Conventional Density Functional Theory (DFT) calculations are limited to a few hundred atoms at most as the computational effort increases with the third power of the number of atoms. To overcome this limitation we have developed a reformulation of DFT with computational cost that increases only linearly, which allows calculations with thousands of atoms. Our linear-scaling DFT method is implemented in the ONETEP program which has been designed to achieve large basis set accuracy and run on parallel supercomputers. Localised orbitals are optimised in situ, and linear-scaling is achieved by taking advantage of the exponential decay of the density matrix according to the physical principle of “near-sightedness of electronic matter” of Nobel Prize winner Walter Kohn. Being able to do large-scale quantum calculations is one thing but using them to solve real problems of industrial relevance is another which throws up other challenges to be overcome. The point is that molecules, biomolecules and nanoparticles are not isolated and not at a temperature of zero Kelvin. On the contrary, they interact heavily with each other and their environment (e.g. solvent) and are in constant thermal motion.
Therefore, this PhD will be focused towards developing new models with which to improve and augment our quantum simulations in order to achieve the required level of realism for the quantum system and its environment. Examples of possible developments include novel DFT methods which connect with wavefunction-based methods, electronic response to external probes as required in the simulation of various spectroscopies, and energy decomposition analysis schemes for intermolecular interactions. Developments could also be in the area of multiscale models which interface a high level of theory for the system of interest with a lower level of theory for its environment (e.g. solvent models, lower level quantum and classical theories, such as polarisable force fields) which are necessary for a realistic description of the environment of the quantum system. The implementation of these highly non-trivial theories will need to be formulated within the localised wavefunction theoretical framework of ONETEP which is required for the linear-scaling computational effort. Modern software engineering principles will need to be used for these developments as they are intended for a high-quality parallel code with a large user and developer base.
The new methods that will be developed during this PhD will open the way for simulations with an unprecedented level of realism in grand-challenge applications such as the simulation of biomolecular association (drug design) or organic photovoltaic materials.
If you wish to discuss any details of the project informally, please contact Professor Chris-Kriton Skylaris, Email: [email protected], Tel: +44 (0) 2380 59 9381.