Contents

3 Point groups and symmetries of molecules 3

3.1 Denition of point groups . . . . . . . . . . . . . . . . . . . 5
Symmetry transformations . . . . . . . . . . . . . . . . . . . 5
Proper point groups: rotations only . . . . . . . . . . . . . . 7
Improper point groups: rotations plus reections . . . . . . . 8
Additional point groups for molecules . . . . . . . . . . . . . 9
3.2 C
2v
group and H
2
O molecule . . . . . . . . . . . . . . . . . . 10
C
2v
group . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Irreps and basis functions . . . . . . . . . . . . . . . . . . . 12
Physical interpretation of the irreps . . . . . . . . . . . . . . 17
3.3 Projection operator method . . . . . . . . . . . . . . . . . . 19
3.4 C
3v
group and the NH
3
molecule. . . . . . . . . . . . . . . . 21
C
3v
group . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Irreps and basis functions . . . . . . . . . . . . . . . . . . . 23
1
2 CONTENTS
Chapter 3
Point groups and symmetries
of molecules
This chapter is an introduction to the point groups and their application
to the study of molecular structure. We demonstrate how to construct and
decompose representations of a point group through an in-depth discussion
on the two simple groups C
2v
and C
3v
and their application to the H
2
O
and NH
3
molecules, respectively. Our focus will be on how to relate the
group representation to the quantum mechanical quantities. In particular
we show how to calculate the wave functions for certain symmetry (irrep),
by applying the projection operator method.
There are three types of motion in molecules, i.e. electronic, vibrational
and rotational. The three types of motion are very dierent in energy, as
is shown in the following table:
Motion type energy in the order of

me
M
N
electronic 0th
vibrational 1st
rotational 2nd
where m
e
and M
N
being the mass of electron and nucleus, respectively.
Therefore, the adiabatic approximation applies and thus the three types of
motion can be studied independently. The electronic and vibrational mo-
tion of a molecule can be described by a point group, since these two types
3
4CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
of motion are governed by the Coulomb eld of the frame that consists of
atoms, and the frame obeys the symmetry of certain point group.
Since all the symmetry transformations of the point group preserve the
shape of the frame, i.e. preserve the Coulomb eld in which the electronic
and vibrational motions are taking place, one has
H = R
1
HR, or [H, R] = 0 (for all R G), (3.1)
where H is the Hamiltonian of the molecule which includes a Coulomb
eld produced by the nuclei that form the frame, and R the symmetry
transformation operators of corresponding point group G. Due to eq. (3.1)
the molecule is said to have the symmetry of group G.
A consequence of the symmetry is that the n

basis functions (
()
1
, . . . ,
()
n
)
of the irrep () are degenerate in energy.
Proof: Due to the invariance (or closure) of the rep-space we have
R
()
i
=
n

j=1

()
j
D
()
ji
(R).
Acting on both sides with H, then we have
l.h.s. = RH
()
i
=
()
i
R
()
i
=
()
i
n

j=1
D
()
ji
(R)
()
j
,
r.h.s. =
n

j=1
D
()
ji
(R)
()
j

()
j
, (3.2)
which leads to
n

j=1
D
()
ji
(R)(
()
i

()
j
)
()
j
= 0. (3.3)
As
()
i
(i = 1, . . . , n

) are linearly independent, the only possible solution

is

()
i
=
()
j
, (i, j = 1, . . . , n

),
i.e. all the states in the irrep are degenerate in energy. This conclusion
is valid for all the systems that have a symmetry described by eq. (3.1).
3.1. DEFINITION OF POINT GROUPS 5
3.1 Denition of point groups
Point groups are discrete subgroups of the rotational group O(3) (see ch.
6 Lie groups), and they are the symmetry groups of molecules and crys-
tals. The elements of point groups are symmetry transformations, which
preserve the distance between all pairs of points of the frame of molecules
or crystals, and bring the frame into coincidence with itself. And all the
transformations leave at least one point in the frame unchanged.
For a molecule, frame means a polyhedron in which the atoms are
situated at vertices. For a crystal a frame is made in such a way that rst
one draws a sphere with unit radius and nd the intersections (poles) of
the normals to the surfaces of the crystal, then make a tangent plane to
the sphere at each pole and obtain a closed polyhedron.
Symmetry transformations
1. Rotation about a n-fold axis (R
n
)
R
n
is a rotation through an angle
2
n
and it brings the frame into
coincidence with itself. The set of transformations {R
n
, R
2
n
, . . . , R
n
n
=
e} forms a cyclic group C
n
, where n is called the order of the axis.
In a point group, if there are more than one axes, the axis with
highest order is called the principal axis (z). The plane that the z
axis lies in or perpendicular to is called vertical (v) or horizontal (h)
plane, respectively. Pure rotations are called proper rotations.
For molecular point groups there is no restriction on the angle of
rotations, while for crystal point group there exist one imposed by
the translational symmetry. Suppose that the axis of rotation R
n
is

OA, where O is the origin and it is at one lattice point. If

t
1
is one
lattice vector that starts from O, then

t
2
= R
n

t
1
,

t
3
= R
n

t
2
, . . . are
also lattice vectors, and they all have the same length. In a plane
that is perpendicular to the axis

OA we depict Figure 3.1 showing
the lattice vectors that are related by rotation:
Suppose that the shortest lattice vector in this plane is

t
0
. Making
rotations R
n
and R
1
n
to

t
0
, respectively, one obtains

t = R
n

t
0
+ R
1
n

t
0
= 2 cos
2
n

t
0
. (3.4)
6CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES

*
H
H
H
H
H
H
H
Hj
- O
=
2
n
=
2
n

t
0
R
n

t
0
R
1
n

t
0
Figure 3.1: Lattice vectors under rotation
Since

t must be a lattice vector, thus its length must be an integer
times of the length of

t
0
, i.e.

