MMEC · IIT HyderabadShelaka Gupta

01 — Research

Catalyst design from first principles

The group uses quantum-mechanical ab initio density functional theory simulations for the rational design of heterogeneous catalysts used in the sustainable production of fuels, materials and chemicals.

Work runs from the elementary step upward: adsorption energies and transition states on a slab, assembled into microkinetic models, then reduced to the handful of descriptors that actually govern a surface’s behaviour. The aim throughout is a mechanistic account specific enough to say which site, which facet, and which second metal.

Biomass to fuels and chemicals

Mechanistic routes for upgrading biomass-derived platform molecules — hydrodeoxygenation on bimetallic PdZn, ring-opening of γ-valerolactone and lactones, retro-Diels–Alder chemistry of partially saturated 2-pyrones, and the development of 6-amyl-α-pyrone as a platform molecule in its own right.

  • HDO
  • Ring-opening
  • retro-DA
  • 6-PP
  • PdZn
SchematicFree energy along a reaction coordinate: bound intermediates separated by transition states. Illustrative of the method; not computed values.

CO2 and small-molecule conversion

Selective hydrogenation of CO2 to clean methanol over MgO-promoted Cu/ZnO, computational design of catalysts for the upgradation of CO2 to dimethyl ether, microkinetic modelling of methane dry reforming on bimetallic nanoparticles, and predictive microkinetics of the acidic oxygen evolution reaction on IrO2.

  • Cu/ZnO
  • DME
  • Dry reforming
  • OER
  • MKM
SchematicDegree of rate control across elementary steps — which step a microkinetic model says the rate actually hangs on.

Alloy surfaces and descriptors

Oxidation and diffusion behaviour of medium- and high-entropy alloys through DFT and molecular dynamics, phase growth in binary metallic systems combining ab initio calculations with diffusion-couple measurements, and machine-learning approximation of CO and OH binding energies on Cu-based bimetallic alloys.

  • HEA/MEA
  • Diffusion
  • ML descriptors
  • Segregation
SchematicActivity against a binding-energy descriptor — too weak and nothing binds, too strong and nothing leaves.