Review of radiative B meson decays in LHCb
Samuel Coquereau
X-CPAN Days, Salamanca, 29 - 31 October 2018
Introduction
I In the Standard Model (SM), the b→sγ process is forbidden at tree level.
W− u, c, t W−γb s
B Using the effective field theory:
Heff =−4GF
√
2 VtbVts∗(C7O7+C70O70) I whereC7 describe the left-handed currents andC70 the right-handed
currents
I The photon polarization can be defined as:
Aγ = P(γL)−P(γR)
P(γL) +P(γR) ALOγ = 1− |CC70
7|2 1 +|CC70
7|2
Samuel Coquereau CPAN 2018 29thOctober 2018 2 / 18
Introduction
I In the Standard Model (SM), the b→sγ process is forbidden at tree level.
W− u, c, t W−γb s
B Using the effective field theory:
Heff =−4GF
√
2 VtbVts∗(C7O7+C70O70) I In the SM C70 = 0 due to absence of right-handed currents
I Leading to:
Aγ = P(γL)−XXP(γRXX)
P(γL) +XXP(γRXX)] = 1+O(ms mb
) ALOγ = 1−Z ZZ
|CC70
7|2 1 +Z
ZZ
|CC70
7|2
= 1
The LHCb experiment
Interaction point
Velo
Tracking stations Tracking
RICH system
Muon chambers ECAL, HCAL Particle Id
Samuel Coquereau CPAN 2018 29thOctober 2018 4 / 18
Analysis in radiative B meson decays at LHCb
1 Photon polarisation at LHCb
2 Bs →φγ
3 Multichannel
Photon polarization at LHCb
I First observation in B+→K+π+π−γ with 3fb−1 (2011-2012) [Phys. Rev. Lett. 112, 161801 (2014)]
B Full amplitude analysis ongoing
I First measurement inB0→K∗e+e− lowq2 with 3fb−1: [JHEP04(2015)064]
A(2)T =−0.23±0.23±0.05 AImT = +0.14±0.22±0.05 B Update with run 1 + run 2
dataset ongoing
] c2
/ [MeV
−)
+e
−e π K+
m(
4800 5000 5200 5400
) 2cCandidates / (30 MeV/
0 5 10 15 20 25
30 Data
Model B0→K*0e+e− e− e+ ) X
*0 K
→( B Combinatorial
LHCb
Samuel Coquereau CPAN 2018 29thOctober 2018 6 / 18
B
s→ φγ (see Clara’s talk)
I Time-dependent decay rate is sensitive to photon polarisation:
Γ(t)∝e−Γst[cosh(∆Γst
2 )−A∆sinh(∆Γst
2 )±Ccos(∆mst)∓Ssin(∆mst)]
I A∆,C andS are dependent to the photon polarisation A∆≈ 2Re(eiφsC7C70)
|C7|2+|C70|2 S ≈ 2Im(eiφsC7C70)
|C7|2+|C70|2 and C ≈0 I giving in the SM (C70 supressed):
A∆SM= 0.047+0.029−0.025 SSM= 0±0.002 CSM≈0.005±0.005
B
s→ φγ: untagged analysis
[Phys. Rev. Lett. 118 (2017) no2, 021801]I Analysis performed using 3fb−1 (Run 1) I 4200Bs →φγ signal events
A∆=−0.98+0.46−0.52+0.23−0.20 B 2σ deviation from SM
B Uncertainty dominated by the statistics
B Systematic uncertainty could be reduce with more statistics
2] c ) [MeV/
γ φ ( m
5000 5500 6000
)2cCandidates / (25 MeV/
0 100 200 300 400
500 Data
Model Signal Peaking Missing kaon Combinatorial LHCb
γ φ
→
0
Bs
[ps]
t
0 5 10
Ratio of candidate yields
0 0.05 0.1 0.15 0.2 0.25 0.3
Data Fit SM LHCb
