Measurement of the Ratio B(b c)/B(b c)

&& Search for Anomalous

Production of Diphoton + e/at CDF

Shin-Shan Yu (余欣珊 )

Fermi National Accelerator Laboratory

Academia Sinica HEP Seminar, January, 2007

Shin-Shan Yu 2

DATA SAMPLE FOR THE ANALYSES

Lambdab relative BRSep 2003 shutdown

Diphoton + e/muMar 2006 shutdown

Measurement of the Ratio B(b c)/B(b c)

Shin-Shan Yu 4

OUTLINE

Why b

CDF Detector, Trigger

b Relative Branching Fractions

budb

Shin-Shan Yu 5

BIG PICTURE : WHY b BARYON?

u d b

Heavy Quark Effective Theory (HQET) simplifies the extraction of CKM elements

Assuming mb >> QCD Corrections expressed in the power

of 1/mb and s(mb) Spin-independent interaction

between light quarks and b quark

Spin of light degrees of freedom = 0, the corrections are simpler than those of b-mesons.

%45.0 )(

% 6.6)(

theorycb

theorycb

B

B

q

ud

0s

Shin-Shan Yu 6

HOW B(b c)/B(b c) ?

Four charged tracks in the final state

Data come from the same trigger, most systematics cancel.

Control samples: similar decays in the B meson system

Relative BR is the yield ratio corrected for the efficiency, e.g:

cbcb

cbcb

N

N

cb

cb

)(

)(

BB

)(

)(0

0

DB

DB

BB

)(

)(*0

*0

DB

DB

BB

and

But since we can not reconstruct neutrinos, several backgrounds can fake our semileptonic signals in the data ….

cbcb

cbc

N

NN XB

cb

cb

)(

)(

)( bgmix

BB

Shin-Shan Yu 7

CDF DETECTOR & TRIGGERSilicon Tracker

|| < 2vertex ~ 30 m

Central Outer Tracker (COT)96 layers drift chamber, up to ||~1

PT /PT~ 0.15% PT

Central Muon Chamber4 layers drift chamber outside the calorimeter|| < 0.6

Two Displaced-track Trigger pT > 2 GeV/c, 120 m ≤ d0 ≤ 1 mm,

Lxy > 200 m, pT > 5.5 GeV/c 150 M events analyzed for this measurement

Shin-Shan Yu 8

FORMULA REMINDER

cbcb

cbc

N

NN XB

cb

cb

)(

)(

)( bgmix

BB

Shin-Shan Yu 9

ANALYSIS REQUIREMENTS

2 out of 4 tracks matched to trigger tracks (two displaced-track trigger)

Muon 120 m ≤ d0 ≤ 1 mm

Pt(4 trks) > 6 GeV/c

Pt(Lc) > 5 GeV/c

Pt(, ) > 2 GeV/c

Pt(proton) > 2 GeV/c

Chi2 of secondary vertex fit < 15

Chi2 of tertiary vertex fit < 14

c of Lb candidate > 250 m

cof Lc candidate > -70 m

3.7 < M(Lc) < 5.64 GeV/c2

Shin-Shan Yu 10

SIGNAL SAMPLE b cX c p+ K- + Inclusive Semileptonic Signal

NDF=36.6/42Prob=70.7%

Inclusive b hadron -> X decays reconstructed as c

Hadronic SignalBackground shapes come from MC

Combinatorial and b-> other c hadron semileptonic decays

Shin-Shan Yu 11

FORMULA REMINDER

cbcb

cbc

N

NN XB

cb

cb

)(

)(

)( bgmix

BB

Shin-Shan Yu 12

MC AND DATA COMPARISON We used MC to obtain relative efficiencies of signals and backgrounds. Compare MC and background subtracted signal distribution in the data. Tune our MC if MC and data disagree, e.g: PT(b)

PT(b) [GeV/c]

BEFORE AFTER

PT(b) [GeV/c]

Shin-Shan Yu 13

FORMULA REMINDER

cbcb

cbc

N

NN XB

cb

cb

)(

)(

)( bgmix

BB

Shin-Shan Yu 14

SOURCES OF SEMILEPTONIC BACKGROUND

Feed-down/Feed-in from single b-hadron decays

Single b-hadron -> …. -> c additional particles e.g:

c

cb

)2593(

Hadrons mis-identified as muons

Real Lambdac and real muon from the decay of two heavy-flavor hadrons

Shin-Shan Yu 15

FEED-DOWN/IN FROM SINGLE B-HADRON Reduced by the M(c) cut

Normalize the amount of background to the measured hadronic signal

hadronichadronic

feedfeed

hadronic

feed

B

Bi

N

N

BR from PDG or estimated from theory and CDF preliminary measurements

Reconstruct as many backgrounds as possible

PDG BR(b -> c X)=9.2%constrain

BR of the other backgrounds

~10% contribution

Shin-Shan Yu 16

?

MIS-IDENTIFIED MUONSp, K, fake muons

cand muon d0 cuts suppress prompt fakes

Our fakes mostly come from b decays

Weight “c+TRKfail” events with the Pfake

~3% contribution

Shin-Shan Yu 17

c FROM TWO HEAVY-FLAVOR HADRONSCharm and from different b- or charmed hadrons

Suppressed due to the cand PT() cuts

Rely on Pythia MC

Most sensitive to hard gluon splitting

~0.2% contributionAssign 100% uncertainties

c+

p+

p+

c+

Shin-Shan Yu 18

SEMILEPTONIC SAMPLE FRACTION

4.3%

15.0%

79.8%

0.9%3.2%

9.8%

86.8%

0.2%

Signal

Feed-in/down

Fake mu

Two different heavy-flavor hadrons

4.9%

40.0% 53.9%

1.2%

Xc XD XD *

Shin-Shan Yu 19

CONSISTENCY CHECK

Shin-Shan Yu 20

UNCERTAINTY SUMMARY fractional uncertainty (%)

Measured BR +3.5 -10.5 8.2 2.3

Estimated BR 2.5 9.2 6.2

CDF Internal

Mass fitting 3.2 4.1 < 0.1

Muon fake 0.9 0.7 0.4

Pt spectrum +1.4-2.5 3.2 2.2

CDF material 1.1 1.7 1.3

scaling 0.4 0.5 0.4

0.2 2.2 1.3

b,c polarizations 1.9 -- --

c Dalitz 0.4 -- --

b lifetime 1.1 -- --

b decay model 2.9 -- --

6.0 6.1 3.4

Statistical 15.0 10.2 13.0

ccbb ,

)

)0

0

DB

DB

B(B(

)

)*0

*0

DB

DB

B(B(

)

)

cb

cb

B(B(

Feed-in/down background and hadronic signal branching fractions

MC modeling of production, decay, efficiency and acceptance

Shin-Shan Yu 21

CONTROL SAMPLE RESULT

(UBR) 0.9(BR) 0.8(syst) 0.6(stat) 0.18.9)

)0

0

DB

DB

B(B(

Consistent with the 2004 world average 7.8 1.0 at the 1level New world average ratio 8.3 0.9

(UBR) 1.1 (BR) 0.4 (syst) 0.6 (stat) 3.27.17)

)*0

*0

DB

DB

B(B(

Consistent with the 2004 world average 19.7 1.7 at the 0.7 level New world average ratio 19.1 1.4

Shin-Shan Yu 22

SIGNAL SAMPLE RESULT

Experimental uncertainties dominated by data sample size

(UBR) 0.5 (BR) 2.1-

0.7 )(2.1 (stat) 0.30.20

)

)

systcb

cb

B(B(

Unexpected and additionalphysics results

Shin-Shan Yu 24

FIRST OBSERVATION of NEW b DECAYSEstimated BR of the reconstructed background based on the first observation

c++

CDF Run II Preliminary 360 pb-1

CDF Run II Preliminary 360 pb-1

c(2593)

c(2625)

CDF Run II Preliminary 360 pb-1

Xcb 0

c0

Xcb

Xcb *

c++,+,0

c(2625)c(2593)

or *

ccc

cc

Shin-Shan Yu 25

084.0236.0)10(

)10(

0

TB

T

P

Pb

I Run CDF

BR) ( 22.0)( 11.0 (stat) 08.082.0)(

)(

)/ 6(

)/ 6( II Run CDF 0

0

ccb

TB

T systDBcGeVP

cGeVPb

BB

CROSS-SECTION RATIO AND B(b c

% )( 08.0

06.0)(stat 19.041.0)(

pTsystcb B

)( 11.0

19.0)(stat 20.056.0

) 6(

) 6(

GeV/c

GeV/c

0

pTsystP

P

TB

Tb

Consistent with the prediction 0.45% (Phys. Lett. B586, 337)

???

Make use of previous CDF measurements

b PT spectrum using fully reconstructed decay was not available for CDF I

Correct the CDF I cross-section ratio using measured PT spectrum

Acceptance correction

Different PT thresholds affect the ratio 10 GeV/c vs. 6 GeV/c

Shin-Shan Yu 26

AND b PT SPECTRA FROM DATA

Shin-Shan Yu 27

WHAT DO WE KNOW ABOUT b NOW?

c l c

20.0 3.7c(2593) +

l seenc(2625)

+l seenc

++ l

seenc

0 l

seen

(4.1 2.0) x 10-

3

Siganature-based Search for Anomalous Production of

Diphoton + e/ at CDF

Shin-Shan Yu 29

MOTIVATION

✔CDF Run I “ee+MET” event: 86 pb-1

✔Dominant SM: WW 810-7 events✔Total: ~10-6 events

✔Personal✔Use calorimeters✔Learn high-pt physics

Phys. Rev. D59, 092002 (1999) by Toback, et al

Shin-Shan Yu 30

RUN II PHOTON ANALYSES

Diphoton + X X = e, , , , met, b-jet, jet

Expand photon analysis to a more systematic search Do not optimize cuts based on any model Only compare kinematic distributions and number of events Look for excess above background prediction Not a blind analysis

Lepton + photon + X X = met, e, , met + b jet

Other model-dependent photon analyses RS graviton to diphoton search more …

+X signaturesSUSY: +MET+X

b`b`bbl*l*ll,

q*q*qqNew dynamics

Shin-Shan Yu 31

A RUN II e EVENT

Is it from SM? Fakes from jets or electrons ? Or new physics?

Shin-Shan Yu 32

ANALYSIS OVERVIEW

Require two central photons and lepton objects

photons || < 1.0

muons || < 1.0

electrons || < 2.0

Specific detector requirements for each lepton type

Data sample ~ 1.1 fb-1 collected with diphoton trigger

2 EM clusters with Et > 12 GeV

isolation, E_had/E_em, shower profile requirements

Compare prediction with observation by applying several levels of loose and tight photon requirements

Includes a wire chamber at showermax

Shin-Shan Yu 33

ANALYSIS REQUIREMENTS

cuts Tight cuts Loose cuts

Photon ET >13 GeV

Showermax fiduciality |XCES|<21.0 cm; 9.0 cm<|ZCES|<230.0 cm

Average CES 2 <20

Had/Em <0.055+0.00045*ET <0.125

Calorimeter Energy Isolation <2.0+0.02(ET-20) or

<0.1*ET if ET<20.0 GeV

<3.0+0.02(ET-20) or

<0.15*ET if ET<20.0 GeV

Track Pt Isolation <2.0+0.005*ET <5.0

Tracks point to photon candidate

0,1

Track Pt if there’s one track pointing

<1.0+0.005*ET <0.25*ET

2nd Showermax cluster (wire or strip)

<0.14*ET if ET<18 GeV

<2.4+0.01*ET if ET>18 GeV

No cut

Shin-Shan Yu 34

DIPHOTON + e/FROM SM AND FAKESW Z : Madgraph MC, weighted with data/MC ID scale factor, luminostiry, trigger efficiency, K factor to correct LO to NLO cross-section

Fake leptons + real photons: apply lepton fake rates to the events with 2 photon candidate + fakeable object, then scale with the true photon fraction obtained from MC and data (showermax and preshower radiator)

True photon fraction ~ 30% for photon Et > 13 GeV

Leptons + fake photons from jets: apply jet faking photon rate to the events with lepton candidate + photon candidate + a central jet

W Z + jet+ 2 jets where 1 jet fake leptonW or Z + 2 jets where both jets fake photonsZ + jet where electron fakes a photon

Leptons + fake photons from electrons: apply electron faking photon rate to the events with lepton candidate + photon candidate + an electron

Z where the electron comes from decay of ZZ + jet where the jet fakes a photon

Shin-Shan Yu 35

DOMINANT BACKGROUND

electron channel• ee from Zwhere one electron fakes the photon

muon channel• Z

Z

Z Z

l

Shin-Shan Yu 36

REJECT FAKE PHOTONS USING SILICON TRKElectron can fake a photon due to track-loss, FSR or BREM

Background from BREM can be reduced by removing photons matched to silicon tracks

tracks seeded by event vertex and a cluster in the EM calorimeter

Measure the fake rateCompare Z peak from ee and eGet Et dependence from MC

Normalize to data

Silicon-track rejections reduce the fake rate by a factor of 3-4

0.36% at 45 GeV

Silicon track rejection applied

Shin-Shan Yu 37

ResultBefore Sili rejection After Sili rejection

Channel e e Prediction 6.82±0.75 0.79±0.11 3.79±0.54 0.71±0.10

Obseverd 3 0 1 0

M(e)=129.5 GeV/c2

M(e1)= 93.0 GeV/c2

M(e2)= 88.3 GeV/c2

The event which survives silicon-track rejection.

= Sum Et of observed objects and missing Et

Shin-Shan Yu 38

CONCLUSION

We have measured the Lb relative branching fraction using 173 pb-1We have made the first observation of 4 Lb semileptonic decaysWe have the first evidence that the Lb pt spectrum is softer than that of the B0 meson. Although Bs mixing is already observed, there’re still interesting B physics to do at the Tevatron. With current available data, we could make a more precise measurements of b-baryon properties or search for unobserved b-baryons.

We have searched for anomalous production of diphoton + e/mu events in 1 fb-1 of data. No excess is observed.

Removing photons matched to silicon tracks reduce the electron faking photon rate by a factor of 4Keep looking …

Back Up Slides

Shin-Shan Yu 40

Control Sample DX, D+ K-+ +

Background shapes come from MC

Hadronic SignalInclusive Semileptonic Signal

Shin-Shan Yu 41

Control Sample D*X, D*+ D0 +, D0 K-+ Inclusive Semileptonic Signal

Background shapes come from MC

Hadronic Signal

Shin-Shan Yu 42

Physics Background

Shin-Shan Yu 43

Dominant Systematics Physics background and hadronic signal branching fractions

Measured: from PDG Estimated: multiply the BR by 2 or 0

Mass fitting model Vary the constant parameters in the fit Several background shapes come from inclusive MC

vary BR of the dominant decays

MC modeling of acceptance and efficiency pT spectrum

affect the efficiency and B (b c b semileptonic decay model

size of the independent MC for reweighting the phase space distribution uncertainty on the predicted form factors

Shin-Shan Yu 44

MC Tuning

Flat phase space

Form factor weighted

Shin-Shan Yu 45

b Polarization

cos1

cos bd

dN