Efficient and stable blue perovskite light-emitting diodes through I-III-VI quantum dot solids as hole transport layer

doi:10.21203/rs.3.rs-5361263/v1

Download PDF

Article

Efficient and stable blue perovskite light-emitting diodes through I-III-VI quantum dot solids as hole transport layer

https://doi.org/10.21203/rs.3.rs-5361263/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Perovskite light-emitting diodes (PeLEDs) have achieved skyrocketing progress in material and device efficiencies. However, inferior stabilities of pure blue devices, remain major hurdles towards full-color displays. Herein, we built PeLEDs based on quasi-two-dimensional (quasi-2D) perovskites using chalcopyrite I-III-VI semiconductor quantum dot (QD) solids as novel inorganic hole transport layer (HTL), to overcome the stability issues in blue PeLEDs. Wide-gap silver-copper-gallium-disulfide (ACGS) QDs were dedicatedly-synthesized aiming for enhanced hole transport efficiency in QD solids through adaptable band structure and surface chemistry engineering, resulting in band-like hole transport with a high mobility of 0.546 cm² V^− 1s^− 1 in the linear working scheme. In addition, the Lewis base group attached to the QD surface (Cl⁻, RS⁻) lower the defect density through buried interface passivation on uncoordinated Pb²⁺ in perovskite, which effectively regulate crystallization kinetics of quasi-2D perovskite. Furthermore, halide interstitial defects were stabilized by Lewis acids group (Zn²⁺) capped on the surface of ACGS QDs, preventing ionic migration and deep-level trap formation. As a result, the champion pure-blue PeLEDs based on ACGS QD solids exhibit preeminent operating lifetime (T₅₀@100 cd/m² = 78 min) for electroluminescence (EL) peak emission wavelength at 471 nm, with maximum external quantum efficiency (EQE) of 10.85%.

Physical sciences/Materials science/Nanoscale materials/Electronic properties and materials

Physical sciences/Nanoscience and technology/Nanoscale devices/Electronic devices

Physical sciences/Materials science/Nanoscale materials/Quantum dots

From the demonstration of first room-temperature electroluminescent perovskite light-emitting diodes (PeLEDs) in 2014, the past full decade have witnessed sustained development of luminescent perovskite materials, as well as skyrocketing progress on highly-efficient and stable devices, shining as promising alternatives to future lighting and displays ^1–10. Despite of significantly-improved device efficiencies in red, green and near-infrared perovskite emitters, the major research and development obstacles relate to the inferior device performances of blue PeLEDs, especially for saturated primary blue color required by displays^11–17. To realize pure- or deep-blue emission (less than 475 nm), spatially-confined excitons as well as compositional engineering are technically required ^18–22. Quasi-two-dimensional (quasi-2D) perovskites offering larger exciton binding energy (hundreds of meV), and lower ion migration mobility compared to 3D counterparts, have gained great success in blue PeLEDs ^23–26. However, the comprehensive performances of blue PeLEDs based on quasi-2D perovskites lag behind quantum-confined nanocrystals. Firstly of all, the distribution of energy domains is inhomogeneous, due to disordered crystallizations of small-n phases (n ≤ 3), which is harmful for exciton energy transfer from small-n phases to large-n phases, where targeted recombination would take place ^{3, 27–30}. Second, the phase separation in Br/Cl mixed halide perovskite due to ionic migration via halide vacancy defect pathways would induce deep-level traps and exciton quenching in quasi-2D perovskite ^31–33. Furthermore, the charge accumulations due to imbalanced electron and hole transport are even more devastating in quasi-2D perovskite due to enormous large cations intercalation, which would cause exciplex emission in hole transport layer (HTL) and interfacial non-radiative recombination. Though intense effort has been spent on alleviating the first two issues through additive engineering, and process optimization to realize efficient exciton energy transfer, only few works focused on balancing charge transport and suppressing interfacial recombination ^34–36.

Regarding to the HTL in solution-processed PeLEDs, conventional organic polymer including poly(9,9-dioctyl-fluorene-co-N-(4-(3-methylpropyl))diphenylamine (TFB), and Poly(N-vinyl carbazole) (PVK) have been widely implemented. However, low hole mobility and stabilities of these organic materials severely affects the performances of blue PeLEDs. To balance the charge transport in PeLEDs, inorganic nickel oxide (NiO_x) has been developed as highly-conductive inorganic HTL in PeLEDs ^37–39. Xie’s group modulated the composition of NiMgO, and fabricated effective green PeLED with a record EQE of 30.84% and operational lifetime of over 100 min at initial luminance of 100 cd/m², revealing enormous potential of inorganic hole transporting materials for high-performance PeLEDs ⁴⁰. Nevertheless, EL efficiencies of NiO_x-based PeLEDs are dramatically limited by inefficient hole injection and long-range exciton quenching at the NiO_x–perovskite interfaces ^{38, 41–43}. Therefore, developing alternative inorganic hole functional materials is necessary to overcome the imbalanced charge transport limitations and interfacial exciton quenching of PeLEDs.

Herein, we seek to realize highly-efficient and stable pure-blue PeLEDs by implementing semiconductor CuGaS₂ (CGS) QDs and their derivatives as novel hole transporting materials. Chalcopyrite I-III-VI material family (CuInSe₂, CuInGaS₂) are intrinsic p-type semiconductors, featuring high hole mobility and tunable bandgap (E_g) ⁴⁴. Compared to NiO_x, of which the carrier transport properties are mainly determined by concentration of Ni³⁺, CGS QDs could be dedicatedly engineered via stoichiometry, doping, QD size and surface chemistry ^45,46. We show that strong electrical coupling could be realized in CGS QD solids by atomistic ligand exchange, leading to exceptionally-high hole concentration (5.81×10¹⁵ cm^− 3) and mobility (0.546 cm² V^− 1s^− 1) in the linear working scheme. Most importantly, the surface X-type ligand (RS⁻, Cl⁻) of QD as Lewis base group could bind to uncoordinated Pb²⁺ in quasi-2D perovskite, which promote the formation of underlying “crystal seeds”, accelerating ordered nucleation and growth of quasi-2D perovskite crystals. In addition, the Z-type surface ligand embedding Lewis acid (Zn²⁺) of QD could stabilize halide interstitials defects, suppress ionic migration and deep-level trap formation. As a result, quasi-2D pure-blue PeLEDs demonstrate improved luminescent performances with peak external quantum efficiencies (EQE) of 10.85% at 471 nm, and operating lifetime of 78 min (T₅₀@100 nit), significantly higher compared to devices based on NiO_x (18 min).

Synthesis and characterizations of I-III-VI semiconductor quantum dot

CGS core QDs were synthesized and assembled into QD solids as HTL in PeLEDs as depicted in Fig. 1. Supplementary Fig. 1a shows the synthetic scheme and material characterizations of chalcopyrite QD prepared in this work. From high-resolution transmission electron microscopy (HR-TEM) images, CGS QD exhibit uniform spherical shape with average size of 4.4 nm (Supplementary Fig. 1b), comparable to the Bohrs exciton radius (~ 4 nm) of chalcopyrite nanocrystals ⁴⁷. X-ray diffraction (XRD) patterns of CGS QD present dominant (112) planes attributed to chalcopyrite phase. Fast Fourier Transform (FFT) analysis reveal single-crystal character of CGS QDs, consistent with chalcopyrite structure in the [1–10] direction.

Recent reports reveal that doping with silver (Ag) into chalcopyrite prompts uniform nucleation and reduces deep-level lattice defects ⁴⁸. Moreover, due to orbital overlap between Ag 3d electrons and Cu 4d electrons, the valence band maximum (VBM) level is lowered, which would facilitate reduced hole transport barrier with quasi-2D perovskite. Herein, Ag doping was conducted by incorporating an appropriate amount of silver iodide (AgI) into metallic precursors. The purified ACGS QD exhibits similar spherical shape without affecting size distributions. Energy-dispersive X-ray spectroscopy (STEM-EDX) show ACG (([Ag]+[Cu])/[Ga]) ratio in the final product is 1.22 as the feed ACG ratio in the precursor is 1:4 (Supplementary Fig. 1c & Supplementary Table 1). Phase evolution from AgGaS₂ to CuGaS₂ by varying the [Ag]/([Ag]+[Cu]) (AAC) ratio was assessed by XRD diffraction (Supplementary Fig. 1d). Significant influences of chemical composition on the bandgap energy (E_g) of ACGS QD solids were reflected in ultra-violet visible (UV-vis) absorption spectrum. E_g increases from 2.17 eV to 2.78 eV as a deficiency in mono-cation ([Cu]/[Ga]) is increased from 1:1 to 1:8 (AAC = 0) ; and increases conspicuously from 2.58 eV to 2.74 eV as AAC increases from 0 to 10% (ACG = 1:4) (Supplementary Fig. 1e-f). Besides, the appearance of excitonic peak indicates mono-dispersion and homogeneous size distribution ⁴⁹. Improved crystallinity on Ag doping can be concluded from observable yellow luminescence under UV excitation at 365 nm, shown in the inset picture of Supplementary Fig. 1g.

It is notable that the monocation-poor precursor condition (ACG ratio or [Cu]/[Ga] ≤ 1:2) is necessary to maintain quaternary/ternary chalcopyrite phase in the final QD. Otherwise, the final product is pure chalcocite Cu₂S as confirmed by both TEM and XRD measurement. (Supplementary Fig. 2) Detailed investigations into the nucleation and growth process of QD affirmed the formation of crystalline Cu₂S/Ag₂S at early stages of synthesis. (Supplementary Fig. 3) As temperature rises, the long-wavelength absorption disappears accompanied by a blueshift of the excitonic peak, suggesting a phase transition from binary low-gap Cu₂S/ Ag₂S, to quaternary wide-gap AgCuGaS₂, due to Ga in-diffusion and Ag out-diffusion (Supplementary Table. 2).

Functional surface ligand engineering and improved hole transport in QD solids

Prior to use ACGS QD solids as HTL in PeLEDs, we developed hybrid ligand exchange technology to replace the long-bulky ligand (RNH_2, RS⁻) on the surface of QDs, by short, atomistic ligand ⁵⁰, as illustrated in Fig. 2a.

At the beginning, ZnCl₂ was intentionally selected as solution-phase ligand treatment due to similar lattice structure between chalcopyrite and zinc-blende ⁵¹. Specifically, it has been widespread reported about passivation effects on uncoordinated Pb²⁺ sites with Lewis base through the formation of Lewis adducts ^{21,25,52–54}. Conceptually, electronegative Lewis base Cl⁻ as X-type ligands provide covalent bond formation with electrophilic surface sites of ACGS QDs (Cu⁺, Ga³⁺) ⁵⁵. In addition, the incorporation of Lewis acids into mixed-halide quasi-2D perovskite could reduce the detrimental effects of chlorine on deep-level defects, and avoid phase separation ^13,56,57. Simultaneously, ZnCl₂ as Z-type ligands could bind to electron-rich Lewis basic sites of ACGS QDs through Lewis acids Zn²⁺ as two-electron acceptors. The effects of QD surface ligand is representatively depicted in Fig. 1b.

ZnCl₂ dissolved in ethanol was added into roughly-purified QD solutions to initiate ligand interactions (see Methods for more detail). Before ligand exchange, pristine ACGS QD exhibit distinguished lattice distortions and stacking faults due to steric hindrances, which could be effectively recovered with improved crystallinities characterized by obvious decrement of inter-dot spacing and uniform distribution after ZnCl₂ treatment, which would promote stronger electrical coupling among dots (Fig. 2b and Supplementary Fig. 4a-c). The crystallography characterizations by XRD presents no significant variations in the diffraction pattern after ligand exchange, indicating the process is a surface-limited interaction, instead of alloying or doping into the core (Supplementary 4d). To understand microscopic origins of Cl⁻ passivation, we used density functional theory (DFT) to calculate the binding energies (E_b) of amines (RNH₂) and thiolates (RS⁻) and Cl⁻ ion on the CGS surface. DFT calculations unveil that Cl⁻ could bind to surface Cu⁺ and Ga³⁺ with strong Lewis acidity, reconstructing ordered cubic structure from distorted crystal surface due to large steric hindrances of aliphatic ligand (Fig. 2c). Furthermore, from density of states (DOS) calculations, the mid-gap trap states are remarkably suppressed and E_b is significantly improved after ZnCl₂ treatment (Fig. 2d and Supplementary 5b). X-ray photoelectron spectroscopy (XPS) studies were conducted to characterize the chemical environment changes of elements in ACGS QD on ligand exchange. As expected, Zn 2p and Cl 2p were distinctly identified in ZnCl₂-treated ACGS QD (Supplementary Fig. 6). In addition, both Cu 2p 1/2 and Ga 3d emission peaks, as well as corresponded auger emissions (Cu LMM, Ga LMM) slightly-shift to higher binding energy regime, due to bonding with electron-rich Cl⁻ ions. The metal nitride peaks, evidenced from resolved Ga 3d (396.8 eV) and N 1s (25.68 eV) core-level emission, are efficiently eliminated in ACGS QDs after ZnCl₂ treatment (Supplementary Table 3). Quantitative analysis from XPS spectra indicate reduced atomic concentration of nitrogen (Supplementary Table. 4). These results clearly indicate effective exchange of surface amines by Cl⁻ ions. Nevertheless, ZnCl₂ cannot sufficiently remove residue thiolates ligand from 1-DDT used to lower the reactivity of Cu⁺, which can be recognized from thermogravimetric analysis (TGA) and fourier transform infrared spectroscopy (FTIR) analysis, in which the methyl and methylene vibrations are still present in ACGS-ZnCl₂ QD. (Supplementary Fig. 7a-b). To fully stripe the un-exchanged alkyl ligand on ACGS QD, we explored solid-phase ligand exchange process to produce homogenous QD solids using 1,2-ethanedithiol (EDT) containing double RS⁻ group ⁵⁸. Likewise, DFT simulations show that EDT exhibits enhanced coordination with surface gallium, leading to increased E_b from 1.049 eV to 9.747 eV, and fully-passivated defect states. High-resolution XPS spectrum suggest trivial influences on chemical bonding and atomic concentrations of elements in ACGS QDs. Meanwhile, the Cu and Ga concentration is declined after ZnCl₂ treatment, as summarized Supplementary Table 3. With observations of decrement of QD size and blue-shifted absorption spectrum of ACGS-ZnCl₂ QD (Supplementary Fig. 7c), it is reasonable to address that the crystallinity of ACGS QD is mainly improved by chlorine ions via exchange of surface long-bulky ligand and desorption of the surface metallic complexes ⁵⁹. As ZnCl₂ removes ligand excessively, ACGS QDs tend to aggregate in the solution, detected by dynamic light-scattering (DLS) measurement (Supplementary Fig. 7d). The influences of ligand density and film washing with EDT on surface morphology of QD solids were monitored through scanning electron microscopy (SEM), and atomic force microscopy (AFM) (Supplementary Fig. 8). As the ligand to QD ratio exceeds 100:1, the deposited films become rough and fuzzy, which is detrimental for light out-coupling. The optimized ligand to QD ratio, is limited to be 100:1 for ZnCl₂ treatment (see Methods). Consequent solid-phase exchange is optimized to be single coating with EDT (1 vol% in acetonitrile). The as-synthesized and exchanged ACGS QDs by ZnCl₂ and EDT is named as ACGS-pristine, ACGS-ZnCl₂, ACGS-ZnCl₂-EDT, respectively.

Next, we built up a thin-film transistor model with bottom-gate-top-contact structure using QD solids as transport channel (QD-TFT) to quantitatively assess the carrier mobility and concentration. Figure 2e depicts the device structure and the picture of SiO₂ substrate as well as SiO₂/ACGS QDs under optical microscope is shown in Supplementary Fig. 9a. By altering the concentration of QD dispersions, the thickness of QD solids can be practically controlled (Supplementary Fig. 9b).

