摘要:
234 J.C.S. Perkin INon-coplanar Double Bonds ; The Existence of AdamanteneBy William Burns, David Grant, M. Anthony McKervey," and George Step, Department of Chemistry,The Queen's University, Belfast BT9 5AGOn exposure to n-butyl-lithium, 1.2-di-iodoadamantane yields t h e highly strained bridgehead olefin adamantene,which can be intercepted in a Diels-Alder reaction with butadiene a t -78 "C: in t h e absence of butadiene itdimerises spontaneously in a [2 + 23 fashion.OF the various statements of the amount of distortion anolefin can sustain and still constitute an isolable substance,much historical interest attaches to that of Bredt,l whopostulated that compounds of the pinane and camphaneclass cannot have a double bond at a bridgehead position.Bredt's rule represents the first qualitative expressionof the limits to the amount of torsional strain an olefincan tolerate.The consequences of this type of distortionin which the 2p orbitals of the n-bond are skewed relativeto each other by an amount proportional to the twist ofthe n-bond have received much scrutiny, and there havebeen several attempts to probe the limits of Bredt'srule.2 trans-Cyclo-octene (1) provides an illustrationof one way of imposing torsional strain in an olefin : themethylene chain linking the trans-substituents in pro-1 J. Bredt, H. Thouet, and J. Schmitz, Annalen, 1924, 437, 1.2 For a recent review of Bredt's rule see G. L. Buchanan,3 K. Ziegler and H. Wilms, Annalen, 1950, 567, 1.4 Cf. N. L. Allinger, J . Amer.Chem. SOC., 1958, 80, 1953;Chem. SOC. Rev., 1974, 3, 41.W. L. Mock, Tetvahedvon Letters, 1972, 476;.jection (2) is of insufficient length to permit completeoverlap of the $-orbitals. Better overlap is attained ifthe vinylic carbon atoms undergo slight rehybridisation,as in (3), with incorporation of some s character into thelobes of the x-bond, and this type of compromise appearsto be quite general in strained 01efins.~ A second, closelyrelated way of imposing torsional strain is to place thedouble bond at a bridgehead position of a bicyclic system,as in bicyclo[3.3.l]non-l-ene (4).536 Olefins (1) and (4)are highly reactive, though isolable substances, and afeature common to both is the presence of a trans-doublebond in an eight-membered carbocycle. These similari-ties led Wiseman to elaborate Bredt's rule to include acomparison of the strain energies of Irans-monocyclo-alkenes with those of the corresponding bicyclic bridge-head alkenes.Thus, trans-cycloheptene ' and its bi-J. R. Wiseman and W. A. Pletcher, J . Amer. Chem. SOL,J. A. Marshall and H. Faubl, J . Amer. Chem. Soc., 1970,E. J. Corey, F. A. Carey, and R. A. E. Winter, J . Amey.1970, 92, 956.92, 948.Chem. SOC., 1965, 87, 9341976cyclic analogue (5) can exist, but cannot be isolated.The extension of this analogy to the six-membered seriesis of particular interest since there is no direct evidencefor the existence of trans-cyclohexene." However, thetrans-cyclohexene analogues bicyclo{2.2.l]hept-l-ene(6) lo and perfluorobicyclo[2.2.l]hept-l-ene (7) l1 havebeen trapped in a Diels-Alder reaction with furan.( 5 ) ( 4 )( 7 )Bredt's rule relates specifically to bridgehead doublebonds in bicyclic systems.Bridged-ring tricyclic systemsof comparable size are much more rigid, the distortionand excess strain produced by a bridgehead double bondshould be much more severe, and there are insufficientdata available to establish whether or not Wiseman'smodification of Bredt's rule applies to such systems.The observation of homoadamantene (8), the tricyclicanalogue of trans-cycloheptene has recently been claimed ;it dimerises spontaneously.12J3 Adamantene (9), thetricyclic analogue of trans-cyclohexene, represents