The Development Of Tight-binding Potential And Firtst Principle Study Of SiGe-based Materials

In the past decades, Si-based materials almost dominated the development of the entire semiconductor industry. Even though the transistor has been successfully fabricated in Ge and other semiconductor materials that may have more superior properties on carrier transport and band gap, silicon devices still account for over97%of all microelectronics. With the advent of the SiGe heterojunction bipolar transistor (HBT) and practical application since1998, SiGe-based materials become more and more important in the field of radio frequency. In addition, CMOS is dominant in semiconductor products due to its advantages of low-power consumption and little-heat emission. If SiGe can be integrated into CMOS in the future, SiGe-based materials will make significant effect on the development of MPU (microprocessor units) devices.The performance of the SiGe device is closely correlated with some building blocks in different dimensions. Previous experiments have focused on the properties of some SiGe systems. However, there are still many unknown physical mechanism which has not been explained theoretically. On the theoretical side, both of the empirical potential and the density functional theory (DFT) possess the limitation in studying the SiGe systems. Although a DFT method can provide very accurate description about the properties of a material, it is not suitable for studying large complex systems due to its huge computational cost. In contrast, an empirical potential can be used to simulate a huge system consisting of a large number of atoms, but it can’t provide electronic information of the system. The tight-binding calculations is a compromise between the DFT method and the empirical potentials:A TB model is more accurate than empirical potential in describing physical properties of a system and can provide electronic information, as well as the TB model has better computational efficiency than DFT method, due to the parameterization of its electronic Hamiltonian. In order to get insight of the physical properties of various complex SiGe systems, we develop a transferable tight-binding potential for germanium and binary systems of silicon and germanium with consideration of the environment-dependent bonding feature, which is presented in this thesis in detail.On the other hand, SiGe nanowires (NWs) have attracted extensive interest recently, due to its novel optical and transport properties. Especially for the transport properties, previous theoretical study has shown that an one-dimensional (1D) electron or hole gas can be realized through the doping of P or B in a thi Gecore/Sishell nanowires, which mainly due to the type-Ⅱ band offset in the nanowires. Activated by the above research, we, based on the DFT calculations, present a systematic study of n-type(N, P),p-type(B, Al) and O doping in Gecore/SiShell NWs, as well as the same doping in the other two kinds of SiGe NWs which possess type-II band ofset, namely, Sicore/Gesheii and fused triangular-prism Si/Ge NWs(FTP Si/Ge NWs).This thesis contains five chapters. The first chapter mainly presents the experimental and theoretical summary of SiGe systems ranging from lower to higher dimensions. Zero-dimensional systems include SiGe binary clusters and quantum dots. One-dimensional systems are mainly about SiGe nanowires and nanotubes. Two-dimensional systems mainly include SiGe quantum well and SiGe heterostructure growing on Si substrates. For three-dimensional systems, we take SiGe crystals as representative.In the second chapter, we introduce our developed environment-dependent tight-binding potential for germanium. Our extensive tests demonstrate that this potential model reproduces the relative stabilities of different crystal structures for germanium, the elastic constants, the phonon dispersion, and the formation energies of monovacancy and interstitial defects in germanium bulk. It is also shown that this model has a good performance in studying low-dimensional systems, such as the Ge(100) and Ge(111) reconstructed surfaces, the [110]-oriented nanowire Ge6o, and some small-sized and medium-sized germanium clusters. As an application, we combine this potential with some search methods, such as compressing liquid and genetic algorithm, to study the structural feature of some large-sized germanium clusters, Ge65, Ge7o and Ge75. Our predicted stable structural feature of these three germanium clusters is in accordance with that from experiment.In the third chapter, we present a transferable tight-binding potential for binary system of silicon and germanium with consideration of the environment-dependent bonding feature. It has been tested for bulk properties such as band structures and energy curves of several crystal structures, bulk modulus and phonon dispersion of diamond Si0.5Ge0.5alloy, and in nearly every case gives good agreement with LDA calculations. The model has also shown good transferability to handling the point defects and Ge atoms adsorbing on the Si(100) and (111) surfaces. In addition, we have performed tight-binding molecular-dynamics simulations to study the effect of concentration of germanium on the melting properties of Si1-xGex alloy and found that the melting point of the Si1-xGex alloy decreases as the concentration of Ge increases. Finally, we have applied the TBMC method to search the low-energy structures of Si0.5Ge0.5alloy and got qualitative agreement with LDA calculations in the energy orders of different configurations.In the fourth chapter, a tight-binding potential for Si-H and Ge-H interaction is briefly described. So far, the fitting database have just included the electronic structure and binding energy curves of some small molecular. In addition, this potential has not been tested for more complex systems or properties. Thus, it needs to be improved in the future.In the fifth chapter, we have studied the doping behavior of n-type(N, P),p-type(B, Al) and O in three different compositionally abrupt SiGe NWs, namely, Gecore/Sishell, Sicore/GeShell and fused-triangular prism Si/Ge NWs based on DFT calculations. Our study indicates that the substitution of Ge with P in all three types of Si/Ge NWs can achieve high density of electron carriers even without thermal excitation. On the other hand, the substitution of Si with Al or B in all three types of Si/Ge NWs can realize high density of hole carriers. Moreover, we find that the geometry of the Si-Ge interface can affect the position of the impurity band due to the local strain induced by the lattice-constant mismatch. For O doping, only the substitution of Si with O at the interface of the fused-triangular prism Si/Ge NW can generate high density of hole carriers. In the other cases, O doping has a little effect on the conductivity of the NWs.