Improving the Stability of α-CsPbI3 Nanocrystals in Extreme Conditions Facilitated by Mn2+ Doping.

The wide application of CsPbI3 nanocrystals (NCs) is limited due to their poor phase stability. We reported that Mn2+-CsPbI3 NCs have better optical performance and phase stability. With a suitable Mn/Pb ratio (5.0%), Mn2+-doped α-CsPbI3 NCs exhibited the best stability under UV irradiation, ethanol addition, and heating. Under UV irradiation and addition of ethanol, photoluminescence (PL) intensities of CsPbI3 NCs could be only preserved up to 35% (22 min UV irradiation) and 10% (ethanol addition), respectively, whereas, Mn2+-doped CsPbI3 (5.0%) exhibited much improved stability, and their intensities could be preserved up to 70% (22 min UV) and 58% (ethanol), respectively. It should be noted that crystal-phase stability could be maintained at least 7 h even at 120 °C. We believe that the improved stability in extreme conditions for α-CsPbI3 NCs can be further applied to optoelectronic devices.

Introduction

All-inorganic cesium lead halide perovskite (ABX3, A site: Cs+, B site: Pb2+, X site: Cl–, Br– or I–) nanocrystals (NCs) have attracted wide attention in the fields of solar cells, light-emitting diodes (LEDs), and lasers due to their unique optical properties, such as high photoluminescence quantum yield (PLQY), high color gamut, and a tunable band gap.[1−11] Generally, their optical properties are directly dependent on their crystal phase.[1,2,12] Therefore, the stability of their crystal phase is quite sensitive to their optical performance, especially for CsPbI3.

Cubic-phase CsPbI3 (α-CsPbI3, Eg = 1.73 eV) having good optoelectronic performance can be easily transformed to an orthorhombic phase (δ-CsPbI3, Eg = 2.25 eV) with poor optoelectronic properties.[13−16] The small-sized Cs+ ions cannot maintain the cubic framework of a corner-sharing PbI64– octahedron, which leads to phase transformation.[1,17,18] To solve this issue, many methods have been used to stabilize α-CsPbI3, such as A,- B-, or X-site doping.[11,19−21] It should be noted that B-site doping can improve the crystal-phase stability of CsPbI3 without any side effects, such as ionic migration, phase separation, or photodecomposition.[20,22] Rogach et al. used SrCl2 as a coprecursor to improve the phase stability of α-CsPbI3 NC solutions after 60 days. Additionally, Sr2+-ion doping and surface Cl–-ion passivation could further enhance their PLQY to 84%.[13] Yao et al. reported novel Sr2+ doping with an I–-ion passivation strategy to stabilize α-CsPbI3 quantum dots. Their PLQY can be maintained above 80% after 60 days and the phase stability of the film is for more than 20 days.[23] Angelis et al. used MnI2 to directly synthesize CsPbMn1–I3 NCs. Their films and solutions are stable for at least 30 days.[14] Nag et al. doped MnI2 via postsynthesis to stabilize α-CsPbI3 NCs, which exhibited good stability in ambient conditions for a month.[24] In summary, B-site doping can improve the optical properties and crystal-phase stability of α-CsPbI3 NCs in ambient conditions, and it has the potential to expand the conditions in extreme conditions, such as ultraviolet (UV), polar solvents, and high temperature.

According to previous studies, MnI2-doped CsPbI3 exhibited improved stability and was mainly attributed to both Mn2+ doping and surface-enriched I–.[23−25] To identify the improvement in stability using only Mn2+ doping, we used manganese acetate (Mn(Ac)2) to eliminate the influence of excess I–. Additionally, MnI2 is expensive and unstable to water and heat, whereas Mn(Ac)2 is relatively much cheaper and quite stable to water and heat. Herein, we used low-cost manganese acetate (Mn(Ac)2) to directly synthesize Mn2+-doped α-CsPbI3 NCs. The as-prepared Mn(Ac)2-doped CsPbI3 was also stable for at least one month (PL decay less than 10%). With increasing Mn/Pb ratios, the PLQY can be up to 98%. More interestingly, their PL and phase stabilities can be greatly improved. When Mn/Pb = 5.0%, Mn2+-doped α-CsPbI3 NCs exhibited the best stability under UV irradiation, ethanol addition, and heating. We believe that the improved stability in extreme conditions for α-CsPbI3 NCs can be further applied to optoelectronic devices.

