3. (Part ii) REACTION MECHANISMS By H. M. R. Hoffmann (Department of Chemistry University College Gower Street London W.C.1) NUMEROUS books’ dealing with aspects of organic reaction mechanisms were published in 1967. Stable Carbonium Ions.-Thanks to n.m.r. spectroscopy and the use of highly acidic media the chemistry of long-lived carbonium ions continues to develop rapidly. Clearly these stable species arouse much less controversy than the more elusive intermediates in solvolyses and a readable review on these ions has appeared.2 One of the simplest carbonium ions and at the same time the most simple aromatic system is the cyclopropenyl cation (l) which has been obtained by two independent route^.^ Carbonium ion salt (la) shows a single peak at T -0.87 in fluorosulphuric acid; its i.r.spectrum is also com- paratively simple as expe~ted.~’ A useful new route to alkylcarbonium ions (3) .H Sb CIS. CHzCIz* Sb C16--20” HYH H CI Ref. 3a (la) .-.-CI SO; + CO + CH,OH I Ref. 3b H CO,CH H (Ib) (a)B. Capon M. J. Perkins and C. W. Rees ‘Organic Reaction Mechanism 1966‘ Interscience London 1967; (b) ‘Aromaticity,’ Chem. SOC. Special Publ. No. 21 London 1967; (c) Adu. Phys. Org. Chem.,1967,5;(d) Progr. Phys. Org. Chem. 1967,4; 1967,5 ;(e) H. J. Shine ‘Aromatic Rearrangements,’ Elsevier Amsterdam 1967 ; (f)S. Ranganathan ‘Fascinating Problems in Organic Reaction Mech- anism,’ Holden-Day San Francisco 1967 ; (9) A. W. Johnson ‘Ylid Chemistry,’ Academic Press New York 1966; (h) R. W. Hoffmann ‘Dehydrobenzene and Cycloalkynes,’ Academic Press New York 1967; (i) M.P. Cava and M. J. Mitchell ‘Cyclobutadiene and Related Compounds,’ Academic Press New York 1967 ;01 ‘l,.l-Cycloaddition Reactions-The Diels-Alder Reaction in Heterocyclic Syntheses,’ ed J. Hamer Academic Press New York 1967; (k) L. L. Muller and J. Hamer ‘1,2- Cycloaddition Reactions The Formation of Three- and Four-Membered Heterocycles,’ Interscience New York 1967; (I) ‘Organic Reactions’ A. C. Cope ed. Vol. 15 New York 1967; (m)‘Organic Photochemistry,’ 0.L. Chapman ed. Marcel-Dekker New York 1967; (n)‘Reactivityof the Photo- excited Organic Molecule,’ Proc. 13th Conference on Chemistry at Brussels Interscience New York 1967. G. A. Olah Chem. Eng. News 1967 March 27th p. 76. (a)R.Breslow J. T. Groves and G. Ryan,J. Amer. Chem. SOC.,1967,89,5048;(b)D. G. Farnum G. Mehta and R. G. Silberman ibid. p. 5048; (c) a number of substituted cyclopropenyl salts have been described by B. Fiihlisch and P. Burgle Annakn 1967,701 67. H.M. R. Hoffinann R,CH ______t FSOlH -SbFS R,C+SbFSFSOT + H2 (21 (3) - CH3 - (CH31,CH I+ CH3CH2CH2CH H,C/'\CH is based on hydride-ion abstraction from alkanes (2) by the extremely strong acid FS03H-SbF,.4 Thus n-butane and isobutane form exclusively the t-butyl cation (4) which may be recovered after boiling (150")in FS03H-SbF,! At room temperature and in the same medium alkanes with seven or more carbon atoms are also converted into this ion (4). Even paraffin wax and poly- ethylene give (4) obviously by complex isomerisation and fragmentation pro- cesses.Salts of (4) can be crystallised from SO or SOzCIF solutions at -80" and are stable at least to room temperature. n-Pentane and isopentane ionise to the t-pentyl cation (5) while neopentane (6)forms ion (5) only at -20" in FS03H-SbF,-S0,CIF; at +25" in FS03H-SbF, the t-butyl cation (4) is formed with liberation of methane. That ion (4) does not arise from (5) follows from the fact that (5) is stable up to 150" under these conditions. For many purposes the combination FS03H-SbF,-S02C1F is more suitable than the system FS0,H-SbF,-SO, since carbonium ions are more stable in it and may be studied over a wider range of temperatures4 (see also discussion of the 2-phenethyl cation below).In HF-SbF solution and at room temperature the reaction of isobutane with protons is reversible (Scheme l).' In contrast. HF-SbFs (CH,),CH + H+ (CH3),C+ + H2 Scheme I G. A. Olah and J. Lukas J. Amer. Chem SOC. 1967,89,4739. A. F.Bickel C. J. Gaasbeek H. Hogeveen J. M. Oelderik and J. C. Platteeuw Chem. Corn. 1967 634; see also H. Hogeveen and A. F. Bickel ibid. p. 635. Part (ii) Reaction Mechanisms the fate of abstracted hydride ion in FS0,H-SbF is not yet completely clear.4 The 1-methylcyclopentyl cation can be obtained by no less than nine different routes (Scheme 2). Noteworthy is the rearrangement of (7) into (8) 0 t FSO,H -SbFs I Scheme 2 which apparently does not involve a primary carbonium ion intermediate whose formation would be energetically unfavourable.6 Simple alkyldicar- bonium ions (9) in which the ionic centres are separated by two or more saturated carbon atoms have been observed in SbF,-S02.However attempts to generate the di-cation from (10) have led to the allylic species (11) only.7 Perhaps not surprisingly di-cation (12) does exist; it is formed on dissolving acetylacetone in the very strong acid HF-SbF,.* In FS0,H-SbF,-S02 solution at -60"aliphatic alcohols,9a ethers,9b as well as thiols and sulphides" are protonated. These species show well resolved n.m.r. spectra with negligible exchange rates at -60". At higher temperatures cleavage to carbonium ions G. A. Olah J. M. Bollinger C. A. Cupas and J. Lukas J. Amer. Chem. SOC.,1967,89,2692. J.M. Bollinger C. A. Cupas K. J. Friday M. L. Woolfe and G. A. Olah J. Amer. Chem. SOC. 1967,89 156; see also J. S. McKechnie and I. C. Paul ibid. p. 5482. D. M. Brouwer Chem. Comm. 1967 515. (a) G. A. Olah J. Sommer and E. Namanworth J. Amer. Chem. SOC. 1967,89 3576; see also E. F. Mooney and M. A. Qaseem Chem. Comm. 1967,230; (b)G. A. Olah and D. H. O'Brien J. Amer. Chem. SOC.,1967.89. 1725. lo G. A. Olah D. H. O'Brien and C. U. Pittman jun. J. Amer. Chem. SOC.,1967,89,2996. H.M.R. Hoffmann 128 R' X = F,CI ,O,S N ctc. (13) (14) occurs. There is growing evidcnce for the existence of vinylic cations which have been discussed mainly as intermediates in electrophilic additions to triple bonds and in triple-bond participation during solvolysis.'' The number of carbonium ions which are stabilised by a heteroatom according to [(13) *(14)] is rapidly becoming legion. The a-fluorostyryl cation (16) can be obtained readily from (15) as indicated. In analogous fashion ion (17) has been prepared while attempts to obtain (18) and (19) were un- successful.l2 The first stable fluorophenylcarbonium ions (20)+22) have been generated from alcohol and fluoride precursors in FS0,H and FS0,H-SbF,-S02 at -60". "F N.m.r. deshielding in (20)relative to the parent alcohol is particularly pronounced in the para-positions pointing to strong contribu- tio 9 from quinoidal forms. The deshielding pattern of the ring fluorine atoms in (21) and (22)is similar.' That carbonium ions (22)and (16)exist is interesting fi For a review see (a)P.E. Peterson and J. E. Duddy J. Amer. Chem. SOC.,1966,88,4990; (b)see also D. S. Noyce M. A. Matesich and P. E. Peterson ibid. 1967 89 6225; W. M. Jones and F. W. Miller ibid. p. 1960; M. Hanack I. Herterich and V. Votf Tetrahedron Letters 1967 3871 ;however cf. also H. R. Ward and P. D. Sherman,jun. J. Amer. Chem. Soe. 1967,89 1962. G. A. Olah R. D. Chambers and M. B. Comisarow J. Amer. Chem. SOC.,1967,89 1268. l3 G. A. Olah and M. B. Comisarow J. Amer. Chem. SOC.,1967,89 1027; see also R. Filler C. S. Wan& M. A. McKinney and F. N. Miller ibid. p. 1026. Part (ii) Reaction Mechanisms I29 I CH3 CH3 (15) (16) FScb ';.c /'bFS -etc. F I H F (A8=6.21ppm) (21) (A6=40-1ppm) (20) since the parent benzyl cation as well as the styryl cation have remained elusive.A full paper on ring-substituted benzyl cations has appearedl4 and n.m.r. spectra of aryl carbonium ions have been discus~ed.~~ A rich variety of protonated carbonyl compounds has been observed by n.m.r. spectroscopy. In media such as HF-SbF5 at -20" protonated acetalde- hyde is present as a mixture of syn-(23) and anti-(24) isomers16-18 in a ratio of ca. 5:1. The dramatic difference in JCH,OH (Jamti> .Isyn)as well as the analysis of fine structure leave no doubt that it is the syn-isomer (23) which is the more stable.16,17 As the methyl group in acetaldehyde is replaced by bulkier sub-stituents formation of the syn-isomer is very much preferred.l6. l7 Protonated ketones have also been investigated.and n.m.r. spectra suggest that the l4 J. M. Bollinger M. B. Comisarow C. A. Cupas and G. A. Olah J. Amer. Chem. SOC.,1967 IE9 5687. l5 D. G. Farnum J. Amer. Chem. SOC. 1967,89 2970. l6 G. k Olah D. H. O'Brien and M. Calin J. Amer. Chem. SOC. 1967,89 3582. '' D. M. Brouwer Rec. Trav. chim. 1967,86 879; H. Hogeveen ibid.,p. 696. l8 M. Brookhart,G. C. Levy and S. Winstein J. Amer. Chem. SOC. 1967,89 1735. 