Interfacial Conditioning for High-Brightness 28 mm2 Active-Area Phenylpropylamine-Capped CsPbBr3 Perovskite Nanocrystal Light-Emitting Diodes Synthesized by a Modified Ligand-Assisted Reprecipitation Method

Abstract Processing-induced interfacial losses limit perovskite nanocrystal light-emitting diodes (PeLEDs) at display-relevant device areas, yet the steps responsible are rarely identified with statistical confidence beyond champion-device reporting. Here, we combine an all-inorganic CsPbBr3 emitter with a replicated factorial study at 28 mm2 active area. Nanocrystals capped with phenylpropylamine (PPA) by a room-temperature modified ligand-assisted reprecipitation (MLARP) route are deposited as two sequential films separated by a ∼2.8 nm polyvinylpyrrolidone (PVP) interlayer, with mild vacuum conditioning before the top layer. Seven of thirteen architectures were replicated over two fabrication sessions, giving 71 functional devices from 84 (84.5% yield). A matched 2 × 2 factorial ANOVA identifies vacuum conditioning as the dominant factor for EQE (partial η2 = 0.90), with a smaller but significant PVP contribution (p = 0.006) and a significant PVP × vacuum interaction for maximum luminance (p < 0.001). The combined route gives a median EQEmax of 12.2% (IQR 0.5%) at 12/12 yield and a champion device with EQEmax = 13.23% at 520 nm, 21,537 cd.m–2, 30.8 lm.W–1 at 100 cd.m–2, 1.7% roll-off at 10 mA.cm–2, and encapsulated τ50 = 3 h at 200 cd.m–2, measured in an integrating sphere. Photoluminescence lifetimes recover from 5.0 to 14.9 ns in the bilayer and the corresponding photoluminescence quantum yield from 8.9% to 28.7%. Across solution, film and both half-device stacks the quantum yield scales linearly with the amplitude-weighted lifetime at a common radiative rate, placing these losses in nonradiative channels and consistent with reduced interfacial quenching. At the same time, transfer-matrix modeling shows outcoupling to be emitter-position dependent (10.4% vs 13.3%), so radiative, transport, and optical contributions remain unseparated. The PVP/vacuum route defines a statistically robust processing window for nanocrystal PeLEDs at a 28 mm2 active area.