[PPT] - Constraining the CKM angle Sneha Malde University of Oxford 25 th PowerPoint Presentation

SLIDE 1

Constraining the CKM angle γ

Sneha Malde University of Oxford 25th January 2017

1 Sneha Malde

SLIDE 2

Many reasons to believe New Physics exists

Sneha Malde 2

The maFer-anH asymmetry that is manifest in our universe is a mystery
There must be a mechanism(s) by which differences between maFer and anH-

maFer are generated.

SLIDE 3

CP ViolaHon and New Physics

Sneha Malde 3

To date only observed in the quark sector, but at levels far below that

required to explain the universe

There must be addiHonal sources of CPV in New Physics models
First ObservaHon of CPV in

1964 in the Kaon system

Nobel prize awarded 1980
Interest in CPV has

conHnued to grow

Observed in B decays in 2001
J. Cronin & V. Fitch

SLIDE 4

CKM Matrix

Sneha Malde 4

u c t ⎛ ⎝ ⎜ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟ ⎟ ← W ± → Vud Vus Vub Vcd Vcs Vcb Vtd Vts Vtb ⎛ ⎝ ⎜ ⎜ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟ ⎟ ⎟ d s b ⎛ ⎝ ⎜ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟ ⎟

b c

W±

By emission or absorpHon of a W± boson , quarks change flavour

SLIDE 5

Unitarity triangle

Sneha Malde 5

Wolfenstein parameterisaHon is commonly used where λ is the sine of the

Cabibbo angle λ≈0.22

The CKM matrix is unitary, and reduces to three rotaHons and one phase.
Phase gives rise to CP violaHon

Vud Vus Vub Vcd Vcs Vcb Vtd Vts Vtb ⎛ ⎝ ⎜ ⎜ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟ ⎟ ⎟ = 1− λ 2 / 2 λ Aλ3(1− ρ −iη) −λ 1− λ 2 / 2 Aλ 2 Aλ3(1− ρ −iη) −Aλ 2 1 ⎛ ⎝ ⎜ ⎜ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟ ⎟ ⎟ +O(λ 4)

(0,0) (1,0) Using the properHes of unitary matrices

0 =1+ Vtb

*Vtd

Vcb

*Vcd

+ Vub

*Vud

Vcb

*Vcd

“Most open” triangle, others are possible (1-λ2/2)(ρ,η) γ γ β β α α

SLIDE 6

Is the triangle a triangle?

Sneha Malde 6

1995 2001 2004 2006 2009 2015

hFp://ckmfiFer.in2p3.fr

Improvements in constraints on triangle apex due to both experiment and theory advances

SLIDE 7

Loop/Tree

Sneha Malde 7

Loop Tree

Loop processes more easily

altered by the presence of New Physics

Constraints on the apex

currently more stringent from loop decay measurements

Largest uncertainty is on γ, a

process accessible at tree level

TheoreHcally clean – uncertainty

from observable to physics parameters ~10-7

Forms a SM benchmark*

*assuming no New Physics in tree decays

hFp://ckmfiFer.in2p3.fr JHEP 01 (2014) 051 PRD 92(3):033002 (2015)

SLIDE 8

γ from indirect determinaHon

Sneha Malde 8

EPJC (2016) 76 197

The unitarity triangle is constructed using mixing and sin(2β) measurements and lamce QCD

EPJC (2016) 76 197 [Blanke, Buras]

(0,0) (1,0) β β γ γ

Length related raHo of Bd and Bs mixing

γ = (62.7± 2.1)! γ = (72.1−5.8

+5.4)!

CombinaHon of all direct measurements (summer 2016) AlternaHve approach from CKM fit excluding all direct measurements of γ

γ = (65.33−2.54

+0.96)!

hFp://ckmfiFer.in2p3.fr

UncertainHes dominated by LQCD, expect to reduce over the next decade Reaching degree level precision from direct measurements is crucial

sin(2β) from BàJ/ψ Ks

SLIDE 9

Why is γ a key goal?

Sneha Malde 9

New Physics must provide a new source of CPV
γ is the least well measured parameter of the

CKM triangle

Only angle easily accessible at tree-level
TheoreHcally prisHne
Provides a SM benchmark against which other

measurements can be compared

With the advent of LHCb the ideal of degree level

precision starts to become reality

SLIDE 10

BàDK

Sneha Malde 10

γ = −arg VudVub

*

VcdVcb

*

$ % & ' ( )

Interference possible if D0 and D0 decay to same final state
Branching fracHon for favoured B decay ~10-4
Fully hadronic final state
Measurements will require high staHsHcs

SLIDE 11

Interference with CP eigenstates “GLW”

Sneha Malde 11

B- D0K-

[KK]DK-

D0K-

rBe

i(δB-γ)

CP eigenstates equally accessible to D0 and D0 Weak phase changes sign for equivalent B+ diagram rB, δB hadronic parameters to be determined alongside γ rB ~0.1 Interested in the rate of observing this decay in B- vs. B+ Interested in the rate of observing this decay vs. one that is not affected by interference, e.g the Cabibbo favoured decay of the D0

Gronau & London, PLB 253 (1991) 483, Gronau & Wyler PLB 265 (1991) 172

SLIDE 12

Interference with CP eigenstates “GLW”

Sneha Malde 12

B- D0K-

[KK]DK-

D0K-

rBe

i(δB-γ)

EquaHons simplified – assume no D mixing For CP+ eigenstates e.g KK, ππ:

N(B−)− N(B+) N(B−)+ N(B+) = ACP+ = 1 RCP+ 2r

B sin(δB)sin(γ)

N(B →[KK]DK)×Γ(D → Kπ) N(B →[Kπ]DK)×Γ(D → KK) = RCP+ =1+r

B 2 + 2r B cos(δB)cos(γ)

SLIDE 13

Interference with flavour specific “ADS”

Sneha Malde 13

B- D0K-

[π-K+]DK-

D0K-

rBe

i(δB-γ)

Larger interference effects as both amplitudes of similar sizes. AddiHonal two parameters rD, δD . External inputs from charm mixing. rD ~ 0.06

rDe

i(δD)

N(B−)− N(B+) N(B−)+ N(B+) = AADS = 1 RADS 2r

Br D sin(δB +δD)sin(γ)

N(B± →[π ±K ∓]DK ±) N(B± →[K ±π ∓]DK ±) = RADS = r

B 2 +r D 2 + 2r Br D cos(δB +δD)cos(γ) Atwood, Dunietz & Soni PRL 78 (1997) 3257, PRD 63 (2001) 036005 (ADS)

SLIDE 14

LHCb detector

Sneha Malde 14

RICH

All except one analysis presented today come from full 2011 and 2012 datasets

SLIDE 15

Detector performance (1)

Sneha Malde 15

Int. J Mod. Phys A 30 (2015) 1530022

VELO: Impact parameter resoluHon Tracking: Momentum resoluHon

SLIDE 16

Detector performance (2)

Sneha Malde 16

Int. J Mod. Phys A 30 (2015) 1530022

RICH detectors Low π misidenHficaHon rate

High kaon idenHficaHon Hardware trigger

hadronic trigger with high efficiency

Sowware trigger

exploits decay topology

SLIDE 17

Datasets

Sneha Malde 17

1 x-1 @ 7 TeV (2011) 2 x-1 @ 8 TeV (2012) Pile up much lower than the GPDs ~ 2 collisions per bunch crossing 0.3 x-1 @ 13 TeV (2015) 1.7 x-1 @ 13 TeV (2016) Pile up reduced to ~1 per bunch crossing Increased cross secHon Analyses today – all but one on Run 1 data

Precision measurements take effort and Hme
2015 data only gives a modest increase.
Most “ Run 1” final results in this talk were produced in 2016

SLIDE 18

SelecHon

Sneha Malde 18

Separate the topology of interest from random combinaHons Use of mulH-variate analysis techniques. Useful variables include: Impact parameters Flight distances from primary. (B travels a ~cm) Flight distances from B – removes e.g BàKππ backgrounds Vertex quality ParHcle ID Specific vetos against parHcular backgrounds

pp collision B D

IP

K π π

All analyses shown here employ similar strategies

SLIDE 19

BàD[Kπ]h – CF control mode

Sneha Malde 19

BàDK

Very large samples, ~ 30K BàDK

BàDπ

BàD*h Difference between the two modes

nly the ID of the bachelor hadron

PID performanceà low crossfeed. B->D*h where a π0 or photon isn’t reconstructed sits to the lew Extremely low level of combinatoric – clean environment Control mode constrains the shapes of signal and backgrounds Control mode also used to measure the B± producHon asymmetry. DetecHon asymmetries calibrated from other data. Results also extracted for BàDπ mode, interference level expected to be ~ magnitude smaller

arXiv:1603.08993

LHCb LHCb

SLIDE 20

BàD(KK)h

Sneha Malde 20

5100 5200 5300 5400 5500

)

