Docs »

Lecture notes »
Lecture 3 : Chern Insulator

Lecture 3 : Chern Insulator

Menu
Quantum Hall Insulator
Half BHZ model
Chern number using h
Chern number as a Skyrmion number
Low-energy Hamiltonian
Chern number as an obstruction
Haldane model

Quantum Hall Insulator

The first Topological Insulator is shown in Integer quantum Hall effect.
For 2D electron gas
(2DEG), by applying a B field perpendicular to the
2D plane, we get plateaus of Hall
conductance
σxy = 1/ρxy which is quantized in ν e2 /h.
In the same time, ρxx = σxx = 0

(which may seems

counterintuitive, but noticing the tensor nature of ρ and
σ). The quantization
of σxy is very
accurate, it is the second best way to measure the fine structure
constant, and it
is inert to interactions and impurities. For a charge
neutral particle in 2D:

i∂t ψ(x, y) = H0 ψ(x, y)

2
^
p
with H0 =
2m
.

⃗  2
^−eA )
(p
In a magnetic field with a vector potential A, we have
H
⃗ 
=
2m
.

For a constant uniform magnetic field, choose Landau gauge:

Ax = 0, Ay = Bx , we have

2 2
ℏ ℏ
2 2
H = (−i∂x ) + (−i∂y − eBx)
2m 2m

H is invariant under translations along the y axis, so

[py , H ] = 0 , we can write the eigen states
of H :
ψ = f (x)exp(iky y) and H ψ = ϵψ . which gives:

ψn,k = ϕn (x − x0 )exp(iky y)
y

ℏky
ϕn (x) satisfies the harmonic oscillator potential
V (x) =
k

2
x
2
with
k 2 2
= e B /m, x0 =
eB
.
The eigen energies are the
famous harmonic one:
 −−
−
k eBℏ
ϵn,k = (n + 1/2)ℏωc = (n + 1/2)ℏ√ = (n + 1/2)
y
m m

Some comments

1. Landau levels are highly degenerated, they form energy bands with
n denotes band index
and ky the momentum, all the
states with the same n (but different ky ) have the
same
energy (flat bands).
2. The degeneracy of each Landau level is
D , where S denotes spin,
and Φ0
Φ
= (2S + 1)
Φ0

is the magnetic flux quanta with

Φ0 = h/e .
By
3. We can also use symmetry gauge:
Ax , though it is more
cumbersome,
Bx
= − , Ay =
2 2

the physics is quite insightful. More on symmetry gauge,

see[1].

Edge states with Quantization of the Hall conductivity

Electrons at the edge can move along the edge in one direction, this
edge states gives quantized
Hall conductivity.

dpdr dp
At top edge:
N = ∫
2πℏ
= L∫
2πℏ
, we get

N dp
n = = ∫
L 2πℏ

kF top
dp kF top − kF μ top − μ
δn top = ∫ = =
kF
2πℏ 2πℏ 2πℏvF top

∂E(k) μtop −μ
We have use the relation
vF = =
ℏ∂k ℏ(kF top −kF )

Net current for top edge:

2
μ top − μ eV /2 e V
Itop = evF top δ top = evF top = e =
2πℏvF top h h 2

We have used μ1 − μ 2 = e(V1 − V2 ) . And, similarly, we have

Ibottom = Itop , that gives

2
e
I = V
h

2
I V e
σ = j/E = / = I /V =
Ly Ly h

If one has n copies of edge states, then total Hall

conductivity is:
 2
e
σ = n
h

Half BHZ model

See the lattice below:

Consider at

each site, there are two

orbital modes coupled together, in the tight-binding approximation, we
can write the Hamiltonian:

†
U (n) † † † †
H = ∑ cn ,ny cn ,ny + c Tx (n)cn ,ny + c Ty (n)cn ,ny + h.c.
x x nx +1,ny x nx ,ny +1 x
2
nx ,ny

Assuming translational invariance along x and y, i.e.

