hep-lat0210040/ft.tex
1: %%only change the website address - 18/4/2000
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3: %%
4: %% ws-p8-50x6-00.tex : 20-11-97
5: %% This Latex2e file rewritten from various sources for use in the
6: %% preparation of the (smaller [8.50''x6.00'']) single-column proceedings 
7: %% Volume, latest version by R. Sankaran with acknowledgements to Susan 
8: %% Hezlet and Lukas Nellen. Please comments to:rsanka@wspc.com.sg
9: %%
10: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
11: %
12: \documentclass{ws-p8-50x6-00}
13: \renewcommand{\rightnote}{}
14: \begin{document}
15: 
16: \title{An introduction to 
17: lattice QCD \\ at non--zero temperature 
18: and density}
19: 
20: \author{Maria-Paola Lombardo}
21: 
22: \address{Istituto Nazionale di Fisica Nucleare -- Sezione di Padova}  
23: 
24: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
25: % You may repeat \author \address as often as necessary      %
26: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
27: 
28: \maketitle
29: 
30: \abstracts{This is an informal overview of methods and results
31: on the QCD phase diagram
32: and lattice termodynamics
33: aimed at specialists in nearby fields.}
34: 
35: \section{Introduction}
36: 
37: QCD at non-zero temperature and density is a rich
38: subject discussed in many talks at this meeting:
39: from the introductory lesson describing the theoretical
40:  discovery  of
41: the quark gluon plasma \cite{cabibbo}  to the subtle interrelations
42: of gauge fields dynamics and charmonium suppresions \cite{satz},
43: from the detailed analysis of the exotic phases at large baryon density
44: \cite{nardulli} to the fascinating aspects of the QCD phase diagram
45: with many flavor \cite{sannino}, from the time-honored theoretical 
46: subjects of the interplay  between topology and confinement
47: \cite{deldebbio} to the challenging aspects of the interrelations
48: of chiral and axial symmetries  \cite{meggiolaro}.
49: These  notes  mostly aimed at the non-specialist
50: might hopefully provide, besides some introductory material,
51: a path  among general idea as well as among the many
52: new results presented in the lively
53: topical session on the QCD phases.
54: 
55: 
56: The material is organised as follows: 
57: Section 2 reviews the imaginary time formulation of 
58: field theory thermodynamics, and gives
59: the functional integral representation
60: of the partition function $\cal Z$. 
61: The representation of $\cal Z$ 
62: naturally leads to a theoretical suggestion:
63: the concept of universality which is singled out in a small
64: Section on its own, because it is so important.
65: Section 4 discusses the QCD chiral transition, 
66: Section 5 the QCD deconfinement transition, and Section 6
67: how to put together the two, eventually describing the
68: complex phenomenology of the real world
69: high temperature QCD phase transition. The equation
70: of state is briefly mentioned here. Section 6 introduces the method
71: used to obtain all of the results of the previous
72: Sections: the lattice regularization. 
73: We shall see that introducing temperature
74: is  straightforward while a non--zero density
75: poses specific problems. We will only mention
76: the main proposals to circumvent such problems, referring to
77: the most recent reviews 
78: \cite{kogut} for details and recent results.
79: 
80: \section{Formulation}
81: Here we will concern ourselves with the path integral 
82: representation of the partition function.
83: The basic property of equilibrium field theory is that one
84: single function $\cal Z$ ( the grand canonical partition function):
85: \begin{equation}
86: \cal Z = \cal Z (V, T, \mu)
87: \end{equation}
88: determines completely
89: the thermodynamic state of a system  according to:
90: \begin{eqnarray}
91: P &=& T \frac {\partial ln {\cal Z}}{\partial V} \\
92: N &=& T \frac {\partial ln {\cal Z}}{\partial \mu} \\
93: S &=& \frac {\partial T ln {\cal Z}}{\partial T} \\
94: E &=& -PV + TS +\mu N 
95: \end{eqnarray}
96: while physical observables $<O>$ can be computed as
97: \begin{equation}
98: <O> = Tr O \hat \rho / {\cal Z}
99: \end{equation}
100: 
101: Any of the excellent books on statistical field theory and thermodynamics
102: can provide a more detailed discussion of these points.
103: I would like to underscore, very shortly,that the  problem is to 
104: learn how to  represent  $\cal Z$ at non--zero temperature 
105: and  baryon density, and how to design a calculational
106: scheme.  
107: 
108: 
109: \subsection{ $\cal Z$ at non--zero temperature and chemical potential }
110: 
111: $\cal Z$ is the trace of the density matrix of the system $\hat \rho$
112: \begin{eqnarray}
113: {\cal Z} &=& Tr \hat \rho \\ 
114: \hat \rho &=& e^{(-H - \mu \hat N)}/T
115: \end{eqnarray}
116: $H$ is the Hamiltonian, $T$ is the temperature and $\hat N$ is any
117: conserved number operator. 
118: 
119: \subsection{Chemical potential, relativistic and non-relativistic}
120: 
121: 
122: 
123: It is worth stressing the main differences 
124: between the non relativistic and relativistic meaning of
125: a chemical potential $\mu$, eqs (5) and (8). 
126: 
127: In a non relativistic setting, 
128: the chemical
129: potential tells us the energy `cost' of adding an extra particle
130: to the system: we thus have a different chemical potential
131: for each type of particle. The term which is added 
132: to the Hamiltonian is $\mu_k N_k$, $N_k$ being indeed the
133: number of particles belonging to the $k$--th species.
