2570 FORSTER AND NEWMAN: THE TRIAZO-GROUP. PART xv.CCLXI 11.-The Trinzo-group. P a r t X V. Triaxo-ethylene ( Vinylazoirnide) and the TriazoethylHalides.By MARTIN ONSLOW FORSTER and SIDNEY HERBERT NEWMAN.MANY cases have now been placed on record in which the atomscomposing a triazo-group in the neighbourhood of an unsaturatedlinking have lent themselves to the formation of a cycloid, owingto a change involving saturation of this linking by the membersof the azoimide nucleus. As recent examples of this principle maybe quoted the spontaneous transformation of allylazoimide into anisomeric diazoamino-compound (Trans., 1908, 93, 1174), whilst anattempt to prepare benzhydroximic azide by interaction of thechloride and sodium azide led t o 1-hydroxy-5-phenyltetrazole (Trans.,1909, 95, 183; compare also Wieland, Ber., 1909, 42, 4199).I n the cases mentioned above, and in others which might beindicated, the environment favourable to the display of additivefunctions on the part of the triazo-group is intramolecular.Whether it is necessarily so cannot be stated with confidence,because, although striking examples of apparently intermolecularaddition have been furnished quite recently by Palazzo and Oliveri-Mandal2 ( A t t i R.Accad. Lincei, 1910, [v], 19, i, 218, 228), whofound that hydrazoic acid converts fulminic acid and methyl-carbylamine into 1-hydroxytetrazole and 1-methyltetrazole respec-tively, whilst Dimroth and Fester (Ber., 1910,43,2219) showed thattetrazole itself is obtainable by the interaction of hydrazoic andprussic acids, it is nevertheless possible that these changes actuallydepend on preliminary addition of H and N, t o unsaturated atoms,and that it is the resulting triazo-compound which undergoes intra,molecular rearrangement, as follows FORSTER ANI) NEWhIBN : T H E TRIAZO-GROUP. PAKT XV.2571I n addition to the experiment just mentioned, Dimroth andPester caused hydrazoic acid to act on acetylene, and although1: 2 : 3-triazole was obtained therefrom, they do not classify thischange with the foregoing ones, as depending on the intermediateformation of vinylazoimide,N-YHCH:CH’ HCiCH+HN,=N,*CH:CH, -+ N<because phenylazoimide is converted into 1-phenyl-2 : 3-triazole bythe action of acetylene. As we have been engaged for some timein experiments lea,ding to the preparation of vinylazoimide, wethink it may be of interest to put forward a description of thissubstance, which was isolated with the object of studying itstendency to undergo the above reari angement.The first experiments in the direction indicated were begun morethan t’wo years ago in continuation of the work on allylazoimide.Attempts were made to bring vinyl bromide into double decom-position with sodium azide, but there was not any evidence of thechange having taken place; in spite of the readiness with whichusually the triazo-group may be substituted for halogen in aliphaticcompounds, the failure was not surprising, as it is well known thathalogen attached to unsaturated carbon takes part in such reactionswith the greatest difficulty.The preparation of triazoethyl alcohol(Trans., 1908, 93, 1865), however, suggested the possibility ofarriving at triazoethylene (vinylazoimide) by the withdrawal ofhalogen hydride from a triazoethyl halide. Accordingly, triazo-ethyl alcohol was converted into triazoethyl bromide by the actionof phosphorus tribromide, and then by double decomposition withsodium iodide, triazoethyl iodide was prepared from the bromide;on acting with alcoholic potassium hydroxide on these compounds,they were readily deprived of halogen hydride without affecting thetriazegroup.Triazoethylene is a pale yellow liquid, heavier than water, boilingat 26O, and having an odour suggesting that of ethylene itself.Itdecolorises bromine water without delay, forming a heavy, oilydibromide. At one time we entertained the hope that it might bepossible to withdraw the elements of hydrogen bromide from thissubstance in such a way as to produce triazoacetylene, the coppe25'72 FOKSTER ANL) NEWMAN: TliE TRIAZO-GROUP. PART XV.derivative of which would probably rank among violent