Thermal rearrangement of aryl(chloro)methylenetetrahydrofuran-2,4,5-triones: its mechanism and the chemistry of the resulting aroylchloro-maleic anhydrides

摘要

146 J.C.S. Perkin IThermal Rearrangement of Aryl (ch1oro)met hylenetet ra hydrof uran -2,4,5-triones : its Mechanism and the Chemistry of the Resulting Aroylchloro-maleic AnhydridesBy Leslie Crombie and Derek P. Reynolds, Department of Chemistry, The University, Nottingham NG7 2RDOn heating above the m.p., aryl(ch1oro) methylenetetrahydrofuran-2.4.5-triones rearrange with double bond shiftand carbonyl oxygen-chlorine interchange giving aroylchloromaleic anhydrides. 14C Labelling studies and otherwork show the integrity of the carbon framework, and a mechanism involving formation of an oxet is proposed :this links mechanistically the formation and the rearrangement of the triones. Aspects of the chemistry of aroyl-chloromaleic anhydrides are reported.ON heating above its m.p., the red trione (1) la undergoesan unusual rearrangement l b involving interchange of acarbonyl oxygen and a chlorine atom, in conjunctionwith a double-bond shift.The product, orange-yellowchloro- (3,4-methylenedioxybenzoyl)maleic anhydride (2)retains the proton resonances of the methylenedioxy-phenyl residue and has vmz (CKC1,) 1 870, 1 780 (cyclicanhydride), and 1 660 cm-l (unsaturated carbonyl).Treatment with water or aqueous sodium hydroxideconverts it into the dicarboxylic acid (3) which gives adimethyl ester (4) with diazomethane and reverts to (2)on heating. In accord with the structures given, both(2) and (4) undergo mass spectral fragmentation pre-dominantly in one step to (5) (base peak), other ionsbeing formed from this.This behaviour contrasts withthat for the trione (1).l"Hydrogenation (Pd or Pt catalyst) of the anhydride(2) leads to saturation of the double bond together withdechlorination and gives the oxo-anhydride (6), hydro-lysed to the known 4-oxo-acid (8). In deuterioacetonesolution (6) exists predominantly in the oxo-form inequilibrium with about 5% of the enol (7). The oxo-enol conversion is shown by addition of D,O: rapidexchange causes disappearance of the quartet at T 4.55and collapse of the multiplet near T 6.6. That thesinglet at 6.03 genuinely belongs to a species in equili-brium with the oxo-form was demonstrated by transfer-ence of spin magnetisation by chemical e~change.~"Thus irradiation at the frequency of the AB multiplet ofthe ketone near 't 6.6 causes saturation of the singlet atT 6.03 due to the enol methylene.On heating, the oxo-anhydride (6) loses carbon dioxide giving the P-un-saturated y-lactone (9) (vmax. 1805 cm-l).Treatment of the chloro-anhydride (2) with aqueoussodium hydroxide followed by hydrogenation of theresulting diacid (3) over palladium yields the hydroxy-acid (lo), accompanied by some of the y-lactone (11)probably formed in the acidic work-up. The acid (10)affords the lactone (ll), identical with an authentic1 (a) M. J. Begley, L. Crombie, R. G. Havard, and D. P.Reynolds, preceding paper; (b) L. Crombie and D. P. Reynolds,J.C.S. Chem. Comm., 1973, 265.B. Willhalm, A. F. Thomas, and F. Gautschi, Tetrahedron,1964, 20, 1185.M.J. Bougault, Ann. Chim. Phys., 1908,15,491.