Phase Engineering Boosts Efficiency of Quasi‑2D All‑Inorganic Perovskite LEDs via Cs Cation Ratio Control

Introduction

Perovskite materials have rapidly attracted attention for thin‑film light‑emitting diodes due to their tunable emission, high ambipolar charge mobility, solution processability, and low cost. Nonetheless, the inherently low exciton binding energy and limited film‑forming ability of three‑dimensional (3D) perovskites often result in sub‑optimal emission characteristics. Over the past decade, a variety of approaches—composition modulation, interface engineering, nanocrystal pinning, solvent optimization, and polymer doping—have been employed to boost the luminous efficiency of perovskite LEDs. Recent devices have achieved external quantum efficiencies (EQE) approaching 20 %, rivaling state‑of‑the‑art OLEDs and underscoring the potential of perovskites for lighting and display technologies. Quasi‑2D perovskites, generally described as L₂(CsPbX₃)_n‑1PbX₄, have emerged as promising emissive materials. Their layered structure, formed by large organic ammonium cations that cannot enter the PbX₆⁴⁻ octahedral framework, creates multiple quantum‑well architectures with enhanced exciton confinement and, consequently, higher photoluminescence quantum yields (PLQY) and improved environmental stability compared to their 3D counterparts. Historically, phenylethylammonium (PEA) and butylammonium (BA) cations have been incorporated into perovskite precursors to produce green PeLEDs with EQEs ranging from 8.8 % to 14.4 % and brightnesses up to 8779 cd m⁻². Similar successes have been reported for sky‑blue devices using mixed PEA and isopropylammonium (IPA) dopants. Despite these advances, quasi‑2D perovskites frequently exhibit a mixed‑phase composition, particularly in stoichiometric n = 3 formulations. Low‑dimensional phases (small n values) are detrimental because they promote non‑radiative exciton‑phonon interactions at room temperature, thus lowering luminous efficiency. Achieving high phase purity remains a central challenge for the field. In this study, we introduce excess cesium cations into the n = 3 perovskite precursor to suppress low‑dimensional phase formation. The resulting quasi‑2D film, composed of phenylpropylammonium bromide (PPABr) and CsPbBr₃, exhibits superior film morphology, reduced defect density, and markedly enhanced photoluminescence. Devices fabricated with this optimized emissive layer achieve a peak brightness of 2921 cd m⁻² and a current efficiency of 1.38 cd A⁻¹, representing a near‑tripling of performance relative to devices based on the stoichiometric n = 3 film.

Methods

All reagents were used as received. Lead bromide (PbBr₂; Alfa Aesar, 99.999 %), dimethyl sulfoxide (DMSO; 99.5 % anhydrous, J&K Chemicals), PEDOT:PSS (Heraeus, VP AI4083), TPBi (≥ 99.9 %), cesium bromide (CsBr; 99.9 %), and phenylpropylammonium bromide (PPABr; ≥ 99.5 %) were sourced from Xi’an Polymer Light Technology Corp. Perovskite precursor solutions were prepared by mixing PPABr, CsBr, and PbBr₂ in DMSO at 60 °C overnight. Four molar ratios were examined: 2:2:3, 2:3:3, 2:3.5:3, and 2:4:3 (PPABr:CsBr:PbBr₂), while maintaining a constant PbBr₂ concentration of 0.15 M. ITO‑coated glass substrates were sequentially cleaned in detergent, deionized water, acetone, and isopropanol for 20 min each, dried at 80 °C for 40 min, and treated in a UV‑Ozone chamber for 20 min. PEDOT:PSS was filtered (0.45 µm PTFE) and spin‑coated at 2900 rpm for 60 s, then baked at 150 °C for 20 min. Substrates were transferred to a nitrogen glovebox for device assembly. Perovskite films were spin‑coated at 3000 rpm for 90 s and annealed at 90 °C for 15 min, yielding ~70 nm thickness. Subsequently, TPBi (40 nm), LiF (1 nm), and Al (100 nm) were thermally evaporated under a base pressure of 4 × 10⁻⁴ Pa. The active area of each PeLED was 0.11 cm². Electrical and optical characterizations were performed in the glovebox without encapsulation. Current density–luminance–voltage (J‑L‑V) curves were recorded using a Keithley 2400 source‑meter and a calibrated silicon photodiode. Electroluminescence (EL) spectra were collected with a PR670 spectrometer. Morphology was examined by FESEM (ZEISS GeminiSEM 300) and AFM (Agilent AFM 5500). Structural analysis employed X‑ray diffraction (XRD; X’Pert PRO, PANalytical). UV‑Visible absorption was measured on an Agilent Cary 5000. Steady‑state photoluminescence (PL) spectra were obtained with a HITACHI F7000, and time‑resolved PL (TRPL) decay curves were recorded using an Edinburgh FLS980 spectrophotometer.

Results and Discussion

Perovskite Film Characterization

Figure 1a displays the absorption spectra of the 3D CsPbBr₃ and 2D PPA₂PbBr₄ reference films. The CsPbBr₃ film shows a peak near 517 nm, whereas the PPA₂PbBr₄ film exhibits a distinct absorption at 400 nm, confirming the presence of n = 1 and n = ∞ phases and illustrating strong quantum confinement in the 2D material. For the mixed‑phase films, multiple absorption peaks indicate coexistence of several n values. In the n = 3 composition with a 2:2 ratio, the absorption associated with low‑n phases dominates, signaling a high proportion of low‑dimensional species. As the CsBr content increases to 2:3 and 2:3.5, absorption features of intermediate‑n phases emerge, indicating that low‑n components are progressively converted into larger‑n phases. XRD patterns (Fig. 1c) reveal two dominant peaks at 15.15° and 30.45°, attributable to the (100) and (200) planes of orthorhombic CsPbBr₃, respectively. The preferential orientation observed across all samples aligns with previous reports. Morphological analysis via SEM and AFM (Figs. 2 and 3) shows that pristine 3D CsPbBr₃ films have a rough surface with numerous voids (RMS = 9.49 nm). Introducing PPABr dramatically improves coverage and reduces grain size, lowering RMS to 2.16 nm at a 2:2 ratio. Further CsBr addition to 2:3 and 2:3.5 ratios maintains a smooth morphology, while a 2:4 ratio reintroduces roughness. These observations confirm that moderate incorporation of CsBr does not compromise film quality. Photoluminescence spectra (Fig. 4a) reveal a blue‑shift from 524 nm (3D CsPbBr₃) to 517 nm (2:2) due to enhanced quantum confinement. Increasing CsBr content induces a slight red‑shift, with the 2:3.5 composition delivering the highest PL intensity under identical excitation conditions. TRPL decay curves (Fig. 4b) were fit with a tri‑exponential model, yielding average lifetimes (τ_avg) that increase from 7.02 ns (pristine 3D) to 32.11 ns for the 2:3.5 film. This significant lifetime extension reflects reduced trap density and improved exciton recombination dynamics, consistent with the suppression of low‑n phases.

LED Device Fabrication

Devices were constructed with the structure ITO/PEDOT:PSS/PPA₂(CsPbBr₃)_n‑1PbBr₄/TPBi/LiF/Al. Current–voltage–luminance measurements (Fig. 6c–e) demonstrate that incorporating PPABr markedly reduces leakage currents at low biases, indicating fewer shunting pathways. The 2:2 film achieves a peak brightness of 1026 cd m⁻² and a current efficiency of 0.80 cd A⁻¹, surpassing the 3D baseline by over two orders of magnitude. The 2:3.5 film further improves performance, reaching 2921 cd m⁻² brightness and 1.38 cd A⁻¹ current efficiency—a near‑tripling of the 2:2 device. EL spectra (Fig. 6e) show a modest red‑shift relative to PL, as commonly observed in operating PeLEDs. These enhancements are attributed to the improved film morphology and the substantial reduction of low‑dimensional phase content achieved by optimizing the CsBr ratio.

Conclusions

We have developed a facile phase‑engineering approach that leverages organic spacers and controlled cesium incorporation to produce quasi‑2D perovskite films with superior morphology and phase purity. The optimized CsBr content suppresses low‑n phase formation, enhances photoluminescence, and extends exciton lifetimes. Consequently, the best‑performing PeLED achieves a peak brightness of 2921 cd m⁻² and a current efficiency of 1.38 cd A⁻¹. This strategy offers a practical route to elevate the efficiency of quasi‑2D perovskite LEDs and paves the way for future high‑performance optoelectronic devices.

Availability of Data and Materials

All datasets are presented in the main text or supplementary files.

Abbreviations

2D:

Two‑dimensional

3D:

Three‑dimensional

AFM:

Atomic force microscope

CE‑V:

Current efficiency‑voltage

CsBr:

Cesium bromide

EQE:

External quantum efficiency

FESEM:

Field‑emission scanning electron microscope

ITO:

Indium tin oxide

J‑V:

Current density‑voltage

L‑V:

Luminance‑voltage

PbBr₂:

Lead bromide

PeLEDs:

Perovskite light‑emitting diodes

PLQY:

Photoluminescence quantum efficiency

PPABr:

Phenylpropylammonium bromide

TRPL:

Time‑resolved photoluminescence

XRD:

X‑ray diffraction

τ_avg:

Average lifetime