Computer Simulations of the Sputtering of Metallic, Organic, and Metal−Organic Surfaces with Bin and C60 Projectiles

This study focuses on the microscopic modeling of 0–25 keV Bi1–3–5 and C60 cluster impacts on three different targets (Au crystal, adsorbed Au nanoparticle, and organic solid), using molecular dynamics simulations, and on the comparison of the calculated quantities with recent experimental results, reported in the literature or obtained in our laboratory. The sputtering statistics are reported, showing nonlinearity of the sputtering yields with the number of cluster atoms at the same incident velocity for Bi1–5 bombardment. They are compared to experiments (especially for the organic target), and the microscopic explanation of the observations is analyzed. The results show that the respective behaviors and performances of the different projectiles are strongly dependent on the target, with clusters of heavy Bi atoms being more efficient at sputtering gold and, conversely, fullerene clusters inducing the largest sputtering yields of the organic material (mass matching). For organic targets, some important and novel conclusions of this work are the following: (i) The increase of the sputtering yield when going from Bi atoms to Bi clusters is insufficient to explain the much larger increase of characteristic ion yields, suggesting a projectile-dependent ionization probability. (ii) The extent of molecular fragmentation follows the order of Bi > Bi3 > Bi5 > C60, that is, softer emission with larger clusters. (iii) Even 5–10 keV Bi atoms create collective molecular motions and craters in the polymeric solid, though the collision cascades are rather dilute. Finally, a second series of simulations performed at low energies predict that 0.1–1 keV Bin clusters should not provide better results for sputtering and depth profiling than isoenergetic single atoms.