High Quantum Yield Blue Emission from Lead-Free Inorganic

wide applications are hindered from toxic lead element which is not environment- and consumer-friendly. Herein, we utilized heterovalent substitution ...
3 downloads 0 Views 4MB Size
High Quantum Yield Blue Emission from LeadFree Inorganic Antimony Halide Perovskite Colloidal Quantum Dots Downloaded via KAOHSIUNG MEDICAL UNIV on July 28, 2018 at 10:26:58 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Jian Zhang, Ying Yang, Hui Deng, Umar Farooq, Xiaokun Yang, Jahangeer Khan, Jiang Tang, and Haisheng Song* Wuhan National Laboratory for Optoelectronics (WNLO) and School of Optical and Electronic Information, Huazhong University of Science and Technology (HUST), 1037 Luoyu Road, 430074 Wuhan, Hubei, P.R. China S Supporting Information *

ABSTRACT: Colloidal quantum dots (QDs) of lead halide perovskite have recently received great attention owing to their remarkable performances in optoelectronic applications. However, their wide applications are hindered from toxic lead element, which is not environment- and consumer-friendly. Herein, we utilized heterovalent substitution of divalent lead (Pb2+) with trivalent antimony (Sb3+) to synthesize stable and brightly luminescent Cs3Sb2Br9 QDs. The lead-free, fullinorganic QDs were fabricated by a modified ligand-assisted reprecipitation strategy. A photoluminescence quantum yield (PLQY) was determined to be 46% at 410 nm, which was superior to that of other reported halide perovskite QDs. The PL enhancement mechanism was unraveled by surface composition derived quantum-well band structure and their large exciton binding energy. The Br-rich surface and the observed 530 meV exciton binding energy were proposed to guarantee the efficient radiative recombination. In addition, we can also tune the inorganic perovskite QD (Cs3Sb2X9) emission wavelength from 370 to 560 nm via anion exchange reactions. The developed full-inorganic lead-free Sbperovskite QDs with high PLQY and stable emission promise great potential for efficient emission candidates. KEYWORDS: lead-free perovskite, inorganic perovskite quantum dots, quantum yield, quantum-well band structure, photostability degradation.13,14 Accordingly, fully inorganic CsPbX3 (X = Cl, Br, I) QDs are emerging as a class of metal halide perovskite QDs, which are less susceptible toward oxygen and moisture. To date, tremendous progress has been achieved in controlling the morphology, size, and halide of CsPbX3 perovskite QDs with PL emission tunable from blue to red.15−17 The left key problem was the lead toxicity, which became the hindrance for inorganic lead halide perovskite applications. Limited effort has been dedicated to address this issue. Thus, the lead-free or less toxic element perovskite becomes the next indispensable work for perovskite QDs.2,18 To obtain low-toxic metal halide perovskites, there were two typical strategies to reduce the utilization of lead. One is the partial substitution of Pb by other low-toxic metal cations. The other is lead-free cation substitution of lead, such as Sn(II), Sn(IV), Mn(II), Bi(III), Sb(III), Cu(II), etc.1,18−20 These low-

O

rganic−inorganic perovskites have attracted great research interest due to their superior electrical and optical properties.1−6 In a remarkably short time, solution-processed metal halide perovskite thin film solar cells have obtained a high certified power conversion efficiency of 22.1% comparable to commercial material (Si, CdTe, CIGS, etc.) based photovoltaics.7 Their single-crystal nanowire lasers show ultralow-threshold room-temperature lasing with tunable emissions and ease of synthesis.8−10 Metal halide perovskite quantum dots (QDs) demonstrate high photoluminescence quantum yield (PLQY > 90%), narrow full width at halfmaximum (fwhm 105/cm), nearly direct band gap, and small effective masses.21 Combing better stability of inorganic perovskite, antimony inorganic perovskite QDs (IPQDs) were expected to further improve their PLQY values. In this work, we successfully synthesized lead-free ⟨111⟩stacked layered Cs3Sb2Br9 IPQDs. The IPQDs were obtained by a modified ligand-assisted reprecipitation (m-LARP) method22 at RT within few second reaction. Their emission QYs could be improved from 20 to 46% with an emission peak of 410 nm by controlling the QD crystallization. Such PLQY value was much higher than that of reported perovskite QDs

RESULTS AND DISCUSSION As shown in Figure 1a,d, the crystal structure of the Cs3Sb2Br9 (P3̅m1, no. 164) is derived from the traditional structure of perovskite CsSbBr3 (Figure 1a) by moving every third Sb layer along ⟨111⟩ to achieve correct charge balance. Using the crystallographic data (Table S1, Supporting Information),25 the unit cell of Cs3Sb2Br9 perovskite was schematically described, as shown in Figure 1b. It consists of bioctahedral (Sb2Br9)3− clusters which are surrounded by cesium cations. To analyze optoelectronic properties of the lead-free Sb-perovskite, we prepare a Cs3Sb2Br9 single crystal with a size of 4.11 mm × 4.11 mm × 2.02 mm by inverse temperature crystallization technique. A yellow, lustrous hexagonal-shaped Cs3Sb2Br9 single crystal (inset of Figure 1c), and its X-ray diffraction 9295

DOI: 10.1021/acsnano.7b04683 ACS Nano 2017, 11, 9294−9302

Article

ACS Nano

Figure 3. Physical properties of colloidal Cs3Sb2Br9 IPQDs. (a) TEM image. (b) HRTEM image of a typical Cs3Sb2Br9 IPQD. (c) Size distribution analysis for the sample in (a). (d) XRD patterns. (e) UV−vis absorption and PL spectra. (f) Time-resolved PL decay and fitting curve of a typical sample.

results are listed in Table S2. The most suitable solvent was selected as DMF or DMSO, and the antisolvent was octane. For QD surface passivation, introduction of capping ligands was a mature strategy to control their morphology. Thus, it inspired us to control the crystallization of Cs3Sb2Br9 IPQDs by varying the surface ligands. Without long-chain ligand assistance or overdose, Sb-perovskite QDs could not be obtained by the reprecipitation method. Utilizing octane and OA mixture as ligand, stable quantum dots could be fabricated. The PLQYs of IPQDs were ∼20%. Through further optimizing the precursor concentration, reaction temperature, and cation ratio, brightly luminescent colloidal IPQDs can be obtained with a precursor concentration of 0.033 mM, an octane/OA ratio of 10:1, and a reaction temperature of RT (Figure S1). The colloidal QD solution photos under sunlight and UV light irradiation are shown in Figure 2c. The clear IPQD solution with uniform emission demonstrated the efficient emission without any aggregation at the bottom. The uniform size distribution of Cs3Sb2Br9 IPQDs was tested by transmission electron microscopy (TEM), as shown in Figure 3a. The obtained colloidal QDs were first characterized by TEM, as shown in Figure 3a,b. The monodispersed Cs3Sb2Br9 IPQDs showed uniform size distribution (Figure 3a) and were highly crystallized (Figure 3b). The average diameter of IPQDs derived from TEM was 3.07 ± 0.6 nm (Figure 3c). The XRD patterns and high-resolution TEM (HRTEM) images were applied to analyze the phase structure of the obtained Cs3Sb2Br9 IPQDs. In Figure 3d, the XRD pattern confirmed the trigonal crystal structure of Cs3Sb2Br9. From the HRTEM image and its corresponding fast Fourier transform (FFT) image in Figure 3b, the interplanar distances were 1.98 and 1.98 Å, respectively, corresponding to (224) and (024) crystal faces with a clip angle of 55°, agreeing well with the identified structure of XRD (Table S3). The absorption and emission spectra of Cs3Sb2Br9 IPQDs are shown in Figure 3e. Their absorption spectra (dashed curve) demonstrated an obvious exciton peak located at approximately 375 nm. In PL spectra

(XRD) spectra are shown in Figure 1c. It holds a trigonal phase with lattice parameters of a = b = 7.930 Å, c = 9.716 Å, which is consistent with the space group P3̅m1 (no. 164) described in Figure 1b,d. To further characterize optical properties of a Cs3Sb2Br9 single crystal, PL and UV−vis diffuse reflectance spectra were measured, as shown in Figure 1e. It emitted at 534 nm with a fwhm of 80 nm. The optical band gap for Cs3Sb2Br9 can be calculated by Tauc’s plot, which is described as (αhν)1/ n = C × (hν − Eg )

