ML-enabled boosting of growth simulations

Carefully controlling the morphology of films/nanostructures during epitaxy is of paramount importance, particularly in view of quantum applications. Yet, this goal is often achiveved based on the grower experience which, in turns, is built largely on trial and errors. There are two reasons for this: first, growth is a complex process and oversimplified models simply do not work. Second, time scales are “human” (minutes, hours), and, as such, typically not reachable by computational approaches good enough to yield a sufficiently good description of growth. Here we shall discuss how some advanced neural-network (NN) architectures, such as convolutional NN and recurrent NN, can be conveniently exploited to yield accelerated simulations of film/nanostructures growth. After having illustrated some relevant examples on strained growth and faceted growth described at the continuum level with deterministic approaches, the possibility to tackle fluctuations at the atomic scale will be introduced by investigating step dynamics at surfaces via Generative Adversarial Newtorks. The talk shall end with a final, critical discussion on the possibility to use experimental data to directly train ML models.