Indication of new physics in B <FONT FACE=Symbol>® F</FONT>Ks decays at the BELLE experiment

Abstract

The BELLE experiment at the asymmetric e + e − (3.5 GeV and 8 GeV) accelerator KEK-B at Tsukuba, Japan, has recorded 152 Million BB meson pairs in 4 years of data taking. In the standard model, the size of CP violation in the B meson system is characterized by a quantity called sin(2ϕ1), in which ϕ1 denotes one of the angles in the unitarity triangle. This quantity was determined precisely to be sin(2ϕ1)=0.733±0.059±0.028, based upon N =1997 events with the decay B→J/ΨKs and N =914 other B→charmonium+Kaon decays [1] [2] [3]. The measurement is consistent with the measurement of the BABAR detector at SLAC [4]. In the standard model, the decay B→ΦKs is expected to show identical CP violation. However, a measurement of N =68 events revealed −0.96±0.50+0.09−0.11 (opposite sign from expectation), which is a 3.5σ deviation from the expected standard model value [1] [5]. The probability of obtaining the observed result for B→ΦKs as a statistical fluctuation of the B→J/ΨKs result is less than 0.1%. In the standard model, B→ΦKs is dominated by a penguin diagram with a W boson and a top quark in a loop. The top quark itself radiates a gluon which creates an ss pair. Two additional decay channels, namely B→η � Ks and B→K + K − Ks are dominated by the same diagram. However, their CP violation seems to be consistent (within still large statistical errors) with the sin(2ϕ1) value from B→J/ΨKs [6] [7]. The difference between the three different decay channels is given by the spin of the Φ meson J P =1 − (vector meson), while all Kaons and the η � are J=0 − (pseudoscalar mesons). In an additional analysis, angular distributions have been investigated. In the case of B→ VV 1 a helicity analysis can be performed. For this purpose, the B→J/ΨK ∗ [8] (pseudoscalar meson Ks replaced by vector meson K ∗ )a ndB→ΦKs [9] were compared. As pointed out recently by Grossman [10], differences between the two decay channels seem visible, and are difficult to be explained within the standard model. BELLE will resume data taking in 10/2003, with an improved Silicon Vertex Detector (4 layers, larger angular coverage). An update of the analysis results is envisaged for summer 2004.