🔬 applied physics

Dual mass milligram-scale torsion oscillator for vibration-free optomechanical sensing

This paper presents a dual milligram-mass torsion oscillator that suppresses environmental vibrations by over an order of magnitude to achieve thermally limited torque sensitivity and high-precision gravimetry, enabling vibration-free optomechanical sensing for applications ranging from commercial devices to fundamental physics experiments.

Original authors: Jack Manley, Thomas Bsaibes, Charles A. Condos, William A. Terrano, Dalziel J. Wilson, Jon R. Pratt

Published 2026-07-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jack Manley, Thomas Bsaibes, Charles A. Condos, William A. Terrano, Dalziel J. Wilson, Jon R. Pratt

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 listen to a single, tiny whisper in the middle of a roaring stadium. That is the daily struggle for scientists who build ultra-sensitive machines to measure the universe's tiniest forces. These researchers work in the world of "optomechanics," a fancy term for devices that use light (optics) to measure the movement of tiny mechanical parts. Think of these devices as microscopic seesaws or spinning tops that are so light and delicate that even the heat of the room or a distant truck rumbling on a highway can make them wobble.

The big question they are trying to solve is: How do we build a sensor small enough to fit on a computer chip, but sensitive enough to detect things like the pull of gravity from a tiny rock or the push of a single beam of light? The problem is that the real world is messy. Our buildings shake, the ground vibrates, and our labs are full of noise. If a sensor is too sensitive to these vibrations, it can't hear the "whisper" of the physics it's trying to study. This is especially true for experiments trying to measure gravity at very short distances, where the signal is incredibly weak and easily drowned out by the "static" of the environment. Scientists have been trying to build these tiny sensors for years, but they often hit a wall where the vibrations from the lab itself are louder than the signals they want to find.

In this paper, a team of researchers from the National Institute of Standards and Technology and several universities introduces a clever new way to build a tiny, vibration-proof sensor. They created a "dual mass" torsion oscillator, which is essentially a microscopic spinning top made of a silicon ribbon with two tiny weights (test masses) attached to opposite ends. You can picture it like a miniature dumbbell spinning on a very thin, stretched rubber band.

The genius of their design lies in how they make it move. Usually, if you shake a table, everything on it wobbles together. But this device has a special "antisymmetric" mode. Imagine two kids on a seesaw: if they both push down at the same time, the whole seesaw bounces up and down (this is the "symmetric" mode, which is very sensitive to vibrations). But if one kid pushes down while the other pulls up, the seesaw twists in the middle without moving up or down as a whole (this is the "antisymmetric" mode). The researchers found that by focusing on this twisting motion, the device becomes almost immune to the shaking of the lab. While the regular shaking mode was easily disturbed by the environment, this special twisting mode suppressed vibrations by more than ten times, allowing the sensor to reach its theoretical limit of sensitivity.

The team put their new device to the test in several ways. First, they checked how quiet it was. They found that in this special twisting mode, the sensor was so quiet that it was only limited by the natural thermal jiggling of atoms (thermomechanical noise), reaching a torque sensitivity of $10^{-18}$ Nm/ $\sqrt{\text{Hz}}$ . To prove it could actually "hear" a force, they used a laser beam to push on the device. By modulating the laser, they applied a tiny, rhythmic push (radiation pressure) equivalent to a torque of $10^{-16}$ Nm and successfully detected it over a 30 Hz bandwidth.

They also tested how well the device could act as a gravity sensor. Because the weights are slightly off-center, the device acts like a pendulum that changes its spinning speed depending on the strength of gravity. They tilted the whole experiment to simulate a change in gravity and measured how the spinning speed shifted. They found that by using the vibration-resistant twisting mode, they could measure changes in gravity with a precision of $10^{-6}g_0$ (where $g_0$ is standard Earth gravity) in just 30 seconds, all while the device was only wobbling by a tiny 100 $\mu$ rad.

The paper explicitly argues against the idea that you need massive, heavy equipment or underground caves to get these measurements. Instead, they show that by engineering the device to have two masses that move in opposite directions, you can cancel out the noise from the environment right on the chip itself. They also note that while the "symmetric" mode (the one that bounces up and down) is very noisy and 40 times worse than the thermal limit, the "antisymmetric" mode is the key to success. This work suggests that we don't need to wait for perfect, vibration-free labs to do high-precision physics; we can build sensors that are smart enough to ignore the noise around them, opening the door for compact, portable devices that could one day be used for everything from finding dark matter to mapping underground resources.

