4 Reaction Mechanisms Part (iii) Electron Spin Resonance Spectroscopy and Free Radical Reactions By A. T. BULLOCK Department of Chemistry University of Aberdeen Old Aberdeen Scotland A89 2UE Of several review articles and books published during the year attention is especi- ally drawn to a book devoted to applications of e.s.r. to polymer research’ and an admirable albeit wholly theoretical article on theories of chemically induced dynamic electron polarization (CIDEP).* Wan and Elliott3 have briefly reviewed CIDEP mechanisms and more importantly have given illustrations of how CIDEP and CIDNP may be used together to study photochemical reactions. E.s.r. studies of triplet states have been reviewed4 with attention drawn to cyclic 47r-electron systems CHz and environmental effects.Especial emphasis was put onto ground state triplets. Finally of relevance to the general field of this report although not specifically about e.s.r. studies is a review of free-radical rearrangements in telomerization. 1 Kinetics and Mechanism Much of the work described in this section deals with radicals having the unpaired electron centred largely on atoms other than carbon. There have been detailed studies of the generation and subsequent reactions of alkyloxy and aryloxy radicals. A possible general route to alkyloxy radicals has been described which involves the photolysis (A = 365 nm) of alkoxy-vanadium(v) chelates.6 The primary step is OR (1) (2) where Q is the 8-quinolyloxo ligand and R represents a series of nineteen alkyl groups.The production of paramagnetic compound (2) containing Vrv was ’ B. Rhby and J. F. Rabek ‘PolymersjProperties and Applications. Vol. 1. E.S.R. Spectroscopy in Polymer Research’ Springer-Verlag Berlin Heidelberg and New York 1977. ’P.W. Atkins and G. T. Evans ‘Advances in Chemical Physics’ XXXV p. 1 el. seq. ed. I. Prigogine and S. A. Rice. J. Wiley and Sons New York. London Sydney and Toronto 1976. J. K. S. Wan and A. J. Elliot Accounts Chem. Res. 1977,10 161. E Wasserman and R. S. Hutton Accounts Chem. Res. 1977,10,27. ’R. Kh. Freidlina and A. B. Terent’ev Accounts Chem. Res. 1977 10 9. S. M. Aliwi and C. H. Bamford J.C.S. Faraday I 1977.73 776. 90 Reaction Mechanisms-Part (iii) Electron Spin Resonance demonstrated directly by e.s.r.whilst comparison of the quantum yields for photo- initiated free radical polymerization of methyl methacrylate in the presence of (1) showed that each molecule of chelate decomposing gave rise to one initiating radical. Subsequent reactions of the alkyloxy radicals were studied by spin-trap- ping experiments using phenyl t-butyl nitrone (PBN) at various concentrations. In the case of primary alkyloxy radicals the e.s.r. spectrum depended on [PBN]. For example for R = Bun a simple spectrum was obtained at [PBN] = 0.008 M which was attributable to the adduct from the isomerized hydroxyalkyl radical (4). PrnCH20. 3 Pr"CH0H (2) (3 (4) At high [PBN] (ca. 0.3 M) both (3) and (4)were trapped. Most secondary aIkyloxy radicals gave the same spectra at high and low concentrations of PBN.The coupling constants as-Hand aNof the resultant nitroxides were characteristic of alkyl adducts. Clearly p -scission occurs according to the general equation R(R)CHO* + R'CHO+.R' (31 However p-scission was found to be less important for R=Pr' cyclopentyl and cyclohexyl where the behaviour was similar to that found for the primary alkyloxy radicals i.e. complex spectra at high [PBN] with and uN suggesting alkyloxy and hydroxyalkyl adducts. Only two derivatives with t-alkyl groups were studied namely R = But and t-pentyl. Both gave simple spectra at high and low [PBN] with splittings characteristic of alkyl adducts. Again this was thought to be a consequence of /3 -scission viz. Me3C0. -+ CH3+ Me2C0 (44 EtMe2CO-+ C2H5+ Me2C0 (4b) Product analyses confirmed this the yield of Me2C0 being 60% for reaction (4a) and 100% for (4b).The reactions of secondary alkyloxy radicals have also been studied in a flow system.' They were produced by Ti"' reduction of secondary alkyl hydroperoxides and cyclic hydroperoxides. No RO-radicals were observed directly but some were trapped by the aci-anion of nitromethane and characterized by the e.s.r. parameters of the ROCH2NP2 adducts. In all cases carbon-centred radicals were observed to be derived from the parent alkyloxy radicals and were explained in terms of the following four reaction pathways. (i) 1,2-hydrogen shift [cf.equation (2) for n-alkyloxy radicals] 1.2-H R'R~CHO--CR'R~OH H20 (ii) 1,5-hydrogen shift 1,s-H e.g.Me(CH&CH(O.)Me -CH2(CH2)2CHMeOH (6) ' B. C. Gilbert R.G. G. Holmes and R. 0.C. Norman J.Chem. Research 1977 (S)l; (M)Ol01. A. T.Bullock (iiia) C-C fragmentation e.g. (oIo*z .CH20(CH2)20CH0 (7) 0 (iiib) C-Me fragmentation FC-Me e.g. MeCH20CH(O*)Me-*CH3 (iv) C-0 fragmentation (iFC-0 e.g. Me2CHOC(0.)Me,-Me2eOH (9) (ii)l.2-H The authors suggest that the 1,2-H shift [equation (5)J is not intramolecular but probably involves the solvent.8 It was concluded that alkyloxy radicals from secondary hydroperoxides show an ease of reaction in the order 1S-H>FC- > 1,2-H>FCkMe.This contrasts with the order for primary alkyloxy radicals uir. 1,2-H "1,5-H>Fc-c. Aryloxy radicals have been obtained by photorearrangement of nitro-compound^.