A simple model for the growth of polycrystalline Si using the kinetic Monte Carlo simulation

Abstract

An extension to the kinetic Monte Carlo simulation technique was developed in order to study thin film deposition and growth of a system approximating polycrystalline silicon. This method was developed to determine the effect of varying the angle of incidence of an atomic beam on the morphology of a poly-Si thin film grown on a crystalline Si substrate. This deposition procedure produced material comprised of individual grains, all with identical orientation; a first step towards modelling poly-Si. The addition of such grains does not significantly affect the bulk film properties relative to the single crystal case. The number of initial grains chosen to represent a set of pre-existing grains on the surface does not affect the gross morphology of the grown film once around 40 monolayers have been deposited. The chief advantage of this polycrystalline-like system is that it allows the observation of both columnar growth (at angles below about 65 •) and dendritic growth at angles above this value; growth of single crystalline material only shows the latter. This fact allows a comparison of results from atomic-scale simulation to existing theories that relate the angle of the morphological features of the grown film to the angle of the incident beam. We show that the simulation data are not particularly well represented by commonly used theories such as the tangent rule, or that due to Tait et al (1993 Thin Solid Films 226 196). Increased angles of incidence cause faster extinction of grains until a steady-state value of the number of grains is reached. When grains are nucleated on a heterogeneous substrate, here chosen as a crude description of Si on glass, increased substrate temperature results in larger grains, and higher angles of incidence result in fewer nucleated grains due to non-local shadowing.

Large artificial anisotropic growth rate in on-lattice simulation of obliquely deposited nanostructures

“…for growth evolution of thick film). Various versions of on-lattice simulation of OAD has been developed and used by different research groups 9,[12][13][14][15][16][17][18][19][20][21][22] .…”

A hybrid kinetic Monte Carlo method for simulating silicon films grown by plasma-enhanced chemical vapor deposition

“…49,50 However, such an approach does not quantitatively account for void formation; in addition, only deposition times spanning less than 100 monolayers can be tracked. 20,[51][52][53][54][55][56] Ruan and Schun 57 have recently proposed an efficient KMC method which includes diffusion processes and allows for the formation of voids and vacancies during film growth. The method is capable of simulating nanocrystalline film with kinetic roughening, vacancy entrapment, grain nucleation and grain evolution.…”

Stochastic modeling of film porosity in thin film deposition

“…A ballistic deposition model is chosen to simulate the evolution of film porosity, which allows vacancies and overhangs to model the microstructural defects in the thin film. Microscopic models with similar rules and structures are reported in simulation works for deposition processes of a variety of materials [12], [24], [25]. The film growth model used in this work is an on-lattice kMC model in which all particles occupy discrete lattice sites.…”

“…Both monocrystalline and polycrystalline kMC models have been developed and simulated [12], [25]. Despite recent significant efforts on modeling of surface roughness, a close study of the existing literature indicates the lack of general and practical methods in the area of modeling thin film porosity using SDEs.…”