Quantum Monte Carlo simulation methods for material science

Seminari_Logo
il giorno 11 Ottobre 2018 alle ore 14.30, in aula 1G09
Il Prof. Dario Alfe' (University College London)
 
 
 
terrà un seminario dal titolo:
 
 
"Quantum Monte Carlo simulation methods for material science"
 
 
 
 
 
Abstract
 
Computer simulations of materials have made enormous strides in the past few decades, mainly thanks to the development in the mid 1960's of a reformulation of quantum mechanics known as density functional theory (DFT). Although still under development, DFT is now the simulation method of reference of most research groups around the world, and has reached a lever of maturity that in most cases can be considered complementary to experiments.
However, the accuracy of DFT relies on a fundamental approximation for the quantity known as the exchange-correlation (XC) potential. It turns out that for a small but significant set of problems current XC approximations are still not
satisfactory, calling for more accurate methods.
Quantum Monte Carlo (QMC) techniques are promising alternatives, and although several orders of magnitude more expensive than DFT, they are methods that can naturally exploit massively parallel computers. After introducing the basics of QMC I will show a number of applications, including the energetics of hydrogen storage materials and the interaction between water and graphene, the latter important for applications such as lubrication and water desalinisation, to name a few.
If time permits I will also touch on recent developments of the method in our group, which have improved the accuracy of QMC and most notably increased its efficiency by 1-2 orders of magnitude.
 
 
 
Proponente
Arturo Tagliacozzo
 
Data: 11/10/2018