1. Introduction
In recent years, zero-dimensional (0D) metal halides have attracted significant attention due to their remarkable photoluminescence (PL) quantum yields (QY) and adjustable emissions. These materials, characterized by the confinement of luminescent metal halide octahedra by organic or inorganic cations, offer great potential for various optoelectronic applications, including LEDs, solar cells, scintillators, sensors, and thermal imaging systems [
1,
2,
3,
4,
5]. In these materials, the Jahn-Teller distortion of metal halide octahedra upon photoexcitation leads to the localization of excitons, enhancing radiative emission by preventing exciton migration to defects. This phenomenon, known as self-trapping of excitons (STEs), contributes to high PL QY and improved stability [
6,
7]. While significant progress has been made in exploring efficient 0D metal halides, the use of lead-based compounds such as Cs
4PbBr
6 is limited by the toxicity of lead, posing environmental and health concerns [
8,
9,
10]. To address this issue, tin (Sn
2+) has emerged as a promising alternative to lead (Pb
2+) due to its similar electronic properties and environmentally-friendly nature [
11,
12,
13]. Numerous studies have probed into the intriguing properties of 0D tin halide perovskites, including Cs
4SnX
6 (where X = Br, I), and demonstrated their potential for optoelectronic applications [
14,
15,
16]. For instance, the work conducted by Kovalenko and his team presented the compelling discovery of an efficient green-yellow emission derived from self-trapped excitons in Cs
4SnBr
6, with the achievement of a notable PL QY of 15±5% at room temperature [
14]. Intriguingly, through strategic substitution of Cs
+ with Rb
+ or K
+ and Br
- with I
-, both the PL peak position and Stokes shift can be simultaneously adjusted [
14]. Quan et al. synthesized high-quality Cs
4SnX
6 (X = Br, I) nanocrystals that exhibit well-defined shapes and narrow size distributions, culminating in an impressive PL QY of up to 21% for Cs
4SnBr
6 nanocrystals [
16]. However, a persistent obstacle facing tin halide perovskites is their lack of stability in air, a problem primarily attributed to the oxidation of Sn
2+ to Sn
4+, leading to a consequential decline in luminescent performance [
17,
18]. To overcome this issue, Zhang et al. implemented an innovative solution by introducing SnF
2 as a tin source, which replaced the easily oxidizable SnBr
2, and successfully enhanced the structural stability of Cs
4SnBr
6 perovskite by utilizing fluorine to suppress Sn
2+ oxidation effectively [
19]. Despite the remarkable PL performance displayed by 0D tin halide perovskites, the quest for their commercialization necessitates considerable efforts to further enhance both their PL efficiency and stability.
0D perovskites generally display wide-band emission with a large Stokes shift. Numerous researchers have employed the STEs model to elucidate the wide-band emission and large Stokes shift induced by octahedral distortions in 0D perovskites upon light excitation [
20,
21,
22,
23,
24]. Strong electron-phonon coupling and a malleable lattice are widely considered the cornerstone factors driving the emission of STEs. Particularly within the context of 0D perovskites, excitons, when photoexcited, induce swift distortions in the lattice of the excited state, culminating in the formation of localized STEs [
22,
23,
24,
25,
26]. Previous investigations have revealed that the lattice distortion in zero-dimensional metal halide perovskites can be manipulated by adjusting factors such as the chemical composition, temperature, and pressure. Such modulation enables precise control of the STE states, which in turn optimizes luminescence performance [
22,
23,
24,
25,
26,
27]. For example, the introduction of varying metal ions into 0D metal halide perovskites can amplify the distortion of [BX6]
4− octahedra and bolster electron-phonon coupling, consequently enhancing the density of STE states and boosting luminescence efficiency [
26]. In previous work, we successfully expanded the emission spectra and amplified the emission efficiency of STEs in Cs
4SnBr
6 through an innovative Mn
2+ doping strategy [
28]. This approach imbued the Mn
2+-doped Cs
4SnBr
6 with remarkably enhanced PL QY of up to ~75%, a broader emission spectrum, and increased thermal stability.
