Quantum Mechanical Laws

Bogdan Mielnik
Oscar Rosas-Ortiz
Physics Department
Cinvestav

EOLSS-UNESCO
http://www.eolss.net
Bogdan Mielnik
Oscar Rosas-Ortiz
Departamento de Física
Cinvestav
Apartado Postal 14-740
México 07000 D.F.
Mexico

E-mail: bogdan@fis.cinvestav.mx
orosas@fis.cinvestav.mx

Distributed free for academic use only (15 April 2010), sale is prohibited.

Copyright (2006) by B. Mielnik and O. Rosas-Ortiz except:

FIGURE 15. Copyright (1996) by the American Physical Society. Reprinted figure with permission from Brune
M., et. al., Phys. Rev. Lett. Vol 76, 1800 (1996). Readers may view, browse, and/or download material for
temporary copying purposes only, provided these uses are for noncommercial personal purposes. Except as
provided by law, this material may not be further reproduced, distributed, transmitted, modified, adapted,
performed, displayed, published, or sold in whole or in part, without prior written permission from the
American Physical Society.

Copyright (1999) by Laurie Grace. The art work is a courtesy of the author and the Scientific
FIGURE 24.

American Journal.

The present is a preprint version of Quantum Mechanical Laws (2007/Rev.2009), a chapter in Fundamentals
of Physics, [Ed. José Luis Morán López], Encyclopedia of Life Support Systems (EOLSS), Developed under
the Auspices of the UNESCO, Eolss Publishers, Oxford, UK, [http://www.eolss.net] ©2007 Eolss Publishers
Co Ltd, Ballamoar Beg, Clengh Road, Sandygate, Ramsey, IM73AE Isle of Man (IOM).

The full chapter (pp 255-326) is also available in the eBook:

Fundamentals of Physics - Volume 1
José Luis Morán-López and Peter Otto Hess (Eds)
Published: 2009
ISBN: 978-1-84826-952-1
 FOREWORD

Shortly after receiving the invitation from EOLSS to write an article on Quantum Mechanical
Laws, we realized that the presenting of the subject in a popular form, without sacrificing the
contents, is a formidable task that we could attend only in an imperfect way. We therefore decided
to center our report on certain aspects which perhaps should not disappear from the science of the
XXI century. An important fact is that Quantum Mechanics is not a closed chapter of the nowadays
knowledge. Its progress in the XX century left a sequence of unfinished discussions, doubts and
paradoxes, which at any moment can lead to some new discoveries. This seems especially
important in the present day science, marked by a high appreciation (if not worship) of the fast
technological results. Indeed, an excessive urgency to obtain the technological "fast food", in some
occasions, can block the progress which it was supposed to stimulate. An example from the past
might illustrate the problem. The photo-electric effect, discovered near the end of the XIX century,
lead to a number of astonishing new steps of the theory and its applications. One of them is the
technique of automatic closing and opening the doors. However, if the scientists of the past were
forced to think too much on how to open and close the doors, perhaps their understanding of the
photoelectric effect would be frustrated, with obvious handicaps for the further progress. So, while
the interests in technologies might stimulate the research some equilibrium must be conserved. The
crowds of today students of course, cannot dedicate their efforts to the fundamental problems. They
must work in technical, mathematical and applied areas. Yet it would be good if the echoes of the
fundamental polemics and unsolved problems were still alive, since now and then they can awake
some sparks of curiosity with unpredictable consequences. Having this in mind, we dedicated our
imperfect report not only to what we already know but also to what we don't...

An interrelated problem might seem secondary. An essential purpose of all interpretational works
and encyclopedia articles is to offer a good bibliography. The trouble is too often avoided by many
authors of interesting popular books, who only occasionally throw some vague reference, e.g.: "Dr.
Smith from Harvard tells that..." (letting the reader guess that the author is very much aware of
Harvard discussions!). However, an inexperienced reader who would like to know, where the
statement of Dr. Smith was published, has very little chances to check (even if he looks at Google,
Wikipedia, Science Citation Index, etc.) We henceforth decided to dedicate some effort to give
some data and comments about certain essential publications from the past. What we could collect,
once again, is desperately incomplete, but if it will give the younger readers just some keys to
begin the search, then the mission of our incomplete bibliography will be fulfilled.

We are grateful to all colleagues for their interest, critical remarks and difficult questions which,
perhaps, someday will find some answers.

Bogdan Mielnik
Oscar Rosas-Ortiz
 QUANTUM MECHANICAL LAWS

Bogdan Mielnik, Departamento de Física, Centro de Investigación y de Estudios

Avanzados, A.P. 14740, México D.F. 07000, Mexico

Oscar Rosas-Ortiz, Departamento de Física, Centro de Investigación y de Estudios

Avanzados, A.P. 14740, México D.F. 07000, Mexico

Summary

The present day quantum laws unify simple empirical facts and fundamental principles
describing the behaviour of micro-systems. A parallel but different component is the
symbolic language adequate to express the ‘logic’ of quantum phenomena. This report is
the sum of both elements. The Sections 1-7 present the decline of the classical theory and
the evidence of the particle-wave duality on an elementary level. In the Sections 8-9 the
origin of the Schrödinger’s wave mechanics and Born’s statistical interpretation are
presented. The following Sections 10-23 are basically addressed to the adepts of exact
sciences. They outline the formal and fundamental problems of Quantum Mechanics,
interpretational paradoxes, concepts of complementarity, teleportation and the idea of
quantum computing. Though using the symbolic language, Quantum Mechanics is not a
hermetic science. It is an open domain, crucial for the present day particle physics,
quantum field theory and contemporary informatics, though carrying a luggage of
unsolved interpretational problems.

Keywords
Indeterminism, Quantum Observables, Probabilistic Interpretation, Schrödinger’s cat, Quantum Control,
Quantum Computing, Entangled States, Einstein-Podolski-Rosen Paradox, Cannonical Quantization, Bell
Inequalities, Path integral, Quantum Teleportation.

CONTENTS:

1. Introduction
2. Black body radiation: the lateral problem becomes fundamental.
3. The discovery of photons
4. Compton’s effect: collisions confirm the existence of photons
5. Atoms: the contradictions of the planetary model
6. The mystery of the allowed energy levels
7. Luis de Broglie: particles or waves?
8. Schrödinger’s wave mechanics: wave vibrations explain the energy levels
9. The statistical interpretation
9.1 Polarized photons: absence of deterministic rules?
9.2 Born’s statistical interpretation.

-i-
10. The Schrödinger’s picture of quantum theory
10.1 States, superpositions, probabilities
10.2 Quantum geometry
10.3 Linear operators, their origin in quantum theory
10.4 Quantum observables, old and new aspects
10.5 Observables as operators: the anatomy of the measurement
10.6 General formalism; the extended Born’s hypothesis
10.7 Functions of an observable
10.8 Transition probabilities: the most elementary statistical law
10.9 Construction of traditional observables
11. The uncertainty principle: instrumental and mathematical aspects.
12. Typical states and spectra
12.1 Harmonic oscillator: the spectral ladder, eigenstates and coherent states
12.2 The angular momentum
12.3 Hydrogen and hydrogen-like atoms
12.4 Pauli’s exclusion principle
12.5 Particle spin: a controversial, non-classical property
12.6 The spectral bands
13. Unitary evolution
14. Canonical quantization: scientific or magic algorithm?
15. The mixed states
16. Quantum control: how to manipulate the particle?
17. Measurement theory and its conceptual consequences
17.1 State reduction (collapse of the wave packet)
17.2 The strong uncertainty principle
17.3 The idea of complementarity
17.4 “Quantum logic”
17.5 Quantum bit (qubit)
18. Interpretational polemics and paradoxes
18.1 Quest for alternative theories
18.2 The measurement problem
18.3 Borderline doctrine
18.4 The paradox of the Schrödinger’s cat
18.5 The doctrine of decoherence
18.6 Ensemble interpretation
18.7 The realistic view
18.8 Many worlds interpretation of Everett
18.9 Quantum jumps and quantum Zeno effect
18.10 The macroscopic superposition
19. Entangled states
19.1 General concepts
19.2 Spin-configuration entanglement
19.3 Many particle entangled states

- ii -
20. Dirac’s theory of the electron as the square root of the Klein-Gordon law
20.1 The Klein-Gordon equation
20.2 Dirac’s equation
20.3 The hypotesis of positron
21. Feynman: the interference of virtual histories
22. Locality problems
22.1 The EPR paradox
22.2 Hidden variables and Bell’s inequalities
22.3 “Seeing in the dark”
22.4 Teleportation
23. The idea of quantum computing and future perspectives
23.1 Quantum cryptography
23.2 Quantum computing
24. Open questions

Glossary

Anomalous Zeeman effect: The doubling of the atomic spectral lines in an external
magnetic field which seemed paradoxical before the magnetic moment of the
electron was discovered.

e
Bohr magneton: The quantity γ = 2 mc = 0.927×10-20 erg/oersted, where e and m are the
electron’s charge and rest mass, c is the light velocity,  = h 2π and h =
6.626×10−34 joules·sec, represents the (anomalous) magnetic moment of the
electron observed in the Zeeman effect, explained by the Dirac’s theory of the
electron.

Boson: Named after Satyendra Nath Bose, bosons are elementary particles with integer
spin and symmetric wavefunctions obeying the Bose-Einstein statistics (any
number of bosons can share the same quantum state). The quanta of light
(photons) are the simplest and most abundant example of bosons.

Complementarity: The fact that certain properties (aspects) of quantum systems cannot be
observed at the same time and make no sense simultaneously.

Copenhagen School: A pragmatic interpretational current following the ideas of N. Bohr,

W. Heisenberg, A. Rosenfeld, and others, centered around the concepts of
uncertainty and complementarity in quantum theories.

Determinism: An idea or demand that the initial conditions should determine uniquely the
final result of a certain process or experiment.

- iii -
Entangled state: The state describing correlations between several independent properties
or components (e.g., different particles) of a quantum system. Represented as a
tensor product of the partial states, acquired a crucial importance in quantum
information, teleportation and computing.

