Introduction

Lead halide based perovskites have become famous semiconductor materials for their industrial prospect in solar cells. Up to date, the efficiency of perovskite solar cells has unceasingly boosted up to 22.1%1,2,3,4,5,6,7,8,9. The application potentials of perovskites in light emitting devices10,11,12,13,14,15,16,17 and lasers18,19 were also demonstrated. In nanometer scale, the huge specific surface area and thus the abundant interface or surface states of the perovskites exhibit prominent effects on the electronic and photoelectronic properties. Therefore, the challenge emerges for chemists of nano-science and -technology to develop synthesis methods to achieve size-controllable perovskite nanocrystals. In the most recent year, colloidal cesium lead halide perovskites nanocrystals and nanowires were successfully synthesized. Kovalenko et al. reported the synthesis of highly luminescent perovskite CsPbX3(X = Cl, Br, I) of 4–15 nm, with the photoluminescence tuned within 410−700 nm20,21. Yang et al. developed a solution synthesis of single-crystalline CsPbX3(X = Cl, Br, I) nanowires22. Prato et al. further found that the composition of perovskite nanocrystals can be feasibly tuned by post-synthesis halide anion exchange23. Despite the success for the all-inorganic cesium lead halide perovskite colloidal nanocrystals, the synthesis methods of colloidal organometal halide perovskite nanocrystals are comparably less developed. Although there are numerous literatures on preparing organic-inorganic perovskite micro/nanocrystals, most of them are synthesized or grow on substrates. For examples, the synthesis of MAPbI3 nanoplates and nanowires on substrates were demonstrated by Jin24, Horvath25, Grätzel and Park et al.26. We reported the synthesis of MAPbI3 crystals on porous TiO2 substrate with the size controllable within 40–700 nm27. The substrates served as a scaffold for the precursors were used to control the perovskite growth kinetics27,28,29,30,31. However, the substrates prevented from scale up the products and the generality for different substrates is limited.

Recently, advances have been made in synthesis of MAPbX3(X = Br, I, Cl) nanocrystals in solution without a substrate32,33,34,35,36. All these reported methods use organic ammonium cation with a long alkyl chain, such as the octylammonium bromide (CH3(CH2)7NH3Br) or octadecylammonium bromide (CH3(CH2)17NH3Br). The ammonium cation serves as the surface capping ligands of the nanocrystals, limiting the crystal growth in one, two or three dimensions. In this article, we use a different surface modulation strategy by applying alkyl carboxylate, the lead oleate (Pb(CH3(CH2)7CH = CH(CH2)7COO)2, abbreviated as Pb(OA)2), as both the lead resource and capping ligands. The as synthesized MAPbBrxI3−x(0 ≤ x ≤ 3) nanocrystals exhibit continuously tunable UV-vis absorption and photoluminescence (PL) spectra across the visible realm, which is attributed to the size related quantum confinement effect with a fixed stoichiometry of the halide composition.

Results and Discussion

The reaction solution contains two solvents, the cyclohexane and the isopropanol. The Pb(OA)2 is soluble in the former, and the MAI or MABr (or the mixture) is soluble in the later. When these two solutions are mixed at room temperature, MAPbBrxI3−x(0 ≤ x ≤ 3) forms immediately which can be easily identified by the color change and the photoluminescence under a portable 365 nm ultraviolet lamp (Fig. 1).

Figure 1
figure 1

(a) Images of typical as synthesized MAPbBrxI3−x(0 ≤ x ≤ 3) colloidal nanocrystals in solution under the day light and (b) images under the ultraviolet excitation (365 nm). From left to right: sample P-I-5, P-I3Br2-5, P-I1Br1-5 and P-Br-5.

Full size image

The XRD spectra of the typical products are shown in Fig. 2 (and Figures S1–S4 in Supporting Information). For the samples P-I-2/3/4/5, strong split peaks at 14.0° and 14.1° that respectively corresponding to (002) and (110) crystal plane and the split peaks of the (004) and (220) peaks can be clearly identified. This split feature verified that the as prepared MAPbI3 are tetragonal perovskite phase (space group I4/mcm)37,38,39,40,41. While for the rest of the samples containing Br−, no split can be identified at (001) or (002) peaks, indicating they are cubic phased perovskite (space group Pm-3m). The perovskite phases via the regulation of I− and Br− are generally observed by various researchers38,39,40,41. For all the samples no impurity of MAI or Pb(OA)2 is found. This may due to the fast reaction rate and excess amount of MAI to consume the Pb(OA)2. Also none of the other known and related MAnPbIm (n, m = 2,4, 3,5, 4,6) phases are observed41.

