1: \documentclass[12pt,epsf,epsfig,psfig]{article}

2: %\documentstyle[12pt]{article}

3: \usepackage{graphicx}

4: \usepackage{epsfig}

5: \oddsidemargin 15pt

6: \topmargin 0pt

7: \headheight 00pt

8: \headsep 00pt

9: \textheight 235mm

10: \textwidth 160mm

11: \voffset=-0.5cm

12: \hoffset=-0.5cm

13: \parindent=0pt

14: 

15: % put your own definitions here:

16: 

17: \def\J{$J/\psi$}

18: \def\j{J/\psi}

19: \def\X{$\chi_c$}

20: \def\x{\chi}

21: \def\P{$\psi'$}

22: \def\p{\psi'}

23: \def\U{$\Upsilon$}

24: \def\u{\Upsilon}

25: \def\C{c{\bar c}}

26: \def\B{b{\bar b}}

27: \def\cg{c{\bar c}\!-\!g}

28: \def\bg{b{\bar b}\!-\!g}

29: \def\b{b{\bar b}}

30: \def\q{q{\bar q}}

31: \def\Q{Q{\bar Q}}

32: \def\e{\epsilon}

33: \def\t{\tau}

34: \def\l{\Lambda_{\rm QCD}}

35: \def\A{$A_{\rm cl}$}

36: \def\a{\alpha}

37: \def\N{$n_{\rm cl}$}

38: \def\n{n_{\rm cl}}

39: \def\S{S_{\C}}

40: \def\s{s_{\rm cl}}

41: \def\bb{\bar \beta}

42: \def\chiral{\psi {\bar \psi}}

43: \def\CMP{{ Comm.\ Math.\ Phys.\ }}

44: \def\NP{{ Nucl.\ Phys.\ }}

45: \def\PL{{ Phys.\ Lett.\ }}

46: \def\PR{{ Phys.\ Rev.\ }}

47: \def\PRep{{ Phys.\ Rep.\ }}

48: \def\PRL{{ Phys.\ Rev.\ Lett.\ }}

49: \def\RMP{{ Rev.\ Mod.\ Phys.\ }}

50: \def\ZP{{ Z.\ Phys.\ }}

51: \def\EPJ{{Eur.\ Phys.\ J.\ }}

52: 

53: \def\be{\begin{equation}}

54: \def\ee{\end{equation}}

55: \def\lsim{\raise0.3ex\hbox{$<$\kern-0.75em\raise-1.1ex\hbox{$\sim$}}}

56: \def\gsim{\raise0.3ex\hbox{$>$\kern-0.75em\raise-1.1ex\hbox{$\sim$}}}

57: 

58: 

59: \begin{document}

60: 

61: 

62: BI-TP 2006/06\hfill 28.\ 2.\ 06

63: 

64: ~\vskip 1cm

65: 

66: \centerline{\Large \bf Charm and Beauty in a Hot Environment$^*$} 

67: 

68: \vskip1cm

69: 

70: \centerline{\large \bf Helmut Satz} 

71: 

72: \vskip 0.5cm

73: 

74: \medskip

75: 

76: \centerline{Universit\"at Bielefeld, Germany}

77: 

78: %\centerline{Fakult\"at f\"ur Physik, Universit\"at Bielefeld}

79: 

80: %\centerline{Postfach 100 131, D-33501 Bielefeld, Germany}

81:  

82: \centerline{and}

83: 

84: \centerline{Instituto Superior T\'ecnico, Lisboa, Portugal}

85: 

86: %\centerline{Centro de F{\'i}sica Teorica de Particulas (CFTP)}

87: 

88: %\centerline{Instituto Superior T\'ecnico, Av. Rovisco Pais, P-1049-001 

89: %Lisboa, Portugal}

90: 

91: \vskip1cm

92: 

93: \centerline{\bf Abstract:}

94: 

95: \vskip0.5cm

96: 

97: We discuss the spectral analysis of quarkonium states in a hot medium of 

98: deconfined quarks and gluons, and we show that such an analysis provides 

99: a way to determine the thermal properties of the quark-gluon plasma.

100: 

101: \vskip1cm

102: 

103: \noindent{\large \bf 1.\ Introduction}

104: 

105: \bigskip

106: 

107: Since some thirty years it is known that besides the almost massless 

108: {\it up} and $down$ quarks of the everyday world, and the still relatively 

109: light {\it strange} quarks required to account for the strange mesons and 

110: hyperons observed in hadron collisions, there are heavy quarks at the other 

111: end of the scale, whose bare masses alone are larger than those of most of 

112: the normal hadrons. These heavy quarks first showed up in the discovery of 

113: the \J ~meson \cite{Ting}, of mass of 3.1 GeV; it is a bound state of a 

114: {\it charm} quark ($c$) and its antiquark ($\bar c$), each having a mass 

115: of some 1.2--1.5 GeV. On the next level there is the \U~meson \cite{Leder}, 

116: with a mass of about 9.5 GeV, made up of a {\it bottom} or {\it beauty} 

117: quark-antiquark pair ($b \bar b$), with each quark here having a mass 

118: around 4.5 - 4.8 GeV. Both charm and bottom quarks can of course also 

119: bind with normal light quarks, giving rise to open charm ($D$) and open 

120: beauty ($B$) mesons. The lightest of these `light-heavy' mesons have 

121: masses of about 1.9 GeV and 5.3 GeV, respectively. 

122: 

123: \medskip

124: 

125: The bound states of a heavy quark $Q$ and its antiquark $\bar Q$ are 

126: generally referred to as quarkonia. Besides the initially discovered 

127: vector ground states \J~and \U, both the $c \bar c$ and the $b \bar b$ 

128: systems give rise to a number of other {\it stable} bound states of 

129: different quantum numbers. They are stable in the sense that their mass 

130: is less than that of two light-heavy mesons, so that strong decays are 

131: forbidden. The measured stable charmonium spectrum contains the $1S$ 

132: scalar $\eta_c$ and vector \J, three $1P$ states \X~(scalar, vector 

133: and tensor), and the $2S$ vector state \P, whose mass is just below 

134: the open charm threshold. There are further charmonium states above the 

135: \P; these can decay into $D \bar D$ pairs, and we shall here restrict 

136: our considerations only to quarkonia stable under strong interactions.

