4 FREE-RADICAL REACTIONS By M. J. Perkins (Department of Chemistry King’s College Strand London W.C.2 ) THISis the first Annual Report for several years in which the organic chemistry of free radicals has been allocated separate space.’ In view of this and the independent publication recently of an annual survey of free-radical reactions which valiantly strives after comprehensive coverage as a prime objective,* it seems apposite to include in the present review some of the more significant developments in the subject during the past 2-3 years. The discussion will be restricted almost entirely to results obtained in the liquid phase. The enormous volume of research publication in this field has recently been paralleled by the growth of an extensive review literature and numerous books3 and reviews4 have appeared during 1968.One of the outstanding problems in radical chemistry concerns the preferred geometry of alkyl radicals. The bulk of experimental data can be accommodated equally well by a planar structure or by one which is pyramidal but rapidly inverting. One approach to this problem has been to investigate the ease of production of bridgehead radicals in cage molecules. For example from a ’ See G. H. Williams Ann Reports 1958 55 156. See B. Capon M. J. Perkins and C. W. Rees ‘Organic Reaction Mechanisms 1965 1966 and 1967,’ Interscience London and New York 1966 1967 and 1968. ‘Advances in Free-radical Chemistry’ ed. G. H. Williams Logos and Academic Press London and New York 1968 vol. 3; A. R.Forrester J. M. Hay and R. H. Thompson Organic Chemistry of Stable Free Radicals,’ Academic Press London and New York 1968; ‘Oxidative Coupling of Phenols,’ eds. W. I. Taylor and A. R Battersby M. Dekker New York 1967; ‘Yuriki no Kagaku’ (Chemistry of Free Radicals) eds. H. Sakurai and K. Tokumaru Nankodo Tokyo 1967; ‘Radical lons,’ eds. E. T. Kaiser and L. Kevan Interscience New York 1968; ‘Organosulphur Chemistry,’ ed. M. J. Janssen Interscience New York 1967; ‘Oxidation of Organic Compounds-1,’ ed. R. F. Gould Advances in Chemistry Series 75 Amer. Chem. SOC. Washington 1968. Introductory Review W. A. Pryor Chem Eng. News Jan. 15 1968 p. 70; Study of Reaction Intermediates by E.S.R. Proc. Roy. SOC. 1968,302 A 287-361; see also G. A. Russell Science 1968 161 423 N.M. Atherton Science Progr. 1968 56 179 Y. S. Lebedev. Uspekhi Khimii 1968 37 934 F. Gerson Chima 1968,22,293 and C. Thompson Quart. Rev. 1968,22,45; Aromatic Substitu- tion D. H. Hey Bull. SOC.chim. France 1968 1591; Radical cyclisation M. Julia and M. Maumy ibid. p. 1603; Reactions of Alkoxy Radicals C. Walling ibid. p. 1609 and K. Heusler Chimia 1967 21 557; Radical Ions M. Szwarc Prog. Phys. Org. Chem. 1968 6 323; Aromatic Radical Anions P. Rempp Bull. SOC. chim. France 1968 1605; Phenoxyl Radicals V. D. Pokhodenko V. A. Khiznyi and V. k Bidzilya Uspekhi Khimii 1968 37 998; Aspects of Autoxidation F. R. Mayo Accounts Chem. Research 1968; 1 193; Photolysis of Aryl Iodides R. K. Sharma and N. Kharasch Angew. Chem. Intern Ed. 1968 7 36; Anode Processes B.E. Conway Pure Appl. Chem. 1968 18 105 N. L. Weinberg and H. R Weinberg Chem. Rev. 1968 68 449 A. K. Vijh and B. E. Conway ibid. 1967 67 623; Biradicals I. D. Morozova and M. E. Dyatkina Uspekhi Khimii 1968 37 865; Tri- alkyltin Hydrides in Radical Reactions Accounts Chem. Research 1968 1 299; Nitro-aromatic Radical Anions E. Buncel A. R Norris and K. E. Russell Quart. Rev. 1968 22 144; Solvent Effects in Radical Abstraction H. Sakurai and A. Hosomi Yuki Gosei Kagaku Kyokai Shi 1967 25 1108; See also the reviews on stereoselection in organic reactions by S. I. Miller Adv. Phys. Org. Chem. 196%8 p. 185; K. Fukui Kyoto Daigaku Nippon Kagakusem Kenkyusho Koenshu. 1966 23 75. 172 M. J. Perkins series of experiments' with bridgehead carbaldehydes in which the relative stability of the radical R* was inferred from the competition between processes (1) and (2); the order of radical stability CCl, RCO* -RCOCl -co CCl RCO.