Multiscale growth and doping simulations of nanostructured devices

点击翻译按钮获取中文摘要

To optimize/design nanostructured devices based on wide-bandgap semiconductors it is crucial to (i) understand and identify growth modes causing self-organization and to (ii) achieve high doping concentrations which are particularly critical for laser devices to realize high injection currents and good contacts. Recent experimental and theoretical studies showed that these two issues are often related: High doping concentrations may significantly modify or even open new growth regimes and the structure/morphology of the surface has significant effect on the doping efficiency. A challenge in simulating such types of interaction is the large range of relevant length and time scales. The features interesting for device design are of the order of 100 - 1000 nm (length) and the time to grow these structures is of the order of seconds. The origin of these effects, however, lies in the atomic processes on the surface which requires a resolution in the length scale of Angstroms and in the time scale of picoseconds. In order to bridge the gap in length and time scales we will develop and apply a multiscale approach which hierarchically combines density functional theory with thermodynamic concepts and elements of statistical physics. Based on this approach we will study the effect of various dopants/impurities on growth mode, surface morphology and nanostructure formation. These results will be used to analyze the experiments of projects II-1 and II-2 and will serve as input parameters for the theory projects III-2 and III-3.