Efficient Near-Infrared Luminescence Based on Double Perovskite Cs₂SnCl₆

Abstract

Cs₂SnCl₆ double perovskite has attracted wide attention as a promising optoelectronic material because of its better stability and lower toxicity than its lead counterparts. However, pure Cs₂SnCl₆ demonstrates quite poor optical properties, which usually calls for active element doping to realize efficient luminescence. Herein, a facile co-precipitation method was used to synthesize Te⁴⁺ and Er³⁺-co-doped Cs₂SnCl₆ microcrystals. The prepared microcrystals were polyhedral, with a size distribution around 1–3 μm. Highly efficient NIR emissions at 1540 nm and 1562 nm due to Er³⁺ were achieved in doped Cs₂SnCl₆ compounds for the first time. Moreover, the visible luminescence lifetimes of Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ decreased with the increase in the Er³⁺ concentration due to the increasing energy transfer efficiency. The strong and multi-wavelength NIR luminescence of Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ originates from the 4f→4f transition of Er³⁺, which was sensitized by the spin-orbital allowed ¹S₀→³P₁ transition of Te⁴⁺ through a self-trapped exciton (STE) state. The findings suggest that ns²-metal and lanthanide ion co-doping is a promising method to extend the emission range of Cs₂SnCl₆ materials to the NIR region.

1. Introduction

The Pb-halide perovskite photovoltaics has been seen in the rapid rise of power conversion efficiency (PCE) in the past several years, expectedly contributing to sustainable development via green energy strategies [1,2]. In addition, Pb-halide perovskites have attracted increasing attention due to their outstanding photoelectric properties, such as their tunable band gap and high photoluminescence quantum yield (PLQY) [3,4,5]. However, there are growing concerns about the lead toxicity to human health and the environment [6,7]. Therefore, various efforts have been made to replace Pb with non-toxic metals, such as tin (Sn), germanium (Ge), bismuth (Bi), and indium (In) [8,9,10,11]. Among those alternative elements, Sn is deemed as a perfect choice because it has the most similar electronic properties in the same group of the periodic table with lead. As expected, tin halide perovskites (CsSnX₃) can also offer an outstanding optoelectronic performance including their narrow band gap, low exciton binding energy, and long carrier diffusion length [12,13,14]. Unfortunately, CsSnX₃ perovskites suffer from rather poor stability under ambient conditions due to the easy oxidation of Sn²⁺ to Sn⁴⁺ [15,16].

In such a context, Sn⁴⁺-based perovskite variants (Cs₂SnX₆) are much more stable than CsSnX₃ perovskites [14]. However, Cs₂SnX₆ exhibits significantly inferior optical properties in the visible and near-infrared (NIR) regions compared to Pb-halide perovskites [17,18]. It is reported that the 6s² electrons of Pb²⁺ play a major role in avoiding the formation of deep trap defects, resulting in highly efficient optoelectronic processes [19,20]. Whereas, the ns² electronic configuration of Sn⁴⁺ is lost in Cs₂SnX₆. Therefore, doping elements with ns² electrons is a very effective method to improve their optoelectronic performance [21]. For example, Bi³⁺-, Sb³⁺-, and Te⁴⁺-doped Cs₂SnX₆ have produced intense blue, red, and yellow emissions, respectively [22,23,24]. In addition, white light emission was obtained from Cs₂SnX₆ that was co-doped with Bi³⁺ and Te⁴⁺ ions [25,26]. Up until now, Cs₂SnX₆ luminescence has almost covered the whole visible region through ion doping. However, to the best of our knowledge, there are few reports of achieving luminescence in the NIR region in Sn⁴⁺-based perovskites, especially for Cs₂SnCl₆ with a wide band gap. While NIR is of significant importance in many applications, including night vision, thermal imaging, bioimaging, and wellness monitoring [27]. Therefore, it is critically demanding to achieve efficient NIR-emitting perovskite derivatives.

To achieve NIR luminescence in Cs₂SnCl₆, lanthanide (Ln³⁺) ions with proper emissions such as Er³⁺, Yb³⁺, and Nd³⁺ are good dopants [28,29,30]. However, a very high excitation energy is required in those Ln³⁺-doped compounds due to the parity forbidden transitions within the 4f^N configurations of Ln³⁺ [31,32]. Fortunately, ns² doping can introduce new light absorption channels and thus act as sensitizers for luminescent Ln³⁺, such as Er³⁺, Yb³⁺, and Nd³⁺. For example, intense and multi-wavelength NIR luminescence was obtained in Cs₂ZrCl₆ at a low excitation energy through Te⁴⁺ co-doping with Er³⁺, Nd³⁺, or Yb³⁺ ions [33].

Herein, we realized the NIR emission in Te⁴⁺- and Er³⁺-co-doped Cs₂SnCl₆ microcrystal following a simple co-precipitation method. Under the low energy excitation at 391 nm, Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ displays an efficient NIR emission peak at ~1540 nm, in contrast to a negligible emission peak at this position in Er³⁺-singly doped Cs₂SnCl₆. The energy transfer processes from Te⁴⁺ to Er³⁺ f-electrons are proposed and discussed in detail based on the experimental findings.

2. Results and Discussion

2.1. Crystal Structure and Characterization

The Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ particles were synthesized through a facile co-precipitation method [33]. To be brief, the precursors SnCl₄, TeO₂, and ErCl₃·6H₂O were mixed with HCl and ethanol and dissolved. Thereafter, Cs₂CO₃ (dissolved in HCl) was added into the reaction mixture, and the perovskite MCs were immediately precipitated. More synthesis details are described in the Supporting Information (SI). As depicted in Figure 1a, the scanning electron microscopy (SEM) image showed that the size of the obtained Cs₂SnCl₆ crystals with a nominal molar concentration of 1.4% Te⁴⁺ and 10% Er³⁺ was mainly in the range of about 1–3 μm (Figure S1). The energy-dispersive spectroscopy (EDS) mapping in Figure 1b–g demonstrated that the constituent elements were uniformly distributed in the microcrystals, and the estimated Cs:(Sn+Te+Er):Cl composition ratio of microcrystals roughly agreed with the stoichiometric ratio of Cs₂SnCl₆, as given in Table S1. Moreover, the exact doping contents of Te⁴⁺ and Er³⁺ were determined to be 1.6% and 2.0% by inductively coupled plasma mass spectrometry (ICP-MS) (Table S2). It is noted that the Te⁴⁺ actual concentration was a little higher than its feeding concentration of 1.4%, which is mainly due to the high solubility of Te⁴⁺ in Cs₂SnCl₆ and the lower formation energy of Cs₂TeCl₆ than that of Cs₂SnCl₆ [34,35].

The XRD pattern in Figure 1h,i confirmed the cubic perovskite-type structure of Cs₂SnCl₆ with space group Fm-3m (PDF no. 07-0197), and no impurity phases were detected. In addition, the Rietveld analysis indicates that the diffraction peaks shifted to lower angles after doping (Figure S2) due to the lattice expansion (Table S3), as Sn⁴⁺ (r = 0.69 Å, CN = 6) was substituted by larger Te⁴⁺ (r = 0.97 Å, CN = 6) and Er³⁺ (r = 0.89 Å, CN = 6). These results indicated that Te⁴⁺, Er³⁺, and Te⁴⁺/Er³⁺ were successfully doped into the Cs₂SnCl₆ crystal lattice. X-ray photoelectron spectroscopy (XPS) was used to analyze the chemical valence state of elements in Cs₂SnCl₆ crystals (Figure 1j). The binding energies of the Cs 3d (Cs 3d_5/2: 723.55 eV, Cs 3d_3/2: 737.39 eV), Sn 3d (Sn 3d_3/2: 495.74 eV, Sn 3d_5/2: 487.24 eV), and Cl 2p (199.06 eV) peaks are consistent with the reported values [36] (Figure S3), proving that the as-prepared Cs₂SnCl₆ crystals are composed of tetravalent Sn. The peaks located at 586.64 eV and 576.38 eV correspond to Te⁴⁺ 3d, and 170.29 eV to Er³⁺ 4d, respectively. The binding energy of Sn 3d was almost same in un-doped and Te⁴⁺-doped samples, while it shifted toward a high energy side in the Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ (Figure S4). After combining the above ICP-MS results, substitutional site of Sn rather than the interstitial site is likely occupied by Te in the doped sample [37].

2.2. Optical Properties

The optical properties of Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ microcrystals were investigated via UV–Vis absorption and photoluminescence (PL) spectra. As shown in Figure 2a, Cs₂SnCl₆ microcrystals showed an optical absorption edge at around 315 nm, which is in agreement with the previous report [22]. While Er³⁺-singly doped Cs₂SnCl₆ has a similar result to the undoped one, interestingly, Te⁴⁺-singly doped and Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ exhibited intense absorption peaks within the region of 280–450 nm. Compared to the pure white color of the sample without Te⁴⁺, these new absorption bands changed the hue of the Te⁴⁺-doped Cs₂SnCl₆ to a luminous yellow (see the photographs in the inset of Figure 2a). In accordance with the absorption spectra, the PL excitation (PLE) spectra also showed peaks between 280 and 450 nm (Figure S5). Thereby, the absorption peaks located at 280–320 nm (A), 320–360 nm (B), and 360–450 nm (C) were derived from the Te⁴⁺-induced ion absorption and could be assigned to the inter-configurational 5s²→5s5p transitions of Te⁴⁺ [36,38].

The PL spectra of undoped and doped Cs₂SnCl₆ upon excitation at 391 nm were given in Figure 2b,c. In the visible region, no emission peaks were observed for both the pristine and Er³⁺-doped Cs₂SnCl₆ microcrystals (Figure 2b). In contrast, an intense yellow emission at about 577 nm with a large Stokes shift of 127 nm occurs after the doping of Te⁴⁺ in Cs₂SnCl₆, and the luminescence intensity of Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ is a little lower than that of the Te⁴⁺-singly doped one. Meanwhile, no emission was observed in the undoped and Te⁴⁺-doped Cs₂SnCl₆ in the NIR region (Figure 2c). Quite weak NIR emissions of Er³⁺-doped Cs₂SnCl₆ were observed in the spectra region from 1450 to 1600 nm, originating from the characteristic ⁴I_13/2→⁴I_15/2 transition of the Er³⁺ ion (Figure S6) [38,39]. The NIR emission intensity of Er³⁺-doped Cs₂SnCl₆ is too weak to evaluate the PLQY. In sharp contrast, Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ displayed an intense NIR emission at 1540 nm and its PLQY was 0.8%. It is also noted that the PL spectrum has other peaks at 1562 nm along with shoulders at around 1480 nm and 1506 nm, which may be caused by the crystal field split manifold of ⁴I_13/2 and ⁴I_15/2 states [40]. The phenomenon that the decreased peak intensity at 577 nm accompanies the enhanced emission at 1540 nm in Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ as compared with the Te⁴⁺singly doped sample suggests that the energy transfer and sensitization take place between the luminescent centers Te⁴⁺ and Er³⁺ in the former.

To achieve the highest NIR emission intensity for Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆, the Er³⁺ doping concentration was first optimized to 10% via monitoring the NIR emission intensity of Er³⁺-doped Cs₂SnCl₆ (Figure S7). Subsequently, the Te⁴⁺ precursor concentrations were varied while keeping a constant Er³⁺ concentration of 10%. It was seen in Figure S8 that the NIR emission intensity gradually increased to a maximum value at 1.4% Te⁴⁺ content and then decreased upon increasing the Te⁴⁺ doping amount (0.2–2.6%). The decrease in PL intensity is due to the concentration quenching effect arising from the energy migration among the ions [39]. As discerned in Figure 2c, Te⁴⁺-doped Cs₂SnCl₆ showed no luminescence at all in the NIR region of 1450–1600 nm. However, the luminescence intensity of Er³⁺ was remarkably enhanced with increasing Te⁴⁺ concentrations below 1.4%, which confirms the sensitization effect of Te⁴⁺ on the NIR luminescence of Er³⁺.

To better understand the sensitization effect of Te⁴⁺ on the Er³⁺ NIR emission, PL and time-resolved PL (TRPL) measurements of Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ were carried out at different Er³⁺ concentrations. As expected, the NIR emission intensity gradually increased with an increasing Er³⁺ concentration from 0 to 10% (Figure 3a), while the visible PL intensity continuously decreased (Figure 3b). The intensity of NIR emissions declines as the Er³⁺ concentration exceeds 10%, which suggests the occurrence of the concentration quenching effect. For Er³⁺-singly doped Cs₂SnCl₆, however, different Er³⁺ doping amounts all lead to a negligible NIR luminescence (Figure S9). At the same time, the visible luminescence lifetimes of Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ also decreased from 4.18 to 3.39 with the increase in the Er³⁺ concentration (Figure 3c, Table S4), which corresponds to the continuously increasing energy transfer efficiency (1 − τ_x/τ₀, where τ₀ is the lifetime of visible luminescence with Er³⁺ doping amount x = 0 [41].) of Te⁴⁺/Er³⁺ from 0.96% to 18.9% and benefits intense NIR emissions.

Furthermore, temperature-dependent PL spectra were studied for the as-prepared Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆. Figure 4a showed that the PL intensities of Te⁴⁺ declined monotonically upon increasing the temperature from 80 K to 300 K. This is attributed to the increased non-radiative transition probability of Te⁴⁺ at higher temperatures and the thermal-enhanced energy transfer from Te⁴⁺ to Er³⁺ [33]. In addition, the full width at half maximum (FWHM) increased as the temperature increased (Figure S10). A high Huang–Rhys factor (S) of 18 was obtained according to the temperature dependence of FWHM, which reveals the strong electron–phonon coupling effect in Cs₂SnCl₆:Te and facilitates the formation of STEs [24]. Nevertheless, the integrated PL intensities of Er³⁺ increased slightly with rising temperature (Figure 4b and Figure S11), which was accompanied by small variations in the PL lifetime at 1540 nm within this temperature range (Figure 4c, Table S5). The temperature-stable PL intensity of the NIR emission reflected a good protection of Er³⁺ 4f electrons by the outer electrons in 5s²5p⁶ shells [42]. It is worth noting that the intensity ratio of I₁₅₄₀/I₁₅₆₂ declined with an increasing temperature (Figure S12). This variation is caused by the population redistribution among crystal field split manifolds of ⁴I_13/2 and ⁴I_15/2 states at different temperatures, which is consistent with the similar observations made in different Er³⁺-doped hosts [39,43].

According to the above optical results, the energy transfer mechanism was described in Figure 4d based on the energy level alignment of Te⁴⁺ and Er³⁺. Thanks to the strong electron-phonon coupling in Cs₂SnCl₆ with soft lattice, transient elastic lattice deformation occurs upon photogeneration, where excitons tend to be self-trapped due to its lower energy and form self-trapped excitons (STEs) [43]. Therefore, carriers excited from ¹S₀ to ³P₁ of Te⁴⁺ ion under 391 nm excitation relax to form STEs, which then recombine to yield the broad band yellow emission at 577 nm in Te⁴⁺ doped Cs₂SnCl₆. For Er³⁺ singly-doped samples, the electrons in ground state ⁴I_15/2 transited to excited state ⁴I_13/2 under low energy excitation, and then generated very weak NIR emission (1540 nm) due to the parity-forbidden transitions within the 4f^N configurations [24,40]. In Te⁴⁺/Er³⁺ co-doped Cs₂SnCl₆, however, partial excitation energy was transferred from STEs to the well-matched ²H_11/2 energy level of Er³⁺ ions in addition to the yellow emission [33,37]. The transferred carriers relaxed non-radiatively to the ⁴I_13/2 energy level, and finally return to the ground states of the ⁴I_15/2 energy level through radiative transition, resulting in the enhanced 1540 nm NIR emissions at the expense of the weakened yellow luminescence from STEs.

2.3. Moisture Stability

Moisture stability is essential for practical applications of perovskite materials. Impressively, the NIR emission was very stable when the samples were exposed to air and even immersed in water. As shown in Figure S13, the XRD pattern of Te⁴⁺/Er³⁺-co-doped Cs₂SnCl₆ microcrystals was basically unchanged after being left in ambient air for 100 days. The PL intensity decreased by only 13% compared to the original data (Figure S14). Moreover, a strong NIR emission of the microcrystals was maintained after being immersed in deionized water for 2 h (Figure S15). Even after the samples were soaked in water for 8 h, the emissions still remained at 30% of the initial level, while the shape and position of XRD peaks remained unchanged (Figure S16). The superior stability in both the structure and NIR luminescence renders Cs₂SnCl₆ microcrystals more promising for practical applications relative to the common lead halide perovskites.

3. Conclusions

In conclusion, intense and multiple NIR emissions at 1540 nm and 1562 nm were achieved in a Sn-based double perovskite Cs₂SnCl₆ through Te⁴⁺/Er³⁺ co-doping. Under a low energy excitation at 391 nm, the NIR luminescence originating from the 4f→4f transition of Er³⁺ was significantly enhanced due to the effective energy transfer from the ¹S₀→³P₁ transition of Te⁴⁺. Furthermore, the Te⁴⁺/Er³⁺ Cs₂SnCl₆ microcrystals prepared via the simple co-precipitation method exhibited excellent emission and moisture stability. These findings bring novel emissive features to Cs₂SnCl₆ double perovskites, thus expanding their optoelectronic properties for future applications, such as NIR biosensors, anti-counterfeit technologies, and optical fiber communication.