JOURNALOFTHE CHEMICAL SOCIETYPERKIN TRANSACTIONS IOrganic and Bio-organic ChemistryReactive Intermediates. Part XXIV. 1 H-Azirine Intermediates in thePyrolysis of 1 H-l,2,3-TriazolesBy Thomas L. Gilchrist,' Geoffrey E. Gymer, and Charles W. Rees," The Robert Robinson Laboratories,Nitrogen has been extruded from several 1,2,3-triazoles by flash vacuum pyrolysis and the fate of the resultingiminocarbenes has been determined. 1 -Alkyl-4,5-diphenyl-1,2,3-triazoles (1 ) gave nitriles (Scheme I ) andisoquinolines and hydroxyisoquinolines (Scheme 2). the former by Wolff rearrangement and the latter by 1.4-hydrogen transfer in the iminocarbene. 1,4-Dimethyl-5-phenyl-1,2,3-triazole (3) and 1.5-dimethyl-4-phenyl-1.2.3-triazole (4) both gave 3-methylisoquinoline ; a mechanism involving a common.1 H-azirine intermediateis proposed (Scheme 4). From both 4- and 5-phenyl-l-(1 -phenylvinyl)-I ,2,3-triazole [(5a) and (6a)l mixturesof 2.4- and 2.5-diphenylpyrrole were isolated ; similarly the corresponding 1 -phenyltriazoles gave mixtures of2- and 3-phenylindole (Scheme 5). 4.5-Diphenyl-v-triazole also gave 2-phenylindole. The formation of theseproducts is rationalised in terms of 1 H-azirine intermediates in the pyrolyses.University of Liverpool, P.O. Box 147, Liverpool L69 3BXTHE status of the unsaturated three-membered hetero-cyclic systems oxiren, thiiren, and 1H-azirine as viablereaction intermediates has recently been the subject ofconsiderable experimental and theoretical work. Thesecompounds can formally be classed as antiaromatic,since they contain four x-electrons in a cyclic system.No derivative of these heterocyclic systems has so farbeen isolated or detected directly, although there is nowconsiderable evidence that oxirens are intermediates inboth photochemical and thermal decompositions ofdiazo-ketones3In an earlier paper we produced evidence for the inter-mediacy of 1H-azirines in the thermal decomposition ofl-phthalimid0-1,2,3-triazoles.~ We now describe theresults of a much more general study of the vacuumpyrolysis of 1,2,3-triazoles, which was principallydesigned to discover whether lH-azirines are involved asintermediates in all cases, or whether the mechanism is aspecial feature of the chemistry of 1-phthalimidotriazolesand related compounds.One problem which arose atthe outset was that, unlike diazo-ketones, very little workhas been done on the thermal or photochemical decom-position of 1,2,3-triazoles. In many cases it was neces-sary to make a preliminary study of the products formedfrom model systems in order to discern the major reactionpathways. l-Alkyl-l,2,3-triazoles have been investi-Part XXIII, T. L. Gilchrist, G. E. Gymer, and C. W. Rees,J.C.S. Perktn I, 1973, 665.Preliminary communication, T. L. Gilchrist, G. E. Gymer,and C. W. Rces, J.C.S. Chem. Comm.. 1973, 835.gated most thoroughly, and representative l-vinyl- andl-phenyl-l,2,3-triazoles have also been prepared andpyrolysed ,1-AZkyZ-I ,2,3-triazoZes.-No compounds of this typehave previously been pyrolysed or photolysed.A seriesof l-alkyl-4,5-diphenyl-1,2,3-triazoles (1) was prepared,either by alkylation of 4,5-diphenyl-v-triazole [com-pounds (la-c)] or by cycloaddition of the alkyl azide todiphenylacetylene [compounds ( l a and d)] . Althoughalkylation gave mixtures of the 1- and the 2-alkyltriazoles(2), it was possible to obtain the l-alkyl derivatives inR111 a; R=Me (2) a; R=Meb; R=Et b; R = E tc; R-Pri C; R=Pr;d; R=CH2Phfair yields by the reaction of the silver or thallium salt of4,5-&phenyltriazole with the appropriate iodoalkane.The l-alkyltriazoles all show an ion (M' - 28) in themass spectrum corresponding to the loss of nitrogen from8 J. Fenwick, G. Frater, K. Ogi, and 0. P. Strausz, J . Amer.Chem.Soc., 1973, 95, 124; P. W. Concannon and J. Ciabattoni,ibid., p. 3284; S. A. Matlin and P. G. Sammes, J.C.S. Pevkin I,1972, 2623; Y. Ogata, Y. Sawaki, and H. Inoue, J. Oyg. Chem.,1973, 38, 10442 J.C.S. Perkin Ithe parent ion, but this is absent from the spectra of the2-alkyltriazoles.Four products were isolated, in good overall yield,from the pyrolysis of l-methyl-4,5-diphenyltriazole.The major product (50-607(0 based on the triazole) wasidentified as 2,2-diphenylpropiononitrile by comparisonwith an authentic specimen. Two minor products wereidentified as 3-phenylisoquinoline (19%) and 4-hydroxy-3-phenylisoquinoline (12%) ; in each case the materialisolated was compared with a specimen prepared by astandard route. Tetraphenylsuccinonitrile (5%) wasdetected as a fourth product.The formation of 2,2-diphenylpropiononitrile can beexplained by the sequence shown in Scheme 1.A singletiminocarbene, generated by thermal extrusion of nitrogenfrom the triazole, may undergo Wolff rearrangement toN-methyldiphenylvinylideneamine ; this in turn can givediphenylpropiononitrile by methyl migration. It wasfound that an independently prepared specimen of N-methyldiphenylvinylideneamine, when pyrolysed underthe same conditions, gave 2,2-diphenylpropiononitrile asthe major product. The other product detected in thepyrolysis of the vinylideneamine was tetraphenyl-succinonitrile, which is the dimer of the stabilised cyano-(dipheny1)methyl radical ; this is strong evidence thatthe rearrangement of the vinylidenearnine involves aradical mechanism.Similar rearrangements of vinyl-ideneamines have been described previ~usly.