1

1 Properties of Spherical Harmonics

1.1 Repetition

In the lecture the spherical harmonics Y ,, m(q, f) were introduced as the

2
eigenfunctions of angular momentum operators l̂ z and l̂ in spherical coordinates.
We found that

l̂ z Y ,, m(q, f) = "mY ,, m(q, f) [1.1]

and

2
l̂ Y ,, m(q, f) = ", ( , + 1 )Y ,, m(q, f) . [1.2]

The spherical harmonics can be defined as

m imf
Y ,, m(q, f) = N ,, m ⋅ P , ( cos q) ⋅ e [1.3]

where , is the quantum number of the orbital angular momentum and m the
magnetic quantum number. There are analytical definitions for the normalization
m
factor N ,, m and the associated Legendre Polynomials P , ( cos q) that allow the
calculations of the spherical harmonics.

The spherical harmonics for , = 0, 1, and 2 are given by

1
Y 0, 0(q, f) = ----------- [1.4]
4p

1 3
Y 1, 0(q, f) = --- --- cos q
2 p
[1.5]
1 3 ± if
Y 1, ±1(q, f) = −
+ --- ------ sin q e
2 2p
2 Chapter 1

1 5 2
Y 2, 0(q, f) = --- --- ( 3cos q – 1 )
4 p
1 15 ± if
Y 2, ±1(q, f) = −
+ --- ------ sin q cos qe . [1.6]
2 2p
1 15 2 ±2if
Y 2, ±2(q, f) = --- ------sin qe
4 2p

Note, that the sign of the functions Y 1, ±1(q, f) and Y 2, ±1(q, f) is defined differently
than in the script of the lecture. The definition here is in agreement with most of the
literature on spherical harmonics.

1.2 Graphical Representation of Spherical Harmonics

The spherical harmonics are often represented graphically since their linear
combinations correspond to the angular functions of orbitals. Figure 1.1a shows a plot
of the spherical harmonics where the phase is color coded. One can clearly see that
Y ,, m(q, f) is symmetric for a rotation about the z axis. The linear combinations
m
1 ⁄ 2 ( Y ,, m(q, f) + ( – 1 ) Y ,, –m(q, f) ) = 2 Y ,, m(q, f) cos ( mf ) , Y ,, 0(q, f) and
m
– i ⁄ 2 ( Y ,, m(q, f) – ( – 1 ) Y ,, –m(q, f) ) = 2 Y ,, m(q, f) sin ( mf ) are always real and
have the form of typical atomic orbitals that are often shown.

1.3 Properties of Spherical Harmonics

There are some important properties of spherical harmonics that simplify

working with them.

1.3.1 Orthogonality and Normalization

The spherical harmonics are normalized and orthogonal, i.e.,

2p p
*
∫ ∫ Y , , m (q, f)Y , , m (q, f) sin q ( dq ) df
1 1 2 2
= d m1, m2 d ,1, ,2 [1.7]
0 0

where the Kronecker delta is defined as

 Properties of Spherical Harmonics 3

a) m = –4 m = –3 m = –2 m = –1 m = 0 m = 1 m = 2 m = 3 m = 4
-π

, = 0

-π/2

, = 1
0

, = 2
π/2

, = 3 π

, = 4

b) m = 4 m = 3 m = 2 m = 1 m = 0 m = 1 m = 2 m = 3 m = 4
-π

, = 0

-π/2

, = 1
0

, = 2
π/2

, = 3 π

, = 4

Figure 1.1: Graphical Representation of the Spherical Harmonics

a) Plot of the spherical harmonics Y ,, m(q, f) where the phase of the function is color coded.
Note that Y ,, m(q, f) is always axially symmetric with respect to a rotation about the z axis
since it depends only on the angle q . The phase of the function changes with a periodicity of
m
m . b) The linear combinations 1 ⁄ 2 ( Y ,, m(q, f) + ( – 1 ) Y ,, –m(q, f) ) = 2 Y ,, m(q, f) cos ( mf ) ,
m
Y ,, 0(q, f) , and – i ⁄ 2 ( Y ,, m(q, f) – ( – 1 ) Y ,, –m(q, f) ) = 2 Y ,, m(q, f) sin ( mf ) are all real
and show only a phase of 0 (positive) and p (negative) and correspond to the typical orbital
shapes.
4 Chapter 1

