Ab initio-based Microscopic Modeling of the Optoelectronic Properties of Semiconductors

In order to find new semiconductor materials for applications, it is desirable to have a theoretical approach that can predict material properties without having the need to test these properties experimentally. This way, when looking for new materials or material combinations it is easier to find promising candidates. In this thesis, DFT is used as a method that can predict electronic and structural properties of periodic semiconductor crystals from ab initio calculations without the need for experimental input. It is combined with the semiconductor Bloch approach, which is a quantum theory including many-body effects used to calculate optical properties. One material class of interest for optoelectronic applications are monolayer TMDCs. We derived an expression for the quasi-2D Coulomb potential, which includes a form factor that contains the difference between the interaction in the ideally 2D case and the quasi-2D interaction in the TMDCs. DFT calculations were performed to obtain the wavefunctions necessary to evaluate the Coulomb interaction in various TMDCs. A MDF Hamiltonian was fitted to the parameters obtained from the DFT calculation. Then, the many-body theory was used to study excitonic properties and band gap renormalization effects of the TMDCs. A density-dependent band gap renormalization occurs due to the dynamic screening of the Coulomb interaction by excited carriers. It was shown that an electron-doped state does not have as strong an effect on the band gap as a symmetric electron-hole population which is usually created from optical excitation. Furthermore, the effect of the dielectric environment on the band gap renormalization was studied for freely suspended monolayers, monolayers on a fused silicon substrate and h-Bn encapsulated monolayers. By solving the Dirac-Wannier equations, exciton resonances and binding energies were calculated. It was found that an increased dielectric constant leads to an enhanced screening of the Coulomb interaction, which decreases the exciton binding energy. The incorporation of dilute amounts of bismuth into III-V semiconductors decreases the band gap, which allows for band gap engineering to design materials for specific applications. In order to predict the material properties of these dilute bismides with DFT, large supercells are needed, which increases the computational cost immensely. Therefore, we applied the LDA-1/2 method in DFT, which improves the band gaps obtained from DFT at a low computational cost. First, we calculated the band gaps of several III-V semiconductors using this method and found good agreement with experimental values. Then, we had to extend the method to be able to perform calculations on III-Bi compounds, which are not semiconductors in a purely binary compound. This allowed us to calculate the properties of Ga(SbBi) as a representative example of the dilute bismides. We performed DFT calculations on different supercell geometries and Bi concentrations. We studied the band gap narrowing and SO splitting increase with increasing Bi concentration and calculated effective band structures of the supercells showing defect states from the localized interaction with the Bi atoms. Absorption and PL spectra were calculated for the different concentrations and showed an earlier onset of absorption and a red-shift of the luminescence peak for increasing Bi concentrations because of the decreasing band gap. Furthermore, we studied the decreasing Auger losses in these materials due to the increasing detuning of the band gap and the SO splitting energy when incorporating higher amounts of Bi. Another interesting candidate for optelectronic applications is tellurium due to its chiral structure and strong nonlinear properties. We used the shLDA-1/2 method within DFT to obtain accurate electronic and structural parameters which compare well to experiments. We analyzed the TDMs of the highest four valence bands and the lowest two conduction bands for the two light polarization directions E parallel c and E perpendicular c. This revelaed that at the direct band gap at the H-point, the two highest valence bands do not couple significantly to the two lowest conduction bands in the E parallel c polarization direction, making transitions between them virtually forbidden. We used our microscopic approach to evaluate the optical properties of this material and show their strong dependence on the polarization direction. The absorption is stronger for the E perpendicular c direction, especially near the band gap, due to the weaker coupling of the valence bands in the E parallel c direction. Optical gain for incoherent carrier populations was calculated, revealing that at higher densities, the peak shifts from the transition to the lowest conduction band to the transition to the second-lowest conduction band in the E perpendicular c direction, while no significant gain was found in the E parallel c direction. Furthermore, we calculated PL spectra and found that the emission is weaker for E parallel c and the peak shifts more than for E perpendicular c, because the coupling increases in the energy regions that are farther away from the band gap. Another intersting property of semiconductors that is studied recently is their ability to generate high harmonic spectra after excitation with non-resonant strong-field THz radiation. We studied this nonlinear optical effect in Te by solving the coupled dynamics of interband polarizations and intraband currents. We illustrated the effect of the phase of the TDMs on the HHG spectra and performed a gauging of the TDMs obtained from DFT by using the phase of triple dipole products in order to obtain smooth and periodic phases across the whole BZ. We found that for the E parallel c direction, the two topmost valence bands do not contribute significantly to the HHG emission. Even harmonics are suppressed in the E parallel c direction due to quantum interference effects. Furthermore, we studied the effect of sample thickness on the spectra using a unidirectional pulse propagation solver. For the E perpendicular c direction, longer sample thicknesses quickly weaken the amplitude of higher harmonics, while for the E parallel c direction, this effect is not as strong because of the lower absorption. Since loss processes are crucial when determining the suitability of a material for applications, we performed another study on the losses in Te through radiative recombination and Auger losses. For this, the input from the DFT calculations with the shLDA-1/2 method is used to calculate radiative lifetimes and Auger rates. In the literature, different values for the splitting between the valence bands H4 and H6 have been reported. Since the splitting energy influences the Auger losses significantly, two different values have been considered in the study. In order to look at losses at different temperatures, the DFT band structure was shifted according to an experimentally found formula. The carrier lifetimes due to radiative and Auger recombination were calculated as a function of the carrier density at different temperatures for both splitting energies. While for low temperature and low density, the lower splitting energy causes the Auger lifetimes to be significantly shorter, at higher temperature and higher densities, the difference between the splitting energies becomes smaller, since holes far away from the band gap can reach the valence band H6. The radiative lifetimes are unaffected by the different splitting energies. The Auger loss coefficient and the bimolecular recombination coefficient for both splitting energies are calculated as a function of the density for different temperatures and analytical fit functions are determined for the temperature-dependent coefficients at the low-density limit.