PhD Defense of Baptiste MARTEAU

The photovoltaic module market is dominated by technologies based on crystalline silicon (c-Si). The use of low temperature (SHJ) or high temperature (TOPCon) passivated contacts leads to record efficiencies (26.8% and 26.2%) close to the theoretical limit of 29.4%. The option explored by the majority of institutes to overcome this limit is to combine c-Si technology with another wide bandgap (EGap) semiconductor material to enable optimum conversion of the solar spectrum over the entire energy range. The theoretical maximum efficiency of such tandem devices can then reach 42%. A two-terminal structure enables easiest module processing leading to reduced production costs. However, this places severe constraints on the interface layers between the two cells. These must provide excellent carrier lifetime in each cell, while ensuring optimal optical (minimal parasitic absorption and reflection) and electrical (efficient and highly conductive recombination junction RJ) properties.
For the c-Si cell, this thesis focuses on TOPCon technology, which is expected to become market mainstream by 2030. This approach, based on poly-Si/SiOx stacks, offers great versatility for the tandem device fabrication processes (stability up to 800°C), and benefits from highly doped layers that are well suited for the formation of RJ. Among the variety of large EGap materials, perovskite (Pk) technology is the most popular solution as it benefits from both high efficiency potential and low production costs. The interface between the two cells (TOPCon and Pk) of the tandem device is usually formed by transparent conductive oxides layers such as ITO (Indium Tin Oxide), which shows excellent electrical and optical properties. However, indium is a critical material that could limit the long-term development of this technology. Therefore, the aim of this thesis is to explore indium-free approaches for the interface of Pk/c-Si tandem cells.
The studies carried out in this work concern Pk/c-Si tandem cells in nip configuration, for which two alternative approaches for interface engineering are investigated. The first one uses no additional interface layer, while the second one integrates an nc-Si (n+) layer to form a silicon tunnel diode, which should provide an optimal recombination current. These two alternative approaches allowed better initial performances than the reference process, mainly by overcoming short-circuit issues in the Pk cell. Tandem devices featuring no additional interface layer show fill factors comparable to those of the world's best devices (>81%) and efficiencies close to 25%, confirming the potential of TOPCon passivated contacts to form indium-free RJ. However, these two indium-free approaches were limited by the appearance of internal series resistance over time. Advanced characterisations explain these degradations by the formation of a SiOx layer between silicon and SnO2 (the electron-selective layer – ESL– of the Pk cell).
In conclusion, TOPCon passivated contacts are particularly well suited to obtain efficient recombination junctions (direct or via silicon tunnel diodes), thus eliminating the need to use indium in the interface layers. As silicon is particularly sensitive to oxidation, the choice of contacting layers (ESL in nip configuration) should be focused on a material that contains no oxygen or has a stronger affinity for oxygen than silicon.
 

• Anne KAMINSKI-CACHOPO, UNIVERSITY PROFESSOR - Grenoble INP : Supervisor
• Pere ROCA I CABARROCAS, RESEARCH DIRECTOR - CNRS DELEGATION ILE-DE-FRANCE SUD : Reviewer
• Marie GUEUNIER-FARRET, ASSOCIATE PROFESSOR  HDR - Bordeaux University: Reviewer
• Stéphane COLLIN, RESEARCH DIRECTOR - CNRS DELEGATION ILE-DE-FRANCE SUD : Examiner
• Daniel BELLET,  UNIVERSITY PROFESSOR - Grenoble INP : Examiner