134: 
135: 
136: In a relativistic setting, particles
137: can be created and destroyed, thus losing their
138: individuality, so to speak: it only makes sense to have a chemical
139: potential coupled to the 0-th component of a conserved current.
140: 
141: In practice, we will mostly concern ourselves with the current 
142: $J_\mu = \bar \psi \gamma_\mu \psi$, 
143: i.e. $J_0 = \bar \psi \gamma_0 \psi = 
144: \psi ^\dagger \psi$  will be the density of
145:  fermion (baryon) number, or,  more precisely, 
146: the difference between the number of fermions and antifermions:
147: $\int J_0 = N - \bar N$ .
148: The chemical potential will then appear in the Lagrangian 
149: as a term 
150: $\mu J_0$ and we recognize immediately
151: that a change of sign for the chemical potential corresponds
152: to an exchange particles--antiparticles. 
153: 
154: A second difference between non-relativistic and relativistic approach
155: concerns the magnitude itself of $\mu$, which in a relativistic
156: theory contains the particle rest mass. 
157: 
158: From a technical point of view, a finite density is easily
159: handled in a non--relativistic setting. From a relativistic field
160: theoretic perspectives there are instead important technical problems.
161: We will discuss these problems later on, and we just anticipate here
162: that the explicit breaking of the particle--antiparticle symmetry
163: induced by a chemical potential induces the loss
164: of positivity of the effective gluonic 
165: Action--i.e. the breakdown of the condition which makes 
166: lattice calculations possible.
167: 
168: 
169: 
170: \subsection{Temperature}
171: 
172: 
173: 
174: Consider the transition amplitude for returning to the original state
175: $\phi_a$ after a time $t$:
176: \begin{equation}
177: <\phi_a| e^{-iHt}| \phi_a> = \int d\pi
178: \int_{\phi(x,0) = \phi_a(x)} ^ {\phi(x,t) = \phi_a(x)} d \phi 
179: e^{i\int_0^tdt \int d^3x (\pi(\vec x, t)
180: \frac {\partial \phi (\vec x, t)} {\partial t} - H (\pi, \phi))}
181: \end{equation}
182: Compare now the above with  expression (2)
183: for $\cal Z$, and make the trace explicit:
184: \begin{equation}
185: {\cal Z} = Tr e^{-\beta ( H - \mu \hat N)} =
186: \int d \phi_a <\phi_a | e^{-\beta (H - \mu N)} | \phi_a >
187: \end{equation}
188: We are naturally lead to the identification
189: \begin{equation}
190: \beta \equiv \frac{1}{T} \rightarrow it
191: \end{equation}
192: 
193: We note - anticipating the discussions of Section below -- 
194: that studying nonzero temperature on a lattice
195: \cite{lattice_rev} is straightforward:
196: one just takes advantage of the finite temporal extent of
197: the lattice, while keeping
198: the space directions much larger than any physical scale 
199: in the system. 
200: 
201: By introducing the integral 
202: $S(\phi, \psi)$ of the Lagrangian density (from now on we will
203: always use 1/T as the upper extreme for the time integration)
204: \begin{equation}
205: S(\phi, \psi) = \int_0^{1/T} dt \int d^3 x {\cal L}(\phi, \psi)
206: \end{equation}
207: $\cal Z$ is  written as
208: \begin{equation} 
209: {\cal Z} = \int d \phi d \psi e^{-S(\phi, \psi)}
210: \end{equation}
211: The only missing ingredient are the boundary conditions for
212: the fields: basically, they follow from the (anti)commuting
213: properties of the (fermi)bose fields
214: which imply
215: \begin{equation}
216: \hat \phi(\vec x, 0) = \hat \phi (\vec x, \beta)
217: \end{equation}
218: for the bosons and
219: \begin{equation}
220: \hat \psi(\vec x, 0) = -\hat \psi (\vec x, \beta)
221: \end{equation}
222: for the fermions.
223: 
224: Fermions and bosons obey antiperiodic and periodic 
225: boundary conditions, respectively,  in the
226: time direction.
227: 
228: 
229: The expression above, together with the boundary
230: conditions just introduced, is the key to field 
231: theory thermodynamics.
232: 
233: 
234: \section{Universality}
235: 
236: It is intuitive that when the smallest
237: significant length scale
238: of the system $l >> 1/T$ the system becomes effectively 
239: d--dimensional. Moreover,  the description of the system can be 
240: effectively `coarse grained', ignoring anything which happens on
241: a scale smaller than  $l$ .
242: 
243: This can become true when
244: the system is approaching a continuous transition:
245: the correlation length of the system $\xi$ is diverging.
246: In such situation all the physics is dominated by long wavelength
247: modes. Not only the system gets effectively reduced, but the
248: coarse graining procedure become doable. As an effect of this
249: procedure, systems which are very different one from another
250: might well be described by the same model, provided that the
251: long range physics is regulated by the same global symmetries: 
252: this is the idea of universality which provides the 
253: theoretical framework for
254: the study of the QCD transition in two interesting (albeit
255: non physical) cases which we review in the next two sections.
