Halide Chemistry in Tin Perovskite Optoelectronics: Bottlenecks and Opportunities

2.1 Formation of I₂ in Sn HaPs

Saidaminov et al. observed that dimethyl sulfoxide (DMSO), a commonly employed solvent in perovskite precursor solutions, can oxidise SnI₂ to SnI₄ upon heating.¹⁹ A detailed investigation by Pascual et al. identified that native hydroiodic acid (HI) can promote the DMSO-mediated degradation of SnI₂ to SnI₄, proposing I₂ formation in this process as the reason behind the oxidation of Sn²⁺ in solution.¹⁵

Recently, we explored Sn HaP film decomposition under ambient environmental conditions (O₂+humidity) and detected SnI₄ as a result of an O₂-driven degradation process,⁸ in agreement with previous reports.²⁰ A key feature of SnI₄ is its rich chemical reactivity which can impact stability. Indeed, we showed that SnI₄ can either i) combine with iodosalts (AI, where A is a monovalent cation) and generate a vacancy-ordered double perovskite analogue (A₂SnI₆) or ii) more importantly, participate in a hydrolysis reaction with H₂O to give HI, which in turn reacts with O₂ to form I₂. The halogen (I₂) was then shown to further decompose perovskite (Figure 1, right) and produce more SnI₄ in a cyclic degradation process. In essence, this mechanism causes the breakdown of perovskite triggered by moisture and O₂, where native iodide ions in the lattice evolve to form aggressive I₂ species.

Further pathways leading to I₂ formation have also been identified by Sanchez-Diaz et al. Specifically, they reported the presence of a light-activated ligand-to-metal charge transfer (LMCT) process in SnI₄ within the perovskite leading to the formation of I₂ and SnI₂.²¹

Altogether, the formation of I₂ can occur through multiple chemical reactions and conditions (e.g. solution and solid-state; inert atmosphere and ambient conditions; dark and under illumination).

2.2 Halide Substitution in Sn HaPs

Despite its key role in determining materials stability, the number of studies addressing the presence and effect of I₂ as an oxidiser within Sn HaPs remains limited. As per Pb HaPs, it is likely that reducing the number of intrinsic point defects and improving the morphology of the film will improve the tolerance to ambient conditions and halogen formation.²² Another approach may be the in situ neutralisation of I₂ through the addition of sacrificial reductants to Sn HaPs, such as NaBH₄.²¹ While these strategies are helpful to improve material and device stability, only limiting the intrinsic ability of Sn HaP to form I₂ is expected to address the root cause of halogen-mediated degradation.

The substitution of X-site I⁻ in the ASnX₃ perovskite structure with bromide (Br⁻) is one approach to mitigate I₂ formation in Sn PSCs. Based on the trend of halogen oxidation strength (F₂>Cl₂>Br₂>I₂), the removal of I⁻ by Br⁻ will most likely result in a weaker tendency of Sn HaP to form halogens. Similarly, Sn⁴⁺ compounds containing Br⁻ are less susceptible to undergo LMCT photoreduction reported for SnI₄, thus limiting halogen evolution.²³ In terms of lattice energy, the smaller ionic radii of Br⁻ lead to stronger Sn−X bonds relative to I⁻, further increasing the stability of the perovskite system. Due to their small size, the remaining halides (Cl⁻ and F⁻) do not easily incorporate into lattice X sites;²⁴ however, the use of SnF₂ or SnCl₂ V_Sn filling additives to sequester Sn⁴⁺ in solution may be beneficial by limiting the presence of the species that can trigger halogen formation as per the mechanisms reviewed above (i.e. SnI₄, HI; see Figure 1, left).

Sn HaPs prepared with high fractions of Br⁻ have been shown to exhibit high stability, with i) Ippili et al. demonstrating that MASnBr₃ (where MA is methylammonium) can retain its crystal structure with minimal changes over 4 months in ambient conditions (Figure 2a)²⁵ and ii) Lee et al. and Zhu et al. reporting higher optical stability after 100 h (Figure 2b) and lower free carrier densities in Br-substituted FASnI₃ (where FA is formamidinium) and MASnI₃, respectively.^{26, 27} This suggests that the evolution of Br₂ does not readily occur in the same way as I₂. In addition to the thermodynamically unfavourable oxidation of Br⁻ and the increased Sn−Br bond strength, the lack of Br₂ most likely originates from a stronger hydrogen bonding between the A-site ammonium/amidinium hydrogens and Br⁻ ions relative to I⁻.²⁸ Furthermore, I⁻ replacement with smaller Br⁻ reduces the unit-cell dimensions, producing a more compact lattice that can potentially block O₂ and moisture ingress; these being key factors in the conversion of SnI₄ to I₂.

