Department of Physics Part II Experimental and

University of Cambridge Theoretical Physics

Theoretical Physics 2
Lecture Notes and Examples

B R Webber and N R Cooper

Lent Term 2004

Preface

In this course, we cover the necessary mathematical tools that underpin modern theoretical physics.
We examine topics in quantum mechanics (with which you have some familiarity from previous
courses) and apply the mathematical tools learnt in the IB Mathematics course (complex anal-
ysis, differential equations, matrix methods, special functions etc.) to topics like perturbation
theory, scattering theory, etc. A course outline is provided below. Items indicated by a * are non-
examinable material. They are there to illustrate the application of the course material to topics
that you will come across in the PartII/Part III Theoretical Physics options. While we have tried
to make the notes as self-contained as possible, you are encouraged to read the relevant sections
of the recommended texts listed below. Throughout the notes, there are “mathematical interlude”
sections reminding you of the the maths you are supposed to have mastered in the IB course. The
“worked examples” are used to illustrate the concepts and you are strongly encouraged to work
through every step, to ensure that you master these concepts and the mathematical techniques.

We are most grateful to Dr Guna Rajagopal for preparing the lecture notes of which these are an
updated version.

Course Outline

• Operator Methods in Quantum Mechanics (2 lectures): Mathematical foundations of

non-relativistic quantum mechanics; vector spaces; operator methods for discrete and contin-
uous eigenspectra; generalized form of the uncertainty principle; simple harmonic oscillator;
delta-function potential; introduction to second quantization.

• Angular Momentum (2 lectures): Eigenvalues/eigenvectors of the angular momentum

operators (orbital/spin); spherical harmonics and their applications; Pauli matrices and
spinors; addition of angular momenta.

• Approximation Methods for Bound States (2 lectures): Variational methods and their
application to problems of interest; perturbation theory (time-independent and time depen-

i
 dent) including degenerate and non-degenerate cases; the JWKB method and its application
to barrier penetration and radioactive decay.

• Scattering Theory (2 lectures): Scattering amplitudes and differential cross-section; par-

tial wave analysis; the optical theorem; Green functions; weak scattering and the Born ap-
proximation; *relation between Born approximation and partial wave expansions; *beyond
the Born approximation.

• Identical Particles in Quantum Mechanics (2 lectures): Wave functions for non-

interacting systems; symmetry of many-particle wave functions; the Pauli exclusion principle;
fermions and bosons; exchange forces; the hydrogen molecule; scattering of identical parti-
cles; *second quantization method for many-particle systems; *pair correlation functions for
bosons and fermions;

• Density Matrices (2 lectures): Pure and mixed states; the density operator and its
properties; position and momentum representation of the density operator; applications in
statistical mechanics.

Problem Sets
The problem sets (integrated within the lecture notes) are a vital and integral part of the course.
The problems have been designed to reinforce key concepts and mathematical skills that you will
need to master if you are serious about doing theoretical physics. Many of them will involve signif-
icant algebraic manipulations and it is vital that you gain the ability to do these long calculations
without making careless mistakes! They come with helpful hints to guide you to their solution.
Problems that you may choose to skip on a first reading are indicated by †.

Books
There is no single book that covers all of material in this course to the conceptual level or mathe-
matical rigour required. Below are some books that come close. Liboff is at the right level for this
course and it is particularly strong on applications. Sakurai is more demanding mathematically
although he makes a lot of effort to explain the concepts clearly. This book is a recommended text
in many graduate schools. Reed and Simon show what is involved in a mathematically rigorous
treatment.

At about the level of the course: Liboff, Quantum Mechanics, 3rd Ed., Addison-Wesley.

At a more advanced level: Sakurai, Quantum Mechanics, 2nd Ed., Addison-Wesley;

Reed and Simon, Methods of Modern Mathematical Physics, Academic Press.
Contents

1 Operator Methods In Quantum Mechanics 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1.1 Mathematical foundations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.1.2 Hilbert space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.1.3 The Schwartz inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.1.4 Some properties of vectors in a Hilbert space . . . . . . . . . . . . . . . . . . 5

1.1.5 Orthonormal systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.1.6 Operators on Hilbert space . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.1.7 Eigenvectors and eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

1.1.8 Observables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

1.1.9 Generalised uncertainty principle . . . . . . . . . . . . . . . . . . . . . . . . . 16

1.1.10 Basis transformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

1.1.11 Matrix representation of operators . . . . . . . . . . . . . . . . . . . . . . . . 19

1.1.12 Mathematical interlude: Dirac delta function . . . . . . . . . . . . . . . . . . 20

1.1.13 Operators with continuous or mixed (discrete-continuous) spectra . . . . . . 21

1.2 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

1.2.1 Harmonic oscillator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

1.2.2 Delta-function potential well . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

iii
 1.3 Introduction to second quantisation . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

1.3.1 Vibrating string . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

1.3.2 Quantisation of vibrating string . . . . . . . . . . . . . . . . . . . . . . . . . . 37

1.3.3 General second quantisation procedure . . . . . . . . . . . . . . . . . . . . . . 38

2 Angular Momentum 41

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

2.2 Orbital angular momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

2.2.1 Eigenvalues of orbital angular momentum . . . . . . . . . . . . . . . . . . . . 47

2.2.2 Eigenfunctions of orbital angular momentum . . . . . . . . . . . . . . . . . . 50

2.2.3 Mathematical interlude: Legendre polynomials and spherical harmonics . . . 53

2.2.4 Angular momentum and rotational invariance . . . . . . . . . . . . . . . . . . 56

2.3 Spin angular momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

2.3.1 Spinors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

2.4 Addition of angular momenta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

2.4.1 Addition of spin- 12 operators . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

2.4.2 Addition of spin- 12 and orbital angular momentum . . . . . . . . . . . . . . . 67

2.4.3 General case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

3 Approximation Methods For Bound States 71

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

3.2 Variational methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

3.2.1 Variational theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

3.2.2 Interlude : atomic units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

3.2.3 Hydrogen molecular ion, H+

2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
 3.2.4 Generalisation: Ritz theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

3.2.5 Linear variation functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

3.3 Perturbation methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

3.3.1 Time-independent perturbation theory . . . . . . . . . . . . . . . . . . . . . . 83

3.3.2 Time-dependent perturbation theory . . . . . . . . . . . . . . . . . . . . . . . 90

3.4 JWKB method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

3.4.1 Derivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

3.4.2 Connection formulae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

3.4.3 *JWKB treatment of the bound state problem . . . . . . . . . . . . . . . . . 101

3.4.4 Barrier penetration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

3.4.5 Alpha decay of nuclei . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

4 Scattering Theory 109

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

4.2 Spherically symmetric square well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

4.3 Mathematical interlude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

4.3.1 Brief review of complex analysis . . . . . . . . . . . . . . . . . . . . . . . . . 111

4.3.2 Properties of spherical Bessel/Neumann functions . . . . . . . . . . . . . . . 113

4.3.3 Expansion of plane waves in spherical harmonics . . . . . . . . . . . . . . . . 115

4.4 The quantum mechanical scattering problem . . . . . . . . . . . . . . . . . . . . . . 116

4.4.1 Born approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

4.5 *Formal time-independent scattering theory . . . . . . . . . . . . . . . . . . . . . . . 126

4.5.1 *Lippmann-Schwinger equation in the position representation . . . . . . . . . 127

4.5.2 *Born again! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

5 Identical Particles in Quantum Mechanics 131

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

5.2 Multi-particle systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

5.2.1 Pauli exclusion principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

5.2.2 Representation of Ψ(1, 2, . . . , N ) . . . . . . . . . . . . . . . . . . . . . . . . . 134

5.2.3 Neglecting the symmetry of the many-body wave function . . . . . . . . . . . 134

5.3 Fermions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

5.4 Bosons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

5.5 Exchange forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

5.6 Helium atom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

5.6.1 Ground state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

5.7 Hydrogen molecule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

5.8 Scattering of identical particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

5.8.1 Scattering of identical spin zero bosons . . . . . . . . . . . . . . . . . . . . . . 149

5.8.2 Scattering of fermions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

5.9 *Modern electronic structure theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

5.9.1 *The many-electron problem . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

5.9.2 *One-electron methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

5.9.3 *Hartree approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

5.9.4 *Hartree-Fock approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

5.9.5 *Density functional methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

5.9.6 *Shortcomings of the mean-field approach . . . . . . . . . . . . . . . . . . . . 156

5.9.7 *Quantum Monte Carlo methods . . . . . . . . . . . . . . . . . . . . . . . . . 157

6 Density Operators 159

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

6.2 Pure and mixed states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

6.3 Properties of the Density Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

6.3.1 Density operator for spin states . . . . . . . . . . . . . . . . . . . . . . . . . . 164

6.3.2 Density operator in the position representation . . . . . . . . . . . . . . . . . 166

6.4 Density operator in statistical mechanics . . . . . . . . . . . . . . . . . . . . . . . . . 168

6.4.1 Density operator for a free particle in the momentum representation . . . . . 170

6.4.2 Density operator for a free particle in the position representation . . . . . . . 171

6.4.3 *Density matrix for the harmonic oscillator . . . . . . . . . . . . . . . . . . . 172

Chapter 1

Operator Methods In Quantum

Mechanics

1.1 Introduction

The purpose of the first two lectures is twofold. First, to review the mathematical formalism of
elementary non-relativistic quantum mechanics, especially the terminology. The second purpose is
to present the basic tools of operator methods, commutation relations, shift operators, etc. and
apply them to familiar problems such as the harmonic oscillator. Before we get down to the operator
formalism, let’s remind ourselves of the fundamental postulates of quantum mechanics as covered
in earlier courses. They are:

• Postulate 1: The state of a quantum-mechanical system is completely specified by a function

Ψ(r, t) (which in general can be complex) that depends on the coordinates of the particles
(collectively denoted by r) and on the time. This function, called the wave function or the
state function, has the important property that Ψ∗ (r, t)Ψ(r, t) dr is the probability that the
system will be found in the volume element dr, located at r, at the time t.

• Postulate 2: To every observable A in classical mechanics, there corresponds a linear Her-

mitian operator  in quantum mechanics.

• Postulate 3: In any measurement of the observable A, the only values that can be obtained
are the eigenvalues {a} of the associated operator Â, which satisfy the eigenvalue equation

ÂΨa = aΨa

1
2 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

where Ψa is the eigenfunction of  corresponding to the eigenvalue a.

• Postulate 4: If a system is in a state described by a normalised wavefunction Ψ, and the

eigenfunctions {Ψa } of  are also normalised, then the probability of obtaining the value a
in a measurement of the observable A is given by
Z ∞
2
Ψ∗a Ψ dr
 
P (a) = 
−∞
(Recall that a function Φ(r) such that
Z ∞
Φ∗ Φ dr = 1
−∞
is said to be normalised.)

• Postulate 5: As a result of a measurement of the observable A in which the value a is

obtained, the wave function of the system becomes the corresponding eigenfunction Ψ a . (This
is sometimes called the collapse of the wave function.)

• Postulate 6: Between measurements, the wave function evolves in time according to the
time-dependent Schrödinger equation
∂Ψ i
= − ĤΨ
∂t h̄
where Ĥ is the Hamiltonian operator of the system.

The justification for the above postulates ultimately rests with experiment. Just as in geometry one
sets up axioms and then logically deduces the consequences, one does the same with the postulates
of QM. To date, there has been no contradiction between experimental results and the outcomes
predicted by applying the above postulates to a wide variety of systems.

We now explore the mathematical structure underpinning quantum mechanics.

1.1.1 Mathematical foundations

In the standard formulation of quantum theory, the state of a physical system is described by a
vector in a Hilbert space H over the complex numbers. The observables and dynamical variables
of the system are represented by linear operators which transform each state vector into another
(possibly the same) state vector. Throughout this course (unless stated otherwise) we will adopt
Dirac’s notation: thus a state vector is denoted by a ket |Ψi. This ket provides a complete de-
scription of the physical state. In the next section we will explore the mathematical properties of
the Hilbert space and learn why it plays such a central role in the mathematical formulation of
quantum mechanics.
1.1. INTRODUCTION 3

1.1.2 Hilbert space

A Hilbert space H,
H = {|ai, |bi, |ci, . . .}, (1.1)

is a linear vector space over the field of complex number C i.e. it is an abstract set of elements
(called vectors) with the following properties

1. ∀ |ai, |bi ∈ H we have

• |ai + |bi ∈ H (closure property)

• |ai + |bi = |bi + |ai (commutative law)

• (|ai + |bi) + |ci = |ai + (|bi) + |ci) (associative law)

• ∃ a null vector, |nulli ∈ H with the property

|ai + |nulli = |ai (1.2)

• ∀ |ai ∈ H ∃ | − ai ∈ H such that

|ai + | − ai = |nulli (1.3)

• ∀ α, β ∈ C

α(|ai + |bi) = α|ai + α|bi (1.4)

(α + β)|ai = α|ai + β|ai (1.5)

(αβ)|ai = α(β|ai) (1.6)

1|ai = |ai (1.7)

2. A scalar product is defined in H. It is denoted by (|ai, |bi) or ha|bi, yielding a complex number.
The scalar product has the following properties

(|ai, λ|bi) = λ(|ai, |bi) (1.8)

(|ai, |bi + |ci) = (|ai, |bi) + (|ai, |ci) (1.9)

(|ai, |bi) = (|bi, |ai)∗ (1.10)

4 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

The last equation can also be written as

ha|bi = hb|ai∗ (1.11)

From the above, we can deduce that

(λ|ai, |bi) = λ∗ (|ai, |bi) (1.12)

= λ∗ ha|bi (1.13)

and

(|a1 i + |a2 i, |bi) = (|a1 i, |bi) + (|a2 i, |bi) (1.14)

= ha1 |bi + ha2 |bi (1.15)

The norm of a vector is defined by q

kak = ha|ai (1.16)

and corresponds to the “length” of a vector. Note that the norm of a vector is a real number ≥ 0.
(This follows from (1.11)).

1.1.3 The Schwartz inequality

Given any |ai, |bi ∈ H we have

kak kbk ≥ |ha|bi| (1.17)

with the equality only being valid for the case

|ai = λ|bi (1.18)

(with λ a complex number) i.e. when one vector is proportional to the other.

Proof

Define a |ci such that

|ci = |ai + λ|bi (1.19)

where λ is an arbitrary complex number. Whatever λ may be:

hc|ci = ha|ai + λha|bi + λ∗ hb|ai + λλ∗ hb|bi (1.20)

≥ 0 (1.21)
1.1. INTRODUCTION 5

Choose for λ the value

hb|ai
λ=− (1.22)
hb|bi
and substitute into the above equation, which reduces to
ha|bihb|ai
ha|ai − ≥0 (1.23)
hb|bi
Since hb|bi is positive, multiply the above inequality by hb|bi to get

ha|aihb|bi ≥ ha|bihb|ai (1.24)

≥ |ha|bi|2 (1.25)

and finally taking square roots and using the definition of the norm we get the required result.
(This result will be used when we prove the generalised uncertainty principle).

1.1.4 Some properties of vectors in a Hilbert space

∀ |ai ∈ H, a sequence {|an i} of vectors exists, with the property that for every  > 0, there exists
at least one vector |an i of the sequence with

k|ai − |an ik ≤  (1.26)

A sequence with this property is called compact.

The Hilbert space is complete i.e. every |ai ∈ H can be arbitrarily closely approximated by a
sequence {|an i}, in the sense that
lim k|ai − |an ik = 0 (1.27)
n→∞

Then the sequence {|an i} has a unique limiting value |ai.

The above properties are necessary for vector spaces of infinite dimension that occur in QM.

1.1.5 Orthonormal systems

Orthogonality of vectors. |ai, |bi ∈ H are said to be orthogonal if

ha|bi = 0 (1.28)

Orthonormal system. The set {|an i} of vectors is an orthonormal system if the vectors are
orthogonal and normalised, i.e.
han |am i = δn,m (1.29)
6 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

where 
 1 m=n
δn,m =
 0 m 6= n

Complete orthonormal system. The orthonormal system {|an i} is complete in H if an arbi-

trary vector |ai ∈ H can be expressed as
X
|ai = αn |an i (1.30)
n

where in general αn are complex numbers whose values are

αm = ham |ai (1.31)

Proof
X 
ham |ai = ham | αn |an i
n

X
= αn ham |an i
n

X
= αn δm,n
n

= αm (1.32)

Thus we can write

X
|ai = |an ihan |ai (1.33)
n
Note that this implies
Iˆ =
X
|an ihan | (1.34)
n
called the “resolution of the identity operator” or the closure relation. The complex
numbers αn are called the an −representation of |ai, i.e. they are the components of the vector
|ai in the basis {|an i}.

1.1.6 Operators on Hilbert space

A linear operator  induces a mapping of H onto itself or onto a subspace of H. (What this means
is that if  acts on some arbitrary vector ∈ H the result is another vector ∈ H or in some subset
of H. Hence
Â(α|ai + β|bi) = αÂ|ai + β Â|bi (1.35)
1.1. INTRODUCTION 7

The operator  is bounded if

kÂ|aik ≤ Ck|aik (1.36)
∀ |ai ∈ H, and C is a real positive constant (< ∞).

Bounded linear operators are continuous, i.e. if

|an i → |ai (1.37)

then it follows that

Â|an i → Â|ai (1.38)

Two operators  and B̂ are equal ( = B̂) if, ∀|ai ∈ H,

Â|ai = B̂|ai (1.39)

The following definitions are valid ∀ |ai ∈ H:

Unit operator, Iˆ
ˆ = |ai
I|ai (1.40)

Zero operator, 0̂
0̂|ai = |nulli (1.41)

Sum operator, Â + B̂
(Â + B̂)|ai = Â|ai + B̂|ai (1.42)

Product operator, ÂB̂

(ÂB̂)|ai = Â(B̂|ai) (1.43)

Adjoint operator, † : Given Â, an adjoint operator, † , exists if ∀ |ai, |bi ∈ H

(|bi, Â|ai) = († |bi, |ai) (1.44)

or
∗
hb|Â|ai = ha|† |bi (1.45)
The adjoint of an operator has the following properties:

(αÂ)† = α∗ † (1.46)

( + B̂)† = † + B̂ † (1.47)

(ÂB̂)† = B̂ † † (1.48)

(† )† =  (1.49)
8 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

If  is Hermitian, then

 = †

hb|Â|bi = hb|† |bi

∗
= hb|† |bi

∗
= hb|Â|bi

= real (1.50)

Unitary operator, U : The operator Û is called unitary if

Û Û † = Û † Û = Iˆ (1.51)

Projection operator, |aiha| : Given any normalised vector |ai, a projection operator P̂ can be
defined as the operator that projects any vector into its component along |ai

P̂ |bi = ha|bi|ai = |aiha|bi (1.52)

We write this symbolically as

P̂ = |aiha| (1.53)

Note that a projection operator is idempotent: its square (or any power) is equal to itself

P̂ 2 = |aiha|aiha| = |aiha| (1.54)

since |ai is normalised. Note that the resolution of the identity (1.34) is a sum of projection
operators.

Commutator, [Â, B̂]

[Â, B̂] = ÂB̂ − B̂ Â (1.55)

Note that in general

ÂB̂ 6= B̂ Â (1.56)

Properties of commutators:
[Â, B̂] = −[B̂, Â] (1.57)

[Â, (B̂ + Ĉ)] = [Â, B̂] + [Â, Ĉ] (1.58)

[Â, B̂ Ĉ] = [Â, B̂]Ĉ + B̂[Â, Ĉ] (1.59)

[Â, [B̂, Ĉ]] + [B̂, [Ĉ, Â]] + [Ĉ, [Â, B̂]] = 0̂ (1.60)
†
[Â, B̂] = [B̂ † , † ] (1.61)
1.1. INTRODUCTION 9

EXAMPLE

Suppose the operators P̂ and Q̂ satisfy the commutation relation

[P̂ , Q̂] = aIˆ

where a is a constant (real) number.

• Reduce the commutator [P̂ , Q̂n ] to its simplest possible form.

Answer: Let
R̂n = [P̂ , Q̂n ] n = 1, 2, · · ·

Then R̂1 = [P̂ , Q̂] = aIˆ and

R̂n+1 = [P̂ , Q̂n+1 ] = [P̂ , Q̂n Q̂] = [P̂ , Q̂n ]Q̂ + Q̂n [P̂ , Q̂]

(We have used [Â, B̂ Ĉ] = B̂[Â, Ĉ] + [Â, B̂]Ĉ). Therefore,

ˆ = R̂n Q̂ + aQ̂n
R̂n+1 = R̂n Q̂ + Q̂n (aI)

which gives R̂2 = 2aQ̂, R̂3 = 3aQ̂2 etc. This implies that

R̂n = [P̂ , Q̂n ] = naQ̂n−1

Note that in general,

∂f
[P̂ , f (Q̂)] = a
∂ Q̂
• Reduce the commutator
[P̂ , eiQ̂ ]

to its simplest form.

Answer: Use results above to get

[P̂ , eiQ̂ ] = iaeiQ̂
10 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

Problem 1: Two operators, Â and B̂ satisfy the equations

 = B̂ † B̂ + 3

 = B̂ B̂ † + 1 (1.62)

• Show that  is self-adjoint

• Find the commutator [B̂ † , B̂]

Answer: −2Iˆ

• Find the commutator [Â, B̂]

Answer: −2B̂

1.1.7 Eigenvectors and eigenvalues

If
Â|ai = a|ai (1.63)

then |ai is an eigenvector of the operator  with eigenvalue a (which in general is a complex
number). The set of all eigenvalues of a operator is called its spectrum, which can take discrete
or continuous values (or both). For the case of Hermitian operators the following is true:

• The eigenvalues are real

• The eigenvectors corresponding to different eigenvalues are orthogonal i.e

Â|ai = a|ai (1.64)

Â|a0 i = a0 |a0 i (1.65)

and if a 6= a0 , then
ha|a0 i = 0 (1.66)

• In addition, the normalised eigenvectors of a bounded Hermitian operator give rise to a

countable, complete orthonormal system. The eigenvalues form a discrete spectrum.
1.1. INTRODUCTION 11

Problem 2: Prove that if Ĥ is a Hermitian operator, then its eigenvalues are real and its eigen-
vectors (corresponding to different eigenvalues) are orthogonal.

Answer: To be discussed in class.

From above, we deduce that an arbitrary |ψi ∈ H can be expanded in terms of the complete,
orthonormal eigenstates {|ai} of a Hermitian operator Â:
X
|ψi = |aiha|ψi (1.67)
a

where the infinite set of complex numbers {ha|ψi} are called the A representation of |ψi.

Problem 3: The operator Q̂ satisfies the equations

Q̂† Q̂† = 0

Q̂Q̂† + Q̂† Q̂ = Iˆ (1.68)

The Hamiltonian for the system is given by

Ĥ = αQ̂Q̂†

where α is a real constant.

• Show that Ĥ is self-adjoint

• Find an expression for Ĥ 2 in terms of Ĥ

Answer: Use the anti-commutator property of Q̂ to get Ĥ 2 = αĤ.

• Deduce the eigenvalues of Ĥ using the results obtained above.

Answer:The eigenvalues are 0 and α.

12 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

Problem 4 : Manipulating Operators

• Show that if |ai is an eigenvector of  with eigenvalue a, then it is an eigenvector of f (Â)

with eigenvalue f (a).

• Show that
(ÂB̂)† = B̂ † † (1.69)

and in general
(ÂB̂ Ĉ . . .)† = . . . Ĉ † B̂ † † (1.70)

• Show that † is Hermitian even if  is not.

• Show that if  is Hermitian, then the expectation value of Â2 are non-negative, and the
eigenvalues of Â2 are non-negative.

• Suppose there exists a linear operator  that has an eigenvector |ψi with eigenvalue a. If
there also exists an operator B̂ such that

[Â, B̂] = B̂ + 2B̂ Â2 (1.71)

then show that B̂|ψi is an eigenvector of  and find the eigenvalue.

Answer: Eigenvalue is 1 + a + 2a2 .

EXAMPLE

• (a) Suppose the operators  and B̂ commute with their commutator, i.e. [B̂, [Â, B̂]] =
[Â, [Â, B̂]] = 0. Show that [Â, B̂ n ] = nB̂ n−1 [Â, B̂] and [Ân , B̂] = nÂn−1 [Â, B̂].

Answer: To show this, consider the following steps:

[Â, B̂ n ] = ÂB̂ n − B̂ n  (1.72)

= ÂB̂ B̂ n−1 − B̂ ÂB̂ n−1 + B̂(ÂB̂)B̂ n−2 − B̂(B̂ Â)B̂ n−3 + · · · B̂ n−1 ÂB̂ − B̂ n−1 B̂ Â

= [Â, B̂]B̂ n−1 + B̂[Â, B̂]B̂ n−2 + · · · + B̂ n−1 [Â, B̂]

1.1. INTRODUCTION 13

Since B̂ commutes with [Â, B̂], we obtain

[Â, B̂ n ] = B̂ n−1 [Â, B̂] + B̂ n−1 [Â, B̂] + · · · + B̂ n−1 [Â, B̂] = nB̂ n−1 [Â, B̂]

as required. In the same way, since [Ân , B] = −[B̂, Ân ] and using the above steps, we obtain

[Ân , B̂] = nÂn−1 [Â, B̂]

as required.
0
• (b) Just as in (a), show that for any analytic function, f (x), we have [Â, f (B̂)] = [Â, B̂]f (B̂),
where f 0 (x) denotes the derivative of f (x).
Answer: We use the results from (a). Since f (x) is analytic, we can expand it in a power
series
P n.
n an x Then

an B̂ n ]
X
[Â, f (B̂)] = [Â, (1.73)
n

an [Â, B̂ n ]
X
=
n

n an B̂ n−1
X
= [Â, B̂]
n

0
= [Â, B̂]f (B̂)

1
• (c) Just as in (a), show that e eB̂ = eÂ+B̂ e 2 [Â,B̂] .
Answer: Consider an operator F̂ (s) which depends on a real parameter s:

F̂ (s) = es esB̂

Its derivative with respect to s is:

! !
dF̂ d s sB̂ d sB̂
= e e + es e (1.74)
ds ds ds

= Âes esB̂ + es B̂esB̂

= Âes esB̂ + es B̂e−s es esB̂

" #
=  + es B̂e−s F̂ (s)

Using part (a), we can write

[es , B̂] = −[B̂, es ] = −s[B̂, Â]es = s[Â, B̂]esÂ

14 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

This means that es B̂ = B̂e−s + s[Â, B̂]es and es B̂e−s = B̂ + s[Â, B̂]. Substituting this
into the equation above, we get
" #
dF̂
= Â + B̂ + s[Â, B̂] F̂ (s)
ds

Since  + B̂ and [Â, B̂] commute, we can integrate this differential equation. This yields
1 2
F̂ (s) = F̂ (0) e(Â+B̂)s+ 2 [Â,B̂]s

ˆ Finally substituting F̂ (0) and s = 1, we obtain the

Setting s = 0, we obtain F̂ (0) = I.
required result.

• (d) Prove the following identity for any two operators  and B̂:
1 1
e B̂e− = B̂ + [Â, B̂] + [Â, [Â, B̂]] + [Â, [Â, [Â, B̂]]] + · · · (1.75)
2! 3!

Answer: To show this, define

f (λ) = eλ B̂e−λÂ

where λ is a real parameter. Then,

f (0) = B̂ (1.76)

f (1) = e B̂e−Â

f 0 (λ) = eλ [Â, B̂]e−λÂ

f 0 (0) = [Â, B̂]

f 00 (λ) = eλ [Â, [Â, B̂]]e−λÂ

f 00 (0) = [Â, [Â, B̂]]

The Taylor expansion of f (λ) is given by

1 2 00
f (λ) = f (0) + λf 0 (0) + λ f (0) + · · ·
2!
This implies
1 2
eλ B̂e−λ = B̂ + λ[Â, B̂] + λ [Â, [Â, B̂]] + · · ·
2!
Now setting λ = 1, we get the required result.
1.1. INTRODUCTION 15

1.1.8 Observables

A Hermitian operator  is an observable if its eigenvectors |ψn i are a basis in the Hilbert space:
that is, if an arbitrary state vector can be written as
D
X
|ψi = |ψn ihψn |ψi (1.77)
n=1

(If D, the dimensionality of the Hilbert space is finite, then all Hermitian operators are observables;
if D is infinite, this is not necessarily so.)

In quantum mechanics, it is a postulate that every measurable physical quantity is described by

an observable and that the only possible result of the measurement of a physical quantity is one
of the eigenvalues of the corresponding observable. Immediately after an observation of  which
yields the eigenvalue an , the system is in the corresponding state |ψn i. It is also a postulate that
the probability of obtaining the result an when observing  on a system in the normalised state
|ψi, is
P (an ) = |hψn |ψi|2 (1.78)
(The probability is determined empirically by making a large number of separate observations of
Â, each observation being made on a copy of the system in the state |ψi.) The normalisation of
|ψi and the closure relation ensure that
D
X
P (an ) = 1 (1.79)
n=1
For an observable, by using the closure relation, one can deduce that
X
 = an |ψn ihψn | (1.80)
n

which is the spectral decomposition of Â.

The expectation value hÂi of an observable Â, when the state vector is |ψi, is defined as the
average value obtained in the limit of a large number of separate observations of Â, each made on
a copy of the system in the state |ψi. From equations (1.78) and (1.80), we have

an |hψn |ψi|2
X X
hÂi = an P (an ) =
n n

X
= an hψ|ψn ihψn |ψi = hψ|Â|ψi (1.81)
n

Let  and B̂ be two observables and suppose that rapid successive measurements yield the results
an and bn respectively. If immediate repetition of the observations always yields the same results
for all possible values of an and bn , then  and B̂ are compatible (or non-interfering) observables.
16 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

Problem 5: A system described by the Hamiltonian Ĥ0 has just two orthogonal energy eigenstates,
|1i and |2i with

h1|1i = 1

h1|2i = 0

h2|2i = 1 (1.82)

The two eigenstates have the same eigenvalues E0 :

Ĥ0 |ii = E0 |ii

for i = 1, 2. Suppose the Hamiltonian for the system is changed by the addition of the term V̂ ,
giving
Ĥ = Ĥ0 + V̂

The matrix elements of V̂ are

h1|V̂ |1i = 0

h1|V̂ |2i = V12

h2|V̂ |2i = 0 (1.83)

• Find the eigenvalues of Ĥ

• Find the normalised eigenstates of Ĥ in terms of |1i and |2i.

Answer: This will be done in class.

1.1.9 Generalised uncertainty principle

Suppose  and B̂ are any two non-commuting operators i.e.

[Â, B̂] = iĈ (1.84)

(where Ĉ is Hermitian). It can be shown that

1  
∆A ∆B ≥ hĈi (1.85)

2
1.1. INTRODUCTION 17

where
h i1
∆A = h(Â − hÂi)2 i 2
(1.86)

and similarly for ∆B. The expectation value is over some arbitrary state vector. This is the
generalised uncertainty principle, which implies that it is not possible for two non-commuting
observables to possess a complete set of simultaneous eigenstates. In particular if Ĉ is a non-zero
real number (times the unit operator), then  and B̂ cannot possess any simultaneous eigenstates.

Problem 6: Prove (1.85).

If the eigenvalues of  are non-degenerate, the normalised eigenvectors |ψn i are unique to within
a phase factor i.e. the kets |ψn i and eiθ |ψn i, where θ is any real number yield the same physical
results. Hence a well defined physical state can be obtained by measuring Â. If the eigenvalues
of  are degenerate we can in principle identify additional observables B̂, Ĉ, . . . which commute
with  and each other (but not functions of  or each other), until we have a set of commuting
observables for which there is no degeneracy. Then the simultaneous eigenvectors |a n , bp , cq , . . .i
are unique to within a phase factor; they are a basis for which the orthonormality relations are

han0 , bp0 , cq0 , . . . |an , bp , cq , . . .i = δn0 n δp0 p δq0 q . . . (1.87)

The observables Â, B̂, Ĉ, . . . constitute a complete set of commuting observables (CSCO).
A well defined initial state can be obtained by an observation of a CSCO.

Problem 7: Given a set of observables Â, B̂, . . . prove that any one of the following conditions
proves the other two:

• Â, B̂, . . . commute with each other,

• Â, B̂, . . . are compatible,

• Â, B̂, . . . possess a complete orthonormal set of simultaneous eigenvectors (assuming no de-
generacy).
18 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

1.1.10 Basis transformations

Suppose {|ψn i} and {|φn i} respectively are the eigenvectors of the non-commuting observables Â
and B̂ of a system. This means that we can use either {|ψn i} or {|φn i} as basis kets for the Hilbert
space. The two bases are related by the transformation

|φn i = Û |ψn i (1.88)

where
X
Û = |φi ihψi | (1.89)
i

Orthonormality of {|φn i} and the closure relation for {|ψn i} ensure that Û is a unitary operator
ˆ
(i.e. Û † Û = I).

Problem 8:

• Prove that Û as defined above is unitary.

• Starting from the eigenvalue equation:

Â|ψn i = an |ψn i (1.90)

show that the operator

Â0 = Û ÂÛ † (1.91)

has Û |ψn i as its eigenvector with eigenvalue an .

• Show also that the inner product, hΨ|Φi is preserved under a unitary transformation.

• If Û is unitary and  is Hermitian, then show that Û ÂÛ † is also Hermitian.

• Show that the form of the operator equation Ĝ = ÂB̂ is preserved under a unitary transfor-
mation.

The problem above shows that a unitary transformation preserves the form of the eigenvalue equa-
tion. In addition, since the eigenvalues of an operator corresponding to an observable are physically
measurable quantities, these values should not be affected by a transformation of basis in Hilbert
space. It therefore follows that the eigenvalues and the Hermiticity of an observable are preserved
in a unitary transformation.
1.1. INTRODUCTION 19

1.1.11 Matrix representation of operators

From the closure relation (or resolution of the identity) it is possible to express any operator as

 = IˆÂIˆ = |n0 ihn0 |Â|nihn|

XX
(1.92)
n n0

where the set {|ni} are a set of basis vectors in the Hilbert space and the complex numbers hn 0 |Â|ni
are a matrix representation of Â. (Note that the matrix representation of † is obtained by
transposing the matrix representation of  and taking the complex conjugate of each element.)
The table below lists various matrix properties:

Matrix Definition Matrix Elements

Symmetric A = AT Apq = Aqp
Antisymmetric A= −AT Apq = −Aqp
Orthogonal A= (AT )−1 (AT A)pq = δpq
Real A = A∗ Apq = A∗pq
Pure Imaginary A = −A∗ Apq = −A∗pq
Hermitian A = A† Apq = A∗qp
Anti-Hermitian A = −A† Apq = −A∗qp
Unitary A = (A† )−1 (A† A)pq = δpq
Singular |A| = 0

where T denotes the transpose of a matrix and |A| denotes the determinant of matrix A.

Problem 9:

• If A, B, C are 3 n × n square matrices, show that T r(ABC) = T r(CAB) = T r(BCA), where

T r denotes the trace of a matrix, i.e. the sum of its diagonal elements.

• Show that the trace of a matrix remains the same (i.e. invariant) under a unitary transfor-
mation.

• Let A be an n × n square matrix with eigenvalues a1 , a2 , . . . , an . Show that |A| = a1 a2 . . . an

and hence that the determinant of A is another invariant property.

• Show that if A is Hermitian, then U = (A + iI)(A − iI)−1 is unitary. (I here is the identity
matrix.)
20 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

• Show that |I + A| = I +  T rA + O(2 ) where A is an n × n square matrix.

• Show that |eA | = eT rA where A is a n × n square matrix.

1.1.12 Mathematical interlude: Dirac delta function

Definition

The Dirac delta function δ(x) is defined as follows


 0 x 6= 0
δ(x) =
 ∞ x=0

Its integral properties are

Z ∞
f (x)δ(x)dx = f (0)
−∞

Z ∞
δ(x)dx = 1
−∞

Z ∞
f (x0 )δ(x − x0 )dx0 = f (x)
−∞

Z ∞
δ(x − x0 )dx0 = 1 (1.93)
−∞

Note that 
Z b  f (0) 0 ∈ [a, b]
f (x)δ(x)dx =
a  0 Otherwise

In three dimensions, the above definition is generalised as follows

Z
f (r)δ(r − a)dr = f (a) (1.94)
all space

In mathematics, an object such as δ(x), which is defined in terms of its integral properties, is called
a distribution.

