53 D. W. Jones stereoselectivity (e.g. exo-endo in the Diels-Alder reaction). It has however been pointed out that in certain cases interaction involving the superjacent orbital of one partner may be imp~rtant.~" In the cyclopentadiene-fulvene addition the addition mode predicted by considering both HOMO-LUMO interactions is that shown in (1)' but the observed mode (2) is correctly predicted if interaction between the cyclopentadiene HOMO and the next-LUMO of fulvene is also considered. The FMO method has been extended to include a rationale for the possible role of catalysts in pericyclic processes as well as insertion reactions and olefin metathe~is.~~ It will be of interest to see if the FMO method can be applied in a quantitative way to reactivity problems in pericyclic reactions involvirg a-bonds.Considerable attention has been devoted to so-called forbidden processes. For the thermal racemization and cis-truns-isomerization of cyclopropanes earlier studies with substituted cyclopropanes cast doubt on Hoffmann's suggestion 4a of a 0,O-trimethylene biradical intermediate (3). Support for (3) is now provided4b by a study using optically active trans-[ 1,2-2H2Jcyclopropane (4)(a365+0.168"!). This = undergoes trans-cis-isomerization 1.07 times faster than racemization in accord with a process involving double methylene rotation e.g. rotation at both C-1and C-2 converts (4) into its enantiomer (3,and it is consistent with the intermediacy of (3). H DD Since an alternative explanation of the results requires a blend of a random biradical intermediate and/or a single methylene rotation process with a large fraction of a triple methylene rotation process the authors4" believe that the double methylene rotation operates exclusively.The same conclusion has been drawn from a study using 1-phenyl[2-2H]cyclopropane.4" The continuous-biradical hypothesis4d is the functional equivalent of a single methylene rotation mechanism and so is inconsis- tent with the present results. As a consequence of through-bond coupling involving the C-1 and C-2 orbitals of (3) and the C-3 methylene orbitals the HOMO of the 0,O-biradical should have the symmetry shown. It should therefore undergo conrotatory rather than disrotatory closure. However no information on this point is currently available.In the related opening of methylenecyclopropanes e.g (6) an intermediate trimethylenemethane could be an orthogonal methylene-ally1 biradical (a)R. Hoffmann J. Amer. Chem. Soc. 1968,90,1475;(6)J. A. Berson and L. D. Pederson ibid. 1975 97,238; (c)J. A. Berson L. D. Pederson and B. K. Carpenter ibid.,p. 240; (d)W. von E. Doering and K. Sachdev ibid. 1974,% 1168. Part (i) Orbital Symmetry Correlations and Pericyclic Reactions (7) or a planar conjugated species (8). Since (6) is thermally converted into its diastereoisomer (9) with partial retention of optical a~tivity,~" the symmetrical (4) (7) (8) (9) orthogonal trimethylenemethane (7) cannot be the only intermediate. The unsym- metrical planar trimethylenemethane (8) must also be involved but to a smaller extent.The energy difference between (7) and (8) thus appears to be smaller than suggested by the calculations of Dewar and Salem5* and more in line with the results of others." Although the stereochemistry of the opening of bicyclobutanes to butadienes e.g. (10)-+ (1l) is nicely rationalized6" in terms of a concerted u2s+ u2a process H-H Me (10) (1 1) biradical intermediates have been suggested. Thus (12)rearranges 2700 times faster than the compound in which the double bond is saturated,6b in accord with allylic stabilization of one radical site in the postulated intermediate (13) (Scheme 1). Scheme 1 MIND0/3 calculations locate an intermediate biradical (14) in the opening of bicyclobutane to butadiene.6c The radical lobes in (14)are coupled by through-space overlap so that (14) is more stable than the classical non-interacting biradical.This serves to explain the stereoselectivity of the ring-opening since an interesting biradical retains a memory