(4av) Real-Time Computational Modelling Based on Machine Learning and External Electric Fields for Enhancing Catalyst Performance Towards Selective Product

Experience in Strategy and Approaches: During my PhD tenure in IIT Indore, I have gained experience in homogeneous catalysis, heterogeneous catalysis, solar cell and machine learning. Majorly, I have performed CO₂ hydrogenation reactions using DFT and applied various machine learning based approaches for screening a huge number of catalysts to obtain the active and selective catalyst. My postdoctoral research at Stanford University focused on developing improved methods for designing heterogeneous catalysts. We have developed a coordination-based α-scheme model that predicts catalytic activity as well as stability of the catalyst.

The more information regarding my research experience, research interests and publications can be found here.

Overview: Catalysis is crucial in industrial processes. However, there is a need for a deeper understanding of surface phenomena and atomic-level chemical reactions. Density functional theory (DFT) calculations provide valuable insights into catalyst activity and stability. Developing computational tools to comprehend complex reactions like dehydrogenation, hydrogenation, and hydrocarbon combustion is imperative. To address the urgent issue of climate change, it is imperative to transition a significant portion of our energy resources to renewable forms. In this regard, catalysis plays a crucial role. Nevertheless, many of these catalytic processes predominantly employ noble metals, which are not only costly but also scarce in availability. Consequently, it is of paramount importance to discover efficient catalysts that utilize abundant earth metals. Scientists have been dedicatedly working in this direction for decades, to uncover cost-effective alternatives to the currently commercialized catalysts. However, the process of discovery is time-consuming, both in experimental and theoretical terms. To address these challenges, researchers are using machine learning and training the dataset to speed up the process. However, several issues exist in this context, such as the availability of data, the fundamental understanding of structure-property relationships, requirement of training data to predict various catalytic properties, selectivity issue of the products and dynamics of the catalyst under reaction condition. I aim to address these issues in my laboratory by dividing into three sections as follows (Figure 1):

1. Machine learning aided catalyst screening: Recently, machine learning and data science is helping to screen huge number of catalysts for various reactions. However, the reason behind obtained results is not clear. If, we can somehow know the reason behind any machine learning properties and link those with the structure we can get a structure property relationship very easily. Therefore, in this thrust of my research, I will investigate the structure property relationship using machine learning and understand the obtained results using physics-based approached to obtain an effective descriptor for catalytic reaction. The step to achieve the above-mentioned research will be as follows:

(i) gather homogeneous and heterogeneous catalysis data from open-source websites (materials project, OQMD, AFLOWLIB, NoMaD cathub etc.) or generate data using computational resources for catalysis (ii) identify important features based on structural, energytical and electronic descriptors such as identity of the metal, ligand sphere, HOMO-LUMO and so on for homogeneous catalysis whereas coordination number, generalized coordination number, d-band center, surface site stability, orbital wise coordination number and so on for heterogeneous catalysis (iv) train vs test validation with the feature importance and machine learning algorithm and find out the correlation among various features and explain the feature importance using physics (v) predict for new catalytic systems and explain using basic feature importance and find out the physics-based model to establish structure property relationship (vi) among all of these find out the best catalytic system and test using DFT and finally validate with experimental collaborators.

2. Oriented external electric field modulated catalysis: Accelerating the catalytic process is crucial in catalyst discovery, alongside directing reactions towards specific products. However, conventional catalyst discovery approaches face limitations in enhancing both catalytic activity and selectivity. For instance, linear scaling relationships often correlate adsorbates, complicating the identification of selective products. To address this challenge, I propose employing an oriented external electric field approach to modulate specific bonds, particularly the transition state intermediates crucial for desired product formation. By applying an external electric field aligned with the transition state structure, we can selectively weaken bonds towards targeted products, thereby enhancing selectivity by lowering activation energies. The steps are as follows:

(i) identify the most important catalyst for any reaction and find out the rate determining step and try to mitigate the energy of the transition state using oriented external electric field. Let’s discuss this with and example such as CO₂hydrogenation reaction in homogeneous Mn(I)-PNP based catalysis where heterolytic hydrogen cleavage is the rate determining step. The increment in the dipole moment of the H-H bond will decrease the activation energy of the rate determining step. The same will be used in heterogeneous catalysis for various useful reaction such as CO₂ reduction reaction on Cu catalysts and nitrogen reduction reaction on Fe-based catalysts.

3. Dynamics of catalysis under operating system: It is widely acknowledged that the synthesis of an efficient catalyst is not sufficient to address energy-related challenges. The stability of a catalyst during a reaction can be compromised by various factors such as temperature, pressure, pH, reaction intermediates and so on. These aspects are equally critical to consider when striving to develop practical and reliable solutions for energy-related issue. To delve deeper into this aspect, I will center our attention on the dynamics of catalysts during reactions. The steps are as follows:

(i) This process is coupled with various factors, but we would like to decouple all the factors in cohesive manner in which we can understand the effect the each parameter, (ii) consider any initial structure based on Wulff construction such as cube nanoparticle, (iii) consider the diffusion barrier of each and every atom, (iv) effect of particle evolution based on temperature, pressure, pH, reaction intermediates and so on.

Teaching Interests

My teaching interests are based on fundamental chemistry education with a focus on Physical, Inorganic, Fundamental of Surface Science, Quantum, Computational Chemistry, and Sustainable Practices

I am passionate about educating students in the fascinating field of chemistry as well as chemical engineering, with a particular interest in Physical Chemistry, Inorganic Chemistry, Quantum Chemistry, Chemical Modeling, Computational Chemistry, Surface Science and Sustainable and Green Chemistry. As a teacher, my primary goal is to foster a deep understanding and appreciation for these diverse branches of chemistry while instilling critical thinking skills and promoting sustainable practices.

• In Physical Chemistry, I aim to guide students through the fundamental principles that govern the behavior of matter and chemical systems. By emphasizing the underlying concepts and their practical applications, I strive to create an engaging learning environment that helps students grasp complex topics such as thermodynamics and chemical kinetics.
• In Inorganic Chemistry, I believe in exploring the unique properties and reactivity of inorganic compounds, including coordination compounds, transition metals, and main group elements. Through hands-on experiments and theoretical discussions, I encourage students to explore the diverse applications of inorganic chemistry in catalysis, materials science, and bioinorganic chemistry.
• Quantum Chemistry represents a fascinating realm where students can delve into the principles that govern atomic and molecular behavior at a fundamental level. By utilizing theoretical models and computational tools, I aim to guide students in understanding the quantum nature of matter and its impact on chemical systems. This includes exploring concepts like molecular orbital theory and electronic structure.
• Chemical Modeling and Computational Chemistry provide powerful tools for simulating and predicting chemical phenomena. In my teaching, I emphasize the importance of computational methods and their applications in understanding complex chemical systems, designing new materials, and optimizing chemical reactions. Through hands-on experiences with software packages and simulations, I aim to equip students with the skills necessary to tackle modern chemical challenges.
• Fundamental of Surface Science plays a crucial role in understanding the behavior of materials at interfaces. I am dedicated to introducing students to the principles of surface science, including surface energy, adsorption, and catalysis. By exploring experimental techniques and theoretical models used in surface science research, I aim to provide students with a solid foundation for investigating surface phenomena and their relevance in areas such as nanotechnology, energy storage, and environmental science.
• Additionally, I am dedicated to integrating Sustainable and Green Chemistry principles into my teaching. I believe it is essential to educate students about environmentally friendly practices and their role in addressing global challenges such as climate change and resource depletion. By highlighting sustainable approaches to synthesis, catalysis, and waste management, I strive to inspire students to make ethical and responsible choices in their future careers.

In summary, my teaching interest lies in fostering a deep understanding of Physical, Inorganic, Quantum, and Computational Chemistry, along with the Fundamental of Surface Science while also instilling a strong awareness of sustainable and green practices. By creating an interactive and inclusive learning environment, I aim to empower students to become knowledgeable and conscientious scientists who contribute to the advancement of chemistry as well as chemical engineering and the well-being of our planet.

Figure 1: Overview of research statement and research directions based on machine learning aided catalyst screening, oriented external electric field modulated catalysis and dynamics of catalysis under operating system.