• No results found

STAR results from the first year at RHIC

N/A
N/A
Protected

Academic year: 2022

Share "STAR results from the first year at RHIC"

Copied!
12
0
0

Loading.... (view fulltext now)

Full text

(1)

—journal of April 2003

physics pp. 627–638

STAR results from the first year at RHIC

HELEN CAINES for the STAR Collaboration

Physics Department, Yale University, New Haven, CT 06520-8124, USA

Abstract. An overview of the latest results from the STAR experiment at RHIC is presented. Pre- liminary measurements ofπ;K;p;ΛandΞ, plus their respective anti-particles at pt<2 GeV/c, where the majority of particle production occurs, allow us to probe the soft processes whilst the harder per- turbative regime can be accessed by studying particle spectra and yields at higher momenta.

1. The STAR experiment and its physics goals

The solenoidal tracker at RHIC (STAR) is a large acceptance tracking detector designed primarily to measure hadronic particle production in high energy nuclear collisions. The ultimate goal of the RHIC experiments is to study the behaviour of strongly interacting matter under extreme conditions. It is hoped that approximately 7.5 times higher center of mass energy produced in RHIC collisions compared to the CERN SPS collider will allow us to establish the presence of a deconfined phase of quarks and gluons for which evidence has been reported by the CERN experiments [1]. The RHIC complex is capable of pro- viding heavy-ion and p–p collisions up tops

NN=200 GeV and STAR intends to explore how particle production and distributions scale with increasing collision energy and parti- cle species. At these high energies particle production at high transverse momentum (pt) is expected to become more pronounced. As these particles originate from hard processes, occurring during the early stages of the collisions, studies of their properties may offer new insight into the state of matter produced.

In the first year of data-taking the detector consisted of a 4 m long time projection cham- ber (TPC) placed within a 0.25 T solenoidal magnet and a ring imaging Cherenkov (RICH) detector as shown in figure 1. The RICH is centered at mid-rapidity and has a narrow ac- ceptance window ofjηj<0.3 and∆φ=20Æ. Collision centrality was determined via the correlation between mid-rapidity charged particles measured by the central trigger barrel (CTB) scintillator surrounding the TPC and two zero degree calorimeters (ZDCs), which detect neutrons, positioned18 m upstream of the TPC center along the beam direction.

The anti-correlation between mid-rapidity charged particles and spectator neutrons allows for centrality discrimination. Peripheral collisions produce fewer mid-rapidity charged par- ticles and a large signal in the ZDC’s while central collisions, being the most violent, cause copious charged particle production and leave few spectator neutrons to continue, undis- turbed, along the beam line. Approximately 700,000 Au–Au events (central and min-bias) were recorded and analyzed during this period with a beam energy ofps

NN=130 GeV.

(2)

ZDC ZDC TPC

CTB

Magnet

Liz

Figure 1. The STAR detector as instrumented in the year 2000,psNN=130 GeV, Au–Au running. The TPC is 4 m in length and the ZDCs are positioned18 m up- stream of the TPC center.

2. The analysis

Charged particle identification was provided via the measurement of specific ionization (dE=dx) of the TPC gas along the charged particle trajectories. The difference in dE=dx for different mass particles with the same momenta allows separation of kaons and pions up to momentum, p0:6 GeV/c and pions and protons up to p1:0 GeV/c. The RICH detector augmented this charged particle identification, through the measurement of the angle of the emitted Cherenkov light, at higher momenta extending the momentum range of identified kaons to 3 GeV/c and protons to 5 GeV/c. The weak decay topology of strange particles, into charged daughters, is also used to reconstruct neutral and charged species out to high momenta. By locating secondary vertices, away from the primary vertex, it is possible to deduce the mass and momentum of the parent particle from the daughter momenta at the decay point. The momentum limit for these measurements is due to the finite momentum resolution of the TPC tracking preventing identification of the secondary vertex at very high momenta.

Detector inefficiencies were estimated via embedding simulated tracks, using a detailed simulation of the detector response, into real events. Events were then analyzed using the same reconstruction algorithms as for the real data and an association code was used to determine the reconstruction efficiency of the embedded particles. This correction factor therefore included both the detector acceptance and efficiency. For primary particles the combined efficiency and acceptance was found to range from 65–90% depending on pt. The data have also been corrected for background resulting from interactions in the detector material, secondary decay products and mis-identified particles.

