Machine learning predictions of surface migration barriers in nucleation and non-equilibrium growth

Abstract

Machine learning is playing an increasing role in the discovery of new materials and may also facilitate the search for optimum growth conditions for crystals and thin films. Here, we perform kinetic Monte-Carlo simulations of sub-monolayer growth. We consider a generic homoepitaxial growth scenario that covers a wide range of conditions with different diffusion barriers (0.4–0.55 eV) and lateral binding energies (0.1–0.4 eV). These simulations are used as a training data set for a convolutional neural network that can predict diffusion barriers and binding energies. Specifically, a single Monte-Carlo image of the morphology is sufficient to determine the energy barriers with an accuracy of approximately 10 meV and the neural network is tolerant to images with noise and lower than atomic-scale resolution. We believe this new machine learning method will be useful for fundamental studies of growth kinetics and growth optimization through better knowledge of microscopic parameters.

Cocrystal Growth in Organic Semiconductor Thin Films: Simulation of Pentacene, Perfluoropentacene, and Their 1:1 Blend Deposited On Graphite

“…It is important to stress that the process of diffusion on formed flat terraces, while considered an important factor for adatoms in inorganic crystal growth, [ 58 ] it is often overlooked in kinetic models of organic crystal growth (see, e.g., ref. [59]), probably because in the typical case of films of standing molecules, barriers are low [ 60 ] and the process is not the rate determining step ruling the growth mechanism.…”

Spiral growth of multicomponent crystals: theoretical aspects

“…In (Elts et al, 2017), this approach was applied to the growth of multicomponent organic crystals. Furthermore, machine learning methods are being actively developed and hold promise for various applications, such as the search for stable crystalline structures (Kang et al, 2022), determination of kinetic parameters of adatoms on the surface (Martynec et al, 2021), and optimization of growth chambers (Schimmel et al, 2022). These new digital tools provide extensive information for the theoretical understanding of growth in multicomponent systems while also offering opportunities to discover new collective phenomena and effects and study them.…”

The Crystallization of Disordered Materials under Shock Is Governed by Their Network Topology

“…Note that because of the existence of thermal fluctuation, the local composition and structure around the nucleus are varied during the nucleation process. Recent simulations show that the local compositional and structural fluctuation can play an important role in the nucleation process, [ 30 , 31 ] even for the stoichiometric composition. As the composition of residual liquid keeps changing during the off‐stoichiometric nucleation process, its compositional fluctuation can be much more dramatic compared with the stoichiometric one.…”