# Introduction to

**Many-body quantum theory in condensed matter physics
**

Henrik Bruus and Karsten Flensberg

Ørsted Laboratory, Niels Bohr Institute, University of Copenhagen Mikroelektronik Centret, Technical University of Denmark

Copenhagen, 15 August 2002

ii

Preface

Preface for the 2001 edition

This introduction to quantum ﬁeld theory in condensed matter physics has emerged from our courses for graduate and advanced undergraduate students at the Niels Bohr Institute, University of Copenhagen, held between the fall of 1999 and the spring of 2001. We have gone through the pain of writing these notes, because we felt the pedagogical need for a book which aimed at putting an emphasis on the physical contents and applications of the rather involved mathematical machinery of quantum ﬁeld theory without loosing mathematical rigor. We hope we have succeeded at least to some extend in reaching this goal. We would like to thank the students who put up with the ﬁrst versions of this book and for their enumerable and valuable comments and suggestions. We are particularly grateful to the students of Many-particle Physics I & II, the academic year 2000-2001, and to Niels Asger Mortensen and Brian Møller Andersen for careful proof reading. Naturally, we are solely responsible for the hopefully few remaining errors and typos. During the work on this book H.B. was supported by the Danish Natural Science Research Council through Ole Rømer Grant No. 9600548.

Ørsted Laboratory, Niels Bohr Institute 1 September, 2001

Karsten Flensberg Henrik Bruus

**Preface for the 2002 edition
**

After running the course in the academic year 2001-2002 our students came up with more corrections and comments so that we felt a new edition was appropriate. We would like to thank our ever enthusiastic students for their valuable help in improving this book.

Karsten Flensberg Ørsted Laboratory Niels Bohr Institute

Henrik Bruus Mikroelektronik Centret Technical University of Denmark

iii

iv

PREFACE

Contents

List of symbols 1 First and second quantization 1.1 First quantization, single-particle systems . . . . . . . . . . . . 1.2 First quantization, many-particle systems . . . . . . . . . . . . 1.2.1 Permutation symmetry and indistinguishability . . . . . 1.2.2 The single-particle states as basis states . . . . . . . . . 1.2.3 Operators in ﬁrst quantization . . . . . . . . . . . . . . 1.3 Second quantization, basic concepts . . . . . . . . . . . . . . . 1.3.1 The occupation number representation . . . . . . . . . . 1.3.2 The boson creation and annihilation operators . . . . . 1.3.3 The fermion creation and annihilation operators . . . . 1.3.4 The general form for second quantization operators . . . 1.3.5 Change of basis in second quantization . . . . . . . . . . 1.3.6 Quantum ﬁeld operators and their Fourier transforms . 1.4 Second quantization, speciﬁc operators . . . . . . . . . . . . . . 1.4.1 The harmonic oscillator in second quantization . . . . . 1.4.2 The electromagnetic ﬁeld in second quantization . . . . 1.4.3 Operators for kinetic energy, spin, density, and current . 1.4.4 The Coulomb interaction in second quantization . . . . 1.4.5 Basis states for systems with diﬀerent kinds of particles 1.5 Second quantization and statistical mechanics . . . . . . . . . . 1.5.1 The distribution function for non-interacting fermions . 1.5.2 Distribution functions for non-interacting bosons . . . . 1.6 Summary and outlook . . . . . . . . . . . . . . . . . . . . . . . 2 The electron gas 2.1 The non-interacting electron gas . . . . . . . . . . . . . . . . 2.1.1 Bloch theory of electrons in a static ion lattice . . . . 2.1.2 Non-interacting electrons in the jellium model . . . . . 2.1.3 Non-interacting electrons at ﬁnite temperature . . . . 2.2 Electron interactions in perturbation theory . . . . . . . . . . 2.2.1 Electron interactions in 1st order perturbation theory v xii 1 2 4 5 6 7 9 10 10 13 14 16 17 18 18 19 21 23 24 25 28 29 29 31 32 33 35 38 39 41

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

vi 2.2.2 Electron interactions in 2nd order perturbation theory . Electron gases in 3, 2, 1, and 0 dimensions . . . . . . . . . . . . 2.3.1 3D electron gases: metals and semiconductors . . . . . . 2.3.2 2D electron gases: GaAs/Ga1−x Alx As heterostructures . 2.3.3 1D electron gases: carbon nanotubes . . . . . . . . . . . 2.3.4 0D electron gases: quantum dots . . . . . . . . . . . . . . . . . . . . . . . . .

CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 44 45 46 48 49 51 52 53 53 56 59 61 63 64 65 68 69 71 73 73 75 78 78 81 85 87 87 88 88 91 91 93 95 95 98 100 102 104 104

2.3

3 Phonons; coupling to electrons 3.1 Jellium oscillations and Einstein phonons . . . . . 3.2 Electron-phonon interaction and the sound velocity 3.3 Lattice vibrations and phonons in 1D . . . . . . . 3.4 Acoustical and optical phonons in 3D . . . . . . . 3.5 The speciﬁc heat of solids in the Debye model . . . 3.6 Electron-phonon interaction in the lattice model . 3.7 Electron-phonon interaction in the jellium model . 3.8 Summary and outlook . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

4 Mean ﬁeld theory 4.1 The art of mean ﬁeld theory . . . . . . . . . . . . . . . . . . . . . 4.2 Hartree–Fock approximation . . . . . . . . . . . . . . . . . . . . . 4.3 Broken symmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Ferromagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1 The Heisenberg model of ionic ferromagnets . . . . . . . . 4.4.2 The Stoner model of metallic ferromagnets . . . . . . . . 4.5 Superconductivity . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.1 Breaking of global gauge symmetry and its consequences . 4.5.2 Microscopic theory . . . . . . . . . . . . . . . . . . . . . . 4.6 Summary and outlook . . . . . . . . . . . . . . . . . . . . . . . . 5 Time evolution pictures 5.1 The Schr¨dinger picture . . . . . . . . . . . . . . . . o 5.2 The Heisenberg picture . . . . . . . . . . . . . . . . 5.3 The interaction picture . . . . . . . . . . . . . . . . . 5.4 Time-evolution in linear response . . . . . . . . . . . 5.5 Time dependent creation and annihilation operators 5.6 Summary and outlook . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

6 Linear response theory 6.1 The general Kubo formula . . . . . . . . . . . . . . . . . . . . 6.2 Kubo formula for conductivity . . . . . . . . . . . . . . . . . 6.3 Kubo formula for conductance . . . . . . . . . . . . . . . . . 6.4 Kubo formula for the dielectric function . . . . . . . . . . . . 6.4.1 Dielectric function for translation-invariant system . . 6.4.2 Relation between dielectric function and conductivity

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

CONTENTS 6.5

vii

Summary and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 . 108 . 111 . 112 . 113 . 113 . 115 . 116 . 116 . 120 . 121 . 121 . 123 . 124 . 125 . . . . . . . . . . . . . . . . . . . . . . . 127 127 128 131 132 132 134 135 136 137 137 141 141 144 145 145 147 147 149 150 151 153 153 153 153

7 Transport in mesoscopic systems 7.1 The S-matrix and scattering states . . . . . . . . . . . . . . . . . . 7.1.1 Unitarity of the S-matrix . . . . . . . . . . . . . . . . . . . 7.1.2 Time-reversal symmetry . . . . . . . . . . . . . . . . . . . . 7.2 Conductance and transmission coeﬃcients . . . . . . . . . . . . . . 7.2.1 The Landauer-B¨ttiker formula, heuristic derivation . . . . u 7.2.2 The Landauer-B¨ttiker formula, linear response derivation . u 7.3 Electron wave guides . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.1 Quantum point contact and conductance quantization . . . 7.3.2 Aharonov-Bohm eﬀect . . . . . . . . . . . . . . . . . . . . . 7.4 Disordered mesoscopic systems . . . . . . . . . . . . . . . . . . . . 7.4.1 Statistics of quantum conductance, random matrix theory . 7.4.2 Weak localization in mesoscopic systems . . . . . . . . . . . 7.4.3 Universal conductance ﬂuctuations . . . . . . . . . . . . . . 7.5 Summary and outlook . . . . . . . . . . . . . . . . . . . . . . . . . 8 Green’s functions 8.1 “Classical” Green’s functions . . . . . . . . . . . . . . . . 8.2 Green’s function for the one-particle Schr¨dinger equation o 8.3 Single-particle Green’s functions of many-body systems . 8.3.1 Green’s function of translation-invariant systems . 8.3.2 Green’s function of free electrons . . . . . . . . . . 8.3.3 The Lehmann representation . . . . . . . . . . . . 8.3.4 The spectral function . . . . . . . . . . . . . . . . 8.3.5 Broadening of the spectral function . . . . . . . . . 8.4 Measuring the single-particle spectral function . . . . . . 8.4.1 Tunneling spectroscopy . . . . . . . . . . . . . . . 8.4.2 Optical spectroscopy . . . . . . . . . . . . . . . . . 8.5 Two-particle correlation functions of many-body systems . 8.6 Summary and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9 Equation of motion theory 9.1 The single-particle Green’s function . . . . . . . . . . . . . . . . . . . 9.1.1 Non-interacting particles . . . . . . . . . . . . . . . . . . . . . . 9.2 Anderson’s model for magnetic impurities . . . . . . . . . . . . . . . . 9.2.1 The equation of motion for the Anderson model . . . . . . . . 9.2.2 Mean-ﬁeld approximation for the Anderson model . . . . . . . 9.2.3 Solving the Anderson model and comparison with experiments 9.2.4 Coulomb blockade and the Anderson model . . . . . . . . . . . 9.2.5 Further correlations in the Anderson model: Kondo eﬀect . . . 9.3 The two-particle correlation function . . . . . . . . . . . . . . . . . . . 9.3.1 The Random Phase Approximation (RPA) . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . 193 . . . .
. . .viii 9. . . . . . . .1 Lowest order approximation . .3. . . . .6. . . a) . . . . . . . . . . . . . .7 Example: polarizability of free electrons . . 161 . . . . . . . 12. . .1 Fourier transform of Matsubara Green’s functions . . . . . . . . . 164 .4 Impurity self-average . . . . . . . . . . . . . . . . . . . . . . . . . .1 Non-interacting particles in external potentials . . 12. 10. . .3 The full Born approximation . . . . . . .2 Connection between Matsubara and retarded functions . . . . . . . . . . . . . 11. . 173 . . . . 12. . . . . . . 10. . . . 11. . . . . . . . . . . . .6.3. . . 189 .2 The Fock self-energy diagram . . . . . . 190 . . . . 12. . . . . . 12. . . . . . . . . . 10. . . . . . . . . . . .4 Self-energy and Dyson’s equation . . 165 . . . . . . 197 . . . . .1 Feynman diagrams for the impurity scattering .1 Matsubara Green’s function for non-interacting particles 10. . 11. . . . . . . .5 Self-energy for impurity scattered electrons . . . . . . . . 12. . . . . . . . . . . . . . . . . 190 . 182 . . . 174 177 . . . . 181 . 10. .3 The cancellation of disconnected Feynman diagrams 12. . . . . . . . . 199 199 199 201 201 202 203 205 206 208 209 209 210 211
10 Imaginary time Green’s functions 10. . . . . . . . . . . . . . . . . . . .2. . . . . . . .3. 167 . . .6 Wick’s theorem . . . . . . 194 . . . 179 . . . 10.4. . . . . . . . . 164 . . 168 . . . . . . . . . . . . . . . .1 Deﬁnitions of Matsubara Green’s functions . . . . . . .4
CONTENTS Summary and outlook . . . . . . .1 The Hartree self-energy diagram . . . . .1. . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Feynman diagrams and external potentials 11. . . . .2 inﬁnite perturbation series!Matsubara Green’s function .1 Advanced functions . 10. . . . . . . . . . . . . . . . .6 Examples of how to evaluate Feynman diagrams . . . . . . . . . . 160 . . . . . . . . . . . . . . . . . . . . . . . . .2 1st order Born approximation . . . . . . . . . . . . . . . . . . . . . . 12. . . . . . . . .8 Summary and outlook . . . . . .3. . 10. . . . . . . .5 Equation of motion . 12. . . . . . . 10. . . . . . . . . . . . . . . . . . . . . . . 157 . .2 Feynman rules for the numerator of G(b. . . .2 Summations over functions with known branch cuts . . . .4 Evaluation of Matsubara sums .3 The Feynman rules for pair interactions . . . . . 156 .2 Elastic scattering and Matsubara frequencies . . . . . . . .1 The perturbation series for G . .3 Single-particle Matsubara Green’s function . 10. . . . . . . . . . . 11. . . 12. . . . . . . . . . . . . . . 163 . . . . . . . .1 Feynman rules for the denominator of G(b. . . . . . . 177 . . . . . . .3 The pair-bubble self-energy diagram . . .5.7 Summary and outlook . . . . . . . . . . .5. . . 10. . . . . . . . . . . . . . . . . 12.5. . . . . .6. . . 11. . . . . . . . .5 The Feynman rules in Fourier space . . .3 Random impurities in disordered metals . . . . . . . . . 184 . . . . . . 11. . . . . 12. . . . . a) . . 12 Feynman diagrams and pair interactions 12. . . . 11. . . . . .5. . . . . . . . . . . .4 The self-consistent Born approximation and beyond 11. . . . . . . .4. 11. . . . . . . 170 . . 161 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 .6 Summary and outlook . . . . . . . . . . . . . . . 11.1 Summations over functions with simple poles . . 10. . .

. . . . . . . . . . .3 RPA resummation of the self-energy . . . . . .
. .3 The ground state energy of the electron gas . 279 . . . . . . . . . . . . .2 The divergence number of self-energy diagrams 13. . . . . . . . . . . .1 Adiabatic continuity .4 The dielectric function and screening . .
. . 220 . 228 . . . . . . . .1 Renormalization of the single particle Green’s function 14. . . . 13. . . . .2 Dynamical screening . . . 14. . . . . . 215 . . . . . . . 16. . .
.3. . .CONTENTS 13 The interacting electron gas 13. . 214 . . . . . . .5 Combined RPA and Born approximation . . . . . . . . . . . . . . . .5 The Cooper instability and Feynman diagrams . . . . . . . . . . . 13. . . .3. . . . . . . . . .2 Electron-phonon interaction and Feynman diagrams . . . . . . . . . . . . . . 15 Impurity scattering and conductivity 15. . .1 Static screening . 13. . . . . . . . . . .2 Landau damping . . . . . 14. . 13. . . . 223 .2. .3 Semi-classical transport equation .
.1 The Green’s function for free phonons . . . . 238 . . 251 . . . . . . . . . . . . . . . .1 Migdal’s theorem . . . . . . . . . .1 The self-energy in the random phase approximation . . . . . . . . 15. . . . . . . . . . . . . 14. . . . . . .3 Mass renormalization? . 14.2 Jellium phonons and the eﬀective electron-electron interaction 16. . . . . . . . . . 15. . . . . . . . . . . . . . . .1 The density dependence of self-energy diagrams 13. . .3 Combining Coulomb and electron-phonon interactions . . . . . . . . 243 . . . .5. . . . . . . . . . . . . . 220 . . .1 Vertex corrections and dressed Green’s functions . . . . . .2. . . . . . . . . . 217 . . . . 16.1. . 13. . . . . 14. . . . . . . 213 . . . . . . . . . . . . .2 Imaginary part of the single particle Green’s function 14. . .2. . . 251 . . . . 275 . . . . .
275 .1. . . . . . . 240 . . . 245 . 231 233 . . . . . . . . .6 Summary and outlook . . . . . . . . . . . . .2 The renormalized Coulomb interaction in RPA . . . . . . . 16. 13.1. . . . . 15. . . .
ix 213 . . . . .3. . .
. . 245 . . . . . . . . 238 . . .4. . . 237 . . 14. . . . . . 13. 235 . . . . . . . . . . . . . 16. . . 227 . . . . . .
. . . . . . . . . . . . . . . . .
. . . .4. . . . 13. . . . .2 The conductivity in terms of a general vertex function 15. . . . . . . . 279 . . . . . . . .2. .5 Plasma oscillations and Landau damping . . .
. . . . . . . 14. . .2 The electron-hole pair interpretation of RPA . . . 233 . . . . . . . . . . . . .1 Plasma oscillations and plasmons . . .4 Microscopic basis of the Fermi liquid theory . . . . . . . . . . . . . . . . . 253 254 259 261 264 273
. .4 Phonon renormalization by electron screening in RPA . . . 215 . . . 276 . . 16. . 14. . . . . . . . . . . . . . . . . . . 281 . . 284
. . .1. . . . . . . . . . . . .2 Semi-classical treatment of screening and plasmons . . . . . . .3 The conductivity in the ﬁrst Born approximation . 218 . . . . . 14. .1 Finite life time of the quasiparticles . . .
16 Green’s functions and phonons 16.1 Calculation of the pair-bubble .
. . . 13. . . . .4 The weak localization correction to the conductivity .5 Outlook and summary .
14 Fermi liquid theory 14. . . . . . . . . . . . . . . . . . . .1 The quasiparticle concept and conserved quantities .5. . . . . 280 . . . . . . . . 230 . . . . . . . . . . . . .4. . . . . . 248 . 13. . . . .
. . . . . . . . . . . .

.
. . . . . . . . . . . . . . . .4 Experimental consequences of the BCS 17. . . . . . . . . .3 Diagonalization of bosonized Hamiltonian . .3. . . . . . . . . . . .
. . . . . 18.spinless case .3. . .3 BCS theory with Green’s functions . . . . . .4 Some useful rules . . . . . .1 Continuous functions in a ﬁnite region . . . . . . . 287 . . . . . . . .6 Tunneling into spinless Luttinger liquid . . . . . 18. . . 17. . . . .4. . . .1 Interacting one dimensional electron system . . . 18. . . .
. .2 speciﬁc heat . . . . . .4 Luttinger liquid with spin . . . . . . . . . . . . . 18. . . .2 Bosonization of Tomonaga model-Hamiltonian . . . . A. . . . . 18. . . . . . . .1 Introduction . . . . . . . . . . . . . 17. .
. .
. . . . . . . . . . . . . . . . .1 The Cooper instability . . . . . . . . . . . . . . . . . . . . 18. . . . . . . . .8. . .
. . . . . 288 . . . . . . . . . . . . . .
18 1D electron gases and Luttinger liquids 18. . . . .
. .
. . .3.
. . . 18. 17. . .2 Carbon Nanotubes . . states . . . . . . . . . . . A Fourier transformations A. . . . . . 17. . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 First look at interacting electrons in one dimension . . . . . . 18. . . . . . . 17. . . . . . . . . . . . A. . . . . . . . .
. . . . 18. . . 18.
CONTENTS 287 . .
. . . . . . . . . 18. . . .5 Green’s functions . . . . . . . . . . . . . . . . . . .4 Real space formulation . . . . . . . . . . . . . 289 289 289 289 289 289 289 289 289 289 290 290 290 290 290 290 290 290 291 291 292 292 292 293 295 326
. . .
. . . . . . .5 Translation invariant systems . . . . . . . . .6. . . . .1 Tunneling density of states . . . .3. . . . . . . . . . . 288 . . . . . . 287 . . . . .
. . . . .2.3 The Luttinger-Tomonaga model . . . . . B Exercises C Index . . . . . . . . . . . . . A. . . . . . . . . . .
. . 18. . . .7 What is a Luttinger liquid? . . . . . . . . . 288 . . . 18. . . . . .x 17 Superconductivity 17. .3 Time and frequency Fourier transforms . . . 287 .8 Experimental realizations of Luttinger liquid physics . . . . . . . . . . . . . . . . . . . .1 One-dimensional transmission line analog . . . . . . . . . . .2 The BCS groundstate . . . .5 Electron operators in bosonized form . . . . . . . . . . . . . . . . . . . . . . . . .8. . . . . . . 288 . . . . . . . . . . . . .2 Continuous functions in an inﬁnite region A. . . . . . . . . . . . . . . . . . . . . . . . . . . .5 The Josephson eﬀect . . . .4. . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .1 Tunneling into the end of Luttinger liquid . . . . . . .
. . . . . . . . .3. . . . .1 Edge states in the fractional quantum Hall eﬀect 18. . . . . 18. . .
. . 18. . . . . . . . .

8. 1 Chap. A0 (k. 7.
6.4 Sec. ikn + iqn ) C R (Q. ω) A. 5. ikn .3. Sec. a† ν a± n a0 A(r.1 Eq. electron) creation operator for particle (fermion.3 10. electron) annihilation/creation operators (state ν) retarded correlation function between A and B (time) advanced correlation function between A and B (time) retarded current-current correlation function (frequency) Matsubara correlation function Cooperon in the Matsubara domain Cooperon in the real time domain speciﬁc heat for ions (constant volume) xi Deﬁnition Sec. t ) A CAB (t. ω).4
. 1
Sec. ω).36) Sec.4. A† b b† b± n B c c† cν . t) A(ν.3. 8. phonon) creation operator for particle (boson. Sec.4 Sec.1
Sec. Sec. 16.1 15. 8.1 10. A(k.1 6.3 Chap. 7. ω) A0 (r.3. (2.4 Sec. Sec. ε) ion CV Meaning operator ♥ in the interaction picture time derivative of ♥ Dirac ket notation for a quantum state ν Dirac bra notation for an adjoint quantum state ν vacuum state annihilation operator for particle (fermion or boson) creation operator for particle (fermion or boson) annihilation/creation operators (state ν) amplitudes of wavefunctions to the left Bohr radius electromagnetic vector potential spectral function in frequency domain (state ν) spectral function (real space. Fourier space) spectral function for free particles phonon annihilation and creation operator annihilation operator for particle (boson. ω) A(r.List of symbols
Symbol ˆ ♥ ˙ ♥ |ν ν| |0 a a† a ν . Sec.4 15. 1. phonon) amplitudes of wavefunctions to the right magnetic ﬁeld annihilation operator for particle (fermion. t ) R (ω) CII CAB C(Q.2 Sec.2. ε.1
Sec. c† ν R CAB (t.

4) Eq. rt ) DR (q. 16 Chap. 16 Chap.11) Chap.9b) Sec.4) Sec. (1. rτ ) D(q. 6. 1. (7. (4.xii Symbol d( ) δ(r) DR (rt. (2. 2nd order perturbation Rydberg energy dispersion relation for BCS quasiparticles energy variable the dielectric constant of vacuum dispersion relation energy of quantum state ν Fermi energy phonon polarization vector dielectric function in real space dielectric function in Fourier space free energy the ﬁlled Fermi sea N -particle quantum state electric potential external electric potential induced electric potential wavefunctions with diﬀerent normalizations wavefunctions in the left and right leads electron-phonon coupling constant (lattice model) electron-phonon coupling constant (jellium model) conductance Deﬁnition Eq. (3.58b)
Eq. (2. t) φext φind ˜ φ. φ ± φLnE .1
. 16 Eq.4 Sec. 1st order perturbation interaction energy of the electron gas. 16 Chap. ω) F |FS φ(r. ν t ) Dαβ (r) ∆k e e2 0 E(r. (4. 3. 7.31) Eq. (9. ω) D(rτ.101)
Eq. φ± RnE gqλ gq G
kλ 0
LIST OF SYMBOLS Meaning density of states (including spin degeneracy for electrons) Dirac delta function retarded phonon propagator retarded phonon propagator (Fourier space) Matsubara phonon propagator Matsubara phonon propagator (Fourier space) retarded many particle Green’s function phonon dynamical matrix superconducting orderparameter elementary charge electron interaction strength electric ﬁeld total energy of the electron gas interaction energy of the electron gas.20) Sec.5
Eq. 6.36) Eq. iqn ) DR (νt.64)
Eq. rt ) ε(k. (1. t) E E (1) E (2) E0 Ek ε εk εν εF ε(rt.4 Sec.

10. 10. r t ) GA (rt.3.3 Sec. γ RA Γ ˜ ˜ ˜ Γx (k.3 Sec. k + q ) Γ0. 12.3 Sec.1 Sec.1 Sec.3
xiii Deﬁnition Sec. r t ) GR (k. 10.LIST OF SYMBOLS Symbol G(rt.2 Sec. r σ τ ) G0 (ντ. r σ τ ) G(ντ.5 Sec. 8.6 Sec.3.1 Sec. 10. ikn ) (n) G0 G(k. ikn ) G0 (ν. 10. k ) G0 (rστ. o free lesser Green’s function free grater Green’s function free advanced Green’s function free retarded Green’s function free retarded Green’s function (Fourier space) lesser Green’s function greater Green’s function advanced Green’s function retarded Green’s function (real space) retarded Green’s function in Fourier space retarded Green’s function (Fourier space) retarded single-particle Green’s function ({ν} basis) Matsubara Green’s function (real space) Matsubara Green’s function ({ν} basis) Matsubara Green’s function (real space four-vectors) Matsubara Green’s function (four-momentum notation) Matsubara Green’s function (real space.32) Sec. 8. ν τ ) G(1. 8. r t ) G> (rt. 8. ν τ ) G0 (k.3 Sec. 11.3. 8.3 Sec. r t ) 0 GA (rt. r t ) 0 G> (rt.3 Sec.3 Eq. (15. 8. 8. 1 ) ˜ ˜ G(k.1 Eq.3 Sec.1 Sec.3. 8. 8. r t ) G< (rt.1 Sec.3.20b)
. ikn ) γ.3 Sec.3 Sec.3 Sec. four vector notation) free (undressed) vertex function a general Hamiltonian unperturbed part of an Hamiltonian perturbative part of an Hamiltonian external potential part of an Hamiltonian interaction part of an Hamiltonian phonon part of an Hamiltonian positive inﬁnitisimal current operator (particle current) electrical current (charge current) Sec. 8.2 Sec. free particles) Matsubara Green’s function (free particles ) n-particle Green’s function (free particles) Matsubara Green’s function (Fourier space) Matsubara Green’s function ({ν} basis. ν t ) G(rστ. r t ) 0 GR (rt. 15.3 Sec. 6. ω) GR (νt.1 Sec. 10. 8.x H H0 H Hext Hint Hph η I Ie Meaning Green’s function for the Schr¨dinger equation o unperturbed Green’s function for Schr¨dingers eq. ikn ) G(ν. 8.3 Sec. ω) GR (k.1 Sec. 10. ω) 0 G< (rt. 10. 10. (8.1 Sec.3.3. r t ) GR (rt. free particles) Matsubara Green’s function (Fourier space. 6. r t ) 0 GR (k.3. free particles) Matsubara Green’s function ({ν} basis. 8. r t ) G0 (rt. frequency domain) scalar vertex function imaginary part of self-energy vertex function (x-component.

(12.4. (6. iqn )
in
Eq.1
Sec. ω) αβ Παβ (q. r t ) αβ ΠR (q. (15.17)
Sec. 1. 1. 14.26) Chap. iωn ) Π0 (q.99a) Eq. 15 Eq. (1. 1. (1.5. 10 Sec. t) Jij kn kF k Meaning current density operator current density operator.4.99a)
Sec.1 Sec.1
0 φ
L λF Λirr m m∗ µ µ n nF (ε) nB (ε) nimp N Nimp ν ω ωq ωn Ω p pn ΠR (rt.5
Eq. diamagnetic term current density operator (momentum space) electric current density operator interaction strength in the Heisenberg model Matsubara frequency (fermions) Fermi wave number general momentum or wave vector variable mean free path or scattering length mean free path (ﬁrst Born approximation) phase breaking mean free path inelastic scattering length normalization length or system size in 1D Fermi wave length irreducible four-point function mass (electrons and general particles) eﬀective interaction renormalized mass chemical potential general quantum number label particle density Fermi-Dirac distribution function Bose-Einstein distribution function impurity density number of particles number of impurities general quantum number label frequency variable phonon dispersion relation Matsubara frequency thermodynamic potential general momentum or wave number variable Matsubara frequency (fermion) retarded current-current correlation function retarded current-current correlation function Matsubara current-current correlation function free pair-bubble diagram
LIST OF SYMBOLS Deﬁnition Eq.5. paramagnetic term current density operator. (1.xiv Symbol Jσ (r) ∆ Jσ (r) A Jσ (r) Jσ (q) Je (r.99a) Eq. 4.2
Chap.34)
.

1.6 Sec.1 Eq. ikn ) ΣF (k. 12. r t ) ΣR (q.5 11. ikn ) σ ΣP (k. 4.1 Sec.5. ikn ) σ ΣH (k. (2. (13. τk uj u(R0 ) uk U ˆ U (t. ikn ) Σk Σ1BA k ΣFBA k ΣSCBA k Σ(l. ikn ) σ ΣRPA (k.6 Sec. 6.96)
Sec.4
Sec. 12. Sec.37) Sec. t ) ˆ U (τ.1
Eq.
11.5. (1. Sec. ω) Σ(q. 7.2
.96) Eq. (14.10)
Sec.LIST OF SYMBOLS Symbol q qn r r r rs ρ ρ0 ρσ (r) ρσ (q) S S σ σαβ (rt. ikn ) σ t t t T τ τ tr τ0 .5. 7.5.39)
Sec. (1. 7.2
Sec. j) Σσ (k. interaction picture Deﬁnition
xv
Sec.5 Eq. 7. Sec. interaction picture imaginary time-evolution operator. 7. τ ) Meaning general momentum variable Matsubara frequency (bosons) general space variable reﬂection matrix coming from left reﬂection matrix coming from right electron gas density parameter density matrix unperturbed density matrix particle density operator (real space) particle density opetor (momentum space) entropy scattering matrix general spin index conductivity tensor retarded self-energy (Fourier space) Matsubara self-energy impurity scattering self-energy ﬁrst Born approximation full Born approximation self-consistent Born approximation general electron self-energy general electron self-energy Fock self-energy Hartree self-energy pair-bubble self-energy RPA electron self-energy general time variable tranmission matrix coming from left transmission matrix coming from right kinetic energy general imaginary time variable transport scattering time life-time in the ﬁrst Born approximation ion displacement (1D) ion displacement (3D) BCS coherence factor general unitary matrix real time-evolution operator.1 Sec. 12.1 11.1 Sec.3 11.6 Eq.

6 Sec.1 Sec.3.xvi Symbol vk V (r). (6. ω) χn (y) ψν (r) ± ψnE ψ(r1 . . iqn ) χR (rt. r2 . r t ) χR (q.5. 13.39) Sec. rn ) Ψσ (r) Ψ† (r) σ θ(x) Meaning
LIST OF SYMBOLS Deﬁnition Sec. . 13.4 Eq.6 Eq.4 Sec.2
BCS coherence factor general single impurity potential Coulomb interaction combined Coulomb and phonon-mediated interaction normalization volume pair interaction Hamiltonian general pair interaction Coulomb interaction RPA-screened Coulomb interaction εk − µ εν − µ Matsubara charge-charge correlation function RPA Matsubara charge-charge correlation function irreducible Matsubara charge-charge correlation function free retarded charge-charge correlation function free Matsubara charge-charge correlation function retarded charge-charge correlation function retarded charge-charge correlation function (Fourier) transverse wavefunction single-particle wave function. W (q) W (r).12)
. W (q) W RPA ξk ξν χ(q. 13. 7.4 Sec. 4. 13.3. iqn ) χirr (q. .4 Sec.1
Sec. r t ) χ0 (q. V (q) V (r). iqn ) χ0 (rt. 13. . V (q) Veﬀ V W W (r). quantum number ν single-particle scattering states n-particle wave function (ﬁrst quantization) quantum ﬁeld annihilation operator quantum ﬁeld creation operator Heaviside’s step function
Sec. 7. 1. 13. (1. 1.2
Sec. iqn ) χRPA (q.2
Sec.

If one in this state measures the dynamical variable associated with the operator A. resulting in the development of quantum electrodynamics (QED) and quantum ﬁeld theory (QFT) in general. A|ψα = α|ψα . from the study of elementary particles on the sub-nucleonic scale to the study of neutron stars and other astrophysical objects on the cosmological scale. one cannot in general predict the outcome with 1
. certain basic concepts are always present in the formulation of quantum theory. Parallel state vectors represent the same physical state. and one therefore mostly deals with normalized state vectors. Schr¨dinger. Quantum theory postulates that all quantum states are represented by state vectors in a Hilbert space. while quantum ﬁeld theory is formulated in the language of second quantization. It has successfully been applied to explain phenomena ranging over many orders of magnitude. Regardless of the representation. By measurement of the type B the quantum state can be prepared to be in an eigenstate |φβ of the operator B. The scale factors are denoted the eigenvalues of the operator. and even the particles themselves got represented by quantized ﬁelds (Jordan and Wigner.Chapter 1
First and second quantization
Quantum theory is the most complete microscopic theory we have today describing the physics of energy and matter. and that all observables are represented by Hermitian operators acting on that space. Historically. It is a fundamental theorem of Hilbert space theory that the set of all eigenvectors of any given Hermitian operator forms a complete basis set of the Hilbert space. Later also the electromagnetic ﬁeld was o quantized (Dirac. By convention. the original form of quantum mechanics is denoted ﬁrst quantization. Any given Hermitian operator A has a number of eigenstates |ψα that up to a real scale factor α is left invariant by the action of the operator. This state can also be expressed as a superposition of eigenstates |ψα of the operator A as |φβ = α |ψα Cαβ . The starting point is the notion of quantum states and the observables of the system under consideration. 1928). Only the inclusion of gravitation stands out as an unsolved problem in fundamental quantum theory. 1927). quantum physics ﬁrst dealt only with the quantization of the motion of particles leaving the electromagnetic ﬁeld classical. In general the eigenstates |ψα and |φβ of two diﬀerent Hermitian operators A and B are not the same. hence the name quantum mechanics (Heisenberg. and Dirac 1925-26). be it ﬁrst or second quantization.

3)
i.g.2
CHAPTER 1. e. and Dzyaloshinskii. t) and A(r. The plane wave representation |k. which also leads to the deﬁnition of adjoint operators. For a given operator A the adjoint operator A† is deﬁned by demanding ψ|A† |φ = φ|A|ψ ∗ for any |ψ and |φ . ↑ . or Shankar. This non-causal element of quantum theory is also known as the collapse of the wavefunction. FIRST AND SECOND QUANTIZATION
certainty. and |ψ and |φ are said to be orthogonal if ψ|φ = 0. and we refer the interested reader to the textbooks by Mahan. through Schr¨dinger’s equation o i ∂t |ψ(t) = H|ψ(t) . If {|ψα } is an orthonormal basis of the Hilbert space. ψ|φ between adjoint “bra” states ψ| and “ket” states |φ . say an electron with charge −e.
1. where electrons orbiting a point-like positively
.1
First quantization. by the inner product of the position bra r| with the state ket. The time evolution of a state vector |ψ(t) is governed by the central operator in quantum mechanics. σ is not always a useful starting point for calculations. Merzbacher. the Hamiltonian H (the operator associated with the total energy of the system). the norm squared of |ψ is given by ψ|ψ . moving in an external electromagnetic ﬁeld described by the potentials ϕ(r. (1.
(1. Landau and Lifshitz. t). In this chapter we will brieﬂy review standard ﬁrst quantization for one and manyparticle systems.2)
An eigenstate describing a free spin-up electron travelling inside a box of volume V can be written as a product of a propagating plane wave and a spin-up spinor. The probability of having any given |ψα as the outcome is given as the absolute square |Cαβ |2 of the associated expansion coeﬃcient. However. between collapse events the time evolution of quantum states is perfectly deterministic. Using the Dirac notation the state ket can be written as |ψk. For example in atomic physics. Fetter and Walecka. t). “bra(c)kets”. One can form inner products. Gorkov. (1. and use standard geometrical terminology. t)
− eϕ(r. The state function (also denoted the wave function) and the ket are related by 1 ψk. σ = √V eik·r χσ (free particle orbital). The corresponding Hamiltonian is H= 1 2m
2 r
i
+ eA(r. and Abrikosov.↑ = |k.1)
Each state vector |ψ is associated with an adjoint state vector (|ψ )† ≡ ψ|. It is only described in probabilistic terms. Based on this we will introduce second quantization. This introduction is not complete in all details. A further connection between the direct and the adjoint Hilbert space is given by the relation ψ|φ = φ|ψ ∗ . For more complete reviews the reader is refereed to the textbooks by Dirac.e. single-particle systems
For simplicity consider a non-relativistic particle. then the above mentioned expansion coeﬃcient Cαβ is found by forming inner products: Cαβ = ψα |φβ . where one simply lists the relevant quantum numbers in the ket.σ (r) = r|k.

m. which in the Landau gauge A = xB ey leads to the Landau eigenstates |n. (1. m. or in the Dirac notation: 1 = dr ν|r r|ν = ν| ( dr |r r|) |ν . ν = (n.1: The probability density | r|ψν |2 in the xy plane for (a) any plane wave ν = (kx . σ). In general a complete set of quantum numbers is denoted ν . where n is an integer. 2. ky . and ν = (n. kz . ky .6)
Similarly. the completeness of a set of basis states ψν (r) means that if a particle is in some state ψ(r) it must be found with probability unity within the orbitals of the basis 2 set: ν | ν|ψ | = 1. ky . σ). 0. The three examples given above corresponds to ν = (kx . .7)
. kz .m (θ. and σ the spin variable. (1. σ = Rnl (r)Yl. φ)χσ (hydrogen orbital). φ) is a spherical harmonic representing angular momentum l with a z component m. A third example is an electron moving in a constant magnetic ﬁeld B = B ez . (b) the hydrogen orbital ν = (4. while Yl.1. σ . Again using the Dirac notation we ﬁnd 1 = ν ψ|ν ν|ψ = ψ| ( ν |ν ν|) |ψ . σ).4)
where Rnl (r) is a radial Coulomb function with n−l nodes. σ) each yielding a state function of the form ψν (r) = r|ν . ky . ky (kz ) is the y (z) component of k. ky . σ = Hn (x/ −ky )e− 2 (x/
1
−ky )2 √
1
Ly Lz
ei(ky y+kz z) χσ
(Landau orbital).5)
where = /eB is the magnetic length and Hn is the normalized Hermite polynomial of order n associated with the harmonic oscillator potential induced by the magnetic ﬁeld.
ν
(1. From this we conclude dr |r r| = 1. σ are much more useful. SINGLE-PARTICLE SYSTEMS
¢ ¤ ¢ £ ¢ ¡
3
Figure 1. Recall that r|n. l. Loosely speaking the normalization condition means that with probability unity a particle in a given quantum state ψν (r) must be somewhere in space: dr |ψν (r)|2 = 1. 1. kz . . and (c) the Landau orbital ν = (3. Examples of each of these three types of electron orbitals are shown in Fig. σ).1. l. m. Recall that r|n. The completeness of a basis state as well as the normalization of the state vectors play a central role in quantum theory.1. charged nucleus are considered. FIRST QUANTIZATION. σ). kz . the hydrogenic eigenstates |n. (1. and we conclude |ν ν| = 1. 0.m (θ. ky . l. kz .

θ(x) = 0. electrons. for k = n. r2 . which (neglecting the spin degree of freedom for the time being) is a complex wave function in 3-dimensional space.↓ n=0
Ly dky −∞ 2π
∞
Lz dkz . say.12)
1. −∞ 2π
∞
(1.
. . and Lz .11)
and Heaviside’s step-function θ(x) for continuous variables. many-particle systems
When turning to N -particle systems. while dr δ(r) = 1.9)
In the mathematical formulation of quantum theory we shall often encounter the following special functions. In that case the symbol ν is to be interpreted as a combination of both summations and integrations. As for one particle this N -particle state function is interpreted as a probability amplitude such that its absolute square is related to a probability: The probability for ﬁnding the N particles N in the 3N −dimensional volume N drj 2 j=1 (1. . a system containing N identical particles. .
(1. (1.4
CHAPTER 1. δk. . The ﬁrst assumption is the natural extension of the single-particle state function ψ(r). For example in the case in Eq. we have
∞
=
ν σ=↑. δ(r) = 0. FIRST AND SECOND QUANTIZATION
We shall often use the completeness relation Eq. (1. rN ). which can be expressed as ψ(r) =
ν
ψν (r)
∗ dr ψν (r )ψ(r )
or
r|ψ =
ν
r|ν ν|ψ . .10)
Dirac’s delta-function δ(r) for continuous variables. r2 . . for r = 0. r2 . Ly . (1. for k = n. rN )| drj = surrounding the point (r1 . to the N -particle state function ψ(r1 .e. for x > 0.13) |ψ(r1 . (1. . i. . . (1. Kronecker’s delta-function δk.5) with Landau orbitals in a box with side lengths Lx . A simple example is the expansion of a state function in a given basis: ψ(r) = r|ψ = r|1|ψ = r| ( ν |ν ν|) |ψ = ν r|ν ν|ψ . 0.7).n for discrete variables. three more assumptions are added to the basic assumptions deﬁning quantum theory. .2
First quantization. which is a complex function in the 3N -dimensional conﬁguration space. 1. .n = 1.8)
It should be noted that the quantum label ν can contain both discrete and continuous quantum numbers. for x < 0. rN ) in j=1 the 3N −dimensional conﬁguration space. .

rN ). . rk .1. It furthermore plays a fundamental role in the studies of the nature of stars and of the scattering processes in high energy physics. rN ) = −ψ(r1 .2.. rk . . However. . . rk . MANY-PARTICLE SYSTEMS
5
1. . . . . . It thus explains the periodic table of the elements. rk . i..15a) (1.1
Permutation symmetry and indistinguishability
A fundamental diﬀerence between classical and quantum mechanics concerns the concept of indistinguishability of identical particles. rj .. .g. . For bosons the assumption is necessary to understand Planck’s radiation law for the electromagnetic ﬁeld.
1 This discrete permutation symmetry is always obeyed. . . superﬂuidity and laser light. rj . (1. . and it has been introduced due to overwhelming experimental evidence. and consequently the starting point in our understanding of atomic physics. . . rN ) (bosons). . . and worse. . ψ(r1 .. . rN ) (fermions).... a so-called Berry phase. The second fundamental assumption for N -particle systems is therefore that identical particles.. . if a number of identical particles are brought to the same region in space. FIRST QUANTIZATION. .. (1. rj . . a colored spot on a billiard ball) without inﬂuencing its behavior. and the corresponding state function can at most diﬀer from the original one by a simple prefactor λ.. . . . thereby soon render it impossible to say which particle is where. rj . . are in principle indistinguishable. and moreover it follows its own continuous path in phase space. In classical mechanics each particle can be equipped with an identifying marker (e. . . . some quasiparticles in 2D exhibit any phase eiφ .2. rk ... Such exotic beasts are called anyons
. .e. the so-called bosons and fermions1 : ψ(r1 . . rj . .. . (1. This is not so in quantum mechanics. . rk . .. From the indistinguishability of particles follows that if two coordinates in an N particle state function are interchanged the same physical state results. . . . . . and spectacular phenomena like Bose–Einstein condensation. rN ) = λ2 ψ(r1 . . . ψ(r1 .. For fermions it immediately leads to the Pauli exclusion principle stating that two fermions cannot occupy the same state. Not even in principle is it possible to mark a particle without inﬂuencing its physical state. If the same two coordinates then are interchanged a second time. condensed matter physics and chemistry.15b) we let rj = rk then ψ = 0 follows. . .14)
and we conclude that λ2 = 1 or λ = ±1. rj . . rj .. . we end with the exact same state function.15b)
The importance of the assumption of indistinguishability of particles in quantum physics cannot be exaggerated. . . upon adiabatic interchange. rN ) = λψ(r1 . because if in Eq. . particles characterized by the same quantum numbers such as mass. . . their wavefunctions will rapidly spread out and overlap with one another. . .. charge and spin. Thus in principle each particle in a group of identical particles can be identiﬁed. . . .. rk . Only two species of particles are thus possible in quantum physics. rN ) = +ψ(r1 .

. FIRST AND SECOND QUANTIZATION
1. .15a) and (1.19)
Like before. .15b) are in Eq.
∗ ψν (r )ψν (r) = δ(r − r ).ν2 .16)
Starting from an arbitrary N -particle state ψ(r1 .
(1.18) leads to ψ(˜1 . rN ) by projecting onto the basis state ψν1 (r1 ): Aν1 (r2 . . . we know from the discussion on indistinguishability that physically it is not a useful basis since the coordinates have to appear in a symmetric way.ν2 . . The symmetry requirements from Eqs. ψ(˜1 . (1.νN
Aν1 .νN . . which therefore ought to be explicitly incorporated in the basis states.. . . rN ): Aν1 . rN ) we form the (N − 1)-particle function Aν1 (r2 . . .21)
where Aν1 . . . . . .. r
(1.2. rN ) = r r
ν1 . Even though the product states N ψνj (rj ) in a mathematical sense form a perfectly j=1 valid basis for the N -particle Hilbert space. . . (1.ν2 .. .. . which upon insertion into Eq.νN is just a complex number. we can invert this expression to give Aν1 in terms of Aν1 . .2
The single-particle states as basis states
We now show that the basis states for the N -particle system can be built from any complete orthonormal single-particle basis {ψν (r)}. ˜2 . . rN ) from Aν1 (r2 .ν2 (r3 . rN ) ≡
∗ dr1 ψν1 (r1 )ψ(r1 . (1. r3 . . . . . rN ). No physical perturbation can ever break the fundamental fermion or boson symmetry. . . . Thus any N -particle state function can be written as a (rather complicated) linear superposition of product states containing N factors of single-particle basis states.
(1. r2 .17)
r This can be inverted by multiplying with ψν1 (˜1 ) and summing over ν1 . . A physical meaningful basis bringing the N coordinates on equal footing in the products ψν1 (r1 )ψν2 (r2 ) . . ψνN (rN ). . . .ν2 (r3 . rN ).. rN ). . .18)
Now deﬁne analogously Aν1 . r2 . ..ν . ..20)
Continuing all the way through ˜N (and then writing r instead of ˜) we end up with r r ψ(r1 . ... . . . . rN ) = r
ν1
ψν1 (˜1 )Aν1 (r2 ..
(1. rN ).
(1. .21) hidden in the coefﬁcients Aν1 . .ν2
ψν1 (˜1 )ψν2 (˜2 )Aν1 . ... . .ν2 (r3 . . . rN ) ≡
∗ dr2 ψν2 (r2 )Aν1 (r2 .. .νN ψν1 (r1 )ψν2 (r2 ) .6
CHAPTER 1. .. r r
(1. ν ∗ dr ψν (r)ψν (r) = δν.. .. rN ) =
ν1 . . . ψνN (rN ) of single-particle state functions is obtained by
.

. ψνN (rN ) + ψν1 (r1 ) ψν2 (r1 ) . . . the Pauli principle. . . .21) gets replaced by the following.
±
. . . . .2. . .e.23)
while the boson case involves a sign-less determinant.and
2
Note that to obtain a normalized state on the right hand side in Eq.
N
ψνj (rp(j) ) . It states that single..24)
Here SN is the group of the N ! permutations p on the set of N coordinates3 .. MANY-PARTICLE SYSTEMS
7
ˆ applying the bosonic symmetrization operator S+ or the fermionic anti-symmetrization ˆ− deﬁned by the following determinants and permanent:2 operator S
N
ˆ S±
j=1
ψνj (rj ) =
ψν1 (r1 ) ψν2 (r1 ) .
N
ψνj (rp(j) ) sign(p).. .. . . . 3 + 2 − 3 − 1 + 2 + 1 −
√1 . . ψν1 (r2 ) ψν2 (r2 ) . (1. is the sign of the permutation p.. p∈SN ψνN (r1 ) ψνN (r2 ) . which in physics are denoted Slater determinants.
ψν1 (r2 ) ψν2 (r2 ) . .. . ψν1 (rN ) . and sign(p). @ 3 A .
.22)
ψνN (r1 ) ψνN (r2 ) . . . . . .. @ 1 A . .22) a prefactor
Q
ν
1 √
must be inserted. FIRST QUANTIZATION. . ψνN (rN ). @ 2 A : .and few-particle operators deﬁned for single. .νN S± ψν1 (r1 )ψν2 (r2 ) . .1 0 signs 0 the1 0 with the 1 of permutations as subscripts.ν2 . . p∈SN ψνN (r1 ) ψνN (r2 ) . ψνN (rN )
where nν is the number of times the state |ν appears in the set {|ν1 . For fermions nν = 0.2.. .
ψν1 (rN ) ψν2 (rN ) . ψν1 (rN ) ... ψν2 (rN ) = . .νN
ˆ Bν1 . ψν1 (r2 ) ψν2 (r2 ) .1.e. ψ(r1 . .3
Operators in ﬁrst quantization
We now turn to the third assumption needed to complete the quantum theory of N particle systems. used in the Slater determinant. .. . .
(1. Using the symmetrized basis states the expansion in Eq. rN ) =
ν1 .
1.
(1.. . |ν2 . . .. i. where the new expansion coeﬃcients Bν1 . . ψν2 (rN ) = . |νN }. . 8N 1 0 1 9 1 1 2 2 3 3 < = S3 = @ 2 A .. 1 (and thus nν ! = 1) so here the prefactor reduces to 3 For0 = 1 we 0 3 have. . .. @ 3 A . . ψνN (rN ) − ψν1 (r1 ) ψν2 (r1 ) . .ν2 . r2 .νN are completely symmetric in their ν-indices..
j=1
(1. . . .. (1. The fermion case involves ordinary determinants. a so-called permanent. i.25)
We need not worry about the precise relation between the two sets of coeﬃcients A and B since we are not going to use it. 0 or 1 for fermions and between 0 and N for bosons. . . @ 1 A . N!
nν !
√1 N!
... . Note how in the fermion case νj = νk leads to ψ = 0.
j=1
(1.

k=j
(1. It is straight forward to extend j this result to the proper symmetrized basis states. rj )
(1.e.32)
.
N
Ttot =
j=1
Tj .
j 1 N
Here the Kronecker delta comes from νa |νnj = δνa .νn |ψνn (r1 ) . introduce the operator of the total interaction energy Vtot .νa νb
=
In the N -particle system we must again take the symmetric combination of the coordinates.
j.νa νb |ψνc (rj ) |ψνd (rk ) ψνa (rj )| ψνb (rk )| (1. such as the Coulomb e2 1 interaction V (rj − rk ) = 4π |r −r | between a pair of electrons.
(1.νb
Tνb νa |ψνb (rj ) ψνa (rj )|.29)
N
=
j=1 νa νb
Tνb νa δνa . We move on to discuss symmetric two-particle operators Vjk .27)
=
ψνa (rj ). .30)
∗ ∗ drj drk ψνc (rj )ψνd (rk )V (rj −rk )ψνa (rj )ψνb (rk ).26) (1. (1. |ψνb (rj ) . . i.8
CHAPTER 1. |ψνn (rN )
1 2 N
(1. .
N
Vtot =
j>k
Vjk
1 = 2
N
Vjk .and two-particle operators. the kinetic
~ energy operator − 2m 2j or an external potential V (rj ). FIRST AND SECOND QUANTIZATION
few-particle states remain unchanged when acting on N -particle states. takes the following form in the r |ν -representation for a single-particle system:
2
Tj where Tνb νa
=
νa .31)
where
Vνc νd . say e.
∗ drj ψνb (rj ) T (rj . . hence the proper kinetic energy operator in this case must be the total (symmetric) kinetic energy operator Ttot associated with all the coordinates.g. In this course we will only work with one.28)
and the action of Ttot on a simple product state is Ttot |ψνn (r1 ) |ψνn (r2 ) . |ψνn (rN ) .
In an N -particle system all N particle coordinates must appear in a symmetrical way. For a two-particle sys0
tem described by the coordinates rj and rk in the |ν -representation with basis states |ψνa (rj ) |ψνb (rk ) we have the usual deﬁnition of Vjk : Vjk =
νa νb νc νd
j
k
Vνc νd .νn . A given local one-particle operator Tj = T (rj . . Let us begin with one-particle operators deﬁned on single-particle states described by the coordinate rj . rj ). .

. Vtot acts as follows: Vtot |ψνn (r1 ) |ψνn (r2 ) . 1.3.
2
HHe =
−
2m
2 1
−
Ze2 1 4π 0 r1
2
+ −
2m
This Hamiltonian consists of four one-particle operators and one two-particle operator.2.3
Second quantization. The starting point of this formalism is the notion of indistinguishability of particles discussed in Sec. 1. BASIC CONCEPTS
¢ £
9
Figure 1. 1.νa νb δνa .2: The position vectors of the two electrons orbiting the helium nucleus and the single-particle probability density P (r1 ) = dr2 1 |ψν1 (r1 )ψν2 (r2 )+ψν2 (r1 )ψν1 (r2 )|2 for the 2 symmetric two-particle state based on the single-particle orbitals |ν1 = |(3. 1. . 2. |ψνn (rN )
1 2 N
=
1 2
N
Vνc νd . |ψνc (rj ) . . As we shall see.
1. |ψνd (rk ) .νn |ψνn (r1 ) .νn δνb .2. SECOND QUANTIZATION.1. |ψνn (rN ) . ↓) depicted in Fig. . ↑) and |ν2 = |(4.1(b). basic concepts
Many-particle physics is formulated in terms of the so-called second quantization representation also known by the more descriptive name occupation number representation. .
j=k
(1.
j=k νa νb νc νd
j k 1 N
A typical Hamiltonian for an N -particle system thus takes the form
N
H = Ttot + Vtot =
j=1
A speciﬁc example is the Hamiltonian for the helium atom. 1.1 combined with the observation in Sec.34)
2 2
−
Ze2 1 4π 0 r2
+
e2 1 .2 that determinants or permanent of single-particle states form a basis for the Hilbert space of N -particle states. see also Fig. 2. 0. 4π 0 |r1 − r2 |
(1. 0.2. . . . Compare with the single orbital |(4.35)
. 2. which in a simple form neglecting spin interactions can be thought of as two electrons with coordinates r = r1 and r = r2 orbiting around a nucleus with charge Z = +2 at r = 0. quantum theory can be formulated in terms of occupation numbers of these single-particle states.
¢
£
¦
¤
¥
¤ £
¦
¤ ¢ ¡ ¥ ¤
(1.33)
Tj +
1 2
N
Vjk . ↓) .

. . Note how by virtue of the direct sum.3. .
j
nνj = N. .
1. | j nνj = N }.25) that in each term only the occupied single-particle states |νnj play a role. . (1.2
The boson creation and annihilation operators
To connect ﬁrst and second quantization we ﬁrst treat bosons. .}. . . (1. nνj+1 . ν (1. . The only non-zero matrix elements of b† j are nνj+1|b† j |nνj . .39)
where B+ (nνj ) is a normalization constant to be determined. nνj+1 . |ν2 . nνj + 1.36) with the ﬁrst quantization basis Eq. It is clear from the form S± ψνn (r1 )ψνn (r2 ) . . the ordering being of paramount ˆ importance for fermions. 0. . . The adjoint of b† j is found by complex conjugation as nνj + 1|b† j |nνj ∗ = ν ν † † † † nνj |(bνj ) |nνj +1 . 1 (fermions) (1. = B+ (nνj ) | . . . nνj+1 . The space spanned by the occupation number basis is denoted the Fock space F.1 some of the fermionic and bosonic basis states in the occupation number representation are shown. . the question arises how to connect the occupation number basis Eq. Consequently. It can be deﬁned as F = F0 ⊕ F1 ⊕ F2 ⊕ . nνj . . = B− (nνj ) | . . . 1. which lowers the occupation number of state |νj by 1. . nν2 . . . . b† j | . (1. while for bosons it can be any non-negative number. . nνj − 1. Given the occupation number operator it is natural to introduce the creation operator b† j that raises the occupation ν number in the state |νj by 1. . . . . bνj | . . Naturally. . In Table. nνj−1 . . . ψνn (rN ) of the 1 2 N basis states in Eq. (bosons). .36)
It is therefore natural to deﬁne occupation number operators nνj which as eigenstates have ˆ the basis states |nνj . . FIRST AND SECOND QUANTIZATION
1. |ν3 ..1
The occupation number representation
The ﬁrst step in deﬁning the occupation number representation is to choose any ordered and complete single-particle basis {|ν1 . N −particle basis states : |nν1 . where FN = span{|nν1 . (1. and as eigenvalues have the number nνj of particles occupying the state νj . . This will be answered in the next section. .
(1. . nνj−1 . . nνj+1 . . nν2 . where for brevity we only explicitly write the occupation ν ν number for νj . . nνj . ˆ We shall show later that for fermions nνj can be 0 or 1. 2. nνj−1 . states containing a diﬀerent number of particles are deﬁned to be orthogonal. 1. . one deﬁnes the annihilation operator bνj ≡ (bνj ) .40)
. .38) nνj = 0.10
CHAPTER 1. This simpliﬁcation is achieved with the occupation number representation. .23).3. The basis states for an N -particle system in the occupation number representation are obtained simply by listing the occupation numbers of each basis state. nνj−1 . nν3 .37) nνj |nνj = nνj |nνj . It must somehow be simpler to formulate a representation where one just counts how many particles there are in each orbital |ν . (1.

. .. . 0. bνk ] = 0. 1. . B] for two operators A and B is deﬁned as [A. 0. i. . However. . 1. . 0. 0.1. b† k ] = δνj .. 0.
11
The creation and annihilation operators b† j and bνj are the fundamental operators in the ν occupation number formalism.e. 1. . . . |0. In summary. |1. . . 0. . . . . 0. ν ν ν
we have [bνj . . 1.42)
By deﬁnition b† and bν are not Hermitian. . . . 0. . . B+ (0) = 1. .3. ν (1. . 0. 2. = | .. BASIC CONCEPTS Table 1. 1. . . . 0. The commutator [A. . . |1. 1. . 0. and hence by Hermitian conjugation that also bνj and bνk commute. |0.. 0. |1. .e. = 0. . . nν2 . 0. . .41)
We demand further that if j = k then bνj and b† k commute. . .. . . . 0. nν3 .. . . 0. . . . nν2 . i. valid for the state |0 . 1. 0. 0. 1. 0. |0. 1. . Since bosons ν are symmetric in the single-particle state index νj we of course demand that b† j and b† k ν ν must commute. . . B] = 0 ⇒ BA = AB. 0. 0. 0. .1: Some occupation number basis states for N -particle systems. . . . .. . . .. 0. . . . 0. But since ν † ∗ = 0|b |1 .. 0. 1. boson basis states |nν1 .42) we show below that this operator in fact is the
. .. . |1. 0. . It is evident that since an unoccupied state can not be emptied further we must demand bνj | . . B (1) = 1. . However. 0. . |1. it also follows that b | . b† k ] = 0. and by ν ν using the operator algebra Eq. . . |2. |0... . nν3 . . i.. 0.. ν ν [bνj . 1. .e. 1. . 1|bνj |0 − νj νj It is clear that bνj and b† j do not commute: bνj b† j |0 = |0 while b† j bνj |0 = 0. . b† j ] |0 = |0 . . . . . . 1. .. |0. 0. (1. |0.. 0. Let us proceed by investigating the properties of b† j and bνj further. 0. . 1. . 0. = | . .e. . 0. . . B] ≡ AB − BA. 1. 0. N 0 1 2 . 0. [bνj ... We assume this commutation relation. .. . . . 0. if j = k we must ν be careful. . 0. ... 0. SECOND QUANTIZATION. i. . 0. |0. |0. . 0. we deﬁne the operator algebra for the bosonic creation and annihilation operators by the following three commutation relations: [b† j . . We also have the freedom to normalize the operators by demanding b† j | . |0. . . 0. 0.. . the product b† bν is. |0. so that [A. 1. . fermion basis states |nν1 . . . . . . . 0. |1.νk ... . (1. As we will demonstrate later any operator can be expressed in terms of them. 2. |1. 1. 1. and we calculate the consequences of this assumption. . . 0. ν also to be valid as an operator identity in general. N 0 1 2 . B− (0) = 0. . 1. |0. . |0. 0.

ν If λ0 is not a non-negative integer this lowering process can continue until a negative eigenvalue is encountered. . nν = 0.44)
i. 2. b† bν |nν = nν |nν . The normalization factors for b† and bν are found from ν ν bν |nν b† |nν ν Hence we arrive at b† bν = nν . Writing |φλ = |nν we have shown that b† bν |nν = nν |nν and ν bν |nν ∝ |nν − 1 . . ˆ ν bν |nν = √ nν |nν − 1 . . (1. ν ν ν (1.e. we ﬁnd that (b† bν )b† |nν = (n + 1)b† |nν . (b† )nν |0 = nν ! |nν . b† bν |φλ = λ|φλ with λ > 0.e.43)
Secondly. 1.3: The action of the bosonic creation operator b† and adjoint annihilation operator ν bν in the occupation number space. Note that b† can act indeﬁnitely. i. . but this violates the condition λ0 > 0. ˆ S+ |ψνn (˜1 ) |ψνn (˜2 ) . Firstly. .42) leads immediately to the following ˆ two very important commutation relations: [b† bν .. 2. Analogously. . bν |φλ0 is also an eigenstate of b† bν . and we conclude that λ = n = 0. Let |φλ be any eigenstate of b† bν . ν ν ν i.Figure 1. r r r ν ν ν
1 2 N 1 2 N
where both sides contain N -particle state-kets completely symmetric in the single-particle state index νnj . . ν ν We ﬁnd that (b† bν )bν |φλ0 = (bν b† − 1)bν |φλ0 = bν (b† bν − 1)|φλ0 = bν (λ0 − 1)|φλ0 . Now choose a particular λ0 and study bν |φλ0 . Eq. 1.46a) (1. (1. FIRST AND SECOND QUANTIZATION
(1. |ψνn (˜N ) = b† n b† n . b† ] = b† . while bν eventually ν hits |0 and annihilates it yielding 0. .48) ν ν
2 2
and we can therefore identify the ﬁrst and second quantized basis states. ν = (b† |nν ν )
†
12
CHAPTER 1. . .46b)
(b† |nν ν
)=
nν |bν b† |nν ν
= nν + 1. bν ] = −bν ν [b† bν . (1.45)
(1.49)
. occupation number operator nν . b† n |0 .e. ν ν ν (1. for any state |φ we note that φ|b† bν |φ is the norm of the state bν |φ and hence ν a positive real number (unless |φ = |0 for which bν |0 = 0).
(1.
§ ¦ ¢ ¥ £ ¢ ¤ £ ¢¨© © ¨ ¨© ¨© ¨© © ¨ © ¨ ¨© ¨© ¢ ¡
= (bν |nν )† (bν |nν ) = nν |b† bν |nν = nν . b† |nν ∝ |nν + 1 .47) ν √ √ b† |nν = nν + 1 |nν + 1 . but with the eigenvalue reduced by 1 to (λ0 − 1).

. .50) (1. respectively. nνj . . if j = k we again must ν be careful. . nνj+1 . . nνj +1. cν ] = −cν ν [c† cν . nνj . (1. The anti-commutator {A.23) it is in the fermionic case not suﬃcient just to list the occupation numbers of the states. But since ν 1|c† j |0 ∗ = 0|cνj |1 it follows that cνj | . so that ν ν ν ν ν ν
. We assume this anti-commutation relation to be valid as ν an operator identity and calculate the consequences.54) and (1. (1. 0. .56)
so that c† and cν steps the eigenvalues of c† cν up and down by one. . so that {A. . . . . (1. . . . . . . . nνj−1 . nνj−1 . C− (0) = 0. = C+ (nνj ) | . . . ν ν ν
i.43) we ﬁnd ˆ [c† cν .3. = | . now denoted c† j and cνj . nνj+1 . C− (1) = 1. We must therefore demand | . It is evident that since an unoccupied state can not be emptied further we must demand cνj | . . . nνj+1 . . . also the order of the occupied states has a meaning. 1. . nνj−1 .3
The fermion creation and annihilation operators
Also for fermions it is natural to introduce creation and annihilation operators. ν It is clear that cνj and c† j do not anti-commute: cνj c† j |0 = |0 while c† j cνj |0 = 0. SECOND QUANTIZATION. ν (1.e. . In summary. . . ν ν ν (1. ν (cνj )2 = 0. nνk = 1.55)
Now. In analogy with Eq. B} for two operators A and B is deﬁned as {A. i.e. . . . . (1. as for bosons we introduce the Hermitian operator c† cν . c† ] = c† . = C− (nνj ) | . . . . . 0. . (1. . c† j } |0 = |0 . ν ν {cνj . nνj = 1.e. (1. . . and hence by Hermitian ν ν conjugation that also cνj and cνk anti-commute. nνj+1 . . However. and by using the operator ν algebra Eq. we deﬁne the operator algebra for the fermionic creation and annihilation operators by the following three anticommutation relations: {c† j . being the Hermitian adjoint of each other: ν c† j | . . B} = 0 ⇒ BA = −AB.53)
For j = k we also demand that cνj and c† k anti-commute. . (1. nνj −1.54)
An immediate consequence of the anti-commutation relations Eq. = −| . . BASIC CONCEPTS
13
1. . .52)
and consequently we must have that c† j and c† k anti-commute.1. (1. . . .e. . i. . . .3. {cνj . . B} ≡ AB + BA. C+ (0) = 1. .55) we have (c† cν )2 = c† (cν c† )cν = c† (1 − c† cν )cν = c† cν . .54) is (c† j )2 = 0. . nνk = 1. .νk . . ν cνj | . . From ν ν Eqs.54) we show below that this operator in fact is the occupation number operator nν . c† k } = 0. . c† k } = δνj . nνj = 1. . . (1. . . .51)
But to maintain the fundamental fermionic antisymmetry upon exchange of orbitals apparent in Eq. . . . = | . = 0. 0. . . 1. . We also have the freedom to normalize the operators by demanding c† j | . . cνk } = 0. . i. . . . nνj−1 . . we have {cνj . .

We have p > 0 since otherwise both sides would yield zero. (1. and c† cν thus only has 0 and 1 as eigenvalues leading to a simple ν ν ν normalization for c† and cν .νn b† n . This rewriting of the ﬁrst quantized operators in Eqs. nν = 0.49) we obtain
N site nj
Ttot b† n . c† cν (c† cν − 1) = 0. . . on the left hand side.e. ν i. ˆ r S− |ψνn (˜1 ) |ψνn (˜2 ) .57) (1. |ψνn (˜N ) = c† n c† n .29) acting on a bosonic N -particle system. the state ν may appear. we have ν c† cν = nν . b† n |0 = ν ν
1 N
where on the right hand side of the equation the operator b† b stands on the site nj . (1.4
The general form for second quantization operators
In second quantization all operators can be expressed in terms of the fundamental creation and annihilation operators deﬁned in the previous two sections. . In summary. . FIRST AND SECOND QUANTIZATION
(1. . Originally. we would like to reinsert the operator b† nj at site nj on the right. ν
and we can readily identify the ﬁrst and second quantized basis states.58)
c† |0 = |1 . To ν make the kets on the two sides of the equation look alike. . Utilizing that Ttot and S+ commute and invoking the basis state identity Eq. (1.3. (1. For simplicity. ν (1. On the right hand ν
¥¦ ¦ ¥ ¨ ¢ ¤ £© §¦ ¥ ©¥§¦ ¨ ¢ ¡
Tνb νa
νa νb j=1
j
14
CHAPTER 1. r r ν ν ν
1 2 N 1 2 N
where both sides contain normalized N -particle state-kets completely anti-symmetric in the single-particle state index νnj in accordance with the Pauli exclusion principle.49) and (1.60)
. 1 c† |1 = 0. . Note that both c† and cν can act at most ν twice before annihilating a state completely. ˆ ν cν |0 = 0.
1.Figure 1. let us ﬁrst consider the single-particle operator Ttot from Eq. p times leading to a contribution (b† )p |0 . ν cν |1 = |0 . b† b . . b† n |0 .59)
δνa . .33) into their second quantized form is achieved by using the basis state identities Eqs. To do this we focus on the state ν ≡ νnj . c† n |0 .59) linking the two representations.29) and (1.4: The action of the fermionic creation operator c† and the adjoint annihilation ν operator cν in the occupation number space. say. c† cν |nν = nν |nν . . In this equation we then act with the bosonic symmetrization operator S+ on both sides. ν ν ν
1 N
(1.

νn yields p identical j contributions cancelling the factor 1/p in Eq. The incoming and outgoing arrows represent initial and ﬁnal states.42). . and we arrive at the simple result Ttot b† n . (1. It is straightforward to generalize this result to two-particle (or any-number-of-particle) operators acting on boson states. b† n |0 .61)
Now.1. which is extremely
£
¡
¦
¦
¨
¦
§
¦
£ © ¡
¨
¦
§
¦
¥ £ ¢ ¡
(1.1) when the necessary care is taken regarding the sign appearing from the anti-commutators in this case. 1. The dashed and wiggled lines represent the transition amplitudes for the one. (1. Note the order of the indices.b
Tνb νa b† b bνa b† n .64)
.3. ν ν ν
1 N
Since this result is valid for any basis state b† n1 . Operators in second quantization are thus composed of linear combinations of products of creation and annihilation operators weighted by the appropriate matrix elements of the operator calculated in ﬁrst quantization.and two-particle operators in second quantization: Ttot =
νi .62)
(1.νk νl a† i a† j aνl aνk . b† nN |0 . . ν ν p p νb ν (1. (1. This is then rewritten ν ν by use of Eqs.60). The sum over j together with δνa . BASIC CONCEPTS
£ ¢ ¡ £ ¤ ¡ £ ¤ ¡
15
Figure 1. respectively. ν 1 Vνi νj . the p operators b† can be redistributed to their original places as they appear on ν the left hand side of Eq.and two-particle processes contained in the operators. (1.48) as b† b (b† )p−1 |0 = b† b ν ν ν 1 † 1 † bν bν (b† )p−1 |0 = b b (b† )p |0 . side the corresponding contribution has changed into b† b (b† )p−1 |0 . ν ν 2νν
i j
Vtot =
νk νl
In Fig. and a similar reasoning can be made for the fermion case (see Exercise 1.and two-particle operators in second quantization. .νj
Tνi νj a† i aνj . SECOND QUANTIZATION.61). . If we let a† denote either a boson operator b† or a fermion operator c† we can state the general form for one. .63) (1. it is actually an operator ν ν † identity stating Ttot = ij Tνi νj bνi bνj .5: A graphical representation of the one. b† n |0 ν ν
1 N
=
a. .5 a graphical representation of these fundamental operator expressions is shown.47) and (1.

66)
In the case of single-particle systems we deﬁne quite naturally creation operators a† and ˜µ a† corresponding to the two basis sets. [˜µ1 . Firstly.
(1. ν
aµ = ˜
ν
˜ ψµ |ψν aν .. ψνn leading to
1 N
a† n a† n . FIRST AND SECOND QUANTIZATION
important in the case of two-particle fermion operators.
(1.νk =
νj νk νj
(1. . .65) ν ν Note how all the permutation symmetry properties are taken care of by the operator algebra of a† and aν . that the basis transformation preserves the bosonic or fermionic particle statistics. which guides us to the transformation rules for creation and ν annihilation operators (see also Fig. (1. . 1.
1. .7) we have the basic transformation law for single-particle states: ˜ |ψµ =
ν
˜ |ψν ψν |ψµ =
ν
˜ ψµ |ψν
∗
|ψν . The ﬁrst quantization matrix element can be read as a transition induced from the initial state |νk νl to the ﬁnal state |νi νj . .70)
. (1. while the ﬁnal state is created by ﬁrst creating state |νj and then state |νi : |0 = aνl aνk |νk νl . that it leaves the total number of particles unchanged. (1.
(1.16
CHAPTER 1. (1. a† aµ = ˜µ ˜
µ µ νj νk
˜ ˜ ψνj |ψµ ψµ |ψνk a† j aνk = ν
νj νk
ψνj |ψνk a† j aνk = ν
νj
a† j aνj .67) lead to two very desirable results. In this section we give the general transformation rules which are to be exploited throughout this course.5
Change of basis in second quantization
Diﬀerent quantum operators are most naturally expressed in diﬀerent representations making basis changes a central issue in quantum physics. . ν
(1.} be two diﬀerent complete and ordered single-particle basis sets. The matrix elements are all in the simple non-symmetrized form of ν Eq. |ψµ2 .68)
The transformation rules Eq. (1. and ﬁnd directly from Eq.67) follows from applying the ﬁrst quantization single-particle ˆ result Eq. a† n |0 = ˜µ ˜µ ˜µ
1 2 N
˜ ψµn |ψνn
νn1
1
1
∗ † aνn 1
.} and {|ψµ1 . From the completeness condition Eq. .µ2 .69) ˜ ˜ ψµ1 |ψνj ψνj |ψµ2 = δµ1 .3. .
νnN
˜ ψµn |ψνn
N
N
∗ † aνn N
|0 . .66) to the N -particle ﬁrst quantized basis states S± |ψνn .
=
and secondly.6): a† = ˜µ
ν
˜ ψµ |ψν
∗
a† . a† k ]± ν ˜ ˜ ψµ1 |ψνj ψνk |ψµ2 δνj . |ψν2 .66) that a† |0 = ˜µ ν ˜ ˜ |ψµ = ν ψµ |ψν ∗ a† |0 . ˜ ˜ Let {|ψν1 . (1.31). .. a† 2 ]± = a ˜µ
νj νk
˜ ˜ ψµ1 |ψνj ψµ2 |ψνk ∗ [aνj . |νi νj = a† i a† j |0 . In second quantization the initial state is annihilated by ﬁrst annihilating state |νk and then state |νl .67)
The general validity of Eq. (1.

(1.3.6
Quantum ﬁeld operators and their Fourier transforms
In particular one second quantization representation requires special attention.5 let the transformed basis set {|ψµ } be the continuous set of position kets {|r } and. Ψ† (r2 )} = δ(r1 − r2 ). as a kind of waves.
(1. Ψ† (r) is the sum of all possible ways to add a particle to the system at position r through any of the basis states ψν (r).e. i. (1. namely the real space representation leading to the deﬁnition of quantum ﬁeld operators.6: The transformation rules for annihilation operators aν and aµ upon change of ˜˜ ˜ } = {|˜ }. Ψ† (r2 )] = δ(r1 − r2 ). On the one hand they are deﬁned as ﬁelds.3.
In some sense the quantum ﬁeld operators express the essence of the wave/particle duality in quantum physics.67) ˜µ Ψ† (r) ≡
ν
r|ψν
∗
a† = ν
ν
∗ ψν (r) a† . second quantization operators have a form analogous to ﬁrst quantization matrix elements.63) one obtains T =
νi νj ∗ dr ψνi (r)Tr ψνj (r) a† i aνj ν ∗ ψνi (r)a† i Tr ν νi νj
=
dr
ψνj (r)aνj
So in the real space representation. BASIC CONCEPTS
17
Figure 1. suppressing the spin index. ν
Ψ(r) ≡
ν
∗ Note that Ψ† (r) and Ψ(r) are second quantization operators. Loosely speaking. using quantum ﬁeld operators. denote a† by Ψ† (r) we obtain from Eq. SECOND QUANTIZATION. (1.1.73)
.
¥ ¤ ¢ ¨ § ¥ ¢£ ¨ © § ¥ ¦ ¢ !
r|ψν aν =
ν
¤¥ ¢£ § ¥ ¢ £ © ¨ § ¥ ¦ ¢£ ¡
ψν (r) aν . while the coeﬃcients ψν (r) and ψν (r) are ordinary ﬁrst quantization wavefunctions. i. but on the other hand they exhibit the commutator properties associated with particles. From Eq. Since Ψ† (r) and Ψ(r) are second quantization operators deﬁned in every point in space they are called quantum ﬁeld operators.
[Ψ(r1 ). {Ψ(r1 ). 1.3. If we in ˜ Sec.71)
boson ﬁelds fermion ﬁelds. basis between {|ψν } = {|ν } and {|ψµ µ
1.
(1. The introduction of quantum ﬁeld operators makes it easy to write down operators in the real space representation.72a) (1.69) it is straight forward to calculate the following fundamental commutator and anti-commutator.71) to the second quantized single-particle operator Eq. By applying the deﬁnition Eq.72b)
=
dr Ψ† (r)Tr Ψ(r). (1.
(1.e.

FIRST AND SECOND QUANTIZATION
Finally. (1. a† ≡ √ −i / 2 2 where x is given in units of the harmonic oscillator length = /mω and p in units of the harmonic oscillator momentum / . k
k
1 Ψ(r) = √ V
eik·r ak . 1 a† = √ q V dr eiq·r Ψ† (r).1
The harmonic oscillator in second quantization
The one-dimensional harmonic oscillator in ﬁrst quantization is characterized by two conjugate variables appearing in the Hamiltonian: the position x and the momentum p. as we shall see during the course. when working with homogeneous systems it is often desirable to transform between the real space and the momentum representations.18
CHAPTER 1. By inspection it can be veriﬁed that the following operators do the job. +i a ≡ √ / 2 2 ⇒ (1.77) we obtain the Hamiltonian. one can think of a as being √ the (1/ 2-normalized) complex number formed by the real part x/ and the imaginary part p/( / ).71) the |ψν basis with the momentum basis |k yields 1 Ψ† (r) = √ V e−ik·r a† .75)
1. Substituting in Eq.
n!
with H|n = ω n +
1 |n . If we as a
.77) i 1 x p p ≡ √ (a† − a).76)
This can be rewritten in second quantization by identifying two operators a† and a satisfying the basic boson commutation relations Eq. speciﬁc operators
In this section we will use the general second quantization formalism to derive some expressions for speciﬁc second quantization operators that we are going to use repeatedly in this course. 1 p 1 x x ≡ √ (a† + a). 1 aq = √ V dr e−iq·r Ψ(r). x] = i . From Eq. H= 1 2 1 p + mω 2 x2 . while a† is found as the adjoint operator to a. and the eigenstates |n : H = ω a† a + 1 2 (a† )n and |n = √ |0 . This picture is particularly useful in the studies of the photon and phonon ﬁelds.42).e. (1. (1. (1.
1. Mnemotechnically.4. i.74)
The inverse expressions are obtained by multiplying by e±iq·r and integrating over r.
k
(1. 2
(1.78)
The excitation of the harmonic oscillator can thus be interpreted as ﬁlling the oscillator with bosonic quanta created by the operator a† . 2m 2 [p. (1. H. to perform a Fourier transformation.4
Second quantization.

The creation operator can also be used to generate the speciﬁc form of the eigenfunctions ψn (x) of the oscillator starting from the groundstate wavefunction ψ0 (x): (a† )n 1 ψn (x) = x|n = x| √ |0 = √
n!
n!
p x x| √ − i ~ √ 2 2
n
|0 = √
1
x
2n n!
−
d dx
n
ψ0 (x). measure of the amplitude of the oscillator in the state with n quanta. 2.λ (r. In most of the applications in this course we shall. c
(1. however.
z). Thus the width of the oscillator wavefunction scales roughly with the squareroot of the number of quanta in the oscillator. Normally.80) are seen to be uk.81)
. 1. ±2. as sketched in Fig. 2. λ = 1. t) = kx =
1 √ ei(k·r−ωk t) . use the squareroot of the expectation value of x2 = 2 (a† a† + a† a + aa† + aa)/2. The quantum nature of the radiation ﬁeld. . λ = 1. the transversality condition ·A = 0 is chosen. t) of the wave equation Eq. treat the electromagnetic ﬁeld classically. The normalized eigenmodes uk. 1.
(1.4.7: The probability density | r|n |2 for n = 0.2
The electromagnetic ﬁeld in second quantization
Historically. and the associated concept of photons play a crucial role in the theory of interactions between matter and light.80)
We assume periodic boundary conditions for A enclosed in a huge box taken to be a cube √ of volume V and hence side length L = 3 V. 2. (1. (same for y and L nx .4. we ﬁnd n|x2 |n = n + 1/2 . SECOND QUANTIZATION. The quantization of the electromagnetic ﬁeld is based on the observation that the eigenmodes of the classical ﬁeld can be thought of as a collection of harmonic oscillators. SPECIFIC OPERATORS
19
Figure 1. Note that the width of the wave function is n|x2 |n = n + 1/2 . These are then quantized.λ (r. 1927). ±1. in which case A is denoted the radiation ﬁeld. (1.7. . In the free ﬁeld case the electromagnetic ﬁeld is completely determined by the vector potential A(r. the electromagnetic ﬁeld was the ﬁrst example of second quantization (Dirac. . ωk = ck V λ 2π nx = 0. t) in a speciﬁc gauge. and 9 quanta in the oscillator state. and we have B = ×A
2
E = −∂t A
·A = 0 1 2 A − 2 ∂t A = 0. The dispersion law is ωk = kc and the two-fold polarization of the ﬁeld is described by polarization vectors λ .79)
1. |n .1.

λ 2 2|Ak.λ and Pk.20
CHAPTER 1.λ ˙ AI = −ωk AR .λ ≡ 2 0 AR k.84) and (1.λ k.85)
From Eqs.λ .λ
(1. k.λ k.λ (a† + ak.λ 2ωk 2 (1.λ |2 = 4 0 ωk
1 2
|AR |2 + |AI |2 .λ ] = i ωk Q = k. which is simply the ﬁeld energy known from electromagnetism.λ
(1. t) = √ V Ak. where each mode are characterized by the conjugate variable Qk.λ = i(a† − ak. Proper k.λ . i.λ k. using Parceval’s theorem and the notation Ak.λ ∂Pk.e. AR k.λ Pk.λ + ).80) we can express H in terms of the radiation ﬁeld A.λ (1. k. Pk.λ (t) = Ak.λ k.λ = −ωk Qk. Pk.λ .λ = AR + iAI for the k. we have H=
2 0 ωk k.λ 2 k.λ √ k. ∂Qk.λ e−iωk t .λ .λ ei(k·r−ωk t) + A∗ e−i(k·r−ωk t) k.83)
In Fourier space. H= 1 2 dr
2 0 |E|
+
1 |B|2 µ0
=
1 2
0
2 dr (ωk |A|2 + c2 k 2 |A|2 ) =
2 0 ωk
dr |A|2 .λ k. k. (1. The ﬁeld A takes only real values and hence it has a Fourier expansion of the form 1 A(r. FIRST AND SECOND QUANTIZATION
The set { 1 . k. [ak.λ ≡ 2ωk 0 Ak.λ . 2 .λ 2 ˙ ˙ ⇒ Q = Pk. and introducing the second quantized Bose operators a† for each k.λ ).87)
. Quantization is now obtained by imposing the usual condition on the commutator of the variables. Qk.λ are therefore introduced: 1 2 H= P 2 + ωk Q2 k.λ real and imaginary part of the coeﬃcients.λ quantized oscillator: 1 H= ωk (a† ak.λ ⇒ [Pk.82) we merge the time dependence with the coeﬃcients.λ k. Using Eq.λ √ I k.λ . (1. up to some normalization constants. the time dependence for the real and imaginary parts are seen to be ˙ AR = +ωk AI k.λ
k.λ
k λ=1.λ ). We now turn to the Hamiltonian H of the system.λ 2 k.λ k. (1. = Qk.λ are the complex expansion coeﬃcients. a† ] = 1.
(1.86)
This ends the proof that the radiation ﬁeld A can thought of as a collection of harmonic oscillator eigenmodes.λ Qk.λ normalized conjugate variables Qk.λ k.82)
where Ak.84)
If in Eq. Ak.λ
k.85) it thus follows that. k/k} forms a right-handed orthonormal basis set. (1.λ ∂H ∂H ˙ ˙ and AI are conjugate variables: ∂AR = −4 0 ωk AI and ∂AI = +4 0 ωk AR .λ k.λ and Pk.2 λ.λ ∂H ∂H ˙ ˙ = −Pk.

and current
In the following we establish the second quantization representation of the four important single-particle operators associated with kinetic energy.74). 2m r 2 k2 k σ |T |kσ = δk . 2m 2
Tr.89)
1. t) = √ V ak.λ ei(k·r−ωk t) + a† e−i(k·r−ωk t) k.
. (1.90b)
Its second quantized forms with spin indices follow directly from Eqs.λ Qk. SPECIFIC OPERATORS
21
To obtain the ﬁnal expression for A in second quantization we simply express Ak.
(1. (1.
(1.93)
In analytical mechanics A enters through the Lagrangian: L = 1 mv 2 − V + qv·A.σ = − k.88)
Substituting this into the expansion Eq. which in turn is expressed in terms of a† and ak. spin. First. we study the kinetic energy operator T . SECOND QUANTIZATION.λ 2 0 2ωk =
0
2 0 ωk
ak.63) and (1.λ in terms of Pk.
with τ =
0 1 1 0
. p) = p·v − L(r. 2 1 TA = dr Ψ† (r) − qA Ψσ (r). density.3
Operators for kinetic energy.
and A∗ → k.λ (1. we treat the spin operator s for electrons.λ Pk. For particles with charge q a magnetic ﬁeld can be included in the expression for the kinetic energy by substituting the canonical momentum p with the kinetic momentum4 p − qA.λ and Qk. But then p = ∂L/∂v = mv + qA.90a) (1.4. Considering inﬁnitesimal variations δA we get δH = H(A + δA) − H(A) = −qv·δA = −q dr J·δA.4.λ Ak.λ .λ .
0 −i i 0
. an expression used to ﬁnd J. k. and particle current density.σ 2 k2
2m
a† ak.σ .92) r σ 2m σ i Next.82) our ﬁnal result is: 1 A(r. momentum representation.1.σ .
(1. (1.λ k.
. (1. and via a Legendre 1 1 transform we get H(r.σ
2
2m
dr Ψ† (r) σ
σ
2 r Ψσ (r)
.91)
The second equality can also be proven directly by inserting Ψ† (r) and Ψ(r) from Eq. v) = 2 mv 2 + V = 2m (pR− qA)2 + V . which is independent of spin and hence diagonal in the spin indices.σ σ = −
real space representation.λ
2 0 ωk
a† . since this by the 2 Euler-Lagrange equations yields the Lorentz force.
k λ=1. the Pauli matrices s=
4
In ﬁrst quantization it is given by 1 0 0 −1
2
τ.73) T =
k.λ
λ.λ : k.k δσ .2
2 0 ωk
(1.λ = AR + iAI → √ + i √ k. In ﬁrst quantization it has the representations
2 δσ . spin. particle density.

FIRST AND SECOND QUANTIZATION
To obtain the second quantized operator we pull out the spin index explicitly in the basis kets. and we obtain H=T+
σ
dr
q A· 2mi
Ψ† (r) Ψσ (r) − Ψ† (r) σ σ
Ψσ (r)
+
q2 2 † A Ψσ (r)Ψσ (r) . (1.σ (r )δ(r − r)ψµ. Variations due to a varying parameter are calculated as derivatives if the parameter appears as a simple factor. we shall take a more general approach based on analytical mechanics. (1. σ σ (1.92) and writing only the A dependent terms of q2 q~ the integrand.σ (r) = dr ψµ. see Eq.95)
From Eq. (1. By analytical mechanics it is found that variations δH in the Hamiltonian function due to variations δA in the vector potential is given by δH = −q dr J·δA (1. But expanding the square in Eq.63) yields ρσ (r) = dr Ψ† (r )δ(r − r)Ψσ (r ) = Ψ† (r)Ψσ (r).75) the momentum representation of this is found to be ρσ (r) = 1 V
† ei(k−k )·r ak σ akσ = kk
1 V
e−iq·r a† k+qσ akσ =
kq
1 V
a† ak+qσ kσ
q k
eiq·r . It is related to the particle density operator ρ(r) through the continuity equation ∂t ρ + ·J = 0.92).92) and the associated footnote. s=
µσµ σ
µ | σ |s|σ |µ c† σ cµσ = µ
2
σ |(τ x . This relationship can be used to actually deﬁne J. and thus the density operator for spin σ is given by ρσ (r) = δ(r − r). τ z )|σ c† cµσ .σ (r)|2 which can also ∗ be written as ρµ. into account. τ y . However. In ﬁrst quantization the fundamental interpretation of the wave function ψµ.98)
. By partial integration this is shifted to Ψ† (r).
(1. 2m (1. (1. −Ψ† (r) 2mi [ ·A + A· ]Ψσ (r) + 2m A2 Ψ† (r)Ψσ (r).σ (r) gives us ρµ.94b) We then turn to the particle density operator ρ(r).22
CHAPTER 1. and obtain with fermion operators the following vector expression. reveals one term where σ σ is acting on A. given by the vector potential A. µ↓ µ↑
µ
(1. The fourth and last operator to be treated is the particle current density operator J(r).97)
We use this expression with H given by the kinetic energy Eq.96) where the momentum transfer q = k − k has been introduced. |ν = |µ |σ . µσ
µ σσ
(1. In second quantization this combined with Eq. (1.σ (r) = |ψµ. This allows us in a simple way to take the magnetic ﬁeld.σ (r ).94a)
with components sx = 2 (c† cµ↑ + c† cµ↓ ) sy = i µ↑ µ↓
µ
2
(c† cµ↑ − c† cµ↓ ) sz = µ↑ µ↓
µ
2
(c† cµ↑ − c† cµ↓ ).

kσ
kq
kq
(1.4. Using the same reasoning that led from Eq.64) with 1 |ν = |k.4. kσ 2 JA (r) = σ −q A(r) mV eiq·r a† ak+q. the exponential is rewritten as
. We can interpret the Coulomb matrix element as describing a transition from an initial state |k1 σ1 .99c)
diamagnetic : JA (r) = − σ
q A(r)Ψ† (r)Ψσ (r).2.σ . (1. σ and ψk. J and JA . SPECIFIC OPERATORS
23
The variations of Eq. σ paramagnetic : Jσ (r) = 2mi Ψ† (r) σ Ψσ (r) −
(1.63) to Eq.σ .31) and Eq. k4 σ2 without ﬂipping any spin. (1.
1.64) to the following quantum ﬁeld operator form of V : V (r2 − r1 ) = 1 2σ dr1 dr2 e2 0 Ψ† (r )Ψ† (r )Ψ (r )Ψ (r ). and we obtain V = 1 2σ 1 2σ k3 σ1 . k |r2 − r1 |
Since r2 − r1 is the relevant variable for the interaction.101)
1 σ2
Here we have introduced the abbreviation e2 = e2 /4π 0 . (1. (1.We can also write the Coulomb 0 interaction directly in the momentum basis by using Eq. k2 σ2 to a ﬁnal state |k3 σ1 . SECOND QUANTIZATION. (1. respectively:
Jσ (r) = Jσ (r) + JA (r). (1. k2 σ2 a† 3 σ1 a† 4 σ2 ak2 σ2 ak1 σ1 k k (1.97) can in Eq. σ m
The momentum representation of J is found in complete analogy with that of ρ Jσ (r) = mV 1 (k + q)eiq·r a† ak+q. |r2 − r1 | σ1 1 σ2 2 σ2 2 σ1 1 (1. (1.98) be performed as derivatives and J is immediately read oﬀ as the prefactor to δA.73) we can go directly from Eq.100) The expression for J in an arbitrary basis is treated in Exercise 1.σ (r) = √V eik·r χσ .1.99a) Ψ† (r) Ψσ (r) . The two terms in the current density operator are denoted the paramagnetic and the diamagnetic term.4
The Coulomb interaction in second quantization
The Coulomb interaction operator V is a two-particle operator not involving spin and thus diagonal in the spin indices of the particles.102)
1 σ2 k1 k2 k3 k4
=
1 σ2
k1 k2 k3 k4
e2 0 V2
dr1 dr2
ei(k1 ·r1 +k2 ·r2 −k3 ·r1 −k4 ·r2 ) † ak3 σ1 a† 4 σ2 ak2 σ2 ak1 σ1 . k4 σ2 |V |k1 σ1 . (1.99b) σ (1.

Let us for simplicity just study two diﬀerent kinds of particles.k1 +q . σ2 . Another example is electrons interacting with the photon ﬁeld. 1. The momenta k3 and k4 of the ﬁnal states can now be written as k3 = k1 + q and k4 = k2 − q. in Fig.24
CHAPTER 1. Likewise we have H2 and {|µ } for the second kind of particles. In this course one important example of such a system is the fermionic electrons in a metal interacting with the bosonic lattice vibrations (phonons). which with the deﬁnitions q ≡ k2 − k4 and r ≡ r2 − r1 become dr1 ei(k1 −k3 +q)·r1 = V δk3 . We study this system in Chap.5 how to calculate the Fourier transform Vq s of the Yukawa potential V ks (r) = k . 3. we just show a graphical representation of the operator. (1. σ1 and |k2 − q. ei[(k1 −k3 )·r1 +(k2 −k4 )·r2 ] = ei(k1 −k3 +k2 −k4 )·r1 ei(k2 −k4 )·(r2 −r1 ) leaving us with two integrals.103)
These integrals express the Fourier transform of the Coulomb interaction5 and the explicit momentum conservation obeyed by the interaction. FIRST AND SECOND QUANTIZATION
Figure 1. Under momentum and spin conservation the incoming states |k1 . The arguments are easily generalized to include more complicated systems. Here we will brieﬂy clarify how to construct the basis set for such composed systems in general. σ1 and |k2 . Here.104)
σ1 σ2 k1 k2 q
We shall study this operator thoroughly in Sec. r q (1.8: A graphical representation of the Coulomb interaction in second quantization. k k (1. The ﬁnal second quantized form of the Coulomb interaction in momentum space is V = 1 2V Vq a† 1 +qσ1 a† 2 −qσ2 ak2 σ2 ak1 σ1 . The starting point is the case where the two kinds of particles do not interact with each other.8. Vq ≡ dr e2 iq·r 4πe2 0 e = 20 . But so far we have not treated systems where diﬀerent kinds of particles are coupled.103) follows by setting ks = 0. The result is Vq s = 4πe0 2 q 2 +ks
2
e
−ks r
from which Eq. Let the ﬁrst kind of particles be described by the Hamiltonian H1 and a complete set of basis states {|ν }.
¥ © £¤ © ¡ § ¥ © £¤ © ¡
!
¥ ¢ £¤ ¨¦ ¢ ¡ § ¥ ¢ £¤ ¢ ¡
.2 in connection with the interacting electron gas.5
Basis states for systems with diﬀerent kinds of particles
In the previous sections we have derived diﬀerent fermion and boson operators. σ2 are with probability amplitude Vq scattered into the outgoing states |k1 + q. For the two decoupled
5 e0 r
2
k We show in Exercise 1.
1. 2.4.

SECOND QUANTIZATION AND STATISTICAL MECHANICS systems an example of separate occupation number basis sets is |ψ (1) |ψ
(2)
25
= |nν1 . We denote the density of states of the total system at a given total energy d(ET ). . where N (E) is the number of states with an energy less than E.81). which is assumed to be smaller than typical time scales of the observation. nµ2 .5
Second quantization and statistical mechanics
The basic assumption of statistical mechanics is the ergodicity assumption. nν2 . .5. . nk2 σ2 . . A general state |Φ can of course be any superposition of the basis states: |Φ =
{νj }{µl }
C{νj }. . . nµj . while the small system and the reservoir have the densities of states ds (Es ) and dr (Er ). . . . . . We let nkσ and Nqλ denote the occupation numbers for electrons and photons.{µl } |nν1 . Nq1 λ1 . . . . nνj . . = |nν1 . . .
(1. e.
(1. nν2 . nµl . .3). . because of the randomness of the system all of the available phase space is covered.105a) (1. .1. . . . .108)
1. = |nµ1 . A basis state |ψ in this representation has the form: |ψ = |nk1 σ1 . . . . . . . .g. . . . nνj . . . . Nql λl . . The time it takes for the system to visit all of the phase is the ergodicity time. |nµ1 . The electronic basis states are the plane wave orbitals |kσ of Eq. . . . nµ2 . . they constitute a closed system with total energy ET . In other words. . Suppose we are interested in some small system connected to the outside world. . It states that as time evolves a system assume all possible states complying with the given external constraints. nµ1 . . . .107)
As a concrete example we can write down the basis states for interacting electrons and photons in the momentum representation. . . nµ1 . and assume that. nkj σj . . ET = Es + Er . .106)
In the last line all the occupation numbers are simply listed within the same ket but the two groups are separated by a semicolon. . . the so-called reservoir. (1. . .
. (1. . nµl . respectively. . nν2 .105b)
When a coupling H12 between the two system is introduced. Based on the ergodicity assumption it is natural to conjecture that the probability for a subsystem to have a deﬁnite energy Es is proportional to the number of ways that the subsystem can have that energy. . nνj . The natural deﬁnition of basis states is the outer product states written as |ψ = |ψ (1) |ψ (2) = |nν1 . . nµ2 . . Nq2 λ2 . . . Let us call the energy of the small system Es and that of the reservoir Er . . . . . . .e. (1. taken as a whole. . . The density of states is deﬁned as d(E) = dN (E)/dE. with a given total energy E. . and the photon states are |qλ given in Eq. nµ2 . (1. i. . . nν2 . . . we need to enlarge the Hilbert space. nµl . nνj .

(1.113) P (Es ) = exp(−βEs ). (1. Therefore the only meaningful interpretation of Eq. FIRST AND SECOND QUANTIZATION
respectively. From the assumption about the probability being proportional to the number of states. This means that if we consider the ratio of two probabilities P (Es ) dr (ET − Es ) = . we have for the probability for the subsystem to have energy Es that P (Es ) ∝ dr (ET − Es ) ∆E. the total density of states is dominated by that of the reservoir and hence d(ET ) ≈ dr (ET ). P (Es ) dr (ET − Es ) (1.114) is that the sum of states runs over eigenstates of the Hamiltonian. (1. For a quantum system with many particles. In conclusion: from statistical mechanics we know that both for classical and a quantum mechanical systems which are connected to a heat bath the probability for a given state s with energy Es to be occupied is given by the Boltzmann distribution 1 (1.111)
When we sum over states. Z where β is the inverse temperature.e. if we make it smaller by cutting it in half by some wall. Since for a given small energy interval ∆E the number of states in the reservoir is much larger than the number states in smaller subsystem.26
CHAPTER 1. P (Es ) dr (ET − Es ) is P (E) ∝ e−βE . (1.112) We have thus arrived at the famous Boltzmann or Gibbs distribution which of course should be normalized. the states s are. unless of course it is an eigenstate of the Hamiltonian. as we have seen. and it is therefore an advantage to have a form which is independent of the choice of basis states. Also for a quantum system it is not clear what is meant by the energy of a given state. But because the energy is only deﬁned up to an additive constant. The only function P (E) that satisﬁes the condition dr (ET − Es ) P (Es ) = = f (Es − Es ). provided of course that it is still much smaller than the new reservoir. nothing should happen to the state of the small system. is the partition function exp(−βEs ). the basis set that we use to compute the energy must be a complete set. we do not expect this probability to be dependent on the size of the reservoir. i. and where the normalization factor. (1. β = 1/kB T . Using the basis states |ν deﬁned by H|ν = Eν |ν . Z.e. we must sum over a set of states which cover the entire space of possible states.114) Z=
s
(1. it can thus only depend on the diﬀerence Es − Es .110)
it must only depend on the energies Es and Es and neither on the total energy ET nor on dr .115)
. i. in general quite complicated to write down.109)
Now.

the symbol H is often used instead of HG for the grand canonical Hamiltonian. the thermal average of any quantum operator A is easily expressed using the density matrix ρ.
(1.e.120)
Unfortunately.114) for the partition function as Z=
ν
ν|ρ|ν = Tr[ρ].116)
We can thus write the expression Eq.117)
Likewise.
(1.119) where the trace now includes states with any number of particles. This ought not cause any problems. Likewise. ρ ≡ e−βH =
ν
|ν e−βEν ν|. They are more than merely normalization factors. Here we have used that the trace is invariant under cyclic permutation. For the grand canonical ensemble the number of particles is not conserved. Z Tr[ρ]
(1. SECOND QUANTIZATION AND STATISTICAL MECHANICS
27
it is now quite natural to introduce the so-called density matrix operator ρ corresponding to the classical Boltzmann factor e−βE . (1.e. we include the term µN in the Hamiltonian. Following the elementary deﬁnition we have A = 1 Z ν|A|ν e−βEν =
ν
1 Tr[ρA] Tr[ρA] = . The partition functions are fundamental quantities in statistical mechanics. The small system is allowed to exchange particles with the reservoir while keeping its average particle number constant. i. and we introduce a chemical potential µ of the reservoir to accommodate this constraint. (1. For example the free energy F ≡ U −
Remember that if tν = Tr [A] is the trace of A in the basis |ν .117) and (1. i. The corresponding density matrix ρG and partition function ZG are deﬁned as: ρG ≡ e−β(H−µN ) . the result obtained from the canonical ensemble is carried over to the grand canonical ensemble by the substitution H → H − µN . which in second quantization translates to the Fock space of the corresponding quantum numbers.
6
. since the sum over states is identiﬁed with the trace operation. so the reader must always carefully check whether H refers to the canonical or to the grand canonical ensemble. Basically. (1. In this book we shall for brevity write H in both cases. it is useful to introduce the Hamiltonian HG corresponding to the grand canonical ensemble. HG ≡ H − µN. ZG = Tr[ρG ].6 This is of course true whatever formalism we use to evaluate the trace.5. then in the transformed basis U |ν we have tU ν = Tr UAU−1 = Tr AU−1 U = Tr [A] = tν . Tr [ABC] = Tr [BCA]. For the canonical ensemble the trace is however restricted to run over states with a given number of particles.1. where N is the particle number operator. (1. In ﬁrst quantization the trace runs over for example the determinant basis. since most of the times we are working in the grand canonical ensemble.118)
Eqs.118) are basis-independent expressions.

important in the canonical ensemble. for H
0
=
Tr[ρ0 H] . (1. −β(εk −µ) 1+e e k +1
(1.124)
This inequality ensures that by minimizing the free energy calculated from the approximate Hamiltonian. 2.125) we see that the natural single-particle energy variable is not εk but rather ξk given by ξk ≡ εk − µ (1. we are guaranteed to make the best possible approximation based on the trial Hamiltonian. which is minimal when the entropy is maximal. This is reﬂected in the single-particle energy variable. Tr[ρ0 ]
(1.121b)
Let us now study the free energy. are directly related to Z and ZG .125)
We shall study the properties of the Fermi–Dirac distribution in Sec. 4. for example the mean ﬁeld approximation in Chap.
.3.121a) . This is based on the following inequality F ≤ H 0 − T S0 . respectively: Z = e−βF ZG = e
−βΩ
(1. Recall that F = U − T S = H − T S. From Eq. FIRST AND SECOND QUANTIZATION
T S. The proper Hamiltonian is therefore HG = H − µN .122) In various approximation schemes. The average occupation nF (εk ) is therefore Tr[ρG nk ] = Tr[ρG ] nk e−β(nk εk −µnk )
nk =0. important in the grand canonical ensemble. we shall use the principle of minimizing the free energy.5. Using the grand canonical ensemble we can derive the famous Fermi–Dirac distribution nF (ε).1
The distribution function for non-interacting fermions
As the temperature is raised from zero in a system of non-interacting fermions the occupation number for the individual energy eigenstates begins to ﬂuctuate rather than being constantly 0 or constantly 1.123) where both H example
0
and S0 are calculated in the approximation ρ ≈ ρ0 = exp(−βH0 ).1
=
0 + e−β(εk −µ) 1 = β(ε −µ) . (1. H0 . Note that the Fermi–Dirac distribution is deﬁned in the grand canonical ensemble.126) For small excitation energies εk varies around µ whereas ξk varies around 0. (1. (1. The state can contain either 0 or 1 electron.28
CHAPTER 1.1.
1. and the thermodynamic potential Ω ≡ U − T S − µN .1
nF (εk ) =
e−β(nk εk −µnk )
nk =0. Consider the electron state |kσ with energy εk .

Consider a bosonic state characterized by its fundamental energy εk . . so for pedagogical reasons we have chosen to start from the usual ﬁrst quantization representation. In Sec. Again using the grand canonical ensemble we derive the equally famous Bose–Einstein distribution nB (ε). perhaps after some measure of acquaintance. 5 will the question be raised of how to treat time evolution in second quantization. the Fermi–Dirac distribution nF (ε). SUMMARY AND OUTLOOK
29
1.63) and (1. In the following three chapters we shall get more acquainted with second quantization through studies of simpliﬁed stationary problems for non-interacting systems or systems where a given particle only interacts with the mean ﬁeld of the other particles. (1. The intricate permutation symmetries are manifestly ensured by the basic (anti-)commutator relations of these fundamental operators. 2. for very small occupation numbers. (1.54)). 1. the representation of quantum mechanics we are going to use throughout this course.42)). With an answer to that question we can proceed with the very interesting but also rather diﬃcult studies of the full time dependent dynamics of many-particle quantum systems.5. (1. The occupation number of the state can be any non-negative integer nk = 0. 1. this main result appears so simple and intuitively clear that one could choose to deﬁne quantum theory directly in second quantization rather than going the cumbersome way from ﬁrst to second quantization.and two-particle operators..127)
The Bose–Einstein distribution diﬀers from the Fermi–Dirac distribution by having −1 in the denominator instead of +1. However.
.
(1.5. In the grand canonical ensemble the average occupation number nB (εk ) is found by writing λk = e−β(εk −µ) and d 1 using the formulas ∞ nλn = λ dλ ∞ λn and ∞ λn = 1−λ : n=0 n=0 n=0
∞
nk e−β(nk εk −µnk ) nB (εk ) =
nk =0 ∞
λk =
d dλk
∞
∞ nk =0 n λkk
λkk =
n
e
nk =0
−β(nk εk −µnk ) nk =0
λk (1−λk )2 1 1−λk
=
1 e
β(εk −µ)
−1
.6.1. In fact. The basic concepts are the occupation number basis states and the fundamental creation and annihilation operators. b† and bν in the bosonic case (see Eq. nk = e−β(εk −µ) . .4 we presented a number of speciﬁc examples of second quantization operators. Both distributions converge towards the classical Maxwell– Boltzmann distribution. First in Chap. . 1. and we got a ﬁrst glimpse of how second quantization leads to a formulation of quantum physics in terms of creation and annihilation of particles and ﬁeld quanta. students usually learn basic quantum theory in ﬁrst quantization. and c† and cν in the fermionic case (see ν ν Eq.2
Distribution functions for non-interacting bosons
Next we ﬁnd the distribution function for non-interacting bosons. It is derived like its fermionic counterpart. The main result of the chapter is the derivation of the general form of one. where the particular particle statistics is not felt very strongly. Eqs.64) and Fig.6
Summary and outlook
In this chapter we have introduced second quantization.
1.

FIRST AND SECOND QUANTIZATION
.30
CHAPTER 1.

the light cloud of the many negatively charged core electrons tightly bound to the nucleus. the outermost few valence electrons. In contrast to the core electrons the Z valence electrons.Figure 2. This is
The electron gas
Chapter 2
free atoms
a solid
31
valence electrons (mass m. The main emphasis will be on the non-interacting electron gas. and if the atom has Z valence electrons the charge of the ion is +Ze. and insulators. semiconductors. such as metals. The valence electrons are freed from their host atoms and form an electron gas holding the ionic lattice together. charge +Ze) core electrons nuclei
. are often free to move away from their respective host atoms forming a gas of electrons swirling around among the ions. Any atom in a metal consists of three parts: the positively charged heavy nucleus at the center. The nucleus with its core electrons is denoted an ion. and ﬁnally. The ions are unchanged during the process where they end up by forming a periodic lattice.
s s r¡r¡ ¡s¡ ss¡sr¡ r¡rrs¡s rs¡¡rssr rss¡¡srrs ¡srs¡ rr¡sr¡ rrss¡rs¡ ¡¡srsr ¡sr¡ rs¡¡srrs ¡rrs¡ rs¡¡srrs srs¡s¡ r¡rsr¡ r¡r¡ ¡¡rrssrrs s¡s¡ rs¡rs¡ tuBtu u pq tB Btu vwBvw w vB Bvw no ¡¡ " "!¡ ¡ " ¡¡¡!¡¡" ¡ !¡ ¡ !!""!"¡"!"¡"!"¡34¡34¡¡34¡¡¡¡¡¡¡ ¡¡¡"!¡¡¡¡¡¡¡¡¡¡ "¡¡¡¡¡ ¡ ¡¡¡¡¡34 ¡¡34 ¡ 343434 ¡343434 343434 !¡ ¡!!"""!!"! ¡ ¡ !"¡ ¡ "¡ ¡3434 ¡¡4 ¡¡¡¡¡¡ ¡!!"¡!!"¡3434¡3434¡¡3434¡¡¡¡¡¡¡ !¡¡¡¡3¡¡¡¡¡¡ ¡¡¡¡34¡¡¡¡¡¡ "¡ !"¡ ¡ "!!!""¡!""¡!!""¡34¡34¡¡a 3434¡34¡¡¡¡8¡8¡ "!!¡ @¡ ¡ " ¡¡¡¡ 34 ¡34 34 ¡¡¡!¡¡34 ¡¡4 ¡¡¡¡¡¡ 9¡¡¡¡¡¡¡3¡¡¡¡¡¡ ¡"!!¡B!39%c &¡@¡4¡b"¡3c!%&& "!"¡4c"!¡¡¡B¡¡¡XYB¡¡78¡78 !"!¡ & ¡ 9@9¡¡ !"¡ %&¡ #$$34 ¡ #$$34 ¡@9@¡9@ %&%& ¡34 %& !""! %&¡ a ¡ #$34 ¡ #$ XY ¡¡!""!"!""!¡ ¡¡¡#$$ %&%&%& `a ¡¡¡7¡7¡¡ "¡ ¡ ¡¡¡ !" ¡¡#$¡ ¡¡¡¡788¡788¡¡ @¡¡ & ¡ ¡ ¡ "!!""¡"!"!"¡B"¡¡¡B¡¡¡XYB¡¡7878¡7788 99@¡¡! 9@9@ ¡ !!""!" ¡ ##$¡$#$ ¡ ¡@99@¡¡ ¡ %&%&¡¡ ¡ ##$¡¡ #$#$ XY ¡¡c %&%&¡%¡¡¡a #$#$¡#¡¡¡¡7¡7¡¡ "¡ %¡ ¡ ¡¡bB"%&&¡¡¡`B#$$¡¡¡B788¡788¡¡ @¡¡ %%&¡ &¡ $# ¡ #$#¡ !"!"!!"¡!"!"!!"¡B!¡B%¡B¡B¡¡¡¡787788¡787788 !"¡ %¡ ¡ 9@¡@99@¡¡ %&%¡%&¡¡ %&%& #$#¡#$#¡ #$#$¡ #$#$ ¡¡¡%¡¡¡¡#¡e ¡¡¡7¡7¡¡ "!¡ ¡ ¡¡b %&&%&¡%&%¡gb!&¡¡¡`g&¡#$$#$¡#$#¡d`$¡B¡¡788¡788¡¡ !¡ & @¡¡@ ¡¡%&%¡ ¡¡¡ 9¡¡!" @9 & ¡!"!" ¡ $ ¡$ de 9@¡9¡!" 9@ ¡&¡ ¡%& ¡¡$¡ #$ @¡@9@¡%¡¡¡%&%¡#¡¡¡¡#$#¡7¡7¡8¡8 9¡¡12 @9@9 %&& 12 '¡¡'( ¡#$$ '( %%&& )¡¡)¡78¡78¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡7¡7 112212¡112212¡&%%&%&&¡%&¡B%'fg '(('(¡$##$#$$¡#$¡B)de )00)0¡0¡¡¡778878¡778878 1& 12¡ %&& $ '(('( $ $ @¡¡ %& 21¡12 ¡ #$ ¡#$ '( # 9¡ @9¡ ( %&¡%&¡%&¡ 0 #¡#¡0 #$¡)0 #$ )0 )0 ¡¡¡%&¡(¡¡¡#$¡)¡¡)0¡78¡78¡¡ 212 ¡¡¡&%¡fg ¡¡¡#¡de ¡¡)¡7¡7¡¡ 9@9¡¡ '(¡ %&&¡ 0 $#$ ¡ ¡9@99@¡211122 9@99@ 211122 '¡ &%& 21122 fg (''(( ¡(''(( %&%& )0¡ $#$ ''(( de 0 #$#$ 80))00 #$#$ 80))00 0))00 @¡¡ ( %&% ¡ % ((¡ 2¡ ¡ 1¡ ' @¡¡%¡¡'(¡¡#¡¡0 #$##$ ¡#$##$ ''(((''( 0¡87¡87¡8¡8 9¡¡¡%¡'¡¡¡#¡0))¡¡)0¡7¡7¡¡ ¡¡%&¡¡¡¡#$¡¡¡¡)¡¡¡7¡7 22112¡22112¡B2¡B%6 (''(((''(¡B(¡¡00))0¡pqB¡¡7878¡7878 1' & 12¡& 1 ' %¡$ ' & ''¡ ) $ ) $ ) ) @¡¡ 12¡ %&¡ (¡ 0 #$# pqpq ¡@¡@ ( %&% ¡% ¡('( %&% hihi '(¡ 0 #$#56 jk 8)¡8¡ 9¡ ¡9@9 212 '¡ 12112212 ¡'(¡'('¡ )¡0¡¡ ¡9@9@9¡¡12¡(¡¡'(''(('(¡'¡56 ¡56 '( ¡)0¡¡ ¡¡212¡21¡(¡¡¡ ¡0¡¡0¡0)¡0¡0 2¡ ¡ ¡¡s %&¡%& 1¡fs1¡'¡556 ''(¡hf&¡i #$¡#$6 '¡))0¡¡)0¡788)0¡788)0)¡)0) 122112¡122112¡B¡5r15l91rB26m@2s 1122112¡6¡s¡66¡¡B¡¡)00))0¡pqB¡¡78¡78 ¡¡¡¡¡¡hB5hi 6¡¡¡B7¡7¡¡ ¡¡¡¡¡¡¡65¡)0¡¡0¡0¡0¡0 ¡¡¡¡¡56¡0¡¡)¡)¡¡ 2¡ '(¡ ¡¡¡¡¡¡i 56¡¡0)¡¡)0¡)0¡)0)¡)0) 1¡ ¡ '( ¡ 12¡ ('(¡ 5¡¡ ¡¡¡¡¡¡6¡5¡)0)0¡¡)0¡)0¡0¡0 2 ('( 1¡6 ( 212 ¡ ('( 66¡656¡656 2¡ ¡ 6 '¡('( 1121221¡1121221¡B¡5r1565565665¡65565665¡¡565565665¡¡¡))0)00)¡¡¡))0)00)¡))0)00) ¡¡r 1122¡¡¡¡¡56¡)¡¡)¡)¡¡ 12¡( ¡11212 ( 565565¡¡ ¡''('¡5655656 ''('(¡''('(¡56 ''('¡¡5655656¡5655656 12¡ ¡''('( '(( ¦ ¦¥¥¦¦ ¦ ¦¥¥¦¦ ¦ ¥¡¡¥¡¡¥¡ ¦¡¡¦¡¡¦¡ ¡ ¥¡¡¥¡¡¥¡ ¡¡ ¦¡¡¦¡¡¦¡¥¦ ¥¦ ¥¦ ¡¡¡¡¡ ¡¡¡¡¡¥¦¥¦ ¡ ¢¡ ¢¡ ¢¡¡ ¡¡¡¡¡ ¥¦¡¡¥¦¡¡¥¦¡ ¢¢ ¢ ¢ ¢ ¢¢ ¡¡¡¡¡ ¥¦¦¡¡¥¦¦¡¡¥¦¡¦¥¦ ¡¡¡¡¡ ¢¢ ¡¡¡¡¡ ¥ ¥ ¥ ¥ ¥ ¡¡¡¡¡ ¡¡ ¡¡¡¡¡ ¥¦ ¥¦ ¡¡¡¡¡¥ ¡¡VW ¡¡¡ ¦¡¡FG ¦¡¡¥¡ ¡ ¢¡ ¢¢ ¡ ¢¡¡ ¡¡B¡¡ ¥¦¥¡¡B¡¡¥¦ ¡¡¡¡¡ ¢¢ ¡ VW ¡¦ ¡¦ FG ¦ ¢¢ ¢ ¢ ¢ ¢¢ ¡¡¡¡¡ ¥ ¥ ¥ ¥ ¦¦¥¦¥¦ ¦¥¡¥¦¦¥¦¥¦ FG ¦¦ ¦ ¢ ¢¢¢ ¢ ¢¡ ¡ VW ¡¡¦¡¡¡¥ ¡ ¢¡ ¡¢¡¡ ¥¡¥¡¡¥¥¡ ¡¢¡ ¢¡¢ AC ¢¢¢ ¡¡VB¡¡ ¡¡B¡¡¥¦¥¦ ¡¡B¡¡ ¢ ¢ ¡¡¡ ¦¡¡¥¦¡¡¦¡ ¡ ¡C ¡ ¢¡¡ ¡¡W ¡¡¡ ¥¦¡¡FG ¡¡¥¦¡ ¡¢¡AB ¡¡ ¡ ¡¡B¡¡ ¥¦¡¡¥¦¡¡¡¥¦ ¡¡¡¡¡ ¢ ¢ xy ¡¡¡¡¡ ¥ ¥ ¥ ¥ ¥ ¡ ¢¡¢ ¢¡ ¢ ¢ ¡¡ ¡¡¡¡¡ ¥¦ ¥¦ ¢¢ ¢¡ ¤¥¦ £¡¦ ¡ ¡ ¢¡ AC ¢ ¢¢ ¡¡¡ ¡¡¡¡¡¥ ¡ ¢¡ ¢¡¢¡¡ ¡¡¡¡¡ ¡¡¡¡¡¥¦ ¡¡¡¡¤ ¡¡ ¢¢ ¡¡¡¡ ¥¡¡¥¡¡¥¡ ¤£¡¡¤¤ ¡B ¢ ¡¡ ¡¡¡ ¦ ¥¦ ¦ ¥¦ ¦ ¡¤ £¤¤££¤ ¡ ¢¡A¤¤ C ¡¢¡¡ ¤ ¤ ¡¢ ¡ ¡¡¡ ¢ ¢ ¡¡¡¡¡ £¤£¡£¤£¡¡££¡££¡£¤ ¡¡ ¢ ¡¡¡¡¡ £¤¤£¡£¤¤£¡¡£¤¡£¤¡ ¢ ¢ ¡¡¡¤¡¤¡¤£¤£ ¡¡ ¢ ¡¡¡¡¡ ¡¡¡¡¡ ¡¡¡¡¡ ¡ ¤£¤£ £¤£¡£¤£¡DB£¡£¡ ¡¡B¤¤¡¤¤¡¤£¤£ ¡¡¡¡¡ ¡¡RSB¡¡ ¡ RS ¤¡¤¡E ¡¤£ DEDE ¤£ ¡¡¡¡¡ ££¤¡££¤¡¡£¡£¡ ¡¡¡¡¡££¤ ££¤¤££¤ ¡¡¡¡¡ ¡ ¤¡¤¡¡£¡£¡ ¡¡B¡¡ © ¡© ¡DE ¡£¡£¡ RS ¡¡RSB¡¡ ©¡¡©¡©¡¡ B¡¡££¤¤ ¡¡ © £¤¡© £¤¡¡£¤¡£¤¡ ¤¡¤¡¡¤ ¤ ¡¡¡¡¡ ©¡¡©¡©¡¡ £¤£¤£¤ £¤¤ £¤¤ ¡¡¡¡¡ ¡¡¡¡¡ ¡¡¡ ¡¡¡¡¡ ¡¡¡¡¡©©©© ¡¡¡£¤£¤£¤ £¤£¡£¤£¡ ¡¡¡¡¡ ¡¡¡¡¡ ¡£¡£¡ ¡¡¡¡¡ ¡¡¡¡¡ ©©¡¡©©¡©©¡¡ ¡¡¡¡¡ ©¡¡©¡©¡¡© ¡£¤¡£¤¡ ©© © ¡©© ¡ ©¡¡B©¡¡ ©©©© ¡ PQPQ ©©©© ©¡¡PQ ©¡©¡¡ ©¡ ©¡¡©¡©¡¡©© ¡¡ © ¡ © ¡¡¡ ¡¡Q ¡¡¡© ¨ ¡¡¡¡¡ ©¡¡PB©¡¡ ¨ §§¨¨§¨ §¡¨ ¨ ¡¡¡¡¡ ¡¡B¡¡© §¡¡¨¡§¡§¡ ¡¡¡¡¡ ¡¡¡¡¡ §¡¡¡§¡§¡ ¡¡B¡¡ ¡¡¡¡¡©©© ¡¡¡¡¡¨§§¨¨§¨ ¡¡¡ © ©©© © © ©©© §¨¡¡§¡§¨¡§¨¡ ¡¡ ¨¡¡¨§¨¡¨¡¨¡ ¡¡¡¡¡ ©¡¡©¡©¡¡ ¨¡¡§¨¡¨¡¨¡ ¡¡TU ¡¡¡ ©¡¡©¡©¡¡ §§¨¡¡§¡§§¨¡§§¨¡ ¡ TU ¨§§¨¨ §¨¡ HI §§¨¡¡HI ¨¡§§¨¡§§¨¡ ¡¡B¡¡¨§§¨ ¡¡¡¡¡ ¡¡¡¡¡ ¡¡TB¡¡ ¡ ¨¡¡¡¨¡¨¡ ¨B ¡¡¡¡¡ ¡ TU
¡¡¡¡¡ ¡¡U ¡¡¡ B §§¨¡¡I ¡§§¨¡§§¨¡ ¡¡HB¡¡¨§§¨ §¨¨§§¨ ¨§¡ HI
¡¡B¡¡ ¨¡¡¨§¡¨¡¨¡ ¡¡¡¡¡ ¡¡¡¡¡ §¨¡¡B§¨¡§¨¡§¨ ¡¡¡¡¡ ¨¡¡§¨¡¨¡¨¡ §¡¡§¡§¡§¡ ¡¡¡¡¡§§¨§¨¨§ §§¨§¨¨§ § §¨¡¡§¨¡§¨¡§¨¡
The study of the interacting electron gas moving in a charge compensating background of positively charged ions is central in this course. with mass m and charge −e. To a large extend the inner degrees of freedom of the ions do not play a signiﬁcant role leaving the center of mass coordinates Rj and total spin Sj of the ions as the only dynamical variables.1: A sketch showing N free atoms merging into a metal. since it will be clear that we need to develop our theoretical tools further to deal with the electron-electron interactions in full. Not only is this system a model of the solids that surrounds us. In this chapter we shall study the basic properties of this system using the formalism of time-independent second quantization as developed in Chap. charge -e) ions (mass M. 1. but historically this system played a major role as testing ground for the development of quantum ﬁeld theory. The ion mass is denoted M .

The formation of a metal from N independent atoms is sketched in Fig. hence no factor 2 . and from there calculate the electronic and phononic properties. where however the discrete nature of the ionic system is approximated by a positively charged. electron-hole excitations (single-particle excitations). is in fact an ab initio calculation.1)
2M
2 R
Ψion (R)
(2.2)
dR1 dR2 Ψ† (R1 )Ψ† (R2 ) ion ion dr Ψ† (r) − σ
2
Z 2 e2 0 Ψ (R2 )Ψion (R1 ). The individual terms are easily written down in second quantization: Tion + Vion−ion = dR Ψ† (R) − ion 1 + 2 Tel + Vel−el =
σ 2
(2. and the Coulomb interaction between the two systems. H = (Tion + Vion−ion ) + (Tel + Vel−el ) + Vel−ion . the equilibrium positions Rj of the ions in the lattice. Later in Sec. |R − r| ion
Note that no double counting is involved in Vel−ion since two diﬀerent types of ﬁelds. etc. Two approximation schemes will be used instead. plasmons (collective electronic charge density waves). THE ELECTRON GAS
true for the alkali metals. ion At zero temperature the ground state of the system is a periodic ion lattice hold together by the cohesive forces of the surrounding electron gas. i. lattice parameters and elasticity constants as input to the theory. |R1 − R2 | ion (2.
2. 2. magnons (spin waves). |r1 − r2 | σ2 (2.2 we attempt to include this interaction. In principle it is possible in ab initio calculations to minimize the energy of the system and ﬁnd the crystal structure and lattice parameters. 2. In this course we will not plunge into such full ﬂedged ab initio calculations.
. One is the phenomenological lattice approach. The other approximation scheme. We take the experimental determination of the crystal structure. Fortunately.4)
2m
2 r
Ψσ (r)
+ Vel−ion =
1 2σ
dr1 dr2 Ψ† 1 (r1 )Ψ† 2 (r2 ) σ σ
1 σ2
drdR Ψ† (r)Ψ† (R) σ ion
σ
(−Ze2 ) 0 Ψ (R)Ψσ (r). most electronic and phononic properties of the system can be derived with good accuracy from the Hamiltonian describing the ion jellium combined with the electron gas.1
The non-interacting electron gas
We ﬁrst study the lattice model and the jellium model in the case of no electron-electron interaction.1. continuous and homogeneous ﬂuid. the ion ’jellium’.e. the so-called jellium model.3) e2 0 Ψ (r2 )Ψσ1 (r1 ). Ψ† (r) σ 1 and Ψ† (R) are involved. From the obtained ground state one can then study the various excitations of the system: phonons (ion vibrations). The Hamiltonian H of the system is written as the sum of kinetic and potential energy of the ionic system and the electronic system treated independently.32
CHAPTER 2.

It is useful to introduce the reciprocal lattice.1
Bloch theory of electrons in a static ion lattice
Let us ﬁrst consider the phenomenological lattice model. a3 }: R = n1 a1 + n2 a2 + n3 a3 . 1. a3 · a1 × a2 (2. m2 . these vibrations can be described in terms of quantized harmonic oscillators (much like the photon ﬁeld of Sec. and the vibrating part. a1 · a2 × a3 b2 = 2π a3 × a1 . also known as the reciprocal space. The latter term must be there since a vibrating ion is giving rise to a vibrating electrical potential inﬂuencing the electrons.2 we study when this is reasonable) one arrives at the Hamiltonian HBloch used in Bloch’s theory of non-interacting electrons moving in a static. Let the static ion lattice be described by the ionic equilibrium positions R in terms of the lattice basis vectors {a1 . and their interaction with both the static part of the lattice. If one furthermore neglects the electron-electron interaction (in Sec. (2.9)
. Vel−latt (r + R) = Vel−latt (r) for any lattice vector R. THE NON-INTERACTING ELECTRON GAS
33
2. m3 ∈ Z.1) into H = (Elatt + Hph ) + (Tel + Vel−el ) + (Vel−latt + Vel−ph ). the phonons. both originating from a combination of Tion . i. b3 } in RL are deﬁned as b1 = 2π a2 × a3 . As will be demonstrated in Chap. X-ray experiments show that the equilibrium positions of the ions form a periodic lattice. (2.8)
where the basis vectors {b1 . (2. RL. (2. n3 ∈ Z.1.7)
Working with periodic lattices it is often convenient to Fourier transform from the direct space to k-space. the electrons are described by their kinetic energy Tel .1. Thus we can drop all the phonon related terms of the Hamiltonian.5)
At zero temperature the ions are not vibrating except for their quantum mechanical zero point motion. and later the phonon problem. b2 . and Vel−ion in the original Hamiltonian Eq.4. RS. The non-interacting part of the phonon ﬁeld is described by a Hamiltonian Hph . n2 .e. At ﬁnite temperature the ions can vibrate around their equilibrium positions with the total electric ﬁeld acting as the restoring force. This lattice has an energy Elatt and an electrical potential Vel−latt associated with it. (2. Finally. Vel−latt . Vion−ion . 3. in RS deﬁned by RL = G ∈ RS eiG ·R = 1 ⇒ G = m1 b1 + m2 b2 + m3 b3 . we have o to understand the Fourier transform of periodic functions. a2 · a3 × a1 b3 = 2π a1 × a2 . periodic ion lattice: HBloch = Tel + Vel−latt (r).1). their mutual interaction Vel−el . Vel−ph . (2. n1 . 2. Thus the Hamiltonian for the phenomenological lattice model changes H of Eq. a2 .2. m1 .2) giving rise to the concept of phonons.6)
To solve the corresponding Schr¨dinger equation.

1
¥ §¦ ¢ ¥ ¤ £¡
¨¦
FBZ = k ∈ RS |k| < |k − G|.13) for ck couples to an inﬁnite.14)
. deﬁned as all k in RS lying closer to G = 0 than to any other reciprocal lattice vector G = 0. Each such inﬁnite family of equations has exactly one representative k ∈ FBZ. ∃k ∈ FBZ. number of similar equations for ck−G .e. which separates in spatial part eik·r and a spin part χσ . for all R ⇔ V (r) =
G ∈RL
The solution of the Schr¨dinger equation HBloch ψ = Eψ can be found in the plane wave o basis |kσ .34
CHAPTER 2. each Schr¨dinger equation of the form Eq. ψnkσ (r) = unk (r)eik·r χσ . but o countable. THE ELECTRON GAS
An important concept is the ﬁrst Brillouin zone. n is the band index. (2. (2. The inﬁnite family of equations generated by a given k ∈ FBZ gives rise to a discrete spectrum of eigenenergies εnk . (2. i. unk (r + R) = unk (r).13)
1 V
cG
G
(nk)
eik·r χσ ≡ unk (r) eik·r χσ . TR ] = 0 ⇒ TR ψ = λR ψ for an eigenstate ψ.
(2. while any k outside FBZ does not give rise to a new set of equations.11)
εk δk. e. with TR f (r) ≡ f (r+R). ∃G ∈ RL : q = k + G.k +G ck .
An alternative derivation of Bloch’s theorem with emphasis on the group theoretic aspects builds on the translation operator TR .15) HBloch ψnkσ = εnk ψnkσ . We get [H.g.k +
G
VG δk. (2. The corresponding eigenfunctions ψnkσ are given by: ψnkσ (r) = 1 V ck+G ei(G +k)·r χσ =
G (n)
According to Eq. any wavevector q ∈ RS can be decomposed (the ﬁgure shows the FBZ for a 2D square lattice):
The Fourier transform of any function periodic in the lattice is as follows: V (r + R) = V (r). FBZ. and thus we end with Bloch’s theorem1 : k ∈ FBZ. Using vectors k ∈ FBZ. (2.11) the function unk (r) is periodic in the lattice. where n ∈ N.
(2.12)
VG ck−G = E ck .
(2. χ↑ = 1 : 0 ψσ (r) ≡ 1 V ck eik ·r χσ ⇒ kσ|HBloch |ψσ =
k k
so the Schr¨dinger equation for a given k is o ck εk +
G
We see that any given coeﬃcient ck only couples to other coeﬃcients of the form ck+G .10)
VG eiG ·r . for all G = 0 ⇓ ∀q. Applying TP after TR leads to λP λR = λP+R ⇒ λR = eik·R ⇒ ψnk (r) = unk (r)eik·r .

If we concentrate on the homogeneous part of the electron gas. i. present in a real lattice becomes the constant potential Vel−jel as sketched in Fig. Also the jellium model results in plane wave solutions.1) we have Vion−ion + Vel−el + Vel−ion = 0. to begin with. when no confusion is possible. so it is most natural to displace all the energy branches into the FBZ (the reduced zone scheme. Ly . In the jellium model the ion charges are imagined to be smeared out to form a homogeneous and. in H of Eq. (2. For many applications it turns out that the Bloch electrons described by ψnkσ (r) can be approximated by plane waves if at the same time the electronic mass m is changed into a material dependent eﬀective mass m∗ . Kittel or Ashcroft and Mermin. +Zρjel . the ion jellium. Vel−latt . (a) The parabolic energy band for free electrons. m∗ is often simply written as m. discard the part of Vel−el that leads to inhomogeneities. We shall use this so-called eﬀective mass approximation throughout this course:2 1 ik·r ψnkσ → √V e χσ The eﬀective mass (2. and Lz and volume V = Lx Ly Lz the single-particle basis states are the simple plane wave solutions to the free particle Schr¨dinger equation o
2
For a derivation of the eﬀective mass approximation see e. THE NON-INTERACTING ELECTRON GAS
35
Figure 2.17)
For a box with side lengths Lx . In other words.1. which are therefore of major interest to study. (c) All wavevectors are equivalent to those in the FBZ.3.2: Bloch’s theorem illustrated for a 1D lattice with lattice constant a. The periodic potential. In the following.
2.2.e.
¨ ©
¦ ¥§
¢ ¤ ¥¦
¨ ©
¢ £
¨ ©
¦¥
¢ ¡
. static positive charge density. (b) The Bloch bands viewed as a break-up of the parabolic free electron band in Brillouin zones (the extended zone scheme. we notice that this part together with the ion jellium forms a completely charge neutral system.2
Non-interacting electrons in the jellium model
In the eﬀective mass approximation of the lattice model the electron eigenstates are plane waves. and we simply end up with Hjel = Tel . The eigenfunctions are seen to be plane waves modulated by a periodic function unk (r) having the same periodicity as the lattice. k ∈ RS). k ∈ FBZ). (2.16) : m → m∗ approximation k unrestricted.g. 2.1.

and the imagined smeared out potential Vel−jel of the jellium model. ↓ . ↓ |k2 . . vF = F . present in a real lattice. We prefer the periodic boundary conditions rather than the Dirichlet boundary conditions ψ(0. y.
))))))))))))) )))))))))))) ))))))))) 23023023023023023023023023023023023023002302302302302302302302302302302302300230230230230230230230230230 1010101010101010101010101001010101010101010101010100101010101010101010 0000000000000123)0000000000000123)0000000000123)
¡ '( ¡ ¡
¡¡ ¡ # ¡¨ ! ¦¡¦ ¡¤ ¡¡ ¡ §¨¡ ¡ ¡ %& © ¡ ¨ !"¡ ¡ £¤¤ #$ ¡ ¡¡ ¡©© ¡§§¨ ¥¥¦¡¥¥¦ ¡££¤ © § £
0
L
0
L
¢ ¢ ¡ ¡¢ ¢¡ ¢¡ ¡¡ ¢
. z) (likewise for the y and z directions). The single-particle basis states are thus 2π kx = Lx nx (same for y and z) 2 k2 1 ik·r Hjel ψkσ = ψ .36
CHAPTER 2. THE ELECTRON GAS
with periodic boundary conditions ψ(L. kσ
(2. from which we obtain the following important rule of great practical value:
→
k
V (2π)3
dk . z) = 0 and ψ(L. since the former gives current carrying eigenstates well suited for the description of transport phenomena. y. . |k1 . λF = (2. . x 2m kσ V V = Lx Ly Lz . Associated with εF is the Fermi wavenumber kF .19)
3
2π 2π 2π Note how the quantization of k means that one state ﬁlls a volume Lx Ly Lz = (2π) in V k-space. ↑ . It is obtained by ﬁlling up the N states with the lowest possible energy. (2. and the Fermi velocity vF : k 2π 1 2mεF . c† 2 ↑ c† 2 ↓ c† 1 ↑ c† 1 ↓ |0 . z) = ψ (0. . . y. ψkσ (r) = √ e χσ . z) = ψ(0. |FS ≡ c† N/2 ↑ c† N/2 ↓ . . . .
(2. ±2. Vel−latt .18) n = 0. y. the Fermi wave length λF . y. |k1 . . εF . . ±1. ↑ . and with this basis we obtain Hjel in second quantization: Hjel =
σ
dr Ψ† (r) − σ
2
2
2m
Ψσ (r)
=
kσ
2 k2
2m
c† ckσ . y.3: A sketch showing the periodic potential.21)
The ground state for N electrons at zero temperature is denoted the Fermi sea or the Fermi sphere |FS . . z) = 0 (likewise for the y and z directions). z) and ψ (L. where εk1 ≤ εk2 ≤ εk3 ≤ . k k k k k k (2.20)
For the further analysis in second quantization it is natural to order the single-particle 2 k2 states ψkσ (r) = |kσ according to their energies εk = ~2m in ascending order. kF m
Vel-latt Vel-jel
Figure 2.23) kF = .22)
The energy of the topmost occupied state is denoted the Fermi energy. (2. |k2 . while the latter yield standing waves carrying no current.

and n ≈ 1029 m−3 follows. 0). and from Eqs. the Fermi wavenumber.25)
and we arrive at the extremely important formula:
3 kF = 3π 2 n. The Fermi sphere is shown as a circle with radius kF . εF . and vF . In A monovalent Cu one electron thus occupies a volume (2×10−10 m)3 . and εF and kF are indicated.23) and (2.4. and the microscopic quantity kF . THE NON-INTERACTING ELECTRON GAS
37
¤ £¢ ¡ ¡
¨¦ ¥§ ¡¢ ¢© © ¤ ¡
. To the left the dispersion relation εk is plotted along the line k = (kx . As a ﬁrst exercise we calculate the relation between the macroscopic quantity n = N/V.03 eV = 81600 K vF = 1. 0). 0. (2.
Thus in |FS all states with εk < εF or |k| < kF are occupied and the rest are unoccupied.27)
θ(x) = 1 for x > 0 and θ(x) = 0 for x < 0 The density can also be estimated as follows. A sketch of |FS in energy.1. 3π 2 F (2.24)
The matrix element is easily evaluated. 2. Filled and empty circles represent occupied and unoccupied states. since nkσ |FS = |FS for |k| < kF and 0 otherwise.4: Two aspects of |FS in k-space.6 nm−1 λF = 0.47×1028 m−3 .
(2. the density. respectively.005 c. ky .
(2.57×106 m/s = 0. ˆ N = FS|N |FS = FS|
kσ
nkσ |FS =
σ
V (2π)3
dk FS|nkσ |FS .and k-space is shown in Fig. This is written in terms of the theta function3 N=
σ
V (2π)3
dk θ(kF −|k|) FS|FS =
2V (2π)3
kF 0
dk k 2
1
2π
d(cos θ)
−1 0
dφ 1 =
V 3 k .
¡¤
¡ ¡ ¤ ¡
2.
(2.26)
This formula allows us to obtain the values of the microscopic parameters kF .26) it thus follows that for copper kF = 13. To the right the occupation of the states is shown in the plane k = (kx . n = 8.Figure 2.46 nm
3 4
εF = 7. The inter-atomic distances are typically 2 ˚. Hall measurements yield the electron density of copper4 .

5π 2 2m F 5
In the last equation we again used Eq. and we need not invoke relativistic considerations.26). (2. thus the low temperature limit of nF (εk ) is of importance: nF (εk ) = 1 e
β(εk −µ)
+1
T →0
−→ θ(µ − εk ). The kinetic energy per particle becomes an important quantity when we in the next section begin to study the Coulomb interaction.125). E (0) = dε ε D(ε).5. and even though the Fermi velocity is large it is still less than a percent of the velocity of light. before turning to the problem of the Coulomb interaction.
−
∂nF β 1 = −→ δ(µ − εk ).
3
(2. N 5 2m 5 2m
(2. and the total energy. Note that to be able to see any eﬀects of the temperature in Fig.003.32)
.28) it can be expressed in terms of n:
2 2 3 2 2 3 2 E (0) = kF = (3π 2 ) 3 n 3 . dε Again using Eq. kT is set to 0. since the system consists of N electrons each with an energy 0 < εk < εF .5. N = dε D(ε).31)
The density of states is a very useful function. (2. We move on to calculate the ground state energy E (0) : E (0) = = FS|Hjel |FS =
kσ 2 k2
2m
1
FS|nkσ |FS = 2
2π
2 V (2π)3 2m
dk k 2 θ(kF − |k|) (2.26) and (2.38
CHAPTER 2. (1. (2. counting the number ∆N of states in the energy interval ∆ε around dε the energy ε.28)
2V (2π)3 2m
2
kF 0
dk k 4
d(cos θ)
−1 0
dφ 1 =
V 2 5 3 k = N εF .30)
and from this dn 1 d(ε) = = 2 dε 2π 2m
2
3 2
ε θ(ε). for ε > 0. and the density of states per volume d(ε) = D(ε)/V = dn . 2. Room temperature yields kT /εF ≈ 0. THE ELECTRON GAS
Note that the Fermi energy corresponds to an extremely high temperature. 2 β ∂εk 4 cosh [ (εk − µ)] T →0 2 (2.
1
(2.03 εF corresponding to T ≈ 2400 K. As temperature is raised from zero the occupation number is given by the Fermi-Dirac distribution nF (εk ). 2. ∆N = D(ε)∆ε.3
Non-interacting electrons at ﬁnite temperature
Finally.26) we ﬁnd
2
εF =
2m
kF =
2
2
2m
(3π ) n
2
2 3
2 3
⇒
1 n(ε) = 2 3π
2m
2
3 2
ε 2 .1. see Eq.
1 2
dN V D(ε) = = 2 dε 2π
2m
2
3 2
ε 2 θ(ε). In the following we shall for example demonstrate how in terms of D(ε) to calculate the particle number. The result is reasonable. which we shall return to shortly. we study some basic temperature dependencies.
2. By Eqs. The main characteristics of this function is shown in Fig.29)
The next concept to be introduced for the non-interacting electron gas is the density of states D(ε) = dN .

This rather high value is chosen to have a clearly observable deviation from the T = 0 case.. nF (εk )d(εk ). corresponding to T = 2400 K in metals. Indeed. We thus exclude the q = 0 term in the following sums. The homogeneous part. 2. ELECTRON INTERACTIONS IN PERTURBATION THEORY
39
Figure 2. which in k-space (see Eqs. only the states within 2kT of εF are actually aﬀected. which is indicated by
¦¥
¡ © ©¨¦ 0) ( ' ¡ § ¥
¦¥
¡ © §¦ ¥ ©
!" £ #$% £ &$%
¦¥
¢ ¡ © ¨¦ ¥ ¡ © § £¤
. and its product with k the density of states.3) into account.27) is around 80000 K for metals. we ﬁnd that even at the melting temperature of metals only a very limited number ∆N of electrons are aﬀected by thermal ﬂuctuations. A careful analysis based on the so-called Sommerfeld expansion combined with the fact that the number of electrons does not change with temperature yields
∞
∂n
n(T = 0) = n(T ) =
0
dε d(ε)f (ε)
⇒
µ(T ) = εF
1−
π2 12
kT εF
2
+ . and more precisely we have ∆N/N = 6kT /εF (≈ 10−3 at room temperature).104)) corresponds to a vanishing wavevector q = 0.2
Electron interactions in perturbation theory
We now apply standard perturbation theory to take the inhomogeneous part of the electron-electron interaction Vel−el of Eq.33)
Because εF according to Eq. where the electron density is much smaller than in metals. which is indicated by the dashed lines. (1. But in fact µ varies slightly with temperature. Now we have at hand an explanation of the old paradox in thermodynamics. has already been taken into account in the jellium model to cancel the homogeneous positive background. Only at temperatures comparable to εF they begin to play a major role.1.5: The Fermi-Dirac distribution nF (εk ). its derivative − ∂ε F .
2.2. (2.
(2. the natural single-particle energy variable in these fundamental expressions actually is ξk = εk − µ and not εk itself.. (2.03 εF . Note that.103) and (1. The electronic degrees of freedom are simply ’frozen’ in. as to why only the ionic vibrational degrees of freedom contribute signiﬁcantly to the speciﬁc heat of solids.2.1 this picture is not true for semiconductors.5. shown at the temperature kT = 0. as mentioned in Sec.3. it is only slightly smeared. 1. At T = 0 the chemical potential µ is identical to εF . The Fermi sphere is not destroyed by heating. As we shall see in Sec.

2 ∂E0 ∂a0
= 0. The prototypical example is of course the hydrogen atom.053 nm. we arrive at E0 = 0 found either by minimization. The values of a and E0 are 0 0 a0
a0 =
me2 0
= 0. The ground state is a spherical symmetric s-wave with a radius denoted the Bohr radius a0 and an energy E0 . The ﬁrst order calculation works well and good physical conclusions can be drawn.29). Consequently. 2a0 1 Ry =
2ma0 2
~2 − e2 . we approach the problem from this limit in the following analysis. a 0 kF
(2.34)
However. as we shall see.36) 2a0
Here we have also introduced the energy unit 1 Ry often encountered in atomic physics as deﬁning a natural energy scale. Due to the Pauli exclusion principle the kinetic energy simply becomes the dominant energy scale in the interacting electron gas at high densities.
(2. We begin the perturbation treatment by establishing the relevant length scale and energy scale for the problem of interacting charges. This is to be compared to the typical 1 ¯ ¯ potential energy of particles with a mean distance d. The following considerations may be helpful mnemotechnically. (2.35)
revealing the following perhaps somewhat counter intuitive fact: the importance of the electron-electron interacting diminishes as the density of the electron gas increases.6 eV.6 eV. where a single electron orbits a proton. and the dimensionless measure rs of the average inter-electronic distance in the electron gas is introduced as the radius in a sphere containing exactly one electron: 4π 1 3π 2 (rs a0 )3 = = 3 3 n kF ⇒ a 0 kF = 9π 4
1 3
−1 rs
⇒
rs =
9π 4
1 3
1 .37)
. q 2 k1 +qσ1 k2 −qσ2 k2 σ2 k1 σ1
(2. THE ELECTRON GAS
4πe2 † 0 c c† c c . The key to the answer lies in the density dependence of the kinetic energy 2 Ekin = E (0) /N ∝ n 3 displayed in Eq.
E0 = −
e2 0 = 13. Lengths are naturally measured in units a0 . A natural question arises: under which circumstances can the non-interacting electron gas actually serve as a starting point for a perturbation expansion in the interaction potential. (2. So we ﬁnd that 0
1 Epot n3 ∝ 2 = n− 3 Ekin n3 1
n→∞
−→
0. Writing E0 as the sum of kinetic energy p2 /2m and potential energy −e2 /a0 . or by using the virial theorem Ekin = − 1 Epot : 2 e2 0 = −13. The typical length scale a0 yields a typical momentum p = /a0 . Epot e2 /d ∝ n 3 . It turns out that to get rid of these divergences the more powerful tools of quantum ﬁeld theory must be invoked. but already in second order the calculation collapses due to divergent integrals. the direct use of this interaction with the tools developed so far becomes the story of the rise and fall of simple minded perturbation theory.40 a prime: Vel−el = 1 2V
k1 k2 q σ1 σ2
CHAPTER 2. But let us see how we arrive at these conclusions.

σ2
2 3
4πe2 0 FS|c† 1 +qσ1 c† 2 −qσ2 ck2 σ2 ck1 σ1 |FS . 2.40) (2. θq . which results in a simple anticommutator yielding the occupation number operators with a minus in front.
FS|c† 1 +qσ1 c† 2 −qσ2 ck2 σ2 ck1 σ1 |FS = δk2 .k2 σ1 .6. 2a0 s rs
(2. Since only one k-vector appears we now drop the index 1. φq ) and (k.k1 +q δσ1 . First note.k1 +q δσ1 .38) 41
.1
This constitutes the zero’th order energy in our perturbation calculation. Second.σ2 θ(kF −|k1 +q|)θ(kF −|k1 |). Second.39) The matrix element is evaluated as follows.where q = 0 leads to k1 + q = k1 . the factor FS| demands that the two creation operators bring us back to |FS . = −δk2 .2.2. θk .
The ﬁrst order energy E (1) is found by the standard perturbation theory procedure:
¡¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡¡ ¡ I ¡¡ ¡F Q VD ¡¡ ¡ H ¡¡ ¡I G ¡¡ ¡P P ¡¡ ¡UE G ¡¡ ¡F TH ¡¡ ¡ FI ¡¡ ¡ D ¡¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡¡ ¡ ¤¥¥¤¥ ¤¥¥¤¥ ¤¥¥¤¥ ¤¥¥¤¥I ¤¥¥¤¥Q DH¤¥¥¤¥ HG¤¥¥¤¥GF¤¥¥¤¥PE D¤¥¥¤¥E F¤¥¥¤¥C H¤¥¥¤¥ I¤¥¥¤¥D ¤¥¥¤¥ ¤¥¥¤¥ ¤¥¥¤¥ ¤¥¥¤¥ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡B ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡A ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡R ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ R ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ B¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ A¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ 02 ) ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡1 4 ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡R ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡)4( ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ )#"#" ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡S6'&'& ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡5 ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ 4 ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ) ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ %$$% ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡0!! ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ 2) ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ 1 ¡ ¡ ¡¡ ¡ ¡ ¡¡ ¡ ¡¡ ¡ ¡ 076)(5 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥2 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥984 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥1 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥674 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥) 3¤¤(@¤¤¤5¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥2 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥089 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥1 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥706 ¤¥¤¥¤¤¥¥¤¥¤¤¥¤¥¥¤¥¤¤¥¥¤¥¤¤¥ ) (5
2.21 0 r ≈ 2 Ry. These possibilities are sketched in Fig.6: The two possible processes in ﬁrst order perturbation theory for two states |k1 σ1 and |k2 σ2 in the Fermi sea. The direct process having q = 0 is already taken into account in the homogeneous part.k1 +q δσ1 . Also the geometry for the k-integration is shown for an arbitrary but ﬁxed value of q. k k 2 q
e2 −2 2. The k.and q-sum are converted into integrals.σ2 FS|c† 1 +qσ1 c† 1 σ1 ck1 +qσ1 ck1 σ1 |FS k k k k FS|Vel−el |FS E (1) 1 = = N N 2VN (a k )2 31 2 2 31 E (0) 3 9π = kF = (a0 e2 ) 0 F = 0 2 N 52m 52 a0 5 4
Electron interactions in 1st order perturbation theory
q
= −δk2 . For q = 0 we therefore end with 2. hence only the exchange process contributes to Vel−el . that the integral is independent of the direction of q 1 2π so that −1 d(cos θq ) 0 dφq = 4π. and polar coordinates (q. ELECTRON INTERACTIONS IN PERTURBATION THEORY
B A 9hg ( 5 ( 8g U H D Sf5 C F T e ¢aHbQE dc E F Q ¢££¢£¢¢£¢££¢¢£¢££¢£¢£ F ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£V7¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£¨¨©©¨©¨¨©¨©©¨¨©¨©©¨©¨© ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£¨¨©©¨©¨¨©¨©©¨¨©¨©©¨©¨©I ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£¨¨©©¨©¨¨©¨©©¨¨©¨©©¨©¨©Q D¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£¨¨©©¨©¨¨©¨©©¨¨©¨©©¨©¨©H ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£P ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£E ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£V F¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£H¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£I¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£D ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£ ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£ ¢¢££¢£¢¢£¢££¢¢£¢££¢£¢£ ¢¢¢¢¢¢¢¢¢¢¢ B `8 £¢£¢¢£¢££¢£ £¢£¢¢£¢££¢£ 5£¢£¢¢£¢££¢£¦¦§§¨©¦§¨¨©¨©©¨© 4£¢£¢¢£¢££¢£¦¦§§¨©¦§¨¨©¨©©¨© £¢£¢¢£¢££¢£¦¦§§¨©¦§¨¨©¨©©¨© £¢£¢¢£¢££¢£¦¦§§¨©¦§¨¨©¨©©¨© £¢£¢¢£¢££¢£ £¢£¢¢£¢££¢£5 £¢£¢¢£¢££¢£ £¢£¢¢£¢££¢£ £¢£¢¢£¢££¢£ £¢£¢¢£¢££¢£ £¢£¢¢£¢££¢£ £¢£¢¢£¢££¢£ £¢£¢¢£¢££¢£ ¢¢¢¢¢ A ¢B Y¢0¢A £¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ £¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£¦§¦§¦¦§§¦§¦¦§¦§§¦§¦¦§§¦§¦§ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£¦§¦§¦¦§§¦§¦¦§¦§§¦§¦¦§§¦§¦§ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£¦§¦§¦¦§§¦§¦¦§¦§§¦§¦¦§§¦§¦§ £¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£¦§¦§¦¦§§¦§¦¦§¦§§¦§¦¦§§¦§¦§ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£X¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£W ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ ¢£¢¢£¢££¢¢£¢££¢£¢¢£¢££¢¢£¢£ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢
Rewriting the energy E (0) of the non-interacting electron gas to these units we obtain:
Figure 2. thus either q = 0 (but that is excluded from Vel−el ) or k2 = k1 + q and σ2 = σ1 . the two annihilation operators can only give a non-zero result if both |k1 | < kF and |k2 | < kF . First. φk ) are employed. only for 0 < q < 2kF does the theta function (2.σ2 FS|nk1 +qσ1 nk1 σ1 |FS
k1 .

now we have learned that the interaction energy per particle is ≈ 1. is the explanation of why the Fermi surface is not destroyed by the strong Coulomb interaction between the electrons. THE ELECTRON GAS
product give a non-zero result.083 Ry = −1. and that for a given cos θk we have q/(2 cos θk ) < k < kF . The last variable is free: 0 < φk < 2π. 2. The geometry of this volume is sketched in Fig. smaller than but certainly comparable to εF .42) 2π rs
The ﬁnal result for the ﬁrst order perturbation theory is thus the simple expression E N −→ E (0) + E (1) = N 2. We found before that thermal smearing occurs but is rather insigniﬁcant compared to the huge Fermi energy.42
CHAPTER 2. For a given ﬁxed value of q the rest of the integral is just the overlap volume between two spheres of radius kF displaced by q. 2 for symmetry.41)
where the prefactors are a factor 2 for spin.6. We thus get E (1) 4πe2 V 0 =− 2(4π) N 2VN (2π)3
2kF 0
dq
q2 2(2π) q2
1
q 2k F
d(cos θk )
V (2π)3
kF
q 2 cos θk
dk k 2 .211 0. E∗ N E N
= −0. since only one electron at a time can be at a given point in space. 4π for q-angles.83.
(2. The negative exchange energy overcomes the positive kinetic energy.29 eV = −0. No external conﬁnement potential is needed to hold the electron gas in the ion ∗ jellium together.44) We note that the negative binding energy is due to the exchange energy of the Coulomb interaction. The integral is elementary and results in k 3 E (1) e2 V kF 4 e2 e2 =− 0 = − 0 (a0 kF ) F = − 0 N 2 N 2π 3 2a0 2π 3 n 2a0 9π 4
1 3
1 rs
3 0.13 eV
(1st order perturbation theory) (experiment on Na)
rs = 3.3 eV. One of the great results of quantum ﬁeld theory. i.916 ≈− Ry. or inter-particle distance rs . and is calculated by noting that q/2kF < cos θk < 1.916 − 2 rs rs Ry. and twice V/(2π)3 for the conversions of k.96.43)
rs →0
This result shows that the electron gas is stable when the repulsive Coulomb interaction is turned on.095 Ry = −1.and q-sums to integrals. 2π for φk . Physically this can be interpreted as an eﬀect of the Pauli exclusion principle: the electrons are forced to avoid each other.
(2. and we can compare the result with experiment: ∂rs (E
∗ rs = 4. There exists an optimal density n∗ . The equilibrium situation is obtained from ∂ (0) + E (1) ) = 0. One may wonder what happens to the Fermi sphere as the interaction is turned on. which we are going to study later in the course. which minimizes the energy and furthermore yields an energy E ∗ < 0. The direct “classical” Coulomb interaction does not take this into account and is therefore over-estimating the energy. (2.
. εF ≈ 7 eV. However. and the exchange part corrects for this by being negative.e. (2.

83 with an ionization ∗ /N = −1.2. |FS is restored by putting the excited electrons back into the holes they left behind. As sketched in Fig. We now proceed to show that the direct interaction process gives a divergent con(2) tribution Edir to E (2) due to the singular behavior of the Coulomb interaction at small momentum transfers q. energy E/N = E
2. ELECTRON INTERACTIONS IN PERTURBATION THEORY
Figure 2. Here we can only reveal what goes wrong.8 this combined with the momentum conserving Coulomb interaction yields intermediate states where two particles are injected out of the Fermi sphere. (2.46) k k To restore |FS the same momentum transfer q must be involved in both ν|Vel−el |FS and FS|Vel−el |ν .7: (a) The energy per particle E/N of the 3D electron gas in ﬁrst order perturbation theory Eq. The matrix elements diverge without giving hope for a simple cure.47) E (0) − Eν
§
© §¨¥¦¤£ £¤ ¢ ¡
43
.43) as a function of the dimensionless inter-particle distance rs . Only two types of processes are possible: the direct and the exchange process. (2.2
Electron interactions in 2nd order perturbation theory
One may try to improve on the ﬁrst order result by going to second order perturbation theory.2. Due ∗ to the exchange interaction the electron gas is stable at rs = rs = 4. and then later learn how to deal correctly with the inﬁnities occurring in the calculations. E (0) − Eν (2.2. For the direct process the constraint |ν = |FS leads to |ν = θ(|k1+q|−kF )θ(|k2−q|−kF )θ(kF−|k1 |)θ(kF−|k2 |)c† 1 +qσ1 c† 2 −qσ2 ck2 σ2 ck1 σ1 |FS . 2. From such an intermediate state. the result is disastrous. and writing Vq = Edir =
(2) 4πe2 0 q2
we ﬁnd
1 V2
q k1 σ1 k2 σ2
1 ( 2 Vq )2 θ(|k1 +q|−kF )θ(|k2 −q|−kF )θ(kF −|k1 |)θ(kF −|k2 |). However.45)
|ν =|FS
where all the intermediate states |ν must be diﬀerent from |FS .29 eV. (2. According to second order perturbation theory E (2) is given by 1 E (2) = N N FS|Vel−el |ν ν|Vel−el |FS .

The exchange process does not lead to a divergence since in this case the momentum 2 transfer in the excitation part is q.3
from which we obtain
The contribution from small values of q to Edir is found by noting that
Electron gases in 3. In fact. as we shall see in Chap.49)
. this is not just an academic exercise. The direct process gives a divergent contribution to E/N while the exchange process gives a ﬁnite contribution. 13. But as will become clear. THE ELECTRON GAS
1 = ln(q) q
q→0
0
∝ ∞. Thus Vq −2 for q → 0. it turns out that one has to consider perturbation theory to inﬁnite order. but in the relaxation part it is k2 − k1 − q. .48a)
(2. 2. which is possible using the full machinery of quantum ﬁeld theory to be developed in the coming chapters.
££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ H P C G F I D E G H C ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ H P C G F I D E G H C ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ IHUE T F S E ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ G F E D BC ££££££££££££££££ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ )¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ( ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ @A ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ @A ££££££££££££££££ ££££££££££££££££ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ &&'' 5 $%%$ 4 03 ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡5 4 73 ) R ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ( ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££75 4 73 ) 95 0 4 03 ) ( ££££££££££££££££ ( ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ !! !! ££££££££££££££££ ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ 5 4 "##" 7823 ) Q5 4 ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ ( ¡¡¡¡¡¡¡¡¡¡¡¡¡¡¡ 0123 ) ( 75 4 ¥¤¥¤ 7823 ) 65 0 4 §¦§¦ ©¨¨© ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ ( ¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢¢ 0123 ££££££££££££££££ )(
2. and 1D.48c)
(2. which is less singular than V 2 ∝ q −4 . To work in various dimensions is a good opportunity to test ones understanding of the basic principles of the physics of electron gases. θ(|k1 +q|−kF )θ(kF −|k1 |) ∝ q.8: The two possible processes in second order perturbation theory for two states |k1 σ1 and |k2 σ2 in the Fermi sea.
Figure 2. We know that physically the energy of the electron gas must be ﬁnite. is replaced by Vq Vk2 −k1 −q ∝ q q This divergent behavior of second order perturbation theory is a nasty surprise.48b)
(2. The only hope for rescue lies in regularization of the divergent behavior by taking higher order perturbation terms into account.
44
We end this chapter on the electron gas by mentioning a few experimental realizations of electron gases in 3D.
q→0
(2. Electron gases at reduced dimensionality is of increasing experimental importance. 1. and 0 dimensions
Edir ∝
2 Vq ∝ k1 (2)
1 q4 E0 − Eν ∝ k2 + k2 − (k1 + q)2 − (k2 − q)2 ∝ q 1 2
.
0
dq q 2
1 1 qq= q4 q
(2)
0
dq
CHAPTER 2. 2D. .

on the other hand. In an insulator εF is at the top of a band. the Maxwell-Boltzmann distribution. In the metallic case εF lies in the middle of a band.1
3D electron gases: metals and semiconductors
Bloch’s theory of non-interacting electrons moving in a periodic lattice provides an explanation for the existence of metals. 2. The system can only be excited by providing suﬃcient energy for the electrons to overcome the energy band gap Egap between the top of the valence band and the bottom of the next empty band. and hence the inability of insulators to conduct electronic thermal and electrical currents. Likewise. The Fermi level εF lies at the top of the valence band resulting in possible excitations energies of at least Egap .e. the so-called conduction band.3.9: (a) A generic bandstructure for a metal. typically less than 2 eV. i. 2.3. and any however small external ﬁeld can excite the system and give rise to a signiﬁcant response. is normally described as a classical gas since for energies εk in the conduction band we have εk−µ > Egap /2 kB T . Consequently there is no energy gap between the last occupied level and the ﬁrst unoccupied level. such that at room temperature a suﬃcient number of electrons are excited thermally up into the conduction band to yield a signiﬁcant conductivity. see
1 k Transport properties are tightly connected to the electron velocity vk = ~ ∂k . and ﬁlled bands does not carry any electrical or thermal current5 . (b) A generic bandstructure for an insulator or a semiconductor. The Fermi level εF lies in the middle of a band resulting in arbitrarily small possible excitations energies.Figure 2. we note that in a typical metal most of the electron states at the Fermi surface are far away from the regions in k-space where the free electron dispersion relation is strongly distorted by the periodic lattice. for the thermal current Jth = 2 k∈FBZ V εk vk = 2 R dk 1 ∂(εk ) . 1. ELECTRON GASES IN 3. and band insulators. The important parameter is the position of the Fermi energy εF relative to the bands as sketched in Fig. semiconductors. the so-called ﬁlled valence band. A semiconductor. and consequently nF (εk ) → e−(εk −µ)/kB T .
2. AND 0 DIMENSIONS
45
©¨©¨ § £ ¦ ¥ £¤¢ ¡
. Therefore one ﬁnds eﬀective masses m∗ . Both currents are integrals over FBZ of gradients of periodic functions and therefore FBZ (2π)3 ~ ∂k zero. The current density R P P 1 dk 1 ∂εk 1 is J = 2 k∈FBZ V vk = 2 FBZ (2π)3 ~ ∂k . the distance up to the unoccupied conduction band. Finally. We emphasize that at room temperature the electron gas in a metal is a degenerate Fermi gas since kB T εF . Semiconductors are insulators at T = 0. This is not possible for small external ﬁelds.9. 5 ∂ε
!©¨ ©¨ § ¦ ¥ ¢ £
2. but their band gap Egap is relatively small.

However.3 the conduction band in Ga1−x Alx As is 300 meV higher than the one in GaAs. In Fig. The main diﬀerence between the two semiconductor crystals is the values of the bottom of the conduction band. The ionized Si donors left in the Ga1−x Alx As provide an electrostatic energy that grows with an increasing number of transferred electrons. (b) A sketch showing the diﬀerent layers in the semiconductor structure as well as some surface gates deﬁning the geometry of the device. The interface between the GaAs and the Ga1−x Alx As semiconductor crystals in the GaAs/Ga1−x Alx As heterostructure can be grown with mono-atomic-layer precision in molecular beam epitaxy (MBE) machines. 2. THE ELECTRON GAS
Figure 2. not all donor electrons will be transferred. Eq. At T = 0 there are of course no free carriers in any of the conduction bands for pure semiconductor systems. close to the vacuum mass m.16). but by doping the Ga1−x Alx As with Si.2
2D electron gases: GaAs/Ga1−x Alx As heterostructures
For the past three decades it has been possible to fabricate 2D electron gases at semiconductor interfaces. and the more resent realization being in the gallium-arsenide/gallium-aluminum-arsenide (GaAs/Ga1−x Alx As) heterostructures. This is because the two semiconductor crystals have nearly the same crystal structure leading to a stress-free interface. Hence the electrons in the former conduction band can gain energy by moving to the latter. In contrast. (2. which by wet etching has gotten a device geometry imprinted in its surface. all the electron states contributing to the transport properties in a semiconductor are close to these regions in k-space.
2. At some point the energy gained by
§ (21 ¡ ¡ 0( ¨ § © '& % ) § ¢ § $ # ©" " ! ¡ ¢ § ¡ © ¡ ¡
§ ¨ §© § © ¡ ¨ § ¦
¦¥ ¤ ¢ £
¢ ¡
. Niels Bohr Institute. The metal contacts and wires are seen attached to the GaAs structure. typically m∗ ≈ 0. conduction electrons are provided. and one ﬁnds strongly modiﬁed eﬀective masses.3. which then accumulate on the GaAs side of the interface due to the energy gain. the key component in integrated electronic circuits.10 a picture of an actual device is shown as well as a sketch of the various layers in a GaAs heterostructure.46
CHAPTER 2.10: (a) A picture of a GaAs-device fabricated at the Ørsted Laboratory. the ﬁrst realization being inversion layers in the celebrated silicon MOSFETs.1 m. For x = 0. and which is essential for the basic research of many quantum eﬀects in condensed matter physics. which is technologically important for high-speed electronics. In the latter system one can obtain extremely long mean free paths (more than 10 µm).

n being the 2D electron density at the interface. The wavefunctions ζn (z) and eigenvalues of the lowest three electron eigenstates in the triangular well. and
© 8 £ § ¡ © § § % ¥ !¡ %& ' ¡¨% ¥ ¡! " %& $ # ¡ £ ! £ ¢ ¡ §8¥¦ 3 1 270 6 ( £¡ ¡¢ ¡¢ ¦¦ ¥¥ 1¤ 33 1 215200 4) (( ¦¤ ¥ ¡¢
.067m. ELECTRON GASES IN 3. yielding zero from the side of the cylindrical box. and since for the reason of charge neutrality. The electrical ﬁeld E at the interface is then found simply by forming a cylindrical Gauss box with its axis along the z direction and one circular ’bottom lid’ at the interface and the other ’top lid’ deep into the GaAs.2. 2 The signiﬁcance of this quantization energy is the following. m∗ = 0.3. Due to the triangular well the 3D free electron wavefunction is modiﬁed. 2. since for symmetry reasons E must be perpendicular to the z axis. The charge density outside this capacitor is zero. We consider the positively charged layer of the ionized Si donors as one plate of a plate capacitor. as calculated from Poisson’s equation: 2 V = −e2 n3D / ∗ . E = 0 at the ’top lid’. where we have used the Bohr radius a0 of Eq. From this we get the typical quantization energy ∆E due to the triangular well: 2 m a ∆E = 13. see Fig. The typical length scale l for the width of the triangular well is found by balancing the ∗/ 1 ~2 0 a0 potential energy and the kinetic quantum energy: eEl = m∗ l2 ⇒ l3 = 4π m∗ /m n .11. 2. The key point to notice is the formation of the almost triangular quantum well at the GaAs side of the interface.36) to bring in atomic units. and typically n = 3 × 1015 m−2 . This is sketched in Fig. Without performing the full calculation we can get a grasp of the order of magnitude by the following estimate. transferring electrons to the GaAs layer is balanced by the growth in electrostatic energy. 1. The well is so narrow that a signiﬁcant size-quantization is obtained. (2. A (2. All the contributions stems from the ’bottom lid’. 1 ψkσ (r) = √ eikx x eiky y eikz z χσ V −→ 1 ψkx ky n σ (r) = √ eikx x eiky y ζn (z) χσ . The conduction band is not ﬂat due to the curvature induced by the charge densities. The experimental input for GaAs is ∗ = 13 0 .11 where the resulting conduction band in equilibrium is shown as function of the position z perpendicular to the interface. Thus at the interface E = en/ ∗ . Note the triangular well forming at the interface.11: The conduction band in a GaAs/GaAlAs heterostructure. which yields l ≈ 5 nm.50)
where ζn (z) is the nth eigenfunction of the triangular well having the eigenenergy εn . 2. Only the z direction is quantized leaving the x and y direction unaltered. while the conduction electrons at the GaAs/GaAlAs interface forms the other plate.6 eV m∗ l0 ≈ 20 meV. AND 0 DIMENSIONS
47
Figure 2.

σ (r) = √ eikx x Rnl (r) Yl (φ) χσ . The electrons are strongly bound to the surface of the cylinder in quantum states arising from the original π-bonds of the graphite system. r. semiconducting or insulating. the so-called carbon nanotube. The corresponding conﬁnement φ ~2 energy is En ∼ 2mλ2 ∼ 1 eV n. The highest occupied state has the energy E0 + εF while the lowest unoccupied state has the energy E1 . two of the three spatial degrees of freedom are frozen in. They are believed to be the strongest material in the world.ky . φ) wavefunctions: 1 ψkσ (r) = √ eikx x eiky y eikz z χσ V −→ 1 ψkx .26) for the 3D case). ∆R ≈ 0. in the azimuthal angle coordinate φ.48 the total energy for all three spatial directions is
2
CHAPTER 2. ζ0 (z). and we arrive at our conclusion: At temperatures T 100 K all occupied electron states have the same orbital in the z direction. i. λn = 2πR0 /n < 2 nm.51)
where we have given the 2D version of the fundamental relation between kF and n (see Exercise 2.53)
. i. since the perimeter must contain an integer number of electron wavelengths λn .
2m
(2. 2.52)
This is of course more than just a mathematical transformation.e. We brieﬂy sketch how this comes about.e. 2D systems do indeed exist in reality. This means that the extension ∆R of the radial wave function Rnl (r) around the mean value R0 is of atomic scale. i. see Fig. and depending on the speciﬁc way the cylinder is rolled up the nanotubes are either metallic. Likewise. there is a strong conﬁnement.
2. This means that theoretical studies of 2D electron gases is far from an academic exercise.1 nm.n. resulting in a radial ~2 R conﬁnement energy E0 ∼ 2m∆R2 ∼ 10 eV.
kF 2 = 2πn ⇒ εF ≈ 10 meV. z) plane waves to cylindrical (x.3
1D electron gases: carbon nanotubes
Since the mid-nineties a new research ﬁeld has developed involving studies of the cylindrically shaped carbon based molecule.12.e. The diﬀerence is E1 − (E0 + εF ) = ∆E − εF ≈ 10 meV ≈ 100 K.l. If this is not provided the system has eﬀectively lost one spatial degree of freedom and is dynamically a 2D system.3. The carbon nanotube can be viewed as a normal graphite sheet rolled up into a cylinder with a radius R0 ≈ 2 nm and a length more than a thousand times R0 .
(2. The cylindrical symmetry of the nanotube makes it natural to change the basis functions from the 3D (x. a metallic nanotube is a nearly ideal 1D wire. There are no severe constraints along the cylinder axis. n i. We therefore end up with a total energy
R φ εkx .e.4 and compare to Eq. L (2. y.n =
2m∗
2 2 kx + ky + ε n . In the same dynamical sense as the GaAs heterostructure is a 2D metal sheet. Any changes of this orbital requires an excitation energy of at least 10 meV.n. (2. These long and thin carbon molecules have some extraordinary material characteristics. THE ELECTRON GAS
εkx .l = E0 + En + 2 2 kx . in the x direction.

10(b).2. This has been realized experimentally in the so-called quantum dot systems. 2. 2. AND 0 DIMENSIONS
¢ ¤ ¢ £ ¢ ¡
49
Figure 2.
2.12: (a) Carbon atoms forming a sheet of graphite with a characteristic hexagonal lattice. A simpliﬁed model of a quantum dot is studied in Exercise 8.4
0D electron gases: quantum dots
Naturally one can think of conﬁning the electrons in all three spatial dimensions.
. 0) band (the position of εF for the metallic case) to the bottom of the (n. also from a pure theoretical point of view do they play an important role. 2 (2. a central quantum model for describing interacting electrons in 1D. Not only are the nanotube very interesting from an experimental point of view. (c) An atomic force micrograph taken at the Ørsted Laboratory. thereby connecting them and allowing for electrical measurements on single molecules. Niels Bohr Institute. (b) A carbon nanotube molecule is formed by rolling up a graphite sheet into a cylindrical geometry. l) = (0. ELECTRON GASES IN 3. for example by using the device shown in Fig. This section will be expanded in the next edition of these notes. 1. In particular that makes the nanotubes a key testing ground for the diagonizable so-called Luttinger liquid model. The nanotubes is one of only a couple of systems exhibiting a nearly ideal 1D behavior. 1) band: 1 φ φ ∆E = (E1 − E0 ) ≈ 1 eV ∼ 12000 K.4.54)
Thus at room temperature the only available degree of freedom is the axial one described by the continuous quantum number kx and the associated plan waves eikx x . showing a bundle of carbon nanotubes placed across a gap between two metal electrodes. with a considerable gap ∆E from the center of the (n.3.3. l) = (0.

50
CHAPTER 2. THE ELECTRON GAS
.

speciﬁc heat.127).Chapter 3
Phonons. The restoring force is the long range Coulomb interaction. The restoring force is the long range Coulomb interaction. In the following sections we shall study the three types of matter oscillation sketched in Fig. (b) slow ion oscillations followed by the electron gas (sound waves. The ions will be treated using two models: the jellium model. (c) Fast electronic plasma oscillations in a static ionic lattice (electronic plasma oscillations). they naturally occur at ﬁnite temperature. The restoring force is the compressibility of the disturbed electron gas. The grayscale represent the electron density and the dots the ions.1: Three types of oscillations in metals. where the ions oscillate around their equilibrium positions forming a regular crystal lattice.1. the Bose-Einstein distribution nB (ε) given in Eq. 1. Phonons play are fundamental role in our understanding of sound. (1. elasticity. where the ions are represented by a smeared-out continuous positive background. 51
.1 to achieve the second quantized form of the corresponding Hamiltonian. (a) Slow ionic density oscillations in a static electron gas (ion plasma oscillations). The quantized vibrations are denoted phonons. acoustic phonons). and the lattice model. 3. a name pointing to the connection between sound waves and lattice vibrations.
¢ ¤ ¢ £ ¢ ¡
Figure 3. 1. so we will therefore often need the thermal distribution function for bosons. they are bosons according to the results of Sec. Since phonons basically are harmonic oscillators. Moreover.4.1. These vibrations are well described by harmonic oscillators and therefore we can employ the results from Sec. coupling to electrons
In this chapter we study the basic properties of ionic vibrations.4. And more surprising may be the fact that the electron-phonon coupling is the cause of conventional superconductivity. and electrical resistivity of solids.

This is explained in details in Secs.3) are not a ion very good description of solids. the exact behavior is described by the Debye-model incorporating phonons with a photon-like dispersion ωq = vs q. the Einstein phonons (also denoted optical phonons.
(3. Although it is correct that CV decreases at low temperature. 3. t) = δρion (r) e−iΩt .1b and in Sec. 3. This force equation is supplemented by the continuity ion equation. We begin as depicted in Fig.3 and 3. However. see Sec. but we hint at the solution of the problem in Fig. which all have the same frequency Ω.2)
In the second equation we have introduced the force density f . Let ρ0 be the particle density of the ion jellium. If we study the limit of small harmonic deviations from equilibrium δρion (r.5.1)
A non-zero δρion corresponds to a charge density Ze δρion and hence is associated with an electric ﬁeld E obeying ·E= Ze
0
δρion
⇒
·f =
Z 2 e2 ρ0 ion
0
δρion . which to ﬁrst order in δρion becomes f = Zeρion E ≈ Zeρ0 E. To fully understand how the optical Einstein phonons get renormalized to become the acoustic Debye phonons requires the full machinery of quantum ﬁeld theory. The ionic oscillations in the continuous jellium model are thus described by harmonic oscillators. 2
(3. the second quantization formalism leads to the following phonon Hamiltonian:
2 ∂t δρion +
1 M
·f =0 ⇒
Ω2 δρion =
Z 2 e2 ρ0 ion δρion ⇒ 0M
Ω=
Z 2 e2 ρ0 ion = 0M
Hph =
q
Ω b† bq + q
1 .e.52
CHAPTER 3. t) = ρ0 + δρion (r) e−iΩt . ∂t ρion + ·(ρion v) = 0. we obtain linear equations of motion with solutions of the form ρion (r.4)
These quantized ion oscillations are denoted phonons. 3. COUPLING TO ELECTRONS
3. where vs is the sound velocity. and a model like this was proposed ion by Einstein in 1906 as the ﬁrst attempt to explain the decrease of heat capacity CV of solids as a function of decreasing temperature (see Sec. ion since the velocity v is already a small quantity. and ion ρel = Zρ0 that of the homogeneous electron gas. which to ﬁrst order in δρion becomes ∂t δρion +ρ0 ·v = 0. 3. which entered the analysis through the Maxwell equation · E = Zeδρion / 0 . 0M (3.3) Ω is the ionic plasma frequency.2. Note that the origin of the ionic plasma frequency is the long-range Coulomb interaction.5). instead of the Einstein dispersion ωq = Ω. Diﬀerentiating this with respect to time and using Newton’s second law f = M ρion ∂t v we obtain Ze2 ρel . Hence. we keep ρel ﬁxed. PHONONS. 3.1
Jellium oscillations and Einstein phonons
Our ﬁrst encounter with phonons will be those arising from a semiclassical treatment of the charge neutral jellium system. 3. These Debye phonons are also denoted acoustical phonons due to their relation to sound propagation.
. i. ion (3.1(a) by ion studying oscillations in the smeared out ion density while neglecting the electron dynamics.

This setup is shown in Fig.1b. and to illustrate its correctness we now use it to estimate the sound velocity in metals. The Hamiltonian is simply the sum of the kinetic energy of the ions and the potential energy of the springs. vs = Zm v . the ions are much heavier.
(3.2.4) is expressed by the Bohm-Staver formula. The equilibrium position 0 of the j’th ion is denoted Rj . 3.2. 2M j 2
[pj1 . more than a factor of 104 .3. 0 0 The lattice spacing is denoted a = Rj − Rj−1 . in which the ions can execute small oscillatory motion around their equilibrium positions. Consequently one would expect the electrons to follow the motion of the ions adiabatically and thereby always maintaining local charge neutrality and thus lowering the high ionic plasma frequency Ω. The surroundings somehow provide the restoring force. 2 The kinetic energy density associated with a sound wave is of the order 1 M vs ρion . which in a more detailed treatment (see Exercise 3. Surprisingly.2
Electron-phonon interaction and the sound velocity
Compared to the light and very mobile valence electrons. 3. which is due to long-range charge Coulomb forces from the charge imbalance.
3. Since the equilibrium system is periodic with the lattice spacing a it is natural
.5)
which for typical numbers yields vs 3000 m/s as found experimentally. while its displacement away from this position is denoted uj . We begin by a simple one dimensional quantum mechanical model consisting of a 1D box of length L containing N ions of mass M each interacting with its two neighbors through a linear force ﬁeld (a spring) with the force constant K. 5 In a stationary state these two energy densities must be of the same order of magnitude. ELECTRON-PHONON INTERACTION AND THE SOUND VELOCITY
53
3. and much slower. This gives an estimate for vs . uN +1 = u1 . while 2 the potential energy density associated with the restoring force must be related to the density dependent energy content of the compressed electron gas.e. 3M F (3. of the order 3 ρel εF . Note how this estimate builds on classical considerations of the ionic motion while using the quantum result for the energy content of a degenerate electron gas. while the ion momentum pj and the displacement uj are canonical variables:
N
Hph =
j=1
1 2 1 p + K(uj − uj−1 )2 . we can nevertheless learn a lot from simply postulating the existence of a periodic ion lattice (as observed in nature). an ordinary macroscopic phenomenon as sound propagation is deeply rooted in quantum physics. i. This situation is depicted in Fig.3
Lattice vibrations and phonons in 1D
Even though we are not yet able to demonstrate how to turn the optical ion plasma oscillations into acoustical phonons. so we have L = N a. uj2 ] =
i
δj1 .j2 .6)
As for the photon model and the jellium model we impose periodic boundary conditions.

in reciprocal space.
k∈FBZ N j=1
0
uj
1 ≡√ N
uk eikRj .0 =
j
1 N 1 N
1 pk ≡ √ N
pj e−ikRj . The reason is that the Hamiltonian in the present case contains products like pk p−k and not p2 as in k the original case. (3. COUPLING TO ELECTRONS
Figure 3. PHONONS. (3. M 2
[pk1 . (3. (3.
0
δk. L a N (3. (3.9) tells us k that uk and p−k form a pair of conjugate variables. the hermiticity of pj and the deﬁnition
¡ 34 ©¡© ¡ ©¡©
eikRj .8)
0
i
δk1 .
0
1 uk ≡ √ N
uj e−ikRj . uk2 ] =
of the Fourier transform lead to p† = p−k .
k∈FBZ N j=1
0
T UQV QV T RQV SRQV $ ¡$ ¡" " ¡ ¡ ¡ ¡ ¡ ¡ CD 78 $ GH !" EF #¡ ¡ ¡#$ ¡!" ¡ ¡ ¡ ¡ AB ¡ ¡ 9@ ¡ #¡# ¡! ! ¡ ¡ ¡ T PQU I S R PQ I PQ I ` T R PQ I `X`XY`X` ¡¨ ¨ ¦ ¡¦ ¤ ¡¤ ¡ 56 ¡ ¢ ¡ ¡ ¢ ¡ ¡ ¡§¨ §¨ 12 ¦ )0 ¥¡ ¡¥¦ ¤ '( £¡ ¡£¤ ¡ ¢ ¢ %& ¡ WXWXYWXW ¡ ¡§ § ¥ ¡¥ W ¡£ £
e−ikRj .−k2 .77) to form creation and annihilation operators. . − + 2∆k. The equilibrium positions shown in the top row are 0 denoted Rj . and a nearest neighbor linear force coupling of strength K.8) into Eq. (1. . 2M k −k 2
ωk =
K ka 2 sin . a a a ∆k = 2π 2π 1 = . we will not use this pair in analogy with x and p in Eq. Instead we combine uk and pk in the deﬁnition of the annihilation and
This looks almost like the Hamiltonian for a set of harmonic oscillators except for some annoying details concerning k and −k. Although the commutator in Eq.2: A 1D lattice of ions with mass M . lattice constant a.
k∈FBZ N j=1
0
δR0 . also the present system of N ions forming a periodic lattice leads to a ﬁrst Brillouin zone. . − + N ∆k . Note that while pj in real space is a nice Hermitian operator. to solve the problem in k-space by performing a discrete Fourier transform.6) we ﬁnd H=
k
1 1 2 p p + M ωk uk u−k .0 =
By straight forward insertion of Eq. By Fourier transformation the N ion coordinates becomes N wave vectors in FBZ: FBZ = − π π π + ∆k.9)
. .54
CHAPTER 3. while the displacements shown in the bottom row are denoted uj .7)
The Fourier transforms of the conjugate variables are: pj 1 ≡√ N pk eikRj . In analogy with electrons moving in a periodic lattice. In fact. pk in k-space is not self-adjoint. FBZ.

Now we break the discrete translational invariance given by the lattice constant a. k
(3. the force constant in the model. These came about as a consequence of breaking the translational invariance of the system by introducing a periodic lattice.9) that ωk −→
k→0 K M ak. (3. 2
[bk1 . 2 −k i √ (b† − bk ). the unit cell is doubled to hold two ions. 3. The sound
K velocity is found to be vs = M a. and to the right the reduced zone scheme.3: The phonon dispersion relation for three diﬀerent 1D lattices.3(a) is broken into sections. The quantized oscillations are denoted phonons. To the left is shown the extended zone scheme.3(a).
so our solution Eq. 2 −k k
k
=
M ωk
. LATTICE VIBRATIONS AND PHONONS IN 1D
55
Figure 3. creation operators bk and b† : −k bk ≡ b† −k ≡ p 1 uk √ +i k . 3. (b) As in (a) but now substituting every second ion of mass M1 with one of mass M2 (white disks) resulting in two ions per unit cell and a doubling of the lattice constant. but the same lattice constant as in (a). as shown in Fig.11)
This is ﬁnally the canonical form of a Hamiltonian describing a set of independent harmonic oscillators in second quantization. Hence the original BZ is halved in size and the original dispersion curve Fig. so upon measuring the value of it.11) does in fact bring about the acoustical phonons. Instead the new lattice constant is 2a. the concept of phonon branches must be introduced. / k 2 k uk ≡ pk ≡
k√
1 † (b + bk ).10)
Note how both the oscillator frequency ωk = ω−k and the oscillator length k = depends on the wavenumber k. In
0
¡ ¡( ¡© ¡ © ¡
'&
!%# 0
'& 0 ¡ )( ¡
'& ¡
# !$ 0
¡ §¨¡§¨ ¥¦¡( ¥ ¡ ¡£ ¢ £¤
'&
# !"
. 3.3(b). (c) As in (a) but now with the addition of mass M2 ions in between the mass M1 ions resulting in two ions per unit cell. (3. b† 2 ] = δk1 .k2 . It is analogous to the Bloch bands for electrons. If. (a) A system with lattice constant a and one ion of mass M1 (black disks) per unit cell. one can determine the value of the free parameter K. Their dispersion relation is shown in Fig. Again by direct insertion it is readily veriﬁed that Hph =
k
−k
ωk b† bk + k
1 . (3.3. It is seen from Eq. / k 2 k p 1 uk √ −i k .3.

The central object in
¦
¦
(3. and in the general case the Hamiltonian Eq. 1 j N U ≈ U0 + 1 2 uα (R0 ) 1
R0 R0 1 2 αβ
∂uα (R0 ) 1
∂2U ∂uβ (R0 ) 2
uβ (R0 ). u(R0 )]. analogous to the band index n for Bloch electrons is introduced to label the diﬀerent branches. PHONONS. should be interpreted as the number of unit cells rather than the number of ions. since no states can be lost. 3. The upper band never approaches zero energy. so to excite these phonons high energies are required.k2 δλ1 . and one ﬁnds the appearance of 1 acoustic branch and (p-1) optical branches. The most notable diﬀerence is the appearance in 3D of polarization in analogy to what we have already seen for the photon ﬁeld. . . We treat the general case of any monatomic Bravais lattice. In the acoustical case the two types of ions oscillate in phase. 2
[bk1 λ1 . while in the optical case they oscillate π radians out of phase. The origin of the energy diﬀerence between an acoustical and an optical phonon at the same wave length is sketched in Fig. the reduced zone scheme in Fig. (3. . • and ◦. The ionic equilibrium positions are denoted R0 and the displacements by u(R0 ) with components j j uα (R0 ). The starting point of the analysis is a second order Taylor expansion j in uα (R0 ) of the potential energy U [u(R0 ). whereas for optical phonons the sign of the displacement alternates between the two types of ions.4: (a) An acoustical and (b) an optical phonon having the same wave length for a 1D system with two ions. b† 2 λ2 ] = δk1 . z.12)
(3.3(b) we of course ﬁnd two branches. It may very well go much beyond the nearest neighbor case studied in the 1D case. COUPLING TO ELECTRONS
¥ ¤ ¢ £
Figure 3.11) is changed into Hph =
kλ
ωkλ b† bkλ + kλ
1 . in Eq. The N appearing above. e. y.13)
. 2
u =0
Note that nothing has been assumed about the range of the potential. so they are known as optical phonons.4
Acoustical and optical phonons in 3D
The fundamental principles for constructing the second quantized phonon ﬁelds established for the 1D case carries over to the 3D case almost unchanged. A branch index λ. . k
3.4 for the case of a two-ion unit cell. The lower branch resembles the original dispersion so it corresponds to acoustic phonons.8). The generalization to p ions per unit cell is straight forward. In fact they can be excited by light.56
¥ ¤ ¢ ¡
CHAPTER 3. 3. (3.g.λ2 . α = x. so we have Nion = pN . per unit cell. For acoustical phonons the size of the displacement of neighboring ions diﬀers only slightly and the sign of it is the same.

R
D(R0 ) = 0. because P P then 0 = R1 R2 d · D(R0 −R0 ) · d = N d · [ 0 D(R0 )] · d.15) . that diagonalizes D(k). hence diagonalizable in an orthonormal basis. The second follows from the fact that U = 0 if all the displacements are the same. 2.16)
Using these symmetries in connection with D(k) we obtain D(k) =
R0
D(R0 ) e−ik·R = 1 2 D(R0 ) e
R0 ik·R0
0
1 2 +e
D(R0 ) e−ik·R +
R0 −ik·R0 R0
0
D(−R0 ) eik·R D(R0 ) sin2
0
=
− 2 = −2
R0
1 k·R0 (3. 2 1 2
(3.
u =0
Dαβ (k) =
R
Dαβ (R) e−ik·R . k. The D-matrix has the following three symmetry properties1 D(R0 )
t 0
= D(R0 ).
λ. the so-called dynamical matrix D(k) with components Dαβ (k): Dαβ (R0 −R0 ) = 1 2 ∂2U ∂uα (R0 ) ∂uβ (R0 ) 1 2 .2 .3. ACOUSTICAL AND OPTICAL PHONONS IN 3D the theory is the force strength matrix
∂2U ∂uα ∂uβ
57
(generalizing K from the 1D case) and its
Fourier transform. The classical equation of motions for the ions are simply M uα (R0 ) = − ¨ 1 ∂U ∂uα (R0 ) 1 ⇒ ¨ 1 −M u(R0 ) =
R0 2
D(R0 −R0 ) u(R0 ). the so-called polarization vectors.3 }.19)
Since D(k) is a real symmetric matrix there exists for any value of k an orthonormal basis set of vectors { k.14)
The discrete Fourier transform in 3D is a straight forward generalization of the one in 1D. The third follows from inversion symmetry 1 2 R always present in monatomic Bravais lattices.λ .
. they are eigenvectors: D(k)
1
kλ
= Kkλ
kλ .0
1 = N
(3.17) .
k∈FBZ N
0
δk. k. t) ∝ ei(k·R
0 −ωt)
⇒
M ω 2 = D(k) .
D(−R0 ) = D(R0 ). j 0
δ
R0 .
(3.
kλ · kλ
= δλ.e.
(3.4.
k∈FBZ N j=1
0 f (R0 ) e−ik·Rj .14). say d. 3. (3.0 j
=
1 N
eik·Rj . λ = 1. i.
e
j=1
−ik·R0 j
Due to the lattice periodicity Dαβ (R0 − R0 ) depends only on the diﬀerence between any 1 2 two ion positions.
(3.18)
We seek simple harmonic solutions to the problem and ﬁnd u(R0 . and for an arbitrary function f (R0 ) we have j 1 f (R0 ) ≡ √ j N 1 f (k) = √ N f (k) eik·Rj .1 . but arbitrary.
(3. 2
Thus D(k) is real and symmetric.20)
The ﬁrst follows from the interchangeability of the order of the diﬀerentiation in Eq.

58

¢ ¡

**CHAPTER 3. PHONONS; COUPLING TO ELECTRONS
**

¥ ¤ ¢ £

Figure 3.5: (a) Three examples of polarization in phonon modes: transverse, longitudinal and general. (b) A generic phonon spectrum for a system with 3 ions in the unit cell. The 9 modes divides into 3 acoustical and 6 optical modes. We have now found the classical eigenmodes ukλ of the 3D lattice vibrations characterized by the wavevector k and the polarization vector kλ : M ω2 = Kkλ ⇒ ukλ (R0 , t) = ei(k·R

0 −ω kλ t)

kλ

kλ

kλ

,

ωkλ ≡

Kkλ . M

**Using as in Eq. (3.10) the now familiar second quantization procedure of harmonic oscillators we obtain ukλ ≡ Hph =
**

kλ kλ √

1 b† + bk,λ 2 −k,λ ωkλ b† bkλ + kλ

kλ ,

kλ

≡

M ωkλ

,

1 , 2

[bkλ , b† λ ] = δk,k δλ,λ . k

Now, what about acoustical and optical phonons in 3D? It is clear from Eq. (3.17) that D(k) ∝ k 2 for k → 0, so the same holds true for its eigenvalues Kkλ . The dispersion relation in Eq. (3.19) therefore becomes ωkλ = vλ (θk , φk ) k, which describes an acoustical phonon with a sound velocity vλ (θk , φk ) in general depending on both the direction of k and the polarization λ. As in 1D the number of ions in the unit cell can be augmented from 1 to p. In that case it can be shown that of the resulting 3p modes 3 are acoustical and 3(p−1) optical modes. The acoustical modes are appearing because it is always possible to construct modes where all the ions have been given nearly the same displacement resulting in an arbitrarily low energy cost associated with such a deformation of the lattice. In Fig. 3.5 is shown the phonon modes for a unit cell with three ions. A 3D lattice with N unit cells each containing p ions, each of which can oscillate in 3 directions, is described by 3pN modes. In terms of phonon modes we end up with 3p so-called phonon branches ωkλ , which for each branch index λ are deﬁned in N discrete points in k-space. Thus in 3D the index λ contains information on both which polarization and which of the acoustical or optical modes we are dealing with.

¦

(3.21)

(3.22) (3.23)

3.5. THE SPECIFIC HEAT OF SOLIDS IN THE DEBYE MODEL

59

3.5

The speciﬁc heat of solids in the Debye model

Debye’s phonon model is a simple model, which describes the temperature dependence of the heat capacitance CV = ∂E of solids exceedingly well, although it contains just one ∂T material dependent free parameter. The phonon spectrum Fig. 3.5(b) in the reduced zone scheme has 3p branches. In Fig. 3.6(a) is shown the acoustic and optical phonon branch in the extended zone scheme for a 1D chain with two ions per unit cell. Note how the optical branch appears as an extension of the acoustical branch. In d dimensions a reasonable average of the spectrum can be obtained by representing all the phonon branches in the reduced zone scheme with d acoustical branches in the extended zone scheme, each with a linear dispersion relation ωkλ = vλ k. Furthermore, since we will use the model to calculate the speciﬁc heat by averaging over all modes, we can even employ a suitable average vD out the polarization dependent velocities vλ and use the same linear dispersion relation for all acoustical branches, ωkλ ≡ vD k ⇒ ε = vD k. (3.24)

Even though we have deformed the phonon spectrum we may not change the number of phonon modes. In the 3D Debye model we have 3Nion modes, in the form of 3 acoustic branches each with Nion allowed wavevectors, where Nion is the number of ions in the lattice. Since we are using periodic boundary conditions the counting of the allowed phonon wavevectors is equivalent to that of Sec. 2.1.2 for plane wave electron states, i.e. Nion = [V/(2π)3 ]×[volume in k-space]. Since the Debye spectrum Eq. (3.24) is isotropic in k-space, the Debye phonon modes must occupy a sphere in this space, i.e. all modes with |k| < kD , where kD is denoted the Debye wave number determined by Nion = V 4 3 πk . (2π)3 3 D (3.25)

Inserting kD into Eq. (3.24) yields the characteristic Debye energy, ωD and hence the characteristic Debye temperature TD : ωD ≡ kB TD ≡ vD kD ⇒ 6π 2 Nion ( vD )3 = V (kB TD )3 . (3.26)

Continuing the analogy with the electron case the density of phonon states Dph (ε) is found by combining Eq. (3.24) and Eq. (3.25) and multiplying by 3 for the number of acoustic branches, Nion (ε) = V 1 ε3 2 ( v )3 6π D ⇒ Dph (ε) = 3 dNion (ε) V 1 = 2 ε2 , dε 2π ( vD )3 0 < ε < kB TD .

(3.27) The energy Eion (T ) of the vibrating lattice is now easily computed using the Bose-Einstein distribution function nB (ε) Eq. (1.127) for the bosonic phonons: Eion (T ) =

kB TD 0

dε εDph (ε)nB (ε) =

V 3 2 ( v )3 2π D

kB TD 0

dε

ε3 . eβε − 1

(3.28)

60

CHAPTER 3. PHONONS; COUPLING TO ELECTRONS

Figure 3.6: (a) The linear Debye approximation to the phonon spectrum with the Debye wave vector kD shown. (b) Comparison between experiment and the Debye model of heat capacitance applied for lead, silver, aluminum, and diamond.

ion It is now straight forward to obtain CV from Eq. (3.28) by diﬀerentiation: ion CV (T ) =

∂Eion T = 9Nion kB ∂T TD

3 0

TD /T

dx

x 4 ex , (ex − 1)2

(3.29)

where the integrand is rendered dimensionless by introducing TD from Eq. (3.26). Note that TD is the only free parameter in the Debye model of heat capacitance; vD dropped out of the calculation. Note also how the model reproduces the classical Dulong-Petit value in the high temperature limit, where all oscillators are thermally excited. In the low temperature limits the oscillators “freeze out” and the heat capacitance drops as T 3 ,

ion CV (T ) −→ T TD

T 12π 4 Nion kB 5 TD

3

,

ion CV (T ) −→ 3Nion kB . T TD

(3.30)

In Fig. 3.6(b) the Debye model is compared to experiment. A remarkable agreement is obtained over the wide temperature range from 10 K to 1000 K after ﬁtting TD for each of the widely diﬀerent materials lead, aluminum, silver and diamond. We end this section by a historical remark. The very ﬁrst published application of quantum theory to a condensed matter problem was in fact Einstein’s work from 1906, reproduced in Fig. 3.7(a), explaining the main features of Weber’s 1875 measurements on diamond. In analogy with Planck’s quantization of the oscillators related to the black body radiation, Einstein quantized the oscillators corresponding to the lattice vibrations, assuming that all oscillators had the same frequency ωE . So instead of Eq. (3.27), Einstein E employed the much simpler Dion (ε) = δ(ε − ωE ), which immediately leads to

ion,E CV (T ) = 3Nion kB

TE T

2

eTE /T , (eTE /T − 1)2

TE ≡ ωE /kB .

(3.31)

While this theory also gives the classical result 3Nion kB in the high temperature limit, ion it exaggerates the decrease of CV at low temperatures by predicting an exponential suppression. In Fig. 3.7(b) is shown a comparison of Debye’s and Einstein’s models. Nowadays, Einstein’s formula is still in use, since it provides a fairly accurate description of the optical phonons which in many cases have a reasonably ﬂat dispersion relation.

¥

( % ¤2 9@5A B BBB

¦§¨© ¡ B B BB B #&78 3 " 456 B B B B B B B B B B ( % ¥2 # )1 " B B B B B B B B ( ¡¤¥ B # )0 " B B B B B B B ( ¥ ' &% #$! " BB BB B B B B B B B BB

¤

¡

£ ¢

E CF IP I Q

HG Q IP

E CD R

3.6. ELECTRON-PHONON INTERACTION IN THE LATTICE MODEL

61

Figure 3.7: (a) The ﬁrst application of quantum theory to condensed matter physics. Einstein’s 1906 theory of heat capacitance of solids. The theory is compared to Weber’s 1875 measurements on diamond. (b) A comparison between Debye’s and Einstein’s model.

3.6

Electron-phonon interaction in the lattice model

In Chap. 2 we mentioned that in the lattice model the electron-ion interaction splits in two terms, one arising from the static lattice and the other from the ionic vibrations, Vel−ion = Vel−latt + Vel−ph . The former has already been dealt with in the HBloch , so in this section the task is to derive the explicit second quantized form of the latter. Regarding the basis states for the combined electron and phonon system we are now in the situation discussed in Sec. 1.4.5. We will simply use the product states given in Eq. (1.108). Our point of departure is the simple expression for the Coulomb energy of an electron density in the electric potential Vion (r − Rj ) of an ion placed at the position Rj ,

N

Vel−ion =

dr (−e)ρel (r)

j=1

Vion (r − Rj ).

As before the actual ion coordinates are given by Rj = R0 + uj , where R0 are the ionic j j equilibrium positions, i.e. the static periodic lattice, and where uj denotes the lattice vibrations. The respective contributions from these two sets of coordinates are separated by a Taylor expansion, Vion (r − Rj ) ≈ Vion (r − R0 ) − r Vion (r − R0 ) · uj , note the sign j j of the second term, and we obtain

N N

Vel−ion =

dr (−e)ρel (r)

j=1

Vion (r−R0 )− j

dr (−e)ρel (r)

j=1

0 r Vion (r−Rj )·uj .

The ﬁrst term is the one entering HBloch in Eq. (2.6), while the second is the electronphonon interaction, also sketched in Fig. 3.8, Vel−ph = dr ρel (r)

j

e uj ·

r Vion (r

− R0 ) . j

Vel−ph is readily deﬁned in real space, but a lot easier to use in k-space, so we will proceed

¦ ¥¢ £ ¤ § $ !" # §

(3.32) (3.33) (3.34)

§ § § ¨ © ¢ ¡

62

CHAPTER 3. PHONONS; COUPLING TO ELECTRONS

Figure 3.8: (a) Being in an eigenstate a Bloch electron moves through a perfect lattice without being scattered. (b) A displaced ion results in an electric dipole relative to the perfect background, and this can scatter Bloch electrons from |k, σ to |k , σ . by Fourier transforming it. Let us begin with the ionic part, the u · V -term. The Fourier transform of uj is given in Eq. (3.22), where we note that the phonon wavevector k is 1 restricted to the Brillouin zone k ∈ FBZ. Deﬁning Vion (r) = V p Vp eip·r , we see that r simply brings down a factor ip. To facilitate comparison to the phonon wavevector k we decompose p as in Eq. (2.10): p = q + G, where q ∈ FBZ and G ∈ RL. All in all we have

r Vion (r

− R0 ) = j uj =

1 V

i(q + G)Vq+G ei(q+G )·(r−Rj ) ,

q∈FBZ G ∈RL kλ √ bk,λ + b† −k,λ 2 kλ

1 √ N

j

k∈FBZ λ

**These expressions, together with e uj ·
**

j r Vel (r

**eik·Rj = N δk,0 , and multiplying by e, lead to
**

i(q+G )·r gq,G ,λ bq,λ + b† , −q,λ e q∈FBZ G ∈RL,λ

− R0 ) = j

1 V

**where we have introduced the phonon coupling strength gq,G ,λ given by gq,G ,λ = ie N (q + G)· 2M ωqλ
**

qλ

The ﬁnal result, Vel−ph , is now obtained by inserting the Fourier representation of the 1 electron density, ρel (r) = V kpσ e−ip·r c† k+pσ ckσ , derived in Eq. (1.96), together with Eq. (3.37) into Eq. (3.34), and utilizing dr eik·r = Vδk,0 : Vel−ph = 1 V

† gq,G ,λ c† k+q+G ,σ ckσ bq,λ + b−q,λ . kσ qλ G

The interpretation of this formula is quite simple. Under momentum conservation (but only up to an undetermined reciprocal lattice vector due to the periodicity of the lattice) and spin conservation the electrons can be scattered from any initial state |k, σ el to the ﬁnal state |k + q + G, σ el either by absorbing a phonon from the state |qλ ph or by emitting a phonon into the state | − qλ ph . A graphical representation of this fundamental process is shown in Fig. 3.9.

¡¤¢¡ £ §¨ © ¦ ¥ © ¨ § ¦

0

¡ ¢ ¥ ¡¨ © § ¡ ¦

(3.35) (3.36)

eik·Rj .

0

(3.37)

Vq+G .

(3.38)

(3.39)

3.7. ELECTRON-PHONON INTERACTION IN THE JELLIUM MODEL

63

Figure 3.9: A graphical representation of the fundamental electron-phonon coupling. The electron states are represented by the straight lines, the phonon states by curly spring-like lines, and the coupling strength by a dot. To the left the electron is scattered by absorbing a phonon, to the right by emitting a phonon. The normal processes, i.e. processes where per deﬁnition G = 0, often tend to dominate over the so-called umklapp processes, where G = 0, so in the following we shall completely neglect the latter.2 Moreover, we shall treat only isotropic media, where qλ is either parallel to or perpendicular to q, i.e. q· qλ in gq,G =0,λ is only non-zero for longitudinally polarized phonons. So in the Isotropic case for Normal phonon processes we have

IN Vel−ph =

1 V

**† gq,λ c† k+q,σ ckσ bq,λ + b−q,λ . kσ qλl
**

l l l

Finally, the most signiﬁcant physics of the electron-phonon coupling can often be extracted from considering just the acoustical modes. Due to their low energies they are excited signiﬁcantly more than the high energy optical phonons at temperatures lower than the Debye temperature. Thus in the Isotropic case for Normal Acoustical phonon processes only the longitudinal acoustical branch enters and we have

INA Vel−ph =

1 V

† gq c† k+q,σ ckσ bq + b−q . kσ q

1 If we for ions with charge +Ze approximate Vq by a Yukawa potential, Vq = Ze q2 +k2 (see s 0 Exercise 1.5), the explicit form of the coupling constant gq is particularly simple:

gq = i

Ze2

0

q2

q 2 + ks

N . 2M ωq

3.7

Electron-phonon interaction in the jellium model

Finally, we return to the case of Einstein phonons in the jellium model treated in Sec. 3.1. The electron-phonon interaction in this case is derived in analogy with the that of normal

There are mainly two reasons why the umklapp processes often can be neglected: (1) Vq+G is small due to the 1/(q + G )2 dependence, and (2) At low temperatures the phase space available for umklapp processes is small.

2

¤ ¢ © ! ¢ © § ¦ ¤¥ £ § ¨ ¡ ¥¦ ¤ £ ¢ ¡

(3.40) (3.41) (3.42)

¤ ¢ ©

¢ © § ¦ ¤¥ £ § ¨ ¡ ¥¦ ¤ £ ¢ ¡

64

CHAPTER 3. PHONONS; COUPLING TO ELECTRONS

lattice phonons in the isotropic case, Eq. (3.41). If we as in Sec. 3.1 neglect the weak dispersion of the Einstein phonons and simply assume that they all vibrate with the ion plasma frequency Ω of Eq. (3.3), the result for N vibrating ions in the volume V is

jel Vel−ph =

1 V

† jel gq c† k+q,σ ckσ bq + b−q , kσ q

(3.43)

with

jel gq = i

Ze2 1 0 q

N . 2M Ω

(3.44)

3.8

Summary and outlook

In this chapter we have derived the second quantized form of the Hamiltonian of the isolated phonon system and the electron-phonon coupling. The solution of the phonon problem actually constitutes our ﬁrst solution of a real interacting many-particle system, each ion is coupled to its neighbors. Also the treatment of the electron-phonon coupling marks an important step forward: here we dealt for the ﬁrst time with the coupling between to diﬀerent kinds of particles, electrons and phonons. The electron-phonon coupling is a very important mechanism in condensed matter systems. It is the cause of a large part of electrical resistivity in metals and semiconductors, and it also plays a major role in studies of heat transport. In Exercise 3.1 and Exercise 3.2 give a ﬁrst hint at how the electron-phonon coupling leads to a scattering or relaxation time for electrons. We shall return to the electron-phonon coupling in Chap. 16, and there see the ﬁrst hint of the remarkable interplay between electrons and phonons that lies at the heart of the understanding of conventional superconductivity. The very successful microscopic theory of superconductivity, the so-called BCS theory, is based on the electron-phonon scattering, even the simple form given in Eqs. (3.41) and (3.42) suﬃces to cause superconductivity.

the electron gas. such as e. 65
£ ¤£ £ £ £ £ ££ £ £ ¢
¡
. which means that the eﬀect of the other particles is included as a mean density (or mean ﬁeld). There exist numerous examples of the success of the mean ﬁeld method and its ability
Figure 4.1: Illustration of the mean ﬁeld idea. which in turn ensure that the method is consistent. This idea is illustrated in Fig. and one talks about a “correlation hole”. Upon performing the mean ﬁeld approximation we can neglect the detailed dynamics and the time-independent second quantization method described in Chap. which is soluble.g. In these cases it suﬃces to include correlations “on the average”.Chapter 4
Mean ﬁeld theory
The physics of interacting particles is often very complicated because the motions of the individual particles depend on the position of all the others. To the right are the interactions felt by the black particle replaced by an average interaction due to the average density of the white particles. 1 suﬃces. leaving an eﬀective single particle problem. There we expect the probability to ﬁnd two electrons in close proximity to be small due to the strong repulsive interaction. due to these correlation eﬀects there is a suppressed density in the neighborhood of every electron. not fully including the correlations.1. The mean ﬁelds are chosen as those which minimize the free energy. or in other words the particles motions become correlated. 4. Consequently. This approximation scheme is called “mean ﬁeld theory”. Left box shows the real physical system where the interaction leads to correlation between the particle motions. This is clearly the case for a system of charged particles interacting by Coulomb forces. as we shall see shortly. gives a good physical model. Nevertheless. in spite of this complicated problem there are a number of cases where a more crude treatment.

which gives
neglected in mean ﬁeld
(4.µµ
Vνµ.1a). In terms of the new basis P † the Hamiltonian then becomes H = i αi i αi .
(4.
1
.ν µ dνν eµµ . a† aν and b† bµ .µµ
Vνµ. such that h is diagonal.
H = H0 + VMF +
νν .5)
The mean ﬁeld Hamiltonian HMF contains only single-particle operators.3) is neglected.2a) (4.µµ
Vνµ.1c)
Vint =
νν . kk Uik hkk Uk j = δij i .µµ
Vνµ. but before going to speciﬁc examples let us discuss the mathematical structure of the mean ﬁeld theory. that the density operators a† aν and b† bµ deviate only little from their average values. In order to do so we deﬁne the deviation operators dνν = a† aν − a† aν . (4. we shall discuss a few examples from condensed matter physics .ν µ a† b† bµ aν . (4. ν ν eµµ = and insert them into (4. The Hamiltonian is H = H0 + Vint . based on physical arguments. and the interaction Vint is approximated by the mean ﬁeld interaction VMF resulting in the so-called mean ﬁeld Hamiltonian HMF given by HMF = H0 + VMF .ν µ a† aν ν
b† bµ .1a) (4. First we consider a system with two kinds of particles. It is then ν µ ν µ natural to use this deviation as a small parameter and perform an expansion.1
To see that a single-particle problem or a quadratic Hamiltonian can always be diagonalized. MEAN FIELD THEORY
to explain various physical phenomena. Let us assume that only interactions between diﬀerent kind of particles are important.2b)
b† bµ µ
−
b† bµ µ
. and { i } are thus the eigenvalues of h.ν µ
a† aν b† bµ + b† bµ a† aν ν µ µ ν
−
νν . write H P P as H = ij a† hij aj . described by operators aν and bµ . H0 =
ν a ξν a† aν + ν µ b ξµ b† bµ .1b) (4.66
CHAPTER 4. and thus the original many-body problem has been reduced to a single-particle problem.3)
where VMF =
νν . αi = j Uij aj . which in principle is always soluble. respectively.4) µ
Because dνν and eµµ are assumed to be small the second term in Eq. with i P † U being a unitary matrix. Now since h is a hermitian matrix there exists a transformation. ν µ
Now suppose that we expect. (4. In this chapter. µ
(4.

(4. 1. Using the expression for the free energy given in Sec.7)
The question is however how to ﬁnd the averages a† aν and b† bµ .9). Thus the two methods are equivalent. which themselves depend on n Next we turn to the alternative route. There are two µ ν possible routes which in fact are equivalent.8a) (4. Method 1: The average is to be determined self-consistently. we get 0= d d 1 FMF = − ln ZMF d¯ a nνν d¯ a nνν β 1 d = Tr e−βHMF a HMF ZMF d¯ νν n 1 Tr e−βHMF Vνµ.4) we can formulate the mean ﬁeld procedure in a diﬀerent way: If we have an interaction term involving two operators A and B given by a product of the two HAB = AB.10)
Eq.ν µ b† bµ − nb = ¯ µµ µ ZMF
µµ
=
µµ
Vνµ.8b)
using the new mean ﬁeld Hamiltonian.11)
This should hold for any pair (ν. when calculating the averages na ≡ a† aν . (4. (4. (4. and n ¯ ¯ ¯ MF and ZMF . i.
.67 Looking at Eq.
(4. (4. Similarly by minimizing with respect to nb ¯ ¯ we get Eq. This means for na (and similarly for nb ) that ¯ ¯ na ≡ a† aν ¯ νν ν
MF
=
1 ZMF
Tr e−βHMF a† aν ν
.
(4.9)
where ZMF is the mean ﬁeld partition function given by ZMF = Tr e−βHMF .ν µ
b† bµ µ
MF
− nb ¯ µµ
.9) and the similar one for nb are called the self-consistency equations (because na ¯ ¯ b are given in terms of H a and nb ). ν ) and hence the last parenthesis has to vanish and we arrive at the self-consistency equation for nb . ¯ νν ν nb ¯ µµ ≡ b† bµ µ . the same answer should come out.e. (4.5.
(4.6) then the mean ﬁeld approximation is given by A coupled to the mean ﬁeld of B plus B coupled to the mean ﬁeld of A and ﬁnally to avoid double counting subtracted by the MF product of the mean ﬁelds (such that HAB = A B ):
MF HAB = A B + A B − A B . Method 2: Use the nνν that minimizes the free energy FMF of the mean ﬁeld Hamiltonian.

we then have c† ck = k 1 V dr dr e−ik ·r eik·r Ψ† (r)Ψ(r ) . This is in essence the approximation done in the mean ﬁeld approach. MEAN FIELD THEORY
We can gain some more understanding of the physical content of the mean ﬁeld approximation if we look at average interaction energy Vint . HMF = HMF + HMF . Here “less important” should be quantiﬁed by checking the validity of the mean ﬁeld approximation. Suppose for example that the Hamiltonian that we are interested in has translational symmetry.
4. Therefore. and then compare the result to a† aν . one must provide some physical insight to reduce the number of mean ﬁeld parameters.
. one has ν either chosen the wrong mean ﬁeld parameter.12)
which is equivalent to assuming that the a and b particles are uncorrelated2 .15)
2 Remember from usual statistics that the correlation function between two stochastic quantities X and Y is deﬁned by XY − X Y .3) as a perturbation. µ
(4.12). one should check that d indeed is small by calculating d . using the neglected term in (4. Often symmetry arguments can help reducing the number of parameters. the average factorizes exactly as in Eq. If it is not small.68
CHAPTER 4. Vint ≈
νν .µµ
Vνµ. With modern computers one can treat hundreds of particles in this way.µ ν a† aν ν
b† bµ . and we get Vint
MF
=
1
Vint
MF
=
νν .ν µ a† aν ν
MF
b† bµ µ
MF .
(4. or the method simply fails and other tools more adequate to deal with the problem at hand must be applied. Suppose we have N diﬀerent quantum numbers.1
The art of mean ﬁeld theory
In practice one has to assume something about the averages a† aν and b† bµ because ν µ even though (4. such that momentum space is a natural choice. That is. A natural approximation would be to evaluate the expectation value of a and b operator separately. which gives N 2 coupled non-linear equations. (4. then there are in principle N 2 diﬀerent combinations. (4. it is out of the question. To see this let us evaluate Vint using the mean ﬁeld Hamiltonian Tr e−βHMF Vint . which of course is only tractable for small systems.9) gives a recipe on how to ﬁnd which averages are important.µµ
Vνµ. (4. but for a condensed matter system. For a system of particles described by operators c and c† . there are simply too many possible combinations.13) ZMF Because the mean ﬁeld Hamiltonian can be separated into a part containing only aa b operators and a part containing only b-operators.14)
The mean ﬁeld approach hence provides a consistent and physically sensible method to study interacting systems where correlations are less important.

17) leading to the possibility of c† ck+Q being k ﬁnite. r + R) (4. namely that Ψ† (r)Ψ(r ) = h(r. Here we go on to formulate the method for like particles.20)
. HARTREE–FOCK APPROXIMATION It is natural to assume that the system is homogeneous.ν µ c† c† cµ cν .2
Hartree–Fock approximation
Above we discussed the mean ﬁeld theory for interactions between diﬀerent particles.19)
is large only when the average ρµµ is non-zero. For most of the combinations νν the average ρνν is zero.1a) we use the approximation to replace a† aν and b† bµ by their average values plus small ν µ corrections.18a) (4. which means Ψ† (r)Ψ(r ) = f (r − r ) ⇒ c† ck = nk δk.16) from line one.17)
with R being a lattice vector. We therefore use the same strategy as in the introduction and write the four operators in the interaction term in terms of a deviation from the average value c† c† cµ − c† cµ ν µ µ cν + c† cν c† cµ .µµ
Vint =
The basic idea behind the mean ﬁeld theory was that the operator ρµµ = c† cµ . because in some cases the symmetry of the system is lower than that of the Hamiltonian. Suppose we have a system of interacting particles described by the Hamiltonian H = H0 + Vint .18b) (4. (4. does not exhaust the possibilities and only includes the so-called Hartree term and now we discuss the more general approximation scheme. however.16)
This assumption about homogeneity is however not always true.4. For interactions between identical particles this. r ) = h(r + R. r ). ν 1 2 Vνµ. (4. h(r.18c)
ξν c† cν . H0 =
ν
(4. For the interaction term in Eq. µ (4.2. like a wave formation then the average Ψ† (r)Ψ(r ) is not function of the r − r only. Instead it has a lower and more restricted symmetry. called the Hartree–Fock approximation. Thus the choice of the proper mean ﬁeld parameters requires physical motivation about which possible states one expects. where R · Q = 2π. (4. ν µ (4. For example if the system spontaneously orders into a state with a spatial density variation. k
69
(4. The kind of crystal structure of course exists in Nature and when it happens we talk about phenomena with broken symmetry. ν µ
νν . Instead we should have started by assuming Eq. It is important to realize that this solution cannot be found if we assumed Eq.
4.k .

and these assumptions must be based on physical knowledge or clever guess-work.ν µ nνν nµµ .ν µ nµν c† cµ + ¯ ν
1 2
Vνµ. This is true except in a set of measure zero. kσ
(4. which gives c† cν − c† cν ν ν c† cµ − c† cµ µ µ + c† cν c† cµ + c† cµ c† cν − c† cµ ν µ µ ν µ c† cν . which means that the expectation value c† ck is diagonal.ν µ nµµ c† cν + ¯ ν
1 2
Vνµ.25a)
V (0) − δσσ V (k − k ) nk σ . indistinguishable particles and there is therefore one combination we is missed be the Hartree approximation. V (k − k )nk σ . We can now read oﬀ the k corresponding Hartree–Fock Hamiltonian. (4. The result is H HF =
kσ HF ξk = ξk + kσ HF ξk c† ckσ . which is proportional to the deviation squared. The third term is the exchange correction.21) ν
If we neglect the ﬁrst term. (1. The mean ﬁeld result for the Fock term thus µ becomes
Fock Vint = −
1 2
Vνµ.22) ¯ ¯
This is the same result we would get if we considered the operators with (ν.ν µ nνµ nµν .ν µ nνµ c† cν − ¯ µ
1 2
Vνµ.18c) also gives a large contribution when c† cµ is ﬁnite.1 from Eq. (4. 2 in condensed matter systems it is normally cancelled out by an equally large term due to the positively charged ionic background. see Exercise 4. namely the so-called exchange or Fock term.7) ν do the same with the combination c† cν .24)
Consider now the example of a homogeneous electron gas which is translation invariant.25b) we assumed for example that the spin
. However. (4.104).25b)
The second term is the interaction with the average electron charge. (4. In deriving Eq. As explained in Chap. This new term appears because the product of four operators in Eq.
kσ
= ξk + V (0)N −
(4. With this assumption we again write c† cν as ν its average value plus a deviation. Again we emphasize that the Hartree–Fock approximation depends crucially on what averages we assume to be ﬁnite. we have arrived at the Hartree approximation for interactions
Hartree Vint =
1 2
Vνµ. this is not the full result because here the operators labelled by µ and ν represent identical.ν µ nνν c† cµ − ¯ µ
1 2
Vνµ. µ ) to be diﬀerent kinds of particles and used the formula from the previous section.23) ¯ ¯
The ﬁnal mean ﬁeld Hamiltonian within the Hartree–Fock approximation is
Fock Hartree H HF = H0 + Vint + Vint
(4. (4. (4. ν ) and (µ. MEAN FIELD THEORY
If the quantum number ν is diﬀerent from µ we can commute cν with the parenthesis. To derive the mean ﬁeld contribution from this possibility we thus ν have to ﬁrst replace c† cµ by its average value and following the principle in Eq.70
CHAPTER 4.

2. This is discussed in Sec.26)
(4. which implies that c† ck↓ = c† ck↑ . Since the operator and the Hamiltonian commute according to the theory of Hermitian operators a common set of eigenstates exists. For a given Hamiltonian with some symmetry (e.2: The energetics of a phase transition. rotational symmetry in real space or in spin space) there exists an operator which reﬂects this symmetry and therefore commutes with the Hamiltonian (e. Below the critical point a two well potential develops and the system has to choose one of the two possibilities.27)
.
4.3. Here T (R) is an operator which displaces all particle coordinates by the amount R. translational symmetry. This fact we can use to “prove” the unphysical result that a density wave can never exist. rotation operator in real space or spin space). 4. The total momentum operator is thus a conserved quantity and it is given by P=
kσ
k c† ckσ . If we allow them k↓ k↑ to be diﬀerent we have the possibility of obtaining a ferromagnetic solution. BROKEN SYMMETRY
71
Figure 4. kσ
We can now choose an orthogonal basis of states with deﬁnite total momentum. symmetry is also maintained.g. kσ
¡
(4. A density wave. Above the critical point the eﬀective potential has a well-deﬁned minimum at the symmetry point. with wave vector Q. translation operator. [H. which means that the translational operator T (R) commutes with Hamiltonian. which indeed happens in some cases.g. T (R)] = 0.4. means that the Fourier transform of the density operator ρ(Q) =
kσ
c† ck+Qσ . Even though the total potential is still symmetric the system will reside only in one well due to the macroscopically large energy barrier and thus the state of the system has “lower symmetry” than the potential. It can be written as T = exp(iR · P). |P .3
Broken symmetry
Mean ﬁeld theory is often used to study phase transitions and thus changes of symmetry. where P is the total momentum operator. Consider for example the case of a liquid of particles where the Hamiltonian of course has translational symmetry.4. and the system is in a state of large symmetry.

superconductivity. The quantity which signals that a phase transition has occurred is called the order parameter. 4. Typical examples are listed in Table 4. also discussed in Sec. they are eﬀectively decoupled due to the large energy barrier it takes to melt and then recrystallize into a new state with a shifted reference (or direction of magnetization). it does not make sense to include them on equal footing in the statistical average as in Eq.
. In those cases where many states of the system are degenerate but separated by large energy barriers. but c† ck+Qσ = kσ 1 Z e−βEP P c† ck+Qσ P = 0.28)
because c† ck+Q |P has momentum P − Q and is thus orthogonal to |P . We are therefore forced to refrain from the fundamental ergodicity postulate of statistical mechanics.5. In the mean ﬁeld approach the trick is to include the order parameter in the choice of ﬁnite mean ﬁelds and.72 Phenomena crystal ferromagnet Bose-Einstein condensate superconductor
CHAPTER 4. one has to built in the possibility of the new phase into the theory.1. In the same way. 1. Even though crystals with diﬀerent spatial reference points (or ferromagnets with magnetization in diﬀerent directions) have formally the same energy. Next we study in some detail examples of symmetry breaking phenomena and their corresponding order parameters. show that the resulting mean ﬁeld Hamiltonian leads to a self-consistent ﬁnite result. has a ﬁnite expectation value.2. When at some critical temperature the thermodynamical state of the system develops a non-zero expectation value of some macroscopic quantity which has a symmetry lower than the original Hamiltonian it is called spontaneous breaking of symmetry.28). What is wrong? The proof above breaks down if the sum of states in the thermodynamical average is restricted. of course. MEAN FIELD THEORY Order parameter physical density wave magnetization population of k = 0 state pair condensate Order parameter mathematical
k
c† ck+Q k c† ck↑ − c† ck↓ k↑ k↓ a† k=0 ck↑ c−k↓
k
Table 4. We have therek fore reached the senseless result that crystals do not exist. In order to arrive at the new phase in a calculation and to avoid the paradox in Eq. (4. This is often illustrated by the double barrier model of phase transitions shown in Fig. kσ
P
(4.28) because they correspond to macroscopically totally diﬀerent conﬁgurations.1: Typical examples of spontaneous symmetry breakings and their corresponding order parameters. and built into the description that the phase space of the system falls into physically separated sections. (4. and other well-known physical phenomena cannot happen. we could “prove” that magnetism.

It depends only on the distance between the ions. Here we consider only the ferromagnetic case. The model discussed here includes interactions between adjacent spins only. The mean ﬁeld decomposition then gives H ≈ HMF = −2
ij
Jij Si · Sj − 2
ij
Jij Si · Sj + 2
ij
3 The term “eﬀective Hamiltonian” has a well-deﬁned meaning. as shown in the right panel. However. It reads H = −2
ij
Jij Si · Sj . (4.
(4.
where Si is the spin operator for the ion on site i and Jij is the strength of the interaction. The ﬁrst case corresponds to the antiferromagnetic case. It means the Hamiltonian describing the important degrees of freedom on the relevant low energy scale. the spins of the individual ions being strongly correlated. between the magnetic moment of the ions on sites i and j. known as the Heisenberg model for interaction between spins in a crystal. while the latter to the ferromagnetic case. while in the ferromagnetic state.30)
We immediately see that if J < 0.4
4. 4. although simple looking.
4. the spins tend to become antiparallel whereas for J > 0.29)
Jij Si · Sj .4. The interaction is generally short ranged and we truncate it so that it. As the model Hamiltonian stands.31)
. shown in the middle. as in Fig.4.1
Ferromagnetism
The Heisenberg model of ionic ferromagnets
In ionic magnetic crystals the interaction between the magnetic ions is due to the exchange interactions originating from the Coulomb interactions.
& % $0 % & () '¡"&% $ # ¡¡ ¡ ¡¡
"! £ ¡ ¡¡ £ £¡ £ £ £¡£¡£ £
¥ §¥ ¢¡ ¤ ¥ ¨ ¦ ¢ ¡ ¥ £ © © © ¢ ¡ ¥© §¥
(4.3: The left ﬁgure shows the Heisenberg model in the disordered state where there is no preferred direction for spins. it is a good example where a mean ﬁeld solution gives an easy and also physical correct answer. J > 0. is only non-zero for nearest neighbors Jij = J0 if i and j are neighbors. FERROMAGNETISM
73
Figure 4. the spins form a collective state with a ﬁnite macroscopic moment along one direction. 0 otherwise.4. it is immensely complicated and cannot be solved in general. it is energetically favorable to for the spins to be parallel.3. Here we will not go into the details of this interaction but simply give the eﬀective Hamiltonian3 .

35)
with one term for each possible spin projection.34)
is now diagonal in the site index and hence easily solved. nJ0 b = nJ0 β. because all directions are equivalent. (4.37)
that there is no solution for b < 1. (4. because of the translational symmetry we expect it to be independent of position coordinate i.
(4.74
CHAPTER 4. MEAN FIELD THEORY
Here Si is the average spin at site i.e. Suppose for simplicity that the ions have spin S = 2 .
But had we reasons to believe that an antiferromagnetic solution (where the spins point in opposite direction on even and odd sites) was relevant (if Jij < 0). (4.33)
j
where n is the number of neighbors. With this simpliﬁcation the mean ﬁeld partition function is ZMF = eβm + e−βm
N
eβN m
Sz
=
eβm + e−βm eβm
2 /2nJ 0
N
. If we choose the z axis along the direction of the magnetization our mean ﬁeld assumption is Si = Sz ez .4 So we assume a ﬁnite but spatial independent average spin polarization. 2 The self-consistency equation is found by minimizing the free energy ∂FMF 1 ∂ =− ln ZMF = −N ∂m β ∂m eβm − e−βm eβm + e−βm +N m = 0.
4
. where d is the dimension. 1 α ≈ bα − (bα)3 .
(4.32) and the magnetic moment m (which by assumption is equal for all sites) felt by each spin thus becomes m=2 Jij Sz ez = 2nJ0 Sz ez . From symmetry arguments we would then expect that the expectation value of this is zero. For a square lattice it is n = 2d. 3 (4. nJ0
which has a solution given by the transcendental equation α = tanh (bα) . The mean ﬁeld Hamiltonian HMF = −2
i
m · Si + mN Sz . i. we would have to assume that also this symmetry was broken. Sz = ± 1 . allow for Si to be non-zero. Here N is the number of sites and 1 m = |m|. α= m . and thus we can determine the critical temperature Tc where the magnetism disappears. Furthermore. But since this is not the right answer we have to assume that the symmetry is broken.36)
It is evident from an expansion for small α. by the condition bc = 1 and hence kB Tc = nJ0 .

g. α ≈
T −T 3 CC . This is simply because the spins are not localized.38)
¦ ¥¡ ¤ ¢£ ¡ ©¡ § ¨ ¦ ¥¡ ¤ ¢£ ¡
.4. At T = 0 where t = ∞ the solution is α = 1 and hence m ≈ nJ0 m = nJ0 . for small α we ﬁnd the solution for the magnetization.36) numerically. Eq. Tc = nJ0 . 4. Hhub =
kσ
ξk c† ckσ + kσ
U 2V
† c† k+qσ ck −qσ ck σ ckσ . For the functional form of the magnetization in the entire range of temperature one must solve Eq. The resulting temperature dependence of the magnetization is shown to the right. FERROMAGNETISM
75
Figure 4. the so-called Hubbard model.4. Metallic magnetism happens e.36) .4: Left two panels show the graphical solution of the mean ﬁeld equation for the Heisenberg model.
4.σσ
We now use the Hartree–Fock approximation scheme on this model but search for a ferromagnetic solution by allowing for the expectation values to depend on the direction of
3(b − 1)/b ⇒ (4.4. Furthermore. corresponding to b = nJ0 /T < 1 there is no solution and hence no ferromagnetic moment. T 1 b
valid close to Tc . The solution is shown in Fig. At temperatures larger than the critical temperature. Typical metals where correlations between conduction band electrons are important are Fe and Ni. The interaction between two particles in those orbitals is stronger than between electrons occupying the more spread out s or p orbitals and hence give a larger correlation between electrons. where this physical fact is reﬂected in a simple but extreme manner: the Coulomb interaction between electrons is taken to be point-like in real space and hence constant in momentum space. (4.4. which of course is a simple task. Since the short range of the interaction is important it is relevant to study a model.2
The Stoner model of metallic ferromagnets
In magnets where the electrons both generate the magnetic moments and also form conduction bands the Heisenberg model cannot explain the magnetism. in transition metals where the conduction bands are formed by the narrower d or f orbitals. (4. For b > 1 a solution exists. k kq.

46)
ζ=
n↑ − n↓ ¯ ¯ n ¯
. ¯
(4.
n = n↑ + n↓ .44) kσ V V
k k
We obtain at zero temperature n↑ = ¯ where are
2 2 ¯ 2m kF ↑ +U n↓
2 k2 dk θ µ− − U n↓ ¯ (2π)3 2m
=
1 3 k . The two equations
2/3 2
2
2m Deﬁne the variables.43a) (4.42)
The full mean ﬁeld Hamiltonian is now given by HMF =
kσ MF ξkσ c† ckσ − kσ
UV 2
nσ nσ + ¯ ¯
σσ
UV 2
n2 . ¯
2m 2mU n1/3 (3π 2 )2/3
2
(6π)2/3 n↓ ¯
2/3
+ U n↑ = µ. ¯ k↓
(4. Using our mean ﬁeld assumptions Eq.39)
and the mean ﬁeld interaction Hamiltonian becomes
MF Vint =
U V
† c† k+qσ ck −qσ ck σ ckσ − k kqσσ
U V
† c† k+qσ ck σ ck −qσ ckσ k kqσσ
U − 2V
† † c† k+qσ ckσ ck −qσ ck σ − ck+qσ ck σ k kqσσ
c† −qσ ckσ k
. (4. ¯ k↑
CHAPTER 4.
(6π)2/3 n↑ ¯
+ U n↓ = µ. ¯ ¯ ¯¯
The mean ﬁeld solution is found by minimization.45)
~
= µ. and of course a similar equation for spin down.47)
.23). (4. 6π 2 F ↑
(4. ¯ ¯ ¯
(4.39) we obtain MF Vint = U kσσ
c† ckσ [¯ σ − nσ δσσ ] − U V n ¯ kσ
σσ
nσ nσ + U V ¯ ¯
σ
n2 .40)
1 The factor 2 disappeared because there are two terms as in Eqs. ¯σ
(4.
γ=
. MEAN FIELD THEORY
c† ck ↓ = δkk nk↓ .22) and (4.76 the spin. (4.41)
where the spin densities have been deﬁned as nσ = ¯ 1 V c† ckσ .43b)
M ξkσF = ξk + U (n↑ + n↓ − nσ ) = ξk + U nσ . ¯σ
σ
(4. The mean ﬁeld parameters are c† ck ↑ = δkk nk↑ . which gives the self-consistency equations 1 1 MF nσ = ¯ c† ckσ M F = nF (ξkσ ). kσ
k
(4.
(4.

This expression has three types of solutions: γ<
4 3 4 3
:
Isotropic solution: ζ = 0
Normal state (4. Thus the interaction energy is lowered by having a polarized ground state. whereas the energy does not depend on the occupation of the same spin direction.4. The polarization is thus a function of the interaction strength. the stronger the interaction the larger the polarization. we get n↑ ¯
2/3
− n↓ ¯
2/3
=
2mU
2
(6π)−2/3 n↑ − n↓ .43a).48)
γζ = (1 + ζ)2/3 − (1 − ζ)2/3 . where it is clear that the mean ﬁeld energy of a given spin direction depends on the occupation of the opposite spin direction. which on the other hand for a ﬁxed density costs kinetic energy. This resulted from two things: the short rang interaction and the exchange term. FERROMAGNETISM
77
Figure 4.5. The Stoner model provides a clear physical picture for how the exchange interactions induce a ferromagnetic phase transition in a metal with strong on-site interactions. if the interaction is short-range. This
765
&' % ! # 2 )§ ¨)1 § 0 4 3 8 9
¥¤¢£ ¡
76 5
' & % $! "! # 2 )§ ¨)1 § 0 ) ¨ ( 8 9
¥¤¢£ ¡ 76 5 8
¨© § ¦ 9
¥¤¢£ ¡
. (4.4. The Stoner model gives a reasonable account of metallic magnets and it is also capable of qualitatively explaining the properties of excitations in the spin polarized states.5: The three possible solutions of the Stoner model. they cannot interact. Then by subtracting the self-consistency conditions (4.49)
< γ < 22/3 : Partial polarization: 0 < ζ < 1 Weak ferromagnet γ> 22/3 : Full polarization ζ = 1 Strong ferromagnet
The diﬀerent solutions are sketched in Fig. (4.46). 4. The competition between the potential and the kinetic energy contributions is what gives rise to the phase transition. One can understand this simply from the Pauli principle which ensures that electrons with the same spin never occupy the same spatial orbital and therefore. This leaves interactions between opposite spin as the only possibility. The possibility for a magnetic solution can be traced back to the spin-dependent energies Eq.

In this section. which means that we can give all electrons the same extra constant phase and still leave the Hamiltonian invariant. the relevant symmetry is the global gauge symmetry. While it is clear why the ﬁnite expectation S gives a magnetization in the case of a ferromagnet. It is probably the most successful theory in condensed matter physics and it has found application in other branches of physics as well. e. Of course.1
Breaking of global gauge symmetry and its consequences
Let us start by understanding what kind of broken symmetry could give rise to a superconducting state. which result from the appearance of a gap in the excitation spectrum. In the same way the phase rotation also changes the superconducting order parameter. Schematically the analogies between superconductors and ferromagnets are as shown in Table 4.g. because it will change the direction of the magnetization. namely loss of global gauge invariance.
4. In that case a broken SO(3) symmetry means that the expectation value S is not invariant under the rotation. distinct single particle properties. but this is not a problem here where we discuss only superconductors connected to electron reservoirs. As was said above. which is of the form cν cν . Below the critical temperature the metal which turns superconducting has no resistance what so ever. we give a short introduction to the BCS theory. which means that magnetic ﬁelds are totally expelled from the interior of the material. to explain the stability of nuclei with an even number of nucleons and also in the standard model of high-energy physics.5. e.78
CHAPTER 4. Besides the superconductivity itself and the Meissner eﬀect superconductors show a number of other characteristics.
4. the number of particles is conserved.g.2. The analog to this in the case of a ferromagnet. These astonishing phenomena result from the superconducting state having a new form of symmetry breaking. which in its spirit is very much like the Hartree–Fock theory presented above.5
Superconductivity
One of the most striking examples of symmetry breaking is the superconducting phase transition. which therefore has a global SO(3) symmetry. and it exhibits perfect diamagnetism (called Meissner eﬀect). MEAN FIELD THEORY
is however outside the scope of this section and the interested reader should consult for example the book by Yosida. but it is also clear that a constant phase cannot have any measurable eﬀect because all expectation values are given by the absolute square
. It is a mean ﬁeld theory but with a quite unusual mean ﬁeld and therefore we begin by discussing the nature of the superconducting phase. Both the new type of phase and the appearance of an excitation gap are explained by the Bardeen-Cooper-Schrieﬀer (BCS) theory from 1957. is that all spins can be rotated by some angle without changing the Hamiltonian. it is not so clear why broken symmetry in the superconducting case leads to a system without resistance. The order parameter for the superconductor thus involves an expectation value of two annihilation operators. We have argued that the superconducting state is sensitive to a change of global phase.

4. but varies extremely slowly.
†
These equations follow from the diﬀerential equation (together with the boundary condi˜ tion Ψϕ=0 = Ψ) δ ˜ ˜ Ψ(r) = iU [ρ(r ).24) in Chap. U would have disappeared all together. Let us therefore assume that we ascribe a phase to the superconducting state which depends on position.51a) (4. For any other non-superconducting system it would not make sense to talk about a quantum mechanical phase diﬀerence over macroscopic distances. maybe of the order of 10 nm or less in metals.50)
(recall that ρ. we get
˜ ˜ Z = Tr U e−βH U −1 = Tr e−β H . superconductor cν → cν eiϕ ⇒ H → H Broken symmetry: 0 = cν cν → cν cν e2iϕ
79 Global SO(3). (5.
(4.5.52)
(See also Eq.2: The analogy between the broken symmetries of a superconductor and a ferromagnet. is the density operator ρ (r) = Ψ† (r) Ψ (r)) because it has the following properties when applied to quantum ﬁeld operators ˜ Ψ(r) = U Ψ(r)U −1 = exp i = Ψ(r) exp (−iϕ (r)) . δϕ (r ) (4. However. of the wave function. SUPERCONDUCTIVITY Global U (1) gauge symmetry.53)
Note here that if we had used the cyclic properties of the trace.) If we transform the density operator and calculate the transformed partition function. 5 which is derived in the same way.51b)
= Ψ (r) exp (iϕ (r)) . we learned above that when dealing with systems with broken
. ϕ (r). phase gradients can have an eﬀect. In order to study the consequences of the phase change let us introduce a unitary transformation which changes the phase U = exp i dr ρ(r)ϕ (r) . such that it takes a macroscopic distance to see any signiﬁcant changes in ϕ (r). However. ferromagnet S → US ⇒ H → H Broken symmetry: 0 = S → U S
Table 4. To argue that the superconducting state depends on phase diﬀerences over macroscopic distances is therefore very unusual and tells us that superconductivity is a macroscopic quantum phenomenon. Ψ (r) = U Ψ (r)U ˜†
† −1
dr ρ(r)ϕ (r) Ψ(r) exp −i
dr ρ(r)ϕ (r) (4. Ψ(r)]U −1 = iΨ(r)δ r − r . (4. simply because quantum coherence is destroyed by all sorts of scattering events on rather short length scales.

as we have argued. Let us therefore minimize the free energy with respect to the phase in order to ﬁnd the condition for the lowest free energy.55) δ ϕ me and hence the energy is minimized if it carries a current. even in equilibrium. and then the superconductor goes into the normal state. me (4. (4. the sum-over-states has to be restricted so that the cyclic properties does not necessarily hold.56)
The meaning of this result is that by forcing a phase gradient onto the system it minimizes its energy by carrying a current even in thermodynamical equilibrium. (4. due to the lack of gauge invariance. This we have anticipated by the symbol Tr which means that the trace is restricted due to the spontaneous symmetry breaking.54) that the energy doesn’t depend on ϕ itself.51) is unchanged under the phase shift transformation. It is clear from (4. The claim above was that contrary to the non-superconducting state in the superconductor the phase is a macroscopic quantity.56) should only be the electrons participating in the superconducting condensate. but as long as this cost is smaller than the alternative which is to go out of the superconducting state. and in fact gapless superconductors do exist. The transformation only changes in the kinetic energy term of the Hamiltonian. MEAN FIELD THEORY
symmetry. but only on the gradient of ϕ. meaning a dissipationless current5 . What we have done so far is to show that if a phase gradient is imposed on a system. Finally.
where the last step closely follows the derivation of the current operator in Sec.
In reality the electron density appearing in Eq. We obtain δF = − J(r) + ρ(r) ϕ(r) = 0. The kinetic part reads ˜ H= 1 2me dr Ψ (r)e dr Ψ (r)
† † iϕ(r) 2
i i
+ eA
e−iϕ(r) Ψ(r)
2
1 = 2me =H−
+ eA −
2
ϕ(r)
Ψ(r).4. the current carrying state is chosen. (4. the impurity scattering term. and the phonon coupling term. given by J= ρ0 ϕ. 1. it should be noted that the appearance of the excitation gap is not the reason for the superconductivity itself.54)
dr ϕ(r) · J(r)+
2me
dr ρ(r)( ϕ(r))2 .80
CHAPTER 4. In the normal state of metals a current is always associated with a non-equilibrium state.
5
. where energy is constantly dissipated from the driving source and absorbed in the conductor. because both the Coulomb interaction term. Of course there is an energy cost for the system to carry the current. the so-called superﬂuid. However. only involve the electron density operator ρ(r) which according to (4. it unavoidably leads to the conclusion that the system will carry a dissipationless current in order to minimize the energy cost of the phase gradient. The critical current is reached when the energies are equal.3. the simple minded derivation presented here assumes that all electrons participates. The superconductivity is. where the energy is assumed to depend on phase diﬀerences on a macroscopic scale.

Furthermore.5. We deﬁne a Cooper pair operator bk = ck↓ c−k↑ . only in 1956 Cooper showed that the Fermi surface of the normal metal state was unstable towards formation of bound pairs of electrons (see Chap. were the principles fully understood. which as we saw in Chap. (4. one would naively expect that the energy scale for the superconducting transition temperature would be of the order ωD /k. Also the idea that the superconductivity was somehow similar to Bose-Einstein condensation. the critical temperatures found in “conventional” low superconductors are in most cases less than 10 K and never more than 30 K.57) HBCS = kσ k↑ −k↓
kσ kk
where Vkk is the coupling strength which is only non-zero for states with energy ξk within ξF ±ωD and furthermore constant and negative. The interaction includes only pair interactions and the remaining interaction is supposed to be included in ξk via a Hartree–Fock term. with the bosons being electron pairs. In Chap. However. It turns out that this eﬀective interaction is most important for electrons occupying time reversed states and in fact they can form a bound state which is the Cooper pair. 16 we will see that the phonon mediated electron-electron interaction (derived from the electron-phonon interaction found in Sec. Because the typical energy exchange due to the attractive interaction is the Debye energy. Vkk = −V0 . in this range. This happens for energy exchanges of order the Debye energy.4. The Cooper pair is thus a bound state ∗ of an electron in state ψν (r) and an electron in state ψν (r) or in the homogeneous case electrons in state k and −k. This trace is felt by the other electrons as an attractive interaction. ωD . attractive. 3 for metals is much smaller than the Fermi energy. The origin of the attractive interaction can intuitively be thought of in the following way: when an electron propagates through the crystal it attracts the positive ions and thus eﬀectively creates a positive trace behind it.6) in fact has a range in frequency and momentum space where it is negative. Subsequently in 1957 when the superconducting state was derived using a variational wavefunction by BCS. from the Cooper instability we know that the phonon-mediated interaction tends to pair electrons with opposite spin and momentum.5. These two physical inputs led BCS to suggest the following remarkably successful model Hamiltonian to explain the superconducting state.
. i. This is however far from the truth because while ωD /k is typically of the order of several hundred kelvin.e. SUPERCONDUCTIVITY
81
4. ωD . 16).2
Microscopic theory
The understanding that superconductivity was closely related to the electron-phonon coupling was clear from the early 1950’ies when for example the isotope eﬀect was discovered. It therefore seems that a new energy scale is generated and we shall indeed see that this is the case. Here we give an outline of the main principles in the BCS theory. had been tried and in fact was the underlying idea of London’s theory in 1935. ωD EF . 3. The BCS eﬀective Hamiltonian model is ξk c† ckσ + Vkk c† c† c−k ↓ ck ↑ .

The way to solve it is by a so-called Bogoliubov transformation.63) (4. |vk |2 = 1 2 1− ξk Ek .61)
which diagonalizes the problem if
† Uk Hk Uk =
Ek 0 ˜ 0 Ek
. (4. k Uk = uk −vk ∗ vk u∗ k . we ﬁnd the following solution for u. Hk = ξk ∆k ∆∗ −ξk k . is that the pair operator has a ﬁnite expectation value and that it varies only little from its average value. For this purpose it is convenient to write the Hamiltonian in matrix notation
MF HBCS = k
c† c−k↓ k↑ ξk +
k
ξk ∆k ∆∗ −ξk k
ck↑ c† −k↓
+ =
k
Vkk c† c† k↑ −k↓ ck ↓ c−k ↑ .64)
2 ˜ ξk + |∆k |2 = −Ek .
(4. (4. (4. It is however somewhat unusual in that terms like c† c† and cc appear.
. + constant.59)
kk † Ak Hk Ak
where Ak = ck↑ c† −k↓ . MEAN FIELD THEORY
The mean ﬁeld assumption made by BCS.62)
˜ After some algebra.58b)
∆k = −
kk
Vkk c−k ↓ ck ↑
The mean ﬁeld Hamiltonian is quadratic in electron operators and should be readily solvable.60)
To bring the Hamiltonian into a diagonal form. we introduce the unitary transformation Bk = γk↑ † γ−k↓ = U−1 Ak .58a) (4. v and the energies.82
CHAPTER 4.
kk
−
k
∆∗ ck↓ c−k↑ k
+
(4. (4. E and E |uk |2 = Ek = 1 2 1+ ξk Ek . The BCS mean ﬁeld Hamiltonian is derived in full analogy with Hartree–Fock mean ﬁeld theory described above
MF HBCS = kσ
ξk c† ckσ − kσ
k
∆k c† c† k↑ −k↓ Vkk c† c† k↑ −k↓ ck ↓ c−k ↑ .

The mean ﬁeld parameter provides an energy gap. The self-consistent solution is found from Eq. the so-called gap equation. Above in the general section on mean ﬁeld theory we saw that this procedure is equivalent to minimizing the free energy with respect to the mean ﬁeld parameter.63). ωD ]. we get |∆| d(EF ) ωD dξ [1 − 2nF (E)] . From (4.64) there are no fermion excitations possible with energy less than |∆|. 2 = V0 d(EF )
ωD 0
dξ
tanh β
ξ 2 + |∆|2 /2 ξ 2 + |∆|2
. (4.69)
. ξk ∈ [−ωD . k
(4.68) |∆| = V0 2 2E −ωD and the gap |∆| is determined by the integral equation.61).66)
As is evident from the new Hamiltonian and the solution in Eq.64) we ﬁnd ∆k = −
k
Vkk c−k ↓ ck ↑ . Vkk
k † u∗ γ−k ↓ − vk γk ↑ k † u∗ γk ↑ + vk γ−k ↓ k
=− =−
k
. (4. by calculating the expectation value of the right hand side using the diagonalized Hamiltonian. This rather unusual particle appears because of the lack of particle conservation in the mean ﬁeld Hamiltonian. (4. which is here bk . and that ωD EF .58b).4.66) the bogoliubons are free fermions and therefore their distribution function is the usual Fermi-Dirac distribution. which is why it is called the superconducting gap. Using Eqs. Now using the approximation that Vkk is ﬁnite only for ξk . (4.67)
where we used in the last step that according to the Hamiltonian Eq. and (4. (4. (4. (4. SUPERCONDUCTIVITY
83
The new fermion operators that diagonalize the Hamiltonian are called bogoliubons and they are superpositions of electrons and holes.65)
and the Hamiltonian is in terms of the new bogoliubons
MF HBCS = k † † Ek γk↑ γk↑ + γk↓ γk↓ + constant.
Vkk
† † u∗ vk γ−k ↓ γ−k ↓ − vk u∗ γk ↑ γk ↑ k k
=−
k
Vkk u∗ vk [1 − 2nF (Ek )] .
(4.
(4.5.61) we have the transformations from old to new operators γk↑ † γ−k↓ = u∗ vk k ∗ −vk uk ck↑ c† −k↓ ⇔ ck↑ c† −k↓ = uk −vk ∗ vk u∗ k γk↑ † γ−k↓ . The excitation gap has a number of important consequences. such that the density of states is constant in the energy interval in question. There are two diﬀerent bogoliubons inherited from the two fold spin degeneracy.

To the left the temperature dependence is shown as it follows from the BCS gap equation in Eq.6: Measured values of the gap parameter for three diﬀerent metals compared to the BCS predictions. Both the gap and the critical temperature are thus reduced by the exponential factor exp (−1/V0 d(EF )) as compared to the bare energy scale of the interaction. but there are many others but the reader is referred to the many very good books on superconductivity. ωD . as we foresaw in
"$ 10 ) & ( ' " # &¥© % $ "# ! © ¡ £ ¢ ¥ ¨
(4.70) 2ωD . The right table shows the measured value of the ratio between twice the gap at zero temperature and critical temperature. At zero temperature the gap.84
CHAPTER 4. This strong renormalization is what generates the new scale. determined from tunneling measurements. The critical temperature is found by setting ∆ = 0 in the integral and one ﬁnds approximately kTC = 1.72)
This is in very good agreement with experimental ﬁndings. is found from 2 = V0 d(EF ) ⇓ ∆0 = ωD ≈ 2ωD e−2/V0 d(EF ) . The theoretical BCS value is given in Eq.5. 4. we get the BCS prediction that the ratio of gap to critical temperature is 2∆0 = 3. which can be solved numerically.71)
ωD 0
dξ
1 ξ2 + ∆2 0
= sinh−1
because in metals V0 d(EF ) is typical a very small number. MEAN FIELD THEORY
Figure 4. Combining (4. ∆0 .6.71).6(b).13ωD e−2/V0 d(EF ) . ∆2 0
¦ ¤§ ¡ A 3@ 8 94 876 5 34 2 ¢
£ ¢
¦ ¤¥
. we show the temperature dependence of the gap together with measured values. see Fig. (4. (4.53. ωD exp (−2/V0 d(EF )). where the ratio typically range between 3 and 4.70) and (4.72).69) together with experimental values. sinh (2/V0 d(EF )) (4. In Fig. 4. This is just one of the successes of the BCS theory. kTC (4.

all spins point in the same direction. Also for the time-dependent case Hartree–Fock type approximations will be invoked. where the important concept of symmetry breaking was introduced.6
Summary and outlook
Mean ﬁeld theories are widely used to study phase transitions in matter and also in e. no matter how many orders where included. Of course we have not covered the vast ﬁelds of both magnetism and superconductivity in detail and the interested reader should consult the book by Yosida to learn more about magnetism. In the remaining part of this course we shall deal with the dynamical properties of many-particle systems. Tinkham and de Gennes to learn about superconductivity. This is in fact a general feature of phase transitions. for the so-called Random Phase Approximation treatment of the dielectric function in Sec. Note that the interaction strength appears in a non-perturbative fashion in this expression. It is not possible by perturbation expansions to cross a phase transition line. e.g. because the function exp (−1/x) has no Taylor expansion at x = 0. The mean ﬁeld approximation is in many cases suﬃcient to understand the important physical features. e. 8.
4.g. Once again we see that there is no automatic way to predict the phase diagram of a given physical system. This is still a very active and interesting area of research. SUMMARY AND OUTLOOK
85
the discussion above. This tells us that the result could never have been derived using perturbation theory in the interaction. 12.4. We have seen examples of this in the case of magnets and superconductors. It is however not clear to what extend the theory can be used to explain the superconductivity of high temperature superconductors and other exotic superconducting materials. The RPA result will later be derived later based on a more rigorous quantum ﬁeld theoretical approach in Chap. and the books by Schrieﬀer. because the two states have no analytic connection.5. It means that the state of the system choose to have a lower symmetry that the original Hamiltonian.
.g. atomic physics to compute the energetics of a ﬁnite size systems. and one must rely on a combination of technical skill and most importantly physical intuition.6. The BCS theory has provided an excellent model for the behavior of low temperature superconductors. at least those that has to do with static properties.

86
CHAPTER 4. MEAN FIELD THEORY
.

But how does one then treat the general case of time dependence in second quantization? That question will be addressed in this chapter. With this notation we can summarize the Schr¨dinger picture with its states |ψ(t) and operators A as: o |ψ(t) = e−iHt |ψ0 . and their time evolution is governed by Schr¨dinger’s equation. and systems where interactions are approximated by time-independent mean ﬁeld theory.1)
In the following we will measure the energy in units of frequency.3)
87
. the Heisenberg picture. The state vectors |ψ(t) does depend on time. may or may not depend on time. To interpret the operator e−iHt we recall that a function f (B) of any operator B is deﬁned by the Taylor expansion of f . states : (5.e.1
The Schr¨dinger picture o
The Schr¨dinger picture is useful when dealing with a time-independent Hamiltonian H. instant processes at a single time t = 0. does not depend on time. n!
(5.
5. where time evolution is discussed using three representations. such that drops out of the time-evolution equations: ε/ → ε and H/ → H.
∞
f (B) =
n=0
f (n) (0) n B . and the o interaction picture.Chapter 5
Time evolution pictures
Using the second quantization procedure. we have so far only treated energy eigenstates with a trivial time dependence eiω t . At the end of the calculations one can easily convert frequencies back to energies. o i. Any other operator A may or may not depend on time.2) The Schr¨dinger picture o operators : A.
i
(5. These representations are used in the following chapters to develop general methods for treating many-particle systems. o The time-independence of H leads to a simple formal solution: i ∂t |ψ(t) = H |ψ(t) ⇒ |ψ(t) = e− ~ Ht |ψ0 . or “pictures”: the Schr¨dinger picture. ∂t H = 0. H.

leaving the state vectors |ψ0 time independent. The Hamiltonian H remains time-independent in the Heisenberg picture. the total time derivative of A in the Heisenberg picture is denoted with a dot. ˙ A. By using Eq.4)
Thus we see that the correspondence between the Heisenberg picture with time-independent state vectors |ψ0 . states : iHt −iHt .
5. A(t).
5. (5. In this way an explicit time-dependence of A is taken into account.3
The interaction picture
The third and last representation. As before the original operator A may be time dependent. (5. |ψ0 ≡ eiHt |ψ(t) . In the general case of time-dependent Hamiltonians. If the matrix elements of any operator between any two states are identical in the two representations. while the explicit time derivative of the original Schr¨dinger operator is denoted ∂t A: o ˙ A(t) = eiHt iHA − iAH + ∂t A e−iHt ⇒ ˙ A(t) = i H. Since H is time independent. A(t) + (∂t A)(t). possibly time-dependent. (5. then the two representations are fully equivalent.5) The Heisenberg picture operators : A(t) ≡ e A e H does not depend on time. and the Schr¨dinger picture is given o by the unitary transformation operator exp(iHt).6)
where X(t) always means eiHt Xe−iHt for any symbol X. in particular for X = ∂t A.7)
. is introduced to deal with the situation where a system described by a time-independent Hamiltonian H0 . Note how carefully the order of the operators is kept during the calculation. interaction V (t). TIME EVOLUTION PICTURES
While the Schr¨dinger picture is quite useful for time-independent operators A. We can do that using the Heisenberg picture. it may o sometimes be preferable to collect all time dependencies in the operators and work with time-independent state vectors. H = H0 + V (t). (5. the interaction picture. but time-dependent operators A(t). (5.2) we obtain the identity ψ (t)|A|ψ(t) = ψ0 |eiHt Ae−iHt |ψ0 ≡ ψ0 |A(t)|ψ0 . The important equation of motion governing the time evolution of A(t) is easily established. with H0 |n0 = εn0 |n0 .2
The Heisenberg picture
The central idea behind the Heisenberg picture is to obtain a representation where all the time dependence is transferred to the operators. with known energy eigenstates |n0 . is perturbed by some. we have to switch to the interaction picture.88
CHAPTER 5. Both the Schr¨dinger and the Heisenberg picture require a time-independent Hamilo tonian.

leaving states that vary only slowly in time due to V (t). (5.8) ˆ ≡ eiH0 t A e−iH0 t . but trivial. U (t. ˆ In the general case with a time-dependent V (t) we must rely on the diﬀerential equation appearing when Eq. t0 ) = eiH0 t e−iH(t−t0 ) e−iH0 t0 .12) ˆ ˆ ˆ From this we observe that U −1 = U † .8) is reduced to ˆ ˆ ˆ i∂t |ψ (t) = V (t) |ψ (t) .10)
ˆ The resulting Schr¨dinger equation for |ψ (t) thus contains explicit reference only to the o ˆ interaction part V (t) of the full Hamiltonian H. i.5). in the interaction picture both ˆ ˆ the state vectors |ψ (t) and the operators A(t) depend on time. ˆ U (t. t0 ) is completely determined by ˆ ˆ operator U ˆ ˆ ˆ |ψ (t) = U (t. We remark that Eq. t0 ) = 1. then one can think of Eq. not the full H. THE INTERACTION PICTURE
89
The key idea behind the interaction picture is to separate the trivial time evolution due to H0 from the intricate one due to V (t).8) as a way to pull out the fast.10). U is indeed a unitary operator. This means that in the interaction picture ˆ the time evolution of a state |ψ (t0 ) from time t0 to t must be given in terms of a unitary ˆ (t. (5.9)
which by Eq. As a result. (5. (5. The deﬁning equations for the interaction picture are ˆ |ψ (t) ≡ eiH0 t |ψ(t) .11)
ˆ When V and thus H are time-independent. The interaction picture and the Heisenberg picture coincide when V = 0. (5. t0 ) = 1 + i
t t0
ˆ ˆ dt V (t ) U (t . (5.5. (5. This is obtained by using only H0 . in the unitary transformation Eq.8): ˆ i∂t |ψ (t) = i∂t eiH0 t |ψ(t) + eiH0 t i∂t |ψ(t) = eiH0 t (−H0 + H)|ψ(t) ..13) ˆ eiH0 t e−iHt |ψ0 = U (t.
(5. i. t0 ). time dependence due to H0 . (5. and we obtain: ⇒ ˆ ˆ ˆ i∂t U (t. (5. t0 ) which also only depends on V (t). t0 ) = V (t) U (t.11) naturally ˆ implies the boundary condition U (t0 .3. t0 ). If V (t) is a weak perturbation. operators : A(t) H0 does not depend on time. (5. The ﬁrst hint of the usefulness of the interaction picture comes from calculating the ˆ time derivative of |ψ (t) using the deﬁnition Eq.e. an explicit form for U (t. in the nonperturbed case. t0 ) = 1.e. t0 ) is obtained by ˆ (t) = eiH0 t |ψ(t) = eiH0 t e−iHt |ψ and |ψ (t0 ) = eiH0 t0 e−iHt0 |ψ into ˆ inserting |ψ 0 0 Eq. (5. t0 ) eiH0 t0 e−iHt0 |ψ0
By integration of this diﬀerential equation we get the integral equation 1 ˆ U (t. (5.11). ˆ U (t0 .11) is inserted in Eq. t0 ) |ψ (t0 ) . (5.14)
. states : The interaction picture (5.

2):
∞
ˆ U (t.15)
Note that in the iteration the ordering of all operators is carefully kept. . t0 ) = 1 + i
t t0
ˆ dt1 V (t1 ) +
1 i2
t t0
ˆ dt1 V (t1 )
t1 t0
ˆ dt2 V (t2 ) + . and we deﬁne ˆ ˆ ˆ Tt [V (t1 )V (t2 ) .
(5. Using the time ordering operator. Consider for example the second order term. In n-th order. . . all n! permutations 1 p ∈ Sn of the n times tj are involved. t0 ) starting from U (t . t0 ) =
n=0
1 1 n! i
n
t t0
t
dt1 .17)
θ(tp(1) − tp(2) ) θ(tp(2) − tp(3) ) .1. 5. . This expression for U (t. A more compact form is obtained by the following rewriting. . The solution is 1 ˆ U (t. Time ordering is easily generalˆ ized to higher order terms. where n factors V (tj ) appear. θ(tp(n−1) − tp(n) ). .18)
ˆ Note the similarity with a usual time evolution factor e−iε t . (5.90
CHAPTER 5. . . t0 ) is the starting point for inﬁnite order perturbation theory and for introducing the concept of Feynman diagrams. V (tn ) = Tt e
Rt
t0 dt
ˆ V (t )
. V (tn )] ≡
p∈Sn
ˆ ˆ ˆ V (tp(1) )V (tp(2) ) . we obtain the ﬁnal compact form (see also Exercise 5.
. . A graphical sketch of the contents of the formula is given in Fig. it is therefore one of the central equations in quantum ﬁeld theory.16)
where we have introduced the time ordering operator Tt . (5.
t0
−i ˆ ˆ dtn Tt V (t1 ) . TIME EVOLUTION PICTURES
ˆ ˆ which we can solve iteratively for U (t. V (tp(n) ) ×
(5. . paying special attention to the dummy variables t1 and t2 :
t t0
ˆ dt1 V (t1 ) 1 = 2 = = ≡ 1 2 1 2 1 2
t t0 t t0 t t0 t t0
t1 t0
ˆ dt2 V (t2 )
t1 t0
ˆ dt1 V (t1 )
t
1 ˆ dt2 V (t2 ) + 2
t t0
ˆ dt2 V (t2 )
t t0
t2 t0
ˆ dt1 V (t1 )
t
dt1 dt1 dt1
1 ˆ ˆ dt2 V (t1 )V (t2 )θ(t1 − t2 ) + 2 t0
t t0 t t0
dt2
t0
ˆ ˆ dt1 V (t2 )V (t1 )θ(t2 − t1 )
ˆ ˆ ˆ ˆ dt2 V (t1 )V (t2 )θ(t1 − t2 ) + V (t2 )V (t1 )θ(t2 − t1 ) ˆ ˆ dt2 Tt [V (t1 )V (t2 )].
1 ˆ ˆ ˆ For n = 3 we have Tt [V (t1 )V (t2 )V (t3 )] = ˆ (t1 )V (t2 )V (t3 )θ(t1 −t2 )θ(t2 −t3 )+V (t1 )V (t3 )V (t2 )θ(t1 −t3 )θ(t3 −t2 )+V (t2 )V (t3 )V (t1 )θ(t2 −t3 )θ(t3 −t1 )+ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ V ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ V (t2 )V (t1 )V (t3 )θ(t2 −t1 )θ(t1 −t3 )+ V (t3 )V (t1 )V (t2 )θ(t3 −t1 )θ(t1 −t2 )+ V (t3 )V (t2 )V (t1 )θ(t3 −t2 )θ(t2 −t1 ). t0 ) = 1. . .

.4.B involving two diﬀerent times t1 and t2 cannot be ν given in a simple closed form: [aν1 (t1 ).5): a† (t) ≡ eiHt a† e−iHt .4 these operators can be taken to be either bosonic or fermionic. and consequently the fundamental ν (anti-)commutator [aν (t1 ). is applicable to a wide range of physical problems. a† ] and [H.3. 1.
. As in Sec. scattering events V .5. which. which happen during the evolution from time t0 to time t. Eq. t0 ) can be viewed as the sum of additional ˆ on top of the trivial phase factors arising from H0 . a† 2 (t2 )]F. (5. aν e
−iHt
(5. ν ν aν (t) ≡ e
iHt
In the case of a general time-independent Hamiltonian with complicated interaction terms.
5. It can therefore be ˆ (t.19)
(5. TIME-EVOLUTION IN LINEAR RESPONSE
91
ˆ Figure 5. 3.1: The time evolution operator U (t.20b)
. . The sum phase factors due to V ˆ contains contributions from processes with 0.
5. 1. a† (t2 )]F. t0 ) ≈ 1 + i
t t0
This simple time evolution operator forms the basis for the Kubo formula in linear response theory.20a) (5. as we shall see in the following chapters. aν ] are not simple.4
Time-evolution in linear response
ˆ In many applications the perturbation V (t) is weak compared to H0 . Let us ﬁrst apply the deﬁnition of the Heisenberg picture. 2. t0 ) by the ﬁrst order approximation justiﬁed to approximate U 1 ˆ U (t. the commutators [H.21)
¥£ ¥£ ¥£ ¥£ ¦ £ ¢ ¦ £ ¢ ¦ £ ¢ ¦ £ ¢ § ¨¦ ¥ £ ¤ £ ¢ ¡ ¦ £¢ ¦ £¢ ! !! " £ £ £ £ ©
ˆ dt V (t ). .B = ν eiHt1 aν1 e−iH(t1 −t2 ) a† 2 e−iHt2 ± eiHt2 a† 2 e−iH(t2 −t1 ) aν1 e−iHt1 ν ν = ?? (5.5
Time dependent creation and annihilation operators
It is of fundamental interest to study how the basic creation and annihilation operators a† and aν evolve in time given some set of basis states {|ν } for a time-independent ν Hamiltonian H.

as we shall see in the following chapters. In fact. H=
ν
εν a† aν .3. C] = A{B.26)
For the diagonal Hamiltonian the time evolution is thus seen to be given by trivial phase factors e±iε t . (5. From Eq. calculating (anti-)commutators like Eq.27)
where H0 is diagonalized in the basis {|ν } with the eigenenergies εν .92
CHAPTER 5. is straightforward:2 aν (t) = i[H. which by Hermitian conjugation leads to a† (t) = e+iεν t a† . (5. γ 1. which are valid for any set of operators. but with the eigenenergies (1 + γ)ε. C} − {A. Eq. aν (t)] ˙ = ieiHt = −iεν e
ν iHt
=
ieiHt [H. Assume that H = H0 + γH0 . ν (5. (5. We can also gain some insight into the interaction picture by a trivial extension of the simple model.24)
=
−iεν aν (t).23) (5. (5.ν aν e−iHt (5. 5.B = e−iεν1 (t1 −t2 ) δν1 .29)
We are using the identities [AB. C]B and [AB. and then use the interaction picture of Sec. Consider ﬁrst a time-independent Hamiltonian H which is diagonal in the |ν -basis. TIME EVOLUTION PICTURES
No further reduction is possible in the general case. aν ]e−iHt = ieiHt
ν
εν a† aν . Note that the ﬁrst identity is particularly useful for bosonic operators and the second for fermionic operators (see Exercise 5.ν2 .22)
The equation of motion. the full Hamiltonian H is also diagonalized in the same basis. Obviously. Let us however try to treat γH0 as a perturbation V to H0 . so |ˆ(t) ν
2
=
eiεν t e−i(1+γ)εν t |ν
=
e−iγεν t |ν .21) is the problem in many-particle physics.4).6).21) can be evaluated directly: [aν1 (t1 ).28)
But we actually know the time evolution of the Schr¨dinger state on the right-hand side o of the equation. ν ν (5. C}B. But let us investigate some simple cases to get a grasp of the time evolution pictures.
. ν (5. ν
(5.8) we then obtain |ˆ(t) = eiεν t |ν(t) . C] = A[B.
By integration we obtain aν (t) = e−iεν t aν .
(5. (5. C] + [A. a† 2 (t2 )]F.25)
In this very simple case the basic (anti-)commutator Eq. aν e−iHt ν aν e
−iHt
εν −δν.

18). Eq. (5.30)
The equation of motion for aν (t) is: aν (t) = i[H.29) directly from Eq. etc.8). The ﬁrst two pictures rely on a time-independent Hamiltonian H. Eq. We have derived an explicit expression.6
Summary and outlook
In this chapter we have introduced the fundamental representations used in the description of time evolution in many-particle systems: the Schr¨dinger picture. t0 ) plays an important role in the formulation of inﬁnite order perturbation theory and the introduction of Feynman diagrams. Which picture to use depends on the problem at hand. SUMMARY AND OUTLOOK
93
Here we clearly see that the fast Schr¨dinger time dependence given by the phase factor o iεν t .
5.5. and we do not know the time evolution of the three-operator product a† 1 +ν2 −ν (t) aν2 (t) aν1 (t). Finally. ν
The problem in this more general case is evident. In the following chapters we will learn various approximate methods to deal with this problem. for the time evolution operator ˆ ˆ ˆ U (t. (5.31)
= −iεν aν (t) + i
ν1 ν2
(Vν2 −ν − Vν−ν1 )a† 1 +ν2 −ν (t) aν2 (t) aν1 (t). The reader can e try to obtain Eq. aν (t) aν2 (t) aν1 (t) ν ν (5. where H0 is a time-independent Hamiltonian with known eigenstates. operators. is replaced in the interaction picture by the slow phase factor eiγεν t . of interacting many-particle systems. nine.19) forms the basis of the widely used linear response theory and the associated Kubo formalism. (5. aν (t)] ˙ = −iεν aν (t) + i
ν1 ν2 q
Vq a† 1 +q (t) a† 2 −q (t). when we are trying to study the full time dependence. (5.2). we brieﬂy point to the complications that arise when the interaction is given by a time-independent operator V not diagonal in the same basis as H0 . (5. ν ν
(5. and how its linearized form Eq.5). and the interaction picture. Finally.
. Eq. (5. or equivalently the full dynamics. This feature is then repeated over and over again generating a never-ending sequence of products containing seven. We shall see in the following chapters how the operator U (t. Consider for example the Coulomb-like interaction written symbolically as H = H0 + V =
ν
εν a† aν + ν
ν1 ν2 q
Vq a† 1 +q a† 2 −q aν2 aν1 .18).6. while the interaction picture involves a timedependent Hamiltonian H of the form H = H0 + V (t). Eq. eleven. t0 ) describing the evolution of an interaction picture state |ψ (t0 ) at time t0 to |ψ (t) ˆ at time t. the o Heisenberg picture. (5. The equation of motion for the single operator aν (t) contains terms with both one and three operators. If we write ν down the equation of motion for this three-operator product we discover that terms are generated involving ﬁve operator products. by studying the basic creation and annihilation operators we have gotten a ﬁrst glimpse of the problems we are facing.

94

CHAPTER 5. TIME EVOLUTION PICTURES

Chapter 6

**Linear response theory
**

Linear response theory is an extremely widely used concept in all branches of physics. It simply states that the response to a weak external perturbation is proportional to the perturbation, and therefore all one needs to understand is the proportionality constant. Below we derive the general formula for the linear response of a quantum system exerted by a perturbation. The physical question we ask is thus: supposing some perturbation H , what is the measured consequence for an observable quantity, A. In other words, what is A to linear order in H ? Among the numerous applications of the linear response formula, one can mention charge and spin susceptibilities of e.g. electron systems due to external electric or magnetic ﬁelds. Responses to external mechanical forces or vibrations can also be calculated using the very same formula. Here we utilize the formalism to derive a general expression for the electrical conductivity and brieﬂy mention other applications.

6.1

The general Kubo formula

Consider a quantum system described by the (time independent) Hamiltonian H0 in thermodynamic equilibrium. According to Sec. 1.5 this means that an expectation value of a physical quantity, described by the operator A, can be evaluated as A = 1 1 Tr [ρ0 A] = Z0 Z0

−βH0

n|A|n e−βEn ,

n

(6.1a) (6.1b)

ρ0 = e

=

n

|n n|e−βEn ,

where ρ0 is the density operator and Z0 =Tr[ρ0 ] is the partition function. Here as in Sec. 1.5, we write the density operator in terms of a complete set of eigenstates, {|n }, of the Hamiltonian, H0 , with eigenenergies {En }. Suppose now that at some time, t = t0 , an external perturbation is applied to the system, driving it out of equilibrium. The perturbation is described by an additional time dependent term in the Hamiltonian H(t) = H0 + H (t)θ(t − t0 ). 95 (6.2)

96

CHAPTER 6. LINEAR RESPONSE THEORY

Figure 6.1: Illustration of the linear response theory. At times before t0 the system is in equilibrium, after which the perturbation is turned on. The system is now evolving according to the new Hamiltonian and is in a non-equilibrium state. The Kubo formula relates the expectation value δ A non−eq in the non-equilibrium state to a equilibrium expectation ˆ ˆ value · · · eq of the more complicated time-dependent commutator [A(t), H (t )]. We emphasize that H0 is the Hamiltonian describing the system before the perturbation was applied, see Fig. 6.1 for an illustration. Now we wish to ﬁnd the expectation value of the operator A at times t greater than t0 . In order to do so we must ﬁnd the time evolution of the density matrix or equivalently the time evolution of the eigenstates of the unperturbed Hamiltonian. Once we know the |n(t) , we can obtain A(t) as A(t) = ρ(t) =

n

1 Z0

n(t)|A|n(t) e−βEn =

n

1 Tr [ρ(t)A] , Z0

|n(t) n(t)|e−βEn .

The philosophy behind this expression is as follows. The initial states of the system are distributed according to the usual Boltzmann distribution e−βE0n /Z0 . At later times the system is described by the same distribution of states but the states are now timedependent and they have evolved according to the new Hamiltonian. The time dependence of the states |n(t) is of course governed by the Schr¨dinger equation o i∂t |n(t) = H(t)|n(t) . (6.4)

Since H is to be regarded as a small perturbation, it is convenient to utilize the interaction picture representation |ˆ (t) introduced in Sec. 5.3. The time dependence in this n representation is given by ˆ |n(t) = e−iH0 t |ˆ (t) = e−iH0 t U (t, t0 )|ˆ (t0 ) , n n where by deﬁnition |ˆ (t0 ) = eiH0 t0 |n(t0 ) = |n . n (6.5)

¡

' @ 9 4 H F ' ¡ D (% G F ¡ E C3 B5@ 9 8 6 7 6 53 2 1 D 2 A 4

¦ § §© ¨ § ¦ ¤ ! ¢¡ "

(6.3a) (6.3b)

0 & ¥ § ) © (& § © ¦ $ ¨ # £ ' % ¦ § §© ¨ § ¦ ¥£ ¤ ¢ P5@ 9 53 2 3 I 4

**6.1. THE GENERAL KUBO FORMULA
**

t

97

ˆ ˆ To linear order in H , Eq. (5.19) states that U (t, t0 ) = 1 − i t0 dt H (t ). Inserting this into (6.3a), one obtains the expectation value of A up to linear order in the perturbation

t

A(t) = A = A

0−i

dt

t0 t

1 Z0

ˆ ˆ ˆ ˆ e−βEn n(t0 )|A(t)H (t ) − H (t )A(t)|n(t0 )

n

0

−i

t0

ˆ ˆ dt [A(t), H (t )] 0 .

(6.6)

The brackets 0 mean an equilibrium average with respect to the Hamiltonian H0 . This is in fact a remarkable and very useful result, because the inherently non-equilibrium quantity A(t) has been expressed as a correlation function of the system in equilibrium. The physical reason for this is that the interaction between excitations created in the non-equilibrium state is an eﬀect to second order in the weak perturbation, and hence not included in linear response. The correlation function that appears in Eq. (6.6), is called a retarded correlation function, and for later reference we rewrite the linear response result as

∞

δ A(t) ≡ A(t) − A where

0

=

t0

R dt CAH (t, t )e−η(t−t ) ,

(6.7)

R CAH (t, t ) = −iθ(t − t )

ˆ ˆ A(t), H (t )

0

.

(6.8)

This is the famous Kubo formula which expresses the linear response to a perturbation, H . We have added a very important detail here: the factor e−η(t−t ) , with an inﬁnitesimal positive parameter η, has been included to force the response at time t due to the inﬂuence of H at time t to decay when t t . In the end of a calculation we must therefore take + . For physical reasons the (retarded) eﬀect of a perturbation must of the limit η → 0 course decrease in time. You can think of the situation that one often has for diﬀerential equations with two solutions: one which increases exponentially with time (physically not acceptable) and one which decreases exponentially with time; the factor e−η(t−t ) is there to pick out the physically relevant solution by introducing an artiﬁcial relaxation mechanism. Kubo formula in the frequency domain It is often convenient to express the response to an external disturbance in the frequency domain. Let us therefore write the perturbation in terms of its Fourier components H (t) =

R such that CAH becomes R CAH (t, t ) = ∞ −∞

dω −iωt e Hω , 2π

(6.9)

dω −iωt R e CAHω (t − t ), 2π

(6.10)

98

CHAPTER 6. LINEAR RESPONSE THEORY

ˆ ˆ because [A(t), Hω (t )] 0 only depends on the diﬀerence between t and t , which can easily be proven using the deﬁnition of the expectation value. When inserted into the Kubo formula, one gets (after setting t0 = −∞, because we are not interested in the transient behavior)

∞ ∞

δ A(t) =

−∞ ∞

dt

−∞

dω −iωt −i(ω+iη)(t −t) R e e CAHω (t − t ) 2π (6.11)

=

−∞

dω −iωt R e CAHω (ω), 2π

**and therefore the ﬁnal result reads in frequency domain
**

R δ Aω = CAHω (ω), R CAHω (ω) = ∞ R dteiωt e−ηt CAHω (t).

(6.12a) (6.12b)

−∞

Note again that the inﬁnitesimal η is incorporated in order to ensure the correct physical result, namely that the retarded response function decays at large times.

6.2

Kubo formula for conductivity

Consider a system of charged particles, electrons say, which is subjected to an external electromagnetic ﬁeld. The electromagnetic ﬁeld induces a current, and the conductivity is the linear response coeﬃcient. In the general case the conductivity may be non-local in both time and space, such that the electric current Je at some point r at time t depends on the electric ﬁeld at points r at times t

α Je (r, t) =

dt

dr

β

σ αβ (rt, r t ) E β (r , t )

(6.13)

where σ αβ (r, r ; t, t ) is the conductivity tensor which describes the current response in direction eα to an applied electric ﬁeld in direction eβ . ˆ ˆ The electric ﬁeld E is given by the electric potential φext and the vector potential Aext E(r, t) = −

r φext (r, t)

− ∂t Aext (r, t).

(6.14)

The current density operator of charged particles in the presence of an electromagnetic ﬁeld was given in Chap. 1. For simplicity we assume only one kind of particles, electrons say, but generalization to more kinds of charge carrying particles is straightforward by simple addition of more current components.1 For electrons Je = −e J . The perturbing term in the Hamiltonian due to the external electromagnetic ﬁeld is given by the coupling

P With more carriers the operator for the electrical current becomes Je (r) = i qi Ji (r), where qi are the charges of the diﬀerent carriers. Note that in this case the currents of the individual species are not necessarily independent.

1

6.2. KUBO FORMULA FOR CONDUCTIVITY

99

of the electrons to the scalar potential and the vector potential. To linear order in the external potential Hext = −e dr ρ(r)φext (r, t) + e dr J(r) · Aext (r, t), (6.15)

where the latter term was explained in Sec. 1.4.3. Let A0 denote the vector potential in the equilibrium, i.e. prior to the onset of the perturbation Aext , and let A denote the total vector potential. Then we have A = A0 +Aext , (6.16)

Again according to Sec. 1.4.3, the current operator has two components, the diamagnetic term and the paramagnetic term J(r) = J (r) + e A(r)ρ(r), m (6.17)

In order to simplify the expressions, we can choose a gauge where the external electrical potential is zero, φext = 0. This is always possible by a suitable choice of A(r, t) as you can see in Eq. (6.14). The ﬁnal result should of course not depend on the choice of gauge. The conductivity is most easily expressed in the frequency domain, and therefore we Fourier transform the perturbation. Since ∂t becomes −iω in the frequency domain we have Aext (r, ω) = (1/iω)Eext (r, ω), and therefore the external perturbation in Eq. (6.15) becomes in the Fourier domain Hext,ω = e iω dr J(r) · Eext (r, ω). (6.18)

In order to exploit the frequency domain formulation of linear response we want to write the deﬁnition of the conductivity tensor in Eq. (6.13) in frequency domain. Because we are only considering linear response the conductivity tensor is a property of the equilibrium system and can thus only depend on time diﬀerences σ αβ (rt, r t ) = σ αβ (r, r , t − t ). The frequency transform of Eq. (6.13) is therefore simply that of a convolution and hence

α Je (r, ω) =

dr

β

σ αβ (r, r , ω) E β (r , ω).

(6.19)

Now since Eq. (6.18) is already linear in the external potential Eext and since we are e only interested in the linear response, we can replace J in Eq. (6.18) by J0 = J + m A0 ρ, thus neglecting the term proportional to Eext · Aext . Eq. (6.18) is therefore replaced by Hext,ω = e iω dr J0 (r) · Eext (r, ω). (6.20)

To ﬁnd the expectation value of the current we write J(r, ω) = J0 (r, ω) + e Aext (r, ω)ρ(r) . m (6.21)

100

CHAPTER 6. LINEAR RESPONSE THEORY

For the last term in Eq. (6.21) we use that to linear order in Aext the expectation value can be evaluated in the equilibrium state e e Aext (r, ω)ρ(r) = Aext (r, ω) ρ(r) m m

0

=

e Eext (r, ω) ρ(r) 0 . iω

(6.22)

For the ﬁrst term in Eq. (6.21) we use the general Kubo formula in Eq. (6.7). Since the equilibrium state does not carry any current, i.e. J0 0 = 0, we conclude that J0 = δ J0 . In frequency domain we should use the results Eq. (6.12a) and substitute J0 (r) for the operator “A”, and Hext,ω for “Hω ”, which leads to J0 (r, ω) = CR0 (r)Hext,ω (ω). Collecting J things we now have J(r, ω) = CR0 (r)Hext,ω (ω) + J Writing out the ﬁrst term CR0 (r)Hext,ω (ω) = J dr

β

e ρ(r) 0 Aext (r,ω). m e β E (r , ω). iω

(6.23)

CR (r)J β (r ) (ω) J

0 0

(6.24)

Comparing with the deﬁnition of the non-local conductivity in Eq. (6.19), we can now collect the two contributions to the conductivity tensor. The ﬁrst term comes from Eq. (6.24) and it is seen to of the same form as (6.12a), in particular the response is non-local in space. In contrast, the second term in Eq. (6.22) stemming from the diamagnetic part of the current operator is local in space. Now collecting the two terms and using that Je = −e J , we ﬁnally arrive at the linear response formula for the conductivity tensor σ αβ (r, r , ω) = ie2 n(r) ie2 R Παβ (r, r , ω) + δ(r − r )δαβ , ω iωm (6.25)

R where we have used the symbol ΠR = CJ0 J0 for the retarded current-current correlation function. In the time domain it is given by R ΠR (r, r , t − t ) = CJ α (r)J β (r ) (t − t ) = −iθ(t − t ) αβ

0 0

ˆα ˆβ J0 (r, t), J0 (r , t )

0

.

(6.26)

Finding the conductivity of a given system has thus been reduced to ﬁnding the retarded current-current correlation function. This formula will be used extensively in Chap. 14.

6.3

Kubo formula for conductance

The conductivity σ is the proportionality coeﬃcient between the electric ﬁeld E and the current density J, and it is an intrinsic property of a material. The conductance on the other hand is the proportionality coeﬃcient between the current I through a sample and the voltage V applied to it, i.e. a sample speciﬁc quantity. The conductance G is deﬁned by the usual Ohm’s law I = GV. (6.27)

ˆ where ξ is a unit vector normal to the surface element daξ and σ is the conductivity tensor. For samples which are inhomogeneous such that this simple relation is not applicable. Because of current conservation we can of course choose any cross section. in contrast to the conductivity. The current I is Ie = = ˆ daξ ξ · Je (ξ. aξ ) = daξ daξ daξ ˆ dr ξ · σ(r.6. ξ . aξ . ω = 0) · ξ E(ξ ). one must use the Kubo formula for conductance rather than that for conductivity. and it is convenient to choose an equipotential surface and to deﬁne a coordinate system (ξ. r . In this coordinate system the electric ﬁeld is directed ˆ ˆ along the ξ-direction. whereby a local description is inadequate. aξ . ω = 0)E(r ). Here we are interested in the DC-response only (or in frequencies where the corresponding wave length is much longer than the sample size). see Figure 6. In order to get the dc-response we should the limit ω → 0 of this expression.2.3. where ξ is a coordinate parallel to the ﬁeld line and where aξ are coordinates on the plane perpendicular to the ξ-direction. KUBO FORMULA FOR CONDUCTANCE
101
Figure 6. and W the area of the cross-section.29)
ˆ ˆ dξ ξ · σ(ξ.2: The principle of a conductance measurement. If
¡
§ § © §¨ ¦ ¥£ ¤
¢
. which together with use of current conservation allows a simple derivation. which. The current passing through the sample is equal to the integrated current density through a cross-section. One example is the so-called mesoscopic conductors.28)
where L is the length of the sample. (6. aξ ). E(r) = ξE(ξ). In the Kubo formula derivation we use a coordinate system given by the equipotential lines. L (6. which are systems smaller than a typical thermalization or equilibration length. is a sample-speciﬁc quantity. For a material where the conductivity can be assumed to be local in space one can ﬁnd the conductance of a speciﬁc sample by the relation G= W σ.

LINEAR RESPONSE THEORY
we furthermore take the real part of (6. φext .30)
ω→0
R where CI(ξ)I(ξ ) is the correlation function between total currents. φtot . (6. The charges are related to the potentials through a set of Poisson equations 2 φtot = − ε1 ρe. Because of current conservation the dc-current may be calculated at any point ξ and thus the result cannot depend on ξ.29) we see that what determines the dc-current is the real part of the ﬁrst term in Eq. The typical experiment is to exert an external potential. as for example the electron gas. We can therefore perform the integration over ξ which is just the voltage diﬀerence V = dξ E(ξ ) = φ(−∞)−φ(∞). and in particular the linear response properties. Consequently the function in side the square brackets in Eq. ext 0 2 1 φ =− ρ
ind ε0 e. and measure the resulting total potential. ξ )E(ξ ).
(6. Since the total particle current at the coordinate ξ is given by ˆ I(ξ) = daξ ξ · J. and we ﬁnally arrive at the for linear response formula for the conductance
G = lim
ie2 R CII (ω).ext .33)
Alternatively to working with the potentials we can work with electric ﬁelds or charges.ind
.4
Kubo formula for the dielectric function
When dealing with systems containing charged particles.
6. ξ ) can be shown to be a symmetric function is cannot depend on ξ either.31)
R Here CII is the retarded current-current function. The total potential is the sum of the external one and the potential created by the induced polarization. This simpliﬁcation is the reason for choosing the skew coordinate system deﬁned by the ﬁeld lines.tot 0 2φ = − ε1 ρe. ω→0 ω
(6. φind . the conductance can instead be written as Ie (ξ) = lim dξ Re ie2 R C (ω) E(ξ ) ≡ ω I(ξ)I(ξ ) dξ G(ξ.34) ρtot = ρext + ρind . (6. This in turn aﬀects the measurements. (6. φtot = φext + φind .102
CHAPTER 6. one is often interested in the dielectric properties of the system.30) cannot depend on ξ. In the time domain it is R ˆ ˆ CII (t − t ) = −iθ(t − t ) [I(t). (6.32)
where the current operator I denote the current through an arbitrary cross section along the sample. since the conductance function G(ξ. (6. Furthermore. I(t )] .25) and hence the retarded correlation function of the current densities. When such a system is subjected to an external electromagnetic perturbation the charge is redistributed and the system gets polarized.

(6. also called the relative permittivity ε φtot = ε−1 φext . t). r t ) assuming linear response theory and for this purpose the induced potential is needed. t ). t) = φext (r. The ratio between the external and the total potential is the dielectric response function. D.42)
In electrodynamics the permittivity is deﬁned as the proportionality constant between the electric displacement ﬁeld.36b)
Our present task is to ﬁnd the dielectric function ε(rt. r t ). In the present formulation. i. KUBO FORMULA FOR THE DIELECTRIC FUNCTION
103
and likewise for electric ﬁelds. which are related to the corresponding charges by a set of Gauss laws. ρe (r. where ρ as deﬁned in Chap.2 However. i. Eext plays the role of the D-ﬁeld. (6. r t ) ≡ χR (rt. or rather its inverse ε−1 (rt. t ). χR . r t )e−η(t−t ) φext (r .38) (6. t). t) = φext (r. 1 deﬁnes particle densities. r t ) φext (r .
(6. and Eind . in reality the permittivity is non-local both in time and space and the general relations between the total and the external potentials are φtot (r. dt ε(rt. t) = dr dr dt ε−1 (rt. r t ) = δ(r − r )δ(t − t ) +
2
dr Vc (r − r )χR (r t. t) = ρe (r.ind (r. (6. · E = ρe /ε0 .
(6. Eext . while Etot is the E-ﬁeld
. Etot . t) =
dr
t0
R dt Cρe ρe (rt. Here we have used the symbols ρe for the charge density.eind (r ). t) 0 = 0) as
∞
ρe. D = ε0 Eext .e. ρ ˆ e
The charge-charge correlation function. r t ) φext (r . ρe (r . t )] 0 . and the electric ﬁeld. The external perturbation is represented as the following term to the Hamiltonian H = dr ρe (r) φext (r.
(6.6.36a) (6. r t ). Once the induced charge is known the potential follows from the Coulomb interaction Vc (r − r ) = 1/( 0 |r − r |) as φind (r) = and hence
∞
dr Vc (r − r ) ρe.35)
which is well-known from classical electrodynamics. r t ) = −iθ(t − t ) [ˆe (r. t ). t ). r t ) φtot (r . is called the polarizability function and it is e an important function which we will encounter many times.41)
From this expression we read oﬀ the inverse of the dielectric function as ε−1 (rt.39)
R Cρe ρe (rt. t) +
dr
dr
t0
dt Vc (r − r )χR (r t.4.37)
The induced charge density follows from linear response theory (if we assume that the system is charge neutral in equilibrium.40)
φtot (r. D = εE. (6.e.

we obtain −iq · σ(q. −iωρ(q. Here we consider again the translational-invariant case. ω (6. ω) + iq · J(q. and therefore the problem is considerably simpliﬁed by going to frequency and momentum space. and using the deﬁnition of conductivity J(q. and the continuity equation.2
Relation between dielectric function and conductivity
Both ε and σ give the response of a system to an applied electromagnetic ﬁeld. e (6. ω) = −iσ(q.48) (6. This method. ω). ω)Eext (q. called linear response theory.4. ω). we arrive at the relation ε−1 (q. After Fourier transformation they become products φtot (q. ω) = 0.46) (6. (continuity equation). ω).1
Dielectric function for translation-invariant system
In the translation-invariant case the polarizability can only depend on the diﬀerences of the arguments. with the dielectric function being ε−1 (q. and one should therefore expect that they were related. In later chapters we will make extensive use of the dielectric function ε and the polarizability χ. t − t ).44) or φext (q. ω)qφext (q. but often we must use some approximation to compute the polarizability. is widely used because many experimental
.4.
6.36) have the form of convolutions. ω). r t ) = χR (r − r . ω).5
Summary and outlook
We have developed a general method for calculating the response to weak perturbations. (6. the conductivity tensor is diagonal. ω) = ωχR (q. This formula also tells us what information about the interactions within a given system can be extracted from measurements of the dielectric properties.ω) = ε(q. using Eq. ω) φext (q. ω) .104
CHAPTER 6. (6.
6. ω) = 1 + Vc (q)χR (q. e (6. i. (6.42) includes all correlation eﬀects. ω)φext (q.e.44) and knowing that for a homogeneous system. where both Eqs. and of course they are. ω) = σ(q. χR (rt. LINEAR RESPONSE THEORY
which ends our derivation. ω) = 1 − i q2 Vc (q)σ(q. (6.45)
So if we know the conductivity we can ﬁnd the dielectric response and vice versa. ω)φtot (q.47) Finally. The dielectric function expressed in Eq.43)
6.ω) = ε−1 (q. ω) = ωρe (q. ω)qφext (q.

6. SUMMARY AND OUTLOOK
105
investigations are done in the linear response regime. In this regime the lack of equilibrium is not important. Also in the next chapter we will make use of the linear response result. 12 and 14.
. The general formula is a correlation function of the quantity that we measure and the quantity to which the weak external perturbation couples. and one can think of this as probing the individual excitations of the systems. when discussing tunneling current between two conductors. and the dielectric response reduces to a charge-charge correlation. Because the perturbation is weak it is not necessary to include interactions between these excitations.5. In the case of conductivity we saw that it was the current-current correlation function. These two will be used later in Chaps.

106
CHAPTER 6. LINEAR RESPONSE THEORY
.

The physics of mesoscopic systems is a vast ﬁeld. Typical mesoscopic structures that we have mind are those which are fabricated on semiconductor chips. there is one relatively simple experiment involving a narrow metallic wire where the conductance as a clear signature of quantum transport decreases in pronounced steps of size 2e2 /h as the wire is stretched and pulled apart. the elastic mean free path. At low temperatures. whereas the more high-tech devices based on semiconductor nanostructures only show quantum eﬀects at low temperatures (see e.Chapter 7
Transport in mesoscopic systems
In this chapter we give an introduction to electronic transport in mesoscopic structures and it is our ﬁrst in-depth use of the Kubo formalism. 0 . This chapter deals with the physics of quantum transport which can be understood by invoking the Fermi liquid picture of non-interacting electrons to be discussed in Chap.2).2). By coherent we mean that the quantum mechanical coherence length is longer than the sample size. The important length scales are the coherence length φ . 7. and of course the sample size. typically the range from 50 mK to 4 K. a0 . 107
. When interactions are important another rich ﬁeld of physics appears. and we shall concentrate on the Landauer-B¨ttiker single-particle approach u to conductance of small nanometer-sized coherent systems. e. However. This even happens at room temperature. They are all clear manifestations of the wave propagation of electrons through the structures.1)
Metallic systems are more diﬃcult to bring into the mesoscopic regime because of their small Fermi wave length. Fig. L.
0
φ
(7. but this we will have to study at some other time. . The ﬁeld of mesoscopic transport is interesting in that it combines physics on many length scales. λF . the atomic in Bohr radius.g. λF ≈ a0 . 2. the length scales for these system are related as a0 λF <L<
in . the energy relaxation length. the Fermi wave length of the electron. 14. by electrostatic conﬁnement of two dimensional electron gases (see e.g.3. Sec. and the phenomena that we discuss in the following all rely on quantum eﬀects.g.

y ∈ [0. and furthermore that the system is two-dimensional.1. The contact is thus required to be reﬂectionless. The transformation from a discrete to a continuous ˜ set of energy levels looks a bit diﬀerent in the two cases. it is assumed that the left and right leads are perfect straight segments with hard walls.
(7. The system is divided into ﬁve regions: left reservoir. In the following φk means a state
. W W
2
n = 1. y) = RnE χn (y) = E= 1 kn (E) 1 kn (E)
1 2 2m px
+
∞. φ . y) = η . This implies that we can consider the electron motion to be quantum mechanically coherent in this region. y) ∈ R. m αn. y) αn. we can safely assume that electrons entering the reservoir will be thermalized at the temperature and chemical potential of the contact before returning to the mesoscopic sample. In this case.2a) (7.
εn =
. TRANSPORT IN MESOSCOPIC SYSTEMS
7. much larger than the entrance to the mesoscopic region. .E
∗
px η φ (x.
(7. it is more natural to label the states in terms of their energy E rather than as usual their k values. The quantum number ±1 represents right and left moving states with wavenumber kn (E) = 2m(E − εn )/ 2 .1
The S-matrix and scattering states
We consider a mesoscopic sample connected to electron reservoirs in the form of macroscopic metal contacts. . N πn 2m W
2 2
2m
2 kn + εn .2b) (7. 7.3)
Because of this normalization. the Hamiltonian and the eigenstates with energy E in the leads are given by HL = HR = φ± (x. and the phase breaking length. . The wavefunctions φ± have been normalized in a particular manner so that they all carry the same absolute probability current in a given cross section:
W 0
dy φη (x. 2. y) = LnE φ± (x.E m
η = ±1. Fig. and right reservoir.
πny 2 sin . For simplicity. By mesoscopic we mean that the size L of the sample region between the two reservoirs is much smaller than the energy relaxation length in . e±ikn (E)x χn (y). In the following we solve for the eigenstates in a geometry similar to Fig. y) ∈ L. (L). Furthermore. mesoscopic region (M). (x.2c) (7.2e)
e±ikn (E)x χn (y).
Here χn denote the transverse wavefunction and W is the width of the leads. left lead. W ] otherwise.2d) (7. 7.108
CHAPTER 7. since the reservoir is a macroscopic conductor.
1 2 2m py . (x. but in practice we can introduce a cut-oﬀ at some large value N without aﬀecting the lowest occupied states. In principle n can be any positive integer. right lead (R). that they are identical as in the ﬁgure.1 illustrates how a contact formed as a “horn” can give a reﬂectionless contact. .

a+ . √ √ ˜ with the usual normalization. where a± and b± are some unknown sets of coeﬃcients. y) ∈ R. . As usual the wavefunction and its 1 2
.1. n bn φRn. ∝ k. The eigenfunctions in the middle region.5) ψE (x. y) + ψM. n an φLn. the system is separated in three regions: L. in the middle region M . y) ∈ L. (x.1: The geometry considered in the derivation of the Landauer formula. THE S-MATRIX AND SCATTERING STATES
109
L
Left reservoir
M
mesoscopic sample
R b+ bRight reservoir
a+ a-
Perfect lead with N channels
Perfect lead with N channels
Figure 7. relating incoming and outgoing electron waves.E (x.7. .E (x. y) ∈ M. which in vector form are written n n as a+ = (a+ . by the density of states. while φk = eikx / k. Let us therefore introduce the so-called scattering matrix or S-matrix formalism. All we will need is the transmission coeﬃcients. R o and M . We can therefore write an eigenstate as − − + + n an φLn.E (x. . y) + − − + + (7.
(7. (x. M . Two reﬂectionless contacts each with N channels connect to a mesoscopic region. A given eigenstate with energy E is some linear combination of φ± and φ± in the RnE LnE leads L and R.E (x. The wave function is written as a superposition of incoming and outgoing wave at the two entrances. ∝ dk/dE.E (x.) and similarly for a− and b± .4)
As we shall see in detail later.E . y). (x. but fortunately we need not specify the wavefunction in the complicated region. ˜ ˜ φk |A|φk → L
k>0 ∞ 0 ∞
dk ˜ ˜ φk |A|φk 2π
=
0 ∞
=
0
dk k φk |A|φk 2π dE k φk |A|φk . y). When solving the Schr¨dinger equation. y) = n bn φRn. 2π dE/dk
∞ 0
=
m 2π 2
dE φk |A|φk . φk = eikx / L. that the quantization of the conductance in units of the universal conductance quantum e2 /h is due to the cancellation of the velocity. are in general not easy to ﬁnd. a feature particular of one dimension. and some unknown complicated function φM. y).

(7. where the superscript ± refers to the direction from which the incoming wave comes. y).E (x.10)
.E (0. It is merely written down in order to illustrate the linear dependence of the coeﬃcients. y) ∈ R. In the plus direction (an incoming wave from the left) the scattering states are + − φLn.E (x.E (x. (x. The coeﬃcients of the scattering matrix are of course energy dependent. to incoming waves. 0. y) = n tn n φLn . 1. i.E (x. The amplitude for transmission in the opposite direction is given by tnn . (7. y) + n rn n φRn . y). y) ∈ L. . φ+ and φ− .9) ψnE (x. The scattering states are denoted ψ ± .110
CHAPTER 7. (7. A particular useful way of representing the n n linear dependence is through the so-called scattering matrix. φ− and φ+ . dy χn (y) ∂x ψM (x. y).e. + + (x. y) ∈ L. Most of the time. and in the minus direction (an electron incoming from the right hand side) − + φRn.
dy χn (y) ∂x ψM. dy χn (y)ψM. y). which relates the amplitudes of outgoing waves.E in the middle region this condition gives 4 × N linearly independent equations to determine a± and b± . − − (x.8)
Here the matrix element tnn represents the transmission amplitude for an incoming wave from the left in state n to be transmitted into state n on the right hand side. L R L R cout ≡ a− b+ = r t t r a+ b− ≡S a+ b− ≡ S cin . . For a given ψM. . Similarly the element rnn gives the amplitude for being reﬂected back into the left lead in state n. ? (x.E (x. . y). cin = (0. {a± } and {b± }. . TRANSPORT IN MESOSCOPIC SYSTEMS
derivative must be continuous. ψnE (x. y) + n rn n φLn . which are states with an incoming wave in one particular lead state. . y) = n tn n φRn . ? (x. or S-matrix.E (x. y)
x=0
.E (L. 0.
(7. we will not have to solve such a system of equations unless we want an exact expression for the wavefunction. y) ∈ M. we suppress this dependence in the notation.E (x. y) ∈ R. We now deﬁne the so-called scattering states. These equations are n n a+ + a− = n n b+ eikn (E)L + b− e−ikn (E)L = n n a+ − a− = n n b+ eikn (E)L − b− e−ikn (E)L = n n kn (E) kn (E) 1 i kn (E) 1 i kn (E) dy χn (y)ψM. y). (x. . y)
x=L
Fortunately.).7)
Here we have deﬁned the important S-matrix to be a matrix of size 2N × 2N with the N × N reﬂection and transmission matrices as block elements S= r t t r . y) ∈ M.

Let us compute I(x) for a state with incoming coeﬃcients cin = (a+ . . which we will make use of below: S† S = 1 and furthermore SS† = 1 ⇔ 1 = r r † +tt† = rr + t t † . First calculate the current in region L
W
IL (x) = =
0
dy a+ · φ+ + a− · φ− L. i.E (7. φ− .1. 0 = r† t + t† r = t † r + r † t. From the unitarity follows some properties of r and t.
7.15)
m
. φ+ .
↔
Jx =
↔
→
2mi
∂x − ∂x .) and φ− = (φ− .2E L.E |a+ |2 − ra+ + t b−
2
∗ ↔ Jx
a+ · φ+ + a− · φ− L. y). .
(7. This we do because we will need the currents later on anyway. and that we can get from the S-matrix.2E L.
←
(7.1E L. The current through a cross section for a given state is. .
W
I(x) =
0
dy Ψ∗ (x. let us look at some properties of the S-matrix. S−1 = S† .
(7. ˙ i. b− ).
where φ+ = (φ+ .1
Unitarity of the S-matrix
Before we calculate the transport properties of a mesoscopic system.13)
We also show the unitarity in a bit more explicit way by calculating the currents on the left and right hand sides of the system.16)
.7.12)
(7. The incoming electron ﬂux n |cin |2 = |cin |2 must equal the outgoing ﬂux n |cout |2 = |cout |2 and therefore c† cout = c† cin out in ⇒ c† (1 − S† S) cin = 0. First of all.1.11)
and hence S† = S−1 . . cf. I(x) cannot depend on x. because to ﬁnd the conductance all we need is the transmission probabilities of electrons. .14)
where the arrows indicate to which side the diﬀerential operators are acting. This is a consequence of probability current conservation. 0 = rt† + t r † = tr† +r t † . (1. Eq. y)Jx Ψ(x.e. For a stationary state.).E obtain IR (x) = m −|b− |2 + ta+ + r b−
2
.E L. i.99b). an eigenstate with energy E. In the same way for R we L. THE S-MATRIX AND SCATTERING STATES
111
The wavefunction in the scattering region is not speciﬁed. .E L. the continuity equation gives ∂x J = −ρ = 0.1E L.E L. in (7.e. it must be unitary.e. .
†
⇔
1 = r † r + t† t = r † r + t † t .

18)
∗ Now. IL = IR . and cnew = c∗ . However. we have cnew = S−B cnew out in which shows that S−B S∗ = 1 B ⇒ S∗ = S† −B B ⇒ SB = ST . since HB = H−B we see that
HB ΨB (r) = E ΨB (r)
⇔
∗ H−B Ψ∗ (r) = E Ψ∗ (r).12) and hence S is unitary. Suppose we have an eigenstate which is a linear combination of incoming and outgoing waves ΨB (r) = (cin φin . (7.
(7. the scattering matrix therefore has an additional symmetry besides being unitary: it is also a symmetric matrix.and outgoing wave functions are cnew = c∗ . and we obtain Eq. cout φout ).
7.21) ⇒ c* = S−B c∗ = S−B S∗ c* . and in this case the Schr¨dinger equation is o
2
HB ΨB (r) =
−
2m
r
e +i A
2
+ V (r) ΨB (r) = E ΨB (r). which is a −B B solution for −B.1. TRANSPORT IN MESOSCOPIC SYSTEMS
m m
(1 − r† r) · a+ − b−
†
t †t
†
b− − 2 Re[ a+
† †
r t b− ] t r b− ] . A non-zero magnetic ﬁeld B = × A breaks time-reversal symmetry. This has some important consequences for the statistics of S-matrices in disordered systems. −B (7.
.2
Time-reversal symmetry
Time-reversal symmetry means that H = H ∗ .19)
or in short: if ΨB (r) is a solution so is Ψ∗ (r). −B −B
(7. In that case the scattering matrix is not only unitary o it is also symmetric. t) is a solution to the Shr¨dinger equation so is Ψ∗ (r. We can therefore construct new eigen−B states by complex conjugation followed by reversal of the magnetic ﬁeld. −t). the new in. because if Ψ(r. because complex conjugation reverses the direction of propagation. which can be seen experimentally by studying transport with and without an applied magnetic ﬁeld. Since out out in in Ψnew is a solution for −B.
(7. This will be of importance when we look at disordered systems below.17a) (7.20)
In case of time-reversal symmetry. in out B in (7. then we can make a new eigenstate by Ψnew (r) = Ψ∗ (r). S = ST .112 or more detailed IL = IR = a+ b+
†
CHAPTER 7.17b)
†
(−1 + r † r )b+ + a+
t† t a+ − 2 Re[ a+
† †
From the continuity equation we know that the current on the two sides must be equal.

However. The number of quanta will be the number of channels connecting the two sides. (7. the electrons entering the reservoir from the sample are thermalized before returning. say. The Landauer-B¨ttiker formula tells us that the conductance of a mesoscopic sample u is quantized in units of 2e2 /h. The answer we ﬁnd. i. CONDUCTANCE AND TRANSMISSION COEFFICIENTS
113
7. This is in fact what happens for the quantum point contact. However. fortunately. but when summing over all incoming states n we in fact get. n. Secondly.e.22) in terms of Tn or Tn as we please. There a particular smooth interface between the two reservoirs ensures that Tn changes in a well-controlled manner between 0 and 1. 7. h
(7.e. i. Thus all electrons entering from the contacts are distributed according to the FermiDirac distribution nF of the given reservoirs.
7. which allows us to calculate the current directly. Therefore it is natural to express the occupation of the scattering ± eigenstates ψnε by two diﬀerent distribution functions f ± and the chemical potentials µL/R
. heuristic derivation u
We argued above that if the reservoirs are much wider than the mesoscopic region and its leads. These ﬂuctuations are universal in the sense that they have an amplitude of the order e2 /h independent of the average conductance. This will be done in two diﬀerent ways: ﬁrst we will argue on physical grounds that the population of the scattering state is given by the equilibrium distribution function of the reservoir.e. since the mesoscopic region is deﬁned to be phase coherent.22)
where Tn are the eigenvalues of the matrix t† t.3.4. 7.2.1
The Landauer-B¨ ttiker formula. The linear response result shows that indeed the ﬁrst method gave the right answer. the celebrated Landauer-B¨ttiker formula. no energy relaxation takes place there. Furthermore.1. which is discussed below in Sec.2
Conductance and transmission coeﬃcients
Next we calculate the conductance. This is discussed in Sec. and consequently electrons originating from. So we can write Eq. we ﬁnd the same result.7. namely in the ﬂuctuations of conductance. we calculate the conductance using linear response theory. by the sum of transmission probabilities G= 2e2 h Tn =
n
2e2 Tr[t† t]. i.2. This probability is Tn = t† t nn . ends up on the other side. since Tn is a number between 0 and 1 one expect this quantization to show up only for some special geometries where Tn is either 0 or 1. one could get in doubt if the Pauli principle is treated correctly. the probability that an electron in a given incoming state. at least in the linear response limit. and.3. the left reservoir maintain their distribution function equal to that of that reservoir. n Tn = n Tn . This should not be confused with the transmission probabilities. there are other examples where the conductance quantum e2 /h shows up. then we can assume reﬂectionless transmission from the leads to the reservoirs. While the ﬁrst method is more physically appealing. is very simple and u physically sensible: the conductance of a mesoscopic sample is given by the sum of all the transmission possibilities a given electron has.

If they are not degenerate the trace must also include a trace over the spin degrees of freedom. For |eV | = |µR −µL | µ. Naturally.17a) and (7. we have stressed the energy dependence of the transmission matrix.
(7.
.22) from ﬁrst principles using the linear response formalism of Chap.17b) +) = δ − by substituting (a n nn for a state moving in the positive direction and (b )n = δnn for a state moving in the negative direction. but rather about the thermal population of eigenmodes. the next section is devoted to a derivation of Eq. Because of current conservation. the current is therefore simply I= e 2π
∞
− Ink = −
dE
n 0
t† t
nn
nF (E − µL ) − t † t
nn
nF (E − µR ) . Once a state is occupied in one lead. In order to dismiss any concern about this point. and the current can be written as I= e 2π
∞ 0
dE Tr t† tE E
nF (E − µL ) − nF (E − µR ) . Let us look at the current in the left lead: I = IL = e
nk ± The currents carried by a scattering state ψnε can be read oﬀ from Eqs. (7. Here we have assumed that the spin u degrees of freedom are degenerate which gives rise to a simple factor of two. as one would normally do in a Boltzmann equation. or M . (7.4). we Taylor expand around µ and ﬁnd after integration
From spin
I=
2
e2 V Tr t† tE E h
⇒
G=
2e2 2e2 Tr t† t = h h
Tn . The unitarity condition Eq.26) m m nn nn Transforming to an energy integral as in Eq. it is automatically also occupied in the other. which means that we should not include any kind of blocking factors (1 − nF ) to ensure that the ﬁnal state is empty. then leads to Tr[t † t ] =Tr[t† t]. 6. We get + Ink =
˜− ˜+ Ink f + (Enk ) + Ink f − (Enk ) .27)
The sum over diagonal elements of t† t is nothing but the trace.25)
t †t = −1 + r † r . (7.
f − (ε) = nF (ε − µR ).28).24)
m
1 − r† r
nn
=
m
t† t
nn
. (7. (7.
CHAPTER 7. Thus we are not talking about a scattering event from one reservoir to the other. R.
(7. we choose to do so in the perfect leads L or R where the wavefunctions are known.23)
Now it is a simple matter to calculate the current through the mesoscopic system.
n
(7.13). The expression Eq.27) for current relies on the fact that the scattering states are eigenstates of the system.29)
This is the famous Landauer-B¨ttiker formula. we can calculate it in either of the regions L. but at low voltages V and temperatures T we can assume Tn to be energy independent and the integral can be done. TRANSPORT IN MESOSCOPIC SYSTEMS
f + (ε) = nF (ε − µL ).
(7.114 of the relevant reservoirs. (7. where µ is the equilibrium electrochemical potential.
(7.
(7.28)
In Eq. (7.

(7.30) yields G(ω) = −
0
= δλλ nF (Eλ ). 0)
0
=
νν
jνν (x )
λλ
jλλ (x )ei(Eλ −Eλ )t/~
c† cλ .
(7. G(ω) = − 2e2 Im ω
∞ −∞
dt ei(ω+iη)t (−i)Θ(t) [I(x. ω → 0. t). n. c† cν ν λ
0
=
λλ
|jλλ (x )|2 ei(Eλ −Eλ )t/~ nF (Eλ ) − nF (Eλ ) . (7. linear response derivation u
Our starting point is Eq. λ → nη 2π ~2 and setting T = 0 such that −∂nF (E)/∂E = δ(E − EF ).31) (7. i.
(7. y). → ←
(7.36)
. We will of course use the scattering states that we found above as our basis. y) ∂ x − ∂ x ψλ (x. η = ±}. (7.) In second quantization the current operator is given by I(x) =
λλ
jλλ (x) c† cλ . Changing the sum over eigenstates to integrals over energy.2
The Landauer-B¨ ttiker formula. (6. ( ω + iη + Eλ − Eλ ) F
2e2 Im ω
(7. (Again we consider the spin degenerate case which is the reason for the factor of two.35)
m dE.2.
(7. and that jλλ (x ) = (jλ λ (x ))∗ .e. We start by calculating the commutator in Eq. t).34)
λλ
and in the dc-limit.n η EF (x )|2 .31) expressing the conductance G in terms of the currentcurrent correlation function. and where jλλ is a matrix element of the current operator. Inserting this |jλλ (x )|2 n (Eλ ) − nF (Eλ ) .7. I(x.33)
where we used that c† cλ λ into Eq. which means that the quantum number λ is speciﬁed by λ = {E. rendering G independent of x.2. 0)] 0 . CONDUCTANCE AND TRANSMISSION COEFFICIENTS
115
7. λ
∗ dy ψλ (x. I(x . the conductance becomes
G(0) = −2 e2 π
m 2π 2
2 nn .ηη
|jnηEF .30)
where the current operator I(x) due to current conservation can be evaluated at any cross section x in the perfect leads.32)
jλλ (x) =
2mi
where we choose {ψλ } as a set of eigenstates.30) I(x . one has G(0) = −2 e2 π
λλ
|jλλ (x )|2 −
∂nF (Eλ ) ∂Eλ
δ (Eλ − Eλ ) .

and if there are N “perfect” channels it is N 2e2 /h. By applying voltage to the gates the width of the constriction could be controlled very accurately. We will now see how this nice eﬀect can happen. each step being of height 2e2 /h. see Fig. The technique they used was a socalled splitgate geometry where a set of metallic gate electrodes was put on top of a twodimensional electron gas such that a narrow contact between the two sides of the 2DEG was formed.
(7.2.ηη 2
m
2
Tr j† j Tr
2
= =2
m m
t† t
2
+ t †t
2
+ r † tt† r + r† t t † r (7.13).1
Electron wave guides
Quantum point contact and conductance quantization
One of the most striking consequences of the Landauer-B¨ttiker formula for conductance is u that the conductance of a perfect channel is 2e2 /h. h (7. (7.n η EF (x )|2 =
nn . until only one remains.
after using the result Eq.38)
Tr t† t . respectively. We have thus seen that it can be derived u microscopically.3
7. The ﬁnal result is therefore I = − 2e2 Tr t† t h dx E(x ) = 2e2 Tr t† t V. 7.n η EF (x ) = t† t m
nn nn
t† r −t † t
nn nn
− t †r
≡
m
j. Hence we get |jnηEF . The relevant wave equation for an electron horn is of course the Schr¨dinger equation. leading to a staircase of conductance. but there is in prino ciple no diﬀerence between the electron wave guide and horn wave guides used in loud
. Suppose there is a smooth constriction between two electron reservoirs. We obtain jnηEF . This has been experimentally tested in numerous experiments and it is now a well-established fact.37)
where the rows and columns correspond to η = +1 and −1.10b. see Fig. As the width decreases quantum channels are squeezed out one by one.
7. 2. and we evaluate them in the L or R region at our convenience. and any doubt about the validity of the treatment of the occupation factor in the heuristic derivation.n η EF (x ) are independent of x . TRANSPORT IN MESOSCOPIC SYSTEMS
Due to current conservation the current matrix elements jnηEF .39)
which again is the Landauer-B¨ttiker formula.116
CHAPTER 7. The ﬁrst experiments showing this was done by groups in Delft (Holland) and Cambridge (England) in 1988.3. Smooth here means a horn-like shape were the curvature at all points is large compared to the wavelength of the wave which is going to be transmitted through the horn. has been removed.

y) Ψ(x.42)
. The device was fabricated and measured at the Ørsted Laboratory. o
2
−
2m
2 ∂y + Vconf (x. y) = E Ψ(x. water waves or other wave phenomena.7. speakers. The right panel shows the conductance versus sidegate voltage. (7. y) = φn (x)χnx (y).
(7.41)
n
This is always possible at any given ﬁxed x since.40)
where Vconf (x.3. we try to separate the motion in longitudinal and transverse motion.2b). but you might say a very striking one. Because the change along the x-direction is assumed to be smooth. being solutions of the transverse Schr¨dinger equation. By clever design this point contact yields a particularly large subband splitting. {χn (x)} forms a complete set. Had the conﬁnement potential been rectangular we would have eigenstates as φ± in Eq. as are the expansion coeﬃcients φn (x). At the lowest temperature (1. Ψ(x. Niels Bohr Institute. Inspired by that we expand the wave function in terms of the transverse eigenstates χnx (y) which however are x-dependent now. ELECTRON WAVE GUIDES
117
Figure 7.3 K) the conductance shows clear steps at integer values of 2e2 /h. y) is the conﬁnement potential. (7. The upper left panel is a picture of the surface of an GaAs chip with an etched point contact structure.2: An experiment on quantized conductance. y). The lower left panel is an zoom-in of this structure recorded in an electron microscope. The Schr¨dinger equation for the quantum point contact geometry is o
2
−
2m
2 2 ∂x + ∂y + Vconf (x.
(7. So the quantized conductance is nothing but a manifestation of the wave nature of a quantum particle. y) χnx (y) = εn (x)χnx (y). which is why the conductance quantization persists up to “high” temperatures of the order 20 K.

y. Inserting Eq. nx 2
(7.40) and multiplying from the left with χ∗ (y) followed by nx integration over the transverse direction.44)
As mentioned. (7. ∂x χn x (y) ≈ 0. Vconf (x. (7.e.e.46)
with the corresponding eigenenergies εn (x) =
2 π2
2m [d(x)]2
n2 .41) into Eq. the maximum of εn (x). In the case of hard walls. the fundamental approximation we wanted to impose was the smooth geometry approximation. (7.43)
δn =
m
n
1 2 dyχ∗ (y) (∂x φn (x)) (∂x χn x (y)) + φn (x)∂x χn x (y) . d(x)/2]. i. (7. It means that the derivative of the transverse mode with respect to longitudinal direction is neglected. it is transmitted through without reﬂection.
(7.118
CHAPTER 7. i. otherwise not.45)
the transverse wavefunctions are the well-known wavefunction for a particle in a box χnx (y) = 2 sin d(x) πn(y − d(x)/2) d(x) . The width of constriction and thereby the height of εn (x) is controlled by a voltage applied to the gate electrodes.
(7.3: Illustration of the adiabatic contact giving rise to an eﬀective one-dimensional barrier. When the energy of the incident electron is larger than the maximum transverse kinetic energy. y) = 0 for y ∈ [−d(x)/2. yields
2
− where
2
2m
2 ∂x + εn (x) φn (x) = Eφn (x) + δn . often referred to as the adiabatic approximation.47)
. TRANSPORT IN MESOSCOPIC SYSTEMS
Closed channel
Energy Gate E
ε n+1(x) ε n(x)
Gate x
Open channel x
Figure 7. ∞ otherwise.

it is smaller than E the mode has suﬃcient energy to pass over the barrier and get through the constriction. In this case we can directly read oﬀ the transmission amplitude because in the notation used for the scattering states. where the conﬁnement potential is modelled by 1 1 2 2 (7. dmin = d(0). φn . Let x = 0 be the position in the constriction where this is most narrow.e. Right at the point where a new channel opens.3. A useful model is the so-called saddle point model for the constriction. (7. (7. i.48)
The transverse direction has thus been translated into an eﬀective 1D barrier. The shape of the smearing will in general depend on the geometry of the constriction and is. (7.
p(x) =
2m(E − εn (x)). ELECTRON WAVE GUIDES
119
Taking the derivative ∂x χn x (y).
(7. of course). 2 2 where V0 is a constant. the mode n cannot transmit (neglecting tunneling through the barrier. (7. where εmax is changed by changing the width of the constricn tion through the voltage of the gate electrodes. at this place is larger than E. we would expect some smearing of the step. εmax ≡ εn (0).49)
which is a solution to Eq.50)
All subbands with energy smaller than EF contribute with one conductance quantum. The conductance is therefore G= 2e2 h Θ(EF − εmax ). Experiments using the splitgate geometry indeed show that the conductance traces (meaning conductance versus gate voltage) are well described by Eq. which obey the 1D Schr¨dinger equation o with an energy barrier εn (x)
2
−
2m
2 ∂x + εn (x) φn (x) = Eφn (x). y) = mωy y 2 − mωx x2 + V0 . If the transverse kinetic energy. For smooth barriers.48) if |p (x)/ p2 (x)| 1. The essence of the adiabatic approximation is that d (x) 1. Obviously the WKB approximation breaks down if p(x) is too small. This is roughly n what is seen experimentally. Using this potential it can be shown that the transmission probability has a particular simple form. The saddle point model can be thought of as a quadratic expansion of the conﬁnement potential near its maximum. we can use the WKB approximation result for the wavefunction φn (x) ≈ φWKB (x) = n 1 p(x)
x
exp i
−∞
dx p(x )/
. 7. If. not universal.3. however. such we end up with an eﬀective onedimensional problem of decoupled modes. n
n
(7. in contrast to the step heights.51) Vconf (x.52) 1 exp π E − V0 − (n + 2 ) ωx / ωy + 1 For this model the smearing of the conductance steps thus has the form of a Fermi function. The barrier is there because some of the kinetic energy is bound into the transverse motion.
.7. which happens when EF = εn (0). will give something proportional to d (x). we have r = 0 and hence |t| = 1. this is illustrated in Fig.52). namely 1 Tn (E) = . which results in a step structure of the conductance as a function of εmax .

35 2.65 2.60 2.50 2.120
CHAPTER 7. 7.70 2.e. (7.3. At small magnetic ﬁelds we can neglect o the orbital changes induced by B in the arms of the ring and absorb the vector potential due to the B-ﬁeld through the hole of the ring as a phase factor ΨB=0 (r) = ΨB=0 (r) exp −i e
r
dl · A .40 2. where the left panel shows the experimental realization.
7.54)
.18). This is shown in the bottom part. Because of the applied B-ﬁeld we must add a vector potential A to the Schr¨dinger equation Eq. The interference is modulated by magnetic ﬂux enclosed by the paths.49).
(7. i.2
Aharonov-Bohm eﬀect
A particular nice example of interference eﬀects in mesoscopic systems is the AharonovBohm eﬀect. (7.4.40) as in Eq.30 -60 -40 -20 0 20 40 60
B (mT)
Figure 7. The geometry is illustrated in Fig. no backscattering. (7.45 2.32 K
G (e 2/h)
2. In that case the transmission coeﬃcient is given by a sum corresponding to the two paths t ∝ exp −i e
r path 1
dl · A + eiφ0 exp −i
e
r
dl · A . Each of the arms in the ring could be an adiabatic wave guide. where an applied magnetic ﬁeld B is used to control the phase of two interfering paths. Both the device fabrication and the measurements have been performed at the Ørsted Laboratory.
path 2
(7. where the wave function can be assumed to be of the form in Eq. while the right panel depicts the conductance versus B-ﬁeld trace.55
T = 0. TRANSPORT IN MESOSCOPIC SYSTEMS
path 1
B path 2
2.4: A device which shows Aharonov-Bohm eﬀect. because of interference between path 1 and path 2.53)
We now approximate the line integral by an integral following the center of the waveguides and furthermore assume ideal adiabatic arms.

The conductance will oscillate as with the applied magnetic.4. we must ﬁrst learn about how to average over S-matrices. In fact. Such ensembles have been studied for a long time. or in other words the distribution P (S)
.1
Statistics of quantum conductance. One cannot precisely neither locate nor control the positions of the impurities. which means that all parts of the phase-space are visited with equal probability. See Fig. The basic assumption being made is that the Hamiltonians describing each of the systems of the ensemble are drawn randomly according to some probability distribution only constrained by the symmetry of the system. which is a mesoscopic region connected to reservoirs just as we have discussed above where the Landauer-B¨ttiker u formula was derived. 7. as a function of any external parameter (e. a signature of quantum interference. or density) the energy levels avoid to cross one another. shape. Experiments have veriﬁed this picture.5. 7. The matrix elements of the Hamiltonians are assumed to follow a Gaussian distribution. The average and the variance of the conductance will turn out to exhibit interesting quantum phenomena.5 shows an example of a disorder mesoscopic system. In order to understand these two phenomena. initially atomic nuclei containing a large number of nucleons. It cannot be a surprise that the classical motion in such a geometry is expected to be chaotic and the system to be ergodic. Again we use the Landauer-B¨ttiker to calculate the u conductance. random matrix theory
Let us consider the statistical properties of some ensemble of disordered or chaotic systems inﬂuenced by some external parameter.
7. magnetic ﬁeld. The experiments we have in mind are e. see Fig.7. Instead one studies the statistical properties of the conductance for an ensemble of systems.4. respectively.g. 7.55)
where Φ is the ﬂux enclosed and Φ0 = h/2e is the ﬂux quantum. DISORDERED MESOSCOPIC SYSTEMS
121
where φ0 is some phase shift due to diﬀerent length of the two arms. which is a manifestation of the non-locality of quantum mechanics.4
Disordered mesoscopic systems
In this section we shall study disordered mesoscopic systems. but because the system is disordered it makes little sense to talk about the conductance for speciﬁc sample geometries.
(7. experiments on disordered “quantum dots”. Fig. Note that the eﬀect persists even if there is no magnetic ﬁeld along the electron trajectories. namely weak localization and universal conductance ﬂuctuations. and from this one can argue that the S-matrix follows the so-called circular ensemble distribution. The transmission probability now becomes |t|2 ∝ 1 + cos φ0 − e
r
dl · A
path 1+2
= 1 + cos φ0 − π
Φpath 1+2 Φ0
.4. Quantum mechanically this means that there are no symmetries and hence no systematic degeneracies of energy levels.g. This means that all unitary matrixes are equally likely. This statistical method is known as random matrix theory (RMT). This important phenomenon is known as level repulsion.
7.

By suitable choice of U0 the various averages can be derived. which for any ﬁxed unitary matrix U0 is a consequence of the constant probability assumption. Uαa Uβb = M 1 ∗ ∗ Uαa Uα a Uβb Uβ b = 2 δαβ δab δα β δa b + δαβ δab δα β δa b M −1 1 δαβ δab δα β δa b + δαβ δab δα β δa b .58)
The method to derive these result is to utilize f (U) = f (U0 U) = f (UU0 ) . We skip the derivation and simply list the ﬁrst few moments of a random unitary matrix of dimension M = 2N : Uαβ = 0. The ﬁrst term in Eq. (7. of scattering matrices S is uniform in the group of unitary matrices of size 2N × 2N.57)
(7. − M (M 2 − 1) (7. where U ∈ U(2N ). This claim can also be justiﬁed by “entropy” considerations. dot boundaries or Fermi energy.122
CHAPTER 7. The averaged over diﬀerent geometries could be an average over positions of impurities. in sense that it is the distribution which maximizes the entropy and hence is the ensemble with “maximal randomness”.56) (7. Here we will not be concerned with the microscopic justiﬁcation for the ensemble averaging. 1 ∗ δαβ δab . but simply say that since we have no information about the scattering matrix the most sensible thing to assume is that all scattering matrices in U(2N ) will appear with equal probability only subject to normalization conditions and symmetry constraints. These correlations however become less important in limit of large M .5: Disordered quantum dot geometry. TRANSPORT IN MESOSCOPIC SYSTEMS
Disordered quantum dot
Impurity
Figure 7.58) is equivalent to assuming the real and imaginary parts of Uαa to be independent. The TR case can be realized by writing S = UUT .
. while the last term corrects for that because the unitarity condition gives some constraints on the elements of U. For the time-reversal symmetry case. we are therefore restricted to symmetric members of U(2N ). denoted U(2N ).

4. B = 0. which we saw from the arguments leading to Eq. Also mesoscopic systems exhibit weak localization. (7. DISORDERED MESOSCOPIC SYSTEMS
123
7.e. .62) (7.59)
The result now depends on whether time-reversal symmetry is present or not. for B = 0.2
Weak localization in mesoscopic systems
In Sec. (7.61) h
n=1 m=N +1 i=1 j=1
and now applying Eq.60) G B=0 = h 2N h 2 The case B = 0 means that in addition to unitarity S is also symmetric. u and we can ﬁnd the weak localization correction not for an individual sample but for an ensemble of samples using random matrix theory of the S-matrix.e.57) directly 2e2 N 2e2 2 1 N = . the series connection between two leads with N channels δG G N = 2 − = 2e2 /h 2e /h 2
N − 2(2N +1) 0
. if a B-ﬁeld is applied or not. hence the conductance is given by the S-matrix through the Landauer-B¨ttiker formula.
.58). Writing S = UUT we get N 2N 2N 2N 2e2 ∗ ∗ G B=0 = Umi Uni Umj Unj . First take the case of broken time-reversal symmetry. (7. The origin of this eﬀect was found to be the constructive interference between time-reversed pairs of paths beginning and ending in the same point in space. (7. for B = 0. (7. The average conductance is therefore 2e2 G = Tr t† t h 2e2 = h
N 2N ∗ Smn Smn . It is important to realize that the weak localization correction survives ensemble averaging.64)
This result clearly shows that quantum corrections. The transition happens when the ﬂux enclosed by a typical trajectory is of order the ﬂux quantum. we have G
B=0
= =
2e2 h
N
2N
2N 2N
(δij + δmn δij ) 1 −
n=1 m=N +1 i=1 j=1
1 2N
1 4N 2 −1
(7. n=1 m=N +1
(7.4 studied the weak localization in self-averaging macroscopic samples.63)
2e2 1 2N 3 2−1 h 4N
1−
1 2N
=
2e2 N 2 .7. (7.55). i. give a reduced conductance and that the quantum coherence is destroyed by a magnetic ﬁeld. h 2N + 1
which is smaller than the B = 0 result. It is natural to compare the conductance with the classical conductance i.
(7. In this case there is no other constraints on S than that it is unitary and there we can use Eq. Of course in reality the transition from the B = 0 to the ﬁnite B-ﬁeld case is a smooth transition. which comes from the last term in Eq. In this case the coherence length is larger than the sample.58).4. 14.

3
Universal conductance ﬂuctuations
The ﬂuctuations of the conductance contains some interesting information about the nature of the eigenstates of a chaotic system. one would expect that if the average conductance is G = N0 (2e2 /h). Naively. In the following we calculate the ﬂuctuations of G using the statistical RMT for the S-matrix as outlined above.
x
(7. This is called the Wigner surmise. we recover the Poisson distribution result. TRANSPORT IN MESOSCOPIC SYSTEMS
7. They were observed experimentally around 1980 and explained theoretically about ﬁve years later.67)
For f constant. In fact one can show from RMT that the statistical measure vanishes when two eigenvalues coincide. Given an eigenvalue x = 0. The ﬂuctuations of the number √ of eigenvalues in a given interval is therefore far from 1/ N . the probability for the next eigenvalue to be at x can be shown to be P (x) = π π x exp − x2 . for N
1 (δnn + δmm ) . we do not expect degeneracies to occur.65) after suitable normalization. corresponding to N0 open channels. (7. then the ﬂuctuations in the number of open channels √ √ would be N0 . then P (x) dx is the probability of ﬁnding an eigenvalue at x.4.66)
and hence P (x) = f (x) exp −
0
dx f (x ) . we get Eq. f (x) dx.
n=1 m=N +1 n =1 m =N +1 2 N 2N N 2N
= =
1 4N 2
2
n=1 m=N +1 n =1 m =N +1 2 4 2 2
−1 1+
1 + δmm δnn − 1 4N 2 . x]:
x
P (x)dx = exp −
0
dx f (x ) f (x)dx. For a completely random system without any symmetries.124
CHAPTER 7. The ﬂuctuation of the conductance in the non-TRS case are G2
B=0
=
2e2 h 2e2 h 2e2 h
2 N
2N
N
2N ∗ ∗ Smn Smn Sm n Sm n .
(7. Assuming “linear repulsion” f (x) ∝ x.68)
N ≈ 4N 2 − 1
2e h
N 2
. which is the physical reason for the “universal” behavior. Suppose that the probability of ﬁnding an eigenvalue in dx is f (x) dx. 2N 1 (7. so that δG = (2e2 /h) N0 . which has given rise to the name universal conductance ﬂuctuations (UCF). times the probability that there was no eigenvalues in the interval [0. the reason being that the transmission probabilities are not independent. Historically the study of these ﬂuctuations were the ﬁrst in the ﬁeld of mesoscopic transport. 2 4 (7.65)
for the case with time-reversal symmetry. and a suggestive derivation is as follows. The number of conducting channels in a given energy window does therefore not follow a Poisson distribution. This is not seen experimentally. It is an experimental fact that the ﬂuctuations turn out to be independent of the size of the conductance itself.

The inset shows the geometry of the quantum dot. which has additional gates by which the shape can be changed. 3876 (1995). Lett. 8
for N
1. Rev.70)
The variance is thus independent of the average value of G and furthermore it is expected to decrease by a factor of 2 when a magnetic ﬁeld is applying. (Cambridge University Press).
7.
(7.69)
A similar calculation for the B = 0 case gives
B=0 (2e2 /h)2
δG2
1 ≈ . Electronic transport in mesoscopic systems. SUMMARY AND OUTLOOK
125
Figure 7.
. After Chan et al.
(7. Datta. 7. 16
for N
1. and the variance is
B=0 (2e2 /h)2
δG2
≈
1 . Text books: 1.6: Variance of the conductance of a quantum dot as a function of magnetic ﬁeld. The decrease of the variance when the time-reversal symmetry is broken by the magnetic ﬁeld is clearly seen and the decreases by approximately a factor of 2 is in agreement with the theory. 1995. The trace is taken at 30 mK.5
Summary and outlook
Below we list a few text books and review papers about mesoscopic physics. 74.6.5. Phys..7. Indeed this is what is seen experimentally for example as shown in Fig. S.

Solid State Physics 44. 1995. Review of Modern Physics 69. (Academic Press). Quantum transport in semiconductor nanostructures. Conductance quantisation in metallic point contacts. Turnbull. C.J. 4. Beenakker. 1991.126
CHAPTER 7. van Houten.M. cond-mat/9910394. 3.M. H. Review papers: 1. Ehrenreich and D. Review of Modern Physics 72.K. Y. J. eds. Alhassid. 2. 895 (2000)
. 731 (1997). Random-matrix theory of quantum transport. The statistical theory of quantum dots. Ferry and S. C.J.W. (Cambridge University Press).W. D. Goodnick. TRANSPORT IN MESOSCOPIC SYSTEMS
2. Beenakker and H. Transport in nanostructures. van Ruitenbeek.

1) is easily veriﬁed by letting 2 act directly on the r integrand and then use Eq. (8.6) 4πε0 |r − r | 127
. 2 . which immediately gives −k 2 G(k) = 1 and hence G(r) = ⇒ G(k) = − 1 .2) r G(r) = δ(r).5)
When inserting this into (8. the electrical potential r follows as 1 dr G(r − r )ρe (r ).Chapter 8
Green’s functions
8. The easiest way to ﬁnd G(r) is by Fourier transformation. (8. To illustrate the idea of the method let us consider the familiar problem of ﬁnding the electrical potential φ given a ﬁxed charge distribution. where δ(r) is the Dirac delta function. we want to solve Poisson’s equation 1 2 φ(r) = − ρe (r). k2 (8.e.3) φ(r) = − ε0 That this is a solution to Eq.2). It has a long history with numerous applications. ρe . i. (8.1) ε0 It turns out to be a good idea instead to look for the solution G of a related but simpler diﬀerential equation 2 (8. This is a good idea because once we have found G(r). 3 k2 (2π) 4πr
(8.1 “Classical” Green’s functions
The Green’s function method is a very useful method in the theory of ordinary and partial diﬀerential equations.4)
dk ik·r e G(k) = − (2π)3
dk eik·r 1 =− .3) we obtain the well-known potential created by a charge distribution 1 ρe (r ) φ(r) = dr . (8. G(r) is called the Green’s function for the Laplace operator. (8.

(8. E (8. E) = E − H0 (r) or G−1 (r.10) (8. where the inverse Green’s function is a function of two arguments.11) (8. The interaction induces transitions from the incoming state to diﬀerent outgoing states. r ) = G0 (r .11) into the G−1 ψE term of Eq. E) = δ(r − r ).9)
This is veriﬁed by inserting ψE from Eq. 0
1
.11) by iteration. (8.e. GREEN’S FUNCTIONS
8. there is a continuum of states and hence we are free to choose any E. The Green’s
R In order to emphasize the matrix structure we could have written this as dr G−1 (r. r ) = (E − H0 (r))δ(r − r ).7)
where we know the eigenstates of H0 .128
CHAPTER 8. 0 and by inspection we see that the solution may be written as an integral equation ΨE (r) = Ψ0 (r) + E dr G0 (r. r . E) − V (r) ΨE = 0. i. r ) G0 (r . G0 (r. (8. The procedure outlined below is then a systematic way of calculating the eﬀect of the interaction between the “beam” and the “target” on the outgoing states. (8. r . Here we consider the case of an open system. One can now solve the integral equation Eq. r .12)
where Ψ0 is an eigenstate to H0 with eigenenergy E. E)V (r )Ψ0 (r ) + O V 2 . E) = δ(r − r ).10) and 0 then using Eq. 0 0 Now the Schr¨dinger equation can be rewritten as o G−1 (r. The next leading terms are also easily found by continuing the iteration procedure. (8.8)
with the boundary condition. (8. In order to solve the Schr¨dinger equation. r). Consider for example the Schr¨dinger equation o [H0 (r) + V (r)] ΨE = EΨE . It is natural to identify the operator [E − H0 (r)] as the inverse of G0 (r. This situation is relevant for scattering problems where a ﬂux of incoming particles (described by H0 ) interacts with a system (described by V ). But in the r-representation it is in fact diagonal G−1 (r. E)V (r )ΨE (r ). r ) and therefore we write1 G−1 (r. (8.2
Green’s function for the one-particle Schr¨dinger equao tion
Green’s functions are particular useful for problems where one looks for perturbation theory solutions. we deﬁne the corresponding Green’s funco tion by the diﬀerential equation [E − H0 (r)] G0 (r. E) G0 (r.9). and where we want to treat V as a perturbation. What we have generated by the E iteration procedure is nothing but the ordinary (non-degenerate) perturbation theory. r ) = 0 δ(r − r ). and up to ﬁrst order in V the solution is ΨE (r) = Ψ0 (r) + E dr G0 (r. r .

17)
which both can be shown by inspection. The time dependent Schr¨dinger equation is o [i∂t − H0 (r) − V (r)] Ψ(r.15a) (8.13) Ψ(r. t.11) we can iterate the solution and get Ψ = Ψ0 + G0 V Ψ0 + G0 V G0 V Ψ0 + G0 V G0 V G0 V Ψ0 + · · · = Ψ0 + G0 + G0 V G0 + G0 V G0 V G0 + · · · V Ψ0 . G0 . t) = Ψ0 (r. t) = i∂t − H0 (r) 0 G
−1
(8. t ) = δ(r − r )δ(t − t ). we continue to study the case of non-interaction particles some more and include time dependence. Similar to Eq.17). (8. t ) = δ(r − r )δ(t − t ). t. (8. t ). as simple building blocks from which the solutions of more complicated problems can be build.
From these building blocks we easily build the solution of the time dependent Schr¨dinger o equation. t) = Ψ0 (r. First we observe that the following self-consistent expression is a solution to Eq. r . (8.2.14b)
(8. GREEN’S FUNCTION FOR THE ONE-PARTICLE SCHRODINGER EQUATION129 function method is thus useful for this kind of iterative calculations and one can regard the Green’s function of the unperturbed system. (8.20)
.1. Again we consider the case where the Hamiltonian has a free particle part H0 of some perturbation V . t) = 0. (8.19)
Noting that the last parenthesis is nothing but G itself we have derived the so-called Dyson equation G = G0 + G0 V G.8) we deﬁne the Green’s functions by [i∂t − H0 (r)] G0 (r.18)
where the integration variables have been suppressed. As for the static case in Eq. t )V (r )Ψ0 (r . t) + dr dt G(r. Before we introduce the many-body Green’s function in the next section. The inverse of the Green’s functions are thus G−1 (r. see Exercise 7. t) + dr dt G0 (r. t ).¨ 8.13)
(8. t.16)
or in terms of the full Green’s function Ψ(r. (8.14a) (8. r . (8. r . we see that the full Green’s function G is given by G = G0 + G0 V G0 + G0 V G0 V G0 + · · · = G0 + G0 V G0 + G0 V G0 + · · · . t) = i∂t − H0 (r) − V (r). [i∂t − H0 (r) − V (r)] G(r. H = H0 + V . r . t )V (r )Ψ(r . (8. By comparison with Eq. t. (8.15b)
(r.

The eigenstates are for this example thus given by ψ(k) = exp(ikx).
(8. given that it was in the state |r at time t . r t ) =
nn
(8.24)
If we choose the basis state |φn as the eigenstates of the Hamiltonian. (8. t exp(ikx).14b) with G0 from the left. t) = dr dt G(rt.22) which indeed is a solution of the partial diﬀerential equation deﬁning the Green’s function.14b). r t ) = −iθ(t − t ) r|e−iH(t−t ) |r .21)
which can be checked by inserting Eq. That the Green’s function is nothing but a propagator is immediately clear when we write is it as G(rt. then the Green’s function becomes G(rt. (8. t) we have deﬁned here is the non-interaction version of the retarded single particle Green’s function that will be introduced in the following section. It is also often called a propagator because it propagates the wavefunction. n t ) φn |r .22) the Green’s function expresses the amplitude for the particle to be in state |r at time t. We could of course calculate the propagator in a diﬀerent basis.21) into the Schr¨dinger equation and using the o deﬁnition Eq. t ). (8.26)
. r t ) = −iθ(t − t )
n
r|φn φn |r e−iEn (t−t ) . (8. (8. see Exercise 7.g.2. Consider an electron incident on a barrier. n t ) = −iθ(t − t ) φn |e−iH(t−t ) |φn . (8. located between x > 0 and x < L. Eq. i. Looking at Eq. suppose it was in a state |φn and time t then the propagator for ending in state |φn is G(nt. As a simple example we end this section by a typical scattering problem in one dimension.14b). for x > L. (8. (8. the incoming wave is for x < 0 given by exp(ikx) while the outgoing wave on the other side x > L is given t exp(ikx). (8.23)
r|φn G(nt. e. if the wavefunction is know at some time then the wavefunction at later times is given by Ψ(r. r t )Ψ(r .e. The Green’s function G(r. the proof being left as an exercise. for x < 0. The Dyson equation can also be derived directly from Eqs. Here t is the transmission amplitude.25)
Propagation from one point to another in quantum mechanics is generally expressed in terms of transmission amplitudes.130
CHAPTER 8.
(8. GREEN’S FUNCTIONS
This equation will play and important role when we introduce the Feynman diagrams later in the course.14) by multiplying Eq. The Green’s function are related by a simple change of basis G(rt.

30b)
We see that the retarded Green’s function can be written in terms of these two functions as GR (rσt. · · · ]B. Ψ† (r t )]B. t. Even though we call these Green’s functions for “single-particle Green’s functions”.3
Single-particle Green’s functions of many-body systems
In many-particle physics we adopt the Green’s function philosophy and deﬁne some simple building blocks. (8.25) we see that the Green’s function for the x > L and x < 0 precisely describes propagator across the scattering region becomes G(xt. The second type of single-particle Green’s functions is the so-called greater and lesser Green’s functions G> (rσt. r σ t ) = −iθ t − t [Ψσ (rt). B} = AB + BA. t ). from which we obtain solutions to our problems. (8.
8.F is deﬁned as [A.2. r σ t ) . x t ) = t G0 (x. (8. The Green’s functions contain only part of the full information carried by the wave functions of the systems but they include the relevant information for the given problem. they are truly many-body objects because they describe the propagation of single particles
.27)
where G0 is the Green’s function in the absence of the scattering potential. SINGLE-PARTICLE GREEN’S FUNCTIONS OF MANY-BODY SYSTEMS
131
When this is inserted into Eq. The retarded Green’s function is deﬁned as GR (rσt. See also Exercise 10. There are various types of single-particle Green’s functions. [A.28) and the one for the propagator for the one particle wavefunction. x > L and x < 0. r σ t ) = −i Ψσ (rt)Ψ† (r t ) . B]F = {A.F . in Eq. For non-interacting particles they are indeed identical. (8. B]B = [A. σ G< (rσt.3.30a) (8.28)
where the (anti-) commutator [· · · . r σ t ) = θ t − t G> (rσt.8.22). σ B : bosons F : fermions (8. x . When we deﬁne the many-body Green’s functions it is not immediately clear that they are solutions to diﬀerential equations as for the Schr¨dinger equation Green’s o functions deﬁned above. also called Green’s functions. r σ t ) − G< (rσt. But as you will see later they are in fact solutions of equations of motions with similar structure justifying calling them Green’s functions. σ (8. (8. B] = AB − BA. Let us simply carry on and deﬁne the diﬀerent types of Green’s functions that we will be working with. r σ t ) = −i (±1) Ψ† (r t )Ψσ (rt) .29)
Notice the similarity between the many-body Green’s function Eq. From this example it is evident that the Green’s function contains information about the transmission amplitudes for the particle.

σt.32)
8. Consider therefore the Hamiltonian for free electrons (or other fermions) H=
kσ
ξkσ c† ckσ . The relation between the real space retarded Green’s function and the corresponding one in a general |ν -basis as deﬁned in Eq. The reason is that they give the amplitude of a particle inserted in point r at time t to propagate to position r at time t. [aνσ (t). because the right hand side cannot explicitly dependent on the origin and on r .3.
kk
(8.31)
where GR (νσt.34b)
GR (k.k ) = δk. kσ
(8.
8.35)
. allowing us to write GR (r − r . ν (8. ν σ t ) = −iθ t − t and similarly for G> and G< . Since the system is translation-invariant G(r.30a). σ t ) = 1 V eik·r GR (kσt. The Green’s functions in Eqs.28). GREEN’S FUNCTIONS
governed by the full many-body Hamiltonian.
k
(8.F . (1.30b) are often referred to as propagators. σt. kσ
The other types of Green’s functions have similar forms.2
Green’s function of free electrons
A particular case often encountered in the theory of quantum liquids is the simple case of free particles.33)
However. (8. a† (t )]B.132
CHAPTER 8. it follows that G(k. Therefore the single-particle functions can include all sorts of correlation eﬀects. σ t ) = −iθ t − t
[akσ (t). σ t ) = 1 V eik·(r−r ) GR (k. ν σ t )ψν (σ r ). r ) can only depend on the diﬀerence r − r and in this case GR (r − r . In this sense GR has its name “retarded” because it is required that t > t . a† σ (t )]B.34a) (8.F . σ t ). k σ t )ei(k−k )·r .k G(k). σr t ) =
νν ∗ ψν (σr)GR (νσt.
(8. and (8.1
Green’s function of translation-invariant systems
For a system with translation-invariance the usual k-representation is a natural basis set. (8.3. k σ t )e−ik ·r . σt. σt.71) is GR (σrt.
kk
1 = V
eik·(r−r ) GR (kσt.

Here |G denotes the groundstate of the free electrons. ρ = |r − r |. we can evaluate G< and ﬁnally GR 0 0 G> (kσ. because it is proportional to the number of electrons. t − t ) = −i G|ck (t)c† (t )|0 = −i G|ck e−iH(t−t ) c† |G eiE0 (t−t ) .37)
and similarly c† (t) = c† eiξk t . t − t ) = −i(1 − nF (ξk ))e−iξk (t−t ) . 0 k k (8. This is intuitively clear from Eqs. we of course have ckσ c† = 1 − nF (ξk ). (8. t − t ) = inF (ξk )e−iξk (t−t ) . By Fourier transforming from the time domain to the frequency domain.3. 0 G< (kσ. i.40)
which precisely is the overlap between a state with an added electron in state k and with a state with an added electron in k and allowing time to evolve from t to t. k .38)
and because the Hamiltonian is diagonal in k and the occupation of free electrons is given by the Fermi-Dirac distribution. t − t ) = −i ckσ c† e−iξk (t−t ) . kω ρ 2m
(8. (8. which we denote G> to indicate that 0 it is the propagator of free electrons. the electron propagator is in the frequency domain G> (kσ. Because the Hamiltonian is diagonal in the quantum numbers k and σ so is the Green’s function and therefore G> (kσ. we get information about the possible energies of the propagating particle. In exactly the kσ same way.39a) (8. t 0 − t ) = −iθ(t − t )e
−iξk (t−t )
(8.
(8. because it requires an empty state while G< gives the propagation of holes. An easy way to remember this is to realize that the factor k k e−iHt to the right of ck must have one more electron in state k than eiHt to the left of ck . This is perhaps more clearly seen if we write the T = 0 deﬁnition of for example G> 0 G> (k. SINGLE-PARTICLE GREEN’S FUNCTIONS OF MANY-BODY SYSTEMS
133
and the corresponding greater function in k-space.39c)
We see that G> gives the propagation of electrons.e.8. 0 GR (kσ. 0 kσ (8. ω) 0 = −2πi dk (1 − nF (ξk ))eik·(r−r ) δ (ξk − ω) (2π)3 2 sin(kω ρ) kω = d(ω) (1 − nF (ω)) . the ﬁlled Fermi sea. = ω. ω) = −2πi [1 − nF (ξk )] δ (ξk − ω) . Now G> becomes G> (kσ.41) 0 The corresponding r-dependent propagator. |G = |FS .39) because the propagators evolve periodically in time with the period given by the energy of the electron. t − t ) = −i ckσ (t)c† (t ) . For example.39b)
. 0 kσ (8.42)
.24)) ckσ (t) = eiHt ckσ e−iHt = ckσ e−iξk t . which expresses propagation of a particle in real space is given by G> (r − r . (5.36)
Because of the simple form of the Hamiltonian we are able to ﬁnd the time dependence of the c-operators (see Eq. (8.

3
The Lehmann representation
A method we will often be using when proving formal results is the so-called Lehmann representation.31). which is just another name for using the set of eigenstates. H.45)
dt ei(ω+iη)t
1 Z
e−βEn
nn
n|cν |n
n |c† |n ei(En −En )t ν
+ 1 = Z = 1 Z
n|c† |n ν e
−βEn
n |cν |n e−i(En −En )t n|cν |n n |c† |n n|c† |n n |cν |n ν ν + ω + En − En + iη ω − En + En + iη .46)
nn
nn
n|cν |n n |c† |n ν e−βEn + e−βEn ω + En − En + iη
.134
CHAPTER 8. t. ω) = −i
0 ∞
(8.3. ω) = −2πi Z e−βEn n|aν |n
nn
n |c† |n δ(En − En + ω). the availability of an empty state (1 − nF ). the interference function sin (x) /x that gives the amplitude of a spherical wave spreading out from the point r . Let us for example study the diagonal Green’s function. If we insert 1 = n |n n| we get G> (ν.3.
e−βEn n |c† |n n|cν |n δ(En − En − ω).
8. ν
(8.43)
In the frequency domain. The retarded Green’s function becomes (again for fermions) GR (ν. The propagation from point r to r of a particle with energy ω is thus determined by the density of states. ω)e−βω . See also Exercise 7. ν
(8. of the full Hamiltonian. (8. c) G< (ν. we obtain G> (ν. GREEN’S FUNCTIONS
where d(ε) = m3/2 ε/2/π 2 is the density of states per spin in three dimensions. d. ω) = = 2πi Z 2πi Z e−βEn n|c† |n ν
nn
n |cν |n δ(En − En − ω).44)
In the same way we have (for fermions. G> (νtνt ). see also Eq. as basis set. ν
nn
= −G> (ν. ν
nn
2πi = Z
e−β(En +ω) n |c† |n n|cν |n δ(En − En − ω). t ) = −i cν (t)c† (t ) = −i ν 1 = −i Z e
nn −βEn
1 Z
n|e−βH aν (t)a† (t )|n ν
n
n|cν |n
n |a† |n ei(En −En )(t−t ) . (2. {|n }.

we get
135
2 Im GR (ν. For example if we look at the retarded propagator for free electrons in Eq.4
The spectral function
The spectral function A(ν. ω). which tells us that an excitation with energy ω can only happen by adding an electron to the state k given by ξk = ω. non-interacting system. ω) = −2 Im GR (ν.3.4
<
(8. ω) = A(ν. ω) [1 − nF (ω)] . ω) = A(ν.e.47)
n|cν |n
nn
n |c† |n e−βEn (1 + e−βω )δ (ω + En − En ) . ω) = 2πδ(ω − ξk ). Again the spectral function is given by a simple delta function A0 (ν. ω). 0 (8.39c) GR (kσ. i.50a) (8. (8. (8. ω) can be thought of as either the quantum state resolution of a particle with given energy ω or as the energy resolution for a particle in a given quantum number ν.52)
Thus for the idealized case of non-interaction free electrons.48)
= −i(1 + e−βω )G> (ν. see Exercise 7.50b)
8. ω − ξk + iη
the corresponding spectral function is A0 (kσ. ω) = 2πδ(ω − ξν ).3. the spectral function is a delta function. This result is true for any quadratic Hamiltonian.53) ν
ν
where ν labels the eigenstates of the system. (8. Deﬁning the spectral function A as A(ν. ν (8.54)
. −iG (ν. ω) = −2 Im GR (kσ. It gives an indication of how well the excitation created by adding a particle in state ν can be described by a free non-interacting particle. we have derived the important general relations iG> (ν.8. as expected. SINGLE-PARTICLE GREEN’S FUNCTIONS OF MANY-BODY SYSTEMS
1 Taking the imaginary part of this and using (ω + iη)−1 = P ω − iπδ(ω). Similar relations hold for bosons.51)
1 . ω) = − =−
2π Z 2π Z
n|cν |n
nn
n |c† |n ν
e−βEn + e−βEn
δ (ω + En − En )
(8. If we for example have H0 = ξν c† cν . ω)nF (ω).49)
(8. ω) = −i 0 =
∞ −∞
dtθ(t − t )eiω(t−t ) e−iξk (t−t )η(t−t ) (8.

g. electronphonon interaction. ω)nF (ω). One possible mechanism of broadening in a metal is by e. the spectral function is similar to the density of states at a given energy. ω) = 1. ω) is also small compared to EF then the Fermi function in (8.3. This follows from Eq.50b) nν = c† cν = −iG< (ν. ω) = − 2π
dω 2 Im GR (ν. GREEN’S FUNCTIONS
Generally. Firstly. see above. e. Eq. ω) −∞ 2π ∞ 1 = dω n|cν |n n |c† |n ν Z −∞
nn
∞
e−βEn + e−βEn
× δ (ω + En − En ) 1 = n|cν |n n |c† |n ν Z =
nn cν c† ν
e−βEn + e−βEn (8. (8.5
Broadening of the spectral function
When interactions are present the spectral function changes from the ideal delta function to a broadened proﬁle. This is evident since the occupation nν of a given state ν is for fermions given by (8. In Chap. Secondly. because if εν EF and the width of A(ν.56)
+ c† cν = cν c† + c† cν = 1. the expected result follows. the deﬁnition of the spectral function. ω) integrates to 2π. We of course expect that if the state |ν is far below the Fermi surface. t = 0) ¯ ν ∞ dω < G (ν. ω) = −i −∞ 2π ∞ dω A(ν. but it may still be a peaked function. ν ν ν
where the last equality follows from the Fermi operator commutation relations.57)
The physical interpretation is that the occupation of a quantum state |ν is an energy integral of the spectral density of single particle states projected onto the state |ν and weighted by the occupation at the given energy.47)
∞ −∞
dω A(ν. it is always positive as one must require. which redistributes the spectral weight because of energy exchange
. Furthermore. due to interactions the spectral function diﬀers from a delta function. which then indicates that the non-interacting approximation is not too far from the truth. then c† cν ≈ 1.55) −∞ 2π This formula is easily derived by considering the Lehmann representation of −2 Im GR in Eq. = −∞ 2π
(8.
8.49) and the fact that n |cν |n n |c† |n = | n |cν |n |2 .57) is approximately unity and since A(ν.g. ν it obeys the sum rule ∞ dω A(ν. This in fact follows from ν the sum rule. εν EF . 13 this is discussed in much more detail. (8. (8. We will now show that the spectral function is a like a probability distribution.136
CHAPTER 8.46). (8.

8. However. The reason for this will be discussed later in the Chap. as we saw in the previous chapter.8. In principle there is only one way to measure the single particle properties. which is to insert/remove a single electron into/out of a many-body system.1. (ω − ξν )2 + (1/τ )2 (8. both being two particle propagators. which is reﬂected in a broadening of the spectral function.
. Another mechanism for broadening is the electron-electron interaction. There are not too many ways for doing this because most experiments measure density or other two-particle properties. This can be achieved by a so-called tunnel junction device or by subjecting the sample to a beam of electrons.4
Measuring the single-particle spectral function
In order to probe the single-particle properties of a many-body system. which. a freely propagating state outside the material. This is illustrated in Fig. (8. the free electron picture is still a good distribution in many cases and in particular for metals. which is quite surprising since the Coulomb interaction between the electrons is a rather strong interaction.4. As a simple example we consider a Green’s function which decays in time due to processes that scatters the particle out of the state ν. The simple notion of single electron propagators becomes less well deﬁned for interacting systems. In the following we study in detail the tunneling case where an electron tunnels from one material to the other and show how the tunneling current is expressed in terms of the spectral functions and thus provides a direct measurement of these. For example the response to an electromagnetic ﬁeld couples to the charge or current. For example when a photon is absorbed and an electron is kicked out from an occupied state to e. measures charge-charge or current-current correlation functions. Such a decaying Green’s function corresponds to a ﬁnite width of the spectral function
∞
A(ν.
8. t) ≈ −iθ(t)e−iξν t e−t/τ . 13. See Chap. a solid state sample say. one must have a way of measuring how the electrons propagate as a function of energy.g. ω) = −2 Im
−∞
dteiωt GR (ν.58) where τ is the characteristic decay time. in some cases also optical experiments approximately measures the single particle density of states. t) ≈ 2 Im i
0
∞
dteiωt e−iξν t e−t/τ =
2/τ . In practice this means taking out or inserting a particle with deﬁnite energy. 13 on Fermi liquid theory.59)
Thus the width in energy space is given by τ −1 . MEASURING THE SINGLE-PARTICLE SPECTRAL FUNCTION
137
between the electron and the phonon system. Amazingly.4. In this situation the retarded Green’s function becomes GR (ν.
8.1
Tunneling spectroscopy
The tunnel experiment set-up consists of two conducting materials brought into close contact such that electrons can tunnel from one to the other.

µ .µ
This is the most general one-particle operator which couples the two systems.
The current passing from 1 to 2 is driven by a shift of chemical potential diﬀerence.138
CHAPTER 8. N1 ] = i[H12 .1: Measurement setup for the tunnel experiment.µ
≡ −i(L − L† ).ν .µ 1.62)
% $ #¢ " 0 & % "$ )' & (
¡ £££ £££££££ £ ££ §¨© ¦ ¤ ¥ ¤ ¤ ¤ ¤ ¤ 4 ¤ ¤ ¤ ¤ ¤ ¤ ¤ ! ¨ ¨2¨ 31 ¢
¤¤¤¤ ¤¤ ¤ ¤ ¤ ¤ ¤ ¤ ££ £ £ £ £ £ §¨© £ ¦ £¥ £ £ ££ ¤
§
2¨ ¨ )1
. The coupling between the system is assumed to be very weak.61) (8. and hence I = i[H. H1 and H2 . GREEN’S FUNCTIONS
Figure 8.60) (8. Ie = −e I . Two systems are brought into close contact.ν and c2. ˙ where I = N1 . c† c1. which gives rise to a term in the Hamiltonian of the form H12 =
νµ ∗ Tνµ c† c2.ν 2. The tunnel matrix element is deﬁned as Tνµ =
∗ dr ψν (r)H(r)ψµ (r). Therefore we calculate the current to lowest order in the coupling. c1. The current through the device is deﬁned by the rate of change of particles. The coupling between the two sides of the junction is due to the ﬁnite overlap of the wavefunctions.µ + Tνµ c† c1.g.
with H(r) being the (ﬁrst quantization) one-particle Hamiltonian.µ + Tνµ c† c1. Systems 1 and 2 are described by their respective Hamiltonians. since the tunnel matrix element is exponentially suppressed with distance between the two systems.ν 2.µ 1.ν 1. In the tunneling Hamiltonian this is modelled by the matrix element Tνν .ν
= −i
νµ
Tνµ c† c2.ν
−
∗ Tνµ c† c1. which means that µ1 = µ2 . e. separated by an insulating material. involving electron operators. This
$ 0 & % "$ 0 ' &
(8. The right panel illustrates the electron wavefunctions in the two subsystems which have a small overlap in the insulator region.ν . N1 ] = i
νµ ν ∗ Tνµ c† c2.ν 2. The current operator itself is already linear in Tνµ and therefore we need only one more order. an oxide or for the so-called scanning tunneling microscope (STM) simply vacuum. 1.

L(t )
∗ Tνµ Tν µ νµ ν µ
eq
dt θ(t − t )
−∞ ∞
c† (t)ˆ1.66b)
with the time dependence of c being given by the Hamiltonian with a common chemical ˜ potential µ. Furthermore.µ c
eq
− c† (t )ˆ1. we are of course allowed to choose a basis set where the Green’s
This is in fact not true for superconductors which are characterized by having a spontaneous breaking of the symmetry corresponding to the conservation of particles and therefore such two-particle tunnel processes are allowed and give rise to the so-called Josephson current.64)
involves terms of the form t . t ).
(8. ˆ ˜ (8. According to the general Kubo formula derived in chap.µ c ˆ1.63a) (8.
eq
= 2 Re
−∞
dt θ(t − t )
νµ ν µ
∗ Tνµ Tν µ
eq
c† (t)ˆ2. L(t )
eq
+ c.ν (t).ν (t)ˆ† (t ) ˆ c1.ν 1.µ
eq
(8.µ (t) c† c2. L(t ) −
eq
= −θ(t − t ) Now the combination ˆ ˆ L(t).8.µ 1.63b)
R ˆ ˆ CIp H12 (t − t ) = −iθ(t − t ) [Ip (t).
2
. MEASURING THE SINGLE-PARTICLE SPECTRAL FUNCTION
139
means that linear response theory is applicable. Naturally the number of particles is a conserved quantity and matrix elements of this type must vanish. L(t )
eq
ˆ ˆ L† (t).µ (t ) ˆ2.ν (t) ˆ1.2 We are therefore left with
∞
Ip (t) = 2 Re = 2 Re
dt θ(t − t )
−∞ ∞
ˆ ˆ L† (t).µ (t )ˆ† (t) ˆ c2.µ (t ) ˆ2.ν
eq
with two electrons created in system 1 and two electrons annihilated in system 2 and therefore is does not conserve the number of particles in each system.4.ν .
c† c2. eq
(8. . c† (t )ˆ2. 6 the particle current is to ﬁrst order in H12 given by
∞
I (t) =
−∞
R dt CIp H12 (t. ˆ ˜ c2 (t) = c2 (t)e−i(−e)V2 t . The correlation function CIH12 can be simpliﬁed a bit as
R CIp H12 (t − t ) = −θ(t − t )
ˆ ˆ ˆ ˆ L(t) − L† (t).ν c c1.c. L(t ) + L† (t ) ˆ ˆ L(t).ν c
eq
c2.66a) (8.65)
Now the time dependence due to the shift in energy by the applied voltages is explicitly pulled out such that c1 (t) = c1 (t)e−i(−e)V1 t . H12 (t )]
where the time development is governed by H = H1 + H2 .

140
CHAPTER 8. and that the density of states peaks near the gap. and we ﬁnally arrive at
∞
Ip =
−∞
dω 2π
|Tνµ |2 A1 (ν.71)
At low temperatures where the derivative of the Fermi function tends to a delta function and (8. ω) A2 (µ. The ν νν particle current then becomes (after change of variable t → t + t)
0
Ip = 2 Re
dt
−∞ νµ
|Tνµ |2 ei(−e)(V1 −V2 )t G> (ν. ω + eV )[nF (ω + eV ) − nF (ω)]. ω + eV ) − G< (ν.69) we see that the current is determined by two factors: the availability of states. which is illustrated in Fig. |Tνµ |2 A2 (ν. ω)G> (µ. ω + eV ) . Therefore by sweeping the voltage across the junction one gets information about A(ν.3 and Exercise 7. (8. ω). The lesser and greater Green’s functions are now written in terms of the spectral function. GREEN’S FUNCTIONS
function of the decoupled system (i. −t )G> (µ. 8. G> = δνν G> .68) 1 2 1 2 −∞ 2π νµ with the voltage given by V = V2 − V1 .
µ
(8. ω). t ) . (8.
. 1 2 1 2
(8.71) becomes dI ∝ A1 (ν. without H12 ) is diagonal. −eV ).
ν
(8.69)
In Eq. given by the diﬀerence of occupation functions. ω + eV ) ≈ const.5.50). −t )G< (µ. i. see Exercise 4.70)
then
dI ∝ dV
∞
dω −
−∞
∂nF (ω + eV ) ∂ω
A1 (ν.67) After Fourier transformation (and reinsertion of the convergence factor eηt ) this expression becomes ∞ dω Ip = |Tνµ |2 G> (ν.72) dV ν So the spectral function can in fact be measured in a rather direct way.e. ω)G< (µ. The tunnel spectroscopy technique amounts to a sweep of an external voltage which controls the chemical potential while measuring the diﬀerential conductance dI/dV . (8. t ) − G< (ν.2. and by the density of states at a given energy. If the other material is a simple material where one can assume the density of states to be more or less constant.e. This is a widely used spectroscopic principle in for example the study of superconductors where it was used to verify the famous prediction of the BCS theory of superconductivity that there is an excitation gap in the superconductor.
νµ
(8. see Eq. (8. Also it is used to study small structures such as quantum dots where the individual quantum levels become visible due to size quantization.

(6. is according to Eq. the
% $ # ¥ ¥ ¥ ¥ ¥¥ ¥ ¥ ! " ! & ¥ §§¥¥©¨§¦§¦§¦¥¥¥ ¤ ¢ ¡£ ¢
Optical spectroscopy
Two-particle correlation functions of many-body systems
&¥ ' '
. The left panel shows the two density of states in the two materials. where there is little variations with energy and the experiment can therefore be used to get information about the density of states of the left material.4. because the dissipation.39)). which was shown in Chap. (8.5. there are cases where it is well approximated by the one-particle spectral function. The two right most panels show the resulting current and the diﬀerential conductance trace.2: The principle used in tunneling spectroscopy. For example.48) given by (take for simplicity the
3
Because the power dissipated at any given point in space and time is P (r. which is the real part of the conductivity3 . we saw that the response to electromagnetic radiation was determined by the auto correlation function of the charge and current densities.8.5
While the single-particle Green’s functions deﬁned above measure the properties of individual particles the higher order Green’s functions give the response of the quantum system to processes involving several particles. we evaluate the polarization function χ = Cρρ for a non-interacting electron gas (see Eq.
8.73)
where A is some two particle operator. which was encountered in the linear response chapter.
8. Typical correlation functions that we will meet are of the type CAA (t. ω). t). A(t ) . t) · E(r.t) = Je (r. Such an example is photo emission spectroscopy. The right one is metal. It is seen how the diﬀerential conductance is a direct measure of ν A1 (ν.2
While the response to an electromagnetic ﬁeld in principle is always given by the dielectric function. (6. t ) = −iθ(t − t ) A(t). In order to treat a speciﬁc case. This function gives for example information about the dissipation due to an applied ﬁeld. 6. One important type of higher order Green’s functions are the correlation functions. TWO-PARTICLE CORRELATION FUNCTIONS OF MANY-BODY SYSTEMS141
Figure 8.

Due to the translation-invariance the result cannot depend on r and one sees that q2 = −q (or formally one can integrate over r and divide by volume to get a delta function. ω) 2π V q 2π V q
. (8. ρ(q2 . the time dependence is given by (see Eq.37)) ρ(q. when inserted into (8.76) χR (q. ω)|2 σ(q. [c† cµ . t − t ) = −iθ(t − t )e2 0 1 V nF (ξk ) − nF (ξk+q ) ei(ξk −ξk+q )(t−t ) . GREEN’S FUNCTIONS
ωe2 Im χR (q. t). ν ν ν ν and we ﬁnd χR (q. kσ
(8.
(8. t − t ) = dr χ(r − r .79) where the subindex “0” indicates that we are using the free electron approximation.ν − c† cµ δν . c† cµ ] = c† cµ δµ. ρ(−q. t)= E (q. (8. t). (1. kσ
(8. t) · Je (r. yields χR (q. (8. t) =
kσ
c† ck+qσ ei(ξk −ξk+q )t . ρ(q2 . t − t ) = −iθ(t − t )e2 0 1 V [c† ck+qσ . t). eiq1 ·r+iq2 ·r e−iq·(r−r ) .µ .142 translation-invariant case) Re σ (q. The commutator is easily evaluated using the formula. t )
q2
ei(q2 +q)·r . c† σ ck −q σ ] ei(ξk −ξk+q )t e kσ k
kk σσ i(ξk −ξk
−q
)t
. ω) = |E(q.96) ρ(q) =
kσ
c† ck+qσ .76). t − t ) = −iθ(t − t )e2 V The Fourier transform of the charge operator was derived in Eq. dr dr ρ(r. ω) · Je (q. t − t )e−iq·(r−r ) . t )
q1 q2
ρ(q.80)
total energy being dissipated is Z Z Z dω 1 X ∗ dω 1 X W = drdt E(r.75)
= −iθ(t − t ) = −iθ(t − t ) 1 = −iθ(t − t ) V
ρ(q1 . t ) .78)
which. t).77)
For free electrons. δq2 +q.0 ) and thus 1 ρ(q. t ) 1 V2 e−iq·(r−r ) .
kσ
(8. q2
(8. ω) = −
CHAPTER 8.74)
In momentum space the polarization is given by χR (q. ω). ρ(r .

(8.81)
(8. when discussing the elementary excitations of the electron gas.
(8. Let us take T = 0 where nF is either zero or one. ω) = −χR (−q. However.3: Absorption of a photon creates an electron-hole pair excitation in the free electron gas. The ﬁrst case corresponds to ω < 0. ω) = −i 0 1 = V
∞ t
dt eiωt
1 V
nF (ξk ) − nF (ξk+q ) ei(ξk −ξk+q )(t−t ) e−η(t−t ) . which means that nF (ξk ) − nF (ξk+q ) is only non-zero if (k > kF and |k + q| < kF ) or (k < kF and |k + q| > kF ).ω) = π V nF (ξk ) − nF (ξk+q ) δ(ξk − ξk+q + ω).5. ω) Eq.8.e. (8. we need only study one case.
kσ
We can now analyze for what q and ω excitations are possible. we will study it in much more detail. which is easily seen from 0 0 Eq. while the latter corresponds to ω > 0.83)
. (8. see Eq. i. The delta function together with the second condition thus imply 0 < ω = q2 1 1 +k·q ⇒ 2m m ωmax = ωmin =
1 2 2m q + vF q 1 2 2m q − vF q
£¢¡
£ ¦ ¥¤ ¡ §
©
¨
3 2 1 0 0 1 2 3 4
(8. (8. and later on.82)
.
This function is known as the Lindhard function.82) is non-zero. for example ω > 0.
kσ
nF (ξk ) − nF (ξk+q )
kσ
ξk − ξk+q + ω + iη
. The possible range of q and ω is given by the dashed area in the right plot. for which (q. The strength of the interaction depends on the imaginary part of the polarization function. we ﬁnd k χR (q. In the frequency space.81).82) because c† ck = nF (ξk ). −ω).
q > 2kF . Within the non-interacting approximation and according to Eq. because of the symmetry χR (q.74) we then have that the dissipation of the electron gas is proportional to − Im χR (q. TWO-PARTICLE CORRELATION FUNCTIONS OF MANY-BODY SYSTEMS143
4
Figure 8.

The physical picture to remember is that the Green’s function G (rσt.144
CHAPTER 8.3. There is a continuum of such excitations given by conditions in (8. This we study in great detail later in this course. ω) spectral function
The spectral function is related to the density of states. σ r t ) = −i Ψσ (rt)Ψ† (r t ) σ G< (rσt. ω) = −2 Im GR (ν. In this chapter we have deﬁned the following many-body Green’s functions GR (rσt. where the change of momentum and energy of an incoming photon is measured. While the electron-hole pair excitations are the only possible source of dissipation in the non-interacting electron gas. The excitation of the electrons gas can be measured by for example inelastic light scattering (Raman scattering). where an electron within the Fermi sea is excited to a state outside the Fermi sea. namely the plasmon excitation.83). Ψ† (r t )]B.
8.8. is illustrated in the hand side of Fig. and secondly the absorption of electromagnetic radiation is given by the charge-charge correlation function. σ r t ) = −i (±1) Ψ† (r t )Ψσ (rt) σ retarded Green’s function greater Green’s function lesser Green’s function
and their corresponding Fourier transforms. and we will use Green’s functions in practically all discussions in the remaining part of the course.6
Summary and outlook
The concept of Green’s functions in many-body physics has been introduced in this chapters. The excitations which give rise to the dissipation are electron-hole pair excitations. this is certainly not true for the interacting case which is more complicated. We have so far seen two examples of this: the density of states is related to the spectral function and it can be measured for example in a tunneling experiment. The Green’s functions describe the dynamical properties of excitations. For non-interacting electrons the spectral function is given by a Dirac delta function A0 (ν.F σ G> (rσt. The important spectral function is in the frequency domain and in a diagonal basis given by A(ν. with initial spin σ and ﬁnal spin σ. ω) = 2πδ(ξν − ω) non-interacting case
. The process discussed here where an electron within the Fermi sea is scattering to an empty state outside the Fermi sea. GREEN’S FUNCTIONS
The possible range of excitations in (q. r σ t ) = −iθ (t − t ) [Ψσ (rt).3. r σ t ) gives the amplitude for propagation from the space-time point r t to rt. There is one particular type of excitation which is immensely important.ω)-space is shown in Fig. 8.

Ψ(r)](t) − [Vint . Ψ† (r t )]B.F . Next. r t ) = (−i) i∂t θ(t − t ) = δ(t − t )δ(r − r )+ + (−i) θ(t − t ) [i∂t Ψ(rt). (8. Ψ† (r t )]B. 145 (9. Ψ(r)] (t) = −[H0 .
9. We shall study examples of both situations in this chapter. Ψ(r)](t). Eq. The basic idea of this method is to generate a series of coupled diﬀerential equations by diﬀerentiating the correlation function at hand a number of times.
Here we used that the derivative of a step function is a delta function and the commutation relations for ﬁeld operators at equal times Ψ(r). Ψ† (r t )]B.2) [Ψ(rt). In many cases we need to calculate the time-dependence of these functions. r t ) = −iθ t − t [Ψ(rt).28) GR (rt. Ψ† (r ) B.1)
We ﬁnd the equation of motion for GR as the derivative with respect to the ﬁrst time argument i∂t GR (rt. let us study the time-derivative of the annihilation operator (throughout this chapter we assume that H is time independent) i∂t Ψ(rt) = − [H.F
+ (−i) θ(t − t ) [i∂t Ψ(rt).F . (9. one needs to invoke physical arguments to truncate the set of equations in a reasonable fashion.Chapter 9
Equation of motion theory
In the previous chapters we saw how various physical observables can be expressed in terms of retarded Green’s functions and correlation functions. (9. one of which is the equation of motion technique. For example one can neglect certain correlations. There are several ways of attacking this problem.F .1
The single-particle Green’s function
Let us consider the retarded Green’s function GR for either fermions or bosons. If these equations close the problem is in principle solvable. and if not. Ψ† (r t )]B.F = δ(r − r ).3)
.

1. Ψ(r)](t). (9.F
In this course we will mainly deal with problems where the Hamiltonian does not depend explicitly on time (linear response was an exception. It is now evident why the many-body functions. ν t ) = δ(t − t )δνν + DR (νt. are called Green’s functions. r t ) = −iθ(t − t ) −[Vint . The Hamiltonian is again written as H = H0 + Vint .F ν (9. ν t ) = −iθ(t − t )
−[Vint .9b)
DR (νt.146
CHAPTER 9.5. a† (t ) ν
B. was deﬁned as LG = delta function. By diﬀerentiation the commutator with H0 is generated −[H0 . for example the kinetic energy.
(9. aν ] = tνν aν . r t ) = δ(t − t )δ(r − r ) + DR (rt. 8. Ψ(r)] = 1 dr Ψ† (r ) 2m 1 2 =− Ψ(r). r t ). a† (t )]B. EQUATION OF MOTION THEORY
where the interaction part of the Hamiltonian includes all the interactions in the given problem. 2m r
2 r
Ψ(r ). If H0 is the usual kinetic energy Hamiltonian of free particles. Often it is convenient to work in some other basis.6)
The diﬀerential equation for the Green’s function in this basis GR (νt.5a) has the structure of the classical Green’s function we saw in Sec. say {ν}. ν
(9. Ψ(r)] we can. GR . ν t ) = −iθ(t − t ) [aν (t). 5. we have −[H0 . as in Sec. Ψ(r) (9. ν t ). DR (rt. but even there the time dependent problem was transformed into a correlation function of a time independent problem). After evaluating [Vint .
ν
(9.5a) (9. Therefore the Green’s function can only depend on the time diﬀerence t−t and in this case
.9a) .7)
is found in exactly the same way as above.F
The function DR thus equals the corrections to the free particle Green’s function. The equation in (9. continue the generation of diﬀerential equations. where the Green’s function of a diﬀerential operator.4)
In this case the equation of motion becomes i∂t + 1 2m
2 r
GR (rt. while H0 describes the quadratic part of the Hamiltonian. L. (9.8)
ν
and hence (iδνν ∂t − tνν ) GR (ν t.5b)
B. aν ](t). where the quadratic part of the Hamiltonian is H0 =
νν
tν ν a † a ν . Ψ† (r t ) .

For a diagonal basis.10b)
DR (ν.14)
which of course agrees with the result found in Eq.
9. 4 we saw that such a term is indeed relevant for superconductors. we proceed by solving a famous model for the appearance of a magnetic moment of impurities of certain
Here we only consider terms of the form c† c but also anomalous terms like cc could be included. ω). we can in fact solve for the Green’s function1 . aν ](t). (9.10a) .2
Anderson’s model for magnetic impurities
In order to exemplify the usefulness of the equation of motion technique.
ν
(9. ω) δνν = 0
1 δ . (8.9. in order to ﬁnd the Green’s function all we need to do is to invert the matrix −1 GR νν . i.2. We can write the equation of motion in frequency domain [δνν (ω + iη) − tνν ] GR (ν ν . to ensure proper convergence. (9.1
Non-interacting particles
For non-interacting particles. For the Green’s function in a superconductor we should therefore solve the linear problem in a way similar to the Bogoliubov transformation introduced in Chap. and that the Fourier transform of a delta function is unity.12)
and in matrix notation Eq.
9. In this case we have (δνν (ω + iη) − tνν ) GR (ν ν . η. As in Sec. ω − εν + iη νν
(9.F
Here it is important to remember that the frequency of the retarded functions must carry a small positive imaginary part. We return to this in Chap. tνν = δνν εν . 8. ω) = δνν (ω + iη) − tνν ≡ GR 0
−1 νν
(9.11)
where the subindex 0 on GR indicates that it is the Green’s function corresponding to a 0 non-interacting Hamiltonian. ω) = δνν 0
ν
(9. ω) = −i
∞ −∞
dtei(ω+iη)(t−t ) θ(t − t )
−[Vint .13)
Therefore. ANDERSON’S MODEL FOR MAGNETIC IMPURITIES
147
it is always useful to work with the Fourier transforms. δ(t) → 1. Recalling that when performing the Fourier transformation of the derivative it becomes ∂t → −iω. ν . which means that the Hamiltonian is bilinear in annihilation or creation operators.11) becomes GR 0
−1
GR = 1. 15.e. ω) = δνν + DR (ν. the solution is 0 GR 0
νν
= GR (ν. 4.51). a† (t ) ν
B. ν . In Chap. 0
(9.1 we deﬁne the inverse Green’s function as GR 0
−1
(νν .1.
1
.

In addition to the bare energy cost for an electron to reside in the d-state. which can be described by an eﬀective non-interacting model Hc =
kσ
For the impurity ion we assume that it has only one spin-degenerate state in the active shell.15)
(9.148
CHAPTER 9. Nb or Mo. which is typically the d shell.17)
. e.g. The crucial input is here dσ the correlation between electrons on the impurity ion. there is an interaction energy that depends on the state being doubly occupied or not.g. e. because the interaction in the narrower d-shell of a magnetic ion is particular strong and this is in fact the reason for the magnetism. The impurity ion Hamiltonian is thus modelled as Hd + HU =
σ
where ndσ = c† cdσ is the number operator for d-electrons.1: The Anderson model describing magnetic impurities embedded in a homogeneous host metal. the d-shell of a Fe ion. k kσ
(9. The coupling occurs because the d-orbital and the conduction band states overlap spatially and also lie close in energy.16)
t∗ c† cdσ . The electrons in the conduction band of the non-magnetic host metal. The bare onsite energy of the state on the magnetic ion is d . has a conduction band. giving rise to a “hybridization” between the two. The states forming the conduction band are primarily s-states that are more extended in space. The electrons occupying the conduction band couple to the outer-most electrons of the magnetic impurity ions. indicated by the dashed areas. magnetic ions embedded in a non-magnetic host metal. as seen in (b). therefore the state with two electrons residing on the ion has energy 2 d + U . and hence interactions are less important for those. The overlapping orbitals leads to a non-diagonal matrix element of the Hamiltonian Hhyb =
kσ
¦ ¨ ¥ '( © £ £¥ " & $% $ # ¨ " ! ¡ ¥ ¦ ( ¥ ¨ ¢ § £ ¥¤¢ ¡
(εk − µ) c† ckσ . dσ tk c† ckσ + dσ
kσ
(9. EQUATION OF MOTION THEORY
Figure 9. The host metal. couple to the level of the magnetic impurity ion. kσ (εd − µ) c† cdσ + U nd↑ n↓ . But the energy of electrons residing on the impurity ion also depends on whether it is doubly occupied or not.

spin up electrons depends on the occupation of spin down electrons. See Fig. it is favorable to ﬁll the orbital by two electrons. but the solution ﬁlls an entire book. c† t } .1 for an illustration.2.9. ω).2. ω) = 1 + U DR (dσ. (8. Although the Anderson model looks simple. is below the chemical potential and from the kinetic energy point of view.21)
The equation of motion are thus found by letting ν in Eq. Due to the hybridization term the Hamiltonian is not diagonal in the d-operators and the equations of motion will involve another Green’s function. t − t ) = −iθ t − t {ckσ (t) . ω) is the spectral function. The sum of these three energy contributions H = Hc + Hd + HU + Hhyb (9. GR . (8. say.2 It turns out that for certain values of the parameters it is energetically favorable for the system to have a magnetic moment (and thus minimizing the on-site interaction energy) while for other values there is no magnetic moment (thus gaining maximum hybridization energy). and it is not possible if 2εd + U > 2µ.18)
is known as the Anderson model. its full solution is very complicated and in fact the model has a very rich phase diagram. The hybridization therefore seems to randomize the spin on the magnetic ion.1
The equation of motion for the Anderson model
The magnetization in the z-direction is given by the expectation value of the diﬀerence n↑ − n↓ between spin up and down occupancy.19) ndσ = 2π where A(dσ.
. ANDERSON’S MODEL FOR MAGNETIC IMPURITIES
149
The bare d-electron energy. 9. However.23)
(ω + iη − εk + µ) GR (kσ. the system gains further kinetic energy by the hybridization.
(9. this costs potential energy. which on the other hand is complicated by the fact that the hopping in and out of the impurity orbital with. ω) −
k
tk GR (kσ. Furthermore. The occupation of a quantum state was found in Eq. k
2 The model in fact has a known exact solution. The physical question we try to answer here is: Under which circumstances is the material magnetic?
9. U . see Eq. c† t } . dσ. dσ. t − t ) = −iθ t − t {cdσ (t) .57) in terms of the spectral function. dσ (9. dσ. The Anderson model has been used to describe numerous eﬀects in the physics of strongly correlated electron systems. ω) − t∗ GR (dσ. (9. (9.49). εd . (9. ω).10 ). ω) = 0. dσ (9. For the d-electron occupation we therefore have dω nF (ω) A(dσ. which follows from the retarded Green’s function.20)
Let us write the equation of motion of this function using Eq.22) (9. namely GR (kσ.10) run over both d and k and we obtain the coupled equations (ω + iη − εd + µ) GR (dσ. All we need to ﬁnd is then GR (dσ. and it is hard to extract useful physical information from this solution.

ω) = 1. d↑ (9. so that it is no longer diagonal in the d-operators. Due to this term the “bare” d-electron energy.29)
and likewise for the spin-down Green’s function. In the time domain the imaginary part translates into a life-time. is seen to be renormalized by two eﬀects: ﬁrst the energy is shifted by U nd↓ due to the interaction with the averaged density of electrons having opposite spin. and solving Eq. (9. ω) = .26)
9. EQUATION OF MOTION THEORY
dtei(ω+iη)(t−t ) θ t − t
−[nd↑ nd↓ . cd↑ ] = −U nd↓ cd↑ . εd . the coupling to the conduction band electrons gives through Σ(ω) an energy shift and most importantly an imaginary part. (9. ω) = −i
∞ −∞
CHAPTER 9. t − t ) = −iθ t − t {nd↓ (t) cd↑ (t). The diagonal modes are instead superpositions of c.23) for GR (d ↑. (9.24)
The commutator in this expression is for σ =↑ [nd↑ nd↓ .22). The ﬁnal answer is 1 GR (d ↑. c† (t )} = nd↓ GR d ↑.28) in Eq.and d-states. cdσ ](t). cd↑ ] = nd↓ [nd↑ . ω − εk + µ + iη
(9.30a) (9. However a mean-ﬁeld approximation still grasps the important physics that the spin-up electron population depends on the spin-down population. and the set of equations does not σ close.150 where DR (dσ. therefore we replace the interaction part HU by its mean-ﬁeld version
M HU F = U nd↑ nd↓ + U nd↓ nd↑ − U nd↑ nd↓ . t − t .
ΣR (ω)
. ω) −
k
|tk |2 GR (d ↑. t − t ) = −iθ(t − t ) nd↓ {cd↑ (t). and secondly.
(9. c† (t ) dσ
.27) (9.25) and likewise we ﬁnd the commutator for spin down by interchanging up and down. We thus face the following more complicated Green’s function DR (d ↑. Inserting (9. ω − εd + µ − U nd↓ − ΣR (ω) ΣR (ω) =
k
(9. since the mean-ﬁeld approximation makes the Hamiltonian quadratic we can include U nd↓ (t) to the energy of the spin-up d-electrons in our equation of motion. ω − εk + µ + iη
The function is our ﬁrst encounter with the concept known as “self-energy”. d↑ In other words. nd↓ (t) d↑ (t)].
(9. The self-energy changes the pole of GR and furthermore gives some broadening to the spectral function.30b)
|tk |2 . the function DR becomes DR (d ↑. (9.26) with respect to time would generate yet another function {[H.28)
With this truncation. c† (t )} .2. ω) gives ω + iη − εd + µ − U nd↓ GR (d ↑. It arises because the coupling to the c-electrons introduces oﬀ diagonal terms in the Hamiltonian. d† (t )} to be determined.2
Mean-ﬁeld approximation for the Anderson model
Diﬀerentiating the function in Eq.

= (ω − ε + µ − U nd↓ )2 + Γ2 ˜
(9. For ω+µ ∈ [−W. the approximation is indeed valid. Let us assume that the product d(ε)|t(ε)|2 is constant within the band limits.31) ω−ε+µ
The density of states d (ε) and the coupling matrix element t (ε) depend on the details of the material.35)
If we neglect the ﬁnite bandwidth. (9. we get
0
W . Note that the spectral function derived here is an example of the Lorentzian form discussed in Sec. ANDERSON’S MODEL FOR MAGNETIC IMPURITIES
151
9.33)
The real part gives a shift of energy and since it is a slowly varying function. the selfenergy Σ is ΣR (ω) = dε d(ε) |t (ε) |2 =P ω − ε + µ + iη dε d(ε) |t (ε) |2 −iπd(ω +µ)|t(ω +µ)|2 .34)
where Γ is the width of the spectral function. ω) = −2 Im GR (d ↑.2. but fortunately it is not important for the present considerations.32)
This approximation is good if the width of the Green’s function (which we shall see shortly is given by Γ) turns out to be small compared to the scale on which d(ε)|t(ε)|2 typically changes. and if we furthermore
nd↑ ≈
dω 2Γ . Since in practice Γ εF . Now the self-consistent mean-ﬁeld equation for nd↑ follows as nd↑ = dω nF (ω)A (d ↑.5. T = 0. W ] we get ΣR (ω) ≈ dε − iΓ −W ω − ε + µ Γ W +ω+µ = − ln − iΓ.9. ω) 2Γ θ (W − |ε|) .3
Solving the Anderson model and comparison with experiments
Assuming that the coupling tk only depends on the length of k and thus on ε. (ω − ε + µ − U nd↓ )2 + Γ2 ˜ −W 2π
(9.3. = − tan−1 2 π Γ
(9. ω) 2π W dω 2Γ = nF (ω) . we simply include it as a shift of ε and deﬁne the new onsite energy ε = ε + Re ΣR . ˜ The spectral function hence becomes A(d ↑.36)
. −W < ε < W.2. (9. 8. which is justiﬁed because Γ consider low temperatures. π W −ω−µ Γ π
W
−W < ω + µ < W.
(9. 2 2 ˜ −∞ 2π (ω − ε + µ − U nd↓ ) + Γ ε − µ + U nd↓ ˜ 1 1 . and deﬁne the important parameter Γ by πd(ε)|t(ε)|2 = Γθ(W − |ε|).

2: The upper part shows the mean ﬁeld solution of the Anderson model with the left panel being magnetization as a function of electron density nel .e. For 4 < nel < 8 the magnetization curve is seen to be quite similar to the prediction of the model.
BC 3 %0%12! 1% 0%3 ( '& # # ¤ £
ga b V f
4 A @ ! % 98 & 3 673 25! 24 1%3 3 ! "3 0%12! ) 1 % #$" ! © ¢¡ ¨ ¦ § ¥ 2B 3 %9 %58 G F ) 3 425! %25E 3 D " 4 X V b Ca ` YP 6W "U e¨ dc c
ch
ST CR Q H7P I H 6 i
Q ¢P I H H
. The latter means that too strong hybridization destroys the magnetization.152
CHAPTER 9. The electron concentration and hence µ is varied by changing the alloy. We see that there is a critical density and a critical U/Γ where the magnetization sets in. EQUATION OF MOTION THEORY
Figure 9. The bottom panel shows experimental results (Clogston et al. i. For nel > 8 the eﬀect of having more than two d-orbitals in the Fe-atoms becomes important and the simple model is no longer adequate. (1962)) for the magnetic moment of Fe embedded in transition metals. the chemical potential. for two diﬀerent Γ-values. while the right panel is the maximum magnetization as function of the correlation energy.

See also Exercise 8.9. i.3. x = − (˜ − µ) /U. at least qualitatively. Such an experiment is for example a tunneling experiment where current is passed through a single atom or a small metallic island which can be thought of as an artiﬁcial atom. because. 6 it was shown that this function is related to the dielectric response function and therefore tells about the screening properties of the material. In an experiment where one probes the actual occupation of the atom this approach would not be suﬃcient. For the electron that wants to enter the island it does matter whether the island is already occupied. As is evident there. THE TWO-PARTICLE CORRELATION FUNCTION We obtain the two coupled equations cot (πn↑ ) = y(n↓ − x). 9. cot (πn↓ ) = y(n↑ − x). ρ(r t )] .5
Further correlations in the Anderson model: Kondo eﬀect
9.2.37b)
The solution of these equation gives the occupation of the d-orbital and in particular tells us whether there is a ﬁnite magnetization. the model describes the observed behavior.3
The two-particle correlation function
The two particle correlation functions.
153
(9. Also for this quantity one can generate a set of equation of motions. But even so they may provide a good starting point for various approximation schemes. such as the density-density correlation. For the case of the electron gas.
9.
9. This means that the energy of a given spin direction is only aﬀected by the average occupation of the opposite spin direction.2. To capture this physics one must go one step beyond the mean-ﬁeld approximation and truncate the equations of motion at a later stage. 6 shown to give the linear response properties.3.3 In Fig. if it is. diﬀerent from the trivial solution n↓ = n↑ .4. (9.e. which it in fact does have. the tunneling barrier is increased by U .1
The Random Phase Approximation (RPA)
A commonly used approximation scheme for correlation functions is the so-called Random Phase Approximation (RPA).
9. This is the topic for Exercise 8. which is one of our
3 We should also convince ourselves that the magnetic solution has lower energy.3.4
Coulomb blockade and the Anderson model
Above we applied the mean-ﬁeld approximation to the interaction.37a) (9. Consider for example the retarded charge-charge correlation function χR (rt. ε y = U/Γ. and as for the single particle Green’s function they are not solvable in general.2 solutions of these equations are shown together with experimental data.
.38)
In Chap. whether there exists a solution n↓ = n↑ . was in Chap. r t ) = −iθ t − t [ρ(rt).

ρ −q. k and for this purpose we need the following commutators c† ck+q . (9. ck ck+q ] = − V (q ) c† c† −q ck ck+q + c† +q c† ck+q ck k k+q k k−q 2
kq
− c† +q c† ck+q+q ck − c† c† −q ck ck+q−q k k k k
.44)
[H0 . Furthermore for each level of the equation of motion a Green’s function with two more electron operators pops up. The two derivations give complementary insight into the physical content of the approximation. k+q k−q
kk q=0
(9. ρ (q) =
k
c† ck+q .t ] .45) is replaced by a mean-ﬁeld expression where pairs of operators are replaced by their average values. ρ −q. In Chap. t − t ) = −iθ t − t [(c† ck+q ) (t) .t )] k
− iθ(t − t ) [−[H. The charge-charge correlation function is χR (q. c† ck −q = c† ck − c† ck+q . c† ck+q ] (t) . but also in general gives a decent description of the interacting electron gas. Using the recipe from Chap. χR (q) = ﬁnd the equation of motion i∂t χR (kq.39)
Furthermore.
.t . k
R k χ (kq).42) a new 6-particle Green’s function is generated.154
CHAPTER 9.12 RPA is derived using Feynman diagrams. it turns out to be better to work with the function χR (kq.
(9. but here we derive it using the equation of motion technique. c† ck+q ] = (ξk − ξk+q ) c† ck+q . k k 1 † [Vint . We will for simplicity work with the translation-invariant electron gas with the Hamiltonian given by the usual kinetic energy plus interaction energy (here we disregard the spin degree of freedom because it is not important) H=
k
ξk c† ck + k
1 2
V (q)c† c† ck ck = H0 + Vint . t − t ) = δ(t − t ) [(c† ck+q )(t). k
(9. t )] .
(9.41) Let us
from which we can easily obtain χ(q) by summing over k. RPA is exact in some limits.40)
However. t − t ) = −iθ t − t 1 V ρ (q.43) (9. ρ (−q) = k c† ck+q . k k k+q k
(9. ρ(−q.t) . At this stage we truncate this series by the random phase approximation which says that the right hand side of (9.45)
When this is inserted into Eq. EQUATION OF MOTION THEORY
main topics in this course.42)
(9. ρ(−q. the q = 0 component is cancelled by the positively charged background.

the interactions add other fundamental excitations. we get [Vint . However.
(9. c† ck+q ] ≈ − k 1 2 V (q ) c† ck+q c† −q ck k k+q
k q =0
155
+ c† ck+q c† −q ck k k+q
+ c† ck+q c† +q ck + c† ck+q c† +q ck k k k−q k−q −c† +q ck k − c† ck+q−q k =
k
c† ck+q+q k c† −q ck k
− c† +q ck k − c† ck+q−q k c† −q ck . 8. 8. ω).49). ω) 0 . by the Fermi-Dirac distribution function. ω) = nk+q − nk 1 − V (q)
k
χR (k q.50)
Replacing the expectation values.k .3) is preserved here. 8. ω) and therefore the structure of the electron-hole excitations of the non-interacting gas (depicted in Fig. 1 − V (q)χR (q. nk .3. ω) was derived in Sec.9. k
c† ck+q+q k c† −q ck + const. The excitations which were shown in Sec. 8.5. Collecting everything and going to the frequency domain the equation of motion becomes. ω) 0
(9.49) and the two results therefore agree nicely. when summed over k. ω) =
k
1 V
nk+q − nk ω + ξk − ξk+q + iδ
1 + V (q) χR (q. ω) indeed was equal to the numerator in (9. 8.47)
which.ω) = 1 + V (q)χR (q. and in the case of a charge liquid these modes are the plasmon modes.48)
χR (q. we recognize the Lindhard function studied in Sec.5. 0
(9. k (9.5 to be related to the imaginary part of χR (q. There we studied a non-interacting electron gas and found that χR (q. but of course the strength is modiﬁed by the real part of the denominator of (9. (9. (ω + iη + ξk − ξk+q ) χR (kq. The free particle polarizability χR (q. ω) . Note that the exchange pairings which we included k in the Hartree-Fock approximation is not included here. The RPA dielectric function becomes 0 εRPA (q.49)
This is the RPA result of the polarizability function. In Sec. The additional
. ω) χR (q. ω) = and hence χR. ω)
−1
= 1 − V (q) χR (q. namely collective modes. ω) .5 we also analyzed the excitation of the non-interacting electrons gas and the analysis there is basically still correct. ω) = 1 V χR (kq. allows us to ﬁnd an equation for χR (q.46)
V (q)
nk+q − nk
where we used that c† ck = nk δk. THE TWO-PARTICLE CORRELATION FUNCTION 4.RPA (q.

we have 0 − Im χR (q. ω) is zero because then 0 there is a possibility of a pole in the polarizability. ω) = −iδ. Therefore approximations are necessary and we saw particular examples of this. and in fact only a very small set of Hamiltonians describing interacting systems can be solved exactly. namely the mean-ﬁeld solution of a magnetic impurity embedded in a metallic host. The set of diﬀerential equation is not soluble in general. ω) 0
+
δ2
= πδ 1 − V (q) Re χR (q. 0
(9. ω) = 0 and this 0 is the plasma oscillation mode. ω) = δ 1−V
2 (q) Re χR (q. If we set Im χR (q. namely the equation of motion method applied to the Green’s functions.156
CHAPTER 9.
9. In the following chapter we use the equation of motion to derive the Green’s functions in the imaginary time formalism and to derive the famous Wick’s theorem. EQUATION OF MOTION THEORY
modes are given by the part where the imaginary part of χR (q. 12.4
Summary and outlook
In this chapter we have seen a method to deal with the dynamical aspects of interacting many-body systems.
. The plasmon is studied in detail in Chap. here we just mention that the condition for the mode turns out to be 2 ω ∝ ωpl + const. Wick’s theorem will then pave the way for introducing the Feynman diagrams.51)
This means that there is a well-deﬁned mode when 1 − V (q) Re χR (q. also called a plasmon. and the RPA approximation for the charge auto correlation function. ω) . q 2 .

(8. In this chapter. The imaginary time formalism is particularly useful when we want to perform perturbation theory. This is much like treating electrical circuit theory with complex numbers even though all currents and voltages are real. which in general are deﬁned as
R CAB (t. from which we can ﬁnd the retarded function as C R = iθ(t−t )(CAB we have 1 CAB (t. or that they can be derived in a simple way from the Green’s functions.1) boson operators mean either single particle operators like b or b† or an even number of fermion operators such as c† c appearing in for example the density operator ρ. and this will eventually lead us to the Feynman diagrams. it is the single particle Green’s function deﬁned in Eq. which we need to learn. and is only a clever mathematical trick. 6. When A and B are two-particle operators. e. For technical reasons it is convenient to use a mapping to a more general Green’s function. the density or current operators. Z 157 (10. B(t )
B.
B : for bosons F : for fermions
. In all the situations we have studied so far the physical observables have been related to the retarded Green’s functions. The present chapter concentrates on the mathematical details of the technique and applications are left for later.
(10. t ) = − A(t)B(t ) . Let us for example look at the deﬁnition of the following correlation function CAB (t. In Eq.g. By deﬁnition (10.1)
When A and B are single particle annihilation and creation operators. t ) = −iθ(t − t )
A(t).F
. The important thing that distinguishes the boson case from the fermion case is the sign change that is obtained upon interchange. This has no real physical meaning.2) CBA ). t ) = − Tr(e−βH A(t)B(t )).3)
. (10.28) from which one could derive the density of states. C R has the form of a retarded correlation function that was shown to give the linear response results of Chap. where the time and frequency arguments are imaginary quantities.Chapter 10
Imaginary time Green’s functions
We have seen that physical observables often have the form of Green’s functions. we introduce a mathematical method to work out the retarded Green’s functions.

In order to treat both U and ρ in one go. But the trouble arises for the density matrix e−βH .e. The solution to this problem is to use imaginary times instead of real times.158
CHAPTER 10. i∂t U = HU while ρ is the solution to ∂β ρ = −Hρ. so A(t) = eitH Ae−itH . (10. t)A(t)U (t. to go back to physical real times from the imaginary time function.4) Z ˆ In Eq.4) result in three time-ordered exponentials. we will see that there is a well-deﬁned method to obtain the physically relevant quantity. If we consider a product of operators A(τ )B(τ ) and write it in terms of the corresponding operators in the interaction representation. (10. (10.8)
Note that we consider only time-independent Hamiltonians in this section. 5. we get ˆ ˆ ˆ ˆ ˆ A(τ )B(τ ) = U (0. you should use the imaginary time deﬁnitions when the time argument is a Greek letter and the usual deﬁnition when the times are written with roman letters. In the end this means that both U and ρ can be treated in just one expansion in powers of V . t ) = − Tr e−βH U (0. where V is the perturbation. i. we can deﬁne the interaction picture for imaginary times as ˆ A(τ ) = eτ H0 Ae−τ H0 . but bare in mind that this is purely a mathematical trick without physical contents. τ )A(τ )U (τ. but we will not cover the more general case of non-linear time dependent response in this course. To make a long story short: this is a mess and a new idea is therefore needed. To employ imaginary time is not as far fetched as it might look.18) we saw also how a single U operator could be expanded as a time-ordered exponential. (10. where τ is real and has the dimension time. As for real time we can deﬁne an imaginary time Heisenberg picture by substituting it by τ . We deﬁne A(τ ) = eτ H Ae−τ H . (5. one cannot use the ordinary equilibrium statistical mechanics but instead one must use a non-equilibrium formalism.5) In this notation.1 They therefore satisfy similar diﬀerential equations: U satisﬁes the Schr¨dinger equation. 5 that the interaction picture provides a systematic way of expanding in powers of V. This we did in the linear response limit in Chap. 0). t a Roman letter. If they are not timeindependent. (10. We could try to utilize this and write CAB as 1 ˆ ˆ ˆ ˆ ˆ CAB (t. τ a Greek letter. which could be collected into a single time-ordered exponential. τ )B(τ )U (τ .
. one replaces the time argument by a imaginary quantity t → −iτ . which should also be expanded in powers of the interaction. the relation between the Heisenberg and the interaction picture in imaginary time follows the arguments in Chap. Then we saw in Chap. o which is known as the Bloch equation.7)
Letting H = H0 + V .
1
(10.6) Similar to the interaction picture deﬁned for real times. t )B(t )U (t . IMAGINARY TIME GREEN’S FUNCTIONS
Suppose the Hamiltonian is H = H0 + V . Furthermore. because both the density operator ρ = e−βH /Z and the time evolution operator U (t) = e−iHt are both exponential functions of the Hamiltonian. 6. 0) . This would in Eq.

(10. 0)) . the operators are ordered such that Tτ (A(τ )B(τ )) is equal to A(τ )B(τ ) for τ > τ and B(τ )A(τ ) when τ > τ . Z Utilizing Eqs.159 ˆ where.8) 1 Tr e−βH Tτ A(τ )B(τ ) . However. (10. we need to relate the correlation functions written in imaginary time and the correlation function with real time arguments. τ ).12)
= Tτ exp −
τ
ˆ dτ1 V (τ1 ) . 0)
0 0
. is of course the same.9) with respect to τ and ﬁnd ˆ ∂τ U (τ. 0)A(τ )B(τ ) = ˆ ˆ Tτ U Tτ A(τ )B(τ ) = Tr e Z ˆ U (β. 0) Tτ (U (0.12).18). τ ) =
n=0
1 (−1)n n!
τ
τ τ
τ
dτ1 · · ·
τ
ˆ ˆ dτn Tτ V (τ1 ) · · · V (τn ) (10.8) and (10. the time-evolution operator U in the interaction picture is ˆ U (τ. First we diﬀerentiate Eq. 0) . τ ) = U (τ. Now the same iterative procedure is applied and we end with
∞
ˆ U (τ. (10. τ ) = eτ H0 e−(τ −τ From this it follows directly that ˆ ˆ ˆ U (τ. Above it was argued that the density operator naturally can be treated within the imaginary time formalism. (5.9)
An explicit expression for U (τ. before we can see the usefulness fully.15) Z This can be written in a much more compact way relying on the properties of Tτ and Eq.3. like in Eq.9) and (10. and where the averages · · · e−βH0 e−βH0
0
depending on appear after normalizing with 1/Z0 = 1/Tr . τ )U (τ . 0)A(τ )B(τ ) 1 −βH0 ˆ (β.11)
ˆ This is analogous to Eq.10) ˆ ˆ ˆ Tτ U (β. and indeed it can.13)
0
Consider now the time ordered expectation value of the pair of operators in Eq. τ ) = 1. 0) = e−βH0 Tτ exp − dτ1 V (τ1 ) .10)
)H −τ H0
e
. τ ) is found in analogy with the derivation of Eq. (10. because by combining Eqs.13) we can immediately expand in powers of V Tτ A(τ )B(τ ) = Tτ A(τ )B(τ ) = (10. (10. τ )A(τ )U (τ. 5.
(10. (10.13) and the boundary condition. This result demonstrates that the trick of using imaginary time indeed allows for a systematic expansion of the complicated looking expression in Eq. (10. (10. (5. U (τ. (10.4).16)
ˆ where we have used Z = Tr e−βH = Tr e−βH0 U (β.
(10. (10.14)
1 ˆ ˆ ˆ ˆ ˆ ˆ Tr e−βH0 U (β. i. τ ).
The time ordering is again the same as deﬁned in Sec.12) we obtain β ˆ ˆ e−βH = e−βH0 U (β. τ ) = eτ H0 (H0 − H)e−(τ −τ
)H −τ H0
e
ˆ ˆ = −V (τ )U (τ.
. τ )B(τ )U (τ . (5.e.

This follows from the cyclic properties of the trace. we have the property CAB (τ ) = ±CAB (τ + β).19)
= CAB (τ − τ ). (10.e. also called Matsubara Green’s function. For τ > τ the equality τ − τ < β is clearly seen if one uses the Lehmann representation in Eq. likewise. IMAGINARY TIME GREEN’S FUNCTIONS
10.20) for τ < 0 is −1 Tr e−βH e(τ +β)H Ae−(τ +β)H B CAB (τ + β) = Z −1 = Tr eτ H Ae−τ H e−βH B Z −1 Tr e−βH Beτ H Ae−τ H = Z −1 = Tr e−βH BA(τ ) Z −1 = ± Tr e−βH Tτ (A(τ )B) Z = ±CAB (τ ).20) which again follows from the cyclic properties of the trace.17) three things are clear. It means that operators are ordered according to history and just like the time-ordering operator seen in Chap.
. CAB (τ. (10. τ ) is a function of the time diﬀerence only. τ ) = Z Tτ A(τ )B(τ ) = θ(τ − τ )A(τ )B(τ ) ± θ τ − τ B(τ )A(τ ). and. the second equality is obtained if τ < τ . (10. τ ) is guaranteed only if −β < τ − τ < β.17) where the time-ordering symbol in imaginary time has been introduced. Secondly.18) The next question is: What values can τ have? From the deﬁnition in Eq. τ ) = CAB (τ − τ ).21) and similarly for τ > 0. convergence of CAB (τ. (10. Thirdly. CAB (τ. τ ) ≡ − Tτ A(τ )B(τ ) . − for fermions. Firstly. The proof of Eq. i. (10. for τ < 0. (10.19) to get a factor exp (− [β − τ + τ ] En ) . 5 with the later “times” to the left + for bosons. We have for τ > τ −1 Tr e−βH eτ H Ae−τ H eτ H Be−τ H CAB (τ. is deﬁned in the following way CAB (τ. = = −1 Tr e−βH e−τ Z
H τH
e
Ae−τ H eτ
H
B
−1 Tr e−βH e(τ −τ Z
)H
Ae−(τ −τ
)H
B (10. (10.
and of course likewise for τ > τ .1
Deﬁnitions of Matsubara Green’s functions
The imaginary time Green’s functions.160
CHAPTER 10.

24)
From now on we use the following notation for the Fourier transforms of the Matsubara Green’s functions
β
CAB (iωn ) =
dτ e
0
iωn τ
CAB (τ ). β
for bosons. 2 0 2 0 β 1 = 1 ± e−iπn dτ eiπnτ /β CAB (τ ). β]. e−iπnτ /β CAB (n). (10.g replace δ(τ ) by δ(τ − 0+ ). we remark that the boundaries of the integral 0 dτ in Eq.
(10.23)
and since the factor 1 ± e−iπn is zero for plus sign and n odd or for minus sign and n even and 2 otherwise.22a) (10.
ωn = ωn =
2nπ β . A consistent choice is always to move the time argument into the interior of the interval [0. due to the symmetry property (10. Note how the information about the temperature is contained in the Matsubara frequencies through β. we take CAB (τ ) to be deﬁned in the interval −β < τ < β. 2 0
(10. for example if CAB (τ ) includes a delta function δ(τ ).2
Connection between Matsubara and retarded functions
We shall now see why the Matsubara Green’s functions have been introduced at all. β Finally. In the frequency domain they are in fact the same analytic function as the usual real times
.10. n is odd for fermions.
1 1 β dτ eiπnτ /β CAB (τ ) + e−iπn dτ eiπnτ /β CAB (τ − β).1
Fourier transform of Matsubara Green’s functions
Next we wish to ﬁnd the Fourier transforms with respect to the “time” argument τ .
10. Because of the properties above.
n is even for bosons.2.21) this can be simpliﬁed as CAB (n) = = 1 2
β 0 β
dτ eiπnτ /β CAB (τ ) +
1 2
0 −β
dτ eiπnτ /β CAB (τ ).
(10. for fermions. (2n+1)π . e.1. CONNECTION BETWEEN MATSUBARA AND RETARDED FUNCTIONS161
10.
(10.25)
The frequency variable ωn is denoted a Matsubara frequency.22b)
n=−∞
However.25) leads to a minor ambiguity of how to treat the boundary τ = 0. and thus according to the theory of Fourier transformations we have a discrete Fourier series on that interval given by CAB (n) ≡ CAB (τ ) = 1 2 1 β
β −β ∞
dτ eiπnτ /β CAB (τ ). we obtain
β
CAB (n) =
0
dτ eiπnτ /β CAB (τ ).

3. because we can generate both CAB (iωn ) and CAB (ω) from the following function deﬁned in the entire complex plane except for the real axis
CAB (z) =
1 Z
nn
n |A| n n |B| n z + En − En
e−βEn − (±) e−βEn
. Therefore the connection between imaginary time functions and retarded real time functions derived here is only valid in this ensemble.26)
The Matsubara function is calculated in a similar way.28)
−1 = Z = −1 Z
e−βEn
nn
n |A| n n |B| n iωn + En − En n |A| n n |B| n iωn + En − En
e−βEn
nn
1 = Z
nn
n |A| n n |B| n iωn + En − En
e−βEn − (±)e−βEn
R Eqs. In other words.27)
and hence
β
CAB (iωn ) =
dτ eiωn τ
0
−1 Z
e−βEn n |A| n
nn
n |B| n eτ (En −En ) .28) show that CAB (iωn ) and CAB (ω) coincide and that they are just R special cases of the same function. the other one follows by analytic continuation. In the general case we get2
R CAB (ω) =
1 Z
nn
n |A| n n |B| n e−βEn − (±)e−βEn ω + En − En + iη
. CAB (iωn ).where η is a positive inﬁnitesimal. eiωn β eβ(En −En ) − 1 . Z
nn
(10.3 we calculated the retarded single particle Green’s function and the result Eq.29)
This function is analytic in the upper (or lower) half plane. (8. Since it is in many cases much easier to compute the Matsubara function.46) can be carried over for fermions. but has a series of poles at En −En along the real axis. For τ > 0. IMAGINARY TIME GREEN’S FUNCTIONS
Green’s functions.26) and (10. Indeed we shall now show that the appropriate analytic R continuation is CAB (ω) = CAB (iωn → ω + iη). that equals CAB (iωn ) on the R imaginary axis and CAB (ω) on the real axis. 8.
2
. According to the theory of analytic functions: if two functions
Note that it is assumed that the grand canonical ensemble is being used because the complete set of states includes states with any number of particles.
(10. this is a powerful method for ﬁnding the corresponding retarded function. This means that once we have one of the two.162
CHAPTER 10. R The relation between the two functions CAB and CAB is proven by use of the Lehmann representation.
(10. there exists an analytic function CAB (z). In Sec. (10. ±eβ(En −En ) − 1 . we have CAB (τ ) = −1 Tr e−βH eτ H Ae−τ H B Z −1 = e−βEn n |A| n n |B| n eτ (En −En ) . where z is a complex frequency argument in the upper half plane. (10.

(10.e.2.1: The analytic continuation procedure in the complex z-plane where the Matsubara function deﬁned for z = iωn goes to the retarded or advanced Green’s functions deﬁned inﬁnitesimally close to real axis. On the real axis coming from above R this function is identical to the retarded function.1
Advanced functions
The function CAB (z) is analytic for all z away from the real axis. 10.1. it is not a simple task to determine CAB (z) unless it has been reduced to an rational function as in Eq. CONNECTION BETWEEN MATSUBARA AND RETARDED FUNCTIONS163
Figure 10. If we na¨ ıvely insert iωn → ω + iη before doing the integral.10. i. coincide in an inﬁnite set of points then they are fully identical functions within the entire domain where at least one of them is a analytic function and. This is illustrated in Fig. we could do the same thing in the lower half
) # " 0 $% # " ( ! ' ' ¥ ©¨ § # " & $% " ©! #
1
' ¢ ' ' ¢¢
¢ ¢ ¢ ¢ ¢
¡¢ ¦ £¤ ¥ ¢
. CAB (z = iωn ) = CAB (iωn ). (10. CAB (z = ω + i0+ ) = CAB (ω). Therefore instead of the continuation in the upper half plane. This means that if we know CAB (iωn ) we can ﬁnd CAB (ω) by analytic continuation: R CAB (ω) = CAB (iωn → ω + iη). To summarize: Using the Lehmann representation we have shown that there exists a function CAB (z) which is analytic for z not purely real and which coincides with the Matsubara function. there is only R one such common function.e. For example look at the deﬁnition in Eq.30) iωn → z → ω + iη gives the right analytic function.28). i. furthermore. where it is evident that the replacement in (10. it is not obvious how to perform the continuation. (10.25).
10. the answer is completely diﬀerent and of course wrong. However. Later we shall see examples of how to perform the analytic continuation correctly. If not.30) Warning: this way of performing the analytic continuation is only true when CAB (iωn ) is written as a rational function which is analytic in the upper half plane.2.

(10. real space.3
Single-particle Matsubara Green’s function
An important type of Matsubara functions are the single-particle Green’s function G. the eﬀect of retardation.36)
(10. t ) = iθ(t − t)
A(t). τ − τ ) = − Tτ cν (τ )c† (τ ) ν .1
Matsubara Green’s function for non-interacting particles
For non-interacting particles the Matsubara Green’s functions can be evaluated in the same way we found the retarded Green’s function in Sec.35)
= −θ(τ − τ ) cν (τ )c† (τ ) − (±) θ(τ − τ ) c† (τ )cν (τ ) ν ν = − θ(τ − τ ) cν c† (±)θ(τ − τ ) c† cν ν ν For fermions this is G0.e.
(10.3.31)
The advanced Green’s function is in the time domain deﬁned as
A CAB (t. τ − τ ) = − θ(τ − τ )(1 − nF (ξν )) − θ(τ − τ )nF (ξν ) e−ξν (τ −τ
)
e−ξν (τ −τ ) . 8.
10. 6. as was explained in Chap.33a) (10. τ ) σ G(ντ.
(10. which gives the so-called advanced Green’s function.F (ν. c† (τ ) = eτ H0 c† e−τ H0 = eξν τ c† . ν τ ) = − Tτ cν (τ )c† (τ ) ν . i. IMAGINARY TIME GREEN’S FUNCTIONS
plane iωn → z → ω − iη. B(t )
B. .37)
.F
. ν
(10. gives the present state of the system as it has evolved from the state at previous times. ν ν ν (10. r στ ) = − Tτ Ψσ (r.
A CAB (ω) = CAB (iωn → ω − iη).τ )Ψ† (r .
10.3. which gives G0 (ν.33b)
{ν} representation. The retarded one.164
CHAPTER 10. Suppose the Hamiltonian is diagonal in the ν-quantum numbers H0 =
ν
ξν c† cν . They are deﬁned as G(rστ.32)
The term “advanced” means that it gives the state of the system at previous times based on the state of system at present times. (10.2.34)
so that cν (τ ) = eτ H0 cν e−τ H0 = e−ξν τ cν .

F (ν. ikn )eikn τ .51). In the frequency representation. 1 . the retarded free particles Green’s functions are for both fermions and bosons GR (ν.41)
10. while the bosonic one becomes qn = 2nπ/β
G0. (10. iqn − ξν (10. τ ). while iqn and iωn are used for boson frequencies. EVALUATION OF MATSUBARA SUMS while the bosonic free particle Green’s function reads G0.
0
1 eiqn β e−ξν β − 1 .10.4
Evaluation of Matsubara sums
When working with Matsubara Green’s functions we will often encounter sums over Matsubara frequencies.42)
. For example sums of the type S1 (ν.4. τ ) = 1 β G(ν. Here we have anticipated the notation that is used later: Matsubara frequencies ikn and ipn are used for fermion frequencies.B (ν.B (ν. ω − ξν + iη (10. ikn − ξν
dτ eikn τ e−ξν τ . (8.39)
−1
because eikn β = −1 and 1 − nF (ε) = e−βε + 1
β
.
ikn
τ > 0.
β
kn = (2n + 1) π/β
= − (1 − nF (ξν )) = − (1 − nF (ξν )) = 1 .38)
G0.F (ν. τ − τ ) = − θ(τ − τ ) (1 + nB (ξν )) + θ(τ − τ )nB (ξν ) e−ξν (τ −τ ) . ω) = 0 in agreement with Eq.
0
1 eikn β e−ξν β − 1 . According to our recipe Eq.30). the fermionic Green’s function is
β
165
(10. iqn ) =
0
dτ eiqn τ G0.B (ν. similar to integrals over frequencies in the real time language. τ ). ikn ) =
0
dτ eikn τ G0. iqn − ξν
dτ eiqn τ e−ξν τ .
β
= − (1 + nB (ξν )) = − (1 + nB (ξν )) = 1 .40)
−1
because eiqn β = 1 and 1 + nB (ε) = − e−βε − 1 .
(10. ikn − ξν (10.

(10. (10.g.45a) nF (z) = βz e +1 1 nB (z) = βz . ikn )G0 (ν2 .49b) C 2πi
.44a) (10. poles for z = i(2n)π/β. n(z). respectively.
ikn
ikn fermion frequency iωn boson frequency
(10. τ ) = 1 β G0 (ν1 .46b) =+ .47)
for fermions and similarly for boson frequencies dz nB (z)g(z) = 2πi Res [nB (z)g(iωn )] =
z=ikn
2πi g(iωn ).43)
This section is devoted to the mathematical techniques for evaluating such sums. is given by dz nF (z)g(z) = 2πi Res [nF (z)g(ikn )] = −
z=ikn
2πi g(ikn ). The imaginary time formalism has been introduced because it will be used to perform perturbation expansions. and therefore the types of sums we will meet are often products of the such free Green’s functions. β
(10. we deﬁne the two generic sums S F (τ ) = S B (τ ) = 1 β 1 β g(ikn )eikn τ . but only regions where g(z) is analytic. the trick is to rewrite them as integrals over a complex variable and to use residue theory. poles for z = i(2n + 1)π/β. ν2 .166
CHAPTER 10. but no singularity of g(z). To evaluate these.46a) βz + 1 z=ikn z→ikn e δ→0 e e +1 β (z − iωn ) δ 1 Res [nB (z)] = lim = lim βiωn βδ (10. These functions turn out to be the well known Fermi and Bose distribution functions 1 .48)
If we therefore deﬁne contours. e. βz − 1 z=iωn z→iωn e δ→0 e β e −1 According to the theory of analytic functions.
ikn
τ > 0. β
(10. In order to be more general.
iωn
and study them for τ > 0. which have poles at z = ikn and z = iωn . IMAGINARY TIME GREEN’S FUNCTIONS
or sums with products of Green’s functions. C. iωn . we can write dz nF (z)g(z)ezτ . (10.
(10. the contour integral which encloses one of these points.44b)
g(iωn )eiωn τ . ikn + iωn )eikn τ . (10. which enclose all point z = ikn in the fermionic case and all points z = iωn in the bosonic case.45b) e −1 The residues at these values are (z − ikn ) δ 1 Res [nF (z)] = lim = lim βikn βδ =− .49a) SF = − C 2πi dz SB = + nB (z)g(z)ezτ . (10. S2 (ν1 . For this we need two functions.

we use the contour integration technique in two special cases.1
Summations over functions with simple poles
Consider a Matsubara frequency sum like Eq. zj .52)
¦¥ ¤ ¢ ©
§¨ ¦¥ ¤ £ ¤ ¢ ¢ ¢ ¢ ¢ ¢
¡¢ ¢ ¢ ¢
¢ ¢ ¤ £ ¢ §! ¦¥ ¤
¤ ¥ §# ¦"¤ £
. (10. has a number of known simple poles. τ > 0.50) S0 (τ ) = β
ikn
where g0 (z).43) but let us take a slightly more general function which could include more free Green’s function. see Fig.2: The contour used to perform the Matsubara sum for a function with known poles. (10.2. (10.43) 1 g0 (z) = . In the following two subsections.4. 10. The contribution from the contour goes to zero as |z| → ∞ and hence the contributions from the z = ikn and z = zj poles add up to zero. eτ Re z → 0. (10. for Re z > 0. Let us therefore consider the sum 1 F g0 (ikn )eikn τ . Furthermore.10. e.51) z − zj
j
where {zj } is the set of known poles and hence g0 (z) is analytic elsewhere in the z-plane. the contour integral itself gives zero because the integrand goes to zero exponentially for z ∈ C∞ (remember 0 < τ < β) nF (z)eτ z = eτ z ∝ eβz + 1 e(τ −β) Re z → 0. in the form of non-interacting Green’s functions like (10.
10.g. Because we know the poles of g0 a good choice for a contour is to take one that covers the entire complex plane C∞ : z = R eiθ where R → ∞. for Re z < 0.4. EVALUATION OF MATSUBARA SUMS
167
Figure 10. Such a contour would give us the contribution for poles of nF (z) plus the contributions from poles of g0 (z).

10. If it is the full Green’s function.168
CHAPTER 10. including for example the inﬂuence of interaction. it is known to be an analytic function in the entire complex plane exempt on the branch cuts.42). For bosons the derivation is almost identical and we get 1 β
B g0 (iωn )eiωn τ = S0 (τ ) = − iωn j z=zj
Res [g0 (z)] nB (zj )ezj τ . IMAGINARY TIME GREEN’S FUNCTIONS
Figure 10.56)
¡¢ ¢
.e. but we do know that it is analytic for z not on the real axis.55) (10. The contribution from the outer parts of the contour goes to zero as |z| → ∞ and hence only the paths parallel to the cut (here the real axis) contribute.53) (10.
10.
The Matsubara sum has thus been simpliﬁed considerably and we shall use this formula several times during the course.(10. we do not know the poles of the Green’s function. i. This general property of the Green’s function was shown in Sec. consider the sum S(τ ) = 1 β g(ikn )eikn τ .54) (10.2
Summations over functions with known branch cuts
The second type of sums we will meet are of the form in Eq.2. In general.3: The contour used to perform the Matsubara sum for a function with known branch cuts.
z=zj
Res [g0 (z)] nF (zj )ezj τ .
¨ § ¦¤ £ ¢ © ¥
© £ ¥ ¤ £ ¥ ¤ ¢ ¢ ¢ ¢ ¢ ¢ ¢
(10. Hence 0=
C∞
dz nF (z)g0 (z)ezτ 2πi g0 (ikn )eikn τ +
ikn F S0 (τ ) = j j z=zj
=− and thus
1 β
Res [g0 (z)] nF (zj )ezj τ .
ikn
τ > 0.4.

49). ikn )e−ikn 0 = S1 (ν. we have in fact found an expression for the expectation value of the occupation. (10.10. can also be used for the Matsubara functions.3. ε) eετ 2πi −∞ ∞ dε = nF (ε)A(ν. 2πi C1 +C2 ∞ 1 =− dε nF (ε) [g(ε + iη) − g(ε − iη)] eετ . 2π
(10. 8 to ﬁnd various Green’s functions. A](τ ) ∂ θ(τ − τ ) A(τ )B(τ ) ± θ(τ − τ ) B(τ )A(τ ) . to a negative inﬁnitesimal 0− .17) with respect to τ . (8. becomes in this way S1 (ν. In the second equality we used that G(ε − iη) = [G(ε + iη)]∗ which follows from Eq. −∞ 2π
(10. 10. They are shifted by an inﬁnitesimal amount η away from the real axis on either side dz S(τ ) = − nF (z)g(z)ezτ .
10. used in Chap. see Eq. (10. ∂τ = δ(τ − τ ) AB − (±)BA + Tτ [H.5
Equation of motion
The equation of motion technique. ε − iη)] eετ . Now setting the time argument in the single particle imaginary time Green’s ν function. Eq. (10.τ ) = −
∞ 1 dε nF (ε) [G(ν. 2πi −∞ ∞ 1 =− dε nF (ε)2i Im GR (ν. (10. (8. 0− ) ν 1 − = G(ν. A]e−τ H = [H.58)
according to the deﬁnition of the spectral function in Eq. A contour which includes all points z = ikn and no singularities of g is therefore C = C1 + C2 depicted in Fig. (10.42).33b).0+ ) β
ikn ∞
=
−∞
dε nF (ε)A(ν. ε + iη) − G(ν. we obtain −∂τ CAB (τ − τ ) =
. ε)eετ .29) with A = cν and B = c† .57).59)
which agrees with our previous ﬁnding. As for the example in the previous section. EQUATION OF MOTION
169
where it is known that g(z) is analytic in the entire complex plane except on the real axis. because c† cν = G(ν. Eq. (10. the part where |z| → ∞ does not contribution to the integral and we are left with the parts of the contour running parallel to the real axis.60) If we diﬀerentiate the Matsubara function Eq. the sum in Eq. (10. In the imaginary time formalism the time derivative of an operator is ∂τ A(τ ) = ∂τ eτ H Ae−τ H = eτ H [H.52).57) 2πi −∞ For example. ε). A](τ )B(τ ) .5.

66)
. and which is used in the example ending this chapter. ν τ ) = δ(τ − τ )δνν .61a) (10. Ψ(r)](τ )Ψ† (r .62)). .e. For example for free particle those given in Eqs.63b)
−∂τ G0 (ντ.62a) (10. r τ ) − dr h0 (r.65)
The average is taken with respect to a non-interacting Hamiltonian H0 (like Eq.states that for non-interacting particles.6
Wick’s theorem
We end this rather technical part by proving an extremely useful theorem. For the single-particle Green’s functions deﬁned in Eqs.
10.33). ν τ ) = δ(τ − τ )δνν + Tτ ([H. (10. r )Ψ(r ). . (10. i. when the Hamiltonian is quadratic. the equations of motion therefore in the two representations reduce to −∂τ G0 (rτ. .64)
This equation together with the boundary condition G(τ ) = ±G(τ + β) gives the solution. whereas the plus sign should be used for boson operators. ν1 τ1 . ν τ ) − or in matrix form
−1 G0 G0 = 1. r τ ) = δ(τ − τ )δ(r − r ). . r )G0 (r τ. IMAGINARY TIME GREEN’S FUNCTIONS
where the minus sign in AB − (±)BA is for fermion operators. The time-evolution is also with respect to H0 and it is given by c(τ ) = eτ H0 c e−τ H0 . cν ](τ )c† (τ )) . −∂τ G(ντ.
(10.61b)
For non-interacting electrons the Hamiltonian is quadratic. r τ ) = δ(τ − τ )δ(r − r ) + Tτ ([H. we then get for both fermion and boson Green’s functions −∂τ G(rτ. higher order Green’s function involving more than one particle can be factorized into products of single-particle Green’s functions. (10. ˆ (10. τ )) . . ν
In this case. νn τn ) = (−1)n Tτ cν1 (τ1 ) · · · cνn (τn )ˆ† (τn ) · · · c† (τ1 ) ˆ ˆ cν ˆν
n 1
(n)
0
.170
CHAPTER 10.
ν
(10.e. of the general form H0 = H0 =
νν
dr
dr Ψ† (r)h0 (r. ν (10.
(10. .37) and (10.called Wick’s theorem .62b)
tνν c† cν . νn τn .
−1 G0 = −∂τ − H0 . . which we have indicated by the subscript 0. tνν G0 (ν τ.
(10.38). The theorem . . which we will need later when doing perturbation theory. Consider an n-particle Green’s function deﬁned as G0 (ν1 τ1 . i.63a) (10.

b) then P = (3. In this case G (n) has the structure
G0
(n)
=
· · · θ(τi − τj ) · · · ± · · · θ(τj − τi ) · · ·
· · · cνi (τi )ˆ† (τj ) · · · ˆ cν
j
· · · c† (τj )ˆνi (τi ) · · · ˆν c j
. τi . For example if n = 2.g. ˆ j ∈ [1. 2)). 2n]. (10.72)
. so that we have
−1 G0i G0 (n) θ = −∂τi G0 . There are two such terms in (10. The easiest way to show Wick’s theorem is through the equation of motion for the n(n) particle Green’s function. This gives two kinds of contributions: the terms coming from the derivative of the theta functions and one term from the derivative of the expectation value itself. . τn . .
(10.63) and (10.69). j2n ) = (−1)n
P ∈S2n (n)
(±1)P θ(σP1 − σP2 ) · · · θ(σPn−1 − σPn )
0
× dP1 (σP1 ) · · · dP2n (σP2n )
. (n)
(10. (10.65) is indeed quite complicated to look at if we write out all the possible orderings and the conditions for that particular ordering. (10.70)
last term
which is similar to the derivation that lead to Eqs.65) are really ordered. .
(10. . we can rewrite G0 as G0 (j1 . Therefore if we (n) sum over all permutations and include the corresponding conditions. there are 4 time arguments which can be ordered in 4! diﬀerent ways. deﬁne the list (a.68)
where Pj denotes the j’th variable in the permutation P (e.64). b. 1.69)
where the factor (±1)P takes into account that for fermions (minus sign) it costs a sign change every time a pair of operators are commuted. Thus we diﬀerentiate G0 with respect to one of time arguments. and they will have diﬀerent order of the permutation.71)
−1 where G0i means that it works on the coordinate νi .6. a. Let us simplify the writing by deﬁning one operator symbol for both creation and annihilation operators dj (σj ) = c† ˆν (τ(2n+1−j) ).10. . corresponding to τi being either smaller or larger than τj . On the right hand side the derivative only acts on the theta functions in Eq. . (2n+1−j) cνj (τj ). The last one gives for example for τ1 − ∂ (n) G ∂τ1 0 ˆ ˆ = − (−1)n Tτ [H0 . τ1 . j ∈ [n + 1. .67)
and furthermore deﬁne the permutations of the 2n operators as P (d1 (σ1 ) · · · d2n (σ2n )) = dP1 (σP1 ) · · · dP2n (σP2n ). WICK’S THEOREM
171
The expression in (10. n]. . c) and the permutation (c.69). cν1 ](τ1 ) · · · cνn (τn )ˆ† (τn ) · · · c† (τ1 ) ˆ cν ˆν
n 1
. (10. Which permutation is the correct one of course depends on how the time arguments in (10. (10. Take now for example the case where τi is next to τj .

. . a constant. . .
without i without j
(10.
(n)
Now Eq. and hence
θ −∂τi G0 (n)
= [· · · ] · · · cνi (τi )ˆ† (τj ) · · · ˆ cν
j
[· · · ] · · · c† (τj )ˆνi (τi ) · · · ˆν c
j
δ(τi − τj ).. νj τj G0
(n−1)
(ν1 τ1 .74) cνi (τi ). 1 ) · · · .F
If τi is next to τj instead of τj .172
CHAPTER 10. G0 (n. . νn τn . With the plus sign it is called a permanent. bosons : (−1) = − (−1) (−1)
x n 1−n
(10. and for fermions (−1)n−i+n−j from moving c† next to cνi . .e.73) We can pull out the equal time commutator or anti-commutator for boson or fermions.n ) =
G0 (1. cνj (τi ) ˆ ˆ
B. IMAGINARY TIME GREEN’S FUNCTIONS
and when this is diﬀerentiated with respect to τi it gives two delta functions. .71) now looks like
−1 G0i G0 (n) n
=
j=1
δνi . . νn τn ). (−1)n from the deﬁnition in (10. (10.77a) (10.. c† (τi ) ˆ ˆν
j
B. (10.νj δ(τi − τj ) (−1)x G0
(n−1)
(ν1 τ1 . In fact. The number of creation and annihilation operators has thus both been reduced by one. .
i ≡ (νi . periodic in the time arguments. . We therefore end up with
(n) G0 (1. . . ν1 τ1 . n )
B. it gives the same result and hence
(n) G0 n
=
j=1
(±)j+i G0 νi τi . (10. ..F
= 0. Besides this undetermined sign.
(10. . τi )
(10.
(10. respectively = δνi . and it leaves a Green’s function which is no longer an n-particle Green’s function but an (n − 1)-Green’s function. we saw a special case of this in Eq.65) [(−1)n−1 ]−1 from the deﬁnition of G (n−1) . . . Hence ˆν ˆ
j
fermions: (−1)x = − (−1)n (−1)1−n (−1)2n−i−j = (−1)j+i .
without i without j
(10.. ν1 τ1 . and this sign denoted (−1)x will (for fermions) depend on the τj in question. . . .78)
By recalling the deﬁnition of a determinants this formula is immediately recognized as the determinant in the case where the minus sign should be used. i. .76) can be integrated and because G0 has the same boundary conditions as G0 .75)
which therefore does not contribute. νn τn ).63) where a one-particle Green’s function was reduced to a zero-particle Green’s function.F
.νj . . .e. n. G0 (n.76)
Let us collect the signs that go into (−1)x : (−1) from (−∂τ ). we get in the same manner the (anti-)commutator cνi (τi ). νn τn .77b)
= 1. our equation of motion (10. i. . 1
. . . .. n ) . 1 ) · · ·
G0 (1. What we have not determined is the sign of the new (n − 1)-Green’s function.79)
.

iqn ) = 1 β 1 V G0 (k + qσ. i.4. τ ) ρ (−q)) 0 .
G0 (k + qσ. ikn ).80)
and expresses it as a two-particle Green’s function χ0 (q.F means that for fermions it is the usual determinant. In order to ﬁnd χ (q. τ ) = − 1 Tτ (ρ (q.
10.82)
By Wick’s theorem this is given by a product of single-particle Green’s functions with all possible pairings and with the sign given by the number of times we interchange two fermion operators.83)
where we consider only q = 0 and use that G0 (k. we do it again here. this is Wick’s theorem. ω) = χ(q. τ )G0 (kσ.
kσ
(10.81) (10.1. the internal frequencies in the two come with the same sign χ0 (q. ikn ) = 1/ (ikn − ξk ). iqn → ω + iη).k . i. EXAMPLE: POLARIZABILITY OF FREE ELECTRONS
173
where we used a shorthand notation with the orbital and the time arguments being collected into one variable. τ ) = − 1 V Tτ c† (τ ) ck+qσ (τ ) c† σ ck −qσ kσ k
k. 10.σ
ck+qσ (τ ) c† σ k
0
Tτ
ck −qσ (0) c† kσ
(τ )
1 − ρ (q) V 0
=0 for q=0 0
ρ (−q) 0 . V (10. ω). 1 χ0 (q. χR (q.e. and where the determinant |·|B.k . to Fourier transform the product (10. The Fourier transform of a product in the time domain is a convolution in the frequency domain. 8. Remembering that G0 (kσ. iqn ) we start from the time-dependent χ χ0 (q. −τ ).54)
.7
Example: polarizability of free electrons
In Sec.84)
ikn
The sum over Matsubara frequencies has exactly the form studied in Sec. k ) ∝ δk.7.k σσ 0
. τ ) = V = 1 V Tτ
k. which follows from the corresponding Matsubara function by χR (q. ikn + iqn )G0 (kσ. Because one function has argument τ while the other has argument −τ .83). The next step is to calculate the frequency dependent function. while for bosons it should be understood as a permanent where all have terms come with a plus sign.10.(10. In order to illustrate the working principle of the imaginary time formalism.
kσ
(10.5 we calculated the polarizability of non-interacting free electrons.e.
(10.σ. The starting point is the physical quantity which is needed: the frequency dependent retarded charge-charge correlation function. we can read oﬀ the answer from Eq.

174

CHAPTER 10. IMAGINARY TIME GREEN’S FUNCTIONS

by inserting the poles of the two G0 (kσ, z) (z = ξk and z = ξk+q − iqn ) and obtain χ0 (q, iqn ) = = 1 V 1 V nF (ξk )G0 (k + qσ, ξk + iqn ) + nF (ξk+q − iqn )G0 (kσ, ξk+q − iqn )

k

nF (ξk ) − nF (ξk+q )

kσ

iqn + ξk − ξk+q

.

(10.85)

**Here we used that nF (ξk+q − iqn ) = 1 eβξk+q e−βiqn +1 = 1 e
**

βξk+q

+1

,

(10.86)

because iqn is a bosonic frequency. After the substitution (10.80) Eq. (10.85) gives the result we found in Eq. (8.81).

10.8

Summary and outlook

When performing calculations of physical quantities at ﬁnite temperatures it turns out that the easiest way to ﬁnd the “real time” introduced in Chap. 7 is often to go via the imaginary time formalism. This formalism has been introduced in this chapter, and in the following chapters on Feynman diagrams it is a necessary tool. There you will see why it is more natural to use the imaginary time Green’s function, also called Matsubara Green’s function. The reason is that the time evolution operator and the Boltzmann weight factor can be treated on an equal footing and one single perturbation expansion suﬃces. In the real time formalism there is no simple way of doing this. We have also derived some very useful relations concerning sums over Matsubara frequencies. The things to remember are the following. Non-interacting particle Green’s function, valid for both bosons and fermions G0 (ν, iωn ) = 1 . iωn − ξv (10.87)

**Matsubara frequency sum over products of non-interacting Green’s functions (for τ > 0) S F (τ ) = 1 β 1 β g0 (ikn )eikn τ =
**

ikn j

Res (g0 (zj )) nF (zj )ezj τ ,

ikn fermion frequency, (10.88a)

S B (τ ) =

**g0 (iωn )eiωn τ = −
**

iωn j

Res (g0 (zj )) nB (zj )ezj τ ,

iωn boson frequency, (10.88b)

with g0 (z) = i 1/ (z − ξi ) . If we perform a sum over functions where the poles are unknown but where the branch cuts are known, we can use a contour depicted in Fig.

10.8. SUMMARY AND OUTLOOK

175

10.3. For example if g(ikn ) is known to be analytic everywhere but on the real axis we get S F (τ ) = 1 β g(ikn )eikn τ = −

ikn ∞ −∞ ∞ −∞

dε nF (ε) [g(ε + iη) − g(ε − iη)] 2πi (10.89)

=−

dε nF (ε) g R (ε) − g A (ε) . 2πi

Finally, we proved an important theorem, Wick’s theorem, which says that for noninteracting an n-particle Green’s function is equal to a sum of products of single-particle Green’s functions, where all possible pairings should be included in the sum. For fermions we must furthermore keep track of the number of factors −1, because each time we interchange two fermion operators we must include a factor -1. The end result was

(n) G0 (1, . . . , n; 1

,...,n ) =

G0 (1, 1 ) · · · . .. . . . G0 (n, 1 ) · · ·

G0 (1, n ) . . . G0 (n, n )

B,F

,

i ≡ (νi , τi ) ,

(10.90)

where G0 (1, . . . , n; 1 , . . . , n ) = (−1)n Tτ c(1) · · · c(n)ˆ† (n ) · · · c† (1 ) ˆ ˆ c ˆ

(n) 0

.

(10.91)

176

CHAPTER 10. IMAGINARY TIME GREEN’S FUNCTIONS

Chapter 11

**Feynman diagrams and external potentials
**

From the previous chapters on linear response theory and Green’s functions, it is clear that complete calculations of thermal averages of time-dependent phenomena in quantum ﬁeld theory are a rather formidable task. Even the basic imaginary time evolution operˆ ˆ ator U (τ ) itself is an inﬁnite series to all orders in the interaction V (r, τ ). One simply faces the problem of getting lost in the myriads of integrals, and not being able to maintain a good physical intuition of which terms are important. In 1948 Feynman solved this problem as part of his seminal work on quantum electrodynamics by inventing the ingenious diagrams that today bear his name. The Feynman diagrams are both an exact mathematical representation of perturbation theory to inﬁnite order and a powerful pictorial method that elucidate the physical content of the complicated expressions. In this chapter we introduce the Feynman diagrams for the case of non-interacting particles in an external potential. Our main example of their use will be the analysis of electron-impurity scattering in disordered metals.

11.1

Non-interacting particles in external potentials

Consider a time-independent Hamiltonian H in the space representation describing noninteracting fermions in an external spin-diagonal single-particle potential Vσ (r): H = H0 + V =

σ

dr Ψ† (r)H0 (r)Ψσ (r) + σ

σ

dr Ψ† (r)Vσ (r)Ψσ (r). σ

(11.1)

As usual we assume that the unperturbed system described by the time-independent Hamiltonian H0 is solvable, and that we know the corresponding eigenstates |ν and 0 Green’s functions Gν . In the following it will prove helpful to introduce the short-hand notation

β

(r1 , σ1 , τ1 ) = (1)

and 177

d1 =

σ1

dr1

0

dτ1

(11.2)

178

CHAPTER 11. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS

for points and integrals in space-time. We want to study the full Green’s function, G(b, a) = − Tτ Ψ(b)Ψ† (a) , governed by ˆ ˆ H, and the bare one, G 0 (b, a) = − Tτ Ψ(b)Ψ† (a) 0 , governed by H0 . We note that since no particle-particle interaction is present in Eq. (11.1) both the full Hamiltonian H and the bare H0 have the simple form of Eq. (10.62), and the equations of motion for the two Green’s functions have the same form as Eq. (10.63):

[−∂τb −H0 (b)] G 0 (b, a) = δ(b−a) [−∂τb −H(b) ] G(b, a) = δ(b−a)

⇔ [−∂τb −H(b)+V (b)] G 0 (b, a) = δ(b−a) ⇔ G(b, a) = [−∂τb −H(b)]−1 δ(b−a),

(11.3a) (11.3b)

where we have also given the formal solution of G, which is helpful in acquiring the actual solution for G. Substituting δ(b − a) in Eq. (11.3b) by the expression from Eq. (11.3a) yields: [−∂τb − H(b)] G(b, a) = [−∂τb − H(b) + V (b)] G 0 (b, a) = [−∂τb − H(b)] G 0 (b, a) + V (b) G 0 (b, a) = [−∂τb − H(b)] G 0 (b, a) + (11.4) d1 δ(b − 1) V (1) G 0 (1, a).

Acting from the left with [−∂τb − H(b)]−1 gives an integral equation for G, the so-called Dyson equation, G(b, a) = G 0 (b, a) + d1 G(b, 1) V (1) G 0 (1, a), (11.5)

where we have used the second expression in Eq. (11.3b) to introduce G in the integrand. By iteratively inserting G itself in the integrand on the left-hand side we obtain the inﬁnite perturbation series G(b, a) = G 0 (b, a) + d1 G 0 (b, 1) V (1) G 0 (1, a) (11.6)

+ d1 d2 G 0 (b, 1) V (1) G 0 (1, 2) V (2) G 0 (2, a)

+ d1 d2 d3 G 0 (b, 1) V (1) G 0 (1, 2) V (2) G 0 (2, 3) V (3) G 0 (3, a) + . . . . The solutions Eqs. (11.5) and (11.6) for G are easy to interpret. The propagator, G, of a fermion in an external potential is given as the sum of all possible processes involving unperturbed propagation, described by G 0 , intersected by any number of scattering events V . So in this simple case there is really no need for further elucidation, but we will anyway proceed by introducing the corresponding Feynman diagrams. The ﬁrst step is to deﬁne the basic graphical vocabulary, i.e. to deﬁne the pictograms representing the basic quantities G, G 0 , and V of the problem. This vocabulary is known

11.2. ELASTIC SCATTERING AND MATSUBARA FREQUENCIES as the Feynman rules: b G(b, a) = G 0 (b, a) =

179

b d1 V (1) . . . = 1 (11.7)

a a Note how the fermion lines point from the points of creation, e.g. Ψ† (a), to the points of annihilation, e.g. Ψ(b). Using the Feynman rules the inﬁnite perturbation series Eq. (11.6) becomes b b b b b 1 = + 1 + 2 + 1 2 3 + ... (11.8)

a a a a a In this form we clearly see how the full propagator from a to b is the sum over all possible ways to connect a and b with bare propagators via any number of scattering events. We can also perform calculations by manipulating the diagrams. Let us for example derive an integral form equivalent to Eq. (11.5) from Eq. (11.8): b b b b b × 1 + 1 1 + 2 1 2 + . . . 3

=

+

=

+

1

(11.9)

a a a a a which by using the Feynman rules can be written as G(b, a) = G 0 (b, a) + d1 G 0 (b, 1) V (1) G(1, a).

a

a (11.10)

The former integral equation Eq. (11.5) for G is obtained by pulling out the bottom part V (n) G 0 (n, a) of every diagram on the right hand side of Eq. (11.8), thereby exchanging the arrow and the double-arrow in the last diagram of Eq. (11.9). This is a ﬁrst demonstration of the compactness of the Feynman diagram, and how visual clarity is obtained without loss of mathematical rigor.

11.2

Elastic scattering and Matsubara frequencies

When a fermion system interacts with a static external potential no energy is transferred between the two systems, a situation referred to as elastic scattering. The lack of energy

180

CHAPTER 11. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS

transfer in elastic scattering is naturally reﬂected in a particularly simple form of the single-particle Green’s function G(ikn ) in Matsubara frequency space. In the following the spin index σ is left out since the same answer is obtained for the two spin directions. First we note that since the Hamiltonian H in Eq. (11.1) is time-independent for static potentials we know from Eq. (10.19) that G(rτ, r τ ) depends only on the time diﬀerence τ − τ , and according to Eqs. (10.22b) and (10.25) it can therefore be expressed in terms of a Fourier transform with just one fermionic Matsubara frequency ikn : G(rτ, r τ ) = 1 β G(r, r ; ikn ) e−ikn (τ −τ ) ,

n β

G(r, r ; ikn ) =

dτ G(rτ, r τ ) eikn (τ −τ ) .

0

(11.11) The Fourier transform of the time convolution dτ1 G 0 (τb − τ1 )V G(τ1 − τa ) appearing in the integral equation of G is the product G 0 (ikn )V G(ikn ). The elastic scattering, i.e. the time-independent V , cannot change the frequencies of the propagators. In Matsubara frequency space the Dyson equation Eq. (11.10) takes the form G(rb , ra ; ikn ) = G 0 (rb , ra ; ikn ) + dr1 G 0 (rb , r1 ; ikn ) V (1) G(r1 , ra ; ikn ). (11.12)

As seen previously, the expressions are simpliﬁed by transforming from the |r -basis to the basis |ν which diagonalizes H0 . We deﬁne the transformed Green’s function in this basis as follows: Gνν ≡ drdr ν|r G(r, r ) r |ν ⇔ G(r, r ) =

νν

r|ν Gνν ν |r .

(11.13)

In a similar way we deﬁne the |ν -transform of V (r) as Vνν ≡ dr ν|r V (r) r|ν . In the |ν, ikn representation the equation of motion Eq. (11.3b) for G is a matrix equation, [(ikn − ξν )δν,ν − Vν,ν ] Gν

ν ,ν

(ikn ) = δν,ν

¯ ¯ or [ikn ¯ − E 0 − V ] G(ikn ) = ¯ (11.14) 1 ¯ 1,

¯ where E 0 is a diagonal matrix with the eigenenergies ξν = εν − µ along the diagonal. We have thus reduced the problem of ﬁnding the full Green’s function to a matrix inversion problem. We note in particular that in accordance with Eq. (10.40) the bare propagator G 0 has the simple diagonal form

0 (ikn − ξν )δν,ν Gν ν ,ν

(ikn ) = δν,ν

⇒

0 Gν,ν (ikn ) =

1 δ . ikn − ξν ν,ν

(11.15)

**We can utilize this to rewrite the integral equation Eq. (11.12) as a simple matrix equation, G(νb νa ; ikn ) = δν
**

b ,νa

G 0 (νa νa ; ikn ) +

νc

G 0 (νb νb ; ikn ) Vνb νc G(νc νa ; ikn ).

(11.16)

¯ We can also formulate Feynman rules in (ν, ikn )-space. We note that G 0 is diagonal in ¯ is a general matrix. To get the sum of all possible quantum processes one must ν, while V

and as a result the range of the potential is ﬁnite. while at low T only the electron-impurity scattering is eﬀective and gives rise to the non-zero value ρ0 of ρxx at T = 0. The frequency argument is suppressed. given by the so-called screening length a.νa
νa
+
νa
νb νc
(11.11. At high temperature the resistivity is mainly due to electron-phonon scattering.νa =
b
δν
b . This will be discussed in detail in Chap.18)
νa
Random impurities in disordered metals
An important example of elastic scattering by external potentials is the case of random impurities in a disordered metal. The valence of the impurity ions is one higher than the host ions. and since the vibrational
Figure 11.1(a) is sketched a number of randomly positioned impurities in an otherwise perfect metal lattice.16) diagrammatically: νb νb νa
0 Gν .ν
b
ikn − ξνa
Vνν =
ν ν
(11.1: (a) A disordered metal consisting of an otherwise perfect metal lattice with a number of randomly positioned impurities giving rise to elastic electron-impurity scattering.νa
=
δνa . (11. RANDOM IMPURITIES IN DISORDERED METALS
181
sum over all matrix indices diﬀerent from the externally given νa and νb . and as a ﬁrst approximation an impurity ion at site Pj gives rise to a simple screened mono-charge Coulomb potential uj (r) = −(e2 /|r − Pj |) e−|r−Pj |/a .3
=
δν
b . ikn )-space we can express Dyson’s equation Eq. At high T the electron-phonon scattering dominates giving rise to a linear behavior.3. νb νb Gνb νa =
νa νa Using these Feynman rules in (ν. (b) The electrical resistivity ρxx (T ) of the disordered metal as a function of temperature. 11. One well-controlled experimental realization of this is provided by a perfect metal Cu lattice with MgII ions substituting a small number of randomly chosen CuI ions.17)
11. The presence of the impurities can be detected by measuring the (longitudinal) resistivity ρxx of the metal as a function of temperature. In Fig. xx
© ©§ ©§¨¥ ¦¥ ¤ ¥ ¥ ¤¢ £
¢ ¡
. 13. The screening is due 0 to the electrons in the system trying to neutralize the impurity charge. . since the Green’s functions are diagonal in ikn .

16) as the unperturbed basis |ν . Finally.22)
Assume that u is only important in a sphere of radius a around the scattering center. typically a few kelvin. Thus u is weak if it is smaller than the smallest of these level spacings. Weak scattering means ˜ 2 /ma2 as follows:1 that u is much smaller than some characteristic level spacing ˜ u ˜ ma2
2
min {1. Now consider Nimp identical impurities situated at the randomly distributed but ﬁxed positions Pj . ikn ) = G (rb−ra . and hence the electron-phonon scattering rate. and the electron density nel is much smaller than unity: nimp nel 1. ra . at the lowest temperatures.
(11. ρ0 . The elastic scattering potential V (r) then acquires the form Nimp
V (r) =
j=1
u(r − Pj ).182
CHAPTER 11. ikn ). say ε = p2 /2m. 11.g.
Pj is randomly distributed. the level spacing is (∂ε/∂p) ∆p = (p/m) ( /a) = ka 2 /ma2 . (2. As a result the resistivity levels oﬀ at some value.19) the Dyson equation Eq. (11. where p = k has been used. We postpone the calculation of the resistivity and in this section just concentrate on studying the electron Matsubara Green’s function G for electrons moving in such a disordered metal.
(11.12) becomes
Nimp
G(rb . only the electron-impurity scattering is left preventing the Bloch electrons in moving unhindered through the crystal.21)
11. the number n of phonons. Exercise 3.3. One is stating that the ratio between the impurity density. We use the plane wave states |kσ from the eﬀective mass approximation Eq. The level spacing for non-perturbed states in that sphere is near the ground state given by the size quantization 2 /ma2 . We assume that the scattering potential u(r − Pj ) diﬀers only signiﬁcantly from 0 for |r − Pj | < a. For high energies around.
~
~
~ ~
. (11. The temperature behavior of the resistivity is depicted in Fig. and that the characteristic value in that region is u. ra . nimp = Nimp /V. (11.2).19)
Two small dimensionless parameters of the system serve as guides to obtain good approximative solutions. kF a}. At lower temperature the phonon degrees of freedom begin to freeze out and the phase space available for scattering also shrinks.1
Feynman diagrams for the impurity scattering
With the random potential Eq.1(b). ikn ) u(r1−Pj ) G(r1 .20)
The other small parameter is stating that the strength of the scattering potential is weak. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS
1 energy ω (n+ 2 ) in thermal equilibrium is proportional to kB T . (11. and consequently the resistivity becomes proportional to some power α of T . the so-called residual resistivity. is also proportional to T (see e. ikn ) +
j=1
1
0
dr1 G 0 (rb−r1 .

. ikn ). and therefore translation-invariant. ikn − ξk
0 Gk (r−r . Not only is it for all practical purposes impossible to know where all the impurities in a given metallic sample de facto are situated..15) Gk of the impurity free. RANDOM IMPURITIES IN DISORDERED METALS
183
where we have used the fact that the unperturbed system is translation-invariant and hence that G 0 (r1 . which may be interpreted as momentum conservaj j−1 j tion in each electron-impurity scattering: the change of the electron momentum is absorbed by the impurity. uk
2 −k1
0 Gk1 uk
.. uq2 Gk1 uq1 Gka e
−i(qn ·Pjn +..
jn
dr1 .kn−1
drn
2 1
(11. then we shall soon ﬁnd an answer. This complicated expression can be simpliﬁed a lot by performing the n spatial integrals. k
(11. problem has the simple form:
0 Gk (ikn ) =
1 .. . To this end we reformulate Dyson’s 0 equation in k space since according to Eq.25)
The Fourier expansion of G (n) (rb . ikn ) in Eq. . The n-th order term G (n) is
Nimp Nimp
G
(n)
(rb .+(k1 −ka )·Pj ]
1
j1 . n=0 where the frequency argument ikn has been suppressed. u(r2 −Pj2 ) G 0 (r2 −r1 ) u(r1 −Pj1 ) G 0 (r1 −ra ). ra ). Naturally.jn
1 Vn
b
q1 .23)
× G 0 (rb −rn ) u(rn − Pjn ) . no simple solution for the Green’s function could be found..
.. .+q2 ·Pj +q1 ·Pj )
×eikb ·(rb −rn ) eiqn ·rn eikn−1 ·(rn −rn−1 ) ... . drj ei(kj −kj−1 −qj )·rj = V δk . ra ) =
j1 . (11. (11.k +q . Utilizing these delta functions in the n q-sums leads to G
(n)
1 (rb ra ) = 2 V
Nimp
e
ka kb
b
ikb ·rb −ika ·ra
e
1 V n−1
k1 . but even if we did..
drn
(11.
q
(11.11. . ... .27)
0 Gka e −i[(kb −kn−1 )·Pjn +. . we can never hope to solve this problem exactly. . ra ) = ∞ G (n) (rb .22) in orders n of the scattering potential u(r−Pj )..kn−1
1 −ka
(11.23) is:
Nimp
G
(n)
(rb . However. First expand the Dyson equation Eq. eiq2 ·r2 eik1 ·(r2 −r1 ) eiq1 ·r1 eika ·(r1 −ra ) .24)
The Fourier transform of the impurity potential u(r−Pj ) is: u(r−Pj ) = 1 V uq eiq·(r−Pj ) =
q
1 V
e−iq·Pj uq eiq·r .qn
1 V2
1
ka kb
V n−1
dr1 . ra . (11. if we are satisﬁed with the answer to the less ambitious and more practical question of what is the average behavior.jn
b −kn−1
0 ×Gk uk
0 Gkn−1 . This n-th order contribution can be interpreted as the sum over all processes involving n scattering events in all possible combination of impurities. We now want to deduce the Feynman rules for constructing diagrams in this situation. ikn ) =
1 V
0 Gk (ikn ) eik·(r−r ) . . ra . ra ) =
j1
. .26)
0 0 0 0 ×Gk uqn Gkn−1 uqn−1 .
k1 . ikn ) = G 0 (r1 −ra ..3.. and obtain G(rb ...

0 go into vertex •1 and G 0 away from vertex •n...31)
=
··· n kb kn−1 k3 3 k2 2 k1 1 ka This diagram is very suggestive.
kb −kn−1
k3 −k2 k2 −k1 k1 −ka
11. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS
(n) b ka
Introducing the Fourier transform Gk G (n) (rb . uk
2 −k1
0 Gk1 . .184
CHAPTER 11. Perform the sums
1 V
over all internal momenta kj . We therefore begin to consider the possibility of performing an average over the random positions Pj of the impurities. e. (11.29) is: Pn Gk
(n) b ka
P3
P2
P1 (11. b ka
1 V2
(11. and
Nimp j1 . If one imagine changing some external parameter.. ra ) as eikb ·rb e−ika ·ra Gk
ka kb (n) .+(k1 −ka )·Pj ]
1
(11. ra ) = with
Nimp (n) Gk ka b
of G (n) (rb .g. 0 Let straight arrows k denote Gk . it is not possible to continue the study of impurity scattering on general grounds without further assumptions. However. Draw n scattering events. as mentioned above.29)
0 × Gk uk
b −kn−1
0 Gkn−1 . each individual
. uk
1 −ka
0 Gka . (n) : b ka
we can now easily deduce the Feynman rules for the diagrams corresponding to Gk (1) (2) (3) (4) (5) (6) (7)
Let dashed arrows j q.28)
=
j1 .jn
1 V n−1
b
e
k1 . Let Gka kb 0 Let Gk go from vertex j to vertex j + 1..jn
over Pjl ..30)
The diagram corresponding to Eq.. . the average electron density or an external magnetic ﬁeld.kn−1
−i[(kb −kn−1 )·Pjn +. . . One can see how an incoming electron with momentum ka is scattered n times under momentum conservation with the impurities and leaves the system with momentum kb .
j
kj
Maintain momentum conservation at each vertex.. (11.4
Impurity self-average
If the electron wavefunctions are completely coherent throughout the entire disordered metal each true electronic eigenfunction exhibit an extremely complex diﬀraction pattern spawned by the randomly positioned scatterers. Pj denote a scattering event uq e−iq·Pj .

when one measures an observable the result is in fact an incoherent average of all these sub-systems. Any n-th order contribution to Gk contains n scattering events. but they need not be on n diﬀerent scatterers.g.4.1 K) lϕ is much smaller than the size of the device. Using modern nano-technology to fabricate small (but still macroscopic) samples. For very large (mm sized) macroscopic samples lϕ is much smaller than the sample size at all experimental realizable temperatures (T > 10 mK for electron gases in metals and semiconductors). As the temperature of a given sample is raised.ka Gka ≡
b
δk
imp
b . In Fig.1 K is seen to be much smoother than the one at 0.32) Here we have anticipated that the impurity averaged Green’s function is diagonal in k due to the restoring of translation-invariance upon average. one can in fact obtain an experimental situation where the electrons can traverse the sample without loosing their quantum-mechanical phase coherence.. 11..
1 V
dPNimp Gka
(11. and we are in the impurity self-averaging case. and standard cryogenic equipment to cool down these samples to ultra-low temperatures.ka
Nsys sys Gka i ∼ δk i=1
1
Nsys
b . This density can be controlled by applying a gate voltage Vg on an external electrode. 11. These ﬂuctuations turn out to be perfectly reproducible as Vg is swept up and down several times.as can be seen from Fig. and one can as well perform the position average over the entire volume of the sample. the impurity average is performed by summing over all the phase-independent coherent sub-systems and dividing by their number Nsys . 11. such as the one shown in Fig. Mathematically.ka V
dP1
1 V
dP2 . and where the many small phase coherent sub-systems of the sample are indicated below the experimental graph. 2. IMPURITY SELF-AVERAGE
185
diﬀraction pattern will of course change drastically due to the sensitivity of the scattering phases of the wavefunctions. this average is the same as an average over the impurity position within a single subsystem . and this eﬀective averaging is consequently denoted self-averaging. In fact. Thus in the following we average over all possible uncorrelated positions Pj of the Nimp impurities for the entire system: 1 Gkb ka V ¯ ≡ δk . This eﬀect is illustrated in Fig. 4. the amount of electron-electron and electron-phonon scattering increases because of an increased phase space for scattering and an increased number of phonons. even on the rather small length scale lϕ the system is already homogeneous.11. At suﬃciently high temperature (e.31 K as a function of the electron density. Some care must be taken regarding the average over the impurity positions Pj . Signiﬁcant quantum ﬂuctuations must therefore occur in any observable at suﬃciently low temperatures. and we can think of the device as being composed as a number of phase-independent small phase coherent sub-systems. and as a result the coherence length lϕ for the electrons diminishes.31 K. The quantum mechanical phase of each individual electron is changed by a small random amount at each inelastic scattering event.2(b) where the conductance trace at 4. Note that this average is imposed by the physical properties of the system itself.2. Therefore.2(a) is shown the conductance trace of a GaAs nano-device. any
. However. The conductance G is seen to ﬂuctuate strongly for minute changes of Vg . at 0.10. But due to the random distribution of the impurities.

30 0.186
¢ ¡
CHAPTER 11.1 K. The sample now contains a large number of independent but phase-coherent sub-systems of size lϕ . down by the small factor nimp /nel .31 K
G (e2/h)
Figure 11.30 0. Then. etc.20 0. 1 ≤ p ≤ n of scatterers could be involved. which suppresses the quantum ﬂuctuations. (b) The same system at T = 4.29) the only reference to i(q ·P +q ·P +.20) we work in the limit of small impurity densities nimp .31 K displaying random but reproducible quantum ﬂuctuations as a function of a gate voltage Vg controlling the electron density.+qn ·Pjn ) .1 K
.50 Vg (V)
T = 0.2: (a) The measured conductance of a disordered GaAs sample at T = 0. As a result a substantial self-averaging occurs. Below is indicated that the phase coherence length lϕ is large compared to the size of the sample. We note that in Eq.29) over impurity positions in this exponential is now ordered according to how many impurities are involved:
¥
¤
¢
£
¥ ¤
5 4 3 2 1 0 0.
number p. follow processes involving two impurities. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS
5 4 G (e2/h) 3 2 1 0 0.40 0.
As mentioned in Eq. with the scattering the impurity positions is the exponential e 1 j1 2 j2 vectors qi = ki − ki−1 .20 0. (11. The ﬂuctuations are almost gone due to the smallness of lϕ at this temperature.40 0. The ﬂuctuations are due to phase coherent scattering against randomly positioned impurities. The sum in Eq. We must therefore carefully sort out all possible ways to scatter on p diﬀerent impurities. (11. For a given ﬁxed number n of scattering events the most important contribution therefore comes from processes involving just one impurity...50 Vg (V)
T = 4. (11.

.33)
Here Q = {q1 .. .
(11.. .3.jn
i
Pn
l=1
ql ·Pj
Nimp
l
=
h1 i(
P
e
i(
P
q ∈Q j1
qj )·Ph
1
1
Nimp Nimp
+
Q1 ∪Q2 =Q h1 h2
e
q ∈Q1 l1
ql )·Ph
1
1
e
P i( q
l2 ∈Q2
ql )·Ph
2
2
Nimp Nimp Nimp
+
Q1 ∪Q2 ∪Q3 =Q h1 h2 h3
e
P i( q
l1
∈Q1
ql )·Ph
1
1
e
P i( q
l2 ∈Q2
ql )·Ph
2
2
e
i(
P
ql ∈Q3 3
ql )·Ph
3
3
+ . Fig. (11. 11.38) the impurity averaged Green’s function is a sum scattering processes against the position-averaged impurities.4. . However. As depicted in Eq. IMPURITY SELF-AVERAGE
187
Nimp
e
j1 . But let us see how these conclusions are reached. while Q1 ∪Q2 ∪.. The impurity average indicated in Eq.11.34)
i
This of course no longer depends on the p impurity positions Phi .35)
i
This error occurs since our approximation amounts to saying that the (p + 1)-st impurity can occupy any of the Nimp impurity sites. q2 . that strictly speaking two diﬀerent impurities cannot occupy the same position. in Eq. qn } is the set of the n scattering vectors.
.32) over each exponential factor independently. (11.P
h=1 h=1 Qh =Q
q ∈Qh hi
qh
. The detailed calculation is straightforward but somewhat cumbersome. All the scattering vectors in one particular subset Qi are connected to the same impurity Phi . The result of the impurity averaging can now be written as
Nimp
e
j1 .2 Since all the p positions Ph now are manifestly diﬀerent we can perform the impurity average indicated in Eq.. the sum of all scattering momenta on the same impurity must be zero.32) over each independent exponential factor results in some Kronecker delta’s meaning that all scattering vectors qhi connected to the same impurity must add up to zero: e
P i( q
∈Qi hi
qh )·Ph
i
i
=
imp
1 V
dPhi e
i(
P
q ∈Qh hi
qh )·Ph
i
i
= δ0. . cf.P
q ∈Qh hi
qh
. Since translation-invariance is restored by the averaging. For the important terms p Nimp and the error is p/Nimp 1. (11.. This introduces a small error of the order 1/Nimp for the important terms in the low impurity density limit involving only a few impurities. the result may perhaps be easier to understand than the derivation. Note... and not just the Nimp −p available sites.
(11. (11.33) we let the j-sums run unrestricted.jn
2
i
Pn
l=1
ql ·Pj
n
p
Sp
l
=
imp p=1
Nimp δ0.
(11. . .∪Qp = Q denotes all possible unions of non-empty disjunct subsets spanning Q. the averaging has restored translation-invariance.

188

CHAPTER 11. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS

**which, when inserted in Eq. (11.29), leads to Gk
**

(n) imp

=

1 V n−1

n k1 ...kn−1 p=1

1

p

Sp

Nimp δ0,P

0 Gk2 . . . uk

n−1 −k

h=1 h=1 Qh =Q 0 Gk .

Qh (khi −k(hi −1) )

0 0 × Gk uk −k Gk1 uk

1 −k2

(11.36)

We note that due to the p factors containing δ-functions there are in fact only n − 1 − p 1 free momenta sums V k to perform. The remaining p volume prefactors are combined with Nimp to yield p impurity density factors nimp = Nimp /V. The Feynman rules for constructing the n-th order contribution Gk purity averaged Green’s function Gk imp are now easy to establish: (1) (2) (3) (4) (5) (7) (8) (9) (10) (11)

(n) imp

to the im-

Let scattering lines q denote the scattering amplitude uq . Let denote a momentum conserving impurity averaged factor nimp δ0,P q . 0 Let fermion lines k denote the unperturbed Green’s function Gk . Draw p impurity stars. Let n1 scattering lines go out from star 1, n2 from star 2, etc, so that the total number n1 + n2 + . . . + np of scattering lines is n. Draw all topological diﬀerent diagrams containing an unbroken chain of n + 1 fermion lines connecting once to each of the n scattering line end-points. 0 Let the ﬁrst and last fermion line be Gk . Maintain momentum conservation at each vertex. Make sure that the sum of all momenta leaving an impurity star is zero. 1 Perform the sum V k over all free internal momenta kj .

j

Sum over all orders n of scattering and over p, with 1 ≤ p ≤ n. (11.37)

**The diagrammatic expansion of Gk Gk
**

imp

imp

has a direct intuitive appeal: +

=

+ + +

+

!

+ + + ··· +

(11.38)

" # $ % &

+ +

In this expression, showing all diagrams up to third order and three diagrams of fourth order, we have for visual clarity suppressed all momentum labels and even the arrows of the scattering lines. For each order the diagrams are arranged after powers of nimp , i.e. the number of impurity stars. In Fig. 11.3 two diagrams with complete labels are shown. In the following section we gain further insight in the solution of Gk imp by rearranging the terms in the diagrammatic expansion, a procedure known as resummation.

'

+ ··· +

(

)

+ ···

Figure 11.3: Two fully labelled ﬁfth order diagrams both with two impurity scatterers. Diagram (a) is a so-called irreducible diagram, i.e. it cannot be cut into two pieces by cutting one internal fermion line. In contrast, diagram (b) is reducible. It consists of two irreducible parts.

11.5

In Fig. 11.3 we introduce the concept of irreducible diagrams, i.e. diagrams in the expansion of Gk imp that cannot be cut into two pieces by cutting a single internal fermion line. We now use this concept to resum the diagrammatic expansion Eq. (11.38) for Gk imp . We remind the reader that this resummation is correct only in the limit of low impurity density. First we deﬁne the so-called self-energy Σk : Σk ≡ = = The sum of all irreducible diagrams in Gk 0 without the two external fermion lines Gk

imp

**Using Σk and the product form of Gk Gk
**

imp

**This algebraic Dyson equation, equivalent to Eqs. (11.9) and (11.18), is readily solved: Gk (ikn )
**

imp

From this solution we immediately learn that Σk enters Gk imp as an additive correction to the original unperturbed energy, ξk → ξk + Σk , hence the name self-energy. The problem

¤ © ¨ ¤ © ¥ ¤ © ¤ © § ¤ © ¤ © ¤ © ¨ ¤ © ¥ ¤ © § ¤ © ¥ ¤ ¦ § ¤ ¤ © ¤ © ¤ ¦ ¨ ¤ £ ¨ ¤ ¦ ¥ ¤ £ ¥ ¤ ¦ ¤ £ § ¤ ¦ ¤ £ ¤ ¦ ¨ ¤ £ ¨ ¤ ¦ ¥ ¤ ¥£ ¤ ¦ ¤ £ § ¤ ¦ ¥ ¤ £ ¤ ¦ § ¤ £ ¢ ¥ ¤ ¦ §¤ £ ¢ ¡

Self-energy for impurity scattered electrons

11.5. SELF-ENERGY FOR IMPURITY SCATTERED ELECTRONS

189

* + , /

+ + +

imp

+

.

+ ···

+ ··· (11.39)

= =

0 0 = Gk + Gk Σk Gk

0 1 2 3 4 5 6

+ + + × +

imp . 0 Gk

in Fourier space, Eq. (11.38) becomes + ...

+ ... (11.40)

=

1−

0 Gk

Σk

=

1

0 (Gk )−1

− Σk

=

1 . ikn − ξk − Σk (ikn )

(11.41)

190

CHAPTER 11. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS

of ﬁnding Gk imp is thus reduced to a calculation of Σk . In the following we go through various approximations for Σk .

11.5.1

Lowest order approximation

One marvellous feature of the self-energy Σk is that even if it is approximated by a ﬁnite number of diagrams, the Dyson equation Eq. (11.40) actually ensures that some diagrams of all orders are included in the perturbation series for Gk imp . This allows for essential changes in Gk imp , notably one can move the poles of Gk imp and hence change the excitation energies. This would not be possible if only a ﬁnite number of diagrams were used in the expansion of Gk imp itself. Bearing in mind the inequalities Eqs. (11.20) and (11.21), the lowest order approximation ΣLOA to Σk is obtained by including only the diagram with the fewest number of k impurity stars and scattering lines, ΣLOA (ikn ) ≡ k

7

= nimp u0 = nimp

dr u(r),

(11.42)

**i.e. a constant, which upon insertion into Dyson’s equation Eq. (11.41) yields
**

LOA Gk (ikn ) =

1 . ikn − (ξk + nimp u0 )

(11.43)

But this just reveals a simple constant shift of all the energy levels with the amount nimp u0 . This shift constitutes a redeﬁnition of the origin of the energy axis with no dynamical consequences. In the following it is absorbed into the deﬁnition of the chemical potential and will therefore not appear in the equations.

11.5.2

1st order Born approximation

The simplest non-trivial low-order approximation to the self-energy is the so-called ﬁrst order Born approximation given by the ’wigwam’-diagram Σ1BA (ikn ) ≡ k

8

k−k k −k k

= nimp

k

|uk−k |2

1 , ikn − ξk

(11.44)

where we have used that u−k = u∗ since u(r) is real. We shall see shortly that Σ1BA = k k Re Σ1BA + i Im Σ1BA moves the poles of Gk imp = away from the real axis, i.e. k k 1BA the propagator acquires a ﬁnite life-time. By Eq. (11.40) we see that Gk is the sum of propagations with any number of sequential wigwam-diagrams:

9

In the evaluation of Σ1BA we shall rely on our physical insight to facilitate the math. k We know that for the electron gas in a typical metal εF ∼ 7 eV ∼ 80 000 K, so as usual only electrons with an energy εk in a narrow shell around εF ≈ µ play a role. For

: ; < = >

1BA

=

+

+

+

+ ···

(11.45)

11.5. SELF-ENERGY FOR IMPURITY SCATTERED ELECTRONS

191

Figure 11.4: (a) The functions nimp |uk |2 and (ω − εk + µ)/[(ω − εk + µ)2 + η 2 ] appearing in the expression for Re Σ1BA (ikn ). (b) The functions nimp |uk |2 and |kn |/[(ω − εk + µ)2 + η 2 ] k appearing in the expression for Im Σ1BA (ikn ). k T < 800 K we have kB T /εF < 10−2 , and for applied voltage drops Vext less than 70 mV over the coherence length lϕ < 10−5 m (the typical size we are looking at), i.e. applied electrical ﬁelds less than 7000 V/m, we have eVext /εF < 10−2 . Thus we are only interested in Σ1BA (ikn ) for k |k| ∼ kF and |ikn → ω + i sgn(kn )η| εF . (11.46) Here we have also anticipated that at the end of the calculation, as sketched in Fig. 10.1, we need to perform an analytical continuation down to the real axis, either from the upper half-plane, where kn > 0, as ikn → ω + iη, or from the lower half-plane, where kn < 0, as ikn → ω − iη. Furthermore, as we shall study in great detail later, the electron gas redistributes itself to screen out the external charges from the impurities, and as a result uk−k varies in a smooth and gentle way for 0 < |k − k | < 2kF . With this physical input in mind we continue: 1 (11.47) Σ1BA (ω + i sgn(kn )η) = nimp |uk−k |2 k (ω − ξk ) + i sgn(kn )η

k

=

k

nimp |uk−k |2

ω − ξk − i sgn(kn ) πδ(ω − ξk ) . (ω − ξk )2 + η 2

Since |uk−k |2 vary smoothly and |ω − ξk | εF ≈ µ we get the functional behavior shown in Fig. 11.4. Since (ω − ξk )/((ω − ξk )2 + η 2 ) is an odd function of ω − ξk and the width η is small, we have3 Re Σ1BA (ikn ) ≈ 0; For the imaginary part of Σ1BA we obtain the k usual delta function for η → 0. Finally, since the spectral function for the unperturbed system forces ω to equal ξk , we obtain: Σ1BA (ikn ) = −iπ sgn(kn ) k

k

3

nimp |uk−k |2 δ(ξk − ξk ) = −i sgn(kn )

Strictly speaking, we only get vanishing real part if the slope of |uk−k |2 is zero near µ. If this is not the case we do get a non-zero real part. However, since |uk−k |2 is slowly varying near µ we get the same real part for all k and k near kF . This contribution can be absorbed into the deﬁnition of µ.

§ ¤ ¥

§ ¨© ¢ £

§ ¦¤ ¥

§ ¨© ¢ ¡

1 , 2τk

(11.48)

192

CHAPTER 11. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS

**where we have introduced the impurity scattering time τk deﬁned as 1 ≡ 2π τk nimp |uk−k |2 δ(ξk − ξk ).
**

k

(11.49)

This result can also be found using Fermi’s golden rule. Now we have obtained the 1st order Born approximation for Gk (ikn ) in Eq. (11.41) and the analytic continuation ikn → z thereof into the entire complex plane: z−ξ 1 i , Im z > 0 k + 2τ 1 1BA 1BA k Gk (ikn ) = −→ Gk (z) = (11.50) 1 ikn − ξk + i sgn(kn ) ikn →z z−ξ − i , Im z < 0. 2τ

k k 2τ k

1BA We see that Gk (z) has a branch cut along the real axis, but that it is analytic separately in the upper and the lower half-plane. This is a property that will play an important role later, when we calculate the electrical resistivity of disordered metals. Note that this behavior is in accordance with the general results obtained in Sec. 9.2 concerning the analytic properties of Matsubara Green’s functions. We close this section by remarking three properties summarized in Fig. 11.5 related to the retarded Green’s function GR,1BA (ω ) = G 1BA (ω + iη). First, it is seen by Fourier R,1BA transforming to the time domain that Gk (t) decays exponentially in time with τk as the typical time scale:

Gk

R,1BA

(t) ≡

e−i(ω +iη)t dω = −i θ(t) e−iξk t e−t/2τk . 2π ω − ξk + i/2τk

(11.51)

−1 Second, exploiting that ω , τk εF , it is seen by Fourier transforming back to real space R,1BA (r, ω ) decays exponentially in space with l ≡ v τ as the typical length scale: that G k F k

GR,1BA (r, ω ) ≡

πd(εF ) ik |r| −|r|/2lk dk eik·r = e F e . (2π)3 ω − ξk + i/2τk kF |r|

(11.52)

**Thirdly, the spectral function A1BA (ω ) is a Lorentzian of width 2τk : k
**

1BA A1BA (ω ) ≡ −2 Im Gk (ω + iη) = k 2 (ω − ξk )2 + 1/4τk

1/τk

(11.53)

In conclusion the impurity averaged 1st Born approximation has resulted in a selfenergy with a non-zero imaginary part. The poles of the Matsubara Green’s function 1BA Gk (ikn ) are therefore shifted away from the real axis, resulting in a both temporal and spatial exponential decay of the retarded Green’s function. This is interpreted as follows: the impurity scattering transforms the free electrons into quasiparticles with a ﬁnite life time given by the scattering time τk and a ﬁnite mean free path given by lk = vF τk . The ﬁnite life time of the quasiparticles is also reﬂected in the broadening of the spectral function. The characteristic sharp δ-function for free electrons, Ak (ω ) = 2πδ(ω − ξk ), is

11.5. SELF-ENERGY FOR IMPURITY SCATTERED ELECTRONS

193

Figure 11.5: (a) The impurity averaged Green’s function Gk (ikn ) imp . The imaginary part of the self-energy is related to the scattering time τk and hence also to the elastic scattering length l = vF τk . (b) In the Born approximation the spectral function Ak (ω ) is a Lorentzian centered around ξk = 0 with a width 1/2τk . broadened into a Lorentzian of width 1/2τk . This means that a particle with momentum k can have an energy ω that diﬀers from ξk with an amount /2τk . This calculation of self-averaged impurity scattering constitutes a ﬁrst and very important example of what can happen in a many-particle system. Note in particular the important role played by the self-energy, and the fact that it can have a non-zero imaginary part. The results is obtained in the 1st order Born approximation, where the self-energy is approximated by a single diagram. But what happens if we take more diagrams into account? The surprising answer is that in the low impurity density limit, nimp nel no qualitative diﬀerence arises by taking more diagrams into account. Only at higher impurity densities where scattering events from diﬀerent impurities begin to interfere new physical eﬀects, such as weak localization, appear. Let us see how this conclusion is reached.

11.5.3

The full Born approximation

A natural extension of the 1st Born approximation is the full Born approximation, which is exact to lowest order in nimp . It is deﬁned by the following self-energy ΣFBA (ikn ), where k any number of scattering on the same impurity is taken into account, i.e. more dashed lines on the wigwam-diagram:

ΣFBA k

≡

?

+

k k−k

kk

@

×

k

+

k

k

A

+

k

+

k

k

B

+

k

+ ···

k

=

C D

+

kk k

δk ,k

E

F

k

G

+ ···

k

In the parenthesis at the end of the second line we ﬁnd a factor, which we denote tk ,k , that is not diagonal in k but with a diagonal that equals the self-energy tk,k = ΣFBA . In k scattering theory tk ,k is known as the transition matrix. When this matrix is known all

@

& RQ © P © P B S

**©¨I © ¨ GH4 & AFB ' E %& $ & 4 AB CD ¤ ¢ @ 9© 8¢ 6 7
**

(11.54)

3 # & 12 45 !()0 ' %& $ !"# ¤ ¢ © ¢ ¡ © ¨ § ¦¤ £ ¥

194

CHAPTER 11. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS

consequences of the complete scattering sequence can be calculated. An integral equation for the transition matrix is derived diagrammatically:

tk

1 ,k2

(ikn ) ≡

δk =

H

+

1 ,k2

k1

I

k1−k

+

k2

k1

J

δk ,k

+ ···

k2

δk

K

+

1 ,k2

k1

L

+

×

k

M

+

2

k

N

+

k

k2

O

+ ···

k2

= nimp u0 δk

1 ,k2

uk

k

1 −k

0 Gk tk ,k .

2

(11.55)

This equation can in many cases be solved numerically. As before the task is simpliﬁed by the fact that we are only interested at electrons moving at the Fermi surface. The real part of the diagonal element tk,k (ikn ), the one yielding the self-energy, is almost constant for |k| ∼ kF and is absorbed into the deﬁnition of the chemical potential µ. We are then 0 left with Imtk,k (ikn ), and by applying the optical theorem,4 Imtk,k = Im k t† Gk tk ,k , k,k we obtain Im ΣFBA (ikn ) k = Im tk,k (ikn ) = Im

k ikn →ω +i sgn(kn )η

|tk,k |2 ikn − ξk |tk,k |2 δ(ω − ξk ).

k

−→

−sgn(kn ) π

(11.56)

This has the same form as Eq. (11.48) with |tk,k |2 instead of nimp |uk−k |2 , and we write ΣFBA (ikn ) = −i sgn(kn ) k 1 , 2τk with 1 ≡ 2π τk |tk,k |2 δ(ξk − ξk ).

k

(11.57)

By iteration of Dyson’s equation we ﬁnd that G FBA is the sum of propagations with any number and any type of sequential wigwam-diagrams:

P Q R S T

FBA

=

+

+

+

+ ···

(11.58)

11.5.4

The self-consistent Born approximation and beyond

Many more diagrams can be taken into account using the self-consistent Born approximation deﬁned by substituting the bare G 0 with the full G in the full Born approximation

Eq. (11.55) states (i): t = u + uG 0 t. Since u† = u the Hermitian conjugate of (i) is (ii): u = −t (G 0 )† u + t† . Insert (ii) into (i): t = u + (t† G 0 t − t† (G 0 )† uG 0 t). Both u and t† (G 0 )† uG 0 t are Hermitian P 0 so Im tk,k = Im k|t† G 0 t|k = Im k t† Gk tk ,k . k,k

† 4

k Σi = Im tk.59)
= nimp u0 δk. 11. since k k the small real part ΣR can be absorbed into µ. The ﬁnal result is ΣSCBA (ikn ) = −i sgn(kn ) k 1 . i.k |2 δ(ω −ξk ). In Fig.55) yields: ΣSCBA k ≡
195
U V W X
+ + + uk−k Gk tk . (11. The only diﬀerence between the full Born and the self-consistent Born approximation is in the case of strong scattering. (11. Only the imaginary part Σk of Σk = ΣR + iΣi plays a role. We again utilize k k that tk. At zero temperature the energy broadening around the Fermi energy εF is given by
. SELF-ENERGY FOR IMPURITY SCATTERED ELECTRONS Eqs.k = Im k
k
|tk.k .k +
a self-consistent equation in ΣSCBA since Gk = (ikn − ξk − ΣSCBA )−1 .k |2 δ(ω − ξk ). |Σk | εF we obtain almost the same result i SCBA as in Eq.54) and (11.11.k |2 ikn −ξk −iΣi k
≈ −sgn(kn −Σi ) π k
k
|tk. Also sketched is the phase space Ω available for the internal momenta k1 and k2 in the two cases. (11.e.62)
+
`
+ ··· +
a
+ ···
We have now resummed most of the diagrams in the diagrammatic expansion of Gk imp with the exception of wigwam-diagrams with crossing lines. where the limiting δ-function in Eq. and if furthermore SCBA the scattering strength is moderate. But this requirement is fulﬁlled by taking Im ΣSCBA (ikn ) ∝ −sgn(kn ) as seen by direct substitution. (11.6 are shown two diﬀerent types of irreducible diagrams of the same order in both nimp and uk .60)
The only self-consistency requirement is thus connected with the sign of the imaginary part.60) may acquire a small renormalization.
k
+ ··· (11. 2τk with 1 ≈ 2π τk |tk.5.56).
k
(11.61)
By iteration of Dyson’s equation we ﬁnd that G SCBA is the sum of propagations with any number and any type of sequential wigwam-diagrams inside wigwam-diagrams but without crossings of any scattering lines:
Y
SCBA
= +
Z [ \ ] ^ _
+ + + + + ···
+ ··· (11.k is only weakly dependent on energy for |k| ≈ kF and ω εF .

It is interest2τk ing to note that in e. Such a suppression factor enters the calculation for each crossing of scattering lines in a diagram. 11. the other is bound to the intersection between two Fermi shells.6: (a) The non-crossing wigwam diagrams. In k-space the broadening ∆k is given by 2 (k + ∆k)2 /2m ≈ ε + /τ which gives ∆k ≈ 1/v τ = 1/l. to a ring with cross section 1/l2 and radius ≈ kF .e. Thus by studying the phase space available for the noncrossed and the crossed processes we have found that the crossed ones are suppressed by a factor Ωb /Ωa ≈ 1/(kF l). 11.63)
In conclusion: all cases where the scattering length l is greater than 1 ˚ we have by A the various Born approximations indeed resummed the perturbation series for Gk (ikn ) imp
taking all relevant diagrams into account and obtained Σk (ikn ) = −i sgn(kn ) . where no crossing of scattering lines occurs. Thus only one of the two internal momenta are free to be anywhere on the Fermi shell. i. Thus the crossing diagram (b) is suppressed relative to the non-crossing diagram (a) with a factor 1/kF l. For ﬁxed k2 the variation of k1 within its Fermi shell is restricted to the intersection between this shell and the Fermi shell of k + k2 − k1 . |Σ| ≈ /τ which relaxes |k1 |. This means that k1 and k2 are both conﬁned to a thin spherical shell of thickness 1/l and radius kF .g. doped semiconductors it is possible to obtain a degenerate electron
#(
¤ © ¦ ¤ © § ¤ © ¦ ¤ §¥ ¤ ¤ © ¦ ¤ §¥ ¤ £ ¦ ¤ ¥ ¤ £ % #$" ! 3 54 ! 3 2
¤ ¦¥ ¤ £
§ ¤ ¦¥ ¤ £
¢ ¤ © 1 0 ¨" # ( )
¤ © ¦ ¤ © § ¤ © ¦ ¤ © ¤ © ¦ ¤ §¥ ¤ ¨§ ¤ ¦¥ ¤ £ £ ¤ ¦¥ ¤ £ ¦ ¤ ¥ ¤ £ % '" ! !&% $" ! ! # #
¢ ¡
. In Fig.196
CHAPTER 11. 11. no further restrictions 2 applies. i. i.6(b).e. the inverse scattering F F F length. for l 1 ˚. so the volume of the available phase space is Ωa = (4πkF /l)2 . |k2 | = kF a bit.e. FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS
Figure 11.6(a).6(b). (b) The crossing wigwam diagram has the same restrictions for k1 and k2 as in (a) plus the constraint that |k+k2 −k1 | ≈ kF . the Feynman rules dictate that one further constraint. 2 So Ωb = (4πkF /l)(2πkF /l2 ). Since for metals 1/kF ∼ 1 ˚ we ﬁnd that A 1 kF l 1. In Fig. where k1 and k2 can take any value on the spherical shell of radius kF and thickness ∆k ≈ 1/l. namely |k + k1 − k2 | ≈ kF . one inside the other. where the scattering lines crosses. The phase 2 space is Ωa ∝ (4πkF ∆k)2 . The phase space 2 is now Ωb ∝ (4πkF ∆k)(2πkF ∆k 2 ). A (11. on a ring with radius ∼ kF and a cross section 1/l2 as indicated in Fig.

63) is violated.k |2 δ(ξk − ξk ). In these systems 1/kF or the Fermi wavelength is much larger than in metals. 15. 15 in the study of the residual resistivity of metals.
k
and the scattering-time broadened spectral function A1BA (ω ) = k
2 (ω − ξk )2 + 1/4τk
1/τk
. by taking higher order correlations into account. Im z < 0. Im z > 0 k + 2τ 1 1BA 1BA k −→ Gk (z) = Gk (ikn ) = 1 sgn(kn ) ikn →z ikn − ξk + i 2τ z−ξ − i . The main result was the determination of the self-energy Σk (ikn ) in terms of the scattering time τk . The weak localization eﬀect is studied in Sec. (11.
k k 2τ k
These results will be employed in Chap. We have applied the diagrammatic technique to an analysis of the single-particle Matsubara Green’s function for electron propagation in disordered metals. One example is the observation of weak localization. which was neglected in calculation presented in this chapter.11. For example is weak localization explained by treating crossed diagrams like the one in Fig.6
Summary and outlook
In this chapter we have introduced the Feynman diagrams for elastic impurity scattering. They can be explained within the theoretical framework presented here. 2τk with 1 ≡ 2π τk |tk. ΣFBA (ikn ) = −i sgn(kn ) k 1 . In this case one may therefore observe deviations from the simple theory presented here. 11. which is an increase in the resistivity due to quantum interference between scattering events involving several impurities at the same time. The theory presented here provides in combination with the Kubo formalism the foundation for a microscopic quantum theory of resistivity. 11.4) and universal conductance ﬂuctuations (see Fig. These more subtle quantum eﬀects are fundamental parts of the modern research ﬁeld known as mesoscopic physics.
The structure in the complex plane of the Green’s function was found to be: z−ξ 1 i . SUMMARY AND OUTLOOK
197
gas with a very low density.
.4.2). and the condition in Eq. 15.6(b). The technique can be extended to the study of quantum eﬀects like weak localization (see Sec.6.
11.

FEYNMAN DIAGRAMS AND EXTERNAL POTENTIALS
.198
CHAPTER 11.

2
The perturbation series for G inﬁnite perturbation series!Matsubara Green’s function
The ﬁeld operators in Eq. The total Hamiltonian H governing the dynamics of the system is as usual given by H = H 0 +W . τa ) . σa .3)
12.1)
while the interaction Hamiltonian W is given by W = 1 2σ dr1 dr2 Ψ† (σ1 . r1 ). this being the case for our coming main examples: electron-electron interactions mediated by Coulomb or by phonon interactions.16) we can immediately transform the expression for G into the 199
. τb . τa ) ≡ − Tτ Ψ(σb . The main goal of this chapter is to derive the Feynman rules for the diagrammatic expansion in orders of W of the full single-particle Matsubara Green’s function Eq. but using Eq. r1 )Ψ† (σ2 . r1 ) Ψ(σ2 .
(12. (10.3) deﬁning G are of course given in the Heisenberg picture. ra . rb . σ1 .σ2
We have specialized to the case where no spin ﬂip processes occur at the vertices. rb . r2 )Ψ(σ1 . r2 . The time-independent Hamiltonian H0 of the unperturbed or non-interacting system is H0 =
σ
dr Ψ† (r)H0 Ψ(r).1 12. (12.Chapter 12
Feynman diagrams and pair interactions
It is in the case of interacting particles and ﬁelds that the power of quantum ﬁeld theory and Feynman diagrams really comes into play.33a) G(σb . Below we develop the Feynman diagram technique for a system of fermions with pair interactions.2)
1 . r2 ) W (σ2 . (12. (10. ra . τb )Ψ† (σa . (12.

a) =
∞
(12.n ˆ ˆ Tτ Ψ† Ψ† 1 1 ˆ ˆ ˆn ˆ Ψ1 Ψ1 . .3). (12..1 . .
0
ˆ ˆ dτn Tτ W (τ1 ) .dndn W1.4)
The subscript 0 indicates that the averages in Eq. (10. To make sure of that we add an inﬁnitesimal time η = 0+ to the time-arguments of Ψ† (1) and Ψ† (2). a) = − n=0
(−1)n n!
∞ n=0
β 0
β
dτ1 . (12. dj . The expansion Eq.j Ψ(j )Ψ(j).8) ˆ ˆ ˆ ˆ ˆn ˆ ˆ ˆ ˆ ˆa d1d1 .4) are with respect to e−βH0 rather ˆ than e−βH as in Eq. W (τn )
0
ˆ Here we need to calculate τ -integrals of W (τ ).1 .
Wj.2) the two creation operators must always be to the left of the two annihilation operators. . . With the short-hand notation (σ1 . τ1 ) = (1) we obtain G(b. . Ψ† Ψ† n ˆ ˆ Ψn Ψn
−
n=0 ∞
(− 1 )n 2 n! (− 1 )n 2 n!
0
. . . . 1 . The τ -integrals of W (τ ) is therefore
β 0
1 ˆ dτj W (τj ) = 2
dj
ˆ ˆ ˆ ˆ dj Ψ† (j+ )Ψ† (j+ ) Wj. Ψ† Ψ† Ψn Ψn Ψb Ψ† 1 1 n d1d1 .. rj . (12.
β 0
0
ˆ ˆ ˆ ˆ dτn Tτ W (τ1 ) . 0)
0 0
. n . (12.. 1. . (12.n Tτ Ψ† Ψ† Ψ1 Ψ1 . (12.Wn. (12.
(12. a) = − Tr e−βH Tτ Ψ(b)Ψ† (a) Tr e−βH = − ˆ ˆ ˆ Tτ U (β. . W (τn )Ψ(b) Ψ† (a)
β
0
(−1)n n!
. (10.. which gives the right ordering when the time-ordering operator Tτ of ˆ Eq.
(12.12) for U is now inserted into Eq. we recognize that the average of the products of ﬁeld operators in the numer(2n+1) ator is the bare (2n+1)-particle Green’s function G0 (b.. ..
dj ≡
σj
dr
0
dτj .j as
β
j+ ≡ (σj . .5) for G: G(b..4) acts.7)
It is only in expressions where the initial and ﬁnal times coincide that the inﬁnitesimal ˆ shift in time of Ψ† plays a role. n ) times
. But one precaution must be taken regarding the ordering of the four operators in the basic two-particle interaction operator. In fact using Eq.4):
∞
G(b. (12.j ≡ W (rj . 1 . FEYNMAN DIAGRAMS AND PAIR INTERACTIONS
interaction picture. .6) for W into Eq.8) is that the average of the ﬁeld operators now involves bare propagation and thermal average both with respect to H0 . 0) Ψ(b) Ψ† (a) ˆ U (β.. r1 .Wn. and Wj. rj ) δ(τj −τj ). τj +η). According to Eq.6)
where we have deﬁned j+ .200
CHAPTER 12. . 1. a.65).
n=0
0
The great advantage of Eq.
(12.5)
dτ1 .dndn W1.
(12. Next insert Eq.

a)
The basic Feynman rules for n’th order term in the denominator of G(b. 1. n ) G G 0 (1.. However. 1) G 0 (n .3.12.j . n ) G
∞ n=0
(− 1 )n 2 n!
d1d1 . We have suppressed. . thus cancels the sign in Eq.3
The Feynman rules for pair interactions
We formulate ﬁrst a number of basic Feynman rules that are derived directly from Eq.. This in turn leads to the formulation of the ﬁnal Feynman rules to be used in all later calculations. n ) G 0 (1 . . .7) is to be shifted inﬁnitesimally to τj + η.6) and (12.n
∞ n=0
(− 1 )n 2 n!
d1d1 . a) = G 0 (b. 1 ) . a) G 0 (1. (12. 1 ) . . −1. it turns out that using these basic rules leads to a proof that the denominator cancels out. . . 1) G 0 (b. 1 ) . 0 (n . j): G(b. 1 . 1 ) . .1 . Now is the time for our main use of Wick’s theorem Eq. a) are (1) (2) (3) (4) (5) Fermion lines: j2 j1 ≡ G 0 (j2 . the fact that the initial time τj in G 0 (l.3. . (12. . n ) . τ1 → τ1 + η..e.. . For each set connect the 2n vertices with 2n fermion lines: one entering and one leaving each vertex.1 . . .dndn W1. G 0 (1. a) G 0 (n . This can be done in (2n)! ways. . a) G 0 (b. G 0 (n . . while in the denominator it is the bare (2n)-particle Green’s function (2n) G0 (1. (10. G 0 (b.1
Feynman rules for the denominator of G(b.10a)
. 1 ) . . .
12. .79): the bare many-particle Green’s functions in the expression for the full single-particle Green’s function are written in terms of determinants containing the bare single-particle Green’s functions G 0 (l. a) G 0 (1 .dndn W1. j1 ). but not forgotten. n ) . .9). . G 0 (n .. 1 . 0 (n . G 0 (1. i. no spin ﬂip. (12. Interaction lines: j j ≡ Wj. (12. j) according to Eqs. 1 ) G 0 (1 ..Wn. .9) . 1) G 0 (1. n . 1) G 0 (1 . THE FEYNMAN RULES FOR PAIR INTERACTIONS
201
(−1)2n+1 = −1. G 0 (1 . j j Draw (2n)! sets of n interaction lines j j. n ) times (−1)2n = 1. G 0 (1 . n ) . .Wn. The resulting sign. sum over internal variables.. (12... 1) G 0 (1. . .
12. . . Vertices: j• ≡ dj δσin .n
This voluminous formula is the starting point for deﬁning the Feynman rules for the diagrammatic expansion of G in terms of the pair interaction W . n ) . 1) G 0 (n . . . 1) G 0 (1 .8). .. . 1 ) G 0 (1.σout .

202
CHAPTER 12..9).
720 terms
12.3..
(12. only Feynman rules (4) and (5) given for the denominator have to be changed to give the rules for the numerator:
. FEYNMAN DIAGRAMS AND PAIR INTERACTIONS
But this is not all.
24 terms
+ . a)
ˆ ˆ ˆ The numerator Tτ [U (β. but indicating the number of diagrams of each order.. and add the resulting (2n)! diagrams of order n. i.2
Feynman rules for the numerator of G(b.G 0 (j2 . 0) 0 of G(b. This raises the dimension of the n’th order determinant from 2n to 2n+1. j1 j2 . j2 )..] = −sgn[.. The product of the diagonal terms in the determinant is per deﬁnition positive and in diagram form it consist of n factors j j .. a) diﬀers from the denominator by the ˆ ˆ presence of the two external ﬁeld operators Ψ(b) and Ψ† (a) that act at the external spacetime points (b) and (a).g.11)
+
+
+ . . F = 2n is even. and at the same time it changes the number of fermion loops by 1.. because what about the sign arising from the expansion of the determinant? Here the concept of fermion loops enters the game.. . +
+ .. a) in Eq.. where F is the number of fermion loops in the coming from the determinant is (−1) given diagram....
+
2 terms
(12. j1 ).e. j1 . A fermion loop is an uninterrupted sequence of fermion lines starting at some vertex j and ending there again j3 j2 .10b) For all n there are (2n)! terms or diagrams of order n in the expansion of the deterˆ minant in the denominator U (β. (12. 0)Ψ(b)Ψ† (a)] 0 of G(b..G 0 (j1 . Suppressing the labels. j1 ). All other diagrams can be constructed one by one simply by pair wise interchange of the endpoints of fermion lines. This changes the determinantal sign of the product since sgn[. An outline of the proof is as follows. F being the number of fermion loops. The overall sign after connecting to other vertices. +
+ .G 0 (j2 .]. Consequently. e. this expansion takes the following form using Feynman diagrams:
ˆ U (β.. becomes . 0)
0
=1 +
+ +
+ .G 0 (j1 . +
+ . e. or j1 F .. j2 ).g. Thus we obtain the last Feynman rule (6)
1 1 Multiply by n! (− 2 )n (−1)F .

. + + . We furthermore note that the parts of the diagrams in Eq. 0)Ψ(b)Ψ† (a)] b + b b
! " #
b b b b + + + + + a a a a a a a b b b b + + + . the so-called disconnected diagrams (e. For each set connect the 2n+2 vertices with 2n+1 fermion lines: one leaving a..12.13) two classes of diagrams appear: those being connected into one piece with the external vertices a and b. the ﬁrst second order diagram). We also note that a diagram containing two or more disconnected parts can be written as a product containing one factor for each disconnected part.13) disconnected from the external vertices are the same as the diagrams appearing in Eq.. A detailed combinatorial analysis (given at the end of this section) reveals that the denominator in G cancels exactly the disconnected parts of
.3
The cancellation of disconnected Feynman diagrams
It looks like we are drowning in diagrams.. (12. and one entering and leaving each internal vertex j. + a a a a + .12)
203
Using these rules we obtain the diagrammatic expansion of the numerator:
ˆ ˆ ˆ Tτ [U (β.g. (12.13)
6 terms
+ . and those consisting of two or more pieces. THE FEYNMAN RULES FOR PAIR INTERACTIONS (4’) (5’)
Draw (2n+1)! sets of n lines j j and 2 external vertices •a and •b. but in fact there is a major reduction at hand. (12..3.3.. the last second-order diagram). We note that in Eq.
0
=
(12.11) order by order..
120 terms
12. the so-called connected diagrams (e.g. one entering b. (12..

1 +
&
+ .15) Pay attention to the fact that only the topology of the diagrams are mentioned. (1) Since all internal vertices have one incoming and one outgoing fermion line. Thus they can at will be stretched. We conclude that all 2n n! diagrams with the same topology relative to the external points give the same value.204
CHAPTER 12.e. i. Multiply each diagram by (−1)F . 2 vertices •a and •b. F being the number of fermion loops.
' ()*+. The proof goes through eight steps. No notion of a time-axis is implied in the imaginary time version of the Feynman diagrams. so that one leaves •a. mirror inverted and otherwise deformed.9). Sum over all the topologically diﬀerent diagrams. one enters •b. j1 ).. so is j due to the interaction line Wj. FEYNMAN DIAGRAMS AND PAIR INTERACTIONS
the diagrams in the numerator leaving only the connected ones: b b
G(b. sum over internal variables. and one enters and leaves each internal vertex •j. a is connected with r W -lines.e. For completeness we give the following proof of the cancellation of the disconnected diagrams..j Vertices: j• ≡ dj δσin . connected diagrams containing n interaction lines j j . the external vertices a and b are always connected. In conclusion.j . (3) In a diagram of order n. (12.. but the reader may skip it since the essential conclusion has already been given above. Interaction lines: j j ≡ −Wj. and 2n+1 fermion lines. a a 1 + + . (12. and for each line there are 2 ways to put a given pair of labels j and j . Except for the sign this 1 factor cancels the prefactor n! (− 1 )n .
+ . . no spin ﬂip j j At order n draw all topologically diﬀerent. b b b b b + + + + a a a a a
$%
+
+ .. . i. We study the numerator of Eq. we are left with a factor of (−1) for each of the 2 n interaction lines..σout . (2) If vertex j somehow is connected to a. a) =
=
Being left with only the connected diagrams we ﬁnd that since now all lines in the diagram are connected in a speciﬁc way to the external points a and b the combinatorics of the permutations of the internal vertex indices is particularly simple.. for pair interactions the ﬁnal version of the Feynman rules for expanding G diagrammatically is: (1) (2) (3) (4) Fermion lines: j2 j1 ≡ G 0 (j2 . (12. There are n! ways to choose the enumeration j of the n interaction lines j j . τ1 → τ1 + η. / 0
.14)
connected
(5) (6)
.

14) deﬁne the concept of irreducible diagrams in G(b. j) as
¡ ¢
¡ ¢ ¢
¡
¡ ¢
(12..4
Self-energy and Dyson’s equation
In complete analogy with Fig. r ]con r! 2
1 −1 m! 2
m
I[r+1. . (12. (a) and (b). SELF-ENERGY AND DYSON’S EQUATION
© ¥ £ § ¥ ¨ £ ¥ ¦ £ ¥ ¤ £
205
Figure 12. where 0 ≤ r ≤ n. . .j . r )] Det G 0
r=0
12. (6) The structure of the sum is now:
∞ n=0
1 n!
∞
−1 2
n
I[1.1.. (5) The r pairs of vertex variables j and j connected to a can be chosen out of the n! available n pairs in r!(n−r)! ways. (r+1) . 11. As depicted in Fig.. we can now based on Eq. i. j ) yield the same result. n..1: Examples of irreducible. a) in the case of pair interactions. (c) and (d). (12.12. 12. n ]
n n r=0 r
=
n=0 ∞
1 −1 n! 2
n! I[1. (r+1) . ˆ (8) The disconnected part is seen to be U (β. connected topological diﬀ.3 for impurity scattering. if as usual Wj. . each choice yielding the same value of the total integral. and so does all the 2n ways of ordering each pair.... such diagrams are the ones that cannot be cut into two pieces by cutting a single fermion line. (r+m) ]discon . r.17)
(2r+1)×(2r+1)
..4. 0) Ψ(b) Ψ† (a)
0
ˆ = U (β. 0) 0 . We thus reach the conclusion
∞
ˆ ˆ ˆ Tτ U (β.
(7) In the connected part all r! permutations of the vertex variable pairs (j.. r ]con I[r+1. The number of disconnected W -lines is denoted m. 1 . 1 . and reducible.[−W (r. r.16)
. Feynman diagrams in the expansion of G(b.j = Wj . one over the 2r variables connected to a and one over the 2m variables disconnected from a. m = n−r. 1 )]. 0)
0
[−W (1. (r+m). n ]discon r!(n − r)!
∞ m=0
=
r=0
1 −1 I[1.e.. n.. (4) In all terms of the expanded numerator the integral factorizes into a product of two integrals. 1 . a) in the presence of pair-interactions. Continuing the analogy with the impurity scattering case we can also deﬁne the self-energy Σ(l. ..

Furthermore. l)
(12. ikn )-space as
.19)
Note how Dyson’s equation in this case is an integral equation.
. We shall shortly see that for a translation-invariant system it becomes an algebraic equation in k-space. (10. In terms of the Fourier transform W (q) = 4πe2 /q 2 the Coulomb 0 interaction W (rτ ..7) the Matsubara frequency iqn appears only in the argument of the exponential function. a). τ )-space to (q. a) without the two external fermion lines G 0 (j.e. < = > ?
δl.iqn
(12. 2. FEYNMAN DIAGRAMS AND PAIR INTERACTIONS
Σ(l.18)
a
+ . j j l j + + + b b b l l l j j j a + b l j × j a + j a a + dl dj G 0 (b.
The sum of all irreducible diagrams in G(b.39) we can express the electronic Green’s function Gσ (rτ.. r .20)
It is important to realize that the Matsubara frequency iqn is bosonic since the Coulomb interaction is bosonic in nature: two fermions are annihilated and two fermions are created by the interaction. i.j + l j + l + l + .2. τ ) is written W (rτ . using Eq. a) G(b.
q.14) and (12.18) we obtain Dyson’s equation for G(b. r τ ) for spin σ in (k. r τ ) = 1 βV W (q) e[iq·(r−r )−iqn (τ −τ )] . 0 Likewise.. we note that due to the factor δ(τ − τ ) in Eq. l) Σ(l.206
CHAPTER 12. a) = = = = b b b b a a a a
= G 0 (b. iqn )-space. a)
1234 5 6 8 9 7 : . Our main example of such a system is the jellium model for Coulomb interacting electrons studied in Sec.. (12..
+
(12.. j) G(j. j) ≡
= =
From Eqs.
12. a) and G 0 (b. one boson object is annihilated and one is created.j only depends on the coordinate diﬀerences rj − rj and τj − τj it is a great advantage to Fourier transform the representation from (r.5
The Feynman rules in Fourier space
For the special case where H0 describes a translation-invariant system and where the interaction Wj. (12.

to even values for the Matsubara frequencies of the interaction lines. In the case of the Coulomb interacting electron gas in the jellium model we thus see 0 that both the Green’s function Gσ and the interaction W depend only on the space and imaginary time diﬀerences r − r and τ − τ . τ )-space points r1 . For each independent momentum a factor 1/βV remains from the corresponding Fourier transform.12. and ˜ ˜ · r ≡ ik · r − ik τ . (12. k. r τ ) =
1 βV
0 Gσ (k. (12. r ≡ (r. r2 . p from r to r2 . and whence the number of independent internal momenta equals n. r3 .21) into Eq. ikn ) is conserved at each ˜ Coulomb scattering vertex: k = p + q . Since each vertex consists of two fermion lines ˜ ˜ and one interaction line. It follows from Eqs.e. ˜ 0 ˜ q ˜˜ kq
(12.23)
From this follows that in Fourier space the four-momentum (k. We therefore end up with the following Feynman rules for the n-order diagrams in the expan-
. ikn ) = 1/(ikn − ξk ) depends on k and ikn . Using this notation we analyze the Fourier transform of the basic ik ˜ n Coulomb scattering vertex r2 ˜ pσ ˜ q ˜ (12. r) Gσ (˜. ˜
r1 ˜
˜ where the (r. τ ). and q from r to r3 . (12. the momentum conservation combined with the odd values of the fermion Matsubara frequencies leads. THE FEYNMAN RULES IN FOURIER SPACE
207
0 Gσ (rτ .5. τ )-space is not changed by the Fourier transform. The topology of the diagram in (r. the order of the diagram. i. On top of their usual meaning the arrows ˜ ˜ ˜ now also indicate the choice of sign for the four-momentum vectors: k ﬂows from r1 to ˜ r.ikn
(12. and r are indicated as well as the wave vectors k.20) and (12. but not on σ. and q to be used in the Fourier transform. r) = r 0 r ˜ 0 r ˜ r ˜ r ˜ r3 . r) r 0 r ˜ 0 r ˜ r ˜ = = = d˜ r 1 (βV)3
0 p r r r r q r r Gσ (˜) Gσ (k) W (˜) ei[˜·(˜2 −˜)+k·(˜−˜1 )+˜·(˜3 −˜)] p 0 ˜ q ˜ ˜˜ kpq p q r d˜ e−i(˜−k+˜)·˜ r ˜ ˜
@
˜ kσ
1 (βV)3 1 (βV)2
0 pr ˜r qr Gσ (˜) Gσ (k) W (˜) ei[˜·˜2 −k·˜1 +˜·˜3 ] p 0 ˜ q ˜ ˜˜ kpq
˜ r r 0 ˜ q r r Gσ (k− q ) Gσ (k) W (˜) ei[k·(˜2 −˜1 )+˜·(˜3 −˜2 )] .22) d˜ Gσ (˜2 .21)
0 where Gσ (k. ikn ). ikn ) e[ik·(r−r )−ikn (τ −τ )] .22) yields with this sign convention d˜ Gσ (˜2 . Inserting the Fourier transforms of Eqs.20) ˜ ˜ ˜ ˜ ˜ ˜ ˜ and (12. r) Gσ (˜. Here ξk ≡ ε − µ. r1 ) W (˜3 . ˜ ˜ ˜ ˜ p. in agreement with our previous remarks. r1 ) W (˜3 . The momentum conservation rule for each of the 2n vertices also leads to 2n delta function constraints on the 2n internal fermion momenta and the n interaction line momenta.21) that ˜ it saves some writing to introduce the four-vector notation k ≡ (k.

41) the self-energy Σσ (k. ikn ) 0 1 − Gσ (k. The factor eikn η follows directly from the Fourier transform when this shift is included.6) and (12. k). ˜ ˜ must also exit it.
(12. ikn )-space the fourth Feynman rule concerning the conservation of four-momentum at the scattering vertices simpliﬁes many calculations. i. k ) = δk. iqn Conserve the spin and four-momentum at each vertex.
0 ˜ 0 ˜ ˜ ˜ ˜ Gσ (k) = Gσ (k) + Gσ (k) Σσ (k) Gσ (k)
with Gσ (k. and they allow one to identify the important
. ikn ) is to be interpreted as the externally given four-vector momentum. q. (12. ikn ). ikn )
E F G
= + =
. ikn )
As in Eq. perform the sum ˜
CD
and
by eikn η .
pσ ˜
. induced here by the Coulomb interaction W .k Σσ (k). ˜˜ ˜ ˜ ˜ Σσ (k) ≡ Σσ (k. ikn ) in the ’same-time’ diagrams
Multiply by
1 βV
for each internal four-momentum p.e. ikn ). kσ.24) Note how the two ’same-time’ diagrams in rule (6) are the only ones where it is relevant to take explicitly into account the inﬁnitesimal shift τj → τj + η mentioned in Eqs.6
Examples of how to evaluate Feynman diagrams
The Feynman diagrams is an extremely useful tool to gain an overview of the very complicated inﬁnite-order perturbation calculation. ikn ). ikn − ξk − Σσ (k. kl = kj . ikn ) = the solution
0 Gσ (k. The self-energy (with spin σ) is thus diagonal in k-space.19) is therefore an algebraic equation.26) (12. ikn ). (12.27)
1 . two external fermion lines Gσ (k. (11.25)
Dyson’s equation Eq. Most noteworthy is the fact that Dyson’s equation becomes an algebraic equation. ip ). and 2n q 0 (p .208
CHAPTER 12. (12. ikn ) Σσ (k. and no spin ﬂipping. FEYNMAN DIAGRAMS AND PAIR INTERACTIONS
sion of Gσ (k.
12.
(12. such as the ones shown in Eq. where (k.7).e. Due to four-momentum conservation a ˜ four-momentum kj entering a self-energy diagram. appears as a direct additive renormalization of the bare energy ξk = εk − µ. Multiply each diagram by (−1)F . In (k. i. At order n draw all topologically diﬀerent connected diagrams containing n 0 oriented interaction lines W (˜). All vertices must contain an incoming internal fermion lines Gσ j j and an outgoing fermion line as well as an interaction line. (1) (2) (3) (4)
0 ≡ Gσ (k. (12. ikn Interaction lines with four-momentum orientation: ≡ −W (q).
Fermion lines with four-momentum orientation:
A B
(5) (6) (7)
0 Multiply Gσ (k. incoming momenta must equal the outgoing. F being the number of fermion loops. ˜ ˜ ˜ Σσ (k.18).

The Feynman rules therefore lead to the following expression for the Hartree self-energy diagram Eq.6. ikn ) ≡ σ
H
−1 βV
σ
0 Gσ (k.18) strip oﬀ the two external fermion lines to obtain the self-energy ΣH :
H Gσ (k. The self-energy diagram is a ﬁrst order diagram. yielding a factor of 1/βV. ipn . (12.24).12.28)
J
=
0 − W (0) Gσ (p. The Matsubara sum can easily be carried out using the method of Sec. In other words we have shown that the tadpole-shaped self-energy diagram is the diagrammatic equivalent of the Hartree mean ﬁeld approximation.28): ΣH (k. while the four-momentum (p. one internal spin. It contains one internal four-momentum. σ
(12. (12. EXAMPLES OF HOW TO EVALUATE FEYNMAN DIAGRAMS
209
processes for a given physical problem. 10. (p. The evaluation of the p-integral is elementary and yields N/2. rule (3) in H Eq.2
The Fock self-energy diagram
F We treat the Fock diagram Gσ and Fock self-energy ΣF similarly: σ
. (12. (12. This renormalization we have calculated by completely diﬀerent means in Sec.6.24). 4. We start with the so-called Hartree diagram Gσ (which is zero in the presence of a charge compensating back-ground).29) exactly equals the Hartree part of the mean ﬁeld energy in Eq.27) the self-energy is the interaction-induced renormalization of the non-interacting single-particle energy. ipn ). (12. We see that the diagrammatic result Eq.1. where we in accordance with Eq.29)
= 2W (0)
dp n (ξ ) (2π)3 F p
Note the need for Feynman rule Eq. F = 1. (12. σ .24)(6) for evaluating this speciﬁc diagram. and one fermion loop. i. V (12.25b). ipn ) and the spin σ in the fermion loop are free to take any value. ikn )
=
0 Gσ (k.
12.2 using the Hartree-Fock mean ﬁeld approximation. ikn ) ≡
The four-momentum transfer along the interaction line is zero.6.e. (12. ikn )
I
0
p. (4. i. According to Eq.4. When this part of the analysis is done one is (hopefully) left with only a few important diagrams that then need to be evaluated. n = 1. ipn ) eipn η p ipn
=
2W (0) β
dp (2π)3
ipn
eipn η ipn − ξp = W (0) N .
12. The spin sum turns into a simple factor 2.1
The Hartree self-energy diagram
To evaluate a given diagram the ﬁrst task is to label the fermion and interaction lines with four-momenta and spin obeying the conservation rules at each vertex.e. We end this chapter by studying the explicit evaluation of three simple Feynman diagrams in Fourier space using the Feynman rules Eq.

g. ikn ) ≡
q. i.e. σ 0 Once more the external fermion lines are written explicitly as two factors Gσ (k.
12. (2π)3
Note that also for this speciﬁc diagram we have used Feynman rule (6). plays a central role in studies of the electron gas. We proceed as in the previous examples: 0 Gσ (k. The four-momentum transferred by σ the interaction line is (k − p. which. We immediately note that this diagram is of second order.28) the internal spin σ is now forced to be equal to the external spin σ. The Matsubara sum can easily be carried out using the method of Sec. as we shall see in Chap. This diagram is a ﬁrst order diagram.11)
.e.6. However. σ ipn q. i. (12. ikn ) leaves us with the pair-bubble selfP . F = 1. FEYNMAN DIAGRAMS AND PAIR INTERACTIONS
F Gσ (k. iqn
0 Removing the two external fermion lines Gσ (k. The spin sum turned into a simple factor 1. no fermion loops are present. i.1.4. Finally.25b) calculated using the Hartree-Fock mean ﬁeld approximation.32)
A full oyster diagram can be seen in e. (4. (12. We have thus shown that the half-oyster self-energy diagram1 is the diagrammatic equivalent of the Fock mean ﬁeld approximation.3
The pair-bubble self-energy diagram
P Our last example is the pair-bubble diagram Gσ . ikn ) ikn −ipn σ
(12.31)
=−
dp W (k−p) nF (ξp ). n = 1.30)
M
=
1 βV −1 β
0 − W (k−p) δσ. 13. At the ﬁrst vertex the incoming momentum
1
N
=
k−q. The evaluation of the p-integral is in principle elementary. ikn ). 10. ikn − ipn ). It contains one internal four-momentum. Eq. yielding a factor 1/βV.e. (p. leaving the Fock self-energy ΣF to be determined. ikn )
k−p G 0 (k. n = 2.30): ΣF (k.e. ikn )
O
p. F = 0. σ ikn −iqn
0 Gσ (k.210
CHAPTER 12. ikn ) P Gσ (k. ipn ). The Feynman rules therefore lead to the following expression for the Fock self-energy diagram Eq. in contrast to Eq. We see that this self-energy diagram exactly equals the Fock part of the energy in Eq. iqn
p+q. i. σ ipn +iqn
(12.σ Gσ (p. (12. ikn ) =
p. energy diagram Σσ containing one fermion loop. ipn ) eipn η σ p ipn
=
dp W (k−p) (2π)3
ipn
eipn η ipn − ξp (12. ikn ) ≡ σ
K L
=
0 Gσ (k. ipn .

ikn − iqn ). Inserting the result for Π0 (q. Π0 (q. iqn ) Gσ (k−q. ikn ). and σ .12. Using the Feynman diagrammatic
. ikn −iqn )
=
dq 0 W (q)2 Π0 (q. ikn−iqn ) continues in the fermion line.34)
12. iqn ) is sent out through the interaction line. (12. The Feynman rules lead to the following expression for the pair-bubble self-energy Eq. iqn ) ≡
The loop contribution Π0 (q. iqn ) is traditionally denoted the pair-bubble. a) or Gσ (k. (12.33)
where we have separated out the contribution Π0 (q. and that the Matsubara sum over ipn can easily be carried out using the method of Sec. ipn ) and continues in a new fermion line as (p+q. iqn ) from the fermion loop.7. the former two yielding a prefactor 1/(βV)2 . We have thereby ensured that the exit momentum equals that of the entrance: (k. iqn ). However. SUMMARY AND OUTLOOK
211
(k. and we shall return to it later. 3 (ip + iq − ξ (2π) n n p+q ) (ipn − ξp )
(12. sending (q.1. while the remainder (k−q. (p. The internal degrees of freedom are (q. At the fermion loop. The Feynman diagram technique is a very powerful tool to use in the context of perturbation theory. The evaluation of the p-integral is in principle elementary.4.7
Summary and outlook
In this chapter we have established the Feynman rules for writing down the Feynman diagrams constituting the inﬁnite-order perturbation expansion of the full single-particle Green’s functions G (b. ikn ) is split. ipn ).
Q
=
−2 β
ipn
dp 1 1 . iqn ) into the pair-bubble self-energy diagram Eq. iqn ) is joined by the internal fermion momentum (p. and we shall study it in more detail in the coming chapters. ikn ) in terms of the pair-interaction W . the calculation can be performed. 10. where it ultimately recombines with the former fermion momentum (k − q. It enables a systematic analysis of the inﬁnitely many terms that need to be taken into account in a given calculation. ipn +iqn ) Gσ (k−q. ikn ) σ ˜ ≡ p
=
−1 (βV)2 1 β
P
q ˜ q ˜
σ pq ipn iqn iqn
− W (q)
2
0 0 0 Gσ (p.32): ΣP (k. iqn ) out through the interaction line. ipn ) Gσ (p+q. (q. (2π)3
(12. At the top of the loop the momentum (q. ipn+iqn ). Our main example is the Coulomb interaction. Here we just note that the spin sum becomes a factor 2. ikn −iqn ).32) leads to a bit more involved Matsubara frequency summation over iqn and momentum integration over q.

as we shall see in the following chapters.34) are the ones that dominate the physics of the interacting electron gas in the high density limit. identify which sub-classes of diagrams that give the most important contributions. We have already given explicit examples of how to evaluate some of the diagrams that are going to play an important role. Indeed. 13 that the diagrams analyzed in Eqs. we show in Chap.
.212
CHAPTER 12. FEYNMAN DIAGRAMS AND PAIR INTERACTIONS
analysis one can. (12.31) and (12. We shall learn how these diagrams determine the ground state energy of the system as well as its dielectric properties such as static and dynamic screening.

2)
The main result of the following diagrammatic calculation is that the dynamics of the interacting system by itself creates a renormalization of the pure Coulomb interaction into a Yukawa-like potential independent of the value of α. The starting point of the theory is the self-energy Σσ (k.2 revealed a divergence in (2) the contribution Edir from the direct processes. see Eq. the second order perturbation analysis of Sec. (2.49). (12. which then without problems can be taken to zero.Chapter 13
The interacting electron gas
In Sec. In analogy with Eq.
13. (2. see Eq. (2. with the Yukawa potential we (2) can obtain a ﬁnite value for Edir in Eq. 2.2 we studied the Coulomb interaction as a perturbation to the non-interacting electron gas in the jellium model. At the end at the calculation we take the limit α → 0 to recover the Coulomb interaction.
(13. where according to Eq.24). ikn ). In this chapter we reanalyze the Coulomb-interacting electron gas in the jellium model using the Feynman diagram technique. and we show how a meaningful ﬁnite ground state energy can be found.35) the interaction energy is negligible. (1. but the divergence reappears as soon as we take the limit α → 0.2.1
The self-energy in the random phase approximation
To construct the diagrammatic expansion of the self-energy Σσ (k. (12. For example. W (r − r ) = |r − r | 4πe2 W (q) = 2 0 2 . ikn ) removing the two external fermion 213
.1)
The range 1/α has no physical origin. q +α (13.49) if α is ﬁnite. To ensure well-behaved ﬁnite integrals during our analysis we work with the Yukawa-potential with an artiﬁcial range 1/α instead of the pure long range Coulomb potential. Edir
(2)
∝
0
dq q 2
1 1 qq (q 2 + α2 )2 q
∼
− ln(α)
α→0
−→
∞. ikn ) in (k. This was expected to be a valid procedure in the high density limit. 2.18) the self-energy is given by the sum of all the irreducible diagrams in Gσ (k. Nevertheless. ikn )-space we use the Feynman rules Eq.103) and the associated footnote. e2 0 e−α|r−r | .

ε ∝ kF .
13. W ()
G 0 () ..and thus rs -dependence:
n n 2n−1
ikn −εk
F
Σ(n) (k. (13. G 0 () .6) are the precise statements for how to identify the most important self-energy diagrams in the high density limit. ikn ) . (2. we have n<n ⇒
(n) Σσ (k. .
(13. ikn ) = σ
∝
˜ dk1 . and then resum the inﬁnite series taking only these terms into account. . ikn ). for rs → 0..6)
Eqs.5) and (13. ikn ) = = 0.214
CHAPTER 13. and β ∝ kF . .37).37) and its degree of divergence in the cut-oﬀ parameter α. . Thus: σ
Σσ (k.
˜ dkn
W () .4)
2n − 1 Green’s fcts
n internal momenta
We then make the integral dimensionless by measuring momenta and frequencies in powers of the Fermi momentum kF and pulling out the corresponding factors of kF . ikn ) of order n:
n interaction terms
Σ(n) (k. The self-energy diagram therefore has W (q) ∝ 2 1 2 ∝ k −2 and G 0 (k) =
q +α F σ
the following kF .1
The density dependence of self-energy diagrams
(n)
Consider an arbitrary self-energy diagram Σσ (k.
. Furthermore. σ
(13. while 1 ˜ ∝ k −2 . . We recall that due to the charge compensating back ground in the jellium model the Hartree self-energy diagram vanishes. ikn )
1
Σ(n ) (k. We have 2+3 2 2 5 1 1 dk ˜ k ∝ kF . ΣH (k. We can conclude that for two diﬀerent orders n and n in the high density limit. . ikn ) ∝ σ
5 kF
−2 kF
−2 kF
=
kF
−(n−2)
∝
n−2 rs .1. THE INTERACTING ELECTRON GAS
0 lines Gσ (k.
(13. This is achieved by noting that each diagram in the expansion is characterized by its density dependence through the dimensionless electron distance parameter rs of Eq. rs → 0. dk1 ∝ β ikn (2π)3 ∝ kF = kF . (13. (2. ikn ) =
+
+
+
+
+ .3)
For each order of W we want to identify the most important terms.5)
where in the last proportionality we have used rs = (9π/4) 3 /(a0 kF ) from Eq.

1) δσ =n
⇒
Σ(n.6) and (13.1.8) the most important terms are those in the diagonal in this table where δ = n. For example (taking α = 0) two lines with the momentum q contributes with W (q)2 which diverges as q −4 for q → 0 independent of the behavior of any other internal momentum p in the diagram.
(13. i.1) (k. ikn ) σ
Σ(n.2)
(n.1)
(n. in a set of measure zero in the integral over q and p. ikn ) .1) in the limit of small q and α leads to a divergent behavior of the self-energy integrals. when all n momenta in the diagram are the same. The ﬁrst few diagrams (without arrows on the
.
(13.e. ikn ) as
(n)
of the self-
(n) δσ ≡
the largest number of interaction lines in (n) Σσ (k. δ)-table. According to Eqs. we now order the self-energy diagrams in a (n.8)
13. In view of this discussion it is natural to deﬁne a divergence number δσ (n) energy diagram Σσ (k. With one notable exception.e.7)
Consider two diagrams Σσ and Σσ of the same order n.1. The exception involves the diagram with the maximal divergence number. In the limit α → 0 this diagram is the largest:
(n. (13. which diverges as q −4 only when both q → 0 and p → 0 at the same time.3
RPA resummation of the self-energy
Using the order n and the divergence number δ. i. In contrast.1. (13.2
The divergence number of self-energy diagrams
The singular nature of the Yukawa-modiﬁed Coulomb potential Eq. The more interaction lines carrying the same momentum there are in a given diagram. it is in general not possible to determine which diagram is the larger based alone on knowledge of the divergence number. ikn ) having the same momentum q. the more divergent is this diagram. THE SELF-ENERGY IN THE RANDOM PHASE APPROXIMATION
215
13. two lines with diﬀerent momenta q and q − p contributes with W (q)W (q − p).2) .2) (k.13. σ
for α → 0 and any n-order diagram Σ(n.

To make the fermion-loop sign from the pair-bubble appear explicitly we prefer to work with χ0 ≡ −Π0 ..
dp 1 1 .e. 3 (ip + iq − ξ (2π) n n p+q ) (ipn − ξp )
.9)
χ0 (q.6. For each given order the diagrams with the highest divergence number are the most important. iqn ) = 2 β
ipn
(13.216
CHAPTER 13. The self-energy in the random phase approximation (RPA) is an inﬁnite sum containing diagrams of all orders n.7.10) q ˜ q ˜ q ˜ Below we are going to analyze parts of the diagrams individually.3. An important part of the self-energy diagrams in Eq. this χ0 is the same correlation function as the one introduced for other reasons in Sec. i.
−
(13. 12.10) is clearly the pair-bubble Π0 (q. since the Feynman rules Eq. THE INTERACTING ELECTRON GAS
interaction lines for graphical clarity) are ˜ Σσ (k) δ=1 n=1 n=2 n=3 n=4
δ=2
−
δ=3
−
−
δ=4
−
−
It is clear that the most important diagrams in the high density limit are those having a low order.11) In fact. iqn ). ΣRPA (k) ≡ q ˜ σ q ˜ p ˜ q ˜ (13. (13. This is straightforward to do. because it ensures that all interaction lines W (q) carry the same momentum q..24) are still valid for each part. It plays a crucial role. 10. iqn ) ≡ already introduced in Sec. (12. but only the most divergent one for each n: q ˜ q ˜ q ˜ q ˜ ˜ ˜ ˜ ˜ k− q ˜ ˜ ˜ ˜ ˜ ˜ k− q k− q p + q k− q ˜ + + + + .
≡ −χ0 (q.

13.13) can be found using a Dyson equation approach. q 2 − 4πe2 χ0 (q. . . iqn ) −→ 4πe2 0 . = . We can now insert the speciﬁc form Eq. omitting
(13.
In (q.12) and (13. iqn ) introduced in Eqs. iqn ) 0 (13.12)
(13.14)
(13.. (13. pulling out the convergent Fock self-energy ×
ΣRPA = σ
In the following we study the properties of the renormalized Coulomb interaction W RPA (˜).16)
+. and let the artiﬁcial cut-oﬀ parameter α tend to zero.2. rewrite the RPA self-energy as ×
ΣRPA = σ
or. . (13. iqn )
α→0
(13. The ﬁnal result is W RPA (q.. .2
The renormalized Coulomb interaction in RPA
The renormalized Coulomb interaction W RPA (q. as + + + + . q
! " # $% &' ( )*
+ + + + .15)
. + . iqn ) ≡ + .1) for the Yukawa-modiﬁed Coulomb interaction.13)
13./ 01 234 567 8 9 :
= + × 1− −W (q) . THE RENORMALIZED COULOMB INTERACTION IN RPA By introducing a renormalized interaction line −W RPA (˜) = q interaction line arrows. −W RPA (q. iqn ) = = =
Note the cancellation of the explicit signs from W and χ0 in the denominator. 1 − W (q) χ0 (q. iqn )-space this is an algebraic equation with the solution −W RPA (q. . = + ≡ + + + = + × + +
217 we can.-. .

We still have to perform the rather involved p-integral.20)
The frequency dependence of the retarded pair-bubble χR can now be found by the 0 usual analytical continuation iqn → ω + iη. χ0 (0. 0). and it is put to zero in the following.
for kB T
εF . ks has been introduced. In the static. long-wave limit at low temperatures χR (q. long-wave limit. long-wave limit q → 0 and iqn = 0. 0) −→ 2 q→0 q 2 + ks where the so-called Thomas-Fermi screening wavenumber. ﬁnd the value of the Thomas-Fermi screening wavenumber ks . THE INTERACTING ELECTRON GAS
W RPA (q. we ﬁnd that W RPA appears in a form identical with the Yukawa-modiﬁed Coulomb interaction.54): χ0 (q. The sum was carried out in Eq. (13. 0). (2π)3 ξp+q − ξp − iqn (13. (12. iqn ). according to Eq. 0) (13.19)
In the following section we calculate the pair-bubble χ0 (q. W RPA (q → 0. and we can 0 perform a Taylor expansion in energy χR (q.18)
In the extreme long wave limit we have W RPA (0. and discuss a physical interpretation of the random phase approximation. 0) = .
(13.218
CHAPTER 13. and thus determine χR (q. 0) −→ 0 q→0 2
F ∂n dp (ξp+q − ξp ) ∂ξp = − dξp d(µ + ξp ) − F 3 (2π) ξp+q − ξp ∂ξp
∂n
−d(εF ). (10.1
Calculation of the pair-bubble
In Eq. (13. iqn ) = 2 dp nF (ξp+q ) − nF (ξp ) . In this limiting case ξp+q → ξp .
2 ks ≡ −4πe2 χ0 (0.2.34) the pair-bubble diagram is given in terms of a p-integral and a Matsubara frequency sum. 0 Note that the pair-bubble is a function of both momentum and frequency. 0) becomes 1 W RPA (q → 0. From now on we no longer need a ﬁnite value of α. a screened Coulomb interaction 4πe2 0 . q → 0 and iqn = 0 + iη. but with the important diﬀerence that the artiﬁcially introduced parameter α in the latter has been replaced with the pair-bubble function −4πe2 χ0 (q. (10. i. 0) is simply minus the density 0 of states at the Fermi level. 0
(13. iqn ) thus has a form similar to W (q).85) using the recipe Eq. However.e.19). (13. 0) = −1 .
13.17) W RPA (q.21)
In the static. iqn ) having its origin in the dynamics of the interacting electron gas. and consequently.22) d(εF )
. it is a simple matter to obtain the static.

ω + iη) =
1
kF 0
dp 2 p 2π
1 a
1 −1
dλ [nF (ξp+q ) − nF (ξp )] δ(ξp+q − ξp −ω ). but limit the calculation of χ0 (q. dp = (2π)3
∞ 0
dp 2 p 4π 2
1
dλ. 0) = 4π e2 d(εF ) = 0 0 0
4 kF . It is therefore useful for numerous applications.23)
a0 being the Bohr radius.
(13. (13.1 nm−1 .20) is Im χ0 (q. x0 ) ≡ 1 − 1 f (x. −x0 ) + 2 8x
2
. x − x2 + x0
(13.27a)
x0 −x x
ln
x + x2 − x0 . THE RENORMALIZED COULOMB INTERACTION IN RPA
219
The Thomas-Fermi screening wavenumber ks is found by combining Eq. 4εF (13.26)
and then the λ-integral followed by the p-integral is carried out using standard logarithmic integrals1 . π a0
(13. In the low temperature limit an analytical expression is obtained by a straightforward but rather tedious calculation. ω + iη) to the low temperature regime kB T εF .31) and (2.
.20) in two terms. substituting p with p − q in the ﬁrst term. (2. and we have λ ≡ cos θ. (13.36).
(13. (13.24) 2m
In the low temperature limit the Fermi-Dirac distribution is a step-function.
2 ks = −4πe2 χR (0. We now turn to the more general case. ω ) = −2 d(εF ) 0 where f (x. x0 ) + f (x.2. For metals ks ≈ 0. (13.28)
Useful integrals are
R
1 dx ax+b =
ln(ax + b) and
dx ln(ax + b) =
1 [(ax a
+ b) ln(ax + b) − ax]. In the p-integral the only angular dependence of the integrand is through cos θ.
R
(13. This result is very important. and the real part of χ0 is most easily calculated by splitting Eq.
−1
ξp−q − ξp =
1 (q 2 − 2pqλ).18) with Eqs.27b)
The imaginary part of χ0 in Eq. such as x ≡ q 2kF and x0 ≡ ω .25) The integrand is now made dimensionless by measuring all momenta in units of kF and all frequencies and energies in units of εF . ω +iη) = −P
kF 0
dp 2 p 2π 2
1 −1
dλ nF (ξp )
1
1 2m
(q 2 −2pqλ)+ω
+
1
1 2m
(q 2 +2pqλ)−ω
. The ﬁnal result for the retarded function χR is 0 Re χR (q. because it relates the screening length 1/ks to microscopic parameters of the electron gas. Finite temperature eﬀects can be obtained by using the Sommerfeld expansion or by numerical integration. and collecting the terms again: Re χ0 (q.13.

For λ = 0 we have the non-interacting electron gas while for λ = 1 we retrieve the full Coulomb interacting electron gas. (13. for |x − x2 | < x0 < x + x2 8x x π x0 Im χR (q.119) the thermodynamic potential Ω ≡ U − T S − µN is given by
. (2π)3 pσ σ Consequently we can interpret χ0 (q.31)
where H0 is the kinetic energy and W the Coulomb interaction Eq. At the later time τ the process is reversed. and deﬁne Hλ ≡ H0 − µN + λW. ω ). The RPA scheme takes interaction processes into account thus changing the dielectric properties of the non-interacting electron gas.30) pσ p+qσ (τ ) cp+qσ 0 . where momentum q and 0 energy ω is absorbed by the electron gas (see also the discussion in Sec. τ > 0) = − c (τ ) c† 0 c† (13. That this is at all possible is perhaps surprising due to the presence of the two-particle Coulomb interaction. But using the equation of motion technique combined with an “integration over the coupling constant” method we obtain the result. THE INTERACTING ELECTRON GAS
Using δ(f [x]) = x0 δ(x − x0 )/|f [x0 ]|.83) we arrive at dp χ0 (q. A careful analysis of when the delta-function and the theta-functions are non-zero leads to π 1 − x0 − x 2 . we study the pair-bubble diagram a little closer in the (q.2
The electron-hole pair interpretation of RPA
We have learned above that the RPA results in a screened Coulomb interaction.29) for 0 < x0 < x − x2 0 2 x. 8. (2.
13. and the electron falls back into the hole state. ω ) = −d(εF ) (13. for other x0 ≥ 0. In the remaining sections of the chapter we study how the eﬀective RPA interaction inﬂuences the ground state energy and the dielectric properties (in linear response) of the electron gas. According to Eq. the λ-integral can be performed. In the time interval from 0 to τ an electron-hole pair is thus present. but this corresponds to a polarization of the electron gas.34). 0. (1.2.3
The ground state energy of the electron gas
We ﬁrst show how to express the ground state energy in terms of the single-particle Green’s functions G(k. τ )-representation. The imaginary part of χR (q. (10. Let λ be a real number 0 ≤ λ ≤ 1. τ > 0) as the sum of all processes of the following type: at τ = 0 an electron is created in the state |pσ and a hole in the state |p + qσ . and we now see the origin of the renormalization of the Coulomb interaction.5).
13. which correspond to an electron jumping from the latter state to the former. describes the corresponding dissipative processes. Choosing τ > 0 in Eq.220
CHAPTER 13. ikn ). To gain some physical insight into the nature of this renormalization. where x0 are the zeros of f [x].

ikn ) through the equation of motion λ for Gσ (k.36)
= δ(τ ) +
kσ
We now let τ = 0− = −η and note that the last term is nothing but the interaction part λW λ of the Hamiltonian.38)
λ λ λ We now utilize that 1 = [Gσ ]−1 Gσ and furthermore that [Gσ ]−1 = ikn − εk − Σλ to obtain σ
λW
λ
=
1 2βV
λ Σλ (k. Furthermore. τ ) using Eqs. using Fourier transforms we can at τ = −η write 1 1 λ λ Gσ (k.39)
. ckσ ](τ ) c† kσ
kσ
λ
λ ε Gσ (k. ikn ) Gσ (k.61b)
− ∂τ
1 V
λ Gσ (k. −η) = β ikn Gσ (k.35)
λ The expectation value λW λ can be related to Gσ (k.
(13. (5.3. We therefore arrive at the following compact expression. σ ikn kσ
(13. whence the ground state energy E of the system can be calculated as E = E 0 + lim
1 T →0 0
dλ λW λ
λ.31) and (10.34)
The subscript λ refers to averaging with respect to Hλ . ikn ) eikn η and δ(−η) = β ikn eikn η . ikn ) − 1 eikn η = 2 λW ikn kσ λ.
(13. β By diﬀerentiating with respect to λ we ﬁnd Ω(λ) = − ∂Ω 1 Tr −βW e−β(H0 −µN+λW) =− ∂λ β Tr e−β(H0 −µN+λW) = W
λ.33)
By integration over λ from 0 to 1 the change in Ω due to the interactions is found:
1
Ω(1) − Ω(0) =
0
dλ λW λ
λ.
(13. THE GROUND STATE ENERGY OF THE ELECTRON GAS
221
1 ln Tr e−β(H0 −µN+λW) . ikn ) eikn η = ikn kσ
1 βV
eikn η + 2 λW
ikn kσ
λ.32)
(13.
(13. τ ) − 2 k kσq
λ W (q) Tτ c† σ (τ ) ck +qσ (τ ) ck−qσ (τ ) c† λ . 1 βV
λ (ikn − εk ) Gσ (k. ikn ) eikn η . τ ) kσ
= δ(τ ) +
1 V 1 V
Tτ [Hλ . At T = 0 we have ∆E = ∆Ω.37)
Collecting the sums on the left-hand side yields 1 βV
λ (ikn − εk )Gσ (k. kσ k 2 (13.13.
(13.

second-order perturbation theory of Sec. ikn ) eikn η . THE INTERACTING ELECTRON GAS
and when this is inserted in Eq. To improve the high-density. We are now in a position to continue the expansion Eq. because it relates the ground state energy of the interacting system to the single-particle Green’s function and the related self-energy. (2.44)
This expression ends the discussion of the ground state energy of the interacting electron gas in the jellium model. This is because only this diagram is divergent without renormalization.35) we ﬁnally arrive at the expression for the ground state energy E = E 0 + lim 1 T →0 2βV
1 ikn kσ 0
dλ λ λ Σ (k. (13.222
CHAPTER 13. (13.8b. we do not have to expand the Green’s function G beyond ﬁrst order: Σλ (k. we know from Eq. Combining Eq.41) (13. The techniques are similar to those employed in the calculation of the pair-bubble diagram in Sec. (13. and Fig. The RPA renormalization of the interaction line in the second diagram in Eq. the most divergent diagram of each of the higher orders. it is a λ remarkable result.43)
−→
2. Moreover.<=> ?@ A BC D E F G
+ + + + + + + 2 + + .0622 log rs − 0. 13.41) needs to be renormalized. 2. 2.43) of E/N in terms of the dimensionless distance parameter rs .
. (13. Since the self-energy Σ contains diagrams from ﬁrst order and up. (13.41) we obtain to (renormalized) second order: E − E 0 ≈ lim
1 T →0 0
dλ λ
1
=
0
dλ λ
Note the similarity between the three diagrams in this expression for E − E 0 and the ones depicted in Fig.42) and (13.40)
This expression allows for an diagrammatic calculation with the additional Feynman rule 1 that limT →0 0 dλ must be performed at the end of the calculation.43) renders the diagram ﬁnite. Since the Thomas-Fermi wavenumber ks replaced α as a cut-oﬀ.6b. ikn ) ≈ σ
λ Gσ (k. (13. By employing the powerful quantum ﬁeld theoretic method.2 we include in Eq.42)
(13. 2.094 − 2 rs rs
(13.40) all diagrams up to second order and. ikn ) ≈
Note that only the second diagram in Eq.2) that this diagram must be proportional to log ks and hence to log rs . (13.
(13. Ry.916 + 0.40) with Eqs.8a. Fig.211 0. 2.1. through RPA. λ σ
(13.2. We will not go through the calculation of these diagrams. ikn ) Gσ (k. E N
rs →0
.

t) 4π 0 |r − r | ⇒ φind (q. (13. ω ) φext (q. ω ) + φind (q.
13.50)
.4. −q. Through the Coulomb interaction this in turn corresponds to an induced potential φind (r. ω )]. as in Sec. i. ω ) is the Fourier transform of the retarded Kubo density-density correlation function χR (q. we will also be able to study other aspects of the interacting electron gas.49)
Here the subscript ’eq’ refers to averaging in equilibrium. (13. with respect to H = H0 + W omitting H . ω ) = 1 + W (q) χR (q.47)
Collecting our partial results we have φind (q. t − t ).48)
where χR (q. Next step is to use the Kubo formula. ω).4.48) the total potential φtot (q. ω) φext (q. 6. see Eqs.
(13. ω ). we could ﬁnally solve the problem posed by the failed second order perturbation theory. ω) φext (q. −qt ) = −iθ(t − t )
1 V
ρ(qt).
(13.4 we study the linear response of the interacting system due to the perturbation H caused by φext . ω)] = (−e)2 Cρρ (q. ω ) φext (q. Using Eq. t − t ) ≡ Cρρ (qt. we write all expressions in Fourier (q. 6. ω)] with the external potential and with the retarded density-density correlator
R [−e ρind (q. (13. φtot (q. ρ(−qt )
eq
. (13. ω ) = φext (q. One suspects that also external potentials φext will be screened similarly.75) and (8. t) = dr −eρind (r .4
The dielectric function and screening
Already from Eq. and indeed. ω) ≡ e2 χR (q. the charge density. Since the unperturbed system even with its Coulomb interacting electrons is translation-invariant. ω ) = 1 W (q) [−e ρind (q.45)
where ρ(r) is the particle density and not. ω ) creates an induced charge density −eρind (q. (13. Having achieved this solution. ω ). (13.15) it is clear that the internal dynamics of the interacting electron gas lead to a screening of the pure Coulomb interaction. ω ) can be written in terms of the polarization function χR . The external potential φext (q.
R χR (q.e. ω )-space. ω ). (8.13. THE DIELECTRIC FUNCTION AND SCREENING
223
in casu resummation of the Feynman diagram series for the single-electron self-energy and Green’s function.76). as we shall see below. t). As in Sec. which relates [−e ρind (r. H = dr [−e ρ(r)] φext (r.46) e2
We divide with e2 since W (q) by deﬁnition contains this factor. this is in fact the case. In the following we focus on the dielectric properties of the system. ω ) = W (q) χR (q.

(10. 0) V τ
eq
. From Eq.52)
where χ(q. and we have divided the contributions to χ into two parts.56) kσ
H I
.53)
We will calculate the latter Green’s function using the Feynman diagram technique. η = 0+ has been inserted to ensure correct ordering. This rule was a direct consequence of the deﬁnition of G(k.96) we can read oﬀ the Fourier transform ρ(±q): ρ(q) =
pσ
c† cp+qσ . 10.24) directly and to write the diagrammatic expansion in (q. ω ) = χ(q. τ ).7 we can obtain χR (q. (1. ω ). iqn → ω + iη). ω ) we can determine ε(q. ε(q. τ ) = − 1 T V τ c† (τ )cp+qσ (τ ) c† k+qσ ckσ pσ
pσ kσ eq
ρq=0
eq
=
1 − ρq=0 V
eq
+
pσ kσ
Tτ cp+qσ (τ )ckσ c† (τ + η)c† k+qσ (η) pσ
connected . ˜ ˜ k k G(k.
(13.54)
Hence χ(q. ω ) (13.30) and the speciﬁc calculation in Sec. It is now possible to apply the Feynman rules Eq. pσ
ρ(−q) =
kσ
c† k+qσ ckσ . (12. iqn ) = χ(˜). (10.81): χ(q. where it for the single-particle Green’s function is stated that the diagrams must contain two Green’s functions with the external momentum k. One part where the two density operators are disconnected from one another. The second term has a structure similar to the simple pair-bubble diagram with an external momentum q ﬂowing through it.55) Here. The disconnected part is zero since the expectation of the charge density in the neutralized and homogeneous jellium model is zero. iqn )-space of χ(q. (12.224
CHAPTER 13. THE INTERACTING ELECTRON GAS
When recalling that φtot corresponds to the electric ﬁeld E. eq
(13. We only have to pay special q attention to rule (4). and the other part where they mix. τ ) given by Eq. as in Eq. ω ) by analytic continuation of the corresponding Matsubara Green’s function χR (q.6).
(13. and φext to the displacement ﬁeld D = 0 εE. But according to Eq. τ ) is seen to be a two-particle Green’s function of the form χ(q.51)
So upon calculating χR (q. τ ) = − Tτ ckσ (τ ) c† ⇒ ··· (13. τ ) ρ(−q. ω ). τ ) = − 1 T ρ(q. iqn ) is the Fourier transform in imaginary time of χ(q. we see that the following expression for the dielectric function or electrical permittivity ε has been derived: 1 = 1 + W (q) χR (q. (13.

except this is a two-particle Green’s function with two operators at each of the external vertices instead of just one.4. +
˜ k (13. we deﬁne the irreducible diagrams in the χ-sum as the ones that cannot be cut into two pieces by cutting any single interaction line :
U
= ≡
L M N O P Q R ST
˜ k ˜ k ˜ ˜ k+ q p ˜ + p−k ˜ ˜ ˜ k + p+q ˜ ˜ ˜ ˜ k+ q + + the sum of all irreducible diagrams in −χ(˜) q + + + + p ˜ ˜ k
(13. p ˜ −χ(˜) ≡ q
p+q ˜ ˜ ≡
˜ ˜ k+ q +
+
In analogy with the self-energy diagrams in Sec.57)
˜ ˜ k+ q
The initial (right) vertex absorbs an external four-momentum q while the ﬁnal (left) vertex ˜ reemits q .4..59)
. THE DIELECTRIC FUNCTION AND SCREENING
225
Likewise for χ(q. We must then have that χ(˜) is the sum of all possible diagrams that connect ˜ q the two ρ-vertices and that involve any number of Coulomb interaction lines.
(13.58)
+ . τ ). τ ) ∼
Tτ c† pσ
(τ )cp+qσ (τ )
ckσ c† k+qσ
⇒
p+q ˜ ˜
J K
p ˜ ..13. One straightforwardly gets the following vertices corresponding to ρ(q) and ρ(−q):
χ(q. 12....
−χirr (˜) ≡ q
p+q ˜ ˜
VWXYZ [
˜ ˜ k+ q
+ ..

iqn ) q
−χRPA (q.
(13. −χ0 (q..
\ ]^_ `abc de g f h
+ + + × + + . THE INTERACTING ELECTRON GAS
Hence we can resum χ(˜) in terms of χirr (˜) and obtain a Dyson equation for it. iqn ) = 1 − e2 irr χ (q. iqn ). iqn ) =
This results in the RPA dielectric function εRPA (q. iqn ). irr (q..61)
1 χirr (q. 1 − W (˜) χ0 (q. iq ) ε(q.65)
1−
.64)
=
=
(13. In RPA χirr (q. iqn ) we can determine the dielectric function. iqn ) = 1 − W (q) χirr (q. iqn ) . iqn ) is approximated by χRPA (q. q q −χ(˜) = q
=
=
=
with the solution
With this result for χ(q.. iqn ) n or more directly
(13. iqn ) 1 = 1 + W (q) = . + = −χ(˜) = q = = 1− −χirr (˜) q 1 − W (˜) χirr (˜) q q ε(q. iqn )
i l k m
−→
RPA
−χirr A (q. iqn ).63)
Note it is e2 and not e2 that appears in the last expression.62)
(13. q q q q
(13. iqn ) = RP
j
= −χ0 (q. 2 0q
+ .226
CHAPTER 13.
−χirr (˜) − χirr (˜) W (˜) χ(˜) .60)
(13. iqn ) is approx0 imated by the simple pair-bubble −χirr (q. iqn ) =
and the full correlation function χ(q. iqn ) 1 − W (q) χ 1 − W (q) χirr (q..

vp → vF . (13.70)
In this regime we see from Eq. 2 0 0q
(13. (13. iqn ) = 1 −
e2 χ (q. The goal of this section is to investigate the frequency dependence of the dielectric function ε(q. vF p → kF . The other is the external impurity potential Eqs. long wave lengths and low temperatures.
+ . ω ) by Eqs.
(13.
−1
ξp+q − ξp ≈ vp qλ. z≡ qvF λ ω 1. (11.68) + .72)
We rewrite the delta-function in energy-space to one in k-space. (13. One is the internal electron-electron interaction potential represented by the screened Coulomb interaction line W RPA . (13. (13. kB T εF .29) that Im χR = 0. and furthermore we introduce a small dimensionless variable z: δ(εp − εF ) = δ(p − kF ) . iqn ). ω − vp qλ
(13.66)
The entire analysis presented in this section leads to the conclusion that the external potentials treated in linear response theory are renormalized (or screened) in the exact same way as the internal Coulomb interactions of the previous section. but to draw some clear-cut physical conclusions. To proceed we adopt the following 0 notation λ ≡ cos θ.21) we obtain Re χR (q. iqn ) 1−
e2 2 0q
χ0 (q.73)
. (13. we conﬁne the discussion to the case of high frequencies.13. all deﬁned by the conditions vF q ω (or x (13. (13.69)
13. PLASMA OSCILLATIONS AND LANDAU DAMPING
227
εRPA (q.67)
This conclusion can be summarized in the following two diagrammatic expansions.25) and (11.20) and Taylor expanding nF as in Eq.29).71)
Utilizing this in Eq.
(13.. ω ) ≈ 0 1 2π 2 dp p2 δ(εp − εF )
1
dλ
−1
vp qλ .5.. iqn ) =
uRPA (q) =
nopq rstu
= + + = + + x0 ). iqn ) = 1 − W (q) χ0 (q. −W RPA (q. iqn )
. We could choose to study the full case described through χR (q. ω ). iqn ) = tot 1 εRPA (q. iqn ) φext (q.5
Plasma oscillations and Landau damping
We now leave the static case and turn on an external potential with frequency ω ..30) represented by the screened electron-impurity line uRPA . q kF (or x 1). iqn ) = φext (q.. dp = (2π)3
∞ 0
dp 2 p 4π 2
1
dλ.27a) 0 and (13. φRPA (q.

THE INTERACTING ELECTRON GAS
This in inserted in Eq. has been introduced.75). and above which they become transparent. (13.66) and (13. putting it in the ultra-violet part of the electromagnetic spectrum. ω ) = 0
2
⇒
2 ω 2 ≈ ωp +
3 (qvF )2 5
⇒
ω (q) ≈ ωp +
2 3 vF 2 q .75) ω 5 ω where the characteristic frequency ωp .0 eV. Typical values are ω ≈ 1016 Hz.74) we ﬁnd the RPA dielectric function in the high-frequency and long-wavelength limit to be 2 ωp 3 qvF 2 εRPA (q. chapter 14. The former parameter can be found by Hall eﬀect measurements. Using the observed parameters for aluminum. the plasma oscillations. The Fermi surface can be mapped out using this technique as described in Ashcroft and Mermin. (13. ω ) ≈ 0 ≈ = 1 2 1 ω k 2π 2 F vF qvF
2 1 kF ω 2 2π 2 qvF n q2 qvF /ω −qvF /ω
dz
z 1−z
+qvF /ω −qvF /ω
1 2 1 3 1 4 1 5 z + z + z + z 2 3 4 5 3 qvF 5 ω
2
m
ω2
1+
. 8. it marks the limit below which metals reﬂects incoming electromagnetic radiation.g. and the smallness of this new variable permits the Taylor expansion z/(1 − z) ≈ z + z 2 + z 3 + z 4 .
(13.115 m0 .16).27 × 1016 Hz = 15.76) ωp ≡ m 0
13. ω ). (13. Note that ε(q. The variable λ is substituted by z.81 × 1029 m−2 and m = 1. Solid State Physics.5. 2055 (1963) regarding aluminum. ω ) φtot (q. φext (q. Re χR (q. (13. Theoretically.1
Plasma oscillations and plasmons
The plasma frequency is an important parameter of the interacting electron gas setting the energy scale for several processes.74)
3 where in the last line we used vF = kF /m and 3π 2 n = kF . well known as the electronic plasma frequency. e. the existence of these oscillations is proved as follows.77)
The de Haas-van Alphen eﬀect is oscillations in the magnetization of a system as the function of an applied external magnetic ﬁeld. For the determination of the electron band mass m in aluminum see N. we obtain ωp = 2. ne2 . 10 ωp
(13. Philos. We are thus about to identify an oscillatory eigenmode for the electron gas. εRPA (q. Al n = 1. A very direct manifestation of the plasmon frequency is the existence of the collective charge density oscillations. ω ) = 1 − 2 1 + .72). while the latter can be determined from de Haas-van Alphen eﬀect2 .
. ω ) = ε(q. (13. ω ) = 0 in fact allows for a situation where the total potential varies in space and time in the absence of any external potential driving these variations. Ashcroft.W.228
CHAPTER 13. Combining Eqs. It is determined by the electron density n and the eﬀective band-mass m of Eq. Mag. (2. Consider the relation D = ε 0 E or the related one. Let us calculate its properties in RPA from Eq.

The ﬁnal energy.1(b). PLASMA OSCILLATIONS AND LANDAU DAMPING
229
Figure 13.
Recall that in the high frequency regime Im χR and consequently Im ε is zero. In the experiment high energy electrons with an initial energy Ei = 20 keV are shot through a 258 nm wide aluminum foil. Phys. here with the emission of three plasmon during the traversal.1(a). is measured on the other side of the foil. The result of the measurement is shown in Fig. The initial energy is Ei = 20 keV. 126. But how could one be convinced of the existence of these oscillations? One beautiful and very direct veriﬁcation is the experiment discussed in Fig. If some eigenmodes exist with a frequency ∼ ωp .8 eV in very good agreement with the value of the plasma frequency of 15. 13.5.0 eV. The energy loss Ei − Ef clearly reveals loss in quanta of ∆E. On the plot electrons exciting as many as seven plasmons are clearly seen. Thus by Eq. and the energy loss Ei − Ef can be plotted. and the ﬁnal energy Ef is measured at zero scattering angle on the other side of the foil. as is the case with any harmonic oscillator. 13. they must be quantized leading to oscillator quanta. Some electrons traverse the foil without exciting any plasmons (the ﬁrst peak). denoted plasmons. They have a simple quadratic dispersion relation ω (q) starting out from ωp for q = 0 and then going up as q is increased. so no 0 damping occurs. 182 (1962).13. Rev.0 eV for bulk aluminum. with a characteristic energy of ωp . but instead a traversal during which two plasmons are excited. The value of the energy loss quantum was measured to be ∆E = 14. (b) A sketch of a typical microscopic process. The energy loss clearly happens in quanta of size ∆E. the plasma oscillations. by Marton et al. others excite one or more as sketched in Fig.77) it is indeed possible to ﬁnd oscillatory eigenmodes. 13.1.
50 234 234
234 @8 5 7 16 0 9 0
234
10
¢ )
¦¦¦¨¦¨¦¤ © § ¥
%% $$ #¤ ¥" ¥" ! &'(
¡¢ £
. then. (13.1: (a) Observation of plasmons in high-energy electron transmission spectroscopy on a 258 nm wide aluminum foil.8 eV in good agreement with the plasma frequency determined by other methods to be 15. The energy quantum ∆E was found to be 14. Ef . Note that the most probable process is not the plasmon-free traversal.

ω ). and this region is shown in Fig. Note that ImχR (q. 13.R φtot (q.230
CHAPTER 13. ω ) 0 . εF = 11. ω ) + iη 0
If. we discuss the damping of excitations. (13. ω + iη)
0
In the case of a vanishing imaginary part Im χ0 we ﬁnd a pole on the real axis: φRPA.2: A gray scale plot of Im χR (q. signaling 0 a temporal decay of the total potential with a decay time proportional to Im χR . ω ). The pure plasma oscillations discussed above are examples of undamped 0 or long-lived excitations. (13.26): x = q/2kF and x0 = ω /4εF . however. ω )-plane of non-vanishing Im χR . ω ) = 0 only in the gray scaled area. which is described by the imaginary part Im χR (q. ω ) RPA. 0 The variables are rescaled according to Eq. ω ) + i eq2 0
In Eq. ω ) = tot φext (q.67) φext (q. which is bounded by the constraint functions 0 given in Eq.2. (13. This can be elucidated by going to the retarded functions in Eq.0 eV. (13. THE INTERACTING ELECTRON GAS
Figure 13. ω ) 2 R χ0 (q.7 eV and ωp = 15.29). Also shown is the plasmon branch with its propagating and damped parts.29) we have within RPA calculated the region the (q. (13.5. ω ) = tot
e2 2 0q
1−
Re
φext (q.79) Im χR (q. ω ) 1−
e2 2 0q
Re χR (q. The physical origin of the non-zero imag0 inary part is the ability for the electron gas to absorb incoming energy by generating
§ ¨ & 2 2 "# ! # D $3 E # ! D $3 3 3 "C ! " % "% 1 3 B A @ B 978© 7& 5 6 % # & % $ 1 3 & 12 1 # 0
V WU T S PQ IG F R H 4 ¡ 4
.R (q. The darker a shade the higher the value.78) e2 1 − q2 χ0 (q. (13. The parameters chosen for this branch are those of aluminum. (13.2
Landau damping
Finally.R (q. ω ) = . 0 φRPA.80)
§¦¥ )' '¤ £¢ ' ' ¡ ("# % ! # & % $
§¦¥ ©¨ £¢
. ImχR = 0 we end up with a usual Lorentzian peak as a function of ω.
13.

2 is shown the dispersion relation for the plasmon excitation. here expressed by their Dyson’s equations −W RPA (q. Another way to understand the eﬀect of a non-vanishing Im χR is to link it to the 0 conductivity σ of the electron gas.47) it follows that 1 e2 Im χR = − q· Re σ ·q. ω )-space. energy and momentum constraints prohibit the excitation of electron-hole pairs. It starts out as a bona ﬁde excitation in the region of the (q. iqn ) = = + +
v wx y z { |}~
× + + + + . = 2.
. (13.211 0. In particular we found the self-energy ΣRPA (k. at some point the dispersion curve crosses into the dissipative Im χ0 = 0 area.
13. when a current J is ﬂowing. ω )-space where the RPA dissipation is 0.13.916 + 0. Outside the appropriate area in (q. 13. . Finally we remark that in Fig. However. The main result was the RPA resummation of diagrams to all orders in perturbation theory valid in the high density limit. It is well-known that the real part of σ is associated with dissipation (Joule heating). and there the plasmon acquires a ﬁnite life time.0622 log rs − 0.81) 0 ω whereby it is explicitly conﬁrmed that a non-vanishing ImχR is associated with the ability 0 of the system to dissipate energy. .6
Summary and outlook
In this chapter we have used the Feynman rules for pair-wise interacting particles to analyze the Coulomb-interacting electron gas in the jellium model. But from Eq. and the electron gas cannot absorb energy by that mechanism. One main result was ﬁnding the screening of the Coulomb interaction both for the internal interaction and for external potentials. In the literature this damping mechanism is denoted Landau damping. Hence the plasmons have inﬁnite life times for small q. SUMMARY AND OUTLOOK
231
electron-hole pairs.6. and they are damped out as a function of time. (6.094 − 2 rs rs uRPA (q) = =
Ry. ikn ) = σ
This result was used to calculate the ground state energy of the electron gas E −E 0 = N + + =
We also used the RPA analysis to study the dielectric properties of the electron gas. In other words for high q-values the plasmonic excitations are not exact eigenmodes of the system.

ω
2 ks 4 kF 2 . ω ) was found in two cases. ω (q) = ωp +
2 3 vF 2 q . where ks = q2 π a0 2 ωp 3 qvF 2 qvF ) = 1 − 2 1 + . long-wave limit and (ii) the high frequency.
. εRPA (q. 0) = 1 + εRPA (q. ω 5 ω
Finally. long-wave limit. and of many other physical phenomena involving the electron gas. electron-phonon interaction. Moreover. 10 ωp
where ωp ≡
ne2 . superconductivity. (i) the static. THE INTERACTING ELECTRON GAS
Explicit expressions for the dielectric function ε(q. we studied the plasma oscillations of the electron gas found from the condition εRPA (q. m 0
The RPA analysis has already given us a good insight in some central physical properties of the electron gas. it plays a crucial role in the studies of electron-impurity scattering. and found the dispersion relation involving the plasma frequency ωp .232
CHAPTER 13. ω ) = 0.

In this description the potential is allowed to vary in space due to some external perturbation or due to interactions with the inhomogeneous density of quasiparticles. Using the Landau transport equation we shall see that the collective modes derived in the previous chapter also come out naturally from a semi-classical description and.
14. This equation is known as the Landau transport equation. The basic conclusion is that a gas of interacting particles can be described by a system of almost non-interacting “quasiparticles”. because it explains why the apparently immensely complicated system of for example interacting electrons in a metal can be regarded as a gas of non-interacting particles. which is calculated from microscopic considerations in Chap. The quasiparticle distribution function satisﬁes a kinetic equation. because it means that one cannot cross a phase
1 See for example the collection of reprints in the book: D.1
Adiabatic continuity
The Fermi liquid theory is based on the assumption that starting from the non-interacting system of particles one can analyze the interacting case by applying perturbation theory. 15. which may include scattering from one state to another for example due to impurity scattering. furthermore.1997). and it is equivalent to the well-known Boltzmann equation from kinetic gas theory. The quasiparticle concept furthermore gives the theoretical foundation of the semiclassical description. The Fermi liquid theory is conceptually extremely important.
233
. This is of course an enormous simpliﬁcation. and this condition will set the limits for the applicability of the theory. This is in fact a rather stringent criterion. AddisonWesley (1961. also is easily understood in terms of scattering of quasiparticles. What we mean by “suﬃciently short” of course has to be quantiﬁed. Pines The Many-body problem. the conductivity.Chapter 14
Fermi liquid theory
The concept of Fermi liquid theory was developed by Landau in 1957-59 and later reﬁned by others1 . These quasiparticles are approximate excitations of the system at suﬃciently short time scales. and it gives the theoretical explanation of why all the results that one gets from the widely used free electron model work so well.

which yields i∂t ψadia (x.3) into Eq. (14. t) ∂V0 (t) ∂V0 (t) ∂t = Hψadia (x.5)
Thus we have an approximate solution if the ﬁrst term dominates over the second term. (14. This is because a phase transition. 4. i.3) is estimated by inserting Eq. (14. t) = EV0 (t) ψadia (x. t) ≈ ψV0 (t) (x) exp −iEV0 (t) t . The phase with broken symmetry can only occur if the ensemble of states in the statistical average is truncated. If you imagine that we start from the non-interacting system excited in some state and then turn on the interaction adiabatically. with o energy EV0 (t) H(t)ψV0 (t) (x) = EV0 (t) ψV0 (t) (x). t) = −V0 (t) exp −x2 /2x2 .234
CHAPTER 14. cannot be reached by perturbation theory starting from the paramagnetic phase.e. (14. The one-dimensional potential will have a number of bound states with discrete eigenenergies and a continuum of eigenenergies corresponding to the delocalized states. As a simple example of adiabatic continuity we now consider a particle trapped in a one-dimensional potential. t) = H(t)ψ(x. t). t) + ∂ψadia (x.2).1) 0 where the depth of the well is time dependent. 2m (14. The accuracy of the solution in Eq. t) = can be approximated by the adiabatic solution ψadia (x.2 If the excitations of the non-interacting system are connected to the excitations of the interacting system by a one-to-one correspondence (at least on short time scales as explained below) the two cases are said to be connected by “adiabatic continuity”. (14. then we would end up in a corresponding excited state of the interacting system. If this change is slow the solution of the Schr¨dinger equation o i∂t ψ(x. FERMI LIQUID THEORY
boundary line.4) Note that both ψV0 (t) (x) and EV0 (t) depend parametrically on the time through V0 (t). t) ψ(x.
. so slowly that the occupation numbers are not changed.2)
where ψV0 (t) (x) is the solution of the static (or instantaneous) Schr¨dinger equation. As an example consider a potential of the form V (x. (14. and let us suppose that it is changing from an initial value V01 to a ﬁnal value V02 . (14. 4.3) p2 + V (x. such as for example the ferromagnetic transition discussed in Chap. What we really are claiming is that the excited states of the interacting system can be labelled by the same quantum numbers as those we used to label the non-interacting system by.3. We now imagine changing the potential slowly. Thus apparently our conclusion is that if the rate of change of V0 (t) is small enough the solution for the new value of V0 = V02 can be found be by starting from the solution
2 This fact you can understand from the concept of broken symmetry explained in Sec. t).

ADIABATIC CONTINUITY
235
with the old value of V0 = V01 and “adiabatically” changing it to its new value. The following arguments are meant to be the full theoretical explanation for the applicability of Fermi liquid theory. such as response functions or occupation numbers we are facing matrix elements between diﬀerent states. For example if the ﬁrst excited state is a bound state. etc) all the way back to the non-interacting case. t > 0.9)
.
(14.1. 0)|(kσ)p t→∞
−→
(k σ )p |(kσ)p 0 . but most importantly it is still the ﬁrst excited state and it is still a bound state. In doing so we gain the fundamental understanding that the interacting and the non-interacting cases have a lot in common at least under some restricting circumstances. 0)|kσ .6)
where |G is the groundstate of the interacting system. but rather to give a physical intuition for the reason why the quasiparticle picture is valid. This may sound completely trivial. It is used to bring the excitations of the interacting case back to the well-known excitations of the non-interacting case.1
The quasiparticle concept and conserved quantities
The principle of adiabatic continuity is now utilized to study a system of interacting particles. then the matrix elements are identical to those of the non-interacting case. (14. In the following this idea is applied to the case of interacting particles. If we now imagine letting time evolve according to a Hamiltonian where the interaction is gradually switched oﬀ at a rate ζ Hζ = H0 + Hint e−ζt . The two states can simply not be connected through small changes of V0 . (kσ )h . let us consider states with single particles or single electron-hole pairs added to the groundstate |(kσ)p = c† |G . and it is not always true.14. for example between states with added particles or added particle-hole pairs. Since we are dealing with the low energy properties of the system. because one is a decaying function and one is an oscillatory function. (14. we can bring the states (|(kσ)p . it will change to a somewhat modiﬁed state with a somewhat modiﬁed energy. but it is not. This turns out to be realized in many systems. kσ |(kσ)p . When calculating physical quantities. For example if the real solution during this change of V0 from V01 to V02 changes from a bound state to an un-bound state (if V02 is small enough there is only one bound state). (5. |(kσ)p . The important message is. This is an example where perturbation theory to any order would never give the right answer. that if we avoid these transitions between diﬀerent kinds of states. under the conditions of adiabaticity.18) the time evolution with the time dependent Hamiltonian is |kσ (t) = e−iH0 t Tt exp −i
t 0
dt Hζ (t ) |kσ ≡ Uζ (t. 0)Uζ (t.1. adiabatic continuity does work.7) then according to Eq. For example
† (k σ )p |(kσ)p = (k σ )p |Uζ (t. thus making computation possible. The ﬁrst term inserts a particle while the second term creates both a particle and hole. (k σ )h = c† ck σ |G .
(14.
14. however. then it does not matter how slowly we change V0 .8)
If. kσ etc.

Thus there is always a temperature range at low temperature where Eq. i. |(kσ)p . (14. 2. and (3) it has excess spin σ. that is τlife ζ −1 . and (3) the total spin S = kσ σnkσ all commute with the Hamiltonian. εkσ ζ.236
CHAPTER 14. or. namely those that are conserved by the Hamiltonian: (1) it has an excess charge e (compared to the groundstate). This leads us to the deﬁnition of quasiparticles:
. FERMI LIQUID THEORY
There are two important assumptions built into this construction:
1. Here −e and m are charge and mass of the electrons. Below we shall see that it is indeed possible to make the approximations consistent.10)
The last condition can in principle always be met at high enough temperatures. or in other words the life-time τlife of the state is long compared to ζ −1 . These properties are all conserved quantities because the corresponding operators (1) the total charge Q = −eN . Most importantly. The adiabatic procedure is valid when the energy of the state is large compared to the rate of change. whereas the ﬁrst one is not necessarily possible. respectively. Next we discuss the properties of the state with an added particle. the adiabatic continuity principle can also be used to calculate the distribution function.
(14. this is equivalent to assuming kB T ζ. namely when we can choose a switch-oﬀ rate ζ such that
−1 τlife
ζ
kB T. 0)|kσ has a number of properties in common with the initial state |(kσ)p . τlife ∝ T 2 . (2) it carries current −e k/m. It is clear that the state where the interaction is switched oﬀ Uζ (∞. and therefore the distribution function c† σ ckσ = (k σ )p |(kσ)p −→ (k σ )p |(kσ)p 0 is a k Fermi-Dirac distribution function.
This apparently leaves an energy window where the idea makes sense. since typical excitation energies are of order of the temperature. (2) the total current Je = −e kσ vk nkσ . because the life-time turns out to be inversely proportional −1 to the square of the temperature.e.10) is fulﬁlled. The interactions do not induce transitions of the states in question.

if the local potential of the quasiparticles is space and time dependent so is then the density of quasiparticles. because they will of course tend to move towards the low potential regions. Finally the quasiparticles are in equilibrium distributed according to the Fermi-Dirac distribution function. The quasiparticle concept only makes sense on time scales shorter than the quasiparticle life time.
3
. t). t). this only makes sense on length scales larger than a typical thermalization length. (k σ )h ). SEMI-CLASSICAL TREATMENT OF SCREENING AND PLASMONS
237
Quasiparticles are the excitations of the interacting system corresponding to the creation or annihilation of particles (for example particle-hole pair state |(kσ)p . Thus we write the resulting local potential φtot (r.11)
The induced potential φind created by the induced density ρind . Consider a uniform electron gas which is subject to an external potential φext (r. (14. This in turn changes the electrical potential because the quasiparticles are charged and therefore the total potential φtot is given by the sum of the external potential φext and the induced potential φind . and they therefore constitute a dilute gas. The quasiparticles can be labelled by the same quantum numbers as the non-interacting case. The induced potential is caused by the excess or deﬁcit of quasiparticles. At low temperatures there are only a few quasiparticles. provided that the corresponding operators commute with the Hamiltonian.2
Semi-classical treatment of screening and plasmons
In Chap. The quasiparticle are thus not to be thought of as the exact eigenstates. For a translation-invariant system of electrons interacting through the Coulomb interaction. because otherwise we could not talk about a local potential.
Note that we are here invoking a new concept namely local equilibrium.14. In the following we make use of the quasiparticle concept to calculate the screening and the transport properties of an electron gas. the quasiparticles quantum numbers are thus k and σ and they carry charge −e and velocity vk = k/m. t) as φtot (r. Here we shall rederive some of this using a less rigorous but maybe physically more appealing approach. must be determined self-consistently. t) + φind (r. We can include the external potential as a local change of the potential felt by the charged quasiparticles. The thermalization length is the length scale on which thermal equilibrium is established. which in turn depends on the total potential. 12 we saw how the collective modes of a charged Fermi gas came out of a rigorous diagrammatical analysis.
14.3 Now.2. Clearly. t) = φext (r.

17)
.2
Dynamical screening
In the dynamical case. (13. The ﬁrst dependence arises from the ﬂow of the distribution function.67) using χR = −d(εF ) from Eq.11) yields φtot (r) = φext (r. To linear order in the local total potential and at low temperatures the induced charge density is given by ρind (r) = 2 V nF ξk + (−e)φtot (r) − nF ξk
k
≈ −(−e)φtot (r)
2 V
−
k
∂nF (ξk ) ∂ξk
≈ −(−e)φtot (r)d(εF ).1
Static screening
First we consider linear static screening. (14.21).2. we expect to ﬁnd collective excitations similar to the plasmons found in Sec. 0 ⇒ φtot (q) = (14. The way to describe this is to look at the deviation of the distribution function nk of a quasiparticle with a given momentum p = k (below we as usual use = 1).16)
The dynamics are controlled by two things: the conservation of charge and the change of momentum with time. (13.5. This depends on both r and t.15) in full agreement with the conclusions of the RPA results Eqs. 1 + W (q)d(εF )
which when inserted into (14.
(14.66 ) and (13. Consequently.14)
14. In order to treat this case we need to reﬁne the analysis a bit to allow for the time it takes the charge to adjust to the varying external potential. (14. The conservation of particles in state k is expressed in the continuity equation nk + ˙
r
· jk = 0. the induced charge density at point r at time t now depends on the total potential at some other point r and at some other (previous) time t . From this we get the induced potential in real space and in q-space as φind (r) = 1 −e dr W (r−r )ρind (r) ⇔ φind (q) = 1 W (q)ρind (q) = −W (q)φtot (q)d(εF ).238
CHAPTER 14. so that nk = nk(t) (r.13) φext (q) . FERMI LIQUID THEORY
14. the number of quasiparticles in each state is conserved. 13.12)
where ξk is quasiparticle energy measured relative to the equilibrium chemical potential and d(εF ) is the density of states at the Fermi level. t) − W (q)d(εF )φtot (q) and hence ε(q) = 1 + W (q) d(εF ). t). −e (14.2.
(14. Because we are interested in times shorter than the life time of the quasiparticles.

(14. based on renormalization by summation of the most important diagrams. (14. gives the same result as the fully microscopic theory.19) we ﬁnd (−iω + iq · vk )nk (q. The second dependence follows from how a negatively charged particle is accelerated in a ﬁeld.2. This is inserted into Eqs. ω) = q· k ξk ω − q · vk − ∂nF (ξk ) ∂ξk (−eφtot (q.19)
Again it is convenient to use Fourier space and introducing the Fourier transform nk (q. ω). (14.24)
which agrees with Eq. (14. We have also gained some physical understanding of this renormalization. We have thus shown that in the long wavelength limit the semi-classical treatment. Using Eq. simply from Newton’s law p = −(−e) ˙
r φtot (r.13) and (14. i. t). 13. (14.18) which is known as the collision-free Boltzmann equation4 . which relies on the Fermi liquid theory.14.22)
k
where the factor 2 comes from to spin degeneracy. because k ξk = vk .
. Note that q drops out because W (q) ∝ q −2 .
From this expression we easily get the induced density by summation over k ρind (q.5. (14. because we saw explicitly how it was due to the screening of the external potential by the mobile quasiparticles. ω) = 2 V q· k ξk ω − q · vk − ∂nF (ξk ) ∂ξk (−eφtot (q.21)
∂nk ∂ξk
φtot (q. ω) = −ie (q ·
k nk ) φtot (q. ω)). At long wavelengths or large frequencies qv ω.
(14.e. ω)).15). (13.75) in Sec. (14.
4 Here r and t are independent space and time variables in contrast to the sometimes used ﬂuid dynamical formulation where r = r(t) follows the particle motion.23)
k
At ω = 0 we recover the static case in Eq.20) To linear order in the potential φtot we can replace the nk on the righthand side by the equilibrium distribution n0 = nF (ξk ) and hence we ﬁnd k nk (q. and hence we get ˙ ∂t (nk ) + k · k nk + vk · r nk = 0.11) and we obtain the dielectric function ε = φext /φtot in the dynamical case ε(q) = 1 − W (q) 2 V q· k ξk ω − q · vk − ∂nF (ξk ) ∂ξk .
(14. ω)
= ie (q·
k ξk )
−
(14. SEMI-CLASSICAL TREATMENT OF SCREENING AND PLASMONS
239
where the current carried of quasiparticles in state k is given by jk = vk nk = ( k/m)nk . ω). we ﬁnd by expanding in powers of q that ε(q) ≈ 1 − W (q) 2 ω2 V (q · vk )2 − ∂nF (ξk ) ∂ξk =1− ωp ω
2
.

(14.25) m
where σ is the conductivity and τp−relax is the momentum relaxation time. the spin is conserved and thus σ = σ . at which scattering from the state |kσ to some other state |k σ occurs. and we get
2
Γ(k σ← kσ) = Γk σ. where |kσ 0 = eik·r / V. The new scattering process thus introduced means that the number of quasiparticles in a given k-state is no longer conserved and we have to modify Eq.3
Semi-classical transport equation
Our last application of the semi-classical approach is the calculation of conductivity of a uniform electron gas with some embedded impurities. In steady state these forces are in balance and hence −(−e)E+ mv τp−relax =0 ⇒ J = −env = e2 nτp−relax E m ⇒ σ= ne2 τp−relax . FERMI LIQUID THEORY
14. We know now that they follow a Fermi-Dirac distribution. but amazingly the result turns out to be the same.17) to take into account the processes that change the occupation number nk . The rate of change is given by the rate. Microscopically the momentum relaxation corresponds to scattering of quasiparticles from one state |kσ with momentum k to another state |k σ with momentum k . The fact that the scattering on an external potential is an elastic scattering is reﬂected in the energy-conserving delta function.27) Vimp (r) =
i
We can then ﬁnd the rate Γ by the adiabatic procedure where the matrix element k σ|Vimp |kσ √ is identiﬁed with non-interacting counterpart k σ|Vimp |kσ 0 . This will in fact lead us to the famous Drude formula.26)
where Vimp is the impurity potential. It can be found from Fermi’s golden rule Γ(k σ← kσ) = 2π k σ|Vimp |kσ
2
δ(ξk − ξk ). 10. 11) u(r − Ri ) (14.
(14. As explained in Chap. (14. A simple minded approach to conductivity would be to say that the forces acting on a small volume of charge is the sum of the external force and a friction force that is taken to be proportional to the velocity of the ﬂuid at the given point. These collisions take momentum out of the electron system.28)
.
(14. Γ(k σ← kσ). thus introducing a mechanism for momentum relaxation and hence resistivity.kσ
2π = 2 V
dr e
j
−ik ·r
u(r − Rj )e
+ik·r
δ(ξk − ξk ). the ﬁnite resistivity of metals at low temperatures is due to scattering against impurities or other imperfections in the crystal structure.240
CHAPTER 14. The total impurity potential is a sum over single impurity potentials situated at positions Ri (see also Chap. the ones considered here. Historically. namely by assuming that the charge carriers form a classical gas. In Sec. For non-magnetic impurities. 15 we will furthermore see how the very same result can be derived in a microscopic quantum theory starting from the Kubo formula and using a diagrammatic approach. the Drude formula was ﬁrst derived in an incorrect way.

k = Wk .
k
(14. (14.29)
Now the change of nk due to collisions is included in the diﬀerential equation Eq.k . The time derivative of nk becomes nk(t) (r.e.k (nk − nk ) . (14.18). The average is done assuming only lowest order scattering.33) V
k
The Boltzmann equation for impurity scattering is rather easily solved in the linear response regime.k (nk − nk ) . i. we obtain nimp ˆ −iωnk + eE · k δ(kF − k) = − V Wk . First we note that p = −eE.k . t) = ˙ d nk dt +
ﬂow−force
∂ nk ∂t
. because the right hand side is then of the form τ −1 nk similar to −iωnk on the left hand side. (14. Because Wk.k − nk (1 − nk )Wk. V
(14.k . We obtain Γk .14. Let us furthermore concentrate on the long wave-length limit such that r nk ≈ 0.30)
where the change due to “ﬂow and force” is given by the left hand side in Eq.31)
The ﬁrst term in the sum represents the rate for being scattered out of the state k and the second term represents the rate for being scattered into to state k from the state k . where k is a unit vector k oriented along k. and multiplied by the impurity potential for a single impurity u(r).
k
(14. (14.k = 2π nimp V dr ei(k−k )·r u(r) δ(ξk − ξk ) ≡
2
nimp Wk . we have nimp ∂ nk =− Wk . which at zero ˆ ˆ temperature becomes k n0 = k θ(kF − k) = −kδ(kF − k). and therefore to linear order in E the ˙ ˙ term k nk multiplying k can be replaced by the equilibrium occupation.k (nk − nk ) .29) times the probability for the initial state to be ﬁlled and the ﬁnal state to be empty.3. SEMI-CLASSICAL TRANSPORT EQUATION
241
Of course we do not know the location of the impurities exactly and therefore we perform a positional average. By going to the frequency domain. Therefore we can simply replace the sum over impurities by the number of scattering centers.
collisions
(14. leaving out interference between scattering on diﬀerent impurities. Nimp = nimp V.32) ∂t V collisions
k
and the full Boltzmann transport equation in the presence of impurity scattering now reads nimp ˙ ∂t (nk ) + k · k nk + vk · r nk = − Wk .34)
Without the nk -term on the right hand side this equation is simple to solve. The total rate is obtained from the Fermi golden rule expression (14. This
. The new collision term is not a derivative but an integral functional of nk ∂ nk ∂t =−
collisions
nimp V
nk (1 − nk )Wk .17) as an additional term.

For a uniform system the result cannot depend on the direction of the electric ﬁeld.37)
(14. and therefore we can put E parallel to k. because it is the time that enters the expression for the conductivity.
k =kF
ˆ ˆ Wk .k ).35) nk (ω) = iω − 1/τ tr where the relaxation time τ tr needs to be determined. (14. The result for the conductivity is 1 ne2 τ tr σ = σ0 .40)
3 where we have used the relation between density and kF .k (1 − cos θk.k kδ(kF − k)−k δ(kF − k ) · E. as we see by calculating the current density J=− =− = 2e V 2e V nk vk
k
τ tr
k
eδ(kF − k) ˆ k k·E tr iω − 1/τ m
1 E 1 ∞ dk k 3 δ(kF − k) d(cos θ) cos2 θ (2π)2 −iω + 1/τ tr m 0 −1 2e2 E 1 e2 n 3 2 = kF = E. −iω + 1/τ tr (2π)2 m 3 (−iω + 1/τ tr ) m
2e2
(14.
|k |=kF
(14.38)
Here θk is angle between the vector k and the electric ﬁeld E. tr V iω − 1/τ
|k |
(14. we can set k = k = kF and remove the common factor δ(kF − k) to get nimp −iω −e ˆ ˆ eE · k + eE · k = tr tr V iω − 1/τ iω − 1/τ which is solved by cos θk nimp 1 = τ tr V Wk .36)
Since Wk . σ0 = .41) 1 − iωτ tr m
. n = kF /3π 2 . (14.39) = τ tr V
k =kF
The time is known as the transport time.k (cos θk − cos θk ).k includes an energy conserving delta function.242
CHAPTER 14.k (k − k ) · E. so let us try the ansatz 1 ˆ eE · kδ (kF − k). That this is in fact a solution is seen by substitution −iω ˆ ˆ eE · kδ(kF − k) + eE · k δ(kF − k) iω − 1/τ tr nimp −e ˆ ˆ = Wk . (14. FERMI LIQUID THEORY
hints that we can obtain the full solution by some imaginary shift of ω. and get nimp 1 Wk .

k . (14.43)
This time expresses the rate for scattering out of a given state k.49 ). Here we investigate the validity of this approach by studying the rate of quasiparticle-quasiparticle scattering. such that the initial and the ﬁnal energies are the same εk + εk = εk+q + εk −q or counting from the chemical potential ξk + ξk = ξk+q + ξk −q .32) and it is therefore incorrect. but it does not tell us how much the momentum is degraded by the scattering process. k − qσ .e. Therefore the transport time is precisely the momentum relaxation time deﬁned in the simple ﬂuid dynamical picture in Eq. cos θ ≈ −1. i.3. The rate for quasiparticle-quasiparticle scattering can be calculated using Fermi’s golden
. This is precisely what the additional cosine-term in Eq. namely the socalled relaxation time approximation. the ﬁnal state will be a state |k + qσ. (14. and τ0 is the relaxation time. k σ scatter. Clearly there is a mechanism for quasiparticle scattering against quasiparticles because they are charged and therefore interact through the Coulomb interaction. (14. but of course both total momentum and total energy are conserved in the scattering event.39) accounts for. On the contrary when the particle is scattered backward. The Coulomb interactions thus introduces a two-particle scattering where momentum and energy are exchanged. But we never included scattering of quasiparticle on other quasiparticles. cos θ ≈ 1. (11.e.42)
where n0 is the equilibrium distribution function. it is tempting to think of the τ0 as the time for scattering out of the state k. This would however only give the ﬁrst term in the right hand side of Eq.25) is indeed applicable to each quasiparticle separately. and the analysis that was applied in the ﬂuid dynamical picture in Eq. which was also calculated in Eq. i. the life time of the state k. i. (14. Often one uses an even simpler approximation for the collision term.14.25) found by the simpliﬁed analysis. is given by the ﬁrst Born approximation 1 τlife = nimp V Wk . In this approximation the collision is replaced by ∂ nk ∂t =−
collisions
nk − n0 k . If two particles in states |kσ. This k approximation in fact gives the correct answer if the relaxation time is identiﬁed with the transport time τ0 = τ tr . The interaction between the particles is screened by the other particles and we should use the RPA result for the interaction.3. the state k is destroyed but the momentum is almost conserved and such a process therefore does not eﬀect the conductivity. The life time. SEMI-CLASSICAL TRANSPORT EQUATION
243
which agrees with Eq. (14. corresponding to a large momentum relaxation. The reason that the two approaches give the same result is that we can treat the quasiparticle as independent. Then we included some ﬁnite life time induced by scattering against impurities. τ0
(14.25).1
Finite life time of the quasiparticles
Above we ﬁrst assumed that the quasiparticles have an inﬁnite life time.
k
(14.e. If the quasiparticle scatters forward.
14. there is a large change in momentum. At ﬁrst sight.

At T = 0 this gives the condition that ξk + ξk − ξk+q > 0. i. (14.
This is a very important result because it tells us that the life-time of the quasiparticles diverges as we approach the Fermi level and thus the notion of quasiparticles is a consistent picture. 14.244 rule
CHAPTER 14. (14. The result for the “life-time” τk of the state |kσ is then given by
screened interaction
Γk+qσ. To ﬁnd that we must multiply Γ with the probability that the state |k σ is occupied and that ﬁnal states are unoccupied and sum over all possible k and q.10) holds as long the temperature is much smaller than the
. We look at situation with a particle above the Fermi surface ξk > 0.
scattering into state k
(14.44) where W RPA (q) is the RPA screened Coulomb interaction. kσ
2
1 2π = 2 2 τk V
spin
kq
W (q) RPA (q. see also Fig.k −qσ . k − q. 0) ε
2
δ ξk + ξk − ξk+q − ξk −q
× nk nk 1 − nk+q 1 − nk −q
scattering out of state k
− 1 − nk 1 − nk nk+q nk −q . FERMI LIQUID THEORY
δ(ξk +ξk −ξk+q − ξk −q ).46)
2
. From this rate we can obtain the total rate for changing the state of a given quasiparticle in state |kσ by the Coulomb interaction. Suppose then we have integrated out the angle dependence. At ﬁnite temperature the typical excitation energy is kB T and ξk is replaced by kB T 1 ∝ T 2 . k −qσ |W RPA (q)|k σ . for T > ξk .47) τk The conclusion from this analysis is: the life-time of the quasiparticles based on Fermi’s golden rule diverges at low temperatures and therefore the condition for the adiabatic approach expressed in Eq. We then have 1 ∼ |W |2 [d(εF )]3 τk = |W |2 [d(εF )]3 = |W |2 [d(εF )]
0 ∞
dξ
−∞ 0 0
dξ Θ(ξk + ξ − ξ )
−∞ 2 3 ξk
dξ ξk + ξ Θ(ξk + ξ ) for T < ξk .45)
The expression (14. A simple phase space argument gives the answer. (14. Then we are left with two energy integrals over ξk ≡ ξ < 0 and ξk −q ≡ ξ > 0.45) can be evaluated explicitly for a particle in state k added to a ﬁlled Fermi sea. which takes care of the delta function.k σ .1. But for now we just want the energy dependence of the life-time. (14. or k + q. nk = 1 and np = θ(kF − p) for p equal to k .kσ = 2π
k+qσ.e.

1
Microscopic basis of the Fermi liquid theory
Renormalization of the single particle Green’s function
The Fermi liquid theory relies on the assumption that the excitation created by adding a particle to the system.14.
14. As illustrated in Fig. Since the particle can only loose energy.4. which in general can be written
. the quasiparticle picture therefore has real physical meaning. because of the Pauli principle the ﬁnal states of both particles have to lie outside the Fermi surface and therefore the phase space volume for the ﬁnal state k + q (white area) and for the initial state k (gray area) both 2 scale with ξk giving rise to a maximum total phase proportional to ξk . This is what we are going to show in this section and thereby give a microscopic foundation of the Fermi liquid theory. The dashed circle indicates the energy of the initial state. Both momentum and energy have to be conserved.4
14. This together with the Pauli principle cause the phase space available for the scattering to be very limited. The function that measures precisely the density of states for adding particles is the retarded Green’s function GR .1: The two-particle scattering event that gives rise to a ﬁnite life time of the quasiparticles. If the retarded Green’s function of the interacting system turns out to be similar to that of free particles. the condition in fact holds for even moderately elevated temperatures. 14. We consider the one-particle retarded Green’s function. These were the quasiparticles. the other particle which is scattered out of state k can only gain energy.4. can be described by a free particle with a long life time. MICROSCOPIC BASIS OF THE FERMI LIQUID THEORY
245
−
ξ
Figure 14. which is illustrated on the right hand ﬁgure. Because the Fermi energy in for example metal is in general a fairly large energy scale.1 the physical reason for the smallness of the scattering rate is that although the Coulomb scattering matrix elements are big there is not much phase space available for scattering due to the Pauli principle. Furthermore. Fermi energy.

The renormalized Fermi wave number k ˜ by the condition that the real part of the energy vanishes ξkF + Re Σ(kF . which leads to GR (k. 0) = 0. where the eﬀective mass by Eq.52) (14. ω)] − i Im ΣR (k. 14. ω) =
CHAPTER 14.48)
where ξk = k 2 /2m −µ is the free electron energy measured with the respect to the chemical potential µ.3. ω) . 0) ˜ ∂k kF . FERMI LIQUID THEORY
1 . meaning ˜F is deﬁned close to the renormalized Fermi-energy. 0))k=kF . ω) is not expanded because we look at its form later. At small ˜ ˜ ˜ energies and for k close to kF .54)
˜ k=kF
In Sec. and where ΣR (k. (14. ω) ≈ ˜ ω − ω∂ω Re ΣR − (k − kF )∂k (ξ + Re ΣR ) − i Im ΣR i 2˜k (ω) τ
−1 −1
˜ ≡ Z ω − ξk + where Z −1 = 1 −
(14. important conclusions can be drawn from the ﬁrst non-trivial approximation.55) (14. we can expand (GR )−1 in powers of k − kF and ω.246 as GR (kσ. 12 was shown to give the exact answer in the high density limit. which of course is not doable in the general case. ω) . ω − ξk − ΣR (kσ. If this also ˜ holds here the spectral function therefore has a Lorentzian shape near k = kF . (14. The ˜ eﬀective energy ξk is usually expressed as ˜ ˜ ˜ ξk = (k − kF )kF /m∗ . ω) is the irreducible retarded self-energy. (14. ∂ω ω=0 ∂ ˜ ˜ ξk = (k − kF )Z (ξk + Re Σ(k. ω)
(14. ω). To calculate the self-energy we should in principle include all possible diagrams. namely the RPA which in Chap. For a very
.53)
The imaginary part of ΣR (k. Let us ﬁrst write the general form of GR by separating the self-energy in real and imaginary parts GR (kσ.49)
˜ We then anticipate the quasiparticle picture by looking at k-values close to the kF . ω) = 1 ω − [ξk + Re ΣR (k. τk (ω) ˜
∂ ˜ Re Σ(kF .51) (14. Fortunately.52) is seen to be m =Z m∗ 1+ m ∂ Re Σ(k. ˜ ∂k
(14.50)
1 = −2Z Im ΣR (k.1 we saw that the life-time goes to inﬁnity at low temperatures.

2π
is not fulﬁlled by Eq. (14. (14. the quasiparticle peak cannot be the whole story. that has an integrated weight given by 1 − Z.56).14. contains more complicated many body excitations not describable by a free electron like peak. (14. Typically Z is found from experiments to be between 0. It contains a distinct peak. the Green’s function and the spectral function ˜ has a sharp peak at ω = ξk . see Exercise 13. 14. ω) ≈ 2πZδ(ω − ξk ). which we denoted A . which is identiﬁed with the quasiparticle. ω). where rs = (3/4πa3 n)−1/3 is the parameter often used to parameterize the density of electron 0 gases. and hence Eq. (14. The peaked spectral function therefore resembles that of a free gas and the pole is identiﬁed as the quasiparticle that was deﬁned in the Fermi liquid theory. ω) = −2Im GR (k. because the general sum rule
∞ −∞
dω A(k. ω) = 2πZ δ(ω − ξk ) + A (k. the book by Mahan.7 and 1 for rs < 3. The constant Z is called the renormalization constant and it is a measure of the quasiparticle weight. This part called Aqp however only carries part of the integrated spectral weight and the rest must therefore be contained in the background function A stemming from other types of excitations. See Fig. However. Therefore we instead write ˜ A(k. small imaginary part we could namely approximate Im ΣR ≈ −η. The renormalization constant shows up for example in the distribution function n(k). ω) = 1.57)
DC C C C E E E C E E E 9 ! @ ' A¦ ¥ £¤ ¡¢ ¤ §
0 () E FE E E E E 3% E E E E 8 4 & 2 ¥1 0 (' ¡¤ ) 6 # ¦ ¥ £¤ ¡¢ ¨© B 666 $! # %& ! 6676 56 "¦ ¥ £¤ ¡¢ ¨© ¤ ¥ £¤ ¡¢ ¥ £¤ ¡¢ ¨© §¦ ¥ £¤ ¡¢
. For a discussion on the measurements of Z using Compton scattering see e.50) gives ˜ A(k. the integral only amounts to Z.56)
This shows that with a small imaginary part.2.4.2.g.2: The spectral function A(k.58) where the remaining contribution A not associated with the pole. MICROSCOPIC BASIS OF THE FERMI LIQUID THEORY
247
Figure 14. ω) as resulting from the analysis of the RPA approximation. There must be another part of the spectral function.
¤
(14. where the jump at the Fermi level is a direct measure of Z.

we are allowed to get the
.3) C = C1 + C2 then we include all the Matsubara frequencies except the one in origin (note that the points shown in Fig. The integrand in analytic everywhere but in z = ξk+q − ikn and for z purely real. ikn + z). σ. however. and this is equivalent to stating that the integration are replaced by the principal part. ikn + ω) − (η → −η) + iη) W (q) nB (ξk+q − ikn ) RPA . 14.12). Therefore we include a loop around the origin so that the contour C = C1 + C2 + C3 shown in Fig. If we therefore make a contour which is like the one in Fig.4. ikn ) = − σ 1 dz nB (z) 2πi V G0 (k εRPA (q. σ. In the Matsubara frequency domain it is given by
.3 is now seen to cancel parts of the counters C1 and C2 so that they are modiﬁed to run between ] − ∞. −δ] and [δ. which we therefore have to subtract again. 14. We now get ΣRPA (kσ. ξk+q − ikn )
In the last term we should use that nB (ξk+q − ikn ) = −nF (ξk+q ) because ikn is a fermion frequency.248
CHAPTER 14.61) ε (q. the RPA self-energy Eq.3 includes all boson Matsubara frequencies z = iωn .59)
where W/εRPA is the screened interaction. in
ΣRPA (kσ. ∞[ only. ikn + iωn ). εRPA (q. FERMI LIQUID THEORY
We still need to show that the assumption of a large τk is valid and we now turn to evaluating the imaginary part of the self-energy. (14. ikn ) = − σ
1 β
iωn
1 V
q
W (q) G0 (k + q. As usual we perform the Matsubara summation by a contour integration ΣRPA (kσ. As seen in Fig.3) are the fermionic Matsubara frequencies). ω
q
W (q) G0 (k + q.3 we. (13.60)
C
q
where C is a suitable contour that encloses all the bosonic Matsubara frequencies z = iωn . z) W (q) + q. when letting δ → 0+ . (10. iωn )
(14. (10. 14. The small loop C3 shown in Fig. (14.2
Imaginary part of the single particle Green’s function
We base our analysis on the most important diagram.
14. σ. Now that we have performed the Matsubara sum. ikn ) = − 1 V
∞
P
q −∞
dω nB (ω) 2πi
× + 1 V
εRPA (q. also enclose the pole in z = ξk+q − ikn .

0 The imaginary part of the polarization function follows from Eq. ξk+q − ε + iη) (14.62) ε (q. ε + ω) = πδ(ε + ω − ξk+q ) and then performed the 0 ω-integration. (14. V q ε (q. σ.63) where we used that − Im GR (k + q. ω + iη) = εRPA (q.60). ω
× (2i) Im − 1 V
q
W (q)GR (k + q.64)
.14. ε) = − 1 V
∞
P
q −∞
dω nB (ω) 2πi 1 εRPA (q. ξk+q − ξk ). ε) = [nB (ξk+q − ε) + nF (ξk+q )] Im RPA . σ. retarded self-energy by the substitution ikn → ε + iη which leads to ΣRPA. ξk+q − ξk ) = 0 2π V
nF (ξk ) − nF (ξk −q ) δ(ξk − ξk −q − ξk+q + ξk ) (14. ξ ε k+q − ξk )
2
(14. ε + ω) 0 + iη) W (q) nF (ξk+q ) RPA .3: The contour C = C1 + C2 + C3 used for integration for a the Matsubara sum that enters the RPA self-energy in Eq. while the that of G0 is the white dot. ξk ) = τk V nB (ξk+q − ξk ) + nF (ξk+q )
q
× Im χR (q. we evaluate the imaginary part in ε = ξk and ﬁnd 1 −2 = −2 Im ΣRPA (kσ. The contour C3 which picks up the contribution from the pole z = 0 cancels the parts of C1 and C2 given by the small loops. ω − iη) . MICROSCOPIC BASIS OF THE FERMI LIQUID THEORY
249
Figure 14.65)
k
§ £ ¡ ©¨¦ ¥ ¤¢ ¤¢ ¡ " # !
W (q) RPA. (13. The poles from the boson frequencies are shown by black dots. The imaginary part of the self-energy becomes 1 W (q) Im ΣRPA (kσ.20) Im χR (q. (14. Since we are interested in the case where a particle with ξk is scattered.4. ξk+q − ε − iη)
∗
because εRPA (q.R (kσ.R (q.

Ferrell.
5 6
J. All together this allows us to write the occupation factors in Eq.67)
At low temperature the ﬁrst term is due to the energy conservation condition non-zero for ξk > 0. The analysis indeed conﬁrmed the physical picture put forward by Landau in his Fermi liquid theory. (13. 812 (1958). We have thus veriﬁed that the imaginary part of the retarded Green’s function indeed goes to zero.250
CHAPTER 14. 14. J. The derivation is rather lengthy and we do not give it here.66) as −nF (ξk )[1 − nF (ξk+q )] 1 − nF (ξk −q ) − [1 − nF (ξk )] nF (ξk+q )nF (ξk −q ) (14. (14. Hence after integration.1 will have even more constrains on the energies.R (q. The ﬁrst term thus corresponds to the scattering out term in Eq. (14.
combined with ξk − ξk+q = −ξk + ξk −q . At least when employing the RPA approximation for the self-energy. (14. This concludes our analysis of the single particle Green’s function.45).20)) and when this is inserted back into Eq. This was done by Luttinger6 who proved that the imaginary part of any diagram for the self-energy goes to zero as ξ 2 or faster. Phys. An explicit calculation of Eq. we obtain 1 4π = −2 Im ΣRPA (kσ. If we furthermore approximate ξk − ξk+q ≈ 0 in εRPA we now see that the life time in Eq. A. (14.45).68) τk 128 εF Going beyond RPA. (14. (14. It is however not hard to imagine that more complicated scattering events than the simple one depicted in Fig. they will result in higher powers of ξk . (14. but the RPA approximation in Chap.
. FERMI LIQUID THEORY
(here we shifted k → k − q as compared with Eq.66) is equivalent to the Fermi’s golden rule expression Eq. Rev. while the second term corresponds to the scattering in term. while the last term is non-zero for ξk < 0. For the second term we use the obvious identity nF ( 1 )[1 − nF ( 2 )] − nF ( 2 )[1 − nF ( 2 ] = nF ( 1 ) − nF ( 2 ) and nF (− ) = 1 − nF ( ).45).66) was done by Quinn and Ferrell5 who got √ 2 3π ξk 2 1 = ωp . J.66)
Let us study the occupation factors in this expression and compare with the Fermi’s golden rule expression Eq. it can in fact be shown that the imaginary part vanishes to all orders in the interaction. 121. For the ﬁrst term in the ﬁrst parenthesis we use the identity nB (
1
−
2 )[nF ( 2 )
− nF ( 2 )] = nF ( 1 )[1 − nF ( 2 )]. ξ ε k+q − ξk )
2
δ(ξk − ξk −q − ξk+q + ξk ). Luttinger. (14. Quinn and R. (14. 942 (1961). Phys. 12 was shown to be exact in the high density limit.M. Rev 112. ε) = − τk V [nB (ξk+q − ξk ) + nF (ξk+q )] nF (ξk ) − nF (ξk −q )
qk
×
W (q) RPA.64).

69)
This equation is extremely useful since it in many situations gives a suﬃciently accurate description of the physics. The conductivity is as we remember from the Kubo formula related to the current-current correlation function.3
Mass renormalization?
In the previous section we saw how the assumption of weakly interacting quasiparticles was justiﬁed by the long life time of the single particle Green’s function. 15 we shall see an explicit example of this by calculating diagrammatically the ﬁnite resistance due to impurity scattering starting from the fully microscopic theory. One can include both electric and magnetic ﬁelds driving the system out of equilibrium.4.e. This approach is based on the Boltzmann equation ˙ ∂t (nk ) + k ·
k nk
+ vk ·
r nk
=
∂nk ∂t
. The result
. For a translation-invariant system the quantum numbers are k and σ. σ = ne2 τ tr /m. Landau’s phenomenological theory was shown to be justiﬁed by a rigors microscopic calculation. using the random phase approximation result for the self-energy. On the right hand side of Eq. Miraculously. provided that the corresponding operators also commute with the full Hamiltonian. In Chap. i. This seems to contradict the postulate of the Fermi liquid theory that the current of the quasiparticles is independent of interactions. These particles we call quasiparticles and they can be labelled by the same quantum numbers of those of the non-interacting system. the interacting system of particles can be described by a gas of non-interacting particles. it is given by k/m and not k/m∗ . (14. The calculation has to be done in a consistent way such that the diagrams included in the irreducible self-energy is also included in the diagrams for the two-particle correlation function. When the same type of diagrams are included both in the self-energy and in the lines that cross the two-particle “bubble” then the mass renormalization exactly cancels.69) we have included colli˙ sions due to impurities and particle-particle collisions . One can also include for example particle-phonon scattering in solids and thus explain the temperature dependence of the diﬀerent transport coeﬃcients. i. How come the renormalized mass m∗ appears in the Green’s function whereas the physically observable conductivity contains the bare mass m? The answer to this question in found by studying how the conductivity is calculated diagrammatically.
collisions
(14. The theory is valid whenever perturbation theory is valid. On long length and time scales we can use a semi-classical approach to study various properties. We also found that the eﬀective mass of the quasiparticle was renormalized due to the interactions. OUTLOOK AND SUMMARY
251
14. See also Exercise 14. The bare velocity of the quasiparticles was important for obtaining the Drude formula for the conductivity.4
14. The driving ﬁelds enter through the Lorentz force as ˙ p = k = (−e)(E + v × B). It has been widely used to explain numerous transport phenomena in gases and solids. when the interaction does not induce a phase transition.e.5
Outlook and summary
We have developed the semi-classical Fermi liquid theory of interacting particles.14.5.

T 2 ).70) This explains why the Fermi liquid theory works: when the imaginary part goes to zero the single particle Green’s function is identical to that of a free particle.
. (14. ε) ∝ max(ε2 . FERMI LIQUID THEORY
of this analysis was that even in the presence of interactions does the Fermi surface persist and near the Fermi surface the imaginary part of the single particle Green’s function rapidly vanish as Im ΣR (kF .252
CHAPTER 14.

and an extended research was initiated on the role of quantum coherence in transport properties. which is the topic in Chap. In three dimensions. we therefore start by deriving the semiclassical result. In one dimension. besides giving a ﬁrst principle justiﬁcation of the Boltzmann equation. 13 the conductivity was calculated within the Boltzmann equation approach. 15. The leading quantum correction is precisely the weak localization eﬀect. We now rederive the Boltzmann equation result starting from a microscopic quantum approach. was discovered in small (∼ µm) phase-coherent structures. at least in two dimensions. The basic physics of impurity scattering was discussed in Chap. which is due to interference between time reversed paths. the situation is again diﬀerent in that at some critical amount of impurity scattering there exists a metal-insulator transition known as the Anderson localization. the so-called universal conductance ﬂuctuations. In this chapter we deal with extended systems and discuss the most important disorder-induced quantum corrections. and it can therefore not be explained semiclassically. where we saw how the single-particle Green’s function acquired a ﬁnite life time after averaging over the positions of the impurities. Weak localization involves coherent scattering on many impurities. 10.Chapter 15
Impurity scattering and conductivity
We now return to the problem of calculating the resistance of a metallic conductor due to scattering against impurities. In Chap. is that it can be extended to include correlation and coherence eﬀects that cannot be described in the semiclassical Boltzmann approach. 253
. In 1979 the weak localization correction to resistivity was observed experimentally in large 2D samples at low temperatures. Then we go on to include the quantum mechanical eﬀect known as weak localization. This is however outside the scope this book. In order to get familiar with the techniques. To understand these smaller systems one must take into account the ﬁnite size of the conductors. The advantage of this microscopic approach. things are more complicated because there all states are localized and one cannot talk about a conductivity that scales in a simple fashion with the length of the system. This discovery started the modern ﬁeld of mesoscopic physics. It was explained theoretically later the same year. A few years later another low-temperature interference eﬀect.

254
CHAPTER 15. αβ ω (15. ω). IMPURITY SCATTERING AND CONDUCTIVITY
Based on the physical picture that emerged from the Fermi liquid description in Chap. As usual we calculate the retarded function starting from the corresponding Matsubara function. Furthermore. because having a electric ﬁeld E(q.ω) where ω = 0 and q ﬁnite is unphysical. ω). term in Eq. r . σαβ = σ δαβ . since it would give rise to an inﬁnite charge built up. we shall see that in order to obtain meaningful results. denoted σ . These corrections cannot be treated evaluating only singleparticle Green’s functions. 10. ω) = − e2 Im ΠR (r. (6.e. In the computation we can choose α to be the x direction. τ ) . (6. V (15.25) in terms of the retarded current-current correlation function Eq. ω) = ω αβ The dc-conductivity is then found by letting1 q → 0 and then ω → 0.
15. the conductivity tensor is isotropic and therefore diagonal. r .3)
In this chapter we consider only homogeneous translation-invariant systems.25) drops out of the real part. ω). For a translation-invariant system we consider as usual the Fourier transform ie2 R Π (q. 14.1)
Note that the last. The Matsubara current-current correlation function is Πxx (q. so-called “diamagnetic”.4. The dc-response at long wavelengths is thus obtained as Re σαβ = −e2 lim lim
ω→0 q→0
1 Im ΠR (q. This means that we shall see how the mass renormalization discussed in Sec. (6.
.25). i.3 is cancelled out. In the second part of the chapter we include electron-electron interactions together with impurity scattering and explicitly demonstrate that the non-interacting approximation is valid.2) σαβ (q. (15. and therefore we take the real part of Eq. 13. (6. αβ ω
(15.4)
1 If in doubt always perform the limit q → 0 ﬁrst. τ − τ ) = − 1 Tτ Jx (q.25) Re σαβ (r. They are thus genuine two-particle correlation eﬀects. In particular we have no magnetic ﬁeld and take A = 0. so-called “paramagnetic”.1
Vertex corrections and dressed Green’s functions
Let us start by the Kubo formula for the electrical conductivity tensor σαβ given in Eq. τ )Jx (−q. (6. term of σ in Eq. Here we shall only look at the dissipative part of the conductivity. which can be described by diagrams where interaction lines “cross” the bubble diagrams. In the following we therefore only include the ﬁrst.26). we assume in the ﬁrst part of this chapter that we can describe the electrons as non-interacting. Note that the system is translation-invariant even in the presence of impurities after performing the position average described in Chap. it is absolutely imperative to include vertex corrections to the current-current correlation bubble diagrams.

7)
iql
1 β
Jx (q. iqn ) = − 1 V
β 0
255
d(τ − τ )eiqn (τ −τ
)
Jx (q.
iqm
This we conveniently rewrite using the four-vector notation q = (iqn . τ ) . iqn ) = − 1 V
β 0
d(τ − τ )eiqn (τ −τ
)1
β
(15. σ
˜ k σ
(15. 10 and (15. iql )Jx (−q. We can now draw diagrams for the current-current correlation function using the Feynman rules. however. q) ˜ Πxx (˜) = − q 1 Jx (˜)Jx (−˜) . τ ) in terms of Jx (q.6) The integration with respect to τ leads to iqn = iql . iqn ) and obtain Πxx (q. τ )Jx (−q.9)
which we draw diagrammatically as a vertex ˜ k Jx (˜) = q ˜ ˜ k+ q The vertex conserves four-momentum.
(15. and whence Πxx (q. VERTEX CORRECTIONS AND DRESSED GREEN’S FUNCTIONS In the frequency domain it is Πxx (q. because τ > τ . Vβ (15. −iqn ) . here we include both the impurity lines from Chap. and thus has the momentum q = (iqn .15.10)
.1.8)
In order to begin the diagrammatical analysis we write the current density Jx (˜) in q four-vector notation
β
Jx (˜) = q
dτ eiqn τ
0
1 1 2m V
(2k + q)x c† (τ )ck+qσ (τ ) kσ
kσ
1 1 = 2m β
ikn
1 V
(2k + q)x c† (ikn )ck+qσ (ikn + iqn ). we must have iqn = −iqm . iqm ) e−iql τ e−iqm τ . Finally. q) ﬂowing ˜ out from it to the left. q q Vβ (15. The procedure is analogous to that for the charge-charge correlation function in Chap. iqn ) = − 1 Jx (q. kσ
kσ
1 11 ≡ 2m β V
˜ ˜ ˜ (2kx +qx )c† (k)cσ (k + q ).5)
where the time-ordering operator Tτ is omitted. since the result cannot depend on τ . 12. iqn )Jx (−q. We can now express Jx (q.

(15.11) and are left with bubble diagrams where the only interaction and impurity lines to be drawn are those connecting the lower and upper electron Green’s functions.11)
the Coulomb interaction lines Πxx (˜) = q
≡
+
+
+
+
+
+
+
+
+
+
We can perform a partial summation of diagrams to all orders by replacing each Green’s function G0 by the full Green’s function G.
.256
CHAPTER 15..
Πxx (˜) = q
+
+
+
Here the double lines represent full Green’s functions expressed by Dyson’s equation as in
+ +
(15. In doing so we have in one step resummed Eq.11) then becomes
+ + + + + + + + + + +
+ ... (15. We obtain (15. 12.12)
+ . Eq.. IMPURITY SCATTERING AND CONDUCTIVITY from Chap.

19)
257
˜ G(k) = =
is the irreducible self-energy. For example in the case where we include where Σirr = impurity scattering within the ﬁrst Born approximation and electron-electron interaction in the RPA approximation.12).15.14)
where RPA means the following screening of all impurity and interaction lines
( ) * + . .15) (15. . These diagrams are denoted vertex corrections.
. To obtain a Dyson equation for Πxx we ﬁrst introduce the irreducible line-crossing diagram Λirr consisting of the sum of all possible diagrams connecting the upper and lower fermion line. = = + + + + +
(15.
! # "
+
(15.
Λirr ≡
.16)
The next step is to organize the diagrams according to the lines crossing the bubbles from the upper to the lower fermion line in a systematic way. and 11 in Eq. (15./01234
≡
+ . (12. (15. VERTEX CORRECTIONS AND DRESSED GREEN’S FUNCTIONS Eq. 10 . which cannot be cut into two pieces by cutting both the upper and the lower line just once2 .13)
1BA + RPA:
˜ Σirr (k) =
% & '
≈ + + +
(15. Diagrams of this type are proportional to q −2 and thus they vanish in the limit q → 0.17)
2 We do not include diagrams like the terms 9. the irreducible self-energy is simply
$
˜ ˜ ˜ ˜ = G0 (k) + G0 (k)Σirr (k)G(k).1.

+ . R. k ). Theory of Superconductivity.“in going”). Addison-Wesley (1964). k) − iq · Γ(k + q .B. and Γ is the current vertex function. 12.
(15.20b)
The notation for the arguments of the Γ functions is Γ = Γ(“out going”. 10 and for the case of interacting particles in Chap. q .x (k .x (k+ q .. k+ q ).258
CHAPTER 15. This relation.20a)
˜ ˜ ˜ ˜ ˜ ˜ ˜ ˜ ˜ ˜ ˜˜ ˜ d˜ Λirr (k.3 is derived using the continuity equation. k) = −G −1 (k + q ) + G −1 (k). We have seen examples of how to choose the physically most important self-energies. called the Ward identity. the answer is simply that there is no freedom left in the choice for the vertex function Γ.. which can be read oﬀ from Eq. (15. q (15. k) +
(15. If we include certain diagrams for the self-energy we must include the corresponding diagrams in the vertex function. k) ≡
? @ A
˜ ˜˜ ≡ Γ0.g.
. 2m = +
+ . This follows from a general relation between the self-energy and the vertex function.
where the function Γ0 is the charge vertex function. q )G(k+ q )G(k+ q + q )Γx (k+ q + q . k + q )G(k )G(k + q )Γx (k + q .18)
where the unperturbed vertex is (15.18) ˜ ˜˜ Γx (k+ q .19)
and the “dressed” vertex function is given by an integral equation.x (k. Consequently... IMPURITY SCATTERING AND CONDUCTIVITY
Using Λirr we see that we can resum all diagrams in Πxx in the following way Πxx (˜) = q + + +
=
≡
≡ −
5678 9 :. In the present case once the approximation for Σirr is chosen. The question is now which diagrams to include in Λirr . Schrieﬀer.
3
The Ward identity reads ˜ ˜˜ ˜ ˜˜ ˜ ˜ ˜ iq0 Γ0 (k + q . both for the impurity scattering problem in Chap. For more details see e. k + q ) = 1 (2kx +qx ).<= >
× + + + ˜ ˜ ˜ ˜ ˜ dk Γ0. ˜ ˜ ˜ ˜˜ ˜ ˜ ˜ Γ0.

2
The conductivity in terms of a general vertex function
Having the expressions for both the single-particle Green’s function G and the vertex function Γ. ikn ). RPA (−˜) ε (˜) ε q q
(15. Here we simply follow the rule as dictated by the Ward identity: if an irreducible diagram is included in Σirr the corresponding diagrams should also be included in Λirr . we can obtain from Eq.15.21)
˜ ˜ ˜ ˜ ˜ ˜ ˜ ˜˜ ˜ d˜ W (˜ )G(k + q )G(k + q + q )Γx (k + q + q . the current-current function is Πxx (0. THE CONDUCTIVITY IN TERMS OF A GENERAL VERTEX FUNCTION259 not fulﬁlling this identity is equivalent to a lack of conservation of particles. and furthermore treat the case q = 0. a name which perhaps becomes clear graphically if Eq.14). Therefore we ﬁrst study a summation
.18) for Πxx .25) The Matsubara sum over ikn is performed in the usual way by a contour integration over z = ikn . and if for clarity we consider only the impurity scattering lines: Πxx (˜) = q
EFGH
+ + + 1 β
ikn
+ . This deﬁnition involves a summation over the internal Matsubara frequency. k. and one uses the term “conserving approximation” for the correct choice for the vertex function. (15.
(15.24)
15. k) = Γ0.x (k. For a derivation and discussion see for example the book by Schrieﬀer. iqn ) = − 1 V Γ0. If we drop the four-vector notation in favor of the standard notation. The presence of two G’s in the summand leads to two branch cuts.2. k)G(k.21) for Λirr is inserted into the ﬁrst line of Eq. k + q ). k)+ where
BCD
≈ +
˜ ≡ W..ikn + iqn )Γx (k. (15.x (k + q . ikn )G(k.23)
This particular approximation is also known as the ladder sum.18) a general formula for the conductivity.. one along z = ε and one along z = −iqn + ε. (15. If we consider the ﬁrst Born approximation and RPA for Σirr as depicted in Eq. (15. (15. Therefore a physically sensible approximation must obey Ward’s identity.22) q ˜ q
u(q) u(−q) ˜ q W (˜) = W RPA (˜) + nimp RPA q . we get for Λirr Λirr =
and in this case the integral function for Γ becomes ˜ ˜˜ ˜ ˜˜ Γx (k + q . with ε being real.ikn + iqn .
k
(15.
(15.

we obtain
R S2F (ω) =
dε [nF (ε) − nF (ε + ω)] f AR (ε. we also take the imaginary part as in Eq. Fur∗ thermore.e. and the second argument is retarded. (15. z + iqn ).27) (15. ε + ω) 2πi −∞ ∞ dε − nF (ε)f RR (ε. ε + iqn ) −∞ 2πi ∞ dε nF (ε − iqn ) f (ε − iqn . ε + ω) − f AR (ε. ε + iqn ) − f (ε − iη.260
CHAPTER 15. 2πi
where the integration contour C is the one shown in Fig. (15.28)
+ f AR (ε − ω.1: The contour used in the frequency summation in Eq. ε ) means that the ﬁrst argument is advanced. ε + ω) − nF (ε + ω)f AA (ε. of the form S2F (iqn ) = 1 β f (ikn .26) (15. Since f AA = f RR . If we shift the integration variable ε → ε + ω in the two last terms. 2πi −∞
∞
Since we are interested in the low frequency limit. ε) . we expand to ﬁrst order in ω. we
©§¥ £ ¨ ¦ ¡ £ ¡ ¤¢
(15. ε + ω) .29)
. i. IMPURITY SCATTERING AND CONDUCTIVITY
Figure 15. ε + iη. ε) − f AA (ε − ω. (15.
with the convention that f AR (ε. and so on. ikn + iqn )
ikn
=−
C
dz nF (z)f (z. ε + ω) 2πi (15. − −∞ 2πi
∞
At the end of the calculation we continue iqn analytically to ω + iη. and ﬁnd
R S2F (ω) = − ∞ −∞
dε nF (ε) f RR (ε. ε − iη. ε + iη) − f (ε − iqn .3).1) made of three contours leading to four integrals over ε S2F (iqn ) = − dε nF (ε) f (ε + iη.26). ε − iη) .

33).33) by the ﬁrst term only. In this case. (15.3. In the following we consider various cases.ikn + iqn . 0. one must look at the speciﬁc physical cases and then solve for the vertex function satisfying Eq. ε) − f RR (ε. The RPA part of the self-energy in Eq. 0)ΓRA (k.3) one obtains
R Im S2F (ω) =
(15. (15. again taking q = 0. THE CONDUCTIVITY IN THE FIRST BORN APPROXIMATION ﬁnd
R Im S2F (ω) = −ω Im ∞ −∞
261
dε 2πi
−
∂nF (ε) ∂ε
f AR (ε.33)
where we have included a factor of 2 due to spin degeneracy. (15.31) − ∂ε and hence ω Re f AA (0.3
The conductivity in the ﬁrst Born approximation
The conductivity was calculated in Sec. (15. 0.
(15.32)
Re σxx = 2 Re
e2 1 2π V
Γ0. the ﬁrst diagram in Eq. Later we discuss what happens if interactions are included. 14. ikn ). x
(15. The starting point in this section is non-interacting electrons scattering on impurities.
15. 0)GR (k.15. 0)ΓAA (k.34)
× G(k + q .21).22) becomes Γx (k.ikn + iqn . where τ is the scattering time for impurity scattering. we may replace the general formula in Eq. ε) . (15.x (k. 0) . 0) x
k
− GA (k. i. 2π By applying this to Eq. the integral equation Eq.e.x (k. To proceed further. (15. 0) − f AR (0. The semiclassical approximation is similar to the ﬁrst Born approximation in that it only includes scattering against a single impurity and neglects interference eﬀects. Therefore we expect to reproduce the semiclassical result if we only include the ﬁrst Born approximation in our diagrammatical calculation. 0) . ikn ) = Γ0.
. We will consider only cases where the disorder is weak and for this case it is shown in the next section that the product GR GA exceeds GA GA by a factor of order 1/τ EF . k. k) GA (k. The vertex function is now solved using the ﬁrst Born approximation. This is how far one can go on general principles. 0)GA (k.14) is not included in this section.30)
At low temperatures. k.20b) and insert the result into (15.25) and then inserting into Eq. Hence in the weak disorder limit. (15. k) + 1 V nimp uRPA (q ) G(k + q . we can approximate the derivative of the Fermi-Dirac function by a delta function ∂nF (ε) ≈ δ(ε) (15. k.3 using a semiclassical approximation for the scattering against the impurities. k+q .ikn + iqn )Γx (k + q . ikn )
q 2
(15.

and subsequently taking the dc-limit ω → 0. For the case of weak impurity scattering the scattering rate τ −1 is small. −2 Im ΣR (k. Remembering that Γx is a component of a vector function Γ and that the unperturbed vertex is Γ0 (k.34). which also appear in Eq. ε)GR (k. (15. One immediately sees that the small factor nimp tends to kill the sum. ε) deﬁned as Γ(k. ε) = k γ(k. as we learned from the Ward identity.262
CHAPTER 15.38)
. ε)
k 2
k · k RX γ (k . We arrive at γ RX (k. k. ikn ) = 1 + 1 V nimp uRPA (k − k) G(k . whence the spectral function is approximately a delta function. (15. ε) 2
≡
(15. Note that there is no internal Matsubara sum because the impurity scattering conserves energy. (15. ε)GR (k. In doing so we in fact use that the system is isotropic which means that only the vector k can give the direction. (15. multiplying by (1/k 2 )k·. and shifting the variable q to q = k − k. ε) = GR (k. The lesson we learn here is that we have to be rather careful with products of Green’s function carrying the same arguments. (15. (15.35)
In the formula Eq.ikn + iqn . also those obtained in the ﬁrst Born approximation. k . In the case of small nimp we therefore get GA (k.33). The Green’s functions G are.35) by letting iqn → ω + iη and ikn → ε ± iη. ε) ≈ 1. ε) 1 1 = Im R (k. (15.37)
where A = −2 Im GR is the spectral function. and where as before the life-time τ is deﬁned by τ −1 = −2 Im ΣR (k. Let us look more carefully into the products GA GR and GR GR .33) for the conductivity both ΓRA and ΓRR appear (or rather x x = (ΓAA )∗ . because in the limit of small nimp . ε)GR (k . It is immediately seen that this is a consistent solution for the imaginary part of both γ RA and γ RR but it turns out that for the real part of γ RA a factor 1/nimp is contained in the Green’s function.ikn + iqn ) ΓRR x
k · k 1BA γ (k . This ﬁrst combination is GA (k. ε) 1 ≡ A(k. ε). and the product of two delta functions has to be deﬁned with care.ikn + iqn . Im GX tends to a delta function. ikn ). ε) ≈ τ 2πδ(ε − ξk ). we deﬁne for convenience a scalar function γ(k. ε)
2 1 ε − ξk − ΣR (k. When inserting this into Eq. ε)/m.23) has been inserted and where uRPA = u/εRPA . k) = k/m. k2
(15. ε). ε) Im Σ ε − ξk − ΣR (k. 0) to zero. Since we do not expect the dynamical screening to be important for the elastic scattering.36) k2
where X = A or R. ikn )
k 2
× G(k . ε) ≡ τ A(k. we set the frequency in εRPA (q. IMPURITY SCATTERING AND CONDUCTIVITY
where the second term in Eq. which we obtain from x Eq. we get γ(k. They satisfy two diﬀerent integral equations. and in the weak scattering limit one should expect the solution of this equation to be simply γ RX (k. ε) = 1 + 1 V nimp uRPA (k − k) GX (k . ε).

In the following we therefore approximate τ with the ﬁrst Born approximation life time τ0
−1 τ −1 ≈ τ0 ≡ 2πnimp k
|u(k − k )|2 δ(ξk − ξk ).41)
The remaining problem is to ﬁnd γ RA (k.44)
. 0. k.42)
Since this equation has no dependence on the direction of k.36) is ﬁnite in the limit nimp → 0. 0. THE CONDUCTIVITY IN THE FIRST BORN APPROXIMATION
263
Because τ ∝ n−1 . ε).3. k2
(15.39)
The last term clearly goes to zero when τ is large and when A is approximated by a delta function. k.15.0) for |k| = kF . k.x (k. (15. ε) ≈ =
R R
ε − ξk − ε − ξk ε − ξk
2
+
(ε − ξk )2 −
2
+
2 i 2τ 1 2 2τ 1 2 2τ 2 + 1 2 2τ
i (ε − ξk ) A(k.43)
1 . But kF is constant. 0) = τ0 γ RA (kF .35) γ RA (k) = 1 + 2π V nimp uRPA (k − k) τ0 δ(ξk )
k 2
k · k RA γ (k ). kF . k)τ0 δ(ξk )ΓRA (k. and since the lengths of both k and k are given by kF . 0) m m m
(15. the product nimp GA GR in Eq. (15. (15. and we get γ RA = 1 + The solution is thus simply γ RA = where λ= 2π V nimp uRPA (k − k) δ(ξk )
k 2
2π V
nimp uRPA (k − k) δ(ξk )
k
2
k·k k2
τ0 γ RA .45)
.33) then becomes Re σxx = 2e2 Re 1 V Γ0.
(15.
(15. 0. The ﬁrst term is a peaked function at ε − ξk = 0. this life time is independent of k. As explained above. we use the ﬁrst Born approximation for the self-energy. 0) x
k
1 = 2e2 τ0 Re V
k
kx e2 n kx δ(ξk ) γ RA (k.
(15.
(15. 1 − λ τ0 k·k = (τ0 )−1 − τ tr k2
−1
(15. γ RA depends only on kF . The solution follows from the integral equation Eq. From these arguments it follows that the terms with GR GR can be omitted and only terms with GR GA are kept. That GR GR is ﬁnite is seen as follows G (k. but the integrated weight is in fact zero as can be checked by performing an integration over ε. imp The combination GR GR on the other hand is not divergent and in fact nimp GR GR → 0 as nimp → 0. ε)G (k. 0. The conductivity Eq.40)
Because all energies are at the Fermi energy.

If an electron interacts with another electron or with a phonon through an inelastic scattering event its energy changes. When inserted back into Eq. (14. m
(15. τ0 (15. and hence the evolution of its phase.
.4 it was hinted that such interference processes are represented by crossing diagrams as in Fig. (11. Here the coherence length means the scale on which the electrons preserve their quantum mechanical phase. and as we know from Chap. spacing 104 -105 atomic lattice spacings. At liquid helium temperature. typical coherence lengths are of the order 1-10 µm. corrections not obtainable in the Boltzmann approach. This is what we expected. At low temperatures the dominant dephasing mechanism is electron-electron scattering. the conductivity formula (15. scattering on diﬀerent impurities can interfere.41) at zero temperature is σ= e2 τ tr 1 m2 V
2 δ(ξk )kx = k
e2 nτ tr . we can go on to calculate various quantum corrections to the Drude formula. IMPURITY SCATTERING AND CONDUCTIVITY
Here the transport time τ tr is deﬁned as τ tr
−1
≡
2π V
nimp uRPA (k − k)
|k |=kF
2
1−
k·k k2
.4
The weak localization correction to the conductivity
The Born approximation includes only scattering on one impurity at a time. The reason is that even the full Born approximation depicted in Eq. which does take into account multiple scattering does so only for multiple scatterings on the same impurity. 13 the scattering rate for these processes is proportional to T 2 . In Sec.6.
(15.44) γ RA becomes γ RA = τ tr . the scale on which the wave function evolves according to the one-particle Schr¨dinger equao tion. As the temperature is lowered we expect quantum mechanical coherence to become more important because the phase coherence length φ increases with decreasing temperature. (15. and its coherence length becomes ﬁnite. i.264
CHAPTER 15. 11.48)
We thus ﬁnd full agreement with the semiclassical result obtained in the previous chapter.5.2 K. Hence φ ∝ T −2 can become very large at suﬃciently low temperatures. Due to these processes the phase of the electron wave acquires some randomization or “dephasing”. 11. In this section we shall study in detail why that is. 4.46)
This expression is precisely the transport time that we derived in the Boltzmann equation approach leading to Eq. and thereby having gained conﬁdence in the mathematical structure of the theory. We saw in Chap.47)
Finally.54). Quantum eﬀects such as interference between scattering on diﬀerent impurities can therefore not be incorporated within the Born approximation scheme. When the coherence length φ exceeds the mean free path imp for impurity scattering. and below. 10 that there was in practice only little diﬀerence between the ﬁrst Born and the full Born approximation.39).e.
15.

2. As in Fig.15. In this regime all kinds of quantum interference processes become important. Interference generally means that the amplitude for two paths t1 and t2 are added as t1 +t2 . so that when taking the absolute square |t1 +t2 |2 = |t1 |2 +|t2 |2 +2|t1 t2 | cos(φ1 −φ2 ). showing that the low-temperature weak localization correction to the resistivity increases logarithmically as the temperature decreases. If the coherence length φ is longer than the mean free path 0 . If we can
£ ¢ ¡§¤ ¢ ¦¡ ¡¥ © ¤ ¢ ¡ £ ¨ ¢£ ¡ @ B A@7 @ 9 78 5 65 4 ¡¤ % $ # "! (')% $& A@7 ¨¤£ 2) 3 0 25 5 41 0
. but now at low temperatures we have added the small but signiﬁcant increase due to the quantum interference known as weak localization. one very important class of interference processes survive φ the self-averaging. 15.2: (a) A sketch of the electrical resistivity ρxx (T ) of a disordered metal as a function of temperature. much more dramatic quantum eﬀects appear in small samples in the so-called mesoscopic regime (see also Chap. (11. as discovered around 1985.1 the linear behavior at high temperatures is due to electron-phonon scattering.2). (b) Experimental data from measurements on a PdAu ﬁlm by Dolan and Osheroﬀ. To picture how averaging over impurity conﬁgurations inﬂuences the interference eﬀects. Only the interference processes which are independent of the impurity positions survive self-averaging. The relative phase φ1 − φ2 determines whether the contributions from the two paths interfere constructively or destructively. 721 (1979).4. but still smaller than the sample size L. Phys. This is because the limit φ L eﬀectively corresponds to averaging over many small independent segments. Lett. 43. that even in the case of large samples. Naturally. 0 L. we follow an electron after it has been scattered to a state with momentum k by an impurity positioned at R1 . THE WEAK LOCALIZATION CORRECTION TO THE CONDUCTIVITY 265
Figure 15. most of the interference eﬀects disappear. (11. the so-called self-averaging illustrated in Fig. the cross-term expresses the interference. shown in Fig.2). and most notably cause the appearance of the universal conductance ﬂuctuations shown in Fig. 7) given by L φ . When the electron hits the next impurity at position R2 it has acquired a phase factor eiφ = eik·(R1 −R2 ) . around 1980 it was found through the observation of the so-called weak localization. In the following we study only the weak localization phenomenon appearing in large samples and not the universal conductance ﬂuctuations appearing in small samples. Rev. However. 11. and it is therefore intuitively clear that these terms vanish when one averages over R1 and R2 . Terms describing interference between the two scattering events will thus contain the factor eik·(R1 −R2 ) .

3: Illustration of the two interfering time-reversed paths discussed in the text. This is indeed possible. Here the external potential is given by the impurity potential.3.50)
GR (r. Ri . 0
i U0 δ(r
− Ri ). r . )U0 GR (Ri .
¥¦¢
¤¢ £ ¡
(15. To study interference eﬀects between scattering at these two impurities we must expand to second order in the impurity potential. Most remarkably. these two paths pick up exactly the same phase factor. and thus their relative phase φ1 −φ2 is always zero independent of the actual positions of the impurities. 15.266
CHAPTER 15. As a consequence there is an enhanced probability for returning to the same point. r . two impurities located at R1 and R2 . 0
Let us look at a speciﬁc process where an electron scatters at. )Uimp (r )GR (r . short range impurities located (15. ) = GR (r.e. ) + 0
i
GR (r. ikn → + iη. First we recall the Dyson equation for the single-particle Green’s functions in an external potential. IMPURITY SCATTERING AND CONDUCTIVITY
Figure 15. we have dr GR (r. Uimp . r . there is a corresponding reverse path which scatters on the same impurities but in the opposite order. i. ﬁnd two paths where the relative phase is independent of the position of the impurities. The key observation is that for each path that ends in the starting point after a speciﬁc sequence of scattering events. and two such paths are shown in Fig. r . 10. hence the name “weak localization”4 . ) + 0 If we for simplicity assume Uimp (r) ≈ at the positions {Ri }. The interesting second order
4 The term “strong localization” is used for the so-called Anderson localization where a metal-insulator transition is induced in three dimensions at a critical strength of the disorder potential. Having realized that the interference between time reversed paths survive impurity averaging. which was derived in Chap. In order to do so we need to identify the corresponding diagrams. ) = GR (r. the cross term would thus survive the impurity average. r . Writing it in the frequency domain and making analytic continuation. ). Thus for two such time-reversed paths there is always constructive interference. r . ). say. and the electrons therefore tend to be localized in space.49)
. r . we have for the retarded Green’s function GR (r. we now want to calculate the resulting correction to the conductivity.

GA
(15.54)
This hints that the interference term coming from time reversed paths can be summed by taking diagrams of this form into account. ω)U0 GR (Q − p3 .51)
These two terms correspond to the transmission amplitudes t1 and t2 discussed above and illustrated in Fig. r . Doing this and averaging over impurity positions R1 and R2 we ﬁnd the quantum correction δ|r|2 to the reﬂection. which we used to derive the Boltzmann equation result. )U0 GR (R1 . ) = GR (r. r . )U0 GR (R2 . R2 . r .52)
Now reﬂection is described by setting r = r . ) 0 0 0 + GR (r. ). )U0 GR (R2 .15. 11. at low temperatures they do play a role as the leading
.61) with the upper lines being retarded and the lower lines being advanced Green’s functions. (15. ) 0 0 0
∗
. r . Nevertheless. THE WEAK LOCALIZATION CORRECTION TO THE CONDUCTIVITY 267 terms (there are also less interesting ones where the electron scatters on the same impurities twice) are GR(2) (r. In k-space one gets δ|r|2
imp
≡ δ|t(r = r )|2 1 ∝ Re 4 V
imp
GR (Q − p1 . R2 .4.3. R1 . )U0 GR (R1 . R2 . the diagram looks like conductivity diagram if we furthermore join the retarded and advanced Green’s function like this GR δ|r|2
imp
=
GA
I
GR GA
GR .4 to be smaller than the Born approximation by a factor 1/kF . r )|2 ∝ Re GR (r. R1 . )U0 GA (p3 . Notice however that contrary to the usual diagram for conductance the Green’s function and the lower and upper branch run in same direction. These crossed diagrams were not included in the Born approximation. R1 . )U0 GR (R1 . )U0 GR (R2 . we set r = r at the end of the calculation. r .
(15. )U0 GR (Q − p2 . ) 0 0 0 × GR (r.5. R2 . and in fact they were shown in Sec. ). R1 . )U0 GR (R2 . )U0 GA (p2 .53)
This formula can be represented by a diagram similar to the last one in Eq. 0 0 0 (15. ) 0 0 0
p1 p2 p3 Q
× GA (p1 . )U0 GR (R1 . First the quantum correction due to interference to the transmission from r to r is δ|t(r. 15. If we however twist the lower branch such that the Green’s function run in opposite directions while the impurity lines cross. The probability for the process is obtained from the absolute square of the Green’s function. and because we want to ﬁnd the correction δ|r|2 to the reﬂection coeﬃcient. 0 0 0
(15.

Let us twist the diagram in Eq. The crossed diagrams we are about to evaluate are also small by the same factor. There is however a trick which allows for a summation just like a ladder diagram. this is a contribution which does not disappear upon self-averaging. We write the current-current correlation function as Π = ΠB + ΠWL where ΠB is the Boltzmann result derived in the previous section. Let us return to the Kubo formula for conductance.e.55)
(15. This reversed
. and where ΠWL (˜) = xx q
+
The full electron Green’s functions in these diagrams are as before the full Green’s function with an appropriately chosen self-energy. (15. (15. This divergence stems from summing the interference of many time-reversed paths. down by a factor 1/kF vF τ0 (see the discussion in Fig. IMPURITY SCATTERING AND CONDUCTIVITY
quantum correction..56) can be written as ˜ ˜ k k WL Πxx (˜) = q C ˜ ˜ ˜ k+q k +q ˜ =− 1 1 2 V2 (2m) ˜ dk
N
M
. This class of diagrams are called the maximally crossed diagrams. they can safely be ignored.268
CHAPTER 15. Most importantly.56) with for example three impurity lines so as to make the impurity lines parallel.58)
˜ ˜ ˜ ˜ ˜ ˜ ˜ ˜ ˜ dk (2kx + qx )G(k)G(k + q )C(k.6). i.
(15. (15.
(15.54) but with more crossing lines. ˜
where the box C is a sum of parallel impurity lines. 11. If we continue this line of reasoning we should include also diagrams where paths scattering on more than two impurities interfere with their time reversed counter parts. We have now identiﬁed which diagrams we need to sum in order to get the leading quantum correction to the conductivity. but as we shall see they nevertheless yield a divergent contribution. analogous to the normal ladder sum of Eq.56)
Then we see that the full series in Eq. However. ΠWL(3) (˜) = q xx
J K L
+ + . and let us sum the maximally crossed diagrams. This sum is diﬀerent from the ladder diagrams that we summed in the Born approximation. k . but now with the fermion lines running in the same direction. since they only give a small contribution.. It is straightforward to see that the corresponding diagrams are of the same type as (15. Since we include crossed diagrams in the vertex function we should in principle also include these in the self-energy. q )G(k )G(k + q )(2kx + qx ).24).57)
(15.

ikn → ε + iη) = 1 . THE WEAK LOCALIZATION CORRECTION TO THE CONDUCTIVITY 269 ladder sum. ikn + iqn )G(p..
(15.4. ikn )G(k .60)
k
V W X Y
C(Q. ikn + iqn )G(p.58) xx ΠWL (0.
. ikn + iqn )
ikn kk
× C(k + k . iqn ) = − xx 1 1 1 2 V2 β (2m) (2kx )G(k. ε − ξk + i/2τ0 (15. which couples two electron lines or two hole lines rather one electron line and one hole line.. ikn ) V
1−
(15. we only study the case q = 0. With these approximations. The solution for the cooperon C is found from the following Dyson-like equation
In order to simplify our calculation. ikn )G(k. It is now straightforward to solve the Dyson-equation for the cooperon ladder and obtain
p W0 G(Q − p. (15. ikn ) =
1 V
Q−p Q−p Q−p + .. while the lower ones carry the frequency ikn .63)
The Green’s function G is here the Born approximation Green’s function which after analytic continuation is GR (k. and denoting k + k ≡ Q the cooperon becomes k
O P Q R S TU
C ≡ + + = + C k Q−p Q−p Q−p C = + + k p p p p p p
+
. (15.59)
(15. ikn )W0 . ikn + iqn )(2kx ). ikn )G(k .64)
where [τ0 ]−1 = 2πW0 d(εF ). ε) = G(k. C. ikn + iqn . It is now simple to ﬁnd the solution for the cooperon C. and furthermore we restrict the analysis to the case of short range impurities so that we can approximate W (q) by a constant. W0 = ni |u0 |2 .. 1 p W0 G(Q − p. ikn + iqn .15.62)
This can then be inserted into the expression for the current-current correlation function ΠWL in Eq.61)
Because the impurity scattering conserves Matsubara frequencies the upper fermion lines all carry the frequency ikn + iqn .
(15. is called a cooperon.

−ξQ−p + i/2τ0 −ξp − i/2τ0 (15. Likewise.33) contributed in the limit of weak scattering and therefore we should replace ikn + iωn by a retarded frequency and ikn by an advanced frequency.e.62) becomes C RA (Q) = ζ(Q) ≡ W0 ζ(Q) . Combining Eqs.64) ζ(Q) becomes ζ(Q) = ni |u0 |2 1 V 1 1 . 0)GA (k . Here small means small compared the width τ0 of the spectral function. we have
WL Re δσxx
e2 = 2× 2π
1 m
2
1 V2
(kx kx )GR (k.270
CHAPTER 15. 0). 0. Instead we evaluate the cooperon as follows. Taking the dc-limit ω → 0 and the low temperature limit T → 0. 0)GA (p. 0). here we cannot replace them by delta functions. After analytical continuation the cooperon in Eq. (15. we study −1 the limit QvF τ0 1 or Q 0 = 1/vF τ0 . However.70)
p
1 (vp · Q)2 . because k and k are connected through C RA (k + k ).69) it follows that C RA diverges in the limit of small Q and small frequency.66) and (15.63) the weak localization correction by the replacements ikn + iqn → ε + ω + iη and ikn → ε − iη. 0)C AR (k + k . IMPURITY SCATTERING AND CONDUCTIVITY
In the previous section we learned that only the GA GR term in Eq. so we need to go to second order in Q ζ(Q) ≈ 1 + ni |u0 |2 1 + ni |u0 | V
2
1 V
p
1 −ξp + i/2τ0 1 −ξp + i/2τ0
3
2
1 −ξp − i/2τ0
−vp · Q+
Q2 2m (15. 0)GA (k. Using Eq. −ξp − i/2τ0
.68) for small −1 Q. 0)GR (k .66) (15.68)
p
To proceed further we must now evaluate the p-sum in ζ(Q).67)
where we have introduced the auxiliary function ζ(Q). by symmetry arguments the term linear in Q vanish. (15. we obtain from (15. we study this contribution by expanding Eq. The dc conductivity is therefore dominated by the contribution from values of Q near zero. 2 + (1/2τ )2 ξ −∞ 0 dξ
∞
(15.69)
where we have used the deﬁnition of the life time τ0 in the Born approximation. V p (15. in which case we have ζ(0) = ni |u0 |2 d(εF ) 1 1 −ξ + i/2τ0 −ξ − i/2τ0 −∞ ∞ 1 = ni |u0 |2 d(εF ) dξ = ni |u0 |2 d(εF )2πτ0 = 1. (15.65) As in the previous section we have factors of GA GR appearing. (15.
kk
(15. Furthermore. We begin by studying Q = 0. i.33). (15. Consequently. (15. 1 − ζ(Q) ni |u0 |2 GR (Q − p. followed by inserting the result into Eq.

and we end up with
ζ(Q) ≈ 1 − where
0
1 Q2 Ndim
2 0
≡ 1 − Dτ0 Q2 .71)
1 2πτ0
dξ
−∞
Q2 vF 2 1 .78)
. 0)GA (k. 0)GR (−k. Ndim
(15. 0)GA (k. δσ W L in Eq. (15.
(15. 0)GA (−k. Closing the contour in the lower part of the complex ξ plan. v F 2 τ0 . 0) =
W0 (1 − Dτ0 Q2 ) W0 1 ≈ .65) becomes δσ
WL
e2 = 2× π
1 m
2
W0 1 τ0 V 2
2 (−kx )GR (k.74)
−1 D being the diﬀusion constant. Now transforming the sum into integrations over ξ and performing the angular integrations. and hence smaller than the width of the 0 spectral function.72)
−1 3 2 To leading order in τ0 . Ndim
(15.76) First we perform the sum over k.4. THE WEAK LOCALIZATION CORRECTION TO THE CONDUCTIVITY 271 where it is indeed seen that the term linear in Q is zero because vp is an odd function of p.
D=
(15. τ0 dominates over τ0 . Since Q < −1 . 0.77)
e2 =− π
kF m
2
2τ0 1 Ndim V
Q<
−1
1 DQ2
(15. we ﬁnd ζ(Q) ≈ 1 + + 1 2πτ0
∞
dξ
−∞ ∞
1 −ξ + i/2τ0 1 −ξ + i/2τ0
3
2
1 −ξ − i/2τ0
Q2 2m (15. we can approximate Q − k by just −k and obtain 1 V
2 kx GR (k. (15.Q<
−1 0
1 GR (Q − k. 0) k. −ξ − i/2τ0 Ndim
where Ndim is the number of dimensions. 2 Dτ0 Q τ0 DQ2
(15.73)
= v F τ0 . With this result for ζ(Q) inserted into (15.73) is only valid for Q 0 . 0)GA (Q − k. 0) = d(εF ) k 2 4πkF 2 kF Ndim ∞
dξ
−∞
1 ξ 2 + (1/2τ0 )2
2
= From this follows δσ
WL
Ndim
3 d(εF )τ0 . DQ2
(15.15.66) we obtain the ﬁnal result for the cooperon
C RA (Q. We emphasize that Eq. 0). we ﬁnd that ζ(Q) ≈ 1 + 2πi 2πτ0 1 i/τ0
2
Q2 + 2m
1 i/τ0
3
Q2 vF 2 .75)
Because the important contribution comes from Q ≈ 0.

Here we choose to do this in a phenomenological fashion instead. Let us suppose that each path in the sum over paths in Eq.83)
−
2π 2 D
tan−1
.81) (15. which does not occur in reality. IMPURITY SCATTERING AND CONDUCTIVITY
We are then left with the Q-integration. 1 − e−γ + e−γ DQ2 τ0 or small γ.e. 0) W0 1 . 0. We can now return to (15. In a real system the electron cannot maintain coherence over arbitrarily long distances due to scattering processes that cause decoherence.79)
It is evident that this integral is divergent in the small Q limit in both one and two dimensions. the function ζ(Q) is changed into ζ(Q. we therefore have C RA (Q. Physically this is because we have allowed interference between path of inﬁnite length. ω) ≈ e−γ 1 − Dτ0 Q2 .61) has a probability of being destroyed by a scattering event and that this probability is proportional to the length of the path. This can be modelled by including a factor e−γ in the impurity potential so that instead of W0 we write W0 e−γ . 3D
. γ = 0 / φ . because paths longer than the sample should not be included. It says that the paths corresponding to a diﬀusion time longer than the phase breaking time cannot contribute to the interference eﬀect.79) and evaluate the integral in one.80)
In the limit of large
φ
(15. If the phase coherence length becomes larger than the sample. The parameter γ is then interpreted as the amount of decoherence experienced within a mean free path. or equivalently to the number of impurity scattering events involved in the path.82)
where τφ = φ/ vF . two and three dimensions. i. τ0 1/τφ + DQ2 (15. the sample size L must of course replace φ as a cut-oﬀ length. This is a physical sensible result. and hence the cooperon gets modiﬁed as W0 . DQ2
(15.
(15. 0) = C RA (Q. Clearly a path with n scatterings will then carry a factor e−nγ . To properly describe the breaking of phase coherence between the time-reversed paths one should include coupling to other degrees of freedom such as coupling to phonons or electronelectron scatterings. respectively 1 1 dQ = Ndim 1/τ + DQ2 (2π) φ
1/ 0
0
dQ
1 1/τφ + DQ2
1 π τφ D
π 1 2π Q 1 Q2 2π 2
tan−1
Dτφ
2 0
. With this modiﬁcation. We must therefore ﬁnd a method to cut-oﬀ these unphysical paths. 0. (15.
1D 2D
Dτφ
2 0
=
1 4Dπ 1 2π 2 D
0
ln 1 +
1 √ Dτφ
Dτφ
2 0
.272
CHAPTER 15. which amounts to 1 V 1 = DQ2
Q<
−1 0
Q<
−1 0
dQ 1 ∝ (2π)Ndim DQ2
Q<
−1 0
dQ
QNdim −1 .

2D 0 (τ )−1/2 . 1D WL τ δσ ∝ (15. and it is in this case that the term “weak” localization makes sense. it does not depend on the details of the material or the impurity concentration. and a logarithmic increase of the conductivity is to be expected.4
. In fact it can be shown that a quantum particle in a one dimensional disordered potential is always localized.5
Combined RPA and Born approximation
This section will be added in the next edition of these notes. one can apply a magnetic ﬁeld which breaks the time-reversal symmetry giving rise to the fundamental interference between time-reversed paths.15.86)
A way to measure this eﬀect is to change the phase coherence time τφ and to look at the change of conductivity. Foremost. See also Exercise 14. COMBINED RPA AND BORN APPROXIMATION
273
which in the limit of large τφ gives us information about the importance of the quantum corrections: − (τφ )1/2 . Indeed δρ ∝ −δσ ∝ ln τφ ∝ − ln T as is measured and shown in Fig.
15. The phase coherence can be changed in two ways. which states that due to the localization correction the conductivity decreases with increasing phase coherence time.5. Two dimensions is in between these two cases. 3D. φ This is an important result. apart from the logarithmic factor. That it is a small correction to the Drude conductivity can be seen from the ratio
WL δσ2D 1 =− σ0 πkF
ln
0
τφ τ0
. because here the correction is small. decreasing the tem−1 perature increases the phase coherence time τφ ∝ T α . (15.
(15. in the one-dimensional case it tells us that in one dimension the localization correction is enormously important and may exceed the Drude result. Furthermore. 15. In three dimensions the situation is more subtle.85) δσ2D ≈ − 2 ln 2π τ0 This result is “universal” since. Secondly. For the two dimensional case we have τφ e2 WL . there a metal-insulator transition occurs at a critical value of the disorder strength.2.84) − ln τφ .

IMPURITY SCATTERING AND CONDUCTIVITY
.274
CHAPTER 15.

We ﬁrst deﬁne and study the Green’s functions for free phonons in both the jellium model and the lattice model. Then we apply the Green’s function technique to the electronphonon interaction problem.1
The Green’s function for free phonons
jel in Eq.λ introduced in Eqs. With these prerequisites the non-interacting phonons are described by Hph and the electron-phonon interaction by Hel−ph as follows: Hph =
qλ
Ωqλ b† bqλ + qλ
1 . (3. that the relevant phonon operators to consider are not the individual phonon creation and annihilation operators. The point of departure is the second quantization formulation of the phonon problem presented in Chap.41) and Hel−ph in Eq. (3. we prove the existence of the so-called Cooper instability of the electron gas.23). 3.g.2)
.43). (3. in particular the bosonic phonon creation and annihilation operators b† −q.10) and (3. e.Chapter 16
Green’s functions and phonons
In this chapter we develop and apply the Green’s function technique for free phonons and for the electron-phonon interaction. We also show how the high frequency Einstein phonons in the free-phonon jellium model become renormalized and become the usual low-frequency acoustic phonons once the electron-phonon interaction is taken into account.σ ckσ Aqλ . We derive the one-electron Green’s function in the presence of both the electron-electron and the electron-phonon interaction. the phonon-induced instability which is the origin of superconductivity. but rather the operators Aqλ and A† deﬁned as qλ
INA It follows from all the Hamiltonians describing electron-phonon interactions. Finally.22) and appearing in the jellium phonon Hamiltonian Eq.
A† ≡ b† + b−qλ = A−qλ . Hel−ph
Aqλ ≡ bqλ + b† −qλ .1)
The phonon operator Aqλ can be interpreted as removing momentum q from the phonon system either by annihilating a phonon with momentum q or by creating one with momentum −q.
16. 2
Hel−ph = 275
1 V
gqλ c† k+q. (3. (3.4) and in the lattice phonon Hamiltonian Eq.
kσ qλ
(16. qλ qλ
(16.λ and bq.

(16. 2
(16.106) the basis states are given in terms of simple outer product states described by the electron occupation numbers nkσ and the phonon occupation numbers Nqλ . Hel and Hph . and thus minimizing the free energy gives ∂F/∂N ≡ µ = 0.1 with the substitutions (ν. (10. for τ < 0. τ ) for free phonons.
for τ > 0.
0 ˆ ˆ Dλ (q. .6)
while in the frequency domain we obtain
0 Dλ (q. GREEN’S FUNCTIONS AND PHONONS
Since Hph does not depend on time.8)
When governed solely by H0 the electronic and phononic degrees of freedom are completely decoupled.18) with a plus-sign. we can in accordance with Eq.
0 Dλ (q.
.
ωn = 2nπ/β. (10. and as in Eq.
(16. τ ). The frequency representation of the free phonon Green’s function follows by applying Eq. iqn ) =
2 Ωqλ 1 1 − = . iqn ) ≡ β 0 0 dτ eiqn τ Dλ (q. |Ψbasis = |nk1 σ1 . nk2 σ2 . .276
CHAPTER 16.9)
1 1 2 2 1 This expression is also valid in the grand canonical ensemble governed by Hph − µN .25). (16.5) deﬁne the ˆ phonon operators Aqλ (τ ) in the imaginary time interaction picture1 ˆ Aqλ (τ ) ≡ eτ Hph Aqλ e−τ Hph . iqn ) are found using the boson results of Sec. (16. τ ) =
− nB (Ωqλ ) + 1 e−Ωqλ τ − nB (Ωqλ ) eΩqλ τ −nB (Ωqλ ) e−Ωqλ τ − nB (Ωqλ ) + 1 eΩqλ τ
. . Nq λ . .3.
16. This is because the number of phonons can vary. . . The unperturbed Hamiltonian is the sum of the free electron and free phonon Hamiltonians. (1. cν ) → (qλ. iqn − Ωqλ iqn + Ωqλ (iqn )2 − (Ωqλ )2
(16. bqλ ). . τ ) ≡ − Tτ Aqλ (τ )A† (0) qλ 0
ˆ ˆ = − Tτ Aqλ (τ )A−qλ (0) 0 . in this section we do not take the Coulomb interaction between the electrons into account.3)
With this imaginary-time boson operator we can follow Eq.7)
where we have used that nB (Ωqλ ) = 1/ exp(βΩqλ ) − 1 .5)
0 0 The speciﬁc forms for Dλ (q.17) and introduce the 0 bosonic Matsubara Green’s function Dλ (q. εν . τ ) and Dλ (q.4)
where Tτ is the bosonic time ordering operator deﬁned in Eq. . Ωqλ . (10. H0 = Hel + Hph =
kσ
εk c† ckσ + kσ
qλ
Ωqλ b† bqλ + qλ
1 .
(16. For clarity.2
Electron-phonon interaction and Feynman diagrams
We next turn to the problem of treating the electron-phonon interaction perturbatively using the Feynman diagram technique. (10. |Nq λ . In the imaginary time domain we ﬁnd 0 Dλ (q. 10.

6) is changed into a P (τ )-integral. ˆ Tτ Aq
1 λ1 1 λ1
ˆ ck+q σ (τ1 )ˆkσ (τ1 ). . . (16. (16.
β 0
1 ˆ dτj P (τj ) = V
dτj
kσ qλ
ˆ gqλ c† ˆk+q. P (τm )
0
ˆ ˆ where the W (τ )-integral of Eq. τ ) = − m=0
(−1)m m!
∞ m=0
β 0
β
dτ1 .1) that only an even number of phonon operators will lead to a non-zero contribution in the equilibrium thermal average.. (12.13)
Note how the thermal average forces the paired momenta to add up to zero. a factor n!/2n from all possible ways to combine the remaining n momenta to the chosen ones and symmetrizing the pairs. And ﬁnally.−qi δλ . .10)
dτ1 .. . τi −τj )δqj . . (12. τ ). However.2. .
β 0
0
ˆ ˆ c ˆkσ dτm Tτ P (τ1 ) .10) can be written as a product of a phononic and an electronic thermal average.
(16.5) and (16.. i j
(16.. In the ﬁnal combinatorics the prefactor (−1)m /m! = 1/(2n)! of Eq. (10. all choices leading to the same result.10) seems to ˆ ˆ be quite diﬀerent since W (τ ) contains four electron operators and P (τ ) only two.Aqm λm (τm )
0
Tτ c† ˆk+q σ (τ1 )ˆkσ (τ1 ).5) we use the interaction picture representation. and substitutes the two-particle interaction Hamiltonian W (τ ) ˆ (τ ) with the electron-phonon interaction P
∞
Gσ (k.σ (τj )ˆkσ (τj ) Aqλ (τj ).. . First we note that because the electronic and phononic degrees of freedom decouple the thermal average of the integrand in the m’th term of say the denominator in Eq. ELECTRON-PHONON INTERACTION AND FEYNMAN DIAGRAMS
277
What happens then as the electron-phonon interaction Hel−ph of Eq. Then a factor (2n)!/(n!n!) appears from choosing the n momenta qj among the 2n to be the independent momenta. A sign (−1)n appears from one minus sign in each of the n factors of the form Eq. .−qi δλ .
0
ˆ ˆ dτm Tτ P (τ1 ) .10) is modiﬁed as follows. we shall now show that the two expressions in fact are very similar.ˆ† m σ (τm )ˆkσ (τm ) c ck+q c
1
.λ
i
j
= −|gq λ |
i i
2
0 Dλ (qi .79) for boson operators to break down the n-particle phonon Green’s function to a product of n single-particle Green’s functions of the form gq λ gq
i i j λj
ˆ ˆ Tτ Aq λ (τi )Aq
i i
j λj
(τj )
0
=
ˆ ˆ |gq λ |2 Tτ Aq λ (τi )A−q λ (τj )
i i i i i i
0
δqj .Aqm λm (τm )ˆ†
1
0
=
0
ˆ Tτ Aq
ˆ (τ1 ). (16. (12.λ .. For each value of n
. In analogy with Eq. we use Wick’s theorem Eq. Hence we end up with the prefactor (−1/2)n /n!. (16.13).. P (τm )ˆkσ (τ ) c† (0)
β
0
(−1)m m!
. but ˆ now in momentum space. c
(16.ˆ† m σ (τm )ˆkσ (τm ) c ck+q c (τ1 ).12)
It is clear from Eq. (16. so we now write m = 2n..2) is turned on? We choose to answer this question by studying the single-electron Green’s function Gσ (k.11)
At ﬁrst sight the two single-electron Green’s functions in Eqs. Next. (16.16.

is the same as for the Coulomb interactions Eq. 0 and 2n internal fermion lines Gσ (pj .24): (1) (2) (3) (4) Fermion lines with four-momentum orientation: Phonon lines with four-momentum orientation: kσ.
∞
Gσ (k. τ ) = − n=0
(−1)n n!
∞ n=0
β 0
β
dτ1 . iqn Conserve the spin and four-momentum at each vertex. the phononmediated interaction is retarded.15)
c† 1 +q. i. . GREEN’S FUNCTIONS AND PHONONS
ˆ the 2n operators P (τi ) form n pairs.e. but while the Coulomb interaction is instantaneous or local in time. iqn ). and we end with the following single-electron Green’s function.
β 0
0
ˆ ˆ dτn Tτ P(τ1 ) . . iqn ) qλ. P(τn )ˆkσ (τ ) c† (0) c ˆkσ
β
0
(−1)n n!
.
0
dτn
ˆ ˆ Tτ P(τ1 ) .σ1 (τj )ˆ† 2 −q. (12. All vertices must contain an incoming and an outgoing fermion line as well as a phonon line. however. ikn
0 ≡ Gσ (k.σ2 (τi )ˆk2 σ2 (τi )ˆk1 σ1 (τj ). The derivation of the Feynman rules in Fourier space.10) is given by the eﬀective two-particle interaction operator
β 0
ˆ dτi P(τi ) =
β 0
β
dτi ×
0
dτj
k1 σ1 k2 σ2 qλ
1 0 |g |2 Dλ (q. i.16) This interaction operator resembles the basic two-particle Coulomb interaction operator Eq. . ikn ).278
CHAPTER 16. . non-local in time. . Multiply each fermion loop by −1. ˆk ck c c V qλ
(16. τi −τj ) c† 1 +q.σ1 (τj )ˆ† 2 −q. τi −τj ). τi −τj ) 2V 2 qλ (16. regarding both the operators and 0 the coupling strength (1/V) |gqλ |2 Dλ (q. . two external fermion lines Gσ (k. P(τn )
0
ˆ ˆ where the P(τ )-integral substituting the original P (τ )-integral of Eq. 1 Multiply by βV for each internal four-momentum p and perform the sum pσλ . ipj ). . (2.σ2 (τi )ˆk2 σ2 (τi )ˆk1 σ1 (τj ). and no spin ﬂipping. (16. .e.14)
dτ1 .34).17)
.
(16. ikn )
(5) (6)
0 1 ≡ − V |gqλ |2 Dλ (q. incoming momenta must equal the outgoing. ˆk ck c c
From this interaction operator we can identify a new type of electron-electron interaction ph Vel−el mediated by the phonons
ph Vel−el =
1 2V
k1 σ1 k2 σ2 qλ
1 0 |g |2 Dλ (q. At order n draw all topologically diﬀerent connected diagrams containing n 0 1 0 oriented phonon lines − V |gqλ |2 Dλ (q. ˜ ˜
(16.

Likewise. 5. heavy ions as parameters in the Schr¨dinger o equation for the fast moving.44) and (16. the question naturally arises of whether to study the inﬂuence of the electrons on the ions before that of the ions on the electrons. 1 1 |gq |2 = V V Ze2 q 0
2
e2 Ω 1 N = = W (q) Ω.18).1 we have already sketched the ion plasma oscillation that occurs. but in Sec. In the case of an ion density ρ0 = N/V we found the ion dispersion-less jellium phonon modes in the long wave length limit. 3. Note that we have dropped in the last equality in accordance with the convention introduced in Sec. (2.e. This theorem is the condensed matter physics analogue to the well-known Born-Oppenheimer approximation of molecular physics.20)
To discuss the role of the electron dynamics we now add the electron-electron Coulomb interaction Vel−el of Eq. light electrons. COMBINING COULOMB AND ELECTRON-PHONON INTERACTIONS
279
16.21)
16. iqn ) = W (q) .19)
which not surprisingly is proportional to the Coulomb interaction W (q). bare. 0M V (16.34) and study the full Hamiltonian H for the electronic and phononic system. i.1 we studied the ion plasma oscillations in the jellium model neglecting the electron dynamics.3. phonon-mediated electron-electron interaction is Ω2 1 |gq |2 D0 (q. V (iqn )2 − Ω2 (16.1
Migdal’s theorem
When the electron-phonon coupling Hel−ph is added. (3. (16. The resulting. The answer is provided by Migdal’s theorem. 2 2 2M Ω 2 0q
(16. In Fig. The latter states that it is a good approximation to consider the coordinates Ri of the slowly moving. which is then solved.3
Combining Coulomb and electron-phonon interactions
We now discuss the eﬀect of the long range Coulomb interactions between electrons and ions and between electrons themselves. A complete calculation is rather tedious.18)
The coupling constant for the electron-electron interaction mediated by these jellium phonons is found by combining Eqs.3. if we consider the interaction between the ions and the electron gas assuming the latter to be homogeneous and completely inert. In the second stage the values of Ri are then changed adiabatically. disregarding all the dynamics of the electrons. it can be proven by phase space arguments that renormalization of the electron-phonon vertex is suppressed at least by a
. Ωq = Ω = Z 2 e2 N .1. H = Hel + Vel−el + Hph + Hel−ph . 3. or vice versa.16. For simplicity we henceforth study only longitudinal phonons and hence drop all reference to the polarization index λ.

and if.
16. get renormalized by electron screening processes to the experimentally observed low frequency acoustic phonons. If these phonons are used we can prove Eq. (12.22) is correct we can prove the assumed phonon renormalization. Eq. yields the following bare. |iqn | ωD . Ω.24)
. The diagram on the left hand side in Eq. 1 0 −W (q) − V |gq |2 D0 (q. Combining the Feynman rules for these two interactions.22) and (3. Furthermore.3. iqn ) −Veﬀ (q. (16. We therefore study only the inﬂuence of the electronic degrees of freedom on the bare phonon degrees of freedom. The important frequencies for acoustic phonons are smaller than the Debye frequency ωD . respectively.24) and (16. Eqs. iqn ) ≡ + . a phase space factor of the order ωD /εF must appear in front of the usual unrestricted contribution from two such lines. χ0 = −d(εF ).
≈
m × M
. where m and M are the masses of the electron and ion.2
Jellium phonons and the eﬀective electron-electron interaction
In more realistic calculations involving interacting electrons we need to consider the sum of the pure electronic Coulomb interaction and the phonon-mediated interaction.22)
The proof builds on a self-consistency assumption. (16. GREEN’S FUNCTIONS AND PHONONS
factor m/M ∼ 10−2 . Consequently.22). (16.21).5) at the ﬁrst and third equality sign. thus we concentrate on phonon frequencies ωq < ωD .17). according to Eq. The result of the analysis is that the assumption for Migdal’s theorem indeed is fulﬁlled. We assume that the high frequency jellium phonons.37) the typical (acoustic) phonon interaction line for low frequencies. (13.280
CHAPTER 16. due to four-momentum conservation. respectively. The ratio between the values of the two diagrams is therefore roughly given by ωD vs k ω W (q) = 1 D =2 × D × d(εF ) = εRPA εF εF 2 vF kF Z 3 m kD ≈ M kF m . (16. the pairbubble of Eq. (16. ωq = vs q. (13. In the following we assume that we can neglect the phonon-induced renormalization of the electron-phonon vertex.
(16. Now. the two internal electron propagators are conﬁned within ωD to the Fermi surface. is −W (q)/εRPA . This combined interaction will be the basis for our analysis of the interacting electron gas henceforth.22) contains one phonon interaction line and two electron propagators more than the diagram on the right hand side.23)
where we have used Eqs. as shown in the following section. eﬀective electron-electron interaction line. We will just outline the proof of Migdal’s theorem here by studying the simplest phonon correction to the electron-phonon vertex. M (16.

(13. ω 2 − Ω2 + i˜ η
(16. (b) The same for the RPA renormalized eﬀective electron-electron RPA interaction Veﬀ (q.4. eﬀective electron-electron interaction.20) into Eq. iqn ) = W (q) + W (q)
Ω2 (iqn )2 = W (q) . the
§ !
© ¨ ¥¦ £ ¢ ¡ §
§
© ¦§¨ ¤ ¥ £ ¢ ¡
. (16. Since Veﬀ (q) is proportional to the bare Coulomb interaction. At high frequencies the normal Coulomb interaction is recovered. 16.
0 The speciﬁc form of Veﬀ is obtained by inserting Eq.
0 Veﬀ (q. ω) = W (q)
ω2 . let us remind ourselves of the following expressions from Chap. (iqn )2 − Ω2 (iqn )2 − Ω2
(16. eﬀective electron-electron interaction Veﬀ (q.
16. in the limit high electron densities are given by RPA. 13. i. ω) → W (q) for ω → ∞. 16. the interaction is attractive for frequencies ω RPA less than the acoustic phonon frequency ωq . and Veﬀ (q.61)–(13.26)
0 The real part of Veﬀ (q. see Sec. Eqs. It is seen that the bare.4. PHONON RENORMALIZATION BY ELECTRON SCREENING IN RPA
281
0 Figure 16.1: (a) The real part of the bare. Now. Before we consider how the phonon propagator is renormalized by the electronic RPA.25)
or going to real frequencies. ω).24).1(a). according to our main result in Chap. iqn → ω + iη. ω) is shown in Fig. ω) as a function of the real frequency ω for a given momentum q. at low frequencies the electron-phonon interaction combined with the originally fully repulsive Coulomb interaction results in an attractive eﬀective electron-electron interaction. these processes. Migdal’s theorem leads us to disregard renormalization due to phonon processes 0 and only to consider the most important electron processes.66) between the dielectric function εRPA .4
Phonon renormalization by electron screening in RPA
The electronic Coulomb interaction renormalizes the bare. ω) → W RPA (q) for ω → ∞. 13. eﬀective electron-electron interaction becomes negative for ω < Ω. (16.e.16. Note that the interaction is 0 attractive for frequencies ω less than the jellium phonon frequency Ω. and that Veﬀ (q.
0 Veﬀ (q.

iqn ) . iqn ) = 1 + = = RPA .33)
= +
∗
. iqn ) (16.30)
=
+
RPA
.28)
RPA The solution for Veﬀ (q.
(16. χ0 (q. −DRPA (q.282
CHAPTER 16.27b) (16. iqn ) = = +
RPA
is the sum of all diagrams between the ﬁrst and the last phonon line. iqn ) = 1 − W (q)χ0 (q.e.
(16. iqn ) χ0 (q.27a) (16. GREEN’S FUNCTIONS AND PHONONS
RPA
density-density correlator −χRPA =
. and then collecting all the diagrams containing only Coulomb interaction lines into one sum (this simply yields the RPA screened Coulomb interaction W RPA ). (13. a physically more transparent form of Veﬀ is obtained by expanding the inﬁnite series Eq. i. iqn ) ε (q. χRPA (q. while the renormalized phonon line DRPA (q.
(16.68) for W RPA and obtain
RPA −Veﬀ (q. 1 − W χ0 1 − W χ0 ε (q.28). and the simple pair-bubble −χ0 =
. we now extend the RPA-result Eq. The solution for the RPA renormalized phonon line is −DRPA (q. iqn )
=
=
+
. iqn ) = = −D0 (q.32)
1−
RPA
. Here the renormalized coupling gq q RPA gq ≡
is the sum of all diagrams between the outgoing left [incoming right] vertex and the ﬁrst [last] phonon line. (16. 1 1 − χRPA (q. iqn ) = 1 − W (q)χ0 (q.27c)
εRPA (q. iqn ). iqn ). iqn ) V |gq |2 D0 (q.
(16. iqn ) = RPA . iqn )
Returning to the electron-phonon problem. iqn ) has the standard form RPA −Veﬀ (q. 0 1 − Veﬀ (q) χ0 (q.31)
. without contributions from the external coupling vertices. iqn )
RPA [ g RPA ]. (16. 1 RPA RPA −Veﬀ (q. iqn )
(16. iqn ) =
= 1−
=
0 −Veﬀ (q) . while collecting the remaining diagrams containing a mix of Coulomb and phonon interaction lines into another sum. iqn ) W χ0 1 1 1 + W (q)χRPA (q.29)
RPA While this expression is correct. iqn ) −W RPA (q) − V |gq |2 DRPA (q.

and as a result the phonon dispersion gets renormalized to the usual low frequency acoustic dispersion ωq = vs q. PHONON RENORMALIZATION BY ELECTRON SCREENING IN RPA Using ﬁrst Eqs.7) and (16. (16. iq ) ε n
=
|gq |2 /V
ωq (q → 0. iq ) (iqn )2 − ωq −Ω n
283
(16.38)
This we recognize as the Bohm-Staver expression Eq.4. RPA (q. (16. (16. 2 2 2 2 ε (iqn )2 − ωq εRPA (iqn ) − ωq (16. iqn ) = Z 2 e2 ρ0 ion = εRPA 0 M Ze2 ρ0 el . iqn ) = − V q
= +
RPA
= (1 + W χRPA ) gq =
gq .34) and (16. (3. (16. In more elementary treatments this spectrum is derived by postulating short range forces following Hooke’s law.35)
is the renormalized phonon frequency due to electronic RPA screening. The signiﬁcance of this result is that starting from the microscopic Hamiltonian Eq. (16. respectively. The long range Coulomb forces of the non-interacting problem resulted in optical jellium phonons with the high frequency Ω.
2 ωq 2Ω W (q) = RPA .36) The ﬁnal form of the RPA screened phonon-mediated electron-electron interaction is now obtained by combining Eqs. Inserting this into 3 2 ρ0 yields the following explicit form of ω : Eq. long wave length limit.5) for the dispersion of acoustic phonons in the jellium model. 3M F
(16.35) and using the relation kF = 3π el q
1 RPA 2 RPA |g | D (q.66) 2 and (13. iqn ) = where ωq ≡ Ω εRPA (q.20) and then Eq.37) We now see that this renormalized propagator is identical to the free phonon propagator Eq.35) for ωq in the static.23) that in this limit εRPA (q.20) where the unscreened phonon frequency Ω and the unscreened Coulomb interaction W (q) have been replaced by their RPA screened counterparts ωq and W (q)/εRPA . we have used the Feynman diagram technique to show how the phonon spectrum gets renormalized by interacting with the electron gas. but now we have proven it from ﬁrst principles.
RPA gq ≡
(16. (16.16. = 2 2 W (q)χRPA (q.27c) leads to DRPA (q. but before doing so we study how also the coupling RPA constant gq gets renormalized in RPA and acquire the value gq . (13.36). A further physical interpretation of this result is obtained by evaluating the expression Eq. (16. iqn ) → ks /q 2 = (4kF /πa0 )/q 2 . We note from Eqs. In a moment we shall interpret this new frequency.21) for the coupled electron and phonon problem. εRPA 0 M (iqn )2 − Ω2 2Ω 2Ω . 0) =
Ze2 ρ0 el q= 2 ks 0 M
Zm v q. By introducing the electron-electron interaction the Coulomb forces get screened. namely the explicit form of the eﬀective electron-electron interaction due to the combination of the Coulomb and the
.34)
(16. We end by stating the main result of this section.

iqn ) rapidly approaches W RPA (q) for frequencies larger than the given acoustic phonon frequency ωq . ˜ q 0 q q ˜ q ˜
(16. The divergence is due to repeated scattering between electron pairs occupying time-reversed states of the form |k ↑ and ˜ |−k ↓ .
=
+
Λ
.5
The Cooper instability and Feynman diagrams
In 1956 Cooper discovered that the electron gas in an ordinary metal would become unstable below a certain critical temperature Tc due to the phonon-induced attractive RPA nature of the eﬀective electron-electron interaction Veﬀ (q. (16. see also Fig. while it becomes attractive instead of repulsive for frequencies below ωq .27). This discovery soon lead Bardeen. We will ﬁrst derive this divergence. RPA Veﬀ (q. p). is proportional to ε2 or ωq for frequencies less than the Debye frequency ωD = vs kD and zero otherwise.40) in the form of a Dyson equation for the pair-scattering vertex Λ.
Λ
which is equivalent to the following integral equation
RPA ˜ ˜ ˜ Λ(k. GREEN’S FUNCTIONS AND PHONONS
!
−˜↓ p
(iqn )2 . the density 2 of phonon states.284
electron-phonon interaction. according to the Debye model of acoustic phonons.42)
RPA To proceed we make a simplifying assumption regarding the functional form of Veﬀ (q. iqn ) from W RPA (q). Using the four-momentum notation k = (k.41)
1 Vβ
0 RPA ˜ − Veﬀ (k− q ) G↑ (˜) G↓ (−˜) Λ(˜..1(b): 1 RPA RPA −Veﬀ (q. iqn ) = + = −W RPA (q)
˜ −k↓
CHAPTER 16. 13 no instabilities arise due to the pure Coulomb interaction. First we note that according to our analysis of the electron gas in Chap.40) ˜ ˜ ˜ ˜ ˜ ˜ ˜ k↑ p↑ ˜ k↑ p↑ ˜ k↑ k1 ↑ p↑ ˜ k↑ k1 ↑ k2 ↑ p↑ ˜ Suppressing all arguments and stripping away the external electron propagators we can recast Eq.39)
16. This means that most of
. 3. and then we will discuss its physical interpretation. According to Eq. iqn ).39) and Fig.5.. 2 (iqn )2 − ωq (16. The instability arises because a certain class of electron-electron scattering processes when added coherently yields a divergent scattering amplitude. Cooper and Schrieﬀer (BCS) to develop the microscopic theory explaining superconductivity. ω) at low frequencies. Dion (ε). In this section we will derive the Cooper instability using Feynman diagrams. (3. 16.1(b). iqn ) −W RPA (q) − V |gq |2 DRPA (q. (16.
(16. ikn ) we consider the following pair ˜ ˜ scattering vertex Λ(k. 16. Thus we are really only interested in RPA the deviations of Veﬀ (q. p) = given by the inﬁnite ladder-diagram sum over scattering events between time-reversed electron pairs:
(16. Sec. Further. see Eq. p) = −Veﬀ (k− p) + ˜
" #$%& '( )
−˜↓ p ˜ −k↓ −˜↓ p ˜ −k↓ ˜ −k1 ↓ ˜ ˜ k−k1 ˜ −k↓ ˜ −k1 ↓ ˜ ˜ k−k1 ˜ −k2 ↓ −˜↓ p Λ
=
˜ p k−˜
+
˜ p k1−˜
+
˜ ˜ ˜ p k2−k1 k2−˜
+ .

44)
The summand on the right hand side does not contain the external momentum k. the entire Fermi-surface becomes unstable at that temperature. p) in the summand we can write Λ(p). It is therefore a reasonable approximation to set ωq = ωD . The conclusion is clear: if the pair-interaction strength Λ diverge for a certain critical temperature Tc .−iqn )
. As T is lowered the denominator in Eq. q
(16. and thus for Λ(q. and the on-set of it marks the transition from a normal metallic state to a superconducting state. (16. since the eﬀective pair-interaction is attractive only for energy exchanges less than ωD .43).45) can approach zero from above resulting in an arbitrarily strong or divergent pair-interaction strength Λ. i. we set the interaction strength to be constant.−iqn ) Λ(q.45) to zero using Gσ (q. Hence we arrive at the model used by Cooper and by BCS: −V. iqn ) = 1/(iqn − εq )
. which we naturally denote Λ. p) thus only involves frequencies less than ωD .e. the resulting pair-interaction Λ equals the attractive pair-interaction strength V from Eq. Now it is furthermore evident that the p-dependence occurs only in the Λ-function. and for those it takes the form 1 Λ(k.e. hence a consistent solution is obtained by taking Λ(k. iqn ) G↓ (−q. p) = Λ(p). In quantum mechanics an inﬁnite scattering amplitude signals a resonance. for the on-set of the Cooper instability is 0 obtained by setting the denominator in Eq. i. iqn ) ≈
(16. |iqn | > ωD .5. or β = βc . The critical temperature T = Tc . in the present case the formation of a bound state between the time-reversed pair of electrons. This instability is called the Cooper instability. and solve for it: V V 1− β
|iqn |<ωD
Λ=
1 V
0 G↑ (q. |iqn | < ωD 0. But in our model this would then happen simultaneously for all electron pairs within a shell of thickness ωD of the Fermi surface. (16. THE COOPER INSTABILITY AND FEYNMAN DIAGRAMS
285
the phonons encountered have a frequency of the order ωD .
(16. p) = V + β
ωD iqn
1 V
0 0 V G↑ (q. β 1/oD.
RPA Veﬀ (q.44) we can take Λ outside the sum. whence for the left hand side we conclude Λ(k. as a last simpliﬁcation. On the right hand side of Eq. (16. (16.16. iqn ) q
0 G↓ (−q.45)
We see that at high temperatures.43)
˜ ˜ The integral equation for Λ(k. Finally. and a new ground state is formed involving bound electron pairs in time-reversed states. p). p) to be a constant.

Fig. the critical temperature Tc for the Cooper instability is lowered to about 1 K by the exponential factor.47)
Two important comments can be made at this stage. 2 2π
βc ωD
1. see e. The ﬁrst is that although the characteristic phonon energy ωD is of the order 100 K. 3.
(16. it is not possible to reach the new ground state resulting from the Cooper instability by perturbation theory in V of the normal metallic Fermi sea.6b. GREEN’S FUNCTIONS AND PHONONS n+
1 2
: 1 V d(εF ) 1 1 V = iqn − εq −iqn − εq βc 2 q |iqn |<ωD 1 β ω 2π c D 1 π 1 − 2 = V d(εF ) |qn | 2 n+ 1 2
n=0 ∞
1=
V βc
dε
−∞
|iqn |<ωD
1 2 qn + ε2
= ≈
V d(εF ) 2βc
|iqn |<ωD
V d(εF ) βc ωD ln 4 .g.
(16. d(εF )/2. since Tc (V ) ∝ exp(−const/V ). The second comment is that Tc is a non-analytic function of the pair-interaction strength V . These problems will be treated in some of the exercise of this chapter and in much greater detail in the next chapter concerning the BCS theory of superconductivity.
.286 and qn =
2π βc
CHAPTER 16. Consequently.46)
where we use the density of states per spin. From this equation Tc is found to be kB Tc ≈ ωD e
2 − V d(ε F )
.

Chapter 17
Superconductivity
17.3
BCS theory with Green’s functions
287
.1
The Cooper instability
17.2
The BCS groundstate
17.

4.4
17.2
Experimental consequences of the BCS states
Tunneling density of states speciﬁc heat
17.1 17.4.288
CHAPTER 17. SUPERCONDUCTIVITY
17.5
The Josephson eﬀect
.

1
Introduction First look at interacting electrons in one dimension
One-dimensional transmission line analog
18.spinless case
Interacting one dimensional electron system Bosonization of Tomonaga model-Hamiltonian Diagonalization of bosonized Hamiltonian Real space formulation Electron operators in bosonized form
289
.2.5
The Luttinger-Tomonaga model .3.2 18.3 18.4 18.1 18.3
18.3.3.Chapter 18
1D electron gases and Luttinger liquids
18.1 18.3.2
18.3.

5 18. 1D ELECTRON GASES AND LUTTINGER LIQUIDS
18.1
Luttinger liquid with spin Green’s functions Tunneling into spinless Luttinger liquid
Tunneling into the end of Luttinger liquid
18.2
What is a Luttinger liquid? Experimental realizations of Luttinger liquid physics
Edge states in the fractional quantum Hall eﬀect Carbon Nanotubes
.6.290
CHAPTER 18.4 18.8.7 18.8.1 18.6
18.8
18.

4)
Note the prefactor 1/V in Eq. .3). . Ly . travelling at the speed v = ω/|k|. .Appendix A
Fourier transformations
Fourier transformation is useful to employ in the case of homogeneous systems or to change linear diﬀerential equations into linear algebraic equations. 291 1 V eik·r = δ(r).0 .1)
A.
V
(A.
k
kx = 2πnx .3)
(A. nx = 0.4). It is our choice to put it there. t) under study on plane wave components. ±2. The idea is to resolve the quantity f (r.ω ei(k·r−ωt) . (A. (A. f (r + Lx ex ) = f (r + Ly ey ) = f (r + Lz ez ) = f (r) can be written as a Fourier series f (r) = where fk = dr f (r) e−ik·r . In all cases the product of the normalization constants should be 1/V.
(A.1
Continuous functions in a ﬁnite region
Consider a rectangular box in 3D with side lengths Lx . The central theorem in Fourier analysis states that any well-behaved function fulﬁlling the periodic boundary conditions.5)
. (A. fk. Lx likewise for y and z.2)
1 V
fk eik·r . or to put 1/ V in front of both equations. ±1. Lz and a volume V = Lx Ly Lz .
k
(A. An extremely important and very useful theorem states dr e−ik·r = V δk. Another choice would √ be to put it in Eq.

292
APPENDIX A.8) (2π)3 Note that the dimensions are okay. 1r ←→ (2π) δ(k).2
Continuous functions in an inﬁnite region
If we let V tend to inﬁnity the k-vectors become quasi-continuous variables.0 . continuous ω. (A.5).
dt eiωt = 2π δ(ω).
(A.6)
Now you see why we choose to put 1/V in front of f (r) = and also dk f eik·r .10) as the Fourier transform of the constant function f = 1 to delta functions (and back): 1r ←→ V δk.
A. −∞ 2π Note again that the dimensions are okay. and in complete analogy with Eq.9)
−∞
dω −iωt e = δ(t).
A.4
Some useful rules
We can think of Eqs. f (r) = 1 V fk eik·r
k V→∞
−→
1 V V (2π)3
dk fk eik·r =
k. (A.
(A.10)
A.11c)
.
3
1k ←→ δ(r). to the case where this period tends to inﬁnity.
dk f eik·r .11b) (A.1) – we have f (t) = and also
∞
dω fω e−iωt .7) – but with the opposite sign of i due to Eq.3) is converted into an integral.8) and (A. Note also that by using Eq.7)
fk =
dk ik·r e = δ(r). (A. Thus t plays the role of r and ω that of k. (A. (A. FOURIER TRANSFORMATIONS
Note the dimensions in these two expressions so that you do not forget where to put the factors of V and 1/V. (A. and the k-sum in Eq.5). 1t ←→ 2π δ(ω). (A. continuous k. (A.3) an reduce by use of Eq.3
Time and frequency Fourier transforms
The time t and frequency ω transforms can be thought of as an extension of functions periodic with the ﬁnite period T .
discrete k. (A. (2π)3 k
(A. (A.
(A. (A. 1ω ←→ δ(t). −∞ 2π
∞
∞
fω =
∞ −∞
dt f (t)eiωt . (2π)3 k
We have dr f (r)e−ik·r . Again it is easy to use these expression to verify that Fourier transforming twice brings you back to the starting point.5) you can prove that Fourier transforming from r to k and then back brings you back to the starting point: insert fk from Eq. dr e−ik·r = (2π)3 δ(k).4) into the expression for f (r) in Eq.11a) (A. 1k ←→ δ(r).

13)
A related rule.12)
or in words: a convolution integral in r-space becomes a product in k-space. f (r. is derived in a similar way (and here given in three diﬀerent versions): dr h(r) g ∗ (r) = dr h(r) g(r) = dr h(r) g(−r) = dk h g∗ .
×.15) (A. r ) of two spatial variables r and r can only depend on the diﬀerence between the coordinates and not on the absolute position of any of them.k ∝ δ(k + k ). The consequences in k-space from this constraint are: dk f eik·(r−r ) ei(k +k)·r .18) (A.5. r ) = f (r−r ).
2
∂t ←→ −iω.A.19)
A. and thus fk.
k
(A. (A.
k
(A. ←→ −k . f (r. For the gradient operator we have
r f (r)
=
r
1 V
fk eik·r =
k
1 V
fk
k
re
ik·r
=
1 V
ikfk eik·r .14) (A. the invariance of inner products going from r to k. Any physical observable f (r. (2π)3 k −k dk h g .
(A. it is obvious from the factor ei(k +k)·r that any reference to the absolute value of r only can vanish if k = −k. By direct application of the deﬁnitions and Eq.21) Since this has to be a function of r − r .17)
Similarly for
2. and ∂t (remember the sign change of i in the latter):
r 2
←→ ik. (2π)3 k k (A. × ←→ ik × .16)
Finally we mention the Fourier transformation of diﬀerential operators.5
Translation invariant systems
We study a translation invariant system.20)
.k
1 V
hk gk eik·r . (A. r ) = dk (2π)3 dk f eik·r eik ·r = (2π)3 k. (2π)3 k k dk h g . ds h(r−s) g(s) ←→ hk gk .8) we ﬁnd f (r) = ds h(r−s) g(s) = ds 1 V2 hk eik·(r−s) gk eik ·s =
k. (2π)3 k.k (A.k dk (2π)3 (A. TRANSLATION INVARIANT SYSTEMS
293
Another useful rule is the rule for Fourier transforming convolution integrals.

we can go through the same arguments or use the formulae from above to get ˜ fk. FOURIER TRANSFORMATIONS
To ﬁnd the proportionality constant.k = Vδk. or in short f (r.26) where A and B are some operators. r ) = g(r − r ). translation−invariant systems. (A.21) and (A.27) (A. For a translation-invariant system we now that g(r.25) This result is used several times in the main text when we consider correlation functions of the form g(r.23)
. (A. and by using the result in Eq. (A. V (A. r ) = A(r)B(r ) .22) we read oﬀ that ˜ fk. r ) = f (r − r ) = dk ˜ ik·(r−r ) f e .k = (2π)3 δ(k + k )fk .−k fk . (A. (2π)3 k (A.22)
and by comparing the two expressions Eqs. (A.−k . r ) ←→ fk.24) For the discrete case. we can also ﬁnd the Fourier transform of f by explicitly using that f only depends on the diﬀerence r − r ˜ f (r.294
APPENDIX A.25) for k = −k we get that g(k) = 1 A(k)B(−k) .

j ): j H = −t
jδ
c† cj . .1
Prove Eq. j+δ
This Hamiltonian is known as the tight-binding Hamiltonian. (1. Use the discrete Fourier transformation a. . (1. . Obtain the ν ν ν ν fermion analogue of Eq. Argue why in this case one has c† b = c† b cνn c† nj . where with the probability amplitude t one valence electron can hop from an ion j to one of the nearest neighbor ions j + δ (as usual {c† . Hint: use the ν basis transformations Eq. What about the ν j fermion anti-commutator sign?
j
νi . cj } = δj. (1. (a) Consider a 1D lattice with N sites. 2.
Exercise 1. A good starting point for analyzing such systems is to describe the kinetic energy by hopping processes. and a lattice constant √ Here j = 1. . cj = (1/ N ) k eikja ck to diagonalize H in k-space and plot the eigenvalues εk as a function of k. N and δ = ±1.63) for fermions: Ttot = fermion operators c† .60) with ν
Exercise 1. Hints: write Eq.62) by moving the pair c† b cνn to the left.Appendix B
Exercises
Exercises for Chapter 1
Exercise 1.67) in the real space representation Eq.99a).3
In some crystals the valence electrons are rather tightly bound to their host ions. (1.νj
Tνi νj c† i cνj . periodic boundary conditions. 295
.2
Find the current density operator J in terms of the arbitrary single particle basis states ψν and the corresponding creation and annihilation operators a† and aν . (1.

Exercise 2. Generalize the 1D model to a 2D square lattice also with the lattice constant a and plot contours of constant energy εkx ky in the kx ky plane. and use this value to determine the microscopic parameters kF .
where [a. 1.287 nm.
Exercise 2. (2.43) is not exact ∗ near the energy minimum density rs = rs = 4. θ.83. In this case the 2D tightbinding model is applicable. φ). where the ions form a 2D square lattice. Calculate the density n of the resulting gas of valence electrons.1
Iron (Fe) in its metallic state has valence II.2
Use the variational principle to argue that although the expression Eq. a† ] = 1. the result E ∗ /N = −1.47) for the contributions from the direct Coulomb interaction processes to the interaction energy in second order perturbation theory. Relate the result to the Coulomb potential. with ks being some real positive k constant with the dimensions of a wavevector. εF . and perform the 0 dφ and −1 d(cos θ) integrals ﬁrst.5
The Yukawa potential is deﬁned as V ks (r) = r0 e−ks r .296
EXERCISES FOR CHAPTER 2
(b) In the high-temperature superconductors the conduction electrons are conﬁned to parallel CuO-planes. while ω and ω0 are positive constants. Diagonalize H by introducing the operator α ≡ a + ω0 /ω and its Hermitian conjugate α† . and λF . What might be the physical origin of the second term in H (see Sec. (2.29 eV nevertheless ensures the stability of the electron gas.34) and (2. and X-ray measurements have revealed that it forms a body-centered-cubic (BCC) crystal with side length a = 0.
Exercises for Chapter 2
Exercise 2.
.4
Consider a bosonic particle moving in 1D with the Hamiltonian H = ω a† a + 1 2 + ω0 a† + a .1)? Compare the result to a classical and a ﬁrst quantized treatment of the problem. Hints: work in polar coordinates
2π +1 ∞ 2 0 r dr
r = (r.4.3
Starting from Eqs. vF . and determine the eigenenergies. (2.
Exercise 1. The remaining integral is a simple integral of the sum of two exponential functions.45) derive the expression Eq. Prove that the Fourier transform is Vq s =
4πe2 0 2 +k 2 q s e2
.
Exercise 1.

3.3 we saw an example of the existence of 1D electron gases in carbon nanotubes. (a) Argue that in this case we need only consider the simple phonon absorption and emission processes given by
abs emi Hel−ph = Hel−ph + Hel−ph = q
gq c† k+q. It can be estimated using Fermi’s Golden Rule (suppressing the unimportant spin index):
1
emi τk
=
2π
f
emi f |Hel−ph |i
2
δ(Ef − Ei ). the relation between the 2D Fermi wave vector kF and the 2D electron density: kF 2 = 2πn.41). (1. i. |i =
† {kσ}i ckσ {q}i
√1
Nq !
b† q
Nq
|0 . and show that
1 c† c b† |i q Nq +1 k+q k
emi for a given q = 0 in Hel−ph the only possible normalized ﬁnal states is √
.
Exercises for Chapter 3
Exercise 3.2 we saw an example of the existence of 2D electron gases in GaAs/Ga1−x Alx As heterostructures.e. the relation between the 1D Fermi wave vector kF and the 1D density of states per length. Use the result to derive the 1D electron density: kF = 2n/π. Assume that |i is a simple but unspeciﬁed product state. d(ε). in analogy with the 3D case. (3.5
In Sec. in analogy with the 3D case.1
We want to study the inﬂuence of electron-phonon scattering on a given electron state INA |kσ using the simple Hamiltonian Hel−ph of Eq.
involving a sum over all possible ﬁnal states with energy Ef = Ei .
.
(c) Show for the state |i that 1
emi τk
=
2π
q
i |gq |2 (Nq + 1) (1 − ni ) ni δ(εk+q − εk + ωq ).
297
Exercise 2. 2.σ ckσ b−q .4
In Sec.
Exercise 2. d(ε). and an initial state |i having the energy Ei and being speciﬁed by the occupation numbers nkσ and Nq for electron states |kσ and phonon states |q (see Eq.
emi (b) The scattering rate corresponding to the emission processes is denoted τk .3.σ ckσ bq +
q
† gq c† k+q. Derive. Use the result to derive the 2D density of states per area. Derive. 2.EXERCISES FOR CHAPTER 3. For simplicity we restrict our study to processes that scatter electrons out of |kσ . k+q k
abs Derive the analogous expression for the scattering rate 1/τk due to absorption.108)).

while emi abs at high T we have 1/τk ≈ 1/τk . We note that since k and k are tied to the Fermi surface the largest q is 2kF . (a) To obtain realistic results we need to use the screened Coulomb or Yukawa potential for the ionic potential Vq (see Eq.42) and Exercise 1. (3. and the electron density 2 ≈ k /a . Now use polar coordinates to obtain k ∝ dεk −1 d(cos θ). (c) As usual. As we shall see in Chap.298
EXERCISES FOR CHAPTER 3 (d) Keeping nk = 1 ﬁxed for our chosen state.5). Show by dimensional analysis 2 involving the Fourier component e2 /( 0 ks ). The electrons redistribute in an attempt to neutralize the ionic potential.42) together with k = k + q change τ emi from Exercise 3. and the corresponding largest phonon energy is denoted 1 ωmax ≡ 2vD kF . With this prove that
1 −1
d(cos θ)δ(εk − εk + ωq ) ∝
2kF 0
q dq δ(εk − εk + ωq ) ∝ εk −εF
ωq . εk − εk < ωmax 0. k . since at low T . k ≈ kF . εk − εk > ωmax
Since dεk = 1
emi τk
dωq show in the limit ωmax
ωmax 0
εF how to obtain
ωmax 0 2 dωq ωq [nB (ωq ) + 1]. the Fermi energy εF . we employ the Debye model of the phonon spectrum (see Sec.
where we here and in the following do not care about the numerical prefactors.5): ωq = vD q.
.1d to
k
1
emi τk
∝
2π
k
ωq [nB (ωq ) + 1] [1 − nF (εk )]δ(εk − εk + ωq ). i. argue why thermal averaging leads to 1 2π = emi τk |gq |2 [nB (ωq ) + 1] [1 − nF (εk+q )] δ(εk+q − εk + ωq ).
q
Exercise 3.2
emi We now determine the temperature dependence of the scattering rate τk in the high and low temperature limits.
∝
2 dωq ωq [nB (ωq ) + 1] [1 − nF (εk )] ≈
(d) Show that the result in (c) leads to the following temperature dependencies: 1 ∝ emi τk T. 1/τk ≈ 0. we are mainly interested in electrons moving relatively close to the Fermi surface (why?). due to the lack of phonons. Furthermore. This immediately gives us the behavior of the total scattering emi abs abs rate 1/τk = 1/τk + 1/τk . (3. and show using q 2 = |k − k|2 that d(cos θ) ∝ q dq. 13 they succeed to do so for distances further away than 1/ks from the ion. for T T 3 + const.. n that ks F 0 1 (b) Show how Eq. 3. for T ωmax /kB ωmax /kB .e.

(a) Verify that the Hamiltonian of the two-atoms-per-unit-cell chain is H=
j
1 2 1 2 1 1 pj + Pj + K(uj − Uj−1 )2 + K(Uj − uj )2 2m 2M 2 2
∂H ∂H ˙ ˙ (b) Use Hamilton’s equations uj = ∂pj and pj = − ∂uj (similar for Uj and Pj ). (3. to ˙ ˙ ¨ obtain the equations for uj and Uj . The masses. . (d) Check that in the limit M = m the dispersion ωk in Eq.
Exercise 3. Verify the dispersion curve ωk displayed in Fig. Pj and Uj . and the displacements of the • ions are denoted m.3 we now want to ﬁnd the eigenmodes of the linear 1D chain with lattice constant a mentioned in Fig. uj+1 .. the momenta.EXERCISES FOR CHAPTER 3
299
Exercise 3. ˙ (b) Take the time derivative and note that π is the force density f . Uk ). ¨ (c) Assume the harmonic solutions uj ≡ uk ei(kja−ω t) and Uj ≡ Uk ei(kja−ω t) to derive a 2 × 2 matrix eigenvalue equation for (uk .
. Uj−1 . uj . Uj+1 . . 3.5). . . (c) Combine the equations and derive the wave equation for ρion . All spring constants are the same. The sites are numbered by j as .4
The task is to prove the Bohm-Staver expression Eq. Uj . (a) Multiply the continuity equation by the ion mass M to obtain M ∂t ρion + · π = 0. We study the situation described in Sec.4. 3.3(c).3
In analogy with the homogeneous 1D chain of Sec.9) of the one-atomper-unit-cell is recovered. respectively. while for the ◦ ions they are denoted M . namely K. The ions are treated as the jellium of Sec.28). which on the other hand is equal to the pressure gradient − P due to the compression of the electron gas following the ionic motion: ˙ π=f =− P = ∂E (0) ∂V . 3.
N
where the electron gas ground state energy E (0) is given in Eq.3(c) and the eigenmode displayed in Fig. 3. pj and uj . (3. where the light and mobile electrons always follow the motion of the slow and heavy ions to maintain local charge neutrality.2. The ionic lattice has a unit cell with two diﬀerent ions • and ◦. . (2. 3. from which the (square of the) sound velocity vs is read oﬀ:
2 M ∂t ρion −
2Z ε 3 F
2
ρion = 0. uj−1 .
where π is the momentum density. .1. 3.

(2. Because the system is charge neutral as a whole. Because we are interested in the low energy excitations we linearize the kinetic such that T = 1 2 dr M ρ0 v 2 (r).34). We deﬁne ρion (r)=ρ0 + δρion (r) ≡ ρ0 +ρ0 ion ion ion ρ(r)=ρ0 ion + δρ(r).
.
ion−el Epot = −
with the usual deﬁnition
el−el Epot =
1 2
drdr δρ(r)V (r − r )δρ(r ). each ion has charge +e. When included it gives rise to the term in Eq. The potential energy contributions which involves the ionic system are the ion-ion interaction and the electron-ion interaction
ion−ion Epot =
1 2
drdr δρion (r)V (r − r )δρion (r ).
We have not explicitly included the electron-electron interaction here. and ρ(r) is the electron density. The classical Lagrangian for this system is L0 = Tion − Vion ion 1 0 Tion = dr M ρion (r)v 2 (r) 2 ion−ion Vion = Epot .
where u is a displacement ﬁeld describing the deviation of the ion density from equilibrium. i. drdr δρ(r)V (r − r )δρion (r ). we need only take into account interactions between deviations from equilibrium.e. where v is the velocity ﬁeld of the ions. We take the case of a monovalent metal.300
EXERCISES FOR CHAPTER 3
Exercise 3. First we quantize the ion system and we start by looking at the isolated ion system.5
Electrons and phonons in the jellium model In this exercise we quantize the jellium model of the ion system in a solid and derive the electron phonon interaction in a way which is somewhat diﬀerent from the method used in the main text. derive the Lagrangian as a functional of u L0 [u] = ion 1 2 dr M ρ0 |u(r)|2 − ion ˙ ρ0 ion 2
2
drdr [
· u(r)] V (r − r )
· u(r ) . · u(r). ion
Using particle conservation show that v(r) = −u(r) ˙ and using this.

Exercise 3. only longitudinal modes exist in the jellium model1 .
1
.pj (r )] = iδij δ(r − r ). one thus has Dαβ (q) = qa qβ V (q) which only has one eigenvector with non-zero eigenvalue. In the jellium case. ion ˙ and show. 2M ρion p−q = i VM ρ0 Ω † ion bq − b−q . 2
q
q
where the polarization vector has the property that q = −q . using the well known relation H =
0 Hion
p · u − L.e. namely q .e. which you can easily check. Using the quantization from the previous exercise verify that the Hamiltonian describing the electron-phonon
In the general non-jellium case the polarization vectors are eigenvectors of the dynamical matrix D(q) in Eq. Deﬁne uq = V † 0 Ω bq + b−q . (3. What dispersion relation do you expect? Next you should quantize the ion system and ﬁnd the Hamiltonian. Ω0 = .
q
The system is quantized by the condition (canonical quantization) [ui (r).14). Furthermore. The polarization is here parallel to q i.pj (−q )] = iδij V δqq . First show that the canonical momentum corresponding to the ﬁeld u is p(r) = M ρ0 u(r). Ω is chosen such that H is diagonal. H= Ω0 b† bq + q 2 M q Explain the physics of this result. Verify that 4πe2 ρ0 1 0 ion . i.6
Bare electron.EXERCISES FOR CHAPTER 3
301
Because the Lagrangian is quadratic in u it is equivalent to a set of harmonic oscillators that describe sound waves of the ion system.phonon interaction in the jellium model. Now follow the standard scheme for diagonalization of the phonons modes. one must choose a positive polarization direction. that ˙ · u(r)] V (r − r ) · u(r )
1 = 2 = 1 2V
ρ0 p2 (r) dr + ion 2 M ρ0 ion
2
drdr [
2
q
ρ0 p(q) · p(−q) + ion 2V M ρ0 ion
V (q) [q · u(q)] [q · u(−q)] . Show that in k-space this becomes [ui (q).

1)
g(q) = iqV (q) and
ρ0 ion . and assume that nk σ = θ (kF − k ). What is the selfconsistency condition? (b) Consider the zero temperature limit. Is it given by ? EHF = εHF nF εHF .1
(a) Consider the Hartree-Fock solution of the homogeneous electron gas in a positive background.25b) and is given by εHF = εk + VHF (k) . k VHF (k) = −
k
V (k − k )nk σ .3)
Aq = bq + b† . e. 2M Ω0
(B. Use this to argue that the guess nk σ = θ (kF − k ) is in fact the correct solution. Is that true as well for the mean ﬁeld solution of the Stoner model?
. (d) The conclusion is so far that Hartree–Fock and ﬁrst order perturbation theory are in this case identical. After the mean-ﬁeld approximation the Hamiltonian can be written as H HF =
k
εHF c† ckσ + constant.g. (3) VHF (k) = − 0 π 2kF k k − kF VHF (k) is increasing monotonously with k (which you might check. which then gives k2 − k2 k + kF e2 kF 1+ F ln .2) (B.39). −q
Exercises for Chapter 4
Exercise 4.
V (q) =
4πe2 0 q2
(2)
The occupation numbers should of course be solved self-consistently.
q
(B. (2. (4. (4) k k
k
and why not? Hint: show that the correct energy reduces to E (1) given in Eq. (c) Now ﬁnd the energy of the electron gas in the Hartree-Fock approximation. graphically). k kσ
(1)
Argue why it is that in this case the Hartree-Fock energy follows from Eq.302 interaction is given by Hion−el = =
EXERCISES FOR CHAPTER 4
drdr δρ(r)V (r − r )δρion (r ) ρ0 ion V ρ(−q)V (q)(iq · u(q))
q
1 =√ V where
g(q)ρ(−q)Aq .

Then write A and C as A (T ) = (T − TC ) α. make a sketch of the speciﬁc heat of the system and show that it is discontinuous at the transition point. ∂T 2
. It also warns us that the single-particle energies derived from a mean-ﬁeld Hamiltonian are not necessary a good approximation of the excitation energies of the system. Consider a transition to a state with a ﬁnite order parameter. η. C (T ) = C.
Exercise 4. At the transition point η vanishes. η) = F0 (T ) + A (T ) η 2 + C (T ) η 4 . The square root singularity that you ﬁnd. You can assume that ∆ is real.2
The Hartree–Fock energies derived in the previous exercise have however some unphysical features. Hint: recall that CV = −T ∂2F .63) and (4. T = TC . This conclusion contradicts both experiments and the Fermi liquid theory discussed in Chap. 14.
Exercise 4. and the excitation energies. TC η (0) = Tc α . First go through the arguments that give the so-called coherence factors. uk and vk . and that A < 0 for T < TC . For second order phase transitions only even terms are present in the free energy expansion F (T. 2C
Finally. d(E).4
In 1937 Landau developed a general phenomenological theory of symmetry breaking phase transitions. (2) diverge at the Fermi level. E. Hint: start with the density of states in k space and translate to a density in energy space.3
In this exercise we calculate the density of states in the superconducting state. ﬁnd the density of states for the excitations in energy space. and use the principle of minimal free energy to show that η= −A = 2C (TC − T ) η (0) . (4. near d(∆).64). Show that it diverges at the “gap-edge”. even if the mean-ﬁeld approach gives a good estimate of the groundstate energy.EXERCISES FOR CHAPTER 4
303
Exercise 4. See also Table 4. has been conﬁrmed in great detail by experiments and is one of the many successes of the BCS theory. Secondly.6. Show that the density of states as derived from Eq. Eqs. Use this to argue that A also vanishes at the transition point. and A > 0 for T > TC . The basic idea is to expand the free energy in powers of the order parameter.

(2. ν
Exercise 5. [AB. C] = A[B . and C. useful for boson operators.4
Due to the equation of motion for operators Eq.22) we have H=
ν
εν a† aν ν
⇒
H(t) =
ν
εν a† (t)aν (t). C] + [A .
.2
ˆ ˆ Show that the third-order term U3 (t. Write down the Heisenberg equations of motion for a† and a. Solve these equations by introducing the operator α† ≡ a† + ω0 /ω . Interpret the change of the zero point energy. In the Heisenberg picture the time evolution of the electron creation and annihilation operators c† and ckσ is governed by the total Hamiltonian H = Hjel + Vel−el . C}B.4. Show the very important relations [AB. (5. t0 ) of U (t. (5.5
In the jellium model of metals the kinetic energy of the electrons is described by the Hamiltonian Hjel of Eq.
Hint: study Eqs. useful for fermion operators.1
We return to the bosonic particle described by the Hamiltonian of Exercise 1.
Exercise 5. (2.16) and (5. (5.
Exercise 5.31) derive the equation of motion for ckσ (t). C] = A{B.34). t0 ) = 3! i
3 t t0 t t
dt1
t0
dt2
t0
ˆ ˆ ˆ dt3 Tt V (t1 )V (t2 )V (t3 ) . t0 ) in Eq.18) indeed has the form 1 1 ˆ U3 (t. for some operators A. B. Express H in terms of α† (t) and α(t).304
EXERCISES FOR CHAPTER 5
Exercises for Chapter 5
Exercise 5. while the interaction energy is given by Vel−el of Eq.6) we will often need to calculate commutators of the form [AB.17) and the associated footnote. (5.19). Apply the Hartree–Fock approximation to the result. C]. C} − {A. In analogy with kσ Eq.
Exercise 5. C]B.3
Use the Heisenberg picture to show that for the diagonal Hamiltonian H of Eq. (5.

4
In this exercise we consider the conductivity of a translation. This means that the conductivity σ(r. Express x in terms of a and a† and calculate the linear response result for the expectation value x(t) .4.EXERCISES FOR CHAPTER 6. and η = 0+ is a positive inﬁnitesimal. where f (x) is any function with a well behaved Taylor expansion around x = 0. t) .and rotational-invariant system. Express your result in both real space and momentum space.2
The spin susceptibility measures the response to a magnetic ﬁeld. see Sec.3
1 We study integrals of the form −∞ dx x+iη f (x). t) · S(r). Neutron scattering experiments are the main source for obtaining experimental information about the distribution of spins in condensed matter systems. Show that 1 in this context x+iη can be decomposed as the following real and imaginary parts ∞
1 1 = P − iπ δ(x).3. Find the response to linear order in B for the induced spin density in the material. Suppose that a piece of some material is perturbed by external magnetic moments. S(r. but now we treat this force as a time-dependent perturbation H = f (t) x.1 we consider a harmonic oscillator inﬂuenced by an external force f (t).
305
Exercises for Chapter 6
Exercise 6.
Exercise 6. Argue that this result is in fact exact. These moments could for example be in the form of a neutron beam in a neutron scattering experiment. The perturbation is in this case given by H = −gµB dr Bext (r. 1.
where S is the spin density operator S(r) = Ψ† (r)sΨ(r). x + iη x Here P means Cauchy principle part: P 1 dx f (x) ≡ x −∞
∞
1 dx f (x) + x −∞
−η
∞ η
1 dx f (x). x
Exercise 6.
Exercise 6. for example by considering the equation of motion for x(t) .1
As in Exercise 5. r ) is a function of r − r only and that
.

(8. where J(q) is the particle current operator in momentum space.1
Verify that the self-consistent equations in Eqs.306
EXERCISES FOR CHAPTER 7
the conductivity tensor is diagonal with identical diagonal components. Derive the expression for the particle current operator in this limit. From this you can derive obtain the long wavelength conductivity σ αβ (0. t). namely Eq. ω). ωm
How does this ﬁt with the Drude result (13.
Exercise 6.14b). Find the relation between the conductivity. ω) = σ(q. (8. ωτ → ∞)? How does the conclusions change for an interacting translation-invariant system?
Exercises for Chapter 7
Exercise 7. and that r|H|φ = H(r) r|φ Which you can see for example by inserting a complete set of eigenstates of H. kσ
kσ
and show that it is time-independent in the Heisenberg picture.22) in fact is identical to the Green’s function by showing that it obey the same diﬀerential equation. t) = 1 m k c† (t) ckσ (t). ω) and the correlation function [J α (q. (8. J α (−q. J(0.13). First consider the conductivity of a non-interacting electron gas at long wave lengths.17) both are solution to the Schr¨dinger equation in Eq. 0)] .2
In this exercise we prove that the propagator in Eq. ω)E(q. Show that in the Fourier domain Je (q. o
Exercise 7. (8.e. Hint: diﬀerentiate (8.16) and (8. where the impurity induced scattering time τ tends to inﬁnity (i. q → 0. i.5
Consider again the conductivity of a translation-invariant and rotational-invariant system. ω) = iδαβ ne2 .e.
. σ(q.42) in the clean limit.22) with respect to time using that the derivative of the theta function is a delta function.

Now calculate the tunneling current-voltage characteristics.66). (4.
Exercise 7.
. t) = −iθ(t) {ck↑ (t). 6. Hints: use the onedimensional version of the particle current operator Ip (x) = mL k+
kqσ
q 2
c† ck+qσ eiqx . Plot the results of I and dI/dV versus V . G> (r. ω) = A(ν. r . ω) = A(ν.65).
Exercise 7. ω) nB (ω). assuming the tunneling matrix element to be approximately constant. which you see from the Hamiltonian Eq.
Exercise 7. kσ
The result for the dc conductance does not depend on where the current is evaluated (why?).42). From Sec.EXERCISES FOR CHAPTER 7
307
Exercise 7.5
The tunneling density of states for a superconductor has a characteristic shape which you ﬁnd in this exercise. Can you suggest an experiment (at least in principle) that measures this propagator. t − t ) = −iθ(t − t ) Ip (xt). ω) [1 + nB (ω)] . First ﬁnd the retarded Green’s function GR (k ↑.5 to ﬁnd that ΠR (x − x .4
Eqs. ω) similar to Eq. iG< (ν. 8. c† } k↑ by expressing the c and c† operators in terms of the diagonal γ-operators called bogoliubons given in Eq.3
Find the greater propagator.6
In this exercise we shall calculate the dc conductance of a perfect one-dimensional wire. Show that the corresponding results for bosons are iG> (ν. (8. ω ΠR (x − x . Now you can use the method in Sec. (4. t − t ) = − iθ(t − t ) × k+ q 2
2 2
mL
nF (εk ) − nF (εk+q )
kqσ
ei(εk −εk+q )(t−t ) eiq(x−x ) . Ip (x t )
where Ip is the operator for the particle current through the wire. Once you have done that the problem is reduced to ﬁnding the Green’s function of a free particle.and two dimensions. but now in one.50) are valid for fermions.3 we have that the conductance is given by G= ie2 R Π (ω). (8.

where i and j can be both be either 1 or 2 and where GR (ij.308
EXERCISES FOR CHAPTER 8
Setting x = x ﬁnd ΠR (0. ω). You may need to know the integral π ds 1 − s2 sin2 (xs) = 8x [x − J1 (2x)]. Don’t j forget to interpret the result. What will change at larger temperatures? The oscillations that you will ﬁnd are called Friedel oscillations.
Do the q-integral ﬁrst and ﬁnd lim Im ΠR (ω) = ωπ
2
m
2
ω→0
m
1 2πL
−
kσ
∂ nF (εk ) ∂εk
k2 |k|
ω 1 = .22) and (9.2
Derive Eqs. Show that lim Im Π (ω) = ωπ
R 2
ω→0
mL
−
kqσ
∂nF (∂εk ) q δ(εk − εk+q ) k + εk 2
2
.7
Consider a 2D electron gas in the xy plane conﬁned to the strip 0 < x < L. t − t ) = −iθ(t − t ) {ci (t). What is the electron density as a function of the distance x from the left edge? Take for simplicity T = 0. you ﬁnd the famous result for the conductance G of a perfect 1D G= 2e2 .
Exercise 8.
. Hints: Use standing waves in the x-direction fulﬁlling the proper boundary conditions. and assume quasi-continuous states with periodic boundary conditions in the y direc† 2 tion. (9.21). 1 2 1 2 Find the Green’s function GR (ij. ω) and study it in the low frequency limit. −βµ + 1 π e In the limit µ channel kT .20) and (9.23) by diﬀerentiating the Green’s functions in (9.
Exercises for Chapter 8
Exercise 8. c† (t )} . where the ν-sum runs√ over the appropriately normalized states |ν . Find the x-dependent density as n(x) = dy ν cν cν | xy|ν | .1
Consider a physical system consisting of fermions allowed to occupy two orbitals. The Hamiltonian is given by H = E1 c† c1 + E2 c† c2 + tc† c2 + t∗ c† c1 . Use the equation of motion method. h
Exercise 7.

kd Derive the equation of motion for F R . Sketch the resulting dI/dV using the expression derived in Chap.3
Consider an atom on a metal surface.30b) to obtain the following equation of motion for GR (d ↑. The electronic states of the atom will hybridize with the conduction electrons in the metal. k k↓ ↓
The last term contains a new type of processes giving rise to higher order correlations (corresponding to spin ﬂips). d† (t )} . (9. If we assume that only a single orbital couples to the metal states. ω): ω + iη − εd + µ − ΣR (ω) GR (d ↑.22). nd↓ ] no new Green’s functions are generated. and (9. Here we go one step further and derive an equation of motion for DR (d ↑. and not the tunneling between atom and metal. In Eq.2. 8 for the tunnel current and the mean ﬁeld expression for the d electron Green’s function. (9. Since the atom is strongly coupled to the metal surface the bottleneck for the current is the tunneling from STM to atom.3. t − t ) = −iθ(t − t ) {(nd↓ ck↑ )(t).
. The ﬁrst term generates a two-particle Green’s function denoted F R given by F R (kd ↑.2 and 9.17). Instead F R is coupled back to DR . nd↓ ] d↑ = nd↓ [Hhyb .18). When you do that the diﬃcult commutator is [Hhyb . t) and truncate it at the two-particle Green’s function level. This constitutes our new and improved approximation. (9. t) is deﬁned in Eq. and show that if you again neglect the term [Hhyb . and the model was then solved at the level of single-particle Green’s functions. ω) = 1 + IU DR (d ↑. (8.60). d↑ ] + [Hhyb .
Exercise 8. which we can describe by a tunneling Hamiltonian as in Eq.2.2. 9. ↑ Note the similarity between F R and the single-particle function GR (kdσ) of Eq. When a scanning tunneling microscope (STM) is placed near the atom current will ﬂow from the STM tip through the atom to the metal. (9. d↑ ] +
kσ
t∗ c† d↓ − tk d† ck↓ d↑ .26). ω). 9.23). It is therefore a good approximation to assume that the atom is in equilibrium with the metal. derived in Sec. Start by combining Eqs. nd↓ d↑ ] = nd↓ [Hhyb .21).28) it was approximated by a product of two single-particle Green’s functions. described by Eq. and to use tunneling theory for the current between tip and atom.4
In this exercise we improve the solution of the Anderson model presented in Secs. and it is therefore omitted. Thus we get a better approximation which takes pair correlations into account. then the atom and the metal can be described by the Anderson model Hamiltonian Eq.EXERCISES FOR CHAPTER 8
309
Exercise 8. (9. First ﬁnd the diﬀerential equation for DR by diﬀerentiation with respect to t. (9. (9. The two-particle Green’s function DR (d ↑.

310
EXERCISES FOR CHAPTER 10
Inset this result into the equation of motion you derived for DR above.5) and (11.3 is changed. R (ω) ω + iη − εd + µ − Σ ω + iη − εd − U + µ − ΣR (ω)
Interpret this result physically. Note that both τ and τ are greater than zero.1
Find the Fermi-Dirac distribution by starting from the Matsubara Green’s function and setting τ = 0− .9).80) to Eq. Finally. Suppose now that the
. for example by considering how the result of Exercise 8. The Dyson equation for otherwise free electrons scattering against an external potential is written in Eqs. τ − τ ) = δ(r − r ) δ(τ − τ ).3
According to Eq. solve for GR and show that the result is GR (d ↑) = 1 − nd↓ nd↓ + . 2m
Show (10. and show that the resulting equation for DR is ω + iη − εd + µ − U − Σ(ω) DR (d ↑) = nd↓ . (10. (10.2
Repeat Exercise 7.
Exercises for Chapter 9
Exercise 9. (10.85). ikn ) = nF (εν ) ikn
−
How would you calculate cν c† ? ν
Exercise 9.
Exercise 9.6 but this time using the imaginary time formalism. (11.63) the equation of motion for the Matsubara Green’s function of a free particle is −∂τ − p2 0 + µ Gσ (r − r .1
Single impurity scattering.
Exercises for Chapter 10
Exercise 10. τ = 0− ) =
1 β
0 e−ikn 0 Gσ (ν.39) by Fourier transforming this equation. Then show that c† cν ν
0 = Gσ (ν. Use the procedure going from Eq.

In for example semiconductor heterostructures one can make quantum-well systems which to a good approximation can be described by a one-dimensional model of free electrons with two tunneling barriers. we can in fact extract the transmission and reﬂection amplitudes t and r. ikn ) ﬁrst and insert that in the Dyson equation for G(xx . σ σ σ where eiφ(x. ω − εk + µ + iη ivω
(do you agree?) where kω = 2m(ω + µ) and v =ω ∂εk /∂k|k=kω . and (2) x → a1 → a2 → a1 → a2 → x. which is determined by the calculation. The transmission is unity when these two paths interfere constructively – as does paths with any number of trips back and forth in the “cavity”. U (x) = U0 δ(x).x ) is a phase factor. e−ika2 · 1 − α −αeiθ −αeiθ 1 − α
−1
·
eika1 eika2
.EXERCISES FOR CHAPTER 10
311
electrons are conﬁned to move in one dimension and that the external potential can be represented by a delta-function impurity potential. Derive the same result using the following simple argument involving two paths for an electron to go from x to x: (1) x → a1 → a2 → x. ω) = Gσ (xx . ikn ).
where H0 is the Hamiltonian for free electrons in one dimension. ω) = σ eik(x−x ) 1+α ivω e−ika1 .
where k = kω (see previous exercise) and where α = U0 /iv and θ = k(a2 − a1 ). ω + iη) = σ
1 L
k
eik(x−x ) 1 ikω |x−x | = e . Since the retarded Green’s function tells us about the amplitude for propagation from point x to point x.
Hint: solve for G(0x .2
Resonant tunneling. ikn )
U0
0 Gσ (0x . ikn ) = Gσ (xx . Write H in x−space and ﬁnd a formal expression for the Matsubara Green’s function using Dyson’s equation. The Hamiltonian is then given by
∞
H = H0 +
−∞
dx ρ(x) U0 δ(x − a1 ) + δ(x − a2 ) . ikn ) + Gσ (x0. From the Dyson equation ﬁnd the retarded Green’s function for x < a1 < a2 < x: GR (xx . ikn ). For x < 0 we have GR (xx . Use this to show that the transmission is unity for the particular values of θ satisfying α = i cot θ. Show that in this case the solution of the Dyson equation becomes
0 0 Gσ (xx . Find r and t and discuss the phase shifts that the electrons acquire when they are scattered. ikn )
1−
U0 0 Gσ (00.x ) G0R (xx .
Exercise 10.
. Here we simplify it somewhat further by representing the tunneling barriers by delta-function potentials situated at a1 and a2 . To ﬁnd the retarded Green’s we thus need the unperturbed Green’s function. ω) θ(−x). ω) = t G0R (xx . which is
0 G0R (xx . ω) θ(x) + 1 + reiφ(x.

As in Eq. (9.14) use the Feynman rules to expand the denominator and the numerator. ¯ σ We introduce the following rather obvious diagrammatic notation for the Matsubara Green’s functions and interactions: ≡ ≡ ≡ G 0 (dσ) G (dσ) G 0 (kσ) U ≡ tk ≡
k
tk
U
. a) = −
0
. (2) Draw topologically identical diagrams only once and multiply with the number of them. (9. the sum of the hybridization Eq. while the interaction part is σ kσ MF given by Hint = Hhyb + HU . 9. We employ the mean-ﬁeld approximation given by Eq.34): Π0 (q.27) where the σ spins only interact with the average density nd¯ of the opposite σ spins. We wish to derive the Dyson equation Eq. which in the interaction picture in the presence of the particle-particle interaction W (r − r ) becomes: ˆ ˆ ˆ Tτ U (β.1
Matsubara frequency summation.2.3
Feynman diagrams and Dyson’s equation for the Anderson model. Use the rule Eq. (12. (12.312
EXERCISES FOR CHAPTER 11
Exercises for Chapter 11
Exercise 11.30) and (12.29) using Feynman diagrams. but now to second order in W .17) and on-site repulsion Eq. 0)
0
G(b. 0) = Tτ exp −
β 0
ˆ dτ W (τ ) .54) for summing over functions with simple poles to perform the Matsubara frequency summation appearing in the following diagrams of Eqs. We return to Anderson’s model for localized magnetic moments in metals. 0) Ψ(b) Ψ† (a) ˆ U (β.2
The cancellation of disconnected diagrams in G(b.27).
ˆ with U (β. (9. iqn ) ≡ ΣF (k. and show explicitly the cancellation of the disconnected diagrams. (1) Start with the simpler denominator (how many terms?). (10. a).
Exercise 11. We study the one-particle Green’s function. (3) Get most of the diagrams in the numerator by cutting open and stretching out a Fermion line in the diagrams from the denominator (how many terms?). The unperturbed Hamiltonian is given by H0 = σ (εd − µ) d† dσ + kσ (εk − µ) c† ckσ . Hints. (9. see Sec. ikn ) ≡ σ
Exercise 11.

2. (9.76) does not contain Planck’s constant and is therefore not a quantum object. Lz ). (13.
313
We write the diagrammatic expansion (here shown up to second order in |tk |2 and U ) for the full d-orbital spin up Green function G (dσ) as: G(d¯ ) σ tk = G (dσ) G 0 (dσ) + G 0 (dσ) G 0 (kσ) G 0 (dσ) G(d¯ ) σ tk + G 0 (dσ) G 0 (kσ) G 0 (dσ) G 0 (kσ) G 0 (dσ) G(d¯ ) σ tk + t∗ k U U t∗ k tk t∗ k + G(d¯ ) σ tk t∗ k +.2)..
Exercise 12. Derive ωp from the following purely classical argument.EXERCISES FOR CHAPTER 12. In the following exercises we consider a translationinvariant electron gas in two dimensions fabricated in a GaAs heterostructure (see Sec. Give a physical interpretation of the resulting motion of the electron gas. U U t∗ k + U G 0 (dσ) G 0 (dσ) G(d¯ ) σ
G 0 (dσ) G 0 (dσ) G 0 (dσ)
+
G 0 (dσ) G 0 (kσ) G 0 (dσ) G 0 (dσ)
G 0 (dσ) G 0 (dσ) G 0 (kσ) G 0 (dσ)
Express the self-energy as a sum of diagrams using a deﬁnition analogous to Eq.
. Treat the ions as an inert.30b).1
A classical treatment of the plasma oscillation. Find the equation of motion for the coordinate ξ using Newtons law and classical electrostatics. The electron gas is then released.2
Interactions in two dimensions. leaving the ion jellium ﬁxed.18). Use the obtained Dyson equation to verify the solution Eqs. (12.30a) and (9.19) Dyson’s equation graphically.3.
Exercises for Chapter 12
Exercise 12. The electronic plasma frequency ωp ≡ ne2 /m 0 introduced in Eq. Consider an electron gas of density n conﬁned in a rectangular box of length Lx in the x direction and having a large surface area Ly Lz in the yz plane (Lx Ly . and derive in analogy with Eq. The resulting system resembles a plate capacitor. (12. The tedious work with the equation of motion has been reduced to simple manipulations with diagrams. Imagine now the electron gas being translated a tiny distance ξ in the x direction (ξ Lx ).. charge compensating jellium background.

z) = 1 eik·r ζ0 (z). the relative permittivity is εr = 13. and low temperatures. RPA 0 with q = (qx .314
EXERCISES FOR CHAPTER 12
The electron mass for this material is m∗ = 0. Find the relation between kF and the electron density.
Hint: use 0 dθ cos (α cos θ) = πJ0 (α). qy ). Show that in 2D the static RPA screened interaction at small wavevectors. For a strictly 2D system. while the electron density ranges from n2D = 1 × 1015 m−2 to 5×1015 m−2 .σ ck −q. n = 0 in Eq. ky ) and r = (x. 0) RPA
e2 . Consider a translation-invariant electron gas in two dimensions fabricated in a GaAs heterostructure.067 m. Write down the interaction part of the Hamiltonian and show that it is of the form H 2D = 1 2A
† W 2D (q) c† k+q. |ζ0 (z)|2 = δ(z). i.4
Static screening in two dimensions. iqn ) = 1 − W 2D (q) χ2D (q. 2D 0 (q + ks )
. Is it larger or smaller than kF for n2D = 2 × 1015 m−2 ?
Exercise 12. q kF . 2D n2D . 0) ≡
W 2D (q) = 2εr ε2D (q. For such a system the RPA dielectric function is given by ε2D (q.50). i. qy ) and where χ2D (q. show that W 2D (q) =
π
e2 2εr
0
q
.σ ck . iqn ) is the 2D version of the 3D pair bubble χ0 0 given in Eq.e. the plasmon dispersion relation is ω = vF ks q/2.
Exercise 12.σ ck. where J0 is the Bessel function of the ﬁrst kind of order zero. (13. while the electron density ranges from n2D = 1 × 1015 m−2 to 5×1015 m−2 .σ kk q σσ
where q = (qx . is given by
2D WRPA (q. Lx Ly
where k = (kx .e. y). kB T εF . Show that at low temperatures.3
Plasmons in two dimensions. and long wave lengths. iqn ). kB T εF . Express ks in terms of the parameters of the electron gas. while ζ0 (z) is the lowest eigenstate in the z direction. where ks is the ThomasFermi screening wavenumber in 2D. 2D 2D q ω/vF . the relative permittivity is εr = 13. The electron wave function for the two dimensional electron gas is restricted to be ψk (r. (2. The electron mass for this material is m∗ = 0.067 m.20).

a) (you do not have to show that). What is the physical interpretation of ∂k ω (k0 )? In conclusion. Then apply Wick’s theorem to obtain the analogue of Eq.e.g. (13. 0) eq as stated in Eq.19). Write down the expression for χ(b. Following arguments similar to those of Eq. (14.53). (14. (12. Are the plasmons damped in the region q < kF in GaAs with the parameters given above?
Exercise 12.5
Damping of two dimensional plasmons. In the estimate you can use the small-q expressions for the plasmon frequency and the polarization.6
Deriving the Feynman diagrams for χ(˜). In the real space formulation χ(b.18). (13. We are not asking q for detailed calculations. τ ) ρ(−q. τ ) = − V Tτ ρ(q. Then apply Wick’s theorem in this situation to obtain the starting point for the diagrammatic expansion directly in q-space. vk = ∂k ωk = 1 ∂k εk .
315
Exercise 12.
.8).EXERCISES FOR CHAPTER 13. i. 2 (∆k)2
Taylor expand ω (k) to ﬁrst order and show that the wave packet can be written as ψ(r. τ ) for a translation-invariant system. The task is to understand the arguments q leading to the diagrammatic expansion for χ(˜) given in Eq. with f (k−k0 ) = exp − (k−k0 )2 . a). and write this in a form analogous to Eq. ω ) in Exercise 6.1
Semi-classical motion. a wave function fairly localized in both space and momentum space: ψ(r.8) for G(b. argue with the help of Appendix A that for a 1 translation-invariant system χ(q. a) ≡ − Tτ ρ(b)ρ(a) eq = − Tτ Ψ† (b)Ψ(b)Ψ† (a)Ψ(a) eq .
Exercises for Chapter 13
Exercise 13.16) it can be shown that the numerator also cancels in the case of χ(b. If the quasiparticles behaves like non-interacting particles why is then the number of quasiparticles conserved on the semi-classical level? To answer this question we introduce the concept of a wave packet. The electron-hole pair continuum is the region in q − ω space where Im χ2D = 0. e. you may start with χ(q.9). t) ≈ ei[k0 ·r−ω (k0 )t] F (r − ∂k ω (k0 ) t). Alternatively. Please note that ρ eq = 0 due to charge neutrality. and (14. Finally.17). a) analogous to Eq. We study Eqs.4.58). t) = dk f (k−k0 ) ei[k·r−ω (k)t] . the wave packet has the energy εk and the velocity vk given by εk = ωk . that you found above. (12. (12. where F is some envelope function. (12. see also the form of σ(q. Find the condition for the plasmons not to be 0 damped by single-particle excitations for q < kF .

while at higher energies. Combine the two classical expressions for the power.316
EXERCISES FOR CHAPTER 13
For external forces F(r. discuss the following simple model for A(k. ω ). Instead of using RPA. q ) from X-ray scattering on sodium. ω ) nF (ω ).
In the so-called impulse approximation for Compton scattering. Show ˜ that ∞ 1 ∞ 1 ∞ dω I(q. q) = N (ω1 . For an interacting electron gas discuss the spectral function A(k. ω ) containing the essential features. ω) is the spectral function and q ≡ mω /q − q/2. the intensity I(ω1 . we can through the power Pk absorbed by the wave packet centered around k deduce the time evolution of k as follows. to show ˙ 1 ˙ k = F. The value of Z can be inferred from X-ray Compton scattering on the electron gas. q2 ˜ q2 ˜ q q −∞ 2π
2m 2m
where A(k. q ) ≡ I(ω1 .
21#! ) % % $ 21#! ) $ $ &) & $ $ &) & $ 0& $ &0 $ &( $ P RSG 6 Q 8 7P I H &( $ G 3E6 W WP 5 Y V X W 6 P FX &' $ 9 3 G 5 7EF 5 D C 5 &' $$ $ B A @ 9 78 6 5 34 " "# ! & % " "# ! & % $ U$ ¢ T $ $ ¢
©¤¥ § ¦¥ £ © ¤ £¨§ ¦ £ £
¤¤ £¥
¢ ¡
. see Fig. (14. I(ω1 . ω . t) = − V (r. ω .
Exercise 13. (b) contains an experi˜ mental determination of I(q. ω . Pk = F · vk and Pk = εk . q). q) of incoming photons of energy ω1 being scattered with the energy and momentum loss ω and q. 2m m
We omit the explicit reference to the ﬁxed ω1 and work with I(q. q ) ∝ ˜ dεk nk = dεk A(k. εk < 4εF . ω ) dk nk δ(ω + εk − εk+q ) ∝ dk nk δ ω − 1 2 1 q − q·k . (a). t) varying slowly in space and time. The experimental result ˜ is compared to theory based on RPA calculations of A(k. Fig.58) and use it to calculate the distribution function nk . Demonstrate the existence of a Fermi surface characterized by the renormalization parameter Z. At low energies. εk > 4εF . is proportional to the number of scattering events on all electrons fulﬁlling the simple kinematic constraint: conservation of energy and momentum. ω ) in Eq. a renormalized quasiparticle pole of weight Z coexists with a broad background of weight 1 − Z.2
Measuring the discontinuity of the distribution function. respectively.

The Boltzmann equation with inclusion of e-e scattering therefore reads ˙ ∂t (nk ) + k ·
k nk
+ vk ·
r nk
=−
1 τk
. long wave limit that ε(0. x
k
.EXERCISES FOR CHAPTER 14.
Exercise 13.
collisions
In the homogenous and static case. and for which is it reﬂected? From X-ray diﬀraction we know that the unit-cell of Na is body-centered cubic (i. no renormalization occurs. that the right hand side is also zero if we use n = nF . absence of external forces. 0. ω ) = Zk 2πδ(ω −ξk ) + (1−Zk ) π θ(W −|ω |).3
Detailed balance. for k > 2kF .
Exercise 13.4
Why are metals shiny? According to Eq. 0)GA (k. and explain why (polished) metals appear shiny. (c). one atom in each corner and one in the center of the cube) with a side-length of 4. k. A It is observed that Na is transparent for UV-light with a wavelength shorter than 206 nm. Hint: Each Na atom donates one electron to the conduction band. i. Set D = 0 (0. i.e. the left hand side is expected to be zero.1
The integral equation for the vertex function in the Born approximation. The conductivity is in the weak scattering limit given by σxx = e2 1 π V kx m GR (k. Show that the usual Fermi-Dirac equation solves the Boltzmann equation in this case. Consider a monochromatic electro−iωt e incident on a metal occupying the half-space x > 0.45) expresses the time between scatterings assuming some unknown distribution function n(k).
ω 2 −ω 2
Exercises for Chapter 14
Exercise 14. Hint: show and use that nF (ε) [1 − nF (ε )] exp (β(ε − ε )) = [1 − nF (ε)] nF (ε ).
317
Here W is the large but unspeciﬁed band width of the conduction band. 0).e. for k < 2kF 1. and the quasiparticle is in fact the bare electron: A(k. (13.75) we have in the semiclassical high 2 frequency. ˆz magnetic wave with E = E(x)e Use the high-frequency limit of Maxwell’s equations in matter. W Zk = Z. For which frequencies does the wave propagate through the metal. Explain Fig. ω) = 1 − ωp /ω 2 . (14. The scattering life time in Eq.e. This exercise deals with the Kubo formula method applied to impurity scattering in metals.23 ˚. 0)ΓRA (k. Explain this. ω)E and prove that 2 E(x) = pc2 E(x). Hint: you may need × ×E = − 2 E.

In this more realistic case. x
k
1. 0.e. i. k . 0) = x kx 1 + m V nimp |u(k − k )|2 GR (k . k. Show that the retarded impurity averaged Green’s function in the Born approximation decays exponentially in time and given a physical interpretation of this result. m
where τ0 is the Born approximation life time
−1 τ0 = 2πd(εF )nimp |u0 |2 . Now σ= e2 nτ tr .318
EXERCISES FOR CHAPTER 14
In the Born approximation the Dyson equation for the vertex function is ΓRA (k. you will ﬁnd for |k| = kF that ΓRA (k) = with |k | = kF and show that dΩ = d(εF ) k + m 2 dΩ |u(k − k )|2 ΓRA (k ). τ0 .
. 0. Now relax the assumption of short range scatterers but assume instead that u(k − k ) −1 is slowly varying on the scale given by the width of the spectral function. 0)ΓRA (k .
2. If the impurity potential is short ranged argue that we can approximate it by a constant |u(k − k )|2 ≈ |u0 |2 . m
e2 nτ0 . m k·k dΩ |u(k − k )|2 1 − 2 4π kF .
where τ tr
−1
= 2πd(εF )nimp
Hint: use the ansatz ΓRA (k) = (k/m)γ.
Exercise 14. Explain the physical meaning of the last term and why it does not appear in the result for point-like impurities. Prove that in this case ΓRA = x and use this to ﬁnd that σ= kx .
dφ dθ sin θ is an integration over the angle of k . 0).2
Life time of the Green’s function in the Born approximation. 0)GA (k .

ikn .
(B.EXERCISES FOR CHAPTER 14
319
Exercise 14.e. ω) is
ζ(Q. ω) ≈ 1 + iωτ0 − Dτ0 Q2 . i. ω) . Show that the frequency provide a small Q cut-oﬀ in the conductivity correction and try to explain why. ε) ≈ .5) ∗ ω − ξk + i τ2∗ If we include the real part in this way (contrary to Sec. The only change in the formula for the weak localization is through the Cooperon which becomes instead C RA (Q.
Exercise 14. ω) = W0 ζ(Q. 0). The impurities only enter in the scattering time. the low frequency ωτ0 of ζ(Q. ω)GA (p.g. Let us for instance consider the lowest order self-consistent Born approximation. If we think about impurity scattering from a microscopic point of view the self energy has both a real and an imaginary part. ω) = |u0 |2 GR (p − Q. V p 1 and long wavelength QvF τ0 1 limit
Show that in this case. when expanding the Green’s function near the Fermi surface we can write Z GR (k. ω) 1 ζ(Q. In this case the self-energy reads
1SCBA
(k.4
Mass renormalization in Drude formula? The mass that appears in the Drude formula is the bare electron mass (possibly including bandstructure eﬀects).3
Weak localization at ﬁnite frequency.4)
Note that it is G and not G0 that enters in the self-consistent approximation. 1 − ζ(Q. vk = k/m and the impurity potential in the collision term. The Drude formula can be derived from e. Here we consider the weak localization correction at ﬁnite frequency.3 and Exercise 5. The real part leads to a renormalization of the mass and to a renormalization of the spectral weight at the Fermi surface. ikn ) = nimp
k
|uk−k |2 G k . the Boltzmann equation. where the mass enters through the velocity. (B. 14. where it is neglected) does it lead to a renormalization of the Drude formula?
.

320
EXERCISES FOR CHAPTER 15
Exercises for Chapter 15
Exercise 15. x > 0. How would the result look if the transverse conﬁnement was given by a square well?
. so that the distribution function for electrons in the states ψk is given − by nF (εk − µL ). where the potential in the transverse direction is a parabolic conﬁnement. nn Using a model electron waveguide.
where Tn (E) is the transmission probability of a given mode Tn (E) = n t∗ tn n . 2m ∂x2
2
eikx + re−ikx .2
Landauer-B¨ ttiker formula at ﬁnite temperatures. the transmission coeﬃcient is supposed to be 1 Tn (E) = Θ(E − En ) = Θ(E − (n + ) ωT ). Hints: vk = 2mi k k
Exercise 15.05. Suppose now that the two ends of the wire have diﬀerent + chemical potential.c. For a general mesoscopic system u show that the linear conductance at ﬁnite temperature is generalized to G= 2e2 h
∞
dE −
n 0
∂nF (E − µ) ∂E
Tn (E). L
and likewise for ψ − . 1 √ teikx . 2 where ωT is the frequency of transverse oscillation. h 1 + (V0 / vF )2
εF . Plot for example G(h/2e2 ) versus µ/ ωT for two diﬀerent temperatures: kT = (0.) = |t|2 = 1 − |r|2 with t = (1 + iz)−1 . x < 0. while the distribution function for electrons in states ψk is nF (εk − µR ). Show that the current at low temperature becomes I = 2e
k>0 + − vk nF (εk − µL ) − vk nF (εk − µR ) =
1 2e2 V.15) ωT . z = mV0 / k. 0. Show that ψ (0+ ) − ψ (0− ) = 2m
2
V0 ψ(0)
and use it to ﬁnd r and t. where V = µR −µL e + ~ ( ψ+ ∗ ∂x ψ+ − c. Find an expression for the conductance G and plot the G as a function of µ.1
Conductance of a delta function barrierAs a model system of a mesoscopic 1D channel take the following Hamiltonian H=− Consider a scattering state
+ ψk (x) = 1 √ L
∂2 + V0 δ (x) .

k >kF . In Chap.
321
Exercises for Chapter 16
Exercise 16. i. We wish to ﬁnd the energy of a pair of electrons added to a ﬁlled Fermi sea state. which is also the starting point used by BCS. i.2
Cooper’s instability In this exercise we shall see that an attractive electron-electron interaction leads to an instability of the Fermi surface. µ + ωD ]. H = H0 + H (1a) H0 =
kσ
(εk − µ) c† ckσ . BCS model Hamiltonian.
. and with interactions according to (2). k −k
(3)
We look at the diﬀerence between the two situations: 1) The electron pair is added to the Fermi surface. (16.1
Phonon Green’s function Prove Eq. k −k (2)
Variational calculation of FS with an added pair of electrons.
Exercise 16.e.EXERCISES FOR CHAPTER 16.e. This observation lead Cooper to study the following model Hamiltonian. (2) becomes H = −V0
k. the sum is further restricted to involve only states outside the Fermi sea.
c† ↑ c† ↓ c−k↓ ck↑ . Thus Eq. in a shell around the Fermi surface. where q is the exchanged momentum and vs is the sound velocity. In order to separate the eﬀect of the interaction on the Fermi sea and on the extra pair of electrons. The scale of this energy is given by the Debye energy.7). i. which means that we choose k = −k and σ = −σ . 2) The electron pair forms a coherent superposition of pairs not necessary at the Fermi surface.e. for energies smaller than qvs . After relabelling we have H = −V0 c† ↑ c† ↓ c−k↓ ck↑ .
(1b)
H =−
V0 2
(1c)
Here the sum is restricted such that all initial and ﬁnal states lie in an interval given by [µ − ωD . Anticipating the physical idea that the electrons due to the attractive interaction will form pairs with zero total momentum and spin. The pairs are thus supposed to consist of electrons with opposite momentum. According to the variational principle the lowest energy of the two is closest to the groundstate energy. we look speciﬁcally at the interaction between such pairs. kσ
† c† k+qσ ck −qσ ck σ ckσ . with |k|. |k | = kF and energy equal to zero. 16 it is shown that the electron-phonon interaction leads to an eﬀective electron-electron interaction. It is attractive for small frequencies.

Find that (reinserting ) E = −2 ωD exp (−1/V0 d(EF )) .1
The Josephson eﬀect. Supercurrent in the equilibrium state. (4) k↑ −k↓
k
Show that αk satisﬁes the following equation. This exercise deals with the supercurrent across a tunnel junction.e. (2εk − EF ) − E
(6)
In order to ﬁnd the energy E you should make use of the following hierarchy of energy scales E ωD EF . (1) kp kσ
kp
where the electron operators for the two sides are called c and f .322
EXERCISES FOR CHAPTER 17
For situation 2 we start by an Ansatz wavefunction. which is a superposition of socalled Cooper pairs |ψ = αk c† c† |F S . two superconductors separated by an insulator. αk (2εk − EF ) − V0
k >kF
αk = Eαk . i.
(5)
and that this leads to a condition for E given by 1 = V0
k >kF
1 . Consider a tunnel junction between two superconductors. The tunnel Hamiltonian is † HT = tkp c† fpσ + t∗ fpσ ckσ . the so-called Josephson eﬀect. In the following we assume for simplicity that in the energy range of interest the tunnel matrix element depends weakly on the states k and p. therefore tkp ≈ t (2)
. respectively. (7) where the validity of the ﬁrst one of course must be checked at the of the calculation. We apply the relations to study the current-voltage characteristic of a tunnel junction in the so-called resistively shunted Josephson junction model. (8) Discuss the following two important issues: • Why does this result indicate an instability of the Fermi surface • Could this result have been reached by perturbation theory in V0 ?
Exercises for Chapter 17
Exercise 17.

f−k↓ fk↑ . −τ ) + c. (4) by the transformation c → eiφc /2 c.c. and ∆ is here taking to be real. F is the free energy and φ = φc − φf is the phase diﬀerence. ikn ) + c. ikn )G12 (p. Verify the following steps G21 (k. The order parameter of each side of the junction have diﬀerent phases. In the following we calculate this so-called supercurrent to second order in the tunneling amplitude.
(3a) (3b)
Hf =
where the two superconductors are assumed to be equal.EXERCISES FOR CHAPTER 17 The Hamiltonian for the two sides are the usual BCS Hamiltonians Hc =
kσ
323
ξk c† ckσ − ∆eiφc kσ
k † ξk fkσ fkσ − ∆eiφf kσ k
c† c† − ∆e−iφc k↑ −k↓
k † † fk↑ f−k↓ − ∆e−iφf k
c−k↓ ck↑ . (2). f → eiφf /2 f. τ )G12 (p. ikn ) ∆d(εF ) 2 ∆d(εF ) 2
∞
=
dξ
−∞
1 (ikn ) − E 2 . (9)
2
=−
π
2 kn + ∆2
. Show that to second order in HT ∂ HT ≈ − ∂φ
β 0
dτ Tτ HT (τ )
0
∂ HT ∂φ
0
=−
1 ∂ 2 ∂φ
β 0
dτ Tτ HT (τ )HT 0 . ∂F ∂ HT = (−2e) . ∂φ ∂φ (5) Show that the equilibrium current running between the two superconductors is IJ = I = (−2e) (6)
˙ where I is the operator for the electrical current I = (−e) Nc . This is a current which runs in thermodynamical equilibrium and hence is dissipationless in the sense that it runs without an applied bias (the chemical potential of the two sides is per deﬁnition identical in equilibrium). (8) ∂φ β
ikn kp
where G12 and G21 are the oﬀ-diagonal Nambu Green’s functions deﬁned in Exercise 3.
(7)
where the expectation value 1 ∂ 2 ∂φ
β 0
dτ Tτ HT (τ )HT
0
means taken with respect to Eqs. Then show that β ∂ t2 eiφ G21 (k. ikn ) =
k p
G12 (p. φc and φf .c. dτ = ∂φ 0 kp ∂ 1 = t2 eiφ G21 (k. The phase diﬀerence between the two side can be absorbed as a phase shift of the tunnel matrix t → e−i(φc −φf )/2 t.

IC dτ τ= 2eIc R t. I. (3) (4) f (t) → f (t)e−iV t/2 .2
RSJ model of a Josephson junction. which corresponds to φ → φ + 2eV t. With a ﬁnite bias voltage across the junction. (5)
(6)
. R 2eR Write this equation in the dimensionless form I = IN + IJ = η= I dφ = + sin φ. The Josephson junction thus acts as a voltage to frequency converter. or 2e ˙ φ = V. (2)
which is called the second Josephson relations. Now we look at the current-voltage characteristic of a Josephson junction in the RSJ model. Consider a current biased setup.e. The second Josephson relation adds interesting dynamics to the Josephson junction because of the intrinsic frequency 2eV / . (1) is still valid. This is known as the ﬁrst Josephson relation.324 and use it to ﬁnd that IJ = − π∆d(εF )t 2
2
EXERCISES FOR CHAPTER 17
1 β
ikn
2 kn
1 + ∆2
e (πd(εF )t) ∆β ∆ tanh 2 2 ∆π ∆β = sin φ tanh . i. i. which has many applications. a current carried by Cooper pairs and the relation IJ = IC sin φ.e. The ﬁnite voltage changes the energy of electrons of the two sides and hence their phase. One can measure this frequency by applying external RF radiation to the junction. a junction with a ﬁxed current.
2
(10)
Exercise 17. 2eRN 2 = sin φ
−1 where the normal state tunnel resistance is given by RN = πe2 d2 t2 / . We can simply include this phase change in the time dependence of by the following substitution c(t) → c(t)eiV t/2 . Thus 1 ˙ V + IC sin φ = φ + IC sin φ. The current is carried by two kinds of electrons: those that are paired and those that are not. one can still have a supercurrent running. The pair current is described by the Josephson relations while the normal current is supposed to be given by Ohm’s law. This current is made up by the sum of the supercurrent and the normal current.

(6) and show that the average voltage becomes V = RIc 0. I > IC (7)
˙ ˙ Hint: ﬁrst ﬁnd solutions for φ = 0 and then a “running” solution where φ = 0. Integrate Eq. For the dφ 1 T 2π last situation the average voltage is V = T 0 dt dt = T . but in a dc measurement one measures the average voltage. Here T is the period of the voltage or the time it takes to increase φ by 2 π. I < IC (I/IC )2 − 1.
.EXERCISES FOR CHAPTER 17
325
The voltage is time dependent.

29. 35 Bloch’s equation. 120 analytic continuation. 2 Brillouin zone bandstructure diagram. 165 many-particle basis. 54
326
. 3 ground state energy E0 . 193 in conductivity. insulator. 158 Bloch’s theorem. 40 in metal. 52 graphical representation. 73 art. 233 with impurity scattering. 78 basis states change in second quantization. 13 anti-symmetrization operator. 40 Boltzmann distribution. 145 anti-commutator. 83 tunneling spectroscopy. 35 deﬁnition. 194 spectral function. 82 Bohm-Staver sound velocity from RPA-screened phonons. 153 Bohr radius a0 . 162 analytic function. 35 metal. 55 adiabatic continuity. 34 for 1D phonons. 82 self-consistent gap equation. 16 complete basis set. 279 Bose-Einstein distribution. 10 deﬁning commutators. 51. 233 advanced function. 163 Aharonov-Bohm eﬀect. 53 Bohr radius a0 . 13 time dependence. 12 bra state. 281 bandstructure diagram extended zone scheme. 2 systems with diﬀerent particles. 241 Born approximation ﬁrst Born approximation. 68 atom artiﬁcial. 5 frequency. 239 introduction. 11 deﬁnition. 33 bandstructure. 259 full Born approximation. 140 Bloch Bloch theory of lattice electrons. 14 orthonormal basis set. 130 many-particle boson systems. density matrix. 285 mean ﬁeld Hamiltonian. 283 semi-classical. 12 many-particle fermion systems. 161 Anderson’s model for magnetic impurities. 1st order. 7 antiferromagnetism. the art of mean ﬁeld theory. 26 Boltzmann equation collision free. 34 Bogoliubov transformation. 45 Bardeen-Cooper-Schrieﬀer (see BCS). 91 time-derivative. 24 BCS theory eﬀective Hamiltonian. 81 interaction potential model. 55 bosons. 40 electron orbitals. 31 attractive pair-interaction. 51 boson creation/annihilation operators. 10 fermions. 56 in second quantization. 261 self-consistent Born approximation. 161. 190.Index
acoustic phonons Debye phonons. 3 Green’s function. 147 annihilation operators 1D phonons. semiconductor. 192 Born-Oppenheimer approximation.

Π deﬁnition. 254 general Kubo formalism. 55 bosons. 59. 22 current-current correlation function. 267 maximally crossed. 31 correlation function
327
charge-charge correlation. 208 conserving approximation. 259 continuity equation for ions in the jellium model. 285 ferromagnetism. 254 d-shell. 13 general deﬁnition. 238 contour integral. 48 charge-charge correlation function. 283 integration over. 181 second quantization. 279 direct process. 59 Debye model. 136 broken symmetry. 10 fermions. C} − {A. jellium model. 27 temperature dependence. 284 cooperons in conductivity. C] + [A. 3 set of quantum numbers ν. 39 collapse of wavefunction. 23 0 electron-phonon. 64 electron-phonon. 268 crossing diagrams deﬁnition. 97 correlation hole around electrons. 65 Coulomb blockade. 99 second quantization. 217. 27 momentum. 153 Coulomb interaction combined with phonons. 121 conductance quantization. 227 screened impurity scattering. 84 crossed diagram deﬁnition. 81 instability. 220 creation operators 1D phonons. 44 in conductivity. C]B. 52 for quasiparticles. 52 Debye energy or frequency ωD .. 269 core electron. 44. 63 electron-phonon. 240 connected Feynman diagrams. 256 RPA renormalization. 100. C] = A{B. 196 current density operator dia. 62 electron-phonon. 264 suppressed in the Born approx. C] = A[B. lattice model. 26 carbon nanotubes. 21 partition function. 103 chemical potential deﬁnition. RPA-screening. 269 introduction. 91 critical temperature Cooper instability. 113 universal ﬂuctuations. 100 mesoscopic system. 166 convergence of Matsubara functions. 81 instability of the Fermi surface. 116 conductivity cooperons. 52.INDEX
broadening of the spectral function. 92 deﬁning bosons. 284
. 100 diagrammatics. 13 time dependence. 71 canonical ensemble. 148 Debye acoustical Debye phonons. 23 Yukawa potential. 11 deﬁning fermions. 3 conductance conductance ﬂuctuations. 213 exchange process. 43 divergence. Feynman diagrams. C}B. 203 conservation of four-momentum. RPA-renormalized. 11 complete basis states. 160 Cooper Cooper pairs. 218 coupling constant electron interaction strength e2 . 104 semi-classical approach. 186 Kubo formalism. 74 superconductivity. general. 103 current-current correlation. 92 [AB. 253 Kubo formalism.and paramagnetic terms. 98 relation to dielectric function. 2 commutator [AB.

31
. 190 for Πxx . 52 elastic scattering Matsubara Green’s function. RPA. BCS. 144. 179 ﬁrst Born approximation. 7 diagonal Hamiltonian. 108. 140 non-interacting electrons. 104 diﬀerential conductance. 223 relation to conductivity. 246. 208 pair interactions in real space. 59 Debye wave number kD . 99 dielectric function ε equation of motion derivation. 2 delta function δ(r). 184 valence electrons. 231 of electron gas. 31 density of states. 269 full Born approximation. 143 distribution function Boltzmann.328
Debye temperature TD . 27 density of states measured by tunneling. 1 Einstein model of speciﬁc heat. 38 phase coherence. 237 electron core electrons. 284 polarization function χ. mesoscopic systems. 189 pair interactions in Fourier space. 29
INDEX
electron reservoir. 178 eﬀective electron-electron interaction Coulomb and phonons. external potential. 22 density matrix operator. 264 Dulong-Petit value for speciﬁc heat. 1 eigenvalue. 226 self-consistent Born approximation. deﬁnition of. 155 irreducible polarization function χirr . 216 the RPA-bubble χRPA . 29 non-interacting fermions. 240. 19 lattice vibrations. 28 Boltzmann. 60 dynamical matrix D(k). 283 eﬀective mass approximation. Debye model. 203 disorder. 133 diamagnetic term in current density. 28 Maxwell-Boltzmann. 226 Kubo formalism. 58 eigenstate deﬁnition. 121 dissipation due to electron-hole pairs. 283 phonon mediated. 179 electric potential classical theory. 1 superposition. 256 impurity averaged electrons. 131 single-particle Green’s function. 26 non-interacting bosons. 205 pair-scattering vertex Λ. 45 donor atoms. 57 Dyson equation Feynman diag. 46 Drude formula. 251. 35 eﬀective mass. Gibbs. Debye model. 59 density of states. 4 disconnected Feynman diagrams. 113 Fermi-Dirac. 102 relation to polarization function χ. 38 phonons. 193 impurity and interaction. 136 density waves.. 226 the RPA-bubble and phonons. 140 diﬀerential equation classical Green’s function. 4 density in second quantization. 59 spectral function. 172 Slater. 251 eigenmodes electromagnetic ﬁeld. 272 determinant ﬁrst quantization. 264. renormalization. 72 density-density correlation function the pair-bubble χ0 ≡ −Π0 . 26 Bose-Einstein. 257 for cooperon. 52 optical phonons. 59 frequency cut-oﬀ. 285 delta function δ(r). 282 dephasing. 146 Dirac bra(c)ket notation for quantum states. RPA. 7 in Wick’s theorem. 194 single-particle in external potential. 280 Coulomb and phonons. jellium. 127 many-body Green’s function. 60 Einstein phonons in the jellium model. 127 external and induced.

171 introduction. 222 Landau damping. 43 thermodynamic potential Ω. 61. 281 Cooper instability. 36 Fermi sea diagrams. 145 ergodic. 36 Fermi’s golden rule. 74 introduction. 276 combined with Coulomb interaction. 36 Fermi wave length λF . 49 1D: carbon nanotubes. 147 Heisenberg operators.INDEX
electron gas. deﬁnition. 281 dielectric properties and screening. 286 Fermi sea. 116 electron-electron scattering attractive interaction. 53 umklapp process. 5 creation/annihilation operators. 228 second order perturbation. 165 many-particle basis. 144. 51 equation of motion Anderson’s model. interacting attractive interaction. non-interacting Bloch theory. 88 in proof of Wick’s theorem. 222 Hartree–Fock mean ﬁeld Hamiltonian. Cooper instability. 35
329
Fermi Fermi energy εF . 264. 45 introduction. 177 static ion lattice. 149 derivation of RPA. 231 electron-phonon interaction adiabatic electron motion. 42 Fermi sea. 36 Fermi sea. 13 deﬁning commutators. 243 electron-hole pairs excitations. 70 inﬁnite perturbation series. 35 motion in external potentials. 155 interpretation of RPA. 121 ergodicity assumption. 63 electronic plasma oscillations graphical representation. 57 frequency domain. 245 Fermi-Dirac distribution. 275 the sound velocity. in general 0D: quantum dots. 31 electron gas. 14 fermion loop. 233 microscopic basis. 177 ﬁnite temperature. 161. 284 dephasing. 13 frequency. 218. 240. 39 ground state energy. 219 Fermi liquid theory introduction. 28. 153 for ions. 33 electron interaction strength e2 . 220 Landau damping. 63 the jellium model. 75 Feynman diagrams cancellation of disconnected diagrams. 203
. 33 density of states. 46 3D: metals and semiconductors. 244. 213 general considerations. 25 extended zone scheme. 272 life-time. 38 ground state energy. 202 ferromagnetism critical temperature. 237 fermion deﬁnition. 51 graphical representation. 147 single-particle Green’s function. 220 electron gas. 53 basis states. 145 Matsubara Green’s function. 38 Feynman diagrams. 63. 223 ﬁrst order perturbation. 275 the lattice model. 276 general introduction. excitations. 72 Stoner model. 43 full self-energy diagram. 36 Fermi wavenumber kF . 48 2D: GaAs heterostructures. 38 jellium model. 169 non-interacting particles. 214 full theory. 250 Thomas-Fermi screening. 230 plasma oscillations. 279 Feynman diagrams. 37 Fermi sea with interactions. 220. 41. 144 Fermi velocity vF . 73 order parameter. 23 0 electron wave guides. 214.

190 full Born approximation. 161 free energy deﬁniton. 54 basic theory. 202 phonon mediated pair interaction. 133 non-interacting particles. many-body system. 26 grand canonical density matrix. 282 Schr¨dinger equation. 132 free phonons. 78 Landau gauge. 127 Lehmann representation. 130 RPA-screened phonons. 70 Fock self-energy for pair interactions. 131 translation-invariant system. one-body system. 46 gauge breaking of gauge symmetry. 209 Hartree–Fock approximation. 52
. 146. 1 Greek letters. 160 introduction. 184 external potential scattering. 127 dressed. 204 pair interactions. external potential. 177 topologically diﬀerent diagrams. 70 Hartree self-energy. 254 free electrons. pair interaction. 69 four-vector/four-momentum notation. 18 second quantization. 205 pair interactions. 189 irreducible diagrams. 39 for ions. 208 interaction line in real space. 135 quadratic. G denominator. imp. 158 Green’s function n-particle. 127 renormalization. 179 impurity averaged Green’s function. 19 Gauss box. 67 GaAs/Ga1−x Alx As heterostructures. 1 single-particle systems. 27 partition function. 69 Hartree–Fock approximation introduction. 182 electron-phonon interaction.330
Cooper instability. 131 retarded. 47 Gibbs distribution. 145 retarded. 10. 19 transversality condition. 27 Hartree–Fock approximation. equation of motion. 2 Fock approximation for interactions. 201 pair interactions. scattering. 132 two particle. 27 ensemble. 141 Hamiltonian diagonal. Matsubara. 204 Feynman rules electron-impurity scattering. 284 electron-impurity scattering. 225 self-consistent Born approximation. 278 ﬁrst quantization many-particle systems. 170 classical. 128 o single-particle. 188 interaction line in Fourier space. 291 Coulomb interaction. many-body system. 276 external potential scattering. 194 single-particle. 3
INDEX
radiation ﬁeld. pair interactions. 131 imaginary time. 193 impurity averaged single-particle. 4 name. 179 ﬁrst Born approximation. 150. 199 polarization function χ. 188 pair interactions in Fourier space. 70 heat capacity for electrons. 275 greater and lesser. 134 Poisson’s equation. 207. 208 pair interactions in real space. 255 Fourier transformation 1D ion vibrations. 170 harmonic oscillator length. 206 equation of motion. 69 mean ﬁeld Hamiltonian. 18 Hartree approximation for interactions. 209 Fock space. 70 the interacting electron gas. 27 in mean ﬁeld theory. 27 gravitation. 135. 245 retarded. 204 irreducible diagrams. 147 Matsubara functions. G numerator.

220 interaction line general pair interaction in real space. 233 damping and plasma oscillations. 52 electron-phonon interaction. 158 Green’s function. 40 imaginary time discussion. 115 u RPA-screening in the electron gas. 97 tunnel current. 88 model of ferromagnetism. 188 in conductivity. 63 full electronic self-energy. 259 Landau and Fermi liquid theory. 189 pair interaction. conductivity. 113 linear response derivation. 33 basis in reciprocal space. 90. 3 gauge. 21 Kronecker’s delta function δk. 90 inﬁnitesimal shift η. 1 hopping. 98. 204 pair interaction in Fourier space.. 149 Hubbard model. 134
. 205 polarization function χirr . 33 lattice vibrations 1D phonon Hamiltonian. 51 forming a static lattice. Hamiltonian. 97 dielectric function. 3 ground state energy E0 . 75 hybridization. 4 Heisenberg Heisenberg picture. 53 electron-phonon interaction. 73 in a metal. 148 hydrogen atom Bohr radius a0 . 73 helium. 208 RPA screened Coulomb line. 144 inﬁnite perturbation series breakdown at phase transistions. 227 RPA screened impurity line. 214 single-particle Green’s function. 40 electron orbitals. 21 kinetic momentum. 35 ket state. 227 inelastic light scattering. 230 eigenstates. 21 second quantization. 33 Heisenberg model. 160 impurities. 100 conductivity. 52 static case. 223 time evolution. 85 electron gas ground state energy. 255 renormalization by RPA-screening. 162 integration over the coupling constant. ionic ferromagnets. magnetic. 264. 184 impurity-scattering line Feynman rules. 115 lattice model basis in real space. 265 ions ionic plasma oscillations. 214 oscillating background. 253 impurity self-average. 280 Einstein phonons.n . 178 ˆ time-evolution operator U (t. GaAs/Ga1−x Alx As. 227 interaction picture imaginary time. 4 Kubo formalism conductance. 88 real space Matsubara Green’s fct. 158 Greek letters. 139 ladder diagram. 3 Landauer-B¨ttiker formula u heuristic derivation. 33 Hamiltonian. 128
331
jellium model eﬀective electron-electron interaction. 222 self-energy for interacting electrons. 46 Hilbert space. 2 kinetic energy operator including a vector potential..INDEX
Heaviside’s step function θ(x). 61 Lehmann representation deﬁnition. 31 irreducible Feynman diagrams impurity scattering. 225 iterative solution. 200 interference. 217. integral eqs. 102 general introduction. 159 introduction. 95 Landauer-B¨ttiker formula. t0 ). 254 correlation function. 148 impurity scattering. 9 heterostructures.

46 mean ﬁeld theory Anderson’s model. 161 Green’s function.332
for G> . 150 BCS mean ﬁeld Hamiltonian. 162 life-time. 261 Matsubara Green’s function. ω). 67 the art of mean ﬁeld theory. in Fourier space. 229 Thomas-Fermi screening in metals. introduction. 147. 135 normalization of quantum states. 65 mean ﬁeld approximation. 31 Hamiltonian. 66 Hartree–Fock mean ﬁeld Hamiltonian. 21 relaxation. 279 molecular beam epitaxy. 219 Migdal’s theorem. 168 Matsubara Green’s function elastic scattering. 91 tunnel current. 275 full Born approximation. 82 broken symmetry. 169 convergence of. 149 magnetization. 240. 107 measuring the spectral function. 262 Lindhard function. 206 interacting electrons in real space. 14 introduction. phase transistions. 282 self-consistent Born approximation. 139 magnetic impurities. 150. 155 linear response theory introduction. 182 ﬁrst Born approximation. evaluation of. 135 in conductivity. 3 normalization. 147 Green’s functions. ω). 233 retarded Green’s function GR (kσ. 107 metal disordering and random impurities. 224 maximally crossed diagrams. 134 Matsubara function. 253 regime. 121 physics. 12 fermions. 71
INDEX
general Hamiltonian HMF . 194 two-particle polarization function χ. G< . 68 mean free path. 31 occupation number operator bosons. 160 Fourier transformation. 143. 74. 190 free phonons. 78 mesoscopic disordered systems. molecular beam epitaxy. with branch cuts. 265 systems. 3 magnetic moment. 243 MOSFET. MBE. 208 interacting electrons in Fourier space. 28 equation of motion. 52 non-interacting particles distribution functions. 193 impurity averaged single-particle. 188 interacting elec. 2 second quantization. 115 u mesoscopic system. 268 MBE. 21 kinetic. 161 frequency. 108 nucleus. 95 Landauer-B¨ttiker formula. 67 partition function ZMF . 135 spectral function A0 (kσ. 181 electrical resistivity. simple poles. 236. 243. 149 many-body system single-particle Green’s function. 164 quasiparticles. 137 Meissner eﬀect. 160 relation to retarded function. 132 Hamiltonian. 199 RPA-screened phonons. 167 sums. 113 time evolution. 181 general description. 46 momentum canonical. equation of motion. 46 Newton’s second law for ions in the jellium model. 131 ﬁrst quantization. 72. 10
. 32 observation of plasmons. and GR . 161 sums. 148 magnetic length. scattering state. 165 sums. 254 Matsubara function. 70 introduction. 9 mass renormalization. 179 electron-impurity scattering.

27
333
in mean ﬁeld theory. 272 eigenmodes in 3D. 72 overlap of wavefunctions localized/extended states. 264 phase coherence for electrons. 133 tunneling. 52 graphical representation. 244. 52
. 56 optical spectroscopy. 59 dephasing. 148 particle propagation. 184 phase coherence length lϕ . 7 in ﬁrst quantization. 275 general introduction. 216 paramagnetic term in current density. 10 operator adjoint. 99 particle-particle scattering in the collision term. 7 Heisenberg equation of motion. 55 relevant operator Aqλ . 7. 204 self-energy in Fourier space. 58 plasma frequency for electron gases in a metals. 87 fermion creation/annihilation. 1 real time ordering Tt . 90 phase coherence. 2 boson creation/annihilation. 282 second quantization. 5. 58 Einstein model of speciﬁc heat. 228 ionic plasma frequency. 51 Hamiltonian for jellium phonons. 208 Dyson equation in real space. 72 list of order parameters. 88 Hermitian. 40 spin matrices. Green’s function. 210 the correlation function χ0 ≡ −Π0 . 281 RPA-renormalized Green’s function. 275 RPA renormalization. 52 phonon branches. 128 ˆ time-evolution operator U (t. 90 perturbation theory ﬁrst order. 264. 55 1D lattice vibrations. interacting electrons. electron gas. 53 electrons. 245 phase transition breakdown of perturbation theory. 85 broken symmetry. 72 pair interactions Dyson equation in Fourier space. 208 self-energy in real space. 89 trace Tr. 41 inﬁnite order. 43 single particle wavefunction. 72 phonons 1D annihilation/creation operators. 21 periodic boundary conditions 1D phonons. 95 second order. 171 permutation group SN . 19 permanent for bosons. iqn ). 26 grand canonical ensemble. 214 linear response. 172 permutation. 205 pair-bubble calculation of the pair-bubble. Debye model. scattering theory. 67 Pauli exclusion principle. 138 pair condensate. electron gas. Kubo formula. ground state energy. 27 optical phonons Einstein phonons. 205 Feynman diagrams. 55. t0 ).INDEX
occupation number representation. 10 electromagnetic ﬁeld. 178 inﬁnite order. 141 optical theorem. 208 Feynman rules in real space. 218 Feynman diagram Π0 (q. 19 expansion of e−iHt . 53 density of states. 13 ﬁrst quantization. 90 second quantization. 7 in Wick’s theorem. 222 inﬁnite order. 211 in the RPA self-energy. 186 phase space. 194 order parameter deﬁnition. 71 order parameters. 216 self-energy diagram. 251 life-time. 199 Feynman rules in Fourier space. 14 ˆ time evolution operator U (t. t0 ). 243 partition function canonical ensemble. 36 photons. 60 free Green’s function.

256 impurity scattering. 178 quadratic Hamiltonian. 35 reﬂection amplitude. 188 the RPA self-energy.. 170 quantum coherence macroscopic in superconductivity. 226 Feynman diagrams. 2 free particle. 25. 103 momentum space. 251 Green’s function. 130 single-particle in external potential. 33 reduced zone scheme. Dyson equation. 51 Landau damping. 173 irreducible Feynman diagrams. 19 probability current conservation. 253. 238. 228 plasmons dynamical screening. 142 relation to dielectric function ε. 2 time evolution. 163 reciprocal lattice basis. 116 quantum state bra and ket state. 136 propagator Green’s function. 184 quantum correction. 233 life-time. 3 orthogonal. 110 reﬂectionless contact. 243 quasiparticle-quasiparticle scattering. 243 radiation ﬁeld. 230 plasmons. 150. 228 ionic plasma oscillations. 107 quantum ﬁeld operator deﬁnition. 2 hydrogen. 19 Raman scattering. 135. 225 free electrons. 2 quasiparticle deﬁnition. 140 quantum eﬀects. 121 random phase approximation (see RPA). 108. 213 rational function. 144 random impurities.334
plasma oscillations electronic plasma oscillations. 153 electron-hole pair interpretation. 127 polarization function χ Dyson equation. 281 reservoir. 243 renormalization constant Z. 229 plasma oscillations. 132 relation to Matsubara function. 33 reciprocal space. 224 polarization vectors phonons. 181 general introduction. 3
INDEX
sum over. 161 Roman letters. 220 Fermi liquid theory. 223 two-particle Matsubara Green’s fct. 186 quantum number ν Feynman rules. 131. 79 single electrons. 17 quantum ﬂuctuations in conductance. 238 experimental observation in metals. 49 tunneling spectroscopy. 47 Green’s function. 147 Green’s function. 235 introduction. 245 of phonons by RPA-screening. 181 random matrix theory. 246. 247 eﬀective mass. 111 probability distribution. 237 Poisson’s equation GaAs heterostructures. 225 Kubo formalism. 236 discussion. 51 interacting electron gas in RPA. 158 RPA for the electron gas Coulomb and impurity lines. 57 photons. 264. 228 semi-classical treatment. 273 quantum dots introduction. 246
. 3 Landau states. 146. 257 deriving the equation of motion. 143. 240 resummation of diagrams current-current correlation. 215 retarded function convergence factor. 108 resistivity (see conductivity). 4 quantum point contact. 113 relaxation time approximation. 17 Fourier transform.

21 statistical mechanics. 192 in sums with branch cuts. 51 space-time. 136 deﬁnition. 150 ﬁrst Born approximation. 259 Rydberg. 194 Schr¨dinger equation. 141 tunneling. 190 Fock diagram for pair interactions. 22 spin. 21
. 2 hydrogen orbital. 218 semiclassical. 226 the polarization function χRPA . 40 scattering length. 208 pair interactions in real space. 3 Slater determinant. 67 self-energy due to hybridization. dynamical. 87 o screening dieelectric properties of the elec. 137 spin Pauli matrices. interacting electrons. 189 interacting electrons. 194 semi-classical approximation. 227 renormalized Coulomb interaction. semi-classical. 237 transport equation. 238 semiclassical. 135 ﬁrst Born approximation. 240 single-particle states as N -particle basis. 151 general mean-ﬁeld theory. unit of energy (Ry). 3 Landau state. 135 physical interpretation. 209 impurity averaged electrons. 55 free phonons in 3D. 135 spectroscopy optical. 193 scattering matrix. 217 resummation of the self-energy. 16 Coulomb interaction. 52 sounds waves. 169 measurement. 215 the dielectric function εRPA . 177 spectral function Anderson’s model. 19 electron-phonon interaction. 193 Schr¨dinger equation o Green’s function. 238 Thomas-Fermi screening. 261 screening. 214 irreducible. 2 Schr¨dinger picture. 39 sound velocity Bohm-Staver formula. 210 RPA self-energy. 21 name. 223 RPA-screened Coulomb interaction. 21 second quantization. 58 harmonic oscillator. 117 scattering theory. 6 free particle state. 112 time-dependent. 128 quantum point contact. pair interactions. 128 o transition matrix.INDEX
plasmons and Landay damping. 257 pair interactions in Fourier space. gas. 9 basis for diﬀerent particles. 128 time reversal symmetry. 205 pair-bubble diag. 283 Bohm-Staver formula. 209 full Born approximation. 108 scattering theory optical theorem. RPA. jellium model. 184 weak localization. 193 Hartree diagram for pair interactions. S. 137 non-interacting particles. 151 broadening. 14 particle current density. 1 operators. points and integrals. 24 change of basis. 108 scattering state.. 25 thermal average. 218 second quantization basic concepts. 18 kinetic energy. 53 Debye model. static. 61 free phonons in 1D. 27 self-average for impurity scattering basic concepts. 23 electromagnetic ﬁeld. 265
335
self-consistent equation Anderson’s model. 7 Sommerfeld expansion. 22 particle density. 226 vertex corrections. 216 self-consistent Born approximation. fermions.

89 time-ordering operator imaginary time Tτ . Kubo. 272 topologically diﬀerent diagrams. 47 truncation Anderson’s model. 254 Green’s function. 91 Heisenberg picture. 78 microscopic BCS theory. 254. 233 transport time. 280 pair-scattering vertex Λ. 88 in linear response. 19 kinetic energy. 24. 277 WKB approximation. 218. t0 ). 204 trace of operators. 257 vertex function. 88 linear response. 72 symmetrization operator. 63 unit cell. 31 vector potential electromagnetic ﬁeld. 81 order parameter. 159 Schr¨dinger picture. 19 triangular potential well. 266. 219. 121. 222 time dependent Hamiltonian. 160 real time Tt . 25 step function θ(x). 255 dressed vertex function. 84 introduction. 242 transversality condition. 90 time-reversal symmetry. 27 transition matrix. 72 statistical mechanics second quantization. 55 unitarity. 258 electron-phonon vertex. 267 transmission coeﬃcients. imaginary time. 154 discussion. 218
. 113 transport equation. 130. 201 phonon Green’s function. 123 Wick’s theorem derivation. 132 transmission amplitude. 253 mesoscopic systems. 264 introduction. 284 vertex correction. 137
INDEX
umklapp process. 138 BCS superconductor. 193 translation-invariant system conductivity. 220 Thomas-Fermi screening.336
spontaneous symmetry breaking breaking of gauge symmetry. scattering theory. 262 wavefunction collapse. 119 Yukawa potential deﬁnition. 138 Stoner model of metallic ferromagnetism. 2 weak localization and conductivity. 4 STM. 213 RPA-screened Coulomb interaction. 140 current. 268 Ward identity. 21 Kubo formalism. 124 valence electrons. 78 introduction. 7 thermal average. 170 interacting electrons. 95 time evolution creation/annihilation operators. S-matrix. 121. 27 thermodynamic potential Ω deﬁnition. 153 derivation of RPA. 111 universal conductance ﬂuctuations. 28 for the interacting electron gas. 99 vertex current vertex. 110. 75 superconductivity critical temperature. 87 o ˆ unitary operator U (t. 258. 91 interaction picture. 78 Meissner eﬀect. 97 operator. 138 spectroscopy. 268. 112 time-reversed paths. 145 tunneling scanning microscope. 150.