Electroweak penguin measurements Marcin Chrząszcz mchrzasz@cernch Universität Zürich, Institute of Nuclear Physics, Polish Academy of Science 5 th KEK Flavour Factory Workshop October 26-27, Tokyo 2015 1 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 1/30 30
Why rare decays? In SM allows only the charged interactions to change flavour Other interactions are flavour conserving One can escape this constrain and produce b s and b d at loop level This kind of processes are suppressed in SM Rare decays New Physics can enter in the loops 2 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 2/30 30
Tools Operator Product Expansion and Effective Field Theory i=1,2 i=3-6,8 H eff = 4G f V V C i (µ)o i (µ) + C 2 }{{} i(µ)o i(µ), i=910 }{{} i i=s left-handed right-handed i=p where C i are the Wilson coefficients and O i are the corresponding effective operators Tree Gluon penguin i=7 Photon penguin EW penguin Scalar penguin Pseudoscalar pen 3 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 3/30 30
LHCb detector - tracking Excellent Impact Parameter (IP) resolution (20 µm) Identify secondary vertices from heavy flavour decays Proper time resolution 40 fs Good separation of primary and secondary vertices Excellent momentum (δp/p 04 06%) and inv mass resolution Low combinatorial background 4 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 4/30 30
LHCb detector - particle identification Excellent Muon identification ϵ µ µ 97%, ϵ π µ 1 3% Good K π separation via RICH detectors, ϵ K K 95%, ϵ π K 5% Reject peaking backgrounds High trigger efficiencies, low momentum thresholds Muons: p T > 176GeV at L0, p T > 10GeV at HLT1, B J/ψX: Trigger 90% 4 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 4/30 30
Recent measurements Branching fractions: B 0,± K 0,± µ µ + LHCb, Mar 14 B 0 K µ µ + CMS, Jul 15 Bs 0 ϕµ µ + LHCb, Jun 15 B ± π ± µ µ + LHCb, Sep 15 Λ b Λµ µ + LHCb, Mar 15 B µ µ + CMS+LHCb, Jun 15 CP asymmetry: B ± π ± µ µ + LHCb, Sep 15 Isospin asymmetry: B Kµ µ + LHCb, Mar 14 Lepton Universality: B ± K ± ll LHCb, Jun 14 Angular: B 0 K ll LHCb, Jan 15 B ± K,± ll BaBar, Aug 15 Bs 0 ϕll LHCb, Jun 15 Λ b Λµ µ + LHCb, Mar 15 5 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 5/30 30
Recent measurements Branching fractions: B 0,± K 0,± µ µ + LHCb, Mar 14 B 0 K µ µ + CMS, Jul 15 Bs 0 ϕµ µ + LHCb, Jun 15 B ± π ± µ µ + LHCb, Sep 15 Λ b Λµ µ + LHCb, Mar 15 B µ µ + CMS+LHCb, Jun 15 CP asymmetry: B ± π ± µ µ + LHCb, Sep 15 Isospin asymmetry: B Kµ µ + LHCb, Mar 14 Lepton Universality: B ± K ± ll LHCb, Jun 14 Angular: B 0 K ll LHCb, Jan 15 B ± K,± ll BaBar, Aug 15 B 0 s ϕll LHCb, Jun 15 Λ b Λµ µ + LHCb, Mar 15 > 2 σ deviations from SM 5 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 5/30 30
B 0 K µ µ + kinematics The kinematics of B 0 K µ µ + decay is described by three angles θ l, θ k, ϕ and invariant mass of the dimuon system (q 2 ) cos θ k : the angle between the direction of the kaon in the K (K ) rest frame and the direction of the K (K ) in the B 0 (B 0 ) rest frame cos θ l : the angle between the direction of the µ (µ + ) in the dimuon rest frame and the direction of the dimuon in the B 0 (B 0 ) rest frame ϕ: the angle between the plane containing the µ and µ + and the plane containing the kaon and pion from the K 6 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 6/30 30
B 0 K µ µ + kinematics The kinematics of B 0 K µ µ + decay is described by three angles θ l, θ k, ϕ and invariant mass of the dimuon system (q 2 ) d 4 Γ dq 2 = dcos θ K dcos θ l dϕ 9 32π [ J 1s sin 2 θ K + J 1c cos 2 θ K + (J 2s sin 2 θ K + J 2c cos 2 θ K ) cos 2θ l +J 3 sin 2 θ K sin 2 θ l cos 2ϕ + J 4 sin 2θ K sin 2θ l cos ϕ + J 5 sin 2θ K sin θ l cos ϕ +(J 6s sin 2 θ K + J 6c cos 2 θ K ) cos θ l + J 7 sin 2θ K sin θ l sin ϕ + J 8 sin 2θ K sin 2θ l sin ϕ ] +J 9 sin 2 θ K sin 2 θ l sin 2ϕ, This is the most general expression of this kind of decay The CP averaged angular observables are defined: S i = J i + J i (dγ + d Γ)/dq 2 6 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 6/30 30
