Water ice plays an important role in many fields of science. You can find it in surprisingly extreme environments in the universe – like comets or the interiors of giant planets, where pressures can exceed 300-600 GPa (3-6 million atmospheres). If we managed to reproduce and study its many different forms in the lab, we could learn a lot more about the physical nature of these bodies. To date, 21 distinct solid phases have been reported, demonstrating the remarkable structural diversity that can occur in this simple molecule. Now, a research team from the Centre for Science at Extreme Conditions (CSEC) at the University of Edinburgh's School of Physics and Astronomy, working at the Extreme Conditions Beamline P02.2 at PETRA III, has discovered the 22^nd phase. They present their work in a recent Nature Materials publication.

 At room temperature, water solidifies into tetragonal ice (called VI) at a pressure of approximately 1 Gigapascal (GPa) or 10,000  times the atmosphere on Earth. Compressed even further, it transforms into cubic ice VII and then into ice X at around 60 GPa (600,000 atmospheres). Theoretical calculations dating back almost 30 years predicted that above 300 GPa (3 million atmospheres), the cubic structure of ice X should give way to a new orthorhombic phase, one of seven different crystal systems in crystallography. The ice we are familiar with from snowflakes and our freezer has a hexagonal crystal lattice, though in very cold conditions it can sometimes take on a cubic structure. As the pressure increases, the crystal lattices of the water molecules become more complex.

Reaching these conditions experimentally, however, is quite a challenge. The scientists squeezed microscale crystallites of ice between the tips of two diamond anvil cells to trigger the transformation from ice X to ice XXII. Another challenge was to obtain good diffraction data of such small quantities of ice, which is an inherently weak X-ray scatterer. This is where DESY's Extreme Conditions Beamline P02.2 at PETRA III came in. Its synchrotron beam can be adjusted with extreme precision to probe very small samples.

“There is no other way of seeing X-ray scattering of ice than with a synchrotron,” DESY beamline scientist Hanns-Peter Liermann says. "The samples are tiny and need to be kept between the diamond anvils. We had to focus the beam down to 2 x 2 microns, and the team from Scotland had to tweak it to reduce effects that contaminate the signal by using a ceramic-plated sample holder. That way, it was possible to reached exactly the right amount of X-ray scattering from ice XXII to study its crystal structure.”

Israel Osmond, first author of the study from the University of Edinburgh, says, “Given that ice barely scatters X-rays at all, this really is pushing high-pressure crystallography to its absolute limits. It’s a real testament to how far diamond anvil cell techniques have come.”

Ross Howie, co-lead of the study, adds, “We inspected the sample by chance as we didn’t expect to observe anything – how wrong we were! Exciting discoveries like this don’t happen every day and it is great that so many early career researchers and students involved in the study got to be a part of it.”

Miriam Pena-Alvarez, co-lead of the study, says, “Most matter in the universe doesn't exist at the atmospheric conditions we are familiar with but exists under high pressure. This is what drives our lab, and this work in particular: recreating those planetary interiors in house, to provide insight on how one of the universe's most abundant molecule behaves under these extreme pressures.

Through a combination of X-ray diffraction and Raman spectroscopy techniques the authors demonstrate that the crystal lattice of ice X distorts upon compression to transition to ice XXII at pressures above 308 GPa, proving a 30-year-old theory. Ice XXII is found to be stable to pressures of 430 GPa, with calculations predicting the next transition to be in excess of 930 GPa. That makes it possibly the last new ice phase discoverable at room temperature and high-pressure with current lab techniques.

The authors expect their finding to open a floodgate of studies in understanding how ice XXII will form at low temperatures, together with how it will meet the hot superionic states. It also forms part of a fast-growing catalogue of solid phases of water that keeps revealing how much structural complexity a “simple” molecule like H₂O can hide under extreme conditions.

With the planned new incarnation of the PETRA light source, PETRA IV, these studies could be improved several times over, with even smaller beam sizes and a more intense flux, which is especially interesting for weakly scattering probes like water or ice.

Original publication:

Osmond, I., Kuzovnikov, M.A., Lewis, D.A., Shuttleworth H>A, Marqueno T, Dixon, M.J, Robertson, C. E. a., Cherepsnalkoski V., Wang B., Hermann A., Peña-Alvarez M., Howie R. Ultrahigh pressure phase transition in ice. Nat. Mater. (2026).

DOI: 10.1038/s41563-026-02732-1