Water's familiar solid form is only one member of a large family of ices. Change pressure and temperature enough, and the molecules rearrange into structures unlike anything in a household freezer. Experiments at pressures exceeding 300 gigapascals have now revealed another member of that family: ice XXII.
At very high pressure, ordinary molecular distinctions in ice change dramatically. In ice X, hydrogen atoms occupy symmetric positions between oxygens, producing a dense ionic-like network. Researchers compressed samples beyond about 308 gigapascals and observed a first-order transition to a different crystal structure, an orthorhombic phase that had been predicted theoretically but not previously established experimentally. The phase remained stable through at least the higher pressures explored in the study.
A pressure of 308 gigapascals is more than three million times atmospheric pressure at sea level. Such conditions are produced in the laboratory with diamond-anvil cells, which squeeze microscopic samples between the tips of diamonds while spectroscopic or diffraction techniques track structural changes. Experiments become increasingly difficult as pressure rises because the sample volume shrinks and the diamonds themselves approach their mechanical limits.
The result matters because water is abundant in planets and moons. In large icy worlds, water can be buried beneath thousands of kilometres of overlying material and compressed into phases that do not exist naturally at Earth's surface. The exact crystal structure influences density, electrical properties, heat transport and the way water interacts with rock or other compounds inside a planet.
Ice XXII therefore adds a new constraint to models of extreme water-rich interiors. Its existence can alter predicted layer boundaries and physical behavior at pressures relevant to massive planets or exoplanets. Researchers will now want to measure properties beyond structure, including conductivity and how the phase behaves at the high temperatures expected inside real planets.
The discovery also tests quantum and materials calculations. Predicting matter at hundreds of gigapascals requires models of chemical bonding under conditions where familiar intuition can fail. Experimental confirmation of a predicted phase shows where those calculations work and provides a benchmark for improving them.
No natural sample of ice XXII is expected to survive a journey to Earth's surface; it would transform as pressure fell. Its significance lies in recreating a state of water that may be ordinary only in places humans cannot directly reach. Each new high-pressure phase extends the map of what one of the simplest molecules can become when planetary conditions push chemical bonds far beyond everyday experience.