1: % Chapter 1
2:
3: \section{Motivations}
4:
5: There are many interesting physical situations in which the system
6: under consideration evolves through a series of highly excited
7: states (i.e., states of finite energy density).
8:
9: As an example consider any model of cosmological inflation, where the
10: inflaton drives the universe exponential evolution by staying for a
11: certain period in states far away from the vacuum
12: \cite{Boyanovsky:1997mq,Kofman:1994rk,Riotto:1998bt}.
13:
14: On the side of particle physics, the ultra-relativistic heavy-ion
15: collisions, scheduled in the forthcoming years at the Relativistic
16: Heavy Ion Collider of Brookhaven National Laboratory (BNL-RHIC) and at
17: the Large Hadron Collider of the European Laboratory for Particle
18: Physics (CERN-LHC), are supposed to produce hadron matter at very high
19: densities and temperatures and out of thermal equilibrium; in such a
20: regime, an approach based on particle scattering at low density cannot
21: be considered a good interpretative tool. To extract sensible
22: information from the theory new computational schemes are necessary.
23:
24: We need, first of all, to go beyond the simple Feynman diagram
25: description at finite order. The use of resummation schemes, like the
26: Hartree--Fock
27: \cite{Kerman:1976yn,cactus} approximation (HF) and the large $N$ limit
28: \cite{largen_exp}, or the Hard Thermal Loop resummation (HTL) for systems at
29: finite temperature
30: \cite{Pisarski:1989cs,Pisarski:1989vd}, can be considered a first step
31: in this direction. They, in fact, enforce a sum over an infinite
32: subset of Feynman diagrams that, in the case of HTL, are dominant in a
33: given region of the parameter space, where the simple truncation of
34: the usual perturbative series at finite order is not justified. In the
35: case of HF, instead, the approximation resums the diagrams which
36: become dominant when the number of spatial dimensions are high. In any
37: case, all of these schemes allow for a non--perturbative treatment of
38: the dynamics under consideration and resummation techniques are
39: already applied to the computation of scattering cross sections and
40: decay width.
41:
42: This is not enough, however, when one wants to describe systems under
43: extreme conditions of density and temperature during their approach to
44: thermal equilibrium. In this case, the suitable approach to QFT
45: consists in setting up an initial value problem. To this end, the
46: standard formulation based on time--ordered Green's functions is not
47: viable, while a Schroedinger picture approach \cite{Eboli:1988fm} may
48: help in clarifying some aspects, by shifting the interest upon the
49: evolution of local field expectation values rather than on the
50: computation of in--out matrix elements. Thus, one specifies an initial
51: state (generally described by a density matrix functional) and then
52: follows its evolution under the time--evolution operator generated by
53: the quantum Hamiltonian.
54:
55: A similarity with a classical mechanics problem may be useful to
56: clarify the difference between this approach and the in--out
57: formalism. Studying out of equilibrium evolution in QFT is like
58: solving a classical dynamics problem, where one follows the time
59: history of the mechanical system, once the initial conditions (for
60: position and velocity) have been specified. On the other hand, one
61: could also fix the mechanical configurations at an initial time
62: $t_{\rm in}$ and a final time $t_{\rm out}$, and then look for the
63: specific trajectory, if any, which connects the two configurations at
64: the two different times; this is exactly what one does to compute
65: cross sections using S matrix elements, where $t_{\rm in} \to
66: -\infty$, $t_{\rm out} \to +\infty$ and the initial and final state
67: describe free particles.
68:
69: As a consequence, the standard functional formalism used in S matrix
70: calculations, is not appropriate to perform the task of finding the
71: solution of an initial value problem in QFT, because the argument of
72: the standard effective action $\Gamma$ is the matrix element of the
73: quantum field operator between an {\em in} and an {\em out} state. In
74: fact, if one tries to derive field equations from such an effective
75: action, complex and non causal quantum corrections appear. On the
76: contrary, the correct order parameter is obtained as the average value
77: of the field operator at a fixed time. Thus, while the effective
78: potential (or the Landau--Ginzburg--Helmholtz free energy) has been
79: very useful in clarifying matters in static configurations, the non
80: equilibrium evolution of the expectation value of quantum fields
81: should be studied from {\em first principles}, resorting to
82: non--equilibrium formalism and addressing issues like dissipative
83: dynamics, decoherence and phase transitions out of equilibrium
84: \cite{Calzetta:1988bw1,Calzetta:1988bw2,Paz:1990jg,Paz:1990sd,Gleiser:1994ea}.