256: 
257: 
258: \section{QCD chiral transition ($m_q = 0$)}
259: 
260: Let us recall the symmetries of the QCD action with
261: $N_f$ flavors of massless quarks, coupled to a $SU(N_c)$ 
262: color group:
263: \begin{equation}
264: SU(N_c)_C \times SU(N_f) \times SU(N_f) \times Z_A(N_f)
265: \end{equation}
266: $SU(N_c)$ is the gauge color symmetry.
267: $SU(N_f) \times SU(N_f) \times Z_A(N_f)$ is
268: the flavor chiral symmetry, after the breaking of the classical
269: $U_A(1)$ symmetry to the discrete $Z_A(N_f)$.
270: 
271: We want to study the realization and pattern(s) of breaking
272: of the chiral symmetries 
273: and we would like to know the interrelation of the above
274: with the possibility of quark liberation predicted 
275: at high temperature and density.
276: 
277: \subsection{Ordinary conditions: zero temperature and density}
278: 
279: In normal conditions (zero temperature and density) the
280: $SU(N_f)_L\times SU(N_f)_R$ chiral symmetry is spontaneously broken
281: to the diagonal $SU(N_f)_{L+R}$.Let us 
282: note the isomorphism:
283: \begin{equation}
284: SU(2) \times SU(2) \equiv O(4)
285: \end{equation}
286: which shows that the symmetry is the same as the one of an O(4) ferromagnet.
287: The relevant degrees of freedom are the three pions, and the sigma
288: particle, and the effective potential is a function of 
289: $\sigma^2 +  |\pi|^2$ in the chiral space. Once a direction in the
290: chiral sphere is selected (say in the $\sigma$ direction) chiral
291: symmetry is spontaneously broken in that direction, according
292: to the equivalent patterns:
293: \begin{eqnarray}
294:  SU(2)_R \times SU(2)_L & \rightarrow & SU(2)_{L + R} \\
295:  O(4) & \rightarrow & O(3) 
296: \end{eqnarray}
297: Massless Goldstone particles (in this case, the three pions) 
298: appear in the direction orthogonal to the one selected by the
299: spontaneous breaking. 
300: 
301: 
302: \subsection{Increasing T}
303: 
304: Disorder increases with temperature. Then, one picture of the high $T$
305: QCD transition can be drawn by using a ferromagnetic analogy of
306: the chiral transition: 
307: $\bar \psi \psi$ can be thought of as a spin
308: field taking values in real space, 
309: but whose orientation is in the chiral sphere. 
310: Chiral symmetry breaking occurs
311: when $< \bar \psi \psi> \ne 0 $, i.e. it corresponds to the ordered phase.
312: By increasing T, $< \bar \psi \psi> \to 0$, and the O(4) symmetry
313: should be restored.
314: 
315: Combining this symmetry analysis with the general idea of dimensional
316: reduction, Pisarski and Wilczek 
317: \cite{piwi} proposed that the high temperature transition
318: in two flavor QCD should be in the universality class of the O(4)
319: sigma model in three dimensions.
320: At high temperature when symmetry is restored there will be just one global
321: minimum for zero value of the fields, and  pion and sigma become eventually 
322: degenerate.
323: 
324: 
325: We have however to keep in mind possible sources of violation of
326: this appealing scenario and, all in all, 
327:  one has to resort to numerical simulations to measure
328: the critical exponents, and verify or disprove the $O(4)$ universality.
329: In turn, this gives information on the issues raised for instance in
330: as well as on the possible restoration of the axial
331: anomaly, see the discussion in Meggiolaro's review.
332: 
333: In practice, one measures the chiral condensate as a function of
334: the coupling parameter $\beta$, which in turns determines the temperature
335: of the system. This gives the exponent $\beta_{mag}$ according to
336: \begin{equation}
337: <\bar \psi \psi> = B (\beta - \beta_c)^{\beta_{mag}}
338: \end{equation}
339: The exponent $\delta$ is extracted from the response at criticality:
340: \begin{equation}
341: <\bar \psi \psi> = A m^{1/\delta} ; \beta = \beta_c
342: \end{equation}
343: The results  for the critical exponents
344: compare favorably with the O(4) results
345: $\beta_{mag} = .38(1)$, $\delta = 4.8(2)$,
346: and definitively rule out mean fields exponents 
347: (which would have characterized
348: a weak first order transition). 
349: However, the results can still be compatible
350: with O(2) exponents, which would signal the persistence of some lattice
351: artifact, and of course it
352: is still possible that the final answer do not fit any of the above
353: predictions, for instance if we were just observing some crossover
354: phenomenon.
355: 
356: 
357: In conclusion the symmetry analysis of the (two flavor) QCD transition
358: gives a definite predictions for the value of the critical exponents,
359: which is possibly (slightly) violated by the numerical results.
360: The numerical relevance of this violation (i.e. possible systematic
361: errors) as well as the its physical implications are an interesting
362: open problem.
363: 
364: The contribution by E. Meggiolaro covers in more detail these points.
365: 
366: 
367: \subsection 
368: {Increasing density}
369: Till two or three years ago, we thought that asymptotic freedom 
370: was the main physical agent behind the pattern of chiral symmetries
371: at high density. Now it has been recognized that, at least at
372: zero temperature,  the
373: main  mechanisms are instability at the Fermi surface leading
374: to color superconductivity \cite{vari}.
375: The review by Nardulli and Sannino cover these points which,
376: unfortunately, are not yet amenable to a lattice study.
377: At non--zero density the lattice approach
378: (see contribution by M. D'Elia for developments presented
379: at this meeting)
380: is still limited to rather  high temperature, where 
381: color superconductivity is most likely lost anyway.
382: 
383: 
384: \section{QCD deconfinement transition ($m_q = \infty$)}
385: 
386: When $m_q = \infty$  quarks are static and  do not
387: contribute to the dynamics: hence, the dynamic of the system is 
388: driven by gluons alone, i.e. we are dealing with a purely Yang-Mills
389: model:
390: \begin{equation}
391: S = F_{\mu \nu}F^{\mu \nu}
392: \end{equation}
393:  In addition to the local gauge symmetry, the action enjoys the global 
394: symmetry associated with the center of the group, $Z(N_c)$. 
395: The order parameter is the Polyakov loop $P$
396: \begin{equation}
397: P = e ^ { i\int_0^{1/T} A_0 dt} 
398: \end{equation}
399: In practice, $P$ is the cost of a static source violating the $Z(N_c)$
400: global symmetry.
401: 
402: The interquark potential V(R,T) (R is the distance, T is the temperature)
403: is
404: \begin{equation}
405: e^{-V(R,T)/T} \propto < P(\vec 0) P^\dagger (\vec R) >
406: \end{equation}
407: Confinement can then be read off the behavior of the interquark potential
408: at large distance.
409: When  $V(R) \propto \sigma R$  it would cost an infinite
410: amount of energy to pull two quarks infinitely apart. Above a certain
411: critical temperature $V(R)$ becomes constant at large distance: i.e.
412: the string tension is zero, confinement is lost. 
413: The implication of this is that $|P|^2 = V(\infty, T)$ 
414: is zero in the confining phase, different from zero otherwise.
415: P plays then a double role, being the order parameter of the
416: center symmetry, and an indicator of confinement. 
417: We learn that in Yang Mills models there is
418: a natural connection between confinement and realization of the
419: $Z(N_c)$ symmetry. Hence,  the confinement / deconfinement
420: transition in Yang Mills systems is amenable to a symmetry
421: description.
422: By applying now the same dimensional reduction argument as above,
423: we conclude that the Universality class expected of
424: the three color model is the same as the one of a three
425: dimensional model with $Z(3)$ global symmetry:
426: Indeed, work by the Columbia, Tsukuba and APE group
427: in the mid 80's -- see again Nicola Cabibbo's talk
428: for comments on this -- found that the transition turns 
429: out to be `almost' second order, i.e. very weakly first order,
430: like the 3d three state Potts model.
431: 
432: 
433: The same reasoning tells us that the two color model is in the universality 
434: class of the three dimensional $Z(2)$ (Ising) model. 
435: This prediction has been checked with
436: a remarkable precision by Engels and collaborators\cite{engels},  
437: and it is a spectacular confirmation of
438: the general idea of universality and dimensional reduction. 
439: 
440: 
441: \section{Summary and Open Questions for the QCD High T Transition}
442: What do we know about the real world: two (nearly) massless quark
443: $m_q << \Lambda_{QCD}$, and one more heavy?
444: 
445: 
446: We can approach then the 'real' world from two sides, 
447: either decreasing the mass from infinity, or increasing the
448: quark mass from zero.
449: 
450: \subsection{Approaching the physical point from infinite mass}
451: Remember that 
452: in the infinite mass limit QCD reduces  to the pure gauge
453: (Yang Mills) model. Yang Mills systems have
454: a deconfining transition  associated with the realization of the global 
455: $Z(N_c)$ symmetry.
456: This places the system in the Ising 3d universality class for two colors,
457: and makes the transition weakly first order (near second, in fact) for
458: three colors. General universality arguments are 
459: perfectly fulfilled by the deconfining transition.
460: 
461: The $Z(N_c)$ symmetry is broken by the kinetic term 
462: of the action when the quarks are dynamic ($m_q < \infty$) :
463: this particular symmetry description of deconfinement
464: only holds for infinite quark mass. 
465: When light quarks enter the game, the global $Z(N_c)$ 
466: symmetry observed at infinite mass is lost, and the simple description 
467: of confinement
468: in terms of such symmetry is not possible any more. 
469: It should however remain true that
470: color forces at large distance should decrease with temperature: the main
471: mechanism, already at work at $T=0$, is the recombination
472: of an (heavy) quark and antiquark with pairs generated by the
473: vacuum: $\bar Q Q \rightarrow \bar q Q + q \bar Q$. 
474: At high temperature it becomes easier to produce light
475: $\bar q q$ pairs from the vacuum, 
476: hence it is easier to `break' the color string  
477: between an (heavy) quark  and antiquark $\bar Q Q$. 
478: In other words, we expect enhanced screening of the color forces,
479: which should be sharp at a  phase transition (or rapid crossover).
480: It is however worth mentioning that, even if the string `breaks' 
481: bound states might well survive giving rise to a complicated, 
482: non--perturbative dynamics above the critical temperature.
483: The physical scale of these phenomena is  the larger physical
484: scale in the system, i.e. the pion radius.
485: 
486: \subsection{Approaching the physical point from zero mass}
487: For zero bare mass the phase transition is chiral.
488: For three colors, two flavors is second order with $T_c
489: \simeq 170$ MeV. The prediction from dimensional reduction + universality
490: --$O(4)$ exponenents-- 
491: is compatible with the data, but the agreement is not perfect.
