J . Chem. SOC., Faraday Trans. I , 1985, 81, 565-578 Radiolytic Preparation of Radical Cations of Nitroalkanes and Related Compounds BY D. N. RAMAKRISHNA RAO AND MARTYN C. R. SYMONS* Department of Chemistry, The University, Leicester LE 1 7RH Received 2nd April, 1984 Exposure of dilute solutions of a range of organic nitro derivatives in trichlorofluoromethane to 6oCo y-rays at 77 K gave radical cations thought to be formed from the expected primary species by some form of relaxation. These cations have e.s.r. spectra similar to that of nitrogen dioxide, and may be the isomeric nitrite cations, RONO+. However, the form and magnitude of the 13C hyperfine coupling for cations derived from 13CH,N0, suggests that the alternative a-structure (R'NO,)+ may be a better description. Cations derived from CH,CH,NO,, CH,CH,CH,NO,, (CH,),CHNO,, (CH,),CHO,, (CH,),C(Cl)NO,, (CH,),C(Br)NO,, (CH,),C(CO,Et)NO, and [(CH,),CNO,], all had very similar e.s.r.spectra, with indications of extra hyperfine splitting from halogen nuclei for the chloro and bromo derivatives. The ethyl derivative was unique in giving a second species in comparable yields which exhibited very large hyperfine coupling and considerable g-value variation. This is tentatively identified as CH,CH,O' radicals complexed with NO+ cations, The ethyl ester derivative gave a second species thought to be the ester cation. Several nitro aromatic compounds gave the normal n-cations initially, but all gave evidence for the rearranged NO,-like cations on annealing. A derivative ArC(Me),NO, with a p-nitro group gave high yields of this NO,-like cation even at 77 K, with no sign of the n-aromatic cation.One vinyl derivative, 1 -nitrocyclohexene, was also studied. This gave a complex spectrum which is assigned to a cation with its SOMO primarily confined to the alkene group. However, conversion into the rearranged cation was again detected on annealing. There is considerable current interest in the study of reactive intermediates of all types.'* The technique of electron spin resonance is well suited for the study of radical intermediates, but it has only recently been applied to the study of radical cations. These have, of course, been extensively studied in the gas phase, but not by magnetic resonance techniques. Two important methods have now been developed for the generation of radical cations in solid matrices.One, apparently best suited to the study of small cations, involves photoionization of the parent molecules during the process of deposition onto a cold-finger using a rare-gas matrix. This has been used to great effect by Knight et al., who have prepared species such as H,O, 'NH3,4 H,C05 and 'CHi6 and studied their structures by e.s.r. spectroscopy. The other, developed originally by Hamill et al. for optical studies,' involves exposing cold, dilute solutions of a substrate in a suitable solvent to ionizing radiation. The most commonly used solvent for e.s.r. studies is trichlorofluoromethane, partly because of its high ionization potential (ca. 1 1.8 eV) and partly because the electron-capture centres formed therefrom give very broad e.s.r.features which do not normally interfere with those 565566 RADICAL CATIONS OF NITROALKANES for the substrate cations. The essential chemistry involved is summarized by the reactions CFCl, -+ CFCli + e- (1) (2) (3) (4) These reactions have been clearly established for a wide range of solutes (S), provided their ionization potentials are -c 1 1.8 eV. In many cases the expected cations have been detected, with no evidence for major modification by the matrix,8-11 but in others hyperfine coupling to 19F l2 or 35Cl and ,'CI 1 3 9 l4 gives evidence for weak bonding to a solvent molecule. In a few cases the e.s.r. results have been unexpected, a noteworthy example being those for the radical cations of nitr0a1kanes.l~ According to simple expectation, photoelectron spectroscopy and theoretical estimates,lG the SOMO for these cations is expected to be confined to the two oxygen atoms.The three possible orbitals [(I)-(111)] CFCl, + e--+ (CFC1,)- + 'CFCl, + C1- CFC1,C + CFCl, f CFCI, + CFCl; CFC1; + S -+ CFCl, + S+. X R Y I Y I I 9 0 ;/ Q I LO, y ' D n n* are close in energy, and there is some controversy relating to their order, but that is not our main concern since such cations either were not detected or were very minor components. The major species detected had e.s.r. spectra so similar to that for 'NO, that we initially supposed this to be the product. Indeed, trapped 'NO, radicals are known to exhibit a rather variable set of e.s.r. parameters, mainly because these small molecules tend to librate in different ways in different solvents,17 so it was not possible to eliminate this theory simply because of lack of precise agreement of parameters.The aim of the present work was to establish the identity of these NO,-like species using WH,NO,, and to extend the study to a range of other nitro derivatives. EXPERIMENTAL The nitro derivatives of methane, ethane, propane and 2-methylpropane were the best commercially available and were generally used without further purification. However, because of the unique results for CH,CH,NO,, this compound was extensively purified by various standard methods. The e.s.r. spectra were unaffected by any of these procedures. Samples of CH,CD,NO, were prepared by dissolving EtNO, in D20 + NaOD solvents, neutralization and extraction after the conversion back into the nitro derivative was complete.The 2-propyl derivatives, Me,C(CI)NO,, Me,C(Br)NO,, Me,C(CO,Et)NO,, Me,C(CMe,NO, and Me,C(Ar)NO, (Ar = p-nitrophenyl) together with 1 -nitrocyclohexene were all purified com- pounds kindly supplied by Dr R. Bowman (Loughborough University). 13CH,N0, (90% enrichment) was from Brochem (B.O.C.) and was used as supplied. Dilute solutions (ca. 0.001 mole fraction) were degassed and frozen as small beads in liquid nitrogen prior to exposure to 6oCo y-rays in a Vickrad cell at 77 K to doses up to ca. 1 Mrad.D. N. R. RAO AND M. C. R. SYMONS 567 E.s.r spectra were measured at 77 K using a Varian El09 spectrometer calibrated with a Hewlett-Packard 5246L frequency counter and a Bruker B-H12E field probe, which were standardized with a sample of diphenylpicrylhydrazyl (DPPH).Samples were annealed in the insert Dewar after decanting the coolant and were recooled to 77 K whenever significant spectral changes were detected. RESULTS AND DISCUSSION In all cases good e.s.r. spectra were obtained in the g = 2 region, which we confidently assign to radical cations derived from the nitro derivatives. The resulting spectra [see fig. 1 of ref. (15) and fig. 5-7 later in this paper] are remarkably similar to the e.s.r. spectra obtained from certain irradiated metal nitrates.18 These are characterized by signals from NO; and 'NO, radicals, that characteristic of 'NO, being the asymmetric triplet with ca. 60 G splitting (B) and that characteristic of NO; being the small triplet feature (ca.3-4 G) occurring ca. 17 G downfield of the central ( M , = 0) line (species A). SPECIES A We expect the spectrum for RNOH+ cations formed by electron loss from oxygen to resemble that for NO; since both have SOMOs confined to oxygen. For NO', there is little doubt that this is an in-plane antibonding combination of oxygen 2p orbitals, but there has been controversy regarding the degree of admixture of the three available atomic orbitals.lg9 2o We therefore expect to find a small, fairly isotropic coupling to 14N in the region of 4 G and a large g-value variation. There can be little doubt that feature B is due to these unmodified cations. However, they were always overshadowed by strong features from species A, so that the full spectrum remains unknown.This means that we are unable to select a precise structure for this species, nor can we extend significantly our previous discussion.15 SPECIES B The similarity in form of the 14N hyperfine coupling and g-tensor components to those for 'NO, radicals must mean that this species has a similar structure. Also, the absence of any proton hyperfine splitting seemed to indicate low spin density on the alkyl groups. These considerations lead,us to suggest the rearranged nitrite form of the cation (IV). Although such cations are unknown, the isoelectronic carbon-centred radicals ROC0 have been As expected, the results show that these are structurally similar to 'COT radicals, and for CH,OCO the proton coupling was only 1.23 G.21 However, our new results for (l3CH3N0,)+ cations make this assignment less compelling.13CH,N0 CATIONS A typical e.s.r. spectrum is shown in fig. 1 . This can be compared with fig. 1 of ref. (1 5) for the corresponding 12CH,N0, cation spectrum. Our best computer fit for this spectrum, using the data given in table 1, was quite acceptable, accommodating all the extra features satisfactorily. The results are surprising in that they establish that the 13C coupling is remarkably large. If the isotropic and anisotropic components568 RADICAL CATIONS OF NITROALKANES 1 32505 I * H 50G , Fig. 1. First-derivative X-band e.s.r. spectrum for 13CH,N0, in CFCl, after exposure to 'j0Co y-rays at 77 K, showing features assigned to the a-radical H,C"Oi (90% enriched in 13C). are analysed in the usual way23 we obtain ca.8% 2s and 38% 2p, character for the orbital on carbon. A similar analysis of the 14N hyperfine components gives ca. 10% 2s and 34% 2pz character. Thus the SOMO seems to be rather evenly divided between carbon and nitrogen, and it is difficult to accommodate that result in terms of the nitrite structure (IV). The fact that the maximum 14N and 13C tensor components appear to be nearly colinear together with g,, which is close to the free-spin value, strongly suggests the alternative a-structure (V), in which the C-N bond has stretched but the CH, unit remains pyramidal and the NO, unit bent. This novel a-structure also accommodates the absence of detectable lH coupling since the coupling for methyl radicals is known to pass through zero as the radical bends.It accords with the a-structure for the cation (Me,CCH,)+ studied by Toriyama et al." It represents the first stage in the dissociation to give 'CH,+NOi radicals, although no such complete dissociation has been detected in the present study even for the t-butyl derivative. It is interesting to compare this a-structure with that found for the isoelectronic system, CH,COg. Lin and K e ~ a n , ~ have used electron spin-echo techniques to show that when this radical breaks down to give 'CH, +C02 in irradiated lithium acetate the two products remain close together. One of us25 has used the results to establish that the only major movements within the crystal are those of the two carbon atoms to give the planar 'CH, unit and linear CO, unit. Thus for H,C---CO, relaxation of both halves of the molecule seems to be complete, whereas for (H,CNO,)+ it is minor for both halves of the cation.Why should this arise? At first sight it might be argued that the C-C bond shouldD. N. R. RAO AND M. C. R. SYMONS 569 Table 1. E.s.r. parameters for radical cations of nitroalkanes and NO2 14N hyperfine coupling constantsa g values radical A , A?/ Az Aiso gz g, gz NO,b CH,NOz C d CH,CH,NO,+ CH,CH,CH,NO; (CH,),CHNOl (CH,),CNOl c1 / \ (35c1) / \ PBr) (CH3)2C NO,+ Br (CH3)2C NO: 46.13 52 120 53 53 52 52 52 52.5 - 0 52 - 4 C02Et 52 / \ I I (CH3)2C NO: (CH3)2C -c(CH3)2 NO,NO; Me Me, I ,NO; C 53 44.8 47 120 48 49 49 48 52 50 - 0 49 - 4 49 48 50 49 48 66.76 66.5 155 66 70 67 67 63 67 5 66 7.5 67 67 68 68 67 52.56 55.2 55.7 57.3 56 55.7 55.7 132 56.5 - - - 56 55.3 57 56.7 55 2.0062 2.0045 2.0045 2.0045 2.005 2.0045 2.005 1.999 2.003 2.004 2.005 2.0045 2.005 2.005 2.004 1.9910 1.994 1.994 1.9935 1.994 1.994 1.994 1.996 1.995 1.995 1.994 1.994 1.993 1.993 1.992 2.0020 2.002 2.002 2.002 2.002 2.002 2.002 2.002 2.002 2.002 2.002 2.002 2.002 2.002 2.002 a 1 G = T; NO, in gas phase; 13C parameters; after annealing. be stronger than the C-N bond (cf the fact that the C-C bond in C,O;- is much stronger than the N-N bond in N,O,).However, reference to the qualitative energy-level scheme in fig. 2 suggests the reverse. We are comparing the one-electron bond between CH,+ and CO, or NO;. For the former, localization on the CH, unit is clearly more favourable than for the latter.Thus we are inclined to favour the a-structure (V) for the major cationic species formed from nitromethane. However, when the irradiated 13CH,N0, sample was annealed the spectrum changed, with complete loss of the 13C features, leaving