We propose a new technique called local deep level transient spectroscopy (local-DLTS), which utilizes scanning nonlinear dielectric microscopy to analyze oxide/semiconductor interface traps, and validate the method by investigating thermally oxidized silicon carbide wafers. Measurements of C-t curves demonstrate the capability of distinguishing sample-to-sample differences in the trap density. Furthermore, the DC bias dependence of the time constant and the local-DLTS signal intensity are investigated, and the results agree to characteristic of interface traps. In addition, the Dit values for the examined samples are estimated from the local-DLTS signals and compared with results obtained using the conventional high-low method. The comparison reveals that the Dit values obtained by the two methods are of the same order of magnitude. Finally, two-dimensional (2D) distributions of local-DLTS signals are obtained, which show substantial intensity variations resulting in random 2D patterns. The 2D distribution of the local-DLTS signal depends on the time constant, which may be due to the coexistence of multiple types of traps with different capture cross sections.
ASJC Scopus subject areas
- Physics and Astronomy(all)