Problem Set 8

8.04 Spring 2013

Solutions
April 17, 2013

Problem 1. (15 points) Superposition State of a Free Particle in 3D

(a) (4 points) Recall from lecture that the energy eigenstates of a free particle in 3D are
given by
= A exp i kk kr = A exp [i (kx x + ky y + kz z)] ,
(1)
where A is some normalization constant. The energy is given by
E=

n2 (kx2 + ky2 + kz2 )

n2 |kk|2
=
.
2m
2m

(2)

In our case, we have

3

2
(kr, 0) =
sin(3x/L)e i(5y+z)/L
3/2
2L
3
2
=
ei3x/L ei3x/L e i(5y+z)/L
3/2
4L
3
2 i(3x+5y+z)/L
=
e
+ ei(3x+5y+z)/L ,
3/2
4L

(3a)
(3b)
(3c)

which we can immediately see is the superposition of two energy eigenstates, one with
kk = (3, 5, 1)/L and the other with kk = (3, 5, 1)/L. In both cases the energy is
E=

n2 |kk|2
35n2
=
,
2m
2mL2

(4)

so this will be our result (with complete certainty) if we measure the energy at t = 0.
(b) (4 points) Since the momentum eigenstates are the same as the energy eigenstates
for a free particle, Equation 3c can also be viewed as a superposition of momentum
eigenstates. Using pk = nkk, we can therefore say that the possible outcomes of a
measurement of momentum are
pk = (3n, 5n, n)/L and pk = (3n, 5n, n)/L.

(5)

Note that in this case the probability of nding the particle in one of these momentum
eigenstates is not simply |cn |2 = 1613 L3 . This is because the wavefunction given in
Equation 3c is not properly normalized (which also implies that the |cn |2 s dont sum to
1), and indeed cannot be normalized. However, since we know that only two outcomes
are possible, and that they are equally likely (since their momentum eigenfunctions
are multiplied by the same coecient), we can say that the probability of measuring
either of the momentum in Equation 5 is 12 .
1

(c) (3 points) We know from before that an energy eigenstate evolves by a multiplicative
phase factor eiEt/n . Since both energy eigenstates in our superposition state (Equation
3c) have the same energy, the time-dependent wavefunction is given by
3

2
2
(kr, t) =
3/2 e
i(3x+5y+z)/L + e
i(3x+5y+z)/L e
iEt/n =
sin(3x/L)e
i(5y+z/L) e
iEt/n ,

2L3/2
4L
(6)
2
2
where E = 35n /(2mL ), as given in Equation 4.
(d) (4 points) Measuring pk = (3n, 5n, n)/L immediately collapses the wavefunction into
the corresponding momentum eigenstate.
(kr, 0) = A exp [i (3x + 5y + z) /L] .

(7)

Since this is also an energy eigenstate, the subsequent time evolution is once again
given by
(kr, t) = A exp [i (3x + 5y + z) /L] eiEt/n ,
(8)
where again E = 35n2 /(2mL2 ). We have left the normalization constant unspecied
because the energy eigenstates of a free particle cannot be normalized.

Problem Set 8
8.04 Spring 2013

Solutions
April 17, 2013
Problem 2. (15 points) Degeneracies

(a)

i. (1 point) The quantum number which determines the energy eigenvalue E is the
wavenumber k. We see that
h in one dimension to a given energy E corresponds
two wavenumbers k =
each energy eigenvalue.

2mE
,
n2

thus there are two linear independent sate for

ii. (1 point) The Hamiltonian of the free particle in one dimension is invariant
under the parity operation, i.e. x x. While the energy remains the same, the
momentum operator transforms like the space coordinate: p p.
(b)

i. (2 points) The allowed energy eigenvalues are simply the sum of the energy eigen
values of the two independent harmonic oscillators (since there is no interaction
between them):

1
1
Enx ,ny = nx nx +
+ ny ny +
= n (nx + ny + 1) .
(9)
2
2
ii. (2 points) Lets make the notation n = nx + ny so that the energy eigenvalue is
En = n (n + 1). For a given eigenenergy En or principal quantum number n, nx
and ny can take only the following values:

nx = 0, 1, . . . , n 1, n
.
(10)
ny = n, n 1, . . . , 0, 1
So we see that there are n + 1 ways to distribute the value of n among nx and
ny . Thus the ground state n = 0 is non degenerate, rst excited state n = 1 is
double degenerate, second excited state n = 2 is triple degenerated, and the third
excited sate n = 3 is four times degenerated.
iii. (1 point) Since the two independent oscillators have the same frequency, a rota
tion of the system around a axis perpendicular to the plane of the oscillators leaves
the system unchanged, or in other words Lz component of angular momentum is
conserved.

