SIMBAD references

Water is usually the main component of ice mantles, which cover the cores of dust grains in cold portions of dense interstellar clouds. When molecular hydrogen is adsorbed onto an icy mantle through physisorption, a common assumption in gas-grain rate-equation models is to use an adsorption energy for molecular hydrogen on a pure water substrate. However, at high density and low temperature, when H₂ is efficiently adsorbed onto the mantle, its surface abundance can be strongly overestimated if this assumption is still used. Unfortunately, the more detailed microscopic Monte Carlo treatment cannot be used to study the abundance of H₂ in ice mantles if a full gas-grain network is utilized. We present a numerical method adapted for rate-equation models that takes into account the possibility that an H₂ molecule can, while diffusing on the surface, find itself bound to another hydrogen molecule, with a far weaker bond than the H₂-water bond, which can lead to more efficient desorption. We label the ensuing desorption ``encounter desorption''. The method is implemented first in a simple system consisting only of hydrogen molecules at steady state between gas and dust using the rate-equation approach and comparing the results with the results of a microscopic Monte Carlo calculation. We then discuss the use of the rate-equation approach with encounter desorption embedded in a complete gas-grain chemical network. For the simple system, the rate-equation model with encounter desorption reproduces the H₂ granular coverage computed by the microscopic Monte Carlo model at 10K for a gas density from 10⁴ to 10¹²cm^–3, and yields up to a factor 4 difference above 10¹²cm^–3. The H₂ granular coverage is also reproduced by a complete gas-grain network. We use the rate-equation approach to study the gas-grain chemistry of cold dense regions with and without the encounter desorption mechanism. We find that the grain surface and gas phase species can be sensitive to the H₂ coverage, up to several orders of magnitude, depending on the species, the density, and the time considered. The method is especially useful for dense and cold environments, and for time-dependent physical conditions, such as occur in the collapse of dense cores and the formation of protoplanetary disks. It is not significantly CPU time consuming, so can be used for example with complex 3D chemical-hydrodynamical simulations.