Heterocyclic studies. Part XXXV. Cleavage of fervenulin and 3-methylfervenulin by nucleophiles

摘要

1818 J.C.S. Perkin IHeterocyclic Studies. Part XXXV.' Cleavage of Fervenulin and 3-Methyl-fervenulin by NucleophilesBy Jim Clark and Michael S. Morton, The Ramage Laboratories, Department of Chemistry and AppliedTreatment of fervenulin {6,8-dirnethylpyrimido5,4-e -as-triazine-5,7(6H,8H) -dione) and 3-rnethylfervenulin withprimary amines gave 6-(3-alkyl-l -methylureido)-5-methylcarbarnoyltriazines by attack at position 7 and cleavageof the 6.7- bond. Reactions with hydrazine and 1.1 -dimethylhydrazine similarly gave triazin-6-ylsemicarbazides.Fervenulin and methylhydrazine gave 5,6-diamino-l,3-dimethyluracil and an unknown compound by cleavage ofthe triazine ring. The 6- (1 -rnethyl-3-alkylureido) -5-rnethylcarbamoyltriazines were converted by cold nitrousacid into open-chain compounds.Chemistry, University of Salford, Salford M5 4WT'H N.rn.r., mass, and U.V.spectra are tabulated and discussed.FERVENULIN (6,8-dimethylpyrimido5,4-e-as-triazine-5,7(6H,8H)-dione) is a crystalline antibiotic first isolatedfrom cultures of Streptomyces feYveizs and unwittinglysynthesised by Pfleiderer and Schundehutte before itsstructure was appreciated. This paper describes the1 Part XXXIV, J . Clark and I. bsol;V. Southon, prccedingpapcr.T. E. E l k , E. C. Olson, C. XI. Langc, and J. W. Shell,'Antibiotics Annual,' 1959-1961, Antibiotics lnc., Kew York,1960, p. 227.3 W. Pflciderer and K. H. Schundehutte, Anualrn, 1958, 615,42.reactions of fervenulin (2; R1 = H) and its 3-methylderivative (2; R1 = Me) with some nucleophiles, as anextension of our work on the cleavage of fused pyrimidinederivatives .4-8* J.Clark, G. Neath, and C . Smith, J. Chem. SOC. ( C ) , 1969,1297.J . Clark and G. Neath, J . Chem. SOC. ( C ) , 1966, 1112.J . Clark and G. Neath, J . Chem. SOC. (C), 1968, 919; J .J . Clark and C. Smith, J . Chem. SOC. (C), 1971, 1948.J . Clark and M. S. Morton, J . C . S . Pevki?t I , 1974, 1812.Clark and C. Smith, ibid., 1969, 27771974 1819Ferveriulin was prepared bj. the published methodbut we were unable to prepare its 3-methyl derivativeby the mctliod described in the same paper. Webelieve that the synthesis failed a t the stage where 6-(2-acetylhydrazino)-l,3-diniethyl-5-nitrosouracil (1 ; X =NO) was to be reduced to the amine (1 ; X = NH,) byhydrogen over Raney nickel.In our hands ammoniawas always produced, probably by hydrogenolysis ofthe hydrazino-group (cf. ref. 9). Reduction of thenitroso-compound (1 ; S := NO) with sodium dithioniteavoided the trouble and 3-methylfervenulin was obtainedon cyclisation and oxidation of the resulting amine (1;X = NH,).Me !!II 'YN NH*NH.CMe0 ( 1 13 ( 5 ) 0 0( 6 )Treatment of fervenuliri and its 3-methyl derivativewith primary amines yielded crystalline products of ringcleavage (Table 1) which were assigned N-triazin-6-ylurea structures (4; R1 = H or Me, R2 == Me, Et,CHMe,, ;CHamp;,CH,, or CH,Ph) in preference to