Mechanistic studies in the chemistry of urea. Part 8. Reactions of urea, 1-methylurea, and 1,3-dimethylurea with some acyloins and butane-2,3-dione (diacetyl) in acid solution

摘要

J.C.S. Perkin I1 Mechanistic Studies in the Chemistry of Urea. Part 8.' Reactions of Urea, I -Methylurea, and 1,3-Dimethylurea with Some Acyloins and Butane-2,3-dione (Diacetyl) in Acid Solution By Anthony R. Butler * and lshtiaq Hussain, Department of Chemistry, The University, St. Andrews, Fife KY16 9ST Urea and 1-methylurea react with both aromatic and aliphatic acyloins to form 4-imidazolin-2-ones but, under the same reaction conditions, 1,3-dimethylurea does not react. However, 1,3-dimethylurea does react with diacetyl to form 4,4'-methylenebis- (1,3,5-trimethyl-4-imidazolin-2-one) (4a). The carbon atom lost in the formation of (4a) is eliminated as formaldehyde. We propose a reaction mechanism which involves formation of an oxetan ring. With 1 -methylurea and urea the main products of reaction with diacetyl are the bicyclic compounds (6a-c).However, it is possible that reactions analogous to that leading to the formation of (4a) also occur and that the carbonium ion (5c) is the coloured species formed in the well known colorimetric procedure for the determination of urea concentrations in biological liquids. INthis series of reports on the reactions of urea and AT-alkylureas with dicarbonyl compounds we come now to consider reactions with butane-'3,3-dione (diacetyl). The unexpected products obtained posed a number of mechanistic problems and occasioned a study of the acid- catalysed reactions of urea and N-methylureas with a number of acyloins. These simpler reactions will be reported first.RESULTS AND DISCUSSION Acyloins are known 293 to react with ureas in the presence of an acid to give 4-imidazolin-2-ones but the success of this reaction appears to depend upon both N-niethylation of the urea and substitution of the acyloin. We therefore investigated the reactions of urea, 1-methylurea, and 1,3--dimethylurea with various acyloins under standard conditions, i.e. in refluxing benzene with added trifluoroacetic acid (TFA). Detailed spectral data of the products are given in the Experimental section. (a) Acetoin (la).-Reaction with urea. proceeds readily to give the expected product 4,5-dime thyl-4-imidazolin-2-one (2a).2 With 1-methylurea we obtained 1,4,5-trimethyl-4-imidazolin-2-one (2b), which was identified mainly from the 13C n.m.r.spectrum which showed the present of two nearly identical methyl groups, one N-methyl group, two olefinic carbons with slightly different shifts, and one carbonyl carbon. The formation of (2b) does not agree with the statement ' the acyloins of the aliphatic series (which) fail to react with sub- stituted ureas '. However, there was no reaction between acetoin and 1,3-diniethylurea. (b) Pro$ionoin (Ib) .--Reaction with urea gave the expected product, viz. (2c), which was identified by spectral data. With 1-methyl- and 1,3-dimethyl-urea unreacted starting material was recovered. Lack of reaction with the former, when compared to reaction with acetoin, must be due to increased steric factors. (c) a-Hydroxybenzyl Methyl Ketone (lc) .--Keaction with urea gave the expected product (2d).All the spectral data are consistent with this structure. There was also reaction with 1-methvlurea but. as the acvloin is unsymmetrical, there is the problem of the position of the N-methyl group. In the IH n.m.r. spectrum of (2b) the protons of the N-methyl group have a chemicd shift of 8 3.12, while those of the present compound are at S 2.48. This upfield shift suggests that the methyl group is adjacent to the phenyl group with consequent magnetic shielding. The product is, therefore, (2e). 