Data-driven Exploration of Metastable Molybdenum Chalcogenides

The goal of this research is first to understand and second predict the properties of useful materials by means of computer simulations. An example is the capacity
of a material to store and transport lithium, which determines the life and maximum electrical power of a battery made from it.
Specifically, this research will develop Machine Learning (ML) approaches to materials modelling, which falls under the EPSRC area of manufacturing the future with
artificial intelligence. ML is the science of computer algorithms that improve with experience, not the explicit intervention of a human programmer. Modern computer simulations of chemical systems can be highly accurate, whether in predicting useful properties such as Lithium capacity or unravelling the structure of complex crystals. However, the simulations usually involve solving (approximately) the equations of quantum mechanics, which requires prohibitively large computing power for models that exceed roughly 1000 atoms in size. We develop methods that 'learn' from quantum-mechanical data to reach that same level of accuracy in simulations, but with far less computing time. This provides access to length-scales in simulations that correspond much more closely to the situation in real devices, giving us new insights to rationally design and discover better materials.
One of the main themes of the project is the construction of the training database: a set of example chemical compounds that the ML models generalise from to predict the properties of new compounds. We will examine the following outstanding questions:
- What kind of data is required for sufficient learning? Is it enough to show the model lots of disordered liquid atoms for it to generalise to all the possible situations an atom might find itself in, or are e.g. solid crystals important too?
- Training data typically includes hundreds of thousands of examples of atoms in different environments. How do we evaluate their information-content systematically? This is important because computer memory limitations restrict the number of data points that can be used directly for learning, so we must select a subset-ideally the most informative ones.
- How do we tell how accurate the ML model is? It is simple to compare the force on an atom predicted by the model versus the training data, which gives us one measure of the error of the model. However, this error does not seem to correlate well with that of predictions of properties of the bulk material, e.g. how well it conducts heat. We likely to be interested in such properties to design a useful material, so it is important that we have confidence in the values we calculate for them.

As a test case for the ML methods development, we will also study the compounds of the elements Molybdenum and Sulfur. The reasons for choosing these compounds in particular are twofold. Firstly, they are used extensively in the chemical industry as a lubricant, to remove the sulfur from crude oil, and to capture Mercury fumes to stop them escaping into the environment. More recently, Lithium has been found to move freely between the sheets of their layered structure, enabling applications in battery materials. The many uses of Molybdenum Sulfides alone makes understanding and predicting the relationship between their structure and properties important.
Secondly, there has been debate among scientists since 1975 over the true structure of MoS3: whether the Molybdenum atoms arrange themselves in triangles or in long chains. Because of the complexity of the disordered network formed by these units, the available experimental evidence is open to interpretation. Computational studies have been able to provide only limited assistance to date in understanding the experimental results because they cannot use models with enough atoms to correspond well to reality. We hope that the new ML methods will be able to resolve the conundrum, in the process proving both their accuracy and utility.

The primary impact of the OxICFM CDT will be the highly-trained world-class scientists that it delivers. This impact will encompass both the short term (during their doctoral studies), the medium term (subsequent employment) and ultimately the longer timescale defined by their future careers and consequent impact on science, engineering and policy in the UK.

The impact of OxICFM students during their doctoral studies will be measured by the culture change in graduate training that the Centre brings about - in working at the interface between inorganic synthesis and manufacturing, and fostering cross-sector industry/academia working practices. By embedding not only from larger companies, but also SMEs, we have developed a training regime that has broader relevance across the sector, and the potential for building bridges by fostering new collaborations spanning enormous diversity in scientific focus and scale. Moreover, at a broader level, OxICFM offers to play a unique role as a major focus (and advocate) for manufacturing engagement with academic inorganic synthetic science in the UK.

From a scientific perspective, OxICFM will be uniquely able to offer a broad training programme incorporating innovative and challenging collaborative projects spanning all aspects of fundamental and applied inorganic synthesis, both molecular and materials based (40+ faculty). These will address key challenges in areas such as energy provision/storage, catalysis, and resource provision/renewal necessary to enhance the capability and durability of UK plc in the medium term. To give some idea of perspective, the output from previous CDTs in Oxford's MPLS Division include two start-up companies and in excess of 30 patents.

It is not only in the industrial and scientific realms that students will have impact during their timeframe of their doctorate. Part of the training programme will be in public engagement: team-based challenges in resource development/training and outreach exercises/implementation will form part of the annual summer school. These in turn will constitute a key part of the impact derived from the CDT by its engagement with the public - both face-to-face and through electronic/web-based media. As the centre matures, our aspiration is that our students - from diverse backgrounds - will act as ambassadors for the programme and promote even higher levels of inclusion from all parts of society.

For our partners, and businesses both large and small in the manufacturing sector, it will be our students who are considered the ultimate output of the OxICFM CDT. Our programme has been shaped by the need of such companies (frequently expressed in preliminary discussions) to recruit doctoral graduates who can apply themselves to a broad spectrum of multi-disciplinary challenges in manufacturing-related synthesis. OxICFM's cohort-based training programme integrates significant industry-led training components and has been designed to deliver a much broader skill set than standard PhD schemes. The current lack of CDT training at the interface of inorganic chemistry and manufacturing (and the relevance of inorganic molecules/materials to numerous industrial sectors) heightens the need for - and the potential impact of - the OxICFM CDT. Our students will represent a tangible and valuable asset to meet the long-term skills demand for scientists to develop new materials and nanotechnology identified in the UK Government's 2013 Foresight report.

In the longer term, the broad and relevant training delivered by OxICFM, and the uniquely wide perspective of the manufacturing sector it will deliver, will allow our graduates to obtain (and thrive in) positions of significant responsibility in industry and in research facilities/institutes. Ultimately we believe that many will go on to be future research leaders, driving innovation and changing research culture, and thereby making a lasting contribution to the UK economy.