the likelyalternatives (5) and (6). The isomers (4) and (6) couldbe formed by attack at the 7-position followed bycleavage of the 6,7- or :amp;bond, respectively; isomer( 5 ) coulcl he formed by attack at the 5-position andcleavage of the 5,G-bond.Me/ aR2NH2 / ( 2 1H MP 0( 3 )0(4 1The niet t1~laxnino-structure (6) was eliminated byU.V.spectroscopy (Table 2) ; the compounds formedshowed no absorption at wavelengths greater than276 nm. The spectra of compounds (6) would havebeen similar to those of 6-amino-5-carbamoyl-as-triazine( 7 ; R == H),l0 its dimethyl derivative (7; R = Ale) (seeC. Ainsworth, J . .4nwv. Clirm. SOC., 1954, 76, 5774.C 'I'cmplc. C. I,. Kussncr, and J. A. Montgomrry, J . Ovg.Chrm , lW!). 34, 2102.later), and the closely related pyrazine derivatives (8;Y = X = R r: H) l1 and (8; X = R = H, Y = Cl),l*which have long wavelength absorptions in the 350-380 nm region (Table 2).Acylation of the amino-groupof these compounds causes marked hypsochromic shiftsso that acylamino-derivatives, e.g. (8; X = H, R r=CHO, Y = C1) and (8; Y = X = H, R = CHO) l3absorb at much shorter wavelengths, as do the com-pounds under investigation.The mass spectra of the compounds were very simpleand strongly supported the proposed structures (4).Molecular ions were absent but all the compounds (4;R1 = H) from fervenulin gave a dominant fragment ionof m/e 166 (base peak) and all those (4; R1 = Me) from3-methylfervenulin gave a corresponding ion of m/e 180(base peak). There was usually no ion of any significanceat a higher m/e value but at high source pressures verysmall ( M t- 1)+ peaks (relative abundance 0-20/)appeared.The ions of m/e 166 ( b ; R1 = H) were shownby accurate mass measurement to have the compositionC6H,N,0 and those of m/e 180 ( b ; R1 = Me) C,H,,,N,O,and were therefore formed by loss of R2NHC0 from the( a I0 1 6 )- McNH*C@CN i +Me-N=C=N-N=C-R'/ C ISCHEME 2n~olecular ions. A peak at m/e 82 (C,H,K,) and anassociated metastable peak (m* 40.5) showed that ion( b ; R1 = H) lost MeKHCOCN to yield ion (c; R1 = H)(Scheme 2). Ions of tn/e 180 from the methyl derivativesdecomposed in a similar fashion to give ions of m/e 96(c; R1 := Me).lH K.m.r. spectra (Table 3) confirmed that the triazinering was intact and were consistent with the proposedstructures (4), but did not help materially in ruling outthe alternative structures ( 5 ) .As a whole the foregoing evidence made it virtuallycertain that the products were the triazin-6-ylureas (4),formed by nucleophilic attack at position 7 of the11 D.D. Pcrrin, J . Ckrm. Soc., 1962, 645.12 A. Albcrt and J . Clark, J . Chem. Soc., 1961, 1666.13 A. Albert and H. I'smamoto, J . CJmn. SOC. (C), 1968, 22891820 J.C.S. Perkin Ipyrimidotriazines (2) (Scheme 1). This is in accord withthe fact that alkaline ring cleavages of closely relatedpyrimido 4,5-e-as-triazines (10) yield compounds such asthe methylamide (ll), which must result from attack atthe corresponding position of the pyrimidine ring.14Furthermore the reported cases of nucleophilic attack on1,3-dimethyl-lurnazines (9) occur at position 2, whichagain corresponds to position 7 of our compound^.