Computing material properties with SIESTA

Abstract

I will give a brief introduction to Density Functional Theory (DFT), a theoretical approach that has allowed a huge advance in the fundamental understanding of the properties of materials. DFT uses quantum mechanics to predict the behavior of materials from first-principles, without having to resort to experimental information. DFT has had a tremendous impact on fundamental research (Prof. Walter Kohn won the 1998 Nobel Prize for Chemistry for its development). Its use in the industrial environment has been less intense, as industrial problems generally involve complex materials properties at different length and time scales, which are not straightforward to address with atomistic methods and models. However, the availability of powerful computers, the development of theoretical methods that allow for DFT to handle systems with a very large number of atoms, and the advances in linking DFT with models that expand to larger scales (via multi-scale approaches) is changing this situation. In this talk, I will present SIESTA, one of the most widely used codes to perform DFT simulations in materials. SIESTA is particularly efficient and is able to handle systems with a very large number of atoms, as its computational cost scales favorably as the number of atoms is increased (even linearly). SIESTA has a number of specific capabilities that also makes it very attractive for industrial applications, such as the possibility of simulating the electric characteristic of nanodevices, or to link to larger scales through QM/MM methodologies. It also performs particularly well in large supercomputers. I will describe the main features of SIESTA, and illustrate them with specific examples from the work that my group has done during the last years. These will include 2D materials, the study of magnetic materials, vibrational spectra, thermal transport, etc.