4 Reaction Mechanisms Part (iii) Free-radical Reactions By R. A. JACKSON School of Chemistry and Molecular Sciences University of Sussex Brighton BN1 9QJ 1 General There has been a flurry of activity on magnetic effects on radical reactions. Three different groups photolysed dibenzyl ketone in solution .' Cage products (recovered dibenzyl ketone and an isomer benzyl p-tolyl ketone) were enriched in I3C whereas radical escape products (dibenzyl CO) were impoverished in I3C. The large magni- tude of the effect which is enhanced by low temperatures and medium viscosities suggests that nuclear magnetic moment and/or magnetic spin isotope effects are involved rather than a kinetic mass isotope effect. The yields of n-butyl t-butyl ether and octane from n-butyl-lithium and t-butyl peroxide were decreased when the reaction was carried out in a magnetic field,* and a magnetic effect was also observed in the photo-induced substitution reaction of 4-methylquinoline-2-carbonitrile in ethanol3 (Scheme 1).The heats of formation of simple aliphatic radicals are still not satisfactorily established. New values of AH,.,, for Et* = 117 f4 and But' = 39 f4 relative Me Me I I + H H EtOH + M~~HOH 1 Me Scheme 1 N. J. Turro Ming-Fea Chow Chao-Jen Chung and B. Kraeutler J. Am. Chem. SOC., 1981,103,3886; N. J. Turro D. R. Anderson Ming-Fea Chow Chao-Jen Chung and B. Kraeutler ibid. p. 3892; L. L. Sterna Report 1980 LBL-10594; Energy Res. Abstr. 1981,6 13 (Chem. Abstr. 1981,95 114 602); A. L. Buchanchenko V.F. Tarasov and V. I. Mal'tsev Zh. Fiz. Khim. 1981 55 1649. Yu. A. Kurskii Yu. N. Baryshnikov G. I. Vesnovskaya N. N. Kaloshina and Yu. A. Alexandrov Dokl. Akad. Nauk SSSR,1981,258,936. N. Hata and M. Hokawa Chem. Lett. 1981,507. 81 R. A. Jackson to Me- = 144 kJ mol-’ have been obtained by e.s.r. measurements of concentration in radical buffers of R* + R’I; the higher value for Et* reconciles conflicting data on the self-reaction of ethyl radical^.^ AHf,300 (But*) has been evaluated as 44 f 4kJ mol-’ from kinetic measurements in both directions of the reaction of hydrogen atoms with isob~tene,~ in reasonable agreement with the radical buffer results. The question of separation of polar and resonance effects of substituents on radical reactions continues to arouse interest.Afanas’ev favours a two-parameter equation for correlating radical reactivity of aromatic compounds.6 u;I values are derived for a substituent X from the effects of X on the spin density of CH,CHX but these values have to be applied in the negative sense for electron-releasing substituents and in the positive sense for electron-withdrawing groups. This Reporter7 has suggested the thermal decomposition of substituted dibenzyl mer- curials [equation (l)]as a model reaction in which the full stabilizing effect of the substituent on the benzyl radical formed should be available in the transition state. Polar effects are allowed for by decomposing meta-substituted compounds and a scale is defined for para-substituents by equation (2).(TO values correlate well DCH2-+ *Hg-CH X with literature data and an excellent correlation is found with Hammett polar substituent constants via the expression (lu*-ul)/n,where u*is u+for electron- releasing and u-for electron-withdrawing substituents and n = 1 for radicals conjugated with an unsaturated centre or n = 2 for substituents conjugated with a lone pair. In the latter case the stabilization for a radical should be approximately half that for a cation whereas in the former case stabilization should be approxi- mately the same in the radical the cation or the anion (see Figure 1).It should n = 1 electron in non-bonding n = 2 electron in anti-bonding orbital in radical orbital in radical Figure 1 Molecular orbital diagrams for the two types of substituted radicals A.L. Castelhano P. R. Marriott and D. Griller J. Am. Chem. SOC.,1981,103,4262. C. E. Canosa and R. M. Marshall Int. J. Chem. Kinet. 1981,13 303. I. B. Afanas’ev Int. J. Chem. Kinet. 1981 13 173. ’ S. Dincturk R. A. Jackson M. Townson H. AgirbaS N. C. Billingham and G. March J. Chem. SOC. Perkin Trans. 