Research Interests

Model Potentials for metal clusters and nanoalloys

We study potential energy surfaces of metal clusters and nanoalloys. Numerous empirical potentials were developed earlier to study metals but none weretransferable to study small sized metals clusters (less that few 100 atoms). This is because of quantum effects that dominate this size regime. To accurately model interactions in metals clusters including quantum effects we build an on-the-fly fitting approach based on Artificial Neural Networks (ANN). ANN is basically a softcomputing technique that is widely used in many non-linear problems. For this we require to train our ANN using variety of structures of Na previously evaluated using DFT. Once trained we can directly use the network to generate PES using molecular simulations.

Model Potentials for small organic molecules

Our aim is to generate potentials functions that are relatively cheap and reliable for small organic molecules. We are particularly interested in AMOEBA (atomic multipole based force field for biomolecular applications). AMOEBA uses charges, dipoles and quadrapoles to study long range interactions. We are trying to develop a standalone code that can generate AMOEBA force field parameters for any organic molecule. At present we are interested in generating parameters for all nucleobases. To test the reliability of the potential parameters we performe global optimizations for small clusters (upto size 4) of all nucleobases. It requires few hours of computing time for a DFT optimization while AMOEBA is computationally very cheap (just take few seconds).

Using computationally cheap yet accurate potential functions we are interested in studying structure/ properties of self-assembled structures of nucleobases and novel magic nano-clusters which may be used as building blocks to make materials with exciting electronic/catalytic/biological properties.

Devolopment of novel computational techniques to study nanoalloy clusters

Our objective is to model nanoalloy clusters, a new class of materials with promising applications in catalysis and biosensing, using new computational modelling methods such as particle swarm optimizations(PSO), multicanonical montecarlo method, infinite swapping method, nested markov chain methods etc.

Dr. Satya S. Bulusu

Assistant Professor

Chemistry (Theoretical and Computational Chemistry)

sbulusu[at]iiti.ac.in

Faculty Profile