Pow(d)ering up: FAPI perovskite nanopowders for air-processed blade coated perovskite solar modules

Pow(d)ering up: FAPI perovskite nanopowders for air-processed blade coated perovskite solar modules† Maurizio Stefanelli, a Muhammed P. U. Haris, b Luigi Vesce, acd Luigi A. Castriotta, a Hafez Nikbakht, a Fabio Matteocci, a Samrana Kazim, bef Alessandro Triolo, g Shahzada Ahmad * bf and Aldo Di Carlo * ag Formamidinium lead iodide (FAPI) is the most prominent perovskite material utilized in the fabrication of singlejunction perovskite solar cells. However, the cubic a-phase perovskite is difficult to retain in precursor solutions for extended periods due to thermodynamic instability, which promotes the formation of the yellow d-phase. In this study, we demonstrate the fabrication of solar cells and modules from FAPI powders synthesised using a single-step method with a non-hazardous solvent and routine purity grade lead iodide. The presynthesised a-FAPI and CsFAPI powders demonstrated considerable potential for scalability and reproducibility. It was observed that similar efficiencies were achieved in small-area cells and mini-modules fabricated using an industrially adaptable blade coating process conducted in an open environment. The enhanced solution rheology and the meticulous control of stoichiometry result in an oriented and less strained crystal lattice, thereby demonstrating superior reproducibility and stability of the perovskite prepared from pre-synthesised powder in comparison to the one produced from high-purity precursor. The methodology developed offers a scalable and cost-effective approach to the production of highperformance and stable perovskite solar modules, with efficiencies reaching 18.5% on a 12.15 cm 2 active area module and a T95 above 1200 h in shelf life stability at 30% RH in a UV-filtered environment. Broader context Halide Perovskite (PVSK) Photovoltaics (PV) are moving from laboratory to pilot/production lines. The success of this transition is based on the use of stable formulation of PVSK and scalable coating techniques preferably in ambient air. Despite big efforts in the scaling up of PVSK technology in the literature, few works faced module fabrication entirely in ambient air. Furthermore, high purity raw materials for PVSK precursor solution and manual handling of them are critical points for the fabrication costs and the reproducibility of the stock solutions. In this work, we show the scaling up of PSC technology by a double step meniscus coating process out of a glovebox with green antisolvent quenching. The entire stack till HTL (nip conguration) is made in ambient air and with scalable techniques. In addition, we make use of pre-synthesised cesium-containg FAPI powder obtained by a single step synthesis with green solvent at room temperature and with low-cost and purity lead iodide. According to our knowledge, this work presents the rst FAPI powder made from low-cost lead iodide employed in a fully ambient air fabrication method. Introduction The growth of single junction perovskite-based photovoltaics is being driven by formamidinium lead iodide (FAPI) based perovskites. The high carrier mobility and optimal bandgap, as predicted by the Schottky–Queisser limit, are reected in exceptional efficiencies above 26% for both p-i-n and n-i-p type solar cell congurations. 1–3 In comparison to other halide perovskites, the absence of bromine ions and the use of methylammonium (MA + ) cation only for cubic a-phase stabilisation (e.g. MACl) contribute to the enhanced stability of the lm. The stability of the lm is considerably improved due to the absence of phase segregation and the higher temperature stability of the formamidiunium cation (FA + ) for MA + . 