explosives,but i t was soon found that the dibromide itself is a dangerousmaterial, and, moreover, breaks up in an interesting manner underthe influence of water. When suspended therein, the oil rapidlydisappears, hydrobromic and hydrazoic acids being set free, whilstbromoacetaldehyde remains dissolved :CH,Br*CHBr*N, + 2H,O = HBr + HN, + CH2Br-CH(OH), -3CH,Br*CH:O + H,O.So far as we know, the dibromide of triazoethylene is the firstcompound in which a triazo-group is attached to an atom of carbonwhich carries also a hdogen, and its peculiar susceptibility towardswater explains the subsidiary decomposition which has always beenobserved to accompany the production of a bistriazo-compound fromthe double decomposition of sodium azide and a.dihalogen derivativeof the aliphatic series in which both halogen atoms are attached tothe same atom of carbon (Trans., 1908, 93, 1070; also this vol.,pp. 126 and 1360). We have not att,empted to isolate and distilthe substance, because a brief experience of aa-bistriazwthane,which exploded with great violence at the laboratory temperature,apparently spontaneously, renders it most probable that ab-dibromo-a-triazoethane would behave in the same way.This seems the morelikely to take place in view of the possibility that the decompositionof triazoethylene dibromide might follow a different course inabsence of water, leading to bromoazoimide :CH2Br*CHBr*N3 -+ BrN3 + CH,Br*CH -+ CHBr:CH,.Since Raschig has shown (Ber., 1908, 41, 4194) that chloro-azoi.mide is frightfully explosive, the foregoing possibility is dis-couraging to a further investigation of aP-dibromo-a-triazoethane.The original purpose with which triazoethylene was prepared,namely, to illustrate its transformation into triazole :has not been fulfilled, the substance having shown itself to besurprisingly stable. After being heated with dilute alcohol underreflux during twelve hours, a considerable proportion was foundto have survived, and although the remainder had changed into adark brown substance, the presence of triazole could not beestablished. Furthermore, on heating an alcoholic solution in asealed tube during twelve hours at 100-llOo, there was produceda dark brown liquid having the odour of a carbylamine, and givinga precipitate with mercuric chloride, but again it was not foundpossible to recognise triazole.The pale yellow colour of triazoethylene is an interesting featureof the compound, and appears to be a genuine property, becausFORSTER AND NEWMAN : THE TRIAZO-GROUP.PART XV. 2573it has been noticed in every specimen, whether prepared from triazeethyl iodide or bromide. It is doubtless due to the concentrationof unsaturated atoms in a small molecule, and although not SOintense as that of diazomethane, it is nevertheless quite distinctive.The comparatively high boiling point was not expected, because therecorded boiling points of vinyl chloride and of vinyl bromide are- 1 5 O and + 1 6 O respectively, and without having made a study ofthe subject, we were under the impression that, whilst the boilingpoint of an aliphatic chlorederivative is unquestionably lower thanthat of the corresponding triazo-compound, the latter would bemore volatile than the brominated substance.As it is often amatter of considerable practical importance to be able to predict,roughly, the boiling point of a new triazo-compound from that ofits haloid parent, we have taken afresh the boiling point of severaltypical aliphatic azoimides described in previous papers, side byside with those of the corresponding bromclderivatives.The resultsshow conclusively that the triazo-group exerts an elevating effecton the boiling point greater than that of the bromine atom.Vinyl bromide .................Vin ylazoimide .....................Ethylene dibromide ............Triazoethyl bromide i.. ..........Ethyl a-bromopropionate ......,, a-triazopropionate ......,, a-bromoisobutyrate ......, , a- triazoisobutyrate ......,, a-bromoisovalerate ......, , a- triazoisoval era te ......B. p. Mni.CH,:CHP,r ..................... 16" 750CH,:CHN,, ................... 