Executive Summary : | The problem of determining the plausible crystal structures only from the information of a molecular structure is a challenge in material chemistry. The motivation for predicting the crystal structure a priory from the molecular information has main advantages. 1) The anticipation of the plausible structures may help to predict a priori the properties of the crystal 2) The predicted structure may act as a meaningful starting point for solving the crystal structure of the molecules, for which it does not form experimental single crystals. 3) The prediction of solid-state phases creates an opportunity for the in-silico design and prior knowledge of the properties of the materials before entering the laboratory. The computer-led prediction and prior knowledge of the untouched areas of chemical space may guide the experimental work.
The proposed work involves an exhaustive search of the chemical phase-space. Searching all the chemical-space without missing any relevant area is a challenge in CSP research. The metric we are using for identifying the most feasible experimental structure in the chemical-space is the lattice energy. The widely used approaches for calculating lattice energy for organic crystals is based on the force-fields. The difference in the predicted lattice energies of structures is less than one kJ/mol in the most general cases. Highly accurate ab initio quantum methods are mandated for a precise calculation of the lattice energies of molecular crystals. However, employing ab initio quantum chemical methods is expensive for the entire search and screening process involved in CSP. Recently, an alternative approach of combining the accurate quantum chemical method with the machine learning method is acquiring acclamation. This proposal is a novel idea to combine the quantum chemical methods with the machine learning algorithms to calculate the lattice energies and predict the energetically feasible crystal structures.
The proposal on developing methods for CSP has extensive ramification and applications in material science. A successful CSP algorithm may have immediate repercussion and, may revolutionize the current ‘heat-beat-and-wait’ approach in material design. The development in CSP research front definitely will influence many sectors like energy, health, environment. These proposed algorithms have wider applications in chemistry, such as in designing catalysts and drug-designing, and may help in reducing the manufacturing cost of materials. Also, CSP research has the potential to improve the living standards of human beings and boost the economy of India. For achieving this vision, I am submitting the details of the proposal on CSP. |