Correction: Dyson orbitals within the fc-CVS-EOM-CCSD framework: theory and application to X-ray photoelectron spectroscopy of ground and excited states

An error was detected, and it is hereby corrected, in the computation of the photoelectron spectra obtained by ionizing core electrons of the two lowest excited states of uracil. A re-examination of the Hartree–Fock molecular orbitals in the 6-311(2+,+)G** (uncontracted) basis (which was used in the calculations) has revealed that the SCF procedure did not converge to the desired lowest triplet states of n p * and pp * characters, but instead found nRy and p Ry solutions. While the character of the hole orbital in these states is the same as in the S 1 and S 2 states, the character of the particle orbital is different. Thus, the XPS reported in ref. 1 do not correspond to the XPS of the S 1 and S 2 excited states. This unfortunate problem occurred because of using a non-standard basis set (uncontracted), which affected the default guess in the SCF procedure. Because of the multiple diffuse functions present in the basis, the SCF calculation of the lowest triplet inadvertently converged to a Rydberg state instead of the valence state. This problem can be easily corrected by using molecular orbitals from the ground-state calculation together with the correct user-defined orbital occupation and the MOM algorithm.

An error was detected, and it is hereby corrected, in the computation of the photoelectron spectra obtained by ionizing core electrons of the two lowest excited states of uracil.A re-examination of the Hartree-Fock molecular orbitals in the 6-311(2+,+)G** (uncontracted) basis (which was used in the calculations) has revealed that the SCF procedure did not converge to the desired lowest triplet states of np* and pp* characters, but instead found nRy and pRy solutions.While the character of the hole orbital in these states is the same as in the S 1 and S 2 states, the character of the particle orbital is different.Thus, the XPS reported in ref. 1 do not correspond to the XPS of the S 1 and S 2 excited states.This unfortunate problem occurred because of using a non-standard basis set (uncontracted), which affected the default guess in the SCF procedure.Because of the multiple diffuse functions present in the basis, the SCF calculation of the lowest triplet inadvertently converged to a Rydberg state instead of the valence state.This problem can be easily corrected by using molecular orbitals from the ground-state calculation together with the correct user-defined orbital occupation and the MOM algorithm.
The results of the calculations of the spectra with the correct orbital occupations are reported below.Fig. 1-3 show the amended spectra at the carbon, nitrogen, and oxygen K-edges, respectively.They replace Fig. 5-7 in the original article.Tables 1  and 2 summarize the raw data (ionization energies and squared Dyson orbital norms) for the S 1 and S 2 excited states, and amend the results in Tables 5 and 6 in the original article.
Below we present the amended discussion of the computed spectra.Importantly, the main conclusion of ref.
1 remains valid: The XPS spectroscopy on the oxygen edge can be used to distinguish different electronic states in uracil.
At all three edges, the XPS spectra of the two excited states are now found significantly closer to that of the ground state.In most cases, the binding energies of the core electrons are slightly red-shifted relative to the corresponding S 0 values, indicating a destabilization of the core electrons in the excited states.
At the C K-edge, see Fig. 1, the ordering of the A and B bands is reversed in the S 2 state, indicating the stabilization (higher binding energy) of the 1s electron of the C atom at position A and destabilization of the 1s electron of the C atom at position B.
At the N K-edge (cf.Fig. 2), the ordering of the A and B bands changes in the S 1 state, primarily due to a larger destabilization of the 1s core electron of the N atom at position B than for the one at position A.
The significant overlap of the C and N edge spectra corresponding to the two excited states makes them less attractive probes since no specific spectral signature region emerges.
At the oxygen K-edge, see Fig. 3, substantial differences between the spectra of the three states (ground state, S 1 and S 2 ) are found.In line with the localized nature of the np* state (see Table 3 in ref. 1), the 1s electron of the oxygen in the A position is strongly stabilized in the S 1 state, versus a very modest destabilization of the one in B position, with a separation between the two IEs of 4-5 eV.This confirms our previous conclusion 1 that the oxygen K-edge region would be the most interesting edge for probing the excited-state dynamics of the S 1 state by UV-pump/XPS-probe spectroscopy.

Table 1
Uracil.Ionization energies (IE, eV) and squared norms of Dyson orbitals for the first excited state (S 1 ).The results are reported at the Franck-Condon geometry (no label) and at state-specific zero-gradient, stationary point (label sp) geometry

Table 2
Uracil.Ionization energies (IE, eV) and squared norms of the second excited state (S 2 ).The results are reported at the Franck-Condon geometry (no label) and at state-specific zero-gradient, stationary point (label sp) geometry