1976 1327Furazans and Furazan Oxides. Part V1.l New Furamano[3,4-d]pyrimidineN-Oxides : Preparation and StructureBy Remus Nusiu and A. John Boulton,' School of Chemical Sciences, University of East Anglia, NorwichSome new furazano[3,4-d] pyrimidine N-oxides have been prepared. The tautomerism of the furazan oxide ringfavours 1 -oxide structures, to the virtual exclusion of the 3-oxides. Some tetrazolopyrimidine intermediatesunderwent hydrolytic cleavage of the pyrimidine ring, leading to tetrazole derivatives.NR4 7TJA NUMBER of examples of the furazano[3,4-d]pyrimidine 7AT-oxide jpyrimidofuroxan) system (1) have been re-p ~ r t e d . ~ ? ~ All carry electron-donor substituents in thepyrimidine ring, and these are apparently required forstability, since the efforts of ourselves and others toobtain derivatives of (1) with simple alkyl or halogensubstituents have been unrewarded.The majority ofthe cases studied are in fact fused aminopyrimidine,pyrimidinonc, or uracil derivatives, (1; R2 = NH,), (2),and (3) , respectively.We embarked upon the present study for two reasons.First, lve swr@t evidence for the separate existence ofisomers a and b,: since corresponding forms can bedetected at low temperature in the n.m.r. spectra offurazano[3,44]pyridine N-oxides? The rates of inter-conversion of isomers can be used to provide a q~alitcktiv~estimate of the aromaticity associated with the ring towhich the furoxan system is fused,l and it was of interestto obtain such information for the pyrimidine, pyrimidin-4-one, and pyrimidine-2,4-dione (uracil) systems.Secondly, both furazano[3,4-d]pyrimidines and their N-oxides have been reported to be of synthetic ~ t i l i t y , ~ , ~and there seemed to be scope for further development inthis area.Pyrimiriim Derivatives (1) ; Ring Cleavage Reactions.-Azidodechlorination was unsuccessfully attempted withthe three chloronitropyrimidines (4)-(6) ; the chlorinewas, however, displaced by hydrazine, giving the corre-sponding hydrazinopyrimidines (7)-( 9).The di-methoxy and methoxy-methyl compounds (7) and (8)gave unstable azides (10) and (11) (and/or their corre-sponding tetrazolopyrimidine cyclised forms 2, withnitrous acid ; these were decomposed without isolationto form the corresponding furoxan derivatives (1 ;R1 = R2 = OMe) and ( 1 ; R1 = Me, R2 = OMe), re-spectively.*4 by-product from the preparation of (1 ; R1 = Me,R2 = OMe) is the tetrazole (12), which had earlier beenisolated by Temple et aL2 from the reaction of B-chloro-5-nitropyrimidin-4-one with sodium azide in aqueousacid medium.When the hydrazine (9) was nitrosated,the sole product isolated was again a tetrazole. Fromthe high carbonyl frequency (1 745 cm-l) of the acetylt Chemical Abstvacts would name the ring system of (1) as[ 1 , 2,5]oxadiszolo[3,4-d]pyrimidine.Througliout this paper, the .I-oxides will be denoted asisomers a, the %oxides as b. When no distinction is intendedbetween the isomers, the designations are omitted.Part V, &I. J.Boulton and D. Middleton, J . Ovg. Chew., 1974,39, 2956.C. Temple, C. L. Kussner, and J. A. Montgomery, J . Org.Chcm., 1068, 33, 2086.group, we prefer the structure (13), rather than theisomeric formula with the acetyl group 0x1 the enaminenitrogen. p-Nitro-enamines have been isolated fromnucleophilic nitropyrimidine ring-cleavage by Clark et a1.'I la10-I l b l(31{41 R':R2: OMe,X r C I(51 R':Me ,R2= OMe, X: C II61R':R': Me,X:CI(71R1=R2: OMe, X-NH.NH2(81 R' :Me , R2:OMe, X : NH. NH2{91R':R2: Me, X = N H * N H 21101R':R2: OMe, X - N 3(11 I R' : Me , R 2= 0 Me , X : N 302 N . C H CO N H 2tO2N\ /NH2 c=cI 'MeI12 I (1 31The lH n.m.r. spectra of the methoxy-compounds (1 ;R1 = R2 = OMe) and (1 ; R1 = Me, R2 = OMe) showedthat in each case one of the two possible forms (la and b)is strongly favoured, and this is assumed by analogywith earlier work4 to be the one (la) in which theN-oxide group is conjugated with the pyrimidinenitrogen atoms, rather than the non-conjugated alter-native (lb).In the case of the dimethoxyfurazano-[3,4-d]pyrimidine N-oxide only two peaks, of equalF. Yoneda and Y. Sakuma, J . Heterocyclic Chem., 1973, 10,993. * A. J. Boulton. P. J. Halls, and A. R. Katritzky, J . Chem.SOL. ( B ) , 1970, 636.E. C. Taylor, Y. Maki, and A. McKillop, J . Org. Chem.,1972,37, 1601.F. Yoneda, Y . Sakuma, and S. Matsumoto, Heterocycles,1975, 3, 113.J. Clark, I. Gelling, I. W. Southon, and M. S. Morton, J .Chenz. SOC. ( C ) , 1970, 4941328 J.C.S.Perkin Iintensity, were seen, both at normal (+38 "C) and atlow (-60 and -80 "C, in CDC1,-CFC1,) temperatures.The 7-methoxy-&methyl compound (1 ; R1 = Me,R2 = OMe) likewise showed only two peaks at normaltemperatures, but on cooling two additional very weaksignals appeared, each to low field of one of the majorpeaks; these are attributed to ca. 1% of the minortautomer (lb). At -28 "C, the chemical shifts were6 4.25 and 2.675 for (la), and 4.305 and 2.70 for (lb)(R1 = Me, R2 = OMe; methoxy and methyl groups,respectively). Coalescence, defined by the disappear-ance of the minor peak, occurred at ca. +5 (OMe) and0 "C (CMe). Although an accurate study of the inter-conversion rate [(la) +- (lb)] is outside the scope ofthis work, qualitatively, at 0 "C the lifetime of the minortautomer (lb) is similar to that of one of the (degenerate)tautomers of benzofuroxan.8 The effect of the N-oxidegroup on the chemical shift of the 7-methoxy group is ashielding one: the methoxy-signal of (la) is 0.055 p.p.m.upfield of that of (lb).Earlier work has shown methylgroups and protons to be shielded similarly (by 0.01-0.03.4 and 0.19-0.35 p.p.m. ,9 respectively).UraciZ Derivatives (3) .-The dimethyl compound (3 ;R1 = R2 = Me) was reported by Yoneda d aL3910 to beformed by the action of either sodium nitrite in aceticacid, or potassium nitrate in acetic and sulphuric acids,on 6-hydroxyamino-l,3-dimethyluracil (14), in yields of65 and 60%, respectively. Our experience is sufficientlydivergent to warrant mention.The ' hydroxylamine(14) ' exists exclusively, so far as can be judged, by i.r.and n.m.r. in [2H6]acetone, in the oxime form (15).When treated with NaN0,-AcOH it gave a mixture ofthe furoxan (3; R1 = R2 = Me) (40%) and the corre-sponding furazan (16) (25y0), with a certain amount ofunidentified material, m.p. 180", of high molecularweight (by mass spectrometry). When the oximeacetate (17) reacted with sodium nitrite in aqueoushydrochloric acid, the furoxan (3; R1 = R2 = Me) (30%)was the only product isolated, while with sodium nitritein acetic acid the furazan (16) was formed in good yield.The monomethyl compound (3; R1 = H, R2 = Me)was reported by Yoneda et aZ.1° as melting above 300",and to be formed from the corresponding hydroxylamineusing the nitrative cyclisation method (KN0,-AcOH-H2S04) as described above for (14).We have preparedthis furoxan by another route [thermal decompositionof the tetrazolopyrimidine (IS)], and find that its m.p.is 211-212". The product was characterised byanalytical and mass spectral data, and by its conversioninto (3; R1 = R2 = Me). We therefore consider thatthe work of Yoneda et aZ.1° is in error on this point.llThe thermal decomposition of the tetrazolopyrimidine(18) hydrate was studied by differential thermogravi-8 K.