Anomalies and deviations in heavy-flavour physics @GreigCowan - - PowerPoint PPT Presentation

SLIDE 1

Anomalies and deviations in heavy-flavour physics

@GreigCowan (Edinburgh)

Birmingham, Dec 2nd 2015

SLIDE 2

• Introduction to the LHCb experiment
• b → sl+l− FCNC decays
• Lepton (non-)universality
• CP violation in the beauty + charm systems

2 / 60

SLIDE 3

The LHC

https://ideas.lego.com/projects/94885

3 / 60

SLIDE 4

The LHCb detector

2008 JINST 3 S08005 4 / 60

Covers 4% of solid angle, but accepts 40% of heavy quark production cross section.

SLIDE 5

A typical LHCb event

[2008 JINST 3 S08005] b ~1cm p p b

nPV s ∼ 2.0 nTracks ∼ 200 σ(pp → bbX) ∼ 80µb σ(cc) ∼ 1500µb

HLT2 DiMuon trigger 5 / 60

SLIDE 6

Run-1 data sample

∼900 physicists from 64 universities/laboratories in 16 countries. O(100k) bb pairs produced/sec. 2010 2011 (1 fb−1 @7TeV) 2012 (2 fb−1 @8TeV) Efficiency > 93%

LHCb designed to run at lower luminosity than ATLAS/CMS.

LHCb tracking/PID is sensitive to pile-up.

LHC pp beams are displaced to reduce instantaneous luminosity - stable running conditions. L2011 ∼ 2.7 × 1032cm−2s−1 L2012 ∼ 4.0 × 1032cm−2s−1

6 / 60

SLIDE 7

Searching for New Physics

ON-SHELL OFF-SHELL Cannot produce particles Higher energy particles can with mc2 > E appear virtually in quantum loops → flavour physics NP?

History: top quark mass predicted by quark mixing

7 / 60

SLIDE 8

Rare (FCNC) b-hadron decays

8 / 60

SLIDE 9

b → s transitions

b → s “penguin” decays are loop/CKM suppressed. FCNC can be crucial to finding out where to look for NP. Model independent effective Hamiltonian, where heavy degrees of freedom have been integrated out in short-distance Wilson coefficients, (Ci). Heff = − 4GF √ 2 VtbV ∗

ts

αe 4π

• i
• CiOi + C′

iO′ i

• [Blake, Gershon, Hiller, Annu. Rev. Nucl. Part. Sci. 2015]

B0 → K∗(892)0µ+µ−

O9(′) = [sγµPL(R)b][lγµl] q2 ≡ m(l+l−)2

9 / 60

SLIDE 10

B0 → K∗(892)0µ+µ−

[LHCb-PAPER-2015-051]

]

2

c ) [GeV/ µ

+

µ

• +

K ( m

5.2 5.3 5.4 5.5 5.6 5.7

]

4

c /

2

[GeV

2

q

2 4 6 8 10 12 14 16 18 20

1 10

2

10

3

10

4

10

LHCb

• Ω ≡ (cos θl, cos θK, φ)

2398 ± 57 events, excluding the charmonia. Di-muon final state is experimentally clean signature, but BR ∼ 10−7. P → V V ′ decay, fully described by q2 ≡ m(µ+µ−)2 and 3 helicity angles. B0 → K∗µ+µ− has rich system of observables (rates, angles, asymmetries) that are sensitive to NP. d4Γ[B0 → K∗0µ+µ−] dq2 d Ω = 9 32π

11

• j=1

Ij(q2)fj( Ω), Ij → Ij for B0 Sj =

• Ij + ¯

Ij dΓ dq2 + d¯ Γ dq2

• ,

Aj =

• Ij − ¯

Ij dΓ dq2 + d¯ Γ dq2

• 10 / 60

SLIDE 11

B0 → K∗(892)0µ+µ−

[LHCb-PAPER-2015-051] ]

2

c ) [MeV/

−

µ

+

µ

−

π

+

K ( m

5200 5400 5600

2

c Events / 5.3 MeV/

50 100 LHCb

4

c /

2

< 6.00 GeV

2

q 1.10 <

]

2

c ) [GeV/

−

π

+

K ( m

0.8 0.85 0.9 0.95

2

c Events / 10 MeV/

50 100 LHCb

4

c /

2

< 6.00 GeV

2

q 1.10 <

Describe m(Kπ) with Breit-Wigner for P- wave and LASS for S- wave K+π−

l

θ cos

• 1
• 0.5

0.5 1

Events / 0.1

50 100 LHCb

4

c /

2

< 6.00 GeV

2

q 1.10 < K

θ cos

• 1
• 0.5

0.5 1

Events / 0.1

50 100 LHCb

4

c /

2

< 6.00 GeV

2

q 1.10 <

[rad] φ

• 2

2

π Events / 0.1

50 100 LHCb

4

c /

2

< 6.00 GeV

2

q 1.10 <

Si, Ai’s extracted using a max likelihood fit. Example fits in ±50 MeV /c2 around K∗(892)0. For the first time the Kπ S-wave is accounted for.

11 / 60

SLIDE 12

B0 → K∗(892)0µ+µ−: some observables

[LHCb-PAPER-2015-051]

]

4

c /

2

[GeV

2

q

5 10 15

L

F

0.2 0.4 0.6 0.8 1

LHCb

SM from ABSZ

S1c ≡ FL

]

4

c /

2

[GeV

2

q

5 10 15

5

S

• 0.5

0.5

LHCb

SM from ABSZ

S6s ≡ 4

3 AFB

]

4

c /

2

[GeV

2

q

5 10 15

3

A

• 0.5

0.5

LHCb ]

4

c /

2

[GeV

2

q

5 10 15

4

A

• 0.5

0.5

LHCb + many other observables not shown

Some observables have physical boundaries ⇒ use Feldman-Cousins for uncertainties. CP-asymmetries consistent with zero, as expected, but some deviations in CP-averaged observables (the Sj’s).

12 / 60

SLIDE 13

B0 → K∗(892)0µ+µ−: the anomaly

[LHCb-PAPER-2015-051]

“Theoretically clean” observables less dependent on hadronic form factors

[Descotes-Genon et al JHEP 05 (2013) 137].

These divide out the hadronic uncertainties to leading order. Tension from the 1 fb−1 LHCb result remains. P ′

i=4,5,6,8 =

Sj=4,5,7,8

• FL(1 − FL)

]

4

c /

2

[GeV

2

q

5 10 15

5

' P

• 1
• 0.5

0.5 1

LHCb

SM from DHMV 2.8σ, 3.0σ from SM A χ2 fit to all CP-averaged

• bservables shows a 3.4σ

shift from SM prediction

13 / 60

SLIDE 14

b → sµ+µ−branching fractions lower than predictions

[JHEP 06 (2014) 133 ]

B0 → K∗(892)0µ+µ−

[JHEP 08 (2013) 131]

B0

s → φµ+µ−

[JHEP 09 (2015) 179]

Λb → Λµ+µ− (Bham)

[JHEP 06 (2015) 115] 14 / 60

SLIDE 15

Observation of B0

s → µ+µ−

CKM suppressed and helicity suppressed ((mµ/mB)2). B(B0

s → µµ)SM = (3.66 ± 0.23) × 10−9

B(B0 → µµ)SM = (1.06 ± 0.09) × 10−10

[PRL 112, 101801 (2014)] Dominant uncertainty will be improved via refined Lattice QCD calcs.

Sensitive to scalar and pseudoscalar NP couplings, e.g., in MSSM B ∝ (tan β)6

d

B0

s → µ+µ−

B0

s

✁

W + W − Z0 t b s µ+ µ−

f

B0

s → µ+µ−

B0

s

✁

X+ W − X0 t b s µ+ µ−

30 years of effort!

15 / 60

SLIDE 16

Observation of B0

s → µ+µ−

[CMS + LHCb, Nature 522, 68-72 (2015)]

]

2

c [MeV/

−

µ

+

µ

m

5000 5200 5400 5600 5800

)

2

c Candidates / (40 MeV/

2 4 6 8 10 12 14 16 Data Signal and background

−

µ

+

µ →

s

B

−

µ

+

µ → B Combinatorial bkg. Semileptonic bkg. Peaking bkg.

CMS and LHCb (LHC run I)

B0

s 6.2σ

B0 3.0σ

Use multi-variate techniques to suppress background. Results consistent with SM at ∼ 2σ. Constrains S and P contributions. One to watch during LHC Run-2.

16 / 60

SLIDE 17

Global fits for Wilson coeffs [Descotes-G et al, arXiv:1510.04239]

2D fit with (CNP

9

, CNP

9′ ) floating

→ 4.5σ deviation from SM Other global fits exist! Inputs from branching fractions and angular observables from b → sll decays, BR(B → Xsγ), BR(B0

s → µ+µ−),. . . .

Many fits performed with different subsets of the observables and different theoretical inputs (form factors, power corrections, charm loops). CNP

9

< 0 plays central role explaining many deviations seen in b → sll transitions. Possible Z′? Leptoquarks? [many authors] How well do we understand QCD-effects? [Lyon, Zwicky]

17 / 60

SLIDE 18

Lepton universality

RK ≡ B(B+→K+µ+µ−)

B(B+→K+e+e−) ,. . .

