Prosiectau Ffiseg ar gyfer myfyrwyr bl.3 a bl.4

Disgrifiad prosiect

Graph based sampling of fully-coordinated geometries of clusters of network-forming materials

(supervisor: Edwin Flikkema)

Nature of project: software, data analysis

Available to students on full-time physics degree schemes or joint students.

Project description and methodology

The project will be within the context of (atomic) clusters of network-forming materials (such as SiO2, TiO2, ZnO), studied by computational methods. These types of clusters have been studied before using empirical force-fields [1] combined with global optimisation methods to obtain stable cluster geometries [2]. Generally, these cluster geometries will exhibit defects in the form of over- or under-coordinated atoms and dangling oxygens. This project will focus on a specific class of cluster geometries which do not have such defects: the so-called fully-coordinated (FC) cluster geometries. The supervisor has already developed a methodology for specifically sampling those FC geometries in the particular case of SiO2 clusters using a particular force-field [3]. The student will attempt to generalise this method to other materials (e.g., TiO2, ZnO) and/or to other types of force-fields. A basic project would be to work with the existing code and apply this to sizes not yet studied before. This involves adapting the simulation parameters in an attempt to improve the sampling efficiency. Large-scale production runs need to be performed and the results need to be analysed to assess the success in finding low-energy fully-coordinated cluster geometries.

A successful project will develop beyond the above in one/some of the following directions:
To take this project further, a new force-field could be implemented. In its simplest form this would mean changing the parameters of the force-field already implemented in the code to model e.g. TiO2. Other types of force-fields could be implemented.

When considering where to take your project, please bear in mind the time available. It is preferable to do fewer things well than to try many and not get conclusive results on any of them. However, sometimes it is useful to have a couple of strands of investigation in parallel to work on in case delays occur.

Additional scope or challenge if taken as a Year-4 project: For a Y4 project, the student can attempt to implement the complicated "environment dependent dynamic charge potential" (EDDQ) potential [4]. Earlier attempts have failed, so this is challenging.

Please speak to Edwin Flikkema (edf) if you consider doing this project.

Initial literature for students:

  1. Flikkema, E. ; Bromley, S. T., Chem. Phys. Lett., 2003, 378, 622.
  2. Bromley, S. T. ; Flikkema, E., Phys. Rev. Lett., 2005, 95, 185505
  3. Flikkema, E. ; Bromley, S. T., Phys. Rev. B, 2009, 80, 035402
  4. Muralidharan, K; Cao, C; Wan, Y; Runge, K; Cheng, H; Chem. Phys. Lett., 2007, 437, 92

Novelty, degree of difficulty and amount of assistance required

The project will involve generalising an existing methodology to be able to apply it to new chemical systems (or to the same SiO2 system with better tuned parameters). This generalisation may involve the implementation of force-fields not yet included in the existing code. Previously a Y4 student has worked on generalizing this approach to TiO2. He has also started on implementing (coding) the EDDQ potential for silica [4], which proved very challenging. The current student can build on this work, e.g. by continuing with the EDDQ potential. In the initial phase of the project the student will need close supervision in order to gain familiarity with the existing computer code. Close supervision is also required if the project is going to include coding. After this initial phase the student should be able to run the code and analyse the results without much assistance.

Project milestones and deliverables (including timescale)

milestoneto be completed by
Familiarisation with existing code.Christmas
Performing test-runs with existing code with existing parameters.end of February
Performing runs with adapted parameters, and/or improved code.mid-March
Analysis of data.Easter