SLIDE # 1

Chapter 2. Principles of quantum mechanics.

Quantum mechanics is needed to understand the basic properties of solids such as energy
bands.

In this chapter, we will cover the basics of quantum mechanics.

Starting with the birth of quantum mechanics, the Schrödinger equation will be introduced as
the governing equation, followed by its applications to closed and open systems. Some real-
world examples will be demonstrated at the end.
 SLIDE # 2

Birth of Quantum Mechanics.

Black body radiation, photoelectric effect, matter wave, hydrogen atom spectrum and
uncertainty principle will be discussed briefly.
 SLIDE # 3

About 100 years ago, scientists believed that the Newton’s mechanics was perfect. But there
appeared some experimental evidences that, maybe, that is not the case. One of the most
popular experiments that the classical mechanics failed to explain and eventually led to the
birth of the quantum mechanics is black body radiation.

When an object or a body is at thermal equilibrium it emits light of certain frequency. The
graph in the right shows the intensity as a function of wavelength of the black body radiation.
As shown in the graph, the classical theory could not explain the experiments, and only after
Planck introduced an assumption that light consists of energy quanta called photons, the
experimental behavior of black body radiation could be explained.
 SLIDE # 4

Another experiment that supported the particle nature of light is the photoelectric effect
experiment.

In the apparatus shown in the left, when light is shone on a metal plate, electrons are emitted
and collected by another metal plate. What was found is that whether the plates emit
electrons or not depends on the wavelength of the light and that the kinetic energy of
emitted electrons depends linearly on the frequency of the light, but not on the intensity of
light.

These were explained only if light behaves like particle having discrete energy quantum.
 SLIDE # 5

So, light seems to have a dual nature, it was long thought be wave but it also behaves like
particle.

Then, how about electrons? Could they be wavelike?

If we apply the Einstein’s relativity theory to photon, we have energy E equal to momentum p
times speed of light c. By equating it with E equal to h nu, we have p equal to h over lambda.

de Broglie proposed that elementary particles like electron should also obey the momentum
wavelength relationship. It was the break through that connects the particle picture with the
wave piecture.
 SLIDE # 6

The figure shows the electron diffraction experiment by Thompson. The result indicated that
an electron beam having the momentum corresponding to de Brogile wavelength comparable
to X-ray show the same diffraction pattern expected for that of X-ray with the same
wavelength.

That is, electrons behave like wave.

The quantum theory of electrons shed light on atomic properties. For example, the hydrogen
spectrum.

Discrete spectral lines were observed in the emission absorption spectra of hydrogen gas,
making scientists puzzled and look for its origin.
 SLIDE # 8

To explain discrete spectra of hydrogen atom, Niels Bohr made a bold proposal that a
stationary orbit is formed for electron if the angular momentum is quantized. The Bohr’s
condition for stationary orbit corresponds to the condition to form standing wave in a circle
as illustrated in the figure.
 SLIDE # 9

Bohr’s theory was successful to explain the discrete energy levels associated with each
stationary orbit and thus discrete emission absorption spectra in a quantitative way, although
it was based on bold postulates that were not completely understood or accepted at that
time.

Equations show the steps to find the energy of an electron orbiting around a hydrogen atom.
Notice that, in the final equation, the energy is discrete with integer n.
 SLIDE # 10

Central to the quantum mechanics lies the uncertainty principle.

In classical wave theory, the dispersion in the position space and the dispersion in the Fourier
transformed space or k space have the reciprocal relationship. That is, delta x delta k is 2 pi.

Quantum mechanically, electron waves have similar properties, which is called the uncertainty
principle.

The uncertainty principle is the most fundamental principle of quantum mechanics, associated
with the concept of wave function and its probabilistic interpretation. One may say that
everything in quantum mechanics breaks down if the uncertainty principle is found to be
invalid.
 SLIDE # 11

There are two forms of the uncertainty principle.

