∗ ECT -05-15 Direct CP violation in B → ρ0(ω)P S O. Leitner1 , Xin-Heng Guo2 , A.W. Thomas3 ∗ † ‡ 6 0 0 1 ECT∗, Strada delle Tabarelle, 286, 38050 Villazzano (Trento), Italy 2 I.N.F.N., Gruppo Collegato di Trento, Trento, Italy n a 2 Key Laboratory of Radiation Beam Technology, Beijing, China J 1 and 3 Material Modification of National Ministry of Education, Beijing, China 1 and v Institute of Low Energy Nuclear Physics, Beijing Normal University, China 4 6 3 Thomas Jefferson National Accelerator Facility 12000 Jefferson Ave., 2 Newport News VA 23606 USA 1 0 6 0 / h p - p e h Abstract : v i X We calculate the direct CP violating asymmetry parameter, a , for B CP r π+π−π and B π+π−K decays, in the case where ρ0 ω mixing effects a→re a taken into accou→nt. We find that the direct CP asymmetr−y for B− π+π−π−, B¯0 π+π−π0, B− π+π−K− and B¯0 π+π−K¯0, reaches its max→imum when the→invariant mass π→+π− is in the vicinit→y of the ω meson mass. The inclusion of ρ0 ω mixing provides an opportunity to erase, without ambiguity, the phase − uncertainty mod(π) in the determination of the CKM angles α in case of b u and → γ in case of b s. → ∗[email protected] †[email protected] ‡[email protected] 1 Direct CP violation Direct CP violating asymmetries in B decays occur through the interference of, at least, two amplitudes with different weak phase, φ, and strong phase, δ. The extraction of the weak phase φ (which is determined by a combination of CKM matrix elements) is made through the measurement of a CP violating asymmetry. However, one must know the strong phase δ which is not still well determined in any theoretical framework. In this regard, the isospin symmetry violating mixing between ρ0 and ω can be extremely important, since it can lead to a large CP violation in B decays such as B ρ0(ω)Y π+π−Y (Y represents a meson) because the strong phase passes through →90o at the→ω resonance. Inany phenomenological treatment oftheweak decays ofhadrons, thestarting point is the weak effective Hamiltonian at low energy. It is obtained by integrating out the heavy fields from the Standard Model Lagrangian. The Operator Product Expansion is used to separate the calculation of the amplitude, A(B F) C (µ) F O B (µ), into i i → ∝ h | | i two distinct physical regimes. One is called hard, represented by C (µ) and calculated i by a perturbative approach. The other is called soft, described by O (µ) and derived i by using a non-perturbative approach. The operators, O (µ), can be understood as i local operators which govern effectively a given decay, reproducing the weak interaction of quarks in a point-like approximation. The Wilson coefficients, C (µ), represent the i physical contributions from scales higher than µ(= m ) and they can be calculated in b perturbation theory because of the property of asymptotic freedom of QCD. Factorization in B decays involves three fundamental scales: the weak interaction scale, M , the b quark mass scale, m , and the strong interaction scale, Λ . The QCD W b QCD factorization (QCDF) approach, based on the concept of color transparency as well as on a soft collinear factorization where the particle energies are bigger than the scale Λ , QCD allows usto write downthe matrixelements F O B (µ)at theleading order inΛ /m i QCD b h | | i and α . The hadronic decay amplitude involves both soft and hard contributions. At s leading order, all the non-perturbative effects are assumed to be contained in the semi- leptonic form factors and the light cone distribution amplitudes. Then, non-factorizable interactions are dominated by hard gluon exchanges and can be calculated perturbatively, inordertocorrectthenaivefactorization(NF)approximation. Ithasbeenalsoshownthat the weak annihilation contributions cannot be neglected in B meson decays even though they are power suppressed in the heavy-quark limit (Λ /m ). Their contributions QCD b are approximated in terms of convolutions of hard scattering kernels with light cone expansions for the final state mesons. Finally, the perturbative calculation of the hard scatteringspectator andannihilationcontributionsisregulatedbyaphysicalscaleoforder Λ . QCD The direct CP violating asymmetry parameter, a , is found to be small for most CP of the non-leptonic B decays when either the naive or QCD factorization framework is applied. However, in the case of B decay channels involving the ρ0 meson, it appears that the asymmetry may be large in the vicinity of ω meson mass. We stress that ρ0 ω − mixing has the dual advantages that the strong phase difference is large (passing rapidly through 90o at the ω resonance) and well known. In the vector meson dominance model, the photon propagator is dressed by coupling to the vector mesons ρ0 and ω. In this regard, the ρ0 ω mixing mechanism has been developed. Knowing the ratio, r, between − the tree and penguin amplitudes, and the strong phase, δ, as well as the weak phase, 1 φ, from the CKM matrix, it is possible to calculate the CP violating asymmetry, a , CP including the ρ0 ω mixing mechanism. More detail for all the results presented here can − be found in Ref. [1]. 2 Isospin symmetry violation and direct CP violation in B decays In Fig. 1, we show the CP violating asymmetry for B− ρ0(ω)π− π+π−π− and B¯0 ρ0(ω)π0 π+π−π0 respectively, as a function of the→energy, √S,→of the two pions com→ing from ρ0→decay, the form factor, FB→π, and the CKM matrix element parameters 1 ρ and η. For comparison, on the same plot we show the CP violating asymmetries, a , CP when NF is applied as well as QCDF where default values for the phases, ϕMi , and H,A parameters, ̺Mi are used. In the latter case, we take ϕMi = 0 and ̺Mi = 1 for all the H,A H,A H,A particles. Focusing first on Fig. 1, where the asymmetry for B− ρ0(ω)π− π+π−π− → → is plotted, we observe that the CP violating asymmetry parameter, a , can be large CP outside the region where the invariant mass of the π+π− pair is in the vicinity of the ω resonance. Thisisthefirstconsequence ofQCDfactorization,sincewithinthisframework, the strong phase can be generated not only by the ρ0 ω mechanism but also by the − Wilson coefficients. Because of the strong phase that is either at the order of α or power s suppressed by Λ /m , the CP violating asymmetry, a , may be small but a large QCD b CP asymmetry cannot be excluded. At the ω resonance, the asymmetry parameter, a , for CP B− π+π−π−, is around0% in our case. In comparison, the asymmetry parameter, a , CP → (stillattheω resonance) obtainedby applying thenaive factorizationgives 10%whereas − it gives 2% in case of QCDF with default values for ϕMi and ̺Mi . The results are quite − H,A H,A different between these approaches because of the strong phase mentioned previously. On the same figure, the asymmetry violating parameter, a , is shown for the decay CP B¯0 π+π−π0. In the vicinity of the ω resonance, the QCDF approach gives an → asymmetry of the order 8%. We obtain 20% and +5% in the case of NF and QCDF − − with the default values for ϕMi and ̺Mi . It appears as well that the asymmetry depends H,A H,A strongly on the CKM matrix parameters ρ and η, as expected. When QCDF is applied, the asymmetry for the decay B− π+π−π−, varies from 12% to 5% outside the region of the ω resonance whereas for t→he decay B¯0 π+π−π0, the asymmetry varies from → 10% down to 20%, depending on the CKM matrix element parameters, ρ and η. In − the vicinity of the ω resonance, the asymmetry, a , takes values from 2% to 5% for CP B− π+π−π− and from 5% to 30% for B¯0 π+π−π0 when ρ an−d η vary. For the d→ecay B− π+π−K−, the asy−mmetry, a ,→in the vicinity of the ω resonance, is CP → about +60% with QCDF, 40% with NF and 45% with QCDF and default values for ϕMi and ̺Mi . For the de−cay B¯0 π+π−K¯0−, when √S is near the ω resonance, the H,A H,A → asymmetry, a is about +70% with QCDF, 60% with NF and 15% with QCDF CP − − and usual default values for ϕMi and ̺Mi . There is no agreement, for the value of the H,A H,A asymmetry between the naive and QCD factorization at the ω resonance except that, in both cases, the CP violating asymmetry, a reaches its maximum in the vicinity of ω. CP Similar conclusions can be drawn to that of previous case regarding the sensitivity of the asymmetry parameter, a , as well as the CKM matrix element parameters, ρ and η. CP We included ρ0 ω mixing in order to investigate its effect on this CP violating − 2 asymmetry. The mixing through isospin violation between ω and ρ0, allows us to obtain a difference of the strong phase reaching its maximum at the ω resonance. ρ0 ω mixing − provides an opportunity to remove the phase uncertainty mod(π) in the determination of two CKM angles, α in the case of B ρ0π and γ in the case of B ρ0K. This → → phase uncertainty usually arises from the conventional determination of sin2α or sin2γ in indirect CP violation. In QCDF, the strong phase can be generated dynamically, however, the mechanism suffers from end-point singularities which are not well controlled. Itisnowapparent that theCabibbo-Kobayashi-Maskawa matrixisthedominant source of CP violation in flavour changing processes in B decays. The corrections to this dominant source coming from beyond the Standard Model are not expected to be large. In fact, the main remaining uncertainty is to deal with the procedure of factorization. The QCDF gives us an explicit picture of factorization in the heavy quark limit. It takes into account all the leading contributions as well as subleading corrections to the naive factorization. The soft collinear effective theory (SCET) has been proposed as a new procedure for factorization. In the last case, it allows one to formulate a collinear factorization theorem in terms of effective operators where new effective degrees of freedom are involved, in order to take into account the collinear, soft and ultrasoft quarks and gluons. All of these investigations allow us to increase our knowledge of B physics and to look for new physics beyond the Standard Model. Acknowledgements This work was supported in part by DOE contract DE-AC05-84ER40150, under which SURA operates Jefferson Lab and by the Special Grants for “Jing Shi Scholar” of Beijing Normal University. References [1] O. Leitner, X. H. Guo and A. W. Thomas, J. Phys. G 31, 199 (2005) [arXiv:hep-ph/0411392] and references therein. 3 80 60 60 40 40 −Κ 20 0Κ +−→ππ 0 +−→ππ 20 −%) : B -20 0%) : B 0 a(cp a(cp-20 -40 -40 -60 -60 -80 770 775 780 785 790 795 770 775 780 785 790 √S (MeV) √S (MeV) 20 15 10 15 5 10 +−−→πππ 5 +−0→πππ -05 −B 0B %) : 0 %) : -10 a(cp a(cp-15 -5 -20 -10 -25 -15 -30 765 770 775 780 785 790 795 770 775 780 785 790 795 √S (MeV) √S (MeV) Figure 1: First row, CP violating asymmetry, a , for B− π+π−K−,B¯0 π+π−K¯0 CP → → for max CKM matrix elements. Solid line (dotted line) for QCDF, dot-dot-dashed line (dot-dash-dashed line) for NF, dot-dashed line (dashed line) for QCDF with default values and for FB→K = 0.35(0.42). Second row, CP violating asymmetry, a , for CP B− π+π−π−,B¯0 π+π−π0, for max CKM matrix elements. Same notation for lines as in→first row with F→B→π = 0.27(0.35). All the figures are given as a function of √S. 4