of the stereoisomer from which it is formed. Presumably the conversion of benzvalene into benzene cannot be a u2s+ a2aprocess leading to trans-benzene. It is surprising therefore that in contrast to the simple bicyclo- butane to butadiene conversion the opening of benzvalene is predicted6d to be concerted proceeding via a very unsymmetrical TS in which one bicyclobutane bond (a) W. R. Roth and G. Wegener Angew. Chem. Intentat. Edn. 1975,14,758;(b)M. J. S. Dewar and J. W. Walton J. Amer. Chem.Soc. 1971,93,3081; W. J. Hehre L. Salem and M. R. Willcott ibid. 1974 96,4328;(c)D. R. Yarkonyand H. F. Schaefer tert. ibid. p. 3754; W. T. Borden ibid. 1975,97,2906. (a) R. B. Woodward and R. Hoffmann 'The Conservation of Orbital Symmetry' Academic Press 1970 p. 77; (b)M. Christl U. Heinmann and W. Kristof J. Amer. Chem. SOC.,1975,97 2299; (c) M. J. S. Dewar and S. Kirschner ibid. p. 2931; (d)ibid. p. 2931. D. W.Jones [‘a’ in (15)] is greatly weakened whilst the other [‘b’ in (15)J is almost intact. Participation of the double-bond and greater stretching of bond ‘a’ than bond ‘b’ may 6 secure a prevailing aromaticity for the TS.6d The forbidden disrotatory opening of bicyclo[2,1 ’OIpent-2-ene to cyclopentadiene is predicted7“ to proceed via a sym- metrical TS as a result of geometrical constraints.The 2-methyl derivative (16) rearranges to a mixture of (17) and (18). This could be due to a secondary (16) (17) (18) rearrangement of (18) to (17) possibly by a ‘hot-molecule’ process since conversion of (16) into (18) is strongly ex other mi^.^' An alternative pathway7d involving cr2s+u2a addition of either the C-1-C-2 and C-4-C-5 or the C-1-C-5 and C- 3-C-4 bonds of (16) is predicted7” to have an activation energy >>250kJ mol-’ and is therefore much less likely. A similar calculation of the TS for the opening of Dewarbenzenes to benzenes7’ allows the suggestion that the TS veers towards the zwitterion (19); this accounts for the more rapid ring opening of (20) compared to Dewarbenzene itself.(19) (20) MIND0/3 Calculations suggest that the reaction of singlet oxygen with olefins is non-concerted proceeding via peroxirans (21) in the case of simple olefins and zwitterions (22) for electron-rich olefins (X = OR or NR2).8a Whilst the peroxirans can rearrange to ene products easily conversion into dioxetans is more difficult. On the other hand zwitterions can readily close to either dioxetans or peroxirans. Thus the best target for detection or trapping would be a peroxiran that is incapable of forming an ene product. Further calculations8’ give support to a proposed mechan- ism for oxiran formation in the reaction of hindered olefins with singlet oxygen (a)M. J. S. Dewar and S. Kirschner,J.C.S. Chem. Comrn. 1975,461; (b) ibid.p. 463; (c)M. C. Flowers and H. M.Frey J. Amer. Chem.Soc. 1972,94,8636;J. I. Brauman W. E. Farneth and M. B. D’Amore ibid. 1973,95 5043; (d)J. E. Baldwin and G. D. Andrews ibid. 1972,94 1775. ((I) M. J. S. Dewar and W. Thiel J. Amer. Chem. Soc. 1975,97,3978; (6)M. J. S. Dewar A. C. Griffin W. Thiel and I. J. Turchi ibid. p. 4439. Part (i) Orbital Symmetry Correlations and Peric yclic Reactions involving ‘reduction’ of a peroxiran intermediate (21) with singlet oxygeq to give ozone and the oxiran. Alternative formulations of the Woodward-Hoff mann rules continue to appear.’ In a very clear exposition Day provides’“ formal proof of the equivalence of the aromaticity and generalized Woodward-Hoffmann rules as well as compact new rules equivalent to the Dewar-Evans-Zimmerman rules.These are particularly easy to apply e.g. analysis of the Diels-Alder addition (23) simply involves choosing an (23) arbitrary connectivity cycle (--). The reaction is thermally allowed because the connectivity cycle crosses no atomic orbital modes (N= 0) and three electron pairs are involved (M= 3) so that M +N = 3 (odd number thermally allowed). Whilst this analysis is equivalent to Zimmerman’s reactivity index,96 it may have pedagogi- cal advantage. It has been notedgd that multicentre reactions e.g. the opening of prismane to benzene which cannot be directly analysed by the generalized Woodward-Hoff mann rules can be so analysed if they are first (artificially) divided into steps e.g. the prismane into benzene conversion is analysed as proceeding via Dewarbenzene and shown to be forbidden.The thermal cyclization of unsaturated carbonyl compounds has been reviewed.26 It involves enolization and intramolecule ene-reaction (Scheme 2). For reactions Scheme 2 leading to five-membered rings the syn-TS (24) giving a cis-relationship between the carbonyl group and the newly formed methyl group is preferred. A neat synthesis of chiral acetic acid” from the ether (25) involves an intramolecular ene-reaction (25; 9 (a)A. C. Day J. Amer. Chem. Suc. 1975,97,2431;(b)H. E. Zimmerman and H. Iwamura ibid. 1970 92,2015; (c) A. Rassat TetrahedronLetters 1975,4081;(d)P. Wieland Tetrahedron,1975,31,1641. lo C. A. Townsend T. Scholl and D. Arigoni J.C.S.am. Comm.1975,921. D. W. Jones arrows) to give (26) which undergoes allylic ether pyrolysis (retro-ene reaction) to (27); Kuhn-Roth oxidation then gives CHDTC02H,of (R)-configuration. 2 Cycloaddition and Cheletropic Reactions In the addition of difluorocarbene to norbornadiene (28; R =H) the normal non- linear cheletropic 1,2-addition to one double bond competes with apparent non- linear addition to the homoconjugated diene system giving (29; X = F; R =H). Steric inhibition of exo-attack on (28; R =Me) leads to predominant 1,4-addition of RR (29) difluorocarbene as well as the hitherto unobserved 1,4-addition of dichlorocar-bene giving (29; X =C1; R =Me).'' Photochemical opening of the azirine (30)gives Me n /Na + PhCEN-C -/ ?Me < ..\ Me Me (30) (31) (32) (33) the 1,3-dipole (31) which can rehybridize to the carbene (32); this then undergoes an unprecedented unturu-addition to the double bond giving (33).12 Loss of sulphur dioxide from (34) and its trans-isomer give respectively truns,trans-hexa-2,4-diene MeeMe (34) and the cis,trans-isomer with >99.9% stereoselectivity. Extrusion of sulphur dioxide from (35) and its trans-isomer is also highly stereoselective but here the polyene formed indicates unturu-elimination of SO2 e.g. (35)gives cis,cis,truns- H Me 02 0 (35) (36) C. W.Jefford W. D. Graham and U. Burger TetrahedronLetters 1975,4717; q.R.A. Moss and C. B. Malion J. Amer. Chem. Soc. 1975,97 344. l2 A. Padwa and P. H. J. Carlsen J. Amer. Chem.Soc. 1975,97,3862. Part (i) Orbital Symmetry Correlations and Pericyclic Reactions octa-2,4,6-triene with ca. 98% stereoselectivity despite the congested environment of one methyl group the antaru-TS arrangement (36).13 ~ In the photochemical (2+2)~ addition of singlet excited trans-stilbene to cis- piperylene there is preference for addition to the more substituted double bond of the diene in accord with FMO consideration~.'~ In an interesting application of 'photochemistry without light' thermolysis of dioxetan (37) gave a mixture of (38) (37) (38) (39) and its known photo-product (39) in accord with cleavage of (37)to triplet excited (38) which then rearranges in part to (39).15The w2s+ T2aaddition of ketens to olefins continues to attract attention.The addition of diphenylketen to cis-ethyl propenyl ether is ca. lo2 times faster than addition to the trans-isomer. This is consistent with less steric hindrance in orientation complex (40) than in (41). In 0 =$ :H -1-'&;/ OEt Ph Ph 0 EtO\ -+. 11 /Me c= =$ H/ \H .a'\-Ph Ph (40) (41) contrast the cis- and truns-ethers react at about the same rate with tetra- cyanoethylene via dipolar intermediates. The TS's for the keten additions are thought to resemble the orientation complexes of the reactants closely with little bond formation. The large negative entropy of activation of the ethyl cis-propenyl ether addition (AS' = -45 e.u.) constitutes 67% of the activation energy (AH' = 7.0 kcal mo1-').16a Some of the adducts produced in this study lose alcohol if they are left to stand over alumina to give cyclobutenones e.g.