iv. (2 points) See the graph below (D stands for degeneracy):

(c)

i. (1 point) Since the two harmonic oscillators are still independent, the energy
eigenvalue is:





1
1
1
Enx ,ny (f) = nx nx +
+ ny ny +
= n (nx + ny + 1) + fn nx +
.
2
2
2
(11)
ii. (1 point) The system is no longer invariant under rotations, so the degeneracy
is lifted.
iii. (3 points) Plotted below are the rst ten eigenenergies in units of n as function
of f (and the quantum numbers nx and ny ):
n = nx + ny = 0
E ; n=nx +n y =0
nx =ny =0
1.04

1.02

1.00

0.98

-0.10

-0.05

0.05

0.10

0.15

 n = nx + ny = 1
E ; n=nx +n y =1
2.15

nx =1 ny =0

2.10
nx =0 ny =1

2.05
2.00
1.95
1.90

-0.10

-0.05

0.05

0.10

0.15

n = nx + ny = 2
E ; n=nx +n y =2
nx =2 ny =0
3.2
nx =1 ny =1
3.1
nx =0 ny =1
3.0

2.9

-0.10

-0.05

0.05

0.10

0.15

 n = nx + ny = 3

E ; n=nx +n y =3
nx =3 ny =0
4.3
nx =2 ny =1
4.2
nx =1 ny =2
4.1
nx =0 ny =3
4.0
3.9
3.8

-0.10

-0.05

0.05

0.10

0.15

[Not required for full credit] In order to observe the crossings between various
energy eigenvalues, below all ten eigenenergies are plotted on the same graph
for a bigger range of f. All the red eigenenergies correspond to nx = 0, all the
green eigenenergies correspond to nx = 1, all the black eigenenergies correspond
to nx = 2, and the blue eigenenergy corresponds to nx = 3.
E

nx =3 ny =0

nx =2 ny =1

nx =1 ny =2
nx =2 ny =0
nx =0 ny =3
nx =1 ny =1

nx =0 ny =1
nx =1 ny =0
3

nx =0 ny =1
2

nx =ny =0
1

-0.5

0.5

1.0

iv. (1 point) As pointed earlier a degeneracy in energy eigenstates indicates that

the system remains invariant under some symmetry transformation, or equiva
lently there is some conserved quantity. To each quantum number describing
6

the conserved observable corresponds one energy eigenfunction, but since the en
ergy eigenvalue does not depend on the conserved quantum numbers the energy
eigenfunctions are degenerated.

Problem Set 8
8.04 Spring 2013

Solutions
April 17, 2013

Problem 3. (20 points) Mathematical Preliminaries: Angular Momentum Operators

z ]:
y, L
(a) (8 points) First we deal with [L
y, L
z ] = [
[L
z px xpz , xpy ypx ]
(12a)
= (
z px xpz )(
xpy ypx ) (
xpy ypx )(
z px xpz )
(12b)
= zpx xpy xpz xpy zpx ypx + xpz ypx xpy zpx + xpy xpz + ypx zpx ypx xpz
(12c)
= (xpz xpy + xpy xpz ) + (
y px zpx zpx ypx ) + (
z px xpy xpy zpx ) + (
xpz ypx ypx xpz ).(12d)
The terms in the rst two sets of parentheses are zero, because all the operators within

a term commute with each other. For example,

xpz xpy + xpy xpz = x2 pz py + x2 py pz = x2 pz py + x2 pz py = 0,

(13)

because [
xa , xb ] = 0, [
pa , pb ] = 0, and[
xa , pb ] = inab . The remaining terms can be

rewritten as follows:

zpx xpy xpy zpx = zpx xpy xzpy px = zpx xpy zxpx py = z (
xp p x) p = inzpy ,
" x -v x " y
=[
x,px ]

(14)

where we have made liberal use of the fact that the position and momentum commute

unless they refer to the same coordinate (i.e. the same x, y, or z). Similarly,

xpz ypx ypx xpz = (

xpx px x)
pz y = inpz y.

