The Xin Group, led by Professor Hongliang Xin in the Department of Chemical Engineering at Virginia Tech, is a research team dedicated to advancing computational catalysis through multiscale modeling, machine learning, and AI-driven materials discovery. The group's work focuses on understanding dynamiX at interfaces—structural and electronic dynamics in nanomaterials—for applications in energy, electronics, and sustainable technologies. With expertise in ab-initio calculations, kinetic simulations, and interpretable machine learning, the Xin Group bridges computational and experimental gaps in heterogeneous catalysis. Professor Xin is a pioneer in agentic science, co-authoring a landmark commentary that positions AI agents as transformative tools for autonomous scientific discovery. This report provides an overview of the group's structure, research activities, and achievements, with emphasis on their contributions to agentic science.
The Xin Research Group is housed in the Department of Chemical Engineering at Virginia Tech, a leading institution for engineering innovation. Established in 2014 under the leadership of Professor Hongliang Xin, the group operates from 271 Goodwin Hall, Blacksburg, VA, with contact details including phone at (540) 231-6156 and email at hxin@vt.edu. Professor Xin, who holds a PhD in Chemical Engineering from the University of Michigan (2011), an MSc from Tsinghua University (2005), and a BSc from Tianjin University (2002), advanced from Assistant Professor (2014–2020) to Associate Professor (2020–2025) and was promoted to Full Professor in 2025. His research has garnered over 12,425 citations on Google Scholar, reflecting expertise in catalysis, quantum chemistry, machine learning, and agentic AI.
The group comprises a small, dynamic team including one postdoc (Hao Deng), three PhD students (Xiangrui Wang, Raul Diaz, Shuyi Cao), and one undergraduate researcher (Melina Panda). No visitors or detailed alumni are currently listed on the group's website. The Xin Group maintains an active presence on X under @XinGroup_VT, where they share updates on research, including calls for postdocs in agentic AI for computational catalysis. Collaborations extend to multidisciplinary initiatives, such as workshops on AI in catalysis and partnerships with institutions like Carnegie Mellon and MIT.
The Xin Group's research portfolio centers on multiscale modeling to understand and design nanoscale materials for energy and electronics applications. Core areas include:
Recent 2025 publications highlight these areas, including "Knowledge graphs in heterogeneous catalysis: Recent advances and future opportunities" (Chinese Journal of Chemical Engineering), "Examining generalizability of AI models for catalysis" (Journal of Catalysis), and "Dissolved Fe species enable a cooperative solid–molecular mechanism for the oxygen evolution reaction on NiFe-based catalysts" (Nature Catalysis). The group's work addresses real-world challenges in sustainability, leveraging AI to accelerate discovery.
Agentic science represents a core interest for the Xin Group, embodying a shift toward AI-driven autonomous research. In a September 2025 commentary in Nature Machine Intelligence, co-authored by Hongliang Xin, John R. Kitchin (Carnegie Mellon), and Heather J. Kulik (MIT), agentic science is defined as a paradigm where AI agents semi-autonomously reason, plan experiments, and interact with digital and physical environments. Xin describes it as an "orchestra" where AI agents act as sections of musicians, with human scientists as conductors, enabling richer discoveries.
This differs from traditional science, where humans operate like solo musicians, by automating tedious tasks and generating novel hypotheses. Impacts are anticipated in environmental stewardship, health, and self-driving labs, where AI serves as the "brain" for robotics and cloud automation. Xin highlights the timeliness due to large language models' ability to digest vast literature and the rise of automated labs. Challenges include AI hallucinations, lack of standardization, and resource demands, addressed through human oversight ("person in the loop") and ethical safeguards. Cross-institutional collaboration is emphasized to avoid silos and enhance synergies. Xin's invited talks, such as at NAM 29 (2025), underscore agentic AI's role in accelerating catalytic materials research.
The Xin Group exemplifies cutting-edge research at the nexus of chemical engineering, AI, and catalysis, with agentic science set to redefine autonomous discovery paradigms. By pioneering AI agents in self-driving labs, the group tackles global challenges in energy and sustainability while advocating responsible AI integration. Future endeavors may further expand agentic applications in catalysis. For more details, visit the official website at https://xingroup.org/.