2 1981 1121; S. Dincturk and R. A. Jackson ibid. p. 1127. Reaction Mechanisms -Part (iii) Free-radical Reactions 83 be emphasized however that in many perhaps most free-radical reactions polar rather than radical-resonance interactions appear to be the dominating influence. 2 Structural Studies Two separate groups report UHF calculations on the t-butyl radical with the 4-31G basis set.8 The radical is calculated to be markedly non-planar with each Me-C* bond predicted to be 22.1" out of the plane defined by the radical centre and the other two methyl group carbon atoms.The inversion barrier is calculated to be 5.1 kJmol-'. The geometry of the radical centre in bicyclo[2.2,l]hept-2-y1 systems has been debated recently. Group IVB (Si Ge Sn) radical adducts to camphor and thiocam- phor are intense enough for the 13C satellites to be observed in natural ab~ndance.~ The results suggest non-planarity at the radical centre but comparison with non-cyclic analogues indicates that the non-planarity is due to the substituents rather than to bond-angle strain. Non-planar geometry is also indicated at the radical centre of 1-hydroxycyclohexyl radicals and substituted derivatives by variable temperature e.s.r.and '3C-labelling studies." The interconversion of the two geometrical isomers of the [1-2H]allyl radical made by the reaction of t-butoxyl radicals with deuteriated alkyl phosphites," has been studied in the temperature range 50-1 10"C [equation (3)].The rotational barrier is 65.7 f 4.2 kJ mol-' (log A = 13.5 f OS) suggesting a value for the ally1 delocalization energy of 58.6-60.7 kJ mol-'. Pent-2-en-4-ynyl radicals can exist in forms (1)and (2) where (1)is just the more stable methane-based stabilization energies for (1)and (2) were estimated to be 112 and 110 kJ mol-' respectively from the observed barriers to rotation r47.5 and 49.2 kJ mol-' for reactions (4) and -(4) respectively].'*" H I (1) (2) [Me3Si-SiMe3]t and [Me3Ge-GeMe3]t have been made by y-irradiation of solid solutions at 77 K.In contrast with [Me3C-CMe3]t [cf. Ann. Rep. Prog. Chem. Sect. B 1980 77 541 which showed a septet indicating a strong interaction with one hydrogen only of each methyl group the silicon and germanium analogues show M. N. Paddon-pow and K. N. Houk J. Am. Chem. SOC. 1981 103 5046; M. Yoshimine and J. Pacansky J. Chem. Phys. 1981,14,5168. A. Alberti M. Guerra and G. F. Pedulli J. Am. Chem. SOC.,1981 103 6604. R. V. Lloyd and J. G. Causey J. Chem. SOC.,Perkin Trans. 2 1981 1143; J. C. Micheau B. Despax N. Paillous A. Lattes A. Castellano J. P. Catteau and A. Lablache-Combier Nouu. J. Chim. 1981 5 257. H.-G. Korth H. Trill and R. Sustmann J. Am.Chem. SOC., 1981,103,4483. (a) C. Roberts and J. C. Walton J. Chem. SOC., Perkin Trans. 2 1981 553; (b) J. T. Wang and F. Williams J. Chem. SOC., Chem. Commun. 1981,666. 84 R. A.Jackson spectra that indicate that interaction occurs with all 18 hydrogen atoms suggesting that the unpaired electron is effectively localized in the M-M u-bonding orbital."' E.s.r. parameters for a range of transient tin-centred radicals have been reported.' In alkyltin radicals R3Sn- @-proton couplings of 3.0-3.1 G are noted; for radicals Ph,Me,-,Sn* only coupling to the methyl protons is observed. Line broadening takes place owing to the rapid exchange reaction of trialkyltin radicals with the corresponding hydride this is confirmed by the observation of line broaden- ing in the n.m.r.spectrum of Me3SnH during irradiation with t-butyl peroxide. A structure with partial phenyl bridging is inferred for the benzyldimethylgermyl radical on the basis of the small temperature-variable CH2 splittings compared with the CH3 proton ~p1ittings.l~ t-Butoxyl radicals abstract hydrogen from BH,- or BH3CN- in fluid solution to give the radical anions BH,' and BH,CN' respectively;" for BH3' lines due to 10 B as well as to "B can be observed. The relatively small "B coupling constants (19.9G and 14.3G respectively) confirm that both species are essentially planar. BH3' resembles silicon-centred radicals in some respects BH,' abstracts halogen readily from n-propyl halides (PrC1 PrBr