YCl3-Substituted CsPbI3 Perovskite Nanorods for Efficient Red-Light-Emitting Diodes

Cesium lead iodide (CsPbI3) perovskite nanocrystals (NCs) are a promising material for red-light-emitting diodes (LEDs) due to their excellent color purity and high luminous efficiency. However, small-sized CsPbI3 colloidal NCs, such as nanocubes, used in LEDs suffer from confinement effects, negatively impacting their photoluminescence quantum yield (PLQY) and overall efficiency. Here, we introduced YCl3 into the CsPbI3 perovskite, which formed anisotropic, one-dimensional (1D) nanorods. This was achieved by taking advantage of the difference in bond energies among iodide and chloride ions, which caused YCl3 to promote the anisotropic growth of CsPbI3 NCs. The addition of YCl3 significantly improved the PLQY by passivating nonradiative recombination rates. The resulting YCl3-substituted CsPbI3 nanorods were applied to the emissive layer in LEDs, and we achieved an external quantum efficiency of ~3.16%, which is 1.86-fold higher than the pristine CsPbI3 NCs (1.69%) based LED. Notably, the ratio of horizontal transition dipole moments (TDMs) in the anisotropic YCl3:CsPbI3 nanorods was found to be 75%, which is higher than the isotropically-oriented TDMs in CsPbI3 nanocrystals (67%). This increased the TDM ratio and led to higher light outcoupling efficiency in nanorod-based LEDs. Overall, the results suggest that YCl3-substituted CsPbI3 nanorods could be promising for achieving high-performance perovskite LEDs.


Introduction
Metal halide perovskites (MHPs) have attracted considerable attention due to their impressive characteristics, such as exceptional color purity, effective radiative recombination, adjustable emission wavelengths, and inexpensive solution processability, making them a promising contender for the next generation of lighting and display [1][2][3][4][5]. All inorganic CsPbI 3 nanocrystals (NCs) are indispensable for these purposes among various MHPs due to their exceptional thermal and chemical durability [5]. Recent strategies, including morphology and interfacial management, the encapsulation of perovskite nanocrystals within polymers or glasses, architecture engineering, and surface chemistry engineering, have been employed to enhance the practical applications of CsPbI 3 NCs and raise the external quantum efficiency (EQE) of CsPbI 3 NC-based light-emitting diodes (LEDs) [3][4][5][6]. Despite the significant advancements, CsPbI 3 NC LEDs have considerable limitations due to their small size (0-dimensional NCs), which is caused by the confinement effect [7]. Due to their small size, the defect-enriched surface of colloidal CsPbI 3 NCs has a detrimental effect on the photoluminescence quantum yield (PLQY) and the low-luminous efficiency of LEDs [8]. Furthermore, the excessive native ligands required to passivate the large surface area of small-sized CsPbI 3 NCs forms an insulation layer that impedes the effective carrier transport capability in the assembled NCs' film, resulting in numerous challenging issues associated with CsPbI 3 NC-based LEDs [2].
The limitations of conventional CsPbI 3 NC-based LEDs can be overcome by using anisotropic one-dimensional (1D) nanorods made from CsPbI 3 NCs. This is due to the combination of two-directional quantum confinement effects and a significantly reduced surface trap density in 1D nanorods, leading to excellent photophysical properties and high aspect ratios [9]. The unique surface morphology of 1D nanorods contributes to their excellent photophysical properties and high aspect ratios [10,11], and their well-defined morphology can restrict the active region of charge carriers and reduce the carrier transport distance [3], making them more conducive to the development of efficient LEDs compared to their counterpart nanocrystals [12].
Although the synthesis of perovskite nanorods from the water-oil transformation of Cs 4 PbBr 6 polyhedrons into nanorods or fragmentation of perovskite nanowires initiated by anion-exchange processes has been reported recently [13,14], these methods produce impure morphologies and defect-enriched surfaces. Hence, it is still challenging to directly synthesize perovskite nanorods with high PLQY and well-defined aspect ratios that exhibit efficient radiative recombination rates. Importantly, the anisotropic nature of a nanorod-based emissive layer is valuable for further increasing the outcoupling efficiency by overcoming the photon losses that become trapped through waveguiding and total internal reflection, as they have high ratios of horizontal transition dipole moments (TDMs) compared to isotopically-oriented nanocubes (nanocrystals) [15][16][17].
Here, we propose a proper method to directly synthesize perovskite nanorods composed of YCl 3 -substituted CsPbI 3 by using Yttrium (III) Chloride Hexahydrate (YCI 3 ·6H 2 O). This approach reduces the crystal size of CsPbI 3 NCs by partially substituting Pb 2+ and I − ions with Y 3+ and Cl − ions. The YCl 3 passivates surface traps and controls the net recombination rates, significantly improving PLQY. Moreover, the environmental durability of the YCl 3 -substituted CsPbI 3 nanorods is significantly enhanced, with only a 28% decrease in PLQY after 45 days of storage under ambient conditions. Importantly, the YCl 3 :CsPbI 3 nanorod-based LEDs exhibit a peak EQE of 3.16%, 1.86 times higher than that of the control CsPbI 3 NC-based device (1.69%). Furthermore, the ratio of horizontal transition dipole moments of the anisotropic YCl 3 :CsPbI 3 nanorods is 75%, higher than that of isotopically-oriented TDMs in CsPbI 3 nanocrystals (67%), resulting in higher light outcoupling efficiency. Thus, our findings suggest that anisotropic nanorods have promising potential in light-emitting devices.

Materials and Methods
All the chemicals were purchased from Sigma-Aldrich.

Cesium Oleate Preparation
Firstly, Cs 2 CO 3 (2.49 mmol), OLA (2.5 mL), and 30 mL ODE were degassed and dried at 120 • C for 1 h. Subsequently, the temperature was raised to 150 • C until a clear solution was obtained under N 2 flow.

Synthesis of CsPbI 3 NCs
PbI 2 (0.174 g, 0.376 mmol) and ODE (10 mL) were loaded into a 100 mL three-neck flask, degassed, and dried under vacuum for 1 h at 120 • C. Then, OA (1 mL) and OLA (1 mL) preheated to 70 • C were injected under the protection of N 2 . After the solution became clear, the temperature was raised to 170 • C, and 0.8~1 mL Cs-oleate (pre-heated to >100 • C) was swiftly injected, which was quenched by immersing the flask in an ice-water bath (5 s later). 0.376 mmol) and OLA were added, and the temperature was raised to 170 • C (30 min later). Then, the prepared Cs-oleate was quickly injected into the mixture and 60~65 s later, the reaction was quenched by immersing the flask in an ice-water bath [12].

Purification
The unsubstituted and YCl 3 -substituted CsPbI 3 NCs were separated by centrifugation at 5000 rpm for 10 min to remove the ODE and unreacted ligands. The precipitate was dispersed in hexane/toluene, and then anti-solvent methyl acetate was mixed, followed by centrifugation at 10,000 rpm for 10 min. The mixture was dispersed in 10 mL hexane/toluene and stored in the refrigerator. After 24 h, the supernatant required colloidal ink for LED application.

LED Fabrication
150 nm thick indium tin oxide (ITO) patterned glass substrates were cleaned by sequential ultrasonication in acetone and isopropanol for 15 min each. Subsequently, the UVozone treatment was carried out for 15 min to improve hydrophilicity before drying with N 2 flow. PEDOT:PSS (filtered with 0.45 µm PVDF filter) was then spin-coated for 1 min at 4000 rpm, which was then annealed at 140 • C for 30 min. All the substrates were transferred into the N 2 -filled glove box. The hole-transporting Poly-TPD (4 mg/mL in CB) layer was spin-coating onto PEDOT:PSS films at 4000 rpm and baked at 120 • C for 15~20 min. The perovskite emissive layer was spin-coated at a speed of 2000 rpm for 45 s. Then, 1,3,5tris(1-phenyl-1H-benzimidazol-2-yl)benzene (TPBi, 40 nm), LiF (1 nm), and Al (130 nm) were sequentially deposited by thermal evaporation into a vacuum deposition chamber. All the devices were encapsulated by a glass lid with a UV-curable resin in an N 2 -filled glovebox. The optoelectronic properties of LEDs were analyzed using a semiconductor parameter analyzer (Keithley 237) connected with a spectrophotometer (Photo Research PR-670). The UV-Vis-IR absorption spectrum of the unsubstituted and YCl 3 -substituted CsPbI 3 NC film was tested using a PerkinElmer LAMBDA-900 spectrophotometer. PLQY was measured by Quantaurus-QY Absolute PL quantum yield spectrometer (Hamamatsu, C11347-11). The structure of the unsubstituted and YCl 3 -substituted CsPbI 3 NCs' thin film was analyzed by an X-ray diffractometer (X'Pert-PRO MRD, Phillips). The shape of the unsubstituted and YCl 3 -substituted CsPbI 3 NCs was confirmed using a field-emission transmission electron microscope (FE-TEM, JEM-2100F) and the cross-sectional TEM images of the LED were obtained from a Cs-corrected TEM (JEM-ARM 200F, JEOL) installed in the Center for University-wide Research Facilities (CURF) at Jeonbuk National University. Chemical analysis was conducted by X-ray Photoelectron Spectrometer (XPS).

Results and Discussion
The control (unsubstituted) and YCl 3 -substituted CsPbI 3 NCs were synthesized following the two-step hot injection method [20]. The YCl 3 -doped CsPbI 3 NRs were realized by adding 0.184 mmol of YCl 3 ·6H 2 O into perovskite medium followed by the injection of Cs-oleate precursor (Scheme 1a). This is unlike the previous reports, which aimed at metal chlorides that used either the identical or adjoining chloride ion in the pristine perovskite nanocrystals [14,21]. This study examines how metal chloride can modulate the shape and optoelectronic properties of CsPbI 3 NCs. Besides changing the surface defect of perovskite NCs, the YCl 3 -doping has an additional effect on inducing the anisotropic growth of the crystals. The intention doping of YCl 3 was carried out in the CsPbI 3 NCs' reaction medium. The distinctive chloride (Cl − ) was not proximate with the iodide (I − ) ions of the CsPbI 3 NCs. The presence of chloride ions on the surface of perovskite nanocrystals (NCs) and the varied bond energies among chloride (Cl − ) ions and iodide (I − ) ions are responsible for the anisotropic growth of perovskite NCs. (Scheme 1b). The transmission electron microscopy (TEM) analyses provide evidence to support this conjecture, as shown in Figure 1. The TEM and high resolution (HR-TEM) morphology of the control and YCl 3 -substituted CsPbI 3 NCs is revealed by TEM ( Figure 1). The control CsPbI 3 NCs contain monodisperse and regular cubic shapes ( Figure 1a). The HR-TEM images show a high crystallinity and lattice spacing of 6.2 Å of the CsPbI 3 NCs, corresponding to the (100) plane of cubic perovskite (Figure 1b,c) [22,23]. The average particle size of the control CsPbI 3 NCs is determined to be~10.05 nm ( Figure 1d). Remarkably, the 0.184 mmol YCl 3 -substituted CsPbI 3 NCs show the one-dimensional nanorods (NRs) (Figure 1e-g). The HR-TEM image of YCl 3substituted CsPbI 3 NRs displays a lattice spacing of 4.5 Å, corresponding to the (110) plane of perovskite (Figure 1g) [12]. The aspect ratio of the YCl 3 -substituted CsPbI 3 NCs is~2.3, and the average sizes of length and diameter are 18.5 and 8.2 nm, respectively. Nanomaterials 2023, 13, x FOR PEER REVIEW 5 of 13 perovskite nanocrystals (NCs) and the varied bond energies among chloride (Cl − ) ions and iodide (I − ) ions are responsible for the anisotropic growth of perovskite NCs. (Scheme 1b). The transmission electron microscopy (TEM) analyses provide evidence to support this conjecture, as shown in Figure 1. The TEM and high resolution (HR-TEM) morphology of the control and YCl3-substituted CsPbI3 NCs is revealed by TEM ( Figure 1). The control CsPbI3 NCs contain monodisperse and regular cubic shapes ( Figure 1a). The HR-TEM images show a high crystallinity and lattice spacing of 6.2 Å of the CsPbI3 NCs, corresponding to the (100) plane of cubic perovskite (Figure 1b,c) [22,23]. The average particle size of the control CsPbI3 NCs is determined to be ~10.05 nm ( Figure 1d). Remarkably, the 0.184 mmol YCl3-substituted CsPbI3 NCs show the one-dimensional nanorods (NRs) (Figure 1e-g). The HR-TEM image of YCl3-substituted CsPbI3 NRs displays a lattice spacing of 4.5 Å, corresponding to the (110) plane of perovskite ( Figure  1g) [12]. The aspect ratio of the YCl3-substituted CsPbI3 NCs is ~2.3, and the average sizes of length and diameter are 18.5 and 8.2 nm, respectively.  Figure 2b,c, the angles of diffraction peaks for the (100) and (200) planes shifted towards higher values attributed to the reduction in the lattice parameters of YCl 3 :CsPbI 3 NRs, which stemmed from the partial substitution lead cation (Pb 2+ ) and iodide ions (I − ) with Y 3+ cation and Cl − ions, respectively [12,24]. The scanning electron microscopy (SEM) images of the unsubstituted and YCl 3 -substituted CsPbI 3 NCs are illustrated in Figure 2d,e. Contrary to perovskite NCs, the SEM image of (YCl 3 -Nanomaterials 2023, 13, 1366 5 of 11 substituted CsPbI 3 ) nanorods are homogeneously distributed on the glass substrate, which indicates that the nanorod film layer has a good foundation for electroluminescence devices. The X-ray diffraction (XRD) patterns were conducted to ascertain the crystal structure of CsPbI3 NCs and YCl3:CsPbI3 NRs, as shown in Figure 2. Both the CsPbI3 NCs and YCl3:CsPbI3 NRs adhere to the reference pattern of the bulk cubic CsPbI3 perovskite (PDF#98-018-1288), and the diffraction peaks, which appear at 14.02°, 20.03°, 28.407°, 31.88°, 35.45°, 41.05°, and 51.63°, are corresponding to cubic planes of (100), (110), (200), (210), (211), (220), and (300), respectively. As shown in Figure 2b,c, the angles of diffraction peaks for the (100) and (200) planes shifted towards higher values attributed to the reduction in the lattice parameters of YCl3:CsPbI3 NRs, which stemmed from the partial substitution lead cation (Pb 2+ ) and iodide ions (I − ) with Y 3+ cation and Cl − ions, respectively [12,24]. The scanning electron microscopy (SEM) images of the unsubstituted and YCl3substituted CsPbI3 NCs are illustrated in Figure 2d,e. Contrary to perovskite NCs, the SEM image of (YCl3-substituted CsPbI3) nanorods are homogeneously distributed on the glass substrate, which indicates that the nanorod film layer has a good foundation for electroluminescence devices. X-ray photoelectron spectra (XPS) analysis was performed to gain insight into the interaction of YCl3 with CsPbI3 nanocrystals (NCs), and the results are presented in Figure  3a. The characteristic XPS signals for Cs 3d, Pb 4f, I 3d, Y 3d, and Cl 2p were observed in YCl3-doped CsPbI3 NCs and N 1s, O 1s, and C 1s signals coupled with native ligand bonding. The high-resolution XPS spectra of Cs 3d, Pb 4f, I 3d, Y 3d, and Cl 2p are displayed in X-ray photoelectron spectra (XPS) analysis was performed to gain insight into the interaction of YCl 3 with CsPbI 3 nanocrystals (NCs), and the results are presented in Figure 3a. The characteristic XPS signals for Cs 3d, Pb 4f, I 3d, Y 3d, and Cl 2p were observed in YCl 3doped CsPbI 3 NCs and N 1s, O 1s, and C 1s signals coupled with native ligand bonding. The high-resolution XPS spectra of Cs 3d, Pb 4f, I 3d, Y 3d, and Cl 2p are displayed in the order in Figure 3b-f. The partial substitution of I − ions by Cl − ions is evidenced by an increase in the binding energies of Pb 2+ 4f 5/2 and Pb 2+ 4f 7/2 from 143.56 and 138.37 eV to 143.64 and 138.7 eV, respectively. The binding energy of I 3d and Cs 3d displays little variation compared to YCl 3 :CsPbI 3 nanocrystals (NCs). More importantly, the binding energy signals of Y 3+ and Cl − can be observed in the YCl 3 -substituted CsPbI 3 nanorods (NRs). These findings provide additional evidence supporting the partial substitution of iodide ions by chloride ions [25]. To gain a better understanding of the effects of the partial substitution of Pb 2+ cation and iodide ions (I − ) with yttrium cation (Y 3+ ) and chloride (Cl − ) ions, the optical properties of both as-synthesized CsPbI3 nanocrystals (NCs) and YCl3:CsPbI3 nanorods (NRs) were analyzed (Figure 4). The corresponding normalized PL and absorption spectra of YCl3:CsPbI3 NRs showed that the peak position exhibited a blue shift owing to the introduction of chloride (Cl − ) ions (Figure 4a,b) [26]. The enlarged bandgap caused the blue shifts of the absorption and PL spectra for YCl3:CsPbI3 NRs due to the partial replacement of lead (Pb 2+ ) cation and iodide (I − ) ions with yttrium cation (Y 3+ ) and chloride (Cl − ) ions [27]. The blue dots in Figure 4b demonstrate the blue shift associated with introducing YCl3 into the CsPbI3 perovskite. The Tauc plot of the unsubstituted and YCl3-substituted CsPbI3 NC films is illustrated in Figure 4c,d, and the corresponding band value of the unsubstituted is 1.76 eV, and the YCl3-substituted is 1.82 eV. The PLQY increased from 51% to 70% for the YCl3 passivated CsPbI3 NCs, suggesting enhanced radiative recombination followed by yttrium chloride doping (Figure 4e). The environmental durability of the YCl3:CsPbI3 solution was noticeably enhanced (Figure 4f), and the YCl3-substituted CsPbI3 NRs maintained (50% out of 70%) a PLQY, with a loss of 28% PLQY after being stored for 45 days under ambient conditions, prized for the effectiveness of yttrium chloride passivation. However, over the same period of time, the PL quantum yield of pristine CsPbI3 NCs nearly approached zero. The inset of Figure 4f shows the images recorded at To gain a better understanding of the effects of the partial substitution of Pb 2+ cation and iodide ions (I − ) with yttrium cation (Y 3+ ) and chloride (Cl − ) ions, the optical properties of both as-synthesized CsPbI 3 nanocrystals (NCs) and YCl 3 :CsPbI 3 nanorods (NRs) were analyzed (Figure 4). The corresponding normalized PL and absorption spectra of YCl 3 :CsPbI 3 NRs showed that the peak position exhibited a blue shift owing to the introduction of chloride (Cl − ) ions (Figure 4a,b) [26]. The enlarged bandgap caused the blue shifts of the absorption and PL spectra for YCl 3 :CsPbI 3 NRs due to the partial replacement of lead (Pb 2+ ) cation and iodide (I − ) ions with yttrium cation (Y 3+ ) and chloride (Cl − ) ions [27]. The blue dots in Figure 4b demonstrate the blue shift associated with introducing YCl 3 into the CsPbI 3 perovskite. The Tauc plot of the unsubstituted and YCl 3 -substituted CsPbI 3 NC films is illustrated in Figure 4c,d, and the corresponding band value of the unsubstituted is 1.76 eV, and the YCl 3 -substituted is 1.82 eV. The PLQY increased from 51% to 70% for the YCl 3 passivated CsPbI 3 NCs, suggesting enhanced radiative recombination followed by yttrium chloride doping (Figure 4e). The environmental durability of the YCl 3 :CsPbI 3 solution was noticeably enhanced (Figure 4f), and the YCl 3 -substituted CsPbI 3 NRs maintained (50% out of 70%) a PLQY, with a loss of 28% PLQY after being stored for 45 days under ambient conditions, prized for the effectiveness of yttrium chloride passivation. However, over the same period of time, the PL quantum yield of pristine CsPbI 3 NCs nearly approached zero. The inset of Figure 4f shows the images recorded at different times for the unsubstituted and YCl 3 -substituted CsPbI 3 NC solution. The control and YCl3-substituted CsPbI3 NCs were employed as emitters to evaluate the potential applications in perovskite LEDs. The LEDs were fabricated based on the configuration of indium tin oxide (ITO)/poly(3,4-ethylenedioxythiophene) polystyrenesulfonate (PEDOT:PSS)/poly(4-butylphenyl-diphenyl-amine) (P-TPD)/unsubstituted CsPbI3 or YCl3:CsPbI3/1,3,5-tris(1-phenyl-1H-benzimidazol-2-yl)benzene (TPBi)/lithium fluoride (LiF)/aluminum (Al). The schematic illustration of PeLED and the corresponding energy levels diagram of the functional layer are illustrated in Figure 5a,b. The functional layers' energy level values are taken from previous literature [12]. The thickness of ITO (150 nm), PEDOT: PSS, P-TPD (60 nm), YCl3:CsPbI3 (55 nm), TPBi (60 nm), and LiF/Al (130 nm) were analyzed by the cross-sectional TEM image, as shown in Figure 5c. The current densityvoltage-luminance (J-V-L) curves of the unsubstituted and YCl3-substituted CsPbI3 NC LEDs with a 4 mm 2 emitting area are displayed in Figure 5d. The turn-on voltage (where the luminance achieved 1 cd/m 2 ) is reduced from ~3.9 V to ~3.6 V for the YCl3-substituted CsPbI3 NC LED, revealing that more balanced carriers injected owing to their matched energy level with the carrier transfer layer and enhanced conductivity induced by the YCl3-substitution that facilitated the efficient charges' injection [12,23]. The CsPbI3 NC and YCl3:CsPbI3 NR-based LEDs showed a maximum luminance of 263.1 cd/m 2 and 421.8 cd/m 2 , respectively. The electroluminescence (EL) spectra of the unsubstituted and YCl3substituted CsPbI3 NC LEDs were observed at 691 nm and 688 nm (Figure 5e). The values are quite different from the PL spectra in the case of the unsubstituted one (Figure 4a). The similar EL spectra of the two devices are attributed to the weak microcavity effect in the LED structure, which is related to the recombination zone of the device. The recombination zone would be at the interface of EML and ETL in both cases due to the dominance of the hole carrier in perovskite EMLs. The YCl3:CsPbI3-based LEDs revealed high color  Figure 5a,b. The functional layers' energy level values are taken from previous literature [12]. The thickness of ITO (150 nm), PEDOT: PSS, P-TPD (60 nm), YCl 3 :CsPbI 3 (55 nm), TPBi (60 nm), and LiF/Al (130 nm) were analyzed by the cross-sectional TEM image, as shown in Figure 5c. The current density-voltage-luminance (J-V-L) curves of the unsubstituted and YCl 3 -substituted CsPbI 3 NC LEDs with a 4 mm 2 emitting area are displayed in Figure 5d. The turn-on voltage (where the luminance achieved 1 cd/m 2 ) is reduced from~3.9 V to~3.6 V for the YCl 3substituted CsPbI 3 NC LED, revealing that more balanced carriers injected owing to their matched energy level with the carrier transfer layer and enhanced conductivity induced by the YCl 3 -substitution that facilitated the efficient charges' injection [12,23]. The CsPbI 3 NC and YCl 3 :CsPbI 3 NR-based LEDs showed a maximum luminance of 263.1 cd/m 2 and 421.8 cd/m 2 , respectively. The electroluminescence (EL) spectra of the unsubstituted and YCl 3 -substituted CsPbI 3 NC LEDs were observed at 691 nm and 688 nm (Figure 5e). The values are quite different from the PL spectra in the case of the unsubstituted one ( Figure 4a). The similar EL spectra of the two devices are attributed to the weak microcavity effect in the LED structure, which is related to the recombination zone of the device. The recombination zone would be at the interface of EML and ETL in both cases due to the dominance of the hole carrier in perovskite EMLs. The YCl 3 :CsPbI 3 -based LEDs revealed high color purity with Commission internationale de l'éclairage (CIE) coordinates of (0.71, 0.26), corresponding to the BT. 2020 color gamut, as shown in Figure 5f. The EQE vs. luminance curves are displayed in Figure 5g. The peak EQE of YCl 3 :CsPbI 3 is 3.16%, 1.86-fold higher than the pristine CsPbI 3 NC (1.69%) based LED. This enhancement is attributed to enhanced PLQY and more balanced carrier transfer in the YCl 3 :CsPbI 3 EML layer.
Nanomaterials 2023, 13, x FOR PEER REVIEW 9 of 12 purity with Commission internationale de l'éclairage (CIE) coordinates of (0.71, 0.26), corresponding to the BT. 2020 color gamut, as shown in Figure 5f. The EQE vs. luminance curves are displayed in Figure 5g. The peak EQE of YCl3:CsPbI3 is 3.16%, 1.86-fold higher than the pristine CsPbI3 NC (1.69%) based LED. This enhancement is attributed to enhanced PLQY and more balanced carrier transfer in the YCl3:CsPbI3 EML layer.  To further investigate the enhanced EQE in the YCl 3 -substituted CsPbI 3 NRs, we conducted angle-dependent photoluminescence (ADPL) measurements to probe the orientation of transition dipole moments (TDMs) in assembled thin film of CsPbI 3 nanocrystals/nanorods. The outcoupling efficiency in PeLEDs has the potential for improvement through controlling the orientation of TDMs [28][29][30]. The optical TDMs of nanoplatelets and nanorods are highly anisotropic, and outcoupling efficiency in planer PeLEDs is profoundly associated with the orientation of emissive TDMs [15,17]. The orientation of the optical TDMs of the unsubstituted and YCl 3 -substituted CsPbI 3 NCs were measured by the ratio of horizontal TDMs (Θ). The experimental data are fitted to the pattern simulated, employing the classical dipole radiation model [31]. The Θ values of CsPbI 3 NCs' and YCl 3 :CsPbI 3 NRs' films are determined to be 67% and 75% (Figure 6a,b). The Θ value in anisotropic nanorods is considerably higher than that in isotropic nanocrystals. Thus, the optical TDMs that are horizontally oriented in anisotropic nanorods are preferred for light outcoupling, resulting in the improved EQE of the LEDs. To further investigate the enhanced EQE in the YCl3-substituted CsPbI3 NRs, we conducted angle-dependent photoluminescence (ADPL) measurements to probe the orientation of transition dipole moments (TDMs) in assembled thin film of CsPbI3 nanocrystals/nanorods. The outcoupling efficiency in PeLEDs has the potential for improvement through controlling the orientation of TDMs [28][29][30]. The optical TDMs of nanoplatelets and nanorods are highly anisotropic, and outcoupling efficiency in planer PeLEDs is profoundly associated with the orientation of emissive TDMs [15,17]. The orientation of the optical TDMs of the unsubstituted and YCl3-substituted CsPbI3 NCs were measured by the ratio of horizontal TDMs (Θ). The experimental data are fitted to the pattern simulated, employing the classical dipole radiation model [31]. The Θ values of CsPbI3 NCs' and YCl3:CsPbI3 NRs' films are determined to be 67% and 75% (Figure 6a,b). The Θ value in anisotropic nanorods is considerably higher than that in isotropic nanocrystals. Thus, the optical TDMs that are horizontally oriented in anisotropic nanorods are preferred for light outcoupling, resulting in the improved EQE of the LEDs.

Conclusions
In this study, we investigated the effects of incorporating metal chloride (YCl3) into CsPbI3 nanocrystals to control the dimensions. We found that the incorporation of YCl3 led to a decrease in the lattice parameters of the CsPbI3 nanocrystals, which resulted from the partial substitution of larger lead cation (Pb 2+ ) and iodide (I − ) ions with smaller Y 3+ and Cl − ions. The presence of Cl − ions on the surface of the NCs, coupled with the difference in bond energies between chloride (Cl − ) and iodide (I − ) ions, led to the anisotropic formation of the CsPbI3 nanocrystals into one-dimensional (1D) nanorods. The YCl3 also significantly improved the photoluminescence quantum yield and storage lifetime of the perovskite solution by passivating nonradiative recombination rates and defects properly. Finally, we used the YCl3-substituted CsPbI3 nanorods as the emissive layer in red LEDs and observed a significant improvement in their performance. The LEDs exhibited an external quantum efficiency of 3.16% which is 1.86-fold higher than the pristine CsPbI3 NC (1.69%) based LED, attributed to the improvement of the ratio of horizontal TDMs in the anisotropic YCl3:CsPbI3 nanorods to 75% from 67% of that of NCs. Overall, the combined characteristics of YCl3-substituted CsPbI3 nanorods show great potential for developing stable and efficient red LEDs.

Conclusions
In this study, we investigated the effects of incorporating metal chloride (YCl 3 ) into CsPbI 3 nanocrystals to control the dimensions. We found that the incorporation of YCl 3 led to a decrease in the lattice parameters of the CsPbI 3 nanocrystals, which resulted from the partial substitution of larger lead cation (Pb 2+ ) and iodide (I − ) ions with smaller Y 3+ and Cl − ions. The presence of Cl − ions on the surface of the NCs, coupled with the difference in bond energies between chloride (Cl − ) and iodide (I − ) ions, led to the anisotropic formation of the CsPbI 3 nanocrystals into one-dimensional (1D) nanorods. The YCl 3 also significantly improved the photoluminescence quantum yield and storage lifetime of the perovskite solution by passivating nonradiative recombination rates and defects properly. Finally, we used the YCl 3 -substituted CsPbI 3 nanorods as the emissive layer in red LEDs and observed a significant improvement in their performance. The LEDs exhibited an external quantum efficiency of 3.16% which is 1.86-fold higher than the pristine CsPbI 3 NC (1.69%) based LED, attributed to the improvement of the ratio of horizontal TDMs in the anisotropic YCl 3 :CsPbI 3 nanorods to 75% from 67% of that of NCs. Overall,