Hole-dominant transport can be surly confirmed from transfer characteristics (I_DS-V_GS) in QD-TFT devices (Fig. 2f). I_DS collected from QD-TFT based on ACGS-pristine is significantly increased from 0.1 nA to 20 nA after ZnCl₂ treatment, and further elevated to hundreds of nano-ampere (nA) after EDT treatment. Corresponding output characteristics (I_DS-V_DS) of QD-TFT based on ACGS-ZnCl₂-EDT QD solids are depicted in Fig. 2g, signifying saturation region while V_DS exceeds − 2 V. The influences of chemical composition on transfer characteristics of QD-TFT based on ACGS QD are reflected in Supplementary Fig. 10a-c. In QD-TFT based on CGS QD, I_DS increases as [Cu]/[Ga] ratio increases. This can be explained as the formation of large amounts of nano-crystalline Cu_xS phase in CGS QD as [Cu]/[Ga] ratio increases. Contrarily, for ACGS QD, I_DS decreases as AAC ratio increases, which can be explained as reduced shallow, acceptor-like V_Cu (copper vacancy) defects therefore decreasing the hole current.

The corresponded hole mobility (µ_h) in the linear working region was calculated based on the transfer characteristics (see Methods). For comparison, the transfer characteristics of NiO_x, TFB, and PVK were also measured and summarized (Supplementary Fig. 10d & Fig. 2h). The calculated µ_h in ACGS (AAC = 10%), and CGS QD solids could reach 0.191 cm² V^− 1s¹ and 0.546 cm² V^− 1s¹, respectively, compared to that of 10^− 5 for organic HTL and 0.127 cm²V^− 1s¹ for NiO_x thin-films. With µ_h calculated, the hole concentration (p_h) was predicted through current-voltage (J-V) characteristics of the two-terminal device structure: glass/ITO/ACGS QD/MoO₃/Ag (Supplementary Fig. 10e). Supplementary Table 5 summarizes µ_h and p_h for chalcopyrite QD solids after hybrid ligand exchange prepared in this work. ACGS QDs exhibits nominal hole mobility but relatively lower hole concentration than organic HTL.

Apart from quantitative analysis of carrier mobility and concentration, the band structure of ACGS QD solids was obtained through ultra-violet spectroscopy (UPS) and kelvin probe force microscopy (KPFM). The effects of chemical stoichiometry, surface chemistry, dimension of QD on work function (WF) and valence band maximum (VBM) were shown in Supplementary Fig. 11a-c. Firstly, Ag doping lowers the VBM energy level due to reduced hole concentration. Specifically, as ACG ratio is 1:2 and the AAC ratio increases from 0 to 10%, the VBM deepens from − 4.53 eV to -4.93 eV. Secondly, increased Cu deficiency leads to remarkably-deepened Fermi-level position (E_F) and the VBM. For instance, as AAC ratio is 0, and the [Cu]/[Ga] ratio decreases from 1:2 to 1:8, the VBM is lowered from − 4.93 eV to -5.37 eV. Then, the quantum confinement effects also affect the band structure; when we synthesized larger ACGS QDs (average size ~ 5.37 nm) using chloride precursors (CuCl, GaCl₃) remaining other parameters unchanged (ACG = 1:2, AAC = 10%), the VBM is upshifted to -4.50 eV compared to -4.93 eV for smaller QDs (Supplementary Fig. 11d-e). Lastly, the surface ligand treatment also elevates VBM as compared to ACGS-pristine. Thus, it is notable that the band structure of chalcopyrite QD solids could be feasibly manipulated through composition engineering, ligand engineering and quantum confinement effects (Supplementary Fig. 11f). Next, the microscopic WF level of quasi-2D perovskite films as well as various HTL were estimated through surface potential mapping measured with kelvin probe microcopy (KPFM) (see Methods). It is calculated that E_F of ACGS QD lies between PEDOT:PSS and perovskite, signifying its critical role as an energy cascade to assist hole injection from PEDOT:PSS to perovskite (Supplementary Fig. 12).

Buried interface passivation and regulated crystallizations of quasi-2D perovskite

After thorough investigations of ACGS QDs and their electrical properties, the influences of ACGS QDs on the interfacial passivation effects on recombination lifetime and crystallizations of quasi-2D perovskite were investigated through time-resolved photoluminescence (TRPL) and XPS. Excitons trapped by defect states attributed to uncoordinated Pb²⁺ are subject to non-radiative recombination ^25,60.

Herein, evidence of chemical coordination between ACGS QDs and Pb²⁺ were assessed by high-resolution XPS on Pb 4f, Br 3d and Cl 2p photoemissions (Fig. 3a-c). The perovskite deposited on PEDOT:PSS substrate is indicated as “control” sample, compared to that deposited on ACGS-ZnCl₂-EDT QD solids. From Fig. 3a, the core-level peaks of Pb 4f shift to higher binding energies, due to increased electron-affinity of Cl⁻ and RS⁻ possessing odd number of valence-shell electrons ⁶⁰. Similarly, the photoemission peaks of Cl 2p _1/2 (199.48 eV) and Cl 2p _3/2 (197.88 eV), as well as Br 3d _1/2 (69.26 eV) and Br 3d _1/2 (68.33 eV) both shift to higher binding energies by 0.3 eV, due to charge disturbance from electron donations of Cl⁻ and RS⁻, and chemical bonding with Zn²⁺ which subsequently results in the change of static interactions with Pb²⁺ cations. In addition, the Zn²⁺ cations attached to the surface of ACGS QDs were detected to diffuse into the perovskite during the annealing process (Supplementary Fig. 13). Therefore, it is concluded that the incorporation ACGS QD solids facilitate in eliminating uncoordinated Pb²_, leading to the formation of “perovskite crystal seeds” at buried HTL/perovskite interface, favoring ordered and homogenous crystallizations of quasi-2D perovskite ^61–63. Furthermore, the stabilization of halide interstitials via Zn²⁺ incoporation could alleviate vacancy defect formation which tends to provide major ionic migration pathways in mixed-halide perovskite ^{32, 64,65}. Similar observations have been reported for alkaline salts, Rb⁺ and K⁺ to suppress the detrimental effects of chlorine in photo-induced phase separation ^13,66. Due to suppressed interfacial exciton quenching and better defect passivation, the overall steady-state photoluminescence (PL) intensity of quasi-2D perovskite films crystalized on ACGS-ZnCl₂-EDT QD solids is significantly improved compared to those on ACGS-pristine, as well as PEDOT:PSS substrate (Fig. 3d).

The charge carrier dynamics at the perovskite/HTL hetero-junction sites and within the quasi-2D perovskite were examined by time-resolved PL (TRPL) analysis, as shown in Fig. 3e. We extracted an average recombination lifetime of 6.46 ns, 1.26 ns an 1.76 ns for perovskite deposited on PEDOT:PSS (control), PEDOT:PSS/ACGS-pristine and PEDOT:PSS/ACGS-ZnCl₂-EDT QD solids, respectively. The extracted PL decay lifetimes (τ₁, τ₂) and their proportionality coefficients (A₁, A₂) are presented in Supplementary Table 6 using second-order exponential functions (see Methods). In typical quasi-2D perovskite, the fast process (τ₁) is usually considered to be the lifetime of excitonic emission ⁶⁷. The proportionality coefficient of fast-decay component (A₁), increases to 0.81 for perovskite deposited on ACGS-ZnCl₂-EDT substrate, compared to 0.50 and 0.75 for that deposited on PEDOT:PSS, and ACGS-pristine substrate, respectively, indicating strengthened radiative recombination in perovskite. In addition,τ₂ fitted for perovskite deposited on ACGS-ZnCl₂-EDT QD solids substrate (3.98 ns) is strongly improved than that on ACGS-pristine QD (1.88 ns), reflect suppressed non-radiative recombination due to the formation of traps within the perovskite. We also performed TRPL analysis on quartz glass substrate to exclude the effects of charge transport (Supplementary Fig. 14 & Supplementary Table 7). As expected, the overall lifetime is shortened from 9.19 ns (quartz glass) to 1.54 ns (ACGS-pristine) and 4.13 ns (ACGS-ZnCl₂-EDT), due to effective charge extraction of ACGS QD solids. Enhanced proportionality of fast-decay component (A₁), as well elevated slow-decay lifetime (τ₂) for perovskite deposited on ACGS-ZnCl₂-EDT QD solids substrate, suggests accelerated excitonic radiative recombination, and suppressed non-radiative recombination due to regulated crystallization kinetics of quasi-2D perovskites on ACGS-ZnCl₂-EDT QD solids. As hydrophilicity of ACGS-ZnCl₂-EDT QD solids is also increased compared to PEDOT:PSS from contact angle measurement (Supplementary Fig. 15), leading to homogeneous film coverage and uniformly-accelerated nucleation process thus significantly-enlarged crystal dimension of quasi-2D perovskite (Fig. 3f ) ⁶⁸. The reduced grain boundaries also contributed to suppressed ionic migration and stable perovskite structure.

UV-vis spectra showed distinct absorption peaks attributed to the small-n (n = 1, 2, 3) phases (Supplementary Fig. 16a). XRD diffraction pattern shows the diffraction peaks at 15.6^o and 31.3^o corresponding to (100) and (200) plane of three-dimensional CsPbBr₃ perovskites on both PEDOT:PSS and ACGS QD substrate. While, diffraction peks at 12.8^o and 27.8^o indicating quasi-2D phases (Supplementary Fig. 16b). The negligible influences of ACGS QD solids on phase distribution of quasi-2D perovskite substantiate that the improved photoluminescence and exciton recombination primarily come from ordered crystallizations and passivated interface with charge transport layer.

Defect physics and highly-efficient, stable blue Pe-LED

To corroborate advantageous effects of ACGS QD solids on carrier transport and interfacial passivation in PeLEDs based on quasi-2D perovskite, blue PeLEDs devices were prepared with the structure: glass/ indium tin oxide (ITO)/PEDOT:PSS/ACGS QD/perovskite/1,3,5-tris(1-phenyl-1H-benzimidazol-2-yl)benzene(TPBi)/LiF/Al (device A), glass/ ITO/PEDOT:PSS/PVK/perovskite/TPBi/LiF/Al (device B) and glass/ ITO/ACGS QD/PVK/perovskite/TPBi/LiF/Al (device C) in which ACGS QD is short for ACGS-ZnCl₂-EDT QD solids. Motivated from recent success of “self-assembled monolayer (SAM)” as HTL in highly-efficient perovskite photovoltaics ⁶⁹, we sought to incorporate a thin-layer of ACGS QD solids on top of PEDOT:PSS. This interfacial layer is supposed to play the role of selective charge transport agent; also be able to suppress the formation of bulk defects through surface modifications.

The device architecture (device A) was optimized based on the band landscape as schematically drawn in Fig. 4a, and the corresponding energy band level of quasi-2D perovskite were obtained from UPS measurement (Fig. 4b & Supplementary Fig. 17). Figure 4c-d & Supplementary Table 9 demonstrate electroluminescent performances of device A, B and C. From Fig. 4c, device C exhibit shunting diode and low EQE. Compared to device B, device A preserves rapidly-increased current density after reaching turn-on voltage (3.3 V), and suppressed efficiency roll-off due to increased hole conductivities of ACGS QD solids, preventing energy loss in HTL and enhancing hole injection into the perovskite emissive layer, thus leading to a maximum EQE and luminance of 10.85%, and 1450 cd/m², respectively. Figure 4e shows the EL spectrum of our champion blue PeLEDs based on ACGS QDs. The EL emission peak at 471 nm with narrow emission bandwidth of 21 nm matches well PL spectrum (Fig. 3d), indicating balanced charge transport dynamics. The spectral stabilities of EL emission strongly support the critical advantages in preventing exciplex emission and stability enhancement of ACGS QD solids. A histogram of the maximum EQE values for 20 devices (Fig. 4f) shows an average EQE of 9.71%, compared to that of 4.04% for device B with PVK as HTL. Figure 4g-h display the device performances of device A with varied AAC ratio of ACGS QD since the applied AAC ratio during QD synthesis is of vital importance in manipulating hole mobility and hole transport barrier of QD solids. To achieve the best EQE, an optimized AAC ratio between 15% and 20% could be verified, above which dramatically-dropped hole mobility is responsible for current and EQE losses for PeLEDs. To highlight the stability enhancement of PeLEDs due to ACGS QDs, we also prepared NiO_x-based PeLEDs (device D), with the structure of glass/ indium tin oxide (ITO)/NiO_x/PVK/perovskite/TPBi/LiF/Al (Supplementary Fig. 18). The measured operational lifetime when the luminance dropped to half of the initial value at 100 cd/m² (T₅₀@100nit) of device A, is significantly extended to 78 min, which is more than four times longer than device D (18 min) (Fig. 4i).The enhanced stabilities indicate optimal charge management and effective blocking effects of impurities from ITO and PEDOT:PSS substrate. The problematic operating lifetime for NiO_x has been reported due to weak attachment to PVK and disordered molecular interactions, which result in significant interfacial defects and exciton quenching ⁶⁸. The overall device performances achieved herein were summarized in supplementary Table 8, and compared to other reported highly-efficient pure blue PeLEDs reported in the literature, manifesting as one of the world top-class devices (Supplementary Table 9).

Despite of comprehensive performance enhancement in blue PeLEDs on introducing ACGS QD solids as HTL, we aim to quantitatively validate their universal effects on hole transport efficiency and defect passivation of quasi-2D perovskite, from device-level electrical diode analysis. Thermal admittance spectroscopy (TAS) and capacitance-profiling (C-V) were conducted on LED devices and hole-only devices (HOD) based on the architecture of device A and B. From raw capacitance-frequency (C-f) characteristics provided in Supplementary Fig. 19, The activation energy (E_A) of deep-level defect could be extracted, which is 0.41 eV (device A) and 0.60 eV (device B), respectively (Fig. 5a). Moreover, the calculated density of states for these deep-level traps is 5.42×10¹⁶ cm^− 3·eV^− 1 (device A), and 6.16×10¹⁷ cm^− 3·eV^− 1 (device B), respectively (Fig. 5b). Supplementary Fig. S20 plot the DOS distribution, from which energetically-narrow defect distribution and reduced maximum defect density revealed for device A suggests substantially-weakened trap-assisted non-radiative recombination. The detailed calculations on E_A and DOS are provided in Methods.

From C-V profiling of HOD in Fig. 5c and Mott-Schottky (MS) analysis in Supplementary Fig. S21a, the built-in potential (V_bi) of HOD based on ACGS QD solids is 0.21 V, profoundly-lower than that based on PVK (0.75 V), due to higher WF level of ACGS QDs resulting in reduced Fermi-level splitting with perovskite (Fig. 5d). Thus, ACGS QD facilitates in overall reduced band-bending and mitigated hole transport barrier. Consistently, the capacitance value of HOD based on ACGS QD (device A) rises dramatically to 3.4 nF compared to that with PVK (device B), which stay almost constant around 1.75 nF until the bias of 3 V, featuring effective charge injection from ACGS QD solids to perovskite. Similar results are also determined in C-V measurement for PeLED devices, certifying lower V_bi for device A (1.8 V) compared to device B (3.0 V). (Supplementary Fig. S21b-e) Consequently the hole current collected from HOD based on ACGS QD (device A) is ten-fold higher than that based on PVK (device B) (Supplementary Fig. S21f) It is thus ascertain that ACGS QD solids HTL provide effective passivation of uncoordinated Pb²⁺ and stabilize halide interstitial defect in quasi-2D perovskite. As a result, the interfacial nucleation kinetics is regulated through microscopic defect control which enhance interfacial charge transport and prevent interfacial recombination.

In summary, we have presented chalcopyrite ACGS QD as potential inorganic hole transport material in highly-efficient, stable blue PeLEDs. Engineering the QDs by size, composition and surface chemistry have realized high-mobility QD solids and suppressed hole injection barrier, allowing for enhanced hole transport and exciton formation in quasi-2D perovskite. The dual-functionalized surface ligand groups of ACGS QD also behave as “interfacial passivator” anchoring to the perovskite through uncoordinated Pb²⁺ and halide interstitial defect, leading to fewer trap states and improved crystallizations of perovskite. With balanced charge transport and inorganic nature of ACGS QDs, the device lifetime is considerably prolonged to hours scale. This work provides creative strategy to replace traditional vacuum-deposited inorganic semiconductor thin films with facile, solution-processable QD solids, with identical electronic properties, thereby allowing us to harness new combinations of optical, electrical and physical properties of QD solids, delivering numerous advantages that cannot be easily-achieved through single inorganic/organic material system.

Most importantly, we bring about interfacial modifications via surface chemistry engineering of QDs interacting with the perovskite, offering a new strategy to overcome phase heterogeneity problems and trap-assisted non-radiative recombination in quasi-2D perovskite. It is feasible that, inorganic chalcopyrite QD can be engineered purposely towards composition, size, shape, and surface chemistry as a promising “self-assembled monolayer (SAM)” at buried hetero-interface with perovskite, supporting enhanced interfacial carrier extraction and transport.

Materials

Oleylamine (OLA, 80%-90%), 1-dodecanethiol (DDT,98%), octane (> 99% GC), cyclo-hexane(> 99% GC), gallium acetylacetonate (99.99% metal basis), gallium chloride (Ga(acac)₃, 99.999% metal basis), cuprous chloride (metal basis 99.95%), cuprous iodide (metal basis 99.95%), sublimed sulfur (AR, ≥ 99.5%(T)), ethanedithiol (97%) were purchased from aladdin®. Zinc chloride (ZnCl₂) was purchased from Alfa Aesar (Ultra dry, 99.999% metal basis). Silver iodide (premion ® metal basis 99.999%), was purchased from thermo scientific. CsBr (99.999%, metal basis), PbBr₂ (99.999%, metal basis) and PbCl₂ (99.999%, metal basis) were purchased from Alfa Aesar. DMSO (99.9%) was purchased from Sigma-Aldrich. PEABr were purchased from Xi’an Polymer Light Technology Corp. TPBi and LiF were purchased from Luminescence Technology Corp. NiO nanoparticles were purchased from Liao’ning You Xuan Renewable Energy Technology Corp.

Synthesis of chalcopyrite QD

In typical synthesis of ACGS QD for PeLEDs applications, 0.025 mmol of AgI, 0.225 mmol of CuI, 1mmol Ga(acac)₃, 10 ml OLA, 10 ml ODE, 2 ml DDT were placed in a 50 ml three-neck flask by stirring, and this mixture was heated to 100°C after degassing for 30 minutes. Then the temperature was further increased to 180°C under Argon (Ar) flow. At 140°C, a sulfur stock, which was obtained by ultrasonically-dissolving 2mmol sulfur in OLA, was rapidly introduced into the mixture to initiate the reaction. During the injection, careful manipulation of the pressure under N₂ flow was required to stabilize the temperature. The growth of QDs maintain at 180°C and last for 5 min. The ACGS cores were isolated by centrifugation using ethanol and were dispersed in cyclohexane for ZnCl₂ ligand exchange.

Hybrid Ligand exchange

The molecular weight of ACGS QDs were estimated as:

$$\:\text{M}={N}_{A}\times\:{\rho\:}{V}_{QD}$$

Where N_A denote Avogadro constant, which is 6.02×10²³. ρ is the density of chalcopyrite CGS, which is 4.35 g/cm³. Assuming the QD is spherical, and the diameter is determined from TEM measurement, which is 4.41 nm. Then volume the QD is calculated as:

$$\:{V}_{QD}=\frac{4}{3}\pi\:{r}^{3}$$

The calculated molecular weight of ACGS QDs is 40 kDa. The as-synthesized ACGS QDs were separated by centrifugation using ethanol and re-dispersed in 2 ml cyclohexane. The concentration of as-synthesized QDs is determined by evacuating the cyclohexane solvent in a 100 ul sample, which is normally in the range between 100 mg/ml to 120 mg/ml for single reaction using precursor amount as described above. 0.1 mmol (20ul of 5M ZnCl₂ in ethanol solution) was added into 400 ul ACGS QDs, which roughly contains 40mg QDs (~ 1 µmol), and the mixture was heated to 60°C and maintained for 30 minutes. In this scenario the ligand density is represented by ligand:QD molar ratio of (0.1 mmol: 1µmol = 100:1). In fact, due to large amount of organic ligand attached onto the surface of QDs, the actual amount of inorganic core is lower than measured, leading to higher ligand to QD ratio than that mentioned in the main text. However, it is still meaningful to estimate the ligand density value in order to optimize the ligand exchange process. The chlorine treated QDs were separated by centrifuging and re-dispersed in octane. All processes were carried out in a fume hood under room atmosphere.

Solid-phase ligand exchange by EDT was conducted through a layer-by-layer approach in the N₂-filled glove box. 1% volume ratio of EDT dissolved in acetonitrile was overspread onto the ACGS QD layer and kept there for 60s; then the film was spun at 1500 rpm for 40s. To remove excess EDT and exchanged DDT ligand, two consecutive rinsing steps were conducted with neat acetonitrile by drop-casting, and the film was baked at 60°C for 30 minutes.

Perovskite precursor preparations

The stoichiometric ratio of CsBr, PbBr₂, PbCl₂ and PEABr (molar ratio of CsBr:PbBr₂:PbCl₂:PEABr = 0.2:0.1:0.1:0.16) were dissolved in dimethyl sulfoxide (DMSO) with a concentration of 0.2 M and stirred overnight in a nitrogen glove box at 60^oC to prepare the mixed precursor. The precursor solution is filtered through a 0.22µm PTFE membrane prior to use.

Device fabrications

For PeLED device fabrications, the patterned ITO substrates were ultrasonic treated with detergent, deionized water, acetone and isopropyl alcohol successively for 15 min. The surface of ITO substrate was treated by oxygen plasma to improve the hydrophilicity. The PEDOT:PSS solution was spun onto the ITO substrate at 4000 rpm and then annealed at 150°C for 20 min. After baking, the substrate was transferred to a nitrogen-filled glove box to deposit HTL. For ACGS QDs-based devices, ACGS-ZnCl₂ (ligand:QD = 100:1) in octane (concentration:4 mg/ml) was first spin-coated at 3000rpm for 30s (thickness ~ 5 nm). Then a film-washing strategy was performed using EDT in acetonitrile (1 vol%). Then the ACGS-ZnCl₂-EDT QD thin-film was annealed at 60°C for 30 min. For PVK-based devices, PVK in chlorobenzene (concentration: 8mg/ml) was spin-coated onto the PEDOT:PSS at 4000 rpm for 40 s, followed by baking at 120°C for 20 min (thickenss ~ 25nm). For NiO_x-based device, NiO nanoparticles were dissolved in deionized water with a concentrtoin of 10 mg/ml. Then the solution was spun onto the ITO substrate at 3000 rpm for 30s, followed by annealing at 150^oC for 15 min. Perovskite films were rotatively coated in a precursor solution at 3000 rpm, and after baking at 50°C for 5 min, the sample were transferred to a vacuum chamber (< 5 × 10^− 4 Pa) and successively deposited with TPBi (40 nm), LiF (1 nm) and Al electrodes (100 nm). During cathode deposition, the active area of 4 mm² is defined by shadow shading. Except for PEDOT:PSS and NiO_x film deposition, other processes were conducted in N₂-filled glove box.

The EL spectra characteristics of the PeLEDs were measured using a Shenzhen Pynect integrated sphere (50 mm in diameter) coupled with a Keithley 2400 sourcemeter. The operation lifetime tests of PeLEDs were conducted on a ZJZCL-1 OLED aging lifespan test instrument. The I-V characteristics of the PeLEDs were collected on a system comprised of a PR-670 Spectra Scan Spectroradiometer coupled with a Keithley 2400 sourcemeter.

The EOD was prepared with the structure ITO (150 nm)/ZnMgO(20 nm)/perovskite/TPBi (40 nm)/LiF (1nm)/Al (100 nm). The HOD was prepared with the structure lTO (150 nm)/PEDOT:PSS (30 nm)/HTL/perovskite/MoO₃(10 nm)/LIF (1 nm)/Ag (100 nm). The deposition process of single carrier devices were exactly the same as that described for LED devices.

Characterizations

The absorption characteristics of QDs were measured with a ultraviolet-visible spectrometer (Model: UV2600, Shimadzu). The QD solution was diluted about 100 times using octane, to reach an optical density of 2 at 300 nm. Crystal structures were analyzed by X-ray diffraction (XRD) (Model: Smartlab SE, Rigaku). Steady-state photoluminescence (PL) and TRPL spectra were recorded on an Edinburgh FLS920 PL spectrometer at an excitation wavelength of 365 nm. TEM images were obtained on JEM-3200FS from JEOL, and the energy-dispersive X-ray (EDX) spectrum were obtained using X-max N 100 LTE spectrometer.

The film morphology were obtained by scanning electron microscopy (SEM) measured on (Model: Sigma 300, ZEISS) with a spectrometer (Xplore 30), inlens and secondary electron detector. The accelerating voltage is 0.02-30 kV with a step of 10 V. Atomic force microscopy (AFM) (cypher ES) were employed for microscopic surface morphology; and KPFM module were employed for suface potential mapping measurement. The sample chamber is specifically protected by N₂ to prevent degradation of perovskites during measurement. The work function (WF) is calculated using the equation: φ_sample-φ_tip = eV_CPD, where φ_sample, φ_tip and V_CPD denotes the WF of sample, the tip, and average contact potential difference. φ_tip (AC 24O TM) is 4.9 eV used in this work.

The Ultra-violet photoelectron spectroscopy (UPS) were measured on ESCALAB XI + from Thermo Fisher Scientific. He (I) ultraviolet radiation source (21.22 eV) from a He discharge lamp was used in the UPS measurements. X-ray photoelectron spectroscopy (XPS) were measured on Nexsa from Thermo Fisher Scientific. Al Kα lines were used as excitation source, and the operation voltage is 12 kV. Survey scan was performed with a range of 100 eV and step of 1 eV, and fine scan was performed with a step of 0.05 eV. The binding energy was calibrated with C 1s emission at 284.8 eV.

FTIR was performed by a Bruker spectrometer with an ATR hemisphere. The transmission mode was used and the background scan is vaccum atmosphere without samples. TGA was performed from 20°C to 600°C at a heating rate of 10°C min under N2 using a Trios V3.2 system (Model: TGA55, TA Instruments:). The hydrodynamic particle sizes of the QDs in solutions were determined using a particle-size analyser (Model: Nano ZS90, Malvern Zetasizer). The actual content of elements in QD were detected using ICP-OES (Model: 720ES (OES), Agilent). The QD solution was dried by nitrogen, and obtained solid was dissolved in 3% nitric acid and deionized water for ICP-OES, respectively. The sample solution was diluted 10,000 times for ICP-OES measurements. Contact angle measurement were obtained on Lauda Scientific instrument using deionized water (Model: LSA100).

DFT simulation details

The passivated surface structures were simulated by constructing a slab model with coupling the (112) CuGaS₂ surface and adsorbed molecules. The slab was separated by a 15 Å vacuum space in the z-direction to avoid interaction between layers.

The structural optimization, total energy, and electronic states of these surface configurations were calculated by employing the Vienna ab initio simulation package (VASP) within the framework of density functional theory. In the calculations, the projector augmented wave method and the exchange-correlation functional of generalized gradient approximation (GGA) were adopted. The energy cut-off was set as 350 eV for the plane-wave basis-set expansion. The k-point of the Brillouin zone was 0.03 Å⁻¹ interval distribution of Monkhorst-Pack for the optimization of structures, and the k-point interval of the total energy self-consistent calculation was 0.02 Å⁻¹ or better. All the structures were fully optimized and convergence thresholds were set as 10^− 5 eV in energy and 10^− 3 eV/Å in force.

Based on the calculated total energies, the ligand binding energy was estimated as E_b = E_{(CuGaS2−ligand)} – E_CuGaS2 – µ_ligand, where E_{(CuGaS2−ligand)} and E_CuGaS2 are the total energy of the passivated and bare slab, respectively, and µ_ligand is the chemical potential of the ligand.

Thin film transistor measurement

The QD-TFT was fabricated as the following steps: 1. Low-resistance (resistivity: 0.001–0.005 Ω•cm) p-type Si wafer with an oxide layer of 200nm was cleaned by SC-1 standard process. 2. NR-1000P (Futurrex) was spin-coated at 4000rpm for 40s, followed by a softbake at 150°C for 60s. 3. The sample was exposed under UV light (365nm) for 30s, and then developed in RD6 for 15s. 4. Cr/Au (10nm/30nm) drain/source electrode was deposited through electron-beam evaporation, followed by lift-off process. 5. QDs solution dispersed (20 mg/ml) in hexane was spin-coated onto the pre-patterned silicon substrate at 1500 rpm for 40s, followed by baked at 60°C for 30min. 6. The gate electrode was probed directly on bottom silicon surface through mechanically-scratching thin film stacks on top.

The device yield measurement of QD transistors were performed using a manual probe station under ambient environment. To avoid short circuit, an insulated glass was placed between the wafer and the sample stage. For the measurement of transfer curves, the bottom gate was connected to on SMU of the Keithley 4200. One drain was connected to another SMU, the source was connected to the ground, and the bulk was left as none. It is important to mention the measurement was conducted under dark relaxed state since the photo-response of CGS QDs is non-ignorable.

The hole mobility (µ_h) in the linear region of working TFTs can be figured out by using the equation:

$\:{\:\:\mu\:}_{h}=\frac{\text{d}{I}_{DS}}{\text{d}{V}_{GS}}\frac{l}{w{C}_{ox}{V}_{DS}}$ Eq. (1)

where l and w are the channel length and width of QD-TFT, which is 50um and 50um, respectively. 200nm-thick SiO₂ was deposited as dielectric layer. C_ox is calculated from the equation: C_ox = ε_ox/t, where the dielectric constant (ε_ox) is estimated to be 3.9 for SiO₂. t refer to the total thickness of insulating oxide. $\:\frac{\text{d}{I}_{DS}}{\text{d}{V}_{GS}}$ could be obtained from the slope of linear approximation of I_DS-V_GS plot at negative bias.

The hole concentration (p_h) is estimated through the equation;

$\:{p}_{h}=\frac{{\sigma\:}_{h}}{q{\mu\:}_{h}}$ Eq. (2)

where σ_h, q and µ_h denote the conductivity, electron charge and hole mobility, respectively. σ_h could be obtained from J-V curves of two-terminal devices based on various HTL. Substituting estimated in Eq. (1), p_h could be calculated.

Thermal admittance and Mott-Schottky analysis

The capacitance spectra of perovskite solar cells was measured using an impedance analyzer from Tonghui electronics (model: TH2838S) together with a home-built cryosystem temperature controller from LakeShore. Each capacitance spectrum was acquired at zero bias within the frequency range between 20 Hz to 1M hz at a given constant temperature. To avoid probe-induced damage to the device, a thin indium (In)-pad was placed underneath the probe, and then placed onto the Al electrode. The connected equivalent circuit model was used for capacitance calculations.

The maximum frequency (ω₀) that a defect at an energy level E_ω within the bandgap can contribute to the junction capacitance is written as:

$\:{\omega\:}_{0}={2N}_{v}{v}_{th}\sigma\:exp\left(-\frac{{E}_{\omega\:}}{kT}\right)={\gamma\:}_{0}{T}^{2}exp\left(-\frac{{E}_{\omega\:}}{kT}\right)$ Eq. (3)

N _v and v_th represent density of state and thermal velocity of valence band holes, respectively. Their product can be generally written as $\:{\gamma\:}_{0}{T}^{2}$. The temperature-independent property $\:{\gamma\:}_{0}$ is denoted as “attempt-to-escape” frequency. ω₀ can be derived by finding the maxima of “scaled derivative” (-ω/kT dC/dω) as a function of frequency axis. The activation energy level (E_A) of defects is estimated by drawing an arrhenius plot between inflection frequency divided by T² (ω₀/T²) versus 1/T .²⁶ E_A is conceived as conjunct effects of migration barrier and defect formation energy ³³.

The electronic density of states (DOS) spectroscopy of defect can be determined from “the scaled derivative”:

$\:{\:N}_{t}\left({E}_{\omega\:}\right)=-\frac{{V}_{bi}}{qw}\frac{\omega\:}{kT}\frac{dC}{d\omega\:}$ Eq. (4)

where V_bi and w represent the built-in electric field and depletion width of the device. V_bi is determined by Mott-Schottky plot of capacitance-voltage profiling measurement, which is 1.8 V, and 3.0 V for device with ACGS QD and PVK as HTL, respectively. The depletion region width at zero bias is determined from capacitance-voltage (C-V) characteristics for LED device with the equation:

$\:C=\frac{{\epsilon\:}_{o}{\epsilon\:}_{r}}{w}$ Eq. (5)

Where ε_o and ε_r denotes the dielectric constant for vacuum and quasi-2D perovskite, respectively. ε_r is estimated using distinct geometry capacitance (C_geo) in TAS measurement under high-frequency excitation. Based on serial connection of HTL, perovskite and ETL, and high capacitance of HTL and ETL due to thin layer, the geometry capacitance of the device is fully determined by the perovskite layer. Therefore, the C_geo is 1.53 nF and 1.60 nF for PeLEDs with PVK and ACGS QDs, respectively. Taken into account of the thickness of perovskite layer, which is roughly estimated to be within the range of 40 nm-50 nm, the ε_r of quasi-2D perovskite is calculated to be 1.63 ~ 1.73, consistent of reported values ⁷⁰. Furthermore, the capacitance at zero bias is measured to be 1.80 nF and 2.11 nF for PeLEDs with PVK and ACGS QDs as HTL, respectively. Thus, the depletion region width for PeLEDs with PVK and ACGS QDs are calculated to be 30.09 nm and 25.69 nm, respectively. Substituting the V_bi, w, and scaled derivative -ω/kT dC/dω, the DOS could be achieved using Eq. (4).

Data availability

The data that support the plots within this paper and the other findings of this study are

available from the corresponding authors upon reasonable request.

Acknowledgements

This work was financially supported by the Guangdong Science and Technology Program (Grant No. 2022A1515011178, 2024A1515011757), the national natural science foundation of China (Grant No. 52173243), the Shenzhen Science and Technology Program (Grant No. JCYJ20220818101402005, JCYJ20210324101801004, RCYX2023121109020901, GYHZ20240218112501002). This work is partially supported by the Guangdong Youth Talent Program (Grant No. 2023TQ07A142), and the Youth Innovation Promotion Association, Chines Academy of Sciences (Grant No. 2023375). The authors acknowledge the financial support from the National Natural Science Foundation of China (No. 62105195, 62174104 and 61735004), the Program of Shanghai Academic/Technology Research Leader (22XD1421200), the Shanghai Natural Science Foundation (23ZR1423300). This work was also supported by the Key Laboratory of Biomedical Imaging Science and System, Chinese Academy of Sciences. C Yang gratefully acknowledge Shanghai University for the support and guidance in device architecture optimizations. We sincerely thank Q. Wu an M. Liu for their assistance in device fabrications. We also acknowledge H Wang and Y Bai for optical measurement in the Institute of Technology for Carbon Neutrality and important coordination during the preparation of manuscript.

Contributions

C Yang and X Zheng conceived the idea and designed the experiment. C Yang, W Hou and Y Bai provided financial support and supervised the work and data acquired. Y Wang contributed in device optimizations, fabrications and characterizations. W Ming was in charge of all morphology and electrical analysis of atomic force microscopy. G Zhong aided in DFT simulations and analysis. Y Shao conducted the photolithography techniques for thin-film transistor, and C Ren assists the corresponding thin-film transistor measurement. X Zheng and Y Wang wrote the original draft of the manuscript. X Yang, M Chen, and H Wang helped to organize the data, figures and deliver meaningful suggestions. C Yang, W Hou provided major revisions. All the authors have discussed the results and comment on the manuscript.

Corresponding Authors

Correspondence to Xue Zheng, Wenjun Hou or Chunlei Yang.

Email: [email protected]

E-mail: [email protected]

Email: [email protected]

Ethics Declarations

Competing interests

The authors declare non conflictions of interests.

Tan ZK et al (2014) Bright light-emitting diodes based on organometal halide perovskite. Nat Nanotechnol 9:687–692
Strank S et al (2015) Metal-halide perovskites for photovoltaic and light-emitting devices. Nat Nanotechnol 10:391–402
Yuan M et al (2016) Perovskite energy funnels for efficient light-emitting diodes. Nat Nanotechnol 11:872–877
Kovalenko MK et al (2017) Properties and potential optoelectronic applications of lead halide perovskite nanocrystals. Science 358:745–750
Service RF (2019) Perovskite LEDs begin to shine. Science 364:918–918
Liu XK et al (2021) Metal halide perovskites for light-emitting diodes. Nat Mater 20:10–21
Han TH et al (2022) A roadmap for the commercialization of perovskite light emitters. Nat Rev Mater 7:757–777
Li J et al (2023) Efficient all-thermally evaporated perovskite light-emitting diodes for active-matrix displays. Nat Photonics 17:435–441
Gao Y et al (2024) Microsecond-response perovskite light-emitting diodes for active-matrix displays. Nat Electron 7:487–496
Editorial (2024) Perovskites light a path forward. Nat Nanotechnol 19:579
Kong L et al (2024) Fabrication of red-emitting perovskite LEDs by stabilizing their octahedral structure. Nature 631:73–79
Sun SQ et al (2024) Enhancing light outcoupling efficiency via anisotropic low refractive index electron transporting materials for efficient perovskite light-emitting diodes. Adv Mater 36:2400421
Gao Y et al (2024) Highly efficient blue light-emitting diodes based on mixed-halide perovskites with reduced chlorine defects. Sci Adv 10:eado5645
Li M et al (2024) Acceleration of radiative recombination for efficient perovskite LEDs. Nature 630:631–635
Ko PK et al (2024) The deepest blue: major advances and challenges in deep blue emitting quasi-2D and nanocrystalline perovskite LEDs. Adv Mater 2407764
Chu Z et al (2024) Blue Light-Emitting Diodes Based on Pure Bromide Perovskites. Adv Mater 2409867
Yang X et al (2023) Focus on perovskite emitters in blue light-emitting diodes. Light: Sci Appl 12:177
Dong Y et al (2020) Bipolar-shell resurfacing for blue LEDs based on strongly confined perovskite quantum dots. Nat Nanotechnol 15:668–674
Karlsson M et al (2021) Mixed halide perovskites for spectrally stable and high-efficiency blue light-emitting diodes. Nat Commun 12:361
Jiang Y et al (2022) Synthesis-on-substrate of quantum dot solids. Nature 612:679–684
Yuan S et al (2024) Efficient blue electroluminescence from reduced-dimensional perovskites. Nat Photonics 18:425–431
Huang Q et al (2024) Enhancing crystal integrity and structural rigidity of CsPbBr₃ nanoplatelets to achieve a narrow color-saturated blue emission. Light: Sci Appl 13:111
Milot R et al (2016) Charge carrier dynamics in 2D hybrid metal-halide perovkites. Nano Lett 16:7001–7007
Lin Y et al (2017) Suppressed ion migration in low-dimensional perovskites. ACS Energy Lett 2:1571–1572
Kong L et al (2021) Smoothing the energy transfer pathway in quasi-2D perovkite films using methaneulfonate leads to highly efficeint light-emitting devices. Nat Commun 12:1246
Li X et al (2021) The 2D Halide Perovskite Rulebook: How the Spacer Influences Everything from the Structure to Optoelectronic Device Efficiency. Chem Rev 4:2230–2291
Quan L et al (2017) Tailoring the Energy Landscape in Quasi-2D Halide Perovskites Enables Efficient Green-Light Emission. Nano Lett 6:3701–3709
Ma D et al (2021) Distribution control enables efficient reduced-dimensional perovskite LEDs. Nature 599:594–598
Wang YK et al (2024) Long-range order enabled stability in quantum dot light-emitting diodes. Nature 629:586–591
Fakharuddin A et al (2022) Perovskite light-emitting diodes. Nat Electron 5:203–216
Shao YC et al (2016) Grain boundary dominated ion migration in polycrystalline organic–inorganic halide perovskite films. Energy Environ Sci 9:1752–1759
Calado P et al (2016) Evidence for ion migration in hybrid perovskite solar cells with minimal hysteresis. Nat Commun 7:13831
Zheng X et al (2021) Correlations of Ionic Migration and Deep-Level Traps Leads to Surface Defect Formation in Perovskite Solar Cells. J Phys Chem C 125:19551–19559
Liu T et al (2019) Tailoring vertical phase distribution of quasi-two-dimensional perovskite films via surface modification of hole-transporting layer. Nat Commun 10:878
Chu Z et al (2023) Blue light-emitting diodes based on quasi-two-dimensional perovskite with efficient charge injection and optimized phase distribution via an alkali metal salt. Nat Electron 6:360–369
Jiang J et al (2024) Efficient pure-red perovskite light-emitting diodes with strong passivation via ultrasmall-sized molecules. Sci Adv 10:eadn5683
Liu Y et al (2019) Efficient blue light-emitting diodes based on quantum-confined bromide perovskite nanostructures. Nat Photonics 13:760–764
Wang HR et al (2020) Boosting the efficiency of NiOx-based perovskite light-emitting diodes by interface engineering. ACS Appl Mater Interfaces 12:53528–53536
Li Z et al (2023) Charge injection engineering at organic/inorganic heterointerfaces for high-efficiency and fast-response perovskite light-emitting diodes. Nat Commun 14:6441
Bai W et al (2023) Perovskite Light-Emitting Diodes with an External Quantum Efficiency Exceeding 30%. Adv Mater 35:2302283
Lee SJ et al (2018) Control of Interface Defects for Efficient and Stable Quasi-2D Perovskite Light-Emitting Diodes Using Nickel Oxide Hole Injection Layer. Adv Sci 5:1801350
Ji H et al (2024) Tailoring Niox/Perovskite Interface via a Multifunctional Self-Assembled Molecule for High-Performance Blue Perovskite Light-Emitting Diodes. Small Struct 5:2400153
Gangishetty MK et al (2018) Reducing Architecture Limitations for Efficient Blue Perovskite Light-Emitting Diodes. Adv Mater 30:1706226
Rau U et al (1999) Electronic properties of Cu(In,Ga)Se₂ heterojunction solar cells–recent achievements, current understanding, and future challenges. Appl Phys A 69:131–147
Dullweber T et al (2000) Study of the effect of gallium grading in Cu(In,Ga)Se₂. Thin Solid Films 361–362:478–481
Sheu HH et al (2016) The effect of Cu concentration in the photovoltaic efficiency of CIGS solar cells prepared by co-evaporation technique. Vacuum 131:278–284
Yang L et al (2023) I–III–VI quantum dots and derivatives: design, synthesis, and properties for light-emitting diodes. Nano Lett 23:2443–2453
Keller J et al (2024) High-concentration silver alloying and steep back-contact gallium grading enabling copper indium gallium selenide solar cell with 23.6% efficiency. Nat Energy 9:467–478
Xie R et al (2009) Formation of high-quality I – III – VI semiconductor nanocrystals by tuning relative reactivity of cationic precursors. J Am Chem Soc 131:5691–5697
Boles MA et al (2016) The surface science of nanocrystals. Nat Mater 15:141–153
Kim J-H et al (2019) Tunable emission of bluish Zn–Cu–Ga–S quantum dots by Mn doping and their electroluminescence. ACS Appl Mater Inter 11:8250–8257
Y. Yu. et al. Regulating perovskite crystallization through interfacial engineering using a zwitterionic additive potassium sulfamate for efficient pure-blue light-emitting iodes. Angew Chem Int Ed 63, e202319730 (2024)
Wang HR et al (2019) Trifluoroacetate induced small-grained CsPbBr₃ perovskite films result in efficient and stable light-emitting devices. Nat Commun 10:665
Kim JS et al (2022) Ultra-bright, efficient and stable perovskite light-emitting diodes. Nature 611:688–694
Owen J (2015) The coordination chemistry of nanocrystal surfaces.Science 347,615–616
Zheng X et al (2020) Chlorine Vacancy passivation in mixed halide perovskite quantum dots by organic pseudohalides enables efficient Rec. 2020 blue light-emitting iodes. ACS Energy Lett 5:793–798
Abdi-Jalebi M et al (2018) Maximizing and stabilizing luminescence from halide perovskites with potassium passivation. Nature 555:497–501
Chuang CH et al (2014) Improved performance and stability in quantum dot solar cells through band alignment engineering. Nat Mater 13:796–801
Uematsu et al (2023). Facile high-yield synthesis of Ag–In–Ga–S quaternary quantum dots and coating with gallium sulfide shells for narrow band-edge emission. Chem Mater 35:1094–1106
Cho HC et al (2015) Overcoming the electroluminescence efficiency limitations of perovskite light-emitting diodes. Science 350:1222–1224
Ye Y-C et al (2024) Interfacial energy level alignment and defect passivation by using a multifunctional molecular for efficient and stable perovskite solar cells. Adv Funct Mater 34:2310136
Ye Y-C et al (2023) Interfacial molecule control enables efficient perovskite light-emitting diodes. Adv Funct Mater 33:2307818
Shen Y et al (2021) Interfacial nucleation seeding for electroluminescent manipulation in blue perovskite light-emitting diodes. Adv Func Mater 31:2103870
Azpiroz, Jon M et al (2015) Defect migration in methylammonium lead iodide and its role in perovskite solar cell operation. Energy Environ Sci 7:8
Kim T et al (2023) Mapping the pathways of photo-induced ion migration in organic-inorganic hybrid halide perovskites. Nat Commun 14:1846
Cao J et al (2018) Interstitial occupancy by extrinsic alkali cations in perovskites and its impact on ion migration. Adv Mater 30:1707350
Liu A et al (2021) Electroluminescence principle and performance improvement of metal halide perovskite light-emitting diodes. Adv Opt Mater 9:2002167
Li Y et al (2024) Modulation of charge transport layer for perovskite light-emitting diodes. Adv Mater 2410535
Yi Z et al (2024) Self-assembled monolayers (SAMs) in inverted perovskite solar cells and their tandem photovoltaics application. Interdiscip Mater 3:203–244
Filip MR et al (2022) Screening of Excitons by organic cations in quasi-two-dimensional organic–inorganic lead-halide perovskites. Nano Lett 22:4870–4878

There is NO Competing Interest.

Supplementaryinformation.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Efficient and stable blue perovskite light-emitting diodes through I-III-VI quantum dot solids as hole transport layer

Status:

Version 1

Abstract

Figures

Introduction

Results

Synthesis and characterizations of I-III-VI semiconductor quantum dot

Functional surface ligand engineering and improved hole transport in QD solids

Buried interface passivation and regulated crystallizations of quasi-2D perovskite

Defect physics and highly-efficient, stable blue Pe-LED

Discussion

Methods

Materials

Synthesis of chalcopyrite QD

Hybrid Ligand exchange

Perovskite precursor preparations

Device fabrications

Characterizations

DFT simulation details

Thin film transistor measurement

Thermal admittance and Mott-Schottky analysis

Declarations

Data availability

References

Additional Declarations

Supplementary Files

Status:

Version 1