@phdthesis{Caprioglio2020,
  author    = {Caprioglio, Pietro},
  title     = {Non-radiative recombination losses in perovskite solar cells},
  doi       = {10.25932/publishup-47763},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-477630},
  school      = {Universit{\"a}t Potsdam},
  pages     = {vi, 242},
  year      = {2020},
  abstract  = {In the last decade the photovoltaic research has been preponderantly overturned by the arrival of metal halide perovskites. The introduction of this class of materials in the academic research for renewable energy literally shifted the focus of a large number of research groups and institutions. The attractiveness of halide perovskites lays particularly on their skyrocketing efficiencies and relatively simple and cheap fabrication methods. Specifically, the latter allowed for a quick development of this research in many universities and institutes around the world at the same time. The outcome has been a fast and beneficial increase in knowledge with a consequent terrific improvement of this new technology. On the other side, the enormous amount of research promoted an immense outgrowth of scientific literature, perpetually published. Halide perovskite solar cells are now effectively competing with other established photovoltaic technologies in terms of power conversion efficiencies and production costs. Despite the tremendous improvement, a thorough understanding of the energy losses in these systems is of imperative importance to unlock the full thermodynamic potential of this material. This thesis focuses on the understanding of the non-radiative recombination processes in the neat perovskite and in complete devices. Specifically, photoluminescence quantum yield (PLQY) measurements were applied to multilayer stacks and cells under different illumination conditions to accurately determine the quasi-Fermi levels splitting (QFLS) in the absorber, and compare it with the external open-circuit voltage of the device (V_OC). Combined with drift-diffusion simulations, this approach allowed us to pinpoint the sites of predominant recombination, but also to investigate the dynamics of the underlying processes. As such, the internal and external ideality factors, associated to the QFLS and V_OC respectively, are studied with the aim of understanding the type of recombination processes taking place in the multilayered architecture of the device. Our findings highlight the failure of the equality between QFLS and V_OC in the case of strong interface recombination, as well as the detrimental effect of all commonly used transport layers in terms of V_OC losses. In these regards, we show how, in most perovskite solar cells, different recombination processes can affect the internal QFLS and the external V_OC and that interface recombination dictates the V_OC losses. This line of arguments allowed to rationalize that, in our devices, the external ideality factor is completely dominated by interface recombination, and that this process can alone be responsible for values of the ideality factor between 1 and 2, typically observed in perovskite solar cells. Importantly, our studies demonstrated how strong interface recombination can lower the ideality factor towards values of 1, often misinterpreted as pure radiative second order recombination. As such, a comprehensive understanding of the recombination loss mechanisms currently limiting the device performance was achieved. In order to reach the full thermodynamic potential of the perovskite absorber, the interfaces of both the electron and hole transport layers (ETL/HTL) must be properly addressed and improved. From here, the second part of the research work is devoted on reducing the interfacial non-radiative energy losses by optimizing the structure and energetics of the relevant interface in our solar cell devices, with the aim of bringing their quasi-Fermi level splitting closer to its radiative limit. As such, the interfaces have been carefully addressed and optimized with different methodologies. First, a small amount of Sr is added into the perovskite precursor solution with the effect of effectively reducing surface and interface recombination. In this case, devices with V_OC up to 1.23 V were achieved and the energy losses were minimized to as low as 100 meV from the radiative limit of the material. Through a combination of different methods, we showed that these improvements are related to a strong n-type surface doping, which repels the holes in the perovskite from the surface and the interface with the ETL. Second, a more general device improvement was achieved by depositing a defect-passivating poly(ionic-liquid) layer on top of the perovskite absorber. The resulting devices featured a concomitant improvement of the V_OC and fill factor, up to 1.17 V and 83\% respectively, reaching efficiency as high as 21.4\%. Moreover, the protecting polymer layer helped to enhance the stability of the devices under prolonged maximum power point tracking measurements. Lastly, PLQY measurements are used to investigate the recombination mechanisms in halide-segregated large bandgap perovskite materials. Here, our findings showed how few iodide-rich low-energy domains act as highly efficient radiative recombination centers, capable of generating PLQY values up to 25\%. Coupling these results with a detailed microscopic cathodoluminescence analysis and absorption profiles allowed to demonstrate how the emission from these low energy domains is due to the trapping of the carriers photogenerated in the Br-rich high-energy domains. Thereby, the strong implications of this phenomenon are discussed in relation to the failure of the optical reciprocity between absorption and emission and on the consequent applicability of the Shockley-Queisser theory for studying the energy losses such systems. In conclusion, the identification and quantification of the non-radiative QFLS and V_OC losses provided a base knowledge of the fundamental limitation of perovskite solar cells and served as guidance for future optimization and development of this technology. Furthermore, by providing practical examples of solar cell improvements, we corroborated the correctness of our fundamental understanding and proposed new methodologies to be further explored by new generations of scientists.},
  language  = {en}
}
@phdthesis{Phung2020,
  author    = {Phung, Thi Thuy Nga},
  title     = {Defect chemistry in halide perovskites},
  doi       = {10.25932/publishup-47652},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-476529},
  school      = {Universit{\"a}t Potsdam},
  pages     = {vi, 231},
  year      = {2020},
  abstract  = {Metallhalogenid-Perowskite haben sich aufgrund ihrer hervorragenden optoelektronischen Eigenschaften zu einer attraktiven Materialklasse f{\"u}r die Photovoltaikindustrie entwickelt. Die Langzeitstabilit{\"a}t ist jedoch noch immer ein Hindernis f{\"u}r die industrielle Realisierung dieser Materialklasse. Zunehmend zeigen sich Hinweise daf{\"u}r, dass intrinsische Defekte im Perowskit die Material-Degradation f{\"o}rdern. Das Verst{\"a}ndnis der Defekte im Perowskit ist wichtig, um seine Stabilit{\"a}t und optoelektronische Qualit{\"a}t weiter zu verbessern. Diese Dissertation konzentriert sich daher auf das Thema Defektchemie im Perowskit. Der erste Teil der Dissertation gibt einen kurzen {\"U}berblick {\"u}ber die Defekteigenschaften von Halogenid-Perowskiten. Anschließend zeigt der zweite Teil, dass das Dotieren von Methylammoniumbleiiodid mit einer kleinen Menge von Erdalkalimetallen (Sr und Mg) ein h{\"o}herwertiges, weniger fehlerhaftes Material erzeugt, was zu hohen Leerlaufspannungen sowohl in der n-i-p als auch in der p-i-n Architektur von Solarzellen f{\"u}hrt. Es wurde beobachtet, dass die Dotierung in zwei Dom{\"a}nen stattfindet: eine niedrige Dotierungskonzentration f{\"u}hrt zum Einschluss der entsprechenden Elemente in das Kristallgitter erm{\"o}glicht, w{\"a}hrend eine hohe Dotierungskonzentration zu einer Phasentrennung f{\"u}hrt. Das Material kann im Niedrigdotierungsbereich mehr n-dotiert sein, w{\"a}hrend es im Hochdotierungsbereich weniger n-dotiert ist. Die Schwelle dieser beiden Regime h{\"a}ngt von der Atomgr{\"o}ße der Dotierelemente ab. Der n{\"a}chste Teil der Dissertation untersucht die photoinduzierte Degradation von Methylammonium-Bleiiodid. Dieser Abbaumechanismus h{\"a}ngt eng mit der Bildung und Migration von defekten zusammen. Nach der Bildung k{\"o}nnen sich diese in Abh{\"a}ngigkeit von der Defektdichte und ihrer Verteilung bewegen. Demnach kann eine hohe Defektdichte wie an den Korngrenzen eines Perowskitfilms die Beweglichkeit von ionischen Punktdefekten hemmen. Diese Erkenntnis ließe sich auf das zuk{\"u}nftige Materialdesign in der Photovoltaikindustrie anwenden, da die Perowskit-Solarzellen normalerweise einen polykristallinen D{\"u}nnfilm mit hoher Korngrenzendichte verwenden. Die abschließende Studie, die in dieser Dissertation vorgestellt wird, konzentriert sich auf die Stabilit{\"a}t der neuesten „dreifach-kationen" Perowskit-basierten Solarzellen unter dem Einfluss einer permanent angelegten elektrischen Spannung. Eine l{\"a}ngere Betriebsdauer (mehr als drei Stunden permanente Spannung) f{\"o}rdert die Amorphisierung im Halogenid-Perowskiten. Es wird hierbei vermutet, dass sich eine amorphe Phase an den Grenzfl{\"a}chen bildet, insbesondere zwischen der lochselektiven Schicht und dem Perowskit. Diese amorphe Phase hemmt den Ladungstransport und beeintr{\"a}chtigt die Leistung der Perowskit-Solarzelle erheblich. Sobald jedoch keine Spannung mehr anliegt k{\"o}nnen sich die Perowskitschichten im Dunkeln bereits nach einer kurzen Pause regenerieren. Die Amorphisierung wird auf die Migration von ionischen Fehlordnungen zur{\"u}ckgef{\"u}hrt, h{\"o}chstwahrscheinlich auf die Migration von Halogeniden. Dieser Ansatz zeigt ein neues Verst{\"a}ndnis des Abbau-Mechanismus in Perowskit-Solarzellen unter Betriebsbedingungen.},
  language  = {en}
}