Transversity amplitudes One can link the angular observables to transversity amplitudes J 1s = (2 + β2 l ) 4 [ A L 2 + A L 2 + A R 2 + A R 2] + 4m2 l (A ) q 2 Re L AR L + A AR, ] J 1c = A L 0 2 + A R 0 2 + 4m2 l [ A q 2 t 2 + 2Re(A L 0 AR 0 ) + β 2 l A S 2, J 2s = β2 l 4 J 3 = 1 2 β2 l [ A L 2 + A L 2 + A R 2 + A R 2], J 2c = β 2 l [ A L 0 2 + A R 0 2], [ A L 2 A L 2 + A R 2 A R 2] [ ], J 4 = 1 β 2 l Re(A L 0 AL R + A 0 AR ), 2 J 5 = 2β l [Re(A L 0 AL R A 0 AR ) m l Re(A L q 2 A S + AR AS )], ] J 6s = 2β l [Re(A L AL R A AR m l ), J 6c = 4β l Re(A L q 2 0 A S + AR 0 AS ), J 7 = 2β l [Im(A L 0 AL R A 0 AR ) + m l ] Im(A L q 2 A S AR AS )), J 8 = 1 [ ] [ ] β 2 l Im(A L 0 AL R + A 0 AR ), J 9 = β 2 l Im(A L L A + AR R A ), 2 7 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 7/30 30
Link to effective operators So here is where the magic happens At leading order the amplitudes can be written as: [ A L,R A L,R A L,R 0 = 2Nm B (1 ŝ) (C eff 9 + C eff 9 ) (C 10 + C 10 ) + 2 ˆm b ŝ ] (Ceff 7 + C eff 7 ) ξ (E K ) [ ] = 2Nm B (1 ŝ) (C eff 9 C eff 9 ) (C 10 C 10 ) + 2 ˆm b ŝ (Ceff 7 C eff 7 ) ξ (E K ) [ ] = Nm B(1 ŝ) 2 ŝ (C eff 9 C eff 9 ) (C 10 C 10 2 ˆm ) + 2 ˆm b(c eff 7 C eff 7 ) ξ (E K ), K where ŝ = q 2 /m 2 B, ˆm i = m i /m B The ξ, are the form factors 8 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 8/30 30
Link to effective operators So here is where the magic happens At leading order the amplitudes can be written as: [ A L,R A L,R A L,R 0 = 2Nm B (1 ŝ) (C eff 9 + C eff 9 ) (C 10 + C 10 ) + 2 ˆm b ŝ ] (Ceff 7 + C eff 7 ) ξ (E K ) [ ] = 2Nm B (1 ŝ) (C eff 9 C eff 9 ) (C 10 C 10 ) + 2 ˆm b ŝ (Ceff 7 C eff 7 ) ξ (E K ) [ ] = Nm B(1 ŝ) 2 ŝ (C eff 9 C eff 9 ) (C 10 C 10 2 ˆm ) + 2 ˆm b(c eff 7 C eff 7 ) ξ (E K ), K where ŝ = q 2 /m 2 B, ˆm i = m i /m B The ξ, are the form factors Now we can construct observables that cancel the ξ form factors at leading order: P 5 J 5 + = J 5 2 (J2 c + J 2 c)(j 2 s + J 2 s) 8 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 8/30 30
Symmetries in B K µµ We have 12 angular coefficients (S i ) There exists 4 symmetry transformations that leave the angular distributions non changed: n i = Un i = [ n = ( A L A R ) e iϕ L 0 0 e iϕ R, n = ] [ cos θ sin θ ( A L ) ( A L A R, n 0 = 0 ) A R 0 sin θ cos θ ] [ cosh i θ sinh i θ sinh i θ cosh i θ Using this symmetries one can show that there are 8 independent observables The pdf can be wrote as: 1 d(γ + Γ) d(γ + Γ)/dq 2 dcosθ l dcosθ k dϕ P = 9 32π [ 34 (1 F L ) sin 2 θ k + F L cos 2 θ k + 1 4 (1 F L) sin 2 θ k cos 2θ l ] n i F L cos 2 θ k cos 2θ l + S 3 sin 2 θ k sin 2 θ l cos 2ϕ + S 4 sin 2θ k sin 2θ l cos ϕ + S 5 sin 2θ k sin θ l cos ϕ + 4 3 A FB sin 2 θ k cos θ l + S 7 sin 2θ k sin θ l sin ϕ + S 8 sin 2θ k sin 2θ l sin ϕ + S 9 sin 2 θ k sin 2 θ l sin 2ϕ ] 9 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 9/30 30
LHCb update of the B 0 K µ µ +, Selection LHCb-CONF-2015-002 PID, kinematics and isolation variables used in a Boosted Decision Tree (BDT) to reject background Reject the regions of J/ψ and ψ(2s) Specific vetos for backgrounds: Λ b pkµµ, B 0 s ϕµµ, etc Using k-fold technique and signal proxy B J/ψK for training the BDT Improved selection allowed for finer binning than the 1fb 1 analysis /c 4 ] 2 [GeV q 2 20 18 16 14 12 10 8 6 4 2 LHCb preliminary 0 52 53 54 - - m(k 55 + µ + µ ) 56 [GeV/c 57 2 ] 10 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 10/30 30 4 10 3 10 2 10 10 1
LHCb update of the B 0 K µ µ +, Selection Signal modelled by a sum of two Crystal-Ball functions Shape is defined using B J/ψK and corrected for q 2 dependency Combinatorial background modelled by exponent Kπ system: Rel Breit Wigner for P-wave Lass model for the S-wave Linear model for background Reduced the systematic compared to previous analysis In total we found 2398 ± 57 candidates in the (01, 19) GeV 2 q 2 region 624 ± 30 candidates in the theoretically the most interesting (11 60) GeV 2 region 11 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 11/30 30
Detector acceptance Detector distorts our angular distribution We need to model this effect 4D function is used: ϵ(cos θ l, cos θ k, ϕ, q 2 ) = c ijkl P i (cos θ l )P j (cos θ k )P k (ϕ)p l (q 2 ), ijkl where P i is the Legendre polynomial of order i We use up to 4 th, 5 th, 6 th, 5 th order for the cos θ l, cos θ k, ϕ, q 2 12 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 12/30 30
Control channel We tested our unfolding procedure on B J/ψK The result is in perfect agreement with other experiments and our different analysis of this decay 13 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 13/30 30
F L B 0 K µµ results 1 08 LHCb preliminary A FB 05 06 04 02 SM from ABSZ 0 0 5 10 15 2 q 2 [GeV /c 4 ] 0-05 LHCb preliminary SM from ABSZ 0 5 10 15 q 2 [GeV 2 /c 4 ] S 3 05 LHCb preliminary S 4 05 LHCb preliminary SM from ABSZ SM from ABSZ 0 0-05 0 5 10 15 q 2 [GeV 2 /c 4 ] -05 0 5 10 15 q 2 [GeV 2 /c 4 ] 14 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 14/30 30
B 0 K µµ results S 5 05 LHCb preliminary S 8 05 LHCb preliminary SM from ABSZ 0 0-05 0 5 10 15 q 2 [GeV 2 /c 4 ] -05 0 5 10 15 q 2 [GeV 2 /c 4 ] S 7 05 LHCb preliminary S 9 05 LHCb preliminary 0 0-05 0 5 10 15 q 2 [GeV 2 /c 4 ] -05 0 5 10 15 q 2 [GeV 2 /c 4 ] 15 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 15/30 30
Results in B K µµ 5 P' 1 05 0 LHCb preliminary SM from DHMV -05-1 0 5 10 15 2 q 2 [GeV /c 4 ] Tension with 3 fb 1 gets confirmed! The two bins deviate both in 28 σ from SM prediction Result compatible with previous result 16 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 16/30 30
Branching fraction measurements of B K ± µµ Despite large theoretical errors the results are consistently smaller then SM prediction 17 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 17/30 30
Branching fraction measurements of B 0 s ϕµµ Recent LHCb measurement, JHEPP09 (2015) 179 Suppressed by fs f d Cleaner because of narrow ϕ resonance 33 σ deviation in SM in the 1 6GeV 2 bin Angular part in agreement with SM (S 5 is not accessible) 18 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 18/30 30
Branching fraction measurements of Λ b Λµµ This years LHCb measurement JHEP 06 (2015) 115 In total 300 candidates in data set Decay not present in the low q 2 19 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 19/30 30
Branching fraction measurements of Λ b Λµµ This years LHCb measurement JHEP 06 (2015) 115 In total 300 candidates in data set Decay not present in the low q 2 19 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 19/30 30
Angular analysis of Λ b Λµµ For the bins in which we have > 3 σ significance the forward backward asymmetry for the hadronic and leptonic system A H F B is in good agreement with SM A l F B always in above SM prediction 20 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 20/30 30
Lepton universality test If Z is responsible for the P 5 anomaly, does it couple equally to all flavours? Challenging analysis due to bremsstrahlung Migration of events modeled by MC Correct for bremsstrahlung Take double ratio with B + J/ψK + to cancel systematics In 3fb 1, LHCb measures R K = 0745 +0090 0074 (stat)+0036 0036 (syst) Consistent with SM at 26σ Phys Rev Lett 113, 151601 (2014) 21 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 21/30 30
Angular analysis of B 0 K ee With the full data set (3fb 1 ) we performed angular analysis in 00004 < q 2 < 1 GeV 2, JHEP 04 (2015) 064 Electrons channels are extremely challenging experimentally: Bremsstrahlung Trigger efficiencies Determine the angular observables: F L, A (2) T, ARe T, AIm T : F L = A 0 2 A 0 2 + A 2 + A 2 A (2) T = A 2 A 2 A 2 + A 2 A Re T = 2Re(A LA L + A RA R ) A 2 + A 2 A Im T = 2Im(A LA L + A RA R ) A 2 + A 2, (1) 22 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 22/30 30
Angular analysis of B 0 K ee Results in full agreement with the SM Similar strength on C 7 Wilson coefficient as from b sγ decays 22 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 22/30 30
Theory implications A preliminary fit prepared by S Descotes-Genon, L Hofer, J Matias, J Virto, presented in arxiv::151004239 Took into the fit: B(B X s γ) = (336 ± 023) 10 4, Misiak et al 2015 B(B µµ), theory: Bobeth et al 2013, experiment: LHCb+CMS average (2015) B(B X s µµ), Huber et al 2015 B(B Kµµ),Bouchard et al 2013, 2015 B (s) K (ϕ)µµ, Horgan et al 2013 B Kee, B K ee and R k 23 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 23/30 30
Theory implications A preliminary fit prepared by S Descotes-Genon, L Hofer, J Matias, J Virto, presented in arxiv::151004239 The data can be explained by modifying the C 9 Wilson coefficient Overall there is around > 4 σ discrepancy wrt SM 24 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 24/30 30
If not NP? We are not there yet! There might be something not taken into account in the theory Resonances (J/ψ, ψ(2s)) tails can mimic NP effects There might be some non factorizable QCD corrections However, the central value of this effect would have to be significantly larger than expected on the basis of existing estimates DStraub, arxiv::150306199 25 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 25/30 30
If not NP? We are not there yet! There might be something not taken into account in the theory Resonances (J/ψ, ψ(2s)) tails can mimic NP effects There might be some non factorizable QCD corrections However, the central value of this effect would have to be significantly larger than expected on the basis of existing estimates DStraub, arxiv::150306199 25 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 25/30 30
There is more! There is one other LUV decay recently measured by LHCb R(D ) = B(B D τ ν) B(B D µν) Clean SM prediction: R(D ) = 0252(3), PRD 85 094025 (2012) LHCb result: R(D ) = 0336 ± 0027 ± 0030, HFAG average: R(D ) = 0322 ± 0022 39 σ discrepancy wrt SM 26 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 26/30 30
Conclusions Clear tensions wrt SM predictions! Measurements cluster in the same direction We are not opening the champagne yet! Still need improvement both on theory and experimental side Time will tell if this is QCD+fluctuations or new Physics: 27 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 27/30 30
Conclusions Clear tensions wrt SM predictions! Measurements cluster in the same direction We are not opening the champagne yet! Still need improvement both on theory and experimental side Time will tell if this is QCD+fluctuations or new Physics: when you have eliminated all the Standard Model explanations, whatever remains, however improbable, must be New Physics prof Joaquim Matias 27 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 27/30 30
Thank you for the attention! 28 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 28/30 30
Backup 29 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 29/30 30
Theory implications 29 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 29/30 30
If not NP? How about our clean P i observables? The QCD cancel as mentioned only at leading order Comparison to normal observables with the optimised ones 30 / Marcin Chrząszcz (Universität Zürich, IFJ PAN) Electroweak penguin measurements 30/30 30