-I. Dahlqvist and S. Forsen, J . Magnetic Resonance, 1970,2, 61.9 A. J. Boulton, A. R. Katritzky, M. J. Sewell, and B. Wallis,J . Chem. SOC. (B), 1967, 914.lo F. Yoneda, Y. Sakuma, and M.Ueno, J . Heterocyclic Chem..1973, 10, 415.l1 F. Yoneda (personal communication) concurs with thisconclusion.metric analysis. Loss of water, and then of nitrogen,was observed. Calculation by the method of Freemanand Carroll l2 gave an estimate of 115-120 kJ mol-lfor the activation energy of the latter process.We attempted to convert the furoxan ring of theuracil derivatives (3; R1 = H or Me, R2 = Me) into ahydroxyimidazole N-oxide by base-catalysed condens-ation with a nitroalkane, in the manner previouslydescribed for benzofuroxans.13 However, no reactionwas observed between nitromethane or 2-nitropropaneand either of these fused furoxans.n 0Me MeI141 R :H 1151 R : H(17 I R : Ac0 0Me \ /N'Nf16) 118)The monomethyl compound (3; R1 = H, R2 = Me)could be converted into a variety of 4-substitutedderivatives (see Table) by alkylation under basic con-ditions. Some of these derivatives hold promise for thesynthesis of further fused heterocyclic systems, byreductive cleavage of the furoxan ring and cyclisationof the residue at the pyrimidine 6-position to the sub-stituent R1; this is under investigation.The insolubility of the 4,6-dimethyl compound (3 ;R1 = R2 = Me) at low temperatures in suitable solventsprecluded the use of the n.m.r.method for investigatingits tautomerism. The 4-n-butyl analogue (3; R1 =Bun, R2 = Me) was freely soluble in CDCl,, but thespectrum at -28 "C was essentially identical with thatat +38 "C; no new peaks were discerned above thenoise level at high amplification, and it is clear that a tlow temperatures there prevails a single isomer, which,on the same grounds of precedent as applied ab~ve,~,l*we expect to be the 1-oxide (3a).It is noteworthy thatthe dimethyl compound (3; R1 = R2 = Me) is sug-gested6 to enter into reaction with aniline in its tauto-meric form (3b). This postulate, though plausible, is,we feel, unnecessary to explain the observed results.EXPERIMENTAL1.r. spectra were taken on a Perkin-Elmer 257 gratingspectrophotometer, and were of KBr disc preparations,unless otherwise specified. N.m.r. spectra at other thanl2 E. S. Freeman and B. Carroll, J . Phys. Chem., 1958, 62, 394.l3 M. J. Abu-el-Haj, J . Org. Chem., 1972, 37, 2519; D. W. S.Latham, 0.Meth-Cohn, and H. Suschitzky, J.C.S. Chem. Comm.,1972, 1040.l4 J. Ackrell and A. J. Boulton, J.C.S. Perkin I , 1973, 351;A. Gasco and A. J. Boulton, J.C.S. Perkin 11, 1973, 16131329normal temperatures (38 "C) were measured with a VarianHA-100 instrument, with V4343 variable temperatureprobe attachment; otherwise, a Perkin-Elmer R12 instru-ment was used. A Unicam SP 800 spectrophotometerprovided the U.V. spectra. Mass spectra (70 eV ionisingpotential), recorded with a Hitachi-Perkin-Elmer RMU 21instrument, were measured and were in agreement withassigned structures for all compounds named in mainheadings below.Light petroleum refers t o the fraction b.p. 60-80 "C.5,7-Dimethoxyfurazano[3,4-d]pyrimidine 1-Oxide ( 1 ;R1 = R2 = OMe).