2

c Events / ( 10 MeV/ 100 200 300 400

K

D

]

K

+

K [

B

LHCb

5100 5200 5300 5400 5500

+

K

D

]

K

+

K [

+

B LHCb 5100 5200 5300 5400 5500 2000 4000 6000

D

]

K

+

K [

B

LHCb ]

2

c ) [MeV/

±

Dh ( m 5100 5200 5300 5400 5500

+

D

]

K

+

K [

+

B LHCb

AK

KK = 0.087± 0.020 ± 0.008

StaHsHcal uncertainty dominant DescripHon of background is the leading systemaHc uncertainty

arXiv:1603.08993

~ 3800 BàDK

SLIDE 21

BàD(ππ)h

Sneha Malde 21

5100 5200 5300 5400 5500

)

2

c Events / ( 10 MeV/ 50 100 150

K

D

]

+
[
B

LHCb

5100 5200 5300 5400 5500

+

K

D

]

+
[
+

B LHCb 5100 5200 5300 5400 5500 500 1000 1500 2000

D

]

+
[
B

LHCb ]

2

c ) [MeV/

±

Dh ( m 5100 5200 5300 5400 5500

+

D

]

+
[
+

B LHCb

AK

ππ = 0.128± 0.037± 0.012

Asymmetry same direcHon as KK mode Combined observaHon of CP violaHon

5σ

arXiv:1603.08993

~ 1160 BàDK

SLIDE 22

BàD[πK]h

Sneha Malde 22

5100 5200 5300 5400 5500

)

2

c Events / ( 10 MeV/ 50 100

K

D

]

+

K

[
B

LHCb

5100 5200 5300 5400 5500

+

K

D

]

K

+

[
+

B LHCb 5100 5200 5300 5400 5500 200 400

D

]

+

K

[
B

LHCb ]

2

c ) [MeV/

±

Dh ( m 5100 5200 5300 5400 5500

+

D

]

K

+

[
+

B LHCb

AK

πK = −0.403± 0.056 ± 0.011

Aπ

πK = 0.100 ± 0.031± 0.009

ObservaHon of CP violaHon in BàDK CPV starts to become visible in BàDπ Combined with DàKK Dàππ significance

8σ 3.9σ

arXiv:1603.08993

~ 550 BàDK

Only observed at LHCb, BF ~10-7 -- a rare decay

SLIDE 23

Comparison of results

Sneha Malde 23

ACP+ Averages

HFAG

Moriond 2016 Moriond 2016

DCP K

BaBar

PRD 82 (2010) 072004

0.25 ± 0.06 ± 0.02 Belle

LP 2011 preliminary

0.29 ± 0.06 ± 0.02 CDF

PRD 81 (2010) 031105(R)

0.39 ± 0.17 ± 0.04 LHCb KK

arXiv:1603.08993

0.09 ± 0.02 ± 0.01 LHCb

arXiv:1603.08993

0.13 ± 0.04 ± 0.01 Average

HFAG

0.13 ± 0.02 BaBar

0.11 ± 0.09 ± 0.01
H F AG

H F A G

Moriond 2016 PRELIMINARY

Moriond 2016

1.2
1
0.8
0.6
0.4
0.2

0.2 0.4 0.6 0.8 1

Moriond 2016
0.4
0.2

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6

HFAG

RCP+ Averages

HFAG

Moriond 2016 Moriond 2016

DCP K

BaBar

PRD 82 (2010) 072004

1.18 ± 0.09 ± 0.05 Belle

LP 2011 preliminary

1.03 ± 0.07 ± 0.03 CDF

PRD 81 (2010) 031105(R)

1.30 ± 0.24 ± 0.12 LHCb KK

arXiv:1603.08993

0.97 ± 0.02 ± 0.02 LHCb

arXiv:1603.08993

1.00 ± 0.04 ± 0.03 Average

HFAG

1.00 ± 0.02 BaBar 1.31 0.13 0.03

H F AG

H F A G

Moriond 2016 PRELIMINARY

AADS Averages

HFAG

Moriond 2016 Moriond 2016

D_K K K

BaBar

PRD 82 (2010) 072006

0.86 ± 0.47 +
.

. 1 1 2 6

Belle

PRL 106 (2011) 231803

0.39 +
.

. 2 2 6 8 +

.

