Effect of stationary ionization of traps near the midgap on the spectrum of thermally stimulated capacitance of semicond
- PDF / 215,969 Bytes
- 4 Pages / 612 x 792 pts (letter) Page_size
- 32 Downloads / 159 Views
CS OF SEMICONDUCTOR DEVICES
Effect of Stationary Ionization of Traps Near the Midgap on the Spectrum of Thermally Stimulated Capacitance of Semiconductor Devices S. V. Bulyarskiœ, A. V. Zhukov^, O. S. Svetukhina, and O. A. Trifonov Ul’anovsk State University, Ul’yanovsk, 432970 Russia ^e-mail: [email protected] Submitted October 25, 2005; accepted for publication January 20, 2006
Abstract—An algorithm for calculating the parameters of deep-level centers in the space-charge region of semiconductor devices on the basis of an analysis of the curves of thermally stimulated capacitance in the presence of recombination centers with the levels located near the midgap is suggested. The method is tested using the commercial AL-107 GaAs-based light-emitting diodes. The results of calculations of the deep-center parameters in the materials under study are consistent with the results of other methods of analysis. PACS numbers: 85.60.Jb; 72.20.Jv; 78.30.Fi DOI: 10.1134/S1063782606090211
1. INTRODUCTION The method of thermally stimulated capacitance (TSC) is widely used for determination of the parameters of deep-level centers in the space-charge region (SCR) of semiconductor devices [1–5]. However, the algorithm of the method is not always applicable to the analysis of specific features of the electronic spectrum related to the deep-level centers in the case where there are recombination levels near the midgap. In this study, we suggest an algorithm for calculating the parameters of the capture and recombination centers in the SCR by separating the nonstationary component from the curve of thermally stimulated variation in the SCR capacitance of semiconductor devices. We chose the commercial AL-107 GaAs light-emitting diodes as the objects for testing. We measured the TSC spectra of the samples in the temperature range from 160 to 360 K at various heating rates in the range from 0.05 to 0.4 K/s. According to the estimates [5], the above temperature range is characteristic of an analysis of deep levels with energies in the range from 0.15 to 0.80 eV. The measurements were carried out using an automated setup that included a resonance capacitance meter based on a Ch4-34 frequency meter. The results obtained were stable and quite reproducible. In Fig. 1 (curve 1), we show the characteristic dependence of the derivative of the TSC on temperature (in the case of heating); this dependence was measured for one of the samples from the AL-107 group. It is shown experimentally that the shape of curve 1 is independent of both the initial occupancy of traps in the SCR and the heating rate in the range from 0.05 to 0.2 K/s (if the heating rate exceeds 0.2 K/s, the
shape of the curve distorts, which may be caused by a number of factors, including nonuniform heating of the samples). We based our preliminary analysis of Fig. 1 on separation of the experimental TSC curve into the components of the thermally stimulated processes as (see [3, 4])
1105
E 1 1 T 2 ∂C ------- = A ⎛ ------⎞ exp -----t ⎛ ------ – ---⎞ ⎝ ⎠ k ⎝ T m T⎠ T ∂T m (1)
E 1 1
Data Loading...