PrI) and displaces alkyl radicals from alkyl isocyanides.The AlH3' radical prepared analogously,16 has a (27Al)= 154.2G which indicates a divergence from planarity similar to that found for SiH3'. The AlH,' radical anion abstracts halogen from Pr"-Hal (Hal = C1 Br or I) and adds to ethylene and benzene to give the corresponding substituted ethyl and cyclohexadienyl radicals. Flash photolysis of N-bromo-3,3-dimethylglutarimidein CCl gave a radical postdated to be the imidyl radical (3) its U.V. and rate of quenching by alkene or alkane is inconsistent with identification of the transient as 'CCl (Scheme 2). The 'precursor' may be an excited state of the precursor molecule or another electronic state of the imidyl radi~a1.l~ 'W0 (3) Scheme 2 Several amidyl radicals RNCOR have been studied by ex.and show a 7r structure in which the electron resides mainly in a N 2p orbital perpendicular to the R-N-C plane.18 Cyclic and acyclic oxyamidyls have been preparedlg and their 170enrichment e.s.r. studies show less delocalization onto the C=O oxygen atom in PhCONOEt compared with *CHCO(CH2$4. A reinvestigation of the structure of 1,3-dialkyltriazenyl radicals indicates that these radicals have a u structure with the electron occupying an anti-bonding three-centre l3 M. Lehnig and K. Doren J. Organomet. Chem. 1981 210 331. '* K. Mochida M. Kira and H. Sakurai Chem. Lett. 1981 645. Is J. R. M. Giles and B. P. Roberts J. Chem. Soc. Chem. Commun. 1981,360. l6 J. R. M. Giles and B. P. Roberts J. Chem. SOC.,Chem. Commun. 1981 1167. R. W. Yip Y.L. Chow and C. Beddard J. Chem. Soc. Chem. Commun. 1981,955. l8 R. Sutcliffe D. Griller J. Lessard and K. U. Ingold J Am. Chem. Soc. 1981 103 624. l9 A. R. Forrester and H. Irikawa J. Chem. Soc. Chem. Commun. 1981,253. Reaction Mechanisms -Part (iii) Free-radical Reactions molecular orbital with greatest spin density on the central nitrogen atom rather than the T structure with the spin density located mainly on the terminal nitrogen atoms as previously suggested.20 Irradiation of the powdered phosphorane at 77 K gave the phosphoranyl radical (4). The electron is equatorial and M-1 Berry pseudorotation takes place with the unpaired electron acting as the pivot. A single-crystal e.s.r. study shows that at 77 K the equatorial radical (5) is formed but on warming to 193 K isomerization to the axial form (6) takes place (Scheme 3).2' Phosphoranyl radicals OCH2CH20$(OR')SR have trigonal bipyramid structures with the SR group and (4) Scheme 3 one of the bridging groups in apical positions.Line-width effects show that apical- equatorial exchange takes place probably uia a (T* intermediate or transition state.22 Dialkyltriazenyl radicals add to trimethyl phosphite to give the cyclic radical (7a). Above 280 K exchange between (7a) and (7b) is fast enough to make N' and N3 magnetically equivalent (Scheme 4).23 ABut-NNN-Bu' +(Me0)3P OMeI .OMe + Bu'-N'-PP(I I OMe NZ-N~-B~~ Me0 ,OMe\ #'5 But-N' -P-OMe I I.N~-N~-BU' (7a) (7b) Scheme 4 A series of overcrowded galvinoxyl radicals has been made.For compounds such as t-butylgalvinoxyl (8 R = But) the bulky t-butyl group attached to the central carbon atom prevents coplanarity of the two rings. Thus (8a) and (8b) are in true equilibrium in contrast to galvinoxyl itself (8,R = H) where structures similar to (8a) and (8b) (but with the benzene rings coplanar) are mere contributions to the overall Photolysis of acetyl hypohalites (CH3C02X X = C1 or Br) in the presence of the corresponding halogen neopentane and dichloromethane gives substantial amounts of CHC1,X and neo-C5HI1X with a selectivity of about 12 in each case. 22 J. C. Brand and B. P! Roberts J. Chem. SOC.,Chem. Commun. 1981,748. " J. H. H. Hamerlinck P. H. H. Hermkens P. Schipper and H. M. Buck J. Chem. SOC.,Chem. Commun. 1981 358; J.H. H. Hamerlinck P. Schipper,.and H. M. Buck ibid. p. 1148. ''J. R. M. Giles and B. P. Roberts J. Chem. SOC.,Perkin Trans. 2 1981 1211. 23 J. C.Brand and B. P. Roberts J. Chem. SOC.,Chem. Commun. 1981 1107. 24 B. Kriste W. Harrer H. Kurreck K. Schubert H. Bauer and W. Gierke J. Am. Chem. Soc. 1981 103,6280. R. A. Jackson In the absence of halogen and with 1,l-dichloroethylene as a trap virtually no halogenation of the neopentane or dichloromethane takes place. It is suggested that two different types of acetoxy radicals may be involved one produced in the presence of halogens which can escape from the cage and abstract hydrogen atoms (probably the 2A27r or 2Bou state) the other produced in scavenged systems which undergoes decarboxylation too fast to take part in hydrogen-abstraction reactions (2A1cr 3 Formation Destruction and Radical Stability The formation of free radicals by a 'spontaneous' reaction between nitroso- compounds and other organic compounds such as alcohols or ethers has often been ascribed to molecule-assisted homolysis (MAH).However Chatgilialoglu and Ingold have shown that in several such reactions the radicals are formed by simple photolysis of the nitroso-compound by laboratory light and it is suggested that the role of light on other 'spontaneous' radical-forming reactions should be investi- gated.26 On the other hand a molecular beam of the reaction between fluorine and methane at 300-400 K indicates that the dominant initiation reaction is the MAH reaction (9 with log(A/l mol-'s-') = 9.3 f 0.3 and E = 47 f 8 kJ mol-'.Stannous chloride is reported to react reversibly with aqueous methylcobalamin by methyl-group abstraction to give CH3SnC12* [equation (6)].276 CH4 + F2 + CH3. + H-F + F* (5) SnC12 + CH3-B12 $ CH3SnC12.+ B12r (6) Very-low-pressure pyrolysis (VLPP) of ethylbenzene isopropylbenzene and t-butylbenzene takes place by p-C-C bond homolysis.28 Derived AH values are +166 kJ mol-' for PhCHMe- and +136 kJ mol-' for PhCMe2* in good agreement with toluene and aniline carrier experiments. VLPP of the appropriate compound R-CH3 gave the substituted propargyl radicals [Re= HCrCCHCH, 25 P. S. Skell and D. D. May J. Am. Chem. SOC.,1981,103,967. 26 C. Chatgilialoglu and K. U. Ingold J. Am. Chem. Soc.1981,103,4833. " (a) C.Seeger G. Rotzoll A. Lubbert and K.Schugerl In?.J. Chem. Kine?. 1981 13 39. (b)Y.-T. Fanchiang and J. M. Wood J. Am. Chem. SOC., 1981,103,5100. 28 D.A.Robaugh and S.E. Stein Int. J. Chem. Kinet. 1981,13,445. Reaction Mechanisms -Part (iii) Free-radical Reactions 87 CH3C=CCHCH3 and CH,CrCC(CH,),] and the 1-methylallenyl radical CH3C=C=CH2. Stabilization energies are 42 f 9 32 f 12 41 f 12 and 33 f 9 kJ mol-' indicating that methyl substituents on propargyl radicals have little influence on stability.29 A shock tube reinvestigation of the thermal decomposition of ethane has been carried out [equation (7)].30 An RRKM model is produced giving excellent agree- ment with other recent data with log k? = 17.2 -380.7 kJ/2.3RT and log k7( 1mol-' s-') = 10.44.CH3-CH3 + 2CH3. (7) U.V. irradiation of pentamethylcyclopentadiene gives the pentamethylcyclo- pentadienyl radical in an unusual reaction involving homolysis of the ring C-H bond.3' A nanosecond laser flash photolysis study of the photosensitized reaction of t-butyl peroxide by ketones or aromatic hydrocarbons such as benzophenone or benzralanthracene suggests that triplet energy transfer to the peroxide leads to a repulsive excited state that decomposes into two Bu'O- radicals.32 The t-butoxyl radical has been generated in aqueous solution by the reaction between Bu'OOH and Ti"' at very low pH the protonated form Bu'OHt is im~licated.~~ Twenty-one bridgehead peresters RC020Bu' have been thermolysed. A transi-tion state of type (9) is invoked and it is thought that the thermolysis constants are determined mainly by the polar effect in the transition state [equation (8)].34 s+ s-RCOOBU' -B [R--C-I-O--OBU']+ R.+ C02 + *OBu' (8) II II 0 0 Di-tin compounds of the type R3Sn-SnR3 (R = Ph cyclohexyl 1-adamantyl) do not dissociate into R3Sn* radicals at temperatures up to 230 "C but for the more hindered compounds R = 2,4,6-trialkylphenyl (alkyl = Me Et Pr') the dissoci- ation temperature can be as low as 20°C with a dissociation energy of only 36 kJ mol-' for the isopropyl omp pound.