Drying and Coating of Perovskite Thin Films: How to Control the Thin Film Morphology in Scalable Dynamic Coating Systems

Simulation of perovskite thin layer crystallization with varying evaporation rates

Perovskite solar cells (PSC) are promising potential competitors to established photovoltaic technologies due to their superior efficiency and low-cost solution processability. However, the limited understanding of the crystallization behaviour hinders the technological transition from lab-scale cells to modules. In this work, advanced phase field (PF) simulations of solution-based film formation are used for the first time to obtain mechanistic and morphological information that is experimentally challenging to access. The well-known transition from a film with many pinholes, for a low evaporation rate, to a smooth film, for high evaporation rates, is recovered in simulation and experiment. The simulation results provide us with an unprecedented understanding of the crystallization process. They show that supersaturation and crystallization confinement effects determine the final morphology. The ratio of evaporation to crystallization rates turns out to be the key parameter driving the final morphology. Increasing this ratio is a robust design rule for obtaining high-quality films, which we expect to be valid independently of the material type.

Retrograde Solubility of Methylammonium Lead Iodide in γ-Butyrolactone Does Not Enhance the Uniformity of Continuously Coated Films

Halide perovskite thin films can be the centerpiece of high-performance solar cells, light-emitting diodes, and other optoelectronic devices if the films are of high uniformity and relatively free of pinholes and other defects. A common strategy to form dense films from solution has been to generate a high density of nuclei by rapidly increasing supersaturation, for example, by timely application of an antisolvent or forced convection. In this work, we examine the role of retrograde solubility, wherein solubility decreases with increasing temperature, as a means of increasing the nucleation density and film coverage of slot-die-coated methylammonium lead iodide (MAPbI3) from γ-butyrolactone (GBL) solution. Coverage was investigated as a function of the substrate temperature and the presence and temperature of an air knife. Results were considered within the framework of the dimensionless modified Biot number, which quantifies the interplay between evaporation and horizontal diffusion. Moderate temperatures and a heated air knife improved film coverage and morphology by enhanced nucleation up to ∼80 °C. However, despite the dense nucleation enabled by retrograde solubility, slow evaporation as a result of the low vapor pressure of GBL, combined with Ostwald ripening at high temperatures, prevented the deposition of void-free, device-quality films. This work has provided a more detailed understanding of the interplay between perovskite processing, solvent parameters, and film morphology and ultimately indicates the obstacles to forming dense, uniform films from solvents with high boiling points even in the presence of rapid nucleation.

Modular slot-die coater for in situ grazing-incidence x-ray scattering experiments on thin films

Multimodal in situ experiments during slot-die coating of thin films pioneer the way to kinetic studies on thin-film formation. They establish a powerful tool to understand and optimize the formation and properties of thin-film devices, e.g., solar cells, sensors, or LED films. Thin-film research benefits from time-resolved grazing-incidence wide- and small-angle x-ray scattering (GIWAXS/GISAXS) with a sub-second resolution to reveal the evolution of crystal structure, texture, and morphology during the deposition process. Simultaneously investigating optical properties by in situ photoluminescence measurements complements in-depth kinetic studies focusing on a comprehensive understanding of the triangular interdependency of processing, structure, and function for a roll-to-roll compatible, scalable thin-film deposition process. Here, we introduce a modular slot-die coater specially designed for in situ GIWAXS/GISAXS measurements and applicable to various ink systems. With a design for quick assembly, the slot-die coater permits the reproducible and comparable fabrication of thin films in the lab and at the synchrotron using the very same hardware components, as demonstrated in this work by experiments performed at Deutsches Elektronen-Synchrotron (DESY). Simultaneous to GIWAXS/GISAXS, photoluminescence measurements probe optoelectronic properties in situ during thin-film formation. An environmental chamber allows to control the atmosphere inside the coater. Modular construction and lightweight design make the coater mobile, easy to transport, quickly extendable, and adaptable to new beamline environments.

Modeling and Fundamental Dynamics of Vacuum, Gas, and Antisolvent Quenching for Scalable Perovskite Processes

Hybrid perovskite photovoltaics (PVs) promise cost-effective fabrication with large-scale solution-based manufacturing processes as well as high power conversion efficiencies. Almost all of today's high-performance solution-processed perovskite absorber films rely on so-called quenching techniques that rapidly increase supersaturation to induce a prompt crystallization. However, to date, there are no metrics for comparing results obtained with different quenching methods. In response, the first quantitative modeling framework for gas quenching, anti-solvent quenching, and vacuum quenching is developed herein. Based on dynamic thickness measurements in a vacuum chamber, previous works on drying dynamics, and commonly known material properties, a detailed analysis of mass transfer dynamics is performed for each quenching technique. The derived models are delivered along with an open-source software framework that is modular and extensible. Thereby, a deep understanding of the impact of each process parameter on mass transfer dynamics is provided. Moreover, the supersaturation rate at critical concentration is proposed as a decisive benchmark of quenching effectiveness, yielding ≈ 10-3 - 10-1s-1 for vacuum quenching, ≈ 10-5 - 10-3s-1 for static gas quenching, ≈ 10-2 - 100s-1 for dynamic gas quenching and ≈ 102s-1 for antisolvent quenching. This benchmark fosters transferability and scalability of hybrid perovskite fabrication, transforming the "art of device making" to well-defined process engineering.

Discovering Process Dynamics for Scalable Perovskite Solar Cell Manufacturing with Explainable AI

Large-area processing of perovskite semiconductor thin-films is complex and evokes unexplained variance in quality, posing a major hurdle for the commercialization of perovskite photovoltaics. Advances in scalable fabrication processes are currently limited to gradual and arbitrary trial-and-error procedures. While the in situ acquisition of photoluminescence (PL) videos has the potential to reveal important variations in the thin-film formation process, the high dimensionality of the data quickly surpasses the limits of human analysis. In response, this study leverages deep learning (DL) and explainable artificial intelligence (XAI) to discover relationships between sensor information acquired during the perovskite thin-film formation process and the resulting solar cell performance indicators, while rendering these relationships humanly understandable. The study further shows how gained insights can be distilled into actionable recommendations for perovskite thin-film processing, advancing toward industrial-scale solar cell manufacturing. This study demonstrates that XAI methods will play a critical role in accelerating energy materials science.

Quasi-2D Model to Predict Solid Microstructure in Drying Thin Films

The microstructure of solid coatings produced by solution processing is highly dependent on the coupling between growth, solute diffusion, and solvent evaporation. Here, a quasi-2D numerical model coupling drying and solidification is used to predict the transient lateral growth of two adjacent nuclei growing toward each other. Lateral gradients of the solute and solvent influence the evolution of film thickness and solid growth rate. The important process parameters and solvent properties are captured by the dimensionless Peclet number (Pe) and the Biot number (Bi), modified by an aspect ratio defined by the film thickness and distance between nuclei. By variation of Pe and Bi, the evaporation dynamics and aspect ratio are shown to largely determine the coating quality. These findings are applied to drying thin films of crystallizing halide perovskites, demonstrating a convenient process map for capturing the relationship between the modified Bi and well-defined coating regimes, which may be generalized for any solution-processed thin film coating systems.

Oriented Perovskite Film from Laser Recrystallization in Magnetic Field

The orientation of crystals on the substrate and the presence of defects are critical factors in electro-optic performance. However, technical approaches to guide the orientational crystallization of electro-optical thin films remain challenging. Here, a novel physical method called magnetic-field-assisted pulse laser annealing (MAPLA) for controlling the orientation of perovskite crystals on substrates is reported. By inducing laser recrystallization of perovskite crystals under a magnetic field and with magnetic nanoparticles, the optical and magnetic fields are found to guide the orientational gathering of perovskite units into nanoclusters, resulting in perovskite crystals with preferred lattice orientation in (110) and (220) perpendicular to the substrate. The perovskite crystals obtained by MAPLA exhibit significantly larger grain size and fewer defects compared to those from pulsed laser annealing (PLA) and traditional thermal annealing, resulting in improved carrier lifetime and mobility. Furthermore, MAPLA demonstrates enhanced device performance, increasing responsivity and detectivity by two times, and photocurrent by nearly three orders compared with PLA. The introduction of Fe2 O3 nanoparticles during MAPLA not only improves crystal size and orientation but also significantly enhances long-term stability by preventing Pb2+ reduction. The MAPLA method has great potential for fabricating many electro-optical thin films with desired device properties and stability.

Controlling Thin Film Morphology Formation during Gas Quenching of Slot-Die Coated Perovskite Solar Modules

Transferring record power conversion efficiency (PCE) >25% of spin coated perovskite solar cells (PSCs) from the laboratory scale to large-area photovoltaic modules requires significant advances in scalable fabrication techniques. In this work, we demonstrate the fundamental interrelation between drying dynamics of slot-die coated precursor solution thin films and the quality of resulting slot-die coated gas-quenched polycrystalline perovskite thin films. Well-defined drying conditions are established using a temperature-stabilized, movable table and a flow-controlled, oblique impinging slot nozzle purged with nitrogen. The accurately deposited solution thin film on the substrate is recorded by a tilted CCD camera, allowing for in situ monitoring of the perovskite thin film formation. With the tracking of crystallization dynamics during the drying process, we identify the critical process parameters needed for the design of optimal drying and gas quenching systems. In addition, defining different drying regimes, we derive practical slot jet adjustments preventing gas backflow and demonstrate large-area, homogeneous, and pinhole-free slot-die coated perovskite thin films that result in solar cells with PCEs of up to 18.6%. Our study reveals key interrelations of process parameters, e.g., the gas flow and drying velocity, and the exact crystallization position with the morphology formation of fabricated thin films, resulting in a homogeneous performance of corresponding 50 × 50 mm2 solar minimodules (17.2%) with only minimal upscaling loss. In addition, we validate a previously developed model on the drying dynamics of perovskite thin films on small-area slot-die coated areas of ≥100 cm2. The study provides methodical guidelines for the design of future slot-die coating setups and establishes a step forward to a successful transfer of solution processes towards industrial-scale deposition systems beyond brute force optimization.

Flexible Microspectrometers Based on Printed Perovskite Pixels with Graded Bandgaps

Miniaturized spectrometers have attracted much attention due to their capability to detect spectral information within a small size. However, such technology still faces challenges including large-scale preparation and performance repeatability. In this work, we overcome these challenges by demonstrating a microspectrometer constructed with a series of pixelized graded-bandgap perovskite photodetectors fabricated with inkjet printing. High-quality perovskite films with minimal pinholes and large grains are deposited by optimizing printing conditions including substrate temperature and surface modification. The resulting perovskite photodetectors show decent photosensing performance, and the different photodetectors based on perovskite films with different bandgaps exhibit various spectral responsivities with different cutoff wavelength edges. Microspectrometers are then constructed with the array of the pixelized graded-bandgap perovskite photodetectors, and incident spectra are algorithmically reconstructed by combining their output currents. The reconstruction performance of the miniaturized spectrometer is evaluated by comparing the results to the spectral curve measured with a commercial bulky spectrometer, indicating a reliable spectral reconstruction with a resolution of around 10 nm. More significantly, the miniaturized spectrometers are successfully fabricated on polymer substrates, and they demonstrate excellent mechanical flexibility. Therefore, this work provides a flexible miniaturized spectrometer with large-scale fabricability, which is promising for emerging applications including wearable devices, hyperspectral imaging, and internet of things.