Professor Martin Dove

Research

Research Interests:

Research interests

My primary scientific interests are in disordered materials – crystalline or non-crystalline – and the way in which the atomic structure and dynamics over both long-range and short-range length scales affects material properties and behaviour.

I am particularly fascinated by network materials. These are materials which can be described as an infinite framework of linked groups of atoms that are strongly bound together. An example on which I have done a lot of work is silica, which exists in a number of different crystalline and amorphous phases. At ambient pressures the structures of all these phases can be described as infinite networks of corner-linked SiO4 tetrahedra, where the bonds within the tetrahedra are really quite stiff but with weak angular forces that define the relative orientations of two linked tetrahedra. With this difference between the strengths of internal and angular forces, the network is actually quite flexible in well-defined ways with regard to motions involving rotations of the tetrahedra, and this flexibility is directly responsible for the existence of displacive phase transitions and anomalous properties in some materials such as negative thermal expansion and becoming softer when compressed.

More recently I have become interested in framework materials where metal cations are linked by organic molecules. Actually silica is a special case of this, where the molecule is a single oxygen atom. The first step up is to have linkages between zinc cations mediated by cyanide ions, forming phases that are direct analogues of some of the silica phases. More complex molecular ions can lead to any number of new phases. For example, Zn(Im)2, where Im is the imidazolate ligand C3N2H2– forms many phases that include direct analogues of a number of zeolite phases.

Other significant interests, which actually are not so far removed from what I have just described, include radiation damage, orientationally disordered crystals, atomic dynamics of glasses, and atomic site order/disorder phase transitions.

Science of network flexibility

One of my most significant achievements is to push forward our understanding of the flexibility of network structures, as encapsulated in our Rigidi Unit Mode (RUM) model. We defined a RUM as a normal model of a material in which the polyhedra can move without distortionn. As such, a RUM will have a relatively low frequency, and can easily be associated with the structure deformation associated with a displacive phase transition. Because RUMs typically involve rotations of polyhedra, they will often act to reduce the volume of a material and hence are often associated with netative thermal expansion, particularly since their low frequency gives them a large ampltidue and hence significant impact.

One of our tools is a method to determine exactly the number of RUMs in a material and their wave vectors. We have also developed tools to quantify the extent of RUM motions in atomic configurations obtained from simulations.

Methods

I have always run a two-track approach, namely combining computer simulations with neutron and x-ray diffraction experiments.

Computer simulations

I use a wide range of simulation approaches, from lattice energy and lattice dynamics calculations through to molecular dynamics and Monte Carlo methods. My preference is to work with empirical representations of the forces between atoms rather than taking a purely quantum mechanical approach, because I prefer to work with large system sizes when possible, but increasingly I am using quantum mechanical approaches to parameterise the force models. For example, in our work on metal-organic structures this is absolutely essential, and works well.

One issue that I have developed a strong interest in is managing the outputs of simulations. With current computational capabilities we can now perform studies that involve hundreds of independent simulations, for example scanning across both temperature and pressure in molecular dynamics simulations. To tackle the problem of accurately extracting the key information from the many data files, we are using XML (specifically the Chemical Markup Language) to represent the data, and have developed tools to parse a large collection of XML files and extract tables of results quickly.

Neutron and x-ray scattering methods

I have long valued the information that can be obtained from scattering experiments, ranging from powder diffraction experiments to give information about the variation of crystal structure with changing external variables (temperature or pressure) through to inelastic scattering experiments that provide information on the atomic dynamics (such as phonons). Much of this work is now performed at the ISIS spallation neutron source.

One of my major interests is in the use of total scattering to provide information about structure on a local length scale. This involves measuring the diffraction pattern accurately (ie free from all background scattering) to very high values of the scattering vector Q, and performing the Fourier transform to obtain a quantity known as the Pair Distribution Function (PDF). Put simply, the PDF is a histogram of interatomic distances, and its value comes from the fact that it provides direct information about the structure on a local length scale and the fluctuations in this local structure. Using the Reverse Monte Carlo method we can build atomic models consistent with the total scattering and PDF data. For crystalline materials we also make explicit use of the information contained within the Bragg scattering, and we have developed the RMCProfile code for this work. We have also developed a module for the GULP lattice simulation code to compute the PDF taking proper account of the contribution of phonons to the breadths of peaks in the PDF.

Much of this work is carried out using the GEM diffractometer at ISIS, but we are developing an interest in total scattering from x-rays. We are members of the consortium developing the XPDF diffractometer at Diamond.

Other contributions have been the development of methods to perform diffraction at simultaneous high pressures and temperatures at ISIS, and I was one of the PI's on the project that developed the MERLIN spectometer. We have developed software to simulate the scattering from instruments such as MERLIN.

Other information

My publication list can be obtained here. This page includes links to papers that I am allowed to post on my own web site, and DOI or email links to the others.

My personal web site is available here. It is however slightly out of date with regard to my move from Cambridge to Queen Mary.

My calendar is available here.

Publications