Fermion: Named after Enrico Fermi, fermions are elementary particles with half-integer
spin and antisymmetric wavefunctions obeying the Fermi-Dirac statistics (no two
fermions can occupy the same quantum mechanical state at the same time). The
most elementary examples are the electron, proton and neutron; today the known
fermions form an ample family of stable and unstable particles as well as atomic
nuclei.

Hidden parameters: The hypothetical classical variables permitting to recover the

deterministic picture of quantum experiments.

Indeterministic effect (phenomenon) : An effect which cannot be precisely predicted on

the basis of fundamentally available initial data.

Magnetic anomaly: The fact that the magnetic moment of the electron (Bohr magneton)
is twice as great as could be explained by the electron spin. It indicates that
neither the spin nor the electron’s magnetic moment admit the simple mechanical
model of an internal electron rotation.

Measurement paradox: The state of a quantum system evolves according to a linear,

deterministic law; yet, the results of the measurements, in general, are
unpredictable (indeterministic).

Neutron interferometry: The application of the neutron interference to determine the

structure of the matter. The principal advantage of the method is the electric
neutrality of the neutrons.

Quantum logic: An abstract approach to quantum theory looking for an analogy between
the simplest quantum mechanical experiments (filters) and propositions of a
certain logical system.

Quantum observable: A non-classical concept of an observable property represented by

an operator, i.e., an operation preformed on quantum states.

Quantum state: According to the old intuitions, a memeber of the discontinuous

(quantized) set of states allowed for a micro-system. In a general theory, a
collection of non-contradictory data (information) permitting to predict in
probabilistic terms the behaviour of a quantum system. If the information is
maximal, the state is called pure, represented by a wavefunction (wave packet),
more generally, by a vector in a linear space (state vector).

- iv -
Quantum superposition: An “undecided” state of a micro-system involving several other
states as tentative options (components) and represented by their linear
combination.

Reduction (collapse) of the wave packet: A hypothetical, unpredictable jump of the

micro-system state to one of the eigenstates of the measured quantity.

Schrödinger’s cat: An emblematic creature in a superposed state of being simultaneously

alive and dead.

Squeezing: An optical or dynamical operation on quantum states causing the contraction

(shrinking) of one canonical observable at the cost of expanding its canonical
counterpart (e.g. by improving the space localization of the wave packet at the
cost of increasing its momentum uncertainty).

Uncertainty principle: The non-commutativity of the operators reflects the limits of the
simultaneous measurability of the corresponding observables. The phenomenon
typical for the complementary quantities.

Wave-particle duality: The micro-particles propagate as waves but are detected as

corpuscles.

-v-
- vi -
1. INTRODUCTION

Quantum Mechanics (QM) was one of the greatest revolutions in physics. Although it did not abolish
(but rather extended) the former classical laws, the generalization was achieved at the cost of adopting
a completely new language of concepts and a new way of thinking at phenomenological and
mathematical levels.

The classical theories were deterministic. For the classical mechanics (Newton, Hamilton) the physical
systems were collections of material points and fields of force moving (evolving) in a completely
predictable way. If one only knew the present state of the system and the external forces, one could
predict with arbitrary precision its future evolution, or alternatively, extrapolate the past. The entire
universe seemed like an enormous, infinitely precise watch, in which the present day state is defined,
up to the smallest detail, by the yesterday state, and this one in turn, at least in principle, by the state of
million years ago (Laplace). On the astronomical scale, these laws permitted to integrate the planetary
and stellar orbits (Newton). In the laboratory experience they granted that the initial conditions of any
experiment determine uniquely its final result. The simplest illustration might be the elastic scattering
of classical balls running between some fixed obstacles (Figure 1) to hit the screen behind. If one
repeats the experiment sending many balls with different initial positions x and momenta p, the places
in which they touch the screen will differ. However, if we start to eliminate the initial differences, the
discrepancy of the final results will tend to disappear. In limit, if we could repeat the experiment with
the initial conditions exactly identical then, according to the classical mechanics, all the balls would hit
the screen in exactly the same point.

FIGURE 1. The deterministic character of the classical collision experiments.

The theory of micro-phenomena which emerged in the first decades of XX century had to abandon this
picture, together with the realistic image of particles, as “material points”, balls, or pebbles. The
objects which replaced them were the tiny corpuscle of electricity called electron (of charge e =
1.0602176462(63) 10 19 Coul and mass me = 9.10938188(72) 10 31 Kg), the corpuscles of the light
(photons) and other particles. The existence of electrons was predicted quite early in the inspired 1744
letter of Benjamin Franklin (c.f. Millikan 1948); proved rigorously by the 1874 experiments of
Thomson with the “cathode rays” (Watson 1947), almost simultaneously in the works of Zeeman (see
Romer 1948). The photons, in a sense, were foreseen by Newton, dismissed by Huygens and Maxwell,
and finally discovered by Einstein (1905). Soon, other micro-objects appeared, all of them showing the
need of a basically new theory replacing the deterministic laws by probabilistic ones.

-1-
2. BLACK BODY RADIATION: THE LATERAL PROBLEM BECOMES FUNDAMENTAL

While developing the deterministic scheme, the classical theories did not neglect the statistical laws.
The scientists of XIX c. knew and were applying the probabilistic descriptions in which not all physical
parameters could be controlled by the experimentalist. The fundamental step was the application of
Boltzman statistics to explain the principles of Thermodynamics. One of the unsolved problems
concerned an idealized physical medium (a perfectly black body) in thermal equilibrium with the
surrounding radiation bath. According to Boltzman, the radiation of such a body should have obeyed
the Raleigh law (see Figure 2), but the experiments were showing a different pattern. Near the end of
XIXc. this discrepancy could seem a lateral trouble, but its fundamental character was soon recognized.
In his polemic with Boltzman, Max Planck assumed that the surface of the black body can be
represented by an ensemble of harmonic oscillators. He then noticed that the false conclusions
disappear if one assumes that the energies exchanged by the radiation field with this surface cannot be
arbitrarily small. By comparing his hypothesis with the known empirical data he found an unexpected
consistency requirement: that the energy portions exchanged with the radiation of frequency should
be h , where h is a natural constant h = 6.626 10-34 Joules sec. Planck attributed the phenomenon to
the structure of the black body material (c.f. Born 1969 Ch VII; Bohm 1951 Ch 1, see also Cervantes-
Cota et al 2005), but soon it turned out that this hypothesis contained only a part of the truth.

FIGURE 2. The Rayleigh-Jeans (green), Wien (red) and Planck (blue) distributions.

3. THE DISCOVERY OF PHOTONS

Since the late XIX c. it was known that the surfaces of certain metals (e.g., silver) when illuminated by
the light emit electrons. Although the existence of this phenomenon did not contradict the classical
electrodynamics, the details did. The energies of the pulled out electrons did not depend on the light
beam intensity. When the intensity decreased, the number of emitted electrons decreased as well, but
their energies did not change; they depended only on the color of the light. For the monochromatic
light beam of frequency , the energies absorbed by the liberated electrons could be easily
reconstructed (c.f. Einstein 1905; Einstein and Infeld 1966; Jammer 1966,1974; Primas 1983;
Cervantes-Cota et al 2005) and turned out proportional to the frequency:

E=h , (3.1)

where h is the same Planck constant which defined the black body radiation. The experiment developed
as if the electrons in the metal were unable to absorb the beam energy smaller than h , depending only

-2-
on the light color but not on the properties of the metallic surface. It looked as if the light beam itself
was split into indivisible portions of energy h . Such was the hypothesis of Albert Einstein in his 1905
paper (Nobel prize 1921).

The idea contradicted classical electrodynamics and henceforth, was not accepted without resistance.
Perhaps, argued the opponents, the light is absorbed continuously, but the detectable effect (the
electron emission) occur only after enough energy is accumulated. Then, however, it would be difficult
to explain why the electrons obtain the energy h , above the minimum necessary to leave the metal.
Despite this, the experimentalists tried to check, whether the light absorption in the photoeffect had no
characteristics of a continuous accumulation. In crucial experiments of Joffe (1913) and Meyer and
Gerlach (1914), the light beam was falling not on a wide metallic surface but on a metallic powder
dispersed in vacuum. According to the classical electrodynamics, the beam energy falling onto each
grain of the powder per time unit could be determined by the Pointing vector integrated over the
exposed grain surface. The beam intensity was so low and the grains were so small that if the beam
were a continuous medium, then the critical energies E = h needed few seconds to accumulate on each
grain; henceforth the photo-effect should start with a visible delay. However, the powder response,
even if weak, was immediate; an effect which could be hardly deduced from the classical theory of the
continuous electromagnetic field. The fundamental discussions are still not over (cf, e.g., Marshall and
Santos 1989), thought the single photons can be already created and dropped one by one onto the
screen (Kim et. al. 1999), and can be “seen without being destroyed” (Maitre et. al. 1997, Nogues et. al.
1999).

4. COMPTON’S EFFECT: COLLISIONS CONFIRM THE EXISTENCE OF PHOTONS

The corpuscular effects of the light were independently detected for the light beams falling onto a gas
of free electrons. According to the classical theory, the plane electromagnetic wave of frequency
falling onto an electron must produce (in the electron rest frame) the spherically diffracted wave of the
same frequency. The scattered wave should just show the Doppler effect in the initial, laboratory
frame. This did not agree with the experiment. In 1923, Compton decided to calculate the same effect
by assuming that the electron of an initial momentum pe collides elastically with a light particle, i.e.,
with a photon of energy E = h = c|p| and momentum p = E n = h n = k, where |n|=1, =c/ is the
c
wavelength, k 2 n/ is the wave-vector and |k| is the wave-number (see Figure 3). A simple
argument based on the energy and momentum conservation proved that the scattered photons cannot
conserve their initial frequency . The dispersed light must have the angle dependent frequency shift in
the electron rest frame. If the scattering angle is , the wavelength of the dispersed light is

- = e(1 - cos ) (4.1)

where e = h
me c 2.43 10-12 m is the Compton wavelength (Born 1969 Ch V-4; Bohm 1951 Ch 2.8;
Messiah 1999 Ch I-5). The prediction fit the experiment. The photoeffect and the effect of Compton in
the first part of XX c. were two basic phenomena unexplainable by the Maxwell’s theory of light. It is
worth reminding that Compton’s argument was sharply criticized by other authors, including C.V.
Raman: “Compton, you are a good debater, but the truth is not in you” (Compton 1961).