^' Cobalamins with an organic group attached to cobalt possessing a P-hydrogen atom decompose by p-elimination in the absence of such a hydrogen atom compounds such as benzyl- or neopentyl- cobalamin decompose spontaneously with Co -C bond homolysis.In the absence of air recombination takes place but in the presence of air the radicals are rapidly oxidized making these compounds highly air-sensitive in aqueous Several groups have carried out photochemistry in micelles :xanthone phenyl alkyl ketones dibenzyl ketone and benzyl phenylacetate have been in~estigated.~' 29 T. T. Nguyen and K. D. King J. Phys. Chem. 1981 85 3130; K. D. King and T. T. Nguyen Znt. J. Chem. Kinet. 1981 13 225. 'O G. B. Skinner D. Rogers and K. B. Patel Int. J. Chem. Kinet. 1981 13 481. 31 A. G. Davies and J. Lusztyk J. Chem. SOC.,Perkin Trans. 2 1981 692. 32 J. C. Scaiano and G. G. Wubbels J.Am. Chem. SOC.,1981,103,640. 33 B. C. Gilbert P. D. R. Marshall R.0. C. Norman N. Pineda and P. S. Williams J. Chem. SOC. Perkin Trans. 2 1981 1392. 34 C. Ruchardt V. Golzke and G. Range Chem. Ber. 1981,114 2769. 3s H.-U. Buschhaus W. P. Neumann and T. Apoussidis Liebigs Ann. Chem. 1981 1190. 36 G. N. Schrauzer and J. H. Grate J. Am. Chem. SOC.,1981,103 541. '' J. C. Scaiano and J. C. Selwyn Can. J. Chem. 1981,59 2368; N. J. Turro Min-Fea Chow Chao-Jen Chung Y. Tanimoto and G. C. Weed J. Am. Chem. SOC.,1981,103 4574; D. Avnir L. J. Johnston P. De Mayo and S. K. Wong J. Chem. SOC.,Chem. Commun. 1981,958. R. A. Jackson In view of the importance of t-butyl peroxide as a radical initiator the rate constant for the 'self-reaction' for the t-butoxyl radical is of considerable interest. The absence of an e.s.r.signal from the t-butoxyl radical makes 2kt difficult to determine directly but by setting up a competition between the self reaction and hydrogen abstraction a value of 2kt = (1.3 f0.5) x lo9M-* s-l at 293 K has been obtained in a flash photolysis-e.s.r. e~periment.~~ This value is close to the diffusion- control limit. The decay of radical (10) is not thought to be diffusion controlled there is an activation energy of 17 kJ mol-' and an intermediate is postulated on the route to the product pinacol (Scheme 5),39 Scheme 5 4 Radical Transfer Fluorination of organic compounds by molecular fluorine is a reaction that is difficult to control. Some of the difficulties are surmounted by fluorinating the hydrocarbon as an aerosol suspension in helium at low temperatures thereby allowing high yields of perfluorinated compounds to be made from the corresponding hydrocarbon with minimum disruption of the carbon ~keleton.~' High positional selectivity of chlorination of hexane at the 2-position is achieved by photolysis with [Pr;NMeC1]'C1O4- (the ratios of 1-:2-:3-chlorohexanes = 7.1 :76.6 :16.5).The enhanced reactivity at the 2-position is attributed to steric difficulty of approach by the bulky aminium radical to the 3-po~ition.~~ The decomposition of a steroidal peracid has been used to achieve a remote hydroxyla- tion (Scheme 6).42 There has been a continuing interest in determining absolute values of rates of radical-transfer reactions. Time-resolved e.s.r. measurements show that the decay of t-butyl radicals in methylcyclopentane solution is of second order but in the presence of chloroform a pseudo-first-order term is added from which the rate constant for the abstraction of a hydrogen or a chlorine atom from chloroform by the t-butyl radical is deduced.43 Laser-flash photolysis has been used to determine the rates of reaction of triethylsilyl radicals with alkyl halides (halogen abstraction) and with ketones (addition to the carbonyl oxygen atom).44 38 S.K. Wong Int. J. Chem. Kinet. 1981 13 433. 39 H. Krohn R. Leuschner and J. K. Dohrmann Ber. Bunsenges. Phys. Chem. 1981,85 139. 40 J. L. Adcock K. Horita and E. B. Renk J. Am. Chem. SOC.,1981,103,6937. 41 S.E.Fu1ler;J. R. Lindsay Smith R. 0.C. Norman and R. Higgens J. Chem. Soc.Perkin Truns. 2 1981,545. 42 J.-P. BCguC D. Lefort and T. D. Thac J. Chem. Soc. Chem. Commun. 1981,1086. 43 H. R.Deutsch and H. Fischer Int. J. Chem. Kinet. 1981,13,527. 