Rapid thermal treatment (RTT) is a widely used technique to modify the micro-nano structure and enhance the optoelectronic properties of materials [
29]. RTT involves rapid heating and cooling, characterized by short heating times and accelerated cooling rates. This transient process can induce changes in the structural order of metal halide octahedra, which plays an important role in the formation of STEs. However, the influence of RTT on the PL properties of 0D Cs
4SnBr
6 has not been fully explored. In this study, we propose a novel strategy based on RTT to enhance the emission of STEs in lead-free Cs
4SnBr
6. We systematically investigate the effects of RTT on the optical properties of Cs
4SnBr
6. Compared to pristine Cs
4SnBr
6, we find that RTT at 200°C for 120 seconds leads to enhanced STE emission, with the PL QY escalating from an initial 50.1% to a substantial 64.7%. The improvement in PL is discussed in terms of electron-phonon coupling and increased binding energies of STEs induced by RTT.
2. Materials and Methods
Cs4SnBr6 samples were synthesized through water-assisted wet ball-milling. Cesium bromide (4 mmol, CsBr, Aladdin, 99.9%), stannous fluoride (1 mmol, SnF2, Macklin, 99.9%), and ammonium bromide (1 mmol, NH4Br, Aladdin, 99.99%) were employed as reactant precursors. To achieve Cs4SnBr6, the molar ratios of CsBr, SnF2, and NH4Br were maintained at 4, 1, and 2 mmol, respectively. Initially, the precursors were loaded into a jar and combined with 60 μL of deionized water. Subsequently, a ball milling process was conducted for 30 min at a speed of 600 rpm. The resulting product was then dried in a vacuum-drying oven for 120 min at room temperature and annealed at various temperatures ranging from 100 to 300 °C using a straightforward RTT process. The RTT process was executed on a rapid thermal processor, heating the sample to the annealing temperature at a rate of 10 °C s−1. After maintaining the annealing temperature for 30, 90 and 120 s, respectively, the system was rapidly cooled to room temperature. Upon cooling, the Cs4SnBr6 powder was acquired via ball milling for 30 min at a speed of 600 rpm. PL measurements were conducted using an Edinburgh Instrument FLS1000 PL spectrometer (Livingstone, UK). The PL spectra were obtained at different temperatures to investigate the temperature-dependent behavior. PL excitation (PLE) spectra and time-resolved PL spectra were also recorded using the same instrument. The crystal structures of Cs4SnBr6 were analyzed using X-ray diffraction (XRD) with a Bruker D8 Advance instrument (Karlsruhe, Germany). XRD measurements were performed at 35 kV and 35 mA to determine the crystal structure of the samples. The compositions of Cs4SnBr6 were determined by energy dispersive spectroscopy (EDS) using a Bruker EDS QUANTAX system (Karlsruhe, Germany). Scanning electron microscopy (SEM) (A Hitachi SU5000 SEM instrument, Tokyo, Japan) was employed to investigate the surface morphology and microstructure of Cs4SnBr6.
3. Results and discussion
The PL spectra of pristine Cs
4SnBr
6 sample and samples subjected to RTT at different temperatures are shown in
Figure 1(a). Both types of samples exhibited a broad emission with a peak at approximately 530 nm. The emission band had a large full width at half maximum (FWHM) of approximately 105 nm. Additionally, a significant Stokes shift of around 1.30 eV was observed, as illustrated in
Figure 1(c) and 1(d). It is evident that RTT at temperatures below 150°C had minimal influence on the PL intensity of Cs
4SnBr
6 samples. However, a notable increase in PL intensity was observed when the RTT temperature was raised to 200°C. Conversely, as the RTT temperature was further increased to 300°C, the PL intensity of Cs
4SnBr
6 samples decreased rapidly.
Figure 1(b) showcases PL spectra of Cs
4SnBr
6 samples annealed at a RTT temperature of 200°C over diverse durations. The observed PL intensity of the Cs
4SnBr
6 samples appears to increment gradually with the prolongation of RTT duration from 30 to 120 seconds. Notably, as depicted in
Figure 1(e) and 1(f), the PL QY takes a significant leap from 50.1% to 64.7% following the annealing of the unprocessed Cs
4SnBr
6 at an RTT temperature of 200°C for a span of 120 seconds.
