3 (Part iii) REACTION MECHANISMS By A. Ledwith (Donnan Laboratories University of Liverpool) Orbital Symmetry Correlations.-‘Orbital symmetry controls in an easily discernible manner the feasibility and ,stereochemical consequences of every concerted reaction’. Conservation of orbital symmetry as exemplified in the Woodward-Hoffman rules’ for organic reactions occurring in a concerted manner has now become an established landmark in the interpretation of reaction mechanisms. Appli- cation of Woodward-Hoffman rules to the most common types of cycloaddi- tion cycloelimination electrocyclic reactions and sigmatropic rearrangements has been further reviewed by Woodward and Hoffman,’ by Volmer and Ser~is,~ and more exhaustively by Gill4 and by Miller.5 The last review is particularly stimulating on two counts.It attempts to place orbital symmetry correlation in perspective with respect to the general development of theories relating to organic reaction mechanisms and it demonstrates clearly defined correspond- ence between the Woodward-Hoffman approach’v ’ and the more theoretical treatments of Longuet-Higgins and Abrahamson6 and Fukui.’ In addition to these general review articles application of orbital symmetry correlation to the simplest example of an electrocyclic reaction-the cyclopropyl -+ ally1 cation rearrangement-has been surveyed by DePuy. An important new application of orbital symmetry correlation has been deduced by McCapra’ for some reactions displaying chemiluminescence. Previous work has demonstrated that chemiluminescence in biochemical systems results from decomposition of four-membered cyclic peroxides.McCapra’ suggests that if decomposition of the peroxide is concerted then conservation of’ orbital symmetry in this electrocyclic reaction ensures the formation of an excited (luminescent) state of one of the carbonyl fragments. -. . ,C-0 ==C-=O -C = 0 bonding -7 ’ j-+ + >LA >c-:o ;C = O* antibonding R. Hoffmann and R. B. Woodward Accounts Chem. Res. 1968,1 17. ’ R. B. Woodward and R. Hoffmann J. Amer. Chem. Soc. 1965,87,395,2046,2511 4388,4389. J. J. Volmer and K. L. Servis J. Chem. Educ. 1968,45,214. G. B. Gill Quart. Rev. 1968,22 338. ’ S. I. Miller Adu. Phys. Org. Chem. 1968 6 185. H. C. Longuet-Higgins and E.W. Abrahamson .I.Amer. Chem. Soc. 1965,87,2045. ’ K. Fukui Tetrahedron Letters 1965. 2009 2427; Bull. Chem. Soc. Japan 1966 39 498; K. Fukui and H. Fujimoto Tetrahedron Letters 1966 251 (see also L. Salem J. Amer. Chem. SOC. 1968,90 543 553). * C. H. DePuy Accounts Chem. Res. 1968,1 33. F. McCapra Chem. Comm. 1968 155. 144 A. Ledwith In essence the Woodward-Hoffman rules predict 'allowed' pathways for concerted reactions having activation energies substantially lower than corres- ponding 'forbidden' pathways. It is not surprising therefore that estimates of the difference in activation energy between these pathways should have attemp- ted where sufficient data exist The first estimate by Braumann and Golden" predicted that the allowed conrotatory mode for isomerisation of bicyclo-[2,1,0]pent-2-ene (1) to cyclopentadiene (2) was favoured by a difference in activation energy of approximately 15 kcal.mole- '.This value was based on estimates of ground and transition state strain energies and some of the assump- (1) (2) tions were later challenged," leading to a revised estimate of approximately 10 kcal.mole-l. A second estimate of the relative rates of conrotatory versus disrotatory ring opening was made by Doorakian and FreedmanI2 from kinetic measurements on the reaction of 1,4-dimethyl-1,2,3,4-tetraphenyl butadienes. Thus for the allowed conrotatory isomerisation (3) +(4)the predicted pathway was shown to be favoured by an absolute minimum of 7 kcal.mole-'. Ph Ph Ph Ph 25"- Phn M e Me Ph (3) (4) cis-cis'trans-In addition to providing a theoretical basis for many well known concerted reactions the Woodward-Hoffman rules make predictions concerning stereo- chemistry ofproducts resulting from sigmatropic rearrangements.The latter are defined as transformations in which a o-bonded atom or group migrates from one end of a conjugated chain to the other. When the highest occupied molecular orbital of the conjugated system across which migration occurs is antisym- metric conservation of orbital symmetry in a concerted reaction requires that the migrating groups must also undergo an antisymmetric transformation. The group either transverses the nodal plane of the skeleton (antarafacial motion) or suffers an inversion of configuration above the nodal plane (supra- facial motion); e.g.for a rl,51 shift of hydrogen lo J. I. Braumann and D. M. Golden J. Amer. Chem. Soc. 1968,90 1920. E. C. Lupton jun. Tetrahedron Letters 1968,4209. l2 G. A. Doorakian and H. H. Freedman J. Amer. Chem. SOC..1968.90.5310 6896. Reaction Mechanisms D suprafacial antarafacial These predictions have stimulated experimental work; that of Berson and his collaborators on [1,3] and [1,5] sigmatropic rearrangements was referred to briefly in last year’s Report13 and has now been reviewed in detail by Berson.14 Particularly impressive is a completed and extensive experimental study of the thermal rearrangement of endo-bicyclo[3,2,0]hept-2-en-6-y1 acetate (5) to exo-norbornenyl acetate (6).Because of the fixed geometry the [1,3] signatropic transformation demonstrated to occur is necessarily suprafacial and in accord with prediction occurs with complete inversion of the configuration of the migrating group. Additional confirmation of the predictive powers of orbital symmetry correlations comes from the work of Brennan and Hill,” who have demonstrated inversion of configuration at migrating carbon in the [1,4] sigmatropic re- arrangement of bicyclo[3,l,0]hexan-3-ones (7) to bicyclo[3,1,O]hex-3-en-2-ones (8). Cycloaddition Reactions.