(a) J. H. Noggle and R. E. Schirmer, ' The Nuclear Over-hauser Effect,' Academic Press, London, 1971; ( b ) L. M. Jack-man and s. Sternhell, 'Applications of Nuclear Magnetic Reson-ance Spectroscopy in Organic Chemistry,' 2nd edn., Pergamon,Oxford, 1969.J. W. Cornforth, G. K. Hughes, and F. Lions, J . Proc. Roy.Soc. New South Wales, 1938, 72, 228.~pecimen,~ when treated with trifluoroacetic acid. How-ever thermal lactonisation of (10) gives a second, isomericlactone (12) which is thermodynamically less stable andC E Oheat Ar CO2R m/e 149 (2) -'0 (7)ABX multiplet (95'/0) singlet, 2H(5%)Tx 4.55 /AX 4.9 HZTA 6.47. j ~ x 9.8 HZ(acetone, 30 "c)r 6.03TB 6.61 JAB 18.6 HZAr L C 0 2 H(8)Ar d -(9)0 iiAr H u T 6.47(11)0I I(121977 147is converted into (11) by treatment with acid: the massspectra of the two lactones are similar.The structuralassignments (11) and (12) may be made on the basisthat there is less steric compression in the former: theAlthough this stereoisomer may be the exclusive productfrom the hydrogenation, it is possible that an isomermixture is produced and that the (RS,RS)-stereoisomerpreferentially undergoes lactonisation to (ll), as thisHAr+C02HH(15)CL 0 0 CLassignments are consistent with n.m.r. evidence as accompanies (10). On heating the lactones (11) andthe 3-proton, cis-vicinal to the aryl group in (11) is (12) above their m.p.s both give the same trans-olefinicshielded46 relative to its counterpart in (12).Ready acid (15) ( J 16.0 Hz for the olefinic protons) by de-acid-catalysed epimerisation of (12) must occur via a carboxylative elimination.0 4L O0 L o 4" '-s8Pi - 1mmol(16) I + I, SOCI22. Ph3P:CH.COflcMe 02C,c, P P h(18) (17)SCHEME 1stabilised 3,4-methylenedioxybenzyl carbocation (13),i.e. A B L l in Ingold's nomenclature.6 . An eliminativemechanism is not operative, since deuterium was notincorporated when trifluoroacetic [2H]acid was em-ployed. Thermal lactonisation, which gives the lessstable isomer, must involve nucleophilic attack on thecarboxy-group (14) and requires the hydroxy-acid (10)to have (SR,RS)-stereochemistry as indicated in (14).Before examining mechanisms for the rearrangement(1) -w (2) it was considered desirable to establish un-equivocally that the a-carbon atom bearing chlorine in(1) becomes the a-carbon atom bearing carbonyl oxygenin (2).To this end, 3,4-methylene&o~yphenyl[3-~~C]-propiolic acid (18) was prepared by using the Wittig13 C. K. Ingold, ' Structure and Mechanism in OrganicChemistry,' Cornell University Press, 2nd edn., 1969148 J.C.S. Perkin Iapproach to acetylene synthesis (Scheme l).' 3,PMethylenedioxyphenylmagnesium bromide * was carb-oxylated with 14C0, and the resulting 3,Pmethylene-dioxy[ l-14C]benzoic acid (16) was converted into theacid chloride and treated with methoxycarbonyl-methylenetriphenylphosphorane. Pyrolysis of the be-t aine ( 17) and hydrolysis then gave [3-14C] - (1 8).Treatment of 3,P-met hylenedio~yphenyl[3-~~C] pro-piolic acid with oxalyl chloride yielded [or-14C]-( 1) whichwas pyrolytically converted into [14C]-(2).The position-ing of the label at the or-position was demonstrated bydegradation to [14C]benzoic acid and extraction of thelabel as 14C02 from the carboxy-group (Scheme 2)./" KMnO4Cu chromitc. * quinolineco21-69pCi mrnol-'SCHEME 2Since all the radioactivity is carried by the latter, theintegrity of the carbon atom attached to the aromaticring is maintained in the rearrangement.In the case being studied (1) (with Ar = 3,4-CH,O,C,H,) it was also observed that migration of