| Summary: | 博士 === 國立臺灣科技大學 === 化學工程系 === 101 === In recent years, computational material science was frequently used to develop new energy materials due to its outstanding accuracy and practicability. Much effort has been done using computational high-throughput method toward the search of new materials as well as deep understanding of physical and chemical phenomena. In this thesis, our main purpose is that using the computational approach to intensively explore the descriptors for specific materials and applications, and achieve the goal of material screening. The followings are the research topics addressed in this dissertation: (I) Computational high-throughput method to screen non-carbon support applied in methanol oxidation reaction (MOR) and oxygen reduction reaction (ORR); (II) Combined computational and experimental approach to develop the bimetallic alloying catalysts applied in H2O2 oxidation reaction.
(I) Computational high-throughput method to screen non-carbon support applied in MOR and ORR: Material screening via computational high-throughput method is an emerging area of materials science. In this study, a methodology based on density functional theory (DFT) calculations was proposed to search TiO2-based non-carbon support materials for fuel cells applications, MOR and ORR, in which the predictions of material properties were carried out by combining electronic structures and thermodynamic results. Several key requirements for the metal oxide supports are good electrical conductivity, reactivity and stability, and we can screen new potential catalyst materials from forming the intermediate states, oxygen vacancy formation energy, Pt adsorption energy and charge variation of deposited Pt. Based on the electrochemical reaction, a support material with good electrical conductivity is a vital criterion, and some anatase-TiMO2 support materials are thus selected to be the available candidates due to much improved electrical conductivity. Combination of oxygen vacancy formation energy and Pt adsorption behavior indicates the place where the strong metal-support interaction (SMSI) occurs, resulting in the variations of the deposited Pt charge. The results of the key requirements are collected by establishing the rules, and then drawing a guide map for screening the available anatase-TiMO2 support materials. Eventually, through the guide map, we can find many non-carbon supports that have the potential to enhance the activity of MOR and ORR. In addition, by direct comparison with experimental observations, we expect that first-principle computational materials screening can efficiently accelerate the design and development of the TiO2-based non-carbon catalysts.
(II) Combined computational and experimental approach to develop the bimetallic alloying catalysts applied in H2O2 oxidation reaction: Using a combined computational and experimental approach to develop a noble metal based catalyst for H2O2 oxidation reaction. A methodology based on DFT calculations was successfully proposed to search for an optimal Pt-M bimetallic catalyst, which was then verified experimentally. In this study, based on the bi-functional mechanism, a series of Pt-M bimetallic systems were first chosen as possible candidates due to the binding strength of the OH group on the second metal (M). The surface Pt d-band center (εd) and the energy barrier with respect to the dehydrogenation kinetics of the chosen bimetallic systems were calculated and used to find an optimal catalyst. A volcano-type relationship between the Pt εd and the energy barrier was found, which implied that a shift in the d-band of surface Pt atoms strongly influences the dehydrogenation kinetics of H2O2. This suggests that an appropriate Pt-based catalyst for H2O2 oxidation should correlate moderate OH adsorption with the middle of the Pt εd. From the inspiration given by the computation results, different carbon-supported Pt-M catalysts were synthesized using a modified Watanabe process and tested for H2O2 oxidation. It was found that Pt-Pd/C demonstrated excellent catalytic activity. The experimental results were in good agreement with computational predications suggesting that the methodology developed for designing Pt-based bimetallic catalysts provides a fast approach to further exploring new catalysts.
|
|---|