4,5 The general route for a CHOSE, Department of Electronic Engineering, University of Rome “Tor Vergata”, 00133 Rome, Italy. E-mail: [email protected] b BCMaterials - Basque Center for Materials, Applications and Nanostructures, Martina Casiano, UPV/EHU Science Park, Leioa, 48940, Spain. E-mail: shahzada.ahmad@ bcmaterials.net c Department of Biomedicine and Prevention, University of Rome “Tor Vergata”, 00133 Rome, Italy d Department of Physics, University of Rome “Tor Vergata”, 00133 Rome, Italy e Materials Physics Center, CSIC-UPV/EHU, Paseo Manuel de Lardizabal 5, 20018 Donostia-San Sebastian, Spain f IKERBASQUE, Basque Foundation for Science, Bilbao, 48009, Spain g ISM-CNR, Consiglio Nazionale delle Ricerche, 00133 Rome, Italy †Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5el00032g Cite this: EES Sol.,2025,1,543 Received 13th March 2025 Accepted 28th June 2025 DOI: 10.1039/d5el00032g rsc.li/EESSolar © 2025 The Author(s). Published by the Royal Society of Chemistry EES Sol.,2025,1,543–553 | 543 EES Solar PAPER Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online View Journal | View Issue FAPI or FA-rich perovskite deposition requires high-purity precursors (e.g. PbI 2 /FAI) dissolved in different solvents depending on whether the anti-solvent or the double step method is used. 6–10 The anti-solvent or double step method is a conventional approach to preparing perovskite solutions. However, errors in the precise weighing of precursor materials may result in the formation of non-stoichiometric species (cations, halides) in the solution, leading to non-radiative losses and reduced efficiencies. 11,12 Additionally, the rheological properties of the solution are a critical consideration. The colloids in perovskite precursors consist of a series of iodoplumbate complexes [PbIn] (n−2) as building blocks for large-size self-assembled lead polyiodides frameworks, which are linked by Pb–I–Pb corner-sharing coordination. 13 It has been observed that, in the standard solution-making route, colloids are much smaller than perovskite solutions prepared from presynthetized powders. The larger lead polyiodides frameworks in the latter case led to a preferable and oriented crystallisation of the perovskite layer. 14 For these reasons, pre-synthesized pure a-FAPI or CsFAPI powders have demonstrated high potential in ensuring device stability and reproducibility as well as efficiency. 15,16 The synthesis can be carried out in different solvent media, including 2-methoxyethanol, acetonitrile, or even in an aqueous solution with high yield (>90%). 1,17–20 Aer the precipitation step, the powder is ltered, annealed at 150 °C, washed with orthogonal solvents (i.e. diethyl ether, chlorobenzene) and kept in a vacuum atmosphere to extract the solvent traces. The synthesised powder can be processed in a typical solvent mixture (DMF/DMSO, 2-ME/NMP) and employed as a perovskite precursor ink. It has been demonstrated that perovskite crystals/powders are superior for exact-stoichiometry perovskite lms with lower defects, resulting in highly uniform and efficient perovskite solar cells exceeding 24% PCE. 21 Furthermore, the precursor solution obtained from the powder exhibits a longer shelf-life than the standard one. 15 The precise control of the stoichiometry of the powder methodology is reected in higher reproducibility and performance compared to the raw precursor's counterpart, thereby demonstrating the benets of obtaining FAPI powders from low-purity precursors (PbI 2 ) without performance loss. 22,23 This approach to FAPI synthesis, involving batch production from low-purity precursors without the need to weigh each one individually from batch to batch, is well suited to an industrial context thanks to its favourable balance of cost and efficiency. Furthermore, compositional engineering is paramount to push stability, and cesium incorporation can allow the formation of Cs-FAPI, which possesses higher stability according to the Goldschmidt tolerance factor. 24 Enhanced operational and moisture stability are demonstrated as well as milder temperature to a-phase transition with respect to the cesium-free counterpart. 18,19,21,24–27 In this context, pre-synthetized a-FAPI powder is emerging as a material for large area devices. Jiang et al. employed low-purity PbI 2 precursor material to synthesize FAPbI 3 in 2-methoxy ethanol solvent (2-ME) via inverse temperature crystallization and fabricated corresponding solar cells and mini-modules through spin-coating inside the glove box. The best-performing device measured an efficiency of 23.9% on small area devices and 19.5% on 13.44 cm 2 mini-module active area. 28 Moreover, the cells retained more than 95% of the initial power conversion efficiency (PCE) aer 400 h of storage in shelf-life conditions. Further, Tong et al. proposed a method to avoid impurities from powder synthesis using hydroiodide acid, formamidiunium acetate, lead iodide, and g-butyrolactone, and defect-less FAPI perovskite lms were fabricated through solubility management. The bestperforming spin-coated device displayed an efficiency of 23% and 16% for a small area and 22 cm 2 active area module, respectively. 