期刊
PHYSICAL CHEMISTRY CHEMICAL PHYSICS
卷 19, 期 3, 页码 2124-2130出版社
ROYAL SOC CHEMISTRY
DOI: 10.1039/c6cp07863j
关键词
-
资金
- DFH/UFA
During the last few decades, many experimental and numerical studies have tried to understand the special dynamics of water at low temperatures by measuring structural relaxation times or shear viscosity, but their conclusions strongly depended on the chosen observable and on the range of temperatures considered. Moreover, recent work [J. Chem. Phys. 2013, 138, 12A526] showed that viscosity and relaxation times could decouple at low temperature in a model binary mixture, raising questions on their equivalence to study supercooled water. Here we used molecular dynamics simulations with the promising TIP4P/2005f water force field to investigate the behavior of both the shear viscosity and the relaxation times of water in a large range of temperatures, in order to get a consistent picture of the dynamics of supercooled water. We show that the TIP4P/2005f model reproduces accurately the experimental values of both the viscosity and the diffusion coefficient over a very large range of temperatures. Focusing first on the structural relaxation dynamics, we observe a decoupling between the so-called alpha- and beta-relaxation times of water at ca. 350 K, suggesting a supercooled-like dynamics over a very large domain of temperatures. By computing shear viscosity over this domain, we compare the accuracy of several phenomenological laws for low temperature dynamics of water to describe both viscosity and alpha-relaxation time. Unlike what is usually admitted, our tests suggest those quantities are not coupled at low temperatures, and thus should not be considered equivalent. In particular, deviations from the Stokes-Einstein relation appear at lower temperatures for the viscosity than for the alpha-relaxation time. These results open new perspectives to understand the dynamics of supercooled water and show the performance of the TIP4P/2005f force field to characterize it.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据