137: 

138: \medskip

139: 

140: \hrule width5cm

141: 

142: \medskip

143: 

144: $*$ Dedicated to Adriano Di Giacomo on the occasion of his 70th birthday

145: 

146: \newpage

147: 

148: The study of quarkonia has played and continues to play a major role

149: in many aspects of QCD: in spectroscopy, in production and decay, and

150: last but not least as probe of hot QCD media \cite{nora}. And in the

151: development of our understanding of and our feeling for the charm and

152: the beauty of strong interactions, Adriano Di Giacomo has played and

153: certainly will continue to play a major role. Neither my competence nor

154: the space and time available here allow me to give an overview of these

155: contributions; so let me just cite a few recent ones which are of

156: particular importance to what I want to discuss here \cite{DG,A-DG}. 

157: 

158: \medskip

159: 

160: Quarkonia are rather unusual hadrons. The masses of the light hadrons,

161: in particular those of the non-strange mesons and baryons, arise almost

162: entirely from the interaction energy of their nearly massless quark 

163: constituents. In contrast, the quarkonium masses are largely determined

164: by the bare charm and bottom quark masses. These large quark masses allow 

165: a very straightforward calculation of many basic quarkonium properties, 

166: using non-relativistic potential theory. It is found that, in particular,

167: the ground states and the lower excitation levels of quarkonia are very

168: much smaller than the normal hadrons, and that they are very tightly bound.

169: Now deconfinement is a matter of scales: when the separation between normal 

170: hadrons becomes much less than the size of these hadrons, they melt to form 

171: the quark-gluon plasma. What happens at this point to the much smaller 

172: quarkonia?  When do they become dissociated? That is the main question

173: we want to address here.

174: We shall show that the disappearance of specific quarkonia signals the 

175: presence of a deconfined medium of a specific temperature \cite{M-S}.

176: Thus the study of the quarkonium spectrum in a given medium 

177: is akin to the spectral analysis of stellar media, where the

178: presence or absence of specific excitation lines allows a determination

179: of the temperature of the stellar interior.

180: 

181: \medskip

182: 

183: We had defined quarkonia as bound states of heavy quarks which are 

184: stable under strong decay, i.e., $M_{c\bar c} \leq 2 M_D$ for charmonia 

185: and $M_{b\bar b} \leq 2 M_B$ for bottomonia. Since the quarks are heavy,

186: with $m_c \simeq 1.2 - 1.5 $ GeV for the charm and $m_b \simeq 4.5 - 4.8$ 

187: GeV for the bottom quark, quarkonium spectroscopy can be studied quite well 

188: in non-relativistic potential theory. The simplest (``Cornell'') confining

189: potential \cite{Cornell} for a $\Q$ at separation distance $r$ has the form

190: \be

191: V(r) = \sigma ~r - {\alpha \over r}

192: \label{cornell}

193: \ee

194: with a string tension $\sigma \simeq 0.2$ GeV$^2$ and a Coulomb-like term

195: with a gauge coupling $\alpha \simeq \pi/12$. The corresponding 

196: Schr\"odinger equation

197: \be

198: \left\{2m_c -{1\over m_c}\nabla^2 + V(r)\right\} \Phi_i(r) = M_i \Phi_i(r)

199: \label{schroedinger}

200: \ee

201: then determines the bound state masses $M_i$ and the wave functions 

202: $\Phi_i(r)$, and with

203: \be

204: \langle r_i^2 \rangle = \int d^3r~ r^2 |\Phi_i(r)|^2 / 

205: \int d^3r~|\Phi_i(r)|^2 .

206: \label{radii}

207: \ee

208: the latter in turn provide the (squared) average bound state ``radii'',

209: here defined as the $\Q$ separation for the state in question.

210: 

211: 

212: \medskip

213: 

214: The solution of eq.\ (\ref{schroedinger}) gives in fact a very good 

215: account of the full (spin-averaged) quarkonium spectroscopy, as seen in

216: Table 1 \cite{D-K-S}. The line labelled $\Delta M$ shows the differences 

217: between the experimental and the calculated values; they are in all cases 

218: less than 1 \%. The input parameters for these results were $m_c=1.25$ GeV,

219: $m_b=4.65$ GeV, $\sqrt \sigma = 0.445$ GeV, $\alpha=\pi/12$.

220: We see that in particular the \J~ and the lower-lying bottomonium

221: states are very tightly bound ($ 2M_{D,B} - M_0 \gg \l)$ and of very

222: small spatial size ($r_0 \ll 2 r_h \simeq 2$ fm). What happens to them

223: in a hot and dense medium?

224: 

225: \vskip0.5cm

226: 

227: \hskip1.5cm

228: \renewcommand{\arraystretch}{1.8}

229: \begin{tabular}{|c|c|c|c|c|c|c|c|c|}

230: \hline

231: {\rm state}& $J/\psi$ & $\chi_c$ & $\psi'$  & $\Upsilon$

232:  & $\chi_b$ & 

233: $\Upsilon'$ & $\chi_b'$ & $\Upsilon''$ \\

234: \hline

235: {\rm mass~[GeV]}&

236: 3.10&

237: 3.53&

238: 3.68&

239: 9.46&

240: 9.99&

241: 10.02&

242: 10.26&

243: 10.36 \\

244: \hline

245: $\Delta E$ {\rm[GeV]}&0.64&0.20&0.05 &1.10&

246: 0.67&0.54&0.31&0.20 \cr

247: \hline

248: $\Delta M$ {\rm[GeV]}&0.02&-0.03&0.03 & 0.06&

249: -0.06&-0.06&-0.08&-0.07 \cr

250: \hline

251: {$r_0$ \rm [fm]}&0.50&0.72&0.90&

252: 0.28&0.44& 0.56& 0.68 &0.78 \cr

253: \hline

254: \end{tabular}

255: 

256: \bigskip

257: 

258: \centerline{Table 1:

259: Quarkonium Spectroscopy from Non-Relativistic Potential Theory \cite{D-K-S}}

260: 

261: \vskip0.8cm

262: 

263: \noindent{\large \bf 2.\ Interaction Range and Colour Screening}

264: 

265: \bigskip

266: 

267: Consider a colour-singlet bound state of a heavy quark $Q$ and its 

268: antiquark $\bar Q$, put into the medium in such a way that we can measure 

269: the energy of the system as function of the $\Q$ separation $r$ (see Fig.\ 

270: \ref{string-b}). The quarks are assumed to be heavy so that they are static 

271: and any energy changes indicate changes in the binding energy. We consider 

272: first the case of vanishing baryon density; at $T=0$, the box is therefore 

273: empty. 

