Design of broadband optical gain in GaSb-based waveguide amplifiers with asymmetric quantum wells

Technical Summary: Design of Broadband Optical Gain in GaSb-based Waveguide Amplifiers with Asymmetric Quantum Wells

Problem Statement
The development of optoelectronic devices operating in the 2–3 µm spectral window is critical for advancing molecular spectroscopy, optical coherence tomography (OCT), and free-space communication. While broadband light sources in this region offer benefits such as reduced scattering and higher axial resolution, there is a need for sources that combine broad spectral coverage, high output power, and electrical pumping. Specifically, applications like broadband tunable lasers and superluminescent diodes require a flat gain spectrum to minimize output power variations. Although bandgap engineering strategies (varying quantum well thicknesses and compositions) have been successful in shorter wavelength systems (GaAs/InP), their application to GaSb-based heterostructures for the mid-infrared requires systematic optimization to manage the complex degrees of freedom in device design.

Methodology
The authors employed a systematic design strategy using the "Harold" self-consistent simulation environment to model GaSb-based semiconductor amplifiers. The approach involved:

• Simulation Framework: The Harold tool solves coupled physics equations, including Poisson's equation, current continuity, carrier capture–escape balance, photon rate equations, and heat flow. It utilizes a parabolic band approximation to evaluate optical gain and accounts for carrier transport in multi-quantum well (MQW) systems.
• Material System: The study focused on type-I compressively strained Ga$_{0.73}$In$_{0.27}$Sb quantum wells (QWs) separated by Al$_{0.25}$Ga$_{0.75}$As$_{0.02}$Sb$_{0.98}$ barriers, lattice-matched to a GaSb substrate.
• Design Variables: The primary design variable was the thickness of the QWs (ranging from 7 nm to 13 nm). The authors prioritized varying thickness over alloy composition to leverage higher fabrication precision.
• Validation: Simulation parameters were validated against experimental data from a double-QW (DQW) structure. Photoluminescence (PL) spectra and light-current (L-I) characteristics of a broad-area laser diode (2 mm cavity, 80 µm stripe) were used to benchmark the model, showing excellent agreement in peak wavelengths (~1950 nm and ~2090 nm) and threshold behavior.

Key Contributions and Results
The study demonstrates that asymmetric MQW structures can achieve flat and wide gain spectra by leveraging carrier density and transition energy tuning across wells of varying thicknesses.

1. Single and Double QW Analysis:

   • Simulations of single QWs (SQW) revealed that gain peaks shift from the E1-HH1 transition to the E1-LH1 transition as carrier density increases. For instance, a 13 nm QW showed a sharp peak shift from 2097 nm to 1800 nm at high carrier densities (~2.6e18 cm⁻³).
   • In DQW structures with asymmetric thicknesses (e.g., a fixed 7 nm top QW and a variable bottom QW), the authors identified specific carrier densities where transitions from both wells contribute simultaneously. A DQW with 7 nm and 13 nm wells achieved a gain bandwidth of ~256 nm FWHM with a flat gain profile.

2. Optimized Multi-QW (MQW) Design:

   • The authors optimized a structure (Structure C) comprising one 7 nm QW and three 13 nm QWs, keeping the total strain within relaxation limits.
   • At a carrier density of ~1.73e18 cm⁻³, this structure produced a flat gain spectrum with a full-width at half-maximum (FWHM) exceeding 340 nm (specifically ~342 nm).
   • The gain spectrum covered the range of approximately 1953 nm to 2140 nm with a maximum gain of ~1253 cm⁻¹ and a gain fluctuation of only ~73 cm⁻¹. This represented a roughly two-fold improvement in maximum gain and a ~100 nm increase in bandwidth compared to the reference DQW structure.
   • Further simulations on a modified waveguide (shorter active region to suppress cavity feedback) allowed access to higher carrier densities, enabling the E1-LH1 transition from the thicker QWs. This resulted in a broader spectral window (1760–2150 nm) with a peak gain of ~1733 cm⁻¹, though with reduced flatness.

3. Temperature Dependence:

   • Simulations indicated that increasing the temperature from 10°C to 100°C broadened the FWHM to approximately 400 nm. However, this came at the cost of a significant reduction in maximum gain (~195 cm⁻¹) and a shift in peak wavelengths toward longer values.

Significance and Claims
The paper claims to present a robust design framework for broadband optical amplifiers and superluminescent diodes in the mid-infrared. By validating the simulation model against experimental data, the authors establish a reliable method for predicting gain features without extensive experimental iteration. The primary significance lies in demonstrating that asymmetric MQW structures with varying thicknesses can effectively broaden the gain spectrum beyond 2 µm while maintaining a flat profile. The study bridges the gap between theoretical potential and practical device design, offering a pathway to develop high-performance broadband sources for sensing and imaging applications. The authors note that while strain-compensation architectures could further optimize these designs, the current work focuses on feasible structures within standard strain limits. Future work is suggested to explore additional QWs and strain profiles to extend the wavelength coverage toward 3 µm.