(42),'6b which undergoes electrocyclic ring-opening (42; arrows) on heating or irradiation. 16' Whilst acyclic (partially) optically active 1,3-dialkylallenes add ketens to give mainly 2-2-alkylidenecyclobutanones(43) having little or no optical activity the l3 W.L. Mock J. Amer. Chem. Soc.,1975,97,3666 3673. l4 F. D. Lewis C. E. Hoyle and D. E. Johnson J. Amer. Chem. Soc. 1975,97,3267. H. E. Zimmerman and G. E. Keck J. Arner. Chem. Soc. 1975,97,3527. (a)R. Huisgen and H. Mayr TetrahedronLetters 1975,2965,2969;(b) H. Mayrand R. Huisgen Angew. Gem. Internat. Edn. 1975,14499;(c) H. Mayr ibid.,p. 500. D. W.Jones (43) (44) (45) addition of optically active 1,3-diphenylallene to t-butylcyanoketen is now shown17a to give the E-isomers (44)and (49,both of which are optically active and therefore formed to some extent via chiral TS’s.The difference between diaryl- and dialkyl- allenes may be rationalized using the usual orientation complex for w2s+ w2a addition. Thus (46; R = Ph) leads to (44)’whilst the more hindered complex with But and CN groups interposed leads to the minor adduct (45). When R = Ph positive charge accumulating at C-1 is stabilized by the phenyl group so that rotation ‘A’ leading to allylic stabilization of this charge can be delayed until far along the reaction path. When R = Alk rotation ‘A’ may run ahead of rotation ‘B’ so that zwitterionic intermediate (47),leading to 2-adducts is formed.Related dipolar 0 (46) (47) intermediates have been invoked to explain the products obtained from cyclo- propenes and ketens. Whilst the concerted ene-reaction of bis-trifluoromethylketen with 1-methylcyclopropene is favoured in the gas phase a dipolar addition becomes competitive in sufficiently polar solvents (Scheme 3).17’ The synthetic potential of 0 -44e &;;3 -(cF4y -::F)-0 CF3 Reagents i (CF,),CCO Scheme 3 additions involving ketens and allenes is indicated by the addition of dimethylketen to the cyclic allene (48) giving (49) which is strongly reminiscent of caryophyl-Iene.17‘ (48) 17 (a) H. A. Bampfield P. R. Brook and W. S. McDonald J.C.S.Gem. Comm. 1975; 132; (b)Dt H. Aue D.F. Shellhamer and G. S. Helwig ibid. pp. 603,604; (c) M. Bertrand R. Maurin J.-L. Gras and G. Gil Terruhedron 1975,31 849; H. Bertrand J.-L.Gras and J. Gore ibid.,p. 857. Part (i) Orbital Symmetry Correlations and Pericyclic Reactions Two fine studies of Diels-Alder diene dimerization differ in their conclusions. Deuterium-labelled piperylene (50) affords a mixture of the endo-(51) and exo-adduct (52) indicating suprafacial addition to both the diene and olefin systems.'8" Me Me [Me Me /Me On the contrary the analogous dimerization of cis,cis- 1,4-dideuteriobutadiene apparently involves ca. 90% supra and 10% antara-addition to the double bond. Whilst the mode of addition to the diene is not established it is suggested that the 10% antarafacial addition may represent a concerted forbidden 7r4s +7r2uprocess the major process being the more familiar 7r4s3-r2s addition."* The intramolecular Diels-Alder reaction has been reviewed and numerous further examples have been provided.196-d The amide (53) undergoes preferred intramolecular addition uia the endo-arrangement shown which best preserves conjugation between the amide group and the diene system. On the other hand amide (54) in which the whole of the amide system is within the connecting bridge (53) (54) prefers the exo-arrangement shown which in this case best maintains analogous ~verlap.''