Articles you may be interested inPerformance of dispersion-corrected double hybrid density functional theory: A computational study of OCShydrocarbon van der Waals complexes Computational and experimental investigation of intermolecular states and forces in the benzene-helium van der Waals complex Comment on "Anisotropic intermolecular interactions in van der Waals and hydrogen-bonded complexes: What can we get from density-functional calculations?" [J.The applicability of various density functional theory ͑DFT͒ methods to describe the anisotropy of the intermolecular potential energy surfaces of hydrogen-bonded ͓OH Ϫ -H 2 O, (H 2 O͒ 2 ͔ and van der Waals ͓CO-H 2 O, He-CO 2 ] complexes has been tested by comparison with supermolecule CCSD͑T͒ ͑coupled-cluster method restricted to single, double, and noniterative triple excitations͒ and perturbational SAPT ͑symmetry-adapted perturbation theory͒ results computed for the same geometries and with the same basis sets. It is shown that for strongly bound ionic hydrogen-bonded complexes, like OH Ϫ -H 2 O, hybrid approaches provide accurate results. For other systems, including the water dimer, the DFT calculations fail to reproduce the correct angular dependence of the potential surfaces. It is also shown that a hybrid functional adjusted to reproduce the CCSD͑T͒ value of the binding energy for the water dimer produces results worse than the standard hybrid functionals for OH Ϫ -H 2 O, and fails to describe the correct anisotropy of the CO-H 2 O interaction.

Calculations have been performed on 10 structures of the cluster H+(H20)5. It is shown that the most stable ones are an open (Eigen) and a cyclic four-membered-ring structure very close in energy and possibly degenerate. This can explain that different structures were proposed by experimentalists. The easy evolution of some structures into others is likely related to the nature of the first solvation shell in larger clusters or solutions. Vibrational frequencies, useful to interpret experimental data, are computed for the two most stable structures. The Problem: Structure of H+(HzO)s and the First Solvation ShellIn 1954, Eigen et al.' proposed as a hydration model for the proton in aqueous solution and ice an oxonium ion H30+ surrounded by three water molecules in a first solvation shell. In some early theoretical work (1956) the presence of a fourth water molecule in the first solvation shell has also been suggested;* three water molecules are hydrogen bonded with the three hydrogen atoms of the ion, while the fourth water molecule is located above the oxygen atom (Figure 1, protondonor structure 1). Since then, there were many proposal^^-^^ on the structure and coordination number, and presently the issue is not settled, neither theoretically nor experimentally. Some of these studies are concemed with the cluster H30+(H20)4, others with larger systems.The first attempt by Newton et aL3 in 1971 to study such systems with ab initio quantum mechanical calculations was restricted to small hydrates involving, at best, four water molecules and the oxonium ion. Using a 4-31G basis set at the Hartree-Fock level, the authors optimized a few structures for the system &O+(H*0)3 and then added to it a fourth molecule. Within these limitations the most stable structure has three water molecules in the Eigen-like structure first shell and a fourth water molecule in the second shell, hydrogen bonded to one water molecule of the first shell with an 0. * 0 distance arbitrarily chosen (Figure 1, structure 2). The interpretation of an infrared spectrum published some time later is based on such a structure.8 However, two experimental paper^^.^ suggested that the fist solvation shell has four water molecules. In particular, from X-ray and thermal neutron studies of hydrochloric acid solutions at 20 "C, Triolo et a1.6 proposed a charge-dipole complex (Figure 1, structure 3). In a further work, ' Newton (1977) extended his studies to structures 1 and 3 in order to check this assumption. He found that neither of these two structures were stabilized with respect to H30+(H20)3 f Hz0, with a more favorable situation for the charge-dipole complex structure 3 than for the hydrogen bonded structure 1. A complete optimization of these structures with ab initio calculations was unfortunately not possible at the time, and the true minima might have been missed. Furthermore, it must be pointed out that the structure of the first shell may be different 'Abstract published in Advance ACS Abstracts, April 15, 1995.

The nature and importance of nonadditive three-body interactions in the ionic OH−(H2O)2 cluster have been studied by supermolecule Mo/ller–Plesset (MP) perturbation theory and coupled-cluster method, and by symmetry-adapted perturbation theory (SAPT). The convergence of the SAPT expansion was tested by comparison with the results obtained from the supermolecule Mo/ller–Plesset perturbation theory calculations through the fourth order (MP2, MP3, MP4SDQ, MP4), and the coupled-cluster calculations including single, double, and approximate triple excitations [CCSD(T)]. It is shown that the SAPT results reproduce the converged CCSD(T) results within 10%. The SAPT method has been used to analyze the three-body interactions in the clusters OH−(H2O)n, n=2,3,4,10, with water molecules located either in the first or the second solvation shell. It is shown that at the Hartree–Fock level the induction nonadditivity is dominant, but it is partly quenched by the Heitler–London and exchange-induction/deformation terms. This implies that the induction energy alone is not a reliable approximation to the Hartree–Fock nonadditive energy. At the correlated level, the most important contributions come from the induction-dispersion and the MP2 exchange energies. The exchange-dispersion and dispersion nonadditivities are much smaller, and for some geometries even negligible. This suggests that it will be difficult to approximate the three-body potential for OH−(H2O)2 by a simple analytical expression. The three-body energy represents only 4%–7% of the pair CCSD(T) intermolecular energy for the OH−(H2O)2 cluster, but can reach as much as 18% for OH−(H2O)4. Particular attention has been paid to the effect of the relaxation of the geometry of the subsystems.