274: 

275: \medskip

276: 

277: \begin{figure}[htb]

278: \hspace*{0.1cm}

279: \centerline{\epsfig{file=box2.eps,width=8cm}}

280: \caption{String breaking for a $\Q$ system}

281: \label{string-b}

282: \end{figure}

283: 

284: \medskip

285: 

286: In vacuum, i.e., at $T=0$, the free energy of the $\Q$ pair is

287: assumed to have the string form \cite{Cornell} 

288: \be

289: F(r) \sim \sigma r

290: \label{string}

291: \ee

292: where $\sigma \simeq 0.16$ GeV$^2$ is the string tension as 

293: determined in the spectroscopy of heavy quark resonances (charmonium

294: and bottomonium states). Thus $F(r)$ increases with separation

295: distance; but when it reaches the value of a pair of dressed

296: light quarks (about the mass of a $\rho$ meson), it becomes energetically

297: favorable to produce a $\q$ pair from the vacuum, break the string

298: and form two light-heavy mesons ($Q\bar q$) and ($\bar Q q$). These

299: can now be separated arbitrarily far without changing the energy of the

300: system (Fig.\ \ref{string-b}). 

301: 

302: \medskip

303: 

304: 

305: The string breaking energy for charm quarks is found to be

306: \be

307: F_0 = 2(M_D - m_c) \simeq 1.2~{\rm GeV};

308: \label{charm-break}

309: \ee

310: for bottom quarks, one obtains the same value,

311: \be

312:  F_0 = 2(M_B - m_b) \simeq 1.2~{\rm GeV},

313: \label{bottom-break}

314: \ee

315: using in both cases the quark mass values obtained in the solution

316: leading to Table 3. Hence the onset of string breaking is evidently 

317: a property of the vacuum as a medium. It occurs when the two heavy 

318: quarks are separated by a distance

319: \be

320: r_0 \simeq 1.2~{\rm GeV}/\sigma \simeq 1.5~{\rm fm},

321: \label{stringbreak}

322: \ee

323: independent of the mass of the (heavy) quarks connected by the string.

324: 

325: \medskip

326: 

327: If we heat the system to get $T>0$, the medium begins to contain light 

328: mesons, and the large distance $\Q$ potential $F(\infty,T)$ decreases, since 

329: we can use these light hadrons to achieve an earlier string breaking through 

330: a kind of flip-flop recoupling of quark constituents \cite{miya}, resulting 

331: in an 

332: effective screening of the interquark force (see Fig.\ \ref{flip}).

333: Near the deconfinement point, the hadron density increases rapidly, and

334: hence the recoupling dissociation becomes much more effective, causing

335: a considerable decrease of $F(\infty,T)$.

336: 

337: \begin{figure}[h]

338: \hspace*{0.1cm}

339: \centerline{\epsfig{file=flipflop.eps,width=9cm}}

340: \caption{In-medium string breaking through recoupling}

341: \label{flip}

342: \end{figure}

343: 

344: \medskip

345: 

346: A further increase of $T$ will eventually bring the medium to the 

347: deconfinement point $T_c$, at which chiral symmetry restoration causes 

348: a rather abrupt drop of the light quark dressing (equivalently, of the 

349: constituent quark mass), increasing strongly the density of constituents. 

350: As a consequence, $F(\infty,T)$ now continues to drop sharply.   

351: Above $T_c$, light quarks and gluons become deconfined colour charges, and 

352: this quark-gluon plasma leads to a colour screening, which limits the range

353: of the strong interaction. The colour screening radius $r_D$, which determines 

354: this range, is inversely proportional to the density of charges, so that

355: it decreases with increasing temperature. As a result, the $\Q$ interaction

356: becomes more and more short-ranged. 

357: 

358: \medskip

359: 

360: In summary, starting from $T=0$, the $\Q$ probe first tests vacuum string 

361: breaking, then a screening-like

362: dissociation through recoupling of constituent quarks, and 

363: finally genuine colour screening. In Fig.\ \ref{OKfree}, we show the 

364: behaviour obtained in full two-flavour QCD for the colour-singlet

365: $\Q$ free energy as a function of $r$ for different $T$ \cite{K-Z}. 

366: 

367: \begin{figure}[htb]

368: \hspace*{-0.3cm}

369: \centerline{\epsfig{file=Ofree.eps,width=9cm}}

370: \caption{The colour singlet $\Q$ free energy $F(r,T)$ vs.\ $r$ at 

371: different $T$ \cite{K-Z}}

372: \label{OKfree}

373: \end{figure}

374: 

375: \medskip

376: 

377: It is evident in Fig.\ \ref{OKfree} that the asymptotic value $F(\infty,T)$, 

378: i.e., the energy needed to separate the $\Q$ pair, decreases with increasing 

379: temperature, as does the separation distance at which ``the string breaks''.

380: For the moment we consider the latter to be defined by the point beyond which 

381: the free energy remains constant within errors, returning in section 4.3 to a 

382: more precise definition. The behaviour of both quantities is shown in Fig.\ 

383: \ref{free-T}. Deconfinement is thus reflected very clearly in the temperature

384: behaviour of the heavy quark potential: both the string breaking energy 

385: and the interaction range drop sharply around $T_c$. The latter decreases 

386: from hadronic size in the confinement region to much smaller values in the 

387: deconfined medium, where colour screening is operative. 

388: 

389: \medskip

390: 

391: \begin{figure}[htb]

392: \mbox{

393: \hskip1cm

394: \epsfig{file=Free-inf.eps,width=5cm}

395: \hskip3cm

396: \epsfig{file=debye.eps,width=5cm,height=3.6cm}}

397: %\vskip0.4cm

398: \caption{String breaking potential and interaction range at different 

399: temperatures}

400: \label{free-T}

401: \end{figure}

402: 

403: The in-medium behaviour of heavy quark bound states thus does serve 

404: quite well as probe of the state of matter in QCD thermodynamics. We

405: had so far just considered $\Q$ bound states in general. Let us now

406: turn to a specific state such as the \J. What happens when the range 

407: of the binding force becomes smaller than the radius of the state?

408: Since the $c$ and the $\bar c$ can now no longer see each other, the

409: \J~ must dissociate for temperatures above this point. Hence the

410: dissociation points of the different quarkonium states provide a 

411: way to measure the temperature of the medium. The effect is illustrated

412: schematically in Fig. \ref{melt-t}, showing how with increasing 

413: temperature the different charmonium states ``melt'' sequentially

414: as function of their binding strength; the most loosely bound state 

415: disappears first, the ground state last.

416: 

417: \medskip

418: 

419: Moreover, since finite temperature lattice QCD also provides the

420: temperature dependence of the energy density \cite{schlad}, the 

421: melting of the different charmonia or bottomonia can be specified 

422: as well in terms of $\e$. In Fig\ \ref{melt-e}, we illustrate this, 

423: combining $\e(T)$ with the force radii shown in Fig.\ \ref{free-T}. 

424: It is evident that although \P~and \X~are both expected to melt 

425: around $T_c$, the corresponding dissociation energy densities could 

426: still differ.

427: 

428: 

429: \begin{figure}[htb]

430: \hskip1cm

431: {\epsfig{file=charm1.eps,width=6.5cm}}

432: \end{figure}

433: \hskip4cm

434: \begin{figure}[htb]

435: \vspace*{-6.4cm}

436: \hspace*{8.5cm}

437: {\epsfig{file=charm2.eps,width=7cm}}

438: %\end{figure}

439: \vspace*{0.5cm}

440: \caption{Charmonium spectra at different temperatures}

441: \label{melt-t}

442: \end{figure}

443: 

444: \medskip

445: 

446: \vspace*{-0.3cm}

447: \begin{figure}[htb]

448: \hspace*{-0.3cm}

449: \centerline{\epsfig{file=debye-L.eps,width=5.5cm}}

450: \caption{Charmonium dissociation vs.\ temperature and energy density}

451: \label{melt-e}

452: \end{figure}

453: 

454: \medskip

455: 

456: To make these considerations quantitative, we thus have to find a way

457: to determine the in-medium melting points of the different quarkonium

458: states. This problem has been addressed in three different approaches:

459: \vskip-0.2cm

460: \begin{itemize}

461: \vskip-0.2cm

462: \item{Model the heavy quark potential as function of the temperature,

463: $V(r,T)$, and solve the resulting Schr\"odinger equation

464: (\ref{schroedinger}).}

465: \vspace*{-0.2cm}

466: \item{Determine the internal energy $U(r,T)$ of a $\Q$ pair at separation 

467: distance $r$ from lattice results for the corresponding free energy $F(r,T)$, 

468: using the thermodynamic relation 

469: \vskip-0.5cm

470: \be

471: U(r,T) = -T^2 \left( {\partial [F(r,T)/T] \over \partial T}\right)

472: = F(r,T) - T \left({\partial F(r,T) \over \partial T}\right),

473: \label{intern}

474: \ee

475: and solve the Schr\"odinger equation with $V(r,T)=U(r,T)$ as the binding 

476: potential.}

477: \vskip-0.5cm

478: \item{Calculate the quarkonium spectrum directly in finite temperature

479: lattice QCD.}

480: \end{itemize}

481: Clearly the last is the only model-independent way, and it will

482: in the long run provide the decisive determination. However, the direct

483: lattice study of charmonium spectra has become possible only quite

484: recently, and so far most results are obtained in quenched QCD 

485: (no dynamical light quarks);

486: corresponding studies for bottomonia are still more difficult.

487: Hence much of what is known so far is based on Schr\"odinger equation

488: studies with different model inputs. To illustrate the model-dependence

489: of the dissociation parameters, we first cite some early work using different

490: models for the temperature dependence of $V(r,T)$, then some recent 

491: studies based on lattice results for $F(r,T)$, and finally summarize

492: the present state of direct lattice calculations of charmonia in finite

493: temperature media.

494: 

495: \vskip0.5cm

496: 

497: \noindent{\large \bf 3.\ Potential Model Studies}

498: 

499: \bigskip

500: 

501: The first quantitative studies of finite temperature charmonium dissociation

502: \cite{K-M-S} were based on screening in the form obtained in one-dimensional 

503: QED, the so-called Schwinger model. The confining part of the Cornell

504: potential (\ref{cornell}), $V(r) \sim \sigma r$, is the solution of the 

505: Laplace equation in one space dimension. In this case, Debye-screening leads 

506: to \cite{Dixit}

507: \be

508: V(r,T) \sim \sigma r \left\{ {1-e^{-\mu r} \over \mu r} \right\}, 

509: \label{schwinger-e}

510: \ee

511: where $\mu(T)$ denotes the screening mass (inverse Debye radius) for the

512: medium at temperature $T$. This form reproduces at least qualitatively 

513: the convergence to a finite large distance value $V(\infty,T) = 

514: \sigma/\mu(T)$, and since $\mu(T)$ increases with $T$, it also gives

515: the expected decrease of the potential with increasing temperature.

516: Combining this with the usual Debye screening for the $1/r$ part of

517: eq.\ (\ref{cornell}) then leads to  

518:  \be

519: V(r,T) \sim \sigma r \left\{ {1-e^{-\mu r} \over \mu r} \right\} 

520: - {\alpha \over r} e^{-\mu r} = {\sigma \over \mu} 

521: \left\{ 1-e^{-\mu r} \right\} - {\alpha \over r} e^{-\mu r}

522: \label{schwinger}

523: \ee

524: for the screened Cornell potential. In \cite{K-M-S}, the screening mass 

525: was assumed to have the form $\mu(T) \simeq 4~T$, as obtained in first

526: lattice estimates of screening in high temperature $SU(N)$ gauge theory.  

527: Solving the Schr\"odinger equation with these inputs, one found that

528: both the \P~and the \X~become dissociated essentially at $T \simeq T_c$,

529: while the \J~persisted up to about $1.2~T_c$. Note that as function of

530: the energy density $\e \sim T^4$, this meant that the \J~really survives

531: up to much higher $\e$.

532: 

533: \medskip

534: 

535: This approach has two basic shortcomings:

536: \vspace*{-0.2cm}

537: \begin{itemize}

538: \item{The Schwinger form (\ref{schwinger-e}) corresponds to the screening

539: of $\sigma r$ in one space dimension; the correct result in three space

540: dimensions is different \cite{Dixit}.}

541: \vspace*{-0.2cm}

542: \item{The screening mass $\mu(T)$ is assumed in its high energy form;

543: lattice studies show today that its behaviour near $T_c$ is quite 

544: different \cite{D-K-K-S}.}

545: \end{itemize}

546: 

547: While the overall behaviour of this approach provides some first insight

548: into the problem, quantitative aspects require a more careful treatment.

549: 

550: \medskip

551: 

552: When lattice results for the heavy quark free energy as function of the 

553: temperature first became available, an alternative description appeared

554: \cite{D-P-S1}. It assumed that in the thermodynamic relation (\ref{intern})

555: the entropy term $-T(\partial F /\partial T)$ could be neglected, thus

556: equating binding potential and free energy,

557: \be

558: V(r,T) = F(r,T) -T(\partial F /\partial T) \simeq F(r,T).

559: \label{DPS}

560: \ee

561: Using this potential in the Schr\"odinger equation (\ref{schroedinger})

562: specifies the temperature dependence of the different charmonium 

563: masses. On the other hand, the large distance limit of $V(r,T)$ 

564: determines the temperature variation of the open charm meson $D$,

565: \be

566: 2 M_D(T) \simeq 2 m_c + V(\infty,T)

567: \label{D-mass}

568: \ee

569: In fig.\ \ref{masses}, we compare the resulting open and hidden charm 

570: masses as function of temperature. It is seen that the \P~mass falls

571: below $2~M_D$ around 0.2 $T_c$, that of the \X~at about 0.8 $T_c$;

572: hence these states disappear by strong decay at the quoted temperatures.

573: Only the ground state \J~survives up to $T_c$ and perhaps

574: slightly above; the lattice data available at the time did not extend 

575: above $T_c$, so further predictions were not possible. 

576: 

577: \begin{figure}[htb]

578: \centerline{\epsfig{file=free.eps,width=6cm}}

579: \caption{Temperature dependence of open and hidden charm masses \cite{D-P-S1}}

580: \label{masses}

581: \end{figure}

582:  

583: The main shortcoming of this approach is quite evident. 

584: The neglect of the entropy term in the potential reduces $V(r,T)$ 

585: and hence the binding. As a result, the $D$ mass drops faster with

586: temperature than that of the charmonium states, and it is this effect

587: which leads to the early charmonium dissociation. Moreover, 

588: in the lattice studies used here, only the colour averaged free energy

589: was calculated, which leads to a further reduction of the binding

590: force.

591: 

592: \medskip

593: 

594: We conclude from these attempts that for a quantitative potential theory 

595: study, the free energy has to be formulated in the correct

596: three-dimensional screened Cornell form, and it then has to be

597: checked against the space- and temperature-dependence of the corresponding

598: colour singlet quantity obtained in lattice QCD. 

599: 

600: \vskip0.5cm

601: 

602: \noindent{\large \bf 4.\ Screening Theory}

603: 

604: \bigskip

605: 

606: The modification of the interaction between two charges immersed in a 

607: dilute medium of charged constituents is provided by Debye-H\"uckel theory, 

608: which for the Coulomb potential in three space dimensions leads to the 

609: well-known Debye screening, 

610: \be

611: {1\over r} ~\to~{1\over r} \e^{-\mu r},

612: \label{debye}

613: \ee

614: where $r_D=1/\mu$ defines the screening radius \cite{Landau}. 

615: Screening can be evaluated more generally \cite{Dixit}

616: for a given free energy $F(r) \sim r^q$ in $d$ space dimensions, with 

617: an arbitrary number $q$. We shall here apply this to the two terms

618: of the Cornell form, with $q=1$ for the string term, $q=-1$ for the gauge 

619: term, in $d=3$ space dimensions \cite{D-K-K-S}.

620: 

621: \medskip

622: 

623: We thus assume that the screening effect can be calculated separately for 

624: each term, so that the screened free energy becomes

625: \be

626: F(r,T) =  F_s(r,T) + F_c(r,T) =

627: \sigma r f_s(r,T) - {\alpha \over r} f_c(r,T).

628: \label{screen-cornell}

629: \ee

630: The screening functions $f_s(r,T)$ and $f_c(r,T)$ must satisfy

631: \begin{eqnarray}

632: f_s(r,T)&=&f_c(r,T)= 1 ~~{\rm for}~ T \rightarrow 0,\nonumber\\

633: f_s(r,T)&=&f_c(r,T)= 1 ~~{\rm for}~ r \!\rightarrow 0,

634: \label{screen-f}

635: \end{eqnarray}

636: 

637: since at $T=0$ there is no medium, while in the short-distance limit

638: $T^{-1} \gg r \to 0$, the medium has no effect. The resulting forms are

639: \cite{Dixit}

640: \be

641: F_c(r,T) = -{\alpha \over r}\left[e^{-\mu r} + \mu r\right]

642: \label{screen-gauge}

643: \ee

644: for the gauge term, and

645: \begin{eqnarray}

646: \hspace{-0.4cm}F_s(r,T) = {\sigma \over \mu}

647: \left[ {\Gamma\left(1/4\right) \over 2^{3/2}

648: \Gamma\left(3/4\right)}-{\sqrt{\mu r} \over 2^{3/4}\Gamma\left(3/4\right)}

649: K_{1/4}[{(\mu r)}^2]\right]

650: \label{screen-string}

651: \end{eqnarray}

652: 

653: for the string term. The first term in eq.\ (\ref{screen-string}) 

654: gives the constant large distance limit due to colour screening; the 

655: second provides a Gaussian cut-off in $x=\mu r$, since 

656: $K_{1/4}(x^2) \sim \exp\{-x^2))$; this is in contrast to the exponential 

657: cut-off given by the Schwinger form (\ref{schwinger}), but in accord

658: with string breaking arguments of Di Giacomo et al.\ \cite{A-DG}. 

