Effect of heterostructure on the hole diffusion length of epitaxial GaAs

Abstract

By measuring the short‐circuit photocurrent in a Schottky barrier diode the minority carrier hole diffusion length in undoped epitaxial layers has been determined of GaAs grown on substrates of GaAs and GaP, and in similar structures with a layer of GaxAl1–xAs between the substrate and the epitaxial GaAs layer. In the heteroepitaxial layers grown directly on GaP the hole diffusion length is less than 0.1 μm. This is considerably smaller than the diffusion length in the homoepitaxial layers, which is between 0.2 and 3.0 μm. In the structures with an intermediate layer of GaxAl1–xAs the hole diffusion length is between 0.4 and 1.6 μm and is not significantly different from the homoepitaxial values. Etching studies have been made on the layers in an effort to correlate defect density with diffusion length. In the structures on GaP there is a high density of dislocation and stacking faults. However only in the heteroepitaxial layers directly on GaP does this significantly affect the diffusion length.

The electrical characterisation of semiconductors

“…The alternative method of measuring J , against W was used by Young and Rowland (1973) to study epitaxial n-type GaAs grown on GaP substrates with and without an intermediate GaAlAs layer and by Smith and Abbott (1972) on both n-and p-type Gap. However, Smith and Rbbott obtained values of I; considerably shorter than those obtained by Dapkus et al(1973) from SEM measurements (see below) and a careful comparison of the two photodiode methods (also on Gap) by Young and Wight (1974) showed that the J , against W method does, indeed, give values of L which are too short.…”