Acid-catalysed hydration of 3-substituted nortricyclanes

摘要

J. CHEM. SOC. PERKIN TRANS. 11 1986 Acid -cat a Iysed Hydration of 3-Substituted Nortricyclanes Martti LajunenDepartment of Chemistry and Biochemistry, University of Turku, SF-20500 Turku 50, Finland Disappearance rates of several 3-X-substituted nortricyclanes X = H, CH,OH, CH,CI, Ac, OH, (CHJOH, OAc, CN, NO,, and 0x01 were measured in aqueous perchloric acid by g.1.c. According to activation parameters, solvent deuterium isotope effects, and log k, versus H, and log k, versus olq correlations, the hydration mechanism is in most cases AS,2 (or AdE2), i.e., protonation of the cyclopropane ring is the rate-determining stage of the reaction. The mechanism is different in the case of 3-methyl-3-hydroxynortricyclane (A-1) and 3-acetoxy- and 3-0x0-nortricyclane (A-2).The slope of the linear log k, versus olq correlation (-1.30 in 1 mol dm-3 HCIO, at 348.2 K) is between those measured for separate protonations of two olefinic carbons of 5-X-su bstituted norbornenes. The contradiction between the normal inductive effects of the 3-X-substituents and the weak effects of the methyl groups at the cyclopropane carbons can be rationalized by formation of an edge-protonated nortricyclane in the rate-determining stage of the reaction. The electrophilic cleavage of cyclopropane, generally initiated dominates ( 2SO, when X = C1 or Br) ' and produces a 7-X-by a proton, has been extensively studied.'q2 Probable substituted 2-norbornyl cation (2). This cation could rearrange intermediates are corner-and/or edge-protonated cyclo- (Wagner-Meerwein) to a 3-X-substituted cation (3), but the propanes whose calculated energies are roughly ~irnilar.~,~ energy of the latter is high due to the location of the electron- The mechanisms of protonation are not changed in the case of symmetrically alkyl-substituted cycl~propanes.~ One such compound is nortricyclane (tricycloC2.2.1 .02v6heptane), the investigation of which is also of interest owing to the formation of a 2-norbornyl cation, whose character (classical or nonclassical) has long been the object of contr~versy.~*~ The proton (deuteron) which initiated the reaction is found mainly at the exo- and endo-6-positions of the product, an exo-2- norbornyl derivative, roughly in 1:1 ratio.'-'' An electron-withdrawing substituent at the 3-position of nortricyclane (l), however, changes the situation markedly (Scheme l)." In this case the rupture of the C(l)-C(6) bond X X .b:+ (2) H withdrawing substituent adjacent to the positive charge. The reaction-initiating proton (deuteron) is now situated mainly at the endo-6 position (290) of product (4), which strongly supports the initial edge protonation of 3-X-substituted nortricyclanes.' ' Characteristic of the electrophilic cleavage of cyclopropane and nortricyclane is a slight rate-increasing effect of methyl (or ethyl) groups at the cyclopropane carbons Chart; a Cl, Br, or OH group at C(3) of nortricyclane does not change the reaction me~hanisrn.~.''-13 This effect can be compared with that of a methyl group at an olefinic carbon of norbornene or bicyclo2.2.1 heptene (5) since this bicyclic substrate also X X (minor ) Scheme 1.1552 J. CHEM. SOC. PERKIN TRANS. 11 1986 krel 1 4 5-6 17 Chart. (7) Scheme 2. produces the 2-norbornyl cation (via the rate-determining protonation ofthe double bond; the A-SE2 or AdE2mechanism) and solely exo-2-norbornyl products (Scheme 2; X = H).6,'3 A methyl group at an olefinic carbon increases the hydration rate by a factor of lo4.', The slight accelerating effects of methyl groups at the cyclopropane ring as well as a poor correlation between the protonation rates and initial strain energies of cyclopropanes have been explained by a slight development of the positive charge at the three-carbon ring and an early transition state.2 On the other hand, in the hydration (protonation) of alkenes the transition state is probably late, i.e., carbocation-like,'5,16 thus a hyperconjugative effect of the methyl group is significant.If this transition state hypothesis is correct, the effects of substituents at other parts of a cyclopropane compound than at the three-carbon ring should also be slight. This gives a goal for the present work, which deals with the effects of 3-substituents on the rate and mechanism of disappearance of nortricyclane in aqueous acid with 5-substituted 2-norbornenes l7 as reference compounds. Results and Discussion Disappearance rates of ten 3-X-substituted nortricyclanes were measured in perchloric acid under different conditions by g.1.c.The rate constants in 1 mol dm-j HClO, at 348.2 K as well as activation parameters, solvent deuterium isotope effects, and slopes for linear log k, uersus H, correlations were calculated. The results are collected in Table 1 together with the inductive substituent constants (o?)of X.' * When X = H, CH20H, CH2Cl, Ac, OH, CN, or NO,, the activation entropies, isotope effects (cf: k,/k, = 1.56 in the hydration of cycl~propane),'~ and slopes for log k, versus H,, are all typical of an A-SE2 or AdE2 mechanism, in which proton transfer from an oxonium ion to a carbon atom of the three- membered ring is the rate-determining stage of rea~tion.'~,~'In the case when X = OH, no general acid catalysis was detected,21 but the results of rate measurements in acidic H20- D20 mixtures were in agreement with the slow proton transfer to a carbon atom.' 2a*' Thus the mechanism does not change in the case of the above mentioned substituents although the hydration rate varies by almost five powers of ten.According to the kinetic parameters (Table 1) an acetoxy or 0x0 group or an a-hydroxy a-methyl function at the 3-position of nortricyclane causes a change of mechanism: protonation now occurs at an oxygen atom of the substituent in a fast pre- equilibrium followed by a rate-limiting bimolecular (OAc* and 0x0) or monomolecular ((",) ~tage."~,~~.~~Paasivirta 24 has also proposed a similar (A-2) mechanism for the reaction of 1-methyl-3-nortricyclanone in sulphuric acid-acetic acid.The results of rate measurements in acidic H20-D,O mixtures are also in agreement with the unimolecular (A-1) mechanism in the case of 3-methylnortricyclan-3-01.~~ Before examination of the effects of the 3-X-substituents on the hydration rates of nortricyclanes, possible protonation of a substituent must be considered. If X is protonated its ability to withdraw electrons increases and thus electrophilic attack on the cyclopropane ring is retarded. Comparison of the basicities The reaction is a normal ester hydrolysis (AAc2),which yields the corresponding alcohol (X = OH).23 J. CHEM. SOC. PERKIN TRANS. II 1986 1553 Table 1. Disappearance rate constants of 3-X-substituted nortricyclanes in 1.00 mol dm-3 aqueous perchloric acid at 348.2 K as well as activation parameters, solvent deuterium isotope effects and slopes for log k, versus H, at the same temperature if not otherwise noted.0,''= inductive su bst it uen t constant Slope for X GIq k,/s-' AH '/kJ mol-' AS'1.l mol-' K-' k,/k, log k, versus H, Ref. H 0 0.1 17 82 -29 -1.17" 13 CH,OH 0.66 8.09 x 92 -21 1.54b This work CH2CI 1.02 3.90 x lW3 90 -34 1.42 20e Ac 1.69 6.31 x 1W' 104 -9 1.35 -1.19' 206 OH 1.74 8.64 x 104 97 -27 1.49 20a E 1.85 4.00 x 10-3 113 + 33 0.48 20a OAc 2.12 2.18 x lW3 63 -113' 0.68 curved This work, 23 CN 3.04 1.23 x lO-' 97 -32 1.38 -1.1 1 206 NO2 3.52 2.05 x 1C6 99 -41 1.22 -1.04, 20c Ox0 3.66g 1.86 x 1W6 105 -41 0.62 -0.65 22 358.2 K.Calculated by means of a linear correlation between the 298.2 K. * 328.2 K. 338.2 K. 318.2 K.'Measured in 60 di~xane-water.~~ substituent constants of Siege1 and Komarmy and the qqvalues.18" lorf X Scheme 3. of X-substituted alkanes and cycloalkanes, however, suggests that the portion of X-protonated substrates is insignificant ( 5 1 in 1 mol dm-3 HC10,).26 The logarithms of the hydration rate constants (at 348.2 K) of the 3-X-substituted nortrkyclanes which react by the A-SE2 mechanism correlate with inductive substituent constants olq * equation (1); r = -0.996; slope = -1.40 and r = -0.994 at 298.2 K; the correlations are almost as good when employing cI values 27. The points for 3-acetoxy-and 3-methyl-3-hydroxy- nortricyclanes deviate about one logarithmic unit upwards from the regression line due to their different mechanisms of reaction, but that for 3-oxonortricyclane is on the line although its mechanism is also different (see above).An explanation for the latter phenomenon is possibly conjugative interaction between the 0x0 group and the three-membered ring2' which reduces the energy of the initial state and retards the protonation of the cyclopropane ring. (A corresponding retardation was observed in the protonation of norbornen-Zone, in which homo-conjugation between the 0x0 group and the carbon-carbon double bond exist~.'~) However, the rate for another mechanism, i.e., protonation of the 0x0 group, is by accident such that the point does not deviate from the regression line calculated for A-SE2 hydration. Now we can consider whether the effects of the substituents are normal or exceptional.A small absolute value of the slope of equation (1) (reaction constant pi) would be in agreement with an early or otherwise slightly charged transition state. Let us take the inductive reaction constants of the A-SE2 hydration(protonation) of the double bond of 5-X-substituted norbor- nenes (5) as reference values (Scheme 2; pi is practically independent of the position, em or endo, of 5-X)." The reaction constants have been determined separately for the protonation of the two olefinic carbons of 5-X-norbornenes: pI = -1.56for C(2) protonation and -0.92 for C(3) protonation under similar conditions as for nortricyclanes.' As we can see, the reaction constant of hydration of the nortricyclanes (-1.30at 348.2 K) is between those measured for the norbornenes and slightly closer to that for C(2) protonation. Let us consider the norbornyl cations formed via rupture of different carbon-carbon bonds of the cyclopropane ring of the 3- X-nortricyclanes (Scheme 3).Cation (6)formed via cleavage of the C(l)-c(2) or C(2)-C(6) bond is similar to that formed uia C(3) protonation of 5-X-norbornene (S); Scheme 21. The formationof another possible cation (9) via cleavage of the same bonds is not probable due to generation of the positive charge 1554 J. CHEM. SOC. PERKIN TRANS. 11 1986 p. f.s.w+-H+ -r.d.s. x Scheme 4. Table 2.Tentative assignments of ,C n.m.r. chemical shifts for 3-X-substituted nortricyclanes in CDCI, (tetramethylsilane as internal reference). The position of X is cis to C(5) X C(1)" C(2)" C(3) C(4) C(5)b C(6)" C(7)b X H 9.7 9.7 33.0 29.6 33.0 9.7 33.0 CH, CH,OH CH,CI COCH, CO,H CN 11.8 11.2 12.0 11.5 11.4 10.7 16.4 12.3 14.0 12.3 12.7 14.2 39.0 47.8 45.2 57.9 49.5 33.9 34.8 30.9 32.1 29.1 33.5 33.9 28.8 29.0 28.9 33.1 34.5 31.1 9.3 9.6 9.7 10.1 10.3 9.7 34.4 34.1 34.0 34.7 30.7 32.9 14.5 62.8 48.1 209.8 and 30.3 180.8 119.0 OHFi,OCOCH, Ox0 13.5 13.2 12.9 19.2 16.2 21.8 13.8 17.3 77.4 82.0 79.9 214.1 35.6 40.5 33.3 37.7 29.4 31.8 30.2 31.6 10.6 12.5 11.2 19.2 30.7 31.8 30.5 31.6 21.8 170.9and 2 1.2 CI 13.7 17.7 65.2 37.0 30.1 11.1 31.6 NO2 13.1 11.5 89.8 35.3 32.4 13.4 30.3 "Signals are possibly to be interchanged.Signals are possibly to be interchanged. adjacent to the electron-withdrawing substituent.' Cation (8) formed oia rupture of the C(l)-C(6) bond is not similar to that (7)formed via C(2) protonation of 5-X-norbornene, but the substituent is, however, at the y position from the positively charged carbon atom in both of them. Thus, the reaction constants of hydration of nortricyclanes is in accord with those measured for the norbornenes. According to Werstiuk and his co-workers,' 'major products are formed via the initial cation (8) and minor products via cation (6)when addition of acetic acid occurs to 3-chloro-or 3-bromo-nortricyclane under catalysis by sulphuric acid.This product analysis is thus also in agreement with the reaction constant measured in this work. What is the explanation for the contradiction between the slight accelerating effect of a methyl group at a cyclopropane carbon and the normal retarding effect of an electron-withdrawing substituent at C(3) upon the hydration (proton- ation) rate of nortricyclane? Evidently it cannot be a slight development of the positive charge in an (early) transition state since the absolute value of the reaction constant should be small in this case. Besides, measurements of hydration rates in acidic H20-D,O mixtures have given results according to which the degree of proton transfer at the transition state (0 5 a 5 1) is very similar in the hydrations of nortricyclanes (a = 0.78 f0.07) and norbornenes (a= 0.75 amp; 0.05) and refers to a late transition state.16 Neither can it be the poor ability of the cyclopropane ring to transmit electronic effects.29 Perhaps the explanation is an existance of different intermediates (lo)and (1I) in the rate-determining (r.d.s.) and product-forming (p.f.s) stages (Scheme 4) as Depuy has proposed for the electrophilic cleavage of substituted cyclopr~panes.~If the former intermediates (10) are the edge- protonated cyclopropanes, the positive charge, although largely developed in the transition state, is not localized at one carbon atom having a methyl substituent as in (11) but delocalized in a two-carbon one-hydrogen system (two-electron three-centre bond).The methyl group has no marked hyperconjugative effect in this case, but the inductive effect of the electron- withdrawing substituent works normally since there is a positive charge at the three-membered carbon ring. The good linearity of the equation suggests that protonation occurs at one edge of the ring in the case of the 3-substituted nortricyclanes and the slope (-1.30) suggests that the edge is the C(lkC(6) bond (Scheme 4), since the positive charge is delocalized between two carbon atoms at the y-positions and one hydrogen atom at the amp;position from the substituent. J. CHEM. SOC. PERKIN TRANS. 11 1986 Experimental Syntheses.-The preparations of 3-X-substituted nortri-cyclanes have been reported.rsquo; 3*20*22 Their purities (by g.1.c.) were mostly 99 or better. In the cases when X = CH,Cl, CN, or NO, the purities were ca.95. The retention times of impurities were such that they did not disturb kinetic measurements by g.1.c. The 13C n.m.r. spectra of the substrates were recorded on a JEOL FX 60 or JEOL JNM-GX-400 spectrometer in CDCl, with tetramethylsilane as internal standard. The chemical shifts are listed in Table 2. They are in the case when X = H, OH, 0x0, or (gv in agreement with published data. Kinetics.-Disappearance rates of the substrates (initial concentration 4 x lo-rsquo; to 3 x lW3 mol drn-,) in thermostatted aqueous perchloric acid were followed by taking samples after appropriate intervals during ca.2.5 half-lives, by neutralizing them with ammonia (and phosphate buffer if necessary) and by analysing them on a Perkin-Elmer F 11 gas chromatograph (FFAP, Carbowax 20 M, or XE 60 packed columns). An inert ketone, mostly norcamphor, was used as internal standard. The peak areas were measured on a Hewlett-Packard 3380s integrator. The fair first-order kinetics was generally observed with standard errors of 1-3, although poor solubility of some substrates (X = H or CH,Cl) decreased accuracy of analyses. In these cases the analyses and runs were repeated several times. Acknowledgements I am grateful to Mr. J. Tallgren for assistance in the kinetic measurements and to Mrs. M. Samppala and Mr. J. Mattinen for recording the n.m.r.spectra. References 1 E.g., reviews: C. H. DePuy, Acc. Chem. Res., 1968,1,33;C. J. Collins, Chem. Rev., 1969,69,543;C. C. Lee, Prog. Phys. Org. Chem., 1970,7, 129;J. L. Fry and G. J. Karabatsas, in lsquo;Carbonium Ions,rsquo; eds. G. A. Olah and P. v. R. Schleyer, Wiley, New York, 1970,vol. 2,ch. 4; M. Saunders, P. Vogel, E. L. Hagen, and R. Rosenfeld, Acc. Chem. Res., 1973,6, 53; C.H. DePuy, Fortschr. Chem. Forsch., 1973,40, 73. 2 (a)K. B. Wiberg and S. R. Kass, J. Am. Chem. Soc., 1985,107,988;(6) K. B. Wiberg, S. R. Kass, and K. C. Bishop 111, ibid., p. 996;(c) K. B. Wiberg, S. R. Kass, A. de Meijere, and K. C. Bishop 111, ibid., p. 1003. 3 R. Sustmann, J. E. Williams, M. J. S. Dewar, L. C. Aen, and P. v. R. Schleyer, J.Am. Chem. Soc., 1969,91,5350K. Raghavachari, R. A. Whiteside, J. A. Pople, and P. v. R. Schleyer, ibid., 1981, 103, 5649. 4 C. H. DePuy, P. C. Funfschilling, A. H. Andrist, and J. M. Olson, J. Am. Chem. SOC.,1977,99,6297. 5 G.A. Olah, A. M. White, J. R. DeMember, A. Commeyras, and C. Y. Lui, J. Am. Chem. Soc., 1970, 92, 4627. 6 G. D. Sargent, in lsquo;Carbonium Ions,rsquo; eds. G. A. Olah and P. v. R. Schleyer, Wiley, New York, 1972,vol. 3, ch. 24;H. C. Brown, lsquo;The Nonclassical Ion Problemrsquo; (with comments by P. v. R. Schleyer), Plenum, New York, 1977;C. A. Grob, Acc. Chem. Res., 1983,16,426; H. C. Brown, ibid., p. 432;G. A. Olah, G. K. S. Prakash, and M. Saunders, ibid., p. 440; C. Walling, ibid., p. 448;V. A. Barkhash, Top. Curr. Chem., 1984, 116/117, 1.7 A. Nickon and J. H. Hammons, J. Am. Chem. SOC.,1964,86,3322;J. H. Hammons, E. K. Probasco, L. A. Sanders, and E. J. Whalen, J. Org. Chem., 1968, 33, 4493; J. Paasivirta, Acta Chem. Scand., 1973, 27, 374. 8 N. H. Werstiuk, D. Dhanoa, and G. Timmins, Can. J. Chem., 1979, 57, 2885; 1983,61, 2403. 9 K.-T. Liu, Tetrahedron Lett., 1978, 1129. 10 J. Passivirta, in lsquo;Lanthanide Shift Reagents in Stereochemical Analysis (MSA Series),rsquo; ed. T. C. Morrill, VCH Publishers, 1986,vol. 5, ch. 4. 11 F. P. Cappelli, G. Timmins, and N. H. Werstiuk, Can.J.Chem., 1972, 50,2163;N. H. Werstiuk, G. Timmins, and F. P. Cappelli, ibid., 1980, 58, 1709;N. H. Werstiuk and F. P. Cappelli, ibid., p. 1725. 12 (a) M.Lajunen and H. Lyytikainen, Finn.Chem. Lett., 1975, 97; (6) M. Lajunen and R. Ollikka, ibid., 1978, 272. 13 M. Lajunen and P. Hirvonen, Finn. Chem. Lett., 1978, 38. 14 M. Lajunen and T. Sura, Tetrahedron, 1978,34,189;M. Lajunen and H. Lyytikainen, Acta Chem. Scand., 1981, A35, 131, 139;M. Lajunen and R. Hiukka, ibid., 1985, A39, 109. 15 V. Gold and M. A. Kessick, J. Chem. Soc., 1965, 6718. 16 M.Lajunen and A. Anderson, Acta Chem. Scand., 1982, A36, 371. 17 M. Lajunen, Acc. Chem. Rex, 1985, 18, 254. 18 (a) C. A. Grob, B. Schaub, and M. G. Schlageter, Helu. Chim. Acta, 1980, 63, 57; (6) S. Siege1 and J. M. Komarmy, J. Am. Chem. SOC., 1960, 82, 2547. 19 R. L. Baird and A. A. Aboderin, J. Am. Chem. SOC.,1964,86, 252. 20 (a) M. Lajunen and P. Hirvonen, Finn.Chem. Lett., 1974,245;(b)M. Lajunen and T. Sura, ibid., 1979, 233; (c) M. Lajunen and H. Kukkonen, Acta Chem. Scand., 1983, A37,447; (d) M. Lajunen and A. Hintsanen, ibid., p. 545;(e) M. Lajunen, ibid., 1985, A39, 85. 21 M. Lajunen and M. Wallin, Finn. Chem. Lett., 1974,251. 22 M. Lajunen and P. Wiksten, Finn. Chem. Lett., 1980, 17. 23 M.Lajunen, Acta Chem. Scand., 1974, AD, 939. 24 J. Paasivirta, Suomen Kemi., 1963, B36, 156. 25 M. Lajunen, Finn. Chem. Lett., 1975,153. 26 E.M. Arnett, Prog. Phys. Org. Chem.,1963,1,223;G. Perdoncin and G. Scorrano, J. Am. Chem. Soc., 1977,99,6983. 27 M. Charton, Prog. Phys. Org. Chem., 1981, 13, 119. 28 N. H.Cromwell and M. A. Graff, J. Org. Chem., 1952, 17,414. 29 V. Mancini, G.Morelli, and L. Standoli, Gazz. Chim. Ital., 1977,107, 47. 30 E. Lippmaa, T. Pehk, and J. Paasivirta, Org. Magn. Reson., 1973, 5, 277;G.C. Levy, R. L. Lichter, and G. L. Nelson, lsquo;Carbon-13 Nuclear Magnetic Resonance Spectroscopy,rsquo; 2nd ed., Wiley, New York, 1980, p. 60 N. H. Werstiuk, R. Taillefer, R. A. Bell, and B. Sayer, Can. J. Chem., 1973,51, 3010; D. P. Kelly and H. C. Brown, J. Am. Chem. SOC.,1975, 97, 3897. Received 23rd December 1985; Paper 512257