a muchmore extreme case of a highly inflexible system in which* The intermediacy of trans-cyclohexene in photolysis ofcis-cyclohexene has been postu1ated.O t After the publication of a preliminary account of this work,lsDr.D. Lenoir, who was independently pursuing the same objec-tive,ls kindly informed us of a superior route to the di-iodide,involving the reaction of protoadamantan-Cone with neatphosphorus tri-iodide. l7 In subsequent work we adopted thisalternative with the slight modification that the reaction wasconducted in hot chloroform whereupon a 92% yield of the di-iodide was obtained.1 Although simple cyclobutanes are susceptible to catalytichydrogenolysis in the gas phase at elevated temperatures,ls thiswas not the case with the cyclobutanes (13) and (14) a t 220 "Cin hydrogen on a platinum-silica catalyst: instead two newC,,H,, hydrocarbons (ratio 2 : 1) were produced quantitatively.Use of rhodium-alumina or palladium-pumice a t 200-220 O Cgave essentially the same results.The structures of these re-arrangement products, which correspond in their g.1.c. retentiontimes with two of the minor products of the aluminium tri-chloride reaction, are under investigation (work in progress withDr. J. J . Rooney).235the opportunity for distributing the large angle strainover several bonds is at a minimum. Nevertheless,adamantene can exist, though it cannot be isolated, aswe now illustrate.The synthetic route chosen involved dehalogenation ofa 1,2-dihalogenoadamantane with an alkyl-lithium.Treatment of a cold solution of protoadamantan-4-onehydrazone (10) and triethylamine in chloroform withiodine l4 yielded a separable mixture of 1,2-di-iodoada-mantane (1 1) (14%) t and 4-iodoprotoadamant-4-ene(12) (25%).The structural assignments are based onanalytical and spectral data (see Experimental section),the mode of formation, and the observation that ex-posure of the iodo-olefin to hot hydriodic acid gave thedi-iodide in 91% yield. Treatment of the di-iodidewith n-butyl-lithium in pentane at 0 "C furnished, inessentially quantitative yield, a C,,H,, hydrocarbon, m.p.151-153", which originally l5 was believed to be the head-to-tail adamantane dimer (13), exclusively on the groundsthat upon exposure to aluminium trichloride in carbondisulphide it underwent ring opening via disproportion-ation, yielding a single C,,H, hydrocarbon (64%) identi-fied as 1,2'-biadamantyl(l5) by g.1.c.-mass spectrometry;neither 1,l'- nor 2,2'-biadamantyl was detected butother unidentified products included a C,H,, hydro-carbon and two C2,H,, hydrocarbons (ca.30% total).We interpreted the exclusive formation of (15) as evidencefor the head-to-tail arrangement (13) rather than thehead-to-head arrangement (14). More recently, how-ever, Lenoir and Firll' have shown by high resolutiong.1.c. that adamantene dimer is a 2 : 1 mixture of (13)and (14) and from 13C n.m.r. analysis have assigned thecisarrangement to (1 3) and the tram-arrangement to(14). Clearly the isomers (13) and (14) differ in theirmode of reaction with aluminium trichloride in carbondisulphide.1Although the formation of [2 + 21 dimers in the de-halogenation of the di-iodide (1 1) suggests the interven-tion of adamantene, we sought more compelling evidenceand attempted to trap the intermediate in a Diels-Alderreaction.Experiments with a large excess of furan inthe reaction mixture were uniformly unsuccessful; theexpected adduct (16) was not detected, though the dimerswere formed as usual. This failure with furan prompted8 J. R. Wiseman and J. A. Chong, J . Amer. Chem. SOC.,10 R. Keese and E. P. Krebs, Angew. Chem. Internat. Edrt.,11 S . F. Campbell, R. Stephens, and J. C . Tatlow, Tetrahedron,12 