Photonic Integrated Circuits for Gas Detection Using Interferometric Sensing

Abstract

This thesis focuses on designing and optimizing a photonic gas sensor based on the Mach-Zehnder Interferometer (MZI) architecture for detecting gases, particularly Carbon Dioxide (CO₂) and Hydrogen (H₂), through refractive index variations. Silicon photonic technologies provide a miniaturized, low-cost, and highly sensitive solution compatible with CMOS fabrication. Polyhexamethylene biguanide (PHMB) and Palladium Tungsten Trioxide (PdWO₃) were employed as sensing layers for CO₂ and H₂, respectively. Simulations using Ansys Lumerical MODE and INTERCONNECT enabled precise modeling and optimization of optical properties, waveguide geometries, and system performance.

The optimized CO₂ sensor demonstrated a sensitivity of 246 pm/ppm and a detection limit of 40 ppb, outperforming similar sensors. The H₂ sensor achieved a sensitivity of 0.2 pm/ppm and a detection limit of 50 ppm. The exposed MZI sensor exhibited a sensitivity of 1443 nm/RIU with a detection limit of 6.93×10⁻⁶ RIU. This work emphasizes the significance of balancing arm lengths to optimize sensitivity and free spectral range (FSR), achieving effective operation over a broad dynamic range.

The fabricated exposed slab sensing arm MZI was characterized by measuring transmission changes when exposed to air and tea tree essential oil vapor, showing an FSR of 0.45 nm and a spectral shift of 0.03 nm. Thermo-optic phase shifters (TOPS) enabled precise phase control by inducing refractive index changes via localized heating through electrical contacts. Optimizing parameters such as heater length, width, placement, and segmentation improved phase shift efficiency and reduced power consumption.

Future directions include developing MZI arrays with diverse functionalized sensing layers for detecting a broader range of gases, including volatile organic compounds (VOCs). Incorporating TOPS for phase optimization and employing machine learning models to analyze sensor outputs can further enhance performance and applicability.

Date of Award	12 Dec 2024
Original language	American English
Supervisor	JAIME Viegas (Supervisor)