-3-
FIGURE 3. Compton effect: the photon scattered in the direction has the new frequency < consistent with (4.1).

5. ATOMS: THE CONTRADICTIONS OF THE PLANETARY MODEL

The disquieting questions arose also about the nature of atoms. Though their existence was already
known (see e.g. Newburgh, Peidle and Rueckner 2006), the exact composition of an atom was ignored.
Since 1911, Rutherford used electrons to bombard the atoms in order to get a more precise information.
The results exhibited a highly inhomogeneous structure: the heavy, positively charged nuclei were
surrounded by the light, negative electrons, so that the total electric charge was vanishing. The image
of a planetary system in miniature, with the electrons substituting the planets and the electrostatic
attraction substituting the gravity (the “planetary model”) was so suggestive, that it persists in
physicists minds until today. However, the model proved wrong.

If the atoms were like the planetary systems the question would arise, why atoms of the same element
(e.g. hydrogen) must have the same size (Feynman, Leighton and Sands 1963). The radius of the
hydrogen atom is r 5.3 10-11 m; so 1 mol of the atomic hydrogen, according to this model, should be
like a galaxy of 6.02252 1023 simple planetary systems composed of one sun and one planet, where
each of 6.02252 1023 planets circulates at the same distance from its sun. The almost surrealistic
vision contradicted the common sense and the probability calculus (the classical theory cannot restrict
the size of the planetary systems!).

Even more serious was the stability problem. The electrostatic Coulomb field is not the only force
acting on the electrons in the atom. Should the classical model be true, the electrons circulating on their
“planetary orbits” should produce the electromagnetic waves (radiation) at the cost of the orbital
energy. Taking away the energy, the emitted radiation should act back on the electrons slowing down
their orbital motion. The effect of this self interaction (the radiative damping) is expressed by the third
time derivative of the trajectory d3x/dt3 and must transform the elliptic orbit into the spirals converging
to the attraction center. The phenomenon exists also in gravitation, for the true planetary motion, but is
incomparably weaker. The clean result was obtained by Dirac (1938), though the problem was known
much before. If resembling the classical point charges, the electrons would spiral down to the nuclei
and the atoms should disappear in a fraction of second. Some authors (Rubinowicz 1968) tell about the
“suicide of a classical theory” (see Figure 4).

-4-
FIGURE 4. The suicide of the classical theory: due to the radiative dumping the atoms should collapse within a fraction of
second.

6. THE MYSTERY OF THE ALLOWED ENERGY LEVELS

The optical phenomena added an extra challenge. The atoms exposed to the oscillating fields or
thermal collisions radiate, but their emission has little to do with the short time continuous burst
predicted by the spiral collapse (Figure 4). The spectroscopy registered the characteristic sequences of
spectral lines emitted without any catastrophe and allowing to identify the atoms of each element.
Thus, the sodium (Na) could be easily recognized by the dominating yellow light of frequency = 503-
520 1012 Hertz, rubidium (Rb) by its red line = 304-482 1012 Hertz. The observed emission-
absorption lines of the atoms typically formed double sequences. Thus, e.g., for the “hydrogen-like
atoms”, the distinguished frequencies were:

1
me2 1 1
nm = cRZ 1 , n, k = 1,2,… (6.1)
M n2 k2

where c = 2.9972 108 m/sec is the speed of light, me = 9.109 10 28 g is the electron’s mass, M and Ze
2
respectively are the mass (M me) and charge of the nucleus, R = m4ec = 4 a0 1.097 107 m 1 is the
-34
Rydberg constant, 2 =1.055 10
h
Joules sec, a0 = 0.529 10-10 m is the Bohr radius and
1/137 was latter called the fine structure constant. The fact was extremely useful in spectroscopy, but
incomprehensible for the planetary model. Moreover, due to the relation between the light color
(frequency) and photon energy, the spectral formula (6.1) suggested that the energies of the emitted
photons correspond to differences between some distinguished energy levels:

2
1 Zc 13.6
En h n = Z2 electron Volts, n = 1,2,… (6.2)
2 n n2

where meM/(me + M) is the reduced mass. Such was the famous observation of Niels Bohr in his
1913 paper (see also Pais 1991). Consistently, Bohr and Sommerfeld proposed a modified planetary
model (Bohr 1913, 1914; Sommerfeld 1916), still for a classical charge circulating in the Coulomb
field, but assuming that only a sequence of particular energies (6.2) was allowed for the electron orbits.
The atomic emissions would correspond to discontinuous jumps of the electron between the permitted
levels (6.2); the energy conservation law explaining the emission spectra. Strangely enough, Bohr
followed the original Planck’s arguments, while opposing strongly the existence of photons (see Pais

-5-
1991, Ch 11(d)), using (6.1-2) to explain the frequency of emitted radiation, but trying to save the idea
of the continuous energy emission. The so formulated theory was later called “old quantum mechanics”
(Rubinowicz 1968). What it could not explain, however, was the very existence of the “allowed” and
“forbidden” orbits.

7. LUIS DE BROGLIE: PARTICLES OR WAVES?

Though this be madness, yet there’s method in’t.
W. Shakespeare, HAMLET (3.2)

While the new facts were arising, the old ones caused new problems. The existence of the interference
pictures was no mystery for the Maxwell’s theory of the light waves. However, if the light was a beam
of particles, then it was hard to understand how the particle trajectories could create the wave pictures.
The simplest case was the traditional double slit experiment for a monochromatic light beam (Figure
5). If the light is granular and if each photon crosses just one of the slits, how can it know that the other
slit exists at all? And it must, since in absence of the second slit, the interference picture does not
appear. Later on, Richard Feynman (see Feynman, Leighton and Sands 1963, p. 37) would qualify the
double slit phenomenon as the “only mystery” of quantum theory (though, perhaps, he unjustly
dismissed the other puzzles; c.f. Section 17).

FIGURE 5. Double slit experiment: The greatest of all mysteries (Feynman).

A paradoxical choice of Luis de Broglie was to face the crisis by making it still deeper. If photons
show a peculiar behavior, then “there is a method in it”. In his 1923-24 papers, de Broglie assumed that
the wave-like patterns cannot be limited to photons, but it must affect all physical particles. He thus
proposed to associate with an electron moving with a momentum p = (p1, p2, p3) and kinetic energy E, a
fictitious (heuristic) image of a wave of frequency = E/h:

p(x,t)= Cei(px – Et)/ , C = constant (7.1)

with the adimensional exponent (px –Et)/ , where px = p1x1 + p2x2 + p3x3, 2 . The idea turned
h

prophetic. The strange vision of de Broglie was confirmed in a sequence of experiments by Davisson
and Germer (1927), Thomson and Reid (1927), and Thomson (1927). The beam of electrons from a
stationary source, with the same carefully defined momenta p, penetrated through a crystalline lattice
(most typically, a thin sheet of metal) falling onto a photographic plate. The electron hits were visible
as the clear points on the emulsion. Their localization was erratic but their accumulations formed a
picture of regular concentric circles, typical for the interference effects (see Figure 6).

-6-
FIGURE 6. The scheme of Davisson, Germer, and Thomson (DGT) experiment. The beam of electrons of the same
momentum p crosses the crystalline lattice and falls onto the screen behind. Differently than for the classical scattering of
Figure A, the points marked by the single electrons are unpredictable, but their accumulation forms a regular pattern
imitating the interference picture of the plane wave of length = h/p (p = |p|).

The diameters and the intensity of the circles (i.e., the density of hits) corresponded to the diffraction of
a plane wave of the length inversely proportional to the momentum, = h/|p|, where h is the same
Planck constant which already appeared in the phenomenon of the black body radiation and the
photoeffect energy (3.1). Moreover, for p = c/ and E = cp (valid for photons), the de Broglie formula
(7.1) reduced neatly to the Einstein’s E = h , indicating that it was a part of a universal law.

The attempts of explaining the phenomenon by the “cooperation” between the beam electrons failed as
the result of the Biberman, Shushkin, and Fabricant (1949) experiments in which the beam intensities
were too low to grant the existence of any ordered electron groups, yet the accumulation of the
erratically dropping electrons created exactly the same diffraction pictures which did not depend on the
beam intensity. The capacity of forming the interference patterns must have been encoded in each
single electron (the conclusion much firmer now due to the techniques which permit the controlled
emission of single electrons and photons, Kim et al 1999).

Step by step it was also verified that the corpuscular-wave duality indeed affects a wider class of
micro-objects. For neutrons it led to the useful techniques of neutron interferometry (Greenberger
1983). The tendency of forming the interference patterns is already checked for heavy micro-objects
(Arndt et al 1999, Hackermüller et al 2003). No known modifications of the classical motion equations
for the beam particles could explain the phenomenon (see also Sec. 18.1). It was the idea of Erwin
Schrödinger to look in a different direction.

