You are on page 1of 13

Quantum Mechanics 1 Homework 5

Ben Sauerwine
Due Friday October 14, 2005

1) (Cohen-Tannoudji) In a two-dimensional vector space, consider the operator


whose matrix in an orthonormal basis {1 , 2 } is written

⎡0 − i ⎤
σy = ⎢ ⎥
⎣i 0 ⎦

a) Is σ y Hermitian? Calculate its eigenvalues and eigenvectors, giving their


normalized expansion in terms of the {1 , 2 } basis.

+
+ ⎡0 − i ⎤ ⎡0 − i ⎤
σy =⎢ ⎥ =⎢ ⎥ =σy
⎣i 0 ⎦ ⎣i 0 ⎦

So this operator is Hermitian. Its eigenvalues are the roots of

⎡− λ − i ⎤
det ⎢ ⎥ = λ2 − 1 : λ ∈ {1,−1}
⎣ i − λ⎦

And so its eigenvectors are

⎡− 1 − i ⎤ ⎡1 i ⎤ 1 ⎡− i ⎤
1 : Null ⎢ ⎥ = Null ⎢ ⎥ = ⎢ ⎥ =
1
(− i 1 + 2 )
⎣ i − 1⎦ ⎣0 0 ⎦ 2⎣1⎦ 2
⎡1 − i ⎤ ⎡1 − i ⎤ 1 ⎡i ⎤
− 1 : Null ⎢ ⎥ = Null ⎢ ⎥ = ⎢ ⎥ =
1
(i 1 + 2 )
⎣i 1 ⎦ ⎣0 0 ⎦ 2 ⎣1⎦ 2

b) Calculate the matrices that represent the projectors onto these eigenvectors
and verify that they satisfy the orthogonality and closure relations.

1 ⎡− i ⎤ 1 − i⎤
e1 e1 = ⎢ ⎥ [i 1] = 1 ⎡⎢ ⎥
1 ⎡i ⎤ 1 ⎡ 1 i⎤
; e−1 e−1 = ⎢ ⎥[− i 1] = ⎢
2⎣ 1 ⎦ 2 ⎣i 1 ⎦ 2 ⎣1⎦ 2 ⎣− i 1⎥⎦
orthogonal
1 1 ⎡1 − i ⎤ ⎡ 1 i ⎤ ⎡0 0⎤
e1 e1 e−1 e−1 = =
2 2 ⎢⎣i 1 ⎥⎦ ⎢⎣− i 1⎥⎦ ⎢⎣0 0⎥⎦
closed
1 ⎡1 − i ⎤ 1 ⎡ 1 i ⎤ ⎡1 0⎤
e1 e1 + e−1 e−1 = + =
2 ⎢⎣i 1 ⎥⎦ 2 ⎢⎣− i 1⎥⎦ ⎢⎣0 1⎥⎦
c) Repeat parts a and b for these matrices:

⎡ 2 i 2⎤
M =⎢ ⎥
⎣− i 2 3 ⎦

eigenvalues:

⎡2−λ i 2 ⎤
⎥ = λ − 5λ + 4 : λ ∈ {1,4}
2
det ⎢
⎣− i 2 3 − λ ⎦

eigenvectors:

⎡ 1
1 : Null ⎢
i 2⎤ ⎡1 i 2 ⎤
⎥ = Null ⎢ ⎥=
1 ⎡− i 2 ⎤
⎢ ⎥=
1
(
−i 2 1 + 2 )
⎣− i 2 2 ⎦ ⎣0 0 ⎦ 3⎣ 1 ⎦ 3
⎡ i ⎤
⎡ − 2 i 2⎤
4 : Null ⎢
− −
⎥ = Null ⎢

1 −
2⎥ =

1 ⎡ i ⎤
⎢ 2⎥ =
1
(
i1 + 2 2 )
⎣ i 2 1 ⎦ ⎣0 0 ⎦ 3 ⎣ ⎦ 3
projectors:

1 ⎡− i 2 ⎤
e1 e1 = ⎢
3⎣ 1 ⎦
[ 1⎡ 2
⎥i 2 1 = ⎢
3 ⎣i 2
] − i 2⎤

1 ⎦
1⎡ i ⎤
e4 e4 = ⎢ ⎥ − i
3 ⎣ 2⎦
[ ]
1⎡ 1
2 = ⎢
3 ⎣− i 2
i 2⎤

2 ⎦
orthogonal
11⎡ 2 − i 2 ⎤⎡ 1 i 2 ⎤ ⎡0 0 ⎤
e1 e1 e4 e4 = ⎢ ⎥⎢ ⎥=⎢ ⎥
3 3 ⎣i 2 1 ⎦ ⎣− i 2 2 ⎦ ⎣0 0 ⎦
closed
1⎡ 2 −i 2⎤ 1 ⎡ 1 i 2 ⎤ ⎡1 0⎤
e1 e1 + e4 e4 = ⎢ ⎥+ ⎢ ⎥=⎢ ⎥
3 ⎣i 2 1 ⎦ 3 ⎣ − i 2 2 ⎦ ⎣0 1 ⎦

⎡ 0 2 0⎤
h⎢ ⎥
Ly = ⎢− 2 0 2⎥
2i ⎢
⎣ 0 − 2 0 ⎥⎦

eigenvalues:
h
=C
2i
⎡ −λ C 2 0 ⎤

det ⎢− C 2 −λ

( ) ( ) ( )
C 2 ⎥ = −λ λ2 + 2C 2 − C 2 λC 2 = −λ λ2 + 4C 2 : λ ∈ {0,2iC ,−2iC}
⎢ 0 − C 2 − λ ⎥⎦

eigenvectors:

⎡ 0 C 2 0 ⎤ ⎡1 0 − 1⎤ ⎡1⎤
⎢ ⎥
0 : Null ⎢− C 2 0 ⎢ ⎥
C 2 ⎥ = Null ⎢0 1 0 ⎥ =
1 ⎢ ⎥
0
⎢ ⎥ =
1
(1 + 3 )
⎢ 0 2 2
⎣ −C 2 0 ⎥⎦ ⎢⎣0 0 0 ⎥⎦ ⎢⎣1⎥⎦

⎡ − 2iC C 2 0 ⎤ ⎡1 0 1 ⎤ ⎡ −1 ⎤

2iC : Null ⎢− C 2 − 2iC

C 2 ⎥ = Null ⎢⎢0 1 ⎥ 1⎢
2
(
2i ⎥ = ⎢− 2i ⎥⎥ = − 1 − 2i 2 + 3
1
2
)
⎢ 0 −C 2 − 2iC ⎥⎦ ⎢⎣0 0 0 ⎥⎦ ⎢⎣ 1 ⎥⎦

⎡ 2iC C 2 0 ⎤ ⎡1 0 1 ⎤ ⎡ −1 ⎤

− 2iC : Null ⎢− C 2 2iC
⎥ ⎢ ⎥ (
C 2 ⎥ = Null ⎢0 1 − 2i ⎥ = ⎢ 2i ⎥⎥ = − 1 + 2i 2 + 3
1⎢
2
1
2
)
⎢ 0 − C 2 2iC ⎥⎦ ⎢⎣0 0 0 ⎥⎦ ⎢⎣ 1 ⎥⎦

projectors:

⎡1⎤ ⎡1 0 1⎤
e0 = ⎢0⎥[1 0 1] = ⎢0 0 0⎥⎥
1⎢ ⎥ 1⎢
e0
2 2
⎢⎣1⎥⎦ ⎢⎣1 0 1⎥⎦

⎡ −1 ⎤ ⎡ 1 − 2i −1 ⎤
e2iC e2iC
4
[
= ⎢− 2i ⎥⎥ − 1
1⎢
] 1⎢
2i 1 = ⎢ 2i
4⎢
2

− 2i ⎥
⎢⎣ 1 ⎥⎦
⎣ −1 2i 1 ⎥⎦

⎡ −1 ⎤ ⎡ 1 2i −1 ⎤
1⎢
4
⎥[ 1⎢
]
e−2iC e−2iC = ⎢ 2i ⎥ − 1 − 2i 1 = ⎢− 2i
4⎢
2

2i ⎥
⎢⎣ 1 ⎥⎦
⎣ − 1 − 2i 1 ⎥⎦
orthogonality:
⎡1 0 1⎤ ⎡ 1 − 2i − 1 ⎤ ⎡0 0 0⎤
1⎢ ⎥ 1⎢ ⎥
e0 e0 e2iC e2iC = ⎢0 0 0⎥ ⎢ 2i 2 − 2i ⎥ = ⎢⎢0 0 0⎥⎥
2 4
⎢⎣1 0 1⎥⎦ ⎢ − 1 2i 1 ⎥⎦ ⎢⎣0 0 0⎥⎦

⎡ 1 − 2i −1 ⎤ ⎡ 1 2i − 1 ⎤ ⎡0 0 0⎤
1⎢ ⎥1⎢ ⎥
e2iC e2iC e−2iC e−2iC = ⎢ 2i 2 − 2i ⎥ ⎢− 2i 2 2i ⎥ = ⎢⎢0 0 0⎥⎥
4⎢ 4
⎣ −1 2i 1 ⎥⎦ ⎢⎣ − 1 − 2i 1 ⎥⎦ ⎢⎣0 0 0⎥⎦
⎡ 1 2i − 1 ⎤ ⎡1 0 1⎤ ⎡0 0 0⎤
1⎢ ⎥1
e−2iC e−2iC e0 e0 = ⎢− 2i 2 2i ⎥ ⎢⎢0 0 0⎥⎥ = ⎢⎢0 0 0⎥⎥
4⎢ 2
⎣ − 1 − 2i 1 ⎥⎦ ⎢⎣1 0 1⎥⎦ ⎢⎣0 0 0⎥⎦
closure:

e0 e0 + e2iC e2iC + e−2iC e−2iC


⎡1 0 1⎤ ⎡ 1 − 2i −1 ⎤ ⎡ 1 2i −1 ⎤
1⎢ ⎥ 1⎢ ⎥ 1⎢ ⎥
= ⎢0 0 0⎥ + ⎢ 2i 2 − 2i ⎥ + ⎢− 2i 2 2i ⎥
2 4⎢ 4⎢
⎢⎣1 0 1⎥⎦
⎣ −1 2i 1 ⎦ ⎥
⎣ − 1 − 2i 1 ⎥⎦
⎡1 0 0 ⎤
= ⎢⎢0 1 0⎥⎥
⎢⎣0 0 1⎥⎦

⎡0 1 ⎤
(Cohen-Tannoudji) Consider the matrix defined by σ x = ⎢ ⎥ . Prove
⎣1 0⎦
eiασ x = I cosα + iσ x sin α .

⎡0 1⎤ even
Notice that my matrix is idempotent: σ x =⎢ ⎥; σ x = I .
odd

⎣1 0⎦
Now consider the Taylor expansion:
eiασ x

=∑
1
(iασ x )n
n =0 n!
∞ ∞
=∑
1
(iασ x )2 n + ∑ 1 (iασ x )2 n+1
n =0 2 n! n=0 (2 n + 1)!
∞ ∞
= I ∑ (− 1) (α )2n + iσ x ∑ (− 1)n 1 (α )2n+1
n1
n =0 2n! n =0 (2n + 1)!
= I cos α + iσ x sin α
2) (Cohen-Tannoudji) In a one-dimensional problem, consider a particle whose
p0 x
i
h
wave function is ψ ( x ) = N
e
where a, p0 are real constants and N is a
x2 + a2
normalization coefficient.

a) Determine N so that ψ ( x ) is normalized.

p0 x p x
∞ ∞ i −i 0
h h
Consider ∫ ψ ( x ) dx = N 2 ∫
e e
dx = 1 . I would like to evaluate this integral in
2

−∞ −∞
x + a2 2

order to determine the normalization constants. I observe that my function has simple
poles at x = ±iα . I choose the curve that goes along the real axis and curves in a wide
arc around infinity: Since the perimeter increases linearly with x and the value of this
function decreases with the square of x, the value of this arc is asymptotically zero.