Some useful properties

δ(x) = δ(−x)
1.1. INTRODUCTION 21

0 0
δ (x) = −δ (−x)

x δ(x) = 0

1
δ(a x) = δ(x)
|a|

1
 
2 2
δ(x − a ) = δ(x − a) − δ(x + a)
|2a|
Z ∞
δ(a − x) δ(x − b)dx = δ(a − b)
−∞

f (x) δ(x − a) = f (a) δ(x − a)

0
x δ (x) = −δ(x)

g(x) 
Z   
g(x) δ f (x) − a dx = (1.95)
|df /dx| x=x0 , f (x0 )=a

These relations can easily be verified by using some arbitrary function. For example, to prove

0
x δ (x) = −δ(x)

we proceed as follows
Z ∞ 0
Z ∞ d
  Z ∞ d
 
f (x)x δ (x)dx = f x δ dx − δ f x dx
−∞ −∞ dx −∞ dx
Z ∞
df
 
= − δ(x) x + f dx
−∞ dx
Z ∞
= − δ(x)f (x)dx (1.96)
−∞

where we have used integration by parts.

1.1.13 Operators with continuous or mixed (discrete-continuous) spectra

There exist operators which do not have a purely discrete spectra, but either have a continuous or
mixed (discrete-continuous) spectrum. An example is the Hamiltonian for the hydrogen atom. In
general, all Hamiltonians for atoms and nuclei have both discrete and continuous spectral ranges.
Usually the discrete spectrum is connected with bound states while the continuous spectrum is
connected with free (unbound) states. The representation related to such operators cause difficulties
22 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

because eigenstates with continuous spectra are not normalizable to unity. (A rigorous discussion
is too difficult so we will just state the results.)

An observable  has a continuous spectrum if its eigenvalues {a}

Â|ai = a|ai

are a continuous set of real numbers. The eigenstates {|ai} can no longer be normalised to unity
but must be normalised to Dirac delta functions:

ha|a0 i = δ(a − a0 ) (1.97)

The resolution of the identity (or closure relation) becomes

Z
da |aiha| = Iˆ (1.98)

and an arbitrary state can |ψi be expanded in terms of the complete set {|ai} via
Z
|ψi = da0 |a0 iha0 |ψi (1.99)

with ha0 |ψi denoting |ψi in the A representation. The inner product for two state vectors |ψi and
|φi is defined as
Z
hψ|φi = da0 hψ|a0 iha0 |φi

Z
= ψ ∗ (a0 )φ(a0 )da0 (1.100)

If the spectrum is mixed, then the expansion of |ψi is

Z
0 0
|a0 iha0 |ψida0
X
|ψi = |a iha |ψi + (1.101)
a0
where the sum is over the discrete eigenvectors and the integral is over the continuous eigenvectors
|ai.

Position and momentum representations for free particles

In one dimension, the eigenvalue equations for x̂ and p̂ read

x̂|x0 i = x0 |x0 i

p̂|p0 i = p0 |p0 i

hx|x0 i = δ(x − x0 )

hp|p0 i = δ(p − p0 ) (1.102)

1.1. INTRODUCTION 23

These definitions, the fundamental commutator

[x̂, p̂] = ih̄ (1.103)

and the above properties of the Dirac delta function can be used to determine the following matrix
elements:

h̄ ∂
hx0 |p̂|x00 i = δ(x0 − x00 )
i ∂x0

h̄ ∂
hp0 |x̂|p00 i = − δ(p0 − p00 )
i ∂p0
2
∂

hx0 |p̂2 |x00 i = − ih̄ δ(x0 − x00 )
∂x0
2
∂

0 2 00
hp |x̂ |p i = ih̄ 0 δ(p0 − p00 ) (1.104)
∂p

Problem 10†1 : Verify the formulae (1.104)

Now consider the eigenvalue problem for the momentum operator in the position representation. If

p̂|p0 i = p0 |p0 i

then we have
Z
hx0 |p̂|p0 i = dx00 hx0 |p̂|x00 ihx00 |p0 i

∂
Z  
00
= dx − ih̄ 0 δ(x0 − x00 ) hx00 |p0 i
∂x

∂
Z
= −ih̄ dx00 δ(x0 − x00 )hx00 |p0 i
∂x0

∂
= −ih̄ hx0 |p0 i (1.105)
∂x0

On the other hand, we also have

hx0 |p̂|p0 i = p0 hx0 |p0 i
1
Problems that you may choose to skip on a first reading are indicated by †.
24 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

Therefore
∂
−ih̄ hx0 |p0 i = p0 hx0 |p0 i (1.106)
∂x0

which implies
1 ip0 x0
 
0 0
hx |p i = √ exp (1.107)
2πh̄ h̄

where we have chosen the normalisation such that

Z
00 0
hp |p i = dx0 hp00 |x0 ihx0 |p0 i

Z
∗
= dx0 hx0 |p00 i hx0 |p0 i

1 i(p0 − p00 )x0

Z  
0
= dx exp
(2πh̄) h̄

= δ(p00 − p0 ) (1.108)

These results can be generalised to three-dimensions. We have

|ri = |x, y, zi

r̂|ri = r|ri

hr0 |r00 i = δ(r0 − r00 )

|pi = |px , py , pz i

p̂|pi = p|pi

hp0 |p00 i = δ(p0 − p00 )

hr0 |p̂|r00 i = −ih̄∇r0 δ(r0 − r00 )

hp0 |r̂|p00 i = ih̄∇p0 δ(p0 − p00 )

1
 
hr|pi = exp ir · p/h̄ (1.109)
(2πh̄)3/2
1.2. APPLICATIONS 25

1.2 Applications

We apply the formalism developed in the previous section to some familiar (and not so familiar!)
examples. Foremost is the quantum mechanical treatment of the simple harmonic oscillator (SHO).
The SHO is ubiquitous in the quantum mechanical treatment of real phenomena where one is
considering the vibrations of a system after a small displacement from its equilibrium position.

1.2.1 Harmonic oscillator

The Hamiltonian for a one-dimensional quantum-mechanical oscillator is

1 2 1
Ĥ = p̂ + mω 2 x̂2 (1.110)
2m 2
where x̂ and p̂ are the one-dimensional position and momentum operators: they are linear Hermitian
operators satisfying the canonical commutation relation

[x̂, p̂] = ih̄ (1.111)

Introduce a dimensionless operator

1
â = 1 (ip̂ + mωx̂) (1.112)
(2mh̄ω) 2
Then we have
1
↠= 1 (−ip̂ + mωx̂) (1.113)
(2mh̄ω) 2
Note that ↠6= â so that â is NOT Hermitian.

Problem 11: Prove the following properties of the operators â and â † :

• [â, ↠] = Iˆ

• We define a dimensionless operator N̂ where N̂ = ↠â. Show that N̂ is Hermitian and that
it satisfies the following commutation relations: [N̂ , â] = −â and [N̂ , aˆ† ] = ↠.

The observables x̂, p̂ and Ĥ are given in terms of â, ↠, N̂ by

1
h̄

2
x̂ = (â + ↠) (1.114)
2mω
26 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

1
1

2
p̂ = i mh̄ω (↠− â) (1.115)
2
1 1 1
     
† †
Ĥ = ââ − h̄ω = â â + h̄ω = N̂ + h̄ω (1.116)
2 2 2

Problem 12: Show that the above expression for Ĥ is correct.

Let |ni denote an eigenvector of N̂ (we assume that such an eigenvector exists) with eigenvalue n:

N̂ |ni = n|ni (1.117)

Then we find from the above expression for Ĥ that

1
 
Ĥ|ni = N̂ + h̄ω|ni
2

1
 
= n+ h̄ω|ni
2
This means that |ni is an energy eigenvector with eigenvalue
1
 
En = n + h̄ω (1.118)
2

(Note that you would expect this since from the definition of N̂ and Ĥ above, it is clear that
[Ĥ, N̂ ] = [N̂ , Ĥ] = 0̂ and so are compatible.) All that we know about n is that it must be a real
number (Why?). To determine n, consider the effect of â and ↠on |ni.

Problem 13: Prove the following results:

N̂ (â|ni) = (n − 1)(â|ni)

N̂ (↠|ni) = (n + 1)(↠|ni)

This indicates that â and ↠act as lowering and raising operators for the quantum number n:

â|ni = α− |n − 1i (1.119)

↠|ni = α+ |n + 1i (1.120)

1.2. APPLICATIONS 27

where α± are numbers to be determined. For this reason ↠and â are often called creation and
annihilation operators, respectively: they create or annihilate one quantum of excitation of the
oscillator (not the oscillator itself!).

1 1
Problem 14: Show that for normalised kets |ni then α+ = (n + 1) 2 and α− = n 2

Therefore we have
1
â|ni = n 2 |n − 1i (1.121)
1
↠|ni = (n + 1) 2 |n + 1i (1.122)

Problem 15: Prove that the possible values n can take are n = 0, 1, 2, . . . i.e. the non-negative
integers.

The ground state, denoted by the ket |0i has energy

1
E0 = h̄ω (1.123)
2
(i.e. the zero-point energy) and has the property

â|0i = 0 (1.124)

All other eigenvectors can be determined from the ground state eigenvector by the repeated appli-
cation of the raising operator ↠:
1
|ni = 1 (↠)n |0i (1.125)
(n!) 2

The kets |ni defined above are orthonormal (since non-degenerate eigenvectors of a Hermitian
operator corresponding to different eigenvalues are orthogonal). They are also complete:
∞
|nihn| = Iˆ
X
(1.126)
n=0

The set {|ni} is a basis for the Hilbert space i.e any state |ψi can be expressed as
∞
X
|ψi = |nihn|ψi (1.127)
n=0

Similarly, any operator  acting on this space can be written

∞ X
∞
|n0 ihn0 |Â|nihn|
X
 = (1.128)
n=0 n0 =0
28 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

Wave functions for the harmonic oscillator

First we determine the ground state wave function using the identity

â|0i = 0

we have

hx|â|0i = 0
Z 
hx|â |x0 ihx0 |dx0 |0i = 0

Z  
dx0 hx|â|x0 ihx0 |0i = 0

Z  
0 0 0
dx hx|â|x iψ0 (x ) = 0 (1.129)

where we have defined ψ0 (x0 ) = hx0 |0i. To evaluate hx|â|x0 i we use (1.112):
1
hx|â|x0 i = 1 hx|(ip̂ + mωx̂)|x0 i
(2mh̄ω) 2

1
 
= 1 ihx|p̂|x0 i + mωhx|x̂|x0 i (1.130)
(2mh̄ω) 2
Using properties of the operators x̂, p̂ in the position representation :

x̂|x0 i = x0 |x0 i

hx|x̂|x0 i = x0 hx|x0 i

= x0 δ(x − x0 )

∂
hx|p̂|x0 i = −ih̄ δ(x − x0 ) (1.131)
∂x
we get
1 ∂
 
0
hx|â|x i = 1 h̄ δ(x − x0 ) + mωx0 δ(x − x0 ) (1.132)
(2mh̄ω) 2 ∂x
Substituting (1.132) into (1.129) we get
∂
Z  
dx h̄ δ(x − x0 ) + mωx0 δ(x − x0 ) ψ0 (x0 ) = 0
0
∂x

∂
 
h̄ + mωx ψ0 (x) = 0 (1.133)
∂x
1.2. APPLICATIONS 29

which has the normalised solution

1
mω mω 2
  
4
ψ0 (x) = exp − x
πh̄ 2h̄
Z ∞
|ψ0 (x)|2 = 1 (1.134)
−∞

To determine the excited state wave functions we proceed as follows: Since

1
↠|ni = (n + 1) 2 |n + 1i

we have
1
hx|↠|ni = (n + 1) 2 hx|n + 1i (1.135)

Defining ψn (x) = hx|ni, (1.135) becomes

Z
hx|↠|ni = dx0 hx|↠|x0 iψn (x0 )

1
= (n + 1) 2 ψn+1 (x) (1.136)

We then have
1
hx|↠|x0 i = 1 hx|(−ip̂ + mωx̂)|x0 i
(2mh̄ω) 2

1
 
0 0
= 1 − ihx|p̂|x i + mωhx|x̂|x i
(2mh̄ω) 2

1 ∂
 
= 1 − h̄ δ(x − x0 ) + mωx0 δ(x − x0 ) (1.137)
(2mh̄ω) 2 ∂x

and (1.136) reduces to

1 
mω h̄ ∂
 
2 1
− + x ψn (x) = (n + 1) 2 ψn+1 (x) (1.138)
2h̄ mω ∂x
which gives the following recursion formula
1 
1 mω h̄ ∂
 
2
ψn+1 (x) = 1 − + x ψn (x) (1.139)
(n + 1) 2 2h̄ mω ∂x

Problem 16† : Work out the details leading to the formula (1.139) starting from (1.137).
30 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

Since we already know ψ0 (x), (1.139) enables us to generate all the excited wave functions (although
they still have to be normalised). This is a “neater” way of solving for the eigenstates of the quantum
harmonic oscillator when compared with the power series method you have come across before. In
fact, this is an example of the so called factorisation method, a very powerful method for solving
second-order differential equations by purely algebraic means. (We will come across this technique
again when we determine the eigenfunctions of the orbital angular momentum operator.)

The theory outlined in the above paragraph provides a complete description of the quantum-
mechanical properties of a one-dimensional harmonic oscillator. For example, matrix elements of
certain functions of the position and momentum operators can easily be evaluated.

Problem 17† : Prove the following results:

•
1 
h̄
 
2 1 1
0
hn |x̂|ni = n δ
2
n0 ,n−1 + (n + 1) δ 2
n0 ,n+1 (1.140)
2mω

•
1 
1
 
2 1 1
0
hn |p̂|ni = i mh̄ω − n 2 δn0 ,n−1 + (n + 1) 2 δn0 ,n+1 (1.141)
2

•
1 1
h̄
  
2 2
hn0 |xˆ2 |ni = n(n − 1) δn0 ,n−2 + (2n + 1)δn,n0 + (n + 1)(n + 2) δn0 ,n+2 (1.142)
2mω

•
1 1
1
   
2 2
hn |pˆ2 |ni = mh̄ω − n(n − 1)
0
δn0 ,n−2 + (2n + 1)δn,n0 − (n + 1)(n + 2) δn0 ,n+2 (1.143)
2

Problem 18:

• Show that if f (↠) is any polynomial in ↠, then

df (↠)
âf (↠)|0i = |0i (1.144)
dâ†

• Show that:
eλâ f (↠)|0i = f (↠+ λ)|0i (1.145)
1.2. APPLICATIONS 31

• Show that:
eλâ f (↠)e−λâ = f (↠+ λ) (1.146)

• Use the preceding relation to prove that

† † 1
eαâ+βâ = eαâ eβâ e− 2 αβ (1.147)

• Show that:
† †
eλâ f (â)e−λâ = f (â − λ) (1.148)

From this, show that

† † 1
eαâ+βâ = eβâ eαâ e 2 αβ (1.149)

• A state |αi that obeys the equation

â|αi = α|αi (1.150)

is called a coherent state. Show that the state |αi may be written as
†
|αi = Ceαâ |0i (1.151)

and determine the value of C. By expanding the state |αi in the basis {|ni} find the probability
that a coherent state contains n quanta. This probability distribution is called a Poisson
distribution. Finally calculate the value of hα|N̂ |αi which corresponds to the average number
of quanta in the coherent state. (This example is relevant in quantum optics.)

1.2.2 Delta-function potential well

As an example of a system with a mixed (discrete-continuous) spectrum, consider a finite potential

well of width a and depth V0 :

V (x) = −V0 for |x| < 21 a

V (x) = 0 elsewhere (1.152)

In the limit that the well becomes very deep and narrow, such that V0 → ∞ and a → 0 while
aV0 ≡ V remains fixed, we may approximate the potential by a Dirac delta function:

V (x) = −V δ(x) (1.153)

(This will also give us some practice at handling the delta function.)
32 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

Negative-energy eigenstates of this system correspond to bound states, which will be normalisable
and have a discrete spectrum. The wave function must fall off exponentially outside the well and
hence must take the form
ψ(x) = Ae−κ|x| (1.154)

where √
−2mE
κ= (1.155)
h̄
√
and (for normalisation) A = κ. Integrating the Schrödinger equation

h̄2 ∂ 2 ψ
Eψ = − − V δ(x)ψ (1.156)
2m ∂x2
between the limits x = − and x = + gives

h̄2
" #
Z + ∂ψ
 
∂ψ

E ψ dx = − − − Vψ(0) (1.157)
− 2m ∂x x=+ ∂x x=−

Now taking the limit  → 0 will make the integral on the left-hand side vanish, since ψ must be
finite and continuous at x = 0. Therefore ψ must have a discontinuity in its derivative at the origin,
such that " #
∂ψ ∂ψ 2mV
  
lim − =− ψ(0) (1.158)
→0 ∂x x=+ ∂x x=− h̄2
Inserting the form (1.154) for the solution, we find that
mV
κ= (1.159)
h̄2
and hence 2
1 V

E=− m (1.160)
2 h̄
Thus for E < 0 there is a unique solution for E, and hence a single bound state.
√
For E > 0, on the other hand, we can obtain plane-wave solutions with wave number k = 2mE/h̄
for any value of E. Since the potential is an even function of x, we can classify the eigenstates
according to parity. Those with odd parity must vanish at the origin and then equation (1.158)
tells us there is no change in the derivative at the origin, just as if the potential well were not there.
Thus the odd-parity eigenfunctions are simply of the form

ψ(x) = C sin kx (1.161)

and any odd-parity wave function of the system can be Fourier decomposed into sine waves, just
as for a free particle. For the usual delta function normalisation hk|k 0 i = δ(k − k 0 ), we require
√
C = 1/ π.
1.2. APPLICATIONS 33

The even-parity eigenstates, on the other hand, need not vanish at the origin, and hence they feel
the presence of the potential well. For E > 0 we can write them in the general form

ψ(x) = C cos(k|x| + φ) (1.162)

where, from equation (1.158), the phase shift φ satisfies the equation
mV κ
tan φ = 2 = k (1.163)
h̄ k
The presence of the phase shift guarantees that the positive-energy even-parity eigenstates (1.162)
are orthogonal to the bound state (1.154). To see this, consider the overlap integral
Z ∞ Z ∞
cos(k|x| + φ)e−κ|x| dx = 2 cos(kx + φ)e−κx dx
−∞ 0

Z ∞ 
= eikx+iφ−κx + e−ikx−iφ−κx dx
0

eiφ e−iφ
= +
κ − ik κ + ik

2
= (κ cos φ − k sin φ)
κ2 + k2

= 0 (1.164)

on account of equation (1.163). Hence any unbound state of the system can be Fourier decomposed
into sine waves and the “kinked cosine” waves (1.162). Conversely, the square modulus of the
overlap between any normalised state and the bound state (1.154) gives the probability that the
particle is bound in the potential well.

Problem 19: Find the probability Pb that a particle with wave function
√
ψ(x) = c e−c|x| (1.165)

will be found in the bound state of the delta-function potential well. Confirm that P b ≤ 1. Find also
the probability P (k)dk that the particle will be found to be unbound, with wave number between
k and k + dk.

Answer:
4cκ 4ck 2 (c − κ)2
Pb = , P (k) =
(c + κ)2 π(κ2 + k 2 )(c2 + k 2 )2

Problem 20† : Confirm that Z ∞

P (k) dk = 1 − Pb (1.166)
0
34 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

1.3 Introduction to second quantisation

Historically, the development of quantum mechanics focused on the physics of entities which in the
classical limit are particles. To these particles was attributed a wave nature (with Schrödinger’s
equation being the wave equation), which shows up in experiments like electron diffraction for
example. This approach may be called first quantisation. This wave-particle duality forms the
foundation on which quantum mechanics was built. Similarly, it is only natural to expect that
waves in classical physics such as electromagnetic radiation could have particle-like attributes,
a conjecture spectacularly confirmed by the Compton effect and Einstein’s explanation of the
photoelectric effect. In the case of the electromagnetic field, it is appropriate to define it not in
terms of its field amplitudes (E or B) but by how many “photons” or radiation quanta are present
in a given mode, and this is especially important when one tries to understand the interaction
of EM radiation with matter, where the number of photons can vary. The procedure whereby
one constructs a particle description of classical waves is known as second quantisation. Note
however that this procedure is not limited to classical waves. It can be applied to the Schrödinger
equation (a wave equation for the probability amplitude of an electron for example). In the second
quantisation procedure, wave fields are “quantised” to describe the problem in terms of “quanta”
in a given mode of vibration. Examples include photons (quanta of EM field), phonons (quanta of
vibrations in solids), plasmons (quanta of collective electron vibrations) etc. The physical states
of this system are characterised by particle occupation numbers, i.e. the number of quanta in a
particular mode. Below, we will use the vibrations of a stretched string (covered in IB Physics!) to
bring out the essence of the second quantisation approach.

1.3.1 Vibrating string

We first review the classical treatment of the vibrating string. Consider a stretched string of uniform
linear density ρ and length L with fixed ends. The transverse displacement as a function of distance
x along the string is denoted by y(x, t). The equation of the vibrating string is given by

∂2y 2
2∂ y
= v (1.167)
∂t2 ∂x2
q
τ
with the boundary conditions y(0, t) = y(L, t) = 0 for all t. Note that v = ρ where τ denotes the
tension in the string. We now use the method of separation of variables to solve equation (1.167)
subject to the stated boundary conditions. Let

y(x, t) = X(x) ψ(t) (1.168)

1.3. INTRODUCTION TO SECOND QUANTISATION 35

Substituting this into (1.167) and dividing throughout by X(x)ψ(t) gives

1 d2 X 1 d2 ψ
v2 = (1.169)
X dx2 ψ dt2
As the left-hand side depends only x and the right-hand side depends only on t, each must be a
constant: let it be −k 2 . We then have the ordinary differential equations

d2 X
+ k2 X = 0
dx2

d2 ψ
+ v2 k2 ψ = 0 (1.170)
dt2
whose boundary conditions become X(0) = X(L) = 0 and corresponding solutions are respectively

X(x) = c1 cos(kx) + c2 sin(kx)

ψ(t) = b1 cos(vkt) + b2 sin(vkt) (1.171)

Inserting the boundary condition gives

c1 = 0, c2 sin(kL) = 0

As c1 and c2 cannot both be zero (otherwise y(x, t) would be identically zero), we have to choose
sin(kL) = 0 or
kL = nπ, n = 1, 2, 3 · · ·

Thus the allowed general solutions are

∞
nπx
X  
y(x, t) = ψn (t) sin
n=1
L

nπvt nπvt
    
ψn (t) = An cos + Bn sin . (1.172)
L L
(We have replaced the integration constants b1 c2 by An etc.) Each term in the above general
solution is an allowed solution corresponding to an allowed mode of vibration. The Fourier expansion
coefficients ψn (t) are called the normal coordinates and for an arbitrary y(x, t) can be determined
by
2
Z L nπx
 
ψn (t) = y(x, t) sin dx (1.173)
L 0 L
A knowledge of all the normal coordinates completely determines y(x, t). Equation (1.172) rep-
resents the superposition of independent modes of vibration i.e. the normal modes, each with its
characteristic frequency given by ωn = nπv/L. An important property of normal coordinates is
36 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

that the kinetic and potential energies are expressible as uncoupled sums of the normal coordinates
and their first derivatives. In the case of the kinetic energy, we have
L 2
1 ∂y
Z 
T = ρ dx
0 2 ∂t
∞
∂y nπx
X  
= ψ̇n sin (1.174)
∂t n=1
L
which implies
∞
LX ∞
1 nπx mπx
Z X    
T = ρ ψ̇n ψ̇m sin sin dx
2 0 n=1 m=1
L L

∞
1
ρL ψ̇n2 .
X
= (1.175)
n=1
4
Similarly, for the potential energy, we have
L 2
1 ∂y
Z 
V = τ dx
0 2 ∂x
∞
1τ
n2 π 2 ψn2
X
= (1.176)
n=1
4L
To derive the equation of motion for the normal coordinates, we construct the Lagrangian L where
∞ 
1 1τ 2 2 2

ρLψ̇n2
X
L=T −V = − n π ψn (1.177)
n=1
4 4L
and using Lagrange’s equation !
∂ ∂L ∂L
− =0 (1.178)
∂t ∂ ψ̇n ∂ψn
we get
n2 π 2 τ
ψ̈n +
ψn = 0 (1.179)
L2 ρ
This is just the equation for a simple harmonic oscillator!

We can equally well describe the vibrating string problem in the Hamiltonian formalism. Recall
that the generalised momentum corresponding to the generalised coordinate ψ n is defined by
∂L 1
Πn = = ρLψ̇n (1.180)
∂ ψ̇n 2
P
The Hamiltonian function, H (where H(p, q) = pq − L) is given by
∞
X
H = Πn ψ̇n − L
n=1

∞
" #
X Π2n n2 π 2 τ 2
= + ψ (1.181)
n=1
ρL 4L n
1.3. INTRODUCTION TO SECOND QUANTISATION 37

which is an infinite sum of harmonic oscillator Hamiltonian functions. The above equation serves
as the starting point for quantizing the vibrations of the string.

1.3.2 Quantisation of vibrating string

To quantize a system, we make the transition from dynamical variables to operators. Define the
quantum mechanical Hamiltonian operator by making the replacement
∂
Πn → −ih̄ (1.182)
∂ψn
Equation (1.181) then becomes
∞
"
h̄2 ∂ 2
#
X τ 2 2 2
Ĥ = − 2
+ n π ψn
n=1
ρL ∂ψn 4L

∞
X
= Ĥn (1.183)
n=1

i.e. the Hamiltonian reduces to a sum of terms, each of which is dependent on a single variable
(ψn ). Therefore the total wave function for this system is given by
∞
Y
Ψ(ψn , t) = ΦNn (ψn , t) (1.184)
n=1

where ΦNn are the eigenfunctions of the SHO problem satisfying

"
h̄2 ∂ 2
#
τ 2 2 2 ∂
− 2
+ n π ψn ΦNn (ψn , t) = ih̄ ΦNn (ψn , t) (1.185)
ρL ∂ψn 4L ∂t

The solutions of the above equation have the generic form

ΦNn (ψn , t) = UNn (ψn ) e−iEn (Nn )t/h̄ (1.186)

where the UNn satisfy

h̄2 d2 UNn τ 2 2 2
− 2
+ n π ψn UNn = E(Nn )UNn (1.187)
ρL dψn 4L
This is just the differential equation for the SHO with the following correspondence
1
“mass” m → ρL
2
τ 2 2
“force constant” k → n π
2L
s
k nπ τ
r
“frequency” ωn = = (1.188)
m L ρ
38 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS

The solutions UNn are the Hermite polynomials discussed earlier in the chapter. So we can draw
the following conclusions:

• The energy in any given normal mode of frequency ωn is

1
 
En (Nn ) = Nn + h̄ωn
2

1 nπh̄ τ
  r
= Nn + (1.189)
2 L ρ
The number Nn = 0, 1, 2, . . . is the occupation number of the mode.

• We can define annihilation and creation operators for each mode as before:
1 1
 
ân = 1 ρLωn ψ̂n + iΠ̂n
(ρLh̄ωn ) 2 2

1 1
 
â†n = 1 ρLωn ψ̂n − iΠ̂n
(ρLh̄ωn ) 2 2

1 
h̄

2 
ψ̂n = ân + â†n
ρLωn

1 1
 
Π̂n = i (ρLh̄ωn ) 2 â†n − ân (1.190)
2
so that the number operator that counts the number of quanta in a particular mode is given
by
N̂n = â†n ân (1.191)

• The state |Nn i indicates that the system has Nn quanta of excitation of the n-th mode, each
with energy h̄ωn . The effect of the creation and annihilation operators on a state |N n i is:
q
â†n |Nn i = (Nn + 1)|Nn + 1i

p
ân |Nn i = Nn |Nn − 1i (1.192)

• In each mode, the lowest energy state does not have zero energy but has energy 21 h̄ωn . This
represents the “vacuum” or ground state of the mode.

1.3.3 General second quantisation procedure

The procedure for quantising the vibrating string described in the previous section can in principle
be applied to any wave equation. The end result is that one is able to “second-quantise” the wave or
1.3. INTRODUCTION TO SECOND QUANTISATION 39

field amplitude. Instead of describing the allowed modes in terms of the amplitude, we now describe
them by the “quanta” or “particle” number in the mode. This procedure can be summarised as
follows:

1. Given a field or wave equation, represent it in normal coordinates.

2. Find the Lagrangian and Hamiltonian functions corresponding to the field equations with
appropriately defined generalised coordinates and momenta.

3. Treat the normal coordinate Hamiltonian obtained above as a quantum mechanical harmonic
oscillator so that

• States are defined by the number of quanta (particles) in each allowed mode
• Creation and annihilation operators as in the SHO case allow particles/quanta to be
added to or removed from the physical state
• A vacuum state with a fixed energy represents the lowest energy state.

We will extend the concept of second quantisation to interacting many-particle systems, in the
process second-quantising the Schrödinger equation, in Chapter 5. We will see that the second
quantisation procedure makes the treatment of interacting many-particle systems more manageable.
40 CHAPTER 1. OPERATOR METHODS IN QUANTUM MECHANICS
Chapter 2

Angular Momentum

2.1 Introduction

As in classical mechanics, the angular momentum (AM) of an isolated quantum mechanical system
is conserved, and this provides a useful general classification of the states of the system in terms
of their angular momenta. However, the mathematics of AM is a good deal more complicated in
QM, because the values of the AM are quantised and because its different components are not
compatible observables.

The conservation of AM is a consequence of the fact that the properties of an isolated physical
system are invariant under rotations of the system in space (or equivalently, under reorientation
of a Cartesian coordinate system used to describe the system). This is an example of a general
connection between invariances and conservation laws, which exists in classical mechanics but is
especially clear in QM.

In the following sections we develop the mathematical treatment of AM and explain its connection
with rotational invariance.

2.2 Orbital angular momentum

In classical mechanics, the AM of a particle (with respect to the origin) is given by the formula:

L=r×p (2.1)

41
42 CHAPTER 2. ANGULAR MOMENTUM

so that,

Lx = ypz − zpy (2.2)

Ly = zpx − xpz (2.3)

Lz = xpy − ypx (2.4)

where r = (x, y, z) and p = (px , py , pz ) are the position and linear momentum vectors, respectively.

The mathematics of AM comes about through the quantum rule of replacing the linear momentum
p of a classical point particle, located at point r, by p̂ = −ih̄∇, thus replacing the classical angular
momentum of a point particle (about the origin of some chosen Cartesian coordinate system) by
the AM operator:
L̂ = −ih̄r × ∇ (2.5)
This implies:

L̂x = y p̂z − z p̂y

h̄ ∂ ∂
 
= y −z
i ∂z ∂y

L̂y = z p̂x − xp̂z

h̄ ∂ ∂
 
= z −x
i ∂x ∂z

L̂z = xp̂y − y p̂x

h̄ ∂ ∂
 
= x −y (2.6)
i ∂y ∂x
Equation (2.6) can be written more compactly as

L̂i = ijk r̂j p̂k

i, j, k = 1, 2, 3 (2.7)

where 1 → x−component, 2 → y−component, 3 → z−component and we have used the Einstein

summation convention (i.e. the expression is summed over repeated indices). ijk is the completely
antisymmetric tensor of the third rank (also called the Levi-Civita tensor)





+1 for even permutations of (123)
ijk = −1 for odd permutations of (123)


 0

otherwise
2.2. ORBITAL ANGULAR MOMENTUM 43

Some useful identities that the ijk satisfy are

 
 δia δib δic 
 

ijk abc =  δja δjb δjc  (2.8)
 

 δka δkb δkc
 


The antisymmetric nature of the ijk means that

ijk = −jik = −kji

The product of two of these tensors, when summed over a pair of indices (i.e. the indices are
contracted), has the following useful properties

ijk ibc = δjb δkc − δkb δjc (2.9)

ijk ijc = 2δkc (2.10)

ijk ijk = 6 (2.11)

which can be proved by expanding out the terms. We will use these results later to prove some
AM operator identities, but familiarity with them will be useful in your future study of special and
general relativity, classical and quantum electrodynamics, and quantum field theory.

The quantal AM properties of a simple one-particle system are then to be inferred from the prop-
erties of these operators and their actions in the associated Hilbert space. In later sections, we
will deduce the eigenvalues and eigenfunctions of these operators. The components of the quantum
mechanical AM operator satisfy the following commutation relations:

[L̂x , L̂y ] = ih̄L̂z

[L̂y , L̂z ] = ih̄L̂x

[L̂z , L̂x ] = ih̄L̂y (2.12)

or
[L̂i , L̂j ] = ih̄ijk L̂k (2.13)

Further identities include

[L̂i , r̂j ] = ih̄ijk r̂k (2.14)

[L̂i , p̂j ] = ih̄ijk p̂k (2.15)

44 CHAPTER 2. ANGULAR MOMENTUM

Problem 1: Verify (2.13)-(2.15) by direct calculation.

Partial solution: Here is the proof for (2.13):

[L̂i , L̂j ] = iαβ jµν [r̂α p̂β , r̂µ p̂ν ] (2.16)

We use the following operator identity to expand out the right hand side of (2.16)

[ÂB̂, Ĉ D̂] = Â[B̂, Ĉ D̂] + [Â, Ĉ D̂]B̂

   
= Â [B̂, Ĉ]D̂ + Ĉ[B̂, D̂] + [Â, Ĉ]D̂ + Ĉ[Â, D̂] B̂

= Â[B̂, Ĉ]D̂ + ÂĈ[B̂, D̂] + [Â, Ĉ]D̂B̂ + Ĉ[Â, D̂]B̂ (2.17)

Making the substitutions

 → r̂α

B̂ → p̂β

Ĉ → r̂µ

D̂ → p̂ν (2.18)

into (2.17) we have

[L̂i , L̂j ] = iαβ jµν [r̂α p̂β , r̂µ p̂ν ]

 
= iαβ jµν r̂α [p̂β , r̂µ ]p̂ν + r̂α r̂µ [p̂β , p̂ν ] + [r̂α , r̂µ ]p̂ν p̂β + r̂µ [r̂α , p̂ν ]p̂β (2.19)

We now make use of the following operator identities

[r̂i , r̂j ] = 0 (2.20)

[p̂i , p̂j ] = 0 (2.21)

[r̂i , p̂j ] = ih̄δij (2.22)

so that (2.16) reduces to

 
[L̂i , L̂j ] = iαβ jµν − ih̄r̂α p̂ν δβµ + ih̄r̂µ p̂β δαν

= −ih̄iαµ jµν r̂α p̂ν + ih̄iνβ jµν r̂µ p̂β (2.23)

2.2. ORBITAL ANGULAR MOMENTUM 45

Since jµν = −jνµ etc. we find that

iαµ jµν = −iαµ jνµ

= δiν δαj − δij δαν

iνβ jµν = δiµ δβj − δij δβν (2.24)

Substituting (2.19) into (2.18) we have

   
[L̂i , L̂j ] = −ih̄ δiν δαj − δij δαν r̂α p̂ν + ih̄ δiµ δβj − δij δβν r̂µ p̂β

   
= −ih̄ δiν δαj − δij δαν r̂α p̂ν + ih̄ δiα δνj − δij δνα r̂α p̂ν (2.25)

Note that in the last term of the above equation we have relabelled the indices µ → α and β →
ν since these are dummy indices (remember, due to the summation convention, summing over
repeated indices α is the same as summing over repeated indices ν). Equation (2.20) reduces to
 
[L̂i , L̂j ] = ih̄ δiα δνj − δiν δαj r̂α p̂ν

= ih̄kij kαν r̂α p̂ν (2.26)

Since kij = −ikj = ijk and L̂k = kαν r̂α p̂ν we finally have the result we want to prove i.e.