(1)

where α is the absorption coefficient, h is Planck’s constant, C is the proportionality constant, ν is the frequency of light, Eg is the band gap, and n is 1/2 or 2 depending on whether it is direct or indirect band gap material. The fitting results confirmed that the obtained Cs3Sb2Br9 perovskite was a direct band gap semiconductor with a band gap of 2.36 eV (Figure 1f). In order to implement the synthesis at RT, we adopted a modified ligand-assisted recrystallization strategy to fabricate Cs3Sb2Br9 IPQDs in order to avoid conventional heating, inert gas, and injection conditions. The m-LARP synthesis could be briefly described as directly adding the reaction precursor solution into a vigorously stirred poor solvent. Figure 2a,b shows the schematic illustrations of the m-LARP technique.22 In a typical fabrication process, a mixture of SbBr3, CsBr, and oleylamine was dissolved in N,N-dimethylformamide (DMF) or dimethylsulfoxide (DMSO) “good” solvent to form a clear precursor solution, and then a fixed amount of the prepared precursor solution was dropped into a mixed solution of octane and oleic acid (OA), which were “poor” solvents for these precursor ions. Under vigorous stirring, a highly supersaturated state could be obtained immediately and induced rapid recrystallization. In order to improve the quality of IPQDs, systematic optimization procedures were designed and implemented to extract the synthesis recipes. First, the precursor solubility in typical solvent was tested, and the 9296

DOI: 10.1021/acsnano.7b04683 ACS Nano 2017, 11, 9294−9302

Article

ACS Nano Table 1. Optical Parameter Comparisons of Halide Perovskite QDs Cs3Sb2Br9 MA3Bi2Br9 CsPbxMn1−xCl3 CsPbCl3 CsSnX3

emission peak (nm)

PLQY (%)

fwhm (nm)

temperature

air stability

photostability

reference

410 430 579 408

46 12 54 10 0.14

41 62

RT RT 170 °C 150 °C

35 days (70%) 7 days (0.01%) 3 months 30 days (95%) 5 min (0.05%)

108 h (50%) 25 h (92%) 60 min (47%) 60 min (38%)

this work 19 1 6 52

12

Figure 4. Composition analysis to Cs3Sb2Br9 IPQDs. (a) EDS spectra. (b−e) High-resolution XPS spectra corresponding to Cs 3d (b), Br 3d (c), Sb 3d (d), and Sb 4d (e) before and after Ar+ etching for Cs3Sb2Br9 IPQDs. (f) Proposed band structure diagram.

the kinetics of exciton recombination in Cs3Sb2Br9 IPQDs, we conducted time-correlated single-photon counting measurements, as shown in Figure 3f. The PL decay curves were fitted with a biexponential decay model (eq 2), in which the PL lifetime was considered as the sum of fast- and slow-decay components that give a short lifetime τ1 and a long lifetime τ2, respectively. The obtained τ1 was 2.4 ns with a percentage of 71%, and τ2 was 8.9 ns with a percentage of 29%, which revealed the high ratio of radiative-to-nonradiative transition. According to organic−inorganic hybrid and all-inorganic perovskite QD reports,6,22,31 the obtained average lifetimes are 4.285 ns much faster than those of traditional perovskites, indicating higher ratio of exciton recombination and less transition at defect states.