The first one, delta x delta p is greater than h bar, means that, if one wants to measure the position of an
electron and its momentum simultaneously, one cannot measure both accurately. That is, if one wants to
measure the position more accurately, he/she has to lose the accuracy of the momentum.

The second one relates the uncertainty in energy and time. What the above uncertainty principle implies is
that, if one wants to measure an energy level of a particle more accurately, the particle has to be stay at the
level for longer time.
 SLIDE # 12

Next, we will introduce the Schrödinger equation and discuss the meaning of the wave function.
 SLIDE # 13

An intuitive derivation of the Schrödinger equation is shown here.

We start with the Helmhotz equation from basic electromagnetics. In the Helmholtz equation,
k is the wave vector which is defined by 2 pi over lambda where lambda is the wave length.
Now, using the de Broglie relationship, we substitute k with p over h bar and divide both the
sides by 2 m.
 SLIDE # 14

Then, since p squared over 2 m is the electron's kinetic energy and the total energy is the sum of kinetic and
potential energies, the right hand side can be written as E minus V as shown in the equation.

After rearranging the terms, we have the final equation, which is called the time-independent Schrödinger
equation.

It should be noted that this is not the derivation of the Schrodinger equation. In fact, there is no way to
derive it from the first principle. Rather it has to be postulated, like the classical Newton's laws are postulated.
 SLIDE # 15

Another intuitive derviation of the Schrodinger equation is as follows:

As we all know, the total energy is written as the sum of the kinetic and potential energies.

In quantum mechanics, physical observables like position, momentum, and energy become so
called operators as shown. You need to know the basics of the quantum mechanics to
understand this. For now, you may take this as granted.

Then, substituting energy, momentum, and position with their corresponding quantum
mechanical operators and introducing the wave function psi x t, we can write the time-
dependent Schrodinger equation as shown.
 SLIDE # 16

Here the Schrodinger's wave equation and the classical wave equation are compared. Note
the similarity and differences.
 SLIDE # 17

From the time-dependent Schrodinger equation, the time independent Schrodinger equation
is derived if the potential V is only a function of position.

If V x t is V x, we can try so called the separation of variables technique where the wave
function is written as a product of the spatial part psi x and the temporal part phi t.

Then the spatial part and the temporal part can be separated as shown, and we introduce a
constant E which is independent of position and time.

Then the temporal part has a simple solution given by phi t equal to exponential minus i E t
over h bar.

The spatial part is expressed as shown, which is the time-independent Schrodinger equation.

The full wave function psi x t is then psi x times phi t where phi t is always given as above.
 SLIDE # 18

What is the wave function? We may get a hint about the meaning of the wave function from
an analogy to electromagnetic waves.

The figure shows a standing electromagnetic wave with the electric field as the wave
amplitude. In the wave picture, the intensity is proportional to the square of the amplitude.
On the other hand, in the particle picture, the intensity should be proportional to the
number of photons in an interval or the numer density.

So, we deduce that the number density of photons should be proportional to the square
amplitude of the wave amplitude.
 SLIDE # 19

Another example that connects the wave amplitude and the number density of photons is the
double slit experiment shown in the figure.

In the double slit experiment, the diffraction pattern appears on the screen and it can be well
explained by the interference of waves that pass through the two slits.

In the particle picture, photons that hit and interact with the photographic film appear as
white dots on the film. As the total number of photons emitted from the source is increased,
the profile of the white dots or the detected photons becomes closer to the diffraction
pattern on the screen.

Therefore, again, we deduce that the number density of photons should be proportional to
the square amplitude of the wave amplitude.
 SLIDE # 20

Our interpretation of the wave function in the Schrodinger equation is that the square of the
wave function corresponds to the probability of finding an electron.

This is similar to the case of photons, as we have seen in the previous slides, where the
probability of finding a photon at any point in space is proportional to the square of the
electric field wave.

In classical mechanics, the position of a particle can be determined precisely, whereas in

From the definition of the probability, the integration of the square of the wave function over
the entire space should be equal to one.
 SLIDE # 21

In addition, the wave function has to meet the two boundary conditions.

The wave function must be finite, single-valued, and continuous.

Its derivative must be also finite, single-valued, and continuous.

In special cases where the potential function, V, becomes infinite in particular regions of space,
the first derivative will not necessarily be continuous, but the remaining boundary conditions
will still hold.
 SLIDE # 22

We will apply the Schrödinger equation to the following four cases: electron moving in free
space, electron confined in infinite barriers, electron transmission over a step barrier, and
electron tunneling through a barrier.
 SLIDE # 23

Firstly, electron moving in free space.

Free space means an empty space. If there is an electron in an empty space, it will move at a certain velocity,
without feeling any potential. That is, the electron has the kinetic energy only. So we set the potential term V
in the Schrödinger equation equal to zero. Then, we have equation (1) and with the introduction of k, which
has the dimension of wave vector, we have Equation (2). A general solution of Equation (2) is written in
Equation (3). You can easily check that the wave function psi of Equation (3) is indeed a solution of Equation
(2) by putting it back to Equation (2).

Now the total wave function including the time part is written as shown in Equation (4). The first term has
the argument of k x minus omega t, which represents a plane wave propagating to the right, while the
second term has the argument of k x plus omega t, which represents a plane wave propagating to the left.

Classically, a free particle would travel with a certain constant speed. Quantum mechanically, a free electron is
described by a wave which propagates with a certain constant wave vector and frequency.
 SLIDE # 24

Let’s consider an electron traveling in the right or positive x direction, with the wave function psi x t equal to
A times exponential i k x minus omega t. k is the wave vector, which was defined in the previous slide, as
square root of 2 m E over h bar squared. Since energy E is p squared over 2 m, we get k equal to p over h
bar. So We can see that de Broglie’s relationship of p equal to h bar k is recovered, or we can say this justifies
our interpretation of k as the wave vector (or wave number in one dimension).

Remember we started with the Schrödinger equation. The solution of the Schrödinger equation for a free
electron in an empty space is a plane wave with the wave vector which is consistent with the de Broglie’s
relationship.

Electron expressed by the plane wave has a define energy E and definite momentum p. See the next slide for
more detail.

Having a definite momentum means that the uncertainty in momentum delta p is zero. This implies that,
from the uncertainty principle, the uncertainty in position is infinite, which means the free electron is spread
all over the space. It is in turn consistent with the fact that the wave function squared, which is A squared, is
constant at every position. This means that the probability to find the particle is the same at every position.
Recall that our interpretation of the wave function is that the square of the wave function corresponds to the
probability to find the particle.
 SLIDE # 25

Classically, a free particle has kinetic energy only, where the kinetic energy is m v squared over 2 or p
squared over 2 m. Quantum mechanically, since momentum p is h bar k, E is hbar squared k squared over 2
m. From this we get wave vector k which is given by square root of 2 m E over h bar and wave length
lambda expressed as h over square root of 2 m E.

Classically, a free particle with a well defined energy has a well defined momentum, through the relationship
between energy and momentum. Quantum mechanically, in the particle picture, a free electron with a well
defined energy also has a well defined momentum. In the wave picture, a free electron with a well defined
energy has a well defined wave vector or wave length, through the de Broglie’s relation of p equal to h bar k.
 SLIDE # 26

There are two categories of the problem when we solve the Schrödinger equation. The first
case is quantum confinement problem where electrons are confined by a potential. This will
lead to so called bound states with discrete energy levels. An example of this is an infinite
potential well problem which we will discuss next. The second category is transmission
problem where electrons are incident on a scattering region and transmit through it. In this
case, we are looking for the transmission probability. An example of this case is the step
barrier scattering problem.

In both the cases, we will start with writing down the Schrödinger equation, then figure out
the boundary conditions, and try to find the wave function. The wave function contains all the
information we want to know.
 SLIDE # 27

As a simple but important example of the electron confinement in a potential, let us consider the infinite
potential well problem.

As shown in the figure, there are three regions in space. Regions 1 and 3 have infinite potential, while Region
2 has a constant, zero potential. Electron is confined in Region 2.

