The benefits of vacuum processed perovskite solar cells will be discussed. Different vacuum based methods are used ranging from thermal co-sublimation, close-space sublimation and vapor assisted flash evaporation.

I will also discuss the benefit of using these perovskites in substrate configuration as well as semitransparent cells for bifacial operation and for building integrated PV. The use of sublimed perovskite in perovskite-silicon tandem cells using fully textured Si bottom cells will be presented, including methods to identify the shortcomings of these tandem cells. 

Using a home build setup to determine the stability for both indoor and outdoor applications we are able to monitor the solar cells stability. First results on outdoor and indoor stability measurements will also be presented demonstrating the importance of encapsulation methods. Best devices have a T95 over 1000 hours mpp at 75 degrees C and 1 sun illumination.

Current research in thin-film deposition emphasizes compatibility with large-scale manufacturing, reproducibility, and uniformity. Although thermal vacuum deposition has long been established in the semiconductor and OLED industries, applying this technique to hybrid perovskite materials remains particularly challenging. The complexity arises primarily from their unconventional sublimation behaviors, characterized by abnormal adsorption dynamics and difficulties in precise process control. In this presentation, I will first elucidate the distinctive sublimation mechanisms and deposition kinetics associated with hybrid perovskites, highlighting how these differ fundamentally from traditional thermal evaporation of conventional inorganic materials. Subsequently, I will introduce our novel solutions to overcome these challenges, including the implementation of a meticulously designed deposition protocol combined with a specifically tailored evaporator system capable of real-time, precise process control. Leveraging these advancements, we successfully fabricated perovskite solar cells with power conversion efficiencies exceeding 25% in conventional n-i-p architectures, demonstrating remarkable batch-to-batch reproducibility. Additionally, I will discuss our recent progress on vacuum-deposited modules exceeding 200 cm², illustrating the significant industrial interest and considerable potential of vacuum-based techniques for perovskite photovoltaic commercialization.

Co-evaporated perovskite holds multiple advantages in comparison to deposition via solution processes, such as conformal deposition on textured substrates and upscalability to large-area devices [1][2]. However, controlling the sublimation rates of organic precursors such as MAI or MACl remains challenging. Their low sticking coefficient on Quartz Crystal Microbalances and their purity-dependent sublimation behavior complicates the monitoring of depositions rates and affects the process reproducibility. This problem becomes especially critical when multiple organic precursor sources are used.

In this work we present a novel approach to sublime organic materials by pressing the precursor powder into a pellet. We observe that using a pellet leads to a lower material consumption as well as a lower pressure inside the chamber during the initial heating stage. Additionally, the pellet can incorporate multiple organic compounds, allowing the simultaneous sublimation of methylammonium iodide and chloride (MAI and MACl) without the requirement of an extra source. We use this pellet alongside a Pb(I_1-­xBr_x)₂ source to deposit a triple-halide MAPIBrCl perovskite with different amount of chloride, determined by the MACl proportion in the pellet. The inclusion of chloride in the perovskite lattice is confirmed by photoluminescence, X-ray diffraction and X-ray fluorescence measurements, in contraste with other reports where MACl helps the crystallization stage of the film but leaves during annealing. Adding chloride proves to be beneficial for the charge transport properties of the film, which translates to a higher fill factor in solar cells. As such, a champion power conversion efficiency of 19.5% was obtained for a 1.66 eV bandgap perovskite.

Low-dimensional Ruddlesden-Popper (RP) structures are becoming ubiquitous in the field of halide perovskites optoelectronics. These materials are tunable between 2D (n = 1) and quasi-2D (1 < n < inf.) phases which determine their optical (bandgap), electronic (binding energy, conductivity) and chemical properties (hydrophobicity) [1]. The tunability has enabled their use as the active layer for light absorption, detection and emission, or at interfaces for efficient charge transport and improved stability. However, solution-processing of such films is challenging since interactions between precursors (organic and metal cations, halide anions) and solvents (such as DMF, DMSO) cause competing crystallization reactions that undermine dimensional purity and impact the optoelectronic properties and device stability [2]. We developed a co-evaporation method to deposit RP films with high control on crystallization by eliminating these precursor-solvent interactions [3,4]. Here, we deposit prototypical 2D (PEA₂PbI₄) and quasi-2D (PEA₂FAPb₂I₇) thin films using phenethylammonium (PEA+) as the organic spacer cation and formamidinium (FA+) as the organic cation for the quasi-2D phase and study the interactions that determine film formation.

