Você está na página 1de 122

Elements of Quantum Chemistry

K L Sebastian

October 10, 2016

Indian Institute of Science


Contents

1 Wave-particle duality of matter 3


1.1 Two-slit experiment with bullets (particles) . . . . . . . . . . . . . 3
1.2 Two-slit experiment with waves . . . . . . . . . . . . . . . . . . . . 5
1.3 Two-slit experiment with electrons . . . . . . . . . . . . . . . . . . 7
1.4 Wave-particle duality . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4.1 de Broglies relation . . . . . . . . . . . . . . . . . . . . . . 12
1.4.2 Heisenbergs uncertainty principle . . . . . . . . . . . . . 12
1.4.3 Standing waves and stationary states . . . . . . . . . . . . 15
1.5 The way in which the electron propagates - the path integral . . . 20
1.5.1 A more complex slit experiment . . . . . . . . . . . . . . . 21
1.5.2 Classical mechanics - only a hand wave away! . . . . . . . 23
1.6 Path integrals and random walks . . . . . . . . . . . . . . . . . . . 24
1.7 The puzzle of quantum mechanics . . . . . . . . . . . . . . . . . . 27
1.8 The basis for the whole of chemistry . . . . . . . . . . . . . . . . . 29

3 The Postulates of Quantum Mechanics 31



3.1 The Schrodinger Equation . . . . . . . . . . . . . . . . . . . . . . . 31
3.2 Postulates of quantum mechanics . . . . . . . . . . . . . . . . . . 32
3.2.1 Postulate I . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.2.2 Postulate II . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2.3 Postulate III . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2.4 Postulate IV . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.5 Notes on the Postulates . . . . . . . . . . . . . . . . . . . . 37
3.3 Stationary States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4 Simple Systems 41
4.1 Continuity of the wave function . . . . . . . . . . . . . . . . . . . . 41
4.2 The particle in a one dimensional box . . . . . . . . . . . . . . . . 41
4.2.1 Energy levels . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.2.2 The wave functions and associated probabilities . . . . . 44
4.3 Particle in a three dimensional box . . . . . . . . . . . . . . . . . . 47
4.4 Cubic box: degeneracy . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.5 Two Systems that do not interact . . . . . . . . . . . . . . . . . . . 54

5 The Hydrogen Atom 57

iii
iv CONTENTS

5.1
The Schrodinger Equation . . . . . . . . . . . . . . . . . . . . . . . 57
5.2 Solution of the equation for () . . . . . . . . . . . . . . . . . . . 61
5.3 The functions () and R(r ) . . . . . . . . . . . . . . . . . . . . . . 62
5.4 The full wave function . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.5 Shapes of atomic orbitals . . . . . . . . . . . . . . . . . . . . . . . . 65
5.5.1 The 1s atomic orbital . . . . . . . . . . . . . . . . . . . . . 65
5.5.2 The 2s atomic orbital . . . . . . . . . . . . . . . . . . . . . 68
5.5.3 The 2p orbitals . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.5.4 The 2p z orbital . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.5.5 The 3s atomic orbital . . . . . . . . . . . . . . . . . . . . . 73
5.5.6 The 3p atomic orbitals . . . . . . . . . . . . . . . . . . . . . 75
5.5.7 The 3d atomic orbitals . . . . . . . . . . . . . . . . . . . . . 75
5.5.8 Note on orthogonality and Independence . . . . . . . . . 76

6 Angular Momentum and the Spin of the Electron 83


6.1 Wave function for an electron in the hydrogen atom . . . . . . . . 88

7 Many electron atoms and the Paulis exclusion principle 89


7.1 Inclusion of the spin part: Paulis principle . . . . . . . . . . . . . 91
7.2 Many-electron atoms . . . . . . . . . . . . . . . . . . . . . . . . . . 92
7.2.1 Effective nuclear charge, Z = Z S . . . . . . . . . . . . 92
7.3 Reading assignments . . . . . . . . . . . . . . . . . . . . . . . . . . 92

8 Chemical Bonds 95
8.1 Why bonds are formed? . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.2 The Hydrogen Molecular Ion . . . . . . . . . . . . . . . . . . . . . . 95
8.3 Wave function for the electron . . . . . . . . . . . . . . . . . . . . . 98
8.3.1 The bonding molecular orbital . . . . . . . . . . . . . . . . 98
8.3.2 The anti-bonding molecular orbital . . . . . . . . . . . . . 100
8.3.3 Valence bond theory . . . . . . . . . . . . . . . . . . . . . . 102
8.4 The hydrogen molecule . . . . . . . . . . . . . . . . . . . . . . . . . 103
8.4.1 Molecular Orbital Theory . . . . . . . . . . . . . . . . . . . 103
8.4.2 Valence Bond Theory . . . . . . . . . . . . . . . . . . . . . 103
8.4.3 MO theory - incorrect dissociation limits . . . . . . . . . . 105
8.5 Hybridization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
8.5.1 sp-hybridization . . . . . . . . . . . . . . . . . . . . . . . . 107
8.5.2 sp 2 -hybridization . . . . . . . . . . . . . . . . . . . . . . . 108
8.5.3 sp 3 -hybridization . . . . . . . . . . . . . . . . . . . . . . . 109

8.6 Huckel Molecular Orbital (HMO) Theory . . . . . . . . . . . . . . 111
8.6.1 Ethylene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
8.6.2 Benzene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

8.6.3 Huckels (4n + 2) rule . . . . . . . . . . . . . . . . . . . . . 114
CONTENTS 1

(a) (b)

Two Views on Quantum Mechanics: (a) frightening, like the Count Dracula or
(b) beautiful like the cine actress Katrina Kaif (images from the internet).
Chapter 1

Wave-particle duality of matter

Microscopic (having mass roughly that of atoms, molecules and smaller) objects
are very peculiar. Their behaviour resembles that of a particle, as well as that of a
wave. In particular, when we detect them, we find them at a particular location as
would be the case for a particle, but when they move, their movement resembles
the propagation of waves. In order to make this clearer, we consider the following
two-slit experiments1 .

1.1 Two-slit experiment with bullets (particles)


First, we imagine an experiment done with bullets, which every body will agree
are particles. Consider a man with a gun shooting out a stream of bullets. He

Figure 1.1 Two slit experiment with bullets. (figure taken from:
http://www.colorado.edu/physics/2000/index.pl)

shoots them in random directions with the consequence that the bullets have
a fairly large angular spread. There is a wall in front of the man, which has two
1 Our discussion follows that of Feynman [1], which may be referred to more details.

3
4 Chapter 1 Wave-particle duality of matter

Figure 1.2 I 1 and I 2 are the patterns produced by bullets on the backstop with only one
of the slits is open while I 12 is the pattern with both the slits open.

openings in the form of rectangular slits. Bullets can pass through them and fall
on the backstop (see Fig. 1.1). The backstop (made of soft wood) retains the
bullets that hit it. Now we keep only one of the slits (slit number 1) open and ask
the man to shoot out a large number of bullets. After he has done so, we look at
the distribution of bullets that have hit the backstop after passing through the slit.
We shall refer to this as the intensity distribution for the bullets. As we have kept
only slit 1 open, we refer to the resultant pattern on the backstop as I 1 (see Fig.
1.2).
The important point to notice is that bullets have the property of lumpiness".
This means that bullets come only in integral numbers - you cannot have a
fraction of a bullet, which is essentially a property of particles. If one keeps only
the slit number 2 open, the result is the intensity distribution I 2 . Now if both
the slits are kept open then the pattern produced on the backstop would be I 12
shown in Figure 1.2. It is obvious that

I 12 = I 1 + I 2 . (1.1)

Thus in the case of particles (bullets), we find that to get the total intensity I 12 that
results when both the slits are kept open, one simply has to add up the individual
intensities I 1 and I 2 . Further particles have the property of lumpiness", that is
they come in integral numbers, like 1, 2, 3 . . . and you do not get something like
3.5 bullets, when they arrive at the backstop (detector).
1.2 Two-slit experiment with waves 5

1.2 Two-slit experiment with waves


Now consider the same experiment done with waves. We have a trough of water
and an electromechanical device which moves up and down on water to produce
circular waves on the surface of water. In front of the source at some distance
we have a partition cutting the surface of water, with two slits and an absorber
behind that. The important fact here is that unlike the bullets the waves have
no lumpiness. The intensity of waves produced is proportional to the extent
of motion (vibration) of the electromechanical device and can be continuously
varied, without any lumpiness (discreteness).
We initially perform the experiment with one of the slits open and the other
closed and the pattern produced by the waves passing through a single slit is
shown in Figure 1.3. If the displacement produced at point by the wave when slit
1 is open is d 1 , then the corresponding intensity would be

I 1 |d 1 |2 . (1.2)

Similarly the intensity for the wave if only slit 2 is open would be

I 2 |d 2 |2 . (1.3)

Now consider both the slits to be kept open. Obviously, the displacement at any
point when both the slits are open is

d 12 = d 1 + d 2 . (1.4)

Hence the intensity of the resultant wave is given by


2
I 12 d 12 = |d 1 + d 2 |2 . (1.5)

At some points the waves from the two slits are in phase - i.e., d 1 and d 2 add up to
give a large net displacement and this this results in constructive interference. At
some other points they are completely out of phase - i.e, d 1 and d 2 are in opposite
directions, cancelling each other. This results in destructive interference. The
pattern we find on the absorber is an interference pattern as shown in Figure 1.4.
Even though we have considered waves on the surface of water, exactly the same
thing happens with electromagnetic waves too. This is actually an important
aspect of wave phenomena. Whenever we have wave phenomena, it is possible
to assign some thing which we call the wave function", which in general we shall
denote by . Its use is to calculate the intensity of the wave, which is given by
I ||2 . It is clear that in the case of waves on the surface of water, is to be
identified as the displacement of the water surface, at the point where we wish to
calculate the intensity.
Unlike particles, waves have

I 12 6= I 1 + I 2 . (1.6)
6 Chapter 1 Wave-particle duality of matter

Figure 1.3 Pattern produced by waves on the backstop with only one slit open (the
green pattern is produced when only slit 1 is open and the red pattern when only slit 2
is open)

Figure 1.4 Interference of waves on the surface of water when both the slits are open.
1.3 Two-slit experiment with electrons 7

In the case of waves, it is the wave function (d in the case of water waves) that is
additive. i.e.
12 = 1 + 2 . (1.7)
As before,
I 12 |12 |2 , I 1 |1 |2 and, I 2 |2 |2 (1.8)

1.3 Two-slit experiment with electrons


We now consider a similar experiment done with electrons. We have a source of
electrons and a wall in front of it with two slits and a detector behind the wall. It
is possible to do this experiment with the intensity of the electron beam reduced
so much that at any instant there is only one electron arriving at the detector.
The electron is found to arrive as a lump, which means that we have either zero
or one electron arriving. Thus electrons have lumpiness", which as we saw is a
characteristic of particles. If we conducted the experiment initially slit 1 (or 2)
open and the other closed and the pattern produced on the absorber is I 1 (or I 2 ),
as shown in Figure 1.8.
Now imagine repeating the experiment with both the slits open and strangely,
the pattern (Figure 1.8) we observe is similar to an interference pattern produced
by waves!
An experiment of this kind was performed at the Hitachi laboratories in Japan,
under the leadership of Tonomura and the results reported in the American
Journal of Physics (American Journal of Physics, 57, 117 (1989)). The experimental
set up is shown in Fig. 1.5, and the intensity distribution of the arrival points of
the electron is shown in Fig. 1.7 (note that this is taken from the video given by
the experimentalists).
This experiment means that electrons, though they have lumpiness, are able
to exhibit interference! Thus we have to associate a wave with the electron, so
that it can exhibit interference. Naturally, there will be a wave function associated
with this wave. As is usual with waves, the square of the magnitude of this will
determine the intensity of electrons arriving at a particular point. In the case
of electrons, intensity simply means their number. One may now think: the
experiment is done with a large number of electrons. So these electrons may
interfere with each other, thus leading to the interference pattern. Is this what is
happening? To check whether this is case, we can do the experiment with single
electrons. If we send out only a single electron from the source, one will see only
a single flash, at a particular location on the detector. If we repeat the experiment
again, the electron, in general does not arrive at the same point! (see the snapshot
from the video, shown in Fig. 1.6, where this is clearly seen). So it appears that the
experiment with single electron is not reproducible! Imagine we collect together
arrival data of a large number of electrons (let us say 100, 000) and superpose
all the data - then we get the interference pattern. If the experiment is repeated,
by collecting together date for another 100, 000 electrons, one reproduces the
same interference pattern. So even though result of the experiment with a single
8 Chapter 1 Wave-particle duality of matter

Figure 1.5 The equipment used by the Tonomura group to perform double slit experi-
ment. There is fine wire of thickness 1/1000 mm the two sides of which form the two
slits. Monoenergetic electrons pass through the two sides, and are detected by a very
sensitive detector, which can produce a flash of light at the point of arrival of a single
electron.

Figure 1.6 Pattern produced by the arrival points of few electrons


1.3 Two-slit experiment with electrons 9

Figure 1.7 Pattern produced by a large number of electrons at the detector. The pat-
tern obtained by collecting together of a large number of such electrons is clearly an
interference pattern.

Figure 1.8 Electrons can be detected one by one and hence are like bullets. But when
both the two slits are open, they show an interference pattern, characteristic of waves!
10 Chapter 1 Wave-particle duality of matter

Figure 1.9 The famous remark of Einstein in which he expressed his disbelief in the
probabilistic nature of quantum theory.

electron is not reproducible, the data for a collection of electrons is. Can one say
something about the outcome if the experiment is done with a single electron?
The answer is very interesting. We can definitely say that the electron will arrive
with more likelihood in regions where there is constructive interference than
in regions where there is destructive interference. That is, instead of talking
about intensities, one has to talk of probabilities, which are proportional to the
intensities in an experiment done with a beam of electrons. Thus we realise
that for a single electron, the probability of detecting it a particular location is
proportional to ||2 where is the value of the wave function evaluated at that
point. That is,
P ||2 = . (1.9)
where stands for the complex conjugate of . The fact that one can only talk
of probabilities for the case of single electron had left even reputed scientists
unhappy. Einstein was the most famous of them. He used to remark God does
not play dice" (see Fig. 1.9) in discussions of quantum theory.
Thus we find that the electron has a dual nature - it exhibits properties of both
a particle and a wave. The same experiment conducted on a single electron will in
general yield various results that are not necessarily the same, and we can merely
give the probability for the occurrence of a particular possibility. However the
result for the study on a large number of electrons, for instance, the interference
pattern, is always exactly reproducible.
1.4 Wave-particle duality 11

Figure 1.10 (a) The molecule C 60 and (b) the interference pattern reported by the
group of Zeilinger in Nature 401, 680-682 (1999).

1.4 Wave-particle duality


Isaac Newton had put forward the idea that light behaves like a beam of particles.
Even before him, Huygens had proposed that light has wave-like nature though
there were not many takers for his proposition, probably due to the scientific
stature of Newton. However later, Maxwell formulated his electromagnetic theory
which firmly established that light has a wave nature. Further, the interference
patterns that one observes in experiments involving light confirms this. Einstein
re-established the concept of the particle nature of light through his findings
on photoelectric effect which can be explained only if we consider light to be
made up of particles, that are referred to as photons. Further evidence came
from the experiments of Compton. Compton studied the inelastic scattering of
light (X-rays or Gamma rays) from matter. In the scattering, part of the energy
of light is lost resulting in a decrease in the energy of light, causing an increase
in its wavelength. The energy lost by the X-/gamma ray is transferred to the
scattering electron, which recoils and is ejected from its atom (which becomes
ionised), and the rest of the energy is taken by the scattered, "degraded" photon)
These experimental results can be explained only if we consider light to be having
particle nature. Thus we conclude that light exhibits wave as well as particle
nature.
It is an experimental fact that not only light, but electrons, neutrons, protons,
atoms, and even molecules like C 60 and phthalocyanines have this dual nature
(see Figs. 1.10 and 1.11). We would like to believe that the wave particle duality is
applicable to even heavier objects, but as of now, these are the heaviest particles
for which experimental verification is available.
12 Chapter 1 Wave-particle duality of matter

Figure 1.11 (a) The molecule pht hal oc y ani ne and (b) the interference pattern re-
ported by Arndt et. al. in Nature Nanotechnology, 7, 297 (2012).

1.4.1 de Broglies relation


We have seen with the help of double slit experiment that an electron has a wave
and a particle nature. In 1924, Louis de Broglie suggested that every particle has a
wave associated with it and that the wave-length of the wave is determined by the
momentum p of the particle. Further, he suggested that the wavelength is given
by
= h/p, (1.10)
with
p = mv, (1.11)
where h is Plancks constant.
de Broglie proposed how his relation could be verified experimentally - one
has to adjust the momentum of electrons such that the wavelength of the wave
associated with it matches the lattice spacing between atoms in a crystal. Then
one would be able to observe diffraction patterns for electrons scattered from
crystals. Three years later, in 1927 Thomson, Davisson and Germer in two inde-
pendent experiments observed the predicted electron diffraction. de Broglie was
given the Nobel prize in 1927 for his hypothesis while Thomson and Davisson
were awarded Nobel prize in 1937 for their experimental verification.

1.4.2 Heisenbergs uncertainty principle


A wave in the x-direction, having a wavelength may be represented mathe-
matically by the function Ae i 2x/ . So if one has a particle with a well defined
1.4 Wave-particle duality 13

momentum p in the x-direction, then the wave function associated with it can be
thought of as
(x) = Ae i 2x/ (1.12)
p
where, is wavelength of the particle and i = 1 and A is a constant. As = h/p,
this may also be written as
(x) = Ae i px/~ , (1.13)
where ~ = h/(2). For such a wave, the probability density will be given by

||2 = Ae i px/~ A e i px/~ = |A|2 . (1.14)

which shows probability of finding the particle is a constant everywhere. The ||2
is spread all over the space, and hence one could find it anywhere along the x-axis!
This is the case where we know the momentum of the particle precisely. That is,
there is no uncertainty in the momentum, which means that the uncertainty in
momentum p = 0. But we find that in this case the position of the particle is
not at all certain. That is, x = . Thus if momentum of particle is completely
specified, the associated wave is spread all over (delocalised) and it is impossible to
predict where the particle would be found.
What if wave is localised and is spread only over a region of size L as shown
in Fig. 1.12? Then clearly, the uncertainty in its position is of the order of L. i.e.
x L. The way to produce such a localised wave packet can be understood
from Fig. 1.13. The figure shows that superposing two waves of slightly different
wave lengths (the waves indicated by the red and green color), to produce a new
wave (indicated by blue color).

Figure 1.12 A wave packet. The uncertainty in position is finite and is of the order of L,
where L is the width of the packet.

Thus adding together two waves of slightly different wavelengths can produce
a new wave, which has larger values in certain regions and smaller values in
other regions. It is easy to realise that adding more and more waves of different
wavelengths with appropriate weights (contributions), one can produce a wave
packet of the shape shown in the Fig. 1.12. We now illustrate this mathematically.
x2

Imagine an electron has the wave function = Ae 2L2 where A is a constant,
assumed to be real. This is appropriate for a wave of the form shown in Fig. 1.12.
14 Chapter 1 Wave-particle duality of matter

Figure 1.13 Adding two waves of slightly different wave lengths. The red and green
waves are added together to get the blue one.