3. Results

3.1 Baryon number and chemistry

One of the first points of interest in studying heavy-ion collisions is the amount of baryon stopping/transport to the mid-rapidity region. As the colliding nuclei consist of only light

(3)

quarks it is possible to probe the net baryon number in the collision region by measuring the anti-baryon/baryon ratio. Figure 2 shows preliminary ¯B=B ratios for the top 11% most central collisions. Also shown are the reported values of the respective ratios measured in Pb–Pb collsions at ps

NN =17 GeV at the SPS [2,3]. It can be seen that while the ratios are significantly higher than the corresponding ratios at the SPS they are not yet unity. This means that even at these higher energies either some baryons are still being transported from the colliding nuclei to the mid-rapidity region or there is significant anti- particle annihilation in the fireball. Exactly how the baryon number is transported over 5 units of rapidity or how such a significant annihilation occurs is still not theroetically understood. As the light quarks in the baryons are replaced by strange quarks the ratio increases as expected. Since no strange quarks exist in the colliding nuclei all strange quarks are produced in s ¯s pairs resulting in a ¯s=s ratio of 1.

Figure 3 shows the ¯p=p ratio as a function of pt. The first point is from the top 14%

central events for (anti-)protons identified via dE=dx in the TPC [4]. The RICH was then used to extend this measurement out to 2.5 GeV/c for events of the same centrality. The error bars indicate the statistical uncertainties and the brackets show the estimated system- atic uncertainties. It can be seen that the ratio remains constant over the measured ptrange.

This is in contradiction with QCD predictions [5] which show a strongly decreasing ¯p=p ratio at high pt. This predicted drop is caused mainly by the difference in (anti-)proton production mechanisms. Most anti-protons are created by gluon fragmentation while the protons come from a mixture of valence quarks and gluon fragmentation. The proton par- ton distributions show that the gluon-to-quark density decreases with the Etof produced jets. As jet Ettranslates directly into particle pt, the ratio of particles produced via gluon

Figure 2. Preliminary anti-baryon to baryon ratios at mid-rapidity measured by STAR for the top 11% most central events. For comparison measurements made at the SPS are included. Errors are statistical only.

(4)

[GeV/c]

0 1 2 P 3

/p ratiop

0 0.2 0.4 0.6 0.8 1

=130 GeV sNN

Central Au-Au,

STAR TPC, |y|<0.3 STAR-RICH, η<0.15

Preliminary

Figure 3. Preliminary ¯p=p ratio as a function of pt. The first point is the value calcu- lated for particles identified via dE=dx measurements in the TPC. The remaining points are from particles identified using RICH.

fragmentation to those produced from valence quarks should also drop. Hence it is pre- dicted that the measured ¯p=p ratio should change significantly at high pt.

As the system cools it first passes through chemical freeze-out. This is defined as the temperature at which inelastic collisions cease and the chemical content of the fire-ball is determined. Various models have been used to estimate the chemical freeze-out tem- perature, Tch, and the light and strange quark chemical potentials, µqandµs (e.g. [6], [7]). While these models differ quite significantly in their detailed assumptions they all re-produce the measured particle ratios, up to and including those using singly strange baryons, and are all in agreement that Tch175 MeV andµq15 MeV. In the modelsµs

is either assumed to be 0 or is calculated to be so within errors.

3.2 Particle multiplicity and spectra

The preliminary dN=dy forπ s for the top 5% most central events is shown in figure 4 [8], error bars are statistical only. The solid circles and squares are the yields obtained when the data are summed over the measured ptregions of 0.2–0.6 GeV/c and 0.05–0.75 GeV/c respectively. The hollow circles are the calculated yield when the distributions are extrapolated over all pt. The function used for extrapolation was the Bose–Einstein distri- bution, dN=dy=A=(emt=Teff 1), where A and Teffare free parameters. The solid circles show that the yield in this region is relatively independant of rapidity. However, a sizable extrapolation must then be made to extract the total yield. The validity of this extrapolation is confirmed via the solid squares. Over this more limited phase space in rapidity the STAR detector has acceptance for a much larger slice of pt. Approximately 85% of the yield is then measured, greatly reducing the possible error in the extrapolation procedure. Yields and slopes obtained, where possible, via fits to both ptranges were the same within errors.

(5)

y

-1 -0.5 0 0.5 1

dN/dy-π

0 50 100 150 200 250 300 350

< 0.6 GeV/c 0.2 < p

< 0.75 GeV/c 0.05 < p

Extrapolated

STAR preliminary Au+Au, sNN = 130 GeV

5% Most Central

Figure 4. Preliminaryπ dN=dy. The solid circles (squares) show the measured yields summed between pt= 0.2–0.6 GeV/c (0.05–0.75). The open circles are the yields obtained by extrapolating the data over all pt.

The fact that the distribution is flat as a function of rapidity, over the region measured, could be taken as an indication of boost-invariance in the system. This is important when compared to most models, where boost-invariance at mid-rapidity is assumed. However a closer look at the extracted inverse slopes, Teff, of the ptspectra, (figure 5) indicates that afterjyj>0:5 the inverse slopes show a significant decrease. Therefore, while the mid- rapidity yield is a constant over one unit of rapidity, the ptdistributions are varying and thus the system is no longer consistant with boost invariance beyondjyj>0:5.

3.3 Thermalization and hydrodynamics

The high number of particles created in these collisions naturally leads to the question: ‘Is the system dense enough to allow hydrodynamical models to become applicable?’ One of the predictions of the hydrodynamic models is a large elliptic flow, which increases with decreasing centrality. As collisions get more peripheral the initial anisotropy of the partial overlap of the colliding nuclei increases. This spatial anisotropy leads, in turn, to a momentum anisotropy in the final state, caused, it is believed, by pressure gradients built up in the early stages of the collision. The azimuthal distributions of the emitted particles are measured with respect to the event plane, determined on an event-by-event basis using low pttracks in the TPC [9]. Figure 6 shows the second Fourier coefficient, v2, of the azimuthal distribution of all charged particles as a function of ptintegrated over all centralities. The data rise, as predicted by hydrodymanics [10], as a function of pt until2 GeV/c. The data then appear to flatten and are systematically below the prediction, shown by curves in figure 6. This deviation from hydrodynamical behavior at high ptwas not unexpected. Fast moving particles may suffer very little re-scattering before leaving the collision zone, and thus have no chance to build up a v2signal. A complete lack of re-scattering would result,

(6)

y

-1 -0.5 0 0.5 1

(MeV)eff T-π

0 50 100 150 200 250 300

STAR preliminary, Au+Au, √sNN = 130 GeV Rapidity Dependence

Slope Parameter π-

Figure 5. Preliminaryπ inverse slopes as a function of rapidity for the 5% most cen- tral events. The error bars are the uncorrelated point-to-point systematic uncertainties on Teff, the overall correlated systematic uncertainty is20 MeV.

[GeV/c]

p

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

)(p2v

0 0.05 0.1 0.15 0.2 0.25

0.3 Charged particles Hydro pions Hydro charged particles Hydro anti-protons

Preliminary

Figure 6. The v2 as a function of pt for charged particles. Also shown are hy- dro-dynamical predictions for charged particles, pions and anti-protons. Error bars are statistical only.

however, in no anisotropic flow and it has therefore been suggested that the observation of significant v2for high ptparticles may be related to in-medium energy loss. This energy loss would also result in an anisotropy in momentum when an initial spatial asymmetry occurs. Calculations have shown that v2measurements at high ptmay therefore allow us to probe the initial gluon density [11].

Transverse mass distributions (mt=

p

p2t +m2) meanwhile, allow us to investigate the degree of thermalization in the system, the thermal freeze-out temperature of each parti- cle species, and the degree of radial flow. Thermal freeze-out is defined as the point at which elastic collisions cease. If there is no radial flow in the system, and it is in thermal equilibrium, the inverse slope of a fit to an exponential gives the temperature, Tth, of the

(7)

2) (GeV/c -mΞ

mT

0 0.5 1 1.5 2 2.5

ydTmdΞN2d evtsNT mπ21

10-3 10-2 10-1

: dN/dy = 3.07 +/- 0.13 Ξ

T = 346 +/- 7 (MeV)

: dN/dy = 2.63 +/- 0.12 Ξ

T = 344 +/- 7 (MeV)

Ξ 0.5 *

STAR Preliminary 0.5 < pt < 3.5 GeV

spectra mT

Ξ &

Ξ corrected

Figure 7. Preliminary mt spectra forΞ (triangles) and ¯Ξ+ (circles) for 14% most central events. The solid curves are exponential fits to the data. Error bars are statistical only.

system. The addition of radial flow,βr, complicates the picture and the inverse slope, Teff, becomes a combination of freeze-out temperature and the radial flow. For lowβr, it has been shown that there is an approximately linear increase in Teffas a function of mass [12].

The transverse mass distributions forΞ and ¯Ξ+are shown in figure 7 for the top 14% cen- tral data. The solid lines represent the results of an exponential fit. Mid-rapidity yields of dN=dy=3:070.13 forΞ and dN=dy=2:60.12 for ¯Ξ+are obtained. It is estimated that the systematic errors are20%. The data have not been corrected for feed-down but this is expected to be a small effect. Inverse slopes of T =3467 and 3447 MeV are obtained for theΞ and ¯Ξ+respectively. This similarity of the inverse slopes extracted for each particle and its anti-particle is repeated for all species measured by STAR. A sum- mary of the Teffextracted from exponential fits to the transverse mass distributions is shown in figure 8 and compared to similar fits applied to SPS data [13,14]. All the STAR data have been fit over similar mt m0ranges. Two features immediately become apparent: (1) The RHIC data is systematically higher than the SPS for all particles masses, indicating a higher radial flow in 200 GeV/c Au–Au collisions. (2) The flattening of the inverse slopes for multi-strange particles first observed at the SPS, continues at RHIC. This behaviour has been attributed to a lower re-scattering cross-section for strange particles, which are therefore less affected by radial flow. However, this large flow velocity leads to a curvature in the transverse mass distribution and an exponential fit is no longer ideal. In fact the ex- tracted inverse slope becomes dependant on the fit range chosen. This effect is dramatically apparent in the extracted inverse slope, of 565 MeV, for the STAR ¯p measurement [15].

These particles are identified via dE=dx and are hence restricted to a much lower mtrange than those particles plotted in figure 8.

It is therefore desirable to move away from these models in order that we may use the full data set available to STAR. The success of hydrodynamics at fitting the v2measurements leads us to explore the applicability of a hydrodynamically inspired model [16]. One of the main features of this model is that the radial flow is described as a function of the radial distance, r, from the centre of the fireball. The STAR data is well-described when a velocity profile ofβr=βs

p

r=R is used, whereβsis the surface velocity at the surface

(8)

2 ) Mass (GeV/c 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Inverse slopes (MeV)

0 100 200 300 400 500 600 700

STAR Preliminary NA44 WA97 NA49

π K pφ Λ Ξ Ω

Figure 8. Preliminary inverse slope measurements as a function of particle mass. Also shown are SPS measurements. The error bars on the STAR data include an estimate of the systematic uncertainties.

0 2

K- P π-

Λ_ _

mt [GeV/c]

Preliminary

- m0

[A.U]

Figure 9. Preliminary fits of the mtspectra for anti-protons, kaons, pions and ¯Λs to a hydrodynamically inspired model, as explained in the text. The fit ranges used are shown by the solid curves and the results of the fit by the dashed curves.

radius R. This model, as shown in figure 9, proves very successful at describing the shape of the particle mtspectra and a combined fit to anti-protons, kaons, pions and ¯Λs shows a common freeze-out temperature of Tth=120+5020MeV and an average radial flow velocity ofhβri=0:52+00:12

:08c. Figure 10 shows the extracted Tthandhβrias a function ofps

NN[17].

It can be seen that there appears to be an asymptote in thermal freeze-out temperature that

(9)

Figure 10. Tthandβras a function of collision energy as calculated via the hydrody- namically inspired model described in the text.

is reached at around AGS energies but that the radial flow continues to increase up to RHIC energies. It would appear that the system cools until a certain temperature is reached, at which point the particle cross-sections are sufficiently low that thermal freeze-out occurs.

The extra energy provided by the RHIC collisions is converted into increasing the radial boost velocity.

3.4 Perturbative regime

Figure 11 shows the corrected inclusive negative hadron transverse momentum distribution when the 5% most central collisions were selected. Also shown is the negative hadron distribution from Pb–Pb collisions at ps

NN=17 GeV from the NA49 collaboration [18]

and the averaged charged hadrons, (h +h+)=2, from ¯p–p collisions at

p

s=200 GeV of AU1 [19]. It can be seen that the STAR data is flatter than the Pb–Pb data resulting in an increase in the mean pt of 18%. A yield of dN=jη=0=2801 (stat)20 (syst), estimated by extrapolating the measured spectrum via a power-law fit [20], is an increase of 52% compared to that reported by NA49.

In order to simply compare the Au–Au data to the scaled ¯p–p data, figure 12 shows the ratio of the STAR data to a scaled parameterization of the UA1 data. This parameteriza- tion tries to take into account the energy difference in the nucleon–nucleon center-of-mass frame [20] between the two data sets and the nuclear overlap integral [21],TAA= 26 mb 1. The horizontal lines shown in figure 12 correspond to the expected value of the ratio under two scaling assumptions. The upper line is the expected value if the Au–Au distribution is a result of the number of binary collisions; due to the scaling of the UA1 data this value is forced to be unity. The lower line corresponds to the expected value of the ratio if the Au–Au collisions scale with the number of wounded nucleons. Particle production is dominated by soft processes at low ptand, as expected, the total pion yield scales with the number of wounded nucleons in the lowest measured ptbin. Below pt=2 GeV/c, the ratio exceeds the wounded nucleon scaling, rising steadily towards the binary collision scaling.

However the ratio never reaches this limit but begins to drop again after pt>2 GeV/c to below 0.5 at pt=6 GeV/c. However, this hard process upper limit is only valid in the absence of nuclear effects, such as collective radial flow, initial state multiple scattering or

(10)

[GeV/c]

p

0 1 2 3 4 5 6

]-2 [ (GeV/c) = 0η)η d/ dp-hN2 d) (1/pevent(1/N

10-6 10-5 10-4 10-3 10-2 10-1 1 10 102 103

= 130 GeV sNN

, Au+Au, STAR h-

= 17.2 GeV sNN

, Pb+Pb, NA49 h-

= 200 GeV s , p )/2, p+

+h+

UA1 (h-

Preliminary

Figure 11. Preliminary negative hadron ptdistribution from the 5% most central data.

Also shown is the same distribution measured atps

NN

=17 GeV in Pb–Pb collisions and the average charged hadron yield from ¯p–p at

p

s=200 GeV.

[GeV/c]

0 1 2 3 4 p5 6

Ratio

0 0.2 0.4 0.6 0.8 1

TAA

× R(130/200)

× (UA1-fit) σ/dp d

(STAR) dN/dp

Binary Collisions Scaling

Wounded Nucleon Scaling

Preliminary

Figure 12. Preliminary ratio of STAR to scaled UA1 ptspectra. The vertical error bars indicate the measurement error, the shaded region the total uncertainty including that of the scaling factors applied to the UA1 data.

(11)

jet-quenching, none of which are present in elementary collisions but for which there is already evidence, reported here, in the RHIC Au–Au collisions.

4. Summary

In summary, STAR has measured a wide range of hadronic probes in both central and min-bias Au–Au collisions at ps

NN

=130 GeV. The pseudorapidity density of negative hadrons in central collisions represents a 50% increase in particle production compared to lower energy results from the CERN SPS.

Studies of the transverse mass distributions indicate a large collective radial flow, which complicates the interpretation of these distributions. Use of a hydrodynamically inspired model seems to reproduce the data and indicates a thermal freeze-out temperature similar to that seen at the CERN SPS and the AGS at BNL.

Statistical thermal models are also able to reproduce the measured particle ratios and imply that the chemical freeze-out of the system occurs at Tch175 MeV. The low, but non-zero, light quark chemical potential suggests there is some transport of baryon number from the colliding nuclei to the mid-rapidity region. The suitability of thermally equili- brated models and the similarity of particle and anti-particle distributions indicates a large amount of re-scattering in the system before thermal freeze-out and the magnitude of el- liptic flow implies that this thermalization occurs early.

One intriguing possibility deserving further study is that the RHIC collisions may be displaying evidence of partonic energy loss in the medium. This idea comes from com- bining the observations of figure 12 with that of figure 6, namely apparent v2saturation at high ptand the suppression of particle production at high ptwhen compared to scaled ¯p–p collisions.

This initial probe of Au–Au collisions has yielded a vast amount of information much of which is still being explored. The completion of this analysis and the first results from the latest Au–Au and p–p runs atps

NN

=200 GeV are eagerly awaited.

Acknowledgments

We wish to thank the RHIC Operations Group and the RHIC Computing Facility at Brookhaven National Laboratory, and the National Energy Research Scientific Comput- ing Center at Lawrence Berkeley National Laboratory for their support. This work was supported by the Division of Nuclear Physics and the Division of High Energy Physics of the Office of Science of the U.S. Department of Energy, the United States National Science Foundation, the Bundesministerium fuer Bildung und Forschung of Germany, the Institut National de la Physique Nucleaire et de la Physique des Particules of France, the United Kingdom Engineering and Physical Sciences Research Council, Fundacao de Amparo a Pesquisa do Estado de Sao Paulo, Brazil, the Russian Ministry of Science and Technology and the Ministry of Education of China and the National Science Foundation of China.

References

[1] U W Heinz and M Jacob, (2000) nucl-th/0002042 [2] E Anderson et al, J. Phys. G25, 171 (1999)

(12)

[3] M Kaneta et al, J. Phys. G23, 1865 (1997) [4] C Adler et al, Phys. Rev. Lett. 86, 4778 (2001) [5] X N Wang, Phys. Rev. C58, 2321 (1998)

[6] P Braun-Munzinger et al, Phys. Lett. B465, 15 (1999) [7] J Solfrank et al, Phys. Rev. C59, 1637 (1999)

[8] M Calderon de la Barca Sanchez, Ph.D Thesis, Yale University (2001) [9] K H Ackermann et al, Phys. Rev. Lett. B86, 402 (2001)

[10] P Huovinen et al, Phys. Lett. B503, 58 (2001)

[11] M Gyulassy, I Vitev and X N Wang, Phys. Rev. Lett. 86, 2537 (2001) [12] I G Bearden et al, Phys. Rev. Lett. 78, 2080 (1997)

[13] F Antorini et al, Europhys. J. C14, 633 (2000) [14] S V Afanasev et al, Phys. Lett. B491, 59 (2000) [15] C Adler et al, Phys. Rev. Lett. 87, 262302 (2001) [16] E Schnedermann et al, Phys. Rev. C48, 2462 (1993) [17] N Xu and M Kaneta, Nucl. Phys. A698, 306 (2002) [18] H Appelshauser et al, Phys. Rev. Lett. 82, 2471 (1999) [19] C Albajar et al, Nucl. Phys. B355, 261 (1990) [20] C Adler et al, Phys. Rev. Lett. 87, 112303 (2001)

[21] K J Eskola, K Kajantie and J Lindfors, Nucl. Phys. B323, 37 (1989)

References

Related documents

The results are discussed in terms of the bulk properties of the hadron production (transverse momentum spectra and particle ratios), collective phenomena exhibited by the

In this energy region, hadron transverse mass spectra first show soften- ing until SPS energies, and re-hardening may emerge at RHIC energies.. Since hadronic matter is expected to

Beam Energy = 100

Second, total particle production in AA collisions per participant pair is the same as the total particle production in e e at the same energy for a broad range of centralities,

The charged particle rapidity density at mid-rapidity has been measured by the four RHIC experiments – BRAHMS, PHENIX, PHOBOS and STAR – with very good agree- ment among their

With higher statistics measurements in the year 2001 at the full RHIC energy of p s NN = 200 GeV, a measurement of the η meson will also be performed, providing the basis for

ALICE (a large ion collider experiment) is the only detector specifically designed for the physics of nuclear collisions at LHC, even though it can also study high

This variance ( var = σ 2 ) is observed (figure 9c) to decrease as a function of h N part i for the 130 GeV data (solid circles) reflecting the fact that the increased