Xin
Xu
a,
Zhongyi
Wang
a,
Qingkun
Kong
b,
Siping
Liu
c,
Ruiling
Zhang
*a,
Xiaojing
Liu
*a and
Keli
Han
ad
aInstitute of Frontier Chemistry, School of Chemistry and Chemical Engineering, Shandong University, Qingdao, 266237, P.R. China. E-mail: liuxj2020@sdu.edu.cn; rlzhang@sdu.edu.cn
bKey Laboratory of Catalytic Conversion and Clean Energy in Universities of Shandong Province, School of Chemistry and Chemical Engineering, Qufu Normal University, Qufu 273165, P. R. China
cGuangxi Key Laboratory of Chemistry and Engineering of Forest Products, School of Chemistry and Chemical Engineering, Guangxi Minzu University, Nanning 530006, P. R. China
dState Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Science, Dalian 116023, P.R. China
First published on 28th June 2024
Low-dimensional lead-free metal halides with efficient self-trapped exciton (STE) emission have attracted tremendous attention in lighting applications. Despite great efforts in manipulating the STE emission properties, it is still a great challenge to achieve controllable STE emission tuning. Herein, we successfully synthesized novel one-dimensional (1D) β-Cs3Cu2Br5 with tunable emission color modulation from blue-greenish to yellow by alloying with different Ag contents. In particular, the sample with a ratio of 12.4% exhibits broadband emission peaking at 508 nm with a high quantum yield of ca. 100%. Temperature-dependent single crystal X-ray diffraction, photoluminescence, and femtosecond transient absorption (TA) measurements further reveal that the Ag-induced lattice distortion of β-Cs3Cu2Br5 contributed to its controllable STE emission properties. Moreover, stable Ag-alloyed crystals have shown potential application prospects in encryption, radiation thermometry and solid-state light emitting devices. This lattice distortion tuning strategy provides a new vision for the controllable STE emission of low-dimensional metal halides.
Since an STE refers to an exciton being trapped in a potential well due to electron–phonon coupling and generating lattice distortion, STE emission is strongly related to the lattice structure.23–26 So far, external high pressure compression has been used to change the material inert structure to modulate the STE emission.27–33 For example, for the double perovskite Cs2Na0.4Ag0.6InCl6,31 the structural shrinkage of octahedra upon applying high pressure strengthened the metal–halogen bond energy and resulted in tunable emission from warm-white to deep-blue. The photoluminescence (PL) efficiency of 2D vacancy-ordered perovskite Cs3Bi2Cl9 nanocrystals (NCs) could be significantly improved by pressure compression.27 The initially non-emissive Cs3Bi2Cl9 NCs could emit bright yellow-greenish light at high pressure. Furthermore, pressure-induced lattice compression in zero-dimensional [(C6H5P)2]4SbCl5 also decreased the distance between organic cations and metal ions and accelerated the charge transfer, which finally promoted the peak position and emission efficiency.33 Although external pressure modulation could cause structural evolution to improve the optical properties of low-dimensional metal halides, tuning the internal intrinsic lattice distortion to modulate the emission properties at ambient pressure is mostly essential for practical solid-state lighting applications.
Element substitution is a strategy to modulate the STE properties of metal halides by adjusting the internal intrinsic structure.34–37 For example, the incorporation of Cu+ in Rb2AgBr3 can improve the lattice strength, significantly enhancing the emission efficiency to 98.8%.38 However, most of these substitutions only either modulate the host emission or generate new emissions, and the emission colors of the doped systems are independent of the dopant concentration.39–41 In addition, the modulation of STE thermal stability is also a key aspect for achieving high-performance solid-state light emitting devices.42–44 Thermally stable STE emission has been reported to be achieved in ion doped 0D metal halides, such as Rb3InCl6:xSb3+ and (BTPP)2MnCl4:2.0%:Sb. In this work, we synthesized a novel 1D β-Cs3Cu2Br5 single crystal stabilized by Ag alloying, which exhibits controllable STE emission from blue-greenish to yellow by modulating the lattice distortion at room temperature and pressure. All the Ag-alloyed β-Cs3Cu2Br5 samples exhibit high photoluminescence quantum yields (PLQYs) of 78–100%. The underlying tunable emission mechanism is revealed by a combination of temperature-dependent PL spectroscopy and structural analysis based on single crystal X-ray diffraction (SCXRD). The results demonstrate that Ag alloying generates lattice distortion and then improves the host STE emission. To the best of our knowledge, this is the first example of achieving controllable STE emission by tuning lattice distortion via ion alloying in low-dimensional metal halides under ambient conditions. The alloyed samples with high efficiency PL and superior stability also showed potential applications in information encryption, radiation thermometry and the solid-state lighting field.