. 4 3

CDF

PRD 84 (2011) 091504

0.82 ± 0.44 ± 0.09

LHCb

arXiv:1603.08993

0.40 ± 0.06 ± 0.01

Average

HFAG

0.41 ± 0.06

Belle

0.41 ± 0.30 ± 0.05

H F AG H F A G

Moriond 2016 PRELIMINARY

HFAG

Moriond 2016

D_K
2
1

1

HFAG

RADS Averages

HFAG

Moriond 2016 Moriond 2016

D_K K

BaBar

PRD 82 (2010) 072006

0.011 ± 0.006 ± 0.002

Belle

PRL 106 (2011) 231803

0.016 ± 0.004 ± 0.001

CDF

PRD 84 (2011) 091504

0.022 ± 0.009 ± 0.003

LHCb

arXiv:1603.08993

0.019 ± 0.001 ± 0.001

Average

HFAG

0.018 ± 0.001

BaBar

0.009 0.008 +0.001

H F AG H F A G

Moriond 2016 PRELIMINARY

HFAG

Moriond 2016

D_K
0.08
0.06
0.04
0.02

0.02 0.04 0.06 0.08 0.1 HFAG

LHCb results dominate world averages

SLIDE 24

MulH-body flavour specific D decays “ADS”

Sneha Malde 24

Treat mulHbody decays inclusively à avoids consideraHon of intermediate

states. ParHcularly useful for 4-

body decay modes

B- D0K-

[π-K+π+π-]DK-

D0K-

rBe

i(δB-γ)

? ?

κe

iδD

f =

Af (x)Af (x)dx

∫

Af

2(x)dx

∫

Af

2(x)dx

∫

AADS = 1 RADS 2r

Br D K3πκ sin(δB +δD K3π )sin(γ)

RADS = r

B 2 +r D 2 + 2r Br D K3πκ cos(δB +δD K3π )cos(γ)

Af Af

Af Af = r

D K3π ~ 0.05

PRD 68 033003 (2003)

SLIDE 25

Measurements of coherence factor

Sneha Malde 25

D0 D0

K-π-π+π-

scillaHon

Interference between mixing and decay determined from Hme-dependent decay rates.

Af = r

D K3πκeiδD

K 3π

Af =1

arXiv:1602.077224

κ,δD

K3π

κ

SLIDE 26

Measurements with CLEO data

Sneha Malde 26

Study ψ(3770) à D0D0 decays
Key: C= -1 for ψ(3770) at threshold
Strong decay, C is conserved
Hence the decays of D0 and D0 are quantum

correlated

This provides the interference to access the phase

informaHon

Study rates where one D meson decays to K3π and

the other to either a CP eigenstate.

Rates are dependent on the κ and δK3π
Synergy between two measurement – sensiHve to

different regions

Strong phase measurements in other decay modes

follow same principles

LHCb + CLEO data

PLB 757 (2016) 520

κ=0.43+0.17

0.13

κ

SLIDE 27

Results DàK3π

Sneha Malde 27

5100 5200 5300 5400 5500

)

2

c Events / ( 10 MeV/ 10 20 30 40

K

D

]

+
+

K

[
B

LHCb

5100 5200 5300 5400 5500

+

K

D

]

+
K

+

[
+

B LHCb 5100 5200 5300 5400 5500 50 100 150

D

]

+
+

K

[
B

LHCb ]

2

c ) [MeV/

±

Dh ( m 5100 5200 5300 5400 5500

+

D

]

+
K

+

[
+

B LHCb

AK

πKππ = −0.313± 0.102 ± 0.038

Complementary informaHon to two body modes.

arXiv:1603.08993

BàDK yield ~160

SLIDE 28

MulH-body self conjugate D decays “quasi-GLW”

Sneha Malde 28

If the CP even fracHon is known then self-conjugate modes can also be used in a similar way to CP eigenstates. Measured at CLEO from quantum correlated data F+

4π= 0.737±0.028

B- D0K-

[π-π+π+π-]DK-

D0K-

rBe

i(δB-γ)

? ?