Figure 2
figure 2

The XRD patterns of some typical as synthesized MAPbBrxI3−x(0 ≤ x ≤ 3) nanocrystals: sample P-I-5, P-I3Br2-2, P-I1Br1-2 and P-Br-5.

Full size image

For all the samples of MAPbBrxI3−x(0 ≤ x ≤ 3) nanocrystals, the typical FTIR vibration modes of organolead halide perovskites are distinctly presented (Figure S5). The 3300–3000 cm−1 broad strong peak is assigned to N-H stretching; the 2950–2820 cm−1 peaks are assigned to symmetric and asymmetric stretching vibrations of CH2 and CH336,42,43,44,45; the peak at 1730–1630 cm−1 corresponds to the COO− modes and the peak at 1440–1360 cm−1 is assigned to the C-H bending44. The symmetric O-H stretch and antisymmetric O-H stretch of H2O were not found in 3600–3800 cm−1 from Figure S5. Thus, no adsorbed H2O was detected, suggesting its good temporary stability in ambient air condition. The good dispersability of the perovskites in cyclohexane also suggests that the surface of MAPbBrxI3−x(0 ≤ x ≤ 3) nanocrystals is coordinated by oleate ligand, like typical oleate or alkylamines modified colloidal nanocrystals. The schematic illustration of a perovskite nanocrystal stabilized by oleate as surface ligands is shown in Fig. 3(a). Figure 3(b,c) are typical TEM and HRTEM images of perovskite nanocrystals of sample P-Br-5, respectively. It indicates that these nanocrystals are well dispersed and the measured crystal lattice matches well with the that of the MAPbBr332,36. Else TEM images of colloidal perovskite nanocrystals of MAPbBrxI3−x(0 ≤ x ≤ 3) can be found in Figures S6–S8.

Figure 3
figure 3

(a) Schematic illustration of a perovskite nanocrystals stabilized by oleate as surface ligands, (b) A typical TEM image and (c) HRTEM image of perovskite nanocrystals of sample P-Br-5.

Full size image

The SEM measurement indicates that with the decrement of MA:Pb from 5:1 to 2:1, the size of all the as synthesized MAPbBrxI3−x(0 ≤ x ≤ 3) nanoparticles increases to ~300 nm (Fig. 4). This trend may find its origin in the nucleation and growth mechanism. Take the nucleation and growth of MAPbI3 as an example, the reaction in the solution is as follows:

Figure 4
figure 4

SEM images of four sets of samples (a–d) P-I-2/3/4/5 from left to right respectively, and the same for (e–h) P-I3Br2-2/3/4/5, (i–l) P-I1Br1-2/3/4/5, and (m–p) P-Br-2/3/4/5 (Scale bar: 1 μm).

Full size image

The alkyl carboxylate, lead oleate Pb(OA)2, acts as not only as the lead resource to react with MAI, but also as the capping ligands. In addition, the reaction rate is higher with the higher MAI concentration. The faster the nucleation bursts, the more completely Pb(OA)2 consumed in solution in a short time, leading to smaller MAPbI3 nanocrystals. While in relatively lower MAI concentration, the nucleation rate is slower and the perovskite growth is dominated following the nucleation, resulting bigger MAPbI3 nanocrystals. Obviously, it shows significance for all these perovskite materials.

The UV-vis absorption spectra and corresponding PL spectra of four sets of samples (P-I-2/3/4/5, P-I3Br2–2/3/4/5, P-I1Br1-2/3/4/5, P-Br-2/3/4/5) are shown in Fig. 5. The band edge absorption peak locates at 764 nm for big MAPbI3 nanocrystals (P-I-2, ~300 nm by SEM). It undergoes a blue shift to 735 nm for smaller MAPbI3 nanocrystals (P-I-5, ~5 nm by TEM). Accordingly, the photoluminescence (PL) peak shifts from 767 nm to 747 nm with the size decreasing. The synthesized MAPbBr3 and MAPbBrxI3−x(0 ≤ x ≤ 3) nanocrystals also show similar UV-vis absorption and PL spectra dependent on the size. This size related monotonic blue shift of both spectra can be induced by the intrinsic quantum confinement effect, for the small crystal size is comparable to the Bohr diameter of MAPbBrxI3−x(0 ≤ x ≤ 3)46,47,48,49,50. There remains only one concern, that since the spectra of P-I3Br2-2/3/4/5 and P-I1Br1-2/3/4/5 are over lapped, it is doubtful if the spectrum shift is caused by the component deviation. By comparing the XRD peaks of P-I3Br2-2/3/4/5 and P-I1Br1-2/3/4/5, this possibility is excluded. As shown in Fig. 6, the (002) and (201) peaks of each set are separated from each other, while almost fixed for a same set. It suggests that the perovskites of the same sets have the same crystal parameter, thus have the same component. Also, the widening of the XRD peaks indicates the size of nanocrystals decreases (Figures S1–S3). These features further support the intrinsic quantum confinement effect for the observed spectra shift.

Figure 5
figure 5

The UV-vis absorption spectra and PL spectra of four sets of samples: (a,b) P-I-2/3/4/5, (c,d) P-I3Br2-2/3/4/5, (e,f) P-I1Br1-2/3/4/5, and (g,h) P-Br-2/3/4/5. The arrows indicate the spectrum evolution of the samples (Excitation: 365 nm).

Full size image
Figure 6
figure 6

The XRD peaks of (002) and (201) for P-I3Br2-2/3/4/5 and P-I1Br1-2/3/4/5.

Full size image

The PL lifetimes of MAPbBrxI3−x(0 ≤ x ≤ 3) were measured (Figures S9–11) to get insight into the exciton recombination dynamics. These PL decays can be well fitted by the bi-exponential function (Figure S12):

where τ1 and τ2 are the fitted decay lifetimes; A1 and A2 are the weighting parameters. The average lifetime τave is calculated by equation (3):

The results are shown in Table 1. It suggests two dynamics, a fast decay (τ1) and a longer-lived component (τ2). With the decreasing size of the nanocrystal or the blue shift of the PL peak, the τave increases for samples of P-I-2/3/4/5, P-I1Br1-2/3/4/5 and P-Br-2/3/4/5. For sample P-I-2, it has a very short lifetime τave of 3.0 ns. The lifetime of the sample P-I-5 is incredibly prolonged to 166.2 ns. This lifetime range covers the ever reported lifetimes of tetragonal MAPbI3 in the form of nanowire, nanorod, film or bulk (Table 2). A very short lifetime of P-I-2 indicates that there exist high effective nonradiative recombination channels in the material50,51,52. It is proposed to be chemical or structural defects in the MAPbI3 nanocrystals. The long lifetime of the sample P-I-5 is unexpected, because it has a big specific surface area which may induce vast surface states. Possibly it suggests that the surface of MAPbI3 nanocrystals is well passivated by the oleate ligands53,54,55. It was noticed that from P-I-2 to P-I-5, both τ1 and τ2 increase drastically. The fast decay τ1 of P-I-5 is 25.5 ns, a value much bigger than the slow decay τ2 of P-I-2 4.8 ns. The mechanisms how the τ1 and τ2 are separately or cohesively tuned are presently unclear. However, it is not likely to originate from quantum confinement effect, but more possible related to the surface or defect states and charge carrier delocalization in MAPbI320. For the samples from P-I1Br1-2 to P-I1Br1-5, the lifetime τave boosts from 1.1 to 42.1 ns. While for the samples from P-Br-2 to P-Br-5, the lifetime τave only increases from 13.0 to 17.1 ns, with the small PL peak shift from 532 to 525 nm. Compared with MAPbBr3 single crystal or film (Table 2), these as prepared MAPbBr3 nanocrystals exhibit shorter lifetime, suggesting more nonradiative recombination effect involved13,56,57. Unfortunately, there was no lifetime data of MAPbBrxI3−x(0 ≤ x ≤ 3) quantum dots synthesized by using ammonium cation with a long alkyl chain as surface capping ligands35,36,37,38,39. Further investigation on how surface ligands type would affect the fluorescence lifetime may give us a deeper understanding of the exciton recombination dynamics, and thus better control of the PL for future use in construction of devices.

Table 1 The fitting decay lifetimes of τ1 and τ2, the average lifetime τave and corresponding PL peak.
Full size table
Table 2 Reported decay lifetimes and corresponding PL peaks for MAPbI3 and MAPbBr3, T = tetragonal, C = cubic.
Full size table

Conclusions

Organometal halide perovskite MAPbBrxI3−x(0 ≤ x ≤ 3) colloidal nanocrystals were synthesized by mixing Pb(OA)2 cyclohexane solution and MAX(X = I or/and Br) isopropanol solution. The size of the perovskites was successfully controlled by the oleate as the ligands, and also by the Pb:MA ratio. The UV-vis absorption and PL spectra show a blue shift as the size decreasing monotonically, which is ascribed for quantum confinement effect. Significantly, these colloidal organometal halide perovskite nanocrystals were dispersed well in nonpolar organic solvent, e.g. toluene or cyclohexane. This would be useful for fabricating perovskite thin films in various substrates such as silicon, polymer or glass. In addition to the method could be generally applied to synthesize other organometal halide perovskite materials, this study would bring chances to the design and fabrication of new photovoltaic and electronic devices.

Materials and Methods

Materials

Lead acetate trihydrate (AR), sodium oleate (>99.5%) and cyclohexane (>99.5%) were bought from Sinopharm. Isopropanol (anhydrous, 99.5%) was purchased from J&K. Hydroiodic acid (HI, 55–58%), hydrobromic acid (HBr, 47.0%) were purchased from Sigma-Aldrich. Methylamine (40% in methanol) was bought from TCI. All chemicals were used as received unless specified otherwise.

Synthesis of Pb(OA)2 (OA = Oleic acid)

The lead acetate solution (1.037 g in 5 mL H2O) was added into the solution of sodium oleate (1.90 g in 25 mL mixture of ethanol and H2O v:v = 1:1) with vigorous stirring. The precipitate (Pb(OA)2) was separated by centrifugation, washed with water and finally dried in oven at 85 °C overnight.

Synthesis of MAX (MA = methylammonium, X = I, Br)

Hydroiodic acid (10 mL, 0.075 mol) or hydrobromic acid (8.6 mL, 0.075 mol) was added to a solution of excess methylamine (24 mL, 0.192 mol) dropwisely at 0 °C under stirring. The mixture was continuously stirred in the ice bath for 2 hrs. The raw product (MAI or MABr) was obtained after evaporation and dried under dynamic vacuum at 60 °C for 12 hrs. The purified MAI or MABr was obtained by recrystallization from a mixed solvent of diethyl ether and ethanol.

Synthesis of MAPbBrxI3−x(0 ≤ x ≤ 3) nanocrystals

Typically, Pb(OA)2 (15.4 mg, 0.02 mmol) was dissolved in cyclohexane (20 mL) to form a solution A. MAI (6.4 mg, 0.04 mmol) was dissolved in isopropanol (20 mL) to make a solution B. Solution A was injected into to solution B with volume ratio of 1:1. The perovskite MAPbI3 nanocrystals can be isolated by centrifugation (3 min at 14000 rpm) and redispersed in solvent of cyclohexane or toluene. The MAPbBrxI3−x(0 ≤ x ≤ 3) nanocrystals was synthesized similarly, by introducing MABr to the reaction. The samples are numbered in the form of P-I-y, P-IaBrb-y, or P-Br-y, where P = perovskite; a and b are the I:Br ratio of a:b such as I1Br1 or I2Br3; y = 2, 3, 4 or 5 means the MA:Pb ratio of y:1 (Table S1). “P-I-2/3/4/5” means four samples of different y values: P-I-2, P-I-3, P-I-4 and P-I-5. To eliminate the effects from the solvent, the volume ratio of cyclohexane to isopropanol is controlled as a constant of 1:1 for all the reactions in the context. The Pb(OA)2 is kept as a constant. The concentration of MA in isopropanol is investigated as a variable.

Instruments

The powder X-ray Diffraction (XRD) spectra were performed on a Bruker D8 Advance instrument with a Cu Kα radiation (λ = 1.5418 Å). The scanning electron microscopy (SEM) images were obtained on the S-4800 SEM at 20 kV. The transmission electron microscopy (TEM) images were measured on a HT7700 (Hitachi) TEM at an acceleration voltage of 100 kV. The FTIR spectra were recorded on a Vector 22 spectrometer with a resolution of 2 cm−1 by using KBr pellets. The absorption spectra were obtained by a Shimadzu UV-2700 UV/Vis spectrophotometer. The PL spectra were collected with a Hitachi F-4600 fluorescence spectrophotometer. PL lifetimes were measured with a Zolix Omini-λ 300 fluorescence spectrophotometer and a picosecond pulsed diode laser (Edinburgh Instruments Ltd.) operating at 379.2 nm.

Additional Information

How to cite this article: Zhao, Y. et al. Colloidal Organometal Halide Perovskite (MAPbBrxI3−x, 0≤x≤3) Quantum Dots: Controllable Synthesis and Tunable Photoluminescence. Sci. Rep. 6, 35931; doi: 10.1038/srep35931 (2016).