Metal-halide perovskites are dominating the landscape of emerging photovoltaics, with efficiency approaching that of conventional high purity semiconductors. Despite the huge success, significant weaknesses still have to be overcome, including light and thermal-induced instabilities, which in turn are likely originated by the peculiar defect chemistry of this materials class. Mastering the photochemistry of defects and impurities is thus of paramount importance to obtain stable materials and consequently devices. Significant efforts have been devoted to the understanding of defect properties of high efficiency Pb-based materials, with general consensus on iodine-related defects (vacancies and interstitials) being related to both charge trapping and ion migration. This latter aspect is strongly associated to perovskite degradation and is thought to mainly occur at the surface or grain boundaries of polycrystalline thin films, where most of the defects are located. Tin-halide perovskites are the only Pb-free material class endowed with comparable optoelectronic properties to lead-iodide perovskites. Despite significant progress, the device efficiency and stability of tin-halide perovskites is still limited by two potentially related phenomena, i.e. self p-doping and tin oxidation. Both processes are likely related to defects, thus understanding tin-halide defect chemistry is a key step towards exploitation of this materials class.

In this talk I shall give an overview of what we know on metal-halide perovskite defect chemistry in relation to improving their long term stability and optoelectronic quality. Our investigations  are based on state of the art density functional theory simulations carried out on prototypical Pb- and Sn-based perovskites, which unveil the mechanistic features and energetics associated to most abundant and potentially harmful defects in both material classes. Possible solutions to defect induced instabilities are discussed in terms of compositional alloying with different metals and surface passivation as the main strategy to obtain efficient and stable metal-halide solar cells.

Excitons are neutral quasiparticles that are formed in semiconductors and insulators upon absorption of photons. Their formation, diffusion, lifetime, and recombination are key for understanding light-conversion processes and designing materials for tailored applications, such as photovoltaics, sensing, lighting, and quantum computing. Metal-halide perovskites are a structurally and chemically diverse class of materials that have been explored in all these application areas.

In this presentation, I will discuss our recent efforts to use first-principles numerical simulation techniques for modelling the formation of excitons in metal-halide perovskites and provide design rules for bespoke material properties. Our calculations allow us to map the complex landscape of electronic properties and excitons, understand the impact of chemical heterogeneity [1 - 4], dimensionality [5 -7] and temperature effects [7], and provide chemically intuitive rules for when to trust canonical models for excitons in these materials.

Excited state dynamics play key roles in numerous molecular and nanoscale materials designed for energy conversion. Controlling these far-from-equilibrium processes and steering them in desired directions require understanding of material’s dynamical response on the nanometer scale and with fine time resolution. We couple real-time time-dependent density functional theory for the evolution of electrons with non-adiabatic molecular dynamics for atomic motions to model such non-equilibrium response in the time-domain and at the atomistic level. The talk will introduce the simulation methodology [1] and discuss several exciting applications among the broad variety of systems and processes studied in our group [2,3], including metal halide perovskites, transition metal dichalcogenides, semiconducting and metallic quantum dots, metallic and semiconducting films, polymers, molecular crystals, graphene, carbon nanotubes, etc. Photo-induced charge and energy transfer, plasmonic excitations, Auger-type processes, energy losses and charge recombination create many challenges due to qualitative differences between molecular and periodic, and organic and inorganic matter. Our simulations provide a unifying description of quantum dynamics on the nanoscale, characterize the timescales and branching ratios of competing processes, resolve debated issues, and generate theoretical guidelines for development of novel systems.

The dynamic response of metal halide perovskite devices shows a variety of physical responses that need to be understood and classified for enhancing the performance and stability and for identifying physical behaviours that may lead to developing new applications. Hysteresis, observed in the current-voltage characteristics of electronic and ionic devices, is a phenomenon in which the shape of the curve is influenced by the speed of measurement. This phenomenon is a result of internal processes that introduce a time delay in the response to external stimuli, causing measurements to depend on past disturbances. This hysteresis effect presents significant challenges, particularly in solution-processed photovoltaic devices like halide perovskite solar cells, where it complicates the assessment of performance quality. Our goal is to classify various manifestations of hysteresis by identifying common elements. Our approach involves examining hysteresis from multiple perspectives, employing simplified models that capture fundamental response patterns. We investigate system behavior using techniques like linear sweep voltammetry, impedance spectroscopy, and the analysis of transient currents resulting from small voltage steps. Our examination uncovers two primary types of hysteresis, characterized by how current responds to rapid sweep rates: capacitive hysteresis and inductive hysteresis. These terms correspond to the dominant component in the equivalent circuit, which governs the transient time response.

Halide perovskite (ABX₃) has recently achieved great success in the field of electronic devices beyond photovoltaics. Compared to conventional memristive metal oxides, halide perovskite possesses multiple chemical elements, abundant defect types, and complete defect chemistry theory that are desired for high-performance memristors. The metal-halide electrochemical reaction mediated by migration of ion defects, the formation of lead clusters, together with the valence-change mechanism driven by halide vacancy can not only meet the demands of two-state, multi-state or even continuous and reversible switching of resistance, but also provide an ideal material model for study of the physical mechanism of memristors.

