Suppression of the dynamic transition in surface water at low hydration levels: A study of water on rutile

Our quasielastic neutron-scattering experiments and molecular-dynamics simulations probing surface water on rutile $\left({\text{TiO}}_2\right)$ have demonstrated that a sufficiently high hydration level is a prerequisite for the temperature-dependent crossover in the nanosecond dynamics of hydration water. Below the monolayer coverage of mobile surface water, a weak temperature dependence of the relaxation times with no apparent crossover is observed. We associate the dynamic crossover with interlayer jumps of the mobile water molecules, which become possible only at a sufficiently high hydration level.

Figure 1

(Color) QENS data (symbols) and fits (solid lines) with Eq. (1) (single Lorentzian fit for the lowest hydration level).

Figure 2

(Color) The characteristic relaxation times calculated from the HWHM $\left({\Gamma }_1\right)$ of the narrow (or, for the lowest hydration level, the only) Lorentzian component of the QENS data fits as $\tau =\hslash /{\Gamma }_1$. Also shown are fits with VFT, Arrhenius, and temperature-independent functions.

Figure 3

(Color) The relaxation times for the slower component obtained from the MD data fits with two Debye components. Also shown are fits of the relaxation times with VFT, Arrhenius, and temperature-independent functions.

Figure 4

(Color) Snapshots of the simulation data [planar views perpendicular to the (110) rutile surface, only ${\text{L}}_2$ and ${\text{L}}_3$ layers are shown for clarity] for the highest hydration level at 260 K, at $t=0$ (after MD equilibration) and $t=1000\text{ps}$. The oxygen atoms of the water molecules that were residing in the ${\text{L}}_2$ and ${\text{L}}_3$ layers at $t=0$ are shown in blue and red, respectively. Top left (a): ${\text{L}}_2$ at $t=0\text{ps}$. Bottom left (b): ${\text{L}}_2$ at $t=1000\text{ps}$. Top right (c): ${\text{L}}_3$ at $t=0\text{ps}$. Bottom right (d): ${\text{L}}_3$ at $t=1000\text{ps}$.