Tailored lead iodide growth for efficient flexible perovskite solar cells and thin-film tandem devices

Flexible perovskite solar cells (PSCs) hold great promise for the low-cost roll-to-roll production of lightweight single- and multijunction photovoltaic devices. Among the different deposition methods used for the perovskite absorber, the two-step hybrid vacuum-solution approach enables precise control over the thickness and morphology of PbI2. However, efficient conversion to perovskite is limited by diffusion of the organic cations in the compact lead halide layer. Herein, a multistage absorber deposition is developed by thermal evaporation of PbI2 and spin coating of CH3NH3I (MAI). The process relies on the different types of growth of vacuum-deposited PbI2 onto amorphous and crystalline surfaces. This approach represents a way to effectively increase the absorber thickness while tackling the limited MAI diffusion in the compact PbI2 film via a two-step deposition method. The efficiency of flexible PSCs is improved from 14.2 to 15.8% with multistage deposition. Furthermore, the use of an amorphous transparent conductive oxide (TCO), InZnO, enhances the mechanical resistance against bending with respect to conventional crystalline TCO-based flexible devices. Near-infrared transparent flexible PSCs are developed with an efficiency of 14.0% and average transmittance of similar to 74% between 800 and 1000 nm. Flexible perovskite/CIGS thin-film tandem devices are demonstrated with an efficiency of 19.6% measured in the four-terminal configuration.

Development of Highly Efficient Perovskite-on-Silicon Tandem Solar Cells

Florent Sahli

Crystalline Silicon (c-Si) solar cells are dominating the photovoltaic (PV) market. Owing to their large manufacturing capacity, reliability and efficiency, c-Si solar cells are now cost-competitive with other non-renewable electricity sources in many places. The price of c-Si modules has been decreasing drastically for the past decades. They now account for less than half of the cost of a PV system. Other costs come from the balance-of-system and these are rather inflexible. And so increasing efficiency of solar modules is the most efficient approach to lower the cost of PV electricity. One issue is that c-Si cells are approaching their efficiency limit. One strategy to increase efficiency beyond this limit relies on adding another solar cell on c-Si to form a tandem solar cell as thermalisation losses are reduced. Metal lead halide perovskite solar cells are promising top cell candidates for c-Si based on their high optoelectronic properties, band gap tunability, high efficiencies, ease of manufacturing and low material costs. By combining both technologies, efficiencies >30% are realistic, well above best-in-class c-Si cells. This thesis aims to produce high efficiency perovskite solar cells for 2-terminal monolithic tandems on c-Si. First, we develop a perovskite fabrication method, which employs thermal evaporation to produce a lead halide template and then organohalide spin-coating. By varying composition at each step, the perovskite band gap can be tuned in the range 1.6-1.8 eV, ideal values for 2-terminal tandems. The use of the template makes the deposition compatible with various substrate textures. Then, we develop a recombination junction that features nanocrystalline hydrogenated silicon layers (nc-Si:H). When used with a font-side polished c-Si bottom cell, it shows a superior optical performance compared to standard transparent conductive oxides thanks to a better matching of refractive indices. Owing to its low conductivity, the top cell leakage current is reduced, enabling to scale-up the cell area. Then, perovskite/c-Si tandems featuring a double-side textured c-Si wafer are demonstrated, achieving a certified efficiency of 25.2%. This efficiency, the highest at the time of publication, is enabled by low reflection losses and efficient light trapping thanks to the c-Si pyramids present on both sides. More importantly, this double-side texture yields an optically close-to-optimum system that is simpler and more efficient compared to alternatives. Furthermore, this top cell process flow does not require any modification to existing c-Si manufacturing production lines as these use front-side textured c-Si. In the second part, we replace the spin-coating of organohalides by a vapour transport deposition (VTD) using a home-made vapour transport deposition setup that offers large processing flexibility. Using a thermally evaporated lead iodide template and methylammonium iodide vapours, perovskite solar cells with an efficiency > 12% are made. Thanks to the presence of a showerhead, a homogeneous perovskite growth is obtained on 6 inch textured c-Si substrates, the industry standard. The VTD of formaminidium iodide (FAI) is more challenging. A trimerisation of FAI forms sym-triazine, which does not react with the template to form the perovskite. Still, in the presence of ammonium cations, sym-triazine can be cleaved to form formamidine, thus offering an alternative pathway to deposit perovskite layers.

EPFL2020

Development of high band gap protocrystalline material for p-i-n thin film silicon solar cells

Fabio Maurizio

High band gap thin film silicon solar cells are of high interest for the use as top cell in triple junctions solar cells. Protocrystalline silicon, in the transition phase between amorphous and microcrystalline silicon, is such a high band gap material. During this project, protocrystalline intrinsic layers have been developed and optimized in p-i-n single junction solar cells by plasma-enhanced chemical vapour deposition (PECVD) in a new multi-chamber deposition system. In the first series, cells and layers were deposited with different dilutions going from silane to hydrogen flux ratios of 1:1 to 1:64. At the dilutions 1:16 and 1:32, protocrystalline ma-terial was obtained which showed amorphous characteristics only up to a certain thickness, where crystallites started to evolve. With dilution 1:16, a second cell series has been deposited with variation of the intrinsic layer thick-ness. The best cells in terms of open-circuit voltage and fill factor product were obtained with an i-layer thickness of 100 nm. These cells have open-circuit voltages and fill factors up to 0.975 V and 75.8 % (0.932 V and 69.9 %) for the initial (degraded) state. These are promising values for the use of this material in top cells. However, the low current density means that the optimum i-layer thickness is probably higher for triple cell application. Further cells have been deposited at different temperatures. These results imply that there is still room for further optimization, when current limitations of the heating system will be overcome.

2012

Fundamentals of perovskite formation for photovoltaics

Amita Ummadisingu

Perovskite solar cells have become strong contenders in the arena of photovoltaics due to the stunning rise in their efficiency from 3 % to over 20 % in just seven years. In this time, numerous device architectures and thin film deposition methods have been explored. The sequential deposition and anti-solvent methods are among the most widely used for preparing perovskite solar cells. Optimizing these perovskite deposition processes to tune the perovskite film morphology plays a key role in the race for the highest efficiencies as homogeneity of the film correlates with superior photovoltaic performance. To date, the factors controlling perovskite film formation in various deposition methods and the precise mechanism are little understood. In light of this, the aim of the thesis is to unravel the salient aspects of perovskite film formation. I first study the formation of methylammonium lead iodide in sequential deposition using confocal laser scanning fluorescence microscopy (CLSM) and scanning electron microscopy (SEM). I discover that illumination during film formation is a major factor in the reaction as it greatly accelerates perovskite formation and tunes the final film morphology by controlling nucleation of lead iodide. The nucleation density increases logarithmically with the illumination intensity. I uncover the mechanism behind this effect and demonstrate that it is a quantum phenomenon and not merely a heating effect. I posit that absorbed light lowers the surface tension, thereby facilitating nucleation. Not only is this effect present in the main deposition methods â sequential deposition and the anti-solvent methods, but also in perovskites of more complex compositions that are employed in state-of-the-art solar cells. Second, I scrutinize the individual stages in the formation of methylammonium lead iodide perovskite in sequential deposition. SEM-cathodoluminescence imaging identifies the presence of mixed crystalline aggregates composed of methylammonium lead iodide perovskite and lead iodide in forming perovskite films. Using cross-sectional CLSM for the first time, I identify the directionality of perovskite formation. Kinetic monitoring and model fitting bring forward the Avrami models as the most suitable to quantify and represent the formation of perovskite under different conditions of temperature, film thickness and illumination. Finally, I identify a maturation effect in the perovskite film in solar cells stored in the dark and investigate the dynamic spontaneous coalescence of crystals in the film. During maturation, small crystals merge to form larger ones and this reduces the number of grain boundaries and the associated trap states, supressing non-radiative recombination. The photovoltaic performance of high efficiency solar cells thus increases and it is accompanied by a reduction in the hysteresis. Overall, apart from providing insights into the dynamic nature of deposited films, I present a fundamental understanding of the perovskite formation process in different deposition methods and for various compositions. Such a comprehensive picture aids in the achievement of high photovoltaic performance.