Introduction

Unlike static lattice model , which deals with average positions of atoms in a crystal,
lattice dynamics extends the concept of crystal lattice to an array of atoms with finite
masses that are capable of motion. This motion is not random but is a superposition of
vibrations of atoms around their equilibrium sites due to the interaction with neighbor
atoms. A collective vibration of atoms in the crystal forms a wave of allowed
wavelength and amplitude.

Just as light is a wave motion that is considered as composed of particles called

photons, we can think of the normal modes of vibration in a solid as being particle-
like. Quantum of lattice vibration is called the phonon.

The problem of lattice dynamics (LD) is to find the normal modes of vibration of a
crystal. In other words, LD seeks to calculate the energies (or frequencies ) of the
phonons as a function of their wave vector's k . The relationship between and k is
called phonon dispersion .

LD offers two different ways of finding the dispersion relation:

1. Quantum-mechanical approach
2. Semiclassical treatment of lattice vibrations

Normal modes of crystal vibrations

There are two possible modes of vibrations of atoms in the crystal:

 longitudinal and
 transverse

In case of longitudinal mode the displacement of atoms from their equilibrium

position coincides with the propagation direction of the wave, whereas for transverse
mode, atoms move perpendicular to the propagation of the wave.

For one atom per unit cell the phonon dispersion curves are represented only by
acoustical branches. However, if we have more than one atom in the unit cell optical
branches will appear additionally. The difference between acoustical and optical
branches arises because of the more options of vibrations for atoms in the unit cell.
For example, atoms A and B of diatomic cell can move together in phase (acoustical
branch) or out of phase (optical branch). Click here to see how it looks like

Generally, for N atoms per unit cell there will be 3 acoustical branches (1 longitudinal
and 2 transverse) and 3N-3 optical branches (N-1 longitudinal and 2N-2 transverse)

Longitudinal Acoustical Mode:

Transverse Acoustical Mode:

Acoustical phonons
Lets consider a linear chain of identical atoms of mass M spaced at a distance , the
lattice constant, connected by invisible Hook's law springs. For simplicity we will
consider longitudinal deformations - that is, displacements of atoms are parallel to
the chain.

Let Un =displacement of atom n from its equilibrium position

Un-1=displacement of atom n-1 from its equilibrium position
Un+1=displacement of atom n+1 from its equilibrium position

The force on atom n will be given by its displacement and the displacement of its
nearest neighbors :

The equation of motion is:

where is a spring constant.

The above equation is not obviously a wave equation, but let us assume a traveling
wave solution, namely,

Let Uno = Uo since if it is a wave, it has to have a definite amplitude.

If we substitute our wave solution into equation of motion we find a phonon's

dispersion relation for linear monatomic chain as follows:

The dispersion curve is shown below.

One important feature of the dispersion curve is the periodicity of the function. For
unit cell length , the repeat period is , which is equal to the unit cell length in
the reciprocal lattice. Therefore the useful information is contained in the waves with
wave vectors lying between the limits

This range of wave vectors is called the first Brillouin zone. At the Brillouin zone
boundaries the nearest atoms of the chain vibrate in the opposite directions and the
wave becomes a standing wave .
 Transverse acoustical standing wave

As k approaches zero (the long-wavelength limit) and we have

where is a phase velocity, which is equivalent to the velocity of a sound in the

crystal.
Phonons with frequency which goes to zero in the limit of small k are known as
acoustical phonons.
"Long" wavelength acoustical vibrations

Optical phonons

Consider a lattice with two kinds of atoms - that is lattice with a basis of two atoms in
the primitive cell. Now we have to write two solutions for the displacement
corresponding to the two masses m and M. The equations of motion are:

and we assume solutions:

where A and B are the amplitudes of vibration of atom m and M respectively

The diatomic case has two solutions of the dispersion relation:

These solutions are plotted in the figure below.

The allowed frequencies of propagation wave are split into an upper branch known as
the optical branch, and a lower branch called the acoustical branch. There is a band of
frequencies between the two branches that cannot propagate. The width of this
forbidden band depends on the difference of the masses. If the two masses are equal,
the two branches join (become degenerate) at . Note that the first Brillouin zone
goes from to just as in the monatomic case if we use the lattice
constant instead of the interatomic spacing . The acoustical branch is qualitatively
similar to the dispersion relation for a monatomic lattice, but the optical branch
represents a completely different form of wave motion.

What is the difference physically between the acoustical and optical branches? It can
be found that for optical branch (in the long wavelength limit) the two atoms in the
unit cell move opposite to each other and the light mass amplitude is greater. For
acoustical branch (in the long wavelength limit) the displacement of both atoms has
the same amplitude, direction and phase. Click here to see the difference between
acoustical and optical modes .

Wave vector of crystal vibrations

The wave vector is a vector which points in the direction of propagation of the wave.
It has the units of cm-1 and defines the vector in reciprocal space (HERE SHOULD
BE THE LINK TO VLAD'S PROJECT). In contrast to the electromagnetic waves the
wave vector, k, of lattice vibrations has discrete spectrum within defined limits.

Lets consider the traveling wave:

where is a wave vector.

The possible values of k depend on the boundary conditions. For periodic boundary
conditions the only requirement is that U(x)=U(x+L), where L is the length of the
chain which gives the values of wave vector:

where n is an integer. Within the first Brillouin zone the upper limit of n is N/2, where
N is a number of atoms in the monatomic chain.

Now we can see that dispersion relation is not strictly speaking a continuous curve but
rather a series of closely spaced points representing the possible modes of vibrations.
Figure below is the dispersion relation for diatomic linear lattice.

The spacing between the dots is equal to

where N is the number of atoms in

the lattice.