CHARACTERIZATION OF PROCESSES CONTRIBUTING TO THE COULOSTATIC-FLASH RESPONSE OF SEMICONDUCTOR - ELECTROLYTE INTERFACES

STEPHEN JOSEPH PARUS, Purdue University

Abstract

The specific characteristics of the semiconductor-solution interface contributing to its coulostatic-flash (transient photopotential) response have been determined. Such interfaces have possible application as a means of converting solar energy into electrical and/or chemical energy. The coulostatic-flash technique is capable of probing photoelectrochemical processes on a very rapid time scale (tens of nanoseconds). The investigation described here was undertaken in order to establish the factors responsible for the experimentally observed signals at semiconductor-solution interfaces. In particular, it is of interest to determine the effect of charge transfer. The ability to study charge transfer at these devices at such a fast rate would provide information useful for the design of more efficient cells. Background material concerning semiconductors, their interfaces, and photoelectrochemical techniques is presented prior to the results. Experimental details are then given, including design of a battery-powered coulostat used to measure low noise photopotentials. The coulostatic response in the absence of charge transfer was examined with passivated p-Si, and with p-type electrodes near flatband. Electron-hole separation and subsequent recombination results in potential changes toward flatband followed by a decay of hundreds microseconds back to the initial potential. As studied at p-GaAs at negative potentials, charge transfer yields a net potential change following the initial excursion. The extent of the initial potential overshoot, due to a transient non-equilibrium carrier distribution, is dependent on light intensity. The lack of coulostatically observable photoreduction of added redox couples is attributed to slow conduction band to surface state electron transfer or to slow charge transfer to solution from surface states. Another mechanistic possibility is that charge transfer proceeds via an initial rapid photoreduction of the electrode surface, followed by slow charge transfer to a solution species, which is not detectable on the short coulostatic time scale. This latter explanation appears most probable, as it is consistent with the evidence of rapid charge transfer obtained with coulostatic-flash experiments for blank aqueous or non-aqueous electrolytes.

Degree

Ph.D.

Subject Area

Analytical chemistry

COinS