Solutions to Quantum Mechanics by Claude

Cohen-Tannoudji, Bernard Diu, and Franck Laloe

Benjamin D. Suh

September 21, 2020

2
Contents

1 Waves and Particles. Introduction to the Fundamental Ideas of Quantum Me-

chanics 5
1.1 Multiple-Slit Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Bound State of a Particle in a ”Delta Function Potential” . . . . . . . . . . . . . . . 7
1.3 Transmission of a ”delta function” potential barrier” . . . . . . . . . . . . . . . . . . 10
1.4 Delta potential, Fourier transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5 Well consisting of two delta functions . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.6 Square Well Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2 The Mathematical Tools of Quantum Mechanics 19

2.1 Hermitian Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2 Pauli Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3 Kets and Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.4 Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5 Orthogonal Projector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.6 Pauli Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.7 Pauli Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.8 2.8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.9 2.9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.10 2.10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.11 Commuting Observables and CSCOs . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.12 Spin Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3 The Postulates of Quantum Mechanics 37

3.1 Wave Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2 Measurement of a one-dimensional particle . . . . . . . . . . . . . . . . . . . . . . . 39
3.3 Free Particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.4 Spreading of a Free Wave Packet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.5 Particle Subject to a Constant Potential . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.6 Three-dimensional wave function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.7 Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.8 Probability current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.9 Probability current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.10 Virial Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.11 Two-particle wave function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3
4 CONTENTS

3.12 Infinite one-dimensional well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

3.13 Infinite two-dimensional well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.14 Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Chapter 1

Waves and Particles. Introduction

to the Fundamental Ideas of
Quantum Mechanics

1.1 Multiple-Slit Experiment

A beam of neutrons of constant velocity, mass Mn (Mn ≈ 1.67x10−27 kg) and energy E,
is incident on a linear chain of atomic nuclei, arranged in a regular fashion as shown
in the figure (these nuclei could be, for example, those of a long, linear molecule).
We call l the distance between two consecutive nuclei, and d, their size (d  l). A
neutron detector D is placed far away, in a direction which makes an angle of θ with
the direction of the incident neutrons.

1.1.a Describe qualitatively the phenomena ob-

Figure 1.1: Multiple-slit
Experiment served at D when the energy E of the incident neu-
trons is varied.

From Complement DI , the fringe separation of the double-slit experi-

λD
ment is given by . Using the Planck-Einstein relations(A-1), we know
l
that as energy increases, the wavelength decreases. Thus, the distance
between successive peaks decreases as well, so we end up seeing more
peaks in a given area.

5
6CHAPTER 1. WAVES AND PARTICLES. INTRODUCTION TO THE FUNDAMENTAL IDEAS OF QUANTUM

1.1.b The counting rate, as a function of E, presents a resonance about

E = E1 . Knowing that there are no other resonances for E < E1 ,
show that one can determine l. Calculate l for θ = 30° and E1 =
1.3 × 10−20 joule.
From complement DI , for a double slit experiment,

l sin(θ) = mλ

Since we are looking at the first interference, we set m = 1. Using the Planck-Einstein
relations(A-1), we can rewrite the right-side of the equation,

hc
l sin(θ) =
E

hc
l=
E sin(θ)

Plugging in the given variables and constants, we find l = 1.61 × 10−14 m. Note that we ignore
d, the diameter of each nuclei since we assumed d  l.

1.1.c At about what value of E must we begin to take the finite size of
the nuclei into account?
Using Heisenburg’s uncertainty relation(C-23), and rewriting the momentum in terms of energy,

~
∆x · ∆p ≥
2

h ~
∆x · ≥
λ 2

hc
∆x ≥
4πE

hc
E≥
4π · ∆x
The size of atomic nuclei are on the order of fermi, 10−15 m, so plugging in those numbers, we
get something on the order of 8 × 10−21 J.
1.2. BOUND STATE OF A PARTICLE IN A ”DELTA FUNCTION POTENTIAL” 7

1.2 Bound State of a Particle in a ”Delta Function Poten-

tial”
Consider a particle whose Hamiltonian H [operator defined by formula (D-10) is:

~2 d 2
H =− − αδ(x)
2m dx2
where α is a positive constant whose dimensions are to be found.

1.2.a Integrate the eigenvalue equation of H between − and +. Letting

 approach 0, show that the derivative of the eigenfunction φ(x)
presents a discontinuity at x = 0 and determine it in terms of α, m,
and φ(0).
Nothing doing, let’s do as the question asks,
Z  Z 
H |φi dx = E |φi dx
− −

  
~2 d2
Z Z Z
− φ(x) dx − αδ(x)φ(x) dx = E φ(x) dx
− 2m dx2 − −

On the right, the integral disappears as  → 0. The delta function only turns on if the integral
includes x = 0, and it picks out that value in that integral,
x=
~2 dφ(x)
− − αφ(0) = 0
2m dx x=−

dφ(x) dφ(x) 2mαφ(0)

− =−
dx x= dx x=− ~2

As  → 0, the left side would disappear unless there were a discontinuity. Since the right side is
not equal to zero, there must be a discontinuity in the derivative of φ(x).

1.2.b Assume that the energy E of the particle is negative(bound state).

φ(x) can then be written:
(
x<0 φ(x) = A1 exp(ρx) + A01 exp(−ρx)
x>0 φ(x) = A2 exp(ρx) + A02 exp(−ρx)
8CHAPTER 1. WAVES AND PARTICLES. INTRODUCTION TO THE FUNDAMENTAL IDEAS OF QUANTUM

Express the constant ρ in terms of E and m. Using the results of the preceding question,
calculate the matrix M defined by:
   
A2 A1
=M
A02 A01

Then, using the condition that φ(x) must be square-integrable, find the possible values
of the energy. Calculate the corresponding normalized wave functions.
We can find ρ by using the Schrodinger equation(B-8),

~2 d 2
− φ(x) = Eφ(x)
2m dx2
We only need to look at φ(x) at a single time, so lets look at the case x < 0,

~2
− (A1 ρ2 exp(ρx) + A01 ρ2 exp(−ρx)) = E(A1 exp(ρx) + A01 exp(−ρx))
2m

~2 ρ 2
− =E
2m

2mE
ρ2 = −
~2
To calculate the required matrix, we use the discontinuity in the derivative,

dφ(x) dφ(x) 2mαφ(0)

− =−
dx x= dx x=− ~2

The first term we will need to use the x > 0 case and x < 0 for the second,
2mα
ρA2 exp(ρ) − A02 ρ exp(−ρ) − A1 ρ exp(ρ) + A01 ρ exp(−ρ) = − (A1 + A01 )
~2
Note that we chose an arbitrary case for x = 0. Because φ(x) must be continuous, A1 + A01 =
A2 + A02 . Letting  → 0, we have,