26 760CH,Br'CH,Rr ..................37 20CH,Br'CH,N, .................. 49 20CH,.CHBr.CO,Et ........... 61 16CH,*CHN;CO,Et ............ 70 16CBr(CH,);CO,Et ............ 62 16CH(CH,),*CHBr'CO,Et ...... 79 16CH(CH,),*CHN,'CO,Et ...... 82 16CN3(CH3);C0,Et ............ 71 16The relationship between the triazo-group and the chlorine atomby the as regards their effect on the boiling point is revealedfollowing data :B. p. Mm.Vinyl chloride .............................. CH,:CHCl .................... - 15" 760Vinylazoimide .......................... CH,:CHN,. .................. 26 7608-Chlorocthyl alcohol.. .................. CH,CI*CH;OH .............. 44 20B-Triazoethyl ,, ..................... CH,N;CH,'OH ............ 73 20Ethyl chloroformate .................... Cl'C0,Et .....................93 760,, triazoacetate ....................... CH,N,*CO,Et ............... 70 20Methyl a-chloromethylacetoacetate .. , CH;CO'CCl(CH,)*CO,Me . 76 13 ,, a-triazomethylacetoacetate ... CH;CO*CN,( CH,)*CO,Me . 76 0 '66We have not, as yet, been very successful in attempts to utilisetriazoethyl iodide as a triazo-ethylating agent. Triazoethyl chloridehas been obtained by heating the iodide with dry mercuric chloride,and in this connexion it is worth noting that the bromide remainsunchanged when treated with the mercury salt under similar con-ditions. isoNitrosocamphor was transformed into an oily triazo-,, triazoforamte ..................... N;CO,Et ..................... 114 769,, chloroacetatc ........................ CH,Cl'CO,Et ..................52 22574 FORSTER AND NEWMAN: THE TRIAZO-GROUP. PART XV.ethyl derivative when heated with triazoethyl iodide and silver oxidein dry ether, but an attempt to prepare bistriazoethyl ether byheating a mixture of triazoethyl alcohol and iodide with dry silveroxide on the water-bath led to the recovery of unaltered triazoethylalcohol after five days. This is the more curious because the actionof ethyl iodide on triazoethyl alcohol in presence of silver oxideleads without difficulty to triazoethyl ether, C,H,*O*CH,-CH,N,.Lastly, we have not been able to produce triazoethylcarbimide bythe action of triazoethyl iodide on lead o r silver cyanate, whilst theinteraction of triazoethyl iodide and silver cyanide, although pro-ducing an odour of carbylamine, was too far from complete toadmit of isolating any definite products.EXPERIMENTAL.Triasoethylene (Vinylaaoimide), C'H,:CHNs.The alcoholic potassium hydroxide used for withdrawing halogenhydride from the triazoethyl halides was a solution of 5 grams in20 grams of water, mixed with 25 grams of absolute alcohol.Thiswas heated on steam in a generating flask, into which was fitteda dropping-funnel and a reflux double-surface condenser ; from thelatter, connexion was made to a small, dry flask, thence to aU- tube filled with calcium chloride, and finally to a large test-tubesurrounded by a freezing-mixture. The temperature of the waterin the condenser having been adjusted to 30°, 5 grams of triazoethyliodide was admitted drop by drop into the alcoholic potassiumhydroxide at the temperature of boiling alcohol, when a pale yellowliquid gradually accumulated in the small flask between thegenerator and the U-tube.All the iodide having been added, theliquid was boiled during fifteen minutes, when i t was found thatthe cooled tube also contained some yellow liquid, and this, beingpresumably free from alcohol and water, was regarded as puretriazoethylene, and found to boil at 26O/760 mm. Although thisexperiment has been made several times, and triazoethyl bromidehas been substituted for the iodide, it has always been observedthat the liquid in both condensing vessels is pale yellow, and thatthe colour does not vary in depth; it therefore seems safe to concludethat this feature is not due to some impurity.Reference has beenmade to the absence of any conclusive evidence that triazoethylenechanges into triazole; in addition to the experiments towards thisend which have been already described, a solution of triazoethylenein petroleum was left in a, stoppered vessel exposed to light duringmany days without giving the faintest indication