4 Reaction Mechanisms Part (i) Pericyclic Reactions By R. J. BUSHBY Department of Organic Chemistry The University Leeds LS2 9JT A simple HMO model for estimating the effect of substituents on the rates of pericyclic reactions has been described and predictions for most cases of monosub-stitution presented in clear tabular form.’ 1 Cycloadditions and Cycloreversions Dewar2 has published details of his calculations on the interconversion of cyclo-hexene and butadiene +ethylene. Despite indications from ab initio calculations3 that this reaction is concerted he remains unconvinced and is still prepared to defend the conclusion of his MIND0/3 method which suggests that it is of the unsymmetrical biradical type. Taken the results illustrate the problems involved in this sort of calculation.At present time-saving approximations have to be made somewhere but whether these are made mainly in the MO method itself or in the method used to search the energy surface the conclusions are never beyond doubt or criticism. Most treatments of reactivity and selectivity in cycloaddition reactions seem to have been in terms of approximate frontier molecular orbital (FMO)theory. Hence although rates of addition of diphenylketen to exocyclic double bonds have been shown not to correlate with alkene ionization potential^,^ a linear relationship has been noted between the ionization potentials of N-acylamino-dienes and their reactivity towards methyl acrylate.’ It has also been shown that the rate of the Diels-Alder reaction between the tetracyclic alkene (1)and tetrazine (2) Py = 2-B.K. Carpenter Terrahedron 1978,34 1877. ’ M. J. S.Dewar S.Olivella and H. S. Rzepa J. Amer. Chem.Soc. 1978,100 5650. L. A. Burke G. Leroy,and M. Sana Theor. Chim.Acfa,1975,40,313; R. E.Townshend G. Ramunni G. Segai W. J. Hehre and L. Salem,J. Amer. Chem. SOC.,1976,98,2190. R.W.Hofhnann and J. Becherer Tetrahedron 1978,34 1193;cf. K.N.Houk Accounts Chem. Res. 1975 11,361. L. E.Ovennan G. F. Taylor K.N. Houk,and L. N. Domelsmith J. Amer. Chem. SOC. 1978,100,3182. 37 R. J. Bushby pyridyl is enhanced by substituents (X =OH or OMe) which interact with the alkene T orbital and raise its energy.6 Secondary orbital interactions (marked *) have been invoked to explain the preferred syn transition state in the addition of phthalimidonitrene to dienes (3)' and LUMO HOMO similar secondary orbital interactions to explain the preference of this nitrene for addition to the 4,5-bond rather than the 2,3-bond of tropone.' The importance of both steric and secondary orbital interactions in determining the stereochemistry of the Diels -Alder reaction has attracted further attention' and is quite nicely illustrated by the work of Martin et a1." Hence the decrease in endo selectivity for the addition of dimethylenecyclobutene to cyclopentadiene (endo :ex0 = 1)relative to the corresponding addition of cyclobutadiene (endo :ex0 >25) can be explained in terms of the smaller orbital coefficients for the centres involved in the secondary orbital interactions as may be seen from the HOMO'Sshown in formulae (4) and (5).(4) For the reaction of cyclobutene derivatives however the decrease in endo selectivity engendered by replacing a methylene by a CMe2 or SO2 grouping is the result of steric effects. Contrary to previous reports the kinetic product of the reaction between furan and maleic anhydride has been shown to be the expected endo adduct" and whereas thiophen is usually unreactive towards this dienophile an adduct is obtained if the reaction is conducted at a pressure of 15 kbar.12 Although the orientation of most photochemical 2 +2 reactions between benzene and alkenes seems to be determined by the natural association between the two reactants it is interesting to note that benzene and vinyl ethers associate in solution in an endo sense but give exo 2 +2 adducts on irradiation.13 M.N. Paddon-Row H. K. Patney and R. N. Warrener J.C.S. Chem. Comm. 1978,296. R. S. Atkinson and J. R. Malpass J.C.S. Perkin I 1977,2242. * D. W. Jones J.C.S. Chem. Comm. 1978,404. J. Kalo D. Ginsburg and J. J. Bloomfield Tetrahedron 1978 34 2153 and following pap&. lo H.-D. Martin R. Iden and H.-J. Schiwek Tetrahedron Letters 1978 3337. M. W. Lee and W. C. Herndon J. Org. Chem. 1978,43,518;cf F. A. L. Anet Tetrahedron Letters 1962 1219. l2 H. Kotsuki S. Kitagawa H. Nishizawa and T. Tokoroyama J. Org. Chem. 1978,43 1471. l3 R. J. Atkins G. I. Fray M. G. B. Drew A. Gilbert and G. N. Taylor Tetrahedron Letters 78,2945. Reaction Mechanisms There seems to have been even more interest than usual in the problem of regioselectivity in the Diels-Alder rea~tion'~*~~ and in other 2 +416*17 and 4 +618 cycloaddition reactions.In many cases these have also been interpreted in terms of FMO theory and in the case of addition of benzonitrile oxide to 3-substituted cyclopentenes the theory has been extended to explain the moderately good linear relationship obtained between selectivity and a function of alkene ionization poten- tial." In view of the number of approximations involved (for example the neglect of product stability coulombic and exchange terms) the degree of success of FMO theory in this area is remarkable and the occurrence of some exceptions not at all surprising. Such anomalies where either the orientation is incorrectly predicted or where large selectivities are only reflected by small differences in the frontier molecular orbitals have however attracted attention and explanations have been sought in terms of the theory itself.Hence Houk has suggested that polarization of the frontier orbitals brought about by the approach of the reagent may be important" and Alston and others have suggested that secondary orbital inter- actions are to blame.".17 Hence the observation that N-t-butyl-nitrone adds to acrylonitrile to give only the adduct (6) but adds to cyanoacetylene to give a 1:1 mixture of adducts (7)and (8)has been attributed to the secondary orbital interaction [marked * in formula (9)] which favours the orientation found in adduct (8) and c/o O\N TCN 6" .-jN + + (6) (7) (8) (9) l4 P.-A.Carrupt M. Avenati D. Quarroz and P. Vogel Tetrahedron Letters 1978,4413;I. J. Westerrnann and C. K. Bradsher J. Org. Chem. 1978 43 3002; T. Imagawa A. Haneda T. Nakagawa and M. Kawanishi. Tetrahedron,1978,34.1893;A. Oliva J. I. Fernandez-Alonso and J. Bertran ibid. p. 2029; P.S.Mariano P. L. Huesmann R.L. Beamer and D. Dunaway-Mariano ibid. p. 2617. lS P.V.Alston R. M. Ottenbrite and T. Cohen J. Org. Chem. 1978,43,1864; T.Cohen R.J. Ruffner D. W. Shull W.M. Daniewski R.M. Ottenbrite and P. V. Alston ibid. p. 4052. l6 J. J. Tufariello and R. C. Gatrone Tetrahedron Letters 1978,2753;D.Mukherjee C. R.Watts and K. N. Houk J. Org. Chem. 1978 43 817; P. Caramella G.Cellerino K.N. Houk F.M.Albini and C. Santiago,ibid. p. 300;G.Cauquis and B. Chabaud Tetrahedron,1978,34,903;H. Benhaoua F.Texier P. Guenot J. Martelli and R.Carrie ibid. p. 1153;E.Stephan Bull. SOC. Chim. France 1978,364;P. Buttioni L.Vo-Quang and Y.Vo-Quang ibid. 1978,401; 415. M. D.Gordon P. V. Alston and A. R.Rossi J. Amer. Chem. SOC. 1978,100,5701. A. Padwa and F. Nobs Tetrahedron Letters 1978.93;L.C. Dunn and K. N. Houk ibid. p. 3411. l9 E.J. McAlduff P. Caramella. and K. N. Houk. 1.Amer. Chem. Soc. 1978,100 105. 