Surface effect that causes a peak in band-edge photocurrent spectra: a quantitative model

TL;DR

This paper presents a quantitative model explaining how surface effects influence the shape of band-edge photocurrent spectra in GaN and GaAs, resolving previous inconsistencies and enabling precise determination of band edge energies.

Contribution

The study introduces a model that accounts for surface effects causing spectral shape variations, allowing accurate extraction of band edge energies from photocurrent spectra.

Findings

The model explains the peak-to-step transition in spectra based on surface effects.

Experimental validation shows the ability to switch spectral shapes in GaAs layers.

The model accurately determines band edge energies despite surface-induced red shifts.

Abstract

Band edge photocurrent spectra are typically observed in either of two shapes: a peak or a step. In this study, we show that the photocurrent band edge response of a GaN layer forms a peak, while the same response in GaN nanowires takes the form of a step, both are red-shifted to the actual band edge energy. Although this apparent inconsistency is not limited to GaN, the physics of this phenomenon has been unclear. To understand the physics behind these observations, we propose a model explaining the apparent discrepancy as resulting from a structure-dependent surface-effect. To test the model, we experiment with a GaAs layer showing that we can deliberately switch between a step and a peak. We demonstrate that using this quantitative model one may obtain the exact band edge transition energy, regardless of the red-shift variance, as well as the density of the surface state charges that cause the red shift. The model thus adds quantitative features to photocurrent spectroscopy.