At the extreme: Breaking the ice mold

water target group (Download Image)

Water cell targets for the Omega experiments. Water is injected into the cell using the fill tubes inserted at the top. Image by Carol Davis/LLNL.

New research involving Lawrence Livermore National Laboratory (LLNL) scientists shows that water can remain liquid in a metastable state when transitioning from liquid to a dense form of ice at higher pressures than previously measured.

Water at extreme conditions has attracted recent attention because of its complex phase diagram, including superionic ice phases having exotic properties that exist at high pressures and densities. To date, 20 unique crystalline ice phases have been found naturally on Earth or in the laboratory. Water also exhibits bizarre metastable phenomena when compressed or cooled very rapidly, which have attracted interest from physicists worldwide for many years.

“If the water is compressed very rapidly, it will remain liquid in a metastable state until finally crystallizing into ice VII at a higher pressure than expected,” said Michelle Marshall, a research scientist at the Laboratory for Laser Energetics (LLE) at the University of Rochester, a former LLNL postdoc and lead author of the study appearing in Physical Review Letters.

Ice VII is the stable polymorph of water at room temperature and at pressures exceeding 2 GPa (more than 19,000 atmospheres]. Recently, ice VII was found naturally on Earth for the first time as inclusions in diamonds sourced deep within the mantle. It may exist inside Jupiter’s icy moons and in water worlds beyond our solar system.

The new research showed how water can remain liquid in a metastable state when undergoing the liquid-to-ice-VII transition at higher pressures than previously measured. Previous experimental work at the giant pulsed-power Z facility showed that the compressed water transforms to ice VII at 7 GPa (69,000 atmospheres) when the water is ramp-compressed over hundreds of nanoseconds. The new experiments instead shifted to use high-power lasers at the Omega Laser Facility to compress water over even shorter timescales (nanoseconds).

Just like in previous LLNL work on gold (Au) and platinum (Pt), the most difficult thing is to compress the water gently enough to avoid forming a shockwave that would ruin the experiment (i.e. realizing a shockless ramp compression). Because water is much more compressible than metals like Au and Pt, creating a ramp compression wave in a micrometer-thin water layer requires increasing the pressure load at a much slower rate.

one shot
Time-integrated image of a laser shot at the Omega Laser Facility to study the liquid to ice VII phase transition in ramp-compressed water. Image by Eugene Kowaluk/LLE.

“Even though the pressures we achieve appear very modest compared to other laser-driven ultrafast dynamic compression experiments, these extremely difficult experiments are really at the frontier of what we can do with giant lasers, and that was an exciting challenge," said LLNL scientist and co-author Marius Millot.

The new data reveal that water can remain liquid to at least 8-9 GPa (79,000-89,000 atmospheres) before crystallizing into ice VII: the freezing pressure increases with the compression rate.

“This means that water can remain liquid to at least 3.5 times higher pressures than expected based on the equilibrium phase diagram,” Marshall said. "It’s really neat to think that we are compressing it so fast that water does not have time to crystallize, so it remains liquid."

“We are at the frontier of experimental ultrafast science,” Marshall said, “and it was great to collaborate with our theory and simulation colleagues to gain a more detailed picture of what was happening. It is remarkable that the most recent theoretical and numerical advances now provide a detailed understanding of the observed phenomena. This could have implications for our general understanding of phase transformations at extreme conditions.”

This work is part of a broader effort to understand phase transition kinetics in dynamically compressed materials. The ubiquitous nature of water and its complex phase diagram make the liquid-to-ice-VII phase transition an interesting test bed for phase-transition kinetics modeling. SAMSA, an LLNL-developed kinetics model, provides a detailed understanding of the experimental results while relying on the fundamentally simple picture of homogeneous nucleation using classical nucleation theory.

Broadly speaking, this work helps improve material models and understanding, which could have interesting implications for other key areas of research at the Laboratory such as advanced manufacturing and 3D printing. Metastable states and complex crystallization of water also are key for atmospheric science and therefore for climate security.

Other Livermore scientists include Dayne Fratanduono, Dane Sterbentz, Philip Myint, Jon Belof, Yong-Jae Kim, Federica Coppari, Suzanne Ali, Jon Eggert, Raymond F. Smith and James McNaney. LLNL’s Carol Davis, Renee Posadas, Eric Folsom and Jim Emig and LLE’s Julie Fooks contributed to the target fabrication and water cell development for this project.