This tutorial is on the use of kinetic simulation methods to describe low-pressure gas flows and plasma processes. These methods include the Direct Simulation Monte Carlo (DSMC) – method for rarefied gas flows as well as the Particle-in-Cell Monte-Carlo (PIC-MC) – method for low-density, non-equilibrium plasmas. For gas flow simulation, the different flow regimes – i.e. molecular flow, continuous flow and transition flow – and their implications on the modelling method are discussed. Shown practical examples include flow conductance determination for vacuum components, evaporation and deposition profile in PVD processes as a function of total pressure. Additionally, the meaning of non-local effects at very low pressure as well as collective effects on the transition between molecular and continuous flow are illustrated.

Reactions between gas phase and surfaces can have significance influence on the global process dynamics. This is demonstrated in a model for reactive sputtering, where the surfaces can be either oxidized or metallic. A 2D model with spatially resolved, dynamic surface coverages illustrates the hysteresis effect in reactive magnetron sputtering.

In the field of plasma simulation, the role of the magnetic confinement in magnetron sputtering is shown together with dynamic plasma features such as rotating spokes, that can be observed in simulation and experiments. Another topic will be the impact of pulse sputtering on the plasma potential and ion energy distribution function at the substrate.

Kinetic simulation of processes in deposition reactors yield the detailed growth conditions at the substrate consisting of the energy and angular distribution functions of ions and neutrals. This tutorial shows how kinetic simulation methods can be connected with atomic film growth simulation models in order to predict not only process dynamics and deposition uniformity but also intrinsic film properties.

Finally – based on the angularly resolved particle flux near the substrate – a fast, ray-tracing based algorithm is shown for prediction of the film thickness profile on curved substrates as a function of uniformity masks and substrate movement trajectory. Such an algorithm can be tuned to certain deposition process conditions and be used to realize a specified film thickness profile by iterative optimization.