Here, the unique ionic-electronic coupled transport properties of nanosized perovskite films will be discussed first. The relation between the microscopic transport mechanism and the macroscopic resistance switching characteristic will be unraveled. These perovskite films are then employed to construct memristors from the bottom up, demonstrating low energy consumption and many important synaptic features. A coupled capacitive and inductive phenomenon originating from charge trapping and ion migration, controlled by amplitude and timing of the programming pulses, is observed for defining the degree of synaptic plasticity. To further improve the performance, a two-terminal synaptic memristor based on phase-pure and well-structured single crystals of perovskites is fabricated. The device exhibits extremely low power consumption (~26 fJ) on par with that of a bio-synapse. In particular, the synapse-like potentiation and depression under continuous stimulations can be reduplicated for more than 20000 cycles with excellent reproducibility, which makes the device one of the most stable and energy efficient artificial synapses reported to date. Finally, inspired by the device behaviors a simple proof-of-concept demonstration of the potential for neuromorphic computing is performed, i.e., classification application using the devices comprising neuron network.

In recent years, cutting-edge research has been conducted worldwide on advanced neuroelectronic materials with synaptic properties that have driven the development of brain-inspired computation [1]. One of the central functions in the human brain is the biological synapsis, which has the essential property of long-term potentiation, where certain patterns of incoming stimuli modify the conductivity of the elements, favoring or inhibiting current flow. This essential phenomena of learning, memory and inference in the cognitive human brain is indispensable for the real-time information-processing and transmission. Thus, the analysis of materials with memory effects is indeed necessary to reproduce the natural functions of potentiation. Nevertheless, the fundamental understanding of the nonlinear relaxation processes to repetitive stimulus is not fully understood yet and cannot be related to basic materials properties.

Here, we study the evolution of the time transient responses that constitute the long-term potentiation mechanism of the popular halide perovskite memristors inspired in artificial neurons [2,3]. We identify the bioelectrical origin of the system’s response in terms of capacitive, inductive, and resistive elements, that can be put in correspondence with the classical ionic-electronic dynamics of perovskite semiconductors. From an advanced mathematical model based on a few nonlinear equations that is extremely successful to reproduce the multiple faces of transient responses in the long-term potentiation, we draw a physical picture that facilitates a sound dynamic interpretation of this important process of the synaptic operation [4]. This new methodology is a very powerful tool for the analysis of different types of synaptic responses and, more generally, for electronic devices with memory effects.

The efficiency of perovskite solar cells is negatively affected by open-circuit voltage losses due to radiative and non-radiative recombination of charges. The exact origin of these voltage losses and the individual contributions of radiative and non-radiative processes are not known. Presently, radiative voltage limits, determined from sensitive photocurrent measurements covering the above- and sub-bandgap regions, often yield unphysically low values. Herein, sensitive photocurrent and electroluminescence spectroscopy are used to probe radiative recombination at sub-bandgap defects in wide-bandgap mixed-halide lead perovskite solar cells. The radiative ideality factor associated with the optical transitions increases from 1, above and near the bandgap edge, to ~2 at mid-bandgap. Such photon energy-dependent ideality factor corresponds to a many-diode model. The radiative open-circuit voltage limit derived from this many-diode model enables the differentiation of radiative and non-radiative voltage losses in perovskite solar cells. The latter are further deconvoluted into contributions from the bulk and interfaces via determining the quasi-Fermi level splitting. The experiments show that while sub-bandgap defects do not contribute to radiative voltage loss, they do affect non-radiative voltage losses.

Quantifying recombination in halide perovskites is essential for controlling and enhancing the performance of perovskite solar cells. Here, we present a comprehensive analysis of recombination dynamics in perovskite solar cells, focusing on the impact of transient measurement techniques on assessing steady-state performance. We emphasize that the decay times extracted from transient photoluminescence measurements cannot generally be reduced to a single value but rather depend strongly on the charge carrier concentration. It follows that the magnitude of the decay time relies on the measurement's sensitivity. We show photoluminescence decay curves with a large dynamic range of more than 10 orders of magnitude demonstrating decay times from tens of nanoseconds up to hundreds of microseconds in perovskite films. Assuming these high decay times correspond to a recombination lifetime in steady state, one would expect a material with perfect photoluminescence quantum yield, which does however not coincide with the experimental findings. Instead, the decay times in transient measurements are prolonged by capacitive effects resulting from charge trapping and detrapping. We quantitatively explain both the transient and steady-state photoluminescence with the presence of a high density of shallow traps without the influence of deep traps. The same characteristic of increasing decay times remains in the full device, indicating capacitive effects from either shallow traps or transport layers and electrodes. With the help of voltage-dependent steady-state photoluminescence measurements we extract the recombination lifetimes during steady-state operation of the device and compare these to the transient decay times. Furthermore, this experiment allows for the determination of the voltage-dependent collection efficiency which illustrates collection losses up to 10% due to recombination even at short circuit. In summary, this study offers a deeper understanding of recombination dynamics in halide perovskite solar cells by analyzing transient and steady-state measurements. Our findings underscore the importance of considering the influence of shallow traps and transport layers in interpreting decay times and optimizing device performance.

In spite of the impressive development in terms of power conversion efficiencies, presently at 26.1%, one of the most problematic issue which still hinders the commercialization concerns the stability of the PSCs. Of critical importance is the detection and mitigation of ion migration, which is evidenced in the hysteretic effects and also in the huge apparent capacitive and inductive effects.

We introduce an equivalent circuit [1], which consistently explains the features in the dynamic J-V characteristics, like the normal and inverted hysteresis, the current bump in the reverse scan following a positive voltage pre-poling, as well as the peculiar capacitive and inductive effects visible in the impedance spectroscopy. Our model is based on the key assumption of ion-modulated recombination current. Here, we discuss the different roles of ionic charge accumulation and ionic charge current in reproducing capacitive and inductive effects, in close connection with the physical processes leading to photo-generated carrier recombination. The simulations are supported by experimental impedance spectroscopy data. Our approach also outlines a possible investigation route of ion migration, which aims to a more robust design of the PSCs.

[1] “Capacitive and Inductive Effects in Perovskite Solar Cells: The Different Roles of Ionic Current and Ionic Charge Accumulation”, N. Filipoiu, A.T. Preda, D.V. Anghel, R. Patru, R.E. Brophy, M. Kateb, C. Besleaga, A.G. Tomulescu, I. Pintilie, A. Manolescu, and G.A. Nemnes, Phys. Rev. Appl. 18, 064087 (2022).
 

Small perturbation techniques are commonly employed to extract device properties of solar cell devices. However, obtaining electronic parameters for diffusion and recombination through impedance spectroscopy in perovskite solar cells has proven elusive thus far, as the measured spectra do not exhibit electron diffusion. In a series of publications, we demonstrate that intensity modulated photocurrent and photovoltage spectroscopies (IMPS and IMVS) reveal a high-frequency spiraling feature that is governed by the diffusion-recombination constants, especially under conditions where carriers are generated far from the collecting contact.

Furthermore, we illustrate how these constants can obscure the diffusion trace in impedance spectra. We present two distinct models and experiments for different configurations: the standard sandwich-contacts solar cell device and the quasi-interdigitated back-contact (QIBC) device for lateral long-range diffusion. Our measurement results yield hole diffusion coefficients and lifetimes that align with previous findings in the literature. Notably, our analysis in the frequency domain effectively separates carrier diffusion (at high frequency) from ionic contact phenomena (at low frequency).

This discovery paves the way for a systematic determination of transport and recombination characteristics under various operando conditions.

We develop an electrical model of the perovskite solar cell (PSC) that explicitly accounts for charge exchange between the perovskite layer and the electrodes, via the relatively low-mobility transport layers.[1] Analysis of the model in the frequency domain shows that the coupling of the transport layer resistance with the perovskite capacitance creates an apparent constant lifetime that can be misinterpreted as a non-radiative recombination lifetime due to a density of defects. Comparison of the model’s predictions in the time and frequency domain further allows identifying a hidden time constant in the frequency domain data, that is clearly observed as the rise time constant in the corresponding time domain measurement.[2] This time constant typically corresponds to the speed of charge extraction from the perovskite layer to the electrodes, which is an important factor determining fill factor losses in PSCs.

The most promising technological application of perovskite solar cells (PSCs) relies on the implementation of single junction perovskite photovoltaic devices in tandem architectures, either as Si-perovskite or all-perovskite. Notoriously, wide-gap perovskites (⁓1.7-1.8 eV) required for such solar cell design are known to suffer from larger open-circuit voltage (V_OC) losses compared to the narrower gap counterparts. Commonly, these types of issues are attributed to an inherent poor material quality, due to either a larger density of traps or to phase instability. However, we demonstrate that these energy losses are associated with strong interface recombination due to an energy misalignment between the perovskite and charge transport layers (CTL), resulting in the external V_OC being lower compared to the internal quasi-Fermi level splitting (QFLS) of the same device. While at the p-interface there is a large variety of transport materials that can be used to mitigate this problem (metal oxides, polymers, and self-assembled monolayers), at the n-interface, fullerenes have been the only successful option so far. Due to this limitation, the non-radiative recombination losses at the n-interface are currently one of the major limitations of wide-gap perovskite solar cells. However, we will show, with different examples, how an appropriate interface design can significantly reduce the non-radiative recombination losses at both interfaces. Consequently, the QFLS-V_OC mismatch is reduced, allowing for accessing the actual thermodynamic potential of the perovskite absorber, even for wide-gap systems. Moreover, we will show how the interfaces can additionally have a strong impact on the energy losses occurring at short-circuit conditions, due to ion movement. As such, by coupling a theoretical and experimental approach, we demonstrate that most of the energy losses in these systems are not related to the absorber itself and particular attention should be focused on the cell architecture in order to achieve a more successful implementation of these devices in tandem configurations.

This contribution discusses experimental results obtained from a variety of transient optoelectronic techniques like transient photoluminescence (TRPL), transient photovoltage (TRPV), impedance spectroscopy (IS), and intensity modulated photovoltage (IMPV) applied to metal-halide materials and solar cells. While for the analysis of bare absorber films TRPL provides clear insight into the recombination kinetics of photogenerated charge carriers, this task gets more complex if the method is applied to layer stacks, including one or more contact layers. For finished devices it turns out that the amount of charge carriers extracted to the contact layers is, and should, be larger than the number of carriers that remains in the absorber. Therefore, a description of TRPL as well as of the other transient methods that can be applied to completed solar cells, needs to take into account a second independent variable, namely the number of charge carriers accumulated on the contacts of the solar cell. The resulting generic two-component model describes the experimental results, especially the fact that TRPV data exhibit an initial rise of the photovoltage followed by a decay.

Metal halide perovskite (HaP) materials, which are the heart of devices, such as solar cells and LEDs, challenge our understanding of semiconductors. We show that for HaPs control of the doping type and density, and properties derived from these, is to a first approximation, via their surfaces. While not unique to HaPs (the effect applies to ALL semiconductors with LOW electronically active bulk and surface defect densities), it is amplified for HaPs, because of their intrinsically low bulk and surface defect densities. Most polycrystalline (<< 1 um grain diameter) thin HaP films have carrier densities below the VOLUME, i.e., BULK densities that will result if even less than 0.1% of the surface sites function as electrically active defects. From single crystal data, we know that other bulk (electrically active) defect densities are orders of magnitude lower.

