Navigation und Service


Can We Build a Universal Framework for Electronic Structure?

Prof. Dr. Mark van Schilfgaarde, Materials and Molecular Modeling, Kings College, London

15.06.2012 11:00 Uhr


How electrons behave and respond to forces, or the \Electronic Structure," is the key to understanding properties of materials at their most fundamental level. Powerful theories accomplish this by solving the fundamental equations of quantum theory ab initio, meaning from rst principles without recourse to empirical or model parameters. By far the most popular are the family of methods based in density functional theory (DFT). The standard approximation to DFT, the local density approximation and its semilocal relatives, have evolved over time, and they are used ubiquitously in almost every branch of science and engineering because they can successfully describe many kinds of materials and their properties. But this approach suers from many well known limitations, which as a result has spawned a veritable zoo of ab initio or quasi-ab initio methods has evolved, to overcome them. The majority of new approaches being developed in this eld attempt to extend the formulations to DFT in one way or another. In this talk I will describe this evolution, and what I see as its successes and failures.
Many extensions are designed to improve on one or another shortcoming, often by singling out a subsystem for improved treatment. But a key point is that because the LDA is based on an ansatz, it is not systematically improvable. Extensions also suer from ambiguities, such as how a subsystem state is singled out, or how double-counting terms should be subtracted. I will present a dierent approach, a self-consistent scheme within the GW approximation. I will brie y describe the GW approximation and why self-consistency is important. This approximation seems to dramatically improve on DFT but has limitations of its own, which I will describe. But because it does not depend on an ansatz, it has the potential to be extended in a relatively systematic manner.


Prof. Dr. Stefan Blügel
Telefon: +49 2461 61-4249
Fax: +49 2461 61-2850