Prospects of searches for invisible $B$-meson decays at… — Plain-Language Explanation

✨

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

Imagine the universe is a giant, high-speed train station where particles are constantly colliding and breaking apart. Physicists at a future facility called the FCC-ee (Future Circular Collider) plan to build the ultimate "particle train station" to study these collisions. Their goal? To catch a very specific, very rare, and very sneaky event: a B-meson (a type of heavy particle) that disappears completely without leaving a single trace.

Here is a breakdown of what the paper says, using simple analogies:

1. The "Ghost" in the Machine

In the Standard Model (our current best rulebook for how the universe works), a B-meson decaying into "nothing" (invisible particles like neutrinos) is so incredibly rare that it's like winning the lottery every day for a million years. It's so suppressed that if we do see it, it's almost certainly proof of New Physics—something our rulebook is missing, like Dark Matter or other hidden particles.

The authors are asking: If we build this massive new collider, can we catch these "ghost" particles before they vanish?

2. The Setup: A Billion Collisions

The paper assumes the FCC-ee will run at a specific energy level (the "Z pole") and produce a staggering 6 trillion (6 × 10¹²) Z bosons.

The Analogy: Imagine firing a cannonball (the Z boson) that instantly splits into two pieces. One piece is a "signal" side where the B-meson might disappear, and the other is a "tag" side where we can see everything clearly.
Because the Z boson is produced at rest, the two pieces fly off in opposite directions, like two skaters pushing off from each other. If one skater suddenly vanishes into thin air, the other skater will still be there, but the balance of the system will be off.

3. The Detective Work: Sorting the Noise

The problem is that the "train station" is incredibly noisy. Most of the time, the Z boson decays into normal particles (quarks, electrons, muons) that create a huge mess of debris. Finding a B-meson that vanishes is like trying to find a single silent whisper in a stadium full of screaming fans.

To solve this, the authors used a two-step strategy:

Step 1: The Bouncer (Preselection): They set up a bouncer at the door to kick out the obvious noise. For example, if they see a clear electron or muon on the "signal" side, they know it's not a ghost event, so they throw it out. They also check that the "tag" side is crowded with enough particles to prove a real collision happened.
Step 2: The AI Detective (The BDT): After the bouncer does its job, they use a sophisticated computer program called a Boosted Decision Tree (BDT). Think of this as a super-smart AI detective. It looks at hundreds of tiny clues:
- How much energy is missing?
- How many tracks are left behind?
- Where did the particles come from?
- Is the "missing energy" on one side balanced by a "crowded" side?
The AI learns to distinguish between three types of events:
1. The Ghost (Signal): The B-meson vanished, leaving a huge energy gap.
2. The Heavy Noise: A messy collision with lots of heavy particles (like bottom or charm quarks).
3. The Light Noise: A collision with lighter particles (like up or down quarks).

4. The Results: How Good is the Search?

The authors ran simulations to see how well this system would work. Here is what they found:

The Goal: They want to prove that if the "ghost" decays happen more often than a certain tiny number, they can find them.
The Limit: If the universe produces these invisible decays more than 7.6 billionths of a billionth (7.6 × 10⁻⁹) of the time, the FCC-ee would be able to say, "We definitely saw something, and it's not just a fluke."
The Discovery: If the rate is slightly higher (around 30 billionths of a billionth), they could actually claim a "discovery" with high confidence.

5. The Catch: Systematic Uncertainties

The paper is very honest about the difficulties. The biggest challenge isn't just the noise; it's knowing exactly how the machine works.

The Analogy: Imagine trying to weigh a feather on a scale that you aren't 100% sure is calibrated correctly. If the scale is off by even a tiny bit, your measurement of the feather is wrong.
In this case, the "scale" is the computer simulation. The authors found that if they don't understand the background noise perfectly (specifically, how often certain particles are produced), their ability to find the "ghost" drops significantly. They estimate they need to know the background noise with a precision of about 2% to get the best results.

6. Separating the Twins

The study also looked at whether they could tell the difference between two types of "ghosts": a B⁰ meson and a B⁰s meson.

The Analogy: It's like trying to tell if a vanishing act was performed by a magician wearing a red hat or a blue hat.
They found they could do this by looking for a specific "partner" particle (a Kaon) that usually travels with the B⁰s. While they can separate them, it's harder and reduces the total number of ghosts they can catch.

The Bottom Line

This paper is a feasibility study. It doesn't claim they found these invisible decays (because they haven't built the machine yet). Instead, it says:

"If we build the FCC-ee and run it as planned, we will have a unique, powerful microscope capable of hunting down these invisible B-meson decays. We can rule out theories that predict these decays happen too often, or we might finally catch a glimpse of new physics hiding in the dark."

It's a roadmap for a future hunt, showing that with the right tools and enough data, the "ghosts" of the particle world might finally be caught.

1. Problem Statement

The search for invisible decays of B mesons ( $B^0_{(s)} \to \text{invisible}$ ) represents a critical probe for physics beyond the Standard Model (SM).

SM Suppression: In the SM, decays like $B^0_{(s)} \to \nu\nu$ are heavily suppressed by loop and helicity factors, resulting in predicted branching fractions of $\mathcal{O}(10^{-25})$ . Even the four-neutrino mode $B^0_{(s)} \to \nu\nu\nu\nu$ is suppressed to $\mathcal{O}(10^{-15})$ .
New Physics (NP) Potential: Many NP scenarios, including Dark Matter, light neutralinos, axion-like particles, and Higgs-portal scalars, can significantly enhance these rates. Unlike transition decays (e.g., $B \to K\nu\nu$ ), invisible $B^0_{(s)}$ decays probe annihilation processes and couplings to all three neutrino generations without hadronic form factor uncertainties.
Current Limits: Existing limits from BaBar ( $< 2.4 \times 10^{-5}$ ) and ALEPH are orders of magnitude above SM predictions. Future $B$ -factories (Belle II) aim for sensitivities around $10^{-6}$ , but the fully invisible nature of the final state makes these searches impossible at hadron colliders like LHCb.
The Gap: There is a need for a facility with massive statistics and a clean experimental environment to push sensitivity down to the $10^{-9}$ range, where many well-motivated NP models reside.

2. Methodology

The authors perform a feasibility study for the Future Circular Collider in electron-positron mode (FCC-ee) operating at the Z-pole ( $\sqrt{s} \approx 91$ GeV).

Experimental Setup

Dataset: Assumes a "Tera-Z" run producing $6 \times 10^{12}$ Z bosons across four experiments. This yields $\sim 720$ billion $B^0$ and $\sim 180$ billion $B^0_s$ mesons.
Detector: Simulations utilize the IDEA (Innovative Detector for Electron-positron Accelerators) concept, featuring a silicon pixel vertex detector, low-mass drift chambers, and a dual-readout calorimeter.
Simulation: Events are generated using Pythia, EvtGen, and Photos. Detector response is simulated via DELPHES. Perfect Particle Identification (PID) is assumed for the study.

Analysis Strategy

The analysis employs a two-stage selection procedure to distinguish signal from background:

Preselection (Rectangular Cuts):
- Hemisphere Definition: Events are split into two hemispheres based on the thrust axis (maximizing $\sum |p_i \cdot \hat{n}|$ ). The "signal hemisphere" is defined as the one with the lowest total reconstructed energy (where the invisible decay occurs).
- Key Cuts:
  - Total event energy $< 85$ GeV (targeting missing energy).
  - No reconstructed electrons or muons in the signal hemisphere (rejects $Z \to \ell\ell$ ).
  - Primary Vertex (PV) invariant mass $< 40$ GeV (isolates heavy quark decays).
  - Signal hemisphere must have at least one charged track (from fragmentation).
  - Non-signal hemisphere particle multiplicity $> 10$ (suppresses $Z \to \tau\tau$ ).
Multivariate Classification (BDT):
- A multiclass Boosted Decision Tree (BDT) using XGBoost is trained to classify events into three categories: Signal ( $B^0_{(s)} \to \nu\nu$ ), Heavy Background ( $Z \to b\bar{b}, c\bar{c}$ ), and Light Background ( $Z \to s\bar{s}, d\bar{d}, u\bar{u}$ ).
- Input Variables (49 total): Includes energy/multiplicity in each hemisphere, displaced vertex counts, thrust vector components, impact parameters, and fit qualities.
- Signal Characteristics: Large missing energy on the signal side, low multiplicity on the signal side, and displaced vertices on the opposite side (from the other $b$ -quark).

Sensitivity Estimation

Three levels of statistical assumption were used to calculate sensitivity:

Single-bin counting: No systematic uncertainties.
Counting with Systematics: Includes uncertainties on efficiency, branching fractions, and fragmentation.
Binned Fits: A binned likelihood fit to the 2D space of BDT outputs ( $P(\text{heavy})$ vs. $P(\text{light})$ ) using pseudoexperiments, accounting for correlated systematic uncertainties ( $\sigma_B$ ).

3. Key Contributions

First FCC-ee Study: This is the first comprehensive feasibility study for invisible $B$ -meson decays at the FCC-ee, establishing the physics reach for the "Tera-Z" run.
Advanced Background Rejection: Demonstrates that a multiclass BDT can effectively separate signal from heavy ( $b\bar{b}$ ) and light ( $q\bar{q}$ ) hadronic backgrounds, which have distinct topological signatures in the signal vs. non-signal hemispheres.
Systematic Uncertainty Analysis: Quantifies the impact of systematic uncertainties, identifying $b$ -quark fragmentation fractions and MC sample size as the dominant limiting factors.
Signal Separation: A preliminary study shows that $B^0$ and $B^0_s$ signals can be partially separated (retaining $\sim 66\%$ of $B^0_s$ while rejecting $\sim 78\%$ of $B^0$ ) by tagging prompt kaons ( $K^\pm, K^0_S$ ) on the signal side.

4. Results

Assuming $6 \times 10^{12}$ Z bosons and optimal BDT cuts, the projected sensitivities are:

Method	90% CL Exclusion Limit	95% CL Exclusion Limit	3 $\sigma$ Discovery	5 $\sigma$ Discovery
Binned Fits (Realistic)	$2.1 \times 10^{-8}$	$2.5 \times 10^{-8}$	$4.2 \times 10^{-8}$	$7.1 \times 10^{-8}$
Counting (No Syst.)	$7.6 \times 10^{-9}$	$9.7 \times 10^{-9}$	$1.8 \times 10^{-8}$	$3.0 \times 10^{-8}$

Optimistic Scenario: If systematic uncertainties are negligible, the experiment could exclude branching fractions above $7.6 \times 10^{-9}$ (90% CL) and discover signals above $3.0 \times 10^{-8}$ (5 $\sigma$ ).
Realistic Scenario: When accounting for systematic uncertainties (specifically requiring $\sigma_B/B \lesssim 2\%$ to match Belle II improvements), the discovery reach is approximately $7.1 \times 10^{-8}$ .
Comparison: These limits represent an improvement of roughly two orders of magnitude over current BaBar limits and significantly surpass the projected sensitivity of Belle II ( $\sim 10^{-6}$ ).

5. Significance

New Physics Probe: The FCC-ee offers a unique, potentially unique, opportunity to search for invisible $B$ decays. A discovery would provide striking evidence for New Physics, particularly models involving dark sectors or long-lived neutral particles that do not necessarily affect $b \to s \nu \nu$ transition rates.
Complementarity: These searches are complementary to $b \to s \nu \nu$ measurements (like the recent Belle II anomaly). While transition decays constrain flavor-changing currents, invisible $B^0_{(s)}$ decays test annihilation processes and couplings to neutrino generations in a theoretically clean environment (free from hadronic form factors).
Feasibility: The study confirms that with the IDEA detector and advanced machine learning techniques, the FCC-ee can achieve the necessary background rejection to probe branching fractions down to the $10^{-8}$ range, provided systematic uncertainties (particularly fragmentation fractions) are controlled to the percent level.

In conclusion, the paper establishes that the FCC-ee is a premier facility for searching for invisible B-meson decays, capable of probing physics scales and models inaccessible to current or near-future $B$ -factories.

Prospects of searches for invisible BBB-meson decays at FCC-ee