17 Recently, Du et al. have emploied CsFAPI single crystals as perovskite precursors underlying the impact of colloids and volatile solvents for effective and homogeneous nucleation. The suppressed secondary nucleation and extended growth of crystals enable high throughput fabrication of full printed carbon-based devices in a controlled environment (RH < 5%) with an efficiency of 19.3% and 16.2% on a small area and 16.84 cm 2 active area mini-module, respectively. 29 FAPI powders are a promising material for module fabrication, but have only been used thus far in a controlled atmosphere within a glove box. There are currently no reports on the scaling up of this process to module size in ambient air. The objective of this work is to address this gap by demonstrating the fabrication of modules in ambient air using pre-synthesized powders. In order to achieve this objective, we have selected Cs x FA 1−x PbI 3 perovskite powder (CsFAPI-pw), with a cesium content (x) of less than 10%, which has been synthesised with one-pot synthesis in a non-hazardous solvent with a yield >93%. We compared powders obtained with low-purity lead iodide sources and those from conventional high-purity precursors (CsFAPI-hp) used for small-area cells and mini-module fabrication. Here, all the layers (except for gold) in both small area (0.5 cm 2 active area) and minimodules (12.5 cm 2 active area) devices were deposited in ambient air by blade coating technique, with a commendable efficiency for small area (0.5 cm 2 ) cells of 19.4% and 19.2% obtained from Cs x FA 1−x PbI 3 pre-synthetized powder and conventional precursor route, respectively. The perovskite lms obtained with the powder method exhibit superior performances in terms of efficiency, stability (ISOS-D1), and reproducibility. We then selected CsFAPI-pw for mini-module fabrication and the device measured 18.5% (less than 5% losses with respect to the small area cell) on a 12.5 cm 2 active area. Results and discussion The deposition of the perovskite layer by solution process is achieved by dissolving the precursors (e.g. FAI and PbI 2 )in a mixture of solvents (e.g. DMF/DMSO) and subsequently depositing them as a thin lm. Although there are reports of using the perovskite powder in the fabrication of perovskite solar cells (PSCs), the use of meniscus coating, such as blade coating, has not been explored in the context of fully ambient air device fabrication. 30,31 In this study, black CsFAPI powder (Fig. 1a and b) was used as a precursor material for perovskite with blade coating technique (Fig. 1c). 27 A conventional perovskite ink was also prepared from high-purity PbI 2 , CsI, and FAI for comparative purposes. 32 The rheology of the two solutions 544 |EES Sol.,2025,1,543–553 © 2025 The Author(s). Published by the Royal Society of Chemistry EES Solar Paper Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online was characterised using dynamic light scattering (DLS) measurement. DLS experiment have been performed to assess the average diameter size of the aggregates present in both CsFAPI-hp and CsFAPI-pw solutions. The population of the colloids in the precursor and powder solutions demonstrates that the size of the particles is inuenced by the starting precursor material. In the case of CsFAPI-hp solution, only small-size colloids (∼1 nm) were detected (Fig. 2a). In contrast, CsFAPI-pw solution exhibited a notable population of colloids in the order of 100 nm, constituting 20% of the total DLS intensity (Fig. 2b). The larger size of the colloids wich are attributed to the iodoplumbate complexes [PbI n ] (n−2) may have an impact on the lm crystallization process, acting as seeds during the fabrication step. 33–35 Besides the small colloids acting as nucleation sites and subsequently undergo a diffusion-controlled growth as described by the LaMer mechanism, 36 the larger colloids in the CsFAPI-pw case can drive oriented growth. 