⎧0 a≠b
d a, b = ⎨ . [1.8]
⎩1 a = b

They form a complete basis set of the Hilbert space of square-integrable functions, i.e.,
every such function can be expressed as a linear combination of spherical harmonics

∞ ,
f (q, f) = ∑ ∑ f ,, m Y ,, m(q, f) . [1.9]
, = 0 m = –,

The coefficients f ,, m can be calculated as

2p p
*
f ,, m = ∫ ∫ f (q, f)Y ,, m(q, f) sin q ( dq ) df . [1.10]
0 0

1.3.2 Product of Two Spherical Harmonics

Since the spherical harmonics form a orthonormal basis set, the product of two
spherical harmonics can again be expressed in spherical harmonics. Let us first look at
a simple example

1 3 1 3 3 2
Y 1, 0(q, f) ⋅ Y 1, 0(q, f) = --- --- cos q ⋅ --- --- cos q = ------cos q . [1.11]
2 p 2 p 4p

Comparing this to the spherical harmonics of Eqs. [1.4]-[1.6] it is immediately clear

that we need the functions Y 2, 0(q, f) and Y 0, 0(q, f) to express the product. We can
make an Ansatz

Y 1, 0(q, f) ⋅ Y 1, 0(q, f) = c 0, 0 Y 0, 0(q, f) + c 2, 0 Y 2, 0(q, f) [1.12]

which leads to

3 2 1 1 5 2
------cos q = c 0, 0 ----------- + c 2, 0 --- --- ( 3cos q – 1 )
4p 4p 4 p
. [1.13]
1 1 5 3 5 2
= c 0, 0 ----------- – c 2, 0 --- --- + c 2, 0 --- --- cos q
4p 4 p 4 p

From this it is immediately clear that

 Properties of Spherical Harmonics 5

3 4 p 1
c 2, 0 = ------ ⋅ --- --- = ----------- [1.14]
4p 3 5 5p

and

1 1 5 1
c 0, 0 = ----------- --- --- ⋅ 4p = ----------- . [1.15]
5p 4 p 4p

For a general product this is of course more complicated but there are a few
simple rules for the general product Y ,1, m1(q, f) ⋅ Y ,2, m2(q, f) . Since the dependence
on f is always given by exp ( imf ) , it is immediately clear that the product function
has to have the magnetic quantum number M = m 1 + m 2 . Using similar arguments,
the orbital angular quantum number can be limited to the range
, 1 – , 2 ≤ L ≤ , 1 + , 2 . In principle, it is not important to know these restrictions since
the Clebsch-Gordan coefficients (or the Wigner 3j symbols) will do the selection
automatically.

The product can in general be written as the following linear combination

( 2, 1 + 1 ) ( 2, 2 + 1 ) ( 2L + 1 )
Y ,1, m1(q, f) ⋅ Y ,2, m2(q, f) = ∑ -------------------------------------------------------------------
4p
L, M
[1.16]
⎛ , , L ⎞ * ⎛ , , L⎞
×⎜ 1 2 ⎟ Y L, M(q, f) ⎜ 1 2 ⎟
⎝ m1 m2 M ⎠ ⎝ 0 0 0⎠

where the Wigner 3j symbols are related to the Racah or Clebsch-Gordan coefficients
by

⎛ ,1 ,2 L ⎞ 1
⎜ ⎟ = ( – 1 ) ,1 – ,2 – M ------------------------- c ( , 1, m 1, , 2, m 2, L, – M ) . [1.17]
⎝ m1 m2 M ⎠ ( 2L + 1 )

The Wigner 3j symbols or the Clebsch-Gordan coefficients can be found in tables in

books about angular momentum or calculated using programs like Matlab, Macsyma,
or Mathematica. Written with Clebsch-Gordan coefficients we obtain for Eq. [1.16]
6 Chapter 1

( 2, 1 + 1 ) ( 2, 2 + 1 )
Y ,1, m1(q, f) ⋅ Y ,2, m2(q, f) = ∑ ---------------------------------------------- Y L, M(q, f)
4p ( 2L + 1 ) . [1.18]
L, M
× c ( , 1, m 1, , 2, m 2, L, M )c ( , 1, 0, , 2, 0, L, 0 )

To calculate the coefficients for the above example, we need the Wigner 3j
symbols for

⎛ 110⎞ 1 ⎛ 1 1 1⎞ ⎛ 1 1 2⎞ 2
⎜ ⎟ = – ------- ⎜ ⎟ = 0 ⎜ ⎟ = ------ [1.19]
⎝ 000⎠ 3 ⎝ 0 0 0⎠ ⎝ 0 0 0⎠ 15

and obtain

2 2
( 2L + 1 ) ⎛ ⎞
Y 1, 0(q, f) ⋅ Y 1, 0(q, f) = ∑ 3 -------------------- ⎜ 1 1 L ⎟ Y L, 0(q, f)
4p ⎝ 0 0 0 ⎠
L=0 . [1.20]
1 1
= ----------- Y 0, 0(q, f) + 0Y 1, 0(q, f) + ----------- Y 2, 0(q, f)
4p 5p

In the same way more complex products can be calculated and decomposed in the
spherical harmonics. This is an iterative way to calculate the functional form of
higher-order spherical harmonics from the lower-order ones. We will discuss this in
more detail in an exercise.

1.3.3 Addition Theorem of Spherical Harmonics

The spherical harmonics obey an addition theorem that can often be used to
simplify expressions

,
2, + 1
∑
*
Y ,, m(q 1, f 1)Y ,, m(q 2, f 2) = ---------------P ,( cos w) [1.21]
4p
m = –,

where w omega describes the angle between two unit vectors oriented at the polar
coordinates ( q 1, f 1 ) and ( q 2, f 2 ) with

cos w = cos q 1 cos q 2 + sin q 1 sin q 2 cos ( f 1 – f 2 ) . [1.22]

 Properties of Spherical Harmonics 7

1.3.4 Integrals Over Spherical Harmonics

The integration over the product of three spherical harmonics can be simplified
using the product rule of Eq. [1.16] and the orthogonality of Eq. [1.7]. This leads to

2p p

∫ ∫ Y , , m (q, f)Y , , m (q, f)Y , , m (q, f) sin q ( dq ) df

1 1 2 2 3 3
0 0

( 2, 1 + 1 ) ( 2, 2 + 1 ) ( 2L + 1 ) ⎛ , 1 , 2 L ⎞ ⎛ , , L ⎞
= ∑ -------------------------------------------------------------------
4p
⎜ ⎟⎜ 1 2 ⎟
⎝ m1 m2 M ⎠ ⎝ 0 0 0 ⎠
L, M
[1.23]
2p p
*
× ∫ ∫ Y L, M(q, f)Y , , m (q, f) sin q ( dq ) df
3 3
0 0

( 2, 1 + 1 ) ( 2, 2 + 1 ) ( 2, 3 + 1 ) ⎛ , 1 , 2 , 3 ⎞ ⎛ , , , ⎞
= --------------------------------------------------------------------- ⎜ ⎟⎜ 1 2 3 ⎟
4p ⎝ m1 m2 m3 ⎠ ⎝ 0 0 0 ⎠

a simple expression involving only a normalization constant and two Wigner 3j

symbols.

1.4 Literature

(1) D. M. Brink, G. R. Satchler, Angular Momentum, third edition, Clarendon Press,

1993.

(2) A. R. Edmonds, Angular Momentum in Quantum Mechanics, Princeton

University Press, 1960.

(3) M. E. Rose, Elementary Theory of Angular Momentum, John Wiley & Sons Inc.,
New York, 1957.
8 Chapter 1