Atomic layer deposition (ALD) is a vapor-phase deposition technique that has attracted increasing attention from both experimentalists and theoreticians in the last few decades. ALD is well-known to produce conformal, uniform, and pinhole-free thin films across the surface of substrates. Due to these advantages, ALD has found many engineering and biomedical applications. However, drawbacks of ALD should be considered. For example, the reaction mechanisms cannot be thoroughly understood through experiments. Moreover, ALD conditions such as materials, pulse and purge durations, and temperature should be optimized for every experiment. It is practically impossible to perform many experiments to find materials and deposition conditions that achieve a thin film with desired applications. Additionally, only existing materials can be tested experimentally, which are often expensive and hazardous, and their use should be minimized. To overcome ALD limitations, theoretical methods are beneficial and essential complements to experimental data. Recently, theoretical approaches have been reported to model, predict, and optimize different ALD aspects, such as materials, mechanisms, and deposition characteristics. Those methods can be validated using a different theoretical approach or a few knowledge-based experiments. This review focuses on recent computational advances in thermal ALD and discusses how theoretical methods can make experiments more efficient.
Keywords: Monte Carlo; atomic layer deposition (ALD); computer-aided molecular design; deposition characteristics density functional theory; group contribution method; lattice Boltzmann method; mechanisms; molecular dynamics; precursors.