therearranging system from C-1 of the aromatic ring toC-6 would pass unnoticed. That such a rearrangementdoes not occur was shown by pyrolytically convertingthe trimethoxy-analogue (19) into (20).The productcontained equivalent aromatic protons (2 H singlet,7 2.55) as required by structure (20).With these possible rearrangements of the carbonframework disposed of, the only reasonable inference ischlorine-oxygen interchange, and any plausible mechan-ism appears to require formation of an oxet [(22) or(26)]. The addition of carbonyl compounds to alk-l-ynyl ethers and sulphides, alk-l-ynylamines, andphenylacetylenes, when catalysed by boron trifluoride,is reported to give ap-unsaturated esters.@,l0 WithG. Markl, Chem. Bey., 1961, 94, 3006; S. T. D. Gough andS. Trippett, J. Chem. Soc., 1962, 2333.C . Feugas, Bull. SOC. chim.France, 1964, 1892.H. Vieregge, H. M. Schmidt, J. Renema, H. J. T. Bos, andJ, F. Arens, Rec. Trav. chim., 1966,85, 929.very electrophilic carbon atoms the catalyst can some-times be omitted and in the uncatalysed addition ofR''c=o/R&-OWRR" OR"''c=c /+IR/f-*RSCHEME 3hexafluoroacetone to ethoxyacetylene an oxet inter-mediate has been isolated.ll In the case of hexachloro-acetone the oxet was too unstable to be isolated but itcould be detected by n.m.r.12 A mechanism as inScheme 3 has been proposed.Cyclisations of arylpropiolyl chloro-oxalyl anhydridesare envisaged in Scheme 4 as giving initially a zwitterionAr -4"6 0 0 J(21) (221 ItSCHEME 4(21). At lower temperatures this yields the red trione(l), but if the chlorine transfer is reversible then, at10 H.G. Viehe, ' Chemistry of Acetylenes,' Dekker, New York,11 W. J. Middleton, J . Org. Chem., 1966, 30, 1307.18 G. van den Bosch, H. J. T. Bos, and J. F. Arens, Rec. Tvav.1969.chim., 1966, 8S, 667; 1970, 89, 1331977 149higher temperatures and under conditions of thermo- involve an ion pair (25) which could itself be interposeddynamic control, ring closure could give the oxet (22), between (21) and (1).and electrocyclic ring opening permits the formation of Two other pathways (Scheme 7) were considered, butthe anhydride (2). A reaction which resembles the appear less attractive. Thus in the zwitterionic inter-lower temperature phase of this reaction is the addition mediate (27), charge is confined on an sp2 orbital and noof carbonyl chloride to alkynyl ethers (Scheme 5 ) in delocalisation is possible: this species, and the tran-which formation of a zwitterion (23) is postulated: the sition state leading to it, are expected to be of higherR"ci (231 (22)SCHEME 6latter then transfers a chloride anion in preference toforming an oxet.12 In the case cited, cis-addition(21) 11(25)SCHEME 6(26) (27)allylicrearrangementArq$o0(22)OR"'/OR'* 'C=CR" ' 'ct0QC\CII(2)SCHEME 7occurs to give exclusively the 2-isomer (24), and thiswas isomerised to a ZE-mixture on distillation. Forthe trione (l), in which the crystals are approximately83% 2, 177, E , the isomerisation mechanism mayenergy than in the case of (21) where the p orbital isunoccupied and charge stabilisation by the aromaticring with its electron-donating substituents is possible.Allylic rearrangement, (26) --+ (22), also seems lesslikely, since this must proceed via an ion pair or a radicalpair, and the energy required for this process is notcompensated for by release of the high strain energy ofthe oxet ring.EXPERIMENTALChloro-( 3,4-methyZenedioxybenzoyl)maleic Anhydride (2) .-The red trione (1) is rearranged by heating above its m.p.Some charring occurs and it is convenient to use a sublim-ation apparatus.For larger batches the following methodwas used. 3,4-Methylenedioxyphenylpropiolic acid (8.5 g)in benzene (50 cm3) was refluxed with oxalyl chloride(8 cm3) for 2 h. The solvent was then evaporated off, andthe crude product transferred to a sublimation apparatus.After heating (180 "C) a t 0.1 mmHg the solid melted andboiled vigorously.The rearranged product distilled ontothe cold finger, and crystallisation from anhydrous benzene-hexane gave the anhydride (2) (9.35 g, 73%) as orange-yellow needles, m.p. 143-145" (Found: C, 51.5; H, 1.95;C1, 12.55%; M+, 279.9773. C,,H5C106 requires C , 51.3;H, 1.8; C1, 12.65%; M , 279.9774), vmX. (CHCl,) 1873mand 1 795s (cyclic anhydride), 1 660 (aryl, afbunsaturatedketone), and 1620 cm-l, Lx. (CHCl,) 257 (E 7 140) and297 nm (7 630), T [(CD,),CO] 2.22 (1 H, dd, .[5,6 8, J2.6 2 Hz,and 3.78 (2 H, s, CH,O,), m/e 282 (28%), 280 (84), and149 (100).Hydrolysis of the Anhydride (2).-(a) The anhydride(1.0 g) was dissolved in cold aqueous sodium hydroxide(10% ; 4 cm3).The solution was cooled in ice, acidifiedwith hydrochloric acid, and extracted with ethyl acetate.The dried (MgSO,) extracts were evaporated ; crystallis-ation from benzene-hexane-ethyl acetate gave crystals ofchZw0-(3,Pmethylenedioxybenzoyl)maleic acid (3) (798 mg,mg, 75%), m.p. 143-145" (decomp.) (Found: C, 48.6;H, 2.6; C1, 12.4. C,,H7C1O7 requires C, 48.3; H, 2.35;C1, 11.9%), v- (mull) 1725 and 1708 (carboxy), 1663(aryl, ap-unsaturated ketone), and 1 632 cm-l, A,, (EtOH)234 (E 17 600), 280 (6 780), and 322 nm (9 270), T [(CD,),CO]H-6), 2.46 (1 H, d, J 2 Hz, H-2), 2.95 (1 H, d, J 8 Hz, H-5),2.42 (1 H, dd, J5,6 8, J2,6 2 Hz, H-6), 2.62 (1 H, d, J2.6 2 Hz,H-2), 2.99 (1 H, d, J 5 , 6 8 Hz, H-5), 3.85 (2 H, S, CHZO,),and 2.6br (2 H, CO,H, exchanged on shaking with D,O)150 J.C.S.Perkin I(b) A suspension of the anhydride (2) (213 mg) in water(10 cm3) was heated on a steam-bath for 10 min (if heatingis prolonged, further reaction occurs). Work-up yieldedthe acid (3) (154 mg, 68%), m.p. 142-144" (decomp.).Sublimation of the acid (3) (12 mg) a t 156 "C and 0.1 mmHggave back the anhydride (2) (7.5 mg), m.p. 142-145' (frombenzene-hexane) .Treatment of the diacid (3) with ethereal diazomethanegave the dimethyl ester (4), a viscous oil (molecular dis-tillation) (Found: c , 51.65; H, 3.6; C1, 10.55%; M+, 326.C1,HllC10, requires C, 51.5; H, 3.4; C1, 10.9%; M , 326),vmX. (CHC1,) 1 730 (ester), 1665 (aryl, aa-unsaturatedketone), and 1620 cm-l, L,, (EtOH) 234 (e 17 SOO), 282(6 750), and 325 nm (9 200), T [(CD,),CO] 2.40 (1 H, dd,J5 6 8, JZ.6 2 Hz, H-6), 2.60 (1 H, d, J2.6 2 Hz, H-2), 2.97(1 H, d, J5.6 8 Hz, H-5), (2 H, S, CH,O,), 6.09 (3 H, S,CO,Me), and 6.23 (3 H, s, CO,Me), m/e 328 (ti%), 326 (la),and 149 (100).Hydrogenation of the Anhydride (2) in Ethyl Acetate.