M. Farcasiu, D. Farcasiu, R. T. Conlin, iM. Jones, jun., and13 B.L. Adams and P. Kovacic, J . Amer. Chem. SOC., 1973,l4 Cf. D. H. R . Barton, R. E. O'Brien, and S. Sternhell,15 D. Grant, M. A. McKervey, J. J. Rooney, N. G. Samman,16 D. Lenoir, Tetrahedron Letters, 1972, 4049.1 7 D. Lenoir and J. Firl, Annalen, 1974, 1467.18 For a recent discussion and leading references see G. hiaireand F. G. Gault, Bull. SOC. cham. France, 1967, 894.1969, 91, 7775.J. A. Marshall, Accounts Chem. Res., 1969, 2, 22.1971, 10, 262; 1972, 11, 618.1065, 21, 3008.P. von R. Schleyer, J. Amer. Chem. Soc., 1973, 95, 8207.95, 8206.J . Chem. SOC., 1962, 470.andG. Step, J.C.S. Chem. Comm., 1972, 1186236 J.C.S. Perkin IWynberg and his co-workers l9 to suggest that adaman-tene is not a transient intermediate in the dehalogenationof the di-iodide on the grounds that such a highly re-active olefin should surely be capable of interception byfuran; they also suggested that the dimers are formedvia astepwise coupling process.None of this is necessarilytrue. In the first place, adamantene should possess avery large torsion angle about the double bond and, asWoodward and Hoffmann have pointed out, the con-comitant twisting of the orbitals of such an olefin pro-vides to a high degree the precise requirements for athe new substance. In this modification the insolubilityof the di-iodide causes adamantene to be generated atvery low concentration, and dirnerisation is therebysuppressed in favour of capture by butadiene. Much ofthe polymerisation was later traced to the work-up pro-cedure; and when this was modified by adding water tothe reaction mixture at -78 "C and pumping off the ex-cess of butadiene at reduced pressure, the extent ofpolymerisation was greatly reduced.The new substance reacted with bromine and it wasisolated from the reaction mixture as a crystallinesymmetry-allowed dimerisation ; in other words, adaman-tene should be especially prone to dimerisation.Second-ly, furan may have been a poor choice as trapping agent :a Dreiding model of the adduct (16) reveals much tor-sional distortion about the conjoining bonds, suggestingthat the molecule should have a very high strain energy.The latter problem should not arise with an acyclicdiene-adamantene adduct. And indeed when a soh-tion of the di-iodide in ether-pentane containing a largeexcess of butadiene at -78 "C was treated with n-butyl-lithium we obtained, in addition to much polymer, ca.15%of a new substance having a g.1.c. retention time appre-ciably shorter than that sf the dimers which were stillthe major products (ca. 85%) of the reaction.21 A simplemodification of this procedure completely changed theproduct ratio, for when a suspension of the di-iodide inneat butadiene at -78 "C (ether having been used in theprevious experiment to bring it into solution) was simi-larly treated, the product ratio was 85 : 15 in favour ofdibromide (18) (Mf 346/348/350) and regenerated withpotassium iodide in hot dry methanol. The liquidolefin (17) was obtained in 67% yield and its molecularweight by mass spectrometry and n.m.r.spectrum wereconsistent with the butadiene-adamantene adductstructure (17). This assignment could be substantiatedas follows. It is known that on exposure to aluminiumbromide the hydrocarbon (20) (a mixture of stereoiso-mers) undergoes rearrangement accompanied by dis-proportionation, yielding as the major product, 1,Z-tetramethyleneadamantane (19) .22 We repeated thispreparation, isolated (19) by preparative g.1.c. and foundit identical with the product of hydrogenation of thebutadiene-adamantene adduct (17).This work provides compelling evidence for the ex-istence of an unstable intermediate in the dehalogenationof 1,Z-di-iodoadamantane, which behaves as if it werethe olefin adamantene (9).More recently, the observa-tion of adamantene has been claimed by two other groups.l9 A. H. Alberts, J. Strating and H. Wynberg, Tetrahedron2o R. B. Woodward and R. Hoffmann, ' The Conservation of21 W. Burns and M. A. McKervey, J.C.S. Chem. Comm., 1974,2z T. M. Gund, E. Osawa, V. 2. Williams, jun., and P. von H.Letters, 1973, 3047. 