An optimized molecular model for ammonia

The central role for that task is played by the molecular model, that determines all of them. Therefore, a balanced modeling procedure, i.e. selection of model type and parameterization, is crucial. Unfortunately, thermophysical properties usually depend on the model parameters in a highly non-linear fashion. So the development of new molecular models of technical quality is a time-consuming task. In this paper a procedure is proposed that uses information from ab initio quantum mechanical calculations to accelerate the modeling process. As an example, ammonia is regarded here. Ammonia is a well-known chemical intermediate, mostly used in fertilizer industries; another important application is its use as a refrigerant. Due to its simple symmetric structure and its strong intermolecular interactions it is also of high academic interest both experimentally and theoretically.

Different approaches can be found in the literature to construct an intermolecular potential for ammonia to be used in molecular simulation. Jorgensen and Ibrahim [4] as well as Hinchliffe et al. [5] used experimental bond distances and angles to place their interaction sites. Jorgensen and Ibrahim fitted a 12-6-3 potential plus four partial charges to results from ab initio quantum mechanical calculations, they derived for 250 orientations of the ammonia dimer using the STO-3G minimal basis set. To yield reasonable potential energies for liquid ammonia compared to experimental results, they had to scale their potential by a factor 1.26.

tractive Morse potential, and four partial charges to construct the intermolecular potential. The parameters were determined by fitting to a total of 61 points on the ammonia dimer energy surface at seven different orientations, which were calculated using the 6-31G* basis set. Hinchliffe et al. have pointed out, that the parameterization is ambiguous concerning the selection of dimer configurations and the used interaction potentials. Also the different models perform different well on various properties.

In a later work Impey and Klein [6] reparameterized the molecular model by Hinchliffe et al. They switched to an “effective” pair potential using one Lennard-Jones potential at the nitrogen nucleus site to describe the dispersive and repulsive interactions. The parameters were optimized to the radial distribution function g N-N of liquid ammonia measured by Narten [7]. Kristóf et al. [8] used this model to predict vapor-liquid equilibrium properties and found systematic deviations in both vapor pressure and saturated densities. So they decided to develop a completely new molecular model. Again they used experimental bond distances and angles to place the interaction sites.

All further parameters of their model, i.e. the partial charges on all atoms and the parameters of the single Lennard-Jones For their simulations, Kristóf et al. used the Gibbs ensemble Monte Carlo (GEMC) technique [9,10] with an extension to the N pH ensemble [11,12]. This methods have some difficulties simulating strongly interacting fluids, yielding to relatively large statistical uncertainties. When applying our methods for the

The modeling philosophy followed here is to keep the molecular model as simple as possible. Therefore, the molecule is assumed rigid and non-polarizable, i.e. a single state-independent set of parameters is used. Hydrogen atoms are not modeled explicitely, a united-atom approach is used.

For both present models, a single Lennard-Jones potential was assumed to describe the dispersive and repulsive interactions. The electrostatic interactions as well as hydrogen bonding were modeled by a total of four partial charges. This modeling approach was found to be appropriate for other hydrogen bonding fluids like methanol [13], ethanol [14], and formic acid [15] and was also followed by Impey and Klein [6] and Kristóf et al. [8] for ammonia. Thus, the potential energy u ij between two ammonia molecules i and j is given by

where a is the site index of charges on molecule i and b the site index of charges on molecule j, respectively. The site-site distances between molecules i and j are denoted by r ij for the single Lennard-Jones potential and r ijab for the four partial charges, respectively. σ and ε are the Lennard-Jones size and energy parameters, while q ia and q jb are the partial charges located at the sites a and b on the molecules i and j, respectively. Finally, ǫ 0 denotes the permittivity of the vacuum.

To keep the modeling procedure as independent as possible from the availability of specific information, no experimental bond lengths or angles were used her

…(Full text truncated)…