Sunteți pe pagina 1din 9

CHM 532

Notes on Angular Momentum Eigenvalues and


Eigenfunctions

In your textbooks, the eigenfunctions and eigenvalues of the angular momentum operators
are determined using differential equations methods. A more powerful approach is to solve
the angular momentum eigenfunction problem using operator methods analogous to the
creation and annihilation operators we have used in the harmonic oscillator problem. These
notes provide details about the operator approach.

1 Preliminaries
We begin by reviewing the angular momentum operators and their commutation relations.
The detailed derivation of these preliminary results can be found in your textbooks. Recall
that the angular momentum is a vector with three components, and there is a Hermitian
operator associated with each component
!
h̄ ∂ ∂
L̂x = y −z , (1)
i ∂z ∂y
!
h̄ ∂ ∂
L̂y = z −x , (2)
i ∂x ∂z
and !
h̄ ∂ ∂
L̂z = x −y . (3)
i ∂y ∂x
Additionally, it is convenient to define the operator corresponding to the square of the length
of the angular momentum vector
L̂2 = L̂2x + L̂2y + L̂2z . (4)
Because the angular momentum is a conserved quantity for systems having spherically sym-
metric potentials, it is important to express these operators as well in spherical polar coor-
dinates !
h̄ ∂ ∂
L̂x = − sin φ − cot θ cos φ , (5)
i ∂θ ∂φ

1
!
h̄ ∂ ∂
L̂y = cos φ − cot θ sin φ , (6)
i ∂θ ∂φ
h̄ ∂
L̂z = , (7)
i ∂φ
and
1 ∂2
!
2 2 1 ∂ ∂
L̂ = −h̄ sin θ + . (8)
sin θ ∂θ ∂θ sin2 θ ∂φ2
It is important to recognize that in spherical polar coordinates, all the angular momentum
operators are independent of the coordinate r, and the eigenfunctions must be functions only
of the angular coordinates; i.e. θ and φ.
Using the operator expressions in either coordinate system (it is most easy to use Carte-
sian coordinates), it is possible to demonstrate the following commutators (see the text for
details)
[L̂x , L̂y ] = ih̄L̂z , (9)
[L̂y , L̂z ] = ih̄L̂x , (10)
[L̂z , L̂x ] = ih̄L̂y , (11)
and
[L̂i , L̂2 ] = 0 (12)
for i = x, y or z. From the commutation relations, it is clear that we can know only one
component of the angular momentum and the square of the angular momentum vector at
the same time. Because L̂z has a particularly simple form in spherical polar coordinates, we
choose to know Lz and L2 simultaneously; i.e. we seek simultaneous eigenfunctions of L̂z
and L̂2 .

2 The Eigenvalues
In the case of the harmonic oscillator, we discovered the eigenvalues of the Hamiltonian by
introducing creation and annihilation operators. In analogy, we make the following defini-
tions:

Definition 1 The step-up operator is defined by

L̂+ = L̂x + iL̂y . (13)

Definition 2 The step-down operator is defined as the adjoint of the step-up operator; i.e.

L̂− = L̂†+ = L̂x − iL̂y . (14)

2
We next introduce and prove a series of lemmas from which we can extract the eigenvalues of
L̂z and L̂2 . We let {Yl,m } represent the common complete orthonormal set of eigenfunctions
of L̂z and L̂2 with m and l respectively the quantum numbers associated with each operator.
We write
L̂z Yl,m = mh̄Yl,m (15)
and
L̂2 Yl,m = f (l)h̄2 Yl,m (16)
where f (l) is some function of the l quantum number the explicit expression for which is
derived below. It is easily verified that the units of h̄ are units of angular momentum, and
we have included factors of h̄ explicitly in Eqs.(15) and (16). The orthonormality condition
is written
hYl,m |Yl0 ,m0 i = δl,l0 δm,m0 (17)
We now introduce and provide proofs for the necessary lemmas.

Lemma 1
L̂z L̂+ = L̂+ (L̂z + h̄). (18)

Proof:
L̂z L̂+ = L̂z (L̂x + iL̂y ) (19)
= L̂x L̂z + [L̂z , L̂x ] + i(L̂y L̂z + [L̂z , L̂y ]) (20)
= L̂x L̂z + ih̄L̂y + i(L̂y L̂z − ih̄L̂x ) (21)
= (L̂x + iL̂y )L̂z + h̄(L̂x + iL̂y ) = L̂+ (L̂z + h̄). (22)
2

Lemma 2
L̂z L̂− = L̂+ (L̂z − h̄). (23)

Proof:
L̂z L̂− = L̂z (L̂x − iL̂y ) (24)
= L̂x L̂z + [L̂z , L̂x ] − i(L̂y L̂z + [L̂z , L̂y ]) (25)
= L̂x L̂z + ih̄L̂y − i(L̂y L̂z − ih̄L̂x ) (26)
= (L̂x − iL̂y )L̂z − h̄(L̂x − iL̂y ) = L̂− (L̂z − h̄). (27)
2

Lemma 3
L̂+ Yl,m = c+ (l, m)Yl,m+1 (28)
where c+ (l, m) is a function only of the quantum numbers l and m.

3
Proof: We must show that L̂+ Yl,m is an eigenfunction of L̂2 and L̂z with the value of the z
quantum number increased by one unit of h̄. Notice that L̂+ commutes with L̂2 so that the
l quantum number must be unaffected by the action of L̂+ . Next

L̂z (L̂+ Yl,m ) = L̂+ (L̂z + h̄)Yl,m (29)

= L̂+ (m + 1)h̄Yl,m = (m + 1)h̄(L̂+ Yl,m ). (30)


2
Lemma 4
L̂− Yl,m = c− (l, m)Yl,m−1 (31)
where c− (l, m) is a function only of the quantum numbers l and m.
Proof: We must show that L̂− Yl,m is an eigenfunction of L̂2 and L̂z with the value of the z
quantum number decreased by one unit of h̄. Notice that L̂− commutes with L̂2 so that the
l quantum number must be unaffected by the action of L̂− . Next

L̂z (L̂− Yl,m ) = L̂− (L̂z − h̄)Yl,m (32)

= L̂− (m − 1)h̄Yl,m = (m − 1)h̄(L̂− Yl,m ). (33)


2
Lemma 5
f (l) ≥ m2 . (34)
Proof: We begin with

hYl,m |L̂− L̂+ |Yl,m i = hL̂+ Yl,m |L̂+ Yl,m i ≥ 0. (35)

Next,
L̂− L̂+ = (L̂x − iL̂y )(L̂x + iL̂y ) = L̂2x + L̂2y + i(L̂x L̂y − L̂y L̂x ) (36)
= L̂2x + L̂2y − h̄L̂z = L̂2 − L̂2z − h̄L̂z . (37)
Then
hYl,m |L̂− L̂+ |Yl,m i = f (l)h̄2 − m2 h̄2 − h̄2 m = [f (l) − m(m + 1)]h̄2 ≥ 0 (38)
or
f (l) − m(m + 1) ≥ 0. (39)
Next,
hYl,m |L̂+ L̂− |Yl,m i = hL̂− Yl,m |L̂− Yl,m i ≥ 0. (40)
Also,
L̂+ L̂− = (L̂x + iL̂y )(L̂x − iL̂y ) = L̂2x + L̂2y − i(L̂x L̂y − L̂y L̂x ) (41)
= L̂2x + L̂2y + h̄L̂z = L̂2 − L̂2z + h̄L̂z . (42)

4
Then
hYl,m |L̂+ L̂− |Yl,m i = f (l)h̄2 − m2 h̄2 + h̄2 m = [f (l) − m(m − 1)]h̄2 ≥ 0 (43)
or
f (l) − m(m − 1) ≥ 0. (44)
Adding Eqs.(39) and (44), we obtain

f (l) − m2 ≥ 0 (45)

or
f (l) ≥ m2 . (46)
2

An implication of Eq.(46) is that for a given l, there exits a maximum value of m which we
denote m and a minimum value of m which we denote m. The next lemma relates m to m.

Lemma 6
m = −m (47)

Proof:
L̂+ Yl,m = L̂− Yl,m = 0. (48)
hYl,m |L̂− L̂+ |Yl,m i = hYl,m |L̂2 − L̂2z − h̄L̂z |Yl,m i = [f (l) − m(m + 1)]h̄2 = 0. (49)
hYl,m |L̂+ L̂− |Yl,m i = hYl,m |L̂2 − L̂2z + h̄L̂z |Yl,m i = [f (l) − m(m − 1)]h̄2 = 0. (50)
Then
f (l) = m(m + 1) = m(m − 1). (51)
Solving for m in terms of m, we obtain

m = −m. (52)

From Eq.(49), we have


f (l) = m(m + 1). (53)
We define l = m; i.e. the maximum value of m for a given l, and we can write

f (l) = l(l + 1). (54)

In the next lemma, we identify what values l can have.

Lemma 7 The quantum number l can be an integer or a 1/2 integer.

5
Proof: We have the following series of equations
L̂− Yl,l = c− (l, l)Yl,l−1 (55)

L̂− Yl,l−1 = c− (l, l − 1)Yl,l−2 (56)


..
.
L̂− Y (l, −l + 1) = c− (l, −l + 1)Yl,−l . (57)
We connect m = l to m = −l by an integer number k of operations. Then
l − (−l) = 2l = k (58)
or
k
l= (59)
2
an integer or a half integer.
2
We summarize the results of these lemmas with the following set of equations and statements:
L̂z Yl,m = mh̄Yl,m (60)

L̂2 Yl,m = l(l + 1)h̄2 Yl,m (61)


−l ≤ m ≤ l (62)
where m and l can either be integers or half integers. Notice for a given l, there are gl = 2l+1
different m-states.
It is interesting that the algebraic commutation relations for the angular momentum
operators allows both integer and half integer quantum numbers for the angular momentum.
When we solve for the eigenfunctions in a subsequent section of these notes, we find that
the half integer quantum numbers are excluded. However, the algebra admits half integer
quantum numbers that, as we can expect, correspond to the spin quantum number. The
spin of particles is derivable from a relativistic treatment of quantum theory and cannot be
understood from the Schrödinger equation alone.
We close this section with two final lemmas that give expressions for c+ (l, m) and c− (l, m).
Lemma 8 q
c+ (l, m) = l(l + 1) − m(m + 1)h̄. (63)
Proof:
hYl,m |L̂− L̂+ |Yl,m i = hL̂+ Yl,m |L̂+ Yl,m i = c2+ (l, m) (64)
= [l(l + 1) − m(m + 1)]h̄2 (65)
or q
c+ (l, m) = l(l + 1) − m(m + 1)h̄. (66)
2

6
Lemma 9 q
c− (l, m) = l(l + 1) − m(m − 1)h̄. (67)

Proof:
hYl,m |L̂+ L̂− |Yl,m i = hL̂− Yl,m |L̂− Yl,m i = c2− (l, m) (68)
= [l(l + 1) − m(m − 1)]h̄2 (69)
or q
c− (l, m) = l(l + 1) − m(m − 1)h̄. (70)
2

3 The Angular Momentum Eigenfunctions


In the algebraic solution for the harmonic oscillator Hamiltonian eigenfunctions, the ground
state eigenfunction is determined by first applying the annihilation operator to the ground
state function and solving the resulting differential equation. The other eigenfunctions are
determined by applying products of the creation operator to the ground state wavefunction.
To find the eigenfunctions of the angular momentum operators, we can use a similar method
starting with the step-up and step-down operators. In what follows, we sketch the treatment
for the angular momentum eigenfunctions, but do not give full details. The detailed treat-
ment given in the text using differential equation methods for determining the eigenfunctions
is in some ways preferable.
We begin with the equation
L̂+ Yl,l (θ, φ) = 0 (71)
where we have explicitly expressed the eigenfunction as a function of the angular variables
in spherical polar coordinates, the most convenient coordinate system for this problem. To
continue, we need an expression for L̂+ in spherical polar coordinates. Using Eqs.(5), (6)
and (13), we obtain !
∂ ∂
L̂+ = h̄eiφ + i cot θ (72)
∂θ ∂φ
so that !
∂ ∂
+ i cot θ Yl,l (θ, φ) = 0. (73)
∂θ ∂φ
We can solve Eq.(73) using separation of variables. We write

Yl,l (θ, φ) = Θl,l (θ)Φl (φ) (74)

and substitute the separated form into Eq.(73). After rearrangement, we find

1 dΘl,l (θ) 1 dΦ
= =s (75)
cot θΘl,l (θ) dθ iΦl (φ) dφ

7
where s is the separation constant. We first solve the differential equation for the variable φ
dΦl
= is dφ (76)
Φl
or
Φl (φ) = Aeisφ . (77)
From Eq.(7) we know that the operator corresponding to the z-component of the angular
momentum depends only on φ, and Eq.(77) must be an eigenfunction of L̂z . Using Eq.(7)

h̄ ∂
L̂z Φl (φ) = lh̄Φl (φ) = Φl (φ) = sh̄Φl (φ) (78)
i ∂φ
so that the separation constant s = l. We next normalize Φl
Z 2π Z 2π
A2 Φ∗l (φ)Φl (φ) dφ = A2 e−ilφ eilφ dφ (79)
0 0

= A2 2π = 1 (80)
or
1
A= √ . (81)

The value of l is determined using the appropriate boundary condition. Because φ is a cyclic
coordinate, we must have
Φl (φ) = Φl (φ + 2π) (82)
or
1 1
√ eilφ = √ eil(φ+2π) (83)
2π 2π
so that
eil2π = 1 (84)
so that l must be an integer. For orbital angular momentum, the values of l are restricted
to l = 0, 1, 2, . . .. As indicated previously, the algebraic solution to the angular momentum
eigenvalue problem admits half integer eigenvalues, but the boundary conditions on the
differential equation exclude the half integer solutions.
We next solve the θ-dependent part of Eq.(75). After rearrangement we have

dΘl,l
= l cot θ dθ (85)
Θl,l

cos θ
=l dθ (86)
sin θ
d sin θ
=l . (87)
sin θ

8
Integrating both sides of Eq.(87)
Θl,l (θ) = B sinl θ (88)
where B is a normalization constant. We normalize using the equation (the extra factor of
sin θ comes from the Jacobian)
Z π
Θ2l,l (θ) sin θ dθ = 1. (89)
0

After some algebra, we find


" #1/2
(2l + 1)!
Θl,l (θ) = sinl θ. (90)
2(2l l!)2

Combining with the solution to the φ-dependent part of Eq.(75) we finally obtain
" #1/2
(2l + 1)! 1
Yl,l (θ, φ) = sinl θeilφ . (91)
4π 2l l!

To obtain expressions for Yl,m (θ, φ) for m 6= l, we apply the step-down operator to Eq.(91).
The result is identical to the result of the differential equation approach to finding the
eigenfunctions that is discussed in your textbooks.
The complete, orthonormal set of functions {Yl,m (θ, φ)} are called spherical harmonics,
and it can be shown that the general expression for the spherical harmonics is given by
#1/2
dl+|m|
"
(2l + 1)(l − |m|)!
Yl,m (θ, φ) = (1 − y 2 )|m|/2 (1 − y 2 )l eimφ (92)
4π(l + |m|)! dy l+|m|

where
sin θ = (1 − y 2 )1/2 . (93)

S-ar putea să vă placă și