[L̂i , L̂j ] = ih̄ijk L̂k (2.27)

Actually we have done the hardest part. The proof for the other identities is much easier!

From these fundamental commutation relations, the entire theory of AM can be deduced. Evidently
L̂x , L̂y , L̂z are incompatible observables so it will be futile to look for states that are simultaneous
eigenfunctions of L̂x and L̂y . From the generalised uncertainty principle we deduce that:

h̄ ˆ
∆Lx ∆Ly ≥ |hLz i| (2.28)
2

On the other hand, the square of the total angular momentum, L̂2 ,

L̂2 ≡ L̂2x + L̂2y + L̂2z (2.29)

does commute with L̂x , L̂y and L̂z .

46 CHAPTER 2. ANGULAR MOMENTUM

Problem 2: Show that:

[L̂2 , L̂x ] = 0 (2.30)

[L̂2 , L̂y ] = 0 (2.31)

[L̂2 , L̂z ] = 0 (2.32)

or more compactly,
[Lˆ2 , L̂] = 0 (2.33)

Problem 3:

• Show that L̂2 = r̂2 p̂2 − (r̂ · p̂)2 + ih̄(r̂ · p̂)

Solution: We use the tensor method.

L̂2 = ijk r̂j p̂k iαβ r̂α p̂β

= (δjα δkβ − δkα δjβ )r̂j p̂k r̂α p̂β

= r̂α p̂β r̂α p̂β − r̂β p̂α r̂α p̂β

ˆ β
= r̂α r̂α p̂β p̂β − ih̄δαβ r̂α p̂β − r̂β (r̂α p̂α − 3ih̄I)p̂

= (r̂ · r̂)(p̂ · p̂) − ih̄(r̂ · p̂) − r̂β r̂α p̂α p̂β + 3ih̄(r̂ · p̂)

= (r̂ · r̂)(p̂ · p̂) − ih̄(r̂ · p̂) − r̂β r̂α p̂β p̂α + 3ih̄(r̂ · p̂) (2.34)

Using

r̂β r̂α p̂β p̂α = r̂β (p̂β r̂α + ih̄δαβ )p̂α

= r̂β p̂β r̂α p̂α + ih̄r̂α p̂α

= (r̂ · p̂)(r̂ · p̂) + ih̄(r̂ · p̂) (2.35)

Putting this into the above equation, we get the desired result. Note that we have used

ijk iαβ = δjα δkβ − δkα δjβ

p̂α r̂α = r̂α p̂α − 3ih̄

p̂k r̂j = r̂j p̂k − ih̄δjk (2.36)

2.2. ORBITAL ANGULAR MOMENTUM 47

• Show that [L̂2 , r̂] = −2ih̄L̂ × r̂ − 2h̄2 r̂

• Show that [L̂2 , p̂] = −2ih̄L̂ × p̂ − 2h̄2 p̂

Problem 4:

• Evaluate the commutators [L̂z , r2 ] and [L̂z , p2 ].

Answer: [L̂z , r2 ] = 0, [L̂z , p2 ] = 0

• Show that the Hamiltonian Ĥ = (p2 /2m) + V commutes with all three components of L̂,
provided that V depends only on r. (Thus Ĥ, L̂2 , and L̂z form a CSCO.)

2.2.1 Eigenvalues of orbital angular momentum

The results of the previous section imply that L̂2 is compatible with each component of L̂ and so
it is possible to find simultaneous eigenstates of L̂2 and L̂z for example (although note that there
is nothing special about the choosing the z-component. We could just as well have chosen the x or
y component). We then have:
L̂2 Y (x, y, z) = λY (x, y, z) (2.37)

and
L̂z Y (x, y, z) = µY (x, y, z) (2.38)

where Y (x, y, z) is an eigenfunction of both L̂2 and L̂z .

To determine the eigenvalues, we use the “ladder operator” technique, similar to the one used to
study the harmonic oscillator. We define:

L̂± ≡ L̂x ± iL̂y (2.39)

which satisfy the following commutation relation

[L̂z , L̂± ] = ±h̄L̂± (2.40)

and
[L̂2 , L̂± ] = 0 (2.41)
48 CHAPTER 2. ANGULAR MOMENTUM

From the result above, we have

L̂2 (L̂± Y ) = L̂± (L̂2 Y )

= L̂± (λY )

= λ(L̂± Y ) (2.42)

So if Y is an eigenfunction of L̂2 then so is L̂± Y , with the same eigenvalue of L̂2 . Now consider
the following:

L̂z (L̂± Y ) = (L̂z L̂± − L̂± L̂z )Y + L̂± L̂z Y

= ±h̄L̂± Y + L̂± (µY )

= (µ ± h̄)(L̂± Y ) (2.43)

This means that (L̂± Y ) is also an eigenfunction of L̂z , but with a different eigenvalue, µ ± h̄. L̂+
is called a “raising” operator because it increases the value of L̂z by h̄, whereas L̂− is called a
“lowering ” operator since it decreases the eigenvalue of L̂z by h̄. Therefore starting from a given
value of λ one obtains a “ladder” of states with each “rung” separated from its neighbours by one
unit of h̄ in the eigenvalue of L̂z . Then to go up the ladder, one applies L̂+ while to go down, L̂− .
This process cannot go on forever though: Eventually we are going to reach a state for which the
z-component of its AM exceeds the total AM of that state and this cannot be!

Problem 5: Prove that if Y is simultaneously an eigenfunction of L̂2 and L̂z then the square of
the eigenvalue of L̂z cannot exceed the eigenvalue of L̂2 . (Hint: Examine the expectation value of
L̂2 ).

The above problem indicates that there must exist a “top rung”, Ytop , such that

L̂+ Ytop = 0 (2.44)

Let the eigenvalue of L̂z for this “top” state to be h̄l. Then we have

L̂z Ytop = h̄lYtop

L̂2 Ytop = λYtop (2.45)

2.2. ORBITAL ANGULAR MOMENTUM 49

Problem 6: Prove that

L̂2 = L̂± L̂∓ + L̂2z ∓ h̄L̂z (2.46)

Using the above result it follows that

L̂2 Ytop = (L̂− L̂+ + L̂2z + h̄L̂z )Ytop

= (0 + h̄2 l2 + h̄2 l)Ytop

= h̄2 l(l + 1)Ytop (2.47)

and hence
λ = h̄2 l(l + 1) (2.48)

which tells us the eigenvalue of L̂2 in terms of the maximum eigenvalue of L̂z . Following the same
reasoning, there is also a “bottom” rung, Ybottom , such that

L̂− Ybottom = 0 (2.49)

with h̄¯l the eigenvalue of L̂z . Following the same method above, we deduce that

λ = h̄2 ¯l(¯l − 1) (2.50)

Problem 7: Verify (2.50). Hence deduce that the only possible value of ¯l is

¯l = −l (2.51)

Evidently the eigenvalues of L̂z are mh̄, where m is an integer going from +l to −l in N integer
N
steps. This implies that l = 2 so that l must be an integer or half-integer. The eigenfunctions of
L̂2 and L̂z are characterised by the numbers l and m, where:

L̂2 Yl,m = h̄2 l(l + 1)Yl,m

1 3
l = 0, , 1, , · · · (2.52)
2 2
50 CHAPTER 2. ANGULAR MOMENTUM

L̂z Yl,m = h̄mYl,m

m = −l, −l + 1, . . . , l − 1, l (2.53)

Note that for a given value of l, there are 2l + 1 different values for m.

Problem 8: The raising and lowering operators change the value of m by one unit, i.e.

L̂± Yl,m = A±
l,m Yl,m±1 (2.54)

What is the value of A±

l,m , if the eigenfunctions Yl,m are to be normalised?

Answer: A±
l,m can take the values:
q
A±
l,m = h̄ l(l + 1) − m(m ± 1)

q
= h̄ (l ∓ m)(l ± m + 1) (2.55)

This result will be of use later when we study the addition of two angular momenta.

2.2.2 Eigenfunctions of orbital angular momentum

We now determine the eigenfunctions for the operators L̂2 and L̂z . The conventional way is to
write out these operators in spherical polar coordinates:

x = r sin θ cos ϕ

y = r sin θ sin ϕ

z = r cos θ (2.56)

from which we can derive:

∂ ∂ 1 ∂ sin ϕ ∂
= sin θ cos ϕ + cos θ cos ϕ −
∂x ∂r r ∂θ r sin θ ∂ϕ

∂ ∂ 1 ∂ cos ϕ ∂
= sin θ sin ϕ + cos θ sin ϕ +
∂y ∂r r ∂θ r sin θ ∂ϕ

∂ ∂ sin θ ∂
= cos θ − (2.57)
∂z ∂r r ∂θ
2.2. ORBITAL ANGULAR MOMENTUM 51

Then the L̂z operator in spherical polar coordinates is:

∂
L̂z = −ih̄ (2.58)
∂ϕ
while the raising/lowering operators are:
∂ ∂
L̂± = h̄e±iϕ (± + i cot θ ) (2.59)
∂θ ∂ϕ

Problem 9: Verify this. From these results determine the value of L̂2 in spherical polar coordinates.
(Hint: You could use the result proven earlier, i.e. L̂2 = L̂+ L̂− + L̂2z − h̄L̂z ).

Since we have already shown that Ylm (θ, ϕ) is a simultaneous eigenfunction of L̂z and L̂2 , from
the above results it is clear that the easiest way to determine Ylm (θ, ϕ) is by using the method of
separation of variables:
Ylm (θ, ϕ) = Θlm (θ)Φm (ϕ) (2.60)

so that
L̂z Ylm (θ, ϕ) = mh̄Ylm (θ, ϕ) (2.61)

implies
dΦm (ϕ)
= imΦm (ϕ) (2.62)
dϕ
which is a first-order differential equation in ϕ. The normalised solutions are:
1
Φm (ϕ) = p eimϕ
(2π)
Z 2π
dϕ |Φm |2 = 1 (2.63)
0

Therefore
Ylm (θ, ϕ) = Θlm (θ)eimϕ (2.64)

To solve for Θlm (θ) we proceed as follows. We use the fact that the maximum value of m is l such
that:
L̂+ Yll = 0 (2.65)

Expanding this out gives the equation

∂ ∂ ∂
   
iϕ
h̄e + i cot θ Θll eilϕ = h̄ei(l+1)ϕ − l cot θ Θll (θ) = 0 (2.66)
∂θ ∂ϕ ∂θ
52 CHAPTER 2. ANGULAR MOMENTUM

whose solution is
Θll (θ) = (sin θ)l (2.67)

Starting from this, we can determine all the other states by applying the lowering operator. This
means that an arbitrary state can be represented by:

Ylm (θ, ϕ) = C L̂l−m l ilϕ

− (sin θ) e (2.68)

where C is a constant.

Problem 10: We investigate the action of L̂− on the “top” state. First verify the following
identity:
∂ 1 d
   
+ l cot θ f (θ) = l
(sin θ)l f (θ) (2.69)
∂θ (sin θ) dθ

Then show that by acting on Yll with L̂− we get:

i(l−1)ϕ 
0e d
 
l l
Yl,l−1 = C − (sin θ) (sin θ)
(sin θ)l dθ

ei(l−2)ϕ d 1 d
 
Yl,l−2 = C 00 (l−1)
(sin θ)2l (2.70)
(sin θ) dθ sin θ dθ

where C 0 , C 00 are constants.

d
If we define a new variable u = cos θ such that − sin1 θ dθ = d
du then the general form for Ylm (θ, ϕ) is

l−m
eimϕ d

Ylm (θ, ϕ) = C [(1 − u2 )l ] (2.71)
(sin θ)m du

These eigenfunctions (the spherical harmonics) are to be normalised over the unit sphere where
the range of integration is 0 ≤ ϕ ≤ 2π and 0 ≤ θ ≤ π. This means
Z Z 2π Z π
2
dΩ |Ylm (θ, ϕ)| = dϕ dθ sin θ |Ylm (θ, ϕ)|2
0 0

Z 2π Z 1
= dϕ du |Ylm (θ, ϕ)|2
0 −1

= 1 (2.72)
2.2. ORBITAL ANGULAR MOMENTUM 53

2.2.3 Mathematical interlude: Legendre polynomials and spherical harmonics

Legendre polynomials

The Legendre polynomials Pl (u) arise when we solve Laplace’s equation in spherical polar coordi-
nates for a problem possessing azimuthal symmetry. They are real functions (polynomials of order
l) defined in the interval [−1, +1] and have the following properties:

The generating function is

∞
1
Pl (u)hl
X
1 = (2.73)
(1 − 2hu + h2 ) 2
l=0

1 dl 2
Pl (u) = (u − 1)l (2.74)
2l l! dul
where −1 ≤ u ≤ 1, and l is a non-negative integer. The Pl (u) arise from a power-series solution of
the 2nd order differential equation, the Legendre equation:
00 0
(1 − u2 )Pl (u) − 2uPl (u) + l(l + 1)Pl (u) = 0 (2.75)

and satisfy the following recurrence relations

(l + 1)Pl+1 (u) = (2l + 1)uPl (u) − lPl−1 (u)

0
(1 − u2 )Pl (u) = −luPl (u) + lPl−1 (u) (2.76)

They are normalised as follows:

Z 1 2
Pl (u)Pm (u)du = δlm (2.77)
−1 2l + 1
The first few polynomials are

P0 (u) = 1

P1 (u) = u

1
P2 (u) = (3u2 − 1)
2

1
P3 (u) = (5u3 − 3u) (2.78)
2
Note that

Pl (u) = (−1)l Pl (−u)

Pl (1) = 1 (2.79)
54 CHAPTER 2. ANGULAR MOMENTUM

Spherical harmonics

The spherical harmonics, Ylm (θ, ϕ), that are conventionally used are:
1
2l + 1 (l − m)!

2
m
Ylm (θ, ϕ) = (−1) Plm (cos θ)eimϕ (2.80)
4π (l + m)!
which is valid for m ≥ 0, while the eigenfunctions corresponding to negative values of m are obtained
from:
Yl,−m = (−1)m Ylm
∗
(2.81)

The functions Plm are the associated Legendre polynomials defined by:
l−m
(l + m)! (1 − u2 )−m/2 d

Plm (u) = (−1)l+m (1 − u2 )l (2.82)
(l − m)! 2l l! du
which is valid for m ≥ 0, and the values for negative m given by:
(l − m)! m
Pl−m (u) = (−1)m P (u) (2.83)
(l + m)! l
(Remember u = cos θ).

Problem 11† : From the above definitions of Ylm (θ, ϕ) and Plm (cos θ), determine the values of
Y00 , Y10 , Y11 , Y20 , Y21 , Y22 . Write out the quantities x, y, z, xy, yz, zx, x2 − y 2 , 2z 2 − x2 − y 2 , y 2 − z 2
q
4π
in terms of the above Ylm (θ, ϕ) (and their complex conjugates). (For example, z = r 3 Y1,0 where
x y
r2 = x2 + y2 + z 2 .) Hence show that r is an eigenfunction of L̂x and r is an eigenfunction of L̂y ,
and find their respective eigenvalues.

When m = 0, (2.80) simplifies to a value independent of the azimuthal angle ϕ

s
2l + 1
Yl0 (θ, ϕ) → Yl0 (θ, 0) = Pl (cos θ) (2.84)
4π
where Pl (cos θ) are the Legendre polynomials.

At the poles θ = 0 and θ = π the azimuthal angle ϕ are indistinguishable so Ylm (θ, ϕ) cannot
depend on ϕ at these angles. Therefore the only non-zero spherical harmonics for these two cases
has to have m = 0 i.e. s
2l + 1
Ylm (0, ϕ) = δm0 (2.85)
4π
and s
l 2l + 1
Ylm (π, ϕ) = (−1) δm0 (2.86)
4π
2.2. ORBITAL ANGULAR MOMENTUM 55

Consider two arbitrary directions in space defined respectively by the angles (θ, ϕ) and (θ 0 , ϕ0 ) and
call the angle between them Θ. The spherical harmonic addition theorem states that

4π m=+l ∗
(θ0 , ϕ0 )
X
Pl (cos Θ) = Ylm (θ, ϕ) Ylm (2.87)
2l + 1 m=−l

Problem 12: Let r12 = |r1 − r2 | be the distance between two points in space so that the r1
direction is (θ1 , ϕ1 ) and r2 direction is (θ2 , ϕ2 ). Prove that

∞ m=+l l
1 X X 4π r< ∗
= Y (θ , ϕ1 )Ylm
l+1 lm 1
(θ2 , ϕ2 ) (2.88)
|r1 − r2 | l=0 m=−l (2l + 1) r>

where r< stands for the smaller of the two distances |r1 | and |r2 |, and r> is the larger of the two
distances.

This result is very useful and we will apply it when we study scattering theory as well as in
variational calculations for the ground state of the helium atom.

Note that the Ylm (θ, ϕ) form a complete set of orthonormal functions, i.e.
Z
dΩ Ylm (θ, ϕ) Yl∗0 m0 (θ, ϕ) = δll0 δmm0 (2.89)

and
∞ m=+l
∗ 1
(θ0 , ϕ0 ) = δ(θ − θ 0 )δ(ϕ − ϕ0 )
X X
Ylm (θ, ϕ)Ylm (2.90)
l=0 m=−l
sin θ

This means that any function of θ and ϕ can be expanded as follows:

∞ m=+l
X X
f (θ, ϕ) = Clm Ylm (θ, ϕ) (2.91)
l=0 m=−l

where Z
∗
Clm = dΩ Ylm (θ, ϕ)f (θ, ϕ) (2.92)

Furthermore, if f (θ, ϕ) is the angular wavefunction of some state, normalised such that
Z
dΩ|f (θ, ϕ)|2 = 1 (2.93)

then the |Clm |2 are the probabilities that simultaneous measurement of L̂2 and L̂z on the state
described by f (θ, ϕ) yields l(l + 1)h̄2 and mh̄ respectively.
56 CHAPTER 2. ANGULAR MOMENTUM

In the Dirac notation, an arbitrary state |ψi can be expanded as:

X
|ψi = Clm |lmi (2.94)
lm

where with the help of the orthonormality condition

hl0 m0 |lmi = δll0 δmm0 (2.95)

we get
Clm = hlm|ψi (2.96)

These eigenstates are also complete so:

∞ m=+l
|lmihlm| = Iˆ
X X
(2.97)
l=0 m=−l

Problem 13: Using hlm0 |Lˆz |lmi = h̄mδm0 m and

 1
2
hlm |Lˆ± |lmi = h̄ l(l + 1) − m(m ± 1)
0
δm0 ,m±1

calculate the matrix representation of L̂x , L̂y and L̂z for the AM state 3/2. Check that the
commutation relations hold.

Problem 14: The Hamiltonian of a rigid rotator with principal moments of inertia I 1 , I2 , I3 is

1 2 1 2 1 2
Ĥ = L̂x + L̂y + L̂
2I1 2I2 2I3 z

Find the eigenvalues of Ĥ if the AM of the system is 1.

Answer: The eigenvalues are h̄2 (α + β), h̄2 (β + γ), and h̄2 (α + γ), where α = 1
2I1 , β= 1
2I2 , γ= 1
2I3 .

2.2.4 Angular momentum and rotational invariance

The fact that the Hamiltonian commutes with the components of the AM is a consequence of
the invariance of the Hamiltonian under reorientation of the coordinate axes used to describe the
system. For example, a rotation of the axes through an angle α around the z-axis changes the
2.2. ORBITAL ANGULAR MOMENTUM 57

coordinate system from S to S 0 where

x0 = x cos α + y sin α

y 0 = −x sin α + y cos α

z0 = z (2.98)

or in terms of spherical polar coordinates

r 0 = r , θ 0 = θ , φ0 = φ − α (2.99)

Notice that by Taylor expansion we can write

∞ n
1 ∂
X 
ψ(r, θ, φ − α) = −α ψ(r, θ, φ)
n=0
n! ∂φ

∂
 
= exp −α ψ(r, θ, φ)
∂φ

i
 
= exp − αL̂z ψ(r, θ, φ) (2.100)
h̄
The operator
i
 
Û = exp − αL̂z (2.101)
h̄
is an example of a rotation operator, in this case a rotation of the axes through an angle α, or
equivalently a rotation of the system through an angle −α, about the z-axis. More generally, a
rotation of the system through an angle ω around the unit vector n will be achieved by the operator
i
 
Û = exp ω L̂ · n (2.102)
h̄
Note that a rotation operator is unitary:
i
 
Û = exp − ω L̂ · n = Û −1
†
(2.103)
h̄

The fact that the components of L̂ commute with the Hamiltonian ensures that any rotation
operator also commutes with Ĥ. It follows that, if ψ is a possible state of the system with energy
E, then so is Û ψ, with the same energy: if Ĥψ = Eψ and Û Ĥ = Ĥ Û , then

Ĥ(Û ψ) = Û Ĥψ = Û Eψ = E(Û ψ) (2.104)

Conversely, if every rotated state Û ψ is an eigenstate of Ĥ with the same energy as ψ, then all
components of L̂ must commute with Ĥ, which implies that L is a conserved quantity.
58 CHAPTER 2. ANGULAR MOMENTUM

2.3 Spin angular momentum

The electron possesses an internal angular momentum called the spin which can assume only the
values + h̄2 and − h̄2 in some arbitrarily chosen direction. In fact, all elementary particles have a spin
degree of freedom. Fermions possess half-integral spin while bosons have integral spin (including
zero). In what follows, we will develop the theory for spin- 21 fermions.

Let the spin operator be Ŝ = (Ŝx , Ŝy , Ŝz ). If n is a unit vector (pointing in some arbitrary direction),
then the Stern-Gerlach experiment indicates that the eigenvalue of the operator Ŝ · n has only two
values (which turns out to be ± h̄2 ) i.e.

h̄
Ŝ · n|n, ±i = ± |n, ±i (2.105)
2
Without loss of generality, one can choose n to point in the z−direction. (Then Ŝ · n = Ŝz ). The
eigenvalue equation then takes the form:
h̄
Ŝz | ↑i = | ↑i
2

h̄
Ŝz | ↓i = − | ↓i (2.106)
2
where | ↑i corresponds to the spin “pointing” in the positive z-axis direction (i.e a spin-up state)
and | ↓i corresponds to the spin “pointing” in the negative z-axis direction (i.e. a spin-down state).
Since spin is a physical observable, Ŝz is Hermitian and the states belonging to distinct eigenvalues
are orthogonal, that is:
h↑ | ↓i = h↓ | ↑i = 0 (2.107)

We further normalise them to unity:

h↑ | ↑i = h↓ | ↓i = 1 (2.108)

The spin operators satisfy the AM commutation relations:

h i
Ŝi , Ŝj = ih̄ijk Ŝk
h i
Ŝz , Ŝ± = ±h̄Ŝ±
h i
Ŝ+ , Ŝ− = 2h̄Ŝz (2.109)

where

Ŝ± = Ŝx ± iŜy

2.3. SPIN ANGULAR MOMENTUM 59

1
Ŝx = (Ŝ+ + Ŝ− )
2

1
Ŝy = −i (Ŝ+ − Ŝ− ) (2.110)
2

For spin S = 12 , Ŝ 2 has the eigenvalue 43 h̄2 :

3
Ŝ 2 | ↑i = h̄2 | ↑i
4

3
Ŝ 2 | ↓i = h̄2 | ↓i (2.111)
4

Problem 15: Show that:

Ŝ+ | ↑i = 0

Ŝ− | ↑i = h̄| ↓i

Ŝ+ | ↓i = h̄| ↑i

Ŝ− | ↓i = 0 (2.112)

We can now represent the spin operators in the basis states | ↑i and | ↓i by the spin matrices
!
h↑ |Ŝi | ↑i h↑ |Ŝi | ↓i
Ŝi → (Si ) = (2.113)
h↓ |Ŝi | ↑i h↓ |Ŝi | ↓i

Problem 16: Show that: !

0 1
Ŝ+ = h̄ (2.114)
0 0
!
0 0
Ŝ− = h̄ (2.115)
1 0
!
h̄ 1 0
Ŝz = (2.116)
2 0 −1
60 CHAPTER 2. ANGULAR MOMENTUM

Introducing the Pauli spin matrices by

h̄
S= σ (2.117)
2
we obtain for them
!
0 1
σx =
1 0
!
0 −i
σy =
i 0
!
1 0
σz = (2.118)
0 −1

Problem 17: Show that:

σx2 = σy2 = σz2 = I (2.119)

[σi , σj ] = 2iijk σk (2.120)

σi σj + σj σi = 2δij (2.121)

T r(σi ) = 0 (2.122)

|σi | = −1 (2.123)

where I is the 2 × 2 unit matrix. Show that the above results can be summarised compactly by the
identity
σi σj = δij + iijk σk (2.124)

Hence, given arbitrary vectors a, b show that

(σ · a)(σ · b) = Ia · b + iσ · (a × b) (2.125)

Problem 18:

Spin state space is a two-dimensional Hilbert space HS , spanned by the orthonormal basis vectors
{| ↑i, | ↓i}. In matrix notation, these basis vectors are defined as:
!
1
| ↑i =
0
!
0
| ↓i = (2.126)
1
2.3. SPIN ANGULAR MOMENTUM 61

These basis vectors are the eigenvectors of the Ŝz , i.e.

h̄
Ŝz | ↑i = + | ↑i
2

h̄
Ŝz | ↓i = − | ↓i (2.127)
2
with

h↑ | ↑i = 1

h↓ | ↓i = 1

h↑ | ↓i = 0

h↓ | ↑i = 0 (2.128)

This basis is complete:

| ↑ih↑ | + | ↓ih↓ | = Iˆ (2.129)

In general, ∀|ψi ∈ HS , we have:

|ψi = α| ↑i + β| ↓i

|α|2 + |β|2 = 1 (2.130)

(Note that in general α and β are complex numbers.) In the {| ↑i, | ↓i} basis, the matrix represen-
tation of Ŝz is diagonal: !
h̄ 1 0
(Sz ) = (2.131)
2 0 −1
while the x− and y−components of the spin operator are:
!
h̄ 0 1
(Sx ) =
2 1 0
!
h̄ 0 −i
(Sy ) = (2.132)
2 i 0

• Show that the normalised eigenvectors of Ŝx (denoted by |±ix ) and Ŝy (denoted by |±iy )
when expanded in the basis {| ↑i, | ↓i} are:
1
|±ix = √ [| ↑i ± | ↓i]
2

1
|±iy = √ [| ↑i ± i| ↓i] (2.133)
2
62 CHAPTER 2. ANGULAR MOMENTUM

• Let u be a unit vector characterised by polar angles ϑ and ϕ with Cartesian components:

u = (sin ϑ cos ϕ, sin ϑ sin ϕ, cos ϑ) (2.134)

Show that if Su = S · u then:

sin ϑe−iϕ
!
h̄ cos ϑ
(Su ) = (2.135)
2 sin ϑeiϕ − cos ϑ
Then show that in the basis {| ↑i, | ↓i}, the eigenvectors of Su are:
ϑ ϕ ϑ ϕ
| ↑iu = cos( )e−i 2 | ↑i + sin( )ei 2 | ↓i (2.136)
2 2

ϑ ϕ ϑ ϕ
| ↓iu = − sin( )e−i 2 | ↑i + cos( )ei 2 | ↓i (2.137)
2 2

2.3.1 Spinors

In the basis {| ↑i, | ↓i}, a general spin state |χi can be written as

|χi = a+ | ↑i + a− | ↓i (2.138)

with complex coefficients a± . Normalisation requires that

|a+ |2 + |a− |2 = 1 (2.139)

The state |χi defined above can also be represented by a two-component column vector called a
spinor whose components are given by the projections onto the basis {| ↑i, | ↓i}:
   
a+ h↑ |χi
|χi =  =  (2.140)
a− h↓ |χi
The basis spinors are    
1 0
χ+ =   , χ− =   (2.141)
0 1
and the completeness relation in this matrix representation is
!
1 0
χ+ χ†+ + χ− χ†− = (2.142)
0 1

Problem 19: An electron is in the spin state

√
1 2 2
|χi = | ↑i + | ↓i
3 3
2.3. SPIN ANGULAR MOMENTUM 63

• What is the probability that a measurement of the z−component of the spin will result in
1
2 h̄? In − 12 h̄?
1
Answer: 9 and 89 .

• What is the expectation value of Ŝz ?

7
Answer: − 18 h̄

• What is the RMS uncertaintity in Ŝz ?

√
2 2
Answer: 9 h̄.

Spin is an additional degree of freedom independent of the spatial degrees of freedom. Spin and
position (or momentum) can assume precise values simultaneously and independently of one another
i.e.
[Ŝ, r̂] = [Ŝ, p̂] = [Ŝ, L̂] = 0 (2.143)

The total quantum state of a particle is constructed from the direct product of the position and
spin eigenstates. The states |ri ⊗ | ↑i and |ri ⊗ | ↓i forms a basis for the Hilbert space (that takes
into account the spatial and spin degrees of freedom). A general state |Ψi in this Hilbert space is
then given by Z  
|Ψi = dr0 ψ+ (r0 ) |r0 i ⊗ | ↑i + ψ− (r0 ) |r0 i ⊗ | ↓i (2.144)

The projection onto position and spin eigenstates are

Z  
hr|Ψi = dr0 ψ+ (r0 ) hr|r0 i| ↑i + ψ− (r0 ) hr|r0 i| ↓i

= ψ+ (r)| ↑i + ψ− (r)| ↓i (2.145)

since hr|r0 i = δ(r − r0 ) and

 
hr| |r0 i ⊗ | ↑i = hr|r0 i| ↑i

 
h↑ | hr|Ψi = ψ+ (r)

 
h↓ | hr|Ψi = ψ− (r) (2.146)

The quantities |ψ± (r)|2 express the probability of finding the particle at position r with z−component
of spin of ± h̄2 . The normalisation condition is
Z   
0 0 ∗ 0 0 ∗ 0 0 0 0 0
hΨ|Ψi = dr hr | ⊗ h↑ |ψ+ (r ) + hr | ⊗ h↓ |ψ− (r ) |r i ⊗ | ↑iψ+ (r ) + |r i ⊗ | ↓iψ− (r )
64 CHAPTER 2. ANGULAR MOMENTUM

Z  
0 0 2 0 2
= dr |ψ+ (r )| + |ψ− (r )|

= 1 (2.147)

2.4 Addition of angular momenta

In general, if we have two angular momentum operators, Ĵ1 and Ĵ2 (where the Ĵi could correspond
to the sum of the orbital AM and the spin AM of one particle, or the spin AM of two electrons,
etc.), we would like to know the possible values of the total AM Ĵ can take where

Ĵ = Ĵ1 + Ĵ2 (2.148)

Assuming that Ĵ1 and Ĵ2 correspond to distinct degrees of freedom, they commute with each other:

[Ĵ1 , Ĵ2 ] = 0 (2.149)

Together with the AM commutation relations for Ĵ1 and Ĵ2 individually, the commutation relations
for Ĵ are given by:
[Jˆi , Jˆj ] = ih̄ijk Jˆk (2.150)

So all the properties of AM and their eigenstates discussed in the sections above also hold for the
total AM.

We start with states |j1 , m1 i and |j2 , m2 i where the two quantum numbers j1 and j2 are fixed and
the mi take the values −ji , · · · , +ji . The corresponding eigenvalue equations are:

Jˆ12 |j1 , m1 i = h̄2 j1 (j1 + 1)|j1 , m1 i

Jˆ1z |j1 , m1 i = h̄m1 |j1 , m1 i (2.151)

Jˆ22 |j2 , m2 i = h̄2 j2 (j2 + 1)|j2 , m2 i

Jˆ2z |j2 , m2 i = h̄m2 |j2 , m2 i (2.152)

From these states we construct the product states

|j1 , m1 , j2 , m2 i = |j1 , m1 i|j2 , m2 i (2.153)

which are eigenfunctions of Jˆz with eigenvalue h̄(m1 +m2 ). However, these are NOT eigenfunctions
of Jˆ2 since
Jˆ2 = Jˆ12 + Jˆ22 + 2Ĵ1 · Ĵ2 (2.154)
2.4. ADDITION OF ANGULAR MOMENTA 65

and the last term does not commute with Jˆ1z or Jˆ2z :

[Ĵ1 · Ĵ2 , Jˆ1z ] = [Jˆ1x , Jˆ1z ]Jˆ2x + [Jˆ1y , Jˆ1z ]Jˆ2y = ih̄(−Jˆ1y Jˆ2x + Jˆ1x Jˆ2y ) (2.155)

and similarly

[Ĵ1 · Ĵ2 , Jˆ2z ] = [Jˆ1x , Jˆ1z ]Jˆ2x + [Jˆ1y , Jˆ1z ]Jˆ2y = ih̄(−Jˆ1x Jˆ2y + Jˆ1y Jˆ2x ) (2.156)

Thus we cannot simultaneously specify the quantum numbers m1 and m2 together with the eigen-
value of Jˆ2 .

The product states (2.153) are eigenstates of the operators

Jˆ12 , Jˆ1z , Jˆ22 , Jˆ2z (2.157)

What we need to do is search for states in which Jˆ2 is also diagonal; that is, we seek eigenfunctions

|j, mj , j1 , j2 i (2.158)

of the four mutually commuting operators

Jˆ2 , Jˆz , Jˆ12 , Jˆ22 (2.159)

with eigenvalues h̄2 j(j + 1), h̄mj , h̄2 j1 (j1 + 1), h̄2 j2 (j2 + 1). At the same time we have to find the
values taken by j (the corresponding mj are then −j, · · · , j), and we have to represent |j, mj , j1 , j2 i
as a linear combination of the product states above. Before considering the general case, we first
look at the addition of two spins, and then at the addition of an orbital AM and a spin.

2.4.1 Addition of spin- 21 operators

Let Ŝ1 and Ŝ2 be two spin- 12 operators whose total spin AM is given by

Ŝ = Ŝ1 + Ŝ2 (2.160)

There are four possible product states:

| + +i = | ↑i| ↑i

| + −i = | ↑i| ↓i

| − +i = | ↓i| ↑i

| − −i = | ↓i| ↓i (2.161)
66 CHAPTER 2. ANGULAR MOMENTUM

in which the first (second) symbol refers to the first (second) spin. These product states are
eigenstates of the operators Ŝ12 , Ŝ22 , Ŝ1z , Ŝ2z .

Problem 20:

• Evaluate the action of the operator Ŝz on each of the product states.

• Show that

Ŝ 2 = Ŝ12 + Ŝ22 + 2Ŝ1 · Ŝ2

3 2
= h̄ + 2Ŝ1z Ŝ2z + Ŝ1+ Ŝ2− + Ŝ1− Ŝ2+ (2.162)
2
Hence show that

Ŝ 2 | − −i = 2h̄2 | − −i

Ŝ 2 | + +i = 2h̄2 | + +i (2.163)

The states | − −i and | + +i therefore have total spin S = 1 and whose z−component of total
spin takes the value of −h̄ and h̄ respectively.

• Given that Ŝ− = Ŝ1− + Ŝ2− show that

1 1
√ Ŝ− | + +i = √ (| + −i + | − +i) (2.164)
h̄ 2 2
1
(The resulting state has been normalised to unity by inserting the factor √
h̄ 2
. Show that the
z−component of this state is zero. Using the notation |S, mi where S designates the total
spin and m its z−component, we have

|1, 1i = | + +i

1
|1, 0i = √ (| + −i + | − +i)
2

|1, −1i = | − −i (2.165)

• There is an additional state |0, 0i which is orthogonal to those given above:

1
|0, 0i = √ (| + −i − | − +i) (2.166)
2
By acting on this state with Ŝ 2 and Ŝz show that this state has spin zero. So we have found
all the eigenstates of Ŝ 2 and Ŝz . The states above corresponding to S = 1 are referred to as
triplet states and that corresponding to S = 0 as a singlet state.
2.4. ADDITION OF ANGULAR MOMENTA 67

• Consider the following projection operators:

3 1
P̂S=1 = + 2 Ŝ1 · Ŝ2
4 h̄

1 1
P̂S=0 = − 2 Ŝ1 · Ŝ2 (2.167)
4 h̄
Show that P̂S=1 projects onto triplet states while P̂S=0 projects onto singlet states.

