Virtually all the ice that we encounter on Earth’s surface, from snowflakes to ice cubes, has a hexagonal crystal structure called ice Ih. But this is not the only form that ice can take. By altering temperature and pressure conditions in the lab, scientists have discovered at least 18 different crystal structures of ice. Puzzlingly, at certain conditions, amorphous solid ice can form without a crystal structure, leading scientists to suggest this form was somehow related to supercooled liquid water.
For about 30 years, scientists have attempted to explain how glassy and pressure-induced amorphous ice forms are thermodynamically related to supercooled water. In a new paper  in Nature, DCO Extreme Physics and Chemistry Community members Chris Tulk (Oak Ridge National Laboratory, USA), Adam Makhluf, Craig Manning (both at University of California Los Angeles, USA), and Dennis Klug (National Research Council of Canada) show that amorphous ice is an experimental side effect of increasing the pressure too fast on a regular ice sample. The amorphous form occurs when ice I experiences a “kinetically arrested transition” – and essentially gets stuck – along the path to a more tightly packed crystal structure. Increasing the pressure more slowly leads to a progression of different equilibrium crystal structures. While such amorphous forms of ice don’t exist naturally on Earth, the study’s findings can help us understand the forms that water and other volatile compounds, such as carbon dioxide, might take on elsewhere in the solar system.
Like many scientific advances, “the discovery was completely serendipitous,” said Tulk. Initially, he was trying to make the amorphous form of ice and then planned to increase the pressure in a step-wise manner to see what happens just before it finally transitions into a more dense crystalline structure.
Tulk froze water onto the anvils of a Paris-Edinburgh press, which is an apparatus that can exert tremendous pressure on a sample. He maintained the sample at almost -200 degrees Celsius with liquid nitrogen inside the Spallation Neutrons and Pressure (SNAP) diffractometer, which is part of the Spallation Neutron Source, a Department of Energy user facility located at the Oak Ridge National Laboratory, USA. The SNAP instrument enables researchers to direct a beam of neutrons at samples maintained under extreme temperatures and pressures and to use the resulting neutron diffraction pattern to determine the sample’s crystal structure. By the time Tulk had set up the instrument, it was late in the evening. So he programmed the system to increase the pressure gradually, starting at low pressure to make sure he didn’t overshoot the transition.
When he returned in the morning, he realized the amorphous ice had never formed. “I have made many amorphous ice samples in my career and I had never seen this behavior,” said Tulk. “It was those lower pressure points that made the difference.” After repeating the results and ruling out problems with his water and equipment he realized that by increasing the pressure very slowly, the ice moved through several different crystalline phases, never forming the amorphous form. These experiments give direct experimental evidence of the energetically underlying crystalline structures that can occur in place of the amorphous form. Meaning that, if the amorphous form were heated under pressure, it would very likely continue on its transition toward the higher density crystal, and not a deeply supercooled liquid.
Working with Makhluf, Tulk used the same setup, but increased the pressure more rapidly. The ice took on the amorphous form, confirming that the speed of the transition was key to avoiding the in-between glassy phase. “So, the amorphous form is not related to ‘melting’ into liquid water, which is an assumption that kicked off 30 years of research,” said Tulk. “It’s more closely related to the crystalline forms, and is simply a kinetically interrupted transition between two crystalline phases.”
Besides clarifying the relationships between amorphous ice, supercooled water and the various crystal structures that ice can take, the findings also solve the problem of how not to make amorphous ice. “With the experimental apparatus that we use, it is very challenging to do things slowly enough at these very low temperatures so that you can avoid the formation of this form of ice,” said Manning. Now that researchers can prevent this experimental side effect, they potentially can apply the same method to studying more complicated systems, such as learning how the presence of dissolved salts or carbon dioxide affect ice crystallization.
While these exotic crystal structures of ice only occur in the lab on Earth, they likely exist elsewhere in the galaxy, such as on icy comets and at the bottom of oceans on ice-crusted moons like Jupiter’s Ganymede, Calisto, and Europa. And since water is a major vehicle for transporting carbon, understanding where and how water exists in the galaxy could also yield insights into how carbon is distributed across planetary bodies.
Main image credit: Jill Hemman/Oak Ridge National Laboratory, U.S. Dept. of Energy