An analytical potential energy model for ammonia–H2 from first principle

Quantum simulations of the hydrogen molecule on ammonia clusters

“…Even in this case, the oscillations in D e and D v versus n are to be attributed to differences in the local structure of the tetrameric ring and, consequently, to the final orientation of rb-H 2 . In many of the cases shown in Figures 2-5, the dominant interaction mode seems to be the one related to the deepest minimum of the rb-H 2 -(NH 3 ) potential, 63 with rb-H 2 donating a hydrogen bond to the lone pair side of vicinal ammonia molecules. The only deviation from a "rhombus" landing site is represented by the presence of an ad-ammonia absorbed nearby in (NH 3 ) 20 (see the top face of H 2 -(NH 3 ) 20 in Figure 5).…”

“…As the interaction energy between the hydrogen and ammonia molecules is rather small in magnitude compared with the one between two ammonia molecules, [61][62][63] it has a negligible influence on the internal structure of NH 3 and H 2 . 63 This ought to be true also for the structure of ammonia clusters, which are held together by stronger intermolecular forces than between H 2 and ammonia. It is however possible that the energy landscape of the mixed clusters does not closely follow the one for the pure species, the chance of reordering being high in our case since the pure systems may present nearly degenerate isomers (vide infra Secs.…”

“…This search is conducted using a fully dimensional (i.e., 5D) rigid-rotor potential energy surface (PES) and its 3D counterpart, spherically averaged over the orientation of H 2 . 63 A selected subset of clusters is subsequently simulated to investigate possible quantum effects. A "vis á vis" comparison of the classical and quantum results coming from the two PESs would also aid in assessing the shortcomings introduced by the spherical average from the quantitative viewpoint.…”

“…Specifically, we opted for using the PES recently developed by us on the basis of MP4 calculations, 63 for which a full 5D treatment of the rigid hydrogen molecule, and a rotationally averaged 3D version (i.e., assuming uniform spatial orientation for the para-H 2 ) are readily available. As to the 3D surface, it is worth mentioning that it represents a pointwise approximation of the true interaction when it is assumed that the rotational motion of an interacting H 2 is identical to its free-rotor (i.e., non-interacting and decoupled) counterpart at any location of its center of mass with respect to NH 3 .…”

“…68 However, the direct polarization term is effectively introduced by means of the multipolar expansion employed to describe the electrostatic component of the surface. The 5D analytical surface 63 presents a global minimum geometry with the hydrogen molecule aligned to the ammonia C 3 axis and on the side of the nitrogen, with a D e of roughly 0.7 kcal/mol (245 cm −1 ). A second minimum (D e ∼ 0.26 kcal/mol or 92.0 cm −1 ) is also found on the H-side of ammonia, with the center of mass of H 2 along a N-H vector and its bond axis being nearly perpendicular to the latter.…”