Characterization of Growth of GaAs on Si/SiO2 via a Thin Layer of Perovskite

Wafers of silicon (Si) are the standard material for deposition of thin layers used for all electronic devices; however, Si has an indirect band-gap and is thus not very efficient for optical applications. Gallium arsenide (GaAs) is much more expensive than Si, but due to its direct band-gap it is widely used for light-emitting and laser diodes. One approach towards an integrated optoelectronics with high efficiency is to grow GaAs on Si. Due to lattice mismatch, direct epitaxy of GaAs on Si leads to strain and so to dislocations and other defects detrimental to high-quality thin layers. There are various approaches for epitaxy of GaAs on Si, and this project investigates one of them.
The purpose of this research has been investigating and characterizing the growth of GaAs on Si via thin layers of perovskites on top of native silicon oxide. The idea that motivated this study was a allegation by Motorola in 2001, aiming at commercialising this route for high frequency wireless and opto-electronic devices.
Pulsed laser deposition (PLD) of thin layers of perovskite materials (SrTiO3 and BaTiO3) on Si(001) substrates via layers of native oxide was performed at the Technische Universität Darmstadt, Germany. The wafers then have been sent back to the University of Sheffield. The as-grown materials have been investigated using different techniques such as atomic force microscopy (AFM), scanning transmission electron microscopy (STEM) and ellipsometry. Then they have been annealed and over-grown at the National Centre for III-V Technologies at the University of Sheffield using a combined molecular beam epitaxy (MBE)-scanning tunnelling microscopy (STM) system with built-in reflection high energy electron diffraction (RHEED). The over-grown specimens have finally again been investigated using AFM, TEM, STEM and energy dispersive X-ray spectroscopy (EDXs) techniques. Here, particular emphasis has been on correlating quantitatively surface topology measurements by AFM with interface roughness measurements on cross-sectioned samples in STEM.
The aim of this study was to find a way to over-grow silicon substrates with GaAs with low dislocation density, however, the perovskite layers and the subsequent GaAs never grew epitaxially but always remained poly-crystalline. This has been attributed to a lack of surface reconstruction of the perovskite layers when annealed under vacuum conditions, as confirmed by in-situ RHEED. Other studies have used sulphide buffer layer or a relaxed buffer layer of BaxSr1−xTiO3 to guarantee the perovskite thin layer to grow epitaxially. The over-growth of GaAs also has been done with the presence of oxygen to prevent the perovskite to evaporate.