RESUMO
We present molecular-frame measurements of the recombination dipole matrix element (RDME) in CO2, N2O, and carbonyl sulfide (OCS) molecules using high-harmonic spectroscopy. Both the amplitudes and phases of the RDMEs exhibit clear imprints of a two-center interference minimum, which moves in energy with the molecular alignment angle relative to the laser polarization. We find that whereas the angle dependence of this minimum is consistent with the molecular geometry in CO2 and N2O, it behaves very differently in OCS; in particular, the phase shift which accompanies the two-center minimum changes sign for different alignment angles. Our results suggest that two interfering structural features contribute to the OCS RDME, namely, (i) the geometrical two-center minimum and (ii) a Cooper-like, electronic-structure minimum associated with the sulfur end of the molecule. We compare our results to ab initio calculations using time-dependent density functional theory and present an empirical model that captures both the two-center and the Cooper-like interferences. We also show that the yield from unaligned samples of two-center molecules is, in general, reduced at high photon energies compared to aligned samples, due to the destructive interference between molecules with different alignments.
RESUMO
We discuss our implementation and application of time-dependent density functional theory (TDDFT) to core-level near-edge absorption spectroscopy, using both linear-response (LR) and real-time (RT) approaches. We briefly describe our restricted excitation window TDDFT (REW-TDDFT) approach for core excitations, which has also been reported by other groups. This is followed by a detailed discussion of real-time TDDFT techniques tailored to core excitations, including obtaining spectral information through delta-function excitation, postprocessing time-dependent signals, and resonant excitation through quasi-monochromatic excitation. We present results for the oxygen K-edge of water and carbon monoxide; the carbon K-edge of carbon monoxide; the ruthenium L3-edge for the hexaamminerutheium(III) ion, including scalar relativistic corrections via the zeroth order regular approximation (ZORA); and the carbon and fluorine K-edges for a series of fluorobenzenes. In all cases, the calculated spectra are found to be in reasonable agreement with experimental results, requiring only a uniform shift ranging from -4 eV to +19 eV, i.e., on the order of a few percent of the excitation energy. Real-time TDDFT visualization of excited state charge densities is used to visually examine the nature of each excitation, which gives insight into the effects of atoms bound to the absorbing center.
RESUMO
The low-lying excited states (La and Lb) of polyacenes from naphthalene to heptacene (N = 2-7) are studied using various time-dependent computational approaches. We perform high-level excited-state calculations using equation of motion coupled cluster with singles and doubles (EOMCCSD) and completely renormalized equation of motion coupled cluster with singles, doubles, and perturbative triples (CR-EOMCCSD(T)) and use these results to evaluate the performance of various range-separated exchange-correlation functionals within linear-response (LR) and real-time (RT) time-dependent density functional theories (TDDFT). As has been reported recently, we find that the range-separated family of functionals addresses the well-documented TDDFT failures in describing these low-lying singlet excited states to a large extent and are as about as accurate as results from EOMCCSD on average. Real-time TDDFT visualization shows that the excited state charged densities are consistent with the predictions of the perimeter free electron orbital (PFEO) model. This corresponds to particle-on-a-ring confinement, which leads to the well-known red-shift of the excitations with acene length. We also use time-dependent semiempirical methods like TD-PM3 and TD-ZINDO, which are capable of handling very large systems. Once reparametrized to match the CR-EOMCCSD(T) results, TD-ZINDO becomes roughly as accurate as range-separated TDDFT, which opens the door to modeling systems such as large molecular assemblies.