Study of few-electron backgrounds in the LUX-ZEPLIN detector

This paper characterizes delayed electron backgrounds in the LUX-ZEPLIN detector and demonstrates that spontaneous grid electron emission can be efficiently rejected using coincident photon tagging, thereby enhancing the experiment's sensitivity to low-energy dark matter signals.

The Gist

The Great Dark Matter Hunt: Cleaning Up the Noise in the LZ Detector

Imagine you are trying to hear a single, tiny whisper from a ghost (Dark Matter) in a room that is already filled with the sound of a thousand buzzing bees. That is the challenge facing the LUX-ZEPLIN (LZ) experiment.

LZ is a massive, ultra-sensitive detector buried deep underground in a gold mine in South Dakota. It is filled with 7 tons of super-cold liquid xenon, acting like a giant, invisible net to catch dark matter particles. When a dark matter particle bumps into a xenon atom, it should create a tiny flash of light and a tiny spark of electricity (an electron).

However, the scientists found a problem: their "whisper" was being drowned out by a lot of static noise. This paper is about figuring out exactly what that noise is and how to filter it out so they can finally hear the ghost.

Here is the story of how they cleaned up the signal, broken down into three main parts:

---

1. The "Ghostly Echo" (Delayed Electrons)

The Problem:
Sometimes, after a big event happens in the detector (like a particle bumping into something), the detector keeps spitting out tiny, random electrons for seconds afterward. It's like knocking over a glass of water; the main splash happens, but then you hear the drip... drip... drip for a long time. These "drips" are called delayed electrons.

The Investigation:
The scientists wanted to know: Where are these drips coming from?

• Theory A: Maybe they are stuck under the surface of the liquid, like bubbles trapped under a lid, waiting to pop out.
• Theory B: Maybe the liquid itself is dirty. Imagine the liquid xenon as a clean swimming pool. If there are tiny bits of dust (impurities) in the water, they might grab onto the electrons as they swim through, hold them for a moment, and then let them go later.

The Discovery:
The scientists did a clever experiment. They looked at how the "drip rate" changed based on how deep the original event was and how "dirty" the liquid was.

• The Analogy: Imagine running through a forest. If the forest is full of sticky burrs (impurities), you get stuck more often. The deeper you run, the more likely you are to get stuck and then get unstuck later.
• The Result: They found that the "drips" were definitely coming from the impurities in the liquid. The electrons were getting caught by these invisible "sticky burrs" and released slowly. This confirmed that the liquid needs to be as pure as possible to stop this echo.

---

2. The "Sparking Wire" (Grid Emission)

The Problem:
Inside the detector, there are metal grids (like a fence) that hold a high voltage to pull the electrons up. Sometimes, these wires start acting like faulty Christmas tree lights. They spontaneously shoot out tiny sparks of electrons, even when nothing else is happening. This is called grid emission.

The Investigation:
The scientists noticed something strange: whenever these wires shot out an electron, they also seemed to shoot out a tiny flash of light (a photon) at the exact same time. It was like a firework that made a pop (electron) and a sparkle (photon) simultaneously.

The Discovery:
This was a game-changer!

• The Analogy: Imagine you are trying to find a specific person in a crowd. Usually, you just look for the person. But if you realize that every time this person sneezes, they also drop a red hat, you can just look for the red hats to find them.
• The Solution: The scientists realized they could use the light (photon) as a "tag" to identify the electron. If they see an electron and a flash of light at the same time, they know, "Ah, this is just a faulty wire spark, not a dark matter signal!" They can then throw that data away.

---

3. The "Noise Cancellation" Headphones

The Goal:
The ultimate goal of the LZ experiment is to find low-mass dark matter. These are very light particles that only create tiny signals (sometimes just 1 or 2 electrons).

The Challenge:
The "faulty wire sparks" (grid emission) create exactly these tiny signals. Before this study, the scientists had to be very aggressive and throw away huge chunks of data just to be safe, which meant they might have thrown away real dark matter signals too.

The New Strategy:
By using the "light tag" method described above, they can now be much smarter.

• Old Way: "If the room is too noisy, shut down the whole experiment."
• New Way: "If we hear a noise and see a red hat, we know it's a wire spark. Ignore it. If we hear a noise without a red hat, keep listening!"

The Result:
They tested this new method on real data. It successfully removed over 90% of the fake "wire spark" noise while keeping almost all of the real signals. It's like putting on high-tech noise-canceling headphones that only cancel out the specific sound of the buzzing bees, leaving the ghost's whisper crystal clear.

---

The Bottom Line

This paper is a victory for the LZ team. They didn't just find the noise; they figured out why it was there (sticky impurities in the liquid) and how to spot it (looking for the accompanying flash of light).

By cleaning up the detector and learning to ignore the "static," they have made the LZ experiment much more sensitive. This brings humanity one step closer to finally catching a dark matter particle and understanding what makes up the invisible 85% of our universe.

Technical Summary

1. Problem Statement

The LUX-ZEPLIN (LZ) experiment is designed to detect Weakly Interacting Massive Particles (WIMPs) via liquid xenon (LXe) time projection chambers (TPCs). While highly sensitive to heavy dark matter, LZ aims to extend its reach to sub-GeV/c² dark matter by detecting low-energy ionization signals (single or few electrons) without requiring a coincident scintillation signal (S1).

The primary challenge in this "ionization-only" search regime is the presence of instrumental backgrounds that mimic dark matter signals. These backgrounds manifest as:

• Delayed electron emission: Spurious single-electron (SE) or multi-electron (ME) pulses appearing milliseconds to seconds after a primary energy deposition.
• Spontaneous grid emission: Electrons emitted from high-voltage electrode grids due to defects, debris, or field emission, often accompanied by photon emission.

These backgrounds obscure the low-energy signal region, limiting the experiment's sensitivity to light dark matter candidates.

2. Methodology

The authors analyzed data from the LZ detector under various conditions, including commissioning runs, grid tuning runs, and the first science run (WS2022). The methodology involved:

• Event Selection:
  • Progenitor Definition: Primary events (progenitors) were selected as S2 pulses with $\ge$ 200 extracted electrons ($e_R$) occurring in the active TPC volume.
  • Background Isolation: Small electron pulses following these progenitors were tracked. Pulses larger than 100 $e_R$ ("vetoing pulses") were used to terminate the tracking window.
  • Spatial Correlation: Pulses were categorized as position-correlated (within 10 cm of the progenitor) or position-uncorrelated (20–30 cm away) to distinguish delayed emission from random noise.
• Normalization: Electron rates were normalized by progenitor size and drift parameters (electron lifetime, drift time) to isolate the effects of liquid impurities and electric fields.
• Grid Emission Study: An "extraction region-only" dataset was used (where the cathode is inactive) to isolate electron emission specifically from the gate grid.
• Coincidence Analysis: The study searched for temporal correlations between grid-emitted electrons and Single Photo-Electrons (SPEs) to identify a tagging mechanism.

3. Key Contributions and Results

A. Characterization of Delayed Electron Backgrounds

• Origin: The study confirms that delayed electron emission is primarily caused by drift electrons captured by impurities in the liquid xenon bulk, rather than electrons trapped under the liquid surface.
  • Evidence: The rate of delayed electrons scales with the number of electrons lost to impurities ($e_L$) and increases with longer drift times (more opportunity for capture).
  • Normalization by electrons lost to impurities ($e_L$) collapses the rate curves across different purity levels, confirming this mechanism.
• Temporal Behavior: The rate of position-correlated delayed electrons follows a power-law decay ($Rate \propto \Delta t^\beta$) with an exponent $\beta \approx -1.13$.
  • This exponent is consistent across different drift fields (110–240 V/cm) and extraction fields (3.4–4.3 kV/cm).
  • The non-integer power law suggests a distribution of impurity species with varying electron affinities, rather than a single trapping mechanism.
• Magnitude: For a typical drift time of 0.85 ms, approximately 1.1% of the progenitor area results in delayed SEs between 3 ms and 1 s.

B. Grid Electron Emission and Photon Tagging

• Observation: Spontaneous electron emission from grid "hot spots" (defects/debris) is a significant source of few-electron backgrounds.
• Photon Correlation: A critical finding is that grid-emitted electrons are frequently accompanied by photon emission.
  • Approximately 8.9% of all S2 pulses from grid emitters have a coincident SPE preceding them within the drift time window (~2.2 µs).
  • Accounting for photon detection efficiency, the authors estimate that ~64% of gate-emitted electron pulses have an associated photon signal.
• Tagging Efficiency: By using a photon coincidence tag (looking for an SPE within the gate drift window), the authors demonstrated a method to identify and reject grid backgrounds.
  • In a test on WS2022 data, this cut suppressed the rate of 2–4 electron pulses (dominated by hot spots) by over an order of magnitude.
  • The cut rejected SEs by a factor of >2 while maintaining a signal acceptance of ~93–98% for genuine dark matter candidate events (S2-only events >10 electrons).

4. Significance

• Improved Sensitivity for Low-Mass Dark Matter: This work provides the necessary tools to suppress the dominant instrumental backgrounds in the few-electron regime. By characterizing the delayed emission mechanism and establishing a photon-tagging cut for grid emission, LZ can significantly lower its energy threshold.
• Validation of Background Models: The confirmation that bulk impurities (not surface traps) drive delayed emission refines the simulation and mitigation strategies for future tonne-scale LXe experiments (e.g., XENONnT, DARWIN).
• New Rejection Strategy: The demonstration that grid emission can be tagged via coincident photons offers a powerful, data-driven background rejection technique that is more efficient than simple rate-based cuts, allowing for the retention of more signal events while removing spurious grid noise.

In summary, this paper establishes a robust framework for understanding and mitigating few-electron backgrounds in LZ, directly enabling more sensitive searches for sub-GeV dark matter and validating new techniques for future dark matter experiments.