Open research questions in Perovskite Materials and Applications

What the literature leaves open

• While [6,6]‐phenyl‐C 61 ‐butyric acid methyl ester (PCBM)/bathocuproine (BCP) is widely used as an electron transport stack in inverted perovskite solar cells (PSCs), its interfacial properties remain insufficiently studied.

  Interface Passivation and Morphology Regulation of Fullerene‐Based Perovskite Solar Cells for Enhancing Device Performance by phenethylammonium iodide · 2026 · DOI

• While widely applicable across diverse perovskite systems, the full potential of spectroscopic techniques remains underutilized, requiring further development to drive progress in perovskite photovoltaic technology. Current spectroscopic techniques face three major challenges that limit their effectiveness. First, complex setups and bulky equipment hinder their applications in dynamic or sensitive environments. For example, highly reactive Sn-based perovskites require inert conditions to avoid oxidation, making open-environment testing impractical. Miniaturization and integration of these systems are therefore critical for flexible testing. Second, nondestructive testing is a key area for improvement. Many optical techniques use lasers that can damage fragile films, such as by 14 Mater.

  Steady-state and time-resolved spectroscopic techniques for investigating defect states in perovskite photovoltaics · 2026 · DOI

• The toxicity of lead-based perovskite solar cells has driven intense research into economical, clean, and sustainable lead-free alternatives to PV technology. Although there are many promising options for LFPSCs, Tin (Sn), bismuth (Bi), Antimony (Sb), and germanium (Ge) are primarily used for this purpose due to their non-toxicity, intrinsic properties, and ecofriendliness. However, many obstacles, such as stability and low efficiency, are holding back their practical application. Our review primarily focuses on strategies to enhance the stability and efficiency of lead-free perovskite solar cells. This paper discusses the limitations of Pb-based perovskites and explores promising lead-free alternatives. It further describes the major issues with lead-free perovskite solar cells and delves into strategies to improve their stability and performance. It also provides an overview of recent developments in lead-free perovskite solar cells. Since the first report on lead-free CsSnI3-based PSC devices, the field has seen rapid growth, and the efficiency of lead-free perovskite solar cells has increased from below 1% to 15%. In this context, lead-free perovskites present a compelling direction due to their intrinsically stable phases and compatibility with existing fabrication technologies. Despite recent advances in synthetic techniques and device engineering, lead-free perovskites still face challenges, including stability and effi- ciency issues, high cost, complex fabrication processes, and poor physicochemical properties. Figure 6 provides a graphical summary of leadfree alternatives and key strategies to improve the efficiency and stability of LFPSCs. Compositional engineering by varying the cation and anion is an effective approach to stabilizing the perovskite structure. In particular, the mixed-cation or mixed-anion strategy yields a high-quality thin film with phase control and defect suppression. Moreover, the performance of leadfree perovskite solar cells is significantly limited by rapid crystallization and poor film quality. The crystallization process can be controlled by solvent interactions, formation of intermediate adducts, and anion substitution. Selective solvents and organic moieties such as DMSO, chlorobenzene, or carboxylic groups retard crystallization and yield compact, pinhole-free thin films. Film quality can be improved through one- or two-step coating techniques and vapor deposition methods. Many atypical synthesis approaches, such as vapor-assisted solution process (VASP) or high-low vacuum deposition (HLVD), are introduced to facilitate the preparation of high-quality thin films. Sn and Ge-based perovskite solar cells are unstable under environmental conditions because they immediately convert from Sn2+ and Ge2+ states into Sn4+ and Ge4+. The use of additives such as SnF2 can suppress oxidation and increase stability.

  Strategic development of stable and efficient lead-free perovskite solar cells · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y Universities (grant no. 501XYGG2025158009). W.Y., Y.F., M.L., and D.L. disclose support for the research of this work from the State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources (grant no. LAPS25005). L.Z. disclose support for the research of this work from the National Natural Science Foundation of China (grant no. 52203208), the Young Elite Scientists Sponsorship Program by CAST (grant no. YESS20240571), and the Beijing Nova Program (grant no. 20230484480). C.P. disclose support for the research of this work from the National Natural Science Foundation of China(grant nos. 52125205, 52532005 and 52250398). X.Y. disclose support for the research of this work from China Postdoctoral Science Foundation (grant nos. BX20250354 and 2025M770177), and Young Elite Scientists Sponsor- ship Program of the Beijing High Innovation Plan (grant no. 20250791).

