A significant process in the formation of the unique atmosphere of Io, a Jovian moon, is collision-induced dissociation of sulfur dioxide. The rarefied nature of the Ionian atmosphere makes its simulation by the Direct Simulation Monte Carlo method (DSMC) the method of choice. However, there is a lack of reliable data on collisions, particularly reaction and collision cross sections needed for DSMC, of Ionian species at the conditions seen in its atmosphere. As such, collisions between SO2 and O are studied through Molecular Dynamics, Quasi-Classical Trajectories using two different methods for determining molecular potential: a polynomial potential defined by Murrell  and a complex tunable potential defined by the ReaxFF force field. Five possible reaction paths are considered: atomization of the SO2 molecule, dissociation to SO, dissociation to O2, and the formation of SO 3, and an exchange reaction leading to SO and O2. Relative velocities and initial SO2 internal energies relevant to Io's atmospheric conditions are used. The results from each chemistry model are analyzed and compared to each other, in particular the reaction cross sections and equivalent variable hard sphere cross sections. In general, higher collision energies are found to cause atomization of the SO2. In addition, dissociation to SO is a significant process for many of the studied cases, but dissociation to O2 is mostly a negligible process. Finally, formation of SO3 occurs only at low impact velocities. The chemistry and collision models developed from the Murrell and ReaxFF methods are then implemented in DSMC 0-D, time-dependant and 2-D axisymmetric simulations under conditions relevant to the Ionian atmosphere. The results of these analyses are examined, and compared to results obtained using the baseline Total Collisional Energy (TCE) model and the previous results of Deng et al . It is found that the new Murrell and ReaxFF based models predict less SO2 dissociation than previous models, and that simulated Ionian atmospheric structure is sensitive to the total cross section model.