J. CHEM. SOC. PERKIN TRANS. 11 1988 Comparison of the Intrinsic Reactivities of Carbon and Oxygen Nucleophiles at the I ,3,5-Trinitro-substituted Aromatic Ring Jonathan P. L. Cox, Michael R. Crampton," and Paul WightDepartment of Chemistry, Durham University, Durham OH 7 3LE Kinetic and equilibrium data are reported for nucleophilic attack by nitroalkane anions at unsubstituted ring positions of 1,3,5-trinitrobenzene and of 2,4,6-trinitrotoluene. The results allow the calculation for this reaction type of values for intrinsic rate coefficients, k,, of 0.18 for CH,N02- and 0.22 for MeCH NO2-. The corresponding value for the malononitrile anion, CH (CN),, is 2.5 x 104,and for the methoxide ion the value is lo3. The results are discussed in terms of the electronic-structural and solvational reorganisation occurring during reaction. There is current interest in the comparison of nucleophilic Table 1.Equilibrium data for reaction of nitromethane" with sodium reactivity in different reaction types. These comparisons are methoxide in methanol at 25 "C helpfully discussed in terms of the intrinsic reactivities 1-4 of nucleophiles in the Marcus sense. Intrinsic rate constants, ko, 10ZINaOMe]/~ Absorbance, 237 nm may be obtained by extrapolation of log k uersus log K plots to 0 O.OO0 K = 1 (AGO = 0), so that thermodynamic driving forces are 0.94 0.087 removed. The reaction type for which most information is 1.89 0.146 available is proton transfer (protonaton of nu~leophiles).'~~*~ 4.73 0.265 9.50 0.345 GbH/lmol-' Increasing data are becoming available for nucleophilic additions to alkenes '77-9 which complement data on alkene- forming elimination reactions,"-' and on nucleophilic additions to carbonyl compounds.',' Common features thought to affect the intrinsic reactivities of nucleophiles are the structural and solvational reorganisation which occurs during the reactions.' Here we report rate measurements for the reversible attack in methanol of nitromethane anion, CH2N0;, and of nitroethane anion, MeCHNO;, on 1,3,5-trinitrobenzene (TNB) and 2,4,6- trinitrotoluene (TNT) as shown in equation (1).The results X X X = H,Me R', R2= H,Me allow the calculation of the intrinsic rate coefficients for reaction of these nucleophiles at an unsubstituted position in the trinitroaromatic ring.We compare these values with those which we calculate from literature data for similar reactions of the malononitrile anion CH(CN), and the methoxide ion. Experimental 1,3,5-Trinitrobenzene was a recrystallised commercial specimen. 2,4,6-Trinitrotoluene, m.p. 82 "C (lit.,I4 82 "C), was a gift from R.A.R.D.E., Waltham Abbey. Nitromethane, nitroethane, and malononitrile with purities > 99% were obtained from Aldrich Chem. Co. A commercial sample of 2-nitropropane was distilled under reduced pressure using a Fischer Spaltrohr column until analysis by gas chromatography indicated a purity of >99%. Solutions of sodium methoxide were prepared by dissolving clean sodium in AnalaR methanol under nitrogen. Cloudiness in the resulting solutions was removed on centrifugation. Concentrations were determined by titration with standardised acid.The AnalaR methanol used as solvent was de-gassed before use. U.v.-visible spectra were measured with Pye-Unicam SP8-18.9 0.414 49.5 0.473 a Concentration is 5.68 x it5M. Calculated as Al(0.51 Me], where A represents absorbance. 22 21 23 22 23 -A) [NaO-Table 2. Equilibrium data for reaction of nitroethane with sodium methoxide in methanol at 25 OC 102[NaOMe]/~ Absorbance, 232 nm K&/1 mol-' 0 0.00 0.10 0.31 470 0.20 0.47 470 0.40 0.67 560 0.60 0.76 600 1.oo 0.78 410 10.0 0.94 20.0 0.96 a Calculated as Al(0.97 - A)[NaOMe].100 or Beckman Lambda 3 instruments. Kinetic measurements were made using a Hi-Tech SF 3L stopped-flow instrument. Rate coefficients were measured under first-order conditions and quoted values are the mean of at least five determinations and are precise to f5%. Results and Discussion Acidities of Nitroalkanes in Methanol.-Anions were gener- ated from the nitroalkanes in methanol by reaction with sodium methoxide, equation (2). After equilibration, values of KCH R1R2CHN0, + MeO-d% R'R2CNO; (2) (1) (2) were determined spectrophotometrically using the strong U.V. absorption of the nitroalkane anions. Data for nitromethane (1; R', R2 = H) and nitroethane (1; R' = H, R2 = Me) are in Tables 1 and 2 respectively. For 2-nitropropane (1; R', R2 = Me) virtually complete conversion into anion was achieved with a base concentration of 0.01~allowing a lower limit of lo3 1 mol-' to be set for KCH.The use of equation (3) together with the value l5 for methanol of pK, 16.92 allows the calculation of the following pK, values in methanol: (1; R', R2 = H) 15.6, (1; R' = H, R2 = Me) 14.2, (1; R', R2 = Me) < 13.9.The corresponding values in water are 10.22, 8.60, and 7.74 respectively.'6 The increase in acidity in this series has been attrib~ted",'~ to the stabilising influence of methyl substituents attached to an sp2-hybridised carbon atom. Monitoring by U.V. of dilute solutions in methanol of the anions (2) showed them to be stable on the timescale required to examine their reactions with nitroaromatics. Reaction of Nitroalkane Anions with TNB.-There is n.m.r.evidence for the formation in dimethyl sulphoxide of a-adducts of structure (3)from TNB and nitroalkane anion^.'^ Our results obtained in methanol containing small excesses of methoxide ions are interpreted by Scheme 1, which allows for the reaction of TNB with methoxide to produce (4). This latter reaction which is known to be very rapid on the stopped-flow timescale2* and for which KoMehas2' the value 17 1 mol-' was minimised by working with [MeO-] \< 0.01~.The major colour-forming reaction was in each case production of (3) with + R'R~CNO;""0""' + OM^- H CR'R~NO~ =g=&3 02"0N02 k-3 '-..0 NO2 NO2 KOMe (3) No2 (4) Scheme 1.J. CHEM. SOC. PERKIN TRANS. 11 1988 absorption maxima in methanol at 450 and 550nm. Under first- order conditions the rate expression appropriate to this process is equation (4).In some cases a further very slow process was observed, but not measured, and may be attributed to di-adduct formation or to slow ionisation of the added nitroalkane group.22 Data for reaction with nitromethane anion are in Table 3. In this case identical kinetic results were obtained on mixing, in the stopped-flow spectrophotometer, either TNB solution with pre- equilibrated nitromethane and methoxide, or TNB plus nitromethane solution with methoxide. This indicates that for nitromethane the equilibrium of equation (2) is established rapidly compared with the reaction forming (3).The kinetic data yield values for k, 800 1 mol-I s-', and k-, 0.011 s-' whose combination gives a value for K3 (= k,/k-,) of 7 x lo4 1 mol-' in good agreement with that obtained independently from absorbance data. Data for reaction with nitroethane anion, in Table 4, lead to values for k, 34 1 mol-' s-I, k-3 0.09s-', and K, 380 1 mol-'. Here measurements were made by mixing TNB solutions with pre- equilibrated nitroethane plus methoxide. Identical values for kinetic parameters were obtained with nitroethane in an excess of methoxide and with methoxide in an excess of nitroethane. 2-Nitropropane was carefully purified since the equilibrium constant for reaction of its anion with TNB has a low value so that it was necessary to avoid the presence of more reactive impurities.The rate coefficients in Table 5 are independent of the concentration of nitropropane anion showing that here equation (4) is dominated by the reverse rate coefficient and leading to a value for k-, of 0.09 s-'. The absorbance values give a value for K, of 4 1 mol-' leading to k, (= K,*k-,) 0.36 1 mol-' s-'. Reaction of Nitroalkane Anions with TNT.-There is good evidence that the major modes of 1 :1 interaction of TNT with bases are o-adduct formation at the 3-position and transfer of a side chain The possibilities for reaction with nitroalkane anions are shown in Scheme 2. Table 3. Kinetic and equilibrium data for reaction of TNB" with nitromethane anion in methanol at 25 "C 104CMeNo21stoich/M 1o3[NaOMe]&,ich/M 1OS[CHZNO;],d,,/M kobs./S-k,",,,.A(550 nm)J 104K;/l mol-' 10 0.1 0.20 0.013 0.013 - - 1.o 4.0 0.81 0.0 17 0.017 0.0077 6.5 2.0 4.0 1.62 0.022 0.023 0.0112 6.2 3.0 4.0 2.43 0.028 0.029 0.0 140 6.9 4.0 4.0 3.24 0.033 0.035 0.0150 6.4 6.0 4.0 4.86 0.045 0.046 0.0170 6.6 1 .o 7.0 1.33 0.0I9 0.020 0.0107 6.9 2.0 7.0 2.66 0.029 0.030 0.0140 6.3 3.0 7.0 4.00 0.038 0.040 0.0160 6.4 4.0 7.0 5.32 0.047 0.048 0.0170 6.0 6.0 7.0 8.00 0.066 0.067 0.0190 7.2 1.o 10 1.8 0.022 0.023 0.0122 6.7 2.0 10 3.6 0.034 0.035 0.0149 5.6 3.0 10 5.4 0.048 0.048 0.0171 6.1 4.0 10 7.2 0.059 0.060 0.0186 7.0 6.0 10 10.8 0.088 0.085 0.0195 6.5 30 10 50 --0.0223 - Concentration is 1 x M.Ionic strength 0.01~with NaCl. p-Bromophenol buffer. Equilibrium concentration, calculated with Km 22 1 mol-'. Calculated from equation (4) with k, 800 1 mol-' s-', k-, 0.011 s-', and KoMe17 1 mol-'. At completion of reaction forming (3). Measured by stopped-flow spectrophotometry in 2 mm pathlength cell. Calculated as A/(0.0223 -A)[CH,NO;]. J. CHEM. SOC. PERKIN TRANS. 11 1988 Table 4. Kinetic and equilibrium data for reaction of TNB" with nitroethane anion in methanol at 25 "C 1O4[EtNO2Istoicd~ 5.0 103[NaoMelstoich/M 2.5 104[MeCHNO;],b,./~ 2.8 kobs./s-l 0.12 kc'alc. 0.10 4500 nm)d 0.0041 k;/l mol-' 340 5.0 3.8 3.2 0.1 1 0.10 0.005 1 380 5.0 5.0 3.6 0.12 0.10 0.0052 360 5.0 7.5 3.9 0.10 0.10 0.0053 360 10 10 8.3 0.11 0.1 1 20 10 16.7 0.14 0.14 40 10 33.4 0.19 0.19 - - 125 2.5 21 0.16 0.16 0.0 1 98 320 150 5.0 42 0.23 0.23 0.0300 360 175 7.5 62 0.3 1 0.30 0.0343 3 50 200 10 84 0.37 0.38 0.0379 370 a Concentration is 2 x lt5M.Equilibrium concentration calculated with KCH 500 1 mol-'. 'Calculated from equation (4) with k, 34 1 mol-' s-', k-,, 0.09 s-l, and KoMe17 1 mol-'. Absorbance due to nitroethane adduct, measured in 2 mm cell. In some cases an additional initial absorbance due to formation of the methoxide adduct was measured. Calculated as A/[CH,CHNO;] (0.050 -Aslow-Afastdue to methoxide adduct). Table 5. Kinetic and equilibrium data for reaction of TNB" with 2-nitropropane anion in methanol at 25 OC Absorbances at 450 nm CCMe2NOileq./M CNaOMeleq./M kObS./s-' &as, &:ow K,"/I mol-' 0.005 0.005 0.1 1 0.0057 0.001 7 3.7 0.010 0.005 0.09 0.0066 0.0036 4.0 0.020 0.005 0.08 0.0077 0.0070 4.1 0.030 0.005 0.08 0.0077 0.0100 4.1 0.040 0.005 0.09 0.0077 0.0140 4.4 0.020 0.010 0.09 0.012 0.0068 4.2 0.030 0.010 0.09 0.012 0.0101 4.4 0.050 0.010 0.10 0.01 3 0.0137 3.8 " Concentration is 2 x lWfM.* Measured by stopped-flow spectrophotometry, 2 mm cell. 'Absorbance at completion of rapid reaction, due to methoxide adduct. Absorbance due to nitropropane-adduct. Calculated as A,,o,/[CMe2NO;] (0.10 -Aslow -Afast).This calculation uses an extinction coefficient of 2.5 x lo4 1 mol-' cm-' for adduct (3; R', R2 = Me) from ref. 19, 22.Table 6. Kinetic and equilibrium data for reaction of TNT" with nitromethane anion in methanol at 25 "C CMeNO2Istoio,/~ CNaOMelstoicd~ 103[CH,NO;],b,./~ kObS./S-' kc'.lc. A(450 nm) K,d/l mol-I 0.05 0.0025 1.28 0.075 0.075 0.0037 72 0.05 0.0050 2.56 0.082 0.08 1 0.0074 78 0.05 0.0100 5.0 0.098 0.092 0.01 17 72 0.065 0.015 8.4 0.11 0.106 0.01 58 67 0.070 0.020 11 0.12 0.117 0.0187 67 0.10 0.0050 3.4 0.092 0.085 0.0080 65 0.10 0.0106 7.2 0.106 0.109 0.0150 72 0.12 0.02 1 1 15 0.130 0.134 0.0210 61 0.13 0.0317 22 0.163 0.165 0.0250 60 0.14 0.042 30 0.200 0.200 0.0270 54 0.15 0.053 37 0.226 0.229 0.0300 58 Concentration is 1 x lO-' M. Equilibrium concentration calculated with KCH 22 1 mol-'. 'Calculated from equation (5) with k, 4.3 1 mol-' s-' and k-, 0.07 s-'.Calculated as Al(0.044 -A)[CH2NO;]. Table 7. Kinetic and equilibrium data for reaction of TNT" with nitroethane anion in methanol at 25 OC LEt No21stoich/M CNaOMelstoicd~ [MecHNo;],b,./~ kobs./s-' A(450 nm)' Kill mol-' 0.2 0.05 0.05 0.25 0.0028 0.66 0.2 0.10 0.10 0.26 0.0052 0.63 0.25 0.15 0.15 0.30 0.0076 0.63 " Concentration is 2 x M. Equilibrium concentration, calculated with KcH 500 1 mol-'. We assume E (450 nm) is 2.2 x lo4 1 mol-' cm-', as found for the corresponding adduct with nitromethane. Calculated as A/(0.088 -A) [MeCHNOJ. Our results show that the initial reactions with nitromethane and nitroethane anions produce species with h,,,. 450 and 540 nm as expected for the adducts (5).There was no evidence in the initial stages for the production of (6) which has a distinctive visible spectrum23with h,,,.at 370, 525, and 650 nm. Nor was competing a-adduct formation by reaction with methoxide ions a problem since the equilibrium constant in methanol for this process has the low value 23 of 0.07 1mol-'. Kinetic results for the initial colour-forming reaction are in Tables 6 and 7 and are interpreted by equation (5). The results 28 J. CHEM. SOC. PERKIN TRANS. I1 1988 Table 8. Comparison of equilibrium and kinetic data for reactions in methanol at 25 "C Reaction with TNB Reaction with TNT A hI 7 r 7 Anion pK," K,/l mol-' k,/l mol-' s-l k-3is-' K,/l mol-I k3/l mol-' s-' k-3ls-l ko CHNO; 15.6 7 x 104 800 0.01 1 62 4.3 0.07 0.18 MeCHNO; 14.2 3 80 34 0.09 0.63 0.16 0.25 0.22 Me,CNO; 13.5 4 0.36 0.09 CH(CN); 14.1 40 3 x los 6.5 x 103 0.05 3.5 x 103 7 x 104 2.5 x 104 MeO-16.9 17d 7 050 305 0.07 280 3000 103 a Standard state is methanol.Estimated from value in water. Values for reaction with TNB from reference 25. Values for reaction with TNT were obtained by stopped-flow spectrophotometry using the methods of reference 25. Values from reference 20. Values from reference 23. 6-5-4-2 3-m -2--0 t b 1-am L1 I I I I I I 01 -4 -3 -2 -1 0 1 2 32 0-log(equi1ibriurn constant 1 -1 -Figure 2. Logarithmic plots of rate coefficient uersus equilibrium constant for attack of methoxide in methanol at unsubstituted ring -positions of activated aromatics at 25 "C.Open circles are for forward -2 reactions and filled circles for reverse reactions. 1.4-Fluoro- l-methoxy- 2,6-dinitrobenzene (ref. 30). 2. 4-Chloro-1-methoxy-2,6-dinitrobenzene (ref. 30). 3. l-Methoxy-2,6-dinitro-4-trifluoromethylbenzene(ref. 30). -2 -1 0 +1 +2 +3 '4 +5 4. l-Trifluoromethyl-3,5-dinitrobenzene(ref. 31). 5. 2,4,6-Trinit-rotoluene (ref. 23). 6. l-Methoxy-4-methylsulphonyl-2,6-dinitroben-log K3 zene (ref. 30). 7. 2,4-Dinitronaphthalene (ref. 32). 8. 4-Cyano-l- methoxy-2,6-dinitrobenzene(ref. 33). 9. l-Chloro-2,4,6-trinitrobenzeneFigure 1. Reactions of nucleophiles at unsubstituted ring positions of (ref. 34). 10. 2,4,6-Trinitroanisole (ref. 35). 11. l-Methoxy-3,5-dinit- TNB (circles) and of TNT (squares).Open symbols refer to k, and ropyridine (ref. 36). 12. 2,4,5-Trinitronaphthalene (ref. 32). 13. N-(n- filled symbols to k-, Butyl)-2,4,6-trini troaniline (ref. 37). 14. 1,3,5-Trinitrobenzene (ref. 20). 15. 2,4,6-Trinitroaniline (ref. 37). 16. 3,5-Dinitropyridine (ref. 38). 17. 1- Methoxy-2,4-dinitro-6-trifluoromethylsulphonylbenzene(ref. 27). 18. 2,4,7-Trinitronaphthalene(ref. 32). Values in methanol have in some cases been extrapolated from literature data in methanol-dimethyl sulphoxide mixtures anion are dominated by the reverse rate constant, giving k-, 0.25 s-'. Knowledge of the value of K3 from equilibrium measurements yields k, (= K,-k-,) 0.16 s-'. Slower reactions resulting in increased absorbance at 500 nm were observed, but not measured, and may be attributed to equilibration with (6)and/or di-adduct f~rrnation.~~-~~ Comparisons.-Rate and equilibrium data for the nitroalkane anions are collected in Table 8 where they are compared with data for the malononitrile anion and for the methoxide ion.Values of K3 for reaction with TNB are ca. a thousand times Scheme 2. higher than corresponding values for reaction with TNT reflecting the unfavourable electronic and steric effects of the methyl ring-substituent on o-adduct f~rmation.~, These values of K,, giving a measure of the carbon basicities26 (thermodynamic affinity for carbon) of the nucleophiles, for nitromethane anion yield k, 4.3 1 mol-' s-l, k-, 0.07 s-', and decrease in the order nitromethane > nitroethane > 2-nitro-K, 62 1 mol-'.The kinetics for reaction with the nitroethane propane and this is the same order as observed for proton J. CHEM. SOC. PERKIN TRANS. 11 1988 basicities, as measured by pK, values. However, the methoxide ion which has the highest proton basicity of the nucleophiles in Table 8 has relatively low carbon basicity. This is consistent with other observations 13*25*26showing that carbon bases have considerably higher carbon basicities than oxygen bases of similar proton basicity. The logarithmic plots of rate coefficient against equilibrium constant shown in Figure 1 allow the determination of intrinsic rate coefficients, k,, for reaction at an unsubstituted ring position in the activated aromatic ring.They show the relatively high reactivity of the malononitrile anion and the low reactivity of the nitroalkane anions. Data for methoxide are limited in Figure 1to reaction with TNB and TNT. However, an extended plot, Figure 2, of literature 22*27 data for reaction of methoxide in methanol at unsubstituted ring positions of ring-activated aromatics gave reasonably good straight lines with a value of k, of ca. 300. These results are profitably discussed in terms of the electronic-structural reorganisation accompanying reaction and the solvent reorganisation during reaction.'^^,^*^*^^*^^ In methanol the nitroalkane anions may be represented by structure (7) in which the negative charge is largely on the oxygen atoms and is strongly solvated. Nucleophilic reaction via the carbon atom, as in equation (l),necessitates movement of charge resulting in considerable structural-electronic reorganisation with accompanying solvent reorganisation.Hence the low reactivity. On the other hand there is evidence that in nitrile anions the negative charge is not strongly delocalised so that malononitrile anion will have more of the character of a true carbanion requiring less reorganisation during reaction and higher reactivity. The ratio of lo5observed here for the two types of anion reacting in methanol can be compared with ratios for reactions in water of ca. lo8 for protonation and ca. lo5for addition at carbonyl carbon.‘ The methoxide ion will be strongly solvated in methanol and the intrinsic reactivity (Table 8) is considerably lower than that for the malononitrile anion reflecting the need for desolvation during bond formation at the aromatic ring.There is a marked difference here with reactivity of the methoxide ion towards protonation where desolvation is not necessary and k, approaches 1O’O. References 1 C. F. Bernasconi, Pure Appl. Chem., 1982, 54, 2335. 2 R. A. Marcus, J. Phys. Chem., 1968, 72,891; A. 0.Cohen and R. A. Marcus, ibid., p. 4249. 3 J. Hine, Adv. Phys. Org. Chem., 1977, 15, 1. 4 W. J. Albery, Annu. Rev. Phys. Chem., 1980, 31, 227. 5 C. F. Bernasconi, Tetrahedron, 1985,41, 3219. 6 F. Terrier, J. Lelievre, A.-P. Chatrousse, and P. Farrell, J. Chem. SOC.,Perkin Trans. 2, 1985, 1479. 7 C. F.Bernasconi and S. A. Hibdon, J. Am. Chem. Soc., 1983, 105, 4343. 8 C. F. Bernasconi, C. I. Murray, J. P. Fox, and D. J. Carre, J. Am. Chem. SOC., 1983, 105, 4349. 9 C. F. Bernasconi, A. Laibelman, and J. L. Zitomer, J. Am. Chem. Soc., 1985, 107, 6563, 6570. 10 C. J. M. Stirling, Ace. Chem. Res., 1979, 12, 198. 11 C. J. M. Stirling and M. Varma, J. Chem. Soc., Chem. Commun., 1981, 553. 12 D. B. Boyd, J. Org. Chem., 1985,50, 885. 13 C. F. Bernasconi, K. A. Howard, and A. Kanavarioti, J. Am. Chem. Soc., 1984, 106,6827. 14 A. McGookin, S. R. Swift, and E. Tittensor, J. SOC.Chem. Ind., London Trans., 1940, 59,92. 15 C. H. Rochester and B. Rossall, J. Chem. Soc., Perkin Trans. 2,1967, 743. 16 D. Turnbull and S. Maron, J. Am. Chem.SOC.,1943,65, 212. 17 F. G. Bordwell, J. E. Bartmess, and J. A. Hautala, J. Org. Chem., 1978,43,3095; F. G. Bordwell and J. E. Bartmess, ibid., p. 3101; F. G. Bordwell, J. E. Bartmess, and J. A. Hautala, ibid., p. 3107. 18 0. A. Reutov, I. P. Beletskaya, and K. P. Butin, ‘CH Acids,’ Pergamon Press, Oxford, 1978. 19 C. A. Fyfe, Can. J. Chem., 1968, 46, 3047. 20 C. F. Bernasconi, J. Am. Chem. SOC., 1970, 92,4682. 21 M. R. Crampton and H. A. Khan, J. Chem. SOC.,Perkin Trans. 2, 1973, 710. 22 E. Buncel, M. R. Crampton, M. J. Strauss, and F. Terrier, ‘Electron Deficient Aromatic- and Heteroaromatic-Base Interactions,’ Elsevier, Amsterdam, 1984. 23 D. N. Brooke and M. R. Crampton, J. Chem. Res., 1980, (3340 (M) 4401. 24 C. A. Fyfe, C. D. Malkiewich, S. W. H. Damji, and A. R. Norris, J. Am. Chem. SOC., 1976, 98, 6983. 25 M. R. Crampton, T. P. Kee, and J. R. Wilcock, Can. J. Chem., 1986, 64, 1714. 26 J. Hine and R. D. Weimar, J. Am. Chem. Soc., 1965,87, 3387. 27 F. Terrier, Chem. Rev., 1982, 82, 77. 28 W. P. Jencks, Chem. Rev., 1985, 85, 51 1. 29 M. J. S. Dewar and D. M. Storch, J. Chem. Soc., Chem. Commun., 1985, 94. 30 F. Terrier, F. Millot, and J. Morel, J. Org. Chem., 1976, 41, 3892. 31 F. Millot and F. Terrier, Bull. SOC.Chim. Fr., 1974, 1823. 32 W. L. Hinze, L. J. Liu, and J. H. Fendler, J. Chem. Soc., Perkin Truns. 2, 1975, 1751. 33 F. Terrier, C. A. Dearing, and R. Schaal, ‘Reaction Transition States,’ Gordon & Breach, London, 1973, 137. 34 L. H. Gan and A. R. Norris, Can. J. Chem., 1974,52, 18. 35 C. F. Bernasconi, J. Am. Chem. SOC.,1971, 93, 6975. 36 F. Terrier, A. P. Chatrousse, and R. Schaal, J. Org. Chem., 1972, 37, 3010. 37 M. R. Crampton and B. Gibson, J. Chem. SOC., Perkin Trans. 2,1980, 752. 38 F. Terrier and A. P. Chatrousse, Bull. Soc. Chim. Fr., 1972, 4549. Received 1st December 1986; Paper 612303