Citation

Jinwen Hu, Chunyi Zhao, Sheng He, Wenming Tian, Ce Hao, Shengye Jin. Carrier dynamics in CsPbI₃ perovskite microcrystals synthesized in solution phase[J]. Chin. Chem. Lett., 2018, 29(5): 699-702  

Jinwen Hu^a,b, Chunyi Zhao^b, Sheng He^b, Wenming Tian^b, Ce Hao^a, Shengye Jin^b    

Organic-inorganic metal halide perovskites (CH₃NH₃PbX₃, X = I^-, Cl^-, Br^-) have received great attentions due to excellent optoelectronic properties such as high charge mobility, long carrier diffusion distance, direct-bandgap semiconductor, tunable bandgap, strong light harvesting ability [1]. These properties have led to their potential applications in the photovoltaic solar cells, light emitting diodes, photodetectors, and lasers [2-17]. Especially, high-efficiency perovskite solar cells have reached power conversion efficiency (PCE) over 22% in 2016 [18]. Despite the exciting progress, the organic-inorganic metal halide perovskite materials still face a huge challenge in practical applications, for that the family of organic-inorganic halide perovskite compounds are unstable, volatile, and hygroscopic in ambient air and are sensitive to the processing conditions on account of their intrinsic structural and thermal instability [19, 20]. To expand the practical applications of perovskite materials, the fabrication of more stable perovskites in the atmospheric environment should be explored.

One strategy to make stable perovskites is to replace the organic cation (CH₃NH₃⁺) with inorganic metal ions cesium (Cs₊) having similar size to organic component, forming all-inorganic cesium lead halide perovskites CsPbX₃ (X = Cl^-, Br^-, I^-) [21]. Compared to the organic-inorganic perovskite CH₃NH₃PbX₃, the all-inorganic perovskites CsPbX₃ show much enhanced thermal and moisture stabilities, without obviously lowering the excellent opticalelectric properties [21-25]. Remarkably, solar cells of all-inorganic perovskites CsPbX₃ have been fabricated successfully with increased device stability while preserving comparable PCE versus organic-inorganic ones [21-23]. Meanwhile, highly luminescent colloidal quantum dot (QD) CsPbX₃ (X = Cl^-, Br^-, and I^- or their mixture) have been synthesized through different methods [16, 26-30]. Despite of these works, we noticed that the synthesis of CsPbI₃ perovskites through facile solution process has encountered a great challenge. It has been known that the perovskite precursors for CsPbI₃ more inclined to grow into yellow orthorhombic non-perovskite phase, rather than black cubic perovskite phase due to inherent structural property under full atmospheric environment. For this reason, CsPbI₃ perovskites were normally fabricated through halide anion exchange reaction in either solution or vapor phases [31-34].

Herein, we reported the synthesis of CsPbI₃ microcrystals via a facile solution process. The elemental composition and crystal structure of the perovskites were confirmed by EDS and XRD measurements. By using PL image microscopy, we also determined the carrier lifetimes of ^·200 ns to 1300 ns, the diffusion coefficient of 0.6-1.2 cm²/s and carrier diffusion length of 4-10 μm in different CsPbI₃ microcrystals.

All chemicals and reagents for the synthesis of CsPbI₃ microcrystals were purchased from Sigma-Aldrich and used as received unless noted otherwise. To grow CsPbI₃ microcrystals, the lead precursor coated substrate was first prepared by dissolving 200 mg PbAc₂·3H₂O/PbCl₂ (the molar ratio of 1:1) in 1 mL anhydrous dimethyl sulfoxide. The lead precursor spins coating evenly on glass substrates. The film was dried at 55 ℃ to remove all the solvent, then cooling naturally before submerged into the beaker. The sealed beaker was kept at room temperature for 48 h, and then the substrate was removed and washed in anhydrous isopropanol for 20 s. After washing, the sample was dried by a N₂ stream immediately. All experimental procedures were carried out under ambient conditions. Inspired by Zhu et al. [15] we provide a solution growth process to synthesize CsPbI₃ microcrystals (with a lateral dimension of a few mircometers) by dipping a PbAc₂/PbCl₂ thin film into a saturated CsI^-isopropanol solution at roomtemperature. The experimental details are in the Supporting information.

From the optical image of as-grown CSPbI₃ microcrystals on a glass coverslip (Fig. 1a), a wide range of microcrystal shapes including wires, plates and cubes can be found. The lateral dimension of these microcrystals varies from several to tens of micrometers. The emission spectrum of the CsPbI₃ microcrystals (Fig. 1b) shows an emission peak centered at 720 nm, corresponding to a bandgap of 1.72 eV. This value is consistent with those of CsPbI₃ colloidal QDs and nanowires and nanoplates synthesized via halide exchange reaction. Unfortunately, we are unable to determine the PL quantum yield due to the limitation of the sample dimension.

