Research Programme
Scientific instruments are often prohibitively expensive. They are also closed source, meaning that once bought, they cannot be modified or extended by the research community. We develop extensible open-source scientific instruments that provide access to important experimental techniques at a fraction of the cost, and share these designs freely to democratise access to cutting edge techniques.
In many cases in materials science, manual workflows mean that only a few samples can be studied at a time. This creates two important problems. Firstly, small sample sizes are susceptible to statistical fluctuations, making many important results difficult to reproduce. Secondly, manual workflows introduce statistical bias through differences in how samples are prepared, measured, and analysed. Taken together, these problems create a huge barrier to using advances in machine learning to drive materials discovery, which relies on large, high quality datasets. We use Python to automate our measurements. This enables us to standardise our experiments and study materials on a larger scale. This allows us to interrogate the statistical strength of our conclusions and to utilise machine learning to more deeply understand our data.
Renewable energy will replace fossil fuels. However, this means that alternatives to fossil driven industrial processes must be found. We study electrocatalytic materials that drive transformations that generate carbon neutral hydrogen, ammonia, and high value hydrocarbons using water and air.
The sun is the most abundant energy source we have access to. We investigate semiconductor photocatalysts that channel solar energy directly into bond-making and bond-breaking: converting light, water, and abundant feedstocks into fuels, fertilisers, and value-added chemicals.