^^ l5The pyrimidotriazines (2; R1 = H or Me) were bothcleaved by hydrazine even at 20" to give similar triazinederivatives whose molecular formulae suggested thatthey were analogues (4; R2 = NH,) of the cleavageproducts already discussed. Fervenulin also reactedwith 1,l-dimethylhydrazine, at loo", to give a triazine,but the 3-methyl compound (2; R = Me) was un-affected by treatment with dimethylhydrazine at 100"for 4 days.Neither pyrimidotriazine reacted withhydroxylamine or methoxyamine at 100".The structure of the product from fervenulin and1,l-dimethylhydrazine (4; R1 = H, R2 = NMe,) wasclearly analogous to those of products of ring-opening byamines.The mass spectrum, with major peaks at m/e166 and 82 linked by a metastable peak (m* 40.5) wascharacteristic, and the lH n.m.r. spectrum (Table 3)showed signals for the two methyl groups and thetriazine proton of the original molecule at very similarchemical shifts to those of the previous triazinylureas(4; R1 = H, R2 = alkyl).The product of cleavage by hydrazine hydrate alsoproved to be a semicarbazide of type (4; R1 = H,R2 = NH,) rather than the other possible isomer (12;R1 = H). I t reacted with nitrous acid in dilute aceticacid to regenerate fervenulin, presumably via the azide(13), and with nitrous acid in hydrochloric acid to givethe methylamide (7; R = Me). Fervenulin could havebeen formed from either isomer r(4; R1 = H, R2 =NH,) or (12; R1 r= H) but the methylamide could onlyhave come from the former.The possibility that it wasformed from the isomer (12; R1 = H) by cyclisation tofervenulin and then ring-cleavage of the latter waseliminated when fervenulin was shown to be stable tonitrous acid.The structure of the methylamide (7; R = Me) wasfirmly established by its U.V. spectrum (Table 2) and byits mass spectrum, which showed losses of 28 (N,) and27 (HCN) mass units, characteristic of the 1,2,4-triazinering,16 and a peak at m/e 82 (MeN=C=N-kCeuro;euro;), expectedby analogy with the other compounds discussed.The mass spectrum of the semicarbazide (4; R1 = H,R2 = NH,) differed from those of all the other cleavageproducts (4; R1 = H) in showing a dominant peak atm/e 167 (C,H,N,O) instead of 166.The loss of MeNHCOl4 L. Heinisch, J. prakt. Chem., 1969, 311, 438; Chem. Bey.,1967, 100, 893; L. Heinisch, W. Ozegowski, and M. Miihlstadt,ibid., 1964, 97, 5.l6 A. Albert, D. J . Brown, and H . C. S. Wood, J . Chem. SOC.,1956,2066; G. P. G. Dick, H. C. S. Wood, and W. R. Logan, ibid.,p. 2131 ; E. C. Taylor, H. M. Loux, E. A. Falco, and G . H. Hit-chings, J. Amer. Chem. SOC., 1965, 77, 2243; E. C. Taylor, C. K.Cain, and H. M. Loux, ibid., 1954, 76, 1874.from the molecular ion (d) is explained by formation of avery stable ion (e) (Scheme 3) which could only beMe( 8 ) ( 9 )0( 7 1MeMe OYyJXO (101* ( 1 4 )MeNH*i MeNHo NGN0( 1 1 1I131 0produced