0 OH 1111 2R-C-CH-R (1) a; R'=R* =Me b; A' =R2 = Et c; R' =Me, R2=Ph d; R' = RZ=Ph R57__rR4 R'h iR3Y All the other spectral data are in agreement with this assignment.1amp;Dimethylurea did not react with a-hydroxybenzyl methyl ketone. (d) Benzoin (Id).-Reaction with urea gave a product which was identical with that (3a)obtained by reaction of urea and benzil. This had been prepared and character- ised previ~usly.~Benzoin is known to be oxidised readily to benzil and this is the simplest explanation of the above observation. With 1-methvlurea the situ- ation is more complicated. The product, which is insoluble in every solvent suitable for recrystallisation, did not melt sharply. This suggested that it was a mixture. Bend is known to react with 1 -methylurea to give (3b) but the IH n.m.r. spectrum of the product had chemical shifts of S 2.96 and 3.14, corresponding to two different X-methyl groups, only one of which is right for (3b).We assume that this is due to concomi-tant formation of (2f) and (3b). The mass spectrum of the product had peaks at ~n/e250 and 322, which are molecular ion peaks for the two products. Other spectral data were not diagnostic. It appears that reaction of benzoin with I-methylurea occurs at a rate similar to that of oxidation to benzil. Under the experiinen tal conditions there was no reaction between benzoin and 1,3-dimethylurea, al- tbougli it is claimed that reaction can occur to give 1,3- din~ethyl-4,5-diphenyl-4-imidazolin-2-one. Our results do not support the view that, in reaction with N-sub- stituted ureas, acylojns of the aliphatic series react differently from those of the aromatic.The simplest 0 Ph-C-C-Ph R31NyNR4 I 0 R~-C =c-CH, -c =c Rrsquo; N~NR, R~N~NR~II I1 0 explanation for the lack ol reactivity of 1,3-dimethylurea is a steric one. We propose a general mechanism, shown in Scheme 1, for the reaction of urea with acyloins in the presence of an acid. As with other reactions in this series, the driving force is the elimination of the elements of water whenever this is possible. Tn Scheme 1 we have shown attack of the hydroxy-bearing carbon atoms preceding attack at the other carbon. The reverse may occur, but the last step in both cases is loss of the elements of water to form a double bond. As formation of (2e) occurs, 311 rather than the other isomer, it appears that the greater basicity of the inethylated nitrogen is more significant than any steric hindrance.All the imidazolin-2-ones prepared, apart froni (Zf), gave intense colours in acid solution. We are not, at present, able to explain this effect. bsol;lsquo;Ere now turn to a consideration of the reaction of 1,3-dime thylurea with butane-2,3-dione. Refluxing the 0 OH 4 ti I R --C-CH-R5 bsol;*.Rrsquo; HN-CQNH, OHrsquo; 0 + 4 II 5R-C-CH-R -t H,O+IRrsquo; N-CONH, 0 OHI R4-C -CH-R HN~NRrsquo; II11 HN NRrsquo; 0 0 SCHEME1 reactants in benzene and TFA resulted in formation of water and a purple solution. Removal of solvent left a viscous mass, which was purified by column chromato- graphy to yield crystals. This material has a molecular formula of C,,H,,N,O, and must be formed from the condensation of two molecules of 1,dimethylurea and two molecules of butane-2,3-dione with elimination of four molecules of water and olze carbon atom.Tn Part 5 we reportedG that a similar reaction occurred between l-phenylpropane-l,2-dioneand 1-methylurea, and Kuono and Ueda reported a similar reaction with t-butylurea. The i.r. spectrum of the product suggested the presence of carbonyl groups and carbon-carbon double bonds. The 360 MHz lH n.m.r. spectrum was unexpectedly simple: two pairs of N-methyl groups (6 2.89 and 2.81), two identical methyl groups (6 1.68), and a methylene group at low field (6 3.25). This gives the correct total of 20 protons. The n.m.r. spectrurn