Raúl I.
Sánchez-Alarcón
a,
Omar E.
Solís
a,
María Cristina
Momblona-Rincon
a,
Teresa S.
Ripolles
*a,
Juan P.
Martínez-Pastor
a,
Rafael
Abargues
*a and
Pablo P.
Boix
*ab
aInstituto de Ciencia de los Materiales-Universidad de Valencia, Catedrático José Beltrán, 2, 46071, Valencia, Spain. E-mail: teresa.ripolles@uv.es; rafael.abargues@uv.es; Pablo.P.Boix@itq.upv.es
bInstituto de Tecnología Química, Instituto de Tecnología Química, Universitat Politècnica València-Consejo Superior de Investigaciones Científicas, Av. dels Tarongers, 46022 València, Spain
First published on 6th May 2024
The lattice arrangement of lead-free perovskites can be altered by synthesis parameters such as concentration, temperature, choice of solvents, and the length of monovalent cations. This effect, often described as dimensionality modulation, can be exploited to develop new materials with better optical properties than their three-dimensional counterpart. This work explores the versatility of the hot-injection synthesis for obtaining 2-thiopheneethylammonium (TEA+) based tin halide perovskite microcrystals. The dimensionality of TEA-based tin bromide perovskite microcrystals can be modulated through changes in the TEA concentration, giving rise to the formation of highly luminescent 0D-TEA4SnBr6 and low-emissive 2D-TEA2SnBr4 microplates. In contrast, 0D-[TEASnCl3][TEACl] and 2D-TEA2SnI4 are thermodynamically preferred as main products when using the chloride and iodide analogues, respectively, limiting the dimensional versatility. It is noteworthy that 0D-TEA4SnBr6 shows the highest PLQY from the low dimensional TEA tin halide family (PLQY = 75% in thin-film) owing to a more confined configuration based on isolated [SnBr6]4− octahedra moieties separated by TEA+ cations, being attractive for optoelectronic applications.
Reducing the dimensionality of perovskite materials by disaggregation of inorganic moieties with bulky organic cations is a promising approach to mitigate the Sn-based perovskite stability issues.5 For example, in two dimensional perovskites, 2D, layered perovskite moieties are hydrogen bonded with voluminous organic spacers, which are self-assembled by van der Waals forces and π-stacking. If layers of 3D perovskites are stuck within the voluminous organic spacers, the physical properties depend on the thickness of inorganic perovskite moieties, being possible to tune the exciton binding energy and band gap of the resulting materials.6 On the other hand, if the density of bulky organic cations is high enough to separate the structure on individual polyhedron moieties, these conduce to the formation of zero-dimensional perovskite structures, 0D.7
This bulky organic layer can act as a protection from external agents, increasing environmental stability. It also leads to a higher confinement of bounded excitons than their three-dimensional material counterpart,8 which can be advantageous for light emitting applications.9 From the fundamental point of view, manipulating the dimensionality of tin-based perovskites can also impact their bandgap, carrier mobility, and other electronic properties due to the spatial confinement of charge carriers. For example, 2D tin halide perovskites present radiative recombination of free excitons by increasing the exciton 2D binding energy.10,11 On the other hand, 0D tin halide perovskites exhibit radiative recombination of self-trapped excitons (STEs) at energy levels originating from the optical distortion of Sn–X bonds.12,13 In particular, STEs refer to the strong electron (exciton) phonon coupling when a soft lattice is distorted upon photoexcitation.14 Interestingly, 0D-tin halide perovskites show higher PLQYs, better environmental stability, higher Stokes shift, and a major exciton binding energy compared to their 2D and 3D counterparts.13,15,16
However, inorganic and hybrid tin halide compounds can adopt multiple dimensionalities depending strongly on the polyhedron type (owing to the high coordination number of Sn(II) ions), inorganic or organic cations, oxidation number, and preparation conditions (pH, temperature, atmosphere, etc).17,18 For example, zero-dimensional tetrabutylammonium (TBAC) tin(II) chloride could be obtained under acidic conditions with stable pyramidal [SnCl3]− anions.19 If the monovalent cation is changed to Cs+, 3D-CsSnCl3 with [SnCl6]4− octahedrons is stable at high temperature because 5s2 orbitals hybridize with 5pz, resulting in hybrid orbitals with the same energy as that of np halogen orbitals.18,20 It should be mentioned that the octahedral tin chloride configuration is also stable for Sn(IV) ions.21 Different inorganic and organic tin(II/IV) halide compounds and dimensionalities are summarized in Table 1.
Cation | Compound | Polyhedral unit | Dimensionality | Ref. |
---|---|---|---|---|
List of acronyms. TBCA+ = tetrabutylammonium; R/S-α-PEA+ = R/S-α-phenethylammonium; MA+ = methylammonium; FA+ = formamidinium; H4BAPP = 1,4-bis(3ammoniumpropyl)piperazinium; C7H8N3+ = 2-amino benzimidazolium; Cs+ = cesium; DAO2+ = 1,8-octyldiammonium; IPA+: isopropylammonium; (CH3)3S+ = trimethyl sulfonium. | ||||
TBAC+ | (TBAC)SnCl3 | [SnCl3]− | 0D | 19 |
R/S-α-PEA+ | R/S-α-PEA SnX3, (X = Cl, Br) | Distorted [SnX6] | 1D | 22 |
MA+, FA+ | ASnI3 (A = MA+, FA+) | [SnI6]4− | 3D | 23 |
PEA+ | PEA2SnI4 | [SnI6]4− | 2D | 24 |
H4BAPP3+ | H4BAPPSnX5 (X = Cl, Br) | [SnX5]3− | 0D | 25 |
C7H8N3+ | (C7H8N3)4SnBr6 | [SnBr6]4− | 0D | 26 |
Cs+ | CsSn2Br5 | [Sn2Br5]− | 1D | 27 |
DAO2+ | DAOSn2I6 | [Sn2I6]2− | 1D | 28 |
IPA+ | μ-[IPA]3Sn2I7 | [Sn2I7]3− | 2D | 29 |
(CH3)3S+ | [(CH3)3S]2SnCl6·H2O | [SnCl6]2− | 0D | 21,27 |
Cs+, MA+, FA+ | A2SnI6 | [SnI6]2− | Undefined | 30 |
Thus, lead-free low-dimensional tin(II) halide perovskites are attracting scientific attention owing to their unique combination of interesting physical properties and lower toxicity.31 They show great potential as visible light emitters on optoelectronic devices32 as well as for enhancing the performance of photovoltaic devices.33 Among multiple organic cations tested on low-dimensional hybrid metal halide perovskites, 2-thiopheneethylammonium (TEA+) has drawn attention as a heteroatomic organic spacer and a passivating agent for lead iodide perovskites, with a stable power conversion efficiency (PCE) of 18.75% for 1000 h34,35 when used in solar cells. Concerning tin-based perovskites, the synthesis of 2D-TEA2SnI4 has been reported at different scales, from nanocrystals36 and thin films37 to single crystals,38 showing a PLQY around 19% for nanodisks in solution prepared by the LARP method36 and 23% for thin films prepared using the antisolvent approach.37 However, to the best of our knowledge, the synthesis and characterization of TEA-based Sn(II) chloride and bromide remain unexplored.
The dimensionality of perovskite nanocrystals synthesized by the hot injection method can be modulated by means of changes in the precursor concentrations.39–41 Dimensional modulation improves the optical properties of lead-free perovskites and their PLQY owing to the separation and distortion of metal halide octahedrons with lone electron pair ns2.42 For instance, Cs-oleate solution precursor concentration is a cornerstone parameter that determines whether 0D-Cs4SnBr6 or 3D-CsSnBr3 nanocrystals are obtained.43,44 Appropriate dimensionality manipulation of lead-free perovskites expands the range of these less toxic materials with a full gamut of wavelength emission depending on the type of inorganic moiety and organic cation. Therefore, the hot injection approach could be re-imagined as a controlled method to achieve high luminescent perovskite nanostructures.39,43
Here, we report the synthesis of 2D and 0D TEA-based Sn(II) halide perovskite microplates by the hot injection method. The control of the [TEA]/[Sn2+] ratio for the bromide-based perovskite adjusts the resulting perovskite dimensionality from 2D to 0D. The PLQY was strongly affected by this factor, ranging from 0.1% for 2D-TEA2SnBr4 to 75% for orange emitting 0D-TEA4SnBr6. Interestingly, TEA/Sn(II)-driven dimensionality control only takes place with TEA-based tin(II) bromide derivatives. Conversely, the reaction between SnCl2 and SnI2 with TEA provides main products pure 0D-[TEASnCl3][TEACl] and 2D-TEA2SnI4 microplates, respectively.
1H and 13C nuclear magnetic resonance (NMR) spectra in the liquid phase for 0D-TEA4SnBr6 and 2D-TEA2SnX4 (X = Br, I) were recorded with a Bruker AV400 spectrometer, with an Oxford Magnet of 400 MHz (9.4 T). 1H and 13C NMR spectra are reported in ppm (δ) relative to the chemical shift of solvent residual signals for DMSO-d6 at 2.51 ppm and 39.99 ppm, respectively.
13C and 119Sn magic angle solid state nuclear magnetic resonance (MAS-NMR) spectra of 0D-[TEASnCl3][TEACl], 0D-TEA4SnBr6, and 2D-TEA2SnX4 (X = Br, I) were measured with a Bruker Advance III 400 WB spectrometer, with an ultra-shielded Widemouth magnet (89 mm), 400 MHz (9.4T), two RF channels and high-power amplifiers for solid-state experiments. For 119Sn MAS-NMR spectra measurements, the SnO powder sample was used as a reference. On the other hand, for 13C and 119Sn MAS-NMR and liquid state NMR, perovskite microcrystals were isolated from n-octane suspension by centrifugation and dried under an inert atmosphere overnight. 119Sn MAS-NMR spectra of 0D-[TEASnCl3][TEACl] were measured at different spinning rates (8 kHz and 10 kHz).
Transmission electron microscopy (TEM) images were acquired from a Hitachi HT7800 microscope with a high-resolution filament of LaB6 with an acceleration voltage of 100 kV. TEA-based tin(II) halide microcrystal suspensions in n-octane were dropped on Cu/C TEM grids.
Energy-dispersive X-ray spectroscopy (EDS) microanalysis was performed using an Oxford Ultim Max 170 detector monitored with Aztec software coupled to a SCIOS2 FESEM microscope from Thermo Scientific. For these measurements, TEA-based tin(II) halide microcrystals were deposited on Si substrates, and sample borders were coated with Ag paste.
Photoluminescence (PL) and photoluminescence excitation (PLE) measurements were conducted in an FLS1000 spectrophotometer from Edinburgh Instruments. A xenon lamp (Xe2) was used as an excitation source. PL spectra were recorded with a PMT-980 detector from Edinburgh Instruments. PL and PLE spectra were measured with a dwell time of 1 s and a wavelength step of 1 nm. The PL quantum yield (PLQY) was measured with an integrating sphere coupled to an FLS1000 spectrophotometer from Edinburgh Instruments. TEA based tin(II) halide microcrystal films were prepared over 0.25 cm2 quartz substrates by drop cast deposition to measure PLE and PL spectra and PLQY.
Fig. 1 Top-view SEM (a) and (d), cross-sectional SEM (b) and (e) and TEM images (c) and (f) of 2D-TEA2SnBr4 (a)–(c) and 0D-TEA4SnBr6 (d)–(f) microcrystals. |
Fig. 2a shows the XRD diffractograms of TEA-based Sn(II) bromide perovskite films for different [TEA]/[Sn2+] ratios. From [TEA]/[Sn2+] < 0.5, the XRD diffractogram shows the multiple reflections peaks at 2θ of 5.62°, 11.32°, 17.02°, 22.78°, 28.64° and 34.56°. These reflections are spaced apart by a 2θ interval of 5.7°, as expected from a Ruddlesden–Popper phase for pure 2D TEA2SnBr4.36 Additionally, EDS analysis (Table S1, ESI†) reveals that Br/Sn and S/Sn atomic ratios are in good agreement with the expected stoichiometry. As the [TEA]/[Sn2+] ratio increases, isolated metal halide octahedra [SnBr6]4− or clusters are generated. From [TEA]/[Sn2+] > 0.5, new non-periodically XRD peaks are observed at 8.52°, 17.13°, 18.23° and 25.84°, which are attributed to 0D-TEA4SnBr6. Atomic percentages of Br are higher than those measured for 2D-TEA2SnBr4 microplates (Table S1, ESI†), which supports the proposed chemical formula.
Concurrently, the increase of TEA molarities results in a decrease in the formation of the 2D-TEA2SnBr4 phase. From [TEA]/[Sn2+] > 2, the diffractograms reveal only reflection peaks of 0D-TEA4SnBr6 along with those of pure TEABr at 2θ values of 4.5°, 9°, 14°, 19°, 24°, 29° and 34°. As expected, 0D-TEA4SnBr6 reflection peaks do not exhibit the same periodic ordering observed in the 2D-TEA2SnBr4.
The differences between 2D-TEA2SnBr4 and 0D-TEA4SnBr6 crystal phases and band structures also generate remarkable differences in their optical properties. PL emission and PL excitation (PLE) spectra of TEA–Sn(II) bromide films synthesized at different molar ratios are shown in Fig. S1 (ESI†). In particular, the [TEA]/[Sn2+] ratio of 0.5, for 2D-TEA2SnBr4 films, is represented in Fig. 2b. Its PL spectra can be deconvoluted into two bands; one sharp band at 470 nm, and another broader band centered at 730 nm. The shape of the PL emission bands is similar to that reported for 2D-PEA2PbI4,45 thus mainly determined by the radiative recombination of free excitons. Additionally, the PLE spectrum exhibits a sharp band at 470 nm. From the spectral difference in the position of the band maximum of the PL and PLE spectra, we measured the Stokes shift energy, which is around 113 meV. The sharp PL and PLE bands shown in Fig. 2b, along with the high order observed in XRD, may be associated with the synthesis of lamellar-like microcrystals. These structures consist of intercalated layers comprising [SnBr6]4− octahedra and organic layers of TEA+ cations, as reported for 2D-TEA2SnI4 nanocrystals and thin films.46 In contrast, the PL emission and PLE spectra of 0D-TEA4SnBr6 films synthesized with a [TEA]/[Sn2+] ratio of 3 are displayed in Fig. 2c. When the perovskite was exposed to UV light (wavelength of 282 nm), 0D-TEA4SnBr6 exhibited a PL spectrum characterized by a band centered at 610 nm and an FWHM of 130 nm. The PLE spectrum shows two sharp bands with a maximum at 282 nm and 310 nm. From PL and PLE spectra, we measured a strong Stokes shift of 2.38 eV, which is attributed to the radiative recombination of self-trapped excitons.47,48
The larger bandgaps and Stokes shifts when transitioning from the 2D to 0D TEA-based Sn bromide perovskite are attributed to a reduction of the orbital overlap between [SnX6]4− octahedra as the 0D phase becomes more predominant with increasing [TEA]/[Sn2+] ratio. This occurs because TEA+ cations separate the [SnBr6]4− octahedra of 2D layers to form isolated [SnBr6]4− octahedra. This results in a higher exciton binding energy, lower carrier mobility, and increased stability. The isolated structure of 0D perovskites also imparts a strong quantum confinement effect, contributing to a higher PL emission compared to that of higher dimensionality.
Fig. 2d shows the 0D/2D phase ratio, PLQY, and Stokes shift of TEA-based Sn(II) bromide perovskite films depending on the [TEA]/[Sn2+] molar ratio. To estimate the phase ratio between 2D and 0D materials, we used eqn (1):
(1) |
Based on our experimental results, the growth of 0D-TEA4SnBr6 is strongly influenced by the isolation of [SnBr6]4− octahedra layers from the 2D-TEA2SnBr4 material onto individual clusters resulting from the increase of TEA+ concentration. Also, the growth of 0D-TEA4SnBr6 is not influenced by coupling with mesitylene or oleic acid molecules as it is the case of the phenethylammonium tin(II) bromide perovskite, and thus the separation of inorganic moieties is influenced mostly by the increment in the TEA+ concentration.49 As a consequence, highly luminescent 0D-TEA4SnBr6 microcrystals can be synthesized by controlling the concentration of TEA.50
A qualitative magic angle solid state (MAS) NMR analysis was carried out to investigate the local environment and chemical bonding of carbon (13C) and tin (119Sn) nuclei in solid-state materials (Fig. 3). 2D-TEA2SnBr4 and 0D-TEA4SnBr6 are compared with TEABr to determine the internal structure of low dimensional TEA-based Sn(II) bromide perovskites.
