UC Irvine

This dissertation advances the field of catalyst synthesis by employing flame spray pyrolysis (FSP) to create nanomaterials with highly controlled and tunable properties. FSP enables precise adjustment of particle size, phase composition, and morphology, which are critical factors in optimizing catalysts for applications such as methane oxidation, water electrolysis, and energy storage. A primary focus of this work is the development of a real-time quality control mechanism during FSP synthesis, achieved by integrating machine learning models trained on laser-induced breakdown spectroscopy (LIBS) data. The balanced accuracies of making a correct prediction of phase information can reach up to 0.89 for tetragonal ZrO2, tetragonal α-MnO2, tetragonal β-MnO2, and tetragonal Mn3O4 phases. The lattice constants and OV %s of the testing samples can be predicted with root-mean-squared errors of 0.04 and 0.05, respectively. This innovative approach allows predictive control over catalyst properties in situ, reducing the need for time-intensive post-synthesis characterization and accelerating catalyst discovery.

Our research demonstrates that key FSP parameters, including oxygen flow rate, liquid feed rate, and precursor concentration, significantly influence catalytic efficiency by modulating factors such as oxygen vacancy (OV) content, metal-support interactions, and phase stability. The oxidizing environment promotes the formation of incorporated Pd2+ structures, while highly dispersed Pd2+ and Pd0 nanoparticles are formed under the reducing synthesis condition. Our findings suggest that the catalysts containing both highly dispersed Pd2+ nanoparticles and incorporated Pd2+ species exhibit superior methane oxidation activity than those containing only one type of Pd2+ structure. In Pd/CexZr1-xO2 catalysts, higher OV% is associated with weaker metal-support interaction and the formation of peroxide species (O22−), while the equivalence ratio dictates Pd speciation. The turnover frequencies of Pd/CexZr1-xO2 catalysts can be adjusted from 0.10 s-1 to 0.21 s-1 by modifying synthesis conditions in the FSP. These results highlight the synergistic interaction among PdO nanoparticles, oxygen vacancies, and peroxide species in improving methane oxidation activity. Additionally, we introduce combustion enthalpy density as a comprehensive synthesis parameter, providing refined control over specific surface areas and morphologies, particularly in the scalable production of silica nanoparticles. This parameterization aids in balancing nanoparticle size with surface properties for applications in thermal insulation and catalysis. Collectively, these findings position FSP as a highly versatile and industrially scalable method for designing advanced catalysts, offering a pathway toward more efficient catalyst production for environmental and energy applications.