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y 59. 60. 61. Xie, T. Large language models as master key: unlocking the secrets of materials science with GPT. Preprint at https://doi.org/ 10.48550/arXiv.2304.02213 (2023). Li, L. et al. Electronic transport of organic-inorganic hybrid perovskites from first-principles and machine learning. Appl. Phys. Lett. 114, 8 (2019). Shetty, P. et al. A general-purpose material property data extraction pipeline from large polymer corpora using natural language processing. Npj Comput. Mater. 9, 52 (2023). 62. Noh, J. et al. Inverse design of solid-state materials via a con- 63. tinuous representation. Matter 1, 1370–1384 (2019). Zhang, C. et al. Machine learning-driven interface material design for high-performance perovskite solar cells with scalability and band-gap universality. Joule 10, 2 (2026). 64. Choe, H., Yoon, H., Kim, I. & Kim, S. Y. Machine learning unlocks new directions in halide perovskite research. ChemElectroChem 13, e202500282 (2026). Laufer, F., Götz, M. & Paetzold, U. W. Deep learning for augmented process monitoring of scalable perovskite thin-film fabrication. Energy Environ. Sci. 18, 1767–1782 (2025). 65. 67. 68. 66. Nathan, A., Chaji, G. R. & Ashtiani, S. J. Driving schemes for a-Si and LTPS AMOLED displays. J. Disp. Technol. 1, 267–277 (2005). Xue, D.-J. et al. Regulating strain in perovskite thin films through charge-transport layers. Nat. Commun. 11, 1514 (2020). Zhao, J. et al. Strained hybrid perovskite thin films and their impact on the intrinsic stability of perovskite solar cells. Sci. Adv. 3, eaao5616 (2017). Sun, C. et al. Multidentate fullerenes enable tunable and robust interfacial bonding for efficient tin-based perovskite solar cells. Adv. Mater. 36, 2410248 (2024). 69. 70. Hui, J. et al. Super strong bonding at the interface between ETL 71. and perovskite for robust flexible optoelectronic devices. Angew. Chem. Int. Ed. 64, e202424483 (2025). Sahoo, K., Harish, V., Ren, H. & Iyer, S. S. A review of die-to-die, dieto-substrate and die-to-wafer heterogeneous integration. IEEE Electron Dev. Rev. 2, 6–31 (2025). Lee, C. Y. et al. 3D integrated process and hybrid bonding of high bandwidth memory (HBM). Electron. Mater. Lett. 21, 395–419 (2025). 73. Gao, G. et al. Scaling package interconnects below 20 µm pitch with hybrid bonding. 2018 IEEE 68th Electronic Components and Technology Conference (ECTC) 314–322 (San Diego, CA, USA, 2018). 72. 75. 74. Chiu, C.-H., Chien, C.-H., Chien, W.-C. & Chen, Y.-H. To Enhance light lumination of OLED lighting by film included fish scales. Dig. Tech. Pap. 47, 1549–1551 (2016). Jastrzebska-Perfect, P. et al. On-site growth of perovskite nanocrystal arrays for integrated nanodevices. Nat. Commun. 14, 3883 (2023). Piferi, C. et al. Hydrophilicity and hydrophobicity control of plasma-treated surfaces via fractal parameters. Adv. Mater. Interfaces 8, 2100724 (2021). 76. 77. Wu, W. et al.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y 11. 12. 13. 14. 15. 16. 17. 18. 19. Leung, T. L. et al. Stability of 2D and quasi-2D perovskite materials and devices. Commun. Mater. 3, 63 (2022). Chen, Y. et al. 2D Ruddlesden–popper perovskites for optoelectronics. Adv. Mater. 30, 1703487 (2018). Zhang, L. et al. High-performance quasi-2D perovskite lightemitting diodes: from materials to devices. Light Sci. Appl. 10, 61 (2021). Lei, Y. et al. Perovskite superlattices with efficient carrier dynamics. Nature 608, 317–323 (2022). Cherniukh, I. et al. Perovskite-type superlattices from lead halide perovskite nanocubes. Nature 593, 535–542 (2021). Liu, Z. et al. Metal–halide perovskite nanocrystal superlattice: selfassembly and optical fingerprints. Adv. Mater. 35, 2209279 (2023). Yang, W. et al. Overcoming charge confinement in perovskite nanocrystal solar cells. Adv. Mater. 35, 2304533 (2023). Jang, K. Y. et al. Efficient deep-blue light-emitting diodes through decoupling of colloidal perovskite quantum dots. Adv. Mater. 36, 2404856 (2024). Shao, W. et al. Molecular templating of layered halide perovskite nanowires. Science 384, 1000–1006 (2024). 21. 20. Wang, S. et al. Morphological engineering for high-performance perovskite field-effect transistors. FlexMat 2, 82–106 (2025). Cheng, X. et al. Principles, fabrication, and applications of halide perovskites-based memristors. FlexMat 1, 127–149 (2024). Zhu, Y. et al. Bilayer metal halide perovskite for efficient and stable solar cells and modules. Mater. Futur. 1, 042102 (2022). Zhang, Q. et al. Nanosecond response perovskite quantum dot light-emitting diodes with ultra-high resolution for active display application. Light Sci. Appl. 14, 285 (2025). 23. 22. 25. for photodetector applications. Light Sci. Appl. 14, 355 (2025). Kim, J. S. et al. Kinetically-controlled intermediate-direct-pinning for homogeneous energy landscapes in quasi-two-dimensional perovskites for efficient and narrow blue emission. Nat. Commun. 16, 9590 (2025). Kong, L. et al. Fabrication of red-emitting perovskite LEDs by stabilizing their octahedral structure. Nature 631, 73–79 (2024). 27. Gao, Y. et al. Microsecond-response perovskite light-emitting 26. diodes for active-matrix displays. Nat. Electron. 7, 487–496 (2024). 28. Meng, Y. et al. Epitaxial growth of α-FAPbI3 at a well-matched heterointerface for efficient perovskite solar cells and solar modules. Adv. Mater. 36, 2309208 (2024). Reo, Y. et al. Vapour-deposited high-performance tin perovskite transistors. Nat. Electron. 8, 403–410 (2025). 29. 30. Du, P. et al. Thermal evaporation for halide perovskite optoelectronics: fundamentals, progress, and outlook. Adv. Opt. Mater. 10, 2101770 (2022). Zhu, H. et al. Long-term operating stability in perovskite photovoltaics. Nat. Rev. Mater. 8, 569–586 (2023). 31. 38. 39. 40. 41. Li, H. et al.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y Fig. 6 | Promising applications of perovskite-based heterogeneous integration devices. a Light management and integration strategies for PeLED-based AR and VR systems. b Ion migration in halide perovskites and their integration in neuromorphic computing applications. A is ion migration. Under electrical or optical stimulation, halide ions redistribute within the perovskite lattice, modulating device con- ductance in a manner analogous to synaptic weight adaptation in biological synapses92,93. Neuromorphic behavior in perovskite-based devices is primarily governed by ion migration, but also arises from the interplay of trap- assisted charge storage and interfacial electrochemical processes. While these mechanisms enable essential synaptic functions such as conductance plasticity, temporal integration, and history-dependent responses, they simultaneously introduce instability. Uncontrolled ion migration accelerates material degradation, promotes trap formation, and induces fluctuations in operating voltage and current. This inherent duality indicates that the same physical processes responsible for neuromorphic functionality can compromise device durability and signal reliability. Achieving reproducible and energy-efficient opera- tion, therefore, requires a transition from empirical material optimi- zation to precise regulation of these dynamic processes. This challenge is particularly critical for backplane-integrated systems, where array- level uniformity and long-term operational stability are as important as the synaptic functionalities themselves. Beyond adopting perovskites as active materials, integrating perovskites with neuromorphic architectures represents a transfor- mative route for next-generation optoelectronics. Such integration enables near-sensor computing and memory, where perception occurs in perovskite devices while memory and processing are realized in neuromorphic circuits. For example, Seung et al. reported an inte- grated system combining synaptic phototransistors with QD LEDs, enabling on-device preprocessing that filters noise and amplifies signal contrast via threshold-driven illumination94. Similarly, Shan et al. developed a sensory-memory processing platform using a tribo- electric nanogenerator as the sensory receptor and QD LEDs as an artificial synapse, demonstrating multi-wavelength synapse-like opti- cal outputs and efficient multimodal information recognition95. Such integrated perovskite-neuromorphic devices offer significant potential for intelligent vision sensors, self-adaptive imaging arrays, and multi- modal AI systems. Therefore, we anticipate that perovskite- neuromorphic integration is poised to enable next-generation adap- tive optoelectronic interfaces and advanced human-machine com- munication technologies.