nucl-ex0409023/bmw.tex
1: \documentclass[a4paper]{jpconf}
2: \usepackage{graphicx}
3: 
4: \def\Journal#1#2#3#4{{#1}{\bf #2}, #3 (#4)}
5: 
6: \def\EPJC{{Eur. Phys. J.}~{\bf C}}
7: \def\IJMPA{{Int. J. Mod. Phys.}~{\bf A}}
8: \def\JPG{{J. Phys}~{\bf G}}
9: \def\NCA{Nuovo Cimento}
10: \def\NIM{Nucl. Instrum. Methods}
11: \def\NIMA{{Nucl. Instrum. Methods}~{\bf A}}
12: \def\NPA{{Nucl. Phys.}~{\bf A}}
13: \def\NPB{{Nucl. Phys.}~{\bf B}}
14: \def\PLB{{Phys. Lett.}~{\bf B}}
15: \def\PLC{Phys. Repts.\ }
16: \def\PRL{Phys. Rev. Lett.\ }
17: \def\PRD{{Phys. Rev.}~{\bf D}}
18: \def\PRC{{Phys. Rev.}~{\bf C}}
19: \def\ZPC{{Z. Phys.}~{\bf C}}
20: \def\ZPA{{Z. Phys.}~{\bf A}}
21: \newcommand \etaL{ {\it et al.} } 
22: 
23: \newcommand \mt{$\langle m_{T}\rangle$ }
24: \newcommand \Nch{$N_{ch}$ }
25: \newcommand \Et{$E_{T}$ }
26: \newcommand \EN{$E_{T}/N_{ch}$ }
27: \newcommand \dNch{$dN_{ch}/d\eta$ }
28: \newcommand \dEt{$dE_{T}/d\eta$ }
29: \newcommand \Np{$N_{p}$ }
30: \newcommand \Nps{$N_{p}$}
31: \newcommand \sqn{$\sqrt{s_{_{NN}}}$ }
32: \newcommand \sqns{$\sqrt{s_{_{NN}}}$}
33: \newcommand \1{19.6~GeV}
34: \newcommand \2{200~GeV}
35: \newcommand \3{130~GeV}
36: \newcommand \6{62.4~GeV}
37: \newcommand \7{17.2~GeV}
38: 
39: \begin{document}
40: 
41: \title{Centrality and \sqn Dependence of the \dEt and \dNch in Heavy Ion Collisions at Mid-rapidity}
42: 
43: \author{A.~Milov for the PHENIX Collaboration\footnote{For the full list of authors and acknowledgments see reference~\cite{ppg19}}}
44: \address{Department of Physics and Astronomy, Stony Brook University, SUNY, Stony Brook, NY 11794, USA}
45: 
46: \begin{abstract}
47: The PHENIX experiment at RHIC has measured transverse energy and charged particle multiplicity 
48: at mid-rapidity in $Au+Au$ collisions at \sqn = 19.6, 130, 62.4 and 200 GeV as a function of centrality.
49: The presented results are compared to measurements from other RHIC experiments, and experiments at 
50: lower energies. The \sqn dependence of \dEt and \dNch per pair of participants is 
51: consistent with logarithmic scaling for the most central events. The centrality dependence of 
52: \dEt and \dNch is similar at all measured incident energies. At RHIC energies the ratio of transverse 
53: energy per charged particle was found independent of centrality and growing slowly with \sqns. 
54: A survey of comparisons between the data and available theoretical models is also presented.
55: \end{abstract}
56: 
57: \section{Introduction}
58: The PHENIX experiment at the Relativistic Heavy Ion Collider (RHIC) at Brookhaven National Laboratory was designed to
59: measure the properties of matter at extremely high temperatures and densities. Under such conditions, there is possibility 
60: to produce states of matter that have yet to be observed or studied in the laboratory. The best known
61: of these is the {\em quark-gluon plasma} (QGP), a form of matter where quarks are not confined within individual baryons
62: but exist as some form of plasma of individual quarks and gluons. It should be emphasized that the exact properties of
63: this matter are not known and that the characterization of the deconfined state, if such a state is produced, is
64: an essential part of the RHIC program.
65: 
66: One fundamental element of the study of ultrarelativistic collisions is the characterization of the interaction in terms
67: of variables such as the energy produced transverse to the beam direction and the number of charged particles. These 
68: two variables are closely related to the collision geometry and are important in understanding global properties of the 
69: system during the collision.
70: 
71: This paper describes a systematic study of \dEt and \dNch at mid-rapidity\footnote{Also referred as \Et and \Nch 
72: in this paper.} by the PHENIX experiment at center-of-mass 
73: energies \sqns~=~19.6, 130, 62.4 and \2. The centrality dependence of \Et and \Nch is characterized by the number of 
74: participants, determined with a Glauber model, and is studied as function of the incident energy.
75: \Et and \Nch results for all four RHIC measurements are included as part of this study.
76: The data taken at 19.6 GeV is particularly interesting because it allows a close comparison to the lower energies 
77: of the CERN SPS program. Comparisons are also made to previous experiments at the Brookhaven AGS 
78: and CERN SPS at center-of-mass energies of 4.8 GeV, 8.7 GeV, and 17.2 GeV. Finally, an extensive set of collision models
79: describing the \Et and \Nch distributions are compared to the existing data.
80: 
81: \section{PHENIX detector and Analysis\label{sec:dtector}}
82: PHENIX is one of four experiments located at RHIC~\cite{rhic}. 
83: The PHENIX detector consists of two central spectrometer arms, 
84: designated east and west for their location relative to the interaction region,
85: and two muon spectrometers, similarly called north and south.  
86: Each central spectrometer arm covers a rapidity range of $|\eta|<0.35$ and subtends $90^{0}$ in azimuth. 
87: The  muon spectrometers both have full azimuthal coverage with a rapidity ranges of 
88: $-2.2<\eta<-1.2$ (south) and $1.2<\eta<2.4$ (north). Additional global 
89: detectors are used as inputs to the trigger and for global event characterization 
90: such as vertex, time of event and centrality determination.
91: A detailed description of the PHENIX detector can be found in~\cite{phenix}. 
92: 
93: The PHENIX detector subsystems relevant for the physics analysis published here are:
94: the Pad Chambers used for the charged particle multiplicity measurement, the 
95: Electromagnetic Calorimeter used to measure transverse energy, the Beam-Beam Counter 
96: and the Zero Degree Calorimeter, the two detectors used for triggering and centrality 
97: determination are described in publications~\cite{pc,emcal,bbc-zdc,zdc}.
98: 
99: The analysis procedures to measure \Et and \Nch are described in details in 
100: publications~\cite{phenix_milov,phenix_bazik,david_thesis,sasha_thesis}. Some additional
101: information concerning the analysis at \sqns~=~\1 can be found in~\cite{ppg19}. The preliminary 
102: results on \dNch at \sqns~=~\6 use the same technique as implemented at other energies. 
103: 
104: For the sake of space we only mention errors relevant to the analyses. Statistical errors are 
105: small and do not exceed 1\% of the measured value. The systematic errors are summarized in 
106: Table~\ref{tab:errors}. The errors for the lowest and the highest incident energy are listed, 
107: whereas at all other energies the are close to these values. The systematic errors 
108: for both measured values are of two types. The first type affects the centrality 
109: dependence. It is listed in the Table~\ref{tab:errors} with the range 
110: (first number corresponds to the most central bin). Errors of the second type contribute 
111: to the overall scaling of the data. In the figures below the centrality dependent errors are 
112: shown as a corridor, and the the full systematic error is shown with the 
113: error bars\footnote{Here and everywhere errors correspond to one standard 
114: deviation.}. Centrality related errors are common to both measured quantities.
115: 
116: \begin{center}
117: \begin{table}[h]
118: \caption{Summary of systematic errors given in \%. When the range is given, 
119: the first number corresponds to the most central bin and the second to the 
120: most peripheral bin.
121: \label{tab:errors}}
122: \centering
123: \begin{tabular}{lcccc}
124: \br
125: \multicolumn{1}{c}{}&\multicolumn{2}{c}{\dEt}  & \multicolumn{2}{c}{\dNch}\\
126: \sqn [GeV]          &  19.6    & 200    & 19.6     & 200     \\
127: \mr
128: Energy resp.        &  4.7     & 3.9    &          &         \\
129: Bkg. / noise        &  0.5-3.5 & 0.2-6  & 1.0      & 1.0     \\
130: Acceptance          &  2.0     & 2.0    & 2.3      & 2.3     \\
131: In- \& outflow      &  3.0     & 3.0    & 5.7      & 2.9     \\
132: Occupancy           &          &        & 1.6-0.3  & 3.5-0.1 \\
133: \hline			       					
134: Centrality          & 2.0      & 0.5    & \multicolumn{2}{c}{same} \\
135: \Np                 & 2.9-6.7  & 2.8-15.& \multicolumn{2}{c}{same} \\
136: Trigger             & 0.4-8.8  & 0.3-16.& \multicolumn{2}{c}{same} \\
137: \br
138: \end{tabular}
139: \end{table}
140: \end{center}
141: 
142: \section{Results \label{sec:result}}
143: \subsection{PHENIX results}
144: The distribution of the raw transverse energy, $E_{T_{EMC}}$, into the fiducial 
145: aperture of two EMCal sectors is shown in the left three panels of 
146: Fig.~\ref{fig:results_raw} for three RHIC energies \sqn = 19.6, 130 and \2. 
147: The lower scale corresponds to the fully corrected \Et normalized to one unit of 
148: pseudorapidity and full azimuthal acceptance. The lower axis in the plot is not labeled beyond \2 
149: to avoid confusion between the true shape of the \dEt distribution 
150: and \Et as measured using the limited acceptance of two EMCal sectors.
