The history of the

• Main Authors: G.Ruocco, F.Sette
• Format: Article
• Language: English
• Published: Institute for Condensed Matter Physics 2008-03-01
• Series: Condensed Matter Physics
• Subjects: water; inelastic x-ray scattering; sound propagation; liquids; glasses
• Online Access: http://dx.doi.org/10.5488/CMP.11.1.29

This paper aims to discuss the present understanding of the high frequency dynamics in liquid water, with particular attention to a specific phenomenon - the so-called fast sound - since its first appearance in the literature up to its most recent explanation. A particular role in this history is played by the inelastic x-ray scattering (IXS) technique, which - with its introduction in the middle '90- allowed to face a large class of problems related to the high frequency dynamics in disordered materials, such as glass and liquids. The results concerning the fast sound in water obtained using the IXS technique are here compared with the inelastic neutron scattering (INS) and molecular dynamics simulation works. The IXS work has allowed us to demonstrate experimentally the existence of two branches of collective modes in liquid water: one linearly dispersing with the momentum (apparent sound velocity of ≈3200 m/s, the "fast sound") and the other at almost constant energy (5..7 meV). It has been possible to show that the dispersing branch originates from the viscoelastic bend up of the ordinary sound branch. The study of this sound velocity dispersion, marking a transition from the ordinary sound, c_o to the "fast sound", c_∞, as a function of temperature, has made it possible to relate the origin of this phenomenon to a structural relaxation process, which presents many analogies to those observed in glass-forming systems. The possibility to estimate from the IXS data the value of the relaxation time, τ, as a function of temperature leads to relating the relaxation process to the structural re-arrangements induced by the making and breaking of hydrogen bonds. In this framework, it is then possible to recognize an hydrodynamical "normal" regime, i. e. when one considers density fluctuations whose period of oscillation is on a timescale long with respect to τ, and a solid-like regime in the opposite limit. In the latter regime, the density fluctuations feel the liquid as frozen and the sound velocity is much higher: this is "fast sound" whose value is equivalent to the sound velocity found in crystalline ice.