J . Phys. Chem. 1991, 95,7313-7319
7313
Isotopic Effects on the Decay Kinetics of the 385-nm Luminescence from Electron-Irradiated Ice C. F. Vernon: A. J. Matich, T. I. Quickenden,* Department of Chemistry, The University of Western Australia, Nedlands, Western Australia 6009. Australia
and D. F. Sangstert Department of Physical and Theoretical Chemistry, The University of Sydney, Sydney, NS W 2006, Australia (Received: February 19, 1991; In Final Form: April 29, 1991)
The effect of isotopic substitution of D for H on the decay kinetics of the 385-nm luminescence peak from electron-irradiated ice has been determined at 78 K. Each luminescence decay was analyzed into a long-lived, first-order component (tip = 690 f 34 ns for H 2 0and 1820 96 ns for D20) and a superimposed short-lived decay which obeyed second-order kinetics. The ratio of the first-order rate constants, k(H20)/k(D20),was 2.7 f 0.3, which agrees with the value of 2.7 for the ratio of the mobilities of protons and deuterons in their respective ices. This suggests that proton migration is the ratedetermining step in the luminescence mechanism. The suggested mechanism involves migration of H+to OH- formed from trapped OH and the subsequent formation of excited water molecules in the associative C'Bl state. The water molecules then decay to the dissociative AIBIstate, emitting broad-band excimer luminescence peaking at 385 nm.
Introduction
stitution on the spectral distribution of the ice luminescence.
It is now well established1-' that ultraviolet and visible region luminescences are emitted when low-temperature H 2 0 ice is exposed to ionizing radiation. Recent publications from this laboratory have shown that the emissions are not due to luminescent impurities' and that they occur in both amorphous and polycrystalline i d as well as in the more frequently studied crystalline ice! It has also been shown that the ice luminescences, with the exception of the observed6 Balmer a emission, are not affected by the extent of ice fragmentati~n.~ At some wavelengths the intensity of the luminescence increases with successive pulses of electrons and finally reaches a plateau level. Studies8 of such dose accumulation effects on the luminescence from crystalline ice have enabled separation of the luminescence into three bands. These bands have the following luminescence ields, G in photons/ 100 eV: 280-340-nm band, G, = 2 X lo-{ 320-6h-nm band, G = 1 X after 50 krad (but rising to a plateau value of 2 X Po-( after a total of 1.2-1 .5 Mrad); 50(M00-nm band, Gp= 1 X lV5. The 320-600-nm band dominates the emission and has a peak at 385 nm. The other two bands do not show the accumulation of intensity with e l d r o n dose exhibited by the 385-nm band. Separation of the luminescences into the above three bands can also be achieved by comparing their decay lifetimes? which are ca. 30 ns for the 280-340-nm emission, ca. 130 ns for the emission band around 385 nm, and ca. 30 ns for the emission between 500 and 600 nm. Kinetic analysis9 suggests that, after 300 ns, the prominent 385-nm luminescence decays according to second-order kinetics with a 2.6:l ratio of initial (Le., zero time) concentrations of the two reactants in a bimolecular rate-determining step. However, the luminescence emitted prior to 300 ns does not fit any simple kinetic law. It has also been shown9 that the 5006CIO-m luminescence decays according to tunneling kinetics, which suggests that the ratdetermining step for that emission involves geminate recombination. In order to more fully elucidate the reactions responsible for the 385-nm luminescence, the present study examines the effect of substituting D for H on the luminescence decay kinetics. The present study also provides an alternative and more satisfactory kinetic analysis of the 385-nm emission. This work follows on from a recently publishedt0 study of the effect of isotopic sub-
* To whom correspondence should be a d d r d .
Racnt address: CSIRO Division of Mineral Products, Curtin University Site. GPO Box U1987. Perth. W.A. 6001. Australia. *Honorary Fellow of the Australian Institute of Nuclear Science and Engineering.
0022-3654/91/2095-7313$02.50/0
Experimental Section The H20and D20used in this study were purified by a multiple oxidative distillation procedure described previously.'.' I Ice samples were prepared as previously describedI2 by freezing vacuum-degassed water in a 5 mm deep copper tray. Such ice comprises7 fragmented crystals of ca. millimeter dimensions. Ice samples held in the tray were cooled with liquid nitrogen and irradiated in an evacuated (
