1: %%%%%%%%%%%%%%%%%%%%%%% file template.tex %%%%%%%%%%%%%%%%%%%%%%%%%
2: %
3: % This is a template file for The European Physical Journal
4: %
5: % Copy it to a new file with a new name and use it as the basis
6: % for your article
7: %
8: %%%%%%%%%%%%%%%%%%%%%%%% Springer-Verlag %%%%%%%%%%%%%%%%%%%%%%%%%%
9: %
10: \begin{filecontents}{leer.eps}
11: %!PS-Adobe-2.0 EPSF-2.0
12: %%CreationDate: Mon Jul 13 16:51:17 1992
13: %%DocumentFonts: (atend)
14: %%Pages: 0 1
15: %%BoundingBox: 72 31 601 342
16: %%EndComments
17:
18: gsave
19: 72 31 moveto
20: 72 342 lineto
21: 601 342 lineto
22: 601 31 lineto
23: 72 31 lineto
24: showpage
25: grestore
26: %%Trailer
27: %%DocumentFonts: Helvetica
28: \end{filecontents}
29: %
30: \documentclass[epj]{svjour}
31: % Remove option referee for final version
32: %
33: % Remove any % below to load the required packages
34: %\usepackage{latexsym}
35: \usepackage{amsmath}% amsmath package
36: \usepackage{graphics}
37: \usepackage{dcolumn}% Align table columns on decimal point
38: \usepackage{bm}% bold math
39: \renewcommand{\labelenumi}{\theenumi)}
40: \newcommand{\lw}[1]{\smash{\lower 1.5ex\hbox{#1}}}
41: % etc
42: %
43: \begin{document}
44: %
45: \title{Analyses of $\bm{k_t}$ distributions at RHIC by means of some
46: selected statistical and stochastic models}
47: %\subtitle{Do you have a subtitle?\\ If so, write it here}
48: \author{M.~Biyajima\inst{1}\thanks{\emph{e-mail:}
49: biyajima@azusa.shinshu-u.ac.jp}
50: \and
51: M.~Kaneyama\inst{1}\thanks{\emph{e-mail:}
52: kaneyama@azusa.shinshu-u.ac.jp}
53: \and
54: T.~Mizoguchi$^2$\thanks{\emph{e-mail:}
55: mizoguti@toba-cmt.ac.jp}
56: \and
57: G.~Wilk$^3$\thanks{\emph{e-mail:}
58: wilk@fuw.edu.pl}
59: % \thanks is optional - remove next line if not needed
60: } % Do not remove
61: %
62: %\offprints{} % Insert a name or remove this line
63: %
64: \institute{Department of Physics, Shinshu University Matsumoto 390-8621, Japan
65: \and Toba National College of Maritime Technology, Toba 517-8501, Japan
66: \and The Andrzej So\l tan Institute for Nuclear Studies, Ho\.za 69, 00681
67: Warsaw, Poland}
68: %
69: \date{Received: date / Revised version: date}
70: % The correct dates will be entered by Springer
71: %
72: \abstract{ The new data on $k_t$ distributions obtained at RHIC are analyzed by means
73: of selected models of statistical and stochastic origin in order to estimate their
74: importance in providing new information on hadronization process, in particular on
75: the value of the temperature at freeze-out to hadronic phase.
76: %
77: \PACS{
78: {25.75.-q}{Relativistic heavy ion collisions} \and
79: {12.40.Ee}{Statistical (extensive and non-extensive) models} \and
80: {02.50.Ey}{Stochastic models}
81: } % end of PACS codes
82: } %end of abstract
83: %
84: \maketitle
85:
86: %
87: % SECTION-1
88: %
89:
90: \section{Introduction} \label{intro}
91:
92: Very recently high $k_t$ distributions at RHIC have been reported in
93: Refs.~\cite{Arsene:2003yk,Adams:2003kv,Adcox:2002pe}. These data are of potentially
94: high interest as a possible source of information on the conditions existing at the
95: freeze-out in heavy-ion collisions. This resulted in a number of works, mostly of
96: statistical or thermal origin \cite{explanat}, stressing different possible dynamical
97: aspects, like the role of resonances or the flow phenomenon. In our work we would
98: like to show that one can account summarily for such (and others) effects considered
99: in the literature by using simple minimal extensions of the known statistical or
100: stochastic models, which were already successfully applied in other analysis of
101: experimental data. They are:
102: \begin{itemize}
103: \item[$(i)$] The modified statistical model inspired by Tsallis statistics
104: \cite{Tsallis:1988aa}, which generalizes the usual Boltzmann-Gibbs statistics to
105: nonextensive systems parametrized by a nonextensivity parameter $q$ (for
106: $q\rightarrow 1$ one returns to the usual Boltzmann-Gibbs extensive scenario); it
107: has been already successfully used in this context \cite{Wilk,Bediaga:1999hv,Wibig}.
108: Parameter $q$ summarizes in such approach all deviations from the Boltzmann-Gibbs
109: statistics including those caused by flow phenomena and resonances \cite{explanat}.
110: \item[$(ii)$] A suitable adaptation of the recently proposed model derived from the
111: Fokker-Planck equation for the Or-stein-Uhlenbeck (O-U) process
112: \cite{Biyajima:2002at,B3,MC} but this time used in the transverse rapidity space,
113: i.e., for $y_t = \frac{1}{2}\ln [(m_t + k_t)/(m_t - k_t)]$ (where $m_t =\sqrt{m^2 +
114: \langle k_t\rangle^2}$), in which one allows for the mass $m$ to be treated as free
115: parameter in order to account for some specific features of data (like flow
116: phenomenon) which cannot be explained in a usual way.