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y mimicking natural architectures, these strategies extend beyond structural replication to introduce intrinsic functionalities, including omnidirectional light extraction and enhanced environmental stability, without compromising the optoelectronic properties of the material. For instance, replicating the hierarchical multilayer structures found in iridescent fish scales can reduce internal reflection and enhance light outcoupling, while simultaneously providing a biomimetic barrier against moisture ingress74. Strategies such as programmed coffee-ring effect, peptide-mediated self-assembly, and biomineralization-inspired growth have demonstrated the ability to regulate perovskite nucleation kinetics with sub-micron precision. These bio-inspired strategies therefore provide a robust and lowdamage route to achieving high-fidelity pixel arrays and long-term operational stability, which are essential for next-generation microdisplays. Patterned surface wettability, inspired by the lotus leaf effect, is one of the most popular techniques for fabricating patterned optoelectronic devices. By creating hydrophilic-hydrophobic regions, the selective deposition of functional materials can be guided. The contrast in surface energy between hydrophilic and hydrophobic domains is critical for patterning accuracy, as solution-processed perovskites are highly sensitive to local wettability. Patterned substrates with controlled wettability and template geometries can induce directional forces that guide nanocrystal positioning during growth, potentially achieving resolutions beyond the conventional lithographic limits75. Diverse approaches to create such patterned wettability include UV Ozone exposure and oxygen plasma treatment. which generate oxygen-rich surface layers to enhance hydrophilicity. These strategies have been successfully applied to fabricate patterned perovskite optoelectronic devices76,77. Besides, chemical functional materials, including poly(methyl methacrylate) (PMMA), poly(N,N’-bis-4-butylphenyl-N,N’-bisphenyl)benzidine (Poly-TPD) and SAMs, have also been adopted to achieve high-contrast wettability patterns78,79. The coffee-ring effect provides another bio-inspired patterning method. During solvent evaporation, non-volatile solutes in a droplet mitigate toward the pinned contact line, forming a characteristic ringshaped pattern. While this effect is typically undesired due to nonuniformity, it can be harnessed to deposit materials selectively when carefully controlled. For example, Li et al. confined solution droplets within pre-patterned hydrophilic regions, enabling nanoparticles to self-assemble into well-defined circuits through the coffee-ring effect, yielding large-area patterns with diverse geometries within one minute while preserving nanoparticle properties80. Thus, such strategies can similarly be applied to perovskite films for high-precision patterning.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y AI-driven methods can extract latent knowledge from the literature, enabling researchers to identify promising directions for materials design. This capability is particularly valuable for developing multi-functional materials, where complex chemical, mechanical and electronic requirements must be simultaneously satisfied. Besides, AIdriven methods can involve diverse data types, including X-ray diffraction, Raman spectra, molecular structures, electronic structures, and morphological information, allowing the construction of unified models that correlate structural, optical, and electronic properties, thereby yielding deeper insights into material behavior. For instance, Xie et al. developed a natural language processing (NLP) task called structured information inference to extract the complex device-level information59. By fine-tuning GPT-3 on an existing PSC FAIR (Findable, Accessible, Interoperable, Reusable) dataset, they achieved > 91% accuracy and expanded the dataset for further analysis. Furthermore, AI methods trained on high-throughput DFT calculations can predict key materials properties, such as bandgap, stability, carrier mobility and defect tolerance. For example, Li et al. combined the data-mining machine learning techniques with the non-equilibrium Green’s function method and DFT to rapidly predict the electronic properties of MAPbI3 60. AI also offers a data-driven, property-guided, and highly efficient approach to materials design, leveraging large amounts of unlabeled data. By identifying promising candidates before conducting costly experiments, AI-driven workflows can drastically reduce unnecessary trials, allowing researchers to focus on the most promising materials. Shetty et al. addressed the challenge of extracting chemistry-structureproperty relationships from the rapidly expanding materials science literature using NLP61. They obtained ~ 300,000 material property entries from ~ 130,000 abstracts within 60 h, spanning fields including fuel cells, supercapacitors, and polymer solar cells, thereby revealing valuable insights. A key advantage of an AI-driven method is the capability to perform inverse design, enabling the prediction of materials with specific target properties. For example, Noh et al. presented an inverse materials design framework that computationally generates new solid-state materials by learning a continuous representation of m known compounds, producing approximately 20,000 hypothetical materials62. Although AI-driven methods are increasingly being introduced into materials discovery, their application to the heterogeneous integration of perovskites with driving backplanes remains at an early stage.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y Fig. 5 | Promising future directions on perovskite heterogeneous integration devices. a AI-driven materials discovery and screening workflow. GIN, graph iso- morphism network; EGNN, equivariant graph neural network; ChemBERTa, che- mical bidirectional encoder representations from transformers; MolFormer, molecular transformer; RNN, recurrent neural network; GraphVAE, graph varia- tional autoencoder; DiffDock, diffusion-based docking; CGCNN, crystal graph convolutional neural network. b Strategies for interfacial coupling and bonding, including die-to-die (D2D), die-to-wafer (D2W), and wafer-to-wafer (W2W). science, device engineering, and AI-driven innovation, future research may enable a new generation of high-performance perovskite optoe- lectronic devices seamlessly integrated with CMOS and/or TFT backplanes. Al-driven materials discovery and screening. Traditionally, materials discovery and screening have relied on trial-and-error experiments and density functional theory (DFT) calculations, both of which are time-consuming and computationally intensive. These conventional approaches typically require iterative synthesis, characterization and parameter optimization, limiting the efficient exploration of the vast chemical space. Besides, advances in characterization techniques have generated increasingly large and complex datasets, posing challenges in uncovering the meaningful structure-property relationship.