~Ph[N\\N ~ -+ Ph2C=C-NMePh NMePh N/Me IJIMeCPhiCN\ Ph2C-T$=NPh2C*CN / Me-IP h2C *CNSCHEME 1Wolff rearrangement is also the major reaction of thecorresponding singlet oxocarbene, so products formed bythis route were to be expected. On the other hand, theformation of isoquinolines involves participation of theN-methyl group, the methyl carbon atJm being in-corporated into the isoquinoline skeleton. The productsare most simply explained by invoking a 1,4-hydrogentransfer from the methyl group of the intermediate imino-carbene, followed by cis-tyans isomerisation and electro-cyclic ring closure to give a dihydroisoquinoline (Scheme2).Such a compound could readily aromatise to 3-phenylisoquinoline. The formation of 4-hydroxy-3-* C. L. Stevens and J. C. French, J. Amer. Chem. SOC., 1954,76, 4398.G. S. Hammond, 0. D. Trapp, R. T. Keys, and D. L. Neff,3. Amer. Chem. SOL, 1959, 81, 4878; H. D. Waits and G. S.Hammond, ibid., 1964,86, 1911; L. A. Singer and P. D. Bartlett,Tetrahedron Letters, 1964, 1887; L. deVries, J . Org. Chem., 1973,38. 4367.J. J. Artus, J.-J. Bonet, and A. E. Pena, J.C.S. Chm. Comm.,1973, 679.phenylisoquinoline may result from autoxidation of thedihydroisoquinoline at the 4-position, followed by de-hydration ; a similar hydroxylation of a dihydropyridinederivative has been observed.6 Attempts to prevent theformation of 4-hydroxy-3-phenylisoquinoline by exclud-ing oxygen during work-up were unsuccessful, andit seems that the autoxidation must occur extremelyreadily.5-+nuSCHEME 2The key feature of the mechanism outlined in Scheme 2is the 1,4-hydrogen transfer. Such reactions of carbenesare rare.Hydrogen transfers which are formallyanalogous have been observed in reactions of vinyl-carbenes ; ' mechanisms involving intramolecular ab-straction by the triplet carbene (or 1,3-diradical) areusually preferred. Similar hydrogen transfers have beenproposed to account for products of photolysis of 1-aminobenzotriazole 8a and of 1-benzylbenzotriazole,8bradical mechanisms again being assumed. Such hydro-gen transfers could also be envisaged in singlet carbeneor 1,3-dipolar structures.The proposed 1,i-hydrogen transfer was investigatedby studying the deuterium distribution in 3-phenyliso-quinoline isolated from the pyrolysis of l-[2H,]methyl-4,5-diphenyl-1,2,3-triazole.If the mechanism depictedin Scheme 2 is operative, deuterium should be found atC-1 and -4. C-1 should be fully deuteriated but C-4may not be since it is likely that this deuterium would bescrambled by interconversion of the dihydroisoquinolinesshown in the Scheme. This was borne out by experi-ment ; the 3-phenylisoquinoline obtained from thispyrolysis was fully deuteriated at C-1, according to then.m.r. spectrum, but the mass spectrum indicated thatonly about one third of the molecules in the specimen con-tained two deuterium atoms.Support for the mechanisms proposed in Schemes 17 G.L. Closs, L. E. Closs, and W. A. Ball, J. Amer. Chem. Soc.,1963, 85, 3796; R. Srinivasan, ibid., 1969, 91, 6260; J. A.Pinock, R. Morchat, and D. R. Arnold, ibid., 1973, 95, 7636;L. A. Wendling and R. G. Bergman, ibid., 1974, 96, 308; R. D.Streeper and P. D. Gardner, Tetrahedron Letters, 1973, 767.8 (a) E. M. Burgess, R. Carithers, and L. McCullagh, J. Amer.Chem. SOC., 1968,90, 1923; (b) M. P. Serve and H. M. Rosenberg,Abstracts 164th National Meeting Amer. Chem. SOC., 1972, Orgn.51975and 2 came from the pyrolysis of the triazoles (1b-d).l-Ethyl-4,5-diphenyl-l,2,3-triazole (lb) gave the pro-ducts expected on this basis, namely-2,2-diphenylbutyro-nitrile and tetraphenylsuccinonitrile (cf.Scheme 1) andl-methyl-3-phenylisoquinoline and 4-hydroxy-l-methyl-3-phenylisoquinoline (cf. Scheme 2). From l-isopropyl-4,6-diphenyl-l,2,3-triazole (lc) deoxybenzoin was iso-lated; in this case, the intermediate formed by the typeof hydrogen transfer shown in Scheme 2 cannot so readilycyclise and aromatise, and may survive, to be hydrolysedto deoxybenzoin during the isolation procedure. De-oxybenzoin was also isolated from the pyrolysis of 1-benzyl-4,5-diphenyl-l,2,3-triazole (Id) ; in this pyrolysisthe major products were bibenzyl and tetraphenyl-succinonitrile, probably formed by radical cleavage of N-benzyldiphenylvinylideneamine (cf. Scheme 1). Twoisoquinolines were also isolated in low yield.One,which could not be fully characterised, had a molecularweight and spectra as expected for 4-hydroxy-1,3-di-phenylisoquinoline, which is a product to be predictedby analogy with the mechanism of Scheme 2. The otherwas 3,4-diphenylisoquinoline; this is probably formed bya variant of the mechanism of Scheme 2, in which theintermediate iminocarbene undergoes direct cyclisationrather than hydrogen transfer. This is depicted inScheme 3. A similar type of reaction has been observedin the photolysis of 1-benzylbenzotriazole.g*SCHEME 3The two major reactions of l-alkyl-4,5-diphenyltri-azoles thus involve Wolff rearrangement and cyclisationto isoquinolines. The second of these was