Adam Obrusnik¹, Gayatri Rane², Martin Kubecka¹, Philipp Dürrenfeld², Kristina Tomankova¹, Uwe Krause2, Denis Shaw^2
1 PlasmaSolve s.r.o., Brno, Czech Republic
² Advanced Energy Industries, Customer Solutions Lab, Karlstein am Main, Germany
Physical vapor deposition (PVD) techniques, particularly in large area coaters such as glass coaters and roll-to-roll (R2R) systems, depend significantly on the control and understanding of plasma dynamics for optimal film deposition. Understanding the response of the PVD plasma is crucial, considering the low tolerances in film thickness and film properties variations, that the large-area coating industry aims for. This study employs 2D physics-based plasma simulations to investigate the impact of pulse shape and anode placement on plasma behavior and film properties in these coaters.
Different pulse shapes, including square-wave, and dynamic reverse pulsing, are analyzed for their effects on plasma characteristics and deposition outcomes. We demonstrate that the model correctly captures the trends in the deposition rate as well as substrate heating, which serves as a solid basis for future correlations of model results with film properties. Ultimately, we identify some qualitative differences between the dynamic reverse pulsing and square-wave mode and each of them clearly has its benefits, which are process-dependent and context-dependent.
Moreover, the placement of anodes within the coater is examined for its potential to modify plasma distribution and energy input, aiming to improve film uniformity and quality. It is demonstrated, both on the modelling side and experimentally, that anodes are indeed a powerful driver in shaping the plasma distribution and even enhancing the deposition rate.
This comparative study enhances the ability to control film properties through precise manipulation of process variables, offering insights that are crucial for advancing PVD technology in industrial applications. It is also discussed how the detailed plasma simulation interfaces with other types of models, such as the global model predicting the hysteresis curve of the system or a 3D model for optimizing the large-scale uniformity.