Comparison of the Total Charged-Particle Multiplicity in High-Energy Heavy Ion Collisions with e+e- and pp/pbar-p Data

5 Pages·0.15 MB·English

by  B. B. Back

Comparison of the Total Charged-Particle Multiplicity in High-Energy Heavy Ion
Collisions with e+e− and pp/pp Data
B.B.Back1, M.D.Baker2, D.S.Barton2, R.R.Betts6, M.Ballintijn4, A.A.Bickley7, R.Bindel7, A.Budzanowski3,
W.Busza4, A.Carroll2, M.P.Decowski4, E.Garc´ia6, N.George1,2, K.Gulbrandsen4, S.Gushue2, C.Halliwell6,
J.Hamblen8, G.A.Heintzelman2, C.Henderson4, D.J.Hofman6, R.S.Hollis6, R.Ho lyn´ski3, B.Holzman2,
A.Iordanova6, E.Johnson8, J.L.Kane4, J.Katzy4,6, N.Khan8, W.Kucewicz6, P.Kulinich4, C.M.Kuo5,
W.T.Lin5, S.Manly8, D.McLeod6, J.Micha lowski3, A.C.Mignerey7, R.Nouicer6, A.Olszewski3, R.Pak2,
I.C.Park8, H.Pernegger4, C.Reed4, L.P.Remsberg2, M.Reuter6, C.Roland4, G.Roland4, L.Rosenberg4,
3 J.Sagerer6, P.Sarin4, P.Sawicki3, W.Skulski8, S.G.Steadman4, P.Steinberg2, G.S.F.Stephans4,
0 M.Stodulski3, A.Sukhanov2, J.-L.Tang5, R.Teng8, A.Trzupek3, C.Vale4, G.J.van Nieuwenhuizen4,
0
2 R.Verdier4, B.Wadsworth4, F.L.H.Wolfs8, B.Wosiek3, K.Wo´zniak3, A.H.Wuosmaa1, B.Wys louch4

n 1 Argonne National Laboratory, Argonne, IL 60439-4843, USA
a 2 Brookhaven National Laboratory, Upton, NY 11973-5000, USA
J 3 Institute of Nuclear Physics, Krak´ow, Poland

8 4 Massachusetts Institute of Technology, Cambridge, MA 02139-4307, USA
2 5 National Central University, Chung-Li, Taiwan
6 University of Illinois at Chicago, Chicago, IL 60607-7059, USA
1 7 University of Maryland, College Park, MD 20742, USA
v 8 University of Rochester, Rochester, NY 14627, USA
7
(Dated: February 5, 2008)
1
0 The PHOBOS experiment at RHIC has measured the total multiplicity of primary charged par-
1 ticles as a function of collision centrality in Au+Au collisions at √sNN = 19.6, 130 and 200 GeV.
0 Above√s 20GeV,thetotalmultiplicityperparticipatingnucleonpair( Nch / Npart/2 )incen-
3 tralevents≈scales with √s in thesameway as Nch in e+e− data. This ishsuggiesthive of aiuniversal
0 h i
mechanism ofparticleproduction instrongly-interactingsystems, controlled mainly bytheamount
/
x of energy available for particle production (perparticipant pair for heavyion collisions). The same
e effect has been observed in pp/pp data after correcting for the energy taken away by leading par-
- ticles. An approximate independence of Nch / Npart/2 on the number of participating nucleons
l h i h i
c isalso observed,reminiscentof“woundednucleon”scaling (Nch Npart),butwiththeconstant of
u proportionality set by themultiplicity measured in e+e− data ra∝ther than by pp/pp data.
n
: PACSnumbers: 25.75.Dw
v
i
X
Central collisions of two gold nuclei at the top en- lisionsbyN /2,datafromheavyioncollisionsmaybe
r part
a ergy of the Relativistic Heavy Ion Collider (RHIC) at directly compared with similar yields in elementary pp,
Brookhaven National Laboratory produce thousands of pp or even the annihilation of e+e− into hadrons.
