Dr Alston Misquitta Project Abstracts

Review Projects (updated Jan 2018)

Topics in intermolecular interactions
This is a big field encompassing all of non-chemical molecular interactions. The intermolecular interactions are responsible for much of the interesting phenomena in the physical world: the existence of gases, liquids and some solids, the self-assembly of complex molecular systems. The Casimir effect is a more general case of the van der Waals (or dispersion) interaction. In this review project you will be able to choose from a number of topics in this field. The level of complexity can vary substantially: some of the more theoretical aspects can be very advanced, while, at the application end, the mathematical level needed is quite low. 

1. General theory of intermolecular interaction : Basic theory; Methods for calculating the interaction energy; Applications to specific systems.
2. Charge-transfer : this is a special interaction seen in hydrogen-bonds. It can be thought of as an incipient chemical bond forming. 
3. The van der Waals interaction : In particular, exploring the link between perturbation theory expressions for van der Waals and the continuum approach developed by Lifshitz. 
4. Methods for handling interactions with metallic systems.
5. Model Breakdown: Investigations into highly non-additive systems where the standard additive models breakdown. Models proposed to fix the problem. Example systems include extended metallic and semi-metallic low-dimensional materials.
6. Unusual interactions: Examples include: Polarons: When free electrons are localised by a combination of the polarization and van der Waals interactions. Low dimensional systems: In which the long wavelength electronic fluctuations dominate and cause strong and highly non-additive dispersion interactions.
7. Models for the interaction energy: A survey of models with the intent of understanding which ones will be expected to work best and in which circumstances.
8. Information theory and its use in the theory of atoms-in-a-molecule: we are now more than ever using information theory to decompose molecular properties into those of the constituent atoms.   
9. Reviews on the uses of various theoretical methods for intermolecular interactions. Benchmarking the methods, etc. 
10. Review on practical applications of intermolecular interactions: technological, bio-molecular, health care, pharma, defence, fundamental physics, etc.

Review projects on other topics may also be possible. I am also interested in various aspects of quantum mechanics, machine learning, complex systems, and even chaos theory. Of course, these choices will depend a lot on what your mathematical background is, and what I think I am capable of advising you on at the time. If you are unsure, just come over and talk to me about it.

MSci/MSc Research/Investigative Project Abstracts (Jan 2018)

Advanced electrostatic models for simulations 

NOTE: This is not one project but many interrelated ones.
The electrostatic interaction is the simplest of the intermolecular and intra molecular interactions. Yet it is also the most important: if you want to get the structure of hydrogen-bonded complexes correct you need to have a reasonably good description of the electrostatic interactions of the species. Likewise, a good description of the electrostatic interactions is essential when modelling complex structures like metal-organic frameworks (MOFs are materials that exhibit unusual properties like negative thermal expansion – see the models in the Physics Museum cabinets). In these projects we will use a state-of-the-art method for partitioning molecules into atomic-like parts to compute electrostatic models. This method, known as the iterative stockholder atoms (ISA) approach was developed by Wheatley and Lillestolen [1] and I have recently developed a robust numerical algorithm to implement it. This model for atoms in a molecule surpasses any available for its numerical properties and the physical nature of the atoms. Consequently it will allow us to explore:

1. Charge-penetration effects: These are the effects of inter-penetrating quantum densities. They are missing from most empirical potentials.
2. Changes in multipole moments as function of vibration: This is particularly important: how do the atomic moments alter as bonds stretch or bend?
3. Comparisons of ISA multipole moments to those from other methods: This will include learning how to calculate reference electrostatic energies using SAPT(DFT). 
4. Variations of multipoles with conformation: How much do they vary? Why? Can we model these variations using simple polarization models?
5. Charge-Transfer: We have state-of-the-art methods for calculating the charge-transfer energy, but we do not understand enough of this important energy to model it sufficiently accurately. Questions to be answered include: How does this energy vary with separation and with angle? Is it proportional to the exchange-repulsion energy? Can we make a formally correct model for this energy? 
6. Distributed polarizabilities using the ISA-Pol method. This is a very new technique that has a lot of potential for dramatically changing the way we calculate distributed polarizabilities and dispersion models. There are a number of projects associated with this technique. 
7. Basin-Hopping: This is a beautiful method for studying the potential energy landscape of complex systems. We have some of the most accurate models for these interactions that can be used in a basin-hopping search for low-energy stable structures. These structures, which are often very beautiful and highly symmetric, are often seen experimentally in molecular beam studies. 
8. Psi4: We have recently started using the Psi4 code in my group. This has opened up a range of new types of calculations and there are many kinds of projects that can be designed around this program. Examples include benchmarking various kinds of DFT methods, comparisons of SAPT (from Psi4) with SAPT-DFT (from CamCASP), calculations with correlated methods like MP2, CCSD(T), etc.
9. Effective polarization models: Can we replicate some of the effects of polarization without an explicit polarization model? This has huge implications in the simulation of complex system for which polarization is important, but an explicit model may cause too large a computational expense. 

Important: Not all projects listed here may be available. A lot depends on your background and ability (mainly mathematical and computational), and also my own interests at the time. Some projects are very demanding and will only be made available to those of you with a good mathematical and computational background.

There may be other projects available at the time you make your selections, so use the above as a guide only. 

Day-to-day: Review of literature on electrostatic methods. Using the CamCASP program to calculate and analyse intermolecular interactions; understand how molecule partitioning methods work; compare models with the literature. We will also use other codes like Psi4 and NWChem/DALTON.

Pre-requisites: (not all essential, but will need to be acquired) The Linux shell, basic Python scripting, basic Fortran90, Quantum Mechanics and a good Mathematics background. We will make extensive use of Linux and some scripting, so you need to be prepared to learn these well during the course of the project.

[1] T. C. Lillestolen and R. J. Wheatley, "Redefining the atom: atomic charge densities produced by an iterative stockholder approach", Chem. Commun., 2008, 5909-5911 (2008). 

[2] A. J. Misquitta, A. J. Stone and F. Fazeli, "Distributed multipoles from a robust basis-space implementation of the iterated stockholder atoms procedure", J. Chem. Theor. Comput. 10, 5405--5418 (2014).

[3] A. J. Misquitta, "Charge-transfer from regularized  symmetry-adapted perturbation theory", J. Chem. Theor. Comput.  9, 5313-5326 (2013).

Charge-Transfer

The hydrogen-bond remains one of the most versatile and often-seen types of intermolecular bonds in biological systems: Base-pairs are bonded by hydrogen bonds in a zipper-like manner; water forms hydrogen bonds to give ice its open hexagonal structure. A characteristic of these bonds is charge- transfer which can be understood as the delocalisation or sharing of electronic charge between the bonded molecules. This sharing leads to stabilization and consequently to the strength of the bond. Charge-transfer poses a considerable problem for interaction models (the simple potentials used to model large proteins) and quantum mechanical methods like density functional theory. The more we understand about charge-transfer the better. In this project we will use state-of-the-art perturbation theory methods to investigate charge transfer in a variety of situations and novel charge-density analysis methods to see the extent of this transfer in complex systems. 

Day-to-day: Review of literature on charge-transfer complexes; review of basic ideas in intermolecular perturbation theory; review of current literature on charge-transfer; get hands-on experience with using the CamCASP program to calculate and analyse intermolecular interactions; identify a set of systems that we will perform calculationson; calculate charge-transfer energies for these systems; understand how charge-density-decomposition methods work; use theseto develop an understanding of the charge flow in the system; compare results with literature; make well-defined statements about the charge-transfer process and, possibly, on the many-body aspects of this process. 

Pre-requisites: (not all essential, but will need to be acquired) The Linux shell, basic Python scripting, basic Fortran90, Quantum Mechanics.