-R* --% RCl adamantan-1-yl > t-butyl -1-bicyclo[2,2,2]octy1 9 norbornan-1-yl was found. Although it is clear from this that a bridgehead norbornyl radical is particularly unstable the remaining data are inconclusive and assumptions implicit in the interpretation of the competition between paths (1) and (2) may not be valid. More reliable information comes from three independent studies6 of perester (RC0,OBu') decomposition. Peresters with the four R-groups mentioned above were studied. All four appeared to decompose by a concerted two-bond scission RC020But + R* + C02 + *OBut After allowing for differences in inductive effects the rates of decomposition followed the order R = t-butyl > adamantan-1-yl > l-bicycl0[2,2,2]octyl > l-norbornyl(1~0,0~4,0~05,0~004). It was concluded that the preferred geometry of an alkyl radical is planar but that distortion from planarity imposes much less strain than is the case for carbonium ions.In contrast however is the observation that the decomposition of 1,l'-azoadamantane requires an activa- tion energy 20 kcal. greater than does 2,2'-azisobutane ! Hydrogen abstraction from adamantane also yields interesting information,* for whilst the bridgehead position is more prone than is the 2-position towards hydrogen abstraction the bridgehead radical (A') once formed appears to be unusually indiscriminate in subsequent reactions with halogen donors.R* + A'H -+ A'* + RH (3) A'. + RHal+ A'Hal + R* (4) In interpreting these results it was suggested that the transition state for reaction (4)might reasonably have much more adamantanyl-radical character than the transition state for reaction (3). Manifestation of any abnormality in the bridgehead radical due to conformation strain should therefore be more pronounced in the halogen transfer (4). D. E. Applequist and L. E. Kaplan J. Amer. Chem. SOC. 1965 87 2194. J. P. Lorand S. D. Chodroff and R W. Wallace J. Amer. Chem. SOC. 1968 90,5266; R C. Fort and R. E. Franklin ibid.,5267; L. B. Humphrey B. Hodgson and R E. Pincock Canad. J. Chem. 46,3099. ' M. Prochazka 0.Ryba and D.Lim Coll. Czech. Chem. Comm. 1968,33,3387;cf:W. Theilacker and K.-H. Beyer Chem Ber. 1961,94,2968. I. Tabushi J. Hamuro and R. Oda J. Amer. Chem. SOC. 1967 89 7127. Free-radical Reactions Two other interesting studies on radical conformation examine the 9-decalyl radi~al.~ A second stereochemical point which has aroused considerable recent interest is the geometry of vinyl radicals. Radical additions to acetylenes are well documented. lo However when stereoisomeric products are possible these are as a rule both observed. There may be stereoselection in the addition step leading preferentially to one or other vinyl radical but these invert more rapidly than the chain transfer (6)to give olefin. The sides of the equilibrium R'.+ HC=CR2 R' R2 >C* / H R1 R' H \ /R2 )c=c< ,C=C \ H H H R2 R' R2 \/ ,c=c \ (5) may be approached independently using appropriate radical precursors such as the peroxides (lc) and (It).From several studies of this kind" it has become clear that equilibration [equation (5)] is indeed rapid similar product mixtures being obtained from both precursors. The relative proportions of the olefinic products appear to be controlled largely by steric factors in the chain- transfer steps (6).'"In the discussion of these effects Kopchic and Kampmeier" were led to the conclusion that the P-phenylstyryl radical might differ from P. D. Bartlett R E. Pincock J. H. Rolston W. G. Schindel and L. A. Singer J. Amer. Chem. Soc. 1965,87,2590; F.D. Greene and N. N. Lowry J. Org. Chem. 1967,32 875 882. For a recent example see R. M. Kopchik and J. A. Kampmeier J. Amer. Chem. Soc. 1968,w 6733. l1 L. A. Singer and N. P. Kong J. Amer. Chem. Soc. 1966,88,5213,1967,89,5251; J. A. Kampmeier and R.M. Fantazier ihid.. 1966 88 5219 G. D. Sargent and M. W. Browne ibid. 1967. 89 2788 0.Sirnamura. K. Tokumaru and H. Yui Trrrohedron Lcqrm. 1966. 5141. 174 M. J. Perkins other vinyl radicals studied by preferring the linear structure (2);however this conflicts with some recent Japanese work in which considerable (>80%) retention of stereochemistry is observed in the stilbenes derived from decom- positions of both (lc) and (It) (R' = R2 = Ph) in cumene or chloroform at 6OO.l An attempt to generate acetylenic radicals' was unsuccessful but it did reveal a new mode of induced-peroxide decomposition Re + PhC=C*CO,OBu'+ PhC==CR*C02*OBu' + PhC=CR + C02 + *OBu' A similar process was demonstrated for cinnamoyl peroxides.Reactions involving the radical (3) on the other hand occur without rupture of the peroxide bond.14 Me Et Et dMe Et \Me Attempts to observe chemical effects of spin correlation in the geminate pairs of radicals produced in azo-compound decompositions have until recently been unsuccessful. Now Bartlett and his colleagues have discovered that attempts to produce triplet-radical pairs reported in the literature,' had systematically failed. This was because sensitized decompositions had proceeded exclusively by singlet-energy transfer.Triplet-energy transfer leads only to isomerisation about the nitrogen-nitrogen double-bond.16 In the cyclic azo-compound (4)[meso or (i-)I where such isomerisation is not possible triplet-sensitized photodecomposition gives a mixture of almost equal parts of meso- and (+)-cyclobutanes (3.''Thermolysis or direct photolysis on the other hand give cyclobutane in which there is substantial retention of the geometry of the precursor. This is entirely consistent with a much longer life- time for the biradical intermediate in the triplet process. Recent work which shows spin-correlation effects in 1,3-biradicals has also been described.' There is an extensive literature detailing kinetic isotope effects in azo-compound decompositions.Seltzer and Mylonakis have now correlated primary nitrogen-15 effects with data on secondary deuterium effects in the l2 K. Tokumaru in 'Yuriki no Kagaku' p. 123. (see ref. 3). l3 N. Muramoto T. Ochiai 0.Simamura and M. Yoshida Chem. Comm. 1968 717 l4 M. M. Schwartz and J. E. Leffler J. Amer. Chem. SOC. 1968,90 1368. l5 P. D. Bartlett and J. M. McBnde Pure Appl. Chem. 1967 15 89. P. D. Bartlett and P. S. Engel J. Amer. Chem. SOC.,1968,90,2960. l7 P. D. Bartlett and N.A. Porter J. Amer. Chem. SOC.,5317. l8 P. Scheiner J. Amer. Chem. SOC.,1968,90 988. Free-radical Reactions pyrolyses of azo-compounds which decompose by one-bond scission or by symmetrical or unsymmetrical two-bond scission.' Other secondary isotope effects have been reported,20 including the demonstration of an increased cage recombination of radicals from perdeuterioazisobutyronitrile compared with those from the non-deuteriated species.21 Cage effects in general have come in for close scrutiny in particular their variation with pressure and vis~osity.2~-~~ An analysis of the viscosity de- pendence was fitted by Pryor and Smith22 to rate data for the decomposition of acetyl peroxide25 and of p-nitrophenylazotriphenylmethanein a range of solvents.Both reactions had scission of only one bond in the rate-determining step; it was argued that cage effects in reactions exhibiting multiple-bond homolysis in the rate-determining step should not lead to reconstitution of the starting material. This is illustrated schematically single-bond scission R1N=NR2 [R'N=N* + R2] +products concerted two-bond scission R'N=NR2 + [R' + N2 + R2] +products Indeed the dependence of decomposition rate on viscosity could be utilised to distinguish between single- and multiple-bond homolyses.In very viscous solvents much more cage recombination occurs than the Pryor-Smith analysis would predict.26 In recent examples of disproportionation of geminate pairs of radicals in a solvent cage,24. 27 an example has been found in which asym- metry at the radical centre is retained in the disproportionation pr~duct.~' Several groups have sought and occasionally found evidence for neigh- bouring-group effects in homolytic reactions. Of course intramolecular (6) (7) S. Seltzer and S.G. Mylonakis J. Amer. Chem SOC.,1967,89 6586. 2o (a) S. E. Scheppele and S. Seltzer J. Amer. Chem. SOC.,1968 90 358; (b) S. G. Mylonakis and S. Seltzer ibid. 5487; S. Rummel H. Hubner and P. Krumbiegel Z. Chem. 1967,7 351. 21 S. Rummel H. Hubner and P. Krumbiegel Z. Chem. 1967 7 392. 22 W. A. Pryor and K. Smith J. Amer. Chem. SOC.,1967,89 1741. 23 C. Walling and H. P. Waits J. Phys. Chem. 1967 71 2361; 0.Dobis J. M. Pearson and M. Szwarc J. Amer. Chem. SOC.,1968,90,278; K. Chakravorty J. M. Pearson and M. Szwarc ibid. 283. 24 R C. Neuman and J. V. Behar Tetrahedron Letters 1968,3281. 25 Cage effects in acetyl peroxide decomposition have been discussed in detail J. W. Taylor and J. C. Martin J. Amer. Chem SOC.,1967,89,6904; see also J. C. Martin J. W. Taylor and E.H. Drew ibid. 129. " H. Kiefer and T. G. Traylor J. Amer. Chem SOC., 6667. " H. M. Walborsky and C. J. Chen J. Amer. Chem. SOC.,1967,89 5499. 176 M. J. Perkins reactions are well known in radical chemistry as for example in functionalisa- tion of steroidal angular methyl groups in the Barton reaction and related processes.2 However kinetic effects which actually assist initial homolysis are less well documented. Examples in which neighbouring double-bonds appear to facilitate peroxide2' and hypochlorite3* decompositions have appeared ; neighbouring sulphur too can assist peroxide decomp~sition,~ though no evidence for intramolecular stabilisation by sulphur could be found in radical (6).32 Reaction of both syn-and anti-7-bromonorbornene with tributyltin deuteride gives ~nti-7-deuterionorbornene.~ This stereospecificity has led to the postula- tion of a possible nonclassical intermediate (7).However alkyl bridging is not normally encountered in radicals as there is no low-lying orbital available to accommodate the unpaired electron. For this reason 1,2-alkyl shifts are almost unknown in radical chemistry. Vinyl migrations,34- 35 on the other hand like phenyl migrations are well documented. The simplest case the rearrangement of allylcarbinyl itself has been probed by isotopic labelling. Experiments 28 For recent examples see E. Wenkert and B. L. Mylori J. Amer. Chem SOC.,1967,89 174; D. H. R Barton R H. Hesse R. E. O'Brien and M. M. Pechet J. Org. Chem. 33 1562; J. E.Baldwin D. H. R Barton I. Dainis and J. L. C. Pereira J. Chem SOC.(C),1968 2283. 29 R C. Lamb L. P. Spadafino R G. Webb E. B. Smith W. E. McNew and J. G. Pacifici J. Org. Chem. 1966,31 147. 'O J. M. Surzur P. Cozzone and M. P. Bertrand Compt. Rend. 1968,267 C 908. 31 T. H. Fisher and J. C. Martin J. Amer. Chem SOC.,1966,88 3382. 32 J. S. Hyde R Breslow and C. DeBoer J. Amer. Chem. SOC.,1966,88,4763. 33 J. Warkentin and E. Sanford J. Amer. Chem SOC.,1968,90,1667. 34 See for example C. K. Alden D. I. Davies and P. J. Rowley J. Chem. SOC.(C) 1968 705 and preceding papers in this series; D. C. Neckers Tetrahedron Letters 1965 1889; L. H. Slaugh J. Amer. Chem SOC.,1965 87 1522; T. A. Halgren M. E. H. Howden M. E. Medof and J. D. Roberts ibid. 1967 89 3051; S.J. Cristol and R V. Barbour ibid. 1968 90 2832; I. S. Lishanskii A. M. Guliev A. G. Zak 0.S. Fomina and A. S. Khachaturov Doklady. Akad. Nauk SSSR 1966,170,1084. 3s L. K Montgomery J. W. Matt and J. R Webster J. Amer. Chem SOC.,1967 89 923; L. K. Montgomery and J. W. Matt ibid. p. 934 3050. Free-radical React ions 177 involving the radical-chain decarbonylation of the aldehyde (8) confirmed that rearrangement was taking place and the isolation of both cis-and trans-[l-2H]but-l-ene from decarbonylation of (9) implies that the cyclo- propylcarbinyl intermediate (10) is sufficiently long-lived for rotation about the carbon-carbon bond to occur.35 An oxygen analogue of the cyclopropyl- carbinyl to allylcarbinyl rearrangement implicit in the above vinyl migrations has been noted [( 11)+ ( and similar behaviour is found in the decompo- sition of cyclopropyl nitrites [e.g.(13)].37 In these reactions however great instability coupled with the effects of substituents point to rearrangement being concerted with the decomposition. At the other extreme examples have come to light of stabilization of radicals by cyclopropyl substituents which do not ring-open in the course of reaction.38 Me/*\ -Me CH=CH Me CH; (13) (nitroso-dimer) Radical bridging is also possible where the bridging group contains a low-lying d-orbital which can accommodate the unpaired electron as for example in the bromine-bridged structures encountered when bromine atoms add to a double bond.39 Evidence for bridging by sulphur4' and tin4' atoms and a 1,Zshift of a silyl can similarly be rationalised in terms of d-orbital participation.Fragmentation of alkoxy-radicals is well known.43 However the reverse process radical 'addition to a carbonyl group is seldom observed. Examples have been encountered in a further analogue of the allylcarbinyl rearrange- ment,44 in additions to perfluoroalkyl ketones,45 and more recently in the 36 S. K. Pradhan and V. M. Girijaballabhan Tetrahedron Letters 1968 3103. 31 C. H. DePuy H. L. Jones and D. H. Gibson J. Amer. Chem. SOC. 1968,90,5306. J. C. Martin J. E. Schultz and J. W. Timberlake Tetrahedron Letters 1967,4629. 39 See for example P. D. Readio and P. S. Skell J. Org. Chem. 1966 31 753 759; for bromine bridging in a radical elimination see D.M. Singleton and J. K Kochi ibid. 1968 33 1027; J. Amer. Chem SOC.,1968,W 1582; Tetrahedron 1968,24 3503. 40 H. H. Szmant and J. J. Rigau Tetrahedron Letters 1967 3337; N. A. Lebel and A. DeBoer J. Amer. Chem. SOC.,1967,89 2784. *' R H. Fish H. G. Kuivila and I. J. Tyminski J. Amer. Chem. SOC.,1967,89,5861. 42 C. G. Pitt and M. S. Fowler J. Amer. Chem. SOC. 1968,90 1928. 43 A recent discussion is given by K. Maruyama and K. Murakami Bull. Chem. SOC.Japan 1968 41 1401. 44 W. Reusch C. K. Johnson and J. A. Manuer J. Amer. Chem. SOC.,1966,88 2803. 45 E. G. Howard P. B. Sargeant and C. G. Krespan J. Amer. Chem. Soc. 1967 89 1422 178 M. J. Perkins reactions of biacetyl as in the synthesis of acetylcyclohexane outlined below.46 0.There have been numerous investigations of polar effects in radical reactions particularly in hydrogen abstraction^.^^-'^ The selectivity exhibited in reac- tions of aminium cation radicals is particularly dramatic. For example radical chlorination of methyl decanoate with N-chlorodialkylamines in strongly acidic media occurs predominantly (ca. 50%) at C-9. Attack by the aminium cation radical is clearly directed away from the protonated ester function by charge repulsion ; however a similar preference for attack at the penultimate carbon atom found with long-chain hydrocarbons may reflect chain-coiling in the highly polar medium; this would render the internal methylene units relatively inac~essible.~~ The ease of chlorination of substituted toluenes under comparable conditions shows an unexpectedly small substituent de- pendence with p = -1.36 (against a+).48This value for p is comparable with the figure for abstraction by Br*.Hydrogen abstraction from 1-substituted adamantanes by trichloromethyl radicals occurs predominantly (ca. 80 %) from the remainining bridgehead positions. By means of competition studies it was possible to determine variations in bridgehead reactivity as a function of the s~bstituent.~~ In this the first study of polar effects on radical reactions in a rigid aliphatic system correlation with Taft a*-values was excellent (p* = -0.4). The importance of bridged intermediates in radical additions has already been noted. Another topic of current interest in radical-addition reactions is the extension to ring synthesis by intramolecular addition.” 52 In particular Julia’s group has recently succeeded in closing two rings by a combination of radical addition and aromatic substitution [( 141 +(15)]?’ The first cyclisation the addition of the resonance-stabilised cyanoacetate radical to the double- bond is reversible.This allows formation of the thermodynamically preferred six-membered ring. Under conditions of kinetic control cyclopentane forma- tion is commonly preferred.’ Two somewhat different conformational arguments have now been advanced to rationalise this. In the first,54 trans- annular overcrowding is considered responsible. In the second,” a stereo- 46 W. G. Bentrude and K. R Darnall J. Amer. Chem. SOC.1968,90,3588;Chem. Comm. 1968,810. 47 R Bernardi R Galli and F. Minisci J. Chem. SOC. (B) 1968 324. 48 R S. Neale and E. Gross J. Amer. Chem. SOC. 1967,89,6579. 49 P. D. Owens G. J. Gleicher and L. M. Smith J. Amer. Chem. SOC. 1968,90,4122. 50 R. D. Gilliom and J. R Howles Canad. J. Chem. 1968 46 2752; G. J. Gleicher J. Ory.Chem. 1968,33 332; E. S. Huyser and K. L. Johnson ibid. 3952; K. H. Lee Tetrahedron 1968,24,4793. 51 M. Julia and J. C. Chottard Bull. SOC. Chim. France 1968 3691 3700. ’* A recent example is given by J. I. G. Cadogan M. Grunbaum D. H. Hey A. S. H. Ong and J. T. Sharp Chem. and Ind. 1968,422; see also refs. 30 and 55. 53 E.y.B. Capon and C. W. Rees Ann Reports 1964 61 261. 54 M. Julia and M. Maumy see ref. 4. ” D. L. Struble A.L. J. Beckwith and G. E. Gream Tetrahedron Letters 1968? 3701. Free-radical Reactions 179 electronic requirement for approach of the adding radical perpendicularly to one end of the double-bond is assumed; this is shown to be more favourable for cyclopentane formation. The cyclisation of farnesyl acetate by benzoyl peroxide and copper ions,56 possibly typifies a synthetically useful procedure ;however in the opinion of the writer the assignment of a radical mechanism to the cyclisation steps of this reaction is open to question. d;'a/cH2 (PhCO,& OAC f Cu'/Cu" Ph CO "H Intramolecular aromatic substitution by the copper-catalysed decomposition of appropriate diazonium salts has long been considered to involve radical intermediates.However compelling evidence'' has only recently been forthcoming. For example,58 the formation of the dimer (18) from (16) can reasonably be envisaged only as following a radical pathway. Cuprous oxide has proved to be a very much more eficient catalyst for effecting similar cyclisations ;furthermore conditions have been found for diverting the radical either by oxidation to phenol or by hydrogen abstraction from the solvent both in excellent yield." Pyrolysis of the dimer (1 8) gives N-methylphenanthridone almost quan- titatively. Presumably this involves dissociation back to (17) followed by rearrangement to (19) and loss of hydrogen.60 The behaviour of derivatives of (18) clearly demonstrates that an aryl group migrates rather than nitrogen.This rearrangement may have a bearing on the mechanism of intermolecular aromatic substitution by aryl radicals. It is known that competition between pathways (7) and (8) open to the cyclohexadienyl radical (20) is essentially independent of the position of the substituent X.Thus if the yizld of biaryls 56 R Breslow S. S. Olin and J. T. Groves Tetrahedron Letrers 1966 4717; ibid. 1968 1837. " R A.Abramovitch and k Robson J. Chem. SOC.(C),1967 1101. " D.H. Hey C. W. Rees and A. R. Todd J. Chem. SOC.(C) 1518. 59 A. H.Lewin and T. Cohen J. Org. Chem. 1967,32 3844. 6o D.Collington D. H. Hey and C. W.Rees,J. Chem SOC.(C) 1968 1017 1026; D.Collington D. H. Hey C. W. Rees and E. le R. Bradley ibid. 1021. 180 M. J. Perkins 9px 4 \o O5 \/ h r-5 4 -", 0 s \ Free-radical Rwetions 181 Ar (7) X I x + other products (21) is increased at the expense of other products by carrying out the reaction in the presence of an oxidising agent [favouring path (7)] the isomer distribu- tion of the biaryls is unaffected.61 One possible explanation of this (a priori unlikely) result could involve a series of relatively rapid intramolecular aryl shifts [equation (9)].X X X The reaction of benzoyl peroxide with benzene has often been quoted as the typical example of radical substitution. It now seems that it may in fact be somewhat exceptional. In this reaction bimolecular interaction of two cyclohexadienyl radicals [path (8) above] is very important. However this does not occur with other familiar phenylating agents because of the build-up in the reaction system of relatively high concentrations of some oxidising species which efficiently divert the cyclohexadienyl radicals to biaryls or other products with phenylazotriphenylmethane it is the trityl radical,62 with N-nitrosoacetanilide it is the nitroxide (22),63and a nitroxide assumes the same role even with benzoyl peroxide if a nitroaromatic is present.64 Ph*+ N=O PhNO-__L -I I PhNAc PhNAc (22) 61 R T.Morrison J. Cazes N. Samkoff and C. A. Howe J. Amer. Chem. SOC.,1962 84 4152; D. H. Hey M. J. Perkins and G. H. Williams Chem and Ind. 1963 83; D. H. Hey K. S. Y. Liang and M. J. Perkins Tetrahedron Letters 1967 1477; however see H. J. M. Dou G.Vernin and J. Metzger ibid. 1968 953. M. J. Perkins J. Chem. SOC.,1964,5932. 63 G. R Chalfont and M. J. Perkins J. Amer. Chem SOC.,1967,89 3054; A. R Forrester Chem. and Id. 1968 1483; for new data on the rearrangement of nitrosoacylarylamines see C. Ruchardt C. C. Tan and B. Freudenberg Tetrahedron Letters 1968,4019. 64 G. R Chalfont D. H. Hey K. S. Y.Liang and M. J. Perkins Chem Comm. 1967 367. 182 M. J. Perkins The homolysis of benzoyl peroxide in benzene initially produces two benzoyloxy-radicals. It has only recently been fully appreciated that these may add reversibly to the benzene before suffering decarboxylation though an excellent analogy has been in the literature for many years.65 The benzoyl- oxycyclohexadienyl radical can be intercepted by oxygen66 or copper (11) ions67 (to give phenyl benzoate) or by coupling with trityl radicals.68 In the last example,.subsequent elimination of benzoic acid gives tetraphenylmethane (the Wieland tritylation reaction).Ever since Gomberg’s pioneering work nearly 70 years ago the hexaphenyl- ethane structure has been written for the dimer of the trityl radical. This now appears to have been mistaken evidence for the para-coupled alternative (23) having been ~ecured.~’ This type of structure rationalises the steric effects Ar,CH*CHAr (24) I2000 Me displayed by meta-substituents on the dimerisation of hindered diarylmethyl radical^.^' In the same work the normal tetra-arylethane (24) was shown to possess high kinetic stability.However (24) when heated to 200” and then cooled gave an equilibrium mixture of (25)and (26). The field of nitroxide radical chemistry is growing so rapidly as almost to merit a review of its own. Indeed excellent cover.age is available in the volume on stable radical^.^ The relative stability of these radicals coupled with an in- creasing range of reactions whereby they may be produced means that they are particularly well suited for spectroscopic study. They are also however turning up as key intermediates in a growing list of homolytic reactions (e.g. refs. 63 and 64).Nitroxides are readily obtained by radical addition to nitrones” 65 D. B. Denney and P. P. Klemchuk J. Amer. Chem SOC.,1958,80 3289. 66 T. Nakata K. Tokumaru and 0.Simamura Tetrahedron Letters 1967,3303.67 M. E. Kurz P. Kovacic A. K. Bose and I. Kugajevsky J. Amer. Chem Soc. 1968,90 1818; M. E. Kurz and P. Kovacic J. Org. Chem 1968,33 266 1950. 68 T. Suehiro A. Kanoya T. Yamauchi T. Komori and S. I. Igeta Tetrahedron 1968,24 1551. 69 H. Lankamp W. T. Nauta and C. MacLean Tetrahedron Letters 1968 249. ’O W. Theilacker and F. Koch Angew. Chem. Internat. Edn. 1966,5 246. ’’ M. Iwamura and N. Inamoto Bull. Chem SOC. Japan 1967,40 703. Free-radical Reactions 183 and to C-nitroso-compounds.7 The latter reaction has been ingeniously applied to probing y-radiation damage in crystals by allowing the irradiated solid to dissolve in a solution of 2-methy1-2-nitro~opropane.~~ Radical fragments from the crystal were trapped by the nitroso-compound to give a nitroxide the e.s.r spectrum of a solution of which was easily interpreted and the structure of the fragment inferred.Use of the same nitroso-compound has been suggested for mechanistic For example in styrene polymerisation by varying the scavenger concentration it was possible to intercept as nitroxides initiator radical 1:1-adduct of this with styrene and growing polymer radicals. These nitroxides were easily distinguishable by their e.s.r. spectra. Use of nitrones for somewhat similar purposes has also been pr~posed.~’ Another e.s.r. technique holds great promise for direct observation of many reactive radicals in solution high-intensity U.V. irradiation of solutions of t-butyl peroxide in hydrogen-donor solvents at low temperatures has been found to give excellent spectra of solvent-derived radicals.76 The nitronylnitroxides (27),77 and the radicals (28)78 and (29)79provide new examples of particularly stable free-radicals and a revised structure (30) has been demonstrated for Banfield and Kenyon’s radical.80 0 Ph I (30) O‘ There has recently been an extensive literature on reactions involving both radicals and compounds of metals which show variable valence.For example oxidative decarboxylations of carboxylic acids (RC0,H) by lead tetra-acetate are though perhaps not invariably,82 believed to proceed uia radicals (R 0). The reactions ofthis oxidant with a wide range of other compounds have also been e~amined.8~ In many instances of carboxylic acid oxidation 72 E.g.A. Mackor,T. A. J. W. Wajer andT. J. de Boer Tetrahedron 1968,24,1623;G. A. Abakumov and G. k Razuvaev Doklady Akad. Nauk. SSSR 1968,187,95. 73 E. Lagercrantz and S. Forshult Nature 1968,218 1247. 74 G. R Chalfont M. J. Perkins and k Horsfield J. Amer. Chem SOC.,1968,90 7141. ’’ E. G. Janzen and B. J. Blackburn ibid. 5909. 76 P. J. Krusic and J. K. Kochl J. Amer. Chem. Sac. 1968,90 7155; J. K. Kochi and P. J. Krusic ibid. p. 7157; J. Q. Adams ibid. p. 5363. 77 J. H. Osiecki and E. F. Ullman J. Amer. Chem. SOC.,1968,90,1078; D. G. B. Boocock R. Darcy and E. F. Ullman ibid. p. 5945; D. G. B. Boocock and E. F. Ullman ibid. p. 6873; k T. Balaban P. J. Halls and A. R Katritzky Chem and Ind. 1968,651;see also L. B. Volodarsky G.A. Kutikova R Z Sagdeev and Y.N. Molin Tetrahedron Letters 1968 1065. ” H.M. Blatter and H. Lukaszewski Tetrahedron Letters 1968 2701. 79 A. T. Balaban P. T. Frangopol M. Frangopol and N. Negoita Tetrahedron 1967,23,4661. R. Foster J. Iball and R. Nash Chem. Comrn. 1968 1414. J. K. Kochi J. Amer. Chem. SOC.,1965,87 3609; J. K. Kochi J. D. Bacha and T. W. Bethea ibid. 1967,89 6538; D. I. Davies and C. Waring J. Chem. SOC.(C),1968 1865. 82 D. I. Davies and C. Waring J. Chem. SOC.(C),1968 2332 2337. 184 M. J. Perkins the intermediate radical may be intercepted by copper@) in a process of oxida- tive eliminati~n.~~ Thus cyclobutanecarboxylic acid is efficiently oxidised by lead tetra-acetate and copper(r1) ions in acetic acid to give cyclobutene. The elimination appears to involve concerted collapse of a cyclobutylcopper species.Similar behaviour is found for the Cu" oxidation of cyclobutyl radicals formed in the Cu'-induced decomposition of bis(cyclobutylformyl)peroxide? Oxidative bis-decarboxylation of a vicinal dicarboxylic acid to an olefin commonly effected by lead tetra-acetate may often be carried out in superior yield by anodic oxidation.86 In this brief survey an attempt has been made to highlight some of the areas of recent or current interest in free-radical chemistry. In such a short space many topics have inevitably not been touched upon. One of these of immense commercial significance is autoxidation. This is mentioned now because a feature evident in more recent radical chemistry is the increasing proportion of work in which absolute rate constants for radical reactions have been esti- mated:' and this exemplified in many studies related to autoxidation notably by Howard and Ingold and their colleagues.88 One final point on autoxidation concerns chain-termination.It has been known for some time that chain termination between pairs of primary or secondary peroxy-radicals is much faster than with tertiary ones. A mechanism proposed by Russell8' to accommodate this involving the electrocyclic process indicated should generate singlet oxygen. That this indeed occurs has been demonstrated by a successful trapping experimentg' 0,(Singlet) 2R2CHO0. R2Co HOCHR CHR 83 E. I. Heiba R M. Dessau and W. J. Koehl J. Amer. Chem SOC.,1968,90 1028 2706; W.H. Starnes ibid. p. 1807; R 0.C. Norman and C. B. Thomas J. Chem SOC.(B) 1967 771; 1968 994; B. C. Gilbert and R 0.C. Norman ibid. 1968 123; R 0.C. Norman and R A. Watson ibid. p. 184; W. H. Starnes J. Org. Chem. 1968 33 2767; G. Just and K. Dahl Tetrahedron 1968 24 5251 ;J. Lhomme and G. Ourisson ibid. pp. 3167 3177 3201; M. L. Mihailovic Z Cekovic V.Andrejevic, R Matic and D. Jeremy ibid. p. 4947; J. B. Aylward and R 0.C. Norman J. Chem. SOC.(C) 1968 2399. 84 J. D. Bacha and J. K. Kochi Tetrahedron 1968,24,2215;J. Org. Chem. 1968,33,2746. 85 J. K. Kochi and A. Bemis J. Amer. Chem SOC. 1968,90,4038; see also J. K. Kochi A. Bemis and C. J. Jenkins ibid. p. 4616. 86 P. Radlick R Klem S. Spurlock J. J. Sims E. E. van Tamelen and T. Whitesides Tetrahedron Letters 1968 5117; H.H. Westberg and H. J. Dauben ibid. p. 5123. 87 A useful compendium of estimated absolute rate constants was given by D. F. DeTar J. Amer. Chem. SOC. 1967,89,4058. ** See for example J. A. Howard and K. U. Ingold Cad. J. Chem. 1967,45 785 793; 1968,46 2655 2661; J. A. Howard K U. Ingold and M. Symonds ibid. 1968,46 1017. 89 G. A. Russell J. Amer. Chem. SOC.,1957 79 3871. Free-radical Reactions Other topics not reviewed here include the chemistry of the monovalent carbon species (:&O,Et) generated recently,” y-radiolysis studies radical ions electron-transfer processes solvent .effects and many aspects of the appli- cation of magnetic resonance techniques. It is hoped that emphasis on some of these topics may be possible in future reports.J. k Howard and K. U. Ingold J. Amer. Chem SOC. 1968,90,1056,1058. 91 T. DoMinh H. E. Gunning and 0.P. Strausz,J. Amer. Chem SOC.,1967,89,6785;0.P.Straw T. DoMihn and J. Font ibid. 1968,90 1930.