hep-ph0309052/apm.tex
1: %%%%%%%%%%%% tex/notes/apm.4.tex %%%%%%%%%%%%%
2: %
3: %	APM: additive Parton Model
4: %	Notes for a Letter
5: %	T.S.Biro	07.08.2003
6: %	09.08.2003	% First draft version of PRL
7: %	09.08.2003	% added References + text corrections
8: %	10.08.2003	% modified constituent philosophy
9: %	13.08.2003	% leave out Fig.3 and Fig.4
10: %			% put in entropy formula + picture
11: %	14.08.2003	% concentrate on constituent dressing
12: %	16.08.2003	% Tsallis distr, relative yields of n=1 and n=2
13: %	28.08.2003	% minor corrections, new refs. about PDF
14: %	01.09.2003	% emphasize the good side, leave out Fig.4, etc.
15: %	03.09.2003.	% small changes in text
16: %	04.09.2003.	% Berndt's text changes, new refs.
17: %
18: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
19: %\documentclass[10pt]{article}
20: %\documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb,floatfix]{revtex4}
21: \documentclass[prl,twocolumn]{revtex4}
22: %\documentclass[prl,preprint]{revtex4}
23: 
24: %%%%%%%%%%%%%%%%%%%%% PACKAGES %%%%%%%%%%%%%
25: 
26: \usepackage{graphicx}% Include figure files
27: \usepackage{dcolumn}% Align table columns on decimal point
28: \usepackage{bm}% bold math
29: 
30: \usepackage{graphics}
31: \usepackage{graphicx}
32: \usepackage[dvips,usenames]{color}
33: 
34: %%%%%%%%%%%%%%%%%%%%%%%% MACROS %%%%%%%%%%%%%%%
35: 
36: %\newcommand{\vs}{\vspace{3mm}}
37: \newcommand{\vs}{\quad}
38: \newcommand{\tc}[2]{\textcolor{#1}{#2}}
39: \newcommand{\cb}[2]{\colorbox{#1}{#2}}
40: \newcommand{\fb}[3]{\fcolorbox{#1}{#2}{#3}}
41: 
42: \newcommand{\be}{\begin{equation}}
43: \newcommand{\ee}{\end{equation}}
44: \newcommand{\ba}{\begin{eqnarray}}
45: \newcommand{\ea}{\end{eqnarray}}
46: \newcommand{\NL}{\nonumber \\}
47: 
48: 
49: \parindent0mm
50: \fboxsep3mm
51: \fboxrule1mm
52: 
53: \begin{document}
54: 
55: %%%%%%%%%%%%%%%%%%%%%%%%% TITLE PAGE %%%%%%%%%%%%%%%%%%
56: 
57: \title{Almost exponential transverse spectra from power law spectra}
58: \author{Tam\'as S. Bir\'o}
59: \affiliation{KFKI, Research Institute for
60: Particle and Nuclear Physics, H-1525 Budapest, Hungary}
61: \author{Berndt M\"uller}
62: \affiliation{Physics Department, Duke University, Durham, NC-27708}
63: \date{printed \today}
64: 
65: 
66: \begin{abstract}
67: \vs
68: {\bf Abstract}
69: We point out that exponential shape of transverse spectra can be
70: obtained as the Fourier transform of the limiting distribution  
71: of randomly positioned partons with power law spectra given by pQCD,
72: which actually realize Tsallis distributions. 
73: Such spectra were used to obtain hadron yields by recombination in relativistic
74: heavy-ion collisions at RHIC energies.
75: 
76: \end{abstract}
77: \pacs{}
78: \maketitle
79: 
80: %%%%%%%%%%%%%%%%%%%%%%%%%  MAIN  TEXT BODY %%%%%%%%%%%%%%%%%%%%%%%
81: 
82: \vs
83: In relativistic heavy-ion collisions
84: %as well as in $pp$ and $pA$ collisions 
85: exponential transverse momentum distributions
86: has been observed at low to mid transverse momenta, $p_T \approx 1 - 4$ GeV
87: \cite{PHENIX,STAR,NA49}.
88: This fact was often interpreted as an indication that the source
89: of the particles can be described by a temperature.
90: This view climaxed in the ''thermal model'' which
91: assumes a situation well described by equilibrium thermodynamics
92: \cite{THERMAL}.
93: 
94: \vs
95: Regarding experimental data the exponential law is most prominent
96: in the transverse mass spectra which fit the form $\exp((m-m_T)/T)$
97: with $m_T=\sqrt{m^2 + p_T^2}$ at mid-rapidity at SPS and RHIC
98: energies quite well. A tendency of increasing slopes with
99: increasing hadron masses in these spectra is usually interpreted
100: as a sign of flow, on the basis of the relation $T = T_0 + m\langle u^2\rangle$
101: \cite{FLOW}.