117: \end{itemize}
118: As a kind of historical reference point we shall use classical statistical model
119: developed long time ago by Hagedorn \cite{Hagedorn:1965aa} in which transverse
120: momentum distribution of produced secondaries is given by the following formula
121: \cite{HR} (with $T_0$ being parameter identified with temperature, $T_h$ denoting the
122: so called Hagedorn temperature \cite{Hagedorn:1965aa,HR} and $m_{\pi}$ being pion
123: mass),
124: \begin{eqnarray}
125: \frac{d^2\sigma}{2\pi k_t dk_t} &=& C\!\! \int_{m_{\pi}}^{\infty}\!\!\!\!\! dm
126: \rho(m)\sqrt{m^2+k_t^2} K_1\!\!\left(\frac{\sqrt{m^2+k_t^2}}{T_0}\right)\!\!; \label{eq:1}\\
127: && \rho(m) = \frac{e^{m/T_h}}{(m^2+m_0^2)^{5/4}}. \label{eq:1a}
128: \end{eqnarray}
129: As one can see in Fig. \ref{fig:1a} although fits to $k_t$ distributions at
130: $\sqrt{s_{NN}} = 200$ GeV obtained by BRAHMS Collaboration \cite{Arsene:2003yk} are
131: quite good, they start to deviate from data at highest values of $k_t$ and became
132: very bad there, what is very clearly seen in Fig. \ref{fig:1b} where we show our fits
133: to STAR data \cite{Adams:2003kv} covering larger span of transverse momenta. Although
134: one can argue that for such large values of $k_t$ statistical approach must give way
135: to some more detailed dynamical calculations \cite{explanat}, there are examples that
136: suitable modifications of statistical approach can lead to quite reasonable results
137: in leptonic, hadronic and nuclear collisions. What we have in mind here are some
138: non-extensive generalizations of statistical model as discussed in
139: \cite{Bediaga:1999hv,Wilk,Wibig}) and some specific realization of stochastic
140: approach as proposed by \cite{Biyajima:2002at,B3,SB}. In what follows we shall
141: therefore apply these two methods to nuclear data of
142: Refs.~\cite{Arsene:2003yk,Adams:2003kv,Adcox:2002pe}.\\
143:
144: In next Section we shall analyze data using nonextensive generalization of
145: statistical model by means of Tsallis statistics. In Section \ref{sec:3} we shall
146: analyze data using stochastic approach in transverse rapidity space. Our conclusions
147: are presented in Section \ref{sec:4}.
148: %
149: \begin{figure*}
150: \begin{center}
151: \resizebox{0.94\textwidth}{!}{\includegraphics{b-hag2.eps}}
152: \end{center}
153: \caption{Results of application of simple statistical model, cf.
154: Eq.~\eqref{eq:1}, to data for $k_t$-distributions at
155: $\sqrt{s_{NN}} = 200$ GeV measured for different
156: centralities by BRAHMS Collaboration~\cite{Arsene:2003yk}.
157: }
158: \label{fig:1a}
159: \end{figure*}
160:
161: \begin{figure}
162: \begin{center}
163: \resizebox{0.50\textwidth}{!}{\includegraphics{s-hag2.eps}}
164: \end{center}
165: \caption{Results of application of simple statistical model, cf.
166: Eq.~\eqref{eq:1}, to data for $k_t$-distributions at
167: $\sqrt{s_{NN}} = 200$ GeV measured for different
168: centralities by STAR Collaboration~\cite{Adams:2003kv}.
169: }
170: \label{fig:1b}
171: \end{figure}
172: %
173: \begin{table}
174: \caption{Values of parameters $C$, $T_h$ and $T_0$ in eq. (\ref{eq:1}) used to obtain
175: results presented in Figs \ref{fig:1a} and \ref{fig:1b}. The values of
176: $\chi^2$/n.d.f. for BRAHMS data are the same for all centralities and equal to
177: $19.3/23$ and $18.3/23$ for $m_0=0.5$ and $0.55$ GeV, respectively. For STAR data
178: they are equal $532/32$ for C.C $=0-5$\%, $249/32$ for C.C$=20-30$\% and $308/32$ for
179: C.C$=60-80$\%. }
180: \begin{center}
181: \begin{tabular}{|c|ccc|}
182: \hline
183: \multicolumn{4}{|c|}{BRAHMS Coll. \cite{Arsene:2003yk} $m_0=0.5$ GeV (fixed)}\\
184: \hline
185: C.C. & $C$ & $T_h$ & $T_0$\\
186: (\%) & & (GeV) & (GeV)\\
187: \hline
188: 0-10 & 177$\pm$11 &0.180$\pm$0.007
189: & 0.169$\pm$0.006\\
190: 10-20 & 127$\pm$9 &0.172$\pm$0.008
191: &0.162$\pm$0.006\\
192: 20-40 & 83$\pm$7 &0.156$\pm$0.008
193: &0.149$\pm$0.007\\
194: 40-60 & 44$\pm$5 &0.133$\pm$0.010
195: &0.128$\pm$0.009\\
196: 60-80 & 177$\pm$11 &0.095$\pm$0.0001
197: &0.093$\pm$0.0001\\
198: \hline
199: \multicolumn{4}{|c|}{BRAHMS Coll. \cite{Arsene:2003yk} $m_0=0.55$ GeV (fixed)}\\
200: \hline
201: 0-10 & 204$\pm$13 &0.172$\pm$0.008
202: &0.162$\pm$0.006\\
203: 10-20 & 146$\pm$10 &0.163$\pm$0.008
204: &0.155$\pm$0.006\\
205: 20-40 & 96$\pm$8 &0.148$\pm$0.008
206: &0.142$\pm$0.007\\
207: 40-60 & 51$\pm$6 &0.124$\pm$0.010
208: &0.121$\pm$0.009\\
209: 60-80 & 204$\pm$13 &0.075$\pm$0.00007
210: &0.075$\pm$0.0001\\
211: \hline
212: \multicolumn{4}{|c|}{STAR Coll. \cite{Adams:2003kv} $m_0=0.5$ GeV (fixed)}\\
213: \hline
214: 0-5 & 816$\pm$15 &0.086$\pm$0.0001
215: &0.085$\pm$0.0001\\
216: 20-30 & 382$\pm$7 &0.077$\pm$0.0001
217: &0.076$\pm$0.0001\\
218: 60-80 & 106$\pm$2 &0.037$\pm$0.00001
219: &0.037$\pm$0.00001\\
220: \hline
221: \end{tabular}
222: \label{tab:a2}
223: \end{center}
224: \end{table}
225: %
226: %
227: % SECTION-2
228: %
229:
230: \section{Analysis of $k_t$ distributions by generalized statistical
231: model based on Tsallis statistics} \label{sec:2}
232:
233: In many fields in physics, which use statistical and stochastic approaches as their
234: tools, it was recognized since some time already that the usual Boltzmann-Gibbs
235: approach encounters serious problems when applied to systems possessing memory
236: effects, correlations (especially long-range correlations but also those caused by
237: the production of resonances in multiparticle production processes or by the flow
238: effects present there) or which phase space has some (multi) fractal structure
239: \cite{Tsallis:1988aa}. Such systems are all, in a sense, \textit{small}, by what we
240: means that the effective range of correlations they experience is of the order of
241: dimension of the system itself. Therefore they will not show property of extensivity
242: leading to Boltzmann-Gibbs form of entropy, which is the basis of any statistical or
243: stochastical model. One can therefore argue that in such cases one has to resort to
244: some dynamical approach in which effects mentioned above would be properly accounted
245: for. The problem is, however, that there is no unique model of this type and usually
246: several approaches are competing among themselves in describing experimental data.
247: The other possibility is to realize that most probably our system is not extensive
248: (in the abovementioned sense) and that this fact should be accounted for by using the
249: non-extensive form of entropy, for example the so called Tsallis
250: \cite{Tsallis:1988aa}. It turns out that such situations are encountered also in
251: domain of multiparticle production processes at high energy collisions (cf.,
252: \cite{Wilk}, to which we refer for all details). In fact, there already exists a
253: number of detailed analysis using a non-extensive approach ranging from $k_t$
254: distributions in $e^+e^-$ annihilations \cite{Bediaga:1999hv} and in $p+\bar{p}$
255: collisions \cite{Wibig} to rapidity distributions in some selected reactions
256: \cite{Wilk}. In \cite{Bediaga:1999hv,Wibig} a kind of non-extensive $q$-version of
257: Hagedorn approach has been used whereas \cite{Wilk} were exploring information
258: theoretical approach to statistical models including as option also its non-extensive
259: version \footnote{It should be mentioned at this point that proper formulation of
260: Hagedorn model using Tsallis $q$-statistics has been proposed in \cite{qHag}. We
261: shall not pursue this problem here.}.
262:
263: In our work we shall apply Tsallis formalism, treated as simplest possible extension
264: of the usual statistical approach with parameter $q$ (the so called nonextensivity
265: parameter or entropic index) summarizing deviations from the usual statistical
266: approach (without, however, specifying their dynamical origin). It leads to ($T_0$
267: denotes temperature):
268: \begin{eqnarray}
269: \hspace{-7mm}\frac{d^2\sigma}{2\pi k_t dk_t} = C\int_0^{\infty}\left[ 1 -
270: (1-q)\frac{\sqrt{k_t^2+k_l^2 + m^2}}{T_0}
271: \right]^{Q}\, dk_l.
272: \label{eq:2}
273: \end{eqnarray}
274: There exist two different formulations leading to slightly different
275: forms of parameter $Q$:
276: \begin{itemize}
277: \item[$(a)$] In first one uses the so-called escort probability
278: distributions \cite{escort}, $P_i = p_i^q/\sum_i p_i^q$ (cf., for
279: example, analysis of $k_t$ distributions in $e^+e^-$ annihilations
280: \cite{Bediaga:1999hv} or in $p\bar{p}$ collisions\cite{Wibig}), in this
281: case $Q = q/(1-q)$.
282: \item[$(b)$] In second approach one uses normal definition of
283: probabilities resulting in $Q = 1/(1-q)$. In this case, as shown in
284: \cite{Wilk:1999dr,ql1}, parameter $q$ is given by the normalized
285: variance of all intrinsic fluctuations present in the hadronizing
286: system under consideration:
287: \begin{equation}
288: q = 1 + \omega =1 + \left( \langle \beta^2\rangle - \langle \beta
289: \rangle^2\right)/\langle \beta\rangle^2 . \label{eq:qq}
290: \end{equation}
291: This conjecture originates from the observation that:
292: %
293: \begin{eqnarray}
294: [1-(1-q)\beta_0 H_0]^{\frac {1}{1-q}} = \int_0^{\infty}e^{-\beta H_0}
295: f(\beta)d\beta
296: \label{eq:6}
297: \end{eqnarray}
298: %
299: where $f(\beta)$ describes fluctuation of parameter $\beta$ and has form of Gamma
300: distribution \cite{Wilk:1999dr,ql1} (in our case $H_0=\sqrt{k_l^2+k_t^2+m^2}$ and
301: fluctuations are in temperature, i.e., $\beta =1/T$ and $\beta_0 =\langle \beta
302: \rangle$ with respect to $f(\beta)$) \footnote{It must be mentioned at this point
303: that this suggestion, which in \cite{Wilk:1999dr} has been derived only for $q>1$
304: case, has been shown to be valid also for $q<1$ case \cite{ql1} and extended to
305: general form of fluctuations leading then to the new concept of
306: \textit{superstatistics} proposed in \cite{superq}. The most recent discussion of
307: physical meaning of $q$ parameter when applied to multiparticle production processes
308: (and in this context also of the possible origin of statistical formulas as well)
309: with relevant references can be found in \cite{Wilk}.}.
310: \end{itemize}
311: %
312: \begin{figure*}
313: \begin{center}
314: \resizebox{0.94\textwidth}{!}{\includegraphics{b-tsa.eps}}
315: \resizebox{0.94\textwidth}{!}{\includegraphics{s-tsa.eps}}
316: \resizebox{0.94\textwidth}{!}{\includegraphics{p-tsa.eps}}
317: \end{center}
318: \caption{Results of application of non-extensive approach given by
319: Eq.~\eqref{eq:2} with $Q=q/(1-q)$ to data for $k_t$-distributions
320: at $\sqrt{s_{NN}} = 200$ GeV measured for different
321: centralities by BRAHMS \cite{Arsene:2003yk} and STAR
322: \cite{Adams:2003kv} Collaborations and to data at
323: $\sqrt{s_{NN}}=130$ GeV as measured by PHENIX Collaboration
324: \cite{Adcox:2002pe}.
325: The results obtained using $Q=1/(1-q)$ instead look
326: essentially the same, therefore they are not shown separately.
327: For differences in values of obtained parameters see
328: Table~\ref{table:q-results}.
329: }
330: \label{fig:2}
331: \end{figure*}
332:
333: %%% TABLE 1 %%%
334:
335: %
336: % For tables use
337: \begin{table*}
338: \caption{Values of characteristic parameters used to fit data on $k_t$-distributions
339: at different centralities by using non-extensive approach as given by
340: Eq.~\eqref{eq:2} with $Q=q/(1-q)$ and $Q=1/(1-q)$ for data at $\sqrt{s_{NN}} = 200$
341: GeV obtained by BRAHMS \cite{Arsene:2003yk} and STAR \cite{Adams:2003kv}
342: Collaborations and at $\sqrt{s_{NN}} = 130$ GeV obtained by PHENIX Collaboration
343: \cite{Adcox:2002pe}. For results obtained using $Q=1/(1-q)$ we provide also explicit
344: values of the corresponding fluctuations of temperature as given by $\Delta
345: T_0=\sqrt{q-1}\cdot T_0$. The order of magnitude of the corresponding errors for
346: $T_0$, $q$ and $\Delta T_0$: $\delta T_0$, $\delta q$ and $\delta \Delta T_0$, ,
347: respectively, are listed below as well. In analysis we have used errors either as
348: provided by experiments (for STAR and PHENIX) or assuming systematic error on the
349: level of $5$\% (for BRAHMS). }
350: \begin{center}
351: \begin{tabular}{|c|cccc|ccccc|}
352: \hline
353: %\multicolumn{10}{|c|}{}\\
354: \multicolumn{10}{|c|}{BRAHMS Coll. \cite{Arsene:2003yk}}\\
355: %\multicolumn{10}{|c|}{}\\
356: %&&&&&&&&&\\
357: \hline
358: &\multicolumn{4}{|c|}{ Eq.~\eqref{eq:2} with $Q=q/(1-q)$}
359: &\multicolumn{5}{|c|}{ Eq.~\eqref{eq:2} with $Q=1/(1-q)$}\\
360: &\multicolumn{4}{|c|}{$\delta T_0=$0.005-0.007, $\delta q
361: =$0.003-0.005.} &\multicolumn{5}{|c|}{$\delta T_0 =$0.005-0.007,
362: $\delta q=$0.002-0.004,}\\
363: &\multicolumn{4}{|c|}{}&\multicolumn{5}{|c|}{ $\delta \Delta
364: T_0 =$0.001-0.002.}\\
365: \hline
366: C.C. & $\chi^2$/n.d.f. & C & $T_0$ & $q$ & $\chi^2$/n.d.f. & C &
367: $T_0$ & $q$ & $\Delta T_0$ \\
368: (\%) & & & (GeV) & & & &
369: (GeV) & & (GeV) \\
370: \hline
371: 0-10 & 11.2/23 & 1033$\pm$78 & 0.232 & 1.043 & 11.2/23 & 1033$\pm$78 &
372: 0.223 & 1.041 & 0.045\\
373: 10-20 & 12.9/23 & 797$\pm$66 & 0.224 & 1.049 & 12.9/23 & 797$\pm$65 &
374: 0.213 & 1.047 & 0.046\\
375: 20-40 & 12.7/23 & 525$\pm$49 & 0.215 & 1.055 & 12.7/23 & 525$\pm$49 &
376: 0.204 & 1.052 & 0.047\\
377: 40-60 & 10.5/23 & 304$\pm$38 & 0.193 & 1.067 & 10.5/23 & 304$\pm$38 &
378: 0.181 & 1.063 & 0.045\\
379: 60-80 & 2.85/22 & 41$\pm$5 & 0.175 & 1.084 & 2.85/22 & 41$\pm$5 &
380: 0.161 & 1.077 & 0.045\\
381: \hline
382: \hline
383: %\multicolumn{10}{|c|}{}\\
384: \multicolumn{10}{|c|}{STAR Coll. \cite{Adams:2003kv}}\\
385: %\multicolumn{10}{|c|}{}\\
386: %&&&&&&&&&\\
387: \hline
388: &\multicolumn{4}{|c|}{ Eq.~\eqref{eq:2} with $Q=q/(1-q)$}
389: &\multicolumn{5}{|c|}{ Eq.~\eqref{eq:2} with $Q=1/(1-q)$}\\
390: &\multicolumn{4}{|c|}{ $\delta T_0 =$0.002-0.003, $\delta q=$
391: 0.001-0.002.} &\multicolumn{5}{|c|}{$\delta T_0=$0.002-0.003, $\delta
392: q\cong$0.001,}\\
393: &\multicolumn{4}{|c|}{} &\multicolumn{5}{|c|}{$\delta \Delta
394: T_0\cong$0.001.}\\
395: \hline
396: C.C. & $\chi^2$/n.d.f. & C & $T_0$ & $q$ & $\chi^2$/n.d.f. & C &
397: $T_0$ & $q$ & $\Delta T_0$ \\
398: (\%) & & & (GeV) & & & &
399: (GeV) & & (GeV) \\
400: \hline
401: 0-5 & 170/32 & 4684$\pm$231 & 0.171 & 1.071 & 170/32 & 4686$\pm$231 &
402: 0.159 & 1.066 & 0.041\\
403: 5-10 & 68/32 & 3393$\pm$184 & 0.176 & 1.068 & 67.8/32 & 3393$\pm$185 &
404: 0.165 & 1.064 & 0.041\\
405: 10-20 & 69/32 & 2767$\pm$144 & 0.171 & 1.073 & 69.2/32 & 2768$\pm$144 &
406: 0.160 & 1.068 & 0.042\\
407: 20-30 & 45/32 & 1928$\pm$102 & 0.169 & 1.075 & 44.7/32 & 1928$\pm$102 &
408: 0.157 & 1.070 & 0.042\\
409: 30-40 & 44/32 & 1391$\pm$78 & 0.165 & 1.078 & 43.9/32 & 1391$\pm$78 &
410: 0.153 & 1.072 & 0.041\\
411: 40-60 & 19/32 & 896$\pm$50 & 0.153 & 1.085 & 19.2/32 & 896$\pm$50 &
412: 0.141 & 1.079 & 0.040\\
413: 60-80 & 14/32 & 414$\pm$25 & 0.137 & 1.095 & 14.2/32 & 413$\pm$25 &
414: 0.125 & 1.087 & 0.037\\
415: $p+p$ & 9.7/29 & 62$\pm$7 & 0.117 & 1.099 & 9.62/29 & 61.9$\pm$7.1 &
416: 0.107 & 1.090 & 0.032\\
417: \hline
418: \hline
419: %\multicolumn{10}{|c|}{}\\
420: \multicolumn{10}{|c|}{PHENIX Coll. \cite{Adcox:2002pe}}\\
421: %\multicolumn{10}{|c|}{}\\
422: %&&&&&&&&&\\
423: \hline
424: &\multicolumn{4}{|c|}{ Eq.~\eqref{eq:2} with $Q=q/(1-q)$}
425: &\multicolumn{5}{|c|}{ Eq.~\eqref{eq:2} with $Q=1/(1-q)$}\\
426: &\multicolumn{4}{|c|}{ $\delta T_0=$0.011-0.016, $\delta
427: q=$0.008-0.010.} &\multicolumn{5}{|c|}{ $\delta T_0=$0.011-0.016,
428: $\delta q=$0.005-0.011,}\\
429: &\multicolumn{4}{|c|}{} &\multicolumn{5}{|c|}{ $\delta \Delta
430: T_0=$0.003-0.005.}\\
431: \hline
432: C.C. & $\chi^2$/n.d.f. & C & $T_0$ & $q$ & $\chi^2$/n.d.f. & C &
433: $T_0$ & $q$ & $\Delta T_0$ \\
434: (\%) & & & (GeV) & & & &
435: (GeV) & & (GeV) \\
436: \hline
437: 0-5 & 5.13/29 & 1694$\pm$409 & 0.201 & 1.049 & 5.1/29 & 1694$\pm$411 &
438: 0.192 & 1.047 & 0.042\\
439: 5-15 & 3.62/29 & 1330$\pm$316 & 0.199 & 1.051 & 3.6/29 & 1330$\pm$316 &
440: 0.190 & 1.048 & 0.042\\
441: 15-30 & 5.55/29 & 846$\pm$206 & 0.196 & 1.054 & 5.6/29 & 846$\pm$206 &
442: 0.186 & 1.051 & 0.042\\
443: 30-60 & 2.63/29 & 433$\pm$113 & 0.178 & 1.066 & 2.6/29 & 433$\pm$113 &
444: 0.167 & 1.074 & 0.045\\
445: 60-80 & 10.6/29 & 139$\pm$48 & 0.152 & 1.080 & 10.6/29 & 139$\pm$48 &
446: 0.141 & 1.062 & 0.035\\
447: 80-92 & 9.10/29 & 74$\pm$45 & 0.121 & 1.098 & 9.1/29 & 74$\pm$42 &
448: 0.110 & 1.089 & 0.033\\
449: \hline
450: \end{tabular}
451: \label{table:q-results}
452: \end{center}
453: \end{table*}
454:
455: We have analyzed BRAHMS \cite{Arsene:2003yk}, STAR
456: \cite{Adams:2003kv} and PHE\-NIX \cite{Adcox:2002pe} data and our
457: results are shown in Fig.~\ref{fig:2} and in
458: Table~\ref{table:q-results}. It turns out that both form of parameter $Q$
459: in Eq.~\eqref{eq:2} result in practically identical curves, therefore here
460: we are showing only results obtained for $Q=q/(1-q)$. The values of
461: parameters are also very close to each other with tendency of $C$,
462: $T_0$ and $q$ estimated by using $Q=1/(1-q)$ being slightly
463: bigger then those obtained for $Q=q/(1-q)$. It is worth to
464: stress at this point that such comparison of these two approaches has
465: been made for the first time here and, as one can see from the
466: presented results, it confirms previous expectation (made in
467: \cite{Wilk}) that in case of only limited phenomenological
468: applications, as is the case of our work, the results from using
469: Eq.~\eqref{eq:2} with $Q=q/(1-q)$ (i.e., parameters: $C^{(a)} = c$,
470: $T_0^{(a)}=l$ and $q^{(a)}=q$) are simply connected
471: to those using $Q=1/(1-q)$ (i.e., to parameters: $C^{(b)} = C$,
472: $T_0^{(b)}=L$ and $q^{(b)} = \hat{Q}$), namely:
473: \begin{equation}
474: \hat{Q} \simeq 1-\frac{1-q}{q},\qquad L\simeq \frac{l}{q},\qquad C\simeq cq
475: . \label{eq:Qq}
476: \end{equation}
477: As one can see from Table~\ref{table:q-results} these relations are indeed satisfied
478: (some small differences present could be attributed to the fact that both sets of
479: results represent results of separate and independent fitting procedures, without
480: making use of Eq.~\eqref{eq:Qq}). It means therefore that in all phenomenological
481: applications one can use either of the two form of parameter $Q$ in
482: Eqs.~\eqref{eq:2}, and, if necessary, to use Eq.~\eqref{eq:Qq} to translate results
483: from one scheme to another. In both cases the pion mass value has been used, $m=0.14$
484: GeV (and we have checked that additional changes in mass $m$ of the type introduced
485: recently in \cite{SB}, would not affect the final results as long as $m$ is limited
486: to, say, $m<0.2$ GeV). The estimated fluctuations of temperature are of the order of
487: $30-45$ MeV. It is interesting to observe that these fluctuations are weaker at small
488: centralities and grow for more peripheral collisions matching very nicely similar
489: fluctuations seen in $p+p$ data \cite{Adams:2003kv} shown here for comparison. One
490: should add here also result from similar analysis of $e^+e^-$ data
491: \cite{Bediaga:1999hv} reporting even higher values of nonextensivity parameter $q$
492: (reaching value of $q\simeq 1.2$), i.e., much stronger fluctuations. These results
493: confirm therefore, for the first time, another expectation made in \cite{Wilk} saying
494: that precisely such trend should be observed. This is because Eq.~\eqref{eq:qq} can
495: be also interpreted as being a measure of the \textit{total heat capacity} $C_h$ of
496: the hadronizing system (cf. \cite{Wilk}):
497: \begin{equation}
498: \frac{1}{C_h} = \frac{\sigma^2(\beta)}{\langle \beta\rangle^2} = \omega
499: = q-1 . \label{eq:cap}
500: \end{equation}
501: As the heat capacity $C_h$ is proportional to the volume, $C_h \sim V$,
502: in our case $V$ would be the volume of interaction (or
503: hadronization), it is expected to grow with volume and,
504: respectively, $q$ is expected to decrease with $V$, which is indeed
505: the case if one puts together results for $e^+e^-$, $p\bar{p}$ and
506: $AA$ collisions.
507:
508: %
509: % SECTION-3
510: %
511:
512: \section{Analysis of $\bm{k_t}$ distributions using stochastic approach in
513: $\bm{y_t}$ space} \label{sec:3}
514:
515: Whereas previous approach was concerned with extension of the purely statistical
516: approach the one presented now will go bit further by modelling hadronization process
517: by a kind of diffusion mechanism Refs.~\cite{Biyajima:2002at,B3,SB} in which the
518: original energy of projectiles is being dissipated in some well defined way into a
519: number of produced secondaries occurring in different part of the phase space
520: \footnote{It should be mentioned here that there exist also non-extensive versions of
521: such diffusion process applied to multiparticle production data \cite{qFP} but we
522: shall not pursue this possibility here.}. In the case considered here it is diffusion
523: process taking place in the $k_t$ space. Actually, it turns out that it is more
524: suitable to consider such diffusion as taking place in the $y_t = \sinh^{-1} (k_t/m)$
525: space. In this case one obtains the following Fokker-Planck equation:
526: %
527: \begin{eqnarray}
528: \frac{\partial P_t (y_t,t)}{\partial t}=\gamma \left[\frac{\partial
529: y_tP_t(y_t,\:t)}{\partial y_t}+\frac{\sigma_t^2}{2\gamma}\frac{\partial^2
530: P_t(y_t,t)}{\partial y_t^2}\right], \label{eq:3}
531: \end{eqnarray}
532: %
533: Its solution can be expressed by Gaussian distribution,
534: \begin{eqnarray}
535: \frac{d^2\sigma}{2\pi k_t dk_t} = CP_t (y_t,t) = \frac C{\sqrt{2\pi
536: V_t^2(t)}}\exp\left[-\frac{y_t^2}{2V_t^2(t)}\right] ,
537: \label{eq:4}
538: \end{eqnarray}
539: with\footnote{See also \cite{Biyajima:2002at}. Actually, Eq.
540: (\ref{eq:4}) is the same as the formula used already long time ago in
541: \cite{MC}.}
542: \begin{equation}
543: 2V_t^2(t) = \frac{\sigma_t^2}{\gamma}(1-e^{-2\gamma t}) .
544: \end{equation}
545:
546: \begin{figure*}
547: \begin{center}
548: \resizebox{0.94\textwidth}{!}{\includegraphics{b-ou.eps}}
549: \resizebox{0.94\textwidth}{!}{\includegraphics{s-ou.eps}}
550: \resizebox{0.94\textwidth}{!}{\includegraphics{p-ou.eps}}
551: \end{center}
552: \caption{Results of application of stochastic approach as given by
553: Eq.~\eqref{eq:4} to data for $k_t$-distributions at
554: $\sqrt{s_{NN}} = 200$ GeV measured for different
555: centralities by BRAHMS~\cite{Arsene:2003yk} and
556: STAR~\cite{Adams:2003kv} Collaborations and at
557: $\sqrt{s_{NN}} = 130$ GeV obtained by PHENIX Collaboration
558: \cite{Adcox:2002pe}. Notice that mass $m$ is treated here
559: as free parameter, in similar way as in \cite{SB}.}
560: \label{fig:3}
561: \end{figure*}
562:
563: %%% TABLE 2 %%%
564:
565: \begin{table}
566: \caption{Values of characteristic parameters used to fit data on
567: $k_t$-distributions at different centralities by using stochastic
568: approach as given by Eq.~\eqref{eq:4} and presented in
569: Fig.~\ref{fig:3} for data at $\sqrt{s_{NN}} = 200$ GeV obtained by
570: BRAHMS \cite{Arsene:2003yk} and STAR \cite{Adams:2003kv}
571: Collaborations and at $\sqrt{s_{NN}} = 130$ GeV obtained by PHENIX
572: Collaboration \cite{Adcox:2002pe}. The order of magnitude of the
573: corresponding errors for $T_0$, $\delta T_0$, are listed below as
574: well.}
575: \begin{center}
576: \begin{tabular}{ccccc}
577: \hline
578: &&&\\\multicolumn{5}{c}{BRAHMS Coll.; $\delta
579: T_0=$ 0.008-0.012; $\delta m =$ 0.024-0.031}\\
580: C.\ C.\ (\%) & $\chi^2$/n.d.f. & $C$ & $T_0$ (GeV) & $m$ (GeV) \\
581: \hline\
582: 0-10 & 39.9/23 & 140$\pm$9 & 0.201 & 0.784\\
583: 10-20 & 24.2/23 & 108$\pm$7 & 0.199 & 0.725\\
584: 20-40 & 20.1/23 & 72$\pm$5 & 0.196 & 0.671\\
585: 40-60 & 11.9/23 & 38$\pm$4 & 0.185 & 0.577\\
586: 60-80 & 4.06/22 & 4.3$\pm$0.5 & 0.184 & 0.515\\
587: \hline
588: \hline
589: &&&\\
590: \multicolumn{5}{c}{STAR Coll.; $\delta T_0=$ 0.004-0.006; $\delta m
591: =$ 0.009-0.014} \\
592: C.\ C.\ (\%) & $\chi^2$/n.d.f. & $C$ & $T_0$ (GeV) & $m$ (GeV) \\
593: \hline
594: 0-5 & 221/32 & 484$\pm$22 & 0.169 & 0.533\\
595: 5-10 & 124/32 & 370$\pm$18 & 0.170 & 0.547\\
596: 10-20 & 121/32 & 310$\pm$14 & 0.168 & 0.513\\
597: 20-30 & 92.9/32 & 217$\pm$10 & 0.168 & 0.498\\
598: 30-40 & 89.9/32 & 158$\pm$8 & 0.165 & 0.473\\
599: 40-60 & 43.9/32 & 99.6$\pm$5.0 & 0.157 & 0.419\\
600: 60-80 & 22.3/32 & 43.2$\pm$2.3 & 0.143 & 0.349\\
601: $p+p$ & 17.4/29 & 5.29$\pm$0.63 & 0.126 & 0.298\\
602: \hline
603: \hline
604: &&&\\
605: \multicolumn{5}{c}{PHENIX Coll.; $\delta
606: T_0=$ 0.020-0.037; $\delta m =$ 0.058-0.078} \\
607: C.\ C.\ (\%) & $\chi^2$/n.d.f. & $C$ & $T_0$ (GeV) & $m$ (GeV) \\
608: \hline
609: 0-5 & 8.06/29 & 161$\pm$34 & 0.185 & 0.734 \\
610: 5-15 & 5.61/29 & 124$\pm$26 & 0.185 & 0.729 \\
611: 15-30& 7.27/29 & 80$\pm$18 & 0.185 & 0.700 \\
612: 30-60& 3.49/29 & 39$\pm$9 & 0.177 & 0.594 \\
613: 60-80& 11.1/29 & 12$\pm$4 & 0.158 & 0.460 \\
614: 80-92& 9.01/29 & 6.1$\pm$3.6 & 0.131 & 0.317 \\
615: \hline
616: \end{tabular}
617: \label{table:stoch}
618: \end{center}
619: \end{table}
620:
621: In Fig.~\ref{fig:3} we show our results of using Eq.~\eqref{eq:4}. It should be
622: noticed that now, following \cite{SB}, we have regarded mass $m$ to be a free
623: parameter. Only then we can obtain good agreement with data. In a sense, variable
624: mass $m$ corresponds in this approach to the non-extensivity parameter $q$ introduced
625: in Section \ref{sec:2} in that it summarily accounts for some additional effects not
626: accounted by simple diffusion process (like, for example, effect of resonances and
627: flow).
628:
629: In the stochastic approach considered here we do not have direct
630: access to the temperature $T_0$. It is accessible only if we
631: additionally assume validity of the Einstein's
632: fluctuation-dissipation relation, which in our case means that
633: measure of the size of diffusion (dissipation), $V^2_t(t)$, can be
634: expressed by the temperature $T_o$ and mass $m$:
635: \begin{equation}
636: V^2(t) = \frac{T_0}{m} . \label{eq:fdt}
637: \end{equation}
638: Therefore our results for $V^2$ shown in Fig.~\ref{fig:3} (see inlets), where
639: $V(t)^2$ increases with increasing centrality, would indicate that temperature $T_0$
640: obtained by applying Einstein's relation with $m$ kept constant would increases with
641: centrality as well, contrary to what has been obtained above by applying
642: $q$-statistical approach. We have allowed then (following \cite{SB}) the mass $m$ to
643: be a free parameter and the best fit is obtained when $m$ decreases with centrality,
644: see inlets in Fig.~\ref{fig:3}. The resulting temperature, $T_0 \simeq m\cdot V_t^2$,
645: behaves then in essentially the same way as function of centrality as in the
646: $q$-statistical approach, cf., Table~\ref{table:stoch} and
647: Fig.~\ref{fig:4}\footnote{It should be stressed here that for constant value of mass,
648: $m=0.14$ GeV as used for $q$-statistics case above, we would have obtained somewhat
649: higher values of $\chi^2$'s. In addition, it is interesting to observe at this point
650: that the fact that we can fit data within modified stochastic approach only by
651: allowing for a kind of "quasiparticles" of mass $m$, different for different
652: centralities, corresponds in a sense to introducing parameter $q$ to the usual
653: statistical model. The possible dynamical origin and meaning of such variable mass
654: is, however, still lacking.}.
655:
656: %
657: % SECTION-4
658: %
659: \section{Concluding remarks} \label{sec:4}
660:
661: We have provided here systematic analysis of recent RHIC data on $k_t$ distributions
662: \cite{Arsene:2003yk,Adams:2003kv,Adcox:2002pe} by using three different kinds of
663: statistical approaches: Hagedorn model \cite{Hagedorn:1965aa}, two versions of the
664: modified statistical based on Tsallis statistics \cite{Tsallis:1988aa} and a suitable
665: adaptation of the stochastic model proposed in \cite{Biyajima:2002at}. We have found
666: that Hagedorn-type model
667: %
668: \begin{figure}
669: \resizebox{0.50\textwidth}{!}{\includegraphics{b-vmt3.eps}}
670: \resizebox{0.50\textwidth}{!}{\includegraphics{s-vmt3.eps}}
671: \resizebox{0.50\textwidth}{!}{\includegraphics{p-vmt3.eps}}
672: \caption{Comparison of temperatures of hadronization obtained by
673: using different approaches as given by: $(a)$ - Eq.
674: \eqref{eq:2} with $Q=q/(1-q)$; $(b)$ - Eq. \eqref{eq:2}
675: and $Q=1/(1-q)$; $(c)$ - Eq. \eqref{eq:4}. In the later
676: case $T_0$ has been obtained from the values of $V_t^2$
677: and $m$ obtained in Fig.~\ref{fig:3} by using Einstein's
678: relation: $T_0=m\cdot V^2(t)$.}
679: \label{fig:4}
680: \end{figure}
681: %
682: \cite{Hagedorn:1965aa} cannot fit data at large $k_t$ (its widely used for quick
683: estimations simplified version with $\rho(m) =1$, which is then just a simple
684: Boltzmann gas model with only one parameter, the temperature $T_0$, fails completely
685: even for smaller $k_t$, cf. Table~\ref{table:brahms}). However, these data can still
686: be reasonably well fitted either by non-extensive extensions of statistical model
687: \cite{Tsallis:1988aa} or by picture of some suitable diffusion process taking place
688: in transverse rapidity space \cite{Biyajima:2002at,B3}. This is specially true if one
689: limits itself to $k_t<5$ GeV/c range as the case of BRAHMS \cite{Arsene:2003yk} and
690: PHENIX \cite{Adcox:2002pe} data, the $k_t <12$ GeV/c range considered in STAR
691: experiment \cite{Adams:2003kv} seems to be already too big to be fitted properly even
692: with these two approaches (the corresponding values of the $\chi^2$ are considerably
693: bigger in this case and the values of parameters obtained for STAR and BRAHMS data,
694: which were taken at the same collision and at the same energy, are also different).
695:
696: %%% TABLE 3 %%%
697:
698: \begin{table*}
699: \caption{Comparison of investigated models: simple statistical model (i.e., Hagedorn
700: model as given by eq. (\ref{eq:1}) but with $\rho(m)=1$, in which case it is just a
701: simple statistical Boltzmann gas model with only one parameter, namely temperature
702: $T_0$), non-extensive Tsallis distribution (NETD) and Ornstein-Uhlenbeck process
703: (O-U), using data on $k_t$ distributions at $\sqrt{s_{NN}} = 200$ GeV obtained by
704: BRAHMS Collaboration \cite{Arsene:2003yk} for smallest and largest centralities.}
705: \begin{center}
706: \begin{tabular}{|cc|ccc|ccc|ccc|}
707: \hline
708: & & \multicolumn{3}{|c|}{Simple statistical model, }&
709: \multicolumn{3}{|c|}{NETD Eq.~\eqref{eq:2}}
710: &\multicolumn{3}{|c|}{ O-U Eq.~\eqref{eq:4}}\\
711: & & \multicolumn{3}{|c|}{Eq.~\eqref{eq:1} with $\rho(m)=1$} &
712: \multicolumn{3}{|c|}{(with $Q=q/(1-q)$)}&\multicolumn{3}{|c|}{}\\
713: %\hline
714: C.C (\%) & & $T_0$ & $q$ & $m$ & $T_0$ & $q$ & $m$ & $T_0$ &
715: $q$ & $m$ \\
716: & & (GeV) & & (GeV) & (GeV) & & (GeV) & (GeV) &
717: & (GeV)\\
718: \hline
719: 0-10 & $\chi^2$/n.d.f& \multicolumn{3}{|c|}{177/23} &
720: \multicolumn{3}{|c|}{10.2/23} & \multicolumn{3}{|c|}{39.9/23}\\
721:
722: & & 0.302 & --- & --- & 0.232 & 1.043 & --- & 0.201 & --- & 0.784 \\
723: \hline
724: 60-80 & $\chi^2$/n.d.f& \multicolumn{3}{|c|}{567/22} &
725: \multicolumn{3}{|c|}{2.76/22} & \multicolumn{3}{|c|}{4.06/22}\\
726: & & 0.325 & --- & --- & 0.175 & 1.084 & --- & 0.184 & --- & 0.515\\
727: \hline
728: \end{tabular}
729: \label{table:brahms}
730: \end{center}
731: \end{table*}
732:
733:
734: As is shown in Fig. \ref{fig:4}, the temperatures $T_0$ obtained in modified
735: statistical and stochastic approaches (with varying mass $m$) are essentially very
736: similar to each other and follow the same dependence on the centrality, namely $T_0$
737: decreases when collision is more peripheral. However, because stochastic approach
738: seems to be more dynamical than $q$-statistical one (where the true dynamical origin
739: of the nonextensivity parameter is not yet firmly established, see
740: \cite{Wilk,Wilk:1999dr}), we regard as the most valuable our finding that stochastic
741: approach \cite{Biyajima:2002at,B3,SB} works so well and can serve to provide first
742: simple estimations of any new data in the future. On the other hand we have also
743: demonstrated that the two possible approaches using $q$-statistics are equivalent to
744: each other, at least in the frame of limited phenomenological approach presented
745: here. One should also stress at this point that $q$ statistical approach offers
746: unique information on fluctuations in the system, which can be translated into
747: information on its volume. Our results for $AA$ and $pp$ collisions taken together
748: with old results for $e^+e^-$ annihilations indicate in this respect distinct growth
749: of the expected volume of interactions from the most elementary annihilation
750: processes to the nuclear collisions.
751:
752: %
753: % ACKNOWLEDGEMENT
754: %
755: \section*{Acknowledgements}
756: We are indebted to various conversations with participants at the RITP meeting (Kyoto
757: University, Aug. 2003) and at the RCNP meetings (Osaka University, Oct. 2003 and
758: November 2004). We also appreciate information on Ref. \cite{MC} provided us by W.
759: Zajc. This study is partially supported by Faculty of Science at Shinshu University.
760: Partial support of the Polish State Committee for Scientific Research (KBN) (grants
761: 621/E-78/SPUB/CERN/P-03/DZ4/99 and 3P03B05724 (GW)) is also acknowledged.
762:
763: \newpage
764:
765: \begin{thebibliography}{}
766:
767: \bibitem{Arsene:2003yk}% [1]
768: I.~Arsene \textit{et al.} [BRAHMS Collaboration],
769: Phys.\ Rev.\ Lett.\ \textbf{91}, (2003) 072305.
770: %[arXiv:nucl-ex/0307003].
771:
772: \bibitem{Adams:2003kv}% [2]
773: J.~Adams \textit{et al.} [STAR Collaboration],
774: Phys.\ Rev.\ Lett.\ \textbf{91}, (2003) 172302.
775:
776: \bibitem{Adcox:2002pe}% [3]
777: K.~Adcox \textit{et al.} [PHENIX Collaboration],
778: Phys.\ Lett.\ B \textbf{561}, (2003) 82.
779:
780: \bibitem{explanat}%[4]
781: Cf., for example, W. Broniowski and W. Florkowski,
782: Phys.\ Rev.\ Lett. \textbf{87}, (2001) 272302 and Acta Phys.\ Polon.\ B
783: \textbf{35}, (2004) 779 and references therein.
784:
785: \bibitem{Tsallis:1988aa}% [5]
786: C.~Tsallis, J.~Stat.~Phys. \textbf{52}, (1988) 479 and in \textit{
787: Nonextensive Statistical Mechanics and its Applications}, S.Abe and
788: Y.Okamoto (Eds.), Lecture Notes in Physics LPN560, Springer (2000).
789: For the most recent developments in this field see
790: http://tsallis.cat.cbpf.br/biblio.htm.
791:
792: \bibitem{Wilk}%[6]
793: W. M. Alberico, A. Lavagno and P.Quarati, Eur.\ Phys.\ J. C \textbf{12},
794: (2000) 499; O. V. Utyuzh, G. Wilk and Z. W\l odarczyk, J.\ Phys.\ G
795: \textbf{26}, (2000) L39; G. Wilk and Z. W\l odarczyk, Physica A
796: \textbf{305}, (2002) 227; F.~S.~Navarra, O.~V.~Utyuzh, G.~Wilk and
797: Z.~W\l odarczyk, Nuovo Cim. C \textbf{24}, (2001) 725;
798: Phys. Rev. D \textbf{67}, (2003) 114002; Physica A
799: \textbf{340}, (2004) 467; Physica A \textbf{344}, (2004) 568 and in Nukleonika
800: \textbf{49} (Supplement 2) (2004) S19.
801:
802: \bibitem{Bediaga:1999hv}%[7]
803: I.~Bediaga, E.~M.~Curado and J.~M.~de Miranda, Physica A \textbf{286},
804: (2000) 156.
805:
806: \bibitem{Wibig}%[8]
807: T.~Wibig and I.~Kurp, Int. J. High Energy Phys. \textbf{0312}, (2003) 039.
808:
809: \bibitem{Biyajima:2002at}%[9]
810: M.~Biyajima, M.~Ide, T.~Mizoguchi and N.~Suzuki, Prog.\ Theor.\
811: Phys.\ \textbf{108}, (2002) 559 [Addendum-ibid.\ \textbf{109}, (2003)
812: 151]; M.~Biyajima and T.~Mizoguchi, Prog.\ Theor.\ Phys.\ {\bf
813: 109}, (2003) 483; see also G.~Wolschin, Eur.\ Phys.\ J.\ A \textbf{5},
814: (1999) 85.
815:
816: \bibitem{B3}%[10]
817: M.~Biyajima, M.~Ide, M.~Kaneyama, T.~Mizoguchi and N.~Suzuki, Prog.\ Theor.\ Phys.\
818: Suppl.\ \textbf{154} (2004) 344; see also G.~Wolschin, Phys.\ Lett.\ B \textbf{569},
819: (2003) 67 and Phys.\ Rev.\ C \textbf{69}, (2004) 024906.
820:
821: \bibitem{MC}%[11]
822: Minh Duong-van and P.~Carruthers, Phys.\ Rev.\ Lett.\ \textbf{31}, (1973) 133.
823:
824: \bibitem{Hagedorn:1965aa}% [12]
825: R.~Hagedorn, Nuovo Cim.\ Suppl. \textbf{3}, (1965) 147; Nuovo Cim.\ A \textbf{52},
826: (1967) 64 and Riv. Nuovo Cim. \textbf{6}, (1983) 1983.
827:
828: \bibitem{HR}%[13]
829: R.~Hagedorn and J.~Ranft, Suppl. Nuovo Cim. \textbf{6}, (1968) 169.
830:
831:
832: \bibitem{SB}%[13]
833: N.~Suzuki and M.~Biyajima, Acta Phys. Polon.\, B \textbf{35}, (2004) 283.
834:
835: \bibitem{Wilk:1999dr}% [14]
836: G.~Wilk and Z.~Wlodarczyk, Phys.\ Rev.\ Lett.\ \textbf{84}, (2000) 2770.
837:
838: \bibitem{qHag}%[15]
839: C.~Beck, Physica\, \textbf{A286}, (2000) 164.
840:
841: \bibitem{ql1}% [16]
842: G.~Wilk and Z.~Wlodarczyk, Chaos,\ Solitons\ and\ Fractals {\bf
843: 13/3}, (2001) 581.
844:
845: \bibitem{superq}%[17]
846: C.~Beck and E.~G.~D.~Cohen, Physica\ A \textbf{322}, (2003) 267.
847:
848: \bibitem{escort}%[18]
849: C.~Tsallis, R.~S.~Mendes and A.~R.~Plastino, Physica\ A \textbf{261}, (1998)
850: 534.
851:
852: \bibitem{qFP}%[19]
853: A.~Lavagno, Physica\ A \textbf{305}, (2002) 238; M.~Rybczy\'nski,
854: Z.~W\l odarczyk and G.~Wilk, Nucl.~Phys.~(Proc.~Suppl.) B\ \textbf{122},
855: (2003) 325.
856:
857: \end{thebibliography}
858:
859:
860: \end{document}
861: