Transforming Metal-Organic Framework Discovery with Data-Driven Approaches

Tackling global challenges like water scarcity and sustainable synthesis demands innovative approaches that integrate functional materials design with AI-driven tools. In this seminar, I will first describe how I engineered porous crystalline materials, in particular metal-organic frameworks (MOFs), for atmospheric water harvesting, thereby addressing the water–energy nexus. By combining gas sorption measurements with structural characterization techniques (e.g., X-ray diffraction and spectroscopic analyses), I established key design rules for hygroscopic MOFs, optimizing pore size, working capacity, energy efficiency, and scalability. These findings led to the development of portable water-capture devices, successfully field-tested in the extreme conditions of Death Valley National Park, underscoring their real-world potential. 

In the second part, I will introduce the integration of large language models (LLMs) into closed-loop porous materials discovery and chemical synthesis planning, respectively. As a prime example of human-AI collaboration, my work has enabled efficient literature data mining, accelerated inverse design, and automated synthesis and characterization with robotics. By streamlining the exploration of synthesis-structure-property relationships, such LLM-assisted workflows not only expedite material development but also hold great promise for advancing self-driving labs, paving the way for scalable, autonomous, and sustainable solutions in materials science.