-Nitric acid (d 1.56; 3 ml) was addeddropwise, with cooling and stirring, t o 4-chloro-2,6-di-methoxypyrimidine 15 (2.0 g) in sulphuric acid (3 ml).Themixture was warmed to 80-90 "C for 2.5 h, with continuousstirring, and was then cooled and poured onto crushed ice(50 g). The nitropyrimidine (4) was filtered off, washedwith water, and recrystallised from ethanol, giving needles(1.7 g, 700&), m.p. 67-69', soluble in CCl,, C&&, andCHCl, (Found: C, 32.6; H, 2.9; N, 18.9. C6H6C1N,04requires C, 32.8; H, 2.7; N, 19.4y0), Lx. (MeOH) 337(E 1 goo), 265 (16 800), and 216 nm (28 000).The nitropyrimidine (4) (2.2 g) in methanol (50 ml) wasadded with stirring to hydrazine hydrate (1 g) in methanol(50 ml) a t 30 OC. After 2 h the yellow solid which separatedwas collected by filtration, washed with water, methanol,and dietliyl ether, and dried in vacuo (P20,).The hydrazine(7) formed small prisms (2.0 g, 97%), m.p. 155-156"(deconip.), from tetrahydrofuran, dioxan, or a large volumeof methanol, but it was sufficiently pure to be used asprepared (Found: C, 33.2; H, 4.4; hT, 32.4. C,H,N,O,requires C, 33.5; H, 4.2; N, 32.6%).Aqueous sodium nitrite (1.5 g in 12 ml) was added to astirred suspension of the hydrazine (7) (2.5 g) in hydro-chloric acid ( 0 . 5 ~ ; 40 ml) at 0 "C. (Ethanol was added toreduce foaming.) The mixture was stirred a t 20 "C for 2 h.The solid, which was filtered off and washed with water, alittle methanol, and then ether, showed i.r. bands at2 150 cm [N, of wide (lo)], and elsewhere others corre-sponding t o the furoxan (1; R1 = R2 = OMe).Thismixture was refluxed 3 h in tetrahydrofuran, the solventwas removed, and the residue was recrystallised fromethanol, giving the furoxan (1; R1 = R2 = OMe) as paleyellow plates or prisms (1.6 g, 70%), m.p. 166-168"(Found: C , 36.7; H, 3.2; N, 28.4. C6H6N,04 requires C,36.4; H, 3.0; N, 28.3%); A,,,. (MeOH) 345 (6 350), 290(21 OOO), 281 (21 200), and 215 nm (E 31 500); vmx. 1 630s,1560s, 1530s, 1340s, 1245m, 1210m, 1 170w, 1075m,and 1 O l O m cm-l; G(CDC1,) 4.12 and 4.22 (OCH,).7-Methoxy-BmethyEfurazanof3,4-dJ~yrimidine I-Oxide (1 ;R1 = Me, R2 = OMe) .-4-Chloro-6-methoxy-2-methyl-5-nitropyrimidine l6 (5) (3.05 g) was treated with an excess ofhydrazine as described for compound (4).The solid whichwas precipitated proved t o be a mixture of the requiredmonopyrimidylhydrazine (8) and the correspondingly di-substituted hydrazine. They were separated by sublim-ation a t 125-130" and 0.3 mmHg, the more volatilesublimate forming orange-yellow needles of the hydrazine(5) (1.6 g, 53y0), m.p. 137-139" (Found: C, 36.1; H, 4.6;N, 35.8. C6H,N,0, requires C , 36.2; H, 4.5; N, 35.2%).The less volatile residue was recrystallised from tetrahydro-furan-light petroleum and sublimed at 180" and 0.3 mmHg,giving yellow needles of 1 ,Z-bis-( 6-methoxy-2-methyZ-5-nitro-l5 H. J. Fisher and T. B. Johnson, J . Amer. Chem. Soc., 1932,54, 727.