130 H.M.R. Hofiann positive charge resides mainly on oxygen (25) the contribution of resonance form (26) being Protonated formic acid exists in the two forms (27) and (28); the ratio (27):(28) is ca. 1 :2 and corresponds to a statistical population of the two forms. Hence their stability must be similar.20 Treatment ,H t-5-47 C-5.74 H R R \+ / + -R?-OH R'?=OH (25) (261 H \ HC+ \, 7 H (27) (28) of carbon monoxide with HF-SbF solution under 10 atmos.pressure did not give any detectable amount of the formyl cation.2' The properties of pro-tonated esters in strongly acidic media are of interest in connection with the mechanism AAcl and AALl of ester hydrolysis; n.m.r. spectra indicate that ion (29) alone is present; no (30) or diprotonated species can be detected. Unfortunately one is still left with the choice of several mechanisms for uni- molecular acid hydrolysis since the ion (30) could be present in very small concentration or else the concerted process (Scheme 3) might apply.22 Pro- tonated dicarboxylic acids and anhydride^,^ acetyl- and benzoyl-pyridinium ions,24 1-formyl and l-a~etyl-azulene~~ have been studied.l9 G. A. Olah M. Calin and D. H. O'Brien J. Amer. Chem. SOC. 1967,89,3586. 2o G. A. Olah and A. M. White J. Amer. Chem. SOC.,1967,89 3591. see also H. Hogeveen Rec. Trau. chim. 1967,86,809. 21 H. Hogeveen A. F. Bickel C. W. Hilbers E. L. Mackor and C. MacLean Rec Trav. chim. 1967,86 687. 22 H. Hogeveen Rec. Trau. chim. 1967 86 816; G. A. Olah D. H. O'Brien and A. M. White J. Amer. Chem. SOC.,1967,89 5694. 23 G. A. Olah and A. M. White J. Amer. Chem. SOC. 1967,89,4752. 24 G. A. Olah and M. Calin J. Amer. Chem SOC. 1967,89,4736. 25 D. Meuche D. Dreyer K. Hafner and E. Heilbronner Helv. Chim. Acta 1967 So 1178. Part (ii) Reaction Mechanisms OH R-c'=O + RdH R'C:4+ + R+ \.OH (31) (321 The methoxycarbonium ion (33) which has been generated from chloro- methyl methyl ether in SbF,-S02 at -60" represents the first stable primary alkoxycarbonium ion.26 The first stable alkyldiazonium ion the 2,2,2-trifluoro- ethyldiazonium ion (35) has been obtained from diazoalkane precursor (34) as shown. Protonation of (34) at the terminal nitrogen atom to give (36) was not detected in the 19Fn.m.r. ~pectrum.~' The reactions of aliphatic diazo- compounds with acids have been reviewed;28 protonated imines have been obtained.29 Halogen-bridged ions have long been recognised as intermediates in certain electrophilic additions and solvolysis reactions to account for the observed stereospecific course and enhanced rates.The ions (37a-c) have now been observed directly on dissolving various 2,3-dihalo-2,3-dimethylbutanesin SbFS-S02 at -60". The n.m.r. spectra reveal the expected ability I > Br > C1 of sustaining positive charge. The fluorine derivative did not form the bridged ion ;instead the rapidly equilibrating pair [(38) +(39)J was observed even at -90°.30 One of the first examples of carbon participation in solvolysis was observed in the phenonium system studied extensively by Cram Winstein and their co-workers. Olah and his group have now shown that the substituted phen- onium ion (41b) can be prepared by direct aryl participation from its precursor 26 G. A. Olah and J. M. Bollinger J. Amer. Chem. SOC. 1967.89.2993;for other alkoxycarbonium ions see H.Hart and D. A. Tomalia Tetrahedron Letters 1967 1347; J. F.King and A. D. Allbutt ibid. p. 49. 27 J. R. Mohrig and K. Keegstra J. Arner. Chem. SOC.,1967,89 5492. 28 R A. More O'FerraU ref l(c),p. 331; see also W. Kirmse and K. Horn Chem. Ber. 1967,100 2699. 29 G.A. Olah and P. Kreienbuhl J. Amer. Chem SOC. 1967,89 4756. 30 G.A. Olah and J. M. Bollinger J. Amer. Chem. SOC.,1967,89,4744. E* 132 H. M.R. Hoffiann + C F3 CH =N=NH / \+/ \ CH, YC X (37a X=CI 1 (37b X=Br) (37c X=I 1 (40).The n.m.r. spectrum of ion (41b) is very simple showing only three types of protons the cyclopropane protons at z 6.53 (relative area 4) the methoxy- protons at T 5.75 (relative area 3) and the AB ring quartets at z 1.88 (relative area 4).31The AB quartets are well separated stressing the benzenonium ion character of (41b) and excluding Brown’s rapidly equilibrating ions (classical or 7~-bridged),~~ which should have phenyl character.In order to appreciate the driving force for phenyl participation (and also some of the experimental difficulties in this field) it is instructive to consider the earlier incorrect report33 that a phenonium ion was formed from threo-and erythro-3-phenylbutan-2-01 (42)in FS03H-SbF5-S02. Under these conditions the benzene ring is sulphin- ated to give the diprotonated ion (43).349 35 Clearly two factors account for the failure to observe a bridged ion in this case :(i) participation by the unsubsti- tuted phenyl group is not sufficiently powerful and (ii) participation is less urgent at a secondary than at a primary centre.Sulphination can be suppressed 31 G. A. Olah M. B. Comisarow E. Namanworth and B. Ramsey J. Amer. Chem. SOC. 1967 89 5259. ’’H. C. Brown R. Bernheimer C. J. Kim and S. E. Scheppele J. Arner. Chem SOC. 1967,89,370. 33 G. A. Olah and C. U. Pittman jun. J. Amer. Chem SOC. 1965,87 3507 3509. 34 M. Brookhart F. A. L. Anet and S. Winstein J. Amer. Chem. SOC.,1966,88,5657; M. Brookhart F. A. L. Anet D. J. Cram and S. Winstein ibid. p. 5659; see also footnote 33 in ref. 35. 35 G. A. Olah C. U. Pittman jun. E. Namanworth and M. B. Comisarow J. Amer. Chem SOC. 1966,88 5571. Part (ii) Reaction Mechanisms 0CH R I I CH2CH2CI (40) (4la:R=H 1 (4lb:R= OCH,) CH3CH-kHCH3 CH3CH-CHCH (42) (431 (43a 1 in the system FSO,H-SbF,-SO2C1F.A great deal of kinetic evidence points to the intermediacy of phenyl-bridged ions.36 Remarkably 2-phenethyl toluene-p-sulphonate is solvolysed 3040 times faster than ethyl tosylate at 75" in trifluoroacetic acid while the corresponding ratio in formic acid is only 2.1 :1 Thus trifluoroacetic acid must be one of the best ionising solvents yet known for solvolytic ~tudies.~' It should be borne in mind that the geometry of the phenonium ion in which the cyclopropyl grouping intersects the six- membered ring at a right angle is not unique; spirodienone (43a) is an inter- mediate of the abnormal Claisen rearrangement and cyclopropylcarbinyl cations prefer the bisected conformation (44).It would be irlteredting l&ooompare these two ions with the hypothetical spiro-cation (45); in this case unlike that of cation (41a) the lowest unoccupied molecular orbital of the cationic system is antisymmetric with respect to the plane through the cyclopropyl ring. Carbonium-Ion Rearrangements and Nucleophilic Substitutions.-The inter-mediates and transition states in various 1,2-rearrangements have been 36 D. J. Cram and J. A. Thompson J. Amer. Chem. SOC. 1967,89 6766; J. L. Coke ibid. p. 135; C. C. Lee and R. J. Tewari Canad. J. Chem. 1967,45 2256; C. C. Lee and B.-S. Hahn ibid. p. 2129; C. C. Lee and L. Noszk6 ibid. 1966 44 2481 2491; C. A. Kingsbury and D. C. Best Tetrahedron Letters 1967 1499; R. Leute and S. Winstein ibid.p. 2475; H. M. R. Hoffmann J. Chem. SOC. 1965 6762. 37 J. E. Nordlander and W. G. Deadman Tetrahedron Letters 1967,4409. H. M. R. Hoflann described by molecular orbital theory.38 Spurred on by the unique tenfold degeneracy of b~llvalene,~' a number of degenerate rearrangements have been reported for carbonium ions and it seems likely that an exciting new chapter of carbonium ion chemistry is about to begin. Doubly degenerate carbonium ions have of course been known for a few years and the ion [(38) + (39)] mentioned above represents just one of several examples. Another case also from the work of Olah4 is ion [(46)+ (47)],in which all five methyl groups are scrambled rapidly even at -180".The barrier for the methyl shift seems to be less than 2-3 kcal./mole and yet there is apparently no tendency to form the static nonclassical ion (48).Even more spectacular is the degeneracy of certain bicyclic and polycyclic systems. Solvolysis of (49) in unbuffered formic acid at reflux yielded a formate in which only 10 +_ 2% deuterium had stayed at C-9 while the remainder was scrambled over the molecule. Thus under forcing conditions the nine carbon atoms of the 9-homocubyl cation (50) achieve complete eq~ivalence.