Simulation system, aqueous interfaces

By far the most common methods of studying aqueous interfaces by simulations are the Metropolis Monte Carlo (MC) technique and the classical molecular dynamics (MD) techniques. They will not be described here in detail, because several excellent textbooks and proceedings volumes (e.g., [2-8]) on the subject are available. In brief, the stochastic MC technique generates microscopic configurations of the system in the canonical (NYT) ensemble the deterministic MD method solves Newton s equations of motion and generates a time-correlated sequence of configurations in the microcanonical (NVE) ensemble. Structural and thermodynamic properties are accessible by both methods the MD method provides additional information about the microscopic dynamics of the system. 

The interface between the aqueous and the organic liquids is more or less flat, of about 10 A wide. Its precise position, defined as the intersection between the density curves of the two solvents, fluctuates during the dynamics In the following, we mostly focus on the location and solvent environment of the solutes. Because of the dynamics nature of the simulated systems, it is not easy to depict shortly their evolution during the simulations. Some remain close to their initial position. For others, major changes may take place. They are illustrated by selected snapshots or cumulated views extracted from the trajectories. 

At the interface, the system organizes differently. All anions are immersed in the aqueous phase, while the cations are at, or close to the two interfaces, which become positively charged. Thus, during the simulation, some cations diffuse through the organic phase (from the liquid-liquid interface to the water / air interface), presumably to reduce their mutual electrostatic repulsions. Although Cl" and NTMA" " never form intimate ion pairs, they display some attractions, as evidenced by the cumulated positions or density curves of Cl" anions of NTMA+CT, compared to those of K+Cl" (Figure 2). 

In this article, some of the recent advances in computer simulations of interfacial phenomena are presented. To keep the focus on biological relevance, only interfaces between water and organic liquids will be discussed. These systems will be hereafter referred to as aqueous interfaces. For broader overviews, which also consider liquid-vapor and liquid-solid interfaces, the reader is referred to several recent articles. 

Computational constraints impose spatial and temporal limitations on simulated systems. The number of atoms considered is typically in the 10 -10 range. The corresponding cross-sectional length of the interface varies between 2 and 4 nm and each lamella is 2 to 5 nm wide. For the aqueous phase, this is equivalent to approximately 7-18 water diameters. The spatial extent of the system is primarily limited by the rapidly growing number of intermolecular interactions. In the pairwise additive approximation, this number is N x (N — l)/2, where N is the number of atoms in the system. In practice, pair interactions of an atom with other atoms are usually truncated spherically. The largest possible truncation distance is half the shortest box edge. 

Modem computer simulation of aqueous interfaces date from only 1985. In this short period, they have yielded new insights into the unique properties of interfacial systems, which distinguish them from bulk phases. Perhaps the most important of these properties is the existence of very different environments, polar and nonpolar, in direct proximity. As a result, aqueous interfaces tend to concentrate and organize organic material. In particular, they provide ideal surroundings for amphiphilic molecules, which can simultaneously have their polar parts immersed in water and nonpolar parts immersed... 

Simulations of water in synthetic and biological membranes are often performed by modeling the pore as an approximately cylindrical tube of infinite length (thus employing periodic boundary conditions in one direction only). Such a system contains one (curved) interface between the aqueous phase and the pore surface. If the entrance region of the channel is important, or if the pore is to be simulated in equilibrium with a bulk-like phase, a scheme like the one in Fig. 2 can be used. In such a system there are two planar interfaces (with a hole representing the channel entrance) in addition to the curved interface of interest. Periodic boundary conditions can be applied again in all three directions of space. 

Two excellent examples of this membrane system have been developed, NS-lOO and PA-300 (5,15). The NS-lOO membrane was made by impregnating a polysulfone support with a 0.67 percent aqueous solution of polyethylenlmine, draining away excess reagent, then contacting the film with a 0.1 percent solution of toluenediisocyanate in hexane. An ultrathln polyurea barrier layer formed at the interface. This membrane was then heat-cured at 110°C. A later version of this membrane was developed (designated NS-101), which used isophthaloyl chloride in place of toluenedilsocyanate, producing a polyamide (16). With either type of membrane, salt rejections in simulated seawater tests at 1000 psi exceeded 99 percent. 

---