2
x
On calculating , one gets the probability distribution = A 2 e L2 which
vanishes rapidly as one goes outside the region |x| < L. Thus it represents a wave
localised in a region of length L. For such a wave packet, the uncertainty in
position
x L. (1.15)
Such a wave packet is a combination of a large number of waves of different wave
lengths as one sees from
p
2
2~ L2 p 2
Z i px
x2 +
= Ae 2L =A d pe 2~2 ~ (1.16)
L

The right hand side of Eq. (1.16) shows that it can be written in terms of a large
number of waves of the form e i px/~ , which corresponds to of momentum p. The
p L2 p 2

importance of that momentum is determined by the weight factor A 2 ~
L e
2~2 .

Note that the contributions to the above integral are essentially from the values
of p ~/L. Thus, there is an uncertainty in the momentum given by

p ~/L. (1.17)

Combining equations (1.15) and (1.17), one gets

xp ~. (1.18)

It is obvious that as one tries to make L smaller, the larger is the range of momenta
that contribute to it. A slightly more rigorous analysis of this leads to the result

xp ~/2. (1.19)

This was originally arrived at by Heisenberg and hence is known as the Heisen-
bergs uncertianty principle, which states it is impossible to have a particle with
both x and p arbitrarily small". The meaning of this statement is that if one
tries to reduce the uncertainty in position, then uncertainty in momentum has to
increase and vice versa.
1.4 Wave-particle duality 15

(a) (b)

Figure 1.14 Wave packet being reflected from a wall. 1,2,3,.... show the successive
positions of the packet resulting in its reflection.

1.4.3 Standing waves and stationary states


The Heisenbergs uncertainty principle, discussed in the previous section is a
very important consequence of the wave particle duality and the de Brogllie
relationship. We now consider another very important consequence of these. To
set the stage, let us consider a long string attached to a wall. Imagine that you
take hold of the other end and stretch it tight.
Now, we have a medium, the stretched string, which can support waves on it.
Suppose move your hand up and bring it back to its initial position very quickly so
as to set up a wave packet on the string (see Fig. 1.14). In the figure, 0 10 shows the
initial wave packet that is created. The packet would move on the string, towards
the wall, hit and be reflected from it. Different instants in its motion towards the
wall and reflection from the wall are shown in (a) and (b) respectively, in the Fig.
1.14. Thus, in this case, we have the wall, which reflects the wave. If the medium
was infinite (for example waves on the surface of water at the centre of a big lake,
which for all practical purposes may be taken to be infinitely large) then the waves
can go on forever, without being reflected back (obvious!). Now suppose we have
obstacles on both the sides so that the wave cannot go out of a finite region of the
medium - then what would happen? Obviously there would be reflections from
both the boundaries. So any travelling wave would get reflected, again and again,
and would go on. (To be accurate: dissipation of energy, which is always there
in macroscopic waves, ensure that the wave dies down after a few reflections.
We consider the ideal case, where there is no such dissipation, and this is the
case with isolated atomic/molecular systems). However, if the disturbance that
we have set up is just right, a very interesting thing happens. The wave and the
reflected wave would reinforce one another, and would form what is referred
to as a standing wave. The simplest possible such pattern for a string stretched
between two walls, is shown in the Fig. 1.15. This is referred to as a normal mode
of the string. The string has a displacement, looking like half a wave (wave length
of which is 2L where L is the separation between the two walls). As time passes,
the whole of the string would move up and down. The shape does not change at
all, though the displacement of the string from the equilibrium position changes
at all points, except at the two end points. The next simplest normal mode too is
16 Chapter 1 Wave-particle duality of matter

Figure 1.15 The different normal modes (standing wave patterns) on a stretched string
between two walls. The n t h mode has n 1 nodes.

Figure 1.16 Normal modes of a stretched membrane. Left hand top corner shows the
membrane. Its periphery is square in shape. Successive modes of increasing complex-
ity are shown. Note that nodes, along which there is no displacement of the membrane,
are lines.

shown in the Fig. 1.15. To understand this, one has to imagine that the string is
divided in to two equal halves. At any instant during this vibration, if one of the
halves is moving up, the other would be moving down and vice versa. Clearly, the
center of the string does not move. This is referred to as a node. It is also possible
to have more complex normal modes of vibration, as shown in the Fig. 1.15. The
number of nodes in these modes are 2 and 3 respectively. It is easy to imagine
that one can have modes with any number of nodes - 0,1,2.... Hence, in principle,
the string has infinite number of such modes. Each such mode is referred to as
a standing wave pattern, because the waves are not going anywhere! Further,
1.4 Wave-particle duality 17

Figure 1.17 Left hand top corner shows the membrane. Five standing wave patterns
with increasing number of nodes are shown. Note that the nodes can be lines or circles.

each is characterised by its number of nodes, the simplest having no nodes (we
do not count the end point as a node, as we have imposed it on the string!), the
next having one node, the next having two etc. Further, in this case, the nodes are
points. That is, if the waves are in a one-dimensional medium, then nodes are
zero dimensional.
Now we think of waves in a two dimensional medium, like the stretched
membrane of a drum (see Fig. 1.16, where we think of a rectangular drum). We fix
the membrane along its periphery and thus restrict the waves to move only in a
limited region. This means that we would have reflections from the boundaries
of the membrane. As before, if one produces an arbitrary disturbance on the
membrane by hitting it, it would move around, undergoing multiple reflections.
As before, if one produces the right kind of disturbance, then the wave and its own
reflections can reinforce one another, thus leading to standing waves, which are
not moving along the membrane. The simplest of these standing wave patterns
are shown in the Fig. 1.16. This has no nodes, while the next two has one node
each, the next has two nodes and so on....Further, in this case, the nodes are lines
along which there is no displacement of the membrane. Thus, if one has waves
in a two dimensional medium, the nodes are one dimensional. If one thinks of a
circular membrane (see Fig. 1.17) the situation is just the same, except that now
nodes can be also be circular, in addition to being lines.
What is the relevance of all this to our matter waves? Exactly the same can
happen with matter waves too. Thus, suppose, one confines an electron to a
rectangular three dimensional box (this is often used as a model for electrons in a
rectangular piece of metal).
The electron is not able to go through the walls of the box which confines it -
18 Chapter 1 Wave-particle duality of matter

Figure 1.18 A particle in a box is a model that is used to describe behavior of electrons
in metals. The walls of the box do not allow the particle to go out and would get re-
flected when it hits a wall. This means that the waves associated with it have to be
confined within the box.

this means that the electron gets reflected from the sides of the box. Naturally the
wave associated with it too is getting reflected. Therefore, one can have travelling
waves, which go on getting reflected and if the conditions are right, standing
wave patterns too can be formed. As in the earlier cases the number of such
standing wave patterns that are possible is infinity. Further, as the waves are in
three dimensions, the nodes have to be two dimensional. That is, they are planes.
The simplest standing wave pattern would have no nodes, the next would have
one node, etc....
In this connection it is interesting to note that in beautiful experiments using
the scanning tunnelling microscope, the group of Eigler et. al. have observed
standing wave patterns of electrons confined to a circular region on the surface
of Cu metal. The confinement was achieved by having Fe atoms sitting at the
boundaries of the circle. This effectively is a two dimensional circular box for
electrons. The standing waves were imaged using the microscope. In the Fig. 1.19,
each peak with an yellow tip is from an Fe atom, and the standing waves inside
the circular region are due to the electrons confined within the box.

Standing waves in atoms - atomic orbitals!

Now we consider the most interesting example: the hydrogen atom. Let us think
of the motion of the electron around the proton. As they are oppositely charged,
they attract one another. This means that the electrons potential energy near the
1.4 Wave-particle duality 19

Figure 1.19 Scanning tunneling microscopy images of a ring of Fe atoms on the surface
of Cu and the resultant standing wave patterns of electrons inside the ring, reported by
Eigler et al. in Nature 363, 524 - 527 (1993).

Figure 1.20 Potential energy of an electron near the proton. In the vicinity of the pro-
ton, the potential energy is very low, and as the electron goes away, it increases, effec-
tively causing the electron to be confined to the vicinity of the proton. It can escape
from the atom only if it is given enough kinetic energy to move away from this trap.

proton is low, as shown in Fig 1.20. As the electron is negatively charged, if we


bring the two together, there is lowering of the energy of the system. If we have
brought them together and left them in this low energy state, then they are happy
to be together and the attraction between them would keep them together. The
electron is of course moving about (otherwise it would fall in to the nucleus and
further it has to satisfy Heisenberg!) but it does not have enough energy to go far
away. This essentially means that we are confining the electron waves to a certain
region in space. Therefore, here also, it is possible for standing wave patterns
to be formed. Each such standing wave leads to what we call an atomic orbital.
20 Chapter 1 Wave-particle duality of matter

Figure 1.21 Atomic orbitals - these are just standing waves formed by the electron
wave!

There would be an infinite number of such standing wave patterns, which means
that the H atom has an infinite number of atomic orbitals.
Remember that the simplest standing wave pattern has no nodes. This is the
reason why the 1s atomic orbital has no nodes. The next would have one node,
which is two dimensional. This is what happens in the 2s atomic orbital. Here,
unlike the rectangular box, one has the interesting situation of having a node
that is not a plane. The node still is two dimensional and is just the surface of a
sphere. It is also possible to have nodes which are planes - 2p x , 2p y and 2p z have
one node each and these are planes (the yz, xz and yz planes respectively). Some
of these atomic orbitals are shown in the Fig.1.21. We will discuss them in great
detail in the chapter on the hydrogen atom.

1.5 The way in which the electron propagates - the path


integral
The most important question that arises out of all our discussions is: how does an
electron move in space? The answer to this is extremely interesting (and puzzling):
through all possible paths that it can take! In this section we discuss how this
1.5 The way in which the electron propagates - the path integral 21

surprising conclusion arises. In order to understand this, we first consider a more


complex thought experiment.

1.5.1 A more complex slit experiment


We now consider a more complex variation of the slit experiment. This being
an imaginary experiment, we can make it as complex as we wish. We imagine
that there are three walls between the source of the electron and the detector, as
shown in the Fig. 1.22.

Figure 1.22 (a) An experimental setup, having three walls, each with two slits. There
are eight different paths that the electron could take. (b) An arrangement with fourteen
walls each with four slits. There are 414 possible paths.

Each of these walls has two slits. Now we imagine that we have an electron
leaving our source at the initial time zero and we want to calculate the probability
of finding the electron at the point P, where we have a detector, after a time t .
To calculate this, we first have to calculate the wave function for the electron at
this point at that time. To reach this point, the wave associated with the electron
could have travelled through any of the eight different possible paths, of which
three are shown in the figure. Each such path makes a contribution to the wave
function. Thus, in this case,

= 1 + 2 + .....8 (1.20)

If now, one imagines there are more walls, each with more slits, then the number
of path that one has to consider increases (see Fig. 1.23). In the limit where
there are a large number of walls (infinity) between the source and the detector,
22 Chapter 1 Wave-particle duality of matter

Figure 1.23 The limit where number of walls and number of slits become infinite.

and each one of them has an infinite number of slits, then the wave function
is a sum of (!) such terms. We can go on increasing the number of slits
on each wall such that the walls eventually disappear - the number of paths,
correspondingly goes on increasing. When we consider the limit in which all the
walls have all disappeared, we are thinking of the propagation of the electron in
free space. Thus, we arrive at the following strange2 conclusion: all the possible
paths that we can imagine, starting at the source at the time zero and ending at
the detector at the time t make a contribution to the wave function. To calculate
the one has to evaluate the contribution of each path and then perform the
sum (remember, it involves terms.). A pictorial representation of the state of
affairs is given in the Fig. 1.22. The sum, involving such a large number of terms
is really frightening, but it has been possible to make sense out of the sum and
perform it, thanks to the genius of Richard Feynman (a similar sum occurs in
the theory of Brownian motion, and was discovered by another genius, Norbert
Wiener and we shall discuss the idea later). It is necessary to rigorously state what
we mean by a path - it is the position of the particle given as a function of time, for
all time between 0 and t , starting at the source at the time zero and ending at the
detector at the time t . We now have a problem: given any path (or sets of paths),
we have to know the contribution that it (they) make to the wave function. This
of course, is determined by the detailed behavior of the path itself. In principle,
even paths which start at the source, go to the moon, and then come back to the
detector have to be considered - but one expects that their contribution has to
be extremely small. The contribution of a path is determined by the action S
associated with that path. What makes this immensely interesting is that action
2 strange to us, mortals!
1.5 The way in which the electron propagates - the path integral 23

is a quantity, well known in classical mechanics. (In fact classical mechanics may
be derived from the principle of least action, which we will discuss in the next
section). Action is defined as follows: Suppose we consider one of these paths,
starting at the source at the time zero and arriving at the detector at the final time
T . At any time t , it is possible to uniquely specify the position of the particle as
the particle followed this, then it would, at any instant of time, have a definite
position. This obviously means that it would also have a well defined velocity,
and consequently, it is possible for us to calculate the kinetic energy, K .E . and the
potential energy P.E . at each instant of time. Thus one can get K .E . P.E . at any
instant during its motion. The integral of this quantity over the particles journey
from source to the detector is defined as the action. Thus,
Z T
S= d t (K .E . P.E .) . (1.21)
0

Once we have calculated the action, we can calculate the contribution of the path
as N e i S/~ and then calculate the wave function as the sum over all paths, as

=N e i S/~ .
X
(1.22)
al l pat hs

The N above is the same for all the paths and ensures that is normalized (The
word normalized means that d has to be equal to unity, so that d
R

may be thought of as the probability of having the system in the volume element
d . More details of this will be given in the chapter on postulates of quantum
mechanics. This approach to quantum mechanics is due to Richard Feynman.
Because one is summing over all the possible paths, Feynman called this the path
integral approach. It makes the motion of quantum objects as physical as it can
be made. The behavior of a quantum particle, we realize, is rather peculiar. All
paths that one can imagine, make a contribution to its wave function, and hence
are important, while if it obeyed classical mechanics, it follows only one of these
paths.

1.5.2 Classical mechanics - only a hand wave away!


It is straight forward to wave your hand, like a magician and get classical mechan-
ics out of this approach, as an approximation. For this, we look at the Fig. 1.24.
As we have seen, the basic ingredient of quantum mechanics is the Heisenbergs
uncertainty principle of Eq. (1.19). This forbids the simultaneous, exact specifica-
tion of position and momentum of a particle. The value of ~ = 1.05456 1034 J s,
is very small and that is the reason why quantum phenomena are important
only for particles of atomic or subatomic masses. If one imagines a Universe,
where ~ was even smaller, let us say by ten orders of magnitude, then classical
mechanics should be applicable to even these particles. Thus mathematically, if
we let ~ 0 then classical mechanics should follow as an approximation from
quantum mechanics. Let us now see what happens to our Eq. (1.22) in this limit.
As we make ~ 0 the quantity S/~ becomes larger and larger, as a result of which
24 Chapter 1 Wave-particle duality of matter

Figure 1.24 Classical mechanics follows from the sum over all paths approach easily.

e i S/~ which is a combination of sine and cosine functions oscillates very rapidly
as one goes from one path to another. Thus the two paths P and P 1 shown in
Fig. (1.24) would have very different values for S/~ and hence in the sum of e i S/~ ,
these two paths would tend to cancel each other. On the other hand suppose
we think of the classical path, shown as C 3 . It has the interesting property that
the action is an extremum. This means that if one considers any path near it (for
example the path C 1 ) the actions for these two are the same, which implies that
e i S/~ for these two paths, instead of cancelling each other would add up. Thus,
only the classical path and a small set of paths near it are important, in the limit
where ~ is made smaller and smaller. As one decreases ~ the number of paths
that contribute become smaller and smaller and eventually in the limit ~ 0 only
the classical path is left.

1.6 Path integrals and random walks


Now let us think of person who is totally drunk (Captain Haddock, see Fig. 1.25)
and executes a walk, in which each step that he takes is in direction independent
of the direction of the previous step. There exists a very interesting analogy
between his walks and the motion of an electron. Before we go in to a discussion
of this, we wish to stress that while the analogy is very useful and interesting,
physically these are two very different situations and the physics of the two
problems are very different.
Imagine that Haddock goes to his favourite haunt in the evening and has
consumed enough of liquor after which he starts walking. He is so drunk that
he has absolutely no sense of direction; on an average, he takes one step per five
second and that essentially executes a random walk, in which each step that he
3 Note: The classical path is the path that a "Newtonian particle" would follow. That is, the

particle obeys the Newtons equation - force = ma. Very interestingly, for this path the action is an
extreme. This is usually known as the principle of least action, though it would be more correct to
refer to it as principle of stationary action. This principle is equivalent to the Newtons equation of
motion, and if one wishes, one can derive Newtons equation of motion from this principle - see
the beautiful lecture of Feynman on the principle of least action.
1.6 Path integrals and random walks 25

Figure 1.25 Drunken walk of Captain Haddock. He leaves the bar at 10 PM. How to
calculate the probability density of finding him near the second bar at 11 PM?

takes is independent of the previous step. We shall assume that the bar is in an
area where there are no buildings, trees or anything else, obstructing his walk,
except another bar, some what away from the first. We shall also imagine that the
size of the steps that he takes is variable, sometimes he takes bigger steps, while at
other times smaller. Suppose, after one hour, we ask where will he be, assuming
that he continues to be in the same drunken state and has continued to execute
his random walk? The answer is that we cannot say with any definiteness where
he will be. However, we can think of probability density of finding him anywhere,
for example, let us say, at the entrance of the next nearest bar. We denote the
position of the second bar by its position vector r, the first bar being taken as the
origin. r is a two dimensional vector. Clearly, this probability density will depend
on r and the time t . Hence we denote it by P (r, t ). Now, one can think of the
following approach to calculate this probability density. We know that initially he
was at the first bar, when he started. We wish to calculate the probability density
of finding him at the door of the next bar after one hour. To reach this bar, he
could have taken any of the different paths that can be drawn on the plane on
which he walks, with the condition that the paths start at the first bar at time zero
and end at the second at the time t . Then it is clear that some of the paths are
more likely (probable) than others. (For example, a path that wanders a distance
of four km from the two bars is very unlikely). Thus, to each path, it must be
possible to assign a probability density. Its value will depend upon the nature (the
shape of the path and how fast Haddock has to walk along it to reach the second
bar in one hour). That is, the probability density depends upon the nature of the
entire path, which we express by saying that it is a functional of the path, which
we denote as P [pat h]. As in the case of the electron, in section (1.5.1), we can
26 Chapter 1 Wave-particle duality of matter

then write the probability density as

P [pat h].
X
P (r, t ) = (1.23)
al l pat hs

This is completely analogous to what we had earlier, except for two very important
differences: The first is that here, we are directly calculating the probability while
in the case of the electron, we are not calculating probabilities, but wave functions
(which are also referred to as probability amplitudes). This is because the walk of
Haddock is completely classical - Haddock is a classical object. Further, each of
the terms P [pat h] is real and positive as they themselves are probabilities. Hence
one term cannot cancel the other, and they can only add up. Hence Haddocks
probability density would never show any interference patterns. In contrast, for
the electron the term N e i S/~ is a complex quantity and leads to interference
patterns.
This type of motion, usually referred to as Brownian motion is very important
and has been the subject of thorough investigations. For example, a colloidal
particle, observed under a microscope exhibits this kind of motion. The phe-
nomenon of diffusion is a direct result of Brownian motion. If we know where a
colloidal particle is initially, then after some time, we can only prescribe a prob-
ability density of finding it at a point r (this time in three dimensional space,
because a colloidal particle is a drunken walker in three dimensions and r is the
vector that specifiedsthe position of the particle). Further, if we denote this prob-
ability density by P (r, t ), then it can be calculated by precisely the same kind of
formula as in Eq. (1.23). However, there is another approach to the same problem
which has been used for a very long time, and hence is very well understood. That
is to solve the diffusion equation
P (r, t )
= D2 P (r, t ). (1.24)
t
The equation is very simple - it says that the probability density of finding the
particle at r changes at a rate Pt
(r,t )
, because it is diffusing into that region, at rate
2
equal to D P (r, t ). Now, this leads to a very interesting point - the propagation
of a quantum mechanical particle is governed by the wave function, which may
be calculated using the path integral technique. The motion of the random
walker too may be described by the same technique. But in the case of the
random walker, the change in the probability density may be written as a partial
differential equation, which is much easier to analyze than the path integral.
Hence, it is clear that the the change in the wave function of quantum mechanical
particle too is described by a partial differential equation. It would be nice to
know this equation, as it would be easier to handle than the path integral in the
Eq. (1.22). In fact it is possible to start from the Eq. (1.22) and derive such an
equation [2]. Though we do not do this here, it is easy to see by our analogy with
the random walk that it should resemble the Eq. (1.24). The equation is

(r, t ) ~2 2
i~ = (r, t ). (1.25)
t 2m
1.7 The puzzle of quantum mechanics 27

This is the celebrated Schrdinger equation for the motion of a free particle, that
is, one whose potential energy is the same at every point - and as it is the same
everywhere, it is convenient to put it equal to zero. The two equations (1.24)
and (1.25) are obviously very similar. Further, the equation (1.25) describes the
situation where the potential energy of the particle is the same everywhere. If the
potential energy changes with position of the particle, then the equation (1.25)
gets modified to (1.26)

(r, t ) ~2 2
i~ = (r, t ) + V (r)(r, t ). (1.26)
t 2m

Figure 1.26 If only one slit is open electrons would arrive at P while if both slits are
open, no electron would arrive there.

