Center for Energy Resources Engineering, Department of Chemical and Biochemical Engineering, Technical University of Denmark, Kgs. Lyngby 2800, Denmark.
Research Institute of Petroleum Exploration & Development (RIPED), PetroChina, Beijing 100083, China.
J Phys Chem B. 2021 Jul 8;125(26):7187-7198. doi: 10.1021/acs.jpcb.1c02816. Epub 2021 Jun 29.
The ability to form hydrogen bonds is one of the most important factors behind water's many anomalous properties. However, there is still no consensus on the hydrogen bond structure of liquid water, including the average number of hydrogen bonds in liquid water. We use molecular dynamics simulations of the polarizable iAMOEBA water model for investigating the hydrogen bond characteristics of liquid water over a wide range of temperatures and pressures. Geometric definitions of a hydrogen bond often use a rectangular region on the plane of hydrogen bond distances and angles. In this work, we find that an elliptical region is more appropriate for the identification of hydrogen bonds, based on statistically favorable molecular configurations. The two-dimensional potential of mean force (PMF) landscape along the hydrogen bond distance (O-H) and angle (O-H-O) is calculated for identifying the statistically favored molecular configurations, which is then used for defining hydrogen bond formation as well as the strength of a hydrogen bond. We further propose a new approach to characterize the hydrogen bonds as strong when the PMF is lower than -2 kT. Using this definition, a consistent explanation for the different average numbers of hydrogen bonds in water is obtained in agreement with the literature. Simulations are also performed with the rigid and nonpolarizable TIP4P/2005 water model. Both water models are qualitatively consistent in predicting the distribution of double-, single-, and non-donor configurations, in line with experimental data, while the iAMOEBA water model yields more quantitatively precise results, including a 10-15% double-donor fraction at 90 °C and 1 atm. The method is also demonstrated to be applicable to the recent, and more general, three-dimensional PMF-based definition of hydrogen bonds.
形成氢键的能力是水的许多异常性质背后最重要的因素之一。然而，对于液态水的氢键结构，包括液态水中氢键的平均数量，仍然没有共识。我们使用极化 iAMOEBA 水模型的分子动力学模拟，在很宽的温度和压力范围内研究液态水的氢键特性。氢键的几何定义通常使用氢键距离和角度平面上的矩形区域。在这项工作中，我们发现基于统计有利的分子构型，椭圆区域更适合氢键的识别。计算了氢键距离（O-H）和角度（O-H-O）的二维平均力势能（PMF）景观，以识别统计有利的分子构型，然后用于定义氢键的形成以及氢键的强度。我们进一步提出了一种新的方法来表征氢键，当 PMF 低于-2 kT 时，氢键为强。使用这个定义，我们得到了与文献一致的关于水中氢键平均数量不同的一致解释。还使用刚性和非极化 TIP4P/2005 水模型进行了模拟。两种水模型在预测双、单和非供体构型的分布方面定性上是一致的，与实验数据一致，而 iAMOEBA 水模型得出了更定量精确的结果，包括在 90°C 和 1 大气压下 10-15%的双供体分数。该方法还被证明适用于最近的、更通用的基于三维 PMF 的氢键定义。