868.Orbital Symmetry,' Verlag Chemie, Weinheim, 1970, p. 76. Schleyer, J . Org. Chem., 1974, 39, 29791976 237Gano and Eizenberg 23 found that photochemical Norrishtype I1 fragmentation of either 1- or 2-adamantyl phenyl-acetate in methanol yielded 1-methoxyadamantane,suggesting the intervention of adamantene, and Alberts~t aZ.19 obtained a trace of an adduct with 2,5-dimethyl-furan in the thermolysis of an adamantane-1,2-diyl-bisperoxy-ester.Surprisingly, no dimers were formedin the latter approach.EXPERIMENTALM.p.s were determined for samples sealed in capillarytubes. Unless otherwise stated i.r. spectral data relateto dispersioiis in potassium bromide discs. lH N.m.r. datawere measured a t 60 MHz with tetramethylsilane as internalstandard. Mass spectrometric data were obtained with an,4.E.I. MS 902 instrument and with an A.E.I. MS 30instrument attached to a Pye-Unicam series 104 gas chroma-tograph. Differential thermal analysis (d.t.a.) was per-formed on a Dupont 900 thermal analyser fitted with ad.s.c. cell. G.1.c. refers to analysis on one of the followingcolumns: (A) 2 m Silicone Gum Rubber on Chromosorb W(5% w/w); (B) 20 m Apiezon L capillary. Preparativeg.1.c.was performed on a 10 f t x Q in column of SiliconeGum Rubber on Chromosorb W (30% w/w). Spence typeH alumina and Whatman SG 31 silica gel were used foradsorption chromatography. Light petroleum had b.p.40-60". The drying agent employed was magnesium sul-phate.Protoadanzantan-4-one.-The ketone was prepared in 65 %yield from adamantan- 1-01 according to the publishedprocedure. 24Protoadamantan-4-one Hydrazone ( 10) .-Protoadamantan-4-one (5.0 g) in ethanolic 10% potassium hydroxide (70 ml)was treated with 98% hydrazine (15 ml). The solution washeated under reflux for 1 h, then cooled, poured into coldwater (250 ml), and extracted with ether (4 x 50 ml), Theextract was washed with water, dried, and concentrated,yielding a waxy solid (5.5 g).1.r. and mass spectral analysisindicated that the product was a mixture of the hydrazone(10) and the corresponding azine. Attempts to isolate thehydrazone by crystallisation of the mixture from ethanol,aqueous ethanol, benzene, xylene, or hexane led to completeconversion into the azine. Accordingly, the mixture wasused directly in the following experiment.4-Iodoprotoadaynant-4-ene (12) and lI2-Di-iodoadaynantane(1 1).-The crude hydrazone (5.2 g) was dissolved in drybenzene (95 ml) and dry triethylamine (1 1.1 ml) was added.The solution was stirred under nitrogen a t room temperaturewhile a solution of iodine (10.0 g) in dry benzene (100 ml) wasadded dropwise over 2 h. After nitrogen evolution hadceased the mixture was stirred for a further 2 h, then washedsuccessively with N-hydrochloric acid, water, saturatedaqueous sodium hydrogen carbonate, and water.The solu-tion was dried and concentrated, yielding an oil which wasplaced on a column of alumina (1 kg) . Elution with benzenegave the iodo-oZe$n (12) (2.0 g, 25y0), b.p. 80' a t 0.1 mmHg(Found: C, 46.5; H, 5.15; I, 48.25%; M+, 260.0064.C,,HI,I requires C, 46.15; H, 5.05; I, 48.8% ; M , 260.0064),z (CDCl,) 3.20-3.32 (1 H, d, vinylic) and 7.9-8.6 (12 H, m),m/e 260 (15%, M+), 135 (15), 133 (16), 105 (lo), 91 (loo),79 (16), and 67 (12). Further elution with benzene gave thedi-iodide (11) (1.7 g, 14y0), m.p. 10&-108" (Found: C, 31.3;H, 3.7%; iW+, 387.9186.C1,H1,12 requires C, 30.95; H,3.65; I, 65.4%; M , 387.9188), T (CDCl,) 4.91 (1 H,s) and 7.15-8.5 (13 H, m), wz/e 388 (llY6, hf+), 261 (loo),134 (44), 105 (20), 91 (59), 79 (38), 77 (28), and 67 (19).Further elution with light petroleum-ether (3 : 1) gavesome protoadamantan-4-one.Reaction of the lodo-olefin (12) with Hydriodic Acid.-Amixture of the iodo-olefin (2.4 g) and constant-boiling hydri-odic acid (40 ml) was heated under reflux for 3 h, then cooled,poured into water (100 ml), and extracted with chloroform(4 x 50 ml). The extract was washed with water, aqueoussodium thiosulphate, and water, and dried. Removal of thesolvent gave the di-iodide (1 1) (3.4 g, 91 yo).Reaction of Protoadaunantan-4-one with Phosphorus Tri-iodide.-To a solution of iodine (450 g) in chloroform (2.4 1)was added red phosphorus (53 g) with stirring.The mix-ture was stirred a t room temperature for l& h, a solution ofprotoadamantan-4-one (1 9.2 g) in chloroform (100 ml), wasadded, and the resulting mixture was heated to 55 "C for 5 h.After cooling, the solution was poured into water (3 1) andfiltered. The organic layer and dichloromethane extracts(3 x 150 ml) of the aqueous layer were combined, washedwith aqueous sodium thiosulphate and water, then dried.Removal of the solvent gave a pale yellow solid (47.2 g)which was placed on a column of silica gel ( 1.3 kg) . Elutionwith light petroleum-ether (9 : 1) gave the di-iodide (11)(45.4 g, 92y0), identified by direct comparison with thematerial described above.Reaction of the Di-iodide (1 1) with n-Butyl-lithium.-n-Butyl-lithium [22 ml of a 23.1% (w/w) solution in hexane]was added dropwise with stirring to a solution of the di-iodide(11) (3.6 g) in pentane under nitrogen a t 0 "C.The solutionwas stirred for a further 3 h and water (150 ml) was added.The organic layer and ethereal extracts (3 x 50 ml) of theaqueous layer were combined, washed with water, and dried.Removal of the solvent gave a semi-solid which was placedon a column of alumina. Elution with pentane gave theadamantene dimers (13) and (14) (1.3 g, loooh). Crystallisa-tion from hexane-methanol gave prisms, m.p. 151-152'(Found: C, 89.4; H, 10.75%; M+, 268.2191. Calc. forCzoHzs: C, 89.5; H, 10.5%; M , 268.2191), T (CDC1,) 7.8-8.8 (complex), m/e 268 (100% M f ) , 225 (33), 211 (13), 135 (26),91 (28), 79 (24), 77 (12), and 67 (12).G.1.c. analysis oncolumns (A) and (B) showed a single peak for dimers (13)and (14).Reaction of the Dinzers (13) and (14) with AluminiumChloride.-Aluminium chloride (0.2 g) was slurried in drycarbon disulphide (100 ml) at room temperature and asolution of the dimers (0.2 g) in carbon disulphide (100 ml)was added. The mixture was stirred for 1 h then pouredonto ice-water. The organic layer and carbon disulphideextracts of the aqueous layer (2 x 50 ml) were combined,washed with water, and dried. Removal of the solventyielded a semi-crystalline solid (0.19 g, 93%) which wasshown by g.1.c. analysis on columns (A) and (B) at 200 "C tobe a mixture of 1,2'-biadamantyl (15) (64%), an unidenti-fied hydrocarbon of molecular weight 266, and two unidenti-fied hydrocarbons of molecular weight 268 (total 36%).Co-injection with authentic samples of 1, 1'-biadamantyl and2,2'-biadamantyl established that these compounds were notD.t.a. showed only one endotherm a t 152 "C.23 J. E. Gano and L. Eisenberg, J . Amev. Ckem. SOC., 1973, 95, ** W. H W. Lunn, J . Cizem. SOC. ( C ) , 1970, 2124; R. 34. Black972. and G. B. Gill, Chern. Comm., 1970, 972238 J.C.S. Perkin Iamong the products. Preparative g.1.c. of the crude productgave 1,2'-biadamantyl (15) as needles, m.p. 263 (lit.,25 266-268') , identical with a sample prepared by Wurtz couplingbetween 1- and 2-bromoadamantane with sodium in toluene.Reaction of the Di-iodide (1 1) with n-Butyl-lithium in LiquidButadiene.-To a suspension of the di-iodide (7.0 g) in drybutadiene (250 ml) a t -78 "C under nitrogen was addeddropwise with stirring n-butyl-lithium in hexane (23.1%wlw; 50 ml), and the mixture was then stirred a t -78 "Cfor 15 h.Water (50 ml) was added dropwise and the mix-ture was allowed to warm to -40 "C. The flask was con-nected to a water aspirator and the excess of butadiene wasremoved by pumping at - 40 "C. Ether (150 ml) and water(250 ml) were added to the residue and the ether layer andethereal extracts of the aqueous layer (3 x 100 ml) werecombined, washed with water, and dried. Removal of thesolvent gave an oil (10.8 g). G.1.c. analysis on column (A)at 215 "C revealed the presence of dimers and a new shorter-retention-time component in the ratio ca.1 : 15. Combinedg.1.c.-mass spectrometric analysis established that the majorcomponent had M+ 188, indicative of an adamantene-buta-diene adduct.Isolation of the Adavvtantene-Butadiene Adduct (17) .-Thecrude product from the previous experiment was taken up inchloroform and 10% bromine in chloroform was added drop-wise with stirring until the red colour persisted. The solu-tion was washed with aqueous sodium disulphite and water,and dried. Removal of the solvent gave a viscous oil (19.8g), g.1.c. analysis of which on column (A) a t 215, "C revealedthe disappearance of the adduct (17). The product wasplaced on a column of silica gel (700 g) .Elution with lightpetroleum gave the dimers (13) and (14) (0.3 g , 13%).Further elution with light petroleum-ether (99 : 1) gave thedibromide (18) (9.3 g) as an oil which slowly crystallised.A sampIe crystsllised from hexane had m.p. 92-94' (Found :C, 48.25; €3, 5.85; Br, 45.9. Calc. for C,,H,,Br,: C, 48.3;H, 5.8; Br, 45.9%), ~t (CDC1,) 7.5-8.8 (12 H,m), and 5.4-6.0 (2 H, m), M+ 350/348/346.The crude product was taken up in dry methanol (300 ml)containing potassium iodide (60 g) and the solution washeated under reflux for 48 h. The cooled solution was di-luted with water (1.5 1) and extracted with dichloromethane(5 x 200 ml). The extracts were washed with aqueoussodium thiosulphate and water, then dried. Removal of thesolvent gave an oil, g.1.c. analysis of which on column (A)a t 215 "C revealed the absence of the dibromide and thepresence of a single component having a retention timeidentical with that of the adduct (17). The product waspassed through a column of alumina (300 g) in light petro-leum, yielding tricyclo[7.3. l. 17~11.01~g]tetradec-3-ene as anoil (2.28 g, 67%). A sample obtained by preparative g.1.c.a t 215 "C (Found: C, 89.45; H, 10.7%; M+, 188.1560.C1,H,, requires C, 89.3; €3, 10.7% ; M , 188.1569) showed T(CDCl,) 9.1-7.8 (18 H) and 4.35 (2 H), rn[e 188 (loo%, M+)131 (38), 91 (22), 79 (20), and 67 (10).Hydrogenation of the Adduct (17).-The olefin (17) (0.184g) in methanol (80 ml) containing Adams' catalyst (0.25 g)was exposed to hydrogen a t 1 atm and room temperature,When uptake had ceased the catalyst was removed and thefiltrate concentrated, yielding the hydrocarbon (19) (0.186 g)as an oil. A sample obtained by preparative g.1.c. at 215 "Cwas identical (g.1.c.; mass and n.m.r. spectra) with asample prepared by aluminium bromide-catalysed rearrange-ment-disproportionation of the hydrogenated norborna-diene dimer (20) .22We thank the Northern Ireland Department of Educationfor a postgraduate award (to W. B.).25 H. J. Storesund, Tetrahedrofi Letters, 1971, 4353. [5/1136 Received, 10th June, 1975
ISSN:1472-7781
DOI:10.1039/P19760000234
出版商:RSC
年代:1976
数据来源: RSC