My Current Research

Materials for alternative energy

Deisgning materials having desired target properties.

• Materials for alternative energy:
  We are focussing on two aspects of this broad area. First, materials for electro-catalytic water splitting: efficient hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) catalysts for both acidic and alkaline media. For acidic media the target is to replace platinum, the best known HER catalyst, by earth abundant, inexpensive materials.
  
  Publications:
  
• Combinatorial design and computational screening of 2D transition metal tri-chalcogenide monolayers: Toward efficient catalysts for hydrogen evolution reaction, P. Sen, K. Alam, T. Das, R. Banerjee and S. Chakraborty, J Phys. Chem. Lett. 11, 3192 (2020).


• Probing active sites on MnPSe₃ and FePSe₃ tri-chalcogenides as a design strategy for better hydrogen evolution reaction catalysts, T. Das, K. Alam, S. Chakraborty and P. Sen, Int. J Hydrogen Energy 46, 37928 (2021).


• Finding the catalytically active sites on the layered tri-chalcogenide compounds CoPS₃ and NiPS₃ for hydrogen evolution reaction, K. Alam, T. Das, S. Chakraborty and P. Sen, Phys. Chem. Chem. Phys. 23, 23967 (2021).




• Designing materials with targeted properties:
  With steady increase in our understanding of the fundamental laws governing natural phenomena, and rapid growth in our computational capabilities, we have now reached a stage where we can attempt to design new materials with precise properties that never existed in nature. We have taken two approaches for this: High Throughput Computation (HTC), and Machine Learning assisted screening and/or design.
  
  In the first approach, a large number of materials are constructed via combinatorial enumeration. Then their properties of interest are calculated in a HTC framework using DFT. Candidate materials meeting the requirements are further analyzed, and depending on the outcome are suggested to experimentalists for synthsis. We have propsed ten possible low-cost, efficient HER catalysts fro acidic media via this method. Attempts to synthesize some of them are currently on.
  
  In the second approach, we train one or more ML models to predict material properties, or to classify them in appropriate classes using available DFT or experimental data. Then the candidate materials are screened using these models to identify the potentially useful ones. These are then taken through DFT for further verification. Those passing this test are suggested for experimental synthesis. We are currently deisgning RE-free PMs in this approach.
  
  In a variant of this approach, we 'inverse' design materials with desried properties using a generative deep newral network (DNN) called variationa autoencoder (VAE).
  
  Publications:
  
• Machine learning assisted hierarchical filtering: A strategy for designing magnets with large moment and anisotropy energy, A. Dutta and P. Sen, J Mater. Chem. C 10, 3404 (2022).



In addition, we work on a number other materials that may or may not be classified into one or the of the above classes. Here are some examples.


•  2D materials:   Ever since the exfoliation of graphene, two-dimensional materials have become an important area of research. We explore a variety of 2D materials motivated both by their fundamental properties, and application potential.
  
  Publications:
  
• Computational design of a robust two-dimensional antiferromagnetic semiconductor, S. Chabungbam and P. Sen, Phys. Rev. B 96, 045404 (2017).


• Electronic structure of MPX3 tri-chalcogenide monolayers in density functional theory: A case study with four compounds (M=Mn, Fe; X=S, Se), P. Sen and R. Chouhan, Electronic Strct. (2020).


• Cobalt Nanoparticles Dispersed Nitrogen-Doped Graphitic Carbon Nanospheres-Based Rechargable High Performance Zinc-Air Batteries, P. Thakur, M. Yeddala, K. Alam, S. Pal, P. Sen, and T. N. Narayanan, ACS Appl. Energ. Mater. (https://dx.doi.org/10.1021/acsaem.0c01200).