derived from cyanodibenzyl ketone, CI-I,Ph*CO*CH( CN)Ph.Later, v. Walther and Schickler (J. pr. Chem., 1897, ii, 55, 350) prepared the same compound by the action of ammonia on the cyanoketone, CH,Ph-CO-CH(CN)Ph, but also failed to obtain i t in the solid condition. The termolecular nitrile cyanbenzyline was thoroughly investigated by Herfeldt ( J . pr. Chem., 1896, ii, 53, 246), who prepared many of its derivatives, and still more recently v. Walthor (J. p. Chem., 1903, ii, 87, 447) has found that i t can be prepared by the action of finely divided sodium on an ethereal solution of berizyl cyanide containing dimethylaniline. Owing to the similarity in constitution between benzyl cyanide and ethyl cyanoacetate, we decided, in order to prepare P-imino-a-cyano-ay-diphenyl propane, CH,Ph* C( :NH)*CH(CN)Ph, to in- vestigate the interaction of benzyl cyanide with its sodium derivative in alcoholic solution, since in the case of ethyl cyanoacetate a good yield of ethyl /3-iminocyanoglut arate, CO, E t C H 2- C( N H) - CH( CN )CO,E t, had been obtained by Baron, Remfry, and Thorpe by this means (Trans., 1904, 85, 1726).We found that the products formed in this condensation varied with the leogth of time during which the heating was continued a t lobo. Thus, when two molecular equivalents of benzyl cyanide and one of N:C'(C H,Ph)>(jphREACTIONS OF IJIINO-CORLPOU,I'DS. P-iRT [I. 1915 sodium ethoxide dissolved in alcohol mere heated on the water- bath for half a n hour, the product consisted of about 70 per ceilt.of p-imino- a-cyano-ay-diphenylpropane, CH2P:i C( :NH) C H(CN)Ph, mixed with unchanged benzyl cyanide. When the heating was prolonged, an odour resembling amnionia became apparent at the mouth of the condenser and O K ~ working up the product at the end of two hours i t was found t o consist of ahoat equal proportions of P-iinino-u-cyano- ay-diphenylpropane, CH2Ph*C( :NH)*C H( CN)Ph, P-keto-a-cymo- ay-diphenylpropane, CH2Ph.CO*CH(CN)Pli, aucl cyanbenzyline, some benzyl cyanide being at the same time recovered unchanged. On prolonging the heating uiitil the oclour reseinbling ammonia ceased t o be perceptible, an operation which usually required twenty- four hours to accoiiiplish, the product was found to consist entirely of P-keto-a-cyano-ay-diplienylpropaue and cyanbeuzy h e .Subsequently i t was found by experiment that the odour resembling ammonia was riot due to this substance itself, but to ethylamine, which was probably foriltecl by the action of the alkaline alcoholic solution on P-imino-a-cyano-uy-dipheny 1 propme, the react ion eviden tl y proceedi ri g uccording to equation (111) : (111). CH,rh*C(:NH)'CH(CIU)Ph + ILOEt -+ CH,PLi*C!O.CH(ON)Ph + EtNW,. It is apparent, therefore, that the first action of benzyl cyanide ou its sodium derivative in alcoholic solution a t 100" is to form p-imino- a-cpno-ay-diphenylpropane according to equation (LV) : (IV). Ph*CH:Na.CN + H*CH(C")Ph -+ P h C HNa C ( : N H) C H ((2 N) P 11, and that if the reaction is stopped at the end of about half an hour n good yield of this compound can be obtained.After this time, the initial product is slowly clecornposed in accordance R. ith equation (111) and P-keto-a-cyano-ay-diphenylpropane is formed ; at the same time, the sodium derivative of P-imino-a-cyano-ay-tliphenylpropane reacts with the unchanged benzyl r.j.:xnide, forming cjanbenzyline in accord- ance with the observation of Watche. Like all previous investigators, we have beeu unable to obtain /3-imino-a-cyano-ay-diphenylpropane (33) in n crystalline condition, although we fiud that it can be distilled without decomposition under a presswe of 20 mm. if the operation is conducted rapidly and only small quantities are used. When treated with cold concentrated sulphuric acid, it behaves in just the same may as ethyl P-imino-a-cy ano-y-phenylbutyrate, being converted into 1 : 3-dinmino-2-phenylnaphthnlene (24) thus : VOL.LXXXIX. 6 li1916 ATKINSON AND THORPE: THE FORMATION AND (23) (24) It is worthy of note that, whereas ethyl 1 : 3-diaminonaphthaIene-2- carboxylate is intense yellow in colour, 1 : 3-diamino-2-phenylnaphthalene is colourless, and a comparison of the three following formuke is of interest as showing the influence of the carbethoxyl, carboxyl, and phenyl groups respectively on the stable forms of similar compounds : CHn Y (Yellow). (Colourless). (Colourless). It is also of interest to note that the same phenomenon occurs in the corresponding hydroxy-compounds, thus : Metzner (Annalen, 1901, 298, 386) finds that ethyl 1 : 3-dihydroxynaphthalene-2-carboxylate ( 2 5 ) is yellow, whereas the corresponding acid (26) is colourless.1 : 3-Di- bydroxy-2-phenylnaphthalene (27), which has been prepared by Volhard (Annulen, 1900, 296, 16) by the action of concentrated sulphuric acid on ethyl diphenylacetoacetate, CH,Ph* CO *CH(Ph)* CO,Et, is also colourless. CH, /\/\GO / V \ O H /\(\OH I I bH*CO,Et I I bO,H I \/\/Ph GO (25) (26) (2'1) OH \/\/ OH \/\/ Yellow (Metzner). Colourless (Metzner). Colonrless (Volhard). The salts of 1 : 3-diamino-2-phenylnaphthalene are colourless, but become intensely coloured on exposure to the air. A full investigation of its properties are in progress. EXPERIMENTAL. Formation of Ethyl p-1Tmino-a-cyano-r-p~~~n~lbutyrate, CH,Ph* C( :NH) CH( CN) CO,E t.This ester mas produced by the condensation of the sodium derivative of ethyl cyanoacetnte with benzgl cyanide in alcoholic solution. Twenty-three grams of sodium were dissolved in 275 grams of alcohol arid mixed with 113 grams of ethyl cyanoacetate. ToREACTIONS OF IMINO-COMPOUNDS. PART 11. 1917 the well-cooled solution 11 7 grams of benzyl cyanide were added and the whole heated on the water-bath. It was found that the yield of the condensation product varied considerably with the length of time during which the mixture was heated and that much care had to be exercised in order t o obtain the above compound in yields sufficiently large for the purposes of the research. This was necessary owing to the fact t h a t soon after being formed, the con- densation product was found to pass rapidly into another compound (see below) which has not yet been identified.This difficulty will be realised when i t is pointed out t h a t after the above mixture had been heated for twelve hours a t 100' only about 10 per cent. of ethyl p-imino-a-cyano-y-phenylbutyrate could be isolated, the majority of the benzyl cyanide being recovered unchanged, whereas if the heating was prolonged at the same temperature for twenty- four hours the product, whilst containing 25 per cent. of the ester, was found to be mixed with as much as 40-45 per cent. of another compound of high molecular weight, and only a relatively small proportion of the benzyl cyanide was recovered uuchanged. On heating for a still longer time the quantity of the complex substance increased and the amount of ethyl P-imino-a-cyano-y-phenylbutyrate formed and of unchanged benzyl cyanide diminished.After numerous experiments carried out under different conditions in which yields of the ethyl ester varying from 10-50 per cent. of the theoi etical amount were obtained, the following method was ultimately adopted as being the most convenient. The molecular mixture described above was heated on the steam- bath for fifteen hours, the flask being vigorously shaken from time to time. It was observed that the white insoluble sodium derivative of ethyl cyanoacetate gradually dissolved and that the solution became deep red in colour and very viscous. After being heated for the specified time the product was cooled, when it set to a thick jelly.Dilute hydrochloric acid was then added and the heavy oil which separated extracted by means of ether, the ether solution being washed with water and then with dilute sodium carbonate solution. During the washing with sodium carbonate a considerable quantity of acid oil was extracted, a description of which will be found below. The ether solution, without being dried, was evaporated free from ether, and the residue subjected to distillation in a current of steam, whereby the whole of the unchanged benzyl cyanide was recovered in the distillate. The residue, which consisted of a viscid brown oil solidifying on cooling, was then separated by filtration, dried on a porous plate, and recrgstallised from absolute alcohol.Ethyl p-imino-a-cyano-y-pheizyZbzLtyrute, purified in this way, was obtained in long, colourless needles, which melted at 125' : 6 K 21918 ATKINSON AND THORPE : THE FORRIATION AND 0 . 1 877 gave 0.4659 CO, and 0.1068 H20. C = 67.69 ; H = 6.32. C1,H,,O,N, requires C = 67.81 ; H = 6.1 per cent. The ethyl ester is insoluble in solutions of alkali carbonates and in aqueous caustic alkalis and is not acted on, even on prolonged boiling, by concentrated hydrochloric acid. It can be recrystallised from either methyl or ethyl alcohol or from benzene, but is almost insoluble in light petrole urn. The sodium carbonate washings from the ether solution were acidified, when a heavy oil separated, which partially solidified, This was filtered by the aid of the pump and the mixture of oil and solid washed with ether, in which solvent the solid was found to be quite insoluble.The residue was dried on a porous plate and found t o consist of a light brown amorphous substance apparently insoluble in all the usual organic solvents with the exception of hot glacial acetic acid in which it was only very sparingly soluble. It was insoluble in concentrated hytlrochloric acid, but dissolved in concentrated sul- yhuric acid apparently uiichanged, since i t was reprecipitated on the addition of water. It is, liowever, readily soluble in dilute solutions of alkali mrbonates and in caustic alkalis aud in the latter case on tlie addition of excess of the alkali :L crystalline alkali salt slowly separates. This salt is, however, difiicult to purify since it is dissociated by water, rind therefore cannot be freed from the excess of alkali.For the purpose of analysis the compound was purified by conversion into the potassium salt, which was effected by first washing the crude product thoroughly with ether and then dissolving it in a moderately dilute solution of caustic potash. On standing a crystal- line yotussiuni salt slowly separated which mas filtered and recon- verted into the higinal acid by treatment with dilute hydrochloric acid. The white amorphous precipitate thus obtained was then re- cryst:illisecl from a laiage quantity of glacial acetic acid, from which solvent it separated in colourless, microscopic needles which me1 ted t o a brown liquid a t 335' : * I. 0.3036 gave 0.5346 CU, arid 0.1416 H,O. C = 73.76 ; H = 5.14.11. 0.2072 ,, 0.5610 CO, ,, 0.0940 H,O. C=73.84 ; H=5*08. 111. 0.2508 ,, 35.0 C.C. nitrogen at 18" arid 767 mm. N = 15.95. C2,Hl,0N, requires C = 73.7 ; H = 5.2 ; N = 16.4 per cent. These figures furnish no clue as to the identity of the substance. Considerable quantities of this compound having accumulated during the course of this research, it is our intention further to investigate its properties. TlJe ethereal layer which floated on the filtrate from the above * This melting p i n t mas taken in cyaiibenzyliiie (see p. 1932).RE-4CTIONS OF ININO-COMPOUNDS. PART 11. 1919 compound after washing it free from oil was separilted in the funnel, washed with water, dried and the ether evaporated. The residue, which consisted of a viscid, dark oil readily soluble in dilute sodium carbonate solution, was, without further purification, mixed with three times its volume of absolute alcohol and an equal volume of con- centrated sulphuric acid and after being allowed to stand for twelve hours, heated on the water-bath for three hours.The mixture mas then poured into water, the oil which separated extracted with ether, and after any unchanged acid had been extracted by means of dilute sodium carbonate solution, the ethereal solution was dried and evaporated free from ether. The brown liquid which remained deposited a considerable quantity of crystals on standing, and these were collected and purified by recrystallisation from hot alcohol, being obtained in silky needles melting a t 178" : 1. 0.1868 gave 0.5262 CO, and 0.0882 H,O.C = 76-65 ; H = 5.28. 111. 0.2284 ,, 24.6 C.C. nitrogen a t 13' and 756 mm. N = 12.78. IV. 0.2264 ,, 24% C.C. ,, ,, 13" ,, 752 mm. N=12*Sl. IT. 0.1950 ,, 0.5472 C02 ,, 0.0908 H20. C = 76.52 ; H= 5.17. C2,HI7ON, requires C == 77.0 ; H = 5-2 ; N = 12.9 per cent. The compound is insoluble in alkalis and in concentrated hydro- Jts constitution has not yet been determined. chloric acid. The Constitution of Ethyl p-lni ino -a - cy ano- y - ph e 2 ~y l b u t yrnte. Formation of Phenylacetic and Malonic Acids bg the Action of illetlryl Alcoholic Potash. The proof o€ the constitution of ethyl /I-imino-a-cyano-y-phenyl- butyrate was derived from a study of the products which mere formed when it was completely hydrolysed by means of an alcoholic solution of potash. As already mentioned in the introduction, this proof is not in itself conclusive, since ethyl P-imino-y-cyano-y-phenylbutyrate, CHPh(CN)*C( :NH)*CH,*CO,Et, which might conceivably be formed in the same reaction would also give the same products on alkaline hydrolysis, but that, taken in conjunction with the other reactions of this ethyl salt, places its constitution beyond question.When ethyl /3-imino-a-cyano-y-phznylbutyrate was mixed with a solution of 1; times the calculated quantity of potash dissolved in methyl alcohol, considerable heat was generated and much ammonia was evolved on heating the solution on the water-bath. After heating for six hours the evolution of ammonia had ceased and a clear solution was obtained on pouring the hydrolysed product into water. The excess of methyl aIcohol was then evaporated and the residue acidified by means of hydrochloric acid, when a white, crystalline substance1920 ATKINSON AND THORPE: THE FORMATION AND separated.This was collected and recrystallised from dilute alcohol, being t.hus obtained in glistening plates melting a t 76". An analysis proved, as its appearance and odour indicated, that it was phenyl- acetic acid : 0.2107 gave 0.5222 CO, and 0.1232 H,O. C7E,0, requires C r= 67.8 ; H = 6-4 per cent. The filtrate from the phenylacetic acid was evaporated to dryness, acidified with concentrated hydrochloric acid and again evaporated, being finally placed in a Soxhlet apparatus and extracted with ether. The ethereal solution on drying and evaporating deposited a quantity of oil which rapidly solidified and then melted a t 132' with decomposi- tion.The analysis, together with the fact that on distillation it yielded acetic acid, showed it to be malonic acid : C = 67-61 ; H= 6 49. 0.2217 gave 0.2796 CO, and 0.0778 H20. C = 34.39 ; H = 3.89. C,H,O, requires C=34*6; 11=3-8 per cent. Formation of Ethyl 1 : 3- Diaminonaphthc~ Zene-2-car box ylute, When dry ethyl /3-imino-a-cyano-y-phenylbutyrate ground to a fine powder is added to three times its weight of cold, concentrated sulphuric acid it dissolves rapidly forming a yellow solution. If the mixture is allowed to becouie warm this solution becomes deep malachite-green in colour, but i t remains yellow if placed in a mixture of ice and salt. If, as soon as the ethyl salt has all dissolved, an operation which usually takes abont one minute, the strong d p h u i i c acid solution is poured on ice, a clear yellow solution is pro- duced from which an almost white crystalline substance separates on standing.This compound is the sulphccte, C,3H140,N,,H,S0,, of the above base and can be obtained in almost colourless plates by recrystallisation from water : 0.1539 gave 0,2706 CO, and 0.0705 H,O. Cl,H,40,N2,H2S0, requires C: = 47.6 ; H = 4.9. The sulphate gradually becomes yellow on exposure to the air, and when dissolved in water forms, in the cold, an almost colourless solution, which on warming becomes bright yellow; this yellow colour is discharged on cooling. Ethyl-1 : 3-diaminonapiithalene-2-carboxylate is precipitated as a yellow, crystalline solid on adding aqueous ammonia to a solution of tbe sulphate aissolved in water.It was collected by filtration and C = 47.95 ; H = 5.12.REACTIONS OF ININO-COMPOUNDS. PART 11. 1921 recrystallised from ether, being obtained in this way in golden yellow plates melting a t 108' : 002244 gave 0.5568 GO, and 0.1262 H,O. 0.3238 ,, 32.4 C.C. of nitrogen at 12" and 769 mm. N = 12-13, C1,H140,N, requires C = 67.8 ; H = 6.1 ; N = 12.2 per cent. C = 67.65 ; H = 6.24. The hydrochloride, C1,H1,O2N2,2HCI, is best prepared by dissolving the base in dilute hydrochloric acid and then adding excess of con- centrated hydrochloric acid to the almost colourless solution. On standing, colourless needles of the above salt separate : 0.2911 gave 0.2734 AgCI.C1= 23.30. Cl,H,40,N2,2HCl requires C1= 23.4 per cent. Ethyl 1 : 3-diaminonaphthalene-2-carboxylate is very readily oxidised by the mildest oxidising agents, being even slowly altered on exposure to the air. The compounds produced in these circumstances as well as the products formed by the action of nitrous acid on this ethyl salt are still under investigation, 1 : 3-DiaminonaphthaZene-2-cccrboxylic acid, I /\/\NH, I Jc02H \A/ NH2 This acid was prepared from the ethyl salt by hydrolysing it with methyl alcoholic potash. 5 grams were added to a solution of 18 times the calculated quantity of caustic potash dissolved in meth3 1 alcohol and the solution warmed on the water-bath. On cooling a crystalline potassium salt, CI1H,O,N2K, separated, which was collected by filtration, purified by washing with a little methyl alcohol, and dried on a porous plate.It is a yellow, crystalline solid readily soluble in water : 0.2195 gave 0.0788 K,SO,. K = 16-09. C,1H,0,N2K requires K = 16.25 per cent. The free acid is precipitated as a white, crystalline solid on adding dilute hydrochloric acid to a solution of the potassium salt in water. It was collected by filtration, dried on a porous plate and recrystallised from warm water. Since the acid rapidly loses carbon dioxide at 80°, it is necessary that the aqueous solution should be considerably below this temperature. Since, however, the acid is readily soluble in warm water and separates again on cooling, it can easily be recrystallised in quantities without raising the temperature of the solution above 60" 1 : 3-Diaminonaphthalene-2-carboxylic acid can be obtained in thisway in colourless needles melting at 8 5 O with immediate elimination of carbon dioxide : 0.2258 gave 0.5374 CO, and 0.1064 H,O.The carboxylic acid does not form salts with acids. C = 64.90 ; H = 5-27, Cl,Hlo0,N2 requires C = 65.3 ; H = 4.9 per cent. 1 ; 3 -AT~6~l~t?~ylenediallzi12e. This base was prepared from the above carboxylic acid by heating i t at 100' until the evolntion of c:trbon dioxide had ceased. The opera- tion was carried out in n small flask which was placed in n bath of sulphuric acid and heated to the requisite temperature. When the evolution of gas had ceased, the product was dissolved in dilute hydro- chloric acid and the deep red solution, after filtering, rendered alkaline with caustic potash.The base, which was then precipitated, was recrystallised from water and obtained in small plates, usually slightly red in colour, which melted at 96": 0.1628 gave 0.4520 CO, and 0.0910 H20. CloHloN2 requires C = 76.0 ; H = 6.3 per cent. The diacetyl derivative, CloH,N,(CO*CH,), mas prepared by dissolving the base in acetic anhydride and heating the solution on the sand-bath for one hour. The excess of the anhydride was then evaporated by placing the solution in an evacuated desiccator over potash until solid. Recrystallised from acetic acid, i t was obtained in fine needles melting at 263-265" (Friedlander, Ber., 1895, 28, 1953) : C = 76.00 ; 13 = 6%. 0.1693 gave 0.3826 CO, and 0.1112 H,O. C = 61-64 ; H= 7.29.C,oH,,O,N, requires C = 6 1.9 ; H = 7.2 per cent. Pornintion of Etlq,! /3-Imiizo-a-cyano- y p l m yl vnlernte, P h C HMe C( : N H) C 13 (CN) CO,F,t. As stated in the introduction, tho methylation of ethyl p-imino- a-cyano-y-phenylbutyriite by means of sodium ethoxide and methyl iodide in alcoholic solution is not easily :~ccompIished. The difficulty is probably due to the fact that the unrenctive char,tcter of the methylene hydrogen atom causes a state of eqnilibrium to be set up between the sodium ethoxide, the ethyl salt and its sodium derivative, i n which only about one-third of the last-named is formed. Thus, by the action of methyl iodide on a cold mixture of equimolecular proportions of sodium ethoxide and ethyl /3-imino-a-cj ano-y-phenyl- butyrate, only about one-third of the ethyl salt is converted into its nietliyl derivative.Therefore, in order t o complete the conversion,REACTIONS OF IMIKO-COMPOUNDS. PART 11. 1923 the process of methylation was repeated four times in the following manner : Ten grams of the ethyl salt were added to a solution containing 1 gram of sodium dissolved i n 1 2 grams of alcohol, and the whole, after being well cooled, treated with 12 grams of methyl iodide. After standing for some time in the cold the mixture was heated on the wat'er-bath until a test portion diluted with water gave a neutral reaction with litmu., when, after the excess of alcohol and unchanged methyl iodide had been distilled off as completely as possible on the water-bath, a further quantity of sodium ethoxicle made by dissolving 1 gram of sodium in 13 grains of alcohol was added, and the solution treated with methyl iodide and again heated until neutral.After this process had been repeated four times, water was added arid the oil, which then separated and which almost immediately solidified, was filtered by the aid of the pump and purified by recrystallisation from nlcohol. Ethyl p-im iiao-a-c~uno-y-pl~enylvcLlernte is obtained in this way in large needles melting at 93' : 0.1596 gave 0.4052 CO, and 0.0959 H,O. C=69.24; H=6*72. C,,H,,U,N, requires C = 68% ; H = % a 6 per cent. Forniation of a-Pheny'propionic Acid and NaZonic Acid j r o n z RthJ ~-1nii~~o-a-cyc~no-y-p~~~nylvnlernte. The proof of the constitution of ethyl P-imino-a-cjano-y-phenyl- y-methylbutymte was derived from the s t d y of the prodiicts formed from i t on alkaline hydrolysis.The operation was carried out as follows : Five grams of the ethyl salt wero mixed with a solution containing one-and-a-half times the cnlculated qiiantity of caustic potash dissolved in methyl alcohol and heated on the water-bath until the odour of ammonia, which was apparen t a t the commencement of the hydrolysis, had entirely disappeared. The prodrict was then diluted with water and evaporated on the water-bath u n t i l free from methyl alcohol, when it was acidified with concentrated hydrochloric acid and allowed to stand, Since the oil which separated on the addition of the acid showed no tendency to crystallise, the solution WIS extracted by means of ether and the ethereal solution,afler drying, evaporated free from ether.The oil which remained was then distilled under the ordinary pressure and was found to boil constantly at 264-265", at which temperature a-phenylpropionic acid, which does not appear to be a solid at the ordinary temperature, boils (Fittig and Wurster, Annalen, 1 S79, 195, 165) : 0.2100 gave 0.5332 CO, and 0.1295 H,O. C = 7134 ; I€ = 6.85. C9Hl,0, requires C = 72.0 ; H = 6-7 per cent.1924 ATKINSON AND THORPE: THE FORMATION AND The aqueous solution after the extraction of the above acid wao evaporated to dryness and extracted in a Soxhlet apparatus with ether. The ethereal solution after drying was evaporated free from ether yielding a solid residue which melted at 132" and gave acetic acid on distillation, evidently therefore consisting of malonic acid : 0.1774 gave 0.2237 CO, and 0.0622 H,O.C,H,O, requires C = 34.6 ; H = 3.8 per cent. The above analysis indicates that the substance is mslonic acid and not methylmnlonic acid, and therefore shows that only one methyl group had entered into the molecule of ethyl P-irnino-a-cyano-y-phenyl- butyrate on its being completely methylated in the manner already described. Owing to there being only a slight difference between the percentage corn positions of ethyl /3-imino-a-cyano- y-phen y 1 bu tyrat e and its met hg 1 derivative, this fact could not be definitely settled on the results of their analysis alone. C = 34.41 ; H = 3.90. Formation of Ethyl 2 ; 4-Diamino- lrnethyhaphthaZene-3-carboxyZate, Me The conversion of ethyl p-imino-a-cyano-y-phenylvalerate into the above derivative of naphthaleno takes place with the same ease as that already described in the case of the unmethylated derivative.Five grams of the finely-ground ethyl salt were added slowly to three times its weight of concentrated sulphuric acid cooled in ice. On each addition, the salt instantly dissolved, and in spite of the low tempera- ture the sulphuric acid solution became deep bluish-green. When all had been added, the solution was allowed to stand for two or three minutes and then poured on ice. The deep bluish-green colour of the solution was instantly discharged on mixing with the ice, forming a clear yellow solution from which the sulphate, C,,H160,N2,H,S0,, of the base slowly separated on standing as a pale yellow, crystalline precipitate : 0.1708 gave 0.3064 CO, and 0.0828 H,O.The sulphate is readily soluble in hot water, but separates from its concentrated solutions in this solvent on cooling. An aqueous solution of the salt at 0" is nearly colourless, but becomes a deep yellow on being heated to the boiling point. On being again cooled, the colour C = 48.92 ; H = 5.42. C1,Hl,O,N,,H,SO, requires C = 49.1 ; H = 5.3 per cent.REACTIONS OF LMINO-COMPOUNDS. PART 11. 1925 is discharged. The sulphate is insoluble in alcohol and this fact can be utilised in its formation, since on pouring the bluish-green sulphuric acid solution into three times its volume of well-cooled absolute alcohol, the salt is completely precipitated and can be isolated by filtration.The free base was prepared by dissolving the sulphate in water and adding ammonia until strongly alkaline. The deep yellow oil which was then precipitated did not8 show any signs OF crystallising, and mas therefore extracted by means of ether, the ethereal solution being washed with a little water, dried with calcium chloride, and evaporated. The yellow syrup which remained instantly solidified on scratching, forming a deep yellow solid, which was recrystallised from methyl alcohol and thus obtained in orange-yellow prisms melting a t 75" : 0.1676 gave 0,4245 CO, and 0.1017 H20. C14H1602N2 requires C = 6 8 3 ; H = 6.6 per cent. Ethyl 2 : 4-dinmino- 1 -methylnaphthalene-3-carboxylate is readily soluble in cold ether and can in this way be distinguished from the lower homologue.It is also soluble in all the usual organic solvents excepting light petroleum, and is readily oxidised, but is not altered on exposure t o the air so rapidly as ethyl 1 : 3-diaminonaphthalene-2- carboxylate. C = 69.08 ; H = 6.74. Formation of 2 : 4-Diamino-l-methyZnaphthalene-3-carboxyZic Acid, Me This acid was prepared from the ethyl salt by the action of alcoholic potash. Ethyl 2 : 4-diamino-l-methylnaphthalene-3-carboxylate was added to a solution containing one and n half times the calculated quantity of potash dissolved in metbyl alcohol and the mixture heated on the water-bath until a test portion diluted with water remained clear, Water was then added and the solution evaporated on the water-bath until free from methyl alcohol, when i t was acidified, and the colourless, glistening leaflets of the acid which then separated collected by filtration.The acid, which is slowly decomposed by boiling water, was recrystallised rapidly from hot water, and obtained in this way in colourless needles which decomposed with evolution of carbon dioxide at 155-160O apparently without melting : C1,H,,O,N, requires C = 66q7 ; H = 5.6 per cent. 0.1697 gave 0.41 39 CO, and 0.0875 H,O. C = 66.52 ; H = 5.73.1926 ATKINSON AND THORPE: THE FORMATION AND Me /\,A I I JNH2 Formation of 2 : $-Dinrnino-l-methylnaphtilalene, I \/\/ NH2 This base was prepared from the carboxylic acid by heating it for some time a t lS0". The finely-ground and purified acid was placed in a small flask and heated in n bath of sulphuric acid at the required temperature until carbonic acid ceased t o be evolved, when the residue was treated with dilute hydrochloric acid, and the deep red solution thus formed boiled with animal charcoal and filtered.On adding dilute aqueous caustic soda to the filtrate, the base was precipitated as a resin, which did not crystallise on standing. It was therefore dissolved in concentrated hydrochloric acid and the solution allowed to stand, when the hydrochloride, CllH12N,,2HCl, slowly separated in white, silky needles : 0.2117 gave 0.2467 AgC1. C1= 28.91. C11H,,N2,2HC1 requires C1= 29.0 per cent. The hydrochloride readily dissolved in water, and the solution on being made alkaline with caustic soda solution deposited the base as a white solid, which, when recrystallised from dilute methyl alcohol, was obtained in fine, white needles melting a t 65' : 0.1912 gave 0.5358 CO, and 0,1245 H,O.The base becomes slowly coloured on exposure to air and light. The platinichloyide, C11HlzN2,H,PtCI,, is precipitated as a yellow, crystalline powder on adding excess of platinic chloride to a solution of the hydrochloride of the base in water : C=76*43 ; H=7*23. C,,H,,N, requires C = 76.7 ; H = 7.0 per cent, 0 2513 gave 0.0834 Pt. Pt =133.38. C,,H,,N,,H2PtCI, requires Pt = 33.5 per cent. Poi-mution of Ztl,,yZ ~-Inaino-a-c~ccno-y-phenyZ-n-he~oc~te, Ph*CHEt-C( 3 H ) CH( QN) C0,Pt. The preparation of the ethyl derivative of ethyl P-imino-a-cyano-y- phenylbutyrate was undertaken in order to show conclusively that only one alkyl group could be introduced into this compound by the action of sodium ethoxide and an alkyl iodide.The process adopted was the same as that already described in the case of the methyl derivative, only in the present instance even greater difficulty was experienced in introducing the alkyl group, and, after the process of ethylationREACTIONS OF IMINO-COMPOUNDS. PART 11. 1927 had been repeated four times, considerable quantities o€ unchanged ethyl salt were recovered. The actual operation mas carried out as follows : 10 grams of ethyl p-imino-a-cyano-y-phenylbutyrate were added to a solution containing 2 grams of sodium in 30 grams of alcohol, which, after being well cooled, was mixed with 24 grams of ethyl iodide and warmed on the water-bath until ;L test portion diluted with water showed a neutral reaction to litmus.The excess of alcohol and ethyl iodide was then distilled off as far as possible on tile waler- bath and the above process again repeated. When the treatment had been carried out four times, water was added, and the oil, which was then precipitated and which solidified on standing, separated by filtration. I n the first place, the solid was recrystallised from dilute methyl alcohol, from which solvent it separated in needles having a melting point of 60-80". These were then ground with cold benzene, in which a considerable quantity dissolved, leaving a residue which melted a t 124-1 25'. The following analysis showed this substance to be unchanged ethyl P-imino-a-cyano-y-phenylbutyrafe : 0,1554 gave 0.3888 CO, and 0.0903 H,O.C,,H1,O,N, requires C = 67.8 ; H = 6.1 per cent. The benzene filtrate on evaporation left a residue which melted atJ 60-65', which was again ground with cold benzene yielding :L further small quantity of unchanged ethyl P-imino-a-cyano-y-pheayl- butyrate. The benzene solution from this second separation gave a product which melted at 60-6 lo. This was fioally recrystallised from methq 1 alcohol, pure ethgl p-imino-a-cytsno-y-phen$-n-?iexoate being thus obtained in small prisms melting a t 60" : C = 63-23 ; I€ = 6.50. 