was equally simple, with only six resonances; this suggested a symmetrical molecule. Apart from the peaks assigned to the methyl groups, the N-methyl groups, and the carbonyl groups, there are two others at S 112.60 and 115.03 p.p.m.These are correct for carbons of an unsymmetrical double bond. The remaining resonance was at 8 19.00 p.p.m., and in the off-resonance spectrum became a triplet and was identified as that due to the methylene group. As the peak in the IH n.m.r. spectrum is not split by geminal coupling, the methylene group appears not to be part of a ring. There are several structures of different degrees of credibility, analogous to those shown in Part 5,6 which are consistent with the spectral data. Fortunately discussion of their relative merits can be avoided as we have an X-ray crystal structure of the product.8 It is 4,4'-methylenebis-(l,3,5-trimethyl-4-imidazolin-2-one) (4a).This molecule has all the features deduced from the spectral data; it is also symmetrical. We can now propose a definite structure for the product of reaction of 1-methylurea and 1-phenylpropane-l,2- dione, which we were not able to do when Part 5 was written. The proposed structure is (4) with R5=R6 = Ph, but it is difficult to be certain about the relative positions of the two N-methyl groups. There are three possibilities, (4b-d). In the 360 MHz lH n.m.r. spectrum two different N-methyl groups are indicated with 8 1.19 and 1.50. This could mean (4d) or a mixture of (4b and c). As the integrals for the two peaks are different, (4d) is not possible and we must have a mixture containing unequal amounts of (4b and c).This is confirmed by two unequal peaks at 8 3.01 and 3.13, corresponding to two different methylene groups. The discussion (see later) concerning the mechanism of form-ation of (44 applies to that of (4b and c). The reaction mixture from which (4a) was obtained was deep purple and we found we could generate this coloured material from (4a). Acidification of an aqueous solution gave a purple solution, the intensity of which increased with time. The rate of colour intensific- ation was increased by blowing oxygen through the solution and by addition of an oxidising agent. Com-pound (4a) is a ' skipped ' diene and therefore readily susceptible to radical oxidation at the methylene group to give a hydr~peroxide.~ Protonation of this and elimination of hydrogen peroxide would give the car- bonium ion (5a), which is related to the cyanine dyes.The positive charge can be delocalised on one of the two outer nitrogens with formation of a conjugated, chromo- phoric system. This, we suggest, is the origin of the colour. The reaction is analogous to the generation of the tropylium ion from cycloheptatriene by autoxidation in strong acid.l0 A solution of (4a) in chloroform is initially colourless but turns purple on standing. Chloroform normally contains some acid due to oxidation. Formation of colour is diminished if light and oxygen are excluded.However, addition of phosphorus pentachloride (a Lewis acid) resulted in immediate production of the colour. Lewis acids generate carbonium ions and are known to convert cycloheptatriene into the tropylium cation.11 Formally (5a) is generated from (4a) by removal of a liydricle ion and further proof of the identity of the J.C.S. Perkin 11 coloured species came from the reaction of (4a) with triphenylmethyl perchlorate.12 The colour appeared immediately. Thus, formation of (5a) is proved beyond doubt and we now know what highly coloured species may be formed by reaction of ureas with a-diketones. IJrca and I-methylurea do not react with butane-2,3- + (5) dione in the same manner as lJ3-dimethylurea. With 1-methylurea a crystalline product was obtained, of formula C,H,,N,O,.The same product