A sensitivity analysis of interstellar ice chemistry in astrochemical models

Interstellar ices play a large role in the chemical evolution of prestellar objects. With new data from the James Webb Space Telescope, we have gained new information about ices in different astronomical environments. Astrochemical models are crucial to bridge the gap between the short timescales of chemical reactions and experiments, and the millions of years of evolution of prestellar objects. Surface chemistry in these models is very complicated as a result of the many required parameters to model it. Many of these parameters, such as binding energies and reaction rate coefficients on grains, are often poorly constrained.

I will present our sensitivity analysis to determine the most important chemical parameters to assess which parameters should get priority for new measurements and calculations. Using randomly sampled binding energies, diffusion barriers, energy barriers for reactions, and kinetic desorption and diffusion prefactors, we determined the correlations of the main ice abundances calculated with UCLCHEM, a gas-grain astrochemical code, on each of these parameters.

By running a wide grid of physical conditions (with varying temperature, density, cosmic-ray ionization rate and UV field strength), we find that the main correlations of abundances of many ice species are the diffusion barriers of small radicals such as H, N, CH and CH₃. Thus, these should be determined more exactly to further increase the accuracy of astrochemical models, leading to a better understanding and interpretation of observations. I will also demonstrate how machine learning can help us calculate these parameters.