Design principle for mitigating moisture induced degradation in 2D halide perovskites

Design principle for mitigating moisture induced degradation in 2D halide perovskites | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Design principle for mitigating moisture induced degradation in 2D halide perovskites Aditya Mohite, Isaac Metcalf, Shanize Forte, Jianlin Zhou, Claudine Katan, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8941587/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract We establish a novel conceptual framework which explains and mitigates moisture-induced degradation in 2D perovskites. Degradation proceeds along defined pathways across the spacer cation iodide (A′I)-PbI2-H2O ternary phase diagram as moisture intercalates into the perovskite. The structure and chemistry of the spacer cations control the availability and formation energies of non-perovskite phases, which in turn determines the degradation mechanism and rate in humid air. Guided by this framework, we establish a spacer cation design principle focused on eliminating low-energy hydrate phases to suppress moisture driven degradation, thereby enhancing 2D perovskite stability. Leveraging this mechanism, we show that replacing linear butylammonium with branched isobutylammonium suppresses moisture-induced degradation of the associated 2D powders by a factor of six. The existence of mixed-organic cation non-perovskite phases can similarly destabilize mixed 3D-2D perovskite films, and suppressing such mixed phases improves stability. Physical sciences/Materials science/Materials for energy and catalysis/Solar cells Physical sciences/Chemistry/Materials chemistry/Electronic materials Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Halide perovskites are emerging semiconductor materials with applications in photovoltaics and optoelectronics. Despite promising device performance, 3-dimensional (3D) halide perovskites suffer from poor durability upon exposure to light, heat, oxygen, and/or humidity, a key barrier to implementation ( 1 ). The incorporation of the related two-dimensional (2D) halide perovskites can markedly improve 3D perovskite durability ( 2 ). However, the origin of the 2D perovskites’ superior stability and its limitations are still not entirely understood, in particular under humidity. The 3D perovskite MAPbI 3 (MA=methylammonium) degrades in humid conditions via the formation of a “monohydrate” phase, MAPbI 3 ⋅H 2 O ( 3 ), and eventually a “dihydrate” phase, MA 4 PbI 6 ⋅2H 2 O ( 4 ). The monohydrate phase consists of “ribbons” of edge-sharing PbI 6 octahedra, whereas the dihydrate phase shows disconnected zero-dimensional (0D) PbI 6 . Both hydrate phases destroy the perovskite sublattice of corner-sharing octahedra, removing the useful optoelectronic properties of MAPbI 3 . At higher humidity levels, the MA cation will exit the lattice to leave the PbI 2 lattice. In contrast, 2D perovskite degradation in humid air is not well-understood. A few multi-layered 2D perovskites have been reported to degrade into lower-n-value (inorganic layer thickness) 2D perovskites and MA hydrates ( 5 – 8 ). Humidity-induced degradation phases have been reported for certain spacer cations ( 9 – 21 ), most famously for phenyldimethylammonium ( 9 ). Other reports have attributed unassigned structural signatures during degradation to potential hydrates ( 22 – 24 ). However, the prevalence and influence of 2D perovskite hydrates are not well understood and as a result, a general framework to describe 2D perovskite degradation mechanisms in humid air remain elusive. Moreover, the relative stability of Ruddlesden-Popper (RP) and Dion-Jacobson (DJ) phase 2D perovskites has been debated for several years. In most contexts, particularly in devices, DJ perovskites appear more stable ( 25 , 26 ); at other times, particularly in unencapsulated films, RP perovskites can show better stability ( 27 ). As state-of-the-art 3D perovskite solar cells continue to incorporate 2D perovskites and 2D forming cations as passivation layers for the synergistic improvement of stability and efficiency, understanding 2D humidity-induced degradation is crucial to predict and enhance the stability of perovskite devices. Here, using linear alkylammonium spacer cations as a model system, we establish a new framework for understanding moisture-induced degradation in 2D perovskites based on the trajectory of the material across the spacer-cation iodide (A′I or A′I₂)–PbI₂–H₂O ternary phase diagram. We show that 2D perovskites incorporating linear alkylammonium spacer cations, NH₃(CH₂)ₙ₋₁CH₃ (n = 3–8), form distinct hydrate phases that are structurally analogous to the monohydrate phase of MAPbI₃; referred to as the “principal hydrates” of 2D perovskites. Notably, signatures of these hydrates have been reported previously, but were misassigned to lower-n 2D perovskite phases ( 5 ). In addition to these principal hydrates, 2D perovskites also form spacer-cation-rich 0D phases that resemble the dihydrate phase of MAPbI₃. Extending beyond this series of linear spacer cations, we observe hydrate and other non-perovskite crystalline phases for nearly all spacer cations examined, across both RP and DJ perovskite families. The emergence of these non-perovskite phases introduces additional degradation pathways under humid conditions, thereby reducing perovskite phase stability. Because the perovskite, hydrate, non-perovskite crystalline, and solvated phases are all composed of A′I (or A′I₂), PbI₂, and H₂O, they can be represented within a single A′I–PbI₂–H₂O ternary phase diagram. Crucially, the molecular structure of the spacer cation dictates which non-perovskite phases populate this diagram and, in turn, governs the stability of the perovskite phase (Figure 1a). These observations point to a new materials-engineering principle: suppressing moisture-induced degradation by simplifying the accessible phase space of 2D perovskites. Throughout this work, two-dimensional (2D) perovskite spacer cations are denoted by the number of carbon atoms in the organic chain. For example, A′ = NH₃(CH₂)₃CH₃ (butylammonium) is labeled “4A,” and the corresponding RP perovskite (4A)₂PbI₄ is denoted as 4A n = 1. An analogous notation is adopted for DJ perovskites: NH₃(CH₂)₄NH₃ (butyldiammonium) is labeled “4DA,” and (4DA)PbI₄ is denoted as 4DA n = 1. Higher-order 2D perovskites (n ≥ 2) are labeled as A–A′ n ≥ 2, where A represents either MA or formamidinium (FA) as the small A-site cation and A′ denotes the spacer cation, with the same convention extended to larger n values. For clarity in selected figures, MA–4A n = 2 and FA–4A n = 2 are abbreviated as M4 and F4, respectively, with analogous abbreviations used for other spacer cations. Results and Discussion 1. Growing and characterizing 2D perovskite hydrates In order to understand the ternary phase diagrams associated with the linear alkylammonium spacer cations, we grew all nonperovskite phases compatible which each cation. Single crystals of principal 2D perovskite hydrates were grown from mixed HI-H 2 O solution. The structures of these hydrate phases are described in Figure 1b-c and Table S1. Ribbons of edge-sharing octahedra having stoichiometry [Pb m I 2m+2 ] 2- (with m being the width of each ribbon in number of octahedra) coordinate with an H 2 O molecule at each edge. The charge of these ribbons is compensated by two alkylammonium spacer cations which occupy the structure’s “organic site”. The crystals share a distinctive needle-like morphology (Figure S1). Their powder X-ray diffraction (XRD) patterns show prominent peaks corresponding to the (100) and (10-1) planes (Figure S2). The distance between spacer cations showed a “zig-zag” behavior with increasing cation length (Figure S3). When incrementing the chain length causes the m-value to increase (e.g. from 3A to 4A), the organic site volume and cation spacing increases. Conversely, when incrementing the chain length does not increase m-value (e.g. from 4A to 5A), the spacing decreases. Since both cation length and inorganic site width are discretized, a few specific cations (e.g. 2A and 7A) cannot be accommodated. Rather than a hydrate, 2A forms face-sharing (2A)PbI 3 (Figure S4 and Table S2). The 7A hydrate showed hair-like crystals which were too thin to characterize via single-crystal diffraction; as such its structure is still unidentified, but 1D XRD pattern confirms it as distinct from the principal hydrates (Figure S5). 9A and 10A were not found to form hydrates, suggesting m=6 may be the upper limit of structural stability. The same synthesis procedure was employed to grow hydrates of the alkyldiammonium DJ perovskites (NH 3 (CH 2 ) N NH 3 )PbI 4 with N = 2 - 10 as shown in Figures 1d and S6 and described in Table S1. These hydrates likewise form ribbons of [Pb m I 2m+2 ] 2- ,but alternate between two H 2 O (e.g. 4DA) and one H 2 O (e.g. 6DA) per organic site. 3DA and 5DA did not conform to the series (Figure S7). Structural distortion is much higher in DJ hydrates than in RP hydrates, as a result of the more rigid organic site bridged by a single spacer (Figure S8). The option of removing an H 2 O allows more spacers to be accommodated despite this rigidity. Interestingly, in addition to the 7DA principal hydrate, the 7DA cation by itself is the perfect length to fill the organic site of (7DA)Pb 2 I 6 , which adopts m=2 ribbons tilted in alternating directions without H 2 O (Figure S9). Compared to RP, DJ perovskites form hydrate phases readily (rather than PbI 2 or perovskite phases) when introduced directly to water, suggesting the DJ hydrates are lower-energy structures than RP hydrates (Figure S10). In order to grow phases with high A′I (RP) or A′I 2 (DJ) concentration, we added PbI 2 to extremely A′I-rich aqueous solutions (Figure S11-S12 and Table S3). A′-rich phases were found for RP cations up to 6A and DJ cations up to 9DA. Many share the 0D structure of MA 4 PbI 6 ·2H 2 O. Hydrates of other common perovskite cations were also grown to explore the space of possible structures. For example, potassium, ammonium, and FA hydrates form tilted m=2 ribbons similar to (7DA)Pb 2 I 6 but with a drastically different organic site (Figure S13). Here H 2 O forms honeycomb-like ribbons reminiscent of the structure of ice, with A′ cations substituting for H 2 O to compensate charge. The ammonium hydrate has been reported previously ( 28 ), as well as a hydronium acid hydrate which adopts a similar structure ( 29 ), but to our knowledge the FA hydrate has not been reported and may be important in the degradation of FAPbI 3 devices. The benzylammonium (BZA) hydrate, which has also been reported ( 15 ), resembles the principal hydrates but with tilted m=3 ribbons (Figure S14). Phenethylammonium and cyclohexylmethylammonium showed no nonperovskite phases. Finally, secondary hydrate phases were observed for some linear spacer cations (Figure S15), termed “transient hydrates”. The 5A transient hydrate (5A) 6 Pb 5 I 16 ⋅4H 2 O shows a mixture of m=1 and m=2 ribbons. The 4A transient hydrate (4A) 2 Pb 3 I 8 ·2H 2 O was only isolated as an oriented film during in-situ degradation and resembles the BZA hydrate. The 10DA transient hydrate (10DA)Pb 3 I 8 ·2H 2 O was inferred from its in-situ Bragg peaks resembling those of (4A) 2 Pb 3 I 8 ·2H 2 O, but this phase was not investigated in detail. With a list of the structures that form from A′I, PbI 2 , and H 2 O, we can hypothesize the PbI 2 - A′I - H 2 O ternary phase diagram for each spacer cation. In the absence of complete knowledge of each phase’s Gibbs free energy, we inferred the phase field connectivity of each phase diagram based on the degradation rate and products observed. Rather than conventional triangle plots, we represent these ternary diagrams as pseudo-binary A′I-PbI 2 phase diagrams, with the x-axis corresponding to relative A′I vs PbI 2 concentration and the y-axis corresponding to absolute H 2 O concentration (Figure S16). In this representation, moving upwards on the diagram corresponds to introducing H 2 O without changing the A′I: PbI 2 ratio. While there have been reports of the 3D perovskite lattice accommodating a small amount of interstitial H 2 O ( 30 , 31 ), for the purposes of this study the position of the solid phase nodes on the diagram was assumed to be fixed at the stoichiometry of the ideal crystal. However, the solvated phase boundary will vary depending on the water solubility of A′I. We measured A′I solubility by adding H 2 O incrementally to A′I powder (Table S9). As shown in Figure S17, MA was found to be less soluble than longer alkylammonium cations, with 4A being the most soluble. Compared to RP, DJ cations showed significantly lower solubility. On all phase diagrams presented, the saturation A′I concentration is given by the intersection of the A′I – H 2 O axis (the left vertical axis) with the first phase boundary from the bottom, which is the upper boundary of the cyan-colored three-phase field (Figure S17). 2. Understanding structure-property relationships for 2D perovskite hydrates i. Exploring the degradation of RP n=1 alkylammonium hydrates We hypothesized a relationship between a 2D perovskite’s phase diagram and its moisture stability, and we tested this hypothesis using in situ XRD during degradation of 2D powders and films in a home-built humidity chamber (Figure 2a and S18). The n=1 2D perovskites with spacer cations 4A - 8A were compared, with degradation rate quantified by the decrease in the out-of-plane (200) perovskite peak intensity. We controlled the humidity in the chamber through the addition of varying amounts of liquid water. Powders with 8mL H 2 O and films with 10mL H 2 O were chosen as representative “less aggressive” and “more aggressive” degradation conditions, respectively. Three regimes could be distinguished in the changing of Bragg peak intensities over time (Figure 2b). At first, the intercalation of H 2 O at grain boundaries appears to facilitate the reorganization of crystallites, increasing (or sometimes sharply decreasing) diffraction intensity. This behavior is consistent with the . Afterwards, intensity decreases linearly with time as the perovskite degrades. We normalize this regime and use it to quantify stability. Finally, intensity tails off approaching complete degradation. Conventional wisdom suggests that the hydrophobicity of spacer cations imparts 2D perovskites with improved moisture stability, so stability should increase monotonically with cation length. Crucially, this relationship was not observed here. Instead, 2D perovskites demonstrated moisture stability in relation to the formation energy of their hydrate phase and the region of 2D phase stability on the ternary A′I-PbI 2 -H 2 O phase diagram. 