Study of $X(3915)\to J/\psi \omega$ in two-photon collisions

We study the process $\gamma \gamma \to J/\psi \omega$ using a data sample of $519.2{\mathrm{fb}}^{-1}$ recorded by the BABAR detector at SLAC at the PEP-II asymmetric-energy $e^{+}{e}^{-}$ collider at center-of-mass energies near the ${\rm Y}(nS)$ ($n=2,3,4$) resonances. We confirm the existence of the charmoniumlike resonance $X(3915)$ decaying to $J/\psi \omega$ with a significance of 7.6 standard deviations, including systematic uncertainties, and measure its mass $(3919.4\pm 2.2\pm 1.6)\mathrm{MeV}/c^2$ and width $(13\pm 6\pm 3)\mathrm{MeV}$, where the first uncertainty is statistical and the second systematic. A spin-parity analysis supports the assignment $J^P=0^{+}$ and therefore the identification of the signal as due to the ${\chi }_{c0}(2P)$ resonance. In this hypothesis we determine the product between the two-photon width and the final state branching fraction to be $(52\pm 10\pm 3)\mathrm{eV}$.

Figure 1

The $p_T$ distribution of selected candidates (solid points). The solid histogram represents the result of a fit to the sum of the simulated signal (dashed line) and background (dotted line) contributions.

Figure 2

Event distribution (solid points) in the $m({\pi }^{+}{\pi }^{-}{\pi }^0)$ versus $m({\ell }^{+}{\ell }^{-})$ plane for the (a) $e^{+}{e}^{-}$ and (b) ${\mu }^{+}{\mu }^{-}$ decay mode of the $J/\psi$. We also show the $J/\psi \omega$ signal region (tile 5) and sidebands (tiles 1–4 and 6–9).

Figure 3

Data (solid points) and normalized MC (histogram) distributions of (a) $m({\ell }^{+}{\ell }^{-})$ for events in the $\omega$ signal region and (b) $m({\pi }^{+}{\pi }^{-}{\pi }^0)$ for events in the $J/\psi$ signal region.

Figure 4

The efficiency-corrected $m(J/\psi \omega )$ distribution of selected events (solid points). The solid line represents the total fit function. The dashed line is the NR contribution. The shaded histogram is the non-$J/\psi \omega$ background defined in the text as $B(5)$ and estimated from sidebands. The vertical dashed (red) line is placed at $m(J/\psi \omega )=3.872\mathrm{GeV}/c^2$.

Figure 5

Diagram illustrating the reference frames involved in the definition of angular variables.

Figure 6

The efficiency distributions in the $X(3915)$ signal region $3890<m(J/\psi \omega )<3950\mathrm{MeV}/c^2$ (solid points) as functions of (a) $\cos {\theta }_{\ell }^{*}$, (b) $\cos {\theta }_n^{*}$, (c) $\cos {\theta }_{ln}$, (d) $\cos {\theta }_h$, and (c) ${\varphi }_l$. The curves show the results from the fits described in the text.

Figure 7

The efficiency-corrected distributions of selected events in the $X(3915)$ signal region $3890<m(J/\psi \omega )<3950\mathrm{MeV}/c^2$ (solid points). (a) $\cos {\theta }_{\ell }^{*}$, (b) $\cos {\theta }_n^{*}$, (c) $\cos {\theta }_{ln}$, and (d) $\cos {\theta }_h$. The solid (red) line represents the expected distribution for the $J^P=0^{\pm }$ assignment and the dashed (blue) line for the $J^P=2^{+}$.

Figure 8

The efficiency-corrected distributions of selected events in the $X(3915)$ signal region $3890<m(J/\psi \omega )<3950\mathrm{MeV}/c^2$ (solid points). The (a) $\cos {\theta }_n$, (b) $\cos {\theta }_l$, and (c) ${\varphi }_l$ distributions are compared with $J^P=0^{+}$ (solid red line) and $J^P=0^{-}$ (dashed blue line) expectations.