Investigation on Photoluminescence of CsPb(Br,I)₃ Perovskite Nanocrystals by Comparison between Hot-Injection and Ion-Exchange Methods

Cs₂CO₃ (99.99%, Mitsuwa Pure Chemical), PbBr₂ (99%, Mitsuwa Pure Chemical), and PbI₂ (99%, Sigma-Aldrich) powders were used as received without further purification. In addition, 1-octadecene (ODE; >90.0%, Tokyo Chemical Industry), oleic acid (OA; >85.0%, Tokyo Chemical Industry), oleylamine (OAm; 80%–90%, Acros Organics), tert-butyl alcohol (>99.0%, Tokyo Chemical Industry), methyl acetate (99.5%, Kanto Chemical), and toluene (99.5%, Kanto Chemical) were dehydrated over molecular sieves (3 A 1/8, Wako Pure Chemical Industries) prior to use.

Cs₂CO₃ (1.25 mmol), ODE (20.0 ml), and OA (1.3 ml) were mixed and degassed for 1 h at 120 °C and then purged with Ar gas. A Cs-oleate precursor solution was obtained by heating the mixture to 150 °C. The resulting solution was preheated at 100 °C before being injected into the reaction flask.

The hot-injection method by Kovalenko's group is the most famous synthetic method for CsPb(Br_1−x I_x )₃ NCs and has been well reported in the literature. ^15,16 ODE (5.0 ml), 0.376 (1–x) mmol PbBr₂, and 0.376 x mmol PbI₂ (x = 0, 0.30, 0.33, 0.40, 0.45, 0.50, 0.55, 0.60, 0.80, and 1.00) were loaded into a four-neck flask and degassed for 1 h at 120 °C and then purged with Ar gas. OA (1.0 ml) and OAm (1.0 ml) were injected into the flask, and after completely dissolving the salt mixture, the temperature was raised to 180 °C. The Cs-oleate precursor solution (0.8 ml) was swiftly injected into the reaction flask and then cooled in an ice water bath after 5 s. Crystallized nanoparticles were precipitated by adding tert-butyl alcohol (25.0 ml). Alternatively, methyl acetate (25.0 ml) was used when x = 1.00. The NCs were collected by centrifugation at ∼11,000 g (10,000 rpm using a 10-cm-diameter rotor) for 5 min and then redispersed in toluene. The purified NC dispersion was used for optical analyses. Subsequently, the NC powder used for X-ray analyses was prepared by drying the collected NCs under vacuum overnight.

Crude dispersions of CsPbB_r3 NC and CsPbI₃ NC prepared by the hot-injection method were used to prepare mixed-halide NCs through halide exchange. PbBr₂ or PbI₂ (1.00 mmol) was dissolved in a mixture of OA (2.0 ml), OAm (2.0 ml), and toluene (6.0 ml) under stirring for 1 h at 80 °C. The desired volume of the obtained solution at room temperature was used to adjust the halide ratio of NCs, as summarized in Table I. After cooling in an ice water bath, the halide solution was added to the crude dispersion (2.0 ml) with vigorous stirring at room temperature. Herein, PbBr₂ and PbI₂ solutions were added to the crude dispersions of CsPbI₃ NC and CsPbBr₃ NC, respectively. After 10 min, ion-exchanged NCs were precipitated by adding tert-butyl alcohol (25.0 ml). Purified toluene dispersions and NC powders were prepared for analyses in the same way as described above.

The elemental ratios of the NC powders were measured using an X-ray fluorescence (XRF) analyzer (ZSXmini II, Rigaku). Elemental compositions were determined by the fundamental parameter method. Powder X-ray diffraction (XRD) profiles were obtained using an X-ray diffractometer (Rint-2200, Rigaku) with a Cu Kα radiation source and monochromator. The NC morphologies were observed by field-emission transmission electron microscopy (TEM; Tecnai G², FEI). The TEM samples were prepared by vacuum drying a drop of NC dispersion on carbon-reinforced collodion-coated copper grids (COL-C10, Oken Shoji) overnight. UV–vis absorption spectra of the NC dispersions were measured using an optical absorption spectrometer (V-570, JASCO). The bandgaps of NCs were estimated from Tauc plots derived from the UV–vis spectra. ¹⁷ The PL spectra of the NC dispersions were measured using a fluorescence spectrometer (FP-6500, JASCO). Each spectral response was calibrated using an ethylene glycol solution of rhodamine B (5.5 g l⁻¹) and a standard light source (ESC-333, JASCO). PL decay curves were recorded on a fluorescence lifetime spectrometer (Quantaurus-Tau C11367, Hamamatsu Photonics) using 405 nm LED. Herein, the PL decay curves were fitted with the following biexponential equation:

$$\begin{eqnarray}&&f\left(t\right)={A}_{1}\exp \left(-\displaystyle \frac{t}{{\tau }_{1}}\right)+{A}_{2}\exp \left(-\displaystyle \frac{t}{{\tau }_{2}}\right),\end{eqnarray} \tag{ 1 }$$

where A₁ and A₂ are the amplitudes, and τ₁ and τ₂ are the PL decay times. The average PL decay time 〈τ〉 was calculated using the following equation.

$$\begin{eqnarray}&&\tau =\displaystyle \frac{{A}_{1}{{\tau }_{1}}^{2}+{A}_{2}{{\tau }_{2}}^{2}}{{A}_{1}{\tau }_{1}+{A}_{2}{\tau }_{2}}\end{eqnarray} \tag{ 2 }$$

Absolute PLQYs were measured using a quantum efficiency measurement system (QE-2000–311C, Otsuka Electronics).