(15)

Putting everything back together again, we get

y, L
z ] = in(
x.
[L
y pz zpy ) = inL

(16)

z, L
x ] and [L
x, L
y ] would be to do the algebra explicitly, much

One way to compute [L

y, L
z ]. An easier way, however, would be to make the following

like we did just now for [L

observation. If we look at the denition of the angular momentum operators,

x = ypz zpy ,
L

y = zpx xpz ,
L

z = xpy ypx
L

(17)

x L
y ,

we see that making the replacements x y, y z, and z x takes L

y L
z , and L
z L
x . There is nothing mysterious about this, since x, y, and z are

L
merely labels for the three coordinate axes, and we chose the replacements carefully so

that the right-handedness of the coordinate system was preserved. Performing these

replacements on Equation 16 yields

z, L
x ] = inL
y
[L

x, L
y ] = inL
z.
and [L
8

(18)

2 = L
2 + L
2 + L
2.
(b) (7 points) We now consider the total angular momentum operator L
x
y
z

The commutator of this operator with Lz is

2 ] = [L
z, L
2 + L
2 + L
2 ] = [L
z, L
2 ] + [L
z, L
2 ] + [L
z, L
2 ] = [L
z, L
2 ] + [L
z, L
2 ]. (19)
z, L
[L
x
y
z
x
y
z
x
y
The rst term can be written as
2 ] =
z, L
[L
x
=
=

zL
2 L
2 L

2

2
L
x
x z = Lz Lx Lx Lz Lx + Lx Lz Lx Lx Lz
zL
x L
xL
z )L
x + L
x (L
zL
x L
xL
z)
(L
z, L
x ]L
x + L
x [L
z, L
x ] = in(L
yL
x + L
xL
y ).
[L

(20a)
(20b)
(20c)

2] = L
zL
2 L
2L

2

2
z, L
[L
y
y
y z = Lz Ly Ly Lz Ly + Ly Lz Ly Ly Lz
zL
y L
yL
z )L
y + L
y (L
zL
y L
yL
z)
= (L

(21a)

Similarly, we have

z, L
y ]L
y + L
y [L
z, L
y ] = in(L
yL
x + L
yL
x ).
= [L

(21b)
(21c)

The two terms cancel, so we get

z, L
2 ] = 0.
[L

(22)

As we argued above, x, y, and z are simply labels for the three coordinate axes. Since
2 is unaected if we switch them, we can say without
their labeling is arbitrary, and L
further calculation that
x, L
2 ] = 0 and [L
y, L
2 ] = 0.
[L

(23)

= L
x iL
y and L
+ = L
x + iL
y . Then
(c) (5 points) Let L
2, L
] = [L
2, L
x iL
y ] = [L
2, L
x ] i[L
2, L
y ] = 0,
[L

(24)

and
] = [L
z, L
x iL
y ] = [L
z, L
x ] i[L
z, L
y ] = inL
y nL
x = nL
.
z, L
[L

(25)

z and L
2 such that
Consider an eigenstate ,m of both L
z ,m = nm ,m ,
L

2 ,m = n2 Q ,m .
L

(26)

The commutation relation (25) tells us that

,m = L
L
z ,m + [L
z, L
] ,m = n(m 1)L
,m ,
zL
L

(27)

,m is an eigenstate of L
z with eigenvalue n(m1). Likewise, the commutation
thus L
relation (24) tells us that
2L
,m = L
L
2 ,m + [L
2, L
] ,m = n2 Q L
,m ,
L

(28)

,m is also an eigenstate of L
2 and it has the same
from which we infer that L
eigenvalue of ,m .

Problem Set 8
8.04 Spring 2013

Solutions
April 17, 2013

2 and L
z
Problem 4. (15 points) Mathematical Preliminaries: Eigenfunctions of L

(a) (5 points) For the angular parts of a wavefunction, the norm of a (, ) is given by
2

(|) =

sin d||2 .

d
0

(29)

Here we have
1
4

3
cos
4

3
sin ei
8

Y0,0 =
Y1,0 =
Y1,1 =

1
4
3
|Y1,0 |2 =
cos2
4
3
|Y1,1 |2 =
sin2
8
|Y0,0 |2 =

(30a)
(30b)
(30c)

So:
2

(Y0,0 |Y0,0 ) =
(Y1,0 |Y1,0 ) =
(Y1,1 |Y1,1 ) =

1
4
0
0
2

3
d
sin d|Y1,0 |2 =
4
0
0
2

3
d
sin d|Y1,1 |2 =
8
0
0
d

sin d|Y0,0 |2 =

d
0

sin d = 1

(31a)

sin d cos2 = 1

(31b)

0
2

d
0

0
2

sin d sin2 = 1, (31c)

d
0

where the last two integrals can be evaluated by making the substitution u cos .
We now check for orthogonality:

3 2
3
d
sin d cos =
sin 2d = 0
(32a)
(Y0,0 |Y1,0 ) =
4 0
4 0
0
(Y0,0 |Y1,1 ) =
=

3
32 2

sin d sin ei
0
0
2

3
2
i
e d
sin d = 0
32 2
0
0
"
-v
"
d

(32b)
(32c)

=0

(Y1,0 |Y1,1 ) =
=

32 2 0

32 2 "

sin d cos sin ei

0

2
i
2
sin cos d = 0,
e d
0
0
-v
"
d

=0

10

(32d)
(32e)

Note that we these three integrals we have exhausted all possible combinations. For
example, it is unnecessary for us to check (Y1,0 |Y0,0 ) because
Z 2 Z



Z 2 Z

(Y1,0 |Y0,0 ) =
d
sin dY1,0 Y0,0 =
d
sin dY0,0 Y1,0 = (Y0,0 |Y1,0 ) = 0,
0

(33)
so if (A |B ) = 0, then (B |A ) is automatically also zero.
2 and L
z , then they must
(b) (4 points) If the spherical harmonics are eigenfunctions of L
satisfy the following eigenvalue equations:



2
1

2
2
= L2
L

n
sin
+
= L2 (34a)
sin

sin2 2

z = Lz

in
= Lz .
(34b)
L

Plugging the spherical harmonics into these equations, we see that they are indeed
2 and L
z , with eigenvalues
eigenfunctions of L
Y0,0 : L2 = 0 and Lz = 0.
Y1,0 : L2 = 2n2 and Lz = 0.
Y1,1 : L2 = 2n2 and Lz = n.
(c) (6 points) From class we know that
= nei ( cot ).
L

(35)

We shall now follow the strategy learnt in class to obtain Yl,l and work out Y42,42 .
First, we require
Y42,42 = P42,42 ()ei42 .
(36)
Imposing
Y42,42 = nei ( 42 cot )P42,42 ()ei42 = 0,
L

(37)

Y42,42 = cl,l (sin )42 ei42 ,

(38)

we nd
and thus,
+ Y42,42 ei ( + cot )(sin )42 ei42 sin41 cos ei41 ,
Y42,41 = L

(39)

and the dependence on and is completely determined. To nd the normalization

factor c42,41 , we impose
Z 2 Z
d
d sin |c42,41 sin41 cos ei41 |2 = 1.
(40)
0

Using Mathematica to work out the integral, we nd

Z 2 Z
d
d sin | sin41 cos ei41 |2 =
0

11

240

,
374606902236028199511433275

(41)

374606902236028199511433275
41
sin cos ei41 .
241

(42)

d sin83 cos2 = 2

= 2
0

and thus

Y42,41 =

12

Problem Set 8
8.04 Spring 2013

Solutions
April 17, 2013

Problem 5. (20 points) Angular Momenta and Uncertainty

2 and L
z , we know that
(a) (5 points) Since Ylm is a normalized eigenstate of L
z = nm.
2 = n2 l(l + 1) and L
L

(43)

+ and L
, which are dened
To nd (Lx ) and (Ly ), we can make use of the operators L
as
L+ = Lx + iLy and L = Lx iLy ,
(44)
so
Lx =

L + L
L+ + L
and Ly =
.
2
2i

(45)

We therefore have
Z
Z
Z
1

(Lx ) =
(L+ + L )d Ylm L+ Ylm d + Ylm
L Ylm d
2
Z
Z

Yl,m1 d = 0,
Ylm Yl,m+1 d + Ylm

(46a)
(46b)

where in the last equality we have used the fact that the spherical harmonics are
orthonormal, that is
(Ylm |Yl m ) = ll mm ,
(47)
so only if the ls and ms both match do we get a non-zero answer. Alternatively, we
could have used notation that is somewhat more compact:
1
(Lx ) = [(|L+ )+(|L+ )] (Ylm |L+ Ylm )+(Ylm |L+ Ylm ) (Ylm |Yl,m+1 )+(Ylm |Yl,m1 ) = 0.
2
(48)
The steps in Equation 48 match those in Equations 46a and 46b exactly, and it is a
good exercise to make sure you can translate between the dierent notations. For (Ly ),
we similarly have
(Ly ) =

