The determination of heterojunction energy band discontinuities in the presence of interface states using capacitance-voltage techniques

Abstract

Effects of trapped interface charge on the determination of heterojunction energy band discontinuity energies using capacitance‐voltage (C‐V) techniques are analyzed both theoretically and experimentally. We show that for shallow traps, whose charge occupancy is unchanged by variations in the applied voltage, the measured conduction band discontinuity energy (ΔE_c) as determined by the depletion technique [Kroemer, Chien, Harris, and Edwall, Appl. Phys Lett. 3 6, 295 (1980)] is a function of trap density (σ_i). This error source in determining ΔE_c is large for small values of ΔE_c due to distortions of the conduction band induced by the trapped interface charge. In addition, the analysis used for determining σ_i and ΔE_c via the depletion technique has been generalized to correct for the effects of both deep and shallow traps. We show that at low measurement frequency, the measured value of ΔE_c is nearly independent of σ_i, and this value is near to the exact value of ΔE_c measured in the absence of interface traps (i.e., for σ_i=0). However, at high measurement frequency (or low temperature), the measured ΔE_c decreases with increasing σ_i or increasing trap energy depth from the conduction band minimum. These deviations from the actual value can be corrected by using the equations developed in this article. We apply these results to the understanding of In_0.53Ga_0.47As/InP heterojunctions. The computer‐simulated apparent free carrier concentration profiles are used to fit experimental data at several temperatures. The best fit value of ΔE_c (=0.22 eV) is in agreement with the value obtained via the emended equations. We explain the complex temperature dependence of these experimental profiles as due to donor traps near the heterojunction.