Investigation of process–structure–property correlations for plasma-enhanced chemical vapor deposited complex disordered solids: towards data-driven computational materials design
No Thumbnail Available
Authors
Meeting name
Sponsors
Date
Journal Title
Format
Subject
Abstract
Complex disordered solids, characterized by multicomponent, multiphase, sub-structured configurations, coupled with a lack of long-range order, represent an application-rich class of materials, exhibiting a vast number of degrees of freedom. However, their existence in an indeterminate number of non-equilibrium meta-stable states, varying with fabrication processes, and the lack of experimental characterization and atomistic modeling methods capable of capturing potential configurations, pose a unique challenge to their structure elucidation and thus design. In the era of artificial intelligence applied to materials science, advances demand an integrated research methodology encompassing experiment, computation, and theory. In this context, the objectives of this dissertation were two-fold: to understand the process–structure–property correlations of selected complex disordered solids and to curate an experimental data repository for these materials to facilitate data-driven computational materials design. Molecular-precursor-derived amorphous hydrogenated silicon carbonitride (a-SiCN:H) films fabricated through plasma-enhanced chemical vapor deposition (PECVD) have been studied as a representative class of complex disordered solids. The role of precursors in dictating the substructures, composition, and properties of a-SiCN:H films was investigated. A full factorial experimental design, coupled with statistical analysis, was performed to understand the influence of PECVD parameters—substrate temperature, plasma power, pressure, and flow rate—on composition and properties, with the additional goal of curating process–structure–property datasets for hexamethyldisilazane-based a-SiCN:H films. Investigations revealed that the composition and properties of the films, such as refractive index and bandgap, varied significantly with PECVD parameters—notably, the substrate temperature and the plasma power. Linear models were formulated based on PECVD parameters to estimate composition and properties within the constraints of factorial design. Structure elucidation of a-SiCN:H films at varied atomic lengths was conducted by infrared, X-ray photoelectron, and solid-state nuclear magnetic resonance spectroscopies, complemented by neutron diffraction. SiCN:H films were identified as amorphous with distinct local order and weak correlations within 3 Å distances. Experimental process–structure–property datasets curated in this dissertation have been contributed to the open-source database, the Materials Project. These experimental inputs will guide the atomistic modeling algorithms to accurate structural models for complex disordered solids. Advancements in modeling and the incorporation of artificial intelligence will have exceptional implications for the predictive design of materials.
Table of Contents
Introduction -- Materials and methods -- Influence of plasma-enhanced chemical vapor deposition parameters on hexamethyldisilazane-derived amorphous hydrogenated silicon carbonitride: a statistical evaluation based on full factorial design -- Charting the structure evolution of amorphous hydrogenated silicon carbonitride with plasma-enhanced chemical vapor deposition parameters -- Catalyzing computational discovery: contributing experimental data to open-source database -- Conclusion and future work
DOI
PubMed ID
Degree
Ph.D. (Doctor of Philosophy)