from a compound bearing the NH*NH, group.In agreement, the product (4; R1 = Me, R2 = NH,)HSCHEME 3broke down in similar fashion to give a dominant peak atm/e 181.Although the mass spectrum could equally well beexplained by loss of MeNHCO from the molecular ion ofthe isomer (12; R1 = H), and hydrazine does sometimesgive cleavages inconsistent with those from other re-agents,* the other evidence shows conclusively thathydrazine attacks at position 7, like other reagents.Treatment of fervenulin with methylhydrazine gaveunexpected results. In benzene solution 5,G-diamino-1,3-dimethyluracil (14) was obtained by an apparentpreferential cleavage of the pyrazine ring.In ethanolicsolution a product, C8HllN502, of uncertain structurebut probably (15) or an isomer with an intact pyrimidinering was formed.3-Methylfervenulin was unchangedby similar treatment with methylhydrazine except forformation of a water soluble x-complex, which revertedto 3-methylfervenulin on acidification.16 W. A. Paudler and R. E. Herbener, J. Heterocyclic Chem.,1967, 4, 224; T. Sasaki, K. Minamoto, M. Nishikawa, and T.Shima, Tetrahedron, 1969, 2!5, 1021 ; W. Paudler and T. K. Chen,J . Heterocyclic Chem., 1971, 8, 3171974 1821Further work is necessary to elucidate the apparentpreferential triazine ring cleavage reactions, but it seemsunlikely that methylhydrazine alone specifically attacksthe triazine ring so a sequence of cleavages and re-cyclisations involving both rings may be involved.Certain as-triazine ring cleavage reactions are known l7and indeed the triazinylureas (4; R1 = H) proved to beamenable to ring-opening.They were stable to boilinghydrazine hydrate, sodium carbonate, and ethanolichydrogen chloride, but were degraded by cold nitrousacid with loss of two nitrogen atoms and gain of twooxygen atoms to give open-chain compounds which wereassigned structures (17). Under similar conditions the3-methyltriazinylureas (4; R = Me) only gave nitratesalts of the unchanged ureas. This suggests that ring-cleavage involves attack at the 3-position a t some stageand is inhibited by the 3-methyl group.Assignment of structures (17) was based mainly onresults of microanalysis, accurate mass measurements,and lH n.m.r. spectra (Table 3). There was no strongabsorption at wavelengths above 225 nm in the u.v.,MeNH-CO*N*HCOI f I I g 1SCHEME 4H I( 1 6 )Me2 N H .c o * N Y oMeNH.COAN -CHO( 1 7 )SCHEME 6and i.r. spectra contained several peaks in the carbonyl(1630-1790 cm-l) and the v m (ca. 3300 cm-l) regions.lH N.m.r. spectra showed one low-field proton ( T ca. 1 4 ) ,two N-methyl groups, two exchangeable protons, andsignals appropriate for RNHCO. Mass spectra showeda loss of RNCO in each case, which was readily explainedby a McLafferty rearrangement (f) - (g) (Scheme 4)and subsequent losses of H20 (D20 if the NH groups werefirst converted to ND) and HCN from ion (g). Alter-natively (g) lost CHO. These losses, which werepostulated on the basis of metastable peaks and massmeasurements, can be accounted for by a series ofrearrangements and cleavages.The function of nitrous acid in the triazine ring-cleavages (Scheme 5) may be to prevent recyclisation andassist hydrolysis of an intermediate (16) by nitrosation.This would explain why the system is stable to hydro-chloric acid but cleaved by nitrous acid.EXPERIMENTALU.V.spectra were measured on a Unicam SP 800 spectro-meter.Mass Spectra.