Irakli Garishvili, Univ. of Tennessee Christine’s burrito conference

1. A measurement of open heavy flavor production at forward angles in Cu+Cu collisions at√sNN=200 GeV • Physics of heavy ion collisions • Heavy flavor production • Single muon measurement • Results NEW! Irakli Garishvili, Univ. of Tennessee Christine’s burrito conference

2. QCD matter confinement de-confinement: Quark Gluon Plasma • early predictions:εC ~ 1 GeV/fm3 (TC ~ 170 MeV) • lattice QCD: εC = 700±200 MeV/fm3

3. Relativistic heavy ion collider (RHIC) PHOBOS BRAHMS PHENIX STAR • 3.83 km circumference • Two independent rings • 120 bunches/ring • 106 ns crossing time • Capable of colliding ~any nuclear species on ~any other species • Energy: - 500 GeV for p-p - 200 GeV for Au-Au(per N-N collision) • Luminosity • Au-Au: 2 x 1026 cm-2 s-1 • p-p : 2 x 1032 cm-2 s-1(polarized)

4. Heavy ion collisions Bjorken energy density πR2 τ0 ~ 0.3 -1.0 fm/c In head on Au+Au collisions at RHIC maximum energy density achieved is at least 15 GeV/fm3!

5. Classifying collisions spectators participants

6. Probing created medium heavy-ion collision p+p collisions • in p+p collisions particle production occurs • in the vacuum • in A+A collisions initially produced particles • encounter dense medium of color charges Nuclear modification factor What is measured in A+A Naïve scaling of QCD RAA = 1 → boring RAA < 1 → suppression RAA > 1 → enhancement

7. Why measure open heavy flavor ? • In Heavy ion collisions • Created at the early stages of the heavy ion collisions, • heavy quarks are considered as important probes of • the created medium ( ) • Precisely measuring open charm production is • important for a better understanding of the • production of J/ψparticles • In p+p collisions • test pQCD predictions (MHQ >> ΛQCD) • baseline measurements for heavy ion collisions τHQ ~ 1/(2MHQ) < 0.1 fm/c

8. Heavy flavor production in p+p collisions ℓ- _ _ D0 K+ ν Hadronization Decay into the Final products Initial HQ production

9. Heavy flavor production in A+A collisions  e,  g c s, d medium g Energy loss, ΔE=Ei-Ef, depends on the medium properties

10. RHIC measurements

11. Tagging heavy quarks ℓ- _ _ D0 D0 K+ K- - ν c c Semi-leptonic decay • Non-photonic leptons: • single electron spectra • single muon spectra Hadronic decay π+

12. The PHENIX detector PHENIX

13. Sources of tracks in muons arms MuID Gaps 0 1 2 3 4 SIGNAL: “Prompt” muons – muons resulting from decays of heavy quarks BACKGROUND: 1.) Decay muons – muons from hadron decays 2.) Punchthrough hadrons – hadrons punching through the entire detector OTHER SOURCES Stopped hadrons – hadrons stopping in the shallow gaps (2 & 3) due to nuclear interaction with the absorber. Gap 0 1 2 3 4 Nosecone absorbers

14. Signal extraction methodology

15. Estimating background with the “Cocktail” • 1. Generation of a “realistic” input PT spectra for mixture of dominant background sources: • 2. Particle propagation through PHENIX geometry • using GEANT: • Hadron shower code selection (FLUKA,GHEISHA) • Nuclear interaction cross section of absorber • 3. “Embedding” simulated tracks into real events to reproduce effects of detector environment during heavy ion collisions

16. Data handles on background Stopped hadrons Using momentum cut, selecting a relatively clean sample of unidentified hadrons from the tracks stopped in the gaps 2 & 3 of the MuID detector. hadrons Slope is proportional to decay muon yield! Decay muons μ± π± π± 40 cm z

17. Adjusting/tuning cocktail to match data GAP 0 1 2 3 4 Normalize and “tune” the cocktail input spectra to match the data in gap 3. Checking MC/data matching at gap 2. The difference is enfolded into systematic errors MC/data

18. Adjusting/tuning cocktail to match data GAP 0 1 2 3 4 Checking MC/data χ2/ndf matching of Z-slopes at gap 4.

19. Inspecting matching between MC and data NORTH SOUTH Gap-2 Gap-3 Slope χ2/ndf Gap-4

20. Determining full background prediction • • • • • • • Weighted average • Entries Cocktail Yield Very conserva- tive bracket • • • Individual package redictions (for a given PT bin) • Background cocktail prediction [invariant yield]

21. Extracting single muon spectra “Raw” spectra: Corrected single muon spectra:

22. Combining two arms

23. Single Muon Spectra 0-20% Cu+Cu 20-40% Cu+Cu 40-94% Cu+Cu p+p collisions

24. Nuclear modification factor 40-94% Cu+Cu 20-40% Cu+Cu 0-20% Cu+Cu

25. Muons (Cu+Cu) vs. electrons (Au+Au)

26. Trying to understand apples vs. oranges Central vs. Central dN/dy vs. y In the collision of the same species Bjorken energy density is expected to be 20-25% different between y=0 and y=1.65 rapidities

27. Forward rapidity Single muons have been measured in both p+p and A+A type collisions. For further more qualitative conclusions need to reduce systematic errors Also need a better control measurement and at least some theoretical predictions for open heavy flavor production at forward rapidity for A+A collisions PHENIX J/Ψdata indicates larger suppression in the forward rapidity so open heavy flavor measurement could provide valuable baseline information.

28. Summary • First measurement of nuclear modification factor of open heavy flavor production in any A+A type collisions • First PHENIX measurement of open heavy flavor production • in Cu+Cu collisions • Significant suppression in the most central Cu+Cu • collisions • Data suggest that the medium effects could be at least as large, or larger, as those in mid-rapidity

29. Outlook • Final global Cu+Cu and p+p analysis underway for the • publication • Reduced systematic errors compared to the preliminary data • expected • PHENIX upgrades (VTX+FVTX) in 2011 • Significant background reduction • Separate D,B-meson measurements Very exciting years to come for heavy flavor community!

30. Back up slides

31. QCD matter: The full picture

32. Radiative energy loss u Q light l in medium, dead cone implies lower energy loss Non-photonic e K+ Hot/Dense Medium c quark e-/- c D0 - transport coeficient radiative energy loss “Dead cone” effect for heavier quarks: in vacuum, gluon radiation suppressed at θ < mQ/EQ

33. enhancement Dissertation Defense Final Collaboration Approval Preliminary Colla-boration Approval saturation Analysis Completion Complete Cu+Cu Dataset available Proposal Defense Cu+Cu data collection First Quarter of 2005 Today End of this Summer December 2006 May 2007 End of Sum- mer of 2007 May 2008

34. Individual package “tuning” scheme • Selecting normalization factor , pretty much by eyeballing data and MC • spectra so that: • 2. Fix MC normalization. As a 0th order iteration applying weights on input pT spectra • for each arm independently thus modifying MC hadron yields in gap3. • Calculating “new” weights . If new weights are not equal to 1. • We repeat step-2 iteratively unless we get perfect matching.

35. Estimating combinatoric background There is irreducible background underneath peak below 0.2. Tracks with high δZ are correlated with track above 0.2 value that are known background source coming from momentum miss-reconstruction due to local occupancy in MuTR. Scaled background shape with tail from 0.2 to 0.3 This procedure is separately applied to MC and data for each pT bin

36. e± pT rather poorly correlated with D pT 130 GeV Au+Au (0-10%) D from PYTHIA D from Hydro B from PYTHIA B from Hydro e from PYTHIA e from Hydro (plot courtesy of M. van Leeuwen) (S.Kelly @ QM`04) Experimental complication... PYTHIA simulation (RHIC energy)

38. Elliptic flow Weak interaction λ→∞ Strong interaction λ→ 0 y f x 2v2 dN/df dN/df f 0 2p f 0 2p

39. Elliptic flow as seen by RHIC • All reconstructed hadrons • exhibit large elliptic flow • Below pT ~ 2 GeV/c the data is • in a good agreement with the • Hydro model prediction • elliptic flow per number of • the constituent quarks follows • universal scaling!

40. Alternative energy loss mechanisms u u K+ c l l D0 K+ Hot/Dense Medium c quark e-/- c D0 D0 meson energy loss Adil & Vitev Hot/Dense Medium e-/- c quark collision energy loss

41. Heavy flavor and elliptic flow PRELIMINARY minimum-bias Rapp & van Hees, PRC 71, 034907 (2005) Run-7 Run-4 Initial preconception was that heavy quarks “too heavy” to participate in the collective behavior. PHENIX measured a large v2 for non-photonic electrons. Suggests that at least charm is partially thermalized. What about bottom? Nobody believes that it can flow. But again who knows?