The water dimer: correlation energy calculations

Calorimetric investigations of hydrogen bonding in binary mixtures containing pyridine and its methyl-substituted derivatives. II. The dilute solutions of methanol and 2-methyl-2-propanol

“…Theoretical predictions usually give the energy from (15.1 to 24.1) kJ AE mol À1 , most commonly ca. 20 kJ AE mol À1 [13,14]. Thus, water-water, as well as methanol-methanol hydrogen bonds are classified as moderate hydrogen bonds [13].…”

Ab initio studies of peroxynitrite anion-water complexes

“…The Mulliken charge on N2 is much more negative and 03 is more positive as a result of the shorter, more linear N2--H8 hydrogen bond. Three factors will make the binding energies nonadditive: (1) charge delocalization per water molecule is less in the two water structures (2) angle strain is significant in structures 6 and 7, and (3) water-water hydrogen bonds (about 5 kcal/mol) [26] form in complexes 6 and 7. Structures 5 and 8 have no significant steric repulsion or extra hydrogen bonding, so the slightly higher binding energies of -2 kcal/mol can be attributed to diminished charge delocalization.…”

The water dimer interaction energy: Convergence to the basis set limit at the correlated level

“…Since the energies of the individual molecules usually are lower when computed within the composite basis of the interacting molecules rather than in the monomer's own basis, it follows that the CP corrected interaction energies are less binding than the uncorrected ones ͑this is strictly true only for variational calculations, e.g., in the case of MP2 theory for the energy expression obtained by minimizing the Hylleraas functional͒. There has been quite a fierce discussion in the literature about whether more reliable interaction energies are obtained if the CP procedure is employed or not, or rather in a modified form, [3][4][5][6][7][8][9][10][11][12][13][14] although the more recent papers on this subject tend to advocate in favor of the application of the full CP correction, as introduced by B&B: e.g., van Duijneveldt et al have shown that the CP correction à la B&B fully removes the BSSE at the FCI level, 9 and there is vast numerical evidence indicating that the same statement holds for other theory levels, like many body perturbation theory, as well. 4,9,14 The extra computational costs due to the CP correction are considerable: nϩ1 computations in the composite basis, rather than a single one, have to be performed for a cluster consisting of n molecules.…”