was obtained by reaction in ethanol saturated with HCl. This compound results from the reaction of two molecules of 1-methylurea with one molecule of the dione and elimin- ation of two molecules of water. The i.r. spectrum indicated the presence of NH and carbonyl groups. The lH n.m.r. spectrum was rather complex with seven singlets. These we ascribe to three methyl groups, two N-methyl groups, and two NH groups. The spectrum could indicate an unsymmetrical molecule or a mixture of isomers. We propose the latter, (6a and b). In (6a) the two methyl groups are identical but in (6b) they differ. Each contains only one type of MeI MeHwMe C-NCONH,I 0 C=NCONH,MeNKNMeI Me (10) (11) N-methyl and NH group, These structures were confirmed by the 13C n.m.r.spectrum of the product. There is evidence for two different carbonyls (one in each isomer) and five peaks in the range 6 16.23--28.00 p.p.m. corresponding to the five different methyl groups. There were three different tertiary carbons (the peaks 1981 remained as singlets in the off-resonance spectrum) at 6 78.56, 82.54, and 86.G2 p.p.m., one in (6a) and two in (6b). We propose a mechanism for tlie formation of (6b) (Scheme 2) which is similar to that suggested pre-00 OH OH II II II Me--C-C-Me + MeHNCONH, -Me-C-C-Me II be recovered unchanged from this solvent. In the 13C n.m.1. spectrum there is only one resonance at low ficlcl (the carbonyl group) but (11)requires a second due to the C=N group.There is, however, a tertiary carbon at S 80.6 p.p.m.. These data are consistent with (6c) which is the analogue of the product obtained from 1-methyl- urea. Other spectral data are consistent with this structure. A mechanism analogous to that of Schemc 2 explains the formation of (6c). 0 The filtrate obtained after the removal of the mixture MeNKNH OHI H, FJ//--L6C0NH, -I Me-C-C-MeI 11 MeNKN0 SCIIEhlE 2 viously for tlie reaction of I-metliylurea with benzil. The crucial intermediate is (8),formed by elimination of water in acid conditions from the cliol (7). If (8) is yrotonated on the hydroxy-group a good leaving group (water) is formed and displacement by a second 1-methylurea molecule forms (9).Ring closure occurs by addition across the carbon-nitrogen double bond, probably by N-protonation ant1 generation of a car-bonium ion which reacts with the nucleopliilic NH, group. Attack of protonated (8)by the other end of the I-methylurea molecule gives (6a),rather than (Gb),and so the mixture of isomers is explained. The intermediacy of (8)explains why 1,3-dimethylurea does not form a product analogous to (6). The diol (10) cannot eliminate water to give a double bond in the ring. Thus, this reaction pathway is blocked and an entirely different one gives (4a) as the main product. A similar effect was noted in the reaction of 1,3-dimetliylurea with henzil and with l-phenylpropane-l,2-dione.GWith the former a hydantoin formed, while the latter gave a spiro-compound.With urea products analogous to (6) were obtained. Reaction of urea with butane-2,3-dione under the same reaction conditions gave a white crystalline material of molecular formula CGH,oN,O,. Lugosi et aZ.13 suggested structure (11) for this material but there is evidence against this. It was difficult to obtain spectral data for the product as it is insoluble in the usual organic solvents. However, it does dissolve in TFA and the material may of (6a and b) and of (Gc) from the reaction mixturc was highly coloured and suggested iormation of an-other reaction product. In view of our identification of the coloured material obtained by reaction of 1,3-dimethylurea as the carbonium ion (5a), it seems pro- bable that the above colours are due to (6b and c).As will be seen later in the discussion of the mechanism of formation of (h),there is nothing to stop urea and 1-methylurea reacting in the same way. However, wc were unsuccessful in isolating runt1 characterising (4e or f) from the coloured filtrates. It is possible that carbonium ion formation is complete and neither (4e or f) is present. With the reaction of 1,3-dimetliylurea some (44 must remain unchanged, although the colour of the solution does indicate that (5a) forms spontaneously. With 1-methylurea and urea the main products of reaction arc, of course (Ga-c); these are highly insoluble and immc-diately removed from the reaction mixture, leaving little material to form (4e and f). Some evidence u7as obtained for formation of (5b).Removal of solvent from the filtrate left a yellow gum which did not crystallise, but a cream solid was obtained on neutralisation of a mcthanolic solution by aqueous ammonia and addition of acetone. As this material was filtered off it became dark brown on contact with air. However, this was avoided when filtration was carried out under dry nitrogen. Once the material was obtained dry it was stable in a tightly sealed container. Addition of a little of this material to water produced a deep golden solution. It was found to be insoluble in all the regular organic solvents and, therefore, we were unable to purify it by recrystallisation.The material was hygroscopic, as well as impure, and no meaningful elemental analysis could be obtained. The 1H n.m.r. spectrum (in D,O) had only peaks cor- responding to methyl groups and N-methyl groups and there was a very large H,O peak, The i.r. spectrum (KBr disc) had a broad peak in the NH-OH region and another in the C=C-C=O region. Alkaline potassium permanganate was decolorised by a solution of the material, suggesting the presence of carbon-carbon double bonds. We suggest that the material isolated is the methanol (12), probably as a hydrate, which, in solution, forms the coloured carbonium ion (5b). There is support for this from the 13C n.m.r. spectrum in aqueous solution. There are peaks centred at 6 26.07, 55.26, 60.41, and 161.51 p.p.m.corresponding to methyl, two iN-methyl groups, and the carbons of the double bond which, because of the positive charge, have all moved down field from the equivalent values in (4a). The carboniuni ion centre and the carbonyl group did not appear in the spectrum. We were equally unsuccessful in obtaining positive direct identification of the coloured product obtained by reaction of urea and butane-2,3-dione. Repetition of the isolation procedure described above yielded a crcani solid which was hygroscopic, air-sensitive, and HH lsquo;crsquo;OH I II M~---C=C-CH-C=C-Me N II IJ I NHMeNKNHMeN YNH I 0 0 (121 (13) soluble only in water. However, we are sure that the colour is due to (k),partly because of the evidence adduced for the formation of (5a),and partly because we have found it impossible to find any other suitable cliromophoric system which can form from urea and an a-diketone.This identification is important as reaction between urea and diacetyl mono-oxime in acid is used as a colorimetric procedure for the measurement of urea concentrations in biological liquids during medical diagn0~is.l~ The method is widely used and appears to be highly specific euro;or urea. Until now the chemistry has been unknown. In acid solution the mono-oxime is liydrolysed to the diketone which then, we suggest, reacts with urea to form (4f). Radical oxidation of (4f) in acid solution generates the coloured species (5c). In the very dilute solutions used in biological assays (6c) is not precipitated.When (Gc) is dissolved in concentrated acid it is converted into (5c). We now come to propose a reaction mechanism for the formation of (4a) from butane-2,3-dione and 1,3-dimethylurea. Anything said in this