Identification of Near-Interface Trap Distribution by Parameter Estimation

To identify the near-interface trap (NIT) distribution of a metal oxide semiconductor (MOS) capacitor, we propose a new evaluation method by parameter estimation through optimization. The MOS capacitor was fabricated with Al/SiO2 (75 nm)/SiC and measured by the capacitance transient (C-t) method. In addition, C-t signals were calculated from the assumed NIT distribution model. Then, the calculated C-t signals were modified to fit the measured signals by optimization of the parameters of the NIT model. The two types of NITs, deep (Ec – Et = 0.57 eV) and shallow (Ec – Et = -0.02 eV or-0.18 eV), were revealed by this method.