Abstract

It has theoretically been shown that in the course of self-trapping exciton surmounts a potential barrier made by lattice distortion (or penetrates it by quantum-mechanical tunneling at low temperatures). Then, how high the potential barrier is has been investigated so far by analyzing the temperature dependence of the quantum yield η (or simply, the indensity) of free-exciton luminescence in many crystals. Three examples of it for RbI measured by Nishimura et al.¹ are shown in the figure of the next page, η decreases with increasing temperature T. This is regarded as originating from the increase of the self­trapping rate W of a free exciton with increasing temperature. Then, assuming that the temperature dependence of W was determined by a factor exp(-E_B/k_BT) describing the classical behavior of surmounting the potential barrier with height E_B, and moreover that other nonradiative and radiative decay processes had a rate independent of temperature, many people have analyzed η by a formula proportional to 1/[C+exp(−E_B/k_BT)] with a constant C. For example, this analysis gives E_B=17~18 meV for RbI.¹ However, the temperature region (15 ~ 30K in RbI) where η most decreases is considerably lower than the Debye temperature (~100K). In such low temperatures, lattice distortion making up the potential barrier cannot be regarded as classical, and the self-trapping process must be intermediate between the classical surmounting and the quantum-mechanical tunneling with rate almost independent of temperature. Therefore the E_B value evaluated so far must be reconsidered: It has been evaluated small.

© 1984 Optical Society of America