Selected by the National Center for Supercomputing Applications last year as a Blue Waters Professor, Hirata was able to take advantage of the computing time set aside for these distinguished researchers to explore a very cold topic: ice. And a resulting article was tapped as an “editor’s pick” by the Journal of Chemical Physics and featured on the journal’s website the week of May 30th.
“Chemistry of water in all three phases largely defines our planetary environment. Its influence is felt in subjects ranging from geology to climate, to biology and ecology, to geopolitics and history. Such a seemingly simple phenomenon—that the water volume collapses upon melting—has an immense impact on every structure and reaction found on the planetary surface and thus on every life form, but it derives from a subtle interplay of water’s peculiar chemical bonding and dynamics,” Hirata wrote in his 2015 Blue Waters annual report. In fact, notes Hirata, computationally determining some of the most delicate thermodynamic and response properties of ice and liquid water, such as melting temperature, from first principles (i.e. with no reference to any experimental data) has not been possible.
Hirata and his Illinois collaborators Soohaeng Y. Willow and Michael A. Salim wrote they “combined the massive computational power of Blue Waters with an algorithmic breakthrough that made ab initio quantum chemistry calculations faster, scalable in parallel, and thus applicable to condensed-phase systems.” The team studied both ice and liquid water, studying properties that are difficult to study experimentally. During their runs on Blue Waters the team also conducted what they believe is most likely the first molecular dynamics simulation for liquid water using on-the-fly atomic forces evaluated by the ab initio electron-correlated molecular orbital method.