Research
Large Time-Scale and Length-Scale Modelling of Lithium Battery
Our research is dedicated to unraveling the complex chemical reaction mechanisms and dynamics within Li-S batteries. We are particularly focused on understanding the stable structural morphologies, electronic structures, and the evolution of chemical bonding throughout the reactions. Another aspect of our work is centered on exploring the recycling process, emphasizing polymer interactions among binders, alkaline reagents, and electrodes. This research is poised to provide valuable insights for achieving high-efficiency recycling rates in battery technology. Furthermore, we are investigating the formation mechanism of lithium dendrites. To understand the shape morphology, a critical concern for lithium battery safety, we employ Molecular Dynamics (MD), Monte Carlo (MC), and coarse-grain methods. This comprehensive approach aims to contribute to advancements in lithium battery safety and overall performance.
Advancing Computational Modeling and Design for Catalysts under Realistic Conditions
Computational modeling plays a pivotal role in unraveling the intricate mechanisms and enhancing the efficiency of catalysts. However, comprehensively discerning the local environment surrounding a catalytic center under authentic conditions, encompassing surface engineering, thermalization, solution effects, electric potentials, catalyst dynamics, revolutions, and the operando characterization of structural changes during catalytic reactions, remains a formidable challenge. Our research group is dedicated to modeling real conditions for a diverse array of catalysts, encompassing cluster-based heterogeneous catalysts, enzymatic catalysts, and supramolecular capsules. This profound understanding of the real nature of the active sites is paramount in the design of high-performance catalysts.
Electronic Structure, Spectroscopy and Chemical Bonding of Size-Selected Metal (3d/4f/5f) Nanoclusters and Nanoalloys
Nanoclusters offer valuable and systematic means to investigate the electronic structure, chemical bonding, and catalytic mechanisms of catalysts and materials. The evolving geometries and increasing sizes of clusters provide insights into their growth mechanisms under different experimental conditions, including confinement on specific surfaces. Within our research group, one of our focal projects involves exploring transition metal, lanthanide, and actinide clusters. Our aim is to unravel intriguing chemical bonding phenomena and spectroscopic characteristics (such as vibrational, electronic, and dynamics of excited states). These investigations lay a theoretical foundation for catalyst design and hold promise for developing novel catalytic materials.
Design of Boron-Based Self-Assembly Nanomaterials
The recent discovery of two-dimensional boron materials, known as “borophene,” similar to carbon’s “graphene,” has led to increased interest in exploring new boron-based materials. Boron clusters, which have received less attention than carbon clusters, present unique challenges due to boron’s electron deficiency and the diverse structures that arise as the clusters grow. Highly symmetrical clusters often serve as building blocks for material design. Therefore, studying the stability, geometry revolution, electronic structure, and chemical bonding of metal-doped boron clusters, particularly those with lanthanide and actinide elements, is crucial. This research can unlock the potential of boron-based materials with exceptional magnetic, optical, and catalytic properties, leading to significant technological advancements.
Machine Learning of Atomic Potential for Study of Descriptor for Materials and Catalyst Design
In this specialized area, our research group is dedicated to harnessing the capabilities of machine learning to advance our comprehension of atomic potentials, encompassing energy and force. By seamlessly integrating these potentials into force fields with chemical accuracy, we can conduct multiscale calculations, employing techniques like ML/MM (machine learning/molecular mechanics), for molecular dynamics or metadynamics simulations across extensive time and length scales. Moreover, we are committed to identifying informative fingerprint descriptors, leveraging insights from quantum chemistry and molecular dynamics, which prove instrumental in designing catalysts with enhanced stability and selectivity. The amalgamation of machine learning techniques and domain-specific expertise holds the promise of expediting the discovery of efficient catalysts and materials, endowed with tailored properties.