-These were measured on A.E.I. MS12and MS902S spectrometers with source temperature ca.220", ionising voltage 70 eV, and resolving power ca. 1000(low resolution spectra) or 10,000 (accurate mass measure-ments). Samples were introduced on direct insertionprobes. (a) Triazine derivatives (relative abundance 25 ;m/e 2 4 0 ) : (4; R1 = H, R2 = Bu) m/e 167 (9), 166 (loo),83 (7), and 82 (14); (4; R1 = Me, R2 = Bu) m/e 181 (lo),180 (loo), 96 (9), and 42 (8); (4; R1 = H, R2 = NMe,)m/e 167 (lo), 166 (loo), 83 (8), 82 (13), 67 (5), 60 (6), 59 (6).and 42 (8); (4; R1 = H, R2 = NH,) m/e 168 (9), 167 (100).166 (5), 82 (6), 70 (9), 58 (7), 57 (5), and 42 (5); (4; R1 =Me, R2 = NH,) m/e 182 (9), 181 (loo), 180 (8), 168 (8).167 (14), 166 (5), 84 (6), 82 (6), 67 (8), 57 (12), 43 (6), 42 (27),and 41 (9); (7; R = Me) (210) m/e 167 (45), 139 (10).112 (18), 111 (18), 82 (loo), 81 ( l l ) , 58 ( l o ) , 55 (23), 54 (12).and 41 (11).(b) Ketones (loyo): (17; R2 = Et) m/e171 (54), 153 (36), 144 (16), 143 (29), 142 (loo), 128 (10).126 (16), 69 (16), 58 (36), 57 (72), 56 (22), 44 (16), 42 (22),and41 (10); (17; R2 = Me) m/e 171 (36), 170 (14), 153 (15).143 (16), 142 (loo), 126 (lo), 83 (13), 69 (14), 58 (36).57 (go), 56 (16), 55 (18), 43 (17), 42 (24), and 41 (21).3,5,7-Trimethyl~yrimido5,4-e-as-triazine-5,7 (6H, 8H) -dione.- 6- (2- Acetylhydrazino) - 1,3-dimethyl-5-nitroso-uracil (3.8 g ) was dissolved in formamide (23 ml) andformic acid (15 ml) a t 100" and reduced by the addition ofsmall portions of sodium dithionite until the colour changedto olive green.The solution was boiled for 30 min, cooled,and poured into water (80 ml) . The product was extractedwith chloroform and the extract was dried (MgSO,) andevaporated t o dryness by a stream of dry air. The resultingoil was chromatographed on a silica column (ethyl acetateas eluant) to yield the pyrimidotriazine (1.5 g ) , m.p.124'(lit.,3 127O).3-Alkyl- l-methyl-l-(5-methylcarbamoyl- 1,2,4-triazin-6-y1)-ureas) .- 5, 7-Dimethylpyrimido5,4-e-as-triazine-5,7-(6H,8H)-dione (fervenulin) (0.25 g) or its 3-methyl derivative(0.26 g), ethanol (10 ml), and the appropriate amine (1 ml)were heated under reflux for 24 h (48 h for benzylamine).Methylamine and ethylamine were in the form of 33aqueous solutions of which 1 ml was used initially and asecond portion (1 ml) was added after several hours underreflux. The mixture was evaporated to dryness underreduced pressure and the triazin-6-ylurea crystallised froma suitable solvent (Table 1).4-Methyl-4- (5-methylcarbamoyl- 1,2,4-triazin-6-yl)serni-carbazide (4; R1 = H; R2 = NH,).