A Novel Arm-Length Stabilization Scheme for Gravitational-Wave Detectors with AlGaAs/GaAs Coated Mirrors

Technical Summary: A Novel Arm-Length Stabilization Scheme for Gravitational-Wave Detectors with AlGaAs/GaAs Coated Mirrors

Problem Statement
Current ground-based gravitational-wave detectors (GWDs), such as Advanced LIGO, utilize an Arm-Length Stabilization (ALS) system to bring arm cavities onto resonance in a controlled manner. The existing ALS scheme employs an auxiliary 1064 nm laser that is frequency-doubled to 532 nm (green light) to pre-stabilize arm cavity lengths. However, future upgrades, specifically the planned A# upgrade to Advanced LIGO and third-generation observatories, intend to incorporate crystalline AlGaAs/GaAs coatings on test masses to reduce coating thermal noise. These coatings possess a bandgap corresponding to approximately 867 nm, causing excessive absorption of the 532 nm auxiliary beam. Consequently, the current frequency-doubled ALS system is unsuitable for detectors utilizing AlGaAs/GaAs mirrors, necessitating a new stabilization scheme that avoids absorption while maintaining precise control.

Methodology
The authors propose and experimentally demonstrate a novel multi-wavelength ALS scheme utilizing a 1596 nm auxiliary laser. This wavelength is chosen because it lies outside the absorption bands of AlGaAs/GaAs coatings. The core of the methodology involves establishing a phase-locked relationship between the 1596 nm auxiliary laser and the primary 1064 nm science laser via their respective harmonic generations at 532 nm.

The experimental setup, realized on a tabletop scale, consists of five subsystems:

1. 1064 nm SHG System: The primary 1064 nm science laser is frequency-doubled to 532 nm using a periodically poled potassium titanyl phosphate (PPKTP) crystal within a bowtie-shaped cavity.
2. 1596 nm THG System: A custom 1596 nm fiber laser serves as the auxiliary source. Its output undergoes single-pass frequency doubling to 798 nm, followed by sum-frequency generation (SFG) with the remaining 1596 nm light to produce a 532 nm beam. This process occurs in a copper oven housing SHG and SFG crystals.
3. 532 nm Phase-Locking System: The 532 nm beams from both the science (SHG) and auxiliary (THG) paths are combined. A beat signal is detected by a high-bandwidth photodetector. A Phase-Locked Loop (PLL) controls the relative phase between the two lasers by applying a frequency offset to the 1596 nm laser, locking the two 532 nm beams together.
4. Cavity Locking Systems: Both the 1596 nm and 1064 nm beams are injected into a linear optical cavity (400 mm long) equipped with dual-wavelength coated mirrors. Standard Pound–Drever–Hall (PDH) techniques are used to generate error signals for locking the cavity to either wavelength.
5. Transition Mechanism: The system allows for a controlled transition where the cavity is initially locked to the 1596 nm beam. By ramping the frequency offset in the PLL, the 1064 nm beam is brought into resonance. Once resonant, the control feedback is switched from the 1596 nm loop to the 1064 nm loop.

Key Contributions

• Novel Scheme Proposal: The paper introduces a multi-wavelength ALS architecture that replaces the 532 nm auxiliary beam with a 1596 nm beam, effectively bypassing the absorption limitations of AlGaAs/GaAs coatings.
• Theoretical Framework: The authors derive the frequency relationships required to ensure that when the 1596 nm laser is locked to a cavity resonance, the 1064 nm science laser can also be brought to co-resonance by tuning the PLL frequency offset by no more than the cavity Free Spectral Range (FSR).
• Experimental Demonstration: The team successfully built a tabletop prototype demonstrating:
  • Stable generation of 532 nm light via single-pass third-harmonic generation (THG) from a 1596 nm source.
  • Phase-locking of the 1596 nm and 1064 nm lasers via their 532 nm harmonics, achieving a beat note linewidth reduction from 145 kHz to 1.4 Hz.
  • Controlled cavity detuning of the 1064 nm beam relative to the cavity resonance via PLL offset adjustment.
  • A robust, repeatable transition from auxiliary (1596 nm) locking to science (1064 nm) locking.

Results

• Cavity Characterization: The experimental cavity exhibited linewidths of 6.09 ± 0.15 MHz (1596 nm) and 1.25 ± 0.06 MHz (1064 nm), corresponding to finesses of 61.6 and 299, respectively, consistent with theoretical predictions.
• Phase Locking Performance: The PLL successfully reduced the beat signal linewidth between the two 532 nm beams to approximately 1.4 Hz. The authors estimate that if frequency noise is dominated by the 1596 nm laser, the resulting frequency noise would be 0.47 Hz, translating to a length noise of ~10 pm for a 4 km arm cavity—well below the 1 nm linewidth of the proposed A# arm cavities.
• Locking Transition: The system demonstrated a successful transition sequence lasting approximately 100 seconds. The cavity was locked to the 1596 nm beam, the 1064 nm beam was swept into resonance by adjusting the PLL offset, and control was seamlessly transferred to the 1064 nm PDH loop. This process was repeated multiple times over several days without failure.

Significance and Claims
The paper claims to provide a viable proof-of-principle solution for arm-length stabilization in future gravitational-wave detectors that utilize AlGaAs/GaAs coated mirrors. The authors assert that their scheme is compatible with the A# upgrade and third-generation observatories (such as Cosmic Explorer and the Einstein Telescope).

The significance of the work lies in its ability to maintain the necessary control over arm cavity lengths without the absorption losses that would render current systems ineffective. While the authors note limitations—specifically the low conversion efficiency of the single-pass THG process and the use of a physical switch for the locking handover—they conclude that the demonstrated stability and robustness confirm the viability of the scheme. They suggest that future implementations could improve performance by using resonant cavities for THG enhancement and implementing smoother control loop blending for the transition phase.