Addressing accuracy by prescribing precision: Bayesian error estimation of point defect energetics

What is it about?

Simulation of the properties of point defects is a well-established field in materials physics and techniques have improved drastically in recent decades. However, little attention has been paid to uncertainty quantification in these simulations – in other words, what are the error bars on the quantities we calculate? In this paper, we explore one approach for assigning these error bars to results of point defect calculations.

Featured Image

Photo by Iewek Gnos on Unsplash

Photo by Iewek Gnos on Unsplash

Why is it important?

Point defects are crucial to many kinds of material behavior. For example, we deliberately introduce special defects (called "dopants") in semiconductors, which is what enables their special behavior that we take advantage of in technologies from microchips to solar cells. Better quantifying our uncertainty in predictions of defect properties will aid in translation theory and simulation work to practice in the lab and, eventually, to technological impact.