The formation of the bare 2D layer (PEA₂PbI₄) and the development of an interface on top of a 3D perovskite film was studied using X-ray diffraction, X-ray photoelectron spectroscopy, electron microscopy and photoluminescence microscopy. The co-evaporated 2D structures were then used as interfacial layers at the hole-transporting interface to yield approx. 22% efficient perovskite solar cells, driven by improvements in the open-circuit voltage and fill factor [4]. Notably, this performance gain is observed over a large RP thickness range, owing to the improved control on RP phase through co-evaporation.

We then studied the crystallization of quasi-2D structures (PEA₂FAPb₂I₇) and observed the impact of phosphonic acid substrate modification using density functional theory calculations, synchrotron-based X-ray scattering and ultrafast pump-probe spectroscopy [5]. Here, favorable interactions between unbound solution-processed  propylphosphonic acid (PPAc) and PEA⁺ increases the relative uptake of PEA⁺ compared to FA⁺, favoring the formation of the 2D phase over the quasi-2D phase. This allows the formation of heterostructures between 2D and quasi-2D phases, controlled by the PPAc concentration and the co-evaporated film thickness.

Taken together, our work demonstrates a new synthesis method to overcome a critical challenge in coating RP perovskites with high phase purity. We demonstrate the applicability of this approach in solar cells as well as present a new way to synthesize controlled RP heterostructures. This work also complements ongoing research in the field of halide perovskites on vapor-based coating methods and presents a scalable method to coat bulk films as well as interfacial layers for high-performance devices.

Metal halide perovskite solar cells have gained significant attention over the last decade due to their low-cost fabrication methods and high efficiency potential. Typically, perovskite films are prepared by solution-based deposition techniques, which offer short deposition times and a broad range of compositions.¹ However, achieving conformal coverage of textured surfaces, highly relevant for monolithic perovskite/silicon tandem solar cells,²  or compositional gradients in the absorber material, to achieve graded Fermi levels,³ can be challenging with these techniques. These limitations can be overcome by co-evaporating the perovskite precursor materials.

Our work focuses on the vacuum-based preparation of perovskite absorbers with a band gap of about 1.68 eV, optimized for monolithic perovskite/silicon tandem solar cells. We show how the choice of hole-transporting material affects the composition of perovskite films in p-i-n solar cells. Our findings reveal that perovskites co-evaporated on spin-coated [2-(3,6-Dimethoxy-9H-carbazol-9-yl)ethyl]phosphonic acid (MeO-2PACz) contain a significantly smaller amount of FAI if the MeO-2PACz layer is washed with ethanol before the perovskite deposition. The reduced FAI content leads to a different morphology and alters the bromine to iodine ratio, impacting the solar cell performance and the band gap of the films.

We use two approaches, namely “seed layers” ⁴ and “loaded hole transport layers” ⁵, for tuning co-evaporated perovskite films through modifying the initial growth stage, yielding larger process windows, reduced sensitivity to substrate properties and improved process stability.

We study how the perovskite composition differs between planar and textured surfaces with random pyramids on silicon heterojunction bottom cells. Finally, fully textured perovskite/silicon tandem solar cells with ~30% PCE (certified) are demonstrated.

Our results illustrate multiple factors influencing the perovskite growth and highlight the potential of co-evaporation processes for the preparation of efficient perovskite/silicon tandem solar cells on textured substrates.

Perovskite-silicon tandem solar cell are highly promising candidates for the industrialization of perovskite based solar cells. We employ several evaporation techniques such as co-evaporation, sequential evaporation, and the hybrid route to fabricate the uniform perovskite films needed for such tandem devices. Each of these approaches has its own challenges and opportunities. In this talk I will share the progress we have made to gain understanding of the mechanisms, crystallization behavior, and the influencing factors that govern the evaporation process. We were able to use these perovskite films in tandem solar cells reaching efficiencies above 31% on 1 cm² cells. Nevertheless it is not fully clear which evaporation approach is best suited for the scaling of perovskite-silicon tandems. Full evaporation approaches like co-evaporation or sequential evaporation have the advantage of having all steps in the same atmosphere, whereas hybrid evaporation / wet-chemical approaches avoid the challenging evaporation of organic halides but introduce additional complexity in the wet-chemical 2^nd step. In this talk I will compare what we have so far learned about all three approaches and will highlight key gaps to be addressed in the future.    

Within the research field of single- and multi-junction solar cells the metal halide perovskites are becoming increasingly appealing for commercialization, primarily due to their record power conversion efficiencies (PCEs). These PCE records are a result of several attractive qualities inherent to perovskites. One of these qualities is its compositional flexibility. A large variety of hybrid compositions, including organic and inorganic precursors and dopants, have led to a range of different band gap semiconductors with increased optoelectronic performance and stability. These improved properties come accompanied by an increased production complexity due to, among other things, the different volatility and solubility of its components.

The most explored production method, with the largest number of publications and the current PCE record, is solution-based deposition. Solution-based deposition techniques allow relativity easy inclusion of compatible precursors without the use of specialized or expensive equipment.

Another method, vapor-phase deposition, in particular co-sublimation of perovskite precursors, has also been successfully shown to lead to high quality perovskite films and solar cells. These perovskite films have the added benefit of the absence of trace amounts of solvents or the necessity of all precursors to be soluble in the same solvent. Additionally, this method can be more seamlessly transferred into already established production lines in the semiconductor industry. However, the number of precursors that can controllably and reproducibly be co-sublimed is limited. Moreover, relatively slow sublimation rates are required to maintain control over the perovskite stoichiometry, resulting in deposition times in the order of hours.

In this presentation a novel gas flow assisted flash evaporation (GFAFE) method will be introduced. This method allows for a high degree of freedom in precursors as the system flash evaporates all tested materials without losing its stoichiometry. The most important novelty of this system is its unprecedented deposition speed of over 1.0 µm/min. This can be used for complex, complete perovskite deposition from a single source, or in a sequential manner, where it tackles the specific need for high deposition speed of organic components as formulated from industry.

This system has initially been used for the production of perovskite solar cells which has led to proof-of-concept devices made with both single source and sequentially deposited perovskites. The performance of the solar cells is enhanced by the use of dopants and additives from the singular powder source.

The system uses a 10x10 cm² sample holder and is designed with upscaling for wafer sized devices in mind by using a showerhead design to homogeneously expand the gas flow. The system uses a powder reservoir that holds up to 5 grams of powder, which has been found to be sufficient for over 10 production batches, making the total production throughput very high, especially for a lab-scale setup. The combination of the qualities above enables us to make high quality, complex perovskites, faster than with conventional methods while decreasing the complexity of perovskite deposition procedure.

Halide perovskites exhibit superior optoelectronic properties and hold great promise in advanced optoelectronic devices. However, they lack precise thickness and interfacial structure control in heterojunctions, critical for modular multilayer architectures such as multiple quantum wells. Vapor-phase deposition method have great advantage in thickness control for perovskites. As a dry processing method, it is also capable of preserving the interfacial structure especially for perovskite heterojunctions. In this work, we demonstrate layer-by-layer heteroepitaxial growth CsPbBr₃ deposition on 2D perovskites by thermal evaporation. The strict heteroepitaxial templating, revealed by reciprocal space maps, significantly improved perovskite film quality by reducing energetic disorder, enhancing photoluminescence quantum yield, and improving carrier transport. Angstrom-level thickness control and sub-Angstrom smooth layers enable quantum-confined photoluminescence of CsPbBr₃ from monolayer, bilayer, and through to bulk.

Futhermore, we demonstrate that the interfacial structure could be tuned by deposition parameters, controls the electronic structure between type-I and type-II heterojunctions. The interfacial-tunable band offset shift could reach much higher values than what has been achieved in III-V semiconductors. Electron transfer from CsPbBr₃ to 2D perovskite for the type-II heterojunction results in observation of charge separation and delayed electron-hole recombination, which is absent in the type-I heterojunction. Our results show that the precise quantum confinement control and large band offset tunability unlock perovskite heterojunctions as platforms for scalable, low-cost modular quantum and superlattice-based optoelectronic applications.

Halide perovskites deposited by physical or chemical vapour deposition techniques are significantly less explored than their solution-processed counterparts. In this talk I will cover our contributions to understanding the hybrid vapour-vapour deposition approach, where one first needs to understand the caveats of the first step (i.e., strong vacuum ~10^-6 mbar PVD of the inorganic template) [1], before studying the transformation of the inorganic template into a 3D perovskite upon exposure to organohalide vapours (e.g., formamidinium iodide or methylammonium bromide) in a close-space-sublimation setting in weak vacuum conditions ~ 0.1 mbar [2]. By tracking X-ray diffraction and photoluminescence in-situ, we identify four different regimes of the transformation and expose critical, previously overlooked, external factors that determine the optoelectronic quality of the perovskite thin-films. In the final part, I will present a new degradation-inversion approach in which we synthesize formamidinium iodide (FAI) in-situ in ambient conditions directly near the inorganic template by applying the principle of Le Chatelier and reacting the as-formed FAI with the inorganic template to form the first proof-of-principle devices with this fast, new, low-cost synthesis approach. [3] 

Metal halide perovskite-based solar cells have transitioned over the past decade from a research innovation to a commercially viable technology. Vapor phase deposition methods for perovskite solar cells (PSCs) are gaining increasing interest in both academia and industry, holding great promise for the commercialization of perovskite-based photovoltaics. Despite these advances, current laboratory-scale vapor-based processes are limited by low deposition rates, creating a significant barrier to achieving the production throughput necessary for industrial adoption. In this contribution, we report on the latest developments and investigations by the team at the Karlsruhe Institute of Technology on advancing vapor phase deposition processes for perovskite photovoltaics. Our research aims to provide practical guidelines to support the transition from research-scale methods to scalable and cost-effective manufacturing.

In 2024, we presented a perspective that conveys a balanced viewpoint from both industry and academia on the prospects of vapor phase deposition of perovskite photovoltaics as part of a large global consortium [1]. Building on this perspective, earlier this year we reported on strategies for achieving high-throughput vapor deposition processes for the industrialization of PSC fabrication in collaboration with industry partners [2]. The latter study addresses the critical challenges of scaling vapor-based processes by evaluating the thermal stability of perovskite precursors, analyzing deposition modes, and conceptualizing a linear sublimation source for production throughput analysis. Together, these studies offer a comprehensive framework for advancing vapor phase deposition methods and accelerating the commercialization of perovskite-based photovoltaic technologies.

Furthermore, this contribution will report on the latest developments of our team on vapor-based perovskite absorber layers and their application in perovskite single junction solar cells as well as perovskite/Si tandem solar cells. This encompasses our recent study on sequential evaporation of inverted formamidinium lead triiodide (FAPI) PSCs that highlights the impact of different hole transport layers (HTLs) on the crystallization and film formation of FAPI perovskite thin films [3]. This study reveals significant changes in PbI₂ crystal orientation depending on the HTL, which in turn affects the subsequent conversion and crystallization processes. We achieve power conversion efficiencies (PCEs) of more than 17%, the highest reported for fully vacuum-processed pure FAPI PSCs in the p-i-n architecture.

1. Abzieher, T. et al. Vapor phase deposition of perovskite photovoltaics: short track to commercialization? Energy Environ. Sci. (2024) doi:10.1039/D3EE03273F.

2. Petry, J. et al. Industrialization of perovskite solar cell fabrication: strategies to achieve high-throughput vapor deposition processes. EES Sol. 1, 404–418 (2025).

3. Diercks, A. et al. Sequential Evaporation of Inverted FAPbI3 Perovskite Solar Cells - Impact of Substrate on Crystallization and Film Formation. ACS Energy Lett. 1165–1173 (2025) doi:10.1021/acsenergylett.4c03315.

We systematically investigated the influence of various substrates, such as PTAA, NiOx, PEDOT:PSS and the self-assembled monolayer MeO-2PACz, on the formation of thermally evaporated FAPbI1Br2 perovskite films. Characterization techniques including X-ray photoelectron spectroscopy (XPS), ultraviolet photoelectron spectroscopy (UPS), scanning electron microscopy (SEM) were utilized to analyze the surface properties and the morphology of the evaporated films. Bulk properties, like crystal structures and optical absorption characteristics of the films, are investigated using X-ray diffraction (XRD) and UV-vis spectroscopy.

Our results reveal distinct substrate-dependent effects on the formation and composition of resulting perovskite thin films. While on PTAA the mixed halide perovskite readily forms, on PEDOT:PSS and NiOx we initially only observe FAPbI3 formation – indicating that a bromide species must become volatile. Only after the deposition of ~30 nm of the perovskite precursors the substrate influence diminishes and bromide becomes incorporated forming the desired FAPbI1Br2. Maybe the most surprising results were obtained for the MeO-2PACz substrates, which are commonly (and successfully) employed in solution processed perovskite devices. Here, no bromide incorporation was seen for any layer thickness up to 200 nm deposition and overall the perovskite formation was hindered. This indicates that SAM layers might present a challenge in thermal evaporation and that substrates can affect the film growth not only close to the interface, but also extending to device-relevant thicknesses.

Vapor phase deposition offers a scalable pathway for translating perovskite solar cell fabrication from laboratory-scale to continuous industrial manufacturing. Successful scale-up necessitates precise control of process parameters as well as robust process repeatability and reproducibility. Controlling the sublimation behavior and deposition kinetics of organic precursor materials remains a significant challenge in vapor-based fabrication processes. While the sublimation properties of methylammonium iodide have been extensively studied, [1-5] analogous studies on formamidinium iodide (FAI) remain scarce. [6,7] This study is the first to systematically investigate how the FAI precursor particle size and the geometry of the sublimation crucible influence the directionality of the emitted vapor flux during deposition.

We demonstrate that conical crucibles lead to beam-focusing of the emitted vapor flux, while cylindrical crucibles exhibit a broader and less directional emission profile. Furthermore, in the case of conical crucibles we observe a strong impact of the FAI particle size on the directionality of the vapor flux, whereas this effect does not occur for cylindrical crucibles. We show that such variations in emission profiles significantly impact the deposited film thickness uniformity, especially with respect to lateral source-to-substrate distance. The inconsistent particle size distribution found in commercial FAI powders in combination/in conjunction with conical crucibles therefore represents a significant challenge for reproducible and repeatable sublimation processes under laboratory conditions. Finally, analysis of commonly used inorganic materials reveals that effusion characteristics are highly material-dependent, adding to the complexity of multi-material deposition processes.

Our findings highlight the critical role of precursor particle size, choice of crucible geometry and spatial arrangement within the vacuum chamber to improve film homogeneity and ensure repeatability and reproducibility of laboratory-scale sublimation processes.

The thermal decomposition of formamidinium iodide (FAI) during vacuum deposition poses a major challenge in the fabrication of high-quality perovskite thin films. In this study, we present a strategy to suppress FAI decomposition by incorporating a minor additive during precursor preparation. The additive modulates the chemical environment, effectively suppressing the formation of volatile decomposition byproducts during deposition, and thereby stabilizes the working pressure during the process. In addition, this approach enabled the formation of wide-bandgap perovskite films with high crystallinity. As a result, single-junction devices based on 1.68 eV bandgap perovskites achieved power conversion efficiency (PCE) of 20.5%, while silicon/perovskite tandem devices exhibited efficiencies of approximately 30%. Our findings provide a practical method for mitigating the thermal instability of organic halide precursors during vacuum deposition. This strategy is expected to advance the advancement of high-efficiency perovskite solar cell technologies using vacuum deposition and facilitate their further commercialization.

In the last ten years, tandem solar cells based on perovskite (PK) materials have shown promising results, surpassing the theoretical limits of single junction Silicon (Si) solar cells. Even though PK/Si tandem solar cells appear capable of achieving >35% of power conversion efficiency, many challenges need to be overcome in order to upscale the PV devices from 1 cm² laboratory scale to larger areas. With that perspective, vapor deposition of the absorber layer seems promising in order to elaborate a conformal and high quality perovskite on top of textured industrial silicon wafers. In the literature, the main industrially compatible techniques to grow the PK layer can be divided into two axes: i) PK deposition by full vapor deposition techniques and ii) PK deposition by hybrid processes (mix of dry and wet processes).

In a first part, we propose pulsed laser deposition (PLD) as full vacuum scalable method to fabricate uniform black phase inorganic perovskites and charge transport layers on 707 cm² substrates (> G12 area). We first developed several PLD-grown contact layers (ITO, SnO2 and NiOx), and then demonstrated the deposition of CsPbI₂Br and CsPbI₃ on 300 mm wafers, exhibiting PL peaks at λ=648 nm (~1.91 eV) and 700 nm (~1.77 eV), with FWHM of 28.9 and 36.5 nm, respectively. The films show excellent uniformity: 0.8% in thickness and 0.2% in PL wavelength. Final device integration and performance measurements are currently underway. Looking ahead, we aim to scale-up the growth rate from the current 8 nm/min at 20 Hz to 125 nm/min using an industrial PLD system operating at 300 Hz, further highlighting PLD as a promising route for large-scale PSC manufacturing.

In a second part, we focus on a hybrid deposition process combining thermal co-evaporation, close space sublimation (CSS) and solvent step for the elaboration of organic-inorganic perovskite layer. Using different characterization techniques (X-ray Diffraction, X-Ray Photoelectron Spectroscopy, Scanning Electron Microscope, etc.), we firstly investigate the structural and chemical properties of the inorganic scaffold. The goal of this study is to understand how the homogeneity, the porosity and the composition of the first layer affect the growth of the final PK film. Simultaneously, we examine the key factors that influence the crystallization mechanism during the second wet step.

Final part will be dedicated to the stability assessment of perovskite/silicon tandem architecture under outdoor and accelerated ageing conditions.