Physical Sciences Division
Researchers Map Water's Transformation from Ice to Liquid
Team determines if models match experimental reality
Results: Two popular computer models used by chemists and physicists to describe water do not correctly predict that ice melts at 32 degrees Fahrenheit, according to a study by scientists at Pacific Northwest National Laboratory and the University of Nebraska, Lincoln. More important, the study suggests that simulations of liquid water at room temperature with those models actually describe a supercooled glassy state, not the regular liquid. The study in the May 2009 Journal of Chemical Physics was the #3 download in June.
"I'm delighted that this article is receiving this kind of attention from the scientific community," said Dr. Sotiris Xantheas, the principal author on the paper. "Hopefully, it will set the standards of how those models should be used in the future to describe water."
The team pointed out the problem in two of the most popular DFT or density functional theory models in predicting the correct phase diagram of water. A DFT model describes the underlying interactions between water at the molecular level, incorporating the interactions between the nuclei and the electrons of the system. Both of those DFT models suggest that the melting temperature of ice is way too high, almost 150°F higher than the value that nature has settled on.
"So if you use those models to describe water at room temperature, you do not get the regular liquid but instead a supercooled glassy state that does not look like nature's most ubiquitous solvent," said Xantheas. This finding will help researchers fine tune DFT models to achieve a closer match to what's observed during experiments.
Why it matters: Whether studying clouds, cancer, or catalysts, the behavior of water is the key factor. Clouds form from water and act as a testbed for a variety of processes that ultimately affect earth's climate. Cancer treatments must work with the water found in cells. And catalysts both speed reactions in water and use the liquid as a feedstock to create renewable energy sources such as hydrogen.
Scientists in all of these fields need to understand, at the molecular level, how water behaves under different conditions and in different environments. For example, how does the behavior of water molecules and sulfur oxide impact the creation of acid rain? The DFT models are used to understand how changes at the atomic and molecular levels influence behaviors we can see and touch. The more realistic these models are, the easier time scientists have understanding and controlling reactions.
What's next: This discovery will assist the team members on making progress on other projects. "If you can't understand water, you can't understand proteins," said Dr. Soohaeng Yoo, one of the authors of this paper, commenting about his upcoming studies.
Acknowledgments: This work was done by Dr. Soohaeng Yoo and Dr. Sotiris Xantheas at Pacific Northwest National Laboratory and Professor Xiao Cheng Zeng at the University of Nebraska at Lincoln. The Department of Energy's EMSL, a national scientific user facility, and Basic Energy Sciences at DOE provided the computer resources.
Reference: Yoo S, XC Zeng, and SS Xantheas. 2009. "On the phase diagram of water with density functional theory potentials: The melting temperature of ice Ih with the Perdew-Burke-Ernzerhof and Becke-Lee-Yang-Parr functionals." Journal of Chemical Physics 130:221102.