Study on the radiative decays of $\Upsilon(nS)\to \eta_b+\gamma$

In this work, we investigate the characteristics of the spin-singlet state $\eta_b$ of the bottomonia family via the radiative decays of $\Upsilon(nS)\to \eta_b+\gamma$. The theoretical estimation of the decay widths is carried out in terms of the light-front quark model (LFQM). Recently CLEO and BaBar collaborations have measured $\mathcal{B}(\Upsilon(3S)\to\gamma\eta_b)$ and the mass of ${\eta_b}$. In terms of the data we fix the concerned input parameters in our calculations of $\Upsilon(nS)\to \eta_b+\gamma$. A special attention is paid on the transition of $\Upsilon(5S)\to \eta_b+\gamma$. The BELLE data showed that the width of $\Upsilon(5S)\to \Upsilon(2S,1S)+\pi\pi$ is two orders larger than that of $\Upsilon(4S)\to \Upsilon(2S,1S)+\pi\pi$, thus some theoretical explanations have been proposed. Among them, it is suggested the inelastic final state interaction (IFSI) $\Upsilon(5S)\to B\bar B\to \Upsilon(1S)+\pi\pi$ may be a natural one. If so, a similar mechanism also applies to $\Upsilon(5S)\to B^{(*)}\bar B^{(*)}\to \eta_b+\gamma$, the precise measurement would serve as a good test whether $\Upsilon(5S)$ possess exotic components. Our calculation in the LFQM indicates that the rate of the direct process $\Upsilon(5S)\to\eta_b+\gamma$ is not anomalous compared to $\Upsilon(mS)\to\eta_b+\gamma (m=1,2,3,4)$, thus if the IFSI does apply, the rate of $\Upsilon(5S)\to\eta_b+\gamma$ should be larger than the others by orders.