© Department of Materials Science and Technology, IIT Delhi
Designing Smart Reagents Through Molecular Modeling
and AI/ML Techniques
Dr . Biswajit Saha
Research & Development, Tata Steel Limited
Abstract
The integration of molecular modeling and machine learning (ML) techniques is revolutionizing mineral processing by enabling deeper insights into mineral– reagent interactions and process optimization. Molecular modeling, including density functional theory (DFT) and molecular dynamics (MD), provides atomistic-level understanding of surface chemistry, adsorption mechanisms, and reaction pathways. These simulations facilitate the rational design of reagents such as collectors, depressants, binders, and frothers tailored to specific mineral surfaces. Complementarily, machine learning leverages experimental and simulation data to uncover patterns, predict process outcomes, and optimize operational parameters. ML algorithms can be trained to predict flotation recovery, reagent efficiency, and process selectivity—significantly reducing the need for costly and time-consuming experiments. By combining the predictive accuracy of molecular modeling with the data-driven adaptability of ML, researchers and industry professionals can develop more efficient, selective, and environmentally sustainable mineral processing strategies. This synergistic approach represents a paradigm shift towards intelligent, simulation-informed mineral processing.
Bio
Dr . Biswajit Saha earned his Ph.D. in Computational Physics from the Indian Association for the Cultivation of Science, India. After completing his Ph.D., he worked at Kassel University, Germany, where he contributed to the development of
plasma models for studying atomic systems within the relativistic multi- configuration Dirac-Fock (MCDF) formalism. Subsequently, he served as a JSPS Fellow at Kyoto University, Japan, and as a Postdoctoral Fellow at Northwestern University, USA, where he extensively studied the formation mechanisms and properties of carbon-based nanomaterials using quantum mechanical methods and atomistic MD simulations. Currently, he is working as a Principal Researcher in
the Research & Development division at Tata Steel Limited. He is involved in the development of various chemicals for mineral processing applications.
Abstract
The integration of molecular modeling and machine learning (ML) techniques is revolutionizing mineral processing by enabling deeper insights into mineral– reagent interactions and process optimization. Molecular modeling, including density functional theory (DFT) and molecular dynamics (MD), provides atomistic-level understanding of surface chemistry, adsorption mechanisms, and reaction pathways. These simulations facilitate the rational design of reagents such as collectors, depressants, binders, and frothers tailored to specific mineral surfaces. Complementarily, machine learning leverages experimental and simulation data to uncover patterns, predict process outcomes, and optimize operational parameters. ML algorithms can be trained to predict flotation recovery, reagent efficiency, and process selectivity—significantly reducing the need for costly and time-consuming experiments. By combining the predictive accuracy of molecular modeling with the data-driven adaptability of ML, researchers and industry professionals can develop more efficient, selective, and environmentally sustainable mineral processing strategies. This synergistic approach represents a paradigm shift towards intelligent, simulation-informed mineral processing.
Bio
Dr . Biswajit Saha earned his Ph.D. in Computational Physics from the Indian Association for the Cultivation of Science, India. After completing his Ph.D., he worked at Kassel University, Germany, where he contributed to the development of
plasma models for studying atomic systems within the relativistic multi- configuration Dirac-Fock (MCDF) formalism. Subsequently, he served as a JSPS Fellow at Kyoto University, Japan, and as a Postdoctoral Fellow at Northwestern University, USA, where he extensively studied the formation mechanisms and properties of carbon-based nanomaterials using quantum mechanical methods and atomistic MD simulations. Currently, he is working as a Principal Researcher in
the Research & Development division at Tata Steel Limited. He is involved in the development of various chemicals for mineral processing applications.
Abstract
The integration of molecular modeling and machine learning (ML) techniques is revolutionizing mineral processing by enabling deeper insights into mineral– reagent interactions and process optimization. Molecular modeling, including density functional theory (DFT) and molecular dynamics (MD), provides atomistic-level understanding of surface chemistry, adsorption mechanisms, and reaction pathways. These simulations facilitate the rational design of reagents such as collectors, depressants, binders, and frothers tailored to specific mineral surfaces. Complementarily, machine learning leverages experimental and simulation data to uncover patterns, predict process outcomes, and optimize operational parameters. ML algorithms can be trained to predict flotation recovery, reagent efficiency, and process selectivity—significantly reducing the need for costly and time-consuming experiments. By combining the predictive accuracy of molecular modeling with the data-driven adaptability of ML, researchers and industry professionals can develop more efficient, selective, and environmentally sustainable mineral processing strategies. This synergistic approach represents a paradigm shift towards intelligent, simulation-informed mineral processing.
Bio
Dr . Biswajit Saha earned his Ph.D. in Computational Physics from the Indian Association for the Cultivation of Science, India. After completing his Ph.D., he worked at Kassel University, Germany, where he contributed to the development of plasma models for studying atomic systems within the relativistic multi- configuration Dirac-Fock (MCDF) formalism. Subsequently, he served as a JSPS Fellow at Kyoto University, Japan, and as a Postdoctoral Fellow at Northwestern University, USA, where he extensively studied the formation mechanisms and properties of carbon-based nanomaterials using quantum mechanical methods and atomistic MD simulations. Currently, he is working as a Principal Researcher in
the Research & Development division at Tata Steel Limited. He is involved in the development of various chemicals for mineral processing applications.
© Department of Materials Science and Engineering, IIT Delhi