First Dark Photon Search Results from the Dandelion Experiment

The Dandelion experiment presents its first results searching for 1 meV dark photon dark matter using a spherical mirror and a 221-KID array, finding no signal and establishing the first millimeter-wavelength upper limits on dark photon kinetic mixing between 0.6 and 1.4 meV after successfully mitigating background noise through de-correlation analysis.

The Gist

Imagine the universe is filled with a mysterious, invisible fog called Dark Matter. For decades, scientists have been trying to figure out what this fog is made of. One popular theory suggests it's made of tiny, ghostly particles called Dark Photons. These particles are like invisible cousins to the light we see every day; they don't interact with normal light, so they pass right through us and our walls without us noticing.

The Dandelion Experiment is a new, high-tech "net" designed to catch these invisible ghosts. Here is how they did it, explained simply:

1. The Magic Mirror (The Net)

The scientists built a giant, shiny aluminum mirror (about the size of a large dining table). They placed it in a super-cold box (colder than outer space!) to stop it from making its own "noise."

The idea is this: If a Dark Photon from the galactic fog hits this mirror, it might magically transform into a regular photon (a tiny packet of light). Because Dark Photons are very light, the light they turn into is a specific color: millimeter waves (invisible to our eyes, but detectable by special sensors).

2. The "Dandelion" Effect (The Direction)

Here is the clever part. The experiment is named "Dandelion" because of how the signal moves.

Imagine you are holding a dandelion seed head and spinning around. The seeds fly out in a specific direction relative to your spin. Similarly, as the Earth rotates, the direction of the "wind" of Dark Photons hitting the mirror changes.

If Dark Photons exist, the spot of light they create on the detector shouldn't stay still. It should trace a predictable, moving path across the camera sensor over the course of a day, just like a shadow moving across a wall. This moving path is the experiment's "signature."

3. The Noise Problem (The Static)

The biggest challenge was that the detector is incredibly sensitive. It's like trying to hear a whisper in a room where a jet engine is running.

• The Whisper: The tiny signal from a Dark Photon.
• The Jet Engine: Heat from the room, the mirror itself, and stray light bouncing around. This creates a massive amount of "background noise" that drowns out the signal.

4. The Detective Work (Separating Signal from Noise)

To solve this, the scientists used a clever trick called Principal Component Analysis (PCA). Think of it like this:

Imagine you are in a crowded room where everyone is talking.

• The Background Noise: Everyone is chatting at once. This noise is the same for everyone in the room; it's a constant hum.
• The Signal: One specific person is walking across the room, whispering a secret to people as they pass.

The scientists looked at all the sensors (the "people" in the room). They noticed that the sensors not on the walking path were all hearing the same "hum" (the background noise). They used a mathematical tool to map out exactly what that "hum" sounded like.

Then, they subtracted that "hum" from the sensors that were on the walking path. If the Dark Photon whisper was there, it would be left behind after the noise was removed.

5. The Result: A Clean Slate

After running the experiment for about 25 hours and analyzing the data with this "noise-canceling" technique, the result was: Silence.

They didn't hear the whisper. The signal they found was consistent with zero.

What does this mean?
It doesn't mean Dark Photons don't exist. It means that if they do exist, they are even more elusive than the scientists hoped. The experiment set a new "speed limit" for how strong the connection between Dark Photons and normal light can be. They ruled out a specific range of possibilities (masses between 0.6 and 1.4 meV).

The Takeaway

The Dandelion experiment proved that its "net" works. It successfully demonstrated a new way to hunt for Dark Matter by looking for a moving shadow in a sea of noise. Even though they didn't catch the fish this time, they proved the boat is seaworthy and gave the scientific community a new, stricter map of where to look next. It's a major step forward in the hunt for the universe's biggest mystery.

Technical Summary

1. Problem Statement

The search for dark matter remains a primary challenge in astroparticle physics. This paper focuses on the Dark Photon (DP) hypothesis, a vector boson of a hidden $U(1)$ symmetry that interacts with the Standard Model photon via kinetic mixing ($\chi$).

• Target: The experiment targets "cold" dark photons in the mass range of 0.6 meV to 1.4 meV (corresponding to frequencies of 150–350 GHz).
• Challenge: Detecting the conversion of these dark photons into standard photons is extremely difficult because the expected signal is a tiny temperature modulation (micro-Kelvin scale) buried within a massive, correlated background noise (thermal emission from room-temperature optics, stray light, and electronic noise).
• Key Differentiator: Unlike omnidirectional searches, the Dandelion experiment exploits the directional signature of dark matter. As the Earth rotates, the galactic dark matter wind creates a predictable, moving signal spot on the detector, distinguishing it from static or isotropic background noise.

2. Methodology

Experimental Setup

• Conversion Mechanism: The experiment uses a 50 cm diameter spherical aluminum mirror (radius of curvature 5 m). Dark photons from the galactic halo convert into standard photons upon hitting the mirror surface.
• Signal Characteristics: The converted photons have an energy equal to the dark photon mass ($E \approx m_X c^2$). Due to the velocity of dark matter in the halo ($v \approx 10^{-3}c$), the emitted photons deviate slightly from the surface normal by an angle $\psi \approx 10^{-3}$ rad, creating a moving spot on the detector.
• Detector Array: The system utilizes the KISS-NIKA camera, containing 221 Kinetic Inductance Detectors (KIDs) cooled to 150 mK. KIDs are superconducting resonators that change resonant frequency when absorbing photons, allowing for precise power measurement.
• Operational Band: The system is optimized for a central frequency of 250 GHz (1.03 meV), with a bandwidth spanning 150–350 GHz.

Data Acquisition Strategy

• Duration: A continuous dataset of 1480 minutes (24.7 hours) was collected on January 11, 2024.
• Modulation Technique: To ensure robustness, the mirror was rapidly switched between two positions every second:
  1. Frontal: Optical axis aligned with the detector center.
  2. Tilted: The mirror was tilted to shift the signal trajectory to a different set of pixels.
  • This allowed for two independent datasets to be concatenated, effectively doubling the integration time and providing cross-validation.

Analysis and Signal Reconstruction

The core of the analysis involves separating the directional signal from the correlated background using a Two-Region Model:

1. Signal Region: Pixels located along the predicted trajectory of the dark photon spot.
2. Background Region: Pixels far from the trajectory, assumed to contain only background noise.

Background Subtraction via PCA:

• The background noise (thermal emission from lenses/mirror) is highly correlated across all pixels.
• The authors applied Principal Component Analysis (PCA) to the "Background-only" region data to identify dominant temporal noise modes ($T^{syst}_i(t)$).
• 10 Principal Components were selected as optimal to model the background without overfitting.
• A linear regression model was fitted to each pixel's time series:
  $$T_k(t) = b_k + \sum a_{k,i} T^{syst}_i(t) + S(x_k, y_k, t)$$
  Where $S$ is the signal component modeled as a 2D Gaussian (approximating the Fresnel diffraction pattern) moving along the trajectory.

3. Key Contributions

• First Directional Search with KIDs: This is the first experiment to use an array of Kinetic Inductance Detectors at millimeter wavelengths to search for dark photons with a directional signature.
• Advanced Noise Mitigation: The paper demonstrates a novel application of PCA-based de-correlation to remove large-scale thermal fluctuations (up to Kelvin variations) that would otherwise swamp the micro-Kelvin signal.
• Validation of the Two-Position Modulation: The study successfully validated that switching the mirror position allows for the separation of instrumental artifacts from true directional signals.

4. Results

• Signal Detection: The analysis found no statistically significant signal. The measured signal amplitude was consistent with zero ($T_{signal} = 0 \pm 6$ mK).
• Upper Limits:
  • An upper limit on the temperature amplitude was set at $\Delta T < 12$ mK (95% confidence level).
  • This was converted into a limit on the incident power and subsequently the kinetic mixing parameter $\chi$.
• Exclusion Curve: The experiment established a new upper limit on the kinetic mixing parameter:
  $$\chi < 8.7 \times 10^{-10}$$
  for dark photon masses between 0.6 meV and 1.4 meV.
• Assumptions: These limits assume dark photons constitute 100% of the local dark matter density ($\rho_{CDM} \approx 0.44 \text{ GeV/cm}^3$).

5. Significance

• New Constraints: These results provide the first constraints on dark photons in the 1 meV mass range using a KID array, filling a gap in the parameter space previously unexplored by this specific technology.
• Feasibility Demonstration: Despite the lack of a detection, the experiment proves the feasibility of using directional detection with millimeter-wave KIDs to search for dark matter. It validates the ability to subtract complex, correlated thermal backgrounds to reach sensitivities required for future discoveries.
• Future Pathway: The methodology established (directional signature + PCA noise subtraction + KID arrays) provides a blueprint for next-generation experiments that could achieve higher sensitivity by increasing integration time or detector count.