Abstract
Liu and co-workers [Phys. Rev. B 82, 161415 (2010)] discussed the long-standing debate regarding whether
H2O molecules on the defect-free (110) surface of rutile (α-TiO2) sorb associatively, or there is dissociation
of some or all first-layer water to produce hydroxyl surface sites. They conducted static density functional
theory (DFT) and DFT molecular dynamics (DFT-MD) investigations using a range of cell configurations and
functionals. We have reproduced their static DFT calculations of the influence of crystal slab thickness on water
sorption energies. However, we disagree with several assertions made by these authors: (a) that second-layer
water structuring and hydrogen bonding to surface oxygens and adsorbed water molecules are “weak”; (b) that
translational diffusion of water molecules in direct contact with the surface approaches that of bulk liquid water;
and (c) that there is no dissociation of adsorbed water at this surface in contact with liquid water. These assertions
directly contradict our publishedwork, which compared synchrotron x-ray crystal truncation rod, second harmonic
generation, quasielastic neutron scattering, surface charge titration, and classical MD simulations of rutile (110)
single-crystal surfaces and (110)-dominated powders in contact with bulk water, and (110)-dominated rutile
nanoparticles with several monolayers of adsorbed water.
