Influence of Nonlinearity on Electronic Transport Characterization of Ion-Implanted Silicon Wafers with Photocarrier Radiometry

Abstract

Two-layer linear and nonlinear models were employed to investigate the influence of the nonlinear dependence of photocarrier radiometry (PCR) signal on excitation power for ion-implanted silicon wafers on the determination of carrier transport parameters of ion-implanted layers in the frequency domain. Three silicon wafers were measured at 830 nm excitation wavelength, with the ion-implantation dose of 1011, 1013, and 1016 cm−2, respectively. The effect of nonlinear coefficient on the multi-parameter fitting for electronic transport parameter determination was analyzed via theoretical simulations and experimental measurements. Both simulation and experimental results showed that the multi-parameter fitted results obtained via the linear PCR model (ignoring the nonlinearity) were significantly different from that obtained via the nonlinear PCR model (taking into account the nonlinearity). With the linear PCR model, the fitting errors in the determination of transport parameters of ion-implanted layers were highly sensitive to the nonlinearity coefficient. For low-implantation-dose silicon wafers with high nonlinearity coefficients close to 2, the square variance increased drastically, and the fitting errors were non-acceptably high. Comparison of the electronic transport parameters determined via the two-layer linear and nonlinear PCR models clearly showed that two-layer nonlinear PCR model is much more suitable than the linear model for describing the PCR signals and is more accurately for characterizing the carrier transport parameters of ion-implanted silicon wafers.