2o K.N.Houk L. N. Domelsmith R.W. Strozier and R. T. Patterson J. Amer. Chem. SOC.,1978,100 6531. R. J. Bushby which should be more important in the acetylene than in the alkene.17 The importance of secondary orbital interactions in determining regioselectivity for the Diels-Alder reaction has however been questioned.21 Isotope effects for the addition of diphenylketen to styrene suggest a concerted mechanism22 and X-ray crystallographic studies have confirmed that the stereo- chemistry of the t-butylcyanoketen + 2-methylbut-2-ene adduct (10) is as expected for a ~2a (keten)+w2s (alkene) process.23 Good evidence has however been obtained for a zwitterionic intermediate in the addition of this keten to activated ~lefins,~~ and other ketens.26 Hence its addition to dimethylketen or the imidate~,~~ decomposition of the 4-azido-5-t-butylcyclopentene-1,3-dione (1 1; R = Me) gives compound (12) but its addition to methylketen or the decomposition of dione (11; R = H) gives the lactone (13).These results are best explained in terms of a common zwitterionic intermediate (14;R = Hor Me) in which the CO’may cyclize either to the carbon or the oxygen and it has been shown that this intermediate can be trapped by methanol. (12) (13) (14) Allenyl cations have been shown to add to diene~~~ in a formal 3 + 4 sense in much the same way as oxyallyl cations although on the basis of the regioselectivity of this latter reaction it has been suggested that the mechanism is stepwise rather than concerted.28 Further examples have appeared relating the geometry of 1,3-dipoles to the extent of carbene character.*’ Hence nitrile imines are thought to be less bent than the corresponding nitrile ylides and show more reactivity in the 1,3 sense and less in the 1,1 sense.3o Additions to the planar singlet trimethylenemethane (15;X = H)have been reinterpreted on the assumption that the HOMO has the symmetry shown in formula ” I.Fleming J. P. Michael L. E. Overman and G. F. Taylor Tetrahedron Letters 1978 1313. ” C. J. Collins B. M. Benjamin and G. W. Kabalka J. Amer. Chem. SOC.,1978 100,2570. 23 P. R.Brook A. M. Eldeeb K. Hunt and W. S. McDonald J.C.S. Chem. Comm. 1978 10. 24 D. Becker and N. C. Bodsky J.C.S. Chem. Comm. 1978,237. 25 H.W. Moore L. Hernandez and R. Chambers J. Amer. Chem. SOC.,1978,100,2245. ’‘ H.W. Moore and D. S. Wilbur J. Amer. Chem. Soc. 1978 100 6523. 27 H.Mayr and B. Grubmiiller Angew. Chem. Internat. Edn. 1978,17,130. 28 H.M.R. Hoffmann and R. Chidgey Tetrahedron Letters 1978,85.29 A.Padwa and A. Ku J. Amer. Chem. SOC.,1978,100,2181;A. Padwa P.H. J. Carlsen and A. Ku ibid. p. 3494;A Padwa and P. H. J. Carlsen. J. Org. Chem. 1978,43,3757. 30 A. Padwa S. Nahm and E. Sato J. Org. Chem. 1978,43,1664. Reaction Mechanisms (16) and the LUMO that in formula (17). Hence reaction with alkenes (YCH= CHY) gives fused adducts (18)bur cyclopentadiene a bridged adduct (19). For the methoxy-substituted TMM (15; X =MeO) however the order of the HOMO and LUMO is reversed and hence reaction with alkenes gives bridged adducts (20).31 (18) (19) (20) Whereas it had originally been claimed that the adducts (21) formed between tropone and fulvenes were the result of a direct 6 (fulvene)+4 (tropone) cyclo- addition as a result of stereochemical studies it is now conceded that they arise through a 3,3-sigmatropic rearrangement of an initial 6 (tropcne)+4 (fulvene) adduct (22).32 RK The barrier to thermal loss of nitrogen from the diazetine (23; R=Me) is unexpectedly high33" and since the reaction is highly exothermic it has been suggested that a triplet electronically excited product could result.This however cannot be the case since nitrogen elimination from (23; R=Et) is stereospecific (cis).33bThermochemical data for the elimination of N2and of N20from compound (24; X = N) and (24; NO) suggest that the relative slowness of elimination of N20 may be related to the lower exothermicity of this reaction rather than to orbital symmetry 31 J. A. Berson R. Siemionko A.Shaw G. O'Connell R. D. Little B. K. Carpenter and L. Shen Tetrahedron Letters 1978 3529. 