
Home
Contact Info
Course Info
Calendar
Homework
Lecture Notes




PHY 521: Condensed Matter Physics I
Prof. S. Teitel stte@pas.rochester.edu  Fall 2012
Problem Set 5
Due Tuesday, November 23, in lecture
 Problem 1
(Ashcroft & Mermin problem 9.5, but a little easier!)
Consider a twodimensional square lattice with lattice constant a.
a) Write down, in units of 2π/a, the radius of a circle that can accommodate m free electrons per primitive cell. Construct a table listing which of the first five Brillouin Zones of the square lattice (see A&M Figs. 9.7 and 9.15a) are completely full, which are partially empty, and which are completely empty for m = 1, 2, . . . , 7. Verify that if m ≤ 7, the occupied levels lie entirely within the first five Brillouin Zones, and that when m ≥ 8, levels in the sixth zone become occupied.
b) Draw pictures of all branches of the Fermi surface, translated back into the first Brillouin Zone, for the cases m = 1, 2, . . . , 5 (note, the example shown in A&M Fig. 9.15b is not in the first Brillouin Zone) and shade the regions corresponding to the occupied states.
 Problem 2
Consider a one dimensional Bravais lattice of lattice constant a . Let x label the position along this lattice, as shown below. Assume there is only a single ion at each BL site. If each ion contributes exactly one conduction electron, we learned in class that we expect the lowest band will be half filled and the system will behave like a metal. However, in the 1930's, Rudolf Peierls showed how such a situation can result in an insulator! This effect is now known as the Peierls instability. The purpose of this problem is to have you figure out the mechanism for this effect.
a) What is the primitive vector for this Bravais lattice? What range of x is the WignerSeitz cell? What are the vectors of the reciprocal lattice? What range of k is the 1st Brillouin Zone?
b) The ions of the lattice give rise to a periodic potential U(x). Assuming the effect of this potential can be treated within the weak potential approximation. Sketch the band structure in the reduce zone scheme, showing the lowest two bands. If each ion of the lattice contributes m conduction electrons, for which values of m will the system be a metal? an insulator? Explain your reasoning.
c) Imagine that the ions of the lattice undergo a distortion as shown below (every other ion moves ot the left an amount δx). How would you now describe the below crystal structure? What is the primitive vector? What range of x is the WignerSeitz cell? What are the vectors of the reciprocal lattice? What range of k is the 1st Brillouin Zone?
d) Sketch the band structure in the reduced zone scheme for this deformed lattice. Show enough bands that your sketch shows the same number of electron states as represented in your sketch of part (b) for the undistorted lattice. If each ion contributes exactly one conduction electron, will the deformed lattice be a metal or an insulator?
e) Assume that the total ion potential U(x) can be written as a sum over the individual potentials from each ion, i.e. U(x) = ∑_{i} V(x  x_{i}), where the x_{i} give the positions of the ions, and V(x) is the potential of a single ion centered at x = 0. If one writes U(x) = U_{o}(x) + δU(x), where U_{o}(x) is the potential of the undistorded lattice and δU(x) is the correction due to the distortion, what is the period of U_{o}(x)? of δU(x)? Write an explicit expression for δU(x) in terms of V(x), assuming the distortion δx is small. How does δU(x) vary with δx?
f) Give a crude estimate of how the total electronic energy density (energy per length) changes due to the lattice distortion. Does it increase or decrease? (Assume one conduction electron per ion as in part (d).) Hint: you should do this by making a physically reasonable estimate, rather than an exact analytical calculation (although the exact calculation can be done and gives an important correction to the naive estimate).
g) Assume we can model the elastic energy on the ionion interaction as a set of springs connecting nearest neighbor ions, i.e. the elastic energy of a pair of ions is (1/2)C(x_{i+1}  x_{i}  a)^{2}, where C is the spring constant. What is the elastic energy density (energy per length) due to the distortion? Assume that the total energy of the sytem is the sum of the elastic and electronic energies. Does the lattice prefer to make the distortion or not?
 Problem 3
Consider a two dimensional square Bravais lattice with lattice constant a. Suppose the periodic ionic potential is given by,
U(x, y) = 4Ucos(2πx/a)cos(2πy/a)
a) Find the Fourier components U_{K} for reciprocal lattice vectors K.
b) For k = (k, 0) along the k_{x} axis, how does the single electron energy ε(k) behave at the boundary of the first Brillouin Zone, (π/a, 0)?
c) For k = (k, k) along the diagonal of the first Brillouin Zone, how does ε(k) behave at the corner of the first Brillouin Zone, (π/a, π/a)?
 Problem 4 (Ashcroft and Mermin problem 9.3)
Consider the point W, k_{W} = (2π/a)(1, 1/2, 0), in the first Brillouin Zone of an fcc structure (see A&M Fig. 9.14). Here three Bragg planes meet. These Bragg planes bisect the reciprocal lattice vectors K_{1} = (2π/a)(1,1,1), K_{2} = (2π/a)(1,1,1), and K_{3} = (2π/a)(2,0,0). Hence the free electron states at wavevectors k_{W}, k_{W} − K_{1}, k_{W} − K_{2}, and k_{W} − K_{3} are degenerate with energies ε_{W}^{0} = (hbar k_{W})^{2}/2m.
a) Consider the effect of a weak ion potential in splitting these degeneracies. Show that the new energies, to lowest order in the ion potential, are given by the four eigenvalues ε of the matrix:
ε_{W}^{0}  U_{1}  U_{1}  U_{2} 
U_{1}  ε_{W}^{0}  U_{2}  U_{1} 
U_{1}  U_{2}  ε_{W}^{0}  U_{1} 
U_{2}  U_{1}  U_{1}  ε_{W}^{0} 
where U_{1} is the Fourier component of the ionic potential at K_{1} and at K_{2} (they are equal by cubic symmetry) and U_{2} is the Fourier component of the ionic potential at K_{3}. Show that these eigenvalues are:
ε= ε^{0}−U_{2} (twice), and ε=ε^{0}+U_{2}±2U_{1}
b) Using a similar method, show that the energies at the point U (see A&M Fig. 9.14), k_{U} = (2π/a)(1, 1/4, 1/4), are
ε=ε_{U}^{0}−U_{2}, ε=ε_{U}^{0}+(1/2)U_{2}±(1/2)(U_{2}^{2}+8U_{1}^{2})^{1/2},
where ε_{U}^{0}=(hbar k_{U})^{2}/2m.

