Spirans. Part 11. A new method for generatingo-quinone methides, and its applications to the synthesis of spirochromans. A note on the autoxidation of a naphthol derivative

摘要

1977 2289Spirans. Part 11 .l A New Method for Generating o-Quinone Methides,and its Applications to the Synthesis of Spirochromans. A Note on theAutoxidation of a Naphthol DerivativeBy Francis M. Dean and David A. Matkin, The Robert Robinson Laboratories, The University of Liverpool,The adducts ( 5 ) and (9) formed from quinonyl carbanions and tetramethyl-1.4-benzoquinone are reduced by zincin acetic acid giving the component phenols. It is considered that o-quinone methides are intermediates, aromatis-ation of the quinonoid ring being the driving force (Scheme I ) . The corresponding ethers (6). ( 7 ) , and (10) giveo-quinone methides that are either readily trapped by acetate ion giving phenolic benzyl acetates (1 3) and (1 7) ordimerise giving the bisnaphthol (25) via the spiran (24).The benzylic acetates form excellent substrates for con-version into spirans such as (4) by standard methods; the spiran (24) can be regenerated from the bisnaphthol byoxidative cyclisation.The naphthol derivative (23) obtained as a by-product is rapidly oxidised by air to give the hydroperoxide (26).Liverpool L69 3BXTHE oxidation of tocopherol and related chromans (1)produces o-quinone methides (2) that rapidly dimerise tothe spirans (3). These have unusual properties, be-having as fluxional molecules and being sensitive to re-duction by ascorbic acid.3 In order to obtain inform-ation as to what extent such properties are determinedby substitution patterns and/or ring fusions we haveprepared several related compounds. We describe herea new method for generating o-quinone methides and itsapplication to the synthesis of the spiran (4) and twoothers.The xanthen derivative (5), ' diduroquinone,' is veryeasily obtained by treating duroquinone (8) with alkali;one molecule of quinone forms a carbanion that appearsto undergo a cycloaddition to a Smith, Tess,and Ullyot studied the reduction of this xanthen withzinc in acetic acid and obtained tetramethylquinol as thesole product.Various methods of reduction failed toproduce any useful result when applied to the ethyl etherof the xanthen. In our earlier work we caused a naph-thoquinonyl carbanion to add to duroquinone (8) givingthe xanthen derivative (9), and found that zinc and aceticacid reduced it to tetramethylquinol and 2,3-dimethyl-naphthalene-1 ,4-dioL6 Evidently both xanthens arecleaved across the heterocycle.A stepwise mechanismPart 10, A. W. Dick, F. M. Dean, D. A. Matkin, and M. L.Robinson, J . C S . Perkin I , 1977, 2204.D. R. Kelan and C. D. Robeson, J . Amer. Chem. SOC., 1962,84, 2963; P. Schudel, H. Mayer, J. Metzger, R. Ruegg, and 0.Isler, Helv. Chim. Acta, 1963, 46, 636; D. McHale and J. Green,Chem. and Ind., 1964, 366; J. L. G. Nilsson, Acta Pharm.SuPcica, 1969, 6, 1 ; J . L. G. Nilsson, J.