U (n) = U , Tx (n) = Tx , Ty (n) = Ty ,
after Fourier transform:
cn
with
1 ikx nx +iky ny
,ny = ∑ ck ,ky e
x kx ,ky x
√N

N = Nx Ny , kx Nx = 2πm1 , ky Ny = 2πm2 , we get

†
H = ∑ c H (k)ck ,ky
kx ,ky x

kx ,ky
with

1
ikx iky
H (k) = U + Tx e + Ty e + h.c.
2

with

,
we get
1 iA 1 iA
U = Δσz , Tx = σz − σx , Ty = σz − σy
2 2 2 2

H (k) = (Δ + coskx + cosky )σz + A(sinkx σx + sinky σy )

Chern number using h

Written the Hamiltonian as: H (k) = h ⋅ σ , we get

Asinkx
⎛ ⎞

h(kx , ky ) = ⎜ Asinky ⎟

⎝ ⎠
Δ + coskx + cosky

and we get the eigen-energies:

−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−
2 2 2 2
E± (kx , ky ) = ±|h(kx , ky )| = ±√A (sin kx + sin ky ) + (Δ + coskx + cosky )
Set A = 1 , for some special parameter Δ, the
dispersion relation shows below:
From the dispersions also the analysis we can see that Dirac cone
appears at

kx , ky = 0, Δ = −2; kx , ky = 0, ±π, Δ = 0; kx , ky = ±π, ±π, Δ = 2 .

It means that the

monopole seats at h = 0, at each region, we

can calculate the Chern number of the valence
band in investigating how
many times does the torus formed by the image of the Brillouin zone
in
the space of h contail the origin. The Torus for different
Δ = −2.5, −1, 1, 2.5 shown below
(for clarity, only half of the
torus is shown):
From h, we get the Chern number as follow: (Set
A = 1 )

|Δ| > 2, Q = 0

−2 < Δ < 0, Q = −1

0 < Δ < 2, Q = 1

As long as the whole band is gapped, we have the same topology and the
same Chern number.
In the region |Δ| > 2 , reminding the
Hamiltonian (In what follows, we always consider A = 1 )

H (k) = A(sinkx σx + sinky σy ) + (Δ + coskx + cosky )σz

It represent Hamiltonians topologically equivalent to the atomic limit,

the limit in which all
hoppings are set to zero, as if the lattice
constant were infinity. The |Δ| > 2 region is
topologically
the same phase as the phase |Δ| → ∞ , which has trivial
(momentum-
independent) eigenstates and zero Chern number.

Chern number as a Skyrmion number

Looking at the vector h, the
(hx (kx , ky ), hy (kx , ky )) shows below

To understand the Chern number, we can see if the hz component

change sign or not in the BZ.
If it changes sign, it corresponds to a
skyrmion, and the Chern number is non-zero, it is the same
as Skyrmion
number. If it does not change sign, then the topology is trivial. The
value of hz (k)
at high-symmetry points in the BZ are shown
below:

From which we can get the Chern number.

Low-energy Hamiltonian
We can also calculate the Chern number using the low-energy Hamiltonian.
At Δ = −2 , the
energy gap collapses at the
Γ = (0, 0) point, near this point, we have
 HΓ+k = kx σx + ky σy + (Δ + 2)σz

For the Hamiltonian H (k) = kx σx + ky σy + mσz , we

can get the monopole field for
E− state is

1
^ ^ ^
Ωij = h ⋅ (∂i h × ∂j h)
2

Then we get

m
Ωij =
2 2 2 3/2
2(kx + ky + m )

Which gives:

1 sign(m)
C = − ∫ Ωij dkx dky = −
2π 2

(kx ,ky ,m)

Alternatively, from
B ,
1
=
2 2 2 2 3/2
(kx +ky +m )

1
C = − ∫ B ⋅ dS
2π

For m > 0 and m < 0 , the integral region and direction

shows below:

Bfield

Note here the Chern number is not an integer, because we used continuous
model, and kx , ky
−−−−−−−−−−−
can range from −∞ to
∞, so the band structure ±√k2x 2
+ ky + m
2

run riot and never turns
back, which should give us another half of the
Chern number (though we have no idea is we
should add or subtract
half-values). But as we can see below, the differences of Chern number
between two states gives the right answer.

So from Δ ,
< −2 → Δ > −2 m < 0 → m > 0 , we get

1 1
C = − sign(m > 0) + sign(m < 0) = −1
2 2

Also we know Δ ,
< −2 C = 0 , so for −2 ,

< Δ < 0 C = −1 .

