Observation of an Altered $a_{0}(980)$ Line shape in $D^{+} \rightarrow \pi^{+}\eta\eta$

Using BESIII data, this paper reports the first amplitude analysis of $D^{+} \rightarrow \pi^{+}\eta\eta$, revealing that the observed $a_{0}(980)$ line shape significantly deviates from conventional models and cannot be satisfactorily described without compromising the physical consistency of the resonance's pole position.

The Gist

Imagine the subatomic world as a giant, chaotic dance floor. In this dance, particles called D+ mesons (the dancers) occasionally split apart into three other particles: a pion and two eta particles.

Physicists at the BESIII detector in China wanted to watch this specific dance move closely. Their goal was to understand the "footwork" of a specific intermediate step in the dance, involving a particle called the a0(980).

Here is the story of what they found, explained simply:

1. The Mystery of the "Footprint"

When the D+ meson splits, it often briefly forms the a0(980) particle before it breaks down further. Think of the a0(980) as a specific dance step. In previous experiments, scientists had a very clear idea of what this step looked like—a specific "footprint" or shape in the data.

However, when the BESIII team watched the D+ → π+ηη dance, the footprint they saw was weird. It didn't match the familiar shape they expected. It was like seeing a dancer do a pirouette that looked slightly different from every pirouette they had ever seen before.

2. Trying to Fix the Puzzle

The scientists tried to explain this weird shape using four different "instruction manuals" (mathematical models) that describe how the a0(980) should behave. These manuals are like different theories on how a dancer should move their feet:

• The Flatté Manual: The standard rulebook.
• The Dispersive Flatté Manual: A slightly tweaked version of the rulebook.
• The T-Matrix Manual: A complex theory involving interactions with other dancers.
• The K-Matrix Manual: Another sophisticated theory about how particles bounce off each other.

The Result: No matter which manual they used, the predicted dance steps did not match the actual video footage. The "footprint" in the data just didn't fit the models.

3. Adding More Dancers?

The team thought, "Maybe there are other dancers we missed!" They tried adding other possible intermediate particles (like f0 or f2 resonances) to their models, hoping these extra dancers would fill in the gaps and make the math work.

The Result: It didn't help. Adding these extra dancers made the math fit the data slightly better, but it created a new problem: the extra dancers weren't actually there in the data. It was like trying to fix a blurry photo by adding random pixels; the picture looked sharper, but the new pixels were fake. The "weird" shape of the a0(980) remained unexplained.

4. The "Magic" Adjustment

Finally, the scientists tried a different approach. Instead of following the rulebooks strictly, they let the a0(980)'s properties float (change freely) to see if they could force the model to match the data.

The Result:

• Success: When they let the numbers change, the model finally matched the video footage perfectly. The "footprint" was explained.
• The Catch: To make the math work, they had to change the fundamental nature of the a0(980). The model required the particle to be much heavier and exist in a state that contradicts everything we know about it. It was like saying, "To explain this dance, the dancer must be made of lead and weigh 500 pounds." While the math worked, the physics made no sense.

5. The Conclusion

The paper concludes that there is a tension (a conflict) between getting a good mathematical fit and having a physically sensible explanation.

• If you use the known rules, the data doesn't fit.
• If you force the data to fit, you break the known rules of physics.

The scientists suggest that the "weird" shape of the a0(980) in this specific decay isn't caused by a new, hidden particle or a simple mistake in the math. Instead, it suggests that our current understanding of how these particles are produced (the "direct-production" picture) might be missing a deeper, more complex mechanism. Something else is happening on the dance floor that we haven't figured out yet.

The Bonus: Measuring the Dance

While solving the mystery of the shape, the team also successfully counted how often this specific dance happens. They measured the branching fraction (the probability of this dance occurring) to be 0.367%. This is a precise number that other scientists can use, even if the "why" behind the shape remains a puzzle.

In short: The scientists found a particle behaving strangely. They tried every known rulebook to explain it, but none worked. When they bent the rules to make it work, the explanation became physically impossible. This suggests our current understanding of how these particles dance is incomplete.

Technical Summary

Technical Summary: Observation of an Altered $a_0(980)$ Line shape in $D^+ \to \pi^+\eta\eta$

Problem and Motivation
The internal structure of light scalar mesons, particularly the $a_0(980)$, remains a subject of active debate in hadron spectroscopy. While recent studies of charm-hadron decays involving the $a_0(980)$ have provided valuable experimental data, the extracted line shapes are often complicated by interference with other intermediate amplitudes. The decay $D^+ \to \pi^+\eta\eta$ offers a relatively clean environment for studying the $a_0(980)$ because the kinematic boundary of the $\pi^+\eta$ invariant mass ($1.322$ GeV/$c^2$) suppresses higher-mass resonances like the $a_2(1320)$, and contributions from the $\eta\eta$ system are expected to be small. This work presents the first amplitude analysis of this decay to investigate the $D^+ \to a_0(980)^+\eta$ intermediate process and the associated $a_0(980)$ line shape.

Methodology
The analysis utilizes $20.3$ fb$^{-1}$ of $e^+e^-$ collision data collected at $\sqrt{s} = 3.773$ GeV with the BESIII detector. The signal is identified using the double-tag (DT) method to suppress backgrounds, selecting events where the $D^-$ is reconstructed in specific tag channels ($K^+\pi^-\pi^-$, $K^0_S\pi^-$, etc.) and the $D^+$ decays to $\pi^+\eta\eta$. A multivariate analysis (BDTG) is employed to further reduce background, resulting in a sample of 1624 candidates with a purity of $(85.1 \pm 0.9)\%$.

An unbinned maximum likelihood fit is performed on the signal events. The amplitude analysis tests four conventional parameterizations for the $a_0(980)$ line shape:

1. Flatté parameterization (with parameters fixed to CLEO values).
2. Dispersively modified Flatté parameterization (with parameters fixed to previous BESIII values).
3. T-matrix formalism (based on coupled-channel final-state interactions).
4. K-matrix formalism (using scattering parameters from literature).

The baseline model includes only the $D^+ \to a_0(980)^+\eta$ chain. To test for missing components, various additional resonant (e.g., $f_0(1370)$, $f_2(1270)$, $f_2(1565)$) and non-resonant amplitudes are added individually and in combination. When the baseline models fail to describe the data, the parameters of the $a_0(980)$ (bare mass $M_0$ and coupling constants $g^2$) are allowed to float in the fit. Pole positions are calculated on the relevant Riemann sheets to characterize the resonance properties.

Key Results

• Dominant Process: The intermediate process $D^+ \to a_0(980)^+\eta$, followed by $a_0(980)^+ \to \pi^+\eta$, is observed as the only significant component. The branching fraction for the decay $D^+ \to \pi^+\eta\eta$ is measured to be $(3.67 \pm 0.12_{\text{stat}} \pm 0.06_{\text{syst}}) \times 10^{-3}$.
• Line Shape Discrepancy: The observed $\pi^+\eta$ mass spectrum exhibits a line shape that differs substantially from those observed in other processes (such as $D_{(s)} \to a_0(980)\pi$ and $D^0 \to a_0(980)^- e^+\nu_e$).
• Failure of Fixed-Parameter Models: None of the conventional descriptions (Flatté, dispersive Flatté, T-matrix, K-matrix) with reference parameters reproduce the observed line shape satisfactorily. Adding small conventional resonant or non-resonant amplitudes does not resolve the discrepancy; apparent improvements in fit quality are attributed to strong correlations and interference with the dominant amplitude rather than stable independent contributions.
• Floating Parameters and Pole Shift: When the $a_0(980)$ parameters are allowed to float in the Flatté and dispersive Flatté models, the fit quality improves significantly. However, this results in a pole mass ($M_{\text{pole}}$) that is driven well above the $K\bar{K}$ threshold (e.g., $M_{\text{pole}} \approx 1.096$ GeV/$c^2$ for the Flatté case). This position is inconsistent with the near-threshold character typically associated with the $a_0(980)$.
• Systematic Checks: Systematic uncertainties, including variations in background shape, reconstruction efficiency, and the inclusion of additional amplitudes, were evaluated. These variations do not bring the pole mass back to the near-threshold region.

Significance and Claims
The paper concludes that there is a tension between achieving a good fit quality and maintaining a physical pole position consistent with the near-threshold nature of the $a_0(980)$ within conventional direct-production amplitude models. The observed distortion in the line shape cannot be explained solely by the choice of parameterization or by interference with small additional conventional amplitudes.

The authors state that these results suggest "mechanisms beyond the conventional tree-level amplitude picture may be relevant" for describing the $a_0(980)$ line shape in $D^+ \to \pi^+\eta\eta$. The study does not claim to have identified a new resonance or a specific new dynamical mechanism but rather highlights a limitation of standard direct-production models in this specific decay channel, indicating that additional dynamical effects may be at play.