Lepton Flavour Universality tests with heavy flavour decays at LHCb - - PowerPoint PPT Presentation

▶
lepton flavour universality tests with heavy flavour
SMART_READER_LITE
LIVE PREVIEW

Lepton Flavour Universality tests with heavy flavour decays at LHCb - - PowerPoint PPT Presentation

Apr 21, 2023 401 likes •688 views

Lepton Flavour Universality tests with heavy flavour decays at LHCb Including a new R K result Thibaud Humair, on behalf of the LHCb collaboration Moriond EW 2019 22 nd March, 2019 LFU and b s + decays 1 ' 5 P SM from DHMV

slide-1
SLIDE 1

Lepton Flavour Universality tests with heavy flavour decays at LHCb

Including a new RK result

Thibaud Humair, on behalf of the LHCb collaboration

Moriond EW 2019

22nd March, 2019

slide-2
SLIDE 2

LFU and b → sℓ+ℓ− decays

Yesterday: in talk presented by Carla Marin:

◮ Interesting discrepancies in b → sµ+µ− decays,

e.g. angular analysis of B0 → K ∗0µ+µ−;

◮ But hadronic uncertainties make interpretation

difficult.

]

4

c /

2

[GeV

2

q

5 10 15

5

' P

1 − 0.5 − 0.5 1 SM from DHMV

LHCb Run 1 analysis

JHEP02(2016)104

Today: test Lepton Flavour Universality in b → sℓ+ℓ− decays, in particular RK and RK ∗: RK (∗) = B(B → K (∗)µ+µ−) B(B → K (∗)e+e−)

SM

= 1.0

◮ All hadronic effects cancel in these ratios: immaculate theoretical predictions of RK (∗)

◮ Small deviation from 1, O(1%), due to radiative corrections (EPJC76(2016)440).

⇒ any statistically significant deviation of these ratios from 1 is a sign of New Physics.

2 Thibaud Humair

slide-3
SLIDE 3

Previous RK ∗ and RK results (LHCb Run 1 data)

LHCb: PRL113(2014)151601 BaBar: PRD86(2012)032012 Belle: PRL103(2009)171801

5 10 15 20

q2 [GeV2/c4]

0.0 0.5 1.0 1.5 2.0

RK∗0

LHCb

LHCb BaBar Belle

LHCb: JHEP08(2017)055

All LHCb results below SM expectations:

◮ RK = 0.745+0.090 −0.074 ± 0.036 for 1.0 < q2 < 6.0 GeV2, ∼ 2.6 σ from SM; ◮ RK ∗ = 0.66+0.11 −0.07 ± 0.03 for 0.045 < q2 < 1.1 GeV2, ∼ 2.2 σ from SM; ◮ RK ∗ = 0.69+0.11 −0.07 ± 0.05 for 1.1 < q2 < 6.0 GeV2, ∼ 2.4 σ from SM;

Together with b → sµµ results, RK and RK ∗ constitute an interesting pattern of anomalies, but the significance is still low.

3 Thibaud Humair

slide-4
SLIDE 4

Previous RK ∗ and RK results (LHCb Run 1 data)

LHCb: PRL113(2014)151601 BaBar: PRD86(2012)032012 Belle: PRL103(2009)171801

5 10 15 20

q2 [GeV2/c4]

0.0 0.5 1.0 1.5 2.0

RK∗0

LHCb

LHCb BaBar Belle

LHCb: JHEP08(2017)055

All LHCb results below SM expectations:

◮ RK = 0.745+0.090 −0.074 ± 0.036 for 1.0 < q2 < 6.0 GeV2, ∼ 2.6 σ from SM; ◮ RK ∗ = 0.66+0.11 −0.07 ± 0.03 for 0.045 < q2 < 1.1 GeV2, ∼ 2.2 σ from SM; ◮ RK ∗ = 0.69+0.11 −0.07 ± 0.05 for 1.1 < q2 < 6.0 GeV2, ∼ 2.4 σ from SM;

Together with b → sµµ results, RK and RK ∗ constitute an interesting pattern of anomalies, but the significance is still low.