Samuel Coquereau CPAN 2018 29thOctober 2018 8 / 18
B
s→ φγ: tagged analysis (see Clara’s talk)
I Analysis with Run 1 well advance (results still blinded) I Fit strategy :
B Simultaneous unbinned fit of Bs →φγ andB0→K∗0γ (control channel) proper-time distribution
B Γ(t) PDF:
n e−Γst0
h
cosh(∆Γ2st0)− A∆sinh(∆Γ2st0) +
κ(1−2ω))Ccos(∆mst0)− Ssin(∆mst0) io
∗[A(ti)×R(t,t0|σt)]
B κ andω are the tagging decision and the mistag respectively B A(ti): binned version of the acceptance function
B R(t,t0|σt): 2-Gaussian model taken from MC
B
s→ φγ: tagged analysis (see Clara’s talk)
I Analysis with Run 1 well advance (results still blinded)
Samuel Coquereau CPAN 2018 29thOctober 2018 10 / 18
B
s→ φγ: tagged analysis (see Clara’s talk)
I Analysis with Run 2:
B More statistics (∼ ×2 Run 1 yield) B Improved flavour tagging (∼10%) B Expected improvements in systematics
B
s→ φγ: tagged analysis (see Clara’s talk)
I Analysis with Run 2:
B More statistics (∼ ×2 Run 1 yield) B Improved flavour tagging (∼10%) B Expected improvements in systematics
I Expected constrains combining A∆ andS (Run 1+ Run 2 Data) I σ(A∆)'σ(S)'0.2
I S sensitive to Im(C70)
Samuel Coquereau CPAN 2018 29thOctober 2018 12 / 18
Multichannel
I The goal:
B Measurement of B(Λ0b→Λ∗0γ) andB(Bs0 →φγ) relative to B(B0→K∗γ)
B Measurement of ACP for Λb →Λ∗γ andB0 →K∗γ I The strategy:
B Use the Run 1 dataset
B 5 simultaneous fits to extract the 5 yields: N(Λb→Λ∗γ), N(Λb→Λ∗γ), N(B0 →K∗γ), N(B0 →K∗γ) and N(Bs0→φγ) B Separate BDT’s for each channel to reduce combinatorial
background
B Particle identification also use to further reduce background
Mass fit strategy
I Signal:
B Double-sided Crystal Ball: αL,nL, αR,nR extracted from MC,σ andµfitted on data
I Background:
B Combinatorial: Chebychev polynomial, Comb(m;p0) = 1 +p0×m
B Peaking Background: Double sided Crystal Ball
B Partially reconstructed background : Argus convoluted with a gaussian,µand σ taken from signal andc andp fixed from MC
Samuel Coquereau CPAN 2018 29thOctober 2018 14 / 18
Mass Fit
B0→K∗0γ
2] c M(BK_B) [MeV/
Candidates / ( 27.5 )
0 200 400 600 800 1000 1200 1400 1600 1800
2000 ABCPd = -0.74271 ± 0.0087
0.022
± = 0.266
*γ
→ K B Ccomb
± 374 = 31547 γ K*
→ NB
0.017
± = 0.145 _B_2012 ρ D0
→ B+
C
0.044
± = 0.234 _B_2012 γ
→ K1 B+
C
0.043
± = 0.177 _B_2012 η K*
C
c2 0.85 MeV/
± = 5281.98
*γ
→ K µB
0.087
± ] = -0.3007 [MeV-1
*γ
→ K B p0
c2 0.82 MeV/
± = 87.47
*γ
→ K σB
0.0087
± = -0.74271 Bd
ACP
0.022
± = 0.266
*γ
→ K B Ccomb
± 374 = 31547 γ K*
→ NB
0.017
± = 0.145 _B_2012 ρ D0
→ B+
C
0.044
± = 0.234 _B_2012 γ
