R Lambert, CERNCERN CPV, 18th June 20101 Flavour-specific asymmetry and LHCb Robert W. Lambert On behalf of the LHCb collaboration

R Lambert, CERN CERN CPV, 18th June 2010 1

Flavour-specific asymmetryand LHCb

Robert W. Lambert

On behalf of the LHCb collaboration

Introduction

D exclusive asymmetry hep-ex 0904.3907

D Inclusive asymmetry hep-ex 1005.2757

What can we do at LHCb?


00 XDB ss

... bb

-31.41.5- 10 (syst)] 9.1(stat) 1.7[ s

fsa

-310 (syst)]46.12.51(stat) 57.9[ bA

http://arxiv.org/abs/0904.3907


Literature

Reading material: LHCb

– CERN-LHCb-2007-054 Public note from old Monte Carlo– CERN-THESIS-2008-045 Paul’s Thesis– CERN-THESIS-2009-001 Rob’s Thesis– CERN-THESIS-2010-076 Ken’s Thesis

Theory– hep-ph 0406300 U. Nierste – hep-ph 0612167 [JHEP] A. Lenz, U. Nierste– hep-ph 0605028 [PRL] Y. Grossman et al.– hep-ph 0604112 [PRL] Z. Ligeti et al.

Other measurements– hep-ex 0505017 [PRD] Belle– hep-ex 0202041 [PRL] Babar– hep-ex 0101006 [PRL] Cleo– Note 9015 CDF inclusive


http://cdsweb.cern.ch/record/1089050?ln=en

http://cdsweb.cern.ch/record/1109349

http://rlambert.web.cern.ch/rlambert/docs/ThesisFinal_Corrected.pdf


http://arxiv.org/abs/hep-ph/0406300


http://iopscience.iop.org/1126-6708/2007/06/072


http://prl.aps.org/abstract/PRL/v97/i15/e151801



http://arxiv.org/abs/hep-ex/0505017

http://prd.aps.org/abstract/PRD/v73/i11/e112002




http://prl.aps.org/abstract/PRL/v86/i22/p5000_1

http://www-cdf.fnal.gov/physics/new/bottom/070816.blessed-acp-bsemil/

Outline

1. Theory .. from an experimentalist

2. Experimental Status

3. Complications at LHCb

4. Why LHCb?

5. Measurements at LHCb

6. Real data highlights

7. Outlook and Prospects

8. Conclusions


Flavour-specific decays Favoured/Allowed Not allowed at tree Through mixing

Flavour specific asymmetry, afs, parameterises CPV in mixing


Flavour-specific asym.

fBBfBB

fBBfBBtAa

qqqq

qqqqqfs

qfs

0000

0000

oror

oror)(

ss DB0

0sB

sD

fBq 0

fBB qq 00

fBq 0

ss DB0

DBd0

The whole shebang

Three different parameters fully constrain b-mixing

afs is very small in the standard model


)(

)(

2)(

)(0

0

0

0

tB

tBiM

tB

tB

dt

di

q

qqq

q

q

q

qqfs Ma

12

12Im ,2,arg2 1212

1212

qqL

qHqq

qqq

LqHq MMMm

M

5

4

104.01.2

101.10.5

SMs

fs

SMdfs

a

a

Discovery Potential

afs is sensitive to new physics (NP): Sensitive to loop contributions Sensitive to new CPV phases

If we allow a single NP phase in the mixing

R Lambert, CERN CERN CPV, 18th June 2010

0sB

0sB?

7

sinRecosIm12

12

12

12SM

SM

SM

SMNP

MMa

sinRecosIm12

12

12

12SM

SM

SM

SMNP

MMa

Discovery Potential




0sB

0sB?

8

SMfsa



Up to 200-times the SM ... but ... (4x10-3)< D measurement

sinRecosIm12

12

12

12SM

SM

SM

SMNP

MMa

Discovery Potential


0sB

0sB?