659: 

660: \begin{figure}[htb]

661: \hspace*{-0.3cm}

662: {\epsfig{file=Oplcf_above.eps,width=8cm}}

663: {\epsfig{file=Oplcf_below.eps,width=8cm}}

664: \hspace*{4cm}(a)\hspace*{7.5cm}(b)

665: \caption{Screening fits to the $\Q$ free energy $F(r,T)$ for $T\geq T_c$

666: (left) and $T\leq T_c$ (right) \cite{D-K-K-S}}

667: \label{free}

668: \end{figure}

669: 

670: At temperatures $T > T_c$, when the medium really consists of unbound

671: colour charges, we thus expect the free energy $F(r,T)$ to have the form

672: (\ref{screen-cornell}), with the two screened terms given by eqs.\ 

673: (\ref{screen-gauge}) and (\ref{screen-string}). In Fig.\ \ref{free}a,

674: it is seen that the results for the colour singlet free energy calculated

675: in two-flavour QCD \cite{OK} are indeed described very well by this form, 

676: with $m_c=1.25$ GeV and $\sqrt \sigma = 0.445$ GeV, as above.

677: The only parameter to be determined, the screening mass

678: $\mu$, is shown in  Fig.\ \ref{sc-mass}, and as expected, it first increases

679: rapidly in the transition region and then turns into the perturbative

680: form $\mu \sim T$.

681: 

682: \medskip

683: 

684: The behaviour of $\Q$ binding in a plasma of unbound quarks and gluons is

685: thus well described by colour screening. Such a description is in fact 

686: found to work well also for $T<T_c$, when quark recombination leads to an

687: effective screening-like reduction of the interaction range, provided one

688: allows higher order contributions in $x=\mu r$ in the Bessel function

689: $K_{1/4}(x^2)$ governing string screening \cite{D-K-K-S}. The resulting fit 

690: to the two-flavour colour singlet free energy below $T_c$ is shown in 

691: Fig.\ \ref{free}b, using

692: $K_{1/4}(x^2+ (1/2) x^4)$ in eq.\ (\ref{screen-string}); the corresponding 

693: values for the screening mass are included in Fig.\ \ref{sc-mass}.   

694: 

695: \begin{figure}[htb]

696: \hspace*{-0.3cm}

697: \centerline{\epsfig{file=mu_ttc.eps,width=7cm}}

698: \caption{The screening mass $\mu(T)$ vs.\ $T$ \protect\cite{D-K-K-S,D-K-S}}

699: \label{sc-mass}

700: \end{figure}

701: 

702: \medskip

703: 

704: With the free energy $F(r,T)$ of a heavy quark-antiquark pair given in 

705: terms of the screening form obtained from Debye-H\"uckel theory, the 

706: internal energy $U(r,T)$ can now be obtained through the thermodynamic 

707: relation (\ref{intern}), and this then provides the binding potential

708: $V(r,T)$ for the temperature dependent version of the Schr\"odinger 

709: equation (\ref{schroedinger}). The resulting solution then specifies

710: the temperature-dependence of charmonium binding as based on the 

711: correct heavy quark potential \cite{D-K-S}.

712: Let us briefly comment on the thermodynamic

713: basis of this approach. The pressure $P$ of a thermodynamic system is

714: given by the free energy, $P=-F=-U+TS$; it is determined by the kinetic

715: energy $TS$ at temperature $T$ and entropy $S(T)$, reduced by the potential

716: energy $U(T)$ between the constituents. In our case, all quantities give

717: the difference between a thermodynamic system containing a $\Q$ pair 

718: and the corresponding system without such a pair. The potential energy

719: of the $\Q$ pair, due both to the attraction of $Q$ and $\bar Q$ and

720: to the modification which the pair causes to the internal energy of 

721: the other constituents of the medium, is therefore given by $U$.

722: To determine the dissociation points for the different quarkonium

723: states, we thus have to solve the Schr\"odinger equation (\ref{schroedinger})

724: with $V(r,T)=U(r,T)$.

725: 

726: \medskip

727: 

728: From eqs.\ (\ref{screen-gauge}) and (\ref{screen-string}) we 

729: get for the $\Q$ potential

730: \be

731: V(r,T)= V(\infty,T) + \tilde V(r,T),

732: \ee

733: with 

734: \be

735: V(\infty,T) = c_1{\sigma \over \mu} - \alpha \mu

736: + T{d\mu \over dT}[c_1 {\sigma \over \mu^2} + \alpha],

737: \ee

738: and $c_1=\Gamma(1/4)/2^{3/2} \Gamma(3/4)$, and where $\tilde V(r,T)$ 

739: contains the part of the potential which vanishes for $r\to \infty$.

740: The behaviour of $V(\infty,T)$ as function of the temperature is

741: shown in Fig.\ \ref{vinf}. It measures (twice) the energy of the cloud

742: of light quarks and gluons around an isolated heavy quark, of an extension 

743: determined by the screening radius, relative to the energy contained in

744: such a cloud of the same size without a heavy quark. This energy difference 

745: arises from the interaction of the heavy quark with light quarks and gluons

746: of the medium, and from the modification of the interaction between the light 

747: constituents themselves, caused by the presence of the heavy charge. --

748: The behaviour of $\tilde V(r,T)$ is shown in Fig.\ \ref{vr} for three

749: different values of the temperature. It is seen that with increasing $T$,

750: screening reduces the range of the potential. 

751: 

752: \begin{figure}[h]

753: %\vskip0.5cm

754: \begin{minipage}[t]{7cm}

755: %\hskip0.1cm 

756: \epsfig{file=vinf.eps,width=7.5cm}

757: %\vskip-0.1cm

758: \caption{The large distance limit of the quarkonium potential $V(r,T)$

759: \protect\cite{D-K-S}}

760: \label{vinf}

761: \end{minipage}

762: \hspace{1.3cm}

763: \begin{minipage}[t]{7cm}

764: \vspace*{-5.3cm}

765: \hskip-0.3cm

766: \epsfig{file=ovr.eps,width=7.5cm}

767: \caption{Variation of $\tilde V(r,T)$ with $r$ for different $T$

768: \protect\cite{D-K-S}}

769: \label{vr}

770: \end{minipage}

771: \end{figure}

772: 

773: \medskip

774: 

775: The relevant Schr\"odinger equation now becomes

776: \be

777: \left\{{1\over m_c}\nabla^2 - \tilde V(r,T)\right\} \Phi_i(r) = 

778: \Delta E_i(T) \Phi_i(r)

779: \label{T-schroedinger}

780: \ee 

781: where

782: \be

783: \Delta E_i(T)  = V(\infty,T) - M_i - 2m_c

784: \ee

785: is the binding energy of charmonium state $i$ at temperature $T$.

786: When it vanishes, the bound state $i$ no longer exists, so that 

787: $\Delta E_i(T)=0 $ determines the dissociation temperature $T_i$ 

788: for that state. The temperature enters only through the $T$-dependence

789: of the screening mass $\mu(T)$, as obtained from the analysis of the 

790: lattice results for $F(r,T)$. -- Equivalently, the divergence of the

791: binding radii, $r_i(T) \to \infty$, can be used to define the 

792: dissociation points $T_i$.

793: 

794: \medskip

795: 

796: The same formalism, with $m_b=4.65$ GeV replacing $m_c$, leads to the

797: bottomonium dissociation points. The combined quarkonium results are 

798: listed in Table 2. They agree quite well with those obtained in a

799: very similar study based on corresponding free energies obtained

800: in quenched lattice QCD \cite{Wong}, indicating that above $T_c$ gluonic 

801: effects dominate. Using a parametrically generalized screened Coulomb

802: potential obtained from lattice QCD results also leads to very compatible 

803: results for the $N_f=2$ and quenched cases\cite{Alberico}. We recall

804: here that the main underlying change, which is responsible for the much

805: higher dissociation temperatures for the quarkonium ground states, is the 

806: use of the full internal energy (\ref{intern}), including the entropy term,

807: as potential in the Schr\"odinger equation: this makes the binding much

808: stronger. 

809: 

810: \medskip

811: 

812: We should note, however, that in all such potential studies

813: it is not so clear what binding energies of less than a few MeV 

814: or bound state radii of several fermi can mean in a medium whose

815: temperature is above 200 MeV and which leads to screening radii 

816: of less than 0.5 fm. In such a situation, thermal activation \cite{K-ML-S}

817: can easily dissciate the bound state.

818: 

819: \medskip

820: 

821: \begin{center}

822: \renewcommand{\arraystretch}{2.0}

823: \begin{tabular}{|c||c|c|c||c|c|c|c|c|}

824: \hline

825:  state & J/$\psi(1S)$ & $\chi_c$(1P) & $\psi^\prime(2S)$&$\Upsilon(1S)$&

826: $\chi_b(1P)$&$\Upsilon(2S)$&$\chi_b(2P)$&$\Upsilon(3S)$\\

827: \hline

828: \hline

829: $T_d/T_c$ & 2.10  & 1.16 & 1.12 & $>4.0$ & 1.76 & 1.60& 1.19 & 1.17 \\

830: \hline

831: \end{tabular}\end{center}

832: 

833: \medskip

834: 

835: \centerline{Table 2: Quarkonium Dissociation Temperatures \cite{D-K-S}}

836: 

837: \vskip0.5cm

838: 

839: \noindent{\large \bf 5.\ Charmonium Correlators}

840: 

841: \bigskip

842: 

843: The direct spectral analysis of charmonia in finite temperature lattice 

844: has come within reach only in very recent years \cite{lattice-charm}.

845: It is possible now to evaluate the correlation functions $G_H(\t,T)$ for

846: hadronic quantum number channels $H$ in terms of the Euclidean time

847: $\t$ and the temperature $T$. These correlation functions are directly

848: related to the corresponding spectral function $\sigma_H(M,T)$,

849: which describe the distribution in mass $M$ at temperature $T$ for

850: the channel in question. In Fig.\ \ref{spec}, schematic results at different

851: temperatures are shown for the \J~and the \X~channels. It is seen that

852: the spectrum for the ground state \J~ remains essentially unchanged

853: even at $1.5~T_c$. At $3~T_c$, however, it has disappeared; the

854: remaining spectrum is that of the $\C$ continuum of \J~quantum numbers

855: at that temperature. In contrast, the \X~is already absent at $1.1~T_c$, 

856: with only the corresponding continuum present. 

857: 

858: \medskip

859: 

860: \begin{figure}[htb]

861: \centerline{\epsfig{file=c-spectral.eps,width=9cm}}

862: \caption{\J~and \X~spectral functions at different temperatures}

863: \label{spec}

864: \end{figure}

865: 

866: These results are clearly very promising: they show that in a foreseeable

867: future, the dissociation parameters of quarkonia can be determined 

868: {\sl ab initio} in lattice QCD. For the moment, however, they remain

869: indicative only, since the underlying calculations were generally

870: performed in quenched QCD, i.e., without dynamical quark loops. Since 

871: such loops are crucial in the break-up of quarkonia into light-heavy mesons, 

872: final results require calculations in full QCD. Some first calculations

873: in two-flavour QCD have just appeared \cite{skullerud} and support

874: the late dissociation of the \J.  The widths of the

875: observed spectral signals are at present determined by the precision of

876: the lattice calculations; to study the actual physical widths, much

877: higher precision is needed. Finally, one has so far only first signals

878: at a few selected points; a temperature scan also requires higher performance

879: computational facilities. Since the next generation of computers, in the

880: multi-Teraflops range, is presently going into operation, the next

881: years should bring the desired results. So far, in view of the mentioned

882: uncertainties in both approaches, the results from direct lattice studies

883: and those from the potential model calculations of the previous section

884: appear quite compatible. 

