SLAC-PUB-12300 BABAR-PROC-06-158 January 2007 Charmless B Decays Contribution to the proceedings of HQL06, Munich, October 16th-20th 2006 7 0 0 2 Wolfgang Gradl (from the BABAR Collaboration) n a The University of Edinburgh J School of Physics 6 Edinburgh EH9 3JZ, UK 1 1 v 2 3 1 Introduction 0 1 0 Rare charmless hadronic B decays are a good testing ground for the standard model. 7 0 The dominant amplitudes contributing to this class of B decays are CKM suppressed / x tree diagrams and b → s or b → d loop diagrams (‘penguins’). These decays can be e used to study interfering standard model (SM) amplitudes and CP violation. They - p are sensitive to the presence of new particles in the loops, and they provide valuable e h information to constrain theoretical models of B decays. v: The B factories BABAR at SLAC and Belle at KEK produce B mesons in the i reaction e+e− → Υ(4S) → BB. So far they have collected integrated luminosities X of about 406fb−1 and 600fb−1, respectively. The results presented here are based on r a subsets of about 200–500fb−1 and are preliminary unless a journal reference is given. 2 ∆S from rare decays Thetime-dependent CPasymmetryinB decays isobservedasanasymmetry between B0 and B0 decay rates into CP eigenstates f Γ(B0 → f)−Γ(B0 → f) A (∆t) = = S sin∆m ∆t−C cos∆m ∆t, (1) cp f d f d Γ(B0 → f)+Γ(B0 → f) where ∆m = 0.502± 0.007 ps−1 and ∆t is the time difference between the decays d of the two neutral B mesons in the event. The coefficients S and C depend on the f f 1 final state f; for the ‘golden’ decay B0 → J/ψK0, for example, which proceeds via S a b → ccs transition, only one weak phase enters, and SJ/ψK0 = sin2β, CJ/ψK0 = 0. S S Here, β ≡ φ is one of the angles of the unitarity triangle of the CKM matrix. 1 Ingeneral, thepresence ofmorethanonecontributing amplitudeforthedecay can introduce additional phases, so that S measured in such a decay deviates from the f simple sin2β. Apart from standard-model amplitudes, particles beyond the standard model may contribute in loop diagrams. There are intriguing hints in experimental data that S is smaller than sin2β in B decays involving the transition b → qqs, like f B0 → φK0, B0 → η′K0, or B0 → π0K0 (see Fig. 1). However, for each of these final states the SM contribution to ∆S ≡ S −sin2β from sub-dominant amplitudes f f needs to be determined in order to draw a conclusion about the presence of any new physics. Typically, models prefer ∆S > 0 [1, 2], while for the final state η′K0, a f S small, negative ∆S is expected [3]. Measuring B decays which are related to the f ones above by approximate SU(3) flavour or isospin symmetries helps to constrain the standard-model expectation for ∆S . f 0 → 0 2.1 B φK The sub-dominant amplitudes contributing to B0 → φK0 can be constrained using SU(3) flavor relations [5]. This requires branching fraction measurements for eleven decay channels (K∗0K0,K∗0K0, and hh′ with h = ρ0,ω,φ and h′ = π0,η,η′). BABAR has measured an upper limit for the sum B(K∗0K0)+B(K∗0K0) < 1.9×10−6 [6] and an updated upper limit for φπ0 of B(φπ0) < 2.8×10−7 [7]. Together with the already known upper limits or branching fractions for the other decays in this list, this allows one to place a bound on |∆SφK0| < 0.43 [6]. 0 → ′ 0 2.2 B η K The decays B0 → η(′)π0,η′η can be used to constrain the SM pollution in B0 → η′K0, The expected branching fractions are between 0.2 and 1 × 10−6 for η(′)π0 and 0.3 - 2×10−6 for η′η. Using 211fb−1 of data, BABAR sets the following upper limits [8] at 90% confidence level (C.L.) in units of 10−6: B(B0 → ηπ0) < 1.3, B(B0 → η′η) < 1.7, B(B0 → η′π0) < 2.1, while Belle [9] measures B(B0 → η′π0) = (2.79+1.02+0.25)×10−6 −0.96−0.34 with 386×106 analysed BB pairs. With these new upper limits, the standard model expectation for ∆Sη′K0 is −0.046 < ∆Sη′K0 < 0.094 [10]. A similar improvement for S S the measurement of sin2α in B0 → π+π− is expected. Belle also measure B(B+ → η′π+) = (1.76+0.67+0.15) × 10−6 and a charge asymmetry in this channel of A = −0.62−0.14 ch 0.20+0.37 ±0.04. −0.36 2 b eff ≡ f eff sin(2 ) sin(2 ) HHFFAAGG 1 DPF/JPS 2006 PRELIMINARY b→ccs World Average 0.68 ± 0.03 f K0 Average 0.39 ± 0.18 h¢ K0 Average 0.61 ± 0.07 K K K Average 0.51 ± 0.21 S S S p 0 K Average 0.33 ± 0.21 S r 0 K Average 0.20 ± 0.57 S w K Average 0.48 ± 0.24 S f K0 Average 0.42 ± 0.17 0 p 0 p 0 K Average -0.84 ± 0.71 S K+ K- K0 Average 0.58 ± 0.13 -3 -2 -1 0 1 2 3 Figure 1: Average of S in the different b → qqs decays [4]. f 2.3 Pure penguin decays There is special interest in decays which only proceed via the b → sss penguin transitions. The b → u amplitudes can only contribute through rescattering. This drastically reduces the standard model ‘pollution’ in these decays, making them a very clean probe for the presence of new particles in the loop. An example for this class of decays is B0 → K0K0K0, in which the CP violating parameters S S S S and C have been measured by both BABAR [11] and Belle [12], with an average of S = 0.51± 0.21, C = −0.23 ± 0.15. BABAR has also searched for the related decay B0 → K0K0K0. The non-resonant contribution (besides B0 → φ(K0K0)K0) to this S S L S L S final state has not been studied before and might be large [13]. Assuming a uniform Dalitz distribution and analysing 211fb−1, BABAR[14] sets a 90% CL upper limit of B(B0 → K0K0K0) < 7.4 × 10−6. Due to a low product of efficiency and daughter S S L branching fraction, this decay is therefore of limited use for the understanding of CP violation in b → qqs decays. 3 3 Measurements related to α Decays containing a b → u transition can be used to measure the angle α ≡ φ in the 2 unitaritytriangle. Ingeneralseveral amplitudeswithdifferent weak phases contribute to these decays, only allowing the direct measurement of an effective parameter α . eff There are several methods to extract the true angle α in presence of this ‘pollution.’ Updated results for the decays B → ρρ, have been presented by Christos Touramanis at this conference. Another new decay studied by BABAR and Belle is B0 → a±π∓, from which α can 1 beextracted upto afour-foldambiguity. Exploiting isospin or approximate SU(3)fla- vorsymmetries thisambiguitycanbeovercome [15]. Thisneedsalsothemeasurement of related axial–vector decays, from which a model-dependent measurement of α can be derived. BABAR searches for B0 → a±π∓ in 211fb−1 and measures [16] a branching 1 fraction of B(B0 → a±π∓) = (33.2±3.8±3.0)×10−6, assuming B(a+ → (3π)+) = 1. 1 1 With about the same luminosity, Belle measures a slightly larger branching fraction of (48.6±4.1±3.9)×10−6 [17]. The next step is to extend this analysis to measure time-dependent CP violation in this decay. BABAR also searched for the related decay B0 → a+ρ−, which also could be used to 1 measure α. In addition, B decays to 5π are an important background for the B → ρρ analyses. In 100fb−1 no significant signal was seen; assuming a fully longitudinal polarisation, the analysis sets a 90% C.L. upper limit of B(B0 → a+ρ−)B(a+ → 1 1 (3π)+) < 61×10−6 [18]. 4 Charmless vector-vector decays For tree-dominated B decays into two vector mesons, helicity conservation arguments together with factorisation suggest that the longitudinal polarisation fraction f is L f ∼ 1 − m2 /m2 , close to unity. Experimentally, this is seen in decays such as L V B B → ρρ, where f ≈ 0.95 is observed. However, there seems to be a pattern emerging L where f is smaller than the expectation in decays dominated by loop diagrams. This L was first seen in the decays B → φK∗, where f is near 0.5 with an uncertainty of L about 0.04 [19, 20]. In the following sections, we describe a number of recent BABAR measurements for several of these vector-vector decays. 4.1 Decays involving an ω meson Toestablishwhethertree-induceddecaysgenerallyhavealargef ,BABARhassearched L for the related decays B → ωV [21], where V = ρ,K∗,ω,φ. The results are sum- marised in Table 1. The only decay with a significant observed yield is B+ → ωρ+ with B(B+ → ωρ+) = (10.6±2.1+1.6)×10−6. The polarisation f is floated in the fit −1.0 L 4 and a large value of f = 0.82±0.11±0.02 is found, as expected for a tree-dominated L decay. → ∗ 4.2 B ρK Conversely, the decays B → ρK∗ are penguin-dominated; some are known to have significant branchingfractionsandf canbemeasured. BABARhaspublishedupdated L measurements of branching fractions, charge asymmetries and polarisation fractions [22]. 4.2.1 B+ → ρ+K∗0 The decay B+ → ρ+K∗0 is particularly interesting because no tree diagramisthought to contribute to this decay. BABAR has a new measurement of the branching fraction, CP asymmetry and polarisation for this decay. The measured branching fraction is B(ρ+K∗0) = (9.6±1.7m1.5)×10−6, A (ρ+K∗0) = −0.01±0.16±0.02. The observed ch polarisation is f = 0.52±0.10±0.04, as expected for a pure penguin decay and in L good agreement with φK∗. 4.2.2 B+ → ρ0K∗+ and B0 → ρ0K∗0 The decays B+ → ρ0K∗+ and B0 → ρ0K∗0 are theoretically less clean because there is a Cabibbo-suppressed tree diagram contributing in addition to the penguin present forallB → ρK∗ decays. Inaddition,B+ → ρ0K∗+ isexperimentally morechallenging because of the smaller branching fraction. For B+ → ρ0K∗+, BABAR measures a branching fraction of (3.6+1.7±0.8)×10−6, −1.6 with a significance of only 2.6σ. The value of f determined by thefit is f = 0.9±0.2 L L although this is not considered a measurement for this decay, as the signal itself is not significant. B0 → ρ0K∗0 is observed with a significance of 5.3σ; the branching fraction is (5.6±0.9±1.3)×10−6 and f = 0.57±0.09±0.08. L 5 B → η(′)K(∗) In B decays to final states comprising η(′)K(∗), the effect of the η–η′ mixing angle combines with differing interference in the penguin diagrams to suppress the final states ηK and η′K∗, and enhance the final states η′K and ηK∗. This pattern has now been experimentally established with rather precise measurements of the branching fractions for η′K and ηK∗ and the observation of the decays η′K∗. These decays are also important in light of measuring S in B0 → η′K0. 5 → ′ 5.1 B η K Belle’s measurements for the branching fractions of B → η′π [9] were already men- tioned above. The same analysis also obtains updated branching fraction measure- ments for the decays B → η′K, with the results B(B0 → η′K0) = (58.9+3.6 ± −3.5 4.3) × 10−6, B(B+ → η′K+) = (69.2 ± 2.2 ± 3.7) × 10−6, A (B+ → η′K+) = ch 0.028±0.028±0.021. → ∗ → 5.2 B ηK and B ηρ BABAR[23]andBelle[24]havepublishedupdatedresultsforthedecaysB → ηK∗(892). Belle also observes the decay B+ → ηρ+ and obtains an upper limit for B0 → etaρ0. These results confirm earlier measurements of B → ηK∗ and ηρ. BABAR also analyses the mass region 1035 < m < 1535 MeV of the Kπ system and obtains branching Kπ fractions for the spin-0 (η(Kπ)∗) and spin-2 (ηK∗) contributions. For these two final 0 2 states no predictions exist so far. The branching fraction results are summarised in Table 2. → ′ ∗ → ′ 5.3 B η K and B η ρ BABAR [25] finds evidence for the decays B → η′K∗ in 211fb−1 and measures branch- ing fractions of B(B+ → η′K∗+) = (4.9+1.9 ± 0.8) × 10−6 and B(B0 → η′K∗0) = −1.7 (3.8±1.1±0.5)×10−6. For the related decays into η′ρ, only B+ → η′ρ+ is seen with B(B+ → η′ρ+) = (8.7+3.1+2.3)×10−6, while B0 → η′ρ0 is small with a 90% C.L. upper −2.8−1.3 limit of B(B0 → η′ρ0) < 3.7×10−6. The direct CP asymmetries in the decays with a significant signal are compatible with zero. Theoretical predictions using SU(3) fla- vor symmetry [26], QCD factorisation [27], and perturbative QCD factorisation [28] agree within errors with the observed branching fractions. The observation of small branching fractions for B → η′K∗ confirms the pattern of enhanced and suppressed decays to η(′)K(∗). → 6 B ππ, πK, KK Updated branching fraction measurements for the two-body decays B → ππ,πK, and KK from BABAR [29, 30, 31, 32] and Belle [33, 34, 35, 33] are summarised in Table 3. Both experiments observe the decays B+ → K0K+ and B0 → K0K0 with a statistical significance > 5σ; decays with b → d hadronic penguins have now been observed. BABAR also studied time dependent CP violation in B0 → K0K0 [31] (recon- structed as B0 → K0K0) which is a pure b → dss penguin decay. Via flavour S S SU(3) symmetry, this decay also allows an estimate of the penguin contribution in 6 B0 → π0π0. Direct CP asymmetry is expected to be zero. The result of the time- dependent fit is S = −1.28+0.80+0.11 and C = −0.40±0.41±0.06. −0.73−0.16 → ′ ′ 6.1 B η η K, φφK Motivated by the large branching fraction for B → η′K and the observation that final states P0P0X0 are CP eigenstates [36], BABAR searched for the decays B → η′η′K. No significant signal was found in 211fb−1, and the upper limits on the branching fractions of B(η′η′K+) < 25×10−6 and B(η′η′K0) < 31×10−6 are set [37]. BELLE searched for the decays B → φφK. In these, direct CP violation could be enhanced in the interference between decays via the η and non-SM decays. In c the analysis [38], charmless decays are selected by requiring that m is below the φφ charm threshold. For these charmless decays, the observed branching fractions are B(φφK+) = (3.2+0.6±0.3)×10−6, B(φφK0) = (2.3+1.0±0.2)×10−6. The measured −0.5 −0.7 direct CP asymmetries are compatible with zero. 7 Summary Charmless hadronic B decays provide a rich field for tests of QCD and the standard model of electroweak interactions. They allow to constrain the SM contribution to ∆S in loop-dominated B decays and precision tests of QCD models. The B factories f have produced a largenumber of new andupdated measurements. Withthe currently analysed statistics, decays with branching fractions of the order of 10−6 are within experimental reach. References [1] M. Beneke, Phys. Lett. B620, 143 (2005), hep-ph/0505075. [2] H.-Y. Cheng, C.-K. Chua, and A. Soni, Phys. Rev. D72, 014006 (2005), hep- ph/0502235. [3] A. R. Williamson and J. Zupan, Phys. Rev. D74, 014003 (2006), hep- ph/0601214. [4] HFAG, http://www.slac.stanford.edu/xorg/hfag, 2006. [5] Y. Grossman et al., Phys. Rev. D68, 015004 (2003), hep-ph/0303171. [6] BABAR, B. Aubert et al., Phys. Rev. D74, 072008 (2006), hep-ex/0606050. [7] BABAR, B. Aubert et al., Phys. Rev. D74, 011102 (2006), hep-ex/0605037. 7 [8] BABAR, B. Aubert et al., Phys. Rev. D73, 071102 (2006), hep-ex/0603013. [9] Belle, J. Schumann et al., Phys. Rev. Lett. 97, 061802 (2006), hep-ex/0603001. [10] M. Gronau, J. L. Rosner, and J. Zupan, Phys. Rev. D74, 093003 (2006), hep- ph/0608085. [11] BABAR, B. Aubert et al., (2006), hep-ex/0607108. [12] Belle, K. F. Chen et al., (2006), hep-ex/0608039. [13] H.-Y. Cheng, C.-K. Chua, and A. Soni, Phys. Rev. D72, 094003 (2005), hep- ph/0506268. [14] BABAR, B. Aubert et al., Phys. Rev. D74, 032005 (2006), hep-ex/0606031. [15] M. Gronau and J. Zupan, Phys. Rev. D73, 057502 (2006), hep-ph/0512148. [16] BABAR, B. Aubert et al., Phys. Rev. Lett. 97, 051802 (2006), hep-ex/0603050. [17] Belle, K. Abe et al., (2005), hep-ex/0507096. [18] BABAR, B. Aubert et al., Phys. Rev. D74, 031104 (2006), hep-ex/0605024. [19] BABAR, B. Aubert et al., Phys. Rev. Lett. 93, 231804 (2004), hep-ex/0408017. [20] Belle, K. F. Chen et al., Phys. Rev. Lett. 94, 221804 (2005), hep-ex/0503013. [21] BABAR, B. Aubert et al., Phys. Rev. D74, 051102 (2006), hep-ex/0605017. [22] BABAR, B. Aubert et al., Phys. Rev. Lett. 97, 201801 (2006), hep-ex/0607057. [23] Belle, K. Abe et al., (2006), hep-ex/0608034. [24] BABAR, B. Aubert et al., Phys. Rev. Lett. 97, 201802 (2006), hep-ex/0608005. [25] BABAR, B. Aubert et al., (2006), hep-ex/0607109. [26] C.-W. Chiang et al., Phys. Rev. D69, 034001 (2004), hep-ph/0307395. [27] M. Beneke and M. Neubert, Nucl. Phys. B675, 333 (2003), hep-ph/0308039. [28] X. Liu et al., Phys. Rev. D73, 074002 (2006), hep-ph/0509362. [29] BABAR, B. Aubert et al., (2006), hep-ex/0608003. [30] BABAR, B. Aubert et al., (2006), hep-ex/0607106. 8 [31] BABAR, B. Aubert et al., Phys. Rev. Lett. 97, 171805 (2006), hep-ex/0608036. [32] BABAR, B. Aubert et al., (2006), hep-ex/0607096. [33] Belle, K. Abe et al., (2006), hep-ex/0609015. [34] Belle, K. Abe et al., Phys. Rev. Lett. 94, 181803 (2005), hep-ex/0408101. [35] Belle, K. Abe et al., (2006), hep-ex/0608049. [36] T. Gershon and M. Hazumi, Phys. Lett. B596, 163 (2004), hep-ph/0402097. [37] BABAR, B. Aubert et al., Phys. Rev. D74, 031105 (2006), hep-ex/0605008. [38] Belle, K. Abe et al., (2006), hep-ex/0609016. 9 B(10−6) S(σ) B U.L ×10−6 f A L ch ωK∗0 2.4±1.1±0.7 2.4 4.2 0.71+0.27 – −0.24 ωK∗+ 0.6+1.4+1.1 0.4 3.4 0.7 fixed – −1.2−0.9 ωρ0 −0.6±0.7+0.8 0.6 1.5 0.9 fixed – −0.3 ωf (980) 0.9±0.4+0.2 2.8 1.5 – – 0 −0.1 ωρ+ 10.6±2.1+1.6 5.7 – 0.82±0.11±0.02 0.04±0.13±0.02 −1.0 ωω 1.8+1.3 ±0.4 2.1 4.0 0.79±0.34 – −0.9 ωφ 0.1±0.5±0.1 0.3 1.2 0.88 fixed – Table 1: Results of the BABAR ωX analysis: measured branching fraction B, sig- nificance including systematic uncertainties S, 90% C.L. upper limit, measured or assumed longitudinal polarisation f , charge asymmetry A . L ch B(10−6) BABAR BELLE B0 → ηK∗0 16.5±1.1±0.8 15.9±1.2±0.9 B+ → ηK∗+ 18.9±1.8±1.3 19.7+2.0 ±1.4 −1.9 B+ → ηρ+ – 4.1+1.4 ±0.34 −1.3 B0 → ηρ0 – < 1.9 B0 → η(Kπ)∗0 11.0±1.6±1.5 – 0 B+ → η(Kπ)∗+ 18.2±2.6±2.6 – 0 B0 → ηK∗0 9.6±1.8±1.1 – 2 B+ → ηK∗+ 9.1±2.7±1.4 – 2 Table 2: Branching fractions for the decays B → ηK∗, ηρ, and η(Kπ). 10