151: 
152: \begin{figure}[h]
153: \includegraphics[width=0.49\linewidth]{etra_raw}
154: \includegraphics[width=0.49\linewidth]{mult_raw}
155: \caption{The distribution of the raw \Et in two EMCal sectors (left) and the number of 
156: tracks in the east arm of the PHENIX detector (right) per MB trigger, measured 
157: at three energies. The lower axis corresponds to mid-rapidity values of 
158: \dEt and \dNch respectively. Distributions of the four 5\% most central bins 
159: are also shown in each plot.\label{fig:results_raw}}
160: \end{figure}
161: 
162: Two EMCal sectors each with azimuthal coverage $\Delta \phi = 27^{o}$ 
163: were used to make a measurement at \sqn = \3. At other energies we used five EMCal sectors.
164: Results obtained with different numbers of sectors at the same energy are consistent within 1.5\%.
165: 
166: The right three panels in Fig.~\ref{fig:results_raw} show the number of tracks 
167: reconstructed in the east arm of the PHENIX detector after the background 
168: subtraction and all other corrections. The lower axis corresponds to measured 
169: distributions normalized to one unit of pseudorapidity and full azimuthal 
170: acceptance. For a similar reason as for the \Et, the lower axis is not labeled above \2 
171: in $dN_{ch}/d\eta$.
172: 
173: For the \Nch measurements at \sqn = \3, only the east arm was used, while for the other 
174: energies the measurements were made using both PHENIX central arms. The 
175: results obtained with two arms at \sqn= 200, 62.4, and \1 are consistent with each 
176: other within 1.5\%.
177: 
178: The distributions shown in Fig.~\ref{fig:results_raw} have a characteristic shape 
179: with a sharp peak that corresponds to the most peripheral events. Missing events caused by the finite 
180: MB trigger efficiency in peripheral events 
181: would make this peak even sharper than measured. 
182: The plateau in all distributions corresponds to mid-central 
183: events and the fall-off to the most central $Au+Au$ events. The shape of the 
184: curves in Fig.~\ref{fig:results_raw} in the fall-off region is a product of the 
185: intrinsic fluctuations of the measured quantities and the limited acceptance of the detector.
186: 
187: The distributions for the four most central bins 0\%-5\% to 15\%-20\% are also shown 
188: in each panel. The centroids of these distributions are used to calculate the 
189: centrality dependence of \Et and $N_{ch}$~\footnote{All plotted and 
190: quoted numbers correspond to the centrality bin-by-bin average values or ratios 
191: of corresponding averages.}. The statistical uncertainty of the mean values (less than or 
192: about 1\%) determined by the width of the distributions are small because of 
193: the large size of the event samples.
194: 
195: The magnitude of \dEt at mid-rapidity 
196: divided by the number of participant pairs as a function of \Np is shown in Fig.~\ref{fig:results_vs_npart}.
197: The right three panels show the same ratio for \dNch at three RHIC energies.
198: \begin{figure}[h]
199: \includegraphics[width=0.49\linewidth]{etra_vs_npart}
200: \includegraphics[width=0.49\linewidth]{mult_vs_npart_bmw}
201: \caption{\dEt (left) and \dNch (right) divided by the number of participant pairs at 
202: different RHIC energies. Errors shown with vertical bars are full systematic 
203: errors. Lines show the part of the systematic error that allows bending or 
204: inclination of the points. Horizontal errors denote the uncertainty in 
205: determination of \Nps. PHENIX preliminary result for \Nch at \sqns~=~\6 is shown 
206: in the right panel.\label{fig:results_vs_npart}}
207: \end{figure}
208: 
209: The horizontal errors correspond to the uncertainty in \Nps, determined within 
210: the framework of the Monte Carlo Glauber model. 
211: The vertical bars show the full systematic errors of the 
212: measurements added quadratically to the errors of \Nps. 
213: The lines denote the corridor in which the points can be inclined or bent. 
214: The statistical errors are smaller than the size of the markers. The upper panel 
215: also shows the results of the two lower panels with open markers for 
216: comparison. 
217: 
218: An important result for Fig.~\ref{fig:results_vs_npart} is the consistency evident 
219: in the behavior of the centrality curves of \Et shown on the left 
220: and \Nch shown on the right for all measured energies. Both values 
221: demonstrate an increase from peripheral (65\%-70\%) to the most central events by (50-70)\%
222: at RHIC energies \sqns=\3 and \2. For the lowest RHIC energy (\sqns=\1) this 
223: increase is at the level of systematic uncertainties of the measurement. 
224: One can note that results from PHOBOS~\cite{phobos_total},
225: show that the total charged particle multiplicity is proportional to \Np
226: while the multiplicity at mid-rapidity over \Np increases with \Nps,
227: indicating that the pseudorapidity distribution gets more narrow for
228: central events. Figure~\ref{fig:results_vs_npart} also show the PHENIX preliminary result for 
229: \Nch at \sqns~=~\6 added to the central panel.
230: 
231: The ratios of the \dEt and \dNch per participant pair measured at different 
232: RHIC energies are shown in Fig.~\ref{fig:ratios_vs_npart}. 
233: In these ratios some common systematic errors cancel.
234: \begin{figure}[h]
235: \includegraphics[width=0.49\linewidth]{etra_ratio_vs_npart}
236: \includegraphics[width=0.49\linewidth]{mult_ratio_vs_npart_bmw}
237: \caption{Ratios of \dEt (left) and \dNch (right) measured at different RHIC 
238: energies. The errors shown with vertical bars 
239: are the full systematic errors. Lines show that part of the systematic error 
240: that
241: allows bending or inclination of the points. The horizontal errors denote 
242: the uncertainty in the determination of \Nps.\label{fig:ratios_vs_npart}}
243: \end{figure}
244: 
245: The increase in the \Et production between \1 and \2 (with an average 
246: factor of 2.3) is larger than for \Nch (with average factor of 1.9). This is consistent
247:  with an increase in the particle production per participant common to both 
248: \Et and \Nch and a $\sim$20\% increase in \mt of produced particles contributing 
249: to the \Et parameter only.
250: 
251: The ratio of \2/\1 shows an increase from peripheral to central events; 
252: however the increase is marginal in comparison to the systematic errors of the 
253: measurement.
254: 
255: The ratio of \2/\3 is flat above \Np $\sim$ 80 and is equal to $1.140\pm0.043$ for 
256: \Et and $1.126\pm0.036$ for \Nch in the most central bin. A rather sharp 
257: increase in the ratios of both quantities between \Np=22 and 83 in the ratios of both quantities 
258: is still at the level of systematic uncertainties.
259: 
260: The preliminary ratio of \2/\6 although shows the decrease with \Np which is within the systematic errors. 
261: The increase in the particle production between these two energies is about 40\%.
262: 
263: The ratio of the transverse energy and charged particle multiplicity at mid-rapidity as 
264: a function of centrality is shown in Fig.~\ref{fig:etra_mult_vs_npart} for 
265: the three energies. The upper plot also shows the results 
266: displayed in the lower panels for comparison.
267: 
268: \begin{figure}[h]
269: \includegraphics[width=0.49\linewidth]{etra_over_mult_vs_npart}\hspace{2pc}%
270: \begin{minipage}[b]{18pc}
271: \caption{\EN vs. \Np at different RHIC energies. The errors shown 
272: with vertical bars are the full systematic errors. Lines show that part of 
273: the systematic error that allows bending or inclination of the points. The 
274: horizontal errors denote the uncertainty in the determination 
275: of \Nps.\label{fig:etra_mult_vs_npart}}
276: \end{minipage}
277: \end{figure}
278: 
279: The ratio $E_{T}/N_{ch}$~\footnote{\EN is used as a shortcut for 
280: $\langle dE_{T}/d\eta \rangle / \langle dN_{ch}/d\eta \rangle$ at $\eta=0$ in C.M.S..}, sometimes 
281: called the ``Global Barometric Observable'', triggered considerable discussion~\cite{gulash,raju}.
282: It is related to the \mt of the produced 
283: particles and is observed to be almost independent of centrality and 
284: incident energy of the collisions within the systematic errors of the previous 
285: measurements. The present paper forges a direct link between 
286: the highest SPS and lowest RHIC energies, making a more quantitative study of 
287: \EN possible.
288: 
289: The results presented in Fig.~\ref{fig:etra_mult_vs_npart} 
290: show that the centrality dependence of \EN is weak and lies within the systematic errors 
291: plotted with lines. There is a clear increase in \EN between 
292: \sqns=\1 and \2. The \sqn dependence of the results is discussed below.
293: 
294: \subsection{Bjorken Energy Density}
295: The Bjorken energy density \cite{bjorken} can be obtained using 
296: \begin{equation} 
297: \epsilon_{Bj} = \frac{1}{A_{\perp} \tau} \frac{dE_T}{dy}, 
298: \label{eq:Bj}
299: \end{equation} 
300: where $\tau$ is the formation time and $A_{\perp}$ is the nuclei 
301: transverse overlap area. 
302: 
303: The transverse overlap area of two colliding nuclei was estimated 
304: using a Monte Carlo Glauber model $A_{\perp} \sim \sigma_x \sigma_y$, 
305: where $\sigma_x$ and $\sigma_y$ are the widths of $x$ and $y$ position 
306: distributions of the participating nucleons in the transverse plane. 
307: The normalization to $\pi R^2$, 
308: where $R$ is the sum of $r_n$ and $d$ parameters in a Woods-Saxon 
309: parameterization (see e.g.:~\cite{ppg19}), was done for the most central 
310: collisions at the impact parameter $b=0$.
311: 
312: For the transformation from $dE_T/d\eta|_{\eta=0}$ to $dE_T/dy|_{y=0}$, 
313: a scale factor of $1.25 \pm 0.05$ was used, see~\cite{ppg19} for the details of conversion procedure.
314: 
315: The Bjorken energy density for three RHIC energies is plotted in the left panel of Fig.~\ref{fig:ebj}.