-0. Bransted, and H.Sjevertsson, ibzd., 1968, 5, 509.was suggested originally,6 but while the reduction of etherlinks adjacent to carbonyl groups is well known, a simi-lar cleavage of carbon-carbon links is rare and does notseem to have been observed for zinc reductions.Wetherefore prefer to regard the cleavage as initiated byreduction of the enedione grouping (Scheme 1) followedby (or perhaps synchronous with) an electrocyclic re-version to quinol (from ring B) and quinone methide(from ring A). If the starting material is the phenol ( 5 ) ,then the quinone methide (11) can tautomerise to aquinone and will therefore also give tetramethylquinol.This accounts for the observations made by Smith et aZ.*and by US.^ I t is clear, however, that if the startingmaterial is the ether (6), the fate of the quinone methidewould be modified. In fact, and in contrast to the re-port by Smith et aZ.,4 the reduction of the methyl ethergave the benzylic acetate (13), derived by the nucleo-philic addition of acetate ion to (12) (Scheme 1).Wetherefore hoped that, by withholding nucleophiles, wewould allow the quinone methide (12) time enough todimerise to the desired spiran (4), but zinc and trifluoro-acetic acid merely reduced it to the quinol derivative(14).H. A. Lloyd, E. A. Sokoloski, B. S. Strauch, and H. M.Fales, Chem. Comm., 1969, 299; W. A. Skinner and P. A. Alau-povic, J . Org. Chem., 1963, 28, 2854; M. S. Chauhan, F. M. Dean,and M. L. Robinson, Chem. Comm., 1971, 1141.L. I. Smith, R. W. H. Tess, and G. E. Cllyot, J . Amer. Chem.SOC., 1944, 66, 1320.K. Chandrasenan and R. H. Thomson, J . Chem. SOC. ( C ) ,1966, 123.F. M. Dean and L. E. Houghton, J . Chem. SOC. ( C ) , 1968,10602290 J.C.S.Perkin IFortunately, the benzylic acetate (13) is itself a suit-able starting material for quinone methide production.When heated in mesitylene (or in the mass spectrometer)it rapidly collapsed by loss of acetic acid to form (12),which then gave the trimer corresponding to theaddition of a third quinone methide molecule to theinto quinone methides, higher temperatures favour trimerf o r r n a t i ~ n , ~ ~ ~ so a low temperature reaction was investi-gated. Reduction of the benzyl acetate with lithiumMe MeORMeMeHO al k y (Me Me( 5 ) R = H( 6 ) R = MeMe 1 MeOMe MeMeMe ' MeOMeCH2( 1 1 ) R = H( 1 2 ) R = Me 02a l k y l Me ( 3 )Me0 M$MeMeORMe H O G M eMe OHMe0Me Me$:: HO MeC H ~ O A C( 1 3 1SCHEME 1( 5 ) R = H(6) R = Me(7) R = Me2CHaluminium hydride gave the alcohol (15), and suddentreatment with hydrogen chloride converted this into thehalide (16).Nucleophilic displacement with sodiumOR MeoMe Me ' CH2XOHMeMeMeMe( 1 3 ) R = Me,X = OAc( 1 4 ) R = Me,X = H( 1 5 ) R = Me,X = OH( 1 6 ) R = Me,X = CL( 1 7 ) R = Me2CH, X = O A c( 1 8 ) R = MeZCH, X = OH( 1 9 ) R = Me2CH, X = C IMeMeMe0MePhPh OMespiro-dimer (4). This trimer is (very tentatively) as-signed structure (20) on the basis of analogy and consis-tent n.m.r. and other spectroscopic result^.^.^ trans-Stilbene trapped the quinone methide as the chromanderivative (21). Since spirans can themselves dissociateS. B. Cavitt, H. Sarrafizadeh, K.Gardner, and P. D.Gardner, J . Org. Chem., 1962, 27, 1211; A. Merijan, B. A. Shoul-ders, and P. T), Gardner, ihid, 1963. 28, 2148.hydrogen carbonate gave the yellow quinone methide (12)which dimerised almost at once to the desired spiran (4).In the same way the xanthen (5) was converted into* Part 9, M. S. Chauhan, F. M. Dean, S. McDonald, and M. L.Part 8, M. S. Chauhan, F. M. Dean, z). Matkin, and M. I,.Robinson, J.C.S. Pevkin I , 1973, 310.Rohinson, T.C.S. Perkin T , 1973. 1201977the isopropyl ether (7), which was carried through thesequence terminating in the halide (19). Treatment withbase again gave the desired spiran (4 ; Me,CHO for MeO) .During the conversion of the alcohol (18) into the halide(19) there seemed to be little loss caused by acid-catalysedremoval of the isopropyl group.In contrast, this lossdominated the pyrolysis of the acetate (17) in mesitylene,for instead of a dimer or trimer the sole product wasduroquinone (8) ; evidently the quinone methide (1 1) andisopropyl acetate are formed, but in the absence of anucleophile or a reducing agent, tautomerism as in theScheme leads to the quinone and there is no furtherreaction.A somewhat different result attended the reduction ofthe naphthalene derivative (10) with zinc. Tetramethyl-quinol was formed as before, but the quinone methide(22) appeared to be partitioned between two sequences,one terminating in the naphthol (23) and the other in thebisnaphthol (25) (Scheme 2).2291the desired spiran, but suffers reduction by zinc to thebisnaphthol.It is readily regenerated by oxidation ofthe bisnaphthol with hexacyanoferrate(I1x) (Scheme 2).Under the same conditions the benzoquinone methidewill tend to make an early capture of any nucleophileprovided. In the reaction in trifluoroacetic acid itaromatised by reduction before it had time to dimerise.Quinone methides have been generated by oxidation loand several other techniques l1 but not previously byreductive methods. The new method also offers a simplemeans of procuring selective asymmetric substitutionpatterns in quinol derivatives when only symmetricalprecursors are mailable. An alternative and more gen-eral solution to this problem has been described recentlybut requires facilities for anodic oxidation.12Note on the Na9hthoZ (23) .-Solutions of this compoundin hydrocarbons proved difficult to handle because of anunusually rapid reaction with air.For this reason thecompound was isolated and characterised as the acetate.0MeMeMeOMeMeHI_I__)B o M e Me HI___)SCHEME 2Me(23 1Thus the naphthoquinone methide dimerises but doesnot add acetate, while the benzoquinone methide addsacetate but does not dimerise. This marked differencecould arise from the relative aromatic stabilisation ener-gies, restitution of aromaticity being less important forthe naphthalene than for the benzene system. Hence thenaphthoquinone methide (22) must be a relatively stableone capable of surviving long enough to dimerise in thereaction medium giving the spiran (24).Indeed, this islo D. A. Bolon, J. Org. Chem., 1970, 35, 715, 3666.l1 A. B. Turner, Quart. Rev., 1964, 18, 347; D. Creed, J.C.S.Chem. Comm., 1976, 121; W. W. Sullivan, D. TTllman, and H.Schechter, TPtrnh,edron Lettrrs, 1969, 457.Left in air, the naphthol itself changed into a compoundassigned structure (26) mainly on spectroscopic grounds.The i.r. spectrum indicated hydroxylic and double con-jugated carbonyl functions; the n.m.r. spectrum indi-cated two vinylic methyl groups (subject to slight long-range coupling) together with a methoxy methyl groupresonating a t rather high fields because it has to rotatethrough the shielding cone of the benzene ring. In 2-naphthol derivatives, hydroperoxidation has been asso-ciated with steric hindrance to planarity of the aromaticl2 M.J . Manning, D. R. Henton, and J. S. Swenton, Tetra-h.edron