PhD Defense of Chloé WULLES

Today's climate challenges encourage the development of more sustainable energy production methods. Among these, the photovoltaic (PV) effect converts the energy offered by solar radiation into electrical power. As a result, research in this field is constantly expanding in a variety of directions. It may involve exploring the potential of new materials and complex structures, or developing ways of improving the performance of more conventional structures. In this context, a better understanding of degradation mechanisms in PV components is essential.
Photovoltaic components are subject to a wide range of losses at various stages of the energy conversion process. These include optical losses, thermalization, resistive effects and recombination of photogenerated carriers. In particular, the presence of defects in PV components encourages recombination processes assisted by electronic trap levels. To limit the influence of these various sources of loss and increase device performance, technologists are introducing innovations. Sometimes, however, this also leads to an increase in the complexity of component physics, not well described by conventional models. Standard characterization methods are sometimes insufficient to interpret the current conduction mechanisms involved in these components optimized for energy production. For these reasons, characterization tools need to be developed to detect the presence of defects and to provide information about the conduction mechanisms operating in PV components.
Low-frequency electronic noise analysis is a non-destructive characterization method that can meet both these needs. It is widely used in the field of microelectronics, in particular because it enables the precise study of generation-recombination or electronic trapping phenomena, which are indicative of the presence of defects. In comparison, it is relatively little used in photovoltaics. This can be explained, among other things, by the highly complex physics of these components, making them less accessible for such studies. However, a number of studies suggest that it could provide unprecedented insights into the operation of PV structures and the presence of defects in them. Thus, it seems important to develop understanding around electronic noise in photovoltaic components, the primary intention of this PhD.
In the course of this work, studies were carried out in various types of PV components developed in specialized laboratories such as CEA-Liten INES, but also in structures optimized for noise measurement designed at IMEP-LaHC. The work carried out on these various components has enabled to identify the signatures of repeated noise mechanisms in PV components and diodes in general. An in-depth study of the modeling of joint voltage and current dependencies of noise has also provided decisive information for the identification of non-elementary conduction mechanisms occurring at the edges of cut Al-BSF cells. Various types of noise were observed and studied during this thesis, mainly 1/f noise, but also white noise (thermal and shot), or even telegraphic noise in the case of heterojunction cells equipped with a thin passivating tunnel oxide layer. All these results have reinforced the relevance of developing the use of electronic noise to characterize PV components.
 

•  Anne KAMINSKI-CACHOPO Grenoble INP : Supervisor
• Jean-Guy TARTARIN Université de Toulouse III : Reviewer
• Jean-Paul KLEIDER CNRS : Reviewer
• Edwige BANO Grenoble INP : Examiner
• Yvan CUMINAL Université de Montpellier : Examiner
• Quentin RAFHAY Grenoble INP UGA :  Guest
• Christoforos THEODOROU CNRS : Guest