One-step synthesis ofcis-5,6-dideuterio-cis-cyclo-octene and its thermal stability towards concerted, intramolecular transfer of hydrogen

摘要

342 J'.C.S. Perkin IOne-step Synthesis of cis-5,6-Dideuterio-cis-cyclo-octene and ItsThermal Stability towards Concerted, Intramolecular Transfer ofHydrogenBy A. J. Bellamy, Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh EH9 3JJcis-5,6-Dideuterio-cis-cyclo-octene has been synthesised in one step from cyclo-octa-I ,5-diene by reductionwith dideuteriodi-imide. On prolonged pyrolysis a t 250". intramolecular transfer of the vicinal 5- and 6- protonsto the double bond does not occur; thus no 1,2-dideuterio-cis-cyclo-octene is formed.SEVERAL examples of the concerted, intermoleculartransfer of two vicinal hydrogen atoms to a doublebond are known,l but few intramolecular exampleshave been reported.2 In all known examples of thistype of reaction a decrease in free energy a t the origin ofmigration occurs (see ref.1). However, given a suitableconfiguration and/or conformation, a thermally neutral,intramolecular hydrogen transfer might conceivablyoccur.Transannular transfer of a single hydrogen atom iscommon in the cyclo-octane series.3 Molecular modelssuggest that cis-cyclo-octene might be able to adopt asuitable conformation for the concerted transfer of twovicinal 5- and 6-protons to the double bond. In theextreme-boat conformation of cis-cyclo-octene, theinternuclear distances between the endo-5- and 6-protonsand C(l) and C(2) respectively, are both ca. 2.2 A.However, in this conformation there are severe non-bonded interactions between the 4- and 7-protons andthe 3- and 8-protons as well as many eclipsed inter-actions.A twist-boat conformation would relievemost of these non-bonded interactions to some degree;in such a conformation, the same hydrogen-carboninternuclear distances are ca. 2-2 and ca. 2-5 A. Inboth conformations the 5- and 6-protons appear closeenough to the double bond to make transfer possible,but the attainment of these conformations may besterically difficult.Since an intramolecular transfer of hydrogen incis-cyclo-octene would produce no observable chemicalchange, cis-5,6-dideuterio-cis-cyclo-octene was studied.S. Hunig, H. R. Muller, and W. Thier, Angew. Chem. Internat.Edn., 1965, 4, 271; W. von E. Doering and J. W. Rosenthal,J . Amer. Chem. SOC., 1967, 89, 4534; I.Fleming and E. Wild;smith, ' Orbital Symmetry Correlations in Organic Reactions,Chemical Society Symposium, Cambridge, 1969.Transfer of the cis-vicinal deuterium atoms in thiscompound would again produce no observable chemicalchange, but transfer of the cis-vicinal 5- and 6-protonswould give 1,2-dideuterio-cis-cyclo-octene. Such atransfer would be detectable by n.m.r. spectroscopy.The material studied was in fact a mixture of cis-5,6-dideuterio-cis-cyclo-octene (72.7y0), 5-deuterio-cis-cyclo-octene (23.6y0), and cis-cyclo-octene (3.7).Neglecting secondary isotope effects, complete equili-bration of cis-5,6-dideuterio-cis-cyclo-octene with 1,2-dideuterio-cis-cyclo-octene by a concerted transfer ofhydrogen should favour the 5,6-isomer by a factor oftwo. Thus, for the material used, the ratio olefinic : al-lylic : other protons (1.95 : 3.95 : 6.39 in the startingmaterial) should change to 1-44 : 4-00 : 6-87 in theequilibrated material.However, heating the cis-5,6-dideuterio-compound at 250" during 7 days producedno significant change in the n.m.r. spectrum. Con-certed transfer of hydrogen does not therefore occur inthis system at this temperature. At higher tempera-tures polymerisation becomes significant.Preparation of cis-5,6-Dideute~io-cis-cyclo-octene.-5,6-Dideuterio-cis-cyclo-octene has been prepared by Copeet aL4 in a three-stage synthesis from cis,-cyclo-octa-1 J-diene monoepoxide. The cis-cyclo-octene obtainedK. MacKenzie, J . Chem. SOC., 1965, 4646.3 A.C. Cope, M. M. Martin, and M. A. McKervey, Quart.4 A. C. Cope, G. A. Berchtold, P. E. Peterson, and S. H.Rev., 1966, 20, 119.Sharman, J . Amer. Chem. SOC., 1960, 82, 63661972 343contained 74 of the 2H2-compound (deuteriumatoms expected to be in a cis-configuration because ofthe method of synthesis), 13 of the C2HJ- and 13of the 2H3-compound the third deuterium atom wasat C(5). A more convenient preparation of cis-5,6-dideuterio-cis-cyclo-octene involves reduction of cis,cis-cyclo-octa-l,5-diene with dideuteriodi-imide. Since di-imide is a cis-reducing agent,5 the cis-cyclo-octeneformed must have the two 5- and 6-deuterium atomsin a cis-relationship. cis-Cyclo-octene is reduced bydi-imide more rapidly than cis,cis-cyclo-octa-l,5-diene ;thus use of a large excess of the diene is necessary forefficient production of cis-cyclo-octene.From the ratio of products, and the isotopic abund-ance in the isolated cyclo-octene and cyclo-octane, itwas calculated that there had been an uptake of 88-1mmol of deuterium atoms and 15.9 mmol of hydrogenatoms.Of the hydrogen, only 2.5 mmol could come fromthe acetic 2Hacid used t o generate the di-imide frompotassium azodicarboxylate. The most likely sourcefor most of the hydrogen is the potassium azodicarb-oxylate, which contained ca. 0.3 hydrogen even afterprolonged drying. To determine what proportion ofthis hydrogen was exchangeable, a weighed amount ofpotassium azodicarboxylate was dissolved in aqueousdeuteriated methanol of known isotopic composition.Integration of the hydroxy-proton signal in the n.m.r.spectrum and comparison with the 13C side-band of thesignal for the methanolic methyl group gave the amountof exchangeable hydrogen liberated by the potassiumazodicarboxylate. We calculated that 21.8 mmol ofexchangeable hydrogen was present in the amount ofthe salt used in the reduction, more than enough toaccount for the reduction results.A reduction product of higher deuterium contentcould probably be obtained by precipitating the saltfrom deuteriated solvents, ensuring that any exchange-able hydrogen in the salt was replaced by deuterium,but this was not necessary for the present work.EXPERIMENTALN.m.r.spectra were determined using a Varian HA 100spectrometer, and mass spectra using an A.E.I.MS 902double-focusing spectrometer. G.1.c. analyses were per-formed with a 50 m, 0.25 mm i.d. capillary column coatedwith Apiezon L, operating at 40".Potassium Azodicarboxylate.6-Azoformamide (70 g) wasadded with stirring to aqueous potassium hydroxide(175 ml; 1 : 1 w/w) at 0". The solid partially dissolved,ammonia was evolved, and the product was precipitated.The latter was filtered off under nitrogen when the evolutionof ammonia had ceased, and dried (water pump). It wasthen dissolved in the minimum volume of water (ca. 400 ml)a t O", and immediately filtered into ethanol (ca. 1.6 1) a t0". The precipitated yellow potassium azodicarboxylatewas filtered off, washed twice with methanol a t Oo, drieda t the pump, and finally dried in vacuo (H2S04); yield* All silver nitrate solutions were saturated with ether.t A control experiment showed that 170/, of the cyclo-octeneThe extraction was lost during the silver nitrate treatment.was followcd by g.1.c.100 g (Found: C, 12.3, 12-2; H, 0.3, 0-25; N, 13.85, 13.9.Calc.for C,N,K,O,: C, 12.35; H, 0.0; N, 14-4y0). Afterfurther drying in vacuo at 100" (P,O,) a 0.5 g sample hadlost 5 mg in weight, and gave the following analyticalfigures: C, 9-7, 9-55, 9.85; H, 0.3, 0-4, 0.25; N, 13.4, 13-45,13.05. These figures suggest that some decompositionof the salt had occurred (lit.,s C, 12.45; H, 0.5; N,14-15y0).Reduction of cis,cis-Cyclo-octa-1,5-diene.-The diene waspurified before use by the formation of its silver nitratecomplex.7 Regeneration by steam-distillation gave mater-ial which was 99.98 pure by g.l.c., and contained lessthan 0-001 cis-cyclo-octene.The purified diene (50-8 g,0-47 mol) was added to a suspension of potassium azo-dicarboxylate (19.3 g, 0.10 mol) in dry ether (200 ml).The apparatus was flushed with nitrogen, then acetic2Hacid (5.0 g, 0.082 mol; 98.5 CaH1) in dry ether(20 ml) was added at the rate of 2-3 drops rnin-l withvigorous stirring. After stirring for 24 h, a further quantityof acetic ZHIacid in ether (5.0 g in 20 ml) was added a tthe same rate. The reduction was followed by g.1.c.(see the Table). After 180 h, the solid was filtered offProduct ratios in the reduction of czs,cis-cyclo-octa-1,B-dienet/h Cyclo-octadiene Cyclo-octene Cyclo-octane24 96.0 3.6 0.484 93.7 5.4 0.9132 92.1 6.4 1.5180 91.0 7.0 2.0and washed with dry ether (100 ml), and the organicsolution was dried (Na,CO,; 6 h).To remove the excessof cyclo-octa-1,5-diene, the solution was first extractedwith aqueous silver nitrate * (100 ml) and water (350 ml),the latter to dissolve the large amount of complex formedin this extraction, and then with saturated aqueous silvernitrate (3 x 100 ml). The resulting ethereal solution,free from cyclo-octa-l,5-diene, t was dried (MgSO,), filtered,and evaporated to give cyclo-octene (75) and cyclo-octane (25).The mixture was separated by preparative g.1.c. O-1ml samples, 10 f t x 0.25 in column, Gas Chrom P (40-60mesh) with a 30 coating of 30 silver nitrate in glycerol,75O, helium carrier gas, 30 lb in-2.The separated productswere shown to have the following isotopic compositionby low-voltage mass spectrometry : cyclo-octene, 2H,72-7y0, 2H,, 23.6y0, 2Ho 3.7 ; cyclo-octane, ,H453-7, 2H3 34.4y0, 2H2 lO-O, and 2Hl 1.9.The n.m.r. spectrum (CC1, solution) of the cyclo-octeneshowed the following absorptions: T 4.45 (1-95H, m,olefinic), 7-9 (3*95H, m, allylic), and 8.55 (6.39H, m, other)(the integrations were based on an isotopic compositionof 73 t2H, and 24 2H, in the third region). Integr-ation of the spectrum of undeuteriated cyclo-octene gavevalues of 1-96H, 3-97H, and 8.03H, respectively. Anotherreduction, where the mixing of reagents was carried outin a dry-box, gave products with the following isotopiccomposition : cyclo-octene, 79.4y0, ,H1 19.4,zHo 1.2 ; cyclo-octane, ,Ha 66.1y0, C2H3 29.0yo,E.J. Corey, D. J. Pasto, and W. L. Mock, J . Amer. C h e wSOC., 1961, 83, 2957.J . Thiele, Annalen, 1892, 271, 127. ' A. C . Cope, C. L. Stevens, and F. A. Hochstein, J . Amer.Chem. Soc.. 1950, 72, 2510; A. C. Cope and F. A. Hochstein,ibid., p. 2515J.C.S. Perkin IL2H2 4.9. The isotopic abundance in cyclo-octane isvery close to that calculated for the reduction of bothdouble bonds by using the isotopic abundance found forthe reduction of the first double bond. This indicatesthat the isotopic abundance in the di-imide must beuniform throughout the reaction.Determination of the Amount of Exchangeable Hydrogenin Potassium AzodicarboxyZate.-Deuterium oxide (4.005 g ;99.7 2H2) and methan2Hl01 (1-304 g ; 98.5 2H1)were mixed in a dry-box, and the hydrogen content wasdetermined by integrating the hydroxy-proton signal andcomparing it with the 13C side-band of the methyl groupsignal in the n.m.r.spectrum. Potassium azodicarboxylate(1.01 1 g) was added slowly to deuterium oxide-methan-2Hlol (2-734 g) in a dry-box, and the resulting solutionwas again analysed by n.m.r. After correction for thehydrogen present in the solvent, this indicated that 1.14mmol of exchangeable hydrogen atoms were present in1.011 g of the salt (i.e., 21.8 mmol of exchangeable hydrogenatoms in the 19.3 g of potassium azodicarboxylate usedin the large-scale reduction of cyclo-octa-l,5-diene).Pyrolysis of cis-5,6-Dideuterio-cis-cyclo-octene.-The cyclo-octene (73 mg), in a sealed tube, was heated at (a) 180"and (b) 250" during 170 h. The product, pure cyclo-octene(by g.l.c.), was dissolved in CCl, and analysed by n.m.r.The average of several integrated spectra gave the followingratios of protons (based on an average of 12.31 protonsper molecule) :T/"C Q Olefinic Allylic Otherb 1.95 3-95 6.39180 1.96 4.02 6.32250 1.90 4-04 6-35Q Polymerisation became significant at 300 "C. b Startingmaterial.1/1365 Received, August 3rd 1971