5 Free Radicals By W. T. DIXON Chemistry Department Bedford College Regent 3 Park London NW? 4NS 1 Kinetics of Free-radical Reactions There are two approaches to the measurement of kinetics in solution one direct in which the appearance and/or disappearance of species is measured as a function of time and the other indirect in which reactions are conducted competitively and relative rates are deduced from the pattern of products. The object of such studies is generally rather more to elucidate reaction mechanisms than to deter- mine individual reaction rates although there may be something of intrinsic interest in the latter. There is still muchactivity in the investigation of the reactions of phenyl radicals which may be generated in a very 'clean' way by the y-radiolysis of diazonium ions ArN,' 'zZ?p Are + N Subsequent reactions of these radicals can then be followed indirectly without the complication of molecular species required to initiate the formation of the phenyl radicals.The rate of disappearance of the diazonium ion (measured photometrically) as a function of dose rate can be used to indicate the mechanism of the overall reaction for the reduction of the diazonium salt to the parent hydrocarbon ArN,+ + AH -+ ArH + A + N + H+ The rate of hydrogen abstraction by the aryl radical is in the order hypophos- phite phosphite =-formate > methanol > formic acid. A chain reaction is in-volved thus generating further aryl radicals Ar. + AH + ArH + AH. ArN,' + AH. + Ar- +A + N + Hi The free-radical phenylation of perfluoroaromatic compounds has been studied with increasing intensity' and the results of classical kinetic studies and of product analyses seem to have shown that the reactions of these perfluorinated compounds are not strictly analogous to those of the corresponding hydrogen compounds.However the analogy is pressed home as far as possible and it seems fairly definite that there is an intermediate '0-complex' of the cyclohexadienyl type. J. E. Packer R. K. Richardson P. J. Soole and D. R. Webster J.C.S. Perkin II 1974 1472. R. Bolton J. P. B. Sandall and G. H. Williams J. Fluorine Chem. 1974 4 347 355. I35 136 W. T. Dixon *aF F F--‘F (1) One intriguing result is the occurrence of small quantities of hexafluorobiphenyl in the products of the reaction of phenyl radicals with hexafluorobenzene.This is analogous to results observed using hexadeuteriobenzene as the substrate3 and is rationalized in terms of a novel exchange in the intermediate (Scheme I). 2 Kinetics Using E.S.R. Although the slow decay of relatively long-lived species such as hindered phenoxyl radicals4 is relatively straightforward in kinetic work mainly ‘steady state’ concentrations of radicals are observed enabling relative rates of generation and termination reactions to be determined. This is true whether the steady-state concentrations are generated in a flow ~ystem,~ electrolytically,6or photolytical-Absolute rate constants can often be calculated from steady-state ex- periments provided that one of the rates involved is known but to do this directly it is necessary to disturb these conditions sufficiently e.g.by means of a rotating sector stopped flow flash photolysis constant current pulse etc. each of which may lead to a meaningful extrapolation to the start of the reaction (when the rate of radical decay is zero !). The photolysis of di-t-butyl peroxide and other peroxides has provided much of the focus in kinetic applications of e.s.r. spectroscopy. On the whole reactions ’E. K. Fields and S. Meyerson J. Amer. Chem. SOC.,1967 89 3224. R. D. Parnell and K. E. Russell J.C.S. Perkin II 1974 161. B. C. Gilbert R. 0.C. Norman and R. C. Sealy J.C.S. Perkin II 1974 824. I. R. Goldberg and A. J. Bond J. Phys. Chem..1974 78 290. ’A. G. Davies R. W. Dennis and B. P. Roberts J.C.S. Perkin 11 1974 1101. D. Griller B. P. Roberts A. G. Davies and K. U. Ingold J. Amer. Chem. SOC.,1974 96 554. D. Griller G. D. Mendenhall W. Van Hoof and K. U. Ingold J. Amer. Chem. SOC. 1974 96 lo V. Malatesta and K. U. Ingold J. Amer. Chem. SOC.,1974 96 3949. I* G. D. Mendenhall D. Griller D. Lindsay T. T. Tidwell and K. U. Ingold J. Amer. Chem. SOC.,1974,96 244 1 D. Lal D. Griller S. Husband and K. U. Ingold J. Amer. Chem. SOC..1974 96 6355. Free Radicals 137 which already have been observed have been quantified such as the dimerization of phosphorus-centred radicals,8 the disproportionation of sterically hindered alkyl radicals,' and the cyclization of the hex-5-enyl radical (2).(2) The more detailed knowledge gained by kinetic investigations shows for example that alkyl radicals leave more easily from an apical than from an equator- ial site of a phosphoranyl radical (RO),PR + (RO),PR + R-Both iminyl' and 1,l-dialkylhydrazyl radicals lo generated by t-butoxyl oxidation decay by dimerization though in the latter case there may be a radical equilib- rium as with hexaphenylethane BuO. -k RC=NH -+ R,C=N.-+ dimer 2R,NNH G R,NNHNHNR L R,NNH + R,N=NH A further important technique in kinetics is to make measurements or conduct experiments first in the presence of scavengers i.e. radical traps and secondly in their absence. This approach has made it possible to find the propagation and termination rate constants in the autoxidation of some hydrocarbon^'^ and to study the rate ofreaction oftriplet carbonyl compounds with some alcohols.14 3 Alkyl Radicals Some further interesting studies on conformation have appeared ;in particular the effect of P-substituents in the ethyl radical appears to lead to a definite preference between conformations (3a) and (3b).15916 When R = Bu' confor-mation (3b) is invariably preferred but when R = H (3a) is preferred for (34 (3b) l3 J.E. Bennett J. A. Eyre and R. Summers J.C.S. Perkin [I 1974 797. l4 P. B. Ayscough and R. C. Sealy J.C.S. Perkin II 1974 1402. D. Griller and K. U. Ingold J. Amer. Chem. Soc. 1974,96 6715; K. S. Chen and J. K. Kochi ibid. p. 794. R. V. Lloyd D. E. Wood and M. T. Rogers J. Amer. Chem. SOC.1974,96 7130. 138 W. T. Dixon X = MR, where M is in the first row of the Periodic Table (e.g. carbon). On the other hand if M is from rows 2,3 or 4 of the Periodic Table (3b) is preferred,' perhaps owing to the size of such groups. In contrast to this when R = Me and X = C1 the eclipsed conformation is preferred,16 whereas when R= Me and X = Br the favoured conformation is (3a). As one expects the stable confor- mations are observed when the speed of internal rotation is low (low temperature) and the usual cos' 8 rule is invoked for the P-proton coupling constants. An interesting attempt to put forward a formula for the angular dependence of P-I3C coupling constants has involved the use of the n.m.r. contact shifts in appropriate nickel complexes of aromatic amines.'' Analysis of the e.s.r. spectra of five- and six-membered-ring radicals'* over the temperature range 545"Chas led to the conclusion that whereas five-membered rings rapidly undergo interconversion between two 'half-chair' conformers the six-membered rings flip relatively slowly from one chair form to the other. The phenacyl radical' has been generated and trapped in matrices at low or high temperatures and the formation of benzyl during its decomposition sug-gests an intricate rearrangement mechanism 4 My1 Radicals Ally1 radicals have always been found to be remarkably resistant to rotation about either C-C bond. However a number of examples have been found where such an isomerization must occur (Scheme 2). Its occurrence has been deduced from product analyses" and it has been observed directly by means of e.s.r.spectroscopy. 9' Scheme 2 l7 G. R. Underwood and H. S. Friedman J. Arner. Chern. SOC.,1974,96,4989; L. M. Stock and M. R. Wasielewski ibid. p. 583. B. C. Gilbert R. 0. C. Norman and M. Trenwith J.C.S. Perkin ZZ 1974 1033. 19 P.H. Kasai D. McLeod,jun. and H. C. McBay J. Amer. Chern. SOC.,1974,96 6864. *O P. S. Engel A. I. Dalton and L. Shen J. Org. Chem. 1974 39 2607; R. M. Hoyle and D. B. Denney ibid. p. 2607; W. P. L. Carter and D. C. Tardy J. Phys. Chem. 1974 78 2201. 21 R.Sustman and H. Trill J. Amer. Chem. SOC.,1974 96 4343. 22 B. E. Smart P. J. Krusic R. Meakin and R. C. Bingham J. Amer. Chem. SOC.,1974 96 7382; see also ibid. p. 621 1.Free Radicals 139 5 Radicals Derived from Aliphatic Alcohols Alkoxy-radicals abstract hydrogen atoms attached to a saturated carbon atom but the presence of a sterically crowded group in an alcohol is liable to increase the reactivity of the hydroxylic hydrogen towards alkoxy- radical^.^ 3y24 Thus compare CH,OH + RO. -+ *CH,OH + ROH Bu'CH,OH + RO-+ Bu'. + ROH + CH,O The observation by e.s.r. of the hindered (or strongly delocalized) radical suggests effective abstraction of the hydroxylic proton which can be inhibited by the presence of fluoride ion with which it forms strong hydrogen bonds R'CH20H + -OR2 -+ R'CH,O. + HOR' L CH20 + R'. Whether or not this whole process is concerted remains to be established since even if it did have an independent existence the lifetime of the radical RCH20-would be too short for it to be observed by trapping.Related to these processes is the thermal decomposition of heavily substituted alcohols25 such as tri-t-butyl- methanol which because of the considerable steric strain within the molecule will fragment homolytically BU:COH + BU. + BU,COH -+ (BUH + B~,CHOH+ B~,c=o) products Similarly results have been obtained from studies of the decomposition of hypochlorites in which the chlorine atoms are trapped by halogen-substituted olefins :26 RCH20CI -+ RCH,O. + C1-3 XCl 1 R. + CH,O There is a marked polar effect in the @-scission of the alkoxy-radical. 50 % yield Scheme 3 23 I. H. Elson and J. K. Kochi J. Org. Chem. 1974,39 2091.24 D. Griller and K. U. Ingold J. Amer. Chem. Soc. 1974 96 630. " J. S. Lomas and J. E. Dubois J. Org. Chem. 1974 39 1776. 26 C. Walling and R. T. Clark J. Amer. Chem. SOC.,1974 96 4530. 140 W. T. Dixon Further examples of 1,4-hydrogen atom shifts27 are found when alkyl hydro- peroxides react with ferrous ion in the presence of cupric acetate (Scheme 3). Such a reaction would be unlikely if the chain were rigid because the formation of the six-membered transition state (which necessitates the S-hydrogen being close to the oxygen) would not be possible. The reaction between organometallic compounds and peroxides has hitherto been believed to be a heterolytic process but product studies and the observa- tion of CIDNP show that radicals are involved (Scheme 4).*’ RCH,CH,MgBr + Bu0,Bu + BuOMgBr +{::::cHz*} polarization step JL RCH,CH,OBu RCH,=CH + BuOH Scheme 4 More comprehensive studies of the interaction of radicals with maleic acid2’ and with allylic compounds30 have been carried out.In the latter case addition to the double bond is invariably preferred to proton abstraction leading to a delocalized allyl-type radical. In many of the adduct radicals long-range coupling constants are observed which are unusually large for non-rigid systems. However by suitably simplified c~mputation,~~ allowing for averaging over the possible conformations it seems that whereas 6-splittings arise from extended hyper- conjugation (of the type implied by the ‘W’ rule) E-splittings are due to a direct interaction between the &-proton and the odd electron cloud on C (4).From a rather different viewpoint the contact shifts of y-and 6-protons have been ex- plained also in terms of a generalized hyperconjugation (5).31 However whereas 6-couplings are invariably positive y-splittings may be positive or negative depending on conformation (6).31-3 This could be the explanation for the absence of y-couplings in some cases where they might have been expected2’ and can be rationalized in valence-bond terms. 2’ Z. Cekovic and M. K. Green J. Amer. Chem. SOC.,1974 96 3000. 28 W. A. Nugent F. Bertini and J. K. Kochi J. Amer. Chem. SOC.,1974 96 4945. 29 W. T. Dixon J. Foxall and G. H. Williams J.C.S. Furuduy II 1974 70 1614. 3o P. Smith R. A.Kaba and P. B. Wood J. Phys. Chem. 1974 78 117. ” G. R. Underwood and H. S. Friedman J. Amer. Chem. SOC.,1974,96,4089. 32 N. L. Bauld and F. R. Farr J. Amer. Chem. SOC.,1974 96 145. 33 G. A. Russell and A. Mackor J. Amer. Chem. SOc. 1974 96 5633. Free Radicals 141 zero spin (5) hyperconjugation (6) homohyperconjugation dependent on conformation A series of papers3L37 has appeared on semidiones RC(O* )=C(O-)R. Acyclic 1,2-semidiones can exist in cis-and in trans-form;34*35the former may be stabilized somewhat by chelation on to a metal cation although the trans con-figuration is still generally preferred. One of the interesting aspects which has appeared in the study of bicyclic aliphatic semidi~nes~~~~~ concerns some clear examples of valence isomerization (Scheme 5).38 In these radicals the odd 0-0 Scheme 5 electron occupies an orbital which is antisymmetric with respect to reflection through the plane of symmetry of the molecule and so we would expect very small if any splittings from the syn-and anti-protons.However a simple two-step model can be used to rationalize the observed splittings. First there is a spin polarization leading to negative spin density in orbital A and secondly there is hyperconjugation (7) of a similar type to that in the vinyl radical. The observed ratio leads one to expect a bridgehead angle of ca. 120" as in vinyl itself and this 34 G. A. Russell D. F. Lawson H. L. Malkus R. D. Stephens G. R. Underwood T. Takano and V. Malatesta J. Amer. Chem. SOC.,1974 96 5830.35 G. A. Russell and J. L. Gerlock J. Amer. Chem. SOC.,1974 96 5838. 36 G. A. Russell P. R. Whittle C. S.C. Chung Y.Kosugi K. Schmitt and E. Goettert J. Amer. Chem. SOC.,1974,96 7053; G. A. Russell G. W. Holland K. Y. Chang R. G. Keske J. Mattox C. S. C. Chung K. Stanley K. Schmitt R. Blankespoor and Y. Kosugi ibid.,p. 7249. 37 G. A. Russell R. L. Blankespoor J. Mattox P. R. Whittle D. Symalla and J. R. Dodd J. Amer. Chem. SOC.,1974 96 7249; R. L. Blankespoor ibid.,p. 6196. G.A. Russell J. R.Dodd T. Ku C. Tanger and C. S. C. Chung J. Amer:Chem. SOC. 1974,96 7255. 142 W. T. Dixon Usyn uanti= 1 2 is borne out by INDO calculations. Assuming an odd-electron density on C-2 of ca. 5 spin-density in A is ca. -5 & (from Q 2.5 mT) and the average splitting of the syn- and anti-protons would be cu.5 x 3%) x mT = 0.08 mT (ob- served 0.12 mT). The order of magnitude is correct and supports the mechanism proposed. * The mechanism of the Kolbe electrolysis is apparently quite strongly depen- dent on interactions between the carboxylate ions and the surface of the anode.39 A reverse-Kolbe reaction can occur at the cathode for activated olefins4' may be reduced and add on carbon dioxide (or vice versa) (Scheme 6). It would appear that a cell could be constructed in which cyclic addition and loss of CO could be achieved! -co CH,=CHX 5 CH,=CHX ee -CHXCH,CO,-'''L Jc-.H' -O,CHXCH,CO -CH XCH,CO,H Scheme 6 6 Simple Aromatic Radicals Some more phenyl-type radicals have been generated in an argon matrix$l i.e.those with 2-OH or 2-OMe substituents (Scheme 7). As might be expected radicals isomerize when the temperature is allowed to rise. Scheme 7 '9 J. P. Coleman R. Lines J. H. P. Utley and B. C. L. Weedon J.C.S. Perkin II 1974 1064. 40 D. A. Tysee and M. M. Baizer J. Org. Chem. 1974 39 2819. 4' P. H. Kasai and D. McLeod jun. J. Amer. Chem. Soc. 1974 96 2338. Free Radicals 143 When phenols or hydroxypyridines are reduced by means of photoelectrons from sodium in an inert matrix the keto-form becomes preferred so that the resultant anions can be regarded as cyclohexadienyl-type radicals (8). In the case of the anion of 4hydroxypyridine two possibilities (9a) and (9b) seem to be realized. The relative yields of phenol and biphenyl during the radiolysis of aqueous solutions of benzene in the presence of N20are affected by the presence of oxidiz- ing ions,43 and the intermediate radicals can also by means of pulses be oxidized ele~trolytically.~~ A large number of aromatic acids also give rise to hydroxy- cyclohexadienyl radicals when present in the Ti1"-H202 flow set-~p.~',~~ The effect of the acid groups seems to be to decrease the rate of acid-catalysed elimination of water46 so that the sulphonated adduct radicals can persist at lower pH's (Scheme 8).Some hydroxyl adducts with naphthalenesulphonic acids have been observed by e.~.r.,~~ and the patterns of spin densities in them are the same as those in corresponding naphthoxyl radicals.3 1 Scheme 8 Phenoxyl radicals have also been exhaustively studied using both in situ radi~lysis~~ and oxidation by Ce'V.48 Well defined trends are observed in the spin distributions of substituted phenoxyl radicals trends which extend smoothly from the nitro-derivatives to the semiquinones. These trends are useful in assign- ing the coupling constants. Different radicals from aminophenols are observed in strongly acidic solutions4* from those seen in neutral although in both cases there are two protons attached to the nitrogen atom. This may be " P. H. Kasai and D. McLeod jun. J. Amer. Chem. SOC.,1974,96 2342. 43 M. K. Eberhardt J. Phys. Chem.. 1974 78 1795; K. Bhatia and R. H. Schuler ibid. p. 2335. 44 K. M. Bansal and A. Henglein J.Phys. Chem. 1974 78 160. 45 G. Filby and K. Gunther J. Phys. Chem. 1974 78 1521. 46 W. T. Dixon and D. Murphy J.C.S. Perkin II 1974 1630. 47 P. Neta and R. W. Fessenden J. Phys. Chem. 1974 78 523. W. T. Dixon M. Moghimi and D. Murphy J.C.S. Furuduy II 1974,70 1713. 144 W. T. Dixon due to protonation on the oxygen atom at lower pH.48 This contrasts with anilino-radicals4’ observed under more alkaline conditions when there is only one proton ‘permanently’ attached to nitrogen. As one might expect the ring- proton coupling constants in the anilino-radical lie between those of phenoxyl and benzyl [(lOH12)] although the negative densities are rather large. To 0,::;; ~~~~~~ 8 066mT 0 1X mT 061 mT 0 82 mT I04 mT (10) (1 1) (12) some extent this trend has been explained by INDO calculation^.^' Benzyl itself can be trapped in an adamantane matrix” at room temperature and in such an environment it has a remarkably long half-time (ca.0.5 h). CIDNP measurementss1 have shown that F splittings of ortho and para fluorine nuclei in fluorinated phenoxyl and anilino-radicals are positive whereas meta fluorine splittings in these radicals are negative. A variety of semiquinones from polyhydr~xynaphthalenes’~ and pyridine diolsS3 have been observed by e.s.r. and the hydroxylation of activated aromatic compounds52 by alkaline hydrogen peroxide (Scheme 9) has been monitored by (observed) Scheme 9 means of autoxidation. Facile dimerization of the radicals from the oxidation of pyridine diols by Pb0,-[Fe(CN),] -leads to diazophenosemiquinones (Scheme These schemes show how e.s.r.can be used as a tool in elucidating mechanisms even of non-radical reaction steps and a further example is the autoxidation of 6-hydroxytropolone. 54 The tropolonyl radical (13)54*55itself has an e.s.r. spectrum very similar to that of phenoxyl [see (14)and (15) for MO models] and it might therefore be expected that the b-hydroxy-derivative might resemble p-benzosemiquinone from the point of view both of its spin distribution and of its stability. In alkaline water 6-hydroxytropolone gave e.s.r. spectra first 49 R.V. Lloyd and D. E. Wood J. Amer. Chem. Soc. 1974 96 659. 50 J. E. Jordon D. W. Pratt and D. E. Wood J. Amer. Chem. Soc. 1974 96 5588.51 M. L. Kaplan M. L. Marion and H. D. Roth J. Phys. Chem. 1974,78 1837. 52 P. Ashworth and W. T. Dixon J.C.S. Perkin II 1974 739. 53 P. Ashworth Biochem. J. 1974 141 577. 54 W. T. Dixon and D. Murphy J.C.S. Perkin II 1974 1430;W. T. Dixon W. E. J. Foster and D. Murphy Mol. Phys. 1974 27 1709. 55 C. A. Russell and J. Kokensgard J. Amer. Chem. SOC.,1967 89 5059. Free Radicals I45 OH semiquinone 0 OH 0-observed (obsened) dimerization + [Ol dimeric product [o] (quinone) -Scheme 10 $i;-. -(JJ / 2.5 rnT 0 I 0 m'l (13) (14) (1 5) of p-benzosemiquinone and later of the semiquinone from 1,2,4-trihydroxyben- zene. In HMPA,* in the absence of water a spectrum of two triplet splittings (0.12,0.2 mT) was observed which eventually decayed to the p-benzosemiquinone spectrum.The sequence of events suggested is as in Scheme 1 1. +co, OH gH (-j6 OH 0-0 O-\ unstable radicals observed Scheme 11 The cyclo-octatrienyne negative ion56 has been observed by e.s.r. in the reduc- tion of monobromocyclo-octatetraene by potassium metal (16) and the effect ofsubstituents on the degenerate orbitals of the planar cyclo-octatetraene negative 56 G. R. Stevenson M. Colon J. G. Concepcion and A. M. Block J. Amer. Chem. SOC. 1974,96 2283. * Hexamethylphosphoramide. 146 W. T. Dixon 0.292 mT 0.406 mT K metal .. ion has been investigated. ’’ As expected electron-releasing groups such as methyl ethyl or butoxy split the degeneracy so that they are attached to positions which have minimum negative charge (and hence maximum spin density) (see Figure).I 11 I higher in energy originally degenerate NBO’s Figure Eleotron-releasing groups decrease the rate of reduction of tropylium ions by zinc metal or chromous ion so that an amino-s~bstituent~~ stops the reaction Scheme 12 altogether (Scheme 12). Chromium(I1) is also an effective one-electron reducing agent for enones :59 R RCH=CHCOR \CHCHR RCOCH2CH2R -0/-A popular way of producing radicals is by photochemical excitation of when the triplet-state species formed can add to double bonds abstract protons (effectively giving ketyl radicals) or even undergo some other intramolecular process. Intramolecular H-abstractions may be dominated by conformational rather than by energy requirements (entropic control).60 ’’ G.R. Stevenson and J. G. Concepcion J. Phys. Chem. 1974 78 90. 58 K. Okarnoto K. Komatsu and 0. Sakaguchi Bull. Chem. Sac. Japan 1974,47 2426 243 1. 59 H. 0. House and E. F. Kinloch J. Org. Chem. 1974 39 1173. 6o F. D. Lewis R. W. Johnson and D. R. Kory J. Amer. Chem. Soc. 1974,96 6456. Free Radicals 147 4- + HBr + GH + Br. 0 0 Radical bromination by N-bromosuccinimide" involves the succinimido- radical (1 8) which acts as a chain carrier. It seems that it is still not clear whether amido-radicals are necessarily of Q-or of ~r-type.~~ Amination of aromatic compounds by analogues of Fenton's reagent (i.e. Fe" and N-chloroamines) takes place via cyclohexadienyl-type intermediates (Scheme 1 3).64 + t Fe" R,NHCl + Fe" -P R,NH termination* R,NH H C1 Scheme 13 A variety of cyclic radicals containing nitrogen have been studied by means of e.s.r.These range from alkyl hydra~yls,~~ e.g. CF,N-N(CF3)R and the related hydrazine positive ions66 of type R,NkNR, which are of ntype to a-radicals of alicycliciminoxy- derivative^,^' characterized by large 14Ncoupling constants. Particularly intriguing are radicals related to 1-aziridylmethyl(1 9).6s " N. Shimizu M. Ishikawa K. Ishikura and S. Nishida J. Amer. Chem. SOC.,1974 96 6456; M. Hoshino S. Arai and M. Imamura J. Phys. Chem. 1974 78 1473; H. A. J. Carless J.C.S. Perkin II 1974 834; H. G. Heine W. Hartmann D. R. Kory J. G. Magyar C.E. Hoyle J. K.McVey and F. D. Lewis J. Org. Chem. 1974 39 691. 62 J. C. Day M. J. Lindstrom and P. S. Skell J. Amer. Chem. SOC.,1974,96 5616; J. G. Trynham and Y. S. Lee ibid. p. 3590. 63 J. N. S. Tam R. W. Yip and Y. L. Chow J. Amer. Chem. SOC.,1974 96 4543; T. Koenig J. A. Hoobler C. E. Klopfenstein G. Hedden F. Sunderman and B. R. Russell ibid. p. 4573. A. Clerici F. Minisci M. Perchinunno and 0. Porta J.C.S. Perkin II 1974 416. 65 L. Lunassi and K. U. Ingold J. Amer. Chem. SOC.,1974 96 5558. " S. F. Nelsen and R. T. Landis jun. J. Amer. Chem. Soc. 1974,96 1788; S. F. Nelsen G. R. Weisman P.J. Hintz D. Olp and M. R. Fahey ibid. p. 2916. 67 G. A. Russell and A. Mackor J. Amer. Chem. SOC.,1974 96 145; H. Caldararn and M. Moran ibid. p. 149. 68 W.C. Daren and C. T. West J. Amer. Chem. SOC.,1974 96 2447. 148 W. T. Dixon The preferred conformation is that with the odd-electron orbital on C eclipsing the lone-pair orbital of the nitrogen (20). This contrasts with the situation in the cyclopropylmethyl radical in which the nodal plane of the odd-electron orbital bisects the ring. Further wide varieties of radicals have been added to nitroso-compounds to a, give nitroxides of the types69 !>N-O where R =alkyl or aryl and X =SO -or [C0(CN),l3- as well as more usual trapping products.'' In several paper^,^^,^^ the mechanisms of decay of nitroxide radicals have been more fully elucidated than hitherto72 (e.g. Scheme 14). In aryl nitroxides addition7 RO. +CF,NO -+ F,C-N-OR I 0 1CF,NO F3C\ N-0-N-CF3 / I RO 0.1 CF,NOR +CF,N02 I N /-'\ RO CF (analysis by m.s.) Scheme 14 occurs via coupling to the aromatic nucleus rather than on to what is usually a sterically hindered nitrogen atom. One suggested path of reaction is shown in Scheme 15. D. Mulvey and W. A. Waters J.C.S. Perkin II 1974 666 772. M. J. Perkins and B. P. Roberts J.C.S.Perkin II 1974,297; A. J. Bard J. C. Gilbert and R. G. Goodin J. Amer. Chem. SOC.,1974,96 620; J. W. Neely G. F. Hatch and R. W. Dreilick ibid. p. 652. A. Calder A. R. Forrester and G. McConnachie J.C.S. Perkin I 1974 2198 2208 2213. D. H. R. Barton R. L. Harris R. H. Hesse M. M. Pecket and F. J. Urban J.C.S. Perkin I 1974 2344. Free Radicals 149 R' Me0 0 N HBu -R' R' Scheme 15 An attempt has been made to rationalize the stability of free radicals in terms of steric hindrance and a push-pull effect designated 'rnerostabili~ation'.~~ The gist of this rationalization is that a compound such as p-nitroaniline (21a) has 'extra' stability due to favourable canonical structures [e.g.(21b)j in which (21b) one substituent donates and the other accepts an electron pair. This idea is carried over to free radicals so that high stability is expected for nitroxides and hydrazyls +' R,N-*O. ++ R2N+O--R,NANR t- R,&-+NR High stability may also be expected for suitably substituted aromatic radicals if appropriate canonical forms can be written down e.g. (22). However it seems doubtful whether anything new has really been gained by introducing the new term since these radical types are fairly well understood already.73 R. W. Baldock P. Hudson A. R. Katritzky and F. Soti J.C.S. Perkin I 1974 1422 1427. 150 W. T. Dixon Scheme 16 Thianthrene (23) can be formed by a free-radical substitution reaction74 as shown in Scheme 16 and its positive radical ion (24) can behave as an electro- ~hile.~~ it has been suggested that radicals previously From solid-state st~dies’~ identified as RS are probably cr* radicals of the type RS’SR . + Finally two systems of biological interest deserve mention. First the cation radicals of some porphyrins (25) have been ob~erved,~’ and restricted rotation is found when R is an alkyl group other than methyl.It is suggested that some -7 t (25) 74 L. Benati P. C. Montevecchi A. Tundo and G. Zanardi J.C.S. Perkin I 1974 1272. ’’ K. Kim V. J. Hull and H. J. Shine J. Org. Chem. 1974 39 2534. 76 M. C. R. Symons J.C.S. PerkinZI 1974 1618. 7’ J. Fajer D. C. Borg A. Forman A. D. Adler and V. Varadi J. Amer. Chem. SOC. 1974,96 1238. Free Radicals 151 I I I CO.C,H; I -m co‘C,H .icl-nt CO .C H; I-m Scheme 17 enzymes connected with the destruction of hydrogen peroxide function cia radicals of this type. Cortisone acetate has been ~ynthesized’~ using a radical ‘relay’ mechanism as shown in Scheme 17. If the p-iodide is used attack is on position 14 instead of position 9. R.Breslow R. J. Corcoran and B. R.Snider J. Amer. Chem. Soc. 1974,96,6791 et seq.