A Machine Learning Framework for Quantum Cascade Laser Design - slideshare 268385094

Aug 5, 2026

Upload by Miles

A machine learning framework is presented for quantum cascade laser (QCL) design optimization, using a multi-layer perceptron to predict a laser figure of merit (FoM) linked to threshold gain from only layer thicknesses and the applied electric field. Training uses designs generated by random perturbations of a baseline 10-layer structure, and validation is performed with a 1D Schrödinger solver. Prediction error is 5–15% when a laser transition exists, and 35–70% otherwise. The method accelerates data collection from 32 hours for 27,000 designs to 8 hours for 907 million, enabling identification of high-FoM structures and layer edits that raise FoM from 94.7 to 141.2 eV ps Å².

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

I. Introduction

Quantum cascade lasers and design goals

Prior computational and ML optimization methods

II. Methods

Identifying starting design and QCL parameters

A Machine Learning Framework for Quantum Cascade Laser Design Andres Correa Hernandez and Claire F. Gmachl Department of Electrical and Computer Engineering, Princeton University, Princeton, NJ, 08545 USA A multi-layer perceptron neural network was used to predict the laser transition figure of merit, a measure of the laser threshold gain, of over 900 million Quantum Cascade Laser designs using only layer thicknesses and the applied electric field as inputs. Designs were generated by randomly altering the layer thicknesses of an initial 10-layer design. Validating the predictions with our 1D Schrödinger solver, the predicted values show 5% to 15% error for structures where a laser transition could occur, and 35% to 70% error for structures where there was no laser transition. The algorithm allowed (i) for the identification of high figure of merit structures, (ii) recognition of which layers should be altered to maximize the figure of merit at a given electric field, and (iii) increased the original design figure of merit of 94.7 to 141.2 eV ps Å2, a 1.5-fold improvement and significant for QC lasers. The computational time for laser design data collection is greatly reduced from 32 hours for 27000 designs using our 1D Schrödinger solver on a virtual machine, to 8 hours for 907 million designs using the machine learning algorithm on a laptop computer. I. Introduction Quantum Cascade Lasers (QCLs) are optoelectronic devices which mainly operate in the mid-infrared [1-3] and THz regimes [4, 5] of the electromagnetic spectrum. These lasers are useful for atmospheric trace chemical and particle detection [6, 7], low-visibility communication [8, 9], and labelfree medical imaging [10, 11], among various other applications. A QCL design consists of alternating well and barrier material such that the bandstructure forms a multi-quantum well heterostructure. Electrons are pumped into the system and photons are emitted through intersubband transitions in the conduction band; between optical transitions the electrons continue traveling across the heterostructure through longitudinal optical (LO) phonon, and other scattering. Ideally, above laser threshold a photon is released for every electron and every period of the active region and injector that is present in the entire active core of the structure. The emission wavelength of the QCL can be tuned by changing specific parameters of the design such as layer thickness, applied electric field, material compositions, and material system. A combination of human intuition and computational analysis using solutions from a Schrödinger solver is a common approach to QCL design. There have been several attempts at optimizing the large design parameter space of QCLs using various computational methods. Use of a genetic algorithm (GA) increased the wall-plug efficiency of a mid-infrared QCL by 7%[12], while another GA method was able to optimize a THz QCL transition frequency over a 2.9 THz range [13], among other applications [14-17] . Simulated annealing on a triple step quantum well design kept the transition energy around 50 meV [18], and also optimized superstructure gratings in QCLs to achieve non-equidistant frequencies [19] . Inverse spectral theory maximizes the gain in a quantum well laser [20] as well as optimizes the active region of a 12 µm QCL [21] . Machine learning (ML) for optimization of semiconductor devices has been applied to defect identification during the fabrication process [22-24] and to designing nanostructures in nanophotonics [25-27] . ML approaches for QCLshave so far looked at improving calculation time for modal gain [28], predicting resonant mode characteristics of QCLs in the THz regime [29], prediction of emission spectra of THz quantum cascade random lasers [30], predicting the threshold gain from higher-order modes given the refractive index profile of a QCL cavity [31], as well as the labeling of relevant wavefunctions [32] . In reference [33] an

FAQs

What inputs does the machine learning model use to predict QCL performance?

The model uses layer thicknesses and the applied electric field to predict the laser figure of merit related to threshold gain.

How is the model validated for predicted designs?

Predicted FoM values are validated using a 1D Schrödinger solver.

What performance gains does the framework achieve in laser design and computation time?

It identifies high-FoM structures and suggests which layers to alter, improving FoM from 94.7 to 141.2 eV ps Å² while reducing design data collection time from 32 hours (27,000 designs) to 8 hours (907 million designs).

Aug 5, 2026

Upload by Miles