102: A fit to pion, kaon and proton spectra at RHIC gives
103: $T_0 \approx 140$ MeV and $\langle u^2 \rangle \approx 0.5 - 0.6$
104: at freeze out\cite{RHIC-FLOW}.
105: A possible interpretation of this fit is a common flow pattern and a
106: {\em common temperature}, $T_0$, for all hadrons produced in the 
107: heavy-ion collision. This value, converted back to no-flow by using
108: the above formula for pions gives about $T \approx 210 $ MeV, 
109: which is somewhat above the color deconfinement
110: temperature predicted by lattice QCD calculations \cite{LQCD}.
111: %Thermal model estimates not considering the radial flow on the other
112: %hand obtain $T \approx 175$ MeV and $\mu_B \approx 45$ MeV at 
113: %RHIC experiments\cite{RECENT-THERMAL}.
114: 
115: \vs
116: Since there is no known hadronic process which could thermalize 
117: on such a short timescale at which this thermal state
118: must have developed, it is tempting to consider {\em quark level} 
119: mechanisms as reason for the observed exponential spectra.
120: 
121: \vs
122: Constituent quark and parton recombination models have been utilized
123: as a picture of hadron formation in relativistic heavy-ion collisions
124: in recent years \cite{ALCOR,RECOMB}. 
125: This idea roots in the constituent quark model of
126: hadrons originally combining mesons from two, baryons from three
127: quark-like partons. In order to obtain an exponential hadron spectrum
128: by this mechanism, one needs to start with parton spectra which
129: are already exponential in the intermediate $p_T$ regime of $2 - 4$
130: GeV. At higher transverse momenta these spectra go over into power laws,
131: as given by pQCD calculations. According to the respective regions of
132: dominance exponential and power law parton distributions are simply
133: added as a phenomenological effort to generate the desired experimental
134: hadron spectra.
135: 
136: \vs
137: In this letter we present a mechanism, by which power law
138: spectra of partons can be combined to give an exponential  spectrum.
139: For this process, one needs several partons to recombine to a
140: constituent quark in contrast to the final step of hadron
141: recombination where only two or three such dressed quarks
142: are used.
143: In the following we show that the exponential can be constructed as the
144: {\em limiting distribution} of cut power laws in the form
145: \be
146:   w(E) = a \left( 1 + \frac{E}{b} \right)^{-c}
147: \label{pQCDPART}
148: \ee
149: with parameters $a$, $b$ and $c$ fitted to pQCD calculation results
150: as in Refs.\cite{PARTON}. Here we modified somewhat the original
151: formula by using $E$, the parton energy, instead of the transverse
152: momentum $p_T$. At zero rapidity it is, however, equal to the transverse
153: mass $m_T$, and for large transverse momenta to $p_T$:
154: \be
155:  E = \cosh y \sqrt{p_T^2 + m^2}
156: \ee
157: For large transverse momenta $w(E)$ goes over into a power law spectrum: 
158: $w \approx ab^c \, |p_T|^{-c}$.
159: 
160: %\vs
161: %In certain cases the limiting distribution is not Gaussian. This happens
162: %if the primordial distribution has a long tail (and therefore some
163: %divergent higher moments, like the Lorentz distribution). A power law
164: %certainly belongs to this category.
165: 
166: \vs 
167: Before turning to the combination of power laws let us briefly recall
168: the familiar case of the central limit theorem \cite{CLIM}. 
169: Short tailed probability
170: distributions all lead to a Gaussian probability distribution for the
171: sum of random deviates in the limit of infinitely many independent
172: draws. It is easy to demonstrate this by considering uniform random deviates
173: $x_i$. Then $w(x_i)=0.5$ if $x_i$ is in the interval $[-1,1]$ and zero
174: otherwise. We seek the distribution $P_n(x)$ of the scaled sum,
175: \be
176: x = \sqrt{\frac{3}{n}} \sum_{i=1}^n x_i.
177: \ee
178: It is given by the n-fold integral
179: \be
180:  P_n(x) \: = \: 
181:    \prod_{i=1}^n \left( \frac{1}{2} \int_{-1}^{1}\!\! dx_i \right) \,
182:    \delta\left(x-\sqrt{\frac{3}{n}}\sum_{j=1}^n x_j\right).
183: \ee
184: In order to derive the Gaussian in the $n \rightarrow \infty$ limit
185: we use its Fourier transform,
186: \be
187:  \tilde{P}_n(k) \: = \: \int_{-\infty}^{+\infty} \!\! dx \,
188:   e^{ikx} \, P_n(x).
189: \ee
190: The $x$-integral is easily done, and the remainder
191: factorizes into $n$ equal contributions,
192: \be
193:  \tilde{P}_n(k) \: = \: \left(\frac{\sin(k\sqrt{3/n})}{k\sqrt{3/n}} \right)^n
194: \ee
195: In the large $n$ limit we arrive at
196: \be
197:  \tilde{P}_{\infty}(k) = \lim_{n\rightarrow\infty} \,
198:  \left( 1 - \frac{k^2}{2n} \right)^n \: = \: e^{-k^2/2}.
