Researchers in the Cavendish Laboratory at the University of Cambridge are perfecting methods of harvesting solar energy using artificial photosynthesis via a special chemical dye. The dye absorbs sunlight, affording sufficient energy to inject one of its electrons into attached titanium dioxide nanoparticles. This process stimulates an electrical circuit which cycles many hundreds of times a second, creating a dye-sensitised solar cell, like a mini chemical engine.

"The nature of the chemical dye is very important because it has to be able to absorb solar energy, so providing the dye with sufficient energy to release electrons into the nanoparticles to kick-start the electrical circuit," says Head of the Structure and Dynamics (S&D) Group at the University of Cambridge, Professor Jacqui Cole.

An exciting application of these solar cells is solar powered 'smart windows' where the cell is embedded barely visibly in the glass. Such glass could power whole buildings in 'smart cities' independently of the electricity grid. There is huge motivation for companies and British research teams, therefore, to find better chemical dyes that perform this dual process very well.

A key part of this research is to undertake big, very complex computations to validate the molecules that have been designed. The S&D Group had some computational knowledge but only with certain people, and it was incomplete. The group required formal training in computational chemistry.

Perfecting this organic chemistry is difficult. Chemical computations, designed to optimise synthetic molecules for performance, help to validate the selected molecules more quickly. Developed together with the EPSRC UK National Service for Computational Chemistry Software at Imperial College London (NSCCS), the computational model tests multiple molecule variants within a huge range of parameters before revealing the best design.

In a lab like the S&D Group, photovoltaics researchers tend to be pure experimentalists more than computational experts. But the researcher's understanding of how the molecule's design can be verified using computations is imperative, as this will reduce the total research hours of complementary experiments. This is a vital consideration in proving the technology commercially and in a sensible timescale, so high-quality training is vitally important.

The S&D Group includes chemistry, materials science and physics PhD students, all very capable of understanding the computations, but with little practice as they were focused on experiments. "Importantly, there was a very good rapport between NSCCS and our research team," Cole says. "They came up with solutions swiftly and pre-empted new features in the computations."

The cost and performance of the molecule also commands powerful computation.

"Researchers struggle to find the best chemical dyes in terms of stability, cost and performance," says Professor Cole. "The economics of solar cell technology are governed by the Price-to-Performance ratio. It needs to be cheap compared with expensive silicon, but needs to raise its performance to get a better P/P ratio than silicon or other materials. We try to make better molecules that perform better - this is molecular engineering and it needs verification."

The calculations for this branch of molecular science used the Gaussian software combined with High Performance Computing, for which NSCCS provides the UK's dedicated facility. Additionally the NSCCS provided a service that helps decipher the specific characteristics of dye-sensitised solar cells quickly. "You absolutely need the computational compliment, to match and check the results of experiments with calculations," says Professor Cole.

The number of calculations required for these projects and the size of the molecules being studied meant that access to a dedicated and integrated facility was essential: the group could not have made progress with only desktop computers, even if the software had been available to them. The training that NSCCS is uniquely able to provide was also essential for the S&D Group's research.

There has been a huge shift in the computational ability in the group. "We have even had students who were full-time experimentalists who now prefer computation. To have the capability of doing world-class experiments and computation under the same roof is compelling," says Cole.

This work has huge commercial potential. Perfected dye-sensitised solar cells have high potential commercial value as a superior alternative to silicon solar cells. They are more sustainable, using organic not semi-metal-based inorganic materials and could be cheaper. Crucially for 'smart windows', dye-sensitised solar cells are transparent. Accordingly, smart buildings are the obvious application, but domestic solar power and eco-factories or even wearable power generation are commercial possibilities.

The S&D Group also works on non-linear optical materials for application in ultra-fast telecoms, where NSCCS computations also accelerate research time. Inorganic materials dominate Photonics in the telecoms sector, but the demand for speed is exposing their limitations. "Future data transfer demands are too fast and the much faster optical response of organics means that they could overcome the bottleneck," says Professor Cole.

Another project with both huge potential and risk for the UK science industry is the US Materials Genome Initiative or MGI, US president Barack Obama's selected science project. One of the four tenets of the MGI is the need to integrate experiment, computation and theory to drive down the current time period from the point of a new material's discovery to its market innovation. MGI is providing the US materials community with advanced tools and techniques, such as computational models, to develop materials faster.

"The MGI is all pervasive in the US, so if academics in the UK are to be globally competitive we need to be aligned to the MGI strategy," says Prof Cole. "The seamless integration of experiment with computation achieved in my research group, for which NSCCS engagement has been instrumental, is very relevant to achieving and maintaining our global competitiveness in materials discovery."

EPSRC EP/P505445/1