@yrirnidin-4-yZ)hydrazineJ m.p. 205-206" (Found : C, 39.2;H, 4.0; N, 30.9.C12Hl,N,0, requires C, 39.3; H, 3.8;N, 30.6%), M+ 366.Aqueous sodium nitrite (0.6 g in 4 ml) was added withstirring t o the hydrazine (5) (1.0 g) in hydrochloric acid(0.501; 16 ml) a t 0 "C. The mixture was stirred at 0 "C for40 min, then for 30 min at 20 "C. It was again cooled to0 "C and filtered. The solid, which was separated from thefiltrate (A), washed with water, and dried (P205), showedan azide i.r. band (2 150 cm-l). It could not be purified,but was heated to reflux (3 h) in tetrahydrofuran (60 ml).After concentration to 20 ml, hot light petroleum (120 ml)was added, and after cooling a yellow precipitate wasremoved. This was dissolved in acetone (25 ml) andfurther light petroleum (120 ml) was added. An un-identified yellow material separated.The light-petroleum-containing mother liquors were combined and evaporatedto dryness, and the residue of the furoxan (1; R1 = Me,R2 = OMe) was recrystallised from water-ethanol (1 : l),giving yellow plates (0.5 g, 55%), m.p. 109-llO", soluble inacetone, chloroform, and ethanol (Found: C, 39.9; H, 3.5;N, 30.4. C6H6N40, requires C, 39.6; H, 3.3; N, 30.8y0),Mf 182; lux. (MeOH) 350 (2 SOO), 283 (3 450), 245 ( 5 600),and 210 nm (E 10 800); vmax. 1 630br,s, 1 580s, 1 545br,s,1510m, 1465m, 1 450m, 1430m, 1 400s, 1 375s, 1 310s,1300s, 1225s, and 1 180m cm-l; G(CDC1,) 2.62 (CH,) and4.21 (OCH,).The filtrate (A) was concentrated to 15 ml and cooled inice. Plates of the tetrazole (12) (0.06 g, 7%) separatedand were filtered off and dried (P205); m.p.187-189"(explosive decomp.) (lit.,2 192-195") ; vmx. 3 410s, 3 300m,3 250br,m, 3 000vbr,m, 1 620s, 1 580s, 1 530m, 1 480s,1400m, 1380w, and 1350br,s cm-l; m/e 172 (lay0, M+),69, 57, 55, 44, 43 (base), 42, 30, 29, and 28.Chloro-2,4-dimethyl-5-nitropyrimidine l7 ( 1.4 g) in methanol(30 ml) was added with stirring to hydrazine hydrate(0.73 g) in methanol (30 ml). After 1 h stirring and afurther 3 h a t 20 "C the methanol was evaporated off andthe residue was washed with water (20 ml) and recrystal-lised from CCl,, giving the (pyrimidine (9) as yellow needles(l.Og, 75%) (Found: C, 39.3; H, 4.9; N, 38.3. C,H,N502requires C, 39.3; H, 5.0; N, 38.3%).l-Acetyl-5-( 2-amino-l-nitruprup-1-enyl)tetrazoZe (13) .-Aqueous sodium nitrite (0.42 g in 3 ml) was added dropwiset o a cooled (-10 "C) stirred solution of the hydrazine (6)(0.6 g) in hydrochloric acid (0.5~; 12 ml).After 30 minstirring the cooling bath was removed and stirring wascontinued for 30 min at 20 "C. The mixture was againcooled, t o 0 O C , and a pale yellow solid was separated byfiltration, washed with cold water, and crystallised fromacetone-light petroleum. The enamino-tetrazole (13) formedyellow needles (0.4 g, 58y0), m.p. 144" (decomp.) (Found:C, 34.0; H, 3.95; N, 40.0. C,H,N,O, requires C, 34.0;H, 3.8; N, 39.6%); Lx. (H,O) 341 (7 650) and 208 nm(E 11 400); vmax. 3 240m, 3 O6Obr,m, 1 745s, 1 620s, 1 555s,1 505w, 1440s, 1 400s, 1 385s, 1 370m, 1 345m, 1 285s,1235m, 1 18Os, 1 140m, 1070m, 1005m, and 995m cm-l;6[(CD,),SO] 2.13 (3 H), 2.25 (3 H), and 7.25br (2 H); m/e212 (loyo, M+), 170, 153, 85, 67, 43 (base), 42, 41, and 30.4,6-DirnethyZfurazano[3,4-d]pyrimidine-5,7 (4H, 6H)-dioute1-Oxide (3; R1 = R2 = Me).-Method (a).6-Hydroxy-amino-l,3-dimethyluracil (15), prepared according t ol6 R. Urban and 0. Schnider, Helv. Chirn. Ada, 1958, 41, 1806.l7 F. L. Rose, J . Chem. Soc., 1954, 4116.6-Hydrazino-2,4-dimethyl-5-nitropyrimidine (9) .