https://orcid.org/0000-0002-8326-3097
This autumn “Simulation Meets Neutron Spectroscopy” (siMol) took place in the Oxfordshire countryside. This was a new format of conference jointly run by the Molecular Spectroscopy Group (MSG) at the ISIS Neutron and Muon Facility and the Ada Lovelace Centre, of the Science and Technologies Facilities Council, UK. The Ada Lovelace centre works with large-scale facilities across the UK to aid in the handling and interpretation of the data which these large scale facilities produce. It has recently been noticed within the MSG that the missing ingredient to overcome the barrier to publication is often the application of simulation methods to provide a fuller understanding of the data. However, despite the abundance of simulation knowhow within the UK, there is often difficulty linking together those in the field of Neutron Spectroscopy, with simulation specialists that share their research interests. The underlying task for the conference was therefore to bring together the two communities of neutron spectroscopy and atomistic simulation, and showcase existing examples of successful collaboration, in order to stimulate fresh discussion and foster new links.
Being a first time conference, the organisers were unsure of what level of interest to expect, but were delighted to have over 80 attendees present. The meeting kicked off with a relaxed evening mixer event, which saw brilliant attendance, and acted as an icebreaker. Speakers came from a broad range of institutions across Europe and represented the disciplines of Chemistry, Energy Materials, Catalysis, Biology and Soft Matter. The balance between experimental and simulation was remarkably good, with many great examples of the crucial role that traditional techniques such as classical molecular dynamics and density functional theory play in the interpretation of neutron data. However, most notably there was also a new interest in machine learning (ML) techniques and the role that they may play in the near future for the rapid and semi-autonomous analysis of data ( and ). In particular, with high-throughput experiments on large classes of materials becoming more common, these ML techniques could play a pivotal role in tackling the current issues with interpreting these vast data sets.
Figure 1. A photo taken during a heated discussion session.
Figure 2. A photo taken during Prof. Simone Meloni’s invited talk.
With such a wide range of research areas represented, a blitz presentation session for both the experimentalists and simulators allowed attendees to quickly identify those with overlapping research areas. Discussions were then able to overspill into a poster session (), where attendees displayed their field of research, before conversation could continue at the conference dinner. The plenary lecture saw long-time simulation advocate; Sir Richard Catlow, present many real world examples within the field of catalysis, where simulation was an essential tool for interpreting experimental results. Focus was also given to the drastically enhanced atomic insight that simulation can provide beyond that obtainable from experiment, provided that the model in question has been rigorously validated by experiment.
A session focusing on major developments within the Molecular Spectroscopy group at ISIS showcased planned upgrades of their instrument suite. The QENS spectrometers OSIRIS and IRIS will both receive modern primary flight path guides, which will result in a significant increase in flux on both instruments. A Si(111) analyser is also being installed in the OSIRIS secondary spectrometer to extend the temporal range to ∼0.4 ns. The secondary spectrometer of the broadband inelastic neutron spectrometer TOSCA will be replaced by an advanced curved analyser with ∼6 sr coverage compared to the present 1 sr. Together with optimization of the Be filter and detectors, this will enable an overall gain factor of ∼10 with lower background and better resolution. The neutron Compton scattering spectrometer, VESUVIO, is being upgraded with increased detector coverage and a dedicated gamma detector for neutron resonant capture analysis, which will mean that heavy elements can also be studied. In addition to these planned upgrades, there was strong support from the community for a proposal for a dedicated QENS instrument with polarisation analysis (SHERPA).
Figure 3. A photo taken during the poster session of the conference.
The question segments generated intense debate, with members of both communities bringing to the floor the issues that may prevent others from working across the two disciplines. In particular, the perceived exotic nature of neutron techniques can overshadow the level of complementarity that it provides to simulation. Equally, the high degree of technical ability required for many simulation methods can be a steep learning curve/barrier for experimentalists wanting to adopt these methods in their own research. A seemingly essential component therefore is to have as many “bi-lingual” advocates of these methods to bridge the required technical expertise.
After the meeting and with some time to reflect on the discussions, it seems there are a few take-home messages. In order to increase both the productivity and quality of research output at neutron facilities in the short term, there must be a concerted effort to set up a platform via which experimentalists can interact/meet simulators with shared research interests. With the Ada Lovelace centre growing as a UK simulation hub, it seems timely to set up these formal links. However, in the longer term, ISIS and other neutron facilities should set aside beam-time dedicated to assessing the role that ML techniques can play as we move into a new era of high-throughput measurement of evermore complex systems.
We would like to acknowledge the funding from both the Molecular Spectroscopy Group and the Ada Lovelace centre in making this conference a reality.