Because HaP interfaces result from contacts to HaP surfaces, the direct implication of this phenomenon is that interface defects will control HaP-based solar cells and LEDs, which involve multi-layered polycrystalline thin-film structures with two interfaces with the HaP layer.

While surface passivation effects on bulk electrical properties are relevant to all semiconductors and have been a crucial step in enabling the use of each of those in today’s electronic devices, they take on greater importance for HaPs, because achieving bulk doping at densities, relevant for electronics, has turned out to be difficult (to say the least), certainly by established doping methods. We show that only by realizing the role of surface/interface defects, will it become possible to control bulk HaP electronic properties.

With the rapid improvements in both of the external quantum efficiency and operating lifetimes of QLEDs, it approaches to the gate of industrialization for flat display applications. Since blue QLED is known to be one of the most important remaining difficult, it has been of great interests to develop the materials and device for OLEDs. In the past three years, we have tried to remove the bastion of blue QLEDs. In this talk, I will introduce our recent progress in the colloidal synthesis large sized ZnSe nanocrystals with pure blue emission as well the introduction of machine learning methodology in device analysis. Large-sized ZnSe nanocrystals with an emission peak of 455-475 nm are synthesized with a general strategy of reactivity-controlled epitaxial growth (RCEG) was developed through sequential injection of high-reactivity and low-reactivity Zn and Se precursors. We further fabricated stable, large-sized ZnSe/ZnS core-shell nanocrystals with photoluminescence quantum yields up to ~60%. Very recently, we build up a machine learning assisted methodology to predict the operational stability of blue QLEDs by analyzing the measurements of over 200 devices. By developing a convolutional neural network (CNN) model, the methodology is able to predict the operation lifetime of QLED.

Organic-inorganic hybrid perovskites (OIHPs) are a fascinating class of semiconductors that can be low-temperature synthesized, crystallized and processed on a large variety of substrates, and at the same time offering outstanding optical and electronic properties, such as broadband spectral tunability, high defect tolerance, high absorption/emission efficiency, room-temperature excitons, etc. These advantages allow OIHPs to complement conventional inorganic semiconductors (e.g., Si and GaAs) in photonics applications that require low-cost, large active area, wide spectral or polarization tunability or flexibility in substrate selection. In this talk I will give a few examples on OIHP based photonic devices, including LEDs, photodetectors and lasers. First, I will discuss how the notorious instability problems of OIHP LEDs may be addressed through surface molecular passivation. By treating the perovskite surface with phenylalkylammonium iodides (PAAIs), molecules consisting of a benzene ring, an alkyl chain and an amine tail we managed to simultaneously achieve a record T50 half-lifetime of 130 hrs under 100 mA/cm² (accelerated testing) and a record radiance of 1282.8 W/(sr × m2) for near-IR OIHP LEDS. The finding of the PAAI was enabled by a fundamental study combining experimental characterization and theoretical modelling, which reveals that the stabilization effect of the passivation is governed by the steric hindrance of the molecules to reconfiguration for accommodating ion migration from the perovskite surface. Secondly, I will focus on the commonly seen gain – bandwidth trade-off problem of photodetectors and introduce a monolithically integrated photovoltaic transistor (PVT) design to solve this dilemma. The PVT exploiting a lead halide perovskite as the photoactive layer achieved a record high gain – bandwidth product of ~ 10¹¹. Finally, I will talk about our recent exploration of using quasi-two-dimensional Dion-Jacobson (DJ) phase perovskites for laser application. With properly selected organic spacers, DJ phase OIHPs offer excellent chemical resistance and thermal processability. These properties allow us to achieve optically pumped perovskite lasers with record-high quality factor, record-low lasing threshold and excellent operational stability.

In this talk, I will give you an overview on the organic and halide perovskites semiconductors based activities taking place in our research group. These semiconductors have emerged as potential photonics materials for future commercial optoelectronic devices, however, face a major challenge of ambient stability. Organic LEDs have become a matured technology by now for small display industries with budget crossing 10s B$. I will discuss the device engineering results on single junction and multi-junction OLEDs.

In halide perovskites; we have results of single junction cell PCE ~ 21% and silicon/perovskite tandem cell PCE > 27%. We also made LEDs with decent luminance efficiency ~ 45 cd/A. In halide perovskites, the two-dimensional (2D) and quasi-2D halide perovskites have shown promising stability against ambient stability. Hence, the understanding of the photo-physics of 2D perovskites can be a tool to get an in-depth insight into these novel halide perovskites. In this talk, we will discuss the origin of broadband illumination and other emission peaks using temperature-dependent time-resolved photoluminescence (PL) spectra of 2D layered halide perovskites (i.e., (PEA)₂PbBr₄ and (PEA)₂PbI₄) semiconductors. We will show that broad emission in (PEA)₂PbBr₄ perovskite is observed due to coupling of electronic states in inorganic well part (PbBr­₆^4-) and organic barrier part (PEA), which is in contrast to a proposed model based on self-trapped-exciton.

In organic photovoltaic devices, current generation results from the sequence of photon absorption, charge separation, and charge collection in competition with recombination. To design higher performance materials and devices, we need models of device output that incoporate both the microscopic, molecular level energy and charge-transfer processes and the macroscopic, device level transport processes. Semiconductor device models are successful in describing charge collection and recombination and the resulting curent-voltage curves, but are unable to relate the processes to milecular properties, whereas molecular scale models are usually limited to small system sizes. In this talk we will present an approach that combines molecular level models and time-resolved device models in a single framework [1,2]. We show how it can explain differnet experimental phenomena in terms of molecular properties, and consider how it can be used to design in desirable moleclar properties. We will dicuss the limitations and possible extensions of the approach.

