AI Agent Expert Interview Series - Dr. Donglin He
May 10, 2026, the Editorial Office of AI Agent had the pleasure of interviewing Dr. He from The Hong Kong University of Science and Technology (Guangzhou), whose research focuses on deep learning, time series forecasting, and intelligent control in energy systems.
In this interview, Dr. He explored how artificial intelligence is reshaping the discovery and design of metal-organic frameworks (MOFs). She discussed AI's growing role in accelerating structure prediction, property optimization, and materials screening, while emphasizing that AI should be viewed as a complementary partner to experiments and theoretical modeling rather than a replacement. Drawing on her research experience in algorithm-driven materials discovery, Dr. He highlighted the value of closed-loop systems that connect machine learning models with experimental feedback, enabling more efficient navigation of the vast MOF design space. She also addressed critical challenges, including limited high-quality data, model interpretability, and the gap between computational predictions and experimental realization. Looking ahead, she shared her vision for AI-enabled MOF research in energy, catalysis, and gas separation applications, and identified autonomous experimentation, explainable AI, and physics-informed learning as key directions that could drive the next generation of intelligent materials discovery.
Interview Questions:
Q1. Your research focuses on algorithm-driven materials discovery. How do you view the role of artificial intelligence in MOF research? How does it relate to traditional experimental and theoretical approaches?
Q2. In MOF research, AI has been applied to areas such as structure prediction and performance screening. In your view, which specific aspects hold the greatest potential for improving research efficiency?
Q3. In MOF research, the experimental parameter space is often highly complex. Your work emphasizes algorithm-driven experimental exploration—could you elaborate on how AI integrates with experiments to form a closed-loop system, thereby improving the efficiency of materials discovery?
Q4. AI for MOF still faces challenges such as data quality, limited interpretability, and gaps between theory and experiment. In your opinion, what is the most critical bottleneck, and how might it be addressed?
Q5. In applications such as gas adsorption, catalysis, and energy conversion, how far do you think AI-driven MOF design is from practical implementation? What are the key breakthroughs needed?
Q6. Looking ahead, AI for MOF is still rapidly evolving. Based on your research background, do you plan to further explore this direction? Which areas or problems do you consider most promising?
About the Interviewee:
Dr. Donglin He joined the Sustainable Energy and Environment Thrust at the Hong Kong University of Science and Technology (Guangzhou) as an Assistant Professor and doctoral supervisor in September 2025.
She completed her Ph.D. in Chemistry at the University of Liverpool (2022), supervised by Professor Andrew I. Cooper (Foreign Member of CAS, FRS, MAE). From 2022 to 2023, she worked as a Research Associate in Professor Leroy Cronin's group at the University of Glasgow, UK. Subsequently, she joined Professor Shuhei Furukawa's group at Kyoto University, Japan, as a JSPS Postdoctoral Fellow.
Her research focuses on the design, synthesis, and application of crystalline porous materials. She is committed to structural optimization, function-oriented synthesis, and algorithm-driven robotic discovery to achieve critical applications in energy and environmental fields, such as gas/vapour separation, pollutant removal, and resource recovery.
As of now, Dr. He has published 15 SCI papers, including 8 as first or co-first author in prestigious international journals such as J. Am. Chem. Soc., Angew. Chem. Int. Ed., Chem. Eur. J., and Chem. Commun.
Editor: Wen Xue
Production Editor: Ting Xu
Respectfully Submitted by the Editorial Office of AI Agent
