1976 21 Radiation Mechanisms. Part VII.l Electron-loss and -capture Processes in Phenylphosphinic Acid and Related Compounds By Shuddhodan B. Mishra and Martyn C. R. Symons," Department of Chemistry, The University, Leicester LEI 7RH Exposure of phenylphosphinic acid and related compounds to 6oCoy-rays at 77 K resulted in electron addition to the aromatic ring, sincesthe characteristic e.s.r. spectra of the alternative phosphoranyl radicals were absent except for that assigned to PhPH (0)OH radicals. The anions were protonated on annealing, or directly at 77 K in methanolic glasses, to give cyclohexadienyl radicals which exhibited a hyperfine coupling of 22 G to slP nuclei. However, the aromatic anion of PhPH(0)OH probably undergoes a 1.2-hydrogen atom shift from phosphorus to the ring carbon, the resulting adduct having A (lH) 40 and A (31P) 200 G.This species is unstable and either breaks down to give benzene and -PO(OH)- radicals, or undergoes a further hydrogen shift to give a normal cyclohexadienyl species. RADIATIONprocesses in aromatic compounds have not been systematically studied using temperature-resolved e.s.r. spectroscopy.2 Various cyclohexadienyl radicals have been detected,,?* which are readily characterised by the large (ca. 50 G) hyperfine coupling exhibited by the methylene protons. Such radicals are also readily formed by direct hydrogen atom additi~n.~Liquid-phase e.s.r. studies of aromatic compounds have been dominated by the ease with which the x-electronic systems can gain or lose electrons.Dixon and Norman,6 in their pioneering studies of hydroxyl radical reactions, obtained the e.s.r. spectrum for hydroxyl radical adducts of aromatic compounds using a flow system. These hy- droxyl adducts have recently been studied during radioly- sis of aqueous solutions of benzene.' Phenyl radicals have also been detected by e.s.r. ~pectroscopy.~~~ Our interest in phenylphosphorus compounds has been twofold, to examine the competition for added electrons between the aromatic ring and the phosphorus atom, and to study adducts of type (I),l0 in the expectation that hyperconjugative interaction with the PR, (or PR,) groups, as first postulated by Eaborn and his co-workers for the corresponding cations,ll would give rise to a relatively large hyperfine coupling to ,IP.l2 In our earlier study of the molecules PhP(O)Cl, and PhP(S)Cl, we found that electron addition was at pho~pliorus.~~However, preliminary studies showed that when chlorine was replaced by oxygen, ring addition dominated.1° In a related study Bockestein et aL1* found that addition of RO-radicals to PhP(OMe), resulted in a radical having a zwitterion structure with the unpaired electron in the x-system of the benzene ring.This has recently been confirmed by Davies et aZ.15 who also found that normal addition to give phosphoranyl radicals occurred when more elect1 onegative substituents such as chlorine were present. [It is noteworthy that the Part VI, S. P. Mishra and 11. C.R. Symons, Internat. J. Radiation Phys. Chew., submitted for publication. I. S. Ginns and M. C. R. Symons, J.C.S. Dalton, 1976, in the press.H. Fischer, Kolloid-Z., 1962, 180, 64. S. Ohnishi, T. Tanei, and I. Nitta, J. Chenz. Phys., 1962, 37, 2402. D. Campbell and M. C. R. Symons, J. Chew. SOC.(A),1969, 2480. W. T. Dixon and R. 0.C. Norman, J. Chem. SOG.,1963, 3119. K. Bhatia, Radiation Res., 1974, 54, 537. J. E. Bennett, B. Mile, and A. Thomas, PYOC.Roy. SOC.,1966, A293, 246. P. H. Kasai, E. Hedaya, and E. B. Whipple, J. Amer. Chew. SOL, 1969, 91, 4364. electron also remained on phosphorus for the adduct ROPH,(Ph).15] Irradiated phenylphosphinic acid and its salts have been studied by Lucken et aL16 who were only interested in the e.s.r.spectrum of the electron-loss centre, PhPO,H(PhPO,-). Their results showed that, in con- trast with the anion PhN0,-, electron-delocalisation in the ring was small. This was in accord with their previous study of Ph,PO radical formed from Ph,PH(O) .l7 EXPERIMENTAL Phenylphosphinic acid, diphenylphosphinic acid, and phenylphosphoric acid were of the highest grade available and were recrystallised from water (D,O) or methanol (CD,OD). Samples were irradiated as fine powders in a 6oCoVickrad y-ray source at a dose rate of 1.7 MCi h-1 for up to 2 h, either at 77 K or at ambient temperature (ca. 40"). E.s.r. spectra were measured with a Varian E3 spectrometer at 77 K. Samples were annealed in the insert Dewar after decanting the liquid nitrogen, and recooled to 77 K whenever significant changes were noticed in the rapidly scanned e.s.r.spectra. Samples irradiated at ambient temperatures were studied both at this temperature and at 77 K. RESULTS AND DISCUSSION Plzenylplzos~lzinicAcid-Exposure at ambient tem- peratures gave e.s.r. spectra of two well defined species, one being for PhPO(0H) radicals, as shown in Figure 1 of ref. 16, and the other being assigned to a cyclohexa-dienyl radical (Figure 1). Data extracted from these spectra are listed in the Table. The most noteworthy aspect of the spectrum for Phl?O(OH) is the extra, isotropic, doublet splitting of ca. 15 G, which must come from the hydroxy proton since it is lost on deuteriation.16 This coupling is larger than is usually observed for hydroxy protons and implies a very specific orientation which must maximise orbital overlap probably by a ' hyperconjugative ' mechanism.lo S. P. Mishra and M. C. R. Symons, Tetrahedron Letters, 1973,41, 4061. l1 R. W. Bott, C. Eaborn, and P. M. Creasley, J. Chew. SOC., 1964, 4804. l2 A. R. Lyons and M. C. R. Symons, J.C.S. Faraday 11,1972, 622. l3 S.P. Mishra and hl.C. R. Symons, J.C.S. Dalton, 1973,1494. l4 G. Bockstein, E. H. J. M. Jansen, and H. M. Buck, J.C.S. Chern. Comrn., 1974, 118. l5 A. G. Davies, M. J. Parrott, and B. P. Roberts, J.C.S. Chem. Comm., 1974, 973. 16 M. Geoffrey and E. A. C. Lucken, MoZ. Phys., 1972, 24, 335. 17 M. Geoffrey and E. A. C. Lucken, MoZ. Phys., 1971, 22, 257.J.C.S. Perkin I1 The cyclohexadienyl radicals probably have structures phosphorus 3s character in the C-P a-bond, induced by (11)and (111). The proton hyperfine coupling constants the electronegative oxygen substitutes. Similar trends are normal, suggesting a spin-density on C-1 (the carbon have been noticed previously and attributed to changes directly bonded to phosphorus) of ca. 3. However, in sf^ hybridisation.20~21 I 32566 C r t PO 2”-1FIGURE First derivative X-band e.s.r. spectrum for phenylphosphonic acid after exposure to 6oCo y-rays: a, at ambient temperature, showing features assigned to cyclohexadienyl radicals (11) and (111); b, at 77 K, showing high-field (MI= -&)features assigned to (A) Phl?O,H radicals, (B) [PhPHO(OH)]- radicals (IV), and (C) cyclohexadienyl radicals of structure (I); and c after annealing, showing features assigned to (PO,H-) radicals, and PhPO(0D) radicals (a) radicals of type R&-PR, have isotropic 31P coupling Irradiation at 77 K gave a complex ear.spectrum, constants in the region of 40 G.lSJs The high value of the more intense centre (g=2) region being dominated by 22 G probably reflects an increase in the proportion of a broad singlet assigned to substituted benzene anions, la E. A. C. Lucken and C.Mazeline, J. Chem. SOC.(A),1966,74; 20’ S. Subramanian, M. C. R. Symons, and H. W. Wardale, J. 1967,439. Chem. SOC.(A),1970, 1239. l8 A. Begum, A. R. Lyons, and M. C. R. Symons, J. Chem. SOC. 21 S. Subramanian and M. C. R.Symons, J. Chem. SOC.(A;,(A),1971, 2388. 1970, 2367. 1976 23 and an asymmetric doublet assigned to 6PH(PH)OH After prolonged irradiation and annealing to room radicals similar to those formed by electron loss from temperature, clear parallel features separated by ca. 270 phosphate ions.20 The outer regions contained many G were detected (Figure lc). These are characteristic of phosphinyl radicals,22 .PL,, having the unpaired electron P in a pure x-level. The perpendicular components for phosphinyl radicals almost coincide in the free-spin region and would be completely hidden by the intense central lines in the present work. The radicals respon- sible for these features must be either l?O,2-[PO(OH)-], or possibly PhpO- (PhPOH), both of which are ex-pected to give rise to similar e.s.r.features.22 We favour H ..!;the former, for reasons given below. In CD,OD solution broad features assigned to PhP0,- radicals were obtained, together with clear wing-lines 0 rml fIYl poorly defined features. The high-field region is shown in Figure lb; the low-field region being very similar. Con-trolled annealing studies established that the A features are broadened perpendicular lines for the MI = -ij part for the deuterium atom adducts (11)and (111). Intense central lines for trapped electrons and D2e0D radicals were also observed. PhenyZphos$?ionic Acid-After irradiation at ambient temperature, central features very similar to those for phenyl phosphinic acid were obtained (Figure 2a).This species is undoubtedly the same type of cyclohexadienyl E.s.r. parameters for radicals formed in y-irradiated phenylphosphorus compounds Compound PhPO (OH) PhPO(OH)2 PhPO (OH) PhPH(0)OH PhPH (0)OH PhPH(0)OHPhPH (0)OH PhPH (0)OH PhPH(0)OH + CD,OD Ph,PO(OH) 0 GI= 10-4 T. 31PHyperfine coupling (G) Radical T/KO All A1 Aieo lH Hyperfine coupling (G) (II), (111) PhP0,- 323 323 ca. 26 ca. 20 22 620 435 497 8 (2 H), 48 (2 H) (7.4, 2 H) 0-P0,Ph 77 ca. 10 PI), (111) PhPO(OH)PopPhP0,-(1)(IN 323 323 323 77 77 77 ca. 26 ca. 20 22 666 475 539 270 ca. 0 ca. 90 660 385 443 ca. 200 ca. 600 9 (2 H), 48 (2 H) (7.4, 2 H) 15 (1 H) 40 (1H), ca. 10 (3 H) ca. 160 (1 H) bPPh,( OH) 77 ca. 10 g Values were close to that of the free-spin (2.00).of the spectrum for PhPO(0H) radicals. The B features comprise a pair of perpendicular lines, separated by 150 G, that decay together on slight annealing. A similar 150 G doublet was lost simultaneously on the low-field side. This 150 G doublet was replaced by poorly defined triplet for the deuteriated compound. The magnitudes of the 31P and 1H hyperfine coupling constants are typical of phosphoranyl radicals having an axial hydrogen ligand (IV), and we therefore identify this species as a parent anion in which the electron is centred on phosphorus rather than on the phenyl group. Another pair of doublets (C), which became momentarily very much better defined during the annealing process prior to decaying together irreversibly, were also shown, by deuteriation, to be caused by coupling to 31P(ca.200 G) and lH (ca. 40 G). The lines were slightly asymmetric, but absence of defined shoulders made any estimate of the anisotropic part of the hyperfine tensor components unreliable. Some indication of a small (ca. 10 G) quartet splitting was also observed for these features (C), which are therefore assigned to the adduct (I). It is significant that use of the partially deuteriated material, PhPH(0)OD gave the same e.s.r. features, so the hydro- gen atom responsible for the 40 G hyperfine coupling must have been originally bound to phosphorus. 0 Temperature of irradiation. radical [(11)and (111)],the 31Phyperfine coupling con- stants being equal within our experimental error.The outer lines (Figure 2b) are assigned to PhP0,- radicals, since the alternative phosphoryl radical PhPO(0H) is expected to have larger 31Pcoupling constants, and, by analogy with the results for phenyl phosphinic acid, an extra doublet splitting from the hydroxy proton. Irradiation at 77 K gave only intense central lines, which were interpreted in terms of a singlet from the aromatic radical anions, and an asymmetric doublet from the OPL, hole-centres. No outer lines of any sort were obtained. On annealing, the room temperature features grew in at the expense of these central features. Solutions in methanol (CD,OD), gave the deuterium atom adducts (11)and (111)directly at 77 K. Diphenylphosfihinic Acid.-Again, the e.s.r.spectra obtained after irradiation at ambient temperature was dominated by that for the cyclohexadienyl radicals (11) and (111)being almost identical with that shown in Figure 2a. However, the central spectral region still contained a strong absorption from radicals thought to be 6P(OH)-Ph,. In the other two compounds these radicals were unstable at ambient temperatures. No outer lines were 22 B. W. Fulham, S. P. Mishra, and M. C. R. Symons, J.C.S. Dalton, 1974, 2145. observed. This was also true after exposure at 77 K, the central features being then assigned to the substituted aromatic anions and the oP(OH)Ph, species. Reaction Mechanisms.-The results show that, at 77 K and room temperature, electron addition at the benzene ring is favoured.Taken in isolation, these results could mean that the distortions required to trap the electron at phosphorus are too slow at 77 K compared with the 1n 32706 ,106,n ml=+l b FIGURE First derivative X-band e.s.r. spectrum for phenyl- 2 phosphonic acid after exposure to 6oCo y-rays at ambient temperature showing : a, features assigned to cyclohexadienyl radicals (11) and (111) and b, features assigned to Phl?O,-radicals minor distortions involved in aromatic anion formation. That is to say that kinetic rather than thermodynamic control could be involved. (This is an intramolecular phenomenon, not one resulting from matrix constraints.) This would accord with the observation that low concen- trations of the phosphoranyl radical with axial hydrogen were formed, since only the hydrogen needs to move extensively and this can be rapid even at 77 K.One would have anticipated that electron transfer from phenyl to phosphorus would then occur on annealing and this was never detected. (A similar electron-transfer from N-0 to Fe was observed for irradiated sodium nitro- prus~ide.~~)However, this might be because ring pro- tonation to give (11) and (111) occurred more rapidly than electron transfer. We do not favour this interpretation, however, because J.C.S. Perkin 11 of the results obtained with the t-butoxy ad duct^.^^^^^ These establish that the reverse electron transfer is in fact favoured for such compounds.Also, the results of Davies et aZ.I5 show that the presence of P-H bonds favours the phosphoranyl species, as we have now ob- served. Thus it seems that electron addition to the aromatic ring is thermodynamically favoured in the present compounds. The reactions of phenylphosphinic acid are noteworthy. It seems that the aromatic anions experience an intra- molecular proton transfer since it is the P-H proton that migrates into the ring. This is surprising from an acid- base viewpoint since this proton is non-acidic. It is unlikely that this is the major mechanism for cyclohexadienyl formation [(11)and (111)] since similar radicals are formed from all three compounds. Also, the adducts of type (I)were only detected for the compound containing a P-H bond.We conclude that normal protonation is not favoured at C-1, but at the ortko- and para-positions of the anion. It is also noteworthy that when protonation occurs at low temperature the methylene group clearly becomes )CHD on replacing protons by deuterons. However, at ambient temperatures an intramolecular scrambling occursso that radicals containing )CH, groups predomin- ate after deuteron addition. This scrambling is slow on the e.s.r. time scale, but fast with respect to the time for measurement (ca. 0.1 h) at room temperature. It can be accomplished by migration of the hydrogen atom between the ortho- and para-positions of the cyclohexa- dienyl radicals. We postulate that the l?L2species (Figure 2c) is PO,- or PO(0H)-, formed from adduct (I)by loss of benzene [reaction (l)]. We were unable to discover any other chemically expected process that would form phos- phinyl radicals. Formation of Ph$’02-radicals from PhPO(OH), probably proceeds via the anion [reaction (Z)]. This process, which only occurs after considerable annealing above 77 K, probably involves incipient or complete formation of the phosphoranyl radical in low concentra- tion, but we were unable to detect this species. We thank the S.R.C. for a grant to M. C. R. S., and the Banaras Hindu University, India, for grantin-g study leave to S. P. M. [5/621 Received, 3rd April, 19751 23 ill. C. R. Symons, D. X. West, and J. G. Wilkinson, J.C.S. Chern. Cornm., 1973, 917.