Condensation of H 2 in the interstellar medium (ISM) has long been seen as a possibility, either by deposition on dust grains or thanks to a phase transition combined with self-gravity. H 2 condensation might explain the observed low efficiency of star formation and might help to hide baryons in spiral galaxies. Our aim is to quantify the solid fraction of H 2 in the ISM due to a phase transition including self-gravity for different densities and temperatures in order to use the results in more complex simulations of the ISM as subgrid physics. We used molecular dynamics simulations of fluids at different temperatures and densities to study the formation of solids. Once the simulations reached a steady state, we calculated the solid mass fraction, energy increase, and timescales. By determining the power laws measured over several orders of magnitude, we extrapolated to lower densities the higher density fluids that can be simulated with current computers. The solid fraction and energy increase of fluids in a phase transition are above 0.1 and do not follow a power law. Fluids out of a phase transition are still forming a small amount of solids due to chance encounters of molecules. The solid mass fraction and energy increase of these fluids are linearly dependent on density and can easily be extrapolated. The timescale is below one second, the condensation can be considered instantaneous. The presence of solid H 2 grains has important dynamic implications on the ISM as they may be the building blocks for larger solid bodies when gravity is included. We provide the solid mass fraction, energy increase, and timescales for high density fluids and extrapolation laws for lower densities.

Molecular clouds typically consist of 3/4 H2, 1/4 He and traces of heavier elements. In an earlier work we showed that at very low temperatures and high densities, H2 can be in a phase transition leading to the formation of ice clumps as large as comets or even planets. However, He has very different chemical properties and no phase transition is expected before H2 in dense interstellar medium (ISM) conditions. The gravitational stability of fluid mixtures has been studied before, but these studies did not include a phase transition. First, we study the gravitational stability of van der Waals fluid mixtures using linearized analysis and examine virial equilibrium conditions using the Lennard-Jones intermolecular potential. Then, combining the Lennard-Jones and gravitational potentials, the non-linear dynamics of fluid mixtures are studied via computer simulations using the molecular dynamics code LAMMPS. Along with the classical, ideal-gas Jeans instability criterion, a fluid mixture is always gravitationally unstable if it is in a phase transition because compression does not increase pressure. However, the condensed phase fraction increases. In unstable situations the species can separate: in some conditions He precipitates faster than H2, while in other conditions the converse occurs. Also, for an initial gas phase collapse the geometry is essential. Contrary to spherical or filamentary collapses, sheet-like collapses starting below 15K easily reach H2 condensation conditions because then they are fastest and both the increase of heating and opacity are limited. Depending on density, temperature and mass, either rocky H2 planetoids, or gaseous He planetoids form. H2 planetoids are favoured by high density, low temperature and low mass, while He planetoids need more mass and can form at temperature well above the critical value.

The existence of substellar cold H2 globules in planetary nebulae and the mere existence of comets suggest that the physics of cold interstellar gas might be much richer than usually envisioned. We study the case of a cold gaseous medium in ISM conditions which is subject to a gas-liquid/solid phase transition. First the equilibrium of general non-ideal fluids is studied using the virial theorem and linear stability analysis. Then the non-linear dynamics is studied by using simulations to characterize the expected formation of solid bodies analogous to comets. The simulations are run with a state of the art molecular dynamics code (LAMMPS). The long-range gravitational forces can be taken into account with short-range molecular forces with finite limited computational resources by using super-molecules, provided the right scaling is followed. The concept of super-molecule is tested with simulations, allowing us to correctly satisfy the Jeans instability criterion for one-phase fluids. The simulations show that fluids presenting a phase transition are gravitationally unstable as well, independent of the strength of the gravitational potential, producing two distinct kinds of sub-stellar bodies, those dominated by gravity ("planetoids") and those dominated by molecular attractive force ("comets"). Observations, formal analysis and computer simulations suggest the possibility of the formation of substellar H2 clumps in cold molecular clouds due to the combination of phase transition and gravity. Fluids presenting a phase transition are gravitationally unstable, independent of the strength of the gravitational potential. Arbitrarily small H2 clumps may form even at relatively high temperatures up to 400 - 600K, according to virial analysis. The combination of phase transition and gravity may be relevant for a wider range of astrophysical situations, such as proto-planetary disks.