In this Communication, for the calculations of the anion [Ba(CO)]− neutral Ba and the carbonyl anion CO− were considered as interacting fragments. The choice was made on the basis of the experimental literature values for the electron affinities (EAs) of Ba (0.108 eV, average value of 2P1/2 and 2P3/2)1 and CO (1.37 eV).2 The authors now learned that the experimental value for the EA of CO is incorrect. High-level quantum chemical calculations suggest that CO has a negative EA,3 i.e. the attached electron in CO− is unbound and the anion is a resonance species that may only be observed as metastable resonating species via sophisticated spectroscopy.6 This means that the bonding analysis of the anion [Ba(CO)]− and the calculation of the bond dissociation energy (BDE) must be carried out using the fragments Ba− and neutral CO. The results of the relevant calculations are reported here. The calculated BDE for the reaction [Ba(CO)]− → Ba− + CO at DLPNO-CCSD(T)/def2-QZVPPD using the CCSD/def2-TZVPP optimized geometries is De = 15.6 kcal mol−1, significantly lower than the previous value of 56.1 kcal mol−1 that was obtained with respect to Ba and CO− at the same level of theory. The numerical results of the EDA-NOCV analysis of [Ba(CO)]− using Ba− in the 2D state and CO as fragments are shown in Table 1, together with the previous data using Ba (1S) and CO−(2Π) as fragments. The intrinsic interaction energy with the former fragments ΔEint = −25.6 kcal mol−1 is also significantly lower than the previous value ΔEint = −71.4 kcal mol−1. BaCO− Fragments Ba− (2D) + CO (1Σ+) Ba (1S) + CO− (2Π) ΔEint −25.6 −71.4 ΔEhybrid[a] 4.2 5.8 ΔEPauli 51.0 52.5 ΔEelstat[a] −36.5 (45.2 %) −66.1 (51.0 %) ΔEorb[a] −44.3 (54.8 %) −63.5 (49.0 %) ΔEorb(1)[b] −27.9 (63.0 %) −41.1 (64.7 %) ΔEorb(2)[b] −11.2 (25.3 %) −16.4 (25.8 %) ΔEorb(rest)[b] −5.2 (11.7 %) −6.0 (9.5 %) ΔEprep 1.9 23.2 BDE 23.7 48.2 The partitioning of the orbital term into the most important pairwise interactions gives in both cases two dominant interactions. The largest contribution ΔEorb(1) comes from the donation of the unpaired π electron at Ba− to the LUMO of CO (Figure 1 a). This was also found in the original work as the strongest orbital interaction in the cation Ba(CO)+, where the electronic reference state of Ba+ cation is the excited 2D state that makes it to become a donor Ba+→CO. The deformation density, which is associated with the second strongest orbital interaction shows that ΔEorb(2) comprises two alterations of the electronic charges (Figure 1 b). There is a polarization at barium, where the occupied 6 s AO of Barium mixes with vacant p(σ) and d(σ) orbitals and a concomitant donation from the HOMO of CO into the σ hybrid AO at Ba. Shape of the most important interacting MOs of fragments and plot of the associated deformation densities Δρ of the most important pairwise orbital interactions ΔEorb(1) and ΔEorb(2) BaCO−. The direction of the charge flow is red→blue. Thus, the most important finding in the original work remains the same: The chemical bonds of barium in the carbonyl complexes [Ba(CO)]+ and [Ba(CO)]− use mainly the 5 d AOs as valence orbitals. The use of the (n−1) d functions as genuine valence orbitals not only by barium but also by the lighter alkaline earth atoms calcium and strontium has in the meantime been supported by the experimental observation of the octacarbonyls E(CO)8 (E = Ca, Sr, Ba), which mimic transition metal complexes.7 The authors are grateful to Prof. Peter Schwerdtfeger, who pointed out the error of choosing the wrong interacting fragments for [Ba(CO)]−.
Read full abstract
Mar 7, 2023
M Krishnan + 1
Open Access