~' The homododecahydryl cation (51) represents a 0 .D (52) (53) '' N. F. Phelan H. H. Jaffk and M. Orchin J. Chem. Educ. 1967,44 626. 39 W. von E. Doering B. M. Ferrier E. T. Fossel J. H. Hartenstein M. Jones jun. G. W. Klumpp R M. Rubin and M. Sunders Tetrahedron 1967,23,3943.40 P. von R. Schleyer J. J. Harper G. L. Dunn V. J. Dipasquo and J. R. E. Hoover J. her. Chem. SOC. 1967,89 699; corrigenda p. 2242; see also J. C. Barborak and R. Pettit ibid. p. 3080. Part (ii) Reaction Mechanisms potentially twenty-one-fold degenerate ion whose synthesis should challenge even the brightest organic chemist.40 After heating [l-l4C)naphthalene in benzene with moist aluminium trichloride for 2 hr. at 60” the radiolabelled carbon was found to be distributed statistically over the carbon skeleton of naphthalene ; this phenomenon has been termed “automeri~ation”.~~ Win-stein and his co-worker~~~~ have described a novel degenerate five-carbon scrambling of the 7-norbornadienyl cation (52). On warming this ion in fluoro- sulphuric acid to -50” atoms C-1 C-6 C-5 C-4 and C-7 become equivalent as shown in (53).A detailed n.m.r. analysis suggests the mechanism in Scheme 4 D Scheme 4 with the first two steps as follows a 1,2-shift by either electron-donating vinylic carbon (C-2 or C-3) to C-7 generates a bicyclo[3,2,0]heptadienyl cation which ring expands immediately either forward (shift of C-3 from C-4 to C-5) or backward (shift of C-2 from C-7 to C-1). Apparently in solvolysis in acetic acid or aqueous acetone the [3,2,0)-system does not leak into the [2,2,1]- cation (nor does the reverse apply).42b Thus the potentially different behaviour of fleeting solvolytic intermediates and long lived carkonium ions should be borne in mind. The possibility of a five-fold degenerate ion [(54) +(55) +etc.] wherein the cyclopropyl ring revolves about the perimeter of the larger ring has also been con~idered.~~ For the first time the energy barrier to “bridge flipping” in various 7- norbornadienyl cations has been determined ;the n.m.r.spectra suggest that all three cations are unsymmetrical [56a-c) +(58a-c)] with flipping barriers of b 19.6 12.4 and <7.6 kcal./mole respectively. Thus the barrier is lowered by an electron-donating group at C-7 which stabilises the symmetrical ion (57b c) but has no appreciable effect on the unsymmetrical species. At higher temperatures the bicycloaromatic (see below) 7-norbornadienyl cation re-arranges to the tropylium ion (59a-c) by a mechanism which remains to be el~cidated.~~ 41 A.T. Balaban and D. FBrcaSiu J. Amer. Chem. SOC. 1967,89 1958. 42 (a) R. K. Lustgarten M. Brookhart and S. Winstein J. Amer. Chem. SOC. 1967 89 6350; (b)Footnote 8 ref. 42a. 43 D. W. Swatton and H. Hart J. Amer. Chem. SOC. 1967,89 5075. 44 M. Brookhart R. K. Lustgarten and S. Winstein J. Amer. Chem. SOC. 1967,89 6352. H.M. R. Hofiann (56a R=H 1 ( 570-c 1 (580-c) (59a-c) (56b:R=CH,) (56~: R=Ph 1 The trishomocyclopropenyl cation (68) was first formulated by Winstein to account for the rate and stereochemical course in the solvolysis of cis-3-bicyclo[3,l,0]hexyl toluene-p-sulphonate (69).45 Compared with other non-classical systems the observed rate enhancement ( 50) is moderate. Extended Huckel calculations indicated that the bridging carbon in the hypothetical cation (66) is stabilised more by bending toward the cyclopropyl group than Relative o.4 I 10'' 6 10" loi4 rates (60) (61) (62) (63) (641 (65) p 01s 45 S.Winstein E. C. Friedrich R. Baker and Y. Lin Tetrahedron 1966 Suppl. No. 8 621 and earlier papers. Part (ii) Reaction Mechanisms to the double bond.46 Two research groups4’ have now demonstrated in- dependently that the tricyclic toluene-p-sulphonate (62) solvolyses lOI4 times faster than its saturated analogue (61);thus the system (62)and (65)“*share the record for nr-participation in solvolysis ! The strikingly different reactivity of (62) and (60) shows again (cf. ions (41a) and (44) above) that cyclopropane interacts only ricr its edge (but not with its “face”) with carbonium ion centre^.^' As regards the different reactivity of (62) and (69) it seems clear that ion (67) is formed much more readily because the trishomocyclopropenyl system in (67) is locked from the beginning into a chair by the bridging ethano-group while considerable energy must be expended to bend the nearly planar five-membered ring in (69) into the chair-like form (68),which is required for maximum orbital overlap.47b Other trishomocyclopropenyl-cation intermediates have been in~estigated.~” However a stable species had remained elusive at the time of writing.That a cyclopropyl group need not necessarily assist ionisation more effec- tively than a double bond can be seen for the toluene-p-sulphonates (70) and (71) which are solvolysed at nearly the same rate in acetic acid.50 Presumably the reduced requirement for participation in the bicyclo[3,3,1]-system as well (70) (71) as the less favourable geometry determine this difference.In the acetolysis of exo-(72) and endo-3-bromobenzenesulphonate(74) more than 90 ”/ bicyclic olefin (73) is formed and both compounds solvolyse some 950 times faster than cyclohexyl bromobenzenesulphonate. In this case relief of nonbounded 46 R. Hoffmann Tetrahedron Letters 1965,3819. 47 (a) H. Tanida T. Tsuji and T. hie J. Amer. Chem SOC. 1967 89 1953; (b) M. A. Battiste C. L. Deyrup R. E. Pincock and J. Haywood-Farmer ibid. p. 1954. 48 S. Winstein and C. Ordronneau J. Amer. Chem. SOC. 1960,82 2084. 49 (a)For further recent studies of the stereoelectronic requirements in cyclopropyl participation see B.Halton M. A. Battiste R. Rehberg C. L. Deyrup and M. E. Brennan J.Amer. Chem. SOC.,1967 89 5964; R. E. Pincock and J. Haywood-Farmer Tetrahedron Letters 1967,4759; (b)W. Broser and D. Rahn Chem. Bm.,1967,100 3472. M. A. Eakin J. Martin and W. Parker Chem. Comm. 1967,955. H.M.R. Hofiann repulsion between the C-3 and C-7 groupings seems responsible for the rate enhan~ement.~ Winstein’s concept of homoconjugation and homoaromaticitys2 discussed above for trishomocyclopropenyl cations has clearly wide applicability particularly to high-energy species (ions excited states and transition states).s3a Conceptually elegant is the extension to “spiroconjugation” and “bicyclo- aromaticity” which has been propounded in three stimulating papers.s3* s4 In Goldstein’s papers4 the concept of aromaticity is no longer tied to planar molecules but clearly taken into the third dimension; as pointed out by this author on the basis of simple MO considerations ions (75)-(78) and higher ”bicycloaromatic‘ 4n systems ”antibicycloarornatic” 4m + 2 systems analogues with 4n x-electrons are expected to possess enhanced thermodynamic stability (h-bicycloaromaticity ’) relative to an appropriate reference compound.Ions (79H82) on the other hand possess 4m + 2 n-electrons and should be antibicycloaromatic in complete reversal of the Hiickel rule. Goldstein has summarised the experimental evidence which is as yet quite meagre ion (75) is the most simple bicycloaromatic species ;regarding potentially antibicyclo- aromatic ions anion (79) could not be obtained on treatment of norbornadiene with amylsodium.Instead only cyclopentadienylsodium and acetylene were observed as primary products (Scheme 5)’’ Furthermore barbaryl toluene-p- sulphonate (83) was solvolysed only in a sluggish reaction despite the neigh- bowing cyclopropyl grouping and two double bonds. Therefore cation (80) ” J. P. Schaefer and C. A. Flegal J. Amer. Chem Soe. 1967,89 5729. ” S. Winstein ref. l(b),p. 5; S. Winstein M. Ogliaruso M. Saki and J. M. Nicholson J. Amer. Chem. SOC. 1967,89 3656; see also J. M. Brown,Chem. Comm. 1967 638. ” (a) H. E. Simmons and T. Fukunaga J. Amer. Chem. Soc. 1967 89 5208; (b)R. Hoffmann A.Imamura and G. D. Zeiss ibid. p. 5215. ” M. J. Goldstein J. Amer. Chem. Soc. 1967,89 6357; see also M. J. Goldstein and B. G. Odell ibid. p. 6356. ” R. A. Finnegan and R. S. McNeeq J. Org. Chem. 1964,2!# 3234. Part (ii) Reaction Mechanisms Scheme 5 cannot be particular stable and all other attempts to generate it were fruitless.3g Chlorination of (84) under mild conditions in ether yielded only rearranged product (85).56Following Goldstein bicycloaromaticity should be measured relative to a reference compound which possesses the same number of trigonal carbon atoms and n-electrons. For example the relative stability of (75) could be gauged by that of the unknown cations (86) and (87); the number k of methylene groups in the reference has to be chosen such as to minimise differ- ences between the o-bond interactions in the bicycloaromatic and the reference.