Highlights from BNL-RHIC [617125]

Highlights from BNL-RHIC
M. J. Tannenbaum
Physics Department, 510c,
Brookhaven National Laboratory,
Upton, NY 11973-5000, USA
[anonimizat]
February 8, 2013
1 Introduction
Since the AGS and SpS xed target heavy-ion programs began operations
in 1986, BNL and CERN have been the principal laboratories for measure-
ments of high energy nucleus-nucleus collisions which create nuclear matter
in conditions of extreme temperature and density [1, 2]. The kinetic energy
of the incident projectiles is dissipated in the large volume of nuclear matter
involved in the reaction. At large energy or baryon densities, a phase tran-
sition is expected from a state of nucleons containing conned quarks and
gluons to a state of \deconned" (from their individual nucleons) quarks
and gluons, in chemical and thermal equilibrium, covering a volume that is
many units of the conning length scale. This state of nuclear matter was
originally given the name Quark Gluon Plasma (QGP) [3], a plasma being
an ionized gas.
The startup of the Relativistic Heavy Ion Collider (RHIC) at BNL in
the year 2000 provided a major advance in the eld, leading to the discovery
of the QGP at RHIC, which was announced on April 19, 2005. The results
at RHIC [1] indicated that instead of behaving like a gas of free quarks and
gluons, the matter created in heavy ion collisions at nucleon-nucleon c.m.
energypsNN= 200 GeV appears to be more like a liquid . This matter in-
teracts much more strongly than originally expected, as elaborated in peer
reviewed articles by the 4 RHIC experiments [4, 5, 6, 7], which inspired the
Research supported by U.S. Department of Energy, DE-AC02-98CH10886.arXiv:1302.1833v1 [nucl-ex] 7 Feb 2013

theorists [8] to give it the new name \sQGP" (strongly interacting QGP).
These properties were quite dierent from the properties of the CERN SpS
xed-target heavy ion results which led to the claim of \new state of mat-
ter", in a press-conference [9] on February 10, 2000, that was neither peer-
reviewed nor published. However, results in the past two years from Pb+Pb
measurements at the CERN-LHC atpsNN= 2760 GeV (2.76 TeV) conrm
the RHIC discoveries [1, 2] and add some new information|notably with
fully reconstructed jets [10, 11].
J= suppression, proposed by Matsui and Satz in 1986 [12] as the `gold-
plated' signature for deconnement, suered from the signicant J= sup-
pression observed in the Cold Nuclear Matter (CNM) of p+A collisions [13]
(A0:92) relative to the point-like scaling ( A1:0) from p-p collisions expected
for the large mass scale. Although, \anomalous" J= suppression, i.e. more
than the estimated CNM eect, was discovered at the CERN SPS heavy
ion program and is its main claim to fame, the later development of J=
measurements has not been concerned with J= suppression as a signature
of deconnement, but rather with the strong c.m. energy dependence of
the CNM eect and the possibility of regeneration of J= from recombina-
tion of the large number of cand cquarks produced in the QGP (e.g. see
ref. [1]). Nevertheless, the search for J= suppression (as well as thermal
photon/dilepton radiation from the QGP) drove the design of the RHIC
experiments [14] and the ALICE experiment at the LHC [15].
2 The RHIC machine in 2012
With the shutdown of the Tevatron at FERMILAB, on September 30,
2011 after 28 years of operation, RHIC at Brookhaven National Labora-
tory (Fig. 1) is the only hadron collider in the U.S. and one of only two
hadron-colliders in the world, the other being the CERN-LHC. Also RHIC
is the world's rst and only polarized proton collider. RHIC is composed
of two independent rings, of circumference 3.8 km, containing a total of
1,740 superconducting magnets. RHIC can collide any species with any
other species and since beginning operation in the year 2000 has provided
collisions at 15 dierent values of nucleon-nucleon c.m. energy,psNN, and
six dierent species combinations. The performance history of RHIC with
A+A and polarized p-p collisions is shown in Fig. 2.
This past year, the performance was even more outstanding than usual
thanks to the new EBIS source and 3-dimensional stochastic cooling [16].
For the rst time in a collider, Cu+Au and U+U collisions atpsNN= 200
2

Figure 1: Aerial view of the RHIC facility [14]. The six crossing points
are labelled as on a clock. The two principal experiments still run-
ning are PHENIX and STAR. Two smaller experiments PHOBOS [5] and
BRAHMS [4] have been completed; a test run, AnDy, occupies the former
location of BRAHMS. The LINAC is the injector for polarized protons into
the Booster/AGS/RHIC chain; with the Jet Target used for precision beam
polarization measurements. The TANDEM injector for Ions has been re-
placed by the Electron Beam Ion Source (EBIS) starting with the 2012 run.
34
Appendix A: History of RHIC Beam Performance

Figure A.1. Delivered integrated luminosity vs. weeks in physics production over the years at
RHIC for heavy -ion runs (left) and polarized proton runs (right). pp-equivalent luminosities are
plotted to facilitate compar ison for different species. The average beam polarizations measured
online are also indicated in the right frame (usually, the beam polarization within collision
constraints imposed by the STAR and PHENIX detectors is several percent higher). The steadily
increasing slopes reflect incremental improvements over the years, including the installation of
bunched beam stochastic cooling systems between 2007 and 2012. The installation of EBIS in
2012 facilitated first U -U collisions and first collisions of asymm etric (Cu -Au) heavy ion species.

