Osamu Mishima

Propagation of the polyamorphic transition of ice and the liquid–liquid critical point

Water has a rich metastable phase behaviour that includes transitions between high- and low-density amorphous ices, and between high- and low-density supercooled liquids. Because the transitions occur under conditions where crystalline ice is the stable phase, they are challenging to probe directly. In the case of the liquids, it remains unclear whether their mutual transformation at low temperatures is continuous, or discontinuous and terminating at a postulated second critical point of water that is metastable with respect to crystallization. The amorphous ices are more amenable to experiments, which have shown that their mutual transformation is sharp and reversible. But the non-equilibrium conditions of these studies make a firm thermodynamic interpretation of the results difficult. Here we use Raman spectroscopy and visual inspection to show that the transformation of high-density to low-density amorphous ices involves the propagation of a phase boundary—a region containing a mixture of both ices. We find that the boundary region becomes narrower as the transformation progresses, and at higher transformation temperatures. These findings strongly suggest that the polyamorphic ice transition is discontinuous; a continuous transformation should occur uniformly over the entire sample. Because the amorphous ices are structurally similar to their supercooled liquid counterparts, our results also imply that the liquids transform discontinuously at low temperatures and thus support the liquid–liquid critical-point theory.

Volume of supercooled water under pressure and the liquid-liquid critical point

The volume of water (H(2)O) was obtained at about 200-275 K and 40-400 MPa by using emulsified water. The plot of volume against temperature showed slightly concave-downward curvature at pressures higher than ≈200 MPa. This is compatible with the liquid-liquid critical-point hypothesis, but hardly with the singularity-free scenario. When the critical point is assumed to exist at ≈50 MPa and ≈223 K, the experimental volume and the derived compressibility are qualitatively described by the modified Fuentevilla-Anisimov scaling equation.

Vitrification of emulsified liquid water under pressure

We cool emulsified liquid water below 130 K at a rate of 103–104u200aKu200as−1 at ≈0.5 GPa and vitrify it under pressure. We recover the produced glassy sample at 1 bar and 77 K, and examine it by Raman spectroscopy, x-ray diffraction, and visual observation. We find that the emulsified glassy water, when heated from 77 K at 1 bar, exhibits a distinct volume expansion around 130 K and becomes a second amorphous state. The second amorphous state becomes ice Ic at 160–170 K with a negligible volume change. This behavior of the high-pressure emulsified glassy water resembles that of high-density amorphous ice made by pressure-induced amorphization of ice Ih.

Application of polyamorphism in water to spontaneous crystallization of emulsified LiCl-H2O solution.

Aqueous solutions are widely explained by the hydration or the bound waterfree water notion. Amorphous polymorphism (polyamorphism) in pure water, which is presently under vigorous discussion, may provide a different view over the solutions. Here, I changed pressure, P, temperature, T, and concentration, C, of emulsified LiCl-H2O solutions and studied their freezing by detecting its heat evolution. It was experimentally indicated that the homogeneous nucleation of low-density crystalline ice I (phase Ih or Ic), in pure water and in solutions, connects to the polyamorphic transition of high-density amorphous ice (HDA) to low-density amorphous ice (LDA). Thus, the polyamorphism of water relates to the phase behavior of aqueous solution. In accordance with the recent simulation result, the nucleation was thought to occur in two stages: the appearance of the LDA-like state and the crystallization. Usefulness of the polyamorphic point of view about the solutions was seen.

Phase separation in dilute LiCl–H2O solution related to the polyamorphism of liquid water

When an emulsified 4.8 mol % LiCl-H2O solution was cooled under a pressure of 0.35 or 0.45 GPa and decompressed to 0.1 GPa at 142 K, slightly above its glass transition temperature (approximately 140 K at 0.1 GPa), its volume increased suddenly. This was regarded as an appearance of the low-density amorphous ice in the liquid solution as suggested by x-ray and Raman measurements, and this appearance corresponded to the high-to-low-density polyamorphic transition of pure H2O. Hysteresis was considered to accompany this volumetric change. The hysteresis of the liquid transition proves its first-order nature and, as for the solution, this suggests that the transition is a polyamorphic phase separation.

The glass-to-liquid transition of the emulsified high-density amorphous ice made by pressure-induced amorphization.

Emulsified high-density amorphous ice, made by pressure-induced amorphization of emulsified ice Ih, was decompressed at about 160 K. The onset of an endothermic event was observed around 0.4 GPa during the decompression. This is consistent with existence of the glass transition to a liquid state, implying the close relationship between melting and amorphization.

Experimentally proven liquid-liquid critical point of dilute glycerol-water solution at 150 K

The experimental and theoretical studies of supercooled liquid water strongly suggest that the two liquid waters and their liquid-liquid critical point (LLCP) exist at low temperature. However, the decisive experimental evidence of the LLCP has not been obtained because of the rapid crystallization of liquid water in the no-mans land. Here, we observed experimentally the pressure-induced polyamorphic transition in the dilute glycerol-water solution which relates to the water polyamorphism. We examined the effect of the glycerol concentration on the liquid-liquid transition, and found its LLCP around 0.12-0.15 mole fraction, 0.03-0.05 GPa, and ~150 K. A 150 K was above, or around, the recently recognized glass transition temperatures of amorphous ices, and the crystallization did not occur, indicating that the direct observation of LLCP is feasible. The low-temperature LLCP has implication to the argument of the relation between the interaction potential of water molecule and the polyamorphic phase diagram.

Raman spectroscopic study of glassy water in dilute lithium chloride aqueous solution vitrified under pressure

We vitrify dilute lithium chloride (LiCl) aqueous solutions (2.0–10.0 molu200a% solute) at about 0.5 GPa using a needle-type pressure device developed for cooling a liquid material rapidly, and measure Raman spectra of the vitrified solution at 1 atm and 25 K. The OH stretching vibrational modes of the vitrified solution seem to consist of a OH stretching vibrational mode of high density amorphous ice (HDA) of pure water and that of glass of highly concentrated LiCl aqueous solution. The Raman profile resembles more and more that of HDA as the solute concentration decreases. When the temperature of the vitrified solution increases at 1 atm, the Raman profile suddenly changes to that of low density amorphous ice around 135 K and the transformed profile goes to that of Ic around 160 K. This transformation behavior seems to correspond to a transformation behavior of HDA. These results suggest that the portion of solvent water in the dilute LiCl aqueous solution vitrified under pressure may form HDA-type molecula...

Differences between pressure-induced densification of LiCl?H2O glass and polyamorphic transition of H2O

We perform volumetric measurements of LiCl aqueous solution up to 1.00xa0GPa in the 100-170xa0K range, examine the pressure-induced vitrification and densification, and draw the pressure-temperature-volume surface. The pressure-induced vitrification of the solution corresponds to the cooling-induced vitrification of the liquid. We found that the volumetric decrease of glassy solution during the densification is continuous and this behavior depends on the glassy state before the compression. Raman profiles of the glassy solutions before and after the densification are similar. In contrast, the polyamorphic transition from low-density amorphous ice (LDA) to high-density amorphous ice (HDA) is discontinuous and their Raman profile before and after the transition is distinct. These results suggest that the densification relates to the structural relaxation and differs intrinsically from the polyamorphic transition. Furthermore, the densification of HDA is observed under high pressure, suggesting that very high-density amorphous ice (VHDA) may be the densified HDA. In order to recognize a polyamorphic transition under a non-equilibrium condition correctly, evidence of not only large volume change but also some distinct structural changes in glassy state is necessary.

Raman Study of the Annealing Effect of Low-Density Glassy Waters

We make three kinds of low density glassy waters, low-density amorphous ice (LDA), amorphous solid water (ASW) and hyperquenched glassy water (HGW), and examine the changes in their Raman spectra of OH-stretching vibrational regions by annealing at 1 bar. The irreversible change in Raman profile for LDA is observed above its calorimetric glass transition temperature ( T g ) of ∼130 K below 144 K. The degree of the change in Raman spectra of LDA at the annealing temperature appears to be different from those for ASW and HGW. These results suggest a possibility that the modification in the hydrogen-bonded network of LDA during the annealing occurs above the calorimetric T g .