We are a computational materials science group interested in exploring primarily low-dimensional materials, starting with the
most abundant and the lightest ones made of elements of the p-block, in search of new paradigms of tunable electronic, magnetic, topological and chemical properties and their interplays,
aimed at proposing new materials towards meeting some of the contemporary scientific and technological challenges like securing
sustainable non-fossil energy resources and pushing the limit of circuit integration and memory devices beyond silicon and metals of the d-block.
We also develop computational methodologies necessary for our exploration which also aids in furthering the frontiers of computing properties of materials.
What exactly we do?
We try to computationally develop understanding of properties of materials from their
microscopic details like relative location of constituent atoms and the resultant
distribution of electrons in orbitals and bonds.
For that purpose we variationally minimize total energy of valence electrons as a function of
relative location of atoms starting from a good guess.
Given a set of locations of atoms we calculate the total energy and spatial distribution of electrons
within the framework of density functional theory, wherein we self-consistently evolve the charge density
from an initial guess density through approximation of many-electron interactions as function of charge density itself.
The approximation works well for simple metals, or more generally for the less localized weakly correlated electrons,
but undermines band-gap quite grossly with increasing level of spatial localization of states even with the lowest
of principal quantum numbers.
We try to recover the correct band-gap by accounting for correlation through
self energy correction of electronic states which we do perturbatively.
Also, often for better resolution of our understanding we sometimes simplify the total Hamiltonian to some model Hamiltonians
in the basis of fewer orbitals and compute their electronic structure within simpler frameworks such as the mean field approximation of
the Hubbard model.
At times when we are stuck by the enormity or complexity of computation for our work using the known techniques, we also forge out
newer methodologies to overcome the difficulties within arguably acceptable levels of simplifications, which often turns out to be
useful directions of research to pursue.
Of course we do all this through computer programs which are partly already available and partly written by our group members.