features570 RADICAL CATIONS OF NITROALKANES \ \ \ \ ,,,- CH; 0,c -\ O,N*-, ''4, I I \ '\,+; I Fig. 2. Qualitative energy-level scheme for Me'CO, and Me'NOl radicals having a single electron in the C-C or N-C a-orbitals. This suggests why the a1 bond is significant for Me'NOi but dissociates to give 'CH, + CO, for Me'CO,. LL-Juu +1 ( B I 0 (6) -l(Bll - .1 1 ( C ) L(C1 - I // ( C ) +1 1 Fig. 3. First-derivative X-band e.s.r. spectrum for CH,CH,NO, in CFCl, after exposure to s°Co y-rays at 77 K, showing features assigned to CH,CH;'NOl cations (B) together with those tentatively assigned to CH,CH26---(NO+) alkoxy radicals (C).very similar to the normal 14N triplet, as obtained from 12CH3N0, solutions. Indeed, in several instances we noticed small changes in the 14N hyperfine and g-values for these cationic species on annealing, and these results suggest that they may occur as a result of a major chemical change. It is tempting to suggest that this is the rearrangement (1) that we originally supposed to occur at 77 K: N / \ R'NO; 0 0-R+. This explains the small change in A(14N) and almost complete loss of 13C coupling. An alternative, that this secondary species is simply 'NO,, seems to us less likely since this requires the concurrent formation of CH; carbocations. We conclude that, at least for CH3N0,, the major cationic species formed at 77 K is the a-cation, H,C"O:, and that this readily rearranges (at ca.130 K) to give the nitrite cation, ONOMe+. This is supported by our results for solutions of amyl nitrite in freon. This gave a spectrum very similar to those obtained from the nitro derivatives. This result helps to support our view that ONOR+ cations should indeed give rise to such spectra.D. N. R. RAO AND M. C. R. SYMONS 57 1 Table 2. E.s.r. parameters for other radical species derived from nitro derivatives radical lH hyperfine coupling/G g values fl-("b' - L 17 (1) 25 (3) 50 (1) ca. 2.0025 ca. 2.0025 ca. 2.0025 Me,C-C 23(2H) 8(2H) ca. 2.0025 \ +OEt CH,CH,O H(l) ca. 106 2.031(11) H(2) ca. 92 2.003 (I) NITROETHANE This compound was unique in giving high yields of a second species whose outer features are shown in fig.3. The main triplet is the normal NO,-like cation, the spectrum for the new species (C) comprising either an anisotropic doublet with A z 198 G or, more probably, a doublet of doublets with A , z 106 G and A , z 92 G. The latter is supported because on annealing the outer lines were lost less rapidly than the triplet features, and two central features were revealed which are probably part of the spectrum for C. These features showed the same extra 4 G splitting seen on the outer perpendicular lines for species C. Thus we require a radical having two strongly coupled protons and strongly asymmetric g tensor (gll = 2.03 1, gl = 2.003) (table 2). All these properties point to the presence of CH,CH,6 radicals.These are expected to have a shifted gll value, strong hyperfine coupling to the two methylene protons and weak coupling to the methyl protons. Although we know of no report giving e.s.r. parameters for this radical, a range of RCH,O' radicals have been studied by e.s.r. spectroscopy, mainly by Box and coworkers, and also the CH,O radical has been studied by this technique.,' Our total proton coupling of ca. 170 G is greater than that normally found (EH = 141-156), whilst the value for g,,, of ca. 2.031 is less than that usually detected (2.054-2.093).28 Very similar results were observed for a radical recently detected by us when attempting to prepare the radical cation of s-tri~xane.,~ The major species was almost certainly the normal cation, but a secondary spectrum FA(PH) = 175 G, g,,, = 2.02861 was assigned to the alkoxy radical indicated in (VI), with weak interaction between the alkoxy oxygen and the cationic centre.In terms of this structure, such an interaction is necessary because, to have a relatively small572 RADICAL CATIONS OF NITROALKANES 1 32505 Fig. 4. First-derivative X-band e.s.r. spectrum for (CH,),CNO, in CFCl, after exposure to s°Co prays at 77 K, showing features assigned to librating (Me,CNO,)+ cations. g shift, the p-orbital degeneracy must be strongly lifted. This is normally accomplished by specific hydrogen bonding,26-28 but this cannot be invoked in the presence case. We therefore suggest structure (VII) to explain the small Ag for the CH,CH,O* radicals thought to be formed from the (EtNO,)+ cation.Note that studies of methyl nitrite and nitromethane by mass-spectrometiic techniques show that a major path for fragmentation gives MeO' radicals and NO+ ions.z8b This lends support to our postulate for species C, but we still do not understand why this species was only detected for nitroethane. In order to check the possibility that this species was formed from an impurity in the nitroethane, a variety of purification procedures was used. These made no difference to the relative yields of the two major products. Furthermore, we were unable to detect any impurities using proton resonance. Despite this identification, we have strong reservations since a rather similar but far weaker spectrum appeared for irradiated freon solutions of Me,C(NO,)C(NO,)Me,, [see fig.7(b)]. In this case, we cannot use any R O formulation to explain the results. [(CH,),CNO,] + CATIONS Although the e.6.r. spectrum differs markedly from those for other nitroalkane cations (fig. 4), the data show that the species is essentially the same, with Aiso andD. N. R. RAO AND M. C. R. SYMONS 1 3250 Ci 1 32SOG 0 1 573 Fig. 5. First-derivative X-band e.s.r. spectrum for (a) Me,CClNO, and (b) Me,CBrNO, in CFCl, after exposure to s°Co y-rays at 77 K, showing features assigned (a) to (Me,CClNO,)+ cations and (b) to (Me,CBrNO,)+ cations.574 RADICAL CATIONS OF NITROALKANES 1 3250 G 1 l u 1 0 Fig. 6. First-derivative X-band e.s.r. spectrum for Me,C(CO,Et)NO, in CFCI, after exposure to 6oCo y-rays at 77 K, showing features assigned to [Me,C(CO,Et).NO,]+ cations and to [Me,C(NO,)-C~,Et]+ cations (D).g,, close to those for the other species. We therefore invoke extensive libration of the -NO, unit about all three axes. This is presumably a consequence of the size of the t-butyl group. However, it may also reflect a change in the type of bonding between the alkyl and -NO, groups, reflecting a weaker bonding and hence greater freedom for NO,. We tentatively suggest that the Me3C- group has become planar or nearly so, the structure approaching the limiting form Me,C+ ---NO,. Very extensive libration for the small molecule NO, would then be expected. CHLORO AND BROMO DERIVATIVES We had hoped to detect well defined hypefine coupling to chlorine and especially bromine nuclei for these cations, as we did for the corresponding anion.,O In fact (fig.5) there are extra small splittings, but these are difficult to correlate with the expected quartet features from 35Cl, ,'Cl, 'OBr and *lBr, which all have Z = i. Our tentative analysis, which accommodates much of the extra splitting, is shown in fig. 5, and the suggested parameters are given in table 1. We stress that the directions of the hyperfine tensor components are unlikely to be coaxial with the 14N and g-value components, so the magnitude of the coupling can only be taken as a rough indication of delocalization. Furthermore, the coupling constants are likely to be smaller than the electrical quadrupole coupling energy, and hence the spectral features are unlikely to be quartets except for fields close to the carbon-halogen axis.Thus there is little point in our endeavouring to obtain a better spectral fit or in considering the significance of the results, except to say that, as expected for the proposed structures (IV) or (V), delocalization onto C1 or Br is small.D. N. R. RAO AND M. C. R. SYMONS 1 32506 1 3200G u - 1 575 I1 II I 1 X +1 X 0 X 1 I I I t 11 Fig. 7. First-derivative X-band e.s.r. spectrum for (a) Me,C(NO,), and (b) Me3C(NOz)- C(NO,)Me, showing features assigned to (a) [Me,C(NO,)”O,]+ cations with only one strongly coupled 14N nucleus and (6) [Me,C(NO,)-C(Me),’NO,]+ cations. [Outer features (a) in (b) are discussed in the text.] Y Z *: +r : r -1 -1576 RADICAL CATIONS OF NITROALKANES Me,C(CO,Et)NOi CATIONS These were of interest because of the potential competition between two localized cationic species, one being an ester cation (RCO,Et)+ and one a nitro cation.We have already established that, in general, ester cations are z, with major spin density on the ester oxygen and strong hyperfine coupling to the a-protons of the ester alkyl group. l4 In fact, both types of spectra were obtained (fig. 6). Analysis of the ester-cation features is difficult because of severe overlap with the 10) features of the nitro-cation spectrum. However, the spectrum was better defined at ca. 140K and gave the parameters listed in table 2. These are typical for an ethyl ester. Thus ethyl formate cations have a(2H) = 22 G and a(2H) = 10 G. The other features are clearly due to the normal form of nitroalkane cations, presumably having the a-structure (V).* Coexistence of these two structures is of considerable interest, since even if one accepts that initial electron loss should occur at either site in a statistical manner, and that no fully delocalized SOMO can be formed, one would expect subsequent electron transfer to give only the more stable cation.The ionization potentials of typical RC0,R' and RNO, molecules are ca. 10.6 and 11.2 eV, respectively. However, the value for nitroalkanes refers to loss from the n ( 0 ) orbitals, rather than from the stretched C-N a-orbital. We need either a fortuitous balance of energies or such effective orthogonality that electron transfer cannot occur despite the short path-length. DINITRO DERIVATIVES For the cation of Me,C(NO,), we find complete localization on one NO, group, with no hyperfine contribution from+the other 14N nucleus (fig.7). This accords with expectation for the a-structure 0,N 'C(Me),NO,, or the rearranged structure. The results establish that any tendency to switch from one NO, group to the other is slow, and that, as with the ester derivative, there is no available orbital which is delocalized onto both nitro groups. This also applies to the cation of tetranitromethane and to the derivative having nitro groups on two adjacent carbon atoms [Me,C(NO,)C(NO,)Me,]. The spectrum for the latter cation shows a small extra doublet splitting on the x features of ca. 4 G. This is tentatively assigned to coupling to the second 14N nucleus. This can appear as a doublet rather than a triplet under conditions of quadrupole control.Outer features {a in fig. 7(b)] are similar to those found in the spectra for EtNO, systems, but in this case they cannot be assigned to Me,C(N0,)6 radicals since there are no p protons. AROMATIC NITRO COMPOUNDS As expected, the initially formed cations have SOMOs based on the ring n - ~ r b i t a l s , ~ ~ that for nitrobenzene being the symmetric orbital with a node through nitrogen (VIII). However, this is not a powerful selection by the nitro group, since a p-methyl group I CH3 * Note added in proof: Recent work suggests that these ester cations may have rearranged structures such as //OH+ R-C ' CH,-CH,.D. N. R. RAO AND M. C. R. SYMONS 577 switches the SOMO so as to place high spin density on the methyl group (IX), as in the toluene However, the most interesting aspect of these results is the fact that, on annealing, an NO,-like spectrum appears, which we.assign either to the a-radical (V) or, more probably, the nitrite derivative, (ArONO)+. The intensities of these features vary strongly with the nature of Ar, as does the temperature at which they are first detected. However, we have not been able to discern any clear pattern. No doubt there are other pathways for decomposition besides this rearrangement. In the light of these results we thought it of interest to study the competition between nitro groups for the cation of the p-nitro derivative (X). Our results, again to our NOz surprise, show that the major product at 77 K is an NO,-like species, with an e.s.r.spectrum similar, for example, to that shown in fig. 7. Since it is formed directly, this is probably the a-radical, ArCMe,'NOz. If initial electron loss from the ring occurs as expected, transfer of the hole into this a-orbital must be facile. This suggests a conformation in which the aliphatic nitro group lies out of the plane of the ring so as to maximize a-n overlap with the C-N bond. Electron transfer then represents the limit of hyperconjugative electron release. It