-In recent years this topic has become increasingly important from both a theoretical and a synthetic viewpoint The growing volume of publications on cycloadditions indicates that ths general type of reaction merits a place beside the better known substitutions eliminations and additions.Prompted by this line of reasoning Huisgen16 l7 has recently published a cogently reasoned consideration of definition classification l3 H. M. R. Hoffman Ann. Reports 1967 4 159. l4 J. A. Berson Accounts Chem. Rex 1968 1 152. l5 T. M. Brennan and R. K. Hill J. Amer. Chem. Soc. 1968,90 5614. l6 R. Huisgen Angew. Chem. Internat. Edn. 1968 7 321. R. Huisgen R. Crashey and J. Sauer in S. Patai ‘The Chemistry of Alkenes,’ Interscience New York 1964 p. 739. 146 A. Ledwith Br cleavaw (7) /of bond a (8) (inversion) (retention) and characterisation of cycloadditions. Accordingly cycldadditions should be defined as inter- or intra-molecular reactions in which the total number of o-bonds increases without elimination of small molecules or ions and in which there is no cleavage of existing o-bonds.Intermolecular processes leading to the formation of four- five and six- membered rings are very common1’ (see later) but the corresponding intra- molecular reactions are not so well characterised. Recent examples include the transformations (9) -+ (10)l8 and those involving (11)” and (12).20 C,H40Me -P C,H40Me -P H H (12) A. Ledwith and D. Parry J. Chem. SOC. R 1967 41; D. W. Adamson and J. W. Kenner J. Chem. SOC. 1935,286; C. D. Hurd and S. C. Lui J. Amer. Chem. SOC.,1935,57,2656. l9 R. Huisgen W. Scheer and H. Huber J. Amer. Chem SOC. 1967 89 1753. *’ A.C. Cope A. C. Haven F. L. Ramp and E. R. Trumbull J. Amer. Chem. SOC. 1952,74,4867. Reaction Mechanisms In further extension of his argument HuisgenI6 has suggested that the pre- vious habit of classfying intermolecular cycloadditions as 1,2-or 1,4- etc. be discontinued. Instead the reaction should be classified according to the numbers of atoms or groups participating in the formation of the new ring. Thus the formation of cyclopropanes (13) from carbenes and olefins is classified as a 2 + 1 (+3) cycloaddition. RCHSH + CBr -+ RCH-CH, \/ CBr (13) Cyclobutanes (14) arise from 2 + 2 (+4) reactions and pyrazolines (15) from 3 + 2 (+5) systems. 2 CF,--CFz + CFZ-CF LF,-LF CH,=CH.CO,Me CH,=N=N++ - (14) C0,Me-0 (15) The new definition and classifications are completely independent of mecha- nism i.e.as to whether the reactions are concerted or multistep processes or whether singlet or triplet excited states are involved in photochemical reactions. For example the most common cycloaddition is the Diels-Alder reaction involving formation of six-membered rings [e.g.(16)]. Diels-Alder additions (‘1,4-cycloadditions’) are undoubtedly concerted in mechanism2’ and would be defined as 4 + 2 (+6) reactions. Six-membered rings (17) may also arise however from the recently discoveredZZ ‘1,4-dipolar additions’ which involve two steps but are also classified as 4 + 2 (-6) reactions. This latter class of cycloadditions has already proved of synthetic value23 J. Sauer Angew. Chem.Internat. Edn. 1967 6 16. ” R. Huisgen K. Herbig and M. Morikawa Chem. kr. 1967 100 1107; R. Huisgen 2. Chem. 1968,8 290. 23 H. Ulrich B. Tucker and A. A. R. Sayigh J. Amer. Chem. SOC. 1968 90 528. 148 A. Ledwith I PhN=C=O -NPh PhNYN'ph (17) 0 in the reactions between sulphonyl isocyanates (18) and carbodi-imides (19). It seems likely that the classification proposed by Huisgen will meet with widespread approval. 2 + 2 +4 Reactions. According to Woodward-Hoffmann rules concerted thermal 2 + 2 cycloadditions are forbidden whereas the corresponding photochemical reactions are allowed.' There are however two main types of 2 + 2 cycloaddition which occur thermally under comparatively mild con- ditions. Cyclobutane formation from halogenated alkene~~~? 25 constitutes one such class of 'forbidden' reaction and addition reactions of ketens the other.26-28 Considerable effort has been expended in mechanistic studies of these types of cycloaddition and a clearer picture of some of the intermediates is beginning to emerge.Bartlett and his collaborators have studied extensively the (gas phase) cycloaddition reactions of halogenated alkene~.~~~ 25 Thus the addition of 1,l- dichloro-2,2-difluoroethylene (20) to butadiene yields the cyclobutane (21) and not (22) suggestjng the intermediate diradical(23). Diradical(23) is about 8 kcal.mole-' lower in energy than (24) and at least 21 kcal.mole-' lower than (25). These energy differences between possible diradical structures if present in the transition states for their formation are great enough to guarantee that as little as 0.001 % of the product will have the reverse orientation if the diradical is an intermediate.Accordingly Bartlett has concluded24 that exclusive head-to- head orientation should result for almost all substituted olefins reacting via diradical intermediates.. Stepwise reactions are not subject to the rules relating to orbital symmetry' and so in this case the apparent anomaly is removed. 24 P. D. Bartlett Science 1968 159 833. 25 P. D. Bartlett G. E. H. Wallbillich A. S. Wingrove J. S. Swenton L. K. Montgomery and B. D. Kramer J. Amer. Chem. Soc. 1968,90,2049;J. S. Swenton and P. D. Bartlett J. Amer. Chem. Soc. 1968,90,2056. 26 R. Huisgen L. A. Feiler and P.Otto. Tetrahedron Letters 1968 4485. 27 R. Huisgen and P. Otto Tetrahedron Letters 1968 4491. 28 G. Binsch L. A. Feiler and R. Huisgen Tetrahedron Letters 1968 4497. Reaction Mechanisms CF2 It CCI P Further evidence for the stepwise diradical nature of cycloadditions in- volving (20) was obtained by a study of its reactions with cis-and trans-double bonds of the isomeric 2,4-hexadiene~.~~~ a stepwise mechanism the 29 In intermediate diradical or dipolar ion may have a lifetime long enough for internal rotation to compete with ring closure. For example in the reactions of (20)with hexadienes (26-28) the cyclobutane products show stereoisomerisa- R jiF2 + CCl CF, II + CCI (2 8) R=Me or C1 29 L. K. Montgomery K.Schueller and P. D. Bartlett J. Amer. Chem. SOC. 1964,86 622. 150 A. Ledwith tion at that double bond which becomes part of the four-membered ring and not at the other one. Stereoquilibration is incomplete and allows a measurement of the competition between rotation and ring closure. The relative rates of these two competing processes were shown to be consistent with reasonable values assigned on the basis of known rotational barriers and known rate constants for the combination of analogous free radicals. In marked contrast to the diradical reactions of halogenated olefins cyclo- additions involving ketens display mechanistic features which in some in- parallel those to be expected for concerted processes (forbidden in this case) and at other times show evidence of dipolar reaction inter- mediates.30*31 Concerted cycloadditions normally have low enthalpies of activation (5-15 kcal.mole-') and very large negative entropies of activation (-25 to -45 e.u.).17 For the reaction between n-butylvinyl ether (29) with diphenylketen (30) to give cyclobutanone (31) the Eyring parameters in benzo- nitrile (31.4") were AH* 9.3 kcal.mole-l AS* -40 e.u. strongly suggesting that the addition was concerted.26 BUT Ph,C=C=O + (3--CH -H' Concerted cycloadditions yield products in which the cis tram-stereochemistry of the reacting double bond is completely retained. In agreement with this cri- terion Huisgen and his collaborators have demonstrated that cycloadducts of diphenylketen and dimethylketen with cis-and trans-propenylpropyl ether (MeCHSH-OPr) are formed with predominant retention of the olefin stereochemistry.28 On the other hand cycloadditions do not normally show regular dependence of rate on solvent polarity.For the reaction between (29) and (30)it was foundz6 that the rate varied linearly with solvent polarity para- meter ET,32although there was only a factor of 50 difference between cyclo- hexane and acetonitrile. Ths rate difference is very small however in com- parison with the corresponding rate ratio of 63,000 for the effect of changing from cyclohexane to acetonitrile for the stepwise cycloaddition of tetracyanoethylene (32) to p-methoxystyrene (33) (CN),C=C( CN) + p-MeO-C,H,-CHSH -+(CN),C-C(CN), II (32) (33) ArCH-CH 'O H.B. Kagan and J. L. Luche Tetrahedron Letters 1968 3093. 31 W. T. Brady and E. D. Dorsey Chem. Comm. 1968 1638. 32 K. Dimroth C. Reichardt T. Siepmann and F. Bohlman Annulen 1963 661 1. 33 D. W. Wiley personal communication quoted by R. Huisgen (ref. 26). Reaction Mechanisms 151 Because of the observed solvent dependence it can be shown26 that the transi- tion state for reaction between (29) and (30) must have a dipole moment greater than that (3.02 D) of the cyclobutanone product (31). Huisgen has suggested26 that all these observations may be rationalised by assuming that reactions of ketens with olefins are concerted but have transition states in which the two new bonds are not formed to the same extent and which have a considerable polarity e.g.(34). Woodward-Hoffmann rules’ would appear to be violated by this interpre- tation of the reaction mechanism and Huisgen has indicated briefly28 how this may be overcome by theoretical reasoning. Full details of the theoretical treat- ment are eagerly awaited especially since there is clear evidence of a two-step procesi for the related cycloadditions of diphenylketen (30) with benzylidene- aniline3’ (35) and with carbodi-imide~~’ (36). PhCH=N-Ph + Ph,C=C===O + PhCH=NPh (35) + I * Ph6H-Tph I I PhEM‘ H-NPh PhCH-NPh A‘ PhZCH-Phz -C=O (37) (39) A particular feature of both the latter reactions is that the dipolar intermediates may be trapped by addition of water or methanol to yield stable adducts (37) and (38) differing from the cycloadducts (39) and (40).3 + 2 -+ 5 Reactions. Cycloadducts of diazoalkanes [e.g. (15)] have been known many years but it was not until the early 1960’s that the classification ‘lY3-dipolar cy~loaddition’~~ 35 became generally accepted. This followed a series of outstanding studies by Huisgen and his collaborator^^^ in which diazoalkanes were shown to represent just one example of a wider class of 1,3-dipolar molecules (abc) which undergo 1,3-cycloadditions and are des- cribed by zwitterionic octet structures e.g. 34 R.Huisgen Angew. Chem. Internat. Edn. 1963,2 565. 35 R. Huisgen. Angew. Chern. Internat. Edn. 1963,2 633. 152 A. Ledwith + +-+-a=b -c i-+a = b = c (b = N) +-+ a = b -~+-+a -b = c (b = NRorO) +-Specific classes of molecular 1,3-dipoles include diazoalkanes (R,C=N=N) nitrile oxides (Arc=&-0) azides (ArN=k=N) nkrones (ArCH=N+ Me- +-0),nitrile imines (ArC=N-N-Ar) sydnones (fkN=CR-CO-0-N) and a great many others.1,3-Dipolar cycloadditions exhibit common mechanistic features :36 they are not markedly influenced as to rate or stereochemistry by solvent polarity they show low enthalpies of activation (5-15 kcal. mole-l) and large negative entropies of activation ( -25 to -45 e.u.) they produce five-membered cyclic compounds in which the stereochemistry of the reacting olefin (dipolarophile) is maintained and finally reaction rates are markedly increased by conjugation of the reacting site in the dipolarophile but reduced by the steric effect of all types of substituent.The synthetic value of 1,3-dipolar cycloadditions has been extensively reviewed by H~isgen,~~ and full experimental details of his nitrile and sydnone~~~ work with nitr~nes,~~ has now been published. Other recent synthetic developments involve use of fluorinated diazoalkanes3' (41) and vinyl diazomethane4' (7) and the formation of high polymers from a variety of 1,3-dipole~.~' For the first time since the conception of concerted dipolar cycloadditions by H~isgen,~~ the reaction mechanism has been challenged in a reasoned and stimulating manner by Fire~tone.~~ Basically the dispute between Fire- stone4 and Hui~gen~~?~~ relates to whether a two-step or concerted addition mode is demanded by experimental observations i.e.jb R. Huisgen R. Grashey H. Hauck and H. Seidl Chem. hr. 1968,101 2043. 3'7 M. Christl and R. Huisgen Tetrahedron Letters 1968 5209. 38 R. Huisgen H. Gotthardt and R. Grashey Chem. hr 1968 101 536; R. Huisgen and H. Gotthardt ibid. pp. 552 1059. 39 J. H. Atherton and R. Fields J. Chem SOC.C,1968 1507. 40 G. Manecke and H. U. Schenck Tetrahedron Letters 1968,2061. 41 J. K. Stille and L. D. Gotter J. Polymer Sci. 4 1968 6 11. 42 R. A. Firestone,J. Org. Chem. 1968,33,2285. 43 R. Huisgen J. Org. Chem. 1968,33,2291. Reaction Mechanisms 153 CH2-CH CF3.CHN + CH2==CH2 I I F3C-CH N .