Intramolecular general base catalysis in the hydrolysis of the ester group of benzoylglycolic acid

摘要

1977 1563 Intramolecular General Base Catalysis in the Hydrolysis of the Ester Group of Benzoylglycolic Acid By Antonio Arcelli and Carlo Concilio," lstituto Chimico ' G. Ciamician ', University of Bologna, Via Selmi 2, 40126 Bologna, Italy The mechanism of hydrolysis of the ester group of benzoylglycolic acid in buffered aqueous solutions at 80 "C in the range where the rate is independent of pH was studied. The uptake of lS0in benzoic acid during hydrolysis in 20 le0enriched water, the interpretation of a Brmnsted correlation of hydrolysis catalysed by some oxy-anions (p 0.49). and the evaluation of activation entropy brought us to the conclusion that the reaction involves the anion of the substrate, and may be interpreted in terms of intramolecular general base catalysis (kamp; x H,O 1.33 x s-l, AH1 22.5 kcal mol-l, ASS -22.2 cal mol-1 K-l).INa previous paper 1 we reported the rate of hydrolysis of the ester bond of benzylglycolic acid between pH 1 and 8. The salient feature was the relatively high rate at pH ca. 5 and we thought it interesting to investigate the reason for this occurrence. The rate constant for spontaneous hydrolysis of the substrate, reported in Figure 1, is pH independent at pH ca. 5. pH-Rate profiles of protogenic esters can be generally interpreted in terms of acid, neutral, and basic hydrolysis of the acid (SH) and ionized (S-.)substrate. A detailed interpretation of the various parts of a curve of this type has already been given by Edwards2 for the hydrolysis of aspirin.According to this treatment, the right-hand side of the diagram represents the alkaline hydrolysis of ionized substrate, while the pH-independent t From rcf. 1 log kp=o = 0.067 + 4.80 GI, where a1 is the inductive constant of X.3 region can be essentially identified with the two kinetic- ally indistinguishable processes (1) and (2). kg;YhCOO*CH,*COOH+ OH ____t PhC0,-+ HO*CH,*CO, 4-H (1) kE;*PhC00CH2*C00 + H20 ___t PhC0,-+ HO*CH,*CO, + H ' (2) 'l'lie rate constants derived from experiment are: k$z--3.57 x lo3 1 mol-l s-l and kamp;, xHzO 1.33 x lop6s-l at 80 "C (Table 3). These values can be compared with those which can be reasonably estimated as follows. From a Taft type treatment of alkaline hydrolysis of benzoic acid esters PhCO,*CH,X, when C.Concilio and A. Arcelli, Ann. Chim. (Italy),1968, 58, 881. L. J, Edwards, Trans. Furaduy SOC.,1950, 48,723. M. Charton, J, Org. Chem., 1964, 29, 1222. 1564 J.C.S. Perkin I1 X = C0,H we calculate ktg-28.4 1 mol-l s-l. From of the substrate in the reaction with OH-(Scheme l), the nucleophilicity ratio for ester carbonyl hydrolysis (ii) nucleophilic, or (iii) general base catalysis by the carboxylate group of the ionized substrate in the reaction with water (Scheme 2 or 3). The participation of an ionized or un-ionized carboxy- group has been postulated for mechanisms of hydrolysis of esters. Invariably the largest rate enhancements are observed in systems where the molecule is more or less rigidly fixed in a favourable conf~rmation.~ In this study we investigate the hydrolysis of benzoyl-glycolic acid and identify the reason for rate enhance- ment in a system in which the vicinal carboxylate group is present in a flexible structure. EXPERIMENTAL Materials and Instruments.-Organic and inorganic reagents (C.Erba AR), 20 1800water (D 10)(Bio-Rad), and 99.75 D,O (Merck) were used as purchased. Benzoylglycolic acid was prepared by a known method.s Water was distilled twice from KMnO, before use. pH 345678910 Determinations were performed at 70-90 "C with a Knick PH KpH 350 pH meter using a glass electrode standardized with phosphate, phthalate, and borax buffers.' IsotopicFIGURE1 pH-Rate profile for spontaneous hydrolysis of benzoylglycolic acid at 80 "C and ionic strength p = 1 (data enrichment was measured with an LKB 9000 mass spectro- from Table 1) meter and spectrophotometric determinations were carried out with Optica CF 4 (single beam) and Perkin-Elmer 402 (kOH-/kH,O 1.5 x lo9)* and kamp; (Table 3), we obtain (double beam) instruments. Methods and Results.-pK, Values (Table 1) were deter- kamp;, x H,O 3.4 x lo-*s-l.* mined by potentiometric titration at 80 f0.2 "C of 0.01~ solutions at ionic strength p = 1 (KCl).? The dissociation constants of benzoylglycolic acid at p = 1 are 1.02 x 10-3 (90 "C), 9.38 x (80 "C), and 8.85 x (70 "C).At HO' / OH 80 "C pKH,o = 14.18 and pKH,o+ = -1.73; pK, was taken t from ref. 10, H,O from ref. 11; H30+ was set as -antilog pH ; OH- was calculated from pK, lo without correction for ionic strength; l2 pD was calculated from pH according to Fife and Bruice.13 P The rate of hydrolysis of benzoylglycolic acid was SCHEME measured in buffered solutions with added KCl to keep the 1 H2 L'L SCHEME2 The values derived from experiment are higher than ionic strength p constant at 1.Slow hydrolyses were those expected, suggesting that the rate enhancement is performed in a sealed glass-ampoules kept in a thermo-due to a catalytic effect. This could be (i) intramolecular 6 C. Concilio and A. Bongini, Ann. Chim. (Itdly), 1966,56, 417. general acid catalysis by the un-ionized carboxy-group ' R. G.Bates, J. Res. Nut. Bureau Stanhds A, 1962,66, 179.8 A. Albert and E. P. Serjeant, ' Ionization Constants of * This value, divided by 54, fits the Brarnsted plot rather well Acips,.a~~l~~~~;'~~~end~~,d~~~~g~~;'amp;;:'Physikalisch-Chemische(see below). t The reproducibility was k0.03 pK units, though the accuracy 10 Landolt-Bornstein, Tabellen,' Berlin, 1923, Hw. 11, p. 1164. 'is lower than this owing to the liquid junction potential.9 Handbook of Chemistry and Physics,' ed. R. C. Weast, The W. P. Jencks and J. Carriuolo, J. Amer. Chem. Soc., 1960,82, Chemical Rubber Co., Cleveland, 1971-1972, pp. F4-F5. 1778. l2 Myung-un Choi and E. R. Thornton, J. Amer. Chem. Soc., A. J. Kirby and P. W. Lancaster, J.C.S.Perkin 11,1972, 1974, 96, 1428. 1206. l3 T. H. Fife and T. C. Bruice, J.Phys. Chem., 1961, 65, 1079. l1 1977 1565 statted bath (f0.05 "C). At set times ampoules were Hz180, and acidified with HC1. Benzoic acid was separated cooled, the content diluted five times with 0.lM-phosphate and sublimed in D~CUO. The solution was neutralized with buffer (pH 7.68), and the optical density was determined. NaOH, lyophilysed, and the salts were chromat~graphed.~~ SCHEME3 TABLEI Rate constants for buffer catalysis (kb) and for spontaneous hydrolysis (A,,) of benzoylglycolic acid Buffer No. of lo6k,/(sodium salt) t/"C pK,". pH btc buffer/^ runs 1:$!:!I b,r! b,e Acetate 70 4.73 0.1-0.8 4 1.25 f0.38 0.489 f0.02 Acetate 70 5.76 0.1-0.8 4 1.00 f0.04 0.538 f0.02 Acetate 80 4.67 4.82 0.1-0.8 4 2.69 f0.16 1.35 f0.08 Acetate 80 4.67 5.79 0.1-0.8 4 2.15 f0.25 1.52 f0.02 Acetate in D20f 80 4.70 0.1-0.8 4 1.43 f0.12 0.75 f0.07 Acetate 90 4.90 0.1-0.8 4 5.08 f0.28 3.19 f0.02 Acetate 90 5.82 0.1-0.8 4 4.55 amp; 0.43 3.73 f0.03 Formate 80 3.99 3.54 0.1-0.8 4 7.63 f0.95 1.07 f0.05 Formate g 80 3.99 4.78 0.08-0.7 4 1.64 f0.03 1.26 f0.01 Succinate 80 5.41 5.76 0.14-0.54 4 9.24 f2.41 1.55 f0.08 Phosphate ' 80 6.61 6.60 0.02-0.20 3 53.9 f4.05 2.63 f0.05 Imidazole 80 6.57 6.32 0.08-0.29 4 35.0 f2.3 1.84 f0.07 Sulphite 80 6.90 6.58 0.06-0.20 4 33.1 f4.9 2.53 f0.77 Borate j 80 9.33 8.84-9.18 0.1-0.64 6 838 amp; 24 pK, of conjugate acid h0.03.6 Ionic strength p = 1 (KC1). C f0.02. Free base; the value of pKa found at 80 "C is used for acetate buffer at 70 and 90 "C.Standard errors determined by the weighed least squares method. fpD = pH + 0.17 = 4.87. In 0.025~-acetate carrier buffer. h p = 1.9. iHP042-. j From the equation k,,b,/Buffer = kb + kamp;OH-/Bufferis also calculated kamp;amp; = 0.922, which is in agreement with the value found for phosphate buffer (cf. Table 3). Fast reactions were followed directly in the cell compart- The fraction containing glycolic acid was treated with ment of the spectrophotometer, started by addition of diazomethane to obtain the methyl ester. 50 p 1 of a stock solution of sodium benzoylglycolate in a thermostatted cuvette containing buffer (3ml). The initial concentration of substrate was between 6 x 4p10-2 and 5 x 10-3~;the wavelength observed was at 275 nm for imidazole and sulphite buffers,* and 232 nm for 3 other buffers.The hydrolyses were followed up at least 75 completion; O.D, values were determined either directly after ten half-lives or after dilution with NaOH. The pseudo-first- order rate constant (bobs) for the disappearance of ester was nevaluated from ln(O.D, -O.D,)/(O.D, -O.D,) ve?ws (Dtime, using a least-squares method.? 0The rate constant for buffer catalysed process (kb) and for spontaneous hydrolysis (KO) were obtained from hobsby means of relation (3). The catalytic rate constant Kh for IIIIIIIIII borate buffer was evaluated from the equation (3)rearranged 0.1 0.2 0.3 04 0.5 0.6 0.7 0.8 0.9 in the form: hobs/B = kb + ktamp; x OH-/B setting Basel/M 2KO = x OH.Both KO and Kb were calculated by the FIGURE Plot of the observed rate constants for hydrolysis of benzoylglycolic acid at 80 "C (in buffered solutions at p = 1least-squares method. Results are shown in Figure 2 and with KC1) versus free base concentration : A, formate pH 4.78;3 and collected in Table 1. B, acetate pH 5.79; C, succinate; D, imidazole; E, sulphite;Hydrolysis in H,180.-(a) AZkaZine. Sodium (3 x g F, phosphate (cf. Table 1) atom) was dissolved in H,l80 (2g) in a nitrogen atmosphere, (b) At pH 5.05. A solution of benzoylglycolic acid (1then added to benzoylglycolic acid (1 mmol). The solution mmol), anhydrous sodium acetate (5mmol), and acetic acid was kept for 1 h at 110 "C, then cooled, lyophilysed to recover l4 R.C. Brasted, ' Comprehensive Inorganic Chemistry,' Van * The sulphite buffer did not show any absorption except after Nostrand, New York, 1961, vol. VIII, p. 153. some hours in the air it formed S,0,2-.14 l5 E. F. Phares, E. H. Mosbach, F. W. Denison, jun., and S. F. 7 CERN computer program E 200 (Geneva). Carson, Analyt. Chem., 1952, 24, 660. 1566 J.C.S. Perkin I1 (1 mmol) in H,180 (2 ml) was kept for 200 h at 110 "C, then Brernsted relationship (4)* plotted in Figure 4. The cooled, neutralized with ca. 1ON-NaOH (20 lyophily-sed and treated as above. 1 I I 1 I 123156 lo3OH-/ Basel FIGURE Plot of kol,JIBasc versus OH-/Base in thc hydro- 3 lysis of licnzoylglycolic acid at 80 *(: in bsol;)orate buffered solu-t1ons TABLE2 Pcrcentage 180 uptakc of benzoic acid and methyl glycolate after hydrolysis of benzoylglycolic acid in HzlW at 110 "CR Hydrolysis Control Natural Alkaline pH 5.05 t, runs * samples Benzoic acid d 19.0 19.5 9.1 0.9 Methyl glycolate 1.o 2.1 1.5 1.1 Mean values over at least five runs in the region of molecular peak approximated to *0.2 In acetate buffer.Experi-ment to evaluate uptake after completion of hydrolysis.Direct inlet. G.1.c. inlet. Procedure (b) was also carried out for benzoic and glycolic acid to ascertain that isotopic exchange was not (3 value calculated by the least-squares method excluding OH -and H20 (Table 3) is 0.49 amp; 0.02. This value is OK.OI1 / 0 2 4 6 8 10 12 14 16 pK,+ log !4 4FIGURE Brernsted plot for gcneral base catalysed hydrolysis of benzoylglycolate anion at 80 "C (data from Table 1 ; acetate pH 5.79, formate pH 4.78).The line of slope 0.49 is the best fit through the rate constants for oxy-anions and imidazole (IM)buffers; KEG-and kamp;-, (1.33 x 10-8/54 1 mol-1 s-1) are not computed typical for general base catalysis ; nucleophilic catalysis by oxy-anions having a pK, much smaller than the leaving group t is in fact characterized by P values close to unity.lg TABLE3 Rate constants and thermodynamic parameters for water, hydroxide, and acetate ion catalysed hydrolysis of benzoylglycolic acid at ionic strength p = 1(KCl) SH Qt/"C 107kamp;oHZOIs-' kamp;/l mol-1 s-16 10-3 koH-/l mol-1 s-1 107 ki;o-/l mol-1 s-10 70 4.82 f0.19 0.55 f0.02 2.02 f0.08 0.971 f0.084 80 13.3 f0.05 0.933 f0.02 3.57 amp; 0.13 2.08 f0.30 90 31.8 f0.02 1.54 f0.05 6.10 f0.04 4.47 f0.53 AH*/kcal mol-1 22.5 f0.4 12.1 f0.02 13.0 f0.01 18.2 f0.3 ASt,,/cal mol-1 K-1 -22.2 f1.1 -24.9 f0.06 -5.9 f0.03 -38.0 amp; 0.8 Obtained from the values of k, (cf.Table 1) in acetate buffer averaged over the pH values investigated, after correction at each pH for the contribution of kamp;-OH-/(l + HaO+/Ka); the difference between the twofiguresis at most 5.8. Obtained in 0.1~-phosphate buffer at pH 8.49 setting boba = kamp;-OH-. In 0.025-0.2~-carbonate buffer at 80 "C (pamp; 9.64), kamp;-0.927 1 mo1-ls-l in good agreement with the value found in phosphate and borate buffers. Calculated from the values of kb (cf. Table 1) according to the comments accompanying expression (5).Calculated from the values of the constants at three temperatures with weighed least square method. From AS* = R lnA/(kTe/h); standard state 1~ at 80 "C. due to ester hydrolysis. Benzoic acid and methyl glycolate give substantial molecular peaks in the mass spectra, allowing a careful determinakion of isotopic enrichment (Table 2). DISCUSSION Bufler Catalysed Hydrolysis (kb).-Catalytic rate constants (Kb in Table 1) establish the existence of the * Statistical p and q factors according to Bell and Evans;17 for imidazole IMH+-IM p = 2 and q = 1. t We calculate from refs. 3 and 18 for the leaving group RCH,-OH: ph', 17.1 (R -C0,-) and pK, 13.0 (R -CO,H) at 26 "C The positive deviation of about two orders of magni-tude from the Brernsted line shown by kamp;-, can be attributed to nucleophilic attack by hydroxide ion on the carbonyl carbon, as already reported by other authors.20 l* C.G. Swain, G.-I. Tsuchihaschi, and L. J. Taylor, Analyt. Chem., 1963, 55, 1415. 17 R. P. Bell and P. G. Evans, Proc. Roy. Soc., 1966, A291,297. l8 S. Takahashi, L. A. Cohen, H. K. Miller, and E. G. Peake, J. Org. Chem., 1971,86, 1205. lQW. P. Jencks and M. Gilchrist, J. Amer. CAem.Soc.. 1968, 90, 2622. 2o h7.Pockcr and E. Green, J.Amer. Ch.em.Soc.. 1973, 95, 113 1977 From the data in Tables 1 and 3 the ratio kamp;-/ kLM= 2.7 x lo5 is obtained which lies within the range observed for general base catalysis (105-106) ; lower values (10-lo3) are normally observed for nucleophilic catalysis.21 Similarly, the ratio k~'/k~Po~'-= 0.65 lies in the range observed for general base (0.25-1.9) rather than for nucleophilic catalysis (lo3).The above evidence substantiates general base Catalysis by the buffers. Additional evidence is pro-vided by the results obtained with acetate buffers. (a) Benzoylglycolic acid can participate in hydrolysis both as a free acid (SH) and as an anion (S ) and the buffer is in principle able to catalyse the reaction both in its acidic (AcOH) and anionic (AcO ) forms. There-fore kl,of equation (3) is given by equation (5)where K, is the dissociation constant of benzoylglycolic acid, kt:o-is the catalytic rate constant of the reaction of acetate ion on the acidic substrate.The meaning of other symbols is obvious. One calculates that on going from pH 5.79 to 4.82, at 80 "C, the k",T.6,xterm increases by a factor of 85, both kinetically indistinguishable kiyo-and k;amp; terms increase by a factor 9.2, while the k;i0-term remains essentially unchanged. On the other hand, k,, is found to increase by a factor of only 1.25 (Table 1). It is then clear that the k:;+ term makes the most significant contribution to the overall rate. Considering also that intermolecular general acid catalysis occurs only in a very limited number of hydrolyses,21 it is reasonable to conclude that, at least in the pH range examined, only the first and the second term of equation (5) make significant contributions to kl,.One then calculates kiL0-2.08 x lov7and kiFo-4.03 x 1 mol-l s-l and a contribution of the respective terms to the overall rate at pH 5.79 of 96.8 and 3.2. Extrapolation of kii0-(2.08 x 1 mol-l s-l) to 109 "C gives 1.6 x 1 mol-l s-l which compares well with the acetate catalysed rate constant for methyl benzoate hydrolysis at the same temperature 22 (1.2 x 1 mol-l s-l). (b) The kinetic isotope effect in the presence of acetate buffer (k,Hzo/k,Dyo1.88) is consistent with general base catalysis, even though it does not allow a clear cut distinction from nucleophilic * From k, in acetate buffer at 80 "C and KPoIK:20 2.48,2A H,O,D,O 1," and pKD,O 13.43.29 Terms in hamp;-and ks;-are neglected.21 S. L. Johnson, Ado. Phys. Org. Chem., 1967, 5, 237. 22 M. L. Bender, F. Chloupek, and bl. C. Neveu, J. Amer. Chem. Soc., 1958, 80, 5384. M. L. Bender, E. J. Pollock, and M. C. Neveu, J. Amer. Chem. SOC.,1962, 84, 595. 24 P. M. Laughton, R. E. Robertson in ' Solute Solvent Inter- actions,' eds. J. F. Coetzee and C. D. Ritchie, Dekker, New York, 1969, p. 473. 25 A. R. Fersht and A. J. Kirby, ,I. Amer. ChPm. Soc., 1967, $9, 4857. catalysis.21pB9a (c) The activation entropy has a highly negative value (-38 cal mol-l K-l) close to those observed for other hydrolyges catalysed by acetate acting as general base (-30.7 foia~pirin,~~-31.2 and -39 for aryl acetates 26) whereas less negative values are reported for nucleophile catalysed hydrolysis.26~~~ In conclusion all evidence on the catalytic effect of acetate ion and other bases is consistent with general base catalysis.This agrees well with the general behaviour of reactions of oxy-anions with esters carrying leaving groups of high basi~ity.~l92~ Spontaaeozts Hydrolysis (KO).-As reported in the introduction, the spontaneous process at pH ca. 5 can be identified with the in tramolecularly catalysed reaction either of the acidic substrate with OH- (Scheme 1) or of the ionized substrate with water (Schemes 2 or 3). Taking into account the contribution of alkaline hydro- lysis of the ionized substrate (which however contributes only 1.5y0 to the rate at pH 4.82) KO of equation (3) is given by equation (6).The values of the rate constants ktH,-and kamp; x H,O calculated from ko in acetate buffer and kamp; are reported in Table 3. The following points are indicative in deciding between the mechanisms outlined above. (a) The failure to fit kii0 (1.33 x 1OP6/54= 2.46 x lo-* 1 mol-l s-l) into the Bransted plot (Figure 4) indicates that the hydrolytic reaction cannot be interpreted in terms of simple general base catalysis by water. However, the value interpolated from the plot (3.06 x 1 mol-l s-l) agrees rather well with that expected from the nucleo- philicity ratio (6.30 x 1 mol-l s-l). (b) The kinetic solvent isotope effects calculated from kf2deg;/k,Dzo = 1.80 * are kts/k:amp; 0.52 and kg;o/kg;o 1.76. These values are compatible with either of the mechanisms and they do not permit a choice between them.(c)Hydrolysis in H,1*O gives rise to an uptake of l*O in the benzoic acid moiety only. If the mechanism in Scheme 2 were to operate, an intermediate benzoic glycolic anhydride would be formed, which would be expected to undergo attack by water preferentially at the glycolic carbonyl carbon. This expectation is based on the behaviour of acetic benzoic anhydride (75 of attack occurs on the acetic carbonyl3O) and on the greater susceptibility to nucleophilic attack of a glycolic with respect to an acetic carbonyl carbon *6 D. G. Oakenfull, T. Riley, and V. Gold, Chem. Comm., 1966, 385. 27 A. J. Kirby, in ' Comprehensive Chemical Kinetics,' eds. C.H. Bamford and C. F. H. Tipper, Elsevier, Amsterdam, 1972, vol. X, p. 146. 2* R. P. Bell, ' The Proton in Chemistry,' Cornell University Press, New York, 1959, p. 188. 29 A. K. Covington, R. A. Robinson, and R. G. Bates, J. Phys.Chem., 1966, 70, 3820. 3O M. L. Bender and M. C. Neveu, .I. Amer. Chem. SOC.,1958, 80. 5388. atom.31 The lack of incorporation of I80 into the gly- colate product provides evidence that the anhydride intermediate is not on the reaction path. (d) Valuable information is given by the activation entropies (Table 3). For the reaction of acidic substrate with OH-, the computed AS$ is -5.9 cal mol-l K-l, too high for bi- molecular alkaline hydrolysis. For such a process a more negative entropy is expected, as is observed for hydrolysis of ethyl benzoate (-25) 32 and of benzoyl- glycolate anion (-24.9) (Table 3).The mechanism shown in Scheme 1 appears to be unlikely on these grounds. For reaction with water AS$ -22.2 cal mol-l K-l is calculated. This value is too negative to be compatible with the mechanism of Scheme 2 since for monomolecular J.C.S. Perkin I1 In conclusion the existence of general base catalysis by oxy-anions, the incorporation of l80, and the values of activation entropies are all consistent with general base catalysis for intramolecular catalysis by the carboxylate anion. This process makes the most important contri- bution to hydrolysis of benzoylglycolate anion at pH ca. 5. The effectiveness of intramolecular assistance is indicated by the following.(i) The intramolecular catalytic rate constant at 80 "C is 6.4~times the inter- molecular acetate catalysed rate in spite of the higher basicity of latter. (ii) Interpolation on the Brgnsted plot of the rate constant for benzoylglycolic acid (pK, 3.03) shows that an effective concentration of 24.4~ of an external anion of that basicity would be required processes values close to zero are generally fo~nd.~'~~ to accomplish the same catalytic effect. This evidence confirms the l80experiments. The value of AS;, --22.2 cal mol-l K-l, agrees with the mechanism of Scheme 3 when compared to the value of -38 found for the activation entropy of intermolecular acetate catalysis. A change from an inter- to an intra-molecular reaction corresponds to an increase from 10 to 35 cal mol 1 K 1.30934 A similar trend is observed in the hydrolysis of aspirin where the activation entropy is -22.5 for intramolecular base catalysis and -30.7 cal mol-l K for intermolecular acetate ~atalysis.2~ 31 E. W. Dean, 2.Centrablat, 1913, 11, 347. 32 E. Tommila, Suomen Kern. (B). 1952,25,37. 33 T. St. Pierre and W. P. Jencks, J. Amer. Chem. Soc., 1968, 90, 3817. It may be observed that the rather low value of the effective concentration agrees with the ' loose transition state structure advocated in Scheme 3. Indeed in a ' tight ' transition state intramolecular catalysis provides much higher rate enhancement^.^^^^^ We are grateful to Professor A. Fava for helpful discus- sions and to Professor C. Castellari for the least squares CERN computer program. 6/1976 Received, 25th October, 19761 34 M. I. Page and W. P. Jencks,Proc. Nut. Acad. Sci. U.S.A., 1971, 88, 1678. 36 M. I.Page and W. P. Jencks, J. Amer. Chem. SOC.,1972, 94, 8818. 30 M. J. Page, Ch.ena.SOC. Rpv., 1973, 2, 295