44 C.Chatgilialoglu K. U. Ingold J. C. Scaiano and H. Woynar J. Am. Chem. Soc. 1981,103 3231. Reaction Mechanisms -Part (iii) Free-radical Reactions (12) + Scheme 6 Hydrogen abstraction by t-butoxyl radicals from the hydroxyl hydrogen of sub- stituted phenols indicates that positive charge is built up on the aromatic system in the transition state. The reaction is slower in polar solvents such as pyridine hydrogen bonding decreases the reactivity of the phenolic OH group.45 Abstraction of chlorine from several N-and O-heteroarylmethyl chlorides by triphenyltin radicals suggests a transition state with contributions both from the heteroarylmethyl radical and the anion.46 Stereoelectronic factors are important in abstraction by t-BuO' radicals from C-H bonds adjacent to oxygen or nitrogen the rate of abstraction is enhanced if the dihedral angle between the C-H bond and the N or 0 lone pair is Tunnelling appears to be important in the abstraction of hydrogen from toluene by bis(trifluoromethy1)nitroxide:the Arrhenius A factor is unusually low and the difference between the activation energies for deuterium compared with protium abstraction is unusually large.48Q Tunnelling is also invoked4*' to explain the difference in AV' observed in the rearrangement of 2,4,6-(Me3C)3C6H2* by transfer of a hydrogen atom from the t-butyl group to the ring (5.3f 1.7 cm3 mol-l) compared with the deuteriated compound 2,4,6-[(C2H3)3C]3C6H*.(-1.2 f 2.0 cm3 mOl-').Photolytic brominolysis of trans-1,2-diarylcyclopropanes involves S,2 attack by a bromine atom at a strained-ring carbon atom (Scheme 7).49The reaction rate is B. + ArAH 5 ArCHBrCH,CHBrAr + Br. H" "Ar Br Scheme 7 enhanced by electron-donating substituents on one or both rings. SH2 attack at saturated oxygen centres appears to be more difficult than at sulphur. However trichlorosilyl radicals react with cyclohexyl octyl ether to give a preponderance of '' P. K. Das M. V. Encinas S. Steenken and J. C. Scaiano J. Am. Chem. Soc. 1981,103,4162. 46 H.Soppe-Mbang and G. J. Gleicher J. Am. Chem. Soc. 1981,103,4100. 47 D. Griller J.A. Howard P.R. Marriott and J. C.Scaiano,J.Am.Chern. SOC.,1981,103,619;V. Malatesta and K. U. Ingold ibid. p. 609. '* (a)V. Malatesta and K. U. Ingold J. Am. Chem. SOC.,1981,103,3094; (6)P.R.Marriott and D. Griller ibid. p. 1521. 49 D. E. Applequist and R. D. Gdanski J. Org. Chem. 1981,46,2502. R. A. Jackson HSiCI, CI3Si’ + OC8Hl + Cl3SiOC8Hl7 + 0. -0 Scheme 8 cyclohexane in accord with Scheme 8 where the more stable cyclohexyl radical is displaced in preference to the octyl radi~al.~’ 5 Addition to Multiple Bonds and Homolytic Aromatic Substitution Alkoxyl alkyl and acyl radicals add to the P=N bond according to reaction (9) the 31P coupling constant of 519G for (13 X = Bu‘O) indicates a pyramidal geometry for the radical.’l Cyclization of the 5-hexen-1-yl radical (14 X = Y = CH,) gives virtually exclusively the less thermodynamically stable cyclopentylmethyl (Me3Si)*N -P=NSiMe3 + X .-+ (Me3Si),N -f’=NSiMe3 (9) I X (13) (14) (15) (16) radical (15 X = Y = CH,). Experiments have now been carried out with analogues containing a Me2Si group at position X or Y. For (14 X = Me,Si Y = CH,) the 5-ring product (15) still predominates but for 14 (X = CH, Y = Me2Si) the 6-ring product (16) dominates the product ratios indicating that it is a reduction in the rate of cyclization to the 5-ring that gives this res~lt.’~ The longer Si-C bond length may make the transition state leading to the 5-ring less favourable. 1-Alkenylmercuric halides react photolytically with organic disulphides to give the 1-alkenyl~ulphide.~~ Addition to the 1-position to give the intermediate radical (17) which may be stabilized by bridging is followed by loss of the mercurio- substituent [equation (lo)].The water soluble nitroso-compound (18 X = H or ,H) has been proposed as a spin trap for aqueous An advantageous simplification of the spectra is achieved in the deuteriated spin trap (18 X = ,H). A study of the kinetics of the reaction of hydroxyl radicals with benzene and toluene shows that below room temperature electrophilic addition to the aromatic nucleus predominates. Above 500 K reversal of the addition reaction reduces the RCHZCH-HgCl + PhS -B RCH-CH-SPh + RCH=CHSPh + .HgCI (10) *.. I *-HgCl (17) 50 R. A. Jackson F. Malek and N.Ozaslan J. Chem. SOC., Chem. Commun. 1981,956. ’* B.P.Roberts and K. Singh J. Chem. SOC.,Perkin Trans. 2 1981 866. 52 J. W.Wilt J. Am. Chem. SOC.,1981,103 5251. 53 G.A. Russell and J. Hershberger J. Am. Chem. SOC., 1980 102,7603. 54 H.Kaur K. H. W. Leung and M. J. Perkins J. Chem. SOC.,Chem. Commun. 1981 142. Reaction Mechanisms -Part (iii) Free-radical Reactions SO<Na+ Br NO (18) amount of aromatic substitution product and abstraction of ring or side-chain hydrogen atoms becomes competiti~e.~~ Competition between ring hydroxylation and side-chain attack has also been studied in the presence of 0,or NO to convert the intermediates into stable Thermal decomposition of Na&08 in aqueous solutions of benzene and nitrobenzene gives phenol biphenyl and 0-and p-nitrophenol.The nitrophenols are not formed in the absence of benzene. Steps (a) and (b) in Scheme 9 are known to be rapid. It is proposed that the hydroxycyclohexadienyl radical can dissociate [step (c)] to benzene and the hydroxy radical which in turn attacks the nitrobenzene to give the nitro phenol^.^' Scheme 9 Improved partial rate factors for homolytic aromatic arylation have been obtained by carrying out the arylation by benzoyl peroxide in the presence of C6H5N0 C6F5N0 or Cu" or FeIII ben~oate.'~ Oxidation of the initial u complex is almost quantitative thus eliminating uncertainties in partial rate factors caused by dimeriz- ation. For eight substituents partial rate factors for the meta-position were close to unity (0.85-1.35); at the para-position all were greater than 1and ranged from 1.1(F)to 4.0 (COPh).6 Fragmentation Arrhenius parameters for alkoxy radical fragmentations have been re-evaluated. For the important reaction (ll) values of log, A(s-') = 14.1 and E = 64.0 kJ mol-' have been ~ecommended.~~ Substituted 9-decalinoxyl radicals frag- ment to give both cyclohexyl and cyclodecyl derivatives which bond cleaves seems to depend on a number of factors including substituent stabilizing effects and the cyclic ketone ring strain.60 (CH3)3C-O. + (CH3)2C=O + CH3. (11) 55 F. P. Tully A. R. Ravishankara R. L. Thompson J. M. Nicovich R. C. Shah N. M. Kreutter and P. H. Wine J. Phys. Chem. 1981,85 2262. 56 R. A. Kenley J. E.Davenport and D. G. Hendry J. Phys. Chem. 1981,85,2740. 57 M. K. Eberhardt J.Am. Chem. SOC.,1981,103 3876. " R. Bolton B. N. Dailly K.Hirakubo K. H. Lee and G. H. Williams J. Chem. SOC.,Perkin Trans.2 1981,1109. '9 Kwang Yul Choo and S. W. Benson Int. J. Chem. Kinet. 1981,13,833. 6o T. L. Macdonald and D. E. O'Dell J. Org. Chem. 1981,46 1501. R. A.Jackson Scheme 10 Cyclobutylmethyl 1-cyclobutyl-1-methylethyl and cyclobut-2-enylmethyl radicals (19) undergo ring-opening reactions (Scheme 10) with log, A(s-')in the range 12.2-13.6 and E/kJ mol-' 42-58. Ring opening of (19) gave the E E-pentadienyl radical (21) presumably via the E 2-form (20). Independent experi- ments show that (20) can be converted into (21) but not vice versa from the rotation barrier a methane-based stabilization energy of 104 kJ mol-' was esti- mated.61 Glycidol derivatives (22 M = R or R3Si) react with t-butoxyl radicals as shown in Scheme ll(a).Ring opening appears to be followed by loss of [H'] and attack by ROO gives the observed acyl radical (24). Where M = acetyl an alternative reaction (b) takes place acyl transfer allows radical (25) to be observed.62 O=CH-CH=CHOM T7-I + c? 0 OM 0 OM M (22) (23) (25) Scheme 11 Tributyltin hydride causes elimination from vicinal dinitro-compounds or p-nitro- sulphones to give the corresponding alkene. The reaction is stereospecific (anti) in the latter but not in the former case. Scheme 12 involving radical anion intermedi- ates is po~tulated.~~ 7 Electron Transfer Both triphenylmethyl cations and anions undergo electron transfer with di-t- butylnitroxide to give triphenylmethyl radicals.If oxygen is admitted to the anion system the di-t-butylnitroxide is regenerated [equation (12)] but when the cation is the precursor the degradation reactions (13) and (14) take place.64 61 K. U. Ingold B. Maillard and J. C. Walton J. Chem. SOC.,Perkin Trans. 2 1981 970; A. G. Davies D. Griller K. U. Ingold D. A. Lindsay and J. C.' Walton ibid.,p. 633. 62 A. G. Davies J. A.-A. Hawari B. Muggleton and Man-Wing Tse J. Chem. SOC.,Perkin Trans. 2 1981 1132. 63 N. Ono H. Miyake R. Tamura I. Hamamoto and A. Kaji Chem. Lett. 1981 1139. 64 H. Singh and J. M. Tedder J. Chem. SOC.,Chem. Commun. 1981,70. Reaction Mechanisms -Part (iii) Free-radical Reactions 93 \ / Bu3SnH or Bu3Sn-+ / + Bu3SnHor Bu3Sn+ + C-C /I -I\ NO; X 'C-C / -+\C-C / +NO X X.+ Bu3SnH + X-H + Bu3Sn-Scheme 12 02 Ph3C-+ BU~NO -+ Bu~N-0-+ Ph3C.++ Ph3COO. B~;NO-+Ph3COO-+ Bu~NO. 02 Ph3C+ + Bu~NO+ Bu$=O + Ph3C. __* Ph3COO-B~;NO. -+ Ph3COO-+ Bu$ = 0 0 II Ph3COO- H-CH2-CMe2p7-But + Ph3COOH + CH2=CMe2 + Bu'NO (14) Nitro-compounds are reduced to alkanes by trialkyltin hydrides in the presence of radical initiators. Electron transfer followed by fragmentation and hydrogen transfer are postulated [reaction (15) compare scheme 12 for di-nitro corn-R~S~H + R. -R$n* + RNO2 -B R:Sn' + RNOS RH + R:Sn* (15) Radical cations can be used to catalyse the Diels-Alder reaction of neutral dienophiles;66 the dienophile is converted into the radical cation that then reacts more rapidly with the diene without loss of the suprafacial stereospecificity of the uncatalysed reaction.For example reaction (16) which uncatalysed gives a 30% yield at 200 "C after 20 h with 4 1 endo exo selectivity takes place in 70% yield at 0 "C in 15min in the presence of [(p-BrC6H4)3N'SbC15-] with 5 1 endo ex0 selectivity. (16) 65 D. D. Tanner E. V. Blackburn and G. E. Diaz J. Am. Chern. SOC.,1981,103,1557. 66 D.J. Bellville D. D. Wirth and N. L. Bauld J. Am. Chem. SOC.,1981,103 718. R. A. Jackson In principle Grignard addition to ketones can take place by a polar route or by single-electron transfer (SET)as shown in Scheme 13.Evidence for the participa- tion of the SET pathway both for primary and for tertiary Grignard reagents was obtained by the use of the unsaturated reagents RMgCl[R = CH2=CH(CHz)3CMez-or CHz=CH(CH2)zCMe2CH2-].In each case some cyclized products were isolated indicating that free R* radicals must have been involved in the reaction.67 ‘RMgX’ + ArzC=O1- Polar --+ ArzC=O %MgX [R + ArzC-OMgXl diffusion\:; reaction in cage ,Ar2C-OMgXIR R + ArzC-OMgX 1 11 RH ArzC-OMgX I Ar2C-OMgX Scheme 13 The SR,l reaction of the nitro-compound (26)with sodium benzenethiolate gives (27),with retention of configuration at high thiolate concentrations though at lower thiolate concentrations some inversion takes place. It is concluded that a pyramidal benzyl radical is formed and trapped (Scheme 14).68 BU,dr6H4N02 Bu‘&:6H4N02 H4N02 (27) Reagents i +e-; ii -NO;; iii PhS-; iv -e-Scheme 14 Russell and co-workers suggest that in the SRNlmechanism [equations (17)-(19)] steps (17) and (18)may merge so that R-is not formed as an intermediate.69 RX‘ -D R-+ X-(17) R*+ N- -D RNT (18) RNT+ RX -+ RXT+ RN (19) 67 E.C. Ashby and J. R. Bowers jun. J. Am. Chem. SOC.,1981,103,2242. 68 R. K. Norris and R. J. Srnyth-King J. Chem. SOC.,Chem. Commun. 1981,79. 69 G. A. Russell B. Mudryk and M. Jawdosiuk J. Am. Chem. SOC.,1981,103,4610. Reaction Mechanisms -Part (iii) Free-radical Reactions When enolate ions E = RC(0-)=CHR’ react with nitro-compounds XCMe2N02 the substitution product ECMe2N02 and the enolate dimer E-E are formed in ratios which do not depend on [XCMe,NO,] or [E-] but which depend strongly on the nature of X(X = C1 NO2 or p-MeC6H4S02).Decomposition of benzoyl peroxide with 4-fluoroanisole in KOAc-HOAc gives 4-methoxyphenyl acetate and benzoate in a maximum ratio of 10:1. Scheme 15 involving a radical cation intermediate is suggested. Initiation can take place with Nu* = PhC02* but in the presence of OAF Nu-= OAc-and the predominant product will be the acetate.” ArX + Nu-+ [ArXNu]. + ArN? + X-T I Scheme 15 Note Readers interested in a computer-readable version of the references in this Section for their own data systems are invited to contact the Reporter. ’O Chem. Commun. 1981 133. L. Eberson and L. Jonsson J. Chem.SOC.