Abstract :
[en] This thesis explores the physics of point defects, conducting a comprehensive investigation through both theoretical and experimental approaches. In a first part, we emphasis the rich behaviour of defects, leading to their detailed classification. Non-radiative defect-mediated recombinations are described based on the Shockley-Read-Hall (SRH) statistics and detailed by explaining the microscopic ori- gin of charge carrier capture coefficients. This discussion unravels the effects of lattice distortions caused by defect incorporation or changes in charge states. Furthermore, in solid-state matter, the rich physics introduced by these lattice imperfections can either enhance or degrade material properties and performances within the desired application. The results presented in this thesis highlight this ambivalent feature of point defects in absorber materials for photovoltaic (PV) applications.
On that basis, we aim to deepen the scientific community’s understanding of defects in earth-abundant photovoltaic materials: Copper oxides Cu2O and kesterites Cu2ZnSnS4. As part of the PV thin film technology, both are Cu-based p-type semiconductors featuring high absorption coefficients and including chalcogen elements (S, Se or O) in their crystalline structures. As such, they are used as absorber layers in solar cell architectures. However, their solar cell efficiencies are limited by low open circuit voltage VOC values due to high charge carriers recombination rates, primarily attributed to defects and crystalline imperfections within the synthesised layers. Despite these similarities, their differing bandgap values de- termine their distinct applications. Copper oxide, as a transparent conducting oxide (TCO), is more suitable for transparent PV, whereas sulphur-based kesterites with their 1.5 eV bandgaps, can be efficiently used in single or tandem approaches with either perovskite or traditional bulk-Si. Furthermore, the complex quaternary chemical nature of kesterites significantly increases the number of possible point defects and secondary phases compared to simpler binary copper oxide materials.
In a first research work, using a first-principles approach, we highlight the im- pact of Sn cationic substitution on the structural, electrical and optical properties of kesterite compounds. By sequentially replacing Sn with two isoelectronic elements, Ge and Si, we report an increase in the kesterite bandgaps and high absorption ab- sorption coefficients of the order of 104 cm−1. Then, ab initio optical results are used as data input to model the electrical power conversion efficiency of kesterite-based solar cells thanks to an improved version of the Shockley-Queisser model. The variation of the solar cell maximum efficiency is studied as a function of the non-radiative recombination rate. We emphasise the suitability of Cu2ZnSnS4 in single or tandem (top cell) solar cells, with a potential efficiency improvement of nearly 10% compared to state-of-the-art values. In addition, Cu2ZnGeS4 appears as an interesting candi- date for top cell absorber layers in tandem approaches, with room for an efficiency improvement of 5 points.
To pursue, based on the supercell approach, the second phase of research in- volves confirming and predicting possible recombination centres in both Cu2ZnSnS4 and Cu2ZnGeS4, respectively. This study is motivated by the previously underlined room for improvement concerning the efficiency of Sn,Ge-based kesterite solar cells. Using a first-principles approach, we investigate the physical behaviour of point defects both upon Ge doping and alloying of Cu2ZnSnS4. The p-type conductivity of both Cu2ZnSnS4 and Cu2ZnGeS4, attributed to VCu and CuZn acceptor defects, is established. We also confirm the detrimental role of the substitutional defects XZn (X=Sn,Ge) acting as recombination centres. However, we emphasise that, in contrast to Sn, the substitution of Zn by Ge results in a defect that is less likely to facilitate pure non-radiative recombinations. This observation could be one potential source of VOC improvement reported in the literature upon Ge incorporation.
Finally, the last research work aims at the experimental investigation of N and Mg doping in Cu2O thin films deposited using RF magnetron sputtering at room temperature. Using a wide range of characterisation techniques (XRD, EDX, VdP-Hall, spectrophotometry, Raman), we correlate the variations of the thin film optoelectrical properties to both crystalline transitions and point defect formations. Starting from a CuO crystalline phase, we unveil a possible synthesis mechanism for a N- doped Cu2O layers. Then, upon N-doping, we report an increased concentration of (N2)Cu shallow acceptor defects explaining the probed enhancement of p-type majority charge carriers observed. Consistent with the established literature, we also confirm the improvement of the sample optoelectrical properties as Mg is intro- duced within split copper vacancy sites VCu,split. Conversely, we demonstrate that co-doping with Mg and N degrades the material crystallinity, leading to a reduction in thin film conductivity, likely due to high nitrogen incorporation.
We believe our results clarified the fundamental mechanisms that operate at the atomic scale via the formation of point defects in both kesterites and copper oxides.