1 ⎛ 1 ⎞ 2 ⎛1⎞
N2 ∫ dx = N 2 2πi⎜ ⎟ = N π⎜ ⎟ =1
−∞
x +a
2 2
⎝ 2ai ⎠ ⎝a⎠
a
N2 =
π
a
N=
π

b) The position of the particle is measured. What is the probability of finding a


a a
result between − and ?
3 3

a a a
3 3
a ⎛1⎞ ⎛ x⎞ 3
∫ ψ (x )
a 1
dx = ∫ dx = ⎜ ⎟ arctan⎜ ⎟
2

a π a x +a
2 2
π ⎝a⎠ ⎝ a ⎠ x=− a
− − 3
3 3

1⎛ ⎛ 1 ⎞ ⎛ −1 ⎞⎞ 1 ⎛ π − π ⎞ 1
= ⎜⎜ arctan⎜ ⎟ − arctan⎜ ⎟ ⎟⎟ = ⎜ − ⎟=
π⎝ ⎝ 3⎠ ⎝ 3 ⎠⎠ π ⎝ 6 6 ⎠ 3

c) Calculate the mean value of the momentum of a particle which has ψ ( x ) for
its wave function.

The momentum operator here amounts to

∞ ∞
⎛ ∂⎞ ⎛ ∂⎞
∫ψ (x )⎜ − ih ⎟ψ ( x )dx = −ih ∫ψ ( x )⎜ ⎟ψ ( x )dx
* *

−∞ ⎝ ∂x ⎠ −∞ ⎝ ∂x ⎠
Evaluating the derivative operator on my wave function, I get

p0 x p0 x p0 x
i i i
∂ ∂ e h h h
ψ (x ) = N ψ (x) − 2 2 ψ (x )
p e xe p0 x
= Ni 0 − =i
∂x ∂x x 2 + a 2
(x ) x +a
3
h x2 + a2 2
+ a2 2
h

Now substituting this into the equation above, I have

∞ ∞
⎛∂⎞
− ih ∫ψ ( x )⎜ ⎟ψ ( x )dx = −ih ∫ i 0 ψ ( x ) − 2 ψ (x ) dx
* p 2 x 2

⎝ ∂x ⎠ x +a 2
−∞ −∞
h
∞ ∞
ψ ( x ) dx + ih ∫ 2 2 ψ ( x ) dx
p0 x
= −ih ∫ i
2 2

−∞
h −∞
x +a
∞ ∞ ∞
x ih 2x ih 1
= p0 + ih ∫ dx = p0 + ∫ dx = p0 − = p0
−∞ (x 2
+ a2 )
2
(
2 −∞ x 2 + a 2 2
)
2 x + a2
2
( ) x =−∞

(Cohen-Tannoudji) The wave function of a free particle, in a one-dimensional


problem, is given at time t = 0 by:

∞ −k

ψ (x,0) = N ∫ dke k0
eikx where k0 , N are constants.
−∞

a) What is the probability P ( p1 ,0 ) that a measurement of the momentum


performed at time t = 0 will yield a result included between − p1
and p1 ? Sketch the function P ( p1 ,0 ) .

Let me start by transforming this into momentum space: from my Fourier transform,
I know that

p = hk
p

∞ −k ∞ h p
ψ (x,0) = N ∫ dke eikx =
1 1 i x
2πh N ∫ dpe k0 e h
k0

−∞ 2πh h −∞
From the form of the Fourier transform, then, I see that
p

h

ψ ( p ,0 ) = Ne k0
. This function is clearly symmetric about p = 0. Therefore, in
h

order to find my probability I can take


p1 p −2 p
4π 2 1 hk0 ⎛ −2 p1

∫ ψ ( p,0) dp = N ∫ e dp = 2πN 2 k0 ⎜1 − e hk0 ⎟
2

h ⎜ ⎟
− p1 0 ⎝ ⎠

For sample variable choices:


1

0.8

0.6

0.4

0.2

1 2 3 4 5

b) What happens to this probability P ( p1 , t ) if the measurement is


performed at time t?

In its general format, this looks like


p
− ⎡p
∞ h p2 ⎤
i ⎢ x− t⎥
ψ (x, t ) = N ∫ dpe k0
e ⎣⎢ h 2 mh ⎦⎥

−∞
using the same strategy as before, I’m left with
p

h p2
2π −i
ψ ( p, t ) = N
t
k0 2 mh
e
h
p1 p −2 p
4π 2 1 hk0 ⎛ −2 p1

∫ ψ ( p , t ) 2
dp = N ∫0 e dp = 2πN 2
k 0


1 − e hk0 ⎟

− p1
h ⎝ ⎠
where the imaginary parts from the formula above have cancelled one another in
the complex conjugate. This indicates that the expected value of momentum is
not a function of time.

c) What is the form of the wave packet at time t = 0? Calculate for this
time the product Δx ⋅ Δp ; what is your conclusion? Describe
qualitatively the subsequent evolution of the wave packet.

Let me now evaluate the integral for real:


∞ −k ∞ −k

ψ (x,0) = N ∫ dke k0
e ikx 2k 0
= N ∫ dke k0
(e ikx
)
+ e −ikx = N
k0 x 2 + 1
2
−∞ 0
At t = 0, the particle’s location is symmetric about x = 0. As you move away
from the x, the wave dies off as an inverse exponential.
Now to find the uncertainty relation, I choose the RMS value of Δx ⋅ Δp .

2

⎛ 2k 0 ⎞
∞∞ 2
Δx = ∫ x ψ (x,0 ) dx = ∫ x ⎜ N 2 2 ⎟ dx = ∫ xN
1 4k 0
2 2
⎜ 2
⎟ 2 2 2

( )
2k0 xdx
⎝ k0 x + 1 ⎠
2
2 −∞ k0 x + 1
2 2
−∞ −∞

2 ∞ ∞ 2 ∞
2k 0 1 4k 1
= − xN + ∫ N 2 2 20 dx = 2k0 N 2 ∫ 2 2 dx =
( ) ( ) ( )
2 2

k0 x 2 + 1 k0 x + 1 −∞ k 0 x + 1
2
x =−∞
2 −∞

⎛1 ⎞ ⎛ π −π ⎞
2k0 N ⎜⎜ arctan(k0 x )⎟⎟ = 2k 0 N 2 ⎜ − ⎟ = 2πk0 N
2 2 2

⎝ k0 ⎠ x=−∞ ⎝2 2 ⎠

Similarly,

p

h

ψ ( p ,0 ) = Ne k0

h
2
⎛ −
p
⎞ −2
p
∞ ⎜ 2π h ⎟ ∞ h ∞ −2 p
⎟ = ∫ p 2 2πN e k0 dp = 4πN
2 2
Δp 2 = ∫ p 2 ⎜ Ne k0 ∫p e
2 hk0
dp
⎜ h ⎟ h h
−∞
⎜ ⎟ −∞ 0
⎝ ⎠

( )
2p
4πN 2 − 1 − hk0 πN 2
= e hk0 h 2 k0 + 2hpk0 + 2 p 2 = h 3k0 = πN 2h 2 k0
2 3 3

h 4 h
p =0

Δx 2 ⋅ Δp 2 = 2π 2 N 4 k04
Now I have , which is certainly not a surprising form.
Δx ⋅ Δp = 2πN 2 k02

As the time increases, the center of this distribution will move according
hk 2
to 0 = kx − t.
2m
6) (Cohen-Tannoudji) Consider the three-dimensional wave function
⎡ x y z ⎤
−⎢ + + ⎥
ψ (x, y, z ) = Ne ⎣ 2 a 2b 2c ⎦

where a, b and c are three positive lengths.

a) Calculate the constant N that normalizes the wave function.


∞ ∞∞∞ x y z ∞ z∞ y∞ x
− − − − − −
∫ ψ ( x, y , z ) = 8N ∫∫∫e e e dxdydz = 8 N ∫e ∫e ∫e
2 2 a b c 2 c b a
dxdydz
−∞ 0 0 0 0 0 0
∞ z∞ y ∞ z
− − −
= 8N 2 ∫ e c
∫ e b adydz = 8 N a ∫ e c bdz = 8N abc = 1
2 2

0 0 0

1
N=
8abc

b) Calculate the probability that a measurement of X will yield a result


included between 0 and a.

∞∞ a
( ) ( )
x y z
− − −
∫ ψ ( x, y , z )
1 1
= 4N ∫∫∫e e e dxdydz = 1 − e −1 abc = 1 − e −1
2 2 a b c

0 0 0
2abc 2

c) Calculate the probability that simultaneous measurements of Y and Z


will yield results included respectively between –b and b and –c and c.

c b∞
( )( ) ( )
x y z
− − −
∫ ψ ( x, y , z )
1
∫∫∫e
2
= 8N 1 − e −1 1 − e −1 abc = 1 − e −1
e e dxdydz =
2 2 a b c

0 0 0
abc
d) Calculate the probability that a measurement of the momentum will yield
a result included in the element dp x dp y dpz centered at the point p x = p y = 0 ,
h
pz =
c
Let me Fourier transform this into momentum space. In three dimensions, it reads
vv
∞ − ip⋅r
v v v
ψ ( p) = ∫ψ (r )e
1
3
h
d 3r
2πh −∞

∞ x y z − ixp x − iyp y − izpz


v − − −
ψ ( p) =
1 1
2πh
3
8abc ∫e
−∞
2a
e 2b
e2c
e h
e h
e h
dxdydz

∞ x y z −ixpx −iyp y −izpz


v − − −
ψ ( p) =
1 1
3
8∫ e 2 a e 2 b e 2 c e h
e h
e h
dxdydz
2πh 8abc 0
v 2h 8 ⎡ ⎤⎡ ⎤⎡ ⎤
ψ ( p) =
a b c
⎢ ⎥⎢ ⎥⎢ ⎥
2πh
3
abc ⎣ h + 2aip x ⎦ ⎢⎣ h + 2bip y ⎥⎦ ⎣ h + 2cipz ⎦

v 2h ⎡ ⎤⎡ ⎤⎡ ⎤
ψ ( p) =
1 1 1
8abc ⎢ ⎥ ⎢ ⎥⎢ ⎥
⎣ h + 2aip x ⎦ ⎣⎢ h + 2bip y ⎦⎥ ⎣ h + 2cipz ⎦
3
πh
This gives me the amplitude with respect to momentum. All that’s left is to find
the integral of this density function in a small element centered at p x = p y = 0 ,
h
pz = .
c

h
dz +
dx d y

∫ ∫ ∫ψ * ( p , p , p )ψ ( p , p , p )dp dp dp
c

x y z x y z x y z
−d x − d y − d + h
z
c
h
dz + 2
⎡ ⎤
2 2
2
dx d y c
⎡ 1 ⎤ 1 ⎡ 1 ⎤
= 8abc ∫ ∫ ∫ h ⎣ h 2 + (2apx )2 ⎥⎦
⎢ ⎢ 2⎥ ⎢ 2⎥
⎣⎢ h + (2bp y ) ⎦⎥
dp x dp y dp z
πh −d x − d y −d z + ⎣ h + (2cpz ) ⎦
c

⎡ ⎛ 2ap x ⎞ ⎤ ⎡ arctan⎛⎜ 2bp y ⎞⎟ ⎤ ⎡ ⎛ 2cpz ⎞ ⎤


⎢ arctan⎜ ⎟ ⎥⎢ ⎜ ⎟ ⎥ ⎢ arctan⎜ ⎟⎥
2
= 8abc ⎢ ⎝ h ⎠ ⎥⎢ ⎝ h ⎠ ⎥⎢ ⎝ h ⎠⎥
πh ⎢ 2ah ⎥⎢ 2bh ⎥⎢ 2ch ⎥
⎢⎣ ⎥⎦ ⎢ ⎥⎢ ⎥⎦
⎣ ⎦⎣ boundaries
h
d x ,d y , d z +
2 ⎛ 2ap x ⎞ ⎛ 2bp y ⎞ ⎛ 2cpz ⎞ c
= arctan⎜ ⎟ arctan⎜⎜ ⎟⎟ arctan⎜ ⎟
πh 4
⎝ h ⎠ ⎝ h ⎠ ⎝ h ⎠ p =− d h
x x , p y =− d y , pz = − d z +
c

r
7) (Cohen-Tannoudji) Let ψ ( x, y , z ) = ψ (r ) be the normalized wave function
r
of a particle. Express in terms of ψ (r ) the probability for:

a) a measurement of the abscissa X, to yield a result included between x1


and x2 .

∞ ∞ x2
v v
∫ ∫ ∫ψ (r )ψ (r )dxdydz
*

−∞−∞ x1

b) a measurement of the component Px of the momentum to yield a result


included between P1 and P2 .
vv
∞ − ip⋅r
v
ψ (P ) =
v v
∫ψ (r )e
1
3
h
d 3r
2πh −∞
∞ ∞ P2
v v
∫ ∫∫ ψ *
(P )
ψ (P )dPx dPy dPz
−∞−∞ P1

c) Simultaneous measurements of X and Pz to yield x1 < x < x2 , Pz ≥ 0


Suppose I only converted the z coordinate into momentum space, and left the
others in position space.

∞ − izPz
ψ ( x, y, Pz ) = ψ (x, y, z )e
1
2πh −∫∞
h
dz

Now integrate
∞ ∞ x2

∫ ∫ ∫ψ (x, y, P )ψ (x, y, P )dxdydP


*
z z z
0 −∞ x1

d) Simultaneous measurements of Px , Py , Pz to yield


p1 < Px < p2 , p3 < Py < p4 , p5 < Pz < p6 .
vv
∞ − ip⋅r
v
ψ (P ) =
v v
∫ψ (r )e
1
3
h
d 3r
2πh −∞
P6 P4 P2
v v
∫ ∫ ∫ψ (P )ψ (P )dPx dPy dPz
*

P5 P3 P1

Show that this probability is equal to the result of b when p3 , p5 → −∞ ,


p 4 , p6 → ∞ .

Clearly, substitution of these values into the integral above will give the result
from b.

e) a measurement of the component U =


1
( X + Y + Z ) of the position to
3
yield a result included between u1 and u2 .

Let me rotate this onto U and then integrate. A valid, orthogonal rotation is
defined by

⎡ 2 2 2 ⎤ ⎡ x ⎤ ⎡U ⎤
1 ⎢ ⎥
⎢ 3 0 − 3 ⎥ ⎢⎢ y ⎥⎥ = ⎢⎢ V ⎥⎥
6⎢
⎣ −1 2 − 1 ⎥⎦ ⎣⎢ z ⎦⎥ ⎣⎢W ⎦⎥

where I have chosen random vectors and used the cross product to find two
orthogonal vectors. Inverting,

⎡ x⎤ ⎡ 2 3 − 1⎤ ⎡U ⎤
⎢ ⎥
⎢ y⎥ = 2 ⎥ ⎢⎢ V ⎥⎥
1
⎢ ⎥ ⎢ 2 0
6⎢
⎢⎣ z ⎥⎦
⎣ 2 − 3 − 1⎥⎦ ⎢⎣W ⎥⎦
Now suppose I wanted to integrate over my U, V, W coordinates but use the wave
function in terms of the x, y, z coordinates. This is no problem.

∞ ∞ u2
⎛ ⎡
⎜ 1 ⎢ 2 3 − 1⎤ ⎞⎟ ⎛⎜ ⎡ 2 − 1⎤ ⎞⎟
3
⎥v 1 ⎢ ⎥v
∫−∞−∫∞u∫ψ ⎜⎜ 6 ⎢ 2 2 ⎥ r ⎟ψ ⎜ 2 ⎥ r ⎟dudvdw
*
0 ⎢ 2 0
⎟ ⎜ 6 ⎟
1 ⎜ ⎢ 2
⎣ − 3 − 1⎥⎦ ⎟ ⎜ ⎢ 2
⎣ − 3 − 1⎥⎦ ⎟
⎝ ⎠ ⎝ ⎠
v
8) (Cohen-Tannoudji) Let J (r ) be the probability current associated with a
v
wave function ψ (r ) describing the state of a particle of mass m.

v
a) Show that m ∫ d 3 rJ (r ) = P , where P is the mean value of momentum.

I know from a strategic manipulation of the Schrodinger equation that

∂ v
∂t
ρ (r , t ) +
h
2mi
[
v v v v
]
ψ * (r , t )∇ 2ψ (r , t ) − ψ (r , t )∇ 2ψ * (r , t ) = 0

And from conservation of probability that

∂ v v v v
ρ (r , t ) + ∇ ⋅ J (r , t ) = 0
∂t
Using integration by parts,

v v r
∇ ⋅ J (r , t ) =
h
2mi
[ v v v v
]
ψ * (r , t )∇ 2ψ (r , t ) − ψ (r , t )∇ 2ψ * (r , t )
v r
J (r , t ) = ∫
h
2mi
[ v v v v
]
ψ * (r , t )∇ 2ψ (r , t ) − ψ (r , t )∇ 2ψ * (r , t ) d 3 r
v v v
⎡ψ * (rv , t )∇ψ (rv, t ) − ∇ψ (rv, t )∇ψ * (rv, t ) ⎤
v r
J (r , t ) =
h ⎢ ∫ ⎥
v * v v v
2mi ⎢− ψ (r , t )∇ψ (r , t ) + ∇ψ (rv , t )∇ψ * (rv, t )⎥
v
⎣ ∫ ⎦
v r
J (r , t ) =
h
[ v v v v v v
] 1 ⎡ v hv v ⎤
ψ * (r , t )∇ψ (r , t ) − ψ (r , t )∇ψ * (r , t ) = Re ⎢ψ * (r , t ) ∇ψ (r , t )⎥
m ⎣
2mi i ⎦
mv r ⎡1 v v v ⎤
J (r , t ) = Re ⎢ ψ * (r , t )∇ψ (r , t )⎥
h ⎣i ⎦

I certainly know that the expected value of momentum is


v⎛ ⎛∂ ∂ ∂ ⎞⎞ v v
∫d rψ * (r )⎜⎜ − ih⎜⎜ xˆ +
3
yˆ + zˆ ⎟⎟ ⎟⎟ψ (r ) = P
⎝ ⎝ ∂x ∂y ∂z ⎠ ⎠
v ⎛1 v ⎞ v v
h ∫ d 3 rψ * (r )⎜ ∇ ⎟ψ (r ) = P
⎝i ⎠
mv v v
h ∫ d 3 r J (r ) = P
h
v v v
m ∫ d 3 rJ (r ) = P

where I have assumed that the integral over the space is real.
v
b) Consider
v v v the operator L (orbital angular momentum) defined by
L = r × p . Are the three components Hermitian operators? Establish the
v v v
[ ]v
relation m ∫ d 3 r r × J (r ) = L .

From the definition,

v ⎡ ∂ ∂⎤ ⎡ ∂ ∂⎤ ⎡ ∂ ∂⎤
L = −ih ⎢ y − z ⎥ xˆ − ih ⎢ z − x ⎥ yˆ − ih ⎢ x − y ⎥ zˆ
⎣ ∂z ∂y ⎦ ⎣ ∂x ∂z ⎦ ⎣ ∂y ∂x ⎦

Each component clearly equals its own adjoint: let me show this for the x
component as an example. Note the change of the coordinate in the operators.

+ ⎡ ∂ ∂⎤ ⎡ ∂ ∂⎤
Lx = ih ⎢ z − y ⎥ = −ih ⎢ y − z ⎥ = Lx
⎣ ∂y ∂z ⎦ ⎣ ∂z ∂y ⎦

Now I substitute starting with the definition of the average value of orbital
angular momentum.

∫ d 3
r ψ (r[
* v v
) r × − ih
v
∇ ψ ]
(r
v
) =
v
L

This final substitution requires me to notice that the constant wave functions
commute through the cross product:

∫ d r [r × −ψ (r )ih∇ψ (r )] =
3 v v * vv v
L

But from part a, this is just

[ v v v
] v
m ∫ d 3 r r × J (r ) = L

You might also like