The benefits of incorporating small fractions of Br⁻/Cl⁻ can also be extended to the performance of fully functioning optoelectronic devices. It is noteworthy that the only two literature examples reporting Sn HaP solar cell efficiencies exceeding 14 % include small fractions of Br⁻ in the absorber layer.^{29, 30} The inclusion of Br⁻ has been shown to deactivate non-radiative carrier recombination and improve the film microstructure. However, only a few reports have assessed the role of halide substitution in the chemical and operational stability of Sn HaP materials and devices. Lee et al. demonstrated that substituting 25 % of I⁻ by Br⁻ in FASnI₃ not only improves device performance but also leads to i) a decrease of three orders of magnitude in free hole concentration and ii) better stability of encapsulated solar cells (84 % of initial efficiency kept after 1000 h under AM1.5 G light exposure).²⁶ Similarly, Yu et al. reported solar cells based on MA_0.25FA_0.75SnI_2.75Br_0.25 with efficiencies of 9.31 % in contrast to 5.02 % in FASnI₃ devices (Figure 2c). Devices with 25 % Br⁻ retained 80 % of their performance after 300 h, outliving I⁻-based devices (Figure 2d).³¹ These improvements in stability have been attributed to an increase in V_Sn formation energy and higher crystal orientation upon Br⁻ incorporation. However, we hypothesise that limiting the I⁻ content in Sn HaPs may also mitigate halogen formation and p-type doping pathways [Figure 1 (left) and Eq. (4) and (5), respectively]. In line with this, Khan et al. proposed that the addition of Br⁻ to Sn HaPs increases the formation energy of iodide vacancies (V_I), limiting harmful HI generation and accounting for the lower Sn⁴⁺ content in Br⁻-substituted films.³²

The addition of Br⁻ leads to a trade-off between light harvesting capability and intrinsic materials stability.³³ Therefore, the need to retain an ideal band gap (for solar cell devices) limits the fraction of native I⁻ that can be replaced by Br⁻. However, opportunities for high Br⁻ fraction, wide band gap perovskites may emerge in tandem solar cells,³⁴ transistors²⁷ and visible light emitters (see Section 4).^{35, 36}

4.1 Two-Dimensional Sn HaP Emitters

Two-dimensional (2D) Sn perovskites offer superior emission properties in the visible region (from ≈450 nm to ≈650 nm) as compared to their 3D counterparts.³⁵ Layered perovskites incorporate bulky organic cations in the A-site (Figure 4a), leading to higher stability and wider band gaps.⁴⁸ Bandgap broadening arises from stronger carrier confinement,⁴⁹ which in turn leads to exciton binding energies of over an order of magnitude higher in 2D perovskites relative to their 3D analogues.⁵⁰ In our previous work, we demonstrated that transitioning from MASnI₃ to (PEA)₂SnI₄ (PEA=phenylethylammonium) causes a blue-shift in the emission from ≈950 nm to ≈630 nm.³⁵ Furthermore, we found that 2D perovskites possess superior optoelectronic properties (i.e. longer charge carrier lifetime, higher photoluminescence quantum yield or PLQY). Hence, we reported the first use of such materials (e.g. (PEA)₂SnI₄) in visible light LEDs. Since our initial work, research efforts within the scientific community have focused on improving the performance of red 2D Sn HaP LEDs. To this end, a number of strategies have been explored including: i) crystal growth engineering to yield more uniform thin films,^51-53 ii) dielectric confinement enhancement through organic cation replacement^{54, 55} and iii) the use of antioxidant additives.^{56, 57} To date, (PEA)₂SnI₄ remains the champion 2D Sn HaP emitter, with record external quantum efficiencies (EQE) in LEDs of 5 % (Figure 4b)⁵⁷ and showing promising upscaling prospects.⁵⁸

To attain efficient visible emission beyond red, quantum size effects in 2D HaPs must be combined with I⁻ substitution by smaller halides (e.g. Br⁻), as shown for three-dimensional (3D) structures. Exploiting this approach, we showed that the emission of (PEA)₂SnI₄ can be tuned from ≈630 nm to ≈470 nm by progressively replacing I⁻ by Br⁻ (Figure 4c).³⁵ This wide emission range makes 2D Sn HaPs interesting candidates for next-generation LEDs with low toxicity. However, PLQY losses and lower photoluminescence (PL) lifetimes persist as unwanted consequences of Br- incorporation,^{35, 59} which make the design of efficient blue Sn HaP LEDs challenging.

4.2 Broadband Sn HaP Phosphors

The use of larger surface-to-volume ratio systems offers a route to more emissive Br⁻-based Sn HaPs. For example, Shellaiah et al. demonstrated a PLQY of over 14 % in ligand-stabilised MASnBr₃ quantum dots.⁶⁰ It is possible that ligand-mediated surface passivation may mitigate Br⁻-related losses by reducing energetic disorder. An additional route to boost light emission in Br⁻-based materials involves the use of strongly confined emitters based on long alkyl chain spacers or even zero-dimensional hybrid structures (Figure 5a). This has led to extremely long-lived emission with PLQYs that often approach unity. For example, Zhou et al. achieved PLQYs around 95 % for 0D (C₄N₂H₁₄Br)₄SnBr₆,⁶¹ whereas Zhang et al. reported PLQYs up to 88 % for 2D (C₁₈H₃₅NH₃)₂SnBr₄.⁶² Efficient radiative recombination arises from stronger electron–phonon coupling in the fully isolated inorganic regions, favouring exciton self-trapping and yielding broadband phosphorescence (Figure 5b,c). Interestingly, this class of Sn HaPs exhibit outstanding stability under a great variety of stressors (i.e. illumination, heat and ambient conditions).⁶³ The substitution of I⁻ by Br⁻ remains an effective route to increase the stability of low-dimensional Sn HaP emitters; Zhou et al. demonstrated that (C₄N₂H₁₄Br)₄SnBr₃I₃ shows higher photostability under illumination vs (C₄N₂H₁₄I)₄SnI₆,⁶⁴ following the same rationale as in their 3D counterparts (see above). Overall, these features make Sn HaP phosphors strong candidates for lighting applications beyond monochromatic light such as white-light phosphorescent devices.

5.1 Halide Defect Management

In the most studied Pb-based perovskite material, i.e. MAPbI₃, it is well established that its constituent ions, and especially I⁻, migrate upon external stimuli (e.g. bias or illumination).^{65, 66} Eames et al. identified that V_I play a key role in this process by enabling hopping of I⁻ (Figure 6a),⁶⁵ while Minns et al. demonstrated that I⁻ can further move through I_i positions via the rearrangement of MA⁺ and even form neutral I₂ within the perovskite lattice.⁶⁷ These processes are known to profoundly affect the electronic properties of Pb HaPs by generating trap states and promoting halogen formation. It is reasonable to suppose that such processes can also occur in Sn HaPs. To overcome this, treatments targeting I⁻-related defects are useful to increase both performance and stability of Pb HaPs; we envisage these strategies to benefit Sn HaPs in a similar manner. Yang et al. introduced additional I⁻ into the organic cation solution to decrease the amount of deep level defects.⁶⁸ This route yielded power conversion efficiencies of over 22 % and 19 % for small and large area devices, respectively. Our own approach consisted of applying post-treatment solutions comprising I⁻ salts onto MAPbI₃ films to fill V_I.⁶⁹ In our previous work, we demonstrated that upon light and O₂ exposure V_I act as preferred sites for the formation of superoxide (O₂⁻) species harmful to perovskite (Figure 6b).^69-72 Deactivating O₂⁻ formation pathways via V_I filling also translates into enhanced operational stability in solar cells.⁶⁹ While O₂⁻ formation in Sn HaPs is unlikely due to their prompt reaction with O₂, passivating halide defects remains important in order to limit potential halogen evolution pathways derived from ion migration.