charged particles. These are the largest particle mul- While bothe+e− andpp/pp collisionsmustultimately
tiplicities generated in man-made subatomic reactions. allowa descriptionbasedonQuantumChromodynamics
The hope is that these complex systems may reveal evi- (QCD), the theory of the strong interaction, the evo-
denceofthecreationanddecayofaQuark-GluonPlasma lution of these two systems tends to be understood in
(QGP), where quarks and gluons are allowed to explore differentways. Thelargemomentumtransfertothe out-
a volume larger than that of a typical hadron. going produced quark and anti-quark in e+e− reactions
Thehighmultiplicities inheavyioncollisionstypically allows the use of perturbative QCD (pQCD) to describe
arisefromthe largenumberofnucleon-nucleoncollisions thespectrumofquarksandgluonsradiatedasthesystem
which occur, with many of the nucleons struck several fragments[3]. Minimumbiascollisionsofhadronsarenot
times as they pass through the oncoming nucleus. Stud- generallythought to be amenable to sucha perturbative
iesofproton-nucleuscollisionsdemonstratedthattheto- description, since the transverse momentum exchanges
talmultiplicity(N )doesnotscaleproportionallytothe involved are typically less than 1 GeV/c. Instead, phe-
ch
number of binary collisions (N ) in the reaction, but nomenologicalapproaches(e.g. PYTHIA[4])areusedto
coll
rather was found to scale more closely with the num- describe most of the (predominantly soft) particles pro-
ber of “wounded nucleons” which participate inelasti- duced in high energy pp or pp collisions.
cally (N ) [1, 2]. For example, the number of par- A basic connection between perturbative and non-
part
ticipants is N = 2 for a proton-proton collision and perturbativephysicshasbeenelucidatedbysimultaneous
part
N =(N +1) for a proton-nucleus collision. Thus, measurementsofthemultiplicityofchargedparticlesand
part coll
by scaling the particle yields measured in heavy ion col- thehigh-momentum“leading”protonsinppcollisionsat
2
Using the data presented in Ref. [7], Fig. 1a shows
dN /dη/ N /2 averagedoverthe forwardandback-
〉 ch part
〈/2 Npart 4 PUAHLAEO5P B(HpOp (S)e 2+2e000-0) G2G0ee0VV GeV w√90as%rNdNChLe=min2hits0ep0rhvGearle)eiVsd.feoTprehntehdseyomsntoeηmsatanctdeicnatrererarslohArosuw(+rneAopnureteshevenentfiitgnsugaraet
η / PHOBOS 19.6 GeV as a shaded band. The Au+Au data are compared with
d Woods-Saxon Fit
/Nch danNdchd/Ndη/dfyornfoorn-es+ineg−lecdolilffisriaocntsiv(ew(iNthSDcu)tpspapcopllliiesdiotnosr[8e-]
d T
, T 2 ject large initial-state photon radiation) [9] at √s= 200
y GeV.The variable y is the rapidity ofchargedparticles
d T
-+ee/ch (a) relativetotheeventthrustaxis,assumingallparticlesto
N have the pion mass.
d
ItisobservedthattheAu+Audataareverysimilarin
0
magnitude and shape to the e+e− data at the same √s,
W-S Fit 1.12 oavnedrsaimlairlagreirnanshgaepinetηo. tThheepdpiffdaertean(caesssbheotwwneeinntFhieg.e+1be−),
ata / 0.8 (b) andAu+AudistributionsshowninFig. 1acanbepartly
D 0.6
attributed to the different kinematic variables. JETSET
0 2 4 6 8
calculations indicate that the y distribution is slightly
T
ηAA (ye+e-) narrower than the corresponding pseudorapidity distri-
T bution in e+e− collisions, with a higher plateau height.
Yet,evenwithouttakingthisintoaccount,thedifference
between the distributions is no more than 10% for η
FIG.1: (a)dNch/dη/ Npart/2 ofchargedparticlesproduced ± | |
h i and y < 4 [4]. However, the same calculations also
in central Au+Au collisions at √sNN = 200 and 19.6 GeV | T|
(shifted by ∆η = 2.32), compared with elementary systems. show that the choice in kinematic variables does not ex-
A Woods-Saxon fit to the 200 GeV Au+Au data is shown. plainthedifferenceintheforwardregion(above η =4),
The e+e− data are plotted as a function of yT, the rapid- althoughthismaynotbesurprising,asthisregion| s|hould
ity relative to the thrust axis. (b) PHOBOS and UA5 data showsomeresidualeffectofthepresenceofthespectator
divided by a Woods-Saxon fit to the200 GeV Au+Audata.
nucleons.
The similarity of the angular distributions indicates
that the total yield of charged particles in e+e− and
the ISR. Basile et al. [5] found that the average multi-
central Au+Au collisions should also be similar for the
plicity N in pp collisions is similar to that for e+e−
ch same √s, when the nuclear data are scaled by the num-
h i
collisions with √se+e− = √seff, where √seff is the pp ber of participant pairs. To correct for the small ac-
center-of-mass energy minus the energy of the leading
ceptancelossesinthePHOBOSapparatus(whichcovers
particles. This is interpreted as a universal mechanism
η <5.4), we have used several methods inspired by the
ofparticleproductioncontrolleddominantlybytheavail- | |
excellent agreement of the lowest energy PHOBOS data
able center of mass energy [5].
withthe higherenergydatawhenshownasafunctionof
In this Letter, we report results from the PHOBOS η′ = η ybeam [7]. PHOBOS data from √sNN = 19.6
−
experiment on the total multiplicity of primary charged GeV for η > 2.5, shifted by ∆η = y200 y19.6 = 2.32
−
particles Nch as a function of Npart for heavy ion col- (the difference in beam rapidities between the two ener-
h i
lisions at √sNN = 19.6, 130 and 200 GeV, where √sNN gies),displaysthelimitingbehaviordiscussedinRef. [7].
is the nucleon-nucleon center-of-mass energy. Compar- This effectively extends the rapidity coverage to η 8.
isons with pp/pp and e+e− data are made to investigate A Woods-Saxon function for dN/dy fit to the Au+∼Au
whether this universalmechanism of particle production data,alsoprovidesareasonabledescriptionofthedN/dη
applies in the context of heavy ion collisions. distribution, and extrapolates through the lower energy
The PHOBOSmultiplicity detector consistsoftwoar- dataaswell. Thus,inonemethod,weintegratedNch/dη
′
rays of silicon detectors which cover nearly the full solid for √sNN = 130 and 200 GeV for η < 0 and then use
′
angle for collision events. The “Octagon” detector sur- the PHOBOSdata at √sNN = 19.6 GeV for η >0. We
rounds the interaction region with a roughly cylindrical alsointegrateWoods-Saxonfits,similartothatshownin
geometry covering η < 3.2. Two sets of three “Ring” Fig. 1a, for η < 8. These two approaches agree within
| | | |
detectors are placed far forwardand backwardof the in- 2% for central events. For the lowest RHIC energy, we
teraction point and surround the beam pipe, covering simply integrate the charged particles in the PHOBOS
3 < η < 5.4. The methods used for measuring the acceptance.
multip|li|city of chargedparticles as well as for extracting In Fig. 2a, data on N from pp,pp, e+e− andcentral
ch
N havebeendescribedinmoredetailinRefs. [6,7]. heavy ion collisions (scaled by N /2) are shown as a
part part
h i
3
functionof√s. Thepp,pp,ande+e− dataanderrorsare
taken from a compilation [10] and no further corrections ) 40
areapplied. Theerrorsshownarethequadraticallycom- 〉2 pp(pp) Data
bined statistical and systematic errors. Heavy ion data /part 30 epp+e(p- pD)a (t@a s/2)
areshownforcentralAu+AueventsatRHIC(thiswork), N Fit to e+e-
〈
GAueV+)Au[11ev]eanntsdfrPobm+EP8b95evaetntthsefAroGmSN(√As4N9Na=t t2h.6e−S4P.S3 〉(/ch 20 PPHHOOBBOOSS interp.
N NA49
(√sNN = 8.6, 12.2 and 17.3 GeV) [12]. A PHOBOS 〈
E895
Au+Au data point at √sNN = 56 GeV has been added 10
by using the measured value at midrapidity [13] and us- (a)
ing the universal limiting distribution described in Ref.
[7] to approximate the shape of the full distribution. All 0
oftheerrorsshownfortheheavyiondataaresystematic. Fit
Perturbative QCD calculations are able to predict the -
e
dependence ofthe totalmultiplicity ine+e− collisionsas + 1
e
a function of √s, Ne+e−(s) = Cαs(s)Aexp(pB/αs(s)), 〉 /
with A and B fully calculable within pQCD[14]. The ch
N
QCD scale ΛQCD is set to 225 MeV, leaving only a con- 〈
stant of proportionalityC free to fit to the experimental 0.5
(b)
data. A fit to the e+e− data has been made with this
expression (“e+e− fit”) and has been used in Fig. 2b to
see how the various systems comparewith e+e− data by 1 10 102 103
scaling all of the data at a given √s by this function.
s(GeV)
Fig. 2bshowsthatthepp/ppdataareabout30%below
e+e− over the full range of energies. However, rescaling
tbhrein√gssthoef deaacthaipnotoinrtebaysonaafbalcetoargroefem1/e2n,t√wsitehfft=he√e+s/e2−, Fe+IGe−. 2a:nd(ac)enTthraeltAotua+l cAhuaregveedntmsualtsipalicfuitnycthiNonchoiffo√rsp.p,Tphpe,
trend,asshownbytheopendiamonds. Thisisconsistent Au+Au data are normalized by Npart/2. The dotted line is
with measurements of leading protons in pp collisions, a perturbative QCD expression fit to the e+e− data. The
which find dN/dxF (where xF = 2pz/√s in the collider diamonds are the pp/pp data with √seff = √s/2. (b) The
reference frame) to be approximately constant for non- datain(a)dividedbythee+e−fit,toallowdirectcomparison
of different data at thesame √s.
diffractive events over a large range of √s [15] and thus
x 1/2.
F
h i∼
Unlike the pp/pp data, the heavy ion data does not
follow the e+e− trend over the whole energy range. In- more of the energy from the forward direction towards
midrapiditythanfoundinanaveragepp/ppcollision-but
stead,theyliebelowtheppdataatAGSenergies,crosses
ultimately limited by the total incident energy. This hy-
throughtheppdatabetweenAGSandSPSenergies,and
joinssmoothlywiththee+e− dataabovethetopSPSen- pothesis should be testable in proton-nucleus collisions,
by measuring particle yields as a function of ν as was
ergy. Thus, at high energies, the multiplicity measured
done in Refs. [16, 17]. The data in those references sug-
per participantpair in Au+Au collisions scales in a sim-
ilar way to e+e− data at the same √s. If we understand gest that pion yields, whether in the projectile region
the lower effective √s in pp collisions as stemming from (y > 0) [17] or integrated over 4π [16] increase rapidly
for ν < 3 and then much more slowly for ν > 3. How-
the “leading particle effect”, where the leading proton
ever, limited experimental acceptances and theoretical
carries off a substantial amount of the available energy,
uncertaintiesprecludemakinganystrongconclusionsre-
the Au+Audata suggesta substantiallyreducedleading
garding the relationship between the energy loss of the
particleeffectincentralcollisionsofheavynucleiathigh
projectile and the total charged multiplicity.
energy.
The alleviationofthe leading particle effect mightnot InFig. 3 Nch / Npart/2 is shownforPHOBOSdata
h i h i
be so surprising in central nuclear collisions. In the at three RHIC energies as a function of Npart. The
Glauber model, each participating nucleon is typically 90% CL systematic error on the centrality dependence
struck4 6timesonaverageasitpassesthroughtheon- of Nch / Npart/2 is shown as a shaded band, and rep-
− h i h i
coming gold nucleus in a central event (the exact value resents a combination of several factors, dominated by
ofν dependingontheenergy-dependentnucleon-nucleon theuncertaintyoftheextrapolationproceduretoextract
inelastic cross section, σNN(s)). One could speculate Nch over the full solid angle.
that the multiple collisions simultaneously excite and Itmighthavebeenexpectedthat,ineventswithlarger
dissociate the participating nucleons, transferring much impact parameters, each participant would have fewer
4
collisions on average and thus not be fully dissociated.
However, within the systematic errors, the total yield 〉

per participant pair is approximately constant (within 2 30
/ 200 GeV
10%)overthemeasuredcentralityrange,65< N < t
part r
h i a
358, which corresponds to 3 < ν < 6, where ν is the p 130 GeV
averagenumberofcollisionsundergonebyeachoncoming N
nucleon. Thus,itappearsthatonlythefirstfewcollisions 20
〈
have any appreciable effect on particle production. It
/

should be noted that this simple scaling is not observed h
for particle yields measured in a limited pseudorapidity c
N 10 19.6 GeV
range near midrapidity [6].
Proton-antiproton data exist at 200 GeV, but not for
e+e-
the other two RHIC energies. We use a parametrization
pp/pp
of pp data from Ref. [18], N = 4.2+4.69s.155, for 0
19.6 and 130 GeV. Severalhmcehaisure−ments exist in e+e− PHOBOS
at 200 GeV, but not for the other two energies. For
0 200 400
these we use the pQCD formula for Ne+e−, the quality
of the fit clearly indicated in Fig. 2b. Fig. 3 shows that
N
the Au+Au dataareconsistentwith“woundednucleon” part
scaling, in that the multiplicity is proportional to N
part
(N N ). However,N clearly does not scale sim-
ch part ch
∝
ply with the multiplicity measuredinpp collisionsatthe FIG. 3: Nch / Npart/2 is shown vs. Npart for √sNN =
h i h i
same energy. Rather, for a large range of impact pa- 19.6,130,and200GeVasclosedsymbols. Theerrorincludes
rameter, the multiplicity scales approximately with the contributions from the uncertainty on overall Nch scale and
total multiplicity in e+e− annihilation at the same √s. Npart scale. The shaded band shows the uncertainty on the
extrapolation procedure. The open symbols show UA5 data
Thus, it appears that the first few collisions per partici-
at 200 GeV and results from an interpolation for the lower
pantaresufficientto liberateasmuchenergyforparticle energies. Thedottedlinesshowthevaluesfrom thee+e− fit.
production as an e+e− reaction.
However, the rapid approach of N / N /2 in
ch part
h i h i
cweanrtdrsalthheeaev+yei−ondactoallicsoiomnpslbicealtoews a√nsyNsNim∼ple20geGoemVettroic- sfocralhinNgc,hsiu/ghNgepsatrstt/h2ai,tareftmerintihsceefinrtstoffe“wwcooullnisdieodnsnpuecrlepoanr”-
interpretation,as allofthe heavy iondata comparedare ticipant, the multiplicity per participant pair saturates
for a similar range of impact parameters. One feature near the value measured in e+e− reactions. Ultimately,
that might point to why the particle yields at the AGS the existence of simple scaling behavior with √seff and
andSPSareperhaps“suppressed”relativeto e+e− data N indicates stronger constraints on particle produc-
part
(and even to pp data at lower energies, as noted in Ref. tionthanpreviouslyconsideredtheoretically. Thus,these
[12])isthemagnitudeoftheratioofnetbaryonstopions resultsmayprovideanewperspectiveonparticleproduc-
in the system. This ratio, which scales approximately tion in heavy ion collisions.
as Npart/Nch, is O(50%) at AGS energies [11], but is This work was partially supported by U.S. DOE
O(1%) at RHIC [19]. In a thermal statistical approach grants DE-AC02-98CH10886, DE-FG02-93ER40802,
[20], this reflectsthe decreaseofthe baryonchemicalpo- DE-FC02-94ER40818,DE-FG02-94ER40865,DE-FG02-
tential with increasing beam energy. 99ER41099, and W-31-109-ENG-38 as well as NSF
In conclusion, the PHOBOS experiment has mea- grants 9603486, 9722606 and 0072204. The Polish
sured the normalized charged-particle multiplicity groups were partially supported by KBN grant 2 PO3B
Nch / Npart/2 in Au+Au collisions at three RHIC en- 103 23. The NCU group was partially supported by
h i h i
ergies as a function of the centrality of the collision. NSC of Taiwan under contract NSC 89-2112-M-008-024.
Above CERN SPS energies, the total multiplicity per
participating nucleon pair, N / N /2 , in central
ch part
events scales with √s in thhe sami eh way asie+e− data.
This is suggestive of a universal mechanism of parti-
cleproductioninstrongly-interactingsystems,controlled [1] J. E. Elias et al. Phys. Rev.Lett. 41, 285 (1978).
[2] A.Bial as,B.Bleszyn´skiandW.Czyz˙,Nucl.Phys.B111,
mainly by the amount of energy available for particle
461 (1976).
production. This may be related to the multiple colli-
[3] Y. L. Dokshitzer, V. A. Khoze, A. H. Mueller and
sions suffered by each participant nucleon, which could
S. I. Troian, “Basics Of Perturbative QCD,” Gif-sur-
substantially reduce the leading particle effect seen pre- Yvette,France: Ed. Frontieres (1991).
viously in pp collisions. The weak centrality dependence [4] PYTHIAmanual,T.Sjostrand,ComputerPhysicsCom-
5
mun.82,74(1994). JETSET7.4iscurrentlypartofthe [11] J. Klay,U.C. DavisPhD. Thesis (2001).
PYTHIAcode. [12] S. V. Afanasiev et al.,Phys. Rev.C 66, 054902 (2002).
[5] M. Basile et al., Phys. Lett. B92, 367 (1980). M. Basile [13] B. B. Back et al.,Phys. Rev.Lett. 88, 022302 (2002).
et al.,Phys. Lett.B95, 311 (1980). [14] A. H.Mueller, Nucl. Phys. B 213, 85 (1983).
[6] B. B. Back et al., Phys.Rev.C 65, 061901 (2002). [15] M. Batista and R. J. M. Covolan, Phys. Rev. D 59,
[7] B.B.Back et al.,submittedtoPhysical ReviewLetters, 054006 (1999).
arXiv:nucl-ex/0210015 (2002). [16] I. Chemakin, et al., arXiv:nucl-ex/9902009 (1999).
[8] G. J. Alner et al.,Z. Phys.C 33, 1 (1986). [17] NA49 Collaboration, CERN/SPSLC/P264 Add. 5
[9] H.Stenzel,ALEPHCollaboration, Contributedpaperto (2000).
ICHEP2000 (2000). [18] H. Heiselberg, Phys. Rept. 351, 161 (2001).
[10] D.E.Groometal.,Eur.Phys.J.C15,1(2000).Withthe [19] K. Adcox et al.,Phys. Rev.Lett. 89, 092302 (2002).
exception of L3 data, the e+e− data are not completely [20] J. Cleymans, arXiv:hep-ph/0201142.
corrected for Λ and Ks production.