Thio-sugars. Part VI. Syntheses of episulphides by reaction of methyl 2,3- and 3,4-anhydroglycopyranosides with potassium thiocyanate or thiourea

摘要

1974 2019Thio-sugars. Part V1.l Syntheses of Episulphides by Reaction of Methyl2,3- and 3,4-Anhydroglycopyranosides with Potassium Thiocyanate orThioureaBy Malini V. Jesudason and Leonard N. Owen," Department of Chemistry, Imperial College, London SW7 2AYEpisulphides, in which the thiiran ring is of opposite configuration to that of the epoxide, have been prepared byreaction of potassium thiocyanate or thiourea with methyl 2,3-anhydro-4,6-0-benzylidene-a-~-taloside, -@-D-taloside, -cc-D-guloside, and -P-D-guloside, methyl 2,3-anhydro-4,6-di-O-methyl-a-~-alloside and -mannoside,methyl 3,4-anhydro-6-deoxy-a-~-taloside, and methyl 3,4-anhydro-6-deoxy-2-O-methyl-~- L-taloside. Some ofthe episulphides were obtained in high yield, but other products, including unsaturated glycosides (shown to beformed from episulphides), were also identified, and predominated in the reactions of the P-D-glycosides.Pro-longed treatment of several of the episulphides with potassium thiocyanate resulted in partial stereomutation of thethiiran ring, whilst prolonged treatment with thiourea led to desulphurisation. Considerable differences in thereactivities of the epoxides, and of the episulphides, are explained by steric and conformational effects.The l H n.m.r. parameters for the epoxides and episulphides are tabulated.THE direct conversion of aliphatic or alicyclic epoxides altroside) (6), from which the genuine episulphide (2) caninto episulphides by treatment with thiocyanate salts or be obtained almost quantitatively by conversion into thethiourea is well known,293 but attempts to effect similar methanesulphonate, followed by reduction with boro-reactions with sugar epoxides have usually been disap- hydride.The Japanese workers also obtained thepointing, and multi-stage processes have been used in- episulphide (2) in very low yield by reaction of the allo-stead; 495 the direct method appears to have been satis- epoxide (3) with ammonium thiocyanate; with potassiumfactory only when the epoxide is external to the sugar thiocyanate the epoxide gave mainly the monosulphideOH( 7 )ring.6 Guthrie and Murphy obtained small yields of theepithio-alloside (4) by reaction of methyl 2,3-anhydro-4,6-0-benzylidene-a-~-mannoside (1) with thiocyanate orthiourea, but reaction of the do-epoxide (3) with thio-urea was said to give the epithio-mannoside (2) (63)together with methyl 4,6-U-benzylidene-2,3-dideoxy-a-~-erytlzro-hex-2-enopyranoside (5).It was subsequentlyshown, however,8 that the supposed episulphide (2)is 2,2'-dithiobis (methyl 4 ,B-O-benzylidene-2-deoxy-a-~-l Part V, J. M. Heap and L. N. Owen, J . Chem. Soc. (C), 1970,712.C. C . J. Culvenor, W. Davies, and K. €3. Pausacker, J . Chem.Soc., 1946, 1050; C. C. J. Culvenor, W. Davies, and N. S. Heath,ibid., 1949, 278; F. G. Bordwell and H. M. Andersen, J . Amer.Chem. Soc., 1953, 75, 4959; L. Goodman and B. R. Baker, ibid.,1969, 81, 4924.M. G. Ettlinger, J . Amer. Chem. Soc., 1950, 72, 4792; E. E.van Tamelen, ibid., 1951, 73, 3444; C. C. Price and P.F. Kirk,ibid., 1953, 76, 2396.(7). Guthrie and Murphy's deduction (from thestability of their supposed manm-episulphide towardsthiourea), that the unsaturated product (5) was notformed through an episulphide, is thus unsubstantiated.Although criticism has been expressed of the viewthat the difficulty in effecting the direct conversion of anepoxide such as (1) into an episulphide is a consequenceof the rigidity imposed on the pyranose ring by thepresence of the benzylidene group, conformationalJ. E. Christensen and L. Goodman, J . Amer. Chem. Soc.,1961, 83, 3827.5 L. Goodman, Chem. Comm., 1968, 219; K. J. Ryan, E. M.Acton, and L. Goodman, J . Org. Chem., 1968, 33, 3727.6 L. D. Hall, L. Hough, and R. A. Pritchard, J . Chem. Soc.,1961, 1537; U.G. Nayak and R. L. Whistler, J . Urg. Chem.,1969, 34, 97.R. D. Guthrie and D. Murphy, J . Chem. Soc., 1965, 6666. * M. Kojima, M. Watanabe, and T. Taguchi, TetrahedrorzLetters, 1968, 8392020 J.C.S. Perkin Imobility, whilst not completely inhibited, is certainlyseverely restricted in this carbohydrate analogue oftrans-decalin, and this could hinder the formation of thetrans-fused five-membered ring in the necessary inter-mediate of type (8) (from thiocyanate) or type (9) (fromthiourea).g If the configuration at C-4 is inverted, thesystem becomes analogous to the much more flexiblecis-decalin, and we have therefore studied the reactivity,towards thiocyanate and thiourea, of 2,3-epoxides con-taining a cis-fused 4,6-0-benzylidene ring, in the expecta-tion that episulphides would be formed more readily thanfrom the 2,3-epoxides (1) and (3).gulo-epoxide (12) reacted slowly to give methyl 4,6-0-benzylidene-2,3-dideoxy-cc-~-threo-hex-2-enopyranoside(14) (42), a similar quantity of epoxide being recovered ;no episulphide was detected.The p-glycosides corresponding to the epoxides (10) and(12) reacted less smoothly.The talo-compound (15)with thiocyanate gave the gzdo-episulphide (18) (33y0),methyl 4,6-0-benzylidene-2,3-dideoxy-~-~-,fhreu-hex-2-enopyranoside (19) (20y0), and 3,3'-thiobis(methyl4,6-0-benzylidene-3-deoxy-p-D-idoside) (20) (40) ; withthiourea, the same products were obtained in yields of9, 22, and 2504, respectively. When the reaction with( 8 ) (9)OMe( 1 0 ) x = 0 (10a)( 1 1 ) x = s,O-CHp " ? o ~ o pMex( 1 5 ) x = 0 ( 1 7 ) x = 0( 1 6 ) X = S ( 1 8 ) X = S( 1 4 ),O-CH2PhCy bMe( 1 9 ),O-CH2OOMe t e.hCQ12s(20)OR OR(23) X = S (251 X =Reaction of methyl 2,3-anhydro-4,6-0-benzylidene-a-D-taloside (10) with potassium thiocyanate gave, in 89yield, the gulo-episulphide (13) ; when thiourea was usedthe yield was only 12 (or 40, allowing for recoveredepoxide).Methyl 2,3-anhydro-4,6-0-benzylidene-cc-~-guloside (12), although requiring a longer reaction time,with thiocyanate gave the talo-episulphide (1 1) (74y0),and, surprisingly, the gulo-episulphide (13) (20) ; theformation of the latter is discussed later. In the pre-ferred OH, conformations, rear approach of the nucleo-phile to open the oxiran ring is more hindered in thegulu- (12a) than in the talo-epoxide (lOa), thus accountingfor the difference in reactivity.With thiourea, theB C. C. J. Culvenor, W. Davies, and W. E. Savige, J . Chem.l o N. R. Williams, Adv. Carbohydmte Chem., 1970;25, 109.SOC., 1962, 4480.s ( 2 7 ) R = Me (29) R = M ethiocyanate was prolonged, no episulphide was isolatedand the yield of the unsaturated product (19) rose to50, clearly demonstrating that the latter compound isderived from the episulphide. The gulo-epoxide (17)with thiocyanate also gave compound (19) (9yo), to-gether with 2,2'-dithiobis (methyl 4,6-O-benzylidene-2-deoxy-p-D-idoside) (21) (31 yo), but, in remarkable con-trast, with thiourea this epoxide gave 59 of the talo-episulphide (16) with 17 of the unsaturated glycoside(19) ; the reason why the use of thiourea is more effectivein this instance than with the other epoxides is probablythat the desired product (16) is particularly resistant tofurther attack by this reagent (see below).In their ring-opening reactions with nucleophiles, theepoxides (15) and (17) normally give products which areformed from the OH5 conformation by diaxial fission,l1974 2021and the constitutions of the monosulphide (20) and thedisulphide (21) are assigned on this basis.Diaxialopening by thiocyanate ion would give the 3-thiocyanato-idoside from (15) and the 2-thiocyanato-idoside from(17), both in the 4C1 conformation, which must change toenable the cyclic intermediate cf.(S) to be formed.The hydroxy- and thiocyanato-groups could becomediequatorial by a change to the lC, conformation, butthis is unlikely because it would result in C-6 and thephenyl group becoming axial. The 4C1 conformationcould, however, change to a boat ( B 2 , J , which wouldpermit cyclisation, but such a change in the p-series (notin the a-series) involves a passing interaction between themethoxy-group and the substituent at C-2; with thep-glycosides, therefore, solvolysis to the hydroxy-thiolcompetes significantly with the formation of the cyclicintcrmediate necessary for production of the episulphide.J2,3 (2-5) and J3,, (7.0 Hz) also support the formulationas the 3,4-aZtro-episulphide (28) rather than the 2,3-guZo-isomer (30).Another product, which could not hepurified, was obtained from the reaction of the epoxide(26); this may have been the 2,3-episulphide (30),because evidence is presented in the following paper thatboth a 2,3- and a 3,4-trithiocarbonate are formed by re-action of this epoxide with sodium methyl xanthate.Lightner and Djerassi l1 observed that steroid episul-phides, prepared from epoxides by reaction with thio-cyanate, were sometimes contaminated with epimers,and 2a,3a-epoxy-9~-methyl-trans-decalin was reported togive a 4 : 1 mixture of the 2p,3p- and the 2a,3a-episul-phide. They suggested that the isomerisation occurredby interaction of the first-formed episulphide withthiocyanate ion, implying. the operation of the illu-strated mechanism (see Scheme).Jankowski andSCHENEThe nionosulphide (20) is then derived by further reactionof the 3-mercapto-compound with the epoxide (15),whilst the more hindered 2-mercapto-compound fails toattack the more hindered epoxide (17) and merely under-goes aerial oxidation to give the disulphide (21).As examples of systems in which no acetal bridge ispresent, nieth yl 2,3-anhydro-4,6-di-O-met hyl-a-~-alloside(24), methyl 2,3-anhydro-4,6-di-O-methyl-a-~-mannoside(22), methyl 3,4-anhydro-6-deoxy-a-~-taloside (26), andmet hy1 3,4-anhydro-6-deoxy-Z-O-methyl-a-~-t aloside(27) werc treated with potassium thiocyanate. Theepisulphides (23), (25), (28), and (29) were isolated inyields of 36, 35, 22, and 15, respectively, no otherproducts being identified.Yields of 16 and 35,respectively, of the episulphides (23) and (25) were ob-tained when thiourea was used instead of thiocyanate.The structure of the episulphide (28), unlike those ofthe other three, is not self-evident, because if thiocyanateion attacked the epoxide (26) at C-3 the thiocyanato-group would be trnns with respect to the hydroxy-fuiictions at both C-2 and C-4. The cyclic intermediate,and consequently the episulphide, could therefore be a2,3-compound. That this was not so was shown by thelH n.m.r. spectrum of the episulphide. A two-protonrnultiplet at 7 5.9-6.3 was assigned to H-2 and H-5,because INDOR and spin-decoupling techniques estab-lished coupling to the anomeric proton (doublet a t 5 5.54)and to the methyl protons (doublet at z 8-52); thisinultiplet also showed coupling to the hydroxy-proton,which therefore is at C-2.The relative magnitudes ofl1 D. A. Lightner and C. Djerassi, Tetrahedron, 1965, 21, 583.l3 I<. Jaiikowski and R . Harvey, Canad. J . CJtewl., 1972, 50,3930.l 3 1C. E. Davis, J. Org. Clzewz., 1958, 23, 1767; R. D. Schuetzand R. L. Jacobs, zbid., p. 1799; 1961, 26, 3467; D. B. Denneyand I. J . Boskin, J , AHEV. Chein. SOL, 1960, 82, 4736.N. 1’. Neureitcr and F. G. Bordwell, J . Atmr. Chenz. SOL,1959, 81, 578.Harvey,12 who did not mention this earlier work, haverecently shown that some cyclohexane episulphides canbe isomerised by heating with thiocyanate. We accord-ingly studied the effect of treating the various sugarepisulphides with thiocyanate or thiourea for longer timesthan had been used in their preparation.In the descrip-tion which follows, total recoveries of chromatographic-ally separated materials were generally ca. 95, includ-ing original episulphide (which accounted for almost allthe balance of the yields of transformation productsquoted).The a-talo-episulphide (1 1), after reaction with potas-sium thiocyanate in boiling aqueous ethanol for 3 days,was partly converted into the gztlo-isomer (13) (38),thus explaining the isolation of the two episulphides (11)and (13) from the reaction of the epoxide (12). Withthiourea, under similar conditions, the episulphide (11)was totally destroyed in 24 h; no isomer was detected,but a 97 yield of the olefinic product (14) was isolated,in complete accord with the reaction of the epoxide (12)with thiourea.There is therefore no doubt that theunsaturated glycosides encountered by ourselves, andby others ‘s8 are derived by desulphurisation of an inter-mediate episulphide. The formation of olefins by nucleo-philic attack on the sulphur atom of a thiiran ring hasbeen effected by a variety of reagents, including trivalentphosphorus corn pound^,^^ 139 l4 alkyl- or aryl-lithium,14~ l5sodium toluene-a-thiolate,16 and sodium alkyl xanth-ate^.^. l7Prolonged reaction of the a-gttlo-episulphide (13) withthiocyanate gave 6 of the isomer (11) (attack at C-2l5 F. G. Bordwell, H. I. Andersen, and B.M. Pitt, J . Amer.C h m . SOC., 1954, 76, 1082; RI. Morton and K. F. Kammereck,ibid., 1970, 92, 3217.l6 J. F. McGhie, W. A. Ross, F. J. Julietti, B. E. Grimwood,G. Usher, and N. I. Waldron, Chenz. and Id., 1962, 1980.l i J. F. McGhie, W. A. Ross, F. J. Julietti, G. Swift, G. Usher,N. 31. Waldron, and B. E. Grimwood, CJzenz. and Ind., 1964, 4602022 J.C.S. Perkin Iin the original episulphide is hindered) and 20 of theolefinic glycoside (la), whilst with thiourea the soletransformation product was again this olefin (79).The corresponding P-glycosides showed interestingdifferences from their a-counterparts. The talo-episul-phide (16) with thiocyanate gave the gzclo-compound (18)(20) with some olefin (19) (6), but although thioureaagain gave no isomer, but only olefin, the conversion wasmerely 15, perhaps because attack on the sulphur atomin the episulphide (16) is more hindered than in the a-isomer (11).On the other hand, the gztlo-episulphide(18), in which attack on carbon is hindered, on treatmentwith thiocyanate gave no isomer, but only the olefin (19)(20), whilst thiourea likewise gave only the olefin, butthe conversion was S9yo, attack on the sulphur atomnow being subject to minimal steric hindrance. Astriking feature, noted with all four episulphides, is thatthiourea is a much more effective desulphurising agentthan thiocyanate, and consequently is best avoided in thepreparation of an episulphide unless the latter is pro-tected by steric hindrance against attack on the sulphuratom.Interconversion of the 4,6-di-O-methyl compounds(23) and (25) with potassium thiocyanate was alsodemonstrated, the allo-episulphide yielding 20 of themanno-compound, whilst the latter was isomerisedunder the same conditions to the extent of 8.Studies of the lH n.m.r.spectra of several 2,3-epoxy-sugars w20 and of the 2,3-episulphide (4) l8 have shownthat the coupling constants between the epoxide-ring(or epithio-ring) protons and their neighbours a t C-1 andC-4 are very small (<1 Hz; often zero) when the rela-tionship is trans, and rather small (1645.5 Hz) whenthe relationship is cis. Parameters for H-1 and Jl,z forthe 2,3-epoxides and 2,3-episulphides prepared in thepresent work are shown in the Table.Data for com-pounds (lo), ( l l ) , (17), (18), (22), and (23), in which H-1and H-2 are trans, and compounds (12), (13), (16), (24),and (25), in which H-1 and H-2 are cis, agree with thisgeneralisation, as also do the values of J3,4 for com-pounds (18), (23), and (25) (the only three of these elevenfor which this coupling could be measured). The zerovalue for the cis-1,2-coupling in the 2,3-anhydro-p-taloside (15) is clearly anomalous, but it may be signifi-cant that the corresponding 2,3-episulphide (16) showsan appreciably smaller cis-l,2-coupling than those ob-served for the episulphides (13) and (25).lH N.m.r. data on 3,4-epoxides are sparse, but in thefew examples where J2,3 and J4,5 have been measurable l9the values accord with the generalisation for J1,2 andJ3,4 in 2,3-epoxides. However, the two 3,4-episulphides(28) and (29) showed Jz,3 2.5 and 3.5 Hz, respectively,and J4,5 4.0 and 6-0 Hz, respectively, though bothcouplings are trans.In deuteriochloroform, the resonance for H-1 in the2,3-epithio-a-~-alloside (25) appeared as a triplet, pre-sumably because of virtual coupling between this protonl8 D.H. Buss, L. Hough, L. D. Hall, and J. F. Manville,Tetrahedron, 1965, 21, 69; cf. F. Sweet and R. I<. Brown, Canad.J. Chem., 1968, 46, 1481.and the thiiran protons, as reported for the spectrum ofmethyl 4,6-0-benzylidene-2,3-dideoxy-2,3-epithio-a-~-alloside,ls but in hexadeuteriobenzene the expecteddoublet was observed. The value for J4,5 (9-0 Hz) forlH N.m.r.parameters for solutions in CDCl, (T values;J in Hz)Com-pound H-1 H- 2 ~ 3 - 3 J1.z Jz.3 J3.4(10) 4.90 (s) 6.88 (d) 6-4 (m) 0.0 3.5(11) 4-83 (s) 6.95 (d) 0.0 6.0(12) 4.88 (d) 6-56 (m) 2.5(13) 4.68 (d) 6.65 (m) 3.2(15) 5-29 ( s ) 6.8 (m) 0.0(17) 5.20 (s) 6.7 (m) 0.0(22) 5.10 (s) 6-81 (9)5.07 (d) 2.0(25) 5.25 (d) 7.0 (m) 4.4 3.2(26) 5.54 (s) 0.0(27) 5.60 (d) 1.0(16) 5.11 (d) 6-5 (m) 2.0 6.0(18) 5.00 (d) 6.77 (4) 0.3 6.5 0.50.00.0 6-8 0.0 5.02 (s) 6.94 (9)(28) 5-54 (d) 6.1 (m) c 5.0 2.5 7.0(29) a 5-45 (d) 6.52 (9) 7.24 (q) 5.0 3.5 7.0a In C6D6. Xlso 6.42 (q, H-4), 6.06 (m, H-5), 6.66 (m, H-6) ;J4,6 9.0, J5,G 3.0, J5,6' 4.0. C -Us0 7.05 (m, H-3 and H-4),6.1 (m, H-5), 8-52 (d, H-6) ; J4 4.0, J 5 7-0. ,41so 7.44 (q,H-4), 6.42 (m, H-5), 8-85 (d, Hi6) ; J4,5 6.0, J5,B 6.0.compound (25) is consistent with the OH5 conformation,which indeed would be expected for all the 2,3-epoxidesand 2,3-episulphides of the a-configuration.The sameconformation is likely to be preferred also for the p-glycosides (15)-(18), in spite of the unfavourableanomeric effect, because the alternative 5 H ~ conforma-tion would result in axial orientations for both C-6 andthe benzylidene phenyl group.l0 For the 3,d-epoxidesand episulphides (26)-(29), the L~HO conformation issterically favoured ; furthermore, the alternative LOH,conformation, with a diaxial relationship between H-1and H-2, could not be reconciled with the values of Jl,2.Further evidence for the constitutions of the newepisulphides is provided by their conversion into thetrithiocarbonates described in the following paper.EXPERIMENTALlH N.m.r.spectra were recorded for solutions in deuterio-chloroform (unless otherwise stated) on Varian A-60 orHA-100 instruments, and the important parameters aregiven in the Table; resonances for aromatic, benzylic,O-methyl protons, etc., were in accord with the constitutionsof the compounds. 1.r. spectra were recorded for allproducts, and were used to assist in identifications andcomparisons, but the absorptions, which were unexceptional,are not given. Optical rotations were measured, for solu-tions in chloroform, with a Perkin-Elmer 141 polarimeter.Kieselgel GF,,, (Merck) was used for t.l.c., and silica M.F.C.(Hopkin and Williams) for column chromatography.Extracts were dried over magnesium sulphate, and solventswere removed under reduced pressure below 50".Petro-leum refers to the fraction of b.p. 40-60".Methyl 2,3-A nhydro-4,6-d i-O-methyl-a-D-nzannoside. - Asolution of methyl 2,3-anhydro-4,6-O-benzylidene-a-~-man-J. G. Buchanan, R. Fletcher, K. Parry, and W. A. Thomas,J . Chem. SOC. ( B ) , 1969, 377.2o R. J. Ferrier and N. Prasad, J . Chem. SOC. ( C ) , 1969, 575;L. Hough, P. A. NLznroe, and -4. C. Richardson, ibid., 1971, 10901974 2023noside 21 (3.4 g) and oxalic acid (10 g) in acetone (325 nil)and water (40 ml) was boiled under reflux for 16 h, and thenneutralised with barium carbonate. The salts were filteredoff and washed with acetone, and the combined filtrate andwashings were evaporated to remove the acetone.Theaqueous residue was extracted once with ether to removebenzaldehyde, and it was then evaporated t o dryness.Extraction of the residue with acetone gave an oil (2-1 g)which was methylated four times with methyl iodide-silveroxide. Distillation of the product gave the 4,6-di-O-methyl comfiound (1.7 g), b.p. 51-52' at mmHg,a,23 + 142" (G lag), T 5.1 (lH, s, H-1) (Found: C, 52.75; H,7-7. C,H1,O, requires C, 52.9; H, 7.9).Methyl 2,3-Anhydro-4,6-O-benzylidene-a-~-guZoside and-taloside.-The mixture of epoxides, prepared by Reich-stein's was separated by preparative t.1.c.(chloroform). The gulo-epoxide had m.p. 175-176" (lit.,22178-179"; lit.,24 174-175"), and the taZo-epoxide, m.p.231-237" (lit.,22 241-242").Reactions of Epoxides.-Except where further details aregiven, the proportions of reagents, and the reaction condi-tions, were based on the following general methods.(i) A solution of the epoxide (1 mmol) and potassiumthiocyanate (0.5 g) in ethanol (ca.25 ml for the 4,6-0-benzylidene compounds; ca. 10 ml for the 4,6-di-O-methylcompounds) and water (2-3 ml) was boiled gently underreflux for the period specified, then cooled, diluted withwater, and extracted with chloroform. The product isolatedfrom these extracts was purified by preparative t.1.c. (chloro-form, unless otherwise specified).(ii) A solution of the epoxide (1 mmol) and thiourea (0.4g) in ethanol (ca. 30 ml for the 4,6-O-benzylidene com-pounds; ca. 10 ml for the 4,6-di-O-methyl compounds) wasboiled under reflux, and worked up as described above.With potassium thiocyanate.(a) Methyl 2,3-anhydro-4,6-O-benzylidene-a-~-taloside (210 nig) after 30 h gavemethyl 4,6-0-benzylidene-2,3-dideoxy-2,3-ep~thio-a-~-guloside(13) (198 mg), m.p. 114-115" (from ether-petroleum),aIDz2 +23-2" (c 0.8) (Found: C, 59.8; H, 6-0; S, 21.6.C14H1,04S requires C, 60.0; H, 5.75; S, 11.4).(b) Methyl 2,3-anhydro-4,6-O-benzylidene-cc-~-guIoside(120 mg) after 170 h gave methyl 4,6-0-benzyZidene-2,3-di-deoxy-2,3-epithio-a-~-taloside (11) (94 mg), m.p. 170-173"(from chloroform-ether), a,21 - 56.3" (c 0-5) (Found : C,60.3; H, 5.8; S, 11.3y0), and the isomeric guio-episulphide(25 mg), m.p.and mixed m.p. 110-113".(e) Methyl 2,3-anhydro-4,6-O-benzylidene-P-~-taloside 2,(100 mg), potassium thiocyanate (200 mg), ethanol (5 ml),and water (1 ml) after 18 h gave methyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epithio-P-~-guZoside (18) (35 mg), m.p. 115-116" (from ether), a,22 -72.0" (c 0-2) (Found: C, 60.1; H,5.2 ; S, 11-3), methyl 4,6-0-benzylidene-2,3-dideoxy-p-~-threo-hex-2-enopyranoside (19 mg), m.p. 124-127" (fromchloroform-ether), - 168" (c 0.8) (lit.,25 m.p. 122-123", - 172"), and 3,3'-thiobis(methyZ 4,6-O-benzylidene-S-deoxy-P-~-idoside) (20) (42 mg), m.p. 139-141" (fromchloroform-ether), @ID2, - 50.4" (c 0.8) (Found: Mf, 562.C2,H3,0,,S requires M , 562), which with acetic anhydride-pyridine gave the bis-2-O-acetate, 1n.p.258-263", a:.,21 ' Methods in Carbohydrate Chemistry,' eds. R. L. Whistlerand M. L. Wolfrom, Academic Press, New York and London, 1963,vol. 2, p. 189.22 E. Sorkin and T. Reichstein, Helu. Chim. Ada, 1946. 28, 1.23 M. Gyr and T. Reichstein, Helv. Chim. Acta, 1945, 28, 226.24 H. Huber and T. Reichstein, Helv. Chim. Acta, 1948,81, 1645.+ 11.2" (C 0.25) (Found: C, 59.2; H, 5.9; S, 5.0. C,,-O,S requires C, 59.4; H, 5.9; S, 4.95). The lH n.m.r.spectrum of the hexenoside agreed with that reported. 25(d) Methyl 2,3-anhydro-4,6-O-benzylidene-P-~-guloside 22(120 mg) after 3 days gave unchanged epoxide (8 mg),methyl 4,6-O-benzylidene-2, 3-dideoxy - B-D-threo-hex-z-eno-side (10 mg), and 2,2'-dithiobis(methyZ 4,6-0-benzyZidene-2-deoxy-P-D-zdoside) (21) (39 mg), m.p.200-202" (from chloro-form-ether), 0(,2~ +46*8" (c 0.4) (Found: C, 56.9; H, 5.2;S, 11.0. C2,H3,01,S2 requires C, 56-55; H, 5-7; S, 10.8y0).(e) Methyl 2,3-anhydro-4, 6-di-O-methyl-cc-~-alloside as(220 mg) after 12 h gave (t.1.c. in ether) methyl 2,3-dideoxy-2,3-epithio-4,6-di-O-methyl-a-~-mannoside (23) ( 1 1 1 mg),b.p. 70" at mmHg, + 126" (c 0.9) (Found: C, 48.9;H, 7-25; S, 14.85. C,H,,O,S requires C, 49.1; H, 7-3; S,14.55).(f) Methyl 2,3-anhydro-4,6-di-O-rnethyl-a-~-mannoside(250 mg) after 18 h gave a crude product which was directlycrystallised from petroleum to give methyl 2,3-dzdeoxy-2,3-epithio-4,6-di-O-metlzy~-a-~-u~~oside (25) (95 mg), m.p.52-55", +249" (c 0.6) (Found: C, 48.9; H, 7.2; S,14.55).(g) Methyl 3,4-anhydro-6-deoxy-a-~-taloside 27 (539 mg),potassium thiocyanate (720 mg) , ethanol (15 ml), and water(10 ml) were stirred together at 80" for 16 h.Dilution withwater and extraction with ether gave a crude product whichwas purified by column chromatography (ether) to givemethyl 3,4,6-trideoxy-3,4-e~ithio-a-~-aZt~oside (28) (I30 mg),m.p. 62-72" (solidified oil) -97" (c 0.3) (Found: C,47.9; H, 6.8; S, 18-1. C,H1203S requires C, 47-7; H, 6.9;S, 18.2).(h) Methyl 3,4-anhydro-6-deoxy-2-O-methyl-a-~-talo-side 28 (250 mg), potassium thiocyanate (310 mg), ethanol(15 ml), and water (5 ml) after 18 h gave methyl 3,4,6-trideoxy-3,4-epithio-2-O-nzethyZ-a-~-aZtvoside (29) (40 mg),m.p. 30", -91" (e 0-6) (Found: C, 50.4; H, 7.2; S,16.6.C,HI4O3S requires C, 50.5; H, 7.4; S, 16.85).With thiourea. (a) Methyl 2,3-anhydro-4,6-O-benzylidene-a-D-taloside (120 mg) after 23 h gave unchanged epoxide (83mg) and methyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epithio-a-D-guloside (13) (16 mg), m.p. and mixed m.p. 113-1 15".(b) Methyl 2,3-anhydro-4,6-0-benzylidene-cc-~-guloside(100 mg) after 42 h gave (t.1.c. with ether-chloroform)unchanged epoxide (42 mg) and methyl 4,6-O-benzylidene-2,3-dideoxy-a-~-threo-hex-2-enopyranoside (14) (39 mg),m.p. 159-161" (from chloroform-ether), - 133" (e0.5) (lit.,25 m.p. 163-164",(c) Methyl 2,3-anhydro-4,6-O-benzylidene-P-~-taloside(200 mg) after 48 h gave methyl 4,6-0-benzylidene-2,3-dideoxy-2,3-epithio-P-~-guloside (18) (24 mg), methyl 4,6-0-benzylidene-2,3-dideoxy-~-~-threo-hex-2-enopyranoside(19) (42 mg), 3,3'-thiobis(methy1 4,6-O-benzylidene-3-deoxy-P-D-idoside) (20) (75 mg), and unchanged epoxide(36 mg), all identified by comparison (m.p., i.r.spectra, andRp) with authentic compounds.(d) Methyl 2,3-anhydro-4,6-0-benzylidene-~-~-guloside(120 ins) after 48 h gave (t.1.c. with ether) unchangedepoxide ( 14 mg) , methyl 4,6-O-benzylidene-2,3-dideoxy-P-D-threo-hex-2-enopyranoside (19 mg), and methyl 4,6-0-benzylidene-2,3-dideoxy-2,3-epithio-P-~-taloside (16) (75 mg) ,25 R. U. Lemieux, E. Fraga, and K. A. Watanabe, Canad. J .Chem.. 1968. 46. 61.- 130').as 6. J. RobertsonandH. G. Dunlop, J . Chem. Soc., 1938, 472.27 J. Jary, K. capek, and J. KovW, Coll. Czech. Chem. Comm.,1963, 28, 2171.28 G.Charalambous and E. E. Percival, J . Chem. SOC., 1954,2443J.C.S. Perkin Im.p. 221-224' (subl.) (from chloroform-ether), E,"- 158" (c 0.6) (Found: C, 59-9; H, 5.8; S, 11-2. CI4H,,O,Srequires C, 60-0; H, 5-75; S, 11.4).(e) Methyl 2,3-anhydro-4,6-di-O-methyl-cc-~-alloside ( 100mg) after 18 h gave (t.1.c. with ether) methyl 2,3-dideoxy-2,3-epithio-4,6-di-O-methyl-a-~-mannoside ( 17 mg) , identicalwith that obtained by the use of thiocyanate.(f) Methyl 2,3-anhydro-4, 6-di-O-methyl-cc-D-mannoside(265 mg) after 18 h gave methyl 2,3-dideoxy-2,3-epithio-4,6-di-O-methyl-a-~-alloside (101 mg), m.p. and mixed m.p.Reactions of EpisuZphides.-Except where further de-tails are given, the proportions of reagents and reactionconditions were as follows.(i) A solution of the episulphide (1 part) and potassiumthiocyanate (2 parts, by weight) in ethanol (60 parts) andwater (10 parts) was boiled under reflux for 72 h, thencooled, diluted with water, and extracted with chloroform.The product was separated into its components by prepara-tive t.1.c.(ether), and these were identified by comparisonwith authentic compounds described in the precedingsections.(ii) A solution of the episulphide (1 part) and thiourea(1-5 parts) in ethanol (100 parts) was boiled under refluxfor 24 h, then treated in the same way.(a) Methyl 4,6-O-benzylid-ene-3,3-dideoxy-2,3-epithio-cc-~-taloside (60 mg) gavemethyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epithio-cc-~-g~ilo-side (23 mg) and starting material (34 mg).(b) Methyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epithio-oc-D-guloside (60 mg) gave methyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epithio-a-~-taloside (3-5 mg) , methyl 4,6-0-benzylidene-2,3-dideoxy-a-~-thveo-hex-2-enopyranoside ( 1 1mg), and starting material (43 mg).52-55'.With fiotassiunz thiocyaizale.(G) Methyl 4,6-0-benzylidene-2,3-dideoxy-2,3-epithio-p-~-taloside (60 mg) gave methyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epithio-p-~-guloside ( 12 mg), methyl 4,6-O-benzylidene-2,3-dideoxy-~-~-thveo-hex-2-enopyranoside (3 mg) , andstarting material (44 mg).(d) Methyl 4,6-0-benzylidene-2,3-dideoxy-3,3-epithio-~-D-guloside (60 mg) gave methyl 4,6-O-benzylidene-2,3-dideoxy-P-~-t~zreo-hex-2-enopyranoside (11 mg) and startingmaterial (47 mg).(e) Methyl 2,3-dideoxy-2,3-epithio-4,6-di-O-methyl-a-~-alloside (112 mg) after 48 h gave methyl 2,3-dideoxy-2,3-epithio-4,6-di-O-methyl-ccc-~-mannoside (23 mg) and start-ing material (82 mg) .(f) Methyl 2,3-dideoxy-2,3-epithio-4,6-di-O-niethyl-cc-r,-mannoside (82 mg) after 48 h gave methyl 2,3-dideoxy-2,3-epithio-4,6-di-O-methyl-a-~-allosicle (7 mg) and startingmaterial (63 mg) .With thiouvea.(a) Methyl 4,6-O-benzylidene-2, Q-dideoxy-2,3-epithio-a-~-taloside (50 mg) gave methyl 4,6-O-benzylid-ene-2,3-dideoxy-a-~-threo-hex-2-enopyranoside (43 mg) .(b) Methyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epithio-a-D-guloside (60 mg) gave the same hex-2-enoside (43 nig) andstarting material (9 nig) .(G) Methyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epithio-~-D-taloside (60 mg) gave methyl 4,6-O-benzylidene-2,3-dideoxy-P-D-threo-hex-4-enopyranoside (8 mg) ancl startingmaterial (44 mg) .(d) Methyl 4,6-O-benzylidene-2,3-dideoxy-2,3-epitliio-P-D-guloside (60 nig) gave the same @-u-hex-2-enoside (45 mg) .'l'he work described in this ancl the following paper wascarried out during a period of study leave granted toM. V. J. by the University of Sri Lanka.4/676 Received, id Afwil, 1974