Introduction to manybody Green s functions


 Teresa Baldwin
 1 years ago
 Views:
Transcription
1 Introduction to manybody Green s functions Matteo Gatti European Theoretical Spectroscopy Facility (ETSF) NanoBio Spectroscopy Group  UPV San Sebastián  Spain ELK school  CECAM 2011
2 Outline 1 Motivation 2 Oneparticle Green s functions: GW approximation 3 Twoparticle Green s functions: BetheSalpeter equation 4 Micromacro connection
3 References Francesco Sottile PhD thesis, Ecole Polytechnique (2003) francesco/tesi_dot.pdf Fabien Bruneval PhD thesis, Ecole Polytechnique (2005) bruneval_these.pdf Giovanni Onida, Lucia Reining, and Angel Rubio Rev. Mod. Phys. 74, 601 (2002). G. Strinati Rivista del Nuovo Cimento 11, (12)1 (1988).
4 Outline 1 Motivation 2 Oneparticle Green s functions: GW approximation 3 Twoparticle Green s functions: BetheSalpeter equation 4 Micromacro connection
5 Motivation Theoretical spectroscopy Calculate and reproduce Understand and explain Predict Exp. at 30 K from: P. Lautenschlager et al., Phys. Rev. B 36, 4821 (1987).
6 Theoretical Spectroscopy Which kind of spectra? Which kind of tools?
7 Why do we have to study more than DFT? Absorption spectrum of bulk silicon in DFT How can we understand this?
8 Why do we have to study more than DFT? Absorption spectrum of bulk silicon in DFT Spectroscopy is exciting!
9 MBPT vs. TDDFT: different worlds, same physics MBPT based on Green s functions oneparticle G: electron addition and removal  GW twoparticle L: electronhole excitation  BSE moves (quasi)particles around is intuitive (easy) TDDFT based on the density response function χ: neutral excitations moves density around is efficient (simple)
10 Response functions External perturbation V ext applied on the sample V tot acting on the electronic system Potentials Dielectric function δv tot = δv ext + δv ind ɛ = δv ext δv tot δv ind = vδρ = 1 v δρ ɛ 1 = δv tot = 1 + v δv ext δv tot δρ δv ext
11 Response functions External perturbation V ext applied on the sample V tot acting on the electronic system Dielectric function P = ɛ = δv ext δv tot = 1 vp ɛ 1 = δv tot δv ext = 1 + vχ δρ δv tot χ = δρ δv ext χ = P + Pvχ P = χ 0 + χ 0 f xc P
12 Micromacro connection MicroscopicMacroscopic connection: local fields χ G,G (q, ω) = P G,G (q, ω) + P G,G1 (q, ω)v G1 (q)χ G1,G (q, ω) ɛ 1 G,G (q, ω) = δ G,G + v G (q)χ G,G (q, ω) ɛ M (q, ω) = 1 ɛ 1 G=0,G =0 (q, ω) Adler, Phys. Rev. 126 (1962); Wiser, Phys. Rev. 129 (1963).
13 Micromacro connection MicroscopicMacroscopic connection: local fields ɛ M (q, ω) = 1 v G=0 (q) χ G=0,G =0(q, ω) χ G,G (q, ω) = P G,G (q, ω) + P G,G1 (q, ω) v G1 (q) χ G1,G (q, ω) v G (q) = 0 for G = 0 v G (q) = v G (q) for G 0 Hanke, Adv. Phys. 27 (1978).
14 Absorption spectra Absorption spectra Abs(ω) = lim q 0 Imɛ M (q, ω) Abs(ω) = lim q 0 Im [v G=0 (q) χ G=0,G =0(q, ω)] Absorption response to V ext + V macro ind
15 Independent particles: KohnSham Independent transitions: ɛ 2 (ω) = 8π2 ϕ Ωω 2 j e v ϕ i 2 δ(ε j ε i ω) ij
16 What is an electron?
17 Outline 1 Motivation 2 Oneparticle Green s functions: GW approximation 3 Twoparticle Green s functions: BetheSalpeter equation 4 Micromacro connection
18 Photoemission Direct Photoemission Inverse Photoemission
19 Why do we have to study more than DFT? adapted from M. van Schilfgaarde et al., PRL 96 (2006).
20 Oneparticle Green s function The oneparticle Green s function G Definition and meaning of G: ig(x 1, t 1 ; x 2, t 2 ) = N T [ ψ(x 1, t 1 )ψ (x 2, t 2 ) ] N for for t 1 > t 2 ig(x 1, t 1 ; x 2, t 2 ) = N ψ(x 1, t 1 )ψ (x 2, t 2 ) N t 1 < t 2 ig(x 1, t 1 ; x 2, t 2 ) = N ψ (x 2, t 2 )ψ(x 1, t 1 ) N
21 Oneparticle Green s function t 1 > t 2 N ψ(x 1, t 1 )ψ (x 2, t 2 ) N t 1 < t 2 N ψ (x 2, t 2 )ψ(x 1, t 1 ) N
22 Oneparticle Green s function What is G? Definition and meaning of G: [ ] G(x 1, t 1 ; x 2, t 2 ) = i < N T ψ(x 1, t 1 )ψ (x 2, t 2 ) N > Insert a complete set of N + 1 or N 1particle states. This yields G(x 1, t 1 ; x 2, t 2 ) = i j f j (x 1 )f j (x 2 )e iε j (t 1 t 2 ) [θ(t 1 t 2 )θ(ε j µ) θ(t 2 t 1 )Θ(µ ε j )]; where: ε j = E(N + 1, j) E(N), ε j > µ E(N) E(N 1, j), ε j < µ f j (x 1 ) = N ψ (x 1) N + 1, j, ε j > µ N 1, j ψ (x 1 ) N, ε j < µ
23 Oneparticle Green s function What is G?  Fourier transform G(x, x, ω) = j Fourier Transform: f j (x)f j (x ) ω ε j + iηsgn(ε j µ). Spectral function: A(x, x ; ω) = 1 π ImG(x, x ; ω) = j f j (x)f j (x )δ(ω ε j ).
24 Photoemission Direct Photoemission Inverse Photoemission Oneparticle excitations poles of oneparticle Green s function G
25 Oneparticle Green s function Oneparticle Green s function From oneparticle G we can obtain: oneparticle excitation spectra groundstate expectation value of any oneparticle operator: e.g. density ρ or density matrix γ: ρ(r, t) = ig(r, r, t, t + ) γ(r, r, t) = ig(r, r, t, t + ) groundstate total energy
26 Oneparticle Green s function Straightforward? G(x, t; x, t ) = i < N T [ ψ(x, t)ψ (x, t ) ] N > N > =??? Interacting ground state! Perturbation Theory? Timeindependent perturbation theories: messy. Textbooks: adiabatically switched on interaction, GellMannLow theorem, Wick s theorem, expansion (diagrams). Lots of diagrams...
27 Oneparticle Green s function Straightforward? G(x, t; x, t ) = i < N T [ ψ(x, t)ψ (x, t ) ] N > N > =??? Interacting ground state! Perturbation Theory? Timeindependent perturbation theories: messy. Textbooks: adiabatically switched on interaction, GellMannLow theorem, Wick s theorem, expansion (diagrams). Lots of diagrams...
28 Oneparticle Green s function Straightforward? G(x, t; x, t ) = i < N T [ ψ(x, t)ψ (x, t ) ] N > N > =??? Interacting ground state! Perturbation Theory? Timeindependent perturbation theories: messy. Textbooks: adiabatically switched on interaction, GellMannLow theorem, Wick s theorem, expansion (diagrams). Lots of diagrams...
29 Functional approach to the MB problem Equation of motion To determine the 1particle Green s function: [ ] i h 0 (1) G(1, 2) = δ(1, 2) i t 1 d3v(1, 3)G 2 (1, 3, 2, 3 + ) Do the Fourier transform in frequency space: [ω h 0 ]G(ω) + i vg 2 (ω) = 1 where h 0 = v ext is the independent particle Hamiltonian. The 2particle Green s function describes the motion of 2 particles.
30 Unfortunately, hierarchy of equations G 1 (1, 2) G 2 (1, 2; 3, 4) G 2 (1, 2; 3, 4) G 3 (1, 2, 3; 4, 5, 6)...
31 Selfenergy Perturbation theory starts from what is known to evaluate what is not known, hoping that the difference is small... Let s say we know G 0 (ω) that corresponds to the Hamiltonian h 0 Everything that is unknown is put in Σ(ω) = G 1 0 (ω) G 1 (ω) This is the definition of the selfenergy Thus, [ω h 0 ]G(ω) Σ(ω)G(ω) = 1 to be compared with [ω h 0 ]G(ω) + i vg 2 (ω) = 1
32 Selfenergy Perturbation theory starts from what is known to evaluate what is not known, hoping that the difference is small... Let s say we know G 0 (ω) that corresponds to the Hamiltonian h 0 Everything that is unknown is put in Σ(ω) = G 1 0 (ω) G 1 (ω) This is the definition of the selfenergy Thus, [ω h 0 ]G(ω) Σ(ω)G(ω) = 1 to be compared with [ω h 0 ]G(ω) + i vg 2 (ω) = 1
33 Oneparticle Green s function Trick due to Schwinger (1951): introduce a small external potential U(3), that will be made equal to zero at the end, and calculate the variations of G with respect to U δg(1, 2) δu(3) = G 2 (1, 3; 2, 3) + G(1, 2)G(3, 3).
34 Hedin s equation Hedin s equations Σ =igw Γ G =G 0 + G 0 ΣG Γ =1 + δσ δg GGΓ P = iggγ W =v + vpw L. Hedin, Phys. Rev. 139 (1965)
35 GW bandstructure: photoemission additional charge
36 GW bandstructure: photoemission additional charge reaction: polarization, screening GW approximation 1 polarization made of noninteracting electronhole pairs (RPA) 2 classical (Hartree) interaction between additional charge and polarization charge
37 Hedin s equation and GW GW approximation Σ =igw Γ G =G 0 + G 0 ΣG Γ =1 P = iggγ W =v + vpw L. Hedin, Phys. Rev. 139 (1965)
38 Hedin s equation and GW GW approximation Σ =igw G =G 0 + G 0 ΣG Γ =1 P = igg W =v + vpw L. Hedin, Phys. Rev. 139 (1965)
39 GW corrections Standard perturbative G 0 W 0 H 0 (r)φ i (r) + H 0 (r)ϕ i (r) + V xc (r)ϕ i (r) = ɛ i ϕ i (r) dr Σ(r, r, ω = E i ) φ i (r ) = E i φ i (r) Firstorder perturbative corrections with Σ = igw : E i ɛ i = ϕ i Σ V xc ϕ i Hybersten and Louie, PRB 34 (1986); Godby, Schlüter and Sham, PRB 37 (1988)
