0% found this document useful (0 votes)
9 views4 pages

CM1 HW01

Uploaded by

noygalron30
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
9 views4 pages

CM1 HW01

Uploaded by

noygalron30
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Problem Set 1

Quantum Theory of Matter - 106035 - Spring 2024


Submit date - June 16

1 Tight Binding in Graphene (15 points)


In the tutorial, we’ve seen the tight-binding band structure of graphene (the honeycomb lattice), for the
case where all atoms are similar (mono-atomic). With on-site energy ε0 = 0, we got the dispersion
p
ε± = ±t 3 + 2 cos(k · a1 ) + 2 cos(k · a2 ) + 2 cos(k · (a1 + a2 )) ,

as plotted in figure 1. Note the points K, K ′ lie on the edge of the Brillouin zone where the spectrum
is gapless. These points are called the Dirac points and they’re given by
 
4π ′ 2π 1
K= (1, 0) K =√ 1, √
3a 3a 3
Now consider the same lattice with two different atoms in the unit cell (diatomic) with different on-site
energies εA , εB . Use the solution we’ve seen in the tutorial (you’ll have to tune it a bit) to argue that a
gap opens at the Dirac points when εA ̸= εB . Find the bandgap ∆ in terms of εA , εB , t. You don’t have
to solve everything from scratch, just use the tutorial to think how the tight binding matrix looks in the
vicinity of the points K, K ′ and diagonalize it.

2
3

1 1
ε/t

0
ε/t

−1 0

−2
K −3
−1
0
K
3π/2
π
−2
π/2
−3π/2 0
−π
a

−π/2
ky

−π/2
0
−π
kx a π/2
π −3π/2
3π/2

Figure 1: The honeycomb lattice dispersion

1
2 Tight Binding on the Kagome Lattice (35 points)
Consider the Kagome lattice, shown in figure 2a.

(a) The Kagome Lattice (b) The Kagome ”Star of David”

Figure 2

1. (5 points) Write down primitive vectors and basis vectors for the lattice. Draw the unit cell that
corresponds to the vectors you’ve chosen. Can the Kagome be split into sub-lattices of the same
type? If yes, state what lattice and how many sub-lattices. If not, explain why.
2. (5 points) Calculate the reciprocal lattice vectors and draw the first Brillouin zone.
3. (10 points) Calculate the dispersion bands using tight binding approximation. You can assume
that all tunneling amplitudes are of magnitude t. Work in nearest neighbors approximation. For
your convenience, you may use any tool you would like, including Mathematica, Python, etc., or
diagonalize it analytically for good sportsmanship.
4. (10 points) Plot the bands as a surface plot within the first Brillouin zone. Use a programming
language of your choice.
5. (5 points) One of the bands you’ve found should be flat, that is, of constant energy. Consider an
electron confined to this flat band. Will it contribute to the current? Why? Find a wavefunction
of such an electron and use it to argue for your statement about the current.
Hint: To find the wavefunction, consider a single Kagome ”star of David” as in figure 2b. Try to
build from the Wannier states (|n, m, α⟩, where α is for the sublattice) a normalized wavefunction
that is zero everywhere apart from the sites that lie on the inner hexagon.

2
3 Peierls Distortion (30 points)
Consider a monoatomic one-dimensional chain with lattice constant a and possible distortion of atoms
rearrange in pairs, as can be seen in figure 3. There are N atoms in the chain (N ≫ 1) and the lattice
constant is a.

t−
t+

Figure 3: One dimensional Peierls chain

In the tight-binding model, the hopping term is a function of the distance between two lattice sites, t(x),
the shorter the bonds the larger the hopping term (magnitude). It means that we can expand

t± = t(a(1 ± δ)) ≃ t(a) ± t′ (a)aδ


t′ (a)a
 
= t(a) 1 ± δ
t(a)
≡ t(1 ± αδ) ,

where α > 0 and δ is dimenssionless and small. The onsite energy, ε0 , is the same for each atom, and
thus can be gauged out. Consider an onsite average filling of 1 electron (with spin degree of freedom).
1. (5 points) Use the tight-binding model to calculate the electron dispersion without the distortion,
ε0 (k).
2. (10 points) Now consider the dimerized set-up (finite δt ), and again, use the tight-binding model
to find the dispersion, εd (k). Is it a metal or an insulaotr? What is the energy gap?
3. (5 points) Find the difference in the electron’s contribution to the ground state energy between the
dimerized to the non dimerized set-ups,
Na
Z  
∆Ee = dk εd (k) − ε0 (k) .
π |k|≤kF

In your answer, it will be helpful to use the complete elliptic integral of the second kind E,
Z π/2 p
E(x) = dθ 1 − x2 sin2 θ ,
0

and can be expanded as


 
1
E(1 − x) = 1 + x c1 − log(x) + O x2

4
4 log(2) − 1
c1 = ≈ 0.443 .
4

4. (5 points) Consider now that nearest neighbors atoms are connected with a spring with an elastic
coefficient K(> 0), as can be seen in figure 4. Find the (classically) energy cost of the deformation
a → a(1 ± δ) of the entire lattice, ∆Eel .

Figure 4: One-dimensional chain with springs

3
5. (5 points) Write the expression for the total energy difference between the distorted and undistorted
set-ups, ∆E(δ) = ∆Ee (δ) + ∆Eel (δ). Show that the minimal value of ∆E with respect to δ is
negative and deduce that the dimerized set-up is energetically favorable over the nondimerized
set-up. It means that in a one-dimensional crystal, lattice distortions destroy the
conductivity.

4 Van-Hove singularities (20 points)


As we’ve seen in the tutorial, extrema in the electronic dispersion lead to a singular behavior in the
density of states g(ε). In class we’ve dealt with the 2D dispersion:

ε(k) = −2t (cos(kx a) + cos(ky a))

We saw that the saddle points with energy ε = 0 at k = (π/a, 0) and k = (0, π/a) contribute a log
divergence to the density of states:  
1 t
g(ε) ∼ 2 2 log
π ta |ε|
1. (10 points) At half-filling (one electron per site) and T = 0, the Fermi energy is εF = 0. What
is the chemical potential as a function of temperature µ(T )? Show that the specific heat at low
temperatures behaves as
t
Cv = αT log + γT
T
Find α, γ. You can leave your answer in an integral form.
Hint: To find the chemical potential,
R∞ you can use the fact that working in the canonical ensemble,
the number of electrons N = −∞ g(ε)fF D (ε) is constant.
2. (10 points) We now turn to a 3D dispersion of the form

ε(k) = A(kx2 + ky2 − kz2 )

Note that this dispersion has a saddle point at k = 0, which will cause a non-analyticity at ε = 0 in
the density of states. What is the form of this non-analyticity? That is, is g(ε) finite? Continuous?
Differentiable? Explain yourself.

You might also like