Atom deposition and sputtering at normal incidence simulated by the Frenkel-Kontorova chain

The impact of a molecule of N atoms with a speed of v_{0} on the free end of the Frenkel-Kontorova chain is numerically simulated. Depending on the values of N and v_{0}, different scenarios of the molecule-chain interaction are observed. Molecules with low speed stick to the chain. At somewhat higher speeds, the molecules bounce off the chain. Further increase in v_{0} results in bouncing off a molecule larger than the incident one. At even higher speed, bouncing of the molecule off the chain takes place simultaneously with the formation of a supersonic crowdion (antikink) propagating along the chain. A very high collision velocity leads to the sputtering of atoms from the chain and the formation of single and multiple supersonic crowdions. Interestingly, the sputtering yield Y as the function of v_{0} demonstrates a nonmonotonous dependence. This is explained by the fact that supersonic crowdions can have a discrete set of propagation velocities. When v_{0} is such that supersonic crowdions are effectively excited, the latter transfer energy deep into the chain, and the sputtering is minimal. For some v_{0} ranges, the formation of supersonic crowdions is suppressed. In these cases, the energy transferred from the impact of the molecule to the chain is spent mainly on the sputtering of atoms. The results obtained qualitatively explain the physics of bombardment of a crystal surface by atomic clusters with applications in physical vapor deposition, ion implantation, ion-beam sputtering, and similar experimental techniques.

Large cluster ions: soft local probes and tools for organic and bio surfaces

Ionised cluster beams have been produced and employed for thin film deposition and surface processing for half a century. In the last two decades, kiloelectronvolt cluster ions have also proved to be outstanding for surface characterisation by secondary ion mass spectrometry (SIMS), because their sputter and ion yields are enhanced in a non-linear fashion with respect to monoatomic projectiles, with a resulting step change of sensitivity for analysis and imaging. In particular, large gas cluster ion beams, or GCIB, have now become a reference in organic surface and thin film analysis using SIMS and X-ray photoelectron spectroscopy (XPS). The reason is that they induce soft molecular desorption and offer the opportunity to conduct damageless depth-profiling and 3D molecular imaging of the most sensitive organic electronics and biological samples, with a nanoscale depth resolution. In line with these recent developments, the present review focuses on rather weakly-bound, light-element cluster ions, such as noble or other gas clusters, and water or alcohol nanodroplets (excluding clusters made of metals, inorganic salts or ionic liquids) and their interaction with surfaces (essentially, but not exclusively, organic). The scope of this article encompasses three aspects. The first one is the fundamentals of large cluster impacts with surfaces, using the wealth of information provided by molecular dynamics simulations and experimental observations. The second focus is on recent applications of large cluster ion beams in surface characterisation, including mass spectrometric analysis and 2D localisation of large molecules, molecular depth-profiling and 3D molecular imaging. Finally, the perspective explores cutting edge developments, involving (i) new types of clusters with a chemistry designed to enhance performance for mass spectrometry imaging, (ii) the use of cluster fragment ion backscattering to locally retrieve physical surface properties and (iii) the fabrication of new biosurface and thin film architectures, where large cluster ion beams are used as tools to transfer biomolecules in vacuo from a target reservoir to any collector substrate.

On Universality in Sputtering Yields Due to Cluster Bombardment

Molecular dynamics simulations, in which atomic and molecular solids are bombarded by Arn (n = 60-2953) clusters, are used to explain the physics that underlie the "universal relation" of the sputtering yield Y per cluster atom versus incident energy E per cluster atom (Y/n vs E/n). We show that a better representation to unify the results is Y/(E/U0) versus (E/U0)/n, where U0 is the sample cohesive energy per atom or molecular equivalent, and the yield Y is given in the units of atoms or molecular equivalents for atomistic and molecular solids, respectively. In addition, we identified a synergistic cluster effect. Specifically, for a given (E/U0)/n value, larger clusters produce larger yields than the yields that are only proportional to the cluster size n or equivalently to the scaled energy E/U0. This synergistic effect can be described in the high (E/U0)/n regime as scaling of Y with (E/U0)(α), where α > 1.