We study the rare hadronic transitions Υ(2S) → Υ(1S)η and Υ(2S) → Υ(1S)π 0 using a sample of 158 ×10 6 Υ(2S) decays collected with the Belle detector at the KEKB asymmetric-energy e + e − collider. We measure the ratios of branching fractions (B) Assuming the value B(Υ(2S) → Υ(1S)π − π + ) = (17.92±0.26)%, we obtain B(Υ(2S) → Υ(1S)η) = (3.57 ± 0.25(stat) ± 0.21(syst)) × 10 −4 and B(Υ(2S) → Υ(1S)π 0 ) < 4.1 × 10 −5 (90% CL).
B(Υ(2S)→Υ(1S)η) B(Υ(2S)→Υ(1S)πPACS numbers: 14.40. Pq,13.25.Gv In recent years, hadronic transitions between quarkonia have led to an impressive series of discoveries [1]: X(3872), Y (4260), as well as h c and h b were observed in transitions either from or to the ψ and Υ states. The phenomenology of these transitions is commonly described with the QCD Multipole Expansion formalism (QCDME) [2,3], which allows one to classify the transitions in a series of chromoelectric and chromomagnetic multiplets. In particular, theoretical predictions for η and π 0 transitions [4,5] among states are being challenged by experimental measurements. The η and π 0 transitions between vector bottomonia should be mediated either by two M1 gluons or by one E1 and one M2 gluon: both cases imply a spin flip of the b quark. The corresponding amplitude should scale as 1/m b , and its measurement yields information about the chromomagnetic moment of the b quark.By scaling from the ψ(2S) → J/ψη transition, one expects a transition width of:, and therefore a ratio of branching fractions R η,. Within the QCDME formalism, one can calculate R η,π + π − = 2.3 × 10 −3 [4] assuming the b quark mass to be m b = 4.67 GeV/c 2 [6]. A further suppression is expected for the π 0 transition, which violates isospin; here, theory predictsThe Υ(2, 3, 4S) → Υ(1S)η transitions have been studied by CLEO [7] and BaBar [8,9]; the measured branching fractions are either unexpectedly large (Υ(4S)) or too small (Υ(2S) and Υ(3S)). The parameters of the quark wave functions must be changed by more than 15% in order to account for these discrepancies [10]. Searches for the π 0 transitions have only yielded upper limits [7,9].We report here a new measurement of the transition Υ(2S) → Υ(1S)η and a search for Υ(2S) → Υ(1S)π 0 using the Belle detector at the KEKB e + e − collider [11]. The Υ(1S) is reconstructed in both the e + e − and µ + µ − decay modes; we reconstruct the η meson via its decay to γγ and π − π + π 0 , and the π 0 in the γγ final state. As a normalization sample, we reconstruct the dominant transition Υ(2S) → Υ(1S)π − π + , which has a branching fraction of (17.92±0.26)% [6]. The data sample for this analysis includes an integrated luminosity of 24.7 fb −1 at the Υ(2S) resonance peak, corresponding to (158±4)×10 6 Υ(2S) decays, and an additional 1.7 fb −1 at √ s=9.993 GeV used to study the QED continuum backgrounds. The Belle detector is described in detail elsewhere [12]. Here, we summarize the features that are relevant to the current analysis. The momentum of each charged track is measured using a four-layer double-sided Sili...