199: \ee
200: The inverse Fourier transformation results in the standard Gaussian
201: \be
202:  P_n(x) = \frac{1}{\sqrt{2\pi}} \, e^{-x^2/2}.
203: \ee 
204: Fig.\ref{FigGauss} demonstrates that this limit is approached very
205: quickly, already the sum of $n=3$  uniform random deviates
206: is distributed nearly Gaussian.
207: 
208: %%%%%%%%%%%%%%%%%%%% FIG1: GAUSS HISTOGRAM FIGURES %%%%%%%%%%%%%%%%%%%%%%
209: %%%%% VERSION 1: 4 pictures in 1
210: %\newpage
211: %\vspace*{-20mm}
212: %\begin{figure}
213: %\begin{center}
214: %\includegraphics[width=0.25\textwidth]{G1.ps}\includegraphics[width=0.25\textwidth]{G2.ps}
215: %\includegraphics[width=0.25\textwidth]{G3.ps}\includegraphics[width=0.25\textwidth]{G4.ps}
216: %\end{center}
217: %\caption{\label{FigGauss}
218: % Comparison of histograms of $m = 200 000$ sums of $n$ uniform
219: %  random deviates in $(-1,1)$ scaled with $\sqrt{3/n}$ and the
220: %   limiting Gauss distribution $\frac{1}{\sqrt{2\pi}}\exp(-x^2/2)$.
221: %   }
222: %\end{figure}
223: 
224: %%% VERSION 2: one picture for all lines
225: %\newpage
226: \begin{figure}[hb]
227: \begin{center}
228: \vspace*{-40mm}
229: \includegraphics[width=0.4\textwidth]{GaussAsSum.ps}
230: \end{center}
231: \caption{\label{FigGauss}
232:  Comparison of histograms of $m = 200 000$ sums of $n$ uniform
233:   random deviates in $(-1,1)$ scaled with $\sqrt{3/n}$ and the
234:    limiting Gauss distribution $\frac{1}{\sqrt{2\pi}}\exp(-x^2/2)$.
235:    }
236: \end{figure}
237: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
238: 
239: For the general case one uses an arbitrary distribution, $w(x_i)$
240: normalized to one, $\int w(x) dx = 1$. The central moments of
241: a distribution are given by higher derivatives of the logarithm
242: of the Fourier transform. For those of the $n$-fold sum we have
243: \be
244:  c_j(n) = \left(\frac{1}{i} \frac{\partial}{\partial k}\right)^j 
245:  	\left. \log \tilde{P}_n(k) \right|_{k=0}
246: \ee
247: Since $\log \tilde{P}_n(k) = n \log \tilde{w}(ka_n)$ when looking
248: for the distribution $P_n(x)$ of the scaled sum $x=a_n \sum_i x_i$,
249: we arrive at the following scaling law for the central moments
250: \be
251:  c_j(n) = n a_n^j c_j(1).
252: \ee
253: With a choice of $a_n \propto 1/\sqrt{n}$ one assures that
254: \be
255:  \lim_{n \rightarrow \infty} \,  c_j(n) = 0, \qquad {\rm for} \quad j>2.
256: \ee
257: Consequently the limiting distribution, $P_{\infty}(x)$, is Gaussian.
258: 
259: \vs
260: Now we consider the dressing of constituent quarks with gluons. 
261: One may  assume that the constituent energy and mass coalesces from
262: $n$ equal, partonic amounts -- a kind of additive parton model -- but
263: this is not necessary.
264: The total mass in this case would be $M = \sum E_i$, and each parton
265: would carry a fraction of the transverse momentum, 
266: $\lambda_i p_T = {\cal O}(p_T/n)$.
267: In the following we assume that the spectral shape of the 
268: constituent partons (recombining later to hadrons) is proportional
269: to the Fourier transform of the composite wave function.
270: Further we assume, that such a constituent quark (e.g. an up quark)
271: is composed of a bare up quark (the seed) and $n$ gluons (in general,
272: sea partons). Since these partons are almost massless, we consider
273: the center of energy as the proper coordinate for the constituent wave
274: function:
275: \be
276:  x = \frac{\sum E_ix_i}{\sum E_i} =  \sum \lambda_i x_i
277: \ee
278: with $\sum \lambda_i = 1$. By equal sharing of the total energy,
279: which is the case whenever the fusion happens at low relative
280: momenta, it is the arithmetic mean (with $1/n$ scaled sum) 
281: of random parton coordinates. 
