21.05.2014 Views

Topological Insulators and Superconductors

Topological Insulators and Superconductors

Topological Insulators and Superconductors

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>Topological</strong> <strong>Insulators</strong> <strong>and</strong> <strong>Superconductors</strong><br />

JHU 2010<br />

Shoucheng Zhang, Stanford University


Colloborators<br />

Stanford group: Xiaoliang Qi, Andrei Bernevig, Congjun Wu, Chaoxing Liu,<br />

Taylor Hughes, Sri Raghu, Suk-bum Chung<br />

Stanford experimentalists: Yulin Chen, Ian Fisher, ZX Shen, Yi Cui, Aharon<br />

Kapitulnik, …<br />

Wuerzburg colleagues: Laurens Molenkamp, Hartmut Buhmann, Markus Koenig<br />

Ewelina Hankiewicz, Bjoern Trauzettle<br />

IOP colleagues: Zhong Fang, Xi Dai, Haijun Zhang, …<br />

Tsinghua colleagues: Qikun Xue, Jinfeng Jia, Xi Chen,…


Outline<br />

Models <strong>and</strong> materials of topological insulators<br />

<strong>Topological</strong> field theory, experimental proposals<br />

<strong>Topological</strong> Mott insulators<br />

<strong>Topological</strong> superconductors <strong>and</strong> superfluids, experimental proposals


The search for new states of matter<br />

The search for new elements led to a golden age of chemistry.<br />

The search for new particles led to the golden age of particle physics.<br />

In condensed matter physics, we ask what are the fundamental states of matter?<br />

In the classical world we have solid, liquid <strong>and</strong> gas. The same H 2<br />

O molecules can<br />

condense into ice, water or vapor.<br />

In the quantum world we have metals, insulators, superconductors, magnets etc.<br />

Most of these states are differentiated by the broken symmetry.<br />

Crystal: Broken<br />

translational symmetry<br />

Magnet: Broken<br />

rotational symmetry<br />

Superconductor: Broken<br />

gauge symmetry


Discovery of the 2D <strong>and</strong> 3D topological insulator<br />

HgTe Theory: Bernevig, Hughes <strong>and</strong> Zhang, Science 314, 1757 (2006)<br />

Experiment: Koenig et al, Science 318, 766 (2007)<br />

BiSb Theory: Fu <strong>and</strong> Kane, PRB 76, 045302 (2007)<br />

Experiment: Hsieh et al, Nature 452, 907 (2008)<br />

Bi2Te3, Sb2Te3, Bi2Se3 Theory: Zhang et al, Nature Physics 5, 438 (2009)<br />

Theory <strong>and</strong> Experiment Bi2Se3: Xia et al, Nature Physics 5, 398 (2009),<br />

Experiment BieTe3: Chen et al Science 325, 178 (2009)


<strong>Topological</strong> Insulator is a New State of Quantum Matter


Chiral (QHE) <strong>and</strong> helical (QSHE) liquids in D=1<br />

ε<br />

ε<br />

k<br />

k<br />

-k F<br />

k F<br />

-k F<br />

k F<br />

The QHE state spatially separates the two<br />

chiral states of a spinless 1D liquid<br />

The QSHE state spatially separates the<br />

four chiral states of a spinful 1D liquid<br />

2=1+1 4=2+2<br />

x<br />

x<br />

Kane <strong>and</strong> Mele (without L<strong>and</strong>au levels)<br />

Benervig <strong>and</strong> Zhang (with L<strong>and</strong>au levels)


Quantum protection by T 2 =-1 (Qi&Zhang, Phys Today, Jan, 2010,<br />

Wu,Bernevig&Zhang, PRL, 2006, Xu&Moore, PRB2006)<br />

Transport experiments:<br />

Molenkamp group<br />

STM experiments:<br />

Yazdani group, Kapitulnik group, Xue<br />

group


From topology to chemistry: the search for the QSH state<br />

• Type III quantum<br />

wells work. HgTe has a<br />

negative b<strong>and</strong> gap!<br />

(Bernevig, Hughes <strong>and</strong><br />

Zhang, Science 2006)<br />

B<strong>and</strong>gap energy (eV)<br />

• Tuning the thickness of<br />

the HgTe/CdTe<br />

quantum well leads to a<br />

topological quantum<br />

phase transition into the<br />

QSH state.<br />

• Sign of the Dirac mass<br />

term determines the<br />

topological term in field<br />

theory<br />

6.0<br />

5.5<br />

5.0<br />

4.5<br />

4.0<br />

3.5<br />

3.0<br />

2.5<br />

2.0<br />

1.5<br />

1.0<br />

0.5<br />

0.0<br />

-0.5<br />

B<strong>and</strong>gap vs. lattice constant<br />

(at room temperature in zinc blende structure)<br />

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.7<br />

lattice constant a [Å] 0


B<strong>and</strong> Structure of HgTe<br />

S<br />

P<br />

P 3/2<br />

S<br />

P 1/2<br />

S<br />

P<br />

S<br />

P 3/2<br />

P 1/2


B<strong>and</strong> inversion in HgTe leads to a topological quantum<br />

phase transition<br />

Let us focus on E1, H1<br />

b<strong>and</strong>s close to<br />

crossing point<br />

HgTe<br />

HgTe<br />

E1<br />

H1<br />

CdTe<br />

H1<br />

CdTe<br />

CdTe<br />

E1<br />

CdTe<br />

normal<br />

inverted


The model of the 2D topological insulator (BHZ, Science 2006)<br />

Square lattice with 4-orbitals per site:<br />

s<br />

,<br />

x y<br />

x y<br />

↑ , s,<br />

↓ , ( p + ip , ↑ , − ( p − ip ), ↓<br />

Nearest neighbor hopping integrals. Mixing matrix<br />

elements between the s <strong>and</strong> the p states must be<br />

odd in k.<br />

H<br />

eff<br />

( k<br />

x<br />

, k<br />

y<br />

)<br />

⎛h(<br />

k)<br />

⎜<br />

⎝ 0<br />

=<br />

∗<br />

h<br />

0 ⎞<br />

⎟<br />

( −k)<br />

⎠<br />

h(<br />

k)<br />

=<br />

⎛<br />

⎜<br />

⎝<br />

A(sin<br />

m(<br />

k)<br />

k<br />

x<br />

+ i sin<br />

k<br />

y<br />

)<br />

A(sin<br />

kx<br />

− i sin<br />

− m(<br />

k)<br />

k<br />

y<br />

) ⎞<br />

⎟<br />

⎠<br />

≡<br />

d<br />

a<br />

a<br />

( k)<br />

τ<br />

⇒<br />

⎛<br />

⎜<br />

⎝<br />

m + Bk<br />

A(<br />

k<br />

x<br />

2<br />

+ ik<br />

y<br />

)<br />

A(<br />

kx<br />

− ik<br />

− m − Bk<br />

y<br />

2<br />

) ⎞<br />

⎟<br />

⎠<br />

Similar to relativistic Dirac equation in 2+1 dimensions, with a mass term<br />

tunable by the sample thickness d! m/Bd c<br />

.


Mass domain wall<br />

Cutting the Hall bar along the y-direction we see a domain-wall structure in the<br />

b<strong>and</strong> structure mass term. This leads to states localized on the domain wall<br />

which still disperse along the x-direction. (similar to Jackiw-Rebbi solition).<br />

y<br />

y<br />

x<br />

x<br />

m/B


Experimental setup<br />

• High mobility samples<br />

of HgTe/CdTe quantum<br />

wells have been<br />

fabricated.<br />

• Because of the small<br />

b<strong>and</strong> gap, about several<br />

meV, one can gate dope<br />

this system from n to p<br />

doped regimes.<br />

• Two tuning parameters,<br />

the thickness d of the<br />

quantum well, <strong>and</strong> the<br />

gate voltage.<br />

•(Koenig et al, Science 2007)


Experimental observation of the QSH edge state<br />

(Konig et al, Science 2007)<br />

x<br />

x


Nonlocal transport in the QSH regime, (Roth et al Science<br />

2009)<br />

25<br />

20<br />

R 14,14<br />

=3/4 h/e 2<br />

R (kΩ)<br />

15<br />

10<br />

1<br />

2<br />

I: 1-4<br />

V: 2-3<br />

5<br />

4<br />

3<br />

R 14,23<br />

=1/4 h/e 2<br />

0<br />

0.0 0.5 1.0 1.5 2.0<br />

V* (V)


3D insulators with a single Dirac cone on the surface<br />

(a)<br />

t 1<br />

t 2<br />

t 3<br />

z<br />

Quintuple<br />

layer<br />

Bi<br />

Se1<br />

Se2<br />

y<br />

x<br />

(b)<br />

(c)<br />

C<br />

A<br />

B<br />

C<br />

A<br />

B<br />

y<br />

x<br />

C


Relevant orbitals of Bi2Se3 <strong>and</strong> the b<strong>and</strong> inversion<br />

(a)<br />

(b)<br />

0.6<br />

Bi<br />

Se<br />

E (eV)<br />

0.2<br />

λ c<br />

-0.2<br />

0 0.2 0.4<br />

λ (eV)<br />

(I) (II) (III)


Bulk <strong>and</strong> surface states from first principle calculations<br />

(a) Sb 2 Se 3 (b) Sb 2 Te 3<br />

(c) Bi 2 Se 3 (d) Bi 2 Te 3


Model for topological insulator Bi2Te3, (Zhang et al, 2009)<br />

Pz+, up, Pz-, up, Pz+, down, Pz-, down<br />

Single Dirac cone on the surface of Bi2Te3<br />

Surface of Bi2Te3 = ¼ Graphene !


Oscillatory crossover from 3D to 2D (Liu et al, 0908.3654)


Arpes experiment on Bi2Se3 surface states, Hasan group


Arpes experiment on Bi2Te3 surface states, Shen group<br />

Doping evolution of the FS <strong>and</strong> b<strong>and</strong> structure<br />

Undoped<br />

Under‐doped<br />

Optimallydoped<br />

Over‐doped<br />

E F (undoped)<br />

BCB bottom<br />

BVB bottom<br />

Dirac point position


AB oscillations in nano-ribbon of Bi2Se3, Cui group


STM experiments on quasi-particle interference, Yazdani,<br />

Kapitulnik, Xue groups, theory Lee et al 0910.1668


General definition of a topological insulator<br />

• Z2 topological b<strong>and</strong> invariant<br />

in momentum space based on<br />

single particle states.<br />

(Fu, Kane <strong>and</strong> Mele, Moore <strong>and</strong> Balents,<br />

Roy)<br />

• <strong>Topological</strong> field theory term<br />

in the effective action. Generally<br />

valid for interacting <strong>and</strong><br />

disordered systems. Directly<br />

measurable physically. Relates<br />

to axion physics! (Qi, Hughes <strong>and</strong><br />

Zhang)<br />

• For a periodic system, the system is<br />

time reversal symmetric only when<br />

θ=0 => trivial insulator<br />

θ=π => non-trivial insulator


θ term with open boundaries<br />

• θ=π implies QHE on the boundary with<br />

2<br />

1 e<br />

σxy<br />

=<br />

2 h<br />

• For a sample with boundary, it is only insulating when a<br />

small T-breaking field is applied to the boundary. The<br />

surface theory is a CS term, describing the half QH.<br />

• Each Dirac cone contributes σ xy<br />

=1/2e 2 /h to the QH.<br />

Therefore, θ=π implies an odd number of Dirac cones on<br />

the surface!<br />

T breaking<br />

M<br />

E<br />

j //<br />

• Surface of a TI = ¼ graphene


Generalization of the QH topology state in d=2 to<br />

time reversal invariant topological state in d>2, in<br />

Science 2001


Time Reversal Invariant <strong>Topological</strong> <strong>Insulators</strong><br />

• Zhang & Hu 2001: TRI topological insulator in D=4. => Root state of all TRI<br />

topological insulators.<br />

• Murakami, Nagaosa & Zhang 2004: Spin Hall insulator with spin-orbit coupled<br />

b<strong>and</strong> structure.<br />

• Kane <strong>and</strong> Mele, Bernevig <strong>and</strong> Zhang 2005: Quantum spin Hall insulator with<br />

<strong>and</strong> without L<strong>and</strong>au levels.<br />

• Fu, Kane & Mele, Moore <strong>and</strong> Balents, Roy 2007: <strong>Topological</strong> b<strong>and</strong> theory based<br />

on Z2<br />

• Qi, Hughes <strong>and</strong> Zhang 2008: <strong>Topological</strong> field theory based on F F dual.<br />

TRB Chern-Simons term in D=2: A 0<br />

=even, A i<br />

=odd<br />

S D<br />

2<br />

3<br />

2<br />

d k da(<br />

k)<br />

d x A(<br />

x)<br />

∧ dA(<br />

x)<br />

= ∫ ∫<br />

TRI Chern-Simons term in D=4: A 0<br />

=even, A i<br />

=odd<br />

S D<br />

4<br />

5<br />

4<br />

d k da(<br />

k)<br />

∧ da(<br />

k)<br />

d x A(<br />

x)<br />

∧ dA(<br />

x)<br />

∧ dA(<br />

x)<br />

= ∫ ∫


Dimensional reduction<br />

• From 4D QHE to the 3D topological insulator<br />

Zhang & Hu, Qi, Hughes & Zhang<br />

S<br />

4DQH<br />

⇒<br />

⇒<br />

∫<br />

S<br />

d<br />

3D<br />

=<br />

3<br />

=<br />

∫<br />

xdt(<br />

∫<br />

d<br />

∫<br />

d<br />

4<br />

xdtε<br />

3<br />

dx<br />

5<br />

A<br />

μνρστ<br />

5<br />

A<br />

μ<br />

( x,<br />

t))<br />

ε<br />

xdtθ<br />

( x,<br />

t)<br />

ε<br />

F<br />

νρ<br />

νρστ<br />

νρστ<br />

F<br />

F<br />

στ<br />

F<br />

νρ<br />

νρ<br />

F<br />

F<br />

στ<br />

στ<br />

x 5<br />

A 5<br />

θ<br />

(x, y, z)<br />

S<br />

• From 3D axion action to the 2D QSH<br />

3D<br />

⇒<br />

⇒<br />

∫<br />

S<br />

=<br />

d<br />

2D<br />

∫<br />

2<br />

μ e μρσ<br />

3 νρστ<br />

J2D<br />

= ε ∂σϕ<br />

∂ θ<br />

d xdtε<br />

Aν<br />

∂<br />

ρθ<br />

∂σ<br />

A<br />

2<br />

ρ<br />

τ<br />

2π<br />

ρστ<br />

xdtε<br />

∫<br />

μ e μσ<br />

( dzAz<br />

( x,<br />

t))<br />

∂<br />

ρθ<br />

∂σ<br />

Aτ<br />

⇔J1<br />

D<br />

= ε ∂σϕ<br />

2π<br />

2 ρστ<br />

= d xdtε<br />

∂ ϕ ∂ θ A<br />

∫<br />

σ<br />

ρ<br />

τ<br />

Goldstone & Wilzcek


<strong>Topological</strong> field theory <strong>and</strong> the family tree<br />

• <strong>Topological</strong> field theory of the QHE: (Thouless et al, Zhang, Hansson <strong>and</strong> Kivelson)<br />

• <strong>Topological</strong> field theory of the TI: (Qi, Hughes <strong>and</strong> Zhang, 2008)<br />

S<br />

S<br />

2<br />

= ∫ d k da(<br />

k)<br />

∫ d<br />

3<br />

= ∫ d k(<br />

a(<br />

k)<br />

∧ da(<br />

k)<br />

+ ..) ∫ d<br />

3<br />

x A(<br />

x)<br />

∧ dA(<br />

x)<br />

4<br />

x dA(<br />

x)<br />

∧ dA(<br />

x)<br />

• More extensive <strong>and</strong><br />

general classification<br />

soon followed (Kitaev,<br />

Ludwig et al)


Equivalence between the integral <strong>and</strong> the discrete<br />

topological invariants (Wang, Qi <strong>and</strong> Zhang, 0910.5954)<br />

LHS=QHZ definition of TI, RHS=FKM definition of TI


RKKY coupling of the surface states (Liu et al, PRL 2009)


Low frequency Faraday/Kerr rotation<br />

(Qi, Hughes <strong>and</strong> Zhang, PRB78, 195424, 2008)<br />

Adiabatic<br />

E g<br />

Requirement:<br />

(surface gap)<br />

normal contribution<br />

<strong>Topological</strong> contribution<br />

θ topo »3.6x10 -3 rad


Seeing the magnetic monopole thru the mirror of a TME<br />

insulator, (Qi et al, Science 323, 1184, 2009)<br />

q<br />

TME insulator<br />

(for μ=μ’, ε=ε’)<br />

similar to Witten’s dyon effect<br />

higher order<br />

feed back<br />

Magnitude of B:


Dynamic axions in topological magnetic insulators<br />

(Li et al, Nature Physics 2010)<br />

• Hubbard interactions leads to anti-ferromagnetic order<br />

• Effective action for dynamical axion


An electron-monopole dyon becomes an anyon!<br />

θ =<br />

2<br />

2α<br />

P3


Spin-plasmon collective mode (Raghu et al, 0909.2477)<br />

• General operator identity:<br />

• Density-spin coupling:


<strong>Topological</strong> Mott insulators<br />

• Dynamic generation of spin-orbit coupling can give rise to TMI (Raghu et al,<br />

PRL 2008).<br />

• Interplay between spin-orbit coupling <strong>and</strong> Mott physics in 5d transition metal Ir<br />

oxides, Nagaosa, SCZ et al PRL 2009, Balents et al, Franz et al.<br />

• <strong>Topological</strong> Kondo insulators (Coleman et al)


<strong>Topological</strong> insulators <strong>and</strong> superconductors<br />

Full pairing gap in the bulk, gapless Majorana edge <strong>and</strong> surface states<br />

Chiral Majorana fermions<br />

Chiral fermions<br />

massless Majorana fermions<br />

massless Dirac fermions<br />

Qi, Hughes, Raghu <strong>and</strong> Zhang, PRL, 2009


Probing He3B as a topological superfluid (Chung <strong>and</strong> Zhang,<br />

2009)<br />

The BCS-BdG model for He3B Model of the 3D TI by Zhang et al<br />

Surface Majorana state:<br />

Qi, Hughes, Raghu <strong>and</strong> Zhang, PRL, 2009<br />

Schnyder et al, PRB, 2008<br />

Kitaev<br />

Roy


General reading<br />

For a video introduction<br />

of topological insulators<br />

<strong>and</strong> superconductors,<br />

see<br />

http://www.youtube.<br />

com/watch?v=Qg8Yu<br />

-Ju3Vw

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!