Silicon Carbide (SiC) has long been considered a promising semiconductor material for high power devices, and has also recently found to be one of the emergent materials for quantum computing. Important for these applications are both the quality and purity of the crystal. In order to be able to engineer components (be it power devices or components for quantum computing), it is necessary to study and understand the behavior of various defects in the crystal.
Deep level defects can greatly influence the semiconducting properties, since they can act as recombination centers by interacting with both holes from the valence band and electrons from the conduction band. Because of this, they may be used to control the charge carrier life time. Besides influencing the electric properties of the materials, deep level defects are also of interest in the field of quantum computing. In this application, the deep level defects can be used as basic units for quantum information – so called qubits.
Deep level defects may also be classified based on their origin, i.e. impurity or intrinsic. An impurity consists of one or more foreign atoms, which means neither carbon nor silicon in the case of SiC. Impurities can be incorporated in the crystal during growth, or through implantation or diffusion. A defect is intrinsic when it does not involve foreign atoms, but instead imperfections in the perfect crystal structure, for example a vacancy, an anti-site or a combinations of these. Intrinsic defects can be created during growth or artificially, using for example electron irradiation.
This thesis is focused on characterization of several deep level defects in SiC using different optical techniques. The optical transitions investigated are in the near-infrared region.
Paper 1 focuses on the possibility to control the concentration of intrinsic defects through the cooling down procedure after high temperature annealing. The temperature of 2300°C is close to the bulk crystal growth temperature. It is shown that it is possible to control the concentration of the silicon vacancy (VSi) and UD-2 (later identified as the divacancy (VCVSi)) by the cooling sequence. Both these defects have later been shown to be promising candidates as qubits and single photon emitters.
Paper 2 gives insight into the electronic structure of the unidentified deep level defect UD-4, which is believed to be of intrinsic origin. The defect is investigated in the polytypes 4H-, 6H-, and 15R-SiC, and the number of optical centers associated with UD-4 follows neither the number of inequivalent sites nor the possible configurations for pair-defects. There are two optical centers in 4H- and 6H-SiC, and three optical centers in 15R-SiC.
Paper 3 investigates several transition metals incorporated in SiC and the formation energies for different possible configurations. This is of importance since several impurity related deep level defects cannot be explained as purely substitutional defects, based on the fact that the number of optical centers does not follow the number of inequivalent sites. This is investigated in detail, and explained using an asymmetric split vacancy (ASV) model. It was found that the formation energy for some transition metals in ASV are lower than the transition metal in a substitutional configuration. Further on, it was shown that the formation energies for transition metals in ASV configurations depend strongly on what kinds of inequivalent sites the ASV can be described by and the lowest formation energy that is found for transition metals in ASV occupying two hexagonal sites.
In paper 4, the optical identification and electronic configuration of the commonly observed deep level defect tungsten (formerly known as UD-1) are reported. The electronic levels involved in the optical transitions of tungsten are deduced and described using group theory techniques.
Paper 5 shows that the above mentioned ASV model can be used to describe the properties of niobium in SiC. In the paper, the optical identification and properties are analyzed and investigated experimentally using photoluminescence, photoluminescence excitation spectroscopy and Zeeman spectroscopy.
In paper 6 the identification of molybdenum (formerly known as I-1) is reported including its electronic configuration. Molybdenum can be well described using the ASV model, and in this paper its local vibrational modes are also investigated in detail. It is shown that using the polarization dependence of local vibration replicas and a simplified defect molecule model, the estimated position of Mo in the ASV is in agreement with the theoretically predicted position reported in paper 3. The usefulness for molybdenum in SiC as a qubit is also investigated.
In paper 7, two different intrinsic nearest pair-neighbor defects are reported: UD-2 (VCVSi) and UD-0 (tentatively assigned as the VCCSi). Their optical properties are analyzed together with their creation and annihilation properties.