The one-electron reduction of carbonium ions. Part 14. Effect of successive introduction of methyl substituents on the reducibility of tropylium ion in chromium(II) ion and cathodic reductions

摘要

J.C.S. Perkin II The One-Electron Reduction of Carbonium Ions. Part 14.1 Effect of Successive Introduction of Methyl Substituents on the Reducibility of fropylium Ion in Chromium(ii) Ion and Cathodic Reductions By Ken'ichi Takeuchi, Takeshi Kurosaki, Yasunori Yokomichi, Yoshihiro Kimura, Yrtsuhiro Kubota, Hiroshi Fujimato, and Kunio Okamoto," Department of Hydrocarbon Chemistry, Faculty of Engineering, Kyoto University, Sakyo-ku, Kyoto 606, Japan The second-order rate constants (k,) of one-electron reduction with CrI' ion of rnethylated tropylium ions C,H,-,,- (CHB),+CIO4-, n = 0-73 were determined in 10 hydrochloric acid at 25.0 "C. Reduction peak potentials deter- mined by means of triangular wave cyclic voltammetry and charge transfer energies with pyrene were compared with the rates of CrII ion reduction in terms of the effect of methyl substituents.The log k, values decrease linearly with an increase in the number of methyl groups up to n = 4. However, the decelerating effect per extra methyl substituent becomes smaller when nis 5 or 6, and the seventh methyl substituent accelerates the reduction. The methyl substituent effect on the reduction peak potentials of the carbocations is similar to that found for the CrT1 ion reduction, giving rise to a linear correlation. A Marcus treatment suggests that CrlI ion reduction proceeds through an outer-sphere mechanism. The charge transfer energies with pyrene are linearly correlated with reduction potentials for ions with n = 0-6, but considerable deviation is noted when n is 7.The charge transfer energies are linearly Correlated with the LUMQ energy levels of the methylated tropylium ions, suggesting that the unexpectedly pronounced reducibilities of hexa- and hepta-methyltropylium ions are attributable to steric factors. The deviation from linearity in the Crl' ion reduction and the reduction potential are reasonably explained by assum-ing that non-bonded repulsive interactions among congested methyl substituents in the carbocations are relieved upon reduction to their corresponding radicals. This explanation is consistent with our previous conclusion from a 1 n.m.r. study that the methyl substituents of 1,2,3,4,5penta-, hexa-, and hepta-methyltropylium ions are highly congested and distorted out of plane.THEst ability-reactivity relationship for carbocations has long been an interesting problem in physical organic chemistry. In previous papers, we demonstrated the existence of linear free energy relations between the rates of one-electron reduction of various substituted cyclo- propenium and tropylium ions with CrII ion or zinc and properties such as pKR+, reduction potential, and charge transfer energy with pyrene.2 Thus it has been shown that the one-electron reducibility is a good measure of the stability of carbocations. (2b)Feldman and Bowie have found that the log k, values for CrII ion reduction of several organic cations of various structural types including the tropylium ion are linearly correlated with their cathodic reduction potentials with a slope of 0.5.3 The result was nicely explained by the Marcus theory which indicates that one-electron reduc- (3a) (3 b)tion by CrII ion proceeds through an outer-sphere rnechani~rn.~ With such background, we were interested in ex-amining various correlations over a wide range of stability of various tropylium ions.For this objective, (4b)the methylated tropylium system appeared to be suitable since (a) the methyl substituent is not so bulky as to exert a steric effect on the reactivity, (b) the methyl substituent effect has been shown to be nearly additive for monomethyl-, 1,2-and 1,3-dimethyl-, and 1,2,3-trimethyl-tropylium ions,l and (c) all the methyl-(5a) (5b)tropylium ions were available (Scheme This paper describes the one-electron reducibilities of all the methyltropylium ions with CrII ion, and their correlations with reduction potentials and charge transfer energies with pyrene.The correlation of the charge transfer energies with LUMO energy levels calculated by the extended HMO method was also examined, SCHEME1 1981 671 RESULTS AND DISCUSSION electron reduction products via carbanions has been Products and Rates of CrT1Ion Reduction.-As yre-reported in the potassium reduction of heptaphenyl-viously described,2a the CrlI ion reduction of the tropylium tropylium ion and the CrII ion reduction of benzyl ions was conducted in 10 hydrochloric acid at 25.0 "C chloride.* Alternatively, the heptamethyltropyl radical, and the reaction was followed by measuring the amount for example, might produce heptamethyltropylidene and of reduction products by U.V.analysis. With the suc- hexamethylheytafulvene via disproportionation, the cessive introduction of methyl groups, the reaction latter being protonated to regenerate (7).7 We are not SCHEME2 became very slow, exhibiting rates not allowing con- in a position to clarify the mechanism for the monomer venient measurements. Therefore, the rates for penta-, formation at present.. hexa-, and hepta-methyltropylium ions were determined The second-order rate constants for Crr ion reduction under pseudo-first-order conditions by the use of a large of the methyltropylium ions were calculated and are excess of CrII ion.However, in the reactions of (5a---c), listed in the Table. Since the above mentioned observ- (6), and (7), considerable amounts of ' apparently-two-ations for (6) and (7) are ambiguous as to the significance electron ' reduction products (monomers) were formed in of their rate data, the data should be taken as approxi- addition to the expected bitropyls (dirners) (Scheme 2). mate measures for their rates of the ('rT1 ion reduction. Rates of Crlr ion reduction at 25.0 OC, reduction peak potentials and charge transfer v,,,,,. value.; with pyrene for metiiyl- tropylium ions Rf Initial concentration Reduction (..----A-potential Cation 1O~R+/M t 02crlII k,/l mo1-l 5-l us. s.c.e./V lo-, ~,,,,,./cm (0) 1.80 3.05 74.0 4 0.2 -0.120 1.81 (1) 1.99 4.13 11.1 amp; 0.2' -0.233 1.88 (24 0.294 1.10 0.908 f 0.01 -0.348 1.98 (2b) 0.334 1.22 1.36 f 0.02 -0.298 1.94 (24 0.313 1.23 0.865 amp; 0.026 -0.312 1.94 (34 0.271 0.866 0.148 Ik 0.002 -0.428 2.00 (3b) 0.300 0.938 0.169 0.005 -0.438 2.01 (34 0,302 1.03 0.114 3-0.001 -0.420 1.99 (3d) 0.300 0.517 0.107 $1 0.004 -0.395 1.99 (4a) 0.742 2.25 (2.07 0.01) X 10 ' -0.500 2.07 (4b) 0.602 3.18 (2.55 f 0.06) x lo-' -0.517 2.07 (44 0.299 1.55 (1.61 f 0.11) x 10P -0.521 2.08 (4d) 0.544 16.0 (2.28 0.01) x 10F -0.495 2.04 (5a) 0.655 12.4 (6.89 * 0.30) x 10P -0.548 2.13 (5b) 0.364 6.90 4.32 x 10-3 -0.549 2.11 (5c) 0.363 12.3 (2.65 amp; 0.02) x -0.604 2.13 0.167 17.9 (2.17 amp; 0.29) x lW3 -0.594 2.17 (7)(6) 0.0697 16.4 (7 f 2) x 10-3 -0.588 2.35 a Data for representative measurements.Average for two data. Data taken from ref. 2b. d Data taken from ref. 1. 6 Sol-vent, CH,CI,; supporting electrolyte, O.lw-Bun,NCIO,; sweep rate, 0.1 V s-l. f With pyrene in CI(CH,),CI: data taken from ref. 5. Although quantitative analysis of the dimer : nionomer In order to examine the additivity of the methyl sub-ratios has not been conducted in each case, the product stituent effect on the rate of CrlI ion reduction, the from (6)was shown by 13C n.m.r. analysis to be composed logarithms of the rate constants were plotted against the of the monomer and the dimer in a mole ratio of ca. 2 : 1. number of the methyl substituents (Figure 1). There From the product mixture from (6) the dimer was are four characteristic features in the plot.First, the isolated in crystalline form, mp. 204.0-205.0 "C, in 20 successive introduction of methyl groups decelerates the yield. Compound (7), in particular, afforded the mono- reduction except in the case of (7). Secondly, the mer as the main product." The formation of two-isomers of di-, tri-, and tetra-methyltropylium ions show * The combination process of the tropyl radical is known to be similar reactivities in each case. Therefore, when the fast (tl,, s) compared with the first one-electron reduction number of the methyl substituent is less than five, the process." In contrast, the highly methylated tropyl radicals are expected to combine at slower rates for steric reasons, giving methyl substituerit effect is nearly additive, and the rise to the two-electron reduction products.In the reaction of more methylated the tropylium ion, the less reactive are (7) the shape of the U.V. spectra of the product changed with time, suggesting the intervention of homolytic cleavage of the f We thank a referee for the wggestion of the possibility of dimer once it is formed. disproportionation. 672 the ions due to increased stabilization. Thirdly, discrepancies among the rates of (5a---c) are evident, (5a) being the most reactive, and (6) reacts six times as fast as the rate predicted from the straight line in the plot. Fourthly, the most unexpected feature is that the introduction of the seventh methyl group accelerates the reduction.The faster rate of (7) than (6) was also ascertained from their competitive reduction with zinc powder in acetonitrile (see Experimental section). 2.c 0.C *N tT4 -2.0 -4.0 01234567 n FIGURE Plot of log k, for CrTTion reduction of methyltropyl- 1 ium ions in 10 HC1 at 26.0 "C against the number of the methyl substituents The lack of correlation for (5a) and (6) might be partly ascribed to the possible levelling of the electron- releasing ability of the methyl groups. However, such an effect as a sole factor cannot explain the abrupt change in the behaviour of (7). The most probable rationaliz- ation is ascribed to steric reasons. Previously, we proposed, on the basis of statistical analysis of 13C n.m.r. chemical shifts of the methyl and ring carbons of methyltropylium ions, that the methyl substituents of (5a), (6), and (7) are distorted out of plane, whereas the seven-membered ring is Koenig and Chang proposed on the basis of helium(1) photoelectron spectroscopy that the tropyl radical has deformed CzCsymmetry.1deg; They suggested that some C-C bonds in the tropyl radical are longer than those in the tropylium ion.Consequently, the pronounced reducibility of (7) may be rationalized by assuming that the non-bonded repulsive interactions among methyl substituents in (7)are relieved by its being convrrtecl into the corresponding radical. J.C.S. Perkin TI Cathodic Reduction.-The CrII ion reduction of the methyltropylium ions vividly showed the effect of the successive introduction of methyl groups on the re-ducibility of the tropylium ion.However, as stated above, there remained some ambiguity in the accuracy of the observed rates since some tropylium ions afforded two-electron reduction products. Consequently, with a view to examining whether the characteristic methyl substituent effect observed in the CrII ion reduction of (6) and (7) is general in the one-electron reduction of methylated tropylium ions, the reduction peak poten- tials were measured in dichloromethane by triangular wave cyclic voltammetry under irreversible conditions. The results are in the Table. The reduction potential decreases by ca. 0.1 V for each methyl introduction over the range from unsubstituted tropylium (0) to tetramethyltropylium ions (4a-4).However, this regular decrease does not hold for (5a-c), and a reversed trend takes place for (6) and (7),coinciding with the trend of the rates of CrT1 ion reduction. This behaviour of hexa-and hepta-methyltropylium ions which is conimon to both the CrII ion reduction and the cathodic reduction may be ascribed to the same origin, most probably to the relief of non-bonded repulsive interactions between the methyl substituents on con-version of the carbocations into their corresponding radicals. The relative rates of CrTr ion reduction designated by RTln(k,/k$O)) = 5.69 log(k,/kJo))/kJ mol-l can be linearly correlated with the reduction potentials over the whole range (a = 0 -7) (Figure 2).t The slope of the correlation line, 0.54 (Y O.SSS), is close to the theoretical red.Ys s.c e / k J mol-' FIGURE Correlation of relative rates for Cr" ion rctluction 2 A.69 log (k,/k,(u))with reduction peak potentials for methyl- tropyliurn ions value, 0.5, which is deduced from the Marcus theory, suggesting that CrJT ion reduction of methyltropylium ions in lo:amp; hydrochloric acid proceeds through an outer-sphere mechani~m.~ This sort of correlation had been observed by Feldnian and Bowie in the CrTT ion t We have previously demonstrated linear correlations between the log k, values for CrrJ ion reduction and the polarographic half-wave potentials for various substituted tropylium ions.a reduction of several organic cations in 75 ethanol in the presence of 2N-perchloric acid.3 Correlation of Redmtion Peak Potentials with Charge Transfer Energies.-According to Mulliken the charge transfer energy (EcT)is a linear function of the electron affinity (Es~.)of an electron acceptor and the ionization potential (Ip)of an electron donor equation (l).11912 Therefore, the EcT values determined for methyi-tropylium ions by the use of pyrene as an electron donor (Table) are expected to be linearly correlated with Eafl.with a slope of -1.0. In the present case we can use the EcT = -Eae. + Ip+ const. (1) reduction peak potential (Ered) in place of E,B.. When the EaT values are plotted as a function of Ered a linear correlation is found to hold for the methylated tropylium ions over the range n = 0-6 with a slope of -0.90 (Y 0.984) in good agreement with prediction (Figure 3).In this correlation also, the point for (6) deviates slightly upward, and that for (7) deviates significantly from the regression line. 280 0 (71 260 c 'd ? 240 bsol; I-V c, 22c 1 I I -60 -40 -20 Fred YS Ste / k J SnOI-' FIGURE Correlation of charge transfer energies with pyrene 3 with reduction peak potentials for iiiethyltropylium ions As mentioned before, the tieviation of one-electron rcducihility for tliese nietliylated tropyliuni ions has been nscribetl to steric reasons. Meanwhile, tlie cliarge trans-fer energy (E(Tr)is determined by the LUMO energy level of an electron acceptor.13 A plot of the LUMO 673 levels calculated by the extended HMO method * against the EcT values affords a fairly good linear correlation with slope 1.01 (Y0.929) (Figure 4).It should be noted -9.4 p 1.01 -9.6 bsol; 0 r 3 1 -9.8 (0) / -10.0 2.2 2.4 2.6 2.8 WV FIGURE Correlation of LUMO energy levels with charge 4 transfer energies with pyrene for methyltropylium ions that the points for hexa- and hepta-methyltropylium ions do not exhibit exceptional deviation. These results SUP-port the aforementioned interpretation that the un-expectedly high reducibilities of (6) and (7) can be attributed to steric reasons. The relief of the repulsive interactions among methyl substituents upon one-electron reduction to form the corresponding radicals provides a reasonable explanation.EXPERIMENTAL U.V. spectra were measured with a Hitachi 200-10 spectrophotometer. lH N.m.r. spectra were recorded on a Hitachi I-24 (60 MHz) instrument. 13C N.m.r. spectra were recorded on a JEOL FXlOO (25.80 MHz) instrument operating in the Fourier transform mode. Cyclic voltam- nietry was conducted with a Hokuto Denko potent io-galvanostat HA-104 and function generator HB-107A. Materials.-All the methyltropylium perchlorates were reported previ~usly.~ All reagents were of reagent grade quality except when otherwise noted. P~oduct of CrII Ion Reduction.--'rhe isolation of the products from Cr Iion reduction of the methyltropylium ions was conducted under conditions similar to those utilized in kinetic studies.Extraction of the hydrocarbon products with chloroform followed by evaporation of the solvent afforded the reduction products in 95-100~0 yields. The *The C-C bond length for the ring was estimated from the formula, 1.405 + n(1.4113 -1.4050) A, where 1.4050 A is the bond length in the tropylium ion, 1.4113 A is the average C-C bond length in the rnethyltropylium ring,14 and n is the number of methyl groups. A regular heptagon was assumed and the bond lengths utilized for Cring-H, C,,,,,-CH,, and C-H of methyl are 1.108,14l.FiOI,'* and 1.095 A,15respectively. The appropriate conformation of methyl substituents was assumed by adoptingGrant's assumptions on methylated benzenes,'" Other necessary data were taken from Hoff niann's paper." crude products were shown to retain the tropylidene moiety by lH n.m.r.analysis. Although the products from (4a-d) were methylated bitropyls (dimers), those from (5a-c) contained lO--ZO of two-electron reduction products (monomers), which were separated by preparative t.1.c. (SO,-hexane) and identified by lH n.m.r. However, the dirners from (5a-c) were unstable and decomposed during separation by t.l.c., exhibiting broad lH n.m.r. signals at 8 0.9-2.0. The crude product from (6)was shown by 13C n.m.r. (CDC1,) to be composed of 1,2,3,4,5,6-hexamethyl-tropylidene as the main monomer and dodecamethylbi- tropyl in a mole ratio of ca. 2 : 1; monomer, 6c (CDC1,) 16.1 (q), 17.7 (q),20.5 (q),42.2 (t),127.0 (s), 129.0 (s), and 135.0 p.p.m.(s); dimer, 8~ (CDCl,) 17.9 (q), 18.2 (q), 24.2 (q), 46.1 (d), 126.7 (s), 133.2 (s), and 133.4 p.p.m. (s); 6~ (CDC1,) 1.53 (12 H, s), 1.68 (12 H, s), 1.90 (12 H, s),and 2.87 p.p.m. (2 H, s). On addition of methanol to the product mixture the dimer was obtained as crystals, m.p. 204.0- 205.0 "C. The product from (7) showed a 13C n.m.r. (CDC1,) spectrum identical with that of heptamethyltropyl- idene, 6~ 14.1 (q),16.2 (q), 17.7 (q), 39.5 (d), 126.5 (s),131.6 (s),and 135.1 p.p.ni. (s). Kinetic Studies.-The second-order rate constants were determined by the batchwise method which was essentially similar to that described previously.1 The percentage reaction of carbocations were determined by measuring the U.V.absorbances of the reduction products at 260 or 265 nm in chloroform. For (4a-d) the absorbance at 10 half-lives were used for that at lOOyo reaction. The ions (5a--c), (6),and (7) reacted too slowly at the CrlI ion concentrations used in kinetic studies to permit the U.V. absorbance at lOOyo reaction to be measured ; therefore, CrII ion concentrations of 0.5-1.0~ were used and the absorbance at 1-2 h was used for that at lOOyoreaction. It was ascertained that the absorbances adopted as those for lO0yo reaction were constant at doubled reaction times. Competitive Reduction of (6) and (7) with Zinc in Aceto-nitrile.-A solution of (6) (5.7 mg, 0.021 mmol) and (7) (4.4mg, 0.015 mmol) in acetonitrile (10ml) was magnetically stirred with zinc powder (167 rng, 2.55 mmol) at 15.5 "C under nitrogen.After 30 min 0.50 ml of the solution was withdrawn, diluted with 10 HC1 to 5-00 ml, and washed with chloroform to remove hydrocarbon products. The U.V. spectrum of the aqueous phase was essentially super-J.C.S. Perkin I1 imposable on that of pure (6). Analysis of the absorptions at 311 E 5 750 for (6);2 050 for (7) and 350 nm E 535 for (6); 5 850 for (7) revealed that (6)and (7) reacted to the extent of 25 and gay0, respectively, indicating that (7) is ca. 10 times as reactive as (6)in the zinc reduction under these conditions. Cyclic Voltammetry.-The reduction peak potentials were measured for 0.001M-cation in 0.1~-Bu~,NC10, in CH,Cl, using a three-electrode cell with platinum wire and auxiliary electrodes and a saturated calomel reference electrode.Scanning was conducted from +0.5 to -1.5 V at a rate of 0.1 V s-l under irreversible conditions. 0/1452 Received, 22nd September, 19801 REFERENCES Part 13, K. Takeuchi, K. Komatsu, K. Yasuda, F. Mikuchi, and K. Okamoto, J. Chem. SOC., Perkin Trans. 2, 1979, 1005. (a) K. Okamoto, K. Komatsu, S. Tsukada, and 0. Murai, Bull. Chem. SOC. Jpn., 1973, 46, 1780; (b) K. Okamoto, K. Komatsu, 0. Murai, 0. Sakaguchi, and Y. Matsui, ibid., p. 1785; (c) K. Okamoto, K. Komatsu, M. Fujimori, and S. Yasuda, ibid., 1974,47, 2426; (d)K. Okamoto, K. Komatsu, and 0.Sakaguchi, ibid., p. 2431; (e) K. Komatsu, K. Takeuchi, N. Abe, K.Yasuda, and K. Okamoto, J. Chem. SOC., Perkin Trans. 2, 1979, 262. W. T. Bowie and M. R. Feldman. J. Am. Chem. SOC., 1977,99, 4721. R. A. Marcus, Can. J. Chem., 1959, 37, 155. 6 K. Takeuchi, Y. Yokomichi, T. Kurosaki, Y. Kimura, and K. Okamoto, Tetrahedron, 1979, 85, 949. P. H. Plesch and A. Stasko, J. Chem. SOC. B, 1971, 2052 R. Breslow and H. W. Chang, J. Am. Chem. SOC., 1965, 87, 2200. 8 (a) J. K. Kochi and D. Buchanan, J. Am. Chem. Soc., 1965,87, 853; (b) J. K. Kochi and D. D. Davis, ibid., 1964, 86, 5264. @ K. Takeuchi, Y. Yokomichi, and K. Okamoto, Tetrahedron, 1980, 36,2945. lo T. Koenig and J. C. Chang, J. Am. Chem. SOC., 1978, 100, 2240. l1 R. S. Mulliken, J. Am. Chem. SOC., 1952, 74, 811. l2 M. Feldman and S. Winstein, Theor. Chim. Acla, 1968, 10, 86. 13 A. Streitwieser, jun., ' Molecular Orbital Theory for Organic Chemists,' Wiley, New York, 1961, ch. 7. 14 M. J. S. Dewar and D. Landman, J. Am. Chem. SOC., 1977,99, 7439. l5 L. Pauling, ' The Nature of the Chemical Bond,' Cornell University Press, Ithaca, 1960, 3rd. edn., ch. 7. l6 W. R. Woolfenden and D. M. Grant, J. Am. Chem. SOC., 1966,88, 1496. K.Hoffmann, J. Chem. Phys., 1963, 39, 1397.