However, we can obtain stable cyan-emissive crystals with high crystallinity (Fig. S4†) by adding a small amount of AgBr (e.g., 0.027 mmol) into the precursor solution. The actual ratio of Ag in crystals was also determined by inductively coupled plasma-mass spectrometry (ICP-MS) and is listed in Table S1.† For clarity, the sample is denoted as Cs3Cu2Br5:xAg+ (x is the molar ratio of Ag/Cu in crystals). Interestingly, the emission color is found to be Ag-content dependent varying from blue-greenish to yellow (Fig. 1a). To acquire the structural information, powder X-ray diffraction (PXRD) is performed. They have very similar diffraction patterns, while the diffraction peaks exhibit a 0.2° shift toward small angles with increasing Ag alloying content up to Cs3Cu2Br5:32.35%Ag+ (Fig. 1b). This suggests that they will have the same host crystal structures, but different lattice expansions due to the involvement of Ag+. The impurity of CsAgBr2 will appear at Ag feeding ratios higher than 1.07 (Fig. S5†). To determine the underlying crystal structure, SCXRD measurement was performed on Cs3Cu2Br5:32.35%Ag+ samples. The resolved crystal structure is shown in Fig. 1c, in which Ag substitutes the position of Cu and they have a disordered distribution in the lattice (the detailed parameters are summarized in Table S2†). The final emissive products are determined to be 1D β-Cs3Cu2Br5 with the involvement of Ag, which belongs to an orthorhombic system with the Cmcm space group. A novel 1D phase structure of β-Cs3Cu2Br5 is generated with the Cu–Br chains consisting of edge-sharing [CuBr4]3− tetrahedra surrounded by Cs+ cations, which is totally different from 0D Cs3Cu2Br5 with isolated [Cu2Br5]3− dimers. The experimental PXRD patterns of β-Cs3(Cu/Ag)2Br5 in Fig. 1b excellently matched the calculated ones from single crystal structure refinement, confirming the novel-synthesized 1D phase structure and the reliability of the refinement results. Moreover, we also resolved the crystal structure of the β-Cs3Cu2Br5:6.52%Ag+ sample, which has the smallest lattice expansion and the closest structure to pure β-Cs3Cu2Br5. As shown in Fig. S6,† the diffraction peaks of the unstable cyan-emissive crystals are a mixture of the simulated peaks based on the crystal structures of 0D Cs3Cu2Br5 and Cs3Cu2Br5:6.52%Ag+. Typically, the diffraction peak at 31.42° belonging to the (131)-oriented plane is in line with the simulated peaks of β-Cs3Cu2Br5:6.52%Ag+. This further confirms that the initially cyan emissive crystals without Ag involvement are pure β-Cs3Cu2Br5. As discussed above, 0D Cs3Cu2Br5 is suggested to be a thermodynamically stable phase and 1D β-Cs3Cu2Br5 is suggested to be a kinetically favored phase. We further provided the formation energy results through density functional theory (DFT) calculations. As shown in Fig. S7,† the calculated formation energy of 0D Cs3Cu2Br5 is −2.519 eV, which is much lower than that of 1D β-Cs3Cu2Br5 (−1.697 eV). This suggests that 0D Cs3Cu2Br5 is a thermodynamically stable product, while 1D β-Cs3Cu2Br5 is a metastable state with a kinetically favored phase. But the incorporation of Ag+ could significantly reduce the formation energy of 1D β-Cs3Cu2Br5 to −2.588 eV, which indicates that the involvement of Ag+ can further stabilize the metastable-state 1D β-Cs3Cu2Br5 to be thermodynamically favorable. The energy-dispersive spectroscopy (EDS) elemental mapping of β-Cs3Cu2Br5:12.40%Ag+ further demonstrates the uniform distribution of Cs, Ag, Cu and Br in β-Cs3(Cu/Ag)2Br5, respectively (Fig. 1d). The ratio of Cs:(Cu + Ag):Br is about 3:2:4.5, which is consistent with the stoichiometric ratio of single crystal refinement results (Fig. S8†). Furthermore, X-ray photoelectron spectroscopy (XPS) was further performed to uncover the electronic structures of β-Cs3(Cu/Ag)2Br5 with different chemical states. As shown in Fig. 1e, two typical Ag+ 3d spectrum peaks at 373.9 eV (3d3/2) and 367.9 eV (3d5/2) were observed, respectively. The peaks at 951.5 eV and 931.6 eV are assigned to Cu+ 2p1/2 and 2p3/2 (Fig. 1f), respectively. All these results support that the Ag and Cu ions are presented in the form of +1 valence chemical states.
To gain a deep understanding of the controllable emission mechanism of Ag-alloyed β-Cs3Cu2Br5, steady-state spectroscopy and time-resolved PL spectroscopy were performed. As shown in Fig. 2a, the β-Cs3Cu2Br5 crystals exhibit broad cyan emission with a peak at 493 nm, a full width at half maximum (FWHM) of 86 nm and a large Stokes shift of 153 nm. Broadband emission with large Stokes shifts is characteristic of STE emission in low-dimensional metal halides.24,25,48 The PL spectrum of β-Cs3Cu2Br5 could be well fitted with two Gaussian-shaped peaks, a high energy (HE) peak located at 490 nm and a low energy (LE) peak centered at 512 nm. This suggests that the emission may originate from two STE states. We subsequently recorded the time-resolved PL spectra of β-Cs3Cu2Br5 at room and low temperatures, respectively. Fig. 2b shows that the PL decay profile at 77 K can be well fitted by using double-exponential functions, yielding two lifetimes, τ1 = 0.53 µs and τ2 = 627.18 µs, respectively. The PL decay profile gives a lifetime of only 139.37 μs (Fig. S9†) at room temperature, suggesting that the two STE states are in thermal equilibrium.49 More importantly, we found that 1D Ag-alloyed β-Cs3Cu2Br5 not only possessed high air and thermal stability under ambient conditions (Fig. S10†), but also exhibited remarkably controllable emission properties depending on the Ag concentration. Fig. 2c displays the tunable bright emission with the light color changing from blue-greenish to yellow with the increase of the Ag ratio from 0.11 to 0.85 under UV irradiation. The corresponding Commission Internationale de l'Eclairage (CIE) chromaticity coordinates are shown in Fig. S11.† Meanwhile, the PL peaks show a continuous red-shift from 502 nm to 545 nm. The FWHM also became broader from 108 to 155 nm (Fig. S12†), suggesting that Ag plays a vital role in tuning the emission of Ag-alloyed β-Cs3Cu2Br5. Fig. 2d shows that the highest PLQY could be up to 100% when the Ag/Cu ratio is 12.40%. The PLQY of pure 1D β-Cs3Cu2Br5 is determined to be 68.2% (Fig. S13†), which is supposed to be lower than the true value due to the rapid phase transformation during the test process. Moreover, similar to the pristine crystal, all PL spectra of Ag-alloyed crystals could also be well fitted by two Gaussian shaped emission peaks (Fig. S14†), a HE one denoted as peak 1 and a LE one denoted as peak 2, and both peak 1 and peak 2 show a continuous red shift with the Ag ratio increasing from 6.52% to 32.35%. This suggests that the incorporation of Ag ions can directly modulate the emission of the inherent host STE state rather than introducing a new emissive STE state.23,50 The time-resolved PL curves of the samples with different Ag contents are presented in Fig. 2e and Table S3,† where all the samples show a double-exponential decay with a short component of 2–33.4 μs and a long component of 12.6–149.2 μs. This further excellently confirms the existence of the two STEs. Specifically, with the increase of the Ag ratio from 0.11 to 0.85, the short component contribution of 33.4 µs is correspondingly enhanced from 8.56% to 61.61% and the relative intensity of peak 2 originating from STE2 is also enhanced. This suggests that the short components of 2–33.4 μs are associated with the STE2 state, while the long components of 12.6–149.2 μs correspond to the STE1 state.