-(a) Fervenulin (0.5 g),l7 J.P. Horwitz in ' Heterocyclic Compounds,' ed. R. C.Elderfield, Wiley, New York and London. 1961, vol. 7, p. 729;J. Klosa, Avch. Pharm., 1966, 228, 4661822TABLE I3-Xlkyl- I-methyl- 1-( 5-methylcarbamoyl- 1,2,4-triazin-6-~1) ureasJ.C.S. Perkin1;ound ("b)C H S42.6 5.5 37.945.5 5.848.2 6.249-8 6-6 32.255.6 5.4 28.442.3 6.2 33.347.7 5.9 33.746.1 7.0 29.451.6 6*!) 30.157.3 5.77 ---L--100") ; B r- I~cIlZ'ne.Required ()r----h---C H K42.9 5.4 37-545.4 5.947.6 6.449.6 6.8 31.656.0 5.4 28.042.2 6.3 32.847.6 6.4 33.346.5 7.1 29.651.4 7.2 30.067.3 5.8ethanol (10 i d ) , and 99Oi;, hydrazine hydrate (2.5 ml) werestirred at 30" for 48 h.The semicarbazide (0.37 g), imp.213--215", was filtered off and crystallised from water(Found: C, 37.1; H, 4-6; N, 43.4. C,H11N,02 requiresC, 37.3; H, 4-9; N, 43.5?/0).TABLE 2U.V. data (H,O)Coinpound L I x . Inn1(2; I' = H) 342,275,238(2; R1 = Me) 349, 275, 238(4; R' = H, 13' = Et) 276, (225(4; R1 : M e , R2 = Et) 272, 225( 7 ; R = Jc) 375, 428(7; I = 11) 358, 239(8; = x =- y L7 13) 11 371, 263(8; K = CHO, X -: Y == H) l3(8;325, 310, 266,235388, 270,I = Y :r H , X = XH-NH,) 5 352,248(8; I = X 7: H, Y = Cl) 12217 infl( 8 ; 3i .:. CHO, X = H, Y = C1) '' 316, 272, 245log cmax.3-61, 3.22, 4.233-66, 3.22, 4.323.313.353.49, 4.093.59, 4.063.83, 3-853-61, 3.75,3.99, 3.843.82, 4.063-80, 3.96, 3-833-73, 3-89, 4.00( b ) Fervenulin (0.2 g) and 997; liydrazine hydrate ( I nil)were stirred at 20" for 45 min; the liydrazide (0.025 g),m.p.213--215", was filtered off and shown to be identicalbsol;bsol;it11 the previous specimen.The semicarbazide (0-05 g) was suspended in cold waterand added dropwise to an ice-cold saturated solution ofsodium nitrite in -acetic acid. Tlic yellow solution wasimmediately extracted with chloroform ( 3 x 20 ml) andthe extract dried (MgSO,) and evaporated to dq-ness, underrcduced pressure, to yield fervenuliii (0.0 1 g) from benzene-light pctroleuni (b.p. 80---100deg;), identical with an authenticspecimen.~~-~~.letJiylniiiino-5-metJiylcarbai~~o~l- I ,2,4-tvzazine.- Theforegoing semicarbazide (0.1 g) was tlissolvcd in 4x-hydro-chloric acid ( 5 ml) and cooled in ice..An excess (starcli--iodide paper) of similarly cooled aqueous sodium nitritc:was added and the mixture Ivas kept on ice for 10 inin antlthen stirred a t room temperature for 0.5 11. 'I'hc niixturewas inale strongly alkalinc with 2x-sodiuni !iytlroxide antlextracted with chloroform ( 3 :.; I 0 1111). The extract wasCompound(4; R1 = H, 1X2 = Me)(4; R1 = H, 12, = Pr*)(4; R' = H, Itz == Et)(4; 13.1 = H, I' == Bun)(4; R' = H, 1' == CH,Ph)(4; R' = R2 = Me)(4; R1 = Me, Ii2 = Et)(4; R1 = hlc, R2 = I+*)(4; R1 = Me, R2 = 13un)(4; R' r= Me, 13' =: CH,Ph)(4; R' -L H, 1Z2 = NH,)(4; R1 = -Me, R2 = NH,)(17; R2 = Me)(17; 11, = Prl)(4; R' H, K2 1 h'M e,)(17; R2 = ICt)(17; 11' = Bun)(17; Re = PhCH,)TABLE 3lH S.m.r.data a7 bsol;-alucs (J in 132)I --- - --- ~ MeN HCO li21?u'HCOS,1Se CHO Solvent R' 1 1 2CD,OI) 2-35 (lH, S ) 7.27 (3H, S ) 7.08 6.93CD,OD 2.33 (lH, S) 8.90 (3H, t, J 7.5). 