A new data structure for accelerating kinetic Monte Carlo method

The kinetic Monte Carlo simulation is a rigorous numerical approach to study the chemistry on dust grains in cold dense interstellar clouds. By tracking every single reaction in chemical networks step by step, this approach produces more precise results than other approaches but takes too much computing time. Here we present a method of a new data structure, which is applicable to any physical conditions and chemical networks, to save computing time for the Monte Carlo algorithm. Using the improved structure, the calculating time is reduced by 80 percent compared with the linear structure when applied to the osu-2008 chemical network at 10 K. We investigate the effect of the encounter desorption in cold cores using the kinetic Monte Carlo model with an accelerating data structure. We found that the encounter desorption remarkably decreases the abundance of grain-surface H-2 but slightly influences the abundances of other species on the grain.