Figure 2a and
Figure 2b display the excitation power dependence of PL for both the pristine sample and sample-120s. The insets of
Figure 2a and
Figure 2b demonstrate that an increase in excitation power, from 160 to 2817 nW, is accompanied by a corresponding enhancement in PL intensity. Nevertheless, the PL peak position maintains consistent, unaffected by variations in the excitation power. Additionally, a distinct linear correlation emerges between the integrated PL intensity and the excitation power within the range of 160 to 2817 nW. The excitation power-dependent PL intensity is commonly employed to determine the underlying mechanism of light emission in semiconductors. As per the literature [
30], the PL intensity (I) can be described by the equation
, where I
0 denotes the excitation power, η symbolizes the emission efficiency, and the exponent k is affiliated with the radiative recombination process. A linear fit of ln(I/η) in contrast to ln(I
0) allows the estimation of the k parameter values as 1.11 and 1.17 for the pristine sample and sample-120s respectively. This deduction strongly suggests that the green emission in both samples is engendered by the recombination of excitons. Given the large Stokes shift of ~1.30 eV, coupled with a broad FWHM of the emission band approximating ~105 nm, and further considering the long radiative lifetime of ~1μs, as illustrated in
Figure 5, the green emission can be ascribed to the radiative recombination STEs, which is prompted by Jahn–Teller distortion of [SnBr
6]
4- octahedra in 0D perovskite [
31,
32,
33].
To understand the enhanced PL, the structure and compositions of Cs
4SnBr
6 samples were evaluated via SEM and EDS, respectively.
Figure 3(a) showcases the SEM image obtained from the pristine Cs
4SnBr
6 sample. The EDS spectrum reveals the presence of Cs, Sn, Br, and F elements in Cs
4SnBr
6, which are uniformly distributed, as demonstrated in the EDS mapping of
Figure 3(c)-3(f). The Cs, Sn, Br, and F elements maintain this uniform distribution even after Cs
4SnBr
6 was annealed at an RTT temperature of 200°C for a duration of 120 s, as shown in
Figure 4(c)-4(f). This result indicates that the elemental distribution of Cs
4SnBr
6 remained unaltered when subjected to RTT at 200°C for a duration of 120 s.
In
Figure 5, we present the XRD patterns obtained for different samples. The XRD pattern of pristine Cs
4SnBr
6 sample (S-0s) reveals the coexistence phase of Cs
4SnBr
6 and CsBr due to incomplete consumption of CsBr powder precursors in the solid-state reaction. Apart from the prominent diffraction peak at 29.7°, attributed to the CsBr phase, we observe that the diffraction peaks of S-0s from the Cs
4SnBr
6 phase are consistent with those reported for Cs
4SnBr
6-SnF
2[
19,
34]. This consistency suggests that the substitution of Br
- with smaller F
- effectively suppresses the oxidation of Sn
2+ in Cs
4SnBr
6. Notably, the diffraction peaks corresponding to crystal planes (110), (300), (131), (223), and (330) of the Cs
4SnBr
6 phase become more pronounced and well-defined as the RTT time increases from 0 s to 120 s [
34]. This strongly indicates an enhanced crystallinity of the Cs
4SnBr
6 powders after the RTT process.
Figure 5.
XRD patterns of the pristine Cs4SnBr6 sample and the samples annealed at a RTT temperature of 200°C for 30s, 90s, and 120s, respectively.
Figure 5.
XRD patterns of the pristine Cs4SnBr6 sample and the samples annealed at a RTT temperature of 200°C for 30s, 90s, and 120s, respectively.
To better comprehend the PL characteristics, we employed an excitation wavelength of 375 nm (facilitated by 70 ps excitation pulses from a laser) to measure the PL decay curves, as illustrated in
Figure 6. The PL decay curve associated with the green emission yielded a sound fit when we employed a biexponential decay function [
35]:
where I
0 represents the background level, τ
1 and τ
2 represent the lifetimes of each exponential decay component, and A
1 and A
2 denote the corresponding amplitudes. The intensity-weighted averaged PL lifetimes are then determined by
[
35]. As depicted in
Figure 6, the green emission in all samples exhibits a slow decay with a long radiative lifetime of approximately 1 μs. Notably, the lifetime remains nearly unchanged despite variations in the RTT conditions, even though the crystallinity of the Cs
4SnBr
6 samples improves after the RTT process. These findings suggest that the PL enhancement does not exclusively arise from a reduction in nonradiative recombination centers.
In order to gain insights into the influence of electron-phonon coupling, the temperature-dependent PL spectra of the samples were measured in the range of 100 to 300 K. As shown in
Figure 7(a) and (b), a decrease in temperature led to an increase in PL intensity and a reduction in the FWHM for both the pristine sample and sample-120s. According to the theory proposed by Stadler [
36], the FWHM of the PL peak is closely related to the electron-phonon coupling and can be described by the following equation:
where S is the Huang-Rhys factor, ℏω is the energy of the phonon mode, T is the temperature, and k
B is Boltzmann's constant. By fitting the temperature-dependent FWHM of the PL peaks using the equation 2, we can calculate the value of the Huang-Rhys factor S, which is commonly used to describe the exciton-phonon coupling. For the pristine sample, it was found that the value of S is as large as 63.7 (see
Figure 7(c)), which is significantly higher than that reported in Cs
4SnBr
6. Generally, a larger S value indicates a stronger electron-phonon coupling, which is more favorable for the formation of STEs. However, a high S value also increases the probability of non-radiative recombination [
24]. The Raman spectrum shown in the inset of
Figure 7(c) reveals two dominant phonon modes, corresponding to the Sn-Br stretching vibrational modes at approximately 130 cm
-1 and 220 cm
-1 [
37,
38], respectively, which may be involved in the electron-phonon coupling and thus result in the S value as large as 63.7. After annealing the sample at a RTT temperature of 200°C for 120 s, the value of S is significantly reduced to 46.1, which closely resembles that observed in Cs
4SnBr
6 [
28]. This reduction can be attributed to the disappearance of the Sn-Br stretching vibrational mode near 240 cm
-1 (see the inset of
Figure 7(d)) resulting from the enhanced crystallinity of the Cs
4SnBr
6 powders after the RTT process, as evidenced by the X-ray diffraction (XRD) patterns (refer to
Figure 5). The fitted data in
Figure 7(d) provide an optical phonon energy (E
LO) of 16.7 meV (134 cm
-1), which corresponds well with the Sn-Br stretching vibrational mode near 130 cm
-1 in Cs
4SnBr
6 (see the inset of
Figure 7(d)) [
37,
38]. This suggests that the electron-phonon coupling only involves the dominant phonon mode near 130 cm
-1, which corresponds to the Sn-Br stretching vibrational mode near 130 cm
-1. On another note, the PL QY of STEs highly depends on the exciton binding energy. The detrapping of STEs due to thermal activation results in a decreased radiative recombination rate. The exciton binding energy of STEs can be determined by analyzing the temperature-dependent integrated PL intensity (I
PL) using the Arrhenius equation [
33]:
where I
PL(T
0) is the integrated PL intensity at 80 K, β is a constant related to the density of radiative recombination centers, k
B is Boltzmann's constant, and E
b is the exciton binding energy. Through fitting the experimental data with the Arrhenius equation, we can obtain the exciton binding energy E
b of 227 meV and 409 meV for the pristine sample and sample S-120, respectively (see
Figure 7(e) and
Figure 5(f)). It is noteworthy that the E
b value in sample S-120 substantially exceeds the 227 meV found in the pristine sample. This infers that the detrapping of STEs via thermal activation has been effectively suppressed in sample S-120, resulting in an augmented emission from the STEs. Therefore, based on the above analyses, we deduce that optimal electron-phonon coupling, compounded by the enhanced exciton binding energy elicited by the RTT, is responsible for the enhanced STE emission discerned in sample S-120.