Search for the production of Higgs-portal scalar bosons in the NuMI beam using the MicroBooNE detector

Using 2.01×10²¹ protons-on-target from the NuMI beam, the MicroBooNE experiment sets the world's strongest limits to date on the mixing angle of Higgs-portal scalar bosons in the 110–155 MeV mass range by searching for their decays into electron-positron pairs.

The Gist

The Big Picture: Hunting for a "Ghost" Particle

Imagine the universe is a giant, bustling city. We know about most of the residents: the people we can see (stars, planets, us) and the ones we can feel but not see (gravity). But physicists suspect there are "ghosts" living in this city—particles that make up Dark Matter. These ghosts are invisible, they don't talk to us, and they barely bump into anything.

For years, scientists have been trying to catch a glimpse of these ghosts. One popular theory suggests these ghosts might be connected to the "Higgs Field" (the invisible field that gives particles their mass) through a secret backdoor. This backdoor is called the Higgs Portal.

This paper is about a team of scientists using a giant, high-tech camera to look for a specific type of ghost particle that might be sneaking through this portal.

The Setup: The "NuMI" Factory and the "MicroBooNE" Camera

To find these ghosts, the scientists needed a factory to make them and a camera to catch them.

1. The Factory (NuMI Beam): Located at Fermilab in Illinois, this is a massive machine that shoots a beam of protons (tiny particles) at a block of graphite. When the protons hit the graphite, they smash into atoms and create a shower of new particles, including Kaons (a type of unstable particle).

   • Analogy: Imagine a high-speed cannon firing ping-pong balls into a wall of bricks. The impact creates a chaotic spray of dust and debris. In this case, the "debris" includes the Kaons we are interested in.

2. The Decay: These Kaons are unstable. As they fly down a long tunnel (the decay pipe), they sometimes break apart. The scientists are looking for a very specific breakup: a Kaon turning into a pion and a Scalar Particle (S).

   • The Mystery: This Scalar Particle (S) is the "ghost." It's invisible at first. But the theory says it shouldn't stay invisible forever. It should eventually decay into an electron and a positron (matter and antimatter twins).

3. The Camera (MicroBooNE): Located 600 meters away from the factory, this is a giant tank filled with Liquid Argon. It's essentially a 3D digital camera that is incredibly sensitive.

   • How it works: When a particle zips through the liquid argon, it knocks electrons off the argon atoms. The camera catches these electrons and builds a 3D picture of the path.
   • The Goal: The scientists are waiting to see a tiny, specific "spark" inside the tank: an electron and a positron appearing out of nowhere, exactly where a ghost particle should have decayed.

The Challenge: Finding a Needle in a Haystack

The problem is that the "ghost" particle is very shy. It interacts so weakly that it rarely shows up. Meanwhile, the camera is constantly being bombarded by other particles (neutrinos) and cosmic rays (particles from space) that look exactly like the ghost particle's signal.

• Analogy: Imagine you are trying to hear a single, quiet whisper in a stadium full of people cheering. The "whisper" is the ghost particle. The "cheering" is the background noise from neutrinos and cosmic rays.

To solve this, the MicroBooNE team used a Boosted Decision Tree (BDT).

• Analogy: Think of the BDT as a super-smart, super-fast security guard. This guard has been trained on millions of computer simulations. It looks at every single event in the camera and asks: "Is this a ghost particle, or is it just noise?" It checks the angle, the energy, and the shape of the spark. If it looks even slightly suspicious, the guard lets it through. If it looks like a normal cheer, the guard ignores it.

The Results: "No Ghosts Found, But We Know Where They Aren't"

The scientists analyzed a massive amount of data (over 200 billion protons hitting the target). They looked for the specific "electron-positron" spark in the liquid argon tank.

The Verdict: They didn't find any ghost particles.

• What does this mean? It doesn't mean the ghosts don't exist. It means that if they do exist, they are even more shy than we thought. They are hiding better than our current "security guard" can catch them.

Because they didn't find them, the scientists set a new limit. They drew a line in the sand and said: "If these ghost particles exist, their connection to the Higgs field (the mixing angle) must be weaker than this number."

• The Achievement: This is the strictest limit ever set for this specific type of ghost particle in the mass range of 110 to 155 MeV (a very specific weight for a particle). They have effectively closed the door on a large chunk of the "hiding spots" where these particles could have been.

Why This Matters

Even though they didn't find the particle, this is a huge victory for science.

1. Ruling Out Possibilities: Science often progresses by saying, "It's not this." By ruling out this specific range of masses and connection strengths, they force theorists to come up with new, better ideas about where the Dark Matter might be hiding.
2. Better Technology: The techniques they used to filter out the noise (the BDTs) and the massive amount of data they processed prove that Liquid Argon detectors are powerful tools for hunting new physics.
3. The Search Continues: The door isn't closed forever; it's just narrower. The next generation of detectors will be even more sensitive, perhaps catching a whisper that this experiment missed.

In summary: The MicroBooNE team built a super-sensitive camera to watch a particle factory, looking for a ghost particle that turns into an electron-positron pair. They didn't see the ghost, but they proved that if it's there, it's incredibly elusive, and they've set the tightest rules yet on how it could possibly behave.

Technical Summary

1. Problem and Motivation

• Physics Context: The Standard Model (SM) does not account for Dark Matter (DM). Many theoretical models propose a "Dark Sector" of particles that interact with SM particles only through new forces or "portals."
• The Higgs Portal: A specific class of models posits that a new scalar singlet particle ($S$) mixes with the SM Higgs boson via a mixing angle $\theta$. This mixing allows $S$ to couple to SM particles, potentially making it a DM candidate or a mediator to a hidden sector.
• The Search Challenge: The mixing angle $\theta$ and the scalar mass $m_S$ are free parameters. Previous experiments have set limits, but a specific mass range around the neutral pion mass ($m_{\pi^0} \approx 135$ MeV) remains challenging due to high backgrounds from $\pi^0$ decays.
• Objective: This paper aims to set the strongest experimental limits to date on the mixing angle $\theta$ for scalar masses in the range 110 MeV < $m_S$ < 155 MeV, utilizing the unique capabilities of the MicroBooNE detector.

2. Methodology

Experimental Setup

• Detector: The MicroBooNE Liquid Argon Time Projection Chamber (LArTPC) at Fermilab. It features high-resolution imaging and particle identification capabilities.
• Beam Source: The Neutrinos from the Main Injector (NuMI) beam, consisting of 120 GeV protons striking a graphite target to produce hadrons (primarily pions and kaons).
• Data Exposure: The analysis uses a total exposure of $2.01 \times 10^{21}$ protons-on-target (POT), combining data from two beam configurations:
  • FHC (Forward Horn Current): Focuses positively charged hadrons.
  • RHC (Reverse Horn Current): Focuses negatively charged hadrons.
• Signal Production Mechanism: The search targets scalar particles ($S$) produced via the decay of Kaons ($K$) in the NuMI beamline:
  • KDIF (Kaon Decay In Flight): Kaons decaying in the 675m decay pipe.
  • KDAR (Kaon Decay At Rest): Kaons stopping and decaying in the NuMI target or the downstream hadron absorber.
• Signal Decay: The analysis searches for the decay channel $S \to e^+e^-$.
  • Kinematic Constraint: The search is restricted to $100 < m_S < 211$ MeV. The paper focuses on $110-155$ MeV because the $S \to \mu^+\mu^-$ channel dominates for $m_S > 2m_\mu$, and lower masses are excluded by previous experiments (E949, NA62).

Event Reconstruction and Selection

• Topology: The $S \to e^+e^-$ decay produces two electromagnetic showers. However, due to the small opening angle of the decay products, simulations show that often only one reconstructed shower is observed (overlapping showers), while others show two distinct showers.
• Sample Division: Events are split into two samples:
  1. One reconstructed shower ($n_{shower}=1$).
  2. Two reconstructed showers ($n_{shower}=2$).
• Background Rejection:
  • Cosmic Rays: Mitigated using the Cosmic Ray Tagger (CRT) and a "cosmic-like" score from the Pandora pattern recognition software.
  • Neutrino Interactions: The primary background comes from neutrino interactions mimicking $e^+e^-$ topologies (e.g., neutral current $\pi^0$ production).
• Machine Learning (BDTs):
  • Eight separate Boosted Decision Trees (BDTs) were trained for each combination of beam polarity (FHC/RHC), shower multiplicity (1 or 2), and production topology (KDIF or KDAR).
  • Input Variables: Key discriminators include reconstructed energy in the TPC collection plane, momentum fractions in $x$ and $z$ directions, and interaction direction.
  • Training: BDTs were trained on Monte Carlo simulations of signal and background, overlaid with real cosmic ray data to ensure realistic noise modeling.

Statistical Analysis

• Method: The CLs (Confidence Level) semi-frequentist method was used, implemented via the pyhf framework.
• Procedure: The six highest-scoring bins from each of the eight BDT distributions were used to simultaneously scale signal contributions and set limits on $\theta$.

3. Key Contributions

1. Expanded Production Sources: Unlike previous MicroBooNE searches that focused solely on KDAR in the absorber, this analysis includes KDAR in the target, KDAR in the absorber, and KDIF in the decay volume. This significantly increases the expected scalar flux.
2. Improved Topological Analysis: The analysis moves beyond a purely topological search by incorporating reconstructed energy and momentum fraction variables, utilizing advanced BDTs to separate signal from background more effectively than previous iterations.
3. Comprehensive Systematics: The study accounts for a wide range of systematic uncertainties, including:
   • NuMI beam flux modeling (PPFX).
   • Kaon production uncertainties (up to 100% for $K_L$ decays).
   • Neutrino cross-section modeling (GENIE).
   • Detector response (ionization, scintillation light yield, electric field maps).

4. Results

• Observation: No significant excess of events over the expected background was observed. The data is consistent with the background-only hypothesis.
• Exclusion Limits: The analysis sets 95% Confidence Level (CL) upper limits on the mixing angle $\theta$.
  • At $m_S = 125$ MeV: $\theta < 3.19 \times 10^{-4}$.
  • At $m_S = 150$ MeV: $\theta < 2.79 \times 10^{-4}$.
• Comparison: These results represent the strongest experimental limits to date in the mass range $110 < m_S < 155$ MeV. They improve upon previous MicroBooNE results and are competitive with or superior to limits from other experiments (Belle-II, LHCb, NA62, ICARUS, PS191) in this specific mass window where $\pi^0$ backgrounds typically degrade sensitivity.

5. Significance

• Dark Matter Constraints: By constraining the mixing angle $\theta$ in the Higgs-portal model, this work significantly narrows the parameter space for light scalar dark matter candidates that couple to the Higgs field.
• Technique Validation: The successful application of LArTPC technology to distinguish overlapping electromagnetic showers and the use of machine learning for complex background rejection demonstrate the maturity of liquid argon detectors for Beyond Standard Model (BSM) physics searches.
• Future Impact: The methodology and limits established here provide a robust foundation for future short-baseline neutrino experiments (such as those in the Short-Baseline Neutrino Program) to continue searching for light, weakly interacting particles.