A computational quantum study on a proton bound heavy noble gas cluster: the [Kr2H]⁺

Proton-bound noble gas complexes have attracted considerable interest due to their unusual chemical bonding and potential relevance in interstellar chemistry. Among these species, the Kr₂H⁺ cation represents an interesting system for investigating the interplay between electronic structure, intermolecular interactions, and quantum nuclear dynamics.

In this work, a computational investigation of the Kr₂H⁺ complex is presented, combining high-level ab initio calculations, machine learning techniques, and quantum nuclear dynamics. Electronic structure calculations were performed at the CCSD(T) level, employing extended correlation-consistent basis sets and accounting for scalar relativistic effects through the Douglas–Kroll–Hess Hamiltonian. The equilibrium geometry and main dissociation pathways were characterized, providing the reference data required to construct three-dimensional potential energy surfaces (PESs).

Two different machine learning approaches, Reproducing Kernel Hilbert Space (RKHS) interpolation and Neural Networks (NNs), were employed to represent the PESs. Their predictive accuracy and ability to reproduce the underlying potential energy landscape were systematically evaluated, with particular attention to the low-energy regions and the strongly anharmonic behavior associated with proton motion.

Finally, quantum nuclear dynamics calculations were performed using the DVR3D program to investigate the vibrational bound states of the complex. The resulting vibrational energy levels and wavefunctions provide insight into its anharmonic dynamics while highlighting the importance of an accurate description of the PES topology for reliable spectroscopic predictions.
Overall, this study illustrates the potential of combining high-level electronic structure methods with machine learning and quantum dynamics to investigate molecular systems characterized by complex intermolecular interactions and large-amplitude nuclear motion