t = k

t
0
,
where k is an integer. Therefore 2 cos
2
n
must be an integer and n
can only take the following values:
2 cos
2
n
= 2 1 0 1 2
=
2
n
= 180 120 90 60 0 (3.5)
n = 2 3 4 6 1
2. Reection in a plane ()
Reections in the vertical (v) and horizontal (h) planes are denoted
as
v
and
h
, respectively. One has

2
v
=
2
h
= e.
In a point group all the rotation axes and reection planes must
intersect at one point (the origin) which is xed in the space. Im-
proper rotation is a product of a proper rotation and a reection,
e.g. R
n

v
, in which the reection plane is perpendicular to the
rotation axis.
3. Space inversion (I)
This is a reection through the origin O, or a product of R
2

h
, where
R
2
is a rotation of and
h
is a reection in a plane that is perpen-
dicular to the rotation axis.
3.1. DEFINITION OF POINT GROUPS 7
Proper point groups: rotations only
Let l
n
be the number of n-fold axes in a point group G. There exist
restrictions on the number of l
n
in a group that has more than one rotation
axes. The sum of all elements of G can be written as
S = e +

n
ln

k=1
S
nk
, (3.6)
where S
nk
is the sum of all rotations along the k-th n-fold axes (excluding
the identity e) and l
n
is the number of n-fold axes. According to eq. (2.35),
the character of S
nk
is
(S
nk
) =
n1

j=1
_
1 + 2 cos
2j
n
_
=
n

j=1
_
1 + 2 cos
2j
n
_
3 = n 3, (3.7)
in which

n
j=1
cos(
2j
n
) = 0 is used. Therefore one has
(S) = 3 +

n
(n 3)l
n
= 3 l
2
+ l
4
+ 3l
6
, (3.8)
in which n takes the value of 2, 3, 4 and 6.
One can prove that (S) = 0 (see Problem 7, Ch 2), which leads to
l
2
= 3 + l
4
+ 3l
6
. (3.9)
The above relation limits the number of proper point groups to only eleven:
1. groups that have only one axis (uni-axial groups): C
n
(n = 1, 2, 3, 4, 6).
2. groups that have one n-fold axis (n = 2, 3, 4, 6) and a set of 2-fold
axes at right angles to it: Dihedral groups D
n
(n = 2, 3, 4, 6).
3. groups that have three 2-fold axes and four 3-fold axis the tetrahedral
group T.
4. group that has three 4-fold axes, four 3-fold and six 2-fold axes: the
Octahedral group O.
8CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
l
2
l
3
l
4
l
6
point group order
0 0 0 0 C
1
1
1 0 0 0 C
2
2
0 1 0 0 C
3
3
0 0 1 0 C
4
4
0 0 0 1 C
6
6
3 0 0 0 D
2
4
3 1 0 0 D
3
6
4 0 1 0 D
4
8
6 0 0 1 D
6
12
3 4 0 0 T 12
6 4 3 0 O 24
Table 3.1: List of proper point groups
Improper point groups: rotations plus reections
1. Adjoin
h
to C
n
:
C
s
, C
2h
, C
3h
, C
4h
, C
6h
(order=2n)
C
s
= C
1h
= {e,
h
}
2. Adjoin
v
to C
n
:
C
2v
, C
3v
, C
4v
, C
6v
(order=2n)
3. Adjoin I to C
n
C
i
= {e, I}
4. Adjoin reection S to C
n
If S
2
= R
2
(S = S
4
), then produce group S
4
= {e, R
2
, S
4
, S
4
R
2
}
If S
2
= R
3
(S = S
6
), then produce group S
6
= {e, R
3
, R
2
3
, S
6
R
3
, S
6
R
2
3
, S
6
},
S
4
and S
6
are Abelian groups.
5. Adjoin
h
to D
n
D
2h
, D
3h
, D
4h
, D
6h
(orderof D
nh
= 4n)
D
nh
= D
n
C
s
3.1. DEFINITION OF POINT GROUPS 9
6. Adjoin
d
to D
n

d
: reection in the plane d that is vertical and it bisects the angle be-
tween two neighboring 2-fold axes. The group D
nd
(n=2,3) contains
n additional reection planes and the group has 4n elements.
The adjoining of the inversion I or a R
n
gives no new groups.
7. Adjoin reection to T and O
T
d
: e, R
3
, R
3
(8), S
4
, S
3
4
(6), S
2
4
= C
2
(3),
d
(6), (order=24)
T
h
= T C
i
, (order=24)
O
h
= {e, R
2
(6), R
3
, R
2
3
(8), R
4
, R
3
4
(6), R
2
4
(3),
I,
v
(6), S
6
, S
5
6
(8), C
4

h
, C
3
4

h
(6),
h
(3)}
O
h
= O C
i
11 proper point groups and 21 improper point groups are all the
possible crystal point groups, which are also symmetry groups for
molecules.
Additional point groups for molecules
For molecules there is no restriction on the order of rotation axis, therefore
there exist the following additional point groups:
C

, C
v
, and D
h
, which can be represented by the three gures.
10CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
h h h
h h h
A
A
A
B
A
A
B
B
A
Figure 3.2: Three additional point groups for molecules
3.2 C
2v
group and H
2
O molecule
C
2v
group
We start with one of the simplest point groups C
2v
and use it to describe
the electronic motion of the molecule H
2
O, the structure of which is shown
in Fig. (3.3).
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@
@I
6
z(R
2
)
-
x