1
[(|L+ )(|L+ )] (Ylm |L+ Ylm )(Ylm |L+ Ylm ) (Ylm |Yl,m+1 )(Ylm |Yl,m1 ) = 0.
2i
(49)

2 ) and (L
2 ), we note that nothing we have done has broken the
(b) (6 points) To nd (L
x
y
z , but
symmetry between the x and y axes. Our wavefunction is an eigenstate of L
nothing we have done so far has made a distinction between x and y. We can thus
2 ) = (L
2 ), and in this case only (it is not true for a general
immediately say that (L
x
y
state):
2 ) + (L
2 ) + (L
2)
2 ) = (L
(L
x
y
z

2 ) = 2(L
2 ) + (L
2)
(L
x
z

13

2
2
2 ) = (L ) (Lz ) .
(L
x
2
(50)

Thus,
2
2 2
2
2
2
2 ) = (|L
2 ) = (|L ) (|Lz ) = l(l + 1)n (|) m n (|) = n [l(l+1)m2 ],
(L
x
x
2
2
2
(51)
where the last equality followed from the fact that is normalized. In conclusion,
then,
2
2 ) = (L
2 ) = n [l(l + 1) m2 ].
(L
(52)
x
y
2
If youre uncomfortable with the symmetry argument, heres a dierent way to ap
2 can be written as
proach the problem. The operator L
x

2 = 1 (L
)(L
+ + L
) = 1 (L
L
+ + L
+L
+ L
2 ).
+ + L
2 + L
L
x

2
4 +

(53)

The rst and last terms are of no consequence because when we sandwich them
between s to nd the expectation value we get zero from orthonomality:
2 ) (Ylm |L
2 Ylm ) (Ylm |Yl,m2 ) = 0.
(|L

(54)

This means

1
+L
L
+ ) + (|L
)
(|L
4

We can deal with whats left by doing a little commutator algebra:

2 ) =
(|L
x

L
L

x iL
y )(L
x iL
y) = L
2 + L
2 iL
yL
x
= (L
x
y
2 L
2 i[L
x, L
y] = L
2 L
2 nL
z,
= L
z

xL
y
iL

(55)

(56a)
(56b)

x, L
y ] = inL
z . Thus,

where in the last equality we used the fact that [L

L
2 ) (|L
2 ) n(|L
z ) = l(l + 1)n2 m2 n2 mn2
,
) = (|L
(|L
z

(57)

2 and L
z . Putting everything together, we get

because is an eigenstate of both L

2

2 ) = n [l(l + 1) m2 ],
(L
x
2

(58)

2 ) and verify that (L

2 ) =
just like before. A similar set of manipulations will give (L
y
x
2 ).
(L
y
x ) = 0 and (L
2 ) = (L
2 ), we have
(c) (4 points) Since (L
x
y
Lx = Ly =

h
h
2 ) (L
y )2 = (L
2 ) = n l(l + 1) m2 .
(L
y
y
2

(59)

The left hand side of the proposed uncertainty relation thus reads

Lx Ly =

n2
[l(l + 1) m2 ].
2

14

(60)

For a xed l, this quantity is minimized when m is as large as possible i.e. when m = l.
In such a situation, we have Lx Ly = ln2 /2. As for the right hand side, we have
n
mn2
(Lz ) =
2
2

(61)

This side is maximized precisely when the other side is minimzed (when m = l), and
we get n2 (Lz ) = ln2 /2. In this case the two sides are equal, but for all other cases
(where m < l) the left hand side is greater. We can therefore conclude
Lx Ly

n
(Lz ).
2

(62)

It is possible for all components of angular momentum to vanish simultaneously. If a

particle is in an eigenstate with l = m = 0, then from the relationships we have proved
2 ) = (L
2 ) = (L
2 ) = 0 as well as (L
x ) = (L
y ) = (L
z ) = 0, so
in this question we have (L
x
y
z
we can say Lx = Ly = Lz = 0.

15

Problem Set 8
8.04 Spring 2013

Solutions
April 17, 2013

Problem 6. (20 points) Lifting the Degeneracy of the Quantum Rigid Rotor

(a) (4 points) From the lecture notes we know that the spherical harmonics Ylm are the
k 2:
eigenfunctions of the angular momentum operator squared L
k 2 Ylm = n2 l(l + 1)Ylm .
L

(63)

Thus the energy eigenvalues of the quantum rigid rotator are:

2
2
k2
lm = L Ylm = n l(l + 1) Ylm Elm = n l(l + 1) .
EY
2I
2I
2I

(64)

(b) (4 points) Although we labeled the energy eigenvalues (64) by both l (controlling the
total angular momentum) and m (controling the z-component of the angular momen
tum) quantum numbers, in the case of the rigid rotator and indeed for all spherically
symmetric systems the energy eigenvalues do not depend on m. From the lecture
notes we know that for a given l the quantum number m runs from +l to l, so every
energy eigenvalue is 2l + 1 degenerate.
(c) (4 points) In this case, the energy operator is:
k2
k 2 Lz 2
Lx 2 + Ly 2
Lz 2
L
Lz 2
L
f Lz 2
E=
+
=
+
=

.
2I
2I(1 + f)
2I
2I(1 + f)
2I
1 + f 2I

(65)

k 2 and Lz , so
The spherical harmonics Ylm are simultaneous eigenfunctions to both L
the energy eigenfunctions of the above quantum energy operator are again the Ylm :

2



k2
f Lz 2
n l(l + 1)
f n2 m2
L

Ylm =

Ylm
EYlm =
2I
1 + f 2I
2I
1 + f 2I
Elm =

f n2 m2
n2 l(l + 1)

.
2I
1 + f 2I

16

(66)

(d) (5 points) Lets take a concrete example, say l = 1 and l = 2:

2I
20

l=2,m=2

15

l=2,m=1

l=1,m=1

10

l=2,m=0
5

l=1,m=0

-1.0

-0.5

0.0

0.5

The dierence between two energy eigenstates El m and Elm is:

El m Elm

f n2 m2 m 2
n2 [l (l + 1) l(l + 1)]
=
+
.
2I
1+f
2I

(67)

The rst fraction on the right hand side of Eq. 67 is a constant and it does not aect
the dierence trend between the eigenenergies; the absolute value of the second fraction
shows how close two eigenenergies can get since it depends on f. We always can assume
2
2
term remaining is
that
J soJexchange El m Elm ) and the Jimportant
J E J m m (if not
J
J
J
. If f 0 then
J E J =
E , but if f f 0 then
J
E J
=
E , so in the former

1+E
1+E
1+E
1E
1+E
case f 0 the two energies are closer. Classically it means that for a given Lz as the
momentum of inertia Iz = I(1 + f) increases the associated rotational energy decreases.
(e) (3 points) From Eq. 66 we see that, for a given l, the m eigenfunctions share the
same eigenvalue, so the degeneracy is only partially lifted. This is because the system
energy is still invariant under Lz Lz , i.e. the rigid rotator starts to spin in the
opposite direction but with the same angular momentum magnitude. The two opposite
rotation directions correspond to +m and m.
We can break this invariance in many ways. For example, we might take Ix 6= Iy
then we could tell whether we were spinning one way or the other by using the right
hand rule with one nger along x and another along y. Unfortunately, the resulting
17

energy eigenvalues cannot be solved for in closed form (this is directly related to the
fact that the totally asymmetric rotor, Ix = Iy = Iz , is classically chaotic).
Instead, we can break this symmetry by adding to the energy operator the term,
E=

k2
f L2z
L

+ qBz Lz .
2I
1 + f 2I

(68)

This term clearly breaks the symmetry Lz Lz . Such a term could arise if we,
say, rubbed Prof. Evans against a cat to charge him with some static electricity, then
k = Bz z in the z direction. The resulting energy eigenvalues
turned on a magnetic eld B
are then,
n2 l(l + 1)
f n2 m2

+ qBz n m
(69)
Elm =
2I
1 + f 2I
Thus breaking the symmetry again lifts the degeneracy.

18

MIT OpenCourseWare
http://ocw.mit.edu

8.04 Quantum Physics I

Spring 2013

For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.