40 GW results M. van Schilfgaarde et al., PRL 96 (2006).
41 Independent (quasi)particles: GW Independent transitions: ɛ 2 (ω) = 8π2 ϕ Ωω 2 j e v ϕ i 2 δ(e j E i ω) ij
42 What is wrong? What is missing?
43 Absorption Twoparticle excitations poles of twoparticle Green s function L Excitonic effects = electron  hole interaction
44 Absorption Twoparticle excitations poles of twoparticle Green s function L Excitonic effects = electron  hole interaction
45 Absorption Twoparticle excitations poles of twoparticle Green s function L Excitonic effects = electron  hole interaction
46 Outline 1 Motivation 2 Oneparticle Green s functions: GW approximation 3 Twoparticle Green s functions: BetheSalpeter equation 4 Micromacro connection
47 Beyond RPA P(12) = ig(12)g(21) = P 0 (12) Independent particles (RPA)
48 Beyond RPA P(12) = ig(13)g(42)γ(342) Interacting particles (excitonic effects)
49 From Hedin s equations to BSE From Hedin... P = iggγ Γ = 1 + δσ δg GGΓ
50 From Hedin s equations to BSE From Hedin......to BetheSalpeter P = iggγ Γ = 1 + δσ δg GGΓ ( δσ ) L = L 0 + L 0 v + i L δg
51 The BetheSalpeter equation Exercise Formal derivation δg(12) L(1234) = i δv = +ig(15) δg 1 (56) ext(34) δv G(62) ext(34) = + ig(15) δ[g 1 0 (56) Vext(56) Σ(56)] G(62) δv ext(34) [ δvh (5)δ(56) = ig(13)g(42) + ig(15)g(62) δσ(56) ] δv ext(34) δv ext(34) [ δvh (5)δ(56) = ig(13)g(42) + ig(15)g(62) δσ(56) ] δg(78) δg(78) δg(78) δv ext(34) [ L(1234) =L 0 (1234) + L 0 (1256) v(57)δ(56)δ(78) + i δσ(56) δg(78) ] L(7834)
52 The BetheSalpeter equation [ L(1234) = L 0 (1234) + L 0 (1256) v(57)δ(56)δ(78) + i δσ(56) δg(78) ] L(7834) Polarizabilities δg(12) L(1234) = i δv ext (34) χ(12) = δρ(1) δv ext (2) L(1122) = χ(12)
53 The BetheSalpeter equation Approximations ( δσ ) L = L 0 + L 0 v + i L δg
54 The BetheSalpeter equation Approximations Σ igw ( δσ ) L = L 0 + L 0 v + i L δg Approximation:
55 The BetheSalpeter equation Approximations Σ igw ( δ(gw )) L = L 0 + L 0 v L δg Approximation: δ(gw ) δg = W + GδW δg W
56 The BetheSalpeter equation Approximations Final result: L = L 0 + L 0 (v W )L
57 The BetheSalpeter equation BetheSalpeter equation L(1234) = L 0 (1234)+ L 0 (1256)[v(57)δ(56)δ(78) W (56)δ(57)δ(68)]L(7834)
58 Absorption spectra in BSE Bulk silicon G. Onida, L. Reining, and A. Rubio, RMP 74 (2002).
59 Solving BSE L(1234) = L 0 (1234)+ L 0 (1256)[v(57)δ(56)δ(78) W (56)δ(57)δ(68)]L(7834) Static W Simplification: W (r 1, r 2, t 1 t 2 ) W (r 1, r 2 )δ(t 1 t 2 ) L(1234) L(r 1, r 2, r 3, r 4, t t ) L(r 1, r 2, r 3, r 4, ω)
60 Solving BSE L(1234) = L 0 (1234)+ L 0 (1256)[ v(57)δ(56)δ(78) W (56)δ(57)δ(68)] L(7834) Static W Simplification: W (r 1, r 2, t 1 t 2 ) W (r 1, r 2 )δ(t 1 t 2 ) L(1234) L(r 1, r 2, r 3, r 4, t t ) L(r 1, r 2, r 3, r 4, ω)
61 Solving BSE L(1234) = L 0 (1234)+ L 0 (1256)[ v(57)δ(56)δ(78) W (56)δ(57)δ(68)] L(7834) Static W Simplification: W (r 1, r 2, t 1 t 2 ) W (r 1, r 2 )δ(t 1 t 2 ) L(1234) L(r 1, r 2, r 3, r 4, t t ) L(r 1, r 2, r 3, r 4, ω)
62 Solving BSE Dielectric function L(r 1 r 2 r 3 r 4 ω) = L 0 (r 1 r 2 r 3 r 4 ω) + dr 5 dr 6 dr 7 dr 8 L 0 (r 1 r 2 r 5 r 6 ω) [ v(r 5 r 7 )δ(r 5 r 6 )δ(r 7 r 8 ) W (r 5 r 6 )δ(r 5 r 7 )δ(r 6 r 8 )] L(r 7 r 8 r 3 r 4 ω) [ ɛ M (ω) = 1 lim v G=0 (q) q 0 ] drdr e iq(r r ) L(r, r, r, r, ω)
63 Solving BSE L(r 1 r 2 r 3 r 4 ω) = L 0 (r 1 r 2 r 3 r 4 ω) + dr 5 dr 6 dr 7 dr 8 L 0 (r 1 r 2 r 5 r 6 ω) [ v(r 5 r 7 )δ(r 5 r 6 )δ(r 7 r 8 ) W (r 5 r 6 )δ(r 5 r 7 )δ(r 6 r 8 )] L(r 7 r 8 r 3 r 4 ω) Transition space How to solve it? L (n1 n 2 )(n 3 n 4 )(ω) = φ n 1 (r 1 )φ n2 (r 2 ) L(r 1 r 2 r 3 r 4 ω) φ n 3 (r 3 )φ n4 (r 4 ) = L
64 Solving BSE L(r 1 r 2 r 3 r 4 ω) = L 0 (r 1 r 2 r 3 r 4 ω) + dr 5 dr 6 dr 7 dr 8 L 0 (r 1 r 2 r 5 r 6 ω) [ v(r 5 r 7 )δ(r 5 r 6 )δ(r 7 r 8 ) W (r 5 r 6 )δ(r 5 r 7 )δ(r 6 r 8 )] L(r 7 r 8 r 3 r 4 ω) Transition space How to solve it? L (n1 n 2 )(n 3 n 4 )(ω) = φ n 1 (r 1 )φ n2 (r 2 ) L(r 1 r 2 r 3 r 4 ω) φ n 3 (r 3 )φ n4 (r 4 ) = L
65 Exercise Calculate: L 0 (r 1, r 2, r 3, r 4, ω) = ij L 0 = (f j f i ) φ i (r 1)φ j (r 2 )φ i (r 3 )φ j (r 4) ω (E i E j ) f n1 f n2 ω (E n2 E n1 ) δ n 1 n 3 δ n2 n 4
66 Solving BSE BSE in transition space We consider only resonant optical transitions for a nonmetallic system: (n 1 n 2 ) = (vkck) (vc) L = L 0 + L 0 ( v W ) L L = [1 L 0 ( v W )] 1 L 0 L = [L 1 0 ( v W )] 1 L (vc)(v c )(ω) = [(E c E v ω)δ vv δ cc + (f v f c ) v W ] 1 (f c f v )
67 Solving BSE L (vc)(v c )(ω) = [(E c E v ω)δ vv δ cc + (f v f c ) v W ] 1 (f c f v ) Spectral representation of a hermitian operator [H exc ωi ] 1 = λ L (vc)(v c )(ω) = λ H exc A λ = E λ A λ A λ A λ E λ ω A (vc) λ A (v c ) λ E λ ω (f c f v )
68 Solving BSE L (vc)(v c )(ω) = [(E c E v ω)δ vv δ cc + (f v f c ) v W ] 1 (f c f v ) L [H exc ωi ] 1 Spectral representation of a hermitian operator [H exc ωi ] 1 = λ L (vc)(v c )(ω) = λ H exc A λ = E λ A λ A λ A λ E λ ω A (vc) λ A (v c ) λ E λ ω (f c f v )
69 Absorption spectra in BSE Independent (quasi)particles Abs(ω) vc v D c 2 δ(e c E v ω) Excitonic effects [H el + H hole +H el hole ] A λ = E λ A λ Abs(ω) λ vc A (vc) λ v D c 2 δ(e λ ω) mixing of transitions: v D c 2 vc A(vc) λ v D c 2 modification of excitation energies: E c E v E λ
70 BSE calculations A threestep method 1 LDA calculation KohnSham wavefunctions ϕ i 2 GW calculation GW energies E i and screened Coulomb interaction W 3 BSE calculation solution of H exc A λ = E λ A λ with: H (vc)(v c ) exc = (E c E v )δ vv δ cc + (f v f c ) vc v W v c excitonic eigenstates A λ, E λ spectra ɛ M (ω)
71 A bit of history derivation of the equation (bound state of deuteron) E. E. Salpeter and H. A. Bethe, PR 84, 1232 (1951). BSE for exciton calculations L.J. Sham and T.M. Rice, PR 144, 708 (1966). W. Hanke and L. J. Sham, PRL 43, 387 (1979). first ab initio calculation G. Onida, L. Reining, R. W. Godby, R. Del Sole, and W. Andreoni, PRL 75, 818 (1995). first ab initio calculations in extended systems S. Albrecht, L. Reining, R. Del Sole, and G. Onida, PRL 80, 4510 (1998). L. X. Benedict, E. L. Shirley, and R. B. Bohn, PRL 80, 4514 (1998). M. Rohlfing and S. G. Louie, PRL 81, 2312 (1998).
72 Continuum excitons Bulk silicon G. Onida, L. Reining, and A. Rubio, RMP 74 (2002).
73 Bound excitons Solid argon F. Sottile, M. Marsili, V. Olevano, and L. Reining, PRB 76 (2007).
74 The Wannier model BetheSalpeter equation H exca λ = E λ A λ H (vc)(v c ) exc = (E c E v)δ vv δ cc + v W Wannier model two parabolic bands E c E v = E g + k 2 2µ 2 2µ no local fields ( v = 0) and effective screened W W (r, r ) = 1 ɛ 0 r r solution = Rydberg series for effective H atom E n = E g R eff n 2 with R eff = R µ ɛ 2 0
75 Exciton analysis Exciton amplitude: Ψ λ (r h, r e ) = vc A (vc) λ φ v(r h )φ c (r e ) Graphene nanoribbon Manganese Oxide D. Prezzi, et al., PRB 77 (2008). C. Rödl, et al., PRB 77 (2008).
76 Outline 1 Motivation 2 Oneparticle Green s functions: GW approximation 3 Twoparticle Green s functions: BetheSalpeter equation 4 Micromacro connection
77 Micromacro connection Observation At long wavelength, external fields are slowly varying over the unit cell: dimension of the unit cell for silicon: 0.5 nm visible radiation 400 nm < λ < 800 nm Total and induced fields are rapidly varying: they include the contribution from electrons in all regions of the cell. Large and irregular fluctuations over the atomic scale.
78 Micromacro connection Observation One usually measures quantities that vary on a macroscopic scale. When we calculate microscopic quantities we need to average over distances that are large compared to the cell parameter small compared to the wavelength of the external perturbation. The differences between the microscopic fields and the averaged (macroscopic) fields are called the crystal local fields.
79 Suppose that we are able to calculate the microscopic dielectric function ɛ, how do we obtain the macroscopic dielectric function ɛ M that we measure in experiments?
80 Micromacro connection Fourier transform In a periodic medium, every function V (r, ω) can be represented by the Fourier series V (r, ω) = qg V (q + G, ω)e i(q+g)r or: V (r, ω) = q e iqr G V (q + G, ω)e igr = q e iqr V (q, r, ω) where: V (q, r, ω) = G V (q + G, ω)e igr V (q, r, ω) is periodic with respect to the Bravais lattice and hence is the quantity that one has to average to get the corresponding macroscopic potential V M (q, ω).
81 Micromacro connection Averages V M (q, ω) = 1 Ω c drv (q, r, ω) V (q, r, ω) = G V (q + G, ω)e igr Therefore: V M (q, ω) = G V (q + G, ω) 1 Ω c dre igr = V (q + 0, ω) The macroscopic average V M corresponds to the G = 0 component of the microscopic V. Example V ext (q, ω) = ɛ M (q, ω)v tot,m (q, ω)
82 Micromacro connection Averages V M (q, ω) = 1 Ω c drv (q, r, ω) V (q, r, ω) = G V (q + G, ω)e igr Therefore: V M (q, ω) = G V (q + G, ω) 1 Ω c dre igr = V (q + 0, ω) The macroscopic average V M corresponds to the G = 0 component of the microscopic V. Example V ext (q, ω) = ɛ M (q, ω)v tot,m (q, ω)
83 Micromacro connection Fourier transforms Fourier transform of a function f (r, r, ω): f (q + G, q + G, ω) = drdr e i(q+g)r f (r, r, ω)e +i(q+g )r f G,G (q, ω) Therefore the relation V tot (r 1, ω) = dr 2 ɛ 1 (r 1, r 2, ω)v ext (r 2, ω) in the Fourier space becomes: V tot (q + G, ω) = G ɛ 1 G,G (q, ω)v ext (q + G, ω)
84 Micromacro connection Fourier transforms Fourier transform of a function f (r, r, ω): f (q + G, q + G, ω) = drdr e i(q+g)r f (r, r, ω)e +i(q+g )r f G,G (q, ω) Therefore the relation V tot (r 1, ω) = dr 2 ɛ 1 (r 1, r 2, ω)v ext (r 2, ω) in the Fourier space becomes: V tot (q + G, ω) = G ɛ 1 G,G (q, ω)v ext (q + G, ω)
85 Micromacro connection Example Macroscopic dielectric function V tot,m (q, ω) = ɛ 1 M (q, ω)v ext(q, ω) V tot(q + G, ω) = G ɛ 1 G,G (q, ω)vext(q + G, ω) V ext is a macroscopic quantity: V M,tot (q, ω) = V tot(q + 0, ω) V tot,m (q, ω) = ɛ 1 G=0,G =0 (q, ω)vext(q, ω) ɛ 1 M (q, ω) = ɛ 1 ɛ M (q, ω) = G=0,G =0 1 (q, ω) ɛ 1 G=0,G =0 (q, ω)
86 Micromacro connection Example Macroscopic dielectric function V tot,m (q, ω) = ɛ 1 M (q, ω)v ext(q, ω) V tot(q + G, ω) = G ɛ 1 G,G (q, ω)vext(q + G, ω) V ext is a macroscopic quantity: V M,tot (q, ω) = V tot(q + 0, ω) V tot,m (q, ω) = ɛ 1 G=0,G =0 (q, ω)vext(q, ω) ɛ 1 M (q, ω) = ɛ 1 ɛ M (q, ω) = G=0,G =0 1 (q, ω) ɛ 1 G=0,G =0 (q, ω)
87 Micromacro connection Example Macroscopic dielectric function V tot,m (q, ω) = ɛ 1 M (q, ω)v ext(q, ω) V tot(q + G, ω) = G ɛ 1 G,G (q, ω)vext(q + G, ω) V ext is a macroscopic quantity: V M,tot (q, ω) = V tot(q + 0, ω) V tot,m (q, ω) = ɛ 1 G=0,G =0 (q, ω)vext(q, ω) ɛ 1 M (q, ω) = ɛ 1 ɛ M (q, ω) = G=0,G =0 1 (q, ω) ɛ 1 G=0,G =0 (q, ω)
88 Micromacro connection Example Macroscopic dielectric function V tot,m (q, ω) = ɛ 1 M (q, ω)v ext(q, ω) V tot(q + G, ω) = G ɛ 1 G,G (q, ω)vext(q + G, ω) V ext is a macroscopic quantity: V M,tot (q, ω) = V tot(q + 0, ω) V tot,m (q, ω) = ɛ 1 G=0,G =0 (q, ω)vext(q, ω) ɛ 1 M (q, ω) = ɛ 1 ɛ M (q, ω) = G=0,G =0 1 (q, ω) ɛ 1 G=0,G =0 (q, ω)
89 Micromacro connection Example Macroscopic dielectric function V tot,m (q, ω) = ɛ 1 M (q, ω)v ext(q, ω) V tot(q + G, ω) = G ɛ 1 G,G (q, ω)vext(q + G, ω) V ext is a macroscopic quantity: V M,tot (q, ω) = V tot(q + 0, ω) V tot,m (q, ω) = ɛ 1 G=0,G =0 (q, ω)vext(q, ω) ɛ 1 M (q, ω) = ɛ 1 ɛ M (q, ω) = G=0,G =0 1 (q, ω) ɛ 1 G=0,G =0 (q, ω)
90 Micromacro connection Macroscopic dielectric function V ext (q + G, ω) = G ɛ G,G (q, ω)v tot (q + G, ω) Remember: V ext is a macroscopic quantity: V ext (q, ω) = G ɛ G=0,G (q, ω)v tot (q + G, ω) V ext (q, ω) = ɛ G=0,G =0(q, ω)v tot,m (q, ω)+ G 0 ɛ G=0,G (q, ω)v tot (q+g, ω) V ext (q, ω) = ɛ M (q, ω)v tot,m (q, ω) ɛ M (q, ω) ɛ G=0,G =0(q, ω)
91 Micromacro connection Macroscopic dielectric function V ext (q + G, ω) = G ɛ G,G (q, ω)v tot (q + G, ω) Remember: V ext is a macroscopic quantity: V ext (q, ω) = G ɛ G=0,G (q, ω)v tot (q + G, ω) V ext (q, ω) = ɛ G=0,G =0(q, ω)v tot,m (q, ω)+ G 0 ɛ G=0,G (q, ω)v tot (q+g, ω) V ext (q, ω) = ɛ M (q, ω)v tot,m (q, ω) ɛ M (q, ω) ɛ G=0,G =0(q, ω)
92 Micromacro connection Macroscopic dielectric function V ext (q + G, ω) = G ɛ G,G (q, ω)v tot (q + G, ω) Remember: V ext is a macroscopic quantity: V ext (q, ω) = G ɛ G=0,G (q, ω)v tot (q + G, ω) V ext (q, ω) = ɛ G=0,G =0(q, ω)v tot,m (q, ω)+ G 0 ɛ G=0,G (q, ω)v tot (q+g, ω) V ext (q, ω) = ɛ M (q, ω)v tot,m (q, ω) ɛ M (q, ω) ɛ G=0,G =0(q, ω)
93 Micromacro connection Spectra ɛ M (q, ω) = 1 ɛ 1 G=0,G =0 (q, ω) Abs(ω) = lim q 0 Imɛ M (ω) = lim EELS(ω) = lim Imɛ 1 M q 0 q 0 Im (ω) = lim q 0 Imɛ 1 1 ɛ 1 G=0,G =0 (q, ω) G=0,G =0 (q, ω)
94 Micromacro connection Spectra ɛ M (q, ω) = 1 ɛ 1 G=0,G =0 (q, ω) Abs(ω) = lim q 0 Imɛ M (ω) = lim EELS(ω) = lim Imɛ 1 M q 0 q 0 Im (ω) = lim q 0 Imɛ 1 1 ɛ 1 G=0,G =0 (q, ω) G=0,G =0 (q, ω)
95 BSE vs. TDDFT: what in common? BSE L = L 0 + L 0 (v + Ξ)L TDDFT χ = χ 0 + χ 0 (v + f xc )χ
96 The Coulomb term v The Coulomb term v = v 0 + v
97 Local fields reloaded MicroscopicMacroscopic connection: local fields χ G,G (q, ω) = P G,G (q, ω) + P G,G1 (q, ω)v G1 (q)χ G1,G (q, ω) G 1 ɛ 1 G,G (q, ω) = δ G,G + v G (q)χ G,G (q, ω) ɛ M (q, ω) = 1 ɛ 1 G=0,G =0 (q, ω) Adler, Phys. Rev. 126 (1962); Wiser, Phys. Rev. 129 (1963).
98 Local fields reloaded MicroscopicMacroscopic connection: local fields ɛ M (q, ω) = 1 v G=0 (q) χ G=0,G =0(q, ω) χ G,G (q, ω) = P G,G (q, ω) + P G,G1 (q, ω) v G1 (q) χ G1,G (q, ω) G 1 v G (q) = 0 for G = 0 v G (q) = v G (q) for G 0 Hanke, Adv. Phys. 27 (1978).
99 Absorption Abs(ω) = lim q 0 Imɛ M (q, ω) Abs(ω) = lim q 0 Im [v G=0 (q) χ G=0,G =0(q, ω)] χ = P + P v χ EELS Absorption response to V ext + V macro ind Eels(ω) = lim q 0 Im[1/ɛ M (q, ω)] Eels(ω) = lim q 0 Im [v G=0 (q)χ G=0,G =0(q, ω)] χ = P + P(v 0 + v)χ Eels response to V ext
100 The Coulomb term v The Coulomb term v = v 0 + v longrange v 0 difference between Abs and Eels
101 Coulomb term v 0 : Abs vs. Eels F. Sottile, PhD thesis (2003)  Bulk silicon: absorption vs. EELS.
102 The Coulomb term v The Coulomb term v = v 0 + v longrange v 0 difference between Abs and Eels what about v?
103 The Coulomb term v The Coulomb term v = v 0 + v longrange v 0 difference between Abs and Eels what about v? v is responsible for crystal localfield effects
104 Coulomb term v: local fields v: local fields ɛ M = 1 v G=0 χ G=0,G =0 Set v = 0 in: χ G,G = χ 0 G,G + G 1 χ 0 G,G 1 v G1 χ G1,G χ G,G = χ 0 G,G Result: ɛ M = 1 v G=0 χ 0 G=0,G =0 that is: no localfield effects! (equivalent to Fermi s golden rule)
105 Coulomb term v: local fields v: local fields ɛ M = 1 v G=0 χ G=0,G =0 Set v = 0 in: χ G,G = χ 0 G,G + G 1 χ 0 G,G 1 v G1 χ G1,G χ G,G = χ 0 G,G Result: ɛ M = 1 v G=0 χ 0 G=0,G =0 that is: no localfield effects! (equivalent to Fermi s golden rule)
106 Coulomb term v: local fields v: local fields ɛ M = 1 v G=0 χ G=0,G =0 Set v = 0 in: χ G,G = χ 0 G,G + G 1 χ 0 G,G 1 v G1 χ G1,G χ G,G = χ 0 G,G Result: ɛ M = 1 v G=0 χ 0 G=0,G =0 that is: no localfield effects! (equivalent to Fermi s golden rule)
107 Coulomb term v: local fields Bulk silicon: absorption
108 Coulomb term v: local fields A. G. Marinopoulos et al., PRL 89 (2002)  Graphite EELS
109 What are local fields? Effective medium theory Uniform field E 0 applied to a dielectric sphere with dielectric constant ɛ in vacuum. From continuity conditions at the interface: P = 3 ɛ 1 4π ɛ + 2 E 0 Jackson, Classical electrodynamics, Sec. 4.4.
110 What are local fields? Effective medium theory Regular lattice of objects dimensionality d of material ɛ 1 in vacuum MaxwellGarnett formulas dot (O D system) wire (1D system) Imɛ 1 (ω) Imɛ M (ω) 9 [Reɛ 1 (ω) + 2] 2 + [Imɛ 1 (ω)] 2 Imɛ M (ω) Imɛ 1(ω) Imɛ Imɛ 1 (ω) M (ω) 4 [Reɛ 1 (ω) + 1] 2 + [Imɛ 1 (ω)] 2
111 What are local fields? F. Bruneval et al., PRL 94 (2005)  Si nanowires S. Botti et al., PRB 79 (2009)  SiGe nanodots
112 MBPT & TDDFT MBPT helps improving DFT & TDDFT DFT & TDDFT help improving MBPT
113 Conclusion (TD)DFT & MBPT... try to learn both!
114 Many thanks!
115 Acknowledgements Silvana Botti Fabien Bruneval Valerio Olevano Lucia Reining Francesco Sottile Valérie Véniard
Università degli Studi Roma TRE. Consorzio Nazionale Interuniversitario per le Scienze Fisiche della Materia
Università degli Studi Roma TRE e Consorzio Nazionale Interuniversitario per le Scienze Fisiche della Materia Dottorato di Ricerca in Scienze Fisiche della Materia XXIII ciclo Challenges for first principles
More informationPQM Supplementary Notes: Spin, topology, SU(2) SO(3) etc
PQM Supplementary Notes: Spin, topology, SU(2) SO(3) etc (away from the syllabus, but interesting material, based on notes by Dr J.M. Evans) 1 Rotations and Noncontractible Loops Rotations in R 3 can
More informationV ( r i, r j ) U[ρ( r i )]. Then particle i feels an average potential which depends on the particle density. ρ( r) = j.
Chapter 8 Mean Field Theory The concept of mean field theory is widely used for the description of interacting manybody systems in physics. The idea behind is to treat the manybody system not by summing
More information1 OneDimensional Magnetism
1 OneDimensional Magnetism HansJürgen Mikeska 1 and Alexei K. Kolezhuk 1,2 1 Institut für Theoretische Physik, Universität Hannover, Appelstaße 2, 30167 Hannover, Germany, mikeska@itp.unihannover.de
More informationThe Kondo Effect. Vijay B. Shenoy (shenoy@physics.iisc.ernet.in) Centre for Condensed Matter Theory Indian Institute of Science
The Kondo Effect Vijay B. Shenoy (shenoy@physics.iisc.ernet.in) Centre for Condensed Matter Theory Indian Institute of Science VBS The Kondo Effect 0 Overview What is Kondo effect? Motivation Why do this?
More informationWhich functional should I choose?
Contents Which functional should I choose? Dmitrĳ Rappoport, Nathan R. M. Crawford, Filipp Furche, and Kieron Burke December 15, 2008 5 Various directions in DFT development 29 5.1 Semilocal functionals................
More information11 The Kondo Effect. Frithjof B. Anders Theoretische Physik II Technische Universität Dortmund. Contents
11 The Kondo Effect Frithjof B. Anders Theoretische Physik II Technische Universität Dortmund Contents 1 Introduction 2 1.1 Resistance minimum............................... 3 1.2 Anderson model..................................
More informationProperties of Electrons, their Interactions with Matter and Applications in Electron Microscopy
Properties of Electrons, their Interactions with Matter and Applications in Electron Microscopy By Frank Krumeich Laboratory of Inorganic Chemistry, ETH Zurich, VladimirPrelogWeg 1, 8093 Zurich, Switzerland
More informationSpaceTime Approach to NonRelativistic Quantum Mechanics
R. P. Feynman, Rev. of Mod. Phys., 20, 367 1948 SpaceTime Approach to NonRelativistic Quantum Mechanics R.P. Feynman Cornell University, Ithaca, New York Reprinted in Quantum Electrodynamics, edited
More informationThe plasmoelectric effect: optically induced electrochemical potentials in resonant metallic structures
The plasmoelectric effect: optically induced electrochemical potentials in resonant metallic structures Matthew T. Sheldon and Harry A. Atwater Thomas J. Watson Laboratories of Applied Physics, California
More informationSingleElectron Devices and Their Applications
SingleElectron Devices and Their Applications KONSTANTIN K. LIKHAREV, MEMBER, IEEE Invited Paper The goal of this paper is to review in brief the basic physics of singleelectron devices, as well as their
More informationOrdering, metastability and phase transitions in twodimensional systems
J. Phys. C: Solid State Phys., Vol. 6, 1973. Erinted in Great Britain. @ 1973 Ordering, metastability and phase transitions in twodimensional systems J M Kosterlit7 and D J Thouless Department of Mathematical
More informationOBSTACLES ON THE WAY TOWARDS THE QUANTIZATION OF SPACE, TIME AND MATTER and possible resolutions
SPIN2000/20 OBSTACLES ON THE WAY TOWARDS THE QUANTIZATION OF SPACE, TIME AND MATTER and possible resolutions Gerard t Hooft Institute for Theoretical Physics University of Utrecht, Princetonplein 5 3584
More informationLearning Objectives for AP Physics
Learning Objectives for AP Physics These course objectives are intended to elaborate on the content outline for Physics B and Physics C found in the AP Physics Course Description. In addition to the five
More informationThe Hamiltonian Mean Field Model: From Dynamics to Statistical Mechanics and Back
The Hamiltonian Mean Field Model: From Dynamics to Statistical Mechanics and Back Thierry Dauxois 1, Vito Latora 2, Andrea Rapisarda 2, Stefano Ruffo 1,3, and Alessandro Torcini 3,4 1 Laboratoire de Physique,
More informationOn the Constitution of Atoms and Molecules. N. Bohr, Dr. phil. Copenhagen (Received July 1913)
N. Bohr, Philos. Mag. 26, 1 1913 Introduction On the Constitution of Atoms and Molecules N. Bohr, Dr. phil. Copenhagen (Received July 1913) In order to explain the results of experiments on scattering
More informationTwolevel states in glasses
Rep. Prog. Phys. 50 (1987) 16571708. Printed in the UK Twolevel states in glasses W A Phillips Cavendish Laboratory, University of Cambridge, Madingley Road, Cambridge CB3 OHE, UK Abstract This review
More informationSYNTHESIS AND PROPERTIES OF NEW SEMICONDUCTOR ALLOYS FOR LONG WAVELENGTH COHERENT EMITTERS ON GALLIUM ARSENIDE
UNIVERSITÀ DEGLI STUDI DI TRIESTE XVI CICLO DEL DOTTORATO DI RICERCA IN FISICA SYNTHESIS AND PROPERTIES OF NEW SEMICONDUCTOR ALLOYS FOR LONG WAVELENGTH COHERENT EMITTERS ON GALLIUM ARSENIDE DOTTORANDO
More informationReinitialization Free Level Set Evolution via Reaction Diffusion
1 Reinitialization Free Level Set Evolution via Reaction Diffusion Kaihua Zhang, Lei Zhang, Member, IEEE, Huihui Song and David Zhang, Fellow, IEEE Abstract This paper presents a novel reactiondiffusion
More informationCan quantum computing solve classically unsolvable problems?
Can quantum computing solve classically unsolvable problems? Andrew Hodges Wadham College University of Oxford Oxford OX1 3PN, U.K. andrew.hodges@wadh.ox.ac.uk Abstract: T. D. Kieu has claimed that a quantum
More informationPhysik Department. Matrix Product Formalism
Physik Department Matrix Product Formalism Diplomarbeit von María Gracia Eckholt Perotti Angefertigt an der Technische Universität München und am MaxPlanckInstitut für Quantenoptik Garching, September
More informationEffect of acceleration on quantum systems
University of Naples Federico II Faculty of Mathematical, Physical and Natural Sciences Master studies in Physics Degree Thesis Effect of acceleration on quantum systems Academic Year 2007 2008 Supervisor
More informationBreakdown of Local Thermodynamic Equilibrium...
II Stellar Atmospheres Copyright (2003) George W. Collins, II 15 Breakdown of Local Thermodynamic Equilibrium... Thus far we have made considerable use of the concepts of equilibrium. In the stellar interior,
More informationTheory of EIT in an Ideal ThreeLevel System
Chapter 3 Theory of EIT in an Ideal ThreeLevel System 3.1 ThreeLevel Systems So far the interaction of a twolevel system (or series of twolevel systems) with a near resonant, monochromatic field has
More informationSpecific Intensity. I ν =
Specific Intensity Initial question: A number of active galactic nuclei display jets, that is, long, nearly linear, structures that can extend for hundreds of kiloparsecs. Many have two oppositelydirected
More informationChapter 9 Unitary Groups and SU(N)
Chapter 9 Unitary Groups and SU(N) The irreducible representations of SO(3) are appropriate for describing the degeneracies of states of quantum mechanical systems which have rotational symmetry in three
More informationUnderstanding the FiniteDifference TimeDomain Method. John B. Schneider
Understanding the FiniteDifference TimeDomain Method John B. Schneider June, 015 ii Contents 1 Numeric Artifacts 7 1.1 Introduction...................................... 7 1. Finite Precision....................................
More informationClassical Mechanics. Joel A. Shapiro
Classical Mechanics Joel A. Shapiro April 21, 2003 Copyright C 1994, 1997 by Joel A. Shapiro All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted
More informationDerivation of Mean Field Equations for Classical Systems
Derivation of Mean Field Equations for Classical Systems Master Thesis iklas Boers Derivation of Mean Field Equations for Classical Systems Master Thesis iklas Boers Fakultät für Mathematik, Informatik
More informationSymmetric Stretch: allows molecule to move through space
BACKGROUND INFORMATION Infrared Spectroscopy Before introducing the subject of IR spectroscopy, we must first review some aspects of the electromagnetic spectrum. The electromagnetic spectrum is composed
More information