9

5101.2 3100.4


Experimental Status

B-factories

Babar, Belle, Cleo all use the di-muon sample Know the initial state (4S) Time-integrated number-counting of di-muons Possible muon detector asymmetry is important to measure


bb

c

m-

n

310)5.51.1(

)(

factbdfs

dfsSL

a

aNN

NNA

lXlbbNNc b

m-

n

hep-ph 0605028


Exclusive measurement

D have also made an exclusive measurement Search for the full decay chain in the data Requires full reconstruction Flavour-tagging provides extra information Lower and more easily understood backgroundsx Extra detector asymmetryx Lower statistics


Bs

Bs

Ds+

m-

n

3-1.41.5-

00

00

10 (syst)] 9.1(stat) 1.7[

sfs

sfs

ss

ssSL

a

afBfB

fBfBA

hep-ex 0904.3907


CDF Inclusive

CDF use only the di-muon sample Don’t know the initial state (pp) Time-integrated number-counting Possible muon detector asymmetry is important to measure


bb

c

m-

n

c b

m-

n

3-10 (syst)]8.69.0(stat) 0.8[

582.0418.0

)(

b

dfs

sfs

b

A

aaNN

NNA

lXlbbNN

2fact-b using10) 9(ip).0 (syst)6.1 (stat)1.20.2( s

fsa

Note 9015

http://www-cdf.fnal.gov/physics/new/bottom/070816.blessed-acp-bsemil/

New D inclusive


bb

c

m-

n

c b

m-

n

D used primarily the di-muon sample Don’t know the initial state (pp) Time-integrated number-counting Possible muon detector asymmetry is important to measure

2fact-b using

3-

10)75.046.1(

10 (syst)]46.12.51(stat) 57.9[

506.0494.0

)(

sfs

b

dfs

sfs

b

a

A

aaNN

NNA

lXlbbNN

hep-ex 1005.2757


New D inclusive

Key Systematics (it’s difficult to be inclusive!) Kaons decaying in-flight Punch-through of hadrons to muons

Key methods Rely on real data, cross-check with well-tuned Monte Carlo Use the inclusive single muon sample to get extra information Reverse magnets to remove most detector asymmetry Do many many cross-checks in different phase-space regions

Result, 3.2 s !


45.06.0 103.2

SM % 0.3) 1( bA

hep-ex 1005.2757


Summary

Before the New D Inclusive Measurement A lot of measurements All consistent with SM +/- 1%

The New D Inclusive Measurement First evidence of departure from the SM Statistics-limited, so may improve over the next year

A lot of theory interest , so, how will LHCb help?

... The situation is significantly more complicated ...



Complications


The simple formula

qqb

q

qqp

qfs

qc

qfsq

fs S

B

t

tmaatA

22/cosh

cos

2222)(

10-3 -> 10 -5

ff

fftAqfs

)(


The simple formula

Polluting asymmetries are much larger than afs

Detector asymmetry dc ~(10-2)

Production asymmetry dp ~(10-2)

Background asymmetry db ~(10-3)

qqb

q

qqp

qfs

qc

qfsq

fs S

B

t

tmaatA

22/cosh

cos

2222)(

1/

/

1)(

)(

1)(

)(

0

0

SB

SB

IN

IN

f

f

b

p

i

ic

10 -2 10 -2 10 -310-3 -> 10 -5

Very Complicated

ff

fftAqfs

)(

LHCb


Detector Asymmetry, dc

Matter detector hadronic interactions are asymmetric

Magnet divides +/- charge, allowing +/- detector asymmetry We need to reverse the magnet regularly


MC Asymmetry in Muons

Left Right

+ve -ve

After 4Tm Magnet

Kaon PDG cross-section

PDG

0

20

40

60

80

100

120

0.1 1 10 100 1000

momentum in lab frame Plab / GeV c

had

ron

ic c

ross

-sec

tio

n /

mb. K-, p

K+, p

-1

Kaon interaction cross-section

K- pK+ p

LHC


An amazing machine

Unfortunately also not CP symmetric

Production Asymmetry, dp

LHC is a proton-proton collider: not CP-symmetric

LHCb is at high rapidity where production asymm. are largest

There is never a simple control channel to measure dp


-0.020

-0.015

-0.010

-0.005

0.000

0.005

0.010

0.015

0 50 100 150 200 250 300 350

Energy, E / GeVP

rod

uc

tio

n A

sy

mm

etr

y,

p

dp (B0 and B0)

Valence QuarkScattering

PYTHIA

Explicitly asymmetric at LHC


Why LHCb?

Why LHCb?

LHCb is a dedicated, precision, b-physics experiment

More statistics: we’re in the forward region, and at LHC


Being Timely

Proper Time: LHCb Velo precise down to 35 fs!


Being Precise

Particle ID: separation handled by dedicated subdetectors

Two RICHes, Calorimetry and Muon system


RICH 1(Vertical)

RICH 2(Horizontal)

MA

GN

ET

TT

T1-T

3

Being Exclusive

Our forte: exclusive, reconstructed, b-decays

In particular, time-dependent measurements


Monte-Carlo!

~100k Ds in 5 fb-1 ~100k Ds in 0.05 fb-1


Measurements

Measurements

Inclusive

Exclusive

Subtraction method

combine and


... bb

00 XDB ss

00 XDBd

ss DB0

00 XDB ss 00 XDBd

Inclusive at LHCb

Channel

Measured by D (see earlier)

Complications Physics! Production asymmetry: N(b)≠N(anti-b) [in acceptance] PYTHIA predicts dp(b) =

Mitigating factors None, difficult to interpret this measurement


... bb

2

dfs

sfsb aa

A

3103.04.3

45.06.0 103.2

SM

~108 per fb-1

Inclusive at LHCb

Channel

Measured by D (see earlier)

Complications Physics! Production asymmetry: N(b)≠N(anti-b) [in acceptance] PYTHIA predicts dp(b) =

Mitigating factors None, difficult to interpret this measurement


... bb

3103.04.3

45.06.0 103.2

SM

~108 per fb-1

2

dfs

sfsb aa

A

Hadronic at LHCb

Channel


KKDB ss0 ~105 per fb-1

CERN-THESIS-2008-045

CERN-LHCb-2007-017

Monte-Carlo!



Hadronic at LHCb

Channel

Measures

Complications Must fix either dc or dp in the fit – (db can be fit beforehand)

Mitigating factors Detector asymmetry small ~10-4 Fit afs and dp. – With excellent proper time resolution (35 fs)


ssb

s

s

sp

sfs

sc

sfss

fs S

B

t

tmaatA

22/cosh

cos

2222)(

KKDB ss0


CERN-LHCb-2007-054

~105 per fb-1



Hadronic at LHCb

Channel

Measures

Complications Must fix either dc or dp in the fit – (db can be fit beforehand)

Mitigating factors Detector asymmetry small ~10-4 Fit afs and dp. – With excellent proper time resolution (35 fs)


ssb

s

s

sp

sfs

sc

sfss

fs S

B

t

tmaatA

22/cosh

cos

2222)(

KKDB ss0


CERN-LHCb-2007-054

~105 per fb-1



Semi-leptonic

Channel (q = s/d)


00 XDB qq ~106 per fb-1

CERN-LHCb-2007-054

Monte-Carlo!

00 XDB ss

00 XDBd

Monte-Carlo!


Semi-leptonic

Channel (q = s/d)

Measures

Complications Missing neutrino makes proper-time resolution worse (≥120 fs?) Detector asymmetry large and difficult to measure

Mitigating factors Lots of statistics, but this makes dc even more important


00 XDB qq

qqb

q

qqp

qfs

qc

qfsq

fs S

B

t

tmaatA

22/cosh

cos

2222)(

CERN-LHCb-2007-054

~106 per fb-1


Semi-leptonic

Channel (q = s/d)

Measures

Complications Missing neutrino makes proper-time resolution worse (≥120 fs?) Detector asymmetry large and difficult to measure

Mitigating factors Lots of statistics, but this makes dc even more important


00 XDB qq

qqb

q

qqp

qfs

qc

qfsq

fs S

B

t

tmaatA

22/cosh

cos

2222)(

CERN-LHCb-2007-054

~106 per fb-1


The subtraction method

Take Bs/Bd with the same final states ( =KK p m) Background asymmetry: 2D fit in mass spectra Production asymmetry: fit with proper-time dependence Detector asymmetry: the same in each decay…

Do a simultaneous time-dependent fit

Measure the difference between Bs and Bd

Very comparible to the D measurement, but orthogonal to it!


f

22

)()(,dfs

sfsc

dfsc

sfsds

fs

aaaaA

45.06.0 105.2

SM


Take Bs/Bd with the same final states ( =KK p m) Background asymmetry: 2D fit in mass spectra Production asymmetry: fit with proper-time dependence Detector asymmetry: the same in each decay…

Do a simultaneous time-dependent fit


Very comparible to the D measurement, but orthogonal to it!


f

22

)()(,dfs

sfsc

dfsc

sfsds

fs

aaaaA

45.06.0 105.2

SM


Real data highlights

R Lambert, CERN 42CERN CPV, 18th June 2010

m-

PV

SV

TV

Bs

Ds+

+,K+,K-

EVT: 49700980RUN: 70684

LHCb Preliminary

mm

Real Data Bs

~800 mb-1

R Lambert, CERN 43CERN CPV, 18th June 2010

PV

SV

Bd

D+

K-,K+,+

TV

m-

EVT: 141526660RUN: 70666

LHCb Preliminary

mm

Real Data Bd

~800 mb-1

All D(s)->KKp


n.b. Selection with no requirement for a muon (need ~40 nb-1)

~2.6 nb-1


Sensitivities and Outlook

Sensitivity estimates

Full MC used to tune toy MC

Massive toy MC studies done in: CERN-LHCb-2007-054 CERN-THESIS-2008-045 CERN-THESIS-2009-001

Scaled to: Latest Monte Carlo efficiencies s(bb) = 500 mb


Stat. Error (500 mb)

100 pb-1 1fb-1

afss (Ds p) 2.1 x10-2 6.8 x10-3

DAfs (Dq mn) 2.0 x10-3 6.3 x10-4

All MC predictions!!Real data will be worse




Current Results


fs

fs

After 1fb-1 of LHCb


AssumeAb central valueand no NP in Bd

fs

fs

c.f. J/Y F


hep-ph 0612167

Directly Measure sin fs

s(fs) = 0.05c in 1 fb-1

Effectively Measures

s() = 0.5c in 1 fb-1

But they constrain NP differently Effective power enhanced NB physical limit of afs is at 4x10-3 < current D result!

/0 JBs

sfsa

sinRecosIm12

12

12

12

MMCurrent measurements (pre D0 paper)

Red : Dms

Yellow: DGs

Black: fs

Blue: afs


Combining D


D and CDF

-2bs=fs


FPCP 2010

http://agenda.infn.it/conferenceOtherViews.py?view=standard&confId=2635

D and CDF

(add my own personal merging)


FPCP 2010

http://agenda.infn.it/conferenceOtherViews.py?view=standard&confId=2635

Conclusion

D have made an astounding new measurement (3.2 s!)

At LHCb the environment is more hostile, concentrate on:

Low detector asymmetry, great proper time resolution

Detector and production asymmetries fitted with the data


ss DB0

Stat. Error (500 mb)

100 pb-1 1fb-1

afss (Ds p) 2.1 x10-2 6.8 x10-3

DAfs (Dq mn) 2.0 x10-3 6.3 x10-4

sfsa from

dsfsA, from 00 XDB qq

All MC predictions!!Real data will be worse

-310 (syst)]46.12.51(stat) 57.9[ bA

(a)

(b)

End

Backups are often required


Cereal loops


Now in bitesize-chunks!

Act now and getyour first 105 decays free!

New Physics inspecial boxes!

Afs DG and Dms in every box!

Two excitingflavours!

Nomenclature

A decay to a given state which cannot be reached by the anti-flavour state (at tree level) = flavour-specific

CP-violating asymmetry in mixing, manifests directly in flavour-specific decays = Afs/afs flavour-specific asymmetry

Since this is most readily observed in semi-leptonic decays it is also referred to as Asl

a final state of given flavour. its charge conjugate.


f f

After 1fb-1 of LHCb


LHCb measurement cuts at right-angles really depends what the value is, and if there is NP!

current favoured value Ab central value and no NP in Bd


Magnet divides +/- charge, allowing +/- asymmetry

by reversing magnet in D0: dc reduced from 3% -> ~0.1%


Asymmetry from Long Muon Tracks Reconstructed in MC

Left Right

-ve +ve

Left Right

+ve -ve

Before 4Tm Magnet After 4Tm Magnet


Matter detector hadronic interactions are asymmetric

Dominant systematic at order 1%


Kaon PDG cross-section

PDG

0

20

40

60

80

100

120

0.1 1 10 100 1000

momentum in lab frame Plab / GeV c

had

ron

ic c

ross

-sec

tio

n /

mb. K-, p

K+, p

-1

Kaon interaction cross-section

K- pK+ p

Resultant charge asymmetry (MC)

Systematics (1)

1. Detector asymmetry • Expect second order, so < 10-4

• Know the magnitude in real data from the pi momentum spectra• Detect any bias by binning in momentum• Correct with MC: Assuming the magnet is reversed!!

2. Production asymmetry• Separate using the time-dependence• Detect large bias from this by binning in eta• Measure production asymmetry, possible but very coarse

3. Background asymmetry • Fit simultaneously, simply, in the B-mass spectrum


ss DB0

Systematics (2)

1. Detector asymmetry only second order, so < 10-4

• Know the magnitude in real data from the daughter momenta• Detect any bias by binning in momentum• Correct with MC: Assuming the magnet is reversed!!

2. Production asymmetry• Separate using the time-dependence • Detect large bias from this by binning in eta• Measure in can help quantify any bias

3. Background asymmetry • Suppressed by the effective B/S < 0.1, in the signal region• Fit simultaneously, 2D fit, in the B-mass and D-mass spectra• Cancel detector-related part also, only production remains• Detect bias by binning in eta


00 XDB qq

ss DB0


Take Bs/Bd with the same final states ( =KK p m)

All production asymmetry is in x2/x3, just throw it away



f

2211,

dfs

sfs

dsdsfs

aaxxA

45.0

6.0 105.2 SM

Subtraction method

Channel (q = s/d)

Measures

Complications Requires weak constraints on DG and Dm

Mitigating factors Production asymmetry fit simultaneously Detector asymmetry cancelled


00 XDB qq

2

dfs

sfs

fs

aaA

45.0

6.0 105.2 SM


KKDqCERN-LHCb-2007-054




Production Asymmetry

Tuned Pythia samples

*=standard decays, †=Stable Bd+Bs

Asymmetries agree with generic bb events

dp x1000 Min Bias (10M)* bb – inclusive (10M)* (20M) †

Pions -(4.23±0.16) -(2.16±0.09) -(2.27±0.07)

Kaons -(17.0±0.5) -(7.73±0.26) -(8.2±0.2)

Muons -- +(2.0±1.2) +(1.0±0.9)

Ds -- -(1.6±1.1) -(1.6±1.1)

Bs -- -(1.9±1.3) -(1.5±0.8)

Bd -- -(3.2±0.7) -(3.2±0.4)

00ds BB

Documents

R Lambert, CERNCERN CPV, 18th June 20101 Flavour-specific asymmetry and LHCb Robert W. Lambert On behalf of the LHCb collaboration