of triazole; onallowing the solvent to evaporate, the odour of a carbylamine wasnoticeableFORSTEH, AND NEWMAN: THE TRIAZO-GROUP. PART xv. 2575The dibrom,ide of triazoethylene was prepared by adding ice-coldbromine water to a well-cooled suspension of the substance in water,the colour of the halogen being immediately destroyed, whilst thelimpid vinylazoimide changed to a heavy, viscous oil.It is necessaryto be most cautious in adding the halogen, because on one occasiona drop of bromine was admitted by accident. to the vessel containingthe triazo-compound, and led to a violent explosion, although theamount of material involved could not have exceeded 1 gram, andthis was diluted with 20 C.C. of water. On allowing the dibromideto remain in contact with water, it rapidly disappeared, and theliquid was found t o contain hydrazoic and hydrobromic acids;moreover, it restored the colour to Schiff's reagent, and when mixedwith ammoniacal silver oxide and filtered, the liquid quicklydeposited silver on warming. In order to make sure that the decom-position of triazoethylene dibromide by water does not follow thepossible alternative course, that, namely, leading to bromoazoimideand vinyl bromide, a specimen of vinylazoimide was converted intothe dibromide with a deficit of bromine, and at once treated withdilute sodium hydroxide, in the expectation that if bromoazoimideis formed, it would behave towards alkali in the manner thatcharacterises chloroazoimide, and that alkali hypobromite would beproduced; we were able to show that hypobromite is not formed,and therefore conclude that the decomposition proceeds only in thedirection of hydrobromic and hydrazoic acids along with bromo-acetaldehyde.The interaction of triazoethylene and concentrated sulphuric acidis mild, gas being evolved slowly, but brisk effervescence occurs witha solution of stannous chloride in hydrochloric acid.The con-clusion that alcoholic potassium hydroxide is without action on thesubstance may be drawn from the fact that on evaporating todryness the liquid contained in the generating flask, no trace ofalkali wide was to be found.8-Chloro-a-triazo e t han e (Trbzo e t h yl Chloride), N,*C%, CH,Cl.The first attempt's to prepare this material were made by addingtriazoethyl alcohol dissolved in absolute ether to the calculatedamount of phosphorus pentachloride covered with the same solvent ;considerable action took place, and was increased by heating underreflux, but the yield of triazoethyl chloride, being only 2 gramsfrom 20 grams of the alcohol, was too disappointing to encouragethe adoption of this method.Thionyl chloride acts vigorously ontriazoethyl alcohol, but, as might be expected, gives a product whichappears to be triazoethyl sulphite, and triazoethyl chloride couldnot be detected, whilst the effect of passing dry hydrogen chlorid2576 FORYTER AND NEWMAN : THE THIAZO-GROUP. PART XV.into a suspension of anhydrous zinc chloride in triazoethyl alcoholwas to liberate hydrazoic acid. It was not until triazoethyl iodidebecame available that the preparation of the chloride was possible.Twenty grams of triazoethyl iodide, mixed with 32 grams ofdried mercuric chloride, were heated at looo during three hours ina small distilling flask, from which the product was then boiledunder 25 mm. pressure, 10 grams, or 90 per cent. of the theoreticalamount, being obtained ; on re-distillation under the same pressure,the substance boiled steadily at 45O:0.0874 gave 30.9 C.C.N, at 2 4 O and 751 mm.0.2709 ,, 0.3642 AgC1. C1=33.25.N = 39.16.C,H,N3C1 requires N = 39.81 ; C! = 33.65 per cent.The substance is limpid and colourless, having a pleasant odoursuggesting that of chloroform ; the density is 1.2885 / 24O. Theaction with a solution of stannous chloride in hydrochloric acid issluggish, gas being evolved only on warming the liquids; theeffervescence with concentrated sulphuric acid becomes brisk onstirring, but the triazo-group appears to be indifferent towardsalkalis, which only liberate triazoethylene. The chloride does notlose its halogen completely when heated with boiling alcoholic