Test of lepton flavour universality in b s + - decays at...
Transcript of Test of lepton flavour universality in b s + - decays at...
Test of lepton flavour universalityin b → s`+`− decays at LHCb
Razvan-Daniel (Dan) Moise,
on behalf of the LHCb collaboration
Moriond EW 2021
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 1 / 12
Lepton Flavour Universality (LFU) & rare decays
Standard Model (SM) couplings of leptons to vectorbosons are flavour-independent.
any sign of lepton flavour non-universalitywould indicate new physics (NP)
Decays of b-quarks are sensitive to LFU.
example: the rare process b → s`+`−— flavour-changing neutral current (FCNC)
SM only allows FCNCs through loop, but NP can enter at tree level or contribute to theloop (even at high mass scales) ⇒ potential LFU violation
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 2 / 12
Anomalous flavour observablesA pattern of interlinked anomalies has emerged in studies of b → s`+`− processes.
branching fractions of e.g.I Bs → µ+µ− (see Flavio’s talk)
I B → K (∗)µ+µ− [JHEP 06 (2014) 133]
angular observables in e.g. B0 → K ∗0µ+µ− [PRL 125 (2020) 011802]
ratios of branching fractions, such as RK∗0 [JHEP 08 (2017) 055]
]4c/2 [GeV2q0 5 10 15 20
]2/G
eV4 c ×
-8 [
102 q
/dBd 0
1
2
3
4
5
LCSR Lattice Data
LHCb
−µ+µ+ K→+B
[JHEP 06 (2014) 133]
0 5 10 15]4c/2 [GeV2q
1−
0.5−
0
0.5
1
5'P
(1S)
ψ/J
(2S)
ψ
LHCb Run 1 + 2016SM from DHMV
[PRL 125 (2020) 011802]
0 1 2 3 4 5 6
q2 [GeV2/c4]
0.0
0.2
0.4
0.6
0.8
1.0
RK∗0
LHCb
LHCb
BIP
CDHMV
EOS
flav.io
JC
[JHEP 08 (2017) 055]
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 3 / 12
The observable RK
One of the anomalous LFU observables is:
RK =
∫ 6.0 GeV2
1.1 GeV2
dB(B+ → K+µ+µ−
)dq2 dq2
∫ 6.0 GeV2
1.1 GeV2
dB(B+ → K+e+e−
)dq2 dq2
SM= 1±O(10−2) EM correction1
Previous LHCb measurement is in tension with SMprediction at level of 2.5 σ [PRL 122 (2019) 191801].
obtained using data collected up until 2016
subject of this talk: update of the previousmeasurement, with full Run 1 and Run 2 data
I adding 2017 & 2018 data effectively doubles the dataset
I following essentially identical procedure
q2 def= dilepton invariant mass squared
[PRL 122 (2019) 191801]
1 [JHEP 06 (2016) 092] [JHEP 07 (2007) 040] [EPJC 76 (2016) 440] [PRD 69 (2004) 074020] [PRD 68 (2003) 094016] [EOS] [flavio]
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 4 / 12
Challenge: muons vs. electrons at LHCb
Bulk of the analysis is dedicated to getting differencesbetween electron and muon detection under control.
different trigger and particle identification strategies
electrons lose significant portions of energy tobremsstrahlung radiation
7 worse mass resolution and reconstruction efficiencycf. muons
3 mitigated by e.g. recovery algorithm (look forphoton clusters in calorimeter that are compatiblewith electron trajectory)
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 5 / 12
Measuring RK
Control electron-muon differences through double ratio between rare B+ → K+`+`− modesand control B+ → K+J/ψ(`+`−) channels:
RK =
(Nµµ
rare
εµµrare/Nee
rare
εeerare
)
/(Nµµ
control
εµµcontrol
/Nee
control
εeecontrol
)︸ ︷︷ ︸
rJ/ψ
=
(Nµµ
rare
Nµµcontrol
/εµµrare
εµµcontrol
)/(Nee
rare
Neecontrol
/εeerare
εeecontrol
)
Five main steps:
1 obtain electron and muon data through tailored selection
2 calibration of simulation → efficiency calculation
3 estimation of systematic uncertainties
4 conduct cross-checks, such as rJ/ψ (known LFU)
5 fit rare data → extraction of RKRare and control selections identical up tocut on q2
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 6 / 12
Measuring RK
Control electron-muon differences through double ratio between rare B+ → K+`+`− modesand control B+ → K+J/ψ(`+`−) channels:
RK =
(Nµµ
rare
εµµrare/Nee
rare
εeerare
)/(Nµµ
control
εµµcontrol
/Nee
control
εeecontrol
)︸ ︷︷ ︸
rJ/ψ
=
(Nµµ
rare
Nµµcontrol
/εµµrare
εµµcontrol
)/(Nee
rare
Neecontrol
/εeerare
εeecontrol
)
Five main steps:
1 obtain electron and muon data through tailored selection
2 calibration of simulation → efficiency calculation
3 estimation of systematic uncertainties
4 conduct cross-checks, such as rJ/ψ (known LFU)
5 fit rare data → extraction of RKRare and control selections identical up tocut on q2
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 6 / 12
Measuring RK
Control electron-muon differences through double ratio between rare B+ → K+`+`− modesand control B+ → K+J/ψ(`+`−) channels:
RK =
(Nµµ
rare
εµµrare/Nee
rare
εeerare
)/(Nµµ
control
εµµcontrol
/Nee
control
εeecontrol
)︸ ︷︷ ︸
rJ/ψ
=
(Nµµ
rare
Nµµcontrol
/εµµrare
εµµcontrol
)/(Nee
rare
Neecontrol
/εeerare
εeecontrol
)
Five main steps:
1 obtain electron and muon data through tailored selection
2 calibration of simulation → efficiency calculation
3 estimation of systematic uncertainties
4 conduct cross-checks, such as rJ/ψ (known LFU)
5 fit rare data → extraction of RKRare and control selections identical up tocut on q2
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 6 / 12
Efficiency calculation and corrections to simulation
Efficiencies estimated from simulated samples, which are calibrated using control data.Calibration procedure is the same as in [PRL 122 (2019) 191801], and it covers:
trigger performance;
particle identification efficiency;
I method described in [EPJ T&I (2019) 6:1)]
B+ kinematics;
resolutions of q2 and m(K+e+e−). ) [MeV]e(TE0 2000 4000 6000 8000 10000
(L0E
lect
ron)
[%
]ε
0
1020
304050
6070
80
90100
LHCb [PRL 122 (2019) 191801]
Example measurement of electron trigger
performance
This leads to %-level control of efficiency ratios.
verified through a host of cross-checks
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 7 / 12
Cross-check example: the single ratio rJ/ψ
rJ/ψ =B(B+ → K+J/ψ(µ+µ−))
B(B+ → K+J/ψ(e+e−))=
Nµµcontrol
εµµcontrol
/Nee
control
εeecontrol
Single ratio requires control of electrons with respect to muons.
stringent cross-check of efficiencies & yields
Result: rJ/ψ = 0.981± 0.020 (stat. & syst.)
Also compute rJ/ψ in bins of variables relevant to detector response(1D rJ/ψ)
if deviations from flatness are genuine, impact on RK is withinestimated systematic uncertainty
]c) [MeV/+B(T
p0 5000 10000 15000
⟩ ψ
J/ r⟨
/ ψ
J/r
0.9
0.95
1
1.05
1.1
LHCb[LHCb-PAPER-2021-004]
1D rJ/ψ , binned in B+pT
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 8 / 12
Cross-check: the double ratio Rψ(2S)
Select data at ψ(2S) resonance and evaluate:
Rψ(2S) =B(B+ → K+ψ(2S)(µ+µ−))
B(B+ → K+ψ(2S)(e+e−))
/B(B+ → K+J/ψ(µ+µ−))
B(B+ → K+J/ψ(e+e−))
expected close to unity
test efficiency corrections at q2 away from J/ψ
independent validation of double ratio
Result: Rψ(2S) = 0.997± 0.011 (stat. & syst.)
]2c [MeV/)−µ+µ+(K(2S)ψm5200 5300 5400 5500 5600
)2 cC
andi
date
s / (
4 M
eV/
5000
10000
15000
20000
25000
-1Data 4 fb
Total fit
+)K−µ+µ(2S)(ψ →+B
Combinatorial
LHCb
Fit to 2017 & 2018 B+ → K+ψ(2S)(µ+µ−) data
]2c [MeV/)−e+e+(K(2S)ψm5200 5400 5600
)2 cC
andi
date
s / (
12 M
eV/
2000
4000
6000
8000
10000-1Data 4 fb
Total fit+)K−e+(2S)(eψ →+B*0)K−e+(2S)(eψ →0B
+)K−e+(eψ J/→+B+K−e+ e→+B
+ X) Kψ(2S) (J/ψ →+B(K X)
s X) Hψ (J/c X→B
Combinatorial
LHCb
Fit to 2017 & 2018 B+ → K+ψ(2S)(e+e−) data
Dilepton mass constrained to mass of ψ(2S)
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 9 / 12
Fits to B+ → K+`+`− dataRK is extracted as a parameter of a simultaneous fit to all B+ → K+`+`− data.
]2c [MeV/)−µ+µ+m(K5200 5300 5400 5500 5600
)2 cC
andi
date
s / (
7 M
eV/
0
100
200
300
400
500
600-1Data 9 fb
Total fit−µ+µ+ K→+B
Combinatorial
LHCb
[LHCb-PAPER-2021-004]
N(K+µ+µ−) ∼ 3850
]2c [MeV/)−e+e+m(K5000 5500 6000
)2 cC
andi
date
s / (
24 M
eV/
020406080
100120140160180200220240
-1Data 9 fbTotal fit
−e+e+ K→+B+)K−e+(eψ J/→+B
Part. Reco.Combinatorial
LHCb
N(K+e+e−) ∼ 1640
correlations between selection efficiencies taken into account
shape parameters derived from calibrated simulation
systematic effect of chosen signal & background models is ∼ 1%
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 10 / 12
Fits to B+ → K+`+`− dataRK is extracted as a parameter of a simultaneous fit to all B+ → K+`+`− data.
]2c [MeV/)−µ+µ+m(K5200 5300 5400 5500 5600
)2 cC
andi
date
s / (
7 M
eV/
0
100
200
300
400
500
600-1Data 9 fb
Total fit−µ+µ+ K→+B
Combinatorial
LHCb
[LHCb-PAPER-2021-004]
N(K+µ+µ−) ∼ 3850
]2c [MeV/)−e+e+m(K5000 5500 6000
)2 cC
andi
date
s / (
24 M
eV/
020406080
100120140160180200220240
-1Data 9 fbTotal fit
−e+e+ K→+B+)K−e+(eψ J/→+B
Part. Reco.Combinatorial
LHCb
N(K+e+e−) ∼ 1640
correlations between selection efficiencies taken into account
shape parameters derived from calibrated simulation
systematic effect of chosen signal & background models is ∼ 1%
And now, the final result
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 10 / 12
RK with full Run 1 and Run 2 LHCb data
The measured value of RK is:
RK = 0.846 +0.042−0.039 (stat.) +0.013
−0.012 (syst.)
dominant systematic effect: fit modelI effects such as calibration of trigger & kinematics
are at permille-level
p-value under SM hypothesis: 0.0010
significance: 3.1σ (evidence for LFU violation inB+ → K+`+`− decays)
0.5 1 1.5KR
-1LHCb 9 fb4c/2 < 6.0 GeV2q1.1 <
Belle4c/2 < 6.0 GeV2q1.0 <
BaBar4c/2 < 8.12 GeV2q0.1 <
[PRD 86 (2012) 02]
[JHEP 03 (2021) 105]
[LHCb-PAPER-2021-004]
0.7 0.8 0.9 1 1.1
KR
0
2
4
6
8
10
12
14 )m
inL
/ L
ln(
−Pr
ofile
of
LHCb-19 fb
[LHCb-PAPER-2021-004]
— 1σ— 3σ— 5σ
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 11 / 12
Other results
0 5 10 15 20]4c/2 [GeV2q
0
1
2
3
4
5
]2/G
eV4 c ×
-8 [
102 q
/dBd
LHCb SM prediction-1electrons 9fb-1muons 3fb
Combining measured RK value with B(B+ → K+µ+µ−) result from [JHEP 06 (2014) 133] gives:
B(B+ → K+e+e−) = (28.6 +1.5−1.4(stat.) ± 1.3(syst.))× 10−9
in the range q2 ∈ [1.1, 6.0] GeV2/c4.
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 11 / 12
Summary
performed most precise measurement of the LFU ratio RK using full LHCb Run 1 andRun 2 data
gained factor ∼ 2 statistics cf. previous result [PRL 122 (2019) 191801]
3.1σ tension with SM prediction1
I evidence for violation of LFU in B+ → K+`+`− decays
LHCb will continue to study flavour anomalies, including LFU-sensitive observables
paper submitted to Nature Physics [LHCb-PAPER-2021-004]
1 [JHEP 06 (2016) 092] [JHEP 07 (2007) 040] [EPJC 76 (2016) 440] [PRD 69 (2004) 074020] [PRD 68 (2003) 094016] [EOS] [flavio]
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
BACKUP
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
K+`+`− final states at LHCb
5.0 5.2 5.4 5.6 5.8 6.0]2c [GeV/)−µ+µ+m(K
0
5
10
15
20
25]4 c/2 [
GeV
2 q
1
10
210
310
410
510LHCb
4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0]2c [GeV/)−e+e+m(K
0
5
10
15
20
25]4 c/2 [
GeV
2 q
1
10
210
310
410
510LHCb
[PRL 122 (2019) 191801]
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
Semileptonic background vetoes
]2c) [MeV/−e+K(m1000 2000 3000 4000 5000
Nor
mal
ised
dis
trib
utio
n
5−10
4−10
3−10
2−10
1−10
−e+e+ K→+B
eν +) eeν −e+ K→(0
D →+B
]−e →[−π )eν −e+ K→(
0D →+B
eν +) e ]−e →[−π+ K→(
0D →+B
LHCb simulation
]2c) [MeV/]−π →[−e+K(m
1700 1800 1900 2000
Nor
mal
ised
dis
trib
utio
n
0.00
0.01
0.02
0.03
0.04
0.05 LHCbsimulation
[LHCb-PAPER-2021-004]
Backgrounds coming from b → c`−ν` transitions removed by mass vetoes.
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
Trigger strategy
Same approach as in the previous analysis:
for µµ channels, trigger on muons: L0Muon
for ee channels, use three exclusive trigger categories: L0Electron,L0Hadron, L0TIS
systematics calculated and cross-checks performed foreach trigger individually
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
Fits to control data: muons
]2c) [MeV/−µ+µ+K(ψJ/m5200 5300 5400 5500 5600
)2 cC
andi
date
s / (
4 M
eV/
10
210
310
410
510 -1Data 5 fbTotal fit
+)K−µ+µ(ψ J/→+B+π)−µ+µ(ψ J/→+B
Combinatorial
LHCb
Previous data
]2c) [MeV/−µ+µ+K(ψJ/m5200 5300 5400 5500 5600
)2 cC
andi
date
s / (
4 M
eV/
10
210
310
410
510 -1Data 4 fbTotal fit
+)K−µ+µ(ψ J/→+B+π)−µ+µ(ψ J/→+B
Combinatorial
LHCb
New data
[LHCb-PAPER-2021-004]
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
Fits to control data: electrons
]2c) [MeV/−e+e+K(ψJ/m5200 5400 5600
)2 cC
andi
date
s / (
12 M
eV/
210
310
410
-1Data 5 fbTotal fit
+)K−e+(eψ J/→+BPart. Reco.
+π)−e+(eψ J/→+BCombinatorial
LHCb
]2c) [MeV/−e+e+K(ψJ/m5200 5400 5600
)2 cC
andi
date
s / (
12 M
eV/
210
310
410
-1Data 4 fbTotal fit
+)K−e+(eψ J/→+BPart. Reco.
+π)−e+(eψ J/→+BCombinatorial
LHCb
Electron trigger
]2c) [MeV/−e+e+K(ψJ/m5200 5400 5600
)2 cC
andi
date
s / (
12 M
eV/
210
310
410-1Data 5 fb
Total fit+)K−e+(eψ J/→+B
Part. Reco.+π)−e+(eψ J/→+B
Combinatorial
LHCb
]2c) [MeV/−e+e+K(ψJ/m5200 5400 5600
)2 cC
andi
date
s / (
12 M
eV/
210
310
410-1Data 4 fb
Total fit+)K−e+(eψ J/→+B
Part. Reco.+π)−e+(eψ J/→+B
Combinatorial
LHCb
Hadron trigger
]2c) [MeV/−e+e+K(ψJ/m5200 5400 5600
)2 cC
andi
date
s / (
12 M
eV/
210
310
410 -1Data 5 fbTotal fit
+)K−e+(eψ J/→+BPart. Reco.
+π)−e+(eψ J/→+BCombinatorial
LHCb
]2c) [MeV/−e+e+K(ψJ/m5200 5400 5600
)2 cC
andi
date
s / (
12 M
eV/
210
310
410 -1Data 4 fbTotal fit
+)K−e+(eψ J/→+BPart. Reco.
+π)−e+(eψ J/→+BCombinatorial
LHCb
Independent trigger
[LHCb-PAPER-2021-004]
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
Cross-check: 2D rJ/ψ
The 1D rJ/ψ cross-check is extended to two variables.
deviations from flatness again covered by systematic uncertainty on RK
rJ/ψ tests show efficiencies understood across phase space
) bin number−l, +l(α ×)) −l(p)), +l(pmax(1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
⟩ ψ
J/ r⟨
/ ψ
J/r
0.9
1
1.1 LHCb[LHCb-PAPER-2021-004]
2D rJ/ψ , binned in max lepton momentum andangle between leptons
K+`+`− & K+J/ψ distributions
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
Additional rJ/ψ checks
]c)) [MeV/−l(T
p), +l(T
pmin(1000 2000 3000 4000 5000
Can
dida
tes
(arb
itrar
y un
its)
0
0.2
0.4
0.6
0.8
1
1.2 LHCb simulation
) [rad]−l, +l(α0 0.1 0.2 0.3 0.4 0.5
Can
dida
tes
(arb
itrar
y un
its)
0
0.2
0.4
0.6
0.8
1
1.2 LHCb simulation
]c)) [MeV/−l(T
p), +l(T
pmin(1000 2000 3000 4000 5000
⟩ ψ
J/ r⟨
/ ψ
J/r
0.9
0.95
1
1.05
1.1
LHCb
) [rad]−l, +l(α0 0.1 0.2 0.3 0.4 0.5
⟩ ψ
J/ r⟨
/ ψ
J/r
0.9
0.95
1
1.05
1.1
LHCb
[LHCb-PAPER-2021-004]Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
Distributions of rare & control samples (I)
) [rad]−l, +l(α0 0.1 0.2 0.3 0.4 0.5
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation
) [rad]−l, +K(α0 0.1 0.2 0.3 0.4 0.5
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation
) [rad]+l, +K(α0 0.1 0.2 0.3 0.4 0.5
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation
]c) [MeV/+K(T
p2000 4000 6000 8000
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation
]c)) [MeV/−l(T
p), +l(T
pmax(2000 4000 6000 8000 10000
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation −e+e+ K→+B−µ+µ+ K→+B
+)K−e+(eψ J/→+B+)K−µ+µ(ψ J/→+B
[PRL 122 (2019) 191801]
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12
Distributions of rare & control samples (II)
)+K(η2 3 4 5
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation
))−l(η), +l(ηmin(2 3 4 5
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation
))−l(η), +l(ηmax(2 3 4 5
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation
))+B(IP2χ(
10log
4− 2− 0 2
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation
))+B(Vtx2χ(
10log
2− 0 2
Can
dida
tes
/ (a.
u.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 LHCb simulation −e+e+ K→+B−µ+µ+ K→+B
+)K−e+(eψ J/→+B+)K−µ+µ(ψ J/→+B
[PRL 122 (2019) 191801]
Dan Moise (Imperial College London) Test of LFU at LHCb 23rd March 2021 12 / 12