1976 605Preparation of Polymer-supported Peroxy-acids and their Use to OxidiseOlefins to EpoxidesBy Charles R. Harrison and Philip Hodge," Department of Chemistry, University of Lancaster, Bailrigg, LancasterTreatment of carboxy-substituted polystyrene resins with hydrogen peroxide in methanesulphonic acid gave resinscontaining aromatic peroxy-acid residues (3.5-4.0 mmol g-l). The resins generally reacted with di- and tri-substituted olefins in tetrahydrofuran at 40 "C to give epoxides (or products derived from epoxides) in yieldsgreater than 50%. In cases where mixtures of stereoisomeric epoxideswere formed the proportions of the isomers were essentially the same as those obtained by using monomericaromatic peroxy-acid reagents.Monosubstituted olefins reacted poorly.POLYMER-SUPPORTED organic reagents have severalattractive features, one of which is that reactions inwhich they are used can often be worked up simply byfiltration or centrifugation.2 We describe here thepreparation of some resins containing peroxy-acidresidues and their use to epoxidise olefins.Severalresins containing aliphatic peroxy-acid residues havebeen described previously but those purely aliphaticresins which reacted with olefins to give good yields ofepoxides3 were expl~sive,~ and those which also con-tained aromatic sulphonic acid residues reacted witholefins to give diols5 or low yields of epoxides.6 Thereagents we now describe contain aromatic peroxy-acidresidues and are not explosive. They react with manydi- and tri-substituted olefins to give good yields ofepoxides.Since our work was completed Frkhet andHaque have described some similar resin^.^ They weremainly concerned with the preparation of the resins andone of their methods was similar to the main method weused. They studied the epoxidatioii of only one olefin.An important difference between their resins and ours isthat ours has ca. 80% of the phenyl residues substitutedwhereas theirs had only ca. 15%. This may result insubstantial differences in the swelling properties of theresins and hence in the availability of the reactive groupsin a given solvent.Preparation of Polymeric Peroxy-acids.-The startingmaterials were cross-linked polystyrene resins containingcarboxy-substituents.These were conveniently pre-pared by introducing carboxy-groups into commercialpolystyrene beads by the sequence shown in the Scheme.*From polystyrenes cross-linked with 1% and also with2% of 9-divinylbenzene (DVB) , carboxy-resins with83% of the rings substituted were prepared. A moreC . R. Harrison and P. Hodge, J.C.S. Chem. Comm., 1974,1009.C. G. Overberger and K. N. Sannes, Angew. Chem. Internat.Edn., 1974, 13, 99; C. C. Leznoff, Chem. Soc. Rev., 1974, 3, 66.T. Takagi, J . Polymer Sci., Part B, Polymer Letters, 1907, 5,1031.T. Takagi and M. Aoyama, J . Polymer Sci., Part B, PolymerLetters, 1974,12,681; C. R. Harrison, P. Hodge, and G. M. Perry,unpublished observations.direct way of preparing the carboxy-resin was by co-polymerisation of methyl 9-vinylbenzoate, styrene, andDVB, followed by hydrolysis of the ester group, but theresin prepared in this way tended to break down underthe conditions used to produce the peroxy-acid.SCHEMEThe carboxy-resins prepared according to the Schemecould be converted into peroxy-acid resins by treatmentwith 85% hydrogen peroxide in methanesulphonic acidat 25 "C for 16 h.The peroxy-acid resins generally hadan activity (measured by iodimetry) of ca. 3 . 5 4 . 0mmol [O] per g, indicating that ca. 70% of the carboxy-groups had been converted into peroxy-acid groups.This oxidative capacity is about 4 times greater thanthat of Frkchet and Haque's resins and is comparableto that (5.0 mmol g-l) of commercial 85% m-chloro-perbenzoic acid.Treatment of the carboxy-resin prepared by thesecond of the above methods with hydrogen peroxideunder the above conditions resulted in breakdown of theresin, but a shorter reaction time (4 h) gave a resin withsatisfactory physical properties and an activity of2.51 mmol [O] per g.This resin was not used in theinvestigations described below.An alternative method of preparing the peroxy-acid6 F. Helfferich and D. B. Luten, J , Appl. Polymer Sci., 1964,6 W. J. Hamilton, K. J. Krause, and H. B. Wagner, J . Appl.J. M. J. Frdchet and K. E. Haque, Macromolecules, 1975, 8,* C. R. Harrison, P. Hodge, J. Kemp, and G. M. Perry,8, 2899.Polymer Sci., 1974, 18, 1039.130.Makromol. Chem., 1975, 178, 267606 J.C.S. Perkin Iresin was to treat the formyl resin (Scheme) in methane-sulphonic acid with hydrogen peroxide.This procedureallows the peroxy-acid to be prepared from the formylresin in one step and it gives material comparable inactivity to that from the carboxy-resin, but it is onlyconvenient to carry the reaction out on a small scaleotherwise the strongly exothermic reaction is difficultt o control.There was no significant loss of activity over severalmonths when the resins were stored at -20 "C, but at20 "C they lost half their activity in ca. 70 days. Assome resins containing aliphatic peroxy-acid residuesare known to be explosive, we made numerous attemptsWe next sought to optimise the reaction time and thequantity of reagent. Treating several substrates withthe resins in THF at 40 "C for various periods showedthat 4 h was the most satisfactory reaction time.Theyield of epoxide did not usually increase significantlyafter this period and in some instances it actually fell.As not all the peroxy-acid residues on the resin werelikely to be accessible to the substrate, an excess ofreagent was used. Usually 2 mol. equiv. were used butsome reactions were also carried out with 3 and 4.5mol. equiv. The results (Table 3) show that increasingthe molar ratio from 2 : 1 to 3 : 1 markedly increased theyields of epoxides but that further increases caused theTABLE 1Effect of solvent on reaction yield aYield of epoxide (%) in various solventsLI \ Substrate THF DMF Dioxan EtOhc CHCl, PhH CH,Cl,Cyclo-octene 95 69 3 2 2C yclododecene 57 45a-Methylstyrene 50 41 28 0 0trans-@-Methylstyrene 52 52 40 7 40 Yields of epoxide (determined by g.1.c.) obtained by treating olefin in solvent at 40 "C for 4 h with 2 mol.equiv. of reagentb The product from a-methylstyrene was a mixture (see text) ; yields given are the prepared from 1% cross-linked polystyrene.total of those for all products. c Reagent prepared from 2% cross-linked polystyrene.to detonate our peroxy-acids by impact, but on nooccasion did an explosion occur.Efect of Various Factors 0% the Eficiency with whichthe Peroxy-acid Resins react with 0Zefins.-Ideally, thereaction conditions used with polymer-supported re-agents should cause the polymer to swell so as tofacilitate access to the reactive groups.A major factordetermining the extent of swelling is the solvent. Wetreated a number of olefins with peroxy-acid resins in arange of solvents under standard conditions and deter-mined the yields of epoxide. Typical results (Table 1)show that tetrahydrofuran (THF) is marginally superiorto dioxan and dimethylformamide (DMF), which in turnare better than ethyl acetate. Benzene, chloroform,and methylene chloride, solvents commonly used withmonomeric peroxy-acids,9 are poor solvents. Interest-ingly this order is approximately the same as that of thesolubility of m-chloroperbenzoic acid (CPBA) in thesesolvents.1°A second major factor in determining the extent ofswelling is the degree of cross-linking; the lower this isthe more the resin swells.Although the chloromethyl-ation reactions used in the preparation of the reagentsalmost certainly increased the extent of cross-linking,llthe peroxy-acids prepared from 1 yo cross-linked poly-styrene usually gave greater yields of epoxide than thoseprepared from 2 yo cross-linked polystyrene. Sometypical results are summarised in Table 2. Otherworkers have reported that decreasing the percentage ofcross-linking improves reaction yields.12H. 0. House, 'Modern Synthetic Reactions,' 2nd edn.,Benjamin, Menlo Park, California, 1972, ch. 6.lo L. F. Fieser and M. Fieser, ' Reagents for Organic Synthesis,'Wiley, New York, 1967, p. 136. m-Chloroperbenzoic acid has asolubility in THF and dioxan of more than 150 g per 100 g at25 "C.yields to fall.The latter effect stems partly from thedifficulty of efficiently recovering the product from therelatively large amount of resin.TABLE 2Effect of percentage of cross-linking on reaction yield 0Substratetrans-oct-2-eneCyclo-octeneCyclo-octeneCyclododecenea-Methylstyrenea-Methylstyrene btrans-P-Me thyls tyreneSolventTHFTHFDioxanTHFTHFDioxanTHF1%cross-linkedreagent e489569575041542%cross-linkedreagent38726651682062Yield of epoxide (determined by g.1.c.) obtained by treatingthe olefin in solvent a t 40 OC with two mol. equiv. of peroxy-acid for 4 h. Reagents prepared from1 and 2% cross-linked polystyrenes. As noted in the text,the actual percentage of cross-linking in the peroxy-acid resinis probably higher.See note b, Table 1.Reaction of the Poiymev-supported Peroxy-acids with aRange of 0lefins.-A range of olefins was treated with theperoxy-acid resins using the best general conditionsestablished by the above work.The polymer wasfiltered or centrifuged off and the filtrate was analysed,usually by g.1.c. Withmost tri- and di-substituted olefins, yields of epoxide inexcess of 50% were obtained, but with monosubstitutedolefins the yields were poor. This is not surprising asthe resin reagent used in THF would be expected to beless reactive than both CPBA and perbenzoic acid usedThe results are given in Table 4.*l J. A. Patterson in ' Biochemical Aspects of Reactions onSolid Supports,' ed.G. R. Stark, Academic Press, New York,1971, p. 204.l2 W. Heitz and R. Michels, Angew. Chem. Infevnak. Ed%., 1972,11, 2981976 607in chloroform, and the latter often react far from that these resin reactions, particularly that of cyclo-quantitatively with tri- and &-substituted olefins and octene, are markedly slowed by the need for the sub-often only poorly with monosubstituted 01efins.~ The strate to diffuse into the resin, and/or by the inaccessi-resin reagent would be expected to be less reactive bility of some of the peroxy-acid residues, and/or bybecause it is an alkyl-substituted perbenzoic acid microenvironmental effects. It is not clear why the(electron-releasing substituents decrease the reactivity 13) polymeric reagent was more stable than the monomeric.TABLE 3Effect of an excess of reagent on reaction yield aYield of epoxide ( o/o)h t I o / / OCross-linking Reaction 2 equiv. 3 equiv.4.6 equiv.Substrate in reagent c time (h) reagent reagent reagenttvaizs-Oct-2-ene 1 4C yclododecene 1 4a-Methylstyrene 2 22trans-P-Methylstyrene 2 44-Xlethoxycarbonylcyclohexene 1 454 76 3 064 82 6874 9152 6536 19Yield of epoxide (determined by g.1.c.) obtained by treating the substrate in THF with various amounts of reagent See noteb, Table 1. C See note c, Table 2.and because it is used in an ethereal s01vent.l~ Thereactions with the resin reagent might also be slowerbecause the substrate needs to diffuse into the resin toreact. In an attempt to throw light on this, cyclo-octene and tram-P-methylstyrene were each separatelyTABLE 4Reactions of peroxy-acid resins with various olefins inTHF at 40 "C for 4 hEquiv. of Yield of UnchangedOlefin reagent epoxide (%) olefin (yo)Met11 ylc yclohexene 2 65 b 0Choles teryl acetate 3 42 C 52tra~s-Oct-2-ene 3 76 9cis-Oct-2-ene 2 60 13Elaidic acid 3 53 40tx-Methylstyrene 2d 50 12tru.ans-p-Methylstyrene 3 62 8Cyclo-octene 2 95 2Cyclododecene 3 82 164-Me thoxycarbonyl- 2 36 284-Carbox ycyclohexene 3d 35 313-Acetoxycyclohexene 3 3 913-H ydroxycyclohexene 3 28 3Norbornene 2d 64 8Styrene 2 7 60Methyl undec-10-enoate 2 1 82cyclohexenea Unless indicated otherwise the reagent was prepared from1 "/; cross-linked polystyrene and yields were determined byg.1.c. b Product was 2-methylcyclohexene.C Products iso-lated by p.1.c.; trio1 monoacetate also obtained (6%). d Re-agent prepared from 2% cross-linked polystyrene. 6 Complexmixture of oxidation products (see Experimental section).treated with p-t-butylperbenzoic acid and with the resinperoxy-acid in THF at 40 "C and the reactions werefollowed by g.1.c. The results (Figure) show that themonomeric reagent reacted faster with both substratesinitially but it also decomposed faster, with the neteffect that higher yields were obtained with the resinreagents. The magnitude of the differences in theinitial rates of the epoxidations indicates, especially asthe monomeric reagent was decomposing more rapidly,l3 B. M. Lynch and I.H. Pausacker, J . Chem. Soc., 1956, 1625.l4 I-'. Renolen and J. Ugelstad, J. Chim. Qhys., 1960, 57, 634;N. X. Schivartz and J . H. Blumbergs, J . Org. Chem., 1964, 29,1976.Use of the resin reagent simplifies the reaction pro-cedure by allowing the excess of and spent reagent to beseparated from the product by filtration. This isparticularly useful when the substrate is acidic. Acommon side reaction in epoxidations is reaction of theepoxide with the carboxylic acid (spent peroxy-acid) togive hydroxy-esters. These also would be removed byUE1t l h2Reactions of peroxy-acids with olefins in THF a t 40" : (A), cyclo-octene and resin peroxy-acid; (B), cyclo-octene and p-t-butylperbenzoic acid ; (C), P-methylstyrene and resin peroxy-acid ; (D), p-methylstyrene and p-t-butylperbenzoic acidfiltration as they would be attached to the polymer, andthis is probably the main reason why not all the startingmaterial was accounted for.In two reactions the initially formed epoxide wasconverted into other products.The epoxide from 1-methylcyclohexene simply rearranged to 2-met hylcyclo-hexanone,15 but in the epoxidation of a-methylstyreneother reactions occurred as well. Thus, acetophenonewas obtained in addition to a-phenylpropionaldehyde.The complexity of the reaction is not a consequence ofSOL, 1959, 81, 658.l5 R. Filler, B. R. Camara, and S. M. Naqvi, J . Amer. Chenz608 J.C.S. Perkin Iusing the polymeric reagent, as the same products wereobtained by using CPBA.A few products, mainly thosefrom the less reactive olefins, contained small amountsof y-butyrolactone produced by oxidation of the THF.When a suspension of the peroxy-acid in THF washeated for 4 h, sufficient y-lactone was formed to accountfor 9% of the oxidation capacity of the resin.Several of the olefins listed in Table 4 could afford amixture of stereoisomers and it was of interest to seewhether the polymeric reagent gave different stereo-chemical results from the monomeric reagents. Thereactions of 4-methoxycarbonyl- and 4-carboxy-cyclo-hexene both gave mixtures of stereoisomers, but in eachcase the ratio (cis : trans 30 : 70 and 33 : 67, respectively)was essentially the same as that (31 : 69) obtained bytreating the ester with CPBA in THF.Previousworkers found a ratio of 34 : 66 from treatment of theester with perbenzoic acid in etherl6 and a ratio of29 : 71 with the ester and CPBA in ch10roform.l~ Un-fortunately 3-acetoxycyclohexene was too unreactivefor a completely satisfactory study to be made, but3-hydrox ycyclohexene gave stereoisomeric epoxides inthe ratio 91 : 9. This is similar to that reported(cis: trans 90: 10) for the reaction of this olefin withperbenzoic acid in ether at 5 "C.18 We could only detectone product, the exo-epoxide, from the norborneneoxidation whether the resin reagent or CPBA was used.This is in agreement with other workers who claim thatthe exo-epoxide is the exclusive l9 or almost exclusiveproduct,20 but Kwart and Takeshita have reported thatboth isomers are formed, the ex0 : endo ratio being94 : 6.21 Hence it appears from this limited number ofexamples that the polymeric reagent gives essentiallythe same stereochemical results as the monomericreagents.EXPERIMENTALThe polystyrene resins used were Biobeads SX1 andS X 2 (Bio-Rad Laboratories, California).These werestyrene-1 % p-divinylbenzene and styrene-2% p-divinyl-benzene copolymers, respectively, in the form of 200-400mesh beads. Resins were filtered off by using no. 4 gradesintered glass filters and were dried in a vacuum oven(0.1 mmHg). G.1.c. was carried out with a Pye 104machine (flame ionisation detector) and a 5 f t columncontaining either Apiezon L or SE30 as stationary phase.Peak areas were determined by triangulation except inthose reactions where a mixture of stereoisomers wasobtained; in these cases a curve resolver was used.Unlessindicated otherwise, ratios were determined by comparisonwith authentic mixtures. All the olefins and most of theepoxides used in this work were obtained commercially.The other epoxides were prepared by standard methods.Introduction of Carboxy-groups into Polystyrene Resins .--This was carried out as described previously * (Scheme).l6 G. Bellucci, F. Marioni, and A. Marsili, Tetrahedron, 1972,28,3393.l7 J. W. Huffman, C. B. S. Rao, and T. Kamiya, J. Org. Chem.,1967, 32, 697.la P. Chamberlain, M. L. Roberts, and G. H. Whitham, J .Chem. Soc. (B), 1970, 1374.From Biobeads SX1 a carboxy-resin with 83% of the ringssubstituted was obtained, Biobeads SX2 also gave aproduct in which 83% of the rings were substituted.Comparable results were obtained on subsequent occasions.C@oZymerisation of MethyZ p- Vinylbenzoate, Styrene, andp-Divinylbenzene : Hydrolysis of the Product.-Methyl p-vinylbenzoate was prepared by methylating P-vinylbenzoicacid 22 with diazomethane.Nitrogen was bubbled througha rapidly stirred solution (ca. 350 rev. min-l) of polyvinylalcohol (3.0 g; l@ 125 000) in water (400 ml) a t 95 "C. Tothis solution was added a mixture of benzoyl peroxide(0.3 g), methyl p-vinylbenzoate (10.0 g), and p-divinyl-benzene (0.7 ml of a 1 : 1 mixture of p-divinylbenzene m dl-ethyl-4-vinylbenzene). Heating and stirring under nitro-gen were continued overnight.The flask was then cooledto 20 "C and water (700 ml) added. The beads wereallowed to settle and the liquid was decanted. The beadswere washed with hot water then transferred to a filter,washed with acetone and methylene chloride, and dried(60 "C). This gave white beads (8.4 g, 81%), vm,(KBr)1730cm-l.The ester groups were hydrolysed by treating the resinfor 16 h with an excess of the potassium butoxide-waterreagent in 1,8-dimethoxyethane heated under reflux. *The carboxy-resin had v,,(KBr) 1 735m and 1 690s cm-1.Preparation of Polymeric Peroxy-acids from Carboxy-resins.-(a) A stirred mixture of the carboxy-resin (5 g;from Biobeads SXl) and methanesulphonic acid (15 ml)was kept a t 25-30 "C while hydrogen peroxide (85%;10 ml) in methanesulphonic acid (10 ml) was carefullyadded dropwise.Control of temperature is important.When the addition was complete the mixture was stirredfor 16 h a t 20 "C. The resin was then filtered off, washedsuccessively with THF and methylene chloride, and driedunder vacuum (20 "C). It was stored a t -20". Todetermine the activity of the product a portion (120-140 mg) was treated with methylene chloride (2 ml),potassium iodide (340 mg), and sulphuric acid (5 ml; 63).The mixture was set aside with occasional shaking for 1 h.It was then titrated against sodium thiosulphate (0. ZN),starch indicator being added near the end-point. The resinhad an activity of 3.90 mmol [O] per g, corresponding to70% of the carboxy-groups being oxidised. Similar re-actions generally gave resins with activities of 3.5-4.0mmal g-l.(b) Similar treatment of the carboxy-resin from BiobeadsSX2 gave a peroxy-acid resin with an activity of 3.82 mmolg-l (69 % of carboxy-groups oxidised) .(c) The carboxy-resin obtained via polymerisation ofmethyl p-vinylbenzoate was treated with hydrogen per-oxide as above.This gave a rubbery product which waspartially soluble in methanesulphonic acid. A reactiontime of 4 h gave a resin with satisfactory physical propertiesand an activity of 2.51 mmol g1 (ca. 41% of carbosy-groups oxidised) .Preparation of Peroxy-acid Resin from Fornzyl Resin .-( a ) A stirred mixture of formyl resin 8 (1.0 g ; from BiobeadsSX2) and methanesulphonic acid (2.5 ml) was treated drop-wise with an ice-cold solution of hydrogen peroxide (1 .O ml .85%) in methanesulphonic acid (7.5 ml) so that the interna1Q S.B. Soloway and S. J . Cristol, J . Org. Chem., 1960,25, 32720 H. C. Brown, J . H. Kawakami, and S. Ikegami, J. Arne?,21 H. Kwart and T. Takeshita, J. Org. Chem., 1963, 28, 670.22 J. R. Leebrick and H. E. Ramsden, J . Org. Chem., 1958,23Chem. SOC., 1970, 92, 6914.9351976 609temperature was kept between 25 and 30 "C. Morehydrogen peroxide (1.0 ml) was then added and the mixturewas stirred at 20 "C for 5 h. The resin was then filteredoff, washed with methanesulphonic acid, and re-treatedwith hydrogen peroxide, but this time the mixture wasstirred overnight. The product was filtered off, washed,and dried as before.The product had an activity of4.29 mmol 8-l (78% of formyl groups oxidised).(b) The formyl resin prepared from Biobeads SXl re-acted similarly.Oxidation of Cyclo-octene by Polymeric Peroxy-acid.-Amixture of cyclo-octene (55 mg, 0.5 mmol), THF (2 ml),and peroxy-acid resin (256 mg of activity 3.90 mmol g-l,1.0 mmol) was stirred at 40 "C for 4 h, cooled, and filtered.The resin was washed several times with small portions ofTHF. G.1.c. analysis of the combined filtrates showed thatthe yield of epoxycyclo-octane was 95% and that 2% of thecyclo-octene was unchanged.Otlzer Oxidations (Tabbs 1-3) .-The reactions sum-msrised in the Tables are only representative. About twiceas many reactions were actually carried out by the aboveprocedure, except that in some instances the resin wasremoved by centrifugation.Different batches of resingave similar results. The reproducibility of the yields wasestimated to be &loyo of the values given. The combinedyields of oxidation products and unchanged startingmaterial usually exceeded SO%, except when 4.5 mol.equiv. of reagent were used.Reactions summavised in the Figure.-fi-t-Butylper-benzoic acid was prepared by the reported method.23 Thesubstrates (0.5 mmol) in THF (10 ml) at 40 "C were separ-ately treated with the peroxy-acids (1 mmol) in the usualmanner. Periodically, samples were removed and theratio of epoxide to olefin was determined by g.l.c., thesamples from the reactions with the monomeric reagentsbeing first treated with sodium sulphite to quench theperoxy-acid and sodium carbonate to remove the carboxylicacid. The g.1.c.analyses showed that the amounts ofepoxide and olefin present were sufficient to account formost of the starting material.Reactions sumnzarised in Table 4.-These were carried outby a procedure similar to that for the oxidation of cyclo-octenc.The product from the oxidation of l-methylcyclo-hexene was identified as 2-methylcyclohexanone by g.l.c.,lH n.m.r., and i.r. analysis (comparison with an authenticsample).The crude product from the oxidation of cholesterylacetate was separated by preparative t.1.c. (KieselgelHF,,, ; chloroform-hexane) . This gave starting material(52%), the epoxide as a solid (42%), and trio1 monoacetate(6%).Each compound was identified by t.1.c. comparisonwith authentic samples. The first two products also hadsatisfactory 1H n.m.r. spectra.The product from one oxidation of a-methylstyrene wasanalysed by g.1.c. and by lH n.m.r. and i.r. spectroscopy(comparison with authentic samples). I t contained a-methylstyrene (27 %), a-phenylpropionaldehyde (2 1 %), andacetophenone (13%). The last was isolated as the 2,4-dinitrophenylhydrazone. Treatment of a-methylstyrene inmethylene chloride at 40 "C with CPBA for 2 h also gave amixture of a-methylstyrene oxide, a-phenylpropionaldehyde,and acetophenone.The product from 4-methoxycarbonylcyclohexene con-tained (g.1.c. analysis on the SE30 column) the cis- andtrans-epoxides in a ratio of 31 : 69 when 2 mol.equiv. ofresin peroxy-acid were used (Table a), and a ratio of 29 : 71when 4.6 mol. equiv. were used (Table 3). The samesubstrate in THF reacted with CPBA at 40 "C to give a40% yield of epoxides with a cis : trans ratio of 31 : 69. Asimilar reaction a t 0 "C gave the epoxides in the ratio30 : 70. Authentic samples of the two epoxides were notavailable and the g.1.c. peaks were assigned by assumingthat the major product (which had the shorter retentiontime) was the trans-isomer. Previous workers using siliconecolumns found this to be the case.16 The product from4-carboxycyclohexene was methylated (diazomethane) priorto g.1.c. analysis.The product from 3-acetoxycyclohexene contained (g.1.c.analysis on the SE30 column) two epoxides in approximatelyequal amounts. The product from 3-hydroxycyclohexene,analysed similarly, contained two epoxides in the ratio91 : 9, the major isomer having the shorter retention time.As authentic samples of the two epoxides were not availablethe peaks could not be assigned unambiguously.The product from norbornene was analysed by g.1.c.(SE30 column). Kwart and Takeshita 21 also used a siliconecolumn. Only one epoxide peak was obtained undervarious operating conditions. It had a retention time thesame as that of authentic exo-epoxide. The 1H n.m.r.spectrum of the crude product by comparison with that ofauthentic exo-epoxide, indicated that this epoxide was themajor, if not the only product. A similar result wasobtained when norbornene was treated with CPBA inTHF.The cis : trans ratio was 33 : 67.The yield of epoxide was 80%.We thank the S.R.C. for financial support and LaporteIndustries Limited for a gift of hydrogen peroxide.[6/2027 Received, 17th October, 1975123 L. S. Silbert, E. Siegel, and D. Swern, J . Org. Chem., 1962,27, 1336
展开▼
