Sub Band Gap Operation Limits for Perovskite Light Emitting Diodes

TL;DR

This work develops a calibrated analytical multi-physics model for PeLEDs to explore the possibility of achieving maximum EQE and ECE at sub $E_g$ biases. By relating current, carrier density, and radiative output while accounting for space-charge effects and recombination channels, the authors derive fundamental limits such as a nearly constant $E_g - V_T$ offset and a sub-$E_g$ path to peak efficiency under mitigated space-charge conditions. They further show that maximum EQE can occur at sub $E_g$ biases once transport limitations are addressed, while maximum ECE occurs at even lower biases, providing concrete expressions and experimental validation across LED technologies. The results offer practical optimization routes (reducing $k_1$, boosting transport-layer conductivity, increasing $V_{BI}$) and propose a framework with broad relevance for designing high-radiance, low-power LEDs beyond PeLEDs.

Abstract

Ultra low voltage operation of Perovskite light emitting diodes (PeLEDs) has been demonstrated in recent years as high radiance with minimal power consumption is a desired feature. However, the light output at such conditions from PeLEDs is typically very low, and the maximum in external quantum efficiency (EQE) and energy conversion efficiency (ECE) are achieved at large biases with significant power consumption. Here, we explore the possibility of achieving maximums in EQE and ECE at sub band gap voltages for PeLEDs. Our analysis consistently interprets otherwise scattered experimental data from literature, identifies the limits for low voltage operation, and elucidates optimization routes for sub band gap high radiance operation of PeLEDs.

Figures (4)

• Figure 1: Variation of EQE/ECE as a function of $J$. The experimental results (symbols) are from Li et al.Li2024. The solid lines are model predictions. The inset shows that the EQE/ECE varies linearly with $V_{app}$ with a slope proportional to $E_g$. The parameters used in model predictions are available in SI.
• Figure 2: Comparison of model predictions with experimental results on ultra low voltage operation of LEDs. The experimental results are from Lian et al.lian2022ultralow for various types of LEDs as denoted in the legend. The theoretical prediction (solid line) correspond to parameters as mentioned in text. As anticipated by the model, $E_g-V_T$ is nearly independent of $E_g$ for a broad class of LED systems. The parameters used in model predictions are available in SI.
• Figure 3: Model predictions for PeLEDs to achieve maximum EQE at sub $E_g$ voltages. Here, the solid line denotes theoretical predictions as per eq. \ref{['eq:vdqe']}. The dashed line represents $V_{EQE}=E_g/q$. The solid symbols are experimental data from recent publications on high performance PeLEDs (red circle from Li et al.Li2024, blue square from Jia et al.jia2021excess, magenta triangle from Zhao et al.zhao2020thermal, black diamond from Zheng et al.zheng2024ultralow, and green triangle from Wang et al.wang2025efficient). The open symbols denote estimated $V_{D,QE}$ as obtained through eqs. \ref{['eq:vdqe']}-\ref{['eq:Vappqe']}. The parameters used in model predictions are available in SI.
• Figure 4: Comparison between $V_{ECE}$ and $V_{EQE}$ for PeLEDs (for $E_g=1.6\,\mathrm{eV}$). The dashed line represent $V_{EQE}=V_{ECE}$ conditions while the solid lines denote model predictions as $k_1$ is varied for various listed $k_2$. The solid symbols are experimental data from recent literature (circle from Li et.Li2024 and triangle from Jia et al.jia2021excess). Here, we find that $V_{ECE}< V_{EQE}$ as predicted by the theoretical model. The parameters used in model predictions are available in SI.