Quasiparticle theory of defects in semiconductors and their excited states
Final Report Abstract
Electronic excitations of defect states in semiconductors are fundamental properties that are commonly characterized in experiments. Such excitations involve the localized defect states in the fundamental band gap of the semiconductor and the extended valence and conduction band or defect resonances. The project pursued a quasi particle theory of defects and their excitation on the basis of the GW-approach and the Bethe-Salpeter equation (BSE) for the treatment of the electron-hole interaction, thus going beyond the the traditional treatment in the frame work of density functional theory (DFT). Although successful in the prediction of several ground state defect properties the DFT was not constructed for the treatment of electronic excitations. Furthermore, the well-known band gap problem of the DFT affects the calculated position of defect levels within the band gap. The central idea was to approach the subject focusing on defect centers with an established identification. The positively charged carbon vacancy in SiC, identified by theory and electron spin resonance experiments (EPR) via the hyperfine tensors was ideal for the purpose of the project. DFT theory predicted a Jahn-Teller system in which the Frank-Condon shifts are crucial ingredients for the interpretation of the electronic excitations. Optical characterization experiments combined with EPR indicated that such shifts should be much smaller if present at all. The GW-approach together with the plasmon pole approximation yielded quasi particle corrections to the vertical ionization levels of the positive carbon vacancy obtained within the DFT on the order of 0.2eV that are geometry dependent thus enhancing the Jahn-Teller distortion of the defect. Although the interpretation of the photo-EPR experiments in terms of a quenching of the EPR signal of V+C by excitation of valence band electrons into the paramagnetic defect level and a latter recovery of the signal by an ionization of V0C seemed plausible from the DFT calculations, the absorption spectra calculated on the basis of the GW- BSE approach favor another interpretation. Besides the neutralization of V+C also a further ionization into a doubly positively charged defect Y2+C destroys the paramagnetic state and quenches the EPR signal. In the relevant range of excitation energies this constitutes the major channel in the spectra of V+C in 3C-SiC. We expect that the calculations underway for the polytype 4H confirm this picture. Development and optimization of the program package SELF employed for these calculations so far allowed defect super cells with 216 lattice sites. The prominent P6/P7 spectrum was another example the project aimed at. The P6/P7 center was originally identified with an excited state of the carbon vacancy-antisite complex. New experiments yielded EPR-spectra with a larger number of hyperfine lines than original one and showed that the signal belonged to a ground state of a defect. A combined experimental and theoretical investigation unambiguously identified the P6/P7 center with the neutral divacancy, a complex of a carbon vacancy with a neighboring silicon vacancy by comparing calculated and measured hyperfine tensor. In the same collaboration we were able to identify the SI5 center, tentatively explained by the di-vacancy, with the negative carbon vacancy-antisite complex. Both defects are important for the understanding of the carrier compensation in SiC. The excitation spectra of the SI5-center was analyzed within the ∆-SCF approach explaining the photo-threshold observed in the photo enhancement of the EPR signal, yet quasi particle and excitonic effects have to be included for a final conclusion. The project demonstrated that the analysis of electronic excitations of defects within the GW and BSE approaches is an essential tool for the understanding of optical spectra.
Publications
- Divacancy and its Identification: Theory. Mater. Sci. Forum 527-529, 523 (2006)
A. Gali, M. Bockstedte, N. T. Son, T. Umeda, J. Isoya, and E. Janzen
- Identification of divacancies in 4H-SiC. Physica B 376-377, 334 (2006)
N. T. Son, T. Umeda, J. Isoya, A. Gali, M. Bockstedte, B. Magnusson, A. Ellison, N. Morishita, T. Oshima, H. Itoh, and E. Janzen
- Identification of the carbon antisite-vacancy pair in 4H-SiC. Phys. Rev. Lett. 96, 145501 (2006)
T. Umeda, N. T. Son, J. Isoya, E. Janzen, T. Ohshima, N. Morishita, H. Itoh, A. Gali, and M. Bockstedte
- Signature of the negative carbon vacancy-antisite complex. Mater. Sci. Forum 527-529, 539 (2006)
M. Bockstedte, A. Gali, T. Umeda, N. T. Son, J. Isoya, and E. Janzen
