X-ray emission spectroscopy: a genetic algorithm to disentangle core–hole-induced dynamics

Abstract

AbstractA genetic algorithm (GA) is developed and applied to make proper connections of final-state potential-energy surfaces and X-ray emission (XES) cross sections between steps in the time-propagation of H-bonded systems after a core–hole is created. We show that this modification results in significantly improved resolution of spectral features in XES with the semiclassical Kramers–Heisenberg approach which takes into account important interference effects. We demonstrate the effects on a water pentamer model as well as on two 17-molecules water clusters representing, respectively, tetrahedral (D2A2) and asymmetric (D1A1) H-bonding environments. For D2A2, the applied procedure improves significantly the obtained intensities, whereas for D1A1 the effects are smaller due to milder dynamics during the core–hole life-time as only one hydrogen is involved. We reinvestigate XES for liquid ethanol and, by properly disentangling the relevant states in the dense manifold of states using the GA, now resolve the important 3a′′ state as a peak rather than a shoulder. Furthermore, by applying the SpecSwap-RMC procedure, we reweigh the distribution of structures in the sampling of the liquid to fit to experiment and estimate the ratio between the main anti and gauche conformers in the liquid at room temperature. This combination of techniques will be generally applicable to challenging problems in liquid-phase spectroscopy.