Synthesis, Crystal Structure, Characterization, and Hydrophobicity Tests of Bismuth(III)– and Silver(I)–Triammionium Bromide Low-Dimensional Perovskites

Abstract

This work describes the synthesis, crystal structure, and hydrophobicity tests of four bismuth(III)– and silver(I)–bromide complexes using the triammonium cations diethylenetriaminonium (H₃DETA³⁺) and N,N,N′,N″,N‴-pentamethyldiethylenetriammonium (H₃PMDTA³⁺). The prepared compounds are the 0D perovskites (H₃DETA)[BiBr₆] (1), (H₃DETA)₂[AgBr₄]Br₃ (2), and (H₃PMDTA)[BiBr₆] (3), as well as the 1D/2D mixed perovskite with minimum formula (H₃PMDTA)[Ag₃Br₆] (4), being the last three novel materials. Compounds 1 and 3 crystallize in the orthorhombic P2₁2₁2₁ space group and are discrete [BiBr₆]³⁻ units with the cation surrounding them. In both compounds, the bismuth(III) metal ion is found in a distorted octahedral coordination geometry. Compound 2 crystallizes in the monoclinic P2₁/c space group, and it is a mixed salt consisting of (H₃DETA)[AgBr₄] and (H₃DETA)Br₃, whereas the silver(I) complexes are also isolated. Finally, compound 4, which crystallizes in the orthorhombic space group Pbcn, is a combination of a 2D and 1D silver–bromide perovskite, with the cations filling the voids. The 2D structure has the minimal formula [Ag₄Br₇]³⁻, with the 1D coordination polymer [Ag₂Br₅]³⁻ being both built up by a combination of bromide ions acting as tetrahedra corner and edge-sharing bridging ligands. The silver(I) in 2 and 4 is found in a tetrahedral coordination geometry. All compounds were deposited on pristine FTO glass, resulting in an increase in the contact angle from 22° to 44°, 36°, 62°, and 54° for films of 1, 2, 3, and 4, respectively. Compounds 1 and 3 were also deposited onto Cs₂AgBiBr₆ film, and the contact angles were observed to be the same as when deposited directly onto the FTO cover glass.

1. Introduction

In recent years, perovskites have attracted the attention of the scientific community due to the outstanding performance of the lead-based perovskites in photovoltaic applications, especially in solar energy generation [1,2]. On the other hand, the presence of lead limits the applications due to its toxicity [3], difficulty in disposal, and degradation upon exposure to moisture [4]. Therefore, lead-free perovskites have been proposed as a replacement for toxic lead. The bismuth(III)-silver(I) perovskite family Cs₂AgBiX₆ (X = Cl, Br) overcomes several drawbacks of lead-based perovskites, such as lower toxicity of the metal atoms in the composition and stability in an open atmosphere [5]. Yet their bandgap (2.7 eV when X = Cl and 2.2 eV when X = Br) limits their efficiency in solar energy conversion and artificial photosynthesis [6,7].

One approach to reduce the band gap of bismuth(III)-silver(I) perovskites is to synthesize them with iodine as the halogen, which theoretically reduces the band gap to 1.0–1.6 eV [7,8]. However, direct synthesis of the Cs₂AgBiI₆ is hindered by its thermodynamic instability under ambient conditions, leading to decomposition into Cs₃Bi₂I₉ and CsAg₂I₃ [9,10]. A promising route to overcome this limitation involves post-synthetic halide exchange on bromide-based nanocrystals [11], but better results for solar cell devices are found using thin films, and nanocrystal synthesis requires hydrophobic surfactants that render the materials more stable but reduce scalability.

