Ice is two-faced under pressure

If you apply enough pressure, water ice will comprise not discernible H₂O molecules but a lattice of hydrogen and oxygen ions. That so-called ice X remains stable up to at least 170 GPa at 300 K. Further heating is expected to weaken the hydrogen bonds and free the H⁺ ions to move through the oxygen lattice—a form of ice known as superionic.

Superionic ice should be present in the planetary interiors of Uranus, Neptune, and other ice giants (see “Ice-giant interiors support superionic water,” Physics Today online, 7 October 2021 ), and its presence could influence the magnetic fields of some of those planets. But theoretical studies have disagreed on whether the oxygen lattice is body-centered cubic (bcc) or face-centered cubic (fcc), when the transition to the superionic phase occurs, and how stable the phase is. Recent experimental works have shown indications of the superionic phase but haven’t been able to resolve those basic questions (see “Superionic ice observed at extreme pressure and temperature,” Physics Today online, 17 May 2019 ).

Now Gunnar Weck of the University of Paris-Saclay in France and his colleagues have obtained clear structural signatures of superionic ice at a broad range of temperatures and pressures. They charted a full phase diagram for stable fcc and bcc phases.

The researchers compressed water in a diamond anvil cell and then heated it with either a carbon dioxide or ytterbium laser. In the case of the CO₂ laser, a small water volume absorbed the light directly and heated up. The technique worked for applied pressures up to 62 GPa, but beyond that, the water’s absorption at the laser wavelength dropped. Higher pressures required indirect heating instead. Weck and his team added a boron-doped diamond cup to the anvil cell, and it absorbed light from the Yb laser and heated the water inside.

The researchers collected x-ray diffraction patterns at a fairly constant pressure while incrementally increasing the temperature from about 500 K to 2500 K. They repeated the process for pressures from 27 GPa to 170 GPa. Both laser-heating strategies produced a gradient of water temperatures, so the diffraction patterns showed peaks for a range of temperatures and phases. But the results still showed a clear transition to bcc superionic ice and from bcc to fcc structures.

Weck and his colleagues plotted a full phase diagram, shown above. The blue shading and blue data points from their and previous work indicate bcc phases, and the yellow area and red data points indicate fcc phases. Density-functional-theory predictions vary in where they expect transitions (dashed lines of various colors). The new experimental results help narrow the range of viable predictions. (G. Weck et al., Phys. Rev. Lett. 128, 165701, 2022 .)