-yq+ (41) R1 R2 +-3, CH,=CH-CH=N=N + R1C=CR2 + c CH N 1\N/ CH I CH2= CH 11 H CH i b ad \c-concerttda/ \c Huisgen I mechanism d===e !\ \,b a c-/ Firestone mechanism d-e.(42) Firestone suggests that the observed independence of rate on solvent polarity is better explained by assuming diradical intermediates (42) as indicated. The observed stereospecific addition would then arise because the energy barrier to rotation around the bond d-e in the diradical(42) is much greater than the activation energy for ring closure or for reversion of (42) to the react- ants. Thus all biradicals (42) which are not formed in the correct conformation for ring closure will revert to starting materials. As noted by Hui~gen,~~ however this premise is in marked contradiction to the high degree of stereo- equilibration observed24 in very similar diradical intermediates formed during reaction of (20) with hexadienes (26-28) (see previous section).Nevertheless workers in other fields44 have previously demonstrated (by calculation) that in free radical polymerisation of methyl methacrylate radicals such as (43) 21' it' -CH2 -CH2-* &O,Me AO,Me (43) have very high barriers to rotation around the CH,-C bond so much so that rotation is energetically less favourable than reaction of the free radical with another molecule of methyl methacrylate. Since similar radicals would be 44 C. E. H. Bawn. W. H. Janes and A. M. North J. Polymer Sci. C 1963.4 427. 154 A. Ledwith involved in dipolar additions via the Firestone mechanism the suggestion that ring closure of diradicals is more rapid than bond rotation would have independent support for at least some systems.If diradical intermediates were involved in 3 + 2 cycloadditions it would be anticipated that the reactivity of typical dipolarophiles towards free radicals would parallel their reactivity in cycloadditions. From studies on copolymerisation of n-butyl maleimide (44;R = Bun) and maleic anhydride (45) with methyl metha~rylate~~ it can CH=CH CH=CH Lo Lo Lo Lo \/ N ‘O/ R (44) (45) be deduced that the double bond in the maleimide is at least one order of magnitude greater in reactivity than that in the anhydride. N-arylmalemides are even more reactive.46 The Firestone mechanism predicts therefore that N-phenylmaleimide should be more reactive than maleic anhydride towards a common 1,3-dipole as found e~perimentally~~ for reactions with aryl +-azides (ArN=N=N).On the other hand for reactions with diphenyl diazo- methane (Ar,C=N=N) maleic anhydride displays a higher reactivity than maleimide N-methylmaleimide or N-~henylmaleimide,~’ arguing against radical intermediates. Firestone suggests4’ that transition states for 3 + 2 dipolar cycloadditions should be coplanar [e.g.(46)]and not having ‘two planes’ [e.g. (47)] as required for an orbital symmetry-allowed concerted process.43 The ‘two planes’ orientation complex (47) was proposed by Hui~gen~~ before orbital symmetry conservation had been recognised as a controlling factor in 45 G. E. Ham ‘Copolymerisation,’ Interscience London 1964. 46 R. C. P. Cubbon Polymer 1965,6,419.‘’ A. Ledwith and A. C. White. unpublished results. Reaction Mechanisms concerted organic reactions and is amply supported by the well established reactivity of sydnones (48) in 1,3-dipolar cycloadditions. Sydnones (48) are R C planar aromatic molecules and only an orientation complex such as (49) is possible. Consideration of the electron distribution in (47) shows that the 1,3-dipolar addition involves interaction of a 4x-electron system with a 2x-electron system-as in the Diels Alder reaction. All aspects of 1,3-dipolar cycloadditions have been surveyed in a spirited defence of the concerted mechanism by H~isgen,~’ although it was admitted that the question of orientation i.e. whether /b\ /b\ a C a C I I I Or I7 d---e e d is not adequately predicted by either concerted or diradical mechanisms.For ths author the experimental evidence points clearly to the Huisgen concerted mechanism for 3 +2 cycloadditions but the challenge by Firestone should stimulate much activity in this area and is all the more welcome. NC CN 6 0 C-C \/ +(NC),C_/\C(CN) *(NC,? \C(CN) NC/\/‘CN 0 (51) 156 A. Ledwith Tetracyanoethylene oxide (50) reacts as a 1,3dip0le,~* via a so called ‘active form’ (51) which may be diradical or dipolar in nature to give mainly tetra- cyano tetrahydrofurans [e.g. (52)l. Although the reaction is one of the most fascinating of 3 + 2 cycloadditions with electron rich olefins and condensed aromatics (2,3-dihydropyran anthra- cene) there is a side reaction involving oxygen transfer e.g.The relative importance of dipolar addition and oxygen transfer has been surveyed by Brown and Cook~on.~’ A novel 1,3dipole has recently been characterised by Turro and his col- laborator~,~~ stimulated by the predictions of Woodward-Hoffman rules. By analogy with ‘active’ tetracyanoethylene oxide (5l) cyclopropanones (53) should react as 1,3dipoles (54) giving rise to concerted 3 + 2 + 5 cyclo-additions or as allylic cations (see later) yielding 3 + 4 + 7 cycloadditions. The former possibility was confirmed for the reaction of chloral (55) with 2,2-dimethylcyclopropanone (53). 0 0-+ \p7 (53) (54) 4 + 2 + 6 Reactions. Diels-Alder cycloadditions are the most important of all cycloadditions and are correctly classified under this heading.From time to time there have been many attempts to explain the stereospecificity lack of clearly defined solvent dependence absence of catalysis etc. by other than a concerted addition process but after carefully considering all the evidence 48 W. J. Linn and R. E. Benson J. Amer. Chem. SOC.,1965,87,3651. 