We further performed femtosecond transient absorption (fs TA) spectroscopy, which can provide a direct observation of STEs with characteristic broadband photo-induced absorption (PIA) signals on the red side of the exciton absorption.48Fig. 3a shows the pseudo-colored TA plot of β-Cs2Cu2Br5:6.52%, 12.40% and 32.35%Ag+ samples with excitation at 340 nm. Immediately after the pump pulse, we observed two broad PIA signals at around 400 and 620 nm. The former has a slow generation process, reaching a maximum after 50 ps. While the latter has a fast formation time within the instrumental response resolution (∼100 fs). Over the next 5 ps, a rapid decay of the PIA signal at 620 nm was observed, accompanied by an increase in the PIA band centered at about 525 nm. This suggests a transformation between two excited-state species. Both the PIA bands at 400 and 525 nm do not decay to zero in the probed time window. As discussed above, the broad emission of the Ag-alloyed samples is ascribed to the recombination of spin-triplet STE1 and STE2 with a microsecond long lifetime.22,45 The observation of two long-lived PIA bands at around 400 and 525 nm agrees well with the assignment and further confirms the presence of STE1 and STE2. Moreover, we also found that the relative intensity of the low-energy PIA bands was enhanced with increasing Ag content (Fig. S15†), which is consistent with the change of the ratio of the fitted peak 2 in the steady-state PL. Hence, we believe that the PIA band at 400 nm is associated with STE1 and the PIA band at 525 nm is related to STE2.
The dynamics at 400 nm exhibits two generation processes with time constants of 247 fs and 7.15 ps and a long-lived decay component (τ > 1 ns), as shown in Fig. 3b. Upon photoexcitation, free excitons were first formed and then trapped by a deformable lattice, producing spin-singlet STEs. Subsequently, spin-triplet STEs were formed via an intersystem crossing (ISC) process. Therefore, we ascribe the first ultrafast component of 247 fs to the trapping of free excitons and assign the second component of 7.15 ps to the ISC process from spin-singlet STE1 to spin-triplet STE1. The short-lived PIA signal at 620 nm is attributed to spin-singlet STE2. Its fast decay represents the ISC process from spin-singlet STE2 to spin-triplet STE2. Lifetime analysis at 617 nm reveals that this process occurs with a time constant of 2.49 ps. The long-lived components of 400 and 617 nm are associated with the decay of the resulting spin-triplet STE states.
To verify the influence of Ag alloying on the crystal structures, we resolved the crystal structures of the samples with different Ag contents. We first focus on the distortion of single [CuBr4]3− units, and the detailed individual angle and bond length are summarized in Table S4.† The tetrahedral parameters Δd (tetrahedral elongation) and σ2 (bond angle variance) were calculated to evaluate the structural deviations of [CuBr4]3− tetrahedra induced by Ag ions.51 As shown in Fig. 3c, as the Ag ratio increases from 6.52% to 32.35%, the bond angle variance (σ2) shows no obvious change, while a decrease in bond distance deviation is observed. Since the Ag ratio of 12.40% corresponds to the highest PLQY, this suggests that the strong Jahn–Teller effect of [CuBr4]3− units may be beneficial for improving the emission efficiency in this system. Subsequently, we also analyzed the distortion between adjacent [Cu2Br6]4− units. Fig. 3d shows the evolution of the [Cu2Br6]4− unit tilt angle in different Ag-alloyed samples; the corner-sharing Br–Br–Br angle decreases from 168.652 to 166.580° as x increases from 6.52% to 32.35% and the smaller tilt angle indicates strong lattice distortion in the crystals.52 As a result of the highly structural distortion in higher Ag alloyed samples, the strength of strong electron–phonon coupling would increase, leading to a change in the ratio of the two intrinsic STEs in β-Cs3Cu2Br5 and a red shift of the emission peak. In addition, the trapped excitons could be easily annihilated by the phonon-assisted nonradiative recombination process due to the strong electron–phonon coupling, and this was the reason for the lower PLQY in highly Ag-alloyed samples. The above results confirm that Ag ions can directly induce lattice distortion and thus control the STE emission properties.
We also recorded the temperature-dependent and time-resolved PL spectra of Ag-alloyed β-Cs3Cu2Br5 to investigate the effect of temperature on the STE emission. As shown in Fig. 3e and S16,† as the temperature decreases from 300 to 80 K, all of the PL spectra of different concentration Ag-alloyed β-Cs3Cu2Br5 become narrower and the PL peaks become gradually blue-shifted. Except for β-Cs3Cu2Br5:32.35%Ag+, the temperature-dependent PL intensity of Ag-alloyed samples exhibits high thermal stability in the range of 80–340 K (Fig. S17†). Furthermore, the Huang–Rhys factor (S) was calculated by fitting the temperature-dependent FWHM of β-Cs3Cu2Br5:x%Ag+, and the S value increased from 28.15 to 61.50 as the Ag ratio increased from 6.52 to 32.35% (Fig. S18†). This indicates the enhancement of the strength of electron–phonon coupling, proving that strong lattice distortion would enhance the strength of electron–phonon coupling.25 The samples with different Ag ratios exhibit the same PL peak at 502 nm at 80 K (Fig. S19†). The PL tail at 550–700 nm increases with increasing Ag content (Fig. S20†). It is notable that 6.52% and 12.40% Ag+-alloyed crystals exhibit different emission colors at room temperature, but have identical PL spectra at 80 K. We measured their cell volumes at 100 K and found that the cell volume of β-Cs3Cu2Br5:6.52%Ag+ (1304.995 Å3) is close to that of β-Cs3Cu2Br5:12.40%Ag+ (1309.705 Å3). The coincidence of the PL profiles of these two samples at 80 K may originate from their similar structures. Hence, we infer that β-Cs3Cu2Br5 has a very soft lattice and the lattice distortion can also be modulated by temperature, leading to tunable emission colors. The temperature-dependent SCXRD analysis results of β-Cs3Cu2Br5:32.35%Ag+ with detailed crystal lattice parameters at different temperatures are provided in Table S5,† which demonstrates that there is lattice shrinkage but no phase transition with decreasing temperature. We also calculated the tetrahedral parameters of [CuBr4]3− at different temperatures (Fig. S21†), and the results indicate that the temperature could largely suppress the lattice distortion of β-Cs3Cu2Br5:32.35%Ag+, leading to the enhancement of the emission intensity and the blue shift of the emission peak. These results also imply the intrinsic feature of the soft crystal lattice of β-Cs3Cu2Br5, which is the reason why Ag alloying can induce significant changes in crystal structures.
Fig. 3f shows the temperature-dependent time-resolved PL spectra of β-Cs3Cu2Br5:12.40%Ag+. The decay curves of β-Cs3Cu2Br5:12.40%Ag+ could be fitted by using bi-exponential functions at different temperatures. The lifetimes are provided in Table S6.† The long component of the lifetimes shows a dramatic decrease from 331.3 µs to 26.55 µs with increasing temperature from 100 to 340 K. Moreover, the two component contribution is strongly related to temperature. At 80 K, the fast component with a lifetime of 8.73 µs only contributes 1.85%, while the ratio of the fast component gradually increases to 31.6% as the temperature increases to 400 K. Similar phenomena also occur in other Ag-alloyed β-Cs3Cu2Br5 crystals (Table S7†). This further suggests that temperature can regulate the ratio of STE1 and STE2.
Based on the tunable emission colors and lifetimes at different temperatures of Ag-alloyed β-Cs3Cu2Br5, we explored its potential application in encryption, radiation thermometry and solid-state lighting devices. As shown in Fig. 4a, Ag-alloyed β-Cs3Cu2Br5 crystals exhibit an obvious emission color change after being heated at 125 °C. β-Cs3Cu2Br5:17.32%Ag+ exhibits an apparent tunable color change from cyan to yellow with the change of temperature (Fig. 4b), and the CIE coordinates of (0.17, 0.43) convert to (0.35, 0.43) when the temperature increases from 80 to 400 K. This motivates us to develop a Morse code-based encryption method by utilizing the temperature-induced tunable emission of Ag-alloyed β-Cs3Cu2Br5, where we define yellow emission as the dash and the other color emission as the dot. Fig. 4c displays the encryption information corresponding to “2023”. Under normal conditions, the wrong information “3894” is shown. It can be decrypted to “2023” after heating. Subsequently, the true information “2023” can be encrypted again with emission color recovery after cooling. As shown in Fig. S22,† the Morse code-based encryption strategy could also be applied to the encryption of complex words. The pattern displays the wrong information “bwljes” at room temperature, while the true information “copper” could only be decrypted after heating. This simple and reversible method provides a new idea for information encryption. In addition, the lifetime of Ag alloyed β-Cs3Cu2Br5 dramatically decreases with increasing temperature, which is available for remote thermometry. The relationship between temperature and PL lifetimes could be well fitted by using the Boltzmann-sigmoid model, , where τ0 is the intrinsic radiative lifetime, T and TB represent the temperature and quenching temperature, and ΔT is the half-width of the temperature sensitive range, and the temperature sensing range is calculated to be TB + 2ΔT.53 The temperature sensing range is found to be dependent on the content of Ag ions, where β-Cs3Cu2Br5:6.52%Ag+ has an ultra-broad sensing range of 119–351 K, whereas the 32.35% Ag+-alloyed sample exhibits a narrower sensing range of 102–246 K. The sensing area is shown in Fig. 4d and Fig. S23† with a color-filled area. The alterable sensing range of Ag-alloyed β-Cs3Cu2Br5 indicates that the phonon-assisted thermal effect could be regulated by Ag-induced lattice distortion,53 which provides the first example for adjusting the temperature sensing range within a single-phase metal halide. As shown in Fig. S24,† the PL lifetime of β-Cs3Cu2Br5:32.35%Ag+ shows a linear relationship with temperature in the range of 80 to 260 K, wherein the rate of change of lifetime is 1.15 µs K−1, indicating that Ag-0.85 is a promising candidate for thermometry applications. The thermometry performance is described by the specific sensitivity, . The maximum value reaches 2.8% K−1 for β-Cs3Cu2Br5:32.35%Ag+ (Fig. 4e). This value is comparable to that of other lifetime-based thermometry (Table S8†),54–57 which implies that Ag-alloyed β-Cs3Cu2Br5 is a promising material for radiation thermometry. Furthermore, the highly efficient Ag-alloyed β-Cs3Cu2Br5 crystals are desirable candidates for fabricating phosphor-converting devices. As shown in Fig. 4f, Ag-alloyed β-Cs3Cu2Br5 crystal powder with different ratios was coated on a commercial 340 nm UV light emitting diode (LED) chip. The devices could display different emission colors ranging from blue-greenish to yellow (Fig. S25†), and the performances of these devices are summarized in Table S9.† In addition, the emission spectra remain unchanged at different voltages (Fig. S26†), indicating the desirable stability of the device. To achieve a white light emission device, pure 0D Cs3Cu2Br5 (blue emission) and β-Cs3Cu2Br5:32.35%Ag+ (yellow emission) crystals were mixed in a ratio of 2:1 and coated on a 275 nm LED chip. The as-fabricated device exhibits cold, pure and white light at different voltages (Fig. S27†), which is due to the PLQY difference between the two samples.
Footnote |
† Electronic supplementary information (ESI) available: Experimental, supplementary Fig. S1–S27, and Tables S1–S9. See DOI: https://doi.org/10.1039/d4qi00970c |
This journal is © the Partner Organisations 2024 |