Results and Discussion

Optical Properties

Mn2+-doped CsPbI3 NCs were synthesized by a hot injection method (see the Section and Figure ).

Figure 1

Illustration of the synthesizing procedure of Mn2+-doped CsPbI3 perovskite NCs.

Their UV–vis absorption spectra (Figure a) showed less effect after Mn2+ doping, which implied that Mn2+ cannot influence the CsPbI3 band gap.[11,26,27] More interestingly, the peak positions of PL show blue shifts from 685 to 679 nm (Figure b and Table S1). Because of the small size of Mn2+ partly replacing Pb2+, a possible lattice shrinking of CsPbI3 NCs leads to a blue shift of their PL peaks. According to the previous studies, using a small-sized ion to replace Pb2+ may lead to lattice contraction and a decrease in size.[23,28] The size of Mn2+ is 0.067 nm, which is much smaller than the size of Pb2+ (0.119 nm). As a result, Mn2+ doping in CsPbI3 NCs leads to a blue shift of their PL spectra due to lattice contraction and size shrinking (Figure S1). To quantitively compare the PL performance of these samples, absolute PLQYs of samples were measured. The results showed that with an increase in the Mn/Pb ratio, PLQY increased from 90 to 98% (Figure c). Furthermore, PL lifetimes were measured by time-resolved PL decay curves (Figure d). The PL decay curves were fitted with a biexponential function (Table ) to calculate the lifetimes of the fast components (corresponding to the nonradiative recombination/τnr), slow components (radiative recombination/τr), the radiation decay rate (kr), and the nonradiation decay rate (knr).[28−30] Compared with CsPbI3 NCs, the average PL lifetimes of Mn2+-doped CsPbI3 NCs increased from 10.18 to 15.12 ns (Table ). Both kr and knr decreased, but knr decreased faster than kr (Table ). These results demonstrate that Mn2+ doping can effectively decrease defects and enhance PLQY.

Figure 2

(a) UV–visible absorption spectra of Mn2+-doped CsPbI3 NCs. (b) PL spectra of Mn2+-doped CsPbI3 NCs. (c) Mn/Pb ratio-dependent PLQYs. The inset shows images of the corresponding samples under UV-light (365 nm) irradiation. (d) PL decays curves for Mn2+-doped CsPbI3 NCs (Mn/Pb = 0, 0.8, 3.1, and 5.0%).

Table 1

Time-Resolved PL Decays of Samples with Mn2+-Doped CsPbI3 NCs (Mn/Pb = 0, 0.8, 3.1, and 5.0%)

	Mn/Pb = 0%	Mn/Pb = 0.8%	Mn/Pb = 3.1%	Mn/Pb = 5.0%
τ1 (ns)	4.27	1.76	1.24	0.86
τ2 (ns)	11.40	11.75	11.65	15.20
A1	0.33 (18%)	0.21 (4%)	0.11 (1%)	0.08 (1%)
A2	0.60 (82%)	0.68 (96%)	0.88 (99%)	0.90 (99%)
τavg (ns)	10.18	11.30	11.50	15.12

Table 2

PL Average Lifetimes (τavg), Radiative Decay Rates (kr), PLQYs, and Nonradiative Decay Rates (knr) of Mn2+-Doped CsPbI3 NCs (Mn/Pb = 0, 0.8, 3.1, and 5.0%)

	Mn/Pb = 0%	Mn/Pb = 0.8%	Mn/Pb = 3.1%	Mn/Pb = 5.0%
τavg (ns)	10.18	11.30	11.50	15.12
PLQY (%)	90	93	95	98
τr (ns)	11.31	12.11	12.15	15.43
kr (×10–2 ns–1)	8.8	8.3	8.2	6.5
τnr (ns)	101.81	161.43	230.00	756.00
knr (×10–3 ns–1)	9.82	6.19	4.35	1.32
kr/knr	8.86	13.41	18.85	49.24