18 / 60

SLIDE 19

Lepton universality (B+ → K+l+l−)

[PRL 113,151601 (2014)]

In the SM only the Higgs boson has non-universal lepton couplings. This results in SM predictions of ∼unity for various decay-rate ratios. RK ≡ B(B+→K+µ+µ−)

B(B+→K+e+e−) SM

= 1 ± O(10−2) 2.6σ deviation Can be described assuming NP only in b → sµµ. Very interesting given indications of non-SM physics in other b → sµµ FCNC decays and 2.4σ excess in H → τµ at CMS [PLB 749 (2015) 337]. Future: Make similar measurements using other decays - R(φ), R(K∗), R(Λ) (Bham).

19 / 60

SLIDE 20

Lepton universality (B0 → D∗+lν)

CKM mechanism well tested, but room for NP if coupling more to 3rd generation (e.g., charged Higgs). B-factories already reporting deviation from theoretically clean SM prediction. Form-factors cancel in the ratio.

Tree-level int., unlike b → sll FCNC

R(D∗) ≡ B(B0→D∗+τντ )

B(B0→D∗+µνµ)

Interesting given hints of non-universality in B+ → K+l+l− decays (RK) and excl/incl measurements of Vub, Vcb.

20 / 60

SLIDE 21

Lepton universality (B0 → D∗+lν)

[PRL 115, 111803 (2015)]

Very challenging measurement at hadron collider (no beam constraints and large backgrounds). B(τ → µνµντ) = (17.41 ± 0.04)% Signal and normalisation have same final state particles. Large samples of events, triggering on charm. Require significant B, D, τ flight distances. Use isolation MVA. Template fit to kinematic variables → R(D∗) ≡ B(B0→D∗+τντ )

B(B0→D∗+µνµ)

) /c (GeV

miss

m )

4

/c

2

(GeV

miss 2

m

• 2

2 4 6 8 10 Pulls

• 2

2

• 2

2 4 6 8 10 1000 2000 3000 4000 LHCb

4

/c

2

< 12.60 GeV

2

9.35 < q )

4

/c

2

Candidates / (0.3 GeV Candidates / (75 MeV) 10 10

* (MeV)

µ

E * (MeV)

µ

E

500 1000 1500 2000 2500 Pulls

• 2

2 500 1000 1500 2000 2500 1000 2000 3000 4000 LHCb

4

/c

2

< 12.60 GeV

2

9.35 < q Candidates / (75 MeV) 21 / 60

SLIDE 22

Lepton universality (B0 → D∗+lν)

[PRL 115, 111803 (2015)]

LHCb R(D∗) = 0.336 ± 0.027 ± 0.030 (2.1σ from SM)

R(D)

0.2 0.3 0.4 0.5 0.6

R(D*)

0.2 0.25 0.3 0.35 0.4 0.45 0.5

BaBar, PRL109,101802(2012) Belle, arXiv:1507.03233 LHCb, arXiv:1506.08614 Average

= 1.0

2

χ ∆

SM prediction ) = 55%

2

χ P(

HFAG

• Prel. EPS2015

3.9σ from SM SM prediction from [PRD 85 (2012) 094025]. Could be explained by enhancement of bL → cLτLνL amplitude. Now using other decay modes to make similar measurements (R(D(s)), R(Λc)).

22 / 60

SLIDE 23

CP violation in the quark sector

23 / 60

SLIDE 24

CP violation in the quark sector

VCKM =   Vud Vus Vub Vcd Vcs Vcb Vtd Vts Vtb   =   1 − λ2/2 λ Aλ3(¯ ρ − i¯ η) −λ 1 − λ2/2 Aλ2 Aλ3(1 − ¯ ρ − i¯ η) −Aλ2 1  +O(λ4)

Wolfenstein parameterisation

3 generations + 1 phase → ¯ η = 0 is

• nly source of CP violation in SM.

CKM picture confirmed up to ∼ 20%. Couplings show strong hierarchy not seen in lepton sector ⇒ “SM flavour puzzle” New Physics should have flavour structure similar to SM. . . . . . or the NP scale is very very large (∼ 100TeV) ⇒ “NP flavour puzzle” Need more precision measurements to look for small deviations.

24 / 60

SLIDE 25

3 types of CP violation

1 Mixing: |q/p| = 1 2 Decay: |Af/Af| = 1 3 Interference between mixing and

decay: φd,s ≡ −arg(λf) ≡ −arg

• q

p Af Af

• = 0

Expect |λf| ≡

• q

p Af Af

• ≈ 1

NP?

|B0

s,L

= p|B0

s + q|B0 s

|B0

s,H

= p|B0

s − q|B0 s

ACP (t) ≡ ΓB

0→f − ΓB0→f

ΓB

0→f + ΓB0→f

= Sf sin(∆m t) − Cf cos(∆m t) cosh(∆Γ t/2) + A∆Γ sinh(∆Γ t/2)

25 / 60

SLIDE 26

CP violation in B0 → J/ψK0

S

[PRL 115 (2015) 031601] 42k tagged events AC

P (t) = SJ/ ψ K0 S

sin(∆mdt) − CJ/

ψ K0 S

cos(∆mdt) SJ/

ψ K0 S

≈ sin 2β SJ/

ψ K0 S

= +0.731 ± 0.035 ± 0.020 CJ/

ψ K0 S

= −0.038 ± 0.032 ± 0.005

Similar precision to B-factories, but LHCb measurement pulled WA up towards indirect determination from global fit. sin 2βWorld Average = 0.691 ± 0.017 sin 2βCKMfitter = 0.748+0.030

−0.032 26 / 60

SLIDE 27

B+ → τ +ντ vs. sin 2β

Winter 2012 Summer 2015

Small tension reduced following: Updated measurement of sin 2β and new measurement of B(B+ → τ +ντ) from Belle

[arXiv:1503.05613].

CKM predictions also changed a bit.

27 / 60

SLIDE 28

CP violation in B0

s → J/ψφ

|B0

s,L

= p|B0

s + q|B0 s

|B0

s,H

= p|B0

s − q|B0 s

∆Γs ≡ ΓL − ΓH φs

SM

= −0.0365 ± 0.0012 rad

[CKMFitter]

σ(φs) ∼ ±0.4 rad

28 / 60

SLIDE 29

CP violation in B0

s → J/ψφ

|B0

s,L

= p|B0

s + q|B0 s

|B0

s,H

= p|B0

s − q|B0 s

∆Γs ≡ ΓL − ΓH φs

SM

= −0.0365 ± 0.0012 rad

[CKMFitter]

Combination

φs = −0.034 ± 0.033 rad ∆Γs = 0.082 ± 0.006 ps−1

Dominated by LHCb [PRL 114 (2015) 041801]

New physics not large. ⇒ need to control SM effects (penguins). Also competitive in gluonic penguin decays (B0

s → φφ). 29 / 60

SLIDE 30

Penguin pollution in φs and sin 2β

30 / 60

SLIDE 31

Controlling penguins pollution

φmeasured

q

= φq + δPenguin + δNew Physics

Enhancement could be caused by non- perturbative hadronic effects that are difficult to calculate in QCD.

[Nierste et al. arXiv:1503.00859], [Liu et al. PRD 89, 094010 (2014)]

1 Measure φs/sin 2β for different polarisation states. 2 Measure δPenguin using decays where penguin/tree ratio is not suppressed. Use SU(3)-flavour relations to link B0

s and B0 (broken at 20-30% level). A(B0

s → (J/ψ φ)f ) =

(1 − λ2/2)A′

f

• 1 + ǫa′

f e iθ′ f eiγ

• Penguin/tree suppressed by ǫ =

|Vus|2 1−|Vus|2 = 0.05

s B0

s

h+h− s J/ψ c b W + c s h− /ψ s B0

s

h+h− s J/ψ c b u, c, t c W + s A(B0

s → (J/ψ K∗0)f ) =

− λAf

• 1 − af eiθf eiγ

Penguin/tree not suppressed (but overall rate suppressed) [Faller et al. PRD 79, 014005 (2009)] [De Bruyn, Fleischer, JHEP1503 (2015) 145] 31 / 60

SLIDE 32

Controlling penguins with B0

s → J/ψK∗0 + B0 → J/ψρ0

Parameter Fitted value ∆φJ/ψ φ

s,0

0.000+0.009

−0.011(stat)+0.004 −0.009(syst)

∆φJ/ψ φ

s,

0.001+0.010

−0.014(stat)+0.007 −0.008(syst)

∆φJ/ψ φ

s,⊥

0.003+0.010

−0.014(stat)+0.007 −0.008(syst) [LHCb-PAPER-2015-034]

Penguin parameters effectively constrained from CP asymmetry measurements. Combined results dominated by B0 → J/ψ ρ0 (access to mixing-induced asymmetry not available in flavour-specific B0

s → J/ψ K∗0 channel).

Penguins are small!

32 / 60

SLIDE 33

CP violation in B0

(s) mixing

(|B0

L,H = p|B0 ± q|B0)

[PRL 114, 081801 (2015)]

ad

sl

as

sl Use semileptonic B0, B0

s decays

asl ≡ Γ(B → B → f) − Γ(B → B → f) Γ(B → B → f) + Γ(B → B → f) ameas(t) = N(f,t)−N(f,t)

N(f,t)+N(f,t) = asl 2

• 1 −

cos(∆mt) cosh(∆Γt/2)

• [Lenz arXiv:1205.1444] - tiny in SM

ad

sl = (−4.1 ± 0.6) × 10−4

as

sl = (+1.9 ± 0.3) × 10−5

ad

sl = −0.0015 ± 0.0017 [HFAG]

as

sl = −0.0075 ± 0.0041 [HFAG]

No tagging needed. Typically time-dep. measurement for B0 system, indep. for B0

s.

Crucial to control production and detection asymmetries using control samples. ∼ 3σ tension with SM from D0 dimuon asymmetry not confirmed or excluded by other experiments. Explanation of D0 dimuon could be due to deviation in value of ∆Γd [PRD 87 074020

(2013)]. 33 / 60

SLIDE 34

New physics in B mixing

[Lenz et al. arXiv:1203.0238v2 + updates]

Introduce generic NP through complex parameter ∆q: MNP,q

12

= MSM,q

12

∆q NP contribution to B0

s mixing is limited to < 30% at 3σ.

But beware of hadronic uncertainties that could mimic small NP. Take-home message: will significantly shrink these contours with Run-2 data and probe BSM contributions @ few % of SM.

34 / 60

SLIDE 35

|Vub| from the CKM unitarity triangle

|Vub| indispensible in CKM unitarity fits. Excellent test of unitarity (and/or NP) by comparing |Vub| (tree-level process) with sin 2β (B0-mixing, loop process).

35 / 60

SLIDE 36

|Vub|

Measure exclusive branching fraction (B0 → πlν, B+ → τντ). Or inclusive sum of states (b → ulν). Each method relies on different theoretical inputs. Long-standing discrepancy between these two approaches using results from BaBar/Belle. Exclusive dominated by B → πlν

36 / 60

SLIDE 37

|Vub| using Λb → pµνµ

Challenging at hadron collider to separate b → uµν and b → cµν processes without beam energy constraint of e+e− machine. Large production of Λb baryons at LHC. Cleaner than B → πlν due to protons in final state. LHCb [JHEP 08 (2014) 143] Aside on b-baryons: No CP violation in the baryon system observed. This is an area where only LHC experiments (particularly LHCb) can contribute.

37 / 60

SLIDE 38

|Vub| using Λb → pµνµ

[Nature Physics 10 (2015) 1038]

To cancel many systematic uncertainties we measure the branching ratio relative to Λb → Λcµνµ, Λc → pKπ. ⇒ Must use global |Vcb| average as input. Lattice QCD input is crucial [Meinel

arXiv:1503.01421].

Fit corrected mass (peaks at m(Λb)) |Vub|2 |Vcb|2 = B(Λb→pµν)q2>15 Ge

V

B(Λb→Λcµν)q2>7 Ge

V RFF

mcorr =

• m2

hµ + p2 T + pT ∼ 18k Λb → pµνµ ∼ 34k Λb → Λcµνµ |Vub| = (3.27 ± 0.15(stat) ± 0.17(syst) ± 0.06(theory)) × 10−3

38 / 60

SLIDE 39

|Vub| using Λb → pµνµ

[Nature Physics 10 (2015) 1038]

|V |

0.003 0.0035 0.004 0.0045 0.005 Inclusive

PDG 2014

Exclusive ) ν l π → (B

PDG 2014 arXiv:1501.05373 RBC/UKQCD arXiv:1503.07839 FNAL/MILC

LHCb ) ν µ p → Λ (

arXiv:1503.01421 Detmold, Lehner, Meinel (using RBC/UKQCD config) b

ub

3.5σ from inclusive

• Syst. limited from Lattice QCD calc. of Λb form-factor (more precise at high q2).

Λb → pµν has different dependence on right-handed currents than other modes. Combination starts to disfavour interpretation of discrepancy in terms of quantity of RHC (ǫR).

39 / 60

SLIDE 40

Charm physics

Only way to study FCNC with u-type quarks. Allows to probe higher energy scales than b decays. Look at time-integrated CP asymmetries. Expect to be small. LHCb measurement of ∆ACP = 0 in 2012 [PRL 108 (2012) 111602]. Wow! Situation now less certain following updates - stay tuned. . .

ACP = Γ(D

0→f)−Γ(D0→f)

Γ(D

0→f)+Γ(D0→f)

∆ACP ≡ ACP (K+K−) − ACP (π+π−)

40 / 60

SLIDE 41

Charm physics

Huge event yields have led to huge progress in CP violation in charm mixing and rare decays. LHCb will take advantage of higher cross-section and new trigger configuration in Run-2.

0.010 0.005 0.000 0.005 0.010

a ind

CP 0.010 0.005 0.000 0.005 0.010

∆a dir

CP

Contours contain 68%, 95%, 99% CL

CDF LHCb prel. LHCb SL BaBar Belle prel. CDF LHCb SL LHCb KK LHCb ππ

no CPV BaBar Belle CDF LHCb HFAG-charm

Winter 2015

aind

C P = (0.058 ± 0.040)%

∆adir

C P = (−0.257 ± 0.104)%

AΓ ≡ τ(D0→h+h−)−τ(D0→h+h−)

τ(D0→h+h−)+τ(D0→h+h−) ≈ −aind CP − adir CP yCP

∆ACP ≈

• 1 + t

τ yCP

• ∆adir

CP + ∆t τ

armind

CP 41 / 60

SLIDE 42

Pentaquarks

[PRL 115 (2015) 072001]

Two pentaquark states observed in Λb → J/ψ pK− 6D amplitude fit performed (coherent sum of resonant states). Fit quality insufficient if only using Λ∗ → pK resonances. Need two Pc states of opposite parity.

Pc(4380)+ Pc(4450)+ JP

3 2 − 5 2 +

Mass [ MeV /c2 ] 4380 ± 8 ± 29 4449.8 ± 1.7 ± 2.5 Width [ MeV /c2 ] 205 ± 18 ± 86 39 ± 5 ± 19 Significance 9σ 12σ Behaves like a resonance Expected Breit-Wigner

Prospect first raised 50 years ago by Gell-Mann, Zweig. LHCb states have quark content ccuud

42 / 60

SLIDE 43

Summary

Exciting indications of non-SM physics in B physics. Crucially, these are in related channels: R(D∗), RK, P ′

5, b → s penguin branching

ratios, (H → τµ). More measurements and theory developments needed to interpret what we are seeing. CKM mechanism holding up to scrutiny, need more precision. Most results statistically limited → looking forward to Run-2 of LHC and start-up of Belle-II ∼2018.

]

4

c /

2

[GeV

2

q

5 10 15

5

' P

• 1
• 0.5

0.5 1

LHCb

SM from DHMV

43 / 60

SLIDE 44

Visualising the P ′

5 discrepancy

44 / 60

SLIDE 45

Run-2 status

97% efficiency 25ns operation

Huge success so far! New trigger configuration commissioned. Offline reconstruction in the trigger! Online calibration + alignment allows physics analyses directly from the trigger. Only tracks and vertices that caused event to trigger are saved (no offline reco). Used for high yield samples (J/ψ , D0, D+ . . . )

45 / 60

SLIDE 46

Run-2 data flow

46 / 60

SLIDE 47

First results at √s = 13 TeV

[arXiv:1509.00771]

J/ψ → µ+µ−

[TeV] s

5 10

b] µ [ σ

1 2 3 4

• b
• from

ψ J/ LHCb FONLL σ 1 ± FONLL,

• b
• from

ψ J/ LHCb FONLL σ 1 ± FONLL,

σ(pp → bbX) = 518 ± 2 ± 53 µb [LHCB-PAPER-2015-041]

charm

σ(pp → ccX) = 2.72 ± 0.01 ± 0.18 ± 0.14 mb (in LHCb)

New trigger and automatic calibration/alignment validated with early measurements (mainly 50ns ramp). First results with Run-2 data! J/ψ and charm cross-sections agree with expectations.

47 / 60

SLIDE 48

Angular analysis of B0

s → J/ψK∗0

NEW!

NB0 = 208700 ± 500 NB0

s = 1800 ± 60

Angular analysis performed in 4 bins around K∗(892)0 → K+π− mass, for B0

s

and B0

s.

Use simulation to get angular efficiency correction (+ correction for lack of S-wave in MC). Account for production and detection asymmetries [PRL 114 (2015) 041601],[PLB

739 (2014) 218], [JHEP 07 (2014) 041].

Parameter Fitted value f0 0.497 ± 0.025 ± 0.025 f 0.179 ± 0.027 ± 0.013 ACP −0.048 ± 0.057 ± 0.020 ACP

• 0.171 ± 0.152 ± 0.028

ACP

⊥

−0.049 ± 0.096 ± 0.025 B(B0

s → J/ψ K∗0) =

(4.13±0.16±0.25±0.24(fs/fd))×10−5

[LHCb-PAPER-2015-034] 48 / 60

SLIDE 49

Controlling penguins with B0

s → J/ψK∗0 NEW!

Use results from angular analysis and branching fraction of B0

s → J/ψ K∗0 to

measure ∆φJ/ψ φ

s,i

for each polarisation i ∈ (0, ⊥, , S). Hi ∝ 1 ǫ

• A′

i

Ai

• 2 B(B0

s → J

/ ψK∗0) B(B0

s → J

/ ψφ) fi f′

i

= 1 − 2ai cos θi cos γ + a2

i

1 + 2ǫa′

i cos θ′ i cos γ + ǫ2a′2 i

ACP

i

= − 2ai sin θi sin γ 1 − 2a′

i cos θ′ i cos γ + a′2 i

SU(3): ai = a′

i, θi = θ′ i

• A′

i

Ai

• computed

with LCSR

[Barucha et al, arXiv:1503.05534]

γ = 73 ± 7◦ [CKM]

[LHCb-PAPER-2015-034] LHCb preliminary

Extract penguin parameters from χ2 fit to Hi and ACP

i

information for each polarisation i ∈ (0, ⊥, , S). Translate to penguin phase shift: Param. Value ± (stat) ± (syst) ±(|A′

i/Ai|)

∆φJ/ψ φ

s,0

0.001+0.087

−0.011 +0.013 −0.008 +0.048 −0.030

∆φJ/ψ φ

s,

0.031+0.049

−0.038 +0.013 −0.013 +0.031 −0.033

∆φJ/ψ φ

s,⊥

−0.046+0.012

−0.012 +0.007 −0.008 +0.017 −0.024

Compare to current experimental precision: σ(φs) = ±0.035 rad, σ(φd) = ±0.028 rad

49 / 60

SLIDE 50

CP violation in B0 → J/ψρ0(770)

PLB 742 (2015) 38-49

) [MeV]

• π

+

π ψ m(J/

5300 5400 5500

Combinations/ (5 MeV)

1000 2000 3000 4000 5000 6000 7000

LHCb B0 → J / ψKπ B0 → J / ψππ ∼ 18k B0

s → J

/ ψππ Sidebands used for bkg modelling K0

S

veto

[PRD 90, 012003 (2014)]

Use ρ0(770) component to measure: φeff

d

= (41.7 ± 9.6+2.8

−6.3)◦, αCP ≡ 1−|λf | 1+|λf | = (−32 ± 28+9 −7) × 10−3

⇒ ∆φd = (−0.9 ± 9.7+2.8

−6.3)◦

(equivalent to 0.016 ± 0.169+0.049

−0.110 rad)

50 / 60

SLIDE 51

Controlling penguins with B0

s → J/ψK∗0 + B0 → J/ψρ0

NEW!

[LHCb-PAPER-2015-034]

Now fit for |A′

i/Ai| to limit

sensitivity to hadronic uncertainties. Assume |A′

i/Ai|(B0 s → J/ψ K∗0) =

|A′

i/Ai|(B0 → J/ψ ρ0)

Parameter Fitted value ∆φJ/ψ φ

s,0

0.000+0.009

−0.011(stat)+0.004 −0.009(syst)

∆φJ/ψ φ

s,

0.001+0.010

−0.014(stat)+0.007 −0.008(syst)

∆φJ/ψ φ

s,⊥

0.003+0.010

−0.014(stat)+0.007 −0.008(syst)

Penguin parameters effectively constrained from CP asymmetry measurements. Combined results dominated by B0 → J/ψ ρ0 (access to mixing-induced asymmetry not available in flavour-specific B0

s → J/ψ K∗0 channel).

Penguins are small!

51 / 60

SLIDE 52

The “Kπ puzzle”

∆ACP ≡ ACP (B+ → K+π0) − ACP (B0 → K+π−) = 0.12 ± 0.02 Naively expect ∆ACP = 0. NP in electroweak penguin loop or QCD effect? Need isospin analysis to understand what is going on (e.g., sum rule proposed by

[Gronau, PLB 627 (2005) 82-88]).

B0 → K+π− measured at BaBar, Belle, CDF, LHCb. B+ → K+π0 at BaBar/Belle. B+ → K+π0 challenging at LHCb (no secondary vertex + photons in final state) but possible [LHCb-CONF-2015-001]. Expect Belle-II to make significant improvements here (including B0 → K0π0).

[Belle, Nature 452, 332-335 (2008)] 52 / 60

SLIDE 53

B0

s → φµ+µ− (φ → K+K−)

[JHEP 09 (2015) 179]

Differential branching fraction and angular analysis (using max likelihood fit). Angular observables in good agreement with SM. dB/dq2 in q2 ∈ [1, 6] GeV /c2 lower than SM by 3.2σ.

Similar story in B+ → K+µ+µ− and B0 → K∗0µ+µ−. SM pred. and wide from [arXiv:1503.05534] SM LQCD from [PRL112(2014)21200 53 / 60

SLIDE 54

Λb → Λµ+µ− (Λ → pπ−)

[JHEP 06 (2015) 115]

Differential branching fraction and first angular analysis (using max likelihood fit). Evidence for decay in first q2 bin, but not in q2 ∈ [1.1, 6] GeV /c2 ⇒ lower than SM. Some angular observables in good agreement with SM, others not. e.g. →

SM prediction [PRD87(2013)074502] 54 / 60

SLIDE 55

B+ → π+µ+µ−

[JHEP 10 (2015) 034]

First measurement of differential branching fraction and CP asymmetry in b → dll transition. dB/dq2 compatible with SM but on the low side.

APR13 [PRD89(2014)094021] HKR15 [arXiv:1506.07760] FNAL/MILC15 [arXiv:1507.01618] 55 / 60

SLIDE 56

Sensitivity prospects

LHCb-PUB-2014-040

Before upgrade. After upgrade. Current theory uncertainty.

56 / 60

SLIDE 57

New physics prospects

[J. Charles et al. PRD 89, 033016 (2014)]

Assume that NP only enters B0 and B0

s mixing: Md,s 12 = (Md,s 12 )SM(1 + hd,se2iσd,s).

Now NP < 30% SM ∼2025 NP < 5% SM h ≈

|Cij|2 |V ∗

tiVtj|2

• 4.5 TeV

Λ

2

57 / 60

SLIDE 58

Controlling penguin pollution in φs

s B0

s

h+h− s J/ψ c b W + c s s B0

s

h+h− s J/ψ c b u, c, t c W + s

Penguin-to-tree suppression: ǫ =

|Vus|2 1−|Vus|2 = 0.05

φmeasured

s

= φs + δPenguin + δNew Physics

Difficult-to-calculate non-perturbative hadronic effects could lead to big enhancement. Measure δPenguin using decays where penguin/tree ratio is enhanced.

[Faller et al. arXiv:0810.4248, De Bruyn & Fleischer, arXiv:1412.6834]

Use SU(3) relations to link B0

s and

B0 (broken at level of 20-30%). |δP| < 1.8◦ c.f. σ(φs) = ±2.0◦, σ(φd) = ±1.4◦ K0

S

veto B0 → J/ψ π+π− [PRD 90, 012003 (2014)]

58 / 60

SLIDE 59

CP violation in charmless B0

s decays [PRD 90 (2014) 052011]

B0

s → φφ: b → s penguin decays sensitive

to NP in the loops. φ → KK: 5 different polarisation amplitudes ⇒ angular analysis. φs = −0.17 ± 0.15 ± 0.03 rad. |λ| = 1.04 ± 0.07 ± 0.03 ⇒ no direct CPV.

b s s s s s W + u,c,t s

SM: |φs| < 0.02 rad

)

2

c Candidates / (4.6 MeV/ 1 10

2

10

2012

LHCb

]

2

c [MeV/

−

K

+

K

−

K

+

K

m 5250 5300 5350 5400 5450 Pull

• 4
• 2

2 4

∼ 4k signal Decay time resolution

Current σt ∼ 46 fs ⇒ D ≈ 0.7 Upgrade σt ∼ 37 fs ⇒ D ≈ 0.8 59 / 60

SLIDE 60

Tree-level measurement of γ

Least well known of the CKM angles. Can be measured entirely from tree decays where there is small residual theory uncertainty |δγ| ≤ O(10−7) [Brod,

Zupan JHEP 1401 (2014) 051]

Use interference between B± → D0K±, D0 → f decay amplitudes. Time-independent B± → D0K± and B0 → DK∗ . . . . . . or time-dependent B0

s → D+ s K (γ − 2βs) ]

4

c /

2

[GeV

2 +

m

1 2 3

]

4

c /

2

[GeV

2 −

m

1 2 3

LHCb

[JHEP 10 (2014) 097] B+ → [K0

Sπ+π−]DK+

]

4

c /

2

[GeV

2 −

m

1 2 3

]

4

c /

2

[GeV

2 +

m

1 2 3

LHCb

B− → [K0

Sπ+π−]DK−

γ = (73+9

−10)◦ [LHCb-CONF-2014-004] B → DK only

Best precision comes from combining many independent decay modes. B-factories: σ(γ) ∼ 15◦; Final LHCb Run-1: σ(γ) ∼ 7◦.

60 / 60