→ K1 B+
C
0.043
± = 0.177 _B_2012 η K*
C
c2 0.85 MeV/
± = 5281.98
*γ
→ K µB
0.087
± ] = -0.3007 [MeV-1
*γ
→ K B p0
c2 0.82 MeV/
± = 87.47
*γ
→ K σB
46005 4800 5000 5200 5400 5600 5800 6000 6200
− 0 5
B0 →K∗0γ
2] c M(BK_Bbar) [MeV/
Candidates / ( 27.5 )
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
0.0087
± = -0.74271 Bd
ACP
0.021
± = 0.232
*γ
→K B Ccomb
± 374 = 31547 γ K*
→ NB
0.016
± = 0.185 _Bbar_2012 ρ D0
→ B+
C
0.044
± = 0.102 _Bbar_2012 γ
→ K1 B+
C
0.043
± = 0.299 _Bbar_2012 η K*
C
c2 0.85 MeV/
± = 5281.98
*γ
→ K µB
0.11
± ] = -0.175 [MeV-1
*γ
→K B 0 p
c2 0.82 MeV/
± = 87.47
*γ
→ K σB
0.0087
± = -0.74271 Bd
ACP
0.021
± = 0.232
*γ
→K B Ccomb
± 374 = 31547 γ K*
→ NB
0.016
± = 0.185 _Bbar_2012 ρ D0
→ B+
C
0.044
± = 0.102 _Bbar_2012 γ
→ K1 B+
C
0.043
± = 0.299 _Bbar_2012 η K*
C
c2 0.85 MeV/
± = 5281.98
*γ
→ K µB
0.11
± ] = -0.175 [MeV-1
*γ
→K B 0 p
c2 0.82 MeV/
± = 87.47
*γ
→ K σB
46005 4800 5000 5200 5400 5600 5800 6000 6200
− 0 5
Mass Fit
] c2 M(Lambda_B) [MeV/
Candidates / ( 33.3333 )
0 100 200 300 400 500
0.017
± = -0.5966 Λb
ACP
0.14
± = 1.88
*γ Λ b→ Λ Ccomb
0.13
± = 5.60 γ Λ
→ RΛ
0.041
± = 0.105 _B_2012 γ
→ K1 B+
C
0.073
± = 0.045 _B_2012 γ φ K C
c2 1.6 MeV/
± = 5622.2 µΛ
c2 1.7 MeV/
± = 83.6 σΛ
0.000063
± = -0.0012936 γ Λ*
→ τΛ
0.017
± = -0.5966 Λb
ACP
0.14
± = 1.88
*γ Λ b→ Λ Ccomb
0.13
± = 5.60 γ Λ
→ RΛ
0.041
± = 0.105 _B_2012 γ
→ K1 B+
C
0.073
± = 0.045 _B_2012 γ φ K C
c2 1.6 MeV/
± = 5622.2 µΛ
c2 1.7 MeV/
± = 83.6 σΛ
0.000063
± = -0.0012936 γ Λ*
→ τΛ
4600 4800 5000 5200 5400 5600 5800 6000 6200 6400 66005
− 0 5
Λb→Λ∗γ B In red: The signal
B In black: The combinatorial background
] c2 M(Lambda_Bbar) [MeV/
Candidates / ( 33.3333 )
0 100 200 300 400
500 AΛCPb = -0.5966 ± 0.017
0.15
± = 1.97
*γ Λ b→ Λ Ccomb
0.13
± = 5.60 γ Λ
→ RΛ
0.045
± = 0.163 _Bbar_2012 γ
→ K1 B+
C
0.077
± = 0.045 _Bbar_2012 γ φ K C
c2 1.6 MeV/
± = 5622.2 µΛ
c2 1.7 MeV/
± = 83.6 σΛ
0.000065
± = -0.0012397
*γ Λ
→ τΛ
0.017
± = -0.5966 Λb
ACP
0.15
± = 1.97
*γ Λ b→ Λ Ccomb
0.13
± = 5.60 γ Λ
→ RΛ
0.045
± = 0.163 _Bbar_2012 γ
→ K1 B+
C
0.077
± = 0.045 _Bbar_2012 γ φ K C
c2 1.6 MeV/
± = 5622.2 µΛ
c2 1.7 MeV/
± = 83.6 σΛ
0.000065
± = -0.0012397
*γ Λ
→ τΛ
4600 4800 5000 5200 5400 5600 5800 6000 6200 6400 66005
− 0 5
Λb→Λ∗γ B In green: B+→K1+γ B In cyan: B+→φKγ
Samuel Coquereau CPAN 2018 29thOctober 2018 16 / 18
First results
B ACP are still blind
B Cross feed contribution fixed in the fit
B Partial. reco. Bkg contribution let free in the fit
I Cross check onB(B0→K∗0γ) and Branching ratio measured in agreement with SM (without syst.
