ALBERT

Notes: Ab-initio pseudopotential two-configuration self-consistent field followed by extensive variational and perturbational second order Møller–Plesset multireference configuration interaction calculations using localized molecular orbitals were performed to characterize the structure and adiabatic potential energy curves of the three lowest (X 1Σ+, 3Σ+, and 1Σ+) purely electronic states of the AgF molecule. Spin-orbit interactions were introduced semiempirically in a second step. The very strong coupling of the neutral Ag(4d105s1)F(2s22p5) and ionic Ag+(4d95s1)F−(2s22p6) configurations at rather short internuclear distance for both excited 3,1Σ+ states is responsible for the appearance of very shallow minima, thus leading to a limited number of stable vibrational levels for these excited states as suggested previously for the AO+ state. In contrast with the CuF molecule, where only the ionic configuration Cu+(3,1D)F−(1S) is present in the 3,1Σ+ states, this coupling of ionic and neutral structures in AgF is explained by the relative positions of the valence orbital energies of the neutral Cu and Ag atoms with respect to the 2p level of the halogen atom. These results lead to the assignment of the observed AO+–X 1Σ+ transition as a 1Σ+–1Σ+ type transition.The very recently observed aΩ1 and A'Ω1 states are shown to be, respectively, the Ω=O− and Ω=1 spin–orbit components of the 3Σ+ state, which justifies the relabeling of aΩ1 into a aΩO−. The calculated spin–orbit-induced splitting between these two components is in excellent agreement with the observed one after reconsidering spectroscopic data. For all these states the calculated spectroscopic constants are in good agreement with available experimental data. The fourth experimental state, BO+, is probably not correlated with the 3Π valence state as previously suggested but it could rather correspond to a Rydberg ionic state involving the Ag+(4d95p)F−(2s22p6) structure. © 1995 American Institute of Physics.

Notes: The two lowest-lying (X and 2)1Σ+ states, the first 3Σ+, 3Π, 1Π, 3Δ, and 1Δ excited states of the AgCl molecule have been studied through extensive complete active space SCF+averaged coupled pair functional calculations, using a 19-active-electron relativistic effective core potential (RECP) for Ag, a 7-active electron RECP for chlorine and large optimized Gaussian basis sets for both atoms. The 2 1Σ+ and 3Σ+ excited states present shallow relative minima very near the equilibrium geometry of the ground state, while the lowest 3,1Π states were found to be totally repulsive in the internuclear range studied. The 3,1Δ states present very shallow minima around 5.2 a.u. The calculated spectroscopic constants for the ground- and excited states are compared with the available experimental data and have been found in good agreement. Even though the 3Π state is repulsive, it lies very close in energy to the 2 1Σ+ one near the equilibrium geometry of the ground state; thus, a strong 3Π–2 1Σ+ mixture through the spin–orbit interaction is predicted to occur that will lead to the fine-structure B state responsible for the recently revised B←X transitions in AgCl. The C–X transition observed at 43 500 cm−1, appears now to arise from a higher-lying root of the 1Σ+ or Π manifolds, perhaps the third 1Σ+ root, while the D–X system seems to arise from the 3Δ←1Σ+ transition around 49 000 cm−1. © 2002 American Institute of Physics.

Notes: Very accurate ab initio electronic + spin-orbit calculations of the lowest-lying states of the Ag atom and Ag+ cation have been performed through the CASSCF + ACPF + EPCISO method, using the Stuttgart small-core (19 active electrons) relativistic effective core potential (RECP) as well as its associated 2D spin-orbit effective potential. An ad hoc spin-orbit P-symmetry pseudopotential for the 2P state adapted to this 19-e RECP and basis set was extracted. The Stuttgart basis set was augmented to a large valence Gaussian basis set (8s8p7d3f3g/6s6p4d3f3g) in order to reproduce at best the experimental 2S-2D and 2S-2P transition energies as well as the ionization potential (IP) of Ag, which play a crucial role for the accurate description of the spectroscopy in silver-containing molecular systems. A detailed discussion on the multiple schemes used to deal with the differential d10 vs d9 electronic correlation for these two excited states is given. The role of the 4s and 4p (core) shells on the 2S-2D and 2S-2P transition energies and the IP is carefully studied and discussed. The core–core correlation is found to play a minor role while an insufficient treatment of the core-valence electronic correlation is responsible for the main differential d10 vs d9 correlation energy error between the 2S-2D and 2S-2P transition energies. For the neutral atom, the 2D5/2-2D3/2 and 2P3/2-2P1/2 splittings are in excellent agreement with the experimental ones. However, the relative calculated energetic ordering for the 2D5/2,2D3/2,2P3/2, and 2P1/2 fine structure components is critically dependent on the J-averaged purely electronic ACPF 2P and 2D energies of the parent states. The 3D fine-structure splitting for the ion is also found in good agreement with the experiment. © 2001 American Institute of Physics.

Notes: Multiconfigurational self-consistent field–multireference configuration interaction calculations followed by the inclusion of spin–orbit corrections were performed to study the 18 lowest (Ω) fine structure states arising from the 9 lowest purely electronic ΛSΣ states of the CuI molecule. An important difference with previous studies on the first excited states of CuF and CuCl is found regarding the relative position of the first neutral states [with the Cu(d10s1)I(s2p2σp3π) configuration] even before taking into account the spin–orbit effects. We have been able to assess the nature of the observed transitions and verified that the first (A) and third (D) systems arise both from a Σ←Π-type transition while the second (C) and fourth (E) systems arise from a Σ←Σ type. The adiabatic potential energy curves for all the Ω states are given as well as the calculated spectroscopic constants of the states having a minimum in the studied range of interatomic distance. We also confirm the hypothesis put forward by Delaval et al.[Chem. Phys. Lett. 139, 212 (1987)] stating that the X←A and X←C transitions are spin forbidden. We explain this through the mixing of singlet and triplet states owing to the spin–orbit coupling of iodine (for the A 1,3Πn state) and that due to copper (for the C 3,1Σ+ state). Finally, an avoided crossing is predicted between the third and fourth Ω=1 fine structure states and a dramatic change of character for the D(1Π) ionic state towards a neutral Π configuration at long interatomic distance.

Notes: Spontaneous radiative lifetimes of the observed electronic states of CuCl have been calculated using a model including spin–orbit interaction mixings within the Cu+(3d94s)Cl−(3s23p6) structure. The required wave functions have been determined semiempirically and locations of the as-yet unobserved components have been estimated. For the A 3Σ1+, C 3Π0e, D 1Π1, E 1Σ0+, and F 3Δ1 components, the lifetimes calculated in the simple single-configuration pure-precession approximation with a slight modification in the Ω=0e block are in good agreement with earlier experimental measurements. On the contrary, for the B 3Π1 component, there is a marked discrepancy that cannot be reduced except by assigning very unlikely values to the off-diagonal spin–orbit parameters.

Notes: The spectroscopic properties of the three lowest-lying (X,2 and 3)1Σ+, the first 3Σ+, the two lowest-lying (1 and 2)3Π, the first 1Π, and the 3,1Δ states of the AgF molecule have been studied through extensive CASSCF (complete active space self-consistent field)+CASPT2 (complete active space second-order perturbational) calculations, using a 19-active-electron relativistic effective core potential for Ag and large Gaussian basis sets for both atoms. Strong mixtures of the Ag+(4d95s1)F−(2s22p6) ionic and Ag(4d95s2)F(2s22p5) or Ag(4d105s1)F(2s22p5) neutral configurations were found for the 3Σ+, 2 1Σ+, and 1 3Π states between 4.0 and 4.4 a.u., while for the higher lying states no evident neutral-ionic crossings were found. This leads to curves that present local maxima at 4.3 a.u. for the 2 1Σ+ and 3Σ+ states as well as for the 1 3Π state at 4.0 a.u. The 2 3Π excited state shows the lowest ionic character of all the states. The calculated spectroscopic constants for all the studied states are reported and found in good accordance with available experimental data. The question of the nature of the electronic parent state of the observed B0+ state, responsible for the most intense transition and which is the shortest lived excited state of AgF, is thoroughly addressed in the light of the present results. They clearly indicate that the B0+ state is not correlated with the Rydberg Ag+(4d95p1)+F−(2s22p6) ionic structure, as previously proposed [J. Chem. Phys. 102, 4482 (1995)]. Since the 2 1Σ+ state has been shown to be the ΛSΣ electronic parent state of the fine-structure A0+ state (these results confirm this idea), and given the difference between the calculated Te (1513 cm−1) of the 2 1Σ+ and 1 3Π states, these calculations point to this latter state as the ΛSΣ parent of the experimental B0+ state. At this level of calculation, the next higher lying state that could contribute (3 1Σ+) through spin–orbit couplings to this B0+ state lies more than 8000 cm−1 away. This, however, is not consistent with the accurately measured radiative lifetimes of 7.1 μs (A′Ω1), 9.1 μs (aΩ1), 240 ns (A0+), 21 ns (B0+) for the four observed excited states, which seem to indicate that the two Ω=0+ excited states are of singlet character. Therefore, only a theoretical study including a substantially more accurate and complete account of the electronic+spin–orbit interactions will yield a reliable answer to this complex problem in the spectroscopy of AgF. © 2000 American Institute of Physics.