8. SCHRÖDINGER’S WAVE MECHANICS: WAVE VIBRATIONS EXPLAIN THE ENERGY

LEVELS

The well known wave phenomena (the strings in guitars, air in flutes, trumpets, etc) are characterized
by certain distinguished frequencies (notes) proper to each stationary oscillation pattern. An inspiration
of Schrödinger was that the electron in an atom might not resemble a classical material point “on the
orbit”, but rather a piece of palpitating continuous medium obeying a certain wave mechanics (the
hydrogen singularity could be a kind of a trumpet generating only specific frequencies). Yet, there
must be a correspondence to the classical theory! Following this idea, Schrödinger turned attention to
the families (“congruencies”) of many simultaneous trajectories described by the generating functions

-7-
S(x,t) of the Hamilton-Jacobi (HJ) equation in classical mechanics. Could it be that these trajectories
are equivalents of the rays in the geometric optics while the surfaces S(x,t) = const are the wave fronts
for a certain new wave phenomenon? In his series of quantization papers (reprinted in Schrödinger
1978), he explored new aspects of the HJ-equation 1 ( S )2 V ( x ) S 0 for the non-relativistic
2m t
p 2

Hamiltonian H V ( x ) . By taking S(x,t) = S(x) –Et and putting S(x) = ln (x), the equation was
2m
)2 (V ( x ) E ) 2
2
reduced to K ( ) ( 0. However, instead of assuming that K( ) vanishes
2m
at every point x, Schrödinger considered the integral I K ( )d 3 x and demanded its stability
R3
under small variations of ( I = 0). By assuming that vanish outside of finite space vicinities
and by applying the standard variational method, he arrived at a new second order differential equation
for :

2
E (x) = (x) + V(x) (x) (8.1)
2m
2 2 2
where = x2
+ y2
+ z2
is the Laplace operator and E still corresponds to the system energy. In his
next works, Schrödinger considered the non-stationary . Looking for consistency with the dynamics
of de Broglie waves (7.1), he replaced the energy E in (8.1) by i , i 1 , obtaining the general
t
propagation equation for the hypothetical wave (x,t):

2
i (x, t) = (x, t) + V(x, t) (x, t), (8.2)
t 2m

evolving in presence of the external potential V(x,t). The time independent equation (8.1) was then
recovered by looking for the stationary (standing wave) solutions (x,t) = e i t (x), = E , with a
time-independent oscillation mode (x). As found later, the similar heuristics could also lead to non-
linear wave equations (see e.g. Rylov 1999; Mosna, Hamilton, and Delle 2005), but the simple guess of
Schrödinger turned specially fortunate. His principal results concerned the static potentials V(x,t)
V(x). By choosing V(x) = e2/r, where r = |x| (the Coulomb potential acting on the hydrogen electron)
he could show that the stationary oscillations (8.1), trapped by the attraction center (with (x) 0 as
|x| + ), admit only a discontinuous sequence of proper energies (6.2) allowed by the self-
interference of in the potential well (Figure 7). Though the physical nature of the complex wave (x)
was still unknown, the central law of quantum mechanics was already discovered!

FIGURE 7. The mechanism of the allowed levels: only for the exceptional energies the wave (8.1) circulating in the
Coulomb well survives the destructive interference.

-8-
The physical sense of (x,t) was a subject of polemic discussions (until today not entirely concluded).
Schrödinger himself was on favor of the straightforward interpretation: that the electron is indeed a
piece of continuous medium represented by a complex function (x,t), which oscillates around the
atomic nucleus, producing the Bohr-Sommerfeld sequence of distinguished frequencies. Yet, this literal
interpretation could not be defended. As objected in the Saclay Conferences, a single wave function
propagating according to (8.1) can reach two distant detecting screens but each electron can be
absorbed by only one of them. Moreover the characteristic property distinguishing the wave
propagation (8.1) from the motion of a classical particle was the tunnel effect: the function of de
Broglie when facing an arbitrarily high potential barrier V(x) 0 is never completely reflected but it is
split into two parts, one reflected one transmitted. Yet, each single electron is never divided: In can be
detected either on one or on the other side of the barrier (Figure 8)

FIGURE 8. An electron and a potential barrier. The Schrödinger’s wave function splits into two parts, but the electron can
be detected just in one of them.

9. THE STATISTICAL INTERPRETATION

9.1. Polarized photons: absence of deterministic rules?

The crucial difficulty is not exclusive for the electrons but concerns particles of all types. For photons,
it is best illustrated by the Dirac’s example of the monochromatic light beams crossing a polarization
filter (e.g., the Nicol prism, Figure 9). Suppose, the incident beam a, is polarized in the direction a. It
then crosses the prism B which selects the light component b polarized in the direction b.

FIGURE 9. Each photon polarized in the direction a is either completely absorbed or completely transmitted by the filter B
of the polarization b. The only prediction concerning the single photon is probabilistic.

If one does not know that the beam is composed of photons, the phenomenon seems easy to explain by
the classical Maxwell-Faraday equations. The beam a is just split by the prism B into two parts: one of
them polarized in the direction b is transmitted, while the other one, orthogonal to b, is absorbed. If the
angle between a and b is (see Figure 9), then cos2 and sin2 respectively are the transmitted and
absorbed fractions of the initial beam intensity. Moreover, the transmitted beam b has now the new

-9-
polarization b determined by the filter B. However, what happens with each single photon? One might
be tempted to assume that each photon repeats exactly the behavior of the polarized plane wave, so that
the part sin2 of it is absorbed, and the rest crosses B. Yet, this cannot be true. Should each photon be
split, the photons emerging from B would be of lower energy, and due to (3.1), the outgoing beam b
would have a different color than a (shifted to red and infrared as increases!) The experiments show
that this is not the case: the color of the transmitted beam is exactly the same as that of the incident one,
though the intensity decreases. This means that (unlikely the entire plane wave) each photon is either
completely absorbed or completely transmitted through the prism B. One might ask what is the
particular reason why some photons are absorbed and some other are not? Are there some “hidden
parameters” (different from the polarization) which could make the difference? However, if such
parameters existed, it would be principally possible to produce a beam polarized in the direction a but
completely transmitted through the filter B. This never happens. In all known experiments the
transmission coefficient is always cos2 and cannot be even slightly affected.

The similar regularity exists in the DGT experiments (Figure 6). Here, from the very beginning it was
known that the electron is not identical with the wave (x,t). The fictitious plane wave, diffracting
around the centers of the crystalline lattice in agreement with (8.1), produces the continuous
interference circles. However, each single electron hits just one point.

9.2. Born’s statistical interpretation

Against Schrödinger, the hypothesis was thus raised that the “wave functions” have probabilistic and
not material sense. Similarly as for the photons, the experiments with the electron interference are not
deterministic. For an electron with a precisely defined momentum in DGT experiments, there is no
information permitting to predict the exact point of the screen which will be hit. If such an information
existed, it should be possible to prepare the beam of electrons in the experiment of Figure 6, all with
the same momentum p, but hitting the screen at a single point or in a reduced screen area instead of
forming an interference pattern. However, this never happened. The element of indeterminism seems
the only way to save the consistency between the wave-like propagation of the particles and their
corpuscular effects (i.e., to explain the wave particle duality, see e.g. Mehra 1974; Rae 1986; Redhead
1987; Gribbin 1996).

Accepting the facts, the quantum mechanics had to desist from the classical concept of state and from a
naïve image of particle as a “material point”, classical ball or a lump of continuous matter. The physics
simply does not known how the electron looks. The particle state in QM (also: the pure state) is the
maximal set of non-contradictory information which one can have about the micro-object (Dirac 1935,
Cohen-Tannoudji, Diu and Lalöe 1977, Ballentine 1998). This information is statistical. What the
Schrödinger equation (8.1) describes is indeed the propagation of probabilities. The wave function
(x,t) is the state of a micro-object, i.e., an evolving “packet of information”. However, how to read
this information? In his 1926 paper sent to Zeitschrift für Physik, Max Born presented his historical
hypothesis: (x, t) defines the probability that the particle detected at the moment t will be found in
vicinity of the space point x. When submitting his manuscript, Born considered (x,t) d3x as the
required probability, where d3x is the volume element. Yet his first conjecture was wrong. The
experience with linearly propagating waves shows that the intensity of the interference patterns on the
screens is quadratic in the wave amplitude. The same was implicit in the propagation equation (8.1)
where the integral of | (x,t)|2 is an absolutely conserved quantity, suggesting the probability
conservation. In the last moment (in his page proof!), Born postulated the probability proportional to
(x,t)2 (he was surely considering real ). The paper (correction included) won the Nobel prize in 1954

- 10 -
(Jammer 1966, footnote 9 on page 302). The idea was subsequently generalized to | (x,t)|2 d3x (See
Born 1926b and Pauli 1927) and turned decisive for the mature formulation of QM.

10. THE SCHRÖDINGER’S PICTURE OF QUANTUM THEORY

10.1. States, state superposition, probabilities

The linear propagation law (8.1) and the quadratic probabilities determine the mathematical form of
QM consistent with the Schrödinger’s idea. The set of the physical pure states (i.e., the maximal
information states or the phase space of the theory) corresponds to the collection of all nontrivial
complex wave functions = { (x)} representing the diffraction/interference patterns of quantum
ensembles. In agreement with Born’s interpretation, | (x)|2 stands for the ensemble density while j =
( * *) is the particle current (Born 1926,1969). Since the states might interfere with
2im
arbitrary phases and intensities the following superposition principle was adopted: whenever two
complex functions 1, 2 represent two possible states of a particle, their arbitrary nontrivial
combination = 1 1 + 2 2 with complex coefficients 1, 2, represents a physical (pure) state as
well (Dirac 1935). The set of all ‘wave functions’ corresponding to the physical states is therefore a
complex linear space. Among them, of special interest are the normalizable ones, for which the integral
3
d 3 x | (x)|2 (denoted || ||2 and called the norm squared of ) is finite. The normalizable wave-
R
functions too form a linear space and admit a simple statistical interpretation. While | (x)|2 is
interpretable as an average ensemble density (observed on the screens), the corresponding probability
density for each single particle must be normalized (the total probability is 1). So, the probability
density is (x) = | (x)|2 || ||2, with the probability integral 3 d 3 x (x) = 1, and consistently, the
R

probability current is j = 1 ( * *). The evolving probability density (x,t) and its
2im || || 2
current are identical for any two proportional functions and (0 ), which therefore
represent the same physical state.