6.80 (2H, q, J 7-5) 7.07 6.93C5DJ 2.33 (lH, S) 8.88 (3H, d, J 7), 8.83 (3H, d, J 7), 6.87 6.77CD,OI) 2-30 ( l H , s) 8-33-9.28 (SH, m), 6.77 (2H q, J 7.5) 7.07 6.925.70 ( l H , sept, J 7)(CD3),S0 2.36 ( l H , s) 5.73 (2H, d, J 4), 2-68 (5H, m), 7.22 747CD,OI) 7.82 (3H, S) 7.26 (3H, S) 7.10 6.95CD,OD 7-80 (3H, s) 8.88 (3H, t, J i . 5 ) , 6.73 (2H, q, J 7.5) 7.05 6.92CD,OD 7.80 (3H, s) 8-33-9.32 (7H, m), 6-76 (2H, (1, J 7.5) 7-06 6-93(CD,),SO 2.62 ( l H , s) 4*35br,~ 2.40br,e 1.60br c 7.48 6-93(CD,),SO 8.17 (lH, s) 4-65br,C 1.90br,c 1-42br 7.50 6.97(CD3)zSO 2.45 ( l H , s) 7.57 (6H, s), l-BObr,C -0.78br c 7-1 8 7.08 d(CD3)2S0 8.98 (3H, t, J 7.5), 6.77 (2i-I, m), 7.38 7.10 1 -851.45br (1H) c(CD,),SO 7.95 (3H.S) 8.99 (3H, d, J 7), 8.83 (311, (1, J 7), 7.2:; 7 - 1 06.16 ( l H , m)1-41br (1H) cCDCI, 7.88 (3H, S) 5.58 (2H, d , J 6), 2.63 (5H, s ) , 7.08 6-96(CD3)2S0 7-36 (3H, d, J 5 ) , 1*50br," 0-98br e 7.38 i - 1 0 1.861.41br,c l.Obr(CD,) ZCO- 8-88 (6H, m), 5.97 ( l H , ni), 1.16 br c 7.32 7.05 1.79CDCI, 8.33-9.13 (7H, m), 6.48-6-73 (2H, m),C 7.17 6.90 1 *62CDCI, 5.50 (2H, d, J 6), 2.55 (5H, s), 2-55," 7.25 6.98 1.63(CD,) ,SO2.30br,c 3-1 1 br c2.30br Ca 1Ieasured on a Varian .460A spectrometer at normal probe temperature using tetramcthylsilanc 3s intcrnal standard.b SingletSignals due to residual ethanol (see Experimental (3H) assignments of methyls may be reversed.section) omitted.c Removed on deuteriation.I3ecomes triplet on cleuteriation1974 1823dried (AIgSo,) and evaporated to yield the yellow methyl-aniide (0.02 g), n1.p. 132--134" from light petroleum (b.p.80--100") (Found: C, 42.0; H, 5.4; K, 42.1. C,H,N,Orequires C, 43. I ; H, 5.4; S, 41.9o/b).4-Methyl-4- (3-inef Iiyl-5-methylcarbamoyl- 1,2,4-triazin-0-y1)sevlticnrbazitle ( 4 ; K1 = Me, R2 = XH,) .-3-Methylfer-venulin (0.5 g ) , cthanol (50 nil), and 99?amp; hydrazine hydrate(2.5 ml) were stirred at 20" for 48 11. The seinicarbaside(0.18 g) was filtered off and crystallised from aqueousethanol; m.p. 207" (1:ound: C, 40.0; H, 5.4; N, 40.8.C,H,,N,O, requires C, 40.15; H, 5.5; N, 41.0).l,l,4-Tri~~iet~i~yl-4-(5-n~e~liylcarbamoyZ- 1,2,4-triazin-6-yl)-seiiiicarbazitle (4; R1 = H, R2 = NMe2).-(a) Fervenulin(0.5 g), l,l-tlinietli~lliy~2razine (2.5 ml), and ethanol (10 ml)were heated on a boiling water-bath for 36 h and the solutionwas then evaporated to dryness under reduced pressure.Chloroform-soluble material was extracted and the whiteunder reflux for 2 11.5,6-Diamino-1,3-dimetliyluracil (0.08g), m.p. 207-209" (from ethanol-ethyl acetate) lit. ,la209" (decomp.), was filtered from the cooled solution andwas identical with an authentic specimen.( b ) Fervenulin (0.2 g) and niethylhydrazine ( 1 nil) werestirred a t 20" for 2 h and the mixture was evaporated todryness. The residue crystallised from water as paleyellow granules, n1.p.234-235" (Found: C, 45.2; H, 5.2;N, 33.4. Calc. for CsHl1X5O2: C, 45.9; H, 5.3; S, 34-5~0).