Physical Sciences Division
It's all in the water . . .
Graphical representation of the liquid at a single time step. Each atom in the simulation is represented by 32 replicas to account for quantum mechanical effects.
Calculated distributions of O-O distances compared to experimental measurements. Red curves are computed quantum mechanically, blue curves are classical, black curves are measured with neutron scattering. R Å is the distance between oxygen atoms), Goo(R) is the normalized distribution.
Same as center figure for the distances between OH radicals.
Water is everywhere—it covers about 75 percent of the earth's surface and it makes up 50-65 percent of the human body: blood is 83% water, muscle is made up of 75% water and even bones have a water content of 22%. Because of its ubiquitous nature and its importance in many biological and chemical processes, both inside and outside of the body, understanding water is critical to understanding our world.
Researchers at Pacific Northwest National Laboratory have created a new model for water based on molecular level information that can predict the properties of water molecules as they interact in small clusters at the microscopic level as well as in large collections at the macroscopic level. The PNNL model is unique in its ability to predict the properties of water molecules at both levels and the range in between.
Led by PNNL Laboratory Fellow Sotiris Xantheas, researchers are trying to answer questions about the structure of liquid water, its energy, its ability to transport ions, and its electrical properties. Understanding water's properties through the entire range from microscopic to macroscopic levels is crucial to better understanding of the role of water in catalyzing chemical reactions occurring both in the human body as well as in the earth's environment.
For instance, understanding water's macroscopic transport properties is important when transporting a medicine to a specific site in the human body because large collections of molecules are involved in this process. Once the medicine arrives at the specific site, it must perform a specific function. Only the closest few water molecules are involved at this specific function that occurs at the microscopic level.
PNNL researchers are using a synergistic approach to gathering information for their model. "We try to get information theoretically by studying the properties of small clusters of water molecules," said PNNL Lab Fellow, Sotiris Xantheas, who is leading the research. "Then, to check for accuracy, we compare this information to experimental information as it becomes available."
In the next step, PNNL scientists used this molecular level information to build a model. "There is no guarantee that if you go along this route, you will get the right answers," Xantheas said. "You have to make sure you have all the correct physics in the model."
The big question for Xantheas and his team was whether their model could go from predicting the properties of water at the microscopic level to predicting them at the macroscopic level.
Using the EMSL supercomputer, they tested the model for accuracy by running their simulations against already known data for macroscopic properties of liquid water. Because the PNNL model is atom based, every atom in the system of liquid water must be described explicitly. Calculations that would have taken more than 20 years on a single processor were completed in 28 days by the EMSL super computer's 256 processors and in-house-developed software.
These results indicate that the PNNL model is able to predict the structural and energetic properties of liquid water through the entire range from clusters to liquid water at ambient conditions. More tests are under way in order to simulate water's anomalous properties such as the density maximum at 4°C, the freezing and melting points as well the relative stability of the various ice forms.
This research was sponsored by the U.S. Department of Energy's Office of Basic Energy Sciences. Scientists contributing to this research include Sotiris Xantheas, former post-doctoral fellow Christian Burnham, current post-doctoral fellow Georgios Fanourgakis, and staff members Gregory Schenter, Wibe de Jong and Edo Apra.