Optical Properties of Materials

The project we propose is aiming to achieve two things simultaneously. One thing is that we have recently implemented parallel TDDFT gradients which allow us to perform geometry optimisations in the excited of molecules. This North-West developed capability is unique in that it is the first implementation that targets large machines. What we don't know at present is how well this method performs for calculating for example fluorescence properties of various materials and clusters. The other thing we don't know is how efficient and scalable the current implementation is. So it is essential that we establish the characteristics of the methodology before we can give this out to general users. The other thing is that we work with Alexei Sokol at UCL to study optical properties of materials. In particular materials that have the potential to be used in solar cells or direct photo-hydrogen generation are of interest. In this category there is a range of materials being Boron Nitride, Aluminium Nitride, and Gallium Nitride. These materials are of interest because they form an iso-electronic series with very similar sodalite cage structures. However their properties do change as the "metal" is replaced going down the periodic table from Boron to Gallium. Gallium Nitride for example is a well known wide band-gap semiconductor. It has applications in blue light emitting diodes, long life-time violet lasers and UV detectors as well as satellite solar cells. Aluminium nitride is an extremely wide band gap semi-conductor. Due to this and its similarities with Gallium nitride it holds promises for the creation of ultraviolet light-emitting diodes. However, at present relatively little is understood about the optical properties of this material. Boron nitride is a semi-conductor at least in the sodalite cage polymorph. Interestingly enough all the allotropes seem to have a very similar band gap.

With the code we have developed it is possible to study the optical properties of these materials and compare how and why the optical properties change as a function of the "metal". Also because the code should run well in parallel it will be possible to study how these properties depend on the size of the clusters. This will require running many parallel calculations, which could most effectively be run in a grid which provides both capability as well as capacity computing, like NWGrid. In particular the eScience infrastructure available on NWGrid will assist significantly in creating the calculations and harvesting data and subsequently comparing results from the various calculation. This will help provide essential insight that will help explain experimental data available for these materials. Hopefully it will also aid in guiding the development of new applications of these materials as either solar cells or optical devices.