⚛️ phenomenology

Fluctuations in atom interferometers as a new tool for dark matter

This paper proposes using the super-binomial variance in atom interferometer count rates as a highly sensitive new signature for detecting dark matter, offering orders-of-magnitude enhanced sensitivity over independent-atom estimates and providing complementary constraints on dark matter interactions with standard model particles.

Original authors: Clara Murgui, Ryan Plestid

Published 2026-03-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Clara Murgui, Ryan Plestid

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

The Big Idea: Listening for a Ghost in the Crowd

Imagine you are trying to detect a ghost. Usually, you'd look for a cold spot or a moving object. But what if the ghost is invisible, doesn't move things, and only whispers? That is the challenge of finding Dark Matter. It makes up most of the universe, but it barely interacts with normal stuff.

Physicists at CERN (Clara Murgui and Ryan Plestid) have proposed a new way to "hear" these whispers. Instead of looking for a single ghost bumping into a single atom, they suggest watching a massive crowd of atoms and listening for a specific kind of "chaos" in their behavior.

The Setup: The Atom Interferometer

Think of an Atom Interferometer as a high-tech race track for atoms.

The Racers: You take a cloud of millions of atoms (like a tiny fog).
The Split: You split this cloud into two paths (Left and Right) and then bring them back together.
The Finish Line: When they rejoin, they create an interference pattern (like ripples in a pond meeting).

Normally, if you run this race, the atoms land in a predictable pattern. If you count how many land on the "Left" side vs. the "Right" side, the results follow a very strict mathematical rule called a Binomial Distribution.

The Analogy: Imagine flipping a coin 1,000 times. You expect roughly 500 heads and 500 tails. Sometimes you get 510, sometimes 490. This natural "wobble" is called shot noise. It's random, but it's predictable. You know exactly how much the number should wiggle.

The New Signal: "Super-Binomial" Noise

The authors propose that if Dark Matter is passing through this cloud of atoms, it won't just bump one atom here or there. Instead, it will create a subtle, invisible "wind" that affects the entire cloud at the same time.

Here is the magic trick:

Normal Noise: If you have a crowd of people flipping coins, and the wind blows randomly on each person's coin, the total number of heads will wiggle a little bit. This is normal.
Dark Matter Noise: If a giant, invisible hand (Dark Matter) gently pushes the entire crowd in the same direction at the exact same moment, the coins will all tilt together.

This creates a correlation. The atoms stop acting like independent individuals and start acting like a synchronized team.

The Result: The "wobble" in the final count becomes much bigger than it should be. The authors call this "Super-Binomial Variance."

Why This is a Game-Changer

The paper highlights two massive advantages of this method:

1. The "Crowd Effect" (N-Enhancement)
In traditional experiments, if you add more atoms, you just get more data, but the signal doesn't get much stronger per atom.

The Paper's Discovery: Because the Dark Matter affects the whole cloud at once, the signal grows with the square of the number of atoms ( $N^2$ ).
The Metaphor: Imagine trying to hear a whisper. If you have one person listening, it's hard. If you have a million people listening, and they all hear the whisper at the exact same time and shout "I heard it!", the signal becomes deafening. This method turns a whisper into a roar by using the collective power of the atom cloud.

2. Immunity to "Fake" Noise
Usually, experiments are ruined by "noise"—vibrations, laser flickers, or temperature changes.

The Problem: These things usually mess up each atom individually.
The Solution: The authors prove that if the noise messes up atoms individually (like a laser flickering), it cannot create this specific "Super-Binomial" pattern. Only a force that connects the atoms (like Dark Matter) can create this specific type of chaos.
The Metaphor: If a crowd of people is jostled by a chaotic mosh pit (laser noise), they move randomly. If a giant invisible wave (Dark Matter) lifts the whole crowd, they move together. The "Super-Binomial" signal is the fingerprint of the wave, not the mosh pit.

What Can We Find?

This new tool is a "Swiss Army Knife" for Dark Matter hunters:

The "Ghostly" Light Stuff: It can find very light Dark Matter particles that move too slowly to trigger standard detectors (which need a hard "thud" to register).
The "Heavy" Stuff: It can find Dark Matter that is so heavy and sticky that it gets stuck in the Earth's atmosphere before reaching underground labs. Standard detectors are blind to this because the Dark Matter never makes it to the door. This method can detect the "traffic jam" of Dark Matter right above our heads.

The Bottom Line

The authors are suggesting we stop looking for Dark Matter as a single "bullet" hitting a target. Instead, we should look for the collective shiver it causes in a cloud of atoms.

By measuring how much the atom counts "wiggle" more than they are supposed to, we can detect Dark Matter that has been hiding in plain sight, invisible to all other methods. It's like realizing that the only way to see a ghost is not to look for it, but to watch how the whole room of people suddenly holds their breath at the same time.

1. Problem Statement

Direct detection of dark matter (DM) faces significant challenges in specific regions of parameter space:

Low Momentum Transfer: Many DM models (very light particles, "dark blobs," feebly coupled mediators, or strongly interacting DM thermalized in the Earth) induce momentum transfers below the thresholds of conventional noble-element detectors ( $\sim$ keV) and even semiconductor devices ( $\sim$ 0.1 eV).
Limitations of Current Atom Interferometry: Existing proposals using atom interferometers typically rely on measuring the mean phase shift (coherent enhancement) or contrast loss (decoherence).
- Phase shifts can vanish for isotropic DM backgrounds or be suppressed by velocity averaging.
- Contrast loss measurements are plagued by "benign" noise sources (laser fluctuations, imperfect path overlap) that mimic the signal, making it difficult to distinguish DM effects from experimental systematic errors.
Strongly Interacting DM (SIDM): For DM with large cross-sections, the particles cannot penetrate the overburden of underground experiments, rendering standard direct detection blind to these candidates.

2. Methodology

The authors propose a novel observable: Super-binomial variance in the atom count rate.

The Null Hypothesis: In the absence of DM correlations, an atom interferometer with $N$ atoms per run behaves as a collection of independent Bernoulli trials. The fraction of atoms detected in the "+" port, $n_+$ , follows a binomial distribution. The variance of this distribution is $\sigma^2_{bin} = \bar{p}(1-\bar{p})/N$ , where $\bar{p}$ is the mean probability.
The Signal Mechanism:
- DM interactions induce correlations between the atoms in the cloud.
- These correlations spoil the independent product state assumption, causing the atoms to deviate from binomial statistics.
- Specifically, the variance of the count rate becomes super-binomial (larger than the binomial prediction).
Key Theoretical Insight:
- The induced variance scales as $N$ (number of atoms per run) relative to the standard shot noise, offering a sensitivity enhancement of orders of magnitude compared to independent-atom estimates.
- Noise Robustness: Crucially, this observable is protected against "P(X)" noise (uncorrelated noise sources like laser phase fluctuations). Such noise alters the individual Bernoulli parameters ( $p_i$ ) but does not induce the specific inter-atom correlations required to generate super-binomial variance. Only genuine many-body correlations (like those induced by DM) can fake the signal.
Experimental Protocol:
1. Run the interferometer $\mathcal{N}$ times with $N$ atoms per shot.
2. Measure the fraction $n_+^{(i)}$ for each shot.
3. Calculate the sample mean $\bar{p}_{obs}$ and sample variance $\sigma^2_{obs}$ .
4. Compare $\sigma^2_{obs}$ against the predicted binomial variance $\bar{p}(1-\bar{p})/N$ .
5. A statistically significant excess in variance indicates DM-induced correlations.

3. Key Contributions

Novel Observable: Introduction of super-binomial fluctuations as a distinct signature for DM, complementary to phase shift and contrast loss measurements.
Enhanced Sensitivity: Demonstration that the signal scales with $N$ (atoms per run) and $1/\bar{p}$ , providing a sensitivity gain of $\sim \sqrt{N}/\bar{p}$ over naive shot-noise limits.
Background Rejection: Identification that this method is immune to uncorrelated experimental noise (laser noise, vibrations) which typically limits contrast-loss searches.
Thermalized DM Detection: A framework to detect DM that has thermalized in the Earth's crust/atmosphere, which is invisible to underground experiments due to shielding.

4. Results and Applications

The authors apply this framework to two specific DM scenarios:

A. Yukawa-Mediated Dark Matter

Model: DM interacts via a long-range Yukawa potential ( $V \propto e^{-m_\phi r}/r$ ) with nucleons.
Findings:
- The method sets competitive constraints on the DM-nucleon coupling $\alpha_{DM}$ .
- For mediator masses $m_\phi \gg p_\chi$ (momentum transfer), the sensitivity scales as $m_\phi^{-4}$ .
- For light mediators ( $m_\phi \ll (\Delta x)^{-1}$ ), the sensitivity becomes flat with respect to $m_\phi$ .
- Comparison: The new observable (solid lines in Fig. 1) complements existing phase-shift searches (dashed lines), covering different regions of the parameter space, particularly where phase shifts are suppressed.

B. Strongly Interacting Dark Matter (SIDM)

Context: DM with cross-sections $\sigma_{\chi N} \gtrsim 10^{-30} \text{ cm}^2$ cannot reach underground detectors.
Findings:
- Earth Thermalization: SIDM thermalizes in the Earth, creating a slow, warm gas population near the surface. Atom interferometers (being threshold-less) can detect this population.
- Traffic Jam Effect: For velocity-dependent cross-sections (e.g., $1/u^4$ ), the thermalized density can be significantly enhanced ( $n_\oplus \gg n_{vir}$ ), leading to large decoherence rates.
- Projected Sensitivity: The paper projects constraints for ground-based experiments (Stanford fountain, AICE at CERN) and space-based missions (AEDGE).
- Impact: This approach fills a critical gap in direct detection, probing cross-sections up to $\sim 10^{-24} \text{ cm}^2$ for GeV-scale DM, a region previously inaccessible to underground experiments and only loosely constrained by astrophysical data.

5. Significance

Paradigm Shift: Moves beyond measuring the mean effect of DM (phase/contrast) to measuring the fluctuations (variance), leveraging quantum statistical properties of the atomic cloud.
Robustness: Offers a "smoking gun" signal that is theoretically distinct from standard experimental noise, reducing false positives.
Broad Reach: Extends the reach of atom interferometry to:
- Very light DM (sub-eV).
- Strongly interacting DM (shielded from underground labs).
- Isotropic DM backgrounds (where phase shifts vanish).
Feasibility: The proposed sensitivity is achievable with near-future technology (e.g., $N \sim 10^8 - 10^{10}$ atoms, phase noise $\sim 1-30$ mrad), making it a viable strategy for upcoming experiments like AICE and AEDGE.

In conclusion, Murgui and Plestid propose that analyzing the variance of atom interferometer counts, rather than just the mean, provides a powerful, noise-resistant, and highly sensitive tool for discovering dark matter across a wide range of masses and interaction strengths.