T. Higashimura, A. Namiki, M. Noda, and H. Hase
87 44
cf. experiment 6, 1s 3 times greater than that in experiment 1. Such results, and the results of experiments 9 and 12, indicate that the benzene yield from silica-alumina irradiated alone, like the yield of trapped positive holes, is limited by the number of preexisting electron traps. The approximate equality of plateau yields in experiments 6, 9 and 12 suggests that (1)0 2 , COz, and NzO trap mobile electrons only at certain surface sites (that are limited in number) and remain bound to, thereby blocking, the site or (2) the number o f preexisting positive-hole traps becomes the limiting factor with the electron scavengers present during irradiation. It is interesting that, as shown in experiments 8 and 10, the presence of electron scavengers during irradiation enhances the yield of only those dealkylation centers that are removed by exposure to Hz. Experiment 7 shows that electrons trapped by irradiation in the presence of 0 2 are not freed by a second irradiation after pumping 0 2 from the reaction cell. On exposure Qf irradiated silica-alumina to co or S&, height of the positive-hole esr signal decreases to -85% of the original vaiue and there is a slight bleaching of the visible co1oration;G as shown in experiments 21 and 23,
there is a corresponding small reduction in benzene yield. Because of the absence of visible coloration on irradiation of silica-alumina in the presence of CO or SO2 and the appearance of a very intense new esr signal attributed to COZ- or SO3-, respectively, it was suggested that CO and SO2 are very effective traps for the mobile positive hole (unstabilized 0 - ) . 6 Thus, equality of the benzene yields in experiments 1, 20, and 22 suggests that the positive holes trapped by CO and SO2 also are effective in isopropylbenzene dealkylation (the total yield of trapped positive holes and, therefore, of benzene being limited by the number of preexisting electron traps). Results of the present study, then, provide strong support for the ideas developed in previous work.Z+ The yield of benzene from contact of isopropylbenzene with irradiated silica-alumina, like the yield of trapped positive holes, does indeed appear to be limited by the number of preexisting electron traps. The total results seem to require that electron transfer from isopropylbenzene to a trapped positive hole (of any kind contributing to the esr spectrum) be the primary process in isopropylbenzene dealkylation on ?-irradiated silica-alumina.
Electronic Spectra of Trapped Electrons in Organic Glasses at 4°K. I 11. Effect of an Electron Scavenger in Ethanol Takenobu Higashimura,* A k h Namiki, Masato Noda, and Hirotomo Hase Research Reactor Institute, Kyoto University, Kumatori-cho, Sennan-gun, Osaka, Japan (Received April 18, 4972) Publication costs assisted by the Research Reactor Institute, Kyoto University
Optical absorption measurements were carried out on et - and esol- in ethanol glasses containing solute BzCl a t 4 and 77°K. The electron scavenging efficiency for the y radiolysis at 4°K is about 4 times larger than for the radiolysis at 77°K and depends slightly on the wavelength of the absorption band. The temperature dependence of the efficiency is tentatively attributed to that of the scavenging cross section obeying the l/u law and/or to that of trapping cross section affected by lattice phonon interaction.
Introduction A t such a very low temperature as 4°K or in a very early stage a t low temperature in organic glasses, electrons are stabilized in shallow traps where surrounding molecular dipoles are not relaxed Such trapped electrons in unrelaxed traps are confirmed by their optical absorption spectra4-10 and esr spectral-3.6 which are much different from those of ordinary solvated electrons observed at 77°K. In glasses containing a small amount of electron scavenger, competitive reactions of electrons occur between the scavenger molecules, positive holes, and traps. Knowledge of the scavenging efficiency in these competeLive processes affords information on the nature of electron traps because the efficiency will depend on both trapping and scavenging cross sections. The efficiency of electron scavenging by biphenyl in 2methyltetrahydrofuran glass was found to be about four Tho Journal of Physical Chemistry, Vol. 76, No. 25, 7972
times as large for irradiation at 4°K as that a t 77"K.ll The result implies that the efficiency of electron scavenging at 4°K is larger than at 77°K. However there remains (1) D.R. Smith and J. J. Pieroni, Can. J. Chem., 45,2723 (1967). (2) H. Yoshidaand T. Higashimura, Can. J. Chem., 48, 504 (1970). (3) T. Higashimura, M. Noda, T. Warashina, and H. Yoshida, J. Chem. Phys., 53, 1152 (1970). (4) H. Hase. M. Noda, and T. Higashimura, J. Chem, Phys., 54, 2975 (1971). (5) H. Hase, M. Noda, T. Higashimura, and K. Fueki, J. Chem. Phys., 55, 5411 (1971). (6) H. Hase, T. Warashina. M. Noda, A. Namiki, and T. Hiaashimura. J. Chem. Phys., 57, 1039 (1972). (7) J. T. Richards and J. K. Thomas, J. Chem. Phys., 53, 218 (1970). (8)J. H. Baxendaie and P Wardman, Nature (London), 230, 449 119711
(9) i.Ke;an, Chem. Phys. Lett., 11, 140 (1971). (10) L. Kevan, J. Chem. Phys., 56, 838 (1972). (11) H. Yoshida, M. Ogasawga, T. Warashina, and T. Higashimura, J. Chem. Phys., 56,4238 (1972).
Spectra of Trapped Electrons in Organic Glasses
,,or;
WAVELENGTH (nrn)
WAVELENGTH 1000 700
3745
(
nm
1
500
400
__
0.5
k-
IO
2'0
Figure 1 . Optical absorption spectrum of ethanol glass containing 0.01 mol YO BzCI. The y irradiation and measurement were carried out at 4°K. T h e total dose was 0.23 Mrad.
another possibility that the higher efficiency may be caused from additional capture in the course of warming the sample from 4 to 77"K, because concentrations of trapped electrons and anions were determined by measurements a t 77°K. In this investigation, effects of solute benzyl chloride (BzC1) on the trapped electron band in ethanol glass at 4 and 77°K are reported and efficiency of electron scavenging is compared for irradiation and measurement at 4°K with those a t 77°K. The temperature dependence of the efficiency is discussed according to the temperature dependence of scavenging and trapping cross sections.
-.._ ---.-._. ---- -.-.. ---.~T.y--~ ............................ ......................
--+ -_ ................. -
-._._,
b-A
25 30 3 5 x l d WAVE NUMBER (crn-'1 15
5
10 15 20 WAVE NUMBER (crn-1)
25x16
Optical absorption spectra of ethanol glass containing 0.01 (- - - -), 0.03 ( - . - .) and 0.1 mol % (. * * ) BzCI, together with t h e spectrum for pure ethanol (-----). T h e y irradiation and measurements were carried out at 4°K. Total y dose was 0.23 Mrad for each sample.
Figure 2.
em----+
-.
-
neutralization
(2)
Experimental Section Reagent grade ethanol was used without further purification. Reagent grade BzC1 was purified by ordinary fractional distillation methods. Disk samples of 0.2 cm thickness were made in liquid nitrogen and were then immersed and fixed in liquid h e l i ~ m Both . ~ irradiation by y rays and optical absorption measurements a t 4°K were performed with the same optical dewar as described before.* The total y (dose was 0.23 Mrad for the irradiation a t 4°K and 0.34 Mrad a t 77°K.
When ethanol glass containing a small amount of BzCl is irradiated and measured a t 4"K, the observed optical absorption band extending from 2000 nm to 300 nm consists oi two well separated bands as shown in Figure 1. The broad band extending from the near-infrared to the visible region is due to electrons in unrelaxed traps, and the sharp absorption band around 310 nm is for benzyl radicals.12 As the concentration of BzCl increases, the band of benzyl radicals increases in intensity. Concomitantly, the electron band decreases in intensity over the whole region as shown in Figure 2. It is noted that the infrared part of the band which is responsible for electrons in shallow, unrelaxed. traps is quenched more efficiently than the visible part responsible for electrons in deep, unrelaxed traps The decrease of the trapped electron band and consequently the increase of the radical bands in intensity are much more sensitive to a change in concentration of BzCl for radiolysis at 4°K than for radiolysis a t 77°K. The electron band is almost eliminated a t a concentration of 0.10 mol % RzCl a t 4"K, whereas the band decreases in intensity to haif of that in pure ethanol glass a t the same concentration of BzCl at 77°K. Because the detrapping process which can compete with the trap relaxation is not important,6 the behavior of mobile electrons, until they are stabilized or disappear, in
The efficiency of electron scavenging, a , is defined as the initial slope of the scavenging curve of [er-] against [SI, and is expressed as
Thus it follows from eq 4 and 5 that
Because the yield of trapped electrons in pure ethanol glass is independent of the temperature of radiolysis,6 the ratio of the efficiency of electron scavenging at 4°K to that at 77°K is given by a(4"K) ____ a(77"K)
-. '
hs(4"K) hi(77"K) Ks(77"K) kt(4"K)
~ 4 4 ° K ) at(7'I"K) g 4 7 T K ) ~r(4010
~~~
where a, and crt are the cross sections of electron scavenging by BzCl molecules and of electron tramping by preformed traps, respectively. For practical evaluation of the value a from the actual scavenging curves, it is better to use the relation
where
[SI112 denotes
the concentration of scavenger mole-
(12) F. S. Dainton, G. A. Salmon, and J. Teply, Proc. Roy. SOC. London. 286, 27 (1965).
The Journal of Physical Chemistry, Vol. 76, No. 25, 1972
3746
. Noda, and H. Hase
T. Higashimura, A. Namiki,
TABLE I:
Efficiency of Electron Scavenging at 4 and 77'K
WAVELENGTH (nm) 340
Temp. of irradiation and measurement 4°K
Wavelength. nm 1500 1250
1000 800 700 540
400 77°K
650 540
45s
Efficiency of electron scavenging, a,
(mol % ) - l 5.3 5.3 4.8 4.2 3.6 2.9 3.3
320
300
I
X 10
f 0.5 f 0.5 f 0.5
f 0.4 f 0.4 f 0.3 f 0.3
0.77 f 0.05 1.OO f 0.05 0.83 f 0.04
cules when G(et-) is half of Go(et-). The values of a thus obtained at the different wavelengths of the absorption band a t 4 and 77'K are listed in Table I. It follows from Table I that a(4"K) depends slightly on the wavelength of the band, while a(77"K) is independent of the wavelength of the band. Thus the value of a(4"K)/a(77"K) lies between 3 and 5 . If the interaction between mobile electrons and scavenger molecules is of induced dipole type, the interaction would be inversely proportional to the distance to the power of 4 and the cross section for electron scavenging would be inversely proportional to the velocity of electrons. If this cross section is averaged over a MaxwellBoltzmann distribution of the velocity of thermalized electrons, the average cross section becomes inversely proportional to the square root of the temperature of the system.l3 Therefore, if the trapping cross section is independent of the velocity of the electrons, the efficiency of electron scavenging would become inversely proportional to the square root of the temperature of the system, and thus the value of a(4"K)/a(77"K) would be about 4.5, agreeing approximately with the experimental result. It should be noted that the previous result for 2-methyltetrahydrofuran gives this ratio as 4.11 The unrelaxed traps might be surrounded by molecular dipoles orienting favorably to form shallow potential wells. When mobile electrons come close t o these traps, they will he interfered with by potential barriers around the traps. Therefore, electrons must penetrate the barriers in order to be trapped. Transmission of this process would depend on the kinetic energy of the electrons, and therefore mobile electrons a t 4°K may have less transmission than that at 77°K.
The Journal of Physical Cnemistry, Vol. 76, No. 25, 7972
0.8
-_ ... WAVE NUMBER c m - 1 )
*
Figure 3. Optical absorption spectra of 0.10 mol % BzCl in ethanol glass irradiated at 4°K. Total y dose was 0.11 Mrad. Solid line spectrum was obtained after trapped electrons were completely photobleached at 4°K. Broken line spectrum was obtained after subsequently warming t h e glass rapidly to 77°K. All absorption measurements were carried out at 4°K.
It is probable that the trapping cross section depends on the interaction of electrons with lattice phonons and that the matrix element of the interaction is smaller at the reduced temperature. This may result in a smaller cross section for trapping a t 4°K than a t 77"K, thereby contributing to the temperature dependence of the efficiency of electron scavenging. Finally, it is noted that the spectrum of the radicals obtained for irradiation and measurement a t 4°K is not identical with that for the radiolysis a t 77"K, as shown in Figure 3. The spectrum a t 4°K has a peak a t 340 nm in addition to the sharp peaks reported in the radiolysis study a t 77"K.12 The new peak disappears a t about 55°K in the course of warming the glass from 4"K, being accompanied by growth of the sharp peaks. At 77"K, intensity of the 318-nm peak grows to about two times the initial intensity. Cooling the glass down again to 4°K causes no further changes in the spectrum. Such a n irreversible change suggests that a precursor of the benzyl radical is responsible for this new peak.
Acknowledgment. The authors wish to express their sincere thanks to Professors T. Watanabe and K. Fueki for their kind discussions and suggestions on the electron trapping mechanism. (13) J. M. Wardman, and M. C. Sauer, Jr., J. Radiaf. Phys. Chem., 3,
273 (1971).
