Development of Tools for the Study of Heavy Element Containing Systems with the CRYSTAL Code and their Applicatio

Please join us for a Grad student seminar this Friday Jan 11 at 3:30 pm in rm 155 Geology presented by Jacques Desmarais, PhD Candidate:

Development of Tools for the Study of Heavy Element Containing Systems with the CRYSTAL Code and their Application

The purest form of fundamental knowledge in Geology and planetary sciences is obtained through knowledge of the properties (i.e. physical or chemical properties) and fundamental states of planetary materials. Such knowledge can be obtained from either direct measurements on natural samples or laboratory experimental simulations, although both may be subjected to severe technical limitations/challenges. This is because, for example experiments can only be carried at a limited range of pressures and temperatures relevant to planetary interiors (typically at best up to lower mantle conditions), or because the experiments can only measure a limited set of observables which may not give a full picture of the chemical or physical process of interest. So some knowledge on the properties and/or states of planetary materials cannot be obtained solely from physical experiments, but can instead be obtained by modelling techniques. Ideally the modelling techniques are built entirely independently from experimental knowledge, that is to say using only first principles (i.e. fully ab initio techniques).

I discuss here the development of such quantum-mechanical ab initio techniques for studying periodic systems in 0D (molecules), 1D (infinite chains), 2D (surfaces, for example mineral surfaces), and 3D (bulk solids, for example minerals), which have broad applications in Geology. The techniques are developed within the CRYSTAL computer program. This is a public code used around the world for ab initio calculations in periodic systems, and has been developed now for over thirty years (principally in Torino, Italy, but also through international collaborations), with the first public version being released in 1988 and the most recent version in 2017. I specifically discuss developments I have made to improve the description of heavy-element containing systems (i.e. transition metals and heavier), as follows.

In general, any ab initio method can be classified according to three factors: 1. its treatment of electronic correlation; 2. the quality of the basis set used to expand the wavefunction and; 3. its treatment of relativistic effects. I discuss work done in order to improve the treatments of points two and three with the CRYSTAL code. For point two, I extend the existing approach in the CRYSTAL code to -type functions (i.e. angular-momentum quantum number  functions) and the development and benchmarking of associated basis sets for the Lanthanide and Actinide series. For point three, I discuss the work done to treat spin-orbit coupling (a type of relativistic effect) in the CRYSTAL code by self-consistently solving the so-called Dirac equation, instead of the already existing approach which only solves the non-relativistic Schrödinger equation.