Laser Frequency Locking Using Regenerative Spectral Hole for Ultrasound Optical Tomography

Abstract

This thesis explores the feasibility of using regenerative spectral hole burning (SHB) in cryogenically cooled Tm³⁺:LaF₃ crystals as a laser frequency reference for Ultrasound Optical Tomography (UOT) systems. UOT is an imaging technique that combines the deep penetration of ultrasound with the high contrast of near-infrared light, offering significant potential for biomedical imaging. A major challenge in such systems is the detection of weak ultrasound-tagged photons, which requires a narrowband and frequency-stable optical filter. While SHB has been employed as the filtering mechanism, the current setup still depends on a Fabry–Pérot cavity for laser stabilization, which adds complexity and is sensitive to environmental disturbances.
To address this, the thesis investigates a cavity-free stabilization scheme in which the regenerative spectral hole itself acts as both filter and frequency reference. A theoretical framework based on Maxwell-Bloch equations, and a three-level population model is reviewed, and the Pound-Drever-Hall technique is used for error signal extraction.
However, in this work, stable locking was not yet achieved experimentally. Two key limitations were identified: Firstly, the mechanical vibrations from the cryostat compressor significantly degraded the spectral hole, broadening its FWHM from 0.99 MHz to 9.25 MHz and reducing its depth by a factor of 3; and secondly, large power fluctuations caused by poor modulation efficiency and multimode fiber coupling resulted in a low signal-to-noise ratio, obscuring the error signal. These findings suggest that regenerative SHB-based locking requires more stringent environmental control and more carefully designed optical setup. While a fully stabilized system was not realized, this work provides critical insight into the constraints of regenerative locking and proposes future directions for achieving robust performance.

Popular Abstract

Imagine trying to look through a thick fog with a flashlight — the light bounces everywhere, and you can’t see what’s behind it. This is what happens when scientists try to use light to look deep into the human body. Although light can carry rich information, it gets scattered too much in tissue to form a clear image.
Ultrasound Optical Tomography (UOT) is a technique which is developed to solve this problem. It uses ultrasound to “tag” certain regions inside the tissue, subtly changing the color of the light that passes through. By detecting these changes, UOT can form high-resolution images several centimeters below the surface — far deeper than conventional optical techniques.
But to detect these subtle changes, scientists need a laser that is extremely stable in its color, or more precisely, its frequency. Even the tiniest drift can make the system miss the tagged light signal. This leads to a central challenge in many optical experiments: how do you keep a laser’s frequency locked precisely to a desired value, even when the environment is noisy or unstable?
A laser’s frequency can be thought of like the pitch of a musical note. But unlike a violin string that naturally stays tuned (at least for a while), a laser’s frequency tends to drift due to temperature changes, mechanical vibrations, or internal noise. “Locking” a laser means continuously comparing its frequency to a stable reference, just like having an automatic tuner constantly adjusting the pitch to keep the note exactly right.
Traditionally, this frequency reference comes from an optical cavity, essentially two mirrors facing each other. The length of the cavity is fixed, which defines a set of fixed resonant frequencies. However, these cavities are bulky (and expensive). In this project, we explore a different approach, which is to use a spectral hole in a cryogenically cooled rare-earth-doped crystal as the reference.
A spectral hole is a narrow dip in the crystal’s absorption profile, created by shining laser on the crystal in just the right way. If the laser frequency matches this spectral hole, the light passes through; otherwise, it gets absorbed. What makes this method special is that the spectral hole itself can regenerate — adapting slightly to the laser while still providing strong feedback, like a dynamic target that keeps guiding the laser frequency back to the center.
To make this feedback precise, we use a technique called the Pound-Drever-Hall (PDH) method, which adds a small modulation to the laser and detects how the light is altered after passing through the crystal. This produces an “error signal” — a measure of how far off-tune the laser is — which is then used to correct it in real time.
In this thesis, we built a theoretical model of this locking system, tested it experimentally using thulium-doped LaF₃ crystals, and analyzed how well it performs under different conditions, including noise and power fluctuations. Our results show that regenerative spectral hole locking is not only feasible but offers a simpler alternative to traditional cavity-based systems — especially valuable for applications like UOT.

• author: Guo, Jiayu
• supervisor: Stefan Kröll; Lars Rippe; Maria Ruchkina
• organization: Atomic Physics; Department of Physics
• course: PHYM03 20242
• year: 2025
• type: H2 - Master's Degree (Two Years)
• subject: Physics and Astronomy
• publication/series: LRAP - Lund reports in Atomic Physics
• report number: 616(2025)
• language: English
• id: 9201944
• date added to LUP: 2025-06-23 08:23:48
• date last changed: 2025-06-23 08:23:48