当前位置：Generative Design of Inorganics and Rapid Experimental Validation via Direct Joule-Heated Synthesis
Generative Design of Inorganics and Rapid Experimental Validation via Direct Joule-Heated Synthesis
Materials discovery is fundamental to advance next-generation technologies as well as for sustainable and circular economy. Beyond computational screening, generative models are efficient at finding materials with desired properties. Here, we introduce a Materials Generative Design and Testing (Mat-GDT) framework. Mat-GDT integrates generative design with accelerated validation through high-throughput simulations and experiments. The central idea of the framework is constructed around a foundation AI model for inorganic materials, which enables Mat-GDT to solve grand challenges in sustainable design of functional materials. We demonstrate the various parts of this framework including a symmetry-aware material representation, Wyckoff based transformer [1] for property-directed generative design [2], a DFT framework for relaxation and phase analysis [3], followed by rapid non-equilibrium experiments through direct joule-heating synthesis (DJS) coupled with machine learning optimization. We showcase this for doped inorganic thermoelectric materials, demonstrating Se/Cd co-doped AgSbTe2 with zT ~2.3 at 575K [4].
[1] Nikita Kazeev et. al. Wyckoff Transformer: Generation of Symmetric Crystals. https://arxiv.org/abs/2503.02407
[2] Shuya Yamazaki et. al. Multi-property directed generative design of inorganic materials through Wyckoff-augmented transfer learning. https://arxiv.org/abs/2503.16784
[3] Wei Nong et. al. CrySPR: A Python interface for implementation of crystal structure pre-relaxation and prediction using machine-learning interatomic potentials. https://doi.org/10.26434/chemrxiv-2024-r4wnq
[4] Chenguang Zhang et. al. Direct Joule-Heated Non-Equilibrium Synthesis Enables High Performing Thermoelectrics. https://arxiv.org/abs/2506.04447Kedar Hippalgaonkar received his Bachelor of Science in Mechanical Engineering from Purdue University in 2006 and completed his PhD in Mechanical Engineering at the University of California, Berkeley in 2014. He is currently an Associate Professor at the School of Materials Science and Engineering at Nanyang Technological University (NTU), Singapore, and a Principal Scientist at the Institute of Materials Research and Engineering (IMRE), A*STAR. His research focuses on the integration of artificial intelligence, high-throughput experimentation, and solid-state physics for the discovery and optimization of functional materials. He leads several multi-PI research programs as lead PI, including the AxCIS (AI xploration of Catalyst Inorganic Surfaces) Lab for catalyst discovery and the Materials Generative Design and Testing (Mat-GDT) Framework for property-directed generative design and foundation models of functional inorganic materials.
He has received several awards, including the National Research Foundation Fellowship (NRFF) and the IES Sustainability Award, and serves as Associate Editor of RSC Digital Discovery , guest editor at Chemical Reviews and Frontiers in materials discovery themed collection at RSC journals. In addition to his academic roles, he is co-founder and Senior Scientific Advisor of Xinterra Pte Ltd, a company focused on AI-enabled materials development. His contributions span thermoelectric materials, machine learning algorithms for materials design, and development of self-driving laboratories.