used as thebasis of a test for the intermediacy of 1H-azirines in thereactions.From the isomeric 1,4-dimethy1-5-phenyl-(3) and 1,5-dimethy1-4-phenyl-l,Z,3-triazole (4), commonproducts might be expected if the intermediate carbenescan undergo interconversion through a 1H-azirine(Scheme 4) ; in particular, 3-methylisoquinoline might beexpected from the pyrolysis of both triazoles although itcan be formed directly from only one of the two carbenes.The triazoles were prepared by the methylation of thesilver salt of 4-methyl-5-phenyl-v-triazole with iodo-methane ; the structures were assigned unambiguouslyby a second synthesis of the triazole (4) from azido-methane, phenylpropanone, and potassium t-butoxide.Two products were isolated from the pyrolysis of 1,4-dimethyl-5-phenyltriazole (3).One was 3-methyliso-quinoline, obtained in 8% yield and identified by com-parison with an authentic specimen. The other witstentatively identified as 2-methyl-Z-phenylpropiono-nitrile, the product expected from Wolff rearrangement.From the pyrolysis of 1,5-dimethyl-4-phenyl-l,2,3-tri-C. Graebe and F. Ullmann, Annulen, 1896, 291, 16; W.Freudenberg, in ' Heterocyclic Compounds,' vol. 3, ed. R. D.Elderfield, Wiley, New York, 1962, p. 298.3azole (4), 3-methylisoquinoline, and 4-hydroxy-3-methyl-isoquinoline were isolated and identified through theirMe[ 4 )SCHEME 4picrates. There was no evidence for the interconversionof the triazoles (3) and (4) under the pyrolysis conditions.The isolation of 3-methylisoquinoline from the pyrolysisof the triazole (3) is thus consistent with the mechanismshown in Scheme 4.1 - ( 1 -Pheny lvinyl) - and 1 -Phenyl-l,2,3-triaxoles .-1 H-1,2,3-Triazoles with an unsaturated substituent at the 1-position have a predictable reaction pathway open tothem for thermal or photochemical decomposition.Thisinvolves cyclisation of the intermediate carbene or 1,3-&radical, leading to products containing a new five-membered ring. Similar reactions of benzotriazoles havelong been known as the basis of the Graebe-Ullmannsynthesis of carbazoles.9 Reactions of this type havealso been observed in monocyclic 1H-triazoles, involvingcyclisation onto a l-phenyl,su l-benzoy1,l0a or l-pyrimi-din-2-yl lo* substituent. If 1H-azirines are involved inthe decomposition of such triazoles, their intermediacyrequires the formation of isomeric cyclisation products(Scheme 5) from unsymmetrically substituted triazoles.SCHEME 5The isomeric 4- and 5-phenyl-l-( l-phenylvinyl)-1,2,3-triazoles (5a) and (6a) were prepared from a-azido-styrene with phenylacetaldehyde and ethyl benzoyl-acetate, respectively; the 5-phenyl isomer (6a) has beenprepared previously.ll The triazole (5a) was pyrolysedlo (a) R.Huisgen and M. Seidel, Ckem. Ber., 1961, 94, 2609;(b) A. J. Hubert and H. Reimlinger, ibid., 1970, 108, 3811.11 G. L'abb6 and A. Hassner, J . HeterocycEic Ckem., 1970, 7,3614 J.C.S. Perkin Iat 570" and 0.01 mmHg, and gave a mixture containing2,4-&phenylpyrrolk, the product to be expected fromdirect cyclisation of the intermediate carbene, as themajor product.2,5-Diphenylpyrrole was also found asa significant product, this being isolated in 14% yieldafter chromatography. A third, minor product, 2,3-diphenylbut-3-enonitrile, can be accounted for by Wolffrearrangement of the intermediate carbene, followed bya 1,3-shift of the l-phenylvinyl group. The isomerictriazole (6a) also gave a mixture of 2,4- and 2,5-diphenyl-pyrrole on pyrolysis, with the 2,5-diphenyl isomer pre-dominating in this case. There was no evidence forinterconversion of the triazoles (5a) and (6a) before de-composition.The isomeric 1,4- and 1,5-diphenyltriazoles l2 (5b) and(6b) gave an analogous pattern of products. The triazole(5b), when pyrolysed at 650", gave a mixture of 2- and 3-phenylindole, with the latter as the major component.The triazole (6b) at 550' gave the same products but with2-phenylindole as the major isomer.These results are apparently in accord with Scheme 5,the initially formed iminocarbene either cyclising directlyor isomerising via the 1H-azirine.An alternative inter-pretation had to be considered when it was found that 2-and 3-phenylindole could be interconverted by pyrolysis,though a t a higher temperature, i.e. 800" and 0.02 mmHg.It thus seemed possible that the triazoles (5) and (6)could all decompose initially to give cyclisation productswithout rearrangement, and that these products thenrearranged to the isomeric pyrroles or indoles in thepyrolysis tube.Further evidence bearing on this prob-lem was obtained by a detailed study of the pyrolysis ofthe diphenyltriazoles (5b) and (6b) using %-labelledcompounds; this is reported in the following paper.134,5-Di~he~tyZ-v-t~iaxole.