10.2. Quantum geometry

It is essential that apart of the norm (probability) conservation, the Schrödinger equation (8.1) grants
also the conservation of a more general bilinear form of any pair = { (x,t)}, = { (x,t)} of the
(simultaneously) evolving states:

| = *(x,t) (x,t) d3x = constant, (10.2.1)

called the inner or scalar product of and , satisfying: (i) Linearity | 1 1+ 1 2 = 1 | 1 +

2 | 2 , 1, 2 . (ii) Hermiticity | = | * (iii) Positivity | 0, | =0 = 0, for
any normalizable , , 1, 2. The time independence of (10.2.1) is an absolute conservation law,
independent of the particular potential in (8.1); it expresses the invariance of | under the arbitrary
evolution transformations. Following Felix Klein (the invariants define the geometry) the inner
products (10.2.1), and the complex linear combinations (superposition) become the fundamental
geometry elements in the set of quantum states. The linear space with the inner product (obeying i-iii),
can be completed to a Hilbert space H (Riesz and Nagy 1990); henceforth, normalizable wave
functions are vectors in a complex Hilbert space, with the vector norm || || defined by || ||2 : | ,
and the vector orthogonality: | = 0. In case of the wave functions in 3 (configuration

- 11 -
states) the dimension of H is infinite (dim H = + ), while the polarization, spin states form the finite
dimensional Hilbert spaces (dim H ) (c.f. Sec.12.4). The general Hilbert space geometry is an
abstract ground on which grows the symbolic language of quantum theories.

10.3. Linear operators: the spirit of interference

The linearity of the evolution law (8.1) and the interference phenomena of quantum ensembles imply
the special role of the linear operators in QM. The linear operator A in H is a function (mapping)
assigning to any vector H a new vector = A so that A( 1 1 + 2 2) 1A 1 + 2A 2. For
† †
every continuous linear operator A there exists exactly one operator A such that A | |A . The
correspondence A A† between operators is the hermitean conjugation generalizing the complex
conjugation z z* for the numbers z . The operator A is Hermitian (real, self-adjoint) if A† = A.
The statistical interpretation of QM privileges the quadratic functions (quadratic forms) of . Each
continuous quadratic form f( ) can be expressed as f( ) = |A , where A is a linear operator: f is real
if A is Hermitian. Furthermore, the stationary state equation (8.2) illustrates the importance of the
eigenvalues and eigenstates of the linear operators. The number is an eigenvalue of A if there
exists a non-trivial function (eigenfunction) u 0, normalizable (or limited by more general boundary
conditions), such that Au = u. An eigenvalue which admits more than one linearly independent u is
called degenerate (for the refinement of these definitions see e.g. Riesz and Nagy 1990). If A is self-
adjoint, is real. If u has the finite norm, is called a discrete eigenvalue of A. The formalism of QM
assigns also an important role to the non-normalizable eigenfunctions of the operators. Most typically,
the square integrable (normalizable) functions represent the bound states, while the non-normalizable
ones are the scattering states generalizing the de Broglie waves. For the states represented by the
normalizable wave functions with || ||2 < + , the theory is most conveniently formulated by
choosing || || = 1. The above concepts form a natural language to describe the measurements and
observable quantities (shortly the observables) in quantum theories.

10.4. Quantum observable: a non-orthodox concept

The concept of an observable is not exclusive for the Schrödinger’s picture of QM; indeed, it first
appeared in the approach of Heisenberg and Dirac (Sec. 14), to become the central element of any
quantum theory.

An observable quantity of classical physics represented an infinitely precise classical measurement

assigning to every state exactly one number and therefore, defining a unique real function on the
classical set of states (phase space). In QM the observable properties of micro-objects are still detected
by the macroscopic measuring devices but their domain of action is different. The measurement is
typically performed on a statistical ensemble of many particles in the same state . Since the result is
unpredictable (even if one knows ), the quantum observable cannot be represented by a function f
with a unique value f( ) for each . It is a random variable, i.e., a whole collection of possible values
j and the corresponding probabilities pj( ) (j = 1, …, n). Yet, the classical random variable can neither
reflect the complexity of the concept. Here, a natural place opens for the linear operators. Since the
QM probability distributions are quadratic in (Born 1926), then quadratic is also the statistical
average of any quantum measurement leading to a unique self-adjoint operator A, which becomes an
optimal candidate to represent the QM observation of particle ensembles.

- 12 -
10.5. Observables as operators: the anatomy of the measurement

Some details of this representation are essential. Without loosing generality, one can imagine a
quantum ensemble as a particle beam entering into the measuring apparatus. The principal part of the
apparatus is a mechanism which splits the beam into n parts of special properties. Though the splitter
might be in one piece, it is convenient to represent its action by a system of linear operators (filters) F1,
F2, …, splitting the wave function into a sequence of linearly independent superposition components
j = Fj (j = 1,2,…) characterizing the measurement. The fact that the second application of the filter
Fj to the already selected component j is irrelevant, Fj j = j, means that Fj2 = Fj and the fact that Fj
disregards all other components i implies FjFi = 0 for i j. In turn, the linear independence of j and
the decomposition = 1 + 2 + + n lead to the identity F1 + + Fn = 1, i.e., the filters Fj are the
projection operators splitting the Hilbert space of states into the linearly independent subspaces
distinguished by the apparatus. The measuring device is also endowed with a system of detectors D1,
D2, …, Dn; each Dj reacting to only one component j and blind to the rest (Figure 10).

FIGURE 10. The scheme of the quantum measurement with a finite number of filters and detection areas. The
corresponding quantum observable is represented by the operator (10.5.1).

The detectors D1, D2, …, Dn, are then labeled by a scale of real numbers 1, …, n defining the
numerical results of the measurement. Since the particles are indivisible, each one can be registered by
only one of the detectors Dj; if this happen, the result of the measurement is considered j. The wave
function should now define the probabilities p1( ), p2( ), …, that the particle will be detected by D1,
D2, … By generalizing the Born’s interpretation, QM assumes that the probabilities depend
quadratically on the magnitudes of the filtered components: if || ||2 = 1, then pj( ) = || j||2 = ||Fj ||2. In
order to grant that the sum of probabilities is 1, it is assumed that the filtered components are mutually
orthogonal, i| j = 0 if i j, implying that Fj are orthogonal projectors, i.e. Fi = Fi† (v. Neumann
1932; Dirac 1933; Riesz and Nagy 1990). The capacity of comprising the entire statistical information
thus possesses the linear operator A defined as the combination of the numerical results j and the
apparatus filters:

- 13 -
 n
A= k Fk . (10.5.1)
k 1

In particular, the statistical average (expectation value) of the measurement is

= nj 1 j|| j||2 = n
j 1 j |Fj| = |A . (10.5.2)

Apart of the average (10.5.2), the operator (10.5.1) defines all remaining aspects of the measurement.
The special role belongs to the components 1, …, n, privileged (filtered) by the apparatus. If the
initial particle state was = k, then with certainty (probability || k||2 = || ||2 = 1) it will be registered
by the detector Dk, and so, the measurement result will be k. Since Fk k = k and Fj k = 0 for j k,
then simultaneously A k = k k (k = 1, 2, …, n). Hence, the possible numerical results of the
experiment are always the eigenvalues of A, and the filtered state components k are its eigenvectors.
When expressing k in terms of the normalized eigenvectors: k = ckuk, where ||uk|| =1, ck , i.e., =
n
k 1 ckuk, the statistical interpretation of acquires the traditional form (Rubinowicz 1968, Cohen-
Tanoudji, Diu, and Lalöe 1977, Ballentine1998):

pj( ) = |cj|2 (j = 1,2,…, n) (10.5.3)

The expression (10.5.1) enjoys some universality: all self-adjoint operators in a Hilbert space are either
special or limiting cases of (10.5.1) with finite, denumerable or continuous eigenvalues. This is granted
by the diagonalization theorem of the Hilbert space theory, where the set of eigenvalues of any
operator A (no matter whether it is the energy) is called the spectrum of A (v. Neumann 1932, Riesz
and Nagy 1990), leading to the general assumptions of QM valid for all types of measurements with
finite, infinite or continuous numerical scales.

10.6. General formalism; the extended Born’s hypothesis

The operator representations (10.5.1-2) and the probabilities (10.5.3) lead to a general formalism which
repeats itself in all pictures of QM (of Schrödinger, Heisenberg or Dirac). An observable is represented
by a self-adjoint, linear operator A = A†. The eigenvectors (eigenfunctions) u of A satisfying Au = u
are special physical states for which the results of the measurement are predictable: if the micro-object
is in the state u , then the result of measuring the observable A must be with the probability 1. If is
a superposition, either countable or continuous, with the coefficients c( ), of many eigenstates u
(properly normalized), then the output of the measurement is uncertain and |c( )|2 yield the probability
distribution for the unpredictable result on the apparatus scale (cf. Dirac 1935).

10.7. Functions of an observable

Some convenient properties of this formalism should be noted. The operator functions f(A) admit an
expression analogous to (10.5.1)

n
f(A) = f( j) Fj. (10.7.1)
j 1

- 14 -
 '
representing the same measuring apparatus of Figure E but with transformed numerical scales i i
= f( i), i = 1,2,…,n. The expression (10.7.1) is specially useful to determine the quadratic (Gaussian)
and higher order statistical moments of the observable A for any state . Thus, e.g., the quadratic error
A (Gaussian uncertainty) differing the single results from the expected value (10.5.2) is given by:

n
2 2
( A)2 = = j pj( ) = < | (A - )2 | > (10.7.2)
j 1

10.8. Transition probabilities: the most elementary statistical law

One of the simplest measuring devices is a filter checking for the micro-systems in a certain given state
H, || || = 1. The corresponding operator A is just the orthogonal projection F onto the 1-
dimensional subspace spanned by the unit vector (In the Dirac’s notation F = | |; c.f. Sec. 15).
One might think that for the initial state vectors , the result of testing on must be negative, but
this is not the case. If is parallel (proportional) to , the output of checking is certainly yes; for
orthogonal to it is certainly not but if is a superposition of both, = c0 + , where , then in
agreement with (10.5.3) the uncertain result depends on the coefficient c0 = | . The number

p( , ) = | | |2, (10.8.1)

is the transition probability; it defines the chance that the system in the initial state will pass the -
checking (compare the polarization filters of Figure 9). Though describing the probabilities of an
instantaneous check, the quantities (10.8.1) are also relevant for the decay and radiation processes
considered in quantum optics (Cohen-Tanoudji, Dupond-Roc and Grynberg 1992, Scully and Zubairy
2002).

