The Belle Experiment Belle recorded 711 fb 1 on the ( 4S ) - - PowerPoint PPT Presentation
Search for B and B and Test of Lepton Universality with R ( K ) at Belle Moriond EW 2019 Markus Prim for the Belle Collaboration | 22nd March 2019 INSTITUT F UR EXPERIMENTELLE TEILCHENPHYSIK (ETP) www.kit.edu KIT
INSTITUT F ¨ UR EXPERIMENTELLE TEILCHENPHYSIK (ETP)
Moriond EW 2019 Markus Prim for the Belle Collaboration | 22nd March 2019
KIT – University of the State of Baden-W¨ urttemberg and National Laboratory of the Helmholtz Association
www.kit.edu
Belle recorded 711 fb−1 on the Υ(4S) resonance.
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 2/23
Inclusive Tag
ǫ = O(100)%
Consistency of Btag ? Semileptonic Tag
ǫ = O(1)%
Knowledge of Btag
B D
ℓ ν
Hadronic Tag
ǫ = O(0.1)%
Exact knowledge of Btag
B
Purity Efficiency ǫ
Exclusive Tagging: The Full Event Interpretation (FEI)
Keck, T., et al. Comput Softw Big Sci (2019) Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 3/23
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 4/23
dΓ dEγ
= αemG2
Fm4
B|Vub|2
48π2 x3
γ(1−xγ)[F 2
A+F 2 V]
with xγ = 2Eγ/mB
Previous Belle result (2015)b:
∆B(B+ → ℓ+νℓγ) < 3.5 × 10−6
upper limit Belle 2015
Form Factors (valid for large photon energies)
FV (Eγ)
=
QumBfB 2EγλB R(Eγ, µ) +
2Eγmb
+ QumBfB (2Eγ)2
=
QumBfB 2EγλB R(Eγ, µ) +
2Eγmb
− QumBfB (2Eγ)2 + QℓfB
Eγ
λB (GeV)
QCD factorization
≈ 0.2
QCD sum rules 0.46 ± 0.11 BaBar (2009)a
> 0.115
Belle (2015)b
> 0.238
a: Phys. Rev. D 80, 111105 (2009) b: Phys. Rev. D 91, 112009 (2015) Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 5/23
Reconstruction & Selection B → ℓνγ with hadronic tagging (FEI) Background Suppression Multivariate Methods B+ → π0ℓ+νℓ B+ → ηℓ+νℓ e+e− → qq Control Region B+ → π0ℓ+νℓ Signal Extraction Likelihood Fit
λB Determination
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 6/23
B+ → e+νeγ B+ → µ+νµγ
ℓ B(B+ → π0ℓ+νℓ) (10−5) ∆B(B+ → ℓ+νℓγ) (10−6) σ
e 8.3+0.9
−0.8 ± 0.9
1.7+1.6
−1.4 ± 0.7
1.1
µ
7.5 ± 0.8 ± 0.6 1.0+1.4
−1.0 ± 0.4
0.8 e, µ 7.9 ± 0.6 ± 0.6 1.4 ± 1.0 ± 0.4 1.4
Previous results for B+ → π0ℓ+νℓ B(B+ → π0ℓ+νℓ) (×10−5) Belle excl.a 8.0 ± 0.8 ± 0.4 PDG 7.80 ± 0.27 a: Sibidanov et al., Phys. Rev. D 88, 032005 (2013) Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 7/23
Rπ = ∆B(B+ → ℓ+νℓγ)
B(B+ → π
0ℓ+νℓ)
= ∆Γ(λB) Γ(B+ → π
0ℓ+νℓ)
Rπ removes dependence of Vub. R
measured π
= (1.7 ± 1.4) × 10−2
based on theoretical input from: Beneke et al., JHEP 07:154 (2018) HFLAV, Eur. Phys. J., C77:895, (2017)
λB > 0.24 GeV
@ 90%C.L.
a a: Phys.Rev. D98 (2018) no.11, 112016
Result of Belle (2015) was λB > 0.238 GeV
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 8/23
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 9/23
B → µνµ (and B → eνe) yet to be observed. Both helicity and CKM suppressed. Pure leptonic channels are a golden mode to extract |Vub|. Highly sensitive to new physics.
B(B → ℓνℓ)SM =
G2
FmBm2 ℓ
8π
m2
ℓ
m2
B
2
f 2
B |Vub|2 τB
B(B → µνµ)SM = (4.26 ± 0.81) × 10−7
Previous Belle result (2018)a:
a: Phys. Rev. Lett. 121, 031801 (2018)
B(B → µνµ) = (6.64 ± 2.22 ± 1.60) × 10−7 @ 2.4 σ
b u µ νµ W+ B+
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 10/23
New Physics Scenario: 2 Higgs Doublet Model. Change of experimental signature:
Branching fraction due to interference.
B(B → ℓνℓ)a
Type II = BSM ×
m2
B tan2 β
m2
H+
B(B → ℓνℓ)b
Type III = BSM ×
m2
B
mbmℓ CR − CL CSM
a: W.-S. Hou, Phys. Rev. D48, 2342 (1993) b: Phys.Rev. D86 (2012) 054014
b u µ νµ W+ B+ b u µ νµ H+ B+
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 11/23
New Physics Scenario: Sterile Neutrino in the final state. Change of experimental signature:
Branching fraction due to additional decay channel:
B(B → µ + missing energy) = B(B → µνµ) + B(B → µN).
Shift in momentum spectrum due to sterile neutrino mass.
b u µ νµ W+ B+ b u µ N H+ B+ b u N µ LQ B+
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 12/23
Reconstruction & Selection B → µνµ with Inclusive Tagging Background Suppression Multivariate Method e+e− → qq Control Region Classifier Sideband Control Channel B → D[→ Kπ]π Signal Extraction Likelihood Fit BR Determination
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 13/23
2.2 2.4 2.6 2.8 3.0 3.2 pB / (GeV) 50 100 150 Entries / (0.050 GeV)
Signal Enriched I
Continuum b c Rare B b u Signal
Data
Preliminary 2.2 2.4 2.6 2.8 3.0 3.2 pB / (GeV) 50 100 150 200 250 Entries / (0.050 GeV)
Signal Enriched II
Continuum b c Rare B b u Signal
Data
Preliminary 2.2 2.4 2.6 2.8 3.0 3.2 pB / (GeV) 200 400 600 Entries / (0.050 GeV)
Background Enriched I
Continuum b c Rare B b u Signal
Data
Preliminary 2.2 2.4 2.6 2.8 3.0 3.2 pB / (GeV) 200 400 600 800 1000 Entries / (0.050 GeV)
Background Enriched II
Continuum b c Rare B b u Signal
Data
Preliminary
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 14/23
B(B → µνµ)
SM 4.26 × 10−7 Belle (2018)a
(6.6 ± 2.2 ± 1.6) × 10−7 @ 2.4 σ
Result (this)
(5.3 ± 2.0 ± 0.9) × 10−7 @ 2.8 σ
Frequentist UL
< 8.64 × 10−7 @ 90% CL
a: Sibidanov et al., Phys. Rev. Lett. 121, 031801 (2018)
Extracted Signal Yield
50 100 150 200
sig
1 2 3 4 5 (
sig)
stat.
sig = 117±48
Preliminary Search for B → ℓν γ and B → µν µ and Test of Lepton Universality with R(K∗) at Belle - Markus Prim 22nd March 2019 15/23
Type II
200 400 600 800 1000 mH + / (GeV) 50 100 150 200 250 300 tan Preliminary
95% CL
B B b s
B(B → ℓνℓ) = BSM ×
m2
B tan2 β
m2
H+
Type III
0.2 0.1 0.0 0.1 0.2 SL 0.2 0.1 0.0 0.1 0.2 SR Preliminary
95% CL
B B
B(B → ℓνℓ) = BSM×
m2
B
mbmℓ CR − CL CSM
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 16/23
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 Sterile Neutrino Mass mN / (GeV) 10
2
10
1
100 local p0 2 1
B N stat. + sys.
Preliminary
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 17/23
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 18/23
Test of lepton flavor universality via: RK∗ = B(B → K∗µ+µ−)
B(B → K∗e+e−) ≈ 1
Clean theoretical predictions. Currently O(2.3σ) tension. New physics can change this ratio. Decay modes used in this analysis: B0 → K∗0ℓ+ℓ− B+ → K∗+ℓ+ℓ− K∗0 → K+π− K∗+ → K+π0 K∗+ → K0
Sπ+ b s d d ℓ− ℓ+ W+ Z0/γ B K∗ b s d d ℓ− ℓ+ Z′ B K∗
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 19/23
Reconstruction & Selection B → K∗ℓ+ℓ− with Hierachical NN Reconstruction Background Suppression Multivariate Method utilizing event topology and NN outputs Control Region B → J/ψK∗ Signal Extraction Likelihood Fit R(K∗) Determination
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 20/23
Example fit for q2 > 0.045 GeV2. 103.0+13.4
−12.7 (139.0+16.0 −15.4) events in the electron (muon) modes.
/ 400 )
2
Events / ( GeV/c 10 20 30 40 50 60 )
2
(GeV/c
bc
M 5.22 5.24 5.26 5.28 Pull 2 − 2 / 400 )
2
Events / ( GeV/c 10 20 30 40 50 60 70 80 90 )
2
(GeV/c
bc
M 5.22 5.24 5.26 5.28 Pull 2 − 2
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 21/23
5 10 15 20
q2 (GeV2/c4)
0.0 0.5 1.0 1.5 2.0
RK ∗
Belle preliminary Data for B 0 and B + modes SM prediction 5 10 15 20
q2 (GeV2/c4)
0.0 0.5 1.0 1.5 2.0
RK ∗
Belle preliminary
Belle LHCb BaBar SM prediction
q2 in GeV2/c4 All modes B0 modes B+ modes [0.045, 1.1] 0.52+0.36
−0.26 ± 0.05
0.46+0.55
−0.27 ± 0.07
0.62+0.60
−0.36 ± 0.10
[1.1, 6] 0.96+0.45
−0.29 ± 0.11
1.06+0.63
−0.38 ± 0.13
0.72+0.99
−0.44 ± 0.18
[0.1, 8] 0.90+0.27
−0.21 ± 0.10
0.86+0.33
−0.24 ± 0.08
0.96+0.56
−0.35 ± 0.14
[15, 19] 1.18+0.52
−0.32 ± 0.10
1.12+0.61
−0.36 ± 0.10
1.40+1.99
−0.68 ± 0.11
[0.045, ] 0.94+0.17
−0.14 ± 0.08
1.12+0.27
−0.21 ± 0.09
0.70+0.24
−0.19 ± 0.07
All measured values are in accordance with the SM and
First measurement of R(K∗+).
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 22/23
B → ℓνγ
Most stringent upper limit on ∆B(B → ℓνγ) < (1.4 ± 1.0 ± 0.4) × 10−6.
λB > 0.24 GeV. B → µνµ
Most precise measurement of B(B → µνµ) = (5.3 ± 2.0 ± 0.9) × 10−7 @ 2.8 σ. Improved Limits on Type III 2HDM. Scan for sterile neutrino.
R(K ∗)
R(K∗) measurement compatible with SM. First measurement of R(K∗+).
R(D(∗))
Measurement of R(D) and R(D*) with a semileptonic tag in Belle by Giacomo Caria.
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 23/23
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 24/23
B(B+ → π0ℓ+νℓ) ∆B(B+ → ℓ+νℓγ)
Source in 10−5 in 10−6 Calibration
±0.49 ±0.09
Reconstruction efficiency
±0.20 ±0.01 LLID efficiency ±0.16 ±0.02
NBB
±0.11 ±0.02
Tracking efficiency
±0.03 ±0.0
Peaking background BDT
±0.02 ±0.24
PDF templates
±0.08 ±0.18
BCL model
±0.25 ±0.01
Reconstructed tag channel
±0.01 ±0.14
B → Xuℓ+νℓ
±0.02 ±0.07
Signal model
±0.00 ±0.03
Combined
±0.62 ±0.36
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 25/23
Source of uncertainty Fractional uncertainty Additive uncertainties B+ → µ+ νµ MC statistics 1.0 % b → u ℓ νℓ modeling 11 % b → c ℓ νℓ modeling 2.5 %
B(b → s) processes
1.0 %
B(B+ → µ+ νµ γ)
0.1 % Continuum modeling 13 % Multiplicative uncertainties NB¯
B
1.4 % Tracking efficiency 0.3 %
LLID efficiency
2.0 % total syst. uncertainty 17 %
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 26/23
q2 in GeV2 / c4 e, µ eff. MC size classifier
tracking peaking bkg. c¯ c bkg. total all modes
0.061 0.004 0.013 0.008 0.016 0.031 0.023 0.075
0.058 0.005 0.029 0.002 0.016 0.054 0.051 0.100
0.090 0.012 0.012 0.014 0.020 0.003 0.003 0.095
0.027 0.006 0.008 0.025 0.009 0.026 0.001 0.047
0.065 0.008 0.048 0.033 0.017 0.070 0.013 0.114 B0 modes
0.073 0.006 0.030 0.018 0.022 0.031 0.021 0.092
0.058 0.006 0.040 0.019 0.017 0.033 0.018 0.084
0.091 0.013 0.007 0.012 0.022 0.007 0.001 0.096
0.024 0.007 0.044 0.005 0.009 0.049 0.001 0.071
0.082 0.010 0.040 0.062 0.021 0.070 0.012 0.133 B+ modes
0.044 0.005 0.032 0.018 0.010 0.025 0.023 0.068
0.060 0.010 0.039 0.040 0.014 0.048 0.107 0.144
0.089 0.028 0.016 0.041 0.021 0.008 0.002 0.106
0.033 0.013 0.067 0.060 0.009 0.006 0.000 0.097
0.045 0.010 0.137 0.060 0.011 0.086 0.009 0.179
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 27/23
FEI performance extracted from a mbc fit on converted Belle data. B+
10 20 30 40 50 60 70 80 90 100 Purity in % 0.00 0.10 0.20 0.30 0.40 0.50 Tag-Side Efficiency in % FEI FR
B0
10 20 30 40 50 60 70 80 90 100 Purity in % 0.00 0.10 0.20 0.30 0.40 0.50 Tag-Side Efficiency in % FEI FR
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 28/23
Low Purity
5.24 5.25 5.26 5.27 5.28 Mbc in GeV 1000 2000 3000 4000 Entries FEI FR
High Purity
5.24 5.25 5.26 5.27 5.28 Mbc in GeV 100 200 300 400 Entries FEI FR
Search for B → ℓν γ and B → µν
µ and Test of Lepton Universality with R(K∗) at Belle -
Markus Prim 22nd March 2019 29/23