3D lattice Monte Carlo modeling of morphology formation of Si/SiOx nanocomposites during phase separation of nonstoichiometric Si oxide films

TL;DR

This study introduces a 3D Monte Carlo lattice model to analyze how the morphology of Si phase in nonstoichiometric SiOx films evolves during phase separation, highlighting the effects of stoichiometry and film thickness.

Contribution

The paper presents a novel 3D lattice Monte Carlo model that incorporates atomic structure and migration dynamics to study morphology formation in SiOx films.

Findings

Isolated Si nanoparticles form at x >= 1.4.

Connected Si networks appear at x <= 0.8.

The percolation threshold shifts with film thickness.

Abstract

In this paper, a three-dimensional lattice model based on the Monte Carlo approach is presented. This model is developed to investigate the kinetics of morphology change during phase separation in nonstoichiometric Si oxide (SiOx, x < 2) films. The model takes into account the SiOx local atomic structure and probabilistic migration of oxygen atoms driven by the tendency of free energy minimization. The influence of the initial SiOx stoichiometry index x and film thickness on the morphology of the precipitated Si phase in the Si oxide matrix is analyzed. The morphology of the Si phase is shown to critically depend on the initial SiOx stoichiometry. Namely, isolated Si nanoparticles form at low excess Si content (x >= 1.4), while interconnected Si networks always appear at x <= 0.8. A dimensional effect on the morphology of the Si phase is revealed. Namely, reducing the film thickness imposes geometric constraints on the Si network formation. The percolation threshold is found to shift from xp ~= 1.35 for the bulk-like SiOx layers to xp ~= 0.85 for the quasi-two-dimensional films. The transition to the bulk material behavior is observed at a SiOx thickness of approximately 4.2 nm.