Interpretation of capacitance vs. voltage measurements of p- n junctions

Abstract

In profiling impurity distributions by use of differential capacitance measurements, it has been customary to interpret d( 1 C 2 )/ dV as proportional to the inverse of impurity concentration at the edge of the depletion region for a one-sided junction. In a recent paper, based on numerical calculations, Kennedy, Murley and Kleinfelder (KMK) observed that this interpretation is incorrect and that d( 1 C 2 )/ dV yields the zero-bias majority carrier distribution. In the present paper, again based on numerical calculations, we show that while the KMK interpretation is approximately correct in that the apparent doping concentration found by applying the C- V relation is an excellent approximation to the zero-bias majority carrier concentration, the two are not exactly equal. This imposes limitations on applying the correction described by Kennedy and O'Brien to obtain the actual impurity concentration from the apparent doping concentration.