Rational Discovery of Inorganic Material Properties

Restricted (Penn State Only)
- Author:
- Katzbaer, Rowan
- Graduate Program:
- Chemistry
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 14, 2023
- Committee Members:
- Kenneth Knappenberger, Major Field Member
Zhiqiang Mao, Co-Chair & Dissertation Advisor
John Asbury, Major Field Member
Ismaila Dabo, Outside Unit & Field Member
Raymond Schaak, Co-Chair & Dissertation Advisor
Philip Bevilacqua, Program Head/Chair - Keywords:
- Catalysis
Inorganic
Chemistry - Abstract:
- Materials for energy generation and storage are essential for society. While many energy materials exist, there is always room for improvements in efficiently and sustainability. Lowering the carbon footprint of energy generation is currently a major driver behind discovery efforts for novel catalytic materials. It is hoped that chemical feedstocks, such as hydrogen, can be generated from clean energy. These catalysts can be discovered rationally, based on the desired properties. Rational property-based material design can be approached in a variety of ways. The first method that will be considered is a data driven approach, where fundamental electronic properties are investigated for their predictive power for catalytic performance. It was determined that topological band structures, with extremely fast and robust carrier mobility, is preserved after catalysis in MoTe2. This case study helped establish the importance of carrier mobility in catalysis. The electronic structure of materials is also suitable for high throughput calculations. This can rapidly narrow the field of materials that are flagged for experimental investigation. However, this method relies on being able to predict band structures accurately and the hypothesis that these correlates to the catalytic performance. We established a theory-experiment feedback loop for a data driven approach to material discovery. This was achieved through the calculation and experimental verification of photocatalyst band gaps and band edges in oxides and sulfides. Material properties can also be designed through an elemental or structural approach. One can try to combine elements known to be catalytically active into stable structures that optimize the bonding environment for a given property (eg. conductivity). This approach was utilized to design high entropy oxides for the oxygen evolution reaction, targeting elements that were predicted to offer higher activity or stability. The last approach to rational property discovery that will be covered is a statistical approach. Where the phase space is less explored and trends less known, one needs a rapid way of assessing the synthetic parameters relation to targeting a material with a given properties. These statistical methods can help design experiments to maximize the information gained from each. This is demonstrated in our optimization of the surface area of catalysts supports through an analysis of variance across several different synthetic parameters. A statistical analysis of results is also helpful for any high throughput or combinatorial synthesis, allowing trends in products and properties to be assessed in a rigorous manner. We took this approach in designing an automated program to analyze nanoparticle heterostructure population distributions. In conclusion, data-driven, structural and statistical approaches can all guild the rational discovery of material properties.