(84) 4 H (86) (87) '' A. S. Kende and T. L. Bogard Tetrahedron Letters 1967 3383. H. M. R. Hohann For a lucid exposition of spiroconjugation the interested reader is referred to the two original papers,53 which should provoke a further flood of experi- mental work. It has been suggested53" that spiroconjugation might extend to kinetic phenomena such as the enhanced rates of solvolysis of certain sulphates and phosphates. The possibility of homoconjugation in certain cumulenes has also been con~idered.~' More than twenty-five different papers have appeared dealing with the norbornyl cation in one way or another; of these only about five seem directly relevant to the solvolytic behaviour of the parent norbornyl cation.An interest- ing model system has been selected by Corey and Glass5* who synthesised the tricyclic exo488) and endo-(89) sulphonates. These two compounds preserve ti (88) (89) Waqncr -Meerrsin rcarranqcrnen H (881 (921 H OH (94) (93) s7 H. Fischer and H. Fischer Chem. Ber. 1967,100,755. s8 E. J. Corey and R. S. Glass J. Amer. Chem. SOC. 1967,89 2600; for a similar approach see R Baker and J. Hudec Chem. Comm. 1967,929. Part (ii) Reaction Mechanisms to a maximum extent the geometry of the norbornyl skeleton and particularly the exact environment at the reaction zone C-1 C-2 and C-6. As expected Wagner-Meerwein rearrangement of the model compound (88) (i.e.,.shift of C-6 from C-1 to C-2) yields (92) which has been estimated to be some 6-7 kcal./mole more strained than the starting material [qualitatively this addition- al strain may be appreciated readily by comparison with the hypothetical cis-and trans-pentalene derivatives (93) and (94)]. If the ion derived from exo-norbornyl toluene-p-sulphonate (90) were to be formulated as a classical ion the rate of ionisation of tricyclic em-sulphonate (88) and (90) should be essentially the same; likewise the rate of solvolysis for endo-norbornyl toluene- p-sulphonate (91)and model (89)would be expected to be similar. The measured rate constants are strikingly in favour of the bridged-ion mechanism 2-endo-norbornyl toluene-p-sulphonate (91) and its tricyclic analogue (89),for which bridging has never been postulated show similar reactivity.However the rate of acetolysis for the tricyclic exo-sulphonate (88)is severely suppressed relative to (90) as one would expect if a bridged nonclassical norbornyl cation is formed from (90) under these conditions. The y-deuterium isotope effects” as well as the activation volume6’ for the solvolysis of exo- and endo-norbornyl bromobenzenesulphonate are also consistent with the intermediacy of a bridged cation. eido -attack (98) (99:R= OH OEt 1 s9 B. L. Murr A. Nickon T. D. Swartz and N. H. Werstiuk J. Amer. Chem. Soc. 1967 SS 1730; J. M. Jerkunica S. BorEiC and D. E. Sunko ibid. p. 1732; Chem. Comm. 1967 1302.6o W. J. le Noble B. L. Yates and A. W. Scaplehorn J. Amer. Chem SOC.,1967,89 3751. 142 H. M.R. Hofiann That the stereospecific exo capture of a norbornyl cation does not necessarily require a bridged precursor has been emphasised by several authors.61 Many reactions of bicyclo[2,2,l]heptenes including electrophilic nucleophilic and free-radical additions as well as 1,3-dipolar cycloadditions proceed with stereo- specific exo-attack.62 It throws an interesting sidelight on the heat of the argument that similar observations had been reported by Alder as early as 1935;63u these have been generally referred to as Alder exo-rule in the German 1iteratu1-e.~~~ An interesting attempt to rationalise this rule has been made by S~hleyer.~~ As an example consider the 1,3-dipolar cycloaddition of phenyl azide to norbornene which yields adduct (96).It may be seen readily from the sideview (97) of norbornene that the hydrogen atoms H-1 and H-2 as well as H-3 and H-4 are partially eclipsed (dihedral angle 20");this is an unfavourable conformation. In reaching the transition state for endo-attack the two hydrogen atoms H-2 and H-3 would have to be bent upwards and the torsional angle to be decreased even further with a concomitant increase in torsional strain. The alternative attack from the exo-side however should be favoured since in approaching the transition state the H-2 and H-3 hydrogens must be bent in endo-direction and torsional strain is relieved. The Alder exo-rule has been excellently reviewed by Klumpp and his co-worker~,~~~ who have also probed experimentally for its limitations.The 3,2-hydride shift observed for the long-lived 2-norbornyl cation can be frozen out in the n.m.r. spectrum by lowering the temperature. However certain substituted norbornyl cations which are usually tertiary and of the C-2 alkylated and arylated type may suffer 3,2-carbon and hydrogen shifts more extensively or even exclusively.64* 65 These shifts appear to proceed with a preference for exo,exo-migration although the first clear example of a 3,2- endqendo-hydride shift has just come to light.66 Presumably bridged tertiary ions are not involved here and in the opinion of the writer Schleyer's torsional hypothesis64 should be considered. Bridgehead-substituted compounds are free from the frustrating ambiguities so often encountered in alicyclic compounds and have long enjoyed special 61 H.C. Brown Chem. Eng. News 1967 Febr. 13th p.87. A. F. Thomas R. A. Schneider and J. Meinwald,J. Amer. Chem SOC. 1967,89,68;H. C. Brown and K.-T. Liu ibid. p. 466 3898 3900; P. von R. Schleyer ibid. p. 3901; see also G. D. Sargenf Quart. Rev. 1966,20 344. (a)K. Alder and G. Stein Annulen 1935,515 185; ibid. J936,525 183 221 ;(b)G. W. Klumpp A. H. Veefkind W. L. de Grad and F. Bickelhaupt Annulen 1967,706,47. 64 P. von R Schleyer J. Amer. Chem. SOC.,1967,89 699; see also ibid. p. 701. 65 C. J. Collins and B. M. Benjamin J. Amer. Chern. SOC.,1967,89,1652; C. J. Collins V. F. Raaen B. M. Benjamin and I. T. Glover ibid.p. 3940; J. A. Berson J. H. Hammons A. W. McRowe R. G. Bergman A. Remanick and D. Houston ibid. p. 2561 ;J. A. Begion A. W. McRowe R. G. Bergman and D. Houston ibid. p. 2563; J. A. Berson and R. G. Bergman ibid. p. 2569; J. A. Berson A. W. McRowe and R. G. Bergman ibid. p. 2573 ;J. A. Berson R. G. Bergman J. H. Hammons and A. W. McRowe ibid. p. 2581 ;J. A. Berson J. H. Hammons A. W. McRowe R. G. Bergmann A. Remanick and D. Houston ibid. p. 2590. 66 A. W. Bushel1 and P. Wilder jun. J. Arner. Chem SOC. 1967,89 5721 ;for another possible example see R. P. Lutz and J. D. Roberts,ibid. 1962,84 3715. Part (ii) Reaction Mechanisms attention in mechanistic studies.67 A spectrum of bridgehead reactivities has emerged and 1-chlorobicyclo[ l,l,l]pentane (98) reacts with particular ease; in 80% aqueous ethanol at 25" this compound is three times more reactive than t-butyl chloride and 1014 times more reactive than 1-chloronorbornane.Unlike its higher homologues compound (98) fragments to give 3-methylene- cyclobutanol and its ethyl ether (99);one possible driving force for this reaction is relief of strain.68 The solvolysis of 8,9-dehydro-2-adamantyl toluene-p- sulphonate has been ~tudied.~' The first saturated bridgehead Grignard reagents (100H102) have been prepared. Compound (102) decomposes in refluxing ether in an ElcB-type reaction (Scheme 6). Not surprisingly the corresponding lithium compound with the greater carbanionic character of the bridgehead is less stable. In striking contrast to the fully fluorinated Grignard reagent (102) com- pounds (100) and (101) are stable in refluxing ether.The reduced stability of Grignard reagent (102) has been ascribed to inside-cage transmission of the F2' MgX F2&) F F2@F2 MgX F2 MgX F2 (100a:X=BrI (IOla-bI (100b X=I I (102a-b 1 lx- Scheme 6 67 R. C. Fort jun. and P. von R. Schleyer Adv. AIicyclic Chem. 1966 1 284; G. J. Gleicher and P. von R. Schleyer J. Amer. Chem. SOC. 1967 89 582; T. McAllister 2. DoleSek F. P. Lossing R. Gleiter and P. von R. Schleyer ibid. p. 5982. K. B. Wiberg and V.Z. Williams,jun.,J. Amer. Chem. SOC.,1967,89 3373. 69 J. E. Baldwin and W. D. Foglesong .I.Amer. Chem. SOC.,1967,89,6372. 