To address this issue, an alternative strategy is proposed: the deposition of a hydrophobic chemically compatible layer on the surface of Cs₂AgBiBr₆ films prior to halide exchange, serving as a protective interface during the conversion to iodide. A similar strategy has been employed to improve the moisture resistance of lead-based devices, depositing a water-repellent film onto the perovskite layer [5,12]. Usually, these hydrophobic compounds must share structural and chemical similarity with the underlying perovskite to avoid introducing lattice mismatch or interfacial defects, while also being hydrophobic enough to slow down degradation in air [13,14]. Low-dimensional perovskite-like materials—particularly those incorporating the same metal cations (Bi³⁺ and/or Ag⁺) and Br⁻ as the host film—are ideal candidates for this role. These materials, when synthesized using bulky organic cations, induce hydrophobicity and limit how many directions in which the metal ions can form a corn-shared polyhedral structure [15,16]. When shared continuously along two axes, the material is called a 2D perovskite [17,18,19]; along one axis, it forms a 1D perovskite [20,21,22]; and if isolated or forming a discrete set of fused polyhedra, it is called a 0D perovskite [23,24,25,26].

With this in mind, we synthesized four low-dimensional bismuth(III)– and silver(I)–bromide perovskites using the organic triammonium cations diethylenetriaminonium (H₃DETA³⁺) and N,N,N’,N”,N”-pentamethyldiethylenetriammonium (H₃PMDTA³⁺), which were deposited on Cs₂AgBiBr₆ films to increase the hydrophobicity of the surface. They were prepared as single crystals, and the crystal structure is presented, as well as the hydrophobicity of the prepared films.

2. Experimental

All starting materials were used as received, without further purification. For film deposition, dimethylsulfoxide (DMSO) and dimethylformamide (DMF) were treated with 3 Å molecular sieves for at least 72 h before use.

General synthesis: For all compounds, the metal precursor was dissolved in an appropriate amount of concentrated HBr at room temperature. The mixture was then heated to 120 °C. In a separate flask, an equimolar amount of the triamine was added to 0.5 mL of concentrated HBr at room temperature and then added dropwise to the metal solution. Concentrated HBr was then slowly added to completely dissolve the resulting material while maintaining the temperature. The final solution was slowly cooled to allow crystallization of the compounds as single crystals. The crystals were filtered off, washed with cooled concentrated HBr, and dried under low pressure for 24 h.

(H₃DETA)[BiBr₆] (1)—950 mg (2.00 mmol) of Bi₂O₃ and 430 µL (4.00 mmol) of DETA. For this compound, the total volume of concentrated HBr used was 40.0 mL. Yield: 65% (2.055 g, 2.58 mmol). Exp. (Calc.) for C₄H₁₆N₃BiBr₆ (794.60 g mol⁻¹): C 6.02 (6.05), H 1.88 (2.03), N 5.08 (5.29), Bi 25.20 (26.30)%. IR (ATR, cm⁻¹): 2958, 1557, 1456, 1167, 1016, 973, 746.

(H₃DETA)₂[AgBr₄]Br₃ (2)—340 mg (2.00 mmol) of AgNO₃.xH₂O and 220 µL (2.00 mmol) of DETA. For this compound, the total volume of concentrated HBr used was 4.0 mL. Yield: 34% (0.204 g, 0.678 mmol). Exp. (Calc.) for C₈H₃₂N₆AgBr₇ (603.81 g mol⁻¹): C 10.80 (10.92), H 3.44 (3.67), N 9.21 (9.55), Ag 11.54 (12.26)%. IR (ATR, cm⁻¹): 2963, 1557, 1423, 1169, 1004, 950, 819, 762, 454.

(H₃PMDTA)[BiBr₆] (3)—480 mg (1.00 mmol) of Bi₂O₃ and 420 µL (2.00 mmol) of PMDTA. For this compound, the total volume of concentrated HBr used was 10.0 mL. Yield: 84% (1.460 g, 1.69 mmol). Exp. (Calc.) for C₉H₂₆N₃BiBr₆ (864.73 g mol⁻¹): C 12.49 (12.50), H 2.83 (3.03), N 4.70 (4.86), Bi 21.74 (24.17)%. IR (ATR, cm⁻¹): 3079 and 2648, 2975, 1445, 1380, 1038, 973, 911, 784.

(H₃PMDTA)[Ag₃Br₆] (4)—340 mg (2 mmol) of AgNO₃.xH₂O and 420 µL (2 mmol) of PMDTA. For this compound, the total volume of concentrated HBr used was 5.0 mL. Yield: 17% (0.195 g, 0.340 mmol). Exp. (Calc.) for C₉H₂₆N₃AgBr₄ (603.81 g mol⁻¹): C 10.85 (11.04), H 3.01 (2.67), N 3.94 (4.29), Ag 32.24 (33.04)%. IR (ATR, cm⁻¹): 2975 (νC–H), 2814 and 2718 (νN–H), 1465, 969.

Thin film deposition: FTO-coated glass was washed, submerging it in water containing a few drops of neutral detergent and placing it in an ultrasonic bath for 15 min. The cleaning step was repeated twice with fresh detergent solution. The FTO glass was then submerged in acetone and placed in the ultrasonic bath for 15 min. The process was repeated two more times. Finally, the glass was heated at 350 °C for 2 h. For the deposition of the perovskite film, an adapted procedure was used [27]. A 0.50 mol L⁻¹ solution of Cs₂AgBiBr₆ in DMSO was applied to the FTO glass to fully cover the surface and then spun at 4000 rpm for 35 s and annealed on a hot plate at 200 °C for 30 min. To deposit the target compounds, saturated solutions were prepared as follows: for compounds 1 and 3, solutions of 0.45 mol L⁻¹ and 0.225 mol L⁻¹, respectively, were prepared in DMF; for compound 2, a solution of 0.225 mol L⁻¹ in DMSO/DMF 1:10 v/v; and for compound 4, a solution of 0.225 mol L⁻¹ in DMSO/DMF 1:2 v/v. Then, 20 µL of the desired solution was deposited onto the FTO-coated glass or Cs₂AgBiBr₆ film, spun at 4000 rpm for 35 s, and annealed on a hot plate at 150 °C for 30 min.

Contact angle experiments: Contact angles were determined from profile images of 20 µL deionized water droplets placed on FTO-coated glass, Cs₂AgBiBr₆ films deposited on FTO, and films of compounds 1–4 deposited on both FTO and pristine Cs₂AgBiBr₆.

3. X-Ray Crystallography

X-ray diffraction data collection on single crystals of all compounds were performed with a Rigaku Synergy diffractometer using Mo-Kα radiations (0.71073 Å). Measurements were performed at room temperature as shown in

Table 1. X-ray diffraction data collection and refinement parameters for the compounds 1, 2, 3, and 4.

Compound	1	2	3	4
Formula	C4H16N3Br6Bi	C8H32N6Br7Ag	C9H23N3Br6Bi	C18H52N6Br12Ag6
Moiety Formula	(C4H16N3)[BiBr6]	(C4H16N3)[AgBr4]·(C4H16N3)Br3	(C9H23N3)[BiBr6]	(C9H26N3)[Ag4Br7]·(C9H26N3)[Ag2Br5]
MM/gmol−1	794.64	879.63	861.68	1958.68
T/°C	27(2)	27(2)	27(2)	20(2)
λ/Å	0.71073	0.71073	0.71073	0.71073
Crystal system	Orthorhombic	Monoclinic	Orthorhombic	Orthorhombic
Space group	P212121	P21/c	P212121	Pbcn
a/Å	7.1619(2)	12.8135(2)	12.5209(3)	21.9327(8)
b/Å	14.0315(5)	8.68440(10)	13.2679(3)	7.0174(2)
c/Å	16.2858(5)	21.2648(3)	13.2685(4)	29.1014(11)
α/°	90	90	90	90
β/°	90	94.2400(10)	90	90
γ/°	90	90	90	90
V/Å3	1636.60(9)	2359.82(6)	2204.24(10)	4479.0(3)
Z	4	4	4	4
ρ/Mg m−3	3.225	2.476	2.606	2.910
μ/mm−1	25.398	20.778	18.869	33.410
F(000)	1416	1656.0	1564.0	3616
Reflections collected (Rint)	12,523 (0.0596)	34,634 (0.0365)	19,708 (0.0477)	41,359 (0.0377)
Independent Reflections	4163	5086	4514	4583
Reflection with I >2σ(I)	3498	4789	4210	3609
R a, wR b [I >2σ(I)]	0.0463, 0.1042	0.0361, 0.0956	0.0317, 0.0695	0.0676, 0.1539
R a, wR b (all data)	0.0600, 0.1089	0.0378, 0.0971	0.0354, 0.0705	0.0882, 0.1649
S c	1.044	1.067	1.042	1.112
ρmax and ρmin/e Å−3	1.263, –2.655	0.789; –1.145	1.010, –1.492	2.760, –1.837
CCDC number	2,434,447	2,434,448	2,434,449	2,434,450

^a R= Σ ||F_o| − |F_c||/Σ |F_o|. ^b wR = [Σ w(|F_o|² − |F_c|²)²/Σ w|F_o|²]^1/2. ^c S = [Σ w(|F_o|² − |F_c|²)²/(n_o − np)]^1/2, where w ∝ 1/σ, n_o = observed, and n_p = fitted parameters.

. Data integration and scaling of the reflections for all compounds were performed with the CRYSALIS suite [28]. Final unit cell parameters were based on the fitting of all reflections’ positions. Analytical absorption corrections and space group identification were performed using the CRYSALIS suite [28]. The structures of all compounds were solved by direct methods using the SUPERFLIP program [29]. For each compound, the positions of all atoms could be unambiguously assigned on consecutive-difference Fourier maps. Refinements were performed using SHELXL [30] based on F² through a full-matrix least-square routine. All non-hydrogen atoms were refined with anisotropic atomic displacement parameters. All hydrogen atoms were located in difference maps and included as fixed contributions according to the riding model [31]. The measured values included C–H = 0.97 Å and Uiso(H) = 1.5 Ueq(C) for methyl groups and C–H = 0.97 Å and Uiso(H) = 1.2 Ueq(C) for methylene carbon atoms. For ammonium nitrogen atoms, N–H = 0.89 Å and Uiso(H) = 1.5 Ueq(N).

4. Results

4.1. Comments on Material Synthesis

All compounds were synthesized as single crystals, ensuring high bulk purity, as confirmed by chemical analyses that closely matched the expected values. Compound 1 was previously synthesized and recrystallized [32], but the present manuscript describes a much more straightforward synthesis of it as single crystals, as well as in a much faster way. In terms of stability, thermogravimetric analysis revealed all four synthesized materials exhibited thermal stability up to 180 °C, making them strong candidates for use in perovskite cells that usually need to be annealed. Compound 1 was stable up to 180 °C, at which point two significant weight loss events were observed. At the end of the thermal analysis, the residue was approximately 4.65% of the initial mass. Compound 2 remained stable when heated up to 235 °C. Then, between 235 and 400 °C, a weight loss event occurred, resulting in approximately 65% mass loss, followed by a near plateau up to 600 °C. At this point, the residue was 31.75% of the initial mass. Compounds 3 and 4 exhibited thermal stability up to 195 °C. For compound 3, between 195 and 600 °C, three overlapping weight loss events occurred, leaving a residue of 6.23% of the initial mass. Finally, for compound 4—similarly to compound 2—an initial 32% weight loss occurred between 195 and 400 °C. This was followed by a slight and continuous weight loss up to 600 °C, resulting in a final residue weighing 65.28% of the initial mass.

The residues for 1 and 3 were very low in mass, because bismuth(III) halides can evaporate [33,34,35]. Thus, it is most likely the [BiBr6]³⁻ complex underwent degradation, evaporating the bismuth(III) bromide and leaving H₃DETABr₃ or H₃PMDTABr₃, respectively. Since the thermal analyses were performed in an oxidizing atmosphere, the organic salt was likely degraded (both by heat and oxidation), resulting in a small amount of remaining mass. For the silver complexes, the degradation was incomplete at 600 °C. Compound 2’s residue was compatible with H₃DETA[AgBr₄] (cal. 33.39%), indicating that the entire H₃DETABr₃ portion of the mixed salt underwent degradation. For 4, the incomplete degradation residue could not be associated with any plausible composition.

4.2. Crystal Structure Description

Compound 1’s crystal structure was presented previously [32], being a 0D perovskite based on a bismuth(III)–bromide complex that crystallizes in the orthorhombic P2₁2₁2₁ space group. In this work, the crystal structure will also be deepened and focused in the organic cation. The asymmetric unit is composed of a [BiBr₆]³⁻ unit and a triprotonated diethylenetriamine cation, as pictured in Figure 1a. There are no solvent molecules in the crystal structure. The bismuth complex is in a distorted octahedral coordination geometry (see Figure 1b), with bond lengths varying from 2.787 Å to 2.942 Å and the angles slightly deviating from 90° (see Table S1). The H₃DETA³⁺ cation is found in an almost perfect zigzag conformation, meaning all C–C and C–N bonds have an anti conformation. In the cation, all non-hydrogen atoms are close to coplanarity, with an average deviation from the mean plane of 0.1298 Å.

The crystal packing of 1 is dictated by hydrogen bonds. The N–H bonds are connected to various [BiBr₆]³⁻ units around the cations (see Table S2). Each bismuth complex is found interacting with six other H₃DETA³⁺ units via hydrogen bonds. The N1 atom is found interacting with three [BiBr₆]³⁻ units, the same number as for N2. Two of the bismuth complexes are shared with these nitrogen atoms. On the other hand, the N3 atom has four [BiBr₆]³⁻ around it, with only one interacting at the same time with N2 and none with N1. As a result, when viewed along the crystallographic a axis, the complexes and cations are equally alternated in the compound packing (see Figure 2).

Compound 2 is also a 0D perovskite, being a mixed salt, that crystallizes in the monoclinic P2₁/c space group. The asymmetric unit consists of one [AgBr₄]³⁻ anion, three bromide anions, and two H₃DETA³⁺ cations (see Figure 3a). That means the co-crystallization of H₃DETA[AgBr₄] and H₃DETABr₃ is essential to form a densely packed structure and to minimize the lattice energy. Additionally, no crystallization solvent molecules were found in the compound. The silver(I) ion is found in a slightly distorted tetrahedral coordination geometry, as highlighted in Figure 3b. The Ag–Br bond lengths ranged from 2.679 Å to 2.867Å and the angles varied from 101.987(18)° to 117.91(2)° (see Table S1). Differently from 1, the H₃DETA³⁺ cation in 2 has all C–C bonds in gauche conformation. The torsion angles for the N–C–C–N set of atoms are 83.30° [N1–C1–C2–N2], 81.92° [N2–C3–C4–N3], 73.56° [N4–C5–C6–N5], and 78.08° [N5–C7–C8–N3].

The crystal packing of 2 is also dictated by hydrogen bonds. The H₃DETA³⁺ cations can interact with both [AgBr₄]³⁻ anions and the bromide anions in all three directions (see Table S3). As a result of the hydrogen bonds, a layered pattern is observed along the crystallographic c axis. This sequence alternates between a layer of bromide anions, followed by a double layer of H₃DETA³⁺, with the [AgBr₄]³⁻ anions mediating the interaction with the cations. The crystal packing of 2 is shown in Figure 4.

Compound 3, similarly to compound 1, is also a 0D perovskite based on a bismuth(III)–bromide complex that crystallizes in the orthorhombic P2₁2₁2₁ space group. The asymmetric unit of 3 is, as in compound 1, composed of one [BiBr₆]³⁻ unit and one triprotonated amine cation, in this case, H₃PMDTA³⁺, as pictured in Figure 5a. The bismuth complex is in a distorted octahedral coordination geometry (Figure 5b), with bond lengths varying from 2.787 Å to 2.942 Å and the angles slightly deviating from 90° (see Table S1). The H₃PMDTA³⁺ in this compound has one of its C–C bonds found in an anti conformation [N2–C3–C4–N3 torsion angle of 9.91°] and one in gauche [N1–C1–C2–N2 torsion angle of 67.49°]. On the other hand, all C–N bonds are in an anti conformation. The torsion angles around C–N bonds are 5.99° [C5–N1–C1–C2], 63.70° [C6–N1–C1–C2], 1.50° [C1–C2–N2–C7], 67.73° [C7–N2–C3–C4], 66.25° [C3–C4–N3–C8], and 8.77° [C3–C4–N3–C9].

The crystal packing of 3 is primarily governed by classic hydrogen bonds (see Table S4). Each H₃PMDTA³⁺ cation interacts via N–H···Br hydrogen bonds to two [BiBr₆]³⁻ units, as seen in Figure 6. It forms a 1D supramolecular structure, parallel to the crystallographic a axis. The 1D supramolecular polymers interact with the neighboring chains via C–H···Br weak hydrogen bonds, involving both CH₃ and CH₂ groups found in H₃PMDTA³⁺, leading to supramolecular interactions in the other directions, granting packing in all directions.

Compound 4 crystallizes in the centrosymmetric orthorhombic Pbcn space group, with its asymmetric unit depicted in Figure 7a. It is a mixed 1D/2D perovskite built from silver(I) and bromide ions. To achieve charge balance, H₃PMDTA³⁺ cations are present. In the 1D coordination polymer, the silver(I) atom adopts a distorted tetrahedral coordination geometry (see Figure 7b and Table S1) with three of its bromide atoms acting as bridging ligands, as seen in Figure 7d. The bromide atom Br2 connects two silver(I) tetrahedra, forming a linear bridge [Ag1–Br2–Ag1ⁱⁱ = 180°, ii = 3/2–x, –1/2+y, z]. The bromide Br3 creates double bridging between the silver(I) atoms, resulting in an edge-sharing tetrahedral motif, presenting a narrow bond angle of 71.32(2)° [Ag1–Br3–Ag1ⁱ, i = 1–x, 2–y, 1–z]. Similarly, the 2D coordination polymer has all silver(I) atoms in a tetrahedral coordination geometry (see Figure 7c,e). One of the bromides, Br5, is found in an almost linear bridging arrangement [155.39(9)° for Ag2–Br5–Ag3], while Br6 and Br7 form a double bridge with a narrow bond angle [Ag2–Br6–Ag3 = 65.58(4)° and Ag2–Br7–Ag3 = 67.41(4)°], and Br4 connects three silver(I) atoms in a T-shaped geometry (see Table S1). The combination of these bonds results in a 2D structure with a 10-membered Ag₆Br₄ macrocyclic ring, where H₃PMDTA³⁺ is found filling the ring void. The positioning of the H₃PMDTA³⁺ cations, occupying the voids of the 2D polymer and the spaces between the 1D chains, is shown in Figure 8a. Similarly to compound 3, the cation has a C–C bond in anti conformation and the other one in gauche, and both C–N bonds are in gauche conformation. The torsion angles for C–C bonds are 3.90° [N1–C1–C2–N2] and 59.77° [N2–C3–C4–N3]. C–N torsion angles are 68.81° [C5–N1–C1–C2], 25.75° [C6–N1–C1–C2], 67.38° [C7–N2–C1–C2], 8.41° [C7–N2–C3–C4], 59.47° [N2–C3–C4–N3], 64.45° [C8–N3–C4–C3], and 70.62° [C8–N3–C4–C3]. This cation interacts via N–H···Br hydrogen bonds with both the 1D and 2D coordination polymers, which bind the components into a continuous three-dimensional supramolecular structure (see all hydrogen bonds’ geometry for this compound in Table S5). In this supramolecular motif, each unit is packed, forming an alternated pattern along the crystallographic c axis. The pattern is composed by a 2D polymer, followed by a H₃PMDTA³⁺ layer, then a layer of 1D polymers, and finally another H₃PMDTA³⁺ layer, as shown in Figure 8b.

4.3. Film Deposition and Water Affinity

All compounds were successfully deposited on FTO, forming colorless films, which indicates their deposition will not interfere in the light absorption of the resulting device. The pristine FTO-coated glass exhibited a contact angle of 22°, indicating strong hydrophilicity. Upon the deposition of films of 1, 2, 3, and 4, the contact angle increased to 44°, 36°, 62°, and 54°, respectively. Contact angle images are shown in Figure 9. Among the H₃DETA³⁺-based compounds, compound 1 shows a higher contact angle than 2, most likely due to the presence of additional bromide anions in the crystal structure, which enhances its interaction with water.

The H₃PMDTA³⁺-based compounds show higher contact angle than the H₃DETA³⁺-based ones. This is attributed to the presence of five methyl groups on the nitrogen atoms, which contribute to increased hydrophobicity. In this respect, compound 3 is more hydrophobic than compound 4. In the crystal packing of 3, the [BiBr₆]³⁻ units are surrounded by the organic cations, whereas in 4, he layered supramolecular structure traps the organic moieties within the voids. Thus, the organic moiety in 4 is less available to interact with water.

Experiments were repeated on Cs₂AgBiBr₆ films deposited on FTO-coated glass, as shown in Figure 10. The pristine perovskite showed a water contact angle of 40°. Compounds 1 and 3 were successfully deposited onto Cs₂AgBiBr₆ films (see insets in Figure 10), and the contact angles were approximately the same as those observed when deposited directly onto FTO-coated glass. This suggests the presence of a thin film of these compounds over the perovskite layer. Silver-based compounds 2 and 4 required a DMSO/DMF solvent mixture for dissolution. However, this solvent combination dissolved the underlying perovskite layer. Thus, it was not possible to analyze their hydrophobicity on the perovskite film so far.

5. Conclusions

This work presented four bismuth(III) and silver(I) low-dimensional perovskites, three of them being novel compounds. Their syntheses were optimized to obtain single crystals, with particularly good yields for the bismuth-based compounds. All compounds exhibited high thermal stability, which is required for the annealing processes typically applied to perovskite films, being stable up to 180 °C. These compounds were analyzed by single crystal X-ray diffraction. Compounds 1, 2, and 3 are 0D perovskites containing isolated [BiBr₆]³⁻ octahedra or [AgBr₄]⁻ tetrahedra, with H₃DETA³⁺ (1 and 2) and H₃PMDTA³⁺ (3) as the cations. Furthermore, it was observed that compound 2 is a mixed salt, as its asymmetric unit presented two H₃DETA³⁺ cations, with the second forming part of the salt H₃DETABr₃. Compound 4 is a mixed 1D/2D perovskite in which the silver(I) forms two distinct coordination polymers, both composed of corner- and edge-sharing tetrahedra. All four compounds were deposited onto FTO-coated glass as films, increasing the hydrophobicity of the surface. However, only compounds 1 and 3 could be deposited onto the Cs₂AgBiBr₆ perovskite film. The film of compound 1 on the perovskite maintained the same contact angle of the pristine perovskite. In contrast, the film of compound 3 made the perovskite surface more hydrophobic, presenting the contact angle slightly higher than when it was deposited directly on FTO-coated glass. These findings contribute to overcoming fabrication challenges associated with the lead-free perovskites that are unstable in humid environments, thereby improving their stability for photovoltaic applications. It is particularly relevant for the promising Cs₂AgBiI₆ perovskite, whose thin films must be fabricated via ion exchange from bromide-based perovskite and which is severely unstable under ambient humidity.