Mechanical loss and stability analysis of NEXCERA in ultra-stable optical cavities

TL;DR

This work first measures the room-temperature mechanical loss of NEXCERA N117B using a gentle nodal-suspension approach, finding a remarkably low minimum loss of $φ = 1.89\times10^{-5}$ across several resonances. Using fluctuation–dissipation theory, the authors translate this loss into Brownian thermal-noise predictions for a 30 cm spacer and compare against ULE and Zerodur, showing NEXCERA offers lower spacer-induced noise due to its high stiffness ($E\approx140\ \text{GPa}$) with comparable loss, and mapping out the impact of mirror-substrate choices. Drift-rate analysis indicates NEXCERA exhibits substantially lower long-term drift than Zerodur, enhancing long-term stability. They further demonstrate that substituting fused-silica or silicon mirror substrates can suppress total cavity noise by over an order of magnitude, with fused silica providing a practical route and silicon offering the lowest intrinsic noise albeit with thermoelastic trade-offs. Overall, the results position NEXCERA as a strong room-temperature spacer candidate for ultra-stable cavities, combining low thermal-noise, low drift, and high stiffness, with significant implications for precision metrology and fundamental-physics tests.

Abstract

NEXCERA has emerged as a ceramic-based material for spacers in ultra-stable optical cavities, with a coefficient of thermal expansion that crosses zero near room temperature. In such cavities, frequency stability is ultimately limited by Brownian thermal noise in the cavity components. A key parameter in this context is the mechanical loss, which has remained unknown for NEXCERA. In this work, we investigate the mechanical loss of NEXCERA N117B at room temperature for various resonances using the gentle nodal suspension technique. We measure a promising minimum mechanical loss of $φ= 1.89\times 10^{-5}$, indicating the suitability of NEXCERA for low-noise optical cavities. Using this value, we calculate the thermal noise of a cavity with a NEXCERA spacer and compare its performance to established materials such as ULE and Zerodur, taking into account different mirror substrate options. Our analysis shows that NEXCERA is a strong candidate for ultra-stable cavities due to its low thermal noise. Combined with its previously reported low linear drift, it offers a highly attractive option for long-term stable optical frequency references.

Figures (7)

• Figure 1: Side view of the nodal suspension setup for mechanical loss measurements. The disk is suspended on a 4.5mm steel sphere due to gravity. A comb-shaped electrode, used to excite the disk via a high-voltage amplifier, is positioned close to the disk using an XYZ stage. For optical readout, a mirror is mounted at an angle of 45° to reflect the incident and reflected laser beams out of the vacuum chamber.
• Figure 2: Frequency scan of the mechanical resonances from 400 to 50 with the corresponding mode pattern simulated with COMSOL. Four different mechanical modes can be observed, which are used for further mechanical loss investigations. The measured rms voltages of the frequency scan are normalized to the maximum rms voltage recorded by the quadrant photodetector in this measurement.
• Figure 3: Exemplary frequency scan of a resonance near 3178.5. The vertical axis shows the normalized oscillation amplitude measured as rms voltages at the quadrant photodetector. The resonance frequency $f_0$ and linewidth $\Delta f$ are used to determine the mechanical loss $\phi$. The inset qualitatively depicts the mode shape, with blue areas indicating nodes (low oscillation amplitude) and red areas indicating antinodes (high oscillation amplitude).
• Figure 4: Mechanical loss results for all three investigated disks and four different resonances. The lowest measured mechanical loss is $\phi=1.894+-0.041e-5$ at $f_0=3.2\kHz$.
• Figure 5: Drift rate for the most common materials used for the ultra-stable optical resonator spacers. Zerodur Keupp2005, ULE Keupp2005 and NEXCERA N117B Ito2017 operate at room temperature, whereas the single-crystal silicon Oelker2019 and sapphire Wiens2016 are used in a cryogenic environment. It is important to note that ULE drift rates vary significantly, with reported values ranging from 2.041e-16s Alnis_2008 to 4.572e-17s Keupp2005. This variation may be due to differences in measurement duration, environmental conditions, and specific glass samples.