Technical Summary: Dual Mass Milligram-Scale Torsion Oscillator for Vibration-Free Optomechanical Sensing

Problem Statement
Chip-scale optomechanical devices hold promise for miniaturizing inertial sensors and fundamental physics experiments, particularly in measuring gravity at short distances (below 100 µm) to probe physics beyond the Standard Model. However, achieving theoretical precision limits is frequently hindered by extraneous environmental vibrations. In traditional torsion balances and mass-loaded silicon nitride (Si $_3$ N $_4$ ) nanoribbon oscillators, incoherent excitation from anthropogenic or seismic activity often overshadows weak signals. Specifically, while mass-loaded oscillators offer ultralow loss through dissipation dilution, their offset center of mass makes them highly sensitive to lateral acceleration noise. Existing mitigation strategies, such as vibration isolation systems or underground laboratories, are often incompatible with compact, field-deployable sensors. Furthermore, using feedback actuation to drive large coherent motion to improve signal-to-noise ratios introduces nonlinear effects.

Methodology
The authors introduce a dual-mass torsion oscillator designed to isolate a specific degree of freedom from environmental vibrations while maintaining coupling to the signal of interest.

Device Fabrication: The device is fabricated from a 400-µm-thick silicon wafer double-coated with high-stress (approx. 1 GPa) Si $_3$ N $_4$ . Two milligram-scale (0.93 mg) silicon test masses are etched to remain suspended on a strained nanoribbon.
Operational Modes: The system supports two torsional modes: a symmetric mode (+, 44 Hz) and an antisymmetric mode (−, 58 Hz). In the antisymmetric mode, the two masses oscillate in opposition, resulting in no net translational motion of the center of mass. This geometry theoretically decouples the mode from the translational accelerations of the device frame.
Experimental Setup: The chip is housed in a vacuum chamber ( $10^{-9}$ Torr) atop a holder actuated by piezoelectric transducers (PZTs). Angular displacement is measured via an optical lever using a 1.55 µm laser and a quadrant photodetector.
Characterization Techniques:
- Quality Factor: Measured via ringdown experiments where one mode is excited while the other is damped to prevent cross-contamination.
- Noise Floor: Analyzed via power spectral density (PSD) of angular displacement, distinguishing between thermomechanical noise, optical readout noise, and laboratory vibrations.
- Torque Sensitivity: Calibrated using oscillating optical radiation pressure applied by the detection laser, with a reference tone used for common noise cancellation.
- Vibration Sensitivity: Characterized by driving the sample holder with out-of-phase PZTs to induce controlled tilts and horizontal accelerations.
- Gravimetry: Evaluated by measuring the parametric shift in resonance frequency due to changes in effective gravity (simulated by tilting the platform).

Key Contributions

Dual-Mass Design: The paper presents a novel implementation of a dual-test-mass system on a single Si $_3$ N $_4$ nanoribbon, specifically engineered to exploit the antisymmetric torsion mode.
Vibration Isolation Mechanism: The authors demonstrate that the antisymmetric mode effectively suppresses sensitivity to translational vibrations by decoupling the center of mass motion from the device frame, a significant improvement over single-mass oscillators.
Thermally Limited Sensitivity: The device achieves a torque sensitivity limited only by thermomechanical noise, bypassing the vibration noise floor that typically plagues such sensors.

Results

Quality Factors: Both the symmetric (44 Hz) and antisymmetric (58 Hz) modes exhibit exceptionally high quality factors of $Q \approx 3.5 \times 10^6$ , indicating ultralow mechanical loss.
Torque Sensitivity: The antisymmetric mode achieves a thermally limited torque sensitivity of $10^{-18}$ Nm/ $\sqrt{\text{Hz}}$ . In contrast, the symmetric mode is dominated by environmental vibrations, resulting in a noise floor approximately 40 times higher than the thermomechanical limit.
Vibration Suppression: Driven response measurements show the antisymmetric mode is 13 times less sensitive to translational vibrations than the symmetric mode. The residual sensitivity is attributed to fabrication imperfections (mass mismatch of ~8%) or twist components in the excitation.
Radiation Pressure Detection: The system successfully detected an optical radiation pressure torque of $10^{-16}$ Nm over a 30 Hz bandwidth, validating the calibration and sensing capability.
Gravimetry Performance: As a frequency-based gravimeter, the device achieved a precision of $10^{-6}g_0$ (where $g_0 = 9.8$ m s $^{-2}$ ) in 30 seconds with an oscillation amplitude of only 100 µrad (0.1 mrad). This represents an order-of-magnitude improvement over previous Si $_3$ N $_4$ -based attempts that relied on large-amplitude driving.

Significance
The paper claims that this work demonstrates a viable technique for overcoming vibration noise in optomechanical sensing without requiring bulky isolation systems or large-amplitude nonlinear driving. By leveraging the antisymmetric mode of a dual-mass oscillator, the device bridges the gap between micro-scale and gram-scale torsion balances, enabling high-precision measurements in compact form factors. The authors highlight the broad implications for applications ranging from commercial inertial sensing to fundamental physics experiments, such as short-range gravity tests and dark matter searches, where vibration noise has historically been a limiting factor. The results suggest that further improvements, such as cryogenic cooling or dual-readout schemes, could further enhance performance.

Technical Summary: Dual Mass Milligram-Scale Torsion Oscillator for Vibration-Free Optomechanical Sensing

More like this