φ production in Pb–Pb collisions at 158 GeV/c per nucleon incident momentum

Abstract

The production of vector mesons φ, ρ and ω has been measured in Pb–Pb collisions at 158 GeV/c per nucleon incident momentum at the CERN/SPS. The muon spectrometer of experiment NA50 detects φ, ρ and ω mesons via their μ+μ− decay channel in the collision center of mass rapidity range 0⩽yCM⩽1. The results reported here show that the relative production of the φ compared to the (ρ+ω) and the φ multiplicity per participant nucleon (Npart) increase with the centrality of the collision. On the other hand, the (ρ+ω) multiplicity per participant does not exhibit any Npart dependence within our errors. The inverse slope parameter as deduced from an exponential fit to the φ transverse mass distribution is 228±10 MeV. Our results are compared with those obtained by experiment NA49 and with theoretical calculations.

Full text

Page 1

Physics Letters B 555 (2003) 147–155
www.elsevier.com/locate/npe
φ production in Pb–Pb collisions at 158 GeV/c per nucleon
incident momentum
NA50 Collaboration
B. Alessandrok, C. Alexad, R. Arnaldik, J. Astruci, M. Atayanm, C. Baglinb,
A. Balditc, S. Beolèk, V. Boldead, P. Bordalog,1, G. Borgesg, A. Bussièreb, L. Capellil,
V. Caponib, C. Castanierc, J. Castorc, B. Chaurandj, I. Chevrotc, B. Cheynisl,
E. Chiavassak, C. Cicalòe, T. Claudinog, M.P. Cometsi, N. Constansj,
S. Constantinescud, P. Cortesea, J. Cruzg, A. De Falcoe, N. De Marcok, G. Dellacasaa,
A. Devauxc, S. Ditad, O. Drapierj, L. Ducrouxl, B. Espagnonc, J. Fargeixc, P. Forcec,
M. Galliok, Y.K. Gavrilovh, C. Gerscheli, P. Giubellinok,2, M.B. Golubevah,
M. Goninj, A.A. Grigorianm, S. Grigoryanm, J.Y. Grossiordl, F.F. Guberh,
A. Guichardl, H. Gulkanyanm, R. Hakobyanm, M. Idzikk,3, D. Jouani,
T.L. Karavitchevah, L. Klubergj, A.B. Kurepinh, Y. Le Borneci, C. Lourençof,
P. Macciottae, M. Mac Cormicki, A. Marzari-Chiesak, M. Maserak,2, A. Masonie,
M. Montenok, A. Mussok, P. Petiauj, A. Piccottik, J.R. Pizzil, W. Prado da Silvak,4,
F. Prinoa, G. Puddue, C. Quintansg, L. Ramelloa, S. Ramosg,1, P. Rato Mendesg,
L. Riccatik, A. Romanaj, H. Santosg, P. Saturninic, E. Scalasa, E. Scomparink,
S. Sercie, R. Shahoyang,5, F. Sigaudok, S. Silvag, M. Sittaa, P. Sondereggerf,1,
X. Tarragoi, N.S. Topilskayah, G.L. Usaie,2, E. Vercellink, L. Villattei, N. Willisi
aUniversità del Piemonte Orientale, Alessandria and INFN-Torino, Italy
bLAPP, CNRS-IN2P3, Annecy-le-Vieux, France
cLPC, Univ. Blaise Pascal and CNRS-IN2P3, Aubière, France
dIFA, Bucharest, Romania
eUniversità di Cagliari/INFN, Cagliari, Italy
fCERN, Geneva, Switzerland
gLIP, Lisbon, Portugal
hINR, Moscow, Russia
iIPN, Univ. de Paris-Sud and CNRS-IN2P3, Orsay, France
jLaboratoire Leprince-Ringuet, Ecole Polytechnique and CNRS-IN2P3, Palaiseau, France
kUniversità di Torino/INFN, Torino, Italy
lIPN, Univ. Claude Bernard Lyon-I and CNRS-IN2P3, Villeurbanne, France
mYerPhI, Yerevan, Armenia
Received 21 October 2002; received in revised form 18 December 2002; accepted 20 December 2002
Editor: L. Montanet
0370-2693/03/$ – see front matter  2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0370-2693(02)03267-7

Page 2

148
NA50 Collaboration / Physics Letters B 555 (2003) 147–155
Abstract
The production of vector mesons φ, ρ and ω has been measured in Pb–Pb collisions at 158 GeV/c per nucleon incident
momentum at the CERN/SPS. The muon spectrometer of experiment NA50 detects φ, ρ and ω mesons via their µ+µ−decay
channel in the collision center of mass rapidity range 0 ? yCM? 1. The results reported here show that the relative production
of the φ compared to the (ρ + ω) and the φ multiplicity per participant nucleon (Npart) increase with the centrality of the
collision. On the other hand, the (ρ + ω) multiplicity per participant does not exhibit any Npartdependence within our errors.
The inverse slope parameter as deduced from an exponential fit to the φ transverse mass distribution is 228 ± 10 MeV. Our
results are compared with those obtained by experiment NA49 and with theoretical calculations.
 2003 Elsevier Science B.V. All rights reserved.
PACS: 25.75.Dw; 12.38.M; 14.40.Cs
Keywords: Quark–gluon plasma; Strangeness enhancement; Phi, rho and omega mesons
1. Introduction
Ultra-relativistic heavy ion collisions provide an
experimental opportunity to study the properties of
hadronic matter at high density and temperature.They
can thus be used to probe the phase transition from
ordinary matter to a deconfined quark–gluon plasma
as theoretically predicted by QCD lattice calculations.
An enhancement of strange particle production has
been predicted as a signature of such a transition
[1], and extensively studied through strange meson
and hyperon production [2,3]. In this context, the φ
meson is of particular interest due to its s¯ s valence
quarkcontent.Makinguseoftheirmuonpairdetection
capabilities optimized for charmonium production
[4,5], experiments NA38 and NA50 have extended
their investigations to the lower part of the dimuon
invariant mass spectrum and measured the features of
φ, ρ and ω meson production in heavy ion collisions
[6–8].
The ρ and ω vector mesons are combinations of
u¯ u and d¯d pairs and have masses close to the φ mass.
This makes it particularly interesting to compare their
production rate both as a function of their transverse
mass and of the centrality of the collision.
1Also at IST, Universidade Técnica de Lisboa, Lisbon, Portugal.
2Also at CERN, Geneva, Switzerland.
3Also at Faculty of Physics and Nuclear Techniques, Academy
of Mining and Metallurgy, Cracow, Poland.
4Now at UERJ, Rio de Janeiro, Brazil.
5On leave of absence from YerPhI, Yerevan, Armenia.
Through the φ/(ρ +ω) ratio we have access to the
s¯ s/(u¯ u+d¯d) ratio [9]. Due to its low interactioncross
section, the φ meson could retain information of the
critical phase of the collision during which the plasma
proceeds to hadronization. Moreover, the identifica-
tion of the vector mesons through their leptonic decay
in the (µ+µ−) channel should have some advantages
as compared to their identification through hadronic
decays (like the K+K−decay) as this latter sample
couldbeaffectedbystrongfinalstateinteractions[10].
Experiment NA38 has studied the φ and ρ + ω
production in p–W, p–U, d–C, d–U, O–U, S–S, S–Cu
and S–U collisions at 200 GeV/c per nucleon [6].
In this Letter, we report results obtained on Pb–Pb
interactions at 158 GeV/c per nucleon incident mo-
mentum. They are based on a very large sample of
events collected by the NA50 experiment during the
run of 1996.
2. Experimental setup
A detailed description of the experimental setup
can be found in [11]. Muon pairs from meson decays
are detected in a spectrometer based on an air-gap
toroidal magnet. The dimuon trigger makes use of the
hits recorded by four scintillation counter hodoscopes
which must include the pattern of 2 muon tracks
originating from the target region and passing through
all the elements of the spectrometer.
Events are selected in the collision center of mass
rapidity window 0 ? yCM? 1 which lies inside the
rapidity acceptance of the spectrometer. The muon

Page 3

NA50 Collaboration / Physics Letters B 555 (2003) 147–155
149
angle defined in the Collins–Soper frame [12] is also
restricted to the interval accepted by the spectrometer,
namely,−0.5? cosθC.S.? 0.5.TheNA50spectrome-
ter, as used for the study reported in this Letter,
was kept in its standard configuration and, in fact, is
optimized for the J/ψ meson detection. For lower
mass resonances, its acceptance starts at a transverse
mass of 1.5 GeV/c2and then rises up to values of
3% and 6.3%, respectively, for the ρ (ω) and for
the φ meson, with these higher values reached for a
transverse mass of 3 GeV/c2.
The mass resolution, for these low mass muon
pairs, is constant and results mainly from the multi-
ple scattering of the muons passing through the spec-
trometer absorbers.It amountsto 70 MeV/c2and thus
does not allow to separate the ω and ρ contributions.
The results given in this Letter will, therefore,refer al-
ways to the sum (ρ +ω).
The centrality of the collision is estimated by an
electromagnetic calorimeter which measures, on an
event by event basis, the transverse energy (ET) of
the neutral particles produced in the collision, in the
laboratory pseudo-rapidity interval 1.1 ?η ?2.3.
A very forward (“zero degree”) hadronic calorime-
ter measures, for each Pb–Pb interaction, the energy
of the non-interacting nucleons (spectators) of the Pb
ion projectile. The same detector is used to trigger
the apparatus on a sample of “minimum bias” events.
This “minimum bias” trigger is completely indepen-
dent from the dimuon trigger. It is obtained by re-
quiring a non-zero energy deposit in the very for-
ward calorimeter. It also requires a minimum value for
the transverse energy recorded in the electromagnetic
calorimeter in order to discard events which have not
interacted in the target region.
The average beam intensity during data collection
was 5.5 × 107Pb ions per burst, with a spill of 4.5 s
nominal duration. A total of 170 million events were
recorded.Amongthem ? 90% werecollectedwith the
dimuon trigger and ? 10% with the minimum bias
trigger. This led to ∼130000 reconstructed φ mesons.
3. Data selection and analysis
Events are selected among a set of preselected runs
satisfying standard conditions of beam quality and
stability together with nominal operation conditions
for every subdetector of the apparatus. Moreover, the
surviving data are subject to a number of selection
cuts in order to ensure that all the measured quantities
are free from potential experimentalbiases. A detailed
description of these selection criteria can be found in
[4,11].
Fig.1displaysa typicalµ+µ−invariantmass spec-
trum. The raw spectrum shows the resonant peaks
corresponding to the φ, ρ and ω mesons, superim-
posed on a mass continuum. This mass continuum is
made of correlated muon pairs arising from several
physicalprocesses, mainlyDalitz decays, q ¯ q annihila-
tions and semileptonic decays of associated D? D pro-
duction. This “physical” continuum is mixed with a
combinatorialbackground(called in short background
from now on) which results from the abundant uncor-
related decays of mainly pions and kaons in the partic-
ular low mass range considered in this Letter. Because
of their uncorrelation, these decays generate both op-
posite and like-sign muon pairs.
The background estimate is based on the “mir-
ror” sample of like-sign muon pair events (Nµ+µ+
and Nµ−µ−) which are necessarily uncorrelated and
which are collected, selected and reconstructed un-
der conditions identical to the opposite-sign pairs.
Fig. 1. Raw experimental mass spectrum (“total”) of the detected
µ+µ−pairs with MT> 1.5 GeV/c2for ET> 5 GeV and
combinatorial background (“comb. b.g.”) mass spectrum.

Page 4

150
NA50 Collaboration / Physics Letters B 555 (2003) 147–155
The corresponding background muon pairs populat-
ing the opposite-sign invariant mass distribution have
obviously the same properties and, in particular, the
same shape as their reflected like-sign pairs, thanks
to their identical acceptance resulting from an appro-
priate fiducial cut. It can be shown that they amount
to Nµ+µ−
bg
correlation is perfectly satisfied, which is the case for
high multiplicity Pb–Pb collisions, and when the dif-
ferent decay processes feeding the background have
similar probabilities of producing a muon detected in
the spectrometer, it can be further shown that R = 1
as both analytically computed [13] and confirmed
through a Monte Carlo simulation [14]. Furthermore,
when estimated without specific assumptions from a
fit to our data sample, the same value R ? 1 is found6
confirming,thereby,thatthe requiredconditionsstated
above are satisfied in our particular case. Numerically,
the resulting uncorrelated background in the reso-
nances mass range (0.5? M ? 1.2 GeV/c2) amounts
to 73% of the total muon pair yield.
Inordertoextractthenumberof φ, ρ andω mesons
from the dimuon mass spectra, we proceed as follows.
We makeuse of a completesimulationof the detec-
tor in order to account for the acceptance and smear-
ing imposed by the apparatus to the decay muons pro-
duced in the collision. Each of the components of the
mass spectrum is simulated according to production
distributions depending on specific parameters which
are adjusted, through an iterative procedure, by com-
parison with the actually collected data in the corre-
sponding mass region, after backgroundsubtraction.
The φ and ω mass distributions are generated us-
ing Breit–Wigner shapes. For the ρ meson, a parame-
trization of the Breit–Wigner formulaincludinga non-
resonant backgroundunder the ρ, as in the HELIOS/3
experiment [15], is used. The transverse mass for
the 3 resonances is generated according to the Hage-
dorn distribution, i.e., dN/dMT∝ M2
[16] where K1is a Bessel function and Tpa parame-
ter. The rapidity is generated according to a Gaussian
= 2 × R ×√Nµ+µ+Nµ−µ−. When non-
TK1(MT/Tp)
6This has been checked by fitting the muon pair invariant mass
distribution in the mass range 1.6–2.5 GeV/c2, assumed as made of
Drell–Yan pairs, semi-leptonic decays of associated D? D production
and uncorrelated background shaped by the corresponding like-
sign pairs. When leaving as free parameters the amount of each
component, the fitted value of R is 1.015 ±0.08.
distribution (centered at mid-rapidity).The continuum
is treated in a phenomenological way according to
dN/dM ∝ 1/Mαe−M/β. Its transverse mass and ra-
pidity distributions are taken with the same analytical
shapes as adopted for the resonances. The parameters
of these generation functions (Tp, the rapidity distrib-
ution width, α and β parameters of the continuum)are
adjusted on the experimental distributions separately
for each component as defined by its corresponding
mass region.
With the generation functions of each of the com-
ponents determined as described above, we generate
and reconstruct each of the components of the mass
spectrum and describe the shape of the detected muon
pairs as seen by the detector. We then fit, after subtrac-
tion of the combinatorial background, the experimen-
tal dimuonmass spectrabetween0.25and1.8GeV/c2
to the following expression:
dNµ+µ−
dM
=Aρ+ω
+AφFφ(M)+ACNTFCNT(M),
where the F functions are the mass distributions
corresponding to each of the components “as seen”
by the detector. Fig. 2 shows the dimuon invariant
mass spectrum superimposed to the fitting function.
The Fωand Fφfunctions are Gaussian and reproduce
?Fρ(M)+RBrFω(M)?
(1)
Fig. 2. Fit of the spectrum of the µ+µ−pairs (combinatorial
background subtracted) with MT>1.5 GeV/c2for ET> 5 GeV.

Page 5

NA50 Collaboration / Physics Letters B 555 (2003) 147–155
151
with good accuracy the resolution of the detector as
determined by Monte Carlo simulation. There are
5 free parameters in the fit: the three amplitudes
(Aρ+ω, Aφand ACNT) and the center position of the
two Gaussians (Fωand Fφ). The same cross section
was assumed for ρ and ω production (as observed
in p–p collisions at 400 GeV/c [17]), so that the
amplitude Aρ+ωis the same for the two resonances.
RBrtakes into account the difference between the ρ
and ω branchingratios into µ+µ−. The numbersof φ,
ρ and ω in the nine transverse energy (ET) bins and
the five dimuon transverse mass (MT) bins considered
in this study are finally extracted using exactly the
same procedure.
The continuum is adjusted on the mass spectrum
below and over the ρ + ω and φ peaks by an ad-
hoc 2 parameter function. It is then interpolated in
the resonance mass region by a smooth variation of
the parameters. The procedure is the same in each
transverse mass bin.
It is worthwhile underlining here that if the com-
binatorial background is subtracted in a different way,
namely, with a different value of R, the function to
whichthe continuumis adjustedis thendifferent.Nev-
ertheless, the resonance peaks remain the same, pro-
vided that the combinatorial background is a smooth
functionofthe mass and doesnot displayany peakun-
der the resonances. This is definitely the case from the
experimental shape as determined from the combina-
torial like-signmass spectrum.Therefore,the numbers
of φ, ρ and ω mesons determined by this method are
not sensitive to the absolute value of R.
The uncertainty of the method on the acceptances
as well as on all the results which are presented
hereafter is carefully determined by slightly varying
the generation parameters [8].
4. Results
The ratio (φ/(ρ + ω))µµ of the produced φ and
ρ +ω resonances decaying into a µ+µ−pair, already
corrected for acceptance, is shown in Fig. 3 as a
function of the transverse mass for the whole ET
domain (ET > 5 GeV). This quantity is independent
of the trigger and reconstruction efficiencies which
cancel out in the ratio. The smaller error bars in Fig. 3
correspondto the statistical and to the fit uncertainties.
Fig. 3. The ratio (φ/(ρ + ω))µµ of the number of produced φ
and ρ + ω mesons and decaying into a µ+µ−pair, as a function
of the dimuon transverse mass (MT) over the whole ET domain
(ET> 5 GeV).
The larger error bars include, added in quadrature, the
method uncertainty which depends on MT. As can be
seen from the figure, the general trend is flat, which
indicates that the φ and ρ + ω production rates have
similar MT dependences.
The same ratio is presented in Fig. 4 as a function
of the number of participant nucleons, for the whole
MT domain, namely MT? 1.5 GeV/c2. The number
of participant nucleons (Npart) is deduced for each ET
bin from the mean ET value, using the Glauber and
the wounded nucleon models [18–20].Again the error
bars correspond to the statistical and to the fit uncer-
tainties. A 9.5% method uncertainty should be added
in quadrature, but it affects all the points coherently
and thus should be disregardedwhen considering only
the trend of the Npartdependence.The (φ/(ρ +ω))µµ
ratio exhibits a 70% increase when going from the
most peripheral to the most central collisions and flat-
tens for Npart? 250. The same behavior is observed
for each MT bin.
φ and (ρ + ω) multiplicities are measured for
the first time in the NA50 experiment. They allow
in particular to check that this increase comes from
an enhancement of the φ production. The φ meson
multiplicity is the number of φ mesons produced per

Page 6

152
NA50 Collaboration / Physics Letters B 555 (2003) 147–155
Fig. 4. The ratio (φ/(ρ + ω))µµas a function of the number of
participant nucleons (Npart) over the whole MT? 1.5 GeV/c2
domain.
Pb–Pb collision. It is calculated from the ratio of the
number of measured φ mesons, Nmeas
acceptance (Accφ), trigger (εtrig) and reconstruction
(εrec) efficiencies, to the number of minimum bias
Pb–Pb collisions:
φ
, corrected for
?Nµµ
φ
=Nµµ
φ(?ET)
Nmeas
φ
NM.B.(?ET)×εtrig×εrec.
On the left-hand side of Fig. 5, are shown the φ and
ρ+ω multiplicities in the µ+µ−channelvs. the num-
ber of participant nucleons. The error bars correspond
to the statistical and the fit uncertainties. A method
uncertainty—of 8.6% for the φ meson and 5.4% for
ρ + ω mesons—and a 2.3% systematic uncertainty
should be added respectively in quadrature and lin-
early. As these errors affect all the points coherently,
they should be disregarded when considering only the
trend of the Npartdependence. The multiplicities in-
crease as Npartincreases. To further quantify this in-
crease, the multiplicities are divided by the number of
=
(?ET)/Accφ
Fig. 5. Multiplicities (left panels) and multiplicities per participant nucleon (right panels), corrected for acceptance, for the φ and ρ +ω mesons
with MT> 1.5 GeV/c2. (The background being different in a minimum bias event and in a dimuon event, the first point has been omitted here
due to the contamination from Pb–air interactions in the minimum bias sample.)

Page 7

NA50 Collaboration / Physics Letters B 555 (2003) 147–155
153
Fig. 6. Fit of the cross-section distributions vs. MTfor φ and ρ +ω, for the whole ETdomain.
Table 1
Effective temperature values for the φ and ρ +ω mesons for each transverse energy bin (ET)
?ET(GeV) 5–2929–42
Tφ(MeV)218±6
Tρ+ω(MeV)221±4
42–53 53–6464–7474–83 83–9393–105
>105
229 ±7
221 ±5
225 ±7
224 ±5
225±7
223±7
231 ±7
230 ±6
229±6
223±6
229 ±6
225 ±7
231±8
232±7
234±7
221±6
participant nucleons, and plotted vs. Npart, as shown
on the right side of Fig. 5. The additional Npartuncer-
tainty, given by the half FWHM of the Npartdistribu-
tioninthecorrespondingcentralitybin,hasbeenintro-
duced in the figure. Within our errors, one can observe
a flat behaviorforthe ρ+ω multiplicity whennormal-
ized by Npart, while there is an increase of ∼ 1.7 of the
φ multiplicity which seems to flatten beyond Npart?
250. This increase of the φ meson multiplicity divided
by Npartis observed in each of the five MT bins.
The φ and ρ + ω production cross sections have
been measured through the dimuon decay channel
(Bµµ×σφ,ρ+ω).TheBµµ×dσφ,ρ+ω/MTdMTquan-
tities are plotted vs. MT on Fig. 6, for the whole ET
domain. The error bars include the statistical, the fit
and the method errors. A systematic uncertainty of
6.9% has to be added linearly. We choose to fit the
Bµµ× dσφ,ρ+ω/MTdMT spectra with MTe−MT/T,
as used by other experiments, which leads to the
inverse slope parameters Tφ? 228 ± 10 MeV and
Tρ+ω? 224± 10 MeV within the studied MT range.
When the same analysis is repeated in each ET bin,
the effective temperatures are found to be independent
of the centrality as can be seen in Table 1.
5. Discussion and conclusions
As previously observed in the NA38 experiment
in S–S, S–Cu and S–U interactions [6], the (φ/(ρ +
ω))µµ ratio increases with increasing centrality in
Pb–Pb reactions. The large data sample collected in
1996 for Pb–Pb collisions allows us to observe a
saturation effect of this ratio for Npart greater than
250. The φ meson multiplicity per participant nucleon
seems to display the same behavior.
This pattern is different from the one measured by
the WA97 experiment for multi-strange baryons for
which production yields saturate at around Npart?
100 [3].
On the other hand, the ratio (φ/(ρ + ω))µµ and
the φ meson multiplicity pattern display the same
behavior as the φ/π and K/π ratios measured by the
NA49 Collaboration [2,21], as well as the K/π ratio
at lower energies (at the AGS in the E866 experiment
[22]).
The φ effective temperature we obtain is very dif-
ferent from the value obtained by the NA49 exper-
iment which, moreover, varies slightly with central-
ity [2]. For the most central collisions, representing

Page 8

154
NA50 Collaboration / Physics Letters B 555 (2003) 147–155
Fig. 7. Effective temperatures vs. particle mass measured by several
experiments in Pb–Pb collisions at 158 GeV/c per nucleon at the
CERN/SPS.
5% of the total cross section, the effective temperature
measured by the NA49 experiment is 305 ± 15 MeV,
whereas we obtain 230 ± 10 MeV. The MT domains
covered by the two experiments are different, but even
in the common MTrange (1.5–2.4GeV/c2) the NA50
invariant slope 216± 15 MeV remains lower than the
NA49 one ? 290 MeV. It is also clear that the two ex-
periments do not measure the φ resonance throughthe
same decay channel, but this does not seem enough to
explain the discrepancy [10].
The effective temperature provides information on
the collective transverse flow and the conditions of
the system at the moment when the particles under
study cease interacting elastically. This is usually de-
scribed by 2 parameters, namely, the thermal freeze-
out temperature Tfo
th, and the mean velocity of the
transverse collective flow vT [23]. For low momenta,
Teffis expected to increase linearly with the particle
mass: Teff? Tfo
effective temperatures Teffvs. the mass of the partic-
ules as measured in Pb–Pb collisions by several exper-
imentsattheCERN/SPS, althoughnotinthesameMT
range. The Teff values corresponding to non-strange
and single-strange hadrons are well reproduced by the
straight line deduced from RQMD calculations [24].
The Ω and?
WA97 experiment[25]is clearly belowthe line. The φ
meson effective temperature extracted from our mea-
th+ 1/2mv2
T. In Fig. 7 are plotted the
Ω effective temperature obtained by the
surementis alsobelowtheline,evenifwe takeintoac-
countthe localslopevariationof ±25MeV onourMT
range. The Ω and?
of strange quarks (or antiquarks) only, have smaller
interaction cross sections in the hadron gas as com-
pared to other hadrons. These particles should decou-
ple earlier from the hadron gas, and therefore be less
affectedbythetransversecollectiveflow,whichwould
explain their low effective temperature [24]. The same
scenario could apply to the φ meson. This idea is sup-
ported by the fact that the MICOR model [26], in
which there are no interactions in the hadron gas, re-
produces the φ, Ω and?
tive temperatures.
Ω (anti)baryons, being composed
Ω, as well as the ρ + ω effec-
References
[1] B. Muller, J. Rafelski, Phys. Rev. Lett. 81 (1986) 1066.
[2] S.V. Afanasiev, et al., NA49 Collaboration, Phys. Lett. B 491
(2000) 59;
S.V. Afanasiev, et al., Nucl. Phys. A 698 (2002) 104c.
[3] F. Antinori, et al., WA97 Collaboration, Nucl. Phys. A 661
(1999) 130c.
[4] M.C. Abreu, et al., NA50 Collaboration, Phys. Lett. B 450
(1999) 456.
[5] M.C. Abreu, et al., NA50 Collaboration, Phys. Lett. B 477
(2000) 28.
[6] C. Baglin, et al., NA50 Collaboration, Phys. Lett. B 272 (1991)
449;
M.C. Abreu, et al., Phys. Lett. B 368 (1996) 239;
J. Astruc, Ph.D. Thesis, Université Paris 6—Pierre et Marie
Curie, 1997;
C. Quintans for the NA50 Collaboration, J. Phys. G: Nucl. Part.
Phys. 27 (2001) 405.
[7] N. Willis for the NA50 Collaboration, Nucl. Phys. A 661
(1999) 534;
M.P. Comets for the NA50 Collaboration, Nucl. Phys. A 663–
664 (2000) 721c;
L. Villatte for the NA50 Collaboration, XXXVIII International
Winter Meeting on Nuclear Physics, Bormio, Italy, supple-
mento No. 116 (2000) 241.
[8] L. Villatte, Ph.D. Thesis, Université Paris 7—Denis Diderot,
2001.
[9] A. Shor, Phys. Lett. 54 (1985) 1122.
[10] S.C. Johnson, B.V. Jacak, A. Drees, Eur. Phys. J. C 18 (2001)
645;
S. Soff, et al., J. Phys. G: Nucl. Part. Phys. 27 (2001) 449.
[11] M.C. Abreu, et al., NA50 Collaboration, Phys. Lett. B 410
(1997) 327.
[12] J.C. Collins, D.E. Soper, Phys. Rev. D 16 (1977) 2219.
[13] S. Papillon, Ph.D. Thesis, Université Paris-Sud, 1993.
[14] M.C. Abreu, et al., NA50 Collaboration, Euro. Phys. J. C 14
(2000) 443;

Page 9

NA50 Collaboration / Physics Letters B 555 (2003) 147–155
155
C. Soave, Ph.D. Thesis, Università degli Studi di Torino, 1998.
[15] A. Angelis, et al., HELIOS3 Collaboration, Eur. Phys. J. C 5
(1998) 63.
[16] R. Hagedorn, Rivista Nuovo Cimento 6 (1983) 10.
[17] M. Aguilar-Benitez, et al., LEBC-EHS Collaboration, Z. Phys.
C 50 (1991) 405.
[18] R.J. Glauber, High Energy Collision Theory, in: Lectures in
Theoretical Physics, Vol. I, Interscience, New York, 1959,
p. 315.
[19] A. Bialas, M. Bleszy´ nski, W. Czyz, Nucl. Phys. B 111 (1976)
461.
[20] D. Kharzeev, C. Lourenço, M. Nardi, H. Satz, Z. Phys. C 74
(1997) 307.
[21] C. Höhne for the NA49 Collaboration, Nucl. Phys. A 661
(1999) 485c.
[22] C.A. Ogilvie for the E866 and E917 Collaborations, Nucl.
Phys. A 638 (1998) 57c.
[23] N. Xu for the NA44 Collaboration, Nucl. Phys. A 610 (1996)
175c.
[24] H. van Hecke, H. Sorge, N. Xu, Phys. Rev. Lett. 81 (1998)
5764.
[25] F. Antinori, et al., WA97 Collaboration, Eur. Phys. J. C 14
(2000) 633.
[26] P. Csizmadia, P. Levai, Phys. Rev. C 61 (2000) 031903(R).