Considerations for capacitance DLTS measurements using a lock-in amplifier

Publisher Website
Google Scholar
Abstract
An outline is presented of how to numerically account for the gate-off time, consisting of both the pulse width and delay times, when performing lock-in amplifier based capacitance deep level transient spectroscopy (DLTS) measurements at frequencies up to 2000 Hz. This frequency is about ten times higher than the maximum frequency normally used for these measurements when using 1-MHz capacitance meters. Further, the relationship between the frequency, pulse width, and delay time was established for which the commonly used value of the normalized thermal emission decay time constant τmax/T0 may be assumed constant (=0.424) without introducing observable errors in the defect parameters calculated from DLTS data. Because the effect of measuring at higher frequencies is to shift the DLTS peaks to higher temperatures, it was found that when using frequencies between 50 and 2000 Hz, defects such as the E2 level in radiation-damaged GaAs, which are usually observed below 77 K at frequencies below 50 Hz, may be conveniently and accurately characterized by capacitance DLTS at temperatures above that of liquid nitrogen.