LettPrs, 1977, 16792292 J.C.S. Perkin Iring induced by a single large substituent such as t - b ~ t y 1 . l ~The present case is essentially similar. Four serried(26)substituents must tend to buckle a ring already of rela-tively low aromaticity thus promoting the change tosp3 hybridisation at one position.EXPERIMENTALI .r. spectra were normally determined on mulls in paraffin ;only the strongest bands are noted. U.V. spectra were de-termined upon ethanolic 10-3-10-4~-sol~tions. N.ni.r.spectra given in the text were recorded under the sameconditions as those in the Tables. Mass spectra were meas-ured a t about 70 eV with a source temperature near 200 "C;m/e values for molecular ions only are noted.Light petro-leum had b.p. 60-80 "C.TABLE 1lH N.ni.r. spectra ( T scale)a of polycyclic quinolderivatives (in CDCl, at 100 MHz)CompoundAssignment (6)ArOCH, 6.46Vinyl OCH,Me,CHOCHAHA 7.087.64CHCH,PhCH.0ArCH, 7.86ArCH, 7.91Vinyl CH, 8.03Vinyl CH,ArCH,CH,ArCH,.CH,ArCH, 7.99Vinyl CH, 8.03CH,C(O.) (CO.) 8.60CH,C(CH,*) (CO.) 8.79(CH3) 2CH(7)6.06-6.286.887.687.827.917.998.358.358.728.90-8.91(4; PriOfor(4) MeO) (2l)c6.39 6.376.456.10-6.34ca. 7.0e7.80 7.96 7.787.88 7.96 7.857.88 8.06 7.877.98 8.298.06 8.298.14 8.297.43- 7.60-7.73 7.887.95- 8.008.25 g 8.30s6.05 f9.0Coupling constants in Hz.All bands had the appropriaterelative intensities. In chlorobenzene. ArH; complexoverlapping multiplets near T 3. J 17 Hz. I11 resolvedmultiplet. f J 8 Hz. g Partially obscured by other bands.J 6 H z .(7sR,* llaS *)-7a,lla-Dihydro-5-hydroxy-6,7a,9,10, l l a -pentamethyl-7H-benzo aIxnnthen-S, 1 1-dione (9) .-The folIow-ing method gave a superior yield to that described previouslyThe requisite tetrahydroindazole (3.0 g), dissolved in tri-chloromethane (34 ml) and methanol (135 ml), was addeddropwise during 2 h to tetramethyl-1,4-benzoquinone (9.1 g ;i.e. a large excess) and sodium acetate trihydrate (2.2 g) inmethanol (330 ml) stirred under nitrogen. After 1 h morethe mixture was cooled to 0 "C and the orange precipitateremoved, washed with cold methanol, and crystxllised frombenzene-trichloromethane to give the xanthendione asorange needles (4.14 g, 85y0), m.p.220". The excess of thetetramethylbenzoquinone was readily retrieved from thesolutions and purified for repeated use.TABLE 2lH N.m.r. spectra. (7 scale) of simple quinol derivatives(in CDCl, at 100 MHz) aCompound - Assign- ~ Ament (13) (14)ArCH,X 5.16Me0 6.66 6.64ArCH, 7.78 7.90ArCH, 7.78 7.90ArCH, 7.92 8.08ArCH, 8.08Me,CHAcO 8.58ArOHCH,OH(CH3)2CH(15) (16) (17) (18) (19)5.1ab 5.26 5.16 5.26b 5.196.42 6.39 6.20e 6.1OC 6.06c7.81 7.68 7.78 7.86 7.727.86 7.78 7.78 7.90 7.847.91 7.94 8.06 7.888.91d 8.78d 8.75d6.20e 6.10 6.06 e8.642.34 f i g 2.307.34 gph 7.04 ggJba Coupling constants in Hz.All bands had the appropriaterelative intensities. Broad ; sharp after contact with D,O.Me,CHO; septet. d, J 6 Hz. m, J 6 Hz. f Relativelysharp; H-bonded. Removed by D,O. Very broad.The methyl ether (10) was obtained from the xanthendione(3.0 g) by treatment with dimethyl sulphate ( 1 ml) andpotassium carbonate (20 g) in refluxing acetone (150 ml) for4 11, and formed yellow needles (from benzene-light petro-leum) (2.8 g), m.p. 161.5-162.5", v,,,. 1675 cm-l (conj.C : 0) (Found: C, 75.8; H, 6.550/6; M+, 364. C,requires C, 75.8; H, 6.6; M , 364).Reduction of the Xunthendione (10) by Zinc.--Zinc dust(10 g) was slowly stirred into a solution of the methoxyxan-thendione (10) (1.0 g) in acetic acid (20 nil) heated on a steam-bath.After about 20 min. the mixture became colourlessand was filtered hot. The residue was extracted with alittle hot acetic acid; the combined solutions on coolinggave a crystalline mass of tetramethylquinone (280 mg).The mother liquor was poured into brine (100 ml) and kepta t 0 "C under carbon dioxide for 2 h. A pink solid separatedwhich was washed with water, dried in uucuo (P205), andpurified from benzene-light petroleum to give 4,4'-di-nzethoxy-3,3'-dimethyZ-2,2'-ethylenebis- 1-naphthol (25) asplates (cu. 180 mg), m.p. 206", vInBx. 3 400br, 1 592, 1 360, and765 cm-l (Found: C, 77.3; H, 6.4; M', 402. C2,H2,04requires C, 77.6; H, 6.5; M , 402.)The benzene-ligh t petroleum mother liquors contained4-methoxy-2,3-dimethyl- 1-naphthol (22) but this deterior-ated when attempts were made to isolate it.In anotherexperiment, these liquors were concentrated in vncuo, theoily residue was treated with acetic anhydride and pyridine,and the product was chromatographed on silica with ben-zene. The chief fraction was an oil that slowly crystallisedfrom benzene, giving 4-methoxy-2,3-dimethyl- 1-nuphthylacetate as needles, m.p. 55", vmax. 1 745, 1 350, 1226, 1081,1025, and 761 cm-l (Found: C , 73.8; H, 6.8; M+, 224.C,,H,,O, requires C, 73.75; H, 6.604, ; M , 244).In a third experiment the benzene--light petroleum motherliquors were left in air to allow the deterioration to continue.During 30 h a solid (130 mg) separated which readily crys-tallised from ether-light petroleum giving 4-hydroperoxy-4-methoxy-2,3-dinzethyZnaphthalen- 1 (4H)-one (26) as needles,l3 P.A . Rradyand J. Carnduff, J.C.S. CAem. Com.m.., 1974, 81619771ii.p. 122 I , v , , ~ , ~ ~ . 3 350 and 1 655 cni- 2 (Found: iL1 ' , 234.0888.C13Hl,04 requires M , 234.0920).chromen-2-spiro- 2 '-naphthaZen- 1 '-one (24) .-The ethylenebisnaphtliol (25) (100 nig) in benzene (20 ml) was slowlyadded to a stirred solution of potassium hexacyanoferrate(II1)(1.0 g ) in water (20 ml), and after 1 h the benzene layer wasseparated, washed with water, and dried (Na,CO,). Thematerial left after evaporation was chroniatographed onsilica from benzene-trichloroniethane (1 : 1) and the chieffraction was purified further from benzene-trichloromethane( I : I), yielding the spivan as yellow plates (50 mg), m.p.180-181.5", v ~ , ; ~ ~ , 1 688, 1595, 1380, 1361, 1090, 783, and775 cm-l (Found: C, 77.4; H, 6.1; I+, 400.16836.C,,$4 requires C, 78.0; H, 6.0; M , 400.16745).3,4-Dihydro-4', 6-dinzethoxy-3', 5-dirnethyl-2H-benzoh -needles (1.66 g), 1ii.y.132", w , , ~ ~ ~ . 1 688, 1 670, 1 655, 1 120,and 945 cm-l (Found: C, 74.8; H , 8.1, M+, 370. C,,H,,-0, requires C, 74.6; H, 8.1; M , 370).Reduction of the Xnnthenedione (6) with Zinc.