-61330 J.C.S. Perkin IPfleiderer and Ferch,ls had m.p. 146-148" (lit.,lS 146-148"); S[(CD,),CO] 3.15 (3 H), 3.21 (3 H), 3.75 (2 H), and9.55 (1 H). Acetylation (Ac,O) gave the oxime acetate(17) as white needles, m.p. 129-130" (lit.,l* 126-128");G(CDC1,) 2.16 (3 H), 3.25 (3 H), 3.48 (3 H), and 3.83 (2 H);v,,,.1 770, 1 736, 1 686, and 1 620 cm-l.The oxime acetate (17) (0.6 g) was dissolved in hydro-chloric acid ( 0 . 5 ~ ; 8 ml), with cautious heating (SO0).The solution was cooled to 0 "C and aqueous sodium nitrite(0.56 g in 6 ml) was added dropwise, with stirring. Themixture was then stirred at 20 "C for 3 h. The solid whichseparated was filtered off, washed with water, and re-crystallised from ethanol, giving platelets or needles, m.p.245" (decomp.) [lit.,, 245" (decomp.)] (0.16 g, 28%). Theproduct was occasionally discoloured yellow or pink, butwithout affecting the m.p. or i.r. spectrum. Although theyield was not high, the product was uncontaminated withthe furazan (16).Method ( b ) . The oxime (15) (0.85 g) was treated withcould be dehydrated (3 h ; 70" and 0.1 mmHg).It de-composes at ca. 155", and then melts 190-192" [Found(anhydrous material): C, 28.15; H, 1.8; N, 39.2.C,H,N,O, requires C, 28.3; H, 1.9; N, 39.6y0]; A,, ( 0 . 1 ~ -HC1) 341 (E 15 750) and 265sh nm (4 100) ; vmx. (of hydrate)3 530m, 3 420m, 1 745s, 1 680s, 1 635s, 1 505w, 1460s,1380w, 1340m, 1315m, 1250m, 1190w, 1035m, and980m cm-l.6-MetJzylfurazano[3,4-d]pyrimidine-5,7(4H,6H)-dione 1-Oxide (3; R1 = H, R2 = Me).-The tetrazolopyrimidine (18)(1.15 g) was heated for 18 h a t 120" and 0.3 mmHg. Theresidue was recrystallised from acetone-light petroleum(1 : l ) , and then from carbon tetrachloride, giving whiteplatelets of the furoxan (0.78 g, 85y0), m.p. 211-212" (de-camp.) (Found: C, 32.6; H, 2.2; X, 30.4.C,H,N,O,requiresC, 32.6; H, 2.2; N, 30.4%) ; A,,,. (dioxan) 312 (E 1 700) and267.5 nm (11 400); vmnx. 3 160w, 3 040w, 1 730m, 1 700s,1 650s, 1 600w, 1545w, 1 445ni, 1 380w, 1 335m, 1 275m,and 1 205m cm-l.N(4)-Alkylation of 6-methylfurazano[3,4-d]pyriniidine-5,7(4H,6H)-dione l-oxide (3 ; R1 = H, R2 = Me)Product (3; R1 = as indicated, R2 = Me)-l ~~~YieldHalide R1X M.P. ("C) (%I Recryst. solvent; form A,,,./nm (E) (in RleOH)Me1 245 70 95% EtOH-H,O; plates 317 (1 550), 268 (11 350)EtI 172-173 40 H,O ; platelets 317 (1 200), 269 (9 750)BunI 104-105 55 50% EtOH-H20; platelets 317 (1 550), 269 (12 300)H,C:CHCH,Br 88-90 45 H,O; plates 315 ( 1 ZOO), 268.5 (8 400)MeCOCH,Br 117-118e 30 H,O; needles 315 (1 ZOO), 267 (10 100)PhCOCH,Br 210-211 50 95% EtOH-H,O; needles 322 ( 2 600), 265 (12 850).248 (17 500)a All new compounds gave satisfactory analytical (C, H, N), i.r., and mass spectral data. Analytical data are available as Supple-mentary Publication No. SUP 21744 (2 pp.) ; for details of Supplementary Publications see Notice to Authors No. 7 (J.C.S. Pevhin I,1975, Index issue). Lit.,3 245". In ethanol; from ref. 3. Inflection. e Softens and partly decomposes ca. 109".aqueous sodium nitrite and hydrochloric acid, followingYoneda and Sakuma's Method A. The yellow solidproduct was separated (t.1.c.) into three components : thefuroxan (3; R1 = R2 = Me) (0.4 g, 40y0), the furazan (16)(0.2 g, 25y0), m.p. 225-226" (lit.,, 225-226"), and anunidentified solid (0.05 g), m.p.180".When the oxime acetate (17) (0.6 g) in acetic acid (1.5 nil)and water (10 ml) was treated dropwise with aqueoussodium nitrite (0.56 g in 6.7 ml) a t 0 "C, then stirred for1 h a t 20 O C , the brown mixture deposited a white crystal-line solid, which formed plates (from cthanol or acetone) ofthe furazan (16) (0.47 g, go%), m.p. 225-226", identicalwith that obtained in method (b).6-MethyZ-8-nitrotetrazolo[ lJ5-c]pyrimidine-5, 7( lH, 6H) -dione (18) .