49 P. Brown and R. C. Cookson Tetrahedron 1968,24,2551. N. J. Turro S. S. Edelson J. R. Williams and T. R. Darling J. Amer. Chem. SOC 1968 90 1926. Reaction Mechanisms available Sauer21 was led to favour strongly the concerted mode. The con- certed mechanism is allowed by Woodward-Hoffman rules as a thermal process but forbidden photochemically. Bryce-Smith and Gilbert’ have recently reported an apparent striking exception to these rules.Addition of trans-stilbene to tetrachloro-o-benzoquinoneat 120” gave exclusively the trans-[4 + 21 adduct (56). Photochemically the same reagents reacted in benzene at 15” (under nitrogen) to give 88% of trans- (56) and only 12% of the corresponding cis-isomer (57). It was concluded that although the photoreaction was a two-step process steric control was similar to that in the concerted mode of addition because of dipolar resonance forms of the singlet diradical intermediate (58). It is highly likely that this interesting observation in a 4 + 2 system is directly related to the ‘forbidden’ concerted 2 + 2 reactions of ketens26-28 discussed in the preceding sections and also to the effect of solvent polarity in orientation in some 2 + 2 photochemical reactions.” 4 + 3 Reactions.Cycloaddition reactions of allylic cations with dienes have recently been demonstrated by Hoffman and J0y.53954 In general terms the reaction may be represented as 0 Electron distribution in the transition state relates to that of the tropylium cation in a manner analogous to the relationship between states for Diels-Alder reactions and benzene. ’’ The experimental approach and typical products are adequately described by the following Scheme.53 D. Bryce-Smith and A. Gilbert Chem. Comm. 1968 1701. ” B. D. Challand and P. de Mayo Chem Comm 1968 982. 53 H. M. R. Hoffman and D. R. Joy J. Chem SOC.(B) 1968 1182. s4 H. M. R. Hoffman D. R. Joy and A. K.Suter J. Chem. SOC.(B) 1968 57. ” M. G. Evans 7’rans.Faraday SOC. 1939,35 824; M. G. Evans and E. Warhurst ibid. 1938,34 614. 158 A. Ledwith solvent-separated ion-pair chair-like T.S. CH, / \ 0 SCHEME Cycloaddition of allylic cations could become an important synthetic route to cycloheptane derivatives. Consequently the need to devise reaction conditions which will permit cycloaddition rather than simple cation addition presents a stimulating area for further development. Cationic polymerisation of diene~~~ is normally considered to involve a mixture of 1,2-and 1,4-addition processes e.g. + + R + CH,--CH.CH=CH + RCH,-CH-CH=CH, 1 C4H6 RCH,-CH=CH-CH,-CH,-CH-CH-rH + RCH,-CH-CH=CH, I+ CH,-CH-CH=CH 56 W. Cooper “The Chemistry of Cationic Polymerisation,” ed P.H. Plesch Pergamon London 2-coCCI Reaction Mechanisms 159 Conditions under which cationic polymerisations are made involve solvents of low dielectric ion-pairing of propagating cations complex ion-pairs formed from Lewis acid catalyst fragments and low temperatures. It seems to the author that conditions for cycloaddition of allylic cations will exist during polymerisation of dienes and might explain the well established deficiency in double bond content of the polymers.56 Plausible intermediates would be similar to (59) and (60).Recent studies of cationic polymerisation of diene~~~ have established the presence of saturated six-membered rings in the polymer chain and mechanisms have been suggested for their formation.In the light of the experimental results of Hoffman and 54 a reinterpretation of both polymer structures and reaction mechanism may be necessary. CH;CH*CH= CH (6 0) Carbenes.-During the past 15 years previously uncharacterised divalent carbon derivatives have become commonplace intermediates in a wide variety of useful organic reactions.58’ 59 This enormous growth in interest and applica- tion stems very largely from pioneering work of the research groups led by 1963 p. 351. 57 N. G. Gaylord I. Kossler and M. Stolka A.C.S. Polymer Preprints 1968 9 No. 2 p. 1254; N. G. Gaylord 1. Kossler M. Stolka J. Vodehnal J. Polymer Sci. 1964 A2 3969; B. Matyska K. Mach J. Vodehnal and I. Kossler Coll. Czech. Chem. Comm. 1965 30 2569; N.G. Gaylord B. Matyska K. Mach and J. Vodehnal J. Polymer Sci. 1966 Al 4 2493. 58 W. Kirmse “Carbene Chemistry,” Academic Press London 1964. 59 ‘4.Ledwith “The Chemistry of Carbenes,” R.I.C. Lecture Series of Monographs. 1964. No. 5. 160 A. Ledwith Doering6’ and Skel16’ during the early 1950’s. Carbenes are generated by two main types of reaction (a) photolysis or thermolysis of reactive molecules such +-as diazoalkanes (R,C=N=N) and ketens (R,==C==C=O) and (b) 1,l-elimination (frequently base-induced) from molecules R,CXY where X and Y are halogens. Considering carbene (CH :) as the simplest example the presence of four low-energy bonding orbitals on the carbon atom automatically makes it an electron-deficient species. Two orbitals are used for bond formation with the hydrogen atoms leaving two orbitals for occupation by the remaining two non-bonding electrons.In triplet carbene the two non-bonding electrons are unpaired in degenerate orbitals whereas in singlet carbene these electrons are paired. Spectroscopic studies62 suggest that carbene has a linear triplet ground state which in the gas phase is rapidly formed by collisional quenching of the initial higher energy bent singlet state. In the liquid phase singlet carbene reacts essentially on collision with its nearest neighbour molecules giving rise to random insertion and addition products. Carbenes have been shown to react with C-H 0-H N-H S-H C-Cl C=C C=O C=N and CkC bonds but apart from the obvious synthetic value of these reactions mechanistic studies in carbene chemistry centre on correlation of reactivity and stereochemistry with the spin multi- plicity and excess energy content of the divalent carbon unit.58* 59 63 In addition the correlation between geometrical structure and spin multiplicity of carbenes has been of great interest to theoretical chemists.The electronic structures of carbenes have recently been discussed at length by Hoffman Zeiss and Van Dine.64 Extended Huckel calculations were used to predict geometries of lowest singlet and triplet states. Some of this work has also been included in a timely review by C10ss~~ whch comprehensively surveys carbene structures and the stereochemistry of 2 + 1 carbene cycloadditions. From a consideration of the early experimental results Skell proposed6’ that singlet carbenes react stereospecifically with ethylenic bonds and insert directly into carbon-hydrogen linkages.Triplet carbene on the other hand should react with alkanes by a typical radical abstraction process and yield isomerised products from cis trans-olefins. Skell further concluded that singlet carbenes were electrophilic reagents and would thus react with olefins in a manner similar to that of carbonium ions resulting in an unsymetrical n-approach of the vacant carbene orbital i.e. (62) rather than the more symetrical approach (61). It is significant 6o W. von E. Doering and A. K. Hoffmann,J. Amer. Chem. SOC.,1954,76,6162;W. Von E. Doering and L. H. Knox ibid. p. 4947. 61 P. S. Skell and R.C. Woodworth J. Amer. Chem SOC.,1956,78,4496;P. S.Skell and A. Y.Garner ibid. p. 3409. G. Hertzberg Proc. Roy. SOC. 1961 A262,291; G. Hertzberg and J. W. C. Johns ibid. 1967 A295,107. 63 H. M. Frey. Prop. Reaction Kinetics. 1964. 2. 131 W. Kirmse. Angew Chem. Internal. Edn. 1965 4 1; G. Kobrich ibid. 1967 6 41. 64 R. Hoffman G. D. Zeiss. and G. W. Van Dine. J. Amer. Chem. SOC..1968.90 1485. 65 G. L. Closs “Topics in Stereochemistry,” ed. N. L. Allinger and E. L. Eliel Interscience 1968 vol. 11 p. 193. Reaction Mechanisms therefore that this early intuititive prediction is fully borne out by theoretical calculations and construction of level and state correlation diagrams.66 \ 1 ' ' =< ,C =c =c top view 'C=C / / I' However it is now suggested66 that singlet carbene ('A ') adds stereospecifically not because it is a singlet but because it can correlate with the lowest singlet configuration of a trimethylene and thus with the ground state of a cyclo- propane.Triplet carbene (3B,) adds non-stereospecifically not because it is a triplet but because its complex with a ground state ethylene must correlate with a triplet state of an excited configuration of the trimethylene one in which there are no barriers to rotation around terminal bonds. Symetrical addition of singlet carbene is allowed provided that the carbene is in an excited linear c~nfiguration.~'Calculations and spectroscopic information place the lowest linear singlet of carbene ('Ag) between 10 and 20 kcal. mole-' above the lowest bent singlet ('A').Reacting carbene generated by photolysis (3660 A) of diazomethane is estimated68 to carry approximately 15 kcal. mole-' excess energy. Consequently photolysis of diazomethane could produce excited linear singlet carbene ('Ag) and hence react stereospecifically with olefins oia a symetrical approach.67 Further theoretical consideration of carbene struc- tures has led to the conclusion69 that singlet ground states are to be expected when the carbene carries a substituent having high-energy occupied orbitals [e.g. :CF, :CCl, :C(NR,),] or when it is conjugated with a polyene system having the aromatic grouping of (4n + 2) x-electrons (e.g. cyclopropenediyl or cyclo heptatrienedi yl). Photolysis of diazomethane or keten in the gas phase yields a mixture of singlet and triplet carbene and permits simultaneous investigation of their relative reactivities." Work in this area continues unabated ; carbene from " R.Hoffman J. Amer. Chetn. Soc. 1968 90 1475. " A. G. Anastassiou Chrni. Conini. 1968 991. 68 H. M. Frey ref. 58 p. 221. 69 R. Gleiter and R. Hoffmann. J. Amer. Chem. SOC.. 1968,M. 5457. lo J. A. Bell Progr. Phys. Org. Chem. 1964,2,1; W. B. de More and S. W. Benson Adti. Photochrm.. 1964 2. 219. 162 A. Ledwith photolysis of diazomethane reacts with cyclobutene7 to give (initially) (63)444). The cyclobutane derivatives (63H65) are thought to arise from singlet carbene and undergo further isomerisation (because of excess of vibrational energy) to a mixture of olefins.Vinylcyclopropane (66) is thought to be the major product of reaction with triplet carbene. Gas phase reaction of carbene with alkyl ethers has e~tablished~~ that singlet carbene inserts into the various C-H bonds in random fashion whereas triplet carbene reacts predominantly at the C-H bond alpha to the oxygen atom. With methyl alkyl ethers singlet carbene appears to undergo a displacement reaction producing dimethyl ether and an olefin. By use of nitrogen-15 it has now been demonstrated that carbene (from a variety of precursors) reacts with nitrogen to regenerate diazomethane.’ Photolysis of keten in hydrogen gives rise to a chain reaction in which methane ethane and ethylene are produced.74 From a detailed kinetic analysis it was concluded that methane is not formed by a direct insertion process.The gas phase reactions of triplet carbene with simple alkanes have been studied in detail by Ring and Rabin~vitch.’~ Primary secondary and tertiary C-H bonds undergo hydrogen abstraction by triplet carbene at the relative rates 1:14 150. The same C-H bonds react with triplet methylene by insertion at the relative rates 1 :2 7. Addition of 3CH2 to the double bond in ethylene occurs 3.5 times faster than abstraction from a tertiary C-H bond. For abstraction from primary C-H 3CH2 shows a primary kinetic isotope effect k,/k = 3.9 whereas insertion by 3CH2 into the same bond shows kH/kD = 2. Reaction between carbene (from photolysis of keten) and ethyl chloride appears to occur exclusively by a radical chain process involving the following abstraction reactions.76 :CH2 + CH,.CH,Cl+ CH2C1 + CH3.CH2* :CH + CH3.CH,C1+ CH,. + *CH2*CH2Cl :CH2 + CH3*CH2C1-+ CH3*+ CH3*cHCI ” C. S. Elliott and H. M. Frey Trans. Faraday SOC. 1968 64 2352. 72 H. M. Frey and M. A. Voisey Trans. Faraday SOC. 1968,64 954. 73 A. E. Shilov A. A. Shteinman and M. B. Tjabin Tetrahedron Letters 1968,4177. ” J. W. Powell-Wiffen and R. P. Wayne Photochem. and PhotobioL 1968,8 131. 75 D. F. Ring and B. S. Rabinovitch Canad. J. Chem. 1968 46 2435. ’‘ C. H. Bamford J. E. Casson and A. N. Hughes Proc. Roy. SOC. 1968 A306 135. Reaction Mechanisms All the expected radical combination and disproportionation products were identified and from the effect of added diluents (N and CO) it was concluded that singlet carbene abstracts chlorine preferentially whereas triplet carbene discriminates in favour of hydrogen abstraction.Singlet fluorocarbene (:CFH)77 and chlorocarbene (:CHC1)78 are formed by recoil reactions of energetic tritium atoms with CH,F and CH,Cl respectively. CH2X2+ T* -+ [CHTX,*] + :CTX + HX Both :CHF and :CHCl react stereospecifically with cis-and trans-but-2-ene but do not insert into C-H bonds. Similar reduced activity of :CHC1 over :CH was previously reported for liquid phase reactions of :CHCl generated by photolysis of chlorodiazomethane.79 Phenylcarbene (PhCH :) may be generated by photolysis of three different precursors (67-69)80 but the relative rates of insertion into the various C-H (iii ) H Ph .yak H Ph (67) (68) linkages of n-pentane are independent of precursor.It was concluded therefore that the initially formed singlet PhCH is rapidly equilibrated to a common vibrational level prior to reaction with the alkane. Phenylcyanocarbene (PhC-CN)81 is produced on photolysis of the oxiran (70) but a more efficient and simpler preparation involves photolysis of the 1,3,2-dioxaphosph(v)olan (71).81 This particular carbene is novel in being a canonical form of the nitrene (73) and is also produced by photolysis of the aide (72).81 -N* PhCd-N -PhCg-N +-+ PhC-C=N l7 Yi-Noo Tang and F. S. Rowland J. Am. Chem. SOC. 1967,89,6420. YI-NooTang and F. S. Rowland J. Amer. Chem. SOC. 1968,90 574. 79 G. L. Closs and J.J. Coyle J. Amer. Chem. SOC.,1962,Sq 4350; 1965,87,4270. *' H. Dietrich G. V. Griffin and R. C. Petterson Tetrahedron Letters 1968 153. P Petrellis and G W Griffin Chem Comm . 1968. 1099 E. Schmitz Angew. Chem. 1964 76 197; H. M. Frey and 1. D. R. Stevens Proc. Chem. SOC. 1362. 79; J. Amer. Chem. SOC. 1962,84 2647. 164 A. Ledwith Diazirines are isomeric forms of diazoalkanes and may be used as carbene precursors although not so conveniently as the corresponding diazoalkane. 82 It has now been reported83 that phenylbromodiazirine is a particularly convenient (photochemical) source of phenylbromocarbene (PhCBr) and photolysis of the diazirine in olefins gives essentially quantitative yields of cyclopropanes with retention of configuration.Thermal decomposition of bromodichloromethylphenylmercury (PhHgCBrCl,) in refluxing benzene has proved to be one of the most convenient method for generating dichlorocarbene. 84 The corresponding mercurial PhHgCC1,F reacts in a similar manner to generate chlorofluorocarbene in the most convenient synthesis yet reported for this reactive intermediate.85 A new synthesis of unsaturated carbenes 86 [e.g. (75)] involves the reaction of bases with nitro-oxazolidinones (74). The corresponding unsaturated diazo- alkane is first formed and the carbene demonstrated to be an intermediate by trapping with cyclohexene or alkyl vinyl ethers. Carbenes and carbonium ions may be regarded as base and conjugate acid respectively i.e. R,C + H+ f R2CH+ Recently a novel and most interesting carbene synthesis has been developed (MeS),kH BF h(MeS),C (76) (MeO),kH BF; a(Me0)2C (77) 83 R.A.Moss Tetrahedron Letters 1967,4905. 84 D.Seyferth and J. M. Burlitch J. Amer. Chem. SOC.,1963,85 2667. 85 D.Seyferth and K. V. Darragh J. Organometallic Chem. 1968,11 9. 86 M.S.Newman and A. 0.M. Okorodudu J. Amer. Chem. SOC. 1968,90,4189. Reaction Mechanisms 165 based on tlus eq~ilibrium.~~ The stable cations (76 and 77) were treated with bases to yield corresponding bisalkylthio- and dialkoxy-carbenes. Neighbouring group participation is now well documented as a contributing factor in the formation and reactivity of electrophilic reagents particularly carbonium ions.88 In a recent publication Robson and Shlechtersg have reported perhaps the first clear case of a similar phenomenon during rearrange- ment to a divalent carbon species.Thermolysis of diazolkanes (78) yields beta-substituted styrenes (79) by rearrangement of initial carbene fragment. A ArIz-CH,X rN+ Are-CH,X +ArCH==CHX 2 (78) (79) X = OMe or NMe In contrast the corresponding mercapto-derivative (80) yields a carbene which rearranges by exclusive (SEt) shift to give the corresponding alpha-substituted styrene (82) A -+ ArC-CH,.SEt + -N2 (80) The authors suggest that increased migratory aptitude of RS (over RO and R2N)results from an 'ylide' transition state (81) involving 3d orbitals of sulphur. Many years ago Chattgo suggested that the adduct of ethylene with platinous chloride was a methylcarbene complex of platinum e.g.[CH,-CH :PtClJ-. Later work made this proposal untenable and led to the presently accepted theory of metal-olefin complexes.9' At that time carbenes were not recognised as reaction intermediates but recent workg2 has shown that carbene-metal complexes may be obtained having a structure similar-to that originally suggested by Chatt.go Carbenes of the type MeOCR and HOCR form crystalline derivatives with carbonyl derivatives of chromium tungsten molybdenum and manganese e.g. Ph(Me0)C :W(CO), Ph(Me0)C :Cr(CO), R(Me0)C Mn(C,H,)(CO),. Crystal structures for some of these complexes have been determinedg3 and it is interesting that for Ph(Me0)C :Cr(CO) the carbene unit is bent with a Ph-C-OMe angle of 104".This compares favourably with the value of 103"for the corresponding angle in singlet carbene.62 '' R.A. Olofson S. W. Walinsky J. P. Marino and J. L. Jernow J. Amer. Chem. SOC. 1968,90 6554. B. Capon Quart. Rev. 1964,28 45. *' J. H. Robson and H. Shechter J. Amer. Chem. SOC. 1967,89 7112. J. Chatt Research 1951 4 180. J. Chatt and L. A. Duncanson J. Chem SOC.,1953,2939. " E. 0.Fischer and A. Riedel Chem Ber. 1968 101 151. 93 0.S. Mills and A. D. Redhouse J. Chem. SOC.(A) 1968 642. 166 A. Ledwith Nitrenes.-Nitrenes (RN :) are isoelectronic with carbenes (R,C :) but have only been recognised as discrete intermediates since the early 1960’s. Like carbenes they possess close-lying singlet and triplet electronic states which control their reactivity.For example singlet ethoxycarbonylnitrene (EtOCON :) undergoes selective and stereospecific insertion into C-H bonds producing amines and adds stereospecifically to olefins to give aziridines. Triplet ethoxy- carbonylnitrene however does not undergo insertion into C-H bonds and adds to olefins with complete loss of the geometric c~nfiguration.’~ Carbenes are generated by photolysis of diazoalkanes (R,CN,)58v 59 and by analogy nitrenes are most conveniently obtained by photolysis of azides (RN,). 