In terms of the one dimensional potential V x, V x is 0 for x greater than 0 and less than a, where a is the
width of the potential well, and V x is infinite otherwise.

Let’s try to solve the Schrödinger equation in this case. As the first step, we write down the Schrödinger
equation. Inside the well, V x is zero, so we have the equation which is exactly the same as the case of the
free electron discussed previously. As before, we set k equal to square root of 2 m E over h bar. Then the
general solution is A exponential i k x plus B exponential minus i k x, where A and B are constants to be
determined.
 SLIDE # 28

As the second step, we apply the boundary conditions.

At the two boundaries of x equal to 0 and x equal to a, psi must be zero. This is because an infinite potential
means an infinite energy, so for an electron to exist in Region 1 or 3, it should have an infinite energy, which
is unrealistic, so we can conclude that the electron cannot exist in each of the infinite energy regions. In other
words, the probability to find the electron in the regions is zero, and since the wave function squared is the
probability, the wave function psi must be zero.

Applying the boundary conditions yields A plus B equal to zero and A exponential minus I k a plus B
exponential I k a equal to zero. We can write the two equations in the form of matrix vector multiplication as
shown in Equation 1. A two by two matrix times a column vector with A and B as the elements and the zero
column vector on the right hand side.

We have trivial solution of A equal to 0 and B equal to 0. Then, the wave function psi is zero, which does not
convey any information. Non-trivial solution exists only if the two by two matrix is not invertible. That means
that the determinant of the matrix, which is exponential i k a minus exponential minus i k a which is 2 times
sin k a is equal to zero as shown in Equation 2.

The solution of Equation (2) gives us k a equal to n times pi where n is an integer, one, two, three, and so on.
With the subscript n attached to k, k n is given by n pi over a, Then the wave length lambda n which is equal
to 2 pi over k n is given by 2 times a over n. In other words, if an integer times half the wave length
coincides with the well width a, we have the solution.
 SLIDE # 29

Since energy is equal to h bar squared k squared over 2 m and k has discrete values k sub n,
energy has also discrete values E sub n. We say that energy is quantized.

If we calculate the four lowest energies, E sub 2 which is the second lowest energy is 4 times
E sub 1. E sub 3 is 9 times E sub 1 and E sub 4 is 16 times E sub 1 and so on.

The corresponding wave functions are shown in the figure. We can see that the lowest energy
state corresponds to the case where half the wave length is equal to a, the next energy state
corresponds to two times half the wave length equal to a, the next energy state corresponds
to three times half the wave length equal to a and so on.

Physically, this is exactly the same phenomena as the occurrence of standing waves on a
guitar string. If we flick a guitar string whose two ends are fixed, initially the string would
seem to vibrate randomly with a lot of mixed frequencies, but at the end, all the frequencies
will decay out but only the frequencies satisfying the standing wave condition, that is, integer
times half the wave length equal to the string length, will survive.

The full wave function including the time part is obtained if we multiply exponential minus i
omega n t to the spatial wave functions.
 SLIDE # 30

We have just seen that the energy is quantized for the infinite potential well problem.

This means that, in the energy-momentum diagram of a free electron which is shown as the solid line in the
figure, the possible values for an electron in the infinite potential are discrete values denoted by the circles in
the figure.
 SLIDE # 31

For a finite potential well where U zero of a potential well is finite, energy is also quantized. But in contrast to
the case of the infinite potential well where the number of quantized energy levels are infinite, there are only
finite number of quantized energy levels. For example, for the potential shown in this slide, there are three
quantized energy levels. The corresponding wave functions have one, two, and three nodes, as shown in the
figure.
 SLIDE # 32

The quantization effect occurs for any potential in which electrons are confined. For example,
the quantized energy levels are shown for the triangular potential and the parabolic potential.
What we can notice in the figures is that, counting from the lowest energy level, the wave
functions have one node, two nodes, three nodes, and so on. This feature is universal for any
confinement potential.
 SLIDE # 33

Now, as the second category in solving the Schrodinger equation, let’s solve a transmission problem.

Contrary to the confinement potential problem, electrons are not confined in the transmission problem.
Rather, they are incident to a scattering region and transmit through it.

As a simplest possible example, let us consider a step barrier scattering problem.

The potential is zero in region 1 and U zero in region 2, with the corresponding Schrodinger equation as
shown. First, let us assume the energy E of the electron is greater than V zero. Then we have simple solutions
for the wave function, which are plane waves.

In region 1, the wave function psi 1 is given by A exponential i k 1 x which represents the incident electron
plus B exponential minus i k 2 x which represents the reflected wave. In region 2, the wave function psi 2 is
given by C exponential i k 2 x which represents the transmitted wave. Note that there is no wave that travels
to the left in region 2. A, B, and C are the amplitudes of the plane waves, which are to be determined as
follows.
 SLIDE # 34

We apply the boundary conditions at x equal to 0. The first boundary condition is that the
wave functions are continuous at the boundary. That is, psi 1 at x zero is equal to psi 2 at x
zero, which leads to equation 1. The second boundary condition is that the first derivative of
the wave function with respect to the position is continuous at the boundary, which leads to
equation 2. The first boundary condition seems to be natural, and the second one seems to
make sense if we think about the current continuity as shown in the inset.

Note that, contrary to the confinement potential problem, equations 1 and 2 together are not
an eigenvalue problem. The incident electron has the continuous energy levels from zero to
infinity. For a given E in the range of zero to infinity we solve for B over A and C over A from
equations 1 and 2.
 SLIDE # 35

From equations 1 and 2, we calculate C over A and B over A.

Then we calculate the incident, reflected, and transmitted current fluxes.

To get the idea of how the current flux is defined, take a look at the inset of the previous
slide.

Finally, we calculate the reflection coefficient R which is the ratio of the reflected flux to the
incident wave and the transmission coefficient T which is the ratio of the transmitted flux to
the incident flux. Note that T plus R is one, which means that current flux is conserved.
 SLIDE # 36

For the case where E is smaller than the step barrier potential V zero, the similar steps are
taken, but the difference is that the wave function of region two, psi 2 is given by an
exponentially decaying function. That is, psi 2 is C exponential minus kappa 2 x where kappa
2 is given as shown. Note that the reflection coefficient is 1 while the transmission coefficient
is 0.
 SLIDE # 37

Let's summarize the results of the step potential problem. For E greater than V zero, the
solution of the Schrodinger equation gives us the plane waves for reflected and transmitted
waves, with the reflection coefficient R of the ratio of the reflected flux to the incident flux
and the transmission coefficient T of the ratio of the transmitted flux to the incident flux.
 SLIDE # 38

Let's try to interpret the meaning of T and R in terms of the particle picture. If R is 0.3 and T
is 0.7, for example, that means that if we do a scattering experiment with electron particle,
out of 1000 trials, we expect to observe particle transmitted 700 times and reflected back by
the step 300 times.

This is in sharp contrast to the classical case. Classically, for any E which is greater than V
zero, T is always one, meaning that every time we do the scattering experiment, electrons are
always transmitted over the step.

Quantum mechanically, electrons behave like wave and our interpretation in terms of the
particle picture should include the probability.
 SLIDE # 39

The curve on the right shows the transmission as a function of energy.

The transmission T becomes 1 as E is increased. That is, as the electron energy becomes
greater relative to the step height, electrons do no feel the presence of the step and little
reflection will take place. This corresponds to the classical limit.

On the other hand, as E becomes closer to the step height, reflection increases sharply.
 SLIDE # 40

For E smaller than V zero, the wave function to the right of the step is an exponentially
decaying function as shown in the figure, which means that electron penetrates into the
barrier region by a certain distance. But it does not propagate to the right. So the
transmission is zero and the reflection is one.

Since the square of the wave function is proportional to the probability to find electrons,
there is nonzero probability to find the electron in the shaded region. But still the reflection R
is always zero if E is less than V zero. An explanation for this is given in the next slide.
 SLIDE # 41

Again, let us imagine we do a scattering experiment with electron particle. For example, let's
say, 900 times out of a total of 1000 times, electrons are directly bounced back without
penetration into the step. That means that, 100 times out of the 1000 trials, electrons will
penetrate into the step. But they will eventually bounce back, making the reflection one.
 SLIDE # 42

There is an open software for the step potential problem which can be accessed by the link
shown. You can check out the wave function, probability, transimission, and so on using the
graphic interface.
 SLIDE # 43

Next, a potential barrier problem.

The procedure is the same for the step barrier case we have just considered. We first write
down the Schrodinger equation for each region, There are three regions, region one, two, and
three as shown in the figure.

The wave functions are again traveling plane waves in regions one and three, but in region
two, the wave functions are exponentially decaying because the electron energy E is smaller
than the barrier height.

We then apply the boundary conditions at x equal to zero and at x equal to a where a is the
potential barrier thickness.

From the four equations, we solve for F over A and finally the transmission T.
 SLIDE # 44

Classically, it is impossible for a particle to tunnel through a barrier if its energy is smaller
than the barrier potential energy. Quantum mechanically, however, since electrons behave like
waves, there is a probabily that they tunnel through the barrier. The transmission T is the
tunneling probability.

From the tunneling probability formula, we can see that as the barrier thickness increases, T
decreases exponentially. For sufficiently thick barrier, the tunneling probability would be
practically zero, which is consistent with the classical result. But as the thickness becomes
smaller, the tunneling probability increases as shown in the figures to the right. For example,
the figure shows that for the barrier with 2 nanometer thickness, the tunneling probability is
substantial and for the 1 nanometer thick barrier tunneling probability becomes close to one.
 SLIDE # 45

For the case of E greater than V zero, the wave functions over the barrier in region 2 are
traveling waves, which is what is different from the previous case of E smaller than V zero. As
a consequence, the sine hyperbolic function changes to sine function in the transmission
formula.
 SLIDE # 46

For the over-the-barrier case, what is noteworthy is the occurrence of the resonance. From the
transmission formula, we can see that if the argument of the sine function is zero the
transmission becomes one. The condition is satisfied when k prime times a is equal to
interger times pi, which corresponds to the standing wave condition over the barrier width.
 SLIDE # 47

Using the same software mentioned before, you can also do potential barrier simulations.
 SLIDE # 48

Some real world examples of confinement potential and an application of the tunneling
phenomena in the scanning tunneling microscopy are discussed in the followings.
 SLIDE # 49

Quantum dots can be formed in an InGaAs GaAs heterostructure as shown in the figure. The
size of the quantum dots are so small that the diameter of a dot is of the order of a
nanometer. Electrons in such a quantum dot are confined in three dimensions, and so called
particle in a box.

If a quantum is modeled by a three dimensional infinite potential well problem, the quantized
energy levels are given as a straight forward extension of the one-dimensional potential well
problem discussed earlier.
 SLIDE # 50

The confinement effect in quantum dots can be applied to displays. In quantum dots, discrete
energy levels are formed and light is emitted when electrons make transistion from an energy
level to another. The wave length or the color of the emitted light is determined by the
energy difference of the two levels. By changing the size of the quantum dot we can control
the energy levels of the dot and hence the color of emitted light.
 SLIDE # 51

The second example is a quantum well, where electrons are confined in a layer sandwiched by
other semiconductor layers as shown. The discrete energy levels formed in the so called
quantum well can be exploited in such applications as lasers and LEDs.
 SLIDE # 52

Another real world example of quantum confinement is electron gas in a metal oxide
semiconductor device. In a MOS device, when a sufficiently large gate electric field is applied,
the energy band in the semiconductor is bent as shown in the figure and electrons become
confined in the potential well.
 SLIDE # 53

An important application of the quantum tunneling phenomena is found in scanning