Synthesis of Germanium-Tin Alloys by Ion Implantation and Pulsed Laser Melting: Towards a Group IV Direct Band Gap Semiconducto

The germanium-tin (Ge1-xSnx) material system is expected to be a direct bandgap group IV semiconductor at a Sn content of 6.5--11 at.%. Hence there has been much interest in preparing such alloys since they are compatible with silicon and they raise the possibility of integrating photonics functionality into silicon circuitry. However, the maximum solid solubility of Sn in Ge is around 0.5 at.% and non-equilibrium deposition techniques such as molecular beam epitaxy or chemical vapour deposition have been used to achieve the desired high Sn concentrations.;In this PhD work, the combination of ion implantation and pulsed laser melting (PLM) is demonstrated to be an alternative promising method to produce a highly Sn concentrated alloy with good crystal quality. In initial studies, it was shown that 100 keV Sn implants followed by PLM produced high quality alloys with up to 6.2 at.%Sn but above these Sn concentrations the crystal quality was poor. The structural properties of the ≤6.2 at.% alloys such as soluble Sn concentration, strain distribution and crystal quality have been characterised by Rutherford backscattering spectrometry (RBS), Raman spectroscopy, X-ray diffraction (XRD) and transmission electron microscopy (TEM). The optical properties and electronic band structure have been studied by spectroscopic ellipsometry. The introduction of substitutional Sn into Ge is shown to either induce a splitting between light and heavy hole subbands or lower the conduction band at the Gamma valley.;However, at higher implant doses needed to achieve >6.2 at.% Sn, ion-beam-induced porosity in Ge starts to occur, which drastically reduces the retained amount of the implanted Sn and such microstructure also hinders good crystallisation of the material during PLM. To solve this problem, it was shown that a nanometer thick SiO2 layer deposited on the Ge substrate prior to the implantation can largely eliminate the formation of porosity. This capping SiO2 layer also helps to increase the retained Sn concentration up to 15 at.% after implantation, as well as significantly improving the crystal quality of the Ge-Sn layer after PLM. With the use of the capping layer, a good quality Ge-Sn layer with ~9 at.% Sn has been achieved using Sn implants at an energy of ~120 keV. However, the thin film alloys produced by 100 keV or 120 keV Sn implantation and PLM are shown to contain compressive strain as a result of the large lattice mismatch between Ge and high Sn content alloys. Such strain compromises the tendency towards a direct bandgap material and hence strain relaxation is highly desirable. A thermal stability study showed that the thin film strained material is metastable up to ~400°C, but thereafter Sn comes out of solution and diffuses to the material surface.;To investigate a possible pathway to the synthesis of strain-relaxed material, a higher Sn implant energy of 350 keV was used to produce thicker alloy layers. XRD/reciprocal space mapping showed that this thicker alloy material is largely relaxed after PLM, which is beneficial for the direct band gap transition and solves the trade-off between higher Sn concentration and compressive strain. However, RBS indicates a sub-surface band of disorder which suggested a possible mechanism for the strain relaxation. Indeed, TEM examination of such material showed the material relaxed via the generation of non-equilibrium threading defects. Despite such defects, a PL study of this relaxed material found photon emission at a wavelength of ~2150 nm for 6--9 at.% Sn alloys. However, the intensity of the emission was variable across different Sn content alloys, presumably as a result of the threading defects. A possible pathway to removing such defects is given that may enable both photodetectors and lasers to be fabricated at wavelengths above 2mum.