Figure 1

	Download: JPG larger image
Fig. 1. (a) The optical image and (b) the emission spectrum of CsPbI3 microcrystals grown on a glass coverslip. (c) The X-ray diffraction (XRD) patterns of the as-grown samples on the FTO glass. Peaks marked by black dots are signals from FTO substrate. Peaks marked by black stars represent the diffraction patterns from the cubic phase CsPbI3 perovskites. Scale bar is 20 μm.

Fig. 1c shows the X-ray diffraction (XRD) patterns of the asgrown CsPbI₃ materials on FTO substrate. The XRD pattern shows strong diffraction peaks at 12.95°, 21.67°, 22.67°, 31.31°, and 39.36°, which can be assigned to the (012), (014), (112), (016), and (027) faces respectively of the undesired nonperovskite orthorhombic crystal. However, there are a few minor peaks (Fig. 1c, peaks indicated by black stars) from the crystal structure of desired black α-CsPbI₃. The characteristic peaks at 14.30°, 20.54°, 28.82°, 32.26°, 35.65°, and 41.17° are ascribed to the (100), (110), (200), (201), (211), and (220) planes of the cubic (Pm-3m) perovskite lattice, respectively. These results indicate that the reaction precursors are more likely to grow into nonperovskite δ-phase compared to desired black perovskite α-phase in ambient air [35-37]. Nevertheless, the presence of XRD patterns of cubic phase CsPbI₃ perovskites, which are identified as the black microcrystals in the optical image (Fig. 1a), confirms the feasibility of using solution process for CsPbI₃ synthesis.

The XRD result indicates that the yield of the CsPbI₃ is still very low in our synthesis. However, the production of the desired black α-CsPbI₃ perovskites implies that CsPbI₃ perovskites should still be able to synthesis even under atmospheric environment, but further optimization is required to improve the production yield. We have found that to improve CsPbI₃ yield, isopropanol used in experiments need to be purified by activated molecular sieve to remove H₂O, and the environment humidity need to be reduced as much as possible.

We further examine the elemental composition of the CsPbI₃ microcrystals using the scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX). Fig. 2a shows a SEM image of a typical CsPbI₃ microcrystal with an ideal rectangular morphology. The EDX mapping of single NP shows uniform distribution of Cs, Pb, and I elements (Figs. 2b-d) with a quantified ratio of ~1:1:3 (Table S1 in Supporting information), which is in good agreement with the stoichiometry, indicating that the singlecrystal has the general structural formula of CsPbI₃.

Figure 2

	Download: JPG larger image
Fig. 2. The SEM image of a typical CsPbI3 microcrystal (a) and EDX mapping of the CsPbI3 microcrystal shows uniform spatial distribution of elements Cs (b), Pb (c), and I (d).

Charge mobility (or diffusion coefficient) and carrier lifetimes are critical photophysical properties in estimating the quality of the perovskite materials. The product of these two parameters determines the carrier diffusion length. In our previous studies, we have successfully determined the carrier diffusion coefficient (D) and carrier lifetime (τ) in CH₃NH₃PbX₃ (X = Br or I) nanowires and nanoplates by using time-resolved and PL-scanned imaging microscopy (the schematic diagram is shown in Fig. S1 in Supporting information) [38-40]. The wavelength of the excitation on sample is 405 nm with 2.5 MHz repetition rate. The excitation laser beam focuses on a corner or an end of the sample through a 100× air objective lens when measuring the diffusion coefficient. The laser intensity at samples can be adjusted by a filter slice. Once the sample is excited, the galvanometer mirror scans the whole sample to collect photons. Each scanning images contains 256 × 256 pixels with the dwell time of ~1 ms at each pixel. The fluorescence signal is collected using a high speed detector (HPM-100-50, Hamamatsu, Japan) with a 660 nm long pass filter and (710 ± 40) nm band pass filter. By coupling with time-correlated single photon counting (TCSPC) model, we can get the PL decay kinetics at any positions in a particle. The diffusion coefficient and the carrier recombination time can be extracted from the fits of the PL kinetics by a diffusion model (Eqs. (1)-(3)).

Applying the above technique, we examined the diffusion coefficient and carrier lifetime in the CsPbI₃ microcrystals. The details about the imaging microscopy are in the experimental section. In Fig. 3a, we show the PL image of a typical CsPbI₃ microcrystal (in a shape of plate) with a focused excitation (at 405 nm, 25.7 mJ/cm²/pulse) locating close to the edge of the particle. The PL image is collected by scanning the PL collection spot over the entire particle. The PL signals from positions other than the excitation spots indicate the presence of carrier diffusion process in the microcrystal. By coupling the time-correlated single photon counting (TCSPC) model, the PL kinetics at any positions in the crystal can be extracted.

Figure 3

Download: JPG larger image

Fig. 3. (a) PL intensity image of a typical CsPbI3 perovskite microcrystal with the excitation locating at a fixed position (solid red circle). The inset is the optical image of the particle where the blue spot indicate the excitation position at 405 nm. The solid white circles mark the positions where the PL kinetics are compared in panel (b). The distances from these positions to the excitation site are labeled beside. (b) PL kinetics extracted at the indicated positions in the CsPbI3 perovskite microcrystal. The black solid lines are the global fits of the kinetics at different positions by Eqs. (1)-(3), yielding the diffusion coefficient and carrier recombination constants of the CsPbI3 microcrystal.

In Fig. 3b, we show the PL kinetics (after subtracting the waveguide component) exacted at positions a few micrometers away to the excitation spot. These kinetics reflect the change of carrier density as a function of time, which can be described by a two-dimensional diffusion model:

(1)

(2)

where ϕ(x, y, t) is the concentration of charge carriers at time t and position (x, y)((x, y) ∈Ω, Ω is the size of the microcrystal); D is the diffusivity; f(ϕ(x, y, t)) is the charge recombination function; k₂ is the radiative electron-hole recombination rate constant and k₁ is defect-induced non-radiative carrier trapping rate constant [38-40]. Further details about the diffusion model are in the Supporting information. This simulation also assumes that the recombination time constants k₁ and k₂ are homogeneously distributed within an individual perovskite crystal. This assumption is reasonable as that we have observed homogeneous distribution of k₁ in CH₃NH₃PbI₃ single crystals in our pervious study [41]. By assuming the same carrier densities for electron and hole, the PL intensity I_PL(x, y, t) is then

(3)

A global fit of the kinetics at different positions by Eqs. (1)-(3) can yield the fitting parameters D, k₁ and k₂. For the typical CsPbI₃ microcrystal shown in Fig. 3, the diffusion coefficient is found to be 0.76 cm²/s. The intrinsic carrier lifetime determined by k₁ is calculated to be 780 ns. The fitting parameters for other selected CsPbI₃ microcrystals are listed in Table S2 in Supporting information.

Statistical analysis from sixteen individual CsPbI₃ microcrystals provides a distribution of diffusion coefficient and intrinsic carrier lifetime (Fig. 4). We find that the D for different microcrystals synthesized under the same experimental condition is consistent, ranging from ^·0.6 cm²/s to 1.2 cm²/s with an averaged value of (0.84± 0.02) cm²/s. The carrier lifetime is found to range from 200 ns to 1300 ns between different microcrystals. Based on the diffusion coefficient, the charge mobility (μ) of electron and hole (when assuming μ_e = μ_h) can then be calculated by Einstein formula μ = D·e/(k_BT), where k_B is the Boltzmann's constant; T is the temperature; e is the electronic charge. It's ready to get the charge mobility μ ranging from 23 cm² V^-1 s^-1 to 41 cm² V^-1 s^-1. With the diffusion coefficient and carrier lifetime, the carrier diffusion length (L_D) can be estimated to be 4 μm to 10 μm by L_D = (Dt)^1/2 for the examined CsPbI₃ microcrystals.

Figure 4

	Download: JPG larger image
Fig. 4. Distribution histograms of the measured (a) diffusion coefficient and (b) lifetime for sixteen individual CsPbI3 microcrystals.

The values of these photophysical parameters are consistent with those of other CsPbI₃ perovskites prepared by halide exchange reactions, and are also similar as in CH₃NH₃PbI₃ single crystals. The carrier lifetime, diffusion coefficient and diffusion length have been believed to be the key parameters that determines that quality and the performance of perovskites when applied in devices. The observed long carrier lifetime (200-1300 ns) indicates a very low defect density in the CsPbI₃ microcrystals. Such long carrier lifetime thus leads to carrier diffusion length up to 10 um in these crystals, indicating that the CsPbI₃ microcrystals synthesized via our solution process retain their high quality features and are also suitable for applications in optoelectronics devices.

In conclusion, we have demonstrated that CsPbI₃ perovskite microcrystals can be directly synthesized by a facile solution method. By probing carrier diffusion and recombination dynamics in different individual microcrystals using time-resolved PLscanned imaging microscopy, we have quantitatively determined the diffusion coefficient, carrier lifetime and diffusion length in the synthesized CsPbI₃ perovskite microcrystals. The values of these important photophysical parameters indicate the high quality of the solution-processed CsPbI₃ perovskite microcrystals, making them ideal candidates for applications in optoelectronic devices.

S. Jin thanks the financial support from the National Natural Science Foundation of China (No. 21473192). C. Hao thanks the financial support from the National Natural Science Foundation of China (Nos. 21373042 and 21677029).

Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.cclet.2017.10.005.