32 I.-M. Tegmo-Larsson and K. N. Houk Tetrahedron Letters 1978 941. 33 (a)P.S. Engel R. A. Hayes L. Keifer S. Szilagyi and J. W. Timberlake J. Amer. Chem. SOC.,1978,100 1876; (b)D. K. White and F. D. Greene J. Amer. Chem. SOC.,1978,100,6760. J. F.M.0th. H. Olsen and J. P. Snyder J. Amer. Chem. SOC.,1977,99,8505. R.J. Bushby 2 Sigmatropic Reactions The skeletal rearrangements and particularly the sigmatropic rearrangements of carbanions have been reviewed.35 It is interesting to note that the 1,4 shift (25)+(26) R = benzyl which is thought to be concerted has a negative volume of activation but the corresponding shift of a benzhydryl group which is thought to involve a radical pair shows a positive volume of a~tivation.~~ OR (25) (26) An example of a thermal 1,3 acyl shift has been described3’ and further evidence presented of the high migratory aptitude of acyl and vinyl groups in 1,5 shifts.38 In one case however it is claimed that 1,5 migration of allyl is faster than that of vinyl.” The kinetics of the rearrangement (27) -B (28)+ (29) have been studied both for (27) (28) (29) the parent hydrocarbon and for its radical anion.4o Quite remarkably the radical anion rearrangement is 10’’ times faster.This could be explained if both reactions involved the intermediate (30;* = -or -) since in the radical anion translation of an electron from an antibonding to a nonbonding orbital would provide the extra driving force.Detailed studies of the interconversions of the bicycloC6.1 .O]nona- trienes (31)-+(32)+etc. show that the 1,7shift is stereospecific with inversion at C-9 which is consistent with a concerted a2a + 7r6s me~hanism.~~ The stereochemistry of 2,3 sigmatropic rearrangements has been the subject of several investigations although no real pattern seems to have emerged. Unlike their open-chain analogues cyclic allyl sulphoxides rearrange through an ex0-transition 35 E. Grovenstein,Angew. Chem. Internat. Edn. 1978,17,313. ” W. J. le Noble and M. R. Daka J. Amer. Chem. SOC.,1978,100 5961. 37 J. M. Janusz L. J. Gardiner and J. A. Berson J. Amer. Chem. Soc. 1978,99,8509. D. J.Field D. W. Jones andG. Kneen J.C.S. Perkin I 1978,1050;M.F. Semme1hack.H.N. Weller and J. Clardy J Org. Chem. 1978,43 3791. ’’ R. D. Miller D. Kaufmann and J. Mayerle J. Amer. Chem. Soc. 1977,H. 8511. *O F. Gerson W. Hukr and K. Mulen Angew. Chem. Internat. Edn. 1978,17,208. 41 F.-G. Klarner and M. Wette Chem. Ber. 1978,111 282. Reaction Mechanisms 43 state (33)."* Also whereas 2,3 rearrangements through the conformation shown in formula (34) leading to trans products are usually preferred exceptions involving conformation (35) and leading to cis products have been reported especially in cyclic sy~tems."~ (33) (34) (35) The mechanism of 3,3-sigmatropic rearrangements and the possible involvement of 1,4-biradicals (36) has remained a major area of interest and controversy. The stereochemistry of the conversion of allylcyclopropene (37) into compounds (38) Ph Ph cis (38) and (39) R=H and Me has been interpreted in terms of a common 1,4-biradical intermediate (40) which gives rise to (38) via a chair-like conformation and (39) (39) (40) through a boat-like onf formation.