Using machine learning to determine the temperature-dependent opacity in plasmas

The free-free opacity in plasmas is fundamental to our understanding of energy transport in stellar interiors and for inertial confinement fusion. However, theoretical predictions in dense plasmas are conflicting and there is a dearth of accurate experimental data for direct model validation. Here we present a novel functional exploration approach to extract the temperature-dependent absorption coefficient of a warm dense aluminium plasma for the first time. The plasma was created via isochoric heating at the XUV free-electron laser FLASH, and was probed with femtosecond time resolution showing the separate contributions to the opacity from hot electrons and ions. We find a pronounced enhancement of the opacity as the plasma electrons are heated to temperatures around the Fermi energy, with further opacity rises observed on ps timescales due to ion heating, melt, and the formation of the warm dense state.

Vinko et al., Phys. Rev. Lett. 124, 225002 (2020), DOI: 10.1103/PhysRevLett.124.225002