0.1564 gave 0.3990 CO, and 0,0996 H,O. C = 69-58 ; H = 7.12. C1,Hl,O2N2 requires C = 69% ; H = '7.0 per cent. Fornicction oj' a-Phenylbutyric Acid, Ph*CH(Et)*CO,H, and Halonic Acid froiri Ethyl ,8-l~~aino-a-cyc~no-y-phen~Z-n-I~exoate. The position of the ethyl group in ethyl P-imino-a-cyano-y -phenyl- ethylbutyrate was determined by the products formed on its com- plete hydrolysis with methyl alcoholic potash.Five grams of the ethyl salt were heated on the water-bath with a solution containing one-and-a-half times the calculated quantity of caustic pot ash dissolved in methyl alcohol until the odour of ammonia has ceased to he apparent. Water was then added and the solution evaporated on the water-bath until free from methyl alcohol, when it was acidified with hydrochloric acid and the product allowed to stand. The oil which was first pre-1928 ATKINSON AND THORPE: THE FORMATION AND cipitated on the addition of the acid became solid on standing and was isolated by filtration. Recrystallised from dilute alcohol it was obtained in crystals melting at 43' which is the correct melting point of a-phenylbutyric acid (Neure, Annalen, 1889, 250, 154) : 0.2001 gave 0-5358 CO, and 0.1333 H,O.CloH,,O, requires C = 73-2 ; H = 7.3 per cent. The aqueous filtrate after extraction with ether was evaporated to dryness on the water-bath, and the residue extracted in a Soxhlet apparatus with ether. The ether solution, after being dried with calcium chloride, was evaporated free from ether, yielding a white solid melting at 132O, which was converted on distillation into acetic acid. It was evidently, therefore, malonic acid, and its formation together with that of a-phenplbutyric acid indicates clearly that the ethyl group in ethyl P-imino-a-cyano-y-phenyl-y-ethylbutyrate is at- tached to the same carbon atom as is the benzene nucleus.C= 73.03 ; H = 7.40. Formation of Ethyl 2 : 4- Diunaino- l-ethyZ.lzapiLtiLnlelze- 3 - carboxylate, Et The conversion of ethyl /3-imino-a-cyano-y-phenyl-n-hexoate into the above derivative of naphthalene was accomplished in the following manner. The finely-ground ethyl salt was added gradually to three times its weight of concentrated sulphuric acid, the acid being kept well cooled in ice through the addition. When all had been added, the sulphuric acid solution, which had become deep indigo blue in colour, was allowed to stand at the ordinary temperature for about two minutes, when it was poured on to ice. On coming in contact with the ice the blue colour of the solution instantly disappeared, forming a light yellow solution from which, however, no separation of the sulphate of the base took place on standing.The clear solution was therefore made alkaline with ammonia and the yellow oil, which was then precipitated, extracted by means of ether. I n order further to purify the base, this ethereal solution was extracted twice with diluto sulphuric acid, and the sulphuric acid extract made alkaline with ammonia. The yellow oil precipitated in this way, and which became solid on standing, was collected by filtration and recrystallised from dilute methyl alcohol, being obtained in small yellow prisms melting a t 63O : 0.1282 gave 0.3298 CO, and 0.0832 H20. C J 70.16 ; H - 7-21, C15H1802N2 requires C = 69.8 ; H = 7.0 per cent.REACTIONS OF IMINO-COMPOUNDS.PART 11. 1929 The hydrochloride, C,,Hl,0,N2,2HC1, is prepared by dissolving the base in concentrated hydrochloric acid, and allowing the solution to stand. C1= 21.67. C1,H,80,N,,2HCl requires C1= 21.5 per cent. It consists of almost colourless needles : 0.2102 gave 0.1838 AgCl. Formation of 2 : 4- Diamino-1 -eth ylnaphtha lene- 3-ca~boxylic Acid, Et I n order to prepare this acid, 4 grams oE the ethyl salt were mixed with a methyl alcoholic solution of one-and-a-half times the calculated quantity of potash, and heated on the water-bath until a test portion was completely soluble in water. The product was then diluted with water and evaporated on the water-bath until free from methyl alcohol, when dilute hydrochloric acid was added and the precipitate, which then separated, isolated by filtration.The acid, like those already described, is rapidly decomposed by boiling water. It was therefore crystallised by dissolving it in water at 60" and allowing the solution to cool, when the acid separated in small leaflets which melted and decomposed a t about 130", the observed melting point, however, varied from 128-133O with the same specimen, depending on the rapidity with which the temperature was raised : 0.1729 gave 0.4281 CO, and 0.0969 H20. C = 67.52 ; H = 6.22. C,,H,,O,N, requires C = 67-8 ; H = 6.1 per cent. Formation of 2 ; 4-Diamino- l-ethylnaphth4zlene. Et This base was prepared by heating the carboxylic acid a few degrees above its melting point. The acid was placed in a small flask and heated in a bath of sulphuric acid at 150' until the evolut,ion of carbon dioxide ceased.The residue, which consisted of a viscid red gum, was dissolved in dilute hydrochloric acid and the solution, after being filtered, made alkaline with dilute aqueous caustic soda. The precipitated base, which solidified on stmding, was then collected by filtration arid recrystallised from dilute metbyl alcohol, being obtained in glistening plates melting a t 74' :1930 ATKINSON AND T'HORPE: THE FOHMATION AND 0.1812 gave 0.5127 GO, and 0.1256 H20. C,,H,,N, requires C = 77.4 ; H = 7.5 per cent. The hydrocldo~ide, C,2H,,N,,2HC1, is precipitated as a white, crystal- line powder on dissolving the base in concentrated hydrochloric acid and allowing the solution t o stand for some time : c! = 77-17 ; H = 7-70. 0.2207 gave 0-2429 AgCl.C1= 27.31. C,,H,,N2,2HC1 requires C1= 27.4 per cent. The plutinicldoyide, C,,H,,N,,H,PtCI,, is formed when a solution of the hydrochloride is mixed with an aqueous solution of plstinic chloride. It is a yellow, crystalline powder : 0,2401 gave 0*078l Pt. Pt =32.53. C,,H1,N2,H,PtCI, requires P t = 32.7 per cent. Codensat i o n of Benxpl Cyanide with its Sodium DerivcLtive, As mentioned in tlie introduction, this condensation was carried out i n three different ways, the nitrile being heated with its sodium deriv- ative in alcoholic solution a t 100" for (1) half an hour, (2) two hours, (3) twenty-four hours. ( 1 ) Forn~cttion of p- t n ~ i n o - a - c y c ~ ~ ~ o - a y ~ i ~ ~ ? ~ e n ~ t ~ ~ r o ~ u ~ CH,Ph* C(: N H)* CH(CN)P h.After a number of preliminary experiments it was ascertained t h a t this substance was produced in t>he greatest yields in the following way. 11.5 grams of sodium were dissolved in 130 grams of alcohol and mixed with 117 grams of benzyl cyanide, the whole being heated on the water-bath for half an hour. After this time the contents of the flask, which had a slight odour resembling ammonia, were mixed with water and the oil which then separated extracted with ether. The ethereal solution after being freed from alcohol by washing with water was, without drying, evaporated free from ether, and then aft'er being mixed with water distilled in a current of steam until all unchanged benzyl cyanide had passed over. The residue, which consisted of a viscid oil, was then extracted by means of ether, and the ethereal solution, after being dried with calcium chloride, evaporated free from ether.The oil which remained mas rapidly distilled under reduced pressure and was found to boil at 274'/20 mm., passing over as a colourless syrup : 0.2016 gave 0.6070 CO, and 0.1111 H,O. C,,H,,N, requires C = SB.0 ; N = 6.0 per cent. The identity of this compound with that prepared by E. v. Meyer was shown by the fact that with hydroxylamine it gavo an oxime melting C=82.12; H=6.17.HEACTIONS OF IMINO-COMPOUNDS. PART 11. 1931 at 107O, identical with that formed by the action of this reagent on the ketone, C!H,Ph*CO*CH(CN)Ph. Even when carefully purified by dis- tillation /3-imino-a-cjano-ay-diphenylpropane is a viscid liquid and shows no tendency to become crystalline. (2) Formation of /3-Irnino-a-cyano-ay-diphnylpropane, /I-&to-a-cyctno-ay-diphenylpropane, CH,Ph-CO*CH(CN)Ph.Cyan- benzyline (6- Amino-5-phenyl-2 : 4-dibenxylpyrimidine) and Ethylamine. In this experiment the reacting substances were mixed in the same proportions as in experiment (1) but were heated on the water-bath for two hours. The odour resembling ammonia which was apparent a t the end of the first half hour, became very marked on further heating, the contents of the flask becoming a t the same time dark red. At the end of the specified time a small quantity of alcohol was distilled over and collected in dilute hydrochloric acid, the solution evaporated on the water-bath to a small bulk and mixed with excess of a solution of platinic chloride.On adding a little alcohol to the mixture a yellow precipitate separated out which was collected by filtration and recrystallised from a little hot water, being thus obtained in yellow cubes. The following analysis showed it to be the platinichloride of ethylamine : CH,Ph*C( :NH)*WH(CN)Ph. 0.2713 gave 0.1054 Pt. (C,H,NH,),,H,PtCl, requires Pt = 39 .O per cent. As soon as the small quantity of alcohol necessary for the above experiment had been distilled over, the alcohol solution without further evaporation was mixed with water, acidified with dilute hydrochloric acid and extracted with ether. The ethereal solution after being washed with water to remom alcohol was thoroughly shaken with a solution of three volumes of concentrated hydrochloric acid to one of water.It was found that by shaking the ethereal solution with this mixture the cyanbenzyline could be completely extracted in the form of its hydrochloride, This hydrochloride, which is a viscid liquid under ordinary conditions, is quite insoluble in both dilute hydrochloric acid and in ether. Therefore on shaking the ethereal solution with the dilute acid three layers are formed, an upper layer consisting of ethereal solution from which the hydrochloride has been extracted, a middle layer consisting of this hydrochloride in the form of a viscid oil and a lower aqueous layer of dilute hydrochloric acid. The three layers were completely separated in the separating funnel, and it was found that on repeating the process three times the whole of the cyan- Pt = 38.85.VOL. LXXXIX. C i L1932 ATKINSOX AND THORPE: THE FORMATION AND benayline present was extracted from the solution. The oily hydro- chloride on warming on the water-bath with excess of sodium carbonate solution yielded cyanbenzyline as a crystalline mass which when recrystallised from absolute alcohol or from light petroleum (b. p. 90-100') was obtained in slender needles melting at 107" : C,,H,,N, requires C = 82.0 ; H = 6.0 per cent. 0.2076 gave 0.6250 GO, and 0.1160 H,O. The properties of cyanbenzyline have already been described by the investigators mentioned in the introduction. In addition to the published data concerning this substance, we find that it can be distilled under diminished pressure without undergoing any decom- position, passing over at 307O (20 mm.) as a colourless oil which sets to a highly refractive colourless glass on cooling.On dissolving this glass in hot alcohol the solution deposits the original cyanbenzyline on cooling.* The hydrochloride, C,,H,,N,,HCl, was obtained as a crystalline solid in the following way : cyanbenzyline was dissolved in absolute alcohol containing a small quantity of concentrated hydrochloric acid in which mixture it is readily soluble. Water was then added until a clear solution just remained, when, on standing, the hydrochloride separated in small, colourless needles : C= 82.11 ; H = 6%. 0.1690 gave 0.0640 AgCl. C1= 9.39. C,,H2,N,,HC1 requires C1= 9.2 per cent. Cyanbenzyline is converted into 6-hydroxy-5-phenyl-2 : 4-dibenzyl- pyrimidine on heating it with a 10 per cent,.sulphuric acid for twenty- four hours. This substance, which was originally prepared by Wache (J. pi-. Chem., 1889, [ii], 39, 251), crystallises from alcohol in small needles and melts a t 187' : 0,1540 gave 0.4646 CO, and 0.0816 H20, 0.2710 ,, 18.8 C.C. nitrogen a t 20' and 752 mm. N=7*88. C,,H,,ON, requires C = 81.8 ; H = 5.7 ; N = 7.9 per cent. The ethereal solution after the extraction of cyanbenzyline was, without drying, evaporated free from ether and the residue distilled in a current of steam until free from the small quantity of unchanged C = 82.28 ; H = 6-93. .Ic The stability of cyanbenzyline is such that it can be distilled a t the ordinary pressure without undergoing decomposition boiling under these conditions above 400" When once melted, it cools to a clear glass, which does not become crystalline except on long standing.We have found this substance extremely useful for determining the melting point8 of compounds between 230" and 380", for which purpose it is well adapted since it does not fume and does not become coloured unless kept near its boiling point for some considerable time.REACTIONS OF IMINO-COMPOUNDS. PART 11. 