Ice's polymorphism, with over 20 crystalline and amorphous phases, is a significant area of research spanning material science and planetary science. This structural variety stems from the flexibility of hydrogen bonds and order-disorder transitions. Hydrogen ordering significantly impacts ice's properties, including molecular rotation and ferro-/antiferroelectric alignment. A key unanswered question is whether a hydrogen-disordered ice phase transitions to only one hydrogen-ordered phase. The known ice phase diagram suggests this, yet the close energies of possible hydrogen configurations due to geometric frustration complicate this. Recent work on high-pressure ice VI revealed an unknown hydrogen-ordered form (β-XV) besides the known ice XV, suggesting the possibility of multiple ordered phases. However, the distinct nature of β-XV remained unclear due to insufficient evidence. This study aimed to investigate the possibility of multiple hydrogen-ordered phases from ice VI using in-situ dielectric and neutron diffraction measurements under high pressure, providing unambiguous evidence for a second hydrogen-ordered phase of ice VI. Understanding the multiplicity of hydrogen-ordered phases is crucial for completing the ice phase diagram and for advancing our knowledge of hydrogen bonding in various materials and conditions. The ability to control the hydrogen ordering through pressure manipulation can have broader implications for material design and our understanding of ice’s physical properties.

Previous research has extensively explored the diverse phases of ice, focusing on the interplay between hydrogen bonding and the resulting structural arrangements. The one-to-one correspondence between disordered and ordered phases was a widely accepted assumption, supported by the then-known phase diagram. However, theoretical studies have suggested the possibility of multiple hydrogen-ordered phases stemming from a single disordered phase due to the inherent geometric frustration in the ice lattice. Recent experimental findings on high-pressure ice VI revealed a new hydrogen-ordered phase, initially labeled β-XV. However, the lack of conclusive experimental data prevented its definitive classification as a distinct phase with a new Roman numeral designation. This study builds on previous theoretical and experimental work, seeking to resolve the ambiguity surrounding β-XV and to explore the potential for greater complexity in ice's phase diagram.

The study employed a combination of in-situ dielectric and neutron diffraction measurements under high pressure. Dielectric measurements were performed using a newly developed pressure cell capable of simultaneously measuring dielectric properties and pressure via the ruby fluorescence method. HCl (or DCl in neutron diffraction experiments) was used as a dopant (10⁻² M concentration) to accelerate hydrogen ordering. Measurements covered pressures from 0.88 to 2.2 GPa and temperatures from 100 to 150 K (dielectric) or 80 to 150 K (neutron diffraction). The dielectric experiments tracked changes in dielectric response (constant and loss) with temperature and pressure to identify phase transitions. The phase transition temperatures were determined based on the sudden weakening of dielectric loss peaks associated with hydrogen ordering. Neutron diffraction experiments were conducted at the PLANET beamline at J-PARC. DCl-doped D₂O was used, and ice VI was prepared through solid-solid phase transitions (ice III → V → VI) to obtain a fine powder sample. The pressure was estimated using Pb as a pressure marker. Neutron diffraction patterns provided structural information, enabling identification of new peaks and refinement of the ice XIX structure. Structural analysis involved determining candidate space groups based on the group-subgroup relationship between ice VI and the newly discovered phase, followed by Rietveld refinement using structural models that adhered to the ice rules. The most plausible space groups for ice XIX were P4 or Pcc2 based on the best fit χ² values, indicating a partially ordered state.

Dielectric measurements revealed a pressure-dependent transition from ice VI to two distinct hydrogen-ordered phases. At lower pressures, the transition to ice XV was observed, while at higher pressures (above 1.5–1.6 GPa), a transition to a new phase, designated ice XIX, occurred. The phase boundary between ice VI and its ordered phases showed a change in slope (dT/dP) from negative to positive around 1.5–1.6 GPa, suggesting that ice XIX has a smaller volume than ice VI and ice XV. This volume contraction thermodynamically stabilizes ice XIX at higher pressures. Neutron diffraction experiments confirmed the existence of ice XIX. The appearance of new peaks in the diffraction patterns, not attributable to ice XV, provided unambiguous evidence of a new crystalline phase. Analysis of the temperature dependence of lattice parameters showed that the unit cell of ice XIX is expanded in *a* and contracted in *c* relative to ice VI, further confirming its distinct nature. The refinement of the ice XIX structure suggests the space group P4 or Pcc2, representing a partially ordered state. These models include deuterium atoms with 50% site occupancy, implying partial hydrogen ordering. The proposed space group indicates a possible fully-ordered phase, underscoring the potential for additional discoveries. The study also points to the significance of centrosymmetry in differentiating the hydrogen configurations of ice XIX and XV. The findings are consistent with a concurrent study by Gasser et al., further solidifying the validity of the observed phase diagram.

The discovery of ice XIX demonstrates that a single hydrogen-disordered ice phase can transform into multiple hydrogen-ordered phases under different pressure conditions. This finding challenges the traditional understanding of ice polymorphism and broadens our understanding of hydrogen bonding in ice. The pressure-induced multiplicity of hydrogen-ordered phases, as demonstrated by the different thermodynamic behavior of ice XV and XIX, represents a novel mechanism for controlling the hydrogen ordering process. The low-temperature region of the ice phase diagram, where ice XIX was found, presents a new frontier for exploring the diverse possibilities of hydrogen ordering in ice. The large number of possible hydrogen-ordered configurations in the ice XIX unit cell highlights the complexity of predicting and modeling the thermodynamic stability of hydrogen-ordered phases in ice, suggesting the need for further advanced computational techniques. The pressure-induced transition between different hydrogen-ordered configurations could potentially influence the physical properties of ice, such as (anti-)ferroelectricity, leading to the exploration of high-pressure and high-electric field conditions. This opens new avenues for controlling the properties of hydrogen-bonded materials.

This research provides compelling experimental evidence for the existence of ice XIX, a second partially-ordered phase of ice VI. This discovery challenges the long-held assumption of a one-to-one correspondence between disordered and ordered ice phases and reveals a new level of complexity in ice's phase behavior. The pressure-dependent hydrogen ordering highlights the potential for controlling the hydrogen configuration and consequently the properties of ice and potentially other hydrogen-bonded materials. Future work could focus on exploring the fully hydrogen-ordered phase predicted for ice VI, further investigating the relationship between pressure, hydrogen ordering, and properties like (anti-)ferroelectricity, and refining structural models of ice XIX through single-crystal neutron diffraction.

The study primarily focused on powder samples; single-crystal studies would provide a more detailed understanding of ice XIX's structure. The acceleration of hydrogen ordering using dopants might influence the observed phase transitions, although similar results were obtained using both H₂O and D₂O samples. The relatively small volume change observed in neutron diffraction might have limited the precision of certain measurements. Further investigation is needed to fully understand the interactions between dopants and the ice structure under high-pressure conditions.