-l-Methylbarbituric acid l9 was converted into4-chloro-l-methyluracil,20 which was then nitrated,21 giving4-chloro-l-methyl-5-nitrouracil [' 6-chloro-2-hydroxy-3-methyl-5-nitro-4( 3H)-pyrimidinone ' 21]. The nitrouracil(2.05 g) was refluxed (7 h) with sodium azide (1.0 g) intetrahydrofuran (100 ml). After cooling, the solid wasfiltered off and washed with tetrahydrofuran (some nitro-uracil was recovered from the solvent).The remainingsolid was soluble in methanol, being precipitated byaddition of benzene. The material showed an azide i.r.band (2 140 cm-l), but i t appeared to isomerise readily tothe tetrazolopyrimidine (1 8 ) , a process which was hastenedby stirring with hydrochloric acid ( 2 ~ ; 25 ml). The whitesolid so obtained (1.7 g, 75%) was soluble in acetone, fromwhich it was precipitated by light petroleum. The tetra-zolopyrimidine formed needles of a monohydrate, whichW. PBeiderer and H. Ferch, Annalen, 1958, 615, 52.l9 H. Blitz and H. Wittek, Ber., 1921,54, 1035.2o G. Niibel and W. Pfleiderer, Chern.Ber., 1962, 95, 1613.Thermal decomposition of the tetrazolopyrimidine ( 18)could also be effected (75% yield) by reflux for 3 h in glacialacetic acid. The synthesis of (3; R1 = H, R2 = Me) wasalso achieved (66% yield from the chloronitrouracil) with-out purification of the intermediate (18).Attempts to effect condensation of the furoxans (3;R1 = H or Me, R2 = Me) with nitromethane or 2-nitro-propane, in tetrahydrofuran, pyridine, or dimethylform-amide, in the presence of triethylamine, gave no recognis-able products, with in most cases recovery of the furoxanin good yield. The tetrahydrofuran solutions from thereactions with the monomethyl furoxan deposited largeyellow prisms, m.p. 96-97", recrystallisable from carbontetrachloride (as needles), which were formed also in theabsence of the nitroalkane, and proved to be the triethyl-ammonium salt of the furoxan (3; R1 = H, R2 = Me)(Found: C, 46.5; H, 6.7; N, 24.5. C,,H,,N,O, requires C,46.6; H, 6.3; N, 24.1%).N (4) -A lkylation of 6-Metlzylfurazano[3,4-d]pyrimidine-5,7(4H, GH)-dione l-Oxide.-The following general pro-cedure was used to obtain the products detailed in the Table.The furazanopyrimidine oxide (3; R1 = H, R2 = Me)(0.46 g) in dry dimethylformamide (DhSF) (25 ml) wastreated with lithium hydride (0.02 g).22 The solution wasstirred a t 70 "C while the halide was dropped in over 10-15 min (in DMF solution if a solid). After further stirringat 70 "C for 1.5 h, the solvent was removed in vncuo andthe residual syrup was triturated with water (20 ml). The21 G. D. Davies, R. I(. Robins, and C. C. Cheng, J . Arner. Chenz.Soc., 1962, 84, 1724.22 Cf. J. A. Vida, Tetvnhedvon Letters, 1972, 3921. Other baseswere not tried; there is no reason to suppose they would be lesssuccessful1976 1331solid so obtained was filtered off and recrystallised asindicated. Purity of recrystallised materials was checkedby t.1.c.Haylett and mass spectra by Dr. B. Gregory. We thankthe Leverhulnie Trust for a Research Fellowship, and theUniversity of Timisoara (Romania) for leave of absence (toR. N.).Microanalyses were performed by Mr. A. W. R. Saunders;i1.m.r. spectra (100 MHz) were obtained by Mr. P. W. [5/2098 Received, 27th October, 1975
展开▼
