PHYS 130A: Angular Momentum

Substitute Lecturer: Emily Nardoni

1 Review of the Hydrogen atom

The hydrogen atom describes an electron orbiting a nucleus in a central potential,

ke2 1
V (r) = − , k= . (1)
r 4π0
We write the wavefunctions for this system in spherical coordinates. Then, the stationary
state wavefunctions take the form
Ψ(~r, t) = ψn`m (~r)e−iωt , ψn`m (~r) = Rn` (r)P`m (θ)Φm (φ). (2)
We put subscripts on the wavefunctions to remind ourselves of which quantum numbers
they depend on. The θ and φ dependence is captured by the functions P`m` (θ) (the
Legendre polynomials) and Φ(φ) = eimφ , which together define the spherical harmonics,
Y`m ,
Y`m (θ, φ) = P`m (θ)Φm (φ). (3)
The spherical harmonics are labeled by the quantum numbers `—the orbital quantum
number–and m—the magnetic quantum number.

The possible energy levels for an electron in the hydrogen atom are
~2 4π0 ~2
En = − a0 = . (4)
2ma20 n2 me2
where n is called the principal quantum number and runs from 1, 2, . . . . The possible
states corresponding to a given energy level labeled by n are degenerate, with
n: ` = 0, 1, 2, . . . , (n − 1), m = 0, ±1, ±2, . . . , ±`.
The radial part of the wavefunction u(r) = rR(r) satisfies a radial version of the time
independent Schrödinger equation,
~2 d2 u(r)
  2
~ `(` + 1) ke2
 
− + − u(r) = Eu(r). (5)
2m dr2 2mr2 r
or written in terms of R(r),
~2 d2 R(r) 2 dR(r)
  2
~ `(` + 1) ke2
 
− + + − R(r) = ER(r). (6)
2m dr2 r dr 2mr2 r
 2 
~ `(`+1) ke2
The term 2mr2
− r is known as the effective potential, V̂eff .

1
2 Angular momentum in quantum mechanics

The main goal for this section is to explain: Why (physically) are the hydrogen wave-
functions labeled by n, `, m?

To do this, we’ll first need to define quantum angular momentum operators. Recall that
the classical definition of angular momentum is L ~ = ~r × p~. To write a quantum operator,
we’re instructed to just replace all the classical variables with operators, and plug in
for those operators. Thus to write the Cartesian coordinates of the angular momentum
operator, just put hats on everything, and plug in for the ~xˆ and p~ˆ operators in the formulae
above. Quantum mechanically, recall that the 3d components of the momentum operator
in Cartesian coordinates are

p̂i = −i~∂i , (7)

or written out in components,

∂ ∂ ∂
p̂x = −i~ , p̂y = −i~ , p̂z = −i~ . (8)
∂x ∂y ∂z
Then, the angular momentum operator is: in matrix form,
 
î ĵ k̂
~ = ~ det  x̂ ŷ ẑ  ,
L (9)
i
∂x ∂y ∂z

or explicitly in components,
 
∂ ∂
L̂x = (ŷ p̂z − ẑ p̂y ) = −i~ ŷ − ẑ ,
∂z ∂y
 
∂ ∂
L̂y = (ẑ p̂x − x̂p̂z ) = −i~ ẑ − x̂ , (10)
∂x ∂z
 
∂ ∂
L̂z = (x̂p̂y − ŷ p̂x ) = −i~ x̂ − ŷ .
∂y ∂x

All of these expressions can be written in spherical coordinates by using the relations
between Cartesian (x, y, z) and spherical (r, θ, φ). This is rather tedious, so we’ll skip
the details here, and give you the answer for the components of the angular momentum
operator as written in spherical coordinates:
 