282: Its distribution, $P_n(x)$, is interpreted in the present model as
283: the constituent quark wave function squared. Its Fourier
284: transform is proportional to the momentum spectrum and is obtained as
285: $n$-fold product of individual parton spectra, each carrying a
286: fraction of the total energy. For the sake of simplicity representing
287: all bare parton distribution $\tilde{w}$ by the same distribution,
288: we have, in the general case, $n$ contributions: 
289: \be
290:  \tilde{P}_n(k) \: = \: \prod_{i=1}^n \tilde{w}(\lambda_i k)
291: \ee
292: with $\sum_1^n \lambda_i = 1$.
293: Considering cut power law parton spectra as given by eq.(\ref{pQCDPART})
294: the $n$-fold spectrum at $y=0$ rapidity becomes
295: \be
296:  \tilde{P}_n(p_T) \: = \: \prod_{i=1}^n 
297:  	a_i \left(1+\frac{\lambda_i M_T}{b} \right)^{-c}.
298: \ee
299: Since all $\lambda_i$ are positive factors and their sum is normalized to one,
300: the limiting distribution for $n \rightarrow \infty$ 
301: enforces that all $\lambda_i \rightarrow 0$ and hence results in 
302: an {\em exponential} distribution, normalized to $\tilde{P}_{\infty}(0)=1$:
303: \be
304:  \tilde{P}_{\infty}(p_T) \: = \: 
305:    \exp \left(\frac{M-M_T}{b/c} \right).
306: \label{MAIN-RESULT}
307: \ee
308: \vs
309: This is the central result of the present letter.  It predicts a spectral slope
310: determined by pQCD as being $T = b/c$. From standard fits to the pQCD parton
311: spectra one gets $T \approx 170 - 210$ MeV, depending on the flavor under
312: consideration and details of the fit to the power law distribution from pQCD. 
313: This slope has nothing to do with the lattice QCD phase transition.
314: 
315: 
316: %%%%%%%%%%%%%%% FIG 2: LIMITING DISTRIBUTION OF CUT POWER LAWS %%%%%
317: \newpage
318: \begin{figure}[htb]
319: \begin{center}
320: \vspace*{-30mm}
321: \includegraphics[width=0.48\textwidth]{nparton.ps}
322: \end{center}
323: \caption{\label{FigLIM}
324:   $\exp(-|p_T|c/b)$,
325:   the limiting distribution of cut power law Fourier transforms
326:   $(1+|p_T|/b)^{-c}$ and its finite $n$ approximations. 
327:   The value {$ n = 20 $} is characteristic for the pion.
328: }
329: \end{figure}
330: 
331: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
332: %%%%%%%%%%%% Insert Tsallis stuff ... %%%%%%%%%%%%%%%
333: \vs
334: Our result can be understood by noticing that the cut power law
335: parton distribution of eq.(\ref{pQCDPART}) is the canonical
336: Tsallis distribution \cite{TSALIS} with a temperature $T = b/c$
337: and $q=1-1/c$. This distribution maximizes the Tsallis entropy
338: with the usual canonical constraint $\sum_i E_i w(E_i) = const$.
339: The down-scaled product of $n$ such distributions is still of the
340: Tsallis form at the same temperature $T=b/c$ but it belongs to
341: the parameter $q = 1 - 1/(cn)$. In the $n \rightarrow \infty$ 
342: limit $q\rightarrow 1$
343: and the usual canonical distribution is recovered with
344: the familiar exponential spectrum. By using the rather large values
345: for $c$ usually obtained by fitting pQCD results, namely $c=8$,
346: one obtains the following series: 
347: $q_1 \approx 0.875$, $q_2 \approx 0.94$, $q_3 \approx 0.96$, etc.
348: 
349: %%%%%%%%%%%%%%%%%%% END TSALLIS insertion %%%%%%%%%%%%%%%%%%%
350: 
351: %%%%%%%%%%%%%%%%% Insertion of discussion %%%%%%%%%%%%%%%%%
352: \vs
353: It is interesting to recall here another physical application
354: of the Tsallis distribution. Imperfectly thermalized systems,
355: e.g. a finite number of oscillators, also lead to a $q$ value
356: not equal to one. In the case of $(n+1)$ oscillators, however,
357: one arrives at a value bigger than one: $q=1+1/n$\cite{TSALIS}.
358: One obtains a distribution reminiscent of the discretized
359: path integral sum,
360: \be
361:  w_i = \frac{1}{Z} \left(1 - \frac{\varepsilon_i}{nT} \right)^n,
362: \ee
363: while primordial parton distributions belong to a $q$ value less than one,
364: \be
365:  w_i = \frac{1}{Z} \left(1 + \frac{\varepsilon_i}{nT} \right)^{-n}.
366: \ee
367: Of course, both distributions approach the Gibbs distribution
368: as $n\rightarrow\infty$.