--(i) Inacetic acid. The methoxy-xanthendione (6) (2.0 g) wasreduced with zince dust in acetic acid, and tetramethyl-quinol removed as described for the analogue (10). Afterremoval of tetramethylquinol, the acetic acid solution waspoured into water (200 nil) and extracted with ether (3 x 80fill).Processed in the conventional way, the extract sup-plied a pale yellow oil largely soluble in light petroleum; fil-tration and concentration supplied a product that graduallycrystallised giving 2-hydroxy-5-wtethoxy-3,4,6-trirnethyl-benzyl acetate (13) as needles (790 mg), m.p. 69-70", w,,,,,3 410, 1700, 1290, and 1082 cm-l (Found: (1, 65.7; H,TABLE 3lH N.m.r. spectra (T scale)' of compounds derived from naphthalene (in CDCl, a t 100 MHz)Compound,2ssignnient,4rH dArHArHCH,O .4rCH,O vinylCHAHsArCH,CH,ArArCH,CH, alkylArCH,*CH, alkylArCH,ArCH,Vinyl CH,Vinyl CH,CH,C(O-) ((20.)CH,C(CH,-) (CO)0 H(10) (24) (23)1.75- 1.93 1.70-1.90 1.84-2.041.93-2.15 1.90-2.26 2.24-2.4 26.45-2.80 2.26-2.76 2.44-2.686.22 6.20 6.166.306.85'7.517.787.98 98.02 98.488.717.16-7.507.66-8.027.74 7.667.637.96fl Coupling constants in Hz. All bands had the appropriate relative intensities.L2H,dimethyl sulphoxide. a-Proton; J ca. 8 and ca. 1 Hz.homoallylic coupling. Removed by contact with D,O.(25) (26)1.71-1.92 1.86-2.101.92-2.17 2.16-2.403.51-3.01 2.40-2.706.23 7.037.057.507.94 !J8.02 g6.43 (2 H) i 1.49* As the acetate; T (AcO) 7.79. With 1e p-Proton; overlapping multiplets. J 17 Hz. g Broadened byOxidation of this spiran (40 mg) with iron(II1) chloride(200 mg) in absolute ethanol (15 ml) heated on the steam-bath for 2 h gave a dark solution that was kept for 12 h thenpoured into water (30 ml) and extracted with chloroform(2 x 25 nil).Isolated from the extract in the usual way,the product separated from chloroform as yellow needles,m.p. 277-278", indistinguishable from authentic 3,3'-di-methy-2,2'-ethylene- 1,4-naphthoquinone.14Ethers of 4a, 9a-Dihydro-7-hydroxy-2,3,4a, 5,6,8,9a-hepta-nzetlzylxanthen- 1,4-dZone.-The hydroxy-dione ( 5 ) (di-duroquinone) ( 5 g) was methylated in the same way as theanalogue ( ! I ) but for 24 h. The product separated fromlight petroleum giving the methyl ether (6) as lemon-yellowneedles ( 5 g), m.p. 130--131.5", w,,,. 1670, 1690sh, 1257,and 1 210 cm-l (Found: C, 73.6; H, 7.90/,; M+, 342. C21-H,@, requires C, 73.7; €1, 7.776; M , 342).The reaction of the hydroxy-dione with 2-iodopropane inrefluxing acetone containing potassium carbonate gave therequired ether but less effectively than the followingmethod.The hydroxy-dione (2.0 g) was dissolved inabsolute ethanol ( 13 ml) containing potassium hydroxide(0.4 g) and heated with 2-iodopropane in absolute ethanol(32 ml) for 4.5 11. The mixture was concentrated to halfits bulk and added t o water (250 nil), and the products wereextracted into ether (3 x 50 ml) and isolated in the usualway as a yellow mass. I'urification of this from lightpetroleum supplied the isopropyl ether (7) as lemon-yellow7.9; M+, 238. C13H1@4 requires C, 65.5; H, 7.6; M ,238).Reduction of the methoxyxan-then (6) (1.0 g) was conducted as in (i) but with tetrahydro-furan (20 ml) and trifluoroacetic acid (4 ml) as cosolventsinstead of acetic acid.The reaction appeared to be com-plete within 5 inin. The hot mixture was filtered and theresidue washed with a little hot tetrahydrofuran. The com-bined filtrates were concentrated to half their bulk and di-luted with light petroleum ( 15 ml) to precipitate tetraniethyl-quinol (370 mg). After removal of this, the solution wasconcentrated to a small volume and mixed with water (25ml) to precipitate the crude product (571 mg) which, afterintensive drying in uucuo (PZO5), was extracted with ben-zene. Concentration of the benzene extract and dilutionwith light petroleum furnished 4-methoxy-2,3,5,6-tetra-methylphenol (14) as needles, m.p. 115" (1it.,l5 116*), wmax.3 390br, 1 258, 1085, and 1015 cm-l.Thwmolysis of the Beizzyl Acetate (1 3) .