1.7 The puzzle of quantum mechanics


In spite of all the above discussion, quantum mechanics is puzzling to the scientist.
Richard Feynman used to say that "nobody understands quantum mechanics".
To see why, let us go back to the two slit experiment. Imagine that only one slit
is kept open. Then electrons will arrive at the point marked P in Fig. 1.26(a).
However, if one kept both the slits open, none of the electrons will arrive at the
point P, as is obvious from 1.26(b).
Therefore, it is clear that every single electron that passes through the system
knows that both the slits are kept open by the experimenter. Therefore, one says
that the electron goes through both the slits, analogous to what happens in Fig.
1.27, which is impossible with large bodies like us. Thus, the electron in this
experiment, knows that both the slits are open! How does it know that? That is the
puzzle of quantum mechanics! For more detailed and very interesting discussion,
see the book by Feynman on path integrals [2].
28 Chapter 1 Wave-particle duality of matter

Figure 1.27 The behavior of the electron in the two slit experiment is similar to the
marks left by the skier in the figure. For a skier this is impossible, but this is exactly
what the electron does!
1.8 The basis for the whole of chemistry 29

1.8 The basis for the whole of chemistry


We end this chapter with a quotation from the famous physicist Paul Dirac given
below:

Figure 1.28 Quantum physicist, P.A.M. Dirac


30 Chapter 1 Wave-particle duality of matter

[3] [4] [2] [5] [6] [7] [8] [9] [10] [11] [4] [12] [13] [14] [15]
Chapter 3

The Postulates of Quantum


Mechanics


3.1 The Schrodinger Equation
We have discussed the path integral approach to the description of the motion of
a particle in the last section. We make it a little bit more precise. Let us imagine
that the particle is known to be at a postion x i at an initial time t = 0. Note that
for simplicity, we think of a particle moving in one dimension. Then the wave
function at the point x and time t is given by1

(x, t ) = N e i S(x,t ;xi )/~ .


X
(3.1)
al l pat hs

The sum over the paths in the above equation has to include sum over all paths
starting at x i at the time t = 0 and ending at x at the time t . It is rather difficult
to perform this sum over all the possible paths. However there is an alternate
approach to calculate (x, t ), which is very convenient. One can start from
Eq. (3.1) and derive a partial differential equation for (x, t ) (see the book by
Feynman and Hibbs [2] for the derivation). This equation was originally suggested

by Schrodinger and hence is known by his name. It is given below:

(x, t ) ~2 2

i~ = + V (x) (x, t ). (3.2)
t 2m x 2

The operator that occurs on the right hand side is known as the Hamiltonian
operator and is denoted by H , and the equation itself is written as

(x, t )
i~ = H (x, t ). (3.3)
t
The Hamiltonian operator determines how the wave function changes with
time. Generalization of equations (3.1) and (3.3) to more than one dimensions is
1 also depends on x , which we have not written explicitly.
i

31
32 Chapter 3 The Postulates of Quantum Mechanics

straightforward. If x denotes all the position co-ordinates needed to describe the


system, then Eq. (3.1) would become

(x, t ) = N e i S(x,t ;xi )/~


X
(3.4)
al l pat hs

and the equation (3.5) becomes2

(x, t ) ~2 2

i~ = + V (x) (x, t ) (3.5)
t 2m x

where 2x involves differentiation with respect to all the position co-ordinates


that are needed for the system. Having thus seen the physical origin of the

Schrodinger equation, we will no longer adopt the path integral approach to
quantum mechanics, but refer the reader to the book by Feynman and Hibbs [2]
for such an approach. We shall now introduce quantum mechanics as a set of
postulates, as is traditional in the usual approaches to it.

3.2 Postulates of quantum mechanics


3.2.1 Postulate I
The state of a system is specified, as fully as is possible, by the wave function
(x, t ). The probability of finding the system in a volume element d is given by
(x, t )(x, t )d .
The wave function is also known by the name state function. x denotes the
collection of variables (x, y, z..), as many as are needed, to specify the position
co-ordinates of all the particles in the system. Also, d = d xd yd z.... is a small
element of volume in the space of position co-ordinates. As the probability
is given by d where d is a volume element, is referred to as the
probability density.
If the system is one dimensional, then one needs only one co-ordinate x and
hence the state function will depend only on x and the time t . An immediate
consequence of this postulate: The probability that the particle is somewhere in
space is unity and hence
Z
d x (x, t )(x, t ) = 1. (3.6)

For a problem involving several dimensions the condition would be d = 1.


R

Further, for physical reasons one expects to be a continuous single-valued


function of the position co-ordinates (if these are not obeyed, then probability
density would also have the same property and that is unsatisfactory).
2 if the masses associated with all the coordinates are the same and equal to m.
3.2 Postulates of quantum mechanics 33

3.2.2 Postulate II
Corresponding to every observable in classical mechanics there is a linear, Hermi-
tian operator in quantum mechanics. In order to find this operator, write down
the classical mechanical expression for the observable and leave the position co-

ordinates unchanged but replace the momenta p x , p y ... with i ~ x , i ~ y , .....

Operators

To understand the postulate, it is necessary to explain the meaning of several


terms. We start with operators. We shall define an operator A to be a short hand
symbol for a set of well defined operations which when performed on a function
f (x) will give a new function g (x). We will write this as

A f (x) = g (x). (3.7)

Examples of operators are given in the table given below. Here c is assumed to be
a real number.

A A f (x) cx
Ae
d d f (x)
dx dx ce cx
d2 d2
d x2 d x2
c 2 e cx
Multiplication by x x f (x) xe cx
(...)2 ( f (x))2 e 2cx
Taking the absolute value |...| | f (x)| e cx

Table 3.1 Examples of operators.

An operator is defined to be linear if it satisfies

f (x) + g (x)) = A f (x) + Ag


A( (x). (3.8)

for any two arbitrary functions f (x) and g (x). It is clear that the last two operators
listed in the table are not linear operators, while all the others are.
If the effect of an operator A on function is equivalent to multiplying the
function by a constant a, then is said to be an eigenfunction of that operator
and a is referred to as the corresponding eigenvalue. Thus the defining equation
for an eigenvalue a and the associated eigenfunction is

= a.
A (3.9)
2
For example, si n(kx) is an eigenfunction of the operator ddx 2 with an eigenvalue
k 2 for any value of k. It is clear that an operator can have several eigenfunctions
each with its own eigenvalue (in this case infinite number).
34 Chapter 3 The Postulates of Quantum Mechanics

An operator is said to be Hermitian if


Z Z

d A = d A . (3.10)

Hermitian operators have the interesting property that all their eigenvalues are
real. The reason for having Hermitian operators only is that eigenvalues are
results of measurements, which necessarily have to be real (see postulate III).

Observables and Associated Operators

An observable in classical mechanics is defined to be any function of the position


co-ordinates and the momenta. For a one dimensional system, x, p x , the potential
energy V (x), kinetic energy p x2 /2m and total energy p x2 /2m + V (x) are examples
of observables. Expressions for the operators corresponding to these, obtained
following the postulate II are given in the table below. The operator corresponding

Observable Operator

x x

px p x = i ~ x
V (x) V (x)
~2 2
Total Energy, E H = 2m x 2 + V (x)

Table 3.2 Observables and the corresponding operators for a one dimensional system.

to energy is of special importance and is referred to as the Hamiltonian operator,


usually denoted by H . Following the above prescription one can write down the
Hamiltonian for a He atom, which has a nucleus of mass M at R and two electrons
each of mass m at r1 and r2 . It is
~2 ~2 ~2 e2 e2 e2
H = 2R 2r1 2r2 + .
2M 2m 2m 40 |r1 R| 40 |r2 R| 40 |r1 r2 |
(3.11)
The first three terms in H are from the kinetic energies of the three particles and
are usually referred to as the kinetic energy operators.
It is of great importance to note that unlike the classical observables x and p x ,
the corresponding quantum mechanical operators do not commute. That is, the
order in which they operate on a function is important. Thus, x p x 6= p x x as
one can easily verify. In fact explicit calculation gives

x p x p x x = i ~ (3.12)

for any arbitrary function . Hence one has x p x p x x = i ~. We say that x and p x
do not commute. x p x p x x is usually written as [x, p x ] and is referred to as the
commutator of x and p x .
We now discuss some mathematical results which are of great importance.
3.2 Postulates of quantum mechanics 35

Properties of Hermitian operators

1. All eigenvalues of Hermitian operators are real numbers.


This implies
Let us say that a is an eigenvalue of a Hermitian operator A.
that
= a
A (3.13)
where is the eigenfunction associated with the eigenvalue a. Multiplying
the equation (3.13) by d and integrating over the entire space leads to
Z Z
d A = a d

(3.14)

As A is a Hermitian operator, the LHS of the above equation, using equation


((3.10)) can be rewritten as
Z Z
d A = d ( A)
(3.15)

Using the fact that is an eigenfunction of A on the RHS of above equation


leads to Z Z Z
d ( A)
= d ( a) = a d ( ) (3.16)

Comparing equations ((3.14)) and ((3.15)) we get


Z Z

a d ( ) = a d ( ) (3.17)

which means
a = a, (3.18)
implying that a has to be a real number.

2. Eigenfunctions belonging to different eigenvalues are orthogonal


Imagine that we have two eigenfunctions, 1 and 2 having the associated
eigenvalues a 1 and a 2 . This means they obey

1 = a 1 1 ,
A (3.19)

and
2 = a 2 2 .
A (3.20)
Multiplying equation (3.21) with 2 d and integrating over the entire
space gives Z Z
d 2 A
1 = a1 d 2 1 , (3.21)

Using the definition of the Hermitian operator given in equation ((3.10)),


and remembering that a 1 and a 2 are both real, we get
Z Z Z
d 2 A1 = d (2 A1 ) = a 2 d 2 1 .

(3.22)
36 Chapter 3 The Postulates of Quantum Mechanics

On comparing equations (3.21) and (3.22), we get


Z Z
a 1 d 2 1 = a 2 d 2 1 .

(3.23)

If a 1 6= a 2 , the above implies that


Z
d 2 1 = 0. (3.24)

Thus the two functions 1 and 2 are orthogonal.


Thus all eigenfunctions of the operator A having different eigenvalues
are orthogonal. In addition to this the eigenfunctions have an additional
property, which is referred to as completeness. This means that if n with
then, any arbitrary
n = 0, 1, 2, . . . form the complete eigenfunctions of A,
acceptable wave function can be uniquely expanded in terms of them as

= c n n .
X
(3.25)
n

This property is referred to as completeness. The proof this is beyond the


scope of our discussions.

3.2.3 Postulate III


The measurement of an observable will give one of the eigenvalues of the associated
operator as the answer. The average of a large number of measurements may be
calculated from the state function and is given by

d A
R

< A >= R . (3.26)
d

< A > is referred to as the expectation value of A.


This is probably the strangest of
all postulates and has been the subject of much discussion. It forms the subject
of measurement theory on which a lot of work is still being done.

An important consequence

Now imagine that the state of the system is given by the wave function , which
naturally is an acceptable wave function, and we will take it to be normalised, i.e.,
Z
d = 1. (3.27)

As the eigenfunctions of A form a complete set, it is possible to expand as in


equation (3.25). Hence we get
Z
< A > = d ( c n n ) A(
c m m )
X X
n m
X X Z
= c n c m a m d n m (3.28)
n m
3.2 Postulates of quantum mechanics 37

Using the orthonormality of eigenfunctions, we get

< A > = c n c m a m nm
X X
n m
c n c n a n
X
=
|c n |2 a n
X
= (3.29)

The last equality means the following. The average of a large number of measure-
ments, is obtained as the weighted average of all the eigenvalues of the operator A
that is being measured. Remembering that each measurement gives a particular
eigenvalue, this means that |c n |2 is the probability that a given measurement will
give a n as the answer. This is a very important conclusion. This also means that
if the state function is actually an eigenfunction l with an eigenvalue a l , the
all c n are zero except c l which is equal to unity. Therefore, all measurements on
the system will lead to a l with unit probability i.e., all measurements will give the
same answer.

3.2.4 Postulate IV

The state function obeys the time dependent Schrodinger equation

(x, t )
i~ = H (x, t ). (3.30)
t

In the above H is the Hamiltonian operator, and this postulate is the reason why
this operator is the most important operator for a quantum mechanical system.
The Eq. (3.30) is a partial differential equation, first order in time. This means
that if we know the wave function at an initial instant of time, say at t = 0, then
we can solve this equation and find the wave function at all future time. Thus one
(x, 0) is known, (x, t ) can be calculated for any time t .

3.2.5 Notes on the Postulates


1. What we have described are the postulates appropriate for non-relativistic
quantum mechanics. To have spin, one has to combine quantum mechan-
ics and relativity theory and that was done by Dirac. As we do not do this,
spin will have to be added on to the approach as an additional postulate,
which we shall do later.

2. There are three equivalent ways to carry out the process of quantization
-i.e. constructing a quantum mechanical description of the system from the

classical description. The first is due to Schrodinger and is the one that we
have discussed in detail and the result is known referred to as wave mechan-
ics. An alternate approach, due to Heisenberg introduces matrices x and
px corresponding to the observables x and p x obeying the commutation
relationship xpx px x = i ~ I, where I is the identity matrix, and then writes
the Hamiltonian (Energy) as a matrix, and calculates the eigenvalues of
38 Chapter 3 The Postulates of Quantum Mechanics

the matrix. This approach is due to Heisenberg and is referred to as matrix


mechanics. The third approach is due to Feynman and it does not work
with strange beasts like operators or matrices. One simply sums over all
possible paths. All the three approaches are equivalent and lead to the same

final answers. The easiest is the one due to Schrodinger and hence is the
most widely used. Matrix mechanics is described in the book by Jordan[16].

3. A question that usually arises when a person encounters the postulates



for the first time is: why one should have p x = i ~ x ? The reason may
be traced to the fact a wave of well defined momentum may be written as
Ae i px/~ (see section 1.4.2) and the operator p x is such that

p x (Ae i px/~ ) = p(Ae i px/~ ). (3.31)

Thus to extract out the value of the momentum from the wave function,
one has to use the operator p x .3

3.3 Stationary States


In the previous sections we have discussed the idea of standing waves. We have
seen that in general, one can have (a) a time dependent wave whose shape is
changing with time or (b) a wave whose shape is not change and the wave is static
(does not move around in space as time passes). This prompts us to look for
similar things in the realm of matter waves. It turns out that it is possible to find

such solutions to the time dependent Schrodinger equation rather easily, if the
potential V does not depend on the time t . We shall assume this in the following.
We use the ansatz
(x, t ) = (x)T (t ). (3.32)
Note that we are assuming that (x, t ), which is a function of two variables x and
t can be written as a product of two separate functions each depending only on
one variable. This approach is known as the method of separation of variables,
and is of great utility. We substitute the expression in Eq. (3.32) in Eq. (3.5) and
divide through out by (x)T (t ) to get
1 d T (t ) ~2 2

1
i~ = + V (x) (x). (3.33)
T (t ) d t (x) 2m x 2
Note that the left hand side of the above equation is a function of just the variable t
and the right hand side only of x, and that both t and x are independent variables.
Such an equation can be satisfied only if each side of Eq. (3.33) is individually
equal to a constant, which we denote by the symbol E (the reason for calling it E
will become clear shortly). Hence we get
1 d T (t )
i~ =E (3.34)
T (t ) d t
3 This is not a rigorous argument and hence should not be considered as a derivation of the

expression for p x .
3.3 Stationary States 39

~2 2

1
+ V (x) (x) = E (3.35)
(x) 2m x 2
Now we have two separate equations, one for T (t ) and the other for (x), and
both are ordinary differential equations. Hence they are easier to solve than the

original time dependent Schrodinger equation of Eq. (3.30). The equation for
T (t ) is easily solved to get
T (t ) = Ae i E t /~ , (3.36)
with A a constant, and Eq. (3.35) may be rearranged to get

~2 2

+ V (x) (x) = E (x). (3.37)
2m x 2
It is usual to write the above equation as

H (x) = E (x), (3.38)

and refer to this equation as the time independent Schrodinger equation. Note
that this says that (x) is an eigenfunction of H and E is the associated eigenvalue.
Thus we find that
(x, t ) = (x)Ae i E t /~ . (3.39)

is a solution of the original time dependent Schrodinger equation (3.30). We have
seen that the wave function usually is written to within a multiplicative constant.
Eventually we will have to normalise it. Therefore we absorb the constant A in
the above equation into the function (x) and write the solution as

(x, t ) = (x)e i E t /~ . (3.40)

In general, an operator would have several (usually, infinite) eigenfunctions each


with its own eigenvalue. Let us therefore denote the eigenfunctions of the Hamil-
tonian operator as 0 (x), 1 (x), 2 (x)....., with the corresponding eigenvalues
E 0 , E 1 , E 2 ..... Then each product of the form n (x)e i E n t /~ with n = 0, 1, 2... sat-
isfies the Eq. (3.30). Thus we have obtain a solution to this equation, for each
eigenfunction of H .
Further, if the state function is

n (x, t ) = n (x)e i E n t /~ , (3.41)


then this is an eigenfunction of the operator H . Therefore from postulate III, it
follows that the measurement of the energy of the system would always lead to
the same answer, viz., E n .
The probability density of finding the particle at a position x is given by
(x, t ) is given by (x, t )(x, t ). So for a state of the form given in Eq. (3.41) this
becomes

n (x, t )n (x, t ) = n (x)e i E n t /~ n (x)e i E n t /~


= n (x)n (x). (3.42)
40 Chapter 3 The Postulates of Quantum Mechanics

Interestingly, we find the probability density to be independent of time t and thus


it does not change with time, even though, n (x, t ) does change! In addition,
it is easy to prove that < A > is independent of time, if A has no explicit time
dependence. Because of these two results, these states are referred to as stationary
states of the system.
Note that the procedure we have adopted, viz., (x, t ) = (x)T (t ) would work
only if V is function of x alone and not of t (why?).
Chapter 4

Simple Systems

4.1 Continuity of the wave function


In this chapter we illustrate the application of the postulates to some simple

problems. These are problems for which it is possible to solve the Schrodinger
equation exactly.

4.2 The particle in a one dimensional box


We now consider a very simplified model in which we assume that a particle
moving in one dimension and is confined to a finite region of space, and solve

the Schrodinger equation for this exactly. The solution will give insights into the
nature of stationary states. Though simplified, the model has found applications
in a variety of physical applications.
We consider a particle moving in one dimension and denote its position co-
ordinate by x. We assume that it is free to move anywhere in the range (0, L)
and that it cannot enter into the region of space outside this range. For this, we
assume that

V (x) = 0 if x (0, L)
= V0 , otherwise. (4.1)

A schematic representation of V (x) is shown in Fig. (4.1). The Hamiltonian


operator for the system is given by

~2 2
H = + V (x). (4.2)
2m x 2

As H is independent of time, stationary states exist and we look for them. This
means we have to solve the equation

~2 d 2

+ V (x) (x) = E (x). (4.3)
2m d x 2

41
42 Chapter 4 Simple Systems

Figure 4.1

It is clear that the particle will never enter into the region outside the box because
there its potential energy would be very large. This leads to the conclusion that
outside the box, the wave function must be zero, i.e., (x) = 0. Continuity of
the wave function implies that (x) = 0 even at the boundaries of the box. Thus
within the box, we have to find a solution to Eq. ((4.3)) subject to the boundary
conditions

(0) = 0, and (4.4)


(L) = 0. (4.5)

As V (x) = 0 within the box, Eq. ((4.3)) becomes

~2 d 2 (x)
= E (x). (4.6)
2m d x2
We rewrite the above as
d 2 (x)
= k 2 (x), (4.7)
d x2
with
k 2 = 2mE /~2 . (4.8)
The most general solution to this equation is

(x) = A sin(kx) + B cos(kx). (4.9)

Using Eq. ((4.4)), we find


(0) = B = 0. (4.10)
Eq. ((4.5)) leads to
(L) = A sin(kL) = 0. (4.11)
Eq. ((4.11)) has two possible solutions. The first is to have A = 0, which would
imply that (x) = 0 for all values of x, which is not acceptable, as we know that
the particle is inside the box. The other is to have

sin(kL) = 0, (4.12)
4.2 The particle in a one dimensional box 43

which implies that k has to have values that satisfies this equation. Obviously,
this implies
n
k= with n = 0, 1, 2, 3 . . . (4.13)
L
However, of the above possible values of n, n = 0 again would lead to a wave
function that vanishes everywhere and hence n = 0 has to be ruled out. Thus we
have
n
k= , with n = 1, 2, 3 . . . (4.14)
L
If k is not equal to one of the above values, one does not have an acceptable
solution. As k is defined by Eq. ((4.8)), this implies that we have acceptable
solutions only if
k 2 ~2 n 2 2 ~2
E= = . (4.15)
2m 2mL 2
We still have the constant A in Eq. ((4.11)) to be determined. For this we use the
normalization condition Z
d x (x)(x) = 1 (4.16)

As the wave function is zero outside the interval (0, L), the above integral need to
be integrated only in this range. Hence we have
Z L nx
A2 d x sin2 ( ) = 1. (4.17)
0 L

Evaluating the integrals gives


L
A2 = 1. (4.18)
2
This gives r
2
A= . (4.19)
L
The wave function and energy both depend on n, we indicate this explicitly by
r
2 nx
n (x) = sin( ), (4.20)
L L
q
The solution A = L2 of equation (4.18) is equivalent to multiplying n (x) of
equation by 1 and is not an independent solution and hence we do not consider
it.
Thus we find that the eigenvalues (allowed energy levels of the particle) of the
Hamiltonian are determined by the number n. Explicitly, the energy of the n t h
level is given by
n 2 2 ~2
En = . (4.21)
2mL 2
Notice also that the number n has arisen automatically in the process of solution

of the Schrodinger equation, and is referred to as a quantum number.
44 Chapter 4 Simple Systems

4.2.1 Energy levels


The allowed energy levels of the particle are shown in Fig. (4.2). The dotted
line represents the energy E = 0. This is not an allowed energy level. The lowest
allowed energy level is E 1 = 2 ~2 /(2mL 2 ). This energy is known as the zero point
energy, as the particle would possess this much energy even at absolute zero.
The existence of zero point energy is contrary to intuition from classical physics.
Classically, the state of lowest energy would be to have the particle at rest inside

Figure 4.2 The first few allowed energy levels for a particle in a one-dimensional box.
The dotted horizontal line represents the energy E = 0. Note that this is not an allowed
energy level. The lowest allowed energy level is E 1 = 2 ~2 /(2mL 2 ). Notice also that the
spacing between successive energy levels increase as the value of n increases.

the box. As the potential energy within the box is zero, its potential energy would
be zero, and as it is not moving its kinetic energy would be zero. Hence the total
energy would be zero. The reason why zero point energy has to be there can
be traced to the Heisenbergs uncertainty principle. If the particle were at rest
within the box, then its position and momentum would both be zero, and hence
certain, violating the principle. The fact that the particle is known to be within box
implies that x L, which in turn implies that p ~/L, which means energy
of the particle has to be of the order of ~2 /(L 2 2m) which gives a rough order of
magnitude estimate of the zero point energy.

4.2.2 The wave functions and associated probabilities


Fig. (4.3) shows plots of the first few wave functions as a function of x. The points
where the wave function changes sign is a node. One sees that n (x) as n nodes.
More interesting are plots of 2n (x) shown in Fig. (4.4). It is seen that in the
ground state, the most probable position for the particle is x = L/2, i.e. the centre
4.2 The particle in a one dimensional box 45

Figure 4.3 Plots of the wave functions 1 (x), 2 (x) , 3 (x) and 4 (x). The function n
has n nodes.

of the box. A look at the probability distributions, given by 2n (x) in Fig. (4.4)
shows that as the value of n increases, the probability becomes almost uniformly
spread within the box, approaching what one would expect classically.

Figure 4.4 Plots of the probability densities, given by 21 (x), 22 (x) , 23 (x) and 24 (x).
The n t h function has n points where the function becomes zero, within in the box.

Example 4.1
A particle in a box of length L is sitting in the ground state. Imagine measuring
position of the particle and calculating the average of all the measurements made.
What would be the average?

Solution: From our postulate III, it is clear that the average of a large number of
measurements is given by the expectation value of the associated operator. As we
are measuring the position (x-co-ordinate) of the particle, its expectation value
would be given by
Z L
< x >= d x1 (x)x1 (x)
0
46 Chapter 4 Simple Systems

q
2
As 1 (x) = L si n(x/L) we get

2
Z L
< x >= d x x si n 2 (x/L)
L 0

On evaluation, we find
< x >= L/2

Example 4.2 : Electronic transitions in conjugated systems

Figure 4.5 The molecule hexatriene. It has a conjugated system of three double bonds,
and there are six -electrons.

A simple model for the motion of a electron in a linear conjugated molecule


like hexatriene (see Fig. (4.5)) is to assume that the electron moves in a one
dimensional box, typically of the order of the length of the molecule. Calculate the
longest wavelength at which the molecule is expected to absorb electromagnetic
radiation to undergo electronic excitation.

Solution:

Figure 4.6 The allowed energy levels of hexatriene according to the particle in a box
model. The six electrons occupy the first three allowed energy levels. It has a conju-
gated system of three double bonds, and there are six -electrons.
4.3 Particle in a three dimensional box 47

If one uses the particle in box model, the allowed energy levels are given by E n =
n 2 2 ~2
2mL 2
The allowed energy levels of the system and the way electrons are distributed
are shown in Fig. (4.6). The transition that needs the smallest energy is from n = 3
to n = 4 and is indicated by the blue dotted arrow. The energy change in the
transition would be
~2 2 2 2
E = E 4 E 3 = (4 3 )
2mL 2
The frequency of the radiation that would bring about the transition is given by

E 7h
= =
h 8mL 2
We take the C = C bond length to be 147pm and C C to be 154pm, from which
we estimate the length of the box to be L = (3 147 + 2 154)pm = 749pm. Hence

7 6.63 1034 J .s
= = 1.13536 1015 /s,
8 9.108 1031 kg (749 1012 m)2

and the corresponding wave length is

2.998 108 m/s


= c/ = = 2.640 107 m = 264nm.
1.13536 1015 /s
In comparison, the experimental value is

= 263nm.

4.3 Particle in a three dimensional box


We now consider a generalisation of the above problem where we take the particle
to be moving in three dimensions, and is confined to a rectangular box of sides
L 1 , L 2 and L 3 (see Fig. (4.7)). The Hamiltonian operator for the system is

~2
H = 2 + V (x, y, z)
2m
We choose

V (x, y, z) = 0, if 0 < x < L 1 and 0 < y < L 2 and 0 < z < L 3 , (4.22)
= otherwise. (4.23)

The fact that the potential is infinitely large outside the box will confine the
particle to within the interior of the box.
The time dependent wave function (x, y, z, t ) of the particle is a function of
three position co-ordinates. As the Hamiltonian is time independent, one can
find stationary states of the form (x, y, z, t ) = (x, y, z)e i E t /~ where obeys the
eigenvalue equation
H (x, y, z) = E (x, y, z). (4.24)
48 Chapter 4 Simple Systems

Figure 4.7 Box having sides of length L 1 , L 2 and L 3 . The potential V (x, y, z) is zero
inside the box and infinity outside. This can be thought of as a model for electrons
within a piece of metal.

As the particle is confined to the inside of the box, the probability of finding it
outside the box is zero, which implies that the wave function outside the box is
zero. As the wave function is a continuous function, this implies that the wave
function must be zero on the walls of the box. Thus we have

(x, y, z) = 0, on the walls of the box (4.25)

As the box has eight walls, this actually implies eight different conditions. We
write two of them:

(0, y, z) = 0 and (L 1 , y, z) = 0 (4.26)

Within the box, V (x, y, z) = 0 and hence the equation (4.24) becomes
~2
2 (x, y, z) = E (x, y, z) (4.27)
2m
We now solve the equation (4.27) subject to the conditions of equation (4.25), we
adopt the method of separation of variables and put

(x, y, z) = X (x)Y (y)Z (z) (4.28)

where X (x), Y (y) and Z (z) are to be determined. Substituting (x, y, z) from
equation (4.28) into the equation(4.27), we get
~2 2 2 2

+ + X (x)Y (y)Z (z) = E X (x)Y (y)Z (z) (4.29)
2m x 2 y 2 z 2
4.3 Particle in a three dimensional box 49

2
Remembering that x 2 affects only functions that depend on x and similarly for x

and y co-ordinates, we get

~2 d 2 X (x) d 2 Y (y) d 2 Z (z)



Y (y)Z (z) + X (x)Z (z) + X (x)Y (y) = E X (x)Y (y)Z (z)
2m d x2 d y2 d z2
(4.30)
Dividing throughout by X (x)Y (y)Z (z) gives

~2 1 d 2 X (x) ~2 1 d 2 Y (y) ~2 1 d 2 Z (z)


=E (4.31)
2m X (x) d x 2 2m Y (y) d y 2 2m Z (z) d z 2

Of the three terms in the left hand side the first depends only on x, the second
only on y and the third on z. The sum of the three is equal to a constant, E . As
x, y and z are independent varaibles, such an equation can be satisfied only if
each of them separately, equal to constants. Thus we get

~21 d 2 X (x)
= E1 (4.32a)
2m X (x) d x 2
~2 1 d 2 Y (y)
= E2 (4.32b)
2m Y (y) d y 2
~2 1 d 2 Z (z)
= E3 (4.32c)
2m Z (z) d z 2
and further, using (4.32) in equation (4.31) leads to

E1 + E2 + E3 = E (4.33)

We now solve the equation (4.32a). As we are considering the region within the
box, this equation is applicable in the region 0 < x < L 1 . The boundary conditions
(4.26) imply that
X (0) = 0 and X (L 1 ) = 0. (4.34)
Equations (4.32a) and (4.34) are simple solve, exactly as in the case of particle in a
one dimensional box. Acceptable solutions exist only if

n 12 2 ~2
E1 = , n 1 = 1, 2, . . . (4.35)
2mL 21

and is given by s
2 n 1 x
X n1 (x) = sin( ) (4.36)
L1 L1
The equations for Y (y) and Z (z) can be solved in an exactly similar fashion and
gives
n 2 2 ~2
E2 = 2 2 , n 2 = 1, 2, . . . (4.37)
2mL 2
s
2 n 2 y
Yn2 (x) = sin( ) (4.38)
L2 L2
50 Chapter 4 Simple Systems

n 32 2 ~2
E3 = , n 3 = 1, 2, . . . (4.39)
2mL 23
s
2 n 3 x
Zn3 (x) = sin( ) (4.40)
L3 L3
Using all the above in equations we get

2 ~2 n 12 n 22 n 32
!
E n1 n2 n3 = + + (4.41)
2m L 21 L 22 L 23

and is given by
s
8 n 1 x n 2 y n 3 z
n1 n2 n3 (x, y, z) = sin( ) sin( ) sin( ) (4.42)
L1L2L3 L1 L2 L3

In the above, the energy E and the wave function depend on three quantum
number n 1 , n 2 and n 3 which we have indicated by adding the subscript n 1 n 2 n 3
to them.
The ground state of the system would be obtained if n 1 = n 2 = n 3 = 1 and
would have an energy !
2 ~2 1 1 1
E 111 = + + (4.43)
2m L 21 L 22 L 23
The associated wave function is
s
8 x y z
111 (x, y, z) = sin( ) sin( ) sin( ) (4.44)
L1L2L3 L1 L2 L3

It is clear that within the box this function never becomes zero, but approaches
zero on the walls of the box.

Figure 4.8 Contour plot of the wave function for the ground state 111 . The surface is
obtained by joining together points in space at which 111 (x, y, z) = 0.1max , where
p
max = maximum possible value of the function, equal to 8/L 1 L 2 L 3
4.4 Cubic box: degeneracy 51

Figure 4.9 Contour plot of the wave function 211 . The surface is obtained by joining
together points in space at which 211 (x, y, z)| = 0.1max , indicated by blue color and
points at which 211 (x, y, z)| = 0.1max , indicated by orange color. It is clear that the
function has one node, at x = L 1 /2, which is a plane.

Figure 4.10 Contour plot of the wave function 121 . The surface is obtained by joining
together points in space at which 121 (x, y, z)| = 0.1max , indicated by blue color and
points at which 121 (x, y, z)| = 0.1max , indicated by orange color. It is clear that the
function has one node, at y = L 2 /2. The node is a plane.

4.4 Cubic box: degeneracy


We now consider the energy levels of a cubic box, for which L 1 = L 2 = L 3 = L.
Then the allowed energy levels are

~2 2
n 12 + n 22 + n 32

E n1 n2 n3 = (4.45)
2mL 2
~ 2 2
6~ 2 2
The lowest energy level is E 111 = 32mL 2 The next energy level is E 211 = 2mL 2 , but

then one realises that E 211 = E 121 = E 112 , and that the wave functions 211 , 121
and 112 are different from one another. Thus one has three different allowed
states, all having the same energy. We say that this energy level is triply degenerate.
52 Chapter 4 Simple Systems

Figure 4.11 Contour plot of the wave function 112 . The surface is obtained by joining
together points in space at which 112 (x, y, z)| = 0.1max , indicated by blue color and
points at which 112 (x, y, z)| = 0.1max , indicated by orange color. It is clear that the
function has one node, at z = L 3 /2. The node is a plane.

Figure 4.12 Contour plot of the wave function 221 . The surface is obtained by joining
together points in space at which 221 (x, y, z)| = 0.1max , indicated by blue color and
points at which 221 (x, y, z)| = 0.1max , indicated by orange color. It is clear that the
function has two nodes, at x = L 1 /2 and y = L 2 /2. The nodes are planes.

The Fig. (4.16) indicates the degneracies of the first six levels of a particle in a
cubic box.

Example 4.3
A cubic box is distorted in the z-direction by 10%. What happens to the degenera-
cies of the allowed energy levels, shown in Fig. (4.16)?

Solution: Putting L 1 = L 2 = L and L 3 = 1.1L in equation (4.41) we find the allowed


4.4 Cubic box: degeneracy 53

Figure 4.13 Contour plot of the wave function 222 . The surface is obtained by joining
together points in space at which 222 (x, y, z)| = 0.1max , indicated by blue color and
points at which 222 (x, y, z)| = 0.1max , indicated by orange color. It is clear that the
function has three nodes, at x = L 1 /2 (plane), at y + L 2 /2 (plane) and at z = L 3 /2 (plane).

Figure 4.14 Contour plot of the wave function 222 . The surface is obtained by joining
together points in space at which 222 (x, y, z)| = 0.1max , indicated by blue color and
points at which 222 (x, y, z)| = 0.1max , indicated by orange color. It is clear that the
function has three nodes, at x = L 1 /2 (plane), at y + L 2 /2 (plane) and at z = L 3 /2 (plane).

energy levels to be given by

n 32
!
~2 2
E n1 n2 n3 = n 12 + n 22 + (4.46)
2mL 2 1.21
~2 2
n 12 + n 22 + 0.826446n 32

=
2mL 2
The energy levels can now be calculated using this equation and the result is shown
in figure (4.15)
54 Chapter 4 Simple Systems

Figure 4.15 Allowed energy levels of a particle in a distorted box. The cubic box is a
highly symmetric object (its point group is O h ). Distorting it in the z-direction reduces
the symmetry to point group D 4h . As a consequence energy levels which were triply de-
generate in O h (for example, 211, 121, 112) split up to two doubly degenerate (211, 121)
and and one singly degenerate level (112). This is similar to what happens in metal
complexes where orbitals that are degenerate are split up when symmetry is lowered.

4.5 Two Systems that do not interact


We now discuss a very interesting result, of great utility. Imagine that one has two
systems which do not interact with each other (see Fig. (??)). For simplicity, we
imagine each to consist of only one particle, whose position can be specified by a
single co-ordinate. Thus, for the first system, position co-ordinate is x, mass m 1

Figure 4.16

and it moves under a potential U (x). Thus its Hamiltonian would be

~2 2
H 1 (x) = +U (x).
2m 1 x 2
4.5 Two Systems that do not interact 55

In an exactly similar fashion the Hamiltonian for the second system would be

~2 2
H 2 (y) = + V (y).
2m 2 y 2

H 1 (x) has the eigenfuctions X 0 (x), X 1 (x), X 2 (x) . . . with energies E 01 , E 11 , E 21 , . . .,


and H 2 (y) has the eigenfunctions Y0 (y), Y1 (y), Y2 (y) . . . with energies E 02 , E 12 , E 22 , . . ..
Now think of the composite system, consisting of the both the particles. Imagine
that the first system is in the stationary state X m (x) and that the second is in the
state Yn (y). Obviously, physical intuition tells us that the energy of the compos-
1
ite system must be E m + E n2 . What would be the wave function (x, y) for the
composite system? The answer is

(x, y) = X m (x)Yn (y)!

The proof of this is easy. All that one needs to note are:

1. As the systems do not interact, the total Hamiltonian H (x, y) is the sum of
the two individual Hamiltonians. Thus

H (x, y) = H 1 (x) + H 2 (y).

2. The functions X m (x) and Yn (y) obey


1
H 1 (x)X m (x) = E m X m (x)

and
H 2 (y)Yn (y) = E n2 Yn (y).

Now letting H operate upon the product X m (x)Yn (y) and noting that H 1 (x) will
affect only X n (x) and H 2 (y) only Yn (y), we find that the product X m (x)Yn (y) is
an eigenfunction of the total Hamiltonian H with the eigenvalue E m 1
+ E n2 . Thus
we find that for two non-interacting systems, the energy is the sum of the two
individual energies, while the wave function is the product.
56 Chapter 4 Simple Systems

(a) 111 (x, y, z). It has no nodes (b) 211 (x, y, z). Node at x = L 1 /2.

(c) 121 (x, y, z). Node at y = L 2 /2. (d) 112 (x, y, z). Node at z = L 3 /2.

(e) 211 (x, y, z). Nodes at x = L 1 /2 (f ) 121 (x, y, z). Nodes at x = L 1 /2,
and y = L 2 /2. y = L 2 /2 and z = L 3 /2

Figure 4.17 Contour surface plots of the wave functions. The surface is obtained by
joining together points in space at which wave function = 0.1max (shown in blue), or
points at which the wave function = 0.1max (shown in orange). max = maximum
p
possible value of the function, equal to 8/L 1 L 2 L 3 . This means that in the blue region
the wave function is positive and in the orange region it is negative. 211 , 121 and 112
have one node each, which are planes, indicated by green planes in the figure. 221 has
two and 222 has three nodes.
Chapter 5

The Hydrogen Atom


5.1 The Schrodinger Equation
The hydrogen atom has two particles (see Fig. 5.1). So one has a Hamiltonian that
contains six position co-ordinates, viz, the co-ordinates of the electron as well as
that of the proton. The Hamiltonian is given by

~2 ~2 e2
H t ot al = 2N 2e , (5.1)
2M 2m e 40 r

where 2N will involve differentiation with respect to the co-ordinates of the


proton (X N , Y N , Z N ) and M denotes its mass and 2e will involve differentiation
with respect to the co-ordinates of the electron (x e , y e , z e ) and m e denotes its
mass. We now have to solve for the wave function of the system t ot al (RN , re )
which obeys
H t ot al t ot al (RN , re ) = E t ot al t ot al (RN , re ). (5.2)
RN and re denote the position vectors of the nucleus and the electron.
The equation looks formidable to solve. However, physical intuition helps.
We are interested in the situation where the two particles form a bound state

Figure 5.1 The position vectors of the nucleus RN and the electron, re with respect to
the origin O. rr el also is shown.

57
58 Chapter 5 The Hydrogen Atom

resulting in an atom. The atom will execute translational motion as a whole


and naturally the mass associated with this would be the total mass of the atom,
equal to M + m e . In addition to this, there will be relative motion which is more
interesting to the chemist, as this determines electronic structure of the atom. In
relative motion, the motion takes place in such a fashion that the centre of mass
of the system does not move. Therefore, both the particles are involved in the
motion and the mass associated with this motion would be a combination of the
masses of the two particles. In fact, proper mathematical formulation leads to
the mass associated with relative motion to be = mmee+M M
. This is known as the
reduced mass of the system. Note that for the Hydrogen atom, as M >> m e , this
is very close to m e . The Hamiltonian for the system may thus be written as

~2 ~2 e2
H t ot al = 2cm 2r el . (5.3)
2(M + m e ) 2 40 r

2cm is the kinetic energy of the centre of mass motion and hence involves dif-
ferentiation with respect to the co-ordinates of the centre of mass. 2r el involves
differentiation with respect to the relative co-ordinates. The most important
thing about this Hamiltonian is that it is the sum of the Hamiltonians for two
non-interacting systems,

~2
H cm = 2cm ,
2(M N + m e )

and
~2 e2
H r el = 2r el .
2 40 r
Therefore, following the section 4.5 we can write the wave function as a product:

t ot al = cm (Rcm )r el (rr el ),

and proceed to separate the variables. Then we get the equation

~2
2cm cm (Rcm ) = E cm cm (Rcm ) (5.4)
2M
which is the Schrodinger equation for a free particle of mass M + m e . As this
simply describes the translational motion of the Hydrogen, and as one can solve
it easily if one assumes that the hydrogen atom is confined to a three dimensional
box (the container), we shall not discuss this equation further.
It is r el (rr el ) that determines how the electron is distributed around the
nucleus and our interest is on this function. It obeys
2
e2

~ 2
r el r el (rr el ) = E r el r el (rr el ). (5.5)
2 40 r

It should be noted that E t ot al = E cm +E r el . In relative motion, one does not expect


the proton to move much as it is 1836 times heavier than the electron. Therefore,

5.1 The Schrodinger Equation 59

Figure 5.2 The Cartesian co-ordinates, (x, y, z) of a point P in space.

it is a very good approximation to neglect the motion of the proton and describe
the motion of the electron with respect to it. This means that we approximate
by m e . rr el is the position vector of the electron with respect to the nucleus. So
we put the proton at the origin of a co-ordinate system (see Fig. 5.2) and denote
the electron by its position vector r (From now on, we will omit the subscript

rel"). We will also drop the subscript of m e . Then the Schrodinger equation for
the internal dynamics of the hydrogen atom is

~2 2 e2

(r) = E (r). (5.6)
2m 40 r

r is the distance of the electron from the proton.


This equation, as it stands cannot be solved if one used Cartesian co-ordinate
system. The reason is that Cartesian co-ordinate system cannot properly handle
systems which have a spherically symmetric potential, like the problem here. In
such a case, it is convenient to change over to spherical polar co-ordinates (see
Fig. 5.3). The three spherical polar co-ordinates are related to the Cartesian ones
by

x = r sin() cos() (5.7)


y = r sin() sin() (5.8)
z = r cos() (5.9)

To cover the entire space it is enough if one lets r [0, ), [0, ] and
[0, 2). Also it is important to note that + 2 is equivalent to as both
correspond to teh same point.
Written in spherical polar co-ordinates, the Hamiltonian is given by

~2 1 2 2 e2

1 1
H = r + sin + (5.10)
2m r 2 r r r 2 sin r 2 sin2 2 40 r
60 Chapter 5 The Hydrogen Atom

Figure 5.3 The polar co-ordinates (r, , ) of a point P in space.

After this change over to spherical polar co-ordinates, one can solve the Eq. (5.6)
by the substitution
(r, , ) = R(r )()(), (5.11)
and using the condition that the function should be normalized. i.e.,
Z
d (r, , )(r, , ) = 1. (5.12)

We briefly sketch the derivation. On putting (5.11) in to equation (5.6) and dividing
through out by R(r )()(), we get

~2 d 2 d R(r ) d d () 1 d 2 () e2

1 1 1
r + sin + =E
2m R(r )r 2 d r dr r 2 () sin d d r 2 sin2 () d 2 40 r
(5.13)
Note that the portion written in red depends only on and that no other term
depends on . Hence one may imagine rearranging the equation such that only
this term is left on the LHS and all other terms are on teh RHS. This would mean
that a function of alone, is equal to a function of r and . As r, and are
independent variables, this means that

1 d 2 ()
= m 2 , (5.14)
() d 2

where m 2 (not to be confused with the mass of the electron!) is a constant. Using
(5.14) back in (5.13), we get

~2 d 2 d R(r ) 1 d d () m2 e2

1 1
r + sin =E
2
2m R(r )r d r dr r () sin d
2 d sin
2 40 r
(5.15)
In the above equation, only the portion written in red colour is dependent on .
Hence this must be equal to a constant, which we denote as . Thus

1 d d () m2
sin = (5.16)
() sin d d sin2
5.2 Solution of the equation for () 61

which may be rearranged to get the differential equation


1 d d () m2

sin + () = 0 (5.17)
sin d d sin2
Using equation (5.16) back in equation (5.18), we get
~2 d 2 d R(r ) e2

1
r =E (5.18)
2m R(r )r 2 d r dr r2 40 r
which on rearrangement gives
~2 1 d 2 d R(r ) e2

r R(r ) R(r ) = E R(r ) (5.19)
2m r 2 d r dr r2 40 r
Thus we have found three ordinary differential equations for R(r ) (5.19), ()
(5.17) and () (5.14). One now has to solve these equations, and get solutions
that are acceptable.

5.2 Solution of the equation for ()


The equation (5.14) is a simple partial differential equation. The solution may be
taken to be
() = Ae i m (5.20)
We now evaluate the value of () on any point with = 0, we get

(0) = A. (5.21)

Any point with = 0 can equally well be said to have = 2. This means that

(0) = (2). (5.22)

If this condition is not satisfied, the function () will be a discontinuous function.


Using the conditions leads to the result

e i m2 = 1.

Thus m cannot be any arbitrary constant. It can have only values that satisfy the
above equation, which are

m = 0, 1, 2, 3 . . . (5.23)

We now determine the constant A such that the function () satisfies the nor-
malisation condition Z 2
d ()() = 1
0
and this gives
1
A= p .
2
Thus we get
1
m () = p e i m . (5.24)
2
62 Chapter 5 The Hydrogen Atom

5.3 The functions () and R(r )


We will not go into the details of the solution process, but only state the results.
The equation (5.17) can be solved for any value of the constant . However, when
one puts the condition that the function that results should be acceptable, one
finds that cannot be arbitrary. It is constrained to be given by

= l (l + 1)

where l is quantum number that can take the values 0, 1, 2 . . .. Further, it is found
that l |m|. The function () is found to depend on the values of l and |m| and
is given by
2l + 1 (l |m|)! 1/2 |m|

l |m| () = (1)|m| P l (cos) (5.25)
2 l + |m|)!
where P l|m| () is known as an associated Legendre function, one of the special
functions of mathematics. These are chosen so as to satisfy the normalisation
condition Z
|m 1 |
d sin P l|m| () P l () = l l 1 |m||m1 |
1
0
. l l 1 is the Kronecker delta function. Details may be found in the book by
Arfken[?].
The equation for R(r ) (5.19) may be solved for any energy E , but when one
puts the acceptability condition on the result, one realises that acceptable solu-
tions exist only for certain values of energy, given by

e2

1
En = , (5.26)
n 2 80 a 0

The solutions turn out to be functions well known to mathematicians as associ-


ated Laugerre functions and are given by
3 1/2
(n l 1)!

2
R nl (r ) = e /2 l L 2l +1
n+l () (5.27)
na 0 2n[(n + l )!]3

where L 2l
n+l
+1
() are the associated Laguerre polynomials [?]. In the above,

2
= r (5.28)
na 0

A list of the radial functions R nl (r ) is given in the Table ??.

5.4 The full wave function


From the above analysis, it is clear that the function (r, , ) depends on the
three quantum numbers n, l and m and may be written as

nl m (r, , ) = R nl (r )Yl m (, ). (5.29)


5.4 The full wave function 63

Figure 5.4 The allowed energy levels for the electron in the Hydrogen atom. The n t h
level has a degeneracy of n 2 . That is, there are n 2 states, all having the same energy.

In the above, the angular part of the function is defined by

Yl m (, ) = l |m| m () (5.30)

and are known as the spherical harmonics. A list of these functions for different
values of l and m is given in the Table 5.2.
The energy of nl m (r, , ) depends only on the principal quantum number
n. These energies are shown in Fig. 5.4. It is obvious that the system has the
lowest possible energy when n = 1 and this we refer to as the ground state of
the hydrogen atom. n = 2 would be the next possible energy level, and there are
four states (wave functions) corresponding to this energy. Thus we say that n = 2
energy level is four fold degenerate while n = 1 is non-degenerate. If n = 3, one
has 9 different possible wave functions. Each one of these wave functions are
referred to as an atomic orbital and represents a possible stationary state of the
electron in the atom.
64 Chapter 5 The Hydrogen Atom

n l R nl
1/2
1
1 0 a 03
2e
1/2 /2
1 e (2)
2 0 a 03
p
2 2
1/2
2 1 1 p
e /2
a 03 2 6
1/2
1 2p /3
22 18 + 27

3 0 a 03
e
81 3
1/2 q
1 1 2 /3
3 1 a 03 27 3e (6 )
1/2 q
1 2 2 /3 2
3 2 a 03 81 15 e

Table 5.1 The radial functions. = r /a 0

l m Yl m (, )
1
0 0 2
p

q
1 i 3
1 1 2e q 2 sin()
1 3
1 0 2 q cos()
1 1 21 e i 2 3
sin()
q
1 2i 15 2
2 2 4e q 2 sin ()
1 i 15
2 1 2e q 2 cos() sin()
1 5
3 cos2 () 1

2 0 4
q
2 1 12 e i 2 15
cos() sin()
q
1 2i 15
2 2 4e sin2 ()
q2
1 3i 35 3
3 3 8e q sin ()
1 2i 105 2
3 2 4e q 2 cos() sin ()
1 i 21 2

3 1 8e q 5 cos () 1 sin()
1 7 3

3 0 4 q 5 cos () 3 cos()
18 e i 21 2

3 1 5 cos () 1 sin()
q
1 2i 105
3 2 4e 2q cos() sin2 ()
3 3 18 e 3i 35 3
sin ()

Table 5.2 The angular part Yl m (, )


5.5 Shapes of atomic orbitals 65

Figure 5.5 One finds incorrect definition of the atomic orbital in many places. An
atomic orbital is not a region of space. It is just the wave function for the electron and
occupies some region of space.

5.5 Shapes of atomic orbitals


In this section, we analyze the atomic orbitals in detail.

5.5.1 The 1s atomic orbital


We now consider the case lowest possible state of the system, i.e., the ground state
of the hydrogen atom. The principal quantum number is n = 1 and this implies
that l = 0 and m = 0. Thus the function is

100 (r, , ) = R 10 (r )Y00 (, ). (5.31)

Using the tables, one finds that is dependent only on the radial co-ordinate r .
Thus !1/2
1
100 (r ) = e r /a0 (5.32)
a 03
As the function has no dependence upon the angles, it is spherically symmetric.
The angular part is Y00 (, ) = p1 . We make a polar plot to make the angulare
4
dependence clear. To make the polar plot, we note that each direction in space (i.e.
a line passing through the origin) has its on value of and . In that direction, we
put a point whose distance from the origin is equal to |Yl m (, )|. The result, for
this orbital is the surface of a sphere of radius p1 , as shown in Fig. 5.6. A plot of
4
R 10 (r ) is shown in the Fig.5.7, which shows that the orbital has its maximum value
at the origin (at the nucleus) and as one goes away, it decreases exponentially. In
principle, the function becomes zero at only. However, one can ask, what is the
region of space, where the electron will be found with a probability of 0.99? This
can be calculated, and is found to be the region within a sphere of radius 4.205a 0
(approximately 2).
66 Chapter 5 The Hydrogen Atom

Figure 5.6 Polar plot for the 1s atomic orbital

What is the probability that the distance of the electron from the nucleus lies
between r and r + d r ? This will happen if the electron is within a spherical shell
between the sphere of radius r and a sphere of radius r +d r . As d r is very small, we
can approximate the volume of the spherical shell by 4r 2 d r . Hence the required
probability is given by 4r 2 2 (r )d r which we write as P (r )d r . P (r ) = 4r 2 2 (r )
gives one an idea as to how probable each distance r is. It is known as the radial
distribution function. We give a plot of P (r ) against r in Fig. 5.8. Obviously there

Figure 5.8 Plot of the radial distribution


function 4r 2 R 10
2
(r ) against r . Units in
Figure 5.7 Plot of R 10 (r ) against r . Units in which a 0 = 1 are used. It has a maximum at
which a 0 = 1 are used. a 0 = 0.52917. r = 1, at a 0 .

is a most probable distance, which can be calculated using calculus. One gets the
distance to be a 0 . Thus, while in Bohr theory, the electron is at a distance a 0 from
the nucleus, according to quantum mechanics, any distance is possible for the
electron; the most probable distance is equal to a 0 .
5.5 Shapes of atomic orbitals 67

Figure 5.9 A plot of (x, y, 0) against (x, y) for the 1s atomic orbital. The orbital has its
maximum value at the nucleus and hence the plot has a single peak at this point. Note
the circular symmetry in the plot, which is a consequence of the spherical symmetry of
the orbital.
68 Chapter 5 The Hydrogen Atom

5.5.2 The 2s atomic orbital


When n = 2, there are four possible atomic orbitals. The first one is the one with
n = 2, l = 0 and m = 0, which is given by

20 (r, , ) = R 20 (r )Y00 (, ). (5.33)

We have already encountered the function Y00 (, ). It is equal to p1 and its


4
polar plot is the surface of a sphere of radius p1 .
Further, we note that for any
4
wave function with l = 0, m has to be equal to zero and then the angular part
of the function would be Y00 (, ). This means that whenever l = 0, the orbital
would be spherically symmetric.
Using the tables we find
!1/2
1
200 (r ) = e r /2a0 (2 r /a 0 ). (5.34)
32a 03

A plot of R 20 (r ) against r is shown in Fig. 5.10. Note that the function becomes
zero when r = 2a 0 , implying that there is a radial node. Thus the surface of
a sphere of radius 2a 0 is a nodal surface. Outside this surface, the function is
negative, as is seen in Fig. 5.12. Radial distribution function for this orbital is
shown in Fig. 5.11, which shows that the function has two maxima, and the
second one is more probable than the first.

Figure 5.11 Plot of the radial distribution


function 4r 2 R 20
2
(r ) against r . Units in
Figure 5.10 Plot of R 20 (r ) against r . which a 0 = 1 are used. It has two maxima.
5.5 Shapes of atomic orbitals 69

Figure 5.12 The 2s atomic orbital. This is a plot of (x, y, 0) against (x, y). The orbital
has its maximum value at the nucleus. So the plot is peaked at this point. For the sake
of showing the negative part of the plot clearly, the top portion had to be clipped off.
Distances are given in units of a 0 , the radius of the first Bohr orbit.

5.5.3 The 2p orbitals


When n = 2, l can take the value 1 too. If l = 1, there are three possible atomic
orbitals, having m = 0, 1 and +1. From the expression 21m (r, , ) = R 21 (r )Y1m ,
it is clear that all these orbitals have the same radial dependence. A plot of R 21 (r )
against r is shown in Fig. 5.22. Interestingly, the value of the wave function is zero
at the origin. Hence the density of the orbital at the nucleus is zero. This is quite
general. It is only the s-orbitals that have a non-zero density at the nucleus.

5.5.4 The 2p z orbital


We now look at m = 0. For this orbital

210 (r, , ) = R 21 (r )Y10 (, ). (5.35)

The angular part Y10 (, ) is not a constant. From the table, we get Y10 (, ) =
3 1/2
4 cos(). In order to understand this function, we construct a polar plot. We
first do it in the YZ plane. In this plane, we consider each possible and for that
direction, we find a point whose distance from the origin is |Y10 (, )|. Joining
70 Chapter 5 The Hydrogen Atom

Figure 5.14 Plot of the radial distribution


function 4r 2 R 21
2
(r ) against r . Units in
Figure 5.13 Plot of R 21 (r ) against r . which a 0 = 1 are used. It has two maxima.

Figure 5.15 Polar plot of p z orbital in the YZ plane. Z axis is the vertical axis. To get the
three dimensional appearence, one has to rotate this figure about the Z-axis.

together all these points leads to the result shown in Fig. 5.16. A nice way of
representing both the radial and angular dependence is to make a contour plot
of the orbital, with colors added, so that one can get an idea of the value of the
orbital too. Such a plot in the YZ plane is shown in Fig. 5.17. In this figure, each
contour joins together points having the same value for the wave function. It is
clear that this orbital resembles a dumbbell and is pointed along the Z-axis. This
orbital is referred to as the 2pz atomic orbital. It is useful to note the shape of this
orbital comes from the factor cos() which may be written as z/r .

The Angular Part for m = 1

From the tables, r


3 i
Y1,1 = e (5.36)
8
Thus the angular parts of these two are complex. There is nothing wrong with
having an atomic orbital that is complex as the probability is determined by ,
5.5 Shapes of atomic orbitals 71

Figure 5.16 Polar plot of p z orbital. It consists of surfaces of two spheres whose centers
are along the Z-axis, and touching each other at the origin. In the polar plots, blue
colour is used to denote the positive lobe, and orange colour for the negative lobe of
the orbital.

which is always real. However, it is usual to construct two new orbitals, that are
real, from these two. The reason why one is allowed to do this is that the two
orbitals have the same energy. (Note that if there are two wave functions (1
and 2 ), having the same energy, then any linear combination of the two, of the
form c 1 1 + c 2 2 is also a solution of the Schrodinger
equation, having the same
energy. c 1 and c 2 are any two arbitrary constants. Hence we take
r
3 1
sin() cos() = p Y1,1 (, ) + Y1,1 (, )

Y1c = (5.37)
4 2
as one new orbital and the other as
r
3 1
Y1,1 (, ) Y1,1 (, ) .

Y1s = sin() sin() = p (5.38)
4 2i
With these as the angular parts, one has two new orbitals

21c (r, , ) = R 21 (r )Y1c (, ) (5.39)


72 Chapter 5 The Hydrogen Atom

Figure 5.17 Contour Plot of p z orbital.

(a) p x , YZ plane is a node (b) p y , XZ plane is a node.

Figure 5.18 Polar plots of the p x and p y -orbitals. Each has one node.

and
21s (r, , ) = R 21 (r )Y1s (, ). (5.40)
Obviously they are real, and further they are orthogonal to each other as well as to
2p z . The construction of polar plots for these orbitals is easy, when one realizes
that their angle dependent parts are sin() cos() = x/r and sin() sin() = y/r .
Remember that angular part for the 2p z is z/r . Hence the polar plots for these
orbitals are similar to that of 2p z except that the lobes are along the X- and Y-axes
respectively.
5.5 Shapes of atomic orbitals 73

5.5.5 The 3s atomic orbital


Just like the 2s orbital, this atomic orbital is spherically symmetric. Plots of R 30
and 4r 2 R 30
2
(r ) are shown in Fig. 5.19 and 5.20 respectively. The Fig. 5.21 shows a

Figure 5.20 Plot of the radial distribution


Figure 5.19 Plot of R 30 (r ) against r . function 4r 2 R 30
2
(r ) against r .

density plot for the 3s orbital. It also shows two circles which are the sections of
the nodal surfaces.

Figure 5.21 The 3s orbital. It has two nodes, and they are represented by the black
circles in the figure.
74 Chapter 5 The Hydrogen Atom

Figure 5.23 Plot of the radial distribution


function 4r 2 R 21
2
(r ) against r . Units in
Figure 5.22 Plot of R 21 (r ) against r . which a 0 = 1 are used. It has two maxima.
5.5 Shapes of atomic orbitals 75

5.5.6 The 3p atomic orbitals


These have n = 3 and l = 1. Hence m = 0, 1. Thus there are three of them, and
their functional forms are easily found from the tables. Their angular depen-
dences are the same as those of the 2p. To illustrate we consider 3p x as a typical
case. For this orbital, the YZ plane is a node, just as for 2p x . 3p x has an additional
node, coming from its radial part, which is just the surface of a sphere. A density
plot of this orbital is given in Fig. 5.24. The nodal sphere too is shown (as a circle).

Figure 5.24 The 3p x orbital. It has two nodes - one the surface of a sphere and the
other, the YZ-plane.

5.5.7 The 3d atomic orbitals


When n = 3, and l = 2, m = 0, 1, 2 and hence there are five different d orbitals.
The first one is 320 (r, , ) = R 32 (r )Y20 (, ). From the tables we find the angle
dependence to be r
1 5
Y20 (, ) = (3 cos2 () 1) (5.41)
4
Just as in the case of 2p z which we analyzed in detail, we can first construct a
section of the polar plot in the YZ plane to get the Fig. 5.25. This may then be
rotated about the Z-axis to get Fig. 5.26. A contour plot for this orbital is given in
Fig. 5.27.
76 Chapter 5 The Hydrogen Atom

Similarly one can analyze the other d orbitals. If one looks at the table,
one sees that the functions can be complex, and in such cases, one takes lin-
ear combinations that are real. This leads to the standard sets of d -orbitals,
d z2 , d x 2 y 2 , d x y , d xz and d y z . As an example, we give the polar plot and contour
plot for d x 2 y 2 in figures 5.29b and 5.28 respectively.
Finally, in figures 5.30,5.31 we give polar plots for the f-orbitals. The orbital
set that we have given are usually known as the cubic set of f-orbitals".

Figure 5.25 Polar plot of d z 2 orbital in the YZ plane. Z axis is the vertical axis. To get the
three dimensional appearance, one has to rotate this figure about the Z-axis

5.5.8 Note on orthogonality and Independence


.
Let us say the student A has taken his co-ordinate system, to get all the orbitals
that we discussed. Suppose another student, B, in a different room takes his own
co-ordinate system, and solves the Schrodinger equation. It is not likely that
the orientation of his XYZ axes is the same as As. For the s-orbitals this is not a
problem, as they will be identical for both A and B. But the two sets of p-orbitals
are different. Do we then have six p-orbitals? To be clear, the ones obtained by
them will be denoted as (p xA , p yA , p zA ) and (p xB , p By , p zB ). The answer is that there
are only three, because any p orbital of the second set (say p xB ) can be expressed
as a linear combination of the three obtained by A and vice versa. Thus the first
and second set are not linearly independent. In fact out of the six orbitals, one can
find only three which are independent of each other. In addition to being linearly
independent, one always chooses orbitals to be orthogonal to one another. We
now explain this idea.

If there are two solutions of the Shrodinger equation 1 and 2 , with different
energies E 1 and E 2 , then one can easily prove:
Z
d 1 2 = 0. (5.42)

The orbitals are said to be orthogonal. This is easy to see for the 1s and 2p z orbitals,
by actually looking at the integral. Thus any atomic orbital of the hydrogen atom
5.5 Shapes of atomic orbitals 77

Figure 5.26 Polar plot of d z 2 orbital. It has the surface of two cones as its nodes. The
for any point on the surface of the cone is 54.73o .

is orthogonal to any other, provided, they have different energies. However, what
about the orbitals 2p x , 2p y , 2p z which have the same energy? The way we have
written them down, they are orthogonal to one another. Making a set of orbitals
orthogonal makes sure that they are independent.
78 Chapter 5 The Hydrogen Atom

Figure 5.27 Contour Plot of d z 2 orbital.

Figure 5.28 Contour plot of d x 2 y 2 orbital. it has two nodal planes, that bisect the angle
between the X and Y axes.
5.5 Shapes of atomic orbitals 79

(a) d z 2 (b) d x 2 y 2

(c) d x y (d) d xz

(e) d y z

Figure 5.29 Polar plots of the five d-orbitals. Each has two nodes.
80 Chapter 5 The Hydrogen Atom

(a) f x 3 (b) f y 3

(c) f z 3 (d) f x y z

(e) f z(x 2 y 2 ) (f ) f y(z 2 x 2 )

Figure 5.30 Polar plots of six of the seven f-orbitals. Each has three nodes.
5.5 Shapes of atomic orbitals 81

Figure 5.31 The seventh orbital, f x(y 2 z 2 )


Chapter 6

Angular Momentum and the Spin


of the Electron

In Bohrs theory, he had to assume that the angular momentum, due to the motion
of the electron around the proton can take only certain values. Interestingly it
leads to the same expression as the one obtained by Schrodinger using wave
mechanics (see Eq. (5.26)). Therefore, it is clear that there is some truth in
Bohrs assumption. So it is worth asking, what would be the value of the angular
momentum of the electron if it is sitting in one of the stationary states discussed
in the last section?
In Bohrs theory, the electron is in a circular orbit, and consequently, the
angular momentum vector points in the direction perpendicular to the circle.
In quantum mechanics, it is not possible for the angular momentum vector to
point in a definite direction. To see this, suppose the angular momentum is
pointing in the Z-direction. Then, the momentum in the Z-direction has to be
zero, so that the uncertainty p z = 0, which means that z = . There is an
infinite uncertainty in the Z co-ordinate, implying that the electron cannot be
within the atom. Hence a confined electron cannot have its angular momentum
pointing in a fixed direction.
The best way to understand how the angular momentum is oriented would be
to look at an experiment for determining the angular momentum of the electron
in the Hydrogen atom. Thinking classically, an electron that is going around the
proton with a speed v constitutes a current running in a loop, the magnitude of the
current being I = ev/(2r ) where v/(2r ) is the number of times the electron goes
around the proton in unit time. We know that a current I , going around a circle of
radius r has a magnetic moment of magnitude = I A = I r 2 where A = r 2 is
the area of the circle. Thus the atom is expected to have an angular momentum,
as well as a magnetic moment. Further, = e(mvr )/(2m) = e/(2m)|l|, where l is
the angular momentum vector. Written as a vectorial equation,
e
= l (6.1)
2m
If such an atom with an angular momentum l and an associated magnetic mo-

83
84 Chapter 6 Angular Momentum and the Spin of the Electron

ment is placed in a uniform external magnetic field of strength B , assumed to


be in the Z-direction, then the magnetic field will exert a torque on the atom. Note
that there is no net force on the atom, but only a torque. But because the atom has
an angular momentum, the angular momentum vector would precess around the
direction of the external magnetic field, like the precessional motion of a spinning
top, caused by the gravitational field (see Fig. 6.1). In the precessional motion,
the component of the magnetic moment, in the direction of the magnetic field
does not change.

Figure 6.1 The top spins about the axis indicated by the dotted line. Due to the torque
exerted by the gravitational force, this axis precesses about the vertical axis (direction
of the gravitational field), and the angle between the two directions does not change. In
the absence of frictional forces, the top will continue to precess forever, with the angle
between the vertical and the spinning axis unchanged.

The external field and the magnetic moment will have an interaction energy
given by
E i nt = z B. (6.2)
However, if the magnetic field is not homogeneous, then the force exerted at the
two poles of the magnetic moment will not be same and there will be a net force
on the atom, trying to move it in the direction of maximum change of the field.
The force will be equal to z B
z . Note that the force exerted will depend on the
orientation of the magnet with respect to the field.
Now let us imagine an experiment in which a beam of hydrogen atoms which
have been put in the n = 2, l = 1 state are made to pass through a region with
inhomogeneous magnetic field as shown in Fig. 6.2. Thinking classically, the
angular momentum of the atom can be oriented in any direction. Hence the
atoms would be deflected by a continuous range of angles and are expected to
form a continuous pattern. However, if this experiment is done, it would be found
that the atoms get deflected by three definite angles only, and form the pattern
shown in 6.2. This means that the magnetic moment of the atom can have only
three different orientation with respect to the applied magnetic field as shown in
Fig. 6.3. This in turn means that when we try to measure the orientation of the
angular momentum with respect to a particular direction, we find that it can have
only certain orientations only. If the experiment is done with hydrogen atoms in
85

the ground state, one one find that there is no deflection of the beam, implying
that there is no angular momentum.
Having got this result, it is of interest to ask what are the values of angular
momentum, in the stationary states that we had found in the last section. As is
typical of quantum mechanics, this requires we should think of the operator asso-
ciated with angular momentum. The classical expression for angular momentum
is
l = rp (6.3)
To find the operator, one will replace the components of momentum p x , p y and

p z by the associated operators i ~ x , i ~ y and i ~ z , and get the operator
l, as well as lx , ly and lz . These would be in cartesian co-ordinates, and can
be expressed easily in spherical polar co-ordinates. The expression for lz is
particularly simple and is i ~ . One can also find the operator l2 which is

the operator corresponding to the square of the angular momentum. Then, on
examining the wave functions of previous section, one sees that nl m (r, , ) has
the term e i m in it, which means that nl m (r, , ) is an eigenfunction of lz , with
an eigenvalue m ~. Also nl m (r, , ) is an eigenfunction of the operator l2 with
an eigenvalue l (l + 1)~2 .
These mean the following:

1. For an electron sitting in the state nl m (r, , ), the value of the square of
the angular momentum l2 , is fixed and is equal to l (l +1)~2 . Hence one says
p
that the length of the angular momentum vector has the value l (l + 1)~.

2. The Z-component of the angular momentum vector has the value m ~. As


m = l , l + 1, ....l , there are (2l + 1) different possible values for m and
hence there are (2l + 1) possible orientations that the angular momentum
vector can take.

In particular, if l = 1 there are 3 different possible orientations that the an-


gular momentum vector can take; each orientation leads to one of the beams
in the experiment that we discussed. The possible ways in which the angular
momentum vector can be oriented in space is shown in Fig. 6.3. Note that each
of the beams corresponds to a definite value of m. Thus this experiment prepares
the Hydrogen atom in each of the three different states, 21,0 , 21,1 and 21,+1 .
These wavefunctions, as we have seen earlier, are normalized and orthogonal to
one another.
This kind of experiment was done by Stern and Gerlach in 1922, using Ag
atoms. Ag has 46 of its 47 electrons paired in a closed shell, with one electron
sitting outside this shell, in the 5s orbital. The net angular momentum of the
closed shell is zero, and the electron in the 5s has no orbital angular momentum.
Hence one does not expect the beam to split up, but in the experiment, the beam
was found to be split into two.
The reason for this splitting is that even though the 5s electron has no orbital
angular momentum, it has a spin angular momentum. (This was suggested by
86 Chapter 6 Angular Momentum and the Spin of the Electron

Figure 6.2 The beam of atoms splits into three, indicating that there are only three pos-
sible orientations that the angular momentum can have with respect to the direction of
the external magnetic field.

p
Figure 6.3 For l = 1, the angular momentum vector has the magnitue 2~ and can be
oriented in three different ways. These are such that the Z-component of the angular
momentum is equal to m ~, with m = 1, 0 and 1. These three different possibilities are
shown in the figure. Only the magnitude and one of the components of the angular
momentum can be specified, this being a consequence of the uncertainty principle.
In the figure, itp
is the Z-component that is specified. As the magnitude of the
p angular
momentum is 2~, the vector itself has to be making an angle of cos1 (m/ 2) with the
Z-direction, and may lie anywhere along the surface of a cone with this as the angle of
the cone. Consequently, the values of X- and Y-components of the angular momentum
cannot be known precisely.
87

Figure 6.4 Electron may be thought of as a charged sphere that is spinning, as a result
of which it has a spin angular momentum. (As spin is a purely relativistic phenomenon
this is not correct.)

p spin quantum number s = 1/2 and the angular momentum has the
Figure 6.5 The
magnitude 3~/2. It can be oriented in two different ways, as shown in the figure.
These are such that the Z-component of the spin angular momentum has either the
value ~/2 or ~/2.
88 Chapter 6 Angular Momentum and the Spin of the Electron

Uhlenbeck and Goudsmit). To think about this physically, one can imagine that
the electron is a charged sphere, spinning about an axis passing through it (see
Fig. 6.4). Due to this spinning motion, it would have a spin angular momentum
which we denote as s. Just as in the case of orbital angular momentum, there
would be a magnetic moment arising from the spin, given by

= e/(2m)2 s (6.4)

(note that there is an extra factor 2 in this formula, in comparison with Eq. 6.1.
(The fact is: spin is a purely relativistic phenomenon and it is not correct to say
that the electron is charged spinning sphere. The proper theory of spin is to be
found in the Dirac equation, which marries Quantum Mechanics and Relativity,
but is out of the scope of this discussion). The Stern-Gerlach experiment shows
that the spin angular momentum can be oriented only in two different ways. As

the properties of spin cannot be obtained from the Schrodinger equation, we
simply postulate its existence and assume the following:
p
1. The magnitude of the spin angular momentum is s(s + 1)~ where s can
have only the value 1/2.

2. The orientation of the spin angular momentum is such that its Z-component
can have only the values m s ~ with m s taking the values +1/2 or 1/2 only.

Just like in the case of orbital angular momentum these two states must have
an associated wave function, which we can imagine is dependent on a spin co-
ordinate . Thus one denotes the wave fucntions for the two states as () for
m s = 1/2 and () for m s = 1/2. In analogy with orbital angular momentum
states, these are normalized and orthogonal, meaning
Z
()()d = 1 (6.5)
Z
()()d = 1 (6.6)

and Z
()()d = 0 (6.7)

6.1 Wave function for an electron in the hydrogen atom


Thus an electron in the hydrogen atom is specified by four quantum numbers -
n, l , m and m s and its wave function can be either nl m (r, , )() or nl m (r, , )().
These are referred to as spin orbitals, as they have an orbital part as well as a spin
part.
Chapter 7

Many electron atoms and the


Paulis exclusion principle

Figure 7.1 The He nucleus is taken to be at the origin. The two electrons are at posi-
tions r1 (x 1 , y 1 , z 1 ) and r2 (x 2 , y 2 , z 2 ). Their distances from the nucleus are denoted
as r 1 and r 2 respectively. The interelectronic distance is r 12 .

Figure 7.2 Hamiltonian for the motion of the two electrons in the He atom. The nu-
cleus is taken to be at the origin. The Hamiltonian has kinetic energy terms from the
two electrons.

89
90 Chapter 7 Many electron atoms and the Paulis exclusion principle

We now think of the He atom, which has two electrons (Fig. ??). Experiments
show that the energy of the ground state of the atom is 78.95 eV . Taking the

nucleus to be at the origin, the Schrodinger equation for the He atom would
involve the Hamiltonian shown in Fig. ??. For this Hamiltonian, the equation is
too complex to be solved analytically to find the allowed energy levels and the
associated wave functions. The reason for the difficulty is that the motion of the
two electrons is correlated - the movement of the first electron is influenced by
the instantaneous position of the other electron, making analytical solution of the
problem impossible. However, a great deal of work has been done on this kind of
problems, to get approximate solutions, using a computer. All these methods are
based on the variation theorem, which states that if is any approximate wave
function for a system, then the quantity E defined by

d H
R
E= R (7.1)
d

obeys the inequality


E E0 (7.2)
where E 0 is the energy of the ground state of the system. If is normalized, then
the denominator in the Eq. (7.1) is unity. The inequality is very useful, because if
one had two approximate functions 1 and 2 with corresponding approximate
energies E 1 and E 2 , then both of them are guaranteed to be greater than E 0 , and
further, the lower of the two is a better approximation to the energy - thus one
has a quantitative procedure to evaluate the energy (see below).
In the case of He atom, if one is prepared to neglect the effect of electron
electron repulsion, then in the ground state of the atom, each electron would sit
3/2
in a 1s atomic orbital of the form 1s(r) = p1 aZ0 e Z r /a0 where Z e is the charge
on the nucleus, and for He atom Z = 2. Thus an approximate wave function for
the two electrons would be of the form given in Eq. (7.4)

(r1 , r2 ) = 1s(r1 )1s(r2 ) (7.3)


1 Z 3 Z r 1 /a0 Z r 2 /a0

= e e (7.4)
a0

with Z = 2. However, the effect of electron-electron repulsion is not included in


the above wave function. Its effect can however be included by a modification of
the wave function. Imagine that the second electrons density is smeared out in
the form of a cloud as it is sitting in a 1s atomic orbital. This cloud will definitely
shield the first electron from the nucleus, as a result of which the effective nuclear
charge felt by the electron is not 2, but less than 2. This means that one would
expect Z to be less than 2. Then how does one find the best value of Z ? This
is where the variation theorem is useful. For any value of Z , the value of E
calculated using the Eq. (7.1) is guaranteed to be greater than E 0 . Hence the best
wave function and energy is obtained by using that value of Z which minimizes
E . For He atom, this kind of calculation is not difficult to perform and it leads
7.1 Inclusion of the spin part: Paulis principle 91

to the result Z = 1.65 and energy 77 eV . The difference between the actual
nuclear charge and the effective nuclear charge is denoted by S and arises from
the shielding. That is,
Z = Z S (7.5)

7.1 Inclusion of the spin part: Paulis principle


The approximate function given in Eq. (7.4) does not take spin of the elctron into
account. It is possible to have four possible spin wave functions for a system with
two electrons. They are

S D = (1 )(2 ) (7.6)
1
SC = p (1 )(2 ) + (1 )(2 )

(7.7)
2
S B = (1 )(2 ) (7.8)
1
S A = p (1 )(2 ) (1 )(2 )

(7.9)
2

Thus the correct form of the approximate wave function can be the spatial part
given by Eq. (7.4) multiplied by any one of the functions S A , S B , SC or S D . It there-
fore appears as if one has four different wave functions, with both the electrons
sitting in the orbital 1s. These would have different values for the total spin angu-
lar momentum and its Z-component. Hence, if we passed a beam of He atoms
through an inhomogeneous magnetic field, the beam should split up depending
on the different orientations of the spin angular momentum. However, no such
splitting up is observed, indicating that all of the four states are not possible. This
is due to the Paulis exclusion principle, which states that in any many electron
system, if one interchanges the co-ordinates of any two electrons, the wave function
should change sign. This implies that one has to choose S A as the spin part, and
the wave function would then be
1
(1, 2) = 1s(r1 )1s(r2 ) p (1 )(2 ) (1 )(2 )

(7.10)
2

This wave function means that there are two electrons in the 1s atomic orbital,
and that one has the spin state while the other has the spin state - note the
electrons are indistinguishable - we cannot say which one is having spin or .
The Pauli principle implies that each orbital can accommodate only two
electrons -i.e., one per spin-orbital. This leads to usual statement of the principle
as applied to atoms - no two electrons can have the same set of values for all the
four quantum numbers.
Usually, we do not write the function of Eq. (7.10) in such detail - instead one
says the electron configuration of the He atom is 1s 2 . When this is said, it means
that the associated wave function has the kind of form given in this equation.
92 Chapter 7 Many electron atoms and the Paulis exclusion principle

7.2 Many-electron atoms


Now we consider a few of the atoms, to illustrate the ideas. Consider Li . The 1s
can accommodate only two electrons, and the third electron can go to 2s, so that
the electron configuration of the ground state of Li is 1s 2 2s 1 . Calculations using
wave functions such as the one in Eq. (7.10) show that this has lower energy than
the case where we put the third electron in a 2p to get the configuration 1s 2 2p 1 .
This means that in a many electron atom, 2s has a lower energy than 2p. This
is not surprising, as we know that a 2s orbital penetrates more into the region
occupied by a 1s orbital than a 2p orbital, and hence has lower energy. The order
in which orbitals are filled up is shown in Fig. ??. Slater has formulated a set of
rules to determine the effective nuclear charge for many electron atoms, given
below, using which the effective nuclear charge felt by an electron sitting in any
orbital of an atom can be calculated. These are very useful as they give an idea
of the energy and size of the atomic orbital - the larger the nuclear charge, the
smaller the size of the orbital, and hence lower the energy.

7.2.1 Effective nuclear charge, Z = Z S


The electrons are divided into groups that have different shielding constants: 1s;
(2s, 2p); (3s,3p); 3d; (4s,4p); 4d; 4f; (5s,5p). s and p form the same group.

Contributions to S for any particular orbital in a group are:

1. Nothing from any group outside it.

2. 0.35 from the same group except for 1s for which it is 0.3.

3. If the group considered is s or p, then 0.85 from electrons further in and 1.0
from all electrons still further in.

4. If it is d or f, then 1.0 from all electrons in the previous groups.

7.3 Reading assignments


1. Portions from the book by Atkins and Jones regarding the Aufbau princi-
ple, periodic table, the variation of ionization energy, electron affinity and
atomic size along the various periods and groups.

2. Calculation of shielding for different atomic orbitals from the book by


Huheey et al.

3. Sections 12.10, 12.11, 12.12 and 12.13 of the book by Zumdahl.


7.3 Reading assignments 93

Figure 7.3 The general order of orbital energies. The actual order may vary depending
upon the occupancies of different orbitals.
Chapter 8

Chemical Bonds

Figure 8.1

8.1 Why bonds are formed?


In this chapter, we use Quantum Mechanics to understand why and how chemical
bonds are formed.

8.2 The Hydrogen Molecular Ion


The simplest species in which a chemical bond exists is the Hydrogen molecular
ion, H2+ . This has two protons which are held together by a single electron.
Imagine that the two nuclei in the molecular ion are kept fixed at a distance R
from each other, as shown in Fig. ??. Then the Hamiltonian operator for the single
electron (as we assume nuclei are held fixed, we do not include kinetic energy

95
96 Chapter 8 Chemical Bonds

Figure 8.2 H2+ molecular ion has two nuclei (A and B) and one electron

operators for the motion of the nuclei) is given by

~2 e2 e2 e2
H = 2 + (8.1)
2m 40 r A 40 r B 40 R


In this case, the Schrodinger equation H = E can actually be solved exactly
and one gets perfect agreement with experiments. This, however is possible only
with a one electron system. Therefore, we look for methods which are of general
applicability.
Imagine that the internuclear distance R = . Then the system would have
dissociated to H and H + . An electron in the hydrogen atom would have an energy
of 13.6 eV due to the electron that sits in its 1s atomic orbital - this is measured
relative to that of a free proton and a free electron, which are taken to have energy
zero. As one brings the proton and the H atom closer by reducing the value of
R from infinity to a finite value, there are two contributions to the energy of the
system. These are: E r due to the repulsive interaction between the two nuclei
and E a arising from the attractive interaction of the electron with the two nuclei.
Thus
E t (R) = E r (R) + E a (R) (8.2)
We know the form of E r (R); it is given by

e2
E r (R) = . (8.3)
40 R
However, we do not know the explicit form of E a - it needs to be calculated, for
which we will use quantum mechanics. Even though we do not know the form
of E a now, we know that it has the value 13.6 eV when R = . We also know
that if we put R = 0, then the two protons would form a single nucleus of charge
2 and this charge is the same as that of a He nucleus. Hence E at t r act i ve (R =
0) = 13.6 4 = 54.4 eV . In the absence of any information on the behavior
of E at t r ac t i ve (R). Using this and the calculations of the next section, we can
construct a plot of E t ot al (R) against R, given in Fig. ??. In constructing the above
curve, we have assumed that the two nuclei are held fixed at a separation of R.
Hence this curve does not include the kinetic energy of nuclear motion, but it just
says how the energy of the bonded system changes as one changes R. This curve
8.2 The Hydrogen Molecular Ion 97

Figure 8.3 Schematic potential energy curve (E t ) for H2+ obtained by adding together
the repulsive (E a ) and attractive (E a ) parts of the energy.

EHRL,eV
6

R ,Angstrom
1 2 3 4

-2

Figure 8.4 Potential Energy Curve for the bonding and antibonding states of H2+ as a
function of the distance between the two nuclei. The red curve is the experimental
result. The dashed blue curve (E b ) is the result of the simple approximation discussed
in the text. The green curve is plot of E a .
98 Chapter 8 Chemical Bonds

will serve as the potential energy for nuclear motion, and hence is referred to as
the potential energy curve.

8.3 Wave function for the electron


Now we try to guess an approximate wave function for the electron. We know
that if the electron is very close to nucleus A, then the effect of nucleus B can be
neglected; then the wave function would resemble a 1s atomic orbital having its
center on this nucleus. We denote this function by 1s A . Explicitly,
3/2
1 1
1s A = p e r A /a0 . (8.4)
a0

Similarly, when the electron is very close to nucleus B, its wave function would
resemble 1s B , given by
3/2
1 1
1s B = p e r B /a0 . (8.5)
a0
Thus, the actual function has to reduce to 1s A for points close to A and reduce to
1s B for points close to B. A function that satisfies these conditions is

= c A 1s A + c B 1s B . (8.6)

where c A and c B are constants. It is possible to find the values of these using the
variation method. However, in this problem we can use physical intuition. It is
obvious that the two atomic orbital must make the same amount of contribution
to the wave function . This means that c 2A = c B2 . Then we have two possibilities;
the first is c A = c B and the second is c B = c A .

8.3.1 The bonding molecular orbital


Let us look at the case with c A = c B . For reasons that will become clear, we will
call this the bonding molecular orbital and denote it by b . On imposing the
normalization condition d 2b = 1, we find c A = p 1 where S = d 1s A 1s B .
R R
2(1+S)
S is called the overlap integral. Thus becomes

1
g 1s = b = p (1s A + 1s B ). (8.7)
2(1 + S)

Note the name g 1s given to . A calculation of the energy of this may be done
using E b = d b H b . It gives
R

H A A + H AB
Eb = (8.8)
1+S

where we have defined H I J = d 1s I H 1s J with I , J = A or B and used the


R

following results: H A A = HB B (this is obvious!) and it is easy to show that H AB =


8.3 Wave function for the electron 99

HB A . Explicit expressions for H A A and H AB can be found and are used to get the
E b given below.
e2 1 J +K

E b (R) = + (8.9)
40 a 0 D 1 +
where D = R/a 0 , J = D1 +e 2D 1 + D1 ,K = e D (1 + D)), and = e D 1 + D + D 2 /3 .

Figure 8.5 Contour plot for the bonding molecular orbital. Usually, the molecular axis
is taken to be the Z-axis. The plot would look the same in any plane containing the Z-
axis. The orbital has no nodes It has cylindrical symmetry about the Z-axis and hence
is referred to as a type orbital. The orbital is symmetric under the inversion operation
about the centre of the molecule and hence is referred to as "gerade". The notation for
this orbital is 1s . The contours are labelled by the value of the wave function along
them.

On making a plot of E b (R) against R we get the blue curve in Fig. ??. This is to
be compared with the experimentally obtained curve shown as the red curve. Note
that the green curve is always above the red one, because of the variation theorem.
The theoretical curve has a minimum at R = 1.32 and at this value E b = 1.77.
Thus calculation leads to an equilibrium internuclear distance R e = 1.32 and
a dissociation energy D e = 1.77eV , in comparison with the experimental values
1.06 and 2.68eV .
A contour plot for the orbital b in the XY-plane is shown in Fig. ??. This
shows that the wave function encompasses both the nuclei -i.e., this is a molecular
orbital (MO). Note that in the above, we have taken a linear combination of the
two atomic orbitals 1s A and 1s B to obtain the MO. This method therefore is known
as the LCAO-MO method.
We can now calculate the probability density that this molecular orbital leads
to. On squaring, we get
1
2b =
2
1s A + 1s B2 + 21s A 1s B

(8.10)
2(1 + S)
100 Chapter 8 Chemical Bonds

The important thing to notice is the term 21s A 1s B which is a product of the two
atomic orbitals. As it is a product, its value is appreciable only in the region
between the two nuclei. This term thus causes an increase of the probability
density in the internuclear region, causing the stabilization of the system, and
hence bonding. A plot of the electron density for this orbital in which the density
of an electron sitting in this molecular orbital is shown in Fig. ??.

8.3.2 The anti-bonding molecular orbital


We now consider the case where c B = c A . On normalising, we get

1
u1s = a = p (1s A 1s B ) (8.11)
2(1 S)

and the corresponding energy is

H A A H AB
E b (R) = (8.12)
1S
which may be evaluated to get

e2 1 J K

E b (R) = + (8.13)
40 a 0 D 1

A plot of this against R gives the green curve in Fig. ??. Interestingly this has no
minimum, implying that in this state H2+ is unstable - there is no bond between
the two species.
A contour plot for this orbital in any plane which contains the Z-axis is shown
in Fig. ??. The orbital has a node, which is a plane passing the centre of the
molecule and perpendicular to the molecular axis. Therefore any electron sitting
in this molecular orbital would have its density concentrated away from the
internuclear region, leading to destabilization of the system - this is the reason
why the orbital is antibonding.
The quantity H A A may be thought of as the energy of an electron if it was
sitting in the orbital 1s A , in the molecule. Of course, the electron does not sit in
this orbital - it sits in an MO. Similarly HB B would be the energy of the electron
if it was forced to sit in 1s B within the molecular ion. Clearly, H A A = HB B . It is
possible to calculate these quantities at any value of R. We can calculate them,
and represent them in a figure (see Fig. ??). The probability density for this orbital
is
1
2a = (1s 2A + 1s B2 21s A 1s B ) (8.14)
2(1 S)
Here, the term 21s A 1s B occurs with a negative sign and hence leads to a decrease
of density in the internuclear region. This is the reason for reduction in the density
in the internuclear region, and consequently the orbital is anti-bonding.
8.3 Wave function for the electron 101

Figure 8.6 Contour plot of the antibonding molecular orbital. Usually, the molecular
axis is taken to be the Z-axis. The plot would look the same in any plane containing the
Z-axis. There is a nodal plane perpendicular to the Z axis, passing through the centre
of the molecule. The orbital has cylindrical symmetry about the Z-axis and hence
is referred to as a type orbital. The orbital is antisymmetric under the inversion
operation about the centre of the molecule and hence is referred to as "ungerade". The
notation for this orbital is 1s .

Figure 8.7 Contour plot of electron density for the antibonding molecular orbital made
at the equilibrium geometry of H2+ molecular ion.
102 Chapter 8 Chemical Bonds

8.3.3 Valence bond theory


.
The wave function given in Eq. (8.7) can be looked in a different fashion. The
part 1s A in it means that the electron is sitting in 1s A associated with nucleus A and
in this function there is no electron associated with nucleus B. Thus this represents
a chemical structure that may be written as H H + . Similarly, the function 1s B
represents the structure H + H . Thus

1
b = p (H H + + H + H ) (8.15)
2(1 + S)

Thus the actual wave function is a linear combination of the wave functions
corresponding to the two structures H H + and H + H . This fact is represented by

H H+ H+H (8.16)

and one says that the actual structure is a resonance hybrid of the two structures
written above. It should be understood that the two structures written above have
no real existence and that the actual molecule is fluctuating between the two
structures. All that we are saying is that the actual wave function has contributions
from the two wave functions corresponding to these two structures.
8.4 The hydrogen molecule 103

8.4 The hydrogen molecule


8.4.1 Molecular Orbital Theory
We now consider the hydrogen molecule. As in the case of H2+ we imagine that
the nuclei A and B are held fixed at a distance R and then the Hamiltonian is given
by

~2 e2 e2
H = 21 (8.17)
2m 40 r A1 40 r B 1
~2 2 e2 e2
2 (8.18)
2m 40 r A2 40 r B 2
e2 e2
+ + (8.19)
40 R 40 r 12

Note that this does not include the kinetic energy of the nuclei. With this Hamil-

tonian, it is impossible to solve the Schrodinger equation

H (r1 , r2 ) = E (R)(r1 , r2 ) (8.20)

The reason why this equation cannot be solved is last term in (8.19), shown in
red color. This is the electron-electron repulsion term which makes the motion
of the first electron depend on where the second electron is at that instant of
time. This makes their motion correlated". If one were to omit this term from
the Hamiltonian, then one has a situation where the two electrons move inde-
pendently of one another. Each electron would feel only the two nuclear charges.
Consequently, each would sit in a molecular orbital covering both the nuclei, and
as the two electrons are independent, the total wave function is a product of
the wave functions for the two electrons. In particular, in the ground state of H2
we expect both the electrons to sit in the g 1s MO that we found in the case of
H2+ in the previous section. Thus the wave function would be

MO (r1 , r2 ) = g 1s(r1 )g 1s(r2 ) (8.21)

Note that this is only an approximate wave function, and that is why we used
the symbol . We can use this , with g 1s given by Eq.(8.7), to calculate the
potential energy curve for H2 molecule. This leads to the curve shown in Fig. 8.8.
Note that the behavior of the curve is totally wrong for large values of R. At large
R the energy should have been that of two separate hydrogen atoms (which we
take to be zero), but the result is very different from that. The reason for this will
be explained later.

8.4.2 Valence Bond Theory


According to Lewis a bond is formed by a pair of electrons. The quantum mechan-
ical version of this idea was formulated by Heitler and London. In the limit of large
internuclear distance the hydrogen molecule would dissociate to two hydrogen
104 Chapter 8 Chemical Bonds

Figure 8.8 Potential energy curves for H2 obtained from simple VB and MO methods.
The experimental values are R e = 0.74 and D e = 4.75 eV. In comparison VB theory
gives R e = 0.87 and D e = 3.16 eV. The MO results are R e = 0.83 and D e = 2.65 eV.

atoms. Hence the wave function may be written as 1s A (r1 )1s B (r2 ). This function
as written means that electron 1 is in the atomic orbital 1s A and electron 2 is in
the orbital 1s B . However, when the distance between the two is not infinitely
large, we cannot say which electron is in which orbital. Hence there is another
possible function, which is 1s A (2)1s B (1). The wave function which accounts for
this fact has to be either the one in Eq. (8.22) or the one in Eq. (8.24).

1
V B (r1 , r2 ) = p (1s A (r1 )1s B (r2 ) + 1s A (r2 )1s B (r1 )) (8.22)
2(1 + S 2 )
The above function, as is clear is symmetric under the interchange of the two
electrons. Therefore, to be acceptable, its spin part has to be antisymmetric.
Therefore the total wave function including spin is

1
V B (r1 , r2 ) p ((1)(2) (2)(1)) (8.23)
2

Using the above function, one gets the potential energy curve shown in Fig. 8.8.
The two spins are antiparallel and hence one has an electron pair bond". The
potential energy curve obtained with this function is better than that of the
molecular orbital theory. The way the wavefunction is chosen, it has no difficulty
in describing dissociation of the molecule. The other possibility is

1
VA B (r1 , r2 ) = p (1s A (r1 )1s B (r2 ) 1s A (r2 )1s B (r1 )) (8.24)
2(1 S 2 )

As the above is antisymmetric under the interchange of the two electrons. the spin
part has to be symmetric; hence the wave function could be any of the following
three
8.4 The hydrogen molecule 105

VA B (r1 , r2 )(1)(2)
VA B (r1 , r2 )(1)(2)
1
VA B (r1 , r2 ) p ((1)(2) + (1)(2)).
2
Each of these functions lead to the same repulsive potential energy curve with no
minimum, predicting that in these states there is no bond formation. The three
states together are referred to as a triplet state. Thus when spins are parallel there
is no bond formed - to have a bond, spins have to be antiparallel.

Improving the VB Theory

An obvious way to improve the above calculation is to include an effective nuclear


charge into the wave function, resulting in a better value for D e . The wave func-
tion can be improved further by realizing that the function in Eq. (8.22) represents
only sharing of a pair of electrons. In addition one can also think of functions of
the form 1s A (1)1s A (2) in which we have put both the electrons into 1s A , which
corresponds to the chemical structure H H + , as well as the situation where both
electrons occupy 1s B , represented as H + H . Thus the approximate would be

= c 1 V B + c 2 (H H + + H + H ). (8.25)

V B represents wave function for the sharing of two electrons and is the quantum
mechanical representation of the ideas of Lewis. c 1 and c 2 are chosen to get the
best possible energy. The contribution of the two ionic structures is expected to
be smaller, than that of the covalent, as the ionic structures have two electrons in
the same orbital, thus increasing their energies. However, one expects that both
would make the same amount of contribution to the final wave function. The fact
that the final wave function has the form of Eq. (8.25) is what is implied when one
uses the word resonance. One says that the actual structure of hydrogen molecule
is a resonance hybrid of the structures

HH+ H H H+H (8.26)

By including the ionic structures the value D e improves to 4.02eV .

8.4.3 MO theory - incorrect dissociation limits


The wave function of MO theory (see Eq. (8.21)) can be written as
1
MO = {(1s A (r1 )1s B (r2 ) + 1s B (r1 )1s A (r2 ))
2(1 + S)
+ (1s A (r1 )1s A (r2 ) + 1s B (r1 )1s B (r2 )) (8.27)

In the above equation, first line represents the covalent part - the electron pair
bond of VB theory. The last two terms, shown in red color are just the ionic terms
106 Chapter 8 Chemical Bonds

that we have discussed earlier. However, in this wave function, ionic terms and
covalent terms occur with the same importance. If this function is used for all
distances, then it predicts that the molecule will dissociate into a pair of hydrogen
atoms or a pair of ions (proton and H ) with equal likelihood, which is not correct.
The dissociation should be to two hydrogen atoms. This is the reason why the
MO wave function leads to wrong asymptotic limit for the energy in the large R
limit. Thus while the simple VB function does not have the ionic terms at all, the
MO function gives it more importance than needed.
The problem of incorrect dissociation limit in MO theory can be resolved by
using a wave function of the form

C I = c 1 g 1s(r1 )g 1s(r2 ) + c 2 u 1s(r1 )u 1s(r2 ). (8.28)

The above is referred to as the method of configuration interaction.

Molecule bond order Bond length (pm) Dissociation


Energy (kJ/mol)
Li 2 1 267 105
B e2 0 245 <10
B2 1 159 289
C2 2 124 599
N2 3 110 942
O2 2 121 494
F2 1 141 154

Table 8.1 Bond orders, bond lengths and dissociation energies of homonuclear di-
atomics of the 2nd row of the periodic table.
8.5 Hybridization 107

8.5 Hybridization
8.5.1 sp-hybridization
The electron configuration of Be is 1s 2 2s 2 , and it is found that it forms compounds
like B e H2 where Be is divalent. As 2p has a larger energy than 2s, the electron
configuration resembles a closed shell and hence one would have expected Be to
zerovalent. The answer to this puzzle is that the energy of 2p is not very much
higher than that of 2s it is possible to have 2p too involved in the bond formation.
Hence one can imagine that an electron is promoted from 2s to 2p. Once this
is done, Be has two unpaired electrons and hence it can be divalent. However,
there is a problem with this - after promoting one electron, one of the electrons
is in a 2s and the other is in a 2p orbital and these are used to form two electron
pair bonds. As one of them involves a 2s and the other a 2p, one does not expect
the bonds to be equivalent. However, experiment shows that B e H2 has a linear
geometry, with the B e atom in the middle; the two B e H are equivalent - they
have the same bond length. Hybridization was suggested as a way out of this.
Let us say that the molecule is lying along the Z-axis. Then only the 2s and 2p z
orbitals of B e can be used to form bonds with the two H atoms. So to form two
bonds, we combine 2s and 2p z to get two new orbitals, which are equivalent to
one another. Thus we write the hybrid orbitals as

1 = a 1 2s + b 1 2p z (8.29)
2 = a 2 2s + b 2 2p z (8.30)

As the contribution of 2s to the two orbitals must be the same, we conclude that

a 12 = a 22 . (8.31)

As there are only two hybrid orbitals, the contribution of 2s to the two of them
together should add up to unity. Hence

a 12 + a 22 = 1. (8.32)

These imply that we can take


p
a 1 = a 2 = 1/ 2. (8.33)

As each hybrid orbital is normalized, we have


Z Z
21 d = 22 d = 1, (8.34)

which imply
a 12 + b 12 = 1. (8.35)
and
a 22 + b 22 = 1. (8.36)
108 Chapter 8 Chemical Bonds

From Eq. 8.33 and 8.35 we get p


b 1 = 1/ 2 (8.37)
p
Note that b 1 could have another possible value, which is 1/ 2. This gives the
orbital 2 , making things easy! Thus

1
1 = p (2s + 2p z ) (8.38)
2
1
2 = p (2s 2p z ) (8.39)
2

are the hybrid orbitals.

Figure 8.9 The two sp hybrid orbitals. Note that the node is the curve that goes to
infinity and that it does not pass through the nucleus.

It may be verified that they satisfy all the conditions that are required. The
shapes of these orbitals are illustrated by the Fig. 8.9. In this case, the 2s and only
one of the 2p orbitals are mixing together to form two equivalent hybrid orbitals.
Hence this is referred to as sp-hybridization.

8.5.2 sp 2 -hybridization
The electron configuration of B is 1s 2 2s 2 2p 1 . We now consider the molecule
BC l 3 , which is known to have the geometry of an equilateral triangle, with the
chlorines sitting at the corners and B sitting at the center. We shall take the B to
be at the origin, with one of the B cl bonds lying along the positive X-direction
(see figure). All the three B C l bonds are found to be equivalent, having the
same bond lengths. Hence in this case, one imagines that the 2s, 2p x and 2p y
mix together to give three equivalent hybrid orbitals, 1 , 2 and 3 , each of which
is normalized and each is orthogonal to the other two. We take the first hybrid
orbital to be along the X-axis. Then it would not have any contribution from 2p y .
Hence we can write
1 = a 1 2s + b 1 2p x (8.40)
8.5 Hybridization 109

As the s-orbital should make equal contribution to all the orbitals, one should
p
have a 12 = 1/3, and therefore we put a 1 = 1/ 3. Normalization of 1 gives a 12 +b 12 =
q
1 and hence b 1 = 23 , so that
1 p
1 = p (2s + 2 2p x ) (8.41)
3
We take the second hybrid orbital to be
2 = a 2 2s + b 2 2p x + c 2 2p y (8.42)
As 2s should make the same amount of contribution as in the case of 2 we should
p
have a 2 = 1/ 3. The orthogonality of 1 and 2 implies
Z p
d 1 2 = 1/3 + b 2 2/3 = 0. (8.43)

which leads to p
b 2 = 1/ 6. (8.44)
Imposing the normalization condition leads to
a 22 + b 22 + c 22 = 1 c 22 = 1/2 (8.45)
Hence
1
c2 = (8.46)
2
Choosing the positive sign leads to 2 and the negative sign gives 3 . Thus
1 1 1
2 = p 2s p 2p x + p 2p y (8.47)
3 6 2
and
1 1 1
3 = p 2s p 2p x p 2p y (8.48)
3 6 2
The orbitals 1 , 2 and 3 lie in the XY plane. They make an angle of 120o with
each other and hence point towards the corners of an equilateral triangle, thus
explaining the geometry of BC l 3 . For shapes of these orbitals see Fig. 8.10.

8.5.3 sp 3 -hybridization
This occurs in tetrahedral molecules, like C H4 . A set of hybrid orbitals satisfying
the normalization and orthogonality conditions are:
1
1 = 2s + 2p x + 2p y + 2p z

(8.49)
2
1
2 = 2s + 2p x 2p y 2p z

(8.50)
2
1
3 = 2s 2p x + 2p y 2p z

(8.51)
2
1
4 = 2s 2p x 2p y + 2p z

(8.52)
2
In space,the orbitals are aligned in the directions: (1, 1, 1), (1, 1, 1), (1, 1, 1)
and (1, 1, 1) and make angles of cos 1 (1/3) = 109o 280 with each other.
110 Chapter 8 Chemical Bonds

Figure 8.10 The three sp 2 hybrid orbitals.

Figure 8.11 The directions of the four sp 3 hybrid orbitals are indicated by the arrows.

8.6 Huckel Molecular Orbital (HMO) Theory 111


8.6 Huckel Molecular Orbital (HMO) Theory
8.6.1 Ethylene
For molecules having a system of conjugated double bonds, a simple version
of MO theory was developed by Huckel. The simplest of these is ethylene. In
the conjugated system, the Carbon atoms are all in the same plane. Each one
of them is sp 2 hybridized, leaving a p-orbital perpendicular to the plane of the
carbon atoms. These orbitals overlap with each other to form what are referred to
as -type molecular orbitals. The simplest such molecule is Ethylene (Ethene).
In this case there are only two carbon atoms. Denoting the two p-orbitals as 1
and 2 and using MO theory leads to a bonding and an anti-bonding molecular
orbitals given by

1
1 = p (1 + 2 ) (8.53)
2(1 + S)
1
2 = p (1 2 ) (8.54)
2(1 S)

The energies of these orbitals are given by

H11 + H12
1 = (8.55)
1+S
H11 H12
2 = . (8.56)
1S
The -type overlap between two atomic orbitals is not large. It is of the order

of 0.2 0.25. Huckel made the approximation of neglecting it in comparison
with unity. Further, in HMO theory, it is usual to adopt the notation H11 = and
H12 = so that the above expressions become

1
1 = p (1 + 2 ) (8.57)
2
1
2 = p (1 2 ) (8.58)
2

The energies of these orbitals are given by

1 = + (8.59)
2 = . (8.60)

Thus the energies of the -type MOs may be represented by the Fig. 8.12. Each
Carbon atom contributes one electron to the system. Hence there are two
electrons and in the ground state of ethylene, the lower orbital is doubly occupied.
The total energy of the system thus E 0 = 2( + ). Note that there should actu-
ally be an additional term coming from the repulsion between the two electrons.
Huckel theory makes additional assumption that such "two"-electron contri-
butions can be neglected. Thus it has only "one"-electron contributions and
112 Chapter 8 Chemical Bonds

Figure 8.12 -type MOs of Ethylene. As is negative, + is lower than .

hence is referred to as a "one"-electron theory. If one allows the system to interact


with electromagnetic radiation of appropriate frequency, then an electron can be
promoted from the doubly occupied orbital to the un-occupied orbital. The en-
ergy change in the process is E = 2. On looking at the electronic-absorption
spectrum of Ethene, one finds an absorption at 170nm. Using this experimental
information, we find that = 90 kc al /mol . This is the usual strategy in the
theory. One does not evaluate the parameters theoretically - they are chosen so as
to fit the experimental data. Therefore the theory is referred to as a semi-empirical
theory.

8.6.2 Benzene
The most interesting example of a molecule with conjugated double bonds is
benzene, having 6 Carbon atoms arranged in the form a perfect hexagon. The
results of HMO theory is summarized in Fig. 8.13
The total -electron energy of benzene is thus 6 + 8. Note that this neglects
electron repulsions. If one imagined that benzene consisted of three separate
ethylene like units, then each unit would have had an energy of 2( + ) and
hence the total energy would be only 6( + ). However, in benzene the energy is
lower because the bonds are not localized to have three ethylene like units. The
MOs of benzene are:
1
1 = p 1 + 2 + 3 + 4 + 5 + 6 .

6
1
2 = p 1 2 23 4 + 5 + 26 .

12
1
3 = 1 + 2 4 5 .

2
1
4 = 1 2 + 4 5 .

2
1
5 = p 1 2 + 23 4 5 + 26 .

12
1
6 = p 1 2 + 3 4 + 5 6 .

6

8.6 Huckel Molecular Orbital (HMO) Theory 113

Figure 8.13 -type MOs of benzene. Note that the lowest and highest levels are non-
degenerate. The remaining two levels are doubly degenerate.As there are 6 -electrons,
the lowest three orbitals are doubly occupied.

Figure 8.14 The -type MOs of benzene. The red color indicates lobes that are positive
and blue is used for negative.
114 Chapter 8 Chemical Bonds

Figure 8.15 The cyclopentadienyl anion and cyclopropenyl cations - both are stable
aromatic systems.

These MOs are shown in Fig. 8.14. One says that the -orbitals of benzene are
delocalized over all the 6 Carbons leading to a lowering of energy, which is usually
referred to as the delocalization energy. Thus the delocalization energy of benzene
is

d el oc al i zat i on ener g y = 6( + ) (6 + 8)
= 2.

It is possible to estimate the delocalization energy of benzene thermochemically.


It is found to be 32 kc al /mol and hence = 18 kc al /mol . However, this value
of is quite different from that found in the previous section using spectroscopic
data. This is usual in semi-empirical theories. The values of parameters de-
termined using different kinds of experiments usually do not agree with each
other! Still the theory is found to be very useful as it gives a very good qualitative
understanding of what is happening in the system.


8.6.3 Huckels (4n + 2) rule
Huckel also analyzed a general ring system having any number of Carbon atoms
arranged in the form of a ring to form a system. He showed that irrespective of
the number of C-atoms, the lowest energy level is always non-degenerate. If the
number of C-atoms is odd, then the remaining orbitals are all doubly degenerate.
If the number of C-atoms is even (benzene is an example) then the highest orbital
is non-degenerate and the remaining are all doubly degenerate. This lead him
to conclude that if there are 2, 4, 6, 10... -electrons in such a ring, it forms a
closed shell and hence is expected to be stable. Thus whenever one has (4n + 2)
electrons with n = 0, 1, 2 . . ., the system is stable. Examples of such systems are:
benzene, cycopentadienyl anion, cyclopropenyl cation etc (see Fig. 8.15 ) Now
consider cylclobutadiene, C 4 H4 . If the four Carbon atoms sit at the corners of
a perfect square, then the MO energy diagram is shown in Fig. 8.16. As there
are 4 -electrons, the system would be a diradical and very reactive. Hence one
does not expect it to be possible to synthesize this system. In fact synthesizing
this molecule was a challenge, which has now been achieved. Actual experiment
shows the system to have a distorted rectangular geometry, with two ethylene like
bonds, thus making the system a closed shell system (How?).

8.6 Huckel Molecular Orbital (HMO) Theory 115

Figure 8.16 The molecular orbitals of cyclobutadiene (all carbon-carbon bond lengths
are assumed to have the same value, and hence the molecule is taken to be having the
shape of a square).
Bibliography

[1] R. P. Feynman, R. B. Leighton, and M. Sands. The Feynman Lectures on


Physics, Volume 3. 1965.

[2] R.P. Feynman and A.R. Hibbs. Quantum Mechanics and Path Integrals.
McGRaw-Hill, New York, 1965. Very nice, but mathematical introduction
to the path integral approach to quantum mechanics, by the person who
developed it.

[3] G. Gamov. Mr. Tompkins in paperback. Cambridge University Press, Cam-


bridgeon, 1965. Very nicely written introduction to quantum mechanics,
relativity theory etc., for the layman.

[4] A. Zeilinger, R. Gahler, C.G. Shull, W. Treimer, and W. Mampe. Single and
double slit diffraction of neutrons. Rev. Mod. Phys., 60:1067, 1988.


[5] J. Gribbin. In Search of Schrodingers Cat. Black Swan, London, 1984. A nice
introduction to the mysteries of quantum mechanics.


[6] J. Gribbin. Schrodingers Kittens and the Search for Reality. Little, Brown &
Co., London, first edition, 1995. Information on experimental verifications
and ideas on the foundations of quantum theory up to 1995.

[7] R.Gilmore. Alice in Quantum Land. Affiliated East West Press, New Delhi,
first edition, 1995. Alice goes to the quantum land, and not to the wonder
land!

[8] T. Hey and P. Walters. The Quantum Universe. Cambridge University Press,
Cambridge, first edition, 1987. Good introduction to the ideas of quantum
mechanics.

[9] G. Venkataraman. Quantum Revolution. Universities Press (India Ltd), 1994.


very readable introduction to the mysteries of quantum physics.

[10] G. Gamov and R. Stannard. The New World of Tompkins. Cambridge Uni-
versity Press, Cambridge, 1999. Tompkins in modern times - update of the
book by Gamov.

[11] O. Carnal and J. Mlynek. Youngs double slti experiment with atoms: A simple
atom interferometer. Phys. Rev. Lett., 66:2689, 1991.

117
118 BIBLIOGRAPHY

[12] R.P. Feynman. QED - The Strange Theory of Light and Matter. Universities
Press (India) Ltd, Hyderabad, 1999. Non-mathematical introduction to the
strange behavior of light and matter by one of the founders of the theory.

[13] W.H. Cropper. The Quantum Physicists. Oxford University Press, Oxford,
1971. Beautiful introduction to the development of quantum mechanics,
along with some biographical details.

[14] I. Marshall and D. Zohar. Whos Afriad of Schroedingers Cat. Quill William
Morrow, New York, first edition, 1998.

[15] R.B. Leighton R. P. Feynman and M. Sands. The Feynman Lectures on


Physics. Addison-Wesley Publishing Company, 1963. Masterly treatment of
physics. Unique discussion of quantum mechanics in vol III.

[16] Thomas F. Jordan. Qunatum Mechanics in Simple Matrix Form. Dover


Publications Inc, New York, 1986. One of the very few books on Matrix
Mechanics.

Você também pode gostar