摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1615 Synthesis of two novel para-extended bisaroxyls and characterization of their triplet spin states Andreas Rebmann, Jinkui Zhou, Paul Schuler, Anton Rieker * and Hartmut B. Stegmann* Institute of Organic Chemistry, University of Tübingen, Auf der Morgenstelle 18, D-72076 Tübingen, Germany 29,3,30,5,59,50-Hexa-tert-butyl-1,19:49,10-terphenyl-4,40-dioxyl 3c and 3,3+,5,5+-tetra-tert-butyl-1,19:49,10:40,1-:4-,1+- quinquephenyl-4,4+-dioxyl 3e, obtained by oxidation of the corresponding biphenols 1c and 1e, show thermally excited triplet spin states.Radical centres, separated by suitable spacers, constitute one of the basic principles in the design of organic molecular ferromagnets. 1 Thus, 2,6-di-tert-butylphenoxyl radical systems,2,3 connected by m-phenylene units in the 4 positions, have been largely used to synthesize oligo-radicals with triplet, quartet and quintet ground states.4–7 We are currently investigating biradical systems 3, in which two phenoxyl radicals are coupled by various p-arylene spacers, in order to clarify the prerequisites necessary for such spacers also to act as ferromagnetic couplers.Whereas the biphenyl and terphenyl species 3a and 3b, respectively, quantitatively exist as extended quinones 4a and 4b (EPR silent, aside from traces of mono-radical impurities in 4b 8), the p-quaterphenylene system d is able to exist in an equilibrium of biradical (3d) and quinone (4d) states.9 In order to favour the biradical/triplet state 3 over the quinonoidal singlet state 4 further, an effective p-orbital overlapping must be prevented, e.g.by twisting or extending the linear aromatic spacer system. Therefore, we synthesized the compound series c and e † and characterized the radical species 2 and 3 by EPR–ENDOR and voltammetry. The EPR spectra of 3c,e in toluene exhibited broad absorption signals (DH = 5 G) at 293 K, resulting from dipole broadening in the biradicals. These signals were superimposed by hyperfine structure signals of the mono-radicals 2c,e with the same g-factors (g2c = 2.00448, g2e = 2.00422) as the biradicals. They presumably originated from a partial oxidation of the corresponding bisphenols 1c,e.ENDOR-spectra in toluene at 233 K gave the following coupling constants (in G, 1 G = 9.3408 × 1025 cm21): for 2c aH2,6 = 1.69, aH69 = 0.50, aH39 = 0.21 and aH-But = 0.09, and for 2e aH2,6 = 1.83, aH29,69 = 1.60, aH39,59 = 0.71, aH20,60 = 0.32, aH30,50 = 0.13 and aH-But < 0.10.† 3,5-Di-tert-butyl-4-(trimethylsilyloxy)phenylboronic acid and 1,4- dibromo-2,5-di-tert-butylbenzene or 4,40-dibromoterphenyl reacted via Pd-catalyzed cross coupling (Suzuki reaction) in toluene to yield 67% of bis(OSiMe3) protected 1c (mp 262 8C) or 53% of bis(OSiMe3) protected 1e (mp 280 8C), respectively. Cleavage of the SiMe3 groups with excess diluted aqueous HCl in boiling THF gave 80% of 1c (mp 355 8C, from toluene) or 98% of 1e (mp 320 8C, from ethyl acetate).The biphenols 1c and 1e show the expected 1H and 13C NMR spectra and elemental analyses. The biradicals 3c and 3e were prepared by oxidation of toluene solutions of 1c and 1e with PbO2 or aqueous KOH–K3Fe(CN)6 in an EPR tube under argon via the intermediary mono-radicals 2c and 2e. For the preparation of the powder samples of 3c,e, 6.0 g of K3Fe(CN)6 and 5.0 g of KOH in 40 ml of water were added to a solution of 0.1 g (0.167 mmol) of 1c in 100 ml toluene, or 0.5 g (0.783 mmol) of 1e in 100 ml toluene, each under argon.After shaking well for 5 min, the organic phase was separated under argon and again oxidized with fresh K3Fe(CN)6 and KOH–H2O. Then, the organic phase was separated, dried over CaCl2 and the toluene removed in vacuo, yielding 3c (83 mg, 83%) or 3e (460 mg, 92%). Toluene solutions of 3c or 3e were frozen in the EPR cavity at 77 K. The resulting matrices revealed characteristic Dms = 1 transitions (zero field splitting, zfs) centred at 3310 G along with forbidden, but strong Dms = 2 transitions (half-field reson- OH R R OH O R R OH O R R O O R R O 2 1 3 4 n n n n • • • O O 1 23 4 5 6 1¢ 2¢ 3¢ 4¢ 5¢ 6¢ 1¢¢ 2¢¢ 3¢¢ 4¢¢ 5¢¢ 6¢¢ c O 1 2 3 4 5 6 1¢ 2¢ 3¢ 4¢ 5¢ 6¢ 1¢¢ 2¢¢ 3¢¢ 4¢¢ 5¢¢ 6¢¢ O 1¢¢¢ 2¢¢¢ 3¢¢¢ 4¢¢¢ 5¢¢¢ 6¢¢¢ 1¢¢¢¢ 2¢¢¢¢ 3¢¢¢¢ 4¢¢¢¢ 5¢¢¢¢ 6¢¢¢¢ a n = 0 b n = 1, R = H c n = 1, R = But d n = 2, R = H e n = 3, R = H Ox Ox e1616 J.Chem.Soc., Perkin Trans. 2, 1997 ance, hfr) at 1655 G (see insets in Fig. 1). The typical zfs pairs of lines corresponding to the orientations of the triplet species (gfactor anisotropy) are described by zfs parameters D and E (3c: |D/hc| = 0.0053 cm21, |E/hc| = 0 cm21; for 3e: |D/hc| = 0.0018 cm21, |E/hc| = 0 cm21). The two signals in the centres of each spectrum of Fig. 1 may again be attributed to the monoradicals 2c,e. The unsymmetrical nature of the matrix spectrum of 3e may be attributed to a g-factor anisotropy.In sealed samples at 25 8C, these central signals increase with time, even under an inert atmosphere and in the dark, with a simultaneous decrease of the zfs signals. Also the total signal intensity decreases, therefore, reversible association of the biradicals by partial spin coupling to double- and oligo-radicals may take place within several days. This would explain the absence of any zfs and hfr in EPR spectra of powders of 3c,e and the rather low susceptibility of those powders as measured by a SQUID magnetometer: for 3c, cmol = 1.54 × 1022 emu mol21 at 5 K; for 3e, cmol = 6.68 × 1023 emu mol21 at 5 K.After redissolution of the powders in toluene, we can again find zfs and hfr of 3c,e in the matrix formed at 77 K. By comparing the radical intensity of 3c,e with that of a 1,3,5-triphenylverdazyl solution by double integration of the zfs we estimate a biradical concentration of 94% of 3c and 96% of 3e at room temperature.Furthermore, we detect a low concentration of the mono-radical (<5% of the total EPR signal) and do not find any indication of a quinonoid structure 4 by IR spectrometry. If the observed D value is explained in terms of a point-dipole approximation 10 [D (G) = 27 810/r3], the distance r between the two radical centres can be estimated for 3c as 7.9 Å and for 3e as 11.4 Å. From X-ray analysis of single crystals of 1e and the bis dimethyl ether of 1c we measured the O? ? ?O distance 14.16 Å for 1c(Me2) and 22.64 Å for 1e.This shows again that the point dipole model is not stringent in the case of biradicals with delocalized spins.11 The value of E = 0 cm21 for Fig. 1 Matrix EPR spectrum of 3c and 3e (zfs) with half-field resonance (hfr) (see insets) both compounds indicates axial symmetry of the biradicals in the matrix.12 The ground state multiplicity and the singlet– triplet gap DET–S = 2 J was determined from the temperature vs.intensity plots (Fig. 2).13 The non-linearity of the curves in the range of 93–293 K clearly indicates singlet ground states for 3c and 3e which are in thermal equilibrium with their triplet states. For this case, the intensity I of the triplet EPR-signal is given by eqn. (1) 14 where J is the exchange coupling between the IT = C[3 exp(22J/RT)]/[1 1 3exp(22J/RT)] (1) unpaired electrons T, the absolute temperature and R = 1.987 cal mol21 K21. Non-linear least-squares curve fitting of the data in Fig. 2 to eqn.(1) yields J = 2156 ± 4 cal mol21 (255 cm21) for 3c and J = 2103 ± 2 cal mol21 (36 cm21) for 3e. The cyclic (CV) and differential-pulse (DPV) voltammetry of 3c (Fig. 3) in pyridine solution show two peak couples (E81 = 20.56 V; E82 = 20.68 V vs. Ag1/Ag). In contrast to 3c and other extended quinones,8,9 3e exhibits only one reduction and re-oxidation peak (E8 = 20.55 V). With increasing scan rate, the reduction and re-oxidation peaks of 3c,3e are shifted towards more negative and more positive potentials, respectively.Controlled-potential electrolysis (CPE) at 21.0 V proves a one-electron transfer for each peak couple of 3c and a formal two-electron transfer for the peak couple of 3e. Apparently, the first and second reduction steps of 3e occur at potentials very close together. This is a consequence of the lower interaction of the electrons in 3e (low D and J values) according to which both radical centres are reduced independently.From the structure of 1c and 1e and the experimental results of CV, DPV and CPE measurements, the electrochemical reduction in both cases can be denoted as a quasi-reversible electron transfer (EE) process. In summary, our results reveal that p-arylene units, although producing an antiferromagnetic ground state, may cause effec- Fig. 2 Curie law plots of EPR intensity (I) vs. 1/T for biradicals 3c,e Fig. 3 DPV of 0.5 mM 3c at a Pt electrode in pyridine solution containing 0.1 M NBu4PF6; reference electrode Ag/Ag1 (0.01 M AgClO4, in MeCN with 0.1 M NBu4PF6), pulse amplitude 50 mV, pulse width 200 ms, scan rate 20 mV s21, potential sweep (a) 0 to 21.2 V, (b) 21.2 to 0 VJ. Chem.Soc., Perkin Trans. 2, 1997 1617 tive ferromagnetic coupling since the triplet state is strongly populated at temperature >200 K if the arylene spacer is twisted (3c) or extended by two or three benzene rings (3d and 3e). Atomic coordinates, bond lengths and angles, and thermal parameters have been deposited at the Cambridge Crystallographic Data Centre (CCDC).For details of the deposition scheme, see ‘Instructions for Authors’, J. Chem. Soc., Perkin Trans. 2, 1997, Issue 1. Any request to the CCDC for this material should quote the full literature citation and the reference number 188/95. Supplementary data are also available (suppl. no. 57269, 4 pp.) from the British Library. For details of the Supplementary Publications scheme see ‘Instructions for Authors’ as above.Acknowledgements We thank the Volkswagenstiftung (I/71009) and the Land Baden-Württemberg for support of this work. References 1 J. S. Miller and A. J. Epstein, Angew. Chem., Int. Ed. Engl., 1994, 33, 385. 2 E. Müller, A. Schick and K. Scheffler, Chem. Ber., 1959, 92, 474. 3 A. Rieker and K. Scheffler, Liebigs Ann. Chem., 1965, 689, 78. 4 G. Kothe, K.-H. Denkel and W. Summernam, Angew. Chem., Int. Ed. Engl., 1970, 9, 906. 5 G. Kothe, E. Ohmes, J. Brickmann and H. Zimmermann, Angew. Chem., Int. Ed. Engl., 1971, 10, 938. 6 D. E. Seeger, P. M. Lahti, A. R. Rossi and J. A. Berson, J. Am. Chem. Soc., 1986, 108, 1251. 7 K. Mukai and N. Inagaki, Bull. Chem. Soc. Jpn., 1980, 53, 2695. 8 J. Zhou and A. Rieker, J. Chem. Soc., Perkin Trans. 2, 1997, 931. 9 A. Rebmann, J. Zhou, P. Schuler, H. B. Stegmann and A. Rieker, J. Chem. Res., (S), 1996, 318 (M), 1996, 1765. 10 T. Mitsumori, N. Koga and H. Iwamura, J. Phys. Org. Chem., 1994, 7, 43. 11 W. Adam, H. M. Harrer, F. Kita, H.-G. Korth, and W. M. Nau, J. Org. Chem., 1997, 62, 1419. 12 E. Wasserman, L. C. Snyder and W. A. Yager, J. Chem. Phys., 1964, 41, 1763. 13 M. Minato and P. M. Lahti, J. Am. Chem. Soc., 1997, 119, 2187. 14 B. Bleaney and K. D. Bowers, Proc. R. Soc., London, Ser. A, 1952, 214. Paper 7/04590E Received 30th June 1997 Accepted 7th July 1997

ISSN:1472-779X    

DOI:10.1039/a704590e

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1619 EVects of Lewis acid on the sign of exchange interaction of photogenerated radical ion pairs Shinji Sekiguchi, Kimio Akiyama and Shozo Tero-Kubota * Institute for Chemical Reaction Science, Tohoku University, Katahira 2-1-1, Aobaku, Sendai 980-77, Japan FTEPR study shows that the sign of the exchange interaction of the radical ion pair including the 4,49-dimethoxybenzophenone radical anion and 1,4-diazabicyclo[2,2,2]- octane (DABCO) radical cation changes from positive to negative by addition of BF3. It is well known that radical pairs usually have a negative exchange interaction (J), i.e.where the singlet radical pair is the lower in energy. However, there have been reports that J is positive in some radical ion pairs.1–4 A conflicting mechanism has also been proposed.5 Recently, we showed unambiguous evidence of a positive J for the radical ion pairs including the radical anions of benzophenone and its derivatives with DABCO~1.4 Note that the neutral radical pairs generated under the same conditions have a negative J.The factors involved in determining the sign of J are unclear. We have studied the effects of Lewis acid on the sign of J of the radical ion pair, since the charge of the radical pair seems to be an important factor in the interradical interactions. It is expected that a Lewis acid can form a weak complex with the photogenerated radical anion, leading to partial shielding of the charge. 4,49-Dimethoxybenzophenone (DMBP, Tokyo Kasei Chemicals) was recrystallized from ethanol. DABCO (Ncalai Tesque) was purified by sublimation. Dimethyl sulfoxide (DMSO, Ncalai Tesque) was used as received. The complexes of DABCO?(BF3)2 and DABCO?BF3 were prepared by the interchange reaction of BF3–diethyl etherate (Wako Chemicals) with DABCO in benzene according to the method reported in the literature.6 The precipitated complex was evaporated and dried under vacuum.FTEPR measurements were carried out using a Bruker ESP 380E EPR spectrometer. The pulse sequence and detection method were reported in previous papers.4,7 The nanosecond laser photolysis spectra were recorded using a multichannel analyzer equipped with a diode array (Princeton Instruments, IRY-700). A Nd:YAG laser (355 nm ) was used as the light source. All measurements were performed at room temperature. Figs. 1(a) and 1(a9) show the parts of the FTEPR spectrum observed at a delay time of 200 ns after the pulsed laser excitation of DMBP (3 × 1023 mol dm23) in the presence of DABCO (0.06 mol dm23) in DMSO at room temperature. The spectra were measured separately because the excitation bandwidth of the microwave pulse was too narrow to excite the whole spectrum.The external magnetic fields were set to the regions of a part of the lower [Fig. 1(a)] and higher [Fig. 1(a9)] fields of the DABCO~1 spectrum. The E/E* (E 1 A/E) polarization observed indicates the contribution of the net Epolarization due to triplet mechanism (TM) and weak A/Eradical pair mechanism (RPM), where E and A denote emission and absorption of microwave radiation, respectively.Since the A/E-RPM is generated from the triplet reaction process, we can conclude that the present radical ion pair has a positive J as reported previously.4 Addition of a Lewis acid into the above system induced a drastic change in the CIDEP pattern, while the hyperfine splitting constants obtained are completely identical to those of free DABCO~1.Figs. 1(b) and 1(b9) depict the parts of the FTEPR spectrum obtained from the laser excitation of DMBP (3 × 1023 mol dm23) in the presence of 0.06 mol dm23 DABCO?(BF3)2 complex in DMSO. The FTEPR spectrum shows that the photoinduced electron transfer definitely occurs to produce free DABCO~1. The E*/A-polarized spectrum due to DABCO~1 is easily interpreted by the contribution of the E-TM and E/A-RPM.These facts suggest that J is negative for this radical ion pair. The time profiles of the RPM polarization are shown in Fig. 2. The RPM components were obtained from the difference between the signal amplitudes of 2mI and 1mI hyperfine lines. In the DMBP–DABCO and DMBP–DABCO?(BF3)2 systems, their A/E and E/A RPM phases were unchanged during the observation. On the other hand, the polarization inversion was observed in the DMBP–DABCO?BF3 system. The A/Epolarization changed to the E/A one at 550 ns, indicating the existence of two kinds of radial ion pairs in this system.The growth and decay curves were analyzed by the sum of exponential functions. The curve fitting shown in Fig. 2 gave a decay time constant of 560 ns for the A/E-polarization. On the other hand, a relatively long time constant (910 ns) was obtained for the decay of E/A-polarization. In order to clarify the structure and dynamics of the reaction intermediates, we measured transient absorption spectra. The excitation of DMBP without the electron donor gave the transient spectrum having a peak maximum at 550 nm and a broad band with an absorption tail up to 750 nm.The spectrum is attributed to the T–T absorption of DMBP. A lifetime of 130 ns was obtained. As shown in Fig. 3(a), the T–T absorption was Fig. 1 Symmetrically located portions (a, b: the low magnetic field lines; a9, b9: the high field lines) of the echo detected FTEPR spectra observed at the delay time of 200 ns after the laser pulse excitation of DMBP (3 × 1023 mol dm23) in the presence of 0.06 mol dm23 DABCO (a) and DABCO?(BF3)2 (b) in DMSO1620 J.Chem. Soc., Perkin Trans. 2, 1997 quenched by the addition of DABCO and resulted in a band at 700 nm. This is easily assigned to DMBP~2.8 The absorption of DABCO~1 would be observed in the wavelength region below 500 nm.9,10 The band at 700 nm decays with a lifetime of 590 ns, which agrees well with the decay time constant of A/EFig. 2 Time variations in the RPM component of the CIDEP spectra calculated by the subtraction of low and high field lines of DABCO~1 generated from the laser photolysis of DMBP (3 × 1023 mol dm23) in the presence of 0.06 mol dm23 DABCO (a), DABCO?BF3 (b) and DABCO?(BF3)2 (c) in DMSO Fig. 3 Transient absorption spectra observed immediately after the laser photolysis of DMBP (3 × 1023 mol dm23) in the presence of 0.06 mol dm23 DABCO (a) and DABCO?(BF3)2 (b) in DMSO polarization in the FTEPR spectrum.It can be deduced that the back electron transfer reaction governs the decay rate of the radicals in the DMBP–DABCO system. Fig. 3(b) shows the transient absorption spectrum observed immediately after laser excitation of the DMBP–DABCO? (BF3)2 system. The spectrum completely coincides with the T–T absorption of DMBP, indicating very low quenching rate by DABCO?(BF3)2. This is probably due to the high oxidation potential of the complex. The slow buildup and small contribution of the TM observed in the FTEPR spectra supported this speculation.No absorption due to free DMBP~2 was observed after the disappearance of the triplet signal, while the FTEPR experiments indicated that photoinduced electron transfer clearly occurred. Since BF3 is a strong electrophilic reagent, a DMBP~2?BF3 complex would be formed after photolysis, leading to a blue shift of the absorption compared with that of free DMBP~2. It can be deduced that the complex formation weakens the Coulomb interaction between the ion radicals and serves to slow down the back electron transfer rate.The present study clearly shows that the radical ion pair including DMBP~2 and DABCO~1 has a positive J and the addition of BF3 induces the inversion of the sign of J. Acknowledgements The present research was partially supported by Grants-in-Aid on Priority-Area-Research on Photoreaction Dynamics (No. 06239103) and of Scientific Research No. 07404040 from the Ministry of Education, Science, Sports and Culture, Japan. References 1 H. Murai and K. Kuwata, Chem. Phys. Lett., 1989, 164, 567. 2 S. N. Batchelor, H. Heikkila, C. W. M. Kay, K. A. McLauchlan and I. A. Shkrob, Chem. Phys., 1992, 162, 29. 3 A. S. Jeevarajan and R. W. Fessenden, J. Phys. Chem., 1992, 96, 1520. 4 S. Sekiguchi, K. Akiyama and S. Tero-Kubota, Chem. Phys. Lett., 1996, 263, 161. 5 N. J. Avdievich, A. S. Jeevarajan and M. D. E. Forbes, J. Phys. Chem., 1996, 100, 5334. 6 J. J. Harris and S. C. Temin, J. Appl. Polym. Sci., 1996, 10, 523. 7 K. Akiyama, S. Sekiguchi and S. Tero-Kubota, J. Phys. Chem., 1996, 100, 180. 8 K. Bhattacharyya and P. K. Das, J. Phys. Chem., 1986, 90, 3987. 9 C. Devadoss and R. W. Fessenden, J. Phys. Chem., 1990, 94, 4540. 10 H. Miyasaka, K. Morita, K. Kamada and N. Mataga, Chem. Phys. Lett., 1991, 178, 504. Paper 7/04366J Received 20th June 1997 Accepted 24th June 1997

ISSN:1472-779X    

DOI:10.1039/a704366j

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1621 Formation of silylketenes via a 1,3-silyl shift. A theoretical study Magali Oblin,a,b Frédéric Fotiadu,c Michel Rajzmann *,b and Jean-Marc Pons *,a a Laboratoire Réactivité en Synthèse Organique (RéSo), UMR 6516, Faculté de St Jérome, boite D12, F-13397 Marseille cedex 20, France b ESA 6009, boite 561, F-13397 Marseille cedex 20, France c Laboratoire de Synthèse Asymétrique, UMR 6516, ENSSPICAM, Faculté de St-Jérôme, F-13397 Marseille cedex 20, France Ab initio and semiempirical calculations show that the formation of silylketenes from silyloxyacetylenes by a 1,3-silyl shift should occur through a concerted closed-shell mechanism involving retention (rather than inversion) of con- figuration at the silicon centre.Silylketenes are stable versatile organic reagents which are involved in numerous important reactions, e.g. Lewis acidpromoted [2 1 2] cycloaddition reactions with carbonyl compounds, Wittig-type reactions, diazomethane insertions and nucleophilic additions of carbon or heteroatomic nucleophiles.1 The first member of this family, trimethylsilylketene, was prepared for the first time in 1964 by Shchukovskaya et al.by thermolysis of ethoxy(trimethylsilyl)acetylene at 120 8C.2 Since this pioneering work, various kinds of silylketenes have been prepared by a number of other methods.3 Sakurai et al., in particular, showed that alkyl(trimethylsilyl)ketenes can be prepared from ethoxyalkynes in the presence of trimethylsilyliodide. 4 This procedure was then generalized5 and we used hexyl(trimethylsilyl)ketene, prepared from ethoxyoctyne, in our syntheses of lipstatin 6 and tetrahydrolipstatin,7 two bioactive b-lactones.The reaction could proceed through the initial formation of a silyl ynol ether which would then undergo a 1,3-silyl shift to yield a silylketene (Scheme 1). Such a mechanism is in agreement with the well known ability of trimethylsilyliodide to cleave ethers 8 and the observed formation of ethyliodide as the reaction by-product.It must however be said that a concerted mechanism cannot be excluded. As part of our theoretical interest in the chemistry of ketenes, formation of b-lactones through [212] cycloaddition 9 and thermolysis of alkoxyalkynes into ketenes,10 we report here our preliminary investigation on the formation of silylketenes via a 1,3-silyl shift from silyl ynol ethers. Calculations were performed both at the ab initio (HF and MP2 calculations with 6-31G* and 6-3111G** bases) 11 and semiempirical (AM1/RHF and AM1/CI)12 levels.Transition states were located by minimizing the gradient norm of the energy and were characterized by one negative eigenvalue of the Hessian matrix. All reaction paths were established unambigu- Scheme 1 O R¢ R O SiMe3 R O • Me3Si R + Me3SiI + R¢I ously by the intrinsic reaction coordinate (IRC) at the semiempirical level. A careful search of the potential energy surface of the parent reaction enabled us to identify two distinct concerted reaction paths (Scheme 2): path (a) involving a silyl shift with retention of configuration on the silicon atom and path (b) involving a silyl shift with inversion of configuration on the same centre.It appears from Tables 1 and 2 that, regardless of the calculation level, reaction path (a) involving retention of configuration of the silicon centre is favoured over reaction path (b) involving inversion of configuration of the same centre. However, note that the difference is much more pronounced in ab initio calculations than in semiempirical ones.This preference for retention path (a) is in full agreement with recent results reported by Yamabe 13 and Kira14 and their co-workers on the 1,3-sigmatropic silyl shift in allylsilane. Yamabe et al. propose that this stems from the fact that the silyl radical adopts a pyramidal form in contrast to the methyl radical and that in path (a) the silicon atom remains sp3-hybridized along the entire reaction path which is not the case in path (b) (Fig. 1). The small impact of the basis extension (from 6-31G* to 6- 3111G**) in ab initio calculations should be noted. This result is in contrast to our calculations on the 1,5-hydrogen shift in silylketene formation.10 Such a difference could result from the lower importance of diffuse orbitals (which are introduced at that extended level) in the 1,3-silyl shift when compared with the 1,5-hydrogen shift.Indeed, since the reactive heart is smaller in the present case, the electron displacement is sufficiently well accounted for by the 6-31G* basis. The relatively small difference between the treatment of the silicon atom in the two bases can also be invoked. It should be noted that AM1/CI calculations (CI = 8) did not lead to significant diminution of the activation energy (Tables 1 and 2). In either case the contribution of the lowest closed-shell configuration remains 99%, indicating clearly the closed-shell character of the mechanism.Furthermore, when calculated with CI = 20, but without optimisation of the CI = 8 geometry, Scheme 2 O SiR3 R¢ R¢ O Si R R R¢ O Si R R R R • R¢ R3Si O 1a–c 2a–c 3a–c 4a–c 1 2 3 4 5 path (a) path (b) retention inversion a R = R¢ = H b R = H, R¢ = Me c R = R¢ = Me1622 J. Chem. Soc., Perkin Trans. 2, 1997 Table 1 Main parameters of transition state 2a Method HF/6-31G* HF/6-3111G** MP2/6-31G* MP2/6-3111G** AM1/RHF AM1/CI = 8 Ea/kcal mol21 51.4 50.3 37.8 37.6 54.4 54.9 DrH/kcal mol21 2.02 21.1 1.25 28.3 228.3 224.4 rO1–C2/Å 1.263 1.253 1.283 1.278 1.266 1.250 rO1–Si/Å 2.057 2.061 1.990 2.036 2.379 2.332 rC3–Si/Å 2.432 2.467 2.434 2.457 2.438 2.485 /C2C3Si/8 59.4 58.5 57.6 58.0 70.7 71.3 /O1C2C3/8 161.4 162.7 160.9 161.8 155.4 150.7 /C2O1Si/8 74.0 74.5 74.0 74.0 72.8 77.7 /O1SiC3/8 65.1 64.2 67.3 66.2 61.0 60.3 /C2O1SiC3/8 0 0.6 20.4 0.3 0 0 Table 2 Main parameters of transition state 3a Method HF/6-31G* HF/6-3111G** MP2/6-31G* MP2/6-3111G** AM1/RHF AM1/CI = 8 Ea/kcal mol21 72.9 73.4 57.1 59.7 58.8 54.5 DrH/kcal mol21 2.02 21.1 1.25 28.3 228.3 224.4 rO1–C2/Å 1.236 1.233 1.263 1.255 1.254 1.267 rO1–Si/Å 2.035 2.066 2.048 2.321 2.324 2.362 rC3–Si/Å 2.431 2.465 2.321 2.455 2.454 2.436 /C2C3Si/8 62.9 63 67.4 71.5 71.7 70.2 /O1C2C3/8 153.9 155 147.0 150.4 150.4 155.5 /C2O1Si/8 79.0 79.0 78.4 77.0 77.0 72.9 /O1SiC3/8 64.1 63.2 67.1 61.1 60.9 61.3 /C2O1SiC3/8 0 0 0 0 0 0 Table 3 Main parameters of transition states 2b,c and 3b,c (AM1/RHF) 2b 3b 2c 3c Ea /kcal mol21 54.2 59.2 46.2 69.0 DrH/kcal mol21 220.5 220.5 213.9 213.9 rO1–C2/Å 1.262 1.251 1.264 1.254 rO1–Si/Å 2.320 2.324 2.392 2.421 rC3–Si/Å 2.482 2.463 2.589 2.535 /C2C3Si/8 68.9 71.8 67.9 71.6 /O1C2C3/8 155.0 150.0 157.7 153.4 /C2O1Si/8 75.1 77.5 75.7 76.2 /O1SiC3/8 61.0 60.7 58.7 58.8 /C2O1SiC3/8 0 0 0 0 Table 4 Main parameters of transition states 2d,e and 3d,e (AM1/RHF) 2d in-plane 2d out-of-plane 3d in-plane 3d out-of-plane 2e in-plane 2e out-of-plane 3e in-plane 3e out-of-plane Ea /kcal mol21 54.0 50.3 55.1 61.2 50.5 49.8 54.1 62.0 DrH/kcal mol21 226.0 226.0 226.0 226.0 225.1 225.1 225.1 225.1 rO1–Si/Å 2.360 2.357 2.358 2.360 2.329 2.363 2.361 2.358 rC2–Si/Å 2.362 2.358 2.405 2.401 2.362 2.363 2.408 2.402 rC3–Si/Å 2.516 2.520 2.485 2.493 2.555 2.534 2.489 2.495 /C2C3Si/8 68.5 68.2 71.6 71.2 67.0 67.8 71.6 71.2 /O1C2C3/8 156.8 157.2 151.5 152.2 157.3 157.8 151.6 152.1 /O1SiC3/8 60.2 60.2 60.0 60.0 60.0 60.0 60.0 60.0 /C2O1SiC3/8 20.4 0 0 20.6 0 0 0 20.8 /C8SiC2O1/8 2155.5 53.5 8.3 292.8 174.9 68.4 0 293.5 contribution of the same configuration only dropped to 94–95% and the activation energy remained quasi-stable (DEa < 4–5 kcal mol21).We then studied at a semiempirical level the effect of methyl substituents. Not surprisingly, the introduction of a methyl group on C3 has little effect on the activation energy of either reaction path.In contrast, the substitution of SiH3 for SiMe3 has a much greater impact on the activation energy of path (b) than of path (a); substitution on the silicon atom is clearly in favour of the retention path because of steric reasons presumably (Table 3). Fig. 1 Structure of transition states 2a (left) and 3a (right) (HF/6- 31G*) We also examined the influence of unsaturated substituents, e.g. vinyl and benzyl, on the competition between the retention vs.inversion of the silicon atom. It appears that regardless of the substituents or its position (in-plane or out-of-plane), the retention path remains the more favoured reaction path (Scheme 3, Table 4). In conclusion, we have shown that the 1,3-silyl shift that might be involved in the formation of silylketenes from alkoxyalkynes in the presence of trimethylsilyliodide, occurs with retention of configuration at the silicon centre.This finding is in agreement with recent calculations dealing with a 1,3-silyl sigmatropic shift in allylsilanes.13,14 Scheme 3 H O H H H O H R H O H H R H O R H H R H 2 in-plane 2 out-of-plane 3 in-plane 3 out-of-plane 8 8 2d, 3d R = vinyl 2e, 3e R = phenylJ. Chem. Soc., Perkin Trans. 2, 1997 1623 Acknowledgements We are grateful to Dr F. Volatron (CNRS, Orsay) and Professor D. Liotard (Université de Bordeaux) for useful discussions. We also thank the CCSJ (Université d’Aix-Marseille), IDRIS (CNRS-Orsay) and the Région ‘Provence-Alpes-Côte d’Azur’ for computational facilities and support. References 1 T.T. Tidwell, Ketenes, Wiley, New York, 1995. 2 L. L. Shchukovskaya, R. I. Pal’chilk and A. N. Lazarev, Dokl. Akad. Nauk. SSSR, 1965, 164, 357. 3 Ref. 1, ch. 4, pp. 348–368. 4 H. Sakurai, A. Shirahata, K. Sasaki and A. Hosomi, Synthesis, 1979, 740. 5 I. V. Efinova, M. A. Kazankova and I. F. Lutsenko, Zh. Obshch. Khim., 1985, 55, 1647. 6 A. Pommier, J.-M. Pons and P. J. Kocienski, J. Org. Chem., 1995, 60, 7334. 7 A. Pommier, J.-M. Pons, P. J. Kocienski and L. Wong, Synthesis, 1994, 1294. 8 G. A. Olah and S. C. Narang, Tetrahedron, 1982, 38, 2225. 9 J.-M. Pons, M. Oblin, A. Pommier, M. Rajzmann and D. Liotard, J. Am. Chem. Soc., 1997, 119, 3333; J.-M. Pons, A. Pommier, M. Rajzmann and D. Liotard, J. Mol. Struct. (THEOCHEM), 1994, 313, 361. 10 M. Oblin, J.-M. Pons and M. Rajzmann, Tetrahedron, 1997, 53, 8165. For a study of the parent reaction, without silicon, see: A. Moyano, M. A. Pericas, F. Serratosa and E. Valenti, J. Org. Chem., 1987, 52, 5532. 11 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman, T. A. Keith, G. A. Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-Latham, V. G. Zakrzewski, J. V. Ortiz, J. B. Foresman, J. Cioslowski, B. B. Stefanov, A. Nanayakara, M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Cheng, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. P. P. Stewart, M. Head-Gordon, C. Gonzalez, J. A. Pople, Gaussian 94, Revision C.2, Gaussian, Inc., Pittsburgh, PA, 1995. 12 M. J. S. Dewar, E. G. Zoebisch, E. F. Healy and J. P. P. Stewart, J. Am. Chem. Soc., 1985, 107, 3902. All calculations were performed with the AMPAC 6.0 code, Semichem, Shawnes, KS 66216, USA. 13 T. Yamabe, K. Nakamura, Y. Shiota, K. Yoshizawa, S. Kawauchi and M. Ishikawa, J. Am. Chem. Soc., 1997, 119, 807. 14 M. Takahashi and M. Kira, J. Am. Chem. Soc., 1997, 119, 1948. Paper 7/03439C Received 19th May 1997 Accepted 10th July 1997

ISSN:1472-779X    

DOI:10.1039/a703439c

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1625 EPR spectra and structure of the radical cations of fluorinated benzenes Akinori Hasegawa,a,* Yoshiteru Itagaki b and Masaru Shiotani b a Department of Chemistry, Kogakkan University, Ise-shi 516, Japan b Department of Applied Chemistry, Faculty of Engineering, Hiroshima University, Higashi-Hiroshima 739, Japan EPR spectra for the radical cations of a series of fluorinated benzenes, generated by irradiation with grays in halocarbon solid matrices, have been observed at low temperatures.The spectra consist of a hyperfine structure with axially symmetric anisotropy mainly due to fluorine nuclei. The observed spectra have been analysed by simulation. Ab initio calculations have been conducted for the cation radicals to obtain their optimized geometries. The results reveal that an unambiguous deformation in geometry is brought about by cationization in each case. INDO calculations have been performed for the optimized geometries of these radical cations to calculate the hyperfine couplings.The calculated couplings strongly support the observed ones. The symmetry of the SOMO for the radical cations resembles that of the HOMO of their neutral mother molecules. The deformed geometries of these radical cations suggests that in the process of releasing an electron from an HOMO, those chemical bonds with bonding nature in the HOMO become elongated and those bonds with antibonding nature become shortened.It is concluded that the structure and symmetry of the SOMO of these radical cations are affected not only by the number of substitutions by fluorine but also by the position of substitution. Introduction Radical cations of benzene formed in Freon (CCl3F) matrices at 4 K have been investigated by means of the EPR method.1 The EPR spectra of the cations observed at the same temperature suggest that the orbital degeneracy of e1g(D6h) for neutral benzene is removed by a static Jahn–Teller effect upon release of one electron and that the unpaired electron occupies a b3g(D2h) SOMO2,3 giving major hyperfine couplings to two H nuclei, as shown in Scheme 1.However, at elevated temperatures (higher than ca. 100 K) dynamic averaging arises resulting in a hyperfine structure (hfs) due to six equivalent H nuclei. Scheme 1 Schematic representation for the electronic state of the benzene cation. When the D6h benzene ring is distorted to a D2h structure by a static Jahn–Teller effect in the process of cationization, the highest occupied e1g orbitals, doubly degenerated, split into b1g and b3g orbitals.Thus, the SOMO is either the b1g orbital for the cation in an elongated D2h geometry or the b3g orbital for the cation in a compressed D2h geometry.3 The b3g SOMO has been strongly supported by the observed EPR spectra with large couplings to two H nuclei.1 The z axis is along the major C2 symmetry axis (dashed line) and the y axis is perpendicular to the ring plane.As regards the ionic radicals of fluorinated aromatic derivatives, a number of spectroscopic and theoretical studies have been carried out and their molecular and electronic structures have been discussed with reference to the substituent effect of fluorine.4 The ionization potentials (IPs) of these derivatives obtained by photoelectron spectroscopy gave information on the energy levels. IPs become higher and the stability of the molecules increases with an increase in the number of fluorines substituted for the hydrogens of benzene,5,6 as shown in Scheme 2.Together with the CNDO/S2 MO calculation results, they were discussed in terms of the ‘perfluoro effect’ 7 resulting from an ‘inductive (stabilization) effect’ for the s bonds of the fluorine atoms and a ‘plus inductive (destabilization) effect’ for the p electrons of the atoms.6 On the other hand, EPR can provide information on spin density distribution and suggest the symmetry of the SOMO which reflects the HOMO and the LUMO of the mother Scheme 2 Energy levels showing the first two ionization potentials from PE spectra with the symmetry of MOs for a series of fluorinated benzenes.5,6 The symmetry of MOs of S or A type depends on whether they are symmetric or antisymmetric under reflection in the plane passing through the C1 and C4 carbons, perpendicular to the ring.See Fig. 5 for the C1 and C4 carbons of o- and m-C6H4F2.1626 J.Chem. Soc., Perkin Trans. 2, 1997 molecules in the case of radical cations and anions, respectively, if the structure of the ionic radicals is similar to that of the mother molecules. These MOs are useful in providing information on, for example, the reactivity of the molecules as the Frontier Electron Theory teaches us.8 The EPR spectra of C6F6 1 generated in a c-C6F12 matrix have been recently investigated by the present authors,9 together with those of the C6F6 anion.The C6F6 1 spectrum at low temperatures consists of an axially symmetric triple quintet with A|| = 1.35 mT and A|| = 9.85 mT due to equivalent two and four 19F nuclei, respectively, and they were successfully interpreted in terms of a D2h structure in a 2B2g state with an elongated ring, although, in addition to the 2B2g state, a 2B3g state with a compressed ring was also suggested from MO calculations. 10 The D2h structure has three C2 symmetry axes, and the z axis was chosen to coincide with the axis perpendicular to the ring in the studies on the C6H6 1 and C6F6 1 cations.3,9 However, when the cations of a series of fluorobenzenes are investigated, it is better to take the z axis along the major C2 axis in the ring plane and to take the y axis perpendicular to the ring.For this choice, B3g(b3g) is unchanged, but B2g(b2g) is changed to B1g(b1g). This axial system is thus used in this paper. It is of interest to note that C6H6 1 has a compressed D2h structure with an unpaired electron in the b3g orbital, while C6F6 1 has an elongated D2h structure with an electron in the b1g orbital, in spite of the similar D6h structures of the neutral molecules.The radical cations of fluoro-substituted pyridines have been also studied by EPR spectroscopy.11 The symmetry of the SOMO for the cations was obtained from an analysis of the hfs due to 19F nuclei with axially symmetric anisotropy. A preliminary EPR study has been carried out for the radical cations of a series of fluoro-substituted benzenes,12 but the details have not yet been reported.Therefore, in this study, full details of the study will be presented, together with the results of MO calculations for the radical cations by ab initio and semiempirical methods. Experimental Commercially available reagents were used for a series of fluorinated benzenes. Solutions containing ca. 1 mol% of these fluorinated benzenes in either fluorotrichloromethane (CCl3F) or perfluorocyclohexane (c-C6F12) were prepared in Spectrosil EPR sample tubes on a vacuum line.The samples were irradiated with g-rays from a 60Co source at 77 K, the typical total absorption dose being ca. 1 Mrad. EPR measurements were carried out with a JEOL JES-PX-1X or Bruker ESP-300E X-band spectrometer operating at 100 kHz modulation and at variable temperatures using an Oxford continuous flow cryostat ESR 900. The field strength was measured using a Bruker EP 035M NMR Gaussmeter. Ab initio calculations were carried out on a Cray J932/24 computer at the Information Processing Center, Hiroshima University, with 3-21 G basis sets, using the GAUSSIAN 94 program,13 to obtain optimized geometries for the radical cations of the fluorinated benzenes.In order to calculate spin densities and hf couplings for atoms in the radical cations, INDO calculations 14 were performed for the geometries optimized by the ab initio method.Results and discussion EPR spectra and their assignments for the cations The radical cations of a series of fluorinated benzenes were detected by the EPR method using either CCl3F (IP1 = 11.8 eV)15 or c-C6F12 (IP1 = 12.9 eV) 16 as matrices at 77 K. The c-C6F12 matrices are more effective for heavily fluorinated benzenes because their IPs increase with an increase in the number of substituted fluorines. The observed anisotropic EPR spectra show well-defined features for the parallel components of axially symmetric hf coupling to 19F nuclei and are photobleached by visible light.The C6H5F1 cation. The unambiguous parallel components of a double doublet were observed in the outermost regions of the EPR spectrum of an irradiated solid solution of C6H5F in CCl3F observed at 93 K, as shown in Fig. 1(A). The main doublet may be assigned to the 19F nucleus and the additional one to the 1H nucleus in the para-position of C6H5F. Thus the observed spectrum may suggest the formation of the radical cation of C6H5F.Spectral simulations were performed for an axially symmetric g tensor, an axially symmetric 19F hf tensor for the main doublet and an almost isotropic tensor for the additional doublet with a smaller coupling due to one 1H, using a second-order treatment. The spectrum simulated with the EPR parameters listed in Table 1 is shown in Fig. 1(A). The discrepancy between the observed and the simulated spectra may result mainly from the fact that the signals of some radicals probably formed from the CCl3F matrix appear rather strongly in the central part of the observed spectrum for this system.Similar EPR spectra were observed for an irradiated solid solution of C6H5F in c-C6F12. Hf coupling to 19F for the cation in the c-C6F12 matrix is smaller than that in the CCl3F matrix, as shown in Table 1. The o-C6H4F2 1 cation. The spectrum of an irradiated solid solution of o-C6H4F2 in a CCl3F matrix showed parallel features with small additional splittings in the wing regions, as shown in Fig. 1(B). The main features were thought to be the outermost parallel components of a 1:2:1 triplet due to two equivalent 19F nuclei and the additional splittings to be a double triplet. Judging from the symmetry of o-C6H4F2, the additional triplet may originate from the two 1H nuclei in the 3,6- or 4,5- positions, but the doublet is difficult to assign to one 1H nucleus in the o-C6H4F2 molecule.The doublet may thus be assigned to the 19F nucleus of CCl3F, as reported in the case of e.g. C2F4 1.17,18 For the assignment of the two H nuclei, the theoretical calculations mentioned below may be helpful. Simulation with the parameters for o-C6H4F2 1 in Table 1 gave satisfactory results, see Fig. 1(B). The m-C6H4F2 1 cation. Parallel features with a well-defined triplet were observed at the wings of the spectra of irradiated solid solutions of m-C6H4F2 in both CCl3F [Fig. 2(A)] and c-C6F12 matrices. The main features may be attributed to a triplet from two equivalent 19F nuclei and the additional triplet to the 1H nuclei in the 4,6-positions of the m-C6H4F2 cation. Fig. 1 First derivative X-band EPR spectra for solid solutions of ca. 1 mol% of fluorobenzene (A) and o-difluorobenzene (B) in trichloro- fluoromethane after irradiation at 77 K, observed at 93 K (A) and 77 K (B), and the lower spectra (dashed lines) are those simulated using the EPR parameters for the corresponding radical cations in Table 1.The signals marked * in (A) correspond to the perpendicular components of the anisotropic hfs. In the observed spectrum, one of the signals is partly masked by signals from matrix radicals.J. Chem. Soc., Perkin Trans. 2, 1997 1627 Table 1 EPR parameters for the radical cations of fluorinated benzenes Radical A/mT Assigned position of cation Matrix CCl3F CCl3F c-C6F12 CCl3F CCl3F c-C6F12 CCl3F c-C6F12 c-C6F12 c-C6F12 c-C6F12 c-C6F12 c-C6F12 T/K 77 4.2 93 77 77 77 77 77 84 140 77 77 93 10 g 2.0028 2.0027 g|| = 2.0032 g^ = 2.0060 g|| = 2.0036 g^ = 2.0051 g|| = 2.003 g^ = 2.008 g|| = 2.002 g^ = 2.006 g|| = 2.002 g^ = 2.003 g|| = 2.002 g^ = 2.006 g|| = 2.003 g^ = 2.004 g|| = 2.002 g^ = 2.006 g|| = 2.002 g^ = 2.006 g|| = 2.002 g^ = 2.006 g|| = 2.002 g^ = 2.006 A|| 15.0 1.0 12.0 1.0 12.0 0.8 0.5 9.6 1.0 9.3 1.0 15.6 5.0 15.0 12.7 6.4 4.94 16.9 12.7 2.8 0.4 10.9 0.5 10.8 10.4 (2)1.1 1.4 9.9 A^ a 1.0 1.2 1.0 1.2 0.3 0.4 ca. 0 0.7 ca. 0 0.7 ca. 0 0.3 ca. 0 0.7 0.7 0.7 ca. 0 0.3 0.3 0.3 ca. 0 0.3 ca. 0 1.3 1.3 (2)0.3 0.0 0.0 aa,b (2)0.4 (2)0.8 (2)0.2 5.7 1.1 4.7 1.1 4.2 0.5 0.2 3.7 0.3 3.6 0.3 5.4 0.2 5.5 4.7 2.6 0.6 1.7 0.7 5.8 4.4 1.1 0.1 3.8 0.2 4.5 4.3 (2)0.6 0.5 3.3 r2pp(19F) c 0.087 0.068 0.072 0.055 0.053 0.092 0.088 0.074 0.035 0.029 0.101 0.076 0.018 0.064 0.059 0.055 20.006 20.009 0.061 19F and 1H 6H H1.4 H2.3.5.6 F1 H4 F1 H4 F1.2 H4.5 F of CCl3F F1.3 H4.6 F1.3 H4.6 F1.4 4H F1 F4 F2 H5 F1.3.5 H2.4.6 F1 F4 F2.6 2H F1.2.4.5 2H F2.6 F3.5 F1 F1.4 F2.3.5.6 Ref. 1 present work present work present work present work present work present work present work present work present work present work present work 9 a Error limit is ±0.3 mT. b Isotropic hf splitting. c Spin density in 2pp(19F) orbital evaluated using the magnetic parameters listed by Goodman and Raynar.20 The spectrum simulated with the parameters for the cation in Table 1 may explain the observed spectrum, as can be seen in Fig. 2(A). Fig. 2 First derivative X-band EPR spectra for solid solutions of ca. 1 mol% of m-difluorobenzene (A) and p-difluorobenzene (B) in trichloro- fluoromethane after irradiation at 77 K, observed at 77 K, and the lower spectra (dashed lines) are those simulated using the EPR parameters for the corresponding radical cations in Table 1 The p-C6H4F2 1 cation. Poorly resolved parallel features were observed in the spectrum of an irradiated solid solution of p-C6H4F2 in CCl3F, as shown in Fig. 2(B). The triplet may be due to two 19F nuclei. Judging from the molecular symmetry, the additional splittings may result from four equivalent 1H nuclei of the p-C6H4F2 cation. The coupling constant was estimated by simulation, as shown in Fig. 2(B). The 1,2,4-C6H3F3 1 cation. An EPR spectrum consisting of many well-defined lines was obtained for an irradiated solid solution of 1,2,4-C6H3F3 in c-C6F12, as shown in Fig. 3(A). This spectrum was almost perfectly reproduced by the simulation performed using a g-tensor and hf tensors for three different doublets, the g and the hf tensors having the same principal direction, and an isotropic tensor for a doublet with a small coupling, as can be seen from Fig. 3(A). The former three doublets may be due to the 19F nuclei of 1,2,4-C6H3F3 1 and the other doublet with the small coupling to a 1H of the cation. Theoretical calculations have to be waited for more detailed assignments.The 1,3,5-C6H3F3 1 cation. Only a broad signal was observed in the spectrum of an irradiated solid solution of 1,3,5-C6H3F3 in c-C6F12 measured at 77 K. The line-width was hardly reduced, even by cooling to 4 K, but did decrease on annealing. Measurement at 140 K gave a spectrum with poorly resolved features. This spectrum may be assigned to the 1,3,5-C6H3F3 11628 J. Chem. Soc., Perkin Trans. 2, 1997 cation which has a structure distorted by the Jahn–Teller effect but is thermally averaged at this temperature, in a similar manner to C6F6 1.This spectrum was tentatively interpreted in terms of three equivalent 19F with a large coupling and three equivalent 1H with a small coupling, listed in Table 1. The 1,2,4,6-C6H2F4 1 cation. The spectrum of an irradiated solid solution of 1,2,4,6-C6H2F4 in c-C6F12 consists of parallel lines attributable to two doublets with different large couplings and a triplet with a smaller coupling, as shown in Fig. 3(B). The symmetry of the molecule suggests that the triplet results from the two equivalent 19F nuclei in the 2,6-positions. The assignment of the two different doublets may be suspended until theoretical calculations. This spectrum was satisfactorily reproduced by simulation with the parameters listed in Table 1, as shown in Fig. 3(B). The 2,3,5,6-C6H2F4 1 cation. An axially symmetric anisotropic spectrum of a quintet was observed for an irradiated solid solution of 2,3,5,6-C6H2F4 in c-C6F12, as shown in Fig. 4(A). The quintet may be due to the four equivalent 19F nuclei of the 2,3,5,6-C6H2F4 cation. This spectrum was perfectly reproduced by simulation for the parameters in Table 1, as can be seen in Fig. 4(A). The C6HF5 1 cation. The spectrum observed for an irradiated solid solution of C6HF5 in c-C6F12 seems to be attributable to Fig. 3 First derivative X-band EPR spectrum for solid solutions of ca. 1 mol% of 1,2,4-trifluorobenzene (A) and 1,2,4,6-tetrafluorobenzene (B) in perfluorocyclohexane after irradiation at 77 K, observed at 84 K (A) and 77 K (B), and the lower spectra (dashed lines) are those simulated using the EPR parameters for the corresponding radical cations in Table 1 Fig. 4 First derivative X-band EPR spectrum for solid solutions of ca. 1 mol% of 2,3,5,6-tetrafluorobenzene (A) and pentafluorobenzene (B) in perfluorocyclohexane after irradiation at 77 K, observed at 77 K (A) and 93 K (B), and the lower spectra (dashed lines) are those simulated using the EPR parameters for the corresponding radical cations in Table 1 a double quintet.However, simulation has revealed that instead of a quintet, two triplets with slightly different couplings listed in Table 1 gave a more satisfactory result, as shown in Fig. 4(B). The doublet with a small coupling may be assigned to the central 19F in the 1-position (para-position with respect to H). Assignment of the two different triplets must also be suspended. The C6F6 1 cations.The authors have already reported on the EPR spectra and structure of the radical cation, C6F6 1, in connection with those of the radical anion, C6F6 2.9 Thus, a summary of the results for C6F6 1 is given here in order to enable us to compare it with the radical cations under investigation. The EPR spectra of the C6F6 1 cation generated in a solid solution of c-C6F12 at 77 K dramatically depended upon the temperature of observation. The 170 K spectrum clearly showed seven equally spaced hyperfine lines due to six equivalent F nuclei, giving the EPR parameters of A|| = 6.77 mT and A^ = ca. 0 mT for the six nuclei and g|| = 2.0020 and g^ = 2.0060. With decreasing temperature, the inner hyperfine lines became broader but the outermost lines remained unchanged. With further decreasing temperature, the outermost bands became sharper and then they split into three lines with an equal splitting of 1.35 mT at 10 K. Thus, the six fluorines of C6F6 1 can be divided into two groups: one is the two equivalent fluorines A|| 1.35 mT, and the other is four equivalent fluorines A|| 9.85 mT.The parent C6F6 molecule has doubly degenerate HOMOs, but the orbital degeneracy is removed in the process of cationization. The symmetry of the SOMO of the radical cation can be determined by the HOMO from which one electron is released. Ab initio calculations for the cation suggested that two distorted D2h structures are possible: an elongated ring structure with a b1g SOMO and a compressed ring structure with a b3g SOMO.3,10 From the character of the SOMOs, a large splitting due to two equivalent F nuclei can be expected from the compressed structure, while a small splitting can be expected from the elongated one.The observed wing features of the triplet with a coupling of 1.35 mT may strongly suggest that the cation has a b1g SOMO and an elongated D2h structure. It may be of interest to note that C6F6 1 takes the elongated D2h structure with the b1g SOMO, whereas C6H6 1 takes the compressed D2h structure with the b3g SOMO.Ab initio calculations gave two results.3 At the UHF level, the elongated structure is 2.0 kcal mol21 more stable than the compressed structure for both C6H6 1 and C6F6 1. On the other hand, at the level including electron correlation for the p electrons by means of second-order Møller-Plesset perturbation,19 the compressed structure for C6H6 1 and the elongated structure for C6F6 1 are more stable, although the two distorted structures are within 0.1 kcal mol21 of each other in both cases.3 Spin densities in the radical cations The anisotropic hyperfine couplings to 19F nuclei listed in Table 1 gave all of the radical cations studied a relation of A||(19F) @ A^(19F) 6 0, which is typical of p-radicals.17 The spin densities in the 2pp orbitals of the 19F atoms can be obtained from the experimental anisotropic hyperfine couplings of A|| and A^ and the theoretical anisotropy in the hyperfine coupling of a 19F nucleus, 2B0 = 108.5 mT.20 Experimental A|| values were determined rather accurately from observed spectra, whereas A^ values were estimated approximately by simulation.In addition, the sign of the values for A^ is not known since the cases of both positive and negative signs were reported for the values of 19F in the a-positions.21 Therefore, both the cases were taken into account but positive signs were tentatively adopted for the calculation of spin densities for the radical cations.The spin densities thus obtained are also given in Table 1. Moreover, the fact that the principal directions of the 19F couplings coincide with one another in a molecule may also suggest planarity in the structure of the radical cations.J. Chem. Soc., Perkin Trans. 2, 1997 1629 Fig. 5 (A) Optimized geometrical structures for the radical cations of (1) C6H5F, (2) o-C6H4F2 and (3) m-C6H4F2 calculated with the GAUSSIAN 94 program at the UHF/3-21G level.The bond lengths are given in Å. (B) The experimental isotropic (aiso) and anisotropic (2B) hf coupling constants to 1H and 19F nuclei (in mT) are compared with the theoretical ones (in parentheses) evaluated by the INDO MO method for the optimized structures. The arrows (�Æ) and (Æ�) stand for the elongated and compressed, respectively, C–C bonds parallel to the axes passing through the C1 and C4 atoms.Major C2 symmetry axes are shown with dashed lines. Note that for the irreducible representation of the SOMO, an axial system was chosen to have the z axis along the major C2 symmetric axis and the y axis perpendicular to the ring, but that for S and A types, reflection was used in the plane passing through the C1 and C4 carbons, perpendicular to the ring. A9 stands for A-like symmetry. Spin densities and hf coupling constants calculated by INDO The EPR data will be discussed at a semiempirical level because high quality MO calculations on such high symmetry openshell systems as these radical cations sometimes lead to pitfalls. 22 In advance of the semiempirical INDO calculations, however, the optimized geometries of these fluorinated benzene cations were obtained by ab initio calculations. Although we have to refrain from quantitative discussion of the results, we may be allowed to qualitatively speculate that a remarkable deformation was brought about by cationization in each case, as shown in Figs. 5(A)–7(A). As for the 1,3,5-C6H3F3 1 cation, the geometry has not yet been optimized, probably because the cation is distorted by the Jahn–Teller effect. This radical cation will therefore be excluded in the following discussion. INDO calculations were carried out for the geometries optimized by the ab initio method. Isotropic hf couplings to 1H and 19F were acquired from the s-spin densities on their nuclei obtained by INDO calculations and the atomic hf constants aiso 0 adjusted for INDO calculations, 53.986 and 4482.920 mT, respectively.14 Anisotropic hf couplings, 2B, to 19F nuclei were obtained from the p-spin densities on the nuclei and the atomic value calculated, 2B0 = 108.6 mT.20 These calculated values are shown with the observed values in Figs. 5(B)–7(B). A comparison between the observed and calculated values strongly supported the assignments, directly made from the observed EPR spectra, not only to 19F of C6H5F1, C6H4F2 1 and 2,3,5,6-C6H2F4 1 but also to 1H in the para-position of C6H5F1, 1H in the 4,6-positions of 1,3-C6H4F2 1, 19F in the 2,6-positions of 1,2,4,6-C6H2F4 1 and 19F in the 1-position (para-position with respect to H) of C6HF5 1.Moreover, as for the assignment of the 1:2:1 triplet to 1H of the 3,6- or 4,5-positions of o- C6H4F2 1, it was concluded that the 4,5-positions were more reasonable. The four different doublets observed for 1,2,4- C6H3F3 1, which could not be identified from the EPR spectrum alone, were assigne each of the 19F nuclei in the 1,2,4- positions and to 1H in the 5-position.Also, couplings to 19F in the 1,4-position in 1,2,4,6-C6H2F4 1 and in the 2,6- and 3,5- positions of C6HF5 1 were finally identified with the help of the calculated results. The observed and assigned coupling constants are shown in Table 1 and the observed and calculated values for isotropic and anisotropic couplings are shown in Figs. 5(B)–7(B). For these cations, the observed isotropic aiso(19F) and anisotropic 2B(19F) couplings are smaller than those calculated. However, good correlations were obtained between the observed aiso(19F) and calculated s-spin density on 19F and between the observed 2B(19F) and the calculated p-spin density on 19F, as shown in Fig. 8. The observed couplings (Y) were related to the calculated spin densities (X) by the linear function Y = CX, in both cases of aiso(19F) and 2B(19F), the constant C being 23.8 × 102 and 72.3 mT for aiso(19F) and 2B(19F), respectively.The former is about one half of the atomic coupling of 4482.920 mT adjusted using only nine data for INDO calculations 14 and the latter is rather smaller than the calculated 2B0 value of 108.6 mT.20 This 2B0 value has not been authorized for INDO calculations, but 50 mT was used as a parameter in the simulation of the EPR spectra of the C6F6 anion radical.9 In addition, one may remark that the values of A^ are not directly determined from the EPR spectra but are estimated by simulation and that the signs were not determined but were recorded as positive.When these C constants obtained were used, the calculated isotropic and anisotropic hf couplings coincide with the observed ones with very high correlation coefficients R of 0.94 and 0.95, respectively. The structure of the radical cations Structural information on the radical cations investigated is already shown in the hf coupling data in Figs. 5(B)–7(B). This consists of the symmetry label of the irreducible representation of the SOMO, the type of SOMO labelled S or A in parentheses, the molecular point group and the arrows standing for the bonds elongated or compressed. These results may not have quantitative meaning, but they may lead us to the following qualitative understanding. The symmetry of the SOMO in each radical cation resembles that of the HOMO in the corresponding neutral molecule.All of the deformed geometries of these radical cations under investig-1630 J. Chem. Soc., Perkin Trans. 2, 1997 Fig. 6 (A) Optimized geometrical structures for the radical cations of (1) p-C6H4F2, (2) 1,2,4-C6H3F3 and (3) 1,2,4,6-C6H2F4 calculated with the GAUSSIAN 94 program at the UHF/3-21G level. The other caption for this figure can be seen in Fig. 5. S9 stands for S-like symmetry. ation imply that in the process of releasing one electron from an HOMO, the chemical bonds giving the bonding nature in the HOMO become elongated and the bonds giving the antibonding nature become shortened.The symmetry of the SOMOs of the radical cations is affected not only by the number of substitutions by fluorine but also by the position of substitution, and whether symmetry is of type S or type A is determined by the balance between the number of fluorines in the 1,4- and the 2,3,5,6-positions. Acknowledgements The authors thank Messrs.H. Kawazoe and H. Toshiro for their assistance in the preliminary stage of this study. The present research was partially supported by the JSPS Program Fig. 7 (A) Optimized geometrical structures for the radical cations of (1) 2,3,5,6-C6H2F4 and (2) C6HF5 calculated with the GAUSSIAN 94 program at the UHF/3-21G level. The other caption for this figure can be seen in Fig. 5. for supporting University-Industry Cooperative Research Project and the Subsidy for Scientific Research of the Ministry of Education in Japan (Grant No. 08240105). Fig. 8 The correlation between the observed hf coupling constants to 19F nuclei and the spin densities calculated by the INDO method for the ab initio optimized geometries. (A) and (B) are the plots for isotropic hf (aiso) vs. 2s spin densities and for anisotropic hf (2B) vs. 2p spin densities, respectively. Constants (C), i.e. the slopes of the fitted lines, and correlation coefficients (R), are given in the Figure.J.Chem. Soc., Perkin Trans. 2, 1997 1631 References 1 M. Iwasaki, K. Toriyama and K. Nunome, J. Chem. Soc., Chem. Commun., 1983, 320. 2 The notation of b2g(D2h) was used in ref. 1. However, the axial system described in the caption of Scheme 1 and used in this paper gives b3g(D2h). Also, see the text. 3 K. Raghavachari, R. C. Haddon, T. A. Miller and V. E. Bondybey, J. Chem. Phys., 1983, 79, 1387. 4 K. S. Chen, P. J. Krusic, P. Meakin and J. K. Kochi, J. Phys. Chem., 1974, 78, 2014; P.J. Krusic, K. S. Chen, P. Meakin and J. K. Kochi, J. Phys. Chem., 1974, 78, 2036; P. J. Krusic and P. Meakin, J. Am. Chem. Soc., 1976, 98, 228; J. T. Wang and F. Williams, Chem. Phys. Lett., 1980, 71, 471; Y. N. Molin and B. A. Anisirnov, Rad. Phys. Chem., 1983, 21, 77; M. B. Yim and D. E. Wood, J. Am. Chem. Soc., 1976, 98, 2053; M. B. Yim, S. DiGroegorie and D. E. Wood, J. Am. Chem. Soc., 1977, 99, 4260; C. R. Brundle, M. B. Robin, N. A. Kuebler and H. Basch, J.Am. Chem. Soc., 1972, 94, 1451; C. R. Brundle, M. B. Robin and N. A. Kuebler, J. Am. Chem. Soc., 1972, 94, 1466. 5 I. D. Clark and D. C. Frost, J. Am. Chem. Soc., 1967, 89, 244; J. R. Frazier, L. G. Carter and H. C. Schweinler, J. Chem. Phys., 1978, 69, 3807; J. A. Sell and A. Kupperman, Chem. Phys., 1978, 33, 367; K. D. Jordan and P. D. Burrow, J. Chem. Phys., 1979, 71, 5384; L. G. Christophorou and H. C. Schweinler, J. Chem. Phys., 1979, 71, 5385 6 C. B. Duke, K. L. Yip, G.P. Ceasar, A. W. Potts and D. G. Streets, J. Chem. Phys., 1977, 66, 256. 7 See, for example: J. W. Rabalais, Principles of Ultraviolet Photoelectron Spectroscopy, Wiley, New York, 1977; D. W. Turner, C. Baker, A. D. Baker, and C. R. Brundle, Molecular Photoelectron Spectroscopy, Wiley, New York, 1970; C. R. Brundle, M. B. Tobin and N. A. Kaubler, J. Am. Chem. Soc., 1972, 94, 1466. 8 K. Fukui, T. Yonezawa and H. Shingu, J. Chem. Phys., 1952, 20, 722. 9 A. Hasegawa, M. Shiotani and Y. Hama, J. Phys. Chem., 1994, 98, 1834. 10 K. Hiraoka, S. Mizuse and S. Yamabe, J. Phys. Chem., 1990, 94, 3689. 11 M. Shiotani, H. Kawazoe and J. Shoma, J. Phys. Chem., 1984, 88, 2220. 12 M. Shiotani, H. Kawazoe and J. Sohma (a) Proceedings of the 7th International Congress of Radiation Research, A1-44, Martinus Nijhoff Publishers, Amsterdam, 1983; (b) Proceedings of the 22nd Japanese ESR Symposium, 1983, p. 127. 13 J. B. Foresman and AE. Frisch, Exploring Chemistry with the Electronic Structure Method, 2nd edn., Gaussian, Inc., Pittsburgh, PA, 1996. 14 J. A. Pople and D. L. Beveridge, Approximate Molecular Orbital Theory, McGraw-Hill, New York, 1970. 15 T. Shida, Y. Nosaka and T. Kato, J. Phys. Chem., 1978, 82, 695. 16 S. Katsumata and M. Shiotani, unpublished data. 17 A. Hasegawa and M. C. R. Symons, J. Chem. Soc., Faraday Trans. 1, 1983, 79, 93. 18 G. W. Eastland, D. N. R. Rao, J. Rideout, M. C. R. Symons and A. Hasegawa, J. Chem. Research (S), 1983, 258. 19 C. Møller and M. S. Plesset, Phys. Rev., 1934, 46, 618; J. S. Binkley and J. A. Pople, Int. J. Quantum Chem., 1975, 9, 229. 20 B. A. Goodman and J. B. Raynor, Adv. Inorg. Chem. Radiochem, 1970, 13, 135. 21 M. Iwasaki, S. Noda and K. Toriyama, Molec. Phys., 1970, 18, 201; M. Iwasaki, Molec. Phys., 1971, 20, 503. 22 The authors thank one of the referees for this comment. Paper 7/03093B Received 6th May 1997 Accepted 22nd May 1997

ISSN:1472-779X    

DOI:10.1039/a703093b

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1633 The proportion of 1,3-migration of a methyl group in the reactions of the iodide (Me3Si)3CSi(CD3)2I with silver salts in alcohols. Mechanistic implications Jarrett R. Black,a Colin Eaborn,b Philip M. Garrity a and Duncan A. R. Happer *,a a Department of Chemistry, University of Canterbury, Christchurch, New Zealand b School of Chemistry, Physics and Environmental Science, University of Sussex, Brighton, UK BN1 9QJ In the reactions of the isotopically labelled (Me3Si)3CSi(CD3)2I, 1-D, with alcohols ROH (R = Me, Et, Pri or But ) in the presence of silver salts AgX (X = ClO4, O3SCF3, O2CCF3, NO3 or BF4) the ratio of the rearranged product (Me3Si)2C[Si(CD3)2Me]SiMe2OR to unrearranged (Me3Si)3CSi(CD3)2OR always falls below the value of 1 :1 expected for capture of an intermediate methyl-bridged cation [(Me3Si)2CSiMe2-Me-Si(CD3)2]1, II, by nucleophilic attack of the alcohol at either end of the bridge.For example, in the reaction with AgClO4 the percentages of rearranged product are 26, 22 and 37 (all values ±8) for R = Me, Et and Pri, respectively, at room temperature and the corresponding figures at the reflux temperature are 29, 40, 41 and 44 (all ±8) for Me, Et, Pri and But.Thus the reaction cannot proceed exclusively by direct formation of a fully free cation II as assumed in the simplest picture. It is concluded that ad hoc modification of this simple picture, for example, by postulating some pre-association of an alcohol molecule at the Si]I bond in 1-D, is preferable to possible alternative mechanisms, such as one involving initial formation of an unbridged cation [(Me3Si)3CSi(CD3)2]1 which can sometimes be captured before conversion into II. The alkoxide products are accompanied by some (Me3Si)3CSi(CD3)2X and (Me3Si)2C[Si(CD3)2Me]SiMe2X (from AgBF4, X = F in both cases) and the corresponding hydroxides, the proportion of rearranged product always being significantly below 50%.In the reactions with AgClO4 or AgO3SCF3 the silicon hydroxides are mainly produced by hydrolysis of initially formed perchlorates or trifluoromethanesulfonates by traces of water in the solvents, but for the other silver salts they arise virtually wholly by capture of the intermediate cation by the water. The type of rearrangement with which this paper is concerned 1 was first observed when the iodide TsiSiPh2I, where Tsi denotes the very bulky tris(trimethylsilyl)methyl (‘trisyl’) group, was treated with silver salts AgX or mercury(II) salts HgX2 in MeOH and found to give mixtures of rearranged products (Me3Si)2C(SiPh2Me)SiMe2X and (Me3Si)2C(SiPh2Me)SiMe2- OMe, an Me group having undergone 1,3-Si-to-Si migration.2,3 Subsequent studies showed that some related iodides, Tsi- SiRR9I, under similar conditions gave mixtures of rearranged and unrearranged products.3,4 The same type of rearrangement, complete in the case of TsiSiPh2I, was found to take place in the reactions with (i) silver salts in inert solvents such as CH2Cl2;5 (ii) ICl in CCl4 [e.g.Tsi- SiEt2I gave TsiSiEt2Cl and (Me3Si)2C(SiEt2Me)SiMe2Cl in ca. 35 : 65 ratio 6]; (iii) CF3CO2H;3 (iv) (in the case of TsiSiPh2I and TsiSiPhMeI) MeOH or PhNH2 under UV irradiation;7 and (v) m-chloroperoxybenzoic acid in MeOH.8 It was suggested that in each case the reaction involved rate-determining formation of a methyl-bridged cation of type I, which could be rapidly attacked by a nucleophile either at the a-site to give an unrearranged product or at the g-site to give a rearranged product.3,9 It was further suggested that in the absence of significant electronic effects of the substituents R9 and R0 the proportion of rearranged product would be determined by the relative degrees of steric hindrance at the a- and g-sites in the cation towards nucleophilic attack, so that, as observed,3 more than 50% of rearranged product would be expected from say TsiSiPhMeI, and very predominantly rearranged product from TsiSiPh2I.10 That the ratio of attack at the a- and g-sites is not determined solely by steric effects was shown by the finding that in the reactions with ICl or silver salts the iodides TsiSiMeYI with Y = Cl or F, gave only rearranged products (although the F atom, in particular, should have a substantially smaller steric effect than an Me group) and TsiSiMe(OMe)I gave exclusively unrearranged products (although OMe would be expected to have a steric effect similar to or slightly larger than an Me group).9 Furthermore the nature of the solvent was found to have a substantial influence on the extent of rearrangement; e.g., the reaction of TsiSiPhHI or TsiSiMeHI with AgBF4 gave only rearranged fluoride in Et2O but a mixture of rearranged and unrearranged fluorides in CH2Cl2.11, † In terms of the simplest picture of the mechanism of the rearrangement, involving formation of a bridged cation of type I that is fully free before being captured by a nucleophile, then in the case of TsiSiMe2I, 1, there should be equal amounts of rearranged and unrearranged products.In order to confirm this we prepared the labelled iodide TsiSi(CD3)2I, 1-D, and first examined its reactions with silver salts AgX in a range of aprotic solvents, which we assume involve intermediate formation of cation II.12 To our surprise the percentage, PR, of rearranged product (Me3Si)2C[Si(CD3)2Me]SiMe2X, formed along with the unrearranged TsiSi(CD3)2X, was in most cases significantly below the expected value of 50, and sometimes below 10, and varied markedly with the solvent.‡ These findings were tentatively attributed to preferential attack at the a-site by † In the reaction of a wide range of compounds TsiSiR9R0I with AgBF4, the proportion of rearranged fluoride has been shown to vary widely with the nature of the solvent and the electronic effects of R and R9 (S.M. Whittaker, PhD Thesis, University of Salford, 1993; P. D. Lickiss, personal communication). ‡ Values of PR > 50 were observed in the reactions with AgO3SCF3 and AgO2CCF3 in CH2Cl2, perhaps because of some return of I2 to the gsite of the bridged cation (as well as to the a-site) before complete separation of AgI.1634 J. Chem. Soc., Perkin Trans. 2, 1997 Table 1 Yields of products TX, TOH and TOR from reactions of Tsi(CD3)2I, 1-D, with alcohols ROH in the presence of AgX, at room temperature or under reflux, and proportion, PR (%), of rearranged product, with estimated uncertainty (±) denoted by the subscript.a Where there is no entry for PR the value could not be even approximately determined.Room temperature Under reflux X In MeOH ClO4 Otf b NO3 Otfac BF4 In EtOH ClO4 Otf b NO3 Otfac BF4 In PriOH ClO4 Otf b NO3 Otfac BF4 In ButOH ClO4 Otf b NO3 Otfac BF4 TX 3 6 14 7 11 1–7 3–30 36 21 52 0 39 78 46 81 <2 10 62 25 83 PR 148 178 83 103 158 148 125 53 63 53 43 23 53 53 <2 23 <23 53 TOH 11 6 1 2 11 35–40 12–35 <2 2 3 75 42 5 4 3 97 87 33 71 17 PR 168 95 105 73 125 63 123 73 <23 <2 <23 <2 TOR 86 88 85 91 77 52 60 64 76 45 25 19 17 50 16 3 2 5 7 <2 PR 268 225 225 285 215 228 165 225 205 295 378 255 235 408 TX <2 5 27 7 14 <2 27 44 21 46 <2 28 64 39 70 0–20 47 79 42 77 PR 218 93 103 258 145 65 173 198 145 73 83 178 205 185 73 175 208 TOH 8 7 <2 <2 7 34 4 <2 2 3 70–80 32 4 4 3 70–95 35 6 22 17 PR 188 188 155 135 11 185 175 135 248 TOR 92 87 73 92 79 66 68 55 77 51 20–30 40 32 57 27 8 14 15 36 6 PR 298 388 338 315 358 40 338 305 275 338 418 368 305 448 448 488 228 308 a TY (Y = X, OH or OR) denotes the mixture of (Me3Si)3CSi(CD3)2Y and (Me3Si)2C[Si(CD3)2Me)]SiMe2Y.b Otf = O2CCF3. c Otfa = O3SCF3. X2 liberated from an ion-pair [Ag1 ? ? ?X2] near that site by formation of AgI. Subsequently reaction of 1-D with ICl in aprotic solvents was found to give less than 50% of rearranged chloride, and this was correspondingly attributed to preferential transfer of Cl2 to the a-silicon centre in the cation from the formed I2Cl2 before it diffused away from that site.13 However a PR value of only 16 was found for the methoxides formed alongside the chlorides in the corresponding reaction in MeOH, and that presented a greater problem, consideration of which was deferred until presentation of the results of the study described below involving reactions of 1-D with silver salts in alcohols. In alcohols, not only should the good solvating ability of the medium be more likely to enable the intermediate to become fully free before taking up the nucleophile but, furthermore, in the formation of alkoxide products the relevant nucleophile, the solvent, should be equally available at the a- and g-sites, giving rise to values of PR of 50.We thus determined the values of PR for formation of the alkoxide [and also for formation of TsiSi(CD3)2Y and its isomer where Y = X or OH] from a variety of silver salts AgX in MeOH, EtOH, PriOH and ButOH (all dried by standard methods but evidently containing traces of water).The results are shown in Table 1. In this table and in the discussion below, for convenience the symbol TY, where Y = X, OH or OR, is used to denote the mixture of unrearranged and rearranged products, TsiSi(CD3)2Y and (Me3Si)2C[Si- (Me3Si)3CSiMe2I 1 (Me3Si)3CSi(CD3)2I 1-D Me Me2 Si R2C Si R¢R¢¢ Me Me2 Si R2C Si (CD3)2 Z H2 Si H2C Si H2 I II III R = Me3Si a g + a g + a g + (CD3)2Me]SiMe2Y. It should also be noted that the products TX formed from AgBF4 are always actually the fluorides.As discussed in the Experimental section, the reliability of the PR value for a particular product depends on the proportion of that product and sometimes also on the degree of separation of the relevant 1H NMR signals. In some cases the estimated uncertainties, shown as subscripts to the PR values in the Tables, are comparable with the values of PR themselves, but in no case are they such as to cast doubt on the validity of the discussion below.Results and discussion We first focus on the results of the reactions in MeOH, initially considering the results in terms of the simple mechanism shown in Scheme 1. The main features with comments, are as follows. (a) Some TOH is formed in every case. For the reactions with AgX where X = NO3, O2CCF3 or BF4, the TOH could not have come from hydrolysis of the initially formed products TX, which are inert under the conditions used, and can be assumed to arise from reaction of the intermediate cation with water.Unfortunately, reliable values of PR for TOH could not be Scheme 1 Simplest representation of the reaction of 1-D with AgX in an alcohol R9OH used as the basis of the discussion, where its inadequacies and possible modifications are considered Me + AgI Me2 Si R2C Si (CD3)2 R3CSi(CD3)2I + AgX slow –X– R = Me3Si R3CSi(CD3)2X + R2C[Si(CD3)2Me]SiMe2X R3CSi(CD3)2OR¢ + R2C[(CD3)2Me]SiMe2OR¢ X– R¢OH 1-D a g +J. Chem.Soc., Perkin Trans. 2, 1997 1635 Table 2 Yields of products TX, TOH and TOMe from reactions of TsiSi(CD3)2X, 1-D, with AgX in MeOH, with or without added H2O, at room temperature, rt, or under reflux, rf, and proportion, PR (%), of rearranged product, with estimated uncertainty (±) denoted by the subscript X ClO4 O3SCF3 T rt rf rt rf rt rf H2Oa 0 0 1.0 31 c 0 1.0 TXb 3 0 0 0 6 0 PR 145 175 TOHb 11 8 30 50 6 27 PR 165 185 115 235 95 105 TOMeb 86 92 70 50 88 73 PR 265 295 245 305 225 215 a Added water, vol%.b Refers to the mixture of rearranged and unrearranged products (see footnote to Table 1). c ca. 50 mol% H2O. obtained in these cases because of overlap of the relevant 1H NMR signals. For the reactions with AgClO4 and AgO3SCF3 however, much of the TOH is likely to come from initially-formed TX, since the compounds TsiSiMe2X in which X is a very good leaving group, viz.OClO3, OSO2CF3 or OCN, select water very efficiently from MeOH (the selectivity factors being ca. 500 and 5700 for OSO2CF3 and OCN, respectively, in molar terms).14 The effects of added water on the outcome of the reaction with AgClO4 (see Table 2) are revealing in this connection. With 1 vol% of added water (i.e. ca. 2.2 mol%) at room temperature the proportion of TOH is raised from 11 to 30%, but on going to 31 vol% of water (50 mol%) the proportion (under reflux) is further raised only to 50%; that is, under these conditions water and methanol compete on roughly equal terms. The results are consistent with the view that the TOH is formed by two distinct processes, the first the hydrolysis of TOClO3, which is highly selective towards water, and the second the direct reaction of water with cation II which, being highly reactive does not discriminate strongly between water and methanol. At the very low water concentrations the value of PR observed for TOH would be expected to be effectively the same as that for TOClO3 (and so relatively low), since the solvolysis of TsiSi(CD3)2OClO3 proceeds without rearrangement, as was confirmed during the present work, giving unrearranged TsiSi(CD3)2OH and TsiSi(CD3)2OMe in ca. 50 : 50 ratio in the ‘anhydrous’ MeOH and solely unrearranged TsiSi(CD3)2OH in an equimolar H2O– MeOH mixture.[The methanolysis of TsiSi(CD3)2OCN,15 the hydrolysis of 1-D,15 and the hydrolysis and methanolysis of TsiSiPh2I and TsiSiEt2I,16 are known not to involve rearrangement.] On the other hand, the PR value for the TOH formed directly from the cation would be expected to be comparable with that observed for TOMe, and the value for TOH formed in 1:1 H2O–MeOH under reflux, viz. 23, is seen to lie between that observed in the ‘anhydrous’ MeOH, i.e. 18, and that for TOMe, i.e. 30. (b) Except for the reaction with AgNO3 under reflux (for which it seems anomalously large, at 27%) the proportion of TX formed is seen to be <15% even if it is assumed that most of the TOH product in the case of X = OClO3 or OSO2CF3 is formed via TX. The PR values for TX vary between 8 and 17 at room temperature and 9 and ca. 25 under reflux. These values fall in the range observed for the reaction with the silver salts in non-hydroxylic solvents,12 and the same explanation could apply, namely that cation II is preferentially attacked at the asite by the X2 liberated near that site within an ion pair [II ? ? ?X2] as AgI is formed.However, it seems appropriate to wonder whether if the cation II ever became fully free it would be expected to react to a significant extent with [Ag1 ? ? ?X2] or X2, especially in the case of the very weakly nucleophilic ClO4 2 or CF3SO3 2 ions, rather than with the MeOH molecules present in much larger concentration and presumably providing solvation at the a- and g-Si sites. In the four-membered ring of cation II these sites would be held close together (calculations on the model bridged ion III with Z = Me gives the separation as 2.55 Å17) and it would require very little movement of X2 over the periphery of the cation within the ion-pair to bring it near to the g-site, where it could attack.If such a migration occurs in the reactions in MeOH then it could be assumed, of course, to occur also for those in the non-hydroxylic media, and, indeed, would be even more likely there.Whether or not there is such migration of liberated X2, it is perhaps somewhat surprising that the yield of TX (when account is taken of subsequent conversion of some of it into TOH for X = OClO3 and OSO2CF3) shows no obvious dependence on the relative nucleophilicities of X2. (c) The presence of LiX for X = NO3 or O2CCF3 leads to significant increases in the product ratio TX :TOMe but these increases are much smaller than would be expected in terms of simple direct competition between X2 and MeOH for the intermediate cation.The concentration of AgX in the reaction medium was ca. 0.025 mol cm23, and so for direct competition the product ratio TX:TOMe for the reaction involving only AgNO3, viz. 1 : 6 (15 : 85), should rise to ca. 3.5 : 1 (300 : 85) and 7 : 1 (600 : 85) in the presence of 0.050 and 0.10 mol cm23 LiNO3, respectively, compared with the observed ratios of 1 : 2.6 and 1 : 1.8.That the observed effects of added X2 are so small can be rationalized in terms of the modification of Scheme 1 suggested above, involving initial attack of the ion pair [Ag1 ? ? ?X2], since it can be reasonably assumed that the proportion of such ion pairs, already high, will not be much raised as the concentration of X2 is increased. (d ) The PR values for TX are mostly raised only by a factor of ca. 1.3 in the presence of 0.050 mol cm23 LiX, and further only by a similar factor on going to 0.10 mol cm23 LiX.To the extent that the rearranged TX is formed by internal transfer of X2 to the g-site within the ion pair [II ? ? ?X2] the presence of additional X2 should not affect the proportion of attack at the g-site, and the small rise in PR could be attributed to the relatively minor proportion of direct attack of X2 at that site. (e) The values of PR for TOMe are mostly substantially below 50, especially in the case of reactions at room temperature.For X = ClO4 or O3SCCF3 some of the TOMe could arise by solvolysis of initially formed TX, leading to low values of PR of <50, but this could not be the case for formation of TOMe from the other TX compounds, which gives similar PR values. In terms of the assumed mechanism this implies that the cation II is not, or not always, fully-free (i.e. solvent-separated from other species involved in the reaction) before the attack by the MeOH, but is in some way rendered unsymmetrical.This asymmetry, while very significant in showing unambiguously that the mechanism of the solvolysis cannot be as simple as that depicted in Scheme 1, is not large; a PR of 25 means that attack at the a-site takes place three times as readily as that at the g-site, and for a PR of 40 the factor is 1.5. As one possible explanation of the asymmetry we suggest that there is some preassociation of an alcohol molecule by hydrogen-bonding to the I atom of the Si]I bond and that as the I atom leaves this molecule is released in a favourable position to attach to the a-site of the bridged cation.This suggestion has the attraction that the currently favoured explanation of remarkable observations on the solvolysis of the highly reactive compounds TsiSiMe2X with X = OClO3, OCN, OSO2- CF3 also involves assumption of similar preassociation,14 and if1636 J. Chem. Soc., Perkin Trans. 2, 1997 it is relevant in those reactions it could be expected to be so for reactions of the iodide 1-D when Ag1 ions are present to assist the departure of I2, although the extent of the hydrogen bonding is likely to be smaller with the I atom than with the O-bonded ligands.Such preassociation would also account for the low values of PR in the reactions of 1-D with ICl in MeOH.§ Another tentative alternative explanation of the low PR values is that some TOR is produced from the cation before the X2 released from [Ag1 ? ? ?X2] has fully separated, and that solvent molecules engaged in solvation of X2, e.g. as [ROH? ? ?X]2, carrying a small negative charge, are slightly more nucleophilic than the bulk solvent molecules near the g-site.Since at higher temperatures that solvated X2 ions would diffuse away from the a-site more rapidly the values of PR for TOR formation would be higher, as observed. ( f ) When account is taken of the uncertainties in the values of PR, for reactions at room temperature there is no significant variation in the values as X in AgX is varied, which is as would be expected if the intermediate cation becomes free from X before reaction with the MeOH.The values appear to be generally higher for reactions under reflux, but again there is no significant variation within them. (g) In the other alcohols, for reactions of AgX with X = NO3, O2CCF3 or BF4, the yield of TX rises, as expected, as the increasing bulk of the alcohol leads to increasing inhibition of attack of the alcohol to give TOR.This is almost certainly also the case for reactions involving AgClO4 and AgO3SCF3, but the formed TOClO3 and TOSO2CF3 are mainly converted into TOH. (In all the media the ratio of TOR: TX for X = OCOCF3, is relatively high, especially in relation to that for X = ONO2; this would not have been predicted since CF3CO2 2 would be expected to be rather more nucleophilic than NO3 2.) The values of PR for TOR could, other things being equal, also be expected to be higher in the bulkier alcohol because the increased steric hindrance to nucleophilic attack allows more time for the formation of the free cation; there is some indication that this is the case for the reactions in PriOH and ButOH, but the effect is rather small, perhaps because the greater bulk of the molecules also leads to lower rates of diffusion of X2 from the reaction site.For each alcohol values of PR for TOR within the set at room temperature or within that under reflux remain reasonably constant within the uncertainty; some seeming possible deviations from this generalization are likely to be unreal; e.g.the value of 16 ± 5 for TOEt formed in the reaction with AgO3SCF3 seems rather low, but is probably simply an extreme reflection of the experimental error, since there is no such anomaly for the reaction under reflux or for reactions in the other alcohols. Possible alternative mechanisms We are not wholly satisfied with our suggested rationalizations of the experimental observations, especially of the fact that the PR values for TOR are always <50, but possible alternative explanations, as considered below, seem to be even less satisfactory.One obvious, and rather attractive, explanation is that the SiMe2 and Si(CD3)2 sites in cation II are not equally prone to nucleophilic attack. Since the C]D bond is shorter than the C]H bond, the CD3 group in II could be significantly smaller than the CH3 group, the difference possibly being sufficiently large in this very crowded species to make attack at the a-site significantly easier than that at the g-site (compare ref. 19). That this may indeed be the case was suggested by our finding that in § The low PR values observed in formation of TX in alcohols could likewise be attributed to preassociation of X2 (perhaps as an ion pair [Ag1 ? ? ?X2]) by a weak nucleophilic interaction at the Si atom of the Si]I bond, and it is noteworthy that the formation of tertiary alkyl azides accompanying solvolysis of simple tertiary alkyl halides in the presence of azide ion is now thought to arise by preassociation of that ion with the starting halide.18 the reaction of 1-D with KSCN in MeCN, which involves unambiguous nucleophilic substitution at the a-site, the rate for TsiSiMe2I was ca. 25% lower than that for TsiSi(CD3)2I. It is impossible to estimate how large such an effect might be for attack of ROH on II; for a given nucleophile the steric effect could be expected to be markedly smaller than that at the silicon atom bearing a large I atom in TsiSiMe2I, but against that the hindrance would be larger for attack by the MeOH molecule than by the small linear SCN2 ion.Our view is that the effect, while significant, is most unlikely to cause the attack at the g-site in II to be ca. 3 times as slow as that at the a-site as it would have to be to account for PR values of ca. 25%, and that major reason for such values must be sought elsewhere. A further possible explanation of the fact that the values of PR for TOR are always <50 is that the TOR products are formed by two distinct processes, one involving attack only at the a-site. Thus formation of an unbridged cation, TsiCSi- (CD3)2 1, could precede that of the bridged ion II. Trapping of the unbridged ion by ROH would give unrearranged product, corresponding to PR = 0, while that of the fully-free bridged ion would give equal amounts of rearranged and unrearranged products, corresponding to PR = 50, and so any value of PR up to 50 could be observed.(Similar considerations would apply, of course, to trapping by X2 or water.) Logically it would have to be assumed that the unbridged cation is also initially formed in reactions of 1-D with AgX in inert solvents or with ICl in methanol or aprotic solvents. Such a mechanism, which we refer to below as the two-stage ionization mechanism, can be ruled out for related reactions involving much better bridging groups, as in reactions of the compounds (Me3Si)2(ZSiMe2)CSiMe2Y with, e.g. Z = Ph, CH]] CH2 or OMe, since in these cases there is clear anchimeric assistance, showing that the bridging is concerted with the departure of Y2,20 but there is no direct evidence for such assistance in the case of bridging Me. Furthermore calculations on the model bridged ion III with Z = Me (in contrast to those for III with Z = Ph) indicate that there is little difference in energy between the bridged and unbridged species.17 Nevertheless we disfavour this mechanism for the following reasons.(i) There is compelling evidence from high-level calculations that sterically-unhindered silylium ions R3Si1 must always be solvated in solution, strongly so even in alkanes, and signifi- cantly so in an argon matrix,17 and even when account is taken of the considerable steric hindrance in a free cation (Me3Si)3- CSi(CD3)2 1 it seems unlikely that this ion would be formed unsolvated in MeOH, or that the solvated ion once formed would fail to collapse to the methoxide rather than undergoing conversion to the bridged ion.In contrast, direct formation of the bridged cation keeps both the a- and g-silicon atoms fourcoordinate, greatly reducing the need for solvation at those centres.17 (ii) Although steric hindrance to nucleophilic attack appears to be rather similar at the functional centres in But 3SiI and TsiSiMe2I, as indicated by their fairly similar reactivities towards KSCN in MeCN, TsiSiMe2I is much the more reactive towards electrophiles, e.g.silver salts or ICl, and this can reasonably be attributed to anchimeric assistance by a bridging Me group.21 It is difficult to envisage any other acceptable explanation of the much greater ease of ionization of (Me3Si)3CSiMe2I; there could be some stabilization of the (Me3Si)3CSiMe2 1 ion by hyperconjugative electron-release from the Me3Si]C bonds (analogous to stabilization of carbocations by b-Me3Si groups) but it seems unlikely that this effect (which in valence bond terms requires double-bond character in the C]SiMe2 bond) would be substantially larger than stabilization of the But 3Si1 ion by the three alkyl groups.(iii) It seems unlikely that the competition between attack on the unbridged ion and that on the bridged ion, which could give rise to a value of zero at one extreme and 50 at the other, would not more often fall near one or other of these extremes forJ.Chem. Soc., Perkin Trans. 2, 1997 1637 formation of TX products in reactions with AgX in inert or protic solvents or of TOR products in reactions with AgX in various alcohols, or reactions with ICl in MeOH or inert solvents. Another possible reaction scheme involves competition between two unrelated processes. In this picture a bridged-ion is formed, and perhaps attacked equally at a- and g-sites, but this reaction is accompanied by a competing one at the a-site involving nucleophilic attack by X2 or ROH concerted with abstraction of I2.This cannot be ruled out, but it seems even more improbable than for the two-stage ionization mechanism that the balance between the two processes, as indicated by PR for TOR, would vary so little for the reactions of the iodide over the range of AgX in alcohols and ICl in MeOH. In particular one would expect that the rate of the reaction involving attack of ROH at the a-site synchronous with leaving of X2 would be greatly reduced on going to the bulkier alcohols, with the rate of the ionization much less affected, giving rise (on the assumption that the cation becomes fully free) to PR values of 50.In summary it seems to us that although the fact that the PR values for formation of the alkoxides TOR all fall below 50 shows that the bridged cation is not always fully free before capture by the alcohols, the discrepancy is not so large as to favour one of the more complex dual processes, especially that of the second type outlined above involving two wholly unrelated reactions, one ionization and the other nucleophilic attack synchronous with leaving of I2. Our suggestions of preassociation of ROH at the I atom of the Si]I bond or of preferential reaction of the intermediate with an alcohol molecule solvating X2 liberated near the a-site provide plausible rationalizations, but others may suggest themselves as more evidence accumulates.We hope to be able to throw further light on the mechanism of reactions of 1-D in due course by synthesizing and studying the reactions of the related compound (Me3Si)2C(SiPhMe2)- Si(CD3)2I, for which there would be no doubt that there is anchimeric assistance by the Ph group and so direct formation of the phenyl-bridged cation. Conclusion Although the simplest mechanistic picture, as shown in Scheme 1, can be reconciled with the experimental data only by making several ad hoc modifications, it seems to be substantially more satisfactory than the possible alternatives that we can suggest.Whatever the mechanism, it is clear that there is a definite preference for substitution at the a-site, and this is especially marked for formation of TX products in both alcoholic and aprotic media. This must be taken into account when the effects of substituents at the a- and g-sites are considered.Thus in the case of the compound TsiSiMeEtI with AgBF4 in CH2Cl2 the small additional steric effect at the a-site seems just to offset the inherent favouring of attack at that site, with ca. 50% of the fluoride product being rearranged. 22 compared with 32–47% in the case of 1-D. However, even in the case of TsiSiMeEtI only ca. 10% of rearranged fluoride is formed in the reaction with AgBF4 in Et2O,22 compared with 5 ± 3 in the case of 1-D, and the remarkable effect of this solvent in inhibiting rearrangement still awaits satisfactory explanation, although we suspect that it may be associated with strong stabilization of, e.g.the ion TsiSiMe2 1 by co-ordinated Et2O (cf. the strong stabilization calculated 17 for [Me3Si?OH2]1). Some other effects of substituents in reactions of the compounds TsiSiR9R0I can be interpreted mainly in terms of electronic effects on the distribution of the formal positive charge in a cation of type I.Thus in the case of TsiSi(OMe)2I and TsiSIMe(OMe)I the fact that only unrearranged products are formed in the reaction with silver salts in CH2CL2 11 or with ICl in CCl4 12 can be attributed to greater interaction of the OMe than of the Me groups with the positive charge at the a- and gsites, respectively (compare the greater stabilization calculated for SiH3 1 on introduction of an OH than of a Me substituent 17). This will lead to a higher proportion of the cationic charge lying at the a- than at the g-site [or, in resonance terms a greater contribution by, for example, the canonical form IV than by V], and so to a higher proportion of nucleophilic attack at the a-site. On the other hand, the fact that only rearranged product is formed in the reaction of TsiSiMeFI with AgOSO2CF3 in CH2Cl2 or with ICl in CCl4 10 even though F should have a smaller steric effect than a Me substituent, can be attributed to substantial lowering of the proportion of the charge at the a-site in the bridged cation, calculations having indicated that FSiH2 1 is considerably less stable than MeSiH2 1.17 In continuing to favour the mechanism shown in Scheme 1, with ad hoc modifications, in spite of the problems this presents, we note that the simple concept of SN1 reactions of alkyl halides and related species had to be progressively modified and refined in terms of multi-stage processes as more and more detailed information became available, and that it has taken studies by several research groups over more than 50 years even to allow a reasonably firm conclusion that solvolysis of secondary alkyl substrates involves an SN1 process (with capture of a nucleophile at various stages of separation of the initially formed ions) rather than competing SN1 and SN2 processes.23 Furthermore, even the question of whether the solvolysis of simple tertiary alkyl halides proceeds by an SN1 mechanism is still a matter for investigation 18 some 60 years after it was first suggested that it does so.Experimental Syntheses TsiSi(CD3)2H. A 0.50 mol dm23 solution (50 cm3) of CD3Li in diethyl ether was added in one portion to solid TsiSiH3 (2 g, 8 mmol) in a small flask. The mixture was stirred under nitrogen for 24 h, then treated dropwise with wet diethyl ether (20 cm3) followed by water. When gas evolution had ceased, concentrated hydrochloric acid (10 cm3) was added, and the ethereal layer was separated, washed with water and then aqueous NaHCO3 and dried (MgSO4).The diethyl ether was evaporated to leave a mixture of TsiSI(CD3)2H and TsiH in ca. 2 : 1 ratio and the TsiH was removed by pumping under vacuum at 60 8C for ca. 1 h. TsiSi(CD3)2I. [This procedure was performed several times on a 0.05–0.1 mmol scale as initial experiments showed that larger scale reactions often gave some TsiSi(CD3)2Cl as a byproduct.] A 0.10 mol dm23 solution of ICl in Cl4 was added dropwise to a rapidly stirred solution of TsiSi(CD3)2H (ca. 50 mg, 0.5 mol) in CCl4 (10 cm3). The progress of the reaction could be monitored by 1H NMR spectroscopy, but it was found with experience that it was possible to detect when an equivalent of ICl had been added by the appearance of a reddish tinge in the solution. At that point the solution was washed, first with 10 cm3 of 0.05 mol dm23 aqueous sodium bisulfite to remove any iodine, and then with water, and the organic layer was separated, dried (MgSO4) and evaporated under reduced pressure to leave the solid product, the purity of which was checked by 1H NMR spectroscopy.It was normally pure enough to be used without further purification, but if necessary contaminants were removed by chromatography on silica gel with light petroleum as eluent. Other TsiSi(CD3)2X derivatives. These were prepared as needed by treating TsiSi(CD3)2I with an excess of the appropriate silver salt in anhydrous diethyl ether.The products obtained Me Me2 Si R2C Si+ (OMe)2 IV Me Me2 Si+ R2C Si (OMe)2 V a g + a g +1638 J. Chem. Soc., Perkin Trans. 2, 1997 in this way contained up to 5% of (Me3Si)2CSi[(CD3)2Me]- SiMe2X, together with traces of TsiSi(CD3)2OEt and TsiSi- (CD3)2OH. Since the samples of TsiSi(CD3)2X were required only for solvolysis studies the presence of these other species was acceptable once the exact composition of the mixture had been established by 1H NMR spectroscopy.Other reagents. The alcohols used were treated by the best available standard methods to remove water but still contained traces of it. The silver salts were supplied (Aldrich) and ‘anhydrous’, except for AgClO4 which was supplied as the monohydrate and used as such. Analysis of product mixtures Compounds were identified, and the composition of mixtures determined, by 1H NMR spectrometry with a Varian Unity 300 MHz spectrometer. For the identification of products the 1H NMR spectra were compared with those of authentic (non-deuterated) samples and the yields were determined by integration.The extent of rearrangement accompanying the formation of products was calculated by use of eqn. (1) where r is the ratio of % rearrangement (PR) = 450/(r 1 1) (1) the area of the peak from all other SiMe groups to that of the peak from the SiMe2Y group. This formula is based on the assumption that the only species contributing to these peaks are (Me3Si)3CSi(CD3)2Y and (Me3Si)2C[SiMe(CD3)2]SiMe2Y.The reliability of the PR values derived in this way depends on how accurately the peak areas can be determined, and on the yield of the product under consideration. The reactions were fairly clean, and the species TX, TOR and TOH together constituted >95% of the product mixture. Any other products gave rise only to signals with chemical shifts <0.20 ppm, thus causing no significant interference. Most analysis problems arose from partial overlap of product peaks, the shifts for which were as follows. (a) For Me3Si 1 Me(CD3)2 protons: (X =) I, 0.34; ClO4, 0.29; O3SCF3, 0.30; NO3, 0.28; O2CCF3, 0.28; F, 0.24; OH, 0.24; OMe, 0.21; OEt, 0.22; OPri, 0.23; OBut, 0.23.(b) For Me2Si protons: (X =) I, 1.06; ClO4, 0.72; O3SCF3, 0.72; NO3, 0.63; O2CCF3, 0.63; F, 0.38; OH, 0.33; OMe, 0.25; OEt, 0.26; OPri, 0.27; OBut, 0.35. The most serious overlap problems involved the SiMe2 peaks of the silyl alkoxides, which lie in the region where the much larger Tsi peaks of other products (especially that of the silanol) usually appear; the Tsi peaks tended to be broad and as a result the signals from the SiMe2 groups in the ethers often appeared as shoulders on them.This made electronic integration unreliable and in such cases areas were normally determined by cutting and weighing. In two cases overlap was so extensive that direct analysis was not possible. One was for TF in the AgBF4 reactions, where the peak for the Me3Si 1 Me protons extensively overlapped that for TsiSi(CD3)2OH, a significant product in these reactions; the best that could be done was to derive estimates for PR by partitioning the combined Tsi 1 Me peak in terms of the relative yields of the fluoride and silanol obtained for the reaction of TsiSiMe2I with AgBF4 under the same conditions.The other was for the reaction with AgOtfa in PriOH, where overlap between the peak for the Me3Si 1 Me protons of the TOtfa and that for the SiMe2OPri was so complete as to rule out even a rough estimate of PR. The yields and PR values for TOBut must be treated with particular caution, since TsiSiMe2OBut has never been isolated, and the assigned shifts are only tentative. Furthermore, the assigned Tsi peak for this derivative lies close to that for the silanol.As a result, the PR values for TOBut in reactions in which the yield of silanol was high are especially uncertain.Our estimates of the reliability of PR values shown in the tables are based on the observed reproducibility and on analysis of a few standard mixtures. The least reliable values are those where the yield of product was small or overlap of peaks was extensive. In spite of the problems we believe that at no point do the uncertainties in the PR values cast doubt on the validity of the conclusions drawn. Reactions of TsiSi(CD3)2I with Ag salts For reactions carried out at room temperature TsiSi(CD3)2I (8 mg, ca. 0.02 mmol) was added to a solution of the silver salt (8 mg) (and where relevant LiX) in the alcohol (2 cm3) contained in a reacti-vial. The vial was filled with dry argon and capped to exclude air and the mixture was stirred until reaction was complete, as indicated by 1H NMR spectroscopic monitoring. The time required varied with the alcohol, from 30–60 min for MeOH to 24 h for PriOH and ButOH. The product solution was worked up by adding it to pentane (3 cm), shaking the organic layer with water to remove the alcohol and any excess of silver salt, drying it over anhydrous sodium sulfate and evaporating the pentane under reduced pressure.The residual solid was dissolved in CDCl3 for recording of the NMR spectrum. Checks showed that neither TsiSiMe2OClO3 nor Tsi- SiMe2O3SCF3 underwent significant hydrolysis during such work-up. For reactions performed under reflux conditions, the experiments were carried out under argon in a 10 mm diameter testtube fitted with a condenser. Reactions were complete in 15–30 min.The scale and the method of work-up were as described for the room temperature reactions. Separate experiments showed that the compounds TsiSiMe2X with X = F, OCOCF3, or ONO2, underwent no appreciable reaction under the conditions used. Solvolysis of TsiSi(CD3)2X (X = OClO3 or OSO2CF3) The relevant compound {ca. 8 mg, containing up to 5% of (Me3Si)2CSi[(CD3)2Me]SiMe2X, together with a little TsiSi- (CD3)2OEt and TsiSi(CD3)2OH as noted above} was dissolved in MeOH under dry argon and the solution was kept at room temperature or under reflux under argon until reaction was judged to be complete.Work-up and analysis were then as described for the reactions in the presence of silver salts, appropriate allowance being made for the presence of the impurities in the starting material. Reaction of TsiSi(CH3)2I and TsiSi(CD3)2I with KSCN in MeCN A solution of the iodide (0.01 mol dm23) and KSCN (0.10 mol dm23) in anhydrous MeCN was kept under dry argon in a vessel immersed in a thermostat held at 35.0 ± 0.05 8C.Samples were removed at appropriate intervals and rapidly cooled, and the solvent was then evaporated off under vacuum. The organic component of the residue was extracted into CDCl3 for determination of the 1H NMR spectrum. The relative integrals of the signals from the (Me3Si)3C groups in the reactant and product indicated the extent of reaction. Good first order kinetics were observed, giving rate constants of 11.10 ± 0.04 and 8.80 ± 0.04 s21, respectively.Acknowledgements We thank Dr A. I. Almansour for important contributions to this study, the EPSRC for support (via C. E.) and Mr R. W. Bott and Dr P. D. Lickiss for helpful comments on the manuscript. References 1 A. R. Bassindale and P. G. Taylor, in The Chemistry of Organosilicon Compounds, ed. S. Patai and Z. Rappoport, Wiley, 1989, pp. 880–886. 2 C. Eaborn, D. A. R. Happer, S. P. Hopper and K. D. Safa, J. Organomet. Chem., 1979, 170, C9. 3 C. Eaborn, D. A. R. Happer, S. P. Hopper and K. D. Safa, J. Organomet. Chem., 1980, 188, 179. 4 C. Eaborn and K. D. Safa, J. Organomet. Chem., 1982, 234, 7.J. Chem. Soc., Perkin Trans. 2, 1997 1639 5 S. S. Dua and C. Eaborn, J. Organomet. Chem., 1981, 204, 21; C. Eaborn, P. D. Lickiss, G. Marquina-Chidsey and E. Y. Thorli, J. Chem. Soc., Chem. Commun., 1982, 1326. 6 C. Eaborn and S. P. Hopper, J. Organomet. Chem., 1980, 192, 27. 7 C. Eaborn, K. D. Safa, A. Ritter and W. Binder, J. Chem. Soc., Perkin Trans. 2, 1982, 1397. 8 C. Eaborn and S. P. Hopper, J. Organomet. Chem., 1980, 192, 27. 9 C. Eaborn, J. Organomet. Chem., 1983, 239, 93. 10 C. Eaborn and D. E. Reed, J. Chem. Soc., Perkin Trans. 2, 1985, 1687. 11 C. Eaborn and D. E. Reed, J. Chem. Soc., Perkin Trans. 2, 1985, 1695. 12 A. I. Almansour, J. R. Black, C. Eaborn, P. M. Garrity and D. A. R. Happer, J. Chem. Soc., Chem. Commun., 1995, 705. 13 A. I. Almansour, H. A. Abubishait and C. Eaborn, J. Organomet. Chem., 1997, 531, 171. 14 Y. Y. El-Kaddar, C. Eaborn, P. D. Lickiss and D. E. Reed, J. Chem. Soc., Perkin Trans. 2, 1992, 1753. 15 A. I. Almansour, personal communication, 1996. 16 S. A. I. Al-Shali, C. Eaborn, F. A. Fattah and S. T. Najim, J. Chem. Soc., Chem. Commun., 1984, 318; C. Eaborn and F. A. Fattah, J. Organomet. Chem., 1990, 396, 1. 17 C. Maerker, J. Kapp and P. v. R. Schleyer, in Organosilicon Chemistry II. From Molecules to Materials, VCH, Weinheim, 1996, pp. 329–359. 18 M. M. Toteva and J. P. Richard, J. Am. Chem. Soc., 1996, 118, 11 434. 19 F. Ruff and I. G. Cszimadia, Organic Reactions. Equilibria, Kinetics and Mechanism, Elsevier, Amsterdam, 1994, pp. 233–239. 20 For leading references see C. Eaborn, P. D. Lickiss and A. D. Taylor, J. Chem. Soc., Perkin Trans. 2, 1994, 1809. 21 C. Eaborn and A. K. Saxena, J. Organomet. Chem., 1984, 271, 33. 22 S. M. Whittaker, PhD Thesis, University of Salford, 1993. 23 P. E. Dietze, in Advances in Carbocation Chemistry, ed. J. M. Coxon, 1995, 2, pp. 179–205. Paper 7/02585H Received 15th April 1997 Accepted 6th June 1997

ISSN:1472-779X    

DOI:10.1039/a702585h

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1641 Physical organic chemistry of transition metal carbene complexes. Part 11.1 Kinetics and mechanism of the hydrolysis of (2-oxacyclopentylidene)pentacarbonylchromium(0) in aqueous acetonitrile Claude F. Bernasconi * and Aquiles E. Leyes Department of Chemistry and Biochemistry, University of California, Santa Cruz, California 95064, USA A kinetic study of the hydrolysis of the title compound, 6, in 50% acetonitrile–50% water (v/v) at 25 8C is reported.The organic products are 2-hydroxytetrahydrofuran in equilibrium with small amounts of 4- hydroxybutanal. There are two possible mechanisms that can account for the hydrolysis. (i) Rate limiting reaction of the conjugate anion of 6 (62) with water, buffer acids and H3O1, followed by (CO)5Cr catalyzed hydrolysis of the resulting 2,3-dihydrofuran. (ii) Ring opening of 6 through nucleophilic substitution on the carbene carbon by OH2, water or by buffer base catalyzed water attack, followed by breakdown of the intermediate substitution product into final hydrolysis products.Kinetic solvent isotope effects can be interpreted by either mechanism. Based on more conclusive isotope effect experiments in the hydrolysis of (CO)5Cr]] C(OR9)CH3 (R9 = CH3 or CH3CH3) and (CO)5Cr]] C(OMe)CH2Ph reported earlier, the first mechanism is preferred by reason of analogy, at least in basic solution. In acidic solution the mechanistic ambiguity could not be resolved, not even for (CO)5Cr]] C(OMe)CH3 which was reinvestigated in HCl and DCl solutions.Introduction Fischer-type carbene complexes of the general structure 1 undergo hydrolysis to form an aldehyde RCH]] O and alcohol R9OH as the organic products in most cases.2–4 However, the mechanism which leads to these products is not the same for all carbene complexes but depends on the nature of the R group. If R does not contain a hydrogen on the carbon a to the carbene carbon, e.g.R = phenyl, the only plausible mechanism available is that shown in Scheme 1.2,4 Scheme 1 represents the reaction in basic solution where 2 is predominantly present as its anion; formation of 2 (22) is quite fast and occurs via a tetrahedral intermediate whose formation is rate limiting. The transformation of 2 (22) into final products is a known,5 but mechanistically poorly understood reaction, which occurs on a much slower timescale than the formation of 2 (22).4 If R contains a hydrogen on the carbon adjacent to the carbene carbon, e.g.R = CH3, there is a competing mechanism which is more efficient than that of Scheme 1. It involves rapid deprotonation of the carbene complex followed by rate limiting reaction of the anion with a proton donor to generate the corresponding vinyl ether.3 This vinyl ether appears to be complexed with (CO)5M which activates it toward rapid hydrolysis to the corresponding aldehyde. Such activation is necessary (CO)5M C OR¢ R 1 (M = Cr or W) Scheme 1 (CO)5M C OH R OR¢ (CO)5M C OH(O–) R RCH O _ (CO)5MOH– + slow 2 (2–) 2 (2–) + R¢OH H2O k2 OH k–1 k1 1 + OH– H2O because vinyl ethers are stable under basic conditions; the effectiveness of this activation has been demonstrated independently in the basic hydrolysis of CH2]] CHOEt.3 The mechanism is shown in Scheme 2 for the reaction in basic solution. A major piece of evidence for this mechanism is the large kinetic solvent isotope effect found in the hydrolysis of 4a,3 4b,3 5a 6 and 5b 6 in 50% acetonitrile–50% water which is inconsistent with the mechanism of Scheme 1.In the case of 5a and 5b the nucleophilic substitution mechanism could also be excluded based on the isolation of PhCH]] CHOCH3 instead of PhCH2- CH]] O as the hydrolysis product. The reason why, in this case, the vinyl ether was not hydrolyzed is attributed to a lower stability and/or shorter lifetime of the complex between (CO)5Cr and PhCH]] CHOMe due to steric crowding.6 Based on recent estimates 4 the pathway via Scheme 2 for the hydrolysis of 4a in basic solution appears to be approximately five-fold faster than the pathway via Scheme 1.Scheme 2 (CO)5M C OR¢ CH (CO)5M C OR¢ C (CO5)M C C H OR¢ – (CO)5MOH– fast + OH– + H2O 3– 3– 3 OH K1 + OH– k2 H2O + ~ CHCH O + R¢OH + H2O (CO)5Cr C OR¢ CH3 (CO)5M C OCH3 CH2Ph 5a (M = Cr) 5b (M = W) 4a (R¢ = CH3) 4b (R¢ = CH3CH2)1642 J. Chem. Soc., Perkin Trans. 2, 1997 In this paper we wish to examine whether the hydrolysis of (2-oxacyclopentylidene)pentacarbonylchromium(0), 6, follows the mechanism of Scheme 1 or Scheme 2.Complex 6, just as 4a, 4b, 5a and 5b contains an acidic proton a to the carbene carbon and hence one might expect its hydrolysis to follow Scheme 2. It will be shown, however, that the results for 6 are less clear cut and are consistent with either mechanism. A reinvestigation of 4a in HCl solution, coupled with a kinetic solvent isotope effect study, also reveals mechanistic ambiguities in acidic solution with this compound. Results General features and product study Unless stated otherwise, all experiments were carried out in 50% MeCN–50% water (v/v) at 25 8C.When 6 is added to a KOH solution at [KOH] > 0.01 M, two reactions are observed by monitoring the UV–VIS spectrum. The faster of the two is in the subsecond range and can be attributed to the reversible deprotonation of 6 that leads to the anion 62.A detailed study of this process has been reported recently.1 The slower process is irreversible and in strongly basic solution occurs on a timescale of several seconds but becomes progressively slower as the pH is decreased. At pH <13.2 ([OH2] < 0.01 M) it is the only visible reaction because the acid–base equilibrium between 6 and 62 strongly disfavors 62, making the proton transfer undetectable. Fig. 1 shows time resolved absorption spectra of the reaction of 6 in a 0.005 M KOH solution.The organic products of the slow process are the hemiacetal 2-hydroxytetrahydrofuran (7) in equilibrium with small amounts of its acyclic form, 4-hydroxybutanal (8). Identification of 7 and 8 was achieved by 1H and 13C NMR spectroscopy. A solution of 0.024 M 6 and 0.01 M NaOD in 75% CD3CN–25% D2O (a higher proportion of organic solvent than in the kinetic experiments was necessary to make 6 more soluble) was left to react and its NMR spectrum compared to that of an authentic sample of 7 in equilibrium with 8.This latter sample was generated by acid hydrolysis of 2,3-dihydrofuran with 0.01 M DCl in 75% CD3CN–25% D2O and then made basic with NaOD, to mimic the conditions of the reaction of 6. Both the 1H and 13C NMR spectra of hydrolyzed 6 were the same as those of the hydrolyzed 2,3-dihydrofuran, except that in the 1H NMR spectrum of the hydrolyzed 6 the signal for the anomeric hydrogen in 7 was missing. The replacement of the anomeric hydrogen by a deuterium atom is consistent with hydrolysis conducted in the presence of D2O.Kinetics Rates of hydrolysis of 6 were measured in the pH range 1.0– 14.2. All kinetic determinations were made under pseudo-firstorder conditions with 6 as the minor component. The ionic strength was maintained at 0.1 M with KCl. The observed pseudo-first-order rate constants, kobsd, are reported elsewhere.7 Experiments were performed in HCl solutions (pH 1.01–1.71), methoxyacetic acid (pH 4.73), acetic acid (pH 5.93), Nmethylmorpholine (N-MeMor) (pH 7.43–8.43) and triethylamine buffers (pH 9.31–12.0), and in KOH solutions (pH O (CO)5Cr O (CO)5Cr 6– 6 – O OH HO O H 8 7 12.18–14.18).In all buffers, general base catalysis was observed. Intercepts of the linear plots of kobsd vs. buffer concentration were combined with the HCl and KOH data to construct the pH–rate profile shown in Fig. 2. The slight downward curvature at the high pH end is consistent with the onset of a shift of the acid–base equilibrium towards the anion (62), as expected on the basis of pKa CH = 14.47 for 6.1 The relatively large scatter in the plateau region of the pH–rate profile is mainly attributed to the slowness of the reaction, which necessitated the use of the inherently less precise initial rates method to evaluate kobsd (see Experimental section).Between pH 4.3 and 8.4 an additional source of error comes from the extrapolation of relatively steep plots of kobsd vs.buffer concentration. Kinetic isotope eVects The deuterated carbene complex, [2H2]6, was reacted with 0.005–0.1 M KOD in 50% MeCN–50% D2O. Just as for the reaction of 6 with KOH in 50% MeCN–50% water, the deuteron transfer and hydrolysis could be observed as separate processes. A plot of kobsd vs. [KOD] for the hydrolysis reaction is Fig. 1 Time resolved absorption spectra for the hydrolysis of 6 in a 0.005 M KOH solution (pH 12.9). Spectra taken every 15 s in a Hewlett- Packard 8452A diode array spectrophotometer. Spectrum of the final product was taken after 8 min.Fig. 2 pH–rate profile of the hydrolysis of 6 in 50% CH3CN–50% H2O [2H2] 6 O (CO)5Cr D DJ. Chem. Soc., Perkin Trans. 2, 1997 1643 shown in Fig. 3. Just as in the pH–rate profile of the reaction of 6 with KOH in 50% MeCN–50% water there is a slight downward curvature at the highest KOD concentrations which is consistent with a shift of the acid–base equilibrium towards the anion.The initial slope of the plot is 5.30 ± 0.18 M21 s21; the corresponding quantity for the reaction of 6 with KOH obtained from the pH–rate profile in 50% MeCN–50% water is 5.20 ± 0.09 M21 s21. Experiments where [2H2]6 was reacted with KOH in MeCN– H2O or where 6 was reacted with KOD in MeCN–D2O did not yield easily interpretable results as discussed elsewhere.1 Attempts at measuring hydrolysis rates in DCl–MeCN–D2O solutions were unsuccessful.As can be seen from the pH–rate profile in Fig. 2, hydrolysis of 6 at low pH is very slow, with kobsd values in the order of 1026 s21. As mentioned earlier, this necessitated the use of the initial rates method which is subject to larger experimental errors than standard first order kinetics. The kobsd values obtained in HCl solutions were nevertheless of acceptable quality. However, experiments in DCl–MeCN–D2O gave erratic results, probably because the rates are considerably lower than in HCl–MeCN–H2O and possible formation of byproducts may be a more serious problem.Hence, we used 4a whose hydrolysis in HCl solution is about 50-fold faster than hydrolysis of 6,3 as a model to determine the kinetic solvent isotope effect (KSIE) in acidic solution. Two types of experiments were performed. (i) Reactions of 4a with DCl in 50% MeCN–50% D2O. The kinetic traces deviated slightly from ideal first-order behavior at short reaction times, probably because H/D exchange occurs in competition with hydrolysis; if [2H]4a, [2H2]4a and [2H3]4a † have slightly different reactivities this could account for the deviations.Nevertheless, approximate kobsd values could be determined. They are summarized in Table 1, along with the corresponding kobsd in HCl–MeCN–H2O. (ii) Reactions of [2H3]4a with DCl in 50% MeCN–50% D2O. In these experiments the kinetic traces showed strict first-order behavior which is consistent with the above interpretation of the slight deviations from first-order kinetics with 4a.The kobsd values are also reported in Table 1. The KSIE calculated for the two types of experiments are 2.98 and 3.19, respectively. Discussion Mechanism In anhydrous solvents, base catalyzed decomposition of Fischer carbene complexes that have acidic hydrogens adjacent to the carbene carbon typically lead to vinyl ethers. This has been Fig. 3 Plot of kobsd vs. [KOD] for the reaction of [2H2]6 in 50% CH3CN–50% D2O † [2H]4a: (CO)5Cr]] C(OCH)3CH2D; [2H2]4a: (CO)5Cr]] C(OCH3)CHD2; [2H3]4a: (CO)5Cr]] C(OCH3)CD3.shown for 4a and other similar complexes in the presence of neat pyridine and N-methylmorpholine, or of quinuclidine in hexane,8 and also for 6 in pyridine which leads to 2,3- dihydrofuran.9 In the presence of water, on the other hand, only 5a leads to the corresponding vinyl ether (PhCH]] CHOCH3) 6 while 4a and 4b lead to the hydrolysis products of the respective vinyl ethers.3 We now find that the decomposition of 6 in 50% MeCN–50% water also leads to the hydrolysis product of 2,3- dihydrofuran, 8 (which rapidly cyclizes to 7), rather than to 2,3- dihydrofuran.Formation of vinyl ethers is easily accounted for by reaction of 32 with a proton donor which leads to protonation of the carbene carbon and cleavage of the bond between the metal and carbene carbon. On the other hand, formation of the hydrolysis products of the expected vinyl ether poses a more complex mechanistic problem and may be explained in two different ways.(i) The reaction does not involve 32 (or 62 in the case at hand) as the intermediate but proceeds by a nucleophilic substitution mechanism, as is the case for 1 with R = Ph in aqueous acetonitrile 4 and for 9 (R = Ph, CH]] CHR and others) in THF in the presence of small amounts of water.2 This mechanism, applied to the hydrolysis of 6, is shown in Scheme 3 where (CO)5CrX is likely to be a mixture with X being either OH2, MeCN or a buffer base.Scheme 3 is essentially the mech- (CO)5Cr C OEt R 9 Scheme 3 C OH(O–) (CO)5Cr OH O C (CO)5Cr OH _ 6 6– 10 10 (CO)5CrX + 7 + 8 fast 11 K1 OH[OH– ] k1 H O + k1 [B] + k1 [OH– ] B OH 2 k–1aH +k–1 [BH] + k–1 H O 2 BH H + k2 H O + k2 [BH] + k2aH 2 + BH Table 1 Rate constants for the hydrolysis of 4a in MeCN–H2O–HCl and MeCN–D2O–DCl, and of [2H3]4a in MeCN–D2O–DCl at 25 8C [LCl]/1022 M kobsd/1025 s21 4a 1 HCl in 50% MeCN–50% H2O 0.58 2.26 4.52 6.80 11.3 average 4.62 4.88 4.91 4.91 5.26 4.92 4a 1 DCl in 50% MeCN–50% D2O 0.52 4.0 8.0 10.0 average 1.63 1.57 1.45 1.88 1.65 a [2H3]4a 1 DCl in 50% MeCN–50% D2O 0.52 2.0 4.0 6.0 10.0 average 1.25 1.47 1.65 1.62 1.67 1.54 b a kobsd(H2O)/kobsd(D2O) = 2.98.b kobsd(H2O)/kobsd(D2O) = 3.19.1644 J. Chem. Soc., Perkin Trans. 2, 1997 anism of Scheme 1, but it accounts for the deprotonation equilibrium of 6 and includes terms not only for reversible OH2 attack (k1 OH[OH2], k21 H2O) and spontaneous breakdown of the tetrahedral intermediate (k2 H2O) but also for reversible nucleophilic attack by water (k1 H2O, kH 21aH1), reversible buffer catalyzed water attack (k1 B[B], k21 BH[BH]), and for the H1 and BH catalyzed breakdown of the intermediate (k2 HaH1, k2 BH[BH]).(ii) Alternatively, the reaction does involve 32 (62 in the present case) as the intermediate and leads to the vinyl ether but the latter is rapidly hydrolyzed because of complexation with (CO)5Cr.This mechanism is shown in Scheme 4 which is an elaboration of Scheme 2 to include k� 2 HaH1 and a k� 2 BH[BH] term.‡ The mechanisms of Schemes 3 and 4 are kinetically indistinguishable. If nucleophilic attack in Scheme 3 is assumed to be rate limiting,§ kobsd is given by eqn. (1), while kobsd for Scheme kobsd = k1 H2O 1 k1 B[B2] 1 k1 OH[OH2] 1 1 K1 OH[OH2] (1) 4 is given by eqn. (2). At pH ! pKa CH (K1 OH[OH2] ! 1), eqns. (1) kobsd = K1 OH[OH2] 1 1 K1 OH[OH2] (k� 2 HaH1 1 k� 2 BH[BH] 1 k� 2 H2O) (2) and (2) simplify to eqns.(3) and (4), respectively; at pH @ pKa CH kobsd = k1 H2O 1 k1 B[B] 1 k1 OH[OH2] (3) kobsd = K1 OHKwk� 2 H 1 K1 OHk� 2 BH[OH2][BH] 1 K1 OHk� 2 H2O[OH2] = Ka CHk� 2 H 1 Ka CH Ka BH k� 2 BH[B] 1 Ka CH aH1 k� 2 H2O (4) (K1 OH[OH2] @ 1; k1 H2O ! k1 OH[OH2]; k� 2 HaH1 ! k� 2 H2O), and in the absence of buffer, eqns. (1) and (2) simplify to eqns. (5) and (6), respectively. kobsd = k1 OH/K1 OH (5) kobsd = k� 2 H2O (6) Due to the high pKa CH of 6 (14.47), most of our kinetic data can be treated by eqns.(3) or (4). The various rate constants (k1 OH, k1 B and k1 H2O according to Scheme 3, or k� 2 H2O, k�2 BH and k� 2 H according to Scheme 4) that give the best fit to the pH–rate profile are summarized in Table 2. Before discussing these rate constants, we ask whether one can distinguish between the two mechanisms. In basic solution, where the k1 OH[OH2] [eqn. (1)] or the k� 2 H2O [eqn.(2)tween the two mechanisms is, in principle, possible on the basis of the kinetic solvent isotope Scheme 4 (CO)5Cr O 6– 6 12 12 (CO)5CrX + 7 + 8 K1 [OH– ] k2 + k2 [BH] + k2 aH OH H O BH H + fast 2 ~ ~ ~ ‡ In Scheme 4 formation of 12 is assumed to be a concerted reaction whereby protonation of the carbene carbon of the anion and metal] carbon bond cleavage are coupled, i.e. occur in a single step. As elaborated upon elsewhere,3 the alternative possibility (in the case of 4a and 4b) of rate limiting protonation of the anion on the metal followed by fast reductive elimination is less attractive.§ Alkoxide ions are typically better leaving groups than OH2 in ester hydrolysis 10 and other reactions;11 hence the reaction of 10 to form 11 is expected to be faster than reaction of 10 to revert back to 6 which makes the nucleophilic attack step rate limiting. effects (KSIE). For Scheme 3, these isotope effects are given by eqns.(7) and (8), respectively; for Scheme 4, the isotope effects are given by eqns. (9) and (10), respectively. kobsd(H2O) kobsd(D2O) = K1 ODk1 OH K1 OHk1 OD (pH @ pKa CH) (7) kobsd(H2O) kobsd(D2O) = k1 OH k1 OD (pH ! pKa CH) (8) kobsd(H2O) kobsd(D2O) = k� 2 H2O k� 2 D2O (pH @ pKa CH) (9) kobsd(H2O) kobsd(D2O) = K1 OHk� 2 H2O K1 ODk� 2 D2O (pH ! pKa CH) (10) Analysis of the KSIE data according to eqns. (7)–(10) affords the results reported in Table 3. The primary KIE reported in the table under ‘Scheme 4’ were estimated by assuming a secondary KSIE of 1.41 in all cases.¶ Focusing first on 4a and 4b, interpretation of the results in terms of Scheme 4 implies sizable primary KIE values, consistent with a mechanism where proton transfer from water to 4a2 or 4b2 is rate limiting or part of the rate limiting step.‡ On the other hand, interpretation in terms of Scheme 3 yields k1 OH/k1 OD ratios that are unrealistically high, especially for 4b2; typically OD2 shows nucleophilic reactivity equal to or somewhat higher than OH2,13,14 e.g., in the hydrolysis of 13 which must proceed by a nucleophilic mechanism, the k1 OH/k1 OD ratio is 1.0 ± 0.04,4 while for 14 it is 0.91 ± 0.05.4 Hence the nucleophilic mechanism is unattractive and can be excluded for 4a and 4b.3 The situation with 6 is less clear cut.For Scheme 4, the results yield a primary KIE of about 2.2 which is rather small, while for the nucleophilic mechanism we obtain k1 OH/k1 OD = 0.98 which is about the same as for the hydrolysis of 13.This suggests that (CO)5W C OEt Ph (CO)5Cr C OMe Ph 13 14 Table 2 Rate constants for rate limiting steps in the hydrolysis of Fischer carbene complexes at 25 8C. Interpretation in terms of Scheme 3 and Scheme 4.a Reactive agent Scheme 3 (k1 OH, k1 B, k1 H2O) OH2 (16.63) c Et3N (10.31) c N-MeMor (7.43) c AcO2 (5.93) c MeOCH2COO2 (4.73) c H2O (21.44) c Scheme 4 (k� 2 H2O, k� 2 BH, k� 2 H) H2O (16.63) d Et3NH1 (10.31) d N-MeMorH1 (7.43) d AcOH (5.93) d MeOCH2COOH (4.73)d H3O1 (21.44) d 6 pKa CH = 14.47 5.20 1.94 × 1024 2.10 × 1025 6.32 × 1025 6.01 × 1026 9.8 × 1027 0.98 2.80 2.2 × 102 2.19 × 104 3.3 × 104 2.9 × 108 4a b pKa CH = 12.50 7.84 × 101 3.49 × 1023 8.61 × 1023 6.01 × 1025 0.16 4.1 × 102 3.2 × 104 1.9 × 108 5a b pKa CH = 10.40 5.24 × 102 4.88 × 1022 1.71 × 1023 2.91 × 1023 1.71 × 1025 8.5 × 1023 6.0 × 1022 1.6 8.6 × 101 4.3 × 105 a k1 OH, k1 B, k� 2 BH and k� 2 H in units of M21 s21; k1 H2O and k� 2 H2O in units of s21.b Ref. 19. c Numbers in parentheses are pKa of conjugate acid. d pKa. ¶ The secondary KSIE is estimated based on (kH/kD)sec = (f2 L2O/fQL2)a with fL2O and fOL2 being the respective fractionation factors.12 It is assumed that the fractionation factors in 50% MeCN–50% L2O are the same as in pure L2O, i.e. 1.0 for L2O and 0.5 for OL2 (ref. 12); a is assumed to be 0.5.J. Chem. Soc., Perkin Trans. 2, 1997 1645 in the case of 6 the mechanism of Scheme 3 might be operative, or the two mechanisms might be in competition with each other; no definite conclusion is possible at this point.Nevertheless, there are a number of features which seem to favor Scheme 4. The first is an analogy with the reactions of 4a and 4b. The second is that in the case of 5a, where formation of PhCH]] CHOMe as the hydrolysis product is difficult to reconcile with Scheme 3 and hence Scheme 4 is the only reasonable mechanism, the KSIE is about as small as for 6.The third is that in the absence of water, base catalyzed decomposition of 6 leads to 2,3-dihydrofuran,9 which implies a mechanism like Scheme 4 and not Scheme 3. In acidic solution, where the k� 2 HaH1 term [eqn. (2)] or k1 H2O term [eqn. (1)] is dominant, no reliable data for the KSIE could be obtained for 6 but the KSIE for hydrolysis of 4a was determined as kobsd(H2O)/kobsd(D2O) = 3.1 (average value). Interpretation in terms of Scheme 4 yields eqn.(11), while interpretkobsd( H2O) kobsd(D2O) = Ka CH(H2O)k� 2 H Ka CD(D2O)k� 2 D (11) ation in terms of Scheme 3 yields eqn. (12). With respect to eqn. kobsd(H2O) kobsd(D2O) = k1 H2O k1 D2O (12) (11), the k� 2 H/k� 2 D ratio may be estimated as follows. The K1 OH/K1 OD ratio was found to be 0.317 based on kinetic measurements.1 Assuming that pKa D2O 2 pKa H2O = 0.86 as in pure water 15 yields Ka CH(H2O)/Ka D2O(CD) = (K1 OH/K1 OD)(Ka H2O/Ka D2O) = 2.3.This, then, affords k� 2 H/k� 2 D = 1.35 via eqn. (11). The k� 2 H/k� 2 D ratio is a composite of a primary KIE and a secondary KSIE. The latter can be estimated to be about 0.69,|| leaving a primary KIE of about 2.24. This is quite a small value but possibly consistent with a very unsymmetrical transition state 16 and/or with a concerted process 16b where proton transfer to the carbene carbon of 4a2 is coupled to carbon–metal bond cleavage. If Scheme 3 prevails, we have k1 H2O/k1 D2O = 3.1.This is the same as the k1 H2O/k1 D2O ratio in the hydrolysis of 13 4 and close to corresponding ratios for reactions of numerous electrophiles with water;13 these isotope effects suggest a mechanism where two water molecules are involved,13 e.g. one acting as nucleophile and the other as base catalyst, as in 15. As shown by the O C O (CO)5Cr H H B (CO)5Cr C CH3 OR O H O H H H 15 d+ d– d– d+ 16 Table 3 Kinetic solvent isotope effects on the hydrolysis of Fischer carbene complexes at 25 8C.Interpretation in terms of Scheme 3 and Scheme 4 Scheme 3 Scheme 4 4a 4b 6 5a k1 OH/k1 OD ca. 1.32 a ca. 2.22 a ca. 0.98 d ca. 0.88 a (k� 2 H2O/k� 2 D2O)exp 4.15 b 7.0 b 3.06 e 2.78 b (k� 2 H2O/k� 2 D2O)prim ca. 2.93 c ca. 4.95 c ca. 2.17 c ca. 1.97 c a Obtained from eqn. (7) after correcting for K1 OD/K1 OH = 3.15. b Obtained directly from eqn. (9). c Estimated assuming a secondary KSIE of 1.41, see text. d Obtained directly from eqn. (8). e Obtained from eqn.(10) after correcting for K1 OH/K1 OD = 0.317. || The secondary KSIE is estimated based on (kH/kD)sec = (f2 L3O1/fL2O)a with fL3O1 and fL2O being the respective fractionation factors.12 They are assumed to be the same as in pure L2O, i.e. 1.0 for L2O and 0.69 for L3O1.12 a = 0.68 is based on the Brønsted plot discussed below. preceding analysis no definite choice between Scheme 4 and 3 can be made with respect to the mechanism of the pHindependent pathway in acidic solution. Rate constants The individual rate constants calculated from the experimental data for Schemes 3 and 4 are summarized in Table 2 along with the corresponding rate constants for the hydrolysis of 4a and 5a.Fig. 4 shows a Brønsted plot of the rate constants for the k� 2 steps according to Scheme 4. The points for the two tertiary ammonium ions and methoxyacetic acid define a straight line of slope a = 0.68 ± 0.05; there is a small positive deviation for acetic acid and a large positive deviation for water while the point for H3O1 is on the line.The positive deviation of the water point and the fact that the point for H3O1 is on the line contrast with the more common observatio from most Brønsted plots.17 It suggests that, due to the bulkiness of the buffer acids, their k� 2 BHvalues are strongly depressed by steric crowding in the transition state. The slight positive deviation for acetic acid may, at least in part, also reflect a smaller steric effect for this less bulky acid.The data for 4a and 5a are more limited but the corresponding Brønsted plots (not shown) display the same features as Fig. 4. An alternative Brønsted plot for base catalysis (not shown) may be constructed based on the rate constants (k1) for Scheme 3. It yields a straight line of slope b = 0.32 ± 0.05 defined by the two amines, methoxyacetate ion and water, and shows a small positive deviation for AcO2 and a large positive deviation for OH2.Again the positive deviations can be understood in terms of steric hindrance, this time of base catalysis by the bulky buffer bases. The rather substantial steric effect which manifests itself in both Brønsted plots is probably more easily understood in terms of Scheme 4 than of Scheme 3. This is because proton transfer from the buffer acid to 62, 4a2 or 5a2 is likely to be direct, i.e. not to involve a bridging water molecule.18 Such direct transfers require a closer approach of the buffer molecule to the carbene complex in the transition state than in a transition state like 16 for base catalyzed water attack.This reasoning is suggestive rather than compelling but it is in agreement with the evidence from kinetic solvent isotope effects, especially for 4a and 4b. On the other hand, no conclusion can be reached, based on the Brønsted plots, regarding the question of whether the hydrolysis pathway might proceed by a different mechanism in acidic solution (Scheme 3) than in basic solution. The dependence of the rate constants on the specific carbene complex calls for comment.In a previous paper 1 the low acidity Fig. 4 Brønsted plot of the k� 2 rate constants (d: k� 2 BH and k� 2 H; s: k� 2 H2O) according to Scheme 4. Slope = 0.68 ± 0.05 based on k� 2 BH for Et3NH1, N-MeMorH1 and MeOCH2COOH.1646 J. Chem. Soc., Perkin Trans. 2, 1997 of 6 (pKa = 14.47) compared to that of 4a (pKa = 12.50) 19 was shown to be mainly the result of stabilization of the acid form (6) rather than to a destabilization of the anion (62).The extra stabilization of 6 was primarily attributed to enhanced pdonation from the oxygen to the carbene carbon (6±), a consequence of the ring structure which locks the oxygen into a position for better p-overlap than is the case for 4a. Irrespective of whether Scheme 3 or 4 prevails, the substrate dependence of the various rate constants lends further support to this conclusion.Assuming Scheme 4 is the correct mechanism, we note that the rate constants for the reaction of 62 and 4a2 with proton donors are quite similar but substantially larger than for the reactions of 5a2 with the same proton donors. This lower reactivity of 5a2 may be attributed to the stabilization of 5a2 by the phenyl group, the same stabilization that is responsible for the higher acidity of 5a (pKa CH = 10.40) 19 compared to that of 4a (pKa CH = 12.50).On the other hand, the comparable reactivities of 62 and 4a2 indicate comparable stabilities of 62 and 4a2, consistent with the notion that the pKa difference between 6 and 4a is not the result of different stabilities of 62 and 4a2, but the result of different stabilities of 6 and 4a. If Scheme 3 is operative, the same conclusion emerges: nucleophilic attack on 4a and 5a occurs with comparable rates and is considerably faster than attack on 6, reflecting the fact that only 6 enjoys extra stabilization from enhanced p-donation by the oxygen atom (6±).Conclusions The hydrolysis of 6 yields 2-hydroxybutanal (8) which rapidly cyclizes to the hemiacetal 7. This finding is consonant with results for the hydrolysis of 4a and 4b which yields acetaldehyde and the respective alcohol under similar conditions. The most likely hydrolysis mechanism is that shown in Scheme 4; as with 4a and 4b, it is consistent with the kinetic solvent isotope effect although this isotope effect is less conclusive for the reaction of 6 than it was for the reactions of 4a and 4b.Tentative support for the mechanism of Scheme 4 also comes from a consideration of steric effects on the hydrolysis rate constants. Irrespective of whether Scheme 3 or 4 prevails, the difference between the individual rate constants (k1 for Scheme 3 or k� 2 for Scheme 4) for 6 and 4a supports an earlier conclusion that the lower acidity of 6 compared to that of 4a is mainly the result of enhanced p-donation by the oxygen in 6 (6±). Experimental Materials (2-Oxacyclopentylidene)pentacarbonylchromium(0), 6, was a gift from Professor Hegedus; it was recrystallized from dry pentane before use, mp 63.5–65.0 8C (lit. 63.5–65.0 8C20). [2H2]6 was prepared by dissolving known amounts of 6 in 1.0 ml of 70% CD3CN–30% D2O in the presence of catalytic amounts of NaOD (<0.005 M). H/D exchange was monitored by 1H NMR spectroscopy until no residual signal for the acidic protons of 6 at 3.6 ppm was observed. The resulting [2H2]6 was not isolated, i.e.the NMR solutions were directly used for the kinetic experiments. (Methoxymethylcarbene)pentacarbonylchromium( 0), 4a, and [2H3]4a were available from a previous study.3 Triethylamine, and N-methylmorpholine were refluxed over sodium for at least 5 h in an argon atmosphere and were then fractionally distilled. Acetic acid and methoxyacetic acid were used as received.HCl and KOH solutions were prepared by diluting prepackaged stock solutions (Baker Analytical). KOD was prepared by dissolving KOH in D2O; the concentration of O (CO)5Cr 6± _ + the resulting solution was determined by potentiometric titration. Acetonitrile was purchased from Fisher Scientific and used as received. Water was obtained from a Millipore water purification system. Water and acetonitrile used in the hydrolysis experiments at pH <7.43 were degassed by the freeze–pump–thaw method to minimize substrate oxidation by dissolved oxygen over long reaction times.CD3CN, D2O and DCl were used as received. Product analysis by NMR All 1H and 13C NMR spectra for the product study were recorded in 75% CD3CN–25% D2O with a 250 MHz Bruker spectrometer. Solutions and pH measurements All kinetic experiments were conducted in 50% CH3CN–50% H2O (v/v) or 50% CH3CN–50% D2O (v/v) solutions at 25 8C, I = 0.1 M (KCl). All pH measurements were made with an Orion 611 pH-meter equipped with a glass electrode and a ‘SureFlow’ (Corning) reference electrode.Actual pH values were calculated by adding 0.18 to the measured pH, according to Allen and Tidwell 21 The pKa values of triethylamine and Nmethylmorpholine were determined by standard potentiometric techniques. The pKa values of the other buffers were known from a previous study.19 Kinetics Typical substrate concentrations were 5–10 × 1025 M. Rates were measured in a Perkin-Elmer Lambda 2 or Hewlett- Packard 8452A diode array UV–VIS spectrophotometer.Kinetics were followed by monitoring the disappearance of the substrate at 364 nm. Rate constants (kobsd) were obtained by computer fit programs (Applied Photophysics and Enzfitter 22). Rates at pH <10.31 were very low and hence kobsd was determined by the initial rates method: reactions were monitored for 1–5 h, after which enough 2M KOH was added to neutralize the acidic component of the buffer and increase the pH of the solution to ª12.Reactions were then further monitored until the infinity value (OD•) was reached. Pseudo-first-order rate constants were calculated according to eqn. (13), where S is the kobsd = S/DOD0 (13) slope of the plot of OD vs. time for the first 1–5 h, and DOD0 = OD• 2 OD0. Acknowledgements This research was supported by Grant No. CHE-9307659 from the National Science Foundation and the donors of the Petroleum Research Fund administered by the American Chemical Society, Grant No. 30444-AC4. We thank Professor Louis Hegedus us with (2-oxacyclopentylidene)- pentacarbonylchromium(0). References 1 Part 10: C. F. Bernasconi and A. E. Leyes, J. Am. Chem. Soc., 1997, 119, 5583. 2 R. Aumann, P. Hinterding, C. Krüger and P. Goddard, J. Organomet. Chem., 1993, 459, 145. 3 C. F. Bernasconi, F. X. Flores and W. Sun, J. Am. Chem. Soc., 1995, 117, 4875. 4 C. F. Bernasconi, F. X. Flores and K. W. Kittredge, J.Am. Chem. Soc., 1997, 119, 2103. 5 E. O. Fischer and A. Maasböl, Chem. Ber., 1967, 100, 2445. 6 C. F. Bernasconi and W. Sun, Organometallics, 1995, 14, 5615. 7 A. E. Leyes, Ph.D. Thesis, University of California, Santa Cruz, 1996. 8 (a) E. O. Fischer and A. Maasböl, J. Organomet. Chem., 1968, 12, P15; (b) E. O. Fischer and W. Plabst, Chem. Ber., 1974, 107, 3326.J. Chem. Soc., Perkin Trans. 2, 1997 1647 9 C. P. Casey and R. L. Anderson, J. Chem. Soc., Chem. Commun., 1975, 895. 10 (a) M. L. Bender, R. D. Ginger and J. P. Unik, J. Am. Chem. Soc., 1958, 80, 1044; (b) S. A. Shain and J. Kirsch, J. Am. Chem. Soc., 1968, 90, 5848; (c) D. DeTar, J. Am. Chem. Soc., 1982, 104, 7205; (d) J. F. Marlier, J. Am. Chem. Soc., 1993, 115, 5953. 11 (a) F. Terrier, Chem. Rev., 1982, 82, 78; (b) C. F. Bernasconi, J. Fassberg, R. B. Killion, D. F. Schuck and Z. Rappoport, J. Am. Chem. Soc., 1991, 113, 4937. 12 (a) R. L. Schowen, Prog. Phys. Org. Chem., 1972, 9, 275; (b) F. J. Alvarez and R. L. Schowen, in Isotopes in Organic Chemistry, ed. E. Buncel and C. C. Lee, Elsevier, New York, 1987, vol. 7, p. 1. 13 P. M. Laughton and R. E. Robertson, in Solute–Solvent Interactions, ed. J. F. Coetzee and C. D. Ritchie, Marcel Dekker, New York, 1969, p. 399. 14 T. E. Casamassina and W. P. Huskey, J. Am. Chem. Soc., 1993, 115, 14. 15 V. Gold and B. M. Lowe, J. Chem. Soc. (A), 1967, 936. 16 (a) F. H. Westheimer, Chem. Rev., 1961, 61, 265; (b) R. A. More O’Ferrall, in Proton Transfer Reactions, ed. E. Caldin and V. Gold, Wiley, New York, 1975, p. 201. 17 (a) A. J. Kresge, Chem. Soc. Rev., 1973, 2, 475; (b) D. J. Hupe and W. P. Jencks, J. Am. Chem. Soc., 1977, 99, 451; (c) W. P. Jencks, S. R. Brant, J. R. Gandler, G. Fendrich and C. Nakamura, J. Am. Chem. Soc., 1982, 104, 7045. 18 (a) F. Hibbert, Compr. Chem. Kinet., 1977, 8, 97; (b) W. J. Albery, in Proton Transfer Reactions, ed. E. Caldin and V. Gold, Wiley, New York, 1975, p. 285; (c) R. A. Bednar and W. P. Jencks, J. Am. Chem. Soc., 1985, 107, 7126. 19 C. F. Bernasconi and W. Sun, J. Am. Chem. Soc., 1993, 115, 12 526. 20 C. P. Casey and R. L. Anderson, J. Organomet. Chem., 1974, 73, C28. 21 A. D. Allen and T. T. Tidwell, J. Am. Chem. Soc., 1987, 109, 2774. 22 Program by R. J. Leatherbarrow, distributed by BIOSOFT, 22 Hills Road, Cambridge, UK CB2 1JP. Paper 7/02286G Received 28th March 1997 Accepted 21th April 1997

ISSN:1472-779X    

DOI:10.1039/a702286g

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1649 Temperature dependent inversion of enantiomer selectivity in the complexation of optically active azophenolic crown ethers containing alkyl substituents as chiral barriers with chiral amines 1 Keiji Hirose, Junichi Fuji, Kimiko Kamada, Yoshito Tobe and Koichiro Naemura * Department of Chemistry, Faculty of Engineering Science, Osaka University, Toyonaka, Osaka 560, Japan Azophenolic crown ethers (S,S)-1, (R,R)-2 and (S,S)-3 have been prepared in enantiomerically pure forms by using (S)-1-(19-adamantyl)ethane-1,2-diol, (R)-3,3-dimethylbutane-1,2-diol and (S)-propane-1,2-diol, respectively, as chiral subunits, and the association constants for their complexes with chiral amines have been determined by 1H NMR or UV–VIS spectroscopic methods at various temperatures.The enantiomer selectivities of crown ethers (S,S)-1 and (R,R)-2 in complexation with 2-aminopropan-1-ol are reversed at ca. 6 8C and increase with increasing temperature above the isoenantioselective temperature.Enantiomer recognition has been widely studied in various types of chemical and biochemical transformations.2 While it is the generally accepted view that lower temperatures enhance enantiomer discrimination in chiral processes, a few papers have reported that the enantiomer selectivity of a chiral process increased with increasing temperature; the enantiospecificity ratio of alcohol dehydrogenase-catalysed oxidation of secondary alcohols increased with increasing temperature above the ‘racemic temperature’,3 the enantiomeric purities of compounds resolved by GLC using a chiral stationary phase was enhanced with increasing column temperature 4 and the optical yield of photochemically induced enantiomeric isomerization improved with increasing irradiation temperature.5 Various types of optically active crown ethers have been prepared and their enantiomer recognition behaviour in complexation with chiral guests has been well documented.6 But as far as we know there has been no report of the temperature dependent inversion of the enantiomer selectivity in complexation of a crown ether with a chiral amine in solution.It is important to seek information on how the temperature might influence the enantiomer selectivity in complexation of crown ethers with chiral amines because temperature dependent inversion of the enantiomer selectivity in complexation is a fundamental problem for the estimation of the configuration of guest species on the basis of the enantiomer selectivity in host–guest complexation.We here describe the preparation of optically active azophenolic crown ethers (S,S)-1, (R,R)-2 and (S,S)-3 containing alkyl substituents as chiral barriers by using (S)-1-(19-adamantyl)ethane-1,2- diol 6, (R)-3,3-dimethylbutane-1,2-diol 8 and (S)-propane-1,2- diol 10, respectively. These crown ethers also possess a phenol moiety bearing an intraannular OH group as a binding site for neutral amines and the additional 2,4-dinitrophenylazo group at its para-position which acts not only as a chomophore but also to enhance the binding ability towards neutral amines.7 The enantiomer selectivity in complexation of the crown ethers 1, 2 and 3 with chiral 2-aminoethanol derivatives in chloroform was evaluated at various temperatures to observe the temperature dependent reversal of the enantiomer selectivity in complexation of crown ethers (S,S)-1 and (R,R)-2 with 2- aminopropan-1-ol at ca. 6 8C. The thermodynamic parameters for complexation were calculated on the basis of the association constants of the complexes determined at different temperatures. Results and discussion We previously reported the preparation of the racemic diol (±)- 6 from 1-adamantylacetic acid but in rather low yield (37% overall) 8 and, therefore, it was our first task to modify the preparation of the diol (±)-6. Treatment of ethyl adamantane- 1-carboxylate with sodium hydride and dimethyl sulfoxide (DMSO) gave compound 4, which was reacted with acetic anhydride and sodium acetate 9 to give the compound 5.Reduction of compound 5 with LiAlH4 gave the diol (±)-6 in 75% overall yield from ethyl adamantane-1-carboxylate. The chiral subunit (S)-7, [a]D 23 120.8 (1021 deg cm2 g21) (CHCl3), was pre- O O O O O OH N (H3C)3C C(CH3)3 N NO2 NO2 N N NO2 NO2 O O O O O OH O O O O O OH CH3 H3C N N NO2 NO2 (S,S)-3 (S,S)-1 (R,R)-21650 J.Chem. Soc., Perkin Trans. 2, 1997 O O O HO OH O O O O O OR OMe O O O HO OH (H3C)3C C(CH3)3 OH HO O O OMe OMe CH3 H3C O O O O O OR OMe C(CH3)3 (H3C)3C O O O O O OR OMe CH3 H3C HO OR H3C RO OH HO OR (H3C)3C CCH2SOCH3 O CHCSCH3 OAc (R,R)-17 R=CH3 (R,R)-18 R=H (S,S)-12 (S,S)-19 R=CH3 (S,S)-20 R=H (R,R)-13 (S,S)-14 (R)-8 R=H (R)-9 R=CPh3 (S)-6 R=H (S)-7 R=CPh3 (S,S)-15 R=CH3 (S,S)-16 R=H 4 5 (S)-10 R=H (S)-11 R=THP O pared from the racemate (±)-6 according to the reported procedures:8 optical resolution of the racemate (±)-6 using (2)-camphanic chloride as a resolving agent followed by the protection of the primary hydroxy group of the resulting diol (S)-6, [a]D 24 118.0 (ethanol).The preparation of the crown ethers (S,S)-1 and (R,R)-2 having the substituents located near the diethylene glycol bridge and the homotopic faces was carried out stepwise; chiral subunits of the same chirality were linked successively with the diethylene glycol unit and with the m-phenylene unit.Condensation of 2 mol equiv. of the chiral subunit (S)-7 with diethylene glycol bis(toluene-p-sulfonate) in the presence of sodium hydride in tetrahydrofuran (THF) gave the 1,11-O-blocked tetraethylene glycol derivative in 74% yield, which was deprotected with methanol containing toluene-p-sulfonic acid to give the C2-diol (S,S)-12, [a]D 22 14.66 (CHCl3), in 98% yield. High dilution condensation of the diol (S,S)-12 with 1,3-bis- (bromomethyl)-2,5-dimethoxybenzene in THF containing sodium hydride and potassium tetrafluoroboranuide gave the crown ether (S,S)-15, [a]D 19 144.3 (CHCl3), in 66% yield.The intraannular methyl group of the crown ether (S,S)-15 was selectively cleaved with sodium ethanethiolate in DMF at 100 8C10 to give the phenolic crown ether (S,S)-16, [a]D 25 123.2 (CHCl3), in 64% yield. Oxidation of the phenolic crown ether (S,S)-16 with cerium(IV) ammonium nitrate (CAN) in acetonitrile gave the corresponding quinone which was immediately treated with 2,4-dinitrophenylhydrazine in a mixture of conc.H2SO4, ethanol and methylene dichloride to give the azophenolic crown ether (S,S)-1 in 82% yield. The diol (R)-8, [a]D 28 225.3 (CHCl3), which was prepared according to the procedures described in the literature 11 was reacted with triphenylmethyl chloride in the presence of triethylamine and 4-dimethylaminopyridine to give the chiral subunit (R)-9, [a]D 28 216.5 (CHCl3), in 70% yield.The C2-diol (R,R)-13, [a]D 25 22.73 (CHCl3), was derived from the chiral subunit (R)-9 in 28% yield by condensation with diethylene glycol bis(toluene-p-sulfonate) followed by deprotection. High dilution condensation of the C2-diol (R,R)-13 with 1,3- bis(bromomethyl)-2,5-dimethoxybenzene gave the crown ether (R,R)-17, [a]D 24 227.0 (CHCl3), in 48% yield. Demethylation of the crown ether (R,R)-17 gave the phenolic crown ether (R,R)-18 in 95% yield, which was transformed to the azophenolic crown ether (R,R)-2 in 50% yield.According to the published route,12 protection of the hydroxy group followed by LiAlH4 reduction, the chiral subunit (S)-11 was prepared from ethyl (S)-lactate as a mixture of two diastereoisomers. Condensation of the chiral subunit (S)-11 with 1,3-bis(bromomethyl)-2,5-dimethoxybenzene followed by deprotection with ethanol and pyridinium toluene-p-sulfonate gave the C2-diol (S,S)-14, [a]D 20 119.2 (CHCl3), in 55% yield.Ring closure of the diol (S,S)-14 with diethylene glycol bis(toluene-p-sulfonate) under high dilution conditions gave the crown ether (S,S)-19, [a]D 24 122.1 (CHCl3), in 50% yield, which was transformed to the azophenolic crown ether (S,S)-3 in 72% yield via the phenolic crown ether (S,S)-20. Enantiomer recognition of the crown ethers (S,S)-1, (R,R)-2 and (S,S)-3 towards chiral amines: 2-amino-3-methylbutan-1-ol 21, 2-aminopropan-1-ol 22 and 1-aminopropan-2-ol 23 was evaluated at various temperatures.Association constants for complexes of (S,S)-1 with 21, 22 and 23, (R,R)-2 with 21, 22 and 23 and (S,S)-3 with 21 were calculated by the non-linear least-squares method on the basis of the 1H NMR spectral data in CDCl3. As association constants for complexes of (S,S)-3 with 22 and 23 were so large that it was difficult to get accurate data at lower temperature by 1H NMR titration, they NH2 OH NH2 Me OH OH Me NH2 Me Me 21 22 23J.Chem. Soc., Perkin Trans. 2, 1997 1651 were determined by the Rose–Drago method13 on the basis of the absorption in the UV–VIS spectrum in CHCl3. The observed Ka-values of the complexes and the thermodynamic parameters for complexation calculated on the basis of Kavalues determined at different temperatures are summarized in Tables 1 and 2. The results show that the association constants of all complexes increased with decreasing temperature but that Ka R/Ka S values were influenced by changes in temperature.In Figs. 1, 2 and 3 are plotted DDG values (DGS 2 DGR) of complexation of the crown ethers (S,S)-1, (S,S)-2 and (S,S)-3, respectively, with the amines 21, 22 and 23 as a function of temperature. The most important features shown in Figs. 1 and 2 are that the sign of DDG values for the complexation of the crown ethers (S,S)-1 and (S,S)-2 with 22 reverses at ca. 6 8C; the isoenantioselective temperature (Tiso) and the S-selectivity towards 22 increased with increasing temperature above Tiso.The enantiomer selectivities in complexation of the other com- Fig. 1 Temperature dependence of DDG (DGS 2 DGR) for the complexation of crown ether (S,S)-1 with amines in chloroform: 2-amino- 3-methylbutan-1-ol (n), 2-aminopropan-1-ol (r) and 1-aminopropan- 2-ol (d) Fig. 2 Temperature dependence of DDG (DGS2DGR) for the complexation of crown ether (S,S)-2 with amines in chloroform: 2- amino-3-methylbutan-1-ol (n), 2-aminopropan-1-ol (r) and 1-aminopropan- 2-ol (d) Fig. 3 Temperature dependence of DDG (DGS 2 DGR) for the complexation of crown ether (S,S)-3 with amines in chloroform: 2-amino- 3-methylbutan-1-ol (n), 2-aminopropan-1-ol (r) and 1-aminopropan- 2-ol (d) binations of crown ethers and amines, except for that of the crown ether (S,S)-2 and 23, showed also an unambiguous temperature dependent enantiomer selectivity; reversal of the selectivity was not observed within the experimental temperature range because of high Tiso values which were estimated by calculation from DH and DS values and are listed in Tables 1 and 2.Next, on the basis of CPK molecular models, we give an explanation for the enantiomer selectivities observed below Tiso which are governed by 2DR,SDH in terms of steric interactions between ligands of the amine and the steric barriers of the crown ether in the complex. For predicting the geometry of the complex of a phenolic crown ether with a 2-aminoethanol derivative, we use as a working hypothesis 14 that the phenolate oxygen atom necessarily serves as a binding site for amines and the hydroxymethyl group of 2-aminoethanol derivatives occupies the area near the phenolate oxygen atom to form the fourth hydrogen bond between the phenolate oxygen atom and the hydroxy group of the guest.Judging from CPK molecular models and the observed enantiomer selectivities, we infer that the pseudo-equatorial substituent at C-5 (open circle in the geometries) makes the methylene group at C-4 and the methine group at C-5 an effective steric barrier on the b-face of the complex: ‘the ethylenoxy barrier’ (shaded ellipse in the geometries 24–29).Figs. 1, 2 and 3 show that the crown ethers (S,S)-1 and (S,S)- 2 having bulky substituents showed better complementarity to (S)-21 than to (R)-21 but the crown ether (S,S)-3 having the small methyl substituents showed the reverse complementarity below Tiso. The predicted geometries 24 and 25 are illustrated for the complexes of the (S,S)-crown ethers 1, 2 and 3 with (R)- 21 and with (S)-21, respectively, on the basis of CPK molecular models of the complexes.The S-selectivity of the crown ethers (S,S)-1 and (S,S)-2 towards 21 is rationalized in terms of steric repulsion between the bulky substituent and the bulky isopropyl group destabilizing the (S,S)-crown ether–(R)-21 complex with the geometry 24 (shaded circle indicates the 1-adamantyl group or the tert-butyl group). As both the chiral substituent barrier and the ligand of the amine are bulky, the crown ethers 1 and 2 recognize 21 by the chiral substituent barrier.On the other hand, the steric interaction between the small substituent and the ligand of the amine is not dominant and the crown ether 3 recognizes 21 by ‘the ethylenoxy barrier’; steric repulsion between ‘the ethylenoxy barrier’ and the isopropyl group made the (S,S)-crown ether–(S)-21 complex with the geometry 25 (shaded circle indicates the methyl group) less stable than the diastereoisomeric complex.Analogously, it is assumed that the crown ethers 1, 2 and 3 recognize 22 by ‘the ethylenoxy barrier’ because the steric interaction between the small ligand of the amine and the substituent of the crown ether is not dominant. The R-selectivity of all (S,S)-crown ethers towards 22 observed below Tiso is rationalized from the geometries 26 and 27 which are illustrated for the stable (S,S)-crown ether–(R)-22 complex and the less stable (S,S)-crown ether–(S)-22 complex, respectively.The Sselectivity of the (S,S)-crown ethers 1 and 2 towards 22 observed above Tiso is not explicable. The geometries 28 and 29 are illustrated for the complexes of the (S,S)-crown ethers with (R)-23 and with (S)-23, respectively. The R-selectivity of all (S,S)-crown ethers towards 23 is assumed to be due to a steric repulsion between the methyl group of the chiral centre of 23 and the substituent of the crown ether making the (S,S)-crown ether–(S)-23 complex with the geometry 29 less stable than the diastereoisomeric complex with the geometry 28.An inversion of the sign of enantiomer selectivity dependent upon temperature is predictable since the enthalpy change and the entropy change compensate each other, as can be seen in Tables 1 and 2, and the entropy change contributes to the stability of the complex.15 The previous failure to observe such a1652 J.Chem. Soc., Perkin Trans. 2, 1997 Table 1 Association constants for the complexes and thermodynamic parameters for complexation of crown ethers (S,S)-1 and (R,R)-2 Crown Ka/M21 ether (S,S)-1 (S,S)-1 (R,R)-2 (R,R)-2 (S,S)-1 (S,S)-1 (R,R)-2 (R,R)-2 (S,S)-1 (S,S)-1 (R,R)-2 (R,R)-2 Amine (R)-21 (S)-21 (R)-21 (S)-21 (R)-22 (S)-22 (R)-22 (S)-22 (R)-23 (S)-23 (R)-23 (S)-23 230 8C (7.90 ± 1.69) × 103 (1.82 ± 0.98) × 104 (1.53 ± 0.48) × 104 (8.64 ± 1.58) × 103 (2.40 ± 1.05) × 104 (6.74 ± 1.43) × 103 (7.49 ± 1.27) × 103 (1.34 ± 0.74) × 104 (1.35 ± 0.32) × 104 (2.65 ± 0.21) × 103 (3.37 ± 0.63) × 103 (4.54 ± 0.52) × 103 215 8C (1.74 ± 0.10) × 103 (3.51 ± 0.48) × 103 (2.82 ± 0.21) × 103 (2.01 ± 0.09) × 103 (4.32 ± 0.50) × 103 (2.49 ± 0.24) × 103 (2.24 ± 0.12) × 103 (2.70 ± 0.26) × 103 (3.72 ± 0.68) × 103 (1.05 ± 0.11) × 103 (9.85 ± 0.74) × 102 (1.63 ± 0.22) × 103 0 8C (4.68 ± 0.20) × 102 (7.11 ± 0.20) × 102 (5.93 ± 0.32) × 102 (4.49 ± 0.24) × 102 (9.56 ± 0.98) × 102 (9.17 ± 0.54) × 102 (7.63 ± 0.35) × 102 (7.80 ± 0.53) × 102 (9.97 ± 0.87) × 102 (3.50 ± 0.28) × 102 (3.60 ± 0.20) × 102 (5.77 ± 0.10) × 102 15 8C (1.64 ± 0.04) × 102 (2.17 ± 0.06) × 102 (1.60 ± 0.07) × 102 (1.35 ± 0.04) × 102 (2.19 ± 0.12) × 102 (2.84 ± 0.06) × 102 (2.35 ± 0.06) × 102 (2.41 ± 0.15) × 102 (2.24 ± 0.22) × 102 (1.29 ± 0.04) × 102 (1.07 ± 0.03) × 102 (1.82 ± 0.05) × 102 30 8C (4.77 ± 0.16) × 101 (6.66 ± 0.19) × 101 (4.38 ± 0.08) × 101 (4.31 ± 0.32) × 101 (4.38 ± 0.81) × 101 (6.94 ± 0.41) × 101 (8.00 ± 0.43) × 101 (5.83 ± 0.08) × 101 (6.57 ± 1.21) × 101 (4.98 ± 0.34) × 101 (2.51 ± 0.13) × 101 (3.29 ± 0.09) × 101 DH/kJ mol21 (25.35 ± 0.26) × 101 (25.96 ± 0.62) × 101 (26.20 ± 0.45) × 101 (25.66 ± 0.49) × 101 (26.63 ± 0.23) × 101 (24.80 ± 0.79) × 101 (24.81 ± 0.34) × 101 (25.66 ± 0.15) × 101 (25.71 ± 0.64) × 101 (24.26 ± 0.41) × 101 (25.10 ± 0.62) × 101 (25.11 ± 1.06) × 101 DS/J K21 mol21 (21.45 ± 0.11) × 102 (21.63 ± 0.26) × 102 (21.74 ± 0.18) × 102 (21.56 ± 0.20) × 102 (21.87 ± 0.09) × 102 (21.22 ± 0.33) × 102 (21.22 ± 0.14) × 102 (21.53 ± 0.06) × 102 (21.54 ± 0.27) × 102 (21.08 ± 0.17) × 102 (21.40 ± 0.26) × 102 (21.37 ± 0.44) × 102 Tiso a/8C 62.1 32.5 5.0 5.3 45.3 — a The predicted isoenantioselective temperatures are calculated from DH and DS values.J.Chem. Soc., Perkin Trans. 2, 1997 1653 Table 2 Association constants for the complexes and thermodynamic parameters for complexation of crown ethers (S,S)-3 Ka/M21 Tiso a/ Amine (R)-21 (S)-21 (R)-22 (S)-22 (R)-23 (S)-23 15 8C (7.05 ± 0.44) × 103 (2.13 ± 0.25) × 103 (1.99 ± 0.01) × 104 (7.27 ± 0.12) × 103 (1.42 ± 0.01) × 104 (7.75 ± 0.03) × 103 25 8C (2.62 ± 0.07) × 103 (9.12 ± 0.42) × 102 (7.04 ± 0.33) × 103 (3.20 ± 0.07) × 103 (5.10 ± 0.09) × 103 (3.18 ± 0.01) × 103 35 8C (1.17 ± 0.12) × 103 (4.66 ± 0.09) × 102 (2.64 ± 0.02) × 103 (1.17 ± 0.01) × 103 (1.92 ± 0.07) × 103 (1.36 ± 0.42) × 102 45 8C (5.20 ± 0.06) × 102 (2.22 ± 0.05) × 102 (1.12 ± 0.02) × 103 (5.63 ± 0.11) × 102 (9.05 ± 0.10) × 102 (6.06 ± 0.19) × 101 DH/kJ mol21 (26.58 ± 0.41) × 101 (25.68 ± 0.42) × 101 (27.33 ± 0.27) × 101 (26.62 ± 0.82) × 101 (27.06 ± 0.66) × 101 (26.47 ± 0.13) × 101 DS/J K21 mol21 (21.55 ± 0.16) × 102 (21.33 ± 0.16) × 102 (21.72 ± 0.10) × 102 (21.56 ± 0.31) × 102 (21.66 ± 0.25) × 102 (21.50 ± 0.05) × 102 8C 148 156 103 a The predicted isoenantioselective temperatures are calculated from DH and DS values.O O O O O H H H H HO H H3C N N NO2 NO2 O O O O O N CH3 H CH2 H H H HO + N NO2 NO2 O O O O O N H H3C CH2 H H H HO 25 N NO2 NO2 - 28 + - + O O O O O N CH H CH2 H H H HO N NO2 NO2 O O O O O N H CH CH2 H H H HO N NO2 NO2 H3C CH3 29 + H3C H3C - + 24 27 26 - O O O O O H H H H H HO CH3 H N N NO2 NO2 + - - temperature dependent inversion of the enantiomer selectivity in host–guest complexation in solution would appear to be due largely to the Tiso value generally being high and the association constant of the complex decreasing markedly with increasing temperature.The present results are the first observed examples of the temperature dependent reversal of the enantiomer selectivity in complexation of crown ethers with amines in solution. Experimental General 1H NMR spectra were obtained on JEOL GSX-270 and JEOL GSX-400 spectrometers for solutions in CDCl3 with SiMe4 as an internal standard, and J values are given in Hz. 13C NMR spectra were recorded on a JASCO JNM-MH-270 spectrometer and chloroform (dC 77.0) was used as a chemical-shift reference.1654 J.Chem. Soc., Perkin Trans. 2, 1997 UV and visible spectra were recorded on a Hitachi 330 spectrometer. Mass spectroscopic analyses were carried out on a JEOL JMS-DX303HF mass spectrometer using m-nitrobenzyl alcohol as a matrix. IR spectra were recorded on a Hitachi 260- 10 spectrometer. Elemental analyses were carried out on a Yanagimoto CHN-Corder, Type 2.HPLC analyses were carried out on a Shimazu GS 8A chromatograph equipped with a UV spectrophotometric detector (wavelength 254 nm) using an Inertsil ODS (GL Sciences) 250 mm × 4.6 mm column. Melting points were measured on a Yanagimoto micro melting point apparatus and are uncorrected. Optical rotations were measured using a JASCO DIP-40 polarimeter and [M]D-values are given in units of 1021 deg cm2 g21. The guest amines (R)- and (S)-2-amino-3- methylbutan-1-ol, (R)- and (S)-2-aminopropan-1-ol and (R)- and (S)-1-aminopropan-2-ol were purchased from Aldrich Chemical Company Inc.and used without further purification. 1-(19-Adamantyl)-2-methylsulfinylethanone 4 A mixture of DMSO (275 g, 3.52 mol) and sodium hydride (16.1 g, 670 mmol) was stirred for 30 min at 65–72 8C under a nitrogen atmosphere. After the evolution of hydrogen had ceased, the reaction mixture was cooled to room temperature. A solution of ethyl adamantane-1-carboxylate (46.0 g, 221 mmol) in dry THF (150 cm3) was added dropwise to the reaction mixture.The reaction mixture was stirred for 40 min at room temperature and then saturated aq. ammonium chloride (10 cm3) was added. The mixture was poured into ice–water, acidified (pH 3–4) with dilute hydrochloric acid and extracted with ethyl acetate. The combined extracts were washed with aq. sodium hydrogen carbonate and water and dried (K2CO3). Removal of the solvent gave a solid, which was recrystallized from hexane to give the title compound 4 (48.4 g, 93%); mp 104–105 8C; nmax(KBr)/cm21 2900, 2850, 1780, 1450, 1340, 1280, 1160, 1040, 1010, 960, 930, 810, 760 and 690; dH(270 MHz; CDCl3) 1.66–1.83 (12H, m, adamantyl CH2), 2.09 (3H, s, adamantyl CH), 2.72 (3H, s, SCH3), 3.78 (1H, d, J 15.0, COCH2SO) and 4.16 (1H, d, J 15.0, COCH2SO), MS (FAB) m/z 240 (M1) and 241 (MH1) (Found: C, 64.81; H, 8.3.C13H20O2S requires C, 64.96; H, 8.39%). S-Methyl 2-acetoxy-2-(19-adamantyl)thioacetate 5 A mixture of compound 4 (60.0 g, 250 mmol), acetic anhydride (540 g, 5.29 mol) and sodium acetate (45.1 g, 550 mmol) was stirred for 4 h at 100–110 8C.Then the reaction mixture was poured into ice–water and extracted with benzene. The combined extracts were washed with water and dried (Na2SO4). Evaporation of the solvent gave a solid, which was recrystallized from ethanol to give the title compound 5 (58.4 g, 83%); mp 86–87 8C; nmax(KBr)/cm21 2930, 2860, 1740, 1700, 1420, 1370, 1240, 1200, 1045 and 970; dH(270 MHz; CDCl3) 1.68–1.73 (12H, m, adamantyl CH2), 1.98 (3H, br s, adamantyl CH), 2.05 (3H, s, COCH3), 2.13 (3H, s, SCH3) and 6.21 (1H, s, CH); MS (FAB) m/z 282 (M1), 283 (MH1) (Found: C, 63.88; H, 7.8.C15H22O3S requires C, 63.80; H, 7.85%). (±)-1-(19-Adamantyl)ethane-1,2-diol 6 A solution of compound 5 (31.4 g, 111 mmol) in dry THF (200 cm3) was added slowly to a suspension of lithium aluminium hydride (6.19 g, 163 mmol) in dry THF (300 cm3) and then the mixture was refluxed for 3 h under a nitrogen atmosphere.The reaction mixture was cooled to 0–5 8C and then ethyl acetate (40 cm3) and saturated aq. ammonium chloride (20 cm3) were successively added. After the deposited solid had been removed by filtration, the solvent was evaporated under reduced pressure to give a solid, which was recrystallized from hexane to give the diol 6 (21.3 g, 97%); mp 128–130 8C; nmax(KBr)/cm21 3300, 2900, 2850, 1090, 1065, 1055 and 1030; dH(270 MHz; CDCl3) 1.52–1.88 (14H, m, adamantyl CH2 and OH), 1.99 (3H, s, adamantyl CH), 3.23 (1H, dd, J 2.8 and 9.0, CH2), 3.56 (1H, dd, J 9.0 and 11.0, CH2) and 3.75 (1H, dd, J 2.8 and 11.0, CH) (Found: C, 73.35; H, 10.2.C12H20O2 requires C, 73.43; H, 10.27%). Optical resolution of (±)-6. According to the procedure reported in our previous paper,8 reaction of the racemate (±)-6 (10.0 g, 50.9 mmol) with (2)-camphanic chloride (24.0 g, 111 mmol) in pyridine (40 cm3) gave a mixture of diastereoisomeric esters as a solid (18.4 g, 65%), [a]D 21 17.79 (c 1.14, acetone), which showed two double doublet signals due to the methine proton of the diol moiety at d 4.95 and 4.77 in its 1H NMR spectrum.The mixture was recrystallized three times from methanol until the signal at d 4.77 disappeared completely to give the diastereomerically pure ester as colorless needles (7.40 g, 26%); mp 193–195 8C; [M]D 24 1201.3 (c 1.01, acetone); nmax(KBr)/cm21 2960, 2910, 2850, 1780, 1750 and 1740 (Found: C, 68.99; H, 8.0.C32H44O8 requires C, 69.04; H, 7.97%). Hydrolysis of the (1)-ester (11.3 g, 20.3 mmol) in a mixture of aq. potassium hydroxide (5%; 100 cm3) and methanol (100 cm3) gave a solid, which was recrystallized from hexane to give the diol (S)-(1)-6 (3.50 g, 89%); mp 125–127 8C, [M]D 24 135.3 (c 0.973, ethanol). The spectral data completely agreed with those of the racemate (±)-6 (Found: C, 73.31; H, 10.2%).(S)-1-(19-Adamantyl)-2-triphenylmethoxyethanol 7 According to the procedure reported in our previous paper,8 the diol (S)-6 (4.28 g, 21.4 mmol) was reacted with triphenylmethyl chloride (13.8 g, 42.8 mmol), triethylamine (4.3 cm3) and 4- dimethylaminopyridine (131 mg, 1.07 mmol) in methylene dichloride (40 cm3) at room temperature for 12 h and silica gel column chromatography of the products gave compound (S)-7 (hexane–diethyl ether, 95 : 5, as eluent) (7.00 g, 75%); mp 138– 140 8C (after recrystallization from ethanol); [M]D 23 191.9 (c 0.937, CHCl3); nmax(KBr)/cm21 3550, 3050, 2900, 2850, 1600, 1480, 1440, 1090, 1070, 770, 760 and 700; dH(270 MHz; CDCl3) 1.40–1.80 (12H, m, adamantyl CH2), 1.90 (3H, br s, adamantyl CH), 2.36 (1H, d, J 2.5, OH), 3.12 (1H, dd, J 8.7 and 8.6, CH2), 3.27 (1H, ddd, J 8.6, 2.7 and 2.5, CH), 3.33 (1H, dd, J 8.7 and 2.7, CH2) and 7.06–7.54 (15H, m, ArH) (Found: C, 84.82; H, 7.8.C31H34O2 requires C, 84.89; H, 7.85%). (R)-3,3-Dimethyl-1-triphenylmethoxybutan-2-ol 9 Treatment of the diol (R)-8, [a]D 28 225.3 (c 0.860, CHCl3) (5.88 g, 49.8 mmol), which was prepared by the procedure reported in the literature,11 with triphenylmethyl chloride (15.5 g, 54.7 mmol), triethylamine (5.5 cm3) and 4-dimethylaminopyridine (168 mg, 1.37 mmol) followed by column chromatography of the products on alumina gave compound (R)-9 (12.5 g, 70%); mp 81–82 8C; [M]D 28 260.1 (c 1.07, CHCl3); nmax(KBr)/cm21 3550, 3050, 2950, 2875, 1600, 1480, 1450, 1230, 1080, 770, 760 and 700; dH(270 MHz; CDCl3) 0.78 [9H, s, C(CH3)3], 2.46 (1H, d, J 2.5, OH), 3.07 (1H, dd, J 9.4 and 8.9, CH2), 3.32 (1H, dd, J 9.4 and 3.0, CH2), 3.24 (1H, ddd, J 8.9, 3.0 and 2.5, CH) and 7.20– 7.45 (15H, m, ArH) (Found: C, 83.57; H, 7.8.C25H28O2 requires C, 83.29; H, 7.83%). (S)-2-Tetrahydropyranyloxypropan-1-ol 11 A mixture of ethyl (S)-lactate (91.3 g, 773 mmol), 3,4-dihydro- 2H-pyran (98.7 ml, 1.08 mol) and three drops of hydrochloric acid was stirred for 12 h at room temperature and then aq.sodium hydrogen carbonate was added to the reaction mixture. The organic phase was separated, washed with water and dried (MgSO4). Removal of the volatile materials under reduced pressure gave ethyl 2-tetrahydropyranyloxypropionate (141 g, 90%) as a colorless oil, which was dissolved in dry THF (400 cm3). The solution was added slowly to a suspension of lithium aluminium hydride (25.0 g, 660 mmol) in dry THF (450 cm3) at 0–5 8C and then the mixture was refluxed for 4 h under a nitrogen atmosphere.The reaction mixture was cooled to 0–5 8C and then saturated aq. ammonium chloride was carefully added to the chilled mixture. After the deposited solid had been removed by filtration, the filtrate was evaporated under reduced pressure.J. Chem. Soc., Perkin Trans. 2, 1997 1655 Distillation of the products gave compound (S)-11 (83.3 g, 75%) as a diastereoisomeric mixture; bp 130–135 8C at 40 mmHg.The diastereoisomeric mixture was used for the next reaction without further separation. (2S,10S)-(1)-2,10-Di(19-adamantyl)-3,6,9-trioxaundecane-1,11- diol 12 A solution of compound (S)-7 (6.10 g, 13.9 mmol) in dry THF (80 cm3) was added dropwise to a suspension of sodium hydride (1.70 g, 70.8 mmol) in dry THF (150 cm3) and then the mixture was refluxed for 4 h. A solution of diethylene glycol bis(toluene-p-sulfonate) (2.60 g, 6.26 mmol) in dry THF (160 cm3) was added slowly to the mixture under reflux and the reaction mixture was refluxed for an additional 10 h.The reaction mixture was cooled to 0–5 8C and then a small amount of cold water was added to the chilled mixture. After the solvent had been evaporated under reduced pressure, the residue was extracted with methylene dichloride. The combined extracts were washed with water and dried (MgSO4). After removal of the solvent, the products were chromatographed on silica gel to give 2,10-di(19-adamantyl)-1,11-bis(triphenylmethoxy)- 3,6,9-trioxaundecane (4.40 g, 74%) (hexane–diethyl ether, 1 : 1, as eluent), which was dissolved in methanol (200 cm3) containing toluene-p-sulfonic acid monohydrate (3.40 g, 17.7 mmol).The solution was stirred at room temperature for 7 h, neutralized with aq. sodium hydrogen carbonate and extracted with chloroform. The combined extracts were washed with aq. sodium hydrogen carbonate and water, dried (K2CO3) and evaporated under reduced pressure.Column chromatography of the residue on silica gel gave the diol (S,S)-12 (2.00 g, 98%) (benzene–ethyl acetate, 1 : 1, as eluent) as a solid; [M]D 22 121.5 (c 0.964, CHCl3); mp 83–85 8C; nmax(KBr)/cm21 3350, 2900, 2850, 1460, 1440, 1420, 1340, 1230, 1100, 990, 940, 930, 810, 740 and 650; dH(270 MHz; CDCl3) 1.67–1.53 (24H, m, adamantyl CH2), 1.95 (6H, br s, adamantyl CH), 3.24–2.83 (2H, m, CH), 3.87–3.53 (12H, m, CH2) and 4.38 (2H, br s, OH), MS (FAB) m/z 463 (MH1) (Found: C, 72.58; H, 9.9.C28H46O5 requires C, 72.69; H, 10.0%). (2R,10R)-(2)-2,10-Di-tert-butyl-3,6,9-trioxaundecane-1,11- diol 13 By a similar procedure to that described for the preparation of the diol (S,S)-12, condensation of compound (R)-9 (11.5 g, 31.3 mmol) with diethylene glycol bis(toluene-p-sulfonate) (5.84 g, 14.1 mmol) followed by hydrolysis gave an oily product, which was chromatographed on silica gel to give the diol (R,R)-13 (1.49 g, 28%) (benzene–ethyl acetate, 1 : 1, as eluent) as a colorless oil; [M]D 25 28.38 (c 2.35, CHCl3); nmax(neat film)/cm21 3350, 3050, 2950, 2875, 1600, 1480, 1450, 1230, 1080, 770, 760 and 700; dH(270 MHz; CDCl3) 0.90 [18H, s, C(CH3)3], 3.07 (2H, dd, J 3.0 and 8.4, CH), 3.50–3.80 (12H, m, CH2) and 4.34 (2H, d, J 6.4, OH); dC(67.9 MHz; CDCl3) 26.1 (q), 34.6 (s), 62.8 (t), 72.8 (t) and 90.7 (d); HRMS m/z 307.2507 (MH1), C16H35O5 requires 307.2485.(S,S)-1,3-Bis(49-hydroxy-2-oxapentyl)-2,5-dimethoxybenzene 14 A solution of compound (S)-11 (10.0 g, 61.2 mmol) in dry THF (150 cm3) was added dropwise to a suspension of sodium hydride (2.94 g, 123 mmol) in dry THF (150 cm3) at room temperature and then the the mixture was stirred for 1 h under reflux.A solution of 1,3-bis(bromomethyl)-2,5-dimethoxybenzene (9.90 g, 30.6 mmol) in dry THF (250 cm3) was added to the reaction mixture under reflux and refluxing was continued for an additional 10 h. The reaction mixture was cooled to 0– 5 8C and then a small amount of cold water was added carefully.After the mixture had been concentrated under reduced pressure, the residue was extracted with a mixture of hexane and ethyl acetate. The combined extracts were washed with water, dried (MgSO4) and evaporated under reduced pressure to give 1,3-bis(49-tetrahydropyranyloxy-2-oxapentyl)-2,5-dimethoxybenzene as a yellow oil, which was dissolved in ethanol (350 cm3) containing pyridinium toluene-p-sulfonate (1.54 g, 6.12 mmol).The solution was heated for 12 h at 60 8C and then concentrated under reduced pressure. The residue was directly chromatographed on silica gel to give the diol (S,S)-14 (5.30 g, 55%) (hexane–ethyl acetate, 1 : 2, as eluent) as a colorless oil; [M]D 24 160.3 (c 1.07, CHCl3); nmax(neat film)/cm21 3400, 2950, 2900, 1480, 1100 and 1070; dH(400 MHz; CDCl3) 1.15 (6H, d, J 6.4, CH3), 2.43 (2H, br s, OH), 3.33 (1H, dd, J 9.5 and 8.2, CH2), 3.51 (1H, dd, J 9.5 and 3.1, CH2), 3.73 (3H, s, CH3O), 3.78 (3H, s, CH3O), 3.95–4.05 (1H, m, CH), 4.58 (4H, s, benzyl CH2), 6.89 (2H, s, ArH); dC(67.9 MHz; CDCl3) 18.6 (q), 55.0 (q), 61.9 (q), 65.8 (d), 67.5 (t), 75.7 (t), 113.9 (d), 131.7 (d), 149.5 (s) and 155.4 (s); HRMS m/z 314.1685 (M1), C16H26O6 requires M, 314.1729.(5S,13S)-5,13-Di(19-adamantyl)-19,21-dimethoxy-3,6,9,12,15- pentaoxabicyclo[15.3.1]henicosane-1(21),17,19-triene 15 A solution of the diol (S,S)-12 (3.80 g, 8.21 mmol) and 1,3- bis(bromomethyl)-2,5-dimethoxybenzene (2.70 g, 8.33 mmol) in dry THF (1000 cm3) was added dropwise to a mixture of sodium hydride (840 mg, 35.0 mmol) and potassium tetrafluoroboranuide (1.20 g, 8.58 mmol) in dry THF (450 cm3) during 50 h under reflux and the reaction mixture was refluxed for an additional 8 h under a nitrogen atmosphere.After the reaction mixture had been cooled to 0–5 8C, a small amount of cold water was added carefully. The reaction mixture was acidified (pH 1) with dilute hydrochloric acid, evaporated under reduced pressure and extracted with diethyl ether.The combined extracts were washed with aq. sodium hydrogen carbonate and water, dried (MgSO4) and evaporated under reduced pressure. Column chromatography of the residue on silica gel gave the crown ether (S,S)-15 (4.40 g, 66%) (hexane–diethyl ether, 1 : 1, as eluent) as a solid; mp 127– 129 8C; [M]D 19 1276.4 (c 0.986, CHCl3); nmax(KBr)/cm21 3020, 2980, 2860, 1500, 1460, 1445, 1260, 1130, 1100 and 1060; dH(270 MHz; CDCl3) 1.53–1.67 (24H, m, adamantyl CH2), 1.95 (6H, s, adamantyl CH), 2.84 (2H, dd, J 2.5 and 7.0, OCH), 3.33–3.76 (12H, m, OCH2), 3.80 (3H, s, OCH3), 3.90 (3H, s, OCH3), 4.34 (2H, d, J 10.8, benzyl CH2), 4.65 (2H, d, J 10.8, benzyl CH2) and 6.85 (2H, s, ArH); MS (FAB) m/z 625 (MH1) (Found: C, 72.70; H, 8.9. C38H56O7 requires C, 73.04; H, 9.03%).(5S,13S)-5,13-Di(19-adamantyl)-19-methoxy-3,6,9,12,15- pentaoxabicyclo[15.3.1]henicosane-1(21),17,19-trien-21-ol 16 Ethanethiol (5.73 g, 92.2 mmol) was added slowly to a suspension of sodium hydride (3.07 g, 128 mmol) in dry DMF (50 cm3) at 0 8C under a nitrogen atmosphere. To the resulting clear solution of sodium ethanethiolate in dry DMF was added slowly a solution of the crown ether (S,S)-15 (2.20 g, 3.52 mmol) in dry DMF (80 cm3) at room temperature and then the reaction mixture was heated for 4 h at 100 8C.The reaction was cooled to 0–5 8C and a small amount of cold water was carefully added.The mixture was acidified (pH 5) with hydrochloric acid and extracted with diethyl ether. The combined extracts were washed with aq. sodium hydrogen carbonate and water and dried (MgSO4). After removal of the solvent, column chromatography of the residue on silica gel gave the phenolic crown ether (S,S)-18 (1.37 g, 64%) (hexane–ethyl acetate, 1 : 1, as eluent); [M]D 25 1141.5 (c 0.865, CHCl3); mp 88–89 8C (from ethyl acetate–hexane); nmax(KBr)/cm21 3340, 2980, 2925, 2850, 1490, 1360, 1255, 1100, 1050, 850, 750 and 660; dH(270 MHz; CDCl3) 1.72–1.55 (24H, m, adamantyl CH2), 1.95 (6H, s, adamantyl CH), 2.95 (2H, dd, J 2.5 and 7.5, OCH), 3.85–3.54 (12H, m, OCH2), 3.75 (3H, s, OCH3), 4.52 (2H, d, J 11.0, benzyl CH2), 4.62 (2H, d, J 11.0, benzyl CH2), 6.69 (2H, s, ArH) and 7.48 (1H, s, OH); MS m/z 610 (M1) (Found: C, 72.58; H, 9.0. C37H54O7 requires C, 72.76; H, 8.91%).1656 J.Chem. Soc., Perkin Trans. 2, 1997 (5R,13R)-5,13-Di-tert-butyl-19,21-dimethoxy-3,6,9,12,15- pentaoxabicyclo[15.3.1]henicosane-1(21),17,19-triene 17 By a procedure similar to that described for the preparation of the crown ether (S,S)-15, condensation of the diol (R,R)-13 (500 mg, 1.55 mmol) with 1,3-bis(bromomethyl)-2,5-dimethoxybenzene (500 mg, 1.55 mmol) followed by silica gel chromatography of the products gave the crown ether (R,R)-17 (355 mg, 48%) (chloroform as eluent) as a colorless oil.HPLC analysis showed a single peak: tR 5.86 min (acetonitrile, 1.0 cm3 min21, as eluent); [M]D 24 2126.4 (c 1.86, CHCl3); nmax(neat film)/ cm21 2950, 2850, 1480, 1360, 1100 and 1050; dH(270 MHz; CDCl3) 0.91 [18H, s, C(CH3)3], 3.00 (2H, dd, J 2.8 and 7.1, CH), 3.31–3.37 (12H, m, CH2), 3.79 (3H, s, OCH3), 3.90 (3H, s, OCH3), 4.34 (2H, d, J 11.0, benzyl CH2), 4.66 (2H, d, J 11.0, benzyl CH2) and 6.85 (2H, s, ArH); dC(69.6 MHz; CDCl3) 26.4 (q), 34.7 (s), 55.6 (q), 63.8 (q), 68.6 (t), 70.5 (t), 71.4 (t), 71.9 (t), 87.7 (d), 116.3 (d), 132.4 (s), 152.0 (s) and 155.0 (s); HRMS m/z 468.3096 (MH1).C26H45O7 requires 468.3087. (5R,13R)-5,13-Di-tert-butyl-19-methoxy-3,6,9,12,15-pentaoxabicyclo[ 15.3.1]henicosane-1(21),17,19-trien-21-ol 18 By a procedure similar to that described for the preparation of the phenolic crown ether (S,S)-16, demethylation of the crown ether (R,R)-17 (100 mg, 0.210 mmol) followed by purification by silica gel column chromatography gave the phenolic crown ether (R,R)-18 (90 mg, 95%) (hexane–ethyl acetate, 1 : 1, as eluent) as a colorless oil, which was immediately used for the next reaction.(5S,13S)-9,21-Dimethoxy-5,13-dimethyl-3,6,9,12,15-pentaoxabicyclo[ 15.3.1]henicosane-1(21),17,19-triene 19 A solution of the diol (S,S)-14 (3.00 g, 9.54 mmol) and diethylene glycol bis(toluene-p-sulfonate) (4.00 g, 9.54 mmol) in dry THF (500 cm3) was added dropwise to a mixture of sodium hydride (912 mg, 38.0 mmol) and potassium tetrafluoroboranuide (1.20 g, 9.54 mmol) in dry THF (200 cm3) over a 9 h period under reflux and the reaction mixture was refluxed for an additional 20 h under a nitrogen atmosphere. After work-up as described above, silica gel column chromatography of the products gave the crown ether (S,S)-19 (1.80 g, 50%) (hexane– ethyl acetate, 1 : 2 as eluent) as a colorless oil.HPLC analysis showed a single peak: tR 5.07 min (acetonitrile, 1.0 cm3 min21, as eluent); [M]D 24 184.9 (c 1.52, CHCl3); nmax(neat film)/ cm21 3300, 2900, 1640, 1600, 1320, 1220 and 1060; dH(270 MHz; CDCl3) 1.12 (6H, d, J 6.5, CH3), 3.39–3.62 (14H, m, CH2 and CH), 3.79 (3H, s, OCH3), 3.94 (3H, s, OCH3), 4.46 (2H, d, J 11.1, benzyl CH2), 4.61 (2H, d, J 11.1, benzyl CH2) and 6.84 (2H, s, ArH); dC(67.9 MHz; CDCl3) 16.8 (q), 55.5 (q), 64.1 (t), 68.4 (t), 68.5 (t), 70.7 (t), 73.3 (t), 74.8 (d), 116.3 (d), 132.3 (s), 152.3 (s) and 154.9 (s); HRMS m/z 384.2185 (M1).C26H44O7 requires 384.2148.(5S,13S)-19-Methoxy-5,13-dimethyl-3,6,9,12,15-pentaoxabicyclo[ 15.3.1]henicosane-1(21),17,19-trien-21-ol 20 By a procedure similar to that described for the preparation of the phenolic crown ether (S,S)-16, demethylation of the crown ether (S,S)-19 (500 mg, 1.30 mmol) followed by purification by silica gel column chromatography gave the phenolic crown ether (S,S)-20 (440 mg, 92%) (hexane–ethyl acetate, 1 : 1, as eluent) as a colorless oil; nmax(neat film)/cm21 3400, 2900, 1660, 1610, 1500, 1480, 1380, 1260, 1100, 1060, 860 and 780; dH(270 MHz; CDCl3) 1.12 (6H, d, J 6.4, CH3), 3.46–3.81 (14H, m, CH2 and CH), 3.79 (3H, s, OCH3), 4.62 (4H, s, benzyl CH2), 6.70 (2H, s, ArH) and 7.59 (1H, br s, OH); MS m/z 370 (M1).This was used for the next reaction without further purification. (5S,13S)-5,13-Di(19-adamantyl)-19-(29,49-dinitrophenylazo)- 3,6,9,12,15-pentaoxabicyclo[15.3.1]henicosane-1(21),17,19- trien-21-ol 1 A solution of the phenolic crown ether (S,S)-16 (950 mg, 1.57 mmol) in acetonitrile (150 cm3) was added to a solution of CAN (2.20 g, 4.01 mmol) in acetonitrile (50 cm3) and then the mixture was stirred for 2 h at room temperature.After water had been added to the reaction mixture, the solvent was evaporated under reduced pressure. The residue was extracted with chloroform and the combined extracts were washed with water and dried (MgSO4). Evaporation of the solvent gave the corresponding quinone (1.00 g) as a yellow oil, which was immediately taken up in a mixture of ethanol (40 cm3) and methylene dichloride (40 cm3).To this solution was added a solution of 2,4-dinitrophenylhydrazine (3.30 g, 8.40 mmol) in a mixture of sulfuric acid (7 cm3) and ethanol (90 cm3) and then the mixture was stirred for 1.5 h at room temperature. After water had been added, the reaction mixture was extracted with chloroform. The combined extracts were washed with aq. sodium hydrogen carbonate and water, dried (MgSO4) and evaporated under reduced pressure.Silica gel column chromatography of the residue gave the azophenolic crown ether (S,S)-1 (982 mg, 82%) (chloroform as eluent) as an orange solid; mp 89–90 8C; lmax(CHCl3)/nm; 403 (e/dm3 mol21 cm21 2.46 × 104); nmax(KBr)/ cm21 3340, 2980, 2925, 2850, 1490, 1360, 1255, 1100, 1050, 850, 750 and 660; dH(270 MHz; CDCl3) 1.56–1.69 (24H, m, adamantyl CH2), 1.96 (6H, s, adamantyl CH), 2.98 (2H, dd, J 2.3 and 7.6, CH), 3.60–3.94 (12H, m, OCH2), 4.67 (2H, d, J 11.4, benzyl CH2), 4.76 (2H, d, J 11.4, benzyl CH2), 7.80 (2H, s, HOArH), 7.82 [1H, d, J 8.9, (O2N)2ArH], 8.49 [1H, dd, J 8.9 and 2.6, (O2N)2ArH], 8.76 [1H, d, J 2.6, (O2N)2ArH] and 8.99 (1H, br s, OH); dC(69.6 MHz; CDCl3) 28.3 (t), 36.9 (t), 37.1 (t), 38.7 (t), 70.2 (t), 70.8 (t), 71.4 (t), 72.1 (t), 87.5 (d), 120.0 (d), 125.2 (s), 126.4 (d), 127.5 (d), 145.7 (s), 146.5 (s), 147.0 (s), 148.9 (s) and 161.5 (s); MS (FAB) m/z 775 (MH1) (Found: C, 64.78; H, 7.0; N, 7.28.C42H54O10N4 requires C, 65.10; H, 7.02; N, 7.23%). (5S,13S)-5,13-Di-tert-butyl-19-(29,49-dinitrophenylazo)-3,6,9, 12,15-pentaoxabicyclo[15.3.1]henicosane-1(21),17,19-trien-21- ol 2 As described for the preparation of the azophenolic crown ether (S,S)-1, oxidation of the phenolic crown ether (R,R)-18 (90 mg, 0.200 mmol) with CAN followed by treatment with 2,4- dinitrophenylhydrazine gave solid products. Silica gel column chromatography gave the azophenolic crown ether (R,R)-2 (60 mg, 50%) (chloroform as eluent) as a red solid; mp 65.0– 66.5 8C; lmax(CHCl3)/nm; 402 (e/dm3 mol21 cm21 2.92 × 104); nmax(KBr)/cm21 3300, 3000, 2900, 1620, 1560, 1490, 1380, 1320, 1140, 940, 870 and 790; dH(270 MHz; CDCl3) 0.94 [18H, s, C(CH3)3], 3.16 (2H, dd, J 2.3 and 7.6, CH), 3.59–3.95 (12H, m, OCH2), 4.68 (2H, d, J 11.2, benzyl CH2), 4.76 (2H, d, J 11.2, benzyl CH2), 7.80 (2H, s, HOArH), 7.82 [1H, d, J 8.9, (O2N)2ArH], 8.49 [1H, dd, J 8.9 and 2.3, (O2N)2ArH], 8.76 [1H, d, J 2.3, (O2N)2ArH] and 8.99 (1H, br s, OH); dC(69.6 MHz; CDCl3) 26.5 (q), 34.9 (s), 70.3 (t), 70.8 (t), 72.1 (t), 72.5 (t), 87.2 (d), 120.0 (d), 120.2 (d), 125.2 (s), 126.4 (d), 127.6 (d), 145.8 (s), 147.0 (s), 147.1 (s), 149.0 (s) and 161.5 (s); MS (FAB) m/z 619 (MH1).(Found: C, 57.90; H, 6.7; N, 8.65. C30H42O10N4 requires C, 58.24; H, 6.84; N, 8.65%). (5S,13S)-5,13-Dimethyl-19-(29,49-dinitrophenylazo)-3,6,9,12,15- pentaoxabicyclo[15.3.1]henicosane-1(21),17,19-trien-21-ol 3 As described for the preparation of the azophenolic crown ether (S,S)-1, oxidation of the phenolic crown ether (S,S)-20 (150 mg, 0.405 mmol) with CAN followed by treatment with 2,4-dinitrophenylhydrazine gave solid products, column chromatography of which on silica gel gave the azophenolic crown ether (S,S)-3 (170 mg, 78%) (chloroform as eluent) as an orange solid.HPLC analysis showed a single peak: tR 5.09 min, (acetonitrile, 1.0 cm3 min21, as eluent); mp 42.0–42.5 8C; lmax(CHCl3)/nm; 403 (e/dm3 mol21 cm21 2.98 × 104); nmax(KBr)/ cm21 3300, 2950, 2900, 1640, 1600, 1500, 1480, 1400, 1320, 1100, 1060, 820 and 740; dH(270 MHz; CDCl3) 1.16 (6H, d,J.Chem. Soc., Perkin Trans. 2, 1997 1657 J 6.3, CH3), 3.16–3.82 (14H, m, CH2 and CH), 4.74 (4H, s, benzyl CH2), 7.80 (2H, s, HOArH), 7.81 [1H, d, J 8.6, (O2N)2ArH], 8.48 [1H, dd, J 8.6 and 2.3, (O2N)2ArH], 8.75 [1H, d, J 2.3, (O2N)2ArH] and 9.04 (1H, br s, OH); dC(67.9 MHz; CDCl3) 16.2 (q), 70.3 (t), 70.5 (t), 74.6 (t), 74.8 (t), 120.0 (d), 120.0 (d), 125.4 (s), 126.4 (d), 127.5 (d), 145.7 (s), 146.6 (s), 147.0 (s), 149.0 (s) and 161.7 (s); HRMS m/z 534.2003 (M1).C24H30O10N4 requires 534.1962. General procedures for evaluation of association constants of complexes Association constants for complexes of (S,S)-1 with 21, 22 and 23, (R,R)-2 with 21, 22 and 23 and (S,S)-3 with 21 were calculated by the non-linear least-squares method on the basis of the 1H NMR spectral data in CDCl3 and those for complexes of (S,S)-3 with 22 and 23 were determined by the Rose–Drago method13 on the basis of the UV–VIS absorptions in CHCl3.In order to evaluate Ka values precisely, the concentrations of the host and the guest were changed in each experiment and measurements were made on solutions ranging in concentration as follows; for 1H NMR titration: 5.29 × 1024–4.88 × 1022 M for the complexes with Ka = 5.20 × 103–1.82 × 104; 1.05 × 1023– 9.76 × 1022 M for the complexes with Ka = 1.60 × 102– 2.40 × 104; 3.18 × 1023–2.64 × 1021 M for the complexes with Ka = 2.51 × 10–1.35 × 102; 1.31 × 1023–7.06 × 1023 M for the complexes with Ka = 4.54 × 103–1.53 × 104; 6.93 × 1024–9.53 × 1023 M for the complexes with Ka = 2.62 × 102–7.49 × 103; for the UV–VIS spectroscopic method: 3.18 × 1025–5.65 × 1024 M for the complexes with Ka = 3.20 × 103–1.99 × 104; 2.76 × 105–9.21 × 1025 M for the complexes with Ka = 2.64 × 103–7.04 × 103; 3.33 × 1024–6.24 × 1023 M for the complexes with Ka = 6.06 × 10–1.74 × 103; 1.11 × 104–8.35 × 1024 M for the complexes with Ka = 1.36 × 102–5.10 × 103.Acknowledgements This work was partially supported by a Grant-in-Aid for Scientific Research (No. 07640718) from the Ministry of Education, Science and Culture of Japan. References 1 A preliminary communication of part of this work has already appeared: K. Naemura, J. Fuji, K. Ogasahara, K. Hirose and Y. Tobe, Chem. Commun., 1996, 2749. 2 J. D. Morrison and H. S. Mosher, Asymmetric Organic Reactions, Prentice-Hall, Englewood Cliffs, 1971; J. Retey and J. A. Robinson, Stereospecificity in Organic Chemistry and Enzymology, Verlag Chemie, Weinheim, 1982; Asymmetric Synthesis, ed. J. D. Morison, vols. 1–5, Academic Press, New York, 1983–1984; Asymmetric Synthesis, ed. G. M. Coppola and H. F. Schuster, Wiley- Interscience, New York, 1987. 3 V. T. Pham, R. S. Phillips and L. G. Ljungdahl, J. Am. Chem. Soc., 1989, 111, 1935. 4 K. Watabe, R. Charles and E. Gil-Av, Angew. Chem., Int. Ed. Engl., 1989, 28, 192; V. Schurig, J. Ossig and R. Link, Angew. Chem., Int. Ed. Engl., 1989, 28, 194. 5 Y. Inoue, T. Yokoyama, M. Yamasaki and A. Tai, Nature, 1989, 341, 225. 6 G. W. Gokel and S. H. Korzeniowski, Macrocyclic Polyether Syntheses, Springer-Verlag, New York, 1982; P. G. Potvin and J.-M. Lehn, Design of Cation and Anion Receptors, Catalysts and Carriers, in Synthesis of Macrocycles: The design of Selective Complexing Agents, ed. R. M. Izatt and J. J. Christensen, Wiley-Interscience, New York, 1987; p. 167; J. F. Stoddart, Topics Stereochem., 1988, 17, 207; J.-M. Lehn, Supramolecular Chemistry, VCH Verlagsgesellschaft, Weinheim, 1995. 7 K. Naemura, Y. Nishikawa, J. Fuji, K. Hirose and Y. Tobe, Tetrahedron: Asymmetry, in press. 8 K. Naemura, T. Mizo-oku, K. Kamada, K. Hirose, Y. Tobe, M. Sawada and Y. Takai, Tetrahedron: Asymmetry, 1994, 5, 1549. 9 S. Iriuchijima, K. Maniwa and G. Tsuchihashi, J. Am. Chem. Soc., 1975, 97, 596. 10 G. I. Feutrill and R. N. Mirrington, Tetrahedron Lett., 1970, 16, 1327. 11 K. Naemura and M. Ueno, Bull. Chem. Soc. Jpn., 1990, 63, 3695; P. Huszthy, J. S. Bradshaw, C. Y. Zhu and R. M. Izatt, J. Org. Chem., 1991, 56, 3330. 12 P. A. Levene and H. L. Haller, J. Biol. Chem., 1926, 67, 329; E. Baer and H. O. L. Fischer, J. Am. Chem. Soc., 1948, 70, 609. 13 N. J. Rose and R. S. Drago, J. Am. Chem. Soc., 1959, 81, 6138. 14 T. Kaneda, K. Hirose and S. Misumi, J. Am. Chem. Soc., 1989, 111, 742. 15 R. M. Izatt, R. E. Terry, B. L. Haymore, L. D. Hansen, N. K. Dalley, A. G. Avondet and J. J. Christensen, J. Am. Chem. Soc., 1976, 98, 7620; Y. Inoue, F. Amano, N. Okada, H. Inada, M. Ouchi, A. Tani, T. Hakushi, Y. Lie and L.-H. Tong, J. Chem. Soc., Perkin Trans. 2, 1990, 1239; Y. Liu, L.-H. Tong, Y. Inoue and T. Hakushi, J. Chem. Soc., Perkin Trans. 2, 1990, 1247. Paper 7/02133J Received 27th March 1997 Accepted 16th May 1997

ISSN:1472-779X    

DOI:10.1039/a702133j

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1659 Stereochemistry of linking segments in the design of helix–helix motifs in peptides. Crystallographic comparison of a glycyl– dipropylglycyl–glycyl segment in a tripeptide and a 14-residue peptide Saumen Datta,a Ramesh Kaul,b,c R. Balaji Rao,c N. Shamala *,a and P. Balaram*,b a Department of Physics, Indian Institute of Science, Bangalore-560012, India b Molecular Biophysics Unit, Indian Institute of Science, Bangalore-560012, India c Department of Chemistry, Banaras Hindu University, Varanasi-221005, India As part of a program to develop synthetic helix–linker–helix peptides the conformational properties of various linking segments are currently being investigated.The propensity of ·,·-di-n-propylglycine (Dpg) residues to adopt backbone conformations in the extended region of the Ramachandran map, suggested by theoretical calculations and supported by experimental observations, prompted us to investigate the utility of the Gly-Dpg-Gly segment as a rigid linking motif.The crystal structure of the achiral tripeptide Boc-Gly-Dpg-Gly-OH 1 revealed a fully extended conformation (� = ±1788, � = ±1718) at Dpg(2), with Gly(1) adopting a helical conformation (� = 12728, � = 12328). The addition of flanking helical segments in the 14 residue peptide Boc-Val-Val-Ala-Leu-Gly-Dpg-Gly-Val-Ala-Leu-Aib-Val-Ala-Leu-OMe 2 resulted in the crystallographic characterization of a continuous helix over the entire length of the peptide. Peptide 1 crystallized in the centrosymmetric space group P21/c with a = 9.505(2) Å, b = 11.025(2) Å, c = 20.075(4) Å, ‚ = 90.198 and Z = 4.Peptide 2 crystallized in space group P212121 with a = 10.172(1) Å, b = 17.521(4) Å, c = 46.438(12) Å and Z = 4. A comparative analysis of Gly-Dpg-Gly segments from available crystal structures indicates a high conformational variability of this segment. This analysis suggests that context and environment may be strong conformational determinants for the Gly-Dpg-Gly segment.Introduction The ability to construct stereochemically well-defined peptide helices, using a-aminoisobutyric acid (Aib) and related a,adialkylated glycines,1–6 has stimulated attempts to assemble helix–linker–helix motifs as models for super secondary structures in proteins.7 The use of nonhelical linkers should facilitate the design of molecules with distinct helical segments. A close packed, approximately antiparallel helix arrangement may then be achieved as a consequence of solvophobic effects, in which release of solvent molecules entropically drives the association of large complementary molecular surfaces.8–10.In the ‘Meccano set’ approach being developed in this laboratory, various linking segments are being investigated. Earlier reports have described attempts to use Gly-Pro units,11, D-amino acids 12 and e-aminocaproic acid (Acp) 13 as linking units between helix pairs.In this paper we describe an analysis of the linking segment Gly-Dpg-Gly (Dpg = a,a-di-n-propylglycine). The choice of Dpg was stimulated by a report that higher a,a-di-n-alkylglycines have pronounced energy minima in the fully extended (f, y ª1808) region of conformational space,14,15 suggesting the utility of this residue in designing stereochemically rigid nonhelical segments. Interestingly, while early crystal structure analyses of homo-oligopeptides containing Dpg provided evidence for the occurrence of the fully extended conformations, 16,17 many subsequent reports provided examples of Dpg in helical conformations.6,18.Both theoretical and experimental studies suggest that two distinct regions of conformational Fig. 1 A stereoview of the tripeptide Boc-Gly-Dpg-Gly-OH 1 structure1660 J. Chem. Soc., Perkin Trans. 2, 1997 Table 1 Crystal data and structure refinement details for peptide 1 and 2 Empirical formula Formula weight Temperature Wavelength Crystal system Space group Unit cell dimensions Volume Z Density (calculated) Absorption coefficient F(000) Crystal size q range for data collection Index ranges Independent reflections Reflections [I > 2s(I)] Data/restraints/parameters Goodness-of-fit Final R indices [I > 2s(I)] R indices (all data) Largest difference peak and hole Peptide 1 C17H31N3O6 373.45 293(2) K 1.541 80 Å Monoclinic P21 /c a = 9.505(2) Å b = 11.025(2) Å c = 20.075(4) Å a = 908 b = 90.198 g = 908 2103.7(7) Å3 4 1.179 Mg m23 0.740 mm21 808 0.4 × 0.4 × 0.4 mm 4.40–74.988 211 < h < 11 0 < k < 13 0 < l < 25 4323 3409 4323/0/265 0.959 R1 = 0.0460 wR2 = 0.1350 R1 = 0.0572 wR2 = 0.1468 0.242 and 20.288 e Å23 Peptide 2 C70H124N14O19 1465.83 293(2) K 1.541 80 Å Orthorhombic P212121 a = 10.172(1) Å b = 17.521(4) Å c = 46.438(12) Å a = 908 b = 908 g = 908 8276(3) Å3 4 1.176 Mg m23 0.704 mm21 3176 0.8 × 0.5 × 0.2 mm 1.90–75.218 0 < h < 12 0 < k < 21 0 < l < 58 9428 7231 9428/0/1052 1.312 R1 = 0.0557 wR2 = 0.1569 R1 = 0.0711 wR2 = 0.1671 0.468 and 20.240 e Å23 Table 2 Backbone dihedral angles for the Gly-Dpg-Gly segment in peptide crystal structures Residue Gly Dpgg Gly Dihedral angles/8 a f y f y f y Segment 1 b 272 232 178 171 263 Segment 2 c 266 251 252 244 263 234 Segment 3 d 294, 296 f 2162, 2153 253, 256 250, 247 264, 265 236, 240 Segment 4 e 72 2166 254 246 2 78 29 Segment 5 280 218 56 32 85 23 a Dihedral angle nomenclature follows that described in ref. 29. b Boc-Gly-Dpg-Gly-OH (this study). c Boc-Val-Val-Ala-Leu-Gly-Dpg-Gly-Val-Ala- Leu-Aib-Val-Ala-Leu-OMe (this study). d Boc-Gly-Dpg-Gly-Val-Ala-Leu-Aib-Val-Ala-Leu-OMe (ref. 27). e Boc-Gly-Dpg-Gly-Gly-Dpg-Gly- NHMe. Segment 4 is the N-terminus tripeptide and segment 5 corresponds to the C-terminus tripeptide (ref. 28). f Two values correspond to the two conformers present in the crystallographic asymmetric unit. g The Dpg sidechain torsion angles in peptide 1 (this study) are c1 (2578, 568), c2 (1708, 1708).space (fully extended and helical) are energetically accessible to Dpg residues. The sequence context and environmental influences presumably determine the precise nature of the conformation adopted. The use of Gly-Dpg-Gly in the present study was dictated by the fact that Gly is highly conformationally flexible and has a relatively low helix propensity. We describe in this report crystal structures of a tripeptide Boc-Gly-Dpg-Gly- OH (peptide 1) and a 14 residue peptide Boc-Val-Val-Ala-Leu- Gly-Dpg-Gly-Val-Ala-Leu-Aib-Val-Ala-Leu-OMe (peptide 2).While the Dpg residue adopts a fully extended conformation in the former, a continuous helix is obtained in the latter. Comparisons with other crystallographically determined Gly-Dpg-Gly segments reveals a significant degree of conformational variability in the sequence. Experimental Peptides were synthesized by conventional solution phase procedures 19 and purified by medium pressure liquid chromatography on a reverse phase C18 (40–60 m) column using methanol–water gradients.Peptides were checked for homogeneity by high performance liquid chromatography on a Table 3 Torsion Angles a (8) in Boc-Val-Val-Ala-Leu-Gly-Dpg-Gly- Val-Ala-Leu-Aib-Val-Ala-Leu OMe (peptide 2) Residue Val (1) Val (2) Ala (3) Leu (4) Gly (5) Dpg (6) Gly (7) Val (8) Ala (9) Leu (10) Aib (11) Val (12) Ala (13) Leu (14) f 261 b 255 261 268 266 252 263 263 262 260 256 278 2106 294 y 223 242 238 239 251 244 234 245 240 251 244 210 26 177 c w 175 180 179 180 173 2176 177 178 176 2169 2172 2178 2179 2176 d c1 264, 63 270, 167 260 269, 170 268, 167 176 67, 259 277 c2 264, 173 178, 2178 66, 2171 250, 2178 a The torsion angles for rotation about bonds of the peptide backbone (f, f, and w) and about bonds of the amino acid side-chains (c1, c2) as suggested by the IUPAC-IUB Commission on Biochemical Nomenclature (ref. 29).Estimated standard deviations ~1.08. b C9(0)–N(1)– Ca(1)–C9(1). c N(14)–Ca(14)–C9(14)–O(OMe). d Ca(14)–C9(14)– O(OMe)–C(OMe).J. Chem. Soc., Perkin Trans. 1661 Table 4 Potential hydrogen bond parameters in Boc-Val-Val-Ala-Leu-Gly-Dpg-Gly-Val-Ala-Leu-Aib-Val-Ala-Leu-OMe Bond length/Å Bond angle/8 Type Intermolecular Intramolecular 4 æÆ 1 b 4 æÆ 1 b 4 æÆ 1 4 æÆ 1 4 æÆ 1 4 æÆ 1 c 4 æÆ 1 4 æÆ 1 4 æÆ 1 4 æÆ 1 4 æÆ 1b 4 æÆ 1 5 æÆ 1 5 æÆ 1 b 5 æÆ 1 b 5 æÆ 1 b 5 æÆ 1 b 5 æÆ 1 c 5 æÆ 1b 5 æÆ 1 b 5 æÆ 1 c 5 æÆ 1 5 æÆ 1c Solvent-peptide Donor O(W) N(1) N(2) N(3) N(4) N(5) N(6) N(7) N(8) N(9) N(10) N(11) N(12) N(13) N(14) N(4) N(5) N(6) N(7) N(8) N(9) N(10) N(11) N(12) N(13) N(14) O(M)d Acceptor O(13) O(W)a O(W)a O(0) O(1) O(2) O(3) O(4) O(5) O(6) O(7) O(8) O(9) O(10) O(11) O(0) O(1) O(2) O(3) O(4) O(5) O(6) O(7) O(8) O(9) O(10) O(3) N? ? ? O 2.83 3.14 3.04 3.01 3.01 3.20 3.76 3.19 3.18 3.23 3.35 3.32 3.70 3.02 3.54 3.79 2.96 3.0 2.94 2.92 3.48 3.04 2.86 3.38 3.67 3.20 2.88 H? ? ? O 2.33 2.50 2.71 3.51 2.81 2.57 2.78 2.98 2.84 3.11 2.23 2.71 3.00 2.16 2.20 2.19 2.20 2.63 2.21 2.06 2.57 3.12 2.69 C]] O? ? ? H 123 100 92 85 84 100 104 86 91 96 112 102 141 157 161 132 146 153 156 146 149 135 135 C]] O? ? ? N 130 114 105 97 99 111 116 99 103 105 121 106 141 164 160 142 153 155 160 152 152 145 169 O? ? ? HN 137 119 117 100 109 129 114 109 117 128 152 163 155 156 172 145 143 167 164 154 158 124 119 a Symmetrically related by the relation (2x 1 ��� , 2y 1 1, z 1 ��� ).b These are the acceptable hydrogen bonds satisfying the criteria of hydrogen bond geometry (ref. 24). c These are the weak hydrogen bonds (ref. 24). d Oxygen atom of CH3OH. Fig. 2 Packing diagram for Boc-Gly-Dpg-Gly-OH 1. The intermolecular hydrogen bonds [O(1) ? ? ? N(3)[2x, 0.5 1 y, 0.5 2 z] = 2.84 Å, O(2) ? ? ? N(1)[2x 1 1, 2y, 2z] = 2.98 Å, O(3) ? ? ? O(L)[2x 1 1, 2y 1 1, 2z] = 2.62 Å] are indicated by broken lines. reversed phase C18 (5 m) column and characterized by 400 MHz 1H NMR spectroscopy.Peptide 2 was obtained as a deletion sequence in the synthesis of a longer symmetrical seventeen residue peptide. Crystals of peptide 1 and 2 were obtained by slow evaporation from a methanol–water solution. X-Ray diffraction data for both the peptide crystals were collected at room temperature, 21 8C, with an automated four-circle diffractometer using Cu-Ka (l = 1.5418 Å) radiation. 25 reflections in the 108 < q<158 range were used for determining the cell constants in both cases. Though the b value (90.198) is close to 908 in the case of peptide 1, the significant difference between hkl and the corresponding h� kl reflections suggests a monoclinic cell.In the case of peptide 1 the structure was determined by the direct phase determination method.20 The structure of peptide 2 was obtained by the vector search method21 followed by partial structure expansion.22 The helical backbone fragment (residue 2 to residue 8) of the sequence Boc-Aib-Val-Ala-Leu- Aib-Val-Ala-Leu-Aib-OMe23 was used in the search method.Both the peptide structures were refined isotropically followed by anisotropic least-squares refinement. Hydrogen atoms were added geometrically and allowed to ride with the corresponding heavy atoms in the final cycle of the refinement. All the relevant crystallographic data collection parameters and structure refinement details for the two peptides are summarized in Table 1.† † Atomic coordinates, bond lengths and angles, and thermal parameters have been deposited at the Cambridge Crystallographic Data Centre (CCDC).For details of the deposition scheme, see ‘Instructions for Authors’, J. Chem. Soc., Perkin Trans. 2, 1997, Issue 1. Any request to the CCDC for this material should quote the full literature citation and the reference number 188/80.1662 J. Chem. Soc., Perkin Trans. 2, 1997 Fig. 3 Stereoview of the crystal structure of peptide 2.The Gly-Dpg-Gly segment is indicated in bold type. The intramolecular hydrogen bonds are indicated by broken lines (see Table 4). Results and discussion Extended Dpg residue in peptide 1 Fig. 1 shows a stereoview of the molecular conformation of tripeptide 1 in crystals. The backbone conformational angles are listed in Table 2, which also provides a comparison with structures of the same segment in larger peptides. In tripeptide 1 the Dpg residue adopts a fully extended conformation while Gly(1) lies in the helical region.The achiral peptide crystallizes in a centrosymmetric space group, with molecules of both helical senses being present in the unit cell. The molecules are held in the crystal by intermolecular hydrogen bonds formed between symmetry related molecules (Fig. 2). Surprisingly, several hydrogen bond donor and acceptor groups do not participate in hydrogen bonding interactions. The peptide helix in the 14 residue peptide 2 Fig. 3 shows a stereoview of the conformation of the 14 residue peptide determined in crystals. The backbone and side-chain torsion angles are listed in Table 3. Intramolecular and intermolecular hydrogen bonds are summarized in Table 4. Hydrogen bond parameters are listed for all potential 4 æÆ 1 and 5 æÆ 1 interactions to provide a ready assessment of helix type. This assumes importance in view of the fact that in helical peptides assignment of 310 and a-helical structures is not always readily apparent.24 The molecule forms an almost completely a-helical structure, stabilized by successive 5 æÆ 1 hydrogen bonds.As frequently observed in peptide helices there is a 310- helical turn at the N-terminus with a 4 æÆ 1 hydrogen bond between the Boc(0)CO and Ala(3)NH groups . A single 310- helical hydrogen bond is also observed near the C-terminus between Leu(10)CO and Ala(13)NH groups. In the centre of the helix there is a evidence of a possible transition between a and 310-helical structures. Gly(5)CO appears to be involved in a 4 æÆ 1 interaction with Val(8)NH, while a corresponding 5 æÆ 1 interaction with Ala(9)NH is definitely weaker as indicated by the N ? ? ? O distances.The molecules pack in the crystal as columns of antiparallel helices, held together in each column by head-to-tail hydrogen bonds mediated by a single bridging water molecule (Fig. 4). A lone methanol molecule is trapped between helical columns and forms a single hydrogen bond with the CO group of Ala(3).This is a relatively rare example of solvation involving bifurcated hydrogen bond formation to a CO group involved in a strong intrahelical hydrogen bond. Such solvent interactions are also observed in protein structures.25 The CH3 group of the CH3OH molecule is in close van der Waals contact with the hydrophobic side chains of Ala(3), Dpg(6), Leu(4) [21 1 x,y,z], Val(8) [21 1 x,y,z], and Aib(9) [1 2 x, ��� 1 y, ��� 2 z] residues (Fig. 5).Such trapped alcohol molecules in helical clusters have also been observed earlier in structures of hydrophobic helices.26 Context dependent Gly-Dpg-Gly conformation Fig. 6 shows an overlay of the structures of the 14 residue peptide 2 and the helical decapeptide Boc-Gly-Dpg-Gly-Val- Ala-Leu-Aib-Val-Ala-Leu-OMe.27 Residues 5–14 of peptide 2 are exactly identical in sequence to the decapeptide. Comparison of the dihedral angles in Table 2 together with Fig. 6J.Chem. Soc., Perkin Trans. 2, 1997 1663 establishes that the Gly-Dpg-Gly segment switches to a completely helical conformation in peptide 2, whereas a nonhelical N-terminus is observed in the decapeptide. Interestingly, Table 2 shows that the 14 residue peptide 2 is the only example where the Gly-Dpg-Gly segment adopts a completely helical conformation. In four out of five peptides listed in Table 2 the Dpg residue adopts helical f, y values, with peptide 1 being the sole exception.However, the overall conformation of the tripeptide segment is nonhelical in all the cases with the exception of peptide 2. In two examples Gly(1) adopts a semi-extended con- Fig. 4 Packing diagram for View down the crystallographic x-axis. Intermolecular hydrogen bonds are indicated by broken lines. W indicates the oxygen molecule of the water and M represents the trapped methanol molecule. Fig. 5 The van der Waals environment of the methanol molecule (M).Atoms which lie within ~4 Å are indicated by the dotted lines. The bold broken line indicates the hydrogen bond between the oxygen atom of methanol and the Ala(3)CO group. formation. The peptide Gly-Dpg-Gly-Gly-Dpg-Gly-NHMe provides an interesting example of a multiple b-turn structure. The N-terminus Gly-Dpg-Gly segment exhibits a type-II (II9) b-turn conformation with Gly(1) and Dpg(2) occupying the i 1 1 and i 1 2 positions.The C-terminus Gly-Dpg-Gly segment forms a type-I (I9) b-turn centred at Dpg(5) and Gly(6). While Gly(4) and Dpg(5) adopt helical f, y values, the signs of the dihedral angles are opposite, indicative of opposing helix senses.28 The above comparison of the Gly-Dpg-Gly conformation in peptides of varying length and sequence suggests that the conformation of this segment may be modulated by subtle environmental effects. Although Dpg residues are constrained to adopt helical or fully extended conformations, the combination of these two stereochemical alternatives with f, y variations at the flanking Gly residues leads to appreciable conformational diversity.Somewhat disappointingly, the Gly- Dpg-Gly segment in the 14-residue peptide 2 favours a helical conformation, resulting in the characterization of a long cylindrical helix in crystals. The overwhelming crystallinity of hydrophobic helical peptide suggests that packing of apolar cylinders into crystalline lattices must be highly favourable. The extent to which the energetics of crystal packing promote the selection of helical conformations in peptide single crystals remains to be established. The present study reaffirms the necessity of interrupting intramolecular hydrogen bonding patterns in order to achieve helix termination in the middle of long hydrophobic sequences.Acknowledgements This research was supported by the Department of Science and Technology, Government of India.R. K. was supported by a Research Associateship of the Department of Biotechnology, Government of India. References 1 I. L. Karle and P. Balaram, Biochemistry, 1990, 29, 6747. 2 P. Balaram, Curr. Opin. Struct. Biol., 1992, 2, 845. Fig. 6 Superposition of the structure of the 14-residue peptide (peptide 2) and the decapeptide Boc-Gly-Dpg-Gly-Val-Ala-Leu-Aib- Val-Ala-Leu-OMe.27 The former is indicated by the solid line, while the latter is represented by a broken line.1664 J.Chem. Soc., Perkin Trans. 2, 1997 3 I. L. Karle, J. L. Flippen-Anderson, R. Gurunath and P. Balaram, Protein Sci., 1994, 4, 1547. 4 A. Banerjee, S. Datta, A. Pramanik, N. Shamala and P. Balaram, J. Am. Chem. Soc., 1996, 118, 9477. 5 I. L. Karle, R. B. Rao, S. Prasad, R. Kaul and P. Balaram, J. Am. Chem. Soc., 1994, 116, 10 355. 6 I. L. Karle, R. Gurunath, S. Prasad, R. Kaul, R. B. Rao and P. Balaram, J. Am. Chem. Soc., 1995, 117, 9632. 7 P. Balaram, Pure Appl. Chem., 1992, 64, 1061. 8 J. A. Bryant, C. B. Knobler and D. J. Cram, J. Am. Chem. Soc., 1990, 112, 1254. 9 J. A. Bryant, J. L. Ericson and D. J. Cram, J. Am. Chem. Soc., 1990, 112, 1254. 10 G. M. Whitesides, E. E. Simanek, J. P. Mathias, C. T. Seto, D. N. Chin, M. Mammen and D. M. Gordon, Acc. Chem. Res., 1995, 28, 37. 11 K. Uma, I. L. Karle and P. Balaram, in Proteins: Structure Dynamics and Design, eds. V. Renugopalakrishnan, P. R. Carey, I. C. P. Smith, S. G. Huang and A. Storer, ESCOM Science Publishers B.V., Leiden, 1991. 12 R. Gurunath and P. Balaram, Biochem. Biophys. Res. Commun., 1994, 202, 241. 13 I. L. Karle, J. L. Flippen-Anderson, M. Sukumar, K. Uma and P. Balaram, J. Am. Chem. Soc., 1991, 113, 3952. 14 E. Benedetti, C. Toniolo, P. M. Hardy, V. Barone, A. Bavoso, B. DiBlasio, P. Grimaldi, F. Lelj, V. Pavone, C. Pedone, G. M. Bonora and I. Lingham, J. Am. Chem. Soc., 1984, 106, 8146. 15 V. Barone, F. Lelj, A. Bavoso, B. Di Blasio, P. Grimaldi, V. Pavone and C. Pedone, Biopolymers, 1985, 24, 1759. 16 G. M. Bonora, C. Toniolo, B. Di Blasio, V. Pavone, C. Pedone, E. Benedetti, I. Lingham and P. Hardy, J. Am. Chem. Soc., 1984, 106, 8152. 17 C. Toniolo, G. M. Bonora, A. Bavoso, E. Benedetti, B. Di Blasio, V. Pavone, C. Pedone, V. Barone, F. Lelj, M. T. Leplawy, K. Kaezmarek and A. Redlinski, Biopolymers, 1988, 27, 373. 18 B. Di Blasio, V. Pavone, C. Isernia, C. Pedone, E. Benedetti, C. Toniolo, P. M. Hardy and I. Lingham, J. Chem. Soc., Perkin Trans. 2, 1992, 523. 19 S. Prasad, R. B. Rao and P. Balaram, Biopolymers, 1995, 35, 11. 20 J. Karle and I. L. Karle, Acta Crystallogr., 1966, 21, 849. 21 E. Egert and G. M. Sheldrick, Acta Crystallogr., Sect. A., 1988, 41, 262. 22 J. Karle, Acta Crystallogr., 1968, 24, 182. 23 I. L. Karle, J. L. Flippen-Anderson, K. Uma and P. Balaram, Int. J. Peptide Protein Res., 1988, 32, 536. 24 S. Datta, N. Shamala, A. Banerjee and P. Balaram, Int. J. Peptide Protein Res., in press. 25 E. N. Baker and R. E. Hubbard, Progr. Biophys. Mol. Biol., 1984, 44, 97. 26 I. L. Karle, J. L. Flippen-Anderson, K. Uma and P. Balaram, Biopolymers, 1990, 29, 1835. 27 I. L. Karle, R. B. Rao, R. Kaul, S. Prasad and P. Balaram, Biopolymers, 1996, 39, 75. 28 I. L. Karle, R. Kaul, R. B. Rao, S. Raghothama and P. Balaram, submitted for publication in J. Am. Chem. Soc. 29 IUPAC-IUB Commission on Biochemical Nomenclature, Eur. J. Biochem., 1970, 17, 193. Paper 7/02109G Received 26th March 1997 Accepted 29th May 1997

ISSN:1472-779X    

DOI:10.1039/a702109g

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1665 Stannanes as free-radical reducing agents: an ab initio study of hydrogen atom transfer from some trialkyltin hydrides to alkyl radicals Dainis Dakternieks,a David J. Henry a and Carl H. Schiesser *,b a School of Biological and Chemical Sciences, Deakin University, Geelong, Victoria, Australia, 3217 b School of Chemistry, The University of Melbourne, Parkville, Victoria, Australia, 3052 Ab initio molecular orbital calculations using a (valence) double-Ó pseudopotential (DZP) basis set, with (MP2, QCISD) and without (SCF) the inclusion of electron correlation, predict that hydrogen atoms, methyl, ethyl, isopropyl and tert-butyl radicals abstract hydrogen atoms from stannane and trimethyltin hydride via transition states in which the attacking and leaving radicals adopt a colinear arrangement.Transition states in which (overall) Sn]C separations of 3.50 Å have been calculated; these distances appear to be independent of the nature of the attacking radical and alkyl substitution at tin.At the highest level of theory (QCISD/DZP//MP2/DZP), energy barriers (ƒE1 ‡) of 18–34 kJ mol21 are predicted for the forward reactions, while the reverse reactions (ƒE2 ‡) are calculated to require 140–170 kJ mol21. These values are marginally affected by the inclusion of zero-point vibrational energy correction. Importantly, QCISD and MP2 calculations predict correctly the relative order of radical reactivity toward reduction by stannanes: tert-butyl > isopropyl > ethyl.By comparison, SCF/DZP, AM1 and AM1(CI = 2) calculations perform somewhat more poorly in their prediction of relative radical reactivity. Introduction Free-radical chemistry has benefited enormously from the invention of tin-based chain-carrying reagents.1–3 Of these, tributyltin hydride and to a lesser extent, triphenyltin hydride, have been the reagents of choice.1 Their ready availability and favourable rate constants for attack of the corresponding tincentred radicals at a variety of radical precursors,4 coupled with useful rate constants for hydrogen transfer 5–8 to alkyl and other radicals, provide for reagents superior to their silicon 9,10 and germanium6,11 counterparts; only tris(trimethylsilyl)silane rivals trialkyltin hydrides in its synthetic utility.10 The transformation of 6-bromohex-1-ene (1) into methylcyclopentane by the action of tributyltin hydride (Scheme 1) typifies the chemistry in question.A knowledge of rate constants is crucial to the successful design of synthetic procedures involving these reagents. Giese points out that stannane chain-carrying reagents are useful because a knowledge of the important rate constants (kc, kH and kBr in Scheme 1) allow, through control of substrate concentration, necessary selectivity criteria to be met.1 Specifically the hex-5-enyl radical (2) must undergo intramolecular addition to form the cyclopentylmethyl radical Scheme 1 • • Br Bu3Sn kc 2 kH 3 Bu3SnH 1 Bu3SnBr kBr • (3), that 3 must abstract a hydrogen atom from tributyltin hydride and that the tributylstannyl radical must abstract the halogen in 1 to form 2.These processes must proceed faster than any competing side reaction. Recently, our interests in the development of modified stannanes for use in free-radical synthesis necessitated our computer modelling of the radical reactions of stannanes and stannyl radicals through the use of ab initio molecular orbital theory.We recently published the results of high-level ab initio investigations into the attack of silyl, germyl and stannyl radicals at the halogen atom in halomethanes and the chalcogen atom in the analogous sulfides, selenides and tellurides.12 These studies predicted that, in accordance with expectation, stannyl radicals react with halogen and chalcogen containing substrates in the order I > Te >> Br > Se >> Cl > S.In addition, reactions involving tellurides are calculated to be reversible, a prediction we have recently verified.13 To the best of our knowledge there are no ab initio reports detailing hydrogen atom transfer from tin to carbon (or any other) centred radical. Beckwith and Zavitsas reported the results of AM1 (semiempirical) calculations on reactivity and diastereoselectivity during stannane reduction of several dioxolanyl radicals.14 These calculations suggest that AM1 is capable of reproducing experimentally observed diastereoselectivities with good levels of correlation with experimentally available data.In order to provide further insight into the intimate details of hydrogen atom transfer from stannanes, we have examined the potential energy surfaces for the attack of hydrogen atom, methyl, ethyl, isopropyl and tert-butyl radicals at the hydrogen atom in stannane (SnH4) with expulsion of stannyl radical, and the analogous reaction of hydrogen atom and methyl radical with trimethyltin hydride (Me3SnH) by ab initio molecular orbital theory and, for comparision in some cases, AM1 (semiempirical) calculations.Methods All ab initio molecular orbital calculations were carried out using the Gaussian 92 15 or Gaussian 94 16 program. Geometry1666 J. Chem. Soc., Perkin Trans. 2, 1997 optimisations were performed using standard gradient techniques at the SCF and MP2 levels of theory using RHF and UHF methods for closed and open shell systems, respectively.17 Further single-point QCISD calculations were performed on each of the MP2 optimised structures. When correlated methods were used calculations were performed using the frozen core approximation.Vibrational frequencies were calculated on each SCF-calculated structure and at the MP2 level on the reactants, products and transition states involved in the reaction of hydrogen atom and methyl radical with stannane (SnH4).Where appropriate, zero-point vibrational energy (ZPE) corrections have been applied. All ab initio calculations were performed using the previously published DZP basis set 12 on a Sun SparcStation 5, Cray YMP4E/ 364 or Cray J916 computer. AM1 and AM1(CI = 2) calculations were performed within Gaussian 92 or AMPAC 5.0 18 on a Sun SparcStation 2 or Sun SparcStation 5. Results and discussion Reaction of hydrogen atom with stannane (SnH4) and trimethyltin hydride (Me3SnH) Species of C3v symmetry (4, 5) were located on the SnH5 and Me3SnH2 potential energy surfaces at the SCF/DZP and MP2/ DZP levels of theory.These structures were found to correspond to the transition states for transfer of hydrogen atom from the tin centre to hydrogen atom (Scheme 2; R = H) and are displayed in Fig. 1, while the calculated energy barriers for these reactions are listed in Table 1 together with the calculated (imaginary) stretching frequency associated with the reaction coordinate in each case.Calculated energies of all structures in this study are listed in Table 2. The data displayed in Table 1 reveal calculated energy barriers of 39.8 (SCF/DZP), 27.3 (MP2/DZP) and 20.6 kJ mol21 (QCISD/DZP//MP2/DZP) for the abstraction of hydrogen atom from stannane (DE1 ‡) with barriers for the reverse reaction (DE2 ‡) of 139.6, 146.6 and 150.1 kJ mol21 at increasing levels of theory respectively. Inclusion of zero-point vibrational energy correction (ZPE) serves to lower slightly the forward barriers (DE1 ‡) by a maximum of 2.2 kJ mol21, while the reverse barriers (DE2 ‡) are also lowered by 6.8–8.8 kJ mol21.These data clearly emphasise the need for inclusion of zeropoint energies in reactions of this type. It is interesting to note that methyl substitution on tin in moving from stannane to trimethyltin hydride serves to lower the barrier for the forward reaction (DE1 ‡) by only 0.4 to 3.1 kJ mol21, with reductions in the reverse barrier (DE2 ‡) of approximately 10 kJ mol21 at each level of theory.Despite this, these reactions are predicted to be significantly exothermic at each level of theory. These data are to be compared with the energy barriers calculated for homolytic substitution by a hydrogen atom at the tin atom in stannane and methylstannane with expulsion of hydrogen atom and methyl radical, respectively.19 Barriers of between Scheme 2 R'SnH2 + R R'3SnH DE2 ‡ + RH 4–11 DE1 ‡ R = H, Me, Et, Pri, But R' = H, Me • • 68.9 kJ mol21 (QCISD/DZP 1 ZPE) and 116.9 kJ mol21 (SCF/ DZP 1 ZPE) for the former reaction with values ranging from 95.3 kJ mol21 (QCISD/DZP) to 109.3 kJ mol21 (SCF/ DZP 1 ZPE) for the latter reaction indicate strongly that, as expected,1 hydrogen abstraction is preferred over homolytic substitution at the tin atom in each case.Inspection of Fig. 1 reveals a pleasing level of correlation between the SCF and MP2 generated transition state structures (4, 5).At the lower level, H]H separations of 1.205 (4) and 1.192 Å (5) are predicted, while Sn]HTS separations of 1.835 and 1.849 Å are calculated for 4 and 5, respectively. Inclusion of electron correlation (MP2) serves to marginally alter the position of the transferring hydrogen atom in each transition state without altering the overall gross transition state structure. Separations of 1.271 and 1.280 Å (H]H in 4 and 5, respectively), coupled with Sn]HTS distances of 1.789 (4) and 1.795 Å (5) lead to overall Sn]Hattack distances of 3.060 and 3.075 Å in structures 4 and 5, respectively.These values are very similar to those calculated at the SCF level of theory, namely 3.040 and 3.041 Å. We also examined the AM1 potential energy surfaces for the reactions described above. Unfortunately, AM1 calculations provided data of questionable quality; values of DE1 ‡ were calculated to be 0.04 and 0.01 kJ mol21 for reactions involving transition states 4 and 5, respectively.Beckwith and Zavitsas also report poor results for the reaction of hydrogen atom with H2, where a negative activation energy is predicted by AM1.14 It seems that AM1 may have problems modelling reactions involving the hydrogen atom in general. Accordingly, we urge caution in the use of AM1 under these circumstances. Reaction of methyl, ethyl, isopropyl and tert-butyl radicals with stannane (SnH4) and trimethyltin hydride (Me3SnH) Extensive searching of the potential energy surfaces for the hydrogen atom transfer reactions involving stannane and methyl, ethyl, isopropyl and tert-butyl radicals, as well as trimethyltin hydride and the methyl radical (Scheme 2; R � H), located structures (6–11) as stationary points at each level of theory.These structures proved to be transition states for the transfer of hydrogen atom and were found to adopt colinear arrangements of attacking and leaving radicals (C3v symmetry) in reactions involving methyl and tert-butyl radical (6, 7, 8, 11).In the remaining cases (9, 10), slight deviations from colinearity are predicted (Cs symmetry) with Sn]HTS]C angles ranging from 174.7 to 178.18. The MP2/DZP calculated transition structures are displayed in Fig. 2. Apart from transition state 6 which prefers to adopt an eclipsed conformation, all structures were found to prefer staggered conformations, except for 9 at the AM1 level of theory, where the eclipsed conformation proved to be of lower energy.† Fig. 1 MP2/DZP calculated transition states (4, 5) (SCF data in parentheses) for hydrogen abstraction by hydrogen atoms from stannane and trimethylstannane 1.789 Å (1.835 Å) 1.795 Å (1.849 Å) 1.271 Å (1.205 Å) 1.280 Å (1.192 Å) 4 ( C ) 3v 5 ( C ) 3v † The eclipsed conformation of 8–11 proved to correspond to secondorder saddle-points at the SCF/DZP level of theory.J. Chem. Soc., Perkin Trans. 2, 1997 1667 Table 1 Calculated energy barriers a for the forward (DE1 ‡) and reverse (DE2 ‡) hydrogen atom abstraction reactions of hydrogen atom with stannane (SnH4) and trimethyltin hydride (Me3SnH) (Scheme 2, R = H) and transition state (imaginary) frequency (n) b of structures (4, 5) R H H R9 H Me TS 4 5 Method SCF/DZP MP2/DZP QCISD/DZPd SCF/DZP MP2/DZP QCISD/DZPd DE1 ‡ 39.8 27.3 20.6 39.4 24.2 18.1 DE1 ‡1ZPVEc 37.7 26.1 [19.4] 37.2 (22.0) (15.9) DE2 ‡ 139.6 146.6 150.1 130.4 136.8 140.2 DE1 ‡1ZPVEc 132.5 139.8 [143.3] 121.6 (128.0) (131.4) n 1766i 1436i — 1711i —— a Energies in kJ mol21.b Frequencies in cm21. c Values in parentheses are estimates based on SCF/DZP ZPE corrections. Values in square brackets are estimates based on MP2/DZP ZPE corrections. d QCISD/DZP//MP2/DZP. It is interesting to note that two transition states (6, 7) were identified for the reaction of the methyl radical with stannane. The eclipsed conformation (6) is predicted to be more stable than the staggered structure (7) by only 0.05 kJ mol21 (SCF/ DZP) while AM1 calculations suggest a preference of 0.01 kJ mol21 for 6.These data suggest significant free rotation during the course of this reaction. In this work, we have extensively examined the eclipsed conformation (6). Ab initio calculated energy barriers for these hydrogen atom transfer reactions (DE1 ‡, DE2 ‡, Scheme 2; R9 � H) are listed in Table 3, while the calculated energies of all structures in this study are found in Table 2.AM1 generated data are included for comparison with the work of Beckwith and Zavitsas.14 Inspection of Table 3 reveals a pleasing degree of convergence in the forward energy barriers (DE1 ‡). For example, attack of the methyl radical at stannane is predicted to have associated barriers of 63.3 (SCF/DZP), 31.5 (MP2/DZP) and 31.7 kJ mol21 (QCISD/DZP//MP2/DZP), suggesting that the MP2 level of theory is able to provide acceptable data; improvement in the level of correlation leads to only a minor decrease in DE1 ‡.Similar trends are observed for the other reactions in this study, with QCISD calculated values of DE1 ‡ lying within 5.8 kJ mol21 of the corresponding MP2 value. We speculate that in some cases the QCISD/DZP calculated potential energy surface may differ enough from the MP2/DZP surface to lead to slight discrepancies in the (single-point) QCISD/ DZP//MP2/DZP data. All reactions are predicted to be significantly exothermic, with reverse barriers (DE2 ‡) ranging from 167.7 (6) to 129.7 kJ mol21 (11) at the QCISD level.As was observed for reactions involving hydrogen atom, zero-point vibrational energy correction (ZPE) leads to slight changes in the predicted values of DE1 ‡ (20.9 to 3.9 kJ mol21), while the reverse reactions (DE2 ‡) are affected more strongly (216.2 to 218.7 kJ mol21). Comparing these data with those associated with homolytic substitution by the methyl radical at the tin atom in SnH4 and MeSnH3 with expulsion of a hydrogen atom and methyl radical, respectively, once again suggests that attack at tin is not competitive with hydrogen abstraction.Energy barriers of around 90 kJ mol21 (QCISD/DZP//MP2/DZP) are predicted for methyl radical attack at tin.19 Inspection of Fig. 2 reveals that the overall structures of transition states 6–11 are relatively unaffected by alkyl substitution on either tin or carbon radical centres, or indeed the level of theory employed.The greatest effect appears to be on the absolute position of the hydrogen atom in the transition state during delivery. For example, at the MP2 level of theory, while the Sn]HTS separation is found to vary between 1.873–1.893 Å and the C]HTS distance is predicted to lie in the range 1.572– Sn H H H H C H H H Sn H H H H C H H H 7 6 1.598 Å, the overall Sn]C separation is found to lie in the narrow range of 3.464–3.476 Å. AM1 calculated structures for transition states 6–11 are very similar to those calculated using the ab initio techniques, with the exception of 9 which is predicted to prefer an eclipsed conformation.The transition state distances are predicted to be somewhat shorter than those calculated using SCF/DZP or MP2/DZP techniques with Sn]HTS and C]HTS separations lying between 1.676–1.740 and 1.649–1.777 Å respectively, resulting in (overall) Sn]C distances of between 3.389 and 3.453 Å, somewhat shorter than the corresponding ab initio separations.The intimate transition state geometries ar agreement with those reported by Beckwith and Zavitsas; C]H and Sn]H distances of 1.720 and 1.699 Å are predicted by AM1 for the transition state involved in the reaction of ethyl radical with trimethyltin hydride.14 Of more significance are the calculated energy barriers (DE1 ‡ and DE2 ‡). Table 3 clearly reveals the AM1 calculated trends in Fig. 2 MP2/DZP calculated transition states (6, 8–11) (SCF data in parentheses) for hydrogen abstraction by various alkyl radicals from stannane and trimethylstannane. AM1 calculated data are included for comparison.AM1 - Optimised Structures aEclipsed conformation (see text). TS r (Sn–HTS)/Å r (C–HTS)/Å (Sn–HTS–C)/° 1.676 1.676 1.697 1.718 1.740 1.764 1.777 1.712 1.674 1.649 180.0 180.0 173.4 174.5 180.0 689 a 10 11 q 1.800 Å (1.873 Å) 1.701 Å (1.598 Å) 6 ( C3v) 1.811 Å (1.893 Å) 1.694 Å (1.583 Å) 8 ( C3v) 10 ( Cs) 1.800 Å (1.883 Å) 1.695 Å (1.581 Å) 174.7° (178.1°) 1.802 Å (1.879 Å) 1.695 Å (1.589 Å) 174.9° (178.1°) 9 ( Cs) 1.795 Å (1.887 Å) 1.703 Å (1.572 Å) 11 ( C3v)1668 J.Chem. Soc., Perkin Trans. 2, 1997 Table 2 SCF, MP2, QCISD,a AM1 and AM1(CI = 2) calculated energies b of the reactants, products and transition states (4–11) in this study Structure H? ?CH3 ?CH2CH3 ?Pri ?But ?SnH3 ?SnMe3 CH4 CH3CH3 CH3CH2CH3 (CH3)3CH SnH4 Me3SnH 45689 10 11 SCF/DZP 20.49764 239.57176 278.61706 2117.66350 2156.71009 24.94363 2122.09584 240.20752 279.24900 2118.29208 2157.33569 25.53930 2122.69479 26.02172 2123.17741 245.08692 2162.23971 284.13184 2123.17819 2162.22498 MP2/DZP — 239.69727 278.88130 2118.06814 2157.25730 25.02396 2122.60693 240.36700 279.54741 2118.73143 2157.91861 25.63952 2123.22440 26.12615 2123.71284 245.32419 2162.90911 284.50998 2123.69900 2162.89095 QCISD/DZP — 239.71891 278.91695 2118.11713 2157.31890 25.04697 2122.66550 240.38949 279.58347 2118.78035 2157.97957 25.66633 2123.28707 26.15551 2123.77781 245.37258 2162.99308 284.57176 2123.77340 2162.97716 AM1c — 0.04771 0.02462 0.00562 20.01031 0.07916 20.02014 20.01402 20.02781 20.03876 20.04692 0.06703 0.01369 —— 0.11944 0.06599 0.10006 d 0.08505 0.07415 AM1(CI = 2)c — 0.01190 0.00691 0.00253 20.00111 0.01937 0.00551 20.00365 20.00671 20.00930 20.01131 0.01590 0.00315 —— 0.03093 0.01858 0.02623 0.02257 0.01991 a QCISD/DZP//MP2/DZP.b Energies in hartrees (1 Eh = 2626 kJ mol21).c Heat of formation. d Eclipsed conformation. Table 3 Calculated energy barriers a for the forward (DE1 ‡) and reverse (DE2 ‡) hydrogen atom abstraction reactions of methyl, ethyl, isopropyl and tert-butyl radicals with stannane (SnH4) and trimethyltin hydride (Me3SnH) (Scheme 2, R � H) and transition state (imaginary) frequency (n)b of structures (6–11) R Me Me Et Pri But R9 H Me H H H TS 6 8 9 10 11 Method SCF/DZP MP2/DZP QCISD/DZPd AM1 AM1(CI = 2) SCF/DZP MP2/DZP QCISD/DZPd AM1 AM1(CI = 2) SCF/DZP MP2/DZP QCISD/DZPd AM1 AM1(CI = 2) SCF/DZP MP2/DZP QCISD/DZPd AM1 AM1(CI = 2) SCF/DZP MP2/DZP QCISD/DZPd AM1 AM1(CI = 2) DE1 ‡ 63.3 31.5 31.7 12.4 34.4 70.5 33.0 33.9 12.0 38.8 64.2 26.9 28.7 22.1 37.7 64.5 21.1 24.8 32.6 45.6 64.0 13.8 19.6 45.8 56.2 DE1 ‡1ZPVEc 67.2 33.7 [33.9] —— 73.7 (36.2) (37.1) —— 66.4 (29.1) (30.9) —— 65.1 (21.7) (25.4) —— 63.1 (12.9) (18.7) —— DE2 ‡ 168.6 175.3 167.7 142.6 167.1 167.1 170.2 162.6 157.2 183.7 159.6 161.2 154.1 127.9 149.1 151.0 148.1 141.6 117.2 137.4 142.7 135.5 129.7 110.1 130.2 DE1 ‡1ZPVEc 152.4 158.8 [151.2] —— 148.4 (151.5) (151.2) —— 141.9 (143.5) (136.4) —— 132.7 (129.8) (123.3) —— 124.2 (117.0) (111.2) —— n 1688i 994i — 533i 1464i 1766i —— 535i 1617i 1688i —— 863i 1488i 1673i —— 1154i 1621i 1652i —— 1364i 1689i a Energies in kJ mol21. b Frequencies in cm21.c Values in parentheses are estimates based on SCF/DZP ZPE corrections. Values in square brackets are estimates based on MP2/DZP ZPE corrections.d QCISD/DZP//MP2/DZP. DE1 ‡. Values of 12.4, 22.1, 32.6 and 45.8 kJ mol21 are predicted for reactions involving transition states 6, 9, 10 and 11, respectively. In other words, in moving from the methyl radical to primary, secondary and tertiary radicals as hydrogen abstracting species, the energy barrier (DE1 ‡) is predicted to undergo increases of up to 33.4 kJ mol21. The value of 22.1 kJ mol21 for the reaction involving the ethyl radical with SnH4 compares well with the previously determined value of 21.8 kJ mol21 for the similar reaction involving trimethyltin hydride.14 It is interesting to compare these data with those calculated using ab initio techniques.While SCF/DZP calculations suggest that DE1 ‡ is about 64–70 kJ mol21 in all cases, inclusion of electron correlation results in decreases in DE1 ‡ in moving through the same set of hydrogen abstracting radicals. Barriers of 31.7, 28.7, 24.8 and 19.6 kJ mol21 are predicted at the QCISD/DZP//MP2/DZP level for reactions involving transition states 6, 9, 10 and 11, respectively.These data can be compared with experimentally determined activation energies associated with hydrogen abstraction by primary, secondary and tertiary radicals from tributyltin hydride. Laser-flash photolytic (LFP) techniques have determined activation energies of 13.5, 15.3, 14.5 and 12.3 kJ mol21J. Chem. Soc., Perkin Trans. 2, 1997 1669 for reactions involving methyl, ethyl, isopropyl and tert-butyl radicals, respectively in isooctane–tert-butyl peroxide.7 These experimentally determined activation energies are some 7–18 kJ mol21 lower than our QCISD data.These discrepancies may be attributed to either solvent effects or differences in alkyl substitution on the stannanes used in the experimental and computational studies, or both.† Indeed, Ingold and co-workers suggest that ‘polar factors’ may be responsible for the activation energies for the isopropyl and tert-butyl radicals which ‘seem to be anomalously low’.7 Perhaps all of the reactions in question are affected by polar factors which would lead to lower than expected activation energies in the LFP reaction solvent [1:1:1 tert-butyl peroxide : trialkylphosphine (or trialkylarsine) :Bu3- SnH in isooctane].Encouragingly, the correlated ab initio methods are generally (apart from methyl) capable of reproducing the experimentally observed trends in DE1 ‡ , while SCF/DZP and AM1 calculations are unable to successfully reproduce these observations. These results highlight the importance of including electron correlation in calculations of this type.Inclusion of correlation into AM1 [AM1(CI = 2)] was suggested by Beckwith and Zavitsas to provide ‘a useful practicable tool for predicting the relative rates, regioselectivity and diastereoselectivity of radical reactions of relatively complex substrates’. 14 In the reactions examined in this study, AM1(CI = 2) calculations serve only to worsen both activation energies and associated trends.Values of DE1 ‡ were calculated to range from 34.4 (R = Me) to 56.2 kJ mol21 (R = But) using AM1(CI = 2). We suggest caution in using AM1 and AM1(CI = 2) in predicting trends associated with hydrogen abstraction reactions from stannanes. Conclusions The results presented above indicate that MP2/DZP and QCISD/DZP//MP2/DZP (ab initio) calculations are generally capable of modelling the relative reactivities of primary, secondary and tertiary radicals toward hydrogen atom abstraction from stannanes.Activation energies for abstraction of hydrogen atom from stannane and trimethyltin hydride (DE1 ‡) are predicted to lie between 18 and 34 kJ mol21 (QCISD) and are some 7–18 kJ mol21 higher than experimentally determined activation energies for analogous reactions with tributyltin hydride. These discrepancies may be attributed to solvent or kyl substitution at tin.Interestingly, SCF/DZP, AM1 and AM1(CI = 2) methods, while predicting similar transition state geometries to the higher-level ab initio methods, perform poorly in predicting energy barriers and relative radical reactivities. We urge caution in the use of these methods in modelling stannane reductions. While our results are not inconsistent with those reported recently by Beckwith and Zavitsas,14 we suspect that the success of AM1 and AM1(CI = 2) calculations in predicting diastereoselectivities in stannane reductions of dioxolanyl radicals is partly due to the geometric insensitivity of the transition states for hydrogen transfer to the nature of the attacking radical and the fact that two faces of the same radical are involved. Errors in activation energy are likely to cancel.Acknowledgements We thank the Australian Research Council for financial support. We also gratefully acknowledge the support of the Ormond Supercomputer Facility, a joint venture of the University of Melbourne and Royal Melbourne Institute of Technology.† A referee suggested that tunnelling effects may also reduce the barrier by an amount which may account for the differences between calculated and experimental data. References 1 B. Giese, Radicals in Organic Synthesis: Formation of Carbon- Carbon Bonds, Pergamon Press, Oxford, 1986. 2 W. P. Neumann, Synthesis, 1987, 665; D. P. Curran, Synthesis, 1988, 417, 489; C.P. Jasperse, D. P. Curran and T. L. Fevig, Chem. Rev., 1991, 91, 1237. 3 For some examples of other synthetically useful stannanes, see: U. Gerigk, M. Gerlach, W. P. Neumann, R. Vieler and V. Weintritt, Synthesis, 1990, 448; J. Light and R. Breslow, Tetrahedron Lett., 1990, 31, 2957; F. Ferkous, D. Messadi, B. De Jeso, M. Degueil- Castaing and B. Maillard, J. Organomet. Chem., 1991, 420, 315; W. P. Neumann and M. Peterseim, React. Polym., 1993, 20, 189; D. P. Curran and S.Hadida, J. Am. Chem. Soc., 1996, 118, 2531; D. P. Curran and D. Nanni, Tetrahedron: Asymmetry, 1996, 7, 2417. 4 C. H. Schiesser and L. M. Wild, Tetrahedron, 1996, 52, 13 265 and references cited therein. 5 D. J. Carlsson and K. U. Ingold, J. Am. Chem. Soc., 1968, 90, 1055; D. J. Carlsson and K. U. Ingold, J. Am. Chem. Soc., 1968, 90, 7047; M. Newcomb, Tetrahedron, 1993, 49, 1151; D. V. Avila, K. U. Ingold, J. Lusztyk, W. R. Dolbier, Jr., H.-Q. Pan and M. Muir, J. Am. Chem. Soc., 1994, 116, 99; S.J. Garden, D. V. Avila, A. L. J. Beckwith, V. W. Bowry, K. U. Ingold and J. Lusztyk, J. Org. Chem., 1996, 61, 805. 6 L. J. Johnston, J. Lusztyk, D. D. M. Wayner, A. N. Abeywickrema, A. L. J. Beckwith, J. C. Scaiano and K. U. Ingold, J. Am. Chem. Soc., 1985, 107, 4594. 7 C. Chatgilialoglu, K. U. Ingold and J. C. Scaiano, J. Am. Chem. Soc., 1981, 103, 7739. 8 O. M. Musa, J. H. Horner, H. Shahin and M. Newcomb, J. Am. Chem. Soc., 1996, 118, 3862. 9 J. Lusztyk, B.Maillard and K. U. Ingold, J. Org. Chem., 1986, 51, 2457; M. Newcomb and S. U. Park, J. Am. Chem. Soc., 1986, 108, 4132; C. Chatgilialoglu, Acc. Chem. Res., 1992, 25, 188; C. Chatgilialoglu, A. Guerrini and M. Lucarini, J. Org. Chem., 1992, 57, 3405; C. Chatgilialoglu, C. Ferreri and M. Lucarini, J. Org. Chem., 1993, 58, 249. 10 C. Chatgilialoglu, D. Griller and M. Lasage, J. Org. Chem., 1988, 53, 3641; M. Ballestri, C. Chatgilialoglu, K. B. Clark, D. Griller, B. Giese and B.Kopping, J. Org. Chem., 1991, 56, 678. 11 J. Lusztyk, B. Maillard, D. A. Lindsay and K. U. Ingold, J. Am. Chem. Soc., 1983, 105, 3578; J. Lusztyk, B. Maillard, S. Deycard, D. A. Lindsay and K. U. Ingold, J. Org. Chem., 1987, 52, 3509. 12 C. H. Schiesser, B. A. Smart and T.-A. Tran, Tetrahedron, 1995, 51, 3327; C. H. Schiesser and B. A. Smart, Tetrahedron, 1995, 51, 6051; see correction in: C. H. Schiesser, B. A. Smart and T.-A. Tran, Tetrahedron, 1995, 51, 10 651. For performance of DZP basis set see: B. A. Smart and C. H. Schiesser, J. Comput. Chem., 1995, 16, 1055. 13 C. H. Schiesser and M. A. Skidmore, Chem. Commun., 1996, 1419. 14 A. L. J. Beckwith and A. A. Zavitsas, J. Am. Chem. Soc., 1995, 117, 607. 15 M. J. Frisch, G. W. Trucks, M. Head-Gordon, P. M. W. Gill, M. W. Wong, J. B. Foresman, B. G. Johnson, H. B. Schlegel, M. A. Robb, E. S. Replogle, R. Gomperts, J. L. Andres, K. Raghavachari, J. S. Binkley, C. Gonzales, R. L. Martin, D. J. Fox, D. J. Defrees, J. Baker, J. J. P. Stewart and J. A. Pople, GAUSSIAN 92, Revision F, Gaussian Inc., Pittsburgh, PA, 1992. 16 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman, T. Keith, G. A. Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-Laham, V. G. Zakrzewski, J. V. Ortiz, J. B. Foresman, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. J. P. Stewart, M. Head-Gordon, C. Gonzalez and J. A. Pople, GAUSSIAN 94, Revision B.3, Gaussian Inc., Pittsburgh, PA, 1995. 17 W. J. Hehre, L. Radom, P.v. R. Schleyer and P. A. Pople, Ab Initio Molecular Orbital Theory, Wiley, New York, 1986. 18 AMPAC 5.0, © 1994 Semichem, 7128 Summitt, Shawnee, KS 66216. 19 C. H. Schiesser, M. L. Styles and L. M. Wild, J. Chem. Soc., Perkin Trans. 2, 1996, 2257. Paper 7/01822C Received 14th March 1997 Accepted 23rd April 1997

ISSN:1472-779X    

DOI:10.1039/a701822c

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

J. Chem. Soc., Perkin Trans. 2, 1997 1671 Synthesis, spectroscopy and electrochemistry of phthalocyanine derivatives functionalised with four and eight peripheral tetrathiafulvalene units Changsheng Wang, Martin R. Bryce,* Andrei S. Batsanov, Claire F. Stanley, Andrew Beeby and Judith A. K. Howard Department of Chemistry, University of Durham, Durham, UK DH1 3LE Metal-free phthalocyanine derivatives 2 and 14 bearing eight and four peripheral tetrathiafulvalene (TTF) units, respectively, have been synthesised, and their solution electrochemistry and optical spectroscopy have been studied.The compounds display redox properties arising from the TTF and from the phthalocyanine groups. 1H NMR and UV–VIS spectroscopic studies in solution show that aggregation is strongly solvent dependent. Quenching of the fluorescence of the phthalocyanine core by the TTF units was observed. The X-ray crystal structure of 4,5-bis(hexylthio)-49,59-bis(hydroxymethyl)-TTF 11, which was synthesised during the course of this work, has been determined.The hydroxy groups of 11 engage in intermolecular (and interstack) hydrogen bonds. Computer modelling studies on phthalocyanine derivatives 2 and 14 are reported. Introduction For many years phthalocyanine derivatives have attracted considerable attention as a consequence of their diverse electronic, optical, structural and coordination properties, which offer applications in the fields of non-linear optics, liquid crystals, Langmuir–Blodgett films, electrochromic devices, molecular metals, gas sensors, photosensitisers and diagnostic and therapeutic agents in pharmacology.1 There is keen current interest in exploring new structural modifications to the phthalocyanine system, including the study of binuclear derivatives, 2 heterocyclic analogues 3 in which the benzene rings are replaced by nitrogen-containing 3a–d or sulfur-containing 3e,f heterocycles, and derivatives with specially-designed ‘active’ peripheral substituents, viz.thiolate groups which coordinate transition metals,4 a fullerene moiety which undergoes electrochemical reduction,5 flexible unsymmetrical substituents which result in glasses,6 mesogenic metal-chelated crown ethers which form nanometre-sized molecular cables,7 glycol side-chains to hinder aggregation and thereby increase the luminescence quantum yield8 and liquid crystalline ferrocenyl- and tetrathiafulvalenyl-phthalocyanine derivatives.9 In a recent communication we reported the synthesis of compound 1 bearing eight tetrathiafulvalene (TTF) substituents which provided the first example of the covalent attachment of phthalocyanine (Pc) and TTF units.10 Subsequently Cook and co-workers described a phthalocyanine system functionalised with one TTF unit which displayed liquid crystalline behaviour.9b We chose TTF as a substituent for our studies because of its well-known ability to form stable cation radicals, 11 and we were interested in exploring the electrochemical and optical properties of the different redox units in Pc–(TTF)x (x = 4 and 8) assemblies.The extreme insolubility of compound 1 in almost all organic solvents severely hindered purification and characterisation studies, but nonetheless 1H NMR spectra were obtained in [2H6]dimethyl sulfoxide ([2H6]DMSO) and optical spectroscopic data and cyclic voltammograms were obtained in dimethylformamide (DMF).In an extension of this work, we now report the synthesis and characterisation of an analogous Pc–(TTF)8 system 2, which by virtue of the sixteen N N N O O HN O O N N N O O NH O O S S S S S S S S S S S S S S S S S S S S S S S S S S S S S S S S SR SR SR SR SR SR SR SR RS RS RS RS RS RS RS RS 1 SR = H 2 SR = S-C6H131672 J. Chem. Soc., Perkin Trans. 2, 1997 S S S S SR SR MeO2C MeO2C O S S S S S SR SR MeO2C MeO2C S S S S SR SR HOH2C HOH2C S S S S SR SR X O O S S S S S S S S SR SR SR SR Y Z S S S S SR SR 3 4 + O O NC NC 5 11 6 X = CO2Me 7 X = CH2OH 8 9 10 Y = Z = Br Y = Z = CN Y = Br; Z = CN NC NC CH2Br CH2Br 2 13 14 12 i ii iii iv v vi vii viii vi Scheme 1 R = C6H13, Reagents and conditions: i, triethyl phosphite, 120 8C; ii, lithium bromide, HMPA, 80 8C; iii, DIBALH, THF, 278 to 20 8C; iv, NaH, THF, 40 8C, then 1,2-dibromo-4,5-bis(bromomethyl)benzene, 20 8C; v, copper cyanide, DMF, 140 8C; vi, lithium pentoxide, pentanol, 125 8C; vii, DIBALH, THF, 210 to 20 8C; viii, NaH, THF, reflux, then compound 12, 20 8C N N N HN N N N NH S S S S SR SR O O S S S S SR RS O O S S S S SR RS O O S S S S RS RS O O 14 R = C6H13 peripheral hexylthio substituents is soluble in organic solvents, and the first example of a Pc–(TTF)4 derivative, viz.compound 14, which also contains solubilising substituents. Results and discussion Synthesis The synthesis of our target molecules 2 and 14 is shown in Scheme 1. The key intermediate for both compounds is TTF derivative 5 which was obtained in 63% yield by the standard phosphite-induced cross-coupling 12 of the 1,3-dithiole-2-one and -2-thione fragments 3 and 4.One of the ester substituents of compound 5 was removed by treatment with lithium bromide in hexamethylphosphoric triamide (HMPA) at 80 8C,13 to afford monoester derivative 6 (88% yield) which was reduced with diisobutylaluminium hydride (DIBALH) to yield the alcohol derivative 7 (86% yield). Compound 7 was deprotonated by treatment with sodium hydride in dry tetrahydrofuran (THF) and the derived alkoxide anion (2 equiv.) reacted with 2,3-dibromo-4,5-bis(bromomethyl)benzene to yield the bis- (TTF) system 8 (64% yield).The conversion of dibromo derivative 8 into dicyano analogue 9 was the most problematical step in the synthesis of 2: reaction with copper cyanide in DMF at 140 8C produced dicyano derivative 9 in an optimised yield ofJ. Chem. Soc., Perkin Trans. 2, 1997 1673 28% after chromatographic separation from the bromo-cyano analogue 10 (22%) which was always present in the product mixture.After much experimentation, we had previously found10 that lithium pentoxide in pentanol 14 at 140 8C were the optimum reagents and conditions for the tetramerisation of the dinitrile precursor to form phthalocyanine derivative 1: we, therefore, adopted similar conditions for the tetramerisation of 9, and obtained phthalocyanine derivative 2 in 56% yield at 125 8C.For the synthesis of compound 14 bearing four TTF units, we first reduced diester 5 with diisobutylaluminium hydride to yield dialcohol 11 (75% yield), the X-ray crystal structure of which is reported below. Compound 11 was then deprotonated by treatment with sodium hydride (2 equiv.) and the dianion was reacted with 4,5-bis(bromomethyl)phthalonitrile 12 (prepared by bromination of dimethylphthalonitrile using Nbromosuccinimide) to yield compound 13 (12% yield).Although the yield of this step is low (predictably for the synthesis of a 10-membered ring) the use of the phthalonitrile derivative had the advantage of circumventing the cyanation step (cf. the inefficient conversion of 8 into 9, discussed above). Compound 13 was tetramerised to afford phthalocyanine system 14 in 48% yield, by direct analogy with the preparation of 2. Compounds 2 and 14 were both isolated as dark green solids which, in marked contrast to analogue 1,10 were soluble in a range of organic solvents (e.g.chloroform, carbon disulfide, dichloromethane and toluene) but insoluble in alcohols, acetonitrile and dimethylformamide. Spectroscopic data, elemental analysis and MALDI TOF mass spectra were entirely consistent with the proposed structures. The 1H NMR spectra of 2 and 14 at room temperature in CDCl3 gave broad peaks with no fine structure; 13C NMR spectra were obtained at 50 8C in CDCl3 and the peaks for compound 2 were much sharper than those for compound 14, suggesting that the latter compound is more aggregated in CDCl3 solution at the relatively high concentration of the NMR sample.A very recent quantitative study emphasises the important effects of concentration and temperature on the aggregation and hence the 1H NMR chemical shifts of octaalkynylphthalocyanines.15 No EPR signals were observed for samples of 2 or 14, suggesting that the broad NMR spectra were not due to radical impurities. UV–VIS and fluorescence spectroscopy UV–VIS spectra of solutions of compounds 2 and 14 in a mixture of toluene and pyridine (99 : 1 v/v) at 20 8C are shown in Fig. 1. The split Q-band, which is characteristic of a metalfree phthalocyanine, is observed at lmax 665 and 700 nm (for compound 2) and at lmax 674 and 709 nm (for compound 14). The presence of an additional broader hypsochromicallyshifted band in the optical spectra at lmax 636 nm (for 2) and 648 Fig. 1 UV–VIS spectra of compound 2 (solid line, 1.1746 × 1025 M) and compound 14 (dashed line, 1.4376 × 1025 M) in toluene–pyridine solution (99 : 1 v/v) nm (for 14) is indicative of aggregated species.6,7a,8 This band was detected even in extremely dilute solutions of both compounds 2 and 14; however its relative intensity decreased while that of the low energy Q-band increased with increasing dilution.If we assume that the relative intensities of the highest energy and lowest energy bands are a qualitative indication of the extent of aggregation of 2 and 14 in solution, then at the same concentration in chloroform, compound 14 was found to be the more highly aggregated, which is in agreement with the NMR results discussed above.In the 300–500 nm region the spectra show superimposed bands of TTF and phthalocyanine species. In dichloromethane solution, the intensity of the lowest wavelength Q-band of compound 2 decreased and was bathochromically shifted to 642 nm (the other bands shifted only very slightly to 667 and 701 nm), whereas no change was observed for compound 14.These data suggest that compound 2 is less aggregated in dichloromethane than in toluene– pyridine. The luminescence emission from compounds 2 and 14 was found to be of very low intensity and fell in the same spectral region as that of unsubstituted metal-free phthalocyanines. Fluorescence quantum yields were <1024 and 1.7 ± 0.3 × 1023, for 2 and 14, respectively. These values are very much lower than that obtained for a metal-free Pc lacking the TTF moieties, cf.Bu4PcH2, F = 0.50.16 However, we have demonstrated that the fluorescence from Bu4PcH2 is efficiently quenched by the addition of TTF to the solution. Thus, in toluene the rate constant for the quenching of Bu4PcH2 by TTF was found to be diffusion controlled, kQ = 1.1 ± 0.1 × 1010 dm3 mol21 s21 (Fig. 2). This quenching is likely to be due to an intermolecular electron transfer reaction between the excited singlet state of the Pc and TTF.Calculations based upon the known redox potentials of TTF (see below) and Pc derivatives 17 indicate that the excited singlet state of the Pc is capable of oxidising TTF. In the tethered compounds 2 and 14, we propose that rapid intramolecular electron transfer occurs between the Pc core and a peripheral TTF. Cyclic voltammetric studies The solution redox properties of phthalocyanine derivatives 2 and 14 were studied by cyclic voltammetry (CV), and these data are presented in Table 1, along with those of the new TTF systems reported herein.In dry benzonitrile, compounds 5–13 exhibited two reversible redox waves typical of TTF derivatives. 18 It is well known that both ester 18a,b and thioalkyl substituents 18c,d raise the oxidation potential of TTF, and accordingly, compounds 5 and 6 are the poorest donors in the present series of compounds. The redox waves for bis(TTF) derivatives 8, 9 and 10 are assumed to be two-electron couples with no interaction between the two TTF systems.It has been shown in previous CV studies on bis(TTF) derivatives linked by a variety of spacer groups19 that there is usually no observable inter- Fig. 2 Stern–Volmer plot for quenching of the fluorescence from Bu4PcH2 by TTF; tf 5.6 ns; lex 630 nm1674 J. Chem. Soc., Perkin Trans. 2, 1997 action between the TTF ring systems unless they are linked via a single-atom spacer 20 or are directly attached, i.e.bi(tetrathiafulvalenyl). 21 The CV data for phthalocyanine systems 2 and 14 are more interesting. In dichloromethane solution, compound 2 exhibits two TTF redox waves at E1 � �� 10.49 V and E2 � �� 10.91 V, with no discernable oxidation (and only a very shallow reduction) of the phthalocyanine core [Fig. 3(a)]. However, in benzonitrile solution, a third oxidation peak is seen for compound 2 at 11.18 V (the corresponding reduction is not seen on the reverse sweep) and on the cathodic sweep a reduction peak is observed at E� �� 20.51 V [Fig. 3(b)]. Based on literature precedents,17 we assign these additional redox waves to a single-electron oxidation and reduction, respectively, of the phthalocyanine core unit. The differential pulse voltammogram (DPV) of 2 [Fig. 3(c)] shows the Pc reduction wave clearly, although the Pc oxidation wave is more apparent in the CV scan. The electrochemistry of 14 was also found to be solvent dependent. In benzonitrile, the redox waves of the Pc unit were weak, and the oxidation wave was partly hidden under the second TTF wave.However, in dichloromethane, in contrast to 2, both the reduction and the oxidation of the Pc group were clearly observed. The CV and DPV for 14 in dichloromethane are shown in Figs. 4(a) and 4(b), respectively. Based on the current passed for the CV peaks of compounds 2 and 14, we suggest that all the TTF units are oxidised, although it is noteworthy that for compound 2 the second TTF wave was less intense than the first wave, possibly due to adsorption phenomena, instability of the fully oxidised TTF units and/or the relatively slow motion (on the CV timescale) of the TTF units around the Pc centre. At lower scan rates the current for the second TTF wave increased slightly, but even at 10 mV s21 this wave was less intense than the second TTF wave. The solvent dependency of the electrochemistry of both 2 and 13 suggests that the conformation and/or extent of aggregation vary considerably with the solvent, as implied by the UV–VIS spectra discussed above.We have recently obtained qualitatively similar data for a pyrazinoporphyrazine system with appended TTF groups.22 Molecular modelling studies Molecular modelling studies of compounds 2 and 14 have been performed. The energy-minimised structure for compound 2, shown in Fig. 5, possesses four TTF units on each side of the mean macrocycle plane.If the molecules of 2 adopt a similar conformation in solution 23 to the computer-optimised structure, the phthalocyanine core will be partly shielded by the TTF units and their attached alkyl chains; this could be an important factor in modulating the electrochemical properties of the core of the molecule, such as in the case of 2 in dichloromethane. In addition, the aggregation of molecules, which is similarly solvent dependent, is another significant factor influencing the electrochemistry of large molecules.The energy-minimized Table 1 Cyclic voltammetry data a Compound TTF 2 5 6 7 8 9 10 11 13 14 b E1 2� 1 /V 0.34 20.51 0.74 0.66 0.52 0.54 0.56 0.55 0.48 0.61 20.50 E2 2� 1 /V 0.73 0.50 1.08 1.04 0.89 0.92 0.92 0.92 0.86 0.96 0.54 E3 2� 1 /V 0.87 0.92 E4 p/V 1.18 1.15 a Electrodes: working: Pt, 1.6 mm diameter; counter: Pt; reference: Ag/ AgCl. Supporting electrolyte: 0.1 M tetrabutylammonium hexafluorophosphate in benzonitrile. Scan rate: 100 mV s21.b In dichloromethane. structure for derivative 14 (Fig. 6) gives a very different conformation for the molecule, again with some shielding of the phthalocyanine core, arising from folding of the ten-membered rings about their O ? ? ? O vectors. X-Ray crystal structure of compound 11 The molecular structure of 11 is shown in Fig. 7. The crystal packing, illustrated in Fig. 8, is relatively loose (22 Å3 per non- H atom vs. ‘normal’ 18 Å3). All atoms of the n-hexyl chains except the C(7) and C(13) (which are directly bonded to sulfur) are disordered, C(8) to C(12) over two positions (A and B) and C(14) to C(18) over three positions (A, B and C).Both positions of the former chain lie in the same area but in different orientations. The same is true about the A and B positions of Fig. 3 (a) Cyclic voltammogram of compound 2 cyclic voltammogram of compound 2 in benzonitrile (for conditions see Table 1); (c) differential pulse voltammogram of compound 2 in benzonitrile: sample width, 8 ms; pulse amplitude, 50 mV; pulse width 10 ms; pulse period, 200 ms; electrodes as in Table 1J.Chem. Soc., Perkin Trans. 2, 1997 1675 the latter chain, while its C position overlaps (in its terminal part) with the A and B positions of the same chain of a neighbouring molecule [see Fig. 8(b)]. This disorder is probably responsible for the rather poor quality of the crystals. No evidence of a superlattice has been found.The TTF moiety of 11 is folded along the S(1) ? ? ? S(2) and S(3) ? ? ? S(4) vectors by 9 and 108, respectively, in a boat fashion. These moieties form a peculiar stack parallel to the crystallographic z direction [Fig. 8(a)] in which the long axes of adjacent molecules are mutually perpendicular (89.78) and their central C2S4 planes parallel within 28 with uniform interplanar separ- Fig. 4 (a) Cyclic voltammogram of compound 14 in dichloromethane (for conditions see Table 1); (b) differential pulse voltammogram of compound 14 in dichloromethane: same conditions as Fig. 3(c) Fig. 5 Energy-minimised structure of compound 2 (side chains omitted for clarity) viewed side-on to the Pc ring ations of ca. 3.64 Å. Both hydroxy groups are simultaneously donors and acceptors of intermolecular (and interstack) hydrogen bonds (Fig. 7).24 O(1)H ? ? ? O(29) bonds link molecules into Fig. 6 Energy-minimised structure of compound 14 (side chains omitted for clarity): (a) view showing folding of the molecule around the Pc core; (b) view onto the plane of the Pc ring Fig. 7 Perspective view of molecule 11, showing the disorder of hexyl groups and intermolecular hydrogen bonding. Primed atoms are related via inversion centre, double-primed via 21 axis. H atoms (except hydroxy ones) are omitted. Selected bond distances (Å): C(1)]C(4) 1.34(1), S(1)]C(1) 1.759(8), S(2)]C(1) 1.754(9), S(1)]C(2) 1.749(11), S(2)]C(3) 1.788(9), S(3)]C(4) 1.747(9), S(4)]C(4) 1.788(8), S(3)]C(5) 1.767(9), S(4)]C(6) 1.755(9), C(2)]C(3) 1.32(1), C(5)]C(6) 1.35(1), O(1) ? ? ? O(29) 2.705(8), O(2) ? ? ? O(10) 2.707(8).1676 J.Chem. Soc., Perkin Trans. 2, 1997 centrosymmetric dimers, while O(2)H ? ? ? O(10) bonds generate infinite chains of molecules (related via the 21 axis) along the x direction [Fig. 8(a)]. The closest intrastack S ? ? ? S contacts are as long as 3.85 Å [shortest sulfur–carbon, S(1) ? ? ? C(6) of 3.51 Å] and there are no direct interstack S ? ? ? S contacts whatsoever.Conclusions We have synthesised new phthalocyanine derivatives bearing peripheral TTF substituents which are soluble in a range of organic solvents. We have demonstrated that the fluorescence from metal-free Pcs is efficiently quenched by the addition of TTF, which would be a consequence of intermolecular electron transfer between H2Pc* and TTF. Quenching of the fluorescence of compounds 2 and 14 is likely to be due to a rapid intramolecular electron transfer reaction between the excited singlet state of the Pc and TTF.Solution electrochemical studies establish that the redox behaviour of these macromolecules is strongly solvent dependent. We have observed the expected two-stage oxidation of the TTF units, and oxidation and reduction of the Pc core. The energy-minimised structures for compound 2 and 14 suggest that the phthalocyanine core is partly shielded by the TTF units and their attached alkyl chains; this could be an important factor in modulating the electrochemical properties of the core of the molecule.Fig. 8 Crystal packing of 11: (a) hydrogen-bonded chains of molecules; parallel projections down z axis, all but one position of the disordered groups and all H atoms are omitted; (b) overlap of different positions of the same hexyl chain between molecules related via 21 axis Experimental General All reagents and solvents were of commercial quality (purchased from Aldrich or Fluka) and were dried where necessary using standard procedures.Light petroleum refers to the fraction with bp 40–60 8C. 1H NMR spectra were obtained on a VXR 200 spectrometer operating at 200.14 MHz: chemical shifts are quoted downfield of tetramethylsilane. J values are in Hz. 13C NMR spectra were obtained on a VXR 500 spectrometer operating at 100.581 MHz. UV–VIS spectra were recorded on a Unicam UV2 spectrometer and the data were processed by a Vision Scan V2.11 program.Mass spectra (EI and CI) were recorded using a VG7070E instrument operating at 70 eV. MALDI TOF mass spectra were obtained on a Kratos IV instrument in the reflection mode, operating with irradiation from a nitrogen laser at 337 nm. The matrix was 2,5- dihydroxybenzoic acid, and spectra were averaged over 100 pulses whilst scanning across the sample: peak half-widths were between 6 and 10 amu. Melting points were obtained on a Kofler hot-stage microscope apparatus and are uncorrected. Electrochemical data were obtained on a BAS CV-50W Voltammetric Analyser.The counter, working and reference electrodes were Pt wire, Pt disc (1.6 mm diameter from BAS) and Ag/AgCl respectively. Cyclic voltammetry was performed under argon using IR compensation. Molecular modelling studies were performed using the Discover 2.9.7/95.0/3.0.0. program in the Insight II 95.0 package and figures were processed by the XP program in the SHELXTL package.25 Fluorescence spectra and quantum yields were recorded on a Perkin-Elmer LS-50B fluorimeter. Quantum yields were measured using the reported method,26 relative to ZnPc in toluene–pyridine solution (99 : 1 v/v), Ff = 0.30,27 and disulfonated aluminium Pc in methanol, Ff = 0.56.28 4,5-Bis(hexylthio)-49,59-bis(methoxycarbonyl)tetrathiafulvalene 5 4,5-Bis(hexylthio)-1,3-dithiole-2-thione 429 (3.67 g, 10 mmol) and 1,3-bis(metoxycarbonyl)-1,3-dithiol-2-one 3 14b (2.34 g, 10 mmol) were mixed with triethyl phosphite (30 ml) and heated to 120 8C with stirring under argon for 50 min.Removal of the phosphite in vacuo gave a deep red oily residue which was chromatographed on a silica column [eluent light petroleum– dichloromethane (1 : 1 v/v)] to give compound 5 as a dark red oil (3.50 g, 63%) (Calc. for C22H32O4S6: C, 47.79; H, 5.83. Found: C, 48.13; H, 5.99%); dH(CDCl3) 3.84 (s, 6H) 2.80 (t, J 7.2, 4H), 1.63 (m, 4H), 1.30 (m, 12H), 0.88 (m, 6H); dC(CDCl3) 160.45, 132.49, 131.14, 130.54, 128.47, 53.86, 36.93, 31.80, 30.20, 28.69, 23.04, 14.53. 4,5-Bis(hexylthio)-49-methoxycarbonyltetrathiafulvalene 6 Compound 5 (3.29 g, 5.95 mmol) and lithium bromide (1.0 g) were mixed with HMPA (40 ml) followed by the addition of 2 drops of water with stirring. The reaction mixture was warmed gradually from room temperature to 80 8C and then stirred at 80 8C for 1.5 h. The mixture was then cooled to room temperature, water (250 ml) and hydrochloric acid (2 drops) were added then the mixture was extracted with chloroform (3 × 30 ml).The organic layer was separated, dried (MgSO4), concentrated in vacuo and then chromatographed [silica column, eluent light petroleum–dichloromethane (1: 1 v/v)] yielding compound 6 as a red–orange oil (2.6 g, 88%) (Calc. for C20H30O2S6: C, 48.54; H, 6.11. Found: C, 48.50; H, 6.13%); dH(CDCl3) 7.35 (s, 1H), 3.82 (s, 3H), 2.81 (dt, J 7.2, 4H), 1.63 (m, 4H), 1.30 (m, 12H), 0.89 (m, 6H). 4,5-Bis(hexylthio)-49-(hydroxymethyl)tetrathiafulvalene 7 Compound 6 (2.6 g, 5.25 mmol) was dissolved in dry THF (50 ml) and the solution was cooled to 278 8C under argon atmosphere.DIBALH (16 ml, 16 mmol, 1.0 M solution in hexane) wasJ. Chem. Soc., Perkin Trans. 2, 1997 1677 added dropwise to the stirred solution. The reaction mixture was left in the cooling bath overnight allowing the temperature to rise gradually to 20 8C: an orange solution formed. Methanol was added very slowly until gas evolution subsided.Solvent (ca. 35 ml) was removed by evaporation in vacuo and 2 M hydrochloric acid (ca. 50 ml) was added to the gel-like orange residue. Extraction of the mixture with chloroform (3 × 30 ml) gave an orange solution which was dried (MgSO4) and chromatographed on a silica column (eluent dichloromethane) affording compound 7 as orange oil (2.11 g, 86%) which crystallised by cooling overnight in a freezer, mp 37.0–38.5 8C (Calc. for C19H30OS6: C, 48.88; H, 6.48.Found: C, 48.88; H, 6.49%); dH(CDCl3) 6.24 (s, 1H), 4.40 (d, J 6.2, 2H), 3.75 (t, J 6.6, 1H), 2.81 (t, J 7.2, 4H), 1.61 (m, 4H), 1.30 (m, 12H), 0.89 (m, 6H). 1,2-Dibromo-4,5-bis[49,59-bis(hexylthio)tetrathiafulvalen-4-ylmethoxymethyl] benzene 8 Compound 7 (2.10 g, 4.50 mmol) was dissolved in dry THF (50 ml) at ambient temperature under argon and sodium hydride (210 mg, 60% dispersion in mineral oil, 5.25 mmol) was added in one portion. The mixture was warmed to 40 8C and stirred for 2 h. 1,2-Dibromo-4,5-bis(bromomethyl)benzene 30 (0.95 g, 2.25 mmol) was added and the mixture was stirred at room temperature for 3 h to afford an orange–yellow solution. The insoluble precipitate was removed by suction filtration and the filtrate was concentrated in vacuo to afford an oil which was chromatographed on a silica column [eluent light petroleum– dichloromethane (1 : 1 v/v)] to give compound 8 as an orange oil (1.72 g, 64%) (Calc. for C46H64Br2O2S12: C, 46.29; H, 5.40.Found: C, 46.38; H, 5.41%); dH(CDCl3) 7.63 (s, 2H), 6.26 (s, 2H), 4.48 (s, 4H), 4.29 (s, 4H), 2.81 (t, J 7.2, 8H), 1.63 (m, 8H), ca. 1.30 (m, 24H), 0.89 (m, 12H). 4,5-Bis[49,59-bis(hexylthio)tetrathiafulvalen-4-ylmethoxymethyl] benzene-1,2-dicarbonitrile 9 and 2-bromo-4,5-bis[49,59- bis(hexylthio)tetrathiafulvalen-4-ylmethoxymethyl]benzene-1- carbonitrile 10 Compound 8 (506 mg, 0.424 mmol) and CuCN (350 mg, 3.9 mmol) in dry DMF (5 ml) were heated to 140 8C and stirred for 12 h under argon.After cooling to room temperature, inorganic solids in the reaction mixture were removed by suction filtration and washed with acetone. A mixture of water (20 ml) and 35% aqueous ammonia solution (2 ml) was added to the filtrate and the brown oil solids which precipitated were collected by decantation and suction filtration and chromatographed [silica column, eluent light petroleum–dichloromethane 1: 1 (v/v)] giving compound 10 as a brownish orange oil (103 mg, 22%) followed by compound 9 as a brown oil (128 mg, 28%).Compound 9 (Calc. for C48H64N2O2S12: C, 53.10; H, 5.94; N, 2.58. Found: C, 52.82; H, 5.96; N, 2.19%); dH(CDCl3) 7.88 (s, 2H), 6.32 (s, 2H), 4.58 (s, 4H), 4.38 (s, 4H), 2.82 (t, J 7.2, 8H), 1.63 (m, 8H), 1.30 (m, 24H), 0.89 (m, 12H). Compound 10 (Calc. for C47H64BrNO2S12: C, 49.53; H, 5.66; N, 1.23. Found: C, 49.59; H, 5.75; N, 1.52%); dH(CDCl3) 7.88 (s, 1H), 7.66 (s, 1H), 6.294 (s, 1H), 6.288 (s, 1H), 4.57 (s, 2H), 4.49 (s, 2H), 4.35 (s, 2H), 4.32 (s, 2), 2.81 (t, J 7.2, 8H), 1.63 (m, 8H), 1.30 (m, 24H), 0.89 (m, 12H). 4,5-Bis(hexylthio)-49,59-bis(hydroxymethyl)tetrathiafulvalene 11 Compound 5 (3.20 g, 5.79 mmol) was dissolved in dry THF (120 ml) under an argon atmosphere and the solution was cooled to 210 8C. DIBALH (35 ml, 35 mmol, 1.0 M solution in hexane) was added dropwise. After stirring at this temperature for 2 h, the cooling bath was removed and stirring was continued for an additional 0.5 h.The orange reaction mixture was evaporated in vacuo until ca. 30 ml of an oily residue remained. Hydrochloric acid (1 M) was added to the residue very slowly at 270 8C until gas evolution ceased, then HCl (1 M, 50 ml) was added. The mixture was extracted with dichloromethane (3 × 50 ml) and the red organic phase was separated, dried (MgSO4) and chromatographed (silica column, eluent ethyl acetate), giving compound 11 as yellow–orange crystals (2.15 g, 75%), mp 97.5–99.0 8C.Recrystallisation from ethyl acetate gave yellow plates which were suitable for the X-ray crystallographic analysis (Calc. for C20H32O2S6: C, 48.35; H, 6.49. Found: C, 48.39; H, 6.51%); dH(CDCl3) 4.41 (d, J 6.0, 4H), 2.81 (t, J 7.2, 4H), 2.50 (br s, 2H), 1.63 (m, 4H), 1.30 (m, 12H), 0.89 (t, J 6.6, 6H). 4,5-Bis(bromomethyl)phthalonitrile 12 4,5-Dimethylphthalonitrile 31 (prepared by reacting 4,5- dibromo-o-xylene with CuCN in DMF; 1.56 g, 10 mmol), N-bromosuccinimide (NBS) (3.80 g, 21.3 mmol) and benzoyl peroxide (ca. 10 mg) were mixed in carbon tetrachloride (10 ml) and the mixture was stirred vigorously and warmed outdoors under bright sunshine for ca. 1.5 h (returning to the shade at intervals when the reaction became too violent) until all the NBS was converted to succinimide which was floating on the surface of the solution. The mixture was filtered by suction and the filtrate was evaporated in vacuo to yield a pale yellow solid residue.The precipitate in the Büchner funnel was triturated with a large amount of water and a second portion of pale yellow solid was obtained. The two portions of solid were combined and chromatographed on a silica column (eluent dichloromethane), giving a white oil (3.0 g), which crystallised in methanol to give 12 as white crystals (2.23 g, 71%), mp 110–122 8C (Calc. for C10H16Br2N2: C, 38.26; H, 1.93; N, 8.92. Found: C, 37.72; H, 1.77; N, 8.43%); m/z (%): 314 [M1] (7.79 EI), 332 [M1 1 18] (0.43, DCI); dH(CDCl3) 7.83 (s, 2H), 4.61 (s, 4H). 2-[4,5-Bis(hexylthio)-1,3-dithiol-2-ylidene]-2H,4H,6H,11H, 13H-[1,3]dithiol[4,5-d][2,7]benzodioxecine-8,9-dicarbonitrile 13 To a solution of compound 11 (737 mg, 1.48 mmol) in dry THF (50 ml) under argon, sodium hydride (120 mg, 3 mmol, 60% dispersion in mineral oil) was added in one portion and the mixture was stirred and refluxed for 4 h. Compound 12 (460 mg, 1.47 mmol) was added and the mixture was stirred at room temperature for 4 h to obtain a brown solution.The precipitate which formed during the reaction was removed by suction filtration through Celite and the filtrate was concentrated and chromatographed (silica column, eluent dichloromethane), affording compound 13 as yellow plates (110 mg, 11.6%), mp 175.0–175.5 8C (Calc. for C30H36N2O2S6: C, 55.52; H, 5.59; N, 4.32. Found: C, 55.29; H, 5.49; N, 4.16%); dH(CDCl3) 7.89 (s, 2H), 4.76 (s, 4H), 4.17 (s, 4H), 2.85 (t, J 7.2, 4H), 1.65 (m, 4H), 1.30 (m, 12H), 0.89 (t, J 6.6, 6H). 2,3,9,10,16,17,23,24-Octakis[49,59-bis(hexylthio)tetrathiafulvalen- 4-ylmethoxymethyl]phthalocyanine 2 To a lithium pentoxide solution [prepared from lithium (180 mg) and dry pentanol (15 ml)], compound 9 (275 mg) was added and the mixture was suddenly heated to 125 8C and vigorously stirred under argon for 5 h. The colour changed from yellow to intense green. The solvent was removed in vacuo and ethanol (20 ml) was added to the green residue.Acetic acid (20 ml) was added and the mixture was left overnight. A dark green solid was obtained by decantation, suction filtration and washing with a large amount of ethanol. The solid was extracted thrice with boiling ethanol then dissolved in dichloromethane (20 ml) to get an intense green solution. This solution was filtered through Celite then evaporated in vacuo, giving compound 2 as a dark green amorphous solid (153 mg, 56%) (Calc. for C192H258N8O8S48: C, 53.07; H, 5.98; N, 2.58.Found: C, 53.32; H, 6.02; N, 2.51%); MALDI-MS, m/z, Found 4308; Calc. for C192H258N8O8S48 4339; dH(CDCl3, 50 8C) 8.53 (br, 1H), 6.65 (br s, 1H), 4.99 (br s, 2H), 4.72 (br s, 2H), 2.85 (t, J 6.8, 2H), 2.78 (t, J 6.8, 2H), 1.66 (m, 4H), 1.31 (m, 12H), 0.90 (t, J 6.8, 6H); dC(CDCl3, 50 8C) 148.03, 137.82, 134.55, 127.96, 122.54, 117.25, 113.62, 107.66, 70.76, 68.40, 36.41, 31.38, 31.29, 29.77,1678 J. Chem. Soc., Perkin Trans. 2, 1997 29.68, 28.29, 28.17, 22.54, 22.43, 13.98, 13.90; lmax/nm (toluene) (e) 324.5 (133 580), 636.0 (45 037), 664.5 (42 654), 700.0 (33 459). 2,14,26,38-Tetrakis[4,5-bis(hexylthio)-1,3-dithiol-2-ylidene]- 4H,6H,10H,12H,16H,18H,22H,24H,28H,30H,34H,36H, 40H,42H,46H,48H-tetrakis([1,3]dithiolo[4,5-c]dioxecino)[7,8- b;79,89-k;70,80-t;7-,8--c9]phthalocyanine 14 Following the procedure for the preparation of 2, compound 14 was synthesised from compound 9 (31 mg) and lithium pentoxide [from lithium (55 mg) and dry pentanol (5 ml)] and isolated as a dark green amorphous solid (yield 15 mg, 48%) (Calc.for C120H146N8O8S24: C, 55.48; H, 5.67; N, 4.31. Found: C, 55.39; H, 5.77; N, 4.48%); MALDI-MS, m/z, Found 2588; Calc. for C120H146N8O8S24 2594; dH(CDCl3) ca. 8.2 (v br), ca. 4.6 (br), ca. 4.1 (br), 2.82 (br), 1.33 and 0.93 (overlapping br peaks); lmax/nm (toluene) (e) 333.0 (85 493), 648.5 (39 303), 674.5 (49 320), 709.0 (43 894) nm. X-Ray crystallography The X-ray diffraction experiment on 11 was carried out at T 150 K on a Siemens 3-circle diffractometer with a CCD area detector, using graphite-monochromated Mo-Ka radiation, l - = 0.710 73 Å.Crystal data: C20H32O2S6, M = 496.8; monoclinic, space group P21/c (No. 14), a = 25.359(2), b = 12.554(1), c = 7.867(1) Å, b = 97.26(1)8, V = 2484.4(4) Å3 (from 442 setting reflections with 10 < q < 238), Z = 4, Dc = 1.33 g cm23, F(000) = 1056, m = 5.7 cm21, crystal size 0.5 × 0.25 × 0.06 mm. 12 298 data with 2q < 488 were collected in w scan mode (0.38 steps); of these 3874 data were unique and 2120 ‘observed’ with I > 2s(I). Semi-empirical absorption correction on Laue equivalents was performed, min/max transmission 0.69/0.98, Rint = 0.16 before and 0.14 after correction.The structure was solved by direct methods and refined by full-matrix least squares against F 2 of 3091 non-negative data, sing SHELXTL software.25 Ordered non-H atoms were refined anisotropically, disordered C atoms isotropically [C(8) to C(12) with 50% occupancies, C(14) to C(18) with 33%] with C]C bonds restrained to 1.53(1) Å.Hydroxy H atoms refined, other H atoms ‘riding’ (265 variables, 25 restraints), converging at wR (F 2, all data) = 0.292, goodness-of-fit 1.17, R(F, obs. data) = 0.099; residual electron density features Drmax = 0.39, Drmin = 20.51 e Å23. Atomic coordinates and displacement parameters, bond distances and angles have been deposited at the Cambridge Crystallographic Data Centre.† Acknowledgements We thank EPSRC for funding this work.References 1 Reviews: (a) Phthalocyanines, Properties and Applications, ed. C. C. Leznoff and A. B. P. Lever, vols. 1–3, VCH, New York, 1989– 1993; (b) H. Schultz, H. Lehmann, M. Rein and M. Hanack, Struct. Bonding (Berlin), 1990, 74, 41; (c) M. Hanack and M. Lang, Adv. Mater., 1994, 6, 819; (d) M. J. Cook, J. Mater. Chem., 1996, 6, 677. 2 D. Lelievre, O. Damette and J. Simon, J. Chem.Soc., Chem. Commun., 1993, 939. 3 (a) F. Fernandez-Lazaro, A. Sastre and T. Torres, J. Chem. Soc., Chem. Commun., 1994, 1525; (b) N. Kobayashi, Y. Higashi and T. Osa, J. Chem. Soc., Chem. Commun., 1994, 1785; (c) P. J. Brach, S. J. Grammatica, O. A. Ossanna and L. Weinberger, J. Heterocycl. Chem., 1970, 7, 1403; (d) B. Mohr, G. Wegner and K. Ohta, J. Chem. † For details of the CCDC deposition scheme, see ‘Instructions for Authors’, J. Chem. Soc., Perkin Trans. 2, 1997, Issue 1.Any request to the CCDC for this material should quote the full literature citation and the reference number 188/84. Soc., Chem. Commun., 1995, 995; (e) M. J. Cook and A. Jafari-Fini, J. Mater. Chem., 1997, 7, 5; ( f ) D. M. Knawby and T. M. Swager, Chem. Mater., 1997, 9, 535. 4 C. S. Velazquez, T. F. Baumann, M. M. Olmstead, H. Hope, A. G. M. Barrett and B. M. Hoffman, J. Am. Chem. Soc., 1993, 115, 9997. 5 T. G. Linssen, K. Dürr, M. Hanack and A. Hirsch, J. Chem. Soc., Chem.Commun., 1995, 103. 6 R. D. George and A. W. Snow, Chem. Mater., 1994, 6, 1587. 7 (a) C. F. van Nostrum, S. J. Picken and R. J. M. Nolte, Angew. Chem., Int. Ed. Engl., 1994, 33, 2173; (b) C. F. van Nostrum and R. J. M. Nolte, Chem. Commun., 1996, 2385; (c) C. F. van Nostrum, Adv. Mater., 1996, 8, 1027. 8 J. Vacus and J. Simon, Adv. Mater., 1995, 7, 797. 9 (a) M. J. Cook, G. Cooke and A. Jafari-Fini, J. Chem. Soc., Chem. Commun., 1995, 1715; (b) M. J. Cook, G. Cooke and A.Jafari-Fini, Chem. Commun., 1996, 1925. 10 M. A. Blower, M. R. Bryce and W. Devonport, Adv. Mater., 1996, 8, 63. 11 (a) F. Wudl, D. Wobschall and E. J. Hufnagel, J. Am. Chem. Soc., 1972, 94, 671; (b) K. Deuchert and S. Hünig, Angew. Chem., Int. Ed. Engl., 1978, 17, 875. 12 For reviews of TTF synthesis see: (a) M. Narita and C. V. Pittman, Synthesis, 1976, 489; (b) A. Krief, Tetrahedron, 1986, 42, 1237; (c) G. Schukat and E. Fanghänel, Sulfur Reports, 1993, 13, 254. 13 For related reactions with TTF derivatives see: (a) P. Blanchard, M.Sallé, G. Duguay, M. Jubault and A. Gorgues, Tetrahedron Lett., 1992, 33, 2685; (b) R. P. Parg, J. D. Kilburn, M. C. Petty, C. Pearson and T. G. Ryan, J. Mater. Chem., 1995, 5, 1609. 14 (a) P. A. Barrett, D. A. Frye and R. P. Linstead, J. Chem. Soc., 1938, 1157; (b) review: C. C. Leznoff in ref. 1(a), 1989, vol. 1, p. 1. 15 D. M. Terekhov, K. J. M. Nolan, C. R. McArthur and C. C. Lenznoff, J. Org. Chem., 1996, 61, 3034. 16 A. Beeby and C. F. Stanley, unpublished data. 17 Review: A. B. P. Lever, E. R. Milaeva and G. Speier in ref. 1(a), vol. 3, p. 1. 18 (a) D. C. Green, J. Org. Chem., 1979, 44, 1476; (b) A. S. Batsanov, M. R. Bryce, J. N. Heaton, A. J. Moore, P. J. Skabara, J. A. K. Howard, E. Ortí, P. M. Viruela and R. Vireula, J. Mater. Chem., 1995, 5, 1689; (c) D. L. Lichtenberger, R. L. Johnston, K. Hinkelmann, T. Suzuki and F. Wudl, J. Am. Chem. Soc., 1990, 112, 3302, and references therein; (d) A. J. Moore and M. R. Bryce, J. Chem. Soc., Chem. Commun., 1991, 1638. 19 (a) Review: M. Adam, K. Müllen, Adv. Mater., 1994, 6, 439; (b) M. Jørgensen, K. A. Lerstrup and K. Bechgaard, J. Org. Chem., 1991, 56, 5684; (c) M. R. Bryce, G. J. Marshallsay and A. J. Moore, J. Org. Chem., 1992, 57, 4859; A. Dolbecq, K. Boubekeur, P. Batail, E. Canadell, P. Auban-Senzier, C. Coulon, K. Lerstrup and K. Bechgaard, J. Mater. Chem., 1995, 5, 1707. 20 (a) M. R. Bryce, G. Cooke, A. S. Dhindsa, D. J. Ando and M. B. Hursthouse, Tetrahedron Lett., 1992, 33, 1783; (b) M. Fourmigué and Y.-S. Huang, Organometallics, 1993, 12, 797. 21 J. Y. Becker, J. Bernstein, A. Ellern, H. Gershtenman and V. Khodorkovsky, J. Mater. Chem., 1995, 5, 1557. 22 C. S. Wang, M. R. Bryce, A. S. Batsanov and J. A. K. Howard, Chem. Eur. J., in the press. 23 We note that molecular dynamics simulations of the interactions of segments of poly(g-benzyl glutamate) in solution suggest that the preferred conformation of macromolecules can be very solvent dependent: J. Helfrich, R. Hentschke, Macromolecules, 1995, 28, 3831. 24 For a review of hydrogen bonding in functionalised TTF derivatives: M. R. Bryce, J. Mater. Chem., 1995, 5, 1481. 25 G. M. Sheldrick, SHELXTL, Version 5/VMS, Siemens Analytical X-Ray Instruments Inc., Madison, WI, USA, 1995. 26 A. T. Rhys-Williams, S. A. Winfield and J. N. Miller, Analyst, 1983, 108, 1067. 27 Z. Xianfu, M. Jinshi and X. Huijan, SPIE J, 1991, 1616, 372. 28 M. S. C. Foley, PhD Thesis, Imperial College, London, 1994. 29 P. Wu, G. Saito, K. Imaeda, Z. Shi, T. Mori, T. Enoki and H. Inokuchi, Chem. Lett., 1986, 441. 30 G. E. F. Gracey and W. R. Jackson, J. Chem. Soc. B, 1969, 1207. 31 A. Fischer, G. N. Henderson, L. M. Iyer and C. J. Chag, Can. J. Chem., 1985, 63, 2390. Paper 7/01703K Received 11th March 1997 Accepted 15th May 1997

ISSN:1472-779X    

DOI:10.1039/a701703k

出版商:RSC    

年代:1997

数据来源: RSC