Electronic Structure And Magnetism Of ZnO-based Magnetic Semiconductors

As we all know, the electron has two fundamental degrees of freedom, e.g. charge and spin. Conventional microelectronics based on electronic charges and transportation properties has greatly promoted the advancement of society and the development of human. However, in conventional microelectronics, we only take advantage of the charge property, neglect the spin degree of freedom of the electron. With the development of society, we pay attention to the spin of electron increasingly. Until 1980s, the discovery of Giant magnetoresistance stimulates researcher's interest in magnetic materials, and this launched the new field of spin electronics- 'spintronics', which aims to combine both spin and charge of electron as the carriers of information, including their transport, management and memory, and is the hotspot field of the condensed matter physics.The magnetic semiconductor is the key materials of the spintronics, which has received much attention in the past few years. The transition metal atoms substitute for the cation of the semiconductors, and the ferromagnetism is induced by the exchange interaction between transition metal atoms. Much effort has been devoted to searching for the magnetic semiconductors with intrinsic ferromagnetic and high Curie temperature. The transition metal ions doped similar Si or Ge is hard to achieve high Curie temperature, which attributes to the low solubility of dopants, consequently the transition metal ions are too far to couple together. In addition, the nearest dopants couple antiferromagnetically with each other usually. Therefore the Curie temperature is too low and it's hard to make great progress with Si or Ge based magnetic semiconductors. In 1990's, the Curie temperature of GaAs based magnetic semiconductor reaches up to 170K. Subsequently, a lot of results onⅡ-Ⅵcompounds (such as ZnO) develop rapidly.ZnO is anⅡ-Ⅵgroup semiconductor with wide band gap (3.4eV) and high exciton bound energy (60meV) and is an attractive multi-functional material in piezoelectricity and optics. Many different explanations about the origin of ferromagnetism in ZnO based magnetic semiconductors have been reported, and the Curie temperature differs greatly, from few K to room temperature. The origin of the ferromagnetic and how to obtain the magnetic semiconductor with intrinsic ferromagnetism and high Curie temperature are the researching key point.In the prime of the 21th century, d0 ferromagnetism was first reported in undoped HfO2, and different lattice defects are considered as the source of magnetism. In the following research, there are a lot of similar reports in oxide, such as SnO2, TIO2, ZnO et al. Most researchers attributed the unexpected undoped ferromagnetism to the lattice defects, such as anion vacancy, cation vacancy, vacancy cluster, interstitial and so on. However, theoretical study predicted that the defect with lowest formation energy is charged in some conditions. Furthermore, the magnetism is sensitive to the charge states of the defects.As a wide gap semiconductor material, the optical of ZnO is sensitive to the strain. Most studies hold that the tensile strain make the optical gap become narrowing, while the compress strain broaden the gap, and there is a gap shift. As a technical means, the strain has some influence on the magnetism of the transition metal doped ZnO, but the relevance is unclear.To theoretically study the magnetic semiconductors, two approaches are usually employed:(1) model Hamiltonian or (2) first-principles calculation based on density functional theory. The latter one has been used in this dissertation. We studied the electronic structures of ZnO-based magnetic semiconductors using first-principles calculation software. According to the band structures, we discussed the magnetic exchange interaction between transition metal atoms and the origin of ferromagnetism. Profiting from the development of density functional theory and computer technology, many first-principles calculation software packages, such as Vasp, Castep, Siesta, Quantum-Espresso, spring up in the last few decades. In our work, we used Quantum-Espresso and Vasp package mainly.In the Mn doped ZnO magnetic semiconductor, Mn atoms tend to close to with each other, and the antiferromagnetic ordering is more stable than the ferromagnetic one, which results in a low Curie temperature. The C can induce spontaneous and extended spin-polarization in the C doped ZnO. The introduction of C can make the nearest and farther Mn atoms couple ferromagnetically. And therefore, we can obtain the intrinsic ferromagnetism and high Curie temperature ZnO based magnetic semiconductor.For the undoped ZnO, we found that only the neutral Zn vacancy (Vzn0) and Vzn1-prefer the spin-polarized states, and the total magnetic moments are 2.06μB and 1.00μB, respectively. Although the Oi defect has a spin-polarized state with a magnetic moment of 1.96μB, its ground state is spin-unpolarized. Other native defects, such as Vzn2-, Vo0, Vo+, Vo2+, and Zni do not induce magnetic moments and thus have no contribution to the unexpected ferromagnetism. Therefore, the Vzn1- vacancies may be responsible for the unexpected room temperature ferromagnetism in un-doped ZnO.Most theoretical and experimental reports indicated that p-type defects can make the ZnMnO ferromagnetic. There are two kinds of common p-type native defects in ZnO, i.e. neutral Zn vacancy (Vzn0) and charged Zn vacancy (Vzn1-). Some researchers considered the VZn0 as the main origin, while others found the VZn1- in ferromagnetic ZnMnO. Most theorists focused on the VZn0, the role of Vzn1- is unclear. We studied the roles of both Vzn0 and Vzn1- on the magnetism in ZnMnO, and found that the Zn(Vzn0)MnO with the lower formation energy than the Zn(Vzn1-)MnO one, is the main cause in ferromagnetic ZnMnO.