Buried dislocation networks designed to organize the growth of III-V semiconductor nanostructures

TL;DR

This study investigates the structure of buried dislocation networks in GaAs crystals and demonstrates how these networks influence the lateral ordering of semiconductor nanostructures during epitaxy.

Contribution

It provides a detailed TEM analysis of dislocation networks at buried interfaces and links their structure to the controlled growth of nanostructures.

Findings

Dislocation networks consist of screw and mixed dislocations accommodating twist and tilt.

Small twists lead to the disappearance of one screw dislocation subnetwork.

Dislocation structures govern the dimensions and orientations of nanostructures during epitaxy.

Abstract

We first report a detailed transmission electron microscopy study of dislocation networks (DNs) formed at shallowly buried interfaces obtained by bonding two GaAs crystals between which we establish in a controlled manner a twist and a tilt around a k110l direction. For large enough twists, the DN consists of a twodimensional network of screw dislocations accommodating mainly the twist and of a one-dimensional network of mixed dislocations accommodating mainly the tilt. We show that in addition the mixed dislocations accommodate part of the twist and we observe and explain slight unexpected disorientations of the screw dislocations with respect to the k110l directions. By performing a quantitative analysis of the whole DN, we propose a coherent interpretation of these observations which also provides data inaccessible by direct experiments. When the twist is small enough, one screw subnetwork vanishes. The surface strain field induced by such DNs has been used to pilot the lateral ordering of GaAs and InGaAs nanostructures during metal-organic vapor phase epitaxy. We prove that the dimensions and orientations of the nanostructures are correlated with those of the cells of the underlying DN and explain how the interface dislocation structure governs the formation of the nanostructures.