Recently, Researchers from the School of Materials Science and Engineering at Shanghai Jiao Tong University, in collaboration with City University of Hong Kong, Central South University, Suzhou Laboratory, and the Max Planck Institute for Sustainable Materials in Germany, has made significant progress in the field of AI-driven materials design. The related findings were published in Science Advances under the title "From scrap to prime: Designing high-performance alloys from dirty feedstocks with machine learning" (https://www.science.org/doi/10.1126/sciadv.aeg9183), with Shanghai Jiao Tong University as the primary affiliation.

The sustainable recycling of metals can dramatically reduce energy consumption and greenhouse gas emissions, while also lessening dependence on scarce and strategically important raw materials. However, recycled metals are highly susceptible to contamination by impurities during the recycling loop, which are difficult to remove and can lead to deteriorated mechanical properties and a strength-ductility mismatch upon remelting. Conventional strategies often address this challenge by diluting impurities with virgin feedstock, which limits scrap utilization rates, narrows the microstructural design space, and undermines the cost and sustainability advantages of recycling. To tackle the challenge of upgrading recycled metals, this study focuses on ductile cast iron and proposes a data-driven paradigm that integrates experimentation with deep neural network optimization to design new cast iron materials with high scrap compatibility. This approach achieves 100% recyclability and an 8.1% reduction in manufacturing cost, while significantly enhancing mechanical performance—ultimate tensile strength improved by up to 15.2% and fracture elongation by up to 73.2%—thereby resolving the strength-ductility trade-off and enabling the high-value utilization of recycled ductile cast iron. By endowing alloys with a practical "scrap-compatibility gene," this strategy offers a new avenue for alloy design and demonstrates, to a certain extent, the potential for the sustainable perpetual recycling of metals.

Dr. Chao Yang, Assistant Professor at the School of Materials Science and Engineering, Shanghai Jiao Tong University, and Dr. Bo Peng from City University of Hong Kong are the co-first authors. The corresponding authors are Assistant Professor Ye Wei from City University of Hong Kong, Professor Liuliu Han from Central South University, and Professor Dierk Raabe from the Max Planck Institute for Sustainable Materials in Germany. Other contributing authors include Professor Jian Lu (Academician) from City University of Hong Kong, Professor Da Shu from the School of Materials Science and Engineering at Shanghai Jiao Tong University, Professor Binhan Sun from East China University of Science and Technology, and Professor Lijun Zhang from Central South University. This work was supported by the 2030 National Major Science and Technology Project, the Shanghai Natural Science Foundation, and the "Science and Technology Revitalizing Inner Mongolia" Special Project of Shanghai Jiao Tong University.