Shin-Shan YuPhysics with Photons at CDF 1 從未知世界來的光 : CDF 實驗的光子物理清華大學高能物理演講西元 2009 年 2 月 26 日余欣珊費米實驗室

Shin-Shan Yu Physics with Photons at CDF 1

從未知世界來的光 :CDF實驗的光子物理

清華大學高能物理演講西元 2009 年 2 月 26日

余欣珊費米實驗室


Invisible Light from the Unknowns: Physics with Photons at CDF

Shin-Shan YuFermi National Accelerator Laboratory

High Energy Physics SeminarNational Tsing Hua University

February 26th, 2009


Standard Model: Success

• Predicted existence of W, Z, and gluon

• Unification of weak and electromagnetic interactions

• Theory and experiments agree over a wide range of measurements Observed all SM particles

except Higgs e.g. W, Z mass with great

precision (better than 0.01%)


Standard Model: Puzzles

1 top quark = 336000 electrons

1. Why are some particles so much heavier than the others? What is the origin of mass? Higgs not found yet

2. Why are there so many kinds of elementary particles?

3. Do all the forces become one? Gravity not included yet

4. How to explain the anti-matter and matter asymmetry of today’s world?

5. How do neutrinos mix?To be continued …


Physics Beyond Standard Model1. Why are some particles so much heavier than the others?

What is the origin of mass? Technicolor

2. Why are there so many kinds of elementary particles? 4th Generation, Compositeness, Leptoquark

3. Do all the forces become one? Supersymmetry, Extra Dimension

To be continued …


Circle of Particle Physics

Puzzles

SM ParametersNew Particles

New Theory Parameters

Measurements of SM

Parameters (Test of SM)

Search for New Phenomena

A Miracle!

Studiesof

New Phenomena

Models That Describe

New ResultsOne Theory?

Theory

Experiment


Outline

Overview of Particle Physics • Tevatron and CDF

Signatures in the Detector

• Why Photons• Measurement of Differential Photon

Production Cross Section• Signature-based Search for New Physics

with Photons

• Outlook


Outline

Overview of Particle Physics

• Tevatron and CDF• Why Photons• Test of Standard Model Parameters• Search for New Phenomena

• Outlook


Fermilab Tevatron and

Collider Detector at Fermilab (CDF)


Fermilab Tevatron~70 km west of Chicago

p 0.98 TeVs=1.96 TeV

p0.98 TeV?

TeV=1012 eV 99.99995% of the speed of light


Facts about Fermilab Tevatron

• Run I (1990-1995)s = 1.8 TeV110/pb per experiment

• Run II (2001- ?)s = 1.96 TeV2.0-2.5/fb in this talk~ 4.8/fb per exp. so far ≥ 8.0/fb per exp. by the end of Run IIRun until 2009-2010 (proposed to run until 2012,

10-15/fb)


Collider Detector at Fermilab (CDF)

• 600+ physicists 61 institutions 15 countries

(including Taiwan)

• 435 publications 180 in Run II

• I am on CDF

2800 Ford Mustangs5 Eiko × 5 Eiko × 5 Eiko

Weight ~ 4500 tonsSize ~ 7.6 m cube


Tracking

System

CDF Sub-Components

Silicon Vertex Detector Drift Chamber

+ +

Solenoid

Muon chamber

EM Cal.

Hadron Cal.

Calorimeter

=


The Magnificent Seven Signatures in the Detector

e jet Beam Axis

Tracking system

EM Layers

Hadronic Layers

Calorimeter

Muon Chamber

b-jet

jet

0)()( ,, jpp yxyx

Momentum imbalance

MET


Why Photons?


Photons in This Talk• Focus on photons with energy > 25 GeV (25×109 eV)

1024 Hz

10-16 m

Photon in this talk

Visible light: 380-750 nm, 400-790 THz


Why I Don’t Like Photons• Hard to get pure sample of photons for calibration

Jets may be misidentified as photons, jet/photon ~ 1000

• Do not know where photon originates from (contamination of non-collision background) Cosmic ray muons: bremsstrahlung due to interaction with

calorimeter

p p

Cosmic ray muon


Why I Like Photons• Copiously produced in p-pbar collisions, only after jets

Test cross sections predicted by theory

• Energy well-measured by the EM calorimeter Measure mass of exotic particles

• Photons do not decay No reduction in rates due to decay branching fractions (BR ~10.8% for Wl and ~ 3.3% for Zll )

• High identification efficiency ~85% (if it does not pair-produce) For comparison, b-jet identification efficiency is 30-50%

• Many theories beyond the standard model predict signatures with photons

EM energy resolution > 2 times better than HAD energy resolution


Production Rates of SM ParticlesRate: N/sec4 x 106

4 x 103

2 x 100

5 x 10-1

5 x 10-4 (2/hour)

cc

Jets (including b-jets)


Models Which Predict Photon Signatures

• Supersymmetry 10 G or 2

0 10

ll jbjjjl displaced

• Compositeness X* X bb, ll, ll

• Technicolor T

T

bb

• Extra dimension G or qq G

• Higgs gg (qq) H (W,Z) (W,Z), qq H±h 4W l ll jj

• 4th Generation b' b bbX

G~


Published CDF Photon AnalysesInclusive Photon Cross Section (QCD)

Angular Distribution (QCD)

PRL 73 (94), PRD 48(93),PRL 68 (92), PRL 71 (93), PRD 65 (02), PRD 70 (04)

Photon + Dijet Dijet Properties (QCD) PRD 57 (98)

Photon + Muon Intrinsic Charm (QCD) PRL 77 (96), PRD 60 (99)

Photon + Jet + X Jet Distribution (QCD) PRD 57 (98)

Photon + Lepton+X Signature-based (Search)

PRD 66 (02), PRL 89 (02), PRL 97 (06), PRD 75 (07)

Photon + Z or W Anomalous Couplings (EWK) Cross Section (QCD)

PRL 74 (95)PRD 73 (06), PRL 94 (05)

Photon + b-jet + X Techni-Omega, Signature-based (Search)

PRL 83 (99), PRD 65 (02)

Photon + Ds W -> photon + Ds (EWK) PRL 77 (96), PRD 58 (98)

Photon + Track W -> Photon + pion (EWK)

PRL 76 (96), PRL 69 (92)PRL 58 (98)

Delayed Photon GMSB (Search) PRL 99 (07), PRD 78 (08)


Published CDF Photon AnalysesExclusive Photon + Met

Extra Dimension (Search)

PRL 89 (02), PRL 101 (08)

Photon + Dilepton Excited lepton (Search) PRL 97 (06), PRL 94 (05)

Exclusive Diphoton Diffractive (QCD) PRL 99 (07)

Diphoton + X Signature-based (Search)

PRL 81 (98), PRD 59 (99)

Diphoton + Met GMSB (Search) PRD 71 (05)

Diphoton Cross Section (QCD) PRL 70 (93), PRL 95 (05)

High-mass Diphoton RS Graviton (Search) PRL 99 (07)

Diphoton + W/Z Fermiophobic Higgs (Search)

PRD 64 (01)• 38 Papers published in Physical Review Letters and

Physical Review D Electroweak and Strong forces (QCD), Physics beyond

Standard Model A big output from a small group



Puzzles



Measurements of SM



A Miracle!

Studies of

New Phenomena




Measurement of Differential Photon Production Cross Section

Work with R. Culbertson (Fermilab), C. Deluca, S. Grinstein, M. Martinez (IFAE, Barcelona)


Test of Strong Interaction

• Test perturbative QCD Use PDF obtained from DIS, Drell-Yan, jet cross-sections

• Potentially provide information on the non-perturbative part (PDF) PDF used to predict X-section of new physics

Protonstructure

(PDF)

SM

BSM

X

X-section

=

perturbativeNon-perturbativeProton Structure

Parton Distribution Function (PDF):fraction of proton momentum carried by parton

Universal Process-dependent


Why Is Gluon PDF Important

• Photon x-section advertised for being useful to gluon PDF Previous photon x-section results are not used in current PDF fits.

More discussions later.

• Dominant Higgs production: gluon gluon fusion


Experimental Motivation• Advantages over jets

Ease of trigger requirements allows access to lower pT

Do not hadronize, no ambiguity due to jet definitions Better energy resolution Less uncertainty on energy scale (1-2% vs. 2-3%)

• Probe photon techniques over a wide energy range

• First CDF Run II measurement Focus on central photons (|| < 1.0) A new method to estimate backgrounds from jets

LUp

fN

ddp

d

T

sigtotal

T

2 fsig: fraction of signal photons

L: number of p pbar per unit x-sectionU: unfolding factor for efficiency, energy scale, energy smearing


LUp

fN

ddp

d

T

sigtotal

T

2

Major Backgrounds

p p

0

Mesons in Jets Decaying to Multiple Photons:0, 0,Ks

0BREM of Cosmic ray muon

Dominant at low pT Dominant at high pT

Signal MCData


Amount of Cosmic Background

MET/ET ()< 0.8 reduces the cosmic background to the level of < 1%


Cosmic Events• Use “photon arrival time at the EM calorimeter – collision

time” (EM Timing)

System intrinsic resolution ~ 0.6 ns


Central Electromagnetic Calorimeter

• EM Calorimeter E)/E =

0.135/ET+0.02

500-cm long ×=0.26 × 0.11

• CMS 0.017 × 0.017

18 X0Electromagnetic Calorimeter

Hadron Calorimeter

EM Cal. unit: 45.5 cm (width) × 22.7 cm (length) × 34.5 cm (depth)

Minimum separation of two photons from 0: 50/ET(0) [cm GeV], e.g. 5 cm for 10 GeV 0

Single EM shower size ~ 3.5 cm


Photon Isolation

Calorimeter isolation

Photons from jets have more surroundingtracks and energy (more isolation)

jet

Isolation energyEnergy of the photon

More isolation energy Photons less isolated

Isolation

background

signal

DATA


Fitting Isolation

DataFit resultSignalBackground

pT = 200-230 GeV

MIleakageconeiso EEEEE


Systematics Study (Signal Fraction)• Use Z electron data for signal template Electrons look like photons in the calorimeter

• Decrease the number of histogram bins to 2 Removing details on the shape

2-bin method2 unknowns2 equations


Fraction of Signal Photons

Signal fraction 70-100%Systematic uncertainty 13% first bin, 5% for the rest


LUp

fN

ddp

d

T

sigtotal

T

2

Unfolding Factor

• Correct detector level back to parton level energy scale, energy

resolution, acceptance, efficiency

• Systematic uncertainties Photon energy scale (6-

13%) ID efficiencies (3-5%)


How Well Do We Know the Energy?

• Photon energy affect the shape of differential x-section• Compare reconstructed Zee mass peak in data and MC

Apply time-dependent energy calibration

• Assign 1.5% systematic uncertainty Fitting model, energy dependence of scale

DatarecTE

trueTE

dataMC ?

Data

Diff

eren

tial X

-sec

tion

ET()


Systematic Uncertainty (X-section)

• Major sources Signal fractions at

low energy Photon energy scale

at high energy

• Total systematic uncertainty 12-15%


Cross Section Result

• X-sections measured over 6 orders of magnitude • Cover the energy range from 30 to 400 GeV • Excess in data below 50 GeV?


Why Are Photon X-sections Not Used in PDF?

• Data have steeper dependence on xT

• Due to mis-modeling of soft gluon emission in the initial state Additional transverse

momentum smearing of initial parton (kT) resolved the data/theory difference

xT=2pT/xT=2pT/s


Remarks

• Effects other than kT at low pT?

• kT is ad-hoc, better theoretical calculation needed Resummation of soft gluon emission Need to understand the difference between Run II CDF and D0

results 10% difference in normalization

• May provide input to gluon PDF for photons pT > 50 GeV?

• QCD Compton process dominant for all pT at LHC

• Low pt region may be useful when the theoretical calculation improves in the future



Puzzles



Measurements of SM



A Miracle!

Studies of

New Phenomena




Search for New Physics with Photons


Two ApproachesModel-dependant search Model New Particles Signatures Supersymmetry Bosonic Higgs 2 photons with a W or Z boson

Optimize selections based on model Report limits on theory parameters (e.g. MH > 106 GeV)

Signature-based search Form 2-object combinations out of the 7 magnificent objects, Start from these 27 base signatures and look for additional objects + X, MET + X

Aim to maximize the chance of discovery Apply a set of various requirements Report event counts and kinematic distributions Signature chosen based on past experience, detector strength, etc.


The Magnificent Seven Signatures in the Detector

e jet Beam Axis

Tracking system

EM Layers

Hadronic Layers

Calorimeter

Muon Chamber

b-jet


CDF Photon Signature-based Search

• + X, X = e,, , MET

• + e or + X, X = b-jet, MET, e, ,

• MET + b-jet + jet

• Displaced + jet

• MET+ n jets

Analyses I worked on

eW H


Search for New Physics in e/

(e)

NP

Data consistent with background prediction. No evidence of new physics, yet.


Search for New Physics in bjMET

Work with R. Culbertson (Fermilab), H. Frisch, D. Krop, C. Pilcher, S. Wilbur (U. Chicago)

NP

j

bM

ET


Major Backgrounds• Combination of real and fake objects

CES/CPR method

for fake photons from jets

Copper Stripson PC board

Au-plated W Wires

Copper Stripson PC board

Au-plated W Wires

Central Shower Profile Detector (CES) Central Preshower Detector (CPR)

2 mm resolution

EM Calorimeter

CES

CPR

Solenoid


CES Method

• A method to estimate how likely a photon candidate is a jet

• Jet has wider profile

Jet

0

Compare measured with expected and form a 2


CPR Method

Jet

0

Jet

0

Jet

0

Jet

0

ee

9

7

1NM

e

CPR

electron CPR efficiency ~ 100%

ee CPR

• A method to estimate how likely a photon candidate is a jet

• Jets have more photons to covert into electrons

CP R


Statistical Technique• CES: data = 1 if 2 < 4; otherwise data = 0

• CPR: data = 1 if there is a CPR hit; otherwise data = 0

• sig and bkg are calibrated

E() [GeV]


Systematic Uncertainties (CES/CPR)

• Uncertainty 10-15%• Dominant source

CPR: hits contributed by backscattered low-energy particle (albedo)

Preshower Detector

EM Calorimeter

Albedos


Kinematic Distributions (bjMET)

HT


Kinematic Distributions (bjMET)


Event Counts

Apply tighter cuts. Then, compare data and background estimate.

Data consistent with background prediction. No evidence of new physics, yet.

Observed 617 events in data


Model Phase• After a large set of signatures are looked for, perform a

global constraint on models

Parameter 1

Pa

ram

ete

r 2

e

bjMET

eeee

eejj


Model-dependant vs. Signature-based• Less sensitive to model A

Maximize the phase space probed

• Background estimate for several sets of cuts Advance techniques

• Report event counts and kinematic distributions Results will not be obsolete

• May miss very peculiar signatures or distributions Guidance from theorists

• Optimized for model A Precise limits

• Background estimate for the optimized cuts Good training ground for

students

• Report limits on new theory parameters Immediately useful for

theorists Easy for comparison

between experiments


Outlook (Tevatron)• Searches with more data

More sensitive to new physics

• Apply isolation and CES/CPR methods to other cross-section measurements and searches

• Explore forward photons Increase di-photon data by at least a factor of 3 Probe parton in a wider x range ? x = M/s exp(±y)

• Advantages over Large Hadron Collider (LHC) experiments Less tracking material (2-5 times smaller), good for photons Access to b-jets (before LHC silicon vertex detectors are fully

commissioned) Access to peculiar signatures

LBS


Outlook (LHC)• Expect to have 1st collision

by the end of 2009 at 10 TeV

• Advantages over Tevatron 5-7 times s 10-100 times instantaneous

luminosity Better detector

• 5x better EM energy resolution

• 10x better momentum resolution

• More coverage• More uniform• Higher granularity


Conclusions

• Presented physics results with photons Test of Standard Model with cross-section

measurement Systematic signature-based search for new physics

with photons

• Both Tevatron and Large Hadron Collider have excellent prospects of discoveries. Will be lots of fun!



Puzzles



Measurements of SM



A Miracle!

Studies of

New Phenomena



Theory

Experiment

Documents

Shin-Shan YuPhysics with Photons at CDF 1 從未知世界來的光 : CDF 實驗的光子物理 清華大學高能物理演講 西元 2009 年 2 月 26 日 余欣珊 費米實驗室

Shin-Shan YuPhysics with Photons at CDF 1 從未知世界來的光 : CDF 實驗的光子物理清華大學高能物理演講西元 2009 年 2 月 26 日余欣珊費米實驗室