-Pyrolysis of the triazole at650" gave a mixture containing diphenylacetonitrile (theproduct to be expected from Wolff rearrangement) andPh PhPh NH Ph HSCHEME 62-phenylindole. Mechanisms for the formation of 2-phenylindole are speculative ; one possibility, involving2,3-diphenyl-lH-azirine and its isomerisation to 2,3-diphenyl-2H-azirineJ is shown in Scheme 6. The ther-mal rearrangement of 2,3-diphenyl-2H-azirine to 2-phenylindole is known to occur readily.14l2 L. Horner and W. Kirmse, Annalen, 1958, 614, 1.T.L. Gilchrist. C. W. Rees, and C. Thomas, following paper.l4 (a) T. L. Gilchrist, C. W. Rees, and E. Stanton, J . Chem.SOC. (C), 1971, 3036; (b) R. Selvarajan and J. H. Boyer, J . Hetevo-cvclic Chem., 1972, 9, 87.In summary, all the 1,2,3-triazoles investigated gaveproducts which were consistent with intermediacy of 1H-azirines, and mechanisms involving 1H-azirines providea common rationalisation for the diversity of productsfound. Dewar and Ramsden have predicted l6 that IH-azirines should be capable of existence as long-livedreaction intermediates, in accord with this rationalis-ation.EXPERIMENTALPreparative layer chromatography was performed usingsilica gel PF 254 (Merck). Vapour phase pyrolyses werecarried out using the apparatus previously described,l6 and acondenser cooled to - 78".Pyrolysis temperatures quotedwere obtained by means of a pyrometer with the thermo-couple inserted into the centre of the oven before applicationof the vacuum.Preparation of Triuzo1es.-(a) 4,5-Dipke~zyZ-v-tazoZe. 1-Amino-4,5-diphenyl- 1,2 , 3-triazole dissolved in acetic acid(10 ml) was stirred rapidly a t 0", sufficient water beingadded to prevent the mixture freezing. Sodium nitrite(0.60 g) in water (2 ml) was added dropwise; the mixturewas then diluted with water (100 ml) and the precipitatefiltered off and dried (1.7 g, 91%). Crystallisation gaveneedles of 4,5-diphenyl-v-triazole, m.p. 135-137" (frombenzene-hexane) (lit.,l7 138").(b) l-MethyE-4,5-di~henyZ-l,2,3-triazoZe (la).4,5-Di-phenyl-v-triazole (1.50 g, 6.8 mmol) in ethanol (25 ml) wasstirred rapidly, and silver nitrate (2.0 g ) in water (3 ml) wasadded dropwise. After 10 min the suspension was neutral-ised with ammonia, and the precipitated silver salt wasfiltered off, washed, and dried. It was then stirred withiodomethane (5 g) in chloroform (25 ml) for 12 h. Silveriodide was filtered off and the filtrate was evaporated toleave a pale cream crystalline solid (1.1 g). Triturationwith ether left a solid (0.7 g, 44%) which was a single sub-stance (t.1.c.). Two crystallisations gave l-methyZ-4,5-diphe?zyl-1,2,3-triazole (la), m.p. 129-130" (from ethylacetate-petroleum) (Found: C, 76.4; H, 5.6; N, 18.0.C1,H1,N3 requires C, 76-6; H, 5-6; N, 17.9yo); T (CDC1,)6.10 (3H) and 2,30-2.90 (lOH, ni) ; m/e 335 (Mf), 221, 207,192 (base), and 165; nz* (235 + 207) 182.5, (207 ----t 192)178, and (192 + 165) 142.The ether solution obtainedafter trituration of the solid contained two components(t .l.c.) : the triazole ( la) and 2-methyl-4,5-diphenyl- 1,2,3-triazole (2a), which is prepared in better yield by the pro-cedure described in the following section.1-[2H,]Methyl-4,5-diphenyl-l ,2,3-triazole (126 mg, 60%)was prepared in an analogous manner from the silver salt of4,5-diphenyl-v-triazole (291 mg) and i~do[~H,]methane(99.5% ",).(c) 2-MetJ~yl-4,5-diphenyl-l , 3,S-tviuzole (2a). 4,5-Di-phenyl-triazole (0.60 g, 2.7 mmol) in ether (25 ml) wasadded to ethereal diazomethane [generated from N-nitrosomethylurea (2.0 g)].The yellow solution was stirredfor 16 h and evaporated to leave an oil (0.63 g). Chromato-graphy (silica; ether-petroleum, 4 : 1) gave an oil (0.45 g,70%) which crystallised. Recrystallisation gave needles of2-methyl-4,5-dipheny1-lJ2,3-triazole (Za), m.p. 60-61" (from15 M. J. S. Dewar and C. A. Ranisden, J.C.S. Chem. Comm.,1973, 688.l6 D. J. Anderson, T. L. Gilchrist, D. C. Honvell, C . W. Rees,and E. Stanton, J.C.S. Pevkin I , 1972, 1317.17 R. StollC, W. Munch, and W. Kind, J . prakt. Chem., 1904,70, 4331976;petroleum) (lit.,l* 61.5-63"); T (CDCl,) 5.75 (3H) and2.30-2.90 (lOH, m). Further elution gave l-methyl-4,5-diphenyl-1,2,3-triazole (0.15 g, 23%).(d) 1- and 2-Ethyl-4,5-diphenyZ-l12,3-triazole [( lb) and(2b)l.4,5-Diphenyl-v-triazole (5-0 g, 22.6 mmol) was con-verted into its silver salt, as described in (a). The salt,suspended in chloroform, was heated under reflux with iodo-ethane (15 g) for 24 h. The mixture was filtered, the filtratewas evaporated, and the residue, an oil (4-0 g), was separatedinto its two components by chromatography [silica (150 g)].(i) Ether-petroleum (1 : 10) eluted 2-ethyl-4,5-diphenyl-1,2,3-triazole (2b) (2-0 g, 35%) as a viscous oil (Found: Mf249.127. C16H15N3 requires M+ 249.128); z (CDCl,) 8.50(3H, t, J 7 Hz), 5.62 (2H, q, J 7 Hz), 2.75-2-90 (6H, m), and2.35-2.75 (4H, m) ; m/e 249 (M+,base), 234, and 220. (ii)Ether-petroleum (1 : 5 ) eluted l-ethyl-4,5-diphenyZ-l12,3-triazole (lb) (1.4 g, 25y0), m.p. 110-111" (from etlier-petroleum) (Found: C, 77.1; H, 5.9; N, 16.9.C,,H,,N,requires C, 77.1; H, 6.1; N, 16.9%); z (CDCl,) 8-65 (3H, t ,J 7 Hz), 5.80 (2H, q, J 7 Hz), and 2-40-2-90 (lOH, m); m/e249 (M+), 221, 193, 192 (base), and 165; m* (249- 221)196, (221 ----t 192) 167, and (192 -t 165) 142.(e) 1- and 2-IsopropyZ-4,5-diphenyl-1,2,3-triazole [( lc) and(Zc)]. 