Charmless hadronic penguin B decays with BaBar: B0$B^0(\overline{B}^0)\rigtarrow \overline{K}^0}K^0$K0 or <FONT FACE=Symbol>h¢ h¢</FONT>

Abstract

Recent results from the BaBar experiment on searches for New Physics using the charmless channels B0 -> $\overline{K}^0}K^0$K0 and B0 ® <FONT FACE=Symbol>h¢ h¢</FONT> are discussed.

B mesons; Rare charmless decays

HEAVY FLAVORS

Charmless hadronic penguin B decays with BaBar: B⁰$B^0(\overline{B}^0)\rigtarrow \overline{K}^0}K^0$K⁰ or h¢ h¢

J. William Gary

Department of Physics and Astronomy, University of California, Riverside, CA 92521 USA

ABSTRACT

Recent results from the BaBar experiment on searches for New Physics using the charmless channels B⁰® >$\overline{K}^0}K^0$K⁰ and B⁰ ® h¢ h¢ are discussed.

Keywords:B mesons; Rare charmless decays

I. INTRODUCTION

The BaBar experiment at SLAC studies e⁺e^- annihilations at the ¡(4S) resonance. The ¡(4S) is a clean, copious source of B mesons. The ¡(4S) decays about half the time to pairs and the other half of the time to B⁺B^- pairs. At the ¡(4S) energy, about 75% of the cross section is e⁺e^-® (q = u, d, s, c) continuum events, with the remaining cross section the e⁺e^-® ¡(4S) ® events. The large continuum background is reduced by considering the kinematic variables m_ES and DE, with DE º - and m_ES º , where and are the CM energy and momentum of the B meson candidate (reconstructed in the decay channel of interest) and is half the CM energy. For B candidates, m_ES is peaked at the B mass and DE is peaked at zero, while for continuum events m_ES and DE do not have any peaking structure. A third key variable to reducing the combinatoric background is the event shape. At the ¡(4S) energy, the B and mesons are produced almost at rest. Thus the event is spherical in momentum space. In contrast, the e⁺e^-® events are jet-like. A Fisher discriminant based on event shape information is used in conjunction with m_ES and DE to separate the B meson candidates from the continuum background.

The interest in charmless decays, corresponding to b ® d and b ® s quark transitions, is that they are loop diagrams. These transitions are called ''penguin'' diagrams. Thus, unlike the much more copious b ® c tree-level transitions, penguin transitions are sensitive to physics beyond the Standard Model through the virtual production of New Physics particles. Among the most important purely hadronic B⁰ decays being studied by BaBar are the tree-level B⁰® J/YK⁰ decay, and the penguin B⁰® fK⁰ and B⁰® h¢K⁰ decays. Feynman diagrams for these decays are shown in Fig. 1.

Of central interest to current studies of the B meson are measurements of the time dependent CP asymmetry in B⁰ and decays to a common final state f, defined by

This has the simple form in terms of the sin and cos functions as shown in eq. (2), with Dm the neutral B_d mass difference. For f a CP eigenstate and for decays dominated by a single weak phase, conditions which hold for the processes shown in Fig. 1, the coefficients S_f and C_f are given by S_f = h_fsin(2b) and C_f = 0, where h_f is the CP eigenvalue (= ±1) and sin(2b) is the phase difference between the B ® f and B ® ® f decay paths. Results for S_f determined from eq. (1) for the three channels shown in Fig. 1 are presented in Fig. 2. The corresponding results for C_f are found to be consistent with zero and thus agree with the Standard Model expectation.

In contrast to C_f, it is seen that the two loop processes (B⁰® fK⁰ and B⁰® h¢K⁰) yield results for sin(2b) which are systematically lower than the result from the tree level J/YK⁰ decay. This deviation is referred to as DS, i.e., DS_f = S_f - sin(2b) with sin(2b) º . Deviations DS ¹ 0 could be caused by New Physics. However, it is also possible that they are caused by sub-dominant Standard Model processes with different weak phases from the dominant diagrams, since this would break the conditions leading to S_f = h_fsin(2b). The dominant diagrams are equivalent to those shown in Fig. 1 with, for example, the quark (but not the quark) as the virtual quark in the propagator loop for B⁰® fK⁰ and B⁰® h¢K⁰. Sub-dominant processes with a different weak phase from the dominant diagrams are referred to as Standard Model pollution. Standard Model pollution to the B⁰® fK⁰ and B⁰® h¢K⁰ decay modes arise from b ® u transitions corresponding to Fig. 1 with the quark as the virtual quark and to the diagrams shown in Fig. 3.

BaBar recently completed studies of the B⁰® K⁰ and B⁰® h¢ h¢. These are published in Refs. [1] and [2], respectively. The principal motivation for studying these two channels is that they can be used to set limits on the b ® u amplitudes (the Standard Model pollution) in B⁰® fK⁰ and B⁰® h¢K⁰, using the technique based on SU(3) flavor symmetry discussed in Ref. [3]. Consider for example, the Feynman diagrams for B⁰® K⁰, which are shown in Fig. 4. These are the same as the diagrams for B⁰® fK⁰ shown in Fig. 1 except that the ® quark transition has been replaced by a ® transition. However, for B⁰® K⁰, there is no suppression of the b ® u propagator term compared to the b ® t and b ® c terms, unlike the case for B⁰® fK⁰. The conservative procedure is then to assume that the B⁰® K⁰ decay rate is dominated by the b ® u term, and to use the observed rate of B⁰® K⁰ and SU(3) flavor symmetry to set an upper limit on the b ® u amplitude (Standard Model pollution) in B⁰® fK⁰. Similarly, the B⁰® h¢ h¢ decay rate is used to set an upper limit on the Standard Model pollution in B⁰® h¢K⁰.

In practice, other charmless, strangeness conserving processes than the two in our study are necessary to set these SU(3) flavor limits on DS in B⁰® fK⁰ and B⁰® h¢K⁰ (see Ref. [3]). However, the two channels of interest for this study have been the limiting factors in this determination. The B⁰® K⁰ channel has not previously been studied, while the B⁰® h¢ h¢ was studied with only a substantially smaller data sample.

II. ANALYSIS AND RESULTS

The

K⁰ analysis is based on 210 fb^-1 of data, corresponding to 232 million pairs. B⁰ candidates are reconstructed through K^*0® K⁺p^- and K⁰® ® p⁺ p^-. The h¢ h¢ analysis is based on 289 fb^-1, corresponding to 324 million pairs. This corresponds to an increase in event statistics of about a factor of four compared to the previous h¢ h¢ study. The h¢ is reconstructed in two channels: the h¢_hpp mode (i.e., h¢ ® hp⁺p^- with h ® gg) and the h¢_rg mode (i.e., h¢ ® r⁰g with r⁰® p⁺ p^-). To reconstruct B⁰ candidates, we use h¢_hpph¢_hpp and h¢_hpph¢_rg combinations. The h¢_rgh¢_rg combinations are not used because of excessive background.

The

K⁰ analysis employs particle identification of the K⁺ and p^-, based on energy loss measurements in the tracking chambers (dE/dx) and radiation in ring imaging Cherenkov detectors. After cutting on event shape measurements to reduce continuum background, an extended maximum likelihood fit is applied to the DE, m_ES and distributions, with the invariant mass of the K^*0 candidate. Projections of the fit results are shown in comparison to the data in Fig. 5. Of the 682 events that survive the preliminary cuts, 660 ± 75 are found to be continuum background and background from events. The number of B⁰® K⁰ events is found to be . The overall detection efficiency is 2.2%. The measured branching fraction is () × 10^-6. We set a 90% confidence level upper limit on the branching fraction of 1.9 × 10^-6. As mentioned above, these are the first results for this channel.

The extended maximum likelihood fit for the h¢h¢ analysis is based on more variables: the event shape information, DE, m_ES and the two h¢ reconstructed masses. For the h¢_rg decay mode, the r helicity angle is also included in the fit. The results of the h¢h¢ fit are given in Table I. The corresponding fit projections are shown in comparison to the data in Fig. 6. The 90% confidence level upper limit we obtain for the branching fraction is 2.4 × 10^-6, a factor of four improvement compared to the previous result.

III. SUMMARY AND CONCLUSIONS

Using the formalism of Ref. [3], we set the following limits on the level of deviation of the effective sin(2b) measurements in B⁰® fK⁰ and B⁰® h¢K⁰ compared to B⁰® J/YK⁰: < 0.38 and < 0.15. The fK⁰ result is the first for this bound. The h¢K⁰ result can be compared to the previous bound of 0.22. The SU(3) bounds we obtain are shown by the horizontal bars in Fig. 2. The observed DS deviations observed for the fK⁰ and h¢K⁰ are seen to be compatible with the sin(2b) result from J/ YK⁰ within these bounds. Therefore, we do not observe evidence for New Physics. BaBar is expected to collect data until the end of 2008 and should have a final data sample of about 1 ab^-1. This will result in an increase by about a factor of three in the available number of events compared to the numbers used in the two studies presented here, corresponding to an expected improvement of about 60% in the DS bounds based on SU(3) flavor symmetry.

Received on 29 October, 2006

• [1] BABAR Collaboration, B. Aubert, et al. Phys. Rev. D 74, 072008 (2006).
• [2] BABAR Collaboration, B. Aubert, et al. Phys. Rev. D 74, 051106(R) (2006),
• [3] Y. Grossman et al., Phys. Rev. D 68, 015004 (2003).

Publication Dates

• Publication in this collection
  10 Aug 2007
• Date of issue
  June 2007

History

• Received
  29 Oct 2006