Electroweak penguins with di-leptons
M. Kreps
Physics Department
www2.warwick.ac.uk
Motivation for EWK penguins
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
In the SM, b → s FCNC decays
SM BF of the order of 10−6
With angular analysis offers
variety of observables
Allows to test some underlying
details of the NP
Form factors make prediction of
some observables less precise
But many observables are free of
form factor uncertainties
Mesons rather well studied,
results on
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
2
0
∗0 + −
B →K µ µ
B 0 → K ∗0 e+ e−
Bs0 → φµ+ µ−
15 July 2013
µ−
µ+
ν
W
W
b
s
q
q
µ+
γ, Z
µ−
s
b
W
q
q
Michal Kreps – Electroweak penguins with di-leptons
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
Λb has non-zero spin
Potential to learn something
new compared to mesons
Can
influence
also
understanding of the mesons
by better understanding of
hadronic physics
First LHCb result on Λb →
Λµ+ µ−
Discussion of some prospects
All results shown today based
on 1 fb−1
dB (Λµ+µ-)/dq2 [10-7(GeV2/c4)-1]
Case of Λb baryon
6
Detmold et al., PRD87, 074502
Aliev et al., PRD81, 056006
Mott, Roberts, Int.J.Mod.Phys. A27, 1250016
Gutsche et al., PRD87, 074031
4
2
0
0
5
10
15
20
q2 [GeV2/c4]
3
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
LHCb detector
4
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Selection with dimuons
Uses BDT or NN to combine
kinematic, topological and PID
inputs
Trained on resonant (J/ψ ) signal
in data
Relatively complicated removal
of cc̄ regions
Decays via J/ψ used for
normalization
4000
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
m(µ+µ−) [MeV/c2]
LHCb
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
m(µ+µ-) [MeV/c2]
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
1200
3500
1000
104
4000
103
3000
2000
102
1000
10
0
5200
5400
5600
1
m(K+π−µ+µ−) [MeV/c2]
3000
800
2500
2000
600
1500
400
1000
LHCb
500
5100
5
5200
15 July 2013
5300
5400
5500
5600
5700
5800
m(K+K µ+µ-) [MeV/c2]
200
0
B 0 → K ∗0 µ+ µ−
LHCb-PAPER-2013-019
Bs0 → φµ+ µ−
LHCb-PAPER-2013-017
Michal Kreps – Electroweak penguins with di-leptons
Peaking backgrounds
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
Most obvious one are decays through charmonia resonances
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
6
Decays through J/ψ and ψ(2S) resonance are vetoed
B 0 → K ∗0 µ+ µ− and Bs0 → φµ+ µ− cross-talk to each other
Similar Λb decays should exists and so they can contribute as well
Thanks to good particle ID at LHCb, they give just small contribution
Many backgrounds of this type can be vetoed rather easily
Typically they are not very difficult to handle for differential branching
fraction measuements
More critical for angular analysis
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
B 0 → K ∗0µ+µ− — q 2 bins
Extended maximum likelihood fit
Signal: double Crystal ball function fixed from J/ψ decays
Bg: combinatorial (exponential), Bs0 → K ∗0 µµ, Bs0 → φµµ and
B 0 → J/ψ K ∗0 from MC
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
0.1 < q < 2 GeV /c 4
40
Signal
Combinatorial bkg
20
Candidates / ( 10 MeV/c 2 )
0
Peaking bkg
5200
5400
5600
mK+π-µ+µ- [
2
2
10.09 < q < 12.86 GeV /c 4
40
168 ± 15
5200
15 July 2013
5400
5600
mK+π-µ+µ- [
2
40
]
73 ± 11
20
5200
5400
5600
mK+π-µ+µ- [
2
14.18 < q < 16 GeV /c 4
40
115 ± 12
20
5200
5400
5600
mK+π-µ+µ- [
LHCb 1.0 fb-1
4.3 < q2 < 8.68 GeV2/c 4
40
]
5200
5400
5600
mK+π-µ+µ- [ MeV/c 2 ]
60
LHCb 1.0 fb-1
16 < q2 < 19 GeV2/c 4
40
116 ± 13
20
0
MeV/c 2
271 ± 19
20
]
LHCb 1.0 fb-1
2
60
0
MeV/c 2
60
0
MeV/c 2
2
2 < q < 4.3 GeV /c 4
]
LHCb 1.0 fb-1
20
LHCb 1.0 fb-1
0
MeV/c 2
60
0
7
2
Candidates / ( 10 MeV/c 2 )
-1
2
60
Candidates / ( 10 MeV/c 2 )
140 ± 13 LHCb 1.0 fb
Candidates / ( 10 MeV/c 2 )
60
Candidates / ( 10 MeV/c 2 )
Candidates / ( 10 MeV/c 2 )
LHCb-PAPER-2013-019
5200
5400
5600
mK+π-µ+µ- [ MeV/c 2 ]
Michal Kreps – Electroweak penguins with di-leptons
Bs0 → φµ+µ− — q 2 bins
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
Extended maximum likelihood fit
Signal: double Gaussian function fixed from J/ψ decays
Bg: combinatorial (exponential)
Clear signal in all q 2 bins
First time we see signal in all q 2 bins
10
LHCb
25.0+5.8
−5.2
5
8
6
2.0 < q2 < 4.3 GeV2/c4
LHCb
14.3+4.9
−4.3
4
Events / 10 MeV/c2
0.1 < q2 < 2.0 GeV2/c4
Events / 10 MeV/c2
Events / 10 MeV/c2
LHCb-PAPER-2013-017
4.3 < q2 < 8.68 GeV2/c4
20
LHCb
41.2+7.5
−7.0
10
2
10
5400
10.09 < q2 < 12.9 GeV2/c4
0
5500
Bs mass [MeV/c2]
LHCb
40.7+7.7
−7.2
5
0
8
5300
5200
5300
8
14.18 < q2 < 16.0 GeV2/c4
6
23.8+5.9
5.3
5400
LHCb
4
2
5200
5300
15 July 2013
5400
5500
Bs mass [MeV/c2]
0
5200
5300
5400
0
5500
Bs mass [MeV/c2]
5500
Bs mass [MeV/c2]
Events / 10 MeV/c2
15
5200
Events / 10 MeV/c2
Events / 10 MeV/c2
0
8
6
5200
5300
5400
16.0 < q2 < 19.0 GeV2/c4
5500
Bs mass [MeV/c2]
LHCb
26.6+5.7
−5.3
4
2
0
5200
5300
5400
5500
Bs mass [MeV/c2]
Michal Kreps – Electroweak penguins with di-leptons
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
Candidates per 10 MeV/c2
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
15
No significant signal below J/ψ resonance
Only highest q 2 bin above 5σ
15
10
2±3
[4.30-8.68] GeV2/ c4
10 10
4±3
5
55
0
15
0
0 0
[10.09-12.86] GeV2/ c415
[14.18-16.00] GeV2/c415
10
10
10
13 ± 5
5
9
[2.00-4.30] GeV2/ c4
5
0
Λb→Λµ+µ-
LHCb
[0.00-2.00] GeV2/ c4
10
15 15
14 ± 4
5
5.4
15 July 2013
5.6
0
5.8 5.4
Candidates per 10 MeV/c2
Λb → Λµ+µ− – q 2 bins
25
20
LHCb
Λb → Λµ+µ-
15
10
5
0
5.4
5.5
5.6
5.7
5.8
M(Λµ µ ) [GeV/c2]
+ -
4±5
[16.00-20.30] GeV2/c4
44 ± 7
5
5.6
5.8
0
5.4
5.6
+ -
M(Λµ µ )
5.8
[GeV/ c2]
Michal Kreps – Electroweak penguins with di-leptons
1.5
Theory
LHCb
Binned
B 0 → K ∗0 µ+ µ−
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
LHCb
1
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
0.5
0
0
LHCb-PAPER-2013-019
5
10
15
q2
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
10
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
N
2
[GeV
20
/ c 4]
sig
ǫrel Bnorm
dB/dq 2 = ∆q1 2 Nnorm
B 0 → K ∗0 µ+ µ− agrees with SM
Two decays have simular shape
For Bs0 → φµµ we measure
integrated rate
+0.64
± 0.17 ± 0.71) × 10−7
(7.07−0.59
15 July 2013
Syst. uncertainties dominated by
normalization BF
Other contributions from Bg
PDFs and data-MC differences
Large fraction of systematic
uncertainty correlated across q 2
bins
dB(Bs→φ µ +µ −)/dq2 [GeV-2c4]
dB /dq 2 [10-7 × c 4 GeV-2]
Differential BF
×10-6
0.1
LHCb
Bs0 → φµ+ µ−
0.05
LHCb-PAPER-2013-017
0
5
10
15
q2 [GeV2/c4]
Michal Kreps – Electroweak penguins with di-leptons
Differential BF
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
Sidenote
on
Λb
normalization:
PDG gives f (b → Λb )×
B(Λb → J/ψΛ)
No good B(Λb → J/ψΛ)
1.5
LHCb
1
0.6
fΛ
b
0.5
(a)
0.5
3<η<5
(b)
2<η<3
0.4
0.3
0.2
LHCb
0
0
0.1
5
10
15
q2
11
f u+f d
d B(Λµ+µ-)/dq2 [10-7(GeV2/c4)-1]
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
Limits at first three bins are Bayesian at 90% C.L.
Non-observation at low q 2 is not entirely by lack of sensitivity
Will need more data to say something more
Integrated BF (excluding J/ψ and ψ(2S) vetos)
(0.96 ± 0.16(stat) ± 0.13(syst) ± 0.21 (norm)) × 10−6
15 July 2013
20
2
[GeV
/c4]
0
s = 7 TeV
5
10
p (charm + µ) (GeV)
T
0
5
10
p (charm + µ) (GeV)
T
Michal Kreps – Electroweak penguins with di-leptons
All experiments
FL
Theory
LHCb
Binned
CDF
BaBar
Belle
ATLAS
CMS
1
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
0.8
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
0.6
LHCb
measurements
most precise
CMS at high q 2 comes
close to LHCb
0.4
0
0
5
10
15
d B(Λµ+µ-)/dq2 [10-7(GeV2/c4)-1]
0.2
3
20
LHCb
LHCb
q 2 [GeV2/ c 4]
CDF
2
1
0
0
5
10
15
20
q2 [GeV2/c4]
12
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Low q 2 region
Low q 2 region particularly sensitive to γ polarization
Decay B 0 → K ∗0 e+ e− can reach lower q 2 as decay with dimuons
Never observed on its own at low q 2
Before moving to angular analysis, measure differential BF in 30 <
m(ℓ+ ℓ− ) < 1000 MeV
Below 30 MeV angles hard to measure
Analysis more challenging by significantly larger brehmstralung
⇒ Try to recover radiated photons
Use of BDT for selection, trained on simulated signal and bg from
data
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
Candidates / (50 MeV/ c 2)
LHCb
L 0 Electron
1000
800
600
400
200
800
LHCb
L 0 TIS
700
600
500
400
300
200
100
0
4500
5000
5500
0
6000
+ −
*0
2
m( e e K ) [MeV / c ]
13
15 July 2013
4500
5000
5500
6000
+ −
*0
m( e e K ) [MeV / c 2]
LHCB-PAPER-2013-005
Candidates / (50 MeV/ c 2)
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
Michal Kreps – Electroweak penguins with di-leptons
B 0 → K ∗0e+e− at low q 2
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
11
00
00
11
000
111
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
Signal: double Crystal ball
Combinatorial bg: exponential
Partially reconstructed bg from
simulation
Significance 4.6σ
Systematic uncertainties well bellow
statistical
Dominant
contributions
from
normalization BF and fit procedure
Resulting B 0 → K ∗0 e+ e− BF
+0.9 +0.2
−7
±
0.3)
×
10
(3.1−0.8
−0.3
+ −
in m(ℓ ℓ ) within 30–1000 MeV
LHCB-PAPER-2013-005
15.0+5.1
−4.5
12
LHCb
L0Electron
10
Data
B → e+e - K *0
8
B → e+e - (K X)
*0
6
B → K *0 γ
Comb. background
4
2
0
4500
5000
5500
6000
m e+e- K *0 [MeV/ c 2]
Candidates / (50 MeV/ c 2)
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
Main physics background from
B 0 → K ∗0 γ with γ conversion
Extended maximum likelihood fit
Candidates / (50 MeV/ c 2)
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
14.1+7.0
−6.3
12
LHCb
L0TIS
10
Data
B → e+e - K
8
*0
B → e+e - (K*0 X)
6
B→K γ
*0
Comb. background
4
2
0
4500
5000
5500
6000
m e+e- K *0 [MeV/ c 2]
14
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Angular analysis
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
Angular distribution
S1s sin2 θK + S1c cos2 θK +
S2s sin2 θK cos 2θℓ + S2c cos2 θK cos 2θℓ +
S3 sin2 θK sin2 θℓ cos 2φ + S4 sin 2θK sin 2θℓ cos φ +
S5 sin 2θK sin θℓ cos φ + S6 cos2 θK cos θℓ +
A7 sin 2θK sin θℓ sin φ + A8 sin 2θK sin 2θℓ sin φ +
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
15
A9 sin2 θK sin2 θℓ sin 2φ
Helicity angles θl and θK defined by
µ+ (µ− ) and K for B 0 (B̄ 0 )
Angle φ is given by
sin φ = (p̂µ+ × p̂µ− ).p̂K ∗0
sin φ = (p̂µ− × p̂µ+ ).p̂K ∗0
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Angular analysis
Assuming q 2 >> 4m2µ and transforming φ as φ + π if φ < 0 one can
write
3
2
FL cos θK + (1 − FL )(1 − cos2 θK ) − FL cos2 θK (2 cos2 θℓ − 1) +
4
1
(1 − FL )(1 − cos2 θK )(2 cos2 θℓ − 1) +
4
4
2
2
S3 (1 − cos θK )(1 − cos θℓ ) cos 2φ̂ + AF B (1 − cos2 θK ) cos θℓ +
3
A9 (1 − cos2 θK )(1 − cos2 θℓ ) sin 2φ̂
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
16
The observables have to satisfy:
|A9 | ≤ 21 (1 − FL ),
|S3 | ≤ 12 (1 − FL )
|AF B | ≤ 34 (1 − FL ),
Those boundary conditions introduce non-trivial effects
The statistical uncertainties estimated using Feldman-Cousins
technique
At lowest q 2 bin one needs also correction from breakdown of q 2 >>
4m2µ assumption
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Bs0 angular analysis
After adding decay rate for Bs and B̄s it looks very similar to B 0
angular distribution
Main difference is that Bs → φµ+ µ− is not self-tagging
→ No sensitivity to AF B unless production flavour is tagged
Due to small statistics, flavour tagging not really possible
Also full angular analysis difficult with current statistics
→ Look to 1D distributions instead
3
3
2
θK : (1 − FL )(1 − cos θK ) + FL cos2 θK ,
4
2
3
3
3 s
2
2
θℓ : (1 − FL )(1 + cos θℓ ) + FL (1 − cos θℓ ) + A6 cos θℓ ,
8
4
4
1
1
1
+
S3 cos 2φ +
A9 sin 2φ
dφ :
2π 2π
2π
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
17
Subsequently fit all three distributions
Should be careful with exact meaning due to non-zero ∆Γs
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Systematic uncertainties
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
Several sources investigated (not all for Bs due to small statistics)
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
18
Angular acceptance (dominant for Bs )
Data-MC differences
Mass model
Peaking backgrounds
S-wave (spin-0 Kπ)
B 0 ↔ B̄ 0 mis-id from K ↔ π mis-id
B 0 , B̄ 0 production and detection asymmetries
Generally no single dominant contribution across all angular
observables
Many of the sources investigated are negligible
Typicall size for Bs of order 20% – 40% of statistical uncertainty
For B 0 typically of order 5% – 20% of statistical uncertainty
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Systematic uncertainties
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
Several sources related to physics where we do not have any
experimental information
With increase in dataset, we gain sensitivity and can start to measure
inputs
Ultimately sources like peaking backgrounds, s-wave or production
asymmetries should be improved with larger statistics
Currently we do not think we can become systematically limited in
those measurements
Source
Acceptance model
Mass model
B 0 → B 0 mis-id
Data-simulation diff.
Kinematic reweighting
Peaking backgrounds
S-wave
B 0 -B 0 asymmetries
19
15 July 2013
AFB
0.02
< 0.01
< 0.01
0.01
< 0.01
0.01
0.01
< 0.01
FL
0.03
< 0.01
< 0.01
0.03
0.01
0.01
0.01
< 0.01
S3
0.01
< 0.01
< 0.01
0.01
< 0.01
0.01
0.02
< 0.01
S9
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
0.01
0.01
< 0.01
A9
< 0.01
< 0.01
0.01
< 0.01
< 0.01
0.01
< 0.01
< 0.01
A2T
0.02
< 0.01
< 0.01
0.03
0.01
0.01
0.05
< 0.01
ARe
T
0.01
< 0.01
< 0.01
0.01
< 0.01
0.01
0.04
< 0.01
Michal Kreps – Electroweak penguins with di-leptons
1
Theory
LHCb
0.8
Binned
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
B 0 → K ∗0 µ+ µ−
LHCb
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
0.6
0.4
0.2
0
0
LHCb-PAPER-2013-019
5
10
15
q2
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
20
[GeV
2
20
/c 4]
First angular analysis of Bs0 →
φµ+ µ−
Within limited statistics consistent
with SM
I would also say that B 0 and Bs look
consistent
15 July 2013
FL is fraction of longitudinal
K ∗0 /φ
Theory affected by form factors
Good agreement between
experiment and SM
FL
FL
FL
1.5
Bs0 → φµ+ µ−
LHCb
1
0.5
0
LHCb-PAPER-2013-017
-0.5
5
10
15
q2 [GeV2/c4]
Michal Kreps – Electroweak penguins with di-leptons
FL
Binned
CDF
BaBar
Belle
ATLAS
CMS
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
1
B 0 → K ∗0 µ+ µ−
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
0.8
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
0.6
0.4
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
0.2
0
0
LHCb-PAPER-2013-019
5
10
15
q2
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
20
2
[GeV
20
/ c 4]
First angular analysis of Bs0 →
φµ+ µ−
Within limited statistics consistent
with SM
I would also say that B 0 and Bs look
consistent
15 July 2013
FL is fraction of longitudinal
K ∗0 /φ
Theory affected by form factors
Good agreement between
experiment and SM
Consistent but more precise than
previous measurments
FL
FL
Theory
LHCb
1.5
Bs0 → φµ+ µ−
LHCb
1
0.5
0
LHCb-PAPER-2013-017
-0.5
5
10
15
q2 [GeV2/c4]
Michal Kreps – Electroweak penguins with di-leptons
AFB
B 0 → K ∗0µ+µ− AF B
Theory
LHCb
1
LHCb
Binned
For Bs0 → φµµ
no sensitivity without
flavour tagging
LHCb-PAPER-2013-019
0.5
0
-0.5
q02 = (4.9 ± 0.9) GeV2
5
10
15
q2
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
21
[GeV
2
20
/c 4]
Probably best known of angular
observables
Early measurements caused some
excitements
With increased statistical power of LHCb,
fully consistent with SM
15 July 2013
Theory
LHCb
AFB
-1
0
Binned
CDF
BaBar
Belle
ATLAS
CMS
1
0.5
0
-0.5
-1
0
5
10
15
20
q 2 [GeV2/ c 4]
Michal Kreps – Electroweak penguins with di-leptons
S3
S3 and A6
1
Theory
LHCb
Binned
B 0 → K ∗0 µ+ µ−
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
LHCb
0.5
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
0
-0.5
LHCb-PAPER-2013-019
-1
0
5
10
15
2
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
2
S3 is asymmetry between two
transverse
K ∗0
amplitudes
(averaged B 0 and B̄ 0 )
A6 is asymmetry between Bs0 and
B̄s0 of real part of transverse
amplitudes interference
Both agree with SM
20
4
1
LHCb
LHCb
0.5
0.5
0
0
-0.5
-0.5
-1
22
A6
1
15 July 2013
Bs0 → φµ+ µ−
5
10
15
2
q [GeV2/c4]
Bs0 → φµ+ µ−
-1
5
10
15
q2 [GeV2/c4]
LHCB-PAPER-2013-017
S3
q [GeV /c ]
Michal Kreps – Electroweak penguins with di-leptons
A9
A9
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
B 0 → K ∗0 µ+ µ−
A9
1
LHCb
CDF
LHCb
0.5
0
-0.5
LHCb-PAPER-2013-019
-1
0
5
10
15
q2
A9
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
T-odd asymmetry from interference
of transverse K ∗0 amplitudes
Sensitive when strong phase would
make CP violation zero
High sensitivity to right handed
currents
LHCb
1
[GeV
2
20
/c 4]
1
LHCb
Bs0 → φµ+ µ−
0.5
0.5
0
0
-0.5
-0.5
0
∗0 + −
B →K µ µ
-1
0
23
5
15 July 2013
10
15
-1
20
q2 [GeV2/c 4]
LHCb-PAPER-2013-017
5
10
q2
15
[GeV2/c4]
Michal Kreps – Electroweak penguins with di-leptons
B 0 reparametrisation
One can define alternative set of observables by
S3 = 21 (1 − FL )A2T and AF B = 43 (1 − FL )ARe
T
have reduced
In large recoil limit the observables A2T and ARe
T
form-factor uncertainties
It is not easy to handle correlations while maintaining solid meaning
of uncertainties
→ Extract also A2T and ARe
T
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
Binned
ARe
T
1
Theory
LHCb
LHCb
1
0.5
0.5
0
0
-0.5
-0.5
Theory
LHCb
Binned
LHCb
-1
0
5
10
15
q2
24
15 July 2013
[GeV
2
20
/c 4]
-1
0
5
10
15
q2
[GeV
2
20
LHCB-PAPER-2013-019
A2T
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
/c 4]
Michal Kreps – Electroweak penguins with di-leptons
Prospects
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
Angular distribution for B 0 +B¯0
S1s sin2 θK + S1c cos2 θK +
S2s sin2 θK cos 2θℓ + S2c cos2 θK cos 2θℓ +
S3 sin2 θK sin2 θℓ cos 2φ + S4 sin 2θK sin 2θℓ cos φ +
S5 sin 2θK sin θℓ cos φ + S6 cos2 θK cos θℓ +
A7 sin 2θK sin θℓ sin φ + A8 sin 2θK sin 2θℓ sin φ +
A9 sin2 θK sin2 θℓ sin 2φ
00
11
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
00
11
000
111
00
11
111
000
00
11
000
111
00
11
000
111
000
111
00
11
111
000
00
11
000
111
00
11
000
111
00
11
000
111
000
111
25
S1c,s and S2c,s related to FL
S6 generates AF B
There are still few terms we did not touch
Watch for EPS talk of Nicola Serra and LHCb-PAPER-2013-037
In several terms it can be interesting to measure both sum of B 0 and
B̄ 0 and also difference between B 0 and B̄ 0
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Prospects
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
AI
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
1
0.5
Isospin asymmetry (difference between B + and B 0 rate) expected to
be close to zero
With 1 fb−1 there is an effect at large q 2 for B → Kµ+ µ−
Working on update of this analysis
In current dataset we expect about 3.7k B + → K + µ− µ+ events
Useful to remember that low mass resonances (φ, ω) decay to two
muons with B ≈ 10−5 –10−4
Also charmonia above open charm threshold can go to two muons
B → K µ +µ -
Candidates / (10 MeV/c2)
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
LHCb
0
-0.5
-1
-1.5
0
5
10
15
20
25
15 July 2013
(e)
LHCb
B+→ K +µ +µ -
250
200
150
1250 ± 42
100
50
0
q2 [GeV2/c4]
26
300
5200
5300
5400
5500
mK +µ +µ -
5600
5700
[MeV/c2]
Michal Kreps – Electroweak penguins with di-leptons
Prospects
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
Candidates per 10 MeV/c2
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
For Λb → Λµ+ µ− yield in highest q 2 is larger than for Bs → φµ+ µ−
Selection is optimized globaly, but signal at high q 2 and bg at low q 2
There is room for improvement
At high q 2 we could start to look for other observables
Easy to do AF B by counting, but doing more needs work to define
observables and corresponding angular distributions
15
[0.00-2.00] GeV2/ c4
15 July 2013
15 15
[2.00-4.30] GeV2/ c4
10
10 10
5
5
55
0
15
Λb→Λµ+µ[4.30-8.68] GeV2/ c4
10
0
2
[10.09-12.86] GeV
/ c415
0 0
2
[14.18-16.00] GeV
/c415
10
10
10
5
5
5
0
27
15
LHCb
5.4
5.6
0
5.8 5.4
5.6
5.8
0
[16.00-20.30] GeV2/c4
5.4
5.6
+ -
M(Λµ µ )
5.8
[GeV/ c2]
Michal Kreps – Electroweak penguins with di-leptons
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
900
Candidates / (5 MeV)
Candidates / (10 MeV)
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
Λb → Λµ+ µ− difficult due to long
Λ lifetime
Should be easier when looking to
excited Λ → pK
In J/ψ channel we see much larger
signal
Unfortunately no region dominated
by single resonance
LHCb
800
700
600
500
400
300
200
28
350
LHCb
Λb → J/ψ Λ
300
250
200
150
100
50
0
5.4
5.5
5.6
5.7
5.8
M(Λµ µ ) [GeV/c2]
+ -
6000
5000
LHCb
4000
3000
2000
1000
100
0
Candidates per 5 MeV/c2
Prospects
1500
15 July 2013
2000
2500
m(pK - ) [MeV]
0
5500
5600
5700
-
5800
m(J/ψpK ) [MeV]
Michal Kreps – Electroweak penguins with di-leptons
Prospects
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
I used to work on CPV in Bs → J/ψφ decays, so one crazy speculative
idea from there
When discussing Bs angular analysis, I mentioned non-zero ∆Γs
Theory typically predicts what happens at t = 0
But experiment measures in this time-integrated mode average over
decay time
Our flavour tagging is not good enough to sensibly separate Bs from
B̄s
In Bs → J/ψφ there is access to CPV even without flavour tagging
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
29
This comes from interference of CP-even with CP-odd final state
and non-zero ∆Γs
With some luck we perhaps might get access to more observables also
in Bs → φµ+ µ− once decay time dependence is taken into account
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons
Conclusions
Presented latest results on rare EWK decays
First time we did multidimensional angular analysis of B 0 →
K ∗0 µ+ µ− (LHCb-PAPER-2013-019)
First convincing signal for B 0 → K ∗0 e+ e− at low q 2
(LHCb-PAPER-2013-005)
First angular analysis of Bs → φµ+ µ− (LHCb-PAPER-2013-017)
Start to be in stage to do more with Λb → Λµ+ µ−
(LHCb-PAPER-2013-025)
All results are worlds best
All consistent with SM
→ But placing strong constraints on new physics
Three new results in this area will be shown at EPS
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
111
000
000
111
000
111
000
111
000
111
000
111
000
111
000
111
000
111
30
B 0 theory based on JHEP 07 (2011) 067, NP B612 (2001) 25, PR D70 (2004)
114005, EPJ C71 (2011) 1635, PR D71 (2005) 014029, JHEP 11 (2008) 032
Bs0 prediction from JHEP 01 (2009) 019, JHEP 0807 (2008) 106, PR D71 (2005)
014029
15 July 2013
Michal Kreps – Electroweak penguins with di-leptons