4,5-Diphenyl-v-triazole silver salt (3.0 g, 9.1 mmol)and 2-iodopropane (10 g) were heated in chloroform (50 ml)under reflux for 24 h. The precipitate was filtered off andthe filtrate evaporated to leave an oil (1.2 g). Chromato-graphy (silica) gave (i) (with ether-petroleum, 1 : 20) an oil( 1.09 g, 42%) which slowly crystallised. Recrystallisationgave 2-isopropyl-4,5-diphenyl-1,2,3-triazole (2c), m.p.64-65" (from ethanol) (Found: C, 77.4; H, 6.7; N, 16.1.Cl,H1,N, requires C, 77.5; H, 6.5; N, 15.9y0); z (CDCl,)8.40 (6H, d, J 7 Hz), 5.18 (lH, septet, J 7 Hz), 2-60-2.90(6H, m), and 2-30-2-60 (4H, m); m/e 263 (M+), 249, 248,and 221. (ii) Elution with ether-petroleum (1 : 5 ) gave asolid (0.2 g , 8%). Crystallisation gave l-iso~ro~yl-4,5-diphenyl-1,2,3-triazole (lc), m.p. 126-128" (from benzene-hexane) (Found: C, 77.6; H, 6.5; N, 15.7%); T (CDCl,)8.55 (6H, d, J 7 Hz), 5-70 (lH, septet, J 7 Hz), and 2-70(lOH, m); m/e 263 (M+), 221, and 178 (base).(f ) l-Isopropyl-4,5-di~henyl- 1,2,3-triazole from 2-azido-propane and diphenylacetylene. 2-Iodopropane (10 g) andsodium azide (10 g) were heated in dimethylformamide (30ml) and water (20 ml) a t 100" for 12 h ; the solution was thendistilled and the distillate collected until the temperatureof the distillation flask reached 140".The distillate wasshaken with toluene (5 ml) and the organic solution wasseparated and dried. The solution, which contained 2-azidopropane (2.4 g; estimated by n.m.r.), was heated withdiphenylacetylene (3.0 g, 17 mmol) a t 135" (sealed tube) for15 h. The toluene was distilled off and the residue trituratedwith ether to give l-isopropyl-4,5-diphenyl-l, 2,3-triazole(1.3 g), m.p. 126-128'. Chromatography of the etherealsolution gave diphenylacetylene (1.8 g) and a second crop ofthe triazole (lc) (0.5 g) (total 1.8 g ; 100% based on di-phenylacetylene consumed).(g) l-Benzyl-4,5-diphenyl- 1,2,3-triazole ( Id).Diphenyl-acetylene and benzyl azide were heated together according toa published procedure 19 to give l-benzy1-4,5-diphenyl- 1,2,3-triazole (93y0), m.p. 112-112-5" (from ethanol) (lit.,lB110-112"); z (CDC1,) 4-65 (2H) and 2-40-3.10 (15H, m).(i) The available 1 mix-ture of l-amino-4-methyl-5-phenyl-1,2,3-triazole and 1-R. R. Fraser, Gurudata, and K. E. Haque, J . Org. Chem.,1969, 34, 4118.(h) 4-MethyZ-5-fihenyl-v-triazole.amino-5-methyl-4-phenyl-1,2,3-triazole (0.38 g) in aceticacid (3 ml) was cooled and stirred while aqueous sodiumnitrite (0.19 g in 1 ml) was added dropwise. Water wasadded occasionally to prevent the mixture solidifying. After1 h water was added to make the volume 25 ml. The finewhite precipitate was filtered off, washed, dried, and crystal-lised to give 4-methyl-5-phenyl-v-triazole (0.32 g, 92%), m.p.160-162" (from aqueous ethanol) (lit.,20 162").(ii) Theavailable 1 mixture of 4-methyl-5-phenyl- and &methyl-4-phenyl- l-p-tolylsulphonylamino- 1,2,3-triazole (4.0 g) ,treated with sodium (1 g) in liquid ammonia (40 ml), gave4-methyl-5-phenyl-v-triazole (1.65 g, 85%), m.p. 160-162".(j) 1,4-DimethyZ-5-phenyl-l, 2,3-triazole (3) , 1,Bdimethyl-4-Phenyl-ll2,3-triazole (4), and 2,4-dimethyZ-5-pJzenyl- 1,2,3-triazole. 4-Methyl-5-phenyl-v-triazole ( 1.65 g, 10.4 mmol)was converted into its silver salt. The dry salt was sus-pended in chloroform (75 ml) and stirred with iodomethane(16 g) a t room temperature for 16 h. The precipitate wasfiltered off and the filtrate was evaporated to leave an oil(1.62 g).Chromatography [silica (80 g)] gave (i) (withether-petroleum, 1 : 10) an oil (0.10 g), tentatively identifiedas 2,4-dimethyl-5-phenyl-l12,3-triazole ; T (CDCl,) 7.53 (3H),and 2.20-2.80 (5H, m) ; m/e 173 (M+, base), 144, 132, 131,and 130. (ii) Elution with ether-petroleum (3: 10) gaveback 4-methyl-5-phenyl-v-triazole (0.21 g) . (iii) Ether-petroleum (1 : 1) eluted 1,4-dimethyl-5-phenyZ- 1,2,3-triazoZe(3) (0.785 g, 50%), m.p. 103-105" (from petroleum)(Found: C, 69.2; H, 6-3; N, 24.5. Cl,Hl,N, requires C,69-3; H, 6-4; N, 24.3%); vmx. 1615w, 1580w, and 1500cm-l;T (CDCl,) 7-70 (3H), 6.09 (3H), and 2*40-2-80 (5H, m); m/e173 (M+), 145, 130, and 103 (base). (iv) Further elutionwith ether-petroleum (1 : 1) gave 1,5-dimethyZ-4-fihenyl-lI2,3-triazole (4) (0.43 g, 27%), m.p.98-100" (from petrol-eum) (Found: C, 69.6; HI 6.4; N, 24.3%); vmX. 1610wand1370 cm-1; 7 (CDC1,) 7-55 (3H), 6.00 (3H), and 2.10--2*80(5H, m); m/e 173 (M+), 145, 144, and 130 (base).(k) 1,5-DiunethyZ-4-phenyZ-l,2,3-triazole from azidomethaneand l-pkenylpropan-2-one. A dry solution of azidomethane(1 g; estimated by n.m.r.) in benzene (6 ml) was preparedby a published procedure.2l 1-Phenylpropan-2-one (2.35 g ,17.5 mmol) and dry tetrahydrofuran (25 ml) were added andthe solution stirred a t 0". Potassium t-butoxide (2.0 g) wasadded in portions during 0-5 h, and the solution was stirredfor a further 6 h. The solvent was distilled off and theresidue washed with ether.The ethereal solution wasevaporated to leave a crystalline solid (1-4 g, 46%). Re-crystallisation gave 1,5-dimethyl-4-phenyl- 1,2,3-triazole (4),m.p. 98-loo", identical (mixed m.p., i.r. spectrum, t.1.c.)with the specimen prepared by methylation of 4-methyl-5-phenyl-v-triazole.(1) 4-Phenyl-l-( l-phenyZvinyl)-1,2,3-triazoZe (5a). Phenyl-acetaldehyde (0.60 g , 5 mmol), a-azidostyrene (0.75 g, 5.2mmol), and potassium t-butoxide (0.7 g) were stirred indry tetrahydrofuran (20 ml) for 12 h. The solvent was dis-tilled off and the residue washed with ether. The etherealsolution was evaporated to leave a gum (0.8 g) which washeated with petroleum. The mixture was filtered and thefiltrate evaporated to leave an oil which slowly crystallised.Recrystallisation gave 4-fihenyl-1-( l-fihenylvinyl)-1,2,3-tri-azoZe (5a) (0.725 g, 60%), m.p.67.5-68.5" (from petroleum)(Found: C, 77.5; HI 5-3; N, 16.9. C,,H,,N, requires C,77.7; H, 5.3; N, 17.0%); T (CDCI,) 4.57 (lH, d, J 1 Hz),lo R. Huisgen and M. Seidel, Chsm. Ber., 1961,94, 2509.2o R. Meier, Chem. Ber., 1953, 86, 1483.21 F. 0. Rice and C . J. Grelecki, J . Phys. Chem., 1957,61, 8306 J.C.S. Perkin I4.29 (lH, d, J 1 Hz), 2.60-2.90 (8H, m), and 2.10-2.40(3H, m); m/e 247 (M+), 219, 218, and 77 (base).(m) 5-Phenyl-l-( l-~henylvinyl)-1,2,3-triazole (6a). Ethylbenzoylacetate (0.70 g, 3.6 mmol) and a-azidostyrene (0.60g, 4.1 mmol) in tetrahydrofuran (5 ml) were added dropwiseto potassium t-butoxide (0.5 g) in tetrahydrofuran (5 ml).The mixture was stirred at room temperature for 20 h,heated under reflux for 6 h, cooled, and poured into water.The precipitate was filtered off, washed with ether, anddigested with aqueous sodium hydroxide ( 5 ~ ; 10 ml).Acidification gave a precipitate of 5-phenyl- 1-( l-phenyl-vinyl)-1,2,3-triazole-4-carboxylic acid (0.20 g, 19yo), m.p.155-1 57" (decomp.) (from chloroform-carbon tetrachloride)[lit.,12 156-157" (decomp.)].The carboxylic acid (0.20 g)was heated at 200" for 2 min. The orange, viscous residueof 5-phenyl-l-(a-styryl)-1,2,3-triazole (6a) l1 was sufficientlypure (t.1.c.) for use in the pyrolysis experiments. In subse-quent pyrolysis experiments the carboxylic acid was de-carboxylated in situ.(n) 1,4- and 1,5-Diphenyl-1,2,3-triazoZe [(5b) and (6b)].These triazoles were prepared by heating azidobenzene withphenylacetylene according to a published procedure.12Preparation of Isoquino1ines.-(a) 3-Phenylisoquinoline.This was prepared by a published procedure 22 involvingreduction of l-chloro-3-phenylisoquinoline, and had m.p.102-104" (from ethanol) (lit.,2a 103-105°) ; T (CDCl,)1-90-2.80 (lOH, m) and 0.80 (1H); mle 205 (N+) (100%and 204 (50%).(b) 3-Phenylisoquinolin-4-ol.3-Phenylisoquinoline (0- 18g ) and aqueous hydrogen peroxide (30%; 0.3 ml) wereheated in acetic acid (3 ml) a t 70" for 40 h. More aqueoushydrogen peroxide (0.2 ml) was added after 12 h. After40 h, t.1.c. (silica; ether) showed that some 3-phenyliso-quinoline still remained. The solution was evaporated todryness at 70" and 20 mmHg, water (3 ml) was added, andthe mixture was again evaporated to dryness.Chloroformwas added to the residue and the solution dried and evap-orated, leaving an oil (0.156 g) which solidified on triturationwith ether, giving a buff solid (0,097 g, 49%). Two re-crystallisations gave 3-phenylisoquinoline N-oxide, m.p.158-160" (from ether), vmx. 3400br and 1570 cm-1. TheN-oxide (68 g) was heated with acetic anhydride (1.5 ml)under reflux for 0.5 h. The acetic anhydride was dis-tilled off and the residue was triturated with ether. Aninsoluble solid (9 mg) was filtered off; v,, 1670 and 1630cm-1, as expected for 2-acetyl-3-phenylisoquinolin-1(2H)-one. The ethereal solution was evaporated to a gum (90 mg)which was heated under reflux with hydrochloric acid (2N;5 ml) for 40 min.The pale yellow solution was made basic(aqueous sodium hydrogen carbonate), and the yellowprecipitate was filtered off and dissolved in ether. Thesolution was evaporated to a yellow gum (22 mg, 32%)which crystallised when triturated with ether-pentane.Recrystallisation gave 3-phenylisoquinolin-4-ol, m.p. 159-161" (yellow prisms from benzene) (Found: M+ 221.082.Cl,HllNO requires M+ 221.084) ; vmx (CHCl,) 3550 (OH),1635, and 1590 cm-l; Lx (EtOH) 255 (E 26,500), 313 (7500),335 (5500), and 402 nm (700); 7 (CDC1,) 4.25br (lH, OH),1.70-2.80 (9H, m), and 1.22 (1H) ; m/e 221 (M+), 220, 205,192, 165, and 85 (base); picrate, m.p. 226-227" (fromethanol).(c) 3-MethylisoquinoZin-4-0l.~~ 3-Methylisoquinoline (7.022 S.Gabriel, Ber., 1886, 18, 3473.23 M. M. Robison and 33. L. Robison, J . Org. Chem., 1956, 21,1337.g) and aqueous hydrogen peroxide (30% ; 8 ml) were heatedin acetic acid (15 ml) a t 70" for 18 h. The product wasisolated as in (b) and gave 3-methylisoquinoline N-oxide (7.0g, go%), m.p. 138-139" (from benzene-cyclohexane) (lit.,23136-139"). The N-oxide (3 g) was heated under reflux inacetic anhydride (30 ml) for 1 h. The acetic anhydride wasdistilled off and the residue was washed with water, thentriturated with ether. This gave a solid tentatively identi-fied as 2-acetyl-3-methylisoquinolin-1(2H)-one (1.1 g), m.p.215-216" (from ethanol) ; vmx. 1670, 1648, and 1610 cm-l.The ethereal solution was added to ethanolic picric acid.The yellow precipitate (1.4 g) was filtered off, aqueoussodium carbonate was added, and the mixture was extractedwith ether.Chromatography (silica) gave an oil (0.65 g)which slowly crystallised; vm, 1766 cm-1, as expected for4-acetoxy-3-methylisoquinoline. This solid (0.20 g) washeated with hydrochloric acid (5% ; 20 ml) under reflux for40 min. An excess of aqueous sodium carbonate was addedand the product was extracted with ether. This gave asolid which was recrystallised to give 3-methylisoquinolin-4-01 (134 mg), m.p. 178-180" (from benzene) (lit.,23 180");vmax 2700br, 1633, and 1584 cm-l: 7 (CDC1,) 7.36 (3H),3-75br (lH, OH), 1.70-2-70 (4H, m), and 1.35 (1H); m/e159 (M+, base) and 158; picrate, m.p.202-203".Pyrolysis of Triazo1es.-(a) l-Methyl-4,B-diPhenyl- 1,2,3-triazole (la). The triazole (100-200 mg portions) waspyrolysed at 600" and 0.01 mmHg. At lower temperatures,some of the triazole was unchanged, and a t higher tempera-tures, a more complex mixture of products was detected.The mixture of pyrolysis products (86 mg from 100 mg) gave,by preparative layer chromatography (silica ; ether-petroleum, 1 : 1) : (i) at RF 0.9, 2,2-diphenylpropiononitrile(45 mg, 51y0), identical (i.r., n.m.r., and mass spectra) withan authentic specimen; (ii) at RF 0.5, 3-phenylisoquinoline(17 mg, 19%), m.p. and mixed m.p. 102-104", identical(i.r. and n.m.r. spectra) with an authentic specimen; and(iii) a t RF 0.4, 3-phenylisoquinolin-4-01 (11 mg, 12y0), m.p.and mixed m.p.160.5-161.5", identical (i.r., u.v., andn.m.r. spectra) with an authentic specimen. If the crudepyrolysis mixture was triturated with ether, tetraphenyl-succinonitrile (4 mg, 5%), m.p. 204-206" (from ethanol)(lit.,2* 215O), remained.To test for the intermediacy of N-methyldiphenylvinyl-ideneamine, the imine 4 (150 mg) was pyrolysed at 600" and0.01 mmHg and gave an amber oil (130 mg). Triturationwith ether gave tetraphenylsuccinonitrile (30 mg, 2 1 yo),m.p. 204-206", and 2,2-diphenylpropiononitrile (95 mg,(b) l-EthyE-4,5-diphenyZ-l,2,3-triazole (lb). The triazole(100 mg) was pyrolysed a t 600-620" and 0-01 mmHg, and ayellow gummy solid (85 mg) collected in the condenser.Preparative layer chromatography (silica; ether-petroleum,1 : 1) gave three products which were not fully characterised;structural assignments were made on the basis of theirspectra, and by analogy with the products from the triazole(la).They were: (i) a t RF 0.7, an oil (44 mg, 50%),identified as 2,2-diphenylbutyronitxitrile ; vmx. (film) 2250 (CN)and 1605w cm-1; 7 (CDCl,) 8.98 (3H, t, J 7 Hz), 7-62 (2H, q,J 7 Hz), and 2.70 (lOH, m); (ii) a t l i p 0.55, l-methyl-3-phenylisoquinoline (5 mg, 6%), m.p. 47-49" after re-sublimation (lit.,26 48-49') ; and (iii) at R p 0-48, 1-methyl-3-~henylisoquinolin-4-02 (12 mg, 13%), rn-p. 105-106" (from24 E. P. Kohler and N. L. Drake, J . Amer. Chem. Soc., 1923, 45,1281.25 S. Goszczynski, Roczniki Chem., 1964, 38, 893.63%)1975benzene); vmX.3000br (OH), 162Ow, and 1598w cm-l;Am= (EtOH) 257, 314, and 338 nm; T (CDC1,) 7.15 (3H) and1*70-2.80 (9H, m) ; m/e 235 (M+, base), 234, and 206; m*(234 --t 206) 182.(c) l-Isopro~yZ-4,5-diphenyZ-1,2,3-triuzoZe (lc). The tri-azole (50 mg) was pyrolysed at 650' and 0.01 mmHg andgave a yellow gum (43 mg). The gum was warmed withdilute hydrochloric acid and the mixture shaken with ether.The ethereal solution was dried and evaporated. The resi-due gave, with 2,4-dinitrophenylhydrazine, the 2,4-dinitro-phenylhydrazone of deoxybenzoin, m.p. and mixed m.p.196-198". No other products were investigated.(d) l-BenzyZ-4,5-diphenyZ-1,2,3-triazoZe (Id). This waspyrolysed in batches (total 1500 mg) at 600" and 0.01 mmHgand gave a complex mixture of products (1250 mg) .Tritur-ation with ether gave tetraphenylsuccinonitrile (300 mg,32y0), m.p. 204-206" (from ethanol) (lit.,24 215") (Found:C, 87-5; H, 5.2; N, 7.1. Calc. for C,,H,,N,: C, 87.5; H,5.3; N, 7.3%), vmx. (CHC1,) 2250w (CN) cm-l; m/e 193(base). The ether solution was evaporated to an oil (950mg). Chromatography (silica ; ether-petroleum, 1 : 20)gave (i) a yellow oil (330 mg) from which crystals of bibenzyl(190 mg, 43%) separated. The residual oil had vmx. 2250cm-1, consistent with the presence of 2,2,3-triphenylpropio-nonitrile, but this was not purified further. (ii) Deoxy-benzoin (225 mg, 27%) was identified by conversion into its2,4-dinitrophenylhydrazone, m.p. and mixed m.p. 19A196O(from ethanol) ; mle 376 (Mf, base). (iii) A yellow gum (80mg) was not fully characterised; the spectral data are con-sistent with a diphenylisoquinolinol ; vmX (CHC1,) 3540 and1590 cm-l; A,, (EtOH) 255, 316, and 342 nm; mle 297(base) 296, and 190.(iv) 3,4-Diphenylisoquinoline (140 mg,10%) had m.p. 159-159-5" (lit.,26 155-156') ; vmx. 1625w,1580w, and 1565w cm-l; 7 (CDC1,) 1-80-3.00 (14H, m) and0.6 (1H); nzle 281 (M+) and 280 (base).(e) 1,4-DimethyZ-5-~henyZ-l,2,3-triazoZe (3). The triazole(100 mg) was pyrolysed at 750" and 0.01 mmHg, and gave ared oil (78 mg). This was triturated with ether, and theethereal solution was evaporated to give an oil (60 mg) ; theether-insoluble residue appeared to be polymeric. Prepara-tive layer chromatography of the oil (silica ; ether-petrol-eum, 3 : 1) gave two products: (i) at RF 0.45, 3-methyliso-quinoline 27 (7 mg, 8%), m.p.and mixed m.p. 60-61";(ii) at Rp 0.8, an oil (16 mg, 19yo), tentatively identified as2-methyl-2-phenylpropiononitrile; vmaX (CHCl,) 2240 cm-1(CN); m/e 145 (Mc) and 129 (base).(f) 1,5-DirnethyZ-4-phenyZ-l,2,3-triazoZe (4). The triazole(100 mg) was pyrolysed at 750" and 0.01 mmHg and gave agum (80 mg). The product mixture was triturated withether to remove a polymeric solid (10 mg); the ether-soluble fraction contained two major components (t.1.c.) ;T (CDCI,) 7.38 and 7.35 (methyl protons), and 1.37 and 1-01(l-H of isoquinolines) . Ethanolic picric acid was added andgave a precipitate (55 mg). Fractional crystallisation fromethanol gave two pure picrates, identical (m.p., mixed m.p.,and i.r.spectra) with authentic specimens of S-methyliso-quinoline picrate and 3-methylisoquinoline-4-01 picrate.The free bases were liberated from the picrates with aqueouspotassium hydroxide. They were extracted by addition ofether and identified (m.p., i.r.) by comparison with authenticspecimens.26 G. Berti and P. Corti, Gazzetta, 1968,88, 704.87 W. H. Mills and J. L. B. Smith, J. Chem. SOC., 1922, 121,28 L. W. Dady, Tetrahedron, 1967, 25, 3606.2732.(g) 4-Phenyl-l-( l-fihenyZvinyl)-1,2,3-triazoZe (5a). Thetriazole (200 mg) was pyrolysed at 570' and 0.01 mmHg andgave a mixture (152 mg) containing three major components,which were isolated by preparative layer chromatography(silica; ether-petroleum, 2 : 3) : (i) at l i p 0.4, 2,4-diphenyl-pyrrole (48 mg, 27%).m.p. and mixed m.p. 178-180" (frompetroleum) (lit.,28 180°), identified by comparison (i.r.) withan authentic specimen (this pyrrole gives an indigo colour withEhrlich's reagent) ; (ii) a t RF 0.6, 2,5-diphenylpyrrole (24 mg,14%), m.p. and mixed m.p. 142-144" (from benzene) (lit.,29l43"), identified by comparison (i.r.) with an authenticspecimen (the pyrrole gives a mauve colour with Ehrlich'sreagent) ; (iii) at R p 0.7, a viscous oil (53 mg) which was nothomogeneous. Two further purifications by layer chromato-graphy gave an oil, identified as 2,3-diphenyZbut-3-enonitriZe(9 mg, 5%); vmaZ 2250w cm-l; T (CDC1,) 5.01 (lH, dd, J 1and 0.75 Hz), 4.41 (lH, d, J 1 Hz), 4-37 (lH, d, J 0.75 Hz),and 2.7&-2-90 (lOH, m); nz/e 219 (MS), 192, and 103 (base).The assignment of structure was supported by the followingreaction.The nitrile (9 mg) was heated in dioxan (2 ml)under reflux with triethylamine (10 mg) for 1 h. This gavean oil (8 mg) which was purified by layer chromatography(silica; ether-petroleum, 1 : 3). One major component wasisolated ; it was 2,3-diphenylcrotononitrile (3 mg), m.p. 79-81" (lit.,30 €!lo), identified by comparison (i.r., t.l.c., n.m.r.)with a specimen prepared by the method of ref. 30; vmx.2215 cm-l; T (CDCI,) 7-43 (3H) and 2.50-3.00 (lOH, m);m/e 219 (M+) and 204 (base).(h) 5-PhenyZ-l-( l-~FrenyZvinyZ)-l,2,3-triuzoZe (6a). 5-Phenyl- 1- (l-phenylviny1)- 1,2,3-triazole-4-carboxylic acid (80mg) was decarboxylated in situ by heating briefly at 200".The residue was pyrolysed a t 575' and 0.1 mmHg and gavea mixture (52 mg) containing two major components.Thesewere separated by layer chromatography (silica ; ether-petroleum 2 : 1); they were (i) at RF 0.5, 2,4-diphenyl-pyrrole (10 mg, 17y0), m.p. and mixed m.p. 177-180°, and(ii) at RF 0-75, 2,5-diphenylpyrrole (27 mg, 45y0), m.p. andmixed m.p. 142-144'.(j) 1,4-l>i~henyZ-1,2,3-triazoZe (5b). The triazole (50 mg)was pyrolysed at 650" and 0.02 mmHg. The product, ayellow gum (38 mg), consisted of 2- and 3-phenylindole asthe major products, the 3-phenyl isomer predominating(t.1.c.). No isolation or quantitative analysis was attempted.(k) 1,5-Di~henyZ-1,2,3-triazoZe (6b) (with C. THOMAS).The triazole (33 mg) was pyrolysed a t 550" and 0.02 mmHg.This gave a yellow gum (27 mg) which consisted of 2- and3-phenylindole, with the 2-phenyl isomer predominating,together with a little starting triazole (t.1.c.). No quantita-tive analysis was performed.In control experiments, both 2- and 3-phenylindole wereunchanged after pyrolysis a t 600" and 0.02 mmHg. Inexperiments performed at 800' and using a tube packed withsilicon carbide chips to increase the contact times, the in-doles were observed to undergo interconversion. 2-Phenylindole (30 mg) gave 3-phenylindole (9.3 mg, 31%)(isolated by preparative layer chromatography) and 2-phenylindole (15.5 mg, 52%). 3-Phenylindole gave 2-phenylindole (18.7 mg, 62%) and 3-phenylindole (4.5 mg,(1) 4,5-DiphenyZ-v-triuzoZe. The triazole (100 mg) waspyrolysed a t 650" and 0.02 mmHg, and gave a yellow gum15%).29 L. F. H. Allen, D. M. Young, and M. R. Gilbert, J . Org.30 T. Kumamoto, K. Hosoi, and T. Mukaiyama, Bull. Chem.Chem., 1937, 2, 227.SOC. Japan, 1968, 41, 2742J.C.S. Perkin I(77 mg). This was crystallised and gave diphenylaceto-nitrile (67 mg, My0), m.p. 70-71" (from ether-hexane)(lit.,16 71-72") ; v,, 2250 cm-l (CN) ; 7 (CDC1,) 4.90 (1H)and 2.70 (lOH, m). Preparative layer chromatography ofthe mother liquors gave diphenylacetonitrile and 2-phenyl-indole (5 mg), m.p. 187-189" (from ethanol), identical (i.r.,mixed m.p.) with an authentic specimen.(m) 1-[2H3]MethyZ-4, 5-diphenyZ-l,2,3-triazoZe (with C.THOMAS). The triazole (60 mg) was pyrolysed a t 630" and0.005 mmHg. 3-Phenylisoquinoline (4.8 mg) was isolatedfrom the products by preparative layer chromatography ;7 (CDCl,) 1.90-2.80 (m), the signal at 7 0.80 observed inundeuteriated specimens being absent; mle 207 (53%) , 206(loo%), 205 (50%), and 204 (10%).We thank Mr. C. Thomas for carrying out the pyrolysesof 2- and 3-phenylindole and l-[2H,]methyl-4,5-diphenyl-1,2,3-triazole, and the S.R.C. for a research studentship(to G. E. G.).[4/1075 Received, 3rd June, 1974