No abstract

A sensitive photoinduced current transient spectroscopy technique using a computer for data acquisition and processing has been worked out. It enables one to determine separately the signatures and concentrations of deep levels in high-resistivity bulk materials. In a previous paper, possible signal processing has been analyzed from a rather theoretical point of view. In the present report, the instrumentation used as well as the experimental procedure and some practical and complementary aspects will be discussed. For the determination of reliable trap signatures, we have proposed and tested a four-gate data treatment that eliminates, or at least reduces, many of the problems encountered when using the usual double-gate procedure provided some experimental precautions are taken. Yet the double-gate method is useful when one wants to estimate the trap concentrations. A standard experimental procedure is presented that takes into account the temperature dependence of the mobility/free-lifetime product of thermal emitted carriers, and its possible variation during the relaxation of the traps.

The use of photo-induced current transients as a means for the detection of deep trapping levels in high-resistivity bulk materials and for the determination of their parameters is discussed. In this perspective some experimental and theoretical problems are discussed on the basis of various models. The conditions which must be satisfied by the kinetics or experimentally in order to observe a quasi-exponential time dependence for the decay of the current corresponding to a single trap are pointed out. Special attention is devoted to various possibilities, depending on the model and on experimental conditions, for the parameters related to the trap or to other centres to be involved in exponential decays and how this can affect the evolution of the transient as a function of temperature. The conclusions of this analysis are of practical importance for the implementation of various possible multigate processing methods of computer recorded transient data, which are evaluated in Part II.

For pt.I see ibid., vol.19, p.57 (1986). The aim of the paper is to describe the principles and the features of several signal processing methods suitable for exponential photocurrent decays and to evaluate each of them for the determination of the temperature dependence of the associated relaxation time and the involved trap parameters under various conditions depending both on the physical model and on the experimental procedure. Double-gate treatments lead to reliable results only in well defined circumstances. The proposed so-called four-gate technique, on the contrary, has a much more extended field of application; thus particular importance is attached to this method. The conditions of use, among others the appropriate choice of the four delay times on the transient, and some specific features have been determined by computer simulation. Computer simulation has also been used in order to evaluate the potential performances of the method in the case of several multi-level configurations. The main conclusion is that it offers increased spectral resolution in comparison with the double-gate processing.