Research
Dense plasmas made with x-rays. An intense x-ray pulse can heat a solid to over a million degrees in tens of femtoseconds, before it has time to expand. We use this to make uniform plasmas at solid density and measure their basic properties: how ionization potentials shift in a dense environment, how fast electrons collide and ionize, and how the electronic structure changes as the system heats. Our experiments run at the Linac Coherent Light Source in California, the European XFEL and FLASH in Hamburg, and SACLA in Japan.
Matter at planetary conditions. We drive shock waves into materials with high-power lasers, and use femtosecond x-ray diffraction and spectroscopy to follow how their structure and electronic state change at the pressures found deep inside planets. Recent work has focused on iron and its oxides, up to the conditions of Earth's outer core.
Theory and simulation of dense matter. We develop first-principles methods for warm dense matter, from density functional theory and quantum Monte Carlo to wave-packet molecular dynamics, together with collisional-radiative atomic kinetics for modelling x-ray spectra. We use these approaches to help interpret our experiments and to predict properties that cannot yet be measured directly.
Machine learning and differentiable physics. We build machine-learning models with the physics built in: density functionals learned from experimental data through fully differentiable DFT, emulators that run millions of times faster than the simulations they replace, and reduced models of electron heat transport for fusion design. We validate these models on real data, both experimental and simulated, and always ask how much we can trust what they predict.
Publication Highlights (full list is on Google Scholar or ORCID)
- C. Crépisson et al., Structural evolution of iron oxides melts at Earth's outer-core pressures, Nat. Commun. 17, 8527 (2026).
- A. Amouretti et al., Isostructural phase transition of Fe₂O₃ under laser shock compression, Phys. Rev. Lett. 134, 176102 (2025).
- M. F. Kasim et al., Building high accuracy emulators for scientific simulations with deep neural architecture search, Mach. Learn.: Sci. Technol. 3, 015013 (2022).
- M. F. Kasim and S. M. Vinko, Learning the exchange-correlation functional from nature with fully differentiable density functional theory, Phys. Rev. Lett. 127, 126403 (2021).
- S. M. Vinko, O. Ciricosta and J. S. Wark, Density functional theory calculations of continuum lowering in strongly coupled plasmas, Nat. Commun. 5, 3533 (2014).
- S. M. Vinko et al., Creation and diagnosis of a solid-density plasma with an X-ray free-electron laser, Nature 482, 59 (2012).