^" The effects of and ~heny1"~ substituents on the rates of 3,3-sigmatropic rearrangements whilst substantial seem less than would be expected for full 1,4-biradical character and studies of the phenyl-substi- tuted systems (41) do not show the expected linear Hammett plot."' Furthermore 42 R.W. Hoffmann R. Gerlach and S. Goldmann Tetrahedron Letters 1978,2599. '' E. Vedejs. M. J. Arw and J. M. Renga Tetrahedron Letters 1978,523; E. Vedejs J. P. Hagen B. L.Roach and K. L. Spear J. Org. Chem. 1978,43 1185; W. C. Still and A. Mitra J. Amer. Chem. Soc. 1978,100,1927. A. Padwa and T. J. Blacklock J. Amer. Chem. SOC.,1978,100 1321. 45 K. J. Shea and S. Wise J. Org Chem. 1978 43 2710. P. Metzner T. N. Pham and J. Vialle J. Chem. Res. (S) 1978,478. 47 E. N. Marvel1 and T. H.-C. Li J. Amer. Chem. SOC.,1978,100,883. R. J. Bushby Gajewski4* has advanced a thermochemical argument that the 1,4-biradical involved in the isomerization and ring opening of bicyclo[2.2.0]hexane cannot also be involved in the Cope rearrangement of hexa-1,5-diene. However the situation is somewhat complicated by the possibility of several geometrically distinct 1’4- biradical~.~’He has also provided evidence from secondary isotope effects which seems to suggest that the Cope reaction is concerted but readily diverted towards radical character by suitable substituents.Hence for compound (42) secondary isotope effects suggest that bond making and breaking are almost equally advanced in the transition state (concerted mechanism?). For compound (43) bond making seems to be in advance of bond breaking (1,4-biradical-like?). For compound (44) bond breaking seems to be in advance of bond making (tending towards two ally1 radicals?). Q It is interesting to note that the 1,2 rearrangement of ylide (45) to give the compound (46) involves 7 1-8 1YO stereoselectivity (for the intramolecular component) whereas the 1,3 rearrangement which gives rise to (47) involves only H H Ph )-Me PhTMePhYMe PhCO-C-NMe2 Ph-C-COMe ph+Me2Me MeI IOH (46) (45) 56-57% stereoselectivity.This observation seems consistent with the idea that the degree of racemization is related to the degree of translation involved within the radical pair.50“ In related ylide rearrangements there seems to be some evidence to support the notion that ‘allowed’ and ‘forbidden’ processes involves distinct types of radical pairs or transition ~tates.”~ It is also interesting to note that whilst ylide (48) apparently prefers the ‘allowed’ 3,2 rearrangement to the ‘forbidden’ 1,2 re-arrangement for ylide (49) the situation is reversed.’’ A number of interesting studies have appeared of the photochemical di-v- methane rearrangement which may be regarded as involving a formal 1,2 sig- matropic shift.Hence Houk has shown that the propensity of bicyclic systems such as compound (50),to undergo this rearrangement is related to the v/vinteraction as 48 J. J. Gajewski and N. D. Conrad J. Amer. Chem. SOC.,1978,100.6268 6269 49 M. J. S. Dewar S. Kirschner H. W. Kollmar and L. E. Wade J. Amer. Chem. Soc. 1974,% 5242. (a)K. Chantrapromma W. D. Ollis and I. 0.Sutherland J.C.S. Chem. Comm. 1978,672; (b) ibid. pages 670 and 673. ’’ W. D. Ollis M. Rey and I. 0.Sutherland J.C.S. Chem. Comm. 1978,675. Reaction Mechanisms (48) (49) (50) estimated by photoelectron spectros~opy.