Optical and electrical optimization by advanced characterization of monolithic perovskite/silicon tandem solar cells

Hybrid metal-halide perovskite solar cells have experienced a tremendous development within a short time period. In 2009 the first perovskite single-junction solar cell with a power conversion efficiency (PCE) of 3.8% was presented. Just eleven years later, in year 2020, a PCE of 25.5% was achieved, which approaches the value of silicon single-junction solar cells. As metal-halide perovskite solar cells can be bandgap-optimized for top cells via compositional tuning, both types of solar cells can be combined into perovskite/silicon tandem solar cells. The synergy can enable higher efficiencies than the single cells alone, while the addition of the perovskite can potentially be done inexpensively, which makes this technology attractive for the photovoltaic market. The first monolithic perovskite/silicon tandem solar cell was presented in 2015 with a PCE of 13.7%. Just two years later, in 2017, when the work on this dissertation started, the PCE was already increased to 23.6%. This dissertation covers the PCE improvement of monolithic perovskite/silicon tandem solar cells to a world record level above 29% and makes an important contribution to the understanding of the principles of device operation and to the techniques used to analyze these cells. Despite the transition to an optically advantageous tandem solar cell design, the optical properties in the multi-layer stack with more than ten different films have a significant influence on the photogenerated current density (JPh) that can be generated in both subcells. Especially high reflection and parasitic absorption reduce the amount of light available for power conversion. In a first study covered by this dissertation, precise adjustments of thickness and deposition conditions of the perovskite layer, selective contact layers and transparent conductive oxide were presented. These optimizations led to an improvement from 25.0% to 26.0% due to reduced reflection and better matching of the photogenerated current densities of the perovskite and silicon subcell. The cumulated photogenerated current density of 39.5mAcm−2 for this tandem cell, which had a flat front side, is today still one of the highest values and even comparable to cells with a textured front side. Although the short-circuit current density (JSC) could be increased by 1.4mAcm−2, the fill factor (FF) decreased by ∼2 percentage points when the subcells operate closer to current matching conditions. Thus, we accounted this reduction to the reduced mismatch and analyzed this in more detail: We artificially induced various mismatch conditions by illuminating the tandem solar cell with different LED-based spectra. We found that the FF reached a minimum close to current matching conditions. If current mismatch between both subcells occurs in the tandem solar cell (i.e. either of the subcells generated a lower current density), the fill factor increased. This effect partially compensates the reduction of the JSC induced by current mismatched subcells and makes monolithic tandem solar cells less sensitive to small mismatch conditions. This is highly important for energy yield analysis and has to be incorporated into yield simulations for more precise analysis. We verified the effect of FF enhancement under current mismatch by simulating the tandem solar cell electrically. Although the simulated FF trend could be well reproduced well as a function of the mismatch, the absolute value of the FF was higher when compared to the measured tandem solar cell. This revealed that further investigation and advanced characterization methods are required to reconstruct the tandem solar cell properly, understand the individual subcell characteristics in more detail and to improve the cell to reach the values obtained by electrical simulations. The second study is linked to the FF losses and addresses the optimization of the FF and open-circuit voltage (VOC) to further improve the PCE further. This was enabled by a self-assembled monolayer (SAM) as a hole-selective contact in the perovskite top cell, which was for the first time utilized in these tandem solar cells. Besides commercially available SAM molecules, which were shown to perform better in single-junction solar cells than the typically used polymer PTAA (poly[bis(4-phenyl)(2,5,6 trimentlyphenyl) amine]), we introduced the molecule Me-4PACz ([4-(3,6-dimethyl-9H-carbazol-9-yl) butyl]phosphonic acid), a new SAM for perovskitebased tandem and single-junction solar cells. In addition to the well-passivated perovskite interface, it enabled a fast hole extraction, which led to high VOC and FF values and an improved photostability for perovskite compositions with high bromide loading to enable 1.68 eV bandgaps, which typically suffer from halide segregation. In tandem solar cells we found similar benefits. All SAMs enabled higher PCE values compared to cells using PTAA as hole-selective contact. Furthermore, Me-4PACz led to high VOC and FF values of up to 1.92V and 81%, respectively. Together with a high JSC, which was optimized in the previous study, a certified world record PCE of 29.15% was enabled. To evaluate the long-term stability of tandem solar cells, we specifically designed and fabricated a light source consisting of two types of LED arrays (193 LEDs in total), enabling precise control to match subcell photocurrent generation with AM1.5g conditions over long periods. Non-encapsulated tandem solar cells were measured continuously for 300 hours in air with relative humidity between 30-40%. The tandem solar cell with Me-4PACz showed highest stability, retaining 95.5% of its initial PCE. As evident from the first study, detailed analysis of the subcell is desired. However, typically it is difficult to evaluate the performance of the individual subcells in series connected tandem solar cells. We used injection-dependent absolute electroluminescence measurements to access the performance of the individual subcells. With this method, we were able to reconstruct the subcell current density-voltage characteristics without the influence of series resistances. From these reconstructions the photovoltaic parameters such as the maximum power point or pseudo fill factor could be determined. Furthermore, the parametrized reconstructed current density-voltage curves were used to simulate the tandem solar cell electrically, which revealed that with this tandem structure a PCE of more than 32% is achievable if series resistance losses can be minimized. The rapid increase in tandem efficiency also increases the need for fast technology transfer into industry. All previously reported high-efficiency tandem solar cells have in common that they are based on >250 μm thick floatzone (FZ) silicon with most of them having a polished front side to enable efficient solution processing of the thin top cell. However, for mass production and thus for economic reasons, it is crucial to use thinner czochralski (CZ) grown wafers, ideally without the need for polishing. Therefore, in the third study, we fabricated tandem solar cells on 100 μm thin CZ silicon based bottom cells with a rough surface and compared the performance to tandem cells fabricated on laboratory-typical bottom cells accompanied by a 1000 hour MPP-track and optical simulations. We found that the median performance of both types was with 27.8% equally high with maximum values of 27.89% and 28.15% for CZ- and FZ-based tandem cells, respectively. However, the individual photovoltaic performance parameters differed: The thinner bottom cell for CZ-based devices enabled higher VOC values due to reduced recombination current density. Furthermore, the photogenerated current density was reduced in the case of thinner silicon subcells, which led to an increased current mismatch. As investigated in the first study, this results in higher FF values of the CZ-based devices. Optical simulations revealed that the reduction of the bottom cell thickness from 280 μm to 100 μm enables to widen the top cell bandgap by ∼0.02 eV and hence further increase the PCE, if current matching conditions should be maintained. Interestingly, this widening is independent of the perovskite’s thickness. On the one hand the widened bandgap could lead to a higher VOC, on the other hand wider bandgaps typically suffer from reduced stability and strong interface recombination. We found the same results for double-side textured tandem solar cells with conformally deposited perovskite and the findings regarding the optimum top-cell bandgap for industrial perovskite/silicon tandem solar cells are highly important for the future development of these fascinating tandem solar cells. This dissertation plays an important role in the development of perovskite/silicon tandem solar cells and hence, will hopefully contribute to the expansion of photovoltaics.

Author dc:creator

• Köhnen, Eike

Advisor dc:contributor.advisor

• Albrecht, Steve

Licence dc:rights.uri

• http://rightsstatements.org/vocab/InC/1.0/

Language dc:language.iso

en

Identifier URI

http://dx.doi.org/10.14279/depositonce-12324

OAI identifier oai:identifier

oai:depositonce.tu-berlin.de:11303/13541