A
A
AA
O
h
1
h
2
i
f f
v
v

y
Figure 3.3: The spacial structure of H
2
O Molecule
In the Figure 3.3 the oxygen atom is at the origin of the coordinate
system x y z. The oxygen atom and the two hydrogen atoms are in
3.2. C
2V
GROUP AND H
2
O MOLECULE 11
e R
2

v

v

e e R
2

v

v

R
2
R
2
e
v

v

v

v

v
e R
2

v

v

v
R
2
e
Table 3.2: Group table of C
2v
the vertical plane v that contains the axes x and z, with the two hydrogen
atoms being on two sides of the zaxis in a symmetric way. The symmetry
group is C
2v
= {e, R
2
,
v
,
v
}, in which e is the identity, R
2
the rotation of
around the z axis, and
v
and

v
represent the reections in the vertical
planes v and v

, respectively. The reection plan v

contains the axes z and

y.
v
v

r
r
A
A

v
r

v
A

R
2
Figure 3.4: Multiplication law
The group table is given in Table 3.2. Through Figure 3.4 we can easily
verify the multiplication law. For example, we rst apply the transforma-
tions
v
and R
2
consecutively to the arbitrary point A, then we get

v
A = A

, R
2
A

= A

, or R
2

v
A = A

.
Since
v
A = A

and A is arbitrary, we have that

v
= R
2

v
. The other
multiplication laws can be veried in the same way.
C
2v
is an Abelian group. Since its order is 4, the number of classes and
the number of non-equivalent irreps are also 4. Its subgroups are
H
1
= {e, R
2
}, H
2
= {e,
v
}, H
3
= {e,
v
}
12CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
which are all invariant subgroups. The factor groups are
C
2v
/H
1
= {H
1
,
v
H
1
}, etc. (3.10)
Irreps and basis functions
Since C
2v
group has 4 classes, the number of non-equivalent irreps is also 4.
The characters and corresponding eigen functions of these irreps are given
in Table 3.2. The table consists of ve parts (columns): the 1st part is
the titles of four irreps, the 2nd the characters of each irrep, the 3rd gives
some mathematical functions that are eigen function of the group elements
with correct eigen value, the 4th and 5th list possible oxygen and hydrogen
wave functions, respectively. In part 3 the variables appearing in the math
functions are those of the coordinate system with origin at the center of
oxygen atom.
The process of calculating characters is as follows. We start with the
identity irrep (A
1
in Table) in which the character of all the group ele-
ments are 1. Then by using the orthogonality and completeness relation
other irreps (A
2
, B
1
and B
2
) can be calculated as alternating representa-
tions (consisting of 1 and -1). A
1
and B
1
are symmetric while A
2
and B
2
anti-symmetric with respect to the rotation R
2
, respectively.
Basis Functions
Irrep e R
2

v

v
math func (O) (H)
A
1
1 1 1 1 z, r, z
2
, r
2
, . . .
2pz
,
2s
, . . .
1

2
(
h
1
+
h
2
) =
(A
1
)
A
2
1 1 -1 -1 xy, x
3
y, . . .
3dxy
, . . .
B
1
1 -1 1 -1 x, xy
2
, x
3
, . . .
2px
, . . .
1

2
(
h
1

h
2
) =
(B
1
)
B
2
1 -1 -1 1 y, x
2
y, y
3
, . . .
2py
, . . .
Table 3.3: The simple characters and wave functions of C
2v
The classication of mathematical functions in the 3rd part of the ta-
ble can be understood by considering the following transformation of the
3.2. C
2V
GROUP AND H
2
O MOLECULE 13
functions:
ex = x, R
2
x = x,
v
x = x,
v
x = x (B
1
),
ey = y, R
2
y = y,
v
y = y,
v
y = y (B
2
),
er = r, R
2
r = r,
v
r = r,
v
r = r, (A
1
), (3.11)
ez = z, R
2
z = z,
v
z = z,
v
z = z, (A
1
),
ex
3
y = x
3
y, R
2
x
3
y = x
3
y,
v
x
3
y = x
3
y,
v
x
3
y = x
3
y, (A
2
)
where x, y, z are the coordinates in a system with oxygen atom at the
origin, and the x and y axes are in the v and v

plane, respectively (as is

shown in Fig.3.5). The sign in each equation shows the eigenvalue (1) of
the function under the action of a group operator. Since all the irreps of
C
2v
are all of one dimension, thus the characters are eigenvalues.
-
x
v
6
y
v

a
o
d
h
1
d
h
2
Figure 3.5: The x y plane
The symbols in parentheses at the end of each row show the irrep that
the function belongs to. It is easy to conrm that the eigen values are the
characters listed in the above table (part 2) for its corresponding irreps.
Therefore we can say that the function x, y and x
3
y have the B
2
, B
1
and
A
2
symmetry, respectively.
To construct the molecular electronic wave functions that belong to
the four irreps (or the four symmetries), we rst build wave functions for
one electron in the eld of oxygen ((O)) and that of the two hydrogens
((H)) respectively, then make coupling between them (which is called the
addition of representations). A molecular wave function made in this way
is the wave function of a single electron moving in the eld produced by
one oxygen nucleus and two hydrogen nuclei.
14CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
First consider the section of oxygen. The 6 valence electrons in oxygen
occupy the 2s and 2p orbits with the following wave functions:

2s
re

r
2ao
,
2pz
ze

r
2ao
,

2px
xe

r
2ao
,
2py
ye

r
2ao
, (3.12)
where the normalizations are neglected. Note that
3d
is too high in energy,
thus is excluded. According to eq. (3.12), those wave functions are eigen
functions of all the operators of group C
2v
.
In the hydrogen sector, we rst consider the wave functions of the two
H atoms separately. The only electron in each H atom occupies the 1s
orbit:

1s
(r
1
) e

r
1
2ao

h
1
,
1s
(r
2
) e

r
1
2ao

h
2
(3.13)
To construct group representation we consider the transformation by group
elements
e
h
1
=
h
1
, R
2

h
1
=
h
2
,
v

h
1
=
h
1
,
v

h
1
=
h
2
e
h
2
=
h
2
, R
2

h
2
=
h
1
,
v

h
2
=
h
2
,
v

h
2
=
h
1
. (3.14)
We can see that the space spanned over the basis (
h1
,
h2
) are closed under
the transformation of C
2v
thus it forms a two-dimensional representation
of this group. The representation matrices of the elements e, R
2
,
v
,
v
are
_
1 0
0 1
_
,
_
0 1
1 0
_
,
_
1 0
0 1
_
,
_
0 1
1 0
_
(3.15)
and their corresponding characters are 2, 0, 2, 0, respectively. We check the
reducibility of the 2-dimensional representation by using the orthogonality
and completeness relations derived in Ch. 2. Since

a
2

=
1
g

i
g
i
|
i
|
2
=
1
4
(2
2
+ 2
2
) = 2, (3.16)
the representation (
h
1
,
h
2
) contains two irreps. The multiplicity of each
irrep is calculated as follows
a
A
1
=
1
4
(2 1 + 0 + 2 1 + 0) = 1,
a
B
1
=
1
4
(2 1 + 0 + 2 1 + 0) = 1,
a
A
2
= a
B
2
= 0,
3.2. C
2V
GROUP AND H
2
O MOLECULE 15
which means that the representation consists in two irreps, i.e. A
1
and
B
1
. To nd the basis function of the irreps of A
1
and B
1
, in general one
needs to use the projection method (explained in section 3.3), however,
for the two-dimensional case one can see immediately that the two linear
combinations

(A
1
)
=
1

2
(
h
1
+
h
2
),

(B
1
)
=
1

2
(
h
1

h
2
), (3.17)
are the basis vectors of irrep A
1
and B
1
, respectively. They are the wave
functions of an electron in a joint eld of two hydrogen nuclei. Note that
from 2p (or even higher) state of hydrogen one can construct wave function
of the irreps A
2
and B
2
, however, these states are higher in energy, thus
are not considered in constructing low-energy molecular wave function.
In the new basis
_

(A
1
)
,
(B
1
)
_
the matrix representation of group C
2v
becomes
e =
_
1 0
0 1
_
,

R
2
=
_
1 0
0 1
_
, =
_
1 0
0 1
_
,
v
=
_
1 0
0 1
_
.
Those matrix representation can be derived through a change of basis (T):
_

(A
1
)
,
(B
1
)
_
=
_

h
1
,
h
2
_
T, thus =
1

2
_
1 1
1 1
_
. (3.18)
It is easy to show that the representation in new basis is

R
2
= T
1
R
2
T =
1

2
_
1 1
1 1
__
0 1
1 0
_
1

2
_
1 1
1 1
_
=
1
2
_
2 0
0 2
_
. (3.19)
We can see that the above process of calculating wave function of irreps
based on characters corresponds to a simultaneous diagonalization process
for operators e, R
2
,
v
,
v
. The advantage of the group theory method is
that it is much simpler than the simultaneous diagonalization of represen-
tation matrices, especially for large matrices or for large group.
We note that the operators e, R
2
,
v
,
v
plus the Hamiltonian H, form
a Complete Set of Commutable Operators (CSCO) in quantum mechanics.
16CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
For general cases the irreps are more than one dimensional and group
elements are not necessarily commutable, then the CSCO are consisting of
Hamiltonian and class operators (see Ch. 2) which are commutable with
all group elements.
Finally we make coupling between the two sections and obtain the
molecular wave functions, i.e. wave functions in a joint eld of one nucleus
O and two nuclei of H. For this we need to consider coupling between the
wave function (O) and the wave function (H). In fact coupling is a
representation addition process which is an inverse process of representa-
tion reduction. Since no group element can couple wave functions between
two non-equivalent irreps, or (
()
, R
()
)

, we only need to consider

couplings between (O) and (H) with the same symmetry (or belong to
the same irreps). Therefore the molecular wave functions are calculated in
an enlarged space with the basis (
2pz
,
(A
1
)
) or (
2py
,
(B
1
)
) for the irreps
A
1
and B
1
, respectively.
In the following we list the molecular wave functions of dierent irreps
irrep A
1
_
_
_

1
=
1

2
(
2s

(A
1
)
)

2
=
2pz
(3.20)
irrep B
1

3
=
1

2
(
2py

(B
1
)
)
Note:
i)
2pz
has little overlap with
(A
1
)
, thus there is little coupling between

1
and
2
.
ii) In
1
and
3
, + means anti-bonding, whereas - bonding (see next
subsection for explanation).
iii) Mixing between orbits with the same symmetry is not governed by
the C
2v
group, rather it is calculated by quantum mechanics. The
benet of a symmetry study is that the relevant conguration space
is much smaller.
iv) For a molecular state every electron is shared by 3 nuclei. Every
molecular orbit can have 2 electrons with opposite spin (Pauli exclu-
sion).
3.2. C
2V
GROUP AND H
2
O MOLECULE 17
Physical interpretation of the irreps
Figure 3.6 shows the electron density distribution of the two states:
(A
1
)
and
(B
1
)
, where
(A
1
)
is a bonding state, that has a electron probability
concentrated in a region between two hydrogen nuclei (marked with the
signs ) and thus its energy is lower, while
(B
1
)
is an anti-bonding
state, whose energy is higher since electron is distributed outside the above
mentioned region.

(A
1
)

h
1
+
h
2

h
1

h
2

h
1
h
2
Bonding State
(Stable)

(B
1
)

h
1

h
2

h
1
h
2
Anti-bonding State
(Unstable)
One electron is shared by the two protons (h
1
and h
2
)
Figure 3.6: Bonding and Anti-bonding states of H
2
system
The electronic spectrum of H
2
O is given in Figure 3.7: The lines in the

2px
(b
1
)

2py
(b
2
)

2pz
(a
1
)

2s
(a
1
)
3a
1
2b
1
1b
2
2a
1
1b
1
1a
1

(B
1
)

