Understanding and controlling polymorphism in molecular solids

Understanding and controlling polymorphism, where molecules crystallise into multiple solid forms, is a major unsolved problem in structural chemistry. Polymorphs often differ significantly in physical properties such as compressibility, solubility and dissolution rate, which majorly affects formulating and manufacturing pharmaceutical products. More fundamentally, polymorphism also tests our understanding of how competing intra- and intermolecular forces determine the kinetics and thermodynamics of molecular assembly during crystallisation.
Exploring the solid-state energy landscape to predict the solid forms of molecular materials is the domain of crystal-structure prediction (CSP).1 Typically, 103-104 candidate crystal structures are generated and their lattice energies evaluated using a parameterised force field model, density-functional theory (DFT), or a combination of both. CSP may find the global energy minimum or a set of energetically similar metastable polymorphs. CSP has evolved with computing capability to become a useful counterpart to experiment. However, even moderately complex systems continue to challenge current CSP methods2, calling for a radical new approach.
Force field models and DFT do not readily decompose energies into different types of interaction. However, Quantum Chemical Topology (QCT) method3, in particular, the modern Interacting Quantum Atoms (IQA) partitioning scheme is an ideal method to obtain chemically rigorous energy contributions.
We aim at a robust protocol for using QCT with our own method called the Relative Energy Gradient (REG) method4 to identify the pertinent interactions. We will then train a machine-learning technique5 on QCT energy contributions to derive an accurate and transferable force field parameterised at the molecular level. We will automate these processes by using the Cambridge Structural Database to select molecules and identify the major degrees of freedom for building the force field. Our ultimate aim is to provide a comprehensive database of parameters and software to allow crystal engineers to perform “point and click” force field calculations with DFT levels of accuracy.

Academic background of candidates:
Applicants are expected to hold, or about to obtain, a minimum upper second class undergraduate degree (or equivalent) in Chemistry (BSc or MChem), Chemical Engineering or Physics. A Masters degree in theoretical/computational chemistry or physics, or experience in machine learning, simulation or condensed matter physics is desirable.