Genetic algorithm optimization of quantum Bragg mirror detectors

Optical absorption in quantum Bragg mirror detectors (QBMDs) relies on transitions between bound states inside a quantum well and bound or quasi-bound states in the continuum in a carefully tailored superlattice. The optimization of QBMDs lattice-matched to InP requires tuning seven specific parameters: the quantum well and barrier thicknesses for the left and right electronic Bragg mirrors (4 parameters), the number of periods in each mirror (2 parameters), and the main quantum well width (1 parameter). We aim to gain insights into the complex QBMD system by mapping transition energies and optimizing heterostructures for enhanced oscillator strength and photocurrent peak amplitude. With all the relevant parameters included in the search space, there are around 50 × 109 possibilities, making it impossible to fully map the search space. Thus, by employing genetic algorithms (GAs) for optimization, heterostructures with optimized oscillator strength and photocurrent amplitude are found, even in vast search spaces. Our investigation demonstrates the feasibility of optimizing QBMD heterostructures using GAs, providing essential design guidelines for improved performance of such devices.