-The anhydride (2.38 g) in anhydrous ethyl acetate (40 cm3)was hydrogenated a t atmospheric temperature and pressureover 10% palladium-charcoal (uptake 1.9 mol.equiv.) .The filtered solution was evaporated, and the residuecrystallised from anhydrous benzene-acetone-hexane asshort needles of 3,4-methylenedioxybenzoylsuccinic anhydride(6) (1.80 g, 85y0), m.p. 145-146" (decomp.) (Found: C,58.4; H, 3.3%; M+, 248.C1,H807 requires C, 58.1; H,3.25%; M , 248), vm, (CHCl,) 1880m and 1 790s (cyclicanhydride), 1 677 (aryl ketone), and 1 608 cm-l, Lx. (CHCl,)239 (E 13 300), 285 ( 5 530), and 327 nm (10 200), T [(CD,),CO]2.17 (1 H, dd, J5.6 8, JZ.6 2 Hz, H-6), 2.47 (1 H, d, JZm6 2 Hz,H-2), 2.97 (1 H, d, J5.6 8 Hz, H-5), 3.81 (2 H, S, CHZO,),4.55 (1 H, dd, X part of ABX multiplet, disappears onaddition of D,O), ca. 6.55 (2 H, m, AB part of ABX multi-plet, analysis l3 gives TA 6.47, TB 6.61; Jax 4.9, JBX 9.8,JAB 18.6Hz), and 6.03 (s, methylene of enol), m/e 248 (11%)204 (lo), 149 (loo), 121 (24), 65 (23), and 63 (24).The anhydride (2) (480 mg) in anhydrous ethyl acetate(20 cm3), hydrogenated over Adams catalyst (48 mg)(uptake 2.0 mol.equiv.) , gave 3,4-methylenedioxybenzoyl-succinic anhydride (6) (260 mg, 61%).Hydrolysis of the Succinic Anhydride (6) .-A solution ofthe anhydride (167 mg) in aqueous sodium hydroxide (10% ;2 cm3) was kept for 16 h, acidified with hydrochloric acid,and extracted with ethyl acetate. The dried (MgSO,)extracts were evaporated. N.m.r. data [(CD,),CO] sug-gested the presence of a mixture (ca. 1 : 3) of 3,4-methylene-dioxybenzoylsuccinic acid and 4-OXO-4- (3,4methylenedioxy-pheny1)butyric acid (8) (7 2.27 (1 H, dd, J5.6 8, J 2 j , 6 2 Hz,H-5'), 3.88 (2 H, s, CH,O,), and ca. 6.7 and 7.3 (4 H, [AB],multiplet, vicinal CH,)). Sublimation, a t 155 "C and0.01 mmHg, completed decarboxylation and gave theoxobutyric acid (8) (99 mg), m.p.136-137" (lit.,s 136")(Found: C, 59.15; H, 4.5%; M+, 222. Calc. forCllH1,O,: C, 59.45; H, 4.55%; M , 222), v,, (mull) 1707(carboxy) and 1667 cm-l (aryl ketone), A,, (EtOH) 229(e 18 000), 273 (7 OlO), and 308 nm (7 740).Thermolysis of the Succinic Anhydride (6) .-Sublimation,at 165 O C and 0.1 mmHg, of the anhydride (6) (389 mg)gave, on crystallisation of the sublimate from acetone, 6-(3,4methyZercedioxy~hercyl)furarc-2( 3H)-one (9) (209 mg) asthin blades,.m.p. 156-158O (Found: C, 64.95; H, 4.1%;M+, 204. CI1H8O, requires C, 64.7; H, 3.95%; M , 204),v,= (CHCI,) 1805 (&-unsaturated y-lactone) and 1 650H-6'), 2.53 (1 H, d, Jz'.s' 2 Hz, H-27, 3.04 (1 H, d, J 8 Hz,cm-l, I,, (EtOH) 213 (e 25 200), 272 (9 906), 285 (6 870),and 289infl cm-l (6 190), r (CDCl,) 2.90 (1 H, dd, J51,&8 8,JZ',S' 8, J2°,6* 2 Hz, H-6'), 3.00 (1 H, d, J2',6' 2 Hz, H-27,3.22 (1 H, d, J5',6' 8 Hz, H-57, 4.06 (2 H, S, CHSO,), 4.44(1 H, t, J 3 Hz, H-4), and 6.67 (2 H, d, J 3 Hz, 3-H,).Hydrogenation of the Chlmo-anhydride (2) in AqueousBase.-The chloro-anhydride (597 mg) in aqueous sodiumhydroxide ( 0 .5 ~ ; 10 cm3) was hydrogenated over 10%palladium-charcoal (50 mg) (uptake 2.8 mol. equiv.).The filtered solution was acidified with hydrochloric acidand extracted with ethyl acetate. The dried (MgSO,)extracts were evaporated and p.1.c. on silica [benzene-dioxan-acetic acid (200 : 25 : a)] gave, as the low RF band,(a-hydroxy-3,4-msthylenedioxybenzyl)succinic acid ( 10) whichafter crystallisation from benzene-ethyl acetate-hexanemelted a t 124-126" and, as heating was continued, re-solidified, and melted again a t 156-159" (Found: C, 54.0;H, 4.65.C12H120, requires C, 53.75; H, 4.5%), v,, (mull)1719 (carboxy carbonyl) and 1610 cm-l, AmX (EtOH) 204(e 25 500), 236 (3 320), and 287 nm (2 970), T [(CD,),CO]3.0br (3 H, OH and CO,H, exchanged on addition of D,O),ca. 3.1 (3 H, m, aromatic), 4.03 (2 H, s, CH,O,), 5.06 (1 H,d, J z , s 7 Hz, a-H), 6.79 (1 H, q, CHCO,H), and 7.55 (2 H,m, CH,*CO,H). The high RF band proved to be trans-2- (3,4-methylenedioxyphenyl) -5-oxotetrahydrofuran-3-carboxylic acid (11) (76 mg), which formed blades (fromethyl acetate-hexane), m.p. 164-165" (decomp.) (lit.,5164-165"), identical (mixed m.p. and i.r.spectrum) withan authentic specimen (Found: C, 57.3; H, 4.05%; M+,250. Calc. for C12H1,06: C, 57.6; H, 4.05%; M , 250),vmx. (mull) 1 745 (y-lactone) and 1 723 cm-l (acid carbonyl),A (EtOH) 203 (32 000), 238 (3 500), and 287 nm (2 900),TT~D~),CO)] 3.0-3.2 (3 H, m, aromatic), 3.99 (2 H, s,CH,-0,), 4.40 (1 H, d, JZ.3 8 Hz, H-2), 6.47 (1 H, m, H-3),and 7.05 (2 H, m, 4-H,).Lactonisation of the Hydroxy-diacid ( 10) .-(a) Thermal.The hydroxy-diacid (216 mg) was heated in an oil-bath at130 "C for 10 min. Crystallisation from ethyl acetate-hexane gave cis-&( 3,4-methylenedioxyphenyl)-5-oxotetra-hydrofuran-3-carboxylic acid (12) (102 mg), m.p. 160-163'(decomp.) (Found: C, 57.65; H, 4.05%; M+, 250.C1,H1,06 requires C, 57.6; H, 4.05% ; M , 250), v,,, (mull)1730 cm-l (acid and lactone carbonyls), A,,,, (EtOH) 202(E 30 300), 238 (3 670), and 287 nm (3 110), T [(CD,),CO]3.15 (3 H, s, aromatic), 4.01 (2 H, s, CH,O,), 4.20 (1 H, d,J,,, 8 Hz, H-2), 6.20 (1 H, q, H-3), and 7.11 (2 H, d, 4Hz).(b) Acid-catalysed.The hydroxy-diacid (10) (71 mg)was dissolved in trifluoroacetic acid (1 ml). After 5 minthe solution was evaporated to give a quantitative yield oftrans- 2- (3,4-methylenedioxyphenyl) - 5-oxotetrahydrofuran-3-carboxylic acid ( l l ) , m.p. 156-159", i.r. spectrumidentical with that of an authentic sample. Repetitionwith trifluoroacetic [,H]acid gave a sample having nodeuterium incorporated (n.m.r.).Epimerisation of the cis-Lactone (12) .-The cis-lactone(12) was dissolved in trifluoroacetic acid.After 5 min thesolution was evaporated and the residue crystallised fromethyl acetate-hexane to afford the trans-isomer (1 l ) , m.p.164-165" (i.r. and n.m.r. comparison). Repetition withtrifluoroacetic [ZHIacid left the n.m.r. spectrum unchanged.Decarboxylation of the Lactone Acids ( 11) and ( 12) .-Thetrans-isomer (1 1) (70 mg) was heated at 180 'C for 30 min.The residue was dissolved in aqueous sodium hydroxidelS E. D. Becker, ' High Resolution NMR,' Academic Press,New York, 19691977 151(10%) ; the solution was washed with chloroform, acidified,and extracted with chloroform. Evaporation of the dried(MgSO,) extracts gave trans-4-( 3,4-methylenedioxyphenyl)-but-3-enoic acid (15) (52 mg), which crystallised fromaqueous ethanol as plates, m.p.115-116" (lit.,5 117-118"),7 (CDC1,) 3.1-3.3 (3 H, m, aromatic), 3.58 (1 H, d, withbroadening due to unresolved allylic coupling, J3,4 16 Hz,CH202), and 6.75 (2 H, d, with broadening due t o un-resolved allylic coupling, J 2 . 3 6.5 Hz, 2-H,).The same procedure with the cis-lactone (12) also gavethe acid (15) (82y0), identical (n.m.r. spectrum) with thatabove.Rearrangement of (u-Chloro-3,4,5-trimethoxybenzylidene)-tetrahydrofuran-2,4,5-trione (19) .-The trione (19) (1.0 g)was heated in a sublimation apparatus a t 180 "C and 0.1mmHg and the product distilled onto the cold finger.The i.r. spectrum of the crude product showed contamin-ation with starting material (band at 1 720 cm-l), and thematerial was resublimed a t 180 "C to complete the con-version.Crystallisation from anhydrous benzene-hexaneyielded chloro- ( 3,4,5-trimethoxybenzoyl)maleic anhydride (20)(523 mg) as orange prisms, m.p. 1OP-105" [Found: C,51.6; H, 3.6; C1, 10.55%; 11/17 (vapour phase osmometry),330.5. Cl4Hl1ClO7 requires C, 51.45; H, 3.6; C1, 10.85% ;M , 326.51, vmx. (CHC1,) 1870m and 1785s (cyclic an-hydride), 1 660 (aryl, up-unsaturated ketone), and 1 621crn-l, (CHCI,) 258 (E 4 880) and 306 nm (6 640),T [(CD,),CO] 2.55 (2 H, s, H-2 and -6) and 6.10 (9 H, s,OMe), m/e 328 (33%), 326 (loo), and 195 (100).14C Labelling Study of the Rearrangement of the Trione (1).-Radiochemical counting techniques. A Nuclear Enter-prises NE 8 130 automatic scintillation counter was used.Samples (1-2 mg) of soluble compounds were accuratelyweighed (Cahn electrobalance) and solutions in dioxan(10 cm3) were prepared.A sample (2 cm3) was intro-duced into a 20 cm3 vial and mixed with dioxan-basedscintillator (NE250) (10 cm3), and the activity was recorded.Counting efficiency was measured by addition of a 14C-labelled internal standard (10 pl; 3.86 x lo5 disint. min-1~ r n - ~ ) , Background radiation was recorded (blank contain-ing scintillator solution only) and averaged 23 counts min-1.Barium carbonate was suspended in a gel composed oftoluene-based scintillator (NE233) and the gelling agent(Carb-O-Sil).3,4-MethyZenedioxyfihenyl[ l-14C] benzoic A cid-Piperonalwas decarbonylated with palladium-charcoal l4 t o givel12-methylenedioxybenzene which, on treatment with N-bromosuccinimide, yielded l-bromo-3,4-methylenedioxy-benzene. 15To stirred magnesium turnings (2.4 g, 0.1 mol) in an-hydrous tetrahydrofuran (40 cm3), under dry nitrogen, afew drops of ethylene dibromide were added, t o initiatethe reaction.l-Bromo-3,4-methylenedioxybenzene (20.1 g,0.1 mol) was added to maintain the temperature between45 and 55 "C, and the mixture was then stirred for 1 h.[14C]Carbon dioxide {from concentrated sulphuric acid(5 cm3) and aqueous sodium [14C]carbonate (ca. 0.25 cm3;0.25 mCi) ] was passed into the Grignard reagent in nitrogen.Carbon dioxide (cylinder) was then passed in. After 2 h,solid carbon dioxide was added; the mixture was keptovernight, then poured into hydrochloric acid (100 cm3)l4 F.Dallacker and H. Zegers, Annulen, 1965, 889, 156.l6 W. J. Gensler and J. E. Stouffer, J . Org. Chem., 1958, 25,H-4), 3.93 (1 H, dt, J 3 , 4 16, J 2 , 3 6.5 Hz, H-3), 4.08 (2 H, S,1958.and ice (300 g). The product was extracted into ethylacetate and washed with aqueous sodium hydroxide (10%).The basic extracts were acidified and the precipitatecrystallised from ethanol t o give 3, 4-methylenedioxy-[ 1-14CI benzoic acid ( 16), further crystallised to constantactivity (1.68 pCi mmol-l).3,4-Methylenedioxyphenyl[ 3-14C]propioli~ A cid .-3,4-Methylenedioxy[ l-14C]benzoic acid (12.2 g ; 1.68 pCi mmol-l)was refluxed with thionyl chloride (25 cm3) for 9 h. Theexcess of thionyl chloride was evaporated off, and theresidue dissolved in benzene.The solution was added t ostirred methoxycarbonylmethylenetriphenylphosphorane 16(49.0 g) in benzene (750 cm3). Next day the precipitatewas collected and washed with water t o remove the phos-phonium salt. The water-insoluble residue was combinedwith the benzene filtrate and the mixture was evaporated.The residue crystallised from aqueous ethanol t o give thebetaine (17) (31.3 g, 89yo), prisms, m.p. 197-198", specificactivity 1.67 pCi mmol-l. Analytical and spectroscopicdata were determined for an inactive sample prepared bythe same procedure (Found: C, 72.05; H, 5.0. C,,H,06Prequires C, 72.25; H, 4.8%), T (CDC1,; 60 MHz) 1.9-2.7(2 H, m, aromatic H-2 and -6), 3.22 (1 H, d, J 8 Hz, H-5),4.11 (2 H, s, CH,O,), and 6.83 (3 H, s, Me0,C).By Markl's pr~cedure,~ the labelled betaine (17) (31.1 g;1.67 pCl mmol-l) was pyrolysed at 0.1 mmHg with a smallflame.The yellow distillate, which solidified on cooling,was dissolved in methanol and aqueous 50% sodiumhydroxide (50 cm3) was added. Next day the solution wasevaporated, the residue treated with water, the insolubletriphenylphosphine oxide filtered off , and the filtratecooled and acidified. Crystallisation from ethyl acetateafforded needles of 3,4-methylenedioxyphenyl[ 3J4C]pro-piolic acid ( 18) (7.7 g, 63y0), specific activity 1.67 pCi mmol-1.Preparation and Rearrangement of the Labelled A ryl (chloro)-methylene Trione (1) .-A suspension of 3,4-methylene-dio~yphenyl[3-~~C]propiolic acid (4.4 g ; 1.67 pCi mmol-1)in benzene (25 cm3) was refluxed for 2 h with oxalyl chloride(10 g).After evaporation a portion of the crude product(ca. 450 mg) was removed and crystallised from benzene t ogive red plates of the trione (1) (377 mg), specific activity1.66 pCi mmol-l. The solution for counting was de-colourised by addition of water (0.1 om3).' The remainder of the product was rearranged, withoutfurther purification, by heating in a sublimation apparatusat 175 "C and 0.2 mmHg. Crystallisation from benzene-hexane gave the anhydride (2) (2.1 g), specific activity1.68 pCi mmol-l [decolourised by water (0.1 ems)].Degradation of the Labelled Chlorovnaleic Anhydride (2) .-Potassium permanganate (1.3 g) in water (40 cm3) wasadded over 30 min to stirred, refluxing, labelled chloro-maleic anhydride (561 mg; 1.68 pCi mmol-l) in water(20 cm3). Aqueous 10% sodium hydroxide was added andthe manganese dioxide was filtered off and washed. Thefiltrate was acidified with hydrochloric acid and extractedwith ethyl acetate. The dried (MgSO,) extracts wereevaporated and the residue (295 mg) purified by p.1.c. onsilica [benzene-dioxan-acetic acid (100 : 20 : 4) as eluant],followed by sublimation (170 "C and 0.1 mmHg), andfinally crystallisation from ethanol to constant activity, t ogive 3,4-methylenedioxy[ l-14C]benzoic acid (85 mg), specificactivity 1.68 pCi mmol-l.Labelled methylenedioxybenzoic acid (54 mg ; 1.68 pCil6 0. Isler, H. Gutmann, M. Montevan, R. Ruegg, G. Ryser,and P. Zeller, Helv. Chim. A d a , 1957, 40, 1242152 J.C.S. Perkin Immol-l) was decarboxylated by refluxing with copperchromite (100 mg) in quinoline (3 cm3) for 4 h in dry,carbon-dioxide-free nitrogen. The exit gases were bubbledthrough barium hydroxide solution and the barium car-bonate was collected by centrifuging, washed thoroughlywith hot water,.and dried in vacuo. The labelled bariumcarbonate (41.0 mg), diluted with inactive material (2.009 0g), was ground, and a portion (ca. 60 mg) was weighed intoa vial for counting. Toluene-based scintillator (10 cm3)was added, followed by gelling agent (0.55 g), and themixture was shaken and then mixed ultrasonically (5 min).The specific activity was 1.69 pCi mmol-l.One of us (D. P. R.) thanks the S.R.C. for a research[6/864 Received, 6th May, 19761studentship