3 Thibaud Humair

Today: update of the RK measurement in 1.1 < q2 < 6.0 GeV2 In this update:

◮ The analysis of 2011 and 2012 data is completely re-optimised,

the analysis strategy re-designed;

◮ 2015 and 2016 LHCb data are added; ◮ In total, updated analysis uses twice as many B’s as the previous analysis.

LHCb-Paper-2019-009

slide-5
SLIDE 5

RK measurement at LHCb

Need two inputs to measure RK: yields and efficiencies. RK = B(B+ → K +µµ) B(B+ → K +ee) = N(K +µµ) N(K +ee) · ε(K +ee) ε(K +µµ) Electron and muon tracks very different in LHCb:

◮ Electrons interact with material and emit

bremsstrahlung;

◮ worse mass and q2 resolution; ◮ lower reconstruction efficiency.

◮ Better PID and trigger performances for muons.

e track µ track Critical aspect in the analysis: get the electron efficiencies fully under control.

4 Thibaud Humair

slide-6
SLIDE 6

RK measurement at LHCb

Need two inputs to measure RK: yields and efficiencies. RK = B(B+ → K +µµ) B(B+ → K +ee) B(B+ → K +J/ψ(µµ)) B(B+ → K +J/ψ(ee)) = N(K +µµ) N(K +J/ψ(µµ)) · N(K +J/ψ(ee)) N(K +ee) · ε(K +J/ψ(µµ)) ε(K +µµ) · ε(K +ee) ε(K +J/ψ(ee)) Electron and muon tracks very different in LHCb:

◮ Electrons interact with material and emit

bremsstrahlung;

◮ worse mass and q2 resolution; ◮ lower reconstruction efficiency.

◮ Better PID and trigger performances for muons.

e track µ track Critical aspect in the analysis: get the electron efficiencies fully under control. ⇒ use double ratio to cancel out most systematic uncertainties.

4 Thibaud Humair

slide-7
SLIDE 7

Efficiency computation

Ratio of efficiencies determined with simulation carefully calibrated using control channels selected from the data:

◮ Calibration of B+ kinematics; ◮ Tracking efficiency calibration; ◮ Particle ID calibration

(method described in EPJ T&I(2019)6:1);

◮ Trigger calibration (right plot); ◮ Calibration q2 and m(Kee) resolution.

Ratio of efficiencies controlled to an excellent level and checked with alternative samples wherever possible. Detailed evaluation of systematic uncertainties shows uncertainties at each step are < 1%

) [MeV] e (

T

E

2000 4000 6000 8000 10000

(L0Electron) [%] ε

10 20 30 40 50 60 70 80 90 100

LHCb Measurement of the electron trigger ef- ficiency using B+ → J/ψ(e+e−)K + data.

5 Thibaud Humair

slide-8
SLIDE 8

Cross-check 1: rJ/ψ in 1D

To check efficiencies are correct, check: rJ/ψ = B(B → K +J/ψ(µµ)) B(B → K +J/ψ(ee)) = 1.0, Result: rJ/ψ = 1.014 ± 0.035 (stat. + syst.)

◮ Check that efficiencies are understood as a

function of any variable: ⇒ differential rJ/ψ demonstrates it is the case: rJ/ψ is flat for all variables examined.

] c )) [MeV/

−

l (

T

p ),

+

l (

T

p min(

1000 2000 3000 4000 5000

〉

ψ J/

r 〈 /

ψ J/

r

0.90 0.95 1.00 1.05 1.10

LHCb

LHCb-Paper-2019-009 Given expected min(pT(ℓ+), pT(ℓ−) spectra, bias expected on RK if deviations are genuine rather than fluctuations is 0.1%.

6 Thibaud Humair

slide-9
SLIDE 9

Cross-check 2: rJ/ψ in 2D

◮ Pick two variables from those that can be used to parametrise the decay in LHCb frame; ◮ Select B+ → J/ψK + events in 2D bins, and compute rJ/ψ in each bin:

4.0 4.5 5.0 5.5

)))

−

l ( p ),

+

p(l (max(

10

log

0.00 0.05 0.10 0.15 0.20 0.25 0.30

) [rad]

−

l ,

+

l ( α

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

LHCb simulation rare J/ψ ) bin number

−

l ,

+

l ( α × ))

−

l ( p ),

+

p(l max(

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

〉

ψ J/

r 〈 /

ψ J/

r

0.9 1.0 1.1

LHCb

LHCb-Paper-2019-009 Flatness of R2D

J/ψ plots gives confidence that efficiencies are understood over all phase-space.