4A and 5A n=1 were significantly less stable in humid air than 6A, 7A, and 8A n=1, both as powders (Figure 2c) and as films (Figure 2d). Moreover, 4A n=1 powder was slightly more stable than 5A n=1. In-situ XRD could also determine the degradation pathway for each sample (Figures S19 and S20). We find that moisture-induced degradation can be envisioned as a quasi-equilibrium process, with samples charting a path across the ternary phase diagram. The effective H 2 O concentration within the sample increases, while simultaneously A′I escapes the perovskite lattice (particularly pronounced in films). This shifts the composition diagonally towards H 2 O and PbI 2 , such that the 2D perovskite begins converting into a mixture of the phases connected by a phase field to the n=1 phase. As shown in the pseudo-binary phase diagrams of Figure 2e, the 4A and 5A systems appear to have a three-phase field (green) connecting their 0D phase to their transient hydrate. This constrains the H 2 O concentration for which n=1 remains a stable phase. As shown in Figure S20, the transient 4A and 5A hydrates emerge first during degradation as expected. As the composition shifts further towards PbI 2 and H 2 O, the principal hydrate phase emerges as well. In contrast, the lack of a transient hydrate, the higher formation energy of the principal 6A hydrate, and the shifted stoichiometry of the 6A 0D phase appear to remove the connection between the hydrate and 0D phases, such that there is a large region (dark blue) of n=1 phase stability above 6A n=1. While A′-rich 0D phases were not observed in-situ during degradation, we hypothesize that the existence of these phases nonetheless influences stability by changing phase diagram connectivity above the n=1 phase. A further corroboration of the quasi-equilibrium picture of hydration can be seen in Figure S21a-b, which shows the sudden emergence of the principal 4A hydrate phase in a degraded 4A n=1 film when the lid is removed from the humidity chamber. The composition moves diagonally upwards across the phase diagram during degradation, and then vertically downward when the lid is removed, ending in a different three-phase field (yellow or red) where the principal hydrate is stable. The phase diagram picture can explain the quicker degradation of 5A n=1 powders compared to 4A n=1. The higher A′I concentration in the transient 5A hydrate compared to the transient 4A hydrate moves the 5A hydrate phase closer to 5A n=1 on the phase diagram. As a result, for the same H 2 O concentration a greater fraction of 5A n=1 will convert to hydrate phase compared to 4A n=1, accelerating degradation despite the increased bulkiness of the 5A cation. Given the 6A phase diagram, it appears contradictory that 6A n=1 films degrade into PbI 2 rather than 6A hydrate phase (Figure S21c-d). In fact, films of more stable 2D perovskite were often found to degrade directly into PbI 2 . One explanation for this behavior could be the existence of a barrier to moisture intercalation (Figure S22). At the outset of degradation, the horizontally-oriented n=1 films show a hydrophobic surface that should oppose moisture ingress. It seems reasonable that the conversion from the n=1 phase to the hydrate phase at the surface of 4A and 5A n=1 films damages the lattice and accelerates the intercalation of H 2 O, which in turn accelerates the phase conversion process in a self-reinforcing manner. Because 6A n=1 would not be expected to form hydrate phase at the earliest stage of moisture intercalation this self-reinforcing degradation might be replaced by a slow solvation of surface 6AI without moisture ingress. As a result, the composition of the system appears to move not diagonally but horizontally towards PbI 2 over time. ii. Rationalizing the poor moisture stability of Dion-Jacobson perovskites The hydrate-phase-mediated moisture degradation picture introduced above helps explain the discrepancy between RP and DJ stability as devices vs as films. DJ hydrates are less structurally complex (i.e. lower m-value) than RP hydrates of the same spacer cation length, perhaps explaining why DJ hydrates form much more readily than RP hydrates (Figure S10). A lower hydrate formation energy would tend to connect the hydrate phase with the left corner of the phase diagram, constraining n=1 phase stability. As a result, DJ perovskite powders degrade into hydrate phases faster than RP perovskites of the same chain length (Figure S23), despite DJ spacer cations being less water-soluble than RP (Figure S17). The increase in moisture stability with chain length is more monotonic for DJ n=1 than for RP n=1, and moreover the hydrate phase (rather than only PbI 2 ) appears in all samples, suggesting that these DJ powders all have similar ternary diagram connectivity (Figure S24). DJ films degraded quickly in humid air, even with only 2mL H 2 O loading (Figure S25). Film stability improved monotonically with cation length except for 7DA, which showed unique instability likely due to additional nonperovskite phases in close proximity to n=1 on its ternary diagram. A similar argument can be made for the instability of 5DA, which despite not having a ribbon-like hydrate phase nonetheless has a comparatively small region of perovskite stability. As a result, while DJ 2D perovskites have been found to show better stability against light and heat, they are relatively more susceptible to humidity-induced degradation compared to RP 2D perovskites. Since humidity has a more pronounced effect for bare films compared to encapsulated devices, DJ devices can tend to show enhanced stability over RP while the trend is reversed for bare films. 3. Degradation of 2D perovskites with n>1 Many applications of 2D perovskites require n-values greater than n=1, in particular for optimal energy alignment to 3D perovskites, and so we also degraded phase-pure n=2 powders of alkylammonium 2D perovskites, (NH 3 (CH 2 ) N-1 CH 3 ) 2 (A)Pb 2 I 7 , using both MA and FA as the A-site cation. MA n=2 perovskites were universally less stable than their n=1 counterparts, while FA n=2 was more stable (Figures 3a and S26). We attribute the greater instability of MA-based 2D perovskites to two sources. First, despite MAI having similar solubility to FAI and lower solubility than alkylammonium spacer cation salts, the low mass of MA + cation appears to allow MAI to readily escape the perovskite lattice and complex with water at the film surface. This promotes the formation of n=1 during degradation. Second, “mixed” MA-A′ hydrate phases introduce new degradation pathways (Figure 3b and Table S4). For the n=2 perovskites (4A) 2 MAPb 2 I 7 (MA-4A n=2), (5A) 2 MAPb 2 I 7 (MA-5A n=2), and (6A) 2 MAPb 2 I 7 (MA-6A n=2) a mixed-cation (MA)(A′)Pb 3 I 8 ·2H 2 O hydrate grows from dilute hydriodic acid solution with a similar structure to (3A) 2 Pb 3 I 8 ·2H 2 O, having an m-value of 3 and a mixed occupation of the organic site by MA and A′. The Pb-Pb interatomic spacing increases with spacer length, from 4.73 Å (4A) to 4.77 Å (5A) to 4.80 Å (6A), compared to 4.77 Å for (3A) 2 Pb 3 I 8 ·2H 2 O. Similarly, mixed MA-7A, MA-8A and MA-9A hydrates were found with composition (MA)(A′)Pb 4 I 10 ·2H 2 O resembling (4A) 2 Pb 4 I 10 ·2H 2 O, and MA-2A hydrates were found with composition (MA)(2A)Pb 2 I 6 resembling (MA) 2 Pb 2 I 6 . Additionally, a mixed MA-4A 0D phase was identified with composition (MA) 2 (4A) 4 PbI 8 (Figure S27a). As seen in Figures 3c, S28, and S29, MA-based n=2 perovskite powders (Figure S28) and films (Figure S29) degrade into a combination of phases including n=1, PbI 2 , and various hydrates. The degradation products are corners of the four-phase field above n=2 on the quaternary MAI-A′I-PbI 2 -H 2 O phase diagram (Figure 3d). For 4A n=2 the degradation products observed are n=1 and the MA-4A hydrate, suggesting the four-phase field shown in Figure 3d. This region is quite narrow in the direction of increasing H 2 O, due to a connection between the mixed MA-4A hydrate and 4A n=1 and as a result MA-4A n=2 has poor phase stability. For MA-5A n=2, the stoichiometry of the transient 5A hydrate blocks a connection between the mixed hydrate and 5A n=1. As a result, the four-phase field appears to be comprised of MA-5A n=2, 5AI, the mixed MA-5A hydrate, and the transient hydrate, with the n=1 phase no longer forming. This widens the four-phase field above 5A n=2, which may improve stability compared to 4A n=2. The increased mass of 5AI may also improve stability by suppressing solvation. Films of MA-5A n=2 show the formation of n=1 and the principal hydrate instead of the transient hydrate, indicating the movement to a different four-phase field away from MAI and towards PbI 2 as the organic cations escape. MA-6A n=2 degrades first into n=1 and subsequently into the principal hydrate, which corresponds to a notably different four-phase field than for 4A or 5A. Crucially, this region connects near pure H 2 O instead of at the 6A 0D phase, significantly widening it in the H 2 O direction and explaining the marked increase in stability between 5A and 6A. The Pb-Pb expansion in the MA-6A mixed hydrate appears to raise its formation energy, reducing its connectivity to other phases. For 6A n=2 films, PbI 2 also emerges as a degradation product suggesting the escape of MAI and 6AI from the sample. MA-8A n=2 degrades via the same pathway as MA-6A n=2, with the formation of the n=1 and principal hydrate phases. However, the degradation occurs notably faster both for powder and film samples in 8A n=2 compared to 6A. MAI appears to escape readily from the 8A n=2 lattice, indicated by the predominance of n=1 and the lack of PbI 2 in degrading 8A n=2 films. This behavior was unique to 8A and is still not fully understood; MA-7A, 9A, and 10A n=2 show exceptional stability (possibly due to their lack of principal hydrates) (Figure S30). Similarly, we studied degradation in FA n=2 powders (Figure S31) and films (Figure S32). Compared to MA, the FA cation has a lower acidity, a greater mass, and a smaller dipole moment ( 32 , 33 ). As a result, perovskites incorporating FA at the A-site have been found to be more chemically stable than those with MA under a variety of conditions, including in the presence of water ( 30 , 34 , 35 ). Unsurpisingly, FA-based 2D perovskites were also universally more stable in this present study. No mixed FA-A′ hydrate phases were observed during degradation, suggesting a less complex phase diagram with fewer degradation pathways. Moreover, suppressed n=1 formation suggests that FAI escapes less readily than MAI. These factors combined gave FA n=2 much higher moisture stability than MA n=2 for all 2D perovskites studied. FA-4A n=2 degraded first into 4A transient hydrate, and subsequently into 4A principal hydrate. In contrast, FA-5A n=2 degraded into the principal hydrate directly. The lack of the transient phase as a degradation pathway gives FA-5A n=2 a markedly higher stability than FA-4A n=2 and suggests the FA-4A and FA-5A phase diagrams have different connectivity. This may be related to the existence of a mixed FA-4A perovskitoid, (FA)(4A)PbI 4 (Figure S27b), although the phase is not observed during degradation. In both FA-4A and FA-5A, the FA hydrate FA 2 Pb 2 I 6 ·4H 2 O appears during early stages of degradation and subsequently fades. For spacer cations longer than 5A, a marked increase in stability is once again observed. FA-6A and FA-7A n=2 degraded into PbI 2 , while FA-8A degraded into n=1 along with a small amount of 8A principal hydrate. FA-9A and 10A likewise formed n=1, suggesting 2D perovskites with longer cations give up FA from the A-site more willingly (Figure S33). 4. Degradation of perovskites with nonlinear spacer cations Next we compared the degradation behavior of the linear cations 4A and 5A to their iso counterparts isobulyammonium (i-4A) and isopentylammonium (i-5A). i-4A forms a tilted ribbon-like hydrate phase with m=2 and stoichiometry (i-4A) 2 Pb 2 I 6 ·2H 2 O, but has no 0D phase. In contrast, i-5A forms the mixed corner-face-sharing perovskitoid (i-5A) 4 Pb 3 I 10 and the 0D phase (i-5A) 6 PbI 8 . The pseudo-binary phase diagrams for i-4A and i-5A are shown in Figure S34. Both i-4AI and i-5AI are less water-soluble than 4AI and 5AI. Compared to 4A and 5A, perovskites with i-4A and i-5A are significantly more stable (Figure S35). This can be attributed to their more favorable phase diagrams, both of which have large regions of phase stability above n=1. i-4A n=1 is moreover more stable than i-5A n=1 likely because the i-4A hydrate stoichiometry is further from n=1 compared to the i-5A perovskitoid. As with other 2D perovskites that showed near-vertical 2D-liquid phase boundaries, the composition of i-4A and i-5A n=1 appears to move horizontally towards PbI 2 ; for i5A this leads to the formation of the H 2 O-free perovskitoid, whereas for i4A this causes the direct formation of PbI 2 . For MA-based i-4A and i-5A n=2 samples (Figure S36), degradation proceeds through the conversion of n=2 to n=1; as these phases have the same position on both phase diagrams, i-5A n=2 shows better stability than i-4A n=2 due to its increased mass. The examples of i-4A and i-5A show how the spacer can be engineered to improve the phase diagram, increasing moisture stability. 5. Degradation of mixed 3D-2D perovskites and 3D-2D stacks Finally, we investigated the moisture stability of 3D-2D perovskite films, incorporating 2D both as an additive to form a 3D-2D composite and also as a capping layer to form a bilayer structure. i. 3D MAPbI 3 with spacer cation salt additives We measured the degradation of MAPbI 3 films doped with 25 mol% A′I salt (spacer length from MA to 10A) using 2mL H 2 O loading (Figures S37 and S38). Four distinct degradation routes were observed: (1) In undoped films and films with MAI and 2AI, the formation of PbI 2 and marginal MA or MA-2A hydrate, respectively; (2) with 4AI and 5AI, the formation of mixed MA-A′ hydrates; (3) with 6AI, the formation of 2D and PbI 2 ; (4) with 7AI and 8AI, the formation of 2D and mixed MA-A′ hydrates (Figure S39a-b). The fact that undoped MAPbI 3 degrades mainly into PbI 2 (not MA hydrate) suggests that MAI escapes rapidly, leading to a path through composition space slanted severely towards PbI 2 (Figure S40). In other words, the degradation behavior mimics that of 6A n=1 in Figure S22, though in this case not because of a barrier to moisture ingress but rather because of the greater tendency of MA to exit the film. Extra MAI shifts the composition further from PbI 2 , promoting the formation of MA hydrate. Additives such as 2AI, 6AI, 7AI, and 8AI show similar stability to undoped MAPbI 3 . Conversely, additives with low-energy hydrates such as 3AI, 4AI, and 5AI (4A and 5A forming mixed MA-A′ hydrates) in fact have significantly worse stability, unlocking faster degradation pathways for MAPbI 3 through hydrate formation (Figure 4b). This picture is corroborated by the behavior of 8AI, 9AI, and 10AI dopants (Figure S38); 9AI and 10AI confer improved stability over 8AI due to the suppression of mixed hydrate phases (9A shows marginal m=4 mixed hydrate, while 10A shows none). Additive molecules longer than 5A formed grains of 2D perovskites in-situ during degradation. Maximal-n-value phases (n=4-6) formed first, with subsequent emergence of lower n-values, consistent with the expected behavior of A′I at grain boundaries intercalating into the 3D lattice. Mixed-hydrate formation also accelerated the degradation of these high-n-value grains. ii. 2D n=1 perovskite capping layers on top of 3D MAPbI 3 Next, we deposited capping layers of 4A to 8A n=1 perovskite atop MAPbI 3 (Figure 4 and S41). As with doped films, the 2D spacer cation strongly influences MAPbI 3 films’ stability and degradation pathway. Distinct degradation routes were observed for a) uncapped MAPbI 3 and MAPbI 3 with 6A n=1, b) MAPbI 3 with 4A and 5A =1, and c) MAPbI 3 with 7A and 8A n=1, as can be seen by plotting the Bragg peak intensities of the hydrates and PbI 2 over time (Figure S42). Once again, mixed hydrate formation allows MAPbI 3 to degrade faster with 4A or 5A n=1 layers than with no cap (Figure 4b), despite the lack of PbI 2 suggesting MA escape is suppressed (Figure 4a). In contrast, a 6A n=1 layer suppressed MAPbI 3 degradation and formed majority PbI 2 . Moreover, intercalation across the interface formed successively higher-n-value 2D up to n=5. The fact that n=2 forms first, rather than n=4-6, suggests the MA cation migrates upwards rather than 6A downwards, consistent with a greater mobility of the MA cation predicted from the formation of PbI 2 in undoped MAPbI 3 films. 7A and 8A capping layers showed a mixture of the degradation pathways seen for 4A/5A and for 6A, with stability intermediate between them. While still showing 3D-2D layer intercalation as with 6A, the degradation into PbI 2 is suppressed by ~10 times and the formation of mixed hydrates is enhanced by ~10 times as with 4A/5A, suggesting that MA-7A and MA-8A hydrates are lower-energy phases than MA-6A hydrate (Figure S39c and S42). The prominence of the MA hydrate in MA-7A and MA-8A compared to MA-4A and MA-5A also may indicate a suppression of MA escape with longer cations. iii. 2D n=1 perovskite capping layers on top of 3D FAPbI 3 Finally, we investigated the degradation of FAPbI 3 films with n=1 capping layers using 8mL water loading (Figure S43). Capped FAPbI 3 peak intensities dropped quickly at the outset, possibly due to film damage during layer deposition. However, with no mixed FA hydrates, all capping layers improved FAPbI 3 stability over longer timeframes. Unlike other samples tested, α-FAPbI 3 is not thermodynamically stable, and its δ-phase appears on the phase diagram instead. Bare FAPbI 3 films degraded principally into PbI 2 with marginal δ-phase and no FA hydrate, suggesting FA escape dominates over moisture intercalation. The 4A and 5A n=1 top layers quickly became FA n=2 during fabrication and degraded similarly to FA n=2 samples. Both 4A and 5A slightly suppressed PbI 2 and promoted δ-phase in the degrading FAPbI 3 layer, suggesting a retention of FAI. 6A and 8A n=1 capping layers remained mostly n=1 after fabrication but converted to n=2 rapidly with humidity. These n=2 layers persisted throughout the measurement period and blocked the formation of both PbI 2 and δ-phase in the FAPbI 3 layer. The nonexistence of FA n>2 2D removes intercalation as a degradation pathway, further improving 3D-2D stability compared to MAPbI 3 -2D films. 7A uniquely remained n=1 for much longer, and its conversion to n=2 appeared to take precedence over degradation of the underlying FAPbI 3 . This may suggest a higher formation energy of FA-7A n=2, or perhaps a sparsely populated FA-7A phase diagram. 