Kristína Tomanková_1,2, Kryštof Mrózek_1,2, Adam Obrusník₁, Alexander Fromm₃, Frank Burmeister_3
1PlasmaSolve s.r.o., Brno, Czech Republic
₂Masaryk University, Brno, Czech Republic
₃Fraunhofer Institute for Mechanics of Materials IWM, Freiburg, Germany
High Power Impulse Magnetron Sputtering (HiPIMS) presents a game-changing technology in the coating industry, creating high quality thin films with high density. As such, it is imperative to make the process efficient, especially on a large scale. A valuable tool in optimizing such complex processes are physics-based numerical simulations due to cost-effectiveness and conservation of valuable materials and energy that would otherwise be needed for a number of experiments under different conditions.
This contribution presents a time-dependent volume-averaged global plasma model of a reactive HiPIMS process that can predict the general behavior of the process and its macroscopic response to variations in process parameters. The model mimics an industrial HiPIMS coater used for deposition of Al₂O₃ and SiO₂ thin films.
Even though a few highly sophisticated models of the HiPIMS process have been developed, works aiming for a detailed model description of HiPIMS for practical and industrial applications are rare in literature. The developed models typically require many non-orthogonal tuning parameters and lack experimental validation.
As opposed to the already published global models, our model was built ab initio while closely investigating the influence and interactions of key physical phenomena characterizing HiPIMS. Utilizing the exceptional computational efficiency of global models, it is possible to clearly demonstrate the importance of various physical processes in a short amount of time. The sensitivity analysis includes effects such as heating of the target, sputter wind or target poisoning whose influence on the process is quantitatively outlined. In this context, our research involves a “stepping back” from already published numerical models that include dozens of source terms, and it is not clear whether all these source terms hold equal importance. The contribution will also describe concrete steps that we are undertaking towards the development of a spatial model of an industrial HiPIMS process and preliminary simulation results, if available.