Learning time-dependent deposition protocols to design thin films via genetic algorithms

Abstract

Designing next generation thin films tailor-made for specific applications relies on the availability of robust process-structure-property relationships. Traditional structure zone diagrams that relate one or two deposition conditions to microstructures are limited to simple mappings, with machine-learning methods only recently attempting to relate multiple processing parameters to the final microstructure. Despite this progress, process-structure relationships are unknown for deposition conditions that vary during thin-film deposition, limiting the range of achievable microstructures and properties. We combine phase-field simulations with a genetic algorithm to identify and design time-dependent deposition protocols that achieve tailor-made microstructures. We simulate the physical vapor deposition of a binary-alloy thin film by employing a phase-field model, where deposition rates and diffusivities of the deposited species vary in time and are controlled via the genetic algorithm. Our genetic-algorithm-guided protocols achieve targeted microstructures with lateral and vertical concentration modulations, as well as more complex, hierarchical microstructures previously not described in classical structure zone diagrams. By elucidating the process-structure mechanisms during physical vapor deposition and using this knowledge to achieve precise thin-film microstructures, our algorithm provides insights to the thin film, physical vapor deposition, and film functionality communities looking for additional avenues to design novel thin-film microstructures.

Keywords

• Physical vapor deposition
• Thin films
• Genetic algorithm
• Phase-field method