Structural Characterization

According to PL spectra, Mn2+ doping in CsPbI3 NCs can be indirectly demonstrated. To directly confirm Mn2+ doping, structural and elemental characterization studies were performed via X-ray diffraction (XRD) and transmission electron microscopy (TEM). In Figure a, two main diffraction peaks located at 14.00 and 28.59° correspond to the (100) and (200) planes of α-CsPbI3 perovskite (PDF # 97-018-1288) NCs. With an increase in the Mn/Pb ratio, both main diffraction peaks slightly shift to higher angles, especially at 28.59° (Figure b). This phenomenon is mainly due to the lattice contraction, which is attributed to the partial substitution of Pb2+ by the small-sized Mn2+.[11] In addition to the demonstration of Mn2+ doping by XRD data at a macroscopic level, TEM was also used to confirm such doping at a microscopic level. As shown in Figure S1, the average sizes of CsPbI3 NCs and Mn2+-doped CsPbI3 NCs (Mn/Pb = 0, 0.8, 3.1, and 5.0%) are 11.39 ± 0.90, 10.80 ± 0.68, 10.20 ± 0.73, and 9.70 ± 1.4 nm, respectively. The small-sized NCs after Mn2+ doping mainly result in lattice contraction and the dynamics of nucleation and growth process.[23,24,28] Additionally, high-resolution TEM (HRTEM) images were used to identify their precise lattice distances. According to the insets in Figure c, the lattice distance of CsPbI3 NCs is 0.310 nm, corresponding to the (200) plane of α-CsPbI3. After Mn2+ doping, the lattice distance of the (200) plane decreases to 0.302 nm, which is attributed to the small-sized Mn2+ substituting Pb2+. According to the XRD test results and Scherer’s formula, the crystal plane spacings of (200) are 0.311 nm (0%), 0.310 nm (0.8%), 0.310 nm (3.1%), and 0.309 nm (5.0%). The tendency of lattice contraction can be confirmed by both XRD and HRTEM results, though they showed measuring errors for both XRD and HRTEM. Furthermore, elemental mapping images and element contents were measured using an energy-dispersive spectrometer (EDS). Elemental analyses confirm that Mn2+ is doped in a CsPbI3 perovskite crystal lattice and the actual Mn doping contents are 0, 0.8, 3.1, and 5.0% for Mn/Pb from 0 to 60% (Figure S2 and Table S2). We chose the elemental mapping area of Mn2+-doped CsPbI3 NCs in high-angle annular dark-field scanning transmission electron microscopy image (HAADF-STEM) (Figure d). On the basis of elemental mapping images (Figure e–h), all of the elements are uniformly distributed in NCs, and the mapping of the Mn element can be clearly observed. Therefore, these results confirm that Mn2+ can be successfully doped in CsPbI3 NCs.

Figure 3

(a) XRD patterns of Mn2+-doped CsPbI3 NCs. (b) Fine XRD patterns in the region of 27–31° (Mn/Pb = 0, 0.8, 3.1, 5.0%). (c) TEM images of CsPbI3 NCs and Mn2+-doped CsPbI3 NCs. Insets: HRTEM images. (d) HAADF-STEM images of Mn2+-doped CsPbI3 NCs and the corresponding elemental mapping of Cs (e), Pb (f), Mn (g), and I (h). (c, d) Mn/Pb = 0 and 5.0%.

