This work presents a machine learning-driven approach integrated with in situ high-energy electron diffraction (RHEED) monitoring to automate the optimization of quantum dot (QD) growth for lasers. A 3.2-fold increase in photoluminescence intensity is achieved alongside a narrow spectral linewidth. The QD lasers fabricated demonstrate efficient lasing at 1240 nm with low threshold currents, highlighting a scalable alternative to manual optimization.
Abstract
Traditional methods for optimizing light source emissions rely on a time-consuming trial-and-error approach. While in situ optimization of light source gain media emission during growth is ideal, it has yet to be realized. In this work, in situ reflection high-energy electron diffraction (RHEED) is integrated with machine learning (ML) to correlate the surface reconstruction with the photoluminescence (PL) of InAs/GaAs quantum dots (QDs), which serve as the active region of lasers. A lightweight ResNet-GLAM model is employed for the real-time processing of RHEED data as input, enabling effective identification of optical performance. This approach guides the dynamic optimization of growth parameters, allowing real-time feedback control to adjust the QDs emission for lasers. InAs QDs on GaAs substrates are successfully optimized, with a 3.2-fold increase in PL intensity and a reduction in full width at half maximum (FWHM) from 36.69 to 28.17 meV. Automated, in situ self-optimized lasers with 5-layer InAs QDs achieved electrically pumped continuous-wave operation at 1240 nm with a low threshold current of 150 A cm−
2 at room temperature, an excellent performance comparable to samples grown through traditional manual multi-parameter optimization methods. These results mark a significant step toward intelligent, low-cost, and reproductive light emitters production.
Advanced Science, EarlyView. Read More
