Dr Kane Shenton is a Computational Scientist for Materials Software, Data and Workflows at STFC Scientific Computing. He is the Project Lead for the Collaborative Computation Project in NMR crystallography (CCP-NC), supporting the community with Chair Professor Paul Hodgkinson.

CCP-NC supports a multidisciplinary community of NMR spectroscopists, crystallographers, materials modellers, and application scientists by developing and integrating software across the area of NMR crystallography. This is an emerging field, recently recognised by the International Union of Crystallography, defined as the combined use of experimental NMR and computation to provide new insight, with atomic resolution, into structure, disorder, and dynamics in the solid state.

Research Community Manager Alison Oliver talks to Dr Shenton about his career so far and his journey from physics to solid-state NMR and eventually working with CoSeC.

1. Your background is in physics, completing your MSci at UCL. What was it that introduced you to physics as a discipline and what did you study for your MSci?

I was always interested in how things worked – I loved taking things apart as a kid but not so much putting them back together – and physics is mostly taking things apart to see why they behave the way they do. In school, physics and maths were the subjects I felt most at home with, I think because I have such a poor memory. If you forget something in most subjects (e.g. history) then that’s that; but with physics and maths you have the chance to work out things yourself through reasoning and experience.

I chose to study physics alongside the philosophy of science for my undergrad (the full title was a real mouthful: natural sciences: condensed matter and nano physics with the history, philosophy and social studies of science), having developed a love for philosophy during my last years at high school. UCL was one of the few places that offered such a combination of subjects at the time, and I very much enjoyed approaching science from these two quite complementary perspectives.

2. Your PhD was split between UCL and the Institute for High Performance Computing in Singapore. What was the focus of your thesis?

My thesis was on understanding the relationships between the crystal structure of bismuth ferrite and its various properties. Bismuth ferrite is a fascinating “multiferroic” material with lots of unusual properties, some of which could underlie the next generation of photovoltaics (amongst other applications). I used density functional theory to try to unpick what it was about the arrangement of the Bi, Fe and O atoms in the crystal that resulted in the most interesting of these properties, and how we might modify the arrangement of atoms to improve these properties. As to how I ended up split between UCL and IHPC in Singapore: my supervisors – David Bowler (UCL) and Wei Li Cheah (IHPC) – had a joint project in this area that benefitted from their complementary skills and knowledge in atomistic simulations and perovskite modelling. I also love travelling and had never been to that part of the world, so I jumped at the chance to visit Singapore.

3. Following your PhD, you moved to Switzerland to become a post doc at ETH Zurich in their Material Theory group. What led to this move?

Nicola Spaldin, the chair of the Materials Theory group at ETH Zurich essentially led the renaissance in multiferroic materials and has been at the forefront of understanding these since. Getting to work with someone that defined a field is what initially attracted me to that group, but it turned out to be a fantastic work environment in many other ways also with a strong emphasis on collaboration, work-life balance and personal development. I ended up working on modelling muons in magnetic materials, together with experimental collaborators in Canada.

4. After your post-doc, please could you explain your next career move and how did you eventually start at STFC?

In Switzerland I worked with an experimental community that benefitted from theoretical support. During this, I got very interested in the tools that enabled me to provide that support (even more so that the particular materials we were applying them to). I saw an opportunity to help develop similar tools and to provide support for a related experimental community (solid-state NMR) in the UK and it seemed like a perfect fit.

5. What led to CCP-NC starting? Please could you explain some background as to is origins?

The CCP-NC was established (long-before my time) in 2011 to support the growing, multidisciplinary field of solid-state NMR (SSNMR) by developing and integrating software tools for a community spanning NMR spectroscopists, crystallographers, materials modellers, and application scientists. Its aim is to streamline workflows from first-principles predictions of NMR parameters to the simulation of spectra, enabling direct comparison with experimental results. The project builds on significant UK investment in solid-state NMR infrastructure and is designed to help maintain the UK’s leading role in this emerging area of research.

6. You support CCP-NC through software development and training. Could you please describe what this entails?

The SSNMR community is primarily an experimental community – i.e. experts in running complex experiments to answer questions about the atomic structure of novel materials. Making sense of their spectra, however, often relies on atomistic simulations of the material and calculations of the NMR parameters. We create software tools and training to help the SSNMR community setting up, running and analysis of such “in silico” experiments. We also work on various community-building and best-practice sharing activities such as defining file format standards and organising community events.

7. Please could you tell us what CCP-NC has achieved during its first five years/or main achievements so far?

The first five years was before my time, but I would say some of the main achievements of the CCP-NC are:

• The magres file format: a standard file format specification that captures the essential information for exchanging DFT-predicted SSNMR data.
• MagresView: a web-based graphical user interface for visualising and interpreting computed NMR parameters from DFT calculations.
• MagresPython and later Soprano: python libraries for dealing with crystal structures and accompanying computed NMR data.
• The Magres Database: an open repository of magres files. Version 1 of this database was our first attempt at encouraging the community to adopt the FAIR principles.

8. What has been the value that CoSeC has brought ti CCP-NC?

CoSeC provides:

• a central pool of expertise to help both with practical project-management and communication.
• by bringing together different research communities under one roof, it’s much easier to see how value can be added, e.g. via small tweaks to a workplan means a tool can support more than one community at the same time, or where joint training would make sense etc.
• providing funding and steering, allowing communities to continue to push the boundaries of cutting-edge research whilst aligning with national objectives.   

• How would you encourage young people to get into physics and computing? What words of advice would you give someone hoping to make it into a career?

I see many ways you could go about encouraging young people to get into physics or computing (or computational physics!) e.g.

• connecting to their interests – maybe physics engines in their favourite games
• show them the fun and delight through some hands-on learning (in my case probably trying to take something apart to see how it worked)
• show them real-world applications – how physics discoveries led to many of the technological solutions they enjoy, or how software developers might have gone about building tools/apps that they use
• Breaking down stereotypes e.g. through highlighting role models

9. What has been the most exciting development that you have worked on during your career?

Since starting at STFC, I’ve most enjoyed making our software more user friendly. This has involved going out to research groups to understand what they do and where their bottlenecks are and then trying to resolve these with better code and documentation. An example of this is a new command-line interface that performs common NMR workflows for our python library, Soprano.

10. What do you think the most important recent developments in the field have been? What do you think will be the most exciting and productive areas of research during the next few years?

Although there’s a lot of hype in this area, the rise of machine-learning is poised to dramatically change what is possible in materials science. Developments such as foundation machine-learned interatomic potentials can dramatically expand the length- and timescales accessible for atomistic simulations. However, I think a missing piece of the puzzle is often the interpretation of ML models – i.e. why the models give the results they give. I suspect that is ‘interpretability’ is where the most important physical insights will come from in the next few years.

11. Who have been the people who have been influential in your career?

For very different reasons, I would include my PhD supervisors (David Bowler and Wei Li Cheah), postdoc advisor (Nicola Spaldin) and current line manager (Gilberto Teobaldi) as having been very influential in my career. Another person that has been hugely influential in terms of my career is my wife, Yasmine Al-Hamdani, an excellent computational materials scientist who always gets me to focus on the “so what” questions.

12. If you had not got involved in the field of software development and training, what do you think you would have done? (Is there another field that you could have seen yourself making an impact on?)

I think I would have been either doing something else in computational materials science or maybe have branched out into data science.