Charmless B decays at LHCbs! !! decay[12]: a) Gluonic penguin, b) singlet penguin, c) colour allowed...

B0(s)→ φφ branching ratio First observation of B0

s → η′η′ First observation of B0→ ρ0ρ0 Summary

Charmless B decays at LHCb

Lars Eklund, on behalf of the LHCb collaboration

University of Glasgow

The 16th International Conference on B-physics at Frontier Machines5 May 2016

Marseille, France

L. Eklund (Glasgow) Charmless B decays at LHCb 5 May 2016 1 / 21



Outline

1 B0(s)→ φφ branching ratioJHEP 10 (2015) 053

2 First observation of B0s→ η′η′

Phys. Rev. Lett. 115 (2015) 051801

3 First observation of B0→ ρ0ρ0

Phys. Lett B747 (2015) 468

4 Summary


http://link.springer.com/article/10.1007/JHEP10(2015)053

http://journals.aps.org/prl/abstract/10.1103/PhysRevLett.115.051801

http://www.sciencedirect.com/science/article/pii/S0370269315004499



Introduction

Why charmless B decays?

|Vub| is small: significant loop contributions

FCNC: forbidden at tree level

Sensitive to contributions beyond the SM

Why LHCb?

Full spectrum of B hadrons: B0, B0s , B+,

B+c , Λ0

b , etc.

Very large production x-section75.3± 5.4± 13 µb @ 7 TeV (Phys. Lett. B694 (2010) 209)

Excellent trigger, tracking and PID

Literature Review - 8

models based on the framework of Minimal Falvour Violation3 (MFV) represent someof the most pessimistic extentions of the Standard Model. The work of Lenz, A. &Nierste, U.(2011)[9] represents a recent example of a MFV study. MFV assumes allflavour changing interactions are governed by the same CKM matrix and that the oneindependent phase in the CKM matrix is the only source of CP violation[10]. In general,the MSSM and models involving MFV are plagued by the SUSY CP problem4. A class ofmodels that does not in general suffer from the SUSY CP problem and does not involveMFV is given by supersymmetric flavour models (the interested reader is directed to thework of Altmannshofer et al. (2010)[11]).

2.3 CP Violation in Bs → φφ

The decay B0s → φφ is an example of a flavour changing neutral current (FCNC) and

hence, is forbidden at tree level in the standard model. The decay is only permittedthrough penguin diagrams (shown in figure 5)[12].

b

s

s

s

s

s

W+

u,c,t

(a)

b

s s

s

s

s

W+

(b)

b

s s

s

s

s

W+

u,c,t

(c)

b

s

s

s

s

s

W+

u,c,t

(d)

Figure 5: Feynman diagrams contributing to the Bs → φφ decay[12]: a) Gluonic penguin,b) singlet penguin, c) colour allowed penguin, d) colour supressed penguin.

The weak phase structure found due to mixing in the previous section for the Bs → J/ψφdecay will also hold for the Bs → φφ decay.

In the penguin diagrams of figure 5, the CKM structure remains the same throughout allamplitudes due to the very high mass of the top quark in the propogators. If the QCDstructure is naively assumed to be the same for both B0

s and B̄0s , the ratio of the decay

amplitudes will be of the form

3MFV models can be accommodated in the Minimal Supersymmetric Standard Model (MSSM), forfurther details (specifically relating to B meson observables), see the work of Ellis et al. (2007)[8].

4This arises from neutron electric dipole moment and FCNC predictions constraining CP violatingphases.

School of Physics and Astronomy Date: 24-6-11







Phys. Rev. Lett. 115 (2015) 051801


Phys. Lett B747 (2015) 468

4 Summary







Motivation: B0(s)→ φφ branching fraction

Measurement of the B0s→ φφ branching fraction

Key decay for CP violation studies

Important normalisation mode

Theoretical predictions 1.5− 2.0× 10−5

Previous measurement: 1.91± 0.31× 10−5

(CDF: PRL 95 (2005) 031801)

Search for B0→ φφ decays

Highly suppressed in the SM

SM predictions 1− 30× 10−9

Could be enhanced in BSM models 10−7

Previous limit: < 2.0× 10−7

(Belle: PRL 101 (2008) 201801)

B0s→ φφ diagram

Literature Review - 8

models based on the framework of Minimal Falvour Violation3 (MFV) represent someof the most pessimistic extentions of the Standard Model. The work of Lenz, A. &Nierste, U.(2011)[9] represents a recent example of a MFV study. MFV assumes allflavour changing interactions are governed by the same CKM matrix and that the oneindependent phase in the CKM matrix is the only source of CP violation[10]. In general,the MSSM and models involving MFV are plagued by the SUSY CP problem4. A class ofmodels that does not in general suffer from the SUSY CP problem and does not involveMFV is given by supersymmetric flavour models (the interested reader is directed to thework of Altmannshofer et al. (2010)[11]).

2.3 CP Violation in Bs → φφ

The decay B0s → φφ is an example of a flavour changing neutral current (FCNC) and

hence, is forbidden at tree level in the standard model. The decay is only permittedthrough penguin diagrams (shown in figure 5)[12].

b

s

s

s

s

s

W+

u,c,t

(a)

b

s s

s

s

s

W+

(b)

b

s s

s

s

s

W+

u,c,t

(c)

b

s

s

s

s

s

W+

u,c,t

(d)

Figure 5: Feynman diagrams contributing to the Bs → φφ decay[12]: a) Gluonic penguin,b) singlet penguin, c) colour allowed penguin, d) colour supressed penguin.

The weak phase structure found due to mixing in the previous section for the Bs → J/ψφdecay will also hold for the Bs → φφ decay.

In the penguin diagrams of figure 5, the CKM structure remains the same throughout allamplitudes due to the very high mass of the top quark in the propogators. If the QCDstructure is naively assumed to be the same for both B0

s and B̄0s , the ratio of the decay

amplitudes will be of the form

3MFV models can be accommodated in the Minimal Supersymmetric Standard Model (MSSM), forfurther details (specifically relating to B meson observables), see the work of Ellis et al. (2007)[8].

4This arises from neutron electric dipole moment and FCNC predictions constraining CP violatingphases.

School of Physics and Astronomy Date: 24-6-11

B0→ φφ diagramb

d̄

ss̄

ss̄

φ

φ






B0(s)→ φφ measurement overview (3 fb−1 Run 1 data)

The B0s→ φφ branching fraction is determined w.r.t. B0→ φK∗0

B(B0s→ φφ)

B(B0→ φK∗0)=

NφφNφK∗0

εselφK∗0

εselφφ

B(φ→ K+π−)

B(φ→ K+K−)· fd

fs

Efficiencies determined from MC, apart from PID determined fromcharm decays

fs/fd hadronisation probability to B0s /B0 mesons

Selection: trigger, pre-selection, boosted decision tree, particle ID

Vetoes against φφ↔ φK∗0 cross feed and D+ and D+s decays




Fits to the B0s→ φφ and B0→ φK∗0 mass spectra

BDT cut optimised to maximise the B0→ φK∗0

significance

Fit to the K+K−K+K− spectrum

Only B0s→ φφ and combinatorial

NB0s→φφ = 2349± 49 candidates

Fit to the K+K−K+π− spectrum

B0(s)→ φK∗0 signal peaks

Combinatorial plus Λ0b→ φpπ− and

Λ0b→ φpK− peaking backgrounds

NB0→φK∗0 = 6680± 86 candidates

]2c) [MeV/−K+K−K+K(m5200 5300 5400 5500 5600

)2 cC

andi

date

s / (

9.0

MeV

/

1

10

210

LHCb

]2c) [MeV/−π+K−K+K(m5200 5300 5400 5500 5600

)2 cC

andi

date

s / (

9.0

MeV

/

1

10

210

310 LHCb




B0s→ φφ branching ratio determination

Non-resonant contributions (from angular analysis)

B0s→ φφ S-wave: 2.1± 1.6%

B0→ φK∗0 S-wave: 26.5± 1.8%

Corrected for, but dominant systematic uncertainty

Final resultB(B0

s→φφ)B(B0→φK∗0)

= 1.84± 0.05 (stat)± 0.07 (syst)± 0.11 (fs/fd )

B(B0s→ φφ) =

(1.84± 0.05 (stat)± 0.07 (syst)± 0.11 (fs/fd )± 0.12 (norm))× 10−5

Consistent with previous measurement and ∼ ×2 more precise




B0→ φφ limit setting

Tighter selection, maximising FoM =ε′S

3/2+√

N′bg

ε′S: BDT selection efficiency

N ′bg: estimated background yield

Signal Yield compatible with zeroNB0→φφ = 5± 6 candidates

Limit set with frequentist approachCLs = CLs+b/CLb

Result: B(B0→ φφ) < 2.8× 10−8 @ 90%

Seven times better than previous limit andexcluding some BSM predictions.

Fit to the K+K−K+K− spectrumB0(s)→ φφ and combinatorial

]2c) [MeV/−K+K−K+K(m5200 5300 5400 5500 5600

)2 cC

andi

date

s / (

9.0

MeV

/

-210

-110

1

10

210

310

LHCb

Expected and measured limit of B0→ φφ

)φφ→0B(BF 2 3 4 5 6 7 8 9

8−10×

sC

L

3−10

2−10

1−10

1

10

)φφ→0B(BF 2 3 4 5 6 7 8 9

8−10×

sC

L

3−10

2−10

1−10

1

10

LHCb






Phys. Rev. Lett. 115 (2015) 051801


Phys. Lett B747 (2015) 468

4 Summary







Motivation: B0s→ η′η′, B+→ η′K+ & B+→ φK+

Search for the B0s→ η′η′ decay

Charmless B0s decays less studied than B0 and B+ decays

U-spin related decays (B+→ η′K+ & B+→ φK+) have large BTheoretical predictions B ∼ 1.4− 5.0× 10−5

CP eigenstate – time dependent CPV studies similar to B0s→ φφ

CP asymmetry of B+→ η′K+ and B+→ φK+ decays

Predicted to have negligible CP violation in the SM

Tree and loop diagrams of comparable size

Sensitive to BSM contributionsBaBar, Belle, CLEO: ACP(B+→ η′K+) = 0.013± 0.017LHCb: ACP(B+→ φK+) = 0.022± 0.023BaBar: ACP(B+→ φK+) = 0.128± 0.046


http://journals.aps.org/prd/abstract/10.1103/PhysRevD.80.112002







B0s→ η′η′ selection and fit (3 fb−1 Run 1 data)

Reconstruction & selection

Reconstruct decays ofη′→ π+π−γ

Selection with rectangular cuts

Maximise FoM = εS

5/2+√

Nbg

Fit to the η′ η′ spectrum

B0s and η′ signal peaks

Combinatorial and partiallyreconstructed backgrounds

NB0s→η′η′ = 36.4± 7.8 (stat)± 1.6 (syst)

]2 [MeV/c'η'ηm5000 5200 5400 5600

)2C

andi

date

s / (

20

MeV

/c

0

5

10

15

20

25

30 LHCb

]2 [MeV/c)1γππ(m900 950 1000

)2C

andi

date

s / (

8 M

eV/c

02468

10121416182022

LHCb

]2 [MeV/c)2γππ(m900 950 1000

)2E

vent

s / (

8 M

eV/c

02468

10121416182022

LHCb

Signal significance: 6.4 σ




B0s→ η′η′ branching ratio determination

B+→ η′K+ normalisation channel

Reconstruction & selection nearidentical to B0

s→ η′η′

NB+→η′K+ = 8672± 114 (stat)

Relative efficiency: B0s→η′η′

B+→η′K+

Particle ID, photon & H/Wtrigger determined from data

Remaining efficiencies from MC

Signal and combinatorial background

]2 [MeV/c'Kηm5000 5100 5200 5300 5400 5500

)2C

andi

date

s / (

10

MeV

/c

0

200

400

600

800

1000

1200

1400

1600 LHCb

]2 [MeV/cγππm900 950 1000

)2C

andi

date

s / (

4 M

eV/c

0

200

400

600

800

1000

1200

LHCb

B(B0s→η′η′)

B(B+→η′K+) = 0.47± 0.09 (stat)± 0.04 (syst)

B(B0s→ η′η′) = (3.31± 0.64 (stat)± 0.28 (syst)± 0.12 (norm))× 10−5




CP asymmetries of B+→ η′K+ & B+→ φK+

Raw asymmetries (Araw ) from fits

For small asymmetries: Araw = ACP +AD,k +AP

AD,k +AP is identical to that of B+→ J/ψK+

Measure the difference in asymmetry

ACP(signal)−ACP(B+→ J/ψK+) = Araw(signal)−Araw(B+→ J/ψK+)

Adding the control mode asymmetry (ACP(B+→ J/ψK+)

ACP(B+→ η′K+) = (−0.2± 1.2 (stat)± 0.1 (syst)± 0.6 (norm)) × 10−2

ACP(B+→ φK+) = (+1.7± 1.1 (stat)± 0.2 (syst)± 0.6 (norm)) × 10−2

Most precise results to date and compatible with CP conservation






Phys. Rev. Lett. 115 (2015) 051801


Phys. Lett B747 (2015) 468

4 Summary







Motivation: Amplitude analysis of B0→ ρ0ρ0

Used in measurement of CKM angle α

Combined analysis of B0→ ρ+ρ−, B+→ ρ+ρ0 and B0→ ρ0ρ0

Measures β + γ = π − αLHCb ideal place to measure B0→ ρ0ρ0

Experimental status

Belle and BaBar found evidence for B0→ ρ0ρ0

B = (0.97± 0.24)× 10−6 (Phys. Rev. D 89 (2014) 119903, Phys. Rev. D 78 (2008) 071104)

Tension between measured values of the longitudinal polarisation

Belle fL = 0.21+0.22−0.26

BaBar fL = 0.75+0.12−0.15






B0→ ρ0ρ0 measurment overview (3 fb−1 Run 1 data)

Selection of B0→ π+π−π+π−

Combine two π+π− pairs consistentwith B0 decays

Pre-selection, BDT and particle ID

Vetoes against J/ψ ,χc0,χc2 and D0

two-body decays and the low massregion of three-body combinations

NB0→π+π−π+π− = 634± 29NB0

s→π+π−π+π− = 101± 13B0

s→ π+π−π+π− significance: > 10 σ

Normalisation channel: B0→ φK∗0

Resonant fraction determined fromangular analysis

]2) [MeV/c-π+π)(-π+πM (5100 5200 5300 5400 5500

)2C

andi

date

s/(1

6 M

eV/c

0

50

100

150

200

250 Data)-π+π)(-π+π (→ 0B)-π+π)(-π+π (→ s

0BCombinatorialPartially reconstructed

)-π+π)(-π+ (K→ 0B

LHCb

]2) [MeV/c-π+)(K-K+M (K5100 5200 5300 5400 5500

)2C

andi

date

s/(1

6 M

eV/c

0200400600800

1000120014001600180020002200

Data

)-π+)(K-K+ (K→ 0B

)+π-)(K-K+ (K→ s0B

Combinatorial

Partially reconstructed

LHCb




Amplitude model for the B0→ π+π−π+π− decays

Amplitude variables

Invariant mass m1,2(π+π−) and anglesθ1,2 and φ

Extracted with sPlot technique

Main components

B0→ VV: B0→ ρ0ρ0 & B0→ ρ0ω

B0→ VS: B0→ ρ0f0(980) & ρ0 + (π+π−)0

B0→ VT: B0→ ρ0f2(1270)

Contamination from B0→ a±1 π±

Definition of angles

−

+

−

θ

B

12θ

ππ

0

π ϕπ

+

Differential decay rate

d5(Γ + Γ)

d cos θ1 d cos θ2 dϕ dm21 dm2

2




Results of the amplitude analyssis

Fit results

Fraction of B0→ ρ0ρ0:P(B0→ ρ0ρ0) = 0.619± 0.072 (stat)± 0.049 (syst)

Longitudinal polarisation:fL = 0.745+0.048

−0.058 (stat)± 0.034 (syst)

]2 [MeV/c1,2

)-π+πM(400 600 800 1000

)2Y

ield

/ (

40 M

eV/c

0

20

40

60

80

100

120

140

160

LHCb

DataTotal

0ρ0ρ →0B long.0ρ0ρ →0B trans.0ρ0ρ →0B

0ρω →0B VS→0B SS→0B ±

π±1 a→0B

VT→0B

1,2θcos-1 -0.5 0 0.5 1

Yie

ld /

( 0.

1 )

0

20

40

60

80

100

120

LHCb

[rad]ϕ-2 0 2

Yie

ld /

( 0.

314

)0

10

20

30

40

50LHCb




Branching ration determination

Efficiency corrections

Integrated across the phase space, taking amplitude model into account

Accounting for MC and data differences

Final result

Approximately 390 B0→ ρ0ρ0 candidates observed

Signal significance 7.1 σ

B(B0→ρ0ρ0)B(B0→φK∗0)

= 0.094± 0.017 (stat)± 0.009 (syst)

B(B0→ ρ0ρ0) = (0.94± 0.17 (stat)± 0.09 (syst)± 0.06 (norm))× 10−6

No evidence found for B0→ ρ0f0(980)

Consistent with previous measurement and ∼ ×12 more precise




Summary

1 B0(s)→ φφ branching ratio

Key decay mode for CPV studiesB(B0

s→ φφ) = (1.84± 0.18)× 10−5

B(B0→ φφ) < 2.8× 10−8 @ 90%

2 First observation of B0s→ η′η′ (6.4σ)

First step towards CPV studies of B0s→ η′η′

B(B0s→ η′η′) = (3.31± 0.71)× 10−5

ACP(B+→ η′K+) = (−0.2± 1.3) × 10−2

ACP(B+→ φK+) = (+1.7± 1.3) × 10−2

3 First observation of B0→ ρ0ρ0 (7.1σ)Used for measuring CKM angle αB(B0→ ρ0ρ0) = (0.94± 0.20) × 10−6

Longitudinal polarisation fL = 0.745+0.059−0.067


Charmless B decays at LHCbs! !! decay[12]: a) Gluonic penguin, b) singlet penguin, c) colour allowed...

Documents

Transcript of Charmless B decays at LHCbs! !! decay[12]: a) Gluonic penguin, b) singlet penguin, c) colour allowed...