Table A.1. Colliding beam species and energies run to date at RHIC, or contemplated for near –
future running. The variety of configurations and energies illustrates RHIC’s versatility.

a) Asymmetric rigidity configurations.
b) For comparison, we list best luminosity performance to date at other colliders,
scaled (ܮןJ) to RHIC energies (255 GeV p, 100 GeV/A heavy ions): Tevatron
݌ҧെ݌ (110u1030 cm2s1); LHC unpolarized p-p (430 u1030 cm2s1 with 50 ns
bunch spacing vs. RHIC’s 107 ns); LHC Pb -Pb (1.6u1030 cm2s1).
Figure 2: a)(left) Au+Au performance, where the nucleon-pair luminosity
is dened as LNN=ABL, whereLis the luminosity and A,Bare
the number of nucleons in the colliding species. b) (right) Polarized p-p
performance. Courtesy Wolfram Fischer.
3

GeV were studied, with the purpose of: i) J= suppression with no-corona
in central collisions (Cu+Au); ii) a large eccentricity of the overlap zone
in central collisions by using a very large deformed nucleus with a prolate
shape, like a rugby ball (U+U). Also, the polarized p-p runs in 2012 included
bothps= 200 and 510 GeV with improved polarization and luminosity for
the purposes of: iii) comparison data for the new silicon vertex detectors
(200 GeV); iv)
avor-identied parton spin distribution functions using the
parity violating single spin asymmetry in Wproduction (500 GeV) [17].
3 Issues in A+A versus p-p collisions
The principal dierence in dealing with collisions of relativistic heavy ions,
e.g. Au+Au, compared to p-p or e-p (or e-A) collisions at the same nucleon-
nucleon c.m. energy,psNN, is that the particle multiplicity is A times
larger in A+A central collisions than in p-p collisions as shown in actual
events from the STAR and PHENIX detectors at RHIC (Fig. 3). This year,
Figure 3: a) (left) A p-p collision in the STAR detector viewed along the
collision axis; b) (center) Au+Au central collision atpsNN= 200 GeV in
STAR; c) (right) Au+Au central collision atpsNN= 200 GeV in PHENIX.
both experiments have benetted from incremental upgrades, with STAR at
present making use of an EM calorimeter with a shower-maximum detector;
and PHENIX with central and \forward" silicon vertex detectors for event-
by-event identication of candbquarks.
A schematic drawing of a collision of two relativistic Au nuclei is shown
in Fig. 4a. In the center of mass system of the nucleus-nucleus collision,
the two Lorentz-contracted nuclei of radius Rapproach each other with
impact parameter b. In the region of overlap, the \participating" nucleons
interact with each other, while in the non-overlap region, the \spectator"
nucleons simply continue on their original trajectories and can be measured,
4

SpectatorsParticipants~15fmddn_ch15fm 0fmb = impact parameter0394N_part
Lars Ewell (BNL) Peripheral(a.u.)Maximum impactparameter ~ 15fmMaximumnumber ofpart. = 394 = 2x197Centralblog scale
Figure 4: a) (left) Schematic of collision in the N-Nc.m. system of two
Lorentz contracted nuclei with radius Rand impact parameter b. The
curve with the ordinate labeled d=dn chrepresents the relative probabil-
ity of charged particle multiplicity nchwhich is directly proportional to the
number of participating nucleons, Npart. b)(right) raw nchdistributions in
p-p to U-U collisions atpsNN= 200 GeV from PHENIX [18].
in principle, by Zero Degree Calorimeters (ZDC), so that the number of
participants, Npart, can be directly determined. The degree of overlap is
called the centrality of the collision, with b0, being the most central and
b2R, the most peripheral. The impact parameter bcan not be measured
directly, so the centrality of a collision is dened in terms of the upper
percentile of nchdistributions, e.g. top 10%-ile, upper 10 20%-ile, from
whichbandNpartcan be deduced.
The energy of the inelastic collision is predominantly dissipated by multi-
ple production of soft particles ( hpTi0:36 GeV/c), where nch, the number
of charged particles produced, is directly proportional to the number of par-
ticipating nucleons ( Npart) as sketched on Fig. 4a. Figure 4b shows the
measured distributions [18] of the charged particle multiplicity, nch, at mid-
rapidity atpsNN= 200 GeV for all the combinations of A+B collisions
measured at RHIC. The increase of nchwith bothAandBis evident.
3.1 Collective Flow
Another unique feature of A+A collisions compared to either p-p or p+A
collisions is the \
ow" observed, which is a collective eect that can not be
obtained from a superposition of independent N-N collisions. Immediately
after an A+A collision, the overlap region dened by the nuclear geometry
of participants is roughly almond shaped (see Fig 5a); with the shortest
axis roughly along the impact parameter vector, which together with the
5

Figure 5: a) (left) Almond shaped overlap zone generated just after an
A+A collision where the incident nuclei are moving along the zaxis and
the impact parameter vector is along the xaxis. b)(right) Measurements of
elliptical-
ow ( v2) and triangular
ow ( v3) for identied hadrons plotted as
vk, divided by the number of constituent quarks nqin the hadron raised to
the powerk=2, as a function of KET=nq[19].
beam (z) axis denes the reaction plane. Fluctuations in the distribution of
participants from event-to-event do not respect the average symmetry of the
almond,!+, so that the event plane of the participants, with angle
k, is relevant in the fourier decomposition of the azimuthal distribution:
Ed3N=dp3= (d3N=2pTdpTdy)[1 +X
k12vkcosk( k)] (1)
which is no longer restricted to even harmonics vkby symmetry so that odd
harmonics such as v3can be produced [1], which was observed (Figs 5b).
The
ow can also be seen in two-particle correlations.1
The fact that the
ow persists for pT>1 GeV/c [20] and that the higher
harmonics are not strongly damped [21, 22] implies that the viscosity is
small, perhaps as small as a quantum viscosity bound from string theory [23],
=s= 1=(4) whereis the shear viscosity and sthe entropy density per unit
volume. This has led to the description of the \sQGP" produced at RHIC
as \the perfect
uid" [8]. It was observed that the
ow is not dominated by
nal state hadrons but is proportional to the number of constituent quarks
nqin the mesons and baryons, in which case v2=nq, as well asv3=n3=2
q[19], as
1If two particles AandBare correlated to the event plane but not otherwise correlated,
the analog of Eq. 1 is dNAB=dAdB/[1 +P
k12vA
kvB
kcosk(AB)].
6

a function of the transverse kinetic energy per constituent quark, KET=nq,
would be universal, as shown in Fig. 5b, where the power k=2 and the
transverse kinetic energy, KETmTm0, are suggested by the relativistic
hydrodynamics of \ideal"
uids [24, 25].
4 Old and New RHIC results
One of the most important innovations at RHIC was the use of hard scat-
tering as an in-situ probe of the medium produced in A+A collisions by the
eect of the medium on outgoing hard-scattered partons produced by the
initial A+A collision. Measurements in p+A (or d+A) collisions, where no
(or negligible) medium is produced, allow correction for any modication of
the nuclear structure function from an incoherent superposition of proton
and neutron structure functions.
Hard-scattering of partons in p-p collisions was discovered at the CERN-
ISR in 1972 as shown at the ICHEP 1972 conference [26] by observation that
thee6pTdependence for pion production at low pTbreaks to a power-law
forpT>2 GeV/c with a characteristicpsdependence (Fig. 6a) which is
more evident from the log-log plot of subsequent 0data (Fig 6b). This
plot, as a function of xT= 2pT=ps(Fig 6b), shows that the cross section
for hard-processes (Fig 6c) obeys the scaling law, called\ xT-scaling":
Ed3
d3p=1
pne
TF(pTps) =1psneG(xT) (2)
forps53:1 GeV,xT0:3 (pT7 GeV/c), where ne(xT;ps)46
gives the form of the force-law between constituents as predicted by Quan-
tum Chromodynamics (QCD) with non-scaling structure and fragmentation
functions and running coupling constant [1].
Figure 6a provided the rst evidence that the electrically charged partons
of e-p deeply inelastic scattering (DIS) interacted with each other much
more strongly than electromagnetically [2]. Note that the parton degrees of
freedom in p-p collisions become evident only for parton-parton scattering
7

Figure 6: a) (left) Plot of invariant single particle inclusive cross sections vs.
pTat the CERN-ISR [26]. b) (center) Log-log plot of CCOR [27] invariant
cross sections vs xT= 2pT=ps. c) (right) CCOR [27] ne(xT;ps) derived
from the combinations indicated. The systematic normalization error atps= 30:6 GeV has been added in quadrature. There is an additional
common systematic error of 0:33 inne.
at largeQ22p2
T>58 (GeV/c)2, while in DIS, only Q2>1 (GeV/c)2is
required.
The use of hard-scattering to probe the thermal or \soft" medium pro-
duced in RHI collisions was stimulated by pQCD studies [28] of the energy
loss of partons produced by hard scattering, \with their color charge fully
exposed", in traversing a medium \with a large density of similarly exposed
color charges". The conclusion was that \Numerical estimates of the loss
suggest that it may be signicantly greater in hot matter than in cold. This
makes the magnitude of the radiative energy loss a remarkable signal for
QGP formation " [28]. In addition to being a probe of the QGP the fully
exposed color charges allow the study of parton-scattering with Q215
(GeV/c)2in the medium where new collective QCD eects may possibly be
observed.
The discovery, at RHIC [29], that 0's produced at large transverse mo-
menta are suppressed in central Au+Au collisions by a factor of 5 com-
pared to point-like scaling from p-p collisions is arguably themajor discovery
in Relativistic Heavy Ion Physics. For 0(Fig. 7a) [30] the hard-scattering
in p-p collisions is indicated by the power law behavior pn
Tfor the invariant
cross section, Ed3=dp3, withn= 8:10:05 forpT3 GeV/c. The Au+Au
8

(GeV/c)Tp1 10)2/GeV2 (cdyTdpN2d
evt NT pπ2 1
−910−710−510−310−11010
(0−10%)AA T×p+p
Au+Au 0−10%
Figure 7: a) (left) Log-log plot of invariant yield of 0atpsNN= 200 GeV
as a function of transverse momentum pTin p-p collisions multiplied by
hTAAifor Au+Au central (0{10%) collisions compared to the Au+Au mea-
surement [30]. b) (right) RAA(pT) for all identied particles so far measured
by PHENIX in Au+Au central collisions atpsNN= 200 GeV.
data at a given pTcan be characterized either as shifted lower in pTbypT
from the point-like scaled p-p data at p0
T=pT+pT, or shifted down in
magnitude, i.e. suppressed. In Fig. 7b, the suppression of the many iden-
tied particles measured by PHENIX at RHIC is presented as the Nuclear
Modication Factor, RAA(pT), the ratio of the yield of e.g. per central
Au+Au collision (upper 10%-ile of observed multiplicity) to the point-like-
scaled p-p cross section, where hTAAiis the average overlap integral of the
nuclear thickness functions:
RAA(pT) =d2N
AA=dpTdyNAA
hTAAid2pp=dpTdy: (3)
The striking dierences of RAA(pT) in central Au+Au collisions for the
many particles measured by PHENIX (Fig. 7b) illustrates the importance of
particle identication for understanding the physics of the medium produced
at RHIC. Most notable are: the equal suppression of 0andmesons by a
constant factor of 5 ( RAA= 0:2) for 4pT15 GeV/c, with suggestion
of an increase in RAAforpT>15 GeV/c; the equality of suppression of
direct-single e(from heavy quark ( c,b) decay) and 0atpT>5 GeV/c;
the non-suppression of direct-
forpT4 GeV/c; the exponential rise
ofRAAof direct-
forpT<2 GeV/c [31], which is totally and dramat-
ically dierent from all other particles and attributed to thermal photon
production by many authors (e.g. see citations in reference [31]). For pT>4
GeV/c, the hard-scattering region, the fact that all hadrons are suppressed,
9

but direct-
are not suppressed, indicates that suppression is a medium ef-
fect on outgoing color-charged partons likely due to energy loss by coherent
Landau-Pomeranchuk-Migdal radiation of gluons, predicted in pQCD [28],
which is sensitive to properties of the medium. Measurements of two-particle
correlations [1] conrm the loss of energy of the away-jet relative to the trig-
ger jet in Au+Au central collisions compared to p-p collisions. However,
there are still many details which remain to be understood, such as the ap-
parent suppression of direct-
forpT>18 GeV/c, approaching that of the
0. Interesting new results this year extend and clarify these observations.
An improved measurement of 0production in Au+Au and p-p collisions
by PHENIX [32] now clearly shows a signicant increase of RAA(decrease
in suppression) with increasing pTover the range 7 < pT<20 GeV/c for
0-5% central Au+Au collisions atpsNN= 200 GeV (Fig. 8a). Interestingly,
despite more than a factor of 20 higherpsNN, the ALICE RAAdata from
LHC [33] are nearly identical to the RHIC measurement for 5 < pT<20
GeV/c. However, since the exponent of the power-law at LHC ( n6) is

atter than that at RHIC ( n8, Fig. 7a), a25% larger shift pT=pTin
the spectrum from p-p to A+A is required at LHC to get the same RAA
(Fig. 8b), likely indicating 25% larger fractional energy loss at LHC in
thispTrange.
Measurements by STAR (Fig. 9a) [34] of the evolution of charged hadron
suppression withpsNNin Au+Au collisions, using the variable RCP=
(R05%
AA=R6080%
AA) which does not require the p-p cross section (see Eq. 3)
468101214161820-11010-5% Centrality Au+Au 200GeV0πPHENIX
Pb+Pb 2.76TeV+/-ALICE h
, alt. p+p ref.+/-ALICE h
[PLB 696(2011)30]AAR
(GeV/c)Tp
(p+p) (GeV/c)Tp4 6 8 10 12 14 16 18 20)T/pTpb > (lossS
-0.0500.050.10.150.20.250.30.350.40.45(global)=0.3%b Pb+Pb 0-5%,
(global)=1.0%b Au+Au 0-5%,
(global)=0.7%b Pb+Pb 70-80%,
(global)=2.9%b Au+Au 70-80%,
Figure 8: a) (left) RAAof0inpsNN= 200 GeV central (0-5%) Au+Au
collisions [32] at RHIC compared to charged hadron RAAinpsNN= 2:76
TeV central (0-5%)Pb+Pb collisions at LHC. b) (right) Shift of pTspectrum
Sloss=pT=pTvs.pT(p-p) calculated by PHENIX [32] for RHIC and LHC.
10

but is usually smaller than RAA[4], show the transition from suppression
(RCP<1) to enhancement ( RCP>1) forpsNN>27 GeV.
(0-5%/60-80%)
(GeV/c)
Tp2 4 6 8 10 12 14 16 18 20AAR
110avirtual
-tagging0/
(d+Au)avirtual Au+Au (MB)
=200 GeVNNs
Figure 9: a) (left) STAR RCP(pT) forhas a function ofpsNNin Au+Au
collisions [34]. b)(right) PHENIX RAA(pT) of direct-
in d+Au and Au+Au
minimum bias collisions atpsNN= 200 GeV [35].
Improved measurements of direct-
production in p-p, dAu and Au+Au
collisions by PHENIX [35] show several interesting results. In Fig. 9b, new
measurements of RAA1 for d+Au using internal conversions in the ther-
mal region, pT<4 GeV, reinforce the uniqueness of the exponential rise
of the Au+Au minimum bias photon spectrum, thus conrming that the
exponential for pT<4 GeV/c in Au+Au is a hot matter eect, i.e. thermal
photon production. Also in Fig. 9b, improved measurements of real direct-

in Au+Au collisions, by eliminating background from
-rays associated with
a second
in the0mass range ( 0tagging), no longer show an \apparent
suppression" but are consistent with RAA= 1 out to 20 GeV/c.
My favorite direct-
result this year is the improved PHENIX measure-
ment in p-p collisions atps= 200 GeV out to pT= 25 GeV/c [36], in
excellent agreement with pQCD. A more direct way to show this without
a detailed theory calculation is to use xTscaling. Figure 10a shows xT
scaling for all presently existing direct-
data2, withne= 4:5, very close
to the pure-scaling parton-parton Rutherford scattering result of ne= 4:0
(Fig. 10b). The deviation of the data points in Fig. 10b from the universal
curve forps>38:7 GeV is an illustration of the non-scaling of the coupling
constant, structure and fragmentation functions in QCD|what I like to call
\QCD in action". Forps38:7 GeV, the deviation of the data from the
2This includes the PHENIX p-p direct-virtual-
measurement down to pT1 GeV/c,
further conrming the absence of a soft production mechanism for direct-
in p-p colli-
sions [1].
11

Tx-210-110)3 c⋅-2 GeV⋅(pb3/dpσ3 Ed⋅n/GeV)s(
61071081091010101110121013101410151016101710181019102010
PHENIX (200) This reportPHENIX (200)ATLAS (7000)CMS (7000)CDF (1800)D0 (1800)
UA2 (630)UA1 (630)UA1 (546)R806 (63)R110 (63)E706 (31.5)E706 (38.7)UA6 (24.3)UA6 (24.3)NA24 (23.8)WA70 (22.3)E704 (19.4)R108 (62.4)R807 (63.0)PHENIX (200)Direct photon (y~0)
n=4.5 )}s{Exp. (
Tx-210-110)3 c!-2 GeV!(pb3/dp"3 Ed!n/GeV)s(!0.57000
61071081091010101110121013101410151016101710181019102010
PHENIX (200) This reportPHENIX (200)ATLAS (7000)CMS (7000)
CDF (1800)D0 (1800)
UA2 (630)UA1 (630)UA1 (546)R806 (63)R110 (63)E706 (31.5)E706 (38.7)UA6 (24.3)UA6 (24.3)NA24 (23.8)WA70 (22.3)E704 (19.4)R108 (62.4)R807 (63.0)PHENIX (200)Direct photon (y~0)n=4.0)}s{Exp. (Figure 10: a)(left)psneEd3=dp3, as a function of xT= 2pT=ps, with
ne= 4:5, for direct-
measurements in p-p and  p-p experiments at the (ps
GeV) indicated. [36]. b)(right) same as (a) with ne= 4:0
universal curve in Fig. 10a (and from pQCD) is claimed to be due to the kT
eect (transverse momentum of the quarks in a nucleon).
Since thep

Tof a direct-
can be measured very precisely, the fragmen-
tation function of the jet from the away quark in the reaction g+q!
+q
can be measured by the direct-
hcorrelations (where hrepresents charged
hadrons opposite in azimuth to the direct-
) because the pTof the away-
quark at production is equal and opposite to p

T, thus known to high pre-
cision (modulo a small kT-smearing eect). This year, improved measure-
ments by PHENIX [37] in both p-p and Au+Au collisions (Fig. 11) now
indicate a signicant modication of the fragmentation function in Au+Au
(0-40%) central collisions compared to p-p (Fig. 11a), with an enhancement
at large=lnzT(lowzT=ph
T=p

T) and a suppression at small (largezT)
which is more clearly seen as IAA(), the ratio of the fragmentation functions
in Au+Au/pp (Fig. 11b). As shown in Fig. 11c,d, restricting the away-side
azimuthal range reduces the large  >0:9 (ph
T<3 GeV/c) enhancement but
leaves the suppression at small  <0:9 relatively unchanged, which shows
that the large enhancement is predominantly at large angles, similar to
the eect observed by CMS with actual jets. [38].
12

3
[rad]φΔ-1 [rad]φΔ dN/dtrig1/NφΔ0 0.5 1 1.5 2 2.5 3 < 2.0ξ 1.6 < × < 9 γT5 < p
(b) 2× uncertaintyn>2 v
φΔ0 0.5 1 1.5 2 2.5 3-0.100.10.20.3 < 2.4ξ 2.0 < × < 9 γT5 < p
(a)
φΔ0 0.5 1 1.5 2 2.5 3 < 1.2ξ 0.8 < × < 9 γT5 < p
(d) 2×0 – 40% Au + Au @ 200 GeV
φΔ0 0.5 1 1.5 2 2.5 3-0.0500.050.10.15 < 1.6ξ 1.2 < × < 9 γT5 < p
(c) – h Au+Auincγ – h Au+Audecγ
φΔ0 0.5 1 1.5 2 2.5 3 < 0.4ξ 0.0 < × < 9 γT5 < p(f) 2× – h Au+Audirγ – h p+pdirγ
φΔ0 0.5 1 1.5 2 2.5 3-0.02-0.0100.010.020.030.040.05 < 0.8ξ 0.4 < × < 9 γT5 < p
(e)FIG. 1:∆φdistribution for various associatedξbins for in-clusive photons (downward triangles), decay photons (upwardtriangles) and direct photons (closed circles) for the 0−40%most central Au+Au collisions. The bottom two panels showthe p+p reference (squares).the opposing jet,zT≈z. To focus on the lowzTregion,188one can plot the fragmentation function as a function of189the variable,ξ=ln(1/zT). Our results support previous190observations thatzTscaling holds approximately over a191range of jet energies [11, 19]. Consequently, the results192for differentpγThave been combined into one 5-9 GeV/c193trigger bin.194Figure 1 shows azimuthal opening angle distributions195for the extracted directγ−hcorrelations in 0-40% cen-196tral Au+Au collisions as well as comparison with the197directγ−hcorrelations inp+p. The first three panels198also show the measured inclusiveγ−hcorrelations and199decayγ−hcorrelations determined from reconstructed200π0−hcorrelations used to determine the directγ−h201correlations in Au+Au as a demonstration of the valid-202ity of the method. On the trigger side (|∆φ|<π/2), the203directγ−hcorrelations in Au+Au show no significant204yield, consitent with thep+pcase where an isolation205cut is applied, indicating that the statistical subtraction206method indeed yields direct photons and that the yield207of fragmentation photons in Au+Au is negligible within208uncertainties.209On the opposing, or away, side the associated particle210yield is visible, and there is siginificant variation when211comparing the correlations in Au+Au top+p.T o f u r -212ther quantify this variation, the yields are integrated over213
ξ dN/dtrig1/N-310-210-1101PHENIX Au+Au 0-40% @ 200 GeV 8.8%± – global sys = PHENIX p+p 8%± – global sys = + h + Xγp+p/Au+Au -> (a) Tz0.20.40.60.81 < 7 GeV/chT < 9 GeV/c x 0.5 < pγT5 < p/2π| < φΔ – π||y| < 0.35
) T = ln(1/zξ0 0.5 1 1.5 2 2.5AAI00.511.522.5-1=7 GeV x 10jetBW-MLLA in medium EYAJEM 9-12 GeV/c(b)FIG. 2: The top panel shows per trigger yield as a functionofξforp+pcollisions (squares) and 0-40% most centralAu+Au collisions (closed circles). The bottom panel showsIAA, the ratio of Au+Au to p+p fragmentation functions. Forreference, the dependence onzTis also indicated.∆φfor|π−∆φ|<π/2, as a function ofξ, to obtain the214effective fragmentation function. The top panel of Fig-215ure 2 shows the integrated away-side yields in Au+Au216andp+pas black circles and blue squares, respectively.217The statistical error bars include the point-to-point un-218correlated systematic uncertainty from the background219subtraction, while the boxes around the points show the220correlated uncertainties. For reference, the dependence221onzTis also indicated in a second scale label.222To study medium modification of the jet fragmenta-223tion function, we take a ratio of theξdistribution in224Au+Au top+p. This ratio, known asIAA,i ss h o w n225in the bottom panel of Figure 2 and can be written as226IAA=YAu+Au/Yp+p. Much of the global scale uncer-227tainties cancel in this ratio, but there is a remaining 6%228uncertainty, as indicated by the shaded band around 1.229In the absence of modification,IAAwould equal 1. The230data instead indicate suppression at lowξand enhance-231ment at higherξ. Including all systematic uncertainties232theχ2/dof value for the highest 4 points compared to233the hypothesis thatIAA= 1 is 17.6/6, corresponding to234a probability thatIAAis 1.0 aboveξ=1 of less than 0.1%.235In order to avoid confusion from effects of partonkT236in nucleons and nuclei when comparing experiment with237theory we useIAA, askTeffects have been observed to be238
4
0 0.5 1 1.5 2 2.5AAI00.511.522.5/2π| < φΔ – π|/3π| < φΔ – π|/6π| < φΔ – π|(a) < 7 GeV/chT < 9 GeV/c x 0.5 < pγT5 < p
0 – 40% Au + Au @ 200 GeV ξ0 0.5 1 1.5 2 2.5ratio00.511.522.533.5/6π| < φΔ – π|/2π| < φΔ – π|(b)FIG. 3: The top panel shows theIAAfor the full away-side(|π−∆φ|<π/2) (closed circles) and for two restricted away-side integration ranges,|π−∆φ|<π/3 (squares) and|π−∆φ|<π/6 (triangles). The bottom panel shows the ratio oftheIAAfor|π−∆φ|<π/2t o|π−∆φ|<π/6.similar inp+pand A+A [9, 13–15], thus they roughly239cancel inIAA. The red curve in the bottom panel of240Figure 2 showsIAAcalculated for 7 GeV jets using the241BW-MLLA model in-medium and in vacuum. The vac-242uum calculation agrees well with the measuredξdistri-243bution ine+e−[18]. The blue curve showsIAApredicted244by YaJEM [17]. Both models show suppression at lowξ245due to parton energy loss in Au+Au, and increasingIAA246with increasingξ. In both cases, this is due to the lost247energy being redistributed into enhanced production of248lower momentum particles.249Though the variation inIAAseen in these models is250consistent with the general trend seen in the data, un-251derstanding the details of the transition from suppression252to enhancement can lead to better understanding of how253lost energy is being redistributed. One such detail is how254the observed variation inIAAdepends on the angular dis-255tribution of particles about the away-side jet axis. The256top panel of figure 3 shows theIAAin three integration257ranges of|π−∆φ|<π/2. Reducing the integration range258from|π−∆φ|<π/2 reduces the observed enhancement259and shifts the effect to higherξ. If the integration range260is restricted to|π−∆φ|<π/6, the enhancement becomes261negligible.262To better quantify the angular range of the enhance-263ment, we can look at the ratio ofIAA’s with different264integration ranges, where some of the statistical and sys-265tematic uncertainties common to allIAAcancel. The266bottom panel of figure 3 shows ratios of the full away-267side integration range to the|π−∆φ|<π/6 case. From268this ratio it is clear that there is a significant variation in269observedIAAas a function of the integration range, and270that the enhancement seen at largeξis predominately at271large angles.272In summary, we have presented evidence for medium273modification of jet fragmentation measured via direct274photon-hadron correlations in√sNN= 200 GeV Au+Au275andp+pcollisions. The ratio of Au+Au top+pyields276indicates that particles are depleted at lowξor high mo-277mentum fraction,zT, due to energy loss of quarks travers-278ing the medium. The ratio exhibits an increasing trend279toward highξ, and excedes one at highξ.R e s t r i c t i n g t h e280away-side integration range reduces the enhancement at281highξsignificantly. This suggests that the medium en-282hances production of soft particles at large angles in par-283ton fragmentation.284We thank the staffof the Collider-Accelerator and285Physics Departments at BNL for their vital contribu-286tions. We acknowledge support from the Office of Nu-287clear Physics in DOE Office of Science and NSF (U.S.A.),288MEXT and JSPS (Japan), CNPq and FAPESP (Brazil),289NSFC (China), MSMT (Czech Republic), IN2P3/CNRS290and CEA (France), BMBF, DAAD, and AvH (Germany),291OTKA (Hungary), DAE and DST (India), ISF (Israel),292NRF (Korea), MES, RAS, and FAAE (Russia), VR and293KAW (Sweden), U.S. CRDF for the FSU, US-Hungary294Fulbright, and US-Israel BSF.295[1] K. Adcox et al., Nucl. Phys.A757,1 8 4( 2 0 0 5 ) .296[2] J. Adams et al., Nucl. Phys.A757,1 0 2( 2 0 0 5 ) .297[3] B. B. Back et al., Nucl. Phys.A757,2 8( 2 0 0 5 ) .298[4] I. Arsene et al., Nucl. Phys.A757,1( 2 0 0 5 ) .299[5] K. Adcox et al., Phys. Rev. Lett.88,0 2 2 3 0 1( 2 0 0 2 ) .300[6] C. Adler et al., Phys. Rev. Lett.89,2 0 2 3 0 1( 2 0 0 2 ) .301[7] U. Wiedmann, Springer-Verlag (2010).302[8] R. Baier, D. Schiff, and B. Zakharov, Annu. Rev. Nucl.303Part. Sci.50,3 7( 2 0 0 0 ) .304[9] S. S. Adler et al. (PHENIX), Phys. Rev. C73,0 5 4 9 0 3305(2006).306[10] X.-N. Wang, Z. Huang, and I. Sarcevic, Phys. Rev. Lett.30777,2 3 1( 1 9 9 6 ) .308[11] A. Adare et al., Phys. Rev. D82,0 7 2 0 0 1( 2 0 1 0 ) .309[12] G. Qin, J. Ruppert, C. Gale, S. Jeon, and G. D. Moore,310Phys.Rev.C80,0 5 4 9 0 9( 2 0 0 9 ) .311[13] C. Stewart et al. (Fermilab E557), Phys. Rev. D42,1 3 8 5312(1990).313[14] M. D. Corcoran et al. (Fermilab E609), Phys. Lett.B259,314209 (1991).315[15] D. Naples et al., Phys. Rev. Lett.72,2 3 4 1( 1 9 9 4 ) .316[16] Zhang, J. Owens, E. Wang, and X. N. Wang, Phys. Rev.317Lett.103,0 3 2 3 0 2( 2 0 0 9 ) .318[17] T. Renk, Phys. Rev. C80,0 1 4 9 0 1( 2 0 0 9 ) .319[18] N. Borghini and U. Wiedemann (2005), hep-ph/0506218.320[19] A. Adare et al., Phys. Rev. C80,0 2 4 9 0 8( 2 0 0 9 ) .321[20] B. Abelev et al., Phys. Rev. C82,0 3 4 9 0 9( 2 0 1 0 ) .322[21] L. Aphecetche et al. (PHENIX), Nucl. Instrum. Meth.323A499,5 2 1( 2 0 0 3 ) .324[22] K. Adcox et al. (PHENIX), Nucl. Instrum. Meth.A499,325489 (2003).326[23] S. Afanasiev et al. (PHENIX) (2012), submitted to Phys.327Rev. Lett., May 2012, 1205.5759.328[24] A. Adare et al., Phys. Rev. Lett.107,2 5 2 3 0 1( 2 0 1 1 ) .329[25] A. Sickles, M. P. McCumber, and A. Adare, Phys. Rev.330C81,0 1 4 9 0 8( 2 0 1 0 ) .331Figure 11: a) =ln(ph
T=p

T) =lnzTdistributions of hadrons opposite
in azimuth (j
j< = 2) to a direct-
trigger with 5 < p

T<9 GeV/c
in p-p and 040% central Au+Au collisions atpsNN= 200 GeV from
PHENIX [37]. b) Ratio of the two distributions, IAA(). c) (right)-(top)
IAA() as in (b) when the away-side azimuthal range is restricted as indi-
cated. d) (right)-(bottom) Ratio of IAAforj
j<= 2 toj
j<= 6.
One of the most exciting discoveries at RHIC, now conrmed at the LHC,
is the suppression of heavy quarks by the same amount as light quarks for
pT>5 GeV/c as indicated at RHIC (Fig. 7b) by direct single- efrom heavy
quark (c,b) decay; and at the LHC by Dmesons from c-quarks [39], and
non-prompt J= fromb-quarks [40] (Fig. 12a). The discovery at RHIC was a
total surprise and a problem since it appears to disfavor the radiative energy
loss explanation [28] of suppression (also called jet-quenching) because heavy
quarks should radiate much less than light quarks or gluons.
Many explanations have been oered including some from string theory;
but the explanation I prefer was by Nino Zichichi [41] who proposed that
since the standard model Higgs Boson, which gives mass to the Electro-
Weak vector Bosons, does not necessarily give mass to Fermions, \it cannot
be excluded that in a QCD coloured world (a QGP), the six quarks are all
nearly massless". If this were true it would certainly explain why light and
heavy quarks appear to exhibit the same radiative energy loss in the medium.
This idea can, in fact, be tested because the energy loss of one hard-scattered
parton relative to its partner, e.g. g+g!b+b, can be measured by
experiments at RHIC and LHC using two particle correlations in which both
13

(GeV/c) tp0 2 4 6 8 10 12 14 16 18AAR
00.20.40.60.811.21.41.61.82
ALICE
0-20% centrality
= 2.76 TeVNNs Pb-Pb,
, |y|<0.5*+, D+, D0Average D
|<0.8d Charged hadrons, |
, |y|<2.4s CMS non-prompt J/
27
Rachid Nouicer Quark Matter 2012be/ ( be+ ce) in 200 GeV Au+Auvsp+pFrom Fit of the DCA distribution
Results:Bottom Production in Au+Au and p+pFigure 12: a) (left) RAAof ALICE [39] D-mesons, charged hadrons, and
CMS [40] non-prompt J= , in central (0-20%) Pb+Pb collisions atpsNN=
2:76 TeV b) (right) b-quark fraction Fb=b!e=(c!e+b!e) of direct
single-ein p-p and Au+Au from PHENIX measurement of the Distance
of Closest Approach (DCA) of the displaced vertex.
the outgoing bandbare identied by measurement of the Distance of Closest
Approach (DCA) of their displaced decay vertices in silicon vertex detectors.
When such results are available, they can be compared to 0-charged hadron
correlations from light quark and gluon jets, for which measurement of the
relative energy loss has been demonstrated at RHIC [1].
Of course, measurement of the Yukawa couplings to Fermions of the can-
didate 125 GeV Higgs Boson at the LHC may be available by the end of
2012; but, already this year, the rst direct measurement of b-quarks in p-p
and Au+Au collisions at RHIC by their displaced vertices was made in the
new PHENIX Silicon VTX detector [42]. Figure 12b shows the measure-
ment of the b-quark fraction Fb=b!e=(c!e+b!e) of direct single e
in p-p and Au+Au collisions atpsNN= 200 GeV, using the PYTHIA cand
bquarkpTdistributions in p-p collisions to calculate the DCA distributions
of theein both p-p and Au+Au. The fact that the Au+Au measurements
for allpe
Tare below the p-p measurements indicates clearly that the b-quark
pTspectrum is modied in Au+Au compared to p-p. However, the correct
conditional DCA distribution requires the actual (modied) b-quarkpTspec-
trum in Au+Au, which must be obtained by iteration. Once the iteration
has converged, the Rb
AA(pe
T) can be calculated from the measured RAA(pe
T)
of the direct-single- eby the relation Rb
AA(pe
T) =RAA(pe
T)FAA
b=Fpp
b. For
example, if the nal FAA
b=Fpp
bthenRb
AA(pe
T) =RAA(pe
T).
14

References
[1] Discoveries and methods at RHIC have also been discussed by
the present author in ISSP proceedings for the years 2008
(arXiv:0906.0745), 2009 (arXiv:1006.5701), 2011(arXiv:1201.5900), as
well as in a textbook to appear soon [2].
[2] Jan Rak and Michael J. Tannenbaum, High-pTPhysics in the Heavy
Ion Era (Cambridge University Press, Cambridge UK, 2013).
[3] E. V. Shuryak, Phys. Repts. 61(1980) 71{158.
[4] I. Arsene et al. (BRAHMS Collab.), Nucl. Phys. A 757(2005) 1{27.
[5] B. B. Back et al. (PHOBOS Collab.), Nucl. Phys. A 757(2005) 28{101.
[6] J. Adams et al. (STAR Collab.), Nucl. Phys. A 757(2005) 102{183.
[7] K. Adcox et al. (PHENIX Collab.), Nucl. Phys. A 757(2005) 184{283.
[8] D. Rischke and G. Levin, eds., Nucl. Phys. A 750(2005) 1{171.
[9] Press Conference, A New State of Matter Created at CERN.
[10] A. Aad, et al. (ATLAS Collab.), Phys. Rev. Lett. 105(2010) 252303.
[11] S. Chatrchyan, et al. (CMS Collab.), Phys. Rev. C 84(2011) 024906.
[12] T. Matsui and H. Satz, Phys. Lett. B 178(1986) 416.
[13] D. M. Alde, et al. E772 Collab., Phys. Rev. Lett. 66(1991) 2285{2288.
M. J. Leitch, Eur. Phys. J. C 43(2005) 157{160 and references therein.
[14] The Relativistic Heavy Ion Collider Project: RHIC and its Detectors ,
Nucl. Instrum. Methods A 499(2003) 235{880.
[15] The Large Hadron Collider: Accelerator and Experiments , JINST 3
(2008) S08001{S08007.
[16] M. Blaskiewicz and W. Fischer, CERN Courier 52(Oct. 2012) 17.
[17] G. Bunce, et al. , Ann. Rev. Nucl. Part. Sci. 50(2000) 525.
[18] J. T. Mitchell et al. (PHENIX), Proc. QM2012, Nucl. Phys. A .
[19] Yi Gu et al. (PHENIX), Proc. QM2012, Nucl. Phys. A .
15

[20] D. Teaney, Phys. Rev. C 68(2003) 034913.
[21] B. H. Alver, et al. , Phys. Rev. C 82(2010) 034913.
[22] A. P. Mishra, et al. , Phys. Rev. C 77(2008) 064902.
[23] P. K. Kovtun, et al. , Phys. Rev. Lett. 94(2005) 111601.
[24] N. Borghini and J.-Y. Ollitrault, Phys. Lett. B 642(2006) 227.
[25] e.g. see R. Lacey, et al. (PHENIX Collab.), J. Phys. G 38(2011) 124048.
[26] R. L. Cool, et al. , inProc. XVI Int. Conf. High Energy Physics , Eds.
J. D. Jackson and A. Roberts (NAL, Batavia, IL, 1973), Vol 3, p. 317.
[27] A. L. S. Angelis, et al. (CCOR), Phys. Lett. B 79(1978) 505.
[28] See R. Baier, D. Schi and B. G. Zakharov, Ann. Rev. Nucl. Part. Sci.
50(2000) 37{69, and references therein.
[29] K. Adcox, et al. (PHENIX Collab.), Phys. Rev. Lett. 88(2002) 022301.
[30] S. S. Adler, et al. (PHENIX Collab.), Phys. Rev. C 76(2007) 034904.
[31] A. Adare, et al. (PHENIX Collab.), Phys. Rev. Lett. 104(2010) 132301.
[32] A. Adare, et al. (PHENIX Collab.), arXiv:1208.2254v1.
[33] K. Aamodt, et al. (ALICE Collab.), Phys. Lett. B 696(2011) 30.
[34] Xin Dong et al. (STAR), Proc. QM2012, Nucl. Phys. A .
[35] A. Adare, et al. (PHENIX Collab.), arXiv:1208.1234v1.
[36] A. Adare, et al. (PHENIX Collab.), Phys. Rev. D 86(2012) 072008.
[37] J. Frantz, et al. (PHENIX), Proc. QM2012, Nucl. Phys. A .
[38] G. Roland et al. (CMS), Proc. QM2012, Nucl. Phys. A .
[39] B. Abelev, et al. (ALICE Collab.), JHEP 09(2012) 112.
[40] S. Chatrchyan, et al. (CMS Collab.), JHEP 05(2012) 063.
[41] A. Zichichi, Nucl. Phys. A 805(2008) 36c.
[42] R. Nouicer, et al. (PHENIX), Proc. QM2012, Nucl. Phys. A .
16