You are on page 1of 30

Universe in a helium droplet

†a a
^ ^

topological Fermi liquids: from Migdal jump to topological Khodel fermion condensate Aalto University G. Volovik
Landau Institute

Migdal-100, Chernogolovka June 25, 2011

* Topological media as topological objects in p-space * Fermi surface as vortex in p-space * Topological protection of Migdal jump and other singularities on Fermi surface * Khodel Fermi condensate as π-vortex * Bulk-surface & bulk-vortex correspondence in topological matter * New life of Fermi condensate: Topologically protected flat bands * Can Fermi surface terminate? Fermi-arc in topological matter with Weyl points * 3He Universe

classes of topological matter as momentum-space objects
metals, normal 3He

Fermi surface: vortex ring in p-space

ω p (p )
y z

pz py

fully gapped topological matter: skyrmion in p-space

p p ∆Φ=2π
F

x

px

H = + c σ .p

3He-B, topological insulators, 3He-A film, vacuum of Standard Model

Weyl point - hedgehog in p-space 3He-A, vacuum of SM, topological semimetals

ω p ,p
y z

py pz
Fermi surface

π

p=p

1

p
π

x

Fermi arc on 3He-A surface & flat band on vortex: Dirac strings in p-space terminating on monopole

p=p

pz
flat band

2

flat band (Khodel state): π-vortex in p-space

Topological stability of Fermi surface
Energy spectrum of non-interacting gas of fermionic atoms

Green's function G-1= iω − ε(p) ω

ε(p)

p 2 p2 p2 = –µ = – F 2m 2m 2m p (p )

ε>0
empty levels

ε<0
occupied levels: Fermi sea

y

z

p p=p
F

p
F

x

Fermi surface

ε=0

∆Φ=2π Fermi surface: vortex ring in p-space phase of Green's function G(ω,p)=|G|e iΦ

is Fermi surface a domain wall in momentum space?

has winding number N = 1 no! it is a vortex ring

Migdal jump & p-space topology
* Singularity at Fermi surface is robust to perturbations: winding number N=1 cannot change continuously, interaction (perturbative) cannot destroy singularity * Typical singularity: Migdal jump

ω

p (p )
y z

n(p) p

p p ∆Φ=2π
F

x

pF

* Other types of singularity: Luttinger Fermi liquid, marginal Fermi liquid, pseudo-gap ...

Fermi surface: vortex line in p-space

G(ω,p) =

Z(p,ω)

G(ω,p)=|G|e iΦ

iω − ε(p)

Z(p,ω) = (ω2 + ε2(p))γ
* Zeroes in Green's function instead of poles ( for γ > 1/2) have the same winding number N=1

From Migdal jump to Landau Fermi-liquid
* Typical singularity in Fermi liquid: Migdal jump * Quasiparticles in Fermi liquid as in Fermi gas

ω Landau theory of Fermi liquid is topologically protected & thus is universal p (p )
y z

p
all metals have Fermi surface ...

p
F

x

∆Φ=2π
Not only metals. Some superconductore too! * Fermi surface in superconductors
"Stability conditions & Fermi surface topologies in a superconductor" Gubankova-Schmitt-Wilczek, Phys.Rev. B74 (2006) 064505

Fermi surface: vortex line in p-space

G(ω,p)=|G|e iΦ

 

quantized vortex in r-space ≡ Fermi surface in p-space homotopy group π1 Topology in r-space
z how is it in p-space ? y x ∆Φ=2π vortex ring Fermi surface

Topology in p-space
ω p (p )
y z

winding number N1 = 1
∆Φ=2π

p

p
F

x

Ψ(r)=|Ψ| e iΦ
scalar order parameter of superfluid & superconductor

G(ω,p)=|G|e iΦ
Green's function (propagator)

classes of mapping S1 → U(1)
manifold of broken symmetry vacuum states

classes of mapping S1 → GL(n,C)
space of non-degenerate complex matrices

non-topological flat bands due to interaction Khodel-Shaginyan fermion condensate JETP Lett. 51, 553 (1990)

GV, JETP Lett. 53, 222 (1991) Nozieres, J. Phys. (Fr.) 2, 443 (1992)

E{n(p)}

δE{n(p)} = ∫ε(p)δn(p)ddp = 0
solutions:

ε(p) = 0

or

δn(p)=0 ε(p) p

n(p)
pF

ε(p) p

n(p) p1 flat band p2 δn(p)=0

δn(p)=0 ε(p) = 0 δn(p)=0

ε(p) = 0

δn(p)=0

splitting of Fermi surface to flat band

Flat band as momentum-space dark soliton terminated by half-quantum vortices

ω p ,p
y z

π

p=p

p
π

1

x

half-quantum vortices in 4-momentum space

p=p

2

ε=0
Khodel-Shaginyan fermionic condensate (flat band)

phase of Green's function changes by π across the "dark soliton"

topological correspondence: topology in bulk protects gapless fermions on the surface or in vortex core bulk-surface correspondence:
2D Quantum Hall insulator & 3He-A film 3D topological insulator superfluid 3He-B superfluid 3He-A, Weyl point semimetal graphene semimetal with Fermi lines chiral edge states Dirac fermions on surface Majorana fermions on surface Fermi arc on surface dispersionless 1D flat band on surface 2D flat band on the surface

bulk-vortex correspondence:
superfluid 3He-A 1D flat band of zero modes in the core

Bulk-surface correspondence: Edge states in quantum Hall topological insulators & 3He-A film
2D non-topological insulator or vacuum 2D topological insulator y
rre nt

y
cu rre nt

~ N=0
empty

2D non-topological insulator or vacuum

~ N=1

~ N=0
x

E(py)

cu

E(py)

empty

0

py

0 occupied right moving edge states

py

occupied left moving edge states

p-space skyrmion

p-space invariant in terms of Green's function & topological QPT

N3=

~

1 e tr ∫ d2p dω G ∂µ G-1 G ∂ν G-1G ∂λ G-1 GV & Yakovenko 24π2 µνλ ~ J. Phys. CM 1, 5263 (1989) N3 = 6 quantum phase transitions in thin 3He-A film ~ N3 = 4
~ transition between plateaus must occur via gapless state!

N3 = 2 gap N3 = 0 a1 QPT from trivial to topological state
~

gap

gap

a film thickness

a2 a3 plateau-plateau QPT between topological states

topology of graphene nodes

N=

1 tr [K o dl H-1 ∂ H] ∫ l 4πi
Ν = +1 Ν = −1 Ν = +1 Ν = +1 Ν = −1

K - symmetry operator,
commuting or anti-commuting with H close to nodes: Ν = −1

HΝ = +1 = τxpx + τypy HΝ = −1 = τxpx − τypy K = τz

exotic fermions: massless fermions with quardatic dispersion semi-Dirac fermions fermions with cubic and quartic dispersion

bilayer graphene double cuprate layer surface of top. insulator neutrino physics

N=

1 tr [ K o dl H-1 ∂ H] ∫ l 4πi
N=0 E p p p
x

N=+1 E p

N=+1 E

N=+2 E p p
x

y x

p

+
E=cp

y

=

y

+
E2 = (p2/ 2m)2

y

p

x

E=cp
Dirac fermions

E2 = 2c2p2 + 4m2
massive fermions

massless fermions with quadratic dispersion

multiple Fermi point
cubic spectrum in trilayer graphene

T. Heikkilä & GV arXiv:1010.0393

N=+1

N= 1+1+1 = 3
N=+1

N=+1

+
E=cp E=cp

+
E=cp

=

E = p3 E = − p3 N= +3
p
x

multilayered graphene

spectrum in the outer layers

N = 1 + 1 + 1 + ...

E = pN E = − pN what kind of induced gravity emerges near degenerate Fermi point?

route to topological flat band on the surface of 3D material

Nodal spiral generates flat band on the surface projection of spiral on the surface determines boundary of flat band

N1 =

1 tr [ K o dl H-1 ∂ H] ∫ l 4πi C

at each (px,py) except the boundary of circle one has 1D gapped state (insulator)

Noutside = 0 trivial 1D insulator Ninside = 1
topological 1D insulator

N1 = 1
C Coutside Cinside

at each (px,py) inside the circle one has 1D gapless edge state this is flat band

Nodal spiral generates flat band on the surface projection of nodal spiral on the surface determines boundary of flat band

lowest energy states: surface states form the flat band

energy spectrum in bulk (projection to px , py plane)

Gapless topological matter with protected flat band on surface or in vortex core

non-topological flat bands due to interaction Khodel-Shaginyan fermion condensate
JETP Lett. 51, 553 (1990) GV, JETP Lett. 53, 222 (1991) Nozieres, J. Phys. (Fr.) 2, 443 (1992)

E(p) flat band p1 p2 p

splitting of Fermi surface to flat band

Bulk-surface correspondence: Flat band on the surface of topological semimetals

flat band on the surface

projection of spiral on the surface determines boundary of the flat band

topologically protected nodal line in the form of spiral

N1 = 1
C

N1 =

1 tr [ K o dl H-1 ∂ H] ∫ l 4πi C

flat band in soliton

H = τ3 (px2+pz2−pF2 )/2m + τ1c(z)pz
nodes at pz = 0 and px = pF
2 2

N=

1 tr [ K o dl H-1 ∂ H] ∫ l 4πi
soliton spectrum in soliton

c

E( px) flat band pF -pF

0

z

Flat band on the graphene edge

N = +1

1 tr [ K o dl H-1 ∂ H] N= ∫ l 4πi

N = +1 N = −1

N=0 flat band N = +1 N=0 N = +1

flat band
px

Surface superconductivity in topological semimetals: route to room temperature superconductivity

Extremely high DOS of flat band gives high transition temperature:
normal superconductors: exponentially suppressed transition temperature

1= g ∫

d 2p 1 2h2 E(p)

flat band superconductivity: linear dependence of Tc on coupling g

Tc = TF exp (-1/gν)
interaction DOS

"Recent studies of the correlations between the internal microstructure of the samples and the transport properties suggest that superconductivity might be localized at the interfaces between crystalline graphite regions of different orientations, running parallel to the graphene planes." PRB. 78, 134516 (2008)

Tc ∼ gSFB
interaction flat band area

multiple Fermi point

Kathryn Moler: possible 2D superconductivityof twin boundaries

Weyl point

pz py

N3 =
px

1 e dSk g . (∂pi g × ∂pj g) = 1 ijk ∫ 8π
over 2D surface S in 3D momentum space

Weyl point: hedgehog in p-space

relativistic quantum fields & gravity emerging near Weyl point
Atiyah-Bott-Shapiro construction: linear expansion near Weyl point in terms of Dirac Γ-matrices emergent relativity H = eak Γa .(pk − p0) k

primary object: tetrad eaµ secondary object: metric gµν = ηab eaµ ebν

all ingredients of Standard Model : chiral fermions & gauge fields emerge in low-energy corner together with spin, Dirac Γ−matrices, gravity & physical laws: Lorentz & gauge invariance, equivalence principle, etc

gµν(pµ- eAµ - eτ .Wµ)(pν- eAν - eτ .Wν) = 0
effective effective effective metric: isotopic spin SU(2) gauge emergent gravity field effective effective electric charge electromagnetic field e = + 1 or −1

From Weyl point to quantum Hall topological insulators N3 = 1 e dSk g . (∂pi g × ∂pj g) ijk ∫ 8π
over 2D surface S in 3D momentum space

top. invariant for Weyl point in 3+1 system pz ∼ N3(pz) = 0
2D trivial insulator Weyl point

hedgehog in p-space ∼ ∼ N3 = N3(pz <p0) − N3(pz >p0)
at each pz one has 2D insulator or fully gapped 2D superfluid

py ∼ N3(pz) = 1

2D topological Hall insulator

px

~ N 3 (p z ) =

1 ∫ dpxdpy g . (∂px g × ∂py g) 4π over the whole 2D momentum space
or over 2D Brillouin zone

skyrmion in p-space

top. invariant for fully gapped 2+1 system

3D matter with Weyl points: Topologically protected flat band in vortex core ∼ N3(pz) = 0
at each pz between two values: 2D topological Hall insulator: zero energy states E (pz)=0, in the vortex core (1D flat band 1011.4665)

vortices in r-space

pz

2D trivial insulator

z

pF cos λ
flat band Weyl point

∼ N3(pz) = 1

2D topological QH insulator

− pF cos λ
Weyl point

∼ N3(pz) = 0

2D trivial insulator GV & Yakovenko (1989)

N3(pz) =

~

1 tr dp dp dω G ∂ G-1 G ∂ G-1G ∂ G-1 ∫ x y ω px py 2 4π

Topologically protected flat band in vortex core of superfluids with Weyl points
E (pz , Q) flat band in spectrum of fermions bound to core of 3He-A vortex (Kopnin-Salomaa 1991) pz E (pz) continuous spectrum bound states flat band of bound states terminates on zeroes of continuous spectrum (i.e on Weyl points) Weyl point flat band Weyl point

pz

3He-A with Weyl points: Topologically protected Fermi arc on the surface pz ∼ N3(pz) = 0

for each |pz | < pF cos λ one has 2D topological Hall insulator with zero energy edge states E (pz)=0 (Dirac valley or Fermi arc PRB 094510,PRB 205101) 2D trivial insulator

pF cos λ
Weyl point Fermi arc 2D topological insulator E (pz)=0

E (pz)=0

Fermi arc

py ∼ N3(pz) = 1 px

− pF cos λ
Weyl point

x=xL

∼ N3(pz) = 0

2D trivial insulator

x=xR

Fermi arc: Fermi surface separates positive and negative energies, but has boundaries
E (py, pz) > 0

Fermi surface py
E (py = 0, |pz| < pF) = 0 E (py, pz) < 0

ordinary Fermi surface has no boundaries Fermi surface of bound states may leak to higher dimension

pz

Fermi surface of localized states is terminated by projections of Weyl points when localized states merge with continuous spectrum

L spectrum of
edge states on left wall
PRB 094510

R spectrum of
edge states on right wall

Conclusion:

p-space topology determines fundamental many-body quantum phenomena

universality classes of quantum vacua effective field theories in these quantum vacua (emergent gravity & QED) topological quantum phase transitions (Lifshitz, plateau, etc.) quantization of Hall and spin-Hall conductivity topological Chern-Simons & Wess-Zumino terms quantum statistics of topological objects bulk-surface & bulk-vortex correspondence exotic fermions: Majorana fermions; flat band; Fermi arc; cubic, quartic etc. dispersion chiral anomaly & vortex dynamics, etc. flat band & room-temperature superconductivity superfuid phases 3He serve as primer for topological matter: quantum vacuum of Standard Model, topological superconductors, topological insulators, topological semimetals, etc. we need: low T, high H, miniaturization, atomically smooth surface, nano-detectors, ... fabrication of samples for room-temperature superconductivity

Emergence & effective theories Weyl, Majorana & Dirac fermions Torsion & spinning strings, torsion instanton Vacuum polarization, screening - antiscreening, running coupling Fermion zero modes on strings & walls Symmetry breaking (anisotropy of vacuum) Antigravitating (negative-mass) string Parity violation -- chiral fermions Gravitational Aharonov-Bohm effect Vacuum instability in strong fields, pair production unity of physics Domain wall terminating on string Casimir force, quantum friction String terminating on domain wall Fermionic charge of vacuum Monopoles on string & Boojums Higgs fields & gauge bosons Witten superconducting string 3He Grand Unification Momentum-space topology Soft core string, Q-balls Hierarchy problem, Supersymmetry Z &W strings Kibble mechanism Gap nodes Neutrino oscillations skyrmions Dark matter detector Low -T scaling physical Chiral anomaly & axions Alice string cosmic Primordial magnetic field Pion string mixed strings vacuum Spin & isospin, String theory Broken state reversal Baryogenesis by textures time CPT-violation, GUT & strings 1/2-vortex, vortex dynamics Cosmological & high-T & chiral Inflation Films: FQHE, Newton constants High Energy supercosmology Branes Statistics & charge of dark energy Physics conductivity matter skyrmions & vortices dark matter creation Effective gravity 1D, 2D systems Edge states; flat band, Fermi arc Bi-metric & vacuum Condensed topological spintronics conformal gravity Gravity 3He gravity Matter 1D fermions in vortex core Graviton, dilaton insulators, Critical fluctuations Spin connection semimetals Mixture of condensates black Rotating vacuum Plasma Phenome Vector & spinor condensates Vacuum dynamics holes QCD BEC BEC of quasipartcles, conformal anomaly Physics nology magnon BEC & laser ergoregion, event horizon neutron meron, skyrmion, 1/2 vortex hydrodynamics Hawking & Unruh effects stars quark nuclear phase General; relativistic; black hole entropy matter physics transitions disorder random spin superfluidity Vacuum instability multi-fluid Superfluidity of neutron star Quark condensate Nuclei vs quantum phase transitions anisotropy rotating superfluid vortices, glitches Nambu--Jona-Lasinio 3He droplet & momentum-space topology Larkin- Imry-Ma Shear flow instability shear flow instability Vaks--Larkin classes of random Shell model Magnetohydrodynamic matrices Color superfluidity Pair-correlations Relativistic plasma Turbulence of vortex lines Savvidi vacuum Collective modes Photon mass propagating vortex front Quark confinement, QCD cosmology Vortex Coulomb velocity independent Intrinsic orbital momentum of quark matter plasma Reynolds number