CO depletion in infrared dark clouds

Context. Infrared dark clouds (IRDCs) are cold, dense structures that are likely representative of the initial conditions of star formation. Many studies of IRDCs employ CO to investigate cloud dynamics, but CO can be highly depleted from the gas phase in IRDCs, which affects its fidelity as tracer. The CO depletion process is also of great interest in astrochemistry because CO ice in dust grain mantles provides the raw material for the formation of complex organic molecules.Aims. We study CO depletion towards four IRDCs to investigate its correlation with the H2 number density and dust temperature, calculated from Herschel far-infrared images.Methods. We used 13CO J = 1 → 0 and 2 → 1 maps to measure the CO depletion factor, fD, across IRDCs G23.46-00.53, G24.49-00.70, G24.94-00.15, and G25.16-00.28. We also considered a normalised CO depletion factor, f′D, which takes a value of unity, that is, no depletion, in the outer lower-density and warmer regions of the clouds. We then investigated the dependence of fD and f′D on the gas density, nH, and dust temperature, Tdust.Results. The CO depletion rises as the density increases and reaches maximum values of f′D ∼ 10 in some regions with nH ≳ 3 × 105 cm−3, although with significant scatter at a given density. We find a tighter, less scattered relation of f′D with temperature that rapidly rise for temperatures ≲18 K. We propose a functional form f′D = exp(T0/[Tdust − T1]) with T0 ≃ 4 K and T1 ≃ 12 K to reproduce this behaviour.Conclusions. We conclude that CO is strongly depleted from the gas phase in cold, dense regions of IRDCs. This means that if it is not accounted for, CO depletion can lead to an underestimation of the total cloud masses based on CO line fluxes by factors up to ∼5. These results indicate a dominant role for thermal desorption in setting near equilibrium abundances of gas-phase CO in IRDCs and provide important constraints for astrochemical models and the chemodynamical history of gas in the early stages of star formation.FullText for HTML: https://doi.org/10.1051/0004-6361/202557102