2.4.2 Addition of spin- 12 and orbital angular momentum

Starting with L̂ and Ŝ, define the total AM as:

Ĵ = L̂ + Ŝ (2.168)

The eigenstates of the operators L̂2 , L̂z , Ŝ 2 , Ŝz are given by:

L̂z |l, ml i = h̄ml |l, ml i

L̂2 |l, ml i = h̄2 l(l + 1)|l, ml i

h̄
Ŝz | ↑i = | ↑i
2

h̄
Ŝz | ↓i = − | ↓i
2

3 2
Ŝ 2 | ↑i = h̄ | ↑i
4

3 2
Ŝ 2 | ↓i = h̄ | ↓i (2.169)
4
where ml = −l, · · · , l. From these states one can form 2(2l + 1) product states:

|l, ml i ⊗ | ↑i

|l, ml i ⊗ | ↓i (2.170)

However, these states are NOT eigenstates of the total AM, Jˆ2 . We therefore seek eigenstates of
Jˆ2 , Jˆz , L̂2 , Ŝ 2 to be obtained by forming linear combinations of the above product states. To this
end, we will make use of the results proved earlier:
q
L̂± |l, ml i = h̄ (l ± ml + 1)(l ∓ ml )|l, ml ± 1i (2.171)
68 CHAPTER 2. ANGULAR MOMENTUM

and
Jˆ2 = L̂2 + Ŝ 2 + 2L̂z Ŝz + L̂+ Ŝ− + L̂− Ŝ+ (2.172)

Denote the eigenstates of Jˆ2 , Jˆz , L̂2 , Ŝ 2 by |j, mj , li (we could have added a fourth index S = 1
2
but this is understood as we are concerned with electron-spin here). Presumably, j, the quantum
number to Jˆ2 has the values:
1 1
j =l+ , l− (2.173)
2 2
(This would give the right number of states since 2(l + 21 ) + 1 + 2(l − 12 ) + 1 = 2(2l + 1)). First
1
consider the case j = l + 2 and the eigenstate at the top of the ladder:
1 1
|j = l + , mj = l + , li = |l, li ⊗ | ↑i (2.174)
2 2
Applying Jˆz we have:

Jˆz |l, li| ↑i = (L̂z + Ŝz )|l, li ⊗ | ↑i

1
= h̄(l + )|l, li ⊗ | ↑i
2

= h̄mj |l, li ⊗ | ↑i (2.175)

Applying Jˆ2 we have:

3 1
Jˆ2 |l, li| ↑i = h̄2 (l(l + 1) + + (2l) )|l, li ⊗ | ↑i
4 2

1 3
= h̄2 (l + )(l + )|l, li ⊗ | ↑i
2 2

= h̄2 j(j + 1)|l, li ⊗ | ↑i (2.176)

This shows that |l, li| ↑i is an eigenstate of Jˆ2 with the eigenvalues j = l + 12 , mj = l + 21 . To obtain
all other states just apply the lowering operator:

Jˆ− = L̂− + Ŝ− (2.177)

repeatedly to |l, li| ↑i.

Problem 21: Show that

q
Jˆ− |l, li| ↑i = (2l)h̄|l, l − 1i ⊗ | ↑i + h̄|l, li ⊗ | ↓i

1 1 q 1 1
Jˆ− |j = l + , mj = l + , li = (2l + 1)|j = l + , mj = l − , li (2.178)
2 2 2 2
2.4. ADDITION OF ANGULAR MOMENTA 69

Hence, we have:
s s
1 1 2l 1
|j = l + , mj = l − , li = |l, l − 1i ⊗ | ↑i + |l, li ⊗ | ↓i (2.179)
2 2 2l + 1 2l + 1

By repeated application of Jˆ− one obtains the general result (which can be verified by mathematical
induction with respect to mj ):
s s
1 l + mj + 12 1 l − mj + 12 1
|j = l + , mj , li = |l, mj − i ⊗ | ↑i + |l, mj + i ⊗ | ↓i (2.180)
2 2l + 1 2 2l + 1 2

where mj takes half integer values in the range (l + 21 ), · · · , −(l + 12 ). The eigenstates corresponding
1
to j = l − 2 are orthogonal to all the states derived above and where mj = (l − 21 ), · · · , −(l − 21 )
are given by
s s
1 l − mj + 12 1 l + mj + 12 1
|j = l − , mj , li = − |l, mj − i ⊗ | ↑i + |l, mj + i ⊗ | ↓i (2.181)
2 2l + 1 2 2l + 1 2

(This can be verified using the same approach as above).

2.4.3 General case

Two arbitrary AM vectors, Ĵ1 and Ĵ2 have AM eigenstates are given by |j1 , m1 i and |j2 , m2 i, respec-
tively. Form the total AM vector Ĵ = Ĵ1 + Ĵ2 . We then have available two different complete sets of
commuting AM operators: Jˆ2 , Jˆ1z , Jˆ2 , Jˆ2z , with eigenstates |j1 , m1 i|j2 , m2 i = |j1 , m1 , j2 , m2 i, and
1 2
Jˆ2 , Jˆz , Jˆ12 , Jˆ22 , with eigenstates |j, m, j1 , j2 i. Both sets of eigenstates form a complete orthonormal
system, which implies that they can be expanded in terms of each other. For example the expansion
of |j, m, j1 , j2 i in the basis {|j1 , m1 , j2 , m2 i} is:
X
|j, m, j1 , j2 i = |j1 , m1 , j2 , m2 ihj1 , m1 , j2 , m2 |j, m, j1 , j2 i (2.182)
m=m1 +m2

The (real) coefficients hj1 , m1 , j2 , m2 |j, m, j1 , j2 i are called Clebsch-Gordan coefficients. (Note
that because the z-component of the AM is conserved, the only composite states that contribute
to the above sum are those for which m = m1 + m2 ). In a commonly applied terminology, one
refers to |j, m, j1 , j2 i as an eigenfunction in the coupled representation and to |j1 , m1 , j2 , m2 i as an
eigenfunction in the the uncoupled representation. The allowed values that j can take (given j 1 and
j2 ) are
j = j1 + j2 , j1 + j2 − 1, . . . , |j1 − j2 | (2.183)
70 CHAPTER 2. ANGULAR MOMENTUM

with the possible values of m being:

m = m1 + m2 = j, j − 1, . . . , −j (2.184)

A general formula for the Clebsch-Gordan coefficients has been derived by Wigner (using group
theory) and numerical tables of these coefficients can be found in books dealing with atomic spec-
troscopy.
Chapter 3

Approximation Methods For Bound

States

3.1 Introduction

There are very few problems in QM that can be solved exactly. To use QM either to verify or
predict the results of experiments, one needs to resort to approximation techniques such as

• Variational methods

• Perturbative methods

• The JWKB method

3.2 Variational methods

Variational methods (usually used to determine the ground state) involve using a parameterised
trial wave function to represent the ground state wave function. The parameters are optimised
to minimise the ground state energy. Consider a physical system whose Hamiltonian Ĥ is time
independent. Assume that the entire spectrum of Ĥ is discrete and non-degenerate.

Ĥ|ψn i = En |ψn i (n = 0, 1, 2, . . .) (3.1)

The energy spectrum is ordered as follows:

E0 < E 1 < E 2 < . . . (3.2)

71
72 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

3.2.1 Variational theorem

If |Ψi represents an arbitrary state of the system, then:

hΨ|Ĥ|Ψi
hĤi = ≥ E0 (3.3)
hΨ|Ψi

with the equality occurring if and only if |Ψi is the ground state eigenvector of Ĥ with eigenvalue
E0 (the ground state energy).

Proof: Expand |Ψi in the basis of the normalised eigenstates of Ĥ:

X
|Ψi = cn |ψn i (3.4)
n

This implies:

hψm |Ĥ|ψn i c∗m cn

XX
hΨ|Ĥ|Ψi = (3.5)
n m

|cn |2 En
X
=
n

and

hψm |ψn ic∗m cn

XX
hΨ|Ψi = (3.6)
n m

|cn |2
X
=
n

Therefore P 2
n |cn | En
hĤi = P 2
(3.7)
n |cn |
Since E0 < E1 < E2 < . . .,
|cn |2 En ≥ E0 |cn |2
X X
(3.8)
n n
so that
|cn |2
P
E0
hĤi ≥ P n 2 = E0 (3.9)
n |cn |
The equality sign holds when c0 = 1 and cn = 0 ∀ n 6= 0, i.e. when |Ψi = |ψ0 i.

In actual applications to determine the ground state properties of bound systems, one first chooses
a trial wave function ψtrial (r, {α, β, . . .}) which depends on the parameters α, β, . . . etc. We then
calculate the expectation value

hψtrial |Ĥ|ψtrial i
E(α, β, . . .) = (3.10)
hψtrial |ψtrial i
3.2. VARIATIONAL METHODS 73

which (from the variational theorem) is an upper bound to the ground state energy of the system.
(The only restriction on ψtrial (r, {α, β, . . .}) is that it obeys the same boundary conditions as the
eigenstates of the Hamiltonian, Ĥ. Otherwise, Eq. (3.4) is not valid.) We of course choose trial
wave functions that are appropriate for the problem at hand.

We then optimise the parameters α, β, . . . by determining those values of α, β, . . . which minimise

E(α, β, . . .) for that particular trial wave function. This means that we have to solve the following
set of equations (linear or non-linear, depending on the functional form of ψ trial (r, {α, β, . . .}) )

∂
E(α, β, . . .) = 0
∂α

∂
E(α, β, . . .) = 0 (3.11)
∂β

etc. Suppose we denote the values of the parameters that minimise E(α, β, . . .) by (ᾱ, β̄, . . .). Then
the minimum value, E(ᾱ, β̄, . . .) constitutes an upper bound to the exact ground state energy,
while the trial wave function for the optimal values, ψ(r, {ᾱ, β̄, . . .}), is an approximation to the
exact ground state wave function.

Problem 1: Consider
E = hψtrial |Ĥ|ψtrial i

with an arbitrary normalised trial wave function, ψtrial . Show that if ψtrial differs from the correct
ground state wave function, ψ0 , by terms of order , then E as defined above differs from the ground
state energy by order 2 .

Problem 2: Use the variational method to show that a one-dimensional attractive potential will
always have a bound state. (Hint: Consider a square well that will fit inside the potential.)

Problem 3: Use the variational method to estimate the ground state energy for the anharmonic
oscillator
P̂ 2
Ĥ = + λx̂4
2m
Compare your result with the exact result

h̄2
 2/3
1/3
E0 = 1.060λ
2m
(Hint: Use a Gaussian trial wave function.)

Answer: Using s
β 1 2 2
ψtrial (x) = √ e− 2 β x
π
74 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

1 2 2 1
the optimal value of β = (6mλ/h̄2 ) 6 giving Emin = 1.083( 2m
h̄
)3 λ3 .

Problem 4† : A particle of mass m (moving in three dimensions) is bound in the ground state of
the exponential potential
4h̄2 −r/a
V (r) = − e
3ma2
Using
ψtrial (r) = A e(−βr/2a)

as a trial function with β as the variational parameter (A is determined by normalisation), obtain

an upper bound for the ground state energy.
2
h̄
Answer: Optimal value is β = 1 and the upper bound for the ground state energy is − 24ma 2.

Problem 5† : Let E1 and E2 be the ground state energies of a particle of mass m moving in the
attractive potentials V1 (r) and V2 (r) respectively. If V1 (r) ≤ V2 (r) for all r one intuitively expects
E1 ≤ E2 . Use a variational argument to derive this result.

3.2.2 Interlude : atomic units

In this section we deal with a system of units called atomic units that is widely used in atomic,
molecular and solid-state physics. What it boils down to is a set of units such that the following
fundamental quantities, h̄, m, |e|, 4π0 are set to 1. Here m is the electron mass, |e| the magnitude
of its charge, and 0 the permittivity of free space. What this means is that the Hamiltonian for a
hydrogen-like atom reduces from
h̄2 2 Ze2
Ĥ = − ∇ −
2m 4π0 r
to
1 Z
Ĥ = − ∇2 −
2 r
which is independent of any physical constants. It turns out that other physical quantities can be
expressed in terms of the above four basic units (i.e. they are derived units). Most important of
these are the atomic unit of length, called the bohr, defined by

4π0 h̄2
a0 = = 1 (au)
me2
where a0 is the Bohr radius of the hydrogen atom in its ground state. (Note that “au” is an
abbreviation for atomic units). In the same way, a more “natural” unit for energy is the hartree,
defined by
me4
1 hartree = = 1 (au)
16π 2 20 h̄2
3.2. VARIATIONAL METHODS 75

such that the ground state energy of the hydrogen atom is − 21 hartree.

Problem 6† : Given the following SI equivalent quantities

m = 9.1091 × 10−31 kg |e| = 1.6021 × 10−19 C

h̄ = 1.0545 × 10−34 Js 4π0 = 1.1126 × 10−10 C2 J−1 m−1

show that

(a) The atomic unit of length, the bohr is 5.29167 × 10−11 m = 0.529167 Å.

(b) The atomic unit of energy, the hartree is 4.35944 × 10−18 J = 27.211 eV.

(c) Express one hartree in units of kilojoules per mole and wavenumbers. Answer: 2625 kJ mol −1 ,
1
ν= λ = 2.195×105 cm−1 . Note that wavenumbers are used commonly in spectroscopy.

(d) Show that the speed of light in atomic units is approximately equal to 137 au.

3.2.3 Hydrogen molecular ion, H+

2

We now apply the variational method to a “real” problem. The simplest molecule is the H +
2 ion. It
so happens that it is possible to solve the Schrödinger equation for this molecule exactly. (Within
the Born-Oppenheimer approximation, the Schrödinger equation for an electron in the field of
the two fixed protons is separable in elliptic coordinates.) Unfortunately, the solutions are quite
complicated and not easy to use. Instead, we will “solve” this problem approximately via the
variational method, where trial wavefunctions are chosen and the energy is minimised with respect
to some variational parameter (which in this case is the inter-proton distance). The Hamiltonian
for the problem, in atomic units, is

1 1 1 1
Ĥ = − ∇2 − − + (3.12)
2 |r − R1 | |r − R2 | |R1 − R2 |

The first term represents the kinetic energy of the electron, the second and third terms describe
the attractive Coulomb interaction between the electron and protons 1 and 2 with position vectors
R1 and R2 respectively. The last term is the repulsive interaction between the two protons. The
vector r denotes the position of the electron and we define the relative vector R = R 1 − R2 . Its
magnitude R is the distance between the two protons. For the trial wavefunction we use a linear
76 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

combination of 1s hydrogen atom wavefunctions localised at the first (ψ 1 ) and second (ψ2 ) proton.
We have
ψ(r) = αψ1 (r) + βψ2 (r) (3.13)

where

1
ψ1 = √ e−|r−R1 | (3.14)
π

1
ψ2 = √ e−|r−R2 | (3.15)
π

Note that there is symmetry under reflection with respect to a plane lying exactly between the two
protons. The wavefunction therefore has the same symmetry and can be classified by its parity.
For positive parity, α = β while for negative parity, α = −β. This means that
" #
ψ+ (r) = N+ ψ1 (r) + ψ2 (r) (3.16)

" #
ψ− (r) = N− ψ1 (r) − ψ2 (r) (3.17)

where N± are normalisation constants which are determined by the condition hψ± |ψ± i = 1, i.e.
Z
N+2 [2 +2 dr ψ1∗ (r)ψ2 (r)] = 1 (3.18)

Z
N−2 [2 −2 ψ1∗ (r)ψ2 (r)] = 1 (3.19)

By defining the overlap integral U (R),

Z
U (R) = dr ψ1∗ (r)ψ2 (r) (3.20)

where R = |R1 − R2 | is the inter-proton distance, the above equation reduces to

1
N± = p (3.21)
2[1 ± U (R)]

Problem 7: Show that !

1
U (R) = 1 + R + R2 e−R (3.22)
3
Hint: Starting from
1
Z
U (R) = dr e−|r−R1 |−|r−R2 |
π
3.2. VARIATIONAL METHODS 77

make a transformation to relative R = R1 −R2 and center-of-mass coordinates, Rcm = 21 (R1 +R2 ),
and making use of the obvious translational invariance to get
1
Z
1 1
U (R) = dr e−|r− 2 R|−|r+ 2 R|
π
The azimuthal symmetry of the integrand about the axis in the direction of R means we can use
elliptical coordinates to carry out the integral. The transformation equations are given by
" #1 " #1
R 2
R 2
R
x= (ξ 2 − 1)(1 − η 2 ) cos φ y= (ξ 2 − 1)(1 − η 2 ) sin φ z= ξη
2 2 2

where ξ ∈ [1, ∞), η ∈ [−1, 1], φ ∈ [0, 2π). We then have

1 R
|r ± R| = (ξ ± η)
2 2
which gives
1 R3 2π 1 ∞
Z Z Z
U (R) = dφ dη dξ (ξ 2 − η 2 ) e−Rξ
π 8 0 −1 1
Carrying out the integrals gives the required result.

The expectation value of the Hamiltonian with respect to ψ± is given by

hψ1 |Ĥ|ψ1 i + hψ2 |Ĥ|ψ2 i + 2hψ1 |Ĥ|ψ2 i

+ (R) =
2(1 + U )

hψ1 |Ĥ|ψ1 i + hψ2 |Ĥ|ψ2 i − 2hψ1 |Ĥ|ψ2 i

− (R) = (3.23)
2(1 − U )

By symmetry we have hψ1 |Ĥ|ψ1 i = hψ2 |Ĥ|ψ2 i so the above equations reduce to

hψ1 |Ĥ|ψ1 i ± hψ1 |Ĥ|ψ2 i

± (R) = (3.24)
(1 ± U )

Problem 8: Show that

Z
hψ1 |Ĥ|ψ1 i = dr ψ1∗ (r)Ĥψ1 (r)

1 |ψ1 (r)|2
Z
= H
1 + − dr
R |r − R2 |
!
1
= H
1 + 1 + R e−2R (3.25)
R
Z
hψ1 |Ĥ|ψ2 i = dr ψ1∗ (r)Ĥψ2 (r)
78 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

!
1 ψ1∗ (r)ψ2 (r)
Z
= H
1 + U (R) − dr
R |r − R2 |
!
1
= H
1 + U (R) + (1 + R)e−R (3.26)
R

where H
1 is the energy of the 1s state of the hydrogen atom (−0.5 hartree).

Hint: Use elliptical coordinates as in the previous problem to carry out the integrations.

Putting all the various integrals together, we determine ± (R) as a function of R. This is shown in
the sketch below.

Note that the state with positive parity (solid line) is bound since + (R) has a minimum while the
state with negative parity (dashed) has no minimum, i.e. that state is never bound. Experimentally
the binding energy is −2.8 eV with R = 1.06 Åwhile our calculations gives −1.76 eV at the
variational minimum value of R = 1.3 Å, which is not bad for such a crude trial function!

3.2.4 Generalisation: Ritz theorem

The variational theorem is generalised as follows:

Theorem: The expectation value of the Hamiltonian is stationary in the neighbourhood of the
3.2. VARIATIONAL METHODS 79

discrete eigenvalues.

Proof: By stationary, we mean that the change in the value of hĤi when the state vector is changed
by an infinitesimal amount is zero (to first order in the change in the state vector). We need to
show that each stationary expectation value is an eigenvalue of Ĥ. Let

hΨ|Ĥ|Ψi
hĤi = (3.27)
hΨ|Ψi
Note that this is a “functional” of the state vector |Ψi. Consider an infinitesimally small change
to |Ψi:
|Ψi → |Ψi + |δΨi (3.28)

We need to determine the corresponding change to hĤi. From (3.27) we have

hΨ|ΨihĤi = hΨ|Ĥ|Ψi (3.29)

Inserting (3.28) into the above equation we have:

(hΨ| + hδΨ|)(|Ψi + |δΨi)[hĤi + δhĤi] = (hΨ| + hδΨ|)Ĥ(|Ψi + |δΨi) (3.30)

which when expanded out gives:

(hΨ|δΨi + hΨ|Ψi + hδΨ|Ψi + hδΨ|δΨi)[hĤi + δhĤi]

= hΨ|Ĥ|δΨi + hΨ|Ĥ|Ψi + hδΨ|Ĥ|δΨi + hδΨ|Ĥ|Ψi (3.31)

Using (3.28), dropping all terms of second order, after some algebra we finally get:

δhĤihΨ|Ψi = hΨ|(Ĥ − hĤi)|δΨi + hδΨ|(Ĥ − hĤi)|Ψi (3.32)

Thus hĤi is stationary, i.e. δhĤi = 0, provided the right hand side of (3.32) is zero:

hΨ|(Ĥ − hĤi)|δΨi + hδΨ|(Ĥ − hĤi)|Ψi = 0 (3.33)

Suppose we define |Φi = (Ĥ − hĤi)|Ψi. Then (3.33) becomes:

hΦ|δΨi + hδΨ|Φi = 0 (3.34)

This must be satisfied for any |δΨi. Let us choose a |δΨi such that:

|δΨi = δλ|Φi (3.35)

where δλ is a number of first order in small quantities. This implies that (3.34) becomes:

2δλhΦ|Φi = 0 (3.36)
80 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

i.e. that the norm of |Φi equals zero. Therefore, |Φi must be zero, i.e.

(Ĥ − hĤi)|Ψi = 0 (3.37)

This implies that:

Ĥ|Ψi = hĤi|Ψi (3.38)

Therefore the right hand side of (3.27) is stationary if and only if the state vector |Ψi corresponds
to an eigenvector of Ĥ, and the stationary values correspond to the eigenvalues of Ĥ.

3.2.5 Linear variation functions

Representations of eigenstates for all except the simplest systems are complicated functions. In
practice, we expand the arbitrary eigenstate |Ψi as a sum of a finite number (N ) of functions
(whose functional form are chosen depending on the type of system under study) so that:

N
X
|Ψi = ci |ii (3.39)
i=1

and these functions are assumed linearly independent (but not necessarily mutually orthogonal).
For example |ii can be plane waves, or Gaussian functions or a mixture of both, etc. Here c i are
complex numbers that are to be determined. The optimal choice for these linear coefficients, from
the variational theorem, are those that make Ĥ stationary. We have

hΨ|Ĥ|Ψi − EhΨ|Ψi = 0 (3.40)

(We set E = hĤi). Substituting (3.39) into (3.40) yields:

N
N X N
N X
c∗j ci Hji − E c∗j ci Sji = 0
X X
(3.41)
i=1 j=1 i=1 j=1

where

Hji = hj|Ĥ|ii

Sji = hj|ii (3.42)

Differentiating (3.41) with respect to c∗l gives:

N
X
(Hli − E Sli ) ci = 0 (3.43)
i=1
3.2. VARIATIONAL METHODS 81

Cramer’s rule tells us that all the ci ’s are zero unless the determinant of the coefficients vanishes:

|H − E S| = 0 (3.44)

Here H is the N × N matrix whose coefficients are Hji defined above, i.e. it is the matrix repre-
sentation of the Hamiltonian in the basis {|ii}. The N × N matrix S whose coefficients are S ji
is called the overlap matrix. Equation (3.44) is called the secular equation for the energy E
and is an N th order polynomial in E. This yields N real roots, some of which may be degenerate.
Arranging these roots in order of increasing value as

Ẽ0 ≤ Ẽ1 ≤ . . . ẼN −1

we can compare them with the exact spectrum of the system (in order of increasing energy)

E0 ≤ E1 ≤ . . . ≤ EN −1 ≤ EN ≤ . . .

From the variation theorem, we know that

E0 ≤ Ẽ0 (3.45)

Moreover, it can be proved1 that

E1 ≤ Ẽ1 , E2 ≤ Ẽ2 , . . . , EN −1 ≤ ẼN −1 (3.46)

Thus the linear variation method provides upper bounds to the energies of the lowest N eigenstates
of the system. The roots of Eq. (3.44) are used as approximations to the energies of the lowest
eigenstates. Increasing the value of N in (3.39) (which corresponds to increasing the number of
functions to represent the eigenstates) can be shown to increase (or at worst cause no change in)
the accuracy of the previously calculated energies. If the set of functions {|ii} form a complete set,
then we will obtain the exact wave functions of the system. Unfortunately, to have a complete set,
we usually need an infinite number of expansion functions!

To obtain an approximation to the ground state wave function, we take the lowest root Ẽ0 of the
secular equation and substitute it into the set of equations (3.43); we then solve this set of equations
for the coefficients c01 , c02 , . . . , c0N , where the superscript is added to indicate that these coefficients
correspond to Ẽ0 . (As Eq. (3.43) constitutes a set of linear, homogeneous equations, we can only
determine ratios of coefficients; we solve the c01 , c02 , . . . , c0N in terms of c01 , and then determine c01
by normalisation).

Having found the c0i ’s, we take

N
c0i |ii
X
|Ψi =
i=1
1
J.K. MacDonald, Phys. Rev. 43, p.830 (1933); R.H. Young, Int. J. Quantum Chem. 6, p.596 (1972).
82 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

as an approximate ground state wave function. Use of the higher roots of Eq. (3.39) in Eq. (3.43)
gives approximations to the excited-state wave functions (which can be shown to be mutually
orthogonal.) This approach forms the basis for most electronic structure calculations in physics
and chemistry to determine the electronic structure of atoms, molecules, solids, surfaces, etc.

Problem 10† : Let V (x) = 0 for −1 ≤ x ≤ +1 and ∞ otherwise (the “particle in a box” problem).
Use

f1 (x) = (1 − x2 )

f2 (x) = (1 − x4 )

to construct the trial function

2
X
|Ψi = ci fi (x)
i=1
Find the approximate energies and wave functions for the lowest two states and compare your
results with the exact solutions to the problem.

Solution: First construct the overlap (S) and Hamiltonian matrix (H), in the above f 1 , f2 basis.
The overlap matrix elements are:
16 64 128
S11 = , S22 = , S12 = S21 =
15 45 105
and the hamiltonian matrix elements are:
4h̄2 16h̄2 8h̄2
H11 = , H22 = , H12 = H21 =
3m 7m 5m
2 2
Now solve the generalised eigenvalue problem to get the eigenvalues 1.23 h̄m and 12.77 h̄m .
2 2
When we compare this with the exact eigenvalues for the first three states, ≈ 1.23 h̄m , 4.93 h̄m and
2
11.10 h̄m , we find that our calculation gives the upper bounds to the first and third state. (Note
that the estimate for the ground state is very close but NOT equal to the exact value.) Since the
basis functions are even functions it is not suprising that we do not get an estimate for the second
state as this is a odd function!

Problem 10: Consider the one-dimensional infinite well of length L. The Hamiltonian for the
2 2
h̄ d
system is Ĥ = − 2m dx2
+V (x) where V (x) = 0 for x ∈ [0, L] and ∞ otherwise. Find the approximate
energies and wave functions for the lowest four states and compare your results with the exact
solutions to the problem. The linear variation function is given by
4
X
|Ψi = ci fi (x)
i=1
3.3. PERTURBATION METHODS 83

with

f1 (x) = x (L − x)

f2 (x) = x2 (L − x)2

1
f3 (x) = x (L − x) ( L − x)
2

1
f4 (x) = x2 (L − x)2 ( L − x)
2

Answer: You must first note that f1 , f2 are even functions while f3 , f4 are odd. This simplifies
evaulation of integrals when we determine the overlap and hamiltonian matrix elements. We get:

S13 = S31 = S14 = S41 = S23 = S32 = S24 = S42 = 0

and the same for the corresponding hamiltonian matrix elements. The other non-zero matrix
elements are given by:

L5 L9 L7 L7 L11 L9
S11 = , S22 = , S12 = S21 = , S33 = , S44 = , S34 = S43 =
30 630 140 840 27720 5040
and
L3 L7 L5 L5 L9 L7
H11 = H22 = , H12 = H21 = , H33 = , H44 = , H34 = H43 =
6 105 30 40 1260 280
h̄2
where the Hamiltonian matrix elements are in units of m.

The secular determinant reduces to a block diagonal form so that instead of having to evaluate
a 4 × 4 determinant, we have two 2 × 2 determinants to work out, which is much easier! The
h̄2
eigenvalues (in units of mL2
) are ≈ 0.125, 0.500, 1.293 and 2.539. The exact eigenvalues are
0.125, 0.500, 1.125, and 2.000, for the first four states. Note that now we have estimates of all four
states as we have included even and odd functions in our basis set.

3.3 Perturbation methods

3.3.1 Time-independent perturbation theory

Consider a time-independent Hamiltonian, Ĥ (whose spectrum is {En }) where

Ĥ = Ĥ0 + λ Ŵ (3.47)
84 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

(0)
in which effects on the system due to Ĥ0 (the unperturbed Hamiltonian whose spectrum is En )
are the dominant ones while those due to λŴ (the perturbation) are relatively weaker. The real
number λ ∈ [0, 1] parameterises the strength of the perturbation. The perturbation will shift energy
levels and/or split degeneracies. We will distinguish between non-degenerate and degenerate cases.
For the non-degenerate case we shall assume

(0)
1. that both Ĥ and Ĥ0 have discrete spectra so the corresponding eigenstates {|ψn i} and {|ψn i}
form complete, orthonormal sets.

2. there is a one-to-one correspondence between the eigenvalues of Ĥ0 and Ĥ (i.e. quantum
states do not appear or disappear when the perturbation is switched off).

3. the perturbation is “weak” (i.e. it does not alter the eigenstates very much).

(0)
The aim of perturbation theory is to estimate the corrections that need to be made to E n and
(0)
|ψn i to give the required En and |ψn i where:

Ĥ|ψn i = En |ψn i (3.48)

Note that this can be an iterative process (i.e. the approximate solutions obtained can be fed back
into the process to obtain a better approximation).

Assume that the eigenvalues and eigenfunctions of Ĥ can be expanded in a power series in λ:

En = En(0) + λEn(1) + λ2 En(2) + · · ·

|ψn i = |ψn(0) i + λ|ψn(1) i + λ2 |ψn(2) i + · · · (3.49)

Note that:

• This series is not guaranteed to converge. However, in many cases it is an asymptotic expan-
sion i.e. the first few terms nevertheless give reliable results.

An asymptotic expansion of a function f (λ),

m
ak λk + Rm (λ)
X
f (λ) = (3.50)
k=0
is characterised by the following behaviour of the remainder:
Rm (λ)
lim = 0
λ→0 λm

lim Rm (λ) = ∞ (3.51)

m→∞
3.3. PERTURBATION METHODS 85

• There exist cases where En and ψn are NOT expandable in λ. For example, the bound states
of a potential cannot be obtained from the continuum states via PT.

Non-degenerate perturbation theory

Substituting (3.49) into (3.48) we get:

(Ĥ0 + λŴ )(|ψn(0) i + λ|ψn(1) i + λ2 |ψn(2) i + · · ·) = (En(0) + λEn(1) + λ2 En(2) + · · ·)(|ψn(0) i

+ λ|ψn(1) i + λ2 |ψn(2) i + · · ·) (3.52)

Comparing coefficients of λ0 , λ1 , λ2 , . . . we have:

0th Order term

Ĥ0 |ψn(0) i = En(0) |ψn(0) i (3.53)

1st Order term

Ĥ0 |ψn(1) i + Ŵ |ψn(0) i = En(0) |ψn(1) i + En(1) |ψn(0) i (3.54)

2nd Order term

Ĥ0 |ψn(2) i + Ŵ |ψn(1) i = En(0) |ψn(2) i + En(1) |ψn(1) i + En(2) |ψn(0) i (3.55)

Problem 11: Verify Eqs. (3.53) - (3.55).

Without loss of generality, we fix the normalisation of |ψn i by

hψn(0) |ψn i = 1 (3.56)

This implies (for any λ)

λhψn(0) |ψn(1) i + λ2 hψn(0) |ψn(2) i + . . . = 0 (3.57)

which means that

hψn(0) |ψn(1) i = hψn(0) |ψn(2) i = . . . = 0 (3.58)
(0)
If we multiply (3.54) by hψn | and make use of (3.53) we end up with

En(1) = hψn(0) |Ŵ |ψn(0) i (3.59)

86 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

(0)
which is the 1st order correction to the energy of the nth eigenstate. Using the fact that {|ψm i}
form a complete orthonormal set, and making use of the result (3.58) we find:

|ψn(1) i = (0)
X
cm |ψm i (3.60)
m6=n

with
(0) (1)
cm = hψm |ψn i (3.61)
(0) (0)
Multiplying (3.55) by hψm | (which is different from hψn |) we find

cm (En(0) − Em
(0) (0)
) = hψm |Ŵ |ψn(0) i (3.62)
(0)
giving the 1st order correction to the state |ψn i as:
(0)
X hψm (0)
|Ŵ |ψn i
|ψn(1) i = (0) (0)
(0)
|ψm i (3.63)
m6=n En − Em

Problem 12: Prove that the 2nd order correction to the energy is given by:

En(2) = hψm
(0)
|Ŵ |ψn(1) i

(0) (0)
X |hψm |Ŵ |ψn i|2
= (0) (0)
(3.64)
m6=n En − E m

Important points to note are:

(2)
• For the ground state, the 2nd order shift in energy, E0 is negative.

• If the matrix elements of Ŵ are of comparable magnitude, neighbouring levels make a larger
contribution than distant levels, in 2nd order PT.

• When the perturbation is switched on, the energy levels spread apart as if they repel one
another.

Problem 13: The following Hamiltonian matrix has been constructed using an orthonormal basis
1 0 0 0 c 0
   

H = 0 3 0 +c 0 0
   

0 0 −1 0 0 c
Here H = H0 + W and c is a constant.
3.3. PERTURBATION METHODS 87

(a) Find the exact eigenvalues of H.

√ √
Answer: E1 = 2 − 1 + c2 , E2 = 2 + 1 + c2 , E3 = c − 2.

(b) Use perturbation theory to determine the eigenvalues correct to second order in c.
Answer: The energy eigenvalues, up to second order are 1 − 12 c2 , 3 + 21 c2 , c − 2.

(c) Compare the results of steps (a) and (b).

Problem 14: A one-dimensional particle-in-a-box has V (x) = 0 for 0 ≤ x ≤ a and ∞ otherwise.

It is perturbed by the potential
W (x) = −λ sin(πx/a)

for 0 ≤ x ≤ a and 0 otherwise. Calculate the approximate ground state energy using first-order
perturbation theory.

h2 8λ
Answer: 8ma2
− 3π .

Problem 15† : For the same particle-in-a-box, consider a perturbation given by


 − 2bx + b, 0 ≤ x ≤ a/2
a
W (x) =
 2bx − b, a/2 ≤ x ≤ a
a

(0)
Determine the eigenvalues of the perturbed system to first order assuming that b  E 1 where
(0)
E1 is the ground-state energy of the unperturbed system.
(1) (1)
Answer: For n odd, we have En = 21 b − 2b
(nπ)2
while for n even, we have En = 12 b.

Problem 16: Evaluate the first and second order corrections to the energy levels of the “slightly”
anharmonic oscillator with potential energy
1
V (x) = mω 2 x̂2 + αx̂3 + β x̂4
2
where the perturbation is
W (x) = αx̂3 + β x̂4

(Hint: Use the results of Chapter 1 Problem 17.)

Answer:
2 3
1 3 h̄ 1 α2 h̄
   
En = n+ h̄ω + β(2n2 + 2n + 1) − (30n2 + 30n + 11)
2 4 mω 8 h̄ω mω
4
1 β2 h̄

− (34n3 + 51n2 + 59n + 21)
8 h̄ω mω
88 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

Perturbation theory for degenerate states

The formulae derived above break down when there is degeneracy in the energy levels of the un-
perturbed Hamiltonian, Ĥ0 , i.e. when there are linearly independent eigenfunctions corresponding
to the same eigenvalue. When this happens, some of the denominators in the 1st order correc-
tion to the eigenfunction or the 2nd order correction to the energy eigenvalues vanish, leading to
divergences. Below we work out a scheme to fix this.
(0)
Suppose the eigenvalue En is g−fold degenerate, i.e. there are g linearly independent eigenfunc-
tions, {|φnα i}, α = 1, . . . , g belonging to this eigenvalue. The {|φnα i} can be chosen to be or-
thonormal.

