Modelling Weak Interactions in the Gas Phase: From Rotational Spectroscopy to Reaction Rates Using Accurate Quantum-Chemical Approaches

During these four years of PhD, my research was focused on structural, energetic and spectroscopic characterisation of stable and reactive systems in the gas phase. A special focus has been put on the theoretical description of non-covalent interactions (NCIs) as they occur in the gas phase. The interest on these chemical bonds arises from the fact that they play a key role in many aspects of life. Indeed, they are responsible for the folding of proteins and characterise the shape of DNA and RNA. The same types of interactions drive self-assembling processes and the interaction between a receptor and its ligands. Furthermore, NCIs can influence chemical reactions by favouring one conformer with respect to others in a given pathway and they can also affect the structure of reactive intermediates and transition states. Usually, these phenomena are not due to a single interaction but to the sum of several hundreds (or thousands) non-covalent contacts occurring simultaneously and in a cooperative manner. Therefore, it is rather difficult to elucidate the type of interactions occurring and their effects on the molecular structures involved. However, one can aim at studying models of such weak bonds through the analysis of prototypical single NCIs occurring in an isolated environment. This idea can be exploited thanks to experimental methods based on rotational spectroscopy, which is an intrinsic high-resolution technique working exclusively in the gas phase, but also thanks to quantum chemistry. Rotational spectroscopy can unveil the interaction occurring in a binary system where two molecules interact and it is able to point out the effects of non-covalent interactions on the molecular structures of the fragments involved. On the other hand, quantum chemical simulations allow for: (i) exploration of the potential energy surface (PES) of the bimolecular system, thus identifying all the possible isomers that can arise from the contacts of two fragments, also in the case of reactive PESs; (ii) accurate energetic studies and decomposition of the energy to unveil the nature of the interaction; (iii) providing ab initio data useful to guide the interpretation of experimental spectra, which can be difficult to analyse due to several factors. Currently, state-of-the-art information on non-covalent complexes are obtained via a strong interplay of rotational spectroscopy and quantum chemistry. However, computational simulations show some limitations due to the challenge of accurately describing NCIs; indeed, they are extremely sensitive to the level of theory employed and an effective compromise between accuracy and computational cost is always difficult. In this context, my PhD thesis aimed at developing accurate computational models to treat medium-sized systems (20-30 atoms) dominated by NCIs such as intermolecular (binary) complexes in the gas phase, to support and/or complement experimental rotational spectroscopy, as well as reactive intermediates and transition states, to accurately describe reaction pathways. The developed models are able to provide reliable estimates for both energies and geometries. These approaches are based on coupled-cluster techniques including single and double excitations and a perturbative treatment of triples (CCSD(T) method). To reduce the computational cost without degrading the accuracy, the CCSD(T) method is employed in conjunction with Møller-Plesset perturbation theory to the second order (MP2) to account for the extrapolation to the complete basis set (CBS) limit and the core-valence correlation effects. Standard methods and their explicitly correlated counterparts, i.e., CCSD(T)-F12 and MP2-F12, have been employed. The thesis will describe how these new computational models have been built based on accurate reference data reported in the literature and the fundamental role played by diffuse functions in the basis sets. Then, the focus will move on the discussion of several examples where the new computational models, namely junChS and junChS-F12, have been used to characterise energies and structures of non-covalent complexes such as the complex between sulfur dioxide and dimethyl sulfide (SO2-DMS), the benzofuran-formaldehyde complex (C8H6OH2CO), and the trifluoroacetophenone-water (CF3COC6H5 –H2O) complex. The thesis will also address the performances of the junChS model in the case of astrochemically relevant reaction pathways, where the energetic barriers play a key role in establishing reliable reaction rate coefficients. Also in this case, a few examples will be given considering the reactions between methanimine (CH2NH) and the CP radical, between oxirane (c-C2H4O) and the CN radical, and between propene (C3H6) and the C3N radical. In these cases, the developed models had to provide accurate energies, not only for closed-shell species (all paired electrons), but also for structures with an unpaired electron (doublet state, open-shell), which are troublesome electronic configurations to describe from the theoretical point of view