🔭 astrophysics

Ultrastable lasers: investigations of crystalline mirrors and closed cycle cooling at 124 K

This paper investigates the photo-thermo-optic and photo-modified birefringence effects in crystalline AlGaAs/GaAs optical coatings across various temperatures and reports the successful transition of a 124 K ultra-stable silicon cavity system from liquid nitrogen to a low-maintenance closed-cycle pulse-tube cryo-cooler.

Original authors: C. Y. Ma, J. Yu, T. Legero, S. Herbers, D. Nicolodi, M. Kempkes, F. Riehle, D. Kedar, J. M. Robinson, J. Ye, U. Sterr

Published 2026-07-01

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: C. Y. Ma, J. Yu, T. Legero, S. Herbers, D. Nicolodi, M. Kempkes, F. Riehle, D. Kedar, J. M. Robinson, J. Ye, U. Sterr

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to keep a perfectly steady heartbeat for a clock so precise it could measure the age of the universe without losing a single second. To do this, scientists use ultra-stable lasers trapped inside a box called an "optical cavity." Think of this cavity like a hallway with mirrors at both ends; the laser light bounces back and forth millions of times. The length of this hallway determines the "note" (frequency) the laser sings. If the hallway expands or shrinks even a tiny bit, the note changes, and the clock becomes inaccurate.

For years, scientists have used mirrors coated with layers of glass-like materials (dielectrics) to keep these lasers stable. However, these mirrors have a tiny flaw: they vibrate slightly due to heat (like a jelly wobbling), which creates "noise" that messes up the laser's note.

To fix this, the researchers in this paper switched to crystalline mirrors made of a special semiconductor material (AlGaAs/GaAs). Think of these as mirrors made from a perfect, rigid crystal instead of a wobbly jelly. These crystals are much stiffer and vibrate less, promising a much quieter, more stable laser.

However, the scientists discovered that these new crystal mirrors have a strange, unexpected personality trait: they react to light in a way that isn't just about heat.

The "Light-Induced Mood Swing"

Usually, when light hits a mirror, it warms it up slightly, causing it to expand. This is a well-known effect called the "photo-thermo-optic effect." But the researchers found that with these crystal mirrors, the light does something else entirely.

When they shone light into the cavity, the mirror's internal structure changed in a way that made it act differently for light vibrating in different directions (a property called birefringence). It's as if the mirror suddenly decided to stretch in one direction but not the other, just because it was being looked at by the laser.

The Temperature Twist: This effect behaved very differently depending on how cold the mirror was. At room temperature, the mirror reacted quickly. But at very cold temperatures (like -150°C), the reaction was incredibly slow, taking hours to settle down, like a heavy door that takes a long time to swing open.
The "External Flashlight" Test: To figure out what was happening, the team shined external lights (LEDs) of different colors onto the back of the mirrors. They found that if the light had enough energy (like blue or green light), it triggered a massive reaction. If the light was too weak (like infrared), nothing happened.

The Secret Mechanism: The "Electron Traffic Jam"

The paper proposes a simple explanation for this behavior, using the properties of semiconductors (the same stuff in computer chips).

Photo-Excitation: When high-energy light hits the crystal, it knocks loose tiny electrical particles called electrons (like knocking marbles loose from a shelf).
The Migration: These loose electrons don't stay put; they migrate through the crystal layers.
The Electric Field: As they move, they create a tiny electric field, like a static shock building up.
The Result: This electric field physically squeezes or stretches the crystal lattice, changing how it handles light (birefringence).

The researchers calculated that this tiny electric field is strong enough to explain the changes they saw. It's a bit like a crowd of people (electrons) moving to one side of a room, causing the floor to tilt slightly.

The "Closed-Loop" Upgrade

The paper also describes a practical upgrade to their lab setup. Previously, to keep their mirrors at a chilly 124 K (-149°C), they had to constantly pour in liquid nitrogen, like refilling a giant thermos every week. This was messy and required constant maintenance.

They replaced this with a closed-cycle cooling system using a pulse-tube cryo-cooler. Imagine this as a high-tech refrigerator that uses a mechanical pump to circulate cold helium gas instead of needing a delivery truck to bring in liquid nitrogen. This new system is much more reliable and "low maintenance," keeping the mirrors at the perfect temperature without the hassle of refills.

Why This Matters

The main takeaway is that while these crystal mirrors are excellent at reducing the "jelly-like" vibrations of the old mirrors, they introduce a new kind of noise caused by the light itself interacting with the electrons in the crystal.

The scientists conclude that to get the best possible performance from these mirrors at room temperature, they need to stabilize the laser power extremely well (to within one part in a million). If they can master this "light-mirror interaction," these crystalline mirrors could lead to the most precise clocks and sensors ever built, capable of detecting things like gravitational waves or dark matter with unprecedented clarity.

Technical Summary: Investigations of Crystalline Mirrors and Closed Cycle Cooling at 124 K

Problem Statement
Ultrastable lasers are critical for quantum sensors and optical clocks, with current state-of-the-art systems achieving fractional frequency instabilities in the mid- $10^{-16}$ range. However, the performance of these systems is increasingly limited by Brownian thermal noise arising from the mechanical loss of conventional dielectric mirror coatings (e.g., $Ta_2O_5/SiO_2$ ). Crystalline AlGaAs/GaAs Bragg reflectors offer a promising alternative, possessing mechanical loss approximately ten times lower than dielectric coatings, which theoretically reduces Brownian noise by a factor of three. Despite this advantage, previous studies at cryogenic temperatures revealed novel noise contributions in crystalline coatings, specifically "intrinsic birefringent noise" and unexpected photo-induced frequency shifts that exceed the standard photo-thermo-optic effects observed in dielectric mirrors. The physical mechanisms driving these photo-induced effects in semiconductor-based coatings, particularly their dependence on temperature, intracavity power, and external illumination, remain not fully understood. Additionally, operational reliability remains a challenge for cryogenic systems, which often rely on liquid nitrogen (LN2) dewars requiring frequent refills.

Methodology
The authors investigated these phenomena using three ultra-stable optical cavities operating at 1.5 µm:

Si5: A 21 cm crystalline silicon cavity with AlGaAs coatings, operating at 124 K.
Si6: A 6 cm crystalline silicon cavity with AlGaAs coatings, operating at 4 K and 16 K.
ULE: A 48 cm fused silica cavity with AlGaAs coatings on fused silica substrates, operating at 297 K.

The experimental approach involved:

Power Step Responses: Sudden changes in intracavity optical power were applied to observe the temporal response of the cavity's resonance frequency and birefringent splitting (the frequency difference between fast and slow polarization eigenmodes).
Normalization: Measurements were normalized to the average light intensity ( $I$ ) at the coating surface to account for differences in cavity finesse and mode size.
External Illumination: The ULE cavity at room temperature was illuminated from outside the vacuum chamber with high-power LEDs of various wavelengths (450 nm to 1.5 µm) to probe the wavelength dependence of the photo-induced effect relative to the GaAs bandgap (~890 nm).
Cooling System Upgrade: To improve reliability, the LN2 cooling system for the 124 K setup was replaced with a closed-cycle pulse-tube cryo-cooler circulating cold helium gas, aiming to eliminate the need for frequent cryogen refills.

Key Results

Photo-Induced Birefringence: A sudden change in optical power induces a modification in the coating birefringence ( $\Delta n_{photo}$ $Δ n_{p h o t o}$ ), leading to a shift in the frequency splitting between polarization eigenmodes. This effect is distinct from thermal expansion.
- Temperature Dependence: The response time varies significantly with temperature. At 4 K and 297 K, the response occurs on the order of seconds. At 124 K, the response is significantly slower, occurring over hours.
- Intensity Dependence: At room temperature, increasing intracavity intensity reduces the line splitting, whereas at 4 K and 124 K, the sign of the effect is opposite. The room temperature response appears highly nonlinear, suggesting saturation at high intensities.
Wavelength Dependence and Mechanism: External illumination experiments revealed that the sensitivity of the birefringence to light intensity drops by five orders of magnitude when photon energies fall below the GaAs bandgap (wavelengths > 1050 nm). Above the bandgap, sensitivity is high.
- Proposed Mechanism: The authors propose a model where photon absorption creates free charge carriers in the first few micrometers of the coating. These carriers migrate through the GaAs/AlGaAs heterostructure, creating an internal electric field. This field induces a change in birefringence via the linear electro-optic (Pockels) effect. The model estimates that the observed maximum birefringence change ( $8.7 \times 10^{-6}$ ) corresponds to an internal electric field of approximately 155 kV/m.
Closed-Cycle Cooling: The transition from LN2 to a pulse-tube cryo-cooler for the 124 K system was successful. The new system maintains the cavity at 124 K with a cooling power of ~35 W at 85 K. The outer radiation shield demonstrates sub-millikelvin temperature stability, comparable to the previous LN2-based system, while significantly reducing maintenance requirements.

Significance and Claims
The paper claims to provide expanded investigations into the photo-induced effects in crystalline AlGaAs coatings, moving beyond previous observations at cryogenic temperatures to include room temperature behavior and broadband external illumination.

Mechanism Identification: The study strongly suggests that the observed photo-induced birefringence is linked to the semiconductor properties of the coatings, specifically the photo-generation of charge carriers and subsequent electro-optic effects, rather than purely thermal mechanisms.
Stability Requirements: Based on the measured sensitivities, the authors calculate that to achieve a photo-induced instability below $1 \times 10^{-17}$ at room temperature, the intracavity power must be stabilized to a relative level of $10^{-6}$ .
Operational Advancement: The successful implementation of a closed-cycle cooling system for a 124 K silicon cavity demonstrates a viable path toward low-maintenance, transportable cryogenic systems, addressing a key barrier for the deployment of ultrastable lasers in field applications.

The authors conclude that while crystalline coatings offer superior thermal noise performance, understanding and mitigating these photo-induced effects is essential for realizing their full potential in next-generation optical clocks and frequency standards. Future work will focus on characterizing the noise contributions with and without external light to determine the ultimate performance limits of these coatings at room temperature.

The "Light-Induced Mood Swing"

The Secret Mechanism: The "Electron Traffic Jam"

The "Closed-Loop" Upgrade

Why This Matters

More like this