J. Phys. Chem. 1991, 95,6259-6264
However, over the past ten years there have been numerous reports of enantioselectivity in thermal and photoinduced, intermolecular electron-transfer rea~tions.'~J' The results presented in these reports were interpreted primarily in terms of differential electron-transfer distances in diastereomeric donoracceptor precursor complexes, and these differential distances were ascribed to enantio-differential steric effects (associated with donor-acceptor contact interactions). However, in a few cases "stereoselective (16) For examples, see: Warren, R. M. L.; Lappin. A. G.; Mehta, B. D.; Neumann, H. M. Inorg. Chem. 1990,29,4185-4189. Marusak, R. A.; Sharp, C.; Lappin. A. G. Inorg. Chem. 1990,29,2298-2302. Sakaki, S.; Nishijima, Y.; Koga, H.; Ohkubo, K. Inorg. Chem. 1989,28,4061-4063. Marusak, R. A,; Osvath, P.; Kempcr, M.; Lappin, A. G. Inorg. Chem. 1989,28,1542-1548. Lappin, A. G.;Martone, D. P.; Owath. P.; Marusak, R. A. Inorg. Chem. 1988, 27,1863-1 868. Gersclowitz, D. A.; Hammenhsi, A,; Taube, H. Inorg. chem. 1987.26, 1842-1845. Porter, G. B.; Sparks, R. H. J. Phorochem. 1980,13, 123-1 31. (17) Marusak, R. A.; Shields, T. P.; Lappin, A. G. In Elecfron Transfer in Biology ond rhe Solid Srare; Johnson, M. K., King, R. B., Kurtz, D. M., Jr.. Kutal, C., Norton, M. L., Scott, R. A., Eds.; Advances in Chemistry 226, American Chemical Society: Washington, DC, 1990; Chapter 13, pp 237-252.
6259
electronic effects" were also invoked to explain enantioselectivity in the electron-transfer rate data. Chiral discriminations in these stereoselective electronic effects would be analogous to those represented by the A&(exch) parameter in our enantioselective energy-transfer model, and they would reflect the relative chiralities of the donor and acceptor electronic wave functions. Comparisons of enantioselectivity in electron-transfer and energy-transfer processes between various types of chiral metal complexes should yield considerable insights regarding the structural determinants of intermolecular chiral discrimination. Many aspects of the enantioselective excited-state quenching model developed in this paper (see sections 3 and 4) may be applied to analyses of enantioselective electron-transfer reaction rate data. Acknowledgment. This work was supported by a grant from the U.S. National Science Foundation (CHE-8820180 to F.S.R.). We also gratefully acknowledge many helpful discussions with Dr.Seth W. Snyder (Argonne National Laboratory) and Dr. James P. Riehl (University of Missouri-St. Louis), both of whom have been important contributors to the theory and measurement of enantioselective quenching kinetics.
Absorption Spectra of Solvated Electrons and Radicals Observed in the Pulse Radiolysis of Cis and Trans Isomers of Methylcyclohexanols and Cyclopentanol to Cyclooctanoi Noboru Fujisaki, Pascal Comte, Pierre P. Infelta, and Tino Caumann* Institute of Physical Chemistry, Federal School of Technology, CH- 1015 Lausanne, Switzerland (Received: December 13, 1990; In Final Form: February 25, 1991)
The absorption spectra of solvated electrons have been determined in cis and trans isomers of 2-, 3-, and 4methylcyclohexano1, cyclopentanol through cyclooctanol, and some other alcohols by microsecond pulse radiolysis at room temperature. With the methylcyclohexanols the wavelength at the absorption maximum, Amax(e;), of solvated electrons for the cis isomer is different from that for the trans isomer by about 100 nm, irrespectiveof the position where a hydrogen atom on the cyclohexanol ring is substituted for a methyl group. It is inferred from the conformational analysis of the methylcyclohexanols that &(e;) for solvated electrons surrounded by the equatorial OH groups is observed at shorter wavelength than that surrounded by the axial OH groups. The optical spectra due to radicals have also been measured in all alcohols studied except tert-butyl alcohol.
Introduction The solvated electron is one of the characteristic reactive intermediates in the radiolysis of polar liquids. The absorption spectra of solvated electrons in polar solvents have been measured by many researchers.' Some spectral data of the solvated electron have been subjected to theoretical interpretations.2 Our un(1) For example: (a) Sauer, Jr., M. C.; Arai, S.; Dorfman, L. M. J . Chem. Phys. 1965, 42, 708. (b) Arai, S.;Sauer, Jr., M. C. J . Chem. Phys. 1966, 44, 2297. (c) Michael, B. D.; Hart, E. J.; Schmidt, K. H. J . Phys. Chem. 1971,75,2798. (d) Jha, K. N.; Bolton, G.L.; Freeman, G. R. J . Phys. Chem. 1972, 76,3816. (e) Leu, A.-D.; Jha, K. N.; Freeman, G. R. Can. J . Chem. 1982,60,2342. (f)Jou, F.-Y.; Freeman, G. R. J . Phys. Chem. 1977,81,909. (e) Hentz, R. R.; Kenney-Wallace,G. A. J. Phys. Chem. 1972, 76,2931. (h) Hentz, R. R.; Kenney-Wallace, G. A. J . Phys. Chem. 1974, 78, 514. (i) Kenney-Wallace, G.A.; Jonah, C. D. J . Phys. Chem. 1982, 86, 2572. ti) Baxendale, J. H.; Wardman, P. J . Chem. Soc., Faraday Trans. I 1973,69, 584. (k) Fujisaki, N.; Comte, P.; Infelta, P. P.; Giumann, T. Helo. Chim. Acta 1990,73,574. (I) Johnson, D. W.; Salmon, G.A. Can. J . Chem. 1977, 55, 2030. (m) Matheson, M. S.; Dorfman, L. M . Pulse Radiolysis; MIT Press: Cambridge, MA, 1969; p 168. (2) For example: (a) Copeland, D. A,; Kcatner, N . R.; Jortner, J. J. Chem. Phys. 1970,53, 1189. (b) Fueki, K.; Feng, D.-F.; Kevan, L. J . Am. Chem. Soc. 1973, 95, 1398. (c) Freeman, G. R. J . Phys. Chem. 1973, 77, 7. (d) Clark, T.; Illing, G. J . Am. Chem. Soc. 1987, 109, 1013. (e) Wallqvist, A,; Martyna, G.;Berne, B. J. J . Phys. Chem. 1988, 92, 1721.
0022-36S4/91/2095-6259$02.50/0
derstanding of the nature of the solvated electron has, ,been deepened by the study of spectral evolution with time"J and spectral shift due to change in temperaturelcvd and pressure." There are a few publications'ca.h where the spectral changes caused by varying molecular structure of the alcohol have been investigated. In particular, a systematic measurement of the spectra of solvated electrons in a variety of alcohols has been made by Hentz and Kenney-Wallace.l@ They have shown that the value of Amax(e;) does not change appreciably on going from n-butanol to n-undecanol in spite of the fact that the static dielectric constants for these alcohols are remarkably different. It has also been noted'gvh that AmaX(e;) for branched alcohols is sensitive to the number and size of branches and distance of the branch point from the O H group. In the present work the spectra of solvated electrons have been measured in cis and trans isomers of methylcyclohexanol (hereafter abbreviated as MCH) and in some other alcohols. Since detailed conformational analysis of the MCH has been carried out by I3C NMR spectroscopy,g the spectral data of solvated electrons can be related to the conformations of MCH. Attention is also paid (3) pbbotin, 0.A.; Sergeyev, N . M.; Chlopkov, V. N.; Nikishova, N . 0.; Bundel, Y . G. Org. Magn. Reson. 1980, 13, 259.
0 1991 American Chemical Society
6260 The Journal of Physical Chemistry, Vol. 95, No. 16, 1991
Fujisaki et al.
to measurement of the spectrum of radical species and to correlation of it with final products formed by radical recombination reactions. Experimental Section Materials. Ethyl alcohol, tert-butyl alcohol, cyclopentanol, cyclohexanol, and cycloheptanol were obtained from Fluka. Cyclopropylmethanol, 1-MCH, cis- and trans-2- and -4-MCH, and cyclooctanol were purchased from Aldrich. cis- and trans3-MCH were furnished by ICN Biomedicals. Since in most cases no spectrum of the solvated electron can be observed in the pulse radiolysis of nonpurified alcohols, all the alcohols were rigorously purified with a temperature-programmed preparative gas chromatograph ( b e , Model 104). The following columns were used: 2- and 5-m columns filled with packing of 20 wt % Apiezon L on Kieselgur and a 5-m column packed with 20 wt % Igepal CO 880 on Kieselgur. Depending on the purity of the purchased alcohols, different purification procedures were adopted. Ethyl and tert-butyl alcohols were simply purified by introducing a small quantity (-0.15 mL) of the alcohol onto the 5-m Igepal column. Cyclooctanol was similarly purified through the use of a 5-m Apiezon L column. Cyclopropylmethanol, cyclopentanol, cyclohexanol, I-MCH, and cis- and trans-2-, -3-, and -4-MCH were first purified through the 5-m Igepal and subsequently through the 5-m Apiezon L column. Cycloheptanol was first purified on the 2 m Apiezon L and then on the 5-m Igepal and finally on the 5-m Apiezon L column. The cis isomer of MCH separated badly from the trans isomer by our preparative GC. The purified cis-2-, -3-, and -4-MCH contained 0.1, 1.0, and 0.1 mol % corresponding trans isomer, respectively. Similarly, trans-2-, -3-, and -4-MCH contained, respectively, 1A0.4, and 0.2 mol % cis isomer after purification. In the purification of high-boiling-point alcohols such as cyclohexanol, thermal decomposition was found to occur in the preparative G C column to give traces (CO.Ol%) of ketonic Compounds. To eliminate the reducing impurities, all the alcohols purified by GC were refluxed with NaBH4 (Fluka, purum) at 70 OC for 24 h under argon. The refluxed alcohols were dried over freshly activated granular CaO for 1 week under vacuum. Water contents in the purified alcohols were determined before and after pulse radiolysis to be C0.03 wt % by an analytical GC equipped with a 50-cm Porapak Q column. The water contents were also occasionally determined with a Karl Fischer titration apparatus. The purified alcohols, especially secondary alcohols, undergo autooxidation to give nonidentified impurities, probably peroxides or ketones, on standing at room temperature. Hence the purified alcohols were stored in a refrigerator at -30 OC. The absorbance of the purified alcohols was determined to be CO.01 in the UV region down to 250 nm using a Beckman spectrophotometer (Model, DK 2-A) and a quartz cell of I-cm optical path. At 230 and 240 nm the absorbances were 0.04 i 0.01 and 0.02 f 0.005, respectively, irrespective of the kind of alcohol. Water used for dosimetry was obtained from Millipore Milli-Q units. The following were used without purification: KSCN (Merck, pro analysi), Ar (L'Air Liquide, 99.9997%), O2 (Carbagas, 99.9%), N 2 0 (Carbagas, 99.99%), and chemicals used to determine the yield of products formed in the steady-state radiolysis of tert-butyl alcohol. Procedures. The pulse radiolysis technique has been described in our previous work," so only the salient features of the approach will be given here. Alcohol (3 mL) was put in a quartz cell (1 X I X 5 cm) and was deaerated by bubbling with argon for 2 h before irradiation with 0.5-ps pulses of 3-MeV electrons from a Van de Graaff accelerator. Care was exercised not to take up adventitious water during the deaeration. The current intensity of the pulse was 250 mA. The irradiation temperature was kept at 25 f 2 OC. For recording the transient absorption, a pulsed 450-W xenon arc lamp and a Bausch & Lomb monochromator were used. The monochromator was coupled with either an RCA 4840 photomultiplier or an RCA C 30808 silicon photodiode. The signals from the photodetector were recorded on a Tektronix 2430 digital oscilloscope.
5 4
3 2 N
0
r
x ' 0)
: 5 (D
.Et
z4
P
a
3 2 1
200
400
600
800
1000
X(nm) Figure 1. Absorption spectra of the transient species measured immediately after the electron pulse: (a) ethyl alcohol ( 0 )and tert-butyl alcohol (0);(b) cyclopropylmethanol(0) and I-methylcyclohexanol (0). Pulse duration, 0.50 1 s ; dose per pulse, 34 Gy; irradiation temperature, 25 & 2 OC.
Dosimetry was carried out by using Orsaturated 5 mM KSCN solution. The dose per pulse was calculated to be 34 Gy taking Gt((SCN),-) = 2.1 X lo4 ( M cm 100 eV)-' at 478 nm." Product Analysis. In the steady-state radiolysis of tert-butyl alcohol, 0.5 mL of the purified alcohol was irradiated after degassing under vacuum with a y-ray source to a total dose of 50 and 100 kGy. The gas chromatographic analysis was performed using a 5-m Column packed with 20 wt % Igepal C O 880 on Kieselgur. The column was held at room temperature for 200 min and then heated a t 1.5 "C/min to 145 OC and held at that temperature. Measurements of Viscosity. The viscosity of the alcohols was measured by using two different Ostwald viscosimeters with a volume of -2 mL. The viscometer was immersed in a stirred water bath thermostated at 25 f 0.01 OC. The density of the alcohols necessary to convert the kinematic to shear viscosity was obtained by using a 2-mL pycnometer. The shear viscosity 1 was found by using 7 = kpt, where p is the density, t the flow time in seconds, and k a calibration constant that we determined by calibrating with water. Some alcohols such as cyclooctanol are thixotropic fluids, and their viscosities depend on the experimental conditions. The absorption spectra of solvated electrons have already been measured in ethanolla*b+*h and terf-butyl alcohol.le*h,mHowever, for e; in ethanol ranges from 68 1le to the reported value of A, at 729Ih nm at room temperature. Likewise, the value of ,A, room temperature in rert-butyl alcohol spreads from 850" to 12531hnm. We have therefore redetermined the absorption bands in ethyl and tert-butyl alcohols with the results shown in Figure la. The absorbances were corrected for day-to-day variation in pulse intensity, but they were not corrected for decay during the pulse. Our A,, values in both ethanol (684 nm) and tert-butyl alcohol (1084 nm) listed in Table I were found to agree well with those determined by Freeman et al.Ic The A, value for e; in tert-butyl alcohol lies close to the detection limit of our equipment (1 100 nm). So the A, value in tert-butyl alcohol may be subjected to larger error than we expect. Figure 1b shows that the spectrum of e;in cyclopropylmethanol is similar to the spectrum in methanol." The absorption band
The Journal of Physical Chemisrry, Vol. 95, No.16, 1991 6261
Solvated Electrons and Radicals
TABLE I: Spectral Characteristics of the Solvated Electron Obsewed on Pulse Radiolysis’ and Some Physicochemical Properties of the Alcohols
alcohol ethyl alcohol tert-butyl alcohol cyclopropylmethanol 1-methylcyclohexanol cis-2-methylcyclohexanol trans-2-methylcyclohexanol cis-3-methylcyclohexanol trans-3-methylcyclohexanol cis-4-methylcyclohexanol trans-4-methylcyclohexanol
L ( e ; ) ,nm 684 1084 632 > I 100 871 768 723 818 837 72 1 784 77 1 763 839
G(e;) WIprb eV
104(100 eV M
1.70
1.26
1.67
1.43 1.07
1.53 1.68
1.06
1.69
1.38 1.37
1.41 1.53 1.42 1.62 1.64
cm)-’
cp 1.08 4.44 3.50 31.3 18.8 36.3 24.8 21.5 26.5 39.2
rl,c
DP 24.6‘ 15.8e
16.9 8.18
1.57 1.33 1.45 9.62 1 9 cyclopentanol cyclohexanol 1.65 1.58 58.3 15 h cycloheptanol 1.71 1.36 51.5 13‘ 1.52 1.59 143.8 121 cyclooctanol ohlse length = 0.50 ps. Dose per pulse = 34 Gy. Irradiation temperature = 25 k 2 O C . bWidth of the absorption band at half-height. ‘Dynamic viscosity in centipoise at 25 O C determined by us. “Static dielectric constant at 25 OC unless otherwise stated. #Reference2c. /Reference 4a; D,at 20 O C . #Reference 4b; 0, at 20 O C . hReference Ih. ‘Reference 5. ]Reference 6. of e; in I-MCH seems to have a peak beyond the long-wavelength limit ( I 100 nm) of photodetector as shown in Figure Ib. In Figure 1 the absorption spectra assigned to radical specieslb are shown in the wavelength region 230-400 nm. rert-Butyl alcohol has no clear absorption band in this region. The radical species are characterized by their slow decay ( r t l z= a few tens of microseconds) compared with rapid decay ( r l 1 2 a few microseconds) of solvated electrons. Parts a-c of Figure 2 show the absorption spectra of solvated electrons and radicals in cis- and rrans-2, -3-, and -4-MCH, for e; in the cis isomer is distinctly different respectively. A, from that in the trans isomer regardless of the location of the methyl group on the cyclohexanol ring. In the case of 2- and 4-MCH, A,,(e;) is longer for cis isomer than for trans isomer. Contrary to this, Xmx(e;) is longer for the trans isomer than for the cis isomer in the case of 3-MCH. The effect of an electron scavenger, N 2 0 , on the transient spectrum was studied and found to be essentially the same for 1-MCH, rrans-2-MCH, cycloheptanol, and cyclooctanol. Only a typical result for rrans-2-MCH-N20 system is displayed in Figure 2a for the sake of simplicity. In the presence of N 2 0 , the absorption band due to e; vanished, but the band due to radicals was scarcely influenced. The spectra of radicals are in contrast to the spectra of e;, insensitive to the variation in the molecular structure of M C H as demonstrated in Figure 2. Amx(e;) barely changes in passing from cyclopentaol to cycloheptanol as presented in Figure 3. Figure 3b shows, however, a shift in &,,(e;) toward longer wavelength occurs on going from cycloheptanol to cyclooctanol. The spectral characteristics of the solvated electrons are summarized in Table 1. The viscosities of the alcohols determined by us are also included in Table I together with the static dielectric constants available.1hJc.c6 Table I1 presents the G values for end products formed in the y-radiolysis of rerr-butyl alcohol. Our G values for acetone and rert-butyl methyl ether agree well with those determined by Verdine7 However, substantial discrepancies were observed between our and Verdin’s7 G values for isobutyraldehyde and 2butanol. It is noted that the yields of acetone and 2-propanol are sensitive to the total dose.
Discussion On the Value of L x ( e , - ) . An interesting point in the present study is that in the case of MCH the &(e,-) values for cis isomers are distinctly different from the values for trans isomers, regardless (4) Riddick, J. A.; Bunger, W. B.; Sakano, T. K. Organic Solornts, 4th ed.;Wiley: New York, 1986; (a) 230, (b) p 231. (5) Poser, U.;Schulte, L.; WLLrlinger. A. Ber. Bunsen-Ges. Phys. Chem. 1985. 89. ~
1275.
( 6 *):&’F U.;WOrflinger, A. Be?. Bunsen-Ges. Phys. Chem. 1!%38,92,765. (7) Verdin, D. Int. J . Rodlat. Phys. Chem. 1910, 2, 201.
TABLE II: C Values of the Products Formed in the Radiolysis of Neat tert-Butyl Alcohol G Droduct ref 7’ this workb acetone 2.55 2.80 0.63
