M A T T E R A T E X T R E M E S

S C I E N T I F I C H I G H L I G H T S

1 4 H I G H L I G H T S 2 0 2 2 I

Observing hot dense superionic water Under high pressure and high temperature, solid water takes a form called superionic ice, in which hydrogen atoms migrate through the oxygen crystal lattice as in a fluid. X-ray diffraction measurements reveal that this superionic ice could occupy a large part of the interiors of the planets Uranus and Neptune.

There is currently great interest in studying the phases and properties of warm dense ice present in the planetary interiors of Uranus, Neptune and other giant icy exoplanets. Under the temperature and pressure conditions of these planetary interiors, part of the ice layer is predicted to be in a superionic (SI) state, where a rigid oxygen lattice coexists with mobile hydrogen atoms, which could influence the magnetic field of Neptune-like planets.

Over the last two decades in the high-pressure physics community, much work has aimed at finding evidence for the existence of SI ice, its lattice structure and the stability of that phase under different pressures and temperatures. It is known that above pressures of 3 GPa, ice structures are based on body-centered cubic (bcc) oxygen sublattices, the first one being ice VII. Above 60 GPa, ice becomes an ionic solid, called ice X, in which the hydrogen atoms are located halfway between the nearest neighbouring oxygen atoms. Upon heating, the hydrogen bond weakens and ice is expected to transform into the superionic (SI) phase, where the hydrogen atoms become mobile. Various structures of superionic ice have been predicted from ab-initio

simulations; however, the properties of these structures and their stability at different pressures and temperatures are still debated.

Recent experimental work using either dynamic [1,2] or static [3,4] compression of water ice more firmly established the existence of superionic ice and demonstrated a transition from an insulating bcc structure to bcc-SI, on one hand, and from the bcc-SI to a face-centered cubic (fcc) superionic phase called ice XVIII, on the other hand. However, the dynamic [2] and static [4] experiments strongly disagreed on the stability of the SI phases. Further measurements were thus essential to resolve these issues.

X-ray diffraction was combined with the laser-heated diamond anvil cell at beamline ID27 to investigate the structural changes in water ice under a range of pressure and temperature conditions. First, water was heated in the diamond anvil cell by a CO2 laser directly absorbed by the H2O sample. X-ray diffraction patterns were collected up to pressures of around 60 GPa, with the resulting structural data demonstrating the presence of fcc superionic ice, as shown in Figure 1. Higher pressures required indirect heating of the H2O sample using an ytterbium laser,

Fig. 1: XRD patterns collected at 57 GPa under CO2 laser heating, showing the presence of fcc-SI diffraction peaks in the spectra collected at 1560 K. The double peak structure of bcc (110) at 1320 K and 1420 K is indicative of the presence of the bcc-SI phase at these temperatures. The inset shows a sketch of a cross- sectional view of the H2O sample cavity under direct CO2 laser heating, with the hot ice volume in red.

Fig. 2: Phase diagram of dense H2O (blue/red circle and diamond symbols) and shock wave measurements (blue/red triangle). The various coloured domains correspond to the stability fields of: in grey, the molecular fluid; in blue, the bcc ices (ice VII, VII , X and bcc-SI); in yellow, the fcc-SI. The white zones represent the uncertainties on the transition lines. The dot lines correspond to recent large-scale first-principle calculations using a machine- learning potential trained on DFT-PBE simulations.