is a major effect rather than a minor perturbation because the C-N bond stretches extensively. We stress that in the absence of 13C data for such radicals we are unable to discover which of the two limiting structures, (IV) or (V), is actually present in these systems.NITROALKENE CATIONS One derivative, 1-nitrocyclohexene (XI), has been studied in freon. The e.s.r. spectrum after irradiation was typical of those for alkene cations and was analysed in terms of one aH at 17 G, three QH at 25 G and one PH at ca. 50 G. These results can be compared with those for the cation of cyclohexene (2aH, 9G; 2QH, 55 G; 2bH, 22 G).32 This suggests that there is a shift of spin away from the nitro group, but that the normal chair form is not adopted. It is interesting that, although the a-proton coupling indicates a high spin density on C,, this is not reflected in an increase in the Q-proton coupling constants. RCoH RADICALS In the light of these results we have searched the literature for evidence of similar processes occurring for RCO; radicals.As stressed above, however, the limiting 'a' structure is now one which resembles R---CO,, with almost no bonding between the two fragments. Although ROC0 radicals are stable in that they can be detected in 20 FAR 1578 RADICAL CATIONS OF NITROALKANES liquid-phase experiments, we found no evidence for their formation from R' and CO, formed by radiolysis of RCO; or RC0,H. Our own attempts to detect such a reaction have also failed. We thank Dr R. Bowman of Loughborough University for kindly supplying several nitro derivatives and for helpful discussions. Reactive Intermediates, ed. R. A. Abramovitch, (Plenum Press, New York, 1980, 1982, 1983), vol. 1,2 and 3. Organic Reactive Intermediates, ed. S . P. McManus (Academic Press, New York, 1973). L.B. Knight and G. Steadman, J. Chem. Phys., 1982, 77, 1750. L. B. Knight and G. Steadman, J. Chem. Phys., 1983, 78, 5940. L. B. Knight and G. Steadman, J. Chem. Phys., 1984, in press. L. B. Knight, G. Steadman, D. Feller and E. R. Davidson, J. Am. Chem. SOC., 1984, 106, 3700. chap. 9, p. 321. T. Kato and T. Shida, J. Am. Chem. SOC., 1979, 101, 6869. M. C. R. Symons and I. G. Smith, J. Chem. Res. (S), 1979, 382. ' W. H. Hamill, Radical Ions, ed. L. Kevan and B. Webster (Wiley Interscience, New York, 1968), lo J. T. Wang and F. Williams, J. Chem. Soc., Chem. Commun., 1981,666. l1 K. Toriyama, K. Nunome and M. Iwasaki, J. Chem. Phys., 1982,77, 5891. l2 A. Hasegawa and M. C. R. Symons, J. Chem. Soc., Faraday Trans. 1, 1983,79, 93. l3 D. Becker, K. Plante and M. D. Sevilla, J. Phys. Chem., 1983, 87, 1648. l4 D. N. R. Rao, J. Rideout and M. C. R. Symons, J. Chem. SOC., Perkin Trans. 2, 1984, 1221. l5 D. N. R. Rao and M. C. R. Symons, Tetrahedron Lett., 1983, 24, 1293. C. N. R. Rao, Indian J. Chem., 1976, 14A, 147. H. Sharp and M. C. R. Symons, J. Chem. Soc. A, 1970, 3075. P. W. Atkins and M. C. R. Symons, J. Chem. Soc., 1962,4794. l9 P. W. Atkins and M. C. R. Symons, The Structure of Inorganic Radicals (Elsevier, Amsterdam, 1967). 2o M. C. R. Symons, J. Chem. Soc., Dalton Trans., 1979, 423. 21 H. Hefter and H. Fischer, Ber. Bunsenges. Phys. Chem., 1970, 74, 493. 22 D. Griller and B. P. Roberts, J. Chem. Soc., Perkin Trans. 2, 1973, 1339. 23 M. C. R. Symons, Chemical and Biochemical Aspects of Electron Spin Resonance Spectroscopy (Van 24 D-P. Linad and L. Kevan, Radiat. Phys. Chem., 1981, 17, 71. 25 M. C. R. Symons, J. Phys. Chem., 1983,87, 1833. 26 J. Y. Lee and H. C. Box, J. Chem. Phys., 1973,59, 2509. 27 M. Iwasaki and K. Toriyama, J. Am. Chem. Soc., 1978, 100, 1964. Nostrand Reinhold, Wokingham, 1978). (a) W. A. Bernard, D. M. Close, J. Hiittermann and H. Zehner, J. Chem. Phys., 1977, 67, 121 1 ; (b) Y. Niwa, S. Tajima and T. Tsuchiya, Int. J. Mass Spectrom. Ion Phys., 1981, 40, 287; (c) J. P. Gilman, T. Hsieh and G. G. Meisels, J. Chem. Phys., 1983, 78, 1174. M. C. R. Symons and B. W. Wren, J. Chem. Soc., Perkin Trans. 2, 1984, 51 1. 30 M. C. R. Symons and W. R. Bowman, Tetrahedron Lett., 1981, 22,4549. 31 M. C. Symons and L. Harris, J. Chem. Res., ( S ) , 1982, 268; (M), 1982,2746. 32 M. Tabata and A. Lund, Chem. Phys., 1983,75, 379. (PAPER 4/538)