369: 
370: \vs
371: We interpret the high power $c \approx 8-9$ in pQCD results as an
372: accidental value simulating an ''almost thermal'' distribution
373: with an effective $q \approx 0.82 - 0.89$ near to one. This explains
374: why the observed hadron spectra at low and intermediate $p_T$
375: appear nearly exponential in heavy-ion and many other collisions.
376: Parton recombination to hadrons amplifies this effect leading to
377: $q = 1 - 1/(cn)$ by  $n$-fold recombination. The Tsallis distribution
378: comes very close to the Gibbs distribution. As we shall argue in
379: the following, recombination processes are faster and more effective
380: with increasing phase space volume.
381: Consequently heavy-ion collisions are the best experiments to study
382: strongly interacting matter (QCD) as close to the canonical distribution
383: as possible. (It is a further interesting question whether the
384: primordial parton distribution is described by a smaller power $c$ in $e^+e^-$
385: collisions than in heavy-ion reactions. This fact could shed a new
386: light onto thermal model interpretation of hadron production in
387: $e^+e^-$ reactions \cite{Becattini}).
388: \vs
389: A remaining question is that how large $n$ has to be for $P_n$  to
390: look exponential. Fig.\ref{FigLIM} shows this behavior with 
391: the parameters $b=2$ GeV, $c=8$ (causing $T = 250$ MeV).
392: Here the convergence is slower than in the case of the sum of uniform
393: random deviates approaching a Gaussian.
394: The value $n=20$ would be characteristic for pion from additive
395: mass estimate $M_{\pi}/m_q \approx 21.5$ with $m_q=(m_u+m_d)/2$.
396: 
397: %%%%%%%%%%%%%%%%%%%%%%%% RATE EQ. for n=2 insertion %%%%%%%%%%%
398: 
399: \vs
400: Another question is in which proportions the constituent partons, 
401: used for recombination to mesons and baryons, may consist of 
402: contributions from clusters of perturbative partons of different
403: complexity $n$.  Our suggestion
404: is, that one may not necessarily need to go far in $n$, i.e. to
405: use already exponential spectra for recombination, in order to
406: have an effective parton distribution close to the usually applied
407: one. In particular, and for the case of simplicity, we discuss
408: the $n=2$ case here: it improves already the original $n=1$
409: distribution a lot in the direction of the assumed exponential.
410: 
411: \vs
412: Fusion of a quark with a gluon and its radiation may be considered
413: as dynamical processes between $n=1$ and $n=2$ parton clusters.
414: In order to find an equilibrium ratio of these two type of partons
415: one inspects a rate equation. 
416: The detailed balance leads to the familiar proportionality applied
417: in recombination models:
418: \be
419:   f_2(E) = \frac{\gamma}{\Gamma} f_1^2(E/2).
420:   \label{BALANCE}
421: \ee
422: The effective spectrum we consider is given by
423: \be
424:  f(E) = A_1 w(E,b,c) + A_2 w^2(E, 2b, 2c).
425: \ee
426: Inspecting  the sum of thermal and original power law parton
427: distribution yields used for recombination by Fries. et al.
428: Ref.\cite{RECOMB}, one concludes that the ratio
429: $A_2/A_1 \approx 0.15$ gives already a surprisingly good approximation
430: in the relevant $p_T$ range from $b \approx 1.5$ GeV to $12$ GeV
431: (cf. Fig.\ref{PYIELD}). 
432: We note that in the original spectra parton energy loss
433: and a blue shift of the temperature of the thermal part
434: due to radial flow have been included. For our figure we 
435: used the same single parton yield, scaled down by a factor $0.87$ and
436: added its square at $p_T/2$ with weight $0.13$. The corresponding
437: Tsallis ''temperature'' is $b/c = 190$ MeV.
438: 
439: 
440: %%%%%%%%%%%%%%%%%%%% FIG 3: PARTON YIELDS %%%%
441: %\newpage
442: \vspace*{-30mm}
443: \begin{figure}[th]
444: \begin{center}
445: \vspace{-10mm}
446: \includegraphics[width=0.45\textwidth]{parton-yield.ps}
447: \end{center}
448: \caption{\label{PYIELD}
449:  The thermal and power law parton yield components used by Fries. 
450:  et al.\cite{RECOMB},
451:  and the sum of the two contributions is compared to the 13-87\% 
452:  mixture of $n=1$ and $n=2$ parton clusters, using the power law
453:  distribution only.
454: }
455: \end{figure}
456: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
457: 
458: 
459: %%%%%%%%%%%%%%%%%%%%%%%%%% END RATE EQ. insertion %%%%%%%%%%%
460: 
461: 
462: \newcommand{\qh}{{q\rightarrow h}}
463: \newcommand{\nqh}{{n_{\qh}}}
464: \vs
465: For realistic meson and baryon formation a further recombination should be
466: applied for each constituent of a hadron, possibly with different flavor,
467: in the final hadronization process. 
468: In a heavy-ion collision event many processes go on
469: in parallel. While some quarks are dressing up to become constituent quarks,
470: some other partons are fragmenting to lower $p_T$ hadrons
471: directly. The hadron $p_T$ spectra then are composed at
472: least from these two components, but maybe also of ''partially dressed''
473: quarks if a moderate $n$ in the production of their center of
474: energy from random positioned gluons appear. These may contribute
475: at intermediate $p_T$ to the hadron formation by recombination
476: as well; their yield is, however, reduced compared to the ones
477: behaving nearly exponentially.
478: 
479: \vs
480: The competition between fragmentation and recombination may in general
481: ease the entropy problem of the pure recombination models,
482: where finally one deals with fewer hadrons than partons. 
483: %In the case of (earlier) quark dressing to a (qg) system the problem
484: %is much less severe, because at this step one does not eliminate
485: %color degrees of freedom by selecting out color singlet clusters,
486: %like one does it for hadronization. 
487: 
488: 
489: \vs
490: We analyze the entropy problem now for the final hadronization step.
491: As an approximation established by the above discussion, we consider
492: the recombination of exponential spectra from exponential precursors.
493: The entropy can be obtained using the Boltzmann formula (the $q=1$
494: case of the more general Tsallis entropy)
495: \be
496: S = \int (-f \ln f) \frac{d^3p}{(2\pi)^3} V
497: \ee
498: while the number of particles are given by
499: \be
500: N = \int \: f  \frac{d^3p}{(2\pi)^3} V.
501: \ee
502: The distribution $f(p)$ we consider is proportional to the Fourier
503: transform of the probabilities. Since they were normalized as $\tilde{P}_n(0)=1$,
504: here we have an unknown factor related to total numbers: $f=a\tilde{P}$.
505: For example for exponential spectra to combine, one gets
506: $f^n(p/n)=a^{n-1}*f(p)$. This leads to the familiar coalescence formula
507: $N_n = C_n N_1^n$ (with $C_n$ being proportional to $(VT^3)^{1-n}$) as well.
508: The $n$-cluster entropy becomes
509: \be
510: {S_n} =  (d-n \ln a) \, N_n,
511: \label{CLUSTER_ENT}
512: \ee
513: due to $\langle p \rangle/T=d$ when
514: considering $d$-dimensional momentum exponentials. At the same time the relative
515: number of $n$-clusters is related to the parameter 
516: $a$ as ${N_n}/{N_1} = a^{n-1}$.
517: 
518: 
519: %%%%%%%%%%%%%%%%%%%% FIG 3: ENTROPY RATIO %%%%
520: %\newpage
521: %\vspace*{-30mm}
522: %\begin{figure}[ht]
523: %\begin{center}
524: %\includegraphics[width=0.45\textwidth]{rec-ent-ratio.ps}
525: %\end{center}
526: %\caption{\label{ENTROPY}
527: %  The ratio of $n$-cluster entropy to the single quark entropy
528: %  for the exponential transverse mass distribution.
529: %}
530: %\end{figure}
531: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
532: 
533: %Fig.\ref{ENTROPY} shows the entropy ratio as a function of the
534: %number ratio for this case. One realizes that the point $N_n=N_1$
535: %is exceptional: here the entropy ratio is one for any cluster.
536: %This $a=1$ case can, however, be never fulfilled purely by recombination.
537: 
538: 
539: \vs
540: The sum of all $n>1$-cluster entropies is then the entropy of
541: the recombined state, $S_H$, while $S_Q=S_1$ is that of the primordial state.
542: From the phase space integrals one gets
543: \be
544: S_H = \sum_{n>1} S_n = d \sum_{n>1} N_n - \ln a \sum_{n>1} (n N_n),
545: \ee
546: where the number of quarks redistributed in hadrons occurs as the
547: last term in the expression:
548: \be
549: \sum_{n>1} n N_n = N_1 = N_Q
550: \ee
551: is the recombination sum rule (counting $n$ quarks in an 
552: $n$-fold cluster). While this rule is quite strict
553: in the recombination of constituents, for dressing up quarks with gluons,
554: of course, such a constraint is not valid. Whenever this
555: sum rule applies, one obtains
556: \be
557: S_H = S_Q + d \left( N_H - \sum_{n>1} n N_n \right),
558: \ee
559: when expressing $ \ln a$ from the $n=1$ case of eq.(\ref{CLUSTER_ENT}).
560: Since the recombined sum goes over $n>1$ indices, and all $N_n$
561: are by definition positive, $\sum S_n < S_1$ is unavoidable
562: by the recombination of exponential spectra. 
563: With other words, after recombination there are fewer hadrons than
564: constituents built into hadrons.
565: The opening of several
566: different flavor channels may ease this problem, but cannot completely solve it.
567: The missing entropy is either generated by i) volume expansion
568: (i.e. the flow must be generated or accelerated by the recombination
569: process) or by ii) fragmentation, when a single parton produces
570: several hadrons (as it is known from $e^+e^-$ 2-jet events),
571: or by iii) the excitation of internal degrees of freedom of the hadrons.
572: 
573: 
574: \vs
575: %{\bf Conclusion} \quad
576: Based on these principles all hadronic transverse momentum distributions
577: measured at RHIC can be worked out as outlined in this letter.
578: The previously assumed exponential parton spectra now are {\em derived} in the 
579: intermediate $p_T$ range by folding pQCD power law spectra. 
580: No assumption of a heat bath, a temperature
581: or thermal equilibrium is required. We present this scenario as an alternative
582: to, and at the same time a possible explanation for, the remarkable
583: phenomenological success of the thermal model.
584: 
585: \vs
586: We do not believe, however, that this result reduces the usefulness of
587: heavy-ion collisions in the search for quark matter. Asymmetric
588: jet quenching and the meson -- baryon scaling of the azimuthal
589: flow ($v_2$) indicate the presence of a collective state
590: with properties of a highly absorptive, perhaps colored, medium. The present
591: explanation of exponential transverse spectra from pQCD and random
592: statistics phenomenology offers an alternative to the 
593: ad hoc assumption of a thermal state. 
594: In fact, the independence of dressing gluons when combined to
595: a constituent parton is an ingredient of getting the
596: exponential $p_T$ distribution. This condition is probably satisfied only
597: in heavy-ion collisions.
598: Our model also calls attention to the fact that
599: ''statistical'' is not always synonymous with equilibrium Gibbs thermodynamics.
600: Certain dynamical phenomena, like e.g. self organizing criticality
601: and -- what we have just discussed -- limiting parton distributions, 
602: belong to a more general framework of statistical physics.
603: 
604: \vs
605: The exponential $m_T$-spectra in fact approach at low $p_T$
606: a Gaussian, $\exp(-p_T^2/(2mT))$, the textbook case of a thermal
607: distribution. It is interesting to note that several recent
608: articles consider a Gauss distributed intrinsic parton $p_T$
609: in the generalized parton distribution functions (GPD-s) \cite{GPD}.
610: The Tsallis distribution predicts $\frac{1}{2}\langle p_T^2 \rangle =
611: m_T(b+m_T)$ leading to $\langle p_T^2 \rangle \approx 1$ GeV$^2$,
612: if using a constituent mass of $m_T=0.3$ GeV and the cut-off $b=1.5$ GeV,
613: in good agreement with recent theoretical analysis of high energy
614: $pp$ data.
615: 
616: \vs
617: We have seen, however, that the parameter $T$ in
618: this distribution is not related to a heat bath or another
619: source of statistical fluctuations. It is combined from two pQCD
620: parameters, from a cut-off $b$ and a power $c$ as being $T=b/c$.
621: The ''heating agent'' in a general sense here is the effective number
622: of independent dressing gluons per constituent, $n$, 
623: which should lead to a Tsallis distribution with $q=1-1/(cn)$
624: quite close to an exponential at low $p_T$.
625: If the transverse positions of the dressing partons are averaged 
626: as independent random
627: variables, the distribution of their center of energy will have a
628: Fourier spectrum, which is nearly exponential up to $p_T\approx 3$ GeV.
629: 
630: \vs
631: {\bf Acknowledgment} \quad
632: Discussions with J. Zim\'anyi and R. Fries are gratefully
633: acknowledged. Special thanks to C.Nonaka for making available
634: pQCD fit parameters and fragmentation function parameterizations.
635: T.S.B. thanks the hospitality of Duke University which was
636: expressed towards him during his stay in Durham.
637: This work has been supported in part by DOE grant
638: FG02-96ER-40945 and by a grant of the
639: Hungarian National Research Fund OTKA (T034269).
640: 
641: 
642: %\vspace{-20mm}
643: 
644: \begin{thebibliography}{xxxxx}
645: 
646: \bibitem{PHENIX} PHENIX collaboration, nucl-ex/0307022, nucl-ex/0307010,
647: 	nucl-ex/0304022, nucl-ex/0212014, Acta Phys.Hung. Heavy Ion Physics
648: 	{\bf 15}, 291, (2002).
649: 
650: \bibitem{STAR} STAR collaboration, hep-ex/0306056, nucl-ex/0306024.
651: 
652: \bibitem{NA49} NA49 collaboration, nucl-ex/0306022.
