|
1. |
Mechanisms of peroxide decomposition. An ESR study of the reactions of the peroxomonosulphate anion (HOOSO3–) with TiIII, FeII, and α-oxygen-substituted radicals |
|
Journal of the Chemical Society, Perkin Transactions 2,
Volume 1,
Issue 8,
1990,
Page 1281-1288
Bruce C. Gilbert,
Preview
|
PDF (1027KB)
|
|
摘要:
J. CHEM. SOC. PERKIN TRANS. 2 1990 Mechanisms of Peroxide Decomposition. An ESR Study of the Reactions of the Peroxomonosulphate Anion (HOOSOJ with Ti"' Fe",and a-Oxygen- substituted Radicals Bruce C. Gilbert' and the late Jonathan K. Stell Department of Chemistry University of York Heslington York YO 1 500 It is shown that decomposition of the peroxymonosulphate anion (HOOS0,-) in aqueous solution with the low-valence metal ions Till' and Fell proceeds via one-electron transfer to yield SO4-' and OH- (rather than Sot-and 'OH) with rate constants in the range lo2-lo5dm3 mol-' s-l. A variety of subsequent reactions of SO4-' with added substrates has been demonstrated. ESR results also reveal the efficacy of HOOS03-as an oxidant for carbon-centred radicals generated in situ.In addition to the occurrence of rapid one-electron oxidation of alkyl radicals (e.g. Et' with k 3.8 x lo5dm3 mol-' s-l) it is also established that certain oxygen-conjugated radicals undergo a novel oxidation reaction involving oxygen-transfer and subsequent fragmentation ('CHMeOH gives Me' with k 1.1 x 1O6 dm3 mol-' s-l). The formation of free radicals during the decomposition of various peroxides by certain transition-metal ions has been effectively demonstrated by continuous-flow ESR experiments which provide information on the first-formed radicals the kinetics of their formation and subsequent radical reactions. l4 Examples recently described include the Fenton reaction and its analogues3 [reaction (l) M"+ = Fe" Ti"'] and related decomposition of the peroxydisulphate anion [reaction (2) M"+ = Fe" Ti'" Cu'].Steady-state analysis (coupled for example with computer simulation) also leads to kinetic information on the initiation reactions and the oxidation or reduction of free radicals with metal ions3- [see e.g. reactions (3) and (4)] as well as the oxidation of relatively nucleophilic free radicals with peroxides [see e.g. reaction (5)].2 'CHMeOH + Fe'"-Fe" + H+ + MeCHO (3) 'CH,CMe,OH + Fe"-Fe"'+ CH2&Me2 + OH-(4) 'CHMeOH + H2O2-+ H+ + MeCHO + HO' + HO-(5) In the work to be described here we set out to investigate the ease with which the low-valent transition-metal ions Ti"' and Fe" react with the peroxomonosulphate ion HO-OSO -and in particular to determine the effect on these reactions of the ligands co-ordinated to the metal. We have previously estab- lished that for Ti'" Ti"'-EDTA and Fe"-EDTA the reaction of hydrogen peroxide is normally considerably faster than that of peroxydisulphate [for example for Ti"' k is ca.2 x lo3 dm3 mol-' s-' compared with 200 dm3 mol-' s-lI4 and have also noted however how certain ligands (e.g. diethylenetriamine-penta-acetic acid DTPA) inhibit the Ti"'-H202 but not the Fe"-H2 0 reaction. Transition-metal decomposition of HOOSO -has been reported to initiate radical polymerization although there is 'v6 little evidence to indicate whether the initial reaction proceeds via reaction (6) or (7). Although the reaction with Co" is believed' to occur via reaction (7) to give SO,-' pulse radiolysis studies8 lead to the conclusion that in this case reaction proceeds almost exclusively (ca.80%) to give the hydroxyl radical [reaction (6)]. Our aim has been to employ ESR spectroscopy to determine whether one-electron transfer to HOOS03-from a low-valent metal ion parallels that achieved by pulse-radiolysis to measure the rates of such reactions and to establish the efficacy and mechanism of oxidation of organic radicals by the peroxide [cf reaction (5)]. HOOSO~-%HO* + ~0~~-(6) HOOSO~-5HO-+ so4-* (7) Results and Discussion Our experiments typically involved the use of a continuous-flow system in which three solutions were mixed shortly (ca. 30 ms) prior to flowing through a sample cell in the cavity of an ESR spectrometer.The three solutions normally contained respec- tively (i) the metal ion (Ti"' Fe") in some cases with added complexing agent (e.g. EDTA) (ii) the peroxymonosulphate anion and (iii) the substrate (see the Experimental section for details). The pH was adjusted by addition of sulphuric acid ammonia or sodium hydroxide to streams (i) or (iii) and in certain experiments hydrogen peroxide was also added to streams (ii) or (iii). (a) The Reaction between HOOS03-and Metal Ions (Ti"' Fe").-To determine whether SO,-' or 'OH (or both) result from reaction between Ti"' and Fe" we carried out experiments in the presence of several alkenes for which the 'OH and SO4-' adducts (obtained from separate experiments with H202 and 10,11 S208,- respectively) are distinguishable by virtue of their ESR parameters. In the reaction of HOOS03 -with Ti"' at pH ca.2 [in the absence or presence of suitable complexing agents EDTA and NTA (nitrilotriacetic acid)] the addition of a variety of alkenes led to the detection of relatively intense and well-resolved ESR spectra which establish unambiguously that SO4-*adducts alone are formed "*''[see reactions (8) and (9) Figure 1 and Table 13 no 'OH adducts were detectable [a limit of ca 5% can be placed on the contribution from reaction (6)]. Further support for this conclusion is provided by results for other substrates (see also Table l) notably the characteristic oxidative decarboxylation of carboxylate anions ('OH reacts g 2.0025al:,n 1 I 2.0mT Figure 1.ESR spectra of the adducts of SO,-' and allyl alcohol detected during the reaction of Ti"'/HOOSO,- and HOCH,CH=CH at pH 2 0,'CH(CH,OH)CH,OSO,-; x ,'CH2CH(OS0,-)CH20H. Table 1. Radicals detected by ESR in the reaction of the Ti"'/HOOSO -couple with aliphatic alkenes and carboxylic acids.a*b Substrate Radicals detected HOCH,CH=CH 'CH(CH,OH)CH,OSO - 'CH,CH(OSO,-)CH,OH HOCH,CH=CHMe 'CHMeCH(OS0 -)CH,OH 'CH(CH20H)CH(OS0,-)Me HOCH,CMe=CH 'CMe(CH,OH)CH,OSO -HOCH,CH=CHCH,OH 'CH(CH,0H)CH(OS0,-)CH20HMe,C====H 'CMe,CH,OSO -MeCH=CHMe 'CHMeCH(OS0 -)Me H0,CCHXHMe 'CHMeCH(OS0 -)CO,H 'CH(CO,H)CH(OSO -)Me HO,CCH=CHCO,H 'CH(CO,H)CH(OSO,-)CO,H HO,CCH=CMeCO,H 'CMe(CO,H)CH(OSO,-)CO,H MeCOzH 'Me 'CH ,CO H EtC0,H 'CH,CH,CO,H 'Et 'CHMeC0,H Me,CHC02H 'CMe,CO,H 'CH,CHMeCO,H 'CHMe Me,CCO,H 'CH,CMe,CO,H Experiments conducted at pH 2-2.5.ESR parameters as previously reported refs. 10-12. via hydrogen-atom abstraction rather than electron-transfer) and the formation of the methyl radical (via Bu'O') from BU'OH.' Ti"' + HOOS03- -,Ti" + HO-+ SO4-' SO4-' + HOCH2CH=CHCH20H+ OSO,-IHOCH2CH<HCH20H (1) OSO,-I2 HOCH2CHkHCH20H -non-radical products The rate constants for reaction (8) for a variety complexes were determined via a steady-state kinetic analysis * See refs. 1 2 and 13 for derivation and for kinetic applications in metal-peroxide systems. Here and elsewhere in the text the concentrations referred to are those after mixing.$ The finding l6 that SO,-* reacts readily with OH- to give 'OH (kca. 5 x lo7 dm3 mol-' s-l) indicates that production of 'OH and SO,'-from peroxymonosuiphate is thermodynamically preferred. J. CHEM. SOC. PERKIN TRANS. 2 1990 and measurement of the concentration of radicals [e.g. (l)] formed by the scavenging of SO4-' with appropriate concen- trations of suitable substrates [e.g. via reaction (9)J2*13It has been shown l3 that for effective scavenging of an intermediate (primary) radical to give a radical [e.g. (l)] which is not itself either oxidized or reduced and which decays simply via bimolecular termination [reaction (lo)] kinetic analysis of the resulting pseudo steady-state condition achieved in the cavity (i.e.after time t) leads to two important equations.* The first indicates that the observed radical concentration should vary with [HOOSO,-] in such a way that it is a maximum for the value of [HOOS03-] indicated in equation (11); the second relates the observed radical concentration to time after mixing (and hence to the flow-rate) 1[HOOS03-] (for R'JmaX) = -(11) k8t k8CHOOSO3 -Itln[R'] = const. -3 (12)L In an initial series of experiments the steady-state concen- trations of SO4-*adducts (to both allyl alcohol and butene-1,4- diol) were monitored as a function of [HOOSO -lo{for fixed [Ti"'] and t and with [HOOS03-lo p [Ti1''l0 to ensure that the concentration of the former remains constant and with alkene concentrations? ca.0.1 mol dm- (to ensure scaveng- ing)) and as a function of flow-rate (and hence t) all other conditions and concentrations being held fixed. The results confirmed the behaviour expected on the basis of the kinetic analysis (a 'peaking' in [R'] as a function of [HOOSO,-] and a linear dependence of ln[R'] on t see ref. 3). Data from both types of experiment were analysed to yield values for k8 (the initiation rate constant for reaction with Ti"') for different complexes (see Table 2); the reaction of the Ti'II-DTPA complex was evidently too slow to follow using these techniques (only weak signals from organic radicals at high [HOOS03-] could be detected) and the value for the rate constant given in Table 2 was derived by monitoring the pseudo first-order decay of the ESR signal (g ca.1.965) from Ti'II-DTPA itself both when the flow was stopped (see ref. 3) and via experiments with allyl alcohol at different flow-rates. Comparison of results for a variety of peroxides (see Table 2) show that for most complexes studied (and for the complexed ion) the rate constants are similar for HOOSO,-and H202 (and somewhat greater than for4*10 S2OS2-). For both HOOSO -and H202 (but more markedly with the latter ,) the use of DTPA leads to a significant reduction in the rate of electron-transfer. Detailed information on the redox potentials of the full range of metal complexes comployed is not apparently available but it might be anticipated that the E" values for these should not differ greatIy:l4 the variation in k8 e.g.the acceleration with NTA and the retardation with DTPA therefore suggests that steric factors are more important than electronic factors in controlling the reaction. In particular we note that the DTPA complex the reaction of which is slow is more crowded than that of EDTA.' This observation coupled with the finding that the reaction yields SO4-* rather than *OH (in contrast with pulse radiolysis results) suggests that electron-transfer proceeds uia an inner-sphere mechanism and decomposition of a metal-peroxide complex.$ Reaction between FeII-EDTA Fe"-NTA and Fe'I-DTPA also yielded SO4-' as judged by the detection of signals from the appropriate adducts. That the relative weakness of the signals under typical flow system conditions reflects the rapidity of the initiating reaction (13) was confirmed by kinetic analysis as indicated above [k13 ca.3 x lo4 dm3 mol-' s-'1 Fe" + HOOSO; --+Fe"' + HO-+ SO,' (13) J. CHEM.SOC. PERKIN TRANS. 2 1990 Table 2. Rate constants (k/dm3 mol-' s-') for the reactions between Ti"' complexes and both HOOSO -and H202.(l*b Complex HOOSO,-H202 Ti"'( aq) 7.0 x lo2 2.1 x 103 Ti"'-EDTA 3.5 x lo2 (pH 2) 4.5 x 103 (PH 3-71 1.3 x lO'(pH8-9) Ti"'-DTP A 1.5 x lo2 5 Ti"'-NTA 1.0 x 104 ca. 1 x 104 "At room temperature and at pH 2 unless indicated otherwise estimated error in k f 15%. For abbreviations see text. Rate constants for H202 are taken from ref. 3 except for Ti"'-NTA (determined in this study).Contrast rate constants for Ti"' with H2P20E2- (2 x lo4 dm3 mol-' s-' ref. 11) and S2OEZ-(2 x lo2 dm3 mol-' s-l ref. 4). c g2'T25 i 2.0 mT -Figure 2. ESR spectra obtained from the reaction of the Ti"'/HOOSO -couple with ethanol ([Ti'"SO,] 1.7 x lW3 mol dm-3 [HOOS03-] 1.0 x lWz mol drn- [EtOH] 0.5 mol dm-3 pH 2) 0 'CHMeOH; x ,'CH2CHzOH;0 Me'. (b) The Oxidation of Radicalsby HOOSO -.-When ethanol was employed as substrate with the Ti"'-HOOS03 -couple three radicals were detected namely the a-and P-radicals 'CHMeOH and 'CH2CH,0H respectively and the methyl radical (see Figure 2). As the concentration of HOOS03- was increased a decrease in the concentration of 'CHMeOH was matched by a marked increase in [Me'] [see Figure 3(a).].This leads to the suggestion that Me' is formed by oxidation of the a-radical by the peroxide for example as depicted in reaction (14); this is in contrast with a formal electron-transfer process [cJ reactions (15a) or (15b) a reaction-type established for radicals of the type 'CR'R'OH with2 H202 and17 Sz082- (see e.g. reaction (5)]. The suggested involvement of an a-hydroxyalkoxyl radical (2) in the novel decomposition is supported in part by the observation" that such species fragment rapidly to alkyl radicals other evidence was sought via the experiments described below. HooSo3-'CHMeOH + MeCHOH Me' + HC02H (14)[a-I-(2) 'CHMeOH MeCHO + H+ +HOP + SO4-' (15a) + HOOSO MeCHO + H+ + HO' + S042-(156) (i) Reactions of Other Oxygen Conjugated Radicals.-Re- sults for a number of alcohols and and ethers with the Ti"'/HOOSO -couple (see Table 3) show behaviour parallel to that for ethanol.For example propan-2-01 gave both 'CMe20H and 'CH,CHMeOH as well as strong signals from Me' which grew at the expense of 'CMe,OH for increased [HOOS03-] the higher ratio of [Me'/['CMe,OH] than [Me']/C'CHMeOH] (for ethanol) for given [HOOS03-] is consistent with a more ready fragmentation of the intermediate for propan-2-01. Reaction of propan-1-01 and pentan-3-01 led to the detection of the ethyl radical as well as radicals expected from SO4-' attack on the substrates. Reaction with butan-2-01 (for which SO4-' itself gives largely 'CMeEtOH) produced Et' and Me' with the former in significantly higher concentration [as might be anticipated from preferential fragmentation of the intermediate (3)to give the more stable alkyl radical reaction (16a)l.The ratio [Et*]/[Me'] was somewhat dependent upon [HOOS03-] in a manner that suggests that the former is itself more readily oxidized by the peroxide (see later). Other examples of C-C cleavage in radicals obtained from alkanols and diols are included in Table 3. o. 85% Et' + MeCO (16a) EtkMeOH ELMe {H00S03-b AH Me' + EtC02H (166)15%(3) Reaction of both diethyl and di-isopropyl ethers also gave strong signals from Me' in addition to 'CHMeOEt and 'CMe20CHMe2 respectively (and weaker signals from the appropriate P-radicals) again [Me'] increased as [HOOSO -1 was increased.Tetrahydrofuran gave the ring-opened radical (4). These observations lend considerable support to the proposal that an alkoxy intermediate is involved [see e.g. reaction (17)] since it has previously been shown that de- composition of the hydroperoxides derived from these ethers by one-electron transfer from Ti"' leads to the fragmentation identical with that observed in the reactions described here [see reactions (18) and (19)].18 These findings strongly suggest that alcohols and ethers react in analogous fashion and that the reactions of the former to give alkyl radicals are not a consequence of direct oxidation of the OH groups to give alkoxyl radicals. 0'IMekHOEt H00S03-* MeCHOEt d Me' + EtOCH=O (17) OOH 0'I IMeCHOEt + Ti"'-+ Ti" + OH-+ MeCHOEtd a:" Me' + EtOCHO (18) + Ti"' +TiIV + OH-+ a;-'CHZCH2CH,OCHO (19) (4) Reaction of 3-phenylpropan-1-01 led to the detection of 'CH2CH2Ph evidently formed uia oxidation of 'CH(0H)- CH,CH,Ph (which is obtained from SO4-' in a manner previously described 19).Reaction of 2-methylpropan- 1-01 led to the detection of the anticipated a- p- and y-radicals but perhaps surprisingly there was only a trace of the prop-2-yl radical the presence of which would have been expected in terms of the mechanism outlined above. No trace of But' could be J. CHEM. SOC. PERKIN TRANS.2 1990 /.' I L 9 a U 0.0 0 10 20 0 10 20 mol dm-3EHOOS0,1/M-3 mol drn-3 CHOOSO~l/10~3 Figure 3. (a) Variation of relative radical concentration with [HOOSO,-) in the reaction of EtOH with Til"/HOOSO,-at pH 2 with mixing time 0.03 s (for conditions see Figure 2) together with simulations based on the kinetic Scheme given in the text and values for k and k given in Table 4; 0 'CHMeOH X 'CH,CH,OH 0 Me' (b) Variation in the concentrations of radicals derived from 'OH (from Ti"'-EDTA and H,O,) and EtOH with [HOOS03-] ([Ti'"] = [EDTA] = 1.7 x rnol drn- [H202] = 1.7 x 10-' mol dm- [EtOH] = 0.5 mol dm-3 pH 2) 0,'CHMeOH; x ,'CH,CH,OH; 0 Me'.Table 3. Radicals detected from reaction of the Ti"'/HOOSO -couple with aliphatic alcohols and ethers." R2* (derived from (a-radical) reaction of R'*) Other radicals detected Substrate R {:E2CMe20HMe,COH -MeOH 'CH,OH EtOH 'CHMeOH Me' 'CH,CH,OH Me,CHOH 'CMe,OH Me' 'CH,CHMeOH 'CHMeCH,OHPrOH 'CHEtOH Et' 'CH,CH,CH,OH 'CH,CH,CHEtOHEt ,CHOH 'CEt,OH Et' 'CHMeCHEtOH 'CH,CH,CHMeOH EtCHMeOH 'CEt MeOH 'CHMeCHMeOH 'CH,CHEtOHn nP Y,OH OH 'CH,CH,CH,CH,CO,H Y OH HOCH,CH,CH,CH,OH 'CH(OH)CH,CH,CH,OH 'CH,CH,CH,OH 'CH(CH,OH)CH,CH,OH 'CMe,CH,OHMe,CHCH,OH 'CH(OH)CHMe 'CHMe,' 'CH,CHMeCH,OH Me,CCH,OH 'CH(OH)CMe -'CH,CMe,CH,OH -PhCH,CH,CH,OH 'CH(0H)CH ,CH ,Ph 'CH,CH,Ph EtOEt 'CHMeOEt Me' 'CH ,CH ,OEt Me,CHOCHMe 'CMe,OCHMe Me' .'CH,CHMeOCHMe 'CH,CH,CH,OCHO C) 0.6) 'CH,OCH,CH,OH 'CH,CH(OH)OMeMeOCH,CH,OH 'CH(OMe)CH,OH 'CH,CHO 'CH,OCH,CH,OMe -MeOCH,CH,OMe 'CH(OMe)CH,OMe HOCH,CH,OH 'CH,CHO " pH 2. * [Et']/[Me'] ca. 7 for [HOOSO,-] 6 x lW3moi dm-,.'Only a trace of this radical detected. J. CHEM.SOC. PERKIN TRANS. 2 1990 detected during the reaction of Ti'"/HOOS03- with 2,2-dimethylpropan- 1-01 (neopentyl alcohol). We believe that secondary and tertiary radicals are more readily oxidized by HOOS03-than methyl and ethyl radicals [as would be expected on the basis of their ionization potentials 2o (Me' 9.34 Et' 8.30 Pr" 7.55 But' 6.93 eV) and as already notedI7 for sz08'-] so that their steady-state concentrations are reduced to or below that required for detection. Consistent with this it was shown that when Et' and Pro were generated by oxidative decarboxylation of the appropriate acids with Ti"'/HOOS03 - signals from the former but more markedly the latter were reduced as [HOOS03-] was increased no signals from But' were obtained from ButCOzH presumably on account of the more rapid oxidation of the tertiary radical. No radicals resulting from fragmentation were observed from some 1 ,2-diols and related a$-dioxygen-substituted compounds (see Table 3) including ethane-1,2-diol 1,4-dioxane 1,2- dimethoxyethane and 2-methoxyethanol.This appears to reflect the effect of a P-oxygen substituent in retarding oxidation of an a-oxygen-substituted radical.,' (ii) Reactions of Ti"'/H202 in the Presence of HOOS03-.-In order to distinguish oxidation of free radicals by peroxymonosulphate from the generation of radicals by SO4-(from HOOS03 -) we utilized the faster rate of reaction of Ti"'- EDTA with H202 than HOOS03- (by a factor of CQ.10) to generate free radicals from 'OH and study their subsequent oxidation in experiments with mixed peroxides (these being in separate streams) with [HzO,] % [HOOS03-1. As Figure 3(b) indicates reaction of 'OH with EtOH gives 'CHMeOH (the major species) and 'CH,CH,OH when [HOOS03 -3 is increased removal of the former and production of 'Me is clearly established. However we also note that the rate of decrease in C'CHMeOH] is faster than the increase in ['Me] which suggests that oxidation to give [Me'] is accompanied by direct oxidation [probably uia reaction (15a) or (15b)l. A full kinetic analysis is described later but we note that in experiments with MeOH (0.6 mol dm-3) and fumaric acid (0.02 mol dm-3) as well as Ti"' and HOOS03- no 'OH adducts of the alkene were observed suggesting that reaction (154 rather than (1 5b) is preferred.(iii) Oxidation of Other Possible Intermediates.-The possi-bilities that the methyl radical (from ethanol) could result from further oxidation of ethanal [formed e.g. uia reactions (15a) or (15b)] was ruled out in rapid-flow experiments in which it was shown that this substrate is unreactive towards HOOS03 -(and Ti"') under these conditions. Similarly propanone (which is believed to be formed by one-electron oxidation of presence of ethanol by the radiation from a 1 kW mercury- xenon lamp also led to the detection of 'CHMeOH 'CH2- CH,OH and Me' (the last of which is not detected in comparable experiments with H,02 and S208,-above). This establishes that metal ions are not required for the unusual oxidation reaction to give alkyl radicals.(v) Other Experiments.-Investigation of the effect of pH on the Ti"*-EDTA/HOOSO3-/EtOH reaction established that relatively low radical concentrations are detected at low pH (possibly on account of the low rate of reaction between HOOS03- and Tinl-EDTA in acid) whereas above pH ca. 10 ['CHMeOH] decreased rapidly accompanied by an increase in [Me']. There are at least two possible explanations for the latter behaviour namely that the ionized counterpart 'CHMeO -[pK ('CHMeOH) is 24 11.61 is more easily oxidized or that the anion -OOS03-[pK (HOOS03-) 9.3 ref. 251 is a more effective oxidant or both. The observation that for 'CHMeOEt (from diethyl ether) there is little decrease in the concentration of the radical or increase in [Me'] above pH 10 suggests that the behaviour of 'CHMeOH is attributable to an increased rate of oxidation of the (unprotonated) ketyl radical.Experiments in which [Ti"'] was increased (21.7 x mol dm-3) led to a reduction in the concentration of alkyl radicals derived by fragmentation (e.g. 'Me from 'CHMeOH). Since alkyl radicals themselves are not effectively reduced by Ti"' under these conditions (but see ref. 26) and as we have also shown alkoxy radicals are readily reduced (with k 6 x 10' dm3 mol-' s-I) we regard this as further support for the proposal that alkoxy radicals are indeed generated in the reactions investigated here.In experiments with Ti" HOOS03- and ethanol addition of acetonitrile to the solvent (33% v/v) led to an increase in ['CHMeOH] and decrease in [Me'] (to ca. 25% of its value in the absence of MeCN). This suggests that the route to the formation of methyl may involve an intermediate in which charge separation is stabilized by solvent polarity or in which hydrogen-bonding confers stability. Alternatively the inter- mediate alkoxy radical may be intercepted by acetonitrile uia a hydrogen-transfer reaction. Finally analysis of the reaction mixture from Ti"'/ HOOS03-/PriOH by GLC indicated that propanone and ethanoic acid are formed in comparable (though low) yields suggesting that the two (competing) oxidation pathways for 'CMe20H [reactions (20) and (21)] are of similar importance under the conditions employed here.+ /-+*CMezoH H+ + Me,CO + HO' + SO4,-(20a)HOOS03-'CMe,OH) failed to react under our conditions; this also rules out the possibility that reactions involved here proceed via a dioxirane intermediate [e.g. (5)] the formation of which has been invoked 22 in the (non-radical) reaction between HOOS03- and ketones. Addition of Fe"' to encourage one- '-+H+ + Me,CO + HO-+ SO4-' (20b) *CMe,OH + HOOSO3-+ SO4,-+ H+ + Me&:& +Me' + MeC0,H (21) (vi) Kinetic Analysis.-In an attempt to determine kinetic parameters both for oxidation of radicals by (formal) one- (5) electron oxidation of 'CHMeOH,3 led to reduction in its concentration and a reduction rather than increase in ['Me] which rules out reaction of +CHMeOH with peroxymono- sulphate as the source of the latter.(iv) Photolysis Experiments.-in situ photolytic decompo- sition of HOOS03- (which gives23 HO' and SO4-') in the electron transfer to HOOS03 -[e.g. reactions (154 and (15b)] and for the effective oxygen-transfer [which results in the production of alkyl radicals via fragmentation e.g. reaction (14)] we have employed both steady-state analysis (as above) as well as computer simulation. Most of our experiments have involved use of the Ti1''-EDTA/H20 couple to generate 'OH and addition of HOOS03 -to study the reaction of first-formed radicals with this substrate. The reactions to be included in the analysis are (22)-(28) reactions (24) and (25) representing the competing modes of oxidation (n.b.the formation of minor J. CHEM. SOC. PERKIN TRANS. 2 1990 0' 1 I0 10 20 1O-3CHOOSO 1 "0 10 20 CR'*l IHOOSO-j1,/10-3 mol dm-3 Figure 4. (a) Variation of k22[Ti1'1],[H,02],[R']-z/dm3mol-' s-' with [HOOSO,-][R']-' in experiments in which HOOSO,- is added to the Ti"'/H,O,/RH system; 0 'CH,OH (from MeOH) 0Me' (from Me,SO) X Et' (from Et,SO). (b)Variation of ['Me][R*],[R'']-' with [HOO- SO -1 [see equation (30)] in experiments in which 'Me was generated by reaction of R" with HOOSO -;R1* x ,'CMe,OH (from Me,CHOH); 0 'CHMeOH (from EtOH); A 'CHMeOEt (from EtOEt). radicals by attack of SO4-' on P-hydrogens is ignored at this stage). Ti"' + H202 -Ti'" + HO-+ 'OH (22) HO' + R'H -R1* + H20 (23) R" + HOOS03-R1+ + HO- SO4-O (24) R" + HOOS03- -R2* (25) R'' R'' R2' + R'' + R2' + R2' __* -1 non-radical products (26) (27) (28) Initial experiments involved kinetic analysis of experiments in which the substrates MeOH Me2S0 and Et2S0 (each of which gives only a single radical R1*) were studied separately signals from 'CH20H Me' and Et' decreased steadily as [HOOS03-] was increased.Appropriate reactions for steady-state analyses for these substrates and under these conditions are equations (22H24) and (26) reaction of Ti"'-EDTA with HOOS03- is negligible (as predicted and also established experimentally). Steady-state analysis for [HO'] and [R"] leads to equation (29) which indicates that if the kinetic analysis is correct then a plot of k22 [Ti111],[H202],/[R"]2 against [HOOS03 -]/[R"] should be linear with a gradient k24 and intercept 2k,.As illustrated in Figure 4(a) the observed behaviour is consistent with that predicted on this basis (in particular the values of 2k which result are ca. 2 x lo9dm3 mol-' s-l which is as expected for small uncharged aliphatic radicals "). and 'Et by HOOS03- (see Table 4) are in accord with expectation based on the relative ionization potentials of these radicals and are also greater than the corresponding rates of the analogous reactions with H202 and S2OS2-.The failure noted above to detect significant concentrations of Pr" and But' in the presence of HOOS0,- is understandable if the rate constants for oxidation of these radicals are bca.1 106dm3mo]-1 s-~ k22[Ti"1],[H202]t = 2k,[Ri']2 + k24[HOOS03-][R1'] (29) For the corresponding reactions of ethanol diethyl ether and propan-2-01 for which there is not only the complication of a small extent of P-radical formation but also as the generation of Me' [via reaction (25)] the above approach was employed (at low [HOOSO,-]) simply to provide estimates for the rate constants for destruction of the first-formed radicals (k24) in the range 2-3 x lo6 dm3 mol-' s-'; these values were refined by computer simulation (see below).Information on the oxygen-transfer reaction [reaction (25)] for the latter substrates was found by a steady-state analysis for [R2'] (i.e.[Me']) this leads to equation (30) in which [R'ITis the total radical concentration (including R1* R2' and any p-radical detected allowance has also been made for termination reactions involving the last-named).Following experiments with Ti'''-EDTA-H2O2 with added HOOS03- and also Ti1",,-HOOS03- in the presence of EtOH Pr'OH and EtOEt it was demonstrated that as expected on the basis of equation (30) [R2'][R']T/[R''] varies linearly with [HOOS03-],* [Figure 4(b)]. Estimates for k25 can be obtained from the gradient (taking k as 2 x lo9 dm3 mol-' s-'); these values were also refined by computer simulation. The resulting rate constants for oxidation of *CH,OH Me' k,[R"][R'] = k25[R"][HOOS03-] (30) * In these experiments [Ti"']o was kept low (<10-3 mol dm-3) to Experimental results for both the Ti"'-EDTA-H,O,-RH- minimize the occurrence of reduction of the intermediate alkoxyl radical HOOS03 -and Ti"',,-HOOS03 -RH systems were satis-by Ti"' (see earlier).factorily simulated with a kinetic program which incorporated J. CHEM. SOC. PERKIN TRANS. 2 1990 Table 4. Rate constants (k/dm3 mol-' s-') for the oxidation of aliphatic radicals by HOOW3- S2OS2-,and H202." Oxidant HOOSO,-S2082-H202 Radical k24 k2 5 k24 k24 Me' 1.2 105 3.3 x 104 Et' 3.8 105 7.4 x 104 'CH,OH 1.8 x lo6 1.3 x 105 3 104 'CHMeOH 2.1 x lo6 1.1 x lo6 6.9 x 105 1.5 105 'CHMeOEt 2.0 x lo6 5.5 x 105 7.5 x 105 5.5 x 104 'CMe,OH 3.0 x lo6 3.4 x lo6 7.1 x 105 5.0 x 105 'CMe,OCHMe 6.0 x lo6 1.9 x lo6 ca. 1.8 x lo6 ca.8.0 x 105 3 104 'Estimated error k k 20% data for S2082- taken from ref. 17 and for Hz02 taken from ref. 2. Calculated on the assumption that the intermediate alkoxyl radical is not reduced by Ti"' (see the text). Values for k25 may be regarded as a lower limit. 6+ 6-HO-OSO; -b L OH I I-H' *OI Me. + HC02H -Me-C-H I OH Scheme 1. 1 the appropriate reactions (22)-(28) (with established rate constants) and other reactions where necessary. For example for Ti"'-EDTA-H202-EtOH system (with added HOOS03 -) we employed k224.5 x lo3dm3 mol-' s-' k('OH + EtOH) 28 1.7 x lo9 dm3 mol-' s-' (for a-attack) and 1.6 x lo8dm3 mol-' (for P-attack) with27 cross-termination k 2 x lo9 dm3 mol-' s-'; allowance was made for oxidation of 'CHMeOH by Ti"-EDTA (k6 x lo8 dm3 mol-' s-l) and the removal of Ti"'- EDTA by H202 (k2 x lo6 dm3 mol-' s-' cf ref.29) as well as the oxidation of Me' by HOOS03- (k 1.2 x lo5 dm3 mol-' s').The best fit to experimental results was obtained with k24 and k25 the rate constants for oxidation of 'CHMeOH by HOOS03-to give MeeHOH and Me' as 2.1 x lo6 and 1.1 x lo6 dm3 mol-' s-' respectively. These values were also incorporated into analogous calculations to simulate the behaviour of a-and P-radicals and Me' from Ti"'-HOOS03- EtOH employing'6 k(S04-* + EtOH) as 1.7 x lo7 dm3 mol-' s-' (a-attack) and 5 x lo5 dm3 mol-' s-' (P-attack) respectively. Again good agreement was obtained [see Figure 3(a)]. Similar analysis for R" and R2'in the reactions of Et20 and Pr'OH led to the values for k24 and k25 shown in Table 4.(c) Summary and Mechanistic Conclusion.-It has been shown that reaction of Ti"' and Fe" complexes with peroxymonosulphate anion generates SO4-* rather than 'OH * We have also considered the possibility that the a-hydroxyalkoxy- radical is the product of reaction of the first-formed oxygen-conjugated radical with SOs-' (from oxidation of HOOSO,- by SO4-' or 'OH). We have dismissed the involvement of SO -* on a kinetic basis (the rate constant for reaction SO4-' with HOOSO,- is reported8 as being less than lo5dm3 mol-' s-I). (295%). The steric inhibition of this reaction by DTPA and the contrast with pulse radiolysis suggests that the overall electron transfer may proceed via an inner-sphere mechanism.Our results also indicate that HOOS03 -is a more powerful one-electron oxidant of radicals than S2OS2-and H202 and as would be expected the rates of reaction increases with decrease in the ionization potential of the appropriate radicals. Although data on E" for the one-electron reduction of HOOS03- are not known these results are certainly not inconsistent with this reaction being an outer-sphere electron-transfer process (governed by Marcus theory).30 We have also established that for certain oxygen-conjugated radicals an alternative and competing mode of oxidation yields alkyl radicals (e.g.Me' from 'CHMeOH) probably uia intermediate alkoxyl radicals. This reaction is best envisaged we believe in terms of a 2-electron oxidation (Scheme 1) involving nucleophilic attack of the radical on the peroxide (cf the well-established heterolytic reaction of amines and sulphides with peroxides to yield N-oxides and sulphoxides respectively ') the resultant a-hydroxyalkoxy radical intermediate would be expected to fragment rapidly as indicated.* A similar oxygen-centred radical has been postulated 32 in the oxidation by S2OS2-of the radical-adduct from 1,3-dimethyluracil though in this case the mechanism proposed involved reaction of a carbonium ion with SO,-' in a solvent cage. It is of course possible that such a reaction mechanism (Scheme 2) accounts for the observations r 1 r 1 L Jcage L JcageI :OHI Me.+ HC02HaMe-C-H I OH Scheme 2.made here and that the occurrence of two reaction pathways [reactions (14) and (IS)] simply reflects the competition (e.g. by +MeCHOH)betweendeprotonationandin-cagetrapping by 'OH. Although our evidence does not point conclusively to either mechanism the absence of similar reactions with either H202or S2082-(even when these oxidations proceed rapidly) the difference in the ratios of the two modes of oxidation with structure (e.g. alcohol-derived radicals us. ether-derived analogues) and the widespread occurrence of related oxidation pathways uia oxygen transfer lends support to the suggestion that Scheme 1 is the more appropriate representation. The observed rate constants (Table 4) could certainly be explained on the basis of Scheme 1 ethers would be expected to react somewhat slower than the corresponding alcohols on the basis of increased steric hindrance whereas the increase in rate on a-alkylation (e.g.'CHMeOH to 'CMe20H is consistent with the expected increase in nucleophilicit y). Experimenta1 ESR spectra were recorded on a Varian E-104 and a Bruker ESP-300 spectrometer each equipped with X-band Klystron and 100 kHz modulation. Hyperfine splittings were measured directly from the field scan [with the ESP-300 by determination with an NMR Gaussmeter ER 035M; with the E-104 this was calibrated with an aqueous solution of Fremy's salt a(N) 1.309 mTj3]; g-values were determined by comparison with that for 'CHMeOH (g 2.003 21 34) obtained from the reaction of 'OH (from T?"-H202) with ethanol.Relative radical concentrations were determined both by spectrum simulation using a program supplied by Dr. M. F. Chiu and by direct double integration (using the ESP-300) absolute radical concentrations were determined by comparison of doubly integrated signals with those from a standard solution of vanadyl sulphate. Flow experiments were conducted by pumping three reagent streams through a mixer which allowed simultaneous mixing ca. 30 ms before passage through the cavity of the spectrometer (using a Watson-Marlowe 502 peristaltic pump). The solutions used were typically as follows stream (i) contained titanium(II1) sulphate (typically [Ti"'] 0.005 mol drn-j) in some cases with added sequestering agent (EDTA DTPA NTA) stream (ii) contained HOOS03 -at concentrations in the range 0.006-0.06 mol dm-3 and stream (iii) contained the substrate (at concen- trations up to 1.5 mol dmb3).The pH was adjusted by addition of sulphuric acid ammonia or sodium hydroxide to streams (i) or (iii) and in certain experiments hydrogen peroxide (0.05 mol drn-j) was added to streams (ii) or (iii). pH Measurements were made using a Pye-Unicam pH meter PW 9410 with the electrode inserted into the effluent stream. All solutions were deoxygenated both before and during use by purging with oxygen-free nitrogen. Photolysis experiments were carried out with the unfiltered radiation from an Hanovia 9778-1 1 kW mercury-xenon compact arc lamp. Solutions containing HOOS03-(ca. 0.03 mol drn-j) and the substrate (ca.1.0 mol dm-3) were slowIy flowed (at a rate of ca. 2 cm3 min-') through the cell during photolysis. A product study was performed uia the addition of titanium(Ir1) chloride solution (0.4 mol dm-3) to an aqueous solution of propan-2-01 and HOOS03-(both 0.05 mol dmw3). Analysis was carried out by flame-ionization GLC via injection of the aqueous reaction mixture onto an 80/100 Carbopack C/O.1% sp-1000 column and identification uia retention times. The kinetic simulation program executed on a DEC VAX computer was kindly provided by Professor D. J. Wad-dington. The source of the peroxymonosulphate anion was the triple salt 2KHSOS.K2SO4-KHSO4 kindly provided by Interox Chemicals Ltd. Other chemicals employed were commercial samples (used as supplied) except for diethyl sulphoxide which was generated in situ as described previously.' AcknowledgementsWe thank Interox Chemicals Ltd. for their support and Dr. M. Jeff and Mr. W. R.Sanderson for helpful discussions. J. CHEM. SOC. PERKIN TRANS. 2 1990 References 1 R. 0.C. Norman Chem. SOC. Rev. 1979,8,1. 2 B. C. Gilbert R. 0.C. Norman and R. C. Sealy,J.Chem. SOC. Perkin Trans.2,1974,824. 3 B. C. Gilbert and M. Jeff 'Free Radicals Chemistry Pathology and Medicine,' eds. C. Rice-Evans and T. Dormandy Richelieu Press London 1988 p. 25. 4 B. C. Gilbert J. K. Stell and M. Jeff J. Chem. SOC.,Perkin Trans. 2 1988,1867. 5 R. K. Samal P. K. Sahoo and S. P. Bhattercharjee J. Mol.Cafal. 1985,33,225. 6 G. Manivannan and P. Maruthamuthu Eur. Polym. J.,1987,23,311. 7 V. A. Lunenok-Burmakina G. G. Lerzina V. B. Emel'yanov and A. G. Miroshnichenko Zh. Fiz. Khim. 1977,51,2831. [Russ. J. Phys. Chem. (Engl. Trans.) 1977,51,1650]. 8 P. Maruthamuthu and P. Neta J.Phys. Chem. 1977,81,937. 9 'Magnetic Properties of Free Radicals,' New Series Group 11 vol. 9 Part b Landolt-Bornstein Springer Verlag Berlin 1977. 10 M. J. Davies and B. C. Gilbert J. Chem. SOC. Perkin Trans. 2 1984 1809. 11 B. C. Gilbert J. K. Stell W. J. Peet and K. J. Radford J. Chem. SOC. Faraday Trans. I 1988,84,3319. 12 M. J. Davies B. C. Gilbert C. B. Thomas and J. Young J. Chem. Soc. Perkin Trans. 2,1985,1199. 13 G. Czapski J. Phys. Chem. 1971,75,2957.14 W. H. Koppenol and J. Butler Adv. Free Rad. Biol. and Med. 1985,1 91. 15 J. Oakes and C. G. van Kralingen J. Chem. SOC. Dalton Trans. 1984 1133. 16 A. B. Ross and P. Neta 'Rate Constants for Reactions of Inorganic Radicals in Aqueous Solution' National Standard Reference Data Systems National Bureau of Standards Washington 1979. 17 M. J. Davies B. C. Gilbert and R. 0. C. Norman J. Chem. SOC. Perkin Trans. 2 1984,503. 18 B. C. Gilbert R. G. G. Holmes and R. 0.C. Norman J. Chem. Res. 1977 (M) 101; (S) 1. 19 B. C. Gilbert C. J. Scarratt C. B. Thomas and J. Young J. Chem. SOC. Perkin Trans. 2 1987,371. I20 'Handbook of Spectroscopy Vol. 1,' ed. J. W. Robinson CRC Press Cleveland 1974; R. W. Kiser 'Introduction to Mass Spectrometry and its Applications,' Prentice-Hall New York 1965.21 See for example data for the reduction of methyl violegen by hydroxyalkyl radicals S. Steenken M. J. Davies and B. C. Gilbert J. Chem. SOC. Perkin Trans. 2 1986 1003. 22 L. Cassidei M. Fiorentino R. Mello 0.Sciacovelli and R. Curci J. Org. Chem. 1986,52,699. 23 P. Kanakaraj and P. Maruthamuthu Int. J. Chem. Kinet. 1983 15 1301. 24 K.-D. Asmus A. Henglein A. Wigger and G. Beck Ber. Bunsenges. Phys. Chem. 1966,70,756. 25 D. F. Evans and M. W. Upton J. Chem. SOC. Dalton Trans. 1985 1151. 26 B. C. Gilbert R. 0.C. Norman and R. C. Sealy,J.Chem. SOC.,Perkin Trans.2,1973,2 1 74. 27 P. Neta Adv. Phys. Org. Chem. 1976,12,224. 28 Farhataziz and A. B. Ross 'Selected Specific Rates of Reactions of Transients from Water in Aqueous Solution. 111. Hydroxyl Radical and Perhydroxy Radical and their Radical Ions,' National Standard Reference Data Systems National Bureau of Standards Washington 1977. 29 B. C. Gilbert R. 0.C. Norman P. S. Williams and J. N. Winter J. Chem. SOC. Perkin Trans. 2,1982,1439. 30 L. Eberson Adv. Phys. Org. Chem. 1982,18,79. 31 (a) R. Curci and J. 0. Edwards 'Organic Peroxides,' vol. 1 ed. D. Swern Wiley-Interscience 1970 pp. 218-240; (b) C. Srinivasan P. Kathalingan and N. Arumagan Can. J. Chem. 1978,56,3043. See also R. Suthakaran P. Subramaniam and C. Srinivasan Proc. Indian Acad. Sci. 1986,97 555. 32 H. P. Schuchmann D. J. Deeble G. Olbrich and C. von Sonntag Int. J.Radiat. Biol. 1987,51,441. 33 R. J. Faber and G. K. Fraenkel J. Chem. Phys. 1967,47,2462. 34 R. Livingston and H. Zeldes J. Chem. Phys. 1966,44 1245. Paper 9/00934E Receiued 3rd March 1989 Accepted 24th April 1989 Correctedproof received 31st May 1990
ISSN:1472-779X
DOI:10.1039/P29900001281
出版商:RSC
年代:1990
数据来源: RSC
|
|