Interaction Energy Analysis of Monovalent Inorganic Anions in Bulk Water Versus Air/Water Interface

Abstract

Soft anions exhibit surface activity at the air/water interface that can be probed using surface-sensitive vibrational spectroscopy, yet the statistical mechanics behind this surface activity remains a matter of debate. Here, we examine the nature of anion–water interactions at the air/water interface using a combination of molecular dynamics simulations and quantum-mechanical energy decomposition analysis based on symmetry-adapted perturbation theory. Results are presented for a set of monovalent anions including Cl–, Br–, I–, CN–, OCN–, SCN–, NO2–, NO3–, and ClOn– (n = 1, 2, 3, 4), several of which are archetypal examples of surface-active species. In all cases, we find that average anion–water interaction energies are systematically larger in bulk water although the difference (with respect to the interaction energy in the interfacial environment) is well within the magnitude of the instantaneous fluctuations. Specifically for the surface-active species Br–(aq), I–(aq), ClO4–(aq), and SCN–(aq), and also for ClO–(aq), the charge-transfer (CT) energy is found to be larger at the interface than it is in bulk water, by an amount that is greater than the standard deviation of the fluctuations. The Cl– ion also has a slightly larger CT energy at the interface but NO3–(aq) does not; these two species are borderline cases where consensus is lacking regarding their surface activity. However, CT stabilization amounts to < 20% of the total induction energy, for all of the ions considered here, and CT-free polarization energies are systematically larger in bulk water, again in all cases, so the role of these effects in soft anion surface activity remains unclear. This analysis complements our recent work suggesting that the short-range solvation structure around these ions is scarcely different at the air/water interface from what it is in bulk water. Together, these observations suggest that changes in first-shell hydration structure around soft anions cannot explain observed surface activities.

State-specific solvation for restricted active space spin-flip (RAS-SF) wave functions based on the polarizable continuum formalism

“…In these calculations, which were performed at the B3LYP/cc-pVTZ level, the charge constraint is implemented using Becke populations with atomic size corrections. 39,40 B. State-relaxed equilibrium procedure…”

Comprehensive Basis-Set Testing of Extended Symmetry Adapted Perturbation Theory

“…(This was seen previously in tests for anion-water clusters. 60 ) However, other good-performing basis sets, including 6-311++(3df,2pd), appear to be converged in the sense that the electrostatic energies are indistinguishable from results obtained in the def2-QZVPD basis set.…”

“…(Again, this is consistent with previous experience for hydrated anions. 60 ) The exchange energy, equivalent to Pauli repulsion between filled orbitals, generally increases in magnitude with the completeness of the basis, as the tails of the density become better described. The Karlsruhe and Dunning basis sets generally afford comparable values of E exch except that the former exhibits slightly larger exchange energies for the strongest hydrogen bonds.…”

“…54,[56][57][58][59] For large systems, the monomer-based nature of XSAPT calculations makes this approach more affordable even than supramolecular DFT. 52,54,55 Some limited basis-set testing of XSAPT methods has been reported in previous work, 42,54,55,60 but only for total interaction energies whereas as herein we examine convergence of individual energy components, which allows us to consider whether earlier tests may have benefited from error cancellation amongst energy components that may exhibit different convergence behavior. The basis-set dependence of traditional SAPT has also been carefully evaluated, 25 however the especially slow convergence of the dispersion energy in the traditional ap-proach means that those tests are not directly applicable to XSAPT.…”

“…Indices along the horizontal axis refer to the way that the dimers were ordered in the original work of Ref 75. and the three color-coded regions delineate the hydrogen-bonded subset (dimers 1-23), the dispersion-bound subset, and the subset of mixed-influence dimers(47)(48)(49)(50)(51)(52)(53)(54)(55)(56)(57)(58)(59)(60)(61)(62)(63)(64)(65)(66). Errors are defined with respect to the CCSD(T)/CBS benchmarks in Ref.75.…”