885: 

886: \vskip0.5cm

887: 

888: \noindent{\large \bf 6.\ Conclusions}

889: 

890: \bigskip

891: 

892: The theoretical analysis of the in-medium behaviour of quarkonia has 

893: greatly advanced in the past decade. Potential model studies based on

894: lattice results for the colour-singlet free energy appear to be converging

895: to results from direct lattice calculations, and within a few more

896: years, the dissociation temperatures for the different quarkonium

897: states should be known precisely. Through corresponding calculations

898: of the QCD equation of state, these temperatures provide the energy

899: density values at which the dissociation occurs. In statistical QCD,

900: quarkonia thus allow a spectral analysis of the quark-gluon plasma.

901: as many of the observed features as possible. 

902: 

903: \medskip

904: 

905: Finally, it seems worthwhile to note that experimental measurements of 

906: the relative dissociation points of the different quarkonium states could

907: be possible in the study of high energy nucleus-nucleus collisions. If

908: they succeed, this might in fact be a unique chance to 

909: compare quantitative {\sl ab initio} QCD predictions directly to data. 

910: 

911: \vskip1cm

912: 

913: \centerline{\large \bf Acknowledgements}

914: 

915: \bigskip

916: 

917: It is a pleasure to thank S.\ Digal, O.\ Kaczmarek,

918: F.\ Karsch, D.\ Kharzeev, P.\ Petreczky, 

919: M.\ Nardi and R.\ Vogt for stimulating and helpful comments.

920: 

921: \begin{thebibliography}{99}

922: 

923: \bibitem{Ting} J.\ J.\ Aubert et al., \PRL 33 (1974) 1404;\\

924: J.\ E.\ Augustin et al., \PRL 33 (1974) 1406.

925: 

926: \bibitem{Leder} S.\ W.\ Herb et al., \PRL 39 (1977) 252.

927: 

928: \bibitem{nora} For a recent survey, see\\

929: N.\ Brambilla et al., {\sl Heavy Quarkonium Physics},

930: Yellow Report CERN-2005-005.

931: 

932: \bibitem{DG} For a recent survey, see\\

933: A.\ Di Giacomo et al., Phys.\ Rep.\ 372 (2002) 319.

934: 

935: \bibitem{A-DG} D.\ Antonov, L.\ Del Debbio and A.\ Di Giacomo, JHEP

936: 0308 (2003) 011;\\

937:  D.\ Antonov and A.\ Di Giacomo, JHEP 0503 (2005) 017.

938: 

939: \bibitem{M-S} T.\ Matsui and H.\ Satz, \PL B 178 (1986) 416.

940: 

941: \bibitem{Cornell} E.\ Eichten et al., \PR D17 (1978) 3090; \PR D 21 (1980)

942: 203.

943: 

944: \bibitem{D-K-S} S.\ Digal, F.\ Karsch and H.\ Satz, in preparation;\\

945: for earlier work at $T=0$ , see\\

946: S.\ Jacobs, M.\ G.\ Olsson and C.\ Suchyta,

947: \PR 33 (1986) 3338.

948: 

949: \bibitem{miya} H.\ Miyazawa, \PR D 20 (1979) 2953;\\

950: V.\ Goloviznin and H.\ Satz, Yad.\ Fiz.\ 60N3 (1997) 523 

951: 

952: \bibitem{K-Z} O.\ Kaczmarek et al., Prog.\ Theor.\ Phys.\ Suppl.\ 153

953: (2004) 287;\\

954: O.\ Kaczmarek and F.\ Zantow, hep-lat/0503017.

955: 

956: \bibitem{schlad} F.\ Karsch, Lect.\ Notes Phys.\ 583 (2002) 209.

957: 

958: \bibitem{K-M-S} F.\ Karsch, M.-T.\ Mehr and H.\ Satz, \ZP C 37 (1988) 617

959: 

960: \bibitem{Dixit} V.\ V.\ Dixit, Mod.\ Phys.\ Lett.\ A5 (1990) 227.

961: 

962: \bibitem{D-K-K-S} S.\ Digal et al., \EPJ C 43 (2005) 71.

963:  

964: \bibitem{D-P-S1} S.\ Digal, P.\ Petreczky and H.\ Satz, \PL B 514 (2001) 57.

965: 

966: \bibitem{Landau} L.\ D.\ Landau and E.\ M.\ Lifshitz, 

967: {\sl Statistical Physics}, Pergamon Press, London, 1958.

968: 

969: \bibitem{OK} O.\ Kaczmarek and F.\ Zantow, \EPJ C 43 (2005) 63.

970: 

971: \bibitem{Wong} C.-Y.\ Wong, hep-ph/0408020

972: 

973: \bibitem{Alberico} W.\ Alberico et al., hep-ph/0507084;\\

974: C.-Y.\ Wong, hep-ph/0509088.

975: 

976: \bibitem{K-ML-S} D.\ Kharzeev, L.\ D.\ McLerran and H.\ Satz, \PL 

977: B 356 (1995) 349.

978: 

979: \bibitem{lattice-charm} T.\ Umeda et al., Int.\ J.\ Mod.\ Phys.\ A16 (2001) 

980: 2215;\\

981: M.\ Asakawa and T.\ Hatsuda, \PRL 92 (2004);\\

982: S.\ Datta et al., \PR D 69 (2004) 094507;\\

983: H.\ Iida et al., hep-lat/0509129

984: 

985: \bibitem{skullerud} R.\ Morrin et al., hep-lat/0509115.

986: 

987: \end{thebibliography}

988: 

989: \end{document}

990: 

991: