We have investigated via ab initio-derived molecular dynamics simulations the influence of up to 4.0 eV of translational excitation on the reactivity of F-2 molecules with the Si(100) surface. We propose a stepwise reaction mechanism which yields two pathways involving Si-F bond formation : F atom abstraction, where one Si-F bond is formed at the expense of the F-F bond while the remaining F atom is ejected from the surface, and dissociative chemisorption, where both F atoms in the incident F-2 molecule form Si-F bonds in a consecutive fashion, once again at the expense of the F-F bond. For the clean surface, we find abstraction to be highly probable at low incident energies, with increasing translational kinetic energy in impinging F-2 molecules leading to increased dissociative chemisorption at the expense of atom abstraction and thus higher reactivity. We do not observe nonreactive scattering for the clean surface, where the incident F-2 molecule reflects from the surface without forming an Si-F bond, but this changes with the introduction of preadsorbed fluorine on the surface. We observe decreasing reactivity with increasing surface fluorine coverage, eventually leading to minimal reactivity for surfaces with 1 ML of fluorine coverage. The reactivity can be increased both by increasing F-2 translational kinetic energy and the introduction of disorder to the silicon surface. We also characterize the products which are ejected from the surface during our simulations. We find that F atoms generated during abstraction have hyperthermal energies which are not characteristic of the surface temperature and exhibit minimal memory of the F-2 incident energy. F-2 molecules which undergo nonreactive scattering do demonstrate memory of their initial conditions, but their exit energy is reduced relative to their incident energy due to the inelastic nature of the scattering. Both types of ejected species have exit angular distributions which are indicative of their nonequilibrated natures. On the basis of our simulations,we predict that translationally activated F-2 molecular beams may be effective for applications involving neutral beam etching of silicon surfaces.