-In neat mesitykne.The benzyl acetate (0.2 g) was heated in refluxing mesitylene(dried over sodium and re-distilled; 20 ml) through whichoxygen-free nitrogen was blown to remove acetic acid con-tinuously.When acetic acid could no longer be detectedin the effluent (about 1 h), the solution was concentratedl4 F. M. Dean, P. G. Jones, R. B. Morton, and P. Sidisunthorn,J . Chem. SOC., 1963, 5336.l5 H. Eilingsfeld and C. Martius, Annulen, 1957, 607, 159.(ii) In trifluoroacetic acidJ.C.S. Perkin 1under reduced pressure to a gum, transformed by additionof light petroleum into a yellowish powder. This crystal-lised from light petroleum to give the trimer (20),3',4',9,9a-tetrahydro-6', 7,9a-trimethoxy- 1,2,4a,5,5',6,7,7',8,8'-nonamethyl-3H-xanthen-4(4aH)-spiro-2'-chromerz-3-one, aslight yellow needles (0.1 g), m.p.190°, vInax. 1 670, 1410, 1 260,1 092, 1 065, and 1 010 cm-l, T (benzene) 6.53 (3 H, s, OMe)6.67 (6 H, s, OMe), 7.71 (3 H, s, ArCH,) 7.75 (6 H, s, ArCH,),7.80 (3H, 2, ArCH,), 7.88 (3 H, s, ArCH,), 7.89 (3 H, s,ArCH,), 8.39 (6 H, s, vinylic Me), 8.58 (3 H, s, angular Me),6.8-7.2 (multiplets overlapped by other bands, ring CH,).(Found: C, 74.15; H, 8.2; M+, 534. C33H4206 requiresC, 74.1; H, 7.9; M , 534).The experiment in(i) was repeated except for the presence of trans-stilbene(2 g). A similar work-up gave a gum that was chromato-graphed on silica from benzene-light petroleum (1 : 1) toremove the excess of stilbene (ca.1.9 g). Elution with neatbenzene then gave the product, which separated from lightpetroleum to furnish trans-6-unethoxy-5,7,8-trimethyl-2,3-di-phenylchroman (21) as needles (20 mg), m.p. 185"; vmax.1 498, 1 410, 1 240, 1096, 777, and 710 cm-l (Found: M+,358.1917. C2,H2,02 requires M , 358.1932). Elution wascompleted using benzene-ether (1 : 1) and supplied the tri-mer (20) (82 mg).Preparation and Thermolysis of 2-Hydroxy-5-isopropoxy-3,4,6-trimethylbenzyl Acetate (17) .-The reduction of the iso-propyl ether (7) in acetic acid occurred exactly as describedfor the methyl ether except that the product was muchmore soluble in the usual organic solvents and was thereforecrystallised from light petroleum by slowly chilling the solu-tion to about -10 "C.Thus obtained, the isopropoxy-acetate formed needles, m.p. 70", vmax. 3 360, 1 700, 1282,1082, and 970 cm-l (Found: C, 67.5; H, 8.5; M+, 266.C15H2,04 requires C, 67.6; H, 8.3; M , 266).Thermolysis of this acetate in a variety of solvents in-cluding xylene, mesitylene, and 2,4,6-trimethylpyridinegave tetramethyl- 1,4-benzoquinone as the only isolableproduct.2-Chloromethyl-4-methoxy-3,5,6-t~imethyl~henol ( 16) .-Themethoxybenzyl acetate (13) (3 g) in ether (sodium-dry; 50ml) was added gradually to a stirred slurry of lithium alu-minium hydride (500 mg) in ether (20 ml) at 0 "C. After afurther 20 min the mixture was quenched with saturatedaqueous sodium potassium tartrate.After the aqueouslayer had been adjusted to neutrality i t was discarded andthe product obtained from the ethereal solution in the usualmanner. The gummy solid crystallised from light petro-leum giving 2-hydroxymethy1-4-rnethoxy-3,5,6-trimethyl-phenol (15) as leaflets (1.1 g), m.p. 123-124"; v, 3 420,1 408, 1002, 825, and 715 cm-l (Found: C, 67.4; H, 8.4;M+, 196. C,lHl,O, requires C, 67.3; H, 8.2; M , 196).Hydrogen chloride was bubbled through sulphuric acidand then into a solution a t 0 "C of the hydroxymethylphenol(ii) I n mesitylene containing stilbene.(0.5 g) in ether (sodiuiii-dry ; 40 1711) containing molecularsieve and protected by a calcium chloride guard-tube. Theflow was kept as rapid as convenience allowed for 40 min,then the mixture was kept at 0 "C for 15 min, and finally per-mitted to reach room temperature.The ether was decantedand the molecular sieve washed with ether (3 x 5 ml). Thecombined ether solutions were concentrated under reducedpressure, the temperature being kept below 5 "C. The pro-duct (0.48 g) which crystallised out during this process waspure enough for further synthetic work. For analysis, theproduct was purified by sublimation (68' and 1.0 mmHg),giving the chloromethylphenol as needles, m.p. 95", v,,,. 3 400,1612, 1280, 1252, 1 165, 1094, 1003, 935, 865, 820, 778,and 721 cm-l (Found: C, 61.8; H, 7.3; C1, 16.9; M+,216/214. CllH,,CIO, requires C, 61.5; H, 7.3; C1, 16.5 ;M , 216/214).5', 6-Dimethoxy-3', 4', 5,6', 7,8-hexamethylchrouan- 2-spiro-1 '-cyclohexa- 3 ', 5'-dien-2'-one (4) .-The chlorome thylphenol(16) (420 mg) in ether (250 nil) was shaken with saturatedaqueous sodium hydrogen carbonate.The ether turneddeep yellow a t once; eventually this colour faded to lime-green and changed no further. Concentration of the ether-eal layer supplied a gum that crystallised from hexanefurnishing the spirodienone as yellow needles (280 mg), m.p.137-138", Amax. 228, 280sh, 286, and 325 nm (log E 4.05,3.56, 3.60, and 3.60), v,,,. 1 663, 1 649sh, I 581, 1 255, 1 098,1 008, and 885 cm-l (Found: C, 74.1; H, 7.7; M f , 356.C22H,,O4 requires C, 74.1 ; H, 7.9 ; M , 356).spiro- 1'-cyclohexa-3', 5'-dien-2'-one (4 ; Me2CH0 for MeO) .-2-Hydroxy-5-isopropoxy-3,4,6-trimethylbenzyl acetate (1 7)(3 g) was reduced with lithium aluminium hydride as des-cribed for the methoxy-analogut:. Obtained thus, 2-hy-droxymethyl-4-isopropoxy- 3,5,6-trinrethylphenol ( 1 8) separ-ated from light petroleum as leaflets (1.0 g), m.p. 97-98",v,,,. 3 500, 1 260, 1 090, and 1 000 cm-l (Found: C, 69.5; H,8.9; M f , 224. Cl,H,,O, requires C, 69.6; H, 9.0; M ,224).Prepared from this phenol by the method described forthe methoxy analogue and in similar yields, 2-chloromethyl-4-isopropoxy-3,5,6-trimethylphenol (19) sublimed (76' and1.5 mmHg) giving needles, m.p. 84-85", v,,,,. 3 460, 1 604,1 248, 1 110, 1 082, 952,846, 810, 765, and 705 cm-l (Found:C, 64.6; 7.7; M+, 244/242. C,,Hl,C1O, requires C,64.3; 7.9; M , 2441242).Treated with sodium hydrogen carbonate in the sameway as the methoxy analogue, the chloromethylphenol (19)(0.5 g) supplied the spirodienone which crystallised fromhexane as yellow needles (0.24 g), m.p. 133", 228, 281sh,287, and 330 nm (log E 4.05, 3.52, 3.56, and 3.53), vmax. 1 665,1 648sh, 1 582, 1 378, 1 329, 1 248, 1 175, 1 092, and 991 cm-l(Found: C, 75.3; H, 8.8; M+, 412.25973. C,#4 re-quires C, 75.7; 8.8; M , 412.26134).5', 6-Di-isopropoxy-3',4', 5,6', 7,8-hexamethyEchroman-2-7/939 Received, 1st June, 1977