∂ ∂
L̂x = i~ sin φ + cot θ cos φ
∂θ ∂φ
 
∂ ∂
L̂y = −i~ cos φ − cot θ sin φ
∂θ ∂φ
(11)
∂
L̂z = −i~
∂φ
1 ∂2
   
2 2 2 2 2 1 ∂ ∂
⇒ L̂ = L̂x + L̂y + L̂z = −~ sin θ + .
sin θ ∂θ ∂θ sin2 θ ∂φ2

2
We’ve also written the total angular momentum squared operator L̂2 , which will come in
handy in a bit.

Back to the main question: The operators Ĥ, L̂2 , L̂z are compatible observables that can
have simultaneous eigenvalues. Those eigenvalues are related to the n, `, m numbers.
The hydrogen wavefunctions are labeled by n, `, m because the hydrogen stationary state
wavefunctions have definite energy E and definite total angular momentum squared L2
(equivalently, magnitude of angular momentum) and definite z-component of angular
momentum Lz .

3 Reinterpreting the radial wave equation for Hy-

drogen

As further motivation, let’s start by noticing that L̂2 was hidden in the previous discussion
of the Hydrogen atom.

Let’s write out the total energy operator for the Hydrogen atom. The potential energy is
the Coulomb potential energy, V (r) = −ke2 /r. The kinetic energy operator in Cartesian
coordinates is
p̂2 ~2
 2
∂2 ∂2 ~2 2

∂
K̂ = =− + + = − ∇. (12)
2m 2m ∂x2 ∂y 2 ∂z 2 2m
Since we’ve been writing the wavefunctions in spherical coordinates, so we should write
the kinetic energy in spherical coordinates as well. This means writing the Laplacian ∇2
in spherical coordinates. Doing so (see Griffiths) yields
~2
 2
1 ∂2
   
∂ 2 ∂ 1 1 ∂ ∂
K̂ = − + + sin θ + (13)
2m ∂r2 r ∂r r2 sin θ ∂θ ∂θ sin2 θ ∂φ2
Comparing with the expressions for angular momentum we wrote in the last section, we
recognize the term in square brackets as precisely the total angular momentum squared
operator. In other words, we can rewrite
~2
 2 
∂ 2 ∂ 1
K̂ = − + + L̂2 (14)
2m ∂r2 r ∂r 2mr2
In spherical coordinates, we’ve shown that there are two components to the kinetic energy
operator: a piece which we’ll call the radial piece (the first term in brackets above), and
a piece which we’ll call the orbital piece,
~2
 2 
∂ 2 ∂ 1
K̂ = K̂radial + K̂orbital , K̂radial ≡ − 2
+ , K̂orbital ≡ L̂2 . (15)
2m ∂r r ∂r 2mr2
Comparing with the way we’d previously solved the Schrodinger equation by separation
of variables, we notice that we had actually been writing
h i
Ĥψ = K̂radial + K̂orbital + V (r) ψ = Eψ, (16)

3
where this goes through if

L̂2 ψ = ~2 `(` + 1)ψ (17)

(and with a similar equation for Lz , which we’ll get to later in this section.)

How do we interpret (17)? It’s the statement that the wavefunction is an eigenfunction of
the total angular momentum squared operator L̂2 , with eigenvalue ~2 `(`+1). Put another
way, in a state with some value of the orbital quantum number `, the magnitude of the
angular momentum |L| ~ in that state is completely specified. Since ` is a positive integer,
` = 0, 1, . . . , (n − 1), this means that the electron’s angular momentum is quantized!

With this idea in mind, we’ll move on to an algebraic approach to obtain the eigenvalues
of these operators and demonstrate why L̂2 , L̂z , Ĥ are good operators to label the states
of the Hydrogen atom.

4 An algebraic approach

Our goal is to find the eigenvalues of the angular momentum operators, which we will do
by computing commutators. We will need to use the canonical commutation relations

[ri , pj ] = i~δij , [ri , rj ] = [pi , pj ] = 0. (18)

The indices run over i = 1, 2, 3, or equivalently i = x, y, z. For instance, r1 ≡ x, p1 ≡ px ,

and so on. δij is zero when the two indices are different, and 1 when the indices are the
same. Also, I’ll stop writing hats on all the operators to save myself on typing.

4.1 ~ does not have a definite direction

L

Let’s compute the following commutator:

[Lx , Ly ] = [ypz − zpy , zpx − xpz ]

= [ypz , zpx ] − [ypz , xpz ] − [zpy , zpx ] + [zpy , xpz ].

The middle two commutators are zero. We are left with

[Lx , Ly ] = ypx [pz , z] + xpy [z, pz ]

= i~(xpy − ypx )
= i~Lz .

The other commutators can be worked out by a cyclic permutation of the indices. The
result can be written compactly as:

[Li , Lj ] = i~ijk Lk . (19)

4
The repeated index k is summed over (this is known as Einstein summation notation).
The notation is that ijk is the completely antisymmetric tensor in 3 dimensions,

123 = 312 = 231 = 1, 132 = 321 = 213 = −1. (20)

These satisfy that switching a pair of indices gets a minus sign ( ijk = −jik ), while
switching two pairs gives a plus sign (ijk = −jik = +jki ). ijk is zero whenever two
indices are equal.

For example, take i = 2(= y), j = 3(= z). This is

X
[L2 , L3 ] = i~ 23k Lk
k
= i~ (231 L1 + 232 L2 + 233 L3 )
= i~ (1 · L1 + 0 + 0)
= i~L1
⇒ [Ly , Lz ] = i~Lx , (21)

and so on. A good exercise to make sure you understand this index notation is to show
that the cross product in three dimensions can equivalently be written

(~a × ~b)k = ijk ai bj . (22)

The point is that we can’t measure two components of L ~ precisely at the same time.
If we measure Lz , we must have an intrinsic uncertainty in Lx , Ly . What this means
is that while classically the direction and magnitude of angular momentum are always
well-defined, but quantum mechanically it does not make sense to talk about L~ as having
a definite direction.

To demonstrate this, we compute

2
~2

1
σL2 x σL2 y ≥ hi~Lz i = hLz i2 (23)
2i 4
~
⇒ σLx σLy ≥ |hLz i|. (24)
2
Another perspective (We didn’t have this discussion in class, but maybe this is another
argument to help your intuition): We can prove by negation that L ~ does not have a
definite direction. Suppose L ~ was precisely known—i.e., we knew all the components
~ = ~r × p~, this implies that the
Lz , Ly , Lz , and let’s say Ly and Lz were zero. Since L
electron must be in the x-y plane. In other words, we would know with certainty that
the electron was at z = 0, so ∆z = 0. But if ∆z = 0, by the uncertainty principle in the
z-direction,
~
∆pz ∆z ≥ ⇒ ∆pz = ∞.
2
So we’d have to have no idea what the electron’s z-component of momentum was. But,
the energy of the electron includes the kinetic energy, p̂2 /2m. So, this would imply that
electron’s energy also was uncertain. And therein lies the contradiction: we already know

5
that an electron in a stationary state of Hydrogen by construction has a definite value
of energy, E. Thus, our assumption that we could pinpoint the direction of the angular
momentum of the electron must be incorrect.

Again, the punchline is that we can’t simultaneously know the x, y, and z components
of the angular momentum of the electron in the Hydrogen atom. Unless all components
are exactly zero, there is no wavefunction for which two or more of these components can
be measured precisely at the same time!

4.2 L2 and Lz commute, and thus label simultaneous eigenstates

Next up: work out commutators with L2 . Let’s check

[L2 , Lx ] = [L2x + L2y + L2z , Lx ].

Of course, Lx commutes with itself. The other two:

[L2y , Lx ] = Ly [Ly , Lx ] + [Ly , Lx ]Ly = −i~(Ly Lz + Lz Ly )

[L2z , Lx ] = Lz [Lz , Lx ] + [Lz , Lx ]Lz = i~(Lz Ly + Ly Lz ).

Together, these add to zero! We get

[L2 , Lx ] = 0. (25)

Similarly, we’ll get [L2 , Ly ] = [L2 , Lz ] = 0, and we can write

[L2 , Li ] = 0. (26)

This means we can simultaneously measure L2 , and ONE of Lx or Ly or Lz . The conven-

tion we choose: L2 and Lz can share simultaneous eigenfunctions. Next class, you will
determine the spectrum of eigenvalues of these operators.