1: % Chapter 4
2:
3: There are common relevant problems in cosmology, astrophysics and
4: ultrarelativistic heavy ion collisions, that ask for a deeper
5: understanding of the physics of the formation of a non--equilibrium
6: plasma of charged particles beginning from an initial state of large
7: energy density, its evolution, the onset of electric and magnetic
8: screening phenomena and the generation of seeds of bulk electric and
9: magnetic fields \cite{Boyanovsky:1999jh}.
10:
11: Taking as a first example the world of elementary particles, one of
12: the main motivations to the use of out of equilibrium techniques comes
13: from the possibility of studying experimentally the formation and
14: evolution of a deconfined phase in QCD: the quark gluon plasma
15: (QGP). It is important, for example, to understand how the quark-gluon
16: plasma forms and equilibrates from the evolution of the parton
17: distribution functions, to correctly compute the time scales for
18: electromagnetic screening that cuts off small angle scattering and how
19: a hydrodynamic picture of the space-time evolution of the plasma
20: emerges from first principles. Also the comprehension of the possible
21: experimental signatures of the quark gluon plasma and chiral phase
22: transition would be very relevant. Electromagnetic probes (photons and
23: dileptons) could provide clear signatures for an out of equilibrium
24: chiral phase transition and the formation of quark phase transition,
25: because they only interact electromagnetically and their mean free
26: path is much bigger that the size of the fireball. In addition, even
27: though these days QCD is generally accepted as the {\em fundamental}
28: theory of strong interactions, its very rich phenomenology and its
29: complex phase structure have not been completely unveiled yet, as
30: recent results at finite temperature and density (color--flavor
31: locking, color superconductivity and quark--hadron continuity) suggest
32: \cite{Wilczek:1999fr,Wilczek:1998ur,Alford:1998mk,Schafer:1998ef}.
33:
34: Regarding cosmology, the fascinating hypothesis of sphaleron induced
35: B-violating processes inside the Standard Model as the basis for the
36: baryon asymmetry, calls for a more accurate analysis of the phenomena
37: involved.
38:
39: The common ingredient in such different situations is the description
40: of the transport properties, like the screening masses and the
41: electrical conductivity, as well as the determination of the
42: relaxation time scales. To this end, a much deeper understanding of
43: collective excitations in an ultrarelativistic plasma is required: to
44: decide if the local thermodynamical equilibrium (LTE) approximation is
45: justified, we must compute the relaxation time scales, given by the
46: damping rate or the lifetime of the excitations in the plasma. The
47: quasi particle description of the collective excitations treats them
48: as narrow resonances characterized by a Breit-Wigner distribution,
49: whose width is related to the imaginary part of the self--energy on
50: the mass shell. In this approximation, the damping rate turns out to
51: be exponential.
52:
53: Of course, in order to reach physically sensible results, we have to
54: improve the models considered so far, by introducing gauge fields: a
55: gauge invariant description of the real--time dynamics of gauge fields
56: is required. Recently, new and unusual features of relaxation of soft
57: degrees of freedom in gauge theories has been discovered
58: \cite{Boyanovsky:1998pg}. The real time evolution of (abelian) gauge
59: fields is very important because photoproduction is expected to be a
60: very clear experimental signature of the chiral phase transition. It
61: may be also relevant in the description of the creation of primordial
62: magnetic fields in the early Universe.
63:
64: In the framework of the non--equilibrium formalism the Hard Thermal
65: Loop (HTL) resummation scheme of Braaten and Pisarsky can be rederived
66: \cite{Boyanovsky:1998pg}. This is a useful scheme, which produce gauge
67: invariant results and infrared finite transport cross sections which
68: renders the damping rate of the excitations in the plasma
69: \cite{Pisarski:1989cs} finite. In fact, the medium effects such as the
70: Debye screening, the collective plasma modes and Landau damping are
71: taken into account in the resummed Green's functions. However, soft
72: transverse (magnetic) photons remain unscreened and induce IR
73: divergences. This characteristic is common to QED, QCD and scalar
74: QED. The HTL resummation scheme may be regarded as giving a Wilsonian
75: effective action (which turns out to be non--local) for soft modes,
76: after the hard modes have been integrated out. Since the soft modes
77: have in general a non perturbative dynamics leading to quite large
78: occupation numbers, it has been hoped to get some information from the
79: numerical simulation of the classical theory. This approach, although
80: justified in scalar field theories, should be considered carefully
81: when applied to gauge theories.
82:
83: With the aim of understanding relaxational dynamics in the quark gluon
84: plasma and electroweak plasma, hot scalar QED has attracted much
85: interest
86: \cite{Boyanovsky:1998pg,Boyanovsky:1998aa}, because it has the same HTL
87: structure as the non abelian case (for the leading term). The real
88: time evolution of inhomogeneous expectation values with soft length
89: scales has been considered in the weak field regime, with linearized
90: equations which admit solutions in closed form through Laplace
91: transform. In such a regime, the results clearly show the dominance of
92: off-shell Landau damping processes. It is thus crucial to understand
93: the time scales of these dissipative processes, their microscopic
94: description and how the relaxation of soft collective excitations
95: proceeds. The analytic structure in the $s-$plane gives the time
96: behavior, which shows a power law tail in the relaxation of transverse
97: field amplitudes and a logarithmic tail for longitudinal fields; this
98: is determined by the behavior of the retarded self--energy at the
99: Landau damping threshold. This is special of HTL at one loop and may
100: change, should high orders be included in the computation. In this
101: linearized regime, it is also possible to compute the Influence
102: Functional \cite{Feynman:1963fq}, and derive an effective Langevin
103: equation for the gauge invariant observables. Also the dissipative and
104: noise correlation functions (kernels) and the fluctuation-dissipation
105: relation between them, is obtained, proving in detail that a Markovian
106: approximation fails to describe correctly the dynamics, when the
107: processes are dominated by Landau damping.
108:
109: Similar techniques have been applied to understand the
110: non--equilibrium dynamics of a thermal plasma in a gravitational field
111: \cite{Calzetta:1994qe,Campos:1998un}. The plasma is described by a
112: massless scalar quantum field and the matter--gravity coupling is
113: treated semiclassically. The quantum matter back reaction on the metric
114: is considered as a noise term in the Einstein equation, which becomes
115: a semiclassical stochastic equation. It is equivalent to a Langevin
116: equation, whose damping and noise kernels satisfy a definite
117: Fluctuation--Dissipation relation.
118:
119: Some results have also been obtained recently, for a condensate of
120: arbitrary amplitude, but uniform in space. In such a situation, a
121: Bloch-Nordsieck resummation of IR divergences, equivalent to a
122: dynamical Renormalization Group resummation, has been applied to
123: scalar QED, in connection with relaxation and damping in real time
124: \cite{Blaizot:1996hd,Boyanovsky:1998aa}. The resummation of
125: quasistatic transverse photons gives an anomalous logarithmic
126: relaxation of the form $\exp(\alpha T t \log t/t_0)$, where $\alpha$
127: is related to the coupling constant and $T$ is the temperature.
128:
129: The dynamical RG resummation to obtain the real time dynamics of
130: relaxation and thermalization is equivalent to the RG resummation of
131: divergences in Euclidean Field Theories. Here, we have the resummation
132: of logarithmic secular terms in the perturbative solution of the
133: evolution equation of expectation values. This analysis implies that a
134: quasi particle description of the relaxation process is allowed only
135: when the perturbative solution displays linear secular terms, while
136: non--linear secular terms in lowest order signal anomalous,
137: non--exponential relaxation.
138:
139: An other important issue is the generation of an out of equilibrium
140: plasma in scalar QED. This problem has been studied within a model
141: containing N charged scalar fields coupled to a $U(1)$ photon, plus a
142: neutral scalar field, which acts as an order parameter and
143: distinguishes between the broken and unbroken symmetry phases. This
144: model provides a convenient description of the chiral phase transition of QCD
145: and allows the study of photon production, the electric and magnetic screening
146: and the build up of electric conductivity. The approximation consists
147: in considering the large $N$ limit in the scalar sector, which takes
148: into account the non perturbative and non linear effects involved in
149: the dynamics out of equilibrium, and the lowest order in $\alpha$,
150: which provides a mechanism of photoproduction.
151:
152: As far as the photon production is concerned, a consistent kinetic
153: equation suitable for strongly out of equilibrium situations has been
154: derived in \cite{Boyanovsky:1999jh}, considering only spontaneous
155: emission of photons (without the inclusion of stimulated
156: processes). The exponential growth in the quantum fluctuations of the
157: matter field, due to parametric amplification or spinodal
158: instabilities, drives the formation of the non equilibrium quantum
159: plasma, and a similar explosive production of photons at early times
160: occurs, while the asymptotic distribution at late times reaches a form
161: \begin{equation}
162: N_{ph}(\omega) \simeq \dfrac{\alpha m^2}{\lambda^2 \omega^3}
163: \end{equation}
164: In case of spinodal instabilities, the asymptotic distribution
165: diverges logarithmically with time and a resummation via the dynamical
166: renormalization group may be in order.
167:
168: The analysis of the photon production allows for an understanding of
169: the mechanism of creation of the electromagnetic field, which shows up
170: in correlated domains of linear size $\xi(t)
171: \simeq \sqrt{t/m}$. The corresponding power spectrum inside the
172: unstable band with an amplitude $\simeq \alpha/\lambda^2$
173:
174: In \cite{Boyanovsky:1999jh} Magnetic and Debye screening masses are
175: defined out of equilibrium, in order to study the dynamical aspects of
176: electric and magnetic screening. While the magnetic mass vanishes out
177: of equilibrium, it is possible to give a definition of a momentum and
178: time dependent magnetic mass which has the advantage of displaying
179: explicitly the different time scales involved in the process. The
180: Debye mass is $\alpha m^2 / \lambda$ for broken symmetry and is
181: proportional to $\sqrt{t}$ for unbroken symmetry, suggesting the
182: necessity of renormalization group resummation for the secular terms,
183: similar to that performed in \cite{Boyanovsky:1998aa}.
184:
185: Such an analysis allows one to follow in detail the onset of the
186: transverse electric conductivity (in the framework of the linear
187: response theory \'a la Kubo) during the non equilibrium evolution and
188: the formation of the quantum plasma. At the end of the early (linear)
189: evolution, the long wavelength conductivity turns out to be
190: proportional to $\alpha m / \lambda$.
191:
192: A comparison with the corresponding equilibrium phenomena is very
193: useful and shows that many results have the same dependence on the
194: particle distribution, which is a time dependent quantity in the case
195: of the non--equilibrium evolution, while is the static Bose--Einstein
196: distribution in case of thermodynamic equilibrium (cfr. also
197: \cite{Niegawa:1997yv}).
198:
199: Scalar QED is very useful to clarify some phenomena which occur in an
200: electromagnetic plasma. On the other hand, most of the non equilibrium
201: phenomena occurring in the ``ordinary'' low--energy matter are
202: described by fermionic QED. In this context, one of the older
203: questions is how the vacuum fluctuations of the photon field affect
204: the effective dynamics of an electron. Moreover, the program of
205: decoherence proposes that the superselection rules coming from the
206: conservation of charges could be induced dynamically: superposition of
207: states with different eigenvalue of the charge operator would rapidly
208: lose their coherence in time. This issue has been considered in
209: \cite{Anastopoulos:1998wh}, where the reduced density matrix for the
210: electron has been computed in perturbation theory by means of the
211: Feynman--Vernon influence functional technique.
212:
213: An other approach is based on the extension of the HTL resummation
214: technique to non equilibrium QFT, by means of the real time formalism
215: in the Keldish CTP representation \cite{Carrington:1997sq}. This
216: extension is based on the assumption of quasi stationary distributions,
217: which occur if the time scale of the microscopic processes in the
218: medium is much smaller than the time scale of relaxation towards
219: equilibrium. In this approximation, the HTL photon self energy, the
220: resummed photon propagator and the damping rate of a hard electron in
221: a QED plasma have the same form out of equilibrium as in equilibrium.
222:
223: Regarding the study of time evolution in non abelian gauge theories, it should
224: be noticed that some work has been done in the framework of classical
225: Yang--Mills theory, solving the evolution equations on a lattice and modeling
226: the two ions by means of classical wave--packets of chromoelectromagnetic
227: field \cite{Krasnitz:1998fy,Krasnitz:1998jb}
228:
229: %{\em introduce problem in QCD (non conservation of BRST charge) and try to
230: %suggest something}