1. Measurement of +- ,e+e- in Ultra peripheral PbPb collisions at 5
Measurement of +- ,e+e- in Ultra peripheral PbPb collisions at 5.5 TeV in CMS
Status Report
Vineet Kumar
NPD BARC

2. Plan of presentation Introduction: UPC PbPb(γPb)¡Pb* in CMS
Brief introduction of Ultra peripheral collisions (UPC)
Physics motivation
UPC PbPb(γPb)¡Pb* in CMS
Theoretical cross section for signal ¡ (1s), ¡ (2s), ¡ (3s)
Theoretical cross section for dileptons continuum (γ γe+e-,µ+ µ-)
¡ reconstruction with trigger in CMS
Low Pt electron reconstruction using Particle Flow technique in CMS
Results
Future Plans

3. Ultra Peripheral Collisions
Ultra Relativistic Heavy ions produce very
high electromagnetic fields due to coherent
action of all protons
UPC are those reactions in which ions interact via their cloud of virtual photons.
An electromagnetic interaction where photons
emitted by ions interact with each other
(b) A photon-nuclear reaction in which a photon
emitted by an ion interacts with other nucleus.

4. Ultra Peripheral Collisions
Experimental Characteristics of UPCs
Low central multiplicities
“cleaner” than hadronic collisions
Zero net charge
Narrow dN/dy peaked at mid-rapidity

5. Physics Motivations in UPC
+AVM+A (VM=J/, ) sensitive to
gluon density squared
Test QED at very high energies ℓ Pb Pb Pb Pb
Unexplored (x,Q2) regime of nPDFs:
Kinematical range covered

6. Predicted cross sections
STARLIGHT Generator
[J.Nystrand,S.KleinNPA752(2005)470]
Signal
PbPb Pb+X
µ+ µ- ,e+e- 
BR %
No. for ʃ L 0.5(nb)-1
(1s)
78 µ b
2.4%
936
(2s)
33 µ b
1.9%
313
(3s)
23 µ b
2.2%
253
 (mass>6Gev)
No. for ʃ L 0.5(nb)-1
0.6 mb
3x105
1.4 mb
7x105
Dilepton continum
  µ+ µ-
   e+e-

7. Muon reconstruction with trigger
Simulated data from starlight
StarlightHepMcGenSim.cfgSim.root
Raw data is made using configuration file HLTrigger/Configuration/test/RelVal_Digi_Digi2Raw_cfg.py
Selected HLT Paths for muons
‘HLT_DoubleMu3’
‘HLT_Mu3’
Partial HLT path from confDB by*
edmConfigFromDB --configName /dev/CMSSW_2_1_X/HLT --format Python --paths HLT_ DoubleMu3> HLTDoubleMu3Config.py
HLTrigger/Configuration/test/RelVal_HLTFromRaw_2E30_cfg.py
Having all the trigger path of trigger table (CMSSW_2_1_X)

8. Trigger efficiency for HLT_DoubleMu3 and HLT_Mu3
  +-
TrigReport Event Summary
TrigReport Events total = passed = 4279 failed = 5721
TrigReport Path Summary
TrigReport Trig Bit# Run Passed Failed Error Name
TrigReport HLTMu3
HLTMu3 Eff  42 %
TrigReport Events total = passed = 1786 failed = 8214
TrigReport HLTDoubleMu3
HLTDoubleMu3 Eff 17%
  +-
TrigReport Event Summary TrigReport
Events total = passed = failed = 48874
TrigReport Path Summary TrigReport
Trig Bit# Run Passed Failed Error Name TrigReport
HLT_DoubleMu3
HLT_DoubleMu3  2%

9. Full reconstruction with HLT
Reconstruction Sequence
GenHlt + Select only those event which pass HLT Path +OfflineReco
HLT Path Used HLT_Mu3
+-
Total events
Triggerd
Reconstructed
Trigger eff %
Reco eff %
__Gen
__OfflineReco
__TriggeredEvents
__Triggered+Reco
+-
Total events
Triggerd
Reconstructed
Trigger eff %
Reco eff %
dielectron cont (Gev/C2 )
Without trigger Reco eff
+ % (4.2%)
+ % (37%)
M+- (Gev/C2)

10.  Family Pt and Y distributions
+-
+-
__triggered
__reco
__triggered
__reco
 rapidity
reco::muon

11.  and dimuon cont scaled for L=0.5 nb-1
Bkg subtracted  
PbPb UPC 5.5 TeV
Full CMS Sim+Reco
__ Pb(+-)
PbPb UPC 5.5 TeV
__ Pb(+-)
__+-
+
+- ’ ’
’’
’’
M+- (Gev/C2)
M+- (Gev/C2)
Upsilon+Dimuoncont Poll4+3Gauss
DimuonCont Poll4
(1s) GeV/C2
(2s) 9.99 GeV/C2
(3s) GeV/C2
 gauss ~ 95 MeV

12. Low Pt electron reconstruction using PF technique in CMS
I tried low Pt electron reconstruction using ParticleFlow Tools.
They claim an increase in electron reconstruction efficiency from 15% to ~ 80% at Pt=4GeV.
These tools are under development and not fully integrated in CMSSW.

13. PF in ||<1.0 gen 8082 --- PF 5101 63.1% gsf 2126 26.3% Electron 
This technique was tested in || < 1.0
I start with 10,000 e+ e- and reconstruct them using particle flow patches.
Technique seems to work efficiency goes up
__ Gen electrons
__PF electrons
__GSF electrons
Electron 
no (|| <1.0)
eff
gen
8082
--- PF
5101
63.1%
gsf
2126
26.3%
__ Gen electrons
__PF electrons
__GSF electrons
Electron Pt (GeV)

14. Electron reconstruction using PF
In full kinematical range.
__ Gen electrons
__PF electrons
__GSF electrons no
efficiency
gen
20,000
--- PF
7944
39.7%
gsf
4403
22.0%
Electron 
__ Gen electrons
__PF electrons
__GSF electrons
Clearly method works best for
|| <1.0
Electron Pt (GeV)

15. PFlow reconstruction of electrons
__GSF
__PF
__GSF
 Inv mass (Gev/C2 )
dielectron cont (Gev/C2 )
Dielctron cont reco eff
PF=1.49%
GSF=0.6%
Up(1s) reco eff
PF=19%
GSF=10%
Mass resolution is better for GSF electron collection
Increase is more prominent for dielctron continuum

16. e+e- continuum (QED Test)
2-photon interaction
Higher order diagrams are required to explain STAR data.
A. J. Baltz, Phys. Rev. Lett. 100, (2008).
Pair pT
Minv
S T A R
- - -Lowest order
__Higher order
- - -Lowest order
__Higher order

17.  and dielectron cont scaled for L=0.5 nb-1
+ dielectron continuum
PbPb UPC 5.5 TeV
Full CMS Sim+Reco
__e+e-
__e+e-
__e+e-__e+e-
inv masse+e- (Gev/C2 )
inv masse+e- (Gev/C2 )
Large cross section of dieletron continuum prevents
us to extract  from continuum.
e+e- continum can be used as signal (QED Test)
We purpose reconstruction of electron continuum to test QED at more energy.

18. Summary and future plans
In PbPb→( Pb)→  Pb* at 5.5 TeV with →µ+µ- whole  family can be reconstructed with good resolution.
This Study in CMS can be used as a tool to study low-x gluon density & evolution in the nucleus.
e+e- continuum can be reconstructed to extend STAR study.
CMS Note

19. Back up slides

20. (1s) or  bbbar Gev KeV
(2s) or ’ bbbar Gev KeV
(3s) or ’’ bbbar Gev KeV

21. Particle Flow :as a user’s point of view
How to run
Install recommended version of Particle Flow.
Then run the following cfg file,which will replay the tracking and the particle flow
cd RecoParticleFlow/Configuration/test
cmsRun fullSimForParticleFlow_cfg.py
Access to particle flow output
1.PAT run PF2PAT+PAT to get pattuples.
2.Access the PFCandidates directly from an ED Analyzer in full framework
3.Acsess the PFCandidates from ROOT+FWLite
cmsRun analyzePFCandidates_cfg.py

22. Particle Flow :as a user’s point of view
The aim of the particle flow is to provide a single list of reconstructed particles which can be of type:
Photon
Charged hadron
Neutral hadron
Electron
Muon
This list will provide a complete description of the event and is as easy to use as the list of true particles from the simulation.
The PF Algorithm
Calorimeter clustering (ECAL,HCAL,PS) PF Clusters
Track reconstruction and extrapolation to the calorimeters (iterative tracking)
Reconstructing blocks of topologically connected elements (tracks, ECAL clusters, HCAL clusters, PreShower clusters) PF Blocks
Analyzing these blocks to reconstruct particles PF Candidates
Particle Flow :as a user’s point of view

23. electron reco in Particle Flow :as a user’s point of view
Reconstruction of track up to ECAL entrance
Associating this track with the electron cluster
Identifying the clusters of emitted photons.
Electron pre identification:
Track cluster matching variables
Tracker only based variables
A multivariate approch is followd .the so called Boosted Decision Trees (BDT) method was chosen.
Journal of physics:Conference Series 120(2008)032039

24. e, measurement in CMS Tracking + ECAL + muon-chambers electrons muons
Tracker+EMCAL
muons
Tracker+muchambers

25.  +e+ e-  and Pt distributions
____ gen
_____rec
____ gen
_____rec
Single e eta
Single e pt
Electrons are peaking outside CMS acceptance
Almost all electrons are concentrated at very low Pt

26. Summary
UPC A+A collisions generate high-energy  beams for photo production studies:  +  ,  +A physics as done at LEP & HERA.
Unique access to nuclear xGA(x,Q2) at small-x [Gluon saturation, non-linear QCD]
Unexplored kinematics regime
Study of PbPb→( Pb)→  Pb* at 5.5 TeV with →µ+µ- ,e+e- in CMS can be used as a tool to study low-x gluon density & evolution in the nucleus

27. Ultra Peripheral Collisions at LHC
Weizsacker -Williams formula for flux radiated by a ion with charge Z at distance r
Here ω=k r/ γL and K0 (ω) and K1 (ω) are modified Bessel functions.
The photo production cross section can be factorized in to the product of photonuclear cross section and the photon flux
Very high photo production cross sections !!
σ(γA)~ Z2 (~104 for Pb), σ(γγ) ~ Z4 (i.e. ~5·107) times larger than e± beams
Characteristics of photon flux in UPC at LHC
Max γ energies :ω< ωmax ~ γ/R ~80 GeV (Coherence condition) Pb-Pb LHC
γA: max. √s γA ≈ 1. TeV ≈ × √s γp (HERA)
γ γ : max. √s γγ ≈ 160 GeV ≈ √s γγ(LEP)

28. Signal + Bkg cross sections
Input MC: STARLIGHT [J. Nystrand, S.Klein, NPA752(2005)470]
Signal
Background ℓ Pb Pb Pb Pb