It is a curious fact that fully quantum mechanical predictions of the properties of isolated molecules can now be made with accuracy which rivals the most precise experimental spectroscopy techniques. However, when looking at extended systems, such as the interaction of a molecule with a solid surface, state-of-the-art computational approaches can often not even reach the accuracy required to deduce correct structures or interaction energies.

The aim of this ambitious research is to make progress in this area – to transfer the accuracy of quantum chemical approaches to the setting of extended systems, by development of new approximations and techniques which use the electronic wavefunction as the central quantity of the simulation. The wavefunction, despite being the first quantum variable which is introduced, is almost entirely neglected within computational simulations of extended systems. This is because of the exponentially large amount of information required to specify it, which has meant that alternatives such as the electron density has generally been used instead.

However, most of this complexity is artificial. For example, within insulating systems, the correlation length between electrons decays exponentially, and so approximations based on locality of electrons or embedding of correlation effects can be introduced, rendering it a tractable computational object. Additionally, parts of the wavefunction have a universal, analytic form (such as when two electrons occupy the same point), and so these parts of the wavefunction can be considered known, and removed from the required parameterization. Furthermore, clever optimization strategies can be developed, including Monte Carlo sampling of the wavefunction, and compact functional forms of the wavefunction, which can dramatically increase the potentiality of this approach.

These new ideas will be developed and then applied to real systems of significant technological interest, where current techniques are lacking, such as correlated transition metal oxide materials, and organic photoactive molecular crystals.

The project will have a significant programming component, where these new methods will need to be coded and tested, before optimization for use on supercomputing resources. Furthermore, the successful candidate should have a strong interest in modeling techniques and quantum many-body physics and/or chemistry. The studentship will be help at King's College London, UK, under the supervision of Dr. George Booth. The project is open to both UK and EU students, and will include a stipend (with London weighting) and cover fees.

If you wish to discuss any details of the project, please contact Dr. George Booth, Email: [email protected], Tel: +44 (0)20 7848 7077