High-performance atomistic modeling of optical thin films deposited by energetic processes

Abstract

In this paper we present a computationally effective approach to classical molecular dynamic simulation of thin film growth with orientation on cluster supercomputing facilities. The goal of the developed approach is to investigate structural heterogeneities of thin films deposited on substrates at a nanoscale level. These heterogeneities depend on the experimental conditions of a deposition process being used. They have essential influence on practical properties of thin films and their modeling is important for achieving further progress in thin film optical technology.The presented research is focused on silicon dioxide thin films growth. A special force field, oriented on the atomistic description of the silicon dioxide deposition on fused silica substrate, has been developed and applied to the molecular dynamic simulation with the GROMACS package. The validity of the developed simulation approach is verified using atomic clusters consisting of up to 10 6 atoms and having characteristic dimensions of up to 30 nm. Its computational efficiency is tested using up to 2048 cores. The dependence of achievable efficiency on model parameters is discussed.

High-performance Full-atomistic Simulation of Optical Thin Films

“…The time of MD simulation of one injection cycle depends linearly on the number of deposited atoms N at and the total time of the deposition process simulation is proportional to the square of the N at [2]. This is a reason for a relatively high computational cost of the simulation procedure as compared to the MD simulation of clusters consisting of fixed numbers of atoms.…”

“…The simulation of the silicon dioxide thin film deposition is organized as a step-by-step procedure [2][3][4][5]. At every deposition step, SiO 2 groups are injected at the top of the simulation box and are directed to the film surface.…”

“…All atoms inside the simulation box move in agreement with the Newton laws. Their interactions are described in the frame of the DESIL force filed [2] that was specially developed for the simulation of the deposition process. The research is carried out using the equipment of the shared research facilities of HPC computing resources at Lomonosov Moscow State University.…”

“…The research is carried out using the equipment of the shared research facilities of HPC computing resources at Lomonosov Moscow State University. Simulation of the deposition process is performed using parallel KUVALDA program [2], for MD part of the algorithm the GROMACS package [6] is used.…”

“…At the same time the progress in high performance computing allows one to perform the investigation of thin films properties on the atomistic level using the classical molecular dynamics (MD) simulation [1]. The MD-based approach for the simulation of deposition of thin films clusters with thicknesses up to hundreds of nanometers was recently developed and was successfully applied to the detailed investigation of silicon dioxide films [2][3][4][5].…”