(147k) Machine Learning-Based Prediction of Thermal Conductivity in Metal-Organic Frameworks

Metal-Organic Frameworks (MOFs) have gained significant attention due to their unique properties and potential applications in gas storage, separation, and catalysis. However, managing the thermal energy generated during the exothermic adsorption process remains a challenge, limiting the practical utilization of MOFs. My research interest lies in unraveling the intricate structure-thermal conductivity relationships in MOFs and exploring strategies to control and tailor their thermal transport properties for specific applications.

To achieve this goal, I am interested in employing a multidisciplinary approach that combines computational simulations, high-throughput screening, machine learning, and deep learning techniques. Classical molecular dynamics simulations allow for a detailed investigation of the thermal transport properties of MOFs by studying the behavior of phonons, the heat-carrying vibrations. By understanding the mechanisms that govern thermal conductivity in MOFs, we can identify key structural characteristics, such as pore size, density, surface area, and metal node connectivity, that influence thermal transport.

Additionally, high-throughput screening of a diverse range of hypothetical MOFs using computational methods provides insights into the relationship between structure and thermal conductivity. By exploring a vast design space of MOFs, considering important compositional and structural features, we can identify MOF structures with enhanced thermal conductivity and uncover the underlying design principles.

Furthermore, the application of machine learning techniques, such as graph convolutional neural networks, can accelerate the discovery and design process by predicting the thermal conductivity of MOFs. This approach enables us to explore a wider range of MOF structures and compositions, reducing the reliance on computationally expensive simulations.

In addition to understanding the impact of structure on thermal conductivity, I am also interested in investigating the role of defects in MOFs. By studying the influence of defects, such as missing linker and missing cluster defects, on thermal transport, we can uncover strategies to enhance or suppress thermal conductivity in MOFs. This knowledge will enable us to engineer MOF materials with tailored thermal properties for specific applications.

In summary, my research interest focuses on elucidating the structure-thermal conductivity relationships in MOFs and developing strategies to control and tailor their thermal transport properties. By combining computational simulations, high-throughput screening, and machine learning techniques, I aim to contribute to the fundamental understanding of MOFs' thermal behavior and pave the way for the design of MOFs with enhanced thermal properties for a wide range of applications.