Outline • introduction • strongly interacting matter • relativistic heavy-ion collisions • probing the hot and dense medium • electrons from heavy flavor at RHIC • reference: p+p collisions • cold nuclear matter: d+Au collisions • hot matter: Au+Au collisions • going beyond single electrons: correlations • summary & outlook
Nuclear matter as QCD laboratory • “ordinary” nuclear matter is made from nucleons • 3 (light) constituent quarks, carrying color charge • quarks interact via the exchange of gluons • gluons carry color charge (“charged photons”)! • key observations • isolated quarks are NEVER observed (“confinement“) • quark masses account for ~1% of the nucleon mass • properties of QCD (Quantum Chromo Dynamics), the theory of strong interaction • learn more → “extraordinary” nuclear matter
The QCD phase diagram • study fundamental properties of matter by • excitation to extreme temperature and/or density • phase transition from nuclear to “quark-gluon“ matter • unique approach: relativistic nuclear collisions • center-of-mass energy: where do you want to go today? • highest temperature at lowest baryon density colliders: RHIC @ BNL and LHC @ CERN • moderate temperature at highest baryon density fixed-target: FAIR @ GSI
STAR RHIC and its experiments • highest CMS energy currently available at • RHIC (Relativistic Heavy-Ion Collider) located at Brookhaven National Laboratory • p+p: √s ≤ 500 GeV (polarized beams!) • A+A: √sNN ≤ 200 GeV (per nucleon-nucleon pair) • experiments with specific focus • BRAHMS (until Run-6) • PHOBOS (until Run-5) • multi purpose experiments • PHENIX • STAR
STAR PHENIX The experimental challenge • ONE central Au+Au collision at max. energy • MANY secondary particles • how to look into the heart of matter?
medium g g A view behind the curtain • “tomography” with scattering experiments • Rutherford: a→ atom discovery of the nucleus • SLAC: electron → proton discovery of quarks • “tomography“ at RHIC • probe has to be “auto generated” in the collision • hard parton (quark, gluon) scattering, leading to • direct photons from quark-gluon Compton scattering • high pT jets • heavy quark-antiquark pairs • once calibrated for p+p collisions modifications observed in p(d)+A & A+A tell about the “medium” • calibration of hard probes • theoretically • perturbative QCD (pQCD) • experimentally • measurement in p+p • in-situ control: direct photons
Nbinary: number of “binary” collisions, determined from the collision geometry (Glauber) p-p Direct photons at √sNN = 200 GeV • photons from quark-gluon Compton scattering Au-Au Medium produced in Au+Au collisions is transparent for direct photons! • no strong final state interaction • direct photons are a calibrated probe
(Light) hadrons at √sNN = 200 GeV • pQCD in reasonable agreement with p+p data • medium modifications in cold (hot) matter: d+Au (Au+Au)? • nuclear modification factor: Quantitative assessment of medium parameters requires in-between (“grey”) probe Medium produced in Au+Au collisions is opaque for light quark and gluon jets! • limiting factor, preventing RAA to drop even further: surface (“Corona”) emission
D mesons , Y’, c Heavy quarks to the rescue? • heavy quarks (cc, bb): mu,d ~ MeV, mc~1.25 GeV, mb~4.5 GeV • hard process (mq >> LQCD) • production at leading order (LO) • mainly gluon fusion • naïve expectation • large mass → small energy loss • confirmed in (most) models • (quantitative) details depend on energy loss mechanism • example: energy loss via gluon radiation • larger parton mass implies less energy loss in forward direction (“dead cone” effect) (Dokshitzer, Kharzeev: PLB 519(2001)199) • partially compensated by medium induced gluon radiation (Armesto, Salgado, Wiedemann: PRD 69(2003)114003) • systematic experimental study • heavy quarks from p+p, p(d)+A, and A+A collisions • in addition: bound states, quarkonia (J/y, U)
Heavy quark data pre-RHIC • total cross section measurements at lower √s • recent review:C. Lourenco & H. Woehri: Phys. Rep. 433 (2006) 127 • charm and bottom cross sections measured at the same √s can differ by more than a factor 10! • differential cross section at higher √s (1.8 TeV) (PRL 91, 241804 (2003)) • CDF: PRL 91, 241804 (2003) • reconstruction of charmed mesons for pT > 5 GeV/c only!
