b physics at LEP

Andrea Sciabà

INFN-CNAF

XXXIII International Symposium on Multiparticle Dynamics

September 5-11, 2003Krakow

ISMD2003Andrea SciabàINFN-CNAF

b physics

Goals: CKM matrix

B oscillations (|Vtd|, |Vts|) B semileptonic BRs (|Vcb|, |Vub|) b lifetimes CP violation

masses Electroweak

sin2 θW (asymmetries) Rb (sensitive to new physics, mtop)

QCD processes fragmentation two-photon events, gluon splitting, etc.

New physics (CP violation, rare decays, etc.) …

B 0 oscillations

V udV ub

*

A B

C

Vtd V

tb *

VcdVcb*

b cℓν

b

uℓν

ISMD2003Andrea SciabàINFN-CNAF

b physics at LEP LEP data sample features and

advantages ~ 3.6×106 Z0 bb events very clean environment high heavy quark tagging power due to

silicon microvertex detectors production of many species of b hadrons high boost of b hadrons

… but many results being superseded by pp colliders and b factories

ISMD2003Andrea SciabàINFN-CNAF

b physics at LEP

LEP data still contributes significantly to b semileptonic BRs Bs oscillations Asymmetries, Rb (no more measurements until Z

factories)

Bs, Λb, excited b hadron masses Fragmentation heavy quark production in 2-photon collisions gluon splitting into heavy quarks

Only a few topics will be covered here

ISMD2003Andrea SciabàINFN-CNAF

Bs oscillationsBs oscillations: Basic concepts

Main motivation:

md |Vtd| gives a large error

md/ ms |Vtd| with a 3 times better theoretical error!

BBss,d,d00BBss,d,d

00

bb

bb

s,ds,d

WW-- WW++

VVtbtb

VVttqq

VVttqq**

VVttbb**

tt

tt

s,ds,d

15% theor. error

22*

qqq BBBtbtqq fBMVVm

2

2

std

ts

B

B

d

s

V

V

m

m

m

m

d

s

|Vcb|2

SU(3) ~ 5% error

)cos( )( 0s

0s BB

tmetP s

t

s

s

1

2

1

Oscillation probability:

fit A for any assumed value ω of Δms

ω <<Δms A 0

ω ≃Δms A 1

ω excluded at 95%CL if A+1.65σA<1

A

2

2

121

1)(exp)(~ tB

As

fNS

ISMD2003Andrea SciabàINFN-CNAF

Strategy Flavour tagging at production

jet charges, vertex charges, kaons, leptons, etc.

typical mistag 25% Flavour tagging at decay

lepton or D charge identify final state (b, bc,

etc.) Distinguish Bs from other hadron

species Measure Bs decay length,

momentum t = (m/p) L (p/p) t

Maximum likelihood fit to amplitude A at given ms

Production flavour tagging

Decay flavour tagging

ISMD2003Andrea SciabàINFN-CNAF

Oscillation analyses Exclusive analyses

Fully reconstructed Bs (A, D)

Semi-exclusive analyses Ds

+- pairs (A, D, O) Ds

+h- pairs (D) Φ- pairs (D)

Inclusive analyses semileptonic decays (A, D, O) inclusive vertices (D)

Very low statistics

High purity

Very high p.t. resolution!

Low statistics

Fair purity

Good p.t. resolution

High statistics

Low purity

Low p.t. resolution

ISMD2003Andrea SciabàINFN-CNAF

Fully reconstructed Bs (ALEPH)

s

1s

s

s

D

aD

D

B 0

KK

KKD

S

0s

0

000 a1

ss DD

,Ds0

,aDs0

1

total

and 0 reconstruction efficiency and purity increase

180 m

p/p 0.5 – 3 %

Well reconstructedMissing ’s or ’s

t0.08 ps

Excellent proper time resolution!

Add a photon

ISMD2003Andrea SciabàINFN-CNAF

Ds+- pairs (DELPHI)

(updated)

KKD

D

D

KKD

KKD

D

*s

s

s

Ss

*s

s

0

0

0

0

KKD

KKD

s

s

;

;eX ss DB0

Semileptonic modes

Hadronic modes Event-by event:

proper time resolution Bs purity

Sensitivity: ms = 8.6 ps-1

ms > 4.9 ps-1 at 95% CL

ISMD2003Andrea SciabàINFN-CNAF

Inclusive analysis (ALEPH) Topological reconstruction

of the D vertex Extensive event-by-event

treatment of: proper time resolutions and

corrections sample composition flavour tags huge sensitivity

enhancement!

Lepton with high p

Topologically rec. “D” track

“B” track

Photon

Primary vertex

Bs

“D”

Sensitivity: ms = 13.6 ps-1

ms > 11.9 ps-1 at 95% CL

ISMD2003Andrea SciabàINFN-CNAF

Oscillation results Almost final LEP results

on Bs oscillations Lower limit on ms: 14.4

ps-1

Now, we must wait for Tevatron to measure ms

ISMD2003Andrea SciabàINFN-CNAF

b quark fragmentation The fragmentation function

models the non-perturbative confinement of the quark into colourless hadrons

different choices of the variable:

not accessible experimentally

ISR, FSR, hard gluon emission effects not unfolded

Model-dependent analyses (fit to a given fragmentation model)

Model-independent analyses (xB spectrum reconstruction)

quark||

hadron||

)(

)(

pE

pEz

beam

hadron

E

ExB

ISMD2003Andrea SciabàINFN-CNAF

Inclusive analyses Event selection based on:

b-tagging secondary vertex

reconstruction Direct xB reconstruction

(OPAL) neural networks to distinguish

particles from b decay and from hadronization

σE≃5 GeV

xB, z reconstructed using neural networks (DELPHI), input variables correlated to: kinematics (track rapidity,

multiplicity, etc.) jet/total energy b-tagging

prelim

inar

y

ISMD2003Andrea SciabàINFN-CNAF

Exclusive analyses (ALEPH) B D(*)ℓν decays

Much lower statistics ~ 3400 events

Eν reconstructed from missing energy

σE ~ 8%

xB spectra not efficiency-corrected

ISMD2003Andrea SciabàINFN-CNAF

Model dependent results Favoured models:

Lund / Lund-Bowler Kartvelishvili

Disfavoured models: Peterson et al. Collins

ISMD2003Andrea SciabàINFN-CNAF

Model independent results

-0.0033

-0.0052

Mean values for xB:

OPAL: 0.7193±0.0016(stat.)+0.0038(syst.)

DELPHI: 0.7153±0.0007(stat.)+0.0049(syst.)

ALEPH: 0.716±0.006(stat.)±0.006(syst.)

All measurements consistent bias w.r.t. old measurements due to

theassumption of the Peterson model

ISMD2003Andrea SciabàINFN-CNAF

AbFB measurements (I)

Forward-backward asymmetry at the Z pole:

Related to sin2θW

AbFB most sensitive to sin2θW

Need to measure: cosθ (from thrust axis) quark charge quark flavour

QCD corrections to be calculated gluon radiation thrust axis approximation most important systematic error

BF

BFqFBA

coscos

cosqFBAd

d

3

81 2

22 1

2

1

2

4

3

)()(qe

qq

qq

ee

ee

AV

AV

AV

AVqFB gg

gg

gg

ggA

AA

f

f

A

Vf

g

g1

4

12effsin

effsin 23

21

q

q

A

V

I

Q

g

g

q

q

e- e+f

f

ISMD2003Andrea SciabàINFN-CNAF

AbFB measurements (II)

Use a lepton tag lepton charge

correlated to the quark charge

need to distinguish

correct for the dilution by B mixing

background

cbb

cbb

c

bobs )( FBFB AA 21

bb FBBF AQQ

Use the jet/vertex/kaon charge b-tagging to enhance

purity purity from data quark charge estimator

Q

charge separation δ from data effect of B mixing

included

ISMD2003Andrea SciabàINFN-CNAF

AbFB with lepton tag

(DELPHI) b-tagging (ηEVT)

Track I.P., sec.vertex mass and energy, track rapidities

Flavour composition vs. ηEVT from data with standard double counting method

Lepton p and p

Charge tagging Lepton charge, opposite jet charge as

cross check Uses Q× Qopp (-/+: right/wrong correlation)

Probability for a lepton of coming from b, bc , c , bckgnd function of (p, p), (ηEVT, Q× Qopp)

2-dim fit of the obs asymmetry to AbFB,

AcFB over bins of (cosT, Pb-Pbc , Pc)

(syst) 0063.0(stat) 0086.00728.0

(syst) 0024.0(stat) 0052.01021.0,0

FB

,0FB

c

b

A

A

New!

ISMD2003Andrea SciabàINFN-CNAF

AbFB with lepton tag

(OPAL) Neural networks to

distinguish b (NETb) and c (NETc) Lepton p and p, lepton jet

energy, (∑p)jet

Jet decay length significances, lepton i.p. significance

Likelihood fit to the number of single-lepton and double-lepton events Bins of cos, NETb and NETc

Background fraction fitted from data

average mixing parameter fitted from data using same-sign dilepton events

AbFB and Ac

FB fitted(syst) .(stat) ..(syst) .(stat) ..

,FB

,FB

003800052006290001500040009810

0

0

c

b

AA

ISMD2003Andrea SciabàINFN-CNAF

AbFB with incl. tag

(DELPHI) high purity b-tagging (b-tag)

lifetime and sec vertex info, rapidities b-tag selection efficiency from data

counting events with 0,1,2 tagged hemi sample composition (b, c, uds)

charge tagging (flavhem) jet charge, sec. vertex charge flavour tag combining P(same Q|B) for

all tracks prob. of charge mistag from data

(unlike/like-sign double tagged events) for charm events, mistag from data using

charge of reconstructed D* (new!) fit to no. of single and double tagged

events in bins of (cosθ, b-tag)

Main systematics from jet charge hem. correlations

(syst) .(stat) ..,FB 0014000300097800 bA

New!

ISMD2003Andrea SciabàINFN-CNAF

AbFB with incl. tag

(OPAL) b events tagged with

NN (secondary vertex info) leptons (p, p, isolation) flavour composition from data (singly

and doubly tagged hemispheres) Pf

charge tagging estimator Q from jet charge (always) sec. vertex charge (if any) kaon charge (from bc s) (if any)

all charge estimators available

f

FBfffBF ACPsQQ flavours

ff BF QQ

4

2 f

ffP flavours

(syst) .(stat) ..,FB 0018000340100200 bA

ISMD2003Andrea SciabàINFN-CNAF

AbFB summary

AbFB measurements all LEP

almost finalized well compatible (also with SLD) sin2lept

eff: 2.9σ discrepancy with ALR (SLD) need to wait for linear colliders

SLD Ab

LEP A bFB

LE

P+

SL

D

Abe

bFB AAA

4

3

ISMD2003Andrea SciabàINFN-CNAF

Open b production in events Test of perturbative QCD

at LO, only direct and single resolved terms (of the same order)

more robust calculations because of the large quark mass

2-3 orders of magnitude suppression w.r.t open charm because of larger mass, smaller charge

photons mostly quasi-real anti-tagged events

Analysis strategy select anti-tagged photon-photon events reconstruct lepton candidates reconstruct jets measure lepton p w.r.t. jet axis

e-

e+

e+

e-

b

b

e-

e+

e+

e-

b

b

direct term

single resolved term

ISMD2003Andrea SciabàINFN-CNAF

bbX in L3 627 pb-1, 189 GeV < s < 209 GeV muons AND electrons large

statistics Fit to p distribution:

p spectra from MC (PYTHIA for )

Nb, Nc, Nuds left free Consistent e and results

Also measurement of

consistent with previous L3 results

(syst)pb.(stat)..)Xbbeeee((syst)pb.(stat)..)Xbbeeee(

electrons

muons

324261232352013

(syst)pb.(stat)..)Xbbeeee( 3271812

muons

electrons

(stat)pb)Xcceeee( 117998

435 events

3850 events

ISMD2003Andrea SciabàINFN-CNAF

bbX in DELPHI (new)

pb (syst) .(stat) ..)Xbbeeee( 4333914

pb (syst) (stat) )Xcceeee( 206191937

463 pb-1, 189 GeV < s < 209 GeV only muons Fit to p distribution:

p spectra from MC (PYTHIA for )

Nc fixed to LEP average for (e+e-cc)

Nuds measured selecting hadrons instead of muons

Nb left free

Study of K-lepton correlations (first time!) bK--X cK-+X increased purity of c (K±) and b (K±±)

events measured cross sections:

651 events

ISMD2003Andrea SciabàINFN-CNAF

bbX in OPAL 371 pb-1, 189 GeV < s < 202

GeV only muons Fit to p distribution:

p spectra from MC (PYTHIA for )

Nc fixed to OPAL measurement for (e+e-cc) with D* mesons

Nuds measured selecting hadrons instead of muons

Nb left free

xT=2pT/Wvis distribution shows the consistency of the 1:1 ratio of direct to single-resolved contributions

pb (syst) (stat) ..)Xbbeeee( 843552214 .-.

444 events

ISMD2003Andrea SciabàINFN-CNAF

bb(cc) summary Consistent results among LEP

experiments Good agreement with Drees

et al. model for charm data

EXCESS for beauty data:4 discrepancy

No satisfactory explanation at present

ISMD2003Andrea SciabàINFN-CNAF

Conclusions

LEP HF physics program almost finalized Bs oscillations (ms> 14.4 ps-1 at 95% CL) heavy quark asymmetries (difficult

interpretation) Heavy quark fragmentation function well

measured and consistent (both in average and shape)

Large discrepancy in bbX rate (but not in ccX rate)

Impossible to show many other results…