Due to their flexible material properties, perovskite materials are a promising candidate for many semiconductor devices such as lasers, memristors, LEDs and solar cells. Despite being the fastest growing photovoltaic technology, commercialization of perovskite devices faces many challenges which need to be overcome such as degradation of the material. Precise mathematical models and numerical simulations help to advance the technology by understanding how charge carriers move through the device. We present new potential-based models which take into account an additional ion species as well as corresponding volume exclusion effects on a perovskite lattice [1,2,3]. We also discuss the impact for simulations. For the simulation we rely on a stable finite volume discretization, as implemented in our flexible 1D/2D/3D Julia tool ChargeTransport.jl [4]. This is joint work with Dilara Abdel and Nicola Courtier.

Drift-diffusion simulations have proven very useful to understand the properties of novel devices such as perovskite solar cells. In order to make such simulations widely accessible and transparent, we have developed an open-software simulation suite called SIMsalabim. SIMsalabim incorporates hysteresis, ions, and tracking of the open-circuit voltage, as well as the ability to include scripting for different measurement techniques such as CELIV, TPC, TPV, and impedance spectroscopy. This makes it possible to fit simulation results to experimental data, including global fits of multiple experiments. To improve user experience and encourage its use by researchers in the field of photovoltaics, we have recently developed a web-version of SIMsalabim that runs in a web-browser [1]. In this talk, a few examples of the applicationof drift-diffusion modelling will be discussed.

Space-charge-limited current (SCLC) measurements have been widely used to study the charge carrier mobility and trap density. However, their applicability to metal halide perovskites is not straightforward, due to the mixed ionic and electronic nature of these materials. Here, we discuss the pitfalls of SCLC for perovskites, and especially the effect of mobile ions. We show, using drift-diffusion simulations, that the ions strongly affect the measurement and that the usual analysis and interpretation of SCLC need to be refined. We highlight that the trap density and mobility cannot be directly quantified using classical methods. We discuss the advantages of pulsed SCLC for obtaining reliable data with minimal influence of the ionic motion. We then show that fitting the pulsed SCLC with DD modeling is a reliable method for extracting mobility, trap, and ion densities simultaneously.

Wide-bandgap perovskite solar cells are plagued by relatively low open-circuit voltages. Here, a number of design rules to increase the open-circuit voltage of wide-bandgap perovskite solar cells are introduced. The combined effects of interface traps, ions, band alignment, and transport properties are introduced to identify the critical parameters for improving the open-circuit voltage. We show that the alignment of energy levels is only part of the story; the effective densities of states are of equal importance. The results pave the way to achieving high open-circuit voltages, despite a significant density of interface defects.

[1] simsalabim-online.com

In this talk, I will present a consistent modelling framework for perovskite solar cells that takes into account the impact of mobile ionic defects on photovoltaic performance. I will explain common perovskite device characteristics through drift-diffusion simulations and reduced-order modelling of both time- and frequency-domain measurements. We use the open-source ‘IonMonger’ code released in 2019 [1] to simulate continuum-level models which describe the effects of material properties and internal mechanisms on the behaviour of a complete device. Version 2.0 [2] can simulate a 100-point electrochemical impedance spectrum in less than a minute on a desktop computer and reproduce the distinctive ionic-electronic features exhibited by perovskite solar cells. Meanwhile, reduced-order models enable the identification of key performance metrics. I will present a simplified description of the steady-state current [3] and frequency response [4] which explains why the classical diode ideality factor is not a reliable indicator of recombination type. Instead, simulations show that we can use two factors (one electronic and one ionic) to reveal the location and type of recombination losses in a planar perovskite solar cell.

Metal halide perovskites are semiconductors that exhibit the rich electronic phenomena known from their more established counterparts, such as interface- and dopant-induced band bending, surface states and surface band bending, and surface photovoltage. But they feature even more complexity due to moderate stability under optical excitation in vacuum that can induce surface states, and reversible p-doping by (ambient) oxygen. The simultaneous occurrence of all these phenomena has initially retarded progress towards a comprehensive understanding of their electronic properties, because the most direct experimental method to assess these properties – photoelectron spectroscopy – has been insufficiently adapted to the needs of the perovskites. Now that several important fundamental questions are resolved, such as those discussed here, we can look forward to obtaining deeper insight into even more complex properties and processes of this fascinating material class. In addition, novel interfacial phenomena have been identified, such as photo-induced energy level re-alignment at charge-selective contacts. This necessitates careful photoelectron spectroscopy studies under operando conditions, in order to reliably link interface energetics and solar cell performance.

Probing Perovskite Solar Cells: Insights into Surface, Interface, and Bulk Characteristics through X-ray Photoelectron Spectroscopy

Authors: Chittaranjan Das, Mayank Kedia and Michael Saliba

Institute for Photovoltaics,

University of Stuttgart, Stuttgart

Germany

Perovskite photovoltaic research has witnessed remarkable advancements in power conversion efficiency, primarily attributable to the outstanding optoelectronic characteristics of the semiconductor absorber layer. Simultaneously, the scientific community has undertaken comprehensive investigations into the fundamental properties that hold a pivotal role in shaping device performance. Among many fundamental properties, the precise positioning of interface band edges between various transport layers and the perovskite absorber layer emerges as a critical determinant, profoundly influencing the flow of photogenerated charges within the layer and thereby exerting a decisive impact on device performance. The positioning of band edges and electronic properties of the charge transport and contact layers are intrinsically governed by their chemical composition and the complex interplay they engage in with adjacent layers. This complex interplay, driven by the chemical nature of the materials involved, plays a decisive role in optimizing charge transport, minimizing recombination losses, and maximizing overall device efficiency.

