🔬 atomic physics

Single-atom detection with a quantum-controlled mechanical oscillator

This paper demonstrates the detection of individual xenon atoms colliding with a levitated nanosphere in its quantum ground state, achieving 96% confidence in observing momentum kicks below 50 keV/c and distinguishing them from thermal background noise.

Original authors: Maxime Perdriat, Maciej Dziewiecki, Josef-Anton Agner, Massimiliano Rossi, Frederic Merkt, Martin Frimmer, Lukas Novotny

Published 2026-08-19

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Maxime Perdriat, Maciej Dziewiecki, Josef-Anton Agner, Massimiliano Rossi, Frederic Merkt, Martin Frimmer, Lukas Novotny

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

In the quiet corners of modern physics, scientists are learning to listen to the faintest whispers of motion. For decades, the dream has been to isolate a tiny object so completely from its surroundings that it behaves not like a billiard ball bouncing off walls, but like a single wave of probability, governed by the strange rules of quantum mechanics. To achieve this, researchers trap microscopic spheres of glass in a vacuum using focused beams of light, creating a cradle where the object can float without touching anything. By cooling these spheres to their lowest possible energy state, the chaotic jiggling caused by heat is silenced, leaving only the fundamental, unavoidable trembling of the universe itself. This extreme sensitivity turns the floating sphere into a detector capable of feeling forces so small they were previously thought impossible to measure, opening a window into the behavior of individual atoms and the invisible interactions that shape our world.

In a recent experiment, a team of researchers at ETH Zurich took this concept a step further, using a levitated glass bead to catch and count individual atoms of xenon. They did not simply wait for atoms to drift by; instead, they engineered a precise, time-gated beam of gas, shooting a stream of xenon atoms at the floating sphere with the accuracy of a sniper. The sphere, a silica particle only one hundred nanometers wide, was held in a vacuum chamber by an optical tweezer—a tight focus of laser light that acts as an invisible trap. The sphere was cooled so effectively that its motion was reduced to its quantum ground state, meaning it was as still as physics allows. When a single xenon atom from the directed beam struck the sphere, it delivered a tiny, sharp kick, transferring a minuscule amount of momentum. The researchers measured the resulting wobble of the sphere with such precision that they could distinguish the impact of a single atom from the background noise of the environment.

The team achieved this by creating a controlled stream of xenon gas. They used a pulsed valve to release the gas in short bursts, allowing the atoms to expand rapidly into a vacuum and form a beam moving at a steady speed of about 310 meters per second. This beam was then directed through a tube connecting two vacuum chambers, aimed directly at the floating sphere. To ensure they were seeing what they thought they were seeing, the researchers compared the results of the directed beam against a control scenario. When they blocked the beam, the number of detected impacts dropped to a low, steady background level caused by random thermal atoms drifting in the chamber. However, when the beam was open, the rate of impacts spiked dramatically at a specific time, matching the travel time of the atoms from the valve to the sphere. This timing confirmed that the detected events were indeed caused by the directed stream of atoms, not random background noise.

The most significant finding was the ability to detect momentum transfers as small as 50 kilo-electron volts per speed of light, a value far below what was previously thought resolvable for such systems. By analyzing the direction of the kicks, the researchers could also tell the difference between the atoms coming from their beam and the random thermal atoms. The atoms from the beam mostly pushed the sphere in one direction, creating an asymmetric pattern of impacts, while the background thermal atoms pushed the sphere equally in all directions. This directional sensitivity allowed the team to identify single-atom collisions with a confidence level of 96%. They observed that the sphere's motion after a collision matched the theoretical prediction for a damped oscillator, confirming that the signal was a genuine physical response to an atomic impact rather than a measurement error.

This work represents a crucial step toward exploring how atoms interact with levitated systems over very short distances. By proving that a single atom can be detected and its momentum measured with high confidence, the experiment demonstrates that these levitated platforms can serve as powerful tools for sensing impulsive forces in regimes that have never been explored before. The researchers suggest that with further improvements, such as using quantum techniques to reduce noise even more, these sensors could detect even lighter atoms and molecules. This capability could eventually lead to new ways of measuring pressure at the atomic level or probing the fundamental forces between matter and light, turning the floating glass bead into a microscope for the invisible world of single-atom collisions.

Technical Summary: Single-atom detection with a quantum-controlled mechanical oscillator

Problem and Motivation
Levitated nanospheres have emerged as a promising platform for sensing static, oscillating, and impulsive forces due to their low mass, high isolation, and controllability under ultra-high vacuum. While their high force sensitivity is being explored for fundamental physics (e.g., macroscopic quantum state preparation, dark matter searches) and technological applications (e.g., field, pressure, and mass sensing), a specific challenge remains in the regime of impulsive force sensing. According to the standard quantum limit (SQL) for impulsive force detection, the minimum resolvable momentum for a nanoparticle with a mass of 1 fg and a trap frequency of 60 kHz is approximately 11.8 keV/c. This value is comparable to the momentum of a thermal gas molecule but significantly smaller than that of an $\alpha$ particle. Consequently, while theory suggests that levitated nanoparticles should be capable of resolving collisions with single atoms, experimental demonstration of this capability—specifically distinguishing directional single-atom impacts from thermal background noise—has been a primary goal.

Methodology
The authors constructed an experimental setup comprising two vacuum chambers: a trapping chamber and an expansion chamber.

Levitated System: A silica nanosphere (100 nm diameter) is trapped in an optical tweezer (1550 nm, 0.7 W) within the trapping chamber at a pressure of $9.0 \times 10^{-9}$ mbar. The particle's center-of-mass motion is measured via homodyne detection. The transverse modes are cooled via parametric feedback, while the axial ( $z$ ) mode is cooled to a phonon occupation of $n_{fb} = 0.80 \pm 0.30$ using cold damping, bringing the system close to its motional quantum ground state.
Atomic Beam Generation: In the expansion chamber, a pulsed supersonic beam of Xenon (Xe) atoms is generated. Xe gas is injected at ~1.2 bar through a pulsed valve with a 100 $\mu$ m nozzle, opening for 200 $\mu$ s with a 200 ms repetition period. The pressure differential drives a supersonic expansion, accelerating atoms to a terminal velocity of $\sim 310$ m/s.
Beam Delivery: The atomic beam is spatially filtered by a 100 $\mu$ m skimmer and directed toward the levitated nanosphere through a connecting tube. The beam is incident at a 45° angle relative to the measurement axis ( $z$ ). A gate valve between the chambers allows for the blocking of the beam to distinguish between directional beam collisions and thermal background collisions.
Detection and Analysis: The system utilizes a matched filter approach to identify impulsive events. The theoretical response function of the damped harmonic oscillator is used as a template to cross-correlate with the measured position time-traces. This allows for the extraction of both the collision time and the magnitude of the momentum transfer ( $\Delta p$ ).

Key Contributions and Results

Single-Atom Detection: The experiment successfully detects individual momentum kicks from single Xe atoms. The maximum momentum transfer along the $-z$ direction is estimated at $\sim 217$ keV/c, corresponding to approximately $25$ times the zero-point fluctuation ( $p_{zpf}$ ).
Signal-to-Noise Ratio (SNR): The system achieves an expected SNR of $\sim 16$ for the strongest momentum kicks. The detection threshold was set at $4 p_{zpf}$ .
Time-Gated Measurements: By triggering measurements with the pulsed valve, the authors observed a distinct temporal profile of collision events. A sharp increase in event rate (reaching $>1000$ counts/s) occurs approximately 1.47 ms after the valve opening, corresponding to the time of flight of the supersonic beam. This is followed by an exponential decay attributed to valve "plunger-rebounces" and skimmer clogging effects.
Background Discrimination: A control experiment involving the partial or complete closure of the gate valve demonstrated that the flat background event rate of $\sim 40$ counts/s observed in the absence of the beam (or after the beam pulse) is attributable to collisions with thermal Xe atoms. When the chamber was purged with nitrogen, the rate dropped to the detector's "dark count rate" of 3.6 counts/s.
Directionality: The distribution of momentum kicks reveals a clear asymmetry for the supersonic beam events (1.5–1.7 ms window), with a dominance of forward-scattering events (negative $z$ -displacement). In contrast, thermal collisions (5.0–9.0 ms window) exhibit a symmetric distribution of forward and backward scattering.
Sensitivity Metric: Based on the event rate analysis, the authors demonstrate the ability to measure collisions with momentum transfers below 50 keV/c with a confidence level of 96%.

Significance
The paper claims that these experiments constitute a first step toward exploring distance-dependent short-range interactions between atoms and levitated systems. Furthermore, the results validate the use of levitated platforms for impulsive force sensing in previously unexplored parameter regimes. The demonstrated sensitivity opens the door for collision-resolved pressure sensing and the study of particles in regimes not previously accessible. The authors note that future enhancements, such as the implementation of squeezing and coherent amplification, could further improve sensitivity to detect even lighter atoms and molecules. Additionally, the work suggests potential applications in neutral-atom interferometry, where the nanosphere could serve as a detector without the need for ionization or preselection of metastable species. The study focuses on center-of-mass degrees of freedom, leaving the exploration of internal atomic states and nonclassical state generation as future directions.

Technical Summary: Single-atom detection with a quantum-controlled mechanical oscillator

More like this