PL Stability

It should be noted that Mn2+-doped CsPbI3 NCs exhibited good PL stability in a long term. After being preserved in air for 36 days, PL intensity decayed less than 10% (Mn/Pb = 5.0%), while the PL intensity of CsPbI3 decayed more than 25% (Figure S3). Therefore, Mn2+-doped CsPbI3 NCs showed potential to improve stability in extreme conditions. To evaluate the stability of samples, UV, polar solvent, and thermal resistance of NCs were systematically studied. First, the stability under UV irradiation (365 nm (8 W) and 254 nm (8 W)) was measured. With increasing UV irradiation time, PL intensities of Mn2+-doped CsPbI3 NCs decreased. It should be noted that CsPbI3 NCs exhibited the poorest UV resistance. With the increase in illumination time, the PL intensity quickly decreased (Figure a). While doping Mn2+, their UV resistance can be dramatically improved (Figure b–d). Figure e presents the images of all samples after 22 min of UV irradiation. It is hard to observe an obvious red emission for samples of Mn/Pb = 0 and 0.8% after 22 min of UV irradiation, whereas the brightness of Mn2+-doped CsPbI3 NCs (Mn/Pb = 3.1 and 5.0%) can be still observed. According to the stability decay curves (Figure f), the PL intensity of Mn2+-doped CsPbI3 NCs (Mn/Pb = 5.0%) only decreased to 70% of the initial one after 22 min of UV irradiation, whereas PL intensity of CsPbI3 NCs decreased to 35% of the initial one in the same condition. These results demonstrated that the UV resistance of CsPbI3 NCs can be improved by Mn2+ doping.

Figure 4

(a–d) Evolution of PL of Mn2+-doped CsPbI3 NCs under 365 nm (8 W) and 254 nm (8 W) UV irradiation. (e) Mn2+-doped CsPbI3 NCs under 365 nm UV irradiation at 0 and 22 min. (f) Evolution of PL areas with increasing irradiation time. (a) Mn/Pb = 0%, (b) Mn/Pb = 0.8%, (c) Mn/Pb = 3.1%, and (d) Mn/Pb = 5.0%. UV resistances of NCs were observed using 365 nm (8 W) and 254 nm (8 W) UV-light sources, and their PL data were collected using an RF600 spectrofluorometer with the excitation wavelength at 500 nm.

It is well known that polar solvents can easily induce α-CsPbI3 NC transformation to an orthorhombic phase, which results in poor PL performance.[31] To characterize their phase stability in polar solvents, a certain amount of ethanol was added to NC solutions, and their PL intensities were used to indirectly identify their phase stability.[32,33] The results (Figures S4 and4a) show that PL intensities for all of the samples quickly decreased after adding ethanol, and gradually tended to be stable. However, with increasing Mn2+ doping, the preserved PL intensities can be increased. The Mn2+-doped CsPbI3 NCs (Mn/Pb = 5.0%) showed the most stability against ethanol, and 58% PL intensity could be preserved. Therefore, PL stability can be greatly improved with addition of ethanol as compared to CsPbI3 NCs (10% PL intensity). Furthermore, the thermal resistance of Mn2+-doped CsPbI3 NCs was studied based on their PL evolution (Figures S5 and 4b). At 80 °C, the preserved PL intensities of samples enhanced with increasing Mn/Pb ratios, and the half-lifetimes for their PL intensities (Mn/Pb = 0, 0.8, 3.1, and 5.0%) are 19, 20.5, 30, and 35 min, respectively. It should be emphasized that the decrease of PL at high temperatures may mainly result in ionic migration and crystal fusion.[12,34,35] Therefore, the precise characterization of phase stability was carried out by XRD.

Crystal-Phase Stability

We further verified the thermal stability of Mn2+-doped CsPbI3 films. All of the films were placed on a hot plate at 120 °C, and we periodically measured their XRD patterns. Figure shows that all of the as-prepared samples exhibit a cubic phase of CsPbI3. CsPbI3 films exhibited the poorest phase stability, and gradually converted from the cubic phase to an orthorhombic phase after heating at 120 °C for 1 h. With increasing the Mn/Pb ratio to 0.8–3.1%, their cubic phase can be maintained for at least 5 h. On further increasing the Mn/Pb ratio to 5.0%, their cubic phase was stable even after 7 h. We also monitored the PL intensities of the solid film evolution at 120 °C. With increasing time, PL intensities of the films decreased, and PL peaks exhibited a red shift of about 2–5 nm (Figure S6), which might be related to the growth of CsPbI3 NCs due to the ionic migration, crystal fusion, and exciton quenching.[12,35] Therefore, the thermal stability of CsPbI3 films can be enhanced via Mn2+ doping (Figure ).

Figure 5

(a) Evolution of PL areas at different times with addition of samples with 20% ethanol. (b) Evolution of PL areas with increasing heating time at 80 °C.

Figure 6

Evolution of the XRD pattern for Mn2+-doped CsPbI3 NC thin films at 120 °C (humidity: 30–40%). (a) Mn/Pb = 0%, (b) Mn/Pb = 0.8%, (c) Mn/Pb = 3.1%, and (d) Mn/Pb = 5.0%.

Conclusions

In this work, we used Mn(Ac)2 as a dopant precursor via a direct synthesis method to improve the phase stability of α-CsPbI3 NCs. Mn2+ doping can dramatically decrease the defects of α-CsPbI3 NCs and enhance their PLQYs. Benefiting from the suitable Mn/Pb ratio (5.0%), their phase stabilities in extreme conditions can be further improved. Under UV irradiation and addition of ethanol, the PL intensities of CsPbI3 NCs could only be preserved up to 35% (22 min UV irradiation) and 10% (ethanol addition), respectively, whereas Mn2+-doped CsPbI3 (5.0%) exhibited much improved stability, and their intensities could be preserved up to 70% (22 min UV irradiation) and 58% (ethanol addition), respectively. Furthermore, the phase stability could be maintained for at least 7 h even at 120 °C. With improved optical performance and phase stability, Mn2+-doped CsPbI3 NC showed potential for application in optoelectronic devices.

Experimental Section

Chemical Materials

Oleic acid (OA, 90%), 1-octadecene (ODE, 90%), oleylamine (OLA, 80–90%), cesium carbonate (Cs2CO3, 99.99%), manganese acetate (Mn (Ac)2, 95%), and PbI2 (99.99%) were purchased from Aladdin. All of the chemicals were used without further purification.

Synthesis of Cs-OA

Cs2CO3 (0.39 g), OA (2.0 mL), and ODE (18.0 mL) were mixed into a 100 mL three-neck flask. After degassing and drying under vacuum for 1 h at 120 °C, the solution was heated at 150 °C under N2. After, a clear solution was obtained, the solution was cooled to 60 °C to obtain a Cs-OA solution.

Synthesis of OLA-HI

Twenty milliliters of OLA and 2 mL of HI were mixed into a 100 mL three-neck flask. Then, the solution was heated at 120 °C for 2 h under N2 to remove water. The solution was then cooled to 60 °C to obtain an OLA-HI solution.

Synthesis of Mn2+-Doped CsPbI3 NCs

Mn2+-doped CsPbI3 NCs were synthesized by the following method. In a typical synthesis, different ratios of PbI2 (0.4 mmol)/Mn(Ac)2 were mixed with 10 mL of ODE in a 100 mL three-neck flask, and the mixture was degassed and dried in vacuum for 1 h at 120 °C. Then, 1.0 mL of OLA, OA, and preheated OLA-HI was injected into the reaction flask, respectively. The mixed solution became clear and was degassed in vacuum for 30 min at 120 °C. Then, the temperature was increased to 260 °C. One milliliter of Cs-OA was swiftly injected. The reaction was stopped at 1 min by moving into an ice bath. The purification process was carried out twice using methyl acetate to precipitate Mn2+-doped CsPbI3 NC.

Characterization

The UV–vis absorption spectra of NC solutions were carried out using a PerkinElmer Lambda 650 S spectrophotometer. PL spectra were collected using an RF600 spectrofluorometer with an excitation wavelength of 425 nm. The crystal structures of NCs were analyzed by X-ray diffraction (XRD, Germany Bruker X-ray diffractometer). The operation voltage and current were 40 kV and 40 mA, respectively, with Cu Kα radiation (λ = 1.5418 Å). The morphology and size of NCs were measured using a transmission electron microscope (TEM) (Hitachi, HT7700) and a high-resolution TEM (HRTEM) (Talos, F200X).