Proton and Antiproton Production in High Energy Heavy Ion Collisions at RHIC (RHICでの高エネルギー重イオン衝突における陽子反陽子生成） 金野正裕 （数理物質科学研究科） 予備審査 (9/27/2007)

Outline 1. Introduction 2. Motivation - Baryon/Meson difference - Hadron production in heavy ion collisions 3. Methods - PHENIX detector - Charged hadron measurement & PID 4. Results & Discussions - Freeze-out properties - (Anti-)Proton production at intermediate pT 5. Conclusion

Introduction - QGP Quarks can exist as an apparent degree of freedom? => Quark Gluon Plasma (QGP) - Matter under high temperature and high energy density. - Quarks and gluons are freely moving in a large volume. Relativistic heavy ion collisions is a method to approach the QGP. QCD transition : Hadron gas <=> QGP Tc ~ 175 MeV εc ~ 0.6 GeV/fm3 QCD phase diagram JHEP 04 (2004) 050 (predicted by lattice QCD calculations) PLB 478 (2000) 447

Relativistic Heavy Ion Collider (RHIC) RHIC located at Brookhaven National Lab (USA) A first collider for heavy ion beam 2 circulating rings (circumference: 3.83 km) Colliding nuclei: Au+Au, Cu+Cu, d+Au, p+p Top energy (Au+Au): sNN = 200 GeV Peak luminosity: ~ 3x1027 cm-2s-1 Experiments: PHENIX, STAR, BRAHMS, PHOBOS

Picture of Relativistic Heavy Ion Collisions Freeze-out Hadronic scatterings Hadron gas QGP Pre equilibrium Incoming nuclei Hadronization Expansion & Cooling Thermalization Hard scattering Initial collision Space-time evolution of a heavy-ion collision (time scale: ~10 fm/c) There are some stages and dynamic changes.

RHIC Findings (1) A bulk system is expanding Rapidity distributions of charged hadron multiplicity (PRL 91, 052303 (2003)) Longitudinal expansion (parallel to beam axis) is dominant. Longitudinal boost invariance partly holds at mid-rapidity. Hydrodynamic calculations reproduce elliptic flow behavior at low pT. Small viscosity (/s) estimated => Nearly perfect fluid A bulk system is expanding longitudinally & transversely. Phys. Rev. C 67 (03) 044903, Phys.Rev.Lett.91 182301 (2003)

RHIC Findings (2) Particles & Medium Effects Baryon enhanced 0.5 1.0 Baryon enhanced B/M Splitting of v2 In central Au+Au collisions, hadrons are suppressed at high pT. - The suppression is a final state effect (parton energy loss). Suppression/Enhancement has particle-type dependence. => Baryon/Meson difference in yields and emission patterns at intermediate pT (2-5 GeV/c).

Hadron Production in RHI Collisions Hadronization Interactions in the medium Low-pT (soft) Thermal emission Quark recombination Thermalization Collective flow High-pT (hard) Jet fragmentation Hard scattering Jet quenching Current understanding: - There are multiple hadronization mechanisms at intermediate pT. - The relative contributions and particle-type dependence are not yet fully understood.

Motivation Outstanding questions: What we should do: - Understanding Baryon/Meson difference at intermediate pT. => What is the origin? - What pT does hydrodynamic contribution exist up to? - Quark recombination process is really necessary? - Can we separate hadron radial flow and quark radial flow ? What we should do: - Measurement of Proton and Antiproton pT Spectra - Sensitive to collective flow due to its relatively large mass. - Indicator of baryon number transport at lower energies. => Enhance the high-pT PID capability with new detector. - Systematic Study - Au+Au/Cu+Cu/p+p collision systems at √sNN = 62.4/200 GeV (system size, energy dependence).

My Activities Construction and Installation of Aerogel Cherenkov detector in PHENIX for high-pT PID upgrade (2002~2004) - Participation in PHENIX experiment during data taking periods (Run3-Run7) - Staying at BNL for ~2 years Data analysis (2005~current): + Calibrations, Software developments + Proton spectra using Aerogel detector Presentations (QM05, HQ06, QM06, etc.) Papers, proceedings (NPA 774 (2006) 461, EPJC 49, 29 (2007)) (Preparing full papers)

PHENIX Global detectors: EM Calorimeter (PID) TOF (PID) (event characterization) - Central Arm Detectors (||<0.35) (magnetic spectrometer) Aerogel Cherenkov (PID) Pad Chambers (tracking) Drift Chamber (momentum meas.) Global detectors: Beam Beam Counter (trigger, centrality, t0, z-vertex, RP) (efficiency: A+A: ~90%, p+p: ~50%) Zero Degree Calorimeter (centrality)

Data Analysis Data sets: Au+Au at sNN = 62.4, 200 GeV (Run-4) Cu+Cu at sNN = 62.4, 200 GeV (Run-5) p+p at sNN = 200 GeV (Run-5) p+p at sNN = 62.4 GeV (Run-6) Analysis methods: (1) Event selection (z-vertex, centrality) (2) Tracking, Momentum determination (3) Track selection (4) Particle Identification (TOF, ACC) (5) MC corrections (acceptance, efficiency) => Invariant yield pT distributions (/K/p) at mid rapidity ||<0.35

Event Selection Participant-Spectator model - Minimum Bias Trigger (BBC coin.) - Centrality determination (BBC, ZDC) Participant-Spectator model Participant Spectator ZDC BBC

Comparison of Au+Au and Cu+Cu (Cu+Cu: b=0.0 fm, Au+Au: b=8.6 fm) <Npart> ~117 Npart (no. of nucleon participants), Ncoll (no. of N-N scatterings) are estimated by Glauber model. Even though Ncoll-Npart relation is almost same between Au+Au and Cu+Cu, the geometrical overlap shape is different. - Cu+Cu: good resolution at smaller Npart. <Npart> ~100

Track Reconstruction  y R x r  z Drift chamber provides 12 hits in (x,y) plane Giving the bending angle  (R=220cm) after passing in magnetic field Giving pT with field-integral value PC1 hits and collision z-vertex fix the polar angle  Momentum resolution: Find intersection points between the trajectory and outer detectors. Projected points are then matched to measured points. x y R  z r 

Track Selection Residual distribution between hit point DC tracks at PC3 Residual distribution between hit point and projection point. Centroid and width are parameterized as a function of pT (position ~8 mm at r = 5m). Require tracks to be within 2. Background subtraction for charged hadron measurement Matching residual distribution has a tail. Asymmetric shape comes from residual bend. Background is subtracted with shape of the distribution. MC study was done. Background sources (dominant at high pT): - e+, e- from  conversion in materials - Weak decays mostly K+, K-

Particle Identification Time of Flight (~120 ps), p(p) ID up to 4 GeV/c m2 distributions (3.5-4.0 GeV/c) + K+ p Clear proton line 2 4 6 8 pT [GeV/c] TOF ACC proton & antiproton ID Aerogel Cherenkov (n=1.011), p(p) ID up to 7 GeV/c Veto for proton ID

Monte Carlo Corrections Acceptance (TOF) Corrections for: - Geometrical acceptance - Decays in flight - Momentum resolution - Detector efficiency - Occupancy effect (tracking efficiency is reduced in high multiplicity environment.) Acceptance (ACC) * MC simulation based on Geant-3. Real data / MC matching: - Dead areas are removed - Detector stability is checked - Same cuts are applied to obtain efficiency Occupancy (TOF, ACC)

Feed-down from weak decays Proton and Lambda pT spectra Feed-down: Weak decays take place very close to the vertex. Since heavier particles take most of the decay momentum, these tracks are inseparable from tracks coming from the vertex of a collision. p+p 200 GeV TOF Evaluation of the fraction: - Effective lambda spectra measured including higher resonances (~33%) - Decays in PHENIX acceptance (MC) - Fraction in measured p(p) : ~15% Fraction of Feed-down ( from ’s)

Systematic Errors Evaluation: Systematic errors are evaluated by varying cut conditions in data analysis. Some parts can be canceled when taking particle ratios etc. Systematic errors (TOF) Systematic errors (ACC)

Baryon Enhancement

Proton and Antiproton pT spectra Au+Au sNN = 200 GeV Cu+Cu sNN = 200 GeV p+p sNN = 200 GeV NOTE: No weak decay feed-down correction applied. pT reach extended up to 6 GeV/c for p(p) with fine centrality bins. (1) Aerogel Cherenkov (2) High statistics

Baryon enhancement at sNN = 200 GeV p/ p/ (Anti-)proton enhancement observed/confirmed in 200 GeV Au+Au/Cu+Cu. Larger than expected from jet fragmentation (measured in pp, e+e-). Clear peak in central events than that in peripheral. p/ (p/) ratios turn over at 2~3 GeV/c ,and fall towards the ratio in p+p. - Indicating a transition from soft to hard at intermediate pT.

Baryon enhancement at sNN = 62.4 GeV p/ p/ (Anti-)proton enhancement observed/confirmed in 62.4 GeV Au+Au/Cu+Cu. Similar pT dependence as at 200 GeV. The lower energy data provides an important information on baryon production and transport at mid-rapidity.

p/ ratio vs. Npart1/3 Cu+Cu vs. Au+Au (200 GeV) Npart scaling of p/ (p/) at same √sNN. The ratios are controlled by the initial overlap size of colliding nuclei, even though overlap region has a different geometrical shape.

* No weak decay feed-down correction applied. Beam energy dependence of enhancement * No weak decay feed-down correction applied. - p/+ ratio : decreasing as a function of sNN. - p/- ratio : increasing as a function sNN. Antiproton is a good probe to study the baryon enhancement.

p/ ratio vs. (dET/d)1/3 No Npart scaling of p/ (p/) in Au+Au between 62.4 and 200 GeV. Transverse energy density dET/d scaling of p/ is favored. - dET/d is a connection key between different √sNN. Proton production at 62.4 GeV is partly from baryon number transport, not only proton-antiproton pair production.

Nuclear Modification Factor RAA Comparison with p+p spectra (reference) in binary collision scaling. Proton, antiproton are enhanced at 1.5 - 4 GeV/c for all centralities. - Suppression is seen for , K.

RAA factor vs. Npart NOTE: Systematic errors (~10%) for overall normalization not shown. - Proton is enhanced for all centralities, while /K are suppressed. - At peripheral, slight enhancement seen as seen in d+Au (Cronin effect). - Similar Npart dependence for Au+Au / Cu+Cu. => Npart scaling of RAA ?

Comparison of RAA in Au+Au/Cu+Cu Pion RAA (pT=2.25 GeV/c) Proton RAA (pT=2.25 GeV/c) RAA (Cu+Cu) > RAA (Au+Au) RAA (Cu+Cu) > RAA (Au+Au) - Geometrical shape : Au+Au more deformed - No. of N-N scatterings per N : narrow peak in Cu+Cu

Summary 1 - Baryon enhancement Proton and antiproton enhancement confirmed at intermediate pT (2-5 GeV/c) in Au+Au/Cu+Cu. A turnover of p/ ratio seen at pT = 2-3 GeV/c. In terms of binary collision scaling, protons and antiprotons are enhanced at pT = 1.5-4 GeV/c, while pions/kaons are suppressed. 62.4 GeV data: At lower energy 62.4 GeV, proton production seems to be more affected by baryon number transport process. => Antiproton is a good indicator of the baryon enhancement. Scaling properties between different systems: Npart scaling of p/ (p/) seen between Au+Au and Cu+Cu at the same energy (sNN= 200/62.4 GeV), even though the overlap region has a different shape. => System volume Npart is a control parameter. (Npart: corresponding to the initial volume of colliding nuclei) - Instead of Npart scaling, transverse energy density dET/d scaling of p/ is favored between different collision energies.

Freeze-out Properties Characterizing bulk properties: - Chemical Freeze-out - Kinetic Freeze-out

Particle Yield dN/dy at mid rapidity Au+Au/Cu+Cu/p+p (sNN = 200 GeV) Au+Au/Cu+Cu/p+p (sNN = 62.4 GeV) Particle yields are (roughly) scaled with Npart btw Au+Au and Cu+Cu. dN/dy(Cu+Cu) > dN/dy(Au+Au) at smaller Npart.

Statistical Model Fit Phys. Rev. C71 054901, 2005 nucl-th/0405068 Extracting chemical freeze-out properties with statistical model fit. Fitting particle ratios of dN/dy (/K/p) at y~0. Assuming local chemical equilibrium of light quarks (u,d,s), s=1. Partial feed-down correction taken into account. - Tch, q : relatively stable s, s : not determined with this set of ratios (/K/p). Strangeness info is short. Au+Au 62.4 GeV (0-5%) -/+: 0.84 +/- 0.04 K+/K-: 1.19 +/- 0.06 p/p: 0.48 +/- 0.03 K-/-: 0.17 +/- 0.02 p/-: 0.08 +/- 0.01 Au+Au 200 GeV (0-5%) -/+: 1.02 +/- 0.05 K+/K-: 1.09 +/- 0.05 p/p: 0.74 +/- 0.05 K-/-: 0.16 +/- 0.02 p/-: 0.08 +/- 0.01 data model data model 1.01 +/- 0.01 1.20 +/- 0.13 0.48 +/- 0.09 0.17 +/- 0.02 0.08 +/- 0.02 1.00 +/- 0.01 1.09 +/- 0.08 0.74 +/- 0.08 0.16 +/- 0.02 0.08 +/- 0.02 Tch: 167 +/- 10 MeV q: 24 +/- 3 MeV 2/ndf: 9.2/2 Tch: 157 +/- 8 MeV q: 9 +/- 1 MeV 2/ndf: 1.1/2

Chemical Freeze-out Temperature Tch ~160 MeV Au+Au/Cu+Cu/p+p (sNN = 200 GeV) Au+Au/Cu+Cu/p+p (sNN = 62.4 GeV) Tch ~160 MeV, flat centrality dependence. Npart scaling of Tch between Au+Au and Cu+Cu. Almost same Tch at √sNN = 62.4, 200 GeV.

Chemical Potential Au+Au/Cu+Cu/p+p (sNN = 200 GeV) Au+Au/Cu+Cu/p+p q (200 GeV) : ~8 MeV, independent of Npart q (62.4 GeV) : increasing with Npart => more baryon stopping at central

Mean Transverse Momentum Au+Au/Cu+Cu/p+p (sNN = 200 GeV) Au+Au/Cu+Cu/p+p (sNN = 62.4 GeV) Clear hadron mass dependence: larger <pT> for heavier particles. => Consistent with radial flow picture. <pT> increases with Npart. it is clearly seen for (anti)proton.

Blast-wave Model Fit 2 map Blast-wave model is a parameterization within a simple boost-invariant model with transverse collective flow. pT spectra reflecting thermal freeze-out temperature and transverse flow at final state. 2 map * Ref: PRC48(1993)2462 Tfo ~120 MeV, T ~0.7 (* Resonance decay feed-down correction not applied. Instead, tighter pT fitting range used. ; 0.6-1.2 GeV/c K; 0.4-1.4 GeV/c, p/pbar; 0.6-1.7 GeV/c) Spectra for heavier particles has a convex shape due to radial flow.

Transverse Flow Velocity Au+Au/Cu+Cu/p+p (sNN = 200 GeV) Au+Au/Cu+Cu/p+p (sNN = 62.4 GeV) <T>: increasing with Npart. - Npart scaling of <T> between Au+Au and Cu+Cu. Almost same <T> at √sNN = 62.4, 200 GeV.

Kinetic Freeze-out Temperature Tfo ~120 MeV Au+Au/Cu+Cu/p+p (sNN = 200 GeV) Au+Au/Cu+Cu/p+p (sNN = 62.4 GeV) Tfo: decreasing with Npart. Npart scaling of Tfo between Au+Au and Cu+Cu. Almost same Tfo at √sNN = 62.4, 200 GeV.

Summary 2 - Freeze-out properties Characterizing bulk properties: - Chemical freeze-out - Kinetic freeze-out => Hadron production at low pT : “Thermal emission + Radial flow” Scaling properties between different systems: - Chemical/kinetic freeze-out properties show similarities between different collision systems. - Npart scaling of freeze-out properties (Au+Au, Cu+Cu), even though the overlapped region has a different shape. => System volume Npart is a control parameter. - Similarity at sNN = 200 and 62.4 GeV. (Only baryon chemical potential shows the difference due to the difference of baryon number transport.)

Two-component Model

Two-component Model (Soft+Hard) Soft component : Thermal emission + Radial flow - Described by Blast-wave model - Npart scaling seen - Thermal distribution extrapolated up to high pT Hard component : Jet fragmentation + Jet suppression - Measured p+p spectra - Ncoll scaling - Constant suppression factor (power-law distribution & fractional energy loss)

Hard component in p+p and
Au+Au p+p sNN = 200 GeV Au+Au sNN = 200 GeV 200 GeV 62.4 GeV Hard component (in p+p) at high pT depends on s. In Au+Au, suppression effect should be taken into account.

Pion pT spectra + - Reproduce the measured pion pT spectra. Blue: data Red: data subtracted by soft Au+Au 200 GeV + - Soft Line Hard Line Reproduce the measured pion pT spectra.

Proton pT spectra p p Reproduce the measured proton pT spectra. Blue: data Red: data subtracted by soft Au+Au 200 GeV p p Soft Line Hard Line Reproduce the measured proton pT spectra.

Fraction of soft and hard components + - p p - Both soft and hard components are necessary to reproduce the hadron spectra at intermediate pT (2-5 GeV/c). - Soft component is extended to higher pT in central. - Intermediate pT: Hard pions vs. Soft protons - Cross point (S=H) vs. pT -

Soft/Hard Separation in p/ - Radial flow can be the origin of the baryon enhancement (pT and centrality dependences). It’s significant. - Hard component is consistent with p+p result, PYTHIA calculation.

Soft/Hard Separation in p/p Data Hard - Hard component is consistent with p+p result, PYTHIA calculation => Universal fragmentation function.

pQCD - q/g jet contribution p(p) in p+p (PYTHIA) p(p) in p+p (NLO pQCD) g jet dominant q jet dominant - Is there is a difference of jet quenching effect between gluon jets and quark jets? - Larger energy loss of gluons is expected than that of quarks.

pQCD - p/ ratio p/+ in p+p (PYTHIA) p/- in p+p (PYTHIA) particle ratios at high pT provide a sensitivity to the difference between quark and gluon fragmentation. p(p) are enhanced in gluon jet than in quark jet, but it’s not large.

pQCD - p/p ratio p/p in p+p (PYTHIA) p/p in Au+Au (Data) - p enhanced in gluon jet than in quark jet in p+p. - In Au+Au, pQCD-based calculation shows a significant effect from energy loss on p/p ratio due to larger energy loss of gluons. - Independent of pT/centrality/system up to 6 GeV/c (0.7 +/- 0.1).

RCP factor vs. pT RCP Above 5 GeV/c, RCP shows similar suppression for pions and (anti)protons, though they have different sensitivities to quark and gluon jets.

Summary 3 - Two-component model - Reproduce the measured pT spectra for pions and protons with a consistent way. - Identify crossover region from soft to hard hadron production at intermediate pT (2-5 GeV/c). Baryon/Meson difference: - Intermediate pT: “Hard” pions vs. “Soft” protons - Origin of baryon enhancement is radial flow. It pushes heavier particles to higher pT. Baryon/Meson difference is trivial? Jet fragmentation and quenching: - Indicating that hard-scattered partons (quarks and gluons) have similar energy loss when traversing the nuclear medium, and parton fragmentation function does not change.

Quark Flow vs. Hadron Flow

Quark recombination At intermediate pT, recombination of One of the hadronization mechanisms. Recombination of thermal quarks in local phase space: qq  Meson, qqq  Baryon At intermediate pT, recombination > fragmentation because quark distribution is thermal: ~exp(-mT/T). At high pT, fragmentation (power-law shape) would be dominant. Fries, R et al PRC 68 (2003) 044902 Greco, V et al PRL 90 (2003) 202302 Hwa, R et al PRC 70(2004) 024905 At intermediate pT, recombination of quarks may be a more efficient mechanism of hadron production than fragmentation.

Applicability of quark recombination model p/ vs. pT v2/n vs. KET/n - Baryon enhancement & quark number scaling of v2 explained by “Quark recombination” - v2 at quark level => Collective flow at quark level - In a simple recombination picture, radial flow cannot be distinguished between hadron and quark phases. => Can we separate hadron flow and quark flow ?

1+1D Adiabatic Expansion - Ideal gas: P=(1/3) Entropy conservation Longitudinal expansion & Transverse expansion - cooling curves - z x y tfo fixed at 10 fm/c at most central bj vs. Np T scaled with (bj)1/4 at t = 1 fm/c Cooling stopped at Tfo

Freeze-out Time & Temperature Freeze-out time vs. Np Freeze-out temperature vs. Np More central collisions freeze out later at lower temperature. Consistent with freeze-out condition: (t)=R(t) Even if quark phase is created before hadronization, hadronic scattering should be taken into account. As expected, Tfo is lower than Tch. Different centrality dependence. Tfo dropping is consistent with 1+1D adiabatic expansion. Tc ~ Tch => the observed chemical eq. not via hadronic scatterings.

Summary 4 - Quark flow & Hadron flow Quark recombination: - In a simple recombination picture, hadron and quark radial flow effects cannot be separated. Since the constituent quark number scaling of elliptic flow v2 is indicative, quark recombination process is thought to be a possible hadronization mechanism. Quark flow vs. Hadron flow: - We see the sum of quark and hadron flow. - The difference of chemical and kinetic freeze-out temperatures shows a finite expansion time at hadronic stage. => Hadron radial flow should exist even though quark flow exist before hadronization.

Conclusions Construction and installation Aerogel Cherenkov counters in PHENIX to enhance PID capability. - Systematic measurement of proton and antiproton pT spectra (Au+Au, Cu+Cu, p+p at sNN = 200/62.4 GeV) Proton and antiproton enhancement confirmed at intermediate pT (2-5 GeV/c). Antiproton is a good indicator for study of the baryon enhancement. - p/ ratio & freeze-out properties show Npart scaling between Au+Au and Cu+Cu at same sNN. The Initial volume (~Npart) of colliding nuclei is a control parameter. - Baryon enhancement is caused by transverse radial flow : - pT and centrality dependences are described by two-component model. - Intermediate pT (2-5 GeV/c): hard pions vs. soft protons - Chemical/Kinetic Freeze-out temperatures provide a hint for further expansion at hadronic stage. - Quarks and gluons have similar energy loss when traversing the nuclear medium, and parton fragmentation function does not change.