7 Thibaud Humair

slide-10
SLIDE 10

Fit to B+ → K +µ+µ− and B+ → K +e+e−

A single fit to the m(K +ℓ+ℓ−) distributions is performed to determine RK from the entire 2011-2016 dataset, taking into account all correlations (LHCb-Paper-2019-009):

]

2

c [MeV/ )

−

µ

+

µ

+

m(K

5200 5300 5400 5500 5600

)

2

c Candidates / (7 MeV/

50 100 150 200 250 300 Data Total fit

−

µ

+

µ

+

K →

+

B Combinatorial

LHCb

Nsig ∼ 1940

]

2

c [MeV/ )

−

e

+

e

+

m(K

5000 5500 6000

)

2

c Candidates / (24 MeV/

20 40 60 80 100 Data Total fit

−

e

+

e

+

K →

+

B

  • Part. Reco.

+

)K

−

e

+

(e ψ J/ →

+

B Combinatorial

LHCb

Nsig ∼ 760 Partially reconstructed background shape in B+ → K +e+e− taken from simulated B0 → K ∗0(K +π−)e+e−, associated systematic is 1%.

8 Thibaud Humair

slide-11
SLIDE 11

RK result with 2011 to 2016 data LHCb-Paper-2019-009

Using 2011 and 2012 LHCb data, RK was: RK = 0.745+0.090

−0.074(stat.) ± 0.036(syst.),

∼ 2.6 σ from SM (PRL113(2014)151601). Adding 2015 and 2016 data, RK becomes:

5 10 15 20

]

4

c /

2

[GeV

2

q

0.0 0.5 1.0 1.5 2.0

K

R

BaBar Belle LHCb Run 1

LHCb

9 Thibaud Humair

slide-12
SLIDE 12

RK result with 2011 to 2016 data LHCb-Paper-2019-009

Using 2011 and 2012 LHCb data, RK was: RK = 0.745+0.090

−0.074(stat.) ± 0.036(syst.),

∼ 2.6 σ from SM (PRL113(2014)151601). Adding 2015 and 2016 data, RK becomes: RK = 0.846 +0.060

−0.054(stat.) +0.016 −0.014(syst.)

∼ 2.5 σ from SM.

]

4

c /

2

[GeV

2

q

5 10 15 20

K

R

0.0 0.5 1.0 1.5 2.0

BaBar Belle LHCb Run 1 LHCb Run 1 + 2015 + 2016

LHCb

9 Thibaud Humair

slide-13
SLIDE 13

RK result with 2011 to 2016 data LHCb-Paper-2019-009

Using 2011 and 2012 LHCb data, RK was: RK = 0.745+0.090

−0.074(stat.) ± 0.036(syst.),

∼ 2.6 σ from SM (PRL113(2014)151601). Adding 2015 and 2016 data, RK becomes: RK = 0.846 +0.060

−0.054(stat.) +0.016 −0.014(syst.)

∼ 2.5 σ from SM.

]

4

c /

2

[GeV

2

q

5 10 15 20

K

R

0.0 0.5 1.0 1.5 2.0

BaBar Belle LHCb Run 1 LHCb Run 1 + 2015 + 2016

LHCb

Dominant systematic uncertainties: Fit shape, trigger calibration, B+ kinematics.

9 Thibaud Humair

slide-14
SLIDE 14

Branching fractions and other results

LHCb-Paper-2019-009 If instead the Run 1 and Run 2 were fitted separately: Rnew

K Run 1 = 0.717+0.083 −0.071 +0.017 −0.016,

RK Run 2 = 0.928+0.089

−0.076 +0.020 −0.017,

Rold

K Run 1 = 0.745+0.090 −0.074 ± 0.036

(PRL113(2014)151601) ,

Compatibility taking correlations into account:

◮ Previous Run 1 result vs. this Run 1 result (new reconstruction selection): < 1 σ; ◮ Run 1 result vs. Run 2 result: 1.9 σ.

B+ → K+µ+µ− branching fraction:

◮ Compatible with previous result (JHEP06(2014)133) at < 1 σ; ◮ Run 1 and Run 2 results compatible at < 1 σ.

B+ → K+e+e− branching fraction: dB (B+ → K +e+e−) dq2 (1.1 < q2 < 6.0 GeV2) = (28.6+2.0

−1.7 ± 1.4) × 10−9 GeV−2

10 Thibaud Humair

slide-15
SLIDE 15

Conclusion and outlook

◮ Updated RK analysis has a significantly improved precision... ◮ ... but SM compatibility unchanged: LFU breaking not confirmed, nor ruled out.

Much remains to be done with the LHCb data in hand:

◮ Update RK and RK ∗ with full Run 2 data

⇒ 2× as many B’s as in present RK update.

◮ Many other observables:

◮ RK and RK∗ in the high q2 bin; ◮ LFU in other b → sℓ+ℓ− decays, e.g. Bs → φℓ+ℓ−, Λb → p+K −ℓ+ℓ−; ◮ Full q2 dependent B0 → K ∗0µ+µ− analysis; ◮ LFU in charged currents (R(D), R(D∗)).

With full LHCb Run 2 data available (up to 2018), the beginning of Belle 2 data taking, and LHCb upgraded detector starting data taking in 2021, we can expect the flavour anomalies to soon be understood.

11 Thibaud Humair

slide-16
SLIDE 16

Thank you

12 Thibaud Humair

slide-17
SLIDE 17

Back-up

13 Thibaud Humair

slide-18
SLIDE 18

2D q2, m plot

5.0 5.2 5.4 5.6 5.8 6.0

]

2

c [GeV/ )

−

µ

+

µ

+

m(K

5 10 15 20 25

]

4

c /

2

[GeV

2

q

1 10

2

10

3

10

4

10

5

10

LHCb

4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0

]

2

c [GeV/ )

−

e

+

e

+

m(K

5 10 15 20 25

]

4

c /

2

[GeV

2

q

1 10

2

10

3

10

4

10

5

10

LHCb

14 Thibaud Humair

slide-19
SLIDE 19

R(ψ(2S))

]

2

c [MeV/ )

−

µ

+

µ

+

(K

(2S) ψ

m

5200 5300 5400 5500 5600

)

2

c Candidates / (4 MeV/

2 4 6 8 10 12 14 16 18 20 22 24

3

10 × Data Total fit

+

)K

−

µ

+

µ (2S)( ψ →

+

B Combinatorial

LHCb ]

2

c [MeV/ )

−

e

+

e

+

(K

(2S) ψ

m

5200 5400 5600

)

2

c Candidates / (12 MeV/

2 4 6 8 10

3

10 × Data Total fit

+

)K

−

e

+

(2S)(e ψ →

+

B

*0

)K

−

e

+

(2S)(e ψ →

+

B H

+

)K

−

e

+

(e ψ J/ →

,0 +

B

+

)K

−

e

+

(e ψ J/ →

+

B

−

e

+

e

+

K →

+

B Combinatorial

LHCb LHCb

Rψ(2S)

K

= 0.986 ± 0.013

15 Thibaud Humair

slide-20
SLIDE 20

Profiled likelihood

0.6 0.8 1 1.2 1.4

K

R

5 10 15 20 25 30

)

min

L / L ln( − Profile of

LHCb

16 Thibaud Humair

slide-21
SLIDE 21

Cascade vetoes

1000 2000 3000 4000 5000

]

2

c [MeV/ )

−

e

+

m(K

5 −

10

4 −

10

3 −

10

2 −

10

1 −

10

Normalised distribution

−

e

+

e

+

K →

+

B ν

+

) e ν

−

e

+

K → ( D →

+

B

]

−

e → [ −

π ) ν

−

e

+

K → ( D →

+

B ν

+

) e

]

−

e → [ −

π

+

K → ( D →

+

B

LHCb simulation

1700 1800 1900 2000

]

2

c [MeV/ )

]

−

π → [ −

e

+

(K

track

m

0.00 0.01 0.02 0.03 0.04 0.05

Normalised distribution LHCb simulation

17 Thibaud Humair

slide-22
SLIDE 22

Distributions (1)

) [rad]

−

l ,

+

l ( α

0.1 0.2 0.3 0.4 0.5

Candidates / (a. u.)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

LHCb simulation

) [rad]

