Benchmark Study\nof Core-Ionization Energies with the\nGeneralized Active Space-Driven Similarity Renormalization Group

X-ray photoelectron spectroscopy (XPS) is a powerful\nexperimental\ntechnique for probing the electronic structure of molecules and materials;\nhowever, interpreting XPS data requires accurate computational methods\nto model core-ionized states. This work proposes and benchmarks a\nnew approach based on the generalized active space-driven similarity\nrenormalization group (GAS-DSRG) for calculating core-ionization energies\nand treating correlation effects at the perturbative and nonperturbative\nlevels. We tested the GAS-DSRG across three data sets. First, the\nvertical core-ionization energies of small molecules containing first-row\nelements are evaluated. GAS-DSRG achieves mean absolute errors below\n0.3 eV, which is comparable to high-level coupled cluster methods.\nNext, the accuracy of GAS-DSRG is evaluated for larger organic molecules\nusing the CORE65 data set, with the DSRG-MRPT3 level yielding a mean\nabsolute error of only 0.34 eV for 65 core-ionization transitions.\nInsights are provided into the treatment of static and dynamic correlation,\nthe importance of high-order perturbation theory, and notable differences\nfrom density functional theory in the predicted energy ordering of\ncore-ionized states for specific molecules. Finally, vibrationally\nresolved XPS spectra of diatomic molecules (CO, N<sub>2</sub>, and\nO<sub>2</sub>) are simulated, showing excellent agreement with experimental\ndata.