(c) Fervenulin (0.5 g), ethanol (35 nil), and niethyl-hydrazine (2.5 ml) were stirred at 20" for 48 11. The mixturewas evaporated to dryness and the residue was crystallisedfrom water to give a product (0-2 g), m.p. 234-235",identical with that described in ( b ) .Treatment of l'riazinylureus with Nityous -4 cid.---Theappropriate triazinylurea (4) (0.1 g ) was dissolved in 4 ~ -hydrochloric acid (5 ml) and cooled to 0". Siniilarly cooledTABLE 4Ureitlopropane derivatives ( I 7)Found (7;) Required (O,,)A Y'iclrl Cryst. f 1(()amp;bsol; M.p. ("C) solvent Formula C H N c euro;I NIt - Me 22 266-268 EtOH C8H 1 zN4', 42.2 5.3 24.6 42.1 5.3 24.6R : Et 60 2 2 8-2 30 EtOH C,H 14NQ_O I 44.9 5.8 23.0 44.6 5.8 23.1I 2 Pr' 56 242-244 EtOH c10H16h -l04 46.7 6.1 21.6 46.9 6-3 21.9R = I3u" 40 171-173 PhH CllHl,N,O, 48.4 0.4 20.6 48.9 6.7 20.71t :: CH,Ph 6 5 194-1 96 PhH C14H16N404 54.8 5.2 18.4 55.3 6-:1 18.4TABLE 53-Alkyl- 1 -methyl- l-(3-niethyl-5-niethylcarbamoyl- 1,2,4-triazin-6-y1) urea nitratesFound ( S o ) Required ( o/o)1 Yield 3I.p.("C) -- ~ p--h-C~JIllpOUI~d (I (decamp.) Formula C H N C H K( 4 ; 13' = R2 L; 'Me),HK03 40 161-162 CBHldN602,HN03 36.3 5.1 32.2 35.9 5.0 32.5(4; I' = M e , I t 2 = l:t),HN03 40 158-160 CloH16N602,HN03 37.5 5.3 30.2 38.1 5.4 31.1(4; K1 = Me, It2 = Pr*),HNOa 57 148 CllHl8N6O2,HNO3 39.8 5.8 29.8 40.1 5-8 29.8(4; R' = hle, It2 = Bu"),HNO, 60 146 C12H2,N602,HN03 42.2 6.1 28.3 42.0 6.2 28.6(4; H' = Mc, R2 =: PhCH,),HSO, 32 154-155 Cl,H,,N60,,H~0, 47.8 4.8 25.3 47.7 5.1 26.0residue crystallised from etlianol-ethyl acetate to yield theseinicurbazide (0.2 g), n1.p.200-202" (Found: C, 43.1;H, 6.3; N, 35.8. CSH,,N,O,,O.5C,H5OH requires C , 43.5;H, 6.6; N, 3.5~4~;).(b) Fervenulin (0.5 g ) and 1,l-dimethylhydrazine (2.5 ml)were heated under reflus for 4 11 before more dimethyl-hydrazine (2.5 nil) was added, and then for a further 20 h.The semicarbazide (0.3 g), 1n.p. 201" (from propan-2-01),was filtered from the cooled solution and shown to beidentical with the previous sample except that elementalanalysis and lH n.m.r. spectroscopy showed the presenceof a little propan-2-01 i n the crystals (Found: C,, 43.7;H, 6.3; N, 35.8. CsH,,N,02,0.33C,H,0 requires C, 43.9;saturated aqueous sodium nitrite was added in excess anclthe mixture was stirred a t 20" for 15-30 min. I n the caseof each 3-unsubstituted triazine (4; R = H) the solutionwas extracted several times with chloroform and the extractwas dried (MgSO,) and evaporated to dryness. Theproduct was crystallised from an appropriate solvent(Table 4). For 3-niethyltriazines (4; R = Me) the aqueousreaction misture was evaporated to dryness and extractedwith boiling acetone. The nitrate salt was crystallisetl fromethanol-ethyl acetate-light petroleum (b.p. 80-100")(Table 5). Similar salts were obtained by allowing solutionsof the :3-niethyltriazinylureas, in dilute nitric acid, toevaporate.l8 I;. F. I3licke and H. C . Goclt, J . Amei,. CJzein. Soc., 1954, 76,2798. 4/418 Received, 4th ilfarch, 1974