85:
86: Of course, the Cauchy problem in QFT is far too difficult to be solved
87: exactly (which would correspond to solve exactly either the time
88: dependent Schroedinger equation for the wave--functional or the
89: quantum Heisenberg equations of motion for the time-dependent field
90: operators). Actually, as we will see in great detail in the following
91: chapters, some sort of approximation is always applied, usually in the
92: form of a dynamics reduced to a subspace of the Hilbert space of the
93: system.
94:
95: To summarize, by {\em Quantum Field Theory Out of Equilibrium} I mean
96: the study of the dynamics of non-equilibrium processes from a
97: fundamental field-theoretical description, by deriving and solving the
98: dynamical equations of motion for the expectation values and
99: correlation functions of physically relevant observables, in the
100: underlying $(3+1)-$dimensional quantum field theory. Typical examples are:
101: phase transitions and particle production out of equilibrium, symmetry
102: breaking and dissipative processes, relaxation and transport processes
103: in extreme environment.
104:
105: The study of real time evolution in Quantum Field Theory dates back to
106: the early 60's and can be rigorously formulated by means of the
107: so--called Closed Time Path (CTP) generating functional
108: \cite{CTP1,CTP2,CTP3,Chou:1985es,Landsman:1987uw} (cfr. section
109: \ref{CTPsec}), or related formalisms like the Feynman--Vernon
110: influence functional technique \cite{Feynman:1963fq} and the Zwanzig's
111: projection method
112: \cite{Anastopoulos:1997qp}. The problem, even if well defined, is so
113: complex that little can be done analytically even in the simplest
114: situation. Thus, the method was used quite rarely until the beginning
115: of the 90's. A decade ago, a large research program was started by
116: different research groups
117: \cite{QFTooE1,Boyanovsky:1993pf,QFTooE3,Baacke:1997se}, who obtained useful
118: results in realistic physical systems, by heavily exploiting the power
119: of the modern computing facilities. In fact, it should be clear that
120: one will have to resort to numerical techniques in order to solve the
121: partial non linear differential equations of any realistic theory;
122: thus, a formulation of the problem practical for numerical methods is
123: crucial. However, it must be also said that the real time dynamics of
124: phase transitions can be studied in a concrete way with presently
125: available computers.
126:
127: I finally want to emphasize from the beginning the interdisciplinary
128: character of such an approach, which can be applied with considerable
129: impact in the fields of cosmology, astrophysics, particle physics,
130: quantum optics and statistical mechanics out of equilibrium. Relevant
131: results have been obtained in the study of pion condensates,
132: supercooled phase transitions (strongly out of equilibrium),
133: inflationary cosmology and early universe dynamics, strong field
134: electrodynamics in astrophysical plasmas, the hadronization stage of
135: the quark gluon plasma, particle production in heavy ion collisions,
136: dynamical symmetry breaking and dissipative processes, decoherence
137: processes in the transition from quantum to classical regime.
138:
139: For the reader's benefit, I will summarize here the main topics
140: addressed and some remarkable results obtained in the framework of
141: Quantum Field Theory Out of Equilibrium.
142:
143: \subsection*{Cosmological issues}
144: The method has been used to study the non--equilibrium aspects of
145: cosmological phase transitions, in an inflationary
146: scenario. The Inflationary paradigm \cite{Guth:1981zm} (for recent
147: developments see also ref. \cite{Kolb:1999ar}) is one of the greatest
148: application of QFT; ``rolling'' equations in the new inflationary model
149: and the corrections introduced by thermal and quantum fluctuations
150: \cite{Boyanovsky:1997mq,Boyanovsky:1994pa,Boyanovsky:1996sv,
151: Cormier:1998nt,Boyanovsky:1999jg} are studied using the methods
152: of non equilibrium quantum statistical mechanics.
153:
154: This formalism is able to give the quantum evolution equations for the
155: inflaton field coupled to Einstein's equation for the metric, allowing
156: for a semiclassical treatment of the coupled gravity-matter problem
157: and showing the existence of inflationary solutions also in this case
158: \cite{Boyanovsky:1997mq}. The subject is relevant to the study of {\em
159: Reheating} of the Universe at the end of inflationary epoch, which can
160: be treated in a full dynamical fashion, without the use of static
161: quantities like the effective potential. Reheating proceeds by means
162: of either parametric (for chaotic models) or spinodal (for new
163: inflation models) instabilities, with profuse particle
164: production. Both are non--perturbative out of equilibrium mechanisms
165: and require a self--consistent non--perturbative approximation scheme
166: to be used, in order to be correctly described in a quantitative
167: manner. The first stage leads to a non thermal momentum spectrum for
168: the produced particles, and is called {\em Preheating} for this
169: reason. After and maybe together with that, thermalization occurs via
170: standard scattering processes among particles \cite{Felder:2000hr}.
171: Moreover, it is widely accepted that our Universe undertook several
172: phase transitions during its cooling down from the initial Big
173: Bang. The last transitions, namely the deconfining and the chiral
174: phase transitions, may be experimentally proved at the modern
175: accelerators, as the BNL-RHIC, which started its activity quite
176: recently, and the CERN-LHC. Even the baryon asymmetry (the
177: predominance of matter over antimatter) may be explained by an
178: electroweak phase transition in non--equilibrium conditions
179: \cite{Sakharov:1967dj}. The efficiency of such an electroweak baryogenesis may
180: be quantitatively measured by the use of the non equilibrium formalism. In
181: fact, while the equilibrium and static properties of phase transitions are by
182: now well understood, we still need to clarify many points about the dynamics
183: of the processes involved as a phase transition proceeds in time and out of
184: thermal equilibrium.
185:
186: Early attempts to analyze the temperature-induced symmetry--changing
187: phase transitions, that are the crucial ingredients in inflationary
188: cosmology, were based on the use of the static, temperature dependent,
189: effective potential, with the temperature changing in time according
190: to some plausible rule. But the methods of thermal field theory are of
191: no help, unless we are able to show that the evolution proceeds so
192: slowly that local thermal equilibrium is maintained all along the
193: phase transition. If this is not the case, it should be noticed that
194: the effective potential is useless, being a quantity defined strictly
195: at equilibrium. As such, it can give information about static
196: properties like possible stable or metastable equilibrium states and
197: critical temperature, but it loses all information about real time
198: processes, like the approach to equilibrium.
199:
200: An other important goal is to study the process of domain formation
201: and growth in an expanding cosmology, which is very relevant for our
202: understanding of the formation of scalar density perturbations during
203: the phase transition, may have left an imprint in the fluctuations of
204: the cosmic microwave background and should have allowed the formation
205: of large scale structures in the Universe.
206:
207: Such theoretical efforts are justified by an intensive experimental
208: work that has transformed Cosmology in a truly observational
209: discipline. Indeed, the precise determination of the power spectrum of
210: the anisotropies in the Cosmic Microwave Background Radiation
211: temperature, yielded first by COBE \cite{cobe}, very recently by
212: BOOMERANG \cite{boomerang} (much more detailed measurements are
213: expected from MAP \cite{map} and PLANCK \cite{planck} missions),
214: allows to test the quantitative predictions of inflationary models
215: based on particle physics scenarios and in the near future will
216: certainly validate or rule out specific proposal.
217:
218: \subsection*{Quantum Decoherence}
219: An other interesting subject is the study of the decoherence process,
220: which means the transition from quantum to classical behavior. Since
221: quantum superpositions of different mean field states are so difficult
222: to observe in nature, the dynamics should be such that the
223: interference between macroscopically distinguishable states is
224: dynamically suppressed
225: \cite{Cooper:1994hr,Habib:1996ee,Anastopoulos:1998wh}. The subject may clarify
226: the dynamics of phase separation in strongly supercooled phase transitions
227: both in Minkowsky
228: \cite{Boyanovsky:1994pa,Cooper:1997ii} (like the formation and evolution of
229: defects in $^4$He after a sudden quench) and in cosmological backgrounds
230: \cite{Boyanovsky:1999jg} and also some aspects of charged particles dynamics
231: \cite{Anastopoulos:1998wh};
232: with this regard, a formulation based on the evolution of a density matrix is
233: very useful because reveals the emergence of a semiclassical stochastic
234: description. In mean field approximations, the decoherence process seems to
235: proceed through a dephasing of the different modes and causes a growth of a
236: suitably defined effective entropy \cite{Cooper:1997ii}.
237:
238: \subsection*{Bose--Einstein Condensation}
239: From the condensed--matter side, the recent experiments on the
240: Bose--Einstein condensation of dilute atomic gases in laser traps at
241: ultracold temperatures have raised a wide interest \cite{science1,
242: science2,science3}. I think that a true understanding of this
243: phenomenon requires the control of non equilibrium properties, since
244: the characteristic time scales and the temperatures are not those
245: typical of thermalization processes.
246:
247: \subsection*{Disoriented Chiral Condensates}
248: One of the phenomena which may occur in hadron physics at high
249: energies, is the possible formation of Disoriented Chiral Condensates
250: (DCC's), which are regions of misaligned vacuum in the isospin
251: internal space of the pions, analogous to misaligned domains in a
252: ferromagnet. Such regions might act as pion lasers, in the sense that
253: they may relax to the standard ground state by coherent emission of
254: pions, with highly non gaussian charge distribution. This may provide
255: an explanation for the Centauro (overwhelmingly charged) and
256: Anti-Centauro (overwhelmingly neutral) cosmic-ray events and no other
257: processes besides a QCD phase transition out of equilibrium have been
258: proposed, which could produce such a signature.
259:
260: The influence of quantum and thermal fluctuations on the dynamics of
261: the formation of Disoriented Chiral Condensate has been studied in the
262: framework of Quantum Field Theory Out of Equilibrium; this subject may
263: be experimentally relevant at present accelerator energies
264: \cite{Boyanovsky:1994pa,Boyanovsky:1995yk}.
265:
266:
267: \section{Layout presentation}
268:
269: Of course, such a broad field of applicability cannot be covered even
270: partially by the issues addressed in a Ph.D. thesis. Many of the topics
271: illustrated in the previous section,
272: however, have been already clarified by means of a thorough analysis
273: of various dynamical aspects in the scalar $\phi^4$ theory in
274: $(3+1)$D. Many physical systems, in fact, at different levels of description,
275: may be modeled using such a field theory. Nevertheless, a complete
276: understanding of the topic is still lacking and a relatively little
277: work has been done to clarify the matter with respect to other field
278: theories.
279:
280: Thus, the specific subject of this thesis is the study of the real
281: time dynamics in two specific models with opposite renormalization
282: properties in the ultraviolet, namely the ``trivial'' $\lambda \phi^4$
283: model in $(3+1)$D and the asymptotically free non linear $\sigma$
284: model in $(1+1)$D, mainly considering the evolution of translation
285: (and rotation) invariant states.
286:
287: Thus, in chapter \ref{sqft}, after a brief introduction to the CTP
288: formalism, an amusing exercise is performed in Quantum Mechanics,
289: considering a harmonic oscillator plus a quartic perturbation. Then,
290: the $\lambda \phi^4$ model in $(3+1)$D is studied in a finite volume
291: and in the infinite $N$ limit both at equilibrium and out of
292: equilibrium, with particular attention to certain fundamental features
293: of the broken symmetry phase. The numerical solution of the dynamical
294: evolution equations shows that the zero--mode quantum fluctuations
295: cannot grow macroscopically large starting from microscopic initial
296: conditions. I conclude that a Bose--Einstein condensation of a
297: ``novel'' form is implied by the non--equilibrium dynamics. On the
298: other hand, out of equilibrium the long--wavelength fluctuations do
299: scale with the linear size of the system, signalling dynamical
300: infrared properties quite different from the equilibrium ones
301: characteristic of the same approximation scheme.
302:
303: Now, it is to be noticed that the large $N$ limit of $O(N)$ $\phi^4$ model is
304: a completely {\em gaussian} theory of {\em transverse} modes. I am
305: interested in going beyond both this approximations and in finding a scheme
306: which allows an interaction between the longitudinal and the transverse modes
307: and is able to retain some non gaussian feature of the complete theory.
308:
309: Consequently, I consider also an other non--perturbative scheme,
310: namely a non--gaussian time dependent Hartree--Fock approximation
311: (tdHF) both at equilibrium and out of equilibrium. I concentrate
312: myself to the simplest case $N=1$, with particular attention to the
313: structure of the ground state and of certain dynamical features in the
314: broken symmetry phase. I derive the mean--field coupled
315: time--dependent Schroedinger equations for the modes of the scalar
316: field and I outline a suitable procedure to renormalize them. A
317: further controlled gaussian approximation of this new tdHF approach is used
318: in order to study the dynamical evolution of the system from
319: non--equilibrium initial conditions characterized by an uniform
320: condensate. I find that, during the slow rolling down, the
321: long--wavelength quantum fluctuations do not grow to a macroscopic
322: size but do scale with the linear size of the system, in accordance
323: with similar results valid for the large $N$ limit of the $O(N)$
324: model. This behavior is at the basis of an argument proving that the
325: gaussian approximation within this tdHF approach is inconsistent. It
326: would be interesting to numerically solve the quantum Schroedinger
327: equations (\ref{H_k}) and (\ref{H_k1}), in order to have a better
328: comprehension of this inconsistencies.
329:
330: In addition, I study the dynamics of the system in infinite volume
331: with particular attention to the asymptotic evolution in the broken
332: symmetry phase. I am able to show that the fixed points of the
333: evolution cover at most the classically metastable part of the static
334: effective potential.
335:
336: As a first step in future directions, I derive the fundamental equations
337: for the case $N>1$ and I outline some problems in completing the
338: renormalization procedure safely.
339:
340: I also relax the hypothesis of spatial uniformity, deriving the
341: evolution equations for a rotationally invariant scalar condensate in
342: the large $N$ limit. A suitable algorithm to solve the partial
343: derivatives equations is presented and a possible implementation on a
344: PC's cluster of the {\em Beowulf} class is outlined.
345:
346: The $\phi^4$ model in $(3+1)$D dimension is a {\em trivial} theory
347: \cite{Wilson:1974jj}: it suffers from complete screening and would
348: have a vanishing coupling constant in the absence of a cutoff. Thus,
349: it must be always considered as an effective theory, valid below a
350: certain scale of energy. It is well known that there exist theories
351: with a completely different ultraviolet behavior, like the non linear
352: $\sigma$ model in $(1+1)$D and QCD. They are asymptotically free
353: theories in the ultraviolet, which means that the running coupling
354: decreases with the increase of the energy scale.
355:
356: With the aim of starting an analysis of the dynamical properties of
357: such theories, the out--of--equilibrium dynamics of the $O(N+1)$
358: nonlinear $\sigma$--model in $1+1$ dimensions is investigated in chapter
359: \ref{smodel}, performing the large $N$ limit. Regarding the nonlinearity as
360: the effect of a suitable large coupling limit of the $O(N+1) \, \phi^4$ model,
361: I first of all verify that the two limits commute, so that the $O(\infty)$
362: nonlinear $\sigma$ model is uniquely defined. Thanks to asymptotic
363: freedom, such model can be completely renormalized also in the
364: out--of--equilibrium context. I numerically study the spectrum of
365: produced particles during the relaxation of an initial condensate and
366: find no evidence for parametric resonance, a result that is consistent
367: with the presence of the nonlinear constraint. Only a weak nonlinear
368: resonance at late times is observed.
369:
370: I present in chapter \ref{gauge} the results obtained so far by other research
371: groups, studying the dynamical evolution in abelian gauge theories like scalar
372: and fermionic QED. The subject is relevant to the comprehension of the
373: formation and evolution out of equilibrium of a plasma of electrically charged
374: particles, as it may happen in certain astrophysical configurations. I also
375: outline the main difficulties which one has to face, when the same approach is
376: applied to the study of non abelian gauge theories.
377:
378: Finally, in chapter \ref{conclusion} I summarize the results and I comment
379: on some open issues in the subject.
380:
381: I close with an appendix containing some details on technical issues
382: about the analytical and numerical computations performed in this work.
383:
384: In conclusion, understanding out of equilibrium processes in Quantum
385: Field Theory requires the use of non perturbative approximation
386: schemes. On the other hand, the results derived by the use of such
387: schemes should be carefully checked with respect to the approximations
388: used, in order to understand to what extent those conclusions can
389: be extended to the full theory and to the phenomena that theory is
390: modeling. With this respect, the use of more and more powerful
391: computers will be of huge help in the near future, in order to
392: investigate quantitatively the phenomenology of QFT out of equilibrium
393: at a deeper and deeper level, especially in the still almost
394: unexplored arena of non abelian gauge theories.
395:
396: \vskip 0.5cm
397: Let me conclude this introductory chapter with the words of Frank Wilczek
398: \cite{Wilczek:1999fr}: {\em As physicists we should not, however, be satisfied
399: with hoarding up formal, abstract knowledge. There are concrete experimental
400: situations and astrophysical objects we must speak to. Hopefully,
401: having mastered some of the basic vocabulary and grammar, we will soon
402: be in a better position to participate in a two-way dialogue with
403: Nature.}
404:
405:
406: