J. CHEM. SOC. FARADAY TRANS., 1994, 90(4), 597-604 Pulse Radiolysis Study of the Reactions of SO;-with some Substituted Benzenes in Aqueous Solution Getahun Merga, C. T. Aravindakumar and B. S. M. Rao* Department of Chemistry, University of Poona , Pune-4 1I 007, India H. Mohan and J. P. Mittal" Chemistry Division, Bhabha Atomic Research Centre, Bombay400 085,India The reactions of SO:-with several substituted benzenes having the general formula, C,H,-,X,Y (where X = F, CI or Br and Y = CH,, CH,CI, CHCI,, CF, or OCH,), have been investigated in aqueous solution by pulse radiolysis. The transient absorption spectra exhibit maxima at 315-330 nm and additional peaks at 270-290 nm with chlorotoluenes and weak peaks around 400 nm with chlorobenzene and 3-chlorofluorobenzene. Only in the case of 3-chloroanisole is the observed spectrum different, exhibiting two distinct peaks at 290 and 475 nm.The second-order rate constants for the reaction of SO:-range from about 10' for 2-chlorobenzotrifluoride to 10" dm3 mol-' s-' for 3-chloroanisole. It is concluded from the Hammett treatment (p' = -1.6) that the reaction mechanism involves both direct electron transfer and addition-elimination reactions. The intermediate radical cation is hydrolysed to give the corresponding 'OH adduct absorbing at 315-330 nm except in the case of 3-chloroanisole where it is stabilized. The formation of a benzyl-type radical by direct H abstraction by SO:-from the CH, group and/or deprotonation of the radical cation is an additional process whose extent is deter-mined by the relative position of the CH, group, the order being para > ortho x meta with monochlorotoluenes.The transient species absorbing around 400 nm is assigned to the phenoxyl-type radical. The differences in reaction mechanism between SO:-and 'OH attack are discussed. Radiation chemical studies on the reactions of primary radicals '-'of water and secondary radicals,'-14 derived from them, with benzene and its substituted derivatives in aqueous solution have contributed to an understanding of structure- reactivity relationships. In the radiolysis of aqueous solu- tions, the reactive species formed are solvated electrons (eJ, 'OH radicals and H atoms : ionizing H20 -eJ2.7), 'OH(2.8), 'H(0.6), H+(2.7), radiation H202(0.6), Hz(0.43) (1) The numbers in the parentheses represent the G values of the species, i.e.the number of molecules per 100 eV (one molecule per 100 eV = 1.036 x lo-' mol J-') of absorbed energy. Because of their high yields, the reactions of both 'OH and ea; with arenes have been widely studied (see ref. 15 and 16 for recent reviews). It is known that the reactivity of e,; with arenes is strongly influenced by the nature of the substituent as was seen" from the increasing rate of its reaction with high-electron-affinity substituents. In the case of halogenated aromatic compounds, the reaction may involve the n orbitals of the aromatic ring or of the substituent halogen The former, a pre- dominant process with fluorocompounds, results in the for- mation of the molecular anion, whereas other aromatic halogenocompounds give an aryl radical and a halide ion.The hydroxyl radical has been shown to react with substi- tuted benzenes first by addition to the benzene ring forming a 7t complex which immediately rearranges to a r~ complex. The different isomeric 'OH adducts',* formed from this rearrangement have absorption maxima in the range 310-340 nm. In our earlier study' on the reactions of the OH radical with substituted benzenes, the reaction rates for the forma- tion of *OH adducts were found to depend on the nature of the substituent. The oxidation of these adducts by K,Fe(CN), was reported6*' to depend on the nature of the substituent, and in the absence of an oxidant, they decay by second-order kinetics.The production of radical cations in aqueous solution by radiation chemical methodsg,' 1*1 relies on the use of oxidi-zing species such as SO;-, Clip, Bri-, TIZ+.Among these, the SO;-radical anion (Eo = 2.5 -3.1 V us. NHE' '*19) is com-e monly used and can be produced from the reactions of ea& and H [reaction (2)] with K2S20, or from 'OH with H2S04.20921 S,O;-+ eJH) -+SO;-+ SOt-(HSO,) (2) Laser flash photolysisZ2 of S,O;-[either by 193 or 248 nm (reaction 3)] has also been shown to be a clean source of so;-: hvs,o;--2so;-(3) ChemicalZ3 and thermal24 methods to produce SO;-were used to elucidate the reaction mechanism by product analysis. The SO:-radical anion reacts with several aromatic com- pounds to give hydroxycyclohexadienyl radicals whose for- mation was explained" either by the addition of SO;-to the benzene ring followed by hydrolysis or by a direct electron transfer from the ring to SO:-followed by hydration of the resulting radical cation.Substituted benzenes of the type C,H, -,X,Y (where X = F, C1 or Br, and Y = CH,, CH,CI, CHCI,, CF, or OCH,) contain different electron-donating and withdrawing groups and it is, therefore, of interest to investigate the reac- tions of SO;-with these systems as they form an ideal class of compounds for structure-reactivity studies. Furthermore, such an investigation would reveal the differences in reaction mechanism between 'OH and SO;-attack.Experimental Preparation of Solutions The substituted benzenes obtained from Fluka were of high purity (>98%) and were used without any further purifi- cation. Other chemicals used were of analytical grade. The solutions were prepared in water purified by the Millipore Milli-Q system and the solutions containing S,Oi-and tert- butyl alcohol were saturated with N, prior to the dissolution of the solute to avoid its volatilization during degassing. Irradiation The reaction of SO:-was studied in N,-saturated aqueous solutions containing K2S208(1.5 x rnol drnp3), solutes (10-4-10-3 mol dm-3) and 0.2 mol dm-3 tert-butyl alcohol. Since the rate constants for the reaction of 'OH radical with the solutes, as determined by us earlier,' are in the range (2- 9) x lo9 dm3 mol-' s-', the reactivity (k[S] = lo8 s-') of OH radical with tert-butyl alcohol (koH+Bu,OH= 5 x lo8 dm3 mol-' s-') is higher at least by an order of magnitude even in the case where the rate constant for the reaction of the OH radical with the solute is the highest.Furthermore, our recent work3 show that the reactivity of el with the compounds used in this study (lo6 s-I), under our experimental condi- tions, is lower by two orders of magnitude than its reactivity with S,Oi-(10' s-'). Therefore, more than 90% of the 'OH radicals will be scavenged by tert-butyl alcohol while e,; reacts quantitatively with S,Oi -to produce the SO;-radical ion [reaction (2)]. The rate of reaction of an H atom with SO:-(k = 2.5 x lo7 dm3 mol-' s-') is lower than that of its reac- tion with benzeneI7 and benzene derivatives (k z 9 x 108 dm3 mol-' s-').Under the conditions employed in this study, only a fraction of H (ca. 30%) is expected to yield SO:-via reaction (2). Thus, G(SO:-) was estimated to be 3.3 molecules per 100 eV [=G(e,i) at high S20i-c~ncentration,~+0.3G(H)]. The contribution of the H atom reaction with the solute, being negligible, is not taken into account. Pulse radiolysis experiments were carried out using high- energy 7 MeV electron pulses (pulse width 50 ns) from a linear accelerator at BARC, Bombay and the details of the set-up are publishedz6 elsewhere. A KSCN dosimeter was used in the optical pulse radiolysis using G~500= 21500 dm3 mol-' cm-' per 100 eV for the transientz7 (SCN);-. The dose per pulse, depending upon the pulse width, was in the range 10-20 Gy.-"I J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 The decay of SO:-in the absence and in the presence of various solutes was monitored at 460 nm. The transient absorption as a function of time was recorded on a storage oscilloscope interfaced to a computer for kinetic analysis.". Results and Discussion Absorption Spectra of the Transients The transient absorption spectra were measured from pulse radiolysis of the solutions containing the solute (1O-j mol dm-3), K,S,08 (1.5 x lo-' mol dm-3) and tert-butyl alcohol (0.2 mol dmP3) by monitoring the absorbance in the range 250-570 nm. The maximum absorbance of the tran- sients was reached within about 2 ps for systems whose rate constants are 2 lo9 dm3 mol-' s-'.Since the SO:-radical anion reacts with tert-butyl alcohol with a half-life of about 5 ps (vide infra) at the concentration employed in our experi- ments, appropriate corrections were applied to the transient absorption spectra with systems having k d 9 x 10' dm3 mol-I s-'. The values of the absorption maxima of the tran- sient species formed are compiled in Table 1. Toluene and Chlorobenzene The transient absorption spectra with the two mono-substituted compounds, toluene and chlorobenzene, are more or less similar with I,,, centred at 315 and 325 nm, respec- tively (Fig. 1). The spectrum in the case of chlorobenzene, in addition, exhibits a weak peak around 400 nm with a shoul- der around 270 nm. As the parent compound absorbs below 280 nm, the shoulder observed may not be characteristic of any intermediate product.The absorbance at 325 nm decays by about 20% over a period of 18 ps after the pulse but the absorbance at 400 nm still persists. Transient absorption spectra were also obtained for the reaction of 'OH with toluene and chlorobenzene by pulsing N,O-saturated solutions of these compounds. These spectra were more or less similar (A,,, x 325 nm) to those obtained in the reaction of SO;-with both systems, though the inten- sities are different in the case of chlorobenzene. Fig. l(a) shows both the spectra obtained in the reaction of 'OH and SO:-with chlorobenzene. Since G(SO:-) = 3.3, the yield of 30 20 10 C r L250 350 450 E 3 0 300 350 400 450 500 550 t 0 Ilnm l/n rn Fig.I Time-resolved transient absorption spectra obtained from the reaction of SO:-with (a)chlorobenzene and (b) toluene: (0)imme-diately, (x) 2 p,(a)10 ps after the pulse. Transient absorption spectrum measured at 2ps after the pulse from the reaction of 'OH with chlorobenzene (A),normalized to G('0H) = G(SOi-) = 3.3 for direct comparison. Dose per pulse =20 Gy, pH 5.5. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 'OH is normalized to 3.3 [from G('0H) = 5.61 for direct comparison with the spectrum observed in the reaction of SO;-. In the reaction of 'OH radical, the absorbance observed at 325 nm is higher with no absorption at 400 nm in contrast to the spectrum measured in the case of SO;-attack.The molar absorptivity of the 'OH adducts of chloro- benzene at 325 nm is determined to be 4600 dm3 mol-' cm-'. The transient species formed at 315 and 325 nm in the reaction of SO;-with toluene and chlorobenzene, respec- tively, may be attributed to their 'OH adducts. This assign- ment is in accord with the results obtained from our present study as well as with those previously rep~rted.~ The formation of 'OH adducts in the reaction of SO;-with benzene derivatives was explained' by both heterolysis [reaction (7), Scheme 11 of the initially formed SO;-adducts [reaction (5)] or via the hydration [reaction(lO)] of the ini- tially formed radical cation [reaction (4)].However, a clear- cut distinction between the two types of mechanisms cannot be made as the life-time of the SO;-adduct to benzene was reported" to be <lo0 ns; even in the case of benzonitrile, a compound with a strongly electron-withdrawing -CN group, the observed rate constant l5 for hydrolysis is >5 x lo6 s-'. The heterolysis rate was found" to increase with the introduction of a methyl substituent in the case of cyclohexene and ally1 alcohol which indicates an S,1 type hydrolysis of the SO;-adduct. Based on a pf value of -2.4 found with some substituted benzenes and benzoates, Neta et al. suggested" that the reac- tion mechanism proceeds by electron transfer from the benzene ring to SO;-. Our p+ value of -1.6 (vide infra) seems to suggest that the reaction proceeds both by direct electron transfer and addition-elimination processes Y Scheme 1 Generalized mechanism for the reaction of SO:-with substituted benzenes of the type C,H,-,X,Y (where X = F, C1 or Br and Y = CH,, CH,CI, CHCl,, CF, or OCH,).The intermediate radical cation is formed by both direct electron transfer [reaction (4)] and S,1 hydrolysis [reaction (9)] of the SO;-adduct formed via reaction (5). The radical cation is hydrolysed to give the correspond- ing 'OH adduct [reaction (lo)] except in the case of 3-chloroanisole. The formation of 'OH adduct radical via reaction (7) may not be likely under the experimental conditions employed. The phenoxyl radical is formed by reaction (8) only to a minor extent in the case of chlorobenzene and 3-chlorofluorobenzene.Reactions (6)and (12) are additional channels for mono- and di-chlorotoluenes. [reactions (4) and (5)]. Since reaction (7) is expected to be base-catalysed, it may not be predominant under our experi- mental conditions (pH 5.5). It can be concluded, therefore, that the formation of 'OH adducts occurs via the hydration of a radical cation formed directly [reaction (4)] or from the SN1 hydroysis [reaction (9)] of the SO:-adduct. The intermediate with a weak peak around 400 nm in the case of chlorobenzene is, possibly, due to the formation of a phenoxyl radical by the elimination of HSO, from the radical adduct [reaction (@I.In another recent study2' involving the reaction of *OH with 3-chloroanisole under acidic conditions, we have observed absorption maxima at 315 and 400 nm which are attributed to the corresponding 'OH adduct and phenoxyl radicals, respectively.It is also reported that phenoxy13' and trichlor~phenoxyl~radicals absorb around 400-430 nm. The alternative possibility of the formation of the radical cation [reaction (9)] of chloroben- zene may be unlikely as its khydtation of benzene. ' should be as high as that Dihalogenobenzenes and 3-Chloroanisole The transient absorption spectrum with 3-chloro-fluorobenzene exhibits two well defined peaks at 315 and 400 nm (Fig. 2). As compared with the spectrum with chloroben- zene, there was an increase in absorbance at 400 nm while the intensity at 3 15 nm decreased, indicating that the elimination of HSO, is more predominant.However, only a band with A,,, at 330, with no additional peak at 400 nm, was noticed in the case of 1,3-dibromobenzene suggesting that SO:-reacts quantitatively leading to the formation of the *OH adduct. The corrected transient absorption spectrum is shown in Fig. 2. The spectrum with 3-chloroanisole, having a strong electron-donating -OCH, group, shows two intense peaks at 290 and 475 nm and a shoulder around 325 nm within 2 ps after the pulse (Fig. 2). The nature of the spectrum clearly demonstrates the formation of the radical cation as observed ~earlier' [E nm ~= 7240~ and E~~~ nm = 3800 dm3 mol-cm -'3 with methoxy- and dimethoxy-benzenes.The radical cation in this case is stabilized owing to the presence of the electron-donating -OCH, group and its hydration [reaction (lo)] is, therefore, unlikely. The calculated molar absorpti- vities are 5600 and 3000 dm3 mol-' cm-' at 290 and 475 nm, respectively. The decay of the absorbance at these wave- lengths was considerable as can be seen from the time- resolved spectra measured at 2 and 16 ps after the pulse. In order to obtain information on the nature of the decay, absorption traces were taken at 290 nm on a longer timescale (100 and 500 ps) both at low (<OSkrad per pulse) and high (1.5 krad per pulse) doses. The fitting of the data in the former case, where the bimolecular decay is negligible, was not satisfactory owing to the poor signal-to-noise radio.The decay of the radical cation at high dose was monitored for three different concentrations (0.05,0.1 and 0.2 mol dm-3) of tert-butyl alcohol. The t1,2 obtained from absorption traces (inset of Fig. 2) is about 13 ps and such a decay indicates a first-order process. We cannot, however, conclude from our data whether the decay of the radical cation is due to its reaction with tert-butyl alcohol or S20i-, or to its rearoma- tization to form the phenyl-type radical [reaction (1 l)]. Mono- and Di-halogenotoluenes The spectral intensities of compounds having both -CH, and -C1 substituents (monochlorotoluenes) seem to depend on their relative positions. The spectra with 2-and 3-chlorotoluenes are more or less similar to that observed in the case of toluene with prominent peaks at 320 and 325 nm, J.CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 40 30 30 s 20 9)0 C C e-2j20 nE m X X m02 10 10 0 I I I I 0 2 3 300 350 400 450 500 550 2 I 300 350 400 450 500 550 Ilnm I/nm Fig. 2 Transient absorption spectra obtained from the reaction of SO;-with (a) 3-chloroanisole, (x) 2 ps and (A) 16 ps after the pulse; (6) 2-chloro-6-fluorotoluene (@), 3-chlorofluorobenzene (m), at 2 ps and 1,3-dibromobenzene (V)at 4 ps after the pulse, corrected for the com- petition reaction of SO>-with tert-butyl alcohol (see text). Dose per pulse =20 Gy, pH 5.5. The inset shows the trace for the decay at 290 nm in the case of 3-chloroanisole at a dose of 15 Gy per pulse and 0.2 mol dm-3 tert-butyl alcohol.respectively. In addition, a weak peak around 260-270 nm is observed with these systems which is considerably enhanced in the case of 4-chlorotoluene. This peak is assigned to the benzyl-type radical which is confirmed from the spectrum obtained by pulse radiolysis of N,O-saturated solutions of 4-chlorotoluene (lo-, mol dm-3) at pH 13, where the 'OH radical exists as 0'-(OHeH' + 0.-,pK = 11.9).,' It is known4 that 0'-reacts preferentially with the methyl group by abstraction of an H atom. The resulting spectrum from the reaction of 0'-with 4-chlorotoluene with absorption at 270 and 315 nm is nearly identical with that reported' in the case of 2-chlorotoluene. Both these spectra have similar fea- tures to those obtained from the reaction of SO;-with 4- chlorotoluene (Fig.3). The formation of benzyl-type radicals in high yield in the case of 4-chlorotoluene is due to H abstraction from the 30 A/nm CH, group [reaction (6)]. It is difficult to explain the ineffi- cient abstraction from 2-and 3-chlorotoluenes, though the steric effect may hinder the abstraction process in the case of 2-chlorotoluene. The formation of a benzyl-type radical, partly by deprotonation of the radical cation [reaction (12)] cannot, however, be ruled out. Pulse conductivity data with variation in pH are needed to obtain more quantitative infor- mation regarding the contribution of each of the two pro- cesses, as our results with absorption spectroscopy at pH 3, 5.5 and 9.6 are not conclusive.The contribution of SO:-leading to the formation of the benzyl-type radical is estimated to be about 10 and 40% for 3-and 4-chlorotoluenes, respectively, based on the reported4 E~~~~~ value of 14000 dm3 mol-' cm-' for the benzyl radical. Our spectral results therefore suggest that the order of the yields of 'OH adducts with monochlorotoluenes is 7c 60 50 9)uC 40 0 9 x 30 m2 20 10 0 250 300 350 400 450 500 550 Alnm Fig. 3 Transient absorption spectra obtained from the reaction of SO;-with (a) 2-chlorotoluene (FJ) and 3-chlorotoluene ( x) (corrected for the competition reaction of SO:-with tert-butyl alcohol). (6) Transient absorption spectra obtained from the reaction of SO;-(FJ) and 0'-( x ) with Cchlorotoluene.All spectra are recorded at 2 ps after the pulse. Dose per pulse =20 Gy. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 30 20 Cm e s X n 0-1c C 1 *I 250 300 350 400 450 500 550 0 300 350 400 450 500 550 l/nm Ifnm Fig. 4 Transient absorption spectra obtained from the reaction of SO;-with (a) 3,4dichlorotoluene (@) and 2,6-dichlorotoluene (m)both 10 ps after the pulse. Both the spectra are corrected for the competition reaction of SO:-with tert-butyl alcohol. (b) 3-Bromotoluene ( x ) 2 ps after the pulse. Dose per pulse =20 Gy, pH 5.5. meta > ortho > para. Our preliminary data32 on product analysis by the HPLC technique from the reaction of SO;-with monochlorotoluenes in the presence of Fe(CN)i -under steady-state conditions have shown that the yields of the pro- ducts formed from benzyl-type intermediates are very small in the case of 2- and 3-chlorotoluenes as compared with 4- chlorotoluene.The extent of the formation of the benzyl-type radical seems to depend on the position of -C1 on the benzene ring. This is evident from the differences in the observed intensities at 270 nm in the spectra with 3,4- and 2,6-dichlorotoluenes. The spectrum with 3,4-dichlorotoluene is more or less similar to that observed with 4-chlorotoluene, whereas the spectrum with 2,6-dichlorotoluene resembles those of 2-and 3-chlorotoluene (Fig. 4). One interesting observation of the absorption spectra from mono- and di-chlorotoluenes is that, when the C1 atom is at position 4 relative to the CH, group (e.g.4-chlorotoluene, 3,4-dichlorotoluene), the formation of the benzyl-type inter- mediate by abstraction of an H atom from the -CH, group is more favoured. The effect of the nature of the substituent on the transient spectra can be seen from a comparison of the spectra of 2- chloro-6-fluorotoluene and 2,6-dichlorotoluene (A,,, = 270, 330, 410 nm) with those of 1,3-dibromobenzene, 3-bromo- toluene and 3-chlorofluorobenzene. The spectrum of 2-chloro-6-fluorotoluene has two overlapping peaks at 300 and 330 nm and two additional small peaks at 390 and 490 nm (Fig. 2). Only peaks (A,,, = 320-330 nm) corresponding to their 'OH adducts [reaction (lo)) are formed with 1,3-dibro- mobenzene and 3-bromotoluene (Fig.4), whereas an addi- tional peak at 410 nm is noticed, as mentioned earlier, in the case of 3-chlorofluorobenzene. a-Fluoro-and a-Chloro-toluenes The spectra with compounds (3,4,a-trichlorotoluene,2-chlorobenzylchloride, 2,6,a,a-tetrachlorotoluene and 2-chlorobenzotrifluoride) having halogen substituents both on the benzene ring and on the side group, have similar features with peaks corresponding to 'OH adducts and weak maxima above 400 nm. Kinetics for the Reactions of SO;-Radicals Decay of SO;-Rates for the reaction of SO;-with substituted benzenes may be determined from the decay of its absorption at 460 nm. N,-saturated S20i-solutions (1.5 x mol dm-j) con- taining 0.2 mol dm-, tert-butyl alcohol were pulse radiolysed (10-20 Gy per pulse) to monitor the decay of SO;-in the absence of solute.A first-order decay of SO;-with tIl2 = 5 ps was observed [(b)of inset of Fig. 5). The possible reactions for the decay of SO;-in the absence of any solute are: (i) its self-reaction, so;-+ so;-+ s,o;-(13) k = 7.6 x lo8 dm3 mol-' s-' ,21 1.6 1.2 time 1 ps -4 c +-....IIu) 02 0.8 -2 -rc 0.4 0 0.2 0.4 0.6 0.8 1.0 [4-chlorotoluene]/l 0-3 mol dw3 Fig. 5 k,, as a function of [4-chlorotoluene]. Inset shows the decay of SO:-at 460 nm, (a) with and (b) without bchlorotoluene mol dm-'). Dose per pulse =20 Gy, pH 5.5. (ii) its reaction with S,Oi-.or any impurity, X-(e.g.C1-) in the solution, SO;-+ S,O;-(X-) --+ SO:-+ S,Og-o(') (14) k(S0;-+ S,Oi-) = 6.6 x lo5 dm3 mol-' s-',,' (iii)and its reaction with the 'OH radical scavenger, tert-butyl alcohol, SO4-+ (CH,),COH -,HSO, + 'CH,(CH,),COH (15) k = 9 x lo5 dm3 mol-' s-'.~~ The tl,, for the decay of SO;-due to reaction (14) under our experimental conditions ([S,Oi-] = 1.5 x lo-, mol dm-3) should be at least 70 ps. The first half-life for the second- order decay of SO;-by reaction (13) is calculated to be about 200 ps considering G(SO;-) = 3.3 and the dose per pulse of 20 Gy. The observed decay of SO;-cannot be ascribed to its reaction with the impurity, Cl-, which is present only to an extent of 0.005% in the K2S208 used by us. Thus, reaction (15) seems to be the major pathway for the observed decay of SO;-in the absence of solute.A value of 1200 dm3 mol-' cm-' was obtained for the molar absorptivity of SO;-at 460 nm measured relative to an assumed G~500nm = 21 500 dm3 mol-' cm-' for (SCN);- and G(SO;-) values of 1100-1600 dm3 mol- ' cm-' for SO;-measured at 440-450 using either pulse radiolysis or flash photoysis tech- niques. Evaluation of Second-order Rate Constants The second-order rate constants for the reaction of SO;-with the substituted benzenes chosen in this study are deter- mined from the least-squares fit of the plots of kobs us. solute concentration. The pseudo-first-order rate constant of SO;-with tert-butyl alcohol (kobs = 1.7 x lo5 s-') is taken as the Y intercept and a typical plot is shown in Fig.5 in the case of Cchlorotoluene. The values of the rate constants obtained with different solutes are summarized in Table 1. The highest second-order rate constant, among the investi- gated compounds, is observed with 3-chloroanisole (k = 1.2 x 10" dm3 mol-' s-'). The rate constant in this = 3.3. Earlier ~~rker~~~*~~-~~ have obtained E s-'3 owing to the deactivation of the benzene ring for SO;-attack by electron-withdrawing groups such as -CH,Cl, -CHCl, and -CF3 . The rate constants for the addition of the 'OH radical to the different solutes used in this study were reported8 by us to be in the range (2-9) x lo9 dm3 mol-' s-', whereas the values for SO;-attack are in the range (108-10'o) dm3 mol-' s-' indicating that the rate of the latter reaction is more influenced by the nature of the substituent.Such an observation was also made by Neta et al." who report rate constant values varying from 5 x lo9 for anisole to <lo6 dm3 mol-' s-l for nitrobenzene. Eflect of Structure on Reactivity by the Hammett Treatment It is known that the SO;-radical anion behaves as an elec- trophile and any substituent effect can be quantitatively cor- J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 case is determined from the growth of the absorbance at 290 nm, instead of that at 460 nm, to avoid interference from the intermediate transient product which also absorbs at 460 nm. The high reactivity of SO;-with 3-chloroanisole, as in the case of dimethoxybenzene~,~ can be attributed to direct elec- tron transfer [reaction (411 rather than to formation of an intermediate adduct to the benzene ring [reaction (5)].The rate constants for the reaction of SO;-with chloro- benzene, toluene, 2-, 3-, and 4-chlorotoluenes, 3-bromo- toluene, 3,4-dichlorotoluene, 3-chlorofluorobenzene and 2- chloro-6-fluorotoluene are 2lo9 dm3 mol-s-'. The second-order rate constants for these systems are determined from at least five different concentrations of the solute C(0.2- 1) x mol dm-3] and their accuracy is within +lo%. A comparison of the rate constants for the reaction of SO;-with chlorobenzene (k = 1.5 x lo9 dm3 mol-' s-') and benzene" (k = 3 x lo9 dm3 mol-' s-') shows that the reac- tion rate is affected by the -C1 group.The rate constants for 2-~hlorobenzotrifluoride,3,4,a-tri-chlorotoluene, 2,6,a,a-tetrachlorotolueneand 2-chloro-ben-zylchloride (whose rates are lower) were measured with relatively high concentrations C(O.6-1) x mol dm-3] ; however, the concentration range is restricted to < mol dm-3 of the solute owing to the solubility limit. These solutes have apparently low reactivity [k = (1-7) x 10' dm3 mol-' Table 1 Bimolecular rate constants (10' dm3 rno1-l s-') obtained in the reaction of OH and SO;-with some substituted benzenes bimolecular rate constant probablesite of substrate OH" so;- absorption maxima/nm UCJll attack chlorobenzene -1.5 270, 325, - 400 toluene -3.1 315,400 -0.31 3-chlorofluorobenzene 4.8 1.o 315, 410 - 3-chloroanisole 9.3 12.0 290,475 -0.58 2-chlorotoluene 6.5 1.7 260, 320 -0.17 3-chlorot oluene 3.5 0.9 260, 325 -0.11 Cchlorotoluene 5.5 1.1 270, 315 -0.31 3,4-dichlorotoluene 2,6-dichlorotoluene 1.7 -0.9 0.4 270, 315 270, 330 -0.11 +0.23 2-chloro-6-fluorotoluene 4.2 1.2 410 300, 330 +0.06 3,4,a-trichlorotoluene 2-chlorobenzyl chloride 2,6,a,a-tetrachlorotoluene 2.5 4.1 4.9 0.2 0.7 0.2 390,490 290, 330 325,460 290, 320 +0.34 +0.10 - 410 2-chlorobenzotrifluoride -0.1 315, 390 - 3-bromotoluene 4.9 1.7 320 -0.10 1,3-dibromobenzene -0.5 330 - " Taken from ref.8. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 related by the Hammett equation. However, each of the substituted benzenes used in our study contains a halogen atom is one of the substituents and an electron-donating (-OCH,, CH,) or an electron-withdrawing (-CF,, -CHCl, ,-CHCl,) group at different positions.A similar procedure, as in the case of 'OH attack,8 involv- ing the algebraic sum (acal)of the Hammett constant^^^-^' (a; or a:) for para and meta substituents and the Taft constant42 (a*)for the ortho substituent was used to deter- mine the more probable site for SO;-attack where Q,,, is expected to be minimum. The minimum aCa1values for differ- ent compounds are shown in Table 1 along with the positions for the more probable site for SO:-attack. The rate constants for diffusion-controlled reactions are not normally expected to obey the Hammett relationship as was seen6 from the levelling in log k us.acalplots for values of log k z9.5 in the oxidation of hydroxycyclohexadienyl rad- icals by IrC1;- and Fe(CN);-. Since the log k values of the compounds chosen in this study are <9.5 except for 3- chloroanisole, the plateau is not apparent. In the case of 3- chloroanisole (k = 1.2 x 10" dm3 mol-' s-'), the selection of the site of attack must occur within the encounter complex as was suggested43 for the reaction of 'OH with biphenyl (k = 1.04 x 10" dm3 mol-' s-') and phenol (10" dm3 mol-' s-I). The Hammett-type plot shown in Fig. 6 with a p+ values of -1.6 has a good linear correlation indicating that a similar reaction mechanism is operative in all cases.Our p+ value of -1.6 is in between the values reported" for 'OH (p' = -0.5) and SO;-(p+ = -2.4) with substituted ben- zenes. Therefore, we suggest that the reaction proceeds by both direct electron transfer as well as addition4imination channels. Though H abstraction is noticed with 4-chlorotoluene and 3,4-dichlorotoluene, they are included in the Hammett plot as the abstraction process is not exclusive. It has been rep~rted~,~~,~~ that the 'OH radical behaves as a weak electrophile and reacts by addition to the aromatic ring with low selectivity (p' = -0.5). The Hammett treat- ment of our data shows that a better correlation is obtained in the case of SO:-reaction than for 'OH attack. This implies that the positions of the probable site of attack (Table 1) may have greater influence on the reaction rate in the former case.However, only from determination of the product distribution of the oxidation of different isomeric 'OH adducts under steady-state conditions can one obtain direct evidence regarding the probable sites of attack predict- lo.Or' -9.5 rn -2 9.0 -8.5 8.01 I I I -0.6 -0.4 -0.2 0 0.2 0.4 OG3l Fig. 6 Hammett plot for the reaction of SO:-with substituted benzenes. 1, 3-Chloroanisole; 2, toluene; 3, Cchlorotoluene; 4, 2-chlorotoluene; 5, 3-chlorotoluene and 3,4-dichlorotoluene; 6, 3-bromotoluene; 7, 2-chloro-6-fluorotoluene; 8, 2-chlorobenzyl-chloride ;9,2,6-dichlorotoluene; 10, 3,4,a-trichlorotoluene. ed by the gCalvalues. Such a product analysis with isomers of monochlorotoluenes is in progress.Conclusion The Hammett analysis and spectral features of the transients formed in the reaction of SO;-with substituted benzenes reveal that the reaction mechanism involves both direct elec- tron transfer and addition4imination processes, eventually leading to the formation of the corresponding hydroxy- cyclohexadienyl radicals. Only with 3-chloroanisole is the formation of a radical cation observed. H abstraction from the CH, group is an additional reaction channel, especially in the case of 4-chlorotoluene and 3,4-chlorotoluene. The extent of each of these processes is dependent on the nature and the relative position of the substituents. The reaction rates [k = (108-10'o) dm3 mol-' s-'1 in the case of SO:-attack are more influenced by the substituent than those of 'OH attack.This is also reflected in the magni- tude of the Hammett constants obtained with SO;-(p' = -1.6) and 'OH (p' = -0.5) and in the differences in the transient absorption spectra obtained with these two reacting species. This suggests that SO;-is a stronger electro- phile and is more selective in its reaction. This study demon- strates that substituted benzenes of the type C,H, -,X,Y (where X = F, Cl or Br and Y = CH, , CH2Cl, CHCl,, CF, or OCH,) are an interesting class of compounds for investi- gation of structure-reactivity relationships by radiation chemical techniques. The authors thank Dr. R.M. Iyer of the BARC and Prof. M.S.Wadia of the University of Poona for their interest in this work. They also thank the referees for their suggestions, particularly regarding the interpretation of the phenoxyl radical absorption spectra, and the DAE for supporting this work. C.T.A. is thankful to the CSIR for the award of a senior research fellowship. References 1 J. Lichtsheidl and N. Getoff, Znt. J. Radiat. Phys. Chem., 1976, 8, 661. 2 S. Gordon, K. H. Schmidt and E. J. Hart, J. Phys. Chem., 1977, 81, 104. 3 H. Mohan, M. Mudaliar, B. S. M. Rao and J. P. Mittal, In?. J. Radiat. Phys. Chem., 1992,40, 513. 4 K. Sehested, H. Corfitzen, H. C. Christensen and E. J. Hart, J. Phys. Chem., 1975,79,310. 5 S. Steenken and N. V. Raghavan, J. Phys. Chem., 1979,83,3191. 6 G.V. Buxton, J. R. Langan and J. R. L. Smith, J. Phys. Chem., 1986,90,6309. 7 P. B. Draper, M. A. Fox, E. Pelizetti and N. Serpone, J. Phys. Chem., 1989,93, 1938. 8 H. Mohan, M. Mudaliar, C. T. Aravindakumar, B. S. M. Rao and J. P. Mittal, J. Chem. SOC., Perkin Trans. 2, 1991, 1387. 9 P. O'Neill, S. Steenken and D. Schulte-Frohlinde, J. Phys. Chem., 1975,79,2773. 10 P. Neta, V. Madhavan, H. Zemel and R. W. Fessenden, J. Am. Chem. SOC.,1977,99,163. 11 S. Steenken, P. O'Neill and D. Schulte-Frohlinde, J. Phys. Chem., 1977,81,26. 12 V. Jagannadham and S. Steenken, J. Am. Chem. SOC., 1988, 110, 2188. 13 Q. Sun, G. N. R. Tripathi and R. H. Schuler, J. Phys. Chem., 1990,94,6273. 14 C. von Sonntag, The Chemical Basis of Radiation Biology, Taylor and Francis, London, 1987, p.40. 15 (a)S. Steenken, J. Chem. SOC., Faraday Trans. I, 1987, 87, 113; (b) S. Steenken, in Free Radicals in Synthesis and Biology, ed. E. Minisci, Nato AS1 Series C-260, Kluwer Academic, Dordrecht, 1989, p. 213. 604 J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 16 P. Neta and A. Harriman, Photoinduced Electron Transfer, ed. 30 N. V. Raghavan and S. Steenken, J. Am. Chem. SOC., 1980, 102, 17 M. A. Fox and M. Chanon, Elsevier, Amsterdam, 1988, p. 110. G. V. Buxton, C. L. Greenstock, W. P. Helman and A. B. Ross, 31 3495. J. W. T. Spinks and R. J. Woods, Introduction to Radiation 18 19 J. Phys. Chem. Re5 Data, 1988, 17, 513. C. T. S. Lian and J. P. Mittal, Radiat. Phys. Chem., 1984,24,209. L. Eberson, Adu.Phys. Org. Chem., 1982,18,79. 32 33 Chemistry, Wiley, New York, 1976, p. 259. G. Merga and B. S. M.Rao, 1993, unpublished. E. Hayon, A. Treinin and J. Wilf, J. Am. Chem. SOC., 1972,94,47. 20 B. Lesigene, C. Ferradini and J. Pucheault, J. Phys. Chem., 1972, 34 W. Roebke, M. Renz and A. Henglein, lnt. J. Radiat. Phys. 76, 3676. Chem., 1969, 1, 39. 21 P. Y. Jiang, Y. Katsumura, R. Nagaishi, M. Domae, K. Ishikawa and K. Ishigure, J. Chem. Soc., Faraday Trans., 1992,88,1653. 35 36 E. Hayon and J. J. McGarvey, J. Phys. Chem., 1976,71,1472. L. Dogliotti and E. Hayon, J. Phys. Chem., 1967,71,2511. 22 W. J. McElroy and S. J. Waygood, J. Chem. SOC., Faraday 37 V. E. Heckel, A. Henglein and G. Beck, Ber. Bunsenges. Phys. Trans., 1990,86, 2557. Chem., 1966,70,149. 23 F. H. Quina and D. G. Whitten, J. Am. Chem. SOC., 1975, 97, 38 0. P. Chawla and R. W. Fessenden, J. Phys. Chem., 1975, 79, 1603. 2693. 24 M. E. Snook and G. A. Hamilton, J. Am. Chem. SOC., 1974, W, 39 P. Neta and L. M. Dorfman, Adv. Chem. Ser., 1968,81,222. 860. 40 M. K. Eberhardt, J. Phys. Chem., 1975,79, 1930. 25 T. I. Balkas, J. H. Fendler and R. H. Schuler, J. Phys. Chem., 41 J. March, Advanced Organic Chemistry Reactions, Mechanism 26 1970, 74, 4497. S. N. Guha, P. N. Moorthy, K. Kishore, D. B. Naik and K. N. and Structure, McGraw Hill, Kogakusha Ltd., New York, 1968, p. 241. Rao, Proc. Indian Acad. Sci., 1987,99, 261. 42 J. Hine, Physical Organic Chemistry, McGraw Hill, New York, 27 E. M. Fielden, The Study of Fast Processes and Transient 1962, p. 97. Species by Electron Pulse Radiolysis, ed. J. H. Baxendale and F. 43 X. Chen and R. H. Schyler, J. Phys. Chem., 1993,97,421. Busi, Reidel, Boston, 1984, p. 59. 44 P. Neta and R. H. Schuler, J. Am. Chem. SOC.,1972,94, 1056. 28 M. S. Panajkar, P. N. Moorty and N. D. Shirke, BARC Rep., 1988, 1410. 29 M. Mohan, Ph. D. Thesis, University of Poona, 1993. Paper 3/05999E; Received 7th October, 1993