silvernitrate, as is the case with the other tqiazoethyl halides, and theabove estimation of chlorine was made by heating in alcohol with30 per cent.aqueous potassium hydroxide, followed by precipitationwith silver nitrate in the solution acidified by nitric acid.When thrown on a hot plate, the substance decrepitates, andburns with a violet flame.P-Bromo-a-t riasoet Tmne (T?.z'uzoe thy1 Bromide) , N3* CH,*CH2Br.Although triazoethyl bromide has been obtained by the action ofbromine on tria.zoethy1 alcohol in presence of amorphous phosphorus,this is not the most convenient method of preparation, the inter-action of phosphorus tribromide and the alcohol, when moderatedby a diluent such as ether or petroleum, leading to more satisfactoryresults.One hundred grams of triazoethyl alcoho,l, covered with100 C.C. of petroleurn (b. p. 40°) in a flask surrounded with meltingice, were treated slowly with 112 grams of phosphorus tribromidein 250 C.C. of the same petroleum, the mixture being shakenvigorously after each portion was added ; phosphorous acidseparated, and a heavy, pale brown, viscous oil, insoluble inpetroleum, constituted a large proportion of the product. Afterthree hours on the water-bath under reflnx, the less dense liquidwas decanted, and the viscous residue shaken several times withsmall quantities of petroleum, which were added to the decantedsolution of triazoethyl bromide ; any excess of phosphorus tribromidFORSTEK AND NEWMAN : THE TRIAZO-GROUP. PART.xv. 2577was destroyed by agitation with small quantities of water, and theliquid having been treated with ignited sodium sulphate, petroleumwas boiled fiway, and the residue distilled under diminishedpressure. The best yield obtained by this process was only 58 grams,representing about 34 per cent. of the amount anticipated :N=28*00. 0-0641 gave 15.9 C.C. N, at 24O and 765 mni.0.4942 ,, 0.6198 AgBr. Br =53*38.C,H,N3Br requires N= 28-02 ; Br = 53.33 per cent.Triazoethyl bromide is a colourless liquid, having the odour ofethylene dibromide, and rapidly becoming yellow when exposed tolight; i t boils at 49O/20 mm., and has the density 1.6675/19'.Action with concentrated sulphuric acid and with a solution ofstannous chloride in hydrochloric acid resembles that of the chloro-compound, but hot alcoholic silver nitrate leads more readily tothe elimination of halogen tha.n in the case of that substance, andthe above determination of bromine was carried out by this agent.The bromide does not become ignited when thrown on a hot plate,merely decrepitating mildly.Attempts have been made t o identify the viscous, brown oil whichaccompanies triazoethyl bromide when prepared by the foregoingmethod, so far without success.It is the production of this sub-stance which is responsible for the disappointing yield, and is par-ticularly inconvenient because triazoethyl bromide is the startingmaterial for the chloride and the iodide. The presence of phos-phorus, bromine, and the triazo-group suggested that the sub-stance might be the bromide of bistriazoethylphosphorous acid,(N3*CH2*CH2dO),PBr, but the bromine content was much too low;it may be a mixture of this substance with triazoethyl phosphite,(N3-CH2*C'H,*~O),P, but an attempt to recover triazoethyl alcoholfrom itl by hydrolysis was not successful.P-lodo-a-t?-iazoethane (Triuzoethyl Iodide), N,-CH,*CH,I.The method employed for this preparation was the one recentlydescribed by Finkelstein (Bey., 1910, 43, 1528), and was found tobe expeditious and economical.Fifty-eight grams of triazoethylbromide were added to a solution of 60 grams of sodium iodide in400 C.C. of dry acetone, sodium bromide being precipitated imme-diately; the mixture having remained at the ordinary temperatureduring the night, action mas completed by heating under reflux,when about two-thirds of the solvent mas distilled off, and theresidue poured into water contained in a separating funnel, fromwhich the heavy, dark brown liquid was then tapped.This wasmixed with the ether used for extracting the triazoethyl iodide fromthe aqueous acetone, and shaken vigorously with a little mercur2578 FORSTER