The emission color of prepared CsPb(Br_1−x I_x )₃ NCs changed from green to red with increasing x, as shown in Fig. 1. Their emission colors revealed that the mixed halides were successfully obtained. XRF results in Fig. 2 show a monotonic increase in actual iodide ratio with the nominal ratio. It should be noted that the measured ratios were lower than the nominal ratios in the Br-rich region regardless of preparation methods. This may be a result of errors derived from matrix effects in the fundamental parameter method of XRF.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Judging from the XRD profiles, the hot-injection method produced single-phase cubic CsPb(Br_1−x I_x )₃ NCs. All of the peaks shifted to lower values of 2θ as the iodide ratio increased, indicating substitution of I⁻ for Br⁻ in CsPb(Br_1−x I_x )₃. The coordination number of the halide ion in cubic CsPb(Br_1−x I_x )₃ is six. ¹⁸ Given that Shannon's ionic radii of Br⁻ and I⁻ are 1.96 Å and 2.20 Å, respectively, ¹⁹ the lattice constant of cubic CsPbI₃ (12.660 Å) is therefore larger than that of cubic CsPbBr₃ (11.880 Å). ²⁰ On the contrary, the NCs prepared by the ion-exchange methods contained a small amount of orthorhombic phase at higher iodide ratio, x. The cubic phase of CsPbI₃ transforms to the more stable orthorhombic phase at room temperature. ²¹ Accordingly, the partial transition of cubic phase to orthorhombic phase at higher x might be attributed to the higher stability of the latter phase. ²²

Judging from TEM images (Fig. 3), the average primary sizes of the prepared NCs were ∼10 nm. Most of the observed CsPbBr₃ NCs and CsPbI₃ NCs were smaller than their calculated exciton Bohr diameters of 7 nm and 12 nm, respectively. ¹⁵ The primary nanoparticles tended to become larger with increasing x because of increased lattice constant.

Zoom In Zoom Out Reset image size

UV–vis absorption spectra and PL spectra are shown in Fig. 4. The absorption edge and emission peak position were shifted to longer wavelength and lower energy, respectively, by decreased bandgap because of the increased iodide ratio. ¹⁰ Bandgaps estimated from the UV–vis spectra and PL peak positions are summarized in Figs. 5A, 5B. Their behaviors overlapped well, regardless of the preparation methods, and no linear relationship was observed. This might be attributed to complex factors determining the bandgap, such as particle size and exciton Bohr diameter.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figure 5C shows the change in FWHM of the PL peak against the iodide ratio. The FWHM for NCs prepared by hot injection achieved a maximum at x = 0.55. In other words, the PL peak width of mixed-halide NCs at x = 0.55 was broader than those of CsPbBr₃ and CsPbI₃. This phenomenon might be attributed to the compositional inhomogeneity among the individual CsPb(Br_1−x I_x )₃ NCs. ⁸ However, the maximum occurred at larger x for NCs prepared by the ion-exchange methods. As mentioned above, a small amount of orthorhombic phase was observed for the NCs at higher x. Partial transition of orthorhombic phase from cubic phase for iodide-rich CsPb(Br_1−x I_x )₃ NCs might result in increased FWHM at higher x. PL decay times estimated from the PL decay curves shown in Fig. 6 were summarized in Table II. PL decay times of the CsPb(Br_1−x I_x )₃ NCs prepared at x = 0.50 were between those of CsPbBr₃ NCs and CsPbI₃ NCs regardless of the preparation method.

Zoom In Zoom Out Reset image size

Change in PLQY with x is shown in Fig. 5D. The minimum PLQY for the hot-injection method was 25.0% at x = 0.50, which was smaller than PLQYs of CsPbBr₃ (70%) and CsPbI₃ (71.0%). The corresponding x value for the minimum PLQY was close to that for the maximum FWHM of the PL peak. The decreased PLQY of mixed-halide NCs could be attributed to lattice distortion induced by the difference in the ion radii of Br⁻ and I⁻. ⁸ In contrast, the minimum PLQYs for halide exchanges from CsPbBr₃ and CsPbI₃ were likely to be located at x = ∼0.3, as described by the dashed lines in Fig. 5D. Furthermore, the minimum PLQYs for both halide exchanges were ∼10%, which was lower than 25.0% observed for the hot-injection method. PLQY is readily affected by internal defects in the NC and surface defects that cause nonradiative relaxation. The addition of a halide solution for ion-exchange might affect the adsorption state of OA and OAm surface modifiers, leading to the decreased PLQY.