Liquid-Phase Transition in Water

From low-temperature high-pressure experiments of amorphous solid form of H₂O, or amorphous ice, and simulations of supercooled liquid water under pressure, two liquid phases of water were expected to exist at low temperature. In addition, existence of a liquid–liquid critical point was expected. The existence of the critical point was theoretically explained and clearly proven by molecular dynamics simulations of water. On the other hand, there has been no rigorous experimental proof of the two-water hypothesis using bulk pure water because the water easily crystallized at low temperatures. Many experiments were performed using emulsified water, amorphous ice, aqueous solutions, confined water, short-term measurements, and more. The results of a lot of indirect experiments were consistent with the hypothesis, and no experiment has clearly denied the hypothesis. Circumstantial evidence showed that real water behaves as if it had a liquid–liquid critical point.

Experimental data on the volumes of low-density amorphous ice, high-density amorphous ice, and liquid water were compared to discuss their phase relationships. It was suggested that liquid water becomes low-density amorphous ice when cooled at low pressure and becomes high-density amorphous ice when cooled at high pressure. It was also suggested that the two amorphous ices are distinctly different. Experiments on the amorphous-amorphous transition, vitrification, and glass transition supported the suggested phase relationships and indicated existence of a discontinuous first-order liquid–liquid transition.

Low-temperature metastable melting lines of crystalline ices were experimentally studied. The melting lines were detected in the no-man’s land, and it suggested existence of supercooled liquid water below the homogeneous nucleation temperature. The melting line of D₂O ice III continued smoothly into the no-man’s land at low pressure, and the line strongly curved at about 230 K and 0.02 GPa. This suggested that liquid water continuously changed from a high-density state to a low-density state along the melting line; the existence of the low-density liquid state was suggested. The existence of the low-density liquid was also suggested by experiments like the short-duration X-ray-diffraction measurement and the infrared-spectra measurement. In contrast to the smooth melting line of Ice III, the slopes of the melting lines of ice IV and ice V appeared to change suddenly. This implied the existence of a first-order liquid–liquid transition and the existence of a liquid–liquid critical point.

To study the relationship between liquid water and two amorphous ices in terms of energy, the Gibbs energy of liquid water was constructed as a function of pressure and temperature using available experimental thermodynamic data of stable liquid water, metastable supercooled water, and amorphous ices. It is examined whether the energy surface of low-temperature liquid water is consistent with the liquid–liquid critical point hypothesis. Even though the data of the bulk pure low-density liquid water at low pressures was missing, most of the data suggested indirectly that the low-density liquid water has low density and low entropy. In this case, the Gibbs energy surface became consistent with the phase separation of liquid water into two liquid phases. From the definition of the thermodynamic properties, all the complex properties of water would be derived in a unified manner from the Gibbs energy surface.

Water is the most familiar and most important liquid. Its complex properties can be explained by the existence of two kinds of liquid water. Research on water will continue.

The melting points of emulsified ices obtained by high-pressure experiments at NIMS are shown in tables.