Ultrafast molecular imaging experiment benchmarks state-of-the-art quantum chemistry predictions

Photochemistry – chemical reactions triggered by the absorption of light – is of huge importance across nature, the atmosphere, technology and medicine. Despite this, theoretically predicting what products will be formed when a molecule absorbs light and the rate at which these products are formed is an enormously challenging task. It requires advanced quantum mechanical simulations of how the nuclei and electrons within a molecule move on ultrafast timescales (a millionth of a billionth of a second). There are many methods to perform these simulations, each of which make different approximations about these complex quantum dynamics. To understand which method works best for a particular problem, we need high-quality experiments which can directly capture these ultrafast motions. The predictions of different simulations can then be ‘benchmarked’ against these experimental results.

Our new paper published in the Journal of Chemical Physics reports experimental results which will benchmark fifteen distinct predictions by leading theoretical quantum chemistry groups submitted as part of a special issue (‘Prediction Challenge: Cyclobutanone Photochemistry’). The theory predictions submitted to the special issue involve dozens of theory groups from around the world. The prediction challenge was ‘blind’ – none of the teams saw any experimental results until our publication. Similarly, we avoided looking at any theory predictions when we analysed the experimental data!

The idea for a theory prediction challenge was born out of discussions at a workshop run by CECAM (Centre Européen de Calcul Atomique et Moléculaire, a European organization dedicated to theoretical chemistry) in 2023, while the idea for the cyclobutanone experiment came from my Marie-Curie Fellowship, which was co-hosted by Dr Thomas Wolf (SLAC) and Dr Rebecca Boll (EuXFEL).

In this work, we used the SLAC Mega-Electronvolt Ultrafast Electron Diffraction (MeV-UED) instrument, part of the Linac Coherent Light Source (LCLS) user facility at SLAC National Accelerator Laboratory in California, to image how the nuclei and electrons rearrange after a prototypical molecule, cyclobutanone, absorbs light. In the experiment, we shine a very short pulse of ultraviolet laser light at cyclobutanone molecules. This is followed by a short burst of electrons, which scatter off the molecules. Electrons which scatter off different atoms within the molecule can interfere, which is reflected in the recorded scattering pattern. Importantly, this interference relates to the distances between atoms in the molecule, and so the experiment directly ‘images’ the structure of the molecule, and how this structure evolves during the reaction. From our experiment, we also observe signal related to the rearrangement of electrons within the molecule after absorbing light. It is this light-induced electronic motion that prompts nuclear motion, rupturing chemical bonds and ultimately breaking the molecule into smaller fragments. Our results give the clearest experimental picture to date of how cyclobutanone evolves after photoexcitation, and so will be used to benchmark the various theory predictions.

The experiment was a highly collaborative effort. It was co-led by Dr Thomas Wolf (SLAC National Laboratory and the Stanford PULSE Institute) and myself (then a Marie-Curie Fellow at SLAC and the EuXFEL facility, and now a Lecturer at Edinburgh), but the collaboration spans many other institutions, including Stanford University, University of Bristol, Brown University, University of Nebraska-Lincoln, Kansas State University. We had to assemble a team with a wide range of expertise to carry out the experiment, which took place over a five-day period of 24 operation at the SLAC MeV-UED facility.

I recently presented the experimental results to a second workshop run by CECAM in Lausanne, Switzerland. This workshop was a great opportunity for the different theoreticians to compare their approaches to the prediction challenge, and to discuss, in light of the experimental data, what worked and what didn’t work.

In my research I am interested in applying advanced experimental techniques to study the photochemistry of gas-phase molecules. Many of these techniques are enabled by the latest technological developments, and promise to allow us to directly ‘image’ ultrafast molecular motion.

I am particularly interested in photochemistry of atmospheric relevance. Cyclobutanone, the focus of the current work, can be seen as a typical volatile organic compound. These species are emitted into the atmosphere by both natural and anthropogenic sources, and understanding how they react following absorption of sunlight is key in assessing their overall impact on the climate and public health.

Acknowledgements

Dr Green was supported by the European Union, through Horizon Europe Project No. 123-CO: 101067645. For full funding acknowledgements for all collaborators in the project, please see the published paper.