2mα
A2 − A02 − A1 + A01 = − (A1 + A01 )
ρ~2

2mα
A2 = − (A1 + A01 ) + A1 − A01 + A02
ρ~2

Using the continuity of the wavefunction,

 
2mα 2mα 0
2A2 = A1 − 2 + 2 − A
ρ~ ρ~2 1
1.2. BOUND STATE OF A PARTICLE IN A ”DELTA FUNCTION POTENTIAL” 9
 
2mα 2mα
2A02 = 0
A1 + A1 2 +
ρ~2 ρ~2

 mα mα 
− +1 −
 ρ~2 ρ~2  A 
 
A2 1
=
A02  A01
 
mα mα
+1
ρ~2 ρ~2
Let’s now normalize the wavefunction,
Z ∞
|φ(x)|2 dx = 1
−∞

If we have a particle coming from the left, we want to set A2 = 0. Furthermore, since we want
this to be square-integrable, we must set A01 = 0. Since the wavefunction must be continuous,
A1 = A02 .
Z 0 Z ∞
A21 exp(2ρx) dx + A21 exp(−2ρx) dx = 1
−∞ 0

 
1 1
A21 + =1
2ρ 2ρ

A21 = ρ
Our normalized wavefunction is given,
( √
x < 0 φ(x) = ρ exp(ρx)
√
x > 0 φ(x) = ρ exp(−ρx)
To find the allowed energy, we go back to the discontinuity equation
dφ(x) dφ(x) 2mαA1
− =−
dx x= dx x=− ~2

2mαA1
ρA1 + ρA − 1 = −
~2

ρ~2 = −mα
Squaring both sides then substituting in our value of ρ,
ρ2 ~4 = m2 α2

−2mE~2 = m2 α2

mα2
E=−
2~2
10CHAPTER 1. WAVES AND PARTICLES. INTRODUCTION TO THE FUNDAMENTAL IDEAS OF QUANTUM

1.3 Transmission of a ”delta function” potential barrier”

Consider a particle placed in the same potential as in the preceding exercise. The
particle is now propagating from left to right along the x axis, with a positive energy
E.

1.3.a Show that a stationary state of the particle can be written:

(
x<0 φ(x) = exp(ikx) + A exp(−ikx)
x>0 φ(x) = B exp(ikx)
where k, A, and B are constants which are to be calculated in terms of the energy E,
dφ
of m and of α(watch out for the discontinuity in at x = 0.
dx
By observation, the given wavefunction is not dependant on time, and it follows the form given
by (D-7), so it is a stationary state. To determine the constants, we can look at the Schrodinger
equation(B-8),
~2 d2
− φ(x) − αδ(x)φ(x) = Eφ(x)
2m dx2
This holds true for all values of x, so let’s look at x > 0. The delta function term dies,
~2
− (−Bk 2 exp(ikx)) = EB exp(ikx)
2m
~2 k 2
=E
2m
2mE
k2 =
~2
At x = 0, we use the wavefunction continuity,
B =1+A
as well as the discontinuity in the first derivative,
dφ(x) dφ(x) 2mαφ(0)
− =−
dx x= dx x=− ~2

2mαB
ikB − ik(1 − A) = −
~2

B(ik~2 + 2mα) = ik~2 (1 − A)

ik~2
B=
ik~2 + mα
mα
A=−
ik~2 + mα
1.3. TRANSMISSION OF A ”DELTA FUNCTION” POTENTIAL BARRIER” 11

1.3.b Set −EL = −mα2 /2~2 (bound state energy of the particle). Calcu-
late, in terms of the dimensionless parameter E/EL , the reflec-
tion coefficient R and the transmission coefficient T of the bar-
rier. Study their variations with respect to E; what happens when
E → ∞? How can this be interpreted? Show that, if the expression
of T is extended for negative values of E, it diverges when E → EL ,
and discuss this result.
To find the reflection coefficient, we look at the parameters of incoming and outgoing particles. A
particle moving to the right has magnitude 1, while a particle moving to the left has magnitude A,

m2 α2
R = |A|2 =
k 2 ~4 + m2 α2

m2 α2
=
m2 α2 (1 + E/EL )

1
R=
1 + E/EL

Similarly, a particle moving to the right after passing through the barrier has magnitude B,

k 2 ~4
T = |B|2 =
k 2 ~4 + m 2 α 2

E/EL
T =
1 + E/EL

We see that T + R = 1. As E → ∞, T → 1 and R → 0, which means the wavefunction has more

and more energy to overcome the barrier. As energy increases, the probability that the particle will
go through the barrier increases. At E = −EL , it has the same energy as the barrier, which is why
we see that discontinuity. Classically, think of this as a ball hitting the top edge of a wall. Small
variations around this point determine if the ball is reflected or transmitted.
12CHAPTER 1. WAVES AND PARTICLES. INTRODUCTION TO THE FUNDAMENTAL IDEAS OF QUANTUM

1.4 Delta potential, Fourier transform

Return to exercise 2, using, this time, the Fourier transform.

1.4.a Write the eigenvalue equation of H and the Fourier transform of

this equation. Deduce directly from this the expression for φ̄(p),
the Fourier transform of φ(x), in terms of p, E, α, and φ(0). Then
show that only one value of E, a negative one, is possible. Only the
bound state of the particle, and not the ones in which it propagates,
is found by this method; why? Then calculate φ(x) and show that
one can find in this way all the results of exercise 2.
Looking at Appendix I, the Fourier transform of the eigenvalue equation is
 2
~2 ip
− φ̄(p) − F [αδ(x)φ(x)] = E φ̄(p)
2m ~
We get this using
Z ∞  
1 ipx
φ̄(p) = √ exp − φ(x) dx
2π~ −∞ ~

Z ∞  
1 ipx
φ(x) = √ exp φ̄(p) dp
2π~ −∞ ~
We note that
2
d2

ip
φ̄(p) = φ̄(p)
dx2 ~
We need to be careful about the Fourier transform of the delta function.
Z ∞  
1 ipx
F [αδ(x)φ(x)] = √ exp − αδ(x)φ(x) dx
2π~ −∞ ~

1
=√ αφ(0)
2π~
Substituting this back in,
p2 αφ(0)
φ̄(p) − √ = E φ̄(p)
2m 2π~

αφ(0) 1
φ̄(p) = √ 2
2π~ (p /2m − E)
1.4. DELTA POTENTIAL, FOURIER TRANSFORM 13

We can reverse our Fourier transform to get φ(x),

 
ipx
Z ∞ exp
αφ(0) ~
φ(x) = dp
2π~ −∞ (p2/2m − E)

 √ 
αφ(0)m −2mE
φ(x) = √ exp − x
~ −2mE ~

To determine the allowed energy, let’s set x = 0 and match sides,

αmφ(0)
φ(0) = √
~ −2mE

In order for this to hold true,

αm
√ =1
~ −2mE

α2 m
E=−
2~2

1.4.b The average kinetic energy of the particle can be written (cf. chap.
III):
Z ∞
1
Ek = p2 |φ̄(p)|2 dp
2m −∞

Show that, when φ̄(p) is a ”sufficiently smooth” function, we also have:

∞
~2 d2 φ
Z
Ek = − φ∗ (x) dx
2m −∞ dx2

These formulas enable us to obtain, in two different ways, the energy Ek for a particle
in the bound state calculated in (a). What result is obtained? Note that, in this case
φ(x) is not ”regular” at x = 0, where its derivative is discontinuous. It is then necessary
to differentiate φ(x) in the sense of distributions, which introduces a contribution of
the point x = 0 to the average value we are looking for. Interpret this contribution
physically: consider a square well, centered at x = 0, whose width a approaches 0 and
whose depth V0 approaches infinity (so that aV0 = α), and study the behaviour of the
wave function in this well.
Starting with the second equation and substituting in the definitions from Appendix I,
∞ 2
d2 φ
Z     
1 ipx ipx ipx
φ∗ (x) = exp − φ̄(p) exp dp
dx2 2π~ −∞ ~ ~ ~
14CHAPTER 1. WAVES AND PARTICLES. INTRODUCTION TO THE FUNDAMENTAL IDEAS OF QUANTUM
Z ∞ Z ∞
1
Ek = p2 |φ̄(p)|2 dp dx
4mπ~ −∞ −∞

The constants in front disappear when we integrate over all x,

Z ∞
1
Ek = p2 |φ̄(p)|2 dp
2m −∞
1.5. WELL CONSISTING OF TWO DELTA FUNCTIONS 15

1.5 Well consisting of two delta functions

Consider a particle of mass m whose potential energy is

V (x) = −αδ(x) − αδ(x − l) α > 0

where l is a constant length.

~2 ρ2
1.5.a Calculate the bound states of the particle, setting E = − . Show
2m
that the possible energies are given by the relation
 
2ρ
exp(−ρl) = ± 1 −
µ
2mα
where µ is defined by µ = 2 . Give a graphical solution of this equation.
~
We can divide this into three regions,

A exp(ρx) x < 0

φ(x) = B(exp(−ρx) + exp(ρ(x − l))) 0 < x < l

A exp(−ρ(x − l)) x > l


We get this because we know that the wavefunction corresponding to a delta function is ex-
ponential on both sides, so we can reasonable expect the wavefunction corresponding to two delta
functions is going to more exponential functions.
To solve for the bound states, let’s start by looking at x = 0. From continuity,

A = B(1 + exp(−ρl))

From problem 2, we know there is a discontinuity in the derivative,

dφ dφ
− = −µφ(0)
dx x= dx x=−

−ρB + ρB exp(−ρl) − ρA = −µA

−ρ + ρ exp(−ρl) − ρ − ρ exp(−ρl) = −µ(1 + exp(−ρl))

−2ρ = −µ(1 + exp(−ρl))

2ρ
exp(−ρl) = 1 −
µ
16CHAPTER 1. WAVES AND PARTICLES. INTRODUCTION TO THE FUNDAMENTAL IDEAS OF QUANTUM

The bound states are given by

 
2ρ
exp(−ρl) = ± 1 −
µ

(i)Ground State. Show that this state is even (invariant with respect to reflection
about the point x = l/2), and that its energy Es is less than the energy −EL introduced
in problem 3. Interpret this result physically. Represent graphically the corresponding
wave function.
1.6. SQUARE WELL POTENTIAL 17

1.6 Square Well Potential

Consider a square well potential of width a and depth V0 (in this exercise, we shall
use systematically the notation of 2-c-α of complement HI ). We intend to study the
properties of the bound state of a particle in a well when its width a approaches zero.

1.6.a Show that there indeed exists only one bound state and calculate
 mV0 a2
its energy E we find E ≈ − , that is, an energy which varies
2~2 
with the square of the area aV0 of the well .
18CHAPTER 1. WAVES AND PARTICLES. INTRODUCTION TO THE FUNDAMENTAL IDEAS OF QUANTUM
Chapter 2

The Mathematical Tools of

Quantum Mechanics

2.1 Hermitian Operator

|φn i are the eigenstates of a Hermitian operator H (H is, for example, the Hamilto-
nian of an arbitrary physical system). Assume that the states |φn i form a discrete
orthonormal basis. The operator U (m, n) is defined by:
U (m, n) = |φm i hφn |

2.1.a Calculate the adjoint U † (m, n) of U (m, n)

Using the definition of the adjoint,
U † (m, n) = |φn i hφm |

2.1.b Calculate the commutator [H, U (m, n)]

Let’s act the commutator on a vector (looking ahead, we’ll set the vector as |φn i,
[H, U ] |φn i = HU |φn i − U H |φn i

= H |φm i hφn |φn i − |φm i hφn |H|φn i

Since H has eigenkets φn and φm , i.e.,
(
H |φm i = m |φm i
H |φn i = n |φn i

[H, U (m, n)] = m − n

19
20 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

2.1.c Prove the relation

U (m, n)U † (p, q) = δnq U (m, p)

Writing this out,
U (m, n)U † (p, q) = |φm i hφn |φq i hφp |
The middle section dies unless n = q, leaving us the delta function,
= δnq |φm i hφp |

U (m, n)U † (p, q) = δnq U (m, p)

2.1.d Calculate Tr{U (m, n)}, the trace of the operator U (m, n)
By definition, the trace is given by
X
Tr U = hα|U |αi
α

where α are the basis states.

X
= hα|φm i hφn |αi
α

Since |φn i form a basis, and since we sum over the basis states, at least one part dies unless
m = n,
Tr U = δmn

2.1.e Let A be an operator, with matrix elements Amn = hφm |A|φn i. Prove
the relation:
X
A= Amn U (m, n)
m,n

Let’s start from the right side. Acting it on |φn i, we only need to sum over m since we can use
orthonormality for n,
X X
Amn U (m, n) |φn i = hφm |A|φn i |φm i hφn |φn i
m,n m

Amn is a scalar, so we can move that around for free,

X
= |φm i hφm |A|φn i
m

Performing the sum, the first part becomes identity, so we can remove it,
X
Amn U (m, n) |φn i = A |φn i
m,n
2.1. HERMITIAN OPERATOR 21

2.1.f Show that Apq = Tr{AU † (p, q)}

We start with the right side. We can write the part inside the trace using the relation we found in
part (e),
X
AU † (p, q) = Amn U (m, n)U † (p, q)
m,n

Using the relation from (c),

X
= Amn δnq U (m, p)
m,n

X
= Amq U (m, p)
m

Taking the trace and using the result from (d),

X
Tr{AU † (p, q)} = Amq δmp
m

Tr{AU † (p, q)} = Apq

22 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

2.2 Pauli Matrices

In a two-dimensional vector space, consider the operator whose matrix, in an orthonor-
mal basis {|1i , |2i}, is written:
 
0 −i
σy =
i 0

2.2.a Is σy Hermitian? Calculate its eigenvalues and eigenvectors (giving

their normalized expansion in terms of the {|1i , |2i} basis).
σy is Hermitian from observation (1.13). Solving the characteristic equation (1.20), we get two
eigenvalues, λ = ±1. The eigenvectors are,
   
1 −i 1 i
|λ = 1i = √ ; |λ = −1i = √
2 1 2 1
In the {|1i , |2i} basis,
√
(
|λ = 1i = 1/ 2(−i |1i + |2i)
√
|λ = −1i = 1/ 2(i |1i + |2i)

2.2.b Calculate the matrices which represent the projectors onto these
eigenvectors. Then verify that they satisfy the orthogonality and
closure relations.
The projection operator (1.16) for λ = 1,

A = |λ = 1i hλ = 1|

= 1/2(−i |1i + |2i)(i h1| + h2|)

 
1 1 −i
A=
2 i 1

Similarly, for λ = −1,

B = |λ = −1i hλ = −1|

 
1 1 i
B=
2 −i 1

We can show that they are orthonormal by multiplying the two together, AB = BA = 0. We
can show completeness by adding them together, A + B = I.
2.3. KETS AND OPERATORS 23

2.3 Kets and Operators

The state space of a certain physical system is three-dimensional. Let {|u1 i , |u2 i , |u3 i}
be an orthonormal basis of this space. The kets |ψ0 i and |ψ1 i are defined by:
√
(
|ψ0 i = 1/ 2 |u1 i + i/2 |u2 i + 1/2 |u3 i
√ √
|ψ1 i = 1/ 3 |u1 i + i/ 3 |u3 i

2.3.a Are these kets normalized?

To tell if these kets are normalized, we need to find the norm (1.4). Let’s start with |ψ0 i,
√ √
hψ0 |ψ0 i = (1/ 2 hu1 | − i/2 hu2 | + 1/2 hu3 |) (1/ 2 |u1 i + i/2 |u2 i + 1/2 |u3 i)
Since the basis kets are orthonormal, we ignore most of the terms,
= 1/2 hu1 |u1 i + 1/4 hu2 |u2 i + 1/4 hu3 |u3 i = 1
For |ψ1 i,
√ √ √ √
hψ1 |ψ1 i = (1/ 3 hu1 | − i/ 3 hu3 |) (1/ 3 |u1 i + i/ 3 |u3 i)

= 1/3 hu1 |u1 i + 1/3 hu3 |u3 i = 2/3

|ψ0 i is normalized, but |ψ1 i is not.

2.3.b
Calculate the matrices ρ0 and ρ1 representing, in the {|u1 i , |u2 i , |u3 i} basis, the projec-
tion operators onto the state |ψ0 i and onto the state |ψ1 i. Verify that these matrices
are Hermitian.
Using equation (1.16),
ρ0 = |ψ0 i hψ0 |

√ √
= (1/ 2 |u1 i + i/2 |u2 i + 1/2 |u3 i)(1/ 2 hu1 | − i/2 hu2 | + 1/2 hu3 |
We can define our orthonormal basis however we want, but for ease, let’s use,
     
1 0 0
|u1 i = 0 ; |u2 i = 1 ; |u3 i = 0
0 0 1
In this basis,
 √ √ 
1/2 −i/2 2 1/2 2

ρ0 =  i/2√2 1/4 i/4 

√
1/2 2 −i/4 1/4
24 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

We use the Hermitian condition (1.13) to see that ρ0 is Hermitian.

Similarly,

ρ1 = |ψ1 i hψ1 |

√ √ √ √
= (1/ 3 |u1 i + i/ 3 |u3 i)(1/ 3 hu1 | − i/ 3 hu3 |)

 
1/3 0 −i/3
ρ1 =  0 0 0 
i/3 0 1/3

Again, we see that ρ1 is Hermitian.

2.4. OPERATORS 25

2.4 Operators
Let K be the operator defined by K = |φi hψ|, where |φi and |ψi are two vectors of the
state space.

2.4.a
Under what condition is K Hermitian Following the Hermitian condition, we want to show

K = K†

Under dual correspondence (1.1), this translates to,

|φi hψ| = |ψi |φi

This is true for |φi = |ψi.

2.4.b
Calculate K 2 . Under what condition is K a projector?

K 2 = |φi hψ|φi hψ|

Comparing to equation (1.16), K is a projector if |φi = |ψi.

2.4.c
Show that K can always be written in the form K = λP1 P2 where λ is a constant to be
calculated and P1 and P2 are projectors
We set P1 to be the |φi projector and P2 to be the |ψi projector (1.16)
(
P1 = |φi hφ|
P2 = |ψi hψ|

Multiplying them together,

P1 P2 = |φi hφ|ψi hψ|

The middle part is just a scalar, so if we want to get rid of it, we need to multiply by a constant,
1
λ=
hφ|ψi
Combining all of this,

λP1 P2 = |φi hψ| = K

26 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

2.5 Orthogonal Projector

Let P1 be the orthogonal projector onto the subspace E1 , P2 be the orthogonal projector
onto the subspace E2 . Show that, for the product P1 P2 to be an orthogonal projector
as well, it is necessary and sufficient that P1 and P2 commute. In this case, what is the
subspace onto which P1 P2 projects?

Let’s say
(
P1 = |φi hφ|
P2 = |ψi hψ|

where |φi and |ψi are in E1 and E2 respectively. The product is thus,

P1 P2 = |φi hφ|ψi hψ|

To show that this is an orthogonal projection, we assume |φi and |ψi are normalized. We can
then use the property of orthogonal projectors that P1 = P1∗ and P2 = P2∗ ,

P1 P2 = P1∗ P2∗

Using equation (1.12),

= (P2 P1 )∗

If P1 P2 is an orthogonal projection, then P1 P2 = P2 P1 .

Alternatively, if we assume P1 P2 commutes,

(P1 P2 )2 = (P1 P2 )(P1 P2 )∗

= P1 P2 P2∗ P1∗

= P1 P2 P1

Since P1 P2 commutes,

= P1 P1 P2 (2.5.1)

(P1 P2 )2 = P1 P2

P1 P2 projects onto the overlap of E1 and E2 . I’m not entirely sure how to rigorously prove this,
but imagine that E1 is the x − y plane and E2 is the y − z plane. If we apply P2 to a vector, it
projects onto that plane. Then, if we apply P1 to that projection, it must project onto the y-axis
(the overlap) since the vector should have no x component after being projected. We can do the
same for the inverse.
2.6. PAULI MATRICES 27

2.6 Pauli Matrices

The σx matrix is defined by:
 
0 1
σx =
1 0

Prove the relation:

exp(iασx ) = I cos(α) + iσx sin(α)

where I is the 2 × 2 unit matrix.

We can expand the left side using a Taylor expansion,

exp(iασx ) = I + iασx + 1/2 (iα)2 σx2 + ...

α2 0
      
1 0 0 1 1 0 1
= + iα − + ...
0 1 1 0 2 1 0 1 0

 
1 − α2/2 + ... iα + ...
= 2
iα + ... 1 − α /2 + ...

Similarly, if we expand the right side,

 
1 − α2/2 + ... 0
I cos(α) =
0 1 − α2/2 + ...

 
0 iα + ...
iσx sin(α) =
iα + ... 0

You can do this out to an arbitrary number of terms until you convince yourself that this relation
holds true.
28 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

2.7 Pauli Matrices

Establish for the σy matrix given in exercise 2, a relation analogous to the one proved
for σx in the preceding exercise. Generalize for all matrices of the form:

σu = λσx + µσy

with

λ2 + µ2 = 1

Calculate the matrices representing exp(2iσx ), (exp(iσx ))2 and exp(i(σx + σy )) Is exp(2iσx )
equal to (exp(iσx ))2 ? exp(i(σx + σy )) to exp(iσx ) exp(iσy )?

Following the methodology in question 6, we expand the exponential,

exp(iασy ) = I + iασy + 1/2 (iα)2 σy2 + ...

 
1 − α2/2 + ... α + ...
=
−α + ... 1 − α2/2 + ...

We can convince ourselves that this is

exp(iασy ) = I cos(α) + iσy sin(α)

For σu , we can’t use the normal rules of exponential multiplication (which answers the last part
of this question). Expanding,

exp(iα(λ sinx +µσy )) = I + iα(λσx + µσy ) + 1/2 (iα)2 (λ2 σx2 + µ2 σy2 + λµσx σy + λµσy σx )

exp(iα(λσx + µσy )) = I cos(α) + iσx sin(αλ) + iσy sin(αµ)

Using the relation found in question 6,

exp(2iσx ) = I cos(2) + iσx sin(2)

(exp(iσx ))2 = I(cos2 (1) − sin2 (1)) + 2iσx cos(1) sin(1)

These are equal using angle addition formulas.

2.8. 2.8 29

2.8 2.8
30 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

2.9 2.9
2.10. 2.10 31

2.10 2.10
32 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

2.11 Commuting Observables and CSCOs

Consider a physical system whose three-dimensional state space is spanned by the
orthonormal basis formed by the three kets |u1 i, |u2 i, |u3 i. In the basis of these three
vectors, taken in this order, the two operators H and B are defined by
   
1 0 0 1 0 0
H = ~ω0 0 −1 0  ; B = b 0 0 1
0 0 −1 0 1 0
where ω0 and b are real constants.

2.11.a Are H and B Hermitian?

By observation, yes.

2.11.b Show that H and B commute. Give a basis of eigenvectors com-

mon to H and B.
To show they commute,
    
1 0 0 1 0 0 1 0 0
HB = ~ω0 b 0 −1 0  0 0 1 = ~ω0 b 0 0 −1
0 0 −1 0 1 0 0 −1 0
    
1 0 0 1 0 0 1 0 0
BH = ~ω0 b 0 0 1 0 −1 0  = 0 0 −1
0 1 0 0 0 −1 0 −1 0
For the eigenvectors, let’s go ahead and solve the characteristic equation (1.20). Doing so gives
eigenvalues λ = ~ω0 , −~ω0 , −~ω0 . We have a degeneracy for −~ω0 , but we can find the easiest
eigenvectors for the other eigenvalues,
   
1 0
1
|λ = ~ω0 i = 0 ; |(λ = −~ω0 )1 i = √ 1
0 2 1

Since we want this basis to be orthonormal,

 
0
1  
|(λ = −~ω0 )2 i = √ 1
2 −1

We can show that these are a common basis,


H |λ = ~ω0 i = ~ω0 |λ = ~ω0 i

H |(λ = −~ω0 )1 i = −~ω0 |(λ = −~ω0 )1 i

H |(λ = −~ω0 )2 i = −~ω0 |(λ = −~ω0 )2 i

2.11. COMMUTING OBSERVABLES AND CSCOS 33

B |λ = ~ω0 i = b |λ = ~ω0 i

B |(λ = −~ω0 )1 i = b |(λ = −~ω0 )1 i

B |(λ = −~ω0 )2 i = −b |λ = −~ω0 )2 i


2.11.c Of the set of operators {H}, {B}, {H, B}, {H 2 , B}, which form a
CSCO?
For {H} and {B}, these cannot be CSCO since they are degenerate, which means that some
eigenvectors must have the same eigenvalue.
{H, B} is a CSCO since no two eigenvectors have the same set of eigenkets.
H 2 is the identity matrix multiplied by some scalar constant, so we can easily convince ourselves
that it commutes with H and B, which means that the eigenvectors of H are shared with H 2 .
Further, since H 2 is basically the identity matrix, all the eigenvalues are going to be the same
(λ = (~ω0 )2 ). This means that {H 2 , B} is not a CSCO since there are two eigenvectors which have
the same eigenvalues.
34 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

2.12 Spin Operators

In the same state space as that of the preceding exercise, consider two operators Lz
and S defined by:
(
Lz |u1 i = |u1 i ; Lz |u2 i = 0; Lz |u3 i = − |u3 i
S |u1 i = |u3 i ; S |u2 i = |u2 i ; S |u3 i = |u1 i

2.12.a Write the matrices which represent, in the {|u1 i , |u2 i , |u3 i} basis,
the operators Lz , L2z , S, S 2 . Are these operators observables?
From observation,
 
1 0 0
Lz = 0 0 0
0 0 −1

 
0 0 1
S = 0 1 0
1 0 0

From this,
 
1 0 0
L2z = 0 0 0
0 0 1

 
1 0 0
S 2 = 0 1 0
0 0 1

These are all Hermitian (1.13) and observables.

2.12.b Give the form of the most general matrix which represents an
operator which commutes with Lz . Same question for L2z , then
for S 2 .
Let’s say we have some matrix,
 
m11 m12 m13
M = m21 m22 m23 
m31 m32 m33

In order for M to commute, it must be diagonal.

2.12. SPIN OPERATORS 35

Acting Lz on it,
 
m11 m12 m13
Lz M =  0 0 0 
−m31 −m32 −m33

 
m11 0 −m13
M Lz =  0 0 0 
m31 0 −m33
In order for these to commute, only m11 and m33 can be non-zero, and we can add a term in
the middle since m22 is unrestrained,
 
m11 0 0
M = 0 m22 0 
0 0 m33

We can repeat the process for L2z ,

 
m11 0 m13
L2z M =  0 0 0 
m31 0 m33

 
m11 0 m13
M L2z =  0 0 0 
m31 0 m33
All four corners survive, and we can again add a term in the middle,
 
m11 0 m13
M = 0 m22 0 
m31 0 m33

Since S 2 is the identity matrix, any matrix will commute with it.

2.12.c Do L2z and S form a CSCO? Give a basis of common eigenvectors.

We’ll solve the characteristic equation (1.20) for L2z , giving us eigenvalues λ = 0, 0, 1. Two eigen-
vectors,
   
1 0
1  
|λ = 1i = √ 0 ; |(λ = 0)1 i = 1
2 1 0
By orthonormality,
 
1
1  
|(λ = 0)2 i = √ 0
2 −1
36 CHAPTER 2. THE MATHEMATICAL TOOLS OF QUANTUM MECHANICS

To see if they form a CSCO, let’s look at the eigenvalues,


2
Lz |λ = 1i = |λ = 1i

L2z |(λ = 0)1 i = 0 |(λ = 0)1 i

 2
Lz |(λ = 0)2 i = 0 |(λ = 0)2 i


S |λ = 1i = |λ = 1i

S |(λ = 0)1 i = |(λ = 0)1 i

S |(λ = 0)2 i = − |(λ = 0)2 i


Since all the eigenvectors have different pairs of eigenvalues, L2z and S form a CSCO.
Chapter 3

The Postulates of Quantum

Mechanics

3.1 Wave Function

In a one-dimensional problem, consider a particle whose wave function is:
exp (ip0 x/~)
ψ(x) = N √
x2 + a2
where a and p0 are real constants and N is a normalization coefficient.

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

To normalize the wavefunction,
Z ∞
ψ ∗ (x)ψ(x) dx = 1
−∞
Z ∞
1
N2 dx = 1
−∞ x2 + a2
Looking up this integral,
a
N2 =
π

3.1.b The position of the particle is measured. What is the probability

√ √
of finding a result between −a/ 3 and +a/ 3?
Using (C-1),
√
Z a/ 3

√
ψ ∗ (x)ψ(x) dx
−a/ 3

37
38 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS
    
a 1 1
= arctan √ − arctan − √
π 3 3

1
P =
3

3.1.c Calculate the mean value of the momentum of a particle which has
ψ(x) for its wave function.
From (C-4),

hpi = hψ|p|ψi

The momentum operator,

∂
p = −i~
∂x
Let’s act the momentum operator on our wave function,
 
∂ exp(ip0 x/~)
p |ψi = −i~N √
∂x x2 + a2

 
ip0/~ exp x exp
= −i~N √ − 2
x2 + a2 (x + a2 )3/2

Z ∞  
ip0 1 x
hψ|p|ψi = −i~N 2 − 2 dx
−∞ ~ x2 + a2 (x + a2 )2

The second term dies since that is an odd function,

a ip0 a p0 a2
hpi = −i~ = 2
π ~ π π
3.2. MEASUREMENT OF A ONE-DIMENSIONAL PARTICLE 39

3.2 Measurement of a one-dimensional particle

Consider, in a one-dimensional problem, a particle of mass m whose wave function at
time t is ψ(x, t).

3.2.a At time t, the distance d of this particle from the origin is measured.
Write, as a function of ψ(x, t), the probability P(d0 ) of finding a
result greater than a given length d0 . What are the limits of P(d0 )
when d0 → 0 and d0 → ∞?
We already know how to write the probability of finding the particle within a certain range. Thus,
in order to find the probability of the particle outside of that range, we subtract the probability
of finding the particle within a certain range from the probability of finding the particle anywhere.
For a normalized wave function, the probability of finding it somewhere is unity, so
Z d0
P(d0 ) = 1 − ψ ∗ (x)ψ(x) dx
−d0

As d0 → 0, the second term goes to zero, so we certain to find the particle outside of that range.
As d0 → ∞, it becomes more difficult to find that particle outside of that range.

3.2.b Instead of performing the measurement of question a, one mea-

sures the velocity v of the particle at time t. Express, as a function
of ψ(x, t), the probability of finding a result greater than a given
value v0
The probability of finding the particle between −p0 and p0 , which are the momenta corresponding
to v0 ,
Z p0
∂
−i~ ψ ∗ (x) ψ(x) dx
−p0 ∂x

Since p = mv, we can easily convert to velocity,

i~ p0 ∗
Z
∂
P(v0 ) = − ψ (x) ψ(x) dx
m −p0 ∂x
40 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

3.3 Free Particle

The wave function of a free particle, in a one-dimensional problem, is given at time
t = 0 by:
Z ∞
ψ(x, 0) = N exp(−|k|/k0 ) exp(ikx) dk
∞

where k0 and N are constants.

3.3.a What is the probability P(p1 , 0) that a measurement of the momen-

tum, performed at time t = 0, will yield a result included between
−p1 and +p1
We start by evaluating the integral,
Z 0 Z ∞ 
ψ(x, 0) = N exp(k/k0 ) exp(ikx) dk + exp(−k/k0 ) exp(ikx) dk
−∞ 0

2N k0
ψ(x, 0) =
1 + k02 x2
Solving for the normalization constant,
Z ∞
2 2 1
4N k0 dx = 1
−∞ (1 + k02 x2 )2

π
4N 2 k02 =1
2k0

r
1
N=
2πk0

r
2k0 1
ψ(x, 0) =
π 1 + k02 x2
To find the probability,
Z p1
P(p1 , 0) = |ψ̄(p, 0)|2 dp
−p1

Where ψ̄ is the Fourier transform,

Z ∞
1
ψ̄(p, 0) = √ ψ(x, 0) exp(−ipx/~) dx
2π~ −∞
3.3. FREE PARTICLE 41

If we compare to the original equation, we see that exp(ikx) cancels out,

1
ψ̄(p, 0) = √ exp(−|p|/p0 )
p0

Z p1  
1 2|p|
P(p1 , 0) = exp − dp
p0 −p1 p0

 
2p1
P(p1 , 0) = 1 − exp −
p0

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

performed at time t? Interpret.
The wave function is determined by the time evolution operator,

ψ(x, t) = U (t)ψ(x, 0)

iP 2 t
 
U (t) = exp −
2m~
Transforming to momentum space is fairly straightforward, P → p,
ip2 t
 
ψ̄(p, t) = exp − ψ̄(p, 0)
2m~
To find the probability,
Z p1
P(p1 , t) = |ψ̄(p, t)|2 dp
−p1

Looking at this, we see

Z p1
= |ψ̄(p, 0)|2 dp
−p1

Which is the same answer as we got in part (a). The probability is time-independent, which
means the energy eigenstates are stationary states.

3.3.c What is the form of the wave packet at time t = 0? Calculate for
this time the product ∆X · ∆P ; what is your conclusion? Describe
qualitatively the subsequent evolution of the wave packet.
As we have found,
r
2k0 1
ψ(x, 0) =
π 1 + k02 x2
42 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

The uncertainty,
(
2
∆X = [hX 2 i − hXi ]1/2 ;
2
∆P = [hP 2 i − hP i ]1/2

Z ∞
hXi = x|ψ(x, 0)|2 dx = 0
−∞

Z ∞
2 1
hX i = x2 |ψ(x, 0)|2 dx =
−∞ k02

Z ∞
hP i = p|ψ̄(p, 0)|2 dp = 0
−∞

∞
p20
Z
hP 2 i = p2 |ψ̄(p, 0)|2 dp =
−∞ 2

 1
∆X = ;
k0




∆P = ~k
√0


2

~
∆X · ∆P = √
2
which follows Heisenberg Uncertainty Principle.
3.4. SPREADING OF A FREE WAVE PACKET 43

3.4 Spreading of a Free Wave Packet

Consider a free particle

3.4.a Show, applying Ehrenfest’s Theorem, that hXi is a linear function

of time, the mean value hP i remaining constant
We apply (D-34) and (D-35),
d 1
hXi = hP i
dt m

d
hP i = − hV 0 (x)i
dt
For a free particle, the potential is constant, so we can see that hP i remains constant. Since
hP i has a constant value, hXi must be linear in time in order to satisfy the above equations.

3.4.b Write the equations of motion for the mean values hX 2 i and hXP + P Xi.
Integrate these equations.
In general, Ehrenfest’s theorem is given by (D-27),
d 1
hAi = h[A, H (t)]i
dt i~
For a free particle,
P2
H =
2m
Let’s start with the second mean value,

[XP + P X, P 2 ] = 4i~P 2

d 2
hXP + P Xi = hP 2 i = 0
dt m
We can see that hP 2 i = 0 since if we plug this into Ehrenfest’s theorem, we end up commuting
2
P with itself.
For hX 2 i,

[X 2 , P 2 ] = 2i~(XP + P X)

d 1
hX 2 i = hXP + P Xi
dt m
We can see that hXP + P Xi is constant in time while hX 2 i varies linearly with time.
44 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

3.4.c Show that with a suitable choice of time origin, the root-mean-
square deviation ∆X is given by:

1
(∆X)2 = (∆P )20 t2 + (∆X)20
m2
where (∆X)0 and (∆P )0 are the root-mean-square deviations of the initial time.
Using the definition of root mean square deviation,
2
(∆X)2 = hX 2 i − hXi

The first term gives

hX 2 i = (∆X)20

The second term gives

1
hXi = hP i t
m
Since hP i is constant in time, hP i = ∆P ,
1
(∆X)2 = (∆P )20 t2 + (∆X)20
m2
3.5. PARTICLE SUBJECT TO A CONSTANT POTENTIAL 45

3.5 Particle Subject to a Constant Potential

In a one-dimensional problem, consider a particle of potential energy V (X) = −f X,
where f is a positive constant [V (X) arises, for example, from a gravity field or a
uniform electric field].

3.5.a Write Ehrenfest’s theorem for the mean values of the position X
and the momentum P of the particle. Integrate these equations;
compare with the classical motion.
As in the previous problem, we turn to (D-27),

d 1
hAi = h[A, H (t)]i
dt i~
This time, because we are in a potential, the Hamiltonian is given by

P2
H (t) = − fX
2m
Let’s calculate some commutation relations,

P2
 
1
X, − fX = [X, P 2 ] − f [X, X]
2m 2m

P2
 
1
P, − fX = [P, P 2 ] − f [P, X]
2m 2m

The first term in each of these we recognize from the equations of motion for a free particle, so
we can use (D-34) and (D-35).

d 1
hXi = hP i
dt m

d
hP i = f
dt
Integrating hP i,

hP i = f t + hP i0

Substituting this in,

1 1
hXi = ( /2f t2 + hP i0 + hXi0 )
m
46 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

3.5.b Show that the root-mean-square deviation of ∆P does not vary

over time.
Using the definition,
2
(∆P )2 = hP 2 i − hP i

We need to calculate hP 2 i, which involves calculating the commutation relation

2
 
2 P
P , − f X = f [X, P 2 ]
2m

d
hP 2 i = 2f hP i
dt
Substituting in hP i that was determined in the last part,

d
hP 2 i = 2f (f t + hP i0 )
dt

hP 2 i = f 2 t2 + 2f t hP i0

2 2
hP 2 i − hP i = f 2 t2 + 2f t hP i0 − (f 2 t2 + 2f t hP i0 + hP i0 )

2
(∆P )2 = − hP i0

3.5.c Write the Schrodinger equation in the {|pi} representation. Deduce

∂ ∂
from it a relation between | hp|ψ(t)i |2 and | hp|ψ(t)i |2 . Integrate
∂t ∂p
the equation thus obtained; give a physical interpretation.
The Schrodinger equation,
∂
i~ |ψ(t)i = H |ψ(t)i
∂t

∂ P2 ∂
i~ |ψ(t)i = |ψ(t)i − i~f |ψ(t)i
∂t 2m ∂p

Multiplying by hp|,

∂ p2 ∂
i~ hp|ψ(t)i = hp|ψ(t)i − i~f hp|ψ(t)i
∂t 2m ∂p
3.6. THREE-DIMENSIONAL WAVE FUNCTION 47

3.6 Three-dimensional wave function

Consider the three-dimensional wave function
  
|x| |y| |z|
ψ(x, y, z) = N exp − + +
2a 2b 2c
where a, b, and c are three positive lengths.

3.6.a Calculate the constant N which normalizes ψ.

To normalize,
∞ ∞ ∞   
|x| |y| |z|
Z Z Z
N2 exp − + + dxdydz = 1
−∞ −∞ −∞ a b c
Let’s look at one of these integrals,
Z ∞   Z 0 Z ∞
|x|
exp − dx = x
exp( /a) dx + exp(−x/a) dx = 2a
−∞ a −∞ 0

Extrapolating this out, our integral gives

N 2 (8abc) = 1

1
N2 =
8abc

3.6.b Calculate the probability that a measurement of X will yield a

result included between 0 and a.
We can simplify this a little. Since we can have y and z values anywhere, we only need to evaluate
Z a
2
4N bc exp(−|x|/a) dx
0

a
1 − e−1
Z
1
= exp(−x/a) dx =
2a 0 2

3.6.c Calculate the probability that simultaneous measurements of Y

and Z will yield results included respectively between −b and b,
and −c and c.
The integral we need to evaluate simplifies to
Z bZ c    
1 |y| |z|
exp − exp − dydz
4bc −b −c b c
48 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

Let’s evaluate one of these integrals,

Z b   Z b Z b
|y|
exp − dy = exp(y/b) dy + exp(−y/b) dy
−b b 0 0

 
1
= 2b 1 −
e

As expected, we get something twice as large as the integral in the previous part.
 2
1
P= 1−
e

3.6.d Calculate the probability that a measurement of the momentum

will yield a result included in the element dpx dpy dpz centered at the
point px = py = 0; pz = ~/c
The first thing to do is transform to momentum space,
Z ∞Z ∞Z ∞      
N ipx x ipy y ipz z
ψ̄(px , py , pz ) = ψ(x, y, z) exp − exp − exp − dxdydz
(2π~)3/2 −∞ −∞ −∞ ~ ~ ~

Let’s evaluate part of this first,

∞  
|x| ipx x
Z
exp − − dx
−∞ 2a ~

0 Z ∞
4a~2
Z      
1 ip 1 ip
= exp x − dx + exp −x + dx = 2
−∞ 2a ~ 0 2a ~ ~ + 4a2 p2

Extrapolating,

N 64abc~6
ψ(px , py , pz ) =
(2π~) (~ + 4a px )(~2 + 4b2 p2y )(~2 + 4c2 p2z )
3/2 2 2 2

For an element dpx dpy dpz centered at the required point, we set px = py = 0 and pz = ~/c,

8
dP =
5(2π~)3/2
3.7. MEASUREMENTS 49

3.7 Measurements
Let ψ(x, y, z) = ψ(~r) be the normalized wave function of a particle. Express in terms of
ψ(~r) the probability for:

3.7.a a measurement of the abscissa X, to yield a result included between

x1 and x2
Z ∞ Z ∞ Z x2
P= ψ ∗ (~r)ψ(~r) dxdydz
−∞ −∞ x1

3.7.b a measurement of the component Px of the momentum, to yield a

result between p~1 and p~2
The first thing to do is Fourier transform to momentum space,
Z ∞Z ∞Z ∞      
1 ipx x ipy y ipz z
ψ̄(~
p) = ψ(~r) exp − exp − exp − dxdydz
(2π~)3/2 −∞ −∞ −∞ ~ ~ ~
Z ∞ Z ∞ Z p2
P= ψ̄ ∗ (~
p)ψ̄(~
p) dpx dpy dpz
−∞ −∞ p1

3.7.c simultaneous measurements of X and Pz to yield:

x1 ≤ x ≤ x2

pz ≥ 0
We can perform this simultaneous measurement because X and Pz commute (note that X and
Px would not). Let’s first perform the measurement of X,
Z ∞ Z ∞ Z x2
Px = ψ ∗ (~r)ψ(~r) dxdydz
−∞ −∞ x1

We now want to Fourier transform to momentum space,

Z ∞ Z ∞ Z x2      
1 ipx x ipy y ipz z
ψ̄(~
p) = ψ(~r) exp − exp − exp − dxdydz
(2π~)3/2 −∞ −∞ x1 ~ ~ ~
Note the different integration bounds because we have made the measurement on X.
Z ∞Z ∞ Z ∞
Pp = ψ̄ ∗ (~
p)ψ̄(~
p) dpx dpy dpz
0 −∞ −∞

P = Px Pp
50 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

3.8 Probability current

3.9. PROBABILITY CURRENT 51

3.9 Probability current

52 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

3.10 Virial Theorem

3.10.a In a one-dimensional problem, consider a particle with the Hamil-

tonian:
P2
H = + V (X)
2m
where
V (X) = λX n
Calculate the commutator [H , XP ]. If there exists one or several stationary states |φi
in the potential V , show that the mean values hT i and hV i of the kinetic and potential
energies in these states satisfy the relation: 2 hT i = n hV i
Let’s split this up into two parts,
1
T = [P 2 , XP ]
2m
i~ 2
T =− P
m
V = λ[X n , XP ]
To solve this, let’s look at the first couple iterations,
= λ([X n , X]P + X[X n , P ]) = λX[X n , P ]

= λX(X[X n−1 , P ] + i~X n−1 )

Continuing, we notice a pattern,
[X n , P ] = i~nX n−1

V = i~nλX n

i~ 2
P + i~nλX n
[H , XP ] = −
m
Using Ehrenfest’s Theorem, and expecting hXP i to be 0 for a stationary state, we can rewrite
this in terms of hT i and hV i.
P2
hT i =
2m
hV i = λX n

0 = −2i~ hT i + i~n hV i

2 hT i = n hV i
3.11. TWO-PARTICLE WAVE FUNCTION 53

3.11 Two-particle wave function

In a one-dimensional problem, consider a system of two particles (1) and (2) with
which is associated the wave function ψ(x1 , x2 ).

3.11.a What is the probability of finding, in a measurement of the posi-

tions X1 and X2 of the two particles, a result such that:

x ≤ x1 ≤ x + dx

α ≤ x2 ≤ β

Z β Z x+dx
P= ψ ∗ (x1 , x2 )ψ(x1 , x2 ) dx1 dx2
α x
54 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

3.12 Infinite one-dimensional well

3.13. INFINITE TWO-DIMENSIONAL WELL 55

3.13 Infinite two-dimensional well

56 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

3.14 Matrices
Consider a physical system whose state space, which is three-dimensional, is spanned
by the orthonormal basis formed by the three kets |u1 i, |u2 i, |u3 i. In this basis, the
Hamiltonian operator H of the system and the two observables A and B are written:
     
1 0 0 1 0 0 0 1 0
H = ~ω0 0 2 0 ; A = a 0 0 1 ; B = b 1 0 0
0 0 2 0 1 0 0 0 1
where ω0 , a, and b are positive real constants.
The physical system at time t = 0 is in the state:
1 1 1
|ψ(0)i = √ |u1 i + |u2 i + |u3 i
2 2 2

3.14.a At time t = 0, the energy of the system is measured. What values

can be found, and with what probabilities? Calculate, for the
system in the state |ψ(0)i, the mean value hH i and the root-
mean-square deviation ∆H .
The possible energy values can be found by finding the eigenvalues of H ,

H |ψi = E |ψi

The eigenvalues are

λ = ~ω0 , 2~ω0 , 2~ω0

The corresponding eigenvectors are

|1i = |u1 i ; |2i = |u2 i ; |20 i = |u3 i

1
P(E = ~ω0 ) = | h1|ψ(0)i |2 =
2

1
P(E = 2~ω0 ) = | h2|ψ(0)i |2 + | h3|ψ(0)i |2 =
2
The mean value,

hH i = hψ(0)|H |ψ0i

Let’s start by writing |ψ(0)i in matrix form,

 √ 
1/ 2

|ψ(0)i =  1/2 
1/2
3.14. MATRICES 57
  √ 
1 0 0 1/ 2
√

hH i = ~ω0 1/ 2 1/2 1/2 0 2 0  1/2 
0 0 2 1/2

3~ω0
hH i =
2
To find the root-mean-square-deviation,
2
∆H = hH 2 i − hH i
We need to calculate,
  √ 
1 0 0 1/ 2
√ 5~2 ω02
hH 2 i = ~2 ω 2


1/ 2 1/2 1/2 0 4 0  1/2  =
1/2 2
0 0 4

10~2 ω02 9~2 ω02 ~2 ω02

∆H = − =
4 4 4

3.14.b Instead of measuring H at time t = 0, one measures A; what

results can be found, and with what probabilities? What is the
state vector immediately after the measurement?
The eigenvalues of A are
λ = a, a, −a
The corresponding eigenvectors,
1 1
|1i = √ (|u2 i + |u3 i); |10 i = |u1 i ; |−1i = √ (|u2 i − |u3 i)
2 2
After the measurement, the state vector will be in A |ψ(0)i,
 √
a/ 2

A |ψ(0)i = a/2 

a/2

3.14.c Calculate the state vector |ψ(t)i of the system at time t

The time evolution operator gives,
 
iH t
|ψ(t)i = exp − |ψ(0)i
~
To first order,
it
H |ψ(0)i
|ψ(t)i = |ψ(0)i −
~
   
1 1 1
|ψ(t)i = √ (1 − iω0 t) |u1 i + − iω0 t |u2 i + − iω0 t |u3 i
2 2 2
58 CHAPTER 3. THE POSTULATES OF QUANTUM MECHANICS

3.14.d Calculate the mean values hAi (t) and hBi (t) of A and B at time t.
What comments can be made?
  √ 
1/ 2(1 − iω t)
√
1 0 0 0
hAi (t) = a 1/ 2(1 + iω0 t) 1/2 + iω0 t 1/2 + iω0 t 0 0 1  1/2 − iω0 t 
0 1 0 1/2 − iω t
0

5ω 2 t2
 
hAi (t) = a 1 + 0
2

  √ 
0 1 0 1/ 2(1 − iω t)
0
√

hBi (t) = b hAi (t) = a 1/ 2(1 + iω0 t) 1/2 + iω0 t 1/2 + iω t 1
0 0 0  1/2 − iω0 t 
0 0 1 1/2 − iω t
0

 
1 1 2
hBi (t) = √ + + 1 + √ ω02 t2
2 4 2
As time goes to infinity, the mean values of A and B increase to infinity.