Chern number as an obstruction

Write
h = |h|(sinθcosϕ, sinθsinϕ, cosθ) .
In principle, for the whole Bloch sphere, we must
at least use two
different bases which are well defined. If for the whole band, we have
hz < 0
or hz > 0A, then, we can use one set of eigen base,
which gives us zero Chern number. But for
the region
−2 < Δ < 0, at the point Γ, we have
hz (0, 0) = Δ + 2 > 0, hz /h = 1 = cosθ,

θ = 0 ; at the point (0, ±π) or

(±π, ±π), we have hz ,
< 0 θ = π ,
means we must use two
set of eigen bases, denotes no-zero Chern number.
The resulting h
^
= h/|h| for
different
Δ = −2.5 → 2.5 shown below:
Haldane Model
Haldane model describe the model of Graphene with real
nearest-neighbor-hopping parameters
but complex
next-nearest-neighbor-hopping parameters resulting from a nonzero
magnetic
field. But the field is periodic and have zero net flux per
plaquette. The real space and reciprocal
space shows below (with unit
length 1):

lattice
lattice

The vectors are:

– –
√3 /2 −√3 /2 0
a1 = ( ) , a2 = ( ) , a3 = ( ) = −(a1 + a2 )
1/2 1/2 −1

– – –
−√3 /2 −√3 /2 −√3
b1 = a 2 − a 3 = ( ) , b2 = a 3 − a 1 = ( ) , b3 = a 1 − a 2 = ( )
3/2 −3/2 0

Set b1 , b2 as two base vectors, then the reciprocal

lattice spanned by b∗1 , b∗2 with

= 2πδ ij , that is
∗
bi ⋅ b
j

– –
−1/√3 −1/√3
∗ ∗
b = 2π ( ),b = 2π ( )
1 2
1/3 −1/3

–
−4√3 π/9
Define
K .
1 ∗ ∗ ′
= (b + b ) = ( ),K = −K
3 1 2
0

We get the Hamiltonian:

† † † † †
H = M ∑ (a aiA − a aiB ) + t ∑ (a ajB + h.c.) + t2 ∑ (a ajA + a ajB + h.c.)
iA iB iA iA iB

i <i,j> <<i,j>>

M breaks inversion symmetry of A and B

sublattice, and t2 is the next-nearest-neighbor
coupling, for a
staggered magnetic field shows below, it changes

e e e h

.
i ∫ A⋅dl i χΦ 0 i χ iϕ
tij → tij e ℏ j→i
= tij e ℏ = tij e ℏ e
= tij e

mag
Use
aiA =
1
∑
k
akA e
ik⋅R iA
, aiB =
1
∑
k
akB e
ik⋅R iB
,
N is the number of unit cell.
√N √N

Then we can get:

† † akA
H = ∑(a a ) H (k) ( )
kA kB
akB
k

We arrived at

H (k) = h0 σ0 + h ⋅ σ

with

h0 = 2t2 cosϕ ∑ cosk ⋅ bi

hx = t ∑ cosk ⋅ a i

hy = −t ∑ sink ⋅ a i

hz = M + 2t2 sinϕ ∑ sink ⋅ bi

Because of the C3 symmetry, the Dirac cone can only happen at

K , K ′ . At the vicinity of K ,
we have

3 –
H (K + q) = −3t2 cosϕ + t1 (q x σx − q y σy ) + (M + 3√3 t2 sinϕ)σz
2

At the vicinity of K ′ , we have

3 –
′
H (K + q) = −3t2 cosϕ − t1 (q x σx − q y σy ) + (M − 3√3 t2 sinϕ)σz
2

where −π < ϕ < π .

Start from M → ±∞, we have zero Chern number, assuming

t2 > 0, ϕ > 0, from the
–
previous results, when M from
+∞ decreases to M = 3√3t2 sinϕ, K ′
band close and
reopen, the Chern number changes to

1 – 1 –
C = − sign((M − 3√3 t2 sinϕ) < 0) + sign((M − 3√3 t2 sinϕ) < 0) = 1
2 2

– –
Then when M > −3√3 t2 sinϕ to
M < −3√3 t2 sinϕ , we have

1 – 1 –
C − 1 = sign((M + 3√3 t2 sinϕ) < 0) − sign((M + 3√3 t2 sinϕ) < 0) ⇒ C = 0
2 2
We can do the same calculation for ϕ < 0 , we get the phase
chart below:

phase chart

Because kx , ky have no dependence on M , t2 , ϕ,

hz in the range of

– – –
(M − 3√3 t2 sinϕ → M + 3√3 t2 sinϕ) , we can draw a
spheroid Σ′ with height 6√3sinϕ
and center
(0, 0, M /t2 ). We can get the Chern number using following
pictures:
spheroid

Also, we can get the Chern number using the whole 3-d picture (set

ϕ = π/2, M /t2 = 6, 0, −6 ):