~~ Similarly the regioselectivity for systems bearing a ring substituent can be rationalized in terms of PMO theory although other treatments give equivalent results.Hence inspection of SOM02for the nitrile (51) and of SOMO1 for the amine (52) shows that bridging to the ortho position (marked +-+)should be favoured whether the substituent is of the acceptor or the donor type.53 Zimmerman has shown that for aryl-substituted 1,4-dienes regioselectivity is moderately well correlated by the substituent constant cand that in the biradical(53) an electron-donating substituent X leads to preferred homolysis of bond b but an electron-accepting substituent cleavage of bond a.54 3 Ene and Related Reactions X-Ray crystallography has confirmed that the product of the ene reaction of a-pinene and chloral as catalysed by ferric chloride has the correct stereochemistry for the least crowded concerted transition state (54),55 but the presence of halogen- H (54) L.N. Dornelsmith P. D. Mollere K. N. Houk R. C. Hahn and R. P. Johnson J. Amer. Chem. Soc. 1978 100,2959. C. Santiago E. J. McAlduff K. N. Houk R. A. Snow and L. A. Paquette J. Amer. Chem. soc. 1978 100,6149. H. E. Zimmerman and T. R. Welter J. Amer. Chem. Suc. 1978 100,4131. M. J. Begley G. B. Gill and B. Wallace J.C.S. Perkin I 1978 93. R. J. Bushby transfer products in the equivalent reaction of 1-substituted and 1,2-disubstituted alkenes has been interpreted in terms of a dipolar intermediate which may react in either of the two modes (55) or (56).56 (55) (56) A P-methoxy substituent has a strong accelerating effect on the rate of decar- boxylation of Py-unsaturated acids,” an observation which is consistent with the changes in charge distribution in the transition state proposed by Dewar (57).58 d+ ;;(::<- H no n-(57) 4 Electrocyclic Reactions STO-3G and 3-41G MO calculation^^^ on the opening of cyclopropylidene to allene have provided a substantially different picture to that based on INDO calculations.60 Kirmse and co-workers have produced an interesting series of papers giving details of their work on the opening of ring-fused cyclopropyldiazonium ions.61 Normally but perhaps not invariably,62 these proceed in the expected disrotatory sense and it seems likely that reaction with solvent occurs before opening to the planar truns,truns-ally1 cation e.g.(58) is complete. Often the highly strained products undergo secondary transformations.Solvolysis of compound (59)61gives among other things cis-3-methoxycycloheptene. This however is formed by iso-merization of the intermediate trans-isomer which can be trapped as its furan adduct t3 pN2+ H@oMe -(58) (59) (60) (60). Thermal ring opening of the aziridines (61)63 proceeds in the expected conrotatory sense to give the azomethine ylide (62),which can be stereospecifically G. B. Gill S. J. Parrott and B. Wallace J.C.S. Chem Comm. 1978 655. ” D. B. Bigley and A. Al-Borno J.C.S. Chem. Comm. 1978,1025. ” M.J. S. Dewar and G. P. Ford J Amer. Chem. SOC.,1977,99 8343. s9 D. J. Pasto M. Haley and D. M. Chipman J. Amer. Chem. SOC.,1978,100 5272. 6o P.W.Dillon and G. R. Underwood J. Amer.Chem. SOC.,1977,99,2435. 6’ W. Kirmse and H. Jendralla Chem. Ber. 1978 111 1857 and following papers. 62 W. Kirmse and U. Richarz Chem. Ber. 1978,111 1895. 63 A. C. Oehlschlager A. S. Yim and M. H. Akhtar Cunad. J. Chem. 1978,56 273. Reaction Mechanisms Ar Ar I I PhAPh Ph Pheyl-trapped by addition to double bonds. As expected for a process dominated by the ArN (HOMO)/(r bond (LUMO) interaction variation of the N-aryl substituent yields a negative Hammett constant (p= -0.80). The equivalent opening of the aziridines (63)is however forced to be disrotatory and seems to be controlled by an ArN (LUMO)/(r bond (HOMO) interaction since the sign of p is reversed (+0.74). It is interesting to note that whilst the potassium salt of the ketone (64)is a stable 107.r-electron aromatic system the lithium salt which is thought to be more covalent undergoes a rapid disrotatory ring closure to give compound (65).64 64 G.Boche and F. Heidenhain Angew. Chem. Internat. Edn. 1978,17,283.