Ĥ0 |φnα i = En(0) |φnα i

hφnα |φnβ i = δαβ (α, β = 1, . . . , g) (3.65)

Let g
(0) X
|ψni i = ciα |φnα i (i = 1, . . . , g) (3.66)
α=1
where the ciα are constants to be determined.

Then we want to find

(1) (2)
Eni = En(0) + λEni + λ2 Eni + · · ·

(0) (1) (2)

|ψni i = |ψni i + λ|ψni i + λ2 |ψni i + · · · (i = 1, . . . , g) (3.67)

such that
Ĥ|ψni i = (Ĥ0 + λŴ )|ψni i = Eni |ψni i (3.68)

In general, the states |ψni i, i = 1, . . . , g will have different energies Eni , i.e. the effect of the
(0)
perturbation Ŵ is to split the g−fold degenerate level Eni into several groups of levels (at most
g of them). Note however that it is possible that Ŵ only shifts the levels without splitting them.
As stated earlier, the perturbation expansions for the states, in the presence of degeneracies in the
spectrum of Ĥ0 , will in general contain terms with zero denominators. Hence these expansions will
only make sense if these terms also have vanishing numerators at the same time. So we demand
that
(0) (0) (0) (0)
hψni |Ŵ |ψnj i = hψni |Ŵ |ψni iδij (i, j = 1, . . . , g) (3.69)
(0) (0)
i.e. the g × g matrix whose elements are hψni |Ŵ |ψnj i must be a diagonal matrix. Our earlier
formula (3.54) becomes:
(1) (1) (0)
(Ĥ0 − En(0) )|ψni i = (Eni − Ŵ )|ψni i (3.70)
3.3. PERTURBATION METHODS 89

(0)
Multiplying the above equation by hψni | gives

(1) (0) (0) (0)

Eni hψni |ψni i = hψni |Ŵ |ψni i (3.71)

Again fixing the normalisation such that

(0)
hψni |ψni i = 1 (3.72)

we have the first order correction as:

(1) (0) (0)

Eni = hψni |Ŵ |ψni i (3.73)

To determine the coefficients cni , starting from (3.70) and multiplying by hφnβ | we get:
   
(1) (1) (0)
hφnβ | Ĥ0 − En(0) |ψni i = hφnβ | (Eni − Ŵ ) |ψni i

(0) (1) (0)

=⇒ hφnβ |Ŵ |ψni i = Eni hφnβ |ψni i

g g
X (1) X
=⇒ ciα hφnβ |Ŵ |φnα i = Eni ciα hφnβ |φnα i (3.74)
α=1 α=1

which reduces to:

g  
X (1)
ciα hφnβ |Ŵ |φnα i − hφnβ |φnα iEni =0 (3.75)
α=1

Using the orthonormality of the {|φnα i}, the above equation reduces to:
g  
X (1)
ciα hφnβ |Ŵ |φnα i − δαβ Eni =0 (3.76)
α=1

This is just a set of homogeneous equations for the g unknown constants ciα . For non-trivial
solutions we must have:  
hφnβ |Ŵ |φnα i − δαβ E (1)  = 0
 
 ni  (3.77)

This determinant yields a g−degree polynomial (with g roots, not necessarily all different). These
(0)
roots correspond to the 1st order correction to the energy En and the values of the ciα can be
determined by substituting the roots back into (3.76).

Problem 17: Given the Hamiltonian matrix H = H0 + W constructed using orthogonal basis
functions where
20 0 0 0 1 0
   

H= 0 20 0  + 1 0 2
   

0 0 30 0 2 0
90 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

(a) Determine the exact eigenvalues.

Answer: The eigenvalues are 20.806, 18.805, and 30.389.

(b) Determine the eigenvalues correct to second order in the perturbation.

Answer: The eigenvalues, correct to second order, are 20.777, 18.818, and 30.404.

Problem 18: A particle of mass m moves in a two dimensional potential box where V (x, y) = 0
when 0 ≤ x ≤ a and 0 ≤ y ≤ a, otherwise V (x, y) = ∞.

(a) Determine the four lowest eigenvalues and display them on a diagram. Indicate the degen-
eracies.
h2 2 h2 2
Answer: , 5h
4ma2 8ma2
(×2), , 5h
ma2 4ma2
(×2).

(b) When the perturbation W (x) = bx2 is applied, the degeneracies are lifted. Assume that
Ŵ  Ĥ0 and determine the eigenvalues of the perturbed system to first order. Illustrate the
shifts on the diagram in part (a).
       
h2 1 1 5h2 1 5h2 h2
Answer: 4ma
2
2 +a b 3 − 2π 2 , 8ma2 +a b
2
3− 2π1 2 , 8ma2
+a2 b 1
3 − 1
8π 2
, ma2
+a2 b 1
3 − 1
8π 2
,
  
5h2 5h2
4ma2
+ a2 b 13 − 2π1 2 , 4ma 2 1 1
2 + a b 3 − 18π 2

Problem 19: A particle of mass m is constrained to move on the x−y plane so that the Hamiltonian
is:
1 1
Ĥ = (p̂2x + p̂2y ) + k(x̂2 + ŷ 2 ) + ax̂ŷ
2m 2

(a) Solve the problem for the case when a = 0. (This is just the two-dimensional harmonic
oscillator.)

(b) List the first few energy levels for the two-dimensional oscillator and give the degeneracies.

(c) Use degenerate perturbation theory to determine the energy splitting for the lowest degenerate
states and the first order corrections to the wave functions for these states when a 6= 0.

Answer: This problem will be done in class.

3.3.2 Time-dependent perturbation theory

In general, if the potential energy function is independent of time (V (r, t) = V (r)), then the
time-dependent Schrödinger equation:
∂Ψ
ĤΨ = ih̄ (3.78)
∂t
3.3. PERTURBATION METHODS 91

is solved by separation of variables:

Ψ(r, t) = ψ(r)e−iEt/h̄ (3.79)

with ψ(r) the solution of the time-independent Schrödinger equation:

Ĥψ(r) = Eψ(r) (3.80)

These are called stationary states since all probabilities and expectation values calculated from
them are constant in time. To allow for transitions between energy levels, we must have a
time-dependent potential. Furthermore, if the time dependent part of the Hamiltonian is “small”
compared to the time-independent part, it can be treated as a perturbation.

Starting form the Hamiltonian,

Ĥ = Ĥ0 + Ŵ (r, t) (3.81)

we seek solutions of the time-dependent Schrödinger equation:

∂|Ψi
ih̄ = Ĥ|Ψi
∂t
 
= Ĥ0 + Ŵ |Ψi (3.82)

Expanding |Ψi in the orthonormal basis of Ĥ0 : {|ψn (r)i}, we have:

e−iEk t/h̄ ck (t) |ψk i

X
|Ψ(r, t)i = (3.83)
k

where
Ĥ0 |ψk i = Ek |ψk i (3.84)

The quantity |ck (t)|2 is the probability of finding the system in state |ψk i at time t. Substituting
(3.83) into (3.82) we get:

∂|Ψi dck (t)

Ek e−iEk t/h̄ ck (t) |ψk i + ih̄ e−iEk t/h̄
X X
ih̄ = |ψk i
∂t k k
dt
 
Ĥ|Ψi = Ĥ0 + Ŵ |Ψi

Ek e−iEk t/h̄ ck (t)|ψk i + e−iEk t/h̄ ck (t) Ŵ |ψk i

X X
= (3.85)
k k

This gives:
dck (t)
e−iEk t/h̄ e−iEk t/h̄ ck (t) Ŵ |ψk i
X X
ih̄ |ψk i = (3.86)
k
dt k
92 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

Multiplying (3.86) by hψi | and using the orthonormality property we finally get:

dci (t) X
ih̄ = Wik eiωik t ck (t) (3.87)
dt k

where

Wik (t) = hψi (r)|Ŵ |ψk (r)i

Ei − E k
ωik = (3.88)
h̄

Note that (3.87) can be written as:

dci (t) i
Wik eiωik t ck (t)
X
= − Wii ci (t) + (3.89)
dt h̄ k6=i

Then, if Wik = 0 ∀i 6= k the above equation is easily solved:

ci (t) = ci (0) e−iWii t/h̄ (3.90)

which indicates that |ci (t)|2 = |ci (0)|2 , i.e. no transitions from the initial state i to other states
take place. Therefore for transitions to take place the perturbation must couple the
different eigenstates of Ĥ0 .

We now develop the perturbation expansion of ci (t). Assume that the system is in state i at
time t = 0. Again introducing the parameter λ to isolate the various orders in the perturbation
expansion we have:

(0) (1) (2)

ci (t) = ci (t) + λci (t) + λ2 ci (t) + . . .

Ŵ (t) → λŴ (t) (3.91)

Assume that the system is in state m at time t = 0:

(0)
ci (0) = δim = ci (0) (3.92)

which also implies

(1) (2)
ci (0) = ci (0) = . . . = 0 (3.93)

Substituting (3.91) into (3.87) and isolating the coefficients of the various powers in λ, we have:
3.3. PERTURBATION METHODS 93

0th Order Term

(0)
dci (t)
= 0
dt
(0)
=⇒ ci (t) = δim (3.94)

In 0th order, there are no transitions.

1st Order Term

(1)
dci (t) iX (0)
= − Wik eiωik t ck (t)
dt h̄ k

i
= − Wim eiωim t
h̄

i
Z t
(1) 0
=⇒ ci (t) = − dt0 Wim (t0 ) eiωim t (3.95)
h̄ 0

This procedure can be carried on for higher order terms. What is the probability that the system,
initially in state m, has made a transition to a state n by time t after the perturbation was switched
on? This transition probability is denoted by Pmn where:
 2
Pmn (t) = hψn |Ψ(r, t)i = |cn (t)|2
 
 (3.96)

To 1st order we therefore have

t
2
1
 Z
0 0 iωnm t0 

Pmn (t) = 2  dt Wnm (t ) e (3.97)
h̄ 0


Fermi’s golden rule

We apply the formula for Pmn to study transitions into a continuous spectrum of final states.
Examples include α−decay (where the final states, characterised by the momenta of the emitted
α−particles, lie in a continuum) and optical transitions (a system in an excited state makes a
transition to a lower state by emitting a photon).

Consider a perturbation that is switched on at time t = 0 and subsequently remains unchanged.

Ŵ (r, t) = V (r) Θ(t) (3.98)

94 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

where Θ(t) is the unit step function defined by:

Θ(t) = 1 t > 0

= 0 t<0 (3.99)

The transition probability Pmn now becomes:

t
2
1 
Z
0 iωnm t0 

Pmn (t) = dt Vnm e
h̄2  0


|Vnm |2  eiωnm t − 1 2
 
=
h̄2  ωnm 
2
|Vnm |2 sin(ωnm t/2)

= (3.100)
h̄2 ωnm /2
Using the following definition of the Dirac delta function:
sin2 (αt)
lim = δ(α) (3.101)
t→∞ πα2 t

For long times, (3.100) becomes:

2π 2
Pmn (t) = t 2 δ(ωnm ) |Vnm |
h̄

2π
= t δ(En − Em ) |Vnm |2 (3.102)
h̄
The transition rate Γnm is defined as the transition probability per unit time and is given by:
Pmn
Γmn =
t

2π
= δ(En − Em ) |Vnm |2 (3.103)
h̄
Since we are treating transitions into the continuum of final states, the transition rate to a group
of final states (whose energies are clustered around En ) is of interest. Assume that the matrix
elements Vnm for all of these nearby final states are equal. Define ρf (En ) as the density of final
states such that ρf (En )dEn gives the number of final states in the interval dEn about En . The
transition rate to this set of states is then:
2π
Z
ρf (Em ) |Vnm |2
X
Γ= Γmn = ρf (En )dEn Γmn = (3.104)
n h̄

Now consider the case of periodic perturbations where the perturbation has been switched on
at t = 0 and varies periodically with time. (An example is a periodic external field such as a EM
wave). In general:
3.3. PERTURBATION METHODS 95

 
Ŵ (r, t) = F̂ e−iωt + F̂ † e+iωt Θ(t) (3.105)

where F̂ is an operator (independent of time). Substituting (3.105) into (3.97) gives:

1  t 0 i(ωnm −ω)t0 2
 Z 
i(ωnm +ω)t0 †
Pmn (t) = dt {e F nm + e F }
nm 

h̄2  0

Fnm = hψn |F̂ |ψm i

†
Fnm = hψn |F̂ † |ψm i (3.106)

For t → ∞ we have
2π
 
Pmn (t) = t |Fnm |2 δ(ωnm − ω) + |Fnm
† 2
| δ(ωnm + ω) (3.107)
h̄2
Therefore the transition rate becomes
2π
 
2 † 2
Γmn = |Fnm | δ(En − Em − h̄ ω) + |Fnm | δ(En − Em + h̄ ω) (3.108)
h̄

Problem 20: Verify (3.107). Note that the two delta functions do not overlap so the cross terms
in the square of the modulus do not contribute.

Problem 21: A hydrogen atom is placed in an electric field E(t) that is uniform and has the time
dependence 
 0 t<0
E(t) =
 E e−γt t>0
0

What is the probability that as t → ∞, the hydrogen atom, if initially in the ground state, makes
a transition to the 2p state?

Answer: The transition probability for 1s → 2p is

|hψ210 |z|ψ100 i|2

e2 E02
(E210 − E100 )2 + h̄2 γ 2

Problem 22: Consider a harmonic oscillator described by

1 2 1
Ĥ = p̂ + mω 2 (t)x̂2
2m x 2
where
ω(t) = ω0 + δω cos(f t)
96 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

and δω  ω0 . Calculate the probability that a transition occurs from the ground state, as a
function of time, given the system is in the ground state at t = 0.

Answer: To be done in class.

Problem 23: Consider a two-level system with E1 < E2 . There is a time-dependent potential
that connects the two levels as follows:

V11 = V22 = 0, V12 = γeiωt , V21 = γe−iωt (3.109)

where γ is real. At t = 0, it is known that only the lower level is populated — that is, c 1 (0) =
1, c2 (0) = 0.

(a) Find |c1 (t)|2 and |c2 (t)|2 for t > 0 by exactly solving the coupled differential equation
2
Vkn (t)eiωkn t cn
X
ih̄ċk =
n=1

for k = 1, 2.

(b) Do the same problem using time-dependent perturbation theory to lowest non vanishing order.
Compare the two approaches for small values of γ. Treat the following two cases separately:
(i) ω very different from ω12 and (ii) ω close to ω12 .

3.4 JWKB method

The JWKB (Jeffreys, Wentzel, Kramers, Brillouin) method is a semi-classical technique for ob-
taining approximate solutions to the one-dimensional Schrödinger equation. It is mainly used in
calculating bound-state energies and tunnelling rates through potential barriers, and is valid in the
h h
limit λ = p = mv → 0 or h̄ → 0 or m → ∞ where m is the mass of the particle, p its momentum
etc.

The key idea is as follows. Imagine a particle of energy E moving through a region where the
potential V (x) is constant. If E > V , the wave function is of the form

ψ(x) = A e±ikx
p
2m(E − V )
k =
h̄
The plus sign indicates particles travelling to the right etc. The wave function is oscillatory, with
constant wavelength λ = 2π/k, and has constant amplitude, A. Consider now the case where
3.4. JWKB METHOD 97

V (x) is not a constant but varies rather slowly in comparison to λ (so that in a region containing
many wavelengths the potentially is essentially constant). Then it is reasonable to suppose that
ψ remains practically sinusoidal except that the wavelength and the amplitude change slowly with
x. This is the central theme of the JWKB method: rapid oscillations are modulated by gradual
variation in amplitude and wavelength.

Similarly, if E < V (with V a constant), then ψ is exponential:

ψ(x) = A e±Kx
p
2m(V − E)
K =
h̄
Now, if V (x) is not constant but again varies slowly in comparison to 1/K, the solution remains
practically exponential except that A and K are now slowly varying functions of x.

There are of course places where this idea breaks down, e.g. in the vicinity of a classical turning
point where E ≈ V . Here, λ (or 1/K) goes to infinity and V (x) can hardly be said to vary “slowly”!
Proper handling of this is the most difficult aspect of the JWKB approximation but the final results
are simple and easy to implement.

3.4.1 Derivation

We seek to solve
d2 ψ
+ k 2 (x)ψ(x) = 0
dx2

2m
k 2 (x) = (E − V (x)) (3.110)
h̄2
The semi-classical limit corresponds to k large. If k were constant, then of course the solutions
would just be e±ikx . This suggests that we try ψ(x) = eiS(x) , where in general S(x) is a complex
function. Then,
dψ
= iS 0 eiS
dx

d2 ψ
= (iS 00 − S 02 )eiS (3.111)
dx2
and the Schrödinger equation reduces to (iS 00 − S 02 + k 2 )eiS = 0, or
q
S 0 = ± k 2 (x) + iS 00 (x)
q
= ±k(x) 1 + iS 00 (x)/k 2 (3.112)
98 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

(Note that if k were a constant, S 00 = 0 and S 0 = ±k.)

We now attempt to solve the above equation by iteration, using S 0 = ±k as the first guess, and as
a second guess we use:
q
S 0 = ±k 1 ± ik 0 (x)/k 2

i k 0 (x)
 
≈ ±k 1 ±
2 k2

i k 0 (x)
≈ ±k + (3.113)
2 k

where we have assumed that the corrections are small. Then, we have

dS i k0
= ±k +
dx 2k
x i x k0
Z Z
S(x) ∼ ± k(x)dx + dx + c (3.114)
2 k

The second integral is a perfect differential (d ln k), so

Z x i
S(x) = ± k(x)dx + ln k + c
2

ψ = eiS
Rx
k(x)dx − 21 ln k
= C e±i e

C Rx
= p e±i k(x)dx (3.115)
k(x)

k0 λ dk
Note that in making the expansion, we have assumed that k2
 1 or 2π dx  k, i.e. that the change
h̄2 k2
in k in one wavelength is much smaller than k. Alternatively, one has λ dV
dx  m so that the
change in V in one wavelength is much smaller than the local kinetic energy.

Note that in the classically forbidden regions, k 2 < 0, one puts k = iK(x) and carries through the
above derivation to get

C Rx
ψ(x) = p e± K(x)dx
K(x)

2m
K2 = (V − E) > 0 (3.116)
h̄2
3.4. JWKB METHOD 99

3.4.2 Connection formulae

In our discussion above, it was emphasised that the JWKB method works when the “short wave-
length approximation” holds. This of course breaks down when we hit the classical turning points
where k 2 (x) = 0 (which happens when E = V ). To overcome this problem, we will derive below
equations relating the forms of the solution to both sides of the turning point.

If the potential can be approximated by an increasing linear potential near the turning point x = a
(the region x > a being classically forbidden), we can write in the vicinity of the turning point

2m ∂V
 
2
k (x) = 2 − (x − a) (3.117)
h̄ ∂x x=a

(If we have a potential which cannot be approximated linearly, we must resort to approximations
with a quadratic term and find a solution in terms of parabolic cylinder functions. We will not go
into the details of this case but you can look it up if you are interested.) The Schrödinger equation
near the turning point becomes

∂V 2m
 
ψ 00 − (x − a)ψ = 0 (3.118)
∂x x=a h̄2
This is a linear potential problem which is solved in terms of Airy functions. If we let

y = α(a − x) ≥ 0

2m ∂V
α3 = ≥0 (3.119)
h̄2 ∂x
then the above differential equation becomes

ψ 00 (y) + y ψ(y) = 0 (3.120)

whose solutions are:

√ √
ψ = A yJ− 1 (z) + B yJ+ 1 (z)
3 3

2 3
z = y2
3
Z a
= k(x) dx (3.121)
x

The procedure is now to make asymptotic expansions for the Bessel functions, match them onto the
JWKB solutions in the classically allowed and forbidden regions and thus obtain formulae relating
the solutions in the two regions.
100 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

For the case y → ∞, i.e. x  a, deep inside the allowed region, we have
r
2 νπ π
 
Jν → cos z − −
πz 2 4
r
2 √ π π √ π π
    
ψ → A y cos z + − + B y cos z − − (3.122)
πz 6 4 6 4
i.e. it oscillates as does the JWKB solution in the left region.

If y → 0, i.e. when x → a, near the turning point, we have

 ν
z
2
Jν ∼
Γ(ν + 1)
√ √
A y( 12 z)−1/3 B y( 12 z)1/3
ψ → + (3.123)
Γ( 23 ) Γ( 43 )

k2 Ra 2 k3
Since we know that y = α and z = x k(x) dx = 3 α3/2 , these can be written as

3 α1/4 a
r
π
 Z 
ψlef t (y → ∞; x  a) → p A cos k(x) dx −
π k(x) x 12
Z a 5π

+ B cos k(x) dx +
x 12

A 31/3 By
ψlef t (near x = a) → 2 + 4 1/3 (3.124)
Γ( 3 ) Γ( 3 ) 3

We now follow a similar procedure in the classically forbidden region, x > a. Let y = α(x − a) > 0.
The Schrödinger equation now becomes ψ 00 − yψ = 0, which has solutions:
√ √
ψ = C y I 1 (z) + D y I− 1 (z)
3 3

Z x
z = K(x) dx
a

2m
 
K 2 (x) = V (x) − E (3.125)
h̄2
The I 0 s are Bessel functions of imaginary argument.

When the same steps described in detail above are followed, we find

3 C α1/4 R x K(x) dx
r  Rx 
5π
ψright (x  a) → p e a + e− a K(x) dx e−i 6
4π K(x)
3.4. JWKB METHOD 101

3 D α1/4 R x K(x) dx
r  Rx 
5π
+ p e a + e− a K(x) dx ei 6
4π K(x)

D 31/3 Cy
ψright (near x = a) → 2 + 4 1/3 (3.126)
Γ( 3 ) Γ( 3 ) 3

We now match these two sets of solutions near x = a. Matching the functions at y = 0 gives us
D = A; matching derivatives gives us B = −C. If we let A = 1, B = 0, we find that the cosine
solutions to the left correspond to dying exponentials to the right. If we manipulate the constant
factors in the above asymptotic solutions, we finally find the connection formulae:

2
Z a π

1 Rx
p cos k(x) dx − ⇐⇒ p e− a K(x) dx
k(x) x 4 K(x)

1
Z a π

1 Rx
p sin k(x) dx − ⇐⇒ − p e a K(x) dx (3.127)
k(x) x 4 K(x)

Rx
This allow a continuation of solutions. A sine solution on the left matches into a −e + a
K(x) dx

solution on the right. Similar formulae for the reverse situation with the turning point on the left
(x < b classically forbidden) give

1 Rb
2
Z x π

p e− x K(x) dx ⇐⇒ p cos k(x) dx −
K(x) k(x) b 4
Z x
1 1 π
Rb 
p e x K(x) dx ⇐⇒ − p sin k(x) dx − (3.128)
K(x) k(x) b 4

Problem 24: Verify Eq. (3.127).

3.4.3 *JWKB treatment of the bound state problem

Given an arbitrary potential V (x), we wish to find the approximate eigenstates. Effectively, this
means that we must find energies such that the JWKB solutions in the potential well match onto
dying exponentials in the classically forbidden regions (i.e. ψ → 0 as x → ± ∞). Consider a
potential as shown in the figure below:
102 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

E
b a

Far to the left, we will have (as x → −∞):

const R x K(x) dx
ψlef t = p e b →0 (3.129)
K(x)
Inside the potential well, this matches to:
2 · const
Z x π

ψinside = p cos k(x) dx − (3.130)
k(x) b 4
We now move along to the next turning point x = a. Rewriting ψ inside (for easy matching) as:
2 · const
Z a Z a π

ψinside = p cos k(x) dx − k(x) dx − (3.131)
k(x) b x 4
We define Z a
φ= k(x) dx (3.132)
b
and applying the connection formulae developed in the previous section, we get
2 · const
 Z a π π

ψinside = p cos − k(x) dx + φ − +
k(x) x 2 4

2 · const

π
  Z a π

= p cos φ − cos k(x) dx −
k(x) 2 x 4

π
  Z a π

− sin −φ sin k(x) dx − (3.133)
2 x 4

The cos solution matches onto the dying exponential, but the sine matches onto a growing expo-
nential. Thus, its coefficient must be zero; i.e. sin( π2 − φ) = 0 or π
2 − φ = −nπ or
Z a 1
k(x) dx = (n + )π (3.134)
b 2
3.4. JWKB METHOD 103

1
This is similar to the Bohr-Sommerfeld Quantization Rule, but with n shifted by 2. This
matching may need to be done many times if we have a complicated potential such as:

Problem 25† : Show that Eq. (3.134) gives the exact energy levels of a simple harmonic oscillator.

3.4.4 Barrier penetration

Given a potential like that shown below:

a b

we know that classically there is no probability of getting through the barrier. However, from
quantum mechanics we find that an incoming wave function of magnitude A will give rise to a
reflected wave function of magnitude B, and a transmitted wave function of magnitude F such
that
A Rx
B Rx
ψlef t = √ ei a k(x) dx + √ e−i a k(x) dx
k k

C Rx
D Rx
ψmiddle = √ e− a K(x) dx + √ e a K(x) dx
K K
104 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

F Rx
G Rx
ψright = √ ei b k(x) dx + √ e−i b k(x) dx (3.135)
k k
Obviously we will set G = 0 at the end (there is no wave coming in from the right). First match
at the turning point x = a. We do this by rewriting ψlef t in such a form as to use the connection
formulae:
1
Z a π

−i π4 i π4
ψlef t = (A e + B e ) √ cos k(x) dx −
k x 4

1
Z a π

−i π4 i π4
+ i (−A e + B e ) √ sin k(x) dx − (3.136)
k x 4
On applying the connection formula, we find:
π π
A e−i 4 + B ei 4 = 2C

π π
−i B ei 4 + i A e−i 4 = D (3.137)

We do the same at the next turning point (x = b). Rewrite ψmiddle as

C Rb
D Rb
ψmiddle = √ Θ−1 e x K(x) dx + √ Θ e− x K(x) dx
K K
Rb
K(x) dx
Θ = e a (3.138)

which on matching yields:

π π
F ei 4 + G e−i 4 = 2DΘ

π π C
i (F ei 4 − G e−i 4 ) = − (3.139)
Θ
These relations can be written in a simple matrix notation as:
π π
  ! 
C 1 e−i 4 ei 4 A
  = π π
 
D 2 i 2 e−i 4 −i 2 ei 4 B

π π
  ! 
C −i Θ ei 4 i Θ e−i 4 F
  = 1 π π
  (3.140)
D 2Θ ei 4 1
2Θ e−i 4 G

After some matrix algebra we get

   
A F
  = T  
B G

1 1
!
1 2Θ+ 2Θ i (2 Θ − 2Θ )
T = 1 1
(3.141)
2 −i (2 Θ − 2Θ ) 2Θ+ 2Θ
3.4. JWKB METHOD 105

1 1
where the matrix T is called the Transfer matrix. If we let G = 0, then we find A = 2 (2 Θ+ 2Θ )F
or

 2 2
F  1

 
 
A = 
Θ + 1

 (3.142)
4Θ

∼ Θ−2 for large Θ.

Therefore for large Θ the transmission coefficient will be

 2
F  Rb
T ≡   ∼ Θ−2 = e−2 a K(x) dx (3.143)
A

Problem 26: Verify Eq. (3.142).

Problem 27: Consider a particle with total energy E = 0 moving in the anharmonic oscillator
potential

1
V (x) = mω 2 (x2 − x4 )
2

Show that, for small , the transmission coefficient of the potential barrier between x = 0 and
√
x = 1/  is

T ∼ e−2mω/3h̄ .

(Notice that this result cannot be obtained using perturbation theory, since the function on the
r.h.s. does not have a power series expansion in .)

3.4.5 Alpha decay of nuclei

The potential that the alpha particle sees is shown below:

106 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES

R r

We assume that Z is the atomic number of the final nucleus, R is the nuclear radius and the
potential is given by:
2Ze2
V (r) = r>R (3.144)
r
i.e. the potential is Coulombic outside the nucleus. Let the energy of the alpha particle be E and
2Ze2
define r0 = E . We need to evaluate
1
1 r0 2Ze2
Z 
2
I = √ −E dr
E R r

Z γR  γR 1
2
= −1 dr
R r

2Ze2
γ = (3.145)
ER
Substitute
r
= sin2 θ
γR

dr = 2γR sin θ cos θ dθ

1
θ0 = sin−1 √ (3.146)
γ
3.4. JWKB METHOD 107

into the above integral to get

Z π
2
I = 2γR cos2 θ dθ
θ0

π 1 (γ − 1)2
 
= γR − sin−1 √ − (3.147)
2 γ γ
Rb
−2 K(x) dx 2m
Since the transmission coefficient is given by T ∼ e a where K 2 (x) = h̄2
(V (r) − E) we
get √
2m √
ln T = −2 I E (3.148)
h̄
2Ze2
For the case that E  R , γ  1 and I ∼ γR[ π2 − √1 ],
γ so we get

T = A e−β
√ 
2m 2Ze2

β = π √
h̄ E
4
√
Ze2 mR
A = e− h̄ (3.149)

Every time the alpha particle is incident on the barrier, it has a probability T to escape. To find
the probability per unit time (Γ) to escape, we multiply by ω, the frequency of oscillation of an
alpha particle in the well of radius R, i.e.

Γ = ω A e−β (3.150)

We estimate ω as follows:
v
ω '
R

∼ ∆p
=
mR

∼ h̄
= (3.151)
2mR2
182
For R = 10 fm, one finds ω = 1020 sec, taking Z = 90, then A = 6 × 10−19 and β = √
E
for E in
MeV. So we have
√
Γ ∼ 0.6 e−182/ E
× 1040 sec−1 (E in MeV) (3.152)

Note that although our approximation is crude (since it is not really true that γ  1), from the
above equation, one finds a very strong variation in Γ as a function of E. For example, doubling
E from 4 to 8 MeV increases the value of Γ by eleven orders of magnitude!
108 CHAPTER 3. APPROXIMATION METHODS FOR BOUND STATES
Chapter 4

Scattering Theory

4.1 Introduction

The path of a moving (incident) particle is altered when it interacts with another (target) particle,
such as an atom or a molecule. Phenomena of this sort are generally called scattering. Scattering
is called elastic when the identities and internal properties of the incident particle and the target
remain unchanged after the collision, and inelastic when the identities or internal properties change
or when other particles are emitted or the two particles form a bound state. Analyses of scattering
give information on the structure and interactions of atoms, molecules, and elementary particles.
We first review the solution to the time independent Schrödinger equation for the spherically
symmetric square well to develop the concepts for the general case of scattering from an arbitrary
potential.

4.2 Spherically symmetric square well

The potential is defined by: (

−V0 r ≤ a;
V (r) = (4.1)
0, r > a.
where V0 > 0. An incident particle will feel an attractive potential when it is within the spherical
region of radius r ≤ a. Note that this potential is spherically symmetric i.e. it depends only on the
distance from the origin. The time independent Schrödinger equation is:

h̄2 2
 
− ∇ + V (r) Ψ(r) = EΨ(r) (4.2)
2m

109
110 CHAPTER 4. SCATTERING THEORY

Since we have spherical symmetry, we use the method of separation of variables to write Ψ(r) as:

Ψ(r) = Rl (r) Ylm (θ, ϕ) (4.3)

where the Ylm (θ, ϕ) are the Spherical Harmonics and the Rl (r) are the solutions to:

h̄2 d2 2 d l(l + 1)
   
− 2
+ − + V (r) Rl (r) = E Rl (r) (4.4)
2m dr r dr r2
Suppose the energy of the incident particles E are such that (E > V ∀r). Define
q
k= 2m(E − V )/h̄

Then (4.4) becomes:

d2 2 d l(l + 1)
 
2
+ − + k 2 Rl (r) = 0 (4.5)
dr r dr r2
Substituting ρ = kr into (4.5) gives

d2 2 d l(l + 1)
 
2
+ − + 1 Rl (ρ) = 0 (4.6)
dρ ρ dρ ρ2

The general solution to (4.6) are the spherical Bessel functions, jl (ρ) and the spherical Neu-
mann functions nl (ρ):
Rl (ρ) = Ajl (ρ) + Bnl (ρ) (4.7)

(The mathematical properties of these functions are discussed in the next section).

When there is no potential (free particle case), the solution to (4.6) which is regular at the origin,
involves only the spherical Bessel function:

Rl (r) = Ajl (kr) (4.8)

√
where now, k = 2mE/h̄. The asymptotic behaviour of this solution far from the origin is:

sin(kr − lπ/2)
Rl (r) r −→ A (4.9)
→∞
kr
If we now have the potential (4.1) acting, with E > 0 then the solution to (4.6) take the form:
(
Ajl (qr) r < a;
Rl (r) =
Bjl (kr) + Cnl (kr), r > a.

where A, B and C are constants to be determined from the boundary conditions and
√
k = 2mE/h̄
q
q = 2m(E + V0 )/h̄ (4.10)
4.3. MATHEMATICAL INTERLUDE 111

From the continuity conditions (i.e. both Rl (r) and its first derivative must be continuous at r = a),
we have
j 0 (qr)  Bjl0 (kr) + Cn0l (kr)
  
q l =k (4.11)
jl (qr) r=a
 Bjl (kr) + Cnl (kr) r=a

from which we can get the ratio C/B. The asymptotic form (see next section) of the solution for
the symmetric spherical well (for r  a) is

B lπ C lπ
    
Rl (r) r −→ sin kr − − cos kr − (4.12)
→∞
kr 2 B 2
Introducing
C
= − tan δl (k) (4.13)
B
the above asymptotic form reduces to:
B 1 lπ
  
Rl (r) r −→ sin kr − + δl (k) (4.14)
→∞
cos δl (k) kr 2

Comparing this with the solution to the free particle case (4.9), we find that the presence of the
potential induces a phase shift δl (k), which depends of the value of l, the energy of the incident
particle (through k) and the strength of the potential (since C/B and therefore δ l (k) depends on
q).

Problem 1: Show that for l = 0 i.e. s-wave scattering, the value of the phase-shift, δ 0 (k) is
determined by:
q cot qa = k cot(ka + δ0 (k)) (4.15)

and hence
k
 
δ0 (k) = arctan tan qa − ka (4.16)
q

4.3 Mathematical interlude

In this section we will review some mathematical tools needed in our study of the quantum mechan-
ical scattering problem. For a more detailed account of the topics discussed here, refer to Arfken’s
book, “Mathematical Methods for Physicists”, Third Edition, which is in the Rayleigh Library.

4.3.1 Brief review of complex analysis

• Analytic functions: A function f of the complex variable z = x + iy, defined in some

region D of the complex plane, is said to be analytic if it satisfies the Cauchy-Riemann
112 CHAPTER 4. SCATTERING THEORY

conditions. Specifically,

f (z) = u(x, y) + i v(x, y)

∂u ∂v
=
∂x ∂y

∂u ∂v
= −
∂y ∂x

(4.17)

Analytic functions have some very important properties among which is that their derivatives
to all orders exist and are also analytic.

• Cauchy theorem: If f (z) is analytic in a simply connected domain D, and if C is a simply

closed curve in D, then I
f (z) dz = 0
C

which can be proved by applying the Cauchy-Riemann conditions. Note that this result is
independent of the contour C.

• Cauchy integral formula: If f (z) is analytic in a simply connected domain D, and if C is

a simply closed curve in D, then

I
f (z)  2πif (z ) if z ∈ D
0 0
dz =
C z − z0  0 otherwise

• Residue theorem: Let f (z) be analytic in some neighbourhood of z = z0 , and let C be a

simple closed contour lying in this neighbourhood and surrounding z = z0 . The quantity

1
I
f (z)dz = Resf (z0 )
2πi C

is independent of the choice of C and is called the residue of f (z) at the point z = z 0 .
Evaluating many line integrals can therefore be reduced to simply finding the residue.