~A cautionary note is sounded by the observation that (55) failed to give the expected product but instead isomerized to (56) which then gave (57).19' Ncm H,CH ,H Me Hcy2CH2 C0,Me CH2I* C0,Me .I Hr=iH (55) (56) (57) (a)J. A. Berson and R. Malherbe J. Amer. Chem. SOC.,1975,97 5910; (6)L. M. Stephenson R. V. Gernrner and S. Current ibid. p. 5909. ly (a)R. G. Carlson Ann. Reports Medicin. Chem. 1974,9 270; (b)W. Oppolzer W. Frostl and H. P. Weber Helv. Chim. Acta 1975,58,590,593;(c)R. F. Borch A. J. Evans and J. J. Wade J. Amer. Chern. Soc. 1975,97,6282;(d)P. G. Sarnrnes and R. A. Watt J.C.S. Chem. Comm. 1975,502;M. T. Cox ibid. p. 903; E. J. Corey and M. Petrzilka Tetrahedron Letters 1975 2537; J. J. S. Bajorek and J. K. Sutherland J.C.S. Perkin I 1975 1559; P. Yates and D. J. Bichan Canad. J. Chem. 1975 53 2045 2054; J. Auerbach and S. M. Weinreb J. Org. Chem. 1975,22,3311;R. H. Martin J. Jespers and N.De Fay Helv. Chirn. Acra 1975,58,776;Y. Nakamura W. E. Oberhansli R. Hollenstein J. Zsindely and H. Schrnid ibid.. p. 1949. D. W. Jones The great reactivity of both cyclobutadiene and cyclopentadienone towards dimerization follows from the small HOMO-LUMO energy gap predicted for these molecules. Frontier MO interactions will be particularly strong in the Diels-Alder dimerization of these species. A related effect is noted2' for the pyridinium betaines (58) which undergo (4+ 2)~dimerization to (59) more \rapidly as the electron- withdrawing effect of the group R increases. The HOMO coefficient at the nitrogen 0-6 no-\+ N R (58) (59) atom of such betaines is small whilst the corresponding LUMO coefficient is large.3u Electron withdrawal by R consequently reduces the HOMO-LUMO separation and promotes dimerization.The cycloaddition of electron-poor olefins to C-2 and C-6 of (58)is mainly determined by the betaine HOMO-dipolarophile LUMO interaction. Replacement of C-2 by a nitrogen atom lowers the HOMO energy and the resulting 3-oxidopyridazinium betaine shows reduced reactivity. The betaines add as 47r-components to fulvenes (60; arrows). Approach of the reactants as in (61) is LUMO 0-% *o 2%oMo lo HOMO R N R (60) (61) (62) favoured over the alternative (62) by the secondary interactions that are indicated between the betaine HOMO and fulvene LUM0.20 The triazine (63) adds dimethyl acetylenedicarboxylate at the indicated posi- tions,21 and so behaves as a 1,ll- rather than a 1,3-dipole.This accords with control +& -0.41t f0.41 by the dipole HOMO which has the Huckel coefficients indicated on (63) at the competing sites.3a 2o N. Dennis B. Ibrahim and A. R. Katritzky J.C.S. Chem. Cornm. 1975,425; see also N. Dennis A. R. Katritzky and M. Ramaiah J.C.S. Perkin I 1975 1506. 2' S. F. Gait M. J. Rance C. W. Rees R. W. Stephenson and R. C. Storr J.C.S. Perkin I 1975 556. Part (i) Orbital Symmetry Correlations and Pericyclic Reactions Secondary orbital interactions can influence the regioselectivity as well as the exo- endo -selectivity of Diels-Alder reactions.22a A pretty example2" is provided by the addition of (64) to (65) which would yield regio-isomer (66) under the control of the coefficients at the termini of the diene system but regioisomer (67)if secondary MeoPh MeON02Ph Ph NO2 (64) (65) (66) (67) interaction between the nitro-group and C-2 or C-3 of the diene is dominant.The uncatalysed reaction gives the exo- and endo-forms of (66) because the secondary interactions are weak. In the catalysed reaction the secondary interactions are expected to be stronger,22C and the product is exclusively the endo-form of regio- isomer (67). The activation volume for the (6 +4)m-addition of tropone to cyclopentadiene is only -7S* 1cm3mol-' compared to values of cu. -35 cm3 mol-' for typical Diels-Alder reactions.23 The small partial molar volume of tropone associated with its dipolar character and consequent electrostriction accounts for this result in terms of a concerted reaction.As for Diels-Alder additions in which secondary interac- tions are possible the TS for this (6 +4)~-addition is smaller than the final state. It is pointed out that a secondary interaction is possible in the exo-TS (68). The 10&kMO secondary interaction shown in (69) has been proposed to explain the syn-selectivity observed for low-temperature addition of amino-nitrenes to diene~.~~ 3 Sigmatropic Reactions In the Stevens rearrangement of optically active (70)to (71)and (72) the degree of retention of configuration in the migrating a-phenylethyl group depends markedly Ph H Ph H Ph Me Me )-Me 'rH Me2yCH2COPh Me,NCHCOPh Me,NCHCOPh (70) (71) (72) (a) P. V. Alston R. M. Ottenbrite and D.D. Shillady J.Org. Chern.,1973,38,4075;(b)P.V.Alston and R. M. Ottenbrite ibid. 1975,40,1111;(c) K.N.Houk and R. W. Strozier J. Amer. Chern. Soc. 1973,95 22 4094. 