Formamidinium-based perovskites: from fundamental properties to solar cell applications

Abstract

Formamidinium (FA)-based metal halide perovskites have enabled perovskite solar cells (PSCs) to achieve certified efficiencies beyond 27%, owing in part to their near-optimal bandgaps and improved thermal stability relative to methylammonium analogs, together with advances in crystallization control and interface engineering. However, commercial deployment is hindered by the metastability of black-phase (α-phase) FAPbI₃, interfacial/defect-assisted degradation, and processing sensitivity. This review is organized around the FA⁺-regulated lattice geometry–electronic structure–defect chemistry–crystallization nexus, linking fundamental mechanisms to device efficiency and operational stability. We summarize key stabilization strategies, including A-site and X-site composition engineering, additive- and intermediate-phase-controlled crystallization, and surface/interface modulation via bulky ions and 2D/3D heterostructures. We further discuss implications for wide-bandgap tandem top cells and MA-free Sn–Pb narrow-bandgap absorbers, where halide segregation, ion migration, and Sn²⁺ oxidation remain critical. For wide-bandgap FA-based perovskites, we further emphasize the persistent tradeoff among bandgap widening, photostability, halide segregation, and quasi-Fermi-level splitting losses. Finally, we outline perspectives for scalable, reproducible, and durable FA-based photovoltaics enabled by mechanism-guided materials design and application-relevant reliability protocols.

Formamidinium-based perovskites combine near-optimal bandgaps with improved thermal stability; understanding their structure, crystallization, defects and interfaces enables efficient and durable solar cells.

1. Introduction

Over the past decade, PSCs have become one of the most influential new photovoltaic technologies. Their rise has been unusually fast: devices that were once laboratory curiosities now reach certified efficiencies above 27%, bringing them close to established solar-cell technologies.^1,2 This rapid progress matters because perovskites can absorb sunlight very strongly and can be made by relatively low-temperature, solution-based processing, opening possibilities for lightweight, flexible, and tandem solar modules.^3–8 At the same time, the field has learned that high efficiency alone is not enough; the material must also remain stable during heat, light, moisture, and long-term electrical operation.^9–12 For this reason, research has increasingly focused on choosing compositions that can deliver both performance and durability. Among the different perovskite compositions, formamidinium (FA)-based perovskites have emerged as a particularly important platform, and they form the focus of this review.

While early breakthroughs were largely driven by methylammonium lead iodide (MAPbI₃), the field has increasingly shifted toward formamidinium-based compositions to improve both efficiency and operational stability.¹³ The transition from MAPbI₃ to FAPbI₃ is primarily motivated by the intrinsic limitations of the methylammonium cation.^14,15 MAPbI₃ has a bandgap of 1.55 eV, which is wider than the ideal value for single-junction photovoltaics, and its thermal stability is compromised by the volatility of MA⁺, which can outgas at elevated temperatures. In contrast, FAPbI₃ exhibits a narrower bandgap (typically 1.43 eV to 1.48 eV), closer to the Shockley–Queisser optimum for single-junction solar cells, thereby enabling a higher theoretical short-circuit current density.^13,16 Moreover, compared with MA⁺, the FA⁺ cation interacts more strongly with the inorganic [PbI₆]⁴⁻ framework, imparting the thermal robustness required for long-term operation.^9,17 Based on these advantages, FA is widely used to construct lead- and tin-based perovskites across a range of bandgaps for photovoltaics.

Despite these optoelectronic advantages, the practical deployment of high-efficiency FA-based PSCs is fundamentally constrained by crystallographic metastability.^18,19 Owing to the larger ionic radius of FA⁺ (2.53 Å) relative to MA⁺ (2.17 Å), the Goldschmidt tolerance factor of FAPbI₃ deviates from the ideal range for a cubic perovskite lattice, introducing tensile strain within the framework.²⁰ Consequently, the photoactive black perovskite phase (α-phase) is thermodynamically metastable at room temperature and can spontaneously undergo a polymorphic transformation into the optically inactive hexagonal yellow phase (δ-phase).^18,19,21 This structural instability is further exacerbated by external stressors, including moisture, defects, and thermal fluctuations, which effectively lower the kinetic barrier to degradation. To overcome these thermodynamic and kinetic challenges, a range of sophisticated stabilization strategies have been developed. These include compositional engineering, such as incorporating smaller cations (Cs⁺ or Rb⁺) to tune the tolerance factor, as well as additive engineering that leverages Lewis bases (e.g., MACl, DMSO) to regulate crystallization kinetics and intermediate-phase chemistry.^22–25 More recent advances in surface-energy modulation and interfacial passivation have further suppressed defect-assisted phase segregation, enabling certified efficiencies of FA-based devices to surpass 27%.^1,26–32

This review aims to provide a comprehensive framework for understanding the materials physics and engineering strategies underpinning state-of-the-art formamidinium-based perovskites. We begin in Section 2 by examining how the molecular features of FA⁺ dictate crystal structure across both lead and tin perovskite families, and surveying the thermodynamic, kinetic, and surface-level strategies available for phase control. Section 3 then discusses how these carefully controlled structures translate into electronic properties, band structure, carrier dynamics, and defect chemistry, and shows that the same FA characteristics that confer outstanding photovoltaic performance also create the vulnerabilities that limit operational lifetime. Section 4 focuses on solar-cell applications, spanning normal-bandgap single-junction devices, wide-bandgap variants for tandem top cells, and MA-free Sn–Pb narrow-bandgap absorbers; we also underscore the pivotal role of interface engineering in minimizing voltage losses and improving operational stability. Finally, Section 5 provides a critical summary and outlook, emphasizing scalable processing and application-relevant reliability protocols.

Unlike previous reviews that mainly focus on compositional engineering or device performance, this Perspective takes the FA cation itself as the central organizing principle. By tracing how the molecular features of HC(NH₂)₂⁺ propagate through crystal structure into electronic properties and ultimately into device function across different bandgap regimes, we highlight which stabilization principles are generalizable and which are composition- or architecture-specific.

2. Structure of formamidinium-dominated perovskites

The unique position of FA-based perovskites among halide semiconductors stems partly from the molecular characteristics of the FA cation itself. The size, shape, hydrogen-bonding capability, and permanent dipole moment of the FA impose a distinctive structural signature on every perovskite it inhabits whether the B-site is lead or tin. This chapter examines how these intrinsic molecular features dictate crystal structure across the FA-based perovskite family, giving rise to systems that are simultaneously near-ideal for photovoltaics and precariously poised on the edge of thermodynamic stability.

2.1. Molecular uniqueness of the FA cation

2.1.1. Size and shape

The effective ionic radius of FA⁺ is approximately 2.53 Å, substantially larger than that of MA⁺ (2.17 Å) and Cs⁺ (1.67 Å).^33–35 This places FA at the upper tolerance-factor boundary for forming a three-dimensional perovskite structure, with a Goldschmidt tolerance factor very close to its upper stability limit.³⁶ Consequently, α-FAPbI₃ is inherently a thermodynamically metastable phase. Unlike the spherical Cs⁺ and the nearly spherical MA⁺, FA⁺ is a planar, resonance-stabilized molecular cation with an anisotropic geometry. Its anisotropic shape introduces a degree of local structural complexity that is absent in the more symmetric cation analogues. The fact that FA⁺ sits precisely at the tolerance-factor boundary means that FA-based perovskites are, by nature, materials poised between structural states.

2.1.2. Hydrogen-bonding capacity

The two –NH₂ termini of FA⁺ act as strong hydrogen-bond donors, forming a network of N–H⋯I interactions with the iodide ions of the [PbI₆]⁴⁻ octahedral cage. These hydrogen bonds are both stronger and more directional than the weaker, more isotropic interactions involving MA. This dual character has profound consequences: the hydrogen-bond network can kinetically trap the metastable α-phase at room temperature, yet the disruption of the same network when FA cations become orientationally disordered also provides a thermodynamic driving force for the transformation to the photoinactive non-perovskite phases. The dual role of hydrogen bonding—simultaneously stabilizing and destabilizing—highlights a recurring theme in FA-based perovskites: no structural feature is purely beneficial or detrimental.

2.1.3. Permanent dipole moment

The FA cation carries a small but finite permanent dipole moment of approximately 0.25 debye. While this value is smaller than that of MA, the planar geometry and the ability to reorient within the cuboctahedral cavity mean that the FA dipole influences the local electrostatic environment and the lattice vibrational modes in a characteristic manner. The FA dipole is often overlooked in device-level discussions, yet its influence on local electrostatics and electron–phonon coupling is pervasive. We see the dipole not merely as a static property but as a dynamic handle—one whose orientation could, in principle, be controlled by external fields to modulate local carrier energetics.

2.2. FA-based lead halide perovskites

The molecular features described in Section 2.1 translate directly into a set of crystal-structure attributes that define FAPbI₃, the most intensively studied FA-based perovskite. In the lead system, the borderline size of FA⁺ produces a rich structural landscape characterized by competing polymorphs and pronounced dynamic disorder.

2.2.1. Polymorphism: the black α-phase versus the yellow δ-phase

FAPbI₃ exists in two radically different structural forms. The photovoltaically active α-phase (black phase) consists of a three-dimensional corner-sharing network of [PbI₆]^4- octahedra, with FA⁺ residing in the interstitial cuboctahedral voids. The inactive δ-phase (yellow phase) is a one-dimensional face-sharing chain structure that crystallizes in a hexagonal non-perovskite lattice (space group P6₃/mmc). The very existence of this polymorphism is a direct consequence of FA⁺ size: the cation is too large to stabilize the corner-sharing network at room temperature, so the system tends to relax into the denser, one-dimensional arrangement of the δ-phase, which is thermodynamically favored. This constitutes the central structural dilemma of FAPbI₃: the phase that delivers the ideal bandgap and superior charge transport is not the thermodynamically stable one. Polymorphism is conventionally framed as the Achilles' heel of FAPbI₃. We counter that this structural duality should instead be viewed as a signature of an exceptionally malleable lattice—one that can access multiple connectivity patterns with modest energy penalties. The same plasticity that enables the undesirable α → δ transition is what permits the broad compositional and strain engineering that are impossible in more rigid perovskites like CsPbI₃.

The phase transition from the yellow non-perovskite phase to the black photoactive perovskite phase typically requires annealing at around 150 °C.^37,38 Moreover, the resulting black phase tends to revert to the yellow phase at room temperature.³⁴

The α to δ phase transformation is nucleation-controlled. It preferentially initiates at grain boundaries, surfaces, and other regions where the coordination environment of FA⁺ is perturbed. Moisture, light, and elevated temperature all accelerate the transition by facilitating the reorganization of the hydrogen-bond network and the collapse of the three-dimensional framework. The fact that phase degradation nucleates at specific sites rather than occurring uniformly suggests a targeted, rather than global, stabilization strategy. If the nucleation sites can be identified and individually passivated through grain-boundary-specific additives or surface terminations, one could effectively arrest the phase transition without altering the desirable bulk properties of the α-phase.

2.2.2. Dynamic disorder and structural flexibility

In the high-temperature cubic or quasicubic α-phase, FA⁺ cations are orientationally highly dynamic, undergoing rapid reorientation motions on the picosecond timescale. This dynamic disorder contributes an entropic stabilization that is essential for preserving the α-phase at elevated temperatures. The average structure, as probed by conventional X-ray diffraction, exhibits high symmetry (cubic or slightly tetragonal). Local structural probes such as pair distribution function analysis, however, reveal substantial instantaneous octahedral tilting distortions that are averaged out in the long-range picture. This globally symmetric, locally distorted character is a direct fingerprint of the FA cation and is absent in the more rigid CsPbI₃ or the less dynamically disordered MAPbI₃. It profoundly affects charge-carrier scattering, as will be discussed in Section 3. This structural dichotomy, globally ordered yet locally fluctuating, is one of the most underappreciated features of FAPbI₃. It implies that properties measured using techniques sensitive to the average structure (e.g., band structure calculations on ideal crystal structure) may fail to capture the functional consequences of local distortions. We argue that intentionally amplifying these local fluctuations, rather than suppressing them, could be a counterintuitive route to enhanced dynamic carrier protection.

2.2.3. High structural tunability – a unique asset of the FA system

Precisely because the FA cation sits at the tolerance-factor limit, the lattice of FAPbI₃ is exceptionally responsive to external chemical pressure. A-site substitution with smaller cations such as Cs⁺ or MA⁺, or X-site substitution with smaller halides such as Br⁻, can dramatically alter the degree of octahedral tilting, the phase-transition temperature, and even lock the α-phase at room temperature. This sensitivity endows the FA-based system with a structural design space that is far wider than that of its MA- or Cs-based counterparts. The very fragility of the α-phase thus becomes a handle for deliberate structural engineering, a theme that will be fully exploited in subsequent chapters. The extreme sensitivity of FAPbI₃ to compositional perturbation means that even minor additives can produce disproportionate structural effects. We see this as an invitation to move beyond the traditional A-site cations (Cs, MA, Rb) and explore a much wider chemical space—including molecular cations with tailored sizes, shapes, and functional groups, that could fine-tune the lattice in ways not possible in less responsive perovskite systems (Fig. 1 and 2).

Fig. 1. (a) Representation of unit cell structures of MAPbI₃ and FAPbI₃, with I⋯H–N hydrogen bonds indicated by dashed lines.³⁹ (b) Dipole moment of MA⁺ (left) and FA⁺ (right) with the corresponding electrostatic potential.⁴⁰ (c) Crystalline structure and polymorphic phase transitions of FAPbI₃. The non-perovskite δ phase transforms into an α phase at high temperature. When the temperature decreases, the tilting of [PbI₆]⁴⁻ can promote the formation of the β and γ phases.⁴¹ Panel (a) reproduced from ref. 39 under the terms of the Creative Commons Attribution 4.0 International licence (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/), copyright 2023, the authors. Panel (b) reproduced from ref. 40 with permission from the American Chemical Society, copyright 2025. Panel (c) reproduced from ref. 41 with permission from the American Chemical Society, copyright 2020.

Fig. 2. (a) Kinetic α–δ phase transition barrier E_b and thermodynamic phase energy difference Δδ–α of pristine FAPbI₃ and those with dopants and impurity interstitials.⁴² (b) Film formation with different Lewis bases. Molecular structures of the Lewis bases (upper panel) and corresponding adduct (middle panel) and perovskite films (lower panel). The perovskite solutions were prepared by mixing 1 mmol of FAI, PbI₂, and corresponding Lewis base in 760 mg of DMF.⁴³ Panel (a) reproduced from ref. 42 under the terms of the Creative Commons Attribution 4.0 International licence (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/), copyright 2024, the authors. Panel (b) reproduced from ref. 43 with permission from the American Chemical Society, copyright 2018.

2.3. FA-based tin halide perovskites

The structural influence of the FA cation is not confined to lead-based systems. When Pb²⁺ is replaced by the smaller Sn²⁺, the same molecular features operate in a different geometric context, providing an instructive comparison that further illuminates FA's role as a structural director.

For tin-based perovskites, the size of the A-site cation strongly influences the Goldschmidt tolerance factor and lattice symmetry as well. Because the ionic radius of Sn²⁺ is smaller than that of Pb²⁺, all-inorganic systems such as CsSnI₃ exhibit a reduced tolerance factor, leading to pronounced lattice distortion. Under inert conditions or with effective encapsulation/additive control, CsSnI₃ can be stabilized as the black orthorhombic γ phase, usually denoted as B-γ-CsSnI₃ at room temperature. The frequently observed coexistence of black and yellow phases is better described as a competition between closely related polymorphs of CsSnI₃ and degradation/oxidation pathways involving Sn²⁺ oxidation, which can generate Cs₂SnI₆. Upon cooling, the black perovskite phase undergoes a displacive symmetry-lowering transition from tetragonal B-β-CsSnI₃ to orthorhombic B-γ-CsSnI₃: the corner-sharing SnI₆ octahedral network is retained, but the octahedra tilt and distort more strongly, lowering the symmetry from tetragonal to orthorhombic. Therefore, the high defect density in CsSnI₃ is more appropriately attributed to Sn-vacancy-induced p-type self-doping and oxidation-related defects, rather than from the B-β to B-γ phase transition itself.⁴⁴

In contrast, FA⁺, with its larger ionic radius, effectively expands the lattice and suppresses octahedral tilting in FASnI₃. This effect favors retention of the photoactive black phase of FASnI₃ at room temperature, indicating favorable crystallographic phase stability. The contrast between CsSnI₃ and FASnI₃ reveals a simple but powerful design principle: the FA cation acts as a “structural expander” that compensates for the small B-site cation, pushing the tolerance factor back toward the ideal range. This size-compensation strategy may be generalizable to other small-metal perovskites beyond tin.

Beyond the geometric size effect, FA⁺ also imparts excellent thermal stability to tin perovskites through its hydrogen-bonding capability. MASnI₃ undergoes a structural phase transition from tetragonal to orthorhombic below 200 K, introducing anisotropic strain, whereas FASnI₃ maintains a single-phase structure even at temperatures as high as 150 °C. This difference arises from the stronger N–H⋯I hydrogen-bonding network formed between FA⁺ and the inorganic framework, which contributes to a more robust hydrogen-bonding environment and suppresses thermally induced structural fluctuations (Fig. 3a and b).^45,46 The thermal robustness of FASnI₃ echoes a theme from Section 2.1 that FA⁺ is not merely a passive spacer but an active structural stabilizer whose directional hydrogen bonds can lock the inorganic framework against thermal distortions. This is especially valuable in tin perovskites, where the inherently softer Sn–I bonds make the lattice more susceptible to thermally activated disorder.

Fig. 3. (a) Crystal structure of MASnI₃. (b) Crystal structure of FASnI₃. (c) Projected densities of states and electron-density distributions for MASnI₃ and FASnI₃, illustrating Sn-5s/I-5p antibonding interactions. Reproduced from ref. 67 with permission from the Royal Society of Chemistry, copyright 2017.

Comparing FA-based lead and tin perovskites reveals a coherent picture of the FA cation's structural role. In both systems, the large size and hydrogen-bonding capability of FA⁺ expand the lattice and suppress octahedral tilting, favoring retention of the three-dimensional perovskite framework. In Pb systems, this structural expansion brings the tolerance factor to the edge of metastability, creating the α/δ polymorphism that is both a challenge and an opportunity. In Sn systems, the same expansion compensates for the smaller B-site cation and stabilizes the black phase against competing polymorphs and thermal distortions. Taken together, these observations lead us to a unifying perspective: the FA cation is best understood as a universal structural director whose effects are systematically modulated by the B-site chemistry. The differences between Pb and Sn systems are not contradictions but complementary manifestations of the same underlying size and hydrogen-bonding principles.

2.4. Phase and crystallization control: from understanding to design

The structural complexity described in both lead and tin systems makes phase control a central challenge for FA-based perovskites. The question is no longer simply what structure does FA produce, but how can we steer the system toward the desired structure and keep it there? The same structural plasticity that creates the phase dilemma also provides the solution: because the FA lattice is so responsive to external perturbations, a diverse toolbox of kinetic, thermodynamic, and defect-based strategies can be deployed to direct crystallization and suppress unwanted phase transitions.

2.4.1. Thermodynamic stabilization through compositional tuning

The most direct approach to phase control is to alter the relative thermodynamic stability of the competing phases. A-site alloying with smaller inorganic cations such as Cs⁺, Rb⁺, and K⁺ increases the effective tolerance factor, contracts the perovskite cavity, and stabilizes the corner-sharing network against the δ-phase. Interestingly, larger organic cations such as guanidinium (GUA⁺) and dimethylammonium (DMA⁺) can also contribute to thermodynamic stabilization, but through a different mechanism: they increase the configurational entropy of the A-site sublattice, thereby reducing the free energy of the α-phase relative to the δ-phase.^47–49 For B-site substitution, both isovalent dopants (such as Zn²⁺, Sn²⁺, Ca²⁺, and Cd²⁺) and heterovalent dopants (such as Nd³⁺, Eu³⁺, and Yb³⁺) have been explored to enhance stability by tuning bond strengths and octahedral geometry.^50–53 For X-site engineering, halides (Br⁻, Cl⁻, and F⁻), pseudohalides (SCN⁻ and OCN⁻), and superhalide anions such as BF⁴⁻ have been introduced.^54–56 X-site substitution is generally believed to improve phase stability by strengthening Pb–X bonds, thereby increasing the activation barrier for phase transition. We note, however, that thermodynamic stabilization through compositional mixing inevitably introduces static disorder and local strain. Whether this compromises the dynamic carrier-protection mechanisms remains an open and important question—one that calls for more systematic studies correlating compositional complexity with polaron dynamics.

2.4.2. Kinetic control through crystallization engineering

Beyond shifting thermodynamic equilibria, kinetic trapping during film formation offers a complementary and often milder route to the α-phase. In FAPbI₃ precursor systems, NMP is effective not simply because of its high solvating ability, but because its carbonyl group can coordinate with PbI₂ and form hydrogen bonds with FA⁺, thereby generating NMP-containing intermediate adducts that regulate nucleation and facilitate the direct formation of dense, phase-pure α-FAPbI₃ films without the need for extensive A-site alloying.^43,57,58 Beyond NMP, coordinating solvents and additives such as 2-methoxyethanol (2ME), γ-butyrolactone (γ-GBL), γ-valerolactone (γ-GVL), and 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone (DMPU) have been shown to similarly improve film morphology and crystallinity by tuning intermediate-phase formation.^43,59 The molecular-level interplay between solvent coordination, FA⁺ hydrogen bonding, and intermediate-phase formation represents an underexploited frontier.

2.4.3. Surface and grain-boundary engineering

Given that the α to δ phase transition nucleates preferentially at surfaces and grain boundaries, local structural passivation at these sites can be disproportionately effective. The mechanism is increasingly understood at the defect level: point defects, such as iodine vacancies (V_I⁺) and iodine interstitials (I_I⁻), are particularly abundant at surfaces,^60–62 where they weaken the local Pb–I framework by increasing the vibrational amplitude of PbI₆ octahedra and lowering the energetic cost for bond distortion and lattice reconstruction.^42,63 This defect-induced “unlocking” of the octahedral network substantially reduces the kinetic barrier for the α to δ transformation.⁶⁴ An iodine-rich growth environment created by introducing excess PbI₂ can suppress V_I⁺ formation, but may concomitantly increase I_I⁻ density, which also promotes phase transition.^65,66 A dual effect highlights the need for balanced defect management rather than simple vacancy suppression. A site-specific strategy to address this challenge is the formation of low-dimensional (2D) or 2D/3D heterostructures using bulky organic spacer cations. These structures create a structurally locked shell around 3D grains, physically blocking the cooperative octahedral rearrangements required for the phase transition while simultaneously passivating the surface defects that would otherwise catalyze it.

In summary, the FA cation, through its size, shape, hydrogen-bonding capacity, and permanent dipole, imposes a distinctive structural signature on every perovskite it inhabits. In lead systems, this signature manifests as a near-ideal but metastable black phase, a globally symmetric yet locally distorted lattice, and an exceptional responsiveness to compositional engineering. In tin systems, the same cation acts as a structural expander that compensates for the small B-site, stabilizes the black phase, and enhances thermal robustness through its hydrogen-bond network. Recognizing that these structural features inevitably bring phase-control challenges, we examined the thermodynamic, kinetic, and surface-level strategies that steer crystallization toward the desired phase and suppress undesired transitions from tolerance-factor tuning and intermediate-adduct engineering to grain-boundary passivation. Across all these dimensions, FA⁺ emerges as far more than a passive A-site occupant: it is a versatile structural director whose influence can be harnessed for deliberate materials design, provided its inherent duality, the simultaneous gift of flexibility and the challenge of instability, is understood and managed.

3. Electronic structure and properties

The properties of a perovskite are encoded in its electronic structure, which in turn is dictated by the crystal framework and the dynamic behavior of the A-site cation. The large FA cation, the corner-sharing network it imperfectly stabilizes, and the dynamic disorder it introduces collectively give rise to a set of electronic characteristics that are simultaneously near-ideal for photovoltaic operation and intrinsically vulnerable to degradation. This section establishes the structure–property links that make the FA system both so attractive and so demanding.

3.1. Band structure: the structural gift of the FA cation

The ground-state electronic structure of FAPbI₃ determines which photons are absorbed, how carriers populate bands, and where the fundamental limits of device efficiency lie. In this subsection, we show that every key feature of the band structure—the gap value, its tunability, and the spin texture—traces back to the size and dynamics of the FA cation.

3.1.1. Structural origin of the near-ideal bandgap lead perovskites

As an important alternative material, FAPbI₃ has garnered significant attention due to its superior thermal stability and a bandgap (approximately 1.48 eV) that is closer to the Shockley–Queisser limit and more suitable (with further band engineering) for perovskite/CIGS tandem structures.⁶⁸ The valence band maximum is primarily derived from the strong hybridization of Pb-s and I-p states, while the conduction band minimum is dominated by empty Pb-p states. FA exerts its influence indirectly through geometric effects on the inorganic sublattice. The large size of FA⁺ expands the perovskite cavity, reducing the tilting of the [PbI₆]⁴⁻ octahedra and pushing the Pb–I–Pb bond angles closer to the ideal 180°. This greater linearity enhances the orbital overlap between Pb and I along the bridging axis, increases the bandwidth, and consequently narrows the bandgap compared to MAPbI₃. Since the bandgap is controlled by Pb–I–Pb bond angles rather than by the chemical identity of FA⁺ itself, it can in principle be modulated by any stimulus that alters the lattice geometry—not just composition but also pressure, epitaxial strain, or electric fields. This raises the intriguing, largely unexplored possibility of dynamically tunable bandgaps in FA-based devices.

3.1.2. Bandgap tunability across Pb and Sn systems

The flexible lattice of FAPbI₃ accommodates a wide range of compositional substitutions without collapsing the three-dimensional perovskite framework. Alloying at the X-site with Br⁻ progressively contracts the lattice and deepens the valence band edge (due to the higher binding energy of Br-4p orbitals), allowing continuous band-gap tuning from 1.5 eV to beyond 2.3 eV. Alloying at the B-site with Sn²⁺ lowers the conduction band edge (Sn-5p orbitals lie lower in energy than Pb-6p) and simultaneously reduces spin–orbit coupling, yielding narrow-gap compositions below 1.3 eV. Crucially, FA-based mixed-halide and mixed-metal systems retain superior structural integrity over wider composition ranges than their MA-based analogues, making FAPbI₃ the material platform of choice for tandem solar cells that require precisely engineered wide and narrow bandgaps. The wide miscibility ranges in FA-based alloys are a direct consequence of the same structural flexibility that produces polymorphism. This reinforces our perspective that the plasticity of the FA lattice is also a resource: it enables access to bandgap regions that are structurally inaccessible in more rigid perovskite families, provided we can manage the associated stability penalties.

An intriguing contrast emerges when comparing Pb and Sn perovskites. In lead-based perovskites, substituting MA⁺ with the larger FA⁺ typically narrows the bandgap, as described above. Tin-based perovskites exhibit the opposite trend: FASnI₃ shows an optical bandgap of 1.41 eV, markedly larger than that of MASnI₃ (from 1.20 to 1.30 eV), and a similar tendency has been reported in bromide analogs, suggesting a general behavior in Sn perovskites. This anomalous dependence originates from the distinct composition of the valence band maximum (VBM) in Sn perovskites, which is dominated by antibonding states derived mainly from Sn-5s and I-5p orbitals.⁶⁹ A smaller MA⁺ contracts the lattice and shortens the Sn–I bond, enhancing orbital overlap and antibonding coupling, thereby pushing the VBM upward and reducing the bandgap. By contrast, the larger FA⁺ expands the lattice, weakens Sn–I overlap and antibonding interaction, and shifts the VBM downward, leading to a wider bandgap.^67,70 This inversion of the bandgap trend is a striking demonstration that the same structural perturbation—A-site cation size increase—can produce opposite electronic outcomes depending on the B-site chemistry. For mixed Pb–Sn compositions, these opposing trends must be carefully balanced.

Beyond bandgap tuning, this VBM downshift is beneficial for suppressing the dominant defect and instability in Sn perovskites. The facile oxidation of Sn²⁺ to Sn⁴⁺ is accompanied by the formation of negatively charged tin vacancies (V_Sn), causing strong p-type self-doping and severe non-radiative recombination. By lowering the VBM, FA⁺ increases the thermodynamic formation energy of V_Sn. Accordingly, both calculations and experiments show that V_Sn forms less readily in FASnI₃ than in MASnI₃ under comparable conditions, resulting in reduced background hole density and improved resistance to oxidation. Consequently, FASnI₃ can be more readily tuned toward a weakly self-doped semiconducting state with reduced background hole density and weaker Fermi-level pinning near the VBM, which is advantageous for achieving higher open-circuit voltage and photovoltaic efficiency.⁷¹

3.1.3. Spin–orbit coupling and Rashba effects

SOC plays an important role in defining the fundamental electronic properties of these perovskite materials.⁷² Theoretical calculations show that when inversion symmetry is broken, strong SOC induces substantial splitting of the conduction band and can introduce a large bandgap correction of approximately 1 eV. From the perspective of device operation, this distinctive electronic structure is considered to potentially suppress non-radiative electron–hole recombination, thereby prolonging carrier lifetimes and improving macroscopic charge-collection efficiency.^73,74 In FAPbI₃, the orientational disorder of FA⁺ breaks local inversion symmetry on a timescale slower than electron scattering, which can enhance the Rashba effect compared to the more static CsPbI₃. A larger Rashba splitting prolongs spin lifetimes and may reduce radiative recombination rates, a double-edged sword. While some reports attribute the long photoluminescence lifetime of FAPbI₃ to this effect, direct experimental verification (e.g., spin-resolved ARPES or time-resolved Faraday rotation) is still lacking. We argue that the Rashba effect in FA-dominated perovskites is an underexplored lever for spin-based optoelectronics, not just a footnote in band-structure calculations.

3.2. Charge carriers: how FA dynamics shape transport and recombination

Once a photon creates an electron–hole pair, the FA cation continues to exert its influence, not through static fields, but through rapid reorientation dynamics. This subsection demonstrates that the dynamic behavior of FA⁺, far from being a nuisance, actively protects carriers and contributes to the exceptional charge-transport properties of FA-based perovskites.

3.2.1. Exciton dissociation and the role of dielectric screening

In FAPbI₃, the exciton binding energy (E_X) is only 10–15 meV, lower than the 20–30 meV reported for MAPbI₃ and substantially below that of CsPbI₃. This exceptionally low E_X arises from enhanced dielectric screening: the infrared dielectric response of the polar inorganic Pb–I framework provides the dominant contribution, while the electronic polarizability of the large FA⁺ cation further reduces the electron–hole Coulomb attraction. The practical consequence is that at room temperature (kT 25 meV), photogenerated carriers in FAPbI₃ are predominantly free—no thermal activation is needed for exciton dissociation.^75,76 This places FAPbI₃ firmly in the free-carrier regime under solar illumination, by passing the exciton diffusion bottleneck that limits organic photovoltaics. This is a critical advantage, but we note that the relative contributions of the inorganic lattice phonons and the organic-cation dynamics to the overall dielectric screening remain incompletely disentangled. A quantitative understanding of these contributions could guide the design of even more efficient screening environments.

3.2.2. High mobility from the octahedral network and dynamic protection

FAPbI₃ exhibits intrinsically high charge-carrier mobility and remarkably long carrier diffusion lengths, frequently exceeding several micrometers in polycrystalline thin films. These properties are again rooted in the structural consequences of the FA cation.

The near-180° Pb–I–Pb bond angles promoted by FA create a highly connected, high-dimensional electronic network. This maximizes the band dispersion and reduces the effective masses of both electrons and holes, leading to intrinsically high mobilities. Compared with MAPbI₃, where stronger octahedral tilting flattens the bands, FAPbI₃ benefits from its more expanded and less distorted inorganic framework. The correlation between bond angle and mobility is well established, but its full implications are underexploited. If mobility can be tuned by controlling octahedral tilting, through A-site engineering, substrate strain, or even light-induced lattice expansion, then carrier transport becomes a design variable rather than a fixed material parameter.

Apart from the near-ideal octahedral framework, the dynamic disorder also plays an important role in achieving the high mobility. Intuitively, one might expect that the large, dynamically disordered FA⁺ would strongly scatter carriers and reduce mobility. However, growing evidence suggests the opposite: the dynamic disorder associated with FA can provide a “dynamic protection” mechanism. The fluctuating electrostatic potential created by the moving dipoles favors the formation of large polarons. These large polarons shield carriers from scattering by charged defects and optical phonons, effectively protecting long-range transport.⁷⁷ The net result is that FAPbI₃ achieves an exceptional combination of high mobility and long carrier lifetime, which is essential for high-efficiency photovoltaics. This dynamic protection mechanism suggests a provocative idea: defect tolerance in FAPbI₃ is not solely an electronic-structure effect but has a dynamic, phonon-related component.

Benefiting from these intrinsic advantages, FAPbI₃ exhibits excellent long-range charge-transport capability.^73,74,78 Experimentally, the effective diffusion lengths of electrons and holes in FAPbI₃ exhibit a notable imbalance, with holes displaying a longer diffusion length than electrons, thereby providing a physical basis for optimizing the thickness of absorber layers in high-efficiency planar heterojunction photovoltaic devices.

Nevertheless, the dynamic dipole of FA—which has a low reorientation barrier of only 21 meV and undergoes rapid rotational motion on the picosecond timescale—creates local electrostatic potential fluctuations. These fluctuations broaden the band tail, increasing the Urbach energy (E_u) of FAPbI₃ to 15 meV, compared to 12 meV for MAPbI₃. A higher E_u is often perceived as detrimental, but in FA-based systems it may be the price for enhanced dielectric screening. The key unanswered question is whether we decouple the beneficial screening from the detrimental tail broadening, possibly by taming the orientational correlation of FA⁺via strain or molecular alloying.

3.2.3. Defect tolerance: static and dynamic contributions

The defect chemistry of perovskites has been discussed extensively; here, we focus on how FA and its related interactions influence defect behavior. The term “defect tolerance” means that even a high density of defects does not cause severe non-radiative recombination, because the defect levels are resonant with the bands or because the lattice screening is extremely efficient.

In FA-based perovskites, static defect tolerance (from the electronic structure—the antibonding character of the VBM and the strong SOC that pushes defect levels toward the bands) is only part of the story. The dynamic A-site motion may influence carrier capture by modulating local electrostatic potentials, but this contribution requires quantitative validation through temperature-dependent capture-cross-section measurements. Here, we call for a concerted effort to measure temperature-dependent capture cross-sections (from 100 K to 350 K) to unambiguously separate static and dynamic contributions. Such experiments would reveal whether the dynamic protection mechanism can be deliberately amplified through A-site engineering.

In tin perovskites, FA⁺ provides an additional defect-suppression mechanism that is conceptually distinct from the dynamic screening described above. The facile oxidation of Sn²⁺ to Sn⁴⁺ is accompanied by the formation of negatively charged V_Sn, causing strong p-type self-doping and severe non-radiative recombination. The VBM downshift induced by FA⁺ (Section 3.1.2) increases the thermodynamic formation energy of V_Sn. Accordingly, both calculations and experiments show that V_Sn forms less readily in FASnI₃ than in MASnI₃ under comparable conditions, resulting in reduced background hole density and improved resistance to oxidation. Consequently, FASnI₃ can be more readily tuned toward a weakly self-doped semiconducting state with reduced background hole density and weaker Fermi-level pinning near the VBM, which is advantageous for achieving higher open-circuit voltage and photovoltaic efficiency.

3.3. The price of dynamism: ion migration and instability

The very same structural features that endow FAPbI₃ with its outstanding properties are also responsible for its most stubborn reliability problems. In the FA system, excellence and instability are two sides of the same coin. This inseparability is the central paradox of FA-based perovskites and, in our view, the most important conceptual message of this Perspective.

3.3.1. Origins of ion migration

Iodide-ion vacancies in FAPbI₃ have low formation energies and low migration barriers (typically 0.1 to 0.6 eV).^79,80 The FA cation exacerbates this propensity. The orientationally disordered FA continuously perturbs the potential energy landscape of neighbouring iodide ions, reducing the barrier for vacancy-mediated hopping. Simultaneously, the repeated breaking and re-forming of N–H⋯I hydrogen bonds assist the iodide-hopping process. The result is a material with an intrinsically high ionic conductivity, which manifests as current–voltage hysteresis, slow photocurrent transients, and screen-charge effects that degrade device performance over time.⁸¹ The FA-assisted ion migration mechanism implies that suppressing ion motion by rigidifying the lattice, the dominant strategy to date, works against the same structural flexibility that enables high mobility and bandgap tunability. To break this, an alternative approach may be as follows: rather than blocking ion migration entirely, one could engineer preferential migration pathways that direct ions toward benign sinks, or harness the ionic current for self-healing functions that repair degraded interfaces during operation.

3.3.2. Defect-mediated phase instability

Beyond the operational hysteresis caused by ion migration, the V_I⁺ defects – formed readily due to FA dynamics – can also catalyse the α to δ phase transition. This transition transforms the high-performance semiconductor into an optically inactive, wide-band-gap (2.8 eV) non-perovskite phase, destroying device function. It is accelerated by moisture, light, and electric fields – all unavoidable operating conditions for a solar cell. First-principles calculations indicate that not all defects accelerate this process; for example, antisite defects may increase the phase-transition barrier.⁸² However, VI⁺ defects have been identified as a key trigger for phase transition. Owing to their low formation energy and low migration barrier, V_I⁺ can form readily at surfaces and can substantially lower the activation energy for conversion from the perovskite to the non-perovskite phase.^42,82 These results suggest that phase stability is governed not only by thermodynamics but also by defect-controlled kinetics. Therefore, suppressing iodide loss and vacancy accumulation, and thereby inhibiting V_I⁺ formation and migration, remains central to stabilizing the black perovskite phase and maintaining long-term device performance.

3.3.3. Light-induced halide segregation in wide-bandgap systems

In FA-based wide-bandgap perovskites such as FAPb(I_1−xBr_x)₃ with high Br content, illumination drives severe halide segregation into I-rich and Br-rich domains.^83,84 Kinetically, the large FA cation expands the lattice and softens the framework, lowering activation barriers for iodide migration. Thermodynamically, photogenerated carriers couple to octahedral breathing modes via electron–phonon interactions, generating strain fields that steer halide flux and stabilize the demixed state. This disordered band landscape funnels carriers into narrow-gap regions, causing severe V_OC losses. Crucially, this is a lattice-breathing problem, not merely ion migration; FA dynamics amplify the resonant coupling that underpins segregation. Specifically, partial substitution of FA⁺ with smaller Cs⁺ induces octahedral tilting and releases the lattice contraction strain otherwise imposed by excessive bromide incorporation. This strain relaxation serves a dual function: it enhances lattice rigidity to kinetically raise migration barriers, while simultaneously widening the bandgap physically and directly attenuating the thermodynamic driving force for photocarrier–phonon coupling. Thus, A-site engineering modulates both kinetic barriers and thermodynamic resonance conditions. Building directly upon these compositional insights, we advocate advancing toward lattice-dynamics engineering—e.g., tuning phonon spectra to detune breathing modes from carrier resonance, or introducing A-site vibrational decouplers—as a complementary frontier that targets the root photochemical driving force rather than kinetics alone.⁸⁵

In summary, the FA cation endows FAPbI₃ with a suite of near-ideal photovoltaic properties, an optimal bandgap, high carrier mobility, long diffusion length, and intrinsic defect tolerance. Yet these very attributes share a common structural origin with the material's key vulnerabilities: thermodynamic metastability, facile ion migration, and photo-induced compositional instability. This inherent duality is the central paradox of FA-based perovskite research. Recognizing that the advantages and weaknesses are homogeneously rooted in the FA cation is the essential starting point for the rational structural design strategies discussed in the subsequent section, where the goal becomes to preserve the benefits while suppressing the detrimental consequences. More fundamentally, this recognition invites a change in mindset that we believe is overdue: from seeing instability as a problem to be eliminated, to seeing it as a consequence of the same structural richness that enables outstanding performance and therefore as something to be navigated, managed, and ultimately designed around, rather than eradicated. The FA cation, in this light, is not a liability to be corrected but a design tool to be mastered (Fig. 4).

Fig. 4. Calculated band structures and densities of states (DOS) of (a) MAPbI₃ and (b) α-FAPbI₃, with the indicated interband-transition energies. Reproduced from ref. 169 with permission from AIP Publishing, copyright 2017.

4. Formamidinium-based solar-cell applications

The preceding sections establish a unified mechanistic picture of FA-based perovskites: the large size, directional hydrogen bonding, and dynamic orientational behavior of FA⁺ reshape the metal-halide framework, enabling favorable bandgaps and efficient carrier transport while simultaneously introducing phase metastability, defect-assisted ion migration, and composition-dependent degradation pathways. Accordingly, the application of FA-based absorbers cannot be understood simply as an optimization of device efficiency. Across normal-bandgap Pb-based cells, wide-bandgap mixed-halide top cells, and narrow-bandgap Sn–Pb bottom cells, the central task is to translate control over lattice geometry, intermediate-phase conversion, vacancy formation, redox chemistry, and interfacial energetics into reduced non-radiative recombination and durable operational stability. The following subsections therefore examine how different device architectures manage the structural and dynamic duality introduced by FA⁺.⁸⁶

4.1. Normal-bandgap perovskite solar cells

FA-based lead halide perovskites with a normal bandgap of approximately 1.48 eV have emerged as core absorber materials for high-performance PSCs, owing to their strong optical absorption, tunable electronic structure, and intrinsically improved thermal stability. Despite the outstanding optoelectronic properties of FAPbI₃, its practical thin-film fabrication is confronted with pronounced crystallographic challenges. The relatively large ionic radius of the FA⁺ cation introduces substantial lattice strain within the [PbI₆]⁴⁻ octahedral framework. According to the Goldschmidt tolerance factor, FAPbI₃ lies close to the upper stability limit of the perovskite structure (t ≈ 1.03), rendering its crystal lattice highly sensitive to structural perturbations.⁸⁷

To address crystallization challenges associated with FAPbI₃, a variety of strategies have been developed to regulate nucleation and crystal-growth pathways. During solution processing, precursor species typically form Lewis acid–base adducts between PbI₂ and polar solvents such as N,N-dimethylformamide (DMF) or dimethyl sulfoxide (DMSO) (e.g., PbI₂·DMSO).⁸⁸ These solvated adducts are stabilized in the early stages of film formation and can be converted into the perovskite lattice upon antisolvent extraction and thermal annealing. By tailoring the solvent system, additives, and processing sequence, the lifetime and conversion pathway of intermediate phases can be controlled, enabling improved film quality and phase purity.

Additives can reduce the critical Gibbs free energy for nucleation and increase the nucleation rate by altering interactions in the precursor solution, leading to more uniform nucleation and oriented crystallization. They can also regulate the crystallization pathway through specific chemical interactions and dynamically passivate halide-vacancy defects at grain boundaries and film surfaces, thereby reducing defect density and improving stability. Growing evidence indicates that chloride-containing additives can promote perovskite crystallization and improve film quality. In 2019, Kim et al. elucidated the mechanistic origin of methylammonium chloride (MACl) as a crystallization-transient agent, highlighting stabilization of a chloride-mediated intermediate phase during film formation.⁸⁹ Specifically, MACl lowers the formation energy of α-FAPbI₃ and induces a stable intermediate, enabling phase-pure α-FAPbI₃ to form with high fidelity even prior to post-annealing via transient A-site cation exchange/substitution. In addition to accelerating grain coarsening, MACl leverages the volatility of Cl⁻-containing species upon annealing to yield impurity-free, highly crystalline α-FAPbI₃ films with prolonged carrier lifetimes. Subsequently, Zhao et al. addressed the instability associated with excess PbI₂ in FAPbI₃ films by converting photochemically active PbI₂ into an inactive secondary phase, (PbI₂)₂RbCl, thereby suppressing ion migration and stabilizing the perovskite layer.⁹⁰ They proposed an inactive secondary phase conversion (ISPC) strategy based on controlled RbCl incorporation, which converts reactive excess PbI₂ at grain boundaries in situ into a chemically inert compound, (PbI₂)₂RbCl. Because of its intercalated structure, this secondary phase exhibits enhanced thermal and photostability and can interrupt PbI₂-initiated degradation cascades while also acting as a barrier to ion transport. Furthermore, a fraction of Cl⁻ can form strong Pb–Cl bonds at the perovskite surface, further anchoring the lattice and suppressing ion migration.

Additionally, pseudohalide anion engineering has also emerged as an effective route to regulate crystallization and suppress defect formation in halide perovskite films. Jeong et al. introduced the pseudohalide formate (HCOO⁻) to mitigate anion-vacancy defects located at grain boundaries and the film surface, thereby improving the crystallinity of FAPbI₃ thin films.⁹¹ Owing to its sufficiently small ionic size and strong affinity to under-coordinated Pb sites, HCOO⁻ can effectively occupy iodide-vacancy-related sites and passivate the ubiquitous halide-vacancy defects in metal-halide perovskites. Such defects are known to accelerate trap-assisted non-radiative recombination of photogenerated carriers, which in turn compromises the open-circuit voltage (V_OC) and fill factor (FF) of PSCs. Similarly, Lu et al. demonstrated that thiocyanate (SCN⁻) introduced from MASCN can promote the rapid formation of the photoactive perovskite phase by lowering the formation energy of key intermediates and growth fronts during crystallization.⁵⁵ The presence of SCN⁻ markedly accelerates grain growth kinetics, leading to enlarged grain size and conformal deposition of phase-pure perovskite films with high crystallinity, reduced defect density, and improved phase purity. Building on these advances, Peng et al. proposed a kinetics-regulation strategy based on in situ surface ion exchange.⁹² By introducing cesium thiocyanate (CsSCN) as an additive into the perovskite precursor, they exploited the partial volatilization of formamidinium species during annealing to in situ generate A-site vacancies at the grain surface. Subsequently, Cs⁺ ions from CsSCN spontaneously occupied these surface vacancies, thereby constructing a thermodynamically more stable, Cs-enriched shell Cs_xFA_1−xPbI₃ on the surface of bulk Cs_0.15FA_0.85PbI₃ grains and forming a distinctive core–shell heterostructure. This Cs-rich surface layer substantially increases the formation energies of surface iodide vacancies and FA vacancies, effectively passivating surface defects. Meanwhile, the compact inorganic-rich composition serves as a robust barrier that suppresses internal ion-migration pathways, contributing to improved operational stability. In addition, the incorporation of SCN⁻ anions suppresses premature nucleation at the early stage of film formation and retards crystallization kinetics, which favors grain coarsening and ultimately enables the synergistic enhancement of both device efficiency and stability.

Distinct from conventional lattice doping, Pan et al. developed a surface-energy regulation strategy using a bulky propeller-like halogenated tertiary ammonium iodide molecule, TCMAI.⁹³ Mechanistic studies showed that TCMAI strongly binds to the perovskite surface via halogen bonding, while its large steric hindrance prevents lattice incorporation and impurity formation. Instead, TCMAI accumulates at the surface and lowers the surface energy of the photoactive α-FAPbI₃ phase, making α-phase nucleation thermodynamically favorable. This enables room-temperature prenucleation of α-phase seeds during spin coating, which subsequently template the direct formation of the black phase during annealing while bypassing the δ-phase pathway. Consequently, residual yellow-phase domains are suppressed, grain coalescence is enhanced, and high-quality films with larger grains and reduced lattice strain are obtained.

More sophisticated crystallization control can be achieved by introducing organic cations that transiently stabilize low-dimensional intermediate structures. For instance, dimethylammonium (DMA⁺) can induce the formation of quasi-two-dimensional intermediates during the initial annealing stage.⁹⁴ Upon further heating, these intermediates gradually decompose and release the templating species, thereby epitaxially guiding the oriented growth of phase-pure α-FAPbI₃. This approach effectively enables a kinetic bypass of the thermodynamically favored but optoelectronically inactive δ-phase, offering a powerful route toward high-purity black-phase films. Nevertheless, for industrial fabrication DMA should be avoided due to its significant health risks.

Building on an improved understanding of phase-transition kinetics, multiple crystallization-regulation strategies have been proposed. On the one hand, multi-cation alloying—incorporating species such as Cs⁺, Rb⁺, or divalent cations like methylenediammonium (MDA²⁺)—has been employed to finely tune lattice parameters and mitigate the lattice strain induced by FA⁺. For example, Kim et al. demonstrated that equimolar co-alloying of Cs⁺ and MDA²⁺ can synergistically neutralize lattice strain, resulting in a pronounced extension of carrier lifetime and enhanced structural stability.⁹⁵ On the other hand, antisolvent engineering has been widely adopted to precisely control solution supersaturation, enabling uniform nucleation and controlled crystal growth while avoiding rapid local crystallization and phase coexistence.

In addition, interfacial energy engineering has emerged as an effective approach for stabilizing the α-phase of FAPbI₃. By introducing bulky organic cations such as phenethylammonium (PEA⁺) or butylammonium (BA⁺) at the film surface or grain boundaries, a thin two-dimensional perovskite overlayer can spontaneously form.⁹⁶ This 2D layer lowers the surface free energy of the α-phase, making it energetically more favorable than the δ-phase, while physically encapsulating grain boundaries to impede moisture/oxygen ingress and phase-propagation pathways. Representative studies include the work by Luo et al.,⁹⁷ who constructed a continuous 2D/3D heterostructure via solid-state in situ growth to stabilize phase-pure α-FAPbI₃ films, and by Tan et al.,⁹⁸ who used self-assembled 2D layers to induce highly oriented (100) growth of underlying FAPbI₃, improving crystalline order and charge-transport properties (Fig. 5).

Fig. 5. (a) Categorization of commonly used solvent systems in solution-processed film preparation. Abbreviations: dimethylformamide (DMF), dimethyl sulfoxide (DMSO), N-methylpyrrolidone (NMP), isopropanol (IPA), acetonitrile (ACN), tetrahydrofuran (THF), diethyl ether (EE), chlorobenzene (CB), toluene (TL).⁵⁹ (b) Calculations of the α-phase FAPbI₃ perovskite structure and that prepared with Cl, MA, or MACl. Formation energies (A) and highest-occupied molecular orbitals (HOMOs).⁸⁹ (c) ¹³C solid-state nuclear magnetic resonance, NMR spectra (recorded at 100 K and 12 kHz magic-angle spinning (MAS)) of FAHCOO (1), δ-FAPbI₃ (2), α-FAPbI₃ (3) and α-FAPbI₃ + 5% FAHCOO (4) (the top trace in (4) is an eightfold magnification).⁹¹ (d) The defect formation energy of Schottky vacancy defects of Cs-random and Cs-rich shell structures.⁹² (e) Schematic illustration of the energy barrier evolution under the different conditions (top); binding energy of α-FAPbI₃ with TCMA⁺, FA⁺, and DMSO (bottom).⁹³ Panel (a) reproduced from ref. 59 with permission from the American Chemical Society, copyright 2025. Panel (b) reproduced from ref. 89 with permission from Elsevier, copyright 2019. Panel (c) reproduced from ref. 91 with permission from Springer Nature, copyright 2021. Panel (d) reproduced from ref. 92 under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence (CC BY-NC-ND 4.0; https://creativecommons.org/licenses/by-nc-nd/4.0/), copyright 2020, the authors. Panel (e) reproduced from ref. 93 with permission from Wiley-VCH GmbH, copyright 2023.

To illustrate how recent crystallization and interfacial engineering strategies translate into actual device performance, Table 1 summarizes the state-of-the-art normal-bandgap FA-based perovskite solar cells. By detailing the absorber compositions, specific additives, and key photovoltaic parameters (V_oc, J_sc, PCE and FF), this table highlights the precise efficacy of current defect-management protocols.

Table 1. A comparative outline of the latest presented perovskite solar cells with normal-bandgap.

Year	Main composition	Additive	V oc (v)	J sc (mA cm−2)	FF (%)	PCE (%) (certified)
Jul-2023	Cs0.05FA0.95PbI3	AICl	1.17	26.22	82.20	25.12 (24.6)99
Jul-2023	Cs0.02FA0.98PbI3		1.18	25.88	82.50	25.1 (24.5)100
Nov-2023	FAPbI3	PbI2·MACl	1.16	26.15	84.13	25.56 (25.2)101
Nov-2023	Cs0.05FA0.95PbI3	MAPbBr3·PbI2·MACl	1.16	26.14	85.74	26.09 (25.8)102
Jan-2024	Cs0.1FA0.9PbI3	PbCl2	1.17	25.67	85.30	25.7 (24.8)103
Jan-24	Cs0.03FA0.97PbI3	CsSCN·TCMAI	1.13	24.98	83.95	23.6 (23.2)93
Feb-2024	FAPbI3	PbI2·MACl	1.18	25.55	83.61	25.10 (ref. 104)
Mar-2024	Cs0.05FA0.95PbI3	AMP	1.17	25.77	84.02	25.35 (24.87)105
Mar-2024	Cs0.05FA0.95PbI3	PbI2	1.18	25.91	81.86	25.03 (24.65)106
May-2024	MA0.7FA0.3PbI3		1.19	25.90	81.40	25.1 (25)107
Jun-2024	Cs0.05FA0.95PbI3	KPF6	1.18	26.11	84.20	26.03 (25.51)108
Jun-2024	Cs0.05FA0.95PbI3	PbI2·MACl	1.20	26.30	84.50	26.69 (26.54)109
Sep-2024	Cs0.05FA0.95PI3	PbCl2	1.17	26.07	85.06	25.92 (25.77)110
Oct-2024	Cs0.05FA0.95PbI2.94Br0.06	PbI2·MACl·PACl	1.19	25.77	81.75	25.07 (ref. 111)
Oct-2024	Cs0.05FA0.95PI3		1.22	24.88	83.90	25.59 (ref. 112)
Oct-2024	Cs0.05FA0.95PbI3		1.18	26.07	84.70	26.1 (25.39)113
Oct-2024	Cs0.05FA0.95PbI3	MACl·PbI2	1.20	26.34	85.20	27.02 (26.88)114
Sep-2024	Cs0.1FA0.9PbI3	PbCl2	1.17	25.61	85.10	25.57 (ref. 115)
Sep-2024	Cs0.1FA0.9PbI3	MACl·PbI2	1.17	25.84	83.68	25.38 (ref. 116)
Oct-2024	Cs0.05FA0.95PbI3	MACl·PbI2	1.19	26.18	85.76	26.81 (26.02)117
Oct-2024	Cs0.05FA0.95PbI3	PbCl2	1.19	26.00	84.27	26.12 (26.01)118
Oct-2024	Cs0.05FA0.95PbI3	PbCl2·MACl	1.18	26.38	85.53	26.38 (ref. 119)
Oct-2024	FAPbI3	MACl	1.20	25.76	85.62	26.37 (26.2)120
Jan-2025	FAPbI3	I2	1.13	26.25	81.20	24.10 (ref. 121)
Jun-2025	Cs0.1FA0.9PbI3	PbI2·MACl	1.21	25.57	83.80	26.01 (ref. 122)
Aug-2025	Cs0.05FA0.95PbI3	M4CH	1.18	25.10	82.70	24.52 (ref. 123)
Oct-2025	Cs0.05FA0.95PbI3	MACl	1.18	26.20	85.93	26.62 (26.4)124

Normal-bandgap FA-based devices provide the most direct demonstration of how the structural features discussed in Sections 2 and 3 translate into photovoltaic performance. Because their high efficiency relies on preserving the metastable α-phase while maintaining a low-defect electronic landscape, crystallization and interface treatments should be assessed not only by their ability to increase the initial PCE, but also by whether they suppress phase reconstruction and ion migration without compromising carrier extraction.⁸²

Current perovskite additives and interfacial treatment strategies generally lack universality and are often effective only for specific perovskite compositions. Fundamentally, the overall requirement can be reduced to two key objectives: reducing the density of bulk and surface defects in the perovskite crystal, and enhancing interfacial charge-carrier extraction. A small amount of PbI₂, MACl, and low-dimensional layers, as well as Lewis bases or salt additives, can regulate crystallization, passivate defects, suppress non-radiative recombination, and inhibit ion migration when their dosage and spatial distribution are properly controlled. At the device level, these effects are typically manifested as improvements in Voc, FF, and operational stability. However, when these species remain in excess, aggregate at the surface or buried interfaces, or form thick insulating secondary phases, their role may shift from defect passivation to recombination centers or transport barriers, resulting in reduced J_sc, FF, and even long-term stability.

The beneficial effect of PbI₂ relies on its presence in small amounts, with a dispersed distribution and controlled spatial location; in contrast, randomly distributed excess PbI₂ may introduce deep-level interfacial states and promote photo-induced degradation.^125,126 MACl is more suitable as a volatile crystallization modulator for FAPbI₃, whereas residual MA⁺ represents a potential risk for thermal and photostability.^127,128 Low-dimensional layers or low-dimensional-templated growth can effectively passivate interfaces and suppress ion migration, but these layers must remain ultrathin and favorably oriented; otherwise, they may impede out-of-plane charge extraction. In addition, 2D/3D structural reconstruction during operation may alter the initially designed interfacial chemistry.^129–131 For Lewis bases, small molecules, and salt additives, the key lies in controlling coordination strength and size matching. Polymer additives, by contrast, are more closely associated with mechanical robustness and moisture–thermal stability engineering, but the formation of continuous insulating phases must be avoided.^132–135

Therefore, the design of additives should ensure, as far as possible, that bulk additives can volatilize effectively without leaving harmful residues, while interfacial additives should be thin and well ordered. Stability evaluation should also be conducted under combined thermal, illumination, and bias stress conditions. Only additives that simultaneously suppress recombination, maintain efficient charge transport, and avoid chemical reconstruction during operation can convert initial efficiency gains into genuine long-term reliability improvements.

4.2. FA-based wide-bandgap perovskite solar cells

FA-based perovskites have become key absorbers for wide-bandgap (WBG) top sub-cells in high-efficiency tandem photovoltaics, largely because of their superior thermal stability relative to MA analogs. Nevertheless, deploying FA-dominated compositions in WBG devices introduces severe crystallographic and thermodynamic challenges. Pure α-phase FAPbI₃ is thermodynamically metastable at room temperature and can spontaneously transform into the non-photoactive yellow δ-phase. This intrinsic instability is exacerbated in mixed-halide (I⁻ and Br⁻) WBG systems, where incorporation of smaller bromide ions induces local lattice strain that can lower the barrier for phase transition and accelerate degradation under humid and thermal stress. Furthermore, mixed-halide WBG perovskites often suffer from photoinduced phase segregation (the Hoke effect) under illumination, where halide migration produces iodide-rich, lower-bandgap domains that act as recombination centers.^83,136 Consequently, stabilizing the FA host while suppressing ion migration remains a critical bottleneck for long-term operational stability.

To address these instability issues, A-site cation engineering, specifically the partial substitution of FA with Cs, to form FAC mixed-cation systems, has proven to be a robust strategy for enhancing structural integrity. From a geometric perspective, incorporating the smaller Cs⁺ into the FA lattice effectively tunes the Goldschmidt tolerance factor towards the ideal cubic regime. This modification, coupled with the increased configurational entropy of the system, thermodynamically stabilizes the photoactive black phase and suppresses the formation of the δ-phase. Even more critical is the role of A-site tuning in mitigating photoinduced phase segregation. The introduction of Cs⁺ induces a tilting of the lead-halide octahedra; this lattice distortion alters the orbital overlap, resulting in a physical widening of the bandgap without necessitating higher bromide concentrations.^137,138

Leveraging the bandgap-widening effect of A-site doping, researchers have successfully adopted a “cesium-rich, bromine-poor” compositional strategy to achieve target bandgaps (1.70 to 1.80 eV). This approach fundamentally weakens the thermodynamic driving force for halide segregation and significantly mitigates non-radiative recombination losses induced by deep-level trap states. In-depth mechanistic studies further reveal that optimizing the FA⁺ to Cs ratio yields a synergistic effect, effectively releasing the lattice contraction strain caused by halide mixing. This strain relaxation enhances lattice rigidity and increases the energy barrier for ion migration. In conclusion, A-site compositional engineering is not merely a method for maintaining crystallographic stability but a central dimension of regulation for balancing bandgap tuning, suppressing phase segregation, and realizing high voltage output in tandem devices (Fig. 6).^34,139

Fig. 6. (a) Variation of internal energy (ΔE), mixing-entropy contribution (−TΔS), and free energy (ΔF = ΔE − TΔS) as a function of Cs content.¹³⁹ (b)(a) Full width at half maximum (FWHM) of the (100) XRD peak in C_syFA_1−yPb(Br_0.4I_0.6)₃ as a function of Cs content (0.0 < y < 0.8); the solid line is a guide to the eye. (b) Effective charge-carrier mobilities of C_syFA_1−yPb(Br_0.4I_0.6)₃ with varying Cs content (error bars show the range obtained). (c) Charge-carrier lifetimes of Cs_yFA_1−yPb(Br_0.4I_0.6)₃ films with varying Cs content obtained from time-correlated single-photon counting (TCSPC) at low excitation fluence (λ_exc = 400 nm); the grey shaded area marks the Cs-content window associated with high crystallinity and stability.³⁴ Panel (a) reproduced from ref. 139 with permission from the Royal Society of Chemistry, copyright 2016. Panel (b) reproduced from ref. 34 with permission from the Royal Society of Chemistry, copyright 2017.

The Albrecht group reported that piperazine iodide (PI), when introduced either as a bulk dopant or as a surface-treatment agent, can effectively reduce residual/excess PbI₂ in perovskite films, thereby lowering non-radiative recombination losses and extending carrier lifetimes.¹⁴⁰ In a related solvent-engineering strategy, Zhao's group showed that partially replacing dimethyl sulfoxide (DMSO) with 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone (DMPU) markedly optimizes perovskite crystallization. Because DMPU exhibits stronger coordination with PbI₂ than DMSO, this tailored solvent system can effectively retard and regulate nucleation and crystal growth, leading to enlarged grains and a substantially reduced defect density.¹⁴¹

The acid dissociation constant (pK_a) plays a pivotal role in determining the strength of N–H bonds and the deprotonation propensity of organic cations—both of which are closely associated with the thermally induced performance degradation of PSCs.^142,143 Taking a “reverse-design” approach, the De Wolf group introduced a THTZ–H⁺ cation capable of forming multidirectional hydrogen bonds with iodide. By in situ generating THTZ–H⁺ through the incorporation of low-pK_a MDACl₂ in the precursor solution, they significantly enhanced the photothermal stability of wide-bandgap perovskites.¹⁴⁴

Beyond bulk and solvent design, interfacial crystallization regulation has also proven highly effective. Ning and co-workers identified severe Br/I halide phase segregation during the initial nucleation of wide-bandgap perovskites on self-assembled monolayer substrates. By introducing a bisphosphonic-acid molecule, p-xylylenebisphosphonic acid (PXA), which strongly coordinates with PbBr₂, they moderated the crystallization-rate mismatch between Br-rich and I-rich domains, suppressed segregation, and improved film crystallinity and device performance.¹⁴⁵

In addition, Li and co-workers incorporated the pseudohalide thiocyanate (SCN⁻) into wide-bandgap perovskites and found that it promotes crystallization and reduces grain-boundary density. A trace amount of SCN⁻ can enter the lattice to form an I/Br/SCN alloy, occupy iodide-vacancy-related sites, and, through steric effects, hinder halide-ion migration, collectively contributing to enhanced operational stability and performance.¹⁴⁶

The performance bottleneck of wide-bandgap mixed-halide PSCs essentially originates from the non-equilibrium crystallization and operational ionic instability of Br/I mixed systems. Although a high Br content can widen the bandgap and meet the requirements of tandem top cells, it also introduces rapid nucleation, non-uniform vertical halide distribution, lattice strain, halide vacancies, and deep defects, thereby triggering Br/I phase segregation and ion migration under illumination and electric fields. The I-rich low-bandgap domains generated by phase segregation can lead to carrier funneling, localized non-radiative recombination, and a reduced effective bandgap. Meanwhile, mobile ions accumulate at interfaces, causing screening of the built-in electric field, QFLS–V_OC mismatch, and interfacial chemical reactions, which ultimately manifest as a severe V_OC deficit and degraded operational stability.¹⁴⁷

Single defect-passivation strategies are no longer sufficient to fully address these issues. A more effective future direction should involve the synergistic design of composition homogenization, synchronized crystallization, enhanced ion-migration barriers, redox stabilization, and optimized selective-contact interfaces (Fig. 7).

Fig. 7. (a) Schematic interactions between S O in DMSO and Pb²⁺ in PbI₂, and between C O in DMPU and Pb²⁺ in PbI₂.¹⁴¹ (b) Expanded 1H NMR signals corresponding to THTZ–H⁺ in scratched perovskite films dissolved in DMSO-d⁶ and in solutions derived from MDACl₂:FAI powder.¹⁴⁴ (c) Calculated coordination diagrams between PXA and lead iodide, lead bromide, and formamidinium ions.¹⁴⁵ (d) Photoluminescence evolution of Cs_0.2I_1.6 films and pseudo-triple-halide-alloy films under 100 sun illumination for 20 min.¹⁴⁶ Panel (a) reproduced from ref. 141 with permission from the Royal Society of Chemistry, copyright 2024. Panel (b) reproduced from ref. 144 with permission from the American Association for the Advancement of Science, copyright 2024. Panel (c) reproduced from ref. 145 with permission from Wiley-VCH GmbH, copyright 2025. Panel (d) reproduced from ref. 146 with permission from Springer Nature, copyright 2024.

4.3. MA-free narrow-bandgap perovskite solar cells

The role of FA⁺ in narrow-bandgap Sn–Pb absorbers is complementary to its role in Pb-only systems. Rather than primarily addressing the α to δ phase transition encountered in FAPbI₃, FA⁺ provides a thermally more robust A-site environment for an oxidation-sensitive Sn²⁺/Pb²⁺ alloy while also helping to mitigate vacancy-related self-doping and structural distortion.¹⁴⁸

While the limitations of MA⁺ are well-documented in pure tin perovskites, its presence in narrow-bandgap Sn–Pb alloyed systems introduces even more severe physicochemical instabilities. The alloying of Sn²⁺ and Pb²⁺ intrinsically generates local microstrain within the lattice due to the atomic size mismatch between the two metal cations. The small ionic radius of MA⁺ provides limited steric stabilization to maintain the structural rigidity of this heavily strained mixed-metal framework.¹⁴⁹ Consequently, MA-containing Sn–Pb perovskites are structurally vulnerable and highly susceptible to phase segregation under operational thermal and illumination stresses.

Furthermore, the thermal volatility of MA⁺ becomes particularly problematic in the Sn–Pb matrix.¹⁵⁰ During crystallization and thermal annealing, the facile volatilization of MA⁺ leaves behind a high density of A-site vacancies, predominantly at grain boundaries and film surfaces. In pure lead systems, such vacancies might act merely as shallow traps; however, in Sn–Pb alloys, these uncoordinated sites serve as facilitated diffusion pathways for oxygen and moisture. Because the chemical environment of Sn²⁺ is already perturbed by adjacent Pb atoms, these MA-induced structural voids lower the activation energy required for oxidation. This accelerates the degradation of Sn²⁺ to Sn⁴⁺, leading to uncontrollable p-type self-doping and rapid device failure.

Driven by these severe degradation pathways, researchers have increasingly pivoted toward MA-free Sn–Pb perovskite compositions. However, the complete removal of MA⁺ introduces a new set of challenges for methylammonium-free Sn–Pb perovskites: crystallization kinetics in the precursor solutions are markedly faster and more difficult to regulate, leading to nonuniform nucleation and inferior film morphologies with small grains, high grain-boundary densities, and incomplete surface coverage.^151,152 Meanwhile, the crystallization mismatch between Sn- and Pb-based perovskites can further aggravate defect formation. The resulting grain boundaries and structural defects accelerate Sn²⁺ oxidation, giving rise to severe p-type self-doping and deep-level trap states.^153–155 Consequently, MA-free Sn–Pb perovskite devices generally exhibit lower efficiencies than their MA-containing counterparts.¹⁵⁰ To address these issues, a variety of strategies have been proposed. Zhao et al. introduced a noncovalent binder that targets open metal sites of coordinatively unsaturated Sn²⁺ solvate species, thereby synchronizing crystallization kinetics and promoting homogeneous Sn–Pb alloying. This approach mitigates chemical heterogeneity and optoelectronic defects that otherwise arise from metal-ion-induced nonuniform nucleation and rapid crystal growth. They further attributed the intrinsically fast crystallization of Sn-based perovskites to the stereochemically active lone pair of Sn²⁺, which disfavors coordination between Lewis-base ligands and metal cations in the precursors. With this treatment, the authors fabricated Sn–Pb single-junction PSCs with a certified PCE of 24.13%.¹⁵⁶

Similarly, to suppress the excessively fast crystallization and the facile oxidation in MA-free Sn–Pb perovskites, Liao et al. employed rubidium acetate (RbAc) as a multifunctional additive. As illustrated in Fig. 8, the Ac⁻ strongly coordinates with Sn²⁺, effectively inhibiting its oxidation and retarding crystallization, whereas Rb⁺ incorporation into the lattice releases microstrain and passivates deep-level defects. This strategy significantly improves carrier lifetime and film quality.¹⁵⁷ Turren-Cruz et al. proposed a multicomponent synergistic approach, in which Rb⁺ was used to optimize crystallization and enlarge grain size; bulky dipropylammonium cations were introduced to form low-dimensional phases at grain boundaries and surfaces as protective barriers, and sodium borohydride served as a reducing agent to directly suppress Sn²⁺ oxidation.¹⁵⁸ Moreover, to overcome the uncontrolled crystallization associated with antisolvent-free processing, Zhang et al. developed a dual-anion synergistic regulation strategy based on FAAc and guanidinium thiocyanate (GuaSCN).¹⁵⁹ Here, Ac⁻ regulates crystallization kinetics by forming intermediate phases and enabling anion-exchange processes while suppressing Sn²⁺ oxidation, whereas thiocyanate (SCN⁻) further promotes large-grain growth and stabilizes Sn²⁺.

Fig. 8. (a) FTIR spectra of neat DMSO, and DMSO solutions of PPD, PbI₂ + PPD, SnI₂ + PPD, and SnI₂ + PbI₂ + PPD.¹⁵⁶ (b) Mechanism of crystallization improvement.¹⁵⁷ (c) X-ray diffraction (XRD) for MA-free CsFA-based perovskite films featuring varying concentrations of rubidium.¹⁵⁸ (d) Curves of the (100) peak intensity during in situ GIWAXS measurements over time.¹⁵⁹ Panel (a) reproduced from ref. 156 under the terms of the Creative Commons Attribution 4.0 International licence (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/), copyright 2024, the authors. Panel (b) reproduced from ref. 157 under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence (CC BY-NC-ND 4.0; https://creativecommons.org/licenses/by-nc-nd/4.0/), copyright 2025, the authors. Panel (c) reproduced from ref. 158 under the terms of the Creative Commons Attribution 4.0 International licence (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/), copyright 2024, the authors. Panel (d) reproduced from ref. 159 with permission from Wiley-VCH GmbH, copyright 2025.

Overall, MA-free Sn–Pb narrow-bandgap perovskites face a central challenge triangle: synchronized Sn–Pb crystallization, suppression of Sn²⁺ oxidation and vacancy-induced self-doping, and interface or grain-boundary passivation without introducing charge-transport barriers. Future progress should therefore move beyond single-additive passivation toward integrated designs that simultaneously control precursor coordination, redox chemistry, microstrain, ion migration, and selective-contact energetics.

4.4. Module fabrication of FA-based perovskites

The current commercialization focus of PSCs is on the reproducible fabrication of large-area modules. For FA-based perovskites, black-phase formation, intermediate-phase evolution, and film crystallization are highly sensitive to humidity, solvent evaporation, and processing atmosphere. Therefore, the industrialization of FA-based perovskites requires not only high-efficiency material systems, but also air-compatible processing, uniform large-area film formation, and module-level stability. Scalable deposition, module architecture, operational stability, and techno-economic viability are therefore central issues for the commercialization of perovskite photovoltaics.^160,161

First, from the perspective of the processing atmosphere, perovskite fabrication is gradually moving from nitrogen-filled gloveboxes toward dry-air or controlled ambient-air processing. Air processing does not imply fabrication under completely uncontrolled open-air conditions; rather, it requires the establishment of a well-defined humidity window. An appropriate amount of moisture can participate in the formation of lead complexes and intermediate phases, thereby regulating nucleation. However, excessive humidity or poorly controlled processing can lead to discontinuous films, increased defect density, and phase instability. Previous studies have shown that ambient water can be converted from a detrimental factor into a beneficial component that promotes uniform nucleation, enabling the fabrication of homogeneous perovskite films under ambient-air conditions.¹⁶² For FA-based systems, the challenge of air processing is further concentrated on the formation of the black α-FAPbI₃ phase and the transformation of intermediate phases. Liu et al. achieved the fabrication of α-FAPbI₃ photovoltaic devices in ambient air through a DMSO-extraction strategy.¹⁶³

Second, large-area deposition represents another key direction for commercialization. Blade coating and slot-die coating offer high material utilization, strong potential for continuous manufacturing, and compatibility with module fabrication. Their major challenges, however, lie in controlling wet-film flow, solvent evaporation, synchronized nucleation, and crystallization uniformity across the entire substrate. Yang et al. proposed a wide-processing-window perovskite ink, highlighting ink engineering as a foundation for scalable high-efficiency fabrication.¹⁶⁴ Deng et al. further improved the photostability of FAC mini-modules through defect compensation, demonstrating that large-area FA-based modules require simultaneous control over crystallization quality, defect formation, and ion migration.¹⁶⁵ More recently, an immobilization strategy in slot-die coating enabled a certified efficiency of 20.3% for 10 cm² inverted perovskite modules, indicating that wet-film immobilization and precursor coordination regulation are crucial for narrowing the cell-to-module efficiency gap.¹⁶⁶

In addition to ink formulation and wet-film immobilization, recent progress in module manufacturing has further highlighted the importance of drying-field and crystallization-field engineering. A representative example is the recent work by Microquanta and collaborators, who developed a three-dimensional laminar-flow-assisted crystallization strategy for square-meter-sized perovskite solar modules. By generating a well-defined laminar airflow over large-area wet films, this method enables rapid and spatially uniform solvent removal, synchronized crystallization, and improved film uniformity over module-scale substrates. The resulting perovskite modules with a small area achieved a certified power conversion efficiency of 15.0% and passed multiple photovoltaic module standard tests, demonstrating that large-area perovskite manufacturing requires not only absorber-composition optimization, but also precise control over coating, drying, crystallization, interconnection, and encapsulation processes.¹⁶⁷

Module fabrication should be regarded as a scaling test of the microscopic structure–property relationships established for FA-based perovskites. In small-area devices, local variations in intermediate-phase conversion, defect density, phase purity, or ion accumulation may be partially masked by statistical averaging. Across module-scale areas, however, these heterogeneities directly translate into spatially distributed voltage loss, non-uniform degradation, and reduced manufacturing yield.¹⁶⁸ Therefore, scalable FA-based processing must provide spatially homogeneous control of both α-phase formation and the subsequent ionic response of the absorber. Future module studies should combine area-resolved mapping of phase purity, photoluminescence or QFLS, ion redistribution, and accelerated operational stability, thereby connecting manufacturing reproducibility with the fundamental FA-regulated degradation mechanisms discussed above.¹¹

5. Conclusions and outlook

5.1. Summary and critical analysis

The shift from MA to FA-based compositions marks a key transition in halide perovskite photovoltaics. As summarized in this review, FAPbI₃ has emerged as a leading absorber for both single-junction and tandem devices, benefiting from enhanced thermal robustness and a near-optimal bandgap approaching the Shockley–Queisser optimum. Yet this optoelectronic advantage is intrinsically coupled to crystallographic metastability: the photoactive α-phase is thermodynamically predisposed to transform into the non-perovskite δ-phase, a propensity commonly associated with a non-ideal Goldschmidt tolerance factor and lattice strain introduced by the relatively large FA cation.

Recent progress can be organized into three complementary stabilization routes:

(1) Kinetic regulation through additives. Volatile additives (like MACl) and pseudohalides (like SCN⁻ and HCOO⁻) can act as crystallization regulators rather than simple compositional modifiers. By reshaping the nucleation-and-growth landscape, either lowering the kinetic barrier to α-phase formation or stabilizing intermediate adducts, these additives can redirect the solidification pathway and suppress competitive δ-phase formation.

(2) Thermodynamic stabilization via alloying. Compositional engineering, most prominently through partial substitution with Cs⁺ or Rb⁺, remains central for improving phase stability. Such alloying tunes the effective tolerance factor and can increase configurational entropy, thereby broadening the stability window of the black perovskite lattice. This approach is especially consequential for wide-bandgap mixed-halide perovskites, where appropriate A-site engineering can also help suppress photoinduced halide segregation.

(3) Surface and interface energy modulation. Stabilization can also be achieved through thermodynamic control at surfaces and interfaces. Bulky organic cations (like TCMAI) and engineered 2D/3D heterostructures illustrate that reducing the surface energy of the α-phase and passivating reactive facets can promote phase-pure film formation, including under less stringent processing environments. These interfacial strategies complement bulk approaches by targeting phase-conversion pathways that initiate at surfaces, grain boundaries, and buried interfaces.

5.2. Future outlook

The structural complexity of FA-based perovskites, long viewed primarily as an obstacle to device stability, is also recognized as a rich design space. Several emerging directions promise to transform our ability to understand, control, and ultimately exploit this complexity, from empirical optimization to mechanistic control. The phase behavior of FA perovskites, whether the α/δ competition in lead systems or the polymorphic landscape in tin analogues, remains governed by a delicate interplay of thermodynamics, kinetics, and defect chemistry. While empirical compositional engineering has achieved remarkable success, truly predictive control will require a deeper mechanistic understanding. The specific questions are increasingly well-defined: How do solvent molecules coordinate with PbI₂ while simultaneously hydrogen-bonding with FA⁺ to template the formation of intermediate adducts that direct nucleation toward the α-phase? How do iodine vacancies and interstitials weaken the local Pb–I framework, increase octahedral vibrational amplitudes, and lower the barrier for the α to δ transition? Answering these questions will shift the field from “obtaining the α-phase” to “dialing in a specific structural state”—with controlled octahedral tilting, local disorder, and defect distributions—that optimizes both performance and stability. This mechanistic understanding is also the prerequisite for moving beyond empirical accelerated aging tests toward mechanism-driven lifetime prediction, a critical step for commercial deployment.

5.2.1. Managing the dynamic duality of the FA cation

Throughout this Perspective, we have emphasized that the dynamic disorder of FA⁺ is simultaneously a source of outstanding carrier protection through large-polaron formation and detrimental ion migration through hydrogen-bond-assisted vacancy hopping. The most profound challenge for the field is to decouple these two consequences. This is fundamentally a lattice-dynamics engineering problem. Can the phonon spectrum be tuned, through targeted local strain, A-site alloying with dynamically active but electronically inert species, or dimensional confinement to strengthen polaron coupling while shifting octahedral breathing modes away from resonance with ion migration pathways? Can the orientational correlation of FA⁺ dipoles be controlled to maintain beneficial electrostatic fluctuations while breaking the collective dipole–ion coupling that drives halide segregation? At the materials-design level, interrupting the cascade from initial defect generation to full device degradation will require introducing targeted additives, redox shuttles, and antioxidant strategies that dynamically repair damage induced by migrating ions, while compositional and strain engineering remain effective approaches for increasing the activation energy for ion migration. These questions point toward a paradigm that moves beyond static structural optimization to the active design of dynamic behavior—a frontier that remains almost entirely unexplored.

5.2.2. Chemical stability of the FA cation itself

Beyond the physical instabilities discussed above lies a deeper chemical vulnerability: the FA cation can undergo irreversible degradation. Under illumination or electrical bias, particularly in the presence of oxidized iodine species generated by halide redox reactions, FA⁺ is susceptible to deprotonation, nucleophilic substitution, and other irreversible transformations. These reactions, often overlooked in favor of the more commonly studied halide processes, are key contributors to the complete collapse of the perovskite lattice. The degradation likely follows a cascade in which ion migration and phase segregation create local environments rich in reactive iodine species, which then chemically attack the FA cation, destroying the very organic component that stabilizes the three-dimensional framework. To capture these transient species and complex interfacial dynamics, future research must develop operando spectroscopic and microscopic techniques capable of monitoring charge distribution and ion accumulation in real time under multiple operational stresses. Understanding the precise reaction pathways—and designing strategies to intercept them, whether through antioxidant additives that scavenge reactive iodine before it attacks FA⁺, or through A-site alloying with degradation-resistant cations—represents a critical frontier for extending operational lifetime.

5.2.3. Computational and experimental tools for mechanistic understanding

Addressing these challenges is becoming increasingly feasible thanks to rapid advances on both computational and experimental fronts. On the computational side, deep-learning-driven methods have achieved transformative breakthroughs: neural-network interatomic potentials trained on DFT data now enable molecular dynamics simulations with near-DFT accuracy at scales of tens of billions of atoms, opening a window into crystallization kinetics from precursor–solvent interactions to nucleation pathways. Deep learning approaches are also being extended to excited-state and transport calculations previously prohibitive for large systems, including GW-level band structures and carrier mobilities, capabilities directly relevant to the dynamic disorder and polaron physics central to FA-based perovskites. On the experimental side, in situ scanning transmission electron microscopy now enables atomic-resolution imaging of point defects at surfaces and grain boundaries, while operando X-ray diffraction and spectroscopy track structural evolution under operating conditions. Coupling these tools with machine-learning-driven data analysis creates a closed loop: simulations predict, experiments validate, and the resulting data refine the models. This predict–validate–iterate cycle will transform the field from empirical optimization toward mechanism-guided, prediction-enabled design.

Phase transitions may serve as a functional asset beyond photovoltaics. The very phase instability that challenges photovoltaic applications may become an asset in other domains. The controllable and reversible α–δ phase transition in FA-based perovskites, with its large contrast in optical bandgap, electrical conductivity, and dielectric constant, is an attractive basis for phase-change memory, memristive devices, and optical switches. Similarly, the mixed ionic–electronic conductivity arising from FA-assisted ion migration, problematic for solar cell stability, could be harnessed for neuromorphic computing elements that rely on voltage-driven ion redistribution to emulate synaptic plasticity. In these contexts, the FA cation's dynamic disorder, structural responsiveness, and facile phase switching are not liabilities but functional features. The key to unlocking such applications lies in the same mechanistic understanding and controlled manipulation of phase behavior that we have advocated for photovoltaic purposes, specifically, the ability to trigger and reverse the phase transition reliably over many cycles without chemical degradation. Viewed from this broader perspective, the inherent duality of the FA cation becomes not a limitation to be overcome, but a versatile design space whose exploration may yield technologies extending far beyond solar energy conversion.

Author contributions

W. L., Z. Y., and G. L. conducted the literature research for the review and contributed to the discussion of the content and writing. D. F. and N. W. provided comments and suggestions for the revision of the manuscript. Q. W. and Z. N. contributed to the discussion, writing, and editing of the manuscript.

Conflicts of interest

The authors declare no competing interests.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.