316: For the 5\% most central collisions, $\epsilon_{Bj} \cdot \tau$ was 
317: $2.2 \pm 0.2$, $4.7 \pm 0.5$ and $5.4 \pm 0.6$ GeV/($fm^2 \cdot c$) for 
318: \sqn=19.6, 130 and 200 GeV, respectively. These values increase by 
319: 2\%, 4\% and 5\%, respectively, for the maximal $N_{part}$=394, as obtained 
320: from extrapolation of PHENIX data points. There is a factor 
321: of 2.6 increase between the ``SPS''-like energy (\sqn=\1) and the top RHIC energy \sqn=\2.
322: The comparison of the only published 
323: $\epsilon_{Bj}$=3.2 GeV/$fm^3$ at SPS~\cite{na49_4} 
324: and top RHIC energies, assuming the same $\tau$=1~fm/c, reveals an increase in 
325: energy density by a factor of only 1.8, which may come from an overestimation 
326: in the SPS measurement, as shown in the left panel of Fig.~\ref{fig:sqn} below.
327: 
328: Another approach is used by STAR in~\cite{star_bj} for the 
329: estimate of the transverse overlap area of the two nuclei $A_{\perp} \sim N_{p}^{2/3}$ 
330: in Eq.~\ref{eq:Bj}. This approach accounts only for the common area of colliding 
331: nucleons, not nuclei. The results differ only in the peripheral bins 
332: as shown in the right panel of Fig.~\ref{fig:ebj}. For a comparison, the same 
333: panel shows the result obtained by STAR which agrees with PHENIX result 
334: within systematic errors, displaying a smaller increase of the energy 
335: density with $N_p$.
336: 
337: \begin{figure}[h]
338: \includegraphics[width=0.49\linewidth]{ebj.eps}
339: \includegraphics[width=0.49\linewidth]{ebj_comp.eps}
340: \caption{$\epsilon_{Bj} \cdot \tau$ deduced from the PHENIX data at three RHIC energies (left) 
341: and using different estimates of the nuclear transverse overlap at \sqn=\3 (right). 
342: \label{fig:ebj}}
343: \end{figure}
344: 
345: Can we derive an estimate for $\tau$? We might say, on general quantum mechanical grounds,
346: that a particle of full energy close to $m_{T}$ can be
347: considered to have ``formed'' after a time $t=\hbar/m_{T}$ since its creation in that frame.
348: 
349: To estimate the average transverse mass, we use the final-state
350: \EN results presented in Fig.~\ref{fig:etra_mult_vs_npart}. As it was mentioned above, this value is 
351: basically independent of centrality and changes from 0.74 GeV to 0.88 GeV for \sqn from 19.6 to 200~GeV.
352: Such estimate would provide us with the formation time $\tau$ which is almost the same for all initial 
353: conditions of the collisions at RHIC.
354: 
355: An approximate factor of 2/3 is used to account for the total number of particles making the estimate of
356: $\langle m_{T} \rangle \simeq$0.57~GeV at \sqn=\2 and \3 and $\langle m_{T} \rangle \simeq$0.50~GeV at \sqn=19.5~GeV.
357: This results in a formation time of $\tau \simeq$0.35~$fm/c$ and 0.40~$fm/c$ at 
358: these energies respectively. Then for the 5\% most central events the estimate of the Bjorken 
359: energy at the three RHIC energies are 5.5~GeV/$fm^3$, 13~GeV/$fm^3$ and 15~GeV/$fm^3$.
360: 
361: \subsection{Comparison to other measurements}
362: Several factors complicate the comparison between the results of PHENIX and 
363: the results of other experiments. AGS and SPS data were taken in the Laboratory (Lab.) system 
364: while the RHIC data are in the Center of Mass (C.M.S.) system. Since
365: $\eta$ and \Et are not boost invariant quantities, the data should be converted into 
366: the same coordinate system. Some experiments provide a complete 
367: set of identified particle spectra from which information about \Et and 
368: \Nch can be deduced, while other experiments require additional assumptions to extract \dEt and \dNch
369: from their results. Publication~\cite{ppg19} discusses it in details
370: 
371: The PHENIX results for \Nch are compared to the data available from the other 
372: RHIC experiments. This comparison is shown in the left panels of 
373: Fig.~\ref{fig:comp_rhic}.
374: \begin{figure}[h]
375: \includegraphics[width=0.49\linewidth]{mult_vs_npart_comp_to_rhic}
376: \includegraphics[width=0.49\linewidth]{mult_vs_npart_for_rhic_aver}
377: \caption{Left panel: \dNch per pair of \Np measured by the four RHIC experiments at 
378: different energies. The shaded area is the PHOBOS systematic error. Right panel: 
379: RHIC average values (including PHENIX) compared to the PHENIX results.\label{fig:comp_rhic}}
380: \end{figure}
381: 
382: There is good agreement between the results of BRAHMS~\cite{brahms1,brahms2}, 
383: PHENIX, PHOBOS~\cite{phobos1,phobos2,phobos3} and STAR~\cite{star1,star2} using 
384: \Np based on a Monte-Carlo Glauber model. This agreement is very impressive because 
385: all four experiments use different apparatuses and techniques to measure the 
386: charged particle production. The systematic errors of all results are uncorrelated, 
387: except for errors due to the same Glauber model which are small.
388: That makes it possible to calculate the RHIC average and reduce 
389: the systematic uncertainty. The averaged results from all four RHIC experiments 
390: are plotted in the right panel of Fig.~\ref{fig:comp_rhic}. The procedure of the 
391: calculating the average is the same as used in~\cite{PDG} and is also explained 
392: in~\cite{ppg19}.
393: 
394: Figure~\ref{fig:etra_star} compares $E_T$ results from the PHENIX and STAR~\cite{star_etra}
395: experiments. The results are consistent for all centralities within systematic errors, though 
396: STAR \dEt per participant pair as shown in the upper panel has a smaller slope for $N_p$ 
397: in going from from semi-peripheral to central collisions and \EN shown in the lower panel is consistent 
398: for all \Nps.
399: \begin{figure}[h]
400: \includegraphics[width=0.49\linewidth]{etra_comp_star}\hspace{2pc}%
401: \begin{minipage}[b]{18pc}
402: \caption{\dEt divided by the number of \Np pairs (top) and  \EN (bottom) measured 
403: by the PHENIX and STAR~\cite{star_etra} experiments at \sqn=200 GeV. PHENIX systematic 
404: errors are explained in the text. The shaded area is the STAR systematic scaling error.\label{fig:etra_star}}
405: \end{minipage}
406: \end{figure}
407: 
408: The RHIC run at \sqns=\1 allows us to make a connection between 
409: RHIC and SPS data. The highest SPS energy of 158A~GeV corresponds to 
410: \sqns~=~\7 in the C.M.S., making a direct comparison of RHIC and SPS results 
411: possible. This comparison is shown in Fig.~\ref{fig:comp_sps}. 
412: The SPS data is taken from~\cite{na49_1,na49_2,na49_3,wa98,ceres_1,na50_1}
413: details of the data compilation are explained in \cite{ppg19}.
414: \begin{figure}[h]
415: \includegraphics[width=0.49\linewidth]{etra_vs_npart_comp_to_sps}
416: \includegraphics[width=0.49\linewidth]{mult_vs_npart_comp_to_sps_bmw}
417: \caption{\dEt (left) and \dNch (right) divided by the number of \Np pairs 
418: measured by PHENIX at \sqns=\1 (solid markers) and recalculated from 
419: the results of the SPS experiments at the highest energy \sqns=\7 (open markers). 
420: The $p+p$ result of NA49 is marked with an open cross.
421: \label{fig:comp_sps}}
422: \end{figure}
423: 
424: Several comments should be made about this comparison. For both measured 
425: parameters the PHENIX results and the SPS results agree. The WA98 results 
426: are systematically higher, especially for $dE_{T}/d\eta$. However the WA98 data has 
427: an additional systematic error common to all points shown for the last bin. 
428: For \Nch the relative spread of the SPS results is 
429: larger than for the RHIC results shown in Fig.~\ref{fig:comp_rhic}, 
430: though overall the \sqns=\7 SPS measurements are consistent with the PHENIX 
431: result at \sqns~=~19.6 GeV. The NA57 results at this energy and at lower energy 
432: are published without systematic errors~\cite{na57}, so they cannot be compared 
433: to other results on the same basis.
434: 
435: Different SPS and AGS experiments made measurements at lower energies. The 
436: combined data of AGS, SPS and RHIC provide a complete picture of the 
437: centrality behavior of \Et and \Nch as a function of the nucleon-nucleon 
438: energy. The centrality dependence of \dNch at mid-rapidity measured at 
439: \sqns=4.8, 8.7 and \7 by different experiments is shown in Fig.~\ref{fig:sps_ags}. 
440: The data is taken form~\cite{na49_2,na49_3,ceres_3,na50_1,ceres_2,e802_1,e802_2,e802_3,e802_5,e917_4,e802_6} 
441: and the details of data compilation can be found in~\cite{ppg19}.
442: 
443: \begin{figure}[h]
444: \includegraphics[width=0.49\linewidth]{mult_vs_npart_for_sps_aver_bmw}\hspace{2pc}%
445: \begin{minipage}[b]{18pc}
446: \caption{\dNch divided by the number of \Np pairs measured by AGS and SPS 
447: experiments and the average taken at different energies recalculated in the C.M.S.
448: \label{fig:sps_ags}}
449: \end{minipage}
450: \end{figure}
451: 
452: At the highest SPS energy the averaging procedure is the same as for RHIC 
453: energies and weighted experimental errors are scaled with the reduced 
454: $\chi^{2}$-like factor $S$ 
455: reaching the value of 1.5 at some points. For the intermediate SPS 
456: energy \sqns=8.7~GeV, two experiments NA45~\cite{ceres_3} and 
457: NA50~\cite{na50_1} reported the centrality dependence of \dNch at 
458: mid-rapidity. The discrepancy in the measurements is approximately three times
459: the quadratic sum of their systematic errors. However the shapes of the two curves are almost 
460: the same. NA49 has published results~\cite{na49_2,na49_3} 
461: which give one point in \dNch at \Np=352. This point favors 
462: the NA45 result. The NA57 results~\cite{na57} also plotted in the Fig.~\ref{fig:sps_ags} 
463: demonstrate even larger discrepancy with the NA50 measurements, but these points 
464: lack the systematic errors, as we mentioned before.
465: 
466: The average centrality curve is produced by taking into 
467: account the shape of the centrality curves reported by NA45 and NA50 
468: and the single NA49 point. The errors are scaled with the 
469: factor $S$, which reaches a value of 2.5 at some points. The AGS results 
470: are presented with a curve produced from the combined results of the E802/E917 
471: experiments (see publication~\cite{ppg19}).
472: 
473: The average SPS centrality dependence at \sqns~=~\7 shown in the upper panel 
474: in Fig.~\ref{fig:sps_ags} and the average curve of the two RHIC 
475: experiments at \sqns=\1 shown in the lower panel in Fig.~\ref{fig:comp_rhic} 
476: are very similar. Less than 5\% increase is expected due to the difference in
477: the incident energies between the highest SPS and the lowest RHIC 
478: energies (see section~\ref{sec:central_collision} below). 
479: 
480: The average values presented in Figs.~\ref{fig:comp_rhic} 
481: and \ref{fig:sps_ags} are summarized in Table~\ref{tab:averages}.
482: 
483: \begin{center}
484: \begin{table}
485: \caption{ Average values of \dNch/(0.5\Nps) at different \sqns. An 
486: additional 5\% error should be added to columns \7 through 4.8~GeV for 
487: the uncertainty related to recalculation to the C.M.S..
488: \label{tab:averages}}
489: \centering
490: \begin{tabular}{cccccccc}
491: \br
492: \Np & \2   & \3   & \6     & \1     & \7    & 8.7~GeV   & 4.8~GeV   \\
493:     & RHIC & RHIC & PHENIX & PHENIX/& SPS   & SPS       & E806/E917 \\
494:     & aver.& aver.& prelim.& PHOBOS & aver. & aver.     & combined  \\
495: \mr
496: 375 & 3.92$\pm$0.13 & 3.41$\pm$0.10 & 	            &               & 1.97$\pm$0.12 & 1.26$\pm$0.11 & 0.92$\pm$0.14\\
497: 350 & 3.81$\pm$0.13 & 3.31$\pm$0.10 & 2.77$\pm$0.21 & 1.91$\pm$0.11 & 1.93$\pm$0.12 & 1.22$\pm$0.11 & 0.89$\pm$0.13\\
498: 325 & 3.72$\pm$0.12 & 3.22$\pm$0.10 & 2.68$\pm$0.20 & 1.89$\pm$0.11 & 1.90$\pm$0.14 & 1.20$\pm$0.11 & 0.85$\pm$0.13\\
499: 300 & 3.65$\pm$0.12 & 3.16$\pm$0.10 & 2.60$\pm$0.20 & 1.88$\pm$0.11 & 1.88$\pm$0.15 & 1.18$\pm$0.10 & 0.81$\pm$0.12\\
500: 275 & 3.56$\pm$0.12 & 3.11$\pm$0.09 & 2.54$\pm$0.19 & 1.87$\pm$0.11 & 1.83$\pm$0.16 & 1.17$\pm$0.10 & 0.78$\pm$0.12\\
501: 250 & 3.51$\pm$0.12 & 3.07$\pm$0.09 & 2.48$\pm$0.19 & 1.87$\pm$0.12 & 1.80$\pm$0.17 & 1.16$\pm$0.10 & 0.76$\pm$0.11\\
502: 225 & 3.45$\pm$0.12 & 3.04$\pm$0.10 & 2.43$\pm$0.19 & 1.85$\pm$0.12 & 1.78$\pm$0.17 & 1.16$\pm$0.10 & 0.75$\pm$0.11\\
503: 200 & 3.38$\pm$0.11 & 3.00$\pm$0.09 & 2.37$\pm$0.19 & 1.83$\pm$0.12 & 1.75$\pm$0.17 & 1.14$\pm$0.10 & 0.74$\pm$0.11\\
504: 175 & 3.34$\pm$0.12 & 2.96$\pm$0.10 & 2.30$\pm$0.19 & 1.81$\pm$0.12 & 1.72$\pm$0.17 & 1.14$\pm$0.09 & 0.72$\pm$0.11\\
505: 150 & 3.27$\pm$0.12 & 2.89$\pm$0.10 & 2.23$\pm$0.20 & 1.76$\pm$0.13 & 1.69$\pm$0.17 & 1.14$\pm$0.09 & 0.71$\pm$0.11\\
506: 125 & 3.20$\pm$0.12 & 2.83$\pm$0.10 & 2.13$\pm$0.20 & 1.72$\pm$0.14 & 1.66$\pm$0.16 & 1.14$\pm$0.09 & 0.70$\pm$0.11\\
507: 100 & 3.14$\pm$0.13 & 2.73$\pm$0.11 & 2.02$\pm$0.22 & 1.68$\pm$0.15 & 1.66$\pm$0.23 & 1.14$\pm$0.09 & 0.67$\pm$0.14\\
508:  75 & 3.03$\pm$0.13 & 2.65$\pm$0.11 & 1.89$\pm$0.24 & 1.62$\pm$0.19 & 1.61$\pm$0.21 &               & 0.64$\pm$0.18\\
509:  50 & 2.73$\pm$0.13 & 2.53$\pm$0.12 & 	            &               & 1.54$\pm$0.19 &               & 0.63$\pm$0.21\\
510:  25 & 2.78$\pm$0.43 & 2.36$\pm$0.30 & 	            &               & 1.45$\pm$0.13 &               &              \\
511: \br				      	     
512: \end{tabular}			                   
513: \end{table}
514: \end{center}
515:  
516: \subsection{Dependence on the incident nucleon energy.}
517: The data compilation made in the previous section allows for
518: a detailed study of the charged particle production in heavy ion reactions 
519: at different incident energies of colliding nuclei. Although the data on
520: transverse energy production is not abundant, a similar comparison can
521: be made~\cite{phenix_milov,phenix_bazik}.
522: 
523: \subsubsection{Central Collisions}
524: \label{sec:central_collision}
525: Figure~\ref{fig:sqn} shows the energy dependence for the \dEt and \dNch 
526: production per pair of participants in the most central collisions 
527: measured by different experiments. See~\cite{ppg19} for the details of the data compilation.
528: 
529: \begin{figure}[h]
530: \includegraphics[width=0.49\linewidth]{etra_vs_sqrts}
531: \includegraphics[width=0.49\linewidth]{mult_vs_sqrts_bmw}
532: \caption{Left panel: \dEt divided by the number of \Np pairs measured in the most 
533: central bin (value given in brackets) as a function of incident nucleon energy. The line is a logarithmic fit. 
534: The band corresponds to a 1$\sigma$ statistical deviation of the fit parameters. 
535: Right panel: the same for $dN_{ch}/d\eta$. The values of \Nch are the average values 
536: corresponding to \Nps~=~350. The single point at \sqn=~56~GeV is based on~\cite{phobos1}.\label{fig:sqn}}
537: \end{figure}
538: The results shown in Fig.~\ref{fig:sqn} are consistent with logarithmic 
539: scaling as described in~\cite{phenix_milov,sasha_thesis,david_thesis}. 
540: Use of the logarithmic function is phenomenological and is suggested 
541: by the trend of the data in the range of available measurements.
542: The agreement of the fits with the data in both panels is very good, especially in 
543: the right panel where the averaged values are used for \Np=350. The single point 
544: of NA49~\cite{na49_4} is excluded from the \Et fit. The results 
545: of the fit $dX/d\eta=(0.5N_{p}\cdot A) ln(\sqrt{s_{NN}}/\sqrt{s_{NN}^{0}})$ are:\\
546: for \Et  $\sqrt{s_{NN}^{0}} = 2.35\pm0.2$~GeV  and $A = 0.73\pm0.03$~GeV \\
547: for \Nch $\sqrt{s_{NN}^{0}} = 1.48\pm0.02$~GeV and $A = 0.74\pm0.01$.\\
548: 
549: The parameter $\sqrt{s_{NN}^{0}}$=2.35~GeV obtained from the \Et fit is 
550: slightly above, although within 3$\sigma$ from the minimum possible 
551: value of \sqns~=~$2\times a.m.u.$=1.86~GeV. The measurement closest to it 
552: at \sqns~=~2.05~GeV done by the FOPI experiment allows
553: to estimate the amount of \dEt produced to be 5.0~GeV in the most central collisions 
554: corresponding to \Np=359. Publication~\cite{ppg19} gives details of the estimate. 
555: This does not disagree with the extrapolation of the fit but 
556: does indicate that the logarithmic parameterization requires higher order 
557: terms to describe how the \Et production starts at very low \sqns.
558: 
559: The right panel of Fig.~\ref{fig:sqn} shows the logarithmic fit to the \Nch data. 
560: It agrees well with all \dNch results plotted for \Nps=350. Unlike that for $E_{T}$, 
561: the fit parameter $\sqrt{s_{NN}^{0}}$ for \Nch is 1.48$\pm$0.02~GeV which is 
562: lower than the minimum allowed \sqns. This suggests that above $2 \times a.m.u.$ 
563: the \Nch production as a function of \sqn should undergo threshold-like behavior, 
564: unlike the \Et production which must approach zero smoothly due to energy 
565: conservation. 
566: 
567: The FOPI measurement at \sqns=1.94~GeV and 2.05~GeV agrees with the 
568: extrapolation of the fit at energy very close to $2 \times a.m.u.$. It is an 
569: interesting result that colliding nuclei with kinetic energies of
570: 0.037~GeV and 0.095~GeV per nucleon in the C.M.S. follow the same particle 
571: production trend as seen at AGS, SPS and RHIC energies.
572: 
573: A fit to the charged particle multiplicity shows a factor of 2.2 increase in \dNch 
574: per participant in the most central events from the highest energy at the AGS (\sqn=4.8~GeV) to the 
575: highest energy at the SPS (\sqn=\7) and a factor of 2.0 from the highest SPS energy to the highest 
576: RHIC energy (\sqn=\2). Assuming the same behavior extends to the LHC highest energy \sqns=5500~GeV one would expect  
577: \dNch = $(6.1 \pm 0.13)\cdot (0.5 N_{p})$ 
578: and the increase in particle production from the highest RHIC energy to be 
579: $\sim$60\% for the most central events. 
580: With the greater energy, the rapidity width should increase by $\sim$60\%
581: i.e. the total charged particle multiplicity would increase by 
582: a factor of $\sim$2.6 from the top RHIC energy.
583: 
584: It is interesting to compare the \sqn dependence shown in Fig.~\ref{fig:sqn} to the similar plot for 
585: $dN_{ch}/dy$ vs. \sqn published in~\cite{kink}. In Fig.4 of that publication the energy dependence 
586: of $dN_{ch}/dy$ for \Nps~=350 changes the slope between AGS 
587: and lower SPS energy. That is not seen in Fig.~\ref{fig:sqn}. The parameterization suggested in 
588: publication~\cite{kink}, $dN_{ch}/dy\propto(\sqrt{s_{NN}})^{0.3}$ is also different from the logarithmic 
589: scaling in Fig.~\ref{fig:sqn}. Although quantities in these two figures are not identical,
590: some contribution to the shape of the curve in Fig.4 in~\cite{kink} may be due to use of logarithmic 
591: scale on Y-axis, producing a bend-like shape out of straight line in a linear scale.
592: 
593: The ratio of \EN for the most central bin as a function of \sqn is 
594: shown in Fig.~\ref{fig:et_nc_sqn}.
595: \begin{figure}[h]
596: \includegraphics[width=0.49\linewidth]{etra_over_mult_vs_sqrts}\hspace{2pc}%
597: \begin{minipage}[b]{18pc}
598: \caption{Ratio of \Et over \Nch for the most central events as a function 
599: of \sqn recalculated into C.M.S.. The line is the ratio of two fits shown 
600: in Fig.~\ref{fig:sqn}. The band corresponds to one standard deviation of 
601: the combined error.\label{fig:et_nc_sqn}}
602: \end{minipage}
603: \end{figure}
604: Note that the line shown in the figure is not the fit to the data 
605: points. Rather, it is calculated as a ratio of the fits plotted in Fig.~\ref{fig:sqn}. 
606: The calculation agrees well with the data.
607: 
608: There are two regions in the plot which can be clearly separated. The
609: region from the lowest allowed \sqn to SPS energy is characterized
610: by a steep increase of the \EN ratio with \sqns. In this region the increase 
611: in the incident energy causes an increase in the \mt of the produced particles. 
612: The second region starts 
613: at the SPS energies. In this region, the \EN ratio is very weakly 
614: dependent on \sqns. The incident energy is converted into particle 
615: production at mid-rapidity rather than into increasing the particle $\langle m_{T} \rangle$. 
616: 
617: The shape of the \EN curve in the first region is governed by the 
618: difference in the $\sqrt{s_{NN}^{0}}$ parameter between \Et and $N_{ch}$. The 
619: second region is dominated by the ratio of the $A$ parameters in the fits; 
620: this ratio is close to 1~GeV. Extrapolating to LHC energies one gets
621: a \EN value of (0.92$\pm$0.06)~GeV.
622: 
623: The \EN parameter is usually associated with the chemical freeze-out temperature $T$. A simple relation 
624: like $2/3 E_{T}/N_{ch}\approx m_{0}+3/2 T$ finds $T\approx$0.22~GeV for a measured value \EN of 
625: 0.88~GeV at top energy at RHIC. Coefficient 2/3 takes into account neutral particle contribution 
626: and the average particle mass $m_{0}\approx0.25$~GeV can be estimated from~\cite{tatsuja}. The freeze-out temperature estimate 
627: of $T\approx$0.22 GeV extracted this way is in an agreement to $T\approx$0.17 GeV measured from the 
628: spectra~\cite{tatsuja} plus some contribution of particle flow. In publications~\cite{Redlich,pbm} 
629: authors suggest the same \EN value close to 1~GeV is valid even at AGS and SIS energies. This is not in 
630: disagreement with the curve in Fig.~\ref{fig:et_nc_sqn} because the \Et definition used 
631: in~\cite{Redlich,pbm}, includes full baryon rest mass. We use different definition
632: of \Et as explained in~\cite{ppg19} where mass of pre-existing baryons is not included.
633: At lower energy where such contribution to the \Et is larger than at RHIC it would bring the \EN ratio closer to one.
634: 
635: The transition between two region with different \EN behavior occurs in the lower SPS energies. There are 
636: other observables which also undergo transition between regimes in the same range of \sqns. For example, the 
637: $\langle K^{+}\rangle /\langle \pi^{+}\rangle$ ratio demonstrates a peak at around \sqns~$\approx$8~GeV 
638: as shown in Fig.4 in reference~\cite{na49_recent}. In the same region, the volume of the colliding system 
639: goes through a minimum as shown in Fig.1 in reference~\cite{ceres}. The asymmetric flow parameter $v_{2}$ 
640: lies in-plane at and above the highest SPS energy and out-of-plane below it, see Fig.8 in 
641: publication~\cite{harry}. These evidences suggest that the energy range of 5--10~GeV has interesting 
642: physics potential which needs to be studied.
643: 
644: \subsubsection{Centrality shape}
645: \label{sec:centrality_shape}
646: Another interesting question is how the shapes of the centrality curves 
647: of \Et and \Nch change with \sqns. 
648: 
649: One approach, previously used in a number of papers is to describe the shape of the 
650: centrality dependence as a sum of ``soft'' and ``hard'' contributions such that the 
651: ``soft'' component is proportional to \Np and the ``hard'' component to the the number of 
652: binary collisions $N_{c}$: $A \times N_{p} + B \times N_{c}$. A disadvantage of this 
653: approach is that the contributions called ``soft'' and ``hard'' do not necessarily 
654: correspond to the physical processes associated with these notations. 
655: Another approach is to assume that the production of \Et and \Nch is proportional to 
656: $N_{p}^{\alpha}$; although the parameter $\alpha$ does not have any physical meaning.
657: 
658: We present the results of $B/A$ and $\alpha$ obtained from the fits 
659: to the data at different \sqn in Table~\ref{tab:alphas}. Although the numbers 
660: tend to increase with beam energy, the values presented are consistent with each other 
661: within the systematic errors.
662: 
663: \begin{center}
664: \begin{table}
665: \caption{``B/A'' ratio and parameter $\alpha$ from the fit to the data. Errors are 
666: calculated assuming a change of the slope of centrality curves within the limits of the ``tilt'' 
667: errors for PHENIX and full errors for the averaged data (Table~\ref{tab:averages}).\label{tab:alphas}}
668: \centering
669: \begin{tabular}{cccc}
670: \br
671: \sqn &        \dEt         &           \dNch       &      \dNch         \\
672:  GeV &        PHENIX       &          PHENIX       &      Average       \\
673: \multicolumn{4}{c}{$B/A$}\\
674: \mr
675: 200  & $0.49^{+.69}_{-.22}$ & $0.41^{+.57}_{-.21}$ &$0.28^{+.18}_{-.15}$\\
676: 130  & $0.41^{+.52}_{-.23}$ & $0.41^{+.45}_{-.23}$ &$0.26^{+.18}_{-.11}$\\
677: 19.6 & $0.37^{+.48}_{-.22}$ & $0.21^{+.30}_{-.15}$ &$0.23^{+.73}_{-.23}$\\
678: 17.2 &			    &                      &$0.31^{+.46}_{-.24}$\\
679: 8.7  &			    &                      &$0.12^{+.64}_{-.20}$\\
680: \multicolumn{4}{c}{parameter $\alpha$}\\
681: \hline
682: 200  & 1.20$\pm$0.07 & 1.18$\pm$0.08 & 1.16$\pm$0.06\\
683: 130  & 1.14$\pm$0.08 & 1.17$\pm$0.08 & 1.14$\pm$0.05\\
684: 19.6 & 1.13$\pm$0.07 & 1.09$\pm$0.06 & 1.10$\pm$0.11\\
685: 17.2 &               &               & 1.11$\pm$0.08\\
686: 8.7  &               &               & 1.06$\pm$0.13\\
687: 4.8  &               &               & 1.20$\pm$0.24\\
688: \br
689: \end{tabular}
690: \end{table}
691: \end{center}
692: 
693: Higher quality data would make it possible to derive a more conclusive 
694: statement about the shape of the curves plotted in Figs.~\ref{fig:comp_rhic}~and~\ref{fig:sps_ags}. 
695: With the present set of data, usually limited to \Np above 50, 
696: a large part of the centrality curve is missing or smeared by systematic errors.
697: To avoid this, we compare $Au+Au$ collisions to $p+p$ (\Nps=2) at the same energy.
698: 
699: \begin{figure}[h]
700: \includegraphics[width=0.49\linewidth]{mult_vs_sqrts_for_diff_npart_new_bmw}\hspace{2pc}%
701: \begin{minipage}[b]{18pc}
702: \caption{The three panels show $dN_{ch}/d\eta/(.5N_{p})$ divided by the logarithmic parameterization from 
703: Fig.~\ref{fig:sqn}. The panels correspond to \Nps~=~350,~100~and~2 ($p+p$) from top to bottom. 
704: $Au+Au$ points are connected with lines also shown in lower panels for comparison. 
705: The $Au+Au$ data is tabulated in Table~\ref{tab:averages}; $p+p$ data and parameterizations 
706: $dN/d\eta=2.5-0.25ln(s_{NN})+0.023ln(s_{NN})^{2}$ (solid line) and 
707: $dN/d\eta=0.27ln(s_{NN})-0.32$ (dashed line) are taken from~\cite{ua5,isr}. 
708: \label{fig:centr}}
709: \end{minipage}
710: \end{figure}
711: 
712: Figure~\ref{fig:centr} shows $dN_{ch}/d\eta/(0.5N_{p})$  
713: divided by the parameterization plotted in the right panel of Fig.~\ref{fig:sqn}.
714: The top panel shows the most central events with $N_{p}\approx$350. All points are consistent 
715: with one demonstrating an agreement of the fit to the data. The points are connected with 
716: a line for visibility. The middle panel shows results for mid-central events with \Nps~=~100 
717: connected with a solid line. The dashed line is the line from the top panel for \Nps~=~350.
718: The points for \Nps~=~100 are lower than \Nps~=~350 by a factor of $0.8-0.9$, over the plotted 
719: range of incident energies. The lower panel shows $p+p$ data corresponding to \Nps~=~2 measured 
720: by several experiments. Dashed lines are the same as appear in the upper two panels for \Nps~=~350~and~100 
721: and the $p+p$ parameterizations from~\cite{ua5,isr}. In the range of RHIC energies these 
722: points are lower by a factor of $0.65-0.75$ than the most central. Dotted lines show data from 
723: the upper two panels. 
724: 
725: These results indicate that the centrality curves normalized to the most central collisions 
726: have a similar shape for all RHIC energies within the errors of available measurements.
727: 
728: At lower energies the $p+p$ data show rise and become equal to the most central results. 
729: That would mean that at the lower SPS 
730: and AGS energies the centrality profile is independent of \Nps. At the same time the 
731: cross over takes place at AGS energy and it would suggest that below it the particle production 
732: per participant decreases with \Nps. There is no measurement supporting such a statement. Also, the 
733: the information about $p+p$ particle production at low energies is recalculated from total number 
734: of produced particles or identified particles sets which require additional assumptions and 
735: systematic uncertainties to be assigned to them. Hence, it is not as reliable as at higher energies.
736: 
737:  The data
738: for of the FNAL and BHMC collaborations are shown in Fig.~\ref{fig:centr} as an indication of the 
739: trend.
740: 
741: \subsection{Comparison to models}
742: A variety of models attempting to describe the behavior of \Et and \Nch as a function 
743: of centrality at different \sqn are available. 
744: An updated set of model results were collected from several theoretical groups 
745: to make a comparison as comprehensive as possible\footnote{For the JAM generator 
746: authors extracted data by running the code and assigning \Np from the Glauber model simulations.}.
747: 
748: Figures~\ref{fig:theo1} 
749: through~\ref{fig:theo_ratio1} show the comparison between the existing theoretical 
750: models\footnote{Models 
751: are presented as the best fit by the polynomial of the lowest degree which 
752: is closer than 1\% to any theoretical point. 
753: The polynomial is plotted in the range where points are provided.}
754: and the data for 19.6, 130 and 200 GeV. Brief descriptions of the models and 
755: their main characteristics are given below.
756: 
757: \begin{figure}[h]
758: \includegraphics[width=0.99\linewidth]{mult_vs_npart_comp_to_theo_1_new_bmw}
759: \caption{\dNch per pair of participants compared to theoretical models. 
760: {\it KLN}~\cite{kln}, {\it SSHM}~\cite{accardi}, {\it EKRT}~\cite{ekrt}, 
761: {\it Minijet}~\cite{minijet} and {\it JAM}~\cite{jam}. The band shows 
762: the range of prediction for the Minijet model.\label{fig:theo1}}
763: \end{figure}
764: 
765: One of the more commonly used Monte Carlo event generators 
766: is {\it HIJING}~\cite{hijing_sim, new_hijing}. This model, like several others,
767: uses pQCD for initial minijet production, and the Lund string 
768: model~\cite{lund_string} for jet fragmentation and hadronization. 
769: {\it HIJING} also includes jet quenching and 
770: nuclear shadowing. 
771: This type of model typically has two components, a soft part  
772: proportional to \Np and a hard part proportional to $N_{coll}$, which
773: partly motivated the discussion in section~\ref{sec:centrality_shape}.  
774: There are also so-called 
775: saturation models which also rely on pQCD and predict that at some fixed scale
776: the gluon and quark phase-space density saturates, thus limiting the number of
777: produced quarks and gluons. An example of this type of model is 
778: {\it EKRT}~\cite{ekrt}, which is referred to as a final state saturation model.
779: %These were the two models the first PHENIX \Nch results were compared with in~\cite{phenix_nch}. 
780: In this paper, comparisons are also made to another parton 
781: saturation type model, {\it KLN}~\cite{kln}, an
782: initial state saturation model, and also to models related to {\it HIJING}, 
783: namely 
784: {\it Minijet}~\cite{minijet} and {\it AMPT}~\cite{ampt}. {\it AMPT} is a 
785: multiphase transport model, and extends {\it HIJING} by including explicit 
786: interactions between initial minijet partons and also final state hadronic interactions. 
787: {\it Minijet} follows the same two-component model
788: as {\it HIJING} but also incorporates an energy dependent cut-off scale,
789: similar to the saturation models.
790: {\it JAM}~\cite{jam} uses {\it RQMD} and {\it UrQMD} inspired ideas for the low energy 
791: interactions and above resonance region soft string excitation is implemented along the lines of {\it HIJING}.
792: 
793: The other models are listed briefly below. 
794: {\it SSHM} and {\it SFM} did not have a designated short identifier, so
795:  they were named somewhat arbitrarily here, based on the physics the models 
796: incorporate.
797: {\it SSHM (Saturation for Semi-Hard Minijet)}~\cite{accardi} is also a 
798: two-component model: pQCD-based for semi-hard partonic interactions, while 
799: for the soft particle production it uses the wounded nucleon model.
800: {\it DSM}~\cite{dsm}, the Dual String Model, is basically the Dual 
801: Parton Model~\cite{dpm}, with the inclusion of strings. 
802: {\it SFM (String Fusion Model)}~\cite{perez}, is a string model which 
803: includes hard collisions, collectivity in the initial state (string fusion), 
804: and rescattering of the produced secondaries. 
805: Finally, there are the hadronic models, {\it LUCIFER}~\cite{kahana}, a cascade model, with input fixed
806:  from lower energy data, and 
807: {\it LEXUS}~\cite{lexus}, a Linear EXtrapolation of Ultrarelativistic 
808: nucleon-nucleon Scattering data to nucleus-nucleus collisions.
809: 
810: The available model results range from predicting (or postdicting) \dNch 
811: at one energy to predicting both
812: \dNch and \dEt at 19.6, 130 and 200 GeV. The models have varying success in 
813: reproducing the data. 
814: 
815: Figure~\ref{fig:theo1} shows that {\it KLN} model is one of the most successful at 
816: describing the \dNch centrality dependence for all three energies. 
817: However, at \sqn=\1 the theoretical curve is steeper than the data. 
818: This results in a reversed 
819: centrality dependence relative to the data for the \2 to \1 ratio.
820: {\it SSHM} describes the 130 and 200 GeV data well for centralities 
821: above \Np$\sim$100, which is the approximate limit of applicability for 
822: this and other 
823: saturation models. For the less central events, the models are lower 
824: than the data. At 19.6 GeV, the model values are 
825: significantly higher than the data. 
826: The saturation model
827: {\it EKRT} describes the central points at both energies but overshoots the 
828: more peripheral data points and thus does not reproduce the general centrality 
829: dependence of the data. 
830: For the non-saturation models included in this figure, {\it Minijet} reproduces both the 
831: overall scale, as well as the centrality and energy dependence of 
832: the data rather well.
833: 
834: The {\it JAM} model shown in Fig.~\ref{fig:theo1} and most of the models included in Fig.~\ref{fig:theo2},
835: provided values for all three energies: 19.6, 130 and 200 GeV. 
836: {\it JAM} is rather successful in describing centrality shape at \sqns~=200~GeV. 
837: It also consistent with the measurements at \3 and partially at 19.6~GeV. In ratios it underestimates the data. 
838: For the \Et results shown in Fig.~\ref{fig:theo_ratio1} {\it JAM} reproduced the centrality shapes well, 
839: but underestimates the \sqn dependence of the data.
840: {\it SFM} is in reasonable agreement with the 130 and 200 GeV data, but gives
841: much larger values than the data at 19.6 GeV.
842: {\it AMPT} is overall in good agreement with the data for the two higher energies,
843: except for the increasing trend in \dNch at the most peripheral events, which
844: is not seen in the data. At the lower energy the \Nch centrality behavior is 
845: underestimated. 
846: {\it LEXUS} severely 
847: overshoots the data for all energies, indicating
848: that nucleus-nucleus effects are not correctly accounted for.
849: {\it LUCIFER} describes the central 
850: points at 130 GeV well, but undershoots the less central values at this energy.
851: The {\it HIJING} models (version 1.37 and a new version with implemented baryon junctions, {\it HIJING B-$\bar{B}$}) 
852: only provide points at 130 and 200 GeV and are in 
853: reasonable agreement with the data at those energies, but generally give 
854: somewhat lower values. The curves shown include quenching and shadowing {\it HIJING}.
855: {\it DSM} describes 19.6 GeV reasonably well for all centralities, and the more
856: central bins for 130 and 200 GeV, but overpredicts the values for 
857: semi-central and peripheral events.
858: 
859: \begin{figure}[h]
860: \includegraphics[width=0.99\linewidth]{mult_vs_npart_comp_to_theo_2_new_bmw}
861: \caption{Theoretical models compared to \dNch per pair of participants. {\it SFM}~\cite{perez}, {\it AMPT}~\cite{ampt}, {\it LEXUS}~\cite{lexus},{\it LUCIFER}~\cite{kahana},{\it HIJING}~\cite{hijing_sim, new_hijing} and {\it DSM}~\cite{dsm}. \label{fig:theo2}}
862: \end{figure}
863: 
864: Figure~\ref{fig:theo_ratio1} shows the results for the models that provide 
865: data for both \dNch and $dE_{T}/d\eta$. For $dE_{T}/d\eta$, {\it LEXUS} and {\it SFM} consistently overshoot the 
866: data for all energies. In the ratio \EN, {\it LEXUS} gives values that are 
867: too low except at the lowest energy 19.6 GeV. That indicates that the 
868: hadronization mechanism allows too little energy per particle. 
869: The {\it SFM} gives values that are too large, except for the most peripheral bin, 
870: this suggests, that the particles are assigned transverse masses that are too large.
871: The {\it HIJING} versions and the related {\it AMPT} model are in reasonable 
872: agreement with the data for both \dEt and $E_{T}/N_{ch}$\footnote{Note that the 
873: {\it HIJING} versions available at the time the data were collected and used
874: for predictions were in worse agreement with the data~\cite{phenix_bazik}. This was
875: before energy loss and minijet separation/cut-off scale parameters were updated.}.
876: 
877: Also shown in Fig.~\ref{fig:theo_ratio1} are the ratios of results at 
878: \2 to \1, and 
879: \2 to \3, for $dE_{T}/d\eta$. These results, especially the comparison of the \2 to \1 data, is intended 
880: to make a 
881: more precise check of the \sqn dependence of the models. 
882: {\it SFM} fails to describe the \1 data and thus can not describe
883: the energy dependence
884: probed by these ratios, unlike {\it LEXUS} which however does not agree
885: well with the individual data curves for 19.6, 130 and 200 GeV.
886: {\it AMPT} and the {\it HIJING} versions reproduce the values of the ratios well,
887: as expected since they are in reasonable agreement with the individual curves.
888: {\it AMPT} and {\it HIJING} are also successful in describing the \EN ratio, as illustrated
889: in the lower panels of Fig.~\ref{fig:theo_ratio1}.
890: 
891: \begin{figure}[h]
892: \includegraphics[width=0.99\linewidth]{etra_vs_npart_comp_to_theo_1_new_bmw}
893: \caption{Theoretical models compared to \dEt per pair of participants (upper panels) and per produced charged particle (lower panels). {\it SFM}~\cite{perez}, {\it AMPT}~\cite{ampt}, {\it LEXUS}~\cite{lexus} and {\it HIJING}~\cite{hijing_sim, new_hijing}.\label{fig:theo_ratio1}}
894: \end{figure}
895: 
896: To summarize, most models reproduce at least some of the data fairly well,
897: but most also fail in describing all the data. 
898: Since the model results typically are given without systematic errors, it is 
899: not entirely straightforward to quantify the level of agreement or 
900: disagreement with the data. Qualitatively, the models that are most successful 
901: in describing both \dEt and \dNch in terms of the overall trends, both 
902: regarding centrality dependence and energy dependence, are {\it AMPT}, 
903: and the {\it HIJING} versions. {\it KLN} and {\it Minijet} unfortunately do not give 
904: information on \dEt but are successful in describing the \dNch results.
905: The \dNch results thus can either be described by the initial state saturation scenario ({\it KLN}) or
906: by the mini-jet models that need an energy-dependent mini-jet   
907: cut-off scale as described in~\cite{new_hijing, minijet} to reproduce the data.
908: 
909: \section{Summary\label{sec:summary}}
910: 
911: This paper presents a systematic study of the energy and centrality
912: dependence of the charged particle multiplicity and transverse energy 
913: at mid-rapidity at \sqn = 19.6, 130 and 200 GeV. 
914: 
915: The yields, divided by the number of participant nucleons, show a consistent
916: centrality dependence (increase from peripheral to central) between \dEt and 
917: \dNch for all energies. Furthermore, the increase in the ratio \EN from \1 to \2 
918: is consistent with a 20\% increase in \mt with increasing \sqns. 
919: The ratio \EN shows only a weak centrality dependence at RHIC energies.
920: 
921: For the \sqn dependence, comparisons were made not only among RHIC results
922: but also including data from lower energy fixed-target experiments at SPS, 
923: AGS and SIS. A phenomenological fit, scaling logarithmically with \sqns, 
924: describes both \dEt and \dNch well for all energies and for the most central collisions.
925: 
926: Using the fit results, one can separate two regions with different 
927: particle production mechanisms. The region below SPS energy is characterized by a 
928: steep increase in \EN$\sim$\mt with \sqns, whereas for the energies above 
929: SPS \EN is found to be weakly dependent on \sqns.
930: 
931: Within the systematic errors of the measurements 
932: the shape of the centrality curves of $dN_{ch}/d\eta/(0.5N_{p})$ vs. \Np 
933: were found to be the same in the range of RHIC energies and to scale with 
934: $\ln(\sqrt{s_{NN}})$. The same trend must be true for \Et because \EN has a very 
935: weak centrality dependence.
936: 
937: Based on the \dEt measurements, the Bjorken energy density 
938: estimates were performed and $\epsilon_{Bj} \cdot \tau$ was determined to be
939: $5.4 \pm 0.6$ GeV/($fm^2 \cdot c$) at \sqn = \2 for the most central bin. 
940: This is in excess of what is
941: believed to be sufficient for a phase transition to the new state of matter. 
942: The energy density increases by about a factor of 2.6 from SPS to RHIC energy.
943: 
944: Finally, a comparison between the RHIC \dNch and \dEt data and a collection 
945: of models was performed. A few models, notably {\it HIJING} and {\it AMPT},
946: reproduce both \dEt and \dNch rather well for several energies.
947: 
948: %\begin{acknowledgments}
949: % put your acknowledgments here.
950: %The authors appreciate the information provided by the model authors and would 
951: %like to thank A. Accardi, S. Barshay, S. Jeon, S. Kahana, D. Kharzeev, Z. Lin, 
952: %N. Armesto Perez, R. Ugoccioni, V. Topor Pop, and X.N. Wang for helpful 
953: %correspondence.
954: %\end{acknowledgments}
955: 
956: \section*{References}
957: \begin{thebibliography}{99}
958: 
959: \bibitem{ppg19} PHENIX Collaboration S.S.~Adler \etaL, 
960:   nucl-ex/0409015
961: 
962: \bibitem{rhic} H.~Hahn \etaL, 
963:   \Journal{\NIMA} {499}{245-263}{2003}.
964: 
965: \bibitem{phenix} PHENIX Collaboration, K.~Adcox \etaL, 
966:   \Journal{\NIMA} {499}{469-479}{2003}.
967: 
968: \bibitem{pc} K.~Adcox \etaL, 
969:   \Journal{\NIMA} {497}{263-293}{2003}.
970: 
971: \bibitem{emcal} L.~Aphechetche \etaL, 
972:   \Journal{\NIMA} {499}{521-536}{2003}.
973: 
974: \bibitem{bbc-zdc} M.~Allen \etaL, 
975:   \Journal{\NIMA} {499}{549-559}{2003}.
976: 
977: \bibitem{zdc} S.~Adler \etaL, 
978:   \Journal{\NIMA} {470}{488-499}{2001}.
979: 
980: %analysis section references
981: \bibitem{phenix_et} PHENIX Collaboration, K.~Adcox \etaL,
982:  	\Journal{\PRL} {87}{052301}{2001}.
983: 
984: \bibitem{phenix_nch} PHENIX Collaboration, K.~Adcox \etaL,
985:    	\Journal{\PRL} {86}{3500}{2001}.
986: 
987: \bibitem{phenix_milov} A.~Milov for the PHENIX Collaboration, 
988: 	\Journal{\NPA} {698}{171}{2002}.
989: 
990: \bibitem{phenix_bazik}  A.~Bazilevsky for the PHENIX Collaboration,
991: 	\Journal{\NPA} {715}{486}{2003}.
992: 
993: \bibitem{david_thesis} D.~Silvermyr, 
994:   Ph.D. Thesis. Lund University (2001).
995: 
996: \bibitem{sasha_thesis} A.~Milov, 
997:   Ph.D. Thesis. The Weizmann Institute of Science (2002).
998: 
999: \bibitem{julia} PHENIX Collaboration, K.~Adcox \etaL,
1000:    	\Journal{\PRL} {88}{242301}{2002}.
1001: 
1002: \bibitem{geant} GEANT 3.2.1, 
1003:   CERN program library.
1004: 
1005: \bibitem{hijing_sim} X.N.~Wang and M.~Gyulassy, 
1006:  	\Journal{\PRD} {44}{3501}{1991}.
1007: 
1008: \bibitem{tatsuja} PHENIX Collaboration, S.S.~Adler \etaL, 
1009:   \Journal{\PRC} {69}{034909}{2004}.
1010: 
1011: \bibitem{na49_1} J.~B\"{a}chler for the NA49 collaboration,
1012: 	\Journal{\NPA} {661}{45}{1999}.
1013: 
1014: \bibitem{na49_2}  M.~van Leeuwen for the NA49 collaboration,
1015:         \Journal{\NPA} {715}{161}{2003}.
1016: 
1017: \bibitem{na49_3}  NA49 Collaboration, S.V.~Afanasiev \etaL, 
1018:         \Journal{\PRC} {66}{054902}{2002}. 
1019: 
1020: \bibitem{PDG} K.~Hagiwara \etaL, http://pdg.lbl.gov,
1021: 	\Journal{\PRD} {66}{010001}{2002}.
1022: 
1023: \bibitem{Hofstadter} B.~Hahn, D.G.~Ravenhall and R.~Hofstadter,
1024: 	Phys. Rev, 101, 1131 (1956), and\\
1025: 	C.W.~De Jager \etaL, Atomic Data and Nuclear Table 24, 479 (1974).
1026: 
1027: \bibitem{phobos6} PHOBOS Collaboration, B.B.~Back \etaL, 
1028:         nucl-ex/0210015.
1029: 
1030: \bibitem{ags_mjt} E-802 Collaboration, T.~Abbott \etaL,
1031:         \Journal{\PRC} {63}{064602}{2001}.
1032: 
1033: \bibitem{mitchell} PHENIX Collaboration, K.~Adcox \etaL,
1034:         \Journal{\PRC} {66}{024901}{2002}.
1035: 
1036: \bibitem{wa98} WA98 Collaboration, M.M. Aggarval \etaL, 
1037: 	\Journal{\EPJC} {18}{651}{2001}.
1038: 
1039: \bibitem{phenix_pp} PHENIX Collaboration, S.S.~Adler \etaL,
1040:         \Journal{\PRL} {91}{241803-1}{2003}.
1041: 
1042: %results section references
1043: 
1044: \bibitem{gulash} M.~Gyulassy, 
1045:   nucl-th/0106072.
1046: 
1047: \bibitem{raju} A.~Krasnitz, Y.~Nara, R.~Venugopalan, hep-ph/0305112,
1048:    	\Journal{\NPA} {727}{427-436}{2003}.
1049: 
1050: \bibitem{bjorken} J.D.~Bjorken, 
1051:         \Journal{\PRD} {27}{140}{1983}.
1052: 
1053: \bibitem{na49_4}  NA49 Collaboration, T.~Alber \etaL, 
1054:         \Journal{\PRL} {75}{3814}{1995}.
1055: 
1056: \bibitem{kink} A.Andronic and P.Braun-Munzinger, hep-ph/04022091
1057: 
1058: \bibitem{Redlich} J.~Cleymans and K.~Redlich,
1059:         \Journal{\PRL} {81}{5284}{1988}.
1060: 
1061: \bibitem{pbm} P.Braun-Munzinger, K.Redlich and J.Stachel 
1062:   nucl-th/0304013
1063: 
1064: \bibitem{na49_recent} M.Gazdzicki for the NA49 Collaboration
1065:   nucl-ex/0403023
1066: 
1067: \bibitem{ceres} CERES Collaboation D.Adamova \etaL,
1068:         \Journal{\PRL} {90}{022301}{2003}.
1069: 
1070: \bibitem{harry} A Appelsh\"{a}user for the CERES Collaboation,
1071:         \Journal{\NPA} {698}{253c-260c}{2002}.
1072: 
1073: \bibitem{star_bj} STAR Collaboration, J.~Adams \etaL, 
1074:   nucl-ex/0311017.
1075: 
1076: \bibitem{brahms1} BRAHMS Collaboration, I.G.~Bearden \etaL, 
1077:    	\Journal{\PLB} {523}{227}{2001}.
1078: 
1079: \bibitem{brahms2} BRAHMS Collaboration, I.G.~Bearden \etaL, 
1080:  	\Journal{\PRL} {88}{202301}{2002}. 
1081: 
1082: \bibitem{phobos1} PHOBOS Collaboration, B.B.~Back \etaL, 
1083:  	\Journal{\PRL} {85}{3100}{2000}.
1084: 
1085: \bibitem{phobos2} PHOBOS Collaboration, B.B.~Back \etaL, 
1086:  	\Journal{\PRC} {65}{061901}{2002}.
1087: 
1088: \bibitem{phobos3} M.D.~Baker \etaL for the PHOBOS Collaboration,, 
1089:         \Journal{\NPA} {715}{65c-74c}{2003}.
1090: 
1091: \bibitem{star1}   T.S.~Ullrich for the STAR Collaboration, 
1092:          nucl-ex/0305018.
1093: 
1094: \bibitem{star2}   T.S.~Ullrich for the STAR Collaboration, 
1095:         \Journal{\NPA} {715}{399}{2003}.
1096: 
1097: \bibitem{star_etra} STAR Collaboration, J~ Adams \etaL, 
1098:         nucl-ex/0407003.
1099: 
1100: \bibitem{ceres_3} D.~Mi\`{s}kowiec for the CERES Collaboration,
1101: Proc. of CIPPQG Palaiseau, September 2001.
1102: 
1103: \bibitem{na50_1}  NA50 Collaboration, M.C.~Abreu \etaL, 
1104:         \Journal{\PLB} {530}{43}{2002}.
1105: 
1106: \bibitem{na57}    NA57 Collaboration, F.~Antinori \etaL, 
1107:         nucl-ex/0406004.
1108: 
1109: %\bibitem{na49_4}  NA49 Collaboration, T.~Alber \etaL, 
1110: %        \Journal{\PRL} {75}{3814}{1995}.
1111: 
1112: \bibitem{phobos_total} PHOBOS Collaboration, B.B.~Back \etaL, 
1113:         nucl-ex/0301017
1114: 
1115: %\bibitem{phobos5} PHOBOS Collaboration, B.B.~Back \etaL, 
1116: %        nucl-ex/0405027
1117: 
1118: \bibitem{ua5}  UA5 Collaboration, G.J.~Alner \etaL,
1119: 	\Journal{\ZPC} {33}{1-6}{1986}.
1120: 
1121: \bibitem{isr}  W.~Thome \etaL, 
1122:  	\Journal{\NPB} {129}{365-389}{1977}. 
1123: 
1124: \bibitem{fnal}  J.~Whitmore,
1125:         \Journal{\PLC} {10}{274-373}{1974}.
1126: 
1127: \bibitem{bhdm} V.~Blobel \etaL,
1128:  	\Journal{\NPB} {69}{454-492}{1974}. 
1129: 
1130: \bibitem{new_hijing} V.~Topor~Pop, M.~Gyulassy, J.~Barrette, C.~Gale,
1131:        X.~N.~Wang, N.~Xu, and K.~Filimonov,
1132:  	\Journal{\PRC} {68}{054902}{2003}.
1133: 
1134: \bibitem{lund_string} B.~Andersson \etaL,  
1135:  	\Journal{\NPB} {281}{289}{2003}.
1136: 
1137: \bibitem{ekrt} K.J.~Eskola \etaL, 
1138: 	\Journal{\NPB} {570}{379}{2000};
1139:    	\Journal{\PLB} {497}{39}{2001}.
1140: 
1141: \bibitem{kln} D.~Kharzeev and M.~Nardi, 
1142:    	\Journal{\PLB} {507}{121}{2001};
1143: 	D.~Kharzeev and E.~Levin, 
1144:    	\Journal{\PLB} {523}{79}{2001}.
1145:  
1146: \bibitem{minijet} S.~Lee and X.N.~Wang,
1147:    	\Journal{\PLB} {527}{85}{2002}. 
1148: 
1149: \bibitem{jam} Y.Nara,
1150:    	\Journal{\NPA} {638}{555c}{1998}. 
1151: 
1152: \bibitem{ampt} Z.~Lin \etaL,
1153:    	\Journal{\PRC} {64}{011902}{2001}. 
1154: 
1155: \bibitem{accardi} A.~Accardi,
1156:    	\Journal{\PRC} {64}{064905}{2001}. 
1157: 
1158: \bibitem{dsm} R.~Ugoccioni,
1159:    	\Journal{\PLB} {491}{253}{2000}. 
1160: 
1161: \bibitem{dpm} A.~Capella {\it et al},
1162:    	\Journal{\PLC} {236}{225}{1994}. 
1163: 
1164: \bibitem{perez} N.~Armesto Perez \etaL,
1165:    	\Journal{\PLB} {527}{92}{2002}; 
1166:    	\Journal{\EPJC} {22}{149}{2001}. 
1167: 
1168: \bibitem{kahana} D.E.~Kahana and S.H.~Kahana,
1169:   nucl-th/0208063.
1170: 
1171: \bibitem{lexus} S.~Jeon and J.~Kapusta,
1172:    	\Journal{\PRC} {63}{011901}{2001}. 
1173: 
1174: %appendix section references
1175: 
1176: \bibitem{ceres_1} F.~Ceretto,  
1177:   Ph.D. Thesis. University of Heidelberg (1998).
1178: 
1179: \bibitem{ceres_2} H.~Appelsh\"{a}user for the CERES Collaboration,
1180:         \Journal{\NPA} {698}{253}{2002}.
1181: 
1182: \bibitem{e802_1} E802 Collaboration, L.~Ahle \etaL, 
1183:  	\Journal{\PRC} {59}{2173}{1999}. 
1184: 
1185: \bibitem{e802_2} E802 Collaboration, L.~Ahle \etaL, 
1186:  	\Journal{\PRC} {58}{3523-3538}{1998}. 
1187: 
1188: \bibitem{e802_3} C.A.~Ogilvie for the E866 and E917 Collaborations,
1189:  	\Journal{\NPA} {638}{57c-68c}{1998}. 
1190: 
1191: \bibitem{e802_5} Y.~Akiba for the E802 Collaboration,
1192:         \Journal{\NPA} {610}{139}{1996}.
1193: 
1194: \bibitem{e917_4} E917 Collaboration, B.B.~Back \etaL, 
1195:  	\Journal{\PRL} {86}{1970-1973}{2001}. 
1196: 
1197: \bibitem{e802_6} E802 Collaboration, L.~Ahle \etaL, 
1198:  	\Journal{\PRC} {57}{R466-R470}{1998}. 
1199: 
1200: \bibitem{e802_7} E802 Collaboration, L.~Ahle \etaL, 
1201:  	\Journal{\PLB} {476}{1}{2000}. 
1202: 
1203: \bibitem{e802_8} E802 Collaboration, L.~Ahle \etaL, 
1204:  	\Journal{\PLB} {490}{53}{2000}. 
1205: 
1206: \bibitem{fopi_1} FOPI Collaboration, W.~Reisdorf \etaL, 
1207:  	\Journal{\NPA} {612}{493}{1997}. 
1208: 
1209: \bibitem{fopi_2} FOPI Collaboration, D.~Pelte \etaL, 
1210:  	\Journal{\ZPA} {357}{215}{1997}. 
1211: 
1212: \bibitem{fopi_3} FOPI Collaboration, B.~Hong \etaL, 
1213:  	\Journal{\PRC} {66}{034901}{2002}. 
1214:  
1215: \end{thebibliography}
1216: 
1217: \end{document}
1218: