Electrical fields in AlN/GaN-nanowires measured with aberration-corrected 4D-STEM

Abstract text (incl. figure legends and references)

The functionality and efficiency of electronic or optoelectronic devices can be critically influenced by electric fields in the structures. Such fields may arise from inhomogeneous doping, spontaneous or piezoelectric polarisation or occur during operation due to an externally applied voltage. A prominent example is the quantum-confined Stark effect, which can strongly diminish the output of light emitting structures even if crystalline quality and composition homogeneity are otherwise excellent. This is especially the case for crystals with an inherent polarisation as found in the wurtzite-structured AlN/GaN material system.

A position resolved measurement of the electric field strength inside of epitaxially grown structures – ideally with the ability to apply voltages to the specimens in-situ – is hence an important tool for the investigation and optimisation of their growth and subsequent operation. As aberration-corrected STEM allows for imaging with atomic resolution, it is an ideal method for such investigations. But while other specimen properties such as composition or geometry can be characterised from strongly integrated STEM signals acquired with large annular detectors, the signal footprint of electric fields in the sample is more delicate: They mostly manifest in a slight redistribution of intensity in the diffraction patterns. For quantitative field measurements, it is hence necessary to acquire and evaluate 4D-STEM datasets, where the diffraction pattern for each scan point is recorded. From these sets, field strengths can be derived under certain conditions and assumptions.

This work focusses on 4D-STEM field measurements in AlN/GaN nanowire heterostructures. These wires are grown by MBE and investigated with a probe-corrected TEM equipped with an ultrafast, pixelated, direct electron detection detector. The measurements are accompanied by multislice simulations to assess their accuracy and precision. Prerequisites for quantitative measurement of polarisation-induced fields are discussed and the management of possible artefacts due to changes in the mean inner potential are addressed. Based on these considerations, experimental results of high-resolution 4D-STEM field-measurements are presented. The conversion into maps of electrostatic potential or charge density is discussed.

Due to the high resolution inherent to STEM, the measured signal is at first dominated by the very large atomic fields that obscure the polarisation-induced fields. An appropriate averaging process is therefore warranted and is discussed in detail. One exemplary evaluation is shown in Fig. 1. The resulting mesoscalic fields are compared to the theoretical expectations.

In order to investigate the influence of external fields, wires are contacted by electron lithography on ultrathin silicon nitride membranes. Different voltages are applied in-situ employing a dedicated electrical biasing specimen holder. These results can be related to photoluminescence spectra of the same wires.

Fig. 1: Measurement of fields in an AlN/GaN layer structure in a nanowire. A 4D-STEM dataset was acquired in the marked area of the overview on the left. It was evaluated yielding a map of electric field strength in growth direction shown in the top right. The different spontaneous and piezoelectric fields in AlN and GaN can be measured.