^ Photochemical rearrangements of nitro-groups are often postulated to proceed uia the nitrite" and it is claimed' that the detection of aryloxy radicals during the photolysis of several hindered nitro-compounds is strong evidence for the intermediate formation of nitrite groups.As an example photolysis of 6-nitro-benzo[a]pyrene (5) gave the spectrum of the aryloxy radical (6) [Scheme (l)].The radical (6) had previously been prepared by H-atom abstraction from benzo[aIpyren-6-01.l1 NO2 (5)=ArN02 Scheme 1 Other oxygen-centred radicals have been studied notably the decay kinetics of cumylperoxy radicals. l2 Previous e.s.r. determinations of the decay kinetics of peroxy radicals have given first second and intermediate orders.The authors point out that in most previous reports the radicals were produced photolytically B. C. Gilbert R. G. G. Holmes H. A. H. Laue and R. 0.C. Norman J.C.S. Perkin II 1976 1047. Y. Ioki J.C.S. PerkinII 1977 1240. lo 0.L. Chapman D. C. Heckert J. W. Reasoner and S. P. Thackaberry J. Amer. Chem. Soc. 1966,88 5550. C. Nagata M. Inomata M. Kodama and Y. Tagashira Gann 1968,59 289. '2 S. Fukuzumi and Y. Ono,J.C.S. Perkin II 1977,622. Reaction Mechanisms-Part (iii) Electron Spin Resonance and suggest that the measured kinetics may have been complicated by the presence of other products of the photolysis. To avoid this the cumylperoxy radicals were produced by flowing a solution (CCl, benzene and cumene solvents) of cumyl hydroperoxide (0.744.70 M) through a tube packed with 20-30 mesh manganese dioxide or cobalt oxide supported on silica.On stopping the flow decay of the cumylperoxy radicals was found to be strictly first-order (kt=0.15* 0.015 s-' at 300 K) irrespective of the initial hydroperoxide and cumylperoxy concentrations and of the nature of the solvent and catalyst. The authors s~bsequently'~ confirmed the first-order decay by measuring the radical concen- trations during the steady-state decomposition of cumene hydroperoxide with Pb02 the latter being held in suspension in the cavity. The decomposition was found to be a radical chain-reaction. The hydroperoxide formed a complex on the surface of Pb02 the complex subsequently dissociated to the peroxy radical which then desorbed into solution.From the kinetic studies it was found necessary to postulate the formation of a 2 1 complex between the hydroperoxide and the peroxy radical. Two pieces of spectral data supported this namely the g-value and the observed line-width both varied monotonically with the initial concentration of cumene hydroperoxide. The kinetic g-value and linewidth results were all accommodated by K =0.50 l2 moF2 for the equilibrium RO2. +2R02H $ (RO2* .* 2RO2H) (10) The same authors have used similar techniques i.e. the heterogeneously catalysed production of radicals to study the liquid-phase autoxidation of cumene with Pb02,14 the determination of cross-propagation rate coefficients in the autoxidation of hydrocarbon^,'^ and the mechanism of the formation of the p-benzosemiquinone anion over manganese dioxide.16 In the autoxidation of cumene (with Pb02) the kinetics were studied over the temperature range 291-393 K.' E.s.r.was used to monitor the concentration of the chain-propagating cumylperoxy radical and the rate of oxygen consumption was measured under the same conditions. Both the radical concentration and the rate of oxygen consumption were constant with time and independent of the catalyst weight liquid volume ratio. Experimentally it was found that [RO2*]= k,[RH][Pb02]* (11) and -d[02]/dr =kb[RH]2[Pb02]0 (12) These results together with the product distribution were shown to be consistent with the following mechanism. Ri Initiation R02H+Pb02 +RO2.k14 R02-+RH -Propagation R. +R02H l3 S. Fukuzumi and Y. Ono J.C.S. Perkin ZI 1977,625. l4 S. Fukuzumi and Y. Ono J.C.S. Perkin ZZ 1977,784. S. Fukuzumi and Y. Ono J. Phys. Chem. 1977,81,1895. l6 Y. Ono T. Matsumura and S. Fukuzumi J.C.S. Perkin II 1977 1421. kl8 ROy d PhCOMe+MeO. fast MeO. +R02. dstable products The value of 2kr7was found to be 5.0~ lo51 mo1-ls-l (291K) in agreement with that found ea~1ier.l~ The following Arrhenius parameters were found loglo(k14/l mol-' s-l) = 5.3-3o/e (20a) and loglo(k17/1 rnoi-' s-') = 10.9-30/e (20b) where 0 = 2.303 RT kJ mol-'. The Pb02-catalysed decomposition of t-butyl hydroperoxide in the presence of hydrocarbons has been studied by the same method." It has been shown that simultaneous measurements of [Bu'Oz*] (by e.s.r.) and the rate of oxygen genera- tion lead to a direct determination of the rate coefficients for cross-propagation reactions (e.g.hydrogen abstraction from hydrocarbons by Bu'02-) and for self- reaction of Bu'Oz*. Rate coefficients for termination chain transfer and cross termination involving Bu'02* were also determined by e.s.r. It was suggested that the technique provides a method for determining absolute rate coefficients for reactions involving Bu'Oz* at various temperatures that is more accurate and simple than the classic hydroperoxide method.17 Before leaving this technique we note that the mechanism of the formation of the p-benzosemiquinone anion over manganese dioxide has been studied.l6 The kinetic results could be fitted to the equation d[SAlldt =k(b[HQ10-[SAl)(c[Mn0~10-[SAI) (21) where k,b and c are constants [Mn02Jo is the Mn02 weight to liquid volume ratio and [Ha] and [SA] are respectively the initial concentrations of hydroquinone and semiquinone anion radical. It was proposed that the active surface species Mn02* abstracts a hydrogen atom from hydroquinone to give the neutral radical which then desorbs and is converted into the semiquinone anion radical. The number of surface species of Mn02* was determined to be 7.7 x 10l8m-, in good agreement with previous values obtained for the oxidations of l,l-diphenyl-2-picrylhydraz-ine18 and cinnamyl alc~hol.'~ l7 J. A. Howard W. A. Schwalm and K. U. Ingold Adu. Chem. Ser.1968 No.75 p. 6. A. T. T. Oei and J. L. Garnett J. Catalysis 1970 19 176. l9 D. Dollimore and K. H. Tonge J. Chem. SOC.(B),1967,1380. Reaction Mechanisms-Part (iii)Electron Spin Resonance 95 The reactions of esters and anhydrides in a Ti"'/HzO2 flow system have been studied in some detail2' following an earlier observation that vinyl acetate in this flow system gave mainly the spectrum of the methyl radicaL2' It has now been found that the alkyl radicals are formed by the route indicated in equations (22)-(24) uiz. RCOzR' + HO2-+ RC020-+ R'OH (22) RC020H+Ti"' -+ RC02*+ OH-+Ti'" (23) The first step is perhydrolysis followed by one-electron reduction of the resultant peroxo-acid. Decarboxylation of the acyloxy radical RC02* then gives the alkyl radical.In some cases the spin trap CH2:NOz- enabled the intermediate acyloxy radicals to be detected in addition to the alkyl radicals. From the relative concen- trations of the spin adducts RC02CH2N02'- and RCH2N02'- it was found that the rates of decarboxylation decrease in the series shown in Scheme 2. A lower limit co; Scheme 2 for the rate of decarboxylation of CH3C02- was estimated to be 2 X lo7s-l whilst an upper limit for PhCO2. is 2.5 x lo5s-l. Decarboxylation reactions have also been observed in the adduct radicals formed by the reaction of photochemically produced acyl radicals with the fumarate dianion.22 The acyl adducts (7) under-went decarboxylation to give the 1-carboxy-3-oxyallyl radical dianions (8) [Scheme 31. For R=Me and Et the e.s.r.spectra of (7) and (8) were observed simul- taneously. Plots of loglo{[RC(O-)=CHCHO2-]/[RCOCH(CO~-)CHCO,-I) us 1/T showed that the activation energies for the decarboxylation step were equal for R = Me and Et to within experimental error and had the value 40 f2 kJ mol-'. The acyl radicals were readily identifiable since hyperfine coupling to protons bonded to 2o B. C. Gilbert R. G. G. Holmes P. D. R. Marshall and R. 0.C. Norman J. Chem. Research 1977 (S)172; (M)1949. D. J. Edge B. C. Gilbert R. 0.C. Norman and P. R. West J. Chem. SOC.(B) 1971 189. 22 S. Steenken and M. Lyda J. Phys. Chem. 1977.81,2201. 96 A. T. Bullock C4 and Cs (C and C of the acyl precursor) of the ally1 radical (8)could be measured. In contrast to these exampleszo*z2 of ready decarboxylations a study of the radical zwitterions formed by the one-electron oxidation of a series of m'ethoxylated benzoic acids by TI2+ Agzt and SO,-has shown that the presence of even one methoxy- group is sufficient to stabilize the radicals with respect to decarb~xylation.~~ The radical zwitterions could also be produced from the addition of *OHto mono- di- and tri-methoxylated benzoic The adducts reacted with H' (k = 10'-lo91mo1-ls-l) to give the zwitterions via elimination of water from the protonated adduct.There have been several reports of sulphur-centred radicals Thus a series of aromatic and aliphatic sulphinyl radicals have been observed and ~haracterized.~' The aromatic sulphinyls were produced by one or other of the three routes shown in Scheme 4,where the -OH radicals were generated in a conventional Ti1I1/H2O2 OH 0 *OH I -H+ // ArSSAr ArSSAr _r Ar-S -ArS .OH H202 -OH //O ArSH -ArS.+ArSOH +ArS 0 hv // ArS(0)CI -+ ArS + C1. Scheme 4 flow system. The aromatic radicals were all 71-radicals with extensive delocaliza- tion on to the aromatic ring. The aliphatic sulphinyl radicals were generated by photolysis of sulphinyl chlorides at low temperatures (ca. -100 "C) and were characterized by g ~2.011 and a'-* in the range 0.8-1.1 mT. Particular attention was paid to MeSO for which g = 2.0100 a(3H)= 1.15 mT and AH the peak-to- peak derivative linewidth was 1.10 mT. The authors suggest that a previous reportz6 of the spectrum of MeSO obtained by the photolysis of t-butyl-methanesulphenate refers to some other sulphur radical [a(3H)= 0.657 mT g = 2.00965 AH = 0.06 mT].Previous failures to detect MeSO and EtSO were attri- buted to efficient spin-rotation relaxation for which Tl q/T. Sulphinyl radicals have also been observed in a study of the photochemical decomposition of organic sulphites and reactions of the sulphites with alkyloxy radical^.^' Dialkyl sulphites (R1R2CH0)zS=0 reacted with Bu'O. in two ways. Addition at the sulphur atom gave sulphuranyloxy radicals (R'RzCHO)zS(0)OB~' while hydrogen abstraction also occurred giving the carbon-centred radical ~CR'RzOS(0)OCHR'Rz. This then fragmented to give alkoxysulphinyl radicals R'R2CHOS0. Photolysis of the dialkyl sulphites in the absence of di-t-butyl peroxide resulted in fission of the S-0 bond to give alkoxysulphinyl and alkyloxy (-0CHR'R') radicals; the latter 23 P.O'Neill S. Steenken and D. Schulte-Frohlinde J. Phys. Chem. 1977,81,26. 24 P. O'Neill S. Steenken and D. Schulte-Frohlinde J. Phys. Chem. 1977 81 31. 25 B. C. Gilbert C. M. Kirk R. 0.C. Norman and H. A. H. Laue J.C.S.Perkin 11 1977,497. '' T. Kawamura P. J. Krusic and J. K. Kochi Tetrahedron Letters 1972 4075. 27 B. C. Gilbert C. M. Kirk and R. 0.C. Norman J. Chem. Research 1977 (S)173; (M)1974. Reaction Mechanisms-Part (iii) Electron Spin Resonance 97 subsequently rearranged to *CR'R20H. The spectrum of HOSO was also obser- ved in this work and possible schemes for its generation were given. The structure and formation of some dialkoxysulphuranyl radicals (10) have been described.28 Two routes seemed to involve sulphenates (9) namely the photolysis of disulphides R'SSR' and thioethers R'SR' in the presence of peroxides.These are shown in Scheme 5. Support for the intermediacy of sulphenates in routes (a) and (b) of Scheme 5 comes from the fact that sulphenates OR2 (a) R'SSR'+*OR~ 4 R'-S / -SR' OR2 .OR2 / CSR (9) -R'-$ R'-S-OR2 (b) R'SR'+-OR2 -+ R'-$-R1 I OR2 Scheme 5 themselves give dialkoxysulphuranyl radicals on photolysis. Evidence for the loss of OR'in Scheme 5(b) was provided by the result for theitan from which the ring-opened radical *CH2CH2CH2SOBu' was detected. The third route to dialk- oxysulphuranyl radicals was shown to be the direct addition of alkyl radicals to dialkylsulphoxylates (R0)2S.Except for R' = CF3 (g = 2,0079) all g values were in the range 2.0090-2.0096. Radicals of the type RCH2 (OBU')~ show spectra consistent with structure (11) whilst Ph 9(OBu'X is essentially a w-radical of structure (12). Deuterium labelling experiments together with e.s.r. have shown that the addition of MeO- radicals to (Me0)2S to give the trimethoxysulphuranyl radical (MeO)& has a high degree of stereosele~tivity.~~ The incoming radical was found to take up an apical site in the adduct. Furthermore neither intra- nor inter-molecular ligand exchange took place to an appreciable extent during the average lifetime of the sulphuranyl radicals. Of radicals carrying a significant spin density on nitrogen atoms nitroxides continue to attract attention.The syntheses e.s.r. spectra and stabilities of two 28 W. B. Gara B. P. Roberts B. C. Gilbert C. M. Kirk and R.0.C.Norman J. Chem. Research 1977 (S)152; (M)1748. 29 J. W. Cooper and B. P. Roberts J.C.S. Chem. Comm. 1977 228. A. T. Bullock 3-oxy- 1,3-diazacyclohexene- 1-oxide radicals [( 13a) (1 3b)l have been de~cribed.~' They were found to be significantly more stable than comparable a-unsubstituted nitr~xides.~' Thus (13a) decomposed with first-order kinetics with k = 6x s-' (298 K). The authors discounted the possibility of a monomer-dimer equilibrium which would show first-order kinetics for decomposition involving either slow bimolecular disproportionation of monomer or unimolecular de-composition of dimer and suggest the possibility of H-atom transfer from C-4 to 0-1 producing a reactive carbon atom.From the coupling constant of uH-4 (1.09 mT) the appropriate dihedral angle was estimated to be 10"in (13a). This is I 0-(13) a; R=H b; R=Me consistent with H-4 having a pseudo-axial conformation two of the three methyl groups being pseudo-equatorial. Kinetic studies have also been carried out on the nitroxides (14) (15) and (16) produced by photolysis of Et3SiH solutions of 2- 3- N(i))OSiEt3 (J/N'"'"" Q oN(0)OSiEt3 (14) (15) (16) and 4-nitropyridines in sit^.^^ Again first-order decays were observed but temperature-jump experiments clearly showed that dimer lay in the reaction pathway.Equation (25) shows the mechanism although a clear choice between decay of monomer and decay of dimer could not be made Products t 2ArN(b)OR $ Dimer -B Products (25) A literature misassignment has been corrected. The e.s.r. signal obtained on the reaction of pentyl nitrite with aniline in benzene had been attributed to the u-radical P~N-NOW.~~ However it has now been that this reaction and the reaction of pentyl nitrite with 1,3-diphenyl-triazine give signals consisting of overlapping spectra. The species responsible were shown to be diphenylnitroxide and phenylpentyloxynitroxide.The reactions between trifluoronitrosomethane and some 1,3-diketones have been shown to generate the nitroxide radicals CF3N(0)CH(COR')COR2 and their tautomers CF,N(O)C(COR') C(OH)R2 30 S.N. Ghriofa R.Daray and M. Conlon J.C.S. Perkin I 1977,651. 31 D. F. Bowman T. Gillan and K. U. Ingold J. Amer. Chem. Soc. 1971,93,6555. 32 L. Lunazzi G. Placucci and N. Ronchi J.C.S. Perkin 11,1977 1132. " A. F. Levit,and I. P. Gragerov Zhur. org. Khim. 1969 5 31. 3A J. I. G. Cadogan R. G. M. Landells R.M. Paton J. T. Sharp and R. U. Weber J. Chem. Research 1977 (S)108. Reaction Mechanisms-Part (iii) Electron Spin Resonance 99 together with the iminoxy radicals (R*CO)(R'CO)C Under similar rezction conditions however diethyl 2-methyl-3-oxosuccinate afforded only the nitroxide CF,N(b)CMe(C02Et)COC02Et.A reaction scheme was proposed which involved hydroxylamine anions CF3N(6)CH(COR2)COR' as intermediates. A series of iminyl and triazenyl radicals have been derived by radical attack on some organic a~ides.~~ Bu'O- radicals were found to react with a series of primary and secondary alkyl azides to give spectra characteristic of the iminyl radicals (g = 2.0028-2.0029 aN=0.95-0.98 mT) uiz.Bu'O. +HC(R')(R2)N-&~N Bu'OH +N2+R'(R2)C=N-(26) The intermediate a-azidoalkyl radicals were not observed. In contrast Et3Si* and Ph3Si. added to primary alkyl azides to give either R2-N=N-fiSiR (17) or R2(R:Si)N-N=N-(18). From a consideration of g-values (2.0010-2.0012) and the three values of aN(1.76-2.0 mT 0.34-0.40 mT and 0.12-0.19 mT) it was argued that (1 8) was the probable structure i.e. the 3,3-disubstituted triazenyl radical. The chain process involved in the decomposition of diphenyldiazomethane induced by copper perchlorate has been studied kinetically and product analyses have teen carried out.The e.s.r. spectra observed suggested that the radical cations Ph2CN2 and Ph2CN NCPh2 are key intermediates in the chain rea~tion.~' Following earlier studies of the u*-T* orbital crossover in a series of fluorinated benzene radical anions,38 the isotropic spectra of several fluorinated pyridine anions have been observed in an adamantane matrix.39 Very large couplings to I9F in penta- and 2,3,4,6-tetrafluoropyridineanions together with small values of a for the more lightly fluorinated species indicate the former to be u-radicals whilst the latter are .rr-radicals. The u*-T* orbital crossover was rationalized in terms of the stabilization of u* and destabilization of T* orbitals due to the inductive effects by back donation from fluorine.Its relevance to the present report lies in the suggestion by the authors that the availability of low-lying u* states should be carefully considered when assessing the chemistry of polyfluoro-aromatics. This is especially true for their reactions with nucleophiles. Phosphorus-centred radicals are represented by studies of some reactions of phosphate radicals4' and of the stereochemical non-rigidity and relative ligand apicophilicities of some phosphoranyl radical^.^**^^ The phosphate radical P642- and its protonated forms HPO4- and H2Pb4 were prepared in situ by photolysis of hu potassium peroxodiphosphate (PzOa4- -2P642-).40 While the phosphate radi- cals themselves were not detected by e.s.r.their adducts to fumaric and maleic acids and to the aci-anion of nitromethane were observed. In general the reac- tions were similar to the related species Sb4- i.e. adduct radicals were obtained 35 B. L. Booth D. J. Edge R. N. Haszeldine and R. G. G. Holrnes J.C.S.Perkin 11,1977 7. 36 J. W. Cooper B. P. Roberts and J. N. Winter J.C.S. Chem. Comm. 1977,320. 37 D. Bethell K. L. Handoo S. A. Fairhurst and L. H. Sutcliffe J.C.S. Chem. Comm. 1977 326. 38 M. B. Yim and D. E. Wood J. Amer. Chem. SOC.,1976,98,2053. 39 M. B. Yim S. DiGregorio and D. E. Wood J. Amer. Chem. SOC.,1977,99,4260. 40 P. Maruthamuthu and H. Taniguchi J. Phys. Chem.. 1977,81 1934. 41 J. W. Cooper M. J. Parrott and B.P. Roberts J.C.S. Perkin 11 1977 730. 42 R. W. Dennis I. H. Elson B. P. Roberts and R. C. Dobbie J.C.S. Perkin 11 1977 889. 100 A. T. Bullock with unsaturated substrates hydroxyalkyl radicals from aliphatic alcohols and inorganic radicals such as to3-from HC03- and P032-from HP032-. However they differed in their reactions with aliphatic and aromatic carboxylic acids. At neutral pH phosphate radicals caused hydrogen abstraction from saturated alipha- tic mono- and dicarboxylic acids giving rise to a-carbon radicals whereas SO,-gave mainly radicals produced by decarboxylation. On the other hand phthalic acid gave a substituted phenyl radical on reaction with the peroxodisulphate system but did not do so with the phosphate radical. The authors concluded that HP0,- is a milder oxidant than the structurally related So4’-radical.A series of phosphoranyl radicals have been found to exhibit linewidth effects which were interpreted in terms of intramolecular ligand exchange at the phos- phorus atom,41 In general it was concluded that the following order of ligand apicophilicity holds F CI RCO2 >RC(O)NR OCN > RO R2N >H > R Other line-shape changes have been observed in some fluoroalkoxy-and fluoroalkyl-phosphoranyl radicals.42 It was concluded that the apicophilicity of -CF3 was less than that of -C1 but usually greater than ROO.Additions of trifluoromethyl radicals to trialkyl phosphites was found to be rever~ible~~ just as an earlier study had demonstrated the reversibility of the addition of methyl radicals.43 Radical addition to tervalent phosphorus when followed by p-scission of an exis- ting ligand has been noted to be the free-radical equivalent of the Arbuzov rea~rangernent.~ Me2N* and phenyl radicals were generated photolytically and allowed to react with a series of cisltrans isomeric five- and six-membered ring phosphites.The reactions were found to be nearly stereospecific and the authors concluded that for the phosphoranyl intermediates permutational isomerization of the Berry or turnstile mechanisms could not compete kinetically with the product- forming p -scission In the field of carbon-centred radicals a flow system has been designed to test for the intermediacy of free radicals in the currently accepted mechanistic scheme for the reaction between benzylic halides and aromatic radical anions.,’ The accepted scheme is RX+NaCloHs + R.-* R-+ products (27) The bis(3,5-di-t-butylphenyl)methylradical was observed in the reaction between its bromide precursor and sodium naphthalenide. The generation of phenyl radi- cals and their subsequent abstraction reactions with a series of aliphatic substrates have been The abstracted atoms were H Br and I. Competitive experiments confirmed the nucleophilicity of the phenyl radical and it has been suggested that the transition state contains a significant contribution from polar structures such as Ph’. -H. . -CHXMe. Radical addition to alkynes and the subsequent intramolecular reactions of the resultant vinyl radicals have been J.W. Cooper and B. P. Roberts J.C.S. Perkin ZZ 1976 808. 44 W. G. Bentrude W. Del Alley N. A. Johnson M. Murakami K. Nishikida and H-W. Tan J. Amer. Chem. Soc. 1977,99,4383. 45 K. Schreimer H. Oehling H. E. Ziegler and I. Angres J. Amer. Chem. SOC.,1977 99 2638. 46 B. Ashworth B. C. Gilbert and R. 0.C. Norman J. Chem. Research 1977 (S)94; (M)1101. Reaction Mechanisms-Part (iii) Electron Spin Resonance 101 de~cribed.~’The vinyl radicals were not detected but were clearly implicated. Intermolecular addition and intramolecular abstraction of the vinyl radicals gave rise to the observed radicals. In the detailed schemes deduced several examples of 1,5-hydrogen atom shifts were observed including an oxygen-to-carbon example. In view of the wide range of reported values for the rate coefficient for mutual termination of t-butyl .radicals the kinetics of this reaction have been studied carefully in isobutane and cyclopentane over the temperature range 170-330 K.48 The radicals were generated by several photolytic routes using a rotating sector- digital signal averaging technique. The results were that 2k = 1.1X 10” 1mol-’ s-’ (298 K) with E = 4.3 kJ mol-’ in both solvents. These are close to recent gas-phase measurements at the same temperature (2k = 9 x lo91 mol-’ s-’)~’ and to those for termination in a series of alkanes” and were well fitted to the Smoluchowski equation suitably modified by the microviscosity theory.” The concordance of the results from several different methods of radical production indicated that the observed rate coefficients were not significantly affected by geminate recombination in the solvent cage.It was pointed that recent gas- and liquid-phase results cast serious doubt on the validity of the currently accepted thermochemistry for alkyl ~adicals.’~ In addition to examples cited earlier there have been several interesting reports on spin-trapping studies. The traps used most widely remain 2-methyl-2-nitroso- propane (MNP) and phenyl-t-butyl nitrone (PBN). There have been two reports on spin-trapping in the y-radiolysis of alcohol^.^^*^^ High concentrations of MNP were used to detect the intermediate methyl radicals in the radiolysis of Bu‘OH MeCH(OH)Me and MeCH20H and the yields of nitroxide adducts were in the same order as the yields of methane.53 However no adduct was observed in the case of MeOH which suggests either that the methyl radical is not a precursor of methane in this case or that it has excess energy and reacts with solvent rather than with the spin trap.PBN was the trap of choice in a study of the y-radiolysis of fluorinated In general both hydrogen-atom and radical adducts were observed. A particularly important observation was that in mixtures of (CF3),CHOH and Me,CHOH the ratio [H adduct]/[radical adduct] increased by more than one hundred-fold as the composition varied from the fully protiated to the pure fluorinated alcohol. This reflected a change in loss pathways for hydrogen atoms especially hydrogen-abstraction reactions in the protiated case.Both MNP and PBN have been used in a study of the thermolysis of aryldiazo alkyl ethers.” When carried out in aromatic solvents the reaction yields biaryls and is thought to involve aryl radicals as intermediates. The reaction is summarized by the equation Ar’-N=N-0-R+Ar2H 4 Ar’Ar2+N2+ROH (28) 47 W. T. Dixon J. Foxall G. H. Williams D. J. Edge B. C. Gilbert H. Kazarians-Moghaddam and R. 0. C. Norman J.C.S. Perkin IZ 1977,827. 48 J. E. Bennett and R. Summers J.C.S. Perkin 11 1977 1504. 49 D. A. Parkes and C. P. Quinn J.C.S. Furuduy I 1976,73 1952. H. Schuh and H. Fischer Internat. J. Chem. Kinetics 1976,8 341. ” A. Spernol and K. Wirtz 2.Nuturforsch. 1953,88,522. 52 R. Hiatt and S. W. Benson Internat. J. Chem. Kinetics 1973,5 385.53 F.P.Sargent and E. M. Gardy J. Phys. Chem. 1977,81 1215. 54 A.C.Ling and L. Kevan J. Phys. Chem. 1977,81 605. ” R.M. Paton and R. U. Weber J.C.S. Chem. Comm. 1977,769. 102 A. T. Bullock Product analyses of isomer ratios the observation of CIDNP and the spin-trapping experiments all confirmed the intermediacy of aryl radicals in this reaction. The phenyl radical has been trapped both by MNP and PBN in the radiolysis of benzene.56 Previous evidence for its presence as an intermediate had been rather indirect. The apparatus and method for trapping radicals produced in a silent electric discharge in the gas phase has been de~cribed.~~" One surprising feature was that trapping seemed to be quite selective only one or two radicals being observed in any one experiment.For example gas chromatographic analysis of the products from 2-methylpropane showed no less than 'seventeen products. However only the 2-methylpropyl radical was trapped in observable quantities. A general review of spin trapping of photolytically-produced radicals in the gas phase has also appeared.57b The kinetics of spin trapping continue to receive attenti~n.~**~~ A combination of pulse radiolysis and spin-trapping data was used to determine the rate of addition of MeO- to MNP for which a rate coefficient of 1.3X 10*1mol-*~-~ (-45 "C) was A more extensive report is concerned with the measure- ment of rate coefficients for spin-trapping primary alkyl radicals with MNP and PBN.59 The method made use of the fact that the 5-hexenyl radical (19) isomerizes to cyclopentylmethyl (20) at a rate k which is reliably known.The spin adducts of the two isomers were distinguished by labelling (19) at the starred position with 13C. Hyperfine coupling to this carbon was only detectable in the adduct of (19) with both traps. The same rate coefficient for trapping kT was assumed for both isomers and was calculated from kT = k,[ 19T.]/[T][20T*] where [19T-] represents the concentration of the adduct of (19) with the trap T and [20T*]is the concentration of the adduct of (20) with T. At 40°C in benzene kT(PBN)= 1.34x lo51 mol-' s-' and k,(MNP) =90.2 X lo51mol-' s-l. The authors suggested extensions of the technique. Other spin-trapping studies include investigations of hydrogen atom abstraction from polystyrene by Bu'O.radicals,60 micellar catalysis of radical reactions,61 and superoxide-alkyl halide reactions.62 56 F. P. Sargent and E. M. Gardy J. Chem. Phys. 1977,67 1793. ''(a)D. B. Hibbert A. J. B. Robertson and M. J. Perkins J.C.S. Faraday I 1977.73 1499; (6)E. G. Janzen Creat. Detect. Excited State 1976,4 83. 58 F. P. Sargent J. Phys. Chem. 1977,81 89. 59 P. Schmid and K. U. Ingold J. Amer. Chem. Soc.,1977,99 6434. 6o N. Ohto E. Niki and Y. Kamiya J.C.S. Perkin ZZ 1977 1416. 61 D. P. Bakalik and J. K. Thomas J. Phys. Chem. 1977,81 1905. 62 M. V. Meritt and R. A. Johnson J. Amer. Chem. Soc.,1977,99,3713. Reaction Mechanisms-Part (iii) Electron Spin Resonance 2 Chemically Induced Dynamic Electron Polarization There have been two reports of S-T, polarization in CIDEP.63*64 Both involve the use of time-resolved e.s.r.typically 2 ps after a radiolysis pulse. In the first,63 -CH2C02H was produced in H20/H2S04 solutions of sodium acetate the other important species being He. At pH 1.3 the low-field line of the triplet from *CH2C02H was in enhanced emission the central line in emission and the high- field line in enhanced absorption (but considerably less intense than the other two lines). However in D20/H2S04 solution an essentially 'normal' CIDEP spectrum was obtained i.e. low-field enhanced emission unpolarized central line and the high-field line in enhanced absorption more intense than the low-field line. The other radical present was Do. These results were explained in the following way.Substantial hyperfine coupling in H*was responsible for making S-T- polariza-tion feasible for radical pairs involving H* atoms. This coupling splits the T+,and T-l levels while Tois unaffected in first order. Thus not only is the S-T-l energy i gap substantially reduced but differentiation of the T,,l levels allows S-Tkl polarization pathways to contribute to the usual S-To polarization. Since Do has a much smaller hyperfine coupling (7.754 mT) than has H* (50.66 mT) there was no appreciable S-T- mixing observed from radical pairs involving Do. Similar mixing was observed by the same authors in an investigation of CIDEP in the pulse radiolysis of aqueous solutions of micelle~.~~ N20-saturated solutions of several anionic and cationic surfactants were studied.Below the critical micelle concen- tration (CMC) the radical pair S-To mixing was dominant in all cases. This mechanism still obtained above the CMC since micelle formation is a dynamic equilibrium but emission and enhanced absorption were observed with emission in substantial excess. Especially noteworthy was the emission from central lines in the spectra (radicals were of the general type R'CH2CHCH2R2). This behaviour noted only above the CMC was ascribed to S-T- mixing consequent upon restricted diffusion in the aggregate phase. CIDEP has been observed in radicals (21) and (22) when solutions of maleimide and the N-ethyl and N-methyl derivatives in various alcohols AH were examined using a pulse photolysis-signal averaging technique.65 The observed spectra showed contributions from both radical pair and triplet mechanisms.The results were accommodated by the mechanism given in equations (30a-e). hv ISC M(S0) M*(S) M*(T) (304 63 A. D. Trifunac and D. J. Nelson J. Amer. Chem. SOC.,1977,99 289. 64 A. D. Trifunac and D. J. Nelson Chem. Phys. Letters 1977,46,346. " P. B. Ayscough T. H. English G. Lambert and A. J. Elliot J.C.S. Furuduy I 1977,73 1302. 104 A. T. Bullock M*(T)+AH -+ MH+A (30b) A+M(s~)__* MA MA,MH -products MA MH +M(So) -polymer (30c) (304 (30e) Evidence that the triplet state was involved in the H-abstraction step (30b) was inferred from the initial polarization of the MH radicals. The counter-radical A reacted within microseconds adding to the double bond of M(So) as evidenced by the transfer of its polarization to MA.Kinetic measurements suggested that MA and MH have second-order termination rate coefficients > lo91mol-' s-l in alco- holic solvents at room temperature and may also undergo further addition to maleimide [reaction (30d)J. Finally a nitrogen laser coupled to a time-resolved e.s.r. spectrometer has been used to study the sensitized and unsensitized photoreductions of biacetyl with triethylamine.66 It was found that the amine quenched both singlet and triplet biacetyl. The reaction of the triplet resulted in efficient radical production but was not rapid enough to produce significant electron polarization. The sensitizer used was benzophenone.In this case polarized spectra of biacetyl radical anions were observed and shown to arise in the following manner. Rapid reaction of the triplet benzophenone with the amine gave rise to polarization in the primary radical MeeHNEt2. This polarization was then transferred via reaction with ground state biacetyl to the secondary radical anions. Estimates of the initial polarization of both primary radicals MeCHNEt and Ph2COH showed these to be equal thus providing the first quantitative evidence for this requirement of the triplet mechanism. K. A. McLaughlan R. C. Sealy and J. M. Wittman J.C.S. Faraday ZZ,1977,73,926.