492: 
493: If the agreement were confirmed, that would be an argument in favor
494: of the non-restoration of the $U_A(1)$ symmetry at the transition,
495: which is also suggested by the behavior of the masses spectrum.
496: Remember in fact that the chiral partner of the pion is
497: the $f_0$, which is in turn degenerate with the scalar
498: $a_0$ with $U_A(1)$ is realized. All in all, 
499: $U_A(1)$ non--restoration across the chiral transition 
500: corresponds to $m_\pi \simeq m_{f_0} 
501: \ne m_{a_0}$ which is the pattern observed in lattice calculations
502: (once more I refer to meggiolaro's talk).
503: 
504: The transition with three (massless) flavor  turns out to be first
505: order. The question is than as to whether the strange quark should be 
506: considered `light' or `heavy'. In general, the real world will be somewhere
507: in between two and three light flavor, and to really investigate
508: the nature of the physical phase transition in QCD one should work
509: as close as possible to the realistic value of the quark masses,
510: which is a very demanding numerical task.
511: 
512: 
513: 
514: \subsection{What do we know on the real QCD phase diagram}
515: 
516: Among the most prominent open questions, there is of course the
517: behavior of `real' QCD, with two light flavor, and a third one
518: of the order of $\Lambda_{QCD}$, so how and when exactly the 
519: $N_f=2$ scenario morfs with the $N_f = 3$? Also, why is
520: $T_\chi$ much smaller that the pure gauge deconfining transition?
521: 
522: At a theoretical level the question is if it is possible to give
523: an unified description of the two transitions,
524: chiral and deconfining. This question is
525: currently under active investigation:
526:  recent work suggests that a symmetry analysis of the deconfining
527: transition can be extended also to theories with dynamical fermions.
528: The physical argument is rooted in a duality transformation
529: which allows the identification of magnetic monopoles as
530: agent of deconfinement. The order parameter for deconfimenent
531: would that be the monopole condensate \cite{deldebbio}.
532: An alternative approach uses percolation as the common 
533: agent driving chiral and confining transitions
534: \cite{satz}.
535: 
536: One unifying description and perhaps the most dramatic evidence
537: of a phase transition away from the `simple' limits $m_q = 0$
538: and $m_q = \infty$  comes from the equation of state: for
539: $T \simeq 180 $ Mev we observe, from lattice calculations,
540: a sharp increase of the internal energy:
541: the behavior of the internal energy is a direct probe of the
542: number of degrees of freedom, and indicates quark and gluon liberation
543: \cite{lattice_rev}.
544: 
545: Finally, the work \cite{criqcd}
546:  arrives at an
547: interesting picture of the phase
548: diagram of QCD by combining symmetry analysis and phenomenological
549: consideration. Particularly interesting is the prediction 
550: of an endpoint of a first order line stemming from zero temperature
551: chiral transition at finite density, which should be experimentally
552: observable.
553: 
554: 
555: \section{Methods for QCD at finite temperature and density}
556: Here we will concern ourselves with computational schemes for QCD.
557: The methods described here are those used to obtain the results
558: reviewed above on the chiral transition, deconfinement and equation
559: of state in QCD. We will give some details on these methods, we 
560: will explain why they are not immediately applicable at finite density and
561: we will close up with a brief assessment of the current situation
562: for finite density QCD.
563: 
564: The question is how to estimate the physical observables
565: $
566: <O> = Tr O \hat \rho / {\cal Z} \nonumber
567: $
568: starting from the representation of the partition function
569: 
570: \begin{equation}
571: {\cal Z}(\mu, T) = \int_{0}^{1/T} dt \int e^{
572:  -S_G + 
573: \bar \psi (\not \partial + m + \mu \gamma_0) \psi}
574: d \bar \psi d \psi dU \nonumber
575: \end{equation}
576: 
577: 
578: The need for  two integrations (over bosons and over
579: fermions) lends itself naturally to two different paths:
580: either integrate gluons first, or fermions.
581: 
582: In the first case (if gluons are integrated out first) 
583: we end up with a purely effective fermionic model:
584: \begin{equation}
585: {\cal Z} (T, \mu, \bar \psi, \psi, U) \simeq {\cal Z} (T, \mu, \bar \psi, \psi)
586: \end{equation}
587: As the integration over gluons 
588: cannot be done exactly 
589: \footnote{ There indeed a a systematic
590: approach to QCD based on the strong coupling expansion which makes
591: this integrations easy, but this is not in the scope of 
592: this introduction},
593:  such effective models are often build with the help of a symmetries'
594: analysis for QCD. The gauge fields enters the the game
595: under the guise of coefficients for such models. 
596: Very important examples
597: of this approach include instanton models, chiral perturbation
598: theory, four fermion models
599: such as those discussed by Nardulli and Sannino at this
600: meeting.
601: 
602: The other way takes advantage of the bilinear nature 
603: of S in the fermionic fields, yielding the exact expression:
604: \begin{equation}
605: {\cal Z}(T, \mu, U) = 
606: \int dU e ^{-(S_g - log(det M))}
607: \end{equation}
608: The above integral needs being regulated, and 
609: scheme for calculating 
610: it has to be devised. Both tasks are accomplished within the
611: lattice approach.
612: 
613: \subsection{ Lattice field theory at T,$\mu \ne$ 0}
614: 
615: Temperature comes for free on a lattice: the lattice has a
616: finite extent $N_t a$, hence temperature is given by
617: $T = 1 / N_t a $. The discretization can be carried
618: out in complete analogy to $T=0$, and most of the techniques
619: developed there (see e.g. the recent review 
620: for an introduction to lattice field theory at zero temperature) 
621: apply at finite temperature as well.
622: 
623: A finite density of baryons $\mu_B J_0$ 
624: \footnote{
625: Remember again that $L(\mu) = L_0 + \mu J_0$,
626: $J_0 = \bar \psi \gamma_0 \psi$, i.e. $ N - \bar N = \int J_0$}
627: poses instead specific problems. 
628: \vskip .5 truecm
629: \noindent
630: \underline{How to discretize $\mu \bar \psi \gamma_0 \psi$}
631: 
632: Recall first the `natural' discretization of the matter fields
633: $\phi(x)$  and their derivatives $\partial_\mu \phi(x)$ 
634: on a regular lattice with spacing $a$:
635: 
636: \begin{eqnarray}
637: \phi_{LATT}(n_1, n_2, n_3, n_4)& = &\phi(n_1a, n_2a, n_3a, n_4a)  \nonumber \\
638: \Delta_\mu
639: \phi_{LATT}(n_1, n_2, n_3, n_4) & = &  
640:  (\phi(n_1a, (n_\mu + 1)a, n_3a, n_4a) - \nonumber \\ && 
641: \phi(n_1a, (n_\mu a, n_3a, n_4a))/2a \nonumber
642: \end{eqnarray}
643: 
644: 
645: This, for instance, is the correct 
646: prescription for the chiral condensate:
647: $ m \bar \psi \psi \rightarrow m \bar \psi_{LATT} \psi_{LATT}$
648: but not, as we will see in a moment, for the baryon density!
649: $ \mu \bar \psi \gamma_0 \psi \rightarrow \mu \bar \psi_{LATT} \gamma_0
650: \psi_{LATT} $ is not the correct lattice form , i.e.
651: {\em The naive discretization is not adequate for baryon density}.
652: 
653: Let us then consider free fermions in the \underline{continuum} 
654: $$S = \int_0^\beta \bar \psi \gamma_mu \partial_\mu \psi
655: + m \bar \psi \psi + \mu \bar \psi \gamma_0 \psi$$
656: 
657: The internal energy 
658: $\epsilon = \frac {1}{V} \frac {\partial}{\partial \beta} ln 
659: { \cal Z} 
660: = \frac {4}{(2 \pi) ^4 } \int d^4 p 
661:  \frac 
662: { (p_0 + i \mu)^2 } 
663: { (p_0 + i \mu)^2 + p^2 + m^2 }$,
664: after subracting the vacuum energy, is finite at 
665: $T=0$: 
666: $ lim _{T \to 0} \epsilon =  \frac {\mu^4}{4 \pi^2}$
667: and gives the expect result
668: 
669: By use of the naive discretization,
670: $\epsilon$ 
671: would instead diverge in the continuum ($a \to 0$):
672: $ L = \bar \psi_x  
673: \gamma_\mu \psi_{x + \mu a} + m \bar \psi_x \psi_x +
674:  \mu \bar \psi_x \gamma_0 \psi_x \epsilon  
675: \propto \frac{\mu^2}{a^2} \to_{a \to 0} \infty$
676: 
677: 
678: \vskip .4 truecm
679: \noindent
680: \underline { The solution \cite{lattice_rev}: $\mu$ is an external field} 
681: 
682: Note the analogy:
683: $\bar \psi \gamma_\mu A_\mu \psi \leftarrow \rightarrow  
684: i \mu \bar \psi \gamma_0 \psi$. It shows us that
685: $\mu$ looks like an external field in the time direction.
686: But we know how to 'put' external field on a lattice: they
687: live on the lattice links: for instance, in electrodynamics 
688: $A  \to \theta = e^{ i A} $  (and, for the gauge fields, this
689: implements in a natural and elegant way gauge invariance)
690: 
691: 
692: These considerations suggest how to put finite density on a lattice:
693: \begin{equation} 
694: L(\mu) = \bar \psi_x \gamma_0 e^{\mu a}\psi_{x + \hat 0} -
695:       \bar \psi_{x + \hat 0} \gamma_0 e^{-\mu a}\psi_{x }
696: \end{equation} 
697: Indeed, this turns out to be the correct prescription: unphysical
698: divergences disappear and the  continuum limit is reproduced.
699: 
700: There is also a very expressive physical interpretation: as we can see,
701: forward propagation is encouraged and backward propagation is suppressed:
702: we are indeed inducing an asymmetry particles-antiparticles.
703: 
704: The 0-th component of the current $j_0$ counts indeed the differences
705: between backward and forward propagation:
706: \begin{equation}
707: J_0 = - \partial_\mu L = \bar \psi_x \gamma_0 e^{\mu a}\psi_{x + \hat 0} +
708:       \bar \psi_{x + \hat 0} \gamma_0 e^{-\mu a}\psi_{x }
709: \end{equation}
710: and reproduces the correct continuum limit.
711: 
712: \vskip .4 truecm
713: \noindent
714: \underline{ Chemical potential and boundary conditions}
715: 
716: By use of an unitary transformation of the fields it is possible
717: to re-express $L$ as
718: 
719: \begin{equation}
720:  L (\mu) = L (0) 
721: \end{equation}
722: with boundary conditions
723: \begin{eqnarray}
724: \phi(x + N_T) &=& e^{\mu N_T} \phi(x) \\
725: \psi(x + N_T) & = & -e^{\mu N_T} \psi(x)
726: \end{eqnarray}
727: 
728: It is of some interest to consider the effect
729: of the chemical potential  on the baryonic propagators in the
730: two cases -- when the chemical potential is included
731: into the Action, and when instead affects only the boundary
732: conditions. 
733: In the first case, 
734: (at zero temperature, and ignoring feedbacks) a term
735: $exp(-N_c \mu)$ multiplies the baryonic propagator: 
736: this produces an apparent decrease of the baryon mass
737: \begin{equation}
738: m_B = m_B - 3 \mu
739: \end{equation}
740: i.e. the baryon becomes massless at $\mu = \mu_c = m_B/3$. 
741: In the same situation, the chiral condensate remains constant till 
742: $\mu_c$, and then suddenly drops to zero.
743: The behavior of the chiral condensate would be
744: suggestive of a strong first order transition, while the behavior of
745: the baryon mass would suggest a second order transition with $\nu = 1$!
746: 
747: In the other formulation
748: (when the chemical potential only affects the boundary conditions), 
749: at zero temperature the baryon mass is constant till $\mu_c$ : 
750: this is consistent with the behavior of the chiral condensate,
751: and with the physical intuition that nothing should happen till
752: the Fermi level is reached. 
753: Of course, nothing is wrong and the apparent contradiction is
754: resolved by noticing that the apparent decrease of the baryon mass
755: below $\mu_c$ merely reflects the change of the reference energy.
756: The relationships
757: above provide a link between the two pictures.
758: 
759: \vskip .4 truecm
760: \noindent
761: \underline{ QCD at nonzero T and $\mu$ at a glance}
762: 
763: The continuum formulation:
764: \begin{eqnarray}
765: \nonumber
766:  L &=& L_{YM} +  \bar \psi  (i \gamma_\mu D_\mu + m) \psi \nonumber \\
767:  \nonumber  & + & \mu \bar \psi \gamma_0 \psi \nonumber \\
768: \end{eqnarray}
769: $\mu$ is explicitly included via the coupling to $\mu J_0$ , and
770: the temperature is the reciprocal of the imaginary time.
771: 
772: 
773: On a lattice:
774: \begin{eqnarray}
775: L & = &  L_{YM} + \sum_{i = 1}^3 \bar \psi_x U \gamma_i \psi_{x + \hat i} -
776:       \bar \psi_{x + \hat i} U^\dagger \gamma_i \psi_{x }  
777: \nonumber \\ & + &  m \bar\psi\psi +
778: \bar \psi_x \gamma_0 e^{\mu a} U \psi_{x + \hat 0} -
779:       \bar \psi_{x + \hat 0} U^\dagger \gamma_0 e^{-\mu a}\psi_{x } \nonumber
780: \end{eqnarray}
781: 
782: $\mu$ appears as a link term, and the temperature is again the reciprocal
783: of the imaginary time. The two formulations coincide in the limit
784: $a \to 0$.
785: 
786: \subsection {From the formulation to the results }
787: 
788: Let us write again:
789: \begin{equation}
790: S = S_{YM}(U) + \bar \psi M (U) \psi
791: \end{equation}
792: By taking advantage of $S$ bilinearity in the fields $\psi, \bar \psi$
793: we can write
794: \begin{equation}
795: { \cal Z } = \int e^{-S_{YM}(U)} det M(U) dU 
796: \end{equation}
797: It is convenient to define an 'effective' Action
798: \begin{equation}
799: S_{eff}(U)= S_{YM}(U) - ln( det M(U)) 
800: \end{equation}
801: Averages of purely gluonic observables can be expressed as
802: \begin{equation}
803: <f(U)> = Z^{-1} \int dU e^{-S_{eff}(U)} f(U) 
804: \end{equation}
805: while fermion bilinears can be evaluated with the help
806: of Grassman algebra
807: \begin{equation}
808: <\bar q q > = Z^{-1} \int dU e^{-S_{eff}(U)} Det M(U)^{-1} 
809: \end{equation}
810: In conclusion, if we know how to treat $\int e^{-S_{eff}(U)}  dU $ 
811: we  have access to all of the gluonic observables, chiral condensate,
812: meson and baryons propagators (masses, decay constant, etc.) etc.
813: much in the same way as at $T=0$.
814: 
815: In practical numerical works lattice discretization is combined with
816: importance sampling: a configuration of gauge fields [U]
817:  is a 'point'
818: in a multidimensional integration space. A Markov chain of points
819: is then created according to the prescription:
820:     $P([U]) \propto e ^{-S_{eff}([U])}$
821: Expectation values are then given by simple averages:
822: $<O> = lim_{N \to \infty} 1/N \sum_{i = 1}^N O(U) $.
823: 
824: The prescription above relies on importance sampling: $S_eff([U])$
825: must be positive \footnote{If this is not the case, one might think of using
826: $S^\dagger$ instead: this has to be used with care because
827: it introduces extra degrees of freedom which might well be dangerous.}
828: In QCD (eq. above)  $M^ \dagger (\mu) = - M  (-\mu)$ : 
829: importance sampling, hence MonteCarlo evaluation of physical observables,
830: works for purely imaginary chemical potential, including, of course, 
831: $\mu=0$.
832: 
833: For QCD with two color, however, as well as for several fermionic models,
834: the positivity condition is met by the Action at non zero $\mu$. All in
835: all, we have a rather elaborate pattern of warnings and possibilities.
836: 
837: \subsection {Results and possibilities for QCD at $\mu \ne 0$ }
838: 
839: I would classify the many questions one can ask along two main lines:
840: 
841: Firstly, what can be done in order to learn about the{\em general} behavior at 
842: nonzero baryon density?
843: 
844: Secondly,  what can be done in order to learn about  {\em real} QCD at 
845: nonzero baryon density?
846: 
847: I would say that to learn 
848: about the effects of a finite density of baryons, we can study
849: two color QCD : here, not only one can study fermionic observables, but it is
850: also possible to observe the effects of a density of baryons on the gauge 
851: fields. This is indeed an active field of research, and I refer to the
852: proceedings of the Lattice meetings for details.
853: 
854: How about {\em real} QCD? there is a growing consensus that at least the high
855: temperature, small density region is accessible: by increasing temperature
856: it is easier to fluctuate light baryons, hence to explore a region with
857: a moderate baryon density. At
858: least three different, complementary approaches stem
859: from this consideration:direct evaluation 
860: of derivatives at $\mu$ = 0, reweighting, analytic continuation from
861: purely imaginary chemical potential~\cite{kogut}. 
862: This last approach has been discussed
863: by M. D'Elia at this meeting. 
864: 
865: \vskip .2 truecm
866: I hope to have managed at least to describe some of the
867: aspects of a rapidly evolving
868: field, and to convey a feeling of its many reasons of interest.
869: A discussion of the recent  results is not in the
870: scope of this cursory introduction: for this, and for a complete
871: set of references, 
872: I refer to the latest reviews \cite{kogut} (not yet available at the time
873: of the Frascati meeting).
874: 
875: 
876: \begin{thebibliography}{99}
877: 
878: \bibitem
879: {cabibbo} N. Cabibbo, introductory remarks, this volume.  
880: \bibitem{satz} H. Satz, ibid.
881: \bibitem{nardulli} G. Nardulli, ibid.; M. Mannarelli, ibid.  
882: \bibitem{sannino} F. Sannino, ibid.; W. Schaefer, ibid.  
883: \bibitem{deldebbio} L. Del Debbio, ibid.; A. Drago, ibid.; D. Cosmai,
884: ibid. ; A. Papa, ibid.  
885: \bibitem{meggiolaro} E. Meggiolaro, ibid.  
886: \bibitem{delia} M. d'Elia, ibid.  
887: \bibitem{kogrev} Introductory material on the lattice approach
888: to field theory can be found in
889: J. B. Kogut, Rev.Mod.Phys.51 (1979) 659;  M. Creutz, 
890: {\em Quarks, Gluons and Lattice}, Cambridge;
891: C. Itzykson-Douffre, {\em Statistical Field Theory},
892: Cambridge,  Chapter 7.
893: \bibitem{vari} M. Alford, 
894: {\tt hep-ph/0102047};  K. Rajagopal 
895: and F. Wilczek, 
896: in 'At the Frontier of Particle Physics/Handbook of QCD', 
897: M. Shifman, ed., (World Scientific); 
898: R. Rapp, T. Schafer, E.V. Shuryak, M. Velkovsky,
899:  Annals Phys.280 (2000)35.
900: \bibitem{lattice_rev} For an
901:  introduction to numerical simulations
902: of lattice QCD thermodynamics see e.g.
903: C. DeTar, in Hwa, R.C. (ed.): 
904: Quark-gluon plasma, vol.2 , pag. 1 or
905: F. Karsch, hep-lat/0206034.
906: \bibitem{piwi} 
907: R. D. Pisarski and F. Wilczek , Phys.Rev. D29 (1984)338.
908: \bibitem{engels} 
909: J. Engels, S. Mashkevich, T. Scheideler, G. Zinovev,
910: Phys.Lett.B365 (1996) 219.
911: \bibitem{criqcd} M. Stephanov, K. Rajagopal, E. Shuryak, Phys. Rev. Lett
912: Phys.Rev.Lett.81 (1998) 4816; 
913: J. Berges and K. Rajagopal, 
914: Nucl.Phys. B538 (1999) 215;
915:  	M.A. Halasz, A.D. Jackson, R.E. Shrock, M.A. Stephanov 
916:  	and J.J.M. Verbaarschot, Phys.Rev.D58
917: (1998) 096007; M. Stephanov, K. Rajagopal, E. Shuryak, 
918: Phys.Rev.D60 (1999) 114028.
919: \bibitem{kogut} Recent reviews on QCD in exteme conditions include:
920: J. B. Kogut, Lattice2002; Z. Fodor, QM2002; K. Kanaya, QM2002;
921: M. Alford, ICHEP2002.
922: \end{thebibliography}
923: 
924: 
925: \end{document}
926: 
927: \end
928: 
929: 
930: 
931: 
932: 
933: 
934: 
935: 
936: 
937: 
938: 
939: 
940: 
941: 
942: 
943: 
944: 
945: 
946: 
947: 
948: 
949: 
950: 
951: 
952: 
953: 
954: 
955: 
956: 
957: