🔬 applied physics

Spectral and Small-Signal Electroluminescence Analysis of Carrier Dynamics in Dual-Color InGaN/GaN Light-Emitting Diodes

This study utilizes spectral analysis and small-signal electroluminescence to reveal that carrier dynamics in dual-color InGaN/GaN LEDs are governed by non-uniform recombination and transport, where green emission dominates at low currents due to bandgap advantages while blue emission prevails at high currents due to reduced quantum-confined Stark effects, ultimately limiting the effective active region to the 2–3 quantum wells nearest the p-GaN side.

Original authors: Xuefeng Li, Rob Armitage, Daniel Feezell

Published 2026-03-23

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Xuefeng Li, Rob Armitage, Daniel Feezell

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: The "Two-Color Light Bulb" Experiment

Imagine you are trying to build the perfect light bulb that can glow both Blue and Green. In the world of LEDs (the lights in your phone or streetlamps), making these colors is tricky. The paper you read is like a detective story where scientists are trying to figure out how electricity (carriers) moves inside a tiny light bulb to decide which color gets to shine the brightest.

They built special "dual-color" LEDs that have layers of blue and green material stacked on top of each other. By watching how these lights behave, they learned how to make better, brighter, and more efficient lights for the future.

The Cast of Characters

To understand the experiment, let's meet the players:

The Carriers (Electrons and Holes): Think of these as tiny delivery trucks carrying packages (energy). Their job is to drive into the light bulb, find a parking spot (a Quantum Well), and drop off their package to create light.
The Quantum Wells (QWs): These are the parking spots. Some spots are painted Blue, and some are painted Green.
- Blue Spots: Harder to park in (higher energy needed), but once you park, you can leave quickly.
- Green Spots: Easier to park in (lower energy needed), but once you park, you tend to get stuck there longer.
The V-Pits (The Secret Tunnels): Usually, the parking spots are separated by thick walls (barriers) that the trucks can't easily cross. However, these scientists built tiny, pyramid-shaped holes in the walls called V-pits. Think of these as secret tunnels or elevators that help the trucks move between the different floors (layers) of the parking garage much faster.

The Story of the Experiment

The scientists tested four different parking garage designs:

Some had Blue spots on the bottom and Green on top.
Some had Green on the bottom and Blue on top.
Some had two of each.

They then turned on the electricity (the traffic) and watched what happened at different speeds (current densities).

1. The "Low Traffic" Rush Hour (Low Current)

When there are only a few trucks on the road (low current), the Green spots get all the action.

Why? The Green spots have a "lower entrance fee" (narrower bandgap). The trucks don't need much energy to get in, so they flock to the Green spots first. Even if the Green spots are on the top floor (farther from the entrance), the trucks find a way there because the Green spots are just easier to enter.

2. The "High Traffic" Rush Hour (High Current)

When they flood the road with millions of trucks (high current), the story changes. The Blue spots start to shine much brighter.

Why? The Green spots get so crowded that the trucks get stuck (a phenomenon called the Quantum Confined Stark Effect). It's like a traffic jam in the Green parking lot.
Meanwhile, the Blue spots are less crowded and have better "exit ramps" (wavefunction overlap). The trucks can park and leave quickly, creating a steady stream of light. So, at high speeds, the Blue light takes over.

3. The "Secret Tunnel" Effect (V-Pits)

The scientists wanted to know: Can the trucks actually travel all the way to the top floor of the garage?

They found that the V-pit tunnels do help! They allow trucks to move between floors better than in normal lights.
However, there is a limit. Even with the tunnels, the trucks mostly stick to the bottom 2 or 3 floors (the ones closest to the entrance). Very few trucks make it all the way to the very top floor.
The Lesson: It's like a busy restaurant. Even if you have a great elevator (V-pits), the tables closest to the door (p-GaN side) get filled first. The tables at the back of the room often stay empty because the waiters (carriers) get tired or stuck before they get there.

The "Speed Test" (Small-Signal Analysis)

To prove their theory, the scientists didn't just look at the light; they flickered the electricity on and off super fast (like a strobe light) and measured how fast the light could react.

Fast Reaction: Means the trucks are parking and leaving quickly (efficient).
Slow Reaction: Means the trucks are stuck in traffic.

They found that the light didn't react uniformly. It was as if the parking garage had different "speed zones." This confirmed that the trucks weren't spread out evenly; they were clustered in specific areas, proving that the "secret tunnels" only help so much.

The Takeaway (Conclusion)

This research teaches us three main things about making better LED lights:

Green wins at low power: If you want a dim, efficient green light, the material naturally prefers the green spots.
Blue wins at high power: If you want a super bright light, the blue spots are better because they don't get as traffic-jammed.
Don't build too many floors: Even with the best "elevators" (V-pits), you can't effectively use more than 2 or 3 layers of parking spots. If you build a 10-story garage, the top 7 floors will just be empty waste.

In short: To make the perfect light, you need to balance the number of layers and use the "V-pit tunnels" wisely, but you shouldn't try to force the trucks to go to the very top of the building if the path is too long.

1. Problem Statement

Improving the quantum efficiency of long-wavelength (green, amber, red) InGaN/GaN Light-Emitting Diodes (LEDs) is critical for solid-state lighting and full-color displays. A major bottleneck in these devices is non-uniform carrier distribution across Multiple Quantum Wells (MQWs).

The Issue: Due to strong hole confinement and slow carrier transport, carriers tend to recombine primarily in the quantum well (QW) closest to the p-GaN layer, leaving deeper QWs underutilized. This leads to high local carrier densities, exacerbating non-radiative Auger-Meitner recombination.
The Challenge: In single-color LEDs, it is experimentally difficult to isolate the emission contributions of specific QWs to study these transport dynamics.
The Context: While V-pit engineering (creating inverted pyramid defects) has been proposed to enhance vertical and lateral carrier transport, the precise extent of its effectiveness in activating deeper QWs in dual-color structures remains unclear.

2. Methodology

The authors employed a combination of Spectral Analysis and Small-Signal Electroluminescence (SSEL) to investigate carrier dynamics in dual-color LEDs.

Device Fabrication:
- Four c-plane InGaN/GaN LED structures were grown via Metal-Organic Chemical Vapor Deposition (MOCVD).
- All devices shared the same epitaxial base but varied in QW arrangement and quantity (2 or 4 QWs).
- Configurations:
  1. n/b/g/p: Blue QW (16% In) followed by Green QW (19% In).
  2. n/g/b/p: Green QW followed by Blue QW.
  3. n/2b/2g/p: Two Blue QWs followed by Two Green QWs.
  4. n/2g/2b/p: Two Green QWs followed by Two Blue QWs.
- All structures utilized V-pit engineering to facilitate carrier transport.
Experimental Techniques:
- Electroluminescence (EL) Spectral Analysis: Emission spectra were measured across varying current densities ( $J$ ). The authors used Fast Fourier Transform (FFT) filtering to remove Fabry–Pérot resonances and modeled the spectra using two Gaussian functions to mathematically separate blue and green emission peaks.
- Photon Counting: The total number of photons ( $N$ ) for each color was calculated by integrating the spectral intensity over specific wavelength ranges, separated by a cutoff at the spectral dip (~462 nm).
- Small-Signal Electroluminescence (SSEL): Modulation response ( $S_{21}$ ) was measured using an avalanche photodiode (APD) and optical filters (shortpass/longpass) to isolate blue or green emission. This allowed the extraction of the -3 dB bandwidth, which serves as a proxy for the effective differential carrier recombination lifetime.

3. Key Contributions

Decoupling QW Contributions: Successfully demonstrated a method to experimentally separate and analyze emission from specific QWs (blue vs. green) in a dual-color stack, bypassing the limitations of single-color LED analysis.
Quantifying V-pit Efficacy: Provided experimental evidence on the limits of V-pit engineering, determining the "effective active region" volume in these specific LED designs.
Dynamic Carrier Redistribution: Revealed how carrier distribution shifts between QWs as a function of current density, driven by the interplay between bandgap energy, Quantum Confined Stark Effect (QCSE), and wavefunction overlap.

4. Key Results

A. Current Density Dependence of Emission

Low Current Density ( $J < 10-20$ A/cm²): Emission is dominated by the Green QW, regardless of its position relative to the p-GaN layer.
- Reason: The green QW has a narrower bandgap (higher Indium content), resulting in a lower turn-on voltage. This allows radiative recombination to occur in the green wells before the quasi-Fermi level separation is sufficient to activate the blue QWs.
High Current Density ( $J > 20$ A/cm²): Emission shifts significantly toward the Blue QW.
- Reason: Blue QWs exhibit a weaker QCSE and larger electron-hole wavefunction overlap compared to green QWs. At high carrier densities, these factors lead to faster radiative recombination rates in the blue wells, overtaking the green emission.
Photon Ratio ( $N_{blue}/N_{green}$ ): The ratio of blue to green photons increases by an order of magnitude as current density rises from 20 to 500 A/cm².

B. Carrier Transport and Distribution

Non-Uniform Recombination: The study confirms that carrier recombination is non-uniform. Even within a single QW type, carriers are not evenly distributed due to strong interwell transport effects.
Reabsorption: The data indicates that photon reabsorption is a weak process in these devices. If reabsorption were dominant, blue emission would decrease at higher currents as it passes through green QWs; however, blue emission increases, suggesting efficient carrier injection into the blue wells.
Effective Active Region:
- In structures with multiple QWs (e.g., n/2b/2g/p), the -3 dB bandwidth for the "far" QWs (those further from p-GaN) did not improve significantly compared to single QW structures.
- Conclusion: The effective active region contributing to strong recombination is limited to roughly 2 to 3 QWs on the p-GaN side. Carrier transport to the 3rd or 4th QW is significantly hindered by the thick barriers (18 nm) and the specific V-pit geometry used.

C. SSEL and Bandwidth Analysis

Roll-off Rate: The modulation response showed a -17 dB/decade roll-off (slower than the typical -20 dB/decade for uniform quasi-SQW LEDs). This confirms an uneven carrier distribution across the QW stack.
Bandwidth Trends:
- In n/b/g/p and n/g/b/p, green emission had a wider -3 dB bandwidth at low $J$ , indicating higher carrier density in the green wells.
- In n/2g/2b/p, the blue emission bandwidth became larger than the green at high $J$ , confirming that carrier redistribution favors the blue wells at high injection levels.

5. Significance

Optimization of Active Regions: The findings suggest that simply adding more QWs to increase efficiency is ineffective if carrier transport cannot reach them. For these specific V-pit engineered structures, the "sweet spot" is 2–3 QWs. Future designs must optimize barrier thickness or V-pit density to activate deeper wells.
Design Guidelines for Long-Wavelength LEDs: The study highlights that while V-pits improve transport, they do not fully solve the hole confinement issue in long-wavelength (high Indium) LEDs. Strategies to enhance hole injection and transport are still required to reduce Auger recombination.
Methodological Advancement: The combination of spectral decomposition and SSEL provides a robust, non-invasive toolset for diagnosing carrier dynamics in complex multi-QW LED structures, offering a pathway to better quantum efficiency modeling without needing to extract exact carrier lifetimes.