1. INTRODUCTION

Halide perovskites are prospective materials for optoelectronics because of the remarkable properties, such as low manufacturing cost, high luminescence efficiency, and tunable light emission characters due to their abundant variety of compositions [1–4]. Among the numerous applications, the white-light emission of organic-inorganic hybrids and all inorganic halide perovskites have received special attention [5–7]. Through flexible regulation of halogen elements in ${\text{CsPbX}}_3$ (X = Cl, Br, I), the white emission with a diverse color temperature and index can be generated [8,9]. However, the anion exchange derived from mixtures of different halide perovskites, and the relatively narrow coverage of lead (Pb) in halide-based perovskites’ luminescent spectra are important factors hindering the development of white lighting [10–12]. Meanwhile, the problems of thermal instability in organic systems and toxicity caused by the heavy Pb limit their further commercial applications [13,14]. In view of this situation, all-inorganic lead-free metal halide perovskites will have more promising prospects in an efficient, stable, and eco-friendly white lighting.

Control of material dimensionality enables them to form various structure types and light emission features [15–17]. The typical 3D and 2D metal halides have been widely investigated because of their excellent optical and electronic characteristics. Nevertheless, low-dimensional 1D and 0D materials exhibit more unique photophysical properties, such as a large Stokes shift, broadband emission, and high photoluminescence quantum yield (PLQY) due to the self-trapped excitons or excited state structural reorganization [18–20]. Recently, Cu-based metal halide perovskites are gradually coming into view because of their abundance, low environmental impact, high efficiency, and low-dimensional structures. For instance, green emission ${\mathrm{Cs}\mathrm{CuCl}}_3$ and ${\mathrm{Cs}}_2{\mathrm{CuCl}}_4$ nanocrystals [21], blue-green luminescence ${\mathrm{Cs}}_2{\mathrm{CuX}}_4$ (X = Cl, Br, Br/I) perovskite quantum dots [22], blue light emission ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{Br}}_{5-x}{\mathrm{I}}_x$ ($0\le \mathrm{x}\le 5$) with near-unity PLQY [23], and ${\mathrm{CsCu}}_2{\mathrm{X}}_3$ (X =Cl, Br, I) with improved air and thermal stability [24] have been recently reported. Specifically, ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ with a 0D electronic structure, large Stokes shift, and strong blue emission has been fabricated and applied for LED devices [25,26]. Another phase of cesium copper iodide, ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, shows broadband yellow emission [24], and the stable yellow LED based on it was successfully realized [27]. Additionally, ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ single crystals with a high-PLQY due to strongly localized 1D excitonic recombination have been reported [28]. As both ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ are pure iodide phases, the general mixed halide’s exchange existing in perovskite’s mixtures could be avoided, and pure white emission can be further achieved by appropriate mixing of these phases. Previously, a pure white-light emission has been successfully realized by adjusting the appropriate mixing ratio of these two phases [29,30]. Nevertheless, these methods always needed to fabricate two pure phases of blue emission ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and yellow emission ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, and then precisely control the mixing ratio of these individual luminescent materials, which was intricate and inconvenient for the practical application. Therefore, developing a simple intrinsic white-light emission of a cesium copper iodide system will become an effective approach. It has been reported that through a controllable CsI–CuI phase transformation by solvent treatment, stable ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ was obtained from ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$, and a single white-emission layer could be prepared [31]. In addition, Shi’s group has reported the electroluminescent white-light emitting diodes in terms of Cu-based halide materials [32].

In this paper, we have successfully prepared the white luminescent material by a one-step doping method, achieving full coverage of the visible spectrum. Through adding impurities into the ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ system, high efficiency and stable ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ was successfully synthesized, and the coexistence of varied high luminescence phases realized the white lighting. By means of the disposable preparation of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ phases, a high-quality and more uniform white luminescence with CIE coordinates of (0.3397, 0.3325) and a CRI reaching 90 could be generated in a simple way.

2. MATERIALS AND METHODS

Cesium iodide (CsI, 99.9%), copper (I) iodide (CuI, 99.999%), neodymium iodide (${\mathrm{NdI}}_3$, 99.9%), terbium iodide (${\mathrm{TbI}}_3$, 99.99%), praseodymium iodide (${\mathrm{PrI}}_3$, 99.9%), bismuth iodide (${\mathrm{BiI}}_3$, 99.9%), $\mathrm{N},\mathrm{N}$-dimethylformamide (DMF, 99.9%), and isopropanol (99.5%) were directly used without further purification.

The cesium copper iodide perovskites were synthesized via an antisolvent infiltration method, which was performed at room temperature by adding the precursor solution within a good solvent into a nonpolar poor solvent. The blend of two various solvents induced a transient supersaturation, leading to the nucleation and form of perovskites. Cesium iodide, copper (I) iodide, and impurity materials (in this paper, ${\mathrm{NdI}}_3$, ${\mathrm{TbI}}_3$, ${\mathrm{PrI}}_3$, and ${\mathrm{BiI}}_3$ were used as impurities, respectively) in different molar ratios were firstly dissolved in DMF to get a precursor solution. Then, the solution was rapidly dropped into the antisolvent of isopropanol to form a precipitate, and the resulting products were the mixture of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$. With the increase in the molar ratio of doping materials, a mixture with different proportions of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ could be obtained. The specific synthesis procedures are as below.

Adding ${\mathrm{NdI}}_3$ into the ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ system. In the synthesis of molar ratios of 9 mol%, CsI (0.6 mmol), CuI (0.364 mmol), and ${\mathrm{NdI}}_3$ (0.036 mmol) were dissolved in DMF (4 mL). The mixture was stirred for 2 h at 80°C, then we let it cool naturally to room temperature. Next the precursor solution was rapidly added into isopropanol (20 mL) with vigorous stirring under air ambient at room temperature, and a precipitate was produced immediately during this process. Then, the resulting precipitate was filtered and washed with isopropanol; the yield of the product was about 75%. The other three concentrations were obtained by three corresponding molar doses: 3 mol% (0.6 mmol CsI, 0.388 mmol CuI, and 0.012 mmol ${\mathrm{NdI}}_3$), 5 mol% (0.6 mmol CsI, 0.38 mmol CuI, and 0.02 mmol ${\mathrm{NdI}}_3$), and 7 mol% (0.6 mmol CsI, 0.372 mmol CuI, and 0.028 mmol ${\mathrm{NdI}}_3$).

Adding ${\mathrm{TbI}}_3$, ${\mathrm{PrI}}_3$, and ${\mathrm{BiI}}_3$ into a ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ system. The impurities were chosen as 9 mol% (0.6 mmol CsI, 0.364 mmol CuI, and 0.036 mmol for ${\mathrm{TbI}}_3$, ${\mathrm{PrI}}_3$, and ${\mathrm{BiI}}_3$, respectively); the synthesis process was the same as above.

Adding ${\mathrm{NdI}}_3$ into the ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ system. The impurities were chosen as 9 mol% (0.3 mmol CsI, 0.546 mmol CuI, and 0.054 mmol ${\mathrm{NdI}}_3$), and the synthesis process was the same as above.

Synthesis of pure ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5\text{:}0.6\mathrm{mmol}$ CsI and 0.4 mmol CuI were used; the synthesis process was the same as above.

Synthesis of pure ${\mathrm{CsCu}}_2{\mathrm{I}}_3\text{:}0.3\mathrm{mmol}$ CsI and 0.6 mmol CuI were used; the synthesis process was the same as above.

Photoluminescence (PL), photoluminescence excitation (PLE), and photoluminescence quantum yield (PLQY) measurements were performed at ambient temperature by an FS5 fluorescence spectrometer equipped with a xenon lamp and an integrating sphere. Powder X-ray diffraction (PXRD) measurements were performed by the Rigaku MiniFlex600 system equipped with a Cu-Kα radiation source ($\lambda =1.5418\mathrm{\mu m}$). All scans were performed at room temperature with a step size of 0.02˚. X-ray photoelectron spectroscopy (XPS) analyses were conducted using an ESCALAB 250Xi spectrometer. Scanning electron microscope (SEM) measurements were performed by a ZEISS SUPRA 55 from Carl Zeiss, Germany. High resolution transmission electron microscopy (HRTEM) images were measured by the JEM-3200FS (JEOL).

3. RESULTS AND DISCUSSION

The ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ system with added ${\mathrm{NdI}}_3$ is first detected by the SEM and TEM techniques. The SEM image [Fig. 1(a)] implies the large quantity and good uniformity of the ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ phases. Energy-dispersive X-ray spectroscopy (EDS) mapping [Fig. 1(b)] confirms that Cs, Cu, I, and Nd are evenly distributed in the material. The SEM and TEM [Fig. 1(c)] images of the as-synthesized ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ composites show a dual-phase morphology characteristic composed of nanoparticles and nanorods, which correspond to ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, respectively. Such morphology features can effectively avoid the color changes of the composites caused by anion exchange, because the two components exist individually without the formation of compact structure configuration, which will make it quite fit to emit stable white light. The TEM image [Fig. 1(c)] shows the micro morphologies of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, in which ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ crystallizes as nanoparticles with an average diameter of $\sim 300\mathrm{nm}$, whereas ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ appears as nanorods with the average length of $\sim 2\mathrm{\mu m}$ and width of $\sim 450\mathrm{nm}$. The TEM lattice fringes and fast Fourier transform (FFT) images of $d$-spacing $\sim 3.24$Å and $\sim 2.1$Å (1 Å = 0.1 nm) coincide with the lattice planes of (213) and (350) for ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, respectively [Figs. 1(d) and 1(e)]. The X-ray photoelectron spectroscopy (XPS) measurements are conducted to validate the chemical states of ions [Figs. 1(f)–1(i)]. The binding energies of 951.8 eV and 932 eV coincide to Cu ${2\mathrm{p}}_{1/2}$ and Cu ${2\mathrm{p}}_{3/2}$, respectively, which demonstrates the monovalent state for copper in the material. Meanwhile, Cs $3\mathrm{d}$ and I $3\mathrm{d}$ correspond to the $+1$ and $-1$ states, which are consistent with published data for cesium copper iodides [25,29]. These results indicate that a system including both high quality ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ has been successfully prepared through the doping method.

 

Fig. 1. (a) SEM, (b) EDS mapping, and (c) TEM images of the sample with ${\mathrm{NdI}}_3$. The lattice planes and fast Fourier transform (FFT) images of (d) ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ nanoparticles and (e) ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ nanorods. (f) XPS analysis of the sample with 9 mol% ${\mathrm{NdI}}_3$, and the respective spectra of (g) Cs $3\mathrm{d}$, (h) Cu $2\mathrm{p}$, and (i) I $3\mathrm{d}$.

Download Full Size | PDF

Figure 2(a) demonstrates the PL (photoluminescence) spectra of pure ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, in which ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ nanoparticles exhibit blue luminescence at 446 nm with a broad FWHM (full width at half maximum) of about 80 nm, while ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ nanorods display yellow emitting at 576 nm with an FWHM of approximately 120 nm. Both samples show large Stokes shifts, and in this way self-absorption by the system can be efficiently prevented. The emissions are possible mostly owing to the self-trapped excitons, deriving from the Jahn-Teller distortion or strong exciton-phonon coupling [23,33]. In addition, broad emissions are usually helpful for enhancing the CRI of white lighting. The PLQYs are ∼$69.94\%$ for pure ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and 9.91% for ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, respectively, and the higher PLQY of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ nanoparticles than ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ nanorods may result from the stronger exciton binding energy in 0D ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$. The PL spectra of the ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ system with various ratios of ${\mathrm{NdI}}_3$ under excitation of 316 nm are shown in Fig. 2(b). At the ratio of 3 mol%, a PL emission involving both ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ has been detected in the material. With the increase of molar ratios, the intensity difference between the emission of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ diminishes, indicating that the proportion of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ gradually increases. As ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ possess comprehensive red, green, and blue components, a mixture of them could form a white lighting. When the intensities of yellow emission from ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and blue emission from ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ reach an almost equal ratio, a normal white lighting can be produced. From Fig. 2(b), it can be seen that at the ratio of 9 mol%, a white emission has been successfully formed. Figure 2(d) shows that the CIE coordinates of it are (0.3397, 0.3325), approaching the standard white light of (0.33, 0.33). In addition, the CRI of this material can reach up to 90 and its PLQY is 16.9%. The PLE (photoluminescence excitation) spectrum for the sample with concentration of 9 mol% is shown in Fig. 2(c). It can be seen that under the excitation wavelength of 316 nm, the PLE efficiency for ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ is almost the same, indicating that a white emission can be motivated at this wavelength, which is consistent with the 9 mol% white PL spectra in Fig. 2(b). Apparently, at this wavelength both of the PL efficiencies for ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ do not reach the optimum, but the yellow emission from ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and blue emission from ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ reach a balance, and a normal white emission can be produced. Figure 2(c) also demonstrates that although the PLQY of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ is much higher than that of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, only the PL efficiency of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$, which has the same amount as that of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, will be effective enough to make a contribution to the form of white lighting. As a result, the emitting intensities will nearly achieve a balance. The large amount of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ in the material with the ratio of 9 mol% will lead to the increase of the balanced effective PL efficiency and the overall powerful white luminescence. To further investigate how the doping ratios influence the formation of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and the proportion between ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, the XRD results are shown in Fig. 2(e). At the low ratio of 3 mol%, the corresponding diffraction peaks of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ appear simultaneously, indicating that both of the two phases are synthesized in the material. Meanwhile, the weak peak intensity of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ signifies that the proportion of it is tiny compared to another phase of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$. With the increase of dopant, the diffraction peak of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ is enhanced, while the diffraction from ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ is gradually weakened, which means that the amount of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ increases accordingly. This result is consistent with the PL spectra [Fig. 2(b)], along with the increase of the doping molar ratio, and the PL intensity difference between emission of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ gradually diminishes.

 

Fig. 2. (a) PL spectra of pure ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$. (b) PL spectra of samples with different concentration of ${\mathrm{NdI}}_3$ under excitation of 316 nm. (c) PLE spectra of emission peaks corresponding to ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ of the sample with 9 mol% ${\mathrm{NdI}}_3$. (d) CIE coordinates of the sample with 9 mol% ${\mathrm{NdI}}_3$. (e) XRD diffraction patterns of samples with different concentration of ${\mathrm{NdI}}_3$ compared to the standard XRD patterns of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$.

Download Full Size | PDF

It has been reported that ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ crystallizes in the Pnma space group of an orthorhombic crystal system. Two types of ${\mathrm{Cu}}^{+}$ appear as tetrahedral sites of ${\mathrm{Cu}}^{+}{\mathrm{I}}_4$ and triangular sites of ${\mathrm{Cu}}^{+}{\mathrm{I}}_3$ in the material, which are edge-connected coming into the isolated ${[{\mathrm{Cu}}_2{\mathrm{I}}_5]}^{3-}$ units [34,35]. In particular, these units are divided by the nearby ${\mathrm{Cs}}^{+}$ ions, forming a zero-dimensional structure [Fig. 3(a)]. In contrast, ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ crystallizes in the Cmcm space group of an orthorhombic system, in which ${\mathrm{Cu}}^{+}{\mathrm{I}}_4$ tetrahedra sites share common faces and edges and form double chains of ${\mathrm{Cu}}_2{\mathrm{I}}_3$ along the $c$-axis separated by cesium, leading to a 1D structure [Fig. 3(b)]. In the research of lead bromine perovskite, it has been reported that ${\text{CsPbBr}}_3$ can gradually transform into another phase of ${\mathrm{Cs}}_4{\mathrm{PbBr}}_6$ by adding ${\mathrm{ZnBr}}_2$ as a revulsive. The process of “survival of the fittest” occurs in ${\text{CsPbBr}}_3$, in which the good quality and stable ${\text{CsPbBr}}_3$ is retained, while the unstable ${\text{CsPbBr}}_3$ is decomposed and likely ripened into the 416-structure by adding the ${\mathrm{ZnBr}}_2$ revulsive [36]. Like the role of ${\mathrm{ZnBr}}_2$ in the system of Cs-Pb-Br, the mechanism of the doping process in the transformation of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ can be attributed to the role the impurities played as a conversion revulsive. The 1D ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ is more stable and easier to generate than the 0D ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$. When the dopants are added into the ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ system, the balance of the whole system will be destroyed, in this process the stable ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ nanocrystals are retained, while the unstable particles with poor quality will be disintegrated and rapidly formed into the ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ structure. The resulting mixture of these two phases can be obtained on this basis. However, when the circumstance is in reverse, namely materials are added into the ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ system, another phase of the ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ does not appear simultaneously (Fig. 4). It can be seen that in every doping concentration, only PL and XRD peaks of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and undissolved CuI (PL emission of 420 nm) exist in the material. The form of nanocrystals requires that the precursor concentration exceeds the critical concentration of nucleation, and a large number of nuclei occur in a very short period of time. When precursor concentration is lower, the nanoparticles will dissolve and disappear, and thus the 0D ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ cannot form concurrently in the ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ system by the doping method.

 

Fig. 3. Crystal structures of (a) ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and (b) ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ with optimized lattice parameters from the top.

Download Full Size | PDF

 

Fig. 4. (a) PL spectra under excitation wavelength of 320 nm. (b) XRD patterns of the samples with ${\mathrm{NdI}}_3$ added into the ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ system with various ratios.

Download Full Size | PDF

Next, other materials like ${\mathrm{TbI}}_3$, ${\mathrm{PrI}}_3$, and ${\mathrm{BiI}}_3$ were added in turn to measure their respective effect on the white emission efficiency. As shown in Fig. 5(a), PL white spectra involving ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ peaks have been detected in every material. The PLQYs for ${\mathrm{NdI}}_3$, ${\mathrm{TbI}}_3$, ${\mathrm{PrI}}_3$, and ${\mathrm{BiI}}_3$ are 16.9%, 14.5%, 9.32%, and 5.52%, respectively. The CIE coordinates in the system of ${\mathrm{NdI}}_3$ and ${\mathrm{TbI}}_3$ are (0.3397, 0.3325) and (0.3477, 0.3368), respectively [Fig. 5(c)], indicating that white emission has been successfully formed within them. As shown in Fig. 5(b), the corresponding XRD peaks of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ appear simultaneously but with different proportions in these samples. Compared to other impurities, the ${\mathrm{NdI}}_3$ sample has the strongest ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ diffraction peak, indicating the largest amount of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ in it. In the reaction process, the solubility and the ionic activity of impurities in the ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ precursor solution will affect the amount and quality of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ to a large extent and will further influence the overall luminescence effect. The inset of Fig. 5(a) shows the ${\mathrm{NdI}}_3$, ${\mathrm{TbI}}_3$, and ${\mathrm{PrI}}_3$ doped ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ solution. It can be seen that the ${\mathrm{NdI}}_3$ sample is completely dissolved and more clarified than others, while the ${\mathrm{TbI}}_3$ and ${\mathrm{PrI}}_3$ samples are like a suspension similar to colloid, which means that these two impurities do not dissolve absolutely in the ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ solution. In addition, Nd is one of the liveliest rare earth metals; it will quickly react in hot solution which is effective to the fast formation of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ and makes positive influence on the overall emission efficiency. Therefore, compared to other dopants, the ${\mathrm{NdI}}_3$ sample has the strongest ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ diffraction peak, indicating the largest amount of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ in it. As discussed previously in Fig. 2, the more amount of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ is beneficial to obtain stronger white emission, and thus the PLQY of the NdI₃ sample is higher than that of other materials. Moreover, the uniformity and quality of the sample could also affect the whole luminescence intensity. The SEM images are shown in Figs. 5(d)–5(g); the uniformity and quality of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ in the ${\mathrm{NdI}}_3$ sample are much better compared to others, corresponding to its higher PL efficiency. In addition, the new generated material is another factor affecting the PLQY. From Fig. 5(b), it can be found that in the ${\mathrm{BiI}}_3$-doped sample new ${\mathrm{Cs}}_3{\mathrm{Bi}}_2{\mathrm{I}}_9$ appears. It consumes a great deal of Cs and I ions which will compose ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$, resulting in the whole material getting few effective luminous centers for white emission and a low PL intensity. Therefore, combined with the above factors, the PLQY of the ${\mathrm{NdI}}_3$ sample is higher than that of other doping materials.

 

Fig. 5. (a) PL spectra of samples with 9 mol% concentration of ${\mathrm{NdI}}_3$ (PL excitation of 316 nm), ${\mathrm{TbI}}_3$ (PL excitation of 317 nm), ${\mathrm{PrI}}_3$ (PL excitation of 319 nm), and ${\mathrm{BiI}}_3$ (PL excitation of 308 nm). Inset: the precursor solution of ${\mathrm{NdI}}_3$, ${\mathrm{TbI}}_3$, and ${\mathrm{PrI}}_3$ samples. (b) XRD patterns of samples with 9 mol% concentration of ${\mathrm{NdI}}_3$, ${\mathrm{TbI}}_3$, ${\mathrm{PrI}}_3$, and ${\mathrm{BiI}}_3$. (c) CIE coordinates of samples with 9 mol% ${\mathrm{NdI}}_3$ and ${\mathrm{TbI}}_3$. (d)–(g) SEM images of the samples with 9 mol% ${\mathrm{NdI}}_3$, ${\mathrm{TbI}}_3$, ${\mathrm{PrI}}_3$, and ${\mathrm{BiI}}_3$, respectively.

Download Full Size | PDF

4. CONCLUSIONS

In conclusion, a controllable one-step doping method was adopted in the cesium copper iodide perovskite’s luminescence, and the results indicated that a system including ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ with high quality had been successfully prepared. Through comparing the PL efficiency of samples under various molar ratios and materials, it has been investigated that the amount of ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ combining with the uniformity and quality of ${\mathrm{Cs}}_3{\mathrm{Cu}}_2{\mathrm{I}}_5$ and ${\mathrm{CsCu}}_2{\mathrm{I}}_3$ was the key factor affecting the white-lighting. Therefore, a high-quality white-emission with CIE coordinates of (0.3397, 0.3325) and CRI of 90 was obtained in a convenient way. This work provides a new approach for the investigation of Cu-based metal halide perovskites and will be helpful for the exploration of lead-free perovskites.