126 FORSTER AND MULLER: THE TRIAZO-GROUP. PART XI.XV. -The Triaxo -group. Part XI. Sub st i t u t edTriaxomalonic m d Phenyltrialxoacetic Acids.By MARTIN ONSLOW FORSTER and ROBERT M~LLER.PURSUING our study of the effect produced by environment on thebehaviour OF the triazo-group, we deal in the present communicationwith substituted triazomalonic acids, for the purpose of comparisonwith the series of monobasic fatty acids already considered in previouspapers (Trans., 1908, 93, 7 2 ; 1909, 95, 191). It has been shownthat the characteristic manner in which triazoacetone is decomposed byalkalis may be deeply modified by exchanging the unsubstitutedmethyl group for ethoxyl, the azidic radicle in triazoacetic ester andits homologues being quite indifferent towards alkali, the attack ofwhich is resisted even by the triazo-acids themselves unless consider-able excess of the agent is employed and the temperature raised to50° or more.From the fact that i n the series quoted, a-triazoiao-butyric acid escaped altogether, it was concluded that the limitingcondition for elimination of t wo-thirds the azidic nitrogen from triazo-acids depends on the association of hydrogen with the carbon atomwhich carries the triazo-group.Accepting ethyl triazoacetate as a standard, the series of substitutedtriazomalonic esters described in this paper may be regarded asderived by replacing one or both atoms of hydrogen with carboxyethylFORSTER AND MULLER: THE TRIAZO-GROUP. PART XI. 127alkyl, phenyl, or triazidic groups, as represented by the followingconstitutional formulz :E t 0 C 0- C N3<g EtO* CO*CN3<C0,E HE thy1 triazoace tate.E tO*CO*CN3<,023, CHEthyl triazomalonate(not isolated).EtO*CO*CN,<C,$t CHEthyl methyltriazomalonate. Ethyl ethyltriazomalonatc.Ethyl phenyltriazomalonate.E t h ylbis triazomalonate.The first point to notice in connexion with the above series is thatthe introduction of a second carboxyethyl group into the niolecule oft riazoacetic ester disturbs so profoundly the equilibrium of thomolecule that ethyl triazomalonate is not capable of separate exist-ence. All attempts to prepare this compound by variations of theoriginal process for obtaining triazo-esters have been fruitless, notfrom want of reactivity between sodium azide and the halogen of thesubstituted malonate, but because under those circumstances in whichdouble decomposition can be brought about, the product immediatelyundergoes profound decomposition, and gives rise to nitrogen, ammonia,and hydrazoic acid, along with a solid nitrogenous compound of highmolecular weight.As soon as the remaining hydrogen atom isreplaced, however, whether by alkyl, phenyl, or the triazo-group,stability returns to the molecule, and the substituted triazornalonicester is sufficiently cohesive not only to withstand the action OF alkali,but actually to undergo hydrolysis, furnishing the substituted triazo-malonic acid.Tn the alkyltriazomalonic acids we have to deal with substancesdirectly comparable with a-triazoisobutyric acid, inasmuch as thetriazotised carbon atom is tertiary.Accordingly they displayunusual resistance towards alkali, which depends for the success of itsattack on the well known power OF substituted malonic acids to loseone molecule of carbon dioxide. Making this alteration in ethyltriazo-malonic acid, for example, leads to a-triazobutyric acid :Consequent on this, there occurs elimination of two-thirds the azidicnitrogen, so that the net result of decomposing ethyltriazomalonic acidwith alkali is propionylformic acid :CH3*CH2*CN3(C0,H)2 + H20 = CH,*CH,*CO*CO,H + 00, + N, + NH,.Proceeding now to consider the effect of the phenyl group, anincrease in the reactivity of the molecule is to be noted, and althoughit has proved possible to hydrolyse the ester to phenyltriazomalonicCH3*CH2*CN3(C0,H), - CO, = CH,*CH2*CHK3*C02H128 FORSTER AND MULLER: THE TRIAZO-GROUP.PART XI,acid, and by the action of ammonia to prepare phenyltriazomalonamide,nevertheless the further breakdown to benzoylformic acid is accom-pIished more readily than the corresponding degradation of the purelyaliphatic molecule :C,H,*CN,(CO,H), + H20 = C,H,*CODC02H + CO, + N, + NH,.This access of reactivity is shown still more clearly by the behaviourof phenyltriazoacetic ester, which we have prepared for comparisonwith triazoacetic ester. Whilst the latter substance may behydrolysed to the acid without risk of eliminating the two-thirdsnitrogen, phenyltriazoacetic ester is so sensitive towards alkalithat the acid cannot be prepared by hydrolysis, but must be derivedfrom sodium phenylchloroacetate by double decomposition with sodiumazide :C,H,*CHCl*CO,Na + NaN, = C,H,*CHN,*CO,Na + NaCl,Ey proceeding carefully it is possible to arrest decomposition ofphenyltriazoacetic acid with alkali at the stage intermediate betweenthe original substance and benzoylformic acid, namely, potassiumphenyliminoacetate :C,H,=CHN,*CO,K --+ C,H,*C( :NH)*CO,K + N,,but this compound readily suffers hydrolysis to potassium benzoyl-formate and ammonia.The instability of phenyltriazoacetic ester isillustrated still better by its behaviour towards ammonia, which con-verts it into phenyliminoacetamide with loss of two-thirds nitrogen andwitbout formation of phenyltriazoacetamide.C,H,*CHN,*CO,*C,H, + NH, =C,H,-C(:NH)*CYO*NH, + C,H,*OH + N,.The latter substance, i n fact, could not be isolated, although the trans-formation of triazoacetic ester [into triazoacetamide by the directaction of ammonia is practically quantitative.Proceeding now to the bistriazo-derivatives, we are confronted withanother illustration of the comparative immunity of the triazo-groupwhen the carbon atom which carries it is tertiary.I n studying theproperties of bistriazoacetic ester (Trans., 1908, 93, 1073), it wasfound that potassium hydroxide readily eliminates hydrazoic acid, andthat even ammonia brings about this change too rapidly to permit theproduction of bistriazoacetamide ; moreover, during the preparation ofthe ester i t mas noticed that the double decomposition between di-chloroacetic ester and sodium azide was accompanied by continuousliberation of hydrazoic and prussic acids, indicating subsidiary changeswhich reduced the yield very much below that required by theory.Bistriazomalonic ester, on the other hand, is a comparatively stablesubstance, and, although exploding with some violence a t lSOo, is muchless dangerous to handle than bistriazoacetic ester.Ammonia convertFORSTER AND MULLER: THE TRIAZO-GROUP. PART XI. 129it into bistriazomalonamide, and although this bhange is accompaniedby elimination of hydrazoic acid, the isolation of a solid amiderepresents a distinct superiority to the behaviour of the bistriazoaceticester. As might be expected, however, the amide is much moresensitive towards concentrated sulphuric acid than.methyltriazomalon-amide and ethyltriazomalonamide, and is completely broken down bypotassium hydroxide, yielding nitrogen and ammonia with prussic andhydrazoic acids.An attempt to produce bistriazomalonic acid by cautious hydrolysisleads us to believe that this substance, like triazomalonic acid, iscapable of existing in solution, but not in the individual state. Anethereal solution was concentrated without heating, but therebyacquired a powerful odour of hydrazoic acid, which remained noticeableduring many days' exposure to reduced pressure; a few crystals ofoxalic acid separated, and the oil, when quite free from hydrazoic acid,gave a nitrogen percentage agreeing fairly well with that of triazo-glycollic acid, this indication being further confirmed by analysis ofthe barium salt. It appears probable, therefore, that bistriazomalonicacid undergoes the following changes in solution :N N;>c<co;H CO H + H,O = H o > ~ ~ * ~ ~ , ~ + CO, + HN,.N3The production of prussic acid during the disruption of the triazo-group appears to be a feature of bistriazo-derivatives exclusively, andwe have searched again for this compound among the products ofdecomposing with alkali those triazo-acids and triazo-esters which havebeen described in previous pBpers, with a negative result ; it is, horn-ever, noteworthy that phenyltriazomalonamide, when warmed withalkali, gives a distinct odour of phenylcarbylamine.It is difficult toexplain the formation of prussic acid from bis triazomalonamide unlessthe production of the substance is preceeded by the loss of carbondioxide, which mould lead t o bistriazoacetamide ; this might beexpected to lose two-thirds the nitrogen of one azidic group, yieldinga highly unstable molecule which mould become resolved into a,mmoniawith hydrazoic, carbonic, and prussic acids :HCN + HN, + CO, + NH,.Such a change appears more probable when it is recalled thatConrad and Bruckner have shown that dichloroacetamide is amongthe products of treating dichloromalonic ester with ammonia (Bey.,1891, 24, 2994).VOL. XCVII. 130 FORSTER AND MULLER: THE TRIAZO-GROUP. PART XI.Much remains to be learned, however, in connexion with thosebistriazo-compounds in which both azidic groups are attached tothe same carbon atom, other points in addition to those mentionedserving to emphasise the characteristic behaviour of such substances.Bistriazomalonic ester, for example, is remarkably stable, and doesnot appear t o undergo alteration with lapse of time, whilst bistriazo-acetic ester develops the odour of prussic acid and deposits, inthe course of a few months, massive, transparent crystals (m.p. 91')containing more carbon, but less nitrogen, than the original material.This rearrangement will be investigated further, as there doubtlessoccurs some change comparable with the triazole formation whichtakes place in the molecule of allylazoimide (Trans., 1908, 93, 1174),and with the production of 1-hydroxy-5-phenyltetrazole from benz-hydroximic chloride and sodium azide (Trans., 1909, 95, 183); therecent observation of Schroeter (Bey., 1909, 42, 2336), who obtaineddiphenyltetrazole from diphenylbistriazomethane, belongs, probably,to the same class of transformation,EXPERIMENTAL.Interaction of Chloromalonic Ester and Sodium Axide.Fifty grams of chloromalonic ester (b.p. 91°/2 mm.) were heatedunder reflux on the water-bath with 25 grams of alcohol, 20 gramsof sodium azide, and sufficient water to maintain the salt insolution; the mixture rapidly became yellow, then dark red, whilebrisk effervescence was set up, the escaping gases being nitrogen,ammonia, and hydrazoic acid. After four hours the liquid wasallowed to cool, and filtered from 3 grams of crystalline material,of which a further quantity about equal to the first was obtained byremoving esters in a current of steam, concentrating the residue t oabout 200 c.c., and then adding dilute sulphuric acid.The substancewas purified by precipitation with sulphuric acid from the solutionin sodium carbonate, followed by successive recry stallisation fromboiling alcohol and ethyl acetate independently, approximately onelitre of the latter solvent being required by 0.5 gram; it crystallisedin minute, colourless needles, became deep red at about 220', anddecomposed completely in the neighbourhood of 240'. Many analyseshave been made, of which the following are typical, without, however,1 evealing the identity of this compound, but they point consistentlyt o a molecule arising by condensation of two or more molecules oftriazomalonic ester :0.2061 gave 0.3577 GO, and 0*0908 H,O.C = 4799 ; H = 4.90,0.1181 ,, 13.2 C.C. N, at 20° and 746.5 mm. N = 12.56.(C9H1105N2)s requires C = 47.58 ; H = 4.84 j N = 12.33 per centFORSTER AND M ~ ~ L L E R : THE TRIAZO-GROUP. PART XI. 131The substance is very sparingly soluble in boiling water, alcohol,chloroform, and ethyl acetate; it does not reduce ammoniacal silveroxide, and does not give any characteristic coloration when ferroussulphate is added to a very dilute solution in sodium hydroxide, buta solution in cold absolute alcohol, which of necessity contains only aminute portion of the substance, develops an intense violet colorationwith ferric chloride.As regards the preparation of triazomalonic ester, the foregoingexperiment was a failure, because not only is the solid p'roduct quitedistinct from a triazo-compound, but *he volatile oil removed bysteam was found not to contain nitrogen.Another fruitless attemptto obtain the substance consisted in shaking during twenty hours a tlaboratory temperature a suspension of chloromalonic ester (50 grams)in aqueous alcohol containing sodium azide (20 grams) ; the mixturebecame dark red, and pressure was developed, but the odour ofammonia or hydrazoic acid was not perceptible, and the heavy oilwhich remained undissolved consisted of original material.Having found that sodium phenyltriazoacetate may be preparedfrom the chloro-compound by the action of aqueous sodium azide,25 grams of bromomalonic acid were neutralised with sodiumcarbonate, and gently warmed with 10 grams of sodium azidedissolved in water.It soon became evident that double decom-position had occurred, because a test with 40 per cent. potassiumhydroxide gave torrents of nitrogen and ammonia, but on extractingthe acidified solution with ether, and evaporating the solvent underreduced pressure, hydrazoic acid was liberated continuously, leavingan oil which no longer evolved nitrogen when treated with alkali.Methykriaxomcdonic Acid, CH,*CN,(CO,H),.Five grams of methyltriazomalonic ester were shaken with 5 gramsof potassium hydroxide dissolved in 5 C.C.of water until, after abouttwenty minutes, the oil had disappeared, when the liquid wasgradually acidified with 50 per cent. sulphuric acid and extractedwith ether. The residue from the latter solidified in the desiccator,and, after crystallisation from warm benzene, was redissolved in itsown weight of ethyl acetate; on adding benzene in approximatelyequal volume, there separated stellate aggregates of long, transparentprisms melting at 8 7 * 5 O :0.1551 gave 37.0 C.C. N, at 22' and 743 mm.C,H50,N, requires N = 26.4 1 per cent.The acid is very hygroscopic, and dissolves freely in ethyl acetate ;benzene dissolves it only sparingly, however, and it is insolublein petroleum, Concentrated sulphuric acid attacks the substanceN = 26.40.K 132 FORSTER AND MOLLER: THE TRIAZO-GROUP. PART XI.very slowly, whilst stannous chloride in hydrochloric acid liberatesnitrogen immediately.The silver salt was precipitated by silver nitrate from aqueousammonium methyltriazomalonate, and, although colourless when fresh,rapidly darkened on attempting to recrystallise it from warm water ;a small quantity of the dried substance detonated with considerableviolence when thrown on a hot plate.Ethyl MethyZtriaxomaZonate, CH,* CN,( C0,- C,H,),.Fifty grams of methylbromomalonic ester in 30 C.C.of absolutealcohol were heated under reflux during ten hours with 22 grams ofsodium azide in 30 C.C. of water; action being then complete, aconsiderable quantity of water was added, and the precipitated oildistilled under diminished pressure, the principal fraction (32 grams)boiling at 69O/0*6 mm.:0.1328 gave 23.0 C.C. N, at 17" and 749 mm.Ethyl methyltriazomalonate is a colourless liquid having sp. gr.1.11695 at 16'/16O ; it has a faint, agreeable perfume, suggestingacetoacetic ester, and the vapour when inhaled with steam producesan effect on the blood-pressure similar t o that of the esters inthe monobasic series. As already indicated, cold concentratedpotassium hydroxide merely hydrolyses the ester, provided that thealkali is not in great excess and the mixture is not heated; evenon evaporating to dryness, the liberation of gas is very slight,and only a small proportion of nitrogen is removed in the formof hydrazoic acid.Action of SuZphuric Acid.-As in the case of triazoacetic ester, theinteraction with concentrated sulphuric acid is very slow, and only onheating the mixture during a considerable period did the volume ofliberated nitrogen reach the calculated amount :0.2243 gave 26.8 C.C.N, at 1 7 O and 747 mm.C,H,,O,N, requires 2/3N = 13.02 per cent.Behaviour towards Stannous Chloride.-A solution of stannouschloride in hydrochloric acid is without action on the ester untilthe temperature reaches about SO0, when a slow but regulareffervescence sets in, and is completed in about two hours,N = 19.80.C,H,,O,N, requires N = 19.53 per cent.N = 13.62.0,3038 gave 36.6 C.C. N, at 1 9 O and 742 mm. N= 13.51.C8H1,O,N3 requires 2/3N = 13.02 per centFORSTER AND MULLER: THE TRIAZO-GROUP.PART XI. 133Methyltriaxomnlonamide, CH,*CN,(CO*NH,),.Five grams of the ester were shaken during two hours with 15 C.C.of concentrated aqueous ammonia, excess of which was removedin the vacuum desiccator after twelve honrs had elapsed. Thecrystalline nmide W R S dissolved in boiling benzene, of which about400 C.C. mere required per gram, separating in long, lustrous needlesmelting at 13'7.5':0.1330 gave 52.5 C.C. N, a t 1 7 O and 742 mm.Met7L?lltriaxonzaZo~zum~de is readily soluble in warm water, alcohol,and petroleum, but is insoluble in cold benzene. A hydrochloricacid solution of stannous chloride attacks the substance rapidlywithout being heated, and liberates nitrogen. Concentratedsulphuric acid behaves in tt most unusual manner, nitrogen beingevolved only slowly even on raising the temperature to 125",below which there is not any effervescence.A parallel with thisremarkable behaviour has been noted quite recently in the case oftriphenylmethylazoimide (Wieland, Ber., 1909, 42, 3027), thesolution of which in concentrated sulphuric acid must be heatedt c 200' before gas evolution becomes vigorous.N = 64.71.C,€1702N, requires N = 44.58 per cent.Ethyltviaxomalonic Acid, C,H,*CN,(C0,H)2.Exactly the same procedure was adopted as in the case of methyl-triazomalonic acid, and, after crystallisation from hot benzene, theacid was obtained in colourless, rhombic, hygroscopic prisms, meltingand decomposing at 105-107° :0.1632 gave 34.4 C.C. N, at 22' and 766 mm.The acid is attacked readily by concentrated sulphuric acid, andN = 24.12.C,H70,N, requires N = 24.28 per cent.by stannous chloride in hydrochloric acid.Ethyl Ethyltriaxomalonute, C,H,*CN,(C0,*C2H,),.The ester was prepared from 42 grams of ethylbromomalonicester, 15 grams of sodium azide, and 45 C.C.of 50 per cent. alcohol,heating under reflux being continued during eight hours ; theproduct was fractionated under reduced pressure, boiling at83.5O/Oo7 mm. :0.1670 gave 2'7.3 C.C. N, at 22' and 769 mm.The colourless liquid has sp. gr. 1 el16 1 at 16"/16O, and the vapour,N = 18-74,C,HISO,N, requires N = 18.38 per cent134 FORSTER AND MeLLER: THE TRIAZO-GROUP. PART XI.although characterised by a pleasant odour, has the disagreeableeffect on the blood-pressure which has become associated with thealiphatic triazo-esters. Whilst the effervescence brought about bystannous chloride in hydrochloric acid is very vigorous, that inducedby concentrated sulphuric acid is very slow.A distinction from thelower homologue is offered by the behaviour towards concentratedpotassium hydroxide, because on heating the ester with excess of thisagent, violent liberation of nitrogen sets in, followed by ammonia, and,on cooling the liquid, there separate crystals of potassium propionyl-formate, containing 28.2 per cent. of potassium (C,H,O,K requiresK = 27.9 per cent.) ; the phenylhydrazone was prepared, and meltedat 148.5" after crystallisation from dilute alcohol.Ethyltriazomalonarnide, C,H5*CNg(CO*NH2),.The substance was obtained by shaking the ester with concentratedaqiieous ammonia, and crystallised from hot benzene in colourless,rhombic plates me1 ting at 167" :0.1027 gave 36.8 C.C.N, at 20° and 763 mm. N= 41.05.C5H,02N5 requires N = 40.93 per cent.Decomposition with stannous chloride in hydrochloric acid readily gavethe calculated amount of nitrogen, but the remarkable behaviour ofthe lower homologue towards concentrated sulphuric acid wasreproduced by etbyltriazomalonamide, which was not decomposedbelow 125".PhenyZtriazomaZotdc Acid, C,H,*CN,(CO,H),.As appears below, it was not found possible to distil phenyltriazo-malonic ester, even under pressure reduced to 0.56 mm., withoutdecomposition, which took place at 150°, and accordingly theundistilled material was employed as the source of the acid.Severalgrams of the ester were shaken with the calculated amount ofpotassium hydroxide in the form of a 20 per cent. aqueous solutionuntil, in the course of about three hours, a clear liquid resulted; thiswas extracted twice with ether, acidified with t,he calculated amountof dilute sulphuric acid, saturated yith solid ammonium sulphate, andfurther extracted five times with ether. After drying with ignitedfiodium sulphate, the residue from evaporation was submitted t odiminished pressure (20 mm.) during thirty-four hours, when itsolidified and became colourless on porous earthenware. Recrystal-lisation from hot benzene gave spherical clusters of snow-whiteneedles melting a t 99" without decomposition :0.1562 gave 25.8 C.C.N, at 18" and 768 mm. N= 19.28.C,H70,N, requires N = 19.00 per centFORSTER AND MULLER: THE TRIAZO-GROUP. PART XT. 135Phenyltriaxomalortic acid effervesces vigorously with concentratedsulphuric acid, while torrents of nitrogen are liberated by a solutioiiof stannous chloride in hydrochloric acid; 40 per cent. aqueouspotassium hydroxide effects immediate disruption of the triazo-groupin the cold, but alkali of half this concentration requires to be heatedbefore bringing about decomposition, when ammonia and nitrogen areliberated without formation of hydrazoic and hydrocyanic acids. Thealkaline liquid from the foregoing experiment gave an immediateprecipitate with phenylhydrazine after being neutralised with dilutesulphuric acid; this was found to be identical with the phenyl-hydrazone of benzoylformic acid obtained by similar procedure frompbenyltriazoacetic acid, whence it follows that the disruption of thetriazo-group is preceded by removal of carbon dioxide from thedibasic acid.Ethy I Phen?/ltriazomatonate, C,H,*CN,(C0,*C2H5)2.Phenylmalonic ethyl ester was prepared according to the method ofWislicenus (Ber., 1894, 27, 1093 ; Annalen, 1888, 246, 315), whichdepends on elimination of carbon monoxide from phenyloxalic esterobtained by the action of ethyl oxalate on the sodium derivativeof ethyl phenylacetate. W islicenus states that ethyl phenylmalonatetends to decompose when boiled under atmospheric pressure at 285',and therefore distilled it at 170-172'/14 mm.; our specimen boiledat 127-1 29O/0*35-0.4 mm. The bromination of phenylmalonic esterhas been described by Wheeler and Johnson (J. Amer. Chem. Soc.,1902, 24, 6$0), who carried out this operation in sealed tubes, but wefind that phenylbromomalonic ester may he prepared in almostquantitative yield by heating the ester with the halogen under ordinarypressure at 140-150' in bright daylight, the product distillingat 141-142"/0*48 mm.Phenylbromomalonic ester (1 8.5 grams), sodium azide (6 grams),and alcohol (10 c.c.), with sufficient water to maintain the saltdissolved, were left in darkness during three weeks with frequentshaking and occasional warming to about 40' ; water was then added,and the precipitated ester extracted and dried, but an attempt todistil the product was fruitless, owing to the decomposition which wasthreatened on raising the temperature of the bath to 150°, when thepressure rose from 0.56 mm.to 1.5 mm. quite suddenly, and theoperation was therefore discontinued. There was not any indicationof distillation taking place, and i t has not been possible, therefore, toobtain the substance in purified condition.The crude ester is a Flightly yellow, heavy oil with a faint, pleasantperfume ; the decomposition with concentrated sulphuric acid take136 FORS'J'ER AND MULLER: THE TRIAZO-GROUP. PART XI,place readily, and nitrogen is liberated also by a solution of stannouschloride in hydrochloric acid. The aotion of potassium hydroxide hasbeen already described.Phen yltriaxonanlonamide, C,H,- CN,( CO * NH,),.On continued shaking with strong aqueous ammonia, phenyltriazo-malonic ester was transformed into a crystalline solid, which dissolvedin hot water, and separated therefrom in snow-white, fern-like leaflets ;the substance melted a t 1890h:0.1038 gave 29.7 C.C.N, at 23O and 760 mm.C,H,O,N, requires N = 31.96 per cent,Phenyltriaxomalonfcmide is moderately soluble in acetone, ethylacetate, and ethyl alcohol ; it dissolves sparingly in boiling benzene,and is insoluble in boiling petroleum.The action of concentrated sulphuric acid on the amide is mildat first, and becomes brisk only on continued stirring, whilst a, solutionof stannous chloride in hydrochloric acid does not liberate nitrogenuntil a few drops of alcohol have been added to complete contact.When heated with 20 per cent, aqueous potassium hydroxide, thecompound evolves nitrogen freely, accompanied by ammonia, theodour of phenylcarbylamine being also noticeable ; hydrazoic andprussic acids, however, were not produced,N = 32.24.Ethyl Bistriaxomalonate, (N,),C(C0,*C2H5)2'Ethyl dichloromalonic ester was obtained as a by-product in thepreparation of ethyl chloromalonic ester on treating malonic ester withchlorine a t SOo, and boiled at 231-234O.It is noteworthy that,although ethyl dichloromalonate may be preserved during manymonths without showing any signs of having undergone change, themonochloro-compound (b, p. 222-223') became transformed into abrownish-grey, fuming liquid, having a marked odour of hydrogenchloride.Twenty grams of rectified dichloromalonic ester, 20 C.C.of absolutealcohol, and 14 grams of sodium azide dissolved in 40 C.C. of waterwere warmed carefully until the liquid became clear, and left in dark-ness during one month, with occasional shaking and gentle warming.On diluting with water and extracting with ether, a colourless oil wasobtained, half a gram of which was heated in an open tube underatmospheric pressure before submitting the whole specimen to distil-lation; no change was observed to take place while the temperatureremained below 175O, but a t the moment of reaching 180° an explosionof very considerable violence occurred. The main quantity of esteFORSTER AND MULLER: THE TRIAZO-GROUP.PART XI. 137was then distilled under 0.81 mm. pressure, boiling steadily at115-115.5':0.1576 gave 47-6 C.C. N, a t 21° and 764 mm.C7HIoO4N6 requires N = 34-71 per cent.Bistriaxontalonnic ester is a colourless oil with a pleasant perfume ; ithas sp. gr. 19136 at 20° compared with water at the same temperature.The decomposition by concentrated sulphuric acid is extremely violent,and the liberation OF nitrogen with stannous chloride in hydrochloricacid torrential :0.3251 gave with SnC1, 62.6 C.C. N, at 21' and 768 mm. N = 22.3.0.1816 ,, ,, H,S04 36.2 C.C. N, ,, 2 1 O ,, 764 mm. N=22*8.C7H,,04N, requires 2/3N = 23.14 per cent.An attempt to prepare bistriazomalonic acid by hydrolysing theester with 10 per cent.alkali was unsuccessful, owing to the readinesswith which the product undergoes spontaneous loss of hjdrazoic acid.Seven grams were shaken with the calculated amount of the agentuntil dissolved, extracted with ether, and treated with the exactquantity of dilute sulphuric acid, after which ammonium sulphate wasadded and the bistriazomalonic acid removed by four extractions withether. The solvent having been evaporated without heating, it wasnoticed that the residual oil, which remained viscous, acquired adistinct odour of hydrazoic acid, and was filled with bubbles whichceased to appear after many days in the desiccator :N = 34.57.0°1410 gave 43.4 C.C. N, at 23' and 757 mm. N = 34.6.C,H,O,N, requires N = 46.1 6 per cent.The latter formula represents triazoglycollic acid, which is verylikely produced by the removal of carbon dioxide and hydrazoic acidfrom bistriazomalonic acid under the influence of water.After manydays, crystals of oxalic acid were noticed in the oil, which was treatedwith a paste of barium carbonate in order to isolate, if possible,barium triazoglycollate. The aqueous filtrate from barium oxalateand unchanged barium carbonate was evaporated to dryness withoutheat, and triturated with absolute alcohol, the insoluble portion beingthen analysed :C,H,O,N, ,, N=35.8 ,,0*0786 gave 0.0498 BaSO,. Ba = 37.25..The substance effervesced vigorously with concentrated sulphuricC,H,0,N6 Ba requires Ba= 37.13 per cent.acid, and was most probably barium triazoglycollnte138 FORSTER AND MULLER: THE TRIAZO-GROUP.PART. XI.Bistriazomalonamide, (N,),C(CO*NH,),.Agitation with strong aqueous ammonia during one hour trans-formed bistriazomalonic ester into a snow-white, crystalline solid,which was recrystallised from boiling water, being only very sparinglysoluble therein ; the amide separated in transparent, colourless prisms,melting a t 162' with vigorous decomposition :0.0866 gave 47.6 C.C. N, at 2 4 O and 765 mm.C,H,O,N, requires N = 60.89 per cent.The substance is readily soluble in hot acetone and ethyl acetate,crystallisiag therefrom in six-sided plates ; it is moderately soluble incold alcohol, but dissolves very sparingly in boiling benzene andin boiling chloroform. Immediate effervescence occurs on mixing theamide with concentrated sulphuric acid, and becomes very vigorous onwarming ; stannous chloride also liberates nitrogen very freely.Thedecomposition with 40 per cent. potassium hydroxide is mostprofound, giving rise t o nitrogen, ammonia, prussic acid, and hydrazoicacid.Phen yltriaxoacet ic Acid, C,H,* CHN, C0,H.Owing to the readiness with which the triazo-group in phenyltriazo-acetic ester undergoes disruption in the presence of alkali, the acidcannot be obtained by hydrolysis. Five grams of phenylchloroaceticacid were therefore exactly neutralised with sodium carbonate in about30 C.C. of water, apd, after admixture with 2.5 grams of sodium azidedissolved in 20 C.C. of water, allowed to remain during two daysprotected from light ; dilute sulphuric acid having been added, and theliquid extracted six times with ether, the latter was dried with sodiumsulphate and evaporated in a vacuum desiccator without being heated.The residue became solid, and was recrystallised three times frombenzene, which deposited the substance in thin, colourless, rhombicplates melting a t 98.5O :N = 60.88.0.1543 gave 31.4 C.C. N, at 20° and 764 mm.C,H70,N, requires N = 23.73 per cent.The decomposition with concentrated sulphuric acid was very violent,and nitrogen was also liberated immediately on mixing the acid withaqueous potassium hydroxide or a solution of stannous chloride inhydrochloric acid.The silver salt could not be analysed, becausereduction took place on attempting to recrystallise it from warmwater.Ethyl Phenyltrhzoacetate, C6H,*CHN,* C02*C,H,.Early attempts to prepare this ester were conducted on the linesfollowed in the case of triazoacetic ester and its higher homologues,N = 23.41FORSTER AND M ~ L L E R : THE TRIAZO-GROUP.PART XI. 139but the results were unsatisfactory ; when alcoholic phenylchloro-acetic ester is heated under refliix with aqueous sodium azide, theliquid rapidly becomes yellow and evolves gas, and, although the oilyproduct of steam distillation answers to the azide test with con-centrated sulphuric acid, it is too complex R mixture to repay furthertreatment, the presence of by-products being best avoided by thefollowing procedure.Eighty grams of phenylchloroacetic ester (b.p. 135"/17 mm.)were mixed with 100 grams of alcohol and 40 grams of sodium azide(1 mol. = 26.2 grams), water and alcohol being then added alternatelyuntil both ester and salt were dissolved ; after remaining a few weeksin the dark, the liquid was found to have deposited a crop of sodiumchloride, and at the end of two months, when the action was judgedto have been complete, water and solid ammonium sulphate wereadded to precipitate the ester, which was then removed, dried in theusual way, and fractionated with the aid of the Gaede pump :0.2204 gave 38.3 C.C. N, a t 16' and 770 mm.C,,H,,O,N, requires N = 20.49 per cent.Ethyl phenpltriazoacetate is a colourless liquid with a very faintpleasant perfume suggesting roses; only on distilling the ester withsteam does the vapour produce the throbbing sensation at the base ofthe forehead and palpitation of the heart characteristic of thealiphatic triazo-esters.It boils at 93'/0-09 mm., and has sp. gr. 1.1434at 2Oo/2O0; the action with concentrated sulphuric acid is veryviolent.Action of Potassium Hydroside.-Contact with dilute aqueous alkaliwas found to bring about an immediate disruption of the triazo-group,and it was therefore useless t o attempt hydrolysis by this means. Onadding 40 per cent. potassium hydroxide drop by drop to 10 grams ofthe ester dissolved in 40 C.C. of alcohol, the liberation of nitrogen witsperfectly regular, and, after excess of alkali had been added, thereseparated a reddish oil which in the vacuum-desiccator rapidlysolidified ; this product, being sparingly soluble in water, was recrps-tallised from the gently warmed liquid, and was obtained in colourless,nacreous plates :N = 20*08.0.3114 gave 0.1470 K,SO,.K=21*16.0.2577 ,, 16.7 C.G. N, at 19' and 768 mm. N=7.53.C:,H,O,NK requires K = 20.85 ; N = 7.49 per cent.I n the course of twenty-four hours' exposure to air, the substancehad begun to undergo hydrolysis, yielding potassium benzoylformate,and when the salt no longer contained nitrogen, benzoylformic acidwas obtained from it and identified by the melting point ( 6 5 O ) and byconversion into the phenylhydrazone, a specimen of which melted a140 FORSTER AND M~LLER: THE TRIAZO-GROUP. PART XI.160" and contained 11.8 per cent.of nitrogen (C,,H,202N, requiresN = 11 *7 per cent,).Action of Ammonia.-Attempts t o prepare phenyltriazoacetnmideby the action of ammonia on phenyltriazoacetic ester were un-successful, owing to disruption of the triazo-group ; this can, however,be so controlled as to permit the isolation of phenyliminoacetamide.Excess of dry ammonia was passed into a well-cooled solution ofphenyltriazoacetic ester in absolute alcohol ; liberation of nitrogenaccompanied the separation of crystals, which increased during twenty-four hours in t.he ice-chest. Recrystallisation from .benzene, repeateduntil the melting point was constant, gave colourless, monoclinicplates melting a t 144' :0.2070 gave 34.3 C.C. N, a t 20" and 770 mm.C,H,ON, requires N = 18.92 per cent.The substance is readily soluble in alcohol and in benzene, beingprecipitated from the respective solutions by water and by petroleum.When exposed to air it produces ammonia, giving the a-amide ofbenzoylformic acid, m.p. 90'. The acid itself mas obtained bycompleting the hydrolysis.N = 19.21.Trriaxoacetophenone (Pkenacylazoimide), N3*CH2*CO*C6H,.Twenty grams of bromoacetophenone dissolved in 50 C.C. ofabsolute alcohol and mixed with 7 grams of sodium azide in theminimum quantity of water were shaken at intervals during sixteenhours in the ice-chest; being very readily fusible, the crystallineproduct was collected on a filter, cooled with chilled brine, and thenprecipitated by petroleum from a dried ethereal solution :0.1695 gave 38.4 C.C.N, at Z O O and 767 mm.C1,H70N3 requires N = 26.09 per cent.The triazoketone crystallises in colourless, lustrous plated, and meltsat 17O; in the course of a few weeks the pale yellow substance becamedark brown, even when protected from light, and the odour ofhydrazoic acid mas perceptible. An attempt t o distil it under 0.1 mm.pressure was unsuccessful, because a threatening decomposition set inwhen the temperature of the bath had reached 130°, the pressurequickly rising to 4 mm.On adding 20 per cent. potassium hydroxide to an alcoholic solutionof triazoacetophenone, immediate liberation of nitrogen takes placefollowed by ammonia on heating the liquid; a deep red coloration isdeveloped, but the indication of hydrazoic acid is trifling.Anattempt to establish the intermediate production of benzoylform-aldehyde on the lines indicated by W. L. Evans (J. Arner. Chem. Soc.,1906, 34, 115) was unsuccessful, the only material isolated beingN=26*16FORSTER AND MULLER: THE TRIAZO-GROUP. PART XI. 141benzoic acid. Two grams in ether were shaken with one gram oflime and 40 C.C. of water, the mixture liberating gas immediately andbecoming dark brownish-red ; after thirty hours the aqueous liquid wasremoved, extracted three times with ether, and acidified, when furtherextraction removed benzoic, not mandelic, acid.Phenacylazoimide reduces ammoniacal silver oxide in the coldsolution.Action of Sulphuric Acid.-The concentrated agent attacks thetriazoketone with great violence, and flame sometimes accompaniesthe small explosion which occurs when they are mixed drop by drop.By using more dilute acid the action can be moderated, until at25 per cent.the yield of nitrogen may be measured :0.2336 gave 34.4 C.C. N, at 18' and 772 mm.C,H70N, requires 2/3N = 17.39 per cent.To complete the decomposition it was found' necessary to use a bathof hot brine, and on cooling the contents of the flask, crystals ofbenzoic acid separated.Behaviour towards Stannous Chloride.-With a cdld 20 per cent.solution of the matal in hydrochloric acid the decomposition was veryslow, and, even after heating during two hours, the quantity of gas wasdeficient :N=17*27.0.1586 gave 22.9 C.C. N, at 18' and 772 mm. N= 16.95.C,H70N, requires 2/3N = 17.39 per cent.The aromatic product was not identified.The Semicarbaxone.-An alcoholic solution of the triazoketone wasmixed with aqueous semicarbazide acetate, the solid which separatedduring twelve hours being recrystallised from warm dilute alcohol :0.0895 gave 29.6 C.C. N, a t 19' and 773 mm.C,H,,ON, requires N = 38.52 per cent.The substance crystallised in colourless, silky needles, melting at127n5-128.50. The reaction with concentrated sulphuric acid wasvery vigorous.The Bromophenyl?i,yhaxone.-The derivative crgstallised from dilutealcohol in lustrous, yellow needles, melting at 114.5' :0.1079 gave 20.1 C.C. N, at. 19' and 768 mm.N = 38.68.N = 21.64.C,,H,,N,Br requires N = 21 -2 1 per cent.Frinxoacet op henoneoxi me, N, C H, C ( : NOH). C6H5,Triazoacetophenone was suspended in an aqueous solution ofhydroxylamine prepared by exactly neutralising with sodium carbonatea slight excess of the hydrochloride; alcohol was then added until,with gentle warming, a clear solution was produced. After thre142 CHALLENGER AND KIPPlNG :days in the dark, a pale yellow oil had separated, and this, afteradding water, was extracted with ether, dried with sodium sulphate,and freed from solvent in the vacuum desiccator. As in the case oftriazoacetone (Trans., 1908, 93, 84), i t was not found possible tocrystallise the oxime, which mas therefore analysed in liquid form :0.1050 gave 28.4 C.C. N, at 1 8 O and 771 mm. N = 31-70.C,H,ON, requires N = 31.82 per cent.An attempt to produce the p-toluenesulphonyl derivative, whichserved to characterise triazoacetoxime (Zoc. cit.), led to an uninvitingblack tar.ROYAL COLLEGE OF SCIEKCE, LONDON.SOUTH KENSINGTON, S.W