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y critical challenge. Future efforts should combine in situ characterization, multiscale modeling, and artificial intelligence (AI)-assisted materials and process optimization to clarify the coupled electronic, ionic, and mechanical processes at heterogeneous interfaces. Such advances will be essential for developing more universal and robust interfacial engineering strategies for system-level integrated perovskite optoelectronics. Scalability and precise patterning. Achieving uniform wafer-scale patterned perovskite films with consistent thickness and morphology remains challenging. Owing to the intrinsic sensitivity of solutionprocessed perovskites to local surface wettability and solvent evaporation kinetics, large-area deposition often results in compositional inhomogeneity and pinhole formation, which induce localized electric- field concentration and degrade device performance. In addition, polar aprotic solvents used in the perovskite precursor inks, such as N,N-dimethylformamide (DMF) and dimethylsulfoxide (DMSO), can corrode metal interconnects or organic passive layers in CMOS or TFT backplanes, thereby imposing constraints on process compatibility. Although chemical vapor deposition and other solvent-free techniques can improve compatibility with driving backplanes, they often suffer from inferior device performance. Continued efforts in vapor-phase deposition are therefore required to balance crystallization kinetics and film uniformity, enabling industrial-scale heterogeneous integration. Precise alignment of perovskite pixels with driving backplanes requires high-resolution and low-damage patterning techniques. Utilizing electron beam etching, researchers have confirmed that perovskite device pixel sizes can be reduced down to ~ 90 nm, corresponding to a record-high pixel density of 127,000 ppi53. This resolution lies within the capability of advanced CMOS manufacturing, thereby enabling the integration of halide perovskites with semiconductor technologies. However, most conventional semiconductor etching methods, although capable of generating high-resolution patterns, remain incompatible with the complex topographies of practical electronic backplanes. TFT and CMOS circuits consist of multilayer interconnects, intricate routing networks and heterogeneous materials, typically resulting in non-planar surface morphologies. Such features complicate perovskite pattern transfer and hinder uniform crystallization. Surface roughness and pixel topography fundamentally govern the thermodynamics of heterogeneous integration by influencing precursor wetting, nucleation kinetics and subsequent film crystallization. Non-planar architectures result in non-uniform precursor distribution and localized thickness variations, which in turn trigger stochastic crystallization and defect formation, particularly at step edges and metal interconnects.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y a d f b c e g Fig. 4 | Integrating PePDs on driving backplanes. a Structural schematic of a perovskite image sensor. RST, reset transistor; SEL, select transistor. b An optical image of bonded and packed perovskite-based image sensors. c Intensity image captured by the prototype perovskite image sensor. d Schematic illustration of a perovskite X-ray detector. X-ray photons generate electron-hole pairs within the perovskite film, and the electrons drift toward the CMOS pixels under an applied bias to finally form an X-ray image. Scale bar, 200 μm. e X-ray image of a standard resolution test pattern. Three regions of interest (ROIs) are highlighted, and the bars with a spatial resolution of 5.0 lp mm−1 in ROI-3 are distinctly visible. f Radiographic images of a mouse. The first column shows an image captured using a commercial indirect-conversion CMOS X-ray detector with a pixel size of 74.8 µm. The second and third columns show images obtained from a direct perovskite- CMOS X-ray detector with an 83.2 µm pixel size at two different dose levels: 100% and 50%. Scale bar, 2.0 mm. g Magnified ROIs. The skull, ribs, and pelvis captured by the direct perovskite-CMOS X-ray detector appear significantly sharper and less noisy, even at 50% of the radiation dose. Panels reproduced with permission from: a–c, ref. 45, Copyright©2024, American Chemical Society. d–g, ref. 46, Copyright©2024, under the terms of the Creative Commons CC BY license.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y a d e b c f g Fig. 3 | Integrating PeLEDs on driving backplanes. a Structural schematic diagram of the display panel by integrating PeLEDs with TFT backplanes. ITO, indium tin oxide; PI, polyimide; NPB, N,N’-bis-(1-naphthalenyl)-N,N’-bis-phenyl-(1,1’-biphenyl)-4,4’-diamine; TPBi, 1,3,5-tri(N-phenylbenzimidazol-2-yl)benzene. b Crosssectional diagram of the TFT structure beneath each sub-pixel, with arrows indicating the current flow. ACT, activated layer; GI, gate insulation layer; ILD, insulation layer dielectric; PLN, planarization layer; PDL, pixel definition layer. c Photograph of the active-matrix PeLED display panel and its operating state. Scale bar, 5 cm. FPC, flexible printed circuit; IC, integrated circuit. d Schematic of the TFT backplane structure. The multilayers of PeLEDs uniformly cover the entire TFT backplane, with the co-cathode connected to grounded (GND). Each pixel integrates three TFTs, address TFT-1, drive TFT-2 and reset TFT-3, combined with a storage capacitor to regulate the PeLED. a-Si, amorphous silicon; n+, n-doped region; Vdata, data voltage; Vscan, scan voltage; VDD, supply voltage. The red dashed box denotes the capacitor area. e–g Digital photographs of the 1-inch red (e), green (f) and blue (g) active-matrix PeLEDs displaying cartoon images. Panel reproduced with permission from: (a–c), ref. 35, Copyright©2023, under exclusive license to Springer Nature Limited. d, ref. 38, Copyright©2024, under exclusive license to Springer Nature Limited. e–g ref. 27, Copyright©2024, under the terms of the Creative Commons CC BY license. as a blanking unit to effectively suppress dark current to near-zero levels, and integrated it on a 64 × 64 In2O3 TFT backplane43. This strategy establishes a generalizable approach for minimizing dark current and enhancing signal-to-noise ratio, advancing high-precision and low-noise perovskite-based X-ray and PD array technologies. Perovskites can also be directly integrated with CMOS to build highperformance imaging and sensing devices. Yen et al. proposed a CMOS-compatible hybrid image sensor architecture by directly integrating PePDs onto a 0.18 μm CMOS readout circuit, realizing a fully integrated 40 × 40 μm2 hybrid perovskite-CMOS array44. This work represented the first demonstration of perovskite-CMOS integration under low-bias operation, showcasing the feasibility of monolithic perovskite-CMOS platforms for compact, low-cost and high-sensitivity imaging. Song et al. demonstrated monolithic integration of PePDs with silicon readout integrated circuits, enabling high-resolution 2D and 3D imaging within a unified architecture (Fig. 4a)45. The hybrid device enables precise 2D imaging and fast top-electrode-controlled indirect time of flight-based 3D depth sensing (Fig. 4b, c), highlighting the potential of in situ integrating perovskite optoelectronic devices on CMOS-compatible substrates. Liu et al.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y Fig. 2 | Fundamentals of perovskite heterogeneous integration. a Schematic illustration of perovskite patterning techniques. b Schematic of interfacial coupling between perovskites and underlying driving backplanes. c Schematic of in situ integration of perovskites with driving backplanes. Typically, the perovskite precursor solution is directly deposited on the backplane, and the MHP forms upon solvent evaporation. d Schematic of ex situ integration of perovskites with driving backplanes. Typically, MHP films or devices are first fabricated on a temporary substrate and subsequently transferred to the target backplanes using a stamp. color PeLED displays (Fig. 3e–g)27. Zhang et al. further developed highperformance sky-blue PeLEDs and successfully integrated them into an active-matrix pixel circuit, demonstrating the feasibility of in situ integrating perovskite optoelectronic devices directly on electronic substrates and advancing their potential for practical display and integrated optoelectronic systems39. Integrating PePDs on driving backplanes. Monolithically heterogeneous integration of PePDs with driving backplanes has unlocked new opportunities for highly miniaturized image sensors. Li et al. presented an all-inorganic perovskite heterojunction film fabricated via vapor deposition that meets TFT integration standards40. Monolithic integration of this heterojunction film with a 64 × 64 TFT array enabled high-resolution, real-time X-ray imaging, demonstrating the feasibility of in situ integrating perovskite materials with TFT backlow-dose flat-panel X-ray detectors. planes for next-generation, Beyond rigid platforms, integrating perovskite PDs with indium gallium zinc oxide (IGZO) TFT panels also enables flexible image sensors. Zou et al. demonstrated a flexible image sensor by integrating PePDs with IGZO TFT panels, providing a low-cost, high-performance alternative to conventional amorphous silicon-based sensors41. To enhance the mechanical anchoring between perovskites and TFT substrates, Deumel et al. adopted a photoresist grid as an adhesion promoter for thick MAPbI3 layers42. Using this method, they integrated perovskite films onto a backplane with a 640 × 480-pixel array and a pixel density of 508 ppi, achieving a pixelated X-ray detector sensitivity of 1060 μC −1 cm−2. Jin et al.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y and defect-free interfaces remain underdeveloped, posing additional barriers of large-scale implementation. Overcoming these challenges will be essential for unlocking the full potential of perovskite-based heterogeneous devices in next-generation optoelectronic technolo- gies, including high-resolution displays, on-chip light sources, and advanced imaging. In this Perspective, we first outline the key requirements for the system-level integration of perovskite optoelectronic devices with driving backplanes. We then summarize recent progress in the het- erogeneous integration of PeLEDs and PePDs with established driving backplanes. We also discuss the key challenges that limit practical deployment, including interfacial coupling, materials compatibility, scalability, precising patterning, and electrical and optical matching. Finally, we highlight promising future research directions and advanced applications of perovskite-based heterogeneous integration devices. Particularly, the convergence of perovskite optoelectronics with neuromorphic architectures presents an exciting emerging ave- nue that could reshape the landscape of next-generation optoelec- tronic technologies.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y Fig. 1 | Crystal structures and device architectures. a Unit cell of perovskite crystal with the chemical formula ABX3, where A is a monovalent cation, B is a divalent cation, and X is a halide anion. b Typical energy band structure of lead iodide perovskites. c Representative dimensionalities of perovskites, including zero-dimensional (0D), one-dimensional (1D), two-dimensional (2D), quasi-2D and three-dimensional (3D) structures. d Illustration of perovskite optoelectronic device configurations. ME, metal electrode; CTL, charge-transporting layer; MHP, metal halide perovskites; TE, transparent electrode; DL, dielectric layer; Si, silicon; CIL, charge-injection layer. e Typical device structures of perovskite LEDs (PeLEDs) or perovskite PDs (PePDs) integrated with driving backplanes such as thin-film transistors (TFTs) and complementary metal-oxide-semiconductor (CMOS) circuits. PA, passivation layer; S, source; D, drain; p Si, p-type silicon; n well, n-type silicon well; n+, heavily doped n-type silicon; p+, heavily doped p-type silicon. network, superlattice structures can be engineered to achieve tailored optical and electronic properties14–16. Moreover, long-chain organic A-site cations can drive the formation of MHP quantum dots (QDs) and/or nanowires, offering further opportunities for tuning material functionality and device performance17–19. Collectively, these structural variations highlight the remarkable versatility of MHPs and their potential for enabling customizable applications in optoelectronic devices. The exceptional optoelectronic properties of MHPs have positioned them as a rising star in next-generation optoelectronic technologies. Up to now, MHPs have been widely applied in PDs, phototransistors, photoconductors and LEDs (Fig. 1d), and their device performance has undergone remarkable advancement over the past decade20–24. Perovskite PDs (PePDs) already exhibit performance metrics comparable to those of commercial counterparts, while perovskite LEDs (PeLEDs) have achieved high external quantum effi- ciencies (EQEs) exceeding 28% in the green and red spectral regions and over 20% for blue emission3,25,26. Moreover, PeLEDs show narrow emission linewidths that are advantageous for achieving high color purity3,25,26. The high color purity and energy-efficient features make i.e., augmented PeLEDs promising for micro-display applications, reality (AR) and virtual reality (VR) near-eye micro-displays. Despite rapidly expanding potential, the commercial translation of PeLEDs ultimately depends on their reliable integration with driving backplanes, such as thin-film transistor (TFT) arrays and complementary metal-oxide-semiconductor (CMOS) chips. for advancing the semiconductor Monolithic heterogeneous integration of electrical driving backplanes with functional optoelectronic devices represents a transformative strategy industry. Representative architectures for perovskite optoelectronic devices enabled by such heterogeneous integration are illustrated in Fig. 1e. Several studies have reported the successful heterogenous integration of PeLEDs or PePDs with TFT or CMOS circuits, achieving uniform, high-resolution pixel arrays and fast switching speeds27. Besides, diverse deposition and growth techniques for MHPs, such as blade coating, epitaxial growth, and thermal evaporation, have been adopted to enable large-area fabrication and compatibility with driving backplanes, thereby broadening the processing toolkit for perovskiteintegration28–30. Despite these advances, based heterogeneous achieving robust heterogeneous integration while maintaining excellent optoelectronic device performance remains a sizable challenge. Presently, major obstacles lie in the substantial mismatches in processing conditions and material properties between perovskites and conventional driving backplanes, particularly in terms of interfacial bonding materials, fabrication environments, and material stability. Such incompatibilities generally lead to increased defect densities, reduced charge-transfer efficiency and overall device degradation. These challenges are further compounded by the pronounced sensitivity of perovskites to processing environments, such as polar solvents, oxygen, and temperature fluctuations, which can exacerbate chemical and structural degradation during integration31,32.

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• https://doi.org/10.1038/s41467-026-74068-y System-level integration of halide perovskite optoelectronics for its commercial deployment Received: 17 December 2025 Accepted: 29 May 2026 Wenqiang Yang 1,2,7, Xiaoyu Yang3,7, Rui Su 4,7, Yuyi Feng 5, Lichen Zhao 6 5, Xinqiang Wang 3, Rui Zhu3, Caofeng Pan 1,2 & Deying Luo 1, Meicheng…

  System-level integration of halide perovskite optoelectronics for its commercial deployment · 2026 · DOI

• We introduce an all-optical technique for characterizing the functional- ity of individual nanocrystals at a large scale. This non-invasive in situ method enables the simultaneous determination of size and emission of thousands of luminescent nanocrystals within minutes, allowing the high-throughput determination of PLQY at the single-particle level. We observe size-dependent phenomena in the life cycle—that is, from birth to death—of archetypical CsPbBr3 perovskite nanocrystals, which were previously masked by bulk averages. In the as-synthesized colloidal sample, we observe a large spread in PLQYs, even among nanocrystals of the same size, with smaller nanocrystals exhibiting higher PLQYs. During post-synthesis solution-phase treatments, we find that larger nanocrystals require additional PbBr2 treatment to match the PLQY of their smaller counterparts, indicating that halide vacancies are the leading cause of the size-dependent PLQY differences. In degrada- tion studies, continuous illumination at high laser powers reveals that although the PL signal is rapidly quenched, a substantial fraction of the original nanocrystal remains. We reveal a strong correlation between photobleaching and volume loss, with degraded nanocrystals retaining their cubic shape and forming metallic lead (Pb0). Counterintuitively, increasing the laser power quenches the degradation process, which we attribute to the formed Pb0 acting as an electron acceptor and absorb- ing the photoinduced electrons. By isolating single-particle effects within colloidal ensembles, our method provides a pathway towards more efficient and stable perovskite nanomaterials and devices. We anticipate that scaling up single-particle methods will be a cornerstone for designing next-generation nanoparticle-based technologies.

  High-throughput in situ sizing and quantum yield determination of individual perovskite nanocrystals · 2026 · DOI

• This section outlines the main methodological and practical constraints of the present multiscale study and delimits the contexts in which the results should be interpreted. The work couples first-principles electronic-structure calculations (WIEN2k) with one-dimensional device simulations (SCAPS-1D) and module/system modelling (PVsyst). While this hierarchical scheme yields useful design directions and comparative trends, it also inherits approximation errors and practical gaps at each level that limit exact quantitative predictivity. 4.1 Limitations of DFT calculations using WIEN2k WIEN2k provides accurate full-potential LAPW results for equilibrium geometry, band dispersion, and frequency-dependent dielectric response within the approximations chosen. Nonetheless, standard density functional approximations (LDA/GGA) typically underestimate band gaps and may misplace defect levels; many-body effects (GW) and excitonic interactions (Bethe–Salpeter) are not included in the independent-particle optical spectra presented here. Temperature effects (thermal expansion, electron–phonon coupling) and dynamical disorder were not considered, so predicted optical onsets and carrier lifetimes represent zero-K, idealized behavior. Convergence choices (k-point density, basis-set 1 3N. B. Nadwi et al. Page 43 of 53 284 parameters, and neglect of spin–orbit coupling (SOC)) influence numerical results; heavy elements such as Cs, In, and I may require explicit spin–orbit coupling to obtain quantitatively reliable band ordering. In the present work, the TB-mBJ (Tran-Blaha modified Becke-Johnson) meta-GGA exchange–correlation potential was employed for band-structure and optical-property calculations rather than the computationally more expensive HSE06 hybrid functional. The TB-mBJ potential was specifically developed to correct the systematic band-gap underestimation of standard GGA/LDA functionals and has been extensively benchmarked across a wide range of semiconductors and insulators, yielding bandgaps in quantitative agreement with experiment and with HSE06 predictions at a fraction of the computational cost (Tran and Blaha 2009). For the Cs2NaInI6 double perovskite system, which possesses a 40-atom conventional unit cell and is treated with the all-electron FP-LAPW basis set in WIEN2k, HSE06 calculations would require approximately one to two orders of magnitude greater computational resources than TB-mBJ, making them impractical within the scope of the present multiscale study. The TB-mBJ functional has been specifically validated for halide perovskite and double perovskite systems and is widely used in the literature for this class of materials. Future work will employ HSE06 corrections on reduced supercells to further validate the TB-mBJ bandgap and effective-mass predictions obtained here.

  Fundamental optoelectronic and device physics of Cs2NaInI6 double halide perovskites from first-principles to multiscale modeling · 2026 · DOI

• The rapid expansion of research on inorganic lead-free halide perovskites has substantially broadened the landscape of viable absorber materials for photovoltaic applications. Driven by both sustainability considerations and the pursuit of long-term operational robustness, increasing attention has shifted toward inorganic compositions capable of retaining favorable optoelectronic characteristics while offering enhanced chemical and structural stability. As understanding of this material landscape has deepened, comparative studies of representative inorganic lead-free perovskites have clarified how coordination environment, structural dimensionality, and defect chemistry influence carrier generation, transport, and recombination. This understanding offers a useful basis for interpreting performance differences across absorbers and for pinpointing the structural and electronic features most critical to efficient photovoltaic operation.

  Stable and eco-friendly inorganic lead-free perovskite solar cells: structural, electronic, and defect engineering · 2026 · DOI

• The stability enhancement mechanism is attributed to 'improved crystallinity and reduced buried-interface defects' hindering 'oxygen and moisture penetration,' but quantitative defect density measurements (using deep-level transient spectroscopy or space-charge-limited current analysis) and direct oxygen/moisture permeation rates through modified versus unmodified SnO2 films are not provided.

  Electron-donor/-acceptor ratio-guided molecular engineering for buried interface optimization in n-i-p perovskite solar cells · 2026 · DOI

• The work identifies that positive surface charges on EDG-rich molecules facilitate electron extraction, but does not quantitatively measure interfacial dipole moments, surface charge density, or energy band alignment shifts across the EDG/EWG ratio spectrum using techniques like ultraviolet photoelectron spectroscopy (UPS) or Kelvin probe force microscopy (KPFM).

  Electron-donor/-acceptor ratio-guided molecular engineering for buried interface optimization in n-i-p perovskite solar cells · 2026 · DOI

• The paper demonstrates a champion PCE of 24.18% with DBA-modified devices but does not report the statistical distribution of device performance across multiple fabrication batches or provide reproducibility data (standard deviation, efficiency range, yield percentage) to assess manufacturing scalability of the EDG/EWG ratio-guided engineering approach.

  Electron-donor/-acceptor ratio-guided molecular engineering for buried interface optimization in n-i-p perovskite solar cells · 2026 · DOI

• The mechanism by which EDG-rich molecules reduce perovskite grain boundary density is attributed to 'retarded crystallization' and 'enhanced film uniformity,' but quantitative crystallization kinetics data (nucleation rates, grain growth dynamics) and direct microscopic comparison of grain boundary evolution between modified and unmodified samples during film formation are absent.

  Electron-donor/-acceptor ratio-guided molecular engineering for buried interface optimization in n-i-p perovskite solar cells · 2026 · DOI

• While the paper establishes that perovskite surface passivation dominates over ETL modification, it does not investigate whether combining optimized EDG-rich dipolar molecules on both the buried interface and perovskite surface (dual-layer passivation) yields synergistic effects beyond single-interface modification, which is mentioned as desirable but not experimentally validated.

  Electron-donor/-acceptor ratio-guided molecular engineering for buried interface optimization in n-i-p perovskite solar cells · 2026 · DOI

Most-cited papers in Perovskite Materials and Applications

• Incorporation of rubidium cations into perovskite solar cells improves photovoltaic performance · Science · 2016 · 3,529 citations
• Perovskite solar cells with atomically coherent interlayers on SnO2 electrodes · Nature · 2021 · 3,089 citations
• Pseudo-halide anion engineering for α-FAPbI3 perovskite solar cells · Nature · 2021 · 3,006 citations
• Efficient perovskite solar cells via improved carrier management · Nature · 2021 · 2,845 citations
• Properties and potential optoelectronic applications of lead halide perovskite nanocrystals · Science · 2017 · 2,469 citations
• Improved charge extraction in inverted perovskite solar cells with dual-site-binding ligands · Science · 2024 · 1,279 citations
• Buried interface molecular hybrid for inverted perovskite solar cells · Nature · 2024 · 967 citations
• Highly efficient and stable perovskite solar cells via a multifunctional hole transporting material · Joule · 2024 · 684 citations
• Reinforcing self-assembly of hole transport molecules for stable inverted perovskite solar cells · Science · 2024 · 498 citations
• Double-side 2D/3D heterojunctions for inverted perovskite solar cells · Nature · 2024 · 417 citations

Most recent work

• Stable and eco-friendly inorganic lead-free perovskite solar cells: structural, electronic, and defect engineering · Communications Materials · 2026
• Transition-Metal-Doped Halide Perovskites for Near-Infrared Emissions: Beyond Oxides · ACS Energy Letters · 2026
• Electron‐Deficient Amines Enable Halide‐Anchoring Hydrogen Bonding for Stable Wide‐Bandgap Perovskites Toward Perovskite/Organic Tandem Solar Cells · Angewandte Chemie International Edition · 2026
• Multifunctional Glutathione Enables ISOS‐Robust Inverted Perovskite Solar Cells via Dipole Engineering and Redox‐Driven Self‐Healing · Advanced Materials · 2026
• Interface engineering with single-walled carbon nanotubes for high-efficiency (> 29%) lead-free BaSnS₃ perovskite solar cells: insights from numerical simulation · Materials for Renewable and Sustainable Energy · 2026
• Fundamental optoelectronic and device physics of Cs2NaInI6 double halide perovskites from first-principles to multiscale modeling · Optical and Quantum Electronics · 2026
• Molecular-templated pre-assembly of self-assembled monolayer for perovskite solar cells and modules with improved reverse-bias stability · Nature Energy · 2026
• Simultaneous optimization of efficiency and stability in tunable HTL-free perovskite solar cells with MWCNT-integrated back contact using a machine learning-derived polynomial regressor · Sustainable Energy Technologies and Assessments · 2026
• Machine learning–enhanced SCAPS optimization of 28.38%-efficient bifacial triple-cation perovskite solar cells for building integration · Surfaces and Interfaces · 2026
• Application of functional group-modulated polymeric hole transport materials in inverted perovskite solar cells · Synthetic Metals · 2026

Related topics in Engineering