144 H.M.R. Hoffmann dipole associated with the carbon fluorine bond of the other bridgehead'O as indicated in formula (102).Perfluorobicyclo[2,2,l]heptanes provide just one example for the marked effect of perfluoroalkyl groups on chemical reactivity. Various workers have proposed that fluorine not only exerts a normal electron-attracting inductive effect but in addition is capable of a conjugative interaction (103b) involving carbon-fluorine no-bond resonance. Since a bridgehead carbanion is forced to remain pyramidal hyperconjugation according to (103b) should be appreci- ably diminished in such a system. 1H-Undecafluorobicyclo[2,2,l]heptane (104) undergoes base-catalysed tritium exchange some five times more readily than tris(trifluoromethy1)methane (105) and it has been suggested that there is no need to invoke such hyperconjugation.If we accept this argument then the tris(trifluoromethy1) anion should have a pyramidal (rather than planar) structure like tri~(trifluoromethy1)amine.~'In the Reporter's opinion fluoro- hydrocarbon (106) would have provided a somewhat better model in this study for the same reason which determines the contrasting reactivity of (101) and (102). According to Bredt's rule a bridgehead double bond cannot exist in a bicyclic system unless the rings are large enough to accommodate the double bond without excessive strain. Bicyclo[3,3,l]non-l-ene (107) has now been obtained by two independent routes. This compound constitutes the smallest ring system yet prepared with a bridgehead double bond; on standing in air it poly- merises.F--c.=-c\/F -C-CNF 'FF 'F ( I07 (I08 1 70 S. F. Campbell J. M. Leach R.Stephens and J. C. Tatlow Tetrahedron Letters 1967,4269. A. Streitwieser jun. and D. Holtz J. Amer. Chem. Soc. 1967,89,692;for a second independent piece of evidence against fluorine hyperconjugation see A. Streitwieser jun. A. P. Marchand and A. H. Pudjaatmaka ibid. p. 693. ''(a) J. A. Marshall and H. Faubl J. Amer. Chem. SOC. 1967 89 5965; (b)J. R. Wiseman ibid. p. 5966. Part (ii) Reaction Mechanisms Neighbouring group participation and rearrangements in cyclopropyl- methyl cyclobutyl and homoallyl systems have been reviewed.73 Since Whitmore described the first authentic neopentyl derivatives the solvolysis of these compounds has been studied with unusual intensity.It is generally accepted and quoted in most textbooks that these reactions entail complete rearrangement of the neopentyl skeleton under carbonium-ion conditions. Fraser and the writer have now analysed the products formed from neopentyl toluene-p-sulphonate in ethanol-water mixtures at 130”. In pure water as much as 10% neopentyl alcohol is formed aside from other products. The distribution of products is solvent dependent and the most simple mechanism embraces at least three intermediates A neopentyl ion-pair the t-pentyl cation and a dimethylcyclopropane precursor.74 Solvolytic studies in water75 and the interpretation of activation parameter^'^ have been reviewed.77 Bedevilling all interpretations of rates particular in water as a solvent is the general question :‘How much do changes in the energy of the initial state (including solvent-solute interactions) and how much do changes in the energy of the transition state contribute to a particular rate ratio?’78 Following Robertson77 this difficulty may be illustrated with the road- grading analogy A road-builder who wishes to lower the height of a hill may do so by (a)scraping offthe top of the hill (b)filling in the valley or perhaps most frequently in real life by (c) a combination of the two operations.lb Fig.1. The road grading analogy It should be pointed out that the internal strain of a molecule which is so frequently invoked as a contributor to the energy of the initial state can be estimated from the measurement of the energy evolved in the thermal re-arrangement of that molecule.Such measurements have been made possible 73 M. Hanack and H.-J. Schneider Angew. Chem. Znternat. Edn. 1967,6,666. 74 G. M. Fraser and H. M. R Hoffmann Chem. Comm. 1967 561; for an elegant demonstration of the role of the counterion in reactions of diazoneopentane with acid see W. Kirmse and K. Horn Tetrahedron Letters 1967 1827. ’’ E. Buncel and P. R. Bradley Canad. J. Chem. 1967 45 515; A. Queen ibid. p. 1619; B. N. Hendy W. A. Redmond and R. E. Robertson ibid. p. 2071; H. S. Golinkin I. Lee and J. B. Him J. Amer. Chem SOC.,1967,89 1307; J. G. Martin and J. M. W. Scott Chem. and Ind. 1967 665. 76 G. Kohnstam ref. l(c) p. 121. ’I7 R. E. Robertson Progr. Phys. Org. Chem. 1967,4 213.(a)See Ann. Reports 1965,62,238; (b)H. M. R. Hoffmann J. Chem. Soc. 1965 6753 6762. 146 H. M.R. Ho_f)Fnann with a temperature-programmed differential calorimeter. Particular attention should be paid to the work of 0th’’’ who by using this instrument has even elucidated the complex steps in the thermal rearrangement of hexamethyl- prismane to hexamethylbenzene. When solving kinetic equations for multistep reactions (e.g. SN1 El and ElcB reactions) one usually assumes that the reactive intermediate (say A) is present in a stationary state (dA/dt = 0). The exact solutions of the kinetic equations however may deviate and criteria for recognising such deviations have been discussed.” The SN2’ reaction is a bimolecular nucleophilic substitution involving allylic rearrangement.That such a reaction involves attack of the nucleophile syn to the leaving group (108) is generally accepted and can also be rationalised by simple HMO theory. For a similar substitution in a system of five carbon atoms (with two conjugated double bonds) one has reached full circle and an anti-relationship of entering and leaving group (cf. SN2 reaction) is predicted. Nucleophilic substitution at an optically active tertiary carbon has received relatively little attention. Displacements in the 2-phenyl-2-butyl system involve an asymmetric ion-pair or its equivalent (formed in the rate-limiting step) which is attacked by the nucleophile in the product-determining step with predominant inversion of configuration.” Similarly the product distribution from the phenyldimethylcarbinyl system implicates the c~unterion,’~ as expected from earlier The principle of hard and soft acids and bases has been discussed for multi- centre (mainly catalysed) reactions8’ and for organic reactions in general.86 When applied with care the concept allows one to appreciate many otherwise unconnected facts ; however the principle does not necessarily invalidate earlier explanations.For example Pearson and Songstad’s reinterpreta- tion86p87 of the ratio koTs/kBr7” has been criticised by Trahanovsky and Doyle.88 Acetolysis of 5-hexenyl toluene-p-sulphonate at 100” leads to open (66 %) and cyclic (34 %) products while the corresponding bromide forms 84 % open and only 16 % cyclic product.Since the olefinic double bond is clearly a soft base and acetic acid is hard one would expect less cyclisation in the acetolysis of the ‘hard’ toluene-p-sulphonate. The results can be understood on the basis that the reaction of the toluene-p-sulphonate involves the more 79 J. F. M. Oth Lecture held at the ‘Symposium on Small Rings’ Louvain Belgium Sept. 1967. K. Frei and H. H. Giinthard Helv. Chim. Act4 1967 SO 1294. W. Drenth Rec. Trav.chim. 1967,86 318. L. H. Sommer and F. k Carey J. Org. Chem. 1967,32 2473 800; see also R. R. Sauers and D. H. Ahlstrom ibid. p. 2233. 83 R. L. Buckson and S. G. Smith J. Org. Chem. 1967,32,634. D. J. Cram and M. R. Sahyun J. Amer. Chem. SOC.,1963,85,1257;M. Cocivera and S. Winstein ibid. p. 1702; P. S. Skell and W.L. Hall ibid. p. 2851. ’’ B. Saville Angew. Chem. Internat. Edn. 1967,6 928. 86 R. G. Pearson and J. Songstad J. Amer. Chem. SOC. 1967,89 1827. R. G. Pearson and J. Songstad J. Org. Chem. 1967,32,2899. W. S. Trahanovsky and M. P. Doyle Chem. Comm. 1967 1021; see also J. Amer. Chem. SOC. 1967,89,4867. Part (ii) Reaction Mechanisms 147 ionic transition state which is also more prone to participation. Other k,,/k, ratios have been determined and discussed.89 Olefin-forming Eliminations.-Sicher and his collaborators continue to make important contributions to the mechanism of bimolecular eliminations. In two detailed papers the syn-anti dichotomy has been described for Hofmann eliminations from cycloalkylammonium and sulphonium salts.In these eliminations the cis-cycloalkenes (n = 5-14 16) are formed by an anti-elimination while the trans-cycloalkenes arise from a syn-route. Remarkably trans-cycloalkenes predominate in rings greater than se~en-rnembered.