Physics 231 Homework 4, Question 4, Graphene
Physics 231 Homework 4, Question 4, Graphene
Physics 231 Homework 4, Question 4, Graphene
Figure 1: The solid lines indicate the crystal structure of graphene. The basis has two atoms, labeled
A and B. The Bravais lattice (consider, e.g, the lattice formed by
√ the A atoms shown by dashed lines)
◦
is triangular with a Bravais lattice spacing 2 × sin 60 × a = 3a, where a is the spacing between
neighboring atoms.
a. There are two atoms per unit cell so 1 band will be filled.
b. The Bravais lattice is the same as the lattice formed
√ by all the A atoms, say. As shown in Fig. 1
this is a triangular lattice with lattice spacing 3a.
c. From Fig. 1, we see that two basis vectors of the Bravais lattice are
√ √
a1 = 3 a (1, 0), a2 = a ( 3/2, 3/2). (1)
1
These basis vectors are of equal length and at 60◦ so the reciprocal lattice is a triangular lattice .
The first Brillouin zone is shown in Fig. 2. The distance OA to the center of the edge of the zone
1 2π 2 2π
is (1/2)b1 = . The distance OC to the corner of the zone is OA/ sin 60◦ = √ .
3 a 3 3 a
B
O C
Figure 2: The solid circles indicate points of the reciprocal lattice. The hexagon is the boundary of the
(first) Brillouin zone.
where R is a lattice point, and δ is the displacement to a neighboring lattice point. In other
words, an electron at site R can hop to any of the neighboring sites, i.e. if | R i denotes a state
with one electron which is at R we have
h R′ |H| R i = 0. (7)
We look for eigenvalues of the Hamiltonian. Consider a state with amplitude ψR for the electron
to be at site labeled by R. The eigenvalue equation is then
ψ R1 ψ R1
ψR ψR
2 2
H . = ǫ . . (8)
.. ..
ψ RN ψ RN
Considering one row of this equation we have
X
−t ψR+δ = ǫψR . (9)
δ
2
We look for amplitudes which vary like eik·r . There will be different amplitudes ψA and ψB on
sublattices A and B, so
ψA eik·r , (R ∈ A) ,
ψR = (10)
ψB eik·r , (R ∈ B) .
The result is plotted in the left part of Fig. 3. We see that the bands meet at the corner of the
zone (the largest value of kx ), point C in Fig. 2. Along the y direction, which is to the center of
the edge of the Brillouin zone, Eq. (12) can be simplified to
ǫ = ±t 5 + 4 cos 23 ky a . (15)
This is plotted in the right part of Fig. 3. At the zone edge the energies are ±1. This corresponds
to point A in Fig. 2.
The maximum and minimum energies are ±3, and occur at k = 0.
e. There are two bands one with positive energy and one with negative energy. They touch at the
corners of the Brillouin zone. Because there are two electrons per unit cell the lower band is
completely filled and the upper one completely empty. Hence the Fermi energy is at ǫF = 0.
We want to compute the dispersion relation in the vicinity of the zone corners where the energy
tends to zero. We write
k=K+q (16)
√
where K is the wavevector at the zone corner, K = (4π/(3 3a), 0) for example, and we will
assume that q is small.
3
Figure 3: The dispersion relation in graphene along two directions in the Brillouin zone. The left figure
is in the direction to the corner of the zone, and the right figure is in the direction to the center of the
edge of the zone.
where
F (kx , ky ) = eiky a + 2e−iky a/2 cos
√
3
2 kx a . (18)
To lowest order in q we have, after a little algebra,
3
F (kx , ky ) = 2 a (qx + iqy ) , (19)
so
H(k) = h̄v σ · q, (20)
where the speed v is given by
3 at
v= , (21)
2 h̄
and σx and σy are Pauli spin matrices. Hence the eigenvalues of the H(k) are
ǫ = ±h̄vq . (22)
Hence we have “massless” excitations with a linear dispersion. The speed v is, of course, much
less than the speed of light. The cone-like energy surface is shown in Fig. 5.
f. There are pockets of low energy excitations near the corner of the Brillouin zone, see Fig. 6
Hence there are are two pockets of low energy excitations, one for each of the two inequivalent
points on the Brillouin zone boundary. Let us compute the density of states in this region (i.e. en-
ergy close to zero). We consider unit area, and, for now, just one pocket and one spin state. As
4
Dispersion relation of graphene
ky
-2
0
2
Out[2]= 0
Ε
-2
-2
kx
Figure 4: The dispersion relation in graphene. There are two bands, one with positive energy, and the
other with negative energy, which touch at the zone corners.
Dispersion
4 Π relation of graphene near a zone corner
kx - 0.10
3 3
0.05
0.00
-0.05
-0.10
0.1
Out[11]=
Ε 0.0
-0.1
-0.10
-0.05
0.00
0.05
ky 0.10
Figure 5: The dispersion relation in graphene near a corner of the Brillouin zone showing the linear
dispersion.
5
C
D’’ C D’’
C’’ D C’’
D
D’ C’ D’ C’
Figure 6: The points C, C ′ , and C ′′ at the corners of the zone are equivalent because they are connected
by a reciprocal lattice vector. Similarly for points D, D′ , and D′′ . However, C is not equivalent to D,
so there are two distinct corners of the Brillouin zone. Consequently, in the repeated zone (right hand
figure) one sees that there are two separate pockets of low energy excitations.
direction) to be
h̄2
ρ(ǫ) = |ǫ| . (24)
2πv 2
Figure 7: The density of states of graphene computed numerically. Note the linear variation near
ǫ = 0. This comes from the linear dispersion near the corners of the Brillouin zone, as discussed in
the text. There are also Van Hove singularities. The discontinuity at the end points ǫ = ±3 comes
from a maximum or minimum of the dispersion relation in two dimensions. (These are at k = 0). The
logarithmic divergence at ǫ = ±1 comes from a saddle point in the dispersion relation at the middle of
the edges of the boundary of the Brillouin zone, point A in Fig. 2.
6
The density of states of graphene, computed numerically, is shown in Fig. 7. The linear density
of states near zero energy is clearly seen, as is the discontinuity at the top of the upper band and
bottom of the lower band (an example of a Van Hove singularity in two dimensions at a maximum
or minimum of the the dispersion relation). One can also see a logarithmic divergence in the
density of states, which is a Van Hove singularity coming from saddle points in the dispersion
relation at points A and B in the figure of the Brillouin zone, Fig. 2.
Hence graphene is on the border between a metal, which has a finite density of states at the Fermi
energy, and an insulator which has a band gap. It is perhaps best thought of as a semi-metal
(which is a metal with a small density of states at the Fermi energy). Alternatively, one could
perhaps view it from the other direction as a semiconductor with a vanishing bandgap.