Relativistic Embedded Equation-of-Motion Coupled-Cluster Approach to the Core-Ionized States of Actinides: A Case Study of Uranyl(VI) in Cs$_2$UO$_2$Cl$_4$

TL;DR

This paper develops a relativistic equation-of-motion coupled-cluster method to accurately compute core-ionization energies of uranyl in different environments, emphasizing the importance of electron correlation, relativistic effects, and environmental influences.

Contribution

It introduces a novel relativistic CVS-EOM-CC approach for actinide core states and evaluates environmental effects using embedding techniques, advancing spectroscopic simulations.

Findings

Relativistic effects significantly influence core-ionization energies.

Embedding models show notable changes in binding energies due to ligands.

Computed spin-orbit splittings agree within 0.7 eV of experimental data.

Abstract

We investigate the core-level ionization energies of the bare uranyl ion (UO$_2^{2+}$) and its interaction with X-rays when it is hosted in the Cs$_2$UO$_2$Cl$_4$ crystalline environment using a recent implementation of the core-valence-separated relativistic equation-of-motion coupled-cluster method (CVS-EOM-CC). Our study evaluates different relativistic Hamiltonians, assesses basis set and virtual space truncation effects, and examines the role of orbital correlation and relaxation in simulating the spectroscopic observables. The results of this investigation highlight the importance of computing two-electron interactions beyond the zeroth-order truncation (i.e. the Coulomb term) when working in the tender and hard X-ray ranges. Additionally, we compare different structural models using the frozen density embedding method (FDE). By contrasting the bare and embedded uranyl models, we observe significant changes in binding energies, highlighting the influence of the equatorial ligands of the uranyl ion on its spectroscopic observables. A comparison between the embedded uranyl and supermolecular systems, excluding the cesium atoms, reveals systematic differences, with binding energy variations from experimental data remaining within 10 eV. Notably, the computed spin-orbit splittings for U $4d$ and $4f$ deviate by less than 0.7 eV, demonstrating the validity of this protocol for computing binding energies in the soft X-ray range.