94 Similarity between carbenes and nitrenes is most marked in their carbonyl derivatives. Carbonylcarbenes are produced from diazo-ketones and yield ketens by concerted (Wolff) rearrangement. Similarly nitrenes from carbonyl azides yield isocyanates by concerted (Curtius) rearrangement,94 e.g.RCO-CHN -RCO*CH -RCH=C=O -N RCON3 -RCO-N -RN=C=O -N For pivaloylnitrene (83),95 rearrangement to t-butyl isocyanate (84)competes favourably with trapping of the nitrene by cyclohexene Me,C.CON Me,C.CON + Me,C-N=C=O (83) (84) (45 %) Aromaticg6 and heterocyclicg7 nitrenes readily undergo ring contraction to the corresponding nitrile e.g. Cyanonitrene (NCN) is generated by thermolysisg8 or photolysisg9 of solutions of cyanogen azide (N,CN). The rules of spin conservation demand 94 W. Lwowski Angew Chem. Internut. Edn. 1967,6 897. 95 G. T. Tisue S. Linke and W. Lwowski J. Amer. Chem. SOC.,1967,89,6303. 96 E. Hedaya M. E. Kent D. W. McNeil F. P. Lossing and T. McAllister Tetrahedron Letters 1968,3415.97 W. D. Crow and C. Wentrup Chem. Comm. 1968 1082 1026. 98 A. G. Anastassiou and H. E. Simmons. J. Amer. Chem. SOC.. 1967.89 3177. 9y H. W. Kroto J. Cheni. Phys. 1966 44 831. Reaction Mechanisms that initially produced cyanonitrene should be in a singlet electronic (excited) state but there has been some confusion over this point since irradiation of N3CN through a Pyrex filter was reported"' to give triplet cyanonitrene. A more recent and detailed investigation of the photolysis of cyanogen azide in the presence of cis-and trans-1,2-dimethylcyclohexane(insertion into the tertiary C-H bonds) has now demonstrated unambiguously that photolysis of cyanogen azide with light ranging from 2100 to 3000 A produces exclusively singlet NCN.''l The earlier results had led to a suggestion1" that the 2750 8 band in the spectrum of N3CN represents a singlet-to-triplet transition and hence photolysis with light of this wavelength yielded triplet NCN directly- a conclusion now shown to be erroneous.1o1 Further studies with cyanogen azide have shownlo2 that three distinct processes occur during thermal reaction with cyclo-octatetrarene (cot).These are a bimolecular reaction leading to alkylidenecyanamide (85) a 2 + 1 cycloaddition of singlet cyanonitrene yielding aziridine (86) and 4 + 1 (stepwise) cycloaddition of triplet cyanonitrene yielding the adduct (87). N3CN cot c -N cot NCN I collisional deactivation I A major difference between carbenes and nitrenes is that the latter readily undergo collisional deactivatioii from excited singlet states to ground triplet states in solution.Consequently there are marked solvent effects on products arising from reactions of nitrenes according to the relative rates of solvent loo L. J. Schoen J. Chem. Phys. 1966,45 2773. lo' A. G. Anastassiou and J. N. Shepelavy J. Amer. Chem. SOC.,1968,90 492. A. G. Anastassiou,J. Amer. Chem. SOC.,1968. 90.1527. 168 A. Ledwith deactivation to triplet electronic states and direct (stereospecific) insertion and addition reactions of the singlet state. Io3 Beckwith and Redm~ndl"~ have evaluated these effects for reactions between ethoxycarbonylnitrene with cis-and trans-but-2-ene. It was found that solvent deactivation of singlet EtO CON was temperature-dependent with an activation energy greater than that for stereospecific reaction with the olefin.In a similar study using anthracene,lo5 the same authors demon- strated a dichotomy of mechanism involving direct substitution (b)with triplet nitrene and intermediate aziridine formation (a)from the singlet derivative. NH*C02 Et a-@JJ \ // H The nature of the stereospecific insertion reactions of nitrenes into C-H bonds continues to be of some interest. A detailed investigation of the reactivity of ethoxycarbonylnitrene towards the various positions of cyclohexane norbornane bicyclo[2,2,2]octane and adamantane led to the conclusionlo6 that both free-radical and nitrene insertion reactivities are governed similarly by structural variations.Benzene reacts with singlet methylsulphonylnitrene to give mainly the sulphonamide (88). Detailed studies by Abramovitch and Uma107 have established that the reaction involves equilibrating intermediates (89) and (90) and a minor product (91). At 120" the reaction yields predominantly (88) via (90). The azepine (91) was trapped (in low yield) as an adduct with tetra- cyanoethylene. '03 J. H. Hall J. W. Hill and J. M. Fargher J. Amer. Chem SOC.,1968,W 5313; A. Mishra S. N. Rice and W. Lwowski J. Org. Chem. 1968,33,481;J. E. Baldwin and R. A. Smith ibid. 1967,32,3506. A. L. J. Beckwith and J. W. Redmond J. Amer. Chem SOC. 1968,90 1351. lo5 A. L. J. Beckwith and J. W. Redmond Chem Comm. 1967 165. lo6 D.S. Breslow E. I. Edwards R. Leone and P. von R. Schleyer J. Amer. Chem. SOC. 1968,90 7097. R. A. Abramovitch and V. Uma Chem. Comm. 1968 797. Reaction Mechanisms 169 0"'"""'"' Nitrenes appear likely to become important synthetic intermediates as demonstrated recently lo8 by the reaction between porphyrins and ethoxy- carbonylnitrene (as shown on p. 170). Aziridines formed by reaction of nitrenes with olefines frequently isomerise to open chain amines during the reaction. With the aminonitrene (92)"' stable a'ziridines are formed by (stereospecific) reaction with a variety of olefins. A main difficulty when using nitrenes as synthetic intermediates lies in the elevated temperatures required to decompose the organic azide precursor.It is highly significant therefore that Dekker and Knox'" report that nitrenes result from metal carbonyl-catalysed decomposition of several precursors. Thus nonacarbonyldi-iron reacts rapidly at room temperature with azido- benzene methyl isocyanate and nitromethane to give the complex adducts (93H95) respectively. (co)~ ,Me\\\//./;;, /Fe (CO) 0 :grR R\yX > /R (CO),Fe --.Fe(CO) (CO)$e--(93) (C0)ff Fe(CO)3 (94) Me' (95) There would be obvious synthetic potential if this type of catalysis could be modified to allow trapping of the nitrene by independent reactants. lo' R. Grigg Chem. Comm. 1967 1238. log R. S. Atkinson and C. W. Rees Chem. Comm. 1967 1230. 'lo M. Dekker and G. R. Knox Chem. Comm. 1967 1243. 170 A.Ledwith