Josja Van Bever, Koen Strijckmans, Diederik Depla, Ghent University, Ghent, Belgium
High-quality coatings with optimized stoichiometry in reactive magnetron sputtering are critical for various industrial applications. Achieving this requires maintaining process conditions within the transition region between metallic and poisoned modes, either through feedback control or high pumping speeds.Yet, the comprehension of the convergence and steady-state conditions of feedback control in this context remains limited.
In our study, we present precise feedback measurements focused on the aluminum-oxygen system. Our investigations reveal the existence of two distinct metastable states, contingent on the system's history. We establish the connection between these states and the double hysteresis phenomenon that is derived from the Reactive Sputtering Deposition (RSD) model on the one hand and prior research relying on IV-characteristic data analysis on the other hand. This linkage is achieved through variations in discharge current density and process parameters closely linked to target poisoning mechanisms.
We delve into the long-term stability of the double hysteresis in feedback control and compare it to stable transition conditions achieved with high-pumping speeds. Our discussion encompasses various factors affecting the long-term stability, including target erosion effects, chamber wall gettering, anode effects, and fluctuations induced by the chosen process control.
We also explore different feedback convergence strategies. Our findings illuminate the path towards an optimal convergence approach, accounting for the stability of each set of transition conditions. This understanding provides a valuable guideline for industry professionals seeking to employ feedback control in a reproducible and time-efficient manner.
This research advances our knowledge of reactive magnetron sputtering, offering insights into the critical interplay between feedback control, metastable conditions, and long-term stability. It promises to enhance the precision and reliability of thin film deposition processes, with implications for a wide range of technological applications.