10.9. The construction of the traditional observables

While the scheme of Figure 10 involves the filters Fj, the traditional QM operators are constructed
rather on the basis of formal or aesthetical arguments, as the “quantum analogues” of classical
quantities. The particle position in QM is represented by the operator triple x = (x1,x2,x3) which
multiplies the wave function (x) by the triple of the space coordinates. Their spectra (sets of
eigenvalues) are continuous, the filters are any devices which yield the particle localization (slits,
screens, Wilson camera, bubble chambers, etc.) Since de Broglie waves (7.1) have the well defined
momenta, they should be the eigenfunctions of the momentum operator, which therefore is defined as p
= i x1 , x2 , x3 . The eigenvalues are continuous, and the filters correspond to the traditional
spectroscopic analysis. An arbitrary function f(p) = f(p1,p2,p3) of the classical momentum variables is
2
replaced by f i x1 , i x2 , i x3 . The kinetic energy operator becomes Ekin = 2pm = - 2 m 2 =
2

2 2 2 2

2m 2
x1 2
x2 2
x3
. The angular momentum operator (relative to the coordinate center x = 0) is
defined as:

M=x p= i x Mx = i (y z z y), My = i (z x x z), Mz = i (x y y x). (10.9.1)

The energy operator (quantum Hamiltonian) for a non-relativistic spinless particle of charge e and mass
m in presence of an external potential V(x,y,z) is

- 15 -
 p2 2
H V ( x) V ( x) (10.9.2)
2m 2m

while in presence of the external magnetic field represented by the vector potential A(x):

2 2
1 e 1 e
H p A V ( x) A V ( x) (10.9.3)
2m c 2m i c

The generic property of the quantum observables is their non-commutativity, i.e., the non-vanishing of
the commutator [A,B] = AB BA. Thus, e.g., the position xj and momentum pk = i xk satisfy:

[xj,pk] = i jk (10.9.4)

where jk is the Kronecker delta. The components of the angular momentum fulfill non-trivial
commutation relations:

[Mx,My] = i Mz, [My,Mz] = i Mz, [Mz,Mx] = i My (10.9.5)

11. THE UNCERTAINTY PRINCIPLE: INSTRUMENTAL AND MATHEMATICAL

ASPECTS

The fact that the quantum states with well defined momenta correspond to de Broglie waves had
important consequences. One of them was the impossibility of simultaneous, sharp measurement of the
position and momentum observables. The first glimpse of this law was due to Dirac and Jordan
(Jammer (1966), Ch. 7.1); its exact formulation belongs to Werner Heisenberg 1927, and henceforth it
is known as the Heisenberg’s uncertainty principle.

To give a realistic argument, Heisenberg devised a thought experiment to measure the position of an
electron by using a single photon in an idealized microscope. When the photon collides with the
electron, the momentum of both particles is changed as ruled by the Compton’s law. The scattered
photon passes through a lens and lands somewhere on a screen while the electron recoils (see Figure
11). The challenge was to deduce the initial position of the observed electron.

FIGURE 11. The Heisenberg microscope introduced as a thought experiment to illustrate the uncertainty in the
simultaneous determination of the position x and the x-component of the momentum of an electron.

- 16 -
A part of the answer was known from the early optical works of Ernst Abbe (1873), who considered
the Numerical Aperture and minimum resolving distance of a microscope (see Volkmann 1966). The
Numerical Aperture (N.A.) is determined by the angle formed by a point-like object and the
microscope objective N.A. = nosin (were no is the refraction index of the medium). The minimum
resolving distance x, in turn, is proportional to the wavelength of the scattered light and inversely
proportional to N.A.:

x /sin . (11.1)

If is given, (11.1) means that the smaller the value of , the finer will be the measurement of the
electron’s position. To determine the position as accurately as possible, the light of the shortest
available wavelength (e.g., gamma-rays) is required. However, a new fact was the Compton’s
scattering law (4.1), implying that the error in the reconstruction of the initial momentum component px
of the electron increases proportionally to sin and to the energy of the scattered photon:

px = 2 (h/ )sin . (11.2)

The simultaneous validity of (11.1) and (11.2) produces a trade-off: the greater the precision in the
measurement of the electron’s position, the greater must be the imprecision of the electron’s
momentum x px h. The original Heisenberg’s paper (1927) opened the general questions about
what can be and what can not be measured in quantum theory (see also v. Neuman 1932, Ch III.4).
Immediately, Bohr (1927, 1928) faced the same problem and improved the Heisenberg’s argument (see
Jammer 1966, 1974). The present day text book versions owe a lot to the Bohr’s interpretation (see,
e.g., Heisenberg 1930, Born 1969, Bohm 1951, Hilgevoord and Uffink 2001).

In mathematical terms, the Heisenberg’s principle reflects the interdependence between (x) and its
Fourier transform. If || ||2 = 1, the probability density for the position measurement is given by Born’s
expression (x) = | (x)|2. To grant the predictable position, the wave packet must be concentrated in a
very small space domain. Inversely, to grant the predictable momentum, must approach the eigenstate
of p, i.e., the widely spread plane wave (7.1). Due to (10.5.3), the probability density (p) for the
momentum measurement is given by the Fourier transform c(p):

1
(x) = e ipx / c(p) d3p (p) = |c(p)|2 (11.3)
3
(2 )

The functions (x) and c(p), in general, show an inverse behavior: the more concentrated one, the more
spread the other. The fact was verified by Heisenberg, Jordan and other authors for Gaussian packets
(Jammer 1966, Ch. 7.1). Indeed, for any normalized the Fourier analysis grants that, the square
(Gaussian) uncertainties xi and pi defined by (10.7.2) obey:

xi pi (i = 1,2,3) (11.4)
2

(both, the microscope version and the formal proof of (11.4) can be found already in v. Neumann 1932,
Ch III.4). As subsequently found, the Heisenberg’s microscope and the simple Fourier argument (11.3-
4) are just a fragment of a general law. An elementary Hilbert space theorem for any two real operators
A, B tells that if [A,B] = iC, then for an arbitrary normalized state (wave packet) the quadratic
(Gaussian) uncertainties (10.7.2) fulfill:

- 17 -
 A B 1 | |C |. (11.5)
2

(see e.g., Rubinowicz 1968; Ballentine 1998, Ch. 8.4). The law (11.5) extends (11.4) and shows that
the uncertainties are a typical phenomenon in quantum domain (the frequently discussed time-energy
uncertainty does not obey so simple arguments; see Muga, Mayato and Egusquiza 2002). Curiously,
the above “quantum fuzziness” is what makes possible the sharply defined (“discrete”) energy levels.

12. TYPICAL STATES AND SPECTRA

12.1. Harmonic oscillator: spectral ladders, eigenstates, coherent states

One of merits of QM is that the discontinuous energy levels (discrete spectrum) did not need to be
artificially assumed. There is nothing discontinuous in the oscillator potential V(x) = 1 m 2 x2
2
considered by Planck. Yet, the corresponding quantum Hamiltonian Hosc admits only a discrete
sequence of the energy levels (Dirac 1935). This is due to the existence of the non-commutating
d d
annihilation and creation operators: A = 1 y , A† = 1 y , where y = m
x, with
2 dy 2 dy
[A,A†] = 1, and

1 d2 y2
Hosc = = (A†A + 1
2 ) (12.1.1)
2 dy 2 2

y2

The multiple application of the creation operator A† to the ground state |0 = c 0 e 2 (c.f. the Dirac’s
symbols, Sec.14) generates the common eigenstates |n of the operator N = A†A (with the ladder of
eigenvalues n = n):

1
|n = (A†)n |0 . (12.1.2)
n!

and of the energy operator Hosc with an equally spaced ladder of eigenvalues: En = (n + 1
2 ) ,
n=0,1,2,… Inversely, the applications of A leads down the ladder (see Figure 12).

FIGURE 12. The spectral levels of the harmonic oscillator generated by the creation operator.

The oscillator (12.1.1) and the operators A, A† are crucial in almost all domains of quantum theory due
to their alternative interpretation: The Hamiltonian (12.1.1) represents the system of many bosons,
occupying the same state (“nest”) of the energy , the eigenvalue n of N means the occupation

- 18 -
number, A† and A increase or decrease by 1 the boson number N, and 1
2 is the vacuum energy
contribution (“empty nest”).

The use of A†, A, is not limited to the construction of the eigenstates of Hosc. It reveals also the
existence of a continuous family of coherent states z , defined as eigenstates of the annihilation
operator A z = z z with complex eigenvalue z C (Schrödinger 1926; Glauber 1963a-c, Glauber
1965; Klauder 1963a-b), symbolically constructed as

z = exp( z 2/2) exp(zA†) 0 (12.1.3)

In terms of the Schrödinger’s wave functions z ( x) x z , the coherent states (12.1.3) are the Gaussian
wave-packets:

1 2
4 | z|2 z 2 1
z ( x) = exp exp 2z x (12.1.4)
2 2

Their important property is the minimal allowed value of the position and momentum uncertainty
x p= 2 . Once again, the importance of the concept exceeds the traditional theory of the quantum
mechanical oscillator (R.J. Glauber, Nobel Prize 2005). In quantum optics, the monochromatic light
forms multi-photon states with the photon number corresponding to the eigenvalues of N =A†A. The
coherent states then represent the “most classical” field forms, where the uncertainty of the field values
is the smallest possible, at the cost of an indefinite photon number. The most important field modes of
this kind are the laser and maser beams, approximating the plane waves of classical electrodynamics
and the coherent, pulsating fields of micro-cavities, both crucial for the experimental techniques of
quantum optics (Cohen-Tannoudji, Dupond-Roc, Grynberg 1992; Scully and Zubary 2002), of
common interest for the quantum control (c.f. Sec 16), particle and molecular physics, and fundamental
problems of quantum theories (see e.g. Ashkin 1970; Wineland and Itano 1979; Golub et al 1979;
Ekstrom and Wineland 1980; Phillips and Metcalf 1987; Cohen-Tannoudji and Phillips 1990; Chu
1992; Cohen-Tannoudji 1998; Arndt et al 1999; Nielsen 2003; Hackermüller et al 2003; Hornberger et
al 2003; Hau 2004)

12.2. The angular momentum

One of the most typical examples where the non-commutativity yields sharply defined (‘discrete’)
spectra is the angular momentum M (10.9.1). In this case, the eigenvalue equation defines only the
angular dependence of the wave function. The common eigenvectors can be shared by M2 and one of
the M components, e.g., Mz. The remaining two components Mx and My then define the ladder
operators b = Mx iMy, b† = Mx + iMy, which intercommunicate the different eigenvalues of Mz. The
eigenvalues of M2 are 2 l (l 1) , labeled by the quantum number l = 0,1,2,… For each fixed l the
corresponding eigensubspace of M2 contains a finite chain of 2l+1 eigenvectors of Mz (“spherical
functions”) generated explicitly by b and b†, labeled by the quantum number m = l, l +1, …, 0,
…, l 1, l.

- 19 -
12.3. Hydrogen and hydrogen-like atoms

2
The attractive Coulomb potential in 3-dimensions V(r) = Zer , r = x 2 y2 z 2 , is the principal
force field acting on the atomic electrons. The corresponding Hamiltonian:

2
p Ze 2 2
Ze 2 2
1 2 1 M2 Ze 2
H= = (12.3.1)
2m r 2m r 2m r r 2 2m r 2 r

possesses the continuous spectrum on the positive energy axis E > 0. The related (unnormalized)
scattering states are analogues of the unbound hyperbolic/parabolic trajectories of the classical Kepler
problem. Besides H has also a discrete spectrum of negative eigenvalues, explaining the existence of
the allowed energy levels without the need of any artificial assumptions. They correspond to the
common eigenvectors of H, M2 and one of the M components, e.g. Mz. In each common eigensubspace
of M2 and Mz, defined by the fixed l, m, l 0, |m| l, the original Hamiltonian H has a sequence of
eigenvectors labeled by the radial quantum number k = 0,1,2,…, corresponding to the eigenvalues:

ZR
Ek+l+1 = (12.3.2)
(k l 1) 2

The eigenvectors with distinct k, l, m, add up to the same degenerate level of H if they share the same
value of n = k + l + 1, called the principal quantum number. The general nth level contains one state
of l = 0, three states of l = 1, five states of l = 2,…, 2n-1 states of l = n 1 and 2n +1 states of l = n;
altogether 1 + 3 + 5 +…+ 2n + 1 = n2 states (see Figure 13), mathematically represented by the
products of spherical and radial functions (c.f. Dirac 1935; Feynman, Leighton and Sands 1963;
Cohen-Tannoudji, Diu, and Lalöe 1977).

FIGURE 13. The schematic representation of the bound states of the hydrogen atom. The states (dark spots) are arranged
on a sequence of expanding triangles (corresponding to the energy levels) with n = 1,2,…, numbering the increasing subsets
of l < n, m = l, …, l. Each spot corresponds to a separable, square integrable solution of (8.2). The numbers of
independent solutions in the subsequent energy levels are 1, 22, 32, …, n2, …, but due to the electron spin (c.f. Sec. 12.5),
the number of independent states on each orbit is twice as great.

The structure, though elementary, shows some deeper group theoretical regularities, as e.g., the SO(4)
symmetries (the accidental degeneracy; see Engelfield 1972), explaining also the Pythagorean links
(the unitary equivalence) between the 3-dimensional oscillator H3osc and the hydrogen Hamiltonian
Hbound (restricted to the subspace of the bound states, with ZR = 21 2 ):

- 20 -
 1
H3osc ~ 1
( H bound ) 2

12.4. Pauli’s exclusion principle

The degenerate structure of the energy levels of Figure 13 was known already in the old QM of Bohr-
Sommerfeld, thanks to the Zeeman’s discovery, that the spectral lines can be split by strong magnetic
fields into the multiplets corresponding to the theoretically predicted quantum numbers l, m. Yet, from
the very beginning some unsolved questions concerned the many-electron atoms. Why cannot all
electrons accumulate simply in the ground state? The total number of electrons and the occupation
numbers of the subsequent orbits for some optically examined atoms were already known. Trying to
explain them, Wolfgang Pauli formulated in 1925 the exclusion principle (Pauli 1925a-b), allowing no
more than one electron in the same quantum state. In his own words: “An entirely non-degenerate
energy level is already <<closed>> if it is occupied by a single electron” (c.f. Pauli 1946).

The law permitted to represent the multi-electron atom as a many level (shell) structure, each level
containing a finite number of cells. Once all cells on the lower shells are occupied by electrons, the
next ones can be added only to the next levels. The effect explained qualitatively the atomic structure,
including the periodic table of elements. However, the electrons detected on the atomic levels were still
too many: twice as much as permitted by the already known quantum numbers. Moreover, an unsolved
problem was the existence of Zeemann’s doublets, which could not be explained on the ground of the
known Bohr-Sommerfeld rules (the anomalous Zeemann effect; c.f. Jammer 1966, Ch. 3.3-4). A semi-
phenomenological research based on the old quantum theory showed that the states of the normal and
“anomalous” Zeemann effects could be conveniently ordered by introducing the generalized quantum
numbers j = l ½, where l is the orbital momentum. Pauli raised the hypothesis about the existence of
a new quantum number allowed to take only two values s = ½, the “two valuedness not describable
classically” (Pauli 1925a). The hypothesis saved the exclusion principle and was consistent with the
1921-22 Gerlach-Stern experiments for the silver atoms (Sec. 19.2). However, what was the new
quantum number?

12.5. Particle spin: a controversial non-classical property

The idea that the “two valuedness” might be due to the intrinsic properties of the electron rather than of
the atomic core, was circulating during several years. In 1921 Compton raised the hypothesis about the
spinning magnetic electron, though Bohr and Pauli were strongly against. In 1925, R. Kronig presented
the hypothesis explaining the new quantum numbers as a property of the electron, but the idea was
rejected by Pauli and Heisenberg. Kronig gave up the publication. In the same 1925, the idea
reappeared in the proposal of Uhlenbeck and Goudsmit. Discouraged by Lorentz, they would drop the
publication, if not the support (and an angry reaction) of Ehrenfest. Their short notes appeared finally
in Naturevissenschaften 1925 and in Nature 1926. Pauli called still the particle spin a “heresy”, but he
gave up in 1926. He then proposed a simple operator representation of the spin observable, as an
internal degree of freedom of an electron. The spin in the direction of any unit vector n = (n1, n2, n3) is
represented by the 2 2 matrix (½)n , where = ( 1, 2, 3) are the Pauli matrices:

0 1 0 i 1 0 2 2 2
1 , 2 , 3 , 1 2 3 1
1 0 i 0 0 1

- 21 -
and (as a historical paradox) the electron spin in the non-relativistic QM has now the name of its most
outstanding opponent (Pauli spin). The spin components (½)n have the identical eigenvalues ½,
and fulfill the characteristic algebraic identities: 1 2 = i 3 = 2 1, 3 1 = i 2 = 1 3, 2 3 = i 1
2 2
= 3 2. If m, n are any two unit vectors, then (n ) = (m ) =1, [n , m ] = 2i(n m) . The spin
½ is the most elementary example of the general spin observables independent of the particle position
and orbital momentum, expressed by the matrix operators acting in the finite dimensional spaces of the
internal degrees of freedom, with the eigenvalues taking 2j+1 values j, j+1, …, j 1, j, where j = 0,
1/2, 1, 3/2, 2, … The photon polarizations correspond to the spin 1 and the hypothetical graviton
should be a particle of the spin 2. In the mathematical representation theory the spin together with the
orbital momentum M, define the generator J = M + of the state transformations representing
2
effects of the rotation group (Corson 1953; Budinich and Trautman 1988; Penrose 2004; Pleba ski
2006).

In the dynamical theories the spin ½ of Pauli is coupled with the external magnetic field B via the
e
magnetic moment = . For the electron, the coefficient is known as the Bohr’s magneton
2mc
(see e.g. Schiff 1968, pp 440) and its peculiarity (magnetic anomaly) is that the ratio of the spin
magnetic moment to the spin rotational momentum is twice as much as for the orbital motion. The
famous result of Gerlach-Stern 1921-22 experiments could be now understood as an effect of the spin
½ of the silver atoms (c.f. Sec. 19.2). A dramatic macroscopic proof of the magnetic anomaly was the
observation of the angular momentum acquired by the ferromagnetic samples in events of
demagnetization (Weyl 1950, Ch. IV.4). In the micro-scale, contributes to the neutral or charged
particle Hamiltonians (10.9.2) and (10.9.3) with an extra (additive) term B, which commutes with
e
both (10.9.2) and (10.9.3). They explain the corrections B in Zeeman’s experiments splitting
2 mc
each energy level of H into a pair of new ones (Dirac 1935, Schiff 1968, Ballentine 1998).

12.6. The spectral bands

The periodic potentials V(x) in the n-dimensional space n (where n = 1,2,3), typical for the crystalline
lattices, are invariant under the symmetry group including n independent space displacements di (i =
1,…, n), V(x + di) V(x). One of the simplest models is an infinite lattice of identical potential centers
extending all over the space. The particle moving in the lattice cannot be trapped by one of the centers,
even if they are attractive, since the tunnel effect opens an escape route to the following ones. The
result are stationary states of motion in form of the Bloch functions resembling the distorted de Broglie
waves:
i
p (x) = (x) exp( px) (12.6.1)
where the free propagation term exp( i px) is corrected (modulated) by a function (x) sharing the
periodicity of V(x). The (numerical) vectors p in (12.6.1), i.e., the quasi-momenta, defining the energy
eigenvalues in (8.2), belong to the Brillouin zones in which the interactions with the lattice centers
permit the creation of the stationary motion patterns. The allowed energy eigenvalues form the
continuous bands (Reed and Simon 1978), interrupted by the spectral gaps (also: forbidden zones). The
motion of the electrons in higher, partly populated spectral bands (conductivity bands) explains the
origin of the electric currents in the metallic lattices, while the motion of the gaps in the electrons
population of the lower bands accounts for the hole conductivity.

- 22 -
The simplest case is the 1-dimensional Schrödinger equation with a periodic potential, called the Hill’s
equation (Magnus and Winkler 1973):

2
d2
V ( x) E , V(x + d) V(x), x . (12.6.2)
2m dx 2

The list of physical applications extends from the electron states of one dimensional DNA molecules,
to the entire solid state technology. If the periodic structure of V(x) is locally affected by impurities
(periodicity defects), it produces the bound states with the new energy levels, explaining the donor and
acceptor structure in semiconductors. The lattice deformations can also form the quantum wells,
quantum spots and contact effects, leading to the technology of transistors, chips and computers,
crucial for the present day informatics.

It might seem that (12.6.2) is limited to the quantum phenomena, but it is not the case. In their Physical
Review Letters’ paper, Avron and Simon (1981) propose to consider the spectral bands of quasi-
periodic potential V(x) as a model for the Saturn rings, an unexpected analogy between the micro and
macro universes (see Figure 14).

FIIGURE 14. The universality of the Schrödinger’s equation with the periodic or quasi-periodic potential: (a) The band
structure of the potential V(x) explains the spectra of the solid state physics; (b) After reinterpreting x as t (time) and E as
the great ellipse axis, the same structure inspired the model of the Saturn rings (Avron and Simon (1981)).

13. UNITARY EVOLUTION

It is important that the energy operator (10.9.2) coincides with the operator postulated by Schrödinger
in his evolution equation (8.1), which therefore reduces to:

i H (13.1)
t

By introducing the evolution operator U(t,t0) such that (x,t) = U(t,t0) (x,t0), one can write the
evolution laws (13.1) in the operator form:

d
i U(t,t0) = H U(t,t0); U(t0,t0) = 1, (13.2)
dt

holding for all quantum systems with time independent or the time dependent Hamiltonians H = H(t).
The fact that the evolution process (8.1, 13.1) conserves the scalar products (10.2.1) was initially

- 23 -
deduced from the particular form of (8.1). However, it is true always, when H is self-adjoint, H = H†.
The evolution operators then are unitary, i.e.:

U(t, t0)U†(t, t0) = U†(t, t0)U(t, t0) =1. (13.3)

If H does not depend on time, the evolution operator in (13.2) admits the simple exponential form:

U(t,t0) = e i (t t 0 ) H / (13.4)

The fact that the evolution equations (8.1), (13.1), (13.2) involve the energy operator (Hamiltonian) is
not accidental: while px, py, pz are the partners of x,y,z coordinates, the energy is the partner of the time,
in either Galileo or Lorentz invariant theories.

14. CANONICAL QUANTIZATION: SCIENTIFIC OR MAGIC ALGORITHM?

As turned out almost simultaneously, the Schrödinger’s wave mechanics was not the only way to
formulate the quantum laws. In 1927 Werner Heisenberg discovered the basic elements of a new theory
in a different form. His idea was to use the set of complex numbers, the hypothetical ‘transition
amplitudes’ between the oscillation modes of micro-systems. At the beginning, Heisenberg was not
aware that he was just describing the infinite matrices (i.e. operators) but soon, his theory proved
equivalent to the Schrödinger’s quantum mechanics. The Heisenberg’s vision, centered on non-
commutative aspects, contributed to the Dirac symbolic formulation (Dirac 1935). In it, the pure states
of a micro-object are not necessarily ‘wave functions’; they are represented by vectors , ,…
(called kets) of an abstract linear space. Their counterparts are vectors , , …, (called bras) of dual
linear space, yielding the numerical information . Dirac assumes the mirror correspondence
; the numbers , the Heisenberg’s transition amplitudes between the states and ,
generalize the inner products | of the Schrödinger’s picture. Instead of looking for differential
operators, Dirac follows the Heisenberg’s idea, fixing attention on the non-commutative aspects. The
classical quantities (c-numbers) in his vision are replaced by ‘non-commuting numbers’ (also: q-
numbers, Dirac 1935). It is observed that the algebraic structure of the commutators [A, B] = AB BA,
f g g f
reproduces some properties of the classical Poisson brackets {f,g} = . This suggests a
qi pi qi pi
quantization procedure. Whenever the classical theory offers a plausible choice of the canonical
1 i j
variables qi, pj, with {qi, pj} = ij = and the well defined Hamiltonian H(qi, pj,t), the quantum
0 i j
theory is constructed by choosing the non-commuting canonical operators qˆ i , pˆ i , and replacing the
classical Poisson brackets by the commutators:

{qi, pj} = 1 [ qˆ i , pˆ i ] = , [ qˆ i , qˆ j ] = [ pˆ i , pˆ j ] = 0, (14.1)

ij ij
i

the procedure known as the canonical quantization. In the dynamical part of the theory, the states ,
, …, are extra-temporal. The time dependence affects the non-commutative quantities A,B, …, and,
once again, it is defined by replacing the Poisson brackets in the classical motion equations by the
commutators:

- 24 -
dA 1
[ A, H ] (14.2)
dt i

Though equivalent to the Schrödinger approach (see e.g., Dirac 1935), the canonical quantization
(14.1) seems more practical to face new problems. So it is in the field theories, which admit a natural
choice of the continuous canonical variables to be ‘redressed’ into the operators. If needed, the process
can be repeated (so it happened with the Dirac wave function which was “second quantized” in
quantum electrodynamics). Some specialists object that the procedure is non-unique and repetitive;
they recommend to formulate a universal quantum field theory and then to deduce the quantum
mechanical and classical levels. However, such a universal theory does not exist. Moreover, the
heuristic methods of canonical quantization with all undefined steps might be more fertile than the pure
deduction method.

Some forgotten ideas unavoidably call attention. According to the old principles of magic, the
seemingly different phenomena may obey the same hidden rules. The laws for some objects or events
may be contagious for another events and objects. The above Middle Age principles of analogy were
long ago dismissed as anti-scientific. Yet, the XX c. quantum mechanics was created by bringing the
classical laws to a new terrain, in which the traditional classical quantities were redressed into the
operators (i.e., ‘infected’ by the quantum structure). At the bottom there seem to be an old idea about
the analogy or correspondence between different levels of natural phenomena (compare with the
mathematical ‘categories’, ‘functors’, and ‘dualities’.) So, were the old intuitions irrelevant?

- 25 -
15. THE MIXED STATES

The Dirac kets representing the pure states are non-unique: two unit vectors , , differing only
by a phase factor, = ei define the same state (c.f. von Neumann 1932, Dirac 1935). Hence, each
pure state corresponds to a whole bunch (ray) of proportional unit vectors. Its unique representant is
the one-dimensional subspace of H, or equivalently, the projection operator . In the dissipative
problems of quantum theory an important role belongs to the mixed statistical ensembles, including
distinct pure components. An average ensemble individual defines then a mixed state. While the pure
states correspond to the 1-dimensional projectors , H, the mixed ones are their finite or
infinite probabilistic combinations:

= n n n , n n = 1, n 0, n = 1, n H (15.1)
n n

where the coefficients n are fractions of the pure components n n participating in the mixture. In
the mathematical terminology (15.1) are positively defined, self adjoint, unit trace operators (Tr =
n n = 1) called the density matrices (Dirac 1935, Cohen-Tannoudji, Diu, and Lalöe 1977). The
decomposition of the mixed state into the pure components in QM is not unique. A customary
representation in terms of the orthogonal pure components, m n = mn, can be always obtained by
the spectral decomposition of in terms of its eigenvectors. However, the representation of the same
mixed state in terms of non-orthogonal components is also possible. In the Schrödinger’s representation
of QM, where every pure state evolves according to (13.1), the density matrix representing the time
evolution of mixed ensemble obeys the differential equation:

d
= i [H, ], (15.2)
dt

where H is the Hamiltonian of the system.

16. QUANTUM CONTROL: HOW TO MANIPULATE THE PARTICLE?

The early quantum theories typically described the micro-systems in fixed external conditions (e.g., in
static fields), with the evolution given by (13.4). However, the real quantum systems exist in a variable
universe where the external conditions can be changed and the evolution controlled. When submitted to
time dependent fields with varying Hamiltonians H(t) in (13.2), the micro-system can perform the
evolution transformations not achievable in static conditions. As known since 1936, in the ion traps the
oscillating electric and magnetic fields can block for a long time the natural spreading and scape of the
wave packets, thus making possible to check the QM and quantum optical effects for the single ions
(Penning 1936; Paul 1993). The ideas of applying the laser light to slow down the particle motion
(laser cooling) were put forward since 1975 by A.L. Schawlow, T.W. Hänsch, D. Wineland and H.G.
Dehemelt; then developed by S. Chu, C. Cohen-Tannoudji and W.D. Phillips. The generation of the
squeezed states for photons and charged fermions by the time dependent Hamiltonians was studied in
Yuen (1976) and followers (see e.g. the review by Dodonov 2002). The laser and maser fields turn
useful to stimulate or inhibit the atomic or molecular transitions as well as to break or catalyze the
chemical links (Brumer and Shapiro 1995). The mathematical results indicate that all simple quantum
systems are controllable, i.e., any unitary transformation in the Hilbert space of states can be in
principle produced by sufficiently rich external fields (Fu, Schirmer and Solomon 2001), though see
also the provocative questions in (Primas 1983, Ch. 1.