(solid curve), their emission showed a sharp and narrow emission peak at 410 nm with a fwhm of 41 nm. The sample showed a relative Stokes shift of 280 meV, implying almost no overlap between absorption and PL emission. Such a weak selfabsorption is beneficial for the usage as phosphors in lighting applications.26 Compared to that of a Cs3Sb2Br9 single crystal, the QD emission peak showed a 120 nm blue shift, demonstrating strong quantum confinement effect. For the PLQY measurements, relative fluorescence test was adopted to obtain the relative value27 using quinine sulfate acid aqueous solution as the standard sample (the measurement procedure is described in the Supporting Information). In order to ensure the accuracy of the calculation method, we utilized the PLQY values of typical CsPbBr3 QD emission at 513 nm for calibration. The PLQY of CsPbBr3 QDs was ∼90% utilizing the relative fluorescence test, consistent with the reported results from Zeng’s group.6 Thus, our present relative fluorescence test was credible to be used for PLQY measurements. The PLQYs of Cs3Sb2Br9 IPQDs were in the range of 20−46%. The optical quality and the characteristic parameters of typical IPQDs are summarized in Table 1. The PLQY values of our Sb-perovskite IPQDs were similar to those of lead-less perovskite QDs and also comparable to those of traditional lead-based IPQDs.6,28,29 It is worth noting that the present PLQY value (46%) obtained a large improvement compared to previous lead-free perovskite QDs.19,20,30 The high PLQY indicated the suppression of nonradiative recombination in Cs3Sb2Br9 IPQDs. To gain more insight into

⎛ −t ⎞ ⎛ −t ⎞ A(t ) = A1 exp⎜ ⎟ + A 2 exp⎜ ⎟ ⎝ τ2 ⎠ ⎝ τ1 ⎠

(2)

Above outstanding PL superiorities, in particular, the 46% blue PLQY and 41 nm linewidth, brought one unavoidable question of why their PLQYs were so high in reference to the core−shell QD structure of CdSe.32−34 It was widely recognized that surface atoms played a vital role in determining the physical properties of colloidal QDs.35,36 Thus, the surface states of Cs3Sb2Br9 IPQDs were explored to figure out the underlying origins. For the Cs3Sb2Br9 structure formula, it has a Br/Sb molar ratio of ∼4.5. The energy-dispersive X-ray spectroscopy (EDS) measurement showed that Cs3Sb2Br9 9297

DOI: 10.1021/acsnano.7b04683 ACS Nano 2017, 11, 9294−9302

Article

ACS Nano

Figure 5. (a) Temperature-dependent PL spectra of the Cs3Sb2Br9 IPQDs. (b) Pseudocolor maps of temperature-dependent PL spectra. (c) Integrated PL intensity and (d) fwhm of the Cs3Sb2Br9 IPQDs as a function of reciprocal temperature from 80 to 300 K.

treatment, whereas it cannot remove inorganic Sb-perovskite. As shown in Figure S2a, the PL intensity had little change with UV-ozone treatment alone. Thus, it manifested that the surface organic ligand was not crucial for high PLQY. In comparison, the PL of the IPQD film sharply decreased after Ar+ etching (Figure S2b). As Ar+ treatment could etch the Br-rich surface, the high PLQY was mainly ascribed to the loss of surface-rich Br atoms. Thus, the surface-rich halogen played a crucial role in high PLQY, which provided a quantum-well band structure to suppress the nonradiative recombination. In addition to the quantum-well band structure of Cs3Sb2Br9 IPQD,40−42 their exciton characteristic was another key parameter for PLQY value. In order to further verify the mechanism of PLQY enhancement, temperature-dependent PL measurement was carried out with the temperature ranging from 300 to 80 K. As shown in Figure 5a, only one emission peak can be resolved with the decrease of temperature, indicating the absence of structural phase transition. A corresponding pseudocolor map of temperature-dependent PL spectra is shown in Figure 5b. At 80 K, the fwhm of the PL peak was as narrow as 20 nm, which presented a typical strong excitonic character with little defect effect. The strong excitonic emission behavior can be verified by performing the powerdependent PL measurement. As shown in Figure S3, a power law dependence (IPL = IEXβ, where β denotes the nonlinear component) had been observed, and the extracted β value was 1.31. It is noted that 1 < β < 2 accounts for the recombination of free excitons and bound excitons, which are associated with the long- and short-lived PL lifetimes.43 With temperature increasing, the PL peaks showed a little blue shift and became weaker (Figure 5a,b). Such blue shift was previously observed from CH3NH3PbX3 and PbS QDs.44 The decreased PL intensities (Figure 5c) could be assigned to the thermally activated nonradiative recombination process.45 In addition, their exciton binding energy (Eb) and exciton−phonon interaction could also be estimated according to temperature-

single crystal held a Br/Sb molar ratio of 4.5 (Table S4), consistent with the stoichiometry ratio. In Figure 4a and Figure S4, the as-synthesized IPQDs had a Br/Sb molar ratio of 5.08, much higher than their stoichiometric ratio. Thus, the obtained IPQDs showed Br-rich composition. In order to clarify the rich Br distribution, we adopted Ar+ etching by combining with X-ray photoelectron spectroscopy (XPS) to probe it. Before Ar+ etching, the typical chemical states for Cs 3d, Br 3d, Sb 3d, and Sb 4d ae shown in Figure 4b−e, respectively. The Br/Sb atomic ratio was approximately 5.05, similar to EDS results. After Ar+ etching for 1 min, XPS spectra evolved into Figure 4b. The main peaks of Cs 3d had not changed, indicating the low bonding interactions of Cs with Br ions either on the surface or in the crystal lattice. On the contrary, the presence of the Sb shoulder peak in the lower binding energy of Sb 3d was obtained, which indicated the deficit of Sb ions on the surface before etching (Figure 4d). The Br 3d peaks kept constant with two peaks located at 68.3 and 69.5 eV corresponding to the inner and surface ions, respectively. According to above XPS analysis, the synthesized Cs3Sb2Br9 IPQDs showed a Br-rich surface, and their band structure was schematically described by Figure 4f. The quasiquantum-well band structure may favor the efficient radiative recombination to enhance PLQY.6 It is widely recognized that the ligand capping effect has been extensively investigated to improve the PLQY.32,34,37 The richness of Br and organic ligands on the Cs3Sb2Br9 IPQD surface may play a crucial role in their high PLQYs.24,38,39 For a rich Br effect, the wider band gap surface could introduce a built-in field to avoid the excited electron trapping by surface defects. For organic ligand function, ultrahigh surface area of QDs could be capped by organic ligands to passivate the surface trapping states and prefer coupling with QDs. To distinguish the main surface passivation mechanism, Cs3Sb2Br9 IPQDs were sequentially treated by UV-ozone and Ar+ ion etching (Figure S2). According to a previous report, the organic ligands on the QD surface could be removed by UV-ozone 9298

DOI: 10.1021/acsnano.7b04683 ACS Nano 2017, 11, 9294−9302

Article

ACS Nano dependent PL spectra. The Eb of the Cs3Sb2Br9 IPQDs can be achieved by the following equation: I0 I (T ) = E 1 + A exp − K bT

(

B

)

Finally, we studied the air stability and photostability of Cs3Sb2Br9 IPQDs without encapsulation, as shown in Figure 6b,c. The air stability of the Cs3Sb2Br9 IPQDs was investigated by measuring the PL evolution as storing time varied under ambient condition with a fluorescent light illumination. At the first few days of storage, the perovskite QD emissions were enhanced for all composition QD solutions (Figure S8b). Such PL enhancement is known as the “photoactivation” phenomenon, which originates from the smoothing of Cs3Sb2Br9 IPQDs and the removal of dangling bonds or other surface defects.50 Upon longer storage time in air, the QD solutions exhibited a gradual decrease in PL intensity. This degradation was similar to that of Pb-based perovskite. It was reported that surface defects caused by oxygen exposure were responsible for the decomposition of perovskite QDs and PL degradation. After aging for 35 days without encapsulation, 70% PLQY can be retained. The corresponding fwhms of Cs3Sb2Br9 IPQDs were gradually increased (Figure 6b). In a control experiment, when the QD solutions were stored in air without light illumination, the PL intensity would exhibit a continuous decrease (Figure S8a). The photostability test of Cs3Sb2Br9 IPQD solution was implemented under continuous UV light (365 nm, 6 W) irradiation, as shown in Figure 6c. The solutions were illuminated for 108 h and exhibited a continuous decrease in PL intensity. After 108 h irradiation, the relative intensity of Cs 3 Sb 2 Br 9 IPQDs kept 50% of its initial value. The corresponding fwhms of the Cs3Sb2Br9 IPQDs solution were located within a narrower range, suggesting their high photostability. Compared to Sn-based perovskites (5 min irradiation,