Library

Notes: Molecular dynamics simulation of 216 water molecules (ST2 model) between charged flat electrodes 2.362 nm apart showed layering with a few molecules at each surface that broke H bonds with the bulk and oriented their charges towards the electrode. Compared to uncharged electrodes, the atomic and molecular distributions were unsymmetric. When a lithium and an iodide ion were substituted at random for two water molecules, the iodide ion contact adsorbed on the anode with no water molecules between it and the electrode. The iodide ion appeared weakly solvated on the solution side to water molecules that preferred to engage in hydrogen bonding with the network of the bulk solvent. In contrast, the lithium ion adsorbed without losing its primary solvation shell of six water molecules and was never observed further than two water molecules removed from the electrode. Its average position corresponded to an ion supported on a tripod of three waters. The average solvation number was not changed upon adsorption in this configuration. These qualitative observations and some quantitative results afford striking confirmation on the one hand and new insight on the other of some aspects of the standard model of the adsorption of ions on electrode surfaces. Time durations for simulations were generally between 200 and 800 ps with a basic integration time step of 2 fs.

Notes: The adsorption geometry and the nature of the interaction of the SCN molecule at an on-top site of the Ag(100) surface have been investigated using ab initio cluster model wave functions. The SCN anion, SCN−, is a bidentate ligand. If the SCN–Ag bond is ionic, we could expect, by analogy with thiocynate–metal complexes, that SCN could be bound to a metal surface through either the N end or the S end. We show that the chemisorption bond for SCN/Ag is very ionic and that the interaction between chemisorbed SCN and the Ag surface is largely electrostatic. The most important bonding mechanism is the polarization of the Ag surface due to the presence of SCN−. However, we do find that there is a small, but non-negligible, covalent interaction. There is a very small energetic cost to change the angle of SCN with respect to the surface between a perpendicular and a parallel orientation. We contrast this with the case of a covalently bonded adsorbate, CO/Ag(100), where the π bond strongly favors orientations near perpendicular. The flat potential energy curve for bending SCN suggests that the adsorption geometry at high SCN coverage may be largely determined by nonbonding interactions between adjacent adsorbates.

Notes: The nature of the bonding between halogen atoms (F, Cl, and Br) and the Ag (111) surface has been investigated by analyzing ab initio Hartree–Fock wave functions for cluster models of the Ag surface and a halogen atom. Using a variety of criteria, we conclude that the bonding is ionic and that the halogen ionicity is essentially −1. The measures of ionicity reported are (a) the expectation value of a projection operator which provides an indication of the total charge associated with the halogen atom, (b) the analysis of the dipole moment curve as function of distance, (c) the effect on the equilibrium bond distances of a uniform external electric field, and (d) the decomposition of the interaction energy into the sum of different contributions. This latter analysis shows that the bonding arises, almost entirely, from two effects: (1) the Coulomb attraction between the charged halogen and the metal and (2) the intraunit polarization of the metal and halogen subunits.

Notes: A new class of potential suitable for modeling the adsorption of water on different metal sites is described. The new potentials are simple in form and convenient for use in computer simulations. In their real space form they comprise three parts: A pairwise sum of spatially anisotropic 12-6 potentials, a pairwise sum of isotropic short range potentials, and an image potential. Two modifications of the potential are developed. In the first, the anisotropic potential acts only on the oxygen atom and not on the protons. In the second, the potential acts on all the atoms of the water molecule. In practical calculations it is convenient to transform the potential to a reciprocal space form in the manner described by Steele [Surf. Sci. 36, 317 (1973)]. Adsorption of water at top, bridge, and hollow sites on (100), (110), and (111) surfaces of Pt, Ni, Cu, and Al were studied using two fitting parameters and the results compared with previous theoretical calculations.

Notes: Constant temperature molecular dynamics has been used to simulate the adsorption of hydrated halide ions X−=F−, Cl−, Br− and I−, and lithium ion Li+ on flat uniformly charged surfaces. The simulations were done with either 214 water molecules and two ions (Li+ and X−) in a box 2.362 nm deep or with 430 water molecules and the two ions in a box 4.320 nm deep. The boxes were periodically replicated in the xy directions. The magnitude of the surface charge on the box ends was ±0.11 e/(nm)2, corresponding to an electric field of 2×107 V/cm. The lateral dimensions of the simulation cell were 1.862 nm×1.862 nm (x×y) in each case. All of the water molecules and ions interacted with the end walls via a weak 9-3 potential. The Stillinger ST2 water model and parameters optimized for alkali halides interacting with the model ST2 water molecule were used in the calculations. Common particles of truncating the interactions at a finite distance (0.82 nm) and switching off Coulomb interactions at small distances were followed. The temperature was set at T=2.411 kJ/mol (290 K). Some of the properties calculated were distribution density profiles for ions and water across the gap important for comparisons with Gouy–Chapman theory, adsorbed ion–water pair correlation functions, and the number of water molecules in the first and second hydration shells of the ions as a function of time. The time spent by a water molecule in the hydration shell was calculated to be approximately ten times longer for lithium than any other ion.The correlation between distance from the electrode and hydration number was studied and generally found to be pronounced for the larger anions. Comparison of the dynamics of the common ion Li+ for different anions revealed the subtle influence of a transcell interaction in the 2.362 nm thick film. In the given field, the smallest ions Li+ and F− remained fully solvated at all times. Chloride behaved quite differently. Part of the time this ion was far enough away from the electrode to be fully hydrated and part of the time it was in physical contact (i.e., physisorbed) on the electrode with no water molecules interposed between it and the electrode. Bromide favored contact adsorption over full hydration most of the time. Iodide was observed to be contact adsorbed almost all of the time. These simulations provide new insights on the behavior of strongly hydrated ions at surfaces and how the transition from noncontact to "contact'' adsorption occurs.