653: 
654: \bibitem{THERMAL}
655: 	J.Letessier, J.Rafelski, A.Tounsi, Phys.Lett.B {\bf 328}, 499, (1999);
656: 	P.Braun-Munziger, J.Stachel, J.P.Wessels, N.Xu,
657: 	Phys.Lett.B {\bf 344}, 43, (1995) and
658: 	Phys.Lett.B {\bf 365}, 1, (1996);
659: 	J.Cleymans, H.Oeschler, K.Redlich, Phys.Lett.B {\bf 485},
660: 	27, (2000);
661: 	K.Redlich, S.Hamieh, A.Tounsi, J.Phys.G {\bf 27}, 413, (2001);
662: 	M.Bleicher, F.M.Liu, A.Keranan, J.Aichelin, S.A.Bass, F.Becattini,
663: 	K.Redlich, K.Werner, Phys.Rev.Lett. {\bf 88}, 202501, (2001);
664: 	W.Broniowski, A.Baran, W.Florkowski, nucl-th/0212053, nucl-th/0212052;
665: 	G.Torrieri, J.Rafelski, New J.Phys. {\bf 3}, 12, (2001).
666: 
667: \bibitem{Becattini}
668: 	F.Becattini, G.Pettini, Nucl.Phys.A {\bf 715}, 557, (2003).
669: 
670: \bibitem{FLOW} E.Schnedermann, J.Sollfrank, U.Heinz, Phys.Rev.C {\bf 48},
671: 	2462, (1994);
672: 	T.Cs\"org\H{o}, B.L\"orstad, Phys.Rev.C {\bf 54}, 1390, (1996).
673: 
674: \bibitem{RHIC-FLOW} e.g. STAR Collaboration, ISMD 2001, hep-ph/0111437;
675: 	A.Ster, T.Cs\"org\H{o}, hep-ph/0112064.
676: 
677: %\bibitem{RECENT-THERMAL} e.g. P.Braun-M\"unziger, K.Redlich, J.Stachel,
678: %	in Quark Gluon Plasma 3, eds.: R.Hwa, X-N.Wang, World Scientific
679: %	Publ.Co., (2003), nucl-th/0304013.
680: 
681: \bibitem{LQCD} Z.Fodor, S.D.Katz, Phys.Lett.B {\bf 534}, 87, (2002);
682: 	Z.Fodor et.al., Nucl.Phys.Proc.Sup. {\bf 106}, 441, (2002);
683: 	F.Csikor, G.I.Egri, Z.Fodor, S.D.Katz, hep-lat/0301027;
684: 	C.R.Altar et.al., Phys.Rev.D {\bf 68}, 014807, (2003);
685: 	F.Karsch, A.Laermann, hep-lat/0305025;
686: 	F.Karsch, K.Redlich, A.Tawfik, hep-ph/0306208.
687: 
688: \bibitem{ALCOR} T.S.Bir\'o, P.L\'evai, J.Zim\'anyi, Phys.Lett.B 
689: 	{\bf 347}, 6, (1995);
690: 	J.Zim\'anyi, T.S.Bir\'o, T.Csorg\H{o}, P.L\'evai, Acta Phys.Hung. Heavy Ion
691: 	Physics {\bf 4}, 15, (1996);
692: 	A.Bialas, Phys.Lett.B {\bf 442}, 449, (1998).
693: 
694: \bibitem{RECOMB} R.J.Fries, B.M\"uller, C.Nonaka, S.A.Bass,
695: 	nucl-th/0301087, nucl-th/0305079, nucl-th/0306027.
696: 
697: \bibitem{PARTON} R.J.Fries, B.M\"uller, D.K.Srivastava, Phys.Rev.Lett.
698: 	{\bf 90}, 132301, (2003);
699: 	D.K.Srivastava, C.Gale, R.J.Fries, Phys.Rev.C {\bf 67}, 034903, (2003).
700: 
701: \bibitem{CLIM} A.R\'enyi, Acta Math.Acad.Sci.Hung. {\bf 10}, 193, (1959);
702: 	R.Botet, M.Ploszajczak: {\em Universal fluctuations: the phenomenology
703: 	of hadronic matter}, World Scientific Lecture Notes in Physics
704: 	{\bf 65}, (2002);
705: 	T.A.Trainor, hep-ph/0001148.
706: 
707: \bibitem{TSALIS} C.Tsallis, J.Stat.Phys. {\bf 52}, 479, (1988),
708: 	Physica A {\bf 221}, 277, (1995); A.Plastino, A.R.Plastino, 
709: 	Brazilian J. Phys. {\bf 29}, 50, (1999).
710: 
711: \bibitem{GPD} G.Papp, G.G.Barnaf\"oldi, P.L\'evai, G.F\'ai, 
712: 	{\em Intrinsic Parton Transverse Momentum in NLO pion production},
713: 	hep-ph/0212249;
714: 	M.A.Kimber, A.D.Martin, M.G.Ryskin, Eur.Phys.J {\bf C12}, 655, (2000);
715: 	G.Watt, D.A.Martin, M.G.Ryskin, hep-ph/0306169;
716: 	A.Gawron, J.Kwiecinski, W.Broniowski, hep-ph/0305219.
717: 
718: \end{thebibliography}
719: \end{document}
720: