v-Triazolo4,5-dpyrimidines (8-azapurines). Part XVI. Preparation of 6-amino-8-azapurines by heating 4-amino-1,2,3-triazole-5-carbonitrile (and itsN-alkyl derivatives) with amidines

摘要

1975 345v-Tr iazol o C4.5-d J py r i m id i nes (8-Aza p u r i nes) . Part XVI . Preparation of6-Amino-8-azapurines by heating 4-Amino-l,2,3-triazole-5-carbonitrile t(and its N-Alkyl Derivatives) with AmidinesBy Adrien Albert, Research School of Chemistry, Australian National University, Canberra, Australia, 26004-Arnino-l.2.3-triazole-5-carbonitrile. and its 1 -, 2-, and 3-methyl, and 3-benzyl derivatives (1 a - e ) , reacted withformamidine, acetamidine. 2,2,2-trichloroacetamidine, and NN'-dibutylforrnamidine to give the corresponding6-amino-, 6-amino-2-methyl-, 6-amino-2-trichloromethyl-, and 6-butylamino-8-azapurines (2). respectively.Yields were usually very good and the reaction has preparative value. 6-Arnino-9-benzyl-2-methyl-8-azapurinewas prepared also by the action of triethyl orthoacetate on 4-amino-3-benzyl-l.2.3-triazole-5-carboxarnidine (3).A 2-trichloromethyl-8-azapurine was conveniently dehalogenated to the 2-dichloromethyl analogue.4-Amino-3-benzyI-l.2.3-triazole-5-carbonitrile (1 e) was converted by free guanidine into 1 -(4-amino-3-benzyl-l,2,3-triazol-5-ylcarbonimidoyl)guanidine (7b), which could be cyclized thermally to 2.6-diamino-9-benzyl-8-azapurine (2p).The mechanisms of the amidine and guanidine reactions are compared.1.r. and lH n.m.r. spectra are recorded and discussed.6-AMINO-8-AZAPURINE (8-aza-adenine) has been found inhibits Gram-positive and -negative bacteria byt o inhibit cancer cells that have become resistant to it greatly reducesment with this drug. The 6-amino-compound also V.A. Gershun, Tsitologiya, 1970, 12, 467.sistently numbered 4, to facilitate comparisons.being incorporated into the RNA;2-amino-8-azapurin-6-one (8-azaguanine) during treat-t In this series, the amino-group of aminotriazoles is con-D. Adomaitiene, T. N. Ignatova, D. Ya. Podgaetskaya, andR. 0. Roblin, J. 0. Lampen, J . P. English, Q. P. Cole, andJ. D. Smith and R. E. F. Matthews, Biochevz. J., 1967, 66,J. R. Vaughan, J. A ~ ~ ~ . ,-hem. sot., 1946, 67, 290.323. 1 Part XV, A. Albert, J . C . S . Perkin I , 1974, 2030346 J.C.S. Perkin 1the rate of multiplication of tobacco mosaic 496 andpotato viruses, and it inhibits the synthesis of adeno-sine triphosphate in the fungus Neurospora crassa.7These inhibitions are typically reversed by adenine.3*sTo assist similar biological studies, a convenientsynthesis of 6-amino-8-azapurines with a variety ofsubstituents was sought.A likely pathway seemedto be the condensation of 4-amino-l,2,3-triazoles withvarious amidines, analogous to the Taylor-Ehrhartreaction in which 4-aminopyrimidine-5-carbonitrile andformamidine, at 1 3 5 O , gave 4-aminopyrimido4,5-d-pyrimidine.* This reaction, which has been extendedto the p y r r ~ l e , ~ oxazole,1° and imidazole lo series, wasreviewedll in 1970, and has since been used to convert2-aminopyrazine-3-carbonitrile into 4-amin0~teridine.l~In the present work, 4-amino-1,2,3-triazole-5-carbo-nitrile (la) and its 1-, 2-, and 3-methyl derivatives(1b-d) were condensed with formamidinium acetate at120" to give very good yields of 6-amino-8-azapurine1(1) ( 2)R R1 R2 K3a H a H H Hb l-Me b 7-Me H Hc 2-Me c 8-Me H Hd 3-Me d 9-Me H He 3-CH,Ph e 9-CH,Ph H Hf 7-Me H Meg 8-Me H Meh 9-CH2Ph H Mei 7-Me H CCI,j 8-Me H CQk 9-CHZPh H cc1,1 8-Me H CHCI,m 7-Me Bun Hn 8-Me Bun Ho 8-Me H NH2p 9-CH2Ph H NH,and its 7-, 8-, and 9-methyl derivatives, respectively(2a-d). The 9-benzyl analogue (2e) was.preparedfrom the corresponding triazole (le) ; the best yield(65) came from a short, incomplete reaction, whereaslonger heating gave a product contaminated by high-melting, feebly soluble material, different from thedimer l3 of the starting material, and suspected to arisefrom a Dains reaction (condensation of an amidine witha reactive methylene group, by elimination of am-rnonia).l4 To pursue this idea further, the N-4-acetyland -formy1 derivatives of the aminotriazole (le) weretreated similarly, but the former remained largely un-changed, and the latter was transformed mainly to the8-azapurine (2e).Some of the aminocyanotriazoles (lb-d) were con-6 I.R. Schneider, Phytopathology, 1964, 44, 243.6 I. P. Zhuk, L. F. Didenko, and N. I. Gorbarenko, Biol.7 H. Urbanek, Acta SOC. bot. Pol., 1967, 36, 347.8 E. C . Taylor and W. A. Ehrhart, J . Amer. Chem. SOL, 1960,8 E. C. Taylor and R. W. Hendress, J . Amer. Chem. Sac., 1966,Nauki, 1970, 108.82, 3138.87, 1996.densed with acetamidinium acetate to give excellentyields of the correspondingly N-alkylated 6-amino-2-methyl-8-azapurines (2f-h) .6-Amino-9-benzyl-2-methyl-8-azapurine (2h) was also prepared by heating4-amino-3-benzyl-l,2,3-triazole-5-carboxa.midine 1 (3)with triethyl orthoacetate ; when acetic anhydride waspresent also, the product was 6-acetamido-9-benzyl-2-methyl-8-azapurine. Moderate yields of three 6-amino-2-trichloromethyl-8-azapurines were obtained bycondensing 2,2,2-trichloroacetamidinium acetate with theaminocyanotriazoles (lb-d); the yield could not beimproved by increasing the reaction temperature (ashad been done with acetamidine) because of decom-position.6-Amino-8-methyl-2-trichloromethyl-8-azapurine (2j)was partly dehalogenated to 6-amino-2-dichloromethyl-8-methyl-8-azapurine (21) by tin(r1) chloride in coldacetone.Mechanism of Condensations with A midines.-Thecondensation (for example) for 4-amino-l-methyl-1,2,3-triazole-5-carbonitrile with formamidinium acetatecould proceed through either the 5-carbonimidoylfox-m-amidine (4) or 4-aminomethyleneamino-l-methyl-1,2,3-triazole-5-carbonitrile (5a).That the latter routewas preferred was indicated by the failure of 4-acet-amido-1 -methyl-1,2,3- triazole-5-carbonitrile to give theCH,Ph( 3 115)a;R=Hb; R= Me4-acetamido-derivative of (4) when heated with form-amidinium acetate under the conditions that convertedthe parent triazole (lb) into the 8-azapurine (2b). Themost likely formation of the postulated intermediate(6a) is by addition of the 4-amino-group of the triazoleto the double bond of the amidinium ion to give atetrahedral intermediate e.g.(S), a course parallel tothe addition l5 of ammonia to the cation of 4-dimethyl-aminomethyleneamino-1 -met hyl-1 ,2,3-triazole-5-carbo-lo J. P. Ferris and L. E. Orgel, J . Amev. Chem. Soc., 1966, 88,3829.l1 E. C. Taylor and A. McKillop, 'The Chemistry of CyclicEnaminonitriles and o-Aminonitriles,' Interscience, New York,1970.l* A. Albert and K. Ohta, J . Chem. SOC. ( C ) , 1971, 3727.18 A. Albert, J . Chem. SOC. ( C ) , 1970, 230.14 F. B. Dains, 0. 0. Malleis, and J. T. Meyers, J . Amer. Chem.Soc., 1913, 35, 970; F. B. Dains, R. Irvin, andC. G. Harrel, ibid.,1921, 43, 613; F. B. Dains, R. Thompson, and W. F. Asendorf,ibid., 1922, 44, 2310; R. M. Roberts, J .Org. Chem., 1949,14, 2771975 347nitrile (5b) as a preliminary to elimination of the di-methylammonium ion and ring closure to 6-amino-7-methy1-8-a~apurine.~~However, the following experimental evidence op-poses such a sequence of cationic forms here. Thetriazole (le) failed to react with acetamidine hydro-chloride under conditions where acetamidinium acetategave a 65 yield. This result made it desirable toexamine a free amidine. In place of free acetamidine(judged too unstable) , trichloroacetamidine was heatedwith the triazole (lc) and gave a 28 yield of the corre-sponding 8-azapurine (2j) (much of this amidine survivedthe reaction); however the acetate gave a 61 yield,and the hydrochloride none at all. Thus acidity ishelpful in one stage (and harmful in another) of thetotal reaction. The optimal acidity is too mild to forcethe eventual ring closure (addition of -NH, to -CN),which is most probably purely thermal.As an alterna-tive to a tetrahedral intermediate, it may be assumedthat a minute proportion of the aminotriazole is con-verted into its cation which combines with the (nucleo-philic) neutral midine much as equimolecular pro-portions of aniline and aniline hydrochloride reactin the industrial synthesis of diphenylamine. However,there is an immense difference in the basic strengths ofacetamidine (pKa 12-4) and these aminotriazolenitriles(e.g. pK, -1.4 for the 2-methyl derivative 16). In thelight of this discussion (which favours a tetrahedralintermediate), the severer conditions required for con-densing acetamidine (compared to formamidine) seemto be mainly electronic, and those encountered withtrichloroacetamidine mainly steric.NN' -Di bu t y If orm amidinium acetate and the t ria zoles(Ib and c) at 200" eliminated butylamine to give 6-butyl-amino-7( and 8)-methyl-8-azapurine (2m and n) respec-tively, by a spontaneous Dimroth rearrangement of thefirst - f ormed 1 -but yl-6-imino- 1,6-dihydro-8-azapurine .lThe hydrochloride of dibutylformamidine did notreact.Condensations with Guanidine.-In the literature,guanidine (always either free or as carbonate) has beenused to annulate o-amino-nitriles to diaminopyrimido-aza-analogues; this has been done in the pyrroline, pyri-midine, and pyrazine series only.ll Intermediates17)a; R=Z-Meb; R-3-CH2Phnamely 1 - (2-aminop yrazin-3-ylcarbonimido yl) guani-dines have been isolated only from the pyrazine-pteridine conversion^.^^ Here, 4-arnino-2-rnethyl-I,2,3-triazole-5-carbonitrile (lc) and free guanidine in boilingethanol readily gave 1- (4-amino-2-meth yl-l ,2,3-t riazol-16 A.Albert, J.C.S. Perkin I, 1972, 461.l6 A. Albert, J.C.S. Perkin I, 1973, 1634.5-ylcarbonimidoy1)guanidine (7a) which decomposedextensively when ring-closure to 2,6-diamino-8-methyl-8-azapurine (20) was attempted in boiling butanol,acetic acid, N-hydrochloric acid, or ethanolic sodiumhydroxide, or with guanidinium acetate in boilingbutanol. 1- (4-Amino-3-benzyl- 1,2,3-triazol-5-ylcarbon-imidoy1)guanidine (7b), similarly prepared, readilycyclized (in boiling butanol) to 2,6-diamino-9-benzyl-8-azapurine (2p).It was found that guanidinium acetate, unlike acet-amidinium acetate, had little tendency to react with theaminocyanotriazoles (1).The greater success of freeguanidine, and the different pathway selected by thisreagent, must depend on the greater ease with whichguanidine forms an anion, thanks to the electronicinfluence of the third amino-group. Such an anionwould readily add to the electropositive carbon of thecyano-group, leading to intermediates of type (7).Physical Constants.-1.r. spectra of 8-azapurines arein Table 1. The spectrum of 2,6-diamino-g-benzyl-8-azapurine is compatible with those of 2,6-diamino-7 (and 8) -met h yl-8-azapurine .l8 The 2- t ric hlor ome t h ylTABLE 11.r. spectra (Nujol)v,,/cm-l8-Azapurines6-Acetamido-9-benzyl-Gmethyl1 , 2,3-Triazoles4-Acetamido-3-benzyl-6-cyano3220, 3150br,w (NH str.),1690br,m (CO, amide I),1590br,s (amide 11), 1316s,1300s (amide 111), 1220m,and 715m3420m, 3315m, 3200m (NHstr.), 1650br,s, 1575s, 1355m,845s, and 795s (C-Cl str.)3290br,m, 3210m, 1625br,s,1225s, and 795m (C-Cl str.)3506w, 3370s, 3290m, 3260m,3130m (NH str.), 1685m.1610s (NH bend), 1585m, and1430m3230s, 3190s (NH str.),2246m ( E N str.), 1675br,s(CO str.). 1585m, 1520s,1490s, 1325m, and 1245m3420m, 3310m (NH str.),1600br,s, 156Om, 1430m, and1185m3416m, 3355m, 3270m (NHstr.), 1610br,s, 1530m,1496m, 1426m, and 1180mderivatives show a clear C-C1 stretching band at 795cm-l.The reversal of strength for amide I and amide I1bands, as seen here in 6-acetamido-9-benzyl-2-methyl-8-azapurine, was discussed re~ent1y.l~ Spectra of some1,2,3-triazoles, including the novel carbonimidoyl-guanidines, are given in the same table. lH N.m.r.spectra of six 6-amino-8-azapurines are given in Table 2.The chemical shifts of 2-H, 6-NH2, and the variousN-alkyl groups are very close to values determined forl7 J. H. Jones and E. C. Cragoe, J. Medicin. Chem., 1968, 11,l9 A. Albert, J.C.S. Perkin I, 1973, 2659.322.A. Albert and H. Taguchi, J.C.S. Perkin I, 1972, 449J.C.S. Perkin Irelated 8-azapurines.15 In the 2,7-dimethyl derivative,the 2-methyl signal occurs much further upfield thanthe 7-methyl signal, as may be expected.Other n.m.r.values are recorded in the Experimental section.TABLE 2lH N.m.r. spectra ca. 34'; solvent (CD,),SOJ of8-azapurines6-Amino-8--azapurine (2)R1 RZ Ri 7 Values 07-Me H H 1.72 (lH, 2-H), 2.26br b (2H, NH,), and7-Me H Me 2.38slbr (2H, NH,), 6-69 (3H, 7-Me),6-66 (3H, 7-Me)and 7.66 (3H, Me)(2H, CH,), 7.36 (3H, 2-Me), and 7.47(3H, Ac) c7-Me H CCI, 1-69br (2H, NH,), and 6-64 (3H, Me)9-CH2Ph H CCI, 1.2vbr (NH,), 2.61 (5H, Ph), and 4.098-Me H CHCI, 1.60brb (2H, NH,), 3.06 (lH, CHCI,),Tetramethylsilane as internal standard; all signals weresinglets. b Exchangeable in D20. 0 Assignments reversiblefor last two peaks.9-CH,Ph Ac Me 0.16 b (lH, CONH), 2.66 (6H, Ph), 4.19(2H, CH,)and 6-66 (3H, Me)EXPERIMENTAL1.r.spectra were taken (for mulls) with a Perkin-Elmer257 grating spectrometer. N.m.r. spectra were obtainedwith a Varian HA 100 (100 MHz) instrument. Specimenssaid to be identical were compared by (i) mixed m.p.determination where applicable, (ii) i.r. spectroscopy,N-hydrochloric acid (4 ml), and the mixture was filtered.The filtrate, mixed with Iw-sodium citrate (0.6 ml) andadjusted to pH 3 with 10N-sodium hydroxide, deposited6-amino-8-azapurine (2a) , identical with authentic material.Example of the Use of Acetamidine.-4Amino- l-methyl-1,2,3-triazole-5-carbonitrile (lb) (0.123 g, 0.001 mol),acetamidinium acetate 13 (0-354 g, 3 equiv.), and hexan-1-01(2 ml) were heated under reflux for 4 h.The solvent wasremoved a t 116" and 25 mmHg. The residue was slumedwith boiling ethanol (2 ml), and set aside at - 10" for 2 h.Filtration furnished 6-amino-2, 7-dimethy1-8-azapuvitte (7-amino-1 ,5-dimethyl-v-triazolo4,5-dpyrimidine) (2f) , m.p.301°, after recrystallization from water with carbon toremove fluorescent material.Alternative Syntheses of 2-MethyZ-8-azapurines (5-Methyl-v-triazolo4,5-d#yrimidines) .-4-Amino-3-benzyl-l, 2,3-tri-azole-5-carboxamidine hydrochloride (0.126 g, 0.0005 mol)and triethyl orthoacetate (3 ml) were strongly heatedunder reflux for 12 h. Volatiles were removed at 100"and 25 mmHg. The residue was boiled with N-aceticacid (2.6 ml) for 30 min, to effect a second stage of thereaction (as revealed by paper chromatography).Themixture deposited, on cooling, 6-amino-9-benzyl-2-methyl-8-azapurine (2h) (83), m.p. 266" (from 65 parts of ethanol).When acetic anhydride (1 ml) was added initially, 6-acet-awido-9-benzyl-2-methyl-8-azapurine (50) was obtained ,m.p. 153" (from 11 parts of ethanol) (Found: C, 59.4;H, 4.9; N, 29-7. C,,H,,N,O requires C, 59.6; H, 5.0;N, 29.8).Example of the Use of 2,2,2-Trichloroacetar~ridine.-4-Amino-2-methyl- 1,2,3-triazole-5-carbonitrile (0.123 g),TABLE 3Preparation of 6-amino-S-azapurines (2) from 4-amino-5-cyano-l,2,3-triazoles (1)Approx. Reflux Fwnd (96) Required ( 1 time Recrystallization Map. YieldProduct (2)Starting Sol- - triazoie vent= tTer* (h) Solvent (Parts) ("C) () - Formula - If7-Me8-Me9-Me7-Me8-Me7-MeS-Me7-Me8-Me9-CHaPh9-CHaPh9-CHaPhHHHHHHHHHHHBunBunHHHHHMeMeMeCCl,CCl,CCl,HHBBBBBHHHHBB0012012012012012016016016016012012020020012221420.522444301 v220266>320303 b18616712677 b p9Obdr88 4;9Obef65 b,'f'80828350614264 b,g20 b,i41.044-360.327.127-342.34.95-25.21.91-92.951.560.735-331-631.224.239.939.530.643.960.026-9641.94.95.01.92-691.836-031-424.439.831.0a B, butan-1-01, H, hexan-1-01, 0, octan-1-01, P, pyridine trihydrate.b Identical with authentic material. e Ref. 3. d A. Albert and K.Tratt, J . Ckem. Soc. (C), 1968.344. 8 Ref. 15. I A. Albert, J . Chem. SOC. (C), 1969,152. o Slight decomp. Much decomp. * Ref. 1.and (iii) comparative chromatography on two WhatmanNo. 1 papers, developed in (a) aqueous 3 NH,Cl, and (b)butanol-5~-acetic acid (7 : 3). Unless otherwise specified,material for analysis was dried a t 80" in air. ' Pyridinetrihydrate refers to the azeotrope, b.p. 92'.4-Amino- 1,2,3-triazole-5-~arbonitrile,~0 and its l-methyl,ls2-methylJ1* 3-methylJ21 and 3-benzyl l3 derivatives ; also4-acetamido- l-methyl- 1,2,3-triazole-5-~arbonitrile,1~ wereprepared as indicated here. Condensations of theseaminonitriles with amidines were carried out as specifiedin Table 3, with the results shown there.Example of the Use of Formamidine.-4-Amino- 1,2,3-triazole-carbonitrile (la) (0.109 g, 0.001 mol), formami-dinium acetate (0.31 g, 0.003 mol), and sieve-driedbutan-1-01 (2 ml) were heated under reflux a t 125' (bathtemp.) for 1 h.The solvent was removed a t 90" and 25mmHg. Water (1 ml) was added (pH 6.5) ; the suspensionwas refrigerated and the solid filtered off, triturated withtrichloroacetamidine 22 (0.483 g , 0.003 mol), acetic acid(0.18 g, 3 mol), and butanol (2 ml) were heated underreflux for 2 h. The solvent was removed a t 90" and 25mmHg; the residue triturated with water (2 mi) and filteredgave 6-amino-8-methyl-2-trichloromethyl-8-azapurine(7-amino-2-methyl-5-trichloromethyl-v-triazolo4, 5-dIpyrinzi-dine).DehaEogenation.-Tin(iI) chloride dihydrate (0.45 g,0.002 mol) dissolved in acetone (4 ml) was added dropwiseto a stirred suspension of 6-amino-8-methyl-2-trichloro-methyl-8-azapurine (0.534 g, 0.002 mol) in acetone (20 ml)a t 24O.After 18 h, the clear solution was evaporated todryness. N-Sodium hydroxide (16 ml) was added to dis-solve the tin. The product was filtered off, boiled brieflywith ethanol (15 ml), then set aside a t 24". Filtration re--4. Albert, J. Chem. SOC. ( C ) , 1969,2379.2o A. Albert and H. Taguchi, J.C.S. Perkin I , 1973, 1629.2z A. Albert and B. Paal, Chenz. and l a d . . 1974, 8741975 349moved some starting material (0.1 g). The filtrate wastaken to dryness and the residue recrystallized from 100parts of benzene (2 crops), giving 6-amino-2-dichloronzethyl-8-methyZ-8-azafiurine ( 7-amino-5-dich Zoromethyl-2-methyZ-v-triazolo4,5-d~~rimidine) (66), m.p.202" prior topartial resolidification (Found: C, 30.8; H, 2-6; C1, 30.5.C6'H,C1,N6 requires C, 30.9; H, 2.6; c1, 30.4).Example of the Use of NN'-Dibutylformamidine.-4-Amino- l-methyl- 1 , 2,3-triazole-5-carbonitrile (0- 123 g),NN'-dibutylformamidine 8 (0.47 g, 3 equiv.), acetic acid(0.18 g , 3 equiv.), and octan-1-01 (2 ml) were heated underreflux for 4 h (the solution remained clear throughout).Solvent was removed a t 135" and 25 mmHg. The crystal-line residue was transferred to a filter with light petroleum,then recrystallized from a little water, giving 6-butyl-amino-7-methyl-azapurine, m.p. 167" (lit.,, 167").Amino-3-benzyl- 1,2,3-triazole-5-carbonitrile (1.0 g, 0.005mol) and freshly prepared acetic formic anhydride (10 ml)were stirred a t 22" for 17 h.The solvent was removed a t40" (in vacuo). The residual crystals, stirred with water(8 ml), then filtered off, gave 3-benzyl-4-formamido-1,2,3-triazole-5-cavbonitrile (9 1 ) which, recrystallized fromethanol (6 parts), or benzene (280 parts), had m.p. 152"(Found: C, 58.4; H, 4.3; N, 31.0. C,,H,N,O requiresC, 58.1; H, 4.0; N, 30*8), 7 (CD,),SO 1.57 (lH, CHO),2.67 (5H, m, Ph), and 4-39 (2H, CH,). The nitrile (le)(1.0 g), acetic anhydride (1.0 g, 2 equiv.), and dried pyridine(10 ml) were heated under reflux for 4 h. The volatileswere removed at 100" and 25 mmHg. The residue, re-crystallized from a little ethanol and then from 150 partsof benzene, gave 4-acetamido-3-benzyl- 1,2,3-triazole-5-carbo-nitvile (75), m.p.153" (Found: C, 59.5; H, 4.8; N, 29.4.C,,H,,N,O requires C, 59.7; H, 4.6; N, 29.0).4-Acylamino-3-benzyl- 1,2,3-t~iazoZe-5-ca~bonitriles.- 4-Condensations with Guanidine.-Guanidine hydrochloride(0.24 g, 0.0025 mol) was boiled for 5 min with ethanolicO-5~-sodium ethoxide (5 ml), cooled, and filtered from NaCl.The filtrate and 4-amino-2-methyl- 1,2,3-triazole-5-carbo-nitrile (0.246 g, 0.002 mol) were heated under reflux for1 h. The ethanol was removed in vacuo a t 30". Theresidue was stirred with water (2 ml) and filtered giving1-( 4-amino-2-methyl- 1,2,3-triazol-5-ylcarbonimidoyl)guani-dine (7a) (85) which, recrystallized from 150 parts ofethanol, softened a t 177", evolved ammonia, and meltedabout 300" Found (for material dried at 24" and 25 mmHg) :C, 33.1; H, 5-7; N, 61.6. C,H,,,N, requires C, 33.0; H,5.5; N, 61.5y0, A,,,, (cation a t pH 7) 231sh and 286 nm(log E 2.97 and 3.92), '(C,D,N), 3.74slbr (2H, 4-NH2,exchangeable) , 5.35mbr (5H, side-chain, exchangeable),and 6-15 (3H, 2-Me). 1-(4-Amino-3-benzyl-1,2,3-triazoZ-5-y1carbonimidoyl)guan~dine (7b), prepared similarly, wastriturated with N-acetic acid (6 ml), filtered from a little2,6-diamino-9-benzyl-8-azapurine (1 3), the filtrate ad-justed to pH > 12 with N-sodium hydroxide, and theamidine (65) filtered off; i t softened, and evolved am-monia when heated, and melted a t ca. 270" Found (formaterial dried a t 24" and 25 mmHg): C, 51.3; H, 5.5;N, 43.2. C,,H14N, requires C, 51-15; H, 5.5; N, 43.4y0.This amidine (1 g) and butanol (8 ml) were heated underreflux for 1 h ; the suspension, taken to dryness a t 95"and 25 mmHg, gave 2,6-dinunino-9-benzyl-8-azapurine(5,7-diamino-3-benzyZ-v-t~iazolo4,5-d~yrimidzne) (2p) (9 1 yo)(from 80 parts of pyridine trihydrate, in 2 crops); verysparingly soluble in boiling ethanol (Found: C, 55.0; H, 4.6;N, 40.3. Cl1Hl1N, requires C, 54.8; H, 4.6; N, 40.6y0).I thank Mr. B. Paal for skilled experimental help.4/1836 Received, 6th September, 1974