23 W. J. le Noble and B. A. Ojosipe J. Amer. Chem. Soc. 1975,97,5939. 24 R.S.Atkinson and J. R. Malpass J.C.S. Chem. Comrn. 1975,555;see too R.S. Atkinson and R. Martin ibid. 1974 386. D. W.Jones on reaction conditions.2s" With sodium hydroxide-water at 0 "Cthere is 99f1Oh net retention of configuration but with sodium methoxide-methanol at 55 "C this figure drops to 56*2% and CIDNP and products of crossed radical coupling are observed. The stereoselectivity and intramolecularity of the reaction both decrease as the viscosity of the solvent decreases in agreement with a caged radical-pair intermediate (73) in which the rate of radical coupling is very fast k ca.lo1' s-'). [ Ph-?Ce] Me,NCHCOPh (73) Some contribution from a concerted process is not excluded by these results or those of a quantitative CIDNP study where the observed enhancement factors were 25% or less. than expected for a reaction proceeding. only by a free-radical path.'" However rapid recombination of the radicals may preclude evolution of the spin wavefunction; it is difficult to distinguish between a concerted pathway and the interaction of radical pairs which do not separate or even The photochemical ring expansion of bicyclo[2,2,l]heptan-2-ones to carbenes e.g. (74) to (75) has been reviewed and structural factors favouring the process have (74) (75) (76) been delineated.26" Unlike the corresponding ring-expansion of cyclobutanones,26b these reactions are believed to involve biradical intermediates e.g.(76) from (74). Thermal 1,3-sigmatropic shifts have been the subject of theoretical ~al~ulation,~~~-~ Epi~tis~~" pointing out that sub-jacent orbital control is important for non-polar but not polar 1,3-shifts and DewarZ7' predicting a biradicaloid forbidden pathway for the vinylcyclopropane to cyclopentene rearrangement. The photochemical 1,3-shift of the optically inactive diastereoisomer (77) affords a mixture of (78) and H H OMe (77) (78) (79) 2s (a)W. D. Ollis M. Rey I. 0.Sutherland and G. L. Closs J.C.S. Chem. Comm. 1975 543; (6) U. H. Dolling G. L. Closs A.H. Cohen and W. D. Ollis ibid.,p. 545. 26 (a)P. Yates and J. C. L. Tam J.C.S. Chem. Comm. 1975,737,739; (b)W.-D. Stohrer P. Jacobs K. H. Kaiser G. Wiech and G. Quinkert Forrschr. Chem. Forsch 1974,46 181. 27 (a)N. D. Epiotis R. L. Yates and F. Bernardi,J. Amer. Chem.Soc. 1975,97,4198;(b)W. W. Schoeller ibid. p. 1978; Chem.Ber. 1975 108 1285; A. Gavezzotti and H. Simonetta Tetrahedron 1975 31 1611; (c) M. J. S. Dewar G.J. Fonken S. Kirschner and D. E. Minter J. Amer. Chem. Sm. 1975,97 6750; (d)J. Gloor and K. Schaffner ibid. p. 4776; (e)G. R. Krow and J. Reilly ibid.,p. 3837; (f) J. J. Eisch and J. E. Galle ibid. p. 4436. Part (i) Orbital Symmetry Correlations and Pericyclic Reactions (79) and is thought to proceed via a geminate radical pair.27d Thermal rearrange- ment of (80; R =H) to (81;R =H) is shown by deuterium labelling to involve retention of configuration at the migrating methylene group.Corresponding rear- rangement of (80; R =CH,Ph) is slow. It is believed that steric effects associated with the group R destabilize those conformations e.g. (82) in which the nitrogen (82) lone-pair can best overlap with the migrating ~r-bond.~~~ The borepin (85),isoelec-tronic with the tropylium cation is formed by heating (83); this could involve a PBPh mph6 /I D Phr P'h Ph Ph6 Ph (83) (84) (85) 1,3-boron shift to (84)followed by electrocyclic ring-~pening.~'~ Thermolysis of (86) gives the exo-isomer (86; Ac and C-5-Me interposed) and the cyclopentene (87). The two products are not formed from a common intermediate and only endo- isomer (86) gives (87).The large negative entropy of activation for the process giving (87) suggests a concerted rearrangement of novel type.28 (84) Functionalization a to a sulphur atom (88)-+(89) required in a biologically patterned synthesis of penicillins was achieved using benzoyl peroxide and may involve a 2,3-sigmatropic shift in the intermediate (90).29a A scheme involving repeated 2,3-sigmatropic shifts has been proposed for the 'growing' of macrocyclic rings.29b x OCOPh I S-CH S-CH "'00 b-r S-CH /\ // /+ \ (88) (89) (90) 28 J. P. Grosclaude H. U. Gonzenbach J. C. Perlberger and K. Schaffner J. Amer. Chem. Soc. 1975,97 4147. 29 (a)J. E. Baldwin,A. Au M. Christie,S.B. Haber and D. Hesson J. Amer. Chem.Soc. 1975,97,5957; (b)E. Vedejs and J. P. Hagen ibid. p. 6878. 66 D. W. Jones Cross-over experiments confirm that the low-temperature Cope rearrangement (<240 "C) of simple hexa-1,5-dienes is concerted. The rearrangement at higher temperatures (259-295 "C) is also concerted but at higher temperatures still (360-390 "C) rearrangement is by a dissociation-recombination mechanism.3oo The syn-anti-isomerization of bicyclononatrienes e.g (91)*(92) is believed to involve Cope rearrangement to intermediate cyclobutenes (93) and (94) which open to the Z,E,Z,Z-nonatetraenes (95) (Scheme 4).30bSince in (95) reverse cyclization to (93) is just as likely as forward-going electrocyclization to (94) the rate of conversion of (91) into (92) should be half the limiting rate at which (95) is trapped by tetra~yanoethylene.~"Thisrate relationship has now been established and provides strong evidence for this mechanism (Scheme 4),30bas opposed to an alternative It Reagents i (NC)&=C(CN) Scheme 4 recently Cope rearrangement of the type (96) -+(97) is accelerated by a factor of 1010-10'7 in the related potassium alkoxides and a further acceleration is observed in the presence of 18-cro~n-6.~~~ Whilst Claisen rearrangement of the geometrical isomers of propenylbutenyl ether (98) proceeds mainly via chair-like 30 (a) D.C. Wigfield and K. Taymaz Tetrahedron Letters 1975,3121;(b)C. P. Lewis and M. Brookhart J. Amer. Chern. Soc. 1975,97,651;(c) G. Boche H. Weber and J.Benz Angew. Chem. Internat. Edn. 1974,13,207;(d)J. M. Brown and M. M. Ogilvy,J. Arner. Chem. Soc. 1974,96,292;(e)D.A. Evans and A. M. Golob ibid. 1975,97,4765. Part (i) Orbital Symmetry Correlations and Pericyclic Reactions TS'S,~'~ * the related rearrangement (99; arrows) adopts a boat-like TS.31 Although cation (100) undergoes the indicated 3,4-sigmatropy the corresponding rearrange- ment in cyclohexadienyl cations is much less imp~rtant.~'" MeCH=CHOCH,CH=CHMe BzO (98) Ph D 0-J (99) (100) There is growing appreciation of the structural factors influencing migratory aptitude in the 1,5-sigmatropic shift.32 Two pathways operate in the thermal conversion of cyclohexadienes (101) into meta -substituted toluenes (102) (Scheme 5).32a As judged by the incorporation of 13Clabel (*C) from the methyl group of 40' H PH2 -Me Scheme 5 (101) into the aromatic ring of the product path (a)is the major one for R =CHO COMe and CO,Me whilst path (6)is followed for R = Ph.This is taken to indicate that the migratory aptitude of the phenyl group is very much less than that of the methoxycarbonyl group. Very rapid 1,Sformy1 migration is observed for the 1-formylindene (103;Y = H) which racemizes at 50 "C by rearrangement to the isoindene (104; Y =H). Since racemization of (105) is twice as fast as its conversion into (103; Y =CDO) each act of racemization involves formation of (104; Y = CDO) in which there is a 50% chance of CHO-CDO exchange.326 Migration of a 31 (a)P. Vittorelli H.-J.Hansen and H. Schmid Heiv. Chim Acta 1975,58 1293; (b)B. Lythgoe and D. A. Metcalfe Tetrahedronhtters 1975,2447;(c)P. Vittorelli J. P. Katalinic G. Mukherjee-Muller H.-J. Hansen and H. Schmid Helv. Chim. Actu. 1975 58 1379. 32 (a)P.Schiess and R. Dinkel TetrahedronLetters 1975,2303;(b)D.J. Field D. W. Jones and G. Kneen J.C.S. Chem. Comm. 1975,754;(c)L.A. Paquette and M. J. Carmody J. Amer. Chem. Soc. 1975,97 5841;(d)J. Backes R. W. Hoffmann and F. W. Steuber Angew. Chem. Znfenurt. Edn. 1975,14,553;(e) A. P.ter Borg H. Kloosterziel and Y. L. Westphal Rec. Trav.chim. 1963,82,717;(e)J. J. McCullough and A. J. Yarwood J.C.S. Chem. Comm. 1975,485;(f)M. R.Willcott and I. M. Rathburn J. Amer. Chem. Soc. 1974,96,938;(g) W.R. Dolbier L. Mdullagh D. Rolison and K.E. Anapolle ibid.,1975 97,934;(h)J. S.Swenton K. A. Burdett D. M. Madigan and P. D. Rosso J. Org. Chem.,1975,40,1280; (i)U.Widmer H. Heimgartner and H. Schmid Heiv. Chim.Actu 1975,58,2210;(j)J. H. M. Hill T. R. Fogg and H. Guttmann J. Org. Chem. 1975,40,2562. D. W.Jones Me /CHO Me Me ,CDC A (103) (104) (105) butadienyl unit also occurs under mild The rate of the ester shift converting (106) into (107) is strongly dependent on the group R at the migration origin; (106; R = NMe,) rearranges 260 times faster than (106; R= Me).32d Similar E&' E E& E E EE (106) E = CO,Me (107) E = C0,Me (108) E = COzMe effects were observed earlier for 1,Shydrogen shifts in cy~loheptatrienes.~~" In the present correlation of rearrangement rate with the HOMO energy of the radical (108) is observed.For the transient isoindenes (109) observed by flash photolysis the indicated 1,5-H shift is faster for (109;Ar=Ph) than for (109) (110) (109; Ar =p-CNC6HJ.32e Alkyl migration in the related isoindene (1lo) is in contrast to that in simple cy~lopentadienes,~'~ a concerted process; ethyl migration is preferred to methyl migration by a factor of 7.32g Most of the data on migratory aptitude agree with accumulation of positive charge at the migration origin and negative charge on the migrating group at the rearrangement TS. Similar charge separation has been for a photochemical 1,7-hydrogen shift. The accelerating effect of phenyl substitution on thermal 1,7-hydrogen shifts has also been and a kinetic study of 1,5-shifts investigates the effect of incorporating the migrating group into rings of varying size.32i 4 Electrocyclic Reactions An ab initio SCF confirms and refines conclusions of an earlier336 VB study of the photochemical disrotatory closure of butadiene to cyclobutene.The generalized electrocyclic reaction (1 11) is predicted to proceed thermally in a 33 (a) D. Grimbert G. Segal and A. Devaquet J. Amer. Chem.SOC.,1975,97,6629; (6) W. Th.A. M. van der Lugt and L. J. Oosterhoff ibid. 1969,91 6042. Part (i) Orbital Symmetry Correlations and Pericyclic Reactions 69 conrotatory way and photochemically in a disrotatory way when Y =2=CH2and X =CH or NH. More surprisingly if X =2=NH and Y =CH2,all variations of the ring-closure are allowed whilst if X =CH and Y =2=NH the reactions are ther- mally forbidden and photochemically allowed for both modes of ring-~pening.~~" These predictions may be relevant to the that the symmetrical anion (112; X=N) closes stereospecifically to (113) but the unsymmetrical species (112; X =CH) undergoes non-stereospecific closure.Evidence favouring disrotatory opening of cyclopropyl radicals has appeared,35n and the opening of related Ph I A?. N-N H Ph Ph aziridinyl radicals The radical-anion (1 14) with cis-phenyl groups is stable to 0°C but if the phenyl groups are trans electrocyclic ring-opening (114; arrows) occurs at low temperature; the preference for conrotatory ring-opening indicated is in agreement with predictions made on the basis of a correlation diagram or INDO calculations and contrary to FMO prediction^.^^ A mechanism for thermal conversion of 5-into 3-substituted pyrones (Scheme 6) is supported by Scheme 6 18 0-labelling experiments the blocking of the reaction by a 6-methyl group and the more ready rearrangement of corresponding ~yran-2-thiones.~~ The anion (1 15) 34 (a)B.Schilling and J. P. Snyder,J. Amer. Chem. Soc. 1975,97,4422;(6)D.H.Hunter and R. P. Steiner Canad. J. Chem. 1975,53,355 35 (a)S. Sustmann and C. Ruchardt Chem.Ber. 1975,108,3043; (6)S.Sustmann R. Sustmann and C. Ruchardt ibid. p. 1527. 36 N. L. Bauld and J. Cessac J. Amer. Chem. SOC.,1975,97 2284. W. H. Pirkle and W. V. Turner J. Org. Chem. 1975,40 1617. 3' D.W.Jones undergoes rapid ring closure to (117) at -4 1 "C,possibly uiu the mono-trans-isomer (116).38 H H 38 S.W.Staley and A. S. Heyn J. Amer. Chern. Soc. 1975,97 3852.