Observation of $\Lambda^+_c\to n\pi^+\eta$ and search for $\Lambda^+_c\to na_0(980)^+$

Using 6.1 fb$^{-1}$ of data from the BESIII detector, this study reports the first observation of the decay $\Lambda_c^+\to n\pi^+\eta$ with a significance of $9.5\sigma$, measures its branching fraction relative to $\Lambda_c^+\to \Lambda\pi^+\eta$, and sets an upper limit on the intermediate process $\Lambda_c^+\to na_0(980)^+$ while employing a Transformer-based deep learning approach for background suppression.

The Gist

The Big Picture: A Cosmic Detective Story

Imagine the universe as a giant, high-speed racetrack. At the BESIII laboratory in Beijing, scientists are running a massive experiment where they smash electrons and positrons (tiny particles of light and matter) together at incredible speeds.

When these particles collide, they don't just disappear; they explode into a shower of new, exotic particles. Think of it like smashing two watches together and seeing what gears, springs, and springs fly out. The scientists are trying to catch specific, rare "gears" that fly out to understand how the universe is built.

In this paper, the team is hunting for a very specific, rare event: a heavy particle called the $\Lambda_c^+$ (Lambda-c-plus) decaying into a neutron, a pion, and an eta particle.

The Mystery: The "Ghost" Particle

The main character in this story is a particle called $a_0(980)$. Physicists have been arguing about what this particle actually is for decades. Is it a simple pair of quarks (like a standard Lego brick)? Is it a complex, four-quark "tetraquark" (like a glued-together Lego structure)? Or is it something else entirely?

Previously, the team found that the $\Lambda_c^+$ particle sometimes turns into a Lambda particle and an $a_0(980)$. But the math didn't add up. The rate at which this happened was way different than what the theories predicted. It was like a magician pulling a rabbit out of a hat, but the rabbit was three times bigger than the hat. Something was missing from the theory.

To solve this, the team decided to try a different trick. Instead of looking at the Lambda particle, they looked at what happens when the $\Lambda_c^+$ turns into a neutron (a neutral particle found in the nucleus of atoms) instead. This is a much harder, rarer trick to pull off.

The Challenge: Finding a Needle in a Haystack

Here is the problem: The "haystack" is huge. When they smash the particles together, they get billions of events. Most of these are boring background noise—particles that look almost like the signal they want, but aren't.

Traditionally, scientists use "cut-based" analysis. Imagine trying to find a specific red marble in a bucket of mixed marbles by saying, "I'll only keep the marbles that are exactly 2cm wide and weigh 5 grams." This works okay, but it's clumsy. You might throw away the red marble because it's slightly squished, or keep a fake one because it happens to be the right size.

In this experiment, the background noise was so loud that the "red marble" (the signal) was completely invisible using old methods. It was like trying to hear a whisper in a rock concert.

The Solution: The AI "Super-Brain"

This is where the paper gets exciting. The team didn't use old-fashioned rules. Instead, they built a Deep Learning system, specifically using a Transformer architecture (the same type of AI technology that powers modern chatbots and translation tools).

Think of the AI as a super-detective who has been trained on millions of examples of "real signals" and "fake noise."

• Old Method: The detective looks at the suspect's height and weight.
• AI Method: The detective looks at the suspect's gait, the way they blink, the texture of their clothes, the smell of their perfume, and the rhythm of their heartbeat. It looks at everything at once.

The AI was trained to look at the "point cloud" of data—the angles, energies, and paths of every single particle flying out of the collision. It learned to spot the subtle, hidden patterns that human-designed rules would miss.

The Results: A New Discovery

Thanks to this AI detective, the team was able to filter out the noise.

1. The Discovery: They found the decay $\Lambda_c^+ \to n\pi^+\eta$ for the first time ever. They were so sure it was real that they gave it a "statistical significance" of 9.5 sigma. In the world of physics, 5 sigma is the gold standard for a discovery; 9.5 is like finding a needle in a haystack, then finding the same needle in a second haystack, and then a third, all in the same spot. It's undeniable.
2. The Measurement: They measured how often this happens compared to the older, known decay. It turns out this "neutron version" happens about 15% as often as the "Lambda version."
3. The Search for the Ghost: They also tried to see if the $a_0(980)$ particle was the "middleman" in this process (i.e., did the $\Lambda_c^+$ turn into a neutron and an $a_0(980)$, which then broke apart?). They didn't see a clear signal. However, they set a strict "speed limit" (an upper limit) on how often this could be happening. This helps theorists know that if the $a_0(980)$ is a tetraquark, it can't be doing this too often.

Why Does This Matter?

This isn't just about counting particles. It's about understanding the rules of the universe.

• The "Standard Model" is incomplete: We know the basic rules of how particles interact, but the behavior of these "exotic" particles (like the $a_0(980)$) is confusing.
• Testing the Theories: By measuring exactly how often these rare decays happen, scientists can tell which theory about the $a_0(980)$ is correct. Is it a simple pair of quarks? A complex tetraquark? Or a "molecule" of two other particles stuck together?
• The AI Revolution: Perhaps the most important takeaway is the tool they used. This paper proves that AI isn't just a buzzword; it is a necessary tool for modern physics. It can see patterns in data that human brains simply cannot, allowing us to discover things that were previously invisible.

The Bottom Line

The BESIII team used a super-smart AI to sift through a mountain of cosmic noise and found a rare, new particle decay. This discovery helps us solve the mystery of what the $a_0(980)$ particle really is, bringing us one step closer to understanding the fundamental building blocks of our universe. And they did it by teaching a computer to "see" the invisible.

Technical Summary

1. Problem Statement and Motivation

The paper addresses two primary physics goals:

• Probing the Structure of the $a_0(980)$ Meson: The light scalar meson $a_0(980)$ has an enigmatic internal structure, with competing hypotheses suggesting it is a conventional $q\bar{q}$ state, a compact tetraquark, a molecular state, or a dynamically generated resonance. Previous BESIII measurements of $\Lambda^+_c \to \Lambda a_0(980)^+$ showed branching fractions (BF) orders of magnitude larger than theoretical predictions, hinting at exotic substructures.
• First Observation of $\Lambda^+_c \to n\pi^+\eta$: While the decay $\Lambda^+_c \to \Lambda\pi^+\eta$ was previously observed, the singly Cabibbo-suppressed (SCS) decay $\Lambda^+_c \to n\pi^+\eta$ (where the final state $\Lambda$ is replaced by a neutron) had not been experimentally investigated. This channel is crucial for testing theoretical models (e.g., SU(3)$_f$ flavor symmetry, chiral unitary approaches) and isolating the intermediate $\Lambda^+_c \to n a_0(980)^+$ process.
• Challenge of Backgrounds: The $\Lambda^+_c \to n\pi^+\eta$ signal is buried under substantial hadronic backgrounds involving various intermediate states. Conventional cut-based analysis methods failed to extract the signal with sufficient statistical significance, necessitating advanced machine learning techniques.

2. Methodology

Data and Simulation

• Dataset: The analysis utilized 6.1 fb$^{-1}$ of $e^+e^-$ collision data collected at center-of-mass energies between $\sqrt{s} = 4.600$ and $4.843$ GeV using the BESIII detector at the BEPCII collider.
• Production Mechanism: $\Lambda^+_c$ baryons were dominantly produced via pair production ($e^+e^- \to \Lambda^+_c \bar{\Lambda}^-_c$).
• Monte Carlo (MC): Extensive MC simulations (using GEANT4, KKMC, EvtGen, and LUNDCHAM) were used to model detector response, estimate backgrounds, and train the deep learning models. Signal MC assumed phase-space (PHSP) distributions for $\Lambda^+_c \to n\pi^+\eta$ and $\Sigma^0\pi^+\eta$, while $\Lambda\pi^+\eta$ used a Partial Wave Analysis (PWA) model.

Analysis Strategy: Double-Tag (DT) Technique

To constrain the kinematics of the signal decay, the collaboration employed a Double-Tag method:

1. Single Tag (ST): The recoil $\bar{\Lambda}^-_c$ baryon was reconstructed in one of 11 exclusive hadronic decay modes (e.g., $\bar{p}K^0_S$, $\bar{p}K^+\pi^-$, etc.).
2. Double Tag (DT): The signal $\Lambda^+_c \to X\pi^+\eta$ (where $X = n, \Lambda, \Sigma^0$) was searched for in the recoil system of the ST candidate.
3. Neutron Reconstruction: The neutron was not directly reconstructed but inferred via missing mass ($M_{miss}$) calculations based on the known beam energy and the reconstructed $\bar{\Lambda}^-_c$, $\pi^+$, and $\eta$.

Deep Learning Signal Identification

To overcome the high background levels that obscured the signal in traditional $M_{miss}$ distributions:

• Architecture: A Transformer-based Deep Neural Network (DNN), adapted from the Particle Transformer (ParT) architecture, was employed.
• Input Features: The model processed data as a "point cloud," ingesting low-level detector information for all charged tracks (angles, momentum, $dE/dx$, TOF) and photon showers (energy, time, crystal count).
• Training: The DNN was trained to classify events into three categories: Signal ($\Lambda^+_c \to X\pi^+\eta$), $\Lambda^+_c$ background, and non-$\Lambda^+_c$ background. An ensemble of 50 DNNs was used to improve robustness.
• Selection: Events were selected based on DNN output scores, significantly reducing background by a factor of ~20 while retaining signal efficiency.

Signal Extraction and Fitting

• Unbinned Maximum Likelihood Fits: Performed on the $M_{miss}$ spectra for both $\eta \to \gamma\gamma$ and $\eta \to \pi^+\pi^-\pi^0$ channels simultaneously.
• Components: The fit included signal components for $n\pi^+\eta$, $\Lambda\pi^+\eta$, and $\Sigma^0\pi^+\eta$, along with background components.
• Branching Fraction Calculation: The relative branching fraction was calculated using the ratio of yields and efficiencies, canceling many systematic uncertainties. The absolute BF was derived using the world average of $\mathcal{B}(\Lambda^+_c \to \Lambda\pi^+\eta)$.

Search for Intermediate Resonance

• $a_0(980)^+$ Search: The $\pi^+\eta$ invariant mass spectrum of the signal events was analyzed to search for the intermediate process $\Lambda^+_c \to n a_0(980)^+$.
• Fit Model: The signal PDF modeled the incoherent superposition of a Flatté line shape (for $a_0(980)^+$) and a non-resonant PHSP distribution.
• Upper Limit: Since no significant signal was found, a Bayesian method was used to set a 90% confidence level (CL) upper limit on the product of branching fractions.

3. Key Contributions

• First Experimental Observation: This is the first observation of the decay $\Lambda^+_c \to n\pi^+\eta$ with a statistical significance of 9.5$\sigma$.
• Advanced Machine Learning Application: The paper demonstrates the successful application of a Transformer-based deep learning architecture to distinguish rare charm baryon decays from overwhelming hadronic backgrounds, a technique that proved superior to traditional cut-based or BDT methods in this context.
• Precision Measurement: Provided a precise measurement of the relative branching fraction and the first absolute branching fraction for this channel.
• Constraint on $a_0(980)$: Set stringent limits on the intermediate $a_0(980)^+$ contribution, providing new data to constrain theoretical models of the scalar meson's structure.

4. Results

Branching Fractions

• Relative Branching Fraction:
  $$\frac{\mathcal{B}(\Lambda^+_c \to n\pi^+\eta)}{\mathcal{B}(\Lambda^+_c \to \Lambda\pi^+\eta)} = 0.155 \pm 0.031_{\text{stat}} \pm 0.012_{\text{syst}}$$
• Absolute Branching Fraction:
  Using the world average $\mathcal{B}(\Lambda^+_c \to \Lambda\pi^+\eta)$, the absolute BF is calculated as:
  $$\mathcal{B}(\Lambda^+_c \to n\pi^+\eta) = (2.94 \pm 0.59_{\text{stat}} \pm 0.23_{\text{syst}} \pm 0.13_{\text{ref}}) \times 10^{-3}$$
  Note: This result is consistent with theoretical predictions from Geng et al. (Ref. [17]) within 2$\sigma$.

Search for $\Lambda^+_c \to n a_0(980)^+$

• Signal Observation: No significant signal for the intermediate $a_0(980)^+$ was observed in the $\pi^+\eta$ invariant mass spectrum.
• Upper Limit: The 90% CL upper limit on the product branching fraction is:
  $$\mathcal{B}(\Lambda^+_c \to n a_0(980)^+) \times \mathcal{B}(a_0(980)^+ \to \pi^+\eta) < 8.4 \times 10^{-4}$$
• Ratio Limit: The upper limit on the ratio relative to the total three-body decay is:
  $$\frac{\mathcal{B}(\Lambda^+_c \to n a_0(980)^+) \times \mathcal{B}(a_0(980)^+ \to \pi^+\eta)}{\mathcal{B}(\Lambda^+_c \to n\pi^+\eta)} < 0.27$$
  This ratio is close to the prediction from the chiral unitary approach (Ref. [18]) but lower than some recent topological SU(3) predictions.

5. Significance

• Theoretical Impact: The measurement provides critical experimental constraints for QCD-based theoretical frameworks describing charm baryon decays. The consistency with SU(3) flavor symmetry predictions (Ref. [17]) supports the validity of these models, while the lack of a strong $a_0(980)$ signal helps refine the understanding of scalar meson dynamics in weak decays.
• Methodological Advancement: The success of the Transformer-based deep learning approach in this analysis highlights a paradigm shift in high-energy physics data analysis. It demonstrates that deep learning can effectively extract signals from complex, high-background environments where traditional methods fail, paving the way for similar applications in future BESIII and other collider analyses.
• Future Prospects: With the anticipated larger data samples at BESIII, the sensitivity to the $a_0(980)$ component will improve, potentially allowing for a definitive determination of the scalar meson's role in charm baryon decays and its internal structure.