Abstract:Nanomaterials and nanostructures with multi-functional properties found widespread applications such as electronics, optics, and coatings that can be fabricated using Atomic Layer Deposition (ALD). ALD is a vapor phase deposition technique to generate thin films of metals and metal oxides on a substrate. In this process, a precursor, which often comprises of organic functional groups that surround the depositing metal, chemisorbs on the substrate or reacts with the surface sites and with each other. Precursor chemisorption on the substrate leads ALD to be a self-limiting process. Thus, the precursor(s) should be chosen in a way to enhance deposition based on the ALD conditions. For a given application, it is practically impossible to carry out a large number of experiments using numerous precursors with varied deposition conditions to find the one that maximizes the growth rate in ALD. In addition, only existing precursors can be tested experimentally. The overall objective of this work is to develop a computational tool for the in-silico design of precursor materials using adsorbate solid solution theory (ASST). In the first part of this paper, we apply the ASST to derive properties of the functional groups present in the precursor using a new Group Contribution Method (GCM). The GCM is successfully derived and compared with the experimental data from ALD. The method shows good agreement and is useful for the design of novel materials. In the second part of this paper, using the thermodynamic properties as obtained from GCM, we develop a computer-aided molecular design (CAMD) framework for the optimal design of novel precursor materials with enhanced deposition properties for the ALD of metal oxides and metals. The metaheuristic efficient ant colony optimization algorithm (EACO) developed in-house is used for both parts. CAMD is a combinatorial optimization methodology, where molecules with optimal desired properties are generated from functional groups. The precursor selection optimization problem is posed as a mixed integer nonlinear programming problem, which is solved using EACO. The ALD growth kinetics is used as an objective of optimization and a solution is validated with thermodynamic constraints. Along with the number of groups, the temperature is also included in the optimization framework as decision variables. 41 novel titanium precursor molecular structures for ALD are generated with growth rates ranging from 1.23 Å/cycle to as high as 1.65 Å/cycle. Thus, these precursors have shown growth rates higher than the known titanium precursors. ALD growth rate is found to be a function of the combination of the precursor functional groups as well as temperature with a complex correlation among them.

Leveraging Bayesian Optimization Software for Atomic Layer Deposition: Single-Objective Optimization of TiO2 Layers

Bayesian optimization for chemical products and functional materials

Bayesian Optimisation for Sequential Experimental Design with Applications in Additive Manufacturing

Bayesian optimization for stable properties amid processing fluctuations in sputter deposition

Bayesian Optimization for Controlled Chemical Vapor Deposition Growth of WS2

Stoichiometric growth of SrTiO3 films via Bayesian optimization with adaptive prior mean

Process Quality Control Through Bayesian Optimization with Adaptive Local Convergence

Unlocking Potential Catalysts: A Machine Learning Approach with Bayesian and Regression Models

Human-in-the-loop Multi-objective Bayesian Optimization for Directed Energy Deposition with in-situ monitoring

Atomic Layer Deposition Process Optimization by Computational Fluid Dynamics

Navigating the Unkown With AI: Multiobjective Bayesian Optimization of Non-Noble Acidic OER Catalysts

An Integrated Method of Bayesian Optimization and D-Optimal Design for Chemical Experiment Optimization

Bayesian optimization with adaptive surrogate models for automated experimental design

A physics informed bayesian optimization approach for material design: application to NiTi shape memory alloys

Bayesian Optimization for Controlled Chemical Vapor Deposition Growth of WS 2

Bayesian optimization acquisition functions for accelerated search of cluster expansion convex hull of multi-component alloys

Optimizing the synthesis and properties of Al-modified anatase catalyst supports by statistical experimental design

Constrained composite Bayesian optimization for rational synthesis of polymeric particles

Practical Path-based Bayesian Optimization

Optimal design of novel precursor materials for the atomic layer deposition using computer-aided molecular design

Adaptive Catalyst Discovery Using Multicriteria Bayesian Optimization with Representation Learning