A Cyan Emitting CsPbBr₃ Perovskite Quantum Dot Glass with Bi Doping

Sample preparation

In this work, 5Na₂CO₃-20ZnO-80H₃BO₃-15SiO₂-10BaCO₃-XBi₂O₃ (X = 0, 6, 12, 24, 48, 72 mol%) borosilicate glass composition with 1.5Cs₂CO₃-3PbO-9NaBr(mol%) perovskite-related sources was used as the raw materials to prepare QDs glass. The X-ray diffraction (XRD) patterns of perovskite quantum dot glasses doped with different Bi₂O₃ molar ratios are shown in Fig. 1a. All diffraction peaks can correspond to the standard card CsPbBr₃ (PDF#54-0752) without any obvious secondary phase. Simultaneously, the diffraction peak did not change significantly with Bi₂O₃ doping. As shown in Fig. 1b, the CsPbBr₃ crystal has a cubic structure with a space group of Pm $\overline{3}$ m. CsPbBr₃ QDs are successfully embedded in the glass matrix as depicted in Fig. 1c. High resolution TEM (HRTEM) micrograph (Fig. 1d) clearly shows the crystallinity and lattice fringes of CsPbBr₃ particles. The interplanar spacing is 2.38 Å, which is consistent with the (211) plane of cubic CsPbBr₃. The results indicate that the CsPbBr₃ QDs were successfully grown in the borosilicate glass matrix, and the doped Bi³⁺ ions did not enter the CsPbBr₃ nanocrystals.

Zoom In Zoom Out Reset image size

The variation of the normalized PL spectra of CsPbBr₃: xBi₂O₃ (x = 0, 0.06, 0.12, 0.24, 0.48, 0.72) excited by 365 nm light are shown in Fig. 2a. Without doped with Bi₂O₃, the CsPbBr₃ perovskite quantum dot glass sample at this time has a maximum emission peak at 523 nm. With the gradual increase of the content of doped Bi₂O₃, the emission peak gradually decreased. When the doping concentration x = 0.48, the emission peak reached the lowest value, and the emission intensity was almost negligible. In addition, by normalizing the emission intensity, it was found that with the increase of the doping concentration of Bi₂O₃, the emission peak has a blue shift from 523 nm to 493 nm. The reason for this phenomenon is that with the addition and increase of Bi³⁺ ions, the matrix of the glass network structure does not change, and Bi³⁺ ions just enter the vacancies of the glass network structure, which will reduce the growth of CsPbBr₃ quantum dot occupying vacancies Probability, which in turn leads to a decrease in emission intensity. At the same time, when Bi³⁺ ions and CsPbBr₃ quantum dot coexist in the same vacancy, the growth space of CsPbBr₃ quantum dot becomes smaller due to the existence of Bi³⁺ ions, resulting in a smaller size of CsPbBr₃ quantum dot, resulting in a blue shift of the emission peak.

Zoom In Zoom Out Reset image size

In further experiments, we measured the UV–vis absorption spectrum of CsPbBr₃ quantum dot glass samples. As shown in Fig. 2b, it can be seen that without doping Bi₂O₃, the absorption curve of the CsPbBr₃ quantum dot sample has a strong absorption at 523 nm, which corresponds to its emission spectrum, which the direct exciton recombination luminescence of CsPbBr₃ quantum dot in the glass. When a small amount of Bi₂O₃ (x = 0.06) is doped, the absorption curve of the CsPbBr₃ quantum dot sample at this time will have a slight blue shift, and the absorption intensity will also decrease accordingly. This is consistent with the performance of the emission spectrum, which is all due to the doping of Bi³⁺ ions into the vacancies of the grid structure of the glass matrix, which leads to the reduction of the size and number of CsPbBr₃ quantum dot and the increase of the band gap. When a certain amount of Bi₂O₃ (x = 0.12, 0.24) is doped, the absorption curve has an obvious blue shift, and the absorption peak intensity is around 480 nm. Through searching for information, we found that the doped Bi³⁺ ion will produce an emission peak at 480 nm. Therefore, the absorption peak produced by the absorption curve at 480 nm is mainly derived from Bi³⁺ ions in the glass matrix. When excessively doped Bi₂O₃ (x = 0.48, 0.72), the absorption curve shows a large-span red shift, which is due to the self-absorption phenomenon of the Bi³⁺ ion in the glass matrix to the CsPbBr₃ quantum dot. This is the reason why CsPbBr₃:xBi³⁺(x = 0.48, 0.72) samples hardly emit light under 365 nm excitation. Figure 2c shows the photo of the CsPbBr₃: xBi³⁺ glass sample under daylight and UV irradiation. When the doping concentration of Bi₂O₃ is 0.72 mmol, the prepared CsPbBr₃: xBi³⁺ sample hardly emits light, which corresponds to the result of the emission spectrum.

In addition, in order to more directly understand the influence of Bi₂O₃ on the luminescence of CsPbBr₃, we measured the photoluminescence quantum yield (PLQY). As shown in Fig. 3a, when the sample is not doped, the quantum yield is 60.8%. As the doping concentration of Bi₂O₃ gradually increases, the quantum efficiency of the sample gradually decreases. When the doping concentration of Bi₂O₃ reaches 0.72 mmol, the quantum efficiency is only 0.8%, the PLQY result is consistent with the emission spectrum. What is more, we monitored the time-resolved PL decay curve of Bi₂O₃ quantum dot glass samples doped with different concentrations as depicted in Fig. 3b. Using the double exponential fitting function to fit the attenuation curve well is as follows:

$$\begin{eqnarray}&&A\,=\,{A}_{0}+{B}_{1}\exp (-t/{\tau }_{1})+{B}_{2}\exp (-t/{\tau }_{2})\end{eqnarray} \tag{ 1 }$$

where A is phosphorescence intensity at time t, and B₁/B₂ and τ₁/τ₂ are the weight coefficient and decay time, respectively. And the fitting formula of average life is:

$$\begin{eqnarray}&&{\tau }_{ave}\,=\,\displaystyle \int A(t)dt/{A}_{0}\end{eqnarray} \tag{ 2 }$$

In the formula, A₀ is the peak intensity, and A(t) is the PL intensity related to the recording time.

Zoom In Zoom Out Reset image size

According to previous studies, ³² the double exponential fitting results of the fluorescence lifetime of perovskite nanocrystals can be divided into fast decay time and slow decay time. The slow decay time may be related to the radiative recombination used for luminescence, and the fast decay time part may be related to the non-radiative recombination when the surface electrons are captured by the defect state and lost in the non-radiative process. As shown in Table I, we can observe that the value of the slow decay time decreased from 9.943 ns (56.66%) for the undoped sample to 2.548 ns (18.37%) for the 0.72 mmol sample. At the same time, as the Bi₂O₃ doping concentration gradually increases, the average life of the sample gradually decreases, which is consistent with the results of luminescence intensity and quantum yield.

Table I. The fluorescence lifetime of the sample after double exponential fitting and the average lifetime under different Bi₂O₃ concentrations.

Doped (mmol)	${\tau }_{{\rm{1}}}$ (ns)	${\tau }_{{\rm{2}}}$ (ns)	${\tau }_{ave}$ (ns)
0	1.2640 (43.34%)	9.943 (56.66%)	6.18152140
0.06	0.9038 (42.65%)	6.223 (57.35%)	3.95436120
0.12	0.5852 (36.93%)	5.346 (63.07%)	3.58783656
0.24	0.2580 (70.61%)	2.351 (29.39%)	0.87373270
0.48	0.2990 (80.60%)	2.666 (19.40%)	0.75819800
0.72	0.2776 (81.63%)	2.548 (18.37%)	0.69467248

The XRD test has confirmed that the doped Bi³⁺ does not enter the CsPbBr₃ quantum dot lattice. In order to verify that the doped Bi³⁺ ions only enter the vacancies of the glass network structure, we measure the infrared spectrum of the glass sample to determine the influence of Bi³⁺on the composition and structure of the glass matrix. As shown in Fig. 4, it shows the FT-IR spectra of a glass sample not doped with Bi₂O₃ and a glass sample doped with 0.24 mmol Bi₂O₃ at 400–4000 cm⁻¹ wavenumber. And it can clearly show the vibration types and wave numbers of [BO₃], [BO₄] and [SiO₄] units in the glass matrix. The absorption peak near 510 cm⁻¹ is attribute to the bending vibration of the B–O–B bond. The absorption peak near 730 cm⁻¹ is comes from the B–O–B bond bending vibration in the [BO₃] triangle. The absorption peak at 1000 cm⁻¹ belongs to the stretching vibration of [SiO₄] tetrahedron. The absorption peak near 1240 cm⁻¹ is mainly ascribed to the vibration of the boron oxygen ring. And the absorption peak near 1375 cm⁻¹ is mainly derived from the asymmetric stretching vibration of the B–O–B bonds in the [BO₃] triangle. Comparing the infrared spectra of a glass sample not doped with Bi₂O₃ and a glass sample doped with 0.24 mmol Bi₂O₃, it can be found that the vibration type and wavenumber of the glass matrix [BO₃], [BO₄] and [SiO₄] groups have almost no change. Therefore, we believe that the doping of Bi₂O₃ does not change the composition and structure of the glass matrix. Motivated by this, it is indirectly confirmed that the doped Bi³⁺ ions only enter the vacancies of the glass network structure.

Zoom In Zoom Out Reset image size

The stability of the prepared CsPbX₃ quantum dot, including light stability, thermal stability and water resistance stability, is the main factor restricting its application. ^33–35 In 2019, Shao et al. have confirmed that using borosilicate as a glass matrix is the most effective measure to improve the stability of borosilicate nanocrystals in water or other environments, and it is also the best way to protect high borosilicate nanocrystals. ³⁶ In order to determine the potential applications of the prepared CsPbBr₃ quantum dot glass in lighting and display, the stability and reliability of the CsPbBr₃:xBi³⁺ glass samples were studied. By heating/cooling the glass sample of CsPbBr₃:xBi³⁺ (x = 0.06), the temperature is increased from room temperature 25 ℃ to 125 ℃, and the measurement is performed every 10 ℃. The PL intensity of the sample changes with temperature is recorded to study the sample's thermal stability, as depicted in Fig. 5a. From the temperature and emission intensity, it can be found that after a heating/cooling cycle, the emission intensity hardly changes. In order to further study the reliability of CsPbBr₃:xBi³⁺ samples, under the same conditions, we conducted 10 heating and cooling shock tests on CsPbBr₃:xBi³⁺ (x = 0.06) samples. As shown in Fig. 5b, the emission peak intensity after 10 consecutive cycles remains basically unchanged. The results of heating and cooling cycling and heating and cooling shock show that the prepared CsPbBr₃:xBi³⁺ sample is very stable, without obvious degradation and size changes.

Zoom In Zoom Out Reset image size

In summary, by doping Bi₂O₃ into the borosilicate glass matrix, the sintering temperature of the CsPbBr₃ quantum dot precursor is reduced, and the precipitation of the CsPbBr₃ quantum dot in the borosilicate glass matrix is realized after heat treatment. The XRD pattern and TEM image structure characterization proved that the CsPbBr₃ quantum dot were indeed crystallized in the borosilicate glass matrix. It also proves that the doped Bi³⁺ ions only exist in the borosilicate glass matrix, but do not participate in the formation of CsPbBr₃ quantum dot. By adjusting the concentration of doped Bi₂O₃, tunable emission from 523 to 493 nm can be achieved. Finally, the CsPbBr₃:xBi³⁺ (x = 0.06) sample was tested for cold and hot cycling and cold and hot impact. The test results showed that the prepared sample had good thermal stability, which provided a good solution for solving the problem of poor thermal stability of perovskite quantum dot.

This work was supported by Science and Technology Planning Project of Zhejiang Province, China (2018C01046), Enterprise-funded Latitudinal Research Projects (J2017-171; J2017-293 and J2017-243), Sponsored by Shanghai Sailing Program (18YF1422500) and Research start-up project of Shanghai Institute of Technology (YJ2018-9).