1. Dileptons and Photons at RHIC Energies
Introduction
Experimental Results
We know our reference … p+p data
Low mass dilepton enhancement in Au+Au
Direct virtual photons in Au+Au
Comparison to Models
Outlook
Summary
C4-1 Y. Akiba: “Dilepton Radiation in PHENIX”
C4-3 Y. Yamaguchi: “Photons from PHENIX”

2. Lepton-Pair Continuum Physics
Modifications due to QCD phase transition
Chiral symmetry restoration
continuum enhancement
modification of vector mesons
thermal radiation
& modified heavy flavor
suppression
(enhancement)
known sources of lepton pairs at s = 200 GeV
Sources “long” after collision:
p0, h, w Dalitz decays
(r), w, f, J/y, y‘ decays
Early in collision (hard probes):
Heavy flavor production
Drell Yan, direct radiation
Baseline from p-p
Thermal (blackbody) radiation
in dileptons and photons
temperature evolution
Medium modifications of meson
pp  r  l+l-
chiral symmetry restoration
Medium effects on hard probes
Heavy flavor energy loss
Large discovery potential also RHIC
Axel Drees

3. Key Challenge for PHENIX: Pair Background
No background rejection  Signal/Background  1/100 in Au-Au
Combinatorial background: e+ and e- from different uncorrelated source
Need event mixing because of acceptance differences for e+ and e-
Use like sign pairs to check event mixing
Unphysical correlated background
Track overlaps in detectors
Not reproducible by mixed events: removed from event sample (pair cut)
Correlated background: e+ and e- from same source but not “signal”
“Cross” pairs  “jet” pairs
Use Monte Carlo simulation and like sign data to estimate and subtract background π0 e+ e- γ X
Subtractions dominate systematic uncertainties
But are well under control experimentally!
Axel Drees

4. Estimate of Expected Sources
Hadron decays:
Fit p0 and p± data p+p or Au+Au
For other mesons h, w, r, f, J/y etc. replace pT  mT and fit normalization to existing data where available
Heavy flavor production:
sc= Ncoll x 567±57±193mb from single
electron measurement
Hadron data follows “mT scaling”
Poster by S. Milov
Predict cocktail of known pair sources
Axel Drees

5. Dilepton Continuum in p+p Collisions
Phys. Lett. B 670, 313 (2009)
Data and Cocktail of known sources represent pairs with e+ and e- PHENIX acceptance
Data are efficiency corrected
Excellent agreement of data and hadron decay contributions
with 30% systematic uncertainties
Axel Drees 5

6. Charm and Bottom Contribution
Subtract hadron decay contribution and fit difference:
Phys. Lett. B 670, 313 (2009)
Consistent with PHENIX single electron measurement
sc= 567±57±193mb
Charm: integration after cocktail subtraction
sc=544 ± 39 (stat) ± 142 (sys) ± 200 (model) mb
Simultaneous fit of charm and bottom:
sc=518 ± 47 (stat) ± 135 (sys) ± 190 (model) mb
sb= 3.9 ± 2.4 (stat) +3/-2 (sys) mb
Axel Drees 6

7. Contribution from Direct (pQCD) Radiation
Measuring direct photons via virtual photons:
any process that radiates g will also radiate g*
for m<<pT g* is “almost real”
extrapolate g*  e+e- yield to m = 0  direct g yield
m > mp removes 90% of hadron decay background
S/B improves by factor 10: 10% direct g  100% direct g*
arXiv:
1 < pT < 2 GeV
2 < pT < 3 GeV
3 < pT < 4 GeV
4 < pT < 5 GeV
pQCD q g
hadron decay cocktail
Small excess at for m<< pT consistent with pQCD direct photons
Axel Drees

8. Au+Au Dilepton Continuum
Excess 150 <mee<750 MeV:
3.4 ± 0.2(stat.) ± 1.3(syst.) ± 0.7(model)
hadron decay cocktail tuned to AuAu
Charm from PYTHIA filtered by acceptance
sc= Ncoll x 567±57±193mb
Charm “thermalized”
filtered by acceptance
sc= Ncoll x 567±57±193mb
Intermediate-mass continuum: consistent with PYTHIA
if charm is modified room for thermal radiation
Axel Drees 8

9. Centrality Dependence of Low Mass Continuum
Excess region: 150 < m < 750 MeV
Yield / (Npart/2) in two mass windows
p0 region: production scales approximately with Npart
Excess region: expect contribution from hot matter
in-medium production from pp or qq annihilation
yield should scale faster than Npart (and it does)
Excess mostly in central AuAu yield increase faster than Npart
p0 region: m < 100 MeV
Axel Drees

10. pT Dependence of Low Mass Enhancement
arXiv:
arXiv:
Au+Au
p+p
0<pT<8.0 GeV/c
0<pT<0.7 GeV/c
0.7<pT<1.5 GeV/c
1.5<pT<8 GeV/c
Low mass excess in Au-Au concentrated at low pT!
Axel Drees

11. Mass Dependent Dilepton pT Spectra
p+p
Au+Au
detector g* e+ e-
B-field
Case A: g* and e+e- in acceptance
Case B: g* in acceptance
e+ and/or e- NOT in acceptance
Apply acceptance correction
correction depends on source
virtual photon polarization
g* different from charm
additional uncertainties!!
p+p consistent with cocktail up to 3 GeV/c
Above mp Au+Au data enhanced for all pT most prominent for pT < 1 GeV/c
Axel Drees

12. Local Slopes of Inclusive mT Spectra
Tdata ~ 240 MeV
Tcocktail ~ 280 MeV
Tdata ~ 120 MeV
Tcocktal ~ 200 MeV
Data have soft mT component not expected from hadron decays
Note: Local slope of all sources!
need more detailed analysis
isolate excess (subtract cocktail)
statistics limited
However: low mT result will not change much!
Soft component below mT ~ 500 MeV:
Teff < 120MeV independent of mass
more than 50% of yield
Axel Drees

13. Dilepton Excess at High pT – Small Mass
arXiv:
arXiv:
Au+Au
p+p
1 < pT < 2 GeV
2 < pT < 3 GeV
3 < pT < 4 GeV
4 < pT < 5 GeV
0<pT<8.0 GeV/c
0<pT<0.7 GeV/c
hadron decay cocktail
0.7<pT<1.5 GeV/c
1.5<pT<8 GeV/c
Significant direct photon excess beyond pQCD in Au+Au
Axel Drees

14. Interpretation as Direct Photon
Relation between real and virtual photons:
Virtual Photon excess
At small mass and high pT
Can be interpreted as
real photon excess
Extrapolate real g yield from dileptons:
Example for one pT range:
Excess*M (A.U).
no change in shape
can be extrapolated
to m=0
Axel Drees

15. First Measurement of Thermal Radiation at RHIC
pQCD
g* (e+e-)
 m=0 g
T ~ 220 MeV
Direct photons from real photons:
Measure inclusive photons
Subtract p0 and h decay photons at S/B < 1:10 for pT<3 GeV
Direct photons from virtual photons:
Measure e+e- pairs at mp < m << pT
Subtract h decays at S/B ~ 1:1
Extrapolate to mass 0
First thermal photon measurement:
Tini > 220 MeV > TC
Axel Drees

16. Comparison to Theoretical Models
A short reminder:
Models for contributions from hot medium (mostly pp from hadronic phase)
Vacuum spectral functions
Dropping mass scenarios
Broadening of spectral function
Broadening of spectral functions worked well at SPS energies (CERES and NA60)
pp annihilation with
medium modified r
works very
well at SPS energies!
Axel Drees

17. In Medium Mesons at RHIC???
Models calculations with broadening of spectral function:
Rapp & vanHees
Central collisions scaled to mb
+ PHENIX cocktail
Dusling & Zahed
Bratkovskaya & Cassing
broadening
broadening and dropping
Au-Au mb
with modified charm
pp annihilation with
medium modified r
insufficient to describe
RHIC data!
closer look at pt spectra
Axel Drees

18. Model Comparison in pT pp-annihilation with meson broadening How about
underestimates range 300 to 500 MeV
maybe does ok in range 500 to 750 MeV
Very different results for 810 to 990 MeV
Different treatment of thermal radiation
and hard contributions
How about
direct radiation?
Axel Drees

19. Thermal Photon Contribution
Theory calculation by Ralf Rapp
Vaccuum EM correlator
Hadronic Many Body theory
Dropping Mass Scenario
QGP (qq annihilation only
q+g  q+g* not included)
q+g  q+g* ?
y=0
pT ~1 GeV/c
Expect virtual photon yield from QGP as large as from Hadron Gas
Axel Drees
Taken from talk by Y.Akiba

20. Calculation of Thermal Photons
Reasonable agreement with data
factors of two to be worked on ..
Initial temperatures and times from theoretical model fits to data:
0.15 fm/c, 590 MeV (d’Enterria et al.)
0.2 fm/c, MeV (Srivastava et al.)
0.5 fm/c, MeV (Alam et al.)
0.17 fm/c, MeV (Rasanen et al.)
0.33 fm/c, MeV (Turbide et al.)
Correlation between T and t0
Tini = 300 to 600 MeV
t0 = 0.15 to 0.5 fm/c
D.d’Enterria, D.Peressounko, Eur.Phys.J.C 46 (2006)
Axel Drees

21. A VERY Naïve Speculation:
Extrapolate direct photons to all mass and pT
Lowest order pQCD qg-Compton
Add to Rapp & vanHess in-medium contribution
ok for 300 to 500 MeV (close to where direct component was fitted)
For higher mass yield over estimated
Miss low mT contribution
pQCD
Maybe not that simple! Needs a theorist to calculate properly!
Axel Drees

22. Outlook into the Future (on Tape)
Cu+Cu data finalized soon
Cover low Npart range
High statistics pp data (4x)
Continuum between J/ and 
Even better reference including bottom
High statistics d+Au
Cold nuclear matter effects
500 GeV data with HBD!
Cu+Cu
Poster by S. Campbell
Poster by J. Kamin
d+Au raw data
Axel Drees

23. Future of the Dilepton Continuum at RHIC
Open experimental issues
Large combinatorial background prohibits precision measurements in low mass region!
Disentangle charm and thermal contribution in intermediate mass region!
signal electron
Cherenkov
blobs
partner positron
needed for rejection e+ e-
qpair
opening
angle
~ 1 m
HBD
Need tools to reject
photon conversions and Dalitz decays
and to identify open charm
PHENIX  hadron blind detector (HBD)
vertex tracking (VTX)
HBD is fully operational
Proof of principle in 2007
Taking data right now with p+p
Expect large Au+Au data set in 2010
Poster by M. Durham
I’m looking forward to competition from STAR once TOF is completed!
Axel Drees

24. Summary Dilepton Data from PHENIX
background subtraction well controlled experimentally
well established p+p reference
discovered a low mass enhancement in central Au+Au
mostly in central collisions
mostly at low mT component with T~ 120 MeV independent of mass
present a first measurement of thermal photons
indicate initial temperature > 220 MeV
Theoretical models: much work is needed!
pp annihilation with collision broadening insufficient to explain data
much of the radiation does not come from the hadronic phase!
more complete calculation of QGP contribution needed
maybe in sQGP perturbative approach not ideal!
thermal radiation from QGP consistent with direct photon data
initial temperature 300 to 600 MeV
uncomfortably large differences between calculations!
Outlook: much more data to come
hopefully not only from PHENIX
Axel Drees

25. Backup Slides
Axel Drees

26. Search for Thermal Photons via Real Photons
The internal conversion
method should also work at LHC!
internal conversions
PHENIX has developed different methods:
Subtraction or tagging of photons detected by calorimeter
Tagging photons detected by conversions, i.e. e+e- pairs
Results consistent with internal conversion method
Axel Drees

27. Combinatorial Background: Like Sign Pairs
Shape from mixed events
Excellent agreements for like sign pairs
also with centrality and pT
Normalization of mixed pairs
Small correlated background at low masses from double conversion or Dalitz+conversion
normalize B++ and B- - to N++ and N- - for m > 0.7 GeV
Normalize mixed + - pairs to
Subtract correlated BG
Systematic uncertainties
statistics of N++ and N--: 0.12 %
different pair cuts in like and unlike sign: 0.2 %
--- Foreground: same evt N++
--- Background: mixed evt B++
Au-Au
Normalization of mixed events:
systematic uncertainty = 0.25%
Axel Drees

28. Au-Au Raw Unlike-Sign Mass Spectrum
arXiv:
Run with added
Photon converter
2.5 x background
Excellent agreement within errors!
Unlike sign pairs data
Mixed unlike sign pairs normalized to:
Systematic errors from background subtraction:
ssignal/signal = sBG/BG * BG/signal
 up to 50% near 500 MeV
0.25%
as large as 200!!
Axel Drees

29. p-p Raw Data: Correlated Background
Cross pairs
Simulate cross pairs
with decay generator
Normalize to like sign
data for small mass
Like Sign Data
Correlated
Signal = Data-Mix
Jet pairs
Simulate with PYTHIA
Normalize to like sign data
Mixed
events
Unlike sign pairs
same simulations
normalization from
like sign pairs
Unlike Sign Data
Alternative methode
Correct like sign correlated background with mixed pairs
Signal: S/B  1
Axel Drees

30. Centrality Dependence of Background Subtraction
Compare like sign data and mixed
background
Evaluation in 0.2 to 1 GeV range
For all centrality bins
mixed event background
and like sign data agree within
quoted systematic errors!!
Similar results for background
evaluation as function pT
Axel Drees

31. Background Description of Function of pT
Good agreement
Axel Drees