section applies equally to the formation of (4b-f) from the appropriate diketone and urea. The first problem is the fate of the missing carbon atom. Evolution of carbon dioxide was detected during the course of the azeotropic distillation but this could have resulted from elimination of formalde- hyde, oxidation to formic acid, and decomposition at the temperature of refluxing benzene. Reaction betweenrsquo; butane-2,3-dione and 1,3-dirnethyl- urea does proceed in aqueous sulphuric acid at room temperature and 13C 1i.m.r.evidence shows that the product is again (4a). U7e attempted to demonstrate formation of formaldehyde in this reaction mixture by blowing through nitrogen and passing this gas into a solution of dimedone, but the result was negative. However, it is known I5 that, in acid solution, formalde- hyde forms a stable trimer and this may not be removed from the reaction mixture by the passage of nitrogen. J.C.S. Perkin I1 However, we were successful in isolating the 2,4- dinitrophenylhydrazone (DNP) of formaldehyde (13) from the reaction mixture. An excess of 1,3-dimethyl- urea was used so that no unchanged butane-2,3-dione, which could itself form a DNP, remained. Addition of 2,4-dini trophen ylh y drazine to the react ion mixture after 1 li yielded a yellow solid which was identified as tlie DNP of formaldehyde by a comparison of its lH n.ni.r.spectrum whth that of an authentic sample. There is a characteristic double doublet centred at 6 6.95 p.p.m. due to gemiiial coupling of the nietliylene group of (13). We think, therefore, that the reaction meclianism involves elimination of formalclehyde. There are pre-cedents for this in related reactions. The liriedel-Crafts reaction of anisole with 2-t-butyloxiran to give 2- (4-me thoxyplien yl) -2-met li yl bu t ane involves loss of a methylene group as formaldehyde.16 Obviously formation of (4a) involves a number of intermediates. We attempted to detect some of these by examination of the 13C n.1m.r.spectra of the reaction mixture taken at intervals during 5 h. However, it transpired that reaction is complete after 0.5 h, the time taken to collect the first spectrum. This spectrum did contain a few peaks in the range 6 84-103 p.p.m., in addition to those expected for (4a), and these can be assigned to polymers of formaldehyde. The same peaks were noticed in the spectrum of acidified formalin. In considering other reactions of urea with diketones we have suggested 5,6 that the first product of reaction is a diol, (10) in this reaction, and that further reaction is governed by tlie tendency of such compounds to eliminate the elements of water in the presence of acid. The intermediate (10) cannot form a double bond in the ring but can form one, or two, exocyclic double bonds, to give (14) or (15).The second of these cannot be the key intermediate as such a compound could not occur in the routes to (4b-d). On the other hand, compounds of type (4)form only if at least one substituent on the cwH2cwMe MeN KNMeMeNKNMe0 0 (14) (15) diketone is a methyl group and so (14) is a possible intermediate. Compounds of type (4)are 4-imidazolin-2-ones wliicli, as we have shown, form readily under our experimental conditions, from acyloins and urea. However, there are two reasons why we reject the intermediacy of acyloins in a mechanism for the formation of (4)from a diketone. First, conversion of a diketone to an acyloin requires reduction and our conditions are, if anything, oxidising. This is evidenced by the reaction of urea with benzoin, where products from reaction with bend were obtained.Second, we were unsuccessful in effecting reaction between any acyloin and 1,3-dimethylurea. Compound (4a) forms readily from 1,3-dirnethylurea. Formation of (4a) does not appear to proceed by condensation of two molecules of butane-2,3-dione before reaction with 1,dirnethyurea. The diketone was refluxed with benzene and TFA for 5 h, much longer than required for formation of (4a), but no water was formed. Apart from some tar, most of the diketone was recovered unchanged. Also, self-condensation is not cat alysed by urea as no reaction occurred when 1,3-dimethylurea was replaced by 1,1,3,3-tetramethylurea, These considerations lead us to suggest that (14)is the crucial intermediate in the formation (4a) and that the departing carbon leaves as formaldehyde.We propose the mechanism shown in Scheme 3. The condensation step is reaction of a molecule of (14) with another mole- cule which has been protonated to provide water as a leaving group. The driving force is the leaving of water and movement of the double bond, which generates an incipient carboniurri ion on the exocyclic methylene group. Reaction of this with the electron-rich dcublc HO CHI 111. Me-C-c CH,=T-C-Me II II 1+ Me-C=C-CHZ-C=C-MeII II f CH,O+H+ MeNKNMe MeNK NMe 0 0 scH EM euro;C 3 Imricl of tlic otlwr niolcculc of (14) arid cyclisation to give an oxetan ring, altliough not readily predicted, is not unreasonable.The process envisaged is not unlike the acid-catalysecl aldol condensation where one molecule of the enol form of acetcme reacts with a molecule of the protonated keto-form. Cjxetans are known to decom- pose readily to give a douhle bond and a keto-compound. In a recent paper l7iriterrnediate formation of an oxetan ring and decomposition to give a double bond in a steroid system has been proposed. Therefore, loss of formalde- hyde from (16) to give (4a) is an expected process. Thus, we can rationalise the formation of (4a) with only the 315 rather unexpected step involving formation of the oxetan ring. We have been unable to find any other reaction mechanism which accounts for the facts in as acceptable a manner. EXPEKIMENTAL The method of azeotropic removal of water has been described previously.Parentheses in the I3C n.ni.r. spectra. indicate the form in the off-resonance spectrum. Acetoin (5 ml), urea (3 g), and TF,4 (3 ml) were refluxed in benzene for 4 11. On cooling a solid separated antl the crystals of 4,5-dimethyl-4-iniidazolin-2-one (2a)were .~c.ashecl with acetone (yield euro;JOY0),m.p. 290" (decomp.) !lit.,2 290" (decomp.), wz/e 112 (M'), vlllaX. (niull) 8 180 (NH), 1 A85 (CxO), and 1665 cm-I (C-C), SII (Tl'A) 2.14 (0 H, s) and 9.96 (1 H, s), 6~ (TFA) 8.7 (q),119.5 (s),antl 151.2 (s)p.p rri. A similar reaction with 1 -methylurea gave 2,4,5-tviiwtlzyl-4-imiduzolin-2-one (2b) (yield 70), 111.1) 185", w/p 126 (Aft), vmax.(niull) 3 150 (NH), 1 685 (GO),and 1 650 c11l (C=C), (CDCI,) 1.94, 1.98 (G W,s),3 12 (3 H, s), and I1 49 (1 H, s), Bc (CIXl,) 8.3, 9.20, 26 9 (q),112.2 (bsol;), 113 8 (s), and 154.6 (5) p.p.m. (Found: C, 564; H, 7 85; N, 22.3 C,H,,N,O requires C, 57.1; H, 8.0; N, 22 2). Reaction of urea with propioilion gave 4,5-diethyl-4-imiduzolin-2-one (2c) (yield 75), m.p. 284", m/e 140 (MI), w,,,. 3 140 (NH), 1 690 (C=O), and 1 670 crn I (C-C), SH ('HJDMSO) 1.02 (6 H, t, J 8 Hz),2.22 (4 H, q, J 8 Hz), 8.52 (1 H, s), and 9.50 (1 H, s), Sc (TFA) 13.9 (q), 18.3 (t). 125.3 (s), and 151.2 (s) p.p.ni. (Found: C, 59.7; 11, 8.65; N, 19.56. C,H,,N,O requires C, 59.95; H, 8.6; N, 20.0:,amp;). Reaction of urea with a-hytiroxybenzyl methyl ketone gave 4-methyE-5-phenyl-4-inzidazo2in-2-~ne(2tl) (yield 45(::,), 11i.y.290" (clecomp.) (lit.,*amp;285O), ni/e 174 (iZIt), wll,,,.3 180 (NH), 1690 (C=O), anti 1650 crn (GC), Sf1 (Tl;,Z) 2 36 (3 H, s) and 7.42 (5 H), (TFA) 10.2, 82.0, 85.3, 127.!f --132.6, 137.2, arid 161.6 p.p.m. With 1-methylurea the product is 3,4-di?nel~lyZ-5-~hen),l-4-iuniduzoZzn-2-one (2e) (yield 55), 1n.p. 170deg;, m/e 188 (Mi),vmax. 3 460 (NH), 1 690 (GO), and 1675 cni (CrC), 81) ('H,I)MSO) 0.78 (3 H, s), 2.48 (3 H, s), 7.18-7.32 (5 H, m),and 5.90 (1 H, s), S(: (2H,ZMSO)24.4, 24.8, 85.6, 91.9, 126.7, 127.95, 128.1, 13!).0, antl 168.7 p.p ni. (I~ountl: (;, 69.95; H, 6.35; N, 14.7. CllHI2N,O reqnires (',, 70.2; H, 6.4; N, la.!)./,). a-Hydroxybenzyl methyl ketone ( lc) was prqxtred from niandelainide by the method of Hey.IB Hut;tne-2,3-tlionc and 1,3-1i1~ietliyl~ire~ were allowed to react under reflux for 40 min, 1)enzene antl TFA removctl by cvaporation, and the residue purifictl by elution from ;in alumina cwlunin (type H) with cliloroform.Ie~novaIof c.lilo~-oforni left a gum which crystallisetl on standing antl the product W;LS recrystallised from nietlirtnol to g~bsol;-c 4,4'-nzethyZenebis-( 1,3,5-trimetkyl-4-zm~~~u~~~i~z-~-one)(41) (yield 54(:6), 111.p. 218*, m/e 262 (A! ') vlllaY. (111~11) 1 iiRO (C=O) and 1 660 cm-1 (C=C), 8,, (CIXL:,) 1.99 ((ieuro;3, s), 3.09 (6 13, s), 8.17 (6 13, s),and 3.58 (2 H, s), 6,; (CI)CI,) 8.41 (q), 1!).00 (t),27.35 (q), 112.60 (s), 115.03 (s), and 153.53 (s) p p.m.(Found: C, 59.05; H, 7.95; N, 21.3. C,,,H,,)N40, requires C, 59.OS; H, 7.6; N, 21.2:/0). The reaction time for 1-methylurea was 2 11. After re-moval of benzene and 'TFA methanol was added to the residue and the solid filtered off. The filtrate was deep orange. The solid was recrystallised from DMSO and waslwtl with acetone to give a mixture of 1,3;t,4,7a-tet.rametI.Lyl-(6a) and J.C.S. Perkin I1 C. F. Winans and 13. Adkins, J. Amer. Chem. Soc., 1933, 55,l13a,6,7a-tet~a~net~cyl-tet~a~~y~~oi~nidazoi4,5-~~~~~iduzole-2,~-4dione (Gb) (yield 62), m p. 305O, nzje 198 (Mt), v,,,~~. 4167. K. Hofmann, lsquo; lmitlazole and Jts Dcrivativc.; I, Intersciencc,3 290 (NH) and 1 730 cn-l (GO),SI1 (L~H~IIMSO;looo) 1.39-1.49 (12 H, 3~),2.61 (3 H, s), 2.78 (9 H, s), 7.33 (3 H, s), and 7.55 (1 H, s), Sc (?rsquo;FA) 16.23, 1!).09, 22.25, 26.25, 28.00, 78.66 (s), 82.34 (s), 86.62 (s), 162.78 (s),ant1 IGB.25 (s) p.p.111.(Found: C, 48 5; H, 7.3; N, 28.2. C,H,,N4C), requires C, 48.45; H, 7.1 ; N, 28.257)).7rsquo;lie xbovc procedure was repeatctl for the reaction of iirca New York, 1953, p. 66. A. R. Butler and E. Leitch, J.C.S. Perkin 11,1980, 103. A. R. Butler, I. Iiussain, and E. Lcitch, J.C.S. Perkin II, 1980, 106. I. Kuono and Y. Uecla, Chem. and Pharm. Bztll. (Jnpu?~), 1071, 19, 110. C. Glidewell antl 1-1.L). tloltlcn, J. Clirirz. I?rsi.arch, in the press. 11. W. S. Chan, C;. Ixvctt, anti J. lsquo;1.lI;Ltthew, J.C.S. Chem. Cotnria,, 197 8, 756.with but,zne-2,3-dione to give :$a,7a-cli?net~~vItetra~ydroiuvl.id-uzo4,5-tlii)I idilzole-2,S-dioiw (tk) (yield 700;)} , 111.p. 348O, /w/e 170 (M +), v,~,lsquo;~~.3 240 (NH), I 720, ant1 1 G70 mi (GO), SH (TFA) 1.80 (6 H, i)and 7.3~-(2 H, s), (TFA) 21.86 (q),80.64 (s), ant1 164.10 (h) p 13 111. (Fonntl: Clsquo;, 42.35; TJ, 6.95; N, 32.95. (rsquo;GHl,,hr402reqiiires C, 42.4; TI, 6.0; N, 32.750/,). h.P. ter Korg, R. v;m bsol;Tcldtn, and A. F. 13ickc1, arc. T;,av. chim., 1061, 81,164. 11 M. E. Volrsquo;pin, I. S. Akhrcn, and D. N. iCur.s,anov, I :vrst. Akud. Nuuk. S.S.S.R. Otdrl Kliam. Nauk., 1957, 760. l2 FI. J. lhubcn, I,. R. EIonncii, and K. hl. llarnion, J, Ory.C/ti.wa., 1960, 25, 1442. l3 K. Lugosi, I. J. Thibert, W. T. Iioll:~nd,and L. I. Lam, Clinical Hioclbcm., 1971, 5, 171. l4 I). R. Wvbenga, j.Ili Ginrgio, and V. J. Irsquo;ilr.ggi, CliJzirol Chm~.,1971, 17,891. j.Schow, J. Irsquo;henz. lrsquo;liys., 1929, 26, 72. l8 M. Inouc, M. Harscln, N. IJmaki, antl I. Ichikawa, Eli//. Chem. Soc. Jupaiz, 1879, 52, 1873. 1rsquo; I. Rtorclli, S. Catalano, V. Scartoni, hl. I;rsquo;crrctti. and .I Tllarsili, I.C.S. PwkztL I, 1979, 16366. lH 1.. Rclir-Hrc~:o~vwslti,Bpr., 1897, 30, 1510. l9 1,. H. lIc.v, J. Crsquo;//PIE~..50c., 1!)30, 1233