The interstellar medium contains dust, in a proportion ~1:1012 gas-phase particles, or 1:100 by mass. The sticking of gas-phase particles onto the grains results in surface chemistry, as reactive species migrate around the grain surface to meet and react. The products of this chemistry are mostly simple hydrides such as water, as well as organic (i.e. carbon-bearing) species, which largely remain on the grains to build up an ice mantle. Later processing of this ice can produce significantly more complex organic molecules that may survive to reach the protoplanetary stage and beyond.
Grain-surface chemistry occurs mainly by the so-called Langmuir-Hinshelwood process; reactive species migrate around the grain surface by a series of thermal hops between binding sites (i.e. local potential minima in the van der Waals interaction between an atom/molecule and the grain surface). Most models of interstellar dust grain-surface chemistry use methods that do not take account of the physical positions of the atoms and molecules (all models except those of Cuppen, in fact). While some models take account of the layering of ice material, they nevertheless implicitly assume a spherical grain and ice mantle.
I have constructed a new chemical kinetic model of dust grain-surface chemistry (Garrod, ApJ, 2013) that begins with a dust grain whose physical size and shape may be defined precisely, in three dimensions. Furthermore, unlike any other astrochemical model, the method does not pre-define the positions of binding sites on the surface. These are calculated “on-the-fly”, as atoms and molecules diffuse around the grain surface. The method finds the potential minimum on the other side of each hop; this allows surface species to take “off-lattice” positions, and allows the ice as a whole to build up a structure according to the binding properties of the reactants.
The model is a Monte Carlo approach, in that it treats the chemical kinetics as a series of consecutive processes chosen randomly according to their rates, following the method of Gillespie (1976). Each hopping, accretion, or evaporation event is considered to be such a process in this treatment.
Because this method traces the positions of each atom/molecule, the results may be easily visualized using direct three-dimensional videos.