TL;DR
This paper presents a detailed study of radiative decays of the $\Upsilon(1S)$ meson into pairs of charged hadrons ($\pi^+\pi^-$, $K^+K^-$, $p\bar{p}$) using data from the CLEO III detector. The findings provide crucial measurements of branching ratios and helicity production for known tensor mesons and set stringent upper limits on the production of hypothetical glueball states, contributing to our understanding of the light-hadron spectrum and the nature of gluonic excitations.
What Problem Does This Paper Address?
Understanding the spectrum of light hadrons, particularly the existence and properties of glueballs (particles composed solely of gluons), is a fundamental challenge in quantum chromodynamics (QCD). Radiative decays of heavy quarkonia, like the $\Upsilon(1S)$, offer a unique window into this spectrum because they proceed through a two-gluon intermediate state, which can then hadronize into various final states. While radiative decays of the $J/\psi$ meson have been extensively studied, those of the $\Upsilon(1S)$ are significantly suppressed due to the larger mass of the bottom quark. This suppression makes experimental studies more challenging but also provides a cleaner probe of the underlying QCD dynamics. Previous studies had observed some $\Upsilon(1S)$ radiative decays, but a more comprehensive analysis with higher statistics was needed to precisely measure branching ratios, confirm spin assignments, and search for exotic states like glueballs, particularly the $f_J(2220)$ and $X(1860)$ which were candidates for glueballs observed in $J/\psi$ decays.
Methods
The researchers utilized a large dataset collected by the CLEO III detector at the Cornell Electron Storage Ring (CESR), specifically focusing on $\Upsilon(1S)$ resonance energy data. The analysis involved several key steps:
- Event Selection: A rigorous selection process was employed to isolate candidate events. This included requiring exactly two charged tracks with good quality fits and $dE/dx$ information, a single unmatched Crystal Calorimeter (CC) shower with significant energy, and a kinematic fit to ensure consistency with the initial $e^+e^-$ system.
- Particle Identification: Charged particles were identified as pions, kaons, or protons using information from the Ring Imaging Cherenkov Detector (RICH) and the $dE/dx$ measurements in the drift chamber.
- Background Subtraction: Continuum background processes (e.g., $e^+e^- \rightarrow \gamma\rho$, $e^+e^- \rightarrow \gamma\phi$) were modeled using data taken at the $\Upsilon(4S)$ energy and subtracted from the $\Upsilon(1S)$ data.
- Signal Extraction: Invariant mass plots for the $\pi^+\pi^-$, $K^+K^-$, and $p\bar{p}$ final states were generated. Signals for known resonances like the $f_2(1270)$ and $f_2'(1525)$ were extracted by fitting these mass distributions with spin-dependent relativistic Breit-Wigner functions.
- Helicity Analysis: To confirm spin assignments and study production mechanisms, the angular distributions of the photon and hadrons were analyzed. These distributions were simultaneously fit to a helicity formalism prediction, allowing for the determination of helicity production ratios.
- Upper Limit Determination: For potential glueball candidates like the $f_J(2220)$ and $X(1860)$, where no significant signal was observed, 90% confidence level upper limits on their product branching ratios were calculated by fitting the invariant mass spectra with a Breit-Wigner function and a flat background.
Datasets
The primary dataset used in this analysis was collected by the CLEO III detector at the $\Upsilon(1S)$ resonance energy ($\sqrt{s} = 9.46$ GeV). This dataset corresponds to approximately 21.2 million $\Upsilon(1S)$ decays. Additionally, data taken at the $\Upsilon(4S)$ energy ($\sqrt{s} = 10.56$ GeV), with an integrated luminosity of 3.49 fb$^{-1}$, was used to model and subtract continuum background processes. Monte Carlo simulations were extensively used for evaluating reconstruction efficiencies and detector response.
Key Insights and Ideas
- Confirmation of Tensor Meson Decays: The study provides strong confirmation of the radiative decays of $\Upsilon(1S)$ into the $f_2(1270)$ and $f_2'(1525)$ tensor mesons, with precise measurements of their branching ratios.
- Helicity Production Dominance: The analysis reveals that these tensor mesons are predominantly produced with helicity 0, aligning with theoretical predictions from non-relativistic QCD and other models. This provides valuable information about the underlying production mechanisms.
- Suppression of $\Upsilon(1S)$ Radiative Decays: The observed branching fractions for $\Upsilon(1S)$ radiative decays are found to be roughly 0.04 times those of the corresponding $J/\psi$ decays, consistent with expectations based on the quark-photon coupling and quark mass differences.
- No Evidence for Glueball Candidates: Crucially, the experiment found no significant evidence for the production of the $f_J(2220)$ or $X(1860)$ in $\Upsilon(1S)$ radiative decays, setting stringent upper limits on their production. This suggests that if these states are indeed glueballs, their production in $\Upsilon(1S)$ radiative decays is highly suppressed compared to known quarkonia states.
- Dominance of Quarkonia: The results indicate that the observed structures in the $\Upsilon(1S)$ radiative decay spectra are consistent with known tensor meson states composed of quarks, rather than exotic glueball states, within the experimental sensitivity.
Key Results
- Branching Ratios:
- $B(\Upsilon(1S) \rightarrow \gamma f_2(1270)) = (10.2 \pm 0.8 \pm 0.7) \times 10^{-5}$
- $B(\Upsilon(1S) \rightarrow \gamma f_2'(1525)) = (3.7^{+0.9}_{-0.7} \pm 0.8) \times 10^{-5}$
- $B(\Upsilon(1S) \rightarrow \gamma K^+K^- \text{ (excess between 2-3 GeV/c}^2)) = (1.14 \pm 0.08 \pm 0.10) \times 10^{-5}$
- Helicity Production Ratios:
- $x^2_{f_2(1270)} = 0.00^{+0.02+0.01}_{-0.00-0.00}$, $y^2_{f_2(1270)} = 0.09^{+0.08+0.04}_{-0.07-0.03}$
- $x^2_{f_2'(1525)} = 0.00^{+0.10+0.01}_{-0.00-0.01}$, $y^2_{f_2'(1525)} = 0.30^{+0.22+0.07}_{-0.22-0.07}$
- Upper Limits for $f_J(2220)$:
- $B(\Upsilon(1S) \rightarrow \gamma f_J(2220)) \times B(f_J(2220) \rightarrow \pi^+\pi^-) < 8 \times 10^{-7}$
- $B(\Upsilon(1S) \rightarrow \gamma f_J(2220)) \times B(f_J(2220) \rightarrow K^+K^-) < 6 \times 10^{-7}$
- $B(\Upsilon(1S) \rightarrow \gamma f_J(2220)) \times B(f_J(2220) \rightarrow p\bar{p}) < 11 \times 10^{-7}$
- Upper Limit for $X(1860)$:
- $B(\Upsilon(1S) \rightarrow \gamma X(1860)) \times B(X(1860) \rightarrow p\bar{p}) < 5 \times 10^{-7}$
Limitations and Future Work
The authors acknowledge that the statistical significance for some of the less significant signal candidates (e.g., $f_0(980)$, $f_4(2050)$, $f_0(1710)$) is below the 5$\sigma$ threshold for a discovery. The analysis of the $p\bar{p}$ invariant mass distribution showed no recognizable structure, particularly no near-threshold enhancement as observed in $J/\psi$ decays, which limits the ability to study this channel.
Future work could involve further theoretical investigations into the production mechanisms and helicity amplitudes of these radiative decays, especially in light of the observed helicity 0 dominance. More precise measurements or searches for other exotic states in $\Upsilon(1S)$ radiative decays could also be pursued with future experiments or improved analysis techniques. The stringent upper limits set on glueball candidates provide important constraints for theoretical models aiming to describe the light-hadron spectrum.
Summary updated