K+ p- Charm measurements at RHIC • ideal (but very challenging in HI environment) • direct reconstruction of charm decays(e.g. ) • STAR (for pT < 3 GeV/c) • helpful to constrain charm cross section D0 K+p- PRL 94, 062301 (2005) • alternative (but indirect, and still challenging) • contribution of semileptonic decays to lepton spectra • PHENIX & STAR • only systematic study: electron spectra at y~0
charged pions: (p+ + p-)/2 neutral pions: p0 (e+ + e-)/2 from heavy flavor electrons: (e+ + e-)/2 e± from heavy flavor: problem I • how to measure a clean spectrum of inclusive e±? • electrons are RARE!
2 central electron/photon/hadron spectrometer arms: |h| 0.35 p 0.2 GeV/c 2 forward muon spectrometers: 1.2 < |h| < 2.4 p 2 GeV/c PHENIX & STAR at RHIC 3 detectors for event characterization: vertex, centrality, reaction plane STAR optimized for hadrons, but can do leptons PHENIX optimized for leptons, but can do hadrons • charged particles • |h| < 1 • pT > 0.15 GeV/c • charged particle ID: • TPC (dE/dx) • Time-of-Flight detector • additional electron ID: EMC • muons • 1.2 < |h| < 2.4 • p > 2 GeV/c • tracking • muon ID: • “absorber” • electrons • |h| < 0.35 • pT > 0.2 GeV/c • tracking • electron ID: • RICH + EMC
PHOTONIC NON PHOTONIC e± from heavy flavor: problem II • there are MANY electrons sources • Dalitz decay of light neutral mesons • most important p0→ g e+e- • but also: h, w, h’, f • conversion of photons in material • main photon source: p0→ gg • in material: g → e+e- • weak kaon decays • Ke3, e.g.: K± → p0 e±ne • dielectron decays of vector mesons • r, w, f → e+e- • direct radiation • conversion of direct photons in material • virtual photons: g* → e+e- • thermal radiation • heavy flavor decays • how to extract e± from heavy flavor decays from the inclusive spectrum?
Extracting e± from heavy flavor • PHENIX • cocktail subtraction method • ALL relevant background sources are measured • background subtraction e± from semileptonic heavy quark decays • converter subtraction method • converter of known thickness added for part of the run • converter multiplies photonic background by KNOWN factor PRL 96(2006)032001 p+p @ √s = 200 GeV • STAR • large acceptance • direct measurement of ~60% of photonic background • rest: extrapolation + cocktail
How well does this work for PHENIX? • test case: p+p at √s = 200 GeV (PRL 97, 252002 (2006)) • how well is the e± background determined? • comparison of two methods • converter measurement • cocktail calculation • excellent agreement • how large is the ratio of signal to background? • S/B > 1 for pT > 2.5 GeV/c • only Dalitz decays and photon conversions are important • PHENIX: conversion ~ 0.5 x Dalitz • STAR: conversion ~ 5 x Dalitz
p+p @ √s = 200 GeV: the reference PRL 97, 252002 (2006) • e± from heavy flavor decays • comparison with FONLL calculation: Fixed Order Next-to-Leading Log pQCD (M. Cacciari, P. Nason, R. Vogt PRL95,122001 (2005)) • theory has uncertainties / parameters too • data are at “upper edge” of theory band • does this look familiar? • total cross section • scc= 567±57(stat)±224(sys) mb
STAR’s e± from p+p collisions • ratio of e± from heavy flavor decays to FONLL pQCD expectation • STAR (scaled down by 25% compared to original preprint) • earlier STAR publication • PHENIX: PRL 97, 252002 (2006) • PHENIX & STAR e± spectra exhibit the SAME shape as predicted by FONLL!! • scale difference: factor ~2 • PHENIX: superior electron measurement • STAR: D meson measurement should help constrain scc
STAR data STAR: D mesons versus e± • D meson and electron measurements at “low” pT • consistent within (large) uncertainties • e± from heavy flavor decays • who is right/wrong? • pro PHENIX: e± data • pro STAR: D data • how to resolve this issue? • PHENIX: D measurement • difficulty: K identification • STAR: reduce material • difficulty: Silicon Vertex Tracker • is this a “show stopper”?
PHENIX PRELIMINARY 1/TABEdN/dp3 [mb GeV-2] Cold nuclear matter: PHENIX • e± spectrum from heavy flavor decays in d+Au at 200 GeV • d+Au data scaled down assuming binary collision scaling • scaled d+Au data are consistent with fit to p+p reference • agreement holds for various d+Au centrality classes • no indication for large cold effects on heavy flavor production at y = 0.
STAR e± data Cold nuclear matter: STAR • nuclear modification factor RdA for e± from heavy quark decay • RdA is consistent with binary scaling • indication for “Cronin” enhancement (initial state scattering, pT broadening) • consistent with PHENIX • PHENIX & STAR • conclude the SAME regarding cold nuclear matter effects on e± from heavy flavor decays! • comparison of PHENIX/STAR d+Au and p+p data • normalization discrepancy cancels in ratio (d+Au)/(p+p)!
PHENIX: PRL 94, 082301 (2005) extrapolation to full phase space • charm cross section per NN collision: 622 ± 57 ± 160 mb • STAR: 1.4 ± 0.2 ± 0.4 mb (d+Au) • central Au+Au collision: ~20 cc pairs! total yield for pT > 0.8 GeV/c Hot matter: e± yield in Au+Au • total yield in Au+Au follows binary collision scaling (as expected for hard probe)! • spectra of e± from heavy flavor decays for different centralities
Binary scaling of “charm” yield at RHIC • PHENIX and STAR measure heavy quark production in various systems • determine sccper binary collision • experiments are self consistent but not consistent with each other • spectral shapes measured by PHENIX & STAR agree in p+p and d+Au → what about Au+Au?
PRL 96, 032301 (2006) PRL 96, 032301 (2006) Discovery of heavy quark energy loss • cocktail analysis of PHENIX Run-2 Au+Au data set • strong modification of heavy quark e± spectra at high pT (similar to p0) • uncertainties too large for stronger conclusions!
Dramatic progress: Run-2 → Run-4 • Run-4 Au+Au data sample: ~109 MB events (~40 x Run-2) • PHENIX: nucl-ex/0611018 • electron measurement extended beyond RICH Cerenkov threshold for pions (pT > 5 GeV/c) • stringent Cerenkov ring selection • “shower shape” cuts in the electromagnetic calorimeter
Dramatic progress: Run-2 → Run-4 • stronger high pT suppression in central collisions • strikingly similar to suppression of light hadrons except for • intermediate pT • highest pT? • careful: decay kinematics! • bottom??? • Run-4 Au+Au data sample: ~109 MB events (~40 x Run-2) nucl-ex/0611018 • indication for light vs. heavy quark mass hierarchy in energy loss at intermediate pT
Heavy flavor e± RAA: PHENIX vs. STAR • is the disagreement between PHENIX & STAR a normalization issue “only”? • use RAA of e± from heavy flavor decays as test case • for d+Au collisions PHENIX & STAR agree in RdA • the same is true for the Au+Au system in • peripheral • mid-central • central collisions • differences between PHENIX & STAR “disappear” in RAA!
Heavy flavor e± RAA: data vs. theory • calculations invoking heavy quark energy loss by gluon radiation • describing the measured suppression is difficult • radiative energy loss of charm and bottom quarks is not enough with typical gluon densities of the produced medium in Au+Au collisions (Djordjevic et al., PLB 632(2006)81) • models involving extreme conditions, implemented via a large transport coefficient q (Armesto et al., PLB 637(2006)362) • agree better with e± data • very “opaque” medium • problems with entropy conservation • there must be something else
Heavy flavor e± RAA: data vs. theory • the return of collisional energy loss • collisional energy loss can be important for heavy quarks • the original idea is old (1982): • J.D. Bjorken (Fermilab-Pub-82/59-THY) • implement collisional energy loss into models • agreement with data gets better, but isn’t perfect yet • collisional + radiative energy loss: Wicks et al., nucl-th/0512076 • additional resonant elastic scattering: van Hees & Rapp, PRC73 (2006) 034913
Heavy flavor e± RAA: data vs. theory • and now for something completely different: collisional dissociation • let’s take heavy quark dynamics serious • what if heavy quarks • fragment inside the medium • form D/B mesons, which then dissociate • Adil & Vitev, hep-ph/0611109 • strong suppression for charm AND bottom at high pT • open questions • how do heavy quarks interact in detail with the medium produced in Au+Au collisions at RHIC • where does bottom decay become important? • need more information
K+ p- Electrons are not “born alone” • e± (m±) from semileptonic heavy quark decays are correlated with products from the original cc pair • hadrons originating from the same parent D/B meson decay • eh correlations (“near side”) → bottom/charm • hadrons originating from the decay of the associated D/B meson • eh correlations (“away side”): too insensitive • leptons from the decay of the associated D/B meson • ee correlations: → energy loss / thermalization • em correlations: → “intermediate” rapidity • mm correlations: → “forward” rapidity
eh correlations in p+p: b vs. c • azimuthal angle correlation of e± from heavy flavor decay with hadrons • “near side” correlation is dominated by decays kinematics • bottom is “wider” than charm due to the larger parent meson mass • assumptions • decays are described properly in PYTHIA • background of (jet) correlations of photonic electrons with hadrons is subtracted properly • ratio of bottom/charm can be determined from line shape analysis • preliminary STAR result agrees with FONLL within large (model dependent) uncertainties • alternative (more direct) approach • invariant mass of eh pairs • pairs with meh>mD ARE from B decays
PHENIX Preliminary Systematic and Normalization Error Dielectrons in Au+Au (I) • invariant mass analysis of e+e- pairs (~870x106 MB events) • problem: HUGE combinatorial background • subtracted via event mixing (sys. error in BG normalization ~0.25 %) • finally • a spectrum with familiar features (J/y) • what else? • where are correlated charm decays? • how does the interaction of charm with the medium manifest itself?
PHENIX Preliminary Dielectrons in Au+Au (II) • intermediate mass region of e+e- continuum (from f to J/y) • expected to be dominated by charm decays • contribution from thermal radiation (not shown) is possible • charm interaction with medium • energy loss • loss of angular correlation • p+p reference is unavailable • RCP: the poor man’s RAA • charm quarks interact strongly with the medium: thermalization?
Y High pressure X Low pressure Z Reaction plane: Z-X plane pY asymmetric pressure gradients (early, self quenching) pX Does charm thermalize? • RAA/RCP << 1 → strong interaction with the medium • large charm mass implies long thermalization time scale • unless interaction with the medium is very strong • collective motion of the medium produced in Au+Au collisions at RHIC • elliptic flow • spatial anisotropy in initial stage • momentum anisotropy in final stage • elliptic flow strength
G. Moore and D. Teaney: PRC 71, 064904 (2005) Interaction of charm with the medium • do charm quarks participate in collective motion? • elliptic flow parameter v2 • momentum aniso- tropy w.r.t. reaction plane orientation • viscous 3-d hydrodynamics calculation • RAA and v2 go hand in hand! • decreasing diffusion coefficient D of charm quarks in the medium • RAA of charm quarks gets smaller at high pT • v2 of charm quarks gets larger • this should still be visible in the e± from semi leptonic decays • where there is energy loss there should be elliptic flow!
χ2 minimum result D->e 2σ 1σ 4σ Does charm flow? • strong elliptic flow of electrons from D meson decays → v2D > 0 • v2c of charm quarks? • recombination Ansatz: (Lin & Molnar, PRC 68 (2003) 044901) • universal v2(pT) for all quarks • simultaneous fit to p, K, e v2(pT) a = 1 b = 0.96 c2/ndf: 21.85/27 • within recombination model: charm flows as light quarks!
Combining RAA and v2 • large suppression and v2 of electrons → charm thermalization • transport models suggest • small heavy quark relaxation time • small diffusion coefficient DHQ x (2pT) ~ 4-6 • this value constrains the ratio viscosity/entropy • h/s ~ (1.5 – 3) / 4p • within a factor 2-3 of conjectured lower quantum bound • consistent with • light hadron v2 analysis (R. Lacey et al., nucl-ex/0609025) • pT fluctuation analysis (S. Gavin & M. Abdel-Aziz, nucl-th/0606061) • while this conclusion is MODEL DEPENDENT it motivates the term “perfect fluid” for the medium produced in Au+Au collisions at RHIC nucl-ex/0611018
Summary: heavy quarks at RHIC • first systematic and comprehensive “heavy quark” measurements in hadronic collisions • heavy quarks are a COMPLEMENTARY hard probe • unique and powerful observables • agreement between PHENIX & STAR is not perfect • many surprising results • challenges for the current theoretical understanding • much more to expect with increasing luminosity and detector upgrades available at RHIC
The future is bright for heavy quark physics • at RHIC • detector upgrades • helpful for electron measurements (in particular for low to intermediate-mass e+e- pairs) • Dalitz and conversion rejection for single e± and e+e- pairs • hadron blind detector (“HBD”) available in PHENIX by Fall 2006! • helpful for improved reaction plane measurements • PHENIX reaction plane detector • needed for optimum heavy quark measurements • measurement of displaced heavy quark decay vertices • silicon vertex trackers are THE cornerstones in the upgrade programs of both PHENIX and STAR • RHIC-II (40 x design luminosity of RHIC) • luminosity matters: • J/y and U spectroscopy and high statistics c & b data • and elsewhere • ALICE / CMS / ATLAS @ LHC: √sLHC ~ 30 x √sRHIC • CBM @ FAIR: “terra incognita” RHIC is CHARMING and the future looks BEAUTIFUL!