Title:	Thermodynamics of metals from first principles
DNr:	SNIC 2018/3-168
Project Type:	SNIC Medium Compute
Principal Investigator:	Mattias Klintenberg <mattias@physics.uu.se>
Affiliation:	Uppsala universitet
Duration:	2018-03-29 – 2019-04-01
Classification:	10304
Keywords:

Abstract

Density functional theory has become the standard method to predict material properties from first principles, however it can in principle only do that at 0 K temperature. Most materials are used at room temperature, many at a lot higher. In order to calculate and predict properties at those finite temperatures a full quantification of the free energy surface of the material is required. Recently the available computational power has risen to the level where it has become possible to calculate all the relevant thermal excitation mechanisms, such as electronic, harmonic and anharmonic vibrational, magnetic and point defect, allowing for theoretical explanation of thermodynamic properties such as heat capacity, thermal expansion coefficient or phase stability. Several techniques such as "thermodynamic integration based on Langevin dynamics" or "temperature-dependent effective potential" have been used in order to reduce the complexity of these calculations, however the accuracy and predictive power of those has not been well established. In addition the "standard" implementation difficulties associated with DFT such as selection of an exchange-correlation functional, k-point sampling, choice of cutoff energies, system size effects and so on further complicate the situation. The aim of the project is to shed light on these matters by systematically investigating the intricate details behind calculating the free energy contributions, comparing different methods and assessing their accuracy, limitations and advangates for different metallic systems.