~~* 91 As an example of an elimination from an open-chain derivative the quarternary ammonium bases (109) and (110) have been studied. Each diastereomer may potentially yield several olefins; the &,trans-pairs shown account for 92 5 % of the products in each case. If (109) were to give the trans-olefin (111) by a syn-elimination and the cis-olefin (112) by the anti-route then these two reactions must proceed with loss of deuterium and should therefore show a H D (I09I H Scheme 7 distinct isotope effect. On the other hand supposing the syn -+ trans and anti + cis dichotomy applies to the corresponding pair (113) and (114)formed from (110) then no pronounced isotope effect would be expected since deuterium is retained in the products.For a variety of conditions e.g. the systems MeOK-MeOH and Bu'OK-Bu'OH it has been found that (113) and (1 14) are indeed formed without appreciable isotope effect (kH/kD 0.9-1-2) 89 D. D. Roberts and J. G. Traynham J. Org. Chem. 1967,32,3177. 90 J. Sicher and J. Zhvada Coll. Czech. Chem. Comm.,1967,32,2122. 9' J. Zkvada and J. Sicher Coll. Czech. Chem. Comm. 1967,32 3701. 148 H. M.R. Hofiann whereas the isotope effect in the analogous reaction of (109) is clearly dis- cernible (kJkD 2.34.7). Thus the original premise is correct i.e.cis-olefins (112) and (114) arise by the anti-route and trans-olefins (111) and (113) by syn-elimination either largely or excl~sively.~~ Sicher’s dichotomy applies also to the formation of cis-and trans-cycloalkenes from cycloalkyl bromides (n = 5-14 16) in Bu‘OK-Bu’OK but not in EtOK-EtOH.93 Several con- clusions follow. Firstly syn-eliminations are much more common than has been hitherto suspected. Medium rings are particularly prone toward this reaction mode open-chain compounds less so and six-membered rings least.94 Secondly syn-eliminations are favoured under Hofmann conditions i.e. with C-H bond breaking in the lead. Typically one uses a powerful base in a solvent which supports ion-pairing ; thus it is visualised that the alkoxide counterion assists the departure of the leaving group electrophilically as shown in Scheme 7.Finally syn-elimination may also intrude when C-X bond breaking is in the lead ; for example solvolytic eliminations from cyclodecyl toluene-p-sulphonates follow the syn-route predominantly if not excl~sively.~~ The challenging question as to the origin of Sicher’s syn-anti elimination dichotomy remains to be answered. Much reinterpretation of earlier data such as k,/k isotope effects Hammett p-values and cis-trans olefin ratios seems imminent . In the light of Sicher’s work a preparative observation by Traynham and collaborator^^^ deserves particular interest When cyclodecyl chloride was treated with Bu‘OK-Me2S0 an olefin mixture (ca. 60 %) was obtained which consisted of almost pure (ca.97%) cis-cyclodecene. On the other hand with lithium dicyclohexylamide in hexane solvent a remarkable reversal occurred ; in this system which seems designed to induce syn-elimination according to Scheme 7 trans-cyclodecene was formed in 96 % purity. syn-Eliminations in the flexible systems discussed above should be dis- tinguished from those occurring in more rigid substrates which cannot attain the anti-periplanar array of H-C-C-X necessary for an anti-elimination. Not surprisingly em-2-norbornyltrimethylammonium hydroxide yields nor- b~rnene~~ in a ‘torsionally enforced’ syn-elimination. In strongly basic media menthyltrimethylammoniumhydroxide forms up to 27 % menth-3-ene by the ~yn-route.~~ In competing substitution and elimination of primary alkyl substrates toluene-p-sulphonates generally give much more product of substitution than the corresponding bromides.Putting this another way koTJkB,(SN2) and 92 M. PBnkova J. Sicher and J. ZBvada Chem. Comm. 1967 394. 93 J. Zivada J. KrupiEka and J. Sicher Chem Comm. 1967 66. 94 M. Svoboda J. Zhvada and J. Sicher Coll. Czech. Chem. Comm. 1967,32 2104. 95 J. G. Traynham D. B. Stone and J. L. Couvillion J. Org. Chem. 1967,32 510; see also G. Wittig and R. Polster Annalen 1958,612 102. 96 (a) J. L. Coke and M. P. Cooke jun. J. Amer. Chem. SOC. 1967 89 2779; (b)These authors appear to have overlooked that practically the same results were described earlier by C. W. Bird R C. Cookson J. Hudec and R. 0.Williams J.Chem. SOC. 1963.410. 97 M. A. Baldwin D. V. Banthorpe A. G. Loudon and F. D. Waller J. Chem. SOC. (B),1967 509. Part (ii) Reaction Mechanisms 149 kOT$kBr(E2) are markedly different from the values observed for pure substi- tution and pure elimination. The observed distortion of koTs/kBr points to an unusual feature of the transition state in these competing processes and a merged mechanism has been proposed;98 such a mechanism has also been considered by Sicher and Zavada" for the elimination of cycloalkyltrimethyl- ammonium salts in Bu'OK-Bu'OH. A novel elimination mechanism the E2cB mechanism in Ingold's termi- nology has been uncovered by Schlosser and Ladenberger;99 on treatment with an organolithium base cis-styryl chloride (115) is converted into the lithium acetylide (1 17).In the first slow step cl-metallated styryl chloride (1 16) The E2c8 mechanism is formed (which can also be identified as an intermediate in tetrahydrofuran at low temperature). In the following fast step a second molecule of base effects dehydrochlorination to (117). The alternative carbenoid route i.e. loss of lithium chloride and hydride shift to give phenylacetylene (118) can be ruled out.99 The ElcB mechanism has been reviewed by McLennan'" and also dis- cussed by the writer.l0' For eliminations in the 2-phenethyl system the ratios koTs/k, increase as the P-proton is rendered more acidic by electron- withdrawing groups (Table). Clearly as C-H bond-breaking (process 'h') increases and the transition state is shifted toward the nearly-ElcB extreme C-X bond-breaking (process 'x') increases in attenuated fashion.Thus the ElcB character of these concerted reactions should be visualised as process 'h'-'x' and it may be appreciated why in general an E2 reaction does not leak into the ElcB route (Scheme 8) merely when the acidity of the P-proton 98 G. M. Fraser and H. M. R. Hoffmann J. Chem. SOC.(B),1967,425. 99 M. Schlosser and V. Ladenberger Chem. Ber. 1967 100 3877 3893 3901 loo D. J. McLennan Quart. Rev. 1967 21,490. lo' H. M. R. Hoffmann Tetrahedron Letters 1967,4393. H.M.R. HofJinann TABLERatios koTs/kBr for E2 reactions in the 2-phenethyl system RC6H4*CH2CH2X Bu*oK~Buto~RC6H4*CH~H2 R kOTslkBr p-Me0 0-15 H 0.22 p-c1 0.44 m-Br 1.19 P-NO2 1.57 is enhanced.'02 As a theoretical handrail for predicting ElcB reactions one should scrutinise the microscopic reverse of the slow step of a potential ElcB reaction ;lo' such nucleophilic additions of typical leaving groups to olefins are rare.However these additions do occur to certain carbenes benzynes other high-energy olefins (cf. Scheme 6) and a,P-unsaturated ketones (cf. The El& mechanism I1 k (slow1 2. -c-c-x 7--II k-i Scheme 8 Michael addition). It is precisely for the formation of these 'olefins' that the ElcB mechanism has been proposed."' The reactions following some carbonyl-methylene condensations furnish instructive examples of ElcB reactions. For instance treatment of 3-nitrobenzaldehyde (1 19) with (120) in the presence of pyridine gives a violet solution (Amax 550 mp) owing to the intermediate carbanion (121); after a few minutes the reaction solution turns cloudy with precipitation of a mixture of (123) and 'olefin' (122).Significantly on adding piperidine to a warm solution of (123) in ethanol the same transient colour (kmax.550 mp) is observed and a mixture of compounds (122) and (123) crystallises on cooling.'03" Because of its insolubility in ethanol olefin (122) can be isolated; in other cases clear proof of an ElcB mechanism is usually more difficult since the olefin formed tends to dimerise and polymeri~e,~~~" as expected for a compound of comparatively high energy."' Presumably the Mannich-Robinson reaction proceeds via an El CBstep also.' 03' lo2 G.M. Fraser and H. M. R. Hoffmann J. Chem. SOC.(B),1967,265. lo' (a) G. Schwenker Arch. Pharrn. 