I. Introduction
Recent years’ improvements in silicon crystal quality and processing schemes have greatly increased the importance of surface passivation layers. In order to profit from high material quality, minimizing the surface recombination has become a very crucial factor. Furthermore, extensive investigations of defects often require surface passivation to be robust to harsh testing conditions while retaining their low recombination activity. The same holds true for next-generation solar cell concepts which require high surface passivation quality being long-term stable under in-field conditions. Aluminum oxide layers appear to meet these requirements for both scientific experiments and large-scale industrial application, exhibiting very low surface recombination velocities (SRV). Stacks of amorphous aluminum oxide layers and amorphous silicon nitride (a-SiNx) or silicon oxide (a-SiO x) are established passivation schemes for rear side passivation in industrially applied solar cell designs such as PERC (passivated emitter and rear cell), e.g., [2]. Many studies have investigated the surface passivation quality of aluminum oxide layers and stack systems on freshly processed samples and demonstrated their quality, e.g., [3] and references therein. Other studies have focused on the resilience of aluminum oxide passivation layers under illumination at room temperature (e.g., [4]), storage in the dark (e.g., [5]), or damp-heat conditions (e.g., [6] ). However, there are less data available for their long-term stability under in-field conditions—i.e., illumination at elevated temperature. Besides their obvious importance for solar cell device operation, such conditions have recently become more common in material studies—e.g., due to the observation of bulk lifetime degradation in current multicrystalline silicon addressed to as LeTID (light and elevated temperature induced degradation) [2] or stabilization of BO defects. The importance of such studies is made evident by the significant performance loss encountered in PERC solar cells and modules in the field [2], [7], [8] which had not been foreseen by stability testing under standard conditions (room temperature). Also, studies by Sperber et al. have demonstrated significant influence of such conditions on the passivation quality [9]– [11]. However, the focus of these studies was on a-SiNx passivation layers and only p-type samples were investigated.