Fig. 3 (a) 13C MAS-NMR spectra of TEABr, 0D-TEA4SnBr6, and 2D-TEA2SnBr4 powders. Crystalline structures and 119Sn NMR of (b) and (c) 2D-TEA2SnBr4 and (d) and (e) 0D-TEA4SnBr6. |
The 13C NMR spectra of 2D-TEA2SnBr4, 0D-TEA4SnBr6, and TEABr are shown in Fig. 3a. A set of four signals, ranging from 145 to 120 ppm, has been ascribed to the aromatic carbons of TEA+. J1, J2, and J3 signals observed in 2D-TEA2SnBr4 are shifted to a higher field (ΔJ = 1.21 ppm) than those of 0D-TEA4SnBr6 (ΔJ = 0.41 ppm) in comparison with TEABr. A characteristic signature of π-stacking interactions often manifests in changes to signal intensity and peak broadening in NMR spectra. The π-stacking interactions among aromatic rings lead to dynamic processes, causing an exchange between stacked and unstacked conformations. This equilibrium results in broader peaks in the NMR spectra due to variations in the chemical environments experienced by the interacting nuclei.
At a low [TEA]/[Sn2+] ratio, TEA+ can be distributed between layers of [SnBr6]4− octahedra, leading to the formation of a 2D tin bromide perovskite. In this structure, [SnBr6]4− octahedra are interconnected through four corner-sharing, creating 2D perovskite monolayers. Under these conditions, we can assume that thiophene rings self-assemble by π–π stacking interaction between thiophenes of adjacent 2D perovskite monolayers in 2D-TEA2SnBr4.51 This combination of structural features contributes to the unique properties and characteristics of the resulting material.52 Conversely, when the [TEA]/[Sn2+] ratio is high, the TEA+ cations form a surrounding layer around individual [SnBr6]4− octahedra, resulting in a lack of interconnection between the octahedra. The high [TEA]/[Sn2+] ratio influences the structural organization, and the isolation between [SnBr6]4− octahedra enables their strong quantum confinement effect, which has implications for the properties of the material, such as radiative recombination by self-trapped excitons.44 The TEA+ cations randomly distributed around the [SnBr6]4− octahedra reduce the π stacking interaction. As a result, the chemical shifts of aromatic carbons observed in 0D-TEA4SnBr6 are very similar to those of TEABr.53
Regarding the 119Sn MAS-NMR spectra, 2D-TEA2SnBr4 (Fig. 3d and Fig. S2, ESI†) shows two different signals at −465 ppm and −1983 ppm assigned to Sn(II) and Sn(IV) respectively,20 with a lower intensity for the latter. In the case of 0D-TEA4SnBr6, no significant signal was observed between −1000 and −2000 ppm. This is fully aligned with the better ambient stability displayed by the 0D-TEA4SnBr6 compared to its bidimensional counterpart.
Interestingly, the 119Sn MAS-NMR spectrum of TEA-based Sn(II) chloride microcrystals (Fig. 4b) revealed a series of peaks between −100 ppm and −800 ppm related to the chemical shift anisotropy (CSA) effect (δiso = −215 ppm) of Sn(II). This can be attributed to the formation of [SnCl3]− trigonal pyramids instead of [SnCl6]4− octahedra.55 To complement 119Sn MAS-NMR measurements, EDS analysis shows that Br/Sn and S/Sn atomic ratios are 4.4 and 2.5, respectively. This may initially suggest a stoichiometry similar to the 2D counterpart. However, the CSA effect depicted in Fig. 4b indicates the presence of a low symmetry site for Sn(II). Therefore, we propose a chemical formula of 0D-[TEASnCl3] [TEACl]. A similar stoichiometry has been reported for different 0D Sn(II) chloride perovskites (A = Cs+, MA+, FA+, (CH3)3S+, C6H22N4Cl3+, and TBA+).19,20,56–58 Some authors have reported the formation of 0D-Cs2SnCl6 with high symmetry [SnCl6]2− octahedra, which showed a single signal on 119Sn NMR around −708 ppm.20,56,59–61 Additionally, other reports suggest a possible phase transition between [SnCl3]− trigonal pyramids to [SnCl6]4− octahedra for Cs-based Sn(II) chloride, with a characteristic NMR signal at −561 ppm observed after annealing the sample above 500 °C under vacuum.56,62 In our case, distinguishing between [SnCl6]4− (from phase transition) and [SnCl6]2− (produced by oxidation of Sn(II) ions) is not directly accessible due to the CSA effect observed in Fig. 4b. Moreover, the signals observed in Fig. S4 (ESI†) change in intensity at different spinning rates in NMR experiments, yet the chemical shifts are the same. As a consequence, we can conclude that TEA-based Sn(II) chloride microcrystals are synthesized as a 0D material with well-defined [SnCl3]− pyramidal anions. EDS analysis (Table S1, ESI†) is in good agreement with the proposed stoichiometry, which confirms that Sn is mainly present as Sn(II).
13C MAS-NMR spectra for 0D-[TEASnCl3][TEACl] (Fig. S5, ESI†) do not show any diminution in the signal's intensity, such as 0D-TEA4SnBr6 compounds, so π-stacking is negligible.51 In Fig. S3 and Table S3 (ESI†), multiple displacements are observed in the PL wavelength of the 0D-[TEASnCl3][TEACl] microcrystal, between 630 nm for [TEA]/[Sn2+] = 1 and 595 nm for [TEA]/[Sn2+] = 2. This displacement may be due to tilting of the [SnCl3]− tetrahedra induced by the increase in TEA+ concentration.56 The increase in the PLQY of 0D-TEA-based tin(II) chloride synthesized at a [TEA]/[Sn2+] ratio between 2 and 4, and the reflections related to TEACl could suggest the formation of [TEASnCl3][TEACl]X (where X is the number of TEACl molecules associated with the principal structure TEASnCl3).55
In the synthesis of TEA-based Sn(II) iodide, only pure 2D-TEA2SnI4 microcrystals were predominantly obtained at all the explored molar ratios (Fig. S7, ESI†), without any evidence of dimensionality change. The XRD diffraction pattern shows multiple peaks at 5.53°, 11.23°, 16.98°, 22.73°, and 28.53° and separated by a period of ∼5.7° (Fig. S7, ESI†), which indicates the formation of the Ruddlesden–Popper phase for 2D-TEA2SnI4.46 These could be confirmed with EDS analysis shown in Table S1 (ESI†). The absence of peaks was observed in the XRD patterns of 0D TEA-based tin(II) iodide. PL and PLE spectra (Fig. S7 and Table S3, ESI†) of TEA-based Sn(II) iodide display similar results for all the [TEA]/[Sn2+] ratios tested. PLE and PL spectra are composed of a narrow band at 620 nm and 660 nm, respectively, and a small Stokes shift of around 80 meV, which is characteristic of radiative recombination of free excitons. The PLQY for 2D-TEA2SnI4 microcrystals (Table S3, ESI†) obtained in this work is relatively low (PLQY = 3%) compared with the record value reported by other authors (18.85% for nanodisk solution in n-hexane46 and 23% in thin films63). This suggested the formation of multiple native superficial defects such as Sn(II) vacancies that affect the emissive properties of 2D-TEA2SnI4 microcrystals, as observed on different metal–organic and inorganic tin(II) halide nanocrystals synthesized by the hot injection method.63,64
A similar dimensional transition was not observable for 0D-[TEASnCl3][TEACl] and 2D-TEA2SnI4. It is worth noting that 0D-[TEASnCl3][TEACl] showed good optical properties with a maximum PLQY of 25%. The PLQY of 0D-TEA4SnBr6 is higher than those of 0D-[TEASnCl3][TEACl] and 2D-TEA2SnI4 because of the confinement effects and separation between inorganic moieties. Although the superior PLQYs for 0D-TEA4SnBr6 and 0D-[TEASnCl3][TEACl] are related to the recombination of self-trapped excitons, the first of them exhibited a higher stable PL emission compared to its chloride analogue. This dimensional modulation can be a useful approach to obtain and design high luminescent lead-free materials with interesting properties for practical applications.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tc00623b |
This journal is © The Royal Society of Chemistry 2024 |