Aq−CP = 1 Rq−CP 2r

B(2F + −1)sin(δB)sin(γ)

Rq−CP =1+r

B 2 + 2r B(2F + −1)cos(δB)cos(γ)

PLB 740 (2015) 1

SLIDE 29

Results Dà4π

Sneha Malde 29

5100 5200 5300 5400 5500

)

2

c Events / ( 10 MeV/ 50 100 150

K

D

]

+
+
[
B

LHCb

5100 5200 5300 5400 5500

+

K

D

]

+
+
[
+

B LHCb 5100 5200 5300 5400 5500 500 1000 1500 2000

D

]

+
+
[
B

LHCb ]

2

c ) [MeV/

±

Dh ( m 5100 5200 5300 5400 5500

+

D

]

+
+
[
+

B LHCb

AK

ππππ = 0.100 ± 0.034± 0.018

First use of this mode -possible due to measurements from CLEO

arXiv:1603.08993

BàDK yield ~1500

SLIDE 30

]

4

c /

2

) [GeV

−

π

S

K (

2

m

1 2 3

]

4

c /

2

) [GeV

+

π

S

K (

2

m

1 2 3

LHCb

D

Self-conjugate D decays using Dalitz plot “GGSZ”

Sneha Malde 30

B- D0K-

[Ksπ+π-]DK-

D0K-

rBe

i(δB-γ)

rDe

i(δD)

? ?

Value of F+ for certain self conjugate decays would be ~0.5 Hence inclusive treatment loses most of the sensiHvity to γ à Analyse the Dalitz plot Best standalone measurement of γ Dalitz Plot encodes all the kinemaHc informaHon of the decay Each point on the Dalitz plot represents a different value of rD and δD

Giri, Gronau, Soffer & Zupan, (GGSZ) PRD 68 (2003) 054018

SLIDE 31

Two methods for accessing the D decay informaHon

Sneha Malde 31

D dalitz plot from B decay will be a superposiHon of D0 and D0
It will differ between B+ and B-
Differences are related to rB δB and γ

Two ways to deal with the varying rD, δD Model dependent rD and δD determined from flavour tagged decays via amplitude model No interference, no direct access to phase informaHon SystemaHc uncertainHes due to model hard to quanHfy Use CLEO data to measure average values of rD and δD in bins Small loss in staHsHcal precision Direct phase informaHon, uncertainHes

n which are easily propagated

Model independent

PRD 82 (2010) 112006

SLIDE 32

Model-independent GGSZ analysis

Sneha Malde 32

Reduces to a counHng experiment in bins of

Dalitz Plot

Bin definiHon designed to minimise staHsHcal

loss ~ 90% of sensiHvity remains

Bin yields + strong phase informaHon à

measurement of x and y

PRD 82 (2010) 112006

SLIDE 33

BàD[Kshh]K (GGSZ)

Sneha Malde 33

]

4

c /

2

[GeV

2 +

m

1 2 3

]

4

c /

2

[GeV

2 −

m

1 2 3

LHCb

]

4

c /

2

[GeV

2 −

m

1 2 3

]

4

c /

2

[GeV

2 +

m

1 2 3

LHCb

x

0.3
0.2
0.1

0.1 0.2 0.3 y

0.3
0.2
0.1

0.1 0.2 0.3

+

B

−

B

LHCb

2γ

Ksππ and KsKK decay modes (not shown)

used. Signal yield ~2400 γ = (62−14

+15) !

JHEP 10 (2014) 097

SeparaHon (x+,y+) , (x-,y-) shows CPV

B+ B-

SLIDE 34

Interplay between different modes

Sneha Malde 34

ADS/GLW/q-GLW observables have non

trivial trigonometric relaHons.

Nuisance parameters rB and δB common to

all modes

Single soluHon selected by GGSZ modes
No single mode dominatesà necessary to

follow all paths

LHCb-CONF-2016-001

SLIDE 35

Other B modes

Sneha Malde 35

B0 B0 b b d d d d u c s c u s D0 K∗0 D0 K∗0 W + W + V ∗

ub

t

Vcs

t

V ∗

cb

t

Vus

t

Favoured and suppressed decay both colour suppressed
rB ~ 0.3 à Larger interference
K* àK+π-, charge of kaon tags flavour of B at decay – no need for Hme

dependent analysis

Yields at LHCb becoming viable for analysis
ADS/GLW analysis already performed on full Run 1 dataset
Different rB and δB

SLIDE 36

SelecHon of B0àDK*

Sneha Malde 36

]

2

c ) [MeV/

±

π

±

DK ( m 5200 5400 5600 5800 )

2

c Candidates / (15.0 MeV/ 20 40 60 80

Total

*0

DK → B

*0

K D →

s

B

*0

K

*0

D →

s

B ρ D → B

±

DK →

±

B Combinatorial

LHCb

]

2

c ) [MeV/

±

π

±

DK ( m 5200 5400 5600 5800 )

2

c Candidates / (15.0 MeV/ 10 20 30

Total

*0

DK → B

*0

K D →

s

B

*0

K

*0

D →

s

B ρ D → B

±

DK →

±

B Combinatorial

LHCb

Yields ~ 90 in Ksππ and 10 in ~KsKK
Twice yield of B factories
Irreducible Bs backgrounds
Width of K*(892) means non-

resonant Kπ decays can contribute to signal peak

Coherence factor dependent on

selecHon

|M(K*)-892| <50 MeV/c2;
|cos(K helicity angle)|>0.4

arXiv: 1604.01525

Ksππ ππ KsKK

SLIDE 37

GGSZ analysis

Sneha Malde 37

Modified binning used for Ksππ – beFer for low yield channels
KsKK split into 2 bins

arXiv: 1604.01525

SLIDE 38

Determining observables

Sneha Malde 38

]

2

c ) [MeV/

−

π

+

DK ( m 5200 5400 5600 5800 )

2

c Candidates / (15.0 MeV/ 2 4 6

Total

*0

DK → B

*0

K D →

s

B

*0

K

*0

D →

s

B ρ D → B

±

DK →

±

B Combinatorial

LHCb

]

2

c ) [MeV/

+

π

−

DK ( m 5200 5400 5600 5800 )

2

c Candidates / (15.0 MeV/ 2 4 6

LHCb

Model dependent fit also performed
rD and δD given by BaBar 2010 amplitude model

arXiv: 1604.01525 arXiv: 1605.01082

Simultaneous fit to all bins to determine x, y
Signal/background shapes fixed from first fit.
Very few signal events per bin
Example fit projecHon of one bin:

B0 B0

SLIDE 39

Results

Sneha Malde 39 ±

x

1 − 1 ±

y

1 − 1

B B LHCb

Model - independent x+ = 0.05± 0.24± 0.04± 0.01 y+ = −0.65−0.23

+0.24 ± 0.08± 0.01

x− = −0.15± 0.14± 0.03± 0.01 y− = 0.25± 0.15± 0.06 ± 0.01

±

x

1

1

±

y

1

1 B B

LHCb

Model - dependent x+ = 0.05± 0.35± 0.02 y+ = −0.81± 0.28± 0.06 x− = −0.31± 0.20 ± 0.04 y− = 0.31± 0.21± 0.05

Good agreement

between methods

UncertainHes from

external strong phase informaHon are ~0.02 for x and ~0.05 for y.

Both methods give

σ(γ)=20°

arXiv: 1604.01525 arXiv: 1605.01082

SLIDE 40

CombinaHon results

Sneha Malde 40

FrequenHst combinaHon using ‘plugin’
method. 71 observables and 32

parameters.

More analyses than shown today
Only “BàDK – like” results included
Only includes the 1x-1 BsàDsK result
Improved precision compared to last

combinaHon (2014) by ~20%

Good agreement with B factory

results

Bayesian interpretaHon is consistent

γ = (72.2−7.3

+6.8)!

arXiv: 1611.03076

BaBar : Belle:

γ = (69−16

+17)!

γ = (73−14

+15)!

PRD 87 (2013) 052015 arXiv:1301.2033

] ° [ γ 1-CL

0.2 0.4 0.6 0.8 1 50 60 70 80 90 68.3% 95.5%

LHCb

SLIDE 41

ContribuHon from different modes

Sneha Malde 41

Common parameter is γ Necessary to pursue different B decays to provide crosschecks Current measurements are dominated by staHsHcal uncertainHes

arXiv: 1611.03076

] ° [ γ 1-CL

0.2 0.4 0.6 0.8 1 50 100 150 68.3% 95.5%

LHCb

SLIDE 42

Run 1 à Run 2

2015 – collected 300 pb-1 @ 13 TeV
2016 – collected 1.7 pb-1 @ 13 TeV
ProducHon cross secHon increases, improved parHcle idenHficaHon, and

slight improvements to trigger mean that yield per pb-1 are 2-3 Hmes larger in Run 2 (depending on decay mode)

StarHng to analyse new modes with Run 1 + Run 2 data – especially ones

that weren’t viable with Run 1 only.

Run 2 target is officially 4°
B+ à DK*+, where K*+ àKSπ

– Should have similar rB to the usual B+àDK+ channel – However expect lower yields due to the KS reconstrucHon efficiency

Sneha Malde 42

SLIDE 43

B+ àDK*+

Sneha Malde 43

Signal well separated from any other physics background. High purity Run 1 (3 x-1) + Run 2 (1 x-1). Yields in each data set are similar Very exciHng for the sensiHviHes we’d be able to achieve in other decay modes

LHCb-CONF-2016-014

]

2

c m(DK*) [MeV/ 4900 5000 5100 5200 5300 5400 5500 5600 )

2

c Candidates / (7.0 MeV/ 20 40 60 80 100 120 140 160 180 200 220 240

*-

K D →

B

*-

) K γ D* (D →

B

*-

) K π D* (D →

B

*-

) K

+

π D* (D → B Combinatorial

LHCb preliminary

SLIDE 44

ADS / GLW analysis

Sneha Malde 44

LHCb-CONF-2016-014

Not enough data to observe the supressed mode, or CPV. Nonetheless remains promising for future due to high purity.

SLIDE 45

SensiHvity to γ

Sneha Malde 45

First CPV measurement to include Run 2 data Add more D decays In the future will provide a valuable cross check against other modes due to the lack of physics background.

SLIDE 46

γ and LHCb upgrade

Sneha Malde 46

LHCb upgrade projecQon (50 S-1) for γ is 0.9° -- no showstoppers forseen If nature is kind, this precision will allow for observaHon of New Physics RUN 2

LS2 Upgrade installaQon

Full upgrade in LS2
Allows for running at higher luminosity in 2021 onwards
L0 hardware trigger à sowware trigger
Increase trigger efficiency for hadronic modes
Dominant experimental systemaHc uncertainHes can be controlled
External inputs will benefit from BES-III data

EPJC (2013) 73:2373

SLIDE 47

End

Sneha Malde 47

SLIDE 48

Bs à Ds K

Sneha Malde 48

Measure CP violaHon in the interference of mixing and decay Both decay amplitudes ~ λ3 à Large interference Tree level process like other analyses shown Time-dependence increases the complexity of the analysis Flavour-tagging also required to know the flavour of the iniHal Bs state

LHCb-CONF-2016-015

SLIDE 49

CP observables

Sneha Malde 49

LHCb-CONF-2016-015

SLIDE 50

Signal/background discriminaHon

Sneha Malde 50

Subsequent parts of the fit only parameterise the signal distribuHons with the use

f the signal weights

LHCb-CONF-2016-014

SLIDE 51

Flavour tagging and Hme dependence

Sneha Malde 51

Efficiency of tagging an event ~ 65.7% EffecHve tagging power ~ 5%

Time acceptance determined from BS à Dsπ Other physics inputs such as BS mixing and lifeHme, and lifeHme difference fixed from

ther measurements

LHCb-CONF-2016-014

SLIDE 52

Fit results and interpretaHon on γ

Sneha Malde 52

LHCb-CONF-2016-014

SLIDE 53

B0 àDKπ Dalitz plot analysis

Sneha Malde 53

B0àDK*, DàCP+, K*àKπ restricts the data to the K* resonance
There is sensiHvity to γ from the full B0àDKπ decay in any Kπ resonance
Amplitude fit of B0àDKπ decay exploits interference between different

resonant contribuHons

Complex amplitudes of the DK* determined relaHve to flavour-specific D2

*K

γ measured from amplitudes and not rates à more informaHon than

standard GLW analysis

New method of measuring γ

PRD 79 (2009) 051301

Favoured (D0àK+π-) mode: CP sensiHve (D0àKK, ππ) modes:

SLIDE 54

B0 àDKπ Dalitz plot analysis

Sneha Malde 54

Favoured (D0àK+π-) mode:

A m2(Dπ),m2(Kπ)

( ) =

cjFj m2(Dπ),m2(Kπ)

( )

j=1 N

∑

CP sensiHve (D0àKK, ππ) modes:

arXiv: 1602.03455

SLIDE 55

Signal yields

Sneha Malde 55

To maximise staHsHcal sensiHvity data

split in bins of MVA output

Data shown with MVA bins combined

weighted according to S/(S+B)

339+/-22 DàKK
168+/-19 Dàππ

The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again. The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again. The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.

arXiv: 1602.03455

SLIDE 56

Dalitz Plot fit

Sneha Malde 56

Fit projecHons of the DàKK and Dàππ samples combined Only results from K*(892) used ProjecHons look very similar

arXiv: 1602.03455

SLIDE 57

Fit Results

Sneha Malde 57

x+ = 0.04± 0.16 ± 0.11 x− = −0.02 ± 0.13± 0.14 y+ = −0.47± 0.28± 0.22 y− = −0.35± 0.26 ± 0.41 κ = 0.958−0.010−0.045

+0.005+0.002

Results for pure K* Also determine the coherence factor

arXiv: 1602.03455

SLIDE 58

B0 combinaHon

Sneha Malde 58

Due to low staHsHcs the B0àDKπ unable to select a single soluHon
In combinaHon with the GGSZ and previous ADS analysis start to constrain

the parameters of interest

LHCb-CONF-2016-001

SLIDE 59

Model-independent GGSZ analysis

Sneha Malde 59

Reduces to a counHng experiment in bins of Dalitz Plot
Bin definiHon designed to minimise staHsHcal loss ~ 90% of sensiHvity remains
Fi determined from B0àD*µν decays (flavour tagged)
ci and si external inputs from CLEO
Arbitrary normalisaHon hb means that insensiHve to producHon asymmetries

PRD 82 (2010) 112006

SLIDE 60

Dalitz Plot efficiency

Sneha Malde 60

]

4

c /

2

) [GeV

−

π

S

K (

2

m

1 2 3

]

4

c /

2

) [GeV

+

π

S

K (

2

m

1 2 3

LHCb

D

VariaHon of efficiency on DP must be taken into

account

B0àD*[D0π] µνX used to determine Fi
Small correcHons required to take care of

selecHon differences between control and signal decay

Determined from simulaHon

The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.

arXiv: 1604.01525

SLIDE 61

Larger phasespace àhigher combinatorics

Sneha Malde 61

Larger phasespace of the Kπ system leads to high combinatorics and larger

amounts of physics bkgs.

To avoid the need to cut hard data is divided into bin of NN output.
Maximises the staHsHcal sensiHvity of the data

arXiv: 1602.03455

SLIDE 62

Combining results -LHCb inputs

Sneha Malde 62

Results discussed today, new or updated since last combinaHon (2014) New results from 2015 Other BàDK ‘like’ results completed in 2014

arXiv: 1605.01082 arXiv: 1602.03455 LHCb-CONF-2016-001

LHCb measurement Type/ Dataset Reference B+àDK+ Dà2h,4h ADS/(q-)GLW (3x-1) arXiv:1603.08993 B0 àDKπ Dalitz (3x-1) arXiv: 1602.03455 B0àDK* DàKsππ GGSZ MD (3x-1) arXiv: 1605.01082 B+àDK+ Dàhhπ0 ADS/q-GLW (3x-1) PRD 91(2015) 112014 B+àDKππ, Dà2h ADS/GLW (3x-1) PRD 92 (2015) 112005 B0àDK* Dà2h ADS (3x-1) PRD 90 (2014) 112002 B+àDK DàKshh GGSZ MI (3x-1) JHEP 10 (2014) 097 B+àDK, DàKsKπ ADS (3x-1) PLB 733 (2014) 36 BsàDsK, Dsàhhh Time dep (1x-1) JHEP 11 (2014) 060

SLIDE 63

Combing results-other inputs

Sneha Malde 63

Parameters Source Reference Charm mixing and CPV in Dàhh HFAG www.slac.stanford.edu/ xorg/hfag/charm/ index.html κ, δD: DàK3π, DàKππ0 LHCb & CLEO data PLB 757 (2016) 520 κ, δD : DàKsKπ CLEO data PRD 85 (2012) 092016 CP fracHon Dà4π, Dàhhπ0 CLEO data PLB 747 (2015) 9 Strong phase informaHon for DàKshh CLEO data PRD 82 (2010) 112006 Constraint on φs LHCb data PRL 114 (2015) 041801

SLIDE 64

Adding “B àDπ” like

Sneha Malde 64

arXiv: 1611.03076

B à Dpi decays usually ignored as rDpi << rDK Don’t like waste! From CKM elements expect rDpi ~ 0.005

] ° [ γ

π D B

r

50 60 70 80 90 0.01 0.02 0.03 0.04 0.05

LHCb

] ° [ γ 1-CL

0.2 0.4 0.6 0.8 1 50 60 70 80 90 68.3% 95.5%

LHCb

With the D modes analysed available the BàDpi and BàDpipipi doesn’t add much in sensiHvity. Aim to extend to other D modes to have a larger impact.

SLIDE 65

ContribuHon from different methods

Sneha Malde 65

Demonstrates the need to pursue all methods BsàDsK

] ° [ γ 1-CL

0.2 0.4 0.6 0.8 1 50 100 150 68.3% 95.5%

LHCb

arXiv: 1611.03076