(A
1
)

h
1

h
2
Figure 3.7: Energy spectrum of H
2
O molecule
middle column of the gure represent the molecular orbits, which are
18CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
orbits of single electron moving in the Coulomb eld produced by nuclei
with electron-electron interaction being neglected. The four lowest orbits
of H
2
O molecule are shown in Figure 3.4 (energy is in eV ). Theoretical
results are from computer calculation with the molecular wave function
given in eq. (3.21). The agreement between theory and experiments is
qualitatively good.
Orbit Exp. Th.
1b
2
12.62 12.50
2a
1
13.78 14.0
1b
1
17.02 17.0
1a
1
32.2 30.0
Table 3.4: Comparison between calculation and experiment
There exist other molecules that have also C
2v
symmetry. For example
the Ethylene Oxide molecule C
2
H
4
O for which the R
2
axis passes the nu-
O
C
a
C
b
H
a
1
H
a
2
H
b
1
H
b
2
r
r r
r r r r
Figure 3.8: The C
2
H
4
O Molecule
cleus O. The atoms C
a
and C
b
are in the plane v and they are each others
image for a reexion in the plane v

. The atoms H
a
and H
b
, H
c
and H
d
are symmetric under the reection
v
.
3.3. PROJECTION OPERATOR METHOD 19
3.3 Projection operator method
For a molecule H
2
O, the decomposition of the representation (
h1
,
h2
) into

(A
1
)
and
(A
1
)
is easy. However, in general cases it is not a easy task to
decompose one representation into irreps, or to obtain wave functions with
certain symmetry. In this section we introduce the projection operator
method as a general approach.
An arbitrary wave function can be expanded in the basis functions of
all irreps:
=

i=1

()
i

()
i
, (3.21)
where the i-th basis function of the irrep has the following transformation
property:
O
R

()
i
=

()
j
D
()
ji
(R) (R G). (3.22)
The operator O
R
is the representation of the group element R in the con-
guration space. Multiplying the above equation with D
()

lm
(R) from left
and sum over R, one obtains

R
D
()

lm
(R)O
R

()
i
=

j
D
()

lm
(R)
()
j
D
()
ji
(R)
=

()
j

R
D
()

lm
(R)D
()
ji
(R)
=

()
j
g
n

lj

mi

=
g
n

()
l

mi

. (3.23)
In the following we study the two cases:
1) Let m = i and = one obtains

()
l
=
n

R
D
()

li
(R)O
R

()
i
(l = 1, . . . , n

), (3.24)
which means that from
()
i
one can produce the rest n

1 basis
functions of irrep . It is easy to show that (
()
1
, . . . ,
()
n
) form a
rep-space for the irrep .
20CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
ii) Let l = m and sum over l, one gets

R
D
()

ll
O
R

()
i
=
g
n

()
i

, (3.25)
and thus one arrives at
n

()

(R)O
R

()
i
=
()
i

. (3.26)
The above equation can be written as
P
()

()
i
=
()
i

, (3.27)
where
P
()
=
n

()

(R)O
R
(3.28)
is a projection operator that can project out components of irrep ()
from any function that is not orthogonal to all the basis functions in
the irrep . In other words, when acting on a general wave function,
the projection operator P
()
eliminates all the components of irreps
other than . Giving that is a wave function as expressed in eq.
(3.21), by acting the operator P
()
from left one gets
P
()
= P
()

i=1

()
i

()
i
=
n

i=1

()
i

()
i
, (3.29)
that is a linear combination of
()
i
, with all the components of

()
i
( = ) being eliminated. It is easy to show that
(P
()
)
2
= P
()
.
Therefore one can construct basis functions
()
i
of irrep () starting
from a known function , which is not orthogonal to all the basis
functions
()
i
(i = 1, 2, , n

). We take the group C

2v
as an exam-
ple:
P
(A
1
)

h
1
=
1
4
(O
e
+O
R
2
+O
v
+O

)
h
1
=
1
4
(
h
1
+
h
2
+
h
1
+
h
2
),
P
(B
1
)

h
1
=
1
4
(O
e
O
R
2
+O
v
O

)
h
1
=
1
4
(
h
1

h
2
+
h
1

h
2
),
3.4. C
3V
GROUP AND THE NH
3
MOLECULE. 21
where the transformation of wave function in eq. (3.14) are used.
After normalization the wave function in eq. (3.17) are reproduced.
Note that if one applies the projection operator P
()
on another wave
function (say
h
2
) the results may be dierent by a linear transfor-
mation.
3.4 C
3v
group and the NH
3
molecule.
C
3v
group
The spacial structure of NH
3
(Ammonia) molecule is shown in Figure
(3.9), where the N atom is at the origin of the coordinate system, and the
three H atoms are at the vertex of a equilateral triangle in a plane that
is perpendicular to the z-axis. While the z-axis passes through the center
(O

) of the triangle h
1
h
2
h
3
, the x-axis is parallel to the line h
1
h
2
and the y-axis parallel to the line h
3
O

. It is obvious that the frame

j
e
e
e
6

* H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Hj
z(R
3
)
x
y
N
o
o

h
1
h
3
h
2
Figure 3.9: The spacial structure of molecule NH
3
of the molecule (or the Coulomb eld produced by the nuclei) is invariant
under the following symmetry transformations:
C
3v
= {e, R
3
, R
2
3
,

}
22CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
where the transformations are (see Fig. 3.10)
e : identity
R
3
: rotation of angule
2
3

: reection in the planes , and , respectively

where the rotation R
3
is set as clockwise around the z axis. The origin (O)
is xed in space.
Figure 3.10 depicts the plane that contains the three hydrogen atoms
(h
1
, h
2
and h
3
). The three lines marked as , and represent the three
vertical planes to which the reections take place. All the three reection
planes contain the z-axis and pass one of the three hydrogen atoms. The
lines marked with (x) and (y) are parallel to the axes x and y, respectively,
that passing through the origin O.
o

r
r r
h
3
h
1
h
2

r
A

r
B

r
A
r
B

r
R
3

(y)
(x)
Figure 3.10: The symmetry plane (axes) of C
3v
group
The group table writes as
It is easy to verify that the transformation R
3

on the point a is
A A

which is equal to the action of

. However,

R
3
on a
is A B

that is identical to the action of

. This means that

R
3

R
3
, therefore C
3v
is a non-Abelian group. C
3v
has the following
subgroups:
H
1
= {e, R
3
, R
2
3
}, H
2
= {e,

}
H
3
= {e,

}, H
4
= {e,

},
3.4. C
3V
GROUP AND THE NH
3
MOLECULE. 23
e R
3
R
2
3

R
3
R
2
3
e

R
2
3
e R
3

e R
3
R
2
3

R
2
3
e R
3

R
3
R
2
3
e
Table 3.5: Group table of C
3v
in which H
1
is an invariant subgroup, since

H
1
=

{E, R
3
, R
2
3
} = {

}
H
1

= {E, R
3
, R
2
3
}

= {

} =

H
1
.
But H
2
, H
3
, H
4
are not invariant. For example, we have
R
3
H
2
= {R
3
,

} while H
2
R
3
= {R
3
,

R
3
} = {R
3
,

}. = R
3
H
2
.
One of the factor groups constructed with H
1
is
C
3v
/H
1
= {H
1
,

H
1
},
which is isomorphic with the subgroups H
2
, H
3
or H
4
. In the same way
one can construct the other factor groups C
3v
/H
2
and C
3v
/H
3
.
Irreps and basis functions
Group C
3v
has three classes:
{e}, {R
3
, R
2
3
}, {

},
thus it has three non-equivalent irreps. For a regular representation one
has

n
2

= 6, therefore the three irreps have dimension 1, 1 and 2, re-

spectively. The characters of C
3v
group is given in Table 3.6. where the
number in parentheses following R
3
and

is the number of elements in

the corresponding class, and the characters in the column under e are also
the dimension of corresponding irreps. The irrep A
1
is an identity irrep
which has number 1 as character for all the classes. A
2
is an alternating
irrep with elements 1 or -1. With orthogonality relation between irreps
24CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
Classes
Irreps e R
3
(2)

(3)
A
1
1 1 1
A
2
1 1 -1
E 2 -1 0
Table 3.6: Character table of group C
3v
one can determine the character for A
2
. Then it is easy to determine the
characters for the 2-dimensional irrep E.
To obtain the molecular orbits we follow the same strategy as for
molecule H
2
O. First we make classication for the wave functions in the
eld of nitrogen ((N)) and those in the eld of three H-atoms ((H)) in
a separate way. The results are listed in Table 3.7. Then make coupling
between the wave functions of (N) and (H) with the same symmetry.
The calculation is described in the following.
Basis Functions
Irreps (N) (H)
A
1

2s
,
2pz
1

3
(
h
1
+
h
2
+
h
3
)
(A
1
)
A
2

(A
2
)
E
_

2px
,
2py
_
_
1

2
(
h
2

h
3
),
1

6
(2
h
1

h
2

h
3
)
_

(E)
1
,
(E)
2
_
Table 3.7: Classication of orbits for nitrogen and hydrogens
Basis functions of N-atom ((N)): N-atom has 7 electron with 5 in
the valence shell, that occupy the orbits
2s
,
2px
,
2py
and
2pz
(maximum
number of electrons is 8). As before only the orbits with lowest energy are
considered.
i) Irrep A
1
: A
1
is an identity irrep, the basis functions can be any one
from

2s
re
r/2a
0
and
2pz
ze
r/2a
0
,
3.4. C
3V
GROUP AND THE NH
3
MOLECULE. 25
which are invariant (with eigenvalue equal to 1) under all operations
of C
3v
.
ii) A
2
is an alternating irrep of the factor group C
3v
/C
3
= {C
3
,

C
3
}
According to Figure 3.10 the application of the C
3v
elements has the
following eects:
R
3
: +
2
3

R
2
3
: +
4
3

:
5
3

:
1
3

in which is the angle between the vector r and the (x)-axis .
To project out the basis function of A
2
, we start from a function
= e
in
. For the equation
O
R
3
e
in
= e
in
to be valid there must be
n = 3m (m = 1, 2, ).
Applying the projection operator P
(A
2
)
(eq. (3.28)) we get

(A
2
)
=
n

(A
2
)

O
T
=
1
6
(O
e
+O
R
3
+O
R
2
3
O

)
=
1
6
{e
i3m
+ e
i3m(+
2
3
)
+ e
i3m(+
4
3
)
e
i3m(
5
3
)
e
i3m()
e
i3m(
1
3
)
}
=
1
2
(e
i3m()
+ e
i3m()
) = cos 3m. (3.30)
This wave function has an energy as high as that of the
3f
state,
thus it will not be considered for the lowest energy spectrum.
iii) Irrep E (2-dimensional)
26CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES

x
R
3

6
-

H
H
H
H
H
H
H
Hj

6
e
y
e
x
Figure 3.11: The R
3
transformation
To derive the irrep E (two-dimensional) of the group C
3v
, we consider
the space spanned over the basis (e
x
, e
y
), in which e
x
and e
y
are unit vectors,
and a general vector is expressed as r = xe
x
+ ye
y
. The transformation of
R
3
on the basis vectors writes
R
3
e
x
=

e

x
=
1
2
e
x

3
2
e
y
,
R
3
e
y
=

e

y
=

3
2
e
x

1
2
e
y
. (3.31)
The action of R
3
on a general vector r is
R
3
r = xR
3
e
x
+ yR
3
e
y
= x
_

1
2
e
x

3
2
e
y
_
+ y
_

3
2
e
x

1
2
e
y
_
=
_

1
2
x +

3
2
y
_
e
x
+
_

3
2
x
1
2
y
_
e
y
Therefore the transformation of R
3
in the conguration space (with the
basis (x, y) writes as follows,
R
3
x =
1
2
x +

3
2
y, R
3
y =

3
2
x
1
2
y. (3.32)
This means that the two functions x and y transform between themselves,
thus (x, y) form a representation space. The matrix representation of R
3
3.4. C
3V
GROUP AND THE NH
3
MOLECULE. 27
and R
2
3
in this space is
O
R
3
=
_

1
2

3
2
3
2

1
2
_
and O
R
2
3
=
_

1
2

3
2

3
2

1
2
_
, (3.33)
respectively
The transformation of

is
O

x = x and O

y = y, (3.34)
therefore we have
O

=
_
1 0
0 1
_
. (3.35)
By using the group table, one obtains all the matrices of the group
elements
O

= O
R
3
O

=
_

1
2

3
2
3
2

1
2
_
_
1 0
0 1
_
=
_
1
2

3
2

3
2

1
2
_
,
O

= O

O
R
3
=
_
1 0
0 1
_
_

1
2

3
2
3
2

1
2
_
=
_
1
2

3
2
3
2

1
2
_
.
It is easy to see that the characters of the classes e, R
3
,

are 2, -1, 0,
respectively, therefore this representation is the irrep E. It is interesting to
note that the sum of all operators of a class (class operator) are diagonal:
O
R
3
+O
R
2
3
=
_
1 0
0 1
_
, O

+O

+O

=
_
0 0
0 0
_
,
O
e
=
_
1 0
0 1
_
.
This is consistent with the Schurs Lemma II, since all the class operators
commute with all the group elements.
Basis function for three H-atom ((H)): Applying group elements
on
h
1
,
h
2
and
h
3
one obtains the results :
e R
3
R
2
3

h
1

h
1

h
2

h
3

h
1

h
2

h
3

h
2

h
2

h
3

h
1

h
3

h
1

h
2

h
3

h
3

h
1

h
2

h
2

h
3

h
1
28CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
It is easy to see from the table that the space spanned over the basis
(
h
1
,
h
2
,
h
3
) is a rep-space of C
3v
. The matrix representation of the
elements e, R
3
and

are
O
e
=
_
_
1 0 0
0 1 0
0 0 1
_
_
, O
R
3
=
_
_
0 0 1
1 0 0
0 1 0
_
_
, O

=
_
_
1 0 0
0 0 1
0 1 0
_
_
, (3.36)
respectively. The corresponding characters for the three classes are (3, 0,
1). One calculates the

a
2

=
1
g

i
g
i

i
= 2 which means that the
space (
h
1
,
h
2
,
h
3
) is reducible. The multiplicity in decomposition can
be calculated as:
a
A
1
=
1
6
(1 1 3 + 2 1 0 + 3 1 1) = 1,
a
A
2
=
1
6
(1 1 3 + 2 1 0 + 3 (1) 1) = 0,
a
E
=
1
6
(1 2 3 + 2 (1) 0 + 3 0 1) = 1,
(3.37)
thus the representation D = D
(A
1
)
D
(E)
.
Now we can use the projection operator method to calculate the basis
functions (H) for the irrep A
1
and E. The projection operator writes
P
()
=
n

()
(R)O
R
,
where stands for the irreps A
1
, A
2
and E, respectively.
For the irrep A
1
, the result is
P
(A
1
)

h
1
(
h
1
+
h
2
+
h
3
+
h
1
+
h
2
+
h
3
)
h
1
+
h
2
+
h
3
.
For the irrep A
2
, the result is
P
(A
2
)

h
1

h
1
+
h
2
+
h
3

h
1

h
2

h
3
= 0,
which means that the space of (
h
1
,
h
2
,
h
3
) contains no component of
the symmetry A
2
. Note that the missing of A
2
irrep here is due to our
restriction on lowest energy states. If one considers as basis function the
3.4. C
3V
GROUP AND THE NH
3
MOLECULE. 29
hydrogen states with higher energies, the operator P
(A
2
)
may project out
components of A
2
irrep. For the irrep E we can have
P
(E)

h
1

2
6
(2
h
1

h
2

h
3
)(
1
)
P
(E)

h
2

2
6
(2
h
2

h
3

h
1
)(

2
)
Let

2
=
1
+ 2

2
=
1
2
(
h
2

h
3
),
then one can have an orthonormal basis for the irrep E

(E)
(H)
_
1
6
(2
h
1

h
2

h
3
),
1

2
(
h
2

h
3
)
_
.
The wave functions of molecule NH
3
: After coupling between (N)
and (H), the molecular orbits are obtained as given in the gure 3.12. It

2px
,
2py
(E)

2pz
(A
1
)

2ps
(A
1
)

(E)
6
?
6
?

(A
1
)
2
6
?

(A
1
)
1
6
?

(E)
1
,
(E)
2

(A
1
)
Figure 3.12: Orbits of molecule NH
3
is worth mentioning that the calculated molecular wave functions are eigen
states of a single electron that moves in the Coulomb eld produced by four
nuclei (one of nitrogen and three of hydrogens). Nitrogen atom has seven
electrons with two being in 1s shell and 5 in valence shell (2p
x
, 2p
y
, 2p
z
, 2s),
whereas each hydrogen atom has one electron. Therefore eight valence elec-
trons are shared by the three protons and the nitrogen nucleus, and one
lls the electrons into the molecular orbits in pairs: with one spin-up and
30CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
Orbits Exp. Theo.

(2)
A
1
-10.2 -11.6

E
-15.0 -17.1

(1)
A
1
-27.0 -31.3
Table 3.8: Low energy spectrum of Molecule NH
3
one spin-down. Thus four orbits are occupied in the order of energy from
low to high:
Calculated energy of molecular states (electronic energy) are given in table
3.8 (in units of eV ). Theoretical results are from computer diagonalization,
which agrees with the experiments reasonably well. We note that for the
coupling between two sections (N) and (H) needs the information of the
wave functions which is beyond the symmetry consideration of group the-
ory. However, a big advantage of the group theory method is that due to
coupling only happens between irreps with the same symmetry, the dimen-
sion of the space in which the coupling takes place is very much reduced.
Decomposition of a three-dimensional representation for (N).
To further demonstrate the use of projection operator, we decompose a
three-dimensional representation (x
2
, xy, y
2
) (in the nitrogen section) into
irreps. By using the transformations eqs (3.32,3.34), we can have
R
3
x
2
=
1
4
x
2

3
2
xy +
3
4
y
2
,
R
3
xy =

3
4
x
2

1
2
xy

3
4
y
2
,
R
3
y
2
=
3
4
x
2
+

3
2
xy +
1
4
y
2
,

x
2
= x
2
,

xy = xy,

y
2
= y
2
.
In the three-dimensional conguration space (x
2
, xy, y
2
) the matrix repre-
sentation of C
3v
group is as follows:
O
R
3
=
_
_
_
1
4

3
4
3
4

3
2

1
2

3
2
3
4

3
4
1
4
_
_
_
,

=
_
_
1 0 0
0 1 0
0 0 1
_
_
3.4. C
3V
GROUP AND THE NH
3
MOLECULE. 31
O
R
2
3
=
_
_
_
1
4

3
4
3
4
3
2

1
2

3
2
3
4

3
4
1
4
_
_
_
, O

= O
R
3
O

=
_
_
_
1
4

3
4
3
4

3
2
1
2

3
2
3
4

3
4
1
4
_
_
_
O

= O

O
R
3
=
_
_
_
1
4

3
4
3
4
3
2
1
2

3
2
3
4

3
4
1
4
_
_
_
. (3.38)
One can easily verify that now the class operators are not diagonal, this
is because that the representation here is not an irrep. To decompose the
representation into irreps we rst calculate the characters, which are 3, 0,
1 for the lasses e, R
3
and

, respectively. Due to that

|a

|
2
=
1
g

i
g
i

i
=
1
6
(1 9 + 0 + 3 1) = 2,
this representation is a composite one, containing two irreps, as is shown
in the following calculation:
a
A
1
=
1
6
(1 1 3 + 0 + 3 1 1) = 1
a
A
2
=
1
6
(1 1 3 + 0 3 1 1) = 0
a
E
=
1
6
(1 2 3 + 0 + 0) = 1.
Therefore the decomposition is
D(R) = D
(A
1
)
+ D
(E)
.
To derive the basis function of each irrep we use the projection operator
eq. (3.28). First, the action of all the elements of C
3v
group, O
R
, on x
2
and xy are listed in the following table:
f = x
2
f = xy
e f x
2
xy
R
3
f
1
4
x
2

3
2
xy +
3
4
y
2

3
4
x
2

1
2
xy

3
4
y
2
R
2
3
f
1
4
x
2
+

3
2
xy +
3
4
y
2

3
4
x
2

1
2
xy +

3
4
y
2

f
1
4
x
2

3
2
xy +
3
4
y
2

3
4
x
2
+
1
2
xy +

3
4
y
2

f x
2
xy

f
1
4
x
2
+

3
2
xy +
3
4
y
2

3
4
x
2
+
1
2
xy

3
4
y
2
32CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
The wave function of irrep A
1
can be calculated by P
(A
1
)
x
2
:

(A
1
)
= P
(A
1
)
x
2
=
1
4
_
e + R
3
+ R
2
3
+

_
x
2
=
1
6
(3x
2
+ 0 xy + 3y
2
) =
1
2
(x
2
+ y
2
).
Two of the function of irrep E
2
can be obtained by applying the projection
operator to x
2
and xy, respectively:

(E
2
)
1
= P
(E
2
)
x
2
=
1
4
_
2e R
3
R
2
3
_
x
2
=
3
8
(x
2
y
2
)

(E
2
)
2
= P
(E
2
)
xy =
1
4
_
2e R
3
R
2
3
_
xy = 3xy.
Therefore the basis for the irrep E is (x
2
y
2
, xy).
3.4. C
3V
GROUP AND THE NH
3
MOLECULE. 33
Problems
1. a) Give examples of functions which belong to the representations A
g
and B
g
of the group C
2h
. Classify the spherical harmonics P
m
l
()e
im
ac-
cording to the representations of C
2h
.
b) Classify the components of an axial vector according to the representa-
tions of C
2h
.
For you reference, the character table is given below:
Irrep e R
2

h
I
A
g
1 1 1 1
A
u
1 1 -1 -1
B
g
1 -1 -1 1
B
u
1 -1 1 -1
where the element R
2
is a rotation of around the z-axis,
h
a reection
in the plane perpendicular to the z-axis and passing the origin, and I the
Inversion.
2. Find the matrices of the irreducible representations of C
3v
. Find func-
tions belonging to each of the representations.
3. Dihedral group D
3
is a symmetry group of a equilateral triangle. The
group has 6 elements: the identity, three reections about axes and two
rotations around the center.
a) Enumerate the group elements and the classes.
b) Enumerate the subgroups, the invariant subgroups and the factor groups.
c) Construct the character table for all the irreducible representations.
d) Let L be the 2-dimensional Euclidean space spanned by e
x
and e
y
. Write
down the matrix representation of elements of D
3
on L with respect to this
Cartesian basis.
e) If h
1
, h
2
and h
3
are three identical atoms at the three vertices of the
triangle, respectively. Show that the functions
h
1
,
h
2
and
h
3
form a
representation of the D
3
group. Find the matrix representation and the
characters of this representation.
f) Determine the reducibility of this representation. If decomposable, nd
the component irreps and corresponding functions.
34CHAPTER 3. POINT GROUPS ANDSYMMETRIES OF MOLECULES
4. Dihedral group D
4
is the symmetry group of a square. The group has
8 elements: the identity, 3 rotations around the center and 4 reections
about the vertical, the horizontal and the diagonal axes.
a) Enumerate the group elements and the classes.
b) Enumerate the subgroups, the invariant subgroups and the factor groups.
c) Construct the character table for all the irreducible representations.
d) Give wave functions that belong to each of the irrep.
Note: The center of the square (O) is set as the origin of the coordinate
system, z-axis is perpendicular to the plane of the square and it passes the
origin.
v
1
v
2

2
@
@
@
@
@
@
@
@
@
@
@
@
v

1
b
O
Figure 3.13: The square for constructing the D
4
group