The correlation between the chemical composition and its impact on the electronic properties of the semiconducting layer has been extensively investigated through X-ray photoelectron spectroscopy (XPS) studies [1]. In the context of perovskite solar cells, these XPS studies hold significant importance as they provide critical insights into the interfacial chemical and electronic properties [2,3].

In this presentation, we will delve into various facets of XPS studies in perovskite solar cells, encompassing examinations from the surface to the bulk properties and spanning across the critical interfaces within the device structure. We will showcase and discuss the principal findings derived from our XPS investigations on perovskite films and solar cell devices. These findings will encompass the electronic nature of the perovskite absorber film and its interactions at the interfaces with both electron and hole-transporting layers[4-6]. Through these discussions, we aim to shed light on the pivotal role that XPS studies play in advancing our comprehension of perovskite solar cell technology, ultimately driving innovation in renewable energy research.

Hybrid organic inorganic metal halide perovskites (MHPs) denote a family of compound semiconductors, which established a novel class of optoelectronics, most prominently known for the perovskite solar cell. While the power conversion efficiency of these photovoltaic devices saw a steep rise in the past decade, tailoring the interfaces between the MHP film and charge transport layer became the major control lever to enhance performance. The use of photoemission spectroscopy to analyze the chemical and electronic properties of these interfaces has been challenging due to many possible chemical reactions at the buried interfaces [1].

Here, I will discuss the use of synchrotron- and lab-based X-ray photoelectron spectroscopy (XPS) experiments to address the particular chemistry of MHP interfaces to adjacent oxide charge transport layers (CTL). At the example of oxide ALD-SnO2 layer grown on top of a double cation mixed halide perovskite film investigated by hard X-ray photoelectron spectroscopy (HAXPES), we find evidence for the formation of new chemical species and changes in the energy level alignment at the interface. The spectra exhibit binding energy shifts that indicate upward band bending of the MHP energy levels. Assuming flat band conditions at the MHP surface prior interface formation, this upward band bending may form an electron transport barrier detrimental to cell performance.

We also employed the methodology to evaluate lead-free halide perovskite films based on formamidinium tin iodide (FASnI₃), for which tin fluoride (SnF₂) is a commonly used additive enabling a retardation of tin oxidation and a reduction of tin vacancies. Our measurements reveal that SnF₂ significantly improves the layer morphology, but preferably precipitates at the PEDOT:PSS/MHP interface where it forms an ultrathin SnS interlayer induced by a chemical reaction with sulfur-containing groups at the PEDOT:PSS surface. Our work adds a new aspect to the discussion of high-efficiency Sn-based perovskite solar cells which still commonly make use of PEDOT:PSS as HTL material in contrast to Pb-based solar cells [2].

I will conclude my talk with a general discussion about the use of PES methods for the analysis of MHP layers and in particular the effect of irradiation-induced beam damage via synchrotron and lab-based X-ray sources.³ By using complementary photoluminescence measurements we are able to reveal beam-induced changes to the optoelectronic properties and track unique physicochemical phenomena such as stimulated self-healing in formamidinium lead bromide (FAPbBr₃).^4,5

Thin-film opto-electrical devices based on perovskite and organic materials possess several interesting advantages compared to their inorganic counterparts. The continuous improvements in performance have led to commercial applications of OLEDs in displays and promising power conversion efficiencies of perovskite solar cells (PSCs) close to the established silicon photovoltaic technology. Progress in those technologies was also facilitated by the increased understanding of device operations and consequently targeted optimization of materials and device structures.

Electro-optical characterization techniques in steady-state, transient and frequency-domain are widely used to gain insights into operation mechanisms. Often, such characterizations are performed on systematically varied devices to understand trends. As a first example, we will show that excess formamidinium precursor in the perovskite layer leads to a faster ion dynamic in the device, as observed through electrical impedance spectroscopy.[1] Second, we analyse the impact of degradation by comparing open-circuit voltage decay (OCVD) measurements on fresh and degraded PSCs.

A quantitative analysis of material parameters by applying analytical formulas requires a careful assessment of the assumptions used in the model. In perovskite devices, the quantification of ionic charge carrier densities and mobilities is of outmost importance. Methods such as transient capacitance,[2] bias-assisted charge extraction[3] and OCVD[4] have been used to determine them. By using drift-diffusion simulations, we show that the traditional analysis of transient current measurements are significantly underestimating the mobile ion densities in systems with high ionic densities, in line with recent findings by Diekmann et al.[3] Based on these results, a slightly adapted method to obtain the correct mobility and density of the ionic charge carriers is proposed.

Combining electro-optical characterization and device simulations offers additional opportunities for the analysis of PSCs. In the last part of this contribution, we employed this strategy to measured electroluminescence (EL) images of a carbon-based mesoporous PSC. While taking transient EL images of 1.4 cm² cells, we found very pronounced - spatially varying - dynamics in the EL signal. To understand the local temporal fluctuation in the EL signal, the complete device stack was modelled with the simulation software Setfos. Through parameter variation of transient simulations, it was found that an increased ion density could reproduce the stronger EL signal, experimentally observed at certain locations. As mobile ions are affecting the long-term stability of PSCs, spots with increased ionic concentrations could catalyse the degradation.