−

l ,

+

K ( α

0.1 0.2 0.3 0.4 0.5

Candidates / (a. u.)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

LHCb simulation

) [rad]

+

l ,

+

K ( α

0.1 0.2 0.3 0.4 0.5

Candidates / (a. u.)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

LHCb simulation

)

+

K ( η

2 3 4 5

Candidates / (a. u.)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

LHCb simulation

))

−

l ( η ),

+

l ( η max(

2 3 4 5

Candidates / (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

+

(e ψ J/ →

+

B

+

)K

−

µ

+

µ ( ψ J/ →

+

B

18 Thibaud Humair

slide-23
SLIDE 23

Distributions (2)

] c )) [MeV/

−

l (

T

p ),

+

l (

T

p max(

2000 4000 6000 8000 10000

Candidates / (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

p min(

1000 2000 3000 4000 5000

Candidates / (a. u.)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

LHCb simulation

))

+

B (

Vtx 2

χ (

10

log

2 − 2

Candidates / (a. u.)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

LHCb simulation

))

+

B (

IP 2

χ (

10

log

4 − 2 − 2

Candidates / (a. u.)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

LHCb simulation

))

−

l ( η ),

+

l ( η min(

2 3 4 5

Candidates / (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

+

(e ψ J/ →

+

B

+

)K

−

µ

+

µ ( ψ J/ →

+

B

19 Thibaud Humair

slide-24
SLIDE 24

Part reco components

]

2

c [MeV/ )

  • e

+

e

+

m(K

4600 4800 5000 5200 5400 5600

Candidates / (a. u.)

5 10

  • e

+

e

*

K → B

  • e

+

e

1 *+

K →

+

B

  • e

+

e

2 *+

K →

+

B )

  • e

+

e → ( ψ J/

+

K →

+

B Y)

+

K → (

s

H ψ J/ →

  • r B

+

X) K ψ J/ → (

c

H → B

LHCb simulation

20 Thibaud Humair

slide-25
SLIDE 25

Fits to resonant mode

]

2

c [MeV/ )

−

µ

+

µ

+

(K

ψ J/

m

5200 5300 5400 5500 5600

)

2

c Candidates / (4 MeV/

10

2

10

3

10

4

10

5

10 Data Total fit

+

)K

−

µ

+

µ ( ψ J/ →

+

B

+

π )

−

µ

+

µ ( ψ J/ →

+

B Combinatorial

LHCb ]

2

c [MeV/ )

−

e

+

e

+

(K

ψ J/

m

5200 5400 5600

)

2

c Candidates / (12 MeV/

2

10

3

10

4

10

5

10 Data Total fit

+

)K

−

e

+

(e ψ J/ →

+

B

  • Part. Reco.

+

π )

−

e

+

(e ψ J/ →

+

B Combinatorial

LHCb

21 Thibaud Humair

slide-26
SLIDE 26

Supplementary rJ/ψ 1D

Candidates / (a. u.)

0.0 0.5 1.0

−

e

+

e

+

K →

+

B

−

µ

+

µ

+

K →

+

B

+

)K

−

e

+

(e ψ J/ →

+

B

+

)K

−

µ

+

µ ( ψ J/ →

+

B

LHCb simulation

dilepton opening angle [rad]

0.1 0.2 0.3 0.4 0.5

〉

ψ J/

r 〈 /

ψ J/

r

0.90 0.95 1.00 1.05 1.10

LHCb

Candidates / (a. u.)

0.0 0.5 1.0 1.5

LHCb simulation

] c ) [MeV/

+

B (

T

p

5000 10000 15000

〉

ψ J/

r 〈 /

ψ J/

r

0.90 0.95 1.00 1.05 1.10

LHCb

22 Thibaud Humair

slide-27
SLIDE 27

Efficiency calibration

(B) [MeV]

T

p

5000 10000 15000

ε K, ee) / ψ J/ → N(B

25 30 35 40 45 50 55 60 65

6

10 ×

Before calibration After calibration

LHCb

(B) [MeV]

T

p

5000 10000 15000

ε ) / µ µ K, ψ J/ → N(B

25 30 35 40 45 50 55 60 65

6

10 ×

Before calibration After calibration

LHCb

23 Thibaud Humair