6. Conclusions This report demonstrates a new theory of 2D perovskite degradation in humid air, introducing the ternary A′I - PbI 2 - H 2 O phase diagram to determine the degradation pathway. Using this framework, new design principles can be conceptualized focusing on suppressing the formation of all but the “unavoidable” degradation phases at the corners of the phase diagram: solid A′I, a solvated liquid phase, and solid PbI 2 . Because of the tendency of organic cations to escape the system through solvation with the ambient humid air, the effective composition of the system tends towards PbI 2 and away from the A′I side over time. In four-component systems such as n≥2 perovskites, the composition will also tend away from the more mobile cation salt, usually leading to a reduction in n-value as the lighter A-site cation preferentially escapes. This theory has been explored using the linear alkylammonium spacer cations from MA to 8A, which share a class of ribbon-like hydrate phases. We can engineer spacer cations to be structurally incompatible with low-energy hydrates, reducing the number of “avoidable” phases in the phase diagram and improving moisture stability, by a factor of 6 when replacing butylammonium with isobutylammonium. We anticipate that this work will accelerate the rational engineering of spacer cations as the field pushes towards ultra-stable 2D perovskites. More broadly, a phase diagram framework may be useful in other aspects of perovskite engineering. For example, the solvent complexes on a 2D perovskite’s A′I – PbI 2 – DMF phase diagram may likewise influence solubility and film formation, as has been explored for MAPbI 3 ( 36 ). We hope this work will inspire deeper investigation into underexplored phases observed during perovskite processing and degradation. Materials and Methods i. Growth of the hydrate phases of 2D perovskites 2D perovskite hydrates were grown by dissolving PbI 2 and A′I salts in hydriodic acid diluted with deionized water. All chemicals were purchased from Sigma Aldrich. PbI 2 and the spacer cation salts MAI, FAI, 3AI, 4AI, 5AI, 6AI, and 8AI were purchased in their iodide salt form. 7AI, 4DAI 2, 5DAI 2 , 6DAI 2 , 7DAI 2 , 8DAI 2 , 9DAI 2 ,10DAI 2 , i-4AI, and i-5AI were synthesized by reacting the associated amine molecule with hydriodic acid in ethanol solvent. The iodide salts were collected as a white precipitate after evaporation of ethanol. To grow the hydrates, PbI 2 and A′I salt were combined with HI and H 2 O in the amounts noted in Tables S5-S8. The mixture was heated to 150 o C under stirring for 1-2 hours, cooled to room temperature, and allowed to sit for several days. The PbI 2 was often found to never fully dissolve in the solution. After some time white or pale yellow needle-like crystals of hydrate phase began growing above the solid PbI 2 . To grow high-A′ nonperovskite crystals, A′I was added to H 2 O near its concentration limit along with a small amount of PbI 2 . Amounts of each precursor are noted in Table S7. The mixture was dissolved at 165 o C over 1-2 hours and allowed to cool slowly to room temperature. In some cases, ~2mL of mineral oil was added to form a buffer layer to prevent water from condensing on the walls of the vial, which was found to improve crystal quality. ii. Synthesis of phase-pure 2D perovskite powders 2D perovskite powders were grown following the established synthesis procedure of the Mohite group. PbO and MACl or FACl powders were dissolved in hydriodic acid stabilized with ~10% hypophosphorus acid (H 3 PO 2 ) under stirring at 100 o C. In a separate vial, the associated amine molecule of the spacer cation was dissolved in HI and reacted at 0 o C to form dissolved A′I. The two vials were combined and the solution was heated to 200 o C under stirring until all components dissolved. The solution was allowed to cool slowly to room temperature (n=1 and MA n=2 2D) or 80 o C (FA n=2 2D). The crystals were collected via vacuum filtration and washed with heptane before being left in a vacuum oven at 70 o C overnight to remove excess HI. iii. Fabrication of phase-pure 2D perovskite films 2D perovskite films were fabricated using the memory-seeds method presented by Sidhik et al. from the Mohite group ( 1 ). Phase-pure 2D perovskite powders synthesized as above were dissolved in DMF with a concentration of 400mg/mL (RP) or dissolved in 4:1 DMF:DMSO with a concentration of 350mg/mL (DJ) under Ar atmosphere. Solutions were stirred at 70 o C for 3 hours before spin coating. 1”x1” glass substrates were cleaned via sonication sequentially in DI water with liquinox detergent, pure DI water, acetone, and 1:1 acetone:ethanol before being dried with an air gun and cleaned with UV-ozone treatment for 30 minutes. To spin-coat films, 70µL of precursor solution was pipetted onto the surface of the substrate and spun at 5000RPM/3500RPMS for 30 seconds without antisolvent (RP) or with 120µL chlorobenzene(CBZ) antisolvent after 20 seconds (DJ) under Ar atmosphere. Films were annealed at 100 o C for 10 minutes in Ar. iv. Fabrication of 3D-2D perovskite bilayer films 3D MAPbI 3 films were fabricated from 1M precursor solutions of 1:1 PbI 2 :MAI dissolved in DMF. 5 mol% MACl was added to improve film quality. Solutions were stirred at 70 o C for 4 hours. Substrates were prepared as described above. MAPbI 3 films were spin-coated for 30 seconds at 5000RPM/3500RPMS with 120µL CBZ antisolvent added after 20 seconds. Films were annealed at 100 o C for 10 minutes. 3D FAPbI 3 films were fabricated from 1M precursor solutions of 1:1 PbI 2 :FAI dissolved in 5:1 DMF:DMSO. 35mol% MACl was added to promote perovskite phase formation and improve film quality as shown by Kim et al. ( 2 ). Solutions were stirred at 70 o C for 4 hours. Substrates were prepared as described above. FAPbI 3 films were spin-coated for 30 seconds at 5000RPM/3500RPMS with 1000 µL diethyl ether antisolvent added after 10 seconds. Films were annealed at 150 o C for 30 minutes. 2D top layers were added following the procedure of Sidhik et al. ( 3 ) 50mg of 2D perovskite powder was dissolved in 1000µL of acetonitrile (ACN) and stirred at room temperature for 1-2 hours. MAPbI 3 films were placed once again on the spin coater chuck and spun at 5000RPM for 30 seconds. After reaching max speed, 50µL of ACN solution was dynamically spin-coated onto the surface of the film. The resulting bilayer films were annealed at 100 o C for 5 minutes. v. Fabrication of 3D perovskite films with 2D additive 3D MAPbI 3 precursor solutions were fabricated as above, with an additional 25 mol% of A′I additive introduced to the solution. The spin-coating and annealing procedure was the same as for the 3D MAPbI 3 layer of 3D-2D bilayer films. vi. Determination of saturation concentration of A′I salts in water A pre-weighed amount of A′I salt was added to a vial containing 5mL of hexane as a noninteracting barrier liquid to prevent water ingress or evaporation. DI water was added to the solution 20µL at a time. After each addition vials were vigorously stirred and sonicated until the partial solvation of A′I salt was complete. This was repeated until no solid phase could be visibly observed in the vial, at which point the total volume of water added was recorded (Table S9). vii. Construction of the phase diagram Each phase diagram was populated with the phases identified in our screening of the ternary composition space, including the pure components PbI 2 , A′I or A′I 2 (A′ = 4A, 5A, 4DA, etc.), and H 2 O; the perovskite phase A′ 2 PbI 4 (RP), A′PbI 4 (DJ), or APbI 3 (3D); the principal and transient hydrate phases; and any other nonperovskite phases observed. The connectivity of the phase diagram was taken from the convex hull of the Gibbs free energy surface across the ternary composition space. To construct the phase diagram, the assumption was made that each observed phase intersects with the convex hull, that the A′I and H 2 O form an ideal solution, and that the energy surface of all solid phases is a Kronniker delta at each phase’s stoichiometry. Measuring the true free energy of the phases was beyond the scope of this project; instead we inferred the connectivity of nodes by the order in which phases were observed during degradation, and with the assumption that phases of similar structure likely had similar free energy. Quaternary phase diagrams were constructed identically, taking the convex hull of the Gibbs Free energy surface across the quaternary composition space. The color-coding of phase fields was kept consistent between phase diagrams. Two-phase fields are colored gray; three-phase fields which include one liquid phase are colored blue or purple; three-phase which include PbI 2 are colored red (if all solid phases) or purple (if mixed solid and liquid phases). Additional three-phase fields, if they exist, are colored green – yellow – brown from left to right (i.e. high-A′I to high-PbI 2 ). viii. Measuring degradation with in-situ 1D-XRD Sample degradation was measured in-situ by means of a home-built humidity chamber described in Figure 1 and Figure S17. The free volume of the humidity chamber was 17mL. The chamber consisted of a “well” with free volume 11mL, and a “lid” with free volume 6mL (accounting for 0.2mL sample volume). Kapton tape was used to form an air-tight seal around the lid which X-rays could pass through. Liquid water (either 2, 8, or 10mL) was added to the chamber and allowed to sit for >3 minutes before beginning the measurement to allow the temperature of the water to equilibrate the temperature inside the XRD. The humidity inside the chamber always reached 100%RH regardless of water volume, but excess water reduced the free volume of air in the chamber and accelerated degradation. In-situ XRD was carried out using a Rigaku Smartlab XRD with Bragg-Brentano focusing using a Cu target and a filter to suppress the Cu K-β signature. Diffraction patterns were collected across a 2θ range of 2 o to 20 o at a speed of 10 degrees per minute (resolution 0.001 o ). After each 2 o to 20 o scan completed the next scan was immediately started. Post-degradation diffraction patterns with the chamber lid removed were collected across an identical angle range and at the same speed. ix. Characterizing crystal structure with single crystal XRD Hydrate single crystals were grown as discussed above. To preserve the liquid water environment of the crystals, 1mL of mineral oil was added directly to the vial of synthesized hydrate in solution. The water was then extracted with a pipette leaving the hydrate and a small layer of water surrounding it submerged in mineral oil. Crystals were placed under a visible microscope and attached to a MiTeGen crystallography sample holder using parabar. Single crystal XRD was carried out using a Rigaku Synergy-S with an MO source and a HyPix-Arc150 detector. x. Characterizing crystal structure of 4A transient hydrate with in-situ GIWAXS The transient 4A hydrate phase was characterized in-situ via GIWAXS conducted at Brookhaven National Lab’s NSLS-II facility 11-BM (CMS). A film of 4A n=1 was loaded into the humidity chamber described above and mounted in CMS’s sample staging area. GIWAXS was conducted with 13.5 keV X-Ray energy using a Pilatus-1M detector positioned maximally close to the sample to increase the q-range of the diffraction. After ~10 minutes of degradation in humid air an oriented film of transient hydrate was observed on the surface of the sample. The measurement continued for 1 hr until the 2D perovskite signature was no longer visible. The q z -axis peak positions seen in GIWAXS were calibrated against the more accurate peak positions determined via 1D XRD. The WAXS pattern suggested an orthorhombic lattice with lattice parameters close to those of the 3A principal hydrate. With reference to other ribbon-like orthorhombic hydrates, we assumed that adjacent ribbons tilted along the b-axis to fix β=90°. The offset of adjacent ribbons along the b-axis was surmised by comparing the measured WAXS pattern with simulated WAXS patterns of space groups Pc2a (no offset), Pnmn (b/2 offset), and F2dd (b/4 offset). The Pc2a space group best aligned with the observed (hkl) Bragg peak intensities for k≠0. References N. Li, X. Niu, Q. Chen, H. Zhou, Towards commercialization: the operational stability of perovskite solar cells. Chem. Soc. Rev. 49 , 8235–8286 (2020). I. Metcalf, S. Sidhik, H. Zhang, A. Agrawal, J. Persaud, J. Hou, J. Even, A. D. Mohite, Synergy of 3D and 2D Perovskites for Durable, Efficient Solar Cells and Beyond. Chem. Rev. 123 , 9565–9652 (2023). G. H. Imler, X. Li, B. Xu, G. E. Dobereiner, H.-L. Dai, Y. Rao, B. B. Wayland, Solid state transformation of the crystalline monohydrate (CH3NH3)PbI3(H2O) to the (CH3NH3)PbI3 perovskite. Chem. Commun. 51 , 11290–11292 (2015). A. M. Leguy, Y. Hu, M. Campoy-Quiles, M. I. Alonso, O. J. Weber, P. Azarhoosh, M. Van Schilfgaarde, M. T. Weller, T. Bein, J. Nelson, Reversible hydration of CH3NH3PbI3 in films, single crystals, and solar cells. Chemistry of Materials 27 , 3397–3407 (2015). B. R. Wygant, A. Z. Ye, A. Dolocan, Q. Vu, D. M. Abbot, C. B. Mullins, Probing the Degradation Chemistry and Enhanced Stability of 2D Organolead Halide Perovskites. J. Am. Chem. Soc. 141 , 18170–18181 (2019). J. Tang, W. Tian, C. Zhao, Q. Sun, C. Zhang, H. Cheng, Y. Shi, S. Jin, Imaging the Moisture-Induced Degradation Process of 2D Organolead Halide Perovskites. ACS Omega 7 , 10365–10371 (2022). B. R. Wygant, G. T. Geberth, A. Z. Ye, A. Dolocan, D. E. Cotton, S. T. Roberts, D. A. Vanden Bout, C. B. Mullins, Moisture-driven formation and growth of quasi-2-D organolead halide perovskite crystallites. ACS Applied Energy Materials 3 , 6280–6290 (2020). J. Schlipf, Y. Hu, S. Pratap, L. Bießmann, N. Hohn, L. Porcar, T. Bein, P. Docampo, P. Müller-Buschbaum, Shedding Light on the Moisture Stability of 3D/2D Hybrid Perovskite Heterojunction Thin Films. ACS Appl. Energy Mater. 2 , 1011–1018 (2019). A. Dučinskas, G. Y. Kim, D. Moia, A. Senocrate, Y.-R. Wang, M. A. Hope, A. Mishra, D. J. Kubicki, M. Siczek, W. Bury, T. Schneeberger, L. Emsley, J. V. Milić, J. Maier, M. Grätzel, Unravelling the Behavior of Dion–Jacobson Layered Hybrid Perovskites in Humid Environments. ACS Energy Lett. 6 , 337–344 (2021). R. N. Nurdillayeva, R. F. Moral, M. Pols, D.-K. Lee, V. Altoe, C. P. Schwartz, S. Tao, C. M. Sutter-Fella, Humidity Disrupts Structural and Chiroptical Properties of Chiral 2D Perovskites. ACS Nano 19 , 11348–11357 (2025). D. G. Billing, A. Lemmerer, Synthesis and crystal structures of inorganic–organic hybrids incorporating an aromatic amine with a chiral functional group. CrystEngComm 8 , 686–695 (2006). A. Dučinskas, G. C. Fish, M. A. Hope, L. Merten, D. Moia, A. Hinderhofer, L. C. Carbone, J.-E. Moser, F. Schreiber, J. Maier, J. V. Milić, M. Grätzel, The Role of Alkyl Chain Length and Halide Counter Ion in Layered Dion−Jacobson Perovskites with Aromatic Spacers. J. Phys. Chem. Lett. 12 , 10325–10332 (2021). Y. Zhang, Z. Wang, S. Hu, P. Yan, H. Li, C. Sheng, Robust and Swiftly Reversible Thermochromic Behavior of a 2D Perovskite of (C6H4(CH2NH3)2)(CH3NH3)[Pb2I7] for Smart Window and Photovoltaic Smart Window Applications. ACS Appl. Mater. Interfaces 13 , 12042–12048 (2021). J. Lai, J. Tan, R. Zhu, W. Lv, S. Ye, Conformational Disordering of Organic Cations during Moisture-Induced Structural Collapse in Organic–Inorganic Hybrid Perovskites. ACS Appl. Energy Mater. 7 , 10521–10527 (2024). I. W. H. Oswald, A. A. Koegel, J. R. Neilson, General Synthesis Principles for Ruddlesden–Popper Hybrid Perovskite Halides from a Dynamic Equilibrium. Chem. Mater. 30 , 8606–8614 (2018). X. Jiang, J. Gu, S. Zheng, T. Zhang, X. Li, Y. Guan, C. Li, J. Wang, W. Zhang, Y. Fu, Mixed-Solvent Crystallization Expands the Chemical Space of 2D Perovskites and Enables Phonon-Limited Exciton Transport. J. Am. Chem. Soc. 147 , 32287–32298 (2025). M.-Y. Zhu, L.-X. Zhang, J. Yin, J.-J. Chen, L.-J. Bie, A fluorescence quenching sensor for Fe3+ detection using (C6H5NH3)2Pb3I8·2H2O hybrid perovskite. Inorganic Chemistry Communications 109 , 107562 (2019). R. Han, Q. Yue, Y. Cao, D. Wei, X. Liu, J. Lin, Highly Sensitive Two-Dimensional Ruddlesden-Popper Perovskites for Thermochromic Smart Windows. ChemNanoMat 10 , e202400306 (2024). W. T. Sun, Z. Xing, A. Sergeev, Y. He, A. M. C. Ng, K. S. Wong, L. Molčanov, J. Popović, A. B. Djurišić, Effects of ambient exposure on photoluminescence of Dion–Jacobson tin-based halide perovskites. J. Mater. Chem. C 11 , 2737–2747 (2023). Y. Zhang, Y. Fu, X. Song, X. Zhang, Releasable Water Charge-Trapping and Water-Resistant Photodetection using 1D Perovskitoid Hydrate Single Crystal. Advanced Materials 34 , 2204710 (2022). I. W. Gilley, H. W. Kwon, C. Liu, Y. Yang, C. Huang, H. Wan, A. S. R. Bati, E. H. Oriel, M. Kepenekian, B. Vishal, S. Zeiske, K. S. Bayikadi, T. E. Wiggins, E. S. Vasileiadou, B. Chen, R. D. Schaller, J. Even, S. De Wolf, E. H. Sargent, M. G. Kanatzidis, Combining Organic Cations of Different Sizes Grants Improved Control over Perovskitoid Dimensionality and Bandgap. J. Am. Chem. Soc. 147 , 7777–7787 (2025). K. Yang, J. Zhao, K. Zhang, B. Shao, F. Cui, X. Yang, Humidity Stability of 2D Dion–Jacobson-Phase Perovskites with Hydrophobic Groups in Diammonium Spacers. Ind. Eng. Chem. Res. 62 , 16154–16162 (2023). Y. Liu, H. Zhou, Y. Ni, J. Guo, R. Lu, C. Li, X. Guo, Revealing stability origin of Dion-Jacobson 2D perovskites with different-rigidity organic cations. Joule 7 , 1016–1032 (2023). K. Meng, X. Wang, Z. Li, Z. Liu, Z. Qiao, C. Wang, Y. Hu, S. Li, L. Cheng, Y. Zhai, G. Chen, Self-passivation of low-dimensional hybrid halide perovskites guided by structural characteristics and degradation kinetics. Energy Environ. Sci. 14 , 2357–2368 (2021). J. Gong, M. Hao, Y. Zhang, M. Liu, Y. Zhou, Layered 2D Halide Perovskites beyond the Ruddlesden–Popper Phase: Tailored Interlayer Chemistries for High-Performance Solar Cells. Angewandte Chemie International Edition 61 , e202112022 (2022). S. Ahmad, P. Fu, S. Yu, Q. Yang, X. Liu, X. Wang, X. Wang, X. Guo, C. Li, Dion-Jacobson Phase 2D Layered Perovskites for Solar Cells with Ultrahigh Stability. Joule 3 , 794–806 (2019). E. S. Vasileiadou, B. Wang, I. Spanopoulos, I. Hadar, A. Navrotsky, M. G. Kanatzidis, Insight on the Stability of Thick Layers in 2D Ruddlesden–Popper and Dion–Jacobson Lead Iodide Perovskites. J. Am. Chem. Soc. 143 , 2523–2536 (2021). L. Prisinzano, D. Delmonte, K. C. Ravaglia, V. Vit, L. Righi, Dehydration process in 1D ammonium lead halide and mixing of organic cations in hybrid perovskites through mechanosynthesis. Mater. Res. Express 7 , 115503 (2020). S. Sobczak, A. M. Fidelli, J.-L. Do, G. P. Demopoulos, A. Moores, T. Friščić, A. Katrusiak, High-pressure observation of elusive iodoplumbic acid in different hydronium-hydrate solid forms. Inorg. Chem. Front. 11 , 735–744 (2024). E. Smecca, Y. Numata, I. Deretzis, G. Pellegrino, S. Boninelli, T. Miyasaka, A. L. Magna, A. Alberti, Stability of solution-processed MAPbI 3 and FAPbI 3 layers. Physical Chemistry Chemical Physics 18 , 13413–13422 (2016). F. Cordero, F. Craciun, A. M. Paoletti, G. Zanotti, Structural Transitions and Stability of FAPbI3 and MAPbI3: The Role of Interstitial Water. Nanomaterials 11 , 1610 (2021). D. H. Fabini, C. C. Stoumpos, G. Laurita, A. Kaltzoglou, A. G. Kontos, P. Falaras, M. G. Kanatzidis, R. Seshadri, Reentrant structural and optical properties and large positive thermal expansion in perovskite formamidinium lead iodide. Angewandte Chemie 128 , 15618–15622 (2016). A. Binek, F. C. Hanusch, P. Docampo, T. Bein, Stabilization of the trigonal high-temperature phase of formamidinium lead iodide. The journal of physical chemistry letters 6 , 1249–1253 (2015). T. Zhang, X. Meng, Y. Bai, S. Xiao, C. Hu, Y. Yang, H. Chen, S. Yang, Profiling the organic cation-dependent degradation of organolead halide perovskite solar cells. J. Mater. Chem. A 5 , 1103–1111 (2017). W. Ke, L. Mao, C. C. Stoumpos, J. Hoffman, I. Spanopoulos, A. D. Mohite, M. G. Kanatzidis, Compositional and Solvent Engineering in Dion–Jacobson 2D Perovskites Boosts Solar Cell Efficiency and Stability. Advanced Energy Materials 9 , 1803384 (2019). A. A. Petrov, A. A. Ordinartsev, K. A. Lyssenko, E. A. Goodilin, A. B. Tarasov, Ternary Phase Diagrams of MAI–PbI2–DMF and MAI–PbI2–DMSO Systems. J. Phys. Chem. C 126 , 169–173 (2022). Additional Declarations There is NO Competing Interest. Supplementary Files HydratesSIsubmission.docx Supplementary Information Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8941587","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":598161970,"identity":"7876b769-1aee-470d-accb-28e2d10748d7","order_by":0,"name":"Aditya Mohite","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAqklEQVRIiWNgGAWjYFCCBAaDDwwSYKYE0VoKZ5Cs5TMPlEmcFv727MTNtnss5PkZmA/e5iGsAWjwmbebjXOeSRjObGBLtiZKC8ON3G3GOQckEgwO8JhJE6VF/kbu9t8WQC32B/i/EafF4EbuBmMGkC0MPGzEaTE883aDYc8BCcMZh9mMLecQo0XueO4Ggx8H6uT525sf3nhDjBYEYCZN+SgYBaNgFIwCfAAA+Nou4XI8y9cAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0001-8865-409X","institution":"Rice University","correspondingAuthor":true,"prefix":"","firstName":"Aditya","middleName":"","lastName":"Mohite","suffix":""},{"id":598161971,"identity":"aa7157ea-ed71-494c-82cf-591c8fade68c","order_by":1,"name":"Isaac Metcalf","email":"","orcid":"","institution":"Rice University","correspondingAuthor":false,"prefix":"","firstName":"Isaac","middleName":"","lastName":"Metcalf","suffix":""},{"id":598161972,"identity":"a3a9af39-e042-403d-94fc-7e9de4dbcf1e","order_by":2,"name":"Shanize Forte","email":"","orcid":"","institution":"Rice University","correspondingAuthor":false,"prefix":"","firstName":"Shanize","middleName":"","lastName":"Forte","suffix":""},{"id":598161973,"identity":"a08ea0bf-2ae4-41bc-bdab-ac0056a6d903","order_by":3,"name":"Jianlin Zhou","email":"","orcid":"","institution":"Rice Universit","correspondingAuthor":false,"prefix":"","firstName":"Jianlin","middleName":"","lastName":"Zhou","suffix":""},{"id":598161974,"identity":"3442ebd1-5640-4d49-a917-628e799bd23f","order_by":4,"name":"Claudine Katan","email":"","orcid":"https://orcid.org/0000-0002-2017-5823","institution":"Univ Rennes","correspondingAuthor":false,"prefix":"","firstName":"Claudine","middleName":"","lastName":"Katan","suffix":""},{"id":598161975,"identity":"38181ccc-dad1-48da-95a4-e43f6f824604","order_by":5,"name":"Esther Tsai","email":"","orcid":"https://orcid.org/0000-0002-5885-6317","institution":"Brookhaven National Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Esther","middleName":"","lastName":"Tsai","suffix":""},{"id":598161976,"identity":"43be70cb-8802-4de7-bad4-73f5eb4a7359","order_by":6,"name":"Jacky Even","email":"","orcid":"https://orcid.org/0000-0002-4607-3390","institution":"Institut FOTON UMR CNRS 6082","correspondingAuthor":false,"prefix":"","firstName":"Jacky","middleName":"","lastName":"Even","suffix":""}],"badges":[],"createdAt":"2026-02-22 21:50:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8941587/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8941587/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103957003,"identity":"aa7fd278-acaa-4b4e-8ba6-8bfdb9738a13","added_by":"auto","created_at":"2026-03-05 03:25:27","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":963199,"visible":true,"origin":"","legend":"\u003cp\u003eFamily of hydrate phases of 2D perovskites with linear alkylammonium spacer cations. a) Concept of generalized phase instability in 2D perovskites. The shape, size, and chemistry of the spacer cation control the stable phases (represented by shaded areas) which appear on the A′I – PbI\u003csub\u003e2\u003c/sub\u003e – H\u003csub\u003e2\u003c/sub\u003eO ternary phase diagram, which influences the rate of humidity-induced degradation. b) The crystal structure of the butylammonium principal hydrate phase (4A)\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e4\u003c/sub\u003eI\u003csub\u003e10\u003c/sub\u003e·2H\u003csub\u003e2\u003c/sub\u003eO, shown along two crystallographic axes. The structure closely resembles that of the MAPbI\u003csub\u003e3\u003c/sub\u003e hydrate MA\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e6\u003c/sub\u003e·2H\u003csub\u003e2\u003c/sub\u003eO, but with four octahedral units per ribbon rather than two. c) The inorganic and organic sites of the principal hydrate phases of linear alkylammonium cations from MA to 8A. The 2A and 7A hydrates do not conform to the series. d) The inorganic and organic sites of the hydrate phases of linear alkyldiammonium cations from 2DA to 10DA. The 3DA and 5DA hydrates do not conform to the series.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-8941587/v1/6b131ba509c8084896c5d2d1.png"},{"id":103957000,"identity":"2e3abbf8-fc4b-4f7d-9abc-cedba6cf8cf1","added_by":"auto","created_at":"2026-03-05 03:25:26","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":327539,"visible":true,"origin":"","legend":"\u003cp\u003eDegradation of 2D n=1 perovskites in humid air. a) Schematic of the in-situ humidity chamber. The chamber is filled with liquid water and sealed with a Kapton film lid, allowing for in-situ characterization of crystal structure via XRD. b) A schematic of a typical perovskite sample’s Bragg peak intensity over time, showing the (i) moisture annealing, (ii) linear degradation, and (iii) exponential tail regions. The fitting procedure employed in this report compares the slope of the linear degradation regime, with the intensity normalized to the extrapolated y-intercept of the linear fit. c) Normalized (002) Bragg peak intensity over time (left) and fitted degradation rates (right) for 2D n=1 powder samples degraded with 8mL H\u003csub\u003e2\u003c/sub\u003eO loading. d) Normalized (002) Bragg peak intensity over time (left) and fitted degradation rates (right) for 2D n=1 film samples degraded with 10mL H\u003csub\u003e2\u003c/sub\u003eO loading. The orange background denotes 2D perovskites with as-gown hydrates of m-value 4, while the green background denotes those with hydrates of m-value 5 or more. e) Pseudo-binary A′I – PbI\u003csub\u003e2\u003c/sub\u003e phase diagrams for 4A (left), 5A (center), and 6A (right) showing the more favorable three-phase field above 6A n=1. The phase field connectivity of each phase diagram is inferred on the degradation rate and products observed. The maximum solubility defines a point on the vertical axis bounding the cyan and grey zones. The 0D phases reported on the horizontal axis are described in Table S3.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-8941587/v1/8753e28ea058ad0aa913fa32.png"},{"id":103957028,"identity":"d3cdbee7-2b4c-4f83-9e56-01be5bca7178","added_by":"auto","created_at":"2026-03-05 03:25:29","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":958568,"visible":true,"origin":"","legend":"\u003cp\u003eDegradation of MA-2D n=2 perovskites in humid air. a) Normalized (002) Bragg peak intensity over time for MA-2D n=2 powder samples degraded with 8mL H\u003csub\u003e2\u003c/sub\u003eO loading. b) The structure of the mixed MA-A′ hydrates. The MA-A′ hydrates with A′=4A, 5A, 6A resemble the 3A principal hydrate (left) and with A′=7A, 8A resemble the 4A principal hydrate (right). c) XRD patterns over time for powders of 4A n=2 (top), 5A n=2(middle) and 6A n=2 (bottom) degraded with 8mL H\u003csub\u003e2\u003c/sub\u003eO loading. d)\u0026nbsp; Quaternary MAI – A′I – PbI\u003csub\u003e2\u003c/sub\u003e – H\u003csub\u003e2\u003c/sub\u003eO phase diagrams for 4A (top), 5A (middle), and 6A (bottom). The four-phase field above n=2 is shown in red.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-8941587/v1/d9a452781b4576932c2a5f9c.png"},{"id":103957016,"identity":"61921e62-0df5-415f-a651-61fa8326fbb1","added_by":"auto","created_at":"2026-03-05 03:25:29","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":283117,"visible":true,"origin":"","legend":"\u003cp\u003eDegradation of mixed 3D-2D perovskites in humid air. a) XRD patterns over time for MAPbI\u003csub\u003e3\u003c/sub\u003e films capped with 5A (left), 7A (middle), and 6A (right) n=1. b) Normalized (100) Bragg peak intensity over time for MAPbI\u003csub\u003e3\u003c/sub\u003e films doped with 25mol% A′I (left) and MAPbI\u003csub\u003e3\u003c/sub\u003e films with a capping layer of A′ n=1 (right), degraded with 2mL H\u003csub\u003e2\u003c/sub\u003eO loading. The fitted degradation rates corresponding to these peak intensities are shown in the respective insets.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-8941587/v1/46c90319a14a4d4c422ebb95.png"},{"id":103957043,"identity":"dd65d250-e00b-4994-b69f-b2e0e0332fa4","added_by":"auto","created_at":"2026-03-05 03:25:36","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3215249,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8941587/v1/35a74ee6-e911-4402-9ed9-68d52c816915.pdf"},{"id":103956999,"identity":"09e59a4a-a474-4c64-83a1-825cde729ffa","added_by":"auto","created_at":"2026-03-05 03:25:22","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":23826816,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"HydratesSIsubmission.docx","url":"https://assets-eu.researchsquare.com/files/rs-8941587/v1/e9a612d984842b06f5577ef0.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Design principle for mitigating moisture induced degradation in 2D halide perovskites","fulltext":[{"header":"Introduction","content":"\u003cp\u003eHalide perovskites are emerging semiconductor materials with applications in photovoltaics and optoelectronics. Despite promising device performance, 3-dimensional (3D) halide perovskites suffer from poor durability upon exposure to light, heat, oxygen, and/or humidity, a key barrier to implementation (\u003cem\u003e1\u003c/em\u003e). The incorporation of the related two-dimensional (2D) halide perovskites can markedly improve 3D perovskite durability (\u003cem\u003e2\u003c/em\u003e). However, the origin of the 2D perovskites’ superior stability and its limitations are still not entirely understood, in particular under humidity. The 3D perovskite MAPbI\u003csub\u003e3\u003c/sub\u003e (MA=methylammonium) degrades in humid conditions via the formation of a “monohydrate” phase, MAPbI\u003csub\u003e3\u003c/sub\u003e⋅H\u003csub\u003e2\u003c/sub\u003eO\u0026nbsp;(\u003cem\u003e3\u003c/em\u003e), and eventually a “dihydrate” phase, MA\u003csub\u003e4\u003c/sub\u003ePbI\u003csub\u003e6\u003c/sub\u003e⋅2H\u003csub\u003e2\u003c/sub\u003eO\u0026nbsp;(\u003cem\u003e4\u003c/em\u003e). The monohydrate phase consists of “ribbons” of edge-sharing PbI\u003csub\u003e6\u003c/sub\u003e octahedra, whereas the dihydrate phase shows disconnected zero-dimensional (0D) PbI\u003csub\u003e6\u003c/sub\u003e. Both hydrate phases destroy the perovskite sublattice of corner-sharing octahedra, removing the useful optoelectronic properties of MAPbI\u003csub\u003e3\u003c/sub\u003e. At higher humidity levels, the MA cation will exit the lattice to leave the PbI\u003csub\u003e2\u003c/sub\u003e lattice.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn contrast, 2D perovskite degradation in humid air is not well-understood. A few multi-layered 2D perovskites have been reported to degrade into lower-n-value (inorganic layer thickness) 2D perovskites and MA hydrates (\u003cem\u003e5\u003c/em\u003e–\u003cem\u003e8\u003c/em\u003e). Humidity-induced degradation phases have been reported for certain spacer cations (\u003cem\u003e9\u003c/em\u003e–\u003cem\u003e21\u003c/em\u003e), most famously for phenyldimethylammonium (\u003cem\u003e9\u003c/em\u003e). Other reports have attributed unassigned structural signatures during degradation to potential hydrates (\u003cem\u003e22\u003c/em\u003e–\u003cem\u003e24\u003c/em\u003e). However, the prevalence and influence of 2D perovskite hydrates are not well understood and as a result, a general framework to describe 2D perovskite degradation mechanisms in humid air remain elusive. Moreover, the relative stability of Ruddlesden-Popper (RP) and Dion-Jacobson (DJ) phase 2D perovskites has been debated for several years. In most contexts, particularly in devices, DJ perovskites appear more stable (\u003cem\u003e25\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e); at other times, particularly in unencapsulated films, RP perovskites can show better stability (\u003cem\u003e27\u003c/em\u003e). As state-of-the-art 3D perovskite solar cells continue to incorporate 2D perovskites and 2D forming cations as passivation layers for the synergistic improvement of stability and efficiency, understanding 2D humidity-induced degradation is crucial to predict and enhance the stability of perovskite devices.\u003c/p\u003e\n\u003cp\u003eHere, using linear alkylammonium spacer cations as a model system, we establish a new framework for understanding moisture-induced degradation in 2D perovskites based on the trajectory of the material across the spacer-cation iodide (A′I or A′I₂)–PbI₂–H₂O ternary phase diagram. We show that 2D perovskites incorporating linear alkylammonium spacer cations, NH₃(CH₂)ₙ₋₁CH₃ (n = 3–8), form distinct hydrate phases that are structurally analogous to the monohydrate phase of MAPbI₃; referred to as the “principal hydrates” of 2D perovskites. Notably, signatures of these hydrates have been reported previously, but were misassigned to lower-n 2D perovskite phases\u0026nbsp;(\u003cem\u003e5\u003c/em\u003e). In addition to these principal hydrates, 2D perovskites also form spacer-cation-rich 0D phases that resemble the dihydrate phase of MAPbI₃. Extending beyond this series of linear spacer cations, we observe hydrate and other non-perovskite crystalline phases for nearly all spacer cations examined, across both RP and DJ perovskite families. The emergence of these non-perovskite phases introduces additional degradation pathways under humid conditions, thereby reducing perovskite phase stability. Because the perovskite, hydrate, non-perovskite crystalline, and solvated phases are all composed of A′I (or A′I₂), PbI₂, and H₂O, they can be represented within a single A′I–PbI₂–H₂O ternary phase diagram. Crucially, the molecular structure of the spacer cation dictates which non-perovskite phases populate this diagram and, in turn, governs the stability of the perovskite phase (Figure 1a). These observations point to a new materials-engineering principle: suppressing moisture-induced degradation by simplifying the accessible phase space of 2D perovskites.\u003c/p\u003e\n\u003cp\u003eThroughout this work, two-dimensional (2D) perovskite spacer cations are denoted by the number of carbon atoms in the organic chain. For example, A′ = NH₃(CH₂)₃CH₃ (butylammonium) is labeled “4A,” and the corresponding RP perovskite (4A)₂PbI₄ is denoted as 4A n = 1. An analogous notation is adopted for DJ perovskites: NH₃(CH₂)₄NH₃ (butyldiammonium) is labeled “4DA,” and (4DA)PbI₄ is denoted as 4DA n = 1. Higher-order 2D perovskites (n ≥ 2) are labeled as A–A′ n ≥ 2, where A represents either MA or formamidinium (FA) as the small A-site cation and A′ denotes the spacer cation, with the same convention extended to larger n values. For clarity in selected figures, MA–4A n = 2 and FA–4A n = 2 are abbreviated as M4 and F4, respectively, with analogous abbreviations used for other spacer cations.\u003c/p\u003e"},{"header":"Results and Discussion","content":"\u003ch3\u003e\u003cstrong\u003e1. Growing and characterizing 2D perovskite hydrates\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eIn order to understand the ternary phase diagrams associated with the linear alkylammonium spacer cations, we grew all nonperovskite phases compatible which each cation. Single crystals of principal 2D perovskite hydrates were grown from mixed HI-H\u003csub\u003e2\u003c/sub\u003eO solution. The structures of these hydrate phases are described in Figure 1b-c and Table S1. Ribbons of edge-sharing octahedra having stoichiometry [Pb\u003csub\u003em\u003c/sub\u003eI\u003csub\u003e2m+2\u003c/sub\u003e]\u003csup\u003e2-\u003c/sup\u003e (with m being the width of each ribbon in number of octahedra) coordinate with an H\u003csub\u003e2\u003c/sub\u003eO molecule at each edge. The charge of these ribbons is compensated by two alkylammonium spacer cations which occupy the structure\u0026rsquo;s \u0026ldquo;organic site\u0026rdquo;. The crystals share a distinctive needle-like morphology (Figure S1). Their powder X-ray diffraction (XRD) patterns show prominent peaks corresponding to the (100) and (10-1) planes (Figure S2).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe distance between spacer cations showed a \u0026ldquo;zig-zag\u0026rdquo; behavior with increasing cation length (Figure S3). When incrementing the chain length causes the m-value to increase (e.g. from 3A to 4A), the organic site volume and cation spacing increases. Conversely, when incrementing the chain length does not increase m-value (e.g. from 4A to 5A), the spacing decreases. Since both cation length and inorganic site width are discretized, a few specific cations (e.g. 2A and 7A) cannot be accommodated. Rather than a hydrate, 2A forms face-sharing (2A)PbI\u003csub\u003e3\u003c/sub\u003e (Figure S4 and Table S2). The 7A hydrate showed hair-like crystals which were too thin to characterize via single-crystal diffraction; as such its structure is still unidentified, but 1D XRD pattern confirms it as distinct from the principal hydrates (Figure S5). 9A and 10A were not found to form hydrates, suggesting m=6 may be the upper limit of structural stability.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe same synthesis procedure was employed to grow hydrates of the alkyldiammonium DJ perovskites (NH\u003csub\u003e3\u003c/sub\u003e(CH\u003csub\u003e2\u003c/sub\u003e)\u003csub\u003eN\u003c/sub\u003eNH\u003csub\u003e3\u003c/sub\u003e)PbI\u003csub\u003e4\u003c/sub\u003e with N = 2 - 10 as shown in Figures 1d and S6 and described in Table S1. These hydrates likewise form ribbons of [Pb\u003csub\u003em\u003c/sub\u003eI\u003csub\u003e2m+2\u003c/sub\u003e]\u003csup\u003e2-\u003c/sup\u003e,but alternate between two H\u003csub\u003e2\u003c/sub\u003eO (e.g. 4DA) and one H\u003csub\u003e2\u003c/sub\u003eO (e.g. 6DA) per organic site. \u0026nbsp;3DA and 5DA did not conform to the series (Figure S7). Structural distortion is much higher in DJ hydrates than in RP hydrates, as a result of the more rigid organic site bridged by a single spacer (Figure S8). The option of removing an H\u003csub\u003e2\u003c/sub\u003eO allows more spacers to be accommodated despite this rigidity. Interestingly, in addition to the 7DA principal hydrate, the 7DA cation by itself is the perfect length to fill the organic site of (7DA)Pb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e6\u003c/sub\u003e, which adopts m=2 ribbons tilted in alternating directions without H\u003csub\u003e2\u003c/sub\u003eO (Figure S9). Compared to RP, DJ perovskites form hydrate phases readily (rather than PbI\u003csub\u003e2\u003c/sub\u003e or perovskite phases) when introduced directly to water, suggesting the DJ hydrates are lower-energy structures than RP hydrates (Figure S10).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn order to grow phases with high A\u0026prime;I (RP) or A\u0026prime;I\u003csub\u003e2\u003c/sub\u003e (DJ) concentration, we added PbI\u003csub\u003e2\u003c/sub\u003e to extremely A\u0026prime;I-rich aqueous solutions (Figure S11-S12 and Table S3). A\u0026prime;-rich phases were found for RP cations up to 6A and DJ cations up to 9DA. Many share the 0D structure of MA\u003csub\u003e4\u003c/sub\u003ePbI\u003csub\u003e6\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHydrates of other common perovskite cations were also grown to explore the space of possible structures. For example, potassium, ammonium, and FA hydrates form tilted m=2 ribbons similar to (7DA)Pb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e6\u003c/sub\u003e but with a drastically different organic site (Figure S13). Here H\u003csub\u003e2\u003c/sub\u003eO forms honeycomb-like ribbons reminiscent of the structure of ice, with A\u0026prime; cations substituting for H\u003csub\u003e2\u003c/sub\u003eO to compensate charge. The ammonium hydrate has been reported previously (\u003cem\u003e28\u003c/em\u003e), as well as a hydronium acid hydrate which adopts a similar structure (\u003cem\u003e29\u003c/em\u003e), but to our knowledge the FA hydrate has not been reported and may be important in the degradation of FAPbI\u003csub\u003e3\u003c/sub\u003e devices. The benzylammonium (BZA) hydrate, which has also been reported (\u003cem\u003e15\u003c/em\u003e), resembles the principal hydrates but with tilted m=3 ribbons (Figure S14). Phenethylammonium and cyclohexylmethylammonium showed no nonperovskite phases.\u003c/p\u003e\n\u003cp\u003eFinally, secondary hydrate phases were observed for some linear spacer cations (Figure S15), termed \u0026ldquo;transient hydrates\u0026rdquo;. The 5A transient hydrate (5A)\u003csub\u003e6\u003c/sub\u003ePb\u003csub\u003e5\u003c/sub\u003eI\u003csub\u003e16\u003c/sub\u003e\u0026sdot;4H\u003csub\u003e2\u003c/sub\u003eO shows a mixture of m=1 and m=2 ribbons. The 4A transient hydrate (4A)\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e3\u003c/sub\u003eI\u003csub\u003e8\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO was only isolated as an oriented film during in-situ degradation and resembles the BZA hydrate. The 10DA transient hydrate (10DA)Pb\u003csub\u003e3\u003c/sub\u003eI\u003csub\u003e8\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO was inferred from its in-situ Bragg peaks resembling those of (4A)\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e3\u003c/sub\u003eI\u003csub\u003e8\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO, but this phase was not investigated in detail.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWith a list of the structures that form from A\u0026prime;I, PbI\u003csub\u003e2\u003c/sub\u003e, and H\u003csub\u003e2\u003c/sub\u003eO, we can hypothesize the PbI\u003csub\u003e2\u003c/sub\u003e - A\u0026prime;I - H\u003csub\u003e2\u003c/sub\u003eO ternary phase diagram for each spacer cation. In the absence of complete knowledge of each phase\u0026rsquo;s Gibbs free energy, we inferred the phase field connectivity of each phase diagram based on the degradation rate and products observed.\u0026nbsp;Rather than conventional triangle plots, we represent these ternary diagrams as pseudo-binary A\u0026prime;I-PbI\u003csub\u003e2\u003c/sub\u003e phase diagrams, with the x-axis corresponding to \u003cem\u003erelative\u003c/em\u003e A\u0026prime;I vs PbI\u003csub\u003e2\u003c/sub\u003e concentration and the y-axis corresponding to \u003cem\u003eabsolute\u003c/em\u003e H\u003csub\u003e2\u003c/sub\u003eO concentration (Figure S16). In this representation, moving upwards on the diagram corresponds to introducing H\u003csub\u003e2\u003c/sub\u003eO without changing the A\u0026prime;I: PbI\u003csub\u003e2\u003c/sub\u003e ratio.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhile there have been reports of the 3D perovskite lattice accommodating a small amount of interstitial H\u003csub\u003e2\u003c/sub\u003eO\u0026nbsp;(\u003cem\u003e30\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e), for the purposes of this study the position of the solid phase nodes on the diagram was assumed to be fixed at the stoichiometry of the ideal crystal. However, the solvated phase boundary will vary depending on the water solubility of A\u0026prime;I. We measured A\u0026prime;I solubility by adding H\u003csub\u003e2\u003c/sub\u003eO incrementally to A\u0026prime;I powder (Table S9). As shown in Figure S17, MA was found to be less soluble than longer alkylammonium cations, with 4A being the most soluble. Compared to RP, DJ cations showed significantly lower solubility. On all phase diagrams presented, the saturation A\u0026prime;I concentration is given by the intersection of the A\u0026prime;I \u0026ndash; H\u003csub\u003e2\u003c/sub\u003eO axis (the left vertical axis) with the first phase boundary from the bottom, which is the upper boundary of the cyan-colored three-phase field (Figure S17).\u003c/p\u003e\n\u003ch3 id=\"_Toc194976036\"\u003e\u003cstrong\u003e2. Understanding structure-property relationships for 2D perovskite hydrates\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003e\u003cstrong\u003ei. Exploring the degradation of RP n=1 alkylammonium hydrates\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe hypothesized a relationship between a 2D perovskite\u0026rsquo;s phase diagram and its moisture stability, and we tested this hypothesis using in situ XRD during degradation of 2D powders and films in a home-built humidity chamber (Figure 2a and S18).\u0026nbsp;The n=1 2D perovskites with spacer cations 4A - 8A were compared, with degradation rate quantified by the decrease in the out-of-plane (200) perovskite peak intensity. We controlled the humidity in the chamber through the addition of varying amounts of liquid water. Powders with 8mL H\u003csub\u003e2\u003c/sub\u003eO and films with 10mL H\u003csub\u003e2\u003c/sub\u003eO were chosen as representative \u0026ldquo;less aggressive\u0026rdquo; and \u0026ldquo;more aggressive\u0026rdquo; degradation conditions, respectively.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThree regimes could be distinguished in the changing of Bragg peak intensities over time (Figure 2b). At first, the intercalation of H\u003csub\u003e2\u003c/sub\u003eO at grain boundaries appears to facilitate the reorganization of crystallites, increasing (or sometimes sharply decreasing) diffraction intensity. This behavior is consistent with the . Afterwards, intensity decreases linearly with time as the perovskite degrades. We normalize this regime and use it to quantify stability. Finally, intensity tails off approaching complete degradation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eConventional wisdom suggests that the hydrophobicity of spacer cations imparts 2D perovskites with improved moisture stability, so stability should increase monotonically with cation length. Crucially, this relationship was not observed here. Instead, 2D perovskites demonstrated moisture stability in relation to the formation energy of their hydrate phase and the region of 2D phase stability on the ternary A\u0026prime;I-PbI\u003csub\u003e2\u003c/sub\u003e-H\u003csub\u003e2\u003c/sub\u003eO phase diagram. 4A and 5A n=1 were significantly less stable in humid air than 6A, 7A, and 8A n=1, both as powders (Figure 2c) and as films (Figure 2d). Moreover, 4A n=1 powder was slightly more stable than 5A n=1.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn-situ XRD could also determine the degradation pathway for each sample (Figures S19 and S20). We find that moisture-induced degradation can be envisioned as a quasi-equilibrium process, with samples charting a path across the ternary phase diagram. The effective H\u003csub\u003e2\u003c/sub\u003eO concentration within the sample increases, while simultaneously A\u0026prime;I escapes the perovskite lattice (particularly pronounced in films). This shifts the composition diagonally towards H\u003csub\u003e2\u003c/sub\u003eO and PbI\u003csub\u003e2\u003c/sub\u003e, such that the 2D perovskite begins converting into a mixture of the phases connected by a phase field to the n=1 phase. As shown in the pseudo-binary phase diagrams of Figure 2e, the 4A and 5A systems appear to have a three-phase field (green) connecting their 0D phase to their transient hydrate. This constrains the H\u003csub\u003e2\u003c/sub\u003eO concentration for which n=1 remains a stable phase. As shown in Figure S20, the transient 4A and 5A hydrates emerge first during degradation as expected. As the composition shifts further towards PbI\u003csub\u003e2\u003c/sub\u003e and H\u003csub\u003e2\u003c/sub\u003eO, the principal hydrate phase emerges as well. In contrast, the lack of a transient hydrate, the higher formation energy of the principal 6A hydrate, and the shifted stoichiometry of the 6A 0D phase appear to remove the connection between the hydrate and 0D phases, such that there is a large region (dark blue) of n=1 phase stability above 6A n=1. While A\u0026prime;-rich 0D phases were not observed in-situ during degradation, we hypothesize that the existence of these phases nonetheless influences stability by changing phase diagram connectivity above the n=1 phase.\u003c/p\u003e\n\u003cp\u003eA further corroboration of the quasi-equilibrium picture of hydration can be seen in Figure S21a-b, which shows the sudden emergence of the principal 4A hydrate phase in a degraded 4A n=1 film when the lid is removed from the humidity chamber. The composition moves diagonally upwards across the phase diagram during degradation, and then vertically downward when the lid is removed, ending in a different three-phase field (yellow or red) where the principal hydrate is stable.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe phase diagram picture can explain the quicker degradation of 5A n=1 powders compared to 4A n=1. The higher A\u0026prime;I concentration in the transient 5A hydrate compared to the transient 4A hydrate moves the 5A hydrate phase closer to 5A n=1 on the phase diagram. As a result, for the same H\u003csub\u003e2\u003c/sub\u003eO concentration a greater fraction of 5A n=1 will convert to hydrate phase compared to 4A n=1, accelerating degradation despite the increased bulkiness of the 5A cation.\u003c/p\u003e\n\u003cp\u003eGiven the 6A phase diagram, it appears contradictory that 6A n=1 films degrade into PbI\u003csub\u003e2\u003c/sub\u003e rather than 6A hydrate phase (Figure S21c-d). In fact, films of more stable 2D perovskite were often found to degrade directly into PbI\u003csub\u003e2\u003c/sub\u003e. One explanation for this behavior could be the existence of a barrier to moisture intercalation (Figure S22). At the outset of degradation, the horizontally-oriented n=1 films show a hydrophobic surface that should oppose moisture ingress. It seems reasonable that the conversion from the n=1 phase to the hydrate phase at the surface of 4A and 5A n=1 films damages the lattice and accelerates the intercalation of H\u003csub\u003e2\u003c/sub\u003eO, which in turn accelerates the phase conversion process in a self-reinforcing manner. Because 6A n=1 would not be expected to form hydrate phase at the earliest stage of moisture intercalation this self-reinforcing degradation might be replaced by a slow solvation of surface 6AI without moisture ingress. As a result, the composition of the system appears to move not diagonally but horizontally towards PbI\u003csub\u003e2\u003c/sub\u003e over time.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eii. Rationalizing the poor moisture stability of Dion-Jacobson perovskites\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe hydrate-phase-mediated moisture degradation picture introduced above helps explain the discrepancy between RP and DJ stability as devices vs as films. DJ hydrates are less structurally complex (i.e. lower m-value) than RP hydrates of the same spacer cation length, perhaps explaining why DJ hydrates form much more readily than RP hydrates (Figure S10). A lower hydrate formation energy would tend to connect the hydrate phase with the left corner of the phase diagram, constraining n=1 phase stability. As a result, DJ perovskite powders degrade into hydrate phases faster than RP perovskites of the same chain length (Figure S23), despite DJ spacer cations being less water-soluble than RP (Figure S17). The increase in moisture stability with chain length is more monotonic for DJ n=1 than for RP n=1, and moreover the hydrate phase (rather than only PbI\u003csub\u003e2\u003c/sub\u003e) appears in all samples, suggesting that these DJ powders all have similar ternary diagram connectivity (Figure S24).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDJ films degraded quickly in humid air, even with only 2mL H\u003csub\u003e2\u003c/sub\u003eO loading (Figure S25). Film stability improved monotonically with cation length except for 7DA, which showed unique instability likely due to additional nonperovskite phases in close proximity to n=1 on its ternary diagram. A similar argument can be made for the instability of 5DA, which despite not having a ribbon-like hydrate phase nonetheless has a comparatively small region of perovskite stability. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAs a result, while DJ 2D perovskites have been found to show better stability against light and heat, they are relatively more susceptible to humidity-induced degradation compared to RP 2D perovskites. Since humidity has a more pronounced effect for bare films compared to encapsulated devices, DJ devices can tend to show enhanced stability over RP while the trend is reversed for bare films.\u0026nbsp;\u003c/p\u003e\n\u003ch3 id=\"_Toc194976037\"\u003e\u003cstrong\u003e3. Degradation of 2D perovskites with n\u0026gt;1\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eMany applications of 2D perovskites require n-values greater than n=1, in particular for optimal energy alignment to 3D perovskites, and so we also degraded phase-pure n=2 powders of alkylammonium 2D perovskites, (NH\u003csub\u003e3\u003c/sub\u003e(CH\u003csub\u003e2\u003c/sub\u003e)\u003csub\u003eN-1\u003c/sub\u003eCH\u003csub\u003e3\u003c/sub\u003e)\u003csub\u003e2\u003c/sub\u003e(A)Pb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e7\u003c/sub\u003e, using both MA and FA as the A-site cation. MA n=2 perovskites were universally less stable than their n=1 counterparts, while FA n=2 was more stable (Figures 3a and S26). We attribute the greater instability of MA-based 2D perovskites to two sources. First, despite MAI having similar solubility to FAI and lower solubility than alkylammonium spacer cation salts, the low mass of MA\u003csup\u003e+\u003c/sup\u003e cation appears to allow MAI to readily escape the perovskite lattice and complex with water at the film surface. This promotes the formation of n=1 during degradation. Second, \u0026ldquo;mixed\u0026rdquo; MA-A\u0026prime; hydrate phases introduce new degradation pathways (Figure 3b and Table S4). For the n=2 perovskites (4A)\u003csub\u003e2\u003c/sub\u003eMAPb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e7\u003c/sub\u003e (MA-4A n=2), (5A)\u003csub\u003e2\u003c/sub\u003eMAPb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e7\u003c/sub\u003e (MA-5A n=2), and (6A)\u003csub\u003e2\u003c/sub\u003eMAPb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e7\u003c/sub\u003e (MA-6A n=2) a mixed-cation (MA)(A\u0026prime;)Pb\u003csub\u003e3\u003c/sub\u003eI\u003csub\u003e8\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO hydrate grows from dilute hydriodic acid solution with a similar structure to (3A)\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e3\u003c/sub\u003eI\u003csub\u003e8\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO, having an m-value of 3 and a mixed occupation of the organic site by MA and A\u0026prime;. The Pb-Pb interatomic spacing increases with spacer length, from 4.73 \u0026Aring; (4A) to 4.77 \u0026Aring; (5A) to 4.80 \u0026Aring; (6A), compared to 4.77 \u0026Aring; for (3A)\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e3\u003c/sub\u003eI\u003csub\u003e8\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO. Similarly, mixed MA-7A, MA-8A and MA-9A hydrates were found with composition (MA)(A\u0026prime;)Pb\u003csub\u003e4\u003c/sub\u003eI\u003csub\u003e10\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO resembling (4A)\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e4\u003c/sub\u003eI\u003csub\u003e10\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO, and MA-2A hydrates were found with composition (MA)(2A)Pb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e6\u003c/sub\u003e resembling (MA)\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e6\u003c/sub\u003e. Additionally, a mixed MA-4A 0D phase was identified with composition (MA)\u003csub\u003e2\u003c/sub\u003e(4A)\u003csub\u003e4\u003c/sub\u003ePbI\u003csub\u003e8\u003c/sub\u003e (Figure S27a).\u003c/p\u003e\n\u003cp\u003eAs seen in Figures 3c, S28, and S29, MA-based n=2 perovskite powders (Figure S28) and films (Figure S29) degrade into a combination of phases including n=1, PbI\u003csub\u003e2\u003c/sub\u003e, and various hydrates. The degradation products are corners of the four-phase field above n=2 on the quaternary MAI-A\u0026prime;I-PbI\u003csub\u003e2\u003c/sub\u003e-H\u003csub\u003e2\u003c/sub\u003eO phase diagram (Figure 3d). For 4A n=2 the degradation products observed are n=1 and the MA-4A hydrate, suggesting the four-phase field shown in Figure 3d. This region is quite narrow in the direction of increasing H\u003csub\u003e2\u003c/sub\u003eO, due to a connection between the mixed MA-4A hydrate and 4A n=1 and as a result MA-4A n=2 has poor phase stability.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor MA-5A n=2, the stoichiometry of the transient 5A hydrate blocks a connection between the mixed hydrate and 5A n=1. As a result, the four-phase field appears to be comprised of MA-5A n=2, 5AI, the mixed MA-5A hydrate, and the transient hydrate, with the n=1 phase no longer forming. This widens the four-phase field above 5A n=2, which may improve stability compared to 4A n=2. The increased mass of 5AI may also improve stability by suppressing solvation. Films of MA-5A n=2 show the formation of n=1 and the principal hydrate instead of the transient hydrate, indicating the movement to a different four-phase field away from MAI and towards PbI\u003csub\u003e2\u003c/sub\u003e as the organic cations escape.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMA-6A n=2 degrades first into n=1 and subsequently into the principal hydrate, which corresponds to a notably different four-phase field than for 4A or 5A. Crucially, this region connects near pure H\u003csub\u003e2\u003c/sub\u003eO instead of at the 6A 0D phase, significantly widening it in the H\u003csub\u003e2\u003c/sub\u003eO direction and explaining the marked increase in stability between 5A and 6A. The Pb-Pb expansion in the MA-6A mixed hydrate appears to raise its formation energy, reducing its connectivity to other phases. For 6A n=2 films, PbI\u003csub\u003e2\u003c/sub\u003e also emerges as a degradation product suggesting the escape of MAI and 6AI from the sample.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMA-8A n=2 degrades via the same pathway as MA-6A n=2, with the formation of the n=1 and principal hydrate phases. However, the degradation occurs notably faster both for powder and film samples in 8A n=2 compared to 6A. MAI appears to escape readily from the 8A n=2 lattice, indicated by the predominance of n=1 and the lack of PbI\u003csub\u003e2\u003c/sub\u003e in degrading 8A n=2 films. This behavior was unique to 8A and is still not fully understood; MA-7A, 9A, and 10A n=2 show exceptional stability (possibly due to their lack of principal hydrates) (Figure S30).\u003c/p\u003e\n\u003cp\u003eSimilarly, we studied degradation in FA n=2 powders (Figure S31) and films (Figure S32). Compared to MA, the FA cation has a lower acidity, a greater mass, and a smaller dipole moment (\u003cem\u003e32\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e). As a result, perovskites incorporating FA at the A-site have been found to be more chemically stable than those with MA under a variety of conditions, including in the presence of water (\u003cem\u003e30\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e). Unsurpisingly, FA-based 2D perovskites were also universally more stable in this present study. No mixed FA-A\u0026prime; hydrate phases were observed during degradation, suggesting a less complex phase diagram with fewer degradation pathways. Moreover, suppressed n=1 formation suggests that FAI escapes less readily than MAI. These factors combined gave FA n=2 much higher moisture stability than MA n=2 for all 2D perovskites studied.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFA-4A n=2 degraded first into 4A transient hydrate, and subsequently into 4A principal hydrate. In contrast, FA-5A n=2 degraded into the principal hydrate directly. The lack of the transient phase as a degradation pathway gives FA-5A n=2 a markedly higher stability than FA-4A n=2 and suggests the FA-4A and FA-5A phase diagrams have different connectivity. This may be related to the existence of a mixed FA-4A perovskitoid, (FA)(4A)PbI\u003csub\u003e4\u003c/sub\u003e (Figure S27b), although the phase is not observed during degradation. In both FA-4A and FA-5A, the FA hydrate FA\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e6\u003c/sub\u003e\u0026middot;4H\u003csub\u003e2\u003c/sub\u003eO appears during early stages of degradation and subsequently fades.\u003c/p\u003e\n\u003cp\u003eFor spacer cations longer than 5A, a marked increase in stability is once again observed. FA-6A and FA-7A n=2 degraded into PbI\u003csub\u003e2\u003c/sub\u003e, while FA-8A degraded into n=1 along with a small amount of 8A principal hydrate. FA-9A and 10A likewise formed n=1, suggesting 2D perovskites with longer cations give up FA from the A-site more willingly (Figure S33).\u003c/p\u003e\n\u003ch3 id=\"_Toc194976038\"\u003e\u003cstrong\u003e4.\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eDegradation of perovskites with nonlinear spacer cations\u0026nbsp;\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eNext we compared the degradation behavior of the linear cations 4A and 5A to their \u003cem\u003eiso\u003c/em\u003e counterparts isobulyammonium (i-4A) and isopentylammonium (i-5A). i-4A forms a tilted ribbon-like hydrate phase with m=2 and stoichiometry (i-4A)\u003csub\u003e2\u003c/sub\u003ePb\u003csub\u003e2\u003c/sub\u003eI\u003csub\u003e6\u003c/sub\u003e\u0026middot;2H\u003csub\u003e2\u003c/sub\u003eO, but has no 0D phase. In contrast, i-5A forms the mixed corner-face-sharing perovskitoid (i-5A)\u003csub\u003e4\u003c/sub\u003ePb\u003csub\u003e3\u003c/sub\u003eI\u003csub\u003e10\u003c/sub\u003e and the 0D phase (i-5A)\u003csub\u003e6\u003c/sub\u003ePbI\u003csub\u003e8\u003c/sub\u003e. The pseudo-binary phase diagrams for i-4A and i-5A are shown in Figure S34. Both i-4AI and i-5AI are less water-soluble than 4AI and 5AI.\u003c/p\u003e\n\u003cp\u003eCompared to 4A and 5A, perovskites with i-4A and i-5A are significantly more stable (Figure S35). This can be attributed to their more favorable phase diagrams, both of which have large regions of phase stability above n=1. i-4A n=1 is moreover more stable than i-5A n=1 likely because the i-4A hydrate stoichiometry is further from n=1 compared to the i-5A perovskitoid. As with other 2D perovskites that showed near-vertical 2D-liquid phase boundaries, the composition of i-4A and i-5A n=1 appears to move horizontally towards PbI\u003csub\u003e2\u003c/sub\u003e; for i5A this leads to the formation of the H\u003csub\u003e2\u003c/sub\u003eO-free perovskitoid, whereas for i4A this causes the direct formation of PbI\u003csub\u003e2\u003c/sub\u003e. For MA-based i-4A and i-5A n=2 samples (Figure S36), degradation proceeds through the conversion of n=2 to n=1; as these phases have the same position on both phase diagrams, i-5A n=2 shows better stability than i-4A n=2 due to its increased mass. The examples of i-4A and i-5A show how the spacer can be engineered to improve the phase diagram, increasing moisture stability.\u003c/p\u003e\n\u003ch3 id=\"_Toc194976039\"\u003e\u003cstrong\u003e5. Degradation of mixed 3D-2D perovskites and 3D-2D stacks\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eFinally, we investigated the moisture stability of 3D-2D perovskite films, incorporating 2D both as an additive to form a 3D-2D composite and also as a capping layer to form a bilayer structure. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ei. 3D MAPbI\u003csub\u003e3\u003c/sub\u003e with spacer cation salt additives\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe measured the degradation of MAPbI\u003csub\u003e3\u003c/sub\u003e films doped with 25 mol% A\u0026prime;I salt (spacer length from MA to 10A) using 2mL H\u003csub\u003e2\u003c/sub\u003eO loading (Figures S37 and S38). Four distinct degradation routes were observed: (1) In undoped films and films with MAI and 2AI, the formation of PbI\u003csub\u003e2\u003c/sub\u003e and marginal MA or MA-2A hydrate, respectively; (2) with 4AI and 5AI, the formation of mixed MA-A\u0026prime; hydrates; (3) with 6AI, the formation of 2D and PbI\u003csub\u003e2\u003c/sub\u003e; (4) with 7AI and 8AI, the formation of 2D and mixed MA-A\u0026prime; hydrates (Figure S39a-b).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe fact that undoped MAPbI\u003csub\u003e3\u003c/sub\u003e degrades mainly into PbI\u003csub\u003e2\u003c/sub\u003e (not MA hydrate) suggests that MAI escapes rapidly, leading to a path through composition space slanted severely towards PbI\u003csub\u003e2\u003c/sub\u003e (Figure S40). In other words, the degradation behavior mimics that of 6A n=1 in Figure S22, though in this case not because of a barrier to moisture ingress but rather because of the greater tendency of MA to exit the film. Extra MAI shifts the composition further from PbI\u003csub\u003e2\u003c/sub\u003e, promoting the formation of MA hydrate.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdditives such as 2AI, 6AI, 7AI, and 8AI show similar stability to undoped MAPbI\u003csub\u003e3\u003c/sub\u003e. Conversely, additives with low-energy hydrates such as 3AI, 4AI, and 5AI (4A and 5A forming mixed MA-A\u0026prime; hydrates) in fact have significantly worse stability, unlocking faster degradation pathways for MAPbI\u003csub\u003e3\u003c/sub\u003e through hydrate formation (Figure 4b). This picture is corroborated by the behavior of 8AI, 9AI, and 10AI dopants (Figure S38); 9AI and 10AI confer improved stability over 8AI due to the suppression of mixed hydrate phases (9A shows marginal m=4 mixed hydrate, while 10A shows none).\u003c/p\u003e\n\u003cp\u003eAdditive molecules longer than 5A formed grains of 2D perovskites in-situ during degradation. Maximal-n-value phases (n=4-6) formed first, with subsequent emergence of lower n-values, consistent with the expected behavior of A\u0026prime;I at grain boundaries intercalating into the 3D lattice. Mixed-hydrate formation also accelerated the degradation of these high-n-value grains.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eii. 2D n=1 perovskite capping layers on top of 3D MAPbI\u003csub\u003e3\u003c/sub\u003e\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNext, we deposited capping layers of 4A to 8A n=1 perovskite atop MAPbI\u003csub\u003e3\u003c/sub\u003e (Figure 4 and S41). As with doped films, the 2D spacer cation strongly influences MAPbI\u003csub\u003e3\u003c/sub\u003e films\u0026rsquo; stability and degradation pathway. Distinct degradation routes were observed for a) uncapped MAPbI\u003csub\u003e3\u003c/sub\u003e and MAPbI\u003csub\u003e3\u003c/sub\u003e with 6A n=1, b) MAPbI\u003csub\u003e3\u003c/sub\u003e with 4A and 5A =1, and c) MAPbI\u003csub\u003e3\u003c/sub\u003e with 7A and 8A n=1, as can be seen by plotting the Bragg peak intensities of the hydrates and PbI\u003csub\u003e2\u003c/sub\u003e over time (Figure S42). Once again, mixed hydrate formation allows MAPbI\u003csub\u003e3\u003c/sub\u003e to degrade faster with 4A or 5A n=1 layers than with no cap (Figure 4b), despite the lack of PbI\u003csub\u003e2\u003c/sub\u003e suggesting MA escape is suppressed (Figure 4a). In contrast, a 6A n=1 layer suppressed MAPbI\u003csub\u003e3\u003c/sub\u003e degradation and formed majority PbI\u003csub\u003e2\u003c/sub\u003e. Moreover, intercalation across the interface formed successively higher-n-value 2D up to n=5. The fact that n=2 forms first, rather than n=4-6, suggests the MA cation migrates upwards rather than 6A downwards, consistent with a greater mobility of the MA cation predicted from the formation of PbI\u003csub\u003e2\u003c/sub\u003e in undoped MAPbI\u003csub\u003e3\u003c/sub\u003e films. 7A and 8A capping layers showed a mixture of the degradation pathways seen for 4A/5A and for 6A, with stability intermediate between them. While still showing 3D-2D layer intercalation as with 6A, the degradation into PbI\u003csub\u003e2\u003c/sub\u003e is suppressed by ~10 times and the formation of mixed hydrates is enhanced by ~10 times as with 4A/5A, suggesting that MA-7A and MA-8A hydrates are lower-energy phases than MA-6A hydrate (Figure S39c and S42). The prominence of the MA hydrate in MA-7A and MA-8A compared to MA-4A and MA-5A also may indicate a suppression of MA escape with longer cations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eiii. 2D n=1 perovskite capping layers on top of 3D FAPbI\u003csub\u003e3\u003c/sub\u003e\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFinally, we investigated the degradation of FAPbI\u003csub\u003e3\u003c/sub\u003e films with n=1 capping layers using 8mL water loading (Figure S43). Capped FAPbI\u003csub\u003e3\u003c/sub\u003e peak intensities dropped quickly at the outset, possibly due to film damage during layer deposition. However, with no mixed FA hydrates, all capping layers improved FAPbI\u003csub\u003e3\u003c/sub\u003e stability over longer timeframes. Unlike other samples tested, \u0026alpha;-FAPbI\u003csub\u003e3\u003c/sub\u003e is not thermodynamically stable, and its \u0026delta;-phase appears on the phase diagram instead. Bare FAPbI\u003csub\u003e3\u003c/sub\u003e films degraded principally into PbI\u003csub\u003e2\u003c/sub\u003e with marginal \u0026delta;-phase and no FA hydrate, suggesting FA escape dominates over moisture intercalation. The 4A and 5A n=1 top layers quickly became FA n=2 during fabrication and degraded similarly to FA n=2 samples. Both 4A and 5A slightly suppressed PbI\u003csub\u003e2\u003c/sub\u003e and promoted \u0026delta;-phase in the degrading FAPbI\u003csub\u003e3\u003c/sub\u003e layer, suggesting a retention of FAI. 6A and 8A n=1 capping layers remained mostly n=1 after fabrication but converted to n=2 rapidly with humidity. These n=2 layers persisted throughout the measurement period and blocked the formation of both PbI\u003csub\u003e2\u003c/sub\u003e and \u0026delta;-phase in the FAPbI\u003csub\u003e3\u003c/sub\u003e layer. The nonexistence of FA n\u0026gt;2 2D removes intercalation as a degradation pathway, further improving 3D-2D stability compared to MAPbI\u003csub\u003e3\u003c/sub\u003e-2D films.\u003c/p\u003e\n\u003cp\u003e7A uniquely remained n=1 for much longer, and its conversion to n=2 appeared to take precedence over degradation of the underlying FAPbI\u003csub\u003e3\u003c/sub\u003e. This may suggest a higher formation energy of FA-7A n=2, or perhaps a sparsely populated FA-7A phase diagram.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003e6. Conclusions\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eThis report demonstrates a new theory of 2D perovskite degradation in humid air, introducing the ternary A\u0026prime;I - PbI\u003csub\u003e2\u003c/sub\u003e - H\u003csub\u003e2\u003c/sub\u003eO phase diagram to determine the degradation pathway. Using this framework, new design principles can be conceptualized focusing on suppressing the formation of all but the \u0026ldquo;unavoidable\u0026rdquo; degradation phases at the corners of the phase diagram: solid A\u0026prime;I, a solvated liquid phase, and solid PbI\u003csub\u003e2\u003c/sub\u003e. Because of the tendency of organic cations to escape the system through solvation with the ambient humid air, the effective composition of the system tends towards PbI\u003csub\u003e2\u003c/sub\u003e and away from the A\u0026prime;I side over time. In four-component systems such as n\u0026ge;2 perovskites, the composition will also tend away from the more mobile cation salt, usually leading to a reduction in n-value as the lighter A-site cation preferentially escapes.\u003c/p\u003e\n\u003cp\u003eThis theory has been explored using the linear alkylammonium spacer cations from MA to 8A, which share a class of ribbon-like hydrate phases. We can engineer spacer cations to be structurally incompatible with low-energy hydrates, reducing the number of \u0026ldquo;avoidable\u0026rdquo; phases in the phase diagram and improving moisture stability, by a factor of 6 when replacing butylammonium with isobutylammonium.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe anticipate that this work will accelerate the rational engineering of spacer cations as the field pushes towards ultra-stable 2D perovskites. More broadly, a phase diagram framework may be useful in other aspects of perovskite engineering. For example, the solvent complexes on a 2D perovskite\u0026rsquo;s A\u0026prime;I \u0026ndash; PbI\u003csub\u003e2\u003c/sub\u003e \u0026ndash; DMF phase diagram may likewise influence solubility and film formation, as has been explored for MAPbI\u003csub\u003e3\u003c/sub\u003e (\u003cem\u003e36\u003c/em\u003e). We hope this work will inspire deeper investigation into underexplored phases observed during perovskite processing and degradation.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003ei. Growth of the hydrate phases of 2D perovskites\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e2D perovskite hydrates were grown by dissolving PbI\u003csub\u003e2\u003c/sub\u003e and A′I salts in hydriodic acid diluted with deionized water. All chemicals were purchased from Sigma Aldrich. PbI\u003csub\u003e2\u003c/sub\u003e and the spacer cation salts MAI, FAI, 3AI, 4AI, 5AI, 6AI, and 8AI were purchased in their iodide salt form. 7AI, 4DAI\u003csub\u003e2,\u0026nbsp;\u003c/sub\u003e5DAI\u003csub\u003e2\u003c/sub\u003e, 6DAI\u003csub\u003e2\u003c/sub\u003e, 7DAI\u003csub\u003e2\u003c/sub\u003e, 8DAI\u003csub\u003e2\u003c/sub\u003e, 9DAI\u003csub\u003e2\u003c/sub\u003e,10DAI\u003csub\u003e2\u003c/sub\u003e, i-4AI, and i-5AI were synthesized by reacting the associated amine molecule with hydriodic acid in ethanol solvent. The iodide salts were collected as a white precipitate after evaporation of ethanol.\u003c/p\u003e\n\u003cp\u003eTo grow the hydrates, PbI\u003csub\u003e2\u003c/sub\u003e and A′I salt were combined with HI and H\u003csub\u003e2\u003c/sub\u003eO in the amounts noted in Tables S5-S8. The mixture was heated to 150\u003csup\u003eo\u003c/sup\u003eC under stirring for 1-2 hours, cooled to room temperature, and allowed to sit for several days. The PbI\u003csub\u003e2\u003c/sub\u003e was often found to never fully dissolve in the solution. After some time white or pale yellow needle-like crystals of hydrate phase began growing above the solid PbI\u003csub\u003e2\u003c/sub\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo grow high-A′ nonperovskite crystals, A′I was added to H\u003csub\u003e2\u003c/sub\u003eO near its concentration limit along with a small amount of PbI\u003csub\u003e2\u003c/sub\u003e. Amounts of each precursor are noted in Table S7. The mixture was dissolved at 165\u003csup\u003eo\u003c/sup\u003eC over 1-2 hours and allowed to cool slowly to room temperature. In some cases, ~2mL of mineral oil was added to form a buffer layer to prevent water from condensing on the walls of the vial, which was found to improve crystal quality.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eii. Synthesis of phase-pure 2D perovskite powders\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e2D perovskite powders were grown following the established synthesis procedure of the Mohite group. PbO and MACl or FACl powders were dissolved in hydriodic acid stabilized with ~10% hypophosphorus acid (H\u003csub\u003e3\u003c/sub\u003ePO\u003csub\u003e2\u003c/sub\u003e) under stirring at 100\u003csup\u003eo\u003c/sup\u003eC. In a separate vial, the associated amine molecule of the spacer cation was dissolved in HI and reacted at 0\u003csup\u003eo\u003c/sup\u003eC to form dissolved A′I. The two vials were combined and the solution was heated to 200\u003csup\u003eo\u003c/sup\u003eC under stirring until all components dissolved. The solution was allowed to cool slowly to room temperature (n=1 and MA n=2 2D) or 80\u003csup\u003eo\u003c/sup\u003eC (FA n=2 2D). The crystals were collected via vacuum filtration and washed with heptane before being left in a vacuum oven at 70\u003csup\u003eo\u003c/sup\u003eC overnight to remove excess HI.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eiii. Fabrication of phase-pure 2D perovskite films\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e2D perovskite films were fabricated using the memory-seeds method presented by Sidhik et al. from the Mohite group (\u003cem\u003e1\u003c/em\u003e). Phase-pure 2D perovskite powders synthesized as above were dissolved in DMF with a concentration of 400mg/mL (RP) or dissolved in 4:1 DMF:DMSO with a concentration of 350mg/mL (DJ) under Ar atmosphere. Solutions were stirred at 70\u003csup\u003eo\u003c/sup\u003eC for 3 hours before spin coating. 1”x1” glass substrates were cleaned via sonication sequentially in DI water with liquinox detergent, pure DI water, acetone, and 1:1 acetone:ethanol before being dried with an air gun and cleaned with UV-ozone treatment for 30 minutes. To spin-coat films, 70µL of precursor solution was pipetted onto the surface of the substrate and spun at 5000RPM/3500RPMS for 30 seconds without antisolvent (RP) or with 120µL chlorobenzene(CBZ) antisolvent after 20 seconds (DJ) under Ar atmosphere. Films were annealed at 100\u003csup\u003eo\u003c/sup\u003eC for 10 minutes in Ar.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eiv. Fabrication of 3D-2D perovskite bilayer films\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e3D MAPbI\u003csub\u003e3\u003c/sub\u003e films were fabricated from 1M precursor solutions of 1:1 PbI\u003csub\u003e2\u003c/sub\u003e:MAI dissolved in DMF. 5 mol% MACl was added to improve film quality. Solutions were stirred at 70\u003csup\u003eo\u003c/sup\u003eC for 4 hours. Substrates were prepared as described above. MAPbI\u003csub\u003e3\u003c/sub\u003e films were spin-coated for 30 seconds at 5000RPM/3500RPMS with 120µL CBZ antisolvent added after 20 seconds. Films were annealed at 100\u003csup\u003eo\u003c/sup\u003eC for 10 minutes. 3D FAPbI\u003csub\u003e3\u003c/sub\u003e films were fabricated from 1M precursor solutions of 1:1 PbI\u003csub\u003e2\u003c/sub\u003e:FAI dissolved in 5:1 DMF:DMSO. 35mol% MACl was added to promote perovskite phase formation and improve film quality as shown by Kim et al. (\u003cem\u003e2\u003c/em\u003e). Solutions were stirred at 70\u003csup\u003eo\u003c/sup\u003eC for 4 hours. Substrates were prepared as described above. FAPbI\u003csub\u003e3\u003c/sub\u003e films were spin-coated for 30 seconds at 5000RPM/3500RPMS with 1000 µL diethyl ether antisolvent added after 10 seconds. Films were annealed at 150\u003csup\u003eo\u003c/sup\u003eC for 30 minutes.\u003c/p\u003e\n\u003cp\u003e2D top layers were added following the procedure of Sidhik et al. (\u003cem\u003e3\u003c/em\u003e) 50mg of 2D perovskite powder was dissolved in 1000µL of acetonitrile (ACN) and stirred at room temperature for 1-2 hours. MAPbI\u003csub\u003e3\u003c/sub\u003e films were placed once again on the spin coater chuck and spun at 5000RPM for 30 seconds. After reaching max speed, 50µL of ACN solution was dynamically spin-coated onto the surface of the film. The resulting bilayer films were annealed at 100\u003csup\u003eo\u003c/sup\u003eC for 5 minutes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ev. Fabrication of 3D perovskite films with 2D additive\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e3D MAPbI\u003csub\u003e3\u003c/sub\u003e precursor solutions were fabricated as above, with an additional 25 mol% of A′I additive introduced to the solution. The spin-coating and annealing procedure was the same as for the 3D MAPbI\u003csub\u003e3\u003c/sub\u003e layer of 3D-2D bilayer films.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003evi. Determination of saturation concentration of A′I salts in water\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA pre-weighed amount of A′I salt was added to a vial containing 5mL of hexane as a noninteracting barrier liquid to prevent water ingress or evaporation. DI water was added to the solution 20µL at a time. After each addition vials were vigorously stirred and sonicated until the partial solvation of A′I salt was complete. This was repeated until no solid phase could be visibly observed in the vial, at which point the total volume of water added was recorded (Table S9).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003evii. Construction of the phase diagram\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEach phase diagram was populated with the phases identified in our screening of the ternary composition space, including the pure components PbI\u003csub\u003e2\u003c/sub\u003e, A′I or A′I\u003csub\u003e2\u003c/sub\u003e (A′ = 4A, 5A, 4DA, etc.), and H\u003csub\u003e2\u003c/sub\u003eO; the perovskite phase A′\u003csub\u003e2\u003c/sub\u003ePbI\u003csub\u003e4\u003c/sub\u003e (RP), A′PbI\u003csub\u003e4\u003c/sub\u003e (DJ), or APbI\u003csub\u003e3\u003c/sub\u003e (3D); the principal and transient hydrate phases; and any other nonperovskite phases observed. The connectivity of the phase diagram was taken from the convex hull of the Gibbs free energy surface across the ternary composition space. To construct the phase diagram, the assumption was made that each observed phase intersects with the convex hull, that the A′I and H\u003csub\u003e2\u003c/sub\u003eO form an ideal solution, and that the energy surface of all solid phases is a Kronniker delta at each phase’s stoichiometry. Measuring the true free energy of the phases was beyond the scope of this project; instead we inferred the connectivity of nodes by the order in which phases were observed during degradation, and with the assumption that phases of similar structure likely had similar free energy. Quaternary phase diagrams were constructed identically, taking the convex hull of the Gibbs Free energy surface across the quaternary composition space.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe color-coding of phase fields was kept consistent between phase diagrams. Two-phase fields are colored gray; three-phase fields which include one liquid phase are colored blue or purple; three-phase which include PbI\u003csub\u003e2\u003c/sub\u003e are colored red (if all solid phases) or purple (if mixed solid and liquid phases). Additional three-phase fields, if they exist, are colored green – yellow – brown from left to right (i.e. high-A′I to high-PbI\u003csub\u003e2\u003c/sub\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eviii. Measuring degradation with in-situ 1D-XRD\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSample degradation was measured in-situ by means of a home-built humidity chamber described in Figure 1 and Figure S17. The free volume of the humidity chamber was 17mL. The chamber consisted of a “well” with free volume 11mL, and a “lid” with free volume 6mL (accounting for 0.2mL sample volume). Kapton tape was used to form an air-tight seal around the lid which X-rays could pass through. Liquid water (either 2, 8, or 10mL) was added to the chamber and allowed to sit for \u0026gt;3 minutes before beginning the measurement to allow the temperature of the water to equilibrate the temperature inside the XRD. The humidity inside the chamber always reached 100%RH regardless of water volume, but excess water reduced the free volume of air in the chamber and accelerated degradation.\u003c/p\u003e\n\u003cp\u003eIn-situ XRD was carried out using a Rigaku Smartlab XRD with Bragg-Brentano focusing using a Cu target and a filter to suppress the Cu K-β signature. Diffraction patterns were collected across a 2θ range of 2\u003csup\u003eo\u003c/sup\u003e to 20\u003csup\u003eo\u003c/sup\u003e at a speed of 10 degrees per minute (resolution 0.001\u003csup\u003eo\u003c/sup\u003e). After each 2\u003csup\u003eo\u003c/sup\u003e to 20\u003csup\u003eo\u003c/sup\u003e scan completed the next scan was immediately started. Post-degradation diffraction patterns with the chamber lid removed were collected across an identical angle range and at the same speed.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eix. Characterizing crystal structure with single crystal XRD\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHydrate single crystals were grown as discussed above. To preserve the liquid water environment of the crystals, 1mL of mineral oil was added directly to the vial of synthesized hydrate in solution. The water was then extracted with a pipette leaving the hydrate and a small layer of water surrounding it submerged in mineral oil. Crystals were placed under a visible microscope and attached to a MiTeGen crystallography sample holder using parabar. Single crystal XRD was carried out using a Rigaku Synergy-S with an MO source and a HyPix-Arc150 detector.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ex. Characterizing crystal structure of 4A transient hydrate with in-situ GIWAXS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe transient 4A hydrate phase was characterized in-situ via GIWAXS conducted at Brookhaven National Lab’s NSLS-II facility 11-BM (CMS). A film of 4A n=1 was loaded into the humidity chamber described above and mounted in CMS’s sample staging area. GIWAXS was conducted with 13.5 keV X-Ray energy using a Pilatus-1M detector positioned maximally close to the sample to increase the q-range of the diffraction. After ~10 minutes of degradation in humid air an oriented film of transient hydrate was observed on the surface of the sample. The measurement continued for 1 hr until the 2D perovskite signature was no longer visible. The q\u003csub\u003ez\u003c/sub\u003e-axis peak positions seen in GIWAXS were calibrated against the more accurate peak positions determined via 1D XRD. The WAXS pattern suggested an orthorhombic lattice with lattice parameters close to those of the 3A principal hydrate. With reference to other ribbon-like orthorhombic hydrates, we assumed that adjacent ribbons tilted along the b-axis to fix β=90°. The offset of adjacent ribbons along the b-axis was surmised by comparing the measured WAXS pattern with simulated WAXS patterns of space groups Pc2a (no offset), Pnmn (b/2 offset), and F2dd (b/4 offset). The Pc2a space group best aligned with the observed (hkl) Bragg peak intensities for k≠0. \u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eN. Li, X. Niu, Q. Chen, H. Zhou, Towards commercialization: the operational stability of perovskite solar cells. \u003cem\u003eChem. Soc. Rev.\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, 8235\u0026ndash;8286 (2020).\u003c/li\u003e\n\u003cli\u003eI. Metcalf, S. Sidhik, H. Zhang, A. Agrawal, J. Persaud, J. Hou, J. Even, A. D. Mohite, Synergy of 3D and 2D Perovskites for Durable, Efficient Solar Cells and Beyond. \u003cem\u003eChem. Rev.\u003c/em\u003e \u003cstrong\u003e123\u003c/strong\u003e, 9565\u0026ndash;9652 (2023).\u003c/li\u003e\n\u003cli\u003eG. H. Imler, X. Li, B. Xu, G. E. Dobereiner, H.-L. Dai, Y. Rao, B. B. Wayland, Solid state transformation of the crystalline monohydrate (CH3NH3)PbI3(H2O) to the (CH3NH3)PbI3 perovskite. \u003cem\u003eChem. Commun.\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, 11290\u0026ndash;11292 (2015).\u003c/li\u003e\n\u003cli\u003eA. M. Leguy, Y. Hu, M. Campoy-Quiles, M. I. Alonso, O. J. Weber, P. Azarhoosh, M. Van Schilfgaarde, M. T. Weller, T. Bein, J. Nelson, Reversible hydration of CH3NH3PbI3 in films, single crystals, and solar cells. \u003cem\u003eChemistry of Materials\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 3397\u0026ndash;3407 (2015).\u003c/li\u003e\n\u003cli\u003eB. R. Wygant, A. Z. Ye, A. Dolocan, Q. Vu, D. M. Abbot, C. B. Mullins, Probing the Degradation Chemistry and Enhanced Stability of 2D Organolead Halide Perovskites. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e141\u003c/strong\u003e, 18170\u0026ndash;18181 (2019).\u003c/li\u003e\n\u003cli\u003eJ. Tang, W. Tian, C. Zhao, Q. Sun, C. Zhang, H. Cheng, Y. Shi, S. Jin, Imaging the Moisture-Induced Degradation Process of 2D Organolead Halide Perovskites. \u003cem\u003eACS Omega\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 10365\u0026ndash;10371 (2022).\u003c/li\u003e\n\u003cli\u003eB. R. Wygant, G. T. Geberth, A. Z. Ye, A. Dolocan, D. E. Cotton, S. T. Roberts, D. A. Vanden Bout, C. B. Mullins, Moisture-driven formation and growth of quasi-2-D organolead halide perovskite crystallites. \u003cem\u003eACS Applied Energy Materials\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 6280\u0026ndash;6290 (2020).\u003c/li\u003e\n\u003cli\u003eJ. Schlipf, Y. Hu, S. Pratap, L. Bie\u0026szlig;mann, N. Hohn, L. Porcar, T. Bein, P. Docampo, P. M\u0026uuml;ller-Buschbaum, Shedding Light on the Moisture Stability of 3D/2D Hybrid Perovskite Heterojunction Thin Films. \u003cem\u003eACS Appl. Energy Mater.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 1011\u0026ndash;1018 (2019).\u003c/li\u003e\n\u003cli\u003eA. Dučinskas, G. Y. Kim, D. Moia, A. Senocrate, Y.-R. Wang, M. A. Hope, A. Mishra, D. J. Kubicki, M. Siczek, W. Bury, T. Schneeberger, L. Emsley, J. V. Milić, J. Maier, M. Gr\u0026auml;tzel, Unravelling the Behavior of Dion\u0026ndash;Jacobson Layered Hybrid Perovskites in Humid Environments. \u003cem\u003eACS Energy Lett.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 337\u0026ndash;344 (2021).\u003c/li\u003e\n\u003cli\u003eR. N. Nurdillayeva, R. F. Moral, M. Pols, D.-K. Lee, V. Altoe, C. P. Schwartz, S. Tao, C. M. Sutter-Fella, Humidity Disrupts Structural and Chiroptical Properties of Chiral 2D Perovskites. \u003cem\u003eACS Nano\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 11348\u0026ndash;11357 (2025).\u003c/li\u003e\n\u003cli\u003eD. G. Billing, A. Lemmerer, Synthesis and crystal structures of inorganic\u0026ndash;organic hybrids incorporating an aromatic amine with a chiral functional group. \u003cem\u003eCrystEngComm\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 686\u0026ndash;695 (2006).\u003c/li\u003e\n\u003cli\u003eA. Dučinskas, G. C. Fish, M. A. Hope, L. Merten, D. Moia, A. Hinderhofer, L. C. Carbone, J.-E. Moser, F. Schreiber, J. Maier, J. V. Milić, M. Gr\u0026auml;tzel, The Role of Alkyl Chain Length and Halide Counter Ion in Layered Dion\u0026minus;Jacobson Perovskites with Aromatic Spacers. \u003cem\u003eJ. Phys. Chem. Lett.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 10325\u0026ndash;10332 (2021).\u003c/li\u003e\n\u003cli\u003eY. Zhang, Z. Wang, S. Hu, P. Yan, H. Li, C. Sheng, Robust and Swiftly Reversible Thermochromic Behavior of a 2D Perovskite of (C6H4(CH2NH3)2)(CH3NH3)[Pb2I7] for Smart Window and Photovoltaic Smart Window Applications. \u003cem\u003eACS Appl. Mater. Interfaces\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 12042\u0026ndash;12048 (2021).\u003c/li\u003e\n\u003cli\u003eJ. Lai, J. Tan, R. Zhu, W. Lv, S. Ye, Conformational Disordering of Organic Cations during Moisture-Induced Structural Collapse in Organic\u0026ndash;Inorganic Hybrid Perovskites. \u003cem\u003eACS Appl. Energy Mater.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 10521\u0026ndash;10527 (2024).\u003c/li\u003e\n\u003cli\u003eI. W. H. Oswald, A. A. Koegel, J. R. Neilson, General Synthesis Principles for Ruddlesden\u0026ndash;Popper Hybrid Perovskite Halides from a Dynamic Equilibrium. \u003cem\u003eChem. Mater.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 8606\u0026ndash;8614 (2018).\u003c/li\u003e\n\u003cli\u003eX. Jiang, J. Gu, S. Zheng, T. Zhang, X. Li, Y. Guan, C. Li, J. Wang, W. Zhang, Y. Fu, Mixed-Solvent Crystallization Expands the Chemical Space of 2D Perovskites and Enables Phonon-Limited Exciton Transport. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e147\u003c/strong\u003e, 32287\u0026ndash;32298 (2025).\u003c/li\u003e\n\u003cli\u003eM.-Y. Zhu, L.-X. Zhang, J. Yin, J.-J. Chen, L.-J. Bie, A fluorescence quenching sensor for Fe3+ detection using (C6H5NH3)2Pb3I8\u0026middot;2H2O hybrid perovskite. \u003cem\u003eInorganic Chemistry Communications\u003c/em\u003e \u003cstrong\u003e109\u003c/strong\u003e, 107562 (2019).\u003c/li\u003e\n\u003cli\u003eR. Han, Q. Yue, Y. Cao, D. Wei, X. Liu, J. Lin, Highly Sensitive Two-Dimensional Ruddlesden-Popper Perovskites for Thermochromic Smart Windows. \u003cem\u003eChemNanoMat\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, e202400306 (2024).\u003c/li\u003e\n\u003cli\u003eW. T. Sun, Z. Xing, A. Sergeev, Y. He, A. M. C. Ng, K. S. Wong, L. Molčanov, J. Popović, A. B. Djuri\u0026scaron;ić, Effects of ambient exposure on photoluminescence of Dion\u0026ndash;Jacobson tin-based halide perovskites. \u003cem\u003eJ. Mater. Chem. C\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 2737\u0026ndash;2747 (2023).\u003c/li\u003e\n\u003cli\u003eY. Zhang, Y. Fu, X. Song, X. Zhang, Releasable Water Charge-Trapping and Water-Resistant Photodetection using 1D Perovskitoid Hydrate Single Crystal. \u003cem\u003eAdvanced Materials\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 2204710 (2022).\u003c/li\u003e\n\u003cli\u003eI. W. Gilley, H. W. Kwon, C. Liu, Y. Yang, C. Huang, H. Wan, A. S. R. Bati, E. H. Oriel, M. Kepenekian, B. Vishal, S. Zeiske, K. S. Bayikadi, T. E. Wiggins, E. S. Vasileiadou, B. Chen, R. D. Schaller, J. Even, S. De Wolf, E. H. Sargent, M. G. Kanatzidis, Combining Organic Cations of Different Sizes Grants Improved Control over Perovskitoid Dimensionality and Bandgap. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e147\u003c/strong\u003e, 7777\u0026ndash;7787 (2025).\u003c/li\u003e\n\u003cli\u003eK. Yang, J. Zhao, K. Zhang, B. Shao, F. Cui, X. Yang, Humidity Stability of 2D Dion\u0026ndash;Jacobson-Phase Perovskites with Hydrophobic Groups in Diammonium Spacers. \u003cem\u003eInd. Eng. Chem. Res.\u003c/em\u003e \u003cstrong\u003e62\u003c/strong\u003e, 16154\u0026ndash;16162 (2023).\u003c/li\u003e\n\u003cli\u003eY. Liu, H. Zhou, Y. Ni, J. Guo, R. Lu, C. Li, X. Guo, Revealing stability origin of Dion-Jacobson 2D perovskites with different-rigidity organic cations. \u003cem\u003eJoule\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 1016\u0026ndash;1032 (2023).\u003c/li\u003e\n\u003cli\u003eK. Meng, X. Wang, Z. Li, Z. Liu, Z. Qiao, C. Wang, Y. Hu, S. Li, L. Cheng, Y. Zhai, G. Chen, Self-passivation of low-dimensional hybrid halide perovskites guided by structural characteristics and degradation kinetics. \u003cem\u003eEnergy Environ. Sci.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 2357\u0026ndash;2368 (2021).\u003c/li\u003e\n\u003cli\u003eJ. Gong, M. Hao, Y. Zhang, M. Liu, Y. Zhou, Layered 2D Halide Perovskites beyond the Ruddlesden\u0026ndash;Popper Phase: Tailored Interlayer Chemistries for High-Performance Solar Cells. \u003cem\u003eAngewandte Chemie International Edition\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e, e202112022 (2022).\u003c/li\u003e\n\u003cli\u003eS. Ahmad, P. Fu, S. Yu, Q. Yang, X. Liu, X. Wang, X. Wang, X. Guo, C. Li, Dion-Jacobson Phase 2D Layered Perovskites for Solar Cells with Ultrahigh Stability. \u003cem\u003eJoule\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 794\u0026ndash;806 (2019).\u003c/li\u003e\n\u003cli\u003eE. S. Vasileiadou, B. Wang, I. Spanopoulos, I. Hadar, A. Navrotsky, M. G. Kanatzidis, Insight on the Stability of Thick Layers in 2D Ruddlesden\u0026ndash;Popper and Dion\u0026ndash;Jacobson Lead Iodide Perovskites. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e143\u003c/strong\u003e, 2523\u0026ndash;2536 (2021).\u003c/li\u003e\n\u003cli\u003eL. Prisinzano, D. Delmonte, K. C. Ravaglia, V. Vit, L. Righi, Dehydration process in 1D ammonium lead halide and mixing of organic cations in hybrid perovskites through mechanosynthesis. \u003cem\u003eMater. Res. Express\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 115503 (2020).\u003c/li\u003e\n\u003cli\u003eS. Sobczak, A. M. Fidelli, J.-L. Do, G. P. Demopoulos, A. Moores, T. Fri\u0026scaron;čić, A. Katrusiak, High-pressure observation of elusive iodoplumbic acid in different hydronium-hydrate solid forms. \u003cem\u003eInorg. Chem. Front.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 735\u0026ndash;744 (2024).\u003c/li\u003e\n\u003cli\u003eE. Smecca, Y. Numata, I. Deretzis, G. Pellegrino, S. Boninelli, T. Miyasaka, A. L. Magna, A. Alberti, Stability of solution-processed MAPbI 3 and FAPbI 3 layers. \u003cem\u003ePhysical Chemistry Chemical Physics\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 13413\u0026ndash;13422 (2016).\u003c/li\u003e\n\u003cli\u003eF. Cordero, F. Craciun, A. M. Paoletti, G. Zanotti, Structural Transitions and Stability of FAPbI3 and MAPbI3: The Role of Interstitial Water. \u003cem\u003eNanomaterials\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1610 (2021).\u003c/li\u003e\n\u003cli\u003eD. H. Fabini, C. C. Stoumpos, G. Laurita, A. Kaltzoglou, A. G. Kontos, P. Falaras, M. G. Kanatzidis, R. Seshadri, Reentrant structural and optical properties and large positive thermal expansion in perovskite formamidinium lead iodide. \u003cem\u003eAngewandte Chemie\u003c/em\u003e \u003cstrong\u003e128\u003c/strong\u003e, 15618\u0026ndash;15622 (2016).\u003c/li\u003e\n\u003cli\u003eA. Binek, F. C. Hanusch, P. Docampo, T. Bein, Stabilization of the trigonal high-temperature phase of formamidinium lead iodide. \u003cem\u003eThe journal of physical chemistry letters\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 1249\u0026ndash;1253 (2015).\u003c/li\u003e\n\u003cli\u003eT. Zhang, X. Meng, Y. Bai, S. Xiao, C. Hu, Y. Yang, H. Chen, S. Yang, Profiling the organic cation-dependent degradation of organolead halide perovskite solar cells. \u003cem\u003eJ. Mater. Chem. A\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1103\u0026ndash;1111 (2017).\u003c/li\u003e\n\u003cli\u003eW. Ke, L. Mao, C. C. Stoumpos, J. Hoffman, I. Spanopoulos, A. D. Mohite, M. G. Kanatzidis, Compositional and Solvent Engineering in Dion\u0026ndash;Jacobson 2D Perovskites Boosts Solar Cell Efficiency and Stability. \u003cem\u003eAdvanced Energy Materials\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1803384 (2019).\u003c/li\u003e\n\u003cli\u003eA. A. Petrov, A. A. Ordinartsev, K. A. Lyssenko, E. A. Goodilin, A. B. Tarasov, Ternary Phase Diagrams of MAI\u0026ndash;PbI2\u0026ndash;DMF and MAI\u0026ndash;PbI2\u0026ndash;DMSO Systems. \u003cem\u003eJ. Phys. Chem. C\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e, 169\u0026ndash;173 (2022).\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8941587/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8941587/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"We establish a novel conceptual framework which explains and mitigates moisture-induced degradation in 2D perovskites. Degradation proceeds along defined pathways across the spacer cation iodide (A′I)-PbI2-H2O ternary phase diagram as moisture intercalates into the perovskite. The structure and chemistry of the spacer cations control the availability and formation energies of non-perovskite phases, which in turn determines the degradation mechanism and rate in humid air. Guided by this framework, we establish a spacer cation design principle focused on eliminating low-energy hydrate phases to suppress moisture driven degradation, thereby enhancing 2D perovskite stability. Leveraging this mechanism, we show that replacing linear butylammonium with branched isobutylammonium suppresses moisture-induced degradation of the associated 2D powders by a factor of six. The existence of mixed-organic cation non-perovskite phases can similarly destabilize mixed 3D-2D perovskite films, and suppressing such mixed phases improves stability.","manuscriptTitle":"Design principle for mitigating moisture induced degradation in 2D halide perovskites","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-05 03:24:45","doi":"10.21203/rs.3.rs-8941587/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"32d700b7-5bb3-4f38-ab4e-743a7f7255ce","owner":[],"postedDate":"March 5th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":63655612,"name":"Physical sciences/Materials science/Materials for energy and catalysis/Solar cells"},{"id":63655613,"name":"Physical sciences/Chemistry/Materials chemistry/Electronic materials"}],"tags":[],"updatedAt":"2026-03-05T03:24:45+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-05 03:24:45","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8941587","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8941587","identity":"rs-8941587","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}