Substitutional dopants in III-V semi-conductors, such as Si atoms in GaAs, are of great interest for the applications in transistors, Schottky diodes, and doping super-lattices which have been widely employed to control the electrical properties of semi-conductors. Although Si doped GaAs systems have been intensively investigated theoretically and experimentally in the last
several decades, some properties are still debated. In order to give a further explanation of Si
doped GaAs systems, we systematically studied DX center in bulk GaAs and in GaAs(110), as
well as the relative stabilities of different charged systems for Si atom replacing Ga atom at the substitutional site near GaAs(110) surface from first principles ground state
method. We show that DX centre is a metastable state in bulk GaAs and completely unstable in
the top few layers of GaAs(110). When Si atom replaces Ga atom at the surface, Charge states have an important influence on the stability of the system, and the additional charge is mainly concentrated on the Si atom for charged system. In addition, we studied the STM images of clean GaAs(110) and
charged Si:GaAs(110) by employing Tersoff-Hamann approximation. The calculated STM
images are in good agreement with experimental results. We show that at the positive bias
voltage the positively charged Si atom presents a bright feature while the negatively charged Si
atom shows a dark feature. In a semi-conductor, all bands are either completely full or completely empty. It is well known that DFT underestimates the band gaps of semi-conductors, a simple rigid shift can be used to
correct the band energies of semi-conductors. Unlike semi-conductor, the fermi energies of
metals lie in some bands. Furthermore, it turned out that some noble metals such as Cu and Ag
depend on the considered band and k point , therefore, the so-called scissors operator can not be
used for the metallic systems. The most successful approach within theoretical method for these
metals is the many body perturbation theory. On the other hand, an interesting study for metals is quasi-particle excitations, which play an important role in a rich variety of physical and
chemical phenomena such as energy transfer in photochemical reaction, desorption and
oxidation of molecules at surfaces, spin transport within bulk metals, across interfaces, and at
surfaces. One of the crucial properties of quasi-particle excitation is their lifetimes which
determine the duration of these excitations. We carried out the calculations of quasi-particle
band-structures and lifetimes for noble metals Cu and Ag within the GW approximation. For
Cu, both the calculated positions of the d bands and the width of the d bands is within 0.1 eV
compared to the experimental results. For Ag, partial core correction should be included in the
pseudo-potential to get reliable positions of the d bands. The calculated lifetime agree with the
experiment in the energy region away from the Fermi level, but deviates from the experimental
results near the Fermi level where short range interactions which GW approach fails to describe
play an important role. For a better description of the lifetime near the Fermi level, higher terms
beyond the GW approximation in the many body perturbation theory need to be considered. In
addition, the image potential state lifetimes in Cu(100) have been calculated using GW
approximation based on the localized Gaussian basis set, and the calculated n=1, 2 imagepotential
state lifetimes are in good agreement with experimental results.
| Identifer | oai:union.ndltd.org:uni-osnabrueck.de/oai:repositorium.ub.uni-osnabrueck.de:urn:nbn:de:gbv:700-201002115394 |
| Date | 11 February 2010 |
| Creators | Yi, Zhijun |
| Contributors | Prof. Dr. Michael Rohlfing, Prof. Dr. Jochen Gemmer |
| Source Sets | Universität Osnabrück |
| Language | English |
| Detected Language | English |
| Type | doc-type:doctoralThesis |
| Format | application/pdf, application/zip |
| Rights | Namensnennung 3.0 Unported, http://creativecommons.org/licenses/by/3.0/ |
Page generated in 0.01 seconds