Physicists Trapped Water Between Soft Membranes So It Could Not Freeze, and Finally Watched It Turn Into Glass

Water has a stretch of temperature that experimenters have never been able to get into. Below about minus 40 Celsius, pure water crystallizes so fast that no instrument can catch it in a liquid state, and the region between that point and the deep-freeze glassy phase has been called no-man's land for decades. An international team has now walked through part of it by giving water nowhere to build ice.

The trick was soft nanoconfinement. Researchers trapped water inside ultra-thin lipid membranes made of phytantriol, a molecule that self-assembles into layered structures with water channels a few nanometers across. Ice needs room to nucleate and grow an ordered crystal; a channel that narrow denies it that room. The membranes themselves stayed fluid throughout, so the water was not pinned against a rigid wall — the confinement was soft, which is closer to the situation inside a living cell than a silica pore is.

The measurements were done at Australia's ANSTO facilities and published in Nature Communications, with Patrick Züblin as lead author. Small- and wide-angle X-ray scattering mapped the structure down to minus 120 Celsius and confirmed no ice formed. Neutron scattering did the harder job: the high-resolution backscattering spectrometer Emu and the time-of-flight spectrometer Pelican track how hydrogen atoms move, on timescales from picoseconds to nanoseconds, which is where a liquid stops behaving like a liquid.

The sharp change showed up between minus 35 and minus 20 Celsius. Across that window the water's molecular dynamics shifted markedly — molecules stopped rearranging on the timescale of the measurement and started rattling in place, the signature of a liquid falling out of equilibrium and locking into a glass. Low-temperature microscopy, nuclear magnetic resonance and computer simulations backed up the scattering data.

Why this matters beyond the phase diagram: the debate over water's low-temperature behavior has run for more than a century and turns on whether liquid water has a second critical point, hidden in the region where ice always intervenes. Some models predict two distinct liquid forms of water, high-density and low-density, separated by a transition line ending in that critical point. Being able to hold water liquid, in a soft environment, across a wider temperature range than before gives those models something to be tested against.

The practical end is cryopreservation. Freezing kills cells mostly by making ice, and vitrification — turning the water in tissue into a glass instead of a crystal — is how sperm, embryos and some tissues survive storage today. It currently requires high concentrations of cryoprotectant chemicals that are themselves toxic. A physical route to vitrification that relies on confinement geometry rather than antifreeze chemistry is a different lever on the same problem, and the same logic applies to food freezing, where ice crystals are what wreck texture.

The team notes that the water in a cell is confined by soft membranes too, which means the behavior they measured may be closer to biological reality than any bulk-water measurement.