High Density Amorphous Ice from Cubic Ice

Abstract

Supercooled water does not behave like a simple liquid, but instead shows a number of anomalous properties. These include the diffusion coefficient [2] or the kinematic viscosity [3] which show anomalous pressure dependencies on compression up to 200MPa, whereas beyond 200 MPa the expected behaviour is observed. To explain these anomalies the second critical point hypothesis has been put forward, in which a first-order-like phase transition between a low(LDL) and high-density liquid (HDL) is thought to occur. LDL and HDL become indistinguishable above the second critical point, which is postulated to be in the “no man’s land” around 180– 220 K and 100–340 MPa, where only crystalline ice has been observed experimentally. This model is supported by the observation that there are (at least) two different phases of amorphous water called low-(LDA) and high-density amorphous ice (HDA), which have been found to show a first-order like transformation in compression/decompression experiments. On heating, LDA and HDA experience a glass-transition to the highly viscous, supercooled liquids denoted low(LDL) and high-density liquid (HDL), respectively. The role of HDA in this model has become unclear since the discovery of a further distinct structural state called very-high density amorphous ice (VHDA). It has lately been argued that only VHDA and LDA are homogeneous disordered structures, whereas HDA does not constitute a particular state of the HDA network. This issue is still under debate, since the transition from HDA to VHDA on isothermal compression at 125 K has been found recently to be similar to the transition from LDA to HDA. The detailed structures of recovered HDA and VHDA have been determined at ambient pressure and 77 K by means of neutron diffraction with isotope substitution. These two studies contain the first experimental determination of the three site-site distribution functions (H H, O O, O H). For both HDA and VHDA there is no evidence in the diffraction pattern to show that it is microcrystalline rather than a genuinely amorphous structure. Herein, we show that HDA produced by pressure-induced amorphization of cubic ice (Ic) at 77 K is not the same material as “traditional” HDA produced by pressure-induced amorphization of hexagonal ice (Ih) at 77 K. The density of both amorphous states recovered at ambient pressure is very similar; however, the X-ray structure factor as well as the phase transition characteristics in differential scanning calorimetry scans, differ in the onset temperature, enthalpy and sharpness of the exotherms. Hexagonal ice was prepared by either pipetting 0.300 mL of deionized water directly into the piston-cylinder apparatus lined with an indium container kept at 77 K, or by heating HDA close to ambient pressure to 260 K and recovering to 77 K and 1 bar, or by Johari’s procedure of decompressing HDA to 0.06 GPa and heating to 235 K. Cubic ice was prepared by hyperquenching small liquid droplets of 3 mm diameter on a cryoplate kept at 190 K in the same way reported earlier. The deposit was then transferred to the piston–cylinder apparatus and filled into the indium–lined cylinder by spooning, under liquid nitrogen. Alternatively, cubic ice was prepared by heating HDA to 185 K at 0.025 GPa and recovering to 77 K and 1 bar. HDA was produced by compressing 300 mg of Ih to 1.5 GPa at 77 K in a similar way to our previous studies. These starting materials for the pressure-induced amorphization process have been characterized using powder X-ray diffractograms, which are depicted in Figure 1. The posi-