AND NEWMAN: THE TRIAZO-GROUP. PART xv.in order to remove dissolved iodine, the residue from the driedether being then distilled under diminished pressure, yielding 43grams :0.0814 gave 15.5 C.C. N, a t 23O and 764 mm. N=21.58.0.4270 ,, 0.5063 AgI. 1=64*12.C&14N31 requires N = 21-32 ; I == 64.45 per cent.Triazoethyl iodide boils at 68O/20 mm., and when freshly distilledis colourless, but quickly becomes pale red; the odour resemblesexactly that of ethyl iodide.It has the density 1*9154/25O, andis able to dissolve mercuric iodide, a property brought to light bythe fact that a specimen which had been decolorised by agitationwith mercury left a considerable residue of the salt on redistillation.The action with stannous chloride in hydrochloric acid is morebrisk than in the case of the other triazoethyl halides, which theiodide resembles, however, in regard to interaction with concentratedsulphuric acid. Behaviour on the hot plate is similar to that oftriazoethyI chloride.It was hoped that a variety of interesting substances might beobtainable from typical compounds containing replaceable hydrogenby triazo-ethylation, but hitherto we have not been successful inthis direction. p-Nitrophenol, for instance, when heated in drybenzene with silver oxide and triazoethyl iodide, gave a brown oilwhich did not invite further examination. isoNitrosocamphor alsogave an oil, remaining liquid during four months, and containing22.35 per cent.of nitrogen (C,,H,,O,N, requires N = 22.4 per cent.).Silver and lead cyanates were heated in ether and in benzene atthe boiling points of these with triazoethyl iodide during manyhours, but triazoethylcarbimide could not be recognised, althoughin absence of a diluent, some action takes place at about looo, asindicated by a mild explosion which occurred. Silver cyanidedeveloped the carbylamine odour when heated with triazoethyliodide during two days on the water-bath, but the proportion ofmaterial remaining unchanged at the end of the experiment wastoo large to hold out any prospect of success.Bistriazoethyl sulphateappears to be formed when triazoethyl iodide is heated in drybenzene with silver sulphate, production of silver iodide beingclearly indicated; the residue left by the solvent on evaporationdid not distil at 140°/1 mm., but .when hydrolysed with 30 percent. potassium hydroxide, the liquid contained potassium sulphate,unmixed with iodide.An attempt was made to prepare bistriazoethyl ether by heating5 grams of triazoethyl alcohol and 11.3 grams of triazoethyl iodidewith 15 grams of dry silver oxide on the water-bath during fivedays, but the entire product distilled at 85O/35 mm., weigheCONDENSATION OF AROMATIC SULPHINIC ACIDS. 25795 grams, and contained 48.3 per cent. of nitrogen; this is theamount required by triazoethyl alcohol itself, whilst bistriazoethylether contains 53.8 per cent., from which it would appear that inthe above experiment the triazoethyl alcohol remained unchanged,whilst the triazoethyl iodide wils transformed into triazoethyleneand diffused out of the apparatus.P-Triaaoethyl Ether, N,*CH,*CH,*O*C2H,.Twenty grams of triazoethyl alcohol and 50 grams of ethyl iodidewere allowed to remain in darkness with 50 grams of dry silveroxide during two days, being then heated on the water-bath withoccasional addition of small quantities of ethyl iodide. After oneweek, the liquid was separated and distilled under diminishedpressure :0.1057 gave 33.5 C.C. N, at 2 l 0 and 763 mm.The substance is a colourless liquid, boiling at 49O/25 mm., andhaving the density 0-9744/24O. The odour resembles that of chloro-ether, and in steam is pungent and sweet. With concentratedsulphuric acid or a solution of stannous chloride in hydrochloricacid, there is a vigorous effervescence, but hot concentrated alcoholicpotassium hydroxide appears to be without action on triazoether,the azoimide nucleus remaining intact. Triazoethyl ether does notexplode when thrown on a hot iron plate. the vapour burning with aluminous, white flame.N=36*39.C,H,ON, requires N = 36.51 per cent.ROYAL COLLEGE OF SCIENCE, LONDON.SOUTH KENSINGTOE, S . W