Answers to problems

Examination in Fys3410 — Condensed Matter Phyics I

Day of examination: 31.03.2004
Examination hours: 09:00 – 12:00

Problem 1 Diffraction

A Bravais lattice is specified by the primitive lattice vectors:

√
1 3
~a1 = ax̂, ~a2 = ax̂ + aŷ, ~a3 = cẑ. (1)
2 2

Where x̂, ŷ, ẑ are the unit vectors along the x, y, and z axes respectively, and
a is the lattice distance. We assume that c > 2a. Notice that the vectors are
not all orthogonal.

1a) Make a sketch of the lattice in the xy-plane. Indicate the nearest
neighbors for a lattice point.

~a2

~a1

1b) Draw the Wigner-Seitz primitive cell for the lattice in the xy-plane.

(Continued on page 2.)

Examination in Fys3410, 31.03.2004 Page 2

1c) Find the primitive reciprocal lattice vectors ~bi corresponding to the
primitive lattice vectors ~ai .

~bi = 2π ~ai+1 × ~ai+2 .

Vc

Vc = ~a1 · (~a2 × ~a3 ).

Vc = ~a1 · (~a2 × ~a3 ) = ~a3 · (~a1 × ~a2 )

√ √
3 3 2
~a1 × ~a2 = ax̂ × aŷ = a ẑ
2 2

√
3 2
Vc = ac
2

2π 4π
=√ 2
Vc 3a c
√ √
~b1 = √4π 1 3 4π 1 3
( acx̂ × ẑ + acŷ × ẑ) = √ (− ŷ + x̂)
3a c 2
2 2 3a 2 2

~b2 = √4π (caẑ × x̂) = √4π ŷ

3a2 c 3a
√
~b3 = √4π (a2 3 x̂ × ŷ) = 2π ẑ
3a2 c 2 c

(Continued on page 3.)

Examination in Fys3410, 31.03.2004 Page 3

1d) Sketch the reciprocal lattice in the xy-plane.

~b2

~b1

We will study X-ray scattering from a crystal with a monatomic basis on

the given Bravais lattice. We will assume that the scattering is elastic. The
experimental geometry is illustrated in the following figure:

y
~k 0

~k θ
x

We assume that the sample is a single crystal and that the axes in the figure
correspond to the axes specified in Eq. 1.

1e) What are the possible scattering vectors ~k 0 when ~k = 2~b1 ?

~k 0 = ~k + G
~

and

(Continued on page 4.)

Examination in Fys3410, 31.03.2004 Page 4

|~k| = |~k 0 |

That is:

~k 0 = 2~b1 + G
~

From the following figure:

We see that all possible vectors ~k 0 of the same length as ~k, and on the
reciprocal lattice, are given as

~k 0 = ±2~b1 , ±2~b2 , ±2(~b1 + ~b2 )

We will then in the following assume that we study a powder of the crystal.
This corresponds to an average over all possible orientations of the crystal
with respect to the given coordinate systems. We ignore effects of finite
crystal and system sizes.

(Continued on page 5.)

Examination in Fys3410, 31.03.2004 Page 5

1f ) Explain Ewald’s construction for this geometry and show that for an
incoming wave of wavelength λ, the angles for intensity maxima for the
scattered wave are given as

λ G
sin(θ/2) = .
2π 2

Where G is the absolute value of a reciprocal lattice vector.

1g) Find the angles θ1 and θ2 corresponding to the smallest and second
smallest scattering angle θ. (θ1 > 0). What is the ratio of the scattered
intensity at these two angles, that is, what is I(θ1 )/I(θ0 )? (Hint: Since
~ in the xy plane.)
c > 2a it is only necessary to consider G

The smallest angles corresponds to the smallest possible G ~ vectors. The

~
smallest G vectors are illustrated in the following figure (blue arrows):

These smallest vectors are:

(Continued on page 6.)

Examination in Fys3410, 31.03.2004 Page 6

~ 1 = ±~b1 , ±~b2 , ±(~b1 + ~b2 )

G

The smallest angle is therefore given by

λ G1
sin(θ1 /2) = .
2π 2

Where G1 = b2 = √4π .
3a

λ
sin(θ1 /2) = √
3a

The second smallest G ~ vectors are illustrated in the figure with red arrows,
and are given as the following 6 vectors:

~ 2 = ±(2~b1 + ~b2 ), ±(2~b2 + ~b1 ), ±(~b2 − ~b2 )

G
√
And the second smallest angle is given by G2 = 3G1 (seen from the figure).
The second smallest angle is therefore given by:

λ √ λ
sin(θ2 /2) = √ 3=
3a a

The structure factor for each of these terms are 1. The intensity for the
powder-diffraction system will be proportional to the number of reciprocal
lattice vectors with the same length. These are both 6, therefore the relative
intensity will be 1.

We will now study the Bravais lattice with a two-atom basis. The atoms are
identical and are placed at positions ~r1 , and ~r2 , given as

1 1
~r1 = 0, ~r2 = ~a1 + ~a2 .
2 2

Where ~a1 and ~a2 are the primitive unit vectors for the lattice given above.

1h) Show that the structure factor SG~ for this basis can be written as

SG~ = f (1 + exp(i(πn1 + πn2 ))).

(Continued on page 7.)

Examination in Fys3410, 31.03.2004 Page 7

where f is the form factor of an atom, and n1 and n2 are integers describing
~
the reciprocal lattice vector G:

~ = n1~b1 + n2~b2 + n3~b3 .

G

The structure factor can be written as

X
SG~ = ~ · ~rj )
fj exp(−iG
j

For the two-atom basis given above, with atom form factors fj = f :

~ · ( 1 ~a1 + 1 ~a2 ))].

~ · ~0) + exp(−iG
SG~ = f [exp(−iG
2 2

We know that ~ai · ~bj = 2πδij , therefore:

X
~ · ( 1 ~a1 + 1 ~a2 ) = ( nj~bj ) · ( 1 ~a1 + 1 ~a2 ) = 1 (n1 + n2 )
G
2 2 j 2 2 2

SG~ = f (1 + exp(i(πn1 + πn2 ))).

~ How do you interpret this?

1i) SG~ = 0 for some values of G.

Problem 2 Lattice Vibrations

We will study lattice vibrations in a monatomic string with next-nearest

neighbor interactions. The system is illustrated in the following figure.

us−2 us−1 us us+1 us+2

c2
c1 m

The equilibrium position of particle s is x = as, where a is the lattice spacing.

For particle s with mass M at position xs , the deviation from the equilibrium
position is us = xs − as.

(Continued on page 8.)

Examination in Fys3410, 31.03.2004 Page 8

Each particle s interacts with its neighbors s − 1 and s + 1 through linear

springs with spring constants c1 . In addition, each particle s interacts with
its next-nearest neighbors s − 2 and s + 2 through linear springs with spring
constant c2 .

2a) Show that the net force Fs on particle s can be written as

Fs = M üs = c1 (us+1 + us−1 − 2us ) + c2 (us+2 + us−2 − 2us ).

2b) Assume that the solution is a planar wave:

us (t) = u0 exp i(ksa − ωt)

2c) Show that ω 2 (k) must satisfy:

c1 ka c2 2ka
ω2 = 4 sin2 + 4 sin2 .
M 2 M 2

The equations of motion give

−ω 2 M us = c1 (eika + e−ika − 2)us + c2 (ei2ka + e−i2ka − 2)us

which gives

c1 c2
ω2 = (2 − 2 cos(ka)) + (2 − 2 cos(2ka))
M M

We note that

ka ka ka
1 − cos(ka) = 1 − cos(2 ) = 1 − (cos2 ( ) − sin2 ( ))
2 2 2

where

ka ka
cos2 ( ) = 1 − sin2 ( )
2 2

which gives the wanted result.

(Continued on page 9.)

Examination in Fys3410, 31.03.2004 Page 9

2d) Find ω(k) when ka ¿ 1.

For ka ¿ 1, sin( ka
2
)' ka
2
, and similarly sin( 2ka
2
) ' ka.

Therefore

s s
c1 1 c2 c1 + 4c2
ω=2 (ka)2 + (ka)2 = ka
M4 M M

Assume that interactions between the m closest neighbors are included. That
is, particle s and s + r interact through a linear spring with force constant
cr for all r ≤ m on the crystal. We ignore the effects of a finite system size.

2e) Find an expression for ω 2 (k) for this case.

m
X cr rka
ω 2 (k) = 4 sin2 .
r=1 M 2