Caution must be exercised upon treating HaPs with I⁻-based salts; Tan et al. showed that I_i defects may arise from such post-treatments, promoting unwanted cubic-to-hexagonal phase transitions in FAPbI₃ absorbers⁷³ and potentially making more I⁻ available for halogen formation in Sn HaPs. As a universal strategy, Jeong et al. recently reported the use of pseudohalide additives for defect elimination in HaP materials.⁷⁴ While Br⁻ and Cl⁻ show superior binding affinity to V_I vs I⁻, formate (HCOO⁻) pseudohalide provides the strongest interaction in FAPbI₃ (Figure 6c), leading to i) certified photovoltaic efficiencies of 25.2 %, ii) high operational stability (≈450 h) and iii) bright electroluminescence with EQE over 10 %. While pseudohalides have been implemented in Sn HaPs to control film growth,⁷⁵ further additive engineering efforts will continue to enable effective halide defect management strategies.⁷⁶ Other methods to mitigate crystallographic defects such as aerosol-assisted solvent treatments are readily applicable in Pb-free perovskite technologies.⁷⁷

5.2 Interfacial Engineering

A wide range of interface engineering strategies have been developed and successfully utilised to improve both the performance and stability of Pb HaP solar cells. For example, regulating Lewis acid–base interactions at HaP/charge transport layer junctions by selecting suitable passivating moieties in the latter is integral to allow efficient charge transfer.⁷⁸ However, such approaches remain virtually unexploited in Sn HaPs. We recently observed the ambient stability of (PEA)_0.2(FA)_0.8SnI₃ deposited onto hole transport layers (HTLs) to correlate with the hole extraction ability of the substrate (Figure 7a,b).⁸ Here we proposed that the fast removal of holes in perovskite may help to mitigate Sn²⁺-to-Sn⁴⁺ oxidation and the evolution of labile SnI₄ and enhance stability; this being consistent with the molecular oxygen and moisture-driven degradation mechanism in Figure 1. Moreover, enhancing anodic contact selectivity in Sn HaP optoelectronics therefore serves as a key device design rule for the realisation of efficient and stable solar cells. A particularly interesting approach involves novel nanomaterials that can extract charges from the device more effectively. Recently, we employed a newly discovered nanomaterial (i.e. phosphorene nanoribbons, PNRs) to improve hole extraction in Pb HaP solar cells (Figure 7c).⁷⁹ MAPbI₃-based solar cells utilising a PNR/poly(triaryl)amine (PTAA) HTL were shown to exhibit power conversion efficiencies over 21 %, a figure comparable to those in single-crystalline MAPbI₃ solar cells.⁸⁰ We also exploited ferrocene as a perovskite surface potential modifier at (FAPbI₃)_0.95(MAPbBr₃)_0.05/spiro-OMeTAD (HTL) junctions.⁸¹ This avenue yielded i) enhanced charge transfer from perovskite to spiro-OMeTAD (Figure 7d,e) and ii) highly efficient solar cells (23.45 %) with high operational stability (70 % performance retention after 1250 h). Similarly, we demonstrated that ultrathin polymethylmethacrylate (PMMA) interlayers can effectively enhance selectivity at PEDOT:PSS/(PEA)_0.2(FA)_0.8SnI₃ interfaces and boost solar cell performance from 6.5 % to 10 % (Figure 7f).⁸² While PMMA interlayers are expected to be beneficial to the operational stability of Sn HaP solar cells, further research in this direction is needed. We expect these approaches to boost hole extraction in archetypal Sn HaP device architectures based on PEDOT:PSS and other less commonly employed HTLs such as nickel oxide (NiO_x)⁸³ or ambipolar tin oxide (SnO_x).⁸⁴ Recently, Song et al. demonstrated the use of self-assembled monolayers (SAMs) on indium tin oxide (ITO) anodes in p-i-n, FASnI₃-based solar cells.⁸⁵ As also shown in Pb and mixed Sn–Pb HaP devices,^{86, 87} we expect chemically tailored SAMs to become next-generation hole-selective contacts in Sn HaP optoelectronics.

Though not being the main focus of this Minireview, the importance of interfacial chemistry at Sn HaP/electron transport layer (ETL) junctions should not be disregarded. Considering that I₂/I₃⁻ can be found at the surface of Sn HaPs even in freshly made samples,²¹ the design of Sn HaP/ETL interfaces tolerant to such degradation products (Figure 1) also warrants future research efforts. For this, the design and application of interlayers that can react with native I₂ in Sn HaPs is a promising direction. Exploiting the tendency of I₂ to bind to polymeric materials⁸⁸ may serve as an opportunity to scavenge the halogen out of Sn HaPs as it is generated. Kang et al. recently employed a polyvinylpyrrolidone–I₂ complex to passivate grain-boundary defects in Pb HaP solar cells;⁸⁹ we propose that the I₂-capturing ability of polyvinylpyrrolidone can similarly be applied in in situ protective layers for Sn HaPs.