Accelerated Sampling of Random Defects in 2D Materials

Friday, May 19, 2017 - 11:20am - 11:40am
Keller 3-180
Petr Plechac (University of Delaware)
We present a Multi-level Monte Carlo technique to accelerate sampling
for approximation of electronic structure
properties in materials with random defects.
The computational efficiency is investigated on test problems given by tight-binding models focusing on different quantities of interest (integrated density of states, density of states, current-current correlation measure).
For the chosen test problems the developed multi-level Monte Carlo estimators significantly reduce the computational time of standard Monte Carlo estimators to obtain a given accuracy. The developed acceleration method is non-intrusive and
can be paired with physics electronic structure codes allowing for sampling in more accurate models (e.g. DFT) where the size of the computational domain is insufficient for obtaining self-averaging on a single realization of the random disorder.