1933 benzyl cyanide, when it was again extracted with ether, the ethereal solution dried by means of caloium chloride and the ether evaporated. The residue, which consisted of a viscid oil, yielded two chief fractions on distillation under diminished pressure. The lower fraction which boiled at 225-230" (20 mm.) solidified on cooling and was purified by recrystallisation from a mixture of benzene and light petroleum (b.p. 70-80") being obtained in needle clusters, melting a t 86' : 0.1566 gave 0.4712 CO, and 0.0800 H,O. 0.2406 ,, 12 C.C. nitrogen at 20" and 756.5 mm. N = 5.70. CI6H,,ON requires C = 81.7 ; H = 5.5 ; N = 6.0 per cent. The analysis and melting point show this substance to be P-keto- a-cyano-ay-diphen ylpropane, originally prepared by v. Meyer (J. p. Chenz., 1895, [ii], 52, 115), by the action of sodium ethoxide on a mixture of ethyl phenylacetate and benzyl cyanide, a view which was confirmed by its conversion into the oxime (m. p. 107O) by the action of hydroxylamine. The higher fraction which boiled constantly at 274' (20 mm.) did not solidify on standing, but from the following analysis evidently consisted of /3-imino-a-cyano-ay-dipheny lpropane : C = 82.06 ; H =5*71.0.2013 gave 0.6042 CO, and 0.1110 H,O. C=81*86; H=6*12. Cl6HI4N2 requires C = 82.0 ; H = 6.0 per cent. ( 3) Format ion of P- Ke to - a - c y ano - a y -d iphen yZpropane and C y anbenz yline. I n this experiment the mixture of benzyl cyanide and sodium ethoxide in alcoholic solution was heated for twenty-four hours at looo, when the odour of ethylamine was hardly apparent. The product was mixed with water and the oil, which then separated, extracted with ether. The ethereal solution, after being washed with water to remove alcohol, was thoroughly washed with a mixture of three parts of concentfated hydrochloric acid to one of water, when the hydrochloride of cyanbenzyline separated as a middle layer in the manner already described.From this oily hydro- chloride the free base was isolated by means of sodium carbonate the yield of cyanbenzyline obtained in this way being about 60 per cent. of that theoretically possible. The ethereal solution, after being dried by means of calcium chloride, was evaporated free from ether and distilled under diminished pressure. It was found to boil con- stantly at 225-230° (20 mm.), and on cooling solidified to a crystal- line cake which when recrystallised from a mixture of benzene and light petroleum (b. p. 70-80O) yielded P-keto-a-cyano-ay-diphenyl- propane, melting at 86O. There is, therefore, apparently no p-irnino-a- cyano-ay-diphenylpropane formed under these oxperirnental conditious, c i L 21934 FORMATION OF IMINO-COMPOUNDS. PART 11.the amount produced in the first instance having either condensed with unchanged benzyl cyanide to form cyanbenzyline, or having passed into P-keto-a-cyano-ay-diphenylpropane through the '' hydrolysis " of the imino-group. The transformation of p-imino-a-cyano-ay-diphenylpropane into 1 : 3-diamino-2-phenyloaphthalene was effected in the following manner: five grams of the liquid nitrile were added gradually to 15 grams of concentrated sulphuric acid, the solution being kept well cooled throughout the addition, since the liquid, which dissolved rapidly in the concentrated acid, caused a considerable rise in temperature. When all had dissolved, the solution, which was dark brown, was allowed to stand at the ordinary temperature for from three to four minutes and then poured into a large volume of water. The solution thus obtained, after being filtered to remove any impurities, was made alkaline with ammonia, and the base, which was then precipitated, collected by filtration and purified by recrystallisation from methyl alcohol or benzene. 1 : 3-Diamino-2-phenylnaphthaZene crystallises in colourless, glistening plates melting at 116O : 0.1564 gave 0.4699 CO, and 0.0853 H,O. 0.3238 ,, 32.6 C.C. nitrogen at 15.8' and 758.4 mm. N= 11.73. C16H14N2 requires C = 82-0 ; H = 6.0 ; N = 12.0 per cent. The base becomes slowly coloured red on exposure to the air. It is easily soluble in alcohol or benzene, sparingly so in cold ether, and insoluble in light petroleum, Its solutions exhibit a marked blue fluorescence. The hydrochloride is precipitated in fine feather-like needles on dis- solving the base in a little dilute hydrochloric acid, and, after adding an equal volume of concentrated hydrochloric acid, allowing the solu- tion to stand : C = 81.94 ; H = 6.10. 0.2211 gave 0.2043 AgC1. C1= 22-92. CI6H,,N,,2HC1 requires C1= 23-1 per cent. The hydrochloride is rapidly coloured red on exposure to light. The dicccetyl derivative, 0, ,H,,N,(Ac),, is best prepared by boiling the base with excess of acetyl chloride until all has passed into solu- tion. A considerable quantity of acetyl chloride is necessary for this purpose, since the acetyl derivative is only sparingly soluble in this reagent. When all had dissolved the solution was cooled and allowedA NEW TRINITROACETAMINOPHENOL. 1935 to atand, the crystalline precipitate which then separated being isolated by filtration and ptirified by dissolving in alcohol and diluting the solution with water. On standing, the wcetyl derivative separated in small needles, melting at 367' : 0.1743 gave 0.4805 CO, and 0.0928 H,O. C = 75-18 ; H = 5.96. C,,H,,0,N2 requires C = 75.5 ; H = 5.7 per cent. The acetyl acetate, C,,H,,N,Ac,CH,*CO,H, was formed in an attempt to prepare the diacetyl derivatim by the action of acetic anhydride on the base. 1 : 3-Diamino-2-phenylnaphthalene was boiled with excess of acetic anhydride for two hours and the solution evaporated in an evacuated desiccator over caustic potash. The gummy residue became completely solid on scratching, and when recrystallised from absolute alcohol was obtained in small prisms melting at 185' : I. 0.1847 gave 0.4864 CO, and 0.0969 H20. C = 71.82 ; H = 5-87 11. 0.1872 ,, 0.4931 CO, ,, 0.0988 H,O. C=71*84; H=5*90. C,,H,,O,N, requires C = 7 1-4 ; H = 5-9 per cent. Other derivatives of this base are still under investigation. Much of the expense entailed by this research has been met by a grant from the Government Grant Committee of the Royal Society, for which we desire to express our indebtedness. MANCHESTER UNIVERSITY.