Quantum chemical protocols for modeling reactions and spectra in astrophysical ice analogs: the challenging case of the C⁺ + H₂O reaction in icy grain mantles.

Abstract

Icy grain mantles that accrete on refractory dust particles in the very cold interstellar medium or beyond the snow line in protoplanetary disks serve as minute incubators for heterogeneous chemistry. Ice mantle chemistry can differ significantly from the gas phase chemistry that occurs in these environments and is often richer. Modeling ices and their chemistry is a challenging task for quantum theoretical methods, but theory promises insight into these systems that is difficult to attain with experiments. Density functional theory (DFT) is predominately employed for modeling reactions in icy grain mantles due to its favorable scalability, but DFT has limitations that risk undercutting its reliability for this task. In this work, basic protocols are proposed for identifying the degree to which DFT methods are able to reproduce experimental or higher level theoretical results for the fundamental interactions upon which ice mantle chemistry depends, including both reactive interactions and non-reactive scaffolding interactions. The exemplar of this study is the reaction of C(+) with H2O, where substantial methodological differences are found in the prediction of gas phase relative energetics for stationary points (about 10 kcal mol(-1) for the C-O bond energy of the H2OC(+) intermediate), which in turn casts doubt about employing it to treat the C(+) + H2O reaction on an ice surface. However, careful explorations demonstrate that B3LYP with small correlation consistent basis sets performs in a sufficiently reliable manner to justify using it to identify plausible chemical pathways, where the dominant products were found to be neutral HOC and the CO(-) anion plus one and two H3O(+) cations, respectively. Predicted vibrational and electronic spectra are presented that would serve to verify or disconfirm the pathways; the latter were computed with time-dependent DFT. Conclusions are compared with those of a recent similar study by McBride and coworkers (J. Phys. Chem. A, 2014, 118, 6991).