Application of the functional renormalization group to Bose systems with broken symmetry [Elektronische Ressource] / von Andreas Sinner

Andreas Sinner, 1.109, 47808

Sujets

• Physik

Informations

• Publié par: goethe_universitat_frankfurt_am_main
• Publié le: 01 janvier 2009
• Nombre de lectures: 9
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

Application of the Functional Renormalization Group
to Bose systems with broken symmetry
Dissertation
zur Erlangung des Doktorgrades
der Naturwissenschaften
vorgelegt beim Fachbereich FB 13 Physik
der Johann Wolfgang Goethe-Universit¨at
in Frankfurt am Main
von
Andreas Sinner
aus Kuibyschew
Frankfurt (2009)
(D 30)vom Fachbereich FB 13 Physik der
Johann Wolfgang Goethe-Universit¨at als Dissertation angenommen.
Dekan: Prof. Dr. Dirk-Hermann Rischke
Gutachter: Prof. Dr. Peter Kopietz
Prof. Dr. Walter Hoffstetter
Datum der Disputation: 6. Juli 2009Abstract
The physics of interacting bosons in the phase with broken symmetry is determined by
the presence of the condensate and is very different from the physics in the symmetric
phase. The Functional Renormalization Group (FRG) represents a powerful investiga-
tion method which allows the description of symmetry breaking with high efficiency. In
the present thesis we apply FRG for studying the physics of two different models in the
broken symmetry phase.
In the first part of this thesis we consider the classical O(1)−model close to the critical
point ofthe second order phase transition. Employing atruncation scheme based onthe
relevance of coupling parameters we study the behavior of the RG-flow which is shown
to be influenced by competition between two characteristic lengths of the system. We
also calculate the momentum dependent self-energy and study its dependence on both
length scales.
In the second part we apply the FRG-formalism to systems of interacting bosons in
arbitrary spatial dimensions at zero temperature. We use a truncation scheme based on
a new non-local potential approximation which satisfy both exact relations postulated
by Hugenholtz and Pines, and Nepomnyashchy and Nepomnyashchy. We study the RG-
flow of the model, discuss different scaling regimes, calculate the single-particle spectral
density function of interacting bosons and extract both damping of quasi-particles and
spectrum of elementary excitations from the latter.
iiiContents
Abstract i
Foreword 1
1 Functional Renormalization Group 5
1.1 Generalized fields and sources . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Generating functionals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 Basics of the Wilsonian RG . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 Exact RG-flow equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.5 Taking symmetry breaking into account. . . . . . . . . . . . . . . . . . . 19
42 Classical φ -model close to criticality 23
2.1 RG-flow equations in the broken symmetry phase . . . . . . . . . . . . . 23
2.2 Probing sharp-cutoff regularization scheme . . . . . . . . . . . . . . . . . 29
2.3 Additive regularization scheme . . . . . . . . . . . . . . . . . . . . . . . . 34
2.4 FRG-enhanced perturbation theory . . . . . . . . . . . . . . . . . . . . . 40
2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3 Interacting bosons at T = 0 47
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2 Functional integral approach to interacting bosons . . . . . . . . . . . . . 49
3.3 Second order perturbation theory. Beliaev damping . . . . . . . . . . . . 54
3.4 Anomalous self-energy at vanishing momenta . . . . . . . . . . . . . . . . 58
3.5 Flowing effective action . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.6 RG-flow equations for interacting bosons . . . . . . . . . . . . . . . . . . 62
3.6.1 RG-flow equation for the condensate . . . . . . . . . . . . . . . . 63
3.6.2 Flow equation for the normal self-energy . . . . . . . . . . . . . . 64
3.6.3 Flow equation for the anomalous self-energy . . . . . . . . . . . . 68
3.6.4 Flow equations for the interaction and dynamical parameters . . . 71
3.6.5 Flowing propagators . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.7 Evaluation of the flow equations for coupling parameters . . . . . . . . . 73
3.8 Renormalized velocity of the Goldstone-mode . . . . . . . . . . . . . . . 79
3.9 Interaction close to the fixed point. Crossover scale . . . . . . . . . . . . 81
3.10 Exact asymptotic propagators in the infrared limit . . . . . . . . . . . . 84
3.11 Hard-core limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.12 Spectral density function . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.13 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
iiiContents
4 Summary 99
A Bosonic FRG in the real field basis 103
A.1 Functional integral approach in the real field basis . . . . . . . . . . . . . 103
A.2 Flow equations for self-energies and coupling parameters . . . . . . . . . 106
B Deutsche Zusammenfassung 109
Bibliography 131
Ver¨offentlichungen 137
Lebenslauf 139
Danksagung 141
ivForeword
In the present thesis we apply the mathematical apparatus of the Functional Renor-
malization Group (FRG) to systems of interacting bosons. At critical temperature Tc
such systems undergo a phase transition. Below T a new macroscopic fraction emergesc
in the system which is referred to as the Bose-Einstein condensate [1, 2]. The critical
temperature of the phase transition depends on both external and internal quantities
like pressure, strength of the magnetic and electric fields, density and mass of particles,
etc. In non-interacting systems, the quantum coherence of single boson wave functions
is responsible for the macroscopic population of the lowest allowed energetic state. In
2/3this case, the critical temperature is proportional to ρ /m [3], where ρ denotes the
density of bosons. Above the critical temperature quantum coherence is destroyed by
thermal fluctuations and quantum effects do not play any important role anymore.
The emergence of the long-range order in interacting Bose systems is due to the inter-
action between particles [4]. Phenomenologically it is taken into account by introducing
an order parameter representing a macroscopic quantity which is non-zero below and
vanishes at and above the critical temperature. Therefore, the system is said to be in
the ordered phase below the critical temperature. Since the emergence ofthe long-range
orderisalwaysrelatedtothereductionofthesymmetry ofthecorrespondingmathemat-
ical model [5] one often speaks about the symmetry broken phase in order to describe
thephasebelowthecriticaltemperature. Onedistinguishes betweenthefirstandsecond
order phase transitions due to the behavior of the system at the critical temperature.
In the case of the second order phase transition, the order parameter vanishes contin-
uously as the system approaches the critical point from below. In the vicinity of the
critical point several observable quantities diverge power-law-like. The powers of these
divergences are called critical exponents. They are the same for systems sharing the
same symmetries, even though these systems might be very different from each other.
Different mathematical models with the same critical exponents are said to belong to
the same universality class.
The calculation of the critical exponents used to be an agenda for the physicists for
several decades. For a long time, the phenomenological Landau theory of the second-
order phase transitions [5, 6] (and mean-field theories in a broader sense) was believed
to provide a correct prediction for the critical exponents and thus to describe phase
transitions correctly. However, the comparison with the few known exact solutions (e. g.
for the Ising model in two spatial dimensions [7]) clearly demonstrated that the mean-
field predictions for the critical exponents is not always correct. In general, the mean-
field exponents are always false in and below some specific spatial dimension D whichc
is called the upper critical dimension of the corresponding mathematical model [8], and
correct in dimensions above D . The reason for the failure of the mean-field theory isc
1Contents
thatitdoesnottakefluctuations intoaccountwhich becomeimportantinthevicinity of
thecriticalpoint. ThelengthscaleatwhichthishappensisdeterminedbytheGinzburg-
criterionandisreferredtoastheGinzburg-scale[8]. Anattempttoimprovetheaccuracy
by calculating perturbative corrections to the mean-field Hamiltonian usually does not
work well, since in this case fluctuations are taken into account at all length scales
simultaneously, i.e. atirrelevantonesaswell. Perturbationtheoryitselfisoftenplagued
by non-physical divergences. Therefore, there is a necessity to go a completely different
way in order to describe phase transitions correctly. This way was explored by Kenneth
Wilson in the form of his Renormalization Group (RG)-approach [9, 10]. The main idea
behind the Wilsonian RG-approach consists of the elimination of fluctuations scale by
scale from the original microscopic model. The effective Hamiltonian which emerges at
the end of this procedure yields astonishingly accurate results for the critical exponents
for a number of models. Reformulated in terms of generating functionals [11, 12, 13],
the Functional Renormalization Group represents a powerful and promising method,
whose applicability reaches far beyond the determination of the critical exponents.
It is of significant interest to apply the Functional Renormalization Group to systems
with broke