Martin Pieper

Quantum
Mechanics
Introduction to Mathematical
Formulation
essentials

Springer essentials
Springer essentials provide up-to-date knowledge in a concentrated form. They
aim to deliver the essence of what counts as "state-of-the-art" in the cur-
rent academic discussion or in practice. With their quick, uncomplicated and
comprehensible information, essentials provide:

• an introduction to a current issue within your field of expertis

• an introduction to a new topic of interest
• an insight, in order to be able to join in the discussion on a particular topic

Available in electronic and printed format, the books present expert knowledge
from Springer specialist authors in a compact form. They are particularly suitable
for use as eBooks on tablet PCs, eBook readers and smartphones. Springer essen-
tials form modules of knowledge from the areas economics, social sciences and
humanities, technology and natural sciences, as well as from medicine, psycho-
logy and health professions, written by renowned Springer-authors across many
disciplines.

More information about this subseries at http://www.springer.com/series/16761

Martin Pieper

Quantum Mechanics
Introduction to Mathematical
Formulation
Martin Pieper
FH Aachen, University of Applied Sciences
Jülich, Germany

ISSN 2197-6708 ISSN 2197-6716 (electronic)

essentials
ISSN 2731-3107 ISSN 2731-3115 (electronic)
Springer essentials
ISBN 978-3-658-32644-9 ISBN 978-3-658-32645-6 (eBook)
https://doi.org/10.1007/978-3-658-32645-6

© Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2021

The translation was done with the help of artificial intelligence (machine translation by the service
DeepL.com). A subsequent human revision was done primarily in terms of content.
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with
regard to jurisdictional claims in published maps and institutional affiliations.

Planung/Lektorat: Iris Ruhmann

This Springer imprint is published by the registered company Springer Fachmedien Wiesbaden
GmbH part of Springer Nature.
The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany
For Elli … the cat’s alive!
What You Can Find in This essential

• Postulates (principles) of quantum mechanics

• Quantum states in Hilbert space
• Operators as observables
• Eigenvalues as measured values
• Born’s rule
• Discrete energy levels with degeneration

vii
Preface

Nils Bohr, Nobel Prize winner and co-founder of quantum mechanics, quotes:
“If quantum mechanics hasn’t profoundly shocked you, you haven’t understood it
yet.” This is certainly irritating at first. What does this mean? The results of quan-
tum mechanics represent a radical break with classical physics. For example, it is
no longer possible to make statements about concrete particle trajectories. Instead,
only probabilities are given. Albert Einstein commented this rather skeptically
with “At any rate, I am convinced that He (God) does not throw dice.”
Nevertheless, modern physics, especially quantum mechanics, is currently
more popular than ever. This is especially true for a broad audience outside the
physics community. This is probably due to numerous popular science books on
the subject, but not least to the success of the television series “Big Bang Theory.”
Thus, many people know the fate of Schrödinger’s cat and one or the other has
surely wondered which hieroglyphics are on Sheldon’s board and what they mean.
These are exactly the questions that the present text wants to answer—at least
partially.
The essential therefore is aimed at an interested readership with a certain basic
education in mathematics, such as that imparted in natural science and engineering
courses. In detail, the education should include linear algebra (vector calculus) in
particular. Previous knowledge of physics, especially quantum mechanics, is not
necessary. Basic knowledge in the context of school physics is sufficient here,
which includes terms such as energy and momentum.
In this text, we can only give an introduction to the subject, focusing in particu-
lar on mathematical formalism. Therefore, while we will go beyond the content
of popular science books, we will of course not cover all the details comple-
tely. For this, we refer to the numerous textbooks. Thus, we try the balancing
act of using as few formulas and abstract terms as possible on the one hand,

ix
x Preface

while providing enough information to understand the mathematical formulation

of quantum mechanics on the other. To this end, we motivate mathematics by
means of well-known examples from linear algebra, that is, vector algebra.
Acknowledgment

The concept for the present text was developed in the context of different electi-
ves in the course of studies of physical engineering at the FH Aachen. I would
therefore like to thank all students who played “guinea pigs” and gave me valua-
ble feedback. For the critical review of the text and numerous hints, I would like
to thank especially my “unwavering first readers” Stephanie Kahmann, Elisabeth
Nierle, Darius Mottaghy, Philipp Weyer and Nadja Hansen, the latter especially
for the support with the graphics. Further thanks are due to Springer Verlag for
giving me the opportunity to write this essential, in particular Mrs Ruhmann and
Mrs Schulz, who accompanied the project.

In July 2019 Martin Pieper

xi
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Quantum Mechanical Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1 Atomic Models According to Bohr and Sommerfeld . . . . . . . . . . . . 3
2.2 Particle in a Box . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2.1 Quantum Mechanical States . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2.2 Discrete Energy Levels and Position Probability . . . . . . . . . 8
2.2.3 Correspondence Principle and Hamiltonian . . . . . . . . . . . . . 9
2.2.4 Preparation and Superposition of States . . . . . . . . . . . . . . . . 10
3 Mathematical formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.1 Heisenberg and Schrödinger Representation . . . . . . . . . . . . . . . . . . . 13
3.2 Postulates of Quantum Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.2.1 Postulate 1: States in the Hilbert Space . . . . . . . . . . . . . . . . . 14
3.2.2 Postulate 2: Measured Values, Operators
and Eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2.3 Postulate 3: Probabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.3 Solution Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4 Examples of Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.1 A Thought Experiment with Toy Blocks . . . . . . . . . . . . . . . . . . . . . . 25
4.2 Schrödinger’s Cat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.3 The Stern-Gerlach Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

xiii
Introduction
1

Quantum mechanics is a subarea of theoretical physics and belongs to modern

physics, which emerged in the twentieth century. The main task of theoretical
physics is to describe natural phenomena by mathematical models. This allows to
explain experiments and predict the results. It investigates which measured varia-
bles are relevant for the description, analyses the universal relationship between
them and relates them to each other by mathematical equations. It is always att-
empted to use as few equations as possible and to find a uniform formulation that
can describe as many phenomena as possible in general terms. This is where theo-
retical physics differs from engineering sciences, for example. In these, the aim
is usually always a direct application of technology, for which specially adapted
models and equations are applied.
Before the twentieth century, the movement of bodies under the influence of
forces, such as celestial mechanics or ball throwing, was described by classical
mechanics. The basis for this are, for example, Newton’s laws, which provide the
trajectories. However, numerous experiments at the end of the nineteenth century
showed that classical mechanics fails in the description of atomic systems (see
e.g. [1]). Many phenomena, such as the photoelectric effect or the spectral lines
of the hydrogen atom, can no longer be explained in classical terms. Therefore, a
new mechanics for atomic systems had to be developed. The result was quantum
mechanics.
The main task of quantum mechanics is therefore to predict the results
of physical measurements on atomic systems (elementary particles, atoms and
molecules). This includes, for example, the movement of atomic particles in elec-
tromagnetic fields, the scattering of atomic particles among themselves or the
structure of atoms and molecules.
In order to be able to describe all these different tasks with only one uni-
form theory, that is, few equations and assumptions, a radical break with classical

© Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2021 1

M. Pieper, Quantum Mechanics, Springer essentials,
https://doi.org/10.1007/978-3-658-32645-6_1
2 1 Introduction

mechanics was made. Numerous classical concepts such as trajectories or deter-

ministic measurement results had to be abandoned. This required a completely
new mathematics, namely functional analysis. It was developed at the beginning
of the twentieth century, almost simultaneously with quantum mechanics, mainly
by David Hilbert in Göttingen.
Exactly this uniform mathematical description is the subject of the essential.
We want to bring the theoretical and mathematical fundamentals of quantum
mechanics closer to a broad readership and thus create an interest in quantum
mechanics.
The essential begins in Chap. 2 with two examples, the atomic models accor-
ding to Nils Bohr and Arnold Sommerfeld and the description of a particle in a
box. The goal is to introduce typical quantum mechanical phenomena like dis-
crete, degenerate energy states and superpositions of states and to discuss them
briefly from a mathematical point of view. The knowledge gained from this will
be used in the following main chapter to motivate the uniform mathematical
formulation and make it plausible.
Chapter 3 forms the core of the essential. We give the basic mathematical
structure of quantum mechanics in the form of three postulates (principles of
theory). In particular, the quantum mechanical states in Hilbert spaces are the
fundament. Furthermore, we deal with Hermitian operators and their eigenvalues,
which stand for possible measured values. We also explain how the associated
probabilities are calculated.
In the last chapter, we will finally discuss three small examples. These are
intended to demonstrate how the rather abstract and theoretical postulates are app-
lied in practice. We start with a thought experiment, on the one hand, to deepen the
notation according to Paul Dirac with bra and ket vectors and on the other hand
to explain and treat degenerate states, as observed in Sect. 2.1 with the hydro-
gen atom. The second example takes up Schrödinger’s cat. We will discuss the
superposition of states and their interpretation. The Stern-Gerlach experiment con-
cludes. Based on experimental observations, we apply the postulates to describe
the system. As a result, we find that the previously unknown property “particle
spin” is the reason for the measurement results in the experiment.
Quantum Mechanical Phenomena
2

2.1 Atomic Models According to Bohr and Sommerfeld

Bohr’s atomic model (1913) assumes a heavy, positively charged nucleus, which
is orbited by negatively charged electrons. However, not all classically possible
orbits are allowed, but only certain orbits to which discrete energy values E n are
assigned. Here the energy level is characterized by the principal quantum number
n ∈ N. If an electron jumps or falls from one orbit to another, electromagnetic
radiation is emitted or absorbed with the frequency f = E/h, whereby E
is called the energy difference of the orbits and h ≈ 6,62 · 10−34 Js is Planck’s
constant (see Fig. 2.1).
We consider the hydrogen atom as an example. Here we can apply the atomic
model to explain, for example, the experimentally determined spectral lines. The
formula according to Nils Bohr corresponds exactly with the empirically found
formula of Johann Jakob Balmer and Johannes Rydberg. Further confirmation of
the model is provided by the Franck-Hertz experiment, in which discrete energy
levels in atoms are also observed.
However, Bohr’s simple model fails if a magnetic field is applied. In this case,
the spectral lines split (Zeeman effect). The reason is that the energy levels E n
are degenerate. This can be explained by the Bohr-Sommerfeld atomic model.
In 1916, Arnold Sommerfeld extended Bohr’s model by also allowing elliptical
orbits on which the electrons orbit the nucleus. A total of three parameters are
required to describe an ellipse, so the quantum numbern is no longer sufficient.
Sommerfeld also introduced the secondary quantum numbers  and m. For the
exact characterization of the quantum mechanical state, the principal quantum
number n and the two secondary quantum numbers  and m are necessary.

© Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2021 3

M. Pieper, Quantum Mechanics, Springer essentials,
https://doi.org/10.1007/978-3-658-32645-6_2
4 2 Quantum Mechanical Phenomena

photon
orbit with E E
1 3
orbit with E E=E -E
2 photon E 3 2
E 2
orbit with E 3
3 E = E
E photon
2
=h f

energy levels
E
1

Fig. 2.1 Emitted photon at the energy jump from E 3 to E 2

Without external fields, the discrete energy levels E n still depend only on the
principal quantum number n and not on  and m. However, the energy state E n
is called degenerate because several possible orbits (states) with energy level E n
exist (see Fig. 2.2). A measurement of the energy value E n therefore does not

Fig. 2.2 Three possible orbits for the energy state E 3

2.2 Particle in a Box 5

uniquely determine the state, as we do not know on which orbit the electron is
located.
If an external magnetic field is present, the discrete energy values also depend
on the magnetic quantum number m (hence the name). In this case, the energy
level E n splits into 2 + 1 equidistant Zeeman levels. Thus, part of the energy
degeneration is cancelled and the state is more precisely defined.
The idea of orbits on which the electrons orbit the atomic nucleus is now
obsolete. In modern quantum mechanics, the specification of a particle trajectory
no longer makes sense. Instead, statements are made about the position probability
of the particle. Nevertheless, the observations on energy degeneration are still
valid.

2.2 Particle in a Box

In this section, we want to understand the occurrence of discrete energy levels.

As an example, we consider a free particle trapped in a one-dimensional box with
infinitely high walls (see Fig. 2.3). Since no forces act on the particle in the box
of length L, the potential energy disappears.
According to the laws of classical mechanics, the particle oscillates in the box
between the walls. The motion is clearly defined by the state, that is, the indication
of the position x(t) and the momentum p(t) at time t. The particle can assume any
energy value E ≥ 0.

2.2.1 Quantum Mechanical States

In microscopic systems described by quantum mechanics, we can no longer

specify the classical state of the particle exactly. The reason is the Heisenberg
uncertainty principle (1927). It states that position and momentum cannot be mea-
sured sharply at the same time. More precisely, the following inequality holds for
the standard deviation of position σx and the standard deviation of momentum σ p


σx · σ p ≥ ,
2

where  ≈ 1,05·10−34 Js denotes the reduced Planck’s constant. So if we mea-

sure the position of the particle very precisely, σx is small. But this means that σ p
must be at least so large that the inequality is satisfied. In particular, σ p cannot
6 2 Quantum Mechanical Phenomena

Fig. 2.3 u 2n (x) for E 1 to E 3 ; position probability for x corresponds to the area

also become arbitrarily small at the same time. Similarly, a very precise measure-
ment of the momentum will result in an uncertainty in the position, corresponding
to the inequality.
In this context, it is important to note that this is a fundamental law of nature,
which does not result from any measuring errors or inaccurate measuring equip-
ment. For a better understanding, we give a semiclassical explanation for the
uncertainty principle, which goes back to Werner Heisenberg himself (see [2] for
further explanations).
We need a measuring device, for example, a microscope, to measure the posi-
tion of a particle. The accuracy is determined by the wavelength of the light used.
A smaller wavelength gives a more accurate result. During the measurement, a
photon hits our particle, is reflected or diffracted and registered in the microscope.
At the moment of the position measurement, when the photon hits the particle, the
momentum of our particle changes unsteadily. This change depends on the energy
of the photon. The smaller the wavelength of the light, that is, the more accurate
the spatial measurement, the greater the change. Therefore, at the moment we
know the position, the momentum of the particle becomes uncertain.
2.2 Particle in a Box 7

We have found that we cannot determine the classical state in atomic sys-
tems. Instead, we introduce a new quantum mechanical state. According to Erwin
Schrödinger, this state is described by a wave function ψ(t, x) that depends on
the position x and the time t. However, it has no direct physical meaning and
serves rather as an “auxiliary quantity” in the calculation. Physical statements are
only possible in connection with Born’s rule:
For the subinterval [a, b] ⊂ [0, L], the probability of the particle being present
at time t in the interval [a, b] can be determined by the integral using the square
of the absolute value |ψ(t, x)|2 of the wave function:

b
W (t, [a, b]) = |ψ(t, x)|2 d x. (2.1)
a

As already mentioned, we note that in quantum mechanics, we can only make

statements about possible measurement results and the probabilities with which
they occur. The question of concrete particle trajectories makes no sense.
In this section, we are not interested in the time-evolution of the state, so we
only consider the position-dependent part u(x) of the wave functions ψ(t, x) =
v(t) · u(x). This is determined by the time-independent Schrödinger equation

2 d 2
− u(x) = Eu(x) (2.2)
2m d x 2

Here m denotes the particle mass. On the left side is the second derivative of
u(x) with respect to x. In addition, E is a constant, which will turn out to be an
energy quantity. The particle can only stay in the box. So it is enough to look at
the equation in the interval [0, L]. Outside the box and at the walls, the position
probability is zero, that is, the wave function u(x) disappears here. This fact is
called boundary condition.
The solution of Eq. 2.2 is discussed in the next section. Before that we will
take a closer look at the constant E. For this purpose, we reformulate Eq. 2.2 and
compare the units:

We recognize that the constant E is an energy quantity, with the unit Joule.
8 2 Quantum Mechanical Phenomena

2.2.2 Discrete Energy Levels and Position Probability

Next, we are interested in the possible energy values for the particle. For this,
we have to solve Eq. 2.2, that is, we are looking for functions that are derived
twice to give a multiple of themselves. The trigonometric functions sin(λx) and
cos(λx), with a constant λ to be determined, are suitable. Because of the boundary
condition, u must disappear at the boundary. For the cosine, however, cos(0) = 1
always holds. As possible solution, only the sine remains.
We now determine λ in such a way that on the one hand the sine fits in our
box, that is, that we have zeros at the walls, and on the other hand that sin(λx)
satisfies the differential Eq. 2.2. We obtain two conditions:

2m E
λ · L = n · π and λ2 = .
2

The discrete energy values are obtained by combining the two expressions:

n2π 2 2m E 2 π 2 2
= 2 ⇒ E= n .
L2  2m L 2

The energy values E are usually marked with an index n: E n . Analogous to the
atomic model, n is called a quantum number and runs through the natural numbers
(without zero). The corresponding solutions also receive an index: u n (x).
According to Born’s rule, the probability of the particle’s position is given
by the square of the wave function u 2n (x). But first we have to normalize: The
integral of u 2n (x) over the entire box must equal 1 for the particle to be safely in
the box with probability of 1. We finally obtain:
  nπ   nπ 
2 2
u n (x) = sin x ⇒ u 2n (x) = sin2 x .
L L L L

Figure 2.3 shows u 2n (x) for the first three energy levels. The area under the
curve indicates the position probability of the particle (see Eq. 2.1). In contrast
to classical mechanics, there are preferred areas (e.g. peaks) where the particle is
located and areas where the position probability is low (e.g. zeros). In addition,
the particle cannot assume any arbitrary energy value E ≥ 0, but only the discrete
energy levels E n .
2.2 Particle in a Box 9

2.2.3 Correspondence Principle and Hamiltonian

We now want to understand Eq. 2.2, that is, especially the mathematical structure.
Because of the constant E, we first consider the total energy of the particle. In our
problem, the energy consists only of the kinetic energy, as the potential energy
disappears in the box according to preconditions:

1 2 p2
E tot = E kin = mv = , (2.3)
2 2m

We denote with v the particle velocity and with p = mv the corresponding

momentum.
In quantum mechanics, the observables of classical mechanics become opera-
tors through correspondence rules (see [3], Chap. 3). From a mathematical point
of view, operators are calculation rules that are applied to states. For identification
purposes, we use a hat above the corresponding letters.
As an example, we consider the position operator x̂ and the momentum
operator p̂:

d
x̂[u(x)] = x · u(x) and p̂[u(x)] = −i u(x).
dx

We see that the operators are applied to the wave function u(x). The position
operator is a multiplication operator that multiplies the wave function by the posi-
tion x. The momentum operator, on the other hand, is a differential operator, in
which the wave function is derived with respect to x.
The correspondence principle now means that we replace the observable posi-
tion and momentum in the classical expressions by their corresponding operators.
In the case of total energy (see Eq. 2.3), we obtain

p2 2 d 2
E tot = ⇒ Ĥ = − ,
2m 2m d x 2

by using the differential operator for the momentum p. The second derivative
occurs because p 2 means that the derivative is applied twice.
In classical mechanics, the Hamiltonian function is closely related to the total
energy. For this reason, we speak of the Hamiltonian, which is referred to as Ĥ .
A comparison with Eq. 2.2 gives:
10 2 Quantum Mechanical Phenomena

2 d 2
− u(x) = Eu(x) ⇔ Ĥ u(x) = Eu(x).
2m d x 2

The stationary Schrödinger equation can therefore be seen as an eigenvalue

problem for the Hamiltonian: We are looking for eigenfunctions u(x), which are
mapped to a multiple when using the Hamiltonian. The multiple is the energy
eigenvalue E. This eigenvalue problem is completely analogous to the search for
eigenvalues and eigenvectors of matrices in linear algebra (see Sect. 3.2.2).

2.2.4 Preparation and Superposition of States

From a mathematical point of view, any linear combination

∞

u(x) = cn u n (x) (2.4)
n=1

of the functions u n (x) satisfies the Schrödinger equation and the boundary
conditions and therefore is a possible state. Here cn ∈ C are series coefficients.
But how can we interpret such linear combinations physically?
u(x) is a superposition of different states u n (x), similar to how different waves
can be superimposed and a new resulting wave is produced. In this case, we
combine the eigenstates u n (x) of the Hamiltonian.
Such superpositions are needed, for example, to describe the experimental
initial conditions that are determined by the preparation of the particle source.
Mathematically, we again use a wave function u(x), which describes this initial
condition at t = 0 as a quantum mechanical state.
In linear algebra, we represent arbitrary vectors by linear combinations of basis
vectors. Analogously, we can expand certain functions in Fourier series, that is,
write them as linear combinations of sine and cosine functions. This also app-
lies to the initial state u(x), which we expand into a series with respect to the
eigenfunctions u n (x) of the Hamiltonian:

∞ ∞   nπ 
  2
u(x) = cn u n (x) = cn sin x .
L L
n=1 n=1

We obtain a linear combination, similar to Eq. 2.4. The coefficients cn are to

be determined in accordance with the function u, analogous to Fourier series.
2.2 Particle in a Box 11

We already understand the meaning of the wave functions u n (x). But how can
we interpret the series coefficients cn physically? To do this, we calculate the
expectation for the energy described by the Hamiltonian (see [4], Sect. 5.4):

L ∞

Ĥ = u(x)∗ Ĥ u(x)d x = |cn |2 E n . (2.5)
0 n=1

The brackets Ĥ indicate the expectation of the energy. The asterisk stands for
the complex conjugate, as wave functions can generally assume complex values.
For a complex number z = a + bi, the complex conjugate is defined by z ∗ =
a − bi.
We compare the expression in Eq. 2.5 with the standard formula from
probability theory for the expectation

E(X ) = pn x n .
n

Here pn is the probability for the random variable X to take the value x n .
By comparing coefficients, we find that the squares of the magnitudes |cn |2 of
the series coefficients cn can be interpreted as probabilities. They indicate the
probability that the particle is in the state u n (x) and takes on the energy value E n
during a measurement, after preparation by u(x).
Mathematical formulation
3

3.1 Heisenberg and Schrödinger Representation

In Sect. 2.2, we have considered a particle that is trapped in a box. The particle
was described by the quantum mechanical state using a wave function u(x). This
approach was developed by Erwin Schrödinger (1926). He conceives particles as
waves to which wave functions are assigned.
At the same time, matrix mechanics was developed by Werner Heisenberg,
Max Born and Pascual Jordan (1925–1926). They describe observables by matri-
ces with an infinite number of entries. The states are correspondingly vectors with
infinitely many components.
Which is the right representation? The answer is: “both,” because they provide
the same physical predictions, that is, they are equivalent. Schrödinger was the
first to show equivalence in 1926. Additional proofs followed. John von Neumann
in particular put quantum mechanics on a strict mathematical basis (see [5]):
The wave functions and the state vectors are only different forms of represen-
tations of a general mathematical structure, the Hilbert space, which is the basis
of quantum mechanics.
We want to understand this point better and consider a 2π periodic function
f (t), which we develop into a Fourier series:

f (t) = cn eint . (3.1)
n∈Z

It does not matter whether we work with the function f (t) or whether we use
cn the Fourier series coefficients. Both representations have the same information.
We have the choice between the function and an infinitely long vector formed
from the expansion coefficients:

© Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2021 13

M. Pieper, Quantum Mechanics, Springer essentials,
https://doi.org/10.1007/978-3-658-32645-6_3
14 3 Mathematical formulation

f (t) ⇔ (. . . , c−2 , c−1 , c0 , c1 , c2 , . . .). (3.2)

Analogously, in quantum mechanics, we have the freedom to choose a concrete

form of representation for the states and observables. What is decisive is the basic
structure of the description, which will turn out to be a separable Hilbert space
on which linear, Hermitian operators are defined.

3.2 Postulates of Quantum Mechanics

In the following sections, we summarize the general description of quantum

mechanics in three basic postulates. We derive these from physically and mathe-
matically reasonable demands and assumptions. As special cases, the postulates
include wave mechanics and matrix mechanics. In particular, we motivate the
abstract terms by well-known examples from linear algebra. We refrain from
formal mathematical definitions and proofs and instead refer to the numerous
literature (e.g., [3, 4] or [5]). In addition, we restrict ourselves to the case of
discrete eigenvalue spectra, which is sufficient for a basic understanding.

3.2.1 Postulate 1: States in the Hilbert Space

In Sect. 2.2, we used wave functions to describe the particle in a box. These
were interpreted as a quantum mechanical state. In physics, states generally form
the basis of any system description. Therefore, we first discuss the necessary
properties of general quantum mechanical states. For motivation, we use the wave
functions.
To develop his wave mechanics, Schrödinger started from de Broglie’s matter
waves. He tried to describe mathematically that particles behave like waves in
certain experiments (see [6], Chap. 2). One of the important wave phenomena is
the so-called superposition principle: different waves superimpose to form a new
resulting wave. In the same way, we can also superimpose wave functions and
obtain general states (see Sect. 2.2.4 and 4.2).
The superposition is mathematically a linear combination. In linear algebra,
we encounter linear combinations when, for example, we express arbitrary vectors
through the standard basis vectors:
3.2 Postulates of Quantum Mechanics 15
     
1 0 2
2· +6· = . (3.3)
0 1 6

Through regular use of vectors, we know that we can add two vectors or
multiply them by a number. As a result, we get a vector again. The same applies
to our quantum mechanical states. Here, a linear combination yields a state, too.
For vectors, we have certain rules of calculation, which also apply to states. For
example, the order of addition is arbitrary: a + b = b + a .
In mathematics, all objects that behave like vectors a , that is, obey the same
rules of calculation as vectors, are grouped together in a so-called vector space or
linear space. Therefore, our states are also abstract vectors in a vector space.
Unfortunately, this is not enough for us. We need a scalar product as an add-on
to calculate expectations. For two-dimensional vectors, the following holds:
 
b1
a · b = (a1 , a2 ) ·
T
= a 1 b1 + a 2 b2 , (3.4)
b2

where a T is the transpose of a , that is, a row vector, if we generally consider a

as column vector.
But where have we used a scalar product in wave mechanics so far? To answer
this question, we first note that the expression

u |v = u(x)∗ · v(x)d x (3.5)

defines a scalar product for the wave functions u(x) and v(x). We now show that
we can express the expectation from Eq. 2.5 using this scalar product. For this
purpose, we replace v by Ĥ u and obtain

u | Ĥ u = u(x)∗ · Ĥ u(x)d x =  Ĥ . (3.6)

The expressions in Eqs. 3.4 and 3.5 fulfill certain calculation rules, which
define a general scalar product in mathematics. However, not every vector space
has a scalar product. Therefore, we call vector spaces with scalar product pre-
Hilbert spaces.
Before we expand the pre-Hilbert space into a Hilbert space, we introduce the
bra-ket notation according to Paul Dirac: Generally speaking, states are described
by ket vectors |u . Here the notation in the brackets is arbitrary. Often the naming
16 3 Mathematical formulation

of states is chosen analogous to the measured values, as for example, |a with the
measured value a. Also common is the use of quantum numbers, which describes
the state |n , which, for example, belongs to the energy value E n of the n-th
Bohr’s orbit.
Dirac’s notation has the advantage that a bra vector a| can be introduced for
each ket vector |a . This is to be seen analogous to the transpose a T . Put together,
the bra and ket vectors form the usual bracket that characterizes the general scalar
product:

a T · b ↔ a|b,

where in the middle, there is only one stroke and no double stroke.
In linear algebra, two vectors are orthogonal if the scalar product disappears:
a T · b = 0. This can be transferred analogously to general states. If we have
a|b = 0, then the two states |a and |b are orthogonal. Unfortunately, this pro-
perty cannot be illustrated as easily for states as for vectors that are perpendicular
to each other. But in the following sections, we will see that it is nevertheless
very useful.
Let us return to the superposition of states and consider Eq. 2.4, now in bra-ket
notation:
∞

|u = cn |u n  . (3.7)
n=1

In contrast to the linear combination of vectors in Eq. 3.3, we sum infinitely

many state vectors. This is because the vector space of the wave functions is infi-
nite, but the vectors a are only two-dimensional. We illustrate this fact with the
function f (t) from paragraph 3.1: We need an infinite number of Fourier coeffi-
cients (see Eqs. 3.1 and 3.2). For the description of a only two vector components
are sufficient.
In principle, we can calculate in infinite-dimensional vector spaces in the same
way as in finite-dimensional ones; there are only a few differences that we have to
take into account. The branch of mathematics that deals with infinite-dimensional
vector spaces is functional analysis (see e.g. [7]). It thus forms the mathematical
basis of quantum mechanics.
We would be finished at this point if we would only consider states from
finite-dimensional vector spaces. The previous properties (vector space with sca-
lar product) are sufficient for description. In quantum mechanics, we are usually
3.2 Postulates of Quantum Mechanics 17

dealing with infinite-dimensional vector spaces, so we must make two additional

demands:
When evaluating the infinite sum in Eq. 3.7, we calculate a limit from the
mathematical point of view. However, it can happen that the limit no longer
belongs to our vector space. Then it is not a state and thus has no physical
meaning.
As an example, we consider the following sequence (see Heron’s method)

xn 1
x0 = 1, xn+1 = + , n = 0, 1, 2, . . . ,
2 xn

which is defined by rational numbers (fractions) and we calculate the first four
elements:

3 17 577
x0 = 1, x1 = = 1.5, x2 = = 1.4166 . . . , x3 = = 1.4142 . . .
2 12 408
√
We suspect the limit value 2 = 1.4142 . . ., which is not a fraction, that is,
does√ not belong to the rational numbers. We must add the irrational numbers (such
as 2) to the rational numbers so that our sequence has a limit. In mathematics,
the real numbers are called complete. Here the limits of all convergent sequences
are again real numbers. The rational numbers are not complete, as the example
shows.
Thus, we must demand that the vector space is complete. Only in this way can
we ensure that infinite sums, as in Eq. 3.7, have a limit in our vector space, that
is, provide a quantum mechanical state. A pre-Hilbert space that is complete is
called Hilbert space.
Let us come to the second additional demand: For motivation, we consider the
linear combination in Eq. 3.3. Here we have used the standard basis vectors e1 =
(1, 0) T and e2 = (0, 1) T . These form a possible basis of the two-dimensional
vector space. We can write any vector uniquely as a linear combination of this
basis vectors. We would like to have something similar for the general quantum
mechanical states.
To construct this basis, we go back to the procedure in Sect. 2.2, where we
solved the eigenvalue problem for the Hamiltonian and found the eigenstates |u n  .
Then, in Eq. 2.4, we formed the arbitrary states by linear combination of the |u n  .
We would now like the eigenstates |u n  to be a basis of our Hilbert space, that
is, we would like to be able to represent all possible states as linear combinations
of the |u n  . For this, we have to make sure that we only have countably infinite
18 3 Mathematical formulation

basis vectors, so that we can form the infinite sum. If we had an uncountable
number of vectors (a “greater” infinity than the countable one), this would not be
possible. A Hilbert space with a countable basis is called separable. Such Hilbert
spaces are in a certain way not “boundlessly large” and thus “manageable.”
We summarize the properties for general quantum mechanical states in the first
postulate:

Postulate 1 (Pure states)

For every quantum mechanical system, there exists a separate Hilbert space H,
so that every pure state of the system can be represented by a vector |a ∈ H of
the Hilbert space, respectively that every vector |a ∈ H from the Hilbert space
corresponds to a possible physical state.

3.2.2 Postulate 2: Measured Values, Operators and Eigenvalues

The examples in Chap. 2 show that quantum mechanics can only make state-
ments about possible measurement results and their probability of occurrence.
The second postulate therefore deals with observables and their measured values.
We generalize the observations from the box example: All quantum mechani-
cal observables are described as operators applied to Hilbert space vectors. The
eigenvalues correspond to the possible measured values and the eigenvectors are
the associated eigenstates. In the following, we will only specify which proper-
ties the operators must have in order for us to interpret them in a physically
meaningful way.
Due to the linear structure of quantum mechanics, only linear operators are
used. The motivation is as follows: We can write states as linear combinations
(see Eq. 3.7). An obvious requirement is then that an operator  should act
individually on the summands:

∞

Â|u = cn Â|u n  .
n=1

But this property characterizes linear operators. In the example above, Â is

applied to the states |u and all |u n  . For the two-dimensional vectors, the linear
operators correspond to the 2 × 2 matrices, which we will use more often for
clarification in the following.
3.2 Postulates of Quantum Mechanics 19

Before we demand another property from the operators, we turn to the mea-
sured values, that is, the eigenvalues of the operators. Eigenvalues are complex
numbers with a specific property: For each eigenvalue a, there exists an eigen-
vector |a = |0 that is not equal to zero. If we apply the operator  to this
eigenvector |a , we obtain a multiple of |a . The multiple corresponds to the
eigenvalue a. We can express this more briefly by mathematical formulas: Â|a =
a|a . For eigenvectors, the mapping is very simple; it consists only of a mul-
tiplication with the eigenvalue. It should be noted that eigenvalues are uniquely
determined, but for eigenvectors, we have certain freedoms, for example, with
respect to the “length.”
For clarification, we look at the matrix
 
12
A= .
21

The eigenvalues of A are a1 = −1 and a2 = 3 and possible eigenvectors are,

for example
       
12 −1 1 −1
· = = (−1) ·
21 1 −1 1
        (3.8)
12 1 3 1
· = = 3· ,
21 1 3 1

but every multiple of the vectors would also be an eigenvector.

Eigenvalues can also be complex numbers, as the following example shows.
The matrix
 
12
B= .
−2 1

has the complex eigenvalues b1,2 = 1 ± 2i.

In quantum mechanics, the eigenvalues correspond to the measured values. In
experiments, however, we do not measure complex numbers, so only real eigen-
values are allowed. We need an additional property of the operators to ensure this
requirement: All operators must be Hermitian operators.
We explain this property for 2 × 2 matrices. A comparison of the two matrices
A and B in the example above shows that A is symmetric: A coincides with the
20 3 Mathematical formulation

transpose A T (A = A T ). The transpose is formed by mirroring the entries on the

diagonal.
For complex-valued matrices A, we have to transpose and additionally form
the complex conjugate of the matrix entries. The resulting matrix A+ is called the
adjoint matrix. We call a matrix self-adjoint or Hermitian, if A = A+ . Similarly,
the definition is extended to general operators by calculating the matrix entries
using a scalar product.
Hermitian operators have a number of useful properties. For us, the following
two are particularly important: All eigenvalues and all expectations calculated
according to Eq. 3.6 are real.
We know almost all properties to describe observables and their measured
values. However, one property is still missing. This concerns the eigenvectors.
For this, we look again at our example for the matrix A. In Eq. 3.8, we have
found that A has the eigenvectors a1 = (−1, 1) T and a2 = (1, 1) T . The scalar
product
 
1
a1T · a2 = (−1, 1) · =0
1

of the eigenvectors is zero, so they are orthogonal. This is not a coincidence,

but applies generally to Hermitian operators. In addition, we can normalize all
eigenvectors, that is, bring them to length one. Such a system is then called
orthonormalized.
In addition, we would like the orthonormal system of eigenvectors to be com-
plete, that is, to form a basis of the corresponding Hilbert space. Then all states
can be written as linear combinations of the system. We demonstrate this fact with
the vector (2, 6) T , which we write as a linear combination of the eigenvectors of
the matrix A (see also Eq. 3.3):
     
−1 1 2
2· +4· = .
1 1 6

Unfortunately, this property is not fulfilled by all Hermitian operators, so we

have to demand it explicitly. Fortunately, it is possible to prove it for all operators
that describe physical observables.
We summarize our observations in the second postulate:
3.2 Postulates of Quantum Mechanics 21

Postulate 2 (Observables and measured values)

All physical observables, except time, are represented by linear, Hermitian opera-
tors. These have a complete system of orthonormalizable eigenstates. The possible
measured values are the eigenvalues of the operators.
Before we derive the third postulate in the next section, we want to look at dege-
nerate states, as we have observed in Bohr’s atomic model. In quantum mechanics,
degeneration is always referred to when there are several different linearly indepen-
dent eigenstates for a measured value (eigenvalue). For clarification, we consider
an observable, which is represented by the matrix
 
30
C=
03

and has the eigenvalue c = 3 with multiplicity two. Two possible eigenvectors are,
the standard basis vectors e1 and e2 . So we find two different linear independent
eigenstates, which belong to the same measured value c = 3. Therefore, the states
are degenerate. By measuring c = 3, we cannot determine whether the system is in
the state e1 or e2 .
Mathematically the degeneration is related to the eigenvalue. We have found an
eigenvalue with multiplicity two. In general, eigenvalues with multiplicity greater
than one are associated with degenerate states.
We need a second observation variable to decide in which state the system is
exactly. If, for example, we have previously only measured the energy, we now also
determine the angular momentum. Here it is important that both observables are
“compatible,” that is, can be measured sharply at the same time. The position and
the momentum would be an example of where this is not the case.
If two quantities can be measured sharply at the same time, it does not matter
which quantity we measure first. The measurements do not influence each other. This
property can be transferred to the corresponding operators: The operators commute
regarding multiplication, which is generally not the case. We consider an example
and introduce an additional observable, which is represented by the matrix
 
13
D= .
31

The eigenvalues of D are d1 = −2 and d2 = 4.

For matrices C and D, C · D = D · C holds, that is, they commute with respect
to matrix multiplication. So we can use D to cancel the degeneration, that is, to
22 3 Mathematical formulation

determine the exact state. For this purpose, we are looking for a common sys-
tem of eigenvectors to the two matrices, that is, two vectors d1 and d2 , which are
eigenvectors to both matrices C and D.
The vectors d1 = (−1, 1) T and d2 = (1, 1) T meet this requirement. So by
specifying the measurement of D, we can now determine exactly what state our
system is in. For example, if we measure c = 3 and d = −2, we know that the
corresponding state vector is d1 . If, on the other hand, we measure c = 3 and d = 4,
we know that the state d2 is present.
We transfer this procedure to general operators. We consider two Hermitian
operators Ĉ and D̂, which commutate, that is, for which the so-called commutator
 
Ĉ , D̂ = Ĉ D̂ − D̂ Ĉ = 0̂

disappears. In this case, we can find a complete set of common eigenstates. The
states are usually designated by their associated eigenvalues: If |c is the eigenstate
belonging to the eigenvalue c of operator Ĉ and |d corresponding to the eigenvalue
d of D̂, then the common eigenstate is denoted by |c, d . The following holds:

Ĉ|c, d = c|c, d and D̂|c, d = d|c, d .

In Sect. 3.3, we will use this fact and give an algorithm how we can accurately cha-
racterize a quantum mechanical system. Another example to illustrate degeneracy
is discussed in Sect. 4.1.

3.2.3 Postulate 3: Probabilities

The third postulate concerns the probabilities with which we obtain the possible
measured values. We already know that in the linear combination

∞

|u = cn |u n 
n=1

the squares of the coefficients |cn |2 indicate the probability of measuring the asso-
ciated eigenstate |u n  (see Sect. 2.2.4). We now consider how to determine these
coefficients for any given state |u .
We consider the following linear combination
3.2 Postulates of Quantum Mechanics 23
     
1 0 0, 6
a = 0, 6 · + 0, 8 · = .
0 1 0, 8

We assume that the standard basis vectors e1 and e2 are eigenstates of an
observable, which is described by a 2 × 2 matrix. Then we obtain the evolu-
tion coefficients in the linear combination by forming the scalar product with the
eigenstates. For example, due to the orthonormality of the vectors we obtain by
   
1 0
e1T · a = 0, 6 · (1, 0) · +0, 8 · (1, 0) · = 0, 6
0 1
 
=1 =0

the coefficient belonging to e1 . The probability of measuring the state e1
2
is: e1T · a .
In quantum mechanics, we can always write general states as linear com-
binations of orthonormalized eigenstates. Therefore, the following holds in
general

∞

u k |u = cn u k |u n  = ck ,
n=1

where we use that u k |u n  gives the value one if n = k and otherwise disap-
pears. The probability is then calculated from the square: |u k |u|2 . We have thus
derived the statement of the third postulate:

Postulate 3 (Probabilities)
The probability w with which the measured value a is encountered during a
measurement at the observable A in the state |u is calculated by

w(a , |u ) = |a|u|2 ,

where |a is the eigenstate associated with eigenvalue a designated by the

corresponding operator Â.
24 3 Mathematical formulation

3.3 Solution Algorithm

We close with a general solution algorithm how to solve quantum mechanical

problems. We consider a quantum mechanical system and assume that we can
describe it by an observable A.
In the first step, we have to write this observable as Hermitian operator Â.
Usually, the classical expression describing the observable is used for this. In it,
all classical quantities are replaced by the known quantum mechanical operators
(correspondence rules).
In the second step, we calculate all eigenvalues a and the corresponding eigen-
states |a . If these are not degenerate, we are finished and can describe our system
uniquely by these eigenstates.
If  is degenerate, that is, has multiple eigenvalues, we must add a second
observable B. This is described analogously as Hermitian operator B̂. We first
check whether the two operators commutate, that is, whether they can be mea-
sured sharply at the same time. For this purpose, the commutator [ Â, B̂] is
calculated. If the commutator disappears, we calculate the common eigenstates
of  and B̂. Again we check whether there are still degenerate eigenvalues. If
the degeneration is completely removed and no more degenerate states exist, we
are done. In this case, Â and B̂ are sufficient for the complete description of the
system. If there are still degenerate states, we have to add another observable C
and proceed analogously.
We iterate this process until there is no more degeneration. This way, we can be
sure that the operators Â, B̂, Ĉ, etc. describe the system completely. In this case,
the measurement of the observables uniquely defines the state. The operators then
form a complete set of commutating operators. The goal of any description of a
quantum system is therefore to find such a complete set.
Examples of Use
4

4.1 A Thought Experiment with Toy Blocks

We begin with a simple thought experiment, in which we look at a toy for child-
ren (Fig. 4.1) and describe it mathematically, within the framework of quantum
mechanics.
The toy consists of toy blocks that differ in shape (round, square and triangular
cross section) and color (white and gray). The child tries to put the toy blocks in
a box, but of course the opening has to be found to match the shape of the block.
How does this toy fit in with quantum mechanics? We interpret the toy blocks
as quantum mechanical particles. The properties shape and color then correspond
to two observables. The box can finally be seen as a measuring device for the
observable shape.
We now want to construct a general basis for a suitable Hilbert space. The
associated operators  and B̂ commute, because the child can determine the
properties shape and color independently. So they form a complete system of
commutating operators, because there are only two properties. We can therefore
find a common basis of eigenvectors to the eigenvalues (measured variables) shape
and color. As usual, we denote these with their corresponding measured value in
bra-ket notation:

Â|r  = r |r  , Â|s = s|s and Â|t = t |t ,

where r stands for round, s for square and t for triangular. |r  is thus, for
example, the eigenstate from  to eigenvalue r, which represents the possible
measurement result round.

© Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2021 25

M. Pieper, Quantum Mechanics, Springer essentials,
https://doi.org/10.1007/978-3-658-32645-6_4
26 4 Examples of Use

Fig. 4.1 Toy for children with toy blocks

Analogously, we define the eigenstates of the operator B̂ that belongs to the

observable color:

B̂|w = w|w and B̂|g = g|g .

Here w stands for the measurement result white and g for gray.
The common eigenstates of  and B̂ are then designated as follows:

|r , w , |r , g , |s, w , |s, g , |t, w and |t, g ,

where the first entry indicates the shape and the second the color.
What happens now when the child plays with the toy blocks? Quantum mecha-
nically it determines the observable form. If a block fits through the round
opening, it is clear that the particle is in a state |r  . However, as long as the
child has not yet determined the color, the building block can be either white or
gray. So it is open whether the particle is in the state |r , w or |r , g . Therefore, a
degenerate state is present. Only when the second observable color is determined,
the degeneration is canceled.
4.3 The Stern-Gerlach Experiment 27

4.2 Schrödinger’s Cat

Schrödinger’s cat is a thought experiment proposed by Erwin Schrödinger in

1935: A cat is locked up in a box. Further there is a radioactive preparation,
a Geiger counter and a poison vial in the box. The preparation decays with a cer-
tain probability and is detected by the Geiger counter. The Geiger counter triggers
a device that breaks the poison vial and the cat dies.
There are two states for the radioactive preparation: decayed |d or not deca-
yed |nd . These superpose to a general state |u = c1 |d + c2 |nd , where |c1 |2
and |c2 |2 indicate the probabilities for the individual states.
Analogously, Schrödinger transfers the situation to the cat: We have the two
states of being alive |a or dead |na . The general state is provided by the linear
combination |v = c1 |na + c2 |a , with the known probabilities for the decay of
the preparation, which correspond to |d and |nd , accordingly.
This superposition is present as long as we do not look into the box and “mea-
sure” whether the cat is alive or dead. By the measurement, the wave function
“collapses” according to the Copenhagen interpretation and we get one of the sin-
gle states. Especially here, the imagination plays a trick on us: the superposition
of both states means that the cat is both alive and dead, which is not possible.
The problem with this example is that we transfer quantum mechanical con-
cepts to the macroscopic world. There are different approaches to solve this
paradox, for example, the many-worlds interpretation or Bohm’s mechanics. A
further explanation can be found in [4] Sect. 10.5.
Paul Dirac warns in this context (see [8], Chap. 4) that the superposition of
quantum mechanical states is something completely new to which no classical
picture should be applied. There is an essential difference to all existing classical
images. Ultimately, it is merely a mathematical theory, that is, a tool that is used
to describe quantum mechanical processes.

4.3 The Stern-Gerlach Experiment

In the Stern-Gerlach experiment, uncharged particles (e.g., silver atoms) are obser-
ved in a vacuum. The particles leave a source and are then focused into a beam,
which is detected on a screen. The result is a spot in the middle. The spot splits
when the beam passes through a magnetic field. If the field runs in the z-direction,
we observe two separate spots at equal distance from the center of the screen,
symmetrically shifted up and down in the z-direction (see Fig. 4.2).
28 4 Examples of Use

Fig. 4.2 Experimental set-up for the Stern-Gerlach experiment

This observation cannot be explained by classical physics, because here the

particles would be distributed uniformly in z-direction between the two observed
spots (see Fig. 4.2). It is therefore a quantum mechanical effect that we want to
describe and investigate with our formalism.
What do we need for the description? First we have to specify the Hilbert
space H to which the states belong. Then we look at the energy and introduce an
operator that describes the corresponding observable.
We observe two beams: deflected upwards (+) or deflected downwards (−).
So we have two states in which a particle can be on the screen. We denote these
with |+ and |− . We obtain any state as a superposition of the two states, that
is, mathematically as a linear combination

|u = u + |+ + u − |−

with appropriate coefficients u + , u − ∈ C whose squares indicate the associated

probabilities.
Two numbers are therefore sufficient to describe the states. We can therefore
write each state as a vector with two components. For this purpose, we identify
the two basic states |± with the standard basis vectors. We obtain for |u the
following:
         
1 0 u+ 1 0
|+ =
ˆ , |− =
ˆ ⇒ |u  =
ˆ = u+ + u− .
0 1 u− 0 1
4.3 The Stern-Gerlach Experiment 29

A simple representation for our Hilbert space H is therefore H = C2 , with

appropriate normalization so that the overall probability is one. This observation
is important, because we now know that the operators on H, which describe the
observables, are 2 × 2 matrices with complex entries.
We will now look at the observable energy: The particle with the magnetic
moment µ  ∈ R3 classically has the interaction energy H = −µ  · B with the

magnetic field B ∈ R (see [9], Chap. 3). The vectors are three-dimensional, that
3

 = (µ1 , µ2 , µ3 )T and B = (B1 , B2 , B3 )T . The magnetic field is fixed in the

is, µ
experiment. We can therefore only quantize the moment µ,  that is, identify it as
an operator within the framework of the correspondence principle. For this, we
calculate the scalar product of the two vectors:

 · B = −(µ1 · B1 + µ2 · B2 + µ3 · B3 ).
H = −µ

The associated Hamiltonian is applied to the state vectors |± ∈ C2 :

 
Ĥ |± = − µ̂1 · B1 + µ̂2 · B2 + µ̂3 · B3 |±
 
= − µ̂1 |± · B1 + µ̂2 |± · B2 + µ̂3 |± · B3 .

The three components µ̂i of the magnetic moment—understood as operators—

are thus 2 × 2 matrices with complex entries. In order to be able to determine
these matrices in more detail, properties resulting from physical observations and
requirements are used. For a detailed calculation, we refer to the literature, for
example, [8], Chap. 37, and only consider energy measurement as an example
here.
It follows from the experiment that we can only measure two discrete values
for H = −µ  ∓µ0 B, where we refer B = | B|
 · B:  to the strength of the magnetic
field and µ0 is derived from the positions of the two observed spots. We have “∓”
as sign, because in the formula for H, there is a minus sign. The measured energy
values correspond to the eigenvalues of the associated operator. So we have

Ĥ |± = ∓µ0 B|±. (4.1)

In our experiment, the magnetic field runs in z-direction, that is, the first and
second components of the magnetic field vector disappear: B1 = B2 = 0. There-
fore, the magnetic field is B3 = B and Ĥ is reduced to Ĥ = −µ̂3 B. We apply
this in Eq. 4.1 and cancel −B:
30 4 Examples of Use

(4.2)

With the matrix

 
1 0
µ̂3 := µ0
0 −1

the condition from Eq. 4.2 is satisfied for the standard basis vectors. By using
further properties, for example, that the operators must be hermitic (µ̂i+ = µ̂i ),
the remaining matrices can also be determined:
   
01 0 −i
µ̂1 := µ0 and µ̂2 := µ0 .
10 i 0

These matrices (without the factor µ0 ) are the Pauli matrices, a first indication
that the spin of the particles is responsible for the observed splitting of the beams.
The commutators
     
µ̂1 , µ̂2 = 2iµ0 µ̂3 , µ̂1 , µ̂3 = −2iµ0 µ̂2 and µ̂2 , µ̂3 = 2iµ0 µ̂1 ,

show that the matrices µ̂i do not commute. The result is typical for angular
momentum quantities. A comparison with Poisson bracket from classical mecha-
nics shows that the result fits to an angular momentum quantity (see [10], Chap. 2
and [11], Chap. 5). This observation suggests that the unknown property (spin)
is an “angular momentum-like” quantity that gives rise to the corresponding
magnetic moment. For more details on spin, we refer to the standard textbooks
again.
What You Learned From This essential

In this introduction to the mathematical formulation of quantum mechanics, you

have…

• used examples to analyze discrete and degenerate energy levels

• established Hermitian operators according to the correspondence rules for the
description of observables
• investigated the significance of quantum mechanical state vectors in Hilbert
space
• interpreted eigenvalues of Hermitian operators as measured values
• calculated probabilities for measured values
• applied three postulates for the mathematical description of quantum mecha-
nical systems to examples

© Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2021 31

M. Pieper, Quantum Mechanics, Springer essentials,
https://doi.org/10.1007/978-3-658-32645-6
Literature

1. Huebener, R.P.; Schopohl, N.: Die Geburt der Quantenphysik. Springer Spektrum:
Wiesbaden, 2016.
2. Orzel, C.: How to Teach Quantum Physics to Your Dog. Scribner: New York, 2010.
3. Nolting, W.: Theoretical Physics 6: Quantum Mechanics—Basics. Springer International
Publishing: Cham, 2017.
4. Rebhan, E.: Theoretische Physik: Quantenmechanik. Spektrum Akademischer Verlag:
Heidelberg, 2008.
5. von Neumann, J.: Mathematical Foundations of Quantum Mechanics. Sixth printing.
Princeton Univers. Press: New Jersey, 1971.
6. Gasiorowicz, S.: Quantum Physics. Third edition. John Wiley & Sons: New York, 2003.
7. Großmann, S.: Funktionalanalysis im Hinblick auf Anwendungen in der Physik. Fifth
Edition. Springer Spektrum: Wiesbaden, 2014.
8. Dirac, P.A.M.: Principles of Quantum Mechanics. Fourth Edition. Oxford University
Press: Oxford, 1958.
9. Nolting, W.: Theoretical Physics 3: Electrodynamics. Springer International Publishing:
Cham, 2016.
10. Nolting, W.: Theoretical Physics 2: Analytical Mechanics. Springer International
Publishing: Cham, 2016.
11. Nolting, W.: Theoretical Physics 7: Quantum Mechanics – Methods and Applications.
Springer International Publishing: Cham 2017.

© Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2021 33

M. Pieper, Quantum Mechanics, Springer essentials,
https://doi.org/10.1007/978-3-658-32645-6