14 The discrepancy between the two population sizes can effectively alter the solution chemistry resulting in the formation of a higher-quality and less strained crystalline lm, Fig. 1 CsFAPI powder synthesis steps (a); black CsFAPI powder (b); process routine for CsFAPI perovskite deposition in ambient air with blade coating technique (c). Fig. 2 Particle size distribution for CsFAPI-hp (a) and CSFAPI-pw solution, and (b) measure by dynamic light scattering technique. © 2025 The Author(s). Published by the Royal Society of Chemistry EES Sol.,2025,1,543–553 | 545 Paper EES Solar Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online as conrmed also by XRD peak analysis. CsFAPI lms in both cases were fabricated via the blade coating technique in an ambient atmosphere with isopropanol green-solvent quenching followed by annealing at 160 °C for 10 min (Fig. 1b) as previously described. 37 The thin lms fabricated from perovskite powder and conventional precursors are labelled hereaer as CsFAPI-pw and CsFAPI-hp, respectively. To compare the quality of CsFAPI-pw to the CsFAPI-hp, we fabricated and characterized a small area (0.5 cm 2 a.a.) PSCs and lms coated on a 5 ×5cm 2 substrate, and were cut into four smaller pieces (2.5 ×2.5 cm 2 ) prior to the thermal evaporation of the back electrode. The morphological analysis (Fig. 3) indicates that the CsFAPI-pw exhibits superior uniformity compared to the CsFAPI-hp. Micron-size grains are detected for CsFAPI-pw presenting a compact and pinholes-free (Fig. 3a and b) lm due to the different crystallization proles of nucleation rate and crystal growth. On the other hand, CsFAPI-hp shows smaller grains and the presence of unreacted PbI 2 (Fig. 3c and d) that can compromise the lm quality and stability over time. The X-ray diffraction (XRD) patterns for both CsFAPI lms (Fig. 3e) show peaks at 2q=14° and 28°, which are indicative of perovskite orientation along (00h) facets that have been identied as the desired orientation for solar cell applications. 38 It should be notice that lead iodide signal at 12.6° is still present for CsFAPI-pw even if there are not possible excess in the stoichiometry of the solution. The reason of that behave in the antisolvent quenching. Infact IPA it's a well know green antisolvent and recently integrated also as passivator additive in other antisolvent system (i.e. chlorobenzene) and eventually used as a washing step to provoke the 2D crystals reorganization on the surface of 3D perovskite. 39–41 In this work bare IPA is used as antisolvent and of curse a large excess of IPA in the antisolvent system can produce some lead iodide 39–42 that is treated and washed out with the passivation strategy with PEAI to form 2D perovskite and wash out the unreacted lead iodide as we already described in a previous work on passivation strategy. 43 While the intensity of diffraction peaks is higher for the CsFAPI-pw as compared to CsFAPI-hp suggesting an enhanced crystallinity as well as a larger lattice constant as indicated by the shitoward lower 2qin the diffractograms (Fig. 3e). In a previous work, we attributed the enhanced crystallinity of the powder-processed FAPbI 3 lms deposited by the spin coating technique, to their conventional counterpart to reduced microstrain. 18 To assess the impact of the microstrain on the blade-coated thin lms and to evaluate how it translates to large-scale fabrications, we performed a detailed analysis of the diffraction peaks employing the Williamson–Hall 21 equation bcos q¼kl Dþ43sin q(1) where b,l,k,D, and 3denote the full width at half-maximum of the diffraction peak, the wavelength of X-ray radiation, the Scherrer constant, the crystallite size, and the microstrain, respectively. We plotted bcos qas a function of 4 sin q, and the slope calculated from the linear tting was estimated as the microstrains (Fig. S1†). We found that, besides the enhanced crystallinity, the CsFAPI-pw lm displays a lower microstrain value of 6 ×10 −4 , while the conventional CsFAPI-hp counterpart shows a higher value of 6.54 ×10 −4 (Fig. S1†). The reduced microstrain will induce long-term stability and the efficiency of CsFAPI-pw devices due to the impact of microstrain on structural stability, defect generation, charge carrier dynamics, and local disorder near the band edges. 18,21,44,45 Fig. 3 SEM top view of CsFAPI-pw (a and b) and CsFAPI-hp (c and d) at lower and higher magnification, respectively; red circles in Fig. 2d highlight unreacted PbI 2 ; XRD diffraction pattern of CsFAPI-hp and CsFAPI-pw deposited by blade coating on bare FTO (e). 546 |EES Sol.,2025,1,543–553 © 2025 The Author(s). Published by the Royal Society of Chemistry EES Solar Paper Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online Photovoltaic performances & dynamic light scattering We evaluated the photovoltaic performances of the small area devices using a planar n-i-p architecture, composed of FTO/ SnO 2 /PVK/PEAI-passivation/Spiro-OMeTAD/Gold. Fig. 4a– d show the statistics on photovoltaic parameters for 15 devices (CsFAPI-pw and CsFAPI-hp). The mean value and the standard deviation of V oc ,J sc , FF, and PCE show a dramatic improvement in the reproducibility and efficiency of CsFAPI-pw with respect to CSFAPI-hp devices (Table 1). Moreover, the absence of unreacted species and the less strained crystal lattice improve Fig. 4 Statistics of photovoltaic parameters for CsFAPI-hp and CsFAPI-pw devices: V oc (a) J sc (b) FF (c) and PCE (d); J–Vcurves for the record devices of CsFAPI-hp and CsFAPI-pw respectively on 0.5 cm 2 active area (e); EQE of the best devices (f). © 2025 The Author(s). Published by the Royal Society of Chemistry EES Sol.,2025,1,543–553 | 547 Paper EES Solar Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online stability over time and robustness during the ambient air fabrication process. 22 The champion devices with 0.5 cm 2 active area for CsFAPI-pw showed open circuit voltage (V oc ) of 1.05 V, short-circuit current density (J sc ) of 23.07 mA cm −2 ,ll factor (FF) of 79.89%, and PCE of 19.40%. While, CsFAPI-hp displayed V oc of 1.05 V, J sc of 22.66 mA cm −2 , FF of 80.4%, and PCE of 19.03%. In both cases, the devices showed negligible hysteresis (Fig. 4e). Short circuit current is in agreement with integrated J sc in EQE measurement where CsFAPI-pw has a higher conversion in the wavelength range 500–820 nm (Fig. 4f). To investigate the improvement of average V oc in CsFAPI-pw concerning CsFAPI-hp and its relation with perovskite properties such as trap densities, we carried out space charge limit current (SCLC) and transient photoluminescence (TRPL) analyses to evaluate and compare the trap state densities of both perovskite lms. SCLC measurements were carried out with an electron-only device architecture, namely FTO/SnO 2 / PVK/PCBM/Ag. 46 The trap-lled limit (TFL) voltage (V TFL ), (Fig. 5a), was determined from the change of slope between the ohmic region (nz1) and the TFL region (n> 2). CsFAPI-pw showed signicantly reduced V TFL (0.233 V) for that of CsFAPIhp (0.442 V). This proves the better quality of the lm also in the bulk and it's reected in the higher overall photovoltaic Table 1 Mean values and standard deviation over 15 devices fabricated with CsFAPI-pw and CsFAPI-hp of photovoltaic properties for the fabricated devices V oc [V] J sc [mA cm −2 ] Fill factor [%] PCE [%] CsFAPI-pw 1.048 0.004 23.000 0.080 79.860 0.079 19.141 0.008 CsFAPI-hp 1.040 0.007 22.800 0.248 79.790 0.902 18.304 0.061 Fig. 5 SCLC analysis (a), transient photoluminescence decay (b), UV-Visible spectra and PL measurements of PVK films (c) and ISOS-D1 at 30% RH in UV-filtered box for CsFAPI-hp and CsFAPI-pw (d), respectively. 548 |EES Sol.,2025,1,543–553 © 2025 The Author(s). Published by the Royal Society of Chemistry EES Solar Paper Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online performances of the CsFAPI-pw devices. The trap state density (N trap ) can be estimated with the following equation: 47 Ntraps ¼2VTFL303 eL2(2) where N traps ,V TFL ,3 0 ,3,e,Lare trap-state density, the trap-led limit voltage, vacuum permittivity (3 0 =8.8 ×10 −12 Fm −1 ), the relative dielectric constant of perovskite (with reference to 62.23), electron charge (e=1.6 ×10 −19 C), the lm thickness of perovskite (800 nm), respectively. Since perovskites with the same stoichiometry and thickness have the same 3 0 ,3,e, and L, the magnitude of N trap is only determined by V TFL .N trap determinations were 4.75 ×10 15 cm −3 for CsFAPI-hp and 2.5 ×10 15 cm −3 for CsFAPI-pw, thus further endorsing the potential of pre-synthetized powder to obtain thin-lm with less trap state density and low hysteresis in PSCs. 48 Further, TRPL characterization was performed on both perovskite lms (sample stack Glass/FTO/PVK) with 405 nm pulsed diode laser, and a double exponential t was applied to determine their average decay times, as reported in Table 2. DT TðtÞ¼A1e  t s1þA2e  t s2þA0(3) save ¼A1s12þA2s22 A1s1þA2s2 (4) Assuming that s ave is affected by the trap density and the corresponding lifetime of the free carriers generated into the lm, 49 is in agreement with SCLC analysis and with the higher peak of photoluminescence (performed on Glass/FTO/PVK sample, excitation wavelength 385 nm) on CsFAPI-pw with respect to the CsFAPI-hp (Fig. 5a–c). Furthermore, we assess the shelf-life stability of both nonencapsulated perovskite devices performing ISOS-D1 at 30% RH in a UV-ltered environment. 50 CsFAPI-hp-based perovskite shows a signicant efficiency drop (Fig. 5d) in the rst 100 hours and presents a large degradation trend starting from 600 hours with a reduction of 20% of initial efficiency (T 80 )of 1100 h. On the other hand, CsFAPI-pw based perovskite, which is characterized by a smooth and compact lm with low defects and low content of unreacted species on its surface, can resist under 30% RH for more than 1200 hours, losing only less than 5% of its initial efficiency. To further test the stability of the perovskite layer we conducted ISOS-L1 stability test in ambient air at 35/40 °C on encapsulated devices with UV-curable resin. The stack is (FTO/ SnO 2 /PVK/PEAI/Spiro-OMeTAD/Au). As shown in the inset of Fig. 5d, the rst drop in efficiency is visible for both perovskite formulations due to the burn-in effect and encapsulation process, which slightly decrease efficiency within the rst 24 hours of light exposure. 51,52 The CsFAPI-hp device reach the T 80 only aer 128 h of light exposure, as result of the poor crystallinity of the PVK lm and the elevated lead iodide excess that affect the operational stability. 53,54 In the CsFAPI-pw case, the stability is more than doubled (T 80 =317 h) respect to the CsFAPI-hp devices, highlighting the improved operational stability of the layer. Moreover, aer 1000 h of test the CsFAPIpw device lose a bit less than 40% of the initial efficiency where instead the CsFAPI-hp device efficiency dropped more than 80% respect the initial efficiency. PVK layer obtained from nano powder still suffers from the pinholes (smaller and in lower quantity, but stil present) into the PVK lm as shown in the SEM top view images (Fig. 3a and b). The possible reasons of that can be found in the massive quantity of IPA used onto the supersaturated PVK lm in the deposition process that extracts part of the formamidiunium that is highly soluble in alcoholic solvents. 39,40,42 Should be pointed out that no additives are used in the PVK layer except for MACl that helps to stabilize the alpha phase in the deposition process. 55 The stability obtained is in line with our previous work on PVK lm obtained from powders with similar architecture 18 but further improvements with antisolvent engineering and additives selection are needed to reach the optimal operational stability. Based on the remarkable results obtained on small-area cells, we scale up the fabrication process used in small-area cells to realize mini-modules. Considering the improved reproducibility and stability for CsFAPI-hp, we focused only on the CsFAPI-pw perovskite for such a study. The mini-module has a total active area of 12.5 cm 2 and a geometrical ll factor (GFF) of 93%, giving an aperture area of 13.44 cm 2 .P1–P2–P3 laser scribing scheme 56 was adopted during the fabrication process to guarantee the series Zconnection among the 5 subcells (2.5 cm 2 active area each) of the mini-module. The SEM/ EDX and the scheme of the P1–P2–P3 laser process are presented (Fig. 6a and b). The relative small signal of lead, carbon, iodine, and the strong signal from tin and oxygen inside P2/P3 etching lines from the EDX mapping conrmed the good quality of the laser process in the removal of the stack without damaging TCO. The best-performing mini-module (Fig. 6c) displayed a V oc of 1064 mV for each cell, 52 mA of current, 76% FF, and a PCE of 18,46% (17.17% on aperture area) in the reverse scan. For the forward scan, the PCE was slightly improved to 18.5% (17.2%). Negligible hysteresis and high reproducibility achieved in this process attest to the potential of our powder engineering strategy for module fabrication. The optimization of the P2 laser process contributed to the highquality contact with the TCO (75% ll factor of), negligible hysteresis, and less than 5% efficiency difference with small area devices. 43,48,57,58 We made preliminary cost estimations for the CsFAPI-pw and CsFAPI-hp, and the data referred to the material usage are normalized to 1 g of perovskite. In the case of CsFAPI-pw, the costs of the solvents are considered for the reaction (acetonitrile, ACN) and for washing the synthesized product Table 2 TRPL double exponential fit for decay of CsFAPI-pw and CsFAPI-hp perovskite films Sample A 1 s 1 [ms] A 2 s 2 [ms] s ave [ms] A 0 CsFAPI-hp 0.955 0.947 0.039 9.706 1.29 0.027 CsFAPI-pw 0.845 1.043 0.070 75.168 6.77 0.033 © 2025 The Author(s). Published by the Royal Society of Chemistry EES Sol.,2025,1,543–553 | 549 Paper EES Solar Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online (diethyl ether) (Tables S1 and S2†). With a rough estimation of the costs and a yield set at 100% for CsFAPI-hp as estimated, considering the “direct synthesis”during the deposition, and 93% for CsFAPI-pw (empirically determined by weighing the converted black FAPI powder and extracting the moles of product useful for the yield calculation moles product / moles reagents ×100), the overall costs for 1 g of perovskites is one fourth for CsFAPI-pw compared to the cost of CsFAPI-hp. This demonstrates the economic sustainability of the FAPI powder made from low-purity lead iodide, which can play a game-changer role in the manufacturing costs. Further, CsFAPI-pw can be produced in a batch production eliminating the need for multiple powder weighing and possible manual errors that can occur in the process and affect the stoichiometry of the solution and will promote reproducibility. Conclusion In conclusion, the utilisation of pre-synthetised Cs-FAPbI 3 powder in conjunction with a scalable deposition technique, blade coating, under ambient air conditions, has been demonstrated for the fabrication of both small-area cells and modules. The photophysical characterisation of the resulting lms demonstrates the superior quality of the perovskite layers produced from the pre-synthetized powder source. The fabricated small-area cells and mini-modules exhibit low hysteresis and efficiency losses of less than 5% when scaling from 0.5 cm 2 to 12.5 cm 2 active areas, along with a shelf-life stability exceeding 1200 hours, retaining over 95% of their initial efficiency. The presented methodology demonstrates that the use of high-purity lead iodide is not a prerequisite for the Fig. 6 Minimodule Image and EDX analysis of P1–P2–P3 scribing process (scale bar is 250 mm) (a); cross view of P1/P2/P3 process in stack and dead area (b) I–Vcurve of the best minimodule made with CsFAPI-pw precursor (c). 550 |EES Sol.,2025,1,543–553 © 2025 The Author(s). Published by the Royal Society of Chemistry EES Solar Paper Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online fabrication of stable and high-efficiency devices, thereby offering a signicant reduction in production costs. The next crucial phase for perovskite solar cell technology is the industrial production of solar modules. In this context, it is essential to prioritise the optimisation of fabrication costs, economic viability and environmental sustainability, alongside the enhancement of device performance. This work represents the inaugural demonstration of the utilisation of low-cost materials, batch production, green solvents (for synthesis and for the antisolvent step), and scalable techniques under ambient conditions, all of which are indispensable elements for the industrialisation of perovskite technology. The powder engineering approach, which employs ambient blade coating, offers a promising pathway to addressing these challenges. The reliability and sustainability of the processes are further reinforced through the use of green solvents for perovskite precursor synthesis (ACN) and antisolvent quenching (Isoprpanol, IPA), as well as the use of low-purity precursors (lead iodide) and ambient air fabrication methods. Data availability Data for this article, including EQE, UV-Vis, JV and DLS measurements are available at ZENODO repository at (https:// doi.org/10.5281/zenodo.14164839). Conflicts of interest There are no conicts to declare. Acknowledgements M. P. U. H. acknowledges the support of the H2020 Project “VIPERLAB”of European Union (no. 101006715). SA acknowledges support for the European Research Council Consolidator grant [MOLEMAT, 726360]. And also the Spanish Ministry of Science and Innovation INTERACTION {PID2021-129085OBI00} and CNS2022-136171 funded by MCIN/AEI/10.13039/ 501100011033 and by the “European Union NextGenerationEU/PRTR”is also acknowledged. L. V. was funded by the European Union's Horizon Europe Programme, through a FET Proactive research and innovation action under grant agreement No. 101084124 (DIAMOND). A. T. acknowledges the Project ECS00000024 “Ecosistemi dell’Innovazione”—Rome Technopole of the Italian Ministry of University and Research, public call no. 3277, PNRR—Mission 4, Component 2, Investment 1.5, nanced by the European Union, Next GenerationEU. A. D. C. acknowledges the Project “Network 4 Energy Sustainable Transition—NEST”, Spoke 1, Project code PE0000021, funded under the National Recovery and Resilience Plan (NRRP), Mission 4, Component 2, Investment 1.3—Call for tender No. 1561 of 11.10.2022 of Ministero dell'Universit` ae della Ricerca (MUR); funded by the European Union—NextGenerationEU. This project has received funding from the European Union's Framework Programme for Research and Innovation Horizon Europe (2021–2027) under the Marie Sklodowska-Curie Grant Agreement No. 101068387 “EFESO”. Notes and references 1 J. Park, et al., Controlled growth of perovskite layers with volatile alkylammonium chlorides, Nature, 2023, 616, 724– 730, DOI: 10.1038/s41586-023-05825-y. 2 Q. Jiang, et al., Surface reaction for efficient and stable inverted perovskite solar cells, Nature, 2022, 611(7935), 278–283, DOI: 10.1038/s41586-022-05268-x. 3 M. A. Green, E. D. Dunlop, J. Hohl-Ebinger, M. Yoshita, N. Kopidakis and X. Hao, Solar cell efficiency tables (Version 58), Prog. Photovolt.: Res. Appl., 2021, 29(7), 657– 667, DOI: 10.1002/pip.3444. 4 H. Choe, D. Jeon, S. J. Lee and J. Cho, Mixed or Segregated: Toward Efficient and Stable Mixed Halide Perovskite-Based Devices, ACS Omega, 2021, 6(38), 24304–24315, DOI: 10.1021/acsomega.1c03714. 5 T. T. Ava, A. Al Mamun, S. Marsillac and G. Namkoong, A review: thermal stability of methylammonium lead halide based perovskite solar cells, Appl. Sci., 2019, 9(1), 188, DOI: 10.3390/app9010188. 6 S. Yang, et al., Thiocyanate assisted performance enhancement of formamidinium based planar perovskite solar cells through a single one-step solution process, J. Mater. Chem. A, 2016, 4(24), 9430–9436, DOI: 10.1039/ c6ta02999j. 7 J. Duan, et al., Planar perovskite FAxMA1-xPbI3 solar cell by two-step deposition method in air ambient, Opt. Mater., 2018, 85,55–60, DOI: 10.1016/j.optmat.2018.07.072. 8 T. Zhang, et al., Spontaneous low-temperature crystallization of a-FAPbI3 for highly efficient perovskite solar cells, Sci. Bull., 2019, 64(21), 1608–1616, DOI: 10.1016/ j.scib.2019.08.029. 9 J. Du, et al., Crystallization Control of MethylammoniumFree Perovskite in Two-Step Deposited Printable TripleMesoscopic Solar Cells, Sol. RRL, 2020, 4, 2000455, DOI: 10.1002/solr.202000455. 10 J. W. Lee and N. G. Park, Two-step deposition method for high-efficiency perovskite solar cells, MRS Bull., 2015, 40(8), 654–659, DOI: 10.1557/mrs.2015.166. 11 P. Fassl, et al., Fractional deviations in precursor stoichiometry dictate the properties, performance and stability of perovskite photovoltaic devices, Energy Environ. Sci., 2018, 11(12), 3380–3391, DOI: 10.1039/c8ee01136b. 12 Z. Yuan, et al., Unveiling the synergistic effect of precursor stoichiometry and interfacial reactions for perovskite lightemitting diodes, Nat. Commun., 2019, 10(1), 1–9, DOI: 10.1038/s41467-019-10612-3. 13 J. Yang, E. L. Lim, L. Tan and Z. Wei, Ink Engineering in Blade-Coating Large-Area Perovskite Solar Cells, Adv. Energy Mater., 2022, 12(28), 1–25, DOI: 10.1002/ aenm.202200975. 14 Y. Zhang, et al., Mechanochemistry Advances HighPerformance Perovskite Solar Cells, Adv. Mater., 2022, 34(6), 1–10, DOI: 10.1002/adma.202107420. 15 G. S. Shin, Y. Zhang and N. G. Park, Stability of Precursor Solution for Perovskite Solar Cell: Mixture (FAI + PbI2) © 2025 The Author(s). Published by the Royal Society of Chemistry EES Sol.,2025,1,543–553 | 551 Paper EES Solar Open Access Article. Published on 04 July 2025. Downloaded on 11/15/2025 5:24:05 PM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online