We are all familiar with the nucleation of water into ice in our daily lives. However, despite being studied since antiquity, scientists still do not fully understand how ice forms at the molecular scale. This is particularly true for so-called heterogeneously catalysed nucleation, in which water is encouraged to form through the presence of an "ice-nucleating agent" – a microscopic "seed" particle of salt or sand, for example.
Michaelides and Morgenstern have now analysed a hexamer using scanning tunnelling microscopy on a sample of water that formed on a hydrophobic metallic surface cooled down to just 5 K. Together with first-principle theoretical calculations running on "massively parallelized" supercomputers, the researchers identified a pattern of alternating shorter and longer bonds between the water molecules. This was a surprising result because water molecules usually bond to each other with equal strength in ice.
"This result tells us something new about hydrogen bonds – which are of the utmost importance in being the physical force that holds water together in its condensed phases," Michaelides told nanotechweb.org. "For the first time, we have identified a clear and specific competition between the ability of water molecules to bond to metal surfaces and to accept hydrogen bonds at the same time."
Michaelides says that the study is a new way of thinking about the structures of supported water clusters that form on solid surfaces. "The result has the potential to impact all disciplines concerned with acquiring a molecular-scale understanding of interfacial water, such as atmospheric chemistry, astrophysics and biology," he added.
The team's interest lies in understanding ice nucleation and how it relates to cloud formation and environmental chemistry. The work will help researchers understand how water interacts with aerosols and microscopic dust particles in the atmosphere. These processes drive cloud formation and are therefore important for our climate.
The researchers reported their work in Nature Materials.