1966,299 131 ;(b)ibid. 1965,298 826. Part (ii) Reaction Mechanisms (122) &OH -& CN (121) A terminology for 1,3-eliminations (Scheme 9) has been proposed by Nickon and Werstiuk.lo4 These authors envisage four principal structures of the transition state and demonstrate two such arrangements experimentally -A-C X-A-0-C-Y + X'Y-Scheme 9 For the concerted formation of nortricyclene (124) from exo-norbornyl toluene-p-sulphonate (90) in Bu'OK-Bu'OH the exo-sickle arrangement (125a) is preferred to the W geometry (125b). For the corresponding reaction of endo-norbornyl toluene-p-sulphonate (91) the initial U geometry (126a) is favoured over the endo-sickle arrangement (126b).Other related reaction^,'^' the Favorsky rearrangement,'06 and the Ramberg-Backlund rearrangement O7 A. Nickon and N. H. Werstiuk J. Amer. Chem. SOC. 1967,89 3914 3915 3917. lo' S. J. Cristol and B. B. Jarvis J. Amer. Chem. ioc. 1967,89 401; ibid. 1966,88 3095. lo6 J. F. Pazos and F. D. Greene J. Amer. Chzm. SOC.,1967,89 1030; H. R. Nace and B. A. Olsen J. Org. Chem. 1967,32 3438; H. 0.House and F. A. Richey jun. ibid. p. 2151; G. W. K. Cavil1 and C. D. Hall Tetrahedron 1967,23 1119; W. Reusch and P. Mattison ibid. p. 1953; C. Rappe and L. Knutsson Acta Chem Scad. 1967,21 163; N. Schamp and W. Coppens Tetrahedron Letters 1967 2697. lo' L. A. Paquette and L. S. Wittenbrook J. Amer.Chem. SOC.,1967,89 4483. 152 H.M.R. Hofiann YOTS H-oTs H H OTs 01s have been Studied. Fragmentation reactions have been reviewed by Grob and Schiess. O8 Nucleophilic Displacement of Vinylic and Acetylenic Halogen.-Until recently three mechanisms have been discussed for the displacement of vinylic and acetylenic halogen (Cl Br and I); threse three mechanisms may be illustrated for the displacement of chlorine from an alkylchloroacetylene (i) Direct Displacement R-CeC-QI &R-CEC-NU 2- (ii) a-Addition and p-elimination R-C'C-CI __C R-C~C-NU 2-(iii) Attack on halogen followed by direct displacement R-C&& 2-[RCEC-+ NU-CI ] -R-CGC-NU Of these three routes the first is considered unlikely and the third viable only in special cases.A new mechanism has been described by Viehe in a lecture :log (iv) P-Addition a-elimination and 'onium rearrangement. R )=2 CI -R-CEC-NU Nu N\ c=c-1 Y lo* C. A. Grob and P. W. Schiess Angew Chem. Internat. Edn. 1967 6 1. Io9 H. G. Viehe Lecture held at the 'Symposium on Small Rings,' Louvain Belgium Sept. 1967; see also H. G. Viehe and S. Y. Delavarenne results reported in H. G. Viehe Angew Chem. Internat. Edn. 1967,6 767. Part (ii) Reaction Mechanisms For example the addition of thiophenolate ion to t-butylchloroacetylene (127) yields the sulphide (128)and has been considered to proceed via mechanism (iv). The last two steps of this mechanism ie. ar-elimination and 'onium rearrange- ment can be observed in other systems also; for instance treatment of (129) Ph S-Na* Bu'-CEC-CI ___f (I271 (I291 (130) with base gives (130) and it remains to be seen whether the phenyl residue is the migrating group ('Fritsch-Buttenberg-Wiechell reaction') or the dimethyl- amino-grouping or both groups.109 Several other questions remain to be answered:"' (i) What in general is the migratory aptitude of groups in the 'onium rearrangement? (ii) Is this rearrangement stereospecifically cis trans or nonstereospecific? (iii) Is a carbene or a carbenoid involved in the rearrange- ment? (iv) Is the new mechanism applicable to displacements in certain halo- olefins and halobenzenes? Truce and his coworkers110 have shown that in contrast to recent claims displacement of chlorine in activated vinylic halides by amines entails complete retention of configuration ;mechanistic details remain to be elucidated.Cyc1oadditions.-A novel cycloaddition principle the 1,4-dipolar cyclo- addition has been clearly exposed by Huisgen in a Since its generalisation by Huisgen in 1959 the 1,3-dipolar cycloaddition has proved to be a fruitful synthetic principle. In the past year alone some thirty papers have dealt with this reaction type and two reviews have appeared.Il3 The 110 W. E. Truce J. E. Parr and M. L. Gorbarty Chem. and Ind. 1967 660. R. Huisgen Lecture held at the 'Symposium on Heterocyclic Chemistry,' Reinhardsbrunn Thuringia DDR Oct. 1967. The lecture will appear in full in Zeitschrift fi Chemie 1968. For published examples of 1,Cdipolar cycloadditions see R.Huisgen M. Morikawa K. Herbig and E. Brunn Chem. Ber. 1967,100 1094; R. Huisgen K. Herbig and M. Morikawa ibid. p. 1107; M. Morikawa and R. Huisgen ibid. p. 1616; E. Winterfeldt and H. Radunq ibid. p. 1680; C. Szhtay and L. Novhk ibid. p. 3038; A. Gomes and M. M. Joullit Chem. Comm. 1967 935. 'I3 R. Huisgen Helu. Chim. Acta 1967,50,2421; R. Huisgen ref. l(b) p. 51. 154 H. M. R. Hofiann Woodward-Hoffmann rules allow concerted photochemical 2 + 2 cyclo-additions but thermally induced reactions should proceed stepwise (Scheme 10). In the wake of this prediction mechanistic interest in the reactions of II + /I t .n F] Or Scheme 10 electron-rich with electron-deficient olefins has been revived.In one of the first detailed investigations of such a cycloaddition Woodward and his collaborators demonstrated that the reaction between diphenylketen (131) and ethoxyacetylene (132) at -25" produced the adducts (134) and (137). The formation of (137) was elegantly rationalised via the dipolar intermediate (133) and the spiro-conjugated species (135) which was visualised to collapse to f phv EtO' Ph,C=C=O + EtOCrCH (131) (132) OEt OEt (135) (136) (l40a X=CI 1 (140b X=OAc) Part (ii) Reaction Mechanisms norcaradiene derivative (136).'l4 Shortly afterwards Huisgen and his co- workers observed a stereospecific addition of ketens onto enol ethers and proposed a concerted mechanism ;'' similar stereospecific reactions were reported by Martin et At the time of writing this Report most authors favoured a spectrum of transition states which was thought to range from a near-concerted' l7 to a stepwise' mechanism' l9 depending on the individual reaction.It should be mentioned that one may replace ketens by vinylic cations in this type of cycloaddition. For instance aluminium trichloride- catalysed trimerisation of but-2-yne yields hexamethyl Dewar benzene in several stages,12' and a similar mechanism appears to account for the forma- tion of aromatic hydrocarbons from alkynes and trifluoroacetic acid.' la For the 2 + 2 cycloaddition of l,l-dichloro-2,2-difluoroethylene to con- jugated dienes a di-radical intermediate had been implicated by the work of Bartlett and his collaborators.'" A recent development is the recognition that transition-metal catalysts may profoundly affect the mechanistic course of cycloadditions.For example the thermally induced 2 + 2 combination (or its microscopic reverse) may become concerted in the presence of metal catalysts.'22 Quadricyclane (138) isomerises smoothly to (139) below 0" in the (1411 (142) (143) presence of ca. 2 mole "/ rhodium(I) palladium(n) and platinum@) complexes ; in the absence of these metals the reaction proceeds with a half-life greater than 14 hr. at 140°.'23 The Reppe synthesis of cyclo-octatetraene (Le. the cyclisation of four molecules of acetylene in the presence of a nickel catalyst) has been considered to proceed in concerted fashion;'24 since a thermally induced metal-free 2 + 2 + 2 + 2 cycloaddition cannot be concerted the '14 J.Druey E. F. Jenny K. Schenker and R. B. Woodward Helv. Chim. Acta 1962,45 600. R. Huisgen L. Feiler and G. Binsch Angew. Chem. Znternat. Edn. 1964,3 753. 'I6 J. C. Martin V. W. Goodlett and R. D. Burpitt J. Org. Chem. 1965,30 4309. 'I' W. T.Brady and 0.H. Waters J. Org. Chem. 1967,32,3703; W. T. Brady and H. R. O'Neal ibid. p. 2704 617. R. Gompper W. Elser and H.-J. Miiller Angew. Chem. Znternat. Edn. 1967 6 453; W. E. Truce D. J. Abraham and P. Son J. Org. Chem. 1967,32,990;L. A. Paquette and M. Rosen J. Amer. Chem. SOC. 1967,89,4102; A. S. Kende Tetrahedron Letters 1967 2661. 'I9 G. Opitz Angew. Chem. Znternat. Edn. 1967,6 107; F. Effenberger and G. Kiefer ibid. p. 951 ; E.V. Dehmlow Chem Ber. 1967,100,3260; I. Fleming and M. H. Karger J. Chem. SOC.(C) 1967,226. W. Schafer and H. Hellmann Angew Chern. Znternat. Edn. 1967,6 518. P. D. Bartlett L. K. Montgomery and B. Seidel J. Arner. Chem. SOC. 1964 86 616; L. K. Montgomery K. Schueller and P. D. Bartlett ibid. p. 622; P. D. Bartlett and L. K. Montgomery ibid. p. 628; see also J. D. Roberts and C. M. Sharts Org. Reactions 1962,12 1. lZ2 F. D. Mango and J. H. Schachtschneider J. Amer. Chem. SOC. 1967,89 2484. 123 H. Hogeveen and H. C. Volger J. Amer. Chem. SOC. 1967 89 2486. G.N. Schrauzer Angew. Chem. Znternat. Edn. 1964,3 185. F H. M.R. Hoffinann transition metal has been considered to provide an orbital pathway for a one- step reaction.lZ2 Electrocyclic reactions.Cyclopropyl-ally1 cation transformations represent the most simple type of electrocyclic reactions and much further work is in accord with the Woodward-Hoffmann rules.'25 If the cyclopropyl cation is stabilised by a second cyclopropyl group as in (140a) ring-opening need not be the exclusive reaction; for example acetolysis of (140a) in the presence of silver acetate yields some 40 % unrearranged product (140b)'26 [the marked stability of the cyclopropylcarbinyl cation (44) has been mentioned above]. What appears to be the first example of the reverse reaction i.e. disrotatory closure of an allylic cation to a cyclopropyl derivative has been reported by Corey and Pirkle.' 27 Bicyclo[2,2,0]pyran-2-one (141) the major product from U.V. irradiation of 2-pyrone forms the tricyclic compound (143) when dissolved in an aprotic solvent at room temperature.Presumably carbon- oxygen fission leads to the substituted cyclobutenyl cation intermediate (142) which is stabilised by 1,3-x-intera~tion'~* and perhaps also by the bidentate interaction with the carboxylate counterion. Covalent collapse at the central carbon of the allylic system furnishes the final product (143).127 Woodward and Hoffmann predicted12' that the opening of cyclopropyl anions to allylic anions should be conrotatory if thermally induced and photochemically disrotatory. By using the isoelectronic aziridines (144) as model13' for a Ma02C C0,Me MU02C H Ar Ar Me0,C H Me0,C C02Me truns 4145 1 cis41451 lZ5 U. Schollkopf K.Fellenberger M. Patsch P. von R. Schleyer T. Su and G. W. Van Dine Tetrahedron Letters 1967 3639; M.S. Baird and C. B. Reese ibid. p. 1379; T. Ando H. Yamanaka S. Terabe A. Horike and W. Funasaka ibid. p. 1123; T. Ando H. Yamanaka and W. Funasaka ibid. p. 2587; L. Ghosez P. Laroche and G. Slinckx ibid. p. 2767; L. Ghosez G. Slinckx M. Glineur P. Hoet and P. Laroche ibid. p. 2773; W. Kutzelnigg ibid. p. 4965; G. H. Whitham and M. Wright Chem. Comm. 1967 294; S. R. Sandler J. Org. Chem. 1967,32,3876; for some possibly anomalous results see W. Kirmse and H. Schiitte,J. Amer. Chem. SOC. 1967,89 1284. lZ6 J. A. Landgrebe and L. W. Becker J. Amer. Chem. SOC.,1967,89 2505. 12' E. J. Corey and W. H. Pirkle Tetrahedron Letters 1967 5255. lZ8 See T. J. Katz and E.H. Gold J. Amer. Chem. SOC.,1964,% 1600. lZ9 R. B. Woodward and R. Hoffmann J. Amer. Chem. SOC.,1965,87 395. Part (ii) Reaction Mechanisms cyclopropyl anion Huisgen and coworker^'^ have provided experimental evidence for this prediction. On heating to loo" cis-(144) is in equilibrium with a small concentration of tran~(145)which can be trapped by efficient dipolarophiles (e.g. dimethyl acetylenedicarboxylate) before leaking into cis-(145).Similarly the azomethine ylid cis-( 145) can be generated stereospecifically from trans-( 144) while the photolytic processes are disrotatory. Opening of cyclobutene to butadiene is symmetry-allowed if conrotatory for the thermal and disrotatory for the photochemical reaction (Scheme 1 l).'" Some time ago dibenzotricyclo-octadiene (146) was reported to form dibenzocyclo- octatetraene (148)on refluxing in o-dichlorobenzene (b.p.180') for 4-5 dR Scheme II (146 1 (I47 1 (I48 1 At room temperature in tetrahydrofuran and in the presence of a molar amount of silver tetrafluoroborate the same isomerisation is complete within 10 sec. The o-xylelene derivative (147)can be trapped in the presence of maleic anhydride as the Diels-Alder adduct (149). Presumably the metal ion and (146) form a complex in which the sterically preferred disrotatory opening is now an allowed process.133 For the opening of the cyclobutenone derivative 30 That amines might not be altogether satisfactory models for carbanions is suggested by work of F. A. L. Anet R. D. Trepka and D.J.Cram J. Amer. Chem SOC.,1967,89,357; see also A. Ratajczak F. A. L. Anet and D. J. Cram ibid. p. 2072. 13' R. Huisgen W. Scheer and H. Huber J. Amer. Chem. SOC. 1967,89 1753. 13' M. Avram D. Diny G. Mateescy and C. D. Nenitzescy Chem. Ber. 1960,93 1789. lJ3 W. Merk and R. Pettit J. Amer. Chem. SOC. 1967,89,4788; see also ibid. p. 4787. H. M.R. Hofmann 0 Ph RG R' Ph M-" Ri MeOD (IS01 Ph H RTy H (150) the terms 'conrotatory' and 'disrotatory' no longer seem to be meaning- ful.' 34 Interestingly thermal and photochemical ring-opening still proceed stereoselectively as indicated; the reason for this finding is not yet clear.'34 The conversion of cis,cis-cyclo-octa-l,3-diene (151)into bicyclo[4,2,00)oct-7-ene (153) on irradiation is a reaction with ample precedent.However the same reaction can also be effected by photosensitisers; in this case (151) is isomerised to the strained cis,trans-cyclo-octa-1,3-diene (152) first which cyclises thermally to (153) in conrotatory fashion. The isomerisation of bicyclohexenyl in the presence of photosensitiser is likely to follow a similar course.'35 The isomeri- sation4' (valence tautomerism) of cycloheptatrienes and n~rcaradienes"~" 134 J. E. Baldwin and M. C. McDaniel J. Amer. Chem. Soc. 1967,89 1537. lJs R. S. H. Liu J. Amer. Chem Soc. 1967 SS,112. IJ6 (a)G. Maier Angew. Chem. Internat. Edn. 1967,6 402; (b)see also E. Ciganek J. Amer. Chem. SOC.,1967,89 1454; T. Mukai H. Kubota and T.Toda Tetrahedron Letters 1967 3581.Part (ii) Reaction Mechanisms 159 and the related processes for benzene oxide and ~xepine'~~ have been reviewed. [161Annulene (154) is conformationally highly mobile. On warming the two cisoid triene systems in (155) are closed to (157) in disrotatory outward disrotatory inward fashion ; similarly the two (photochemically induced) conrotatory closures to (158)proceed in an alternate sense.138 The thermal A A -hY W (155 1 (1571 hv ___t (156) (158) cyclisation of tetraenes to cyclo-octatrienes is a known rea~tion,'~' but the stereochemistry had not been investigated. A preliminary report suggesting a conrotatory closure for cis,cis,cis,trans-8-decadienehas now appeared.I4' Sigmatropic rearrangements. In a concerted thermal 1,3-sigmatropic shift the migrating hydrogen can establish bonding interactions with C-1 and C-3 only if the rearrangement is antarafacial (159).14' A transition state such as (159) is not readily accessible on geometric grounds and therefore such 1,3- hydrogen shifts are rare.14' If however the migrating atom is a carbon atom a suprafacial shift might be possible (160).14' Such a shift must necessarily entail inversion of configuration of the migrating carbon.'42 Some elegant experiments pertinent to these predictions have been described by Berson and his coworker^.'^^ Heating (161) in decalin solution to above 300" gives (162) in which the deuterium and acetoxy group are now both syn to each other.'The (1 61) -+(1 62) rearrangement necessarily suprafacial thus occurs with highly specific inversion of the migrating group C-7.This result is difficult to rationalise in terms of a stepwise mechanism passing over an 137 E. Vogel and H. Giinther Angew Chem. Internat. Edn. 1967,6,385. 13' G. Schroder W. Martin and J. F. M. 0th Angew. Chem Internat. Edn. 1967,6 870. 13' W. Ziegenbein Chem. Ber. 1965,98 1427; H. Meister ibid.,1963,96 1688. 140 E. N. Marvell and J. Seubert J. Amer. Chem. SOC. 1967 89 3377; however see R. Huisgen A. Dahmen and H. Huber ibid.,p. 7130 for a highly stereospecific cyclisation of three geometrically isomeric decatetraenes. R. B. Woodward and R. Hoffmann J. Amer. Chem. SOC. 1965,87,2511. J. A. Berson and G. L. Nelson J. Amer. Chem. SOC. 1967,89 5503; see also J. A.Berson and R. J. Wood ibid.,p. 1043. H. M. R. HofJinann 4fq-0 +;Ac LH,qH AcO D OAc H intermediate in which the C-7-C-1 bond is broken but no significant bonding of C-7-C-3 exists. Such a process would be expected to result in retention or randomisation of configuration. Apparently the preferred approach to the transition state is by compression of the C-2-C-1-C-7 angle and torsion about the C-5-C-6 and C-6-C-7 bonds until oppositefaces of C-7 can bond simultaneously to C-1 and C-3 as in (160). That progress along the reaction co-ordinate should consist of this complex set of motions demonstrates the predictive power of orbital symmetry consideration^."^^ Thermal 1,5-sigma- tropic shifts in medium rings have been investigated further.'43* 144 143 A.P. ter Borg H. Kloosterziel and Y. L. Westphal Rec. Trao.chim. 1967,86,474; K. W. Egger J. Amm. Chem. SOC. 1967,89 3688; J. K. Crandall and L.-H. Chang J. Org. Chem. 1967 32 532; J. K. Crandall and R. J. Watkins Tetrahedron Letters 1967 1717. 144 I thank Mr. J. B. Cresswell for his superb drawings.