Recombination in semiconductors is often quantified using the concept of a charge-carrier lifetime. This is particularly sensible either in doped semiconductors or in intrinsic semiconductors with dominant recombination via deep defects. In both cases, the carrier density will decay exponentially after a pulsed excitation and assigning a characteristic time to this exponential decay is then a sound approach to analyze the data. However, in halide perovskites, intrinsic defects are predominantly shallow. Furthermore, in the popular lead-halide perovskites with a significant fraction of I as anion, the doping density is extremely low. Thus, the decay is not always exponential. We show that if care is taken to maximize the dynamic range of transient photoluminescence measurements, the decay is often highly non-exponential and in fact more resembling a power law of photoluminescence intensity being approximately proportional to 1/time². This has important consequences for the interpretation of the data. The concept of a lifetime becomes less helpful as the experimentally observed decay time is a strong function of carrier density and therefore laser fluence. Furthermore, repetition rate becomes very important as the decay time depends also on the time (approximately linearly) at which the decay time is determined. We discuss the alternative approach to introduce a recombination coefficient instead and explain how to relate this coefficient to the properties of the shallow traps.  

Recently we suggested a new solar cell loss analysis by using the absorptance or simple external quantum efficiency (EQE) measured with sufficiently high sensitivity to also account for defects. Unlike common radiative-limit methods, where the impact of deep defects is ignored by exponential extrapolation of the Urbach absorption edge, our loss analysis considers the full EQE including states below the Urbach edge and uses corrections for band-filling and light trapping. We validate this new metric on a whole range of photovoltaic materials and verify its accuracy by electrical simulations. Any deviations between this newly established metric and experimental open circuit voltage are due to the presence of spatially localized defects and are explained as violations of the assumption of flat quasi-Fermi levels through the device [1]. Newly, the reciprocity between photoluminescence and absorptance will be discussed and the method of experimental determination of light trapping correction factor will be introduced and experimentally validated on the perovskite layers deposited on substrates with different roughness.

Understanding the role of mobile ionic defects in hybrid perovskite devices is one of the most important challenges in the field of perovskite optoelectronics. Progress in this direction relies on the development of appropriate experimental methods able to probe mixed conduction in these materials, and of models that can provide an accessible, yet accurate, interpretation of device function. [1]

In this contribution, we highlight important aspects in the charge carrier chemistry of mixed conducting halide perovskites at equilibrium. First, we present a systematic approach to the investigation of mixed conduction in methylammonium lead iodide (MAPI) thin films based on horizontal device structures. The results highlight the importance of electrode charging in the long time scale polarization behavior of MAPI devices, [2] consistent with the dynamics of electric field screening in MAPI devices probed with spectroscopic and optoelectronic methods. [3] Further ionic interactions occurring at interfaces at equilibrium are presented, emphasizing the role of mobile ionic defects in the determination of space charge equilibrium in mixed conducting solar cells. [4] To facilitate the discussion of these aspects, we refer to generalized energy diagrams, where the representation of both the ionic and the electronic properties are combined. [5,6] This also highlights the importance of the halogen partial pressure, as a knob for the control of stoichiometry and, therefore of mixed conductivity. We conclude by giving a perspective on how the understanding of the material’s behavior at equilibrium can be extended to the situation under light and voltage bias.

The efficiency of solution-processed bulk-heterojunction organic solar cells has continued to increase in recent years and is now approaching 20%. This continued improvement in cell efficiency is maintaining substantial academic as well as commercial interest in this technology. Compared to their inorganic counterparts, the non-radiative energy loss is higher in organic solar cells which limits their open-circuit voltage (V_OC) and thus the power conversion efficiency. Therefore, understanding the physical nature of non-radiative energy loss and V_OC of organic solar cells is of substantial interest to the community. Studying the temperature dependence of V_OC can provide unique insights into the physical origin of non-radiative loss because non-radiative recombination is temperature dependent. In addition, understanding the temperature dependence of V_OC is of practical interest because the extrapolation of the V_OC vs. temperature curve to 0 K may provide an estimate of the maximum achievable V_OC for organic solar cells. In this study, the authors provide a coherent description of the temperature dependence of open-circuit voltage of organic solar cells, combining experimental measurements and numerical simulations. Specifically, the authors have experimentally measured the temperature dependence of charge carrier mobilities which have then been used to simulate the temperature-dependent V_OC of organic solar cells. Significantly, it is found that the experimentally measured temperature dependence of open-circuit voltage can be correctly reproduced only if the temperature-dependent mobilities are included. Other factors that can complicate the analysis of the temperature-dependent V_OC  will also be covered in the talk including the leakage current,^[1] and the different energetic levels at the donor:acceptor interfaces of bulk-heterojunction and bilayer organic solar cells.^[2]

Lead halide perovskites have attracted significant attention over the last decade,[1] particularly in light harvesting.[2] Since the ground-breaking works demonstrated the potential of perovskite materials for photovoltaics, the field has advanced rapidly, with record efficiencies exceeding 25% within less than a decade.[3] In parallel to the extensive study of solar cells, lead halide perovskite thin film transistors (TFTs) are being developed. Reports on perovskite TFTs show dominant ionic effects causing large hysteresis, gate voltage screening, device degradation, and lack of saturation in the current. Several attempts to minimize these effects are made, such as measurements in cryogenic conditions to reduce ionic conductivity, or pulsed mode measurements to minimize the slow ionic response. Still, most of the experimental data on perovskite TFTs lack current saturation.[4-9]

Namely, the field is still in its infancy, with the entrance barrier being the ability to control the ion-related effects. To address this ion migration phenomenon and provide physical reasoning for the various experimental reports, we present a 2D device simulation of lead-halide perovskite-based TFTs containing mobile charged species using Sentaurus Device by Synopsys®. Examining the transistor output performance for different average ion densities and scan durations. We found that the sign of the mobile specie creates a difference between electron-channel (negative channel) and hole-channel (positive channel) TFTs. Moreover, we found that the incomplete saturation is due to ions’ effect on the charge extraction through the contacts and not to channel effects (as screening etc.). Building on the importance of the JV scan protocol for solar cells, we devote significant attention to the role of the voltage scan speed.

Since, comparing electron-channel and hole-channel TFTs can help decipher the sign of the dominant mobile ion, its density, and its dynamic within the film, utilizing the same perovskite materials as in solar cells would allow researchers to improve the understanding of the mechanisms governing solar PVs and better their performance.

Shamalia, D. and N. Tessler, Device Simulations of Perovskite Transistors Containing Mobile Iodide. Submitted.

Device modeling is extensively used in solar cell research going from simple models such as the Shockley diode equation or partial differential equations (PDE) to more complex models like Monte-Carlo or drift-diffusion (DD). These models are often used as a diagnostic tool to understand and quantify the main losses for a given device by reproducing experimental measurements.

However, one of the main criticisms about using complex models as a means to quantify material properties is the many fitting (10 to 40) parameters that need to be estimated. Many argue that with so many fitting parameters one could fit almost any model to the experimental data.

In this presentation, the challenges of using high-dimensional physical models to quantify material properties accurately will be addressed. As well as, how machine learning methods such as Bayesian optimization (BO) combined with high-throughput (HT) experimental data can be leveraged to study material properties at scale.

The potential of using modeling, BO and HT data will be illustrated by several case studies using different physical models (PDE or DD) and experimental methods such as (i) transient absorption spectroscopy, (ii) transient photoluminescence and microwave conductivity, (iii) light-intensity-dependent current-voltage characteristics.

The open-source package BOAR (Bayesian Optimization for Automated Research) will be introduced here as a flexible package to address many challenges of solar cell research such as smart experimental planning or fitting of high-dimensional models to experimental data. This package combined with PDE solvers and open-source DD package SIMsalabim provides an easy-to-use and flexible toolbox for solar cell research.

Bayesian inference methods are useful tools to distinguish between a high number of correlated parameters in a system. They have been previously used in the field of solar cells for well-studied technologies like silicon solar cells, which are already well studied such that a lot of material parameters are well known. It has also been used for material systems like SnS solar cells where the number of unknown parameters is small [1–3]. However, previous implementations were limited by the following challenges - (1) not easily scalable to higher dimensions, (2) solution depending on the initialization condition of the sampler and (3) the speed of operation was also limited by the device model.

In this presentation, we will address these challenges and discuss the strategies we have implemented to overcome them. In our implementation of Bayesian inference, we solve the issue of scalability by using the state-of-the-art Markov chain Monte Carlo (MCMC) sampling technique. However, even though MCMC makes Bayesian inference scalable to higher dimensions, the solution is highly sensitive to the initialization condition of the samplers. To solve this problem, we introduce a hybrid MCMC method coupled with optimization algorithms such that we can maintain robustness of initialization condition from one run to another. This method of initialization is also robust at finding multiple minima/maxima in the solution space. We also incorporated a neural network based surrogate model to replace the device model and hence not limited by the speed of the device model. Overall, as an effect of the three improvements, we achieve improvement in scalability, robustness, and speed even in 15 dimensional problems. The improvements discussed have an overarching impact on high dimensional parameter estimation even in needle-in-haystack like situations.

[1]        R. E. Brandt, R. C. Kurchin, V. Steinmann, D. Kitchaev, C. Roat, S. Levcenco, G. Ceder, T. Unold, and T. Buonassisi, Joule 1, 843 (2017).

[2]        R. Kurchin, G. Romano, and T. Buonassisi, Comput. Phys. Commun. 239, 161 (2019).

[3]        R. C. Kurchin, J. R. Poindexter, V. Vähänissi, H. Savin, ¶ Carlos Del Cañizo, and T. Buonassisi, How Much Physics Is in a Current-Voltage Curve? Inferring Defect Properties from Photovoltaic Device Measurements (2019).

In our study of carbon-based hole-transporter-free PSCs with mesoporous layers, we observed pronounced local variations in Electroluminescence (EL) of turn-on dynamics, consistent with prior research but with notable local enhancements in EL intensity. To elucidate these variations, we employed a 1D drift-diffusion simulation (Setfos 5.3) to model the device stack, fitting material parameters and using a recombination-coupled emission model to simulate transient EL signals.

Our simulations revealed that locally increased ion densities could account for the stronger EL signal at specific positions. The temporal behavior of this phenomenon is strongly influenced by the mobility of iodine vacancies. The results of the 1D drift-diffusion model were integrated into the 2D+1D FEM software tool Laoss to account for sheet resistances on the top and bottom electrodes [1], thereby explaining potential drops detectable in EL images. The FEM tool also facilitated qualitative comparisons with EL images by displaying simulated transient luminescence results in 2D.

In summary, our study of mesoporous perovskite solar cells' transient EL images revealed local ion density variations, reproduced through a combination of 1D drift-diffusion and 2D+1D FEM simulations. These findings underscore the impact of mobile ionic charge carriers on the long-term performance of PSCs, with spots exhibiting increased ionic concentrations potentially serving as vulnerable points susceptible to degradation.