uncert. yet)
I Using the PDG value for B(B0→K∗0γ) we can extract:
2] c M(Bs) [MeV/
Candidates / ( 25.8333 )
0 100 200 300 400
500 = 0.439 ± 0.026
γ φ s→ B Ccomb
0.16
± = 7.16 γ φ s→ RB
0.0098
± = 0.0970 _2012 Kγ φ
→ B+
C
c2 1.7 MeV/
± = 5366.3 µBs
0.043
± ] = -0.5955 [MeV-1 γ φ
→ B p0
c2 1.7 MeV/
± = 88.4 σBs
0.026
± = 0.439 γ φ s→ B Ccomb
0.16
± = 7.16 γ φ s→ RB
0.0098
± = 0.0970 _2012 Kγ φ
→ B+
C
c2 1.7 MeV/
± = 5366.3 µBs
0.043
± ] = -0.5955 [MeV-1 γ φ
→ B p0
c2 1.7 MeV/
± = 88.4 σBs
46005 4800 5000 5200 5400 5600 5800 6000
− 0 5
Bs0→φγ
Conclusion
I Radiative decays are sensitive to the photon polarisation
B Allowing to constrain the C70 complex plane and to test the SM prediction
I Many analysis not shown here:
B B0 →K∗γ isospin asymmetry, Search for Λ0b→N∗γ, ...
I Potential to improve the current results:
B Analysis using Run 2 dataset of LHCb are in progress B Improvement of reconstruction and analysis techniques I New measurement are coming:
B S(Bs0 →φγ),ACP for Λb →Λ∗γ B b-baryon decays (see Luis Miguel’s talk)
Samuel Coquereau CPAN 2018 29thOctober 2018 18 / 18
Thank you for your attention
BACKUP
Samuel Coquereau CPAN 2018 29thOctober 2018 18 / 18
Photon polarization in B
0→ K
∗e
+e
−low q
2I At lowq2 B0 →K∗e+e− is dominated by the photon pole:
1 d(Γ + ¯Γ)/dq2
d3(Γ + ¯Γ) dcosθ`dcosθKdφ= 9
16π 3
4(1−FL) sin2θK+FLcos2θK+ 1
4(1−FL) sin2θK−FLcos2θK cos 2θ`+ 1
2(1−FL)A(2)T sin2θKsin2θ`cos 2φ+ (1−FL)AReT sin2θ`cosθ`+ 1
2(1−FL)AImT sin2θKsin2θ`sin 2φ
.
] c2 / [MeV
−) +e
−e π K+ m(
4800 5000 5200 5400
) 2cCandidates / (30 MeV/
0 5 10 15 20 25
30 Data
Model B0→K*0e+e− e− e+ ) X
*0 K
→( B Combinatorial
LHCb
q2∈[0.002,1.120]
GeV2/c4
A(2)T (q2 →0) = 2Re(C7C70∗)
|C7|2+|C70|2, AImT (q2 →0) = 2Im(C7C70∗)
|C7|2+|C70|2