• Isolated singularity: A function f (z), analytic in the neighbourhood of z = z0 with the

exception of the point z = z0 itself, is said to have an isolated singularity at z = z0 . If there
are a finite number of isolated singularities at z = zi within C, then the integral is given by
the sum of residues, I X
f (z)dz = 2πi Resf (zi )
C i
4.3. MATHEMATICAL INTERLUDE 113

• Laurent series: If f (z) is analytic over a region r1 < |z − z0 | < r2 , it may be represented
there by a generalisation of Taylor’s series called a Laurent series:
∞
An (z − z0 )n
X
f (z) =
n=−∞

where the expansion coefficients are given by

1 f (z)
I
An = dz
2πi C (z − z0 )n+1
Then Resf (z0 ) is equal to the coefficient A−1 .

Example: The function f (z) = exp(z) + exp(1/z) − 1 is analytic for |z| > 0 and has Laurent
expansion
∞
X 1 n
f (z) = z
n=−∞
|n|!
Therefore in this case Resf (0) = 1.

4.3.2 Properties of spherical Bessel/Neumann functions

The spherical Bessel/Neumann functions are the solutions to the 2nd −order linear differential
equation (4.6):
l 
1 d sin ρ
 
l
jl (ρ) = (−ρ)
ρ dρ ρ
l 
1 d cos ρ
 
l
nl (ρ) = −(−ρ) (4.18)
ρ dρ ρ
where l = 0, 1, 2, . . .. They can be expressed by the ascending power series
1 2
ρl 2ρ ( 12 ρ2 )2
 
jl (ρ) = 1− + − ···
(2l + 1)!! 1!(2l + 3) 2!(2l + 3)(2l + 5)

1 2
(2l − 1)!! 2ρ ( 21 ρ2 )2
 
nl (ρ) = − 1 − + − ··· (4.19)
ρl+1 1!(1 − 2l) 2!(1 − 2l)(3 − 2l)
where (2l + 1)!! = 1 · 3 · 5 · · · (2l + 1).
(1) (2)
The spherical Hankel functions hl (ρ) and hl (ρ) are defined in terms of jl and nl :
(1)
hl (ρ) = jl (ρ) + inl (ρ)

(2) (1)
hl (ρ) = [hl (ρ)]∗ (4.20)
114 CHAPTER 4. SCATTERING THEORY

The explicit forms for the first few values of l are:

sin ρ
j0 (ρ) =
ρ

cos ρ
n0 (ρ) = −
ρ

(1) eiρ
h0 (ρ) =
iρ

sin ρ cos ρ
j1 (ρ) = −
ρ2 ρ

cos ρ sin ρ
n1 (ρ) = − −
ρ2 ρ

eiρ i
 
(1)
h1 (ρ) = − 1+ (4.21)
ρ ρ
Note that in general, nl (ρ) is singular (i.e. diverges) at the origin (while jl (ρ) is regular). This
means that if the interval over which the solution to (4.6) is sought includes the origin, then we
have to drop the nl (ρ) part. All the functions and their first derivatives defined above satisfy the
following recursion formulae:
2l + 1
zl (ρ) = zl−1 (ρ) + zl+1 (ρ)
ρ

1
 
0
zl (ρ) = lzl−1 (ρ) − (l + 1)zl+1 (ρ) (4.22)
(2l + 1)
(1) (2)
where the zl (ρ) are any of the functions jl (ρ), nl (ρ), hl (ρ), hl (ρ). The spherical Bessel functions
have the integral representation:

(−i)l 1
Z
jl (ρ) = dzPl (z)eiρz (4.23)
2 −1

where the Pl (z) are the Legendre polynomials. (You can easily verify this for l = 0, 1 etc. This
formula will be used later when we derive the expansion of plane-waves in terms of spherical waves).
The functions have the following asymptotic behaviour:

For small arguments (ρ  1, l):

ρl
jl (ρ) →
(2l + 1)!!

(2l − 1)!!
nl (ρ) → − (4.24)
ρl+1
4.3. MATHEMATICAL INTERLUDE 115

For large arguments (ρ  l):

1 lπ
 
jl (ρ) → sin ρ −
ρ 2

1 lπ
 
nl (ρ) → − cos ρ −
ρ 2

(1) i
hl (ρ) → − ei(ρ−lπ/2) (4.25)
ρ

4.3.3 Expansion of plane waves in spherical harmonics

In this section, we derive a formula relating a plane wave to an expansion involving the spherical
Bessel functions and the spherical harmonics. This will be used later in the partial-wave analysis
of spherically symmetric potentials.

The set of spherical wave solutions to the free particle Schrödinger equation, {j l (kr)} is complete.
We can therefore expand a plane wave, given by eik·r in terms of these solutions:
∞ X
l
eik·r =
X
clm (k) jl (kr) Ylm (θ, ϕ) (4.26)
l=0 m=−l

where the expansion coefficients clm (k) are to be determined.

First let k point along the z-axis and let θ be the angle between r and the z-axis. Then we have

k · r = kr cos θ

Note that there is no ϕ dependence as we have azimuthal symmetry about the z-axis. This implies
that

clm (k) → Al

1
2l + 1

2
Ylm (θ, ϕ) → Pl (cos θ) (4.27)
4π

and
∞ 1
2l + 1

2
eikr cos θ =
X
Al jl (kr) Pl (cos θ) (4.28)
l=0
4π

Using the orthonormality of the Legendre polynomials:

Z 1 2
d cos θ Pl (cos θ)Pl0 (cos θ) = δll0 (4.29)
−1 2l + 1
116 CHAPTER 4. SCATTERING THEORY

we have
1 1
Z 1
Al jl (kr) = [4π(2l + 1)] 2 dzPl (z)eikrz (4.30)
2 −1

where z = cos θ. Using the integral representation of the spherical Bessel function, (4.30) reduces
to
1
Al = il [4π(2l + 1)] 2 (4.31)

Therefore (4.26) becomes:

∞
eikr cos θ = il (2l + 1) jl (kr) Pl (cos θ)
X
(4.32)
l=0

For an arbitrary k, all we need to do to generalise (4.32) is use the addition theorem of spherical
harmonics. If θ is the angle between the vectors k and r then this theorem tells us that:
l
4π X
Pl (cos θ) = Y ∗ (Ωk ) Ylm (Ωr ) (4.33)
2l + 1 m=−l lm

where Ωk ≡ (θk , ϕk ) and Ωr ≡ (θr , ϕr ). Substituting (4.33) into (4.32) gives the general expansion:
∞ X
l
ik·r
il jl (kr) Ylm
∗
X
e = 4π (Ωk ) Ylm (Ωr ) (4.34)
l=0 m=−l

4.4 The quantum mechanical scattering problem

In a typical scattering experiment, one might measure the number of particles that are scattered
dσ
by an angle (θ, ϕ) into the element dΩ of the solid angle. The differential cross section dΩ is
defined by:
dσ number of particles scattered into dΩ per unit time
dΩ =
dΩ number of incident particles crossing unit area per unit time
In terms of the incident and scattered fluxes of particles we have:

dσ r2 dΩ|jscatt |2
dΩ = (4.35)
dΩ |jincid |2

where jincid and jscatt are the incident and scattered flux densities respectively. The total cross
section σtot is given by
dσ
Z
σtot = dΩ
unit sphere dΩ
Z 2π Z 1 dσ
= dϕ d cos θ (4.36)
0 −1 dΩ
4.4. THE QUANTUM MECHANICAL SCATTERING PROBLEM 117

The experimental setup is such that we have a beam of particles incident from z = −∞ travelling in
the +z direction (characterised by a plane-wave eikz ) scattered by the target particles (represented
mathematically by a potential V (r)), so that we have a spherical wave that emanates from the target
that describes the scattered particles. We therefore seek a solution to the Schrödinger equation
that has the asymptotic form:
eikr
ψ(r) r −→ eikz + f (θ, ϕ) (4.37)
→∞
r
The function f (θ, ϕ) is called the scattering amplitude and is related to the differential cross
section by:
dσ
= |f (θ, ϕ)|2 (4.38)
dΩ

Partial wave analysis

We now apply the tools developed in the previous sections to study scattering from a spherically
symmetric potential V (r). We assume that V (r) is short ranged in the sense that:
∞
Z 
r2 V (r) dr  < ∞
 

 (4.39)
0

(This ensures that we not only have a well defined scattering problem but also ensures the con-
vergence of the various expressions we shall encounter below.) The time independent Schrödinger
equation (4.2) for a particle in such a spherically symmetric potential is given by:

[∇2 + k 2 − U (r)]ψ(r) = 0

2mV (r)
U (r) =
h̄2

2mE
k2 = (4.40)
h̄2
We now decompose the ψ(r) into spherical waves:
∞
il (2l + 1) Rl (r) Pl (cos θ)
X
ψ(r) = (4.41)
l=0

where we have taken advantage of the spherical symmetry of V (r). Substituting (4.41) into (4.40)
gives the radial equation for Rl (r)
d2 2 d l(l + 1)
 
+ − + k 2 − U (r) Rl (r) = 0 (4.42)
dr2 r dr r2
The boundary condition for Rl (r) is that for r → 0, we want Rl (r) to be finite. Equation (4.42)
reduces for r → 0 to
d2 2 d l(l + 1)
 
2
+ − Rl (r) = 0 (4.43)
dr r dr r2
118 CHAPTER 4. SCATTERING THEORY

and the solution that is finite at the origin is

Rl (r) r−→
→0
Arl (4.44)

The remaining boundary condition is determined by the physics of the situation. For an incoming
beam eik·r (where it is understood that k points along the +z-axis), we have
∞
eik·r = il (2l + 1)jl (kr) Pl (cos θ)
X

l=0

∞
1X
 
(2) (1)
= il (2l + 1) hl (kr) + hl (kr) Pl (cos θ) (4.45)
2 l=0

The effect of the potential is to cause scattering and hence to modify the amplitude of each of the
outgoing spherical waves. Therefore, asymptotically, the full solution must be of the form
∞
1 (2)
 
X
l (1)
ψ(r) r → ∞
−→ i (2l + 1) hl (kr) + Sl (k)hl (kr) Pl (cos θ) (4.46)
l=0
2

where Sl contains the total effect of the scattering. (Note that we have incorporated into (4.46) the
fact that, for r → ∞, ψ(r) must satisfy the free particle Schrödinger equation due to the short-range
nature of the scattering potential as well as the fact that the scattering effects only the outgoing
spherical waves.) Equation (4.46) can now be rewritten as:
∞
1 X
 
l (1) (2) (1)
ψ(r) −→ i (2l + 1) hl (kr) + hl (kr) + [Sl (k) − 1]hl (kr) Pl (cos θ)
r→∞
2 l=0

∞
1
 
X
l (1)
= i (2l + 1) jl (kr) + [Sl (k) − 1]hl (kr) Pl (cos θ)
l=0
2

∞
1 (1)
eik·r + il (2l + 1)
X
= [Sl (k) − 1] hl (kr) Pl (cos θ) (4.47)
l=0
2

(1)
Since this is the solution for large r, we replace hl (kr) by its asymptotic form to get the final
asymptotic form for ψ(r)
eikr
ψ(r) r −→ eik·r + f (θ) (4.48)
→∞
r
where
∞
1 X
f (θ) = (2l + 1) [Sl (k) − 1] Pl (cos θ) (4.49)
2ik l=0

(Note that f (θ) also depends on the energy of the incident particle). Since from (4.38), the dif-
dσ
ferential cross section dΩ depends only on the scattering amplitude f (θ), our scattering problem
4.4. THE QUANTUM MECHANICAL SCATTERING PROBLEM 119

will be solved if we find the f (θ) or alternatively the Sl (k). The complex numbers Sl (k) can be
expressed in terms of real numbers δl (k) called phase shifts which are defined by

Sl (k) = e2iδl (k) (4.50)

(The factor of 2 is conventional). Using this definition of the phase-shifts, the scattering amplitude
(4.49) reduces to,
∞
1X
f (θ) = (2l + 1) eiδl (k) sin δl (k) Pl (cos θ) (4.51)
k l=0
and the differential cross section is
dσ
= |f (θ)|2
dΩ
∞ X ∞
1 X
= (2l + 1) (2l0 + 1) sin δl sin δl0 cos(δl − δl0 ) Pl (cos θ) Pl0 (cos θ) (4.52)
k 2 l=0 l0 =0

The total cross section, σtot is given by

dσ
Z
σtot = dΩ
unit sphere dΩ

∞
4π X
= (2l + 1) sin2 δl (4.53)
k 2 l=0

Problem 2: Verify equation (4.53).

The total cross section can also be written as

∞
X
σtot = σl (4.54)
l=0

where the lth partial cross section σl is given by

4π
σl = (2l + 1) sin2 δl (4.55)
k2
For elastic scattering, the phase shifts are real so it follows from (4.55) that for elastic scattering
4π
σl ≤ (2l + 1) (4.56)
k2
The value 4π(2l + 1)/k 2 is known as the unitarity bound and is reached only for
1
 
δl = n + π, (n = 0, 1, . . .) (4.57)
2
120 CHAPTER 4. SCATTERING THEORY

This is the condition for resonance which shows up as a local maximum in the cross section for
the corresponding partial wave.

On the other hand,

∞
1 X
Imf (θ) = (2l + 1) Re[1 − Sl ] Pl (cos θ)
2k l=0

∞
1 X
= (2l + 1) [2 sin2 δl ] Pl (cos θ) (4.58)
2k l=0

Setting θ = 0 and making use of Pl (1) = 1, we get

∞
1X
Imf (θ = 0) = (2l + 1) sin2 δl (4.59)
k l=0

Comparing this with (4.53), we obtain the optical theorem:

4π
σtot = Imf (θ = 0) (4.60)
k

Problem 3: Consider the elastic scattering of a low energy particle from a “hard sphere” potential
(V (r) = ∞ for r ≤ a, 0 otherwise.) Derive an expression for tan δl and show that for l = 0,
tan δ0 = − tan(ka) where k 2 = 2mE/h̄2 . Show that as k → 0 the total cross section approaches
4πa2 . Hence obtain an expression for the s-wave contribution to the forward scattering amplitude
f (θ = 0) and verify the optical theorem for ka  1.

4.4.1 Born approximation

Another technique for determining the scattering amplitude, valid for scattering potentials that
are weak (and so may be regarded as a perturbation), is by means of the Born approximation. To
derive the expression for the scattering amplitude in this case, we first express the time independent
Schrödinger equation in integral form.

Integral form of the Schrödinger equation

The time independent Schrödinger equation

"
h̄2 2
#
− ∇ + V (r) ψ(r) = Eψ(r) (4.61)
2m
4.4. THE QUANTUM MECHANICAL SCATTERING PROBLEM 121

can be written as
[∇2 + k 2 ]ψ(r) = Q(r) (4.62)

where
√
k = 2mE/h̄

2m
Q(r) = V (r) ψ(r) (4.63)
h̄2

Note that

• V (r) in (4.61) need not be spherically symmetric,

• Q(r), the “inhomogeneous source term” itself depends on ψ(r)

• (4.62) is the inhomogeneous Helmholtz equation.

Suppose there exists a function, G(r − r0 ), that is the solution to the following differential equation

[∇2 + k 2 ]G(r − r0 ) = δ(r − r0 ) (4.64)

where the Dirac delta function δ(r − r0 ) is defined by

1
Z
0 0
δ(r − r ) = eis·(r−r ) ds (4.65)
(2π)3 all space

In what follows, it is understood that the Laplacian ∇2 acts on the argument r. We can then
express ψ(r) as an integral Z
ψ(r) = G(r − r0 )Q(r0 )dr0 (4.66)
all space

Note that (4.66) is a integral equation since the unknown function ψ(r) appears under the integral
sign. To show that (4.66) satisfies (4.62) consider
Z
[∇2 + k 2 ]ψ(r) = [∇2 + k 2 ] G(r − r0 )Q(r0 )dr0
all space

Z
= [(∇2 + k 2 )G(r − r0 )]Q(r0 )dr0
all space

Z
= δ(r − r0 )Q(r0 )dr0
all space

= Q(r) (4.67)
122 CHAPTER 4. SCATTERING THEORY

The function G(r − r0 ) is called the Green’s function for the equation (4.62). In general, the
Green’s function for a given differential equation represents the “response” to a delta-function
source. The easiest way to determine the G(r − r0 ) here is by taking the Fourier transform (which
turns the differential equation into a algebraic equation). Let
1
Z
G(r) = eis·r g(s)ds (4.68)
(2π)3/2 all space

Then
1
Z
(∇2 + k 2 )G(r) = [(∇2 + k 2 )eis·r ]g(s)ds
(2π)3/2 all space

1
Z
= (k 2 − s2 )eis·r g(s)ds (4.69)
(2π)3/2 all space

Problem 4: Verify Eq (4.69).

This means that

1 1
Z Z
(k 2 − s2 )eis·r g(s)ds = eis·r ds (4.70)
(2π)3/2 all space (2π)3 all space

which implies
1 1
g(s) = (4.71)
(2π) (k − s2 )
3/2 2

so that
1 1
Z
G(r) = eis·r ds (4.72)
(2π)3 all space (k 2 − s2 )
To evaluate the integral (4.72) we first note that r is fixed as far as the s integration is concerned.
It is advantageous to do the integration in spherical coordinates such that s = (s, θ, ϕ) and with
the polar axis pointing along r. Then since s · r = sr cos θ and ds = s2 sin θ dθ dϕ ds, (4.72) becomes

1 2π ∞ s2 ds 1
Z Z Z
G(r) = dϕ d cos θ eisr cos θ (4.73)
(2π)3 0 0 (k 2 − s2 ) −1

The ϕ integration is trivial (= 2π) and the θ integration yields

Z 1 2 sin(sr)
d cos θ eisr cos θ = (4.74)
−1 sr
Thus
1 2
Z ∞ s sin(sr)
G(r) = ds
(2π)2 r 0 (k 2 − s2 )

1
Z ∞ s sin(sr)
= ds (4.75)
4π 2 r −∞ (k 2 − s2 )
4.4. THE QUANTUM MECHANICAL SCATTERING PROBLEM 123

1 isr
Writing sin(sr) = 2i (e − e−isr ) (4.75) becomes
i
G(r) = [I1 − I2 ]
8π 2 r
∞ seisr
Z
I1 = ds
−∞ (s − k)(s + k)

∞ se−isr
Z
I2 = ds
−∞ (s − k)(s + k)

(4.76)

The integrals I1 and I2 can be evaluated using Cauchy’s integral formula.

Problem 5: In (4.76), the integration is along the real axis and passes right over the pole singu-
larities at ±k. How to deal with integration in the vicinity of these singularities is fixed by the
boundary conditions; the result is that the contour of integration should go over the singularity at
−k and under the singularity at +k. We must close the contour in such a way that the semicircle
at infinity contributes nothing.

• Show that for I1 , we must close the contour above the real axis. Hence, show that

I1 = iπeikr

• For I2 show that the contour must be closed below the real axis so that

I2 = −iπeikr

(Remember that when you go round the contour in the clockwise direction, you pick up a
minus sign).

Using the results of the above problem, we have

eikr
G(r) = − (4.77)
4πr
so that the general solution to (4.66) takes the form
0
m eik|r−r |
Z
ψ(r) = ψ0 (r) − V (r0 )ψ(r0 )dr0 (4.78)
2πh̄2 all space |r − r0 |
where the ψ0 (r) satisfies the free particle Schrödinger equation

[∇2 + k 2 ]ψ0 (r) = 0 (4.79)

124 CHAPTER 4. SCATTERING THEORY

Equation (4.78) is the integral form of the Schrödinger equation and is equivalent to the differential
form plus boundary conditions.

Problem 6: Check that (4.78) satisfies (4.62) by direct substitution.

Hint: The following identity will be useful:

1
 
∇2 = −4πδ(r − r0 )
|r − r0 |
The ∇ acts on the r argument.

First Born approximation

Suppose V (r0 ) is localised about r0 = 0. (This means that the potential drops to zero outside
some finite region). We want to determine ψ(r) at points far away (i.e. r  r 0 ) from the scattering
centre. For this case we have
2r · r0
 
|r − r | 0 2 ∼
= r2 1 −
r2

=⇒ |r − r0 | ∼
= r − r̂ · r0 (4.80)

where r̂ denotes a unit vector pointing along r. Let

k0 ≡ kr̂

then
0 0 0
eik|r−r | ∼
= eikr e−ik ·r (4.81)

and therefore 0
eik|r−r | ∼ eikr −ik0 ·r0
= e (4.82)
|r − r0 | r
In the case of scattering where the incident beam is along the z−axis, we require

ψ0 (r) = eikz (4.83)

In the asymptotic limit (large r), (4.78) reduces to

m eikr
Z
0 0
ψ(r) ∼
= eikz − e−ik ·r V (r0 )ψ(r0 )dr0 (4.84)
2πh̄2 r all space

from which we can read off the scattering amplitude

m
Z
0 0
f (θ, ϕ) = − e−ik ·r V (r0 )ψ(r0 )dr0 (4.85)
2πh̄2 all space
4.4. THE QUANTUM MECHANICAL SCATTERING PROBLEM 125

This expression for f (θ, ϕ) is exact. We now invoke the Born approximation: Suppose the
incoming plane wave is not substantially altered by the potential (i.e. the scattering potential is
weak), than it makes sense to substitute
0 0
ψ(r0 ) ≈ ψ0 (r0 ) = eikz = eik·r (4.86)

where k = kẑ, into the integral (4.84). The scattering amplitude in the Born approximation then
reduces to
m
Z
0 0
f (k0 , k) = f (θ, ϕ) = − e−i(k −k)·r V (r0 )dr0 (4.87)
2πh̄2 all space

(Note: k and k0 both have magnitude k but the former points in the direction of the incident beam
while the latter points towards the detector). Equation (4.87) indicates that f (θ, ϕ) in the Born
approximation is just the Fourier transform of the scattering potential.

For a spherically symmetric potential, V (r) = V (r), the Born approximation for f (θ) reduces to
2m
Z ∞
f (θ) ∼
=− rV (r) sin(qr)dr (4.88)
qh̄2 0

where

q = k0 − k

q = |k0 − k|

= 2k sin(θ/2)

(k0 − k) · r0 = qr0 cos θ 0 (4.89)

For low-energy scattering, where the de Broglie wavelength of the scattered particle is much
larger than the extent of the scattering region, the Born approximation simplifies to
m
Z
f (θ, ϕ) ∼
=− V (r0 )dr0 (4.90)
2πh̄2 all space

Problem 7: A potential of considerable utility in both nuclear physics and chemistry is the Yukawa
or “screened” Coulomb potential:
e−r/r0
V (r) = −A
r
where r0 > 0. Determine the scattering amplitude in the first Born approximation for this potential.
Hence, obtain the differential scattering cross section in the limit r0 → ∞.
126 CHAPTER 4. SCATTERING THEORY

Answer: The scattering amplitude is given by

!
2Am r02
f (θ) =
h̄2 1 + q 2 r02

and the differential cross section at the limit r0 → ∞ is

dσ 4A2 m2
= 4 4
dΩ h̄ q

Problem 8† : Determine the scattering amplitude in the first Born approximation for the spheri-
cally symmetric square well.

Answer: For an attractive square well, we have

2mV0
f (θ) = (sin qa − qa cos qa)
h̄2 q 3

Problem 9† : Determine the scattering amplitude in the first Born approximation for the potential
2
V (r) = −Ae−br

where A, b > 0.

Answer: The scattering amplitude is given by

√
Am π −q2 /4b
f (θ) = 2 3/2 e
2h̄ b

4.5 *Formal time-independent scattering theory

We now develop a time-independent formulation of scattering processes. Assume that the Hamil-
tonian for the scattering problem is
Ĥ = Ĥ0 + V̂ (4.91)

where Ĥ0 stands for the kinetic energy operator

p̂2
Ĥ0 = (4.92)
2m

In the absence of a scatterer, V̂ would be zero and the energy eigenstate would be just a free particle
state |pi where
p2
Ĥ0 |pi = |pi (4.93)
2m
4.5. *FORMAL TIME-INDEPENDENT SCATTERING THEORY 127

The presence of V̂ therefore causes the energy eigenstate to be different from the free-particle state.
For elastic scattering, where no change in energy is involved, we need to find solutions to Eq. (4.91)
with the same energy eigenvalue as Eq. (4.93).

In general, let |φi be an energy eigenstate of Ĥ0

Ĥ0 |φi = E|φi (4.94)

where in general, |φi stands for either plane waves or free spherical wave states. We want to solve
 
Ĥ|ψi = Ĥ0 + V̂ |ψi = E|ψi (4.95)

with both Ĥ0 and Ĥ exhibiting continuous energy spectra. We look for solutions to Eq. (4.95)
such that
|ψi → |φi as V̂ → 0 (4.96)

where |φi is the solution to the free particle Schrödinger equation Eq. (4.94) with the same energy
eigenvalue.

The desired solution (at least in a formal sense) is

 −1
|ψi = |φi + E − Ĥ0 V̂ |ψi (4.97)

although there are complications arising from the singular nature of the operator (E − Ĥ0 )−1 .

Problem 10† : Verify that Eq. (4.97) is indeed a solution to Eq. (4.95)

Moreover, the prescription for dealing with the singular nature of (E − Ĥ0 )−1 is to treat E as a
complex variable and to add an infinitesimal imaginary number to it, i.e.
1
|ψi = |φi + V̂ |ψi (4.98)
(E − Ĥ0 + i)
Eq. (4.98) is called the Lippmann-Schwinger equation and is a vector equation independent of
any particular representations.

4.5.1 *Lippmann-Schwinger equation in the position representation

We now focus on expressing the operator Lippmann-Schwinger equation Eq. (4.98) in the position
representation, which enables us to obtain the scattering amplitude in the (first) Born approxima-
tion. In the position basis (4.98) becomes
1
Z
hr|ψi = hr|φi + dr0 hr| |r0 ihr0 |V̂ |ψi (4.99)
(E − Ĥ0 + i)
128 CHAPTER 4. SCATTERING THEORY

which is an integral equation for scattering because the unknown ket |ψi appears under the
integral sign. The quantity
h̄2 1
G(r, r0 ) = hr| |r0 i (4.100)
2m (E − Ĥ0 + i)
is just the Green’s function (equation (4.77)) derived in the previous section. It was worked out to
be 0
0 1 eik|r−r |
G(r, r ) = −
4π |r − r0 |
where E = h̄2 k 2 /2m. Equation (4.98) then just reduces to the integral equation (4.66). By looking
at the large distance behaviour of r, we again get the expression for the scattering amplitude
describing an incident beam of particles in the direction k being scattered in the direction k 0 :
m
f (k0 , k) = − hk0 |V̂ |ψi (4.101)
2πh̄2
which is exact.

4.5.2 *Born again!

We now define the transition operator T̂ , such that

V̂ |ψi = T̂ |φi (4.102)

Multiplying the Lippmann-Schwinger equation Eq.(4.98) by V̂ , we obtain

1
T̂ |φi = V̂ |φi + V̂ T̂ |φi (4.103)
(E − Ĥ0 + i)
which is supposed to hold for any |φi (taken to be any plane-wave state for example, which we
know are complete). Therefore the following operator equation is satisfied:
1
T̂ = V̂ + V̂ T̂ (4.104)
(E − Ĥ0 + i)

The scattering amplitude can now be written as

m
f (k0 , k) = − 0
2 hk |T̂ |ki (4.105)
2πh̄
which shows that to determine the scattering amplitude, it is sufficient to know the transition
operator T̂ .

An iterative solution for T̂ is obtained as follows:

1 1 1
T̂ = V̂ + V̂ V̂ + V̂ V̂ V̂ + · · · (4.106)
(E − Ĥ0 + i) (E − Ĥ0 + i) (E − Ĥ0 + i)
4.5. *FORMAL TIME-INDEPENDENT SCATTERING THEORY 129

Correspondingly, we can expand f (k0 , k) as follows:

∞
f (k0 , k) = f (n) (k0 , k)
X
(4.107)
n=1

where n is the number of times the V̂ operator enters. We have

m
f (1) (k0 , k) = − hk0 |V̂ |ki
2πh̄2

m 1
f (2) (k0 , k) = − 0
2 hk |V̂ V̂ |ki (4.108)
2πh̄ (E − Ĥ0 + i)

etc., which are just the first, second etc. order Born approximations and is called the Born series.
A physical interpretation is that the scattering process is viewed as a multi-step process with,
for example, f (2) (k0 , k) being viewed as the incident wave (with wave vector k undergoing two
sequential interactions before being scattered into the direction k0 and so on. In this context,
the Green’s function is called the propagator – it tells us how a disturbance propagates between
one interaction and the next. The Born series was the inspiration for Feynman’s formulation of
relativistic quantum mechanics, which is expressed entirely in terms of vertex functions V̂ and
propagators G, connected together in Feynman diagrams. It is this technique that forms the
basis for the treatment of the quantum theory of fields and its application to a wide range of
phenomena including elementary particle and condensed matter physics.
130 CHAPTER 4. SCATTERING THEORY
Chapter 5

Identical Particles in Quantum

Mechanics

5.1 Introduction

Consider a system of N identical particles. In classical physics it is possible (at least in principle), to
keep track of the individual particles at each instant of time. In this sense, the particles retain their
individuality and are distinguishable. In QM the situation is very different. When two identical
particles come together, interact, and then separate, their individuality is lost. This is because we
cannot ‘tag’ the particles nor can we follow their trajectories (since this would entail a position
measurement at each instant in time which necessarily disturbs the system). Therefore, whereas
in classical physics there is no necessity to treat distinguishable or non-distinguishable particles
differently, in QM the fact that identical particles are indistinguishable must be “built in” to the
theory.

5.2 Multi-particle systems

Consider the Hamiltonian for the N −particle system

N 
h̄2 2
X 
Ĥ(1, 2, . . . , N ) = − ∇ + vext (i) + Vint (1, 2, . . . , N ) (5.1)
i=1
2m i

where the labels 1, 2, . . . represent all of particle i’s internal (for example spin) and external (for
example position) degrees of freedom, collectively. vext (i) is the external potential (for example due

131
132 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

to an electric/magnetic field) acting on the ith particle, and Vint (1, 2, . . . , N ) represents the mutual
interaction between the particles. If, as is usually the case, the interaction between the particles is
due to two-body interactions (as in the case of the Coulomb interaction between charged particles)
then
N
1X
Vint (1, 2, . . . , N ) = Vint (i, j) (5.2)
2 i6=j

where due to Newton’s 3rd law

Vint (i, j) = Vint (j, i) (5.3)

Therefore the N −particle Hamiltonian is given by

N  N
h̄2 2 1X
X 
Ĥ(1, 2, . . . , N ) = − ∇ + vext (i) + Vint (i, j) (5.4)
i=1
2m i 2 i6=j

Note that this Hamiltonian in invariant under the interchange of particles. Let P̂ij be a
particle interchange (or permutation) operator which interchanges the labels i ↔ j such that its
action on f (1, 2, . . . , N ) is

P̂ij f (1, 2, . . . , i, . . . , j, . . . , N ) = f (1, 2, . . . , j, . . . , i, . . . , N ) (5.5)

Note that from (5.5) it is apparent that P̂ij2 = Iˆ which implies that P̂ij has ±1 as eigenvalues. (In
fact, P̂ij is both Hermitian and unitary). Let P̂ denote some arbitrary particle interchange, not
necessarily pair-wise. (Note that P̂ can be represented as the product of pair-wise interchanges,
P̂ij . P̂ is called even(odd) if it is made up of an even(odd) number of the P̂ij ). The property
P̂ 2 = Iˆ still holds. Then by the symmetry of the Hamiltonian under any number of interchanges
of identical particles, we have
[Ĥ, P̂ ] = 0 (5.6)

Theorem: Let Ψ(1, 2, . . . , N ) denote the N −body wave function for this system. If Ψ(1, 2, . . . , N )
is an eigenfunction of Ĥ(1, 2, . . . , N ) with eigenvalue E, then so is P̂ Ψ.

Proof: Given ĤΨ = EΨ, from (5.6) we have

Ĥ(P̂ Ψ) = P̂ (ĤΨ) = E(P̂ Ψ) (5.7)

i.e. P̂ Ψ is also an eigenstate corresponding to the same eigenvalue E.

This means that we can use the eigenvalues of P̂ (which are ±1) together with the energy E to
label the eigenstates of the Hamiltonian. In general for any symmetrical operator (i.e. one which
is even under interchange of particle labels), Ŝ(1, 2, . . . , N ) we have

[P̂ , Ŝ] = 0 (5.8)

5.2. MULTI-PARTICLE SYSTEMS 133

If |Φi = P̂ |Ψi for some arbitrary |Ψi then

hΦ|Ŝ|Φi = hΨ|P̂ † Ŝ P̂ |Ψi

= hΨ|P̂ † P̂ Ŝ|Ψi

= hΨ|Ŝ|Ψi (5.9)

using the unitary property of P̂ . Therefore the expectation value (and more generally the matrix
elements) of the states |Ψi and P̂ |Ψi of a symmetrical operator Ŝ are equal. Since identical particles
are influenced equivalently by any physical process, all physical operators are symmetric and
thus the states Ψ(1, 2, . . . , N ) and P̂ Ψ(1, 2, . . . , N ) cannot be distinguished from one another. As
we have seen, we must in fact have

P̂ Ψ(1, 2, . . . , N ) = ±Ψ(1, 2, . . . , N ) (5.10)

All experimental observations to date imply that the sign factor is either always +1 (i.e. Ψ is com-
pletely symmetric and represents bosons) or else it is ±1 for even/odd P̂ (i.e. Ψ is antisymmetric
with respect to any pair-wise interchange and represents fermions). Furthermore, the nature of a
given type of particle is intimately connected with its intrinsic spin. This connection between spin
and symmetry follows from the spin-statistics theorem of quantum field theory which states
that all elementary particles with integer spin (including spin 0) are bosons (obeying Bose-Einstein
statistics) and all particles with half-odd integer spin are fermions (obeying Fermi-Dirac statistics).

5.2.1 Pauli exclusion principle

Let ΨA (1, 2, . . . , N ) denote a totally antisymmetric wave function. Then

P̂ij ΨA (1, . . . , i, . . . , j, . . . N ) = ΨA (1, . . . , j, . . . , i, . . . N )

= −ΨA (1, . . . , i, . . . , j, . . . N ) (5.11)

If we set i = j in (5.11) we get

ΨA (1, . . . , i, . . . , i, . . . N ) = −ΨA (1, . . . , i, . . . , i, . . . N ) (5.12)

i.e.
ΨA (1, . . . , i, . . . , i, . . . N ) = 0
Thus two fermions in the same spin state cannot occupy the same position in space. (Remember that
i includes both the spatial and spin degrees of freedom). This is the Pauli Exclusion Principle.
Note that this is true independent of whether the particles interact with one another or not. The
key point is that they are identical fermions.
134 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

5.2.2 Representation of Ψ(1, 2, . . . , N )

For N identical non-interacting particles, the Hamiltonian is

N
X
Ĥ = Ĥ(i) (5.13)
i=1

i.e the sum of N −identical one-particle Hamiltonians Ĥ(i). Assume that the eigenfunctions of Ĥ(i)
are given by the solution to the following one-particle Schrödinger equation

Ĥ(i)ψαi (i) = Eαi ψαi (i) (5.14)

where i labels the particles (and is in the range 1 → N ), while αi labels the different one-particle
eigenstates (and ranges from 1 → ∞). We can use these one-particle eigenfunctions to construct
the N −particle wave function in various ways.

5.2.3 Neglecting the symmetry of the many-body wave function

If we set
Ψ(1, 2, . . . , N ) = ψα1 (1)ψα2 (2) · · · ψαN (N ) (5.15)

then
N
X
ĤΨ = Ĥ(i)Ψ
i=1

N
X  
= Ĥ(i) ψα1 (1)ψα2 (2) · · · ψαN (N )
i=1

N
X  
= ψα1 (1) · · · Ĥ(i)ψαi (i) · · · ψαN (N )
i=1

N
X  
= ψα1 (1) · · · Eαi ψαi (i) · · · ψαN (N )
i=1

N
X
= E αi Ψ
i=1

= EΨ (5.16)
PN
which implies E = i=1 Eαi . These states are in general neither symmetric nor antisymmetric,
and therefore cannot describe real systems of identical particles.
5.3. FERMIONS 135

5.3 Fermions

For two fermions, the antisymmetric state is

1
 
ΨA (1, 2) = √ ψα1 (1)ψα2 (2) − ψα1 (2)ψα2 (1) (5.17)
2

Problem 1: Show that if the ψαi (i)’s are orthonormal, then ΨA (1, 2) defined above is normalised
to one.

In general, for N identical fermions the many-body wave function is given by

1 X
 
ΨA (1, · · · N ) = √ (−1)P P̂ ψα1 (1)ψα2 (2) · · · ψαN (N )
N! P
 
 ψα1 (1) ψα1 (2) ··· ψα1 (N )
 

1  .. .. .. ..

= √  . . . .

 (5.18)
N !  
 ψαN (1) ψαN (2) · · · ψαN (N ) 


√
The normalisation factor of 1/ N ! is introduced because the determinant consists of N ! mutually
orthogonal terms. (Actually orthogonality is not really necessary, all that is required is that the
ψαi (i)’s are linearly independent.) Such determinants of one-particle states are called Slater
determinants. Since the interchange of any two particle coordinates corresponds to interchanging
columns in the Slater determinant (which introduces a factor of –1), the antisymmetry is already
built in. Also, the ΨA (1, · · · N ) defined above is zero if we set αi = αj for any i 6= j, i.e. no state
can be multiply occupied (this is another statement of the Pauli exclusion principle).

5.4 Bosons

For two bosons, the symmetric state for α1 6= α2 is

1
 
ΨS (1, 2) = √ ψα1 (1)ψα2 (2) + ψα1 (2)ψα2 (1) (5.19)
2
whereas for α1 = α2 we have simply

ΨS (1, 2) = ψα1 (1)ψα1 (2) (5.20)

In general, for N bosons the many-body wave function is defined as

s
N1 !N2 ! . . . X 0
 
ΨS (1, · · · N ) = P̂ ψα1 (1)ψα2 (2) · · · ψαN (N ) (5.21)
N! P0
136 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

P
where N1 is the multiplicity of state α1 etc. The summation P0 is only over those permutations
leading to distinct terms and includes N !/N1 !N2 ! . . . different terms.

5.5 Exchange forces

We now examine the physical impact of the symmetrization requirement for identical particles.
Consider a simple one-dimensional case with two non-interacting identical particles, one in state
ψa and the other in state ψb with these two states being orthonormal.

If the two particles are distinguishable, with particle 1 in state ψa (x1 ) and particle 2 in state
ψb (x2 ), then the total wave function for the two particle system is

Ψ(x1 , x2 ) = ψa (x1 )ψb (x2 ) (5.22)

If the two particles are identical bosons, then the total normalised wave function is
1
 
ΨS (x1 , x2 ) = √ ψa (x1 )ψb (x2 ) + ψa (x2 )ψb (x1 ) (5.23)
2
and if they are identical fermions it is
1
 
ΨA (x1 , x2 ) = √ ψa (x1 )ψb (x2 ) − ψa (x2 )ψb (x1 ) (5.24)
2
What is the expectation value of the square of the separation distance between the two particles
X12 ? Here X12 is defined as

X12 = h(x1 − x2 )2 i
Z
= dx1 dx2 Ψ∗ (x1 , x2 )(x1 − x2 )2 Ψ(x1 , x2 )

= hx21 i + hx22 i − 2hx1 x2 i (5.25)

where the expectation value is with respect to the total two particle wave function. There are three
cases.

Case 1: Distinguishable particles

Using (5.22), we have

Z
hx21 i = dx1 dx2 Ψ∗ (x1 , x2 )x21 Ψ(x1 , x2 )

Z Z
= x21 |ψa (x1 )|2 dx1 |ψb (x2 )|2 dx2

= hx2 ia (5.26)
5.5. EXCHANGE FORCES 137

i.e. this is the expectation value of x2 in the one-particle state ψa . (Also remember that ψa and ψb
are normalised). Similarly,
Z Z
hx22 i = x22 |ψb (x2 )|2 dx2 |ψa (x1 )|2 dx1

= hx2 ib (5.27)

and
Z Z
2
hx1 x2 i = x1 |ψa (x1 )| dx1 x2 |ψb (x2 )|2 dx2

= hxia hxib (5.28)

which implies that

X12 = hx2 ia + hx2 ib − 2hxia hxib (5.29)

Note that the above result would have been the same if particle 1 had been in state ψ b and particle
2 had been in ψa .

Case 2: Identical particles Using (5.23) and (5.24) we have

1
Z Z 
hx21 i = x21 |ψa (x1 )|2 dx1 |ψb (x2 )|2 dx2
2

1
Z Z 
+ x21 |ψb (x1 )|2 dx1 |ψa (x2 )|2 dx2
2

1
Z Z 
± x21 ψa∗ (x1 )ψb (x1 )dx1 ψb∗ (x2 )ψa (x2 )dx2
2

1
Z Z 
± x21 ψb∗ (x1 )ψa (x1 )dx1 ψa∗ (x2 )ψb (x2 )dx2
2

1
 
= hx2 ia + hx2 ib ± 0 ± 0
2

1
 
= hx2 ia + hx2 ib (5.30)
2
where the + sign corresponds to bosons and the − sign to fermions. Similarly,
1
 
hx22 i = hx2 ib + hx2 ia (5.31)
2

Problem 2: Verify (5.31) above. (Note that you would expect hx1 i = hx2 i since one can’t tell the
particles apart.) Then show that

hx1 x2 i = hxia hxib ± |hxiab |2 (5.32)

138 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

where Z
hxiab = xψa∗ (x)ψb (x)dx (5.33)

Therefore, we have





hx2 ia + hx2 ib − 2hxia hxib distinguishable particles
X12 =

hx2 i + hx2 i − 2hxi hxi − 2|hxi |2 bosons
a b a b ab
 hx2 ia + hx2 ib − 2hxia hxib + 2|hxiab |2

fermions


This indicates that identical bosons tend to be somewhat closer together and identical fermions
somewhat farther apart than distinguishable particles in the same pair of states. Note that hxi ab
vanishes unless the two wave functions actually overlap. As a practical matter, it is reasonable to
assume that identical particles with non-overlapping wave functions (if for example they are very far
apart) are distinguishable. If there is some overlap of the wave functions, then the system behaves
as though there were a “force of attraction” between identical bosons and a “force of repulsion”
between identical fermions (which tends to keep them apart). It is called the exchange force,
although it is not really a force at all as no physical agency is acting on the particles; rather it
is a purely geometrical consequence of the symmetrization requirement. It is a strictly quantum
mechanical phenomenon with no classical counterpart.

Problem 3: Imagine two noninteracting particles each of mass m in an infinite square well. If one
of the particles is in state ψn where
r
2 nπx
 
ψn (x) = sin
L L
and the other is in state ψm , calculate X12 assuming (a) they are distinguishable, (b) they are
identical bosons and (c) they are identical fermions.

Answer: To be done in class.

5.6 Helium atom

The accurate determination of the properties of atoms is a difficult task. The two main sources of
the difficulty are:

• the Coulomb interaction between the electrons, and

• the spin-orbit interaction between the electron spins and the electric fields in the atom.
5.6. HELIUM ATOM 139

In this section we deal with the simplest atom after hydrogen, the helium atom. (Note that other
two-electron ions such as H− , Li+ , Be++ etc. can be treated in the same way.) The spin-orbit
interaction can be regarded as a weak perturbation; to a first approximation it can be neglected
entirely and we only consider the Coulomb interactions between the electrons and the between
electrons and the positively charged nuclei. The Hamiltonian (in atomic units) is given by
1 1 Z Z 1
Ĥ = − ∇21 − ∇22 − − + (5.34)
2 2 r1 r2 r12
where for generality we write the Hamiltonian in terms of Z rather then setting it to 2 for the specific
case of helium. In the equation above, the first two terms correspond to the kinetic energies of the
two electrons, the next two terms are the electron-nuclei coulombic attraction, and the last term
is the coulombic repulsion between the electrons. We will use both perturbation theory and the
variational method to “solve” the time-independent Schrödinger equation to determine the ground
state of this atom.

5.6.1 Ground state

We apply perturbation theory to this problem by considering the inter-electronic repulsion term to
be a perturbation. In this case, the unperturbed hamiltonian Ĥ0 is given by
1 1 2 Z Z
Ĥ0 = − ∇21 − ∇2 − − (5.35)
2 2 r1 r2
and the zeroth-order ground state wave function by

ψ (0) (r1 , r2 ) = ψ1s (r1 ) ψ1s (r2 ) (5.36)

!1
Z3 2
ψ1s (r) = e−Zr (5.37)
π
Here we are using the fact that in the ground state the two electrons are a state of total spin S = 0,
which is already antisymmetric with respect to interchange of their spins (see Chapter 2), so that
they can have the same spatial wave function.

Since the ground state energy of a hydrogen-like system is −Z 2 /2, the energy E (0) in atomic units
is given by
Z2 Z2
E (0) = − − = −Z 2 (5.38)
2 2
The first-order correction to E (0) (see Chapter 3) is
1
Z Z
E (1) = dr1 dr2 ψ (0)∗ (r1 , r2 ) ψ (0) (r1 , r2 )
r12

1
Z Z
= dr1 dr2 ψ1s (r1 ) ψ1s (r2 ) ψ1s (r1 ) ψ1s (r2 ) (5.39)
r12
140 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

The integral is evaluated in the problem below.

Problem 4: We evaluate the integral in (5.39) as follows. Let r1 and r2 be the radius vectors of
electron 1 and 2 respectively and let θ be the angle between these two vectors. In general, this
is not the polar angle from the z-axis but if we choose on of the radius vectors, say r 1 , to be the
z-axis, then we can treat θ as the polar angle in spherical coordinates. From the cosine rule, we
have !1
2
r12 = r12 + r22 − 2r1 r2 cos θ

Show that E (1) becomes

Z6 ∞ ∞
Z Z
E (1) = e−2Zr1 4πr12 dr1 e−2Zr2 r22 dr2
π2 0 0

Z 2π Z π sin θ dθ
× dφ 1 (5.40)
0 0 (r12 + r22 − 2r1 r2 cos θ) 2
Letting x = cos θ, show that the integral over θ is
Z π sin θ dθ
Z 1 dx
1 = 1
0 (r12 + r22 − 2r1 r2 cos θ) 2 −1 (r 2
1 + r22 − 2r1 r2 x) 2

2
= r1 > r 2
r1

2
= r1 < r 2 (5.41)
r2
Substituting this result into E (1) , show that
!
Z ∞ 1
Z r1 Z ∞
(1) 6 −2Zr1
E = 16Z e r12 dr1 e−2Zr2 r22 dr2 + e −2Zr2
r2 dr2
0 r1 0 r1

" !#
Z ∞ 1 1
3 −2Zr1
= 4Z e r12 dr1 − e−2Zr1 Z +
0 r1 r1

5
= Z (5.42)
8

The final result is

5
Z E (1) = (5.43)
8
so that the ground state energy of the helium atom through first order is
5
E = E (0) + E (1) + · · · = −Z 2 + Z + ··· (5.44)
8
5.6. HELIUM ATOM 141

Setting Z = 2, we find that E = − 11

4 = −2.75 or −74.83 eV, while the experimental value is −79.00
eV or −2.9033 au. This means that first-order perturbation theory already gives a result within 5
% of the experimental (exact) value. Note that more extensive perturbation calculations to many
orders by Scheer and Knight (Rev. Mod. Phys. 35, 426 (1963)) give
5 0.008698679
E = −Z 2 + Z − 0.157666254 + + ··· (5.45)
8 Z
which yields a value of −2.9037 au, in excellent agreement with experiment.

Problem 5: In this problem, we study how to get the ground state energy of the helium atom
using the variational method. We start with the trial wave function (5.36) with Z as a variational
parameter. Note that this wave function is already normalised. Thus we need to evaluate
Z
E= dr1 dr2 ψ ∗ Ĥψ

with the Hamiltonian defined above. The evaluation of the integral is greatly simplified if one
recalls that the ψ1s is an eigenfunction of a hydrogen-like Hamiltonian with a nucleus having a
charge Z. It is convenient to rewrite this Hamiltonian as
1 Z 1 Z Z −2 Z −2 1
Ĥ = − ∇21 − − ∇2 − + + + (5.46)
2 r1 2 2 r2 r1 r2 r12
where ! !1 !1
1 Z Z3 2
Z2 Z3 2
− ∇21 − e−Zr =− e−Zr (5.47)
2 r1 π 2 π
Show that
"
Z6 Z2 Z2
Z Z
−Z(r1 +r2 )
E(Z) = e − −
π2 2 2
#
Z −2 Z −2 1 −Z(r1 +r2 )
+ + + e dr1 dr2
r1 r2 r12

Z3 e−2Zr 5
Z
2
= −Z + 2(Z − 2) dr + Z
π r 8

27
= Z2 − Z (5.48)
8
Now minimise E with respect to Z and show that

E = −2.8477 au

27
with Zmin = 16 . This value can be interpreted as an effective nuclear charge. The fact that this
value is less than 2 reflects the fact that each electron partially screens the nucleus from the other.
142 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

While the agreement between the above result and experiment looks impressive, it can be mislead-
ing. Consider the ionisation energy (IE) for example, which is the difference between the energy
of a singly ionised helium atom and the energy of the helium atom in its ground state. From our
calculation above we get this to be IE = −2.0−(− 11
4 ) = 0.750 au = 20.4 eV, while the experimental
value is 24.6 eV. This is indeed a big difference and we should be able to do better. One way is to
use a more general form of the trial wave function (still using the variational method). We could
go beyond the one-electron product form used above and use a functional form that includes the
inter-electronic distance explicitly. This was first done by Hylleras in 1930, who used
 
−Zr1 −Zr2
ψ(r1 , r2 , r12 ) = e e 1 + g(r1 , r2 , r12 ) (5.49)

with g chosen to be a polynomial in r1 , r2 , r12 and whose expansion coefficients are variationally
optimised. Using 14 terms for g, he obtained a value of −2.9037 for the ground state energy (exact
to five significant figures). A more extensive calculation using this approach, by Pekeris in 1959
yielded −2.903724375 au with 1078 terms!

While this showed that one could obtain essentially exact energies using the variational method
with more elaborate trial wave functions in which the inter-electronic distance r 12 is incorporated
explicitly, this approach is computationally very demanding and does not lend itself to the study
of large atoms and molecules. More state-of-the-art methods will be discussed later.

5.7 Hydrogen molecule

We now consider the ground state of the hydrogen molecule, the main point of this calculation being
to study the effect of the exchange energy on the chemical bonding between atoms. This energy
is entirely a consequence of the antisymmetry of the electron wave function under the interchange
of the two electrons.

The hydrogen molecule consists of two hydrogen atoms bound together. Since the protons are much
more massive than the electrons, we neglect their motion and treat them as fixed centres of force
(i.e. the Born-Oppenheimer approximation.) With this approximation, the Hamiltonian for a
hydrogen molecule becomes
h̄2 2 e2 h̄2 2 e2 e2 e2 e2 e2
Ĥ(1, 2) = − ∇1 − − ∇2 − + + − − (5.50)
2m r1 2m r2 r12 R r1B r2A
where the first and third term corresponds to the kinetic energies of the electrons; the second and
fourth terms correspond to the potential energy of interaction between electron 1 and nucleus A
5.7. HYDROGEN MOLECULE 143

and electron 2 and nucleus B, respectively; the fifth and sixth term correspond to the potential
energy of inter-electronic and inter-nuclear repulsions; the seventh and eighth term corresponds to
the potential energy of interaction between electron 1 and nucleus B and electron 2 and nucleus A
respectively, (the various coordinates are labelled in the figure below).

e r
1 12
e
2

r1 r2

r r
2A 1B
A B
R

We write the Hamiltonian as

Ĥ = Ĥ0 + Ŵ (5.51)

where Ĥ0 is the “hydrogen atom” part of the Hamiltonian

h̄2 2 e2 h̄2 2 e2
− ∇1 − − ∇ − (5.52)
2m r1 2m 2 r2

and
e2 e2 e2 e2
Ŵ = + − − (5.53)
r12 R r1B r2A
e2 2 e2
Here the two-body potential is r12 , whereas −( er1 + r1B ) is the “external” potential for particle 1
e2 e2 e2
and −( r2 + r2A ) is the “external” potential for particle 2. The repulsive potential R between the
two protons is just a constant i.e. independent of the electron positions. To determine the ground
state energy of the hydrogen molecule, we treat Ŵ as a perturbation. In this case we only need to
find a good approximation for the ground state wave function. The energy is then approximated
by the expectation value of Ĥ in this approximate ground state. For R sufficiently large, we have
two unperturbed hydrogen atoms with the Hamiltonian Ĥ0 , and we use the eigenfunctions of Ĥ0
as our approximate ground state.

Since the electrons are fermions, the total wave function (which must be a direct product of the
spatial and spin parts), must be antisymmetric in the particle labels (which include the spin degrees
of freedom). From the discussion of the addition of angular momenta for two spin− 21 particles in
144 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

Chapter 2, we know that the possible values of the total spin angular momentum for the electrons
are the S = 1 triplet state, χ(t) where



 |1, 1i = | + +i

χ(t) = |1, 0i = √1 (| + −i + | − +i)

 2
 |1, −1i = | − −i


and the S = 0 singlet state, χ(s) with

1
χ(s) = |0, 0i = √ (| + −i − | − +i) (5.54)
2
Note that the triplet state is even under the interchange of the spin coordinates while the singlet
state is odd. The corresponding spatial wave functions are therefore
 
Ψs (r1 , r2 ) = As ψA (1) ψB (2) + ψA (2) ψB (1)

 
Ψt (r1 , r2 ) = At ψA (1) ψB (2) − ψA (2) ψB (1) (5.55)

where As and At are normalisation constants, and the ψA (i), ψB (i) represent hydrogen atom ground
state wave functions centred at A and B respectively. Thus
1 1
ψA (1) = √ e−r1 /a , ψA (2) = √ e−r2A /a
πa3 πa3

1 1
ψB (1) = √ e−r1B /a , ψB (2) = √ e−r2 /a (5.56)
πa 3 πa 3

where a is the Bohr radius, a = h̄2 /2me2 .

Problem 6: Show that the normalisations As and At in equations (5.55) are given by
 − 1
2
2
As = 2(1 + A )

 − 1
2
2
At = 2(1 − A ) (5.57)

where the overlap integral, A, is given by

1
Z
A= dr1 e−(r1 +r1B )/a = hψA |ψB i (5.58)
πa3

To evaluate integral (5.58), we transform to elliptical coordinates

r1 + r1B
ξ =
R
r1 − r1B
η = (5.59)
R
5.7. HYDROGEN MOLECULE 145

and φ (which is the angle of rotation about the line joining the two protons. The volume element
in these coordinates is
R3 2
dr1 = (ξ − η 2 ) dξ dη dφ (5.60)
8
with the range of integration being

1 ≤ ξ < ∞, −1 ≤ η ≤ +1, 0 ≤ φ ≤ 2π (5.61)

Problem 7: Verify (5.60) and, using (5.60) and (5.61), show that
2 
R 1 R
 
A= 1+ + e−R/a (5.62)
a 3 a

To the extent that Ψs and Ψt are good wave functions, the ground state energy of the hydrogen
molecule can be approximated in the singlet and triplet states by
Z
Es (R) = Ψ∗s (r1 , r2 ) Ĥ Ψs (r1 , r2 ) dr1 dr2

Z
Et (R) = Ψ∗t (r1 , r2 ) Ĥ Ψt (r1 , r2 ) dr1 dr2 (5.63)

These integrals can be rewritten with a little algebra in the form

J +K
Es (R) =
1 + A2

J −K
Et (R) = (5.64)
1 − A2
where
e2 e2 e2 e2
Z  
2
J = |ψA (1)ψB (2)| + − − dr1 dr2
r12 R r1B r2A

e2 e2 e2
Z Z
= − |ψA (1)|2 dr1 − |ψB (2)|2 dr2
R r1B r2A

e2
Z
+ |ψA (1)|2 |ψB (2)|2 dr1 dr2 (5.65)
r12
and
e2 e2 e2 e2
Z  
K = ψA (1) ψB (2) + − − ψA (2) ψB (1) dr1 dr2
r12 R r1B r2A

e2 2 e2 e2
Z Z
= A −A ψB (2) ψA (2)dr2 − A ψA (1) ψB (1)dr1
R r2A r1B

e2
Z
+ ψA (1)ψB (2) ψA (2)ψB (1) dr1 dr2 (5.66)
r12
146 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

Problem 8: Work out the steps leading to (5.64) - (5.66).

The integral J (5.65) is called the Coulomb interaction integral and contains the following
terms. The first term gives the Coulomb repulsion of the two protons. The second term gives
the energy due to the interaction of the proton at B with the “charge density” −e|ψ A (1)|2 due
to electron 1, while the third term gives the interaction energy between the proton at A and the
“charge density” −e|ψB (2)|2 due to electron 2. (Since protons, like electrons, are identical particles,
the second and third terms must be equal.) The last term gives the interaction between the two
“charge densities” −e|ψA (1)|2 and −e|ψB (2)|2 centred at A and B. The integral K is called the
exchange integral and results strictly from the indistinguishability of the two electrons. It is this
type of term that gives rise to covalent bonding, ferromagnetism, etc.

Problem 9: The various integrals in (5.65) and (5.66) can be evaluated by the use of elliptical
coordinates.

• (a) Verify that

e2 πa3 R
Z   
2
|ψA (1)| dr1 = 1 − e−2R/a 1 + (5.67)
r1B R a

• (b) The last term in (5.65) can be written as

e2
Z Z
|ψA (1)|2 |ψB (2)|2 dr1 dr2 = ρA (1)ΦB (1) dr1 (5.68)
r12

where
ρA (1) = −e |ψA (1)|2 (5.69)

is the “charge density” due to electron 1 centred at A and ΦB (1) is the “potential” at r1 due
to electron 2 with charge density −e|ψB (2)|2 centred at B. Thus ΦB (1) satisfies the Poisson
equation
∇2 ΦB (r1 ) = 4πe |ψB (2)|2 (5.70)

Evaluate this potential and hence show that

2 3 
e2 5 3 R 1 R
  
J = e−2R/a 1 + − − (5.71)
R 8 4 a 6 a

For the exchange integral K, the terms involving only one integration are again obtained by going
to elliptical coordinates. The last term cannot be expressed in terms of elementary functions. It
5.7. HYDROGEN MOLECULE 147

can, however, be expressed in terms of the exponential integral

∞ e−t
Z
Ei(x) = − dt (5.72)
−x t
whose properties can be found on p.228 of the “Handbook of Mathematical Functions” by Abramowitz
and Stegun (in the Rayleigh Library). The final result for K (which you are welcome to verify but
the details can be found in the paper by Sugiura, Z. Physik, Vol. 45, p.484 (1927)) is:
e2 2 6 R
 
K = A 1 + (C + ln )
R 5 a
2 3 
e2 −2R/a 11 103 R 49 R 11 R
  
+ e + + +
a 8 20 a 15 a 15 a
2  2
6 e2 2R/a R 1 R
 
+ e 1− + Ei(−4R/a)
5R a 3 a
2 
12 R/a R 1 R
 
− Ae 1− + Ei(−2R/a) (5.73)
5 a 3 a
where C = 0.577215 . . . is Euler’s constant. A plot of Es and Et as functions of R is shown below.

E
6

Et

Es
R

Although quantitative agreement with experimental values for the equilibrium bond length and
dissociation energy is not very good, we find that Et is always higher in energy than Es . In fact
148 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

the singlet state is called the bonding state while the triplet state is called the anti-bonding
state. Better agreement with experiment is obtained if we take into account the correlation
between the electrons i.e. the fact that electrons avoid each other as much as possible. One can
“build” this into the ground state wave function by constructing the two-electron wave function as
a function of r1 , r2 and r12 (with variational parameters) and then using the variational principle
to get the best estimate for these parameters.

5.8 Scattering of identical particles

If the particles in the target and the incident beam are identical, it is impossible to tell if we are
observing a scattered beam particle or a recoiling target particle, as shown in the following figure:

1 2

1 2 1 2

2 1

There is no way of distinguishing between a deflection of a particle through an angle θ and a

deflection of (π − θ) in the centre-of-mass frame, since momentum conservation demands that if
one of the particles scatters through θ then the other goes in the direction (π − θ). For classical
scattering, there is an increase in the scattering cross section relative to the non-identical particle
case since the number of counts at a certain detector will be the sum due to both particles. The
consequence is that the classical cross section σcl (θ) is symmetric about θ = π/2:

σcl (θ) = σ(θ) + σ(π − θ) (5.74)

In QM there is no way of distinguishing the two final states. For quantum scattering, interference
occurs at the scattering amplitude level, i.e. between the scattering amplitudes f (θ) and f (π − θ),
which can result in dramatic changes.
5.8. SCATTERING OF IDENTICAL PARTICLES 149

5.8.1 Scattering of identical spin zero bosons

The differential scattering cross section, for the scattering of two identical spin zero bosons (for
example, α-particles) is given by

dσ
= |f (θ) + f (π − θ)|2
dΩ
 
= |f (θ)|2 + |f (π − θ)|2 + 2Re f ∗ (θ)f (π − θ) (5.75)

which differs from the classical result (5.74) by the interference term (the last term on the r.h.s.).
It leads to an enhancement of the differential cross section at θ = π/2

dσ 

= 4 × |f (θ = π/2)|2 (5.76)
dΩ θ=π/2

as compared to the result that would have been obtained without interference:
dσ 

= 2 × |f (θ = π/2)|2 (5.77)
dΩ θ=π/2

Problem 10: Show that in the partial wave analysis of the cross section for boson-boson scattering,
only the even l values contribute to the differential cross section. (This means that it is impossible
to measure or even define odd-l phase shifts for boson-boson scattering).

5.8.2 Scattering of fermions

When the scattering of two spin− 12 particles is considered (for example electron-electron or proton-
proton scattering), then the scattering amplitude should reflect the basic antisymmetry of the total
wave function under the interchange of the two particles. If the two particles are in a spin singlet
state, then the spatial wave function is symmetric (even) and

dσs
= |f (θ) + f (π − θ)|2 (5.78)
dΩ
and we have the same situation as the boson case discussed above. If the two particles are in a spin
triplet state, then the spatial wave function is antisymmetric and
dσt
= |f (θ) − f (π − θ)|2
dΩ
 
2 2 ∗
= |f (θ)| + |f (π − θ)| − 2Re f (θ)f (π − θ) (5.79)
150 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

In the scattering of two unpolarised protons for example (where all spin states are equally likely),
the probability of finding two protons in a triplet state is three times as large as finding them in
the singlet state. Therefore the differential cross section for this case is

dσ 3 dσt 1 dσs
= +
dΩ 4 dΩ 4 dΩ
 
2 2 ∗
= |f (θ)| + |f (π − θ)| − Re f (θ)f (π − θ) (5.80)

Analogously to the boson case discussed in the last problem, the singlet term in (5.80) is a sum
over only even l’s while the triplet term is a sum over the odd l’s. The resulting differential cross
section therefore has mixed symmetry with respect to θ.

5.9 *Modern electronic structure theory

In this section we review the theoretical foundations of some of the computational techniques that
are currently used in electronic structure calculations for real systems. The relative merits of
the different techniques for calculating the electronic structure of atoms, molecules and solids are
considered. The progress made over the past decade in developing quantum Monte Carlo (QMC)
methods as a tool for tackling realistic continuum electronic structure problems is described.

5.9.1 *The many-electron problem

Within the Born-Oppenheimer approximation, the time independent Schrödinger equation for a
fully interacting many-electron system is

N
h̄2 2 X
!
X Zα e2 1 XX e2
ĤΨ = − ∇i − Ψ+ Ψ = EΨ , (5.81)
i=1
2m ri − d~α |
α 4π0 |~ 2 i j6=i 4π0 |~ri − ~rj |

where Ψ is the N -electron wave functions, ~ri are the electron positions, d~α are the positions of the
ions and Zα are the ionic charges. This equation is impossible to solve exactly so approximate
solutions must be sought. One of the main challenges of condensed matter physics is to try to
find good, workable approximations that contain the essence of the physics involved in a particular
problem and to obtain the most accurate solutions possible. For the rest of this chapter all equations
will be written in atomic units, e = me = h̄ = 4π0 = 1.
5.9. *MODERN ELECTRONIC STRUCTURE THEORY 151

5.9.2 *One-electron methods

One of the most common ways of dealing with the many-electron problem is to assume that each
electron can be considered separately. Each electron is treated as moving in a mean field potential,
U (~r). This potential models the effects of all the other particles in the system, as well as any
external potential acting on the system.

The one-electron equations are of the form

1
− ∇2 ψi (~r) + U (~r)ψi (~r) = i ψi (~r) , (5.82)
2
where ψi (~r) is a one-electron wave function and i are Lagrange multipliers which arise from the
fact that the one-electron wave functions are normalised. Choosing an appropriate U (~r) for the
single electron is still a very complicated problem. U (~r) depends upon the interactions between
the electrons and therefore on the one-electron wave functions. Since initially neither of these
quantities, U (~r) or ψi (~r), is known, it is necessary to solve Eq.(5.82) in a self-consistent manner.
What this means is that we make a guess for the ψi , then construct the U (how one does this is
described below) and proceed to solve the above equation to get the new value for ψ i . One then
repeats this procedure until the ψi have converged in some prescribed manner.

5.9.3 *Hartree approximation

This approximation starts from the one-electron equations (5.82). U (~r) is chosen to try to model
the interaction terms in this equation. The ions contribute a potential
X Zα
Uion (~r) = − . (5.83)
α |~r − d~α |
All the other electrons in the system also contribute to the potential. The potential due to the
electrons is approximated by the electrostatic interaction with all the others, which can be written
in terms of the electron density, ρ(~r), as
1
Z
UH (~r) = d~r0 [ρ(~r0 ) − ρi (~r0 )] , (5.84)
|~r − ~r0 |
where the self-interaction potential due to electron i has been removed.

To actually calculate the Hartree potential it is necessary to know the electronic charge distribution
of the system. If the electrons are assumed to be independent of each other, then it is straightforward
to construct ρ(~r) from the single electron eigenstates

|ψi (~r)|2 ,
X
ρ(~r) = (5.85)
i
152 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

where the summation over i includes all occupied states. Using this charge density the total one-
electron potential is
XZ 1
Ui (~r) = Uion (~r) + d~r0 |ψj (~r0 )|2 . (5.86)
j6=i
|~r − ~r0 |
The potential Ui (~r) is different for each orbital, and therefore the orbitals are not orthogonal. Note
that Ui (~r) depends on all the other orbitals, ψj , and so the solution of Eq.(5.82) must be found
self-consistently.

The choice of U (~r) in Eq.(5.86) all seems a bit like guesswork, but it can also be derived using
the variational principle. If the electrons are assumed to be non-interacting, and so the N -electron
wave function is just the product of the one-electron wave functions,

Ψ = ψ1 (~r1 )ψ2 (~r2 ) . . . ψN (~rN ) . (5.87)

then using the variational principle, by minimising the expectation of Ĥ with respect to the single-
particle orbitals, one can derive the Hartree equations:
1 |ψj (~r0 )|2
  Z
− ∇2 + Uion (~r) ψi (~r) + d~r0
X
ψi (~r) = i ψi (~r) , (5.88)
2 j(6=i)
|~r − ~r0 |

5.9.4 *Hartree-Fock approximation

The Hartree-Fock approximation is an extension of the above Hartree approximation to include the
permutation symmetry of the wave function, which leads to the exchange interaction. Exchange is
due to the Pauli exclusion principle, which states that the total wave function for the system must
be antisymmetric under particle exchange. This means that when two arguments are swapped the
wave function changes sign as follows:

Ψ(~x1 , ~x2 , . . . , ~xi , . . . , ~xj , . . . , ~xN ) = −Ψ(~x1 , ~x2 , . . . , ~xj , . . . , ~xi , . . . , ~xN ) , (5.89)

where ~xi includes coordinates of position and spin. Therefore no two electrons can have the same
set of quantum numbers, and electrons with the same spin cannot occupy the same spatial state
simultaneously.

Instead of using the simple product form of the wave function shown in Eq.(5.87), a Slater deter-
minant wave function which satisfies antisymmetry is used
 
 ψ1 (~ x1 ) ψ1 (~x2 ) ··· ψ1 (~xN ) 
 
 
 ψ2 (~ x1 ) ψ2 (~x2 ) · · · ψ2 (~xN ) 

D =  .. ..

 , (5.90)

 . . 

 ψN (~ x1 ) ψN (~x2 ) · · · ψN (~xN ) 
 
5.9. *MODERN ELECTRONIC STRUCTURE THEORY 153

where ψi (~xj ) are the one-electron wave functions.

Following exactly the same method of minimising the expectation value of Ĥ with respect to the
one-electron wave functions as was used in the derivation of the Hartree equations, results in the
following set of one-electron equations, the Hartree-Fock equations;

1 |ψj (~r0 )|2

  Z
− ∇2 + Uion (~r) ψi (~r) + d~r0
X
i ψi (~r) = ψi (~r)
2 j
|~r − ~r0 |

Z
ψj∗ (~r0 )ψi (~r0 )
d~r0
X
− δ si sj ψj (~r) , (5.91)
j
|~r − ~r0 |

where si labels the spin of particle i. Note the self-interaction cancels out from the second and third
terms. The extra term in these equations, when compared to Eq.(5.88), is known as the exchange
term and is only non-zero when considering electrons of the same spin. The effect of exchange on
the many-body system is that electrons of like spin tend to avoid each other. As a result of this,
each electron has a “hole” associated with it which is known as the exchange hole (or the Fermi
hole). This is a small volume around the electron which like-spin electrons avoid. The charge
contained in the exchange hole is positive and exactly equivalent to the absence of one electron.

Unlike all the other terms acting on ψi , the exchange term is a non-local integral operator and
this makes the Hartree-Fock equations hard to solve in all but a few special cases.

Hartree-Fock calculations, which include the exchange interaction between electrons, are most useful
for performing calculations on relatively small systems as they are considerably more computation-
ally expensive than Hartree and DFT-LDA calculations (see below), due to the non-local exchange
term. Even for atoms, however, Hartree-Fock theory is not ideal. For example, H − is predicted to
be unstable in contradiction to reality.

Various improvements to Hartree-Fock theory have been attempted. Unrestricted Hartree-Fock

theory ignores some of the simplifying restrictions which are normally applied to Hartree-Fock
wave functions. The exchange interaction is allowed to make the spatial parts of spin up and spin
down electron wave functions different for the same state. However, although for some systems
this results in an improvement, especially for open shell systems, it also sometimes produces worse
results than conventional Hartree-Fock theory. In general, Hartree-Fock theory is most useful as a
tool for providing qualitative answers. It is also used as the starting point for other methods, such
as some quantum Monte Carlo calculations.
154 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

5.9.5 *Density functional methods

Computations based on density functional structure theory are the most popular approach to
determining the electronic properties of real materials in physics, chemistry and material science.
Density functional theory is a formally exact theory based on the charge density of a system.
Kohn-Sham density functional theory is a formally exact one-electron theory. Working within the
Born-Oppenheimer approximation, the many-body Schrödinger equation,
N
!
1 Zα 1 XX 1
− ∇2i −
X X
ĤΨ = Ψ+ Ψ = EΨ , (5.92)
i=1
2 ri − d~α |
α |~ 2 i j6=i |~ri − ~rj |

where Ψ is the many-body wave function, is replaced by a set of N one-electron equations of the
form
1
 
− ∇2 + V (~r) ψi (~r) = ψi (~r) , (5.93)
2
where ψi (~r) is a single-electron wave function. These one-electron equations contain a potential
V (~r) produced by all the ions and the electrons. Density functional theory properly includes all
parts of the electron-electron interaction, i.e. the Hartree potential
ρ(~r0 )
Z
VH (~r) = d~r0 , (5.94)
|~r − ~r0 |
where ρ is the charge density of all the electrons, a potential due to exchange and correlation effects,
VXC (~r), and the external potential due to the ions, Vext (~r),

V (~r) = Vext (~r) + VH (~r) + VXC (~r) . (5.95)

Hohenberg and Kohn (Phys. Rev. 136, B864 (1964)) originally developed density functional theory
for application to the ground state of a system of ‘spinless fermions’. In such a system the particle
density is given by Z
ρ(~r) = N |Ψo (~r, ~r2 , . . . , ~rN )|2 d~r2 . . . d~rN , (5.96)

with Ψ0 being the many-body ground state wave function of the system. It can be shown that
the total ground state energy of the system is a functional of the density, E[ρ(~r)], and that if the
energy due to the electron-ion interactions is excluded the remainder of the energy is a universal
functional of the density, F [ρ(~r)] (i.e. F [ρ(~r)] does not depend on the potential from the ions).

Kohn and Sham (Phys. Rev. 140, A1133 (1965)) introduced a method based on the Hohenberg-
Kohn theorem that enables one to minimise the functional E[ρ(~r)] by varying ρ(~r) over all densities
containing N electrons. This constraint is introduced by the Lagrange multiplier, µ, chosen so that
R
ρ(~r)d~r = N ,
δ
 Z 
E[ρ(~r)] − µ ρ(~r)d~r = 0
δρ(~r)
5.9. *MODERN ELECTRONIC STRUCTURE THEORY 155

δE[ρ(~r)]
⇒ =µ . (5.97)
δρ(~r)
Kohn and Sham chose to separate F [ρ(~r)] into three parts, so that E[ρ(~r)] becomes
1 ρ(~r)ρ(~r0 )
Z Z Z
E [ρ(~r)] = Ts [ρ(~r)] + d~rd~r0 + EXC [ρ(~r)] + ρ(~r)Vext (~r)d~r , (5.98)
2 |~r − ~r0 |
where Ts [ρ(~r)] is defined as the kinetic energy of a non-interacting electron gas with density ρ(~r),
N
1X
Z
Ts [ρ(~r)] = − ψi∗ (~r)∇2 ψi (~r)d~r . (5.99)
2 i=1

Eq.(5.98) also acts as a definition for the exchange-correlation energy functional, E XC [ρ(~r)]. We
can now rewrite Eq.(5.97) in terms of an effective potential, Veff (~r), as follows
δTs [ρ(~r)]
+ Veff (~r) = µ , (5.100)
δρ(~r)
where
ρ(~r0 )
Z
Veff (~r) = Vext (~r) + d~r0 + VXC (~r) , (5.101)
|~r − ~r0 |
and
δEXC [ρ(~r)]
VXC (~r) = . (5.102)
δρ(~r)

Now, if one considers a system that really contained non-interacting electrons moving in an external
potential equal to Veff (~r), as defined in Eq.(5.101), then the same analysis would lead to exactly
the same Eq.(5.100). Therefore, to find the ground state energy and density, E 0 and ρ0 (~r) all one
has to do is solve the one-electron equations
1
 
− ∇2i + Veff (~r) − i ψi (~r) = 0 . (5.103)
2
As the density is constructed according to
N
|ψi (~r)|2 ,
X
ρ(~r) = (5.104)
i=1

these equations (5.101-5.103) must be solved self-consistently with Eq.(5.104).

The above derivation assumes that the exchange-correlation functional is known. At present nu-
merical exchange-correlation potentials have only been determined for a few simple model systems,
and so most current density functional calculations use the local density approximation (LDA).
The LDA approximates the XC functional by a simple function of the density at any position, ~r.
The value of this function is the XC energy per electron in a uniform homogeneous electron gas of
density n(~r). The LDA expression for EXC [n(~r)] is
Z
EXC [n(~r)] ≈ XC (n(~r))n(~r)d~r . (5.105)
156 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

The LDA is only expected to be accurate for systems with slowly varying electronic charge densities,
which is not the case in most real systems, but despite this it has been surprisingly successful. The
success is currently understood to be due to two main features:

• The sum rule on the exchange-correlation hole is conserved, i.e. within the LDA, the exchange-
correlation hole contains exactly an equal and opposite amount of charge to the electron it
surrounds, and

• The exchange-correlation energy only depends on the spherical average of the exchange-
correlation hole, i.e.
1 ρ(~r) ρXC (~r, |~r − ~r0 |)
Z Z
EXC [ρ] = d~r d~r0 , (5.106)
2 |~r − ~r0 |
where ρXC (~r, |~r − ~r0 |) is the spherical average of ρXC (~r, ~r0 ) about ~r0 = ~r for each ~r.

Therefore, the fact that in the LDA ρXC is constrained to be spherically symmetric about ~r 0 = ~r
is not a handicap.

The LDA is remarkably accurate, but often fails when the electrons are strongly correlated, as
in systems containing d and f orbital electrons. In strongly correlated systems, the correlations
may change the whole nature of the ground state and the local density approximation, derived
from homogeneous electron gas results, is not successful. For example, the high T c superconductor
La2 CuO4 is an anti-ferromagnetic insulator but the LDA finds it to be metallic. Also FeO, MnO
and NiO all have Mott metal-insulator transitions but the LDA predicts that they are either
semiconductors or metals. Other failings of the LDA are that it tends to overbind atoms in solids,
that it finds stable negative ions to be unstable and that it predicts iron to be fcc paramagnetic,
when it is actually bcc ferromagnetic.

5.9.6 *Shortcomings of the mean-field approach

The main problem with Hartree, Hartree-Fock and LDA methods is the approximations they in-
troduce in the process of reducing the many-body problem to a one-electron problem. Hartree and
Hartree-Fock calculations do not, in general, provide satisfactory results and are best used as a
qualitative guide to the expected ground state properties. The configuration interaction method
(popular in quantum chemistry), while in principle exact, is in practice only useful for small systems;
for condensed matter systems it is not of practical value.

Density functional theory within the LDA provides the current staple method of performing elec-
tronic structure calculations and for many purposes gives good results. However, it fails for highly
5.9. *MODERN ELECTRONIC STRUCTURE THEORY 157

correlated systems and tends to underestimate band gaps and overestimates cohesive energies and
hence is not ideal.

A straightforward, accurate approach to many-body systems that is now gaining popularity in the
electronic structure theory community is the quantum Monte Carlo method, described in the next
section.

5.9.7 *Quantum Monte Carlo methods

Applying QMC techniques to study the electronic structure of real systems is relatively new. The
variational principle provides the starting point with the aim of finding an approximate solution
to Schrödinger’s equation. As we saw in Chapter 3, the expectation value of a Hamiltonian, Ĥ,
calculated using a trial wave function, ΨT , is never lower in value than the true ground state energy,
0 , which is the expectation value of Ĥ calculated using the true ground state wave function, Ψ0 .
The success of the method depends on making a physically plausible guess at the form of the
ground state wave function, ΨT , of the Hamiltonian, Ĥ. This guess will be referred to as the
trial/guiding wave function. The “trial” part of the name refers to the use of the wave function as
a guess of the true ground state wave function to be used as the input wave function in a variational
quantum Monte Carlo (VMC) calculation. The “guiding” part refers to the use of the same wave
function as an input wave function in the diffusion quantum Monte Carlo (DMC) algorithm as part
of the mechanism to introduce importance sampling. The trial/guiding wave function depends
on a number of variable parameters which can be adjusted to minimise the energy expectation
value. If the guessed values of these parameters are good and the chosen functional form builds in
enough variational freedom to adequately describe the physics of the system being studied, then
very accurate estimates of the ground state energy can be obtained. Variational quantum Monte
Carlo (VMC) calculations are direct applications of the above variational principle.

The variational quantum Monte Carlo (VMC) method is the simpler of the two quantum Monte
Carlo methods. It is based on a combination of the variational principle and Monte Carlo evaluation
of integrals using importance sampling based on the Metropolis algorithm.

Within the Born-Oppenheimer approximation, the Hamiltonian for a many-body system can be
written as
N
1 Zα 1 XX 1 1 X X Zα Zβ
− ∇2i −
X XX
Ĥ = + + . (5.107)
i=1
2 i ri − d~α | 2 i j6=i |~ri − ~rj | 2 α β6=α |d~α − d~β |
α |~

The VMC method relies on one being able to construct a trial wave function, Ψ T , that is a reason-
ably good approximation to the true ground state wave function, Ψ0 . The energy associated with
158 CHAPTER 5. IDENTICAL PARTICLES IN QUANTUM MECHANICS

the trial wave function is given by

hΨT |Ĥ|ΨT i
ET =
hΨT |ΨT i

Z ~ ĤΨT (R)d
Ψ∗T (R) ~ R ~
= . (5.108)
~ T (R)d
Ψ∗ (R)Ψ ~ R~
T

~ denotes the 3N -dimensional vector of electronic positions. The variational principle ensures
where R
that ET is a rigorous upper bound to the true ground state energy, E0 .

The VMC method is a Monte Carlo method for evaluating the multi-dimensional integral in
Eq.(5.108). This is achieved by rewriting Eq.(5.108) in the following form:
 2 ~
~  ĤΨT (R)
Z ΨT (R) ~
~ dR
ΨT (R)
ET = R  2 . (5.109)
~ 
ΨT (R)

~ in configuration space. At each

The Metropolis algorithm is used to sample a series of points, R,
~
ĤΨT (R)
of these points the “local energy”, ~ , is evaluated. After a sufficient number of evaluations
ΨT (R)
of the local energy have been made, the average is taken:
N ~i)
1 X ĤΨT (R
EV M C = −→ ET . (5.110)
~i)
N i=1 ΨT (R N →∞

~ i are sampled
(The Metropolis algorithm is a method of ensuring that, in the limit of large N , the R
~ 2 .)
from |ΨT (R)|
Chapter 6

Density Operators

6.1 Introduction

A quantum-mechanical wave function (or state vector), when it exists, conveys the maximum
amount of information permitted by quantum mechanics concerning the properties of a physical
system in the state described by the wave function. Situations in which we have accurate wave
functions for a physical system are actually quite rare. More often, the complexity of the system
owing to its many degrees of freedom precludes the possibility of constructing a wave function.
It is then necessary to resort to statistical methods. When the state of an incompletely prepared
system is only partially known, we resort to assigning a probability to all possible state vectors that
the system could be in. The synthesis of this statistical nature with the probabilities arising from
the quantum mechanics of state vectors can be made using a mathematical entity, the density
operator. The density operator formalism was introduced independently by Landau and von
Neumann, and enables us to reformulate the laws of quantum mechanics more generally than with
the formalism using state vectors or wave functions alone. All predictions – of a statistical nature –
that one can make at a given time about a physical system can be found once we know its density
operator. Furthermore, the density operator formalism enables us to extend quantum mechanics to
the description of statistical mixtures representing systems which are not well known or to describe
the state of a part of a quantum system (i.e. a subsystem). Here we will examine the properties of
this operator and its application to quantum statistical mechanics.

159
160 CHAPTER 6. DENSITY OPERATORS

6.2 Pure and mixed states

In classical mechanics, the dynamical state of a system is completely determined once the values of
the positions and momenta of all the particles are known. The state of the system at any subsequent
time can be predicted with certainty (since all one needs to do is integrate the equations of motion).
In quantum mechanics however this is not the case. A precise simultaneous measurement of two
physical variables is only possible if the two operator corresponding to the two variables commute.
The largest set of mutually commuting independent observables, {Â, B̂, . . .} that can be found will
give the most complete characterisation possible. (This is just the complete set of commuting
observables (CSCO) discussed in Chapter 1.) The measurement of another variable whose
operator is not contained in the above set of operators necessarily introduces uncertainty into at
least one of those already measured. This means that it is not possible to give a more complete
specification of the system. In general, the maximum information which can be obtained on a system
consists of the eigenvalues of the CSCO. The system is then completely specified by assigning the
state vector |a, b, . . .i in a Hilbert space H to it. If the measurement of the observables { Â, B̂, . . .} on
the state |a, b, . . .i is immediately repeated, we get the same values a, b, . . . again. The existence of
such a set of experiments (for which the results can be predicted with certainty) gives a necessary
and sufficient characterisation for the state of “maximum knowledge”. The states of maximum
knowledge are called pure states. Pure states represent the ultimate limit of precise observation
as permitted by the uncertainty principle and are the quantum mechanical analog of classical states
where all positions and momenta of all particles are known.

In practice, the state of a system is not pure and cannot be represented by a single state vector.
However, it can be described by stating that the system has certain probabilities p 1 , p2 , . . . of being
in the pure states |Ψ1 i, |Ψ2 i, . . ., respectively. Therefore in the case of incomplete knowledge about
the state of the system, it is necessary to use a statistical description in the same sense as classical
statistical mechanics. Systems which cannot be characterised by a single-state vector are called
mixed states.

Consider an ensemble of particles in the pure state |Ψi. If this state is not one of the eigenstates of
the observable  then measurements of the corresponding physical quantity will produce a variety
of results, each of which is an eigenvalue of Â. If similar measurements are made on a very large
number of particles, all of which were in the same state |Ψi, then, in general, all the possible
eigenvalues of |Ψi would be obtained. The average of these values is given by the expectation
value hÂi of the observable corresponding to  which is defined by the matrix element

hÂi = hΨ|Â|Ψi (6.1)

6.3. PROPERTIES OF THE DENSITY OPERATOR 161

where we have assumed that |Ψi is normalised.

In order to obtain hÂi for a mixture of states, |Ψ1 i, |Ψ2 i, . . ., the expectation values hΨi |Â|Ψi i of
each of the pure state components must be calculated and then averaged by summing over all pure
states multiplied by its corresponding statistical weight pi :
X
hÂi = pi hΨi |Â|Ψi i (6.2)
i

where we have again assumed that the |Ψn i are normalised. Note that statistics enter into Eq.
(6.2) in two ways: First of all in the quantum mechanical expectation value hΨ i |Â|Ψi i and secondly
in the ensemble average over these values with the weights pi . While the first type of averaging is
connected with the perturbation of the system during the measurement (and is therefore inherent in
the nature of quantisation), the second averaging is introduced because of the lack of information as
to which of the several pure states the system may be in. This latter averaging closely resembles that
of classical statistical mechanics and it can be conveniently performed by using density operator
techniques.

6.3 Properties of the Density Operator

The density operator is defined by

X
ρ̂ = pi |Ψi ihΨi | (6.3)
i

where pi is the probability of the system being in the normalised state |Ψi i and the sum is over all
states that are accessible to the system. The probabilities pi satisfy

p2i ≤ 1
X X
0 ≤ pi ≤ 1, pi = 1, (6.4)
i i

For a pure state there is just one pi (which is equal to unity) and all the rest are zero. In that case

ρ̂ = |ΨihΨ| (pure state) (6.5)

Let {|ψi i} be a complete orthonormal set which serves as a basis for the expansion of |Ψ i i (and
from which we can construct the matrix representation of state vectors and operators). We have
X
|Ψi i = cni |ψn i (6.6)
n

and from the orthonormality of the {|ψi i},

cni = hψn |Ψi i (6.7)

162 CHAPTER 6. DENSITY OPERATORS

We now construct the density matrix which consists of the matrix elements of the density operator
in the {|ψi i} basis:
X
hψn |ρ̂|ψm i = pi hψn |Ψi ihΨi |ψm i
i

pi cni c∗mi
X
= (6.8)
i

which characterises ρ̂ as a Hermitian operator since

hψn |ρ̂|ψm i = hψm |ρ̂|ψn i∗ (6.9)

(given that the pi are real), i.e we have

ρ̂ = ρ̂† (6.10)

From Eq. (6.8), the probability of finding the system in the state |ψn i is given by the diagonal
element
pi |cni |2
X
hψn |ρ̂|ψn i = (6.11)
i

which gives a physical interpretation of the diagonal elements of the density operator. Because
probabilities are positive numbers, we have

hψn |ρ̂|ψn i ≥ 0 (6.12)

The trace of ρ̂ (i.e. the sum of the diagonal matrix elements) is

X
Tr ρ̂ = hψn |ρ̂|ψn i
n

XX
= pi hψn |Ψi i hΨi |ψn i
i n

X
= pi hΨi |Ψi i
i

X
= pi
i

= 1 (6.13)

(Since the trace of an operator is an invariant quantity, the above result is independent of the basis.)
As ρ̂ is Hermitian, the diagonal elements hψn |ρ̂|ψn i must be real and from Eq. (6.8) it follows that

0 ≤ hψn |ρ̂|ψn i ≤ 1 (6.14)

6.3. PROPERTIES OF THE DENSITY OPERATOR 163

Note that for a pure state, hψn |ρ̂|ψn i = |cn |2 , which is the probability of finding the system in the
state ψn .

Consider the matrix elements of ρ̂2 :

hψn |ρˆ2 |ψm i =

X
hψn |ρ̂|ψk i hψk |ρ̂|ψm i
k

XXX
= pi pj hψn |Ψi ihΨi |ψk i hψk |Ψj ihΨj |ψm i (6.15)
i j k

where we have used Eq. (6.3).

Problem 1: Using (6.15), show that

Tr ρ̂2 ≤ 1 (6.16)

For a pure state, there is only one pi and it is equal to unity. Therefore

Tr ρ̂2 = 1 (pure state) (6.17)

and

ρ̂2 = |ΨihΨ|ΨihΨ|

= |ΨihΨ|

= ρ̂ (pure state) (6.18)

i.e. ρ̂ is idempotent for a pure state. Thus whether a state is pure or not can be established by
testing whether (6.17) or (6.18) is satisfied or not.

We now derive the expectation value of an operator  for pure as well as mixed states. Let

hÂii = hΨi |Â|Ψi i (6.19)

and
X
hÂi = pi hÂii (6.20)
i

The distinction between hÂii and hÂi is that the former is a quantum-mechanical average or the
expectation value of an operator  when the system is definitely in the state |Ψi i. On the other
hand, hÂi is a statistical or ensemble average which from (6.20), is seen to be the weighted average
of hÂii taken over all states that the system may occupy. For pure states, we have

hÂi = hÂii (pure state) (6.21)

164 CHAPTER 6. DENSITY OPERATORS

Now consider the operator ρ̂Â. From (6.3) we have

X
ρ̂Â = pi |Ψi ihΨi | Â (6.22)
i

In the {|ψi i} basis,

X
hψn |ρ̂Â|ψm i = pi hψn |Ψi ihΨi |Â|ψm i (6.23)
i

Taking the trace of ρ̂Â,

X
Tr ρ̂Â = hψn |ρ̂Â|ψn i
n

XX
= pi hψn |Ψi ihΨi |Â|ψn i
i n

X
= pi hΨi |Â|Ψi i
i

= hÂi (6.24)

Thus the average value of an operator for a system in either a pure or mixed state, is known as soon
as the density operator is known. Therefore the density operator contains all physically
significant information on the system.

To summarise, the density operator ρ̂ has the following properties:

• ρ̂ is Hermitean: ρ̂ = ρ̂† . This follows from the fact that the pi are real. This property means
that the expectation value of any observable is real.

• ρ̂ has unit trace: Tr ρ̂ = 1.

• ρ̂ is non-negative: hΦ|ρ̂|Φi ≥ 0 ∀ |Φi ∈ H

• The expectation value of an operator  is given by hÂi = Tr ρ̂Â.

6.3.1 Density operator for spin states

Suppose the spin state of an electron is given by

|Ψi = | ↑i (6.25)

so that the density operator is

ρ̂ = | ↑ih↑ | (6.26)
6.3. PROPERTIES OF THE DENSITY OPERATOR 165

In the basis {| ↑i, | ↓i} (i.e. the eigenstates of Ŝz , the z-component of the spin angular momentum
of the electron), the density matrix is !
1 0
ρ̂ = (6.27)
0 0

Problem 2: Verify (6.27) and hence show that the expectation values of the operators Ŝx , Ŝy , Ŝz
are 0,0 and 21 h̄ respectively.

More generally, if the electron is in a state described by

|Ψi = a1 | ↑i + a2 | ↓i (6.28)

with
|a1 |2 + |a2 |2 = 1

the density operator is

|a1 |2 a1 a∗2
!
ρ̂ = (6.29)
a2 a∗1 |a2 |2
which indicates that the diagonal elements |a1 |2 and |a2 |2 are just the probabilities that the electron
is the state | ↑i and | ↓i respectively.

Another useful form for the density matrix for spin- 21 particles is obtained by writing

ρ̂ = c0 I + c1 Ŝx + c2 Ŝy + c3 Ŝz (6.30)

where I is the unit 2 × 2 matrix and the ci ’s are real numbers. The density matrix becomes
c0 + 12 c3 1
!
2 (c1 − ic2 )
ρ̂ = 1
(6.31)
2 (c1 + ic2 ) c0 − 12 c3
(where we have set h̄ = 1).

Problem 3: Verify (6.31) using the definition of the spin operators in terms of the Pauli matrices.
1
Show that c0 = 2 and the expectation values of Ŝx , Ŝy , Ŝz are given by 12 c1 , 12 c2 , 12 c3 respectively.
Hence show that the density operator can be written compactly as
1
 
ρ̂ = I + hσ̂i · σ
2
where σ = (σx , σy , σz ) is the vector whose components are the Pauli matrices.

Problem 4: By analogy with the polarisation of the spin- 12 case discussed in this section,
!
the
a
polarisation of a light quantum can be described by a two-component wave function , where
b
166 CHAPTER 6. DENSITY OPERATORS

|a|2 and |b|2 are the probabilities that the photon is polarised in one or the other of two mutually
perpendicular directions (or that the photon is right- or left-hand circularly polarised). If we want
to determine the polarisation of a photon, we could, for instance, use a filter, which we shall call a
detector (although strictly speaking it is not a detector but a device to prepare for a measurement).
Such a filter could correspond to a pure state, described by a wave function

Ψdet = cdet det

1 Ψ1 + c2 Ψ2

where ! !
1 0
Ψ1 = , Ψ2 = (6.32)
0 1
are the wave functions corresponding to the two polarisation states. This pure state corresponds
to a 2 × 2 detector density matrix ρ̂det given by its matrix elements

ρdet det det ∗

ij = ci · (cj )

Find an expression for the probability of a response of a detector described by ρ̂ det to a photon in
a state described by a density matrix ρ̂.

6.3.2 Density operator in the position representation

The density operator in the position representation is defined by

ρ(x0 , x) = hx0 |ρ̂|xi

pi Ψi (x0 ) Ψ∗i (x)

X
= (6.33)
i
which, for a pure state becomes

ρ(x0 , x) = Ψ(x0 ) Ψ∗ (x) (pure state) (6.34)

The expectation value for an operator  is then given by

hÂi = Tr ρ̂ Â
Z
= dx hx|ρ̂Â|xi

Z Z 
0 0 0
= dx hx|ρ̂ dx |x ihx | Â|xi

Z Z
= dx0 dx hx|ρ̂|x0 i hx0 |Â|xi

Z Z
= dx0 dx ρ(x, x0 ) A(x0 , x) (6.35)
6.3. PROPERTIES OF THE DENSITY OPERATOR 167

Problem 5: Show that

(a) When  = x̂ , i.e. the position operator, then

Z
hx̂i = dx x ρ(x, x)

(b) When  = p̂ , i.e. the momentum operator, then

h̄ ∂
Z  
hp̂i = − dx ρ(x, x0 )
i ∂x x0 =x

Problem 6: Often one is dealing with a system which is part of a larger system. Let x and q
denote, respectively, the coordinates of the smaller system and the coordinates of the remainder of
the larger system. The larger system will be described by a normalised wave function Ψ(x, q) which
cannot necessarily be written as a product of functions depending on x and q only. Let  be an
operator acting only on the x variables, let Ĥ be the Hamiltonian describing the smaller system,
and let the density operator ρ̂ be defined in the position representation by the equation
Z
0
hx|ρ̂|x i = Ψ∗ (q, x0 ) Ψ(q, x) dq (6.36)

where the integration is over all the degrees of freedom of the remainder of the larger system.

(a) Express the expectation value of  in terms of ρ̂ for the case where the larger system is
described by the wave function Ψ(q, x).

(b) What is the normalisation condition for ρ̂?

(c) Find the equation of motion for ρ̂.

Problem 7: If the wave function Ψ(q, x) of the preceding problem can be written in the form

Ψ(q, x) = Φ(q) χ(x) (6.37)

we are dealing with a pure state. Prove that the necessary and sufficient condition for the pure
state is that ρ̂ is idempotent, i.e. that
ρ̂2 = ρ̂ (6.38)
168 CHAPTER 6. DENSITY OPERATORS

6.4 Density operator in statistical mechanics

Let {Ψn } be a complete set of orthonormal functions (eigenstates of the Hamiltonian Ĥ for the
system), that satisfy
ĤΨn = En Ψn (6.39)

and let the states be occupied according to the Boltzmann distribution

pn = N e−βEn (6.40)

where pn is the probability of finding the system in the eigenstate Ψn with energy En , β = 1/kT
with k the Boltzmann constant, T the absolute temperature and N a normalisation constant chosen
to ensure that
X
pn = 1
n

Condition (6.40) defines thermal equilibrium and the corresponding density operator ρ̂(β) is
(from (6.3)):
e−βEn |Ψn ihΨn |
X
ρ̂(β) = N (6.41)
n
Since
e−β Ĥ |Ψn i = e−βEn |Ψn i (6.42)

this enables us to write

e−βEn |Ψn ihΨn |

X
ρ̂(β) = N
n

e−β Ĥ |Ψn ihΨn |

X
= N
n

= N e−β Ĥ
X
|Ψn ihΨn |
n

= N e−β Ĥ (6.43)

To determine N , we note that

Tr ρ̂(β) = N Tr e−β Ĥ

= 1 (6.44)

which implies
1
N= (6.45)
Tr e−β Ĥ
6.4. DENSITY OPERATOR IN STATISTICAL MECHANICS 169

Hence the density operator under thermal equilibrium is

e−β Ĥ
ρ̂(β) =
Tr e−β Ĥ

1 −β Ĥ
= e (6.46)
Z
where
Z = Tr e−β Ĥ (6.47)

is known as the canonical partition function. (Note that the partition function is a function of
absolute temperature, T , the volume V , and the number of particles that make up the system, N .)
We see that from the knowledge of the density operator in any representation, one can determine
the partition function and therefore all thermodynamic properties of the system. For instance, the
average of an observable  is given by

hÂi = Tr ρ̂ Â
 
Tr e−β Ĥ Â
=   (6.48)
Tr e−β Ĥ

The mean energy of the system (i.e. the internal energy) is given by U where

U = hĤi
 
Tr e−β Ĥ Ĥ
=  
Tr e−β Ĥ

∂
 
= − ln Tr e−β Ĥ
∂β

∂
= − ln Z(T, V, N ) (6.49)
∂β
From the partition function, we obtain all thermodynamic observables:

S = −k Tr (ρ̂ ln ρ̂) (entropy)

= k β hĤi + k ln Z(T, V, N )

F = U − TS (Helmholtz free energy)

170 CHAPTER 6. DENSITY OPERATORS

= −kT ln Z(T, V, N )
  
−β Ĥ
= −kT ln Tr e (6.50)

We now calculate the density operator (in various representations) for some concrete cases.

6.4.1 Density operator for a free particle in the momentum representation

We determine the density operator in the momentum representation for a free particle in a box of
volume L3 with periodic boundary conditions. The Hamiltonian is given by Ĥ = p̂2 /2m and the
energy eigenfunction are plane waves;

Ĥ|ψk i = E|ψk i (6.51)

with
h̄2 k2
E= (6.52)
2m
and |ψk i defined by

1
ψk (r) = √ eik·r
V

2π
k = (nx , ny , nz )
L

ni = 0, ±1, ±2, . . . (6.53)

Note that the energy eigenvalues are discrete but their mutual separation for macroscopic volumes
is so small that one may treat them as essentially continuous. The advantage of the formulation
using a box and periodic boundary conditions is that one has automatically introduced into the
formalism a finite volume for the particles, which is not the case for free plane waves we have used
so far (in scattering theory for example). The functions ψk (r) are orthonormalized,

hψk0 |ψk i = δk,k0

= δnx0 ,nx δny0 ,ny δnz ,nz (6.54)

and complete,
ψk∗ (r0 )ψk (r) = δ(r0 − r)
X
(6.55)
k
6.4. DENSITY OPERATOR IN STATISTICAL MECHANICS 171

The canonical partition function is

Z(T, V, 1) = Tr e−β Ĥ

hψk |e−β Ĥ |ψk i

X
=
k

βh̄2 2
e− 2m k
X
= (6.56)
k

Since the eigenvalues k are very close together in a large volume, we can replace the sum in (6.56)
by an integral.
V βh̄2
Z
2
Z(T, V, 1) = dk e− 2m k
(2π)3
3/2
V 2mπ

=
(2π)3 βh̄2

V
= (6.57)
λ3
where λ is called the thermal wavelength. The matrix elements of the density operator thus becomes

λ3 − βh̄2 k2
hψk0 |ρ̂|ψk i = e 2m δk,k0 (6.58)
V
which is a diagonal matrix.

6.4.2 Density operator for a free particle in the position representation

We look for the canonical density operator in the position representation for a free particle in a
box of volume V and periodic boundary conditions. We have

hr0 |ρ̂|ri = hr0 |k0 i hk0 |ρ̂|ki hk|ri

X

k,k0

λ3 − βh̄2 k2
 
ψk0 (r0 ) e 2m δk,k0 ψk∗ (r)
X
=
k,k0
V

λ3 1 βh̄2 2
Z  
= dk exp − k + ik · (r0 − r) (6.59)
V (2π)3 2m
172 CHAPTER 6. DENSITY OPERATORS

Problem 8: Show that Eq. (6.59) reduces to

3/2
λ3 1 m 2mπ
 
hr0 |ρ̂|ri = exp − (r0 − r)2
V (2π)3 2βh̄2 βh̄2

1 π 0
 
= exp − (r − r)2 (6.60)
V λ2

Hence in the position representation, the density matrix is no longer a diagonal matrix, but a Gaus-
sian function in (r0 − r). The diagonal elements of the density matrix in the position representation
can be interpreted as the density distribution in position space i.e.
1
hr0 |ρ̂|ri = ρ(r) = (6.61)
V
The non-diagonal elements r 6= r0 can be interpreted as the transition probability of the particle
to move from a position r to a new position r0 (though these transitions are restricted to spatial
regions having the size of the thermal wavelengths.) For large temperatures (λ → 0) this is hardly
observable, but for low temperatures λ may become very large, which implies that quantum effects
play an especially large role at low temperatures.

6.4.3 *Density matrix for the harmonic oscillator

Here, we determine the density matrix for the one-dimensional quantum harmonic oscillator in the
position representation. This result is of great importance in quantum statistical mechanics and
the mathematical steps involved in deriving the final result carry over to other areas of theoretical
physics. We shall use the expression for the energy eigenfunction in the position representation
derived before:
1/4
Mω H (x) 1
  
Ψn (q) = √n exp − x2
πh̄ 2n n! 2
s
Mω
x = q (6.62)
h̄
and the energy eigenvalues are En = h̄ω(n + 21 ), and the Hermite polynomials are defined by
n
d

n x2 2
Hn (x) = (−1) e e−x
dx
2
ex +∞
Z
= √ (−2iu)n exp{−u2 + 2ixu} du (6.63)
π −∞
6.4. DENSITY OPERATOR IN STATISTICAL MECHANICS 173

The density operator in the energy representation is trivial:

hm|ρ̂|ni = ρn δmn

1 1
 
ρn = exp − β h̄ω(n + )
Z 2

n = 0, 1, 2, . . . (6.64)

where −1
1
 
Z(T, V, 1) = 2 sinh βh̄ω (6.65)
2

Problem 9† : Verify Eq. (6.64)–(6.65).

In the position representation, we have

hq 0 |ρ̂|qi = hq 0 |n0 ihn0 |ρ̂|nihq|ni

X

nn0

Ψn0 (q 0 )ρnn0 Ψ∗n (q)

X
=
nn0

1 X 1
 
= exp − β h̄ω(n + ) Ψ∗n (q) Ψn (q 0 )
Z n 2

1/2
1 Mω 1
  
= exp − (x2 + x02 )
Z πh̄ 2
∞
1 1
 
− β h̄ω(n + ) Hn (x) Hn (x0 )
X
× n
exp (6.66)
n=0
2 n! 2

Problem 10† : Verify the steps leading to (6.66).

Hint: We have twice inserted the complete set of energy eigenfunctions.

Using the integral representation of the Hermite polynomials we get:

1/2
1 Mω 1
  
0
hq |ρ̂|qi = exp + (x2 + x02 )
Zπ πh̄ 2
∞
+∞ +∞ (−2uv)n 1
Z Z  
exp{−u2 + 2ixu}
X
× du dv exp − β h̄ω(n + )
−∞ −∞ n=0
n! 2

× exp{−v 2 + 2ix0 v} (6.67)

174 CHAPTER 6. DENSITY OPERATORS

The summation over n can be carried out as follows:

∞
(−2uv)n 1
X  
exp − β h̄ω(n + )
n=0
n! 2

∞ n
1 1
  X 
= exp − βh̄ω − 2uv exp(−βh̄ω)
2 n=0
n!

1
   
−βh̄ω
= exp − βh̄ω exp − 2uv e (6.68)
2
Then Eq. (6.67) becomes
1/2
1 Mω 1
  
0
hq |ρ̂|qi = exp + (x2 + x02 − βh̄ω)
Zπ πh̄ 2
Z +∞ Z +∞  
2 2 0 −βh̄ω
× du dv exp − u + 2ixu − v + 2ix v − 2uv e (6.69)
−∞ −∞

The argument in the exponent is a general quadratic form, which can be rewritten in the form
1
−u2 + 2ixu − v 2 + 2ix0 v − 2uv e−βh̄ω = − wT · A · w + ib · w (6.70)
2
where
e−βh̄ω
!
1
A = 2
e−βh̄ω 1
!
x
b = 2
x0
!
u
w = (6.71)
v

We now use the general formula

1 (2π)n/2 1
Z    
dn w exp − wT · A · w + ib · w = 1 exp − bT · A−1 · b (6.72)
2 [detA] 2 2

which holds if A is an invertible symmetric matrix.

Problem 11† : Verify (6.72).

Using (6.72) we get

1
e− 2 βh̄ω
1/2
1 Mω

0
hq |ρ̂|qi = 1
Z πh̄ [1 − e−2βh̄ω ] 2
6.4. DENSITY OPERATOR IN STATISTICAL MECHANICS 175

1 2
 
× exp (x + x02 ) − [1 − e−2βh̄ω ]−1 (x2 + x02 − 2xx0 e−βh̄ω )
2
1
1 Mω 1 xx0
  
2
= exp − (x2 + x02 ) coth(βh̄ω) + (6.73)
Z 2πh̄ sinh(βh̄ω) 2 sinh(βh̄ω)

Using the identity

1 cosh(βh̄ω) − 1 sinh(βh̄ω)
 
tanh βh̄ω = = (6.74)
2 sinh(βh̄ω) 1 + cosh(βh̄ω)
one finally gets
1
1 Mω

2
0
hq |ρ̂|qi =
Z 2πh̄ sinh(βh̄ω)

Mω 1 1
     
× exp − (q + q 0 )2 tanh βh̄ω + (q − q 0 )2 coth βh̄ω (6.75)
4h̄ 2 2

The diagonal elements of the density matrix in the position representation yield directly the average
density distribution of a quantum mechanical oscillator (at temperature T ):
 1
Mω 1 Mω 1
     
2
ρ(q) = tanh βh̄ω exp − tanh βh̄ω q 2 (6.76)
πh̄ 2 h̄ 2
which is a Gaussian distribution with width
 1
2
 h̄ 
σq =    (6.77)
1
 
2M ω tanh 2 βh̄ω

Problem 12† : Show that in the limit of high temperatures, βh̄ω  1,

1
mω 2 M ω2 q2
  
2
ρ(q) ≈ exp − (6.78)
2πkT 2kT
and at low temperature βh̄ω  1,
1
mω M ωq 2
  
2
ρ(q) ≈ exp − (6.79)
πh̄ h̄
Therefore the density matrix thus contains, for high temperatures, the classical limit, and for very
low temperatures, the quantum mechanical ground state density.