Quantitative electric field determination in semiconductor homo- and hetero-interfaces using electron diffraction and 4D-STEM

Abstract text (incl. figure legends and references)

Introduction:

Nanometer scale characterization of electric fields at semiconductor interfaces is necessary for most electronic devices such as lasers, transistors, solar cells and batteries in order to understand the performance of those devices. Various experimental methods such as electron holography [1], differential phase contrast (DPC) imaging [2], the Fresnel method [3] and the Focault method [4] have been reported in the past for the imaging of built-in electric fields. Recently, we reported the quantitative evaluation on the electric field across a GaAs-based pn-junction using the center of mass (COM) shift of the central beam by 4D-STEM [5].

Objectives:

We aim for the quantification of long-range electric fields that build up at the homo- and hetero-interfaces by measuring the COM shift of the central beam disc by combining precession electron diffraction (PED) with 4D-STEM.

Materials and methods:

In the present study, we use a Nanomegas ASTAR PED system attached to a double aberration corrected JEOL JEM2200FS for quantifying the fields across both homo- and hetero-interfaces. Therefore, we use the parallel beam scanning mode where the convergence angles are in the range of 1-3mrad. Accordingly, dynamical diffraction effects, which destroy the signal at hetero-junctions are reduced as we precess the beam around the optical axis of the microscope. In this way, we manage to get rid of the signal from the atomic fields improving the resolution for the much weaker long-range fields.

Results:

GaAs-based pn-junctions with different doping concentrations are analyzed to show the potential of this method. We optimize the parameters for observing the COM shift of the central beam disc which is determined by the built-in electric field, the polarity and the doping concentrations of the pn-junction. Fig.1 shows the COM evaluations performed on the GaAs-based pn-junction. Further, we apply this method to hetero-interfaces such as AlAs/GaAs and GaP/Si. Apparently, the observed COM shift at the hetero-interfaces also relates to the change in the mean inner potential (MIP). Fig.2 shows the experimental results across the AlAs/GaAs hetero-interface.

Conclusions:

Our results clearly indicate that the presence of electric fields caused by doping or interatomic bonding at the interface can be detected. The presentation will explain the PED-4D STEM experiments, the optimum parameters for those as well as will give some experimental examples.

Fig 1: Experimental measurements of a GaAs-based pn-junction (a) Virtual Bright Field (BF) image with an inset showing the central beam disc with precession of 0.45° (b) COM shift across the depletion region perpendicular to the growth direction [001] (c) COM shift and electric field profiles at the depletion region and linear fits to the rise and fall of the peak in electric field.

Fig 2: Experimental measurements of a AlAs/GaAs heterointerface (a) Virtual BF image with an inset showing the central beam disc with precession 0.45° (b) COM shift at the interface perpendicular to the growth direction [001] (c) COM shift and electric field profiles across the interface.

References:

Sasaki et al. Microscopy 63, pp.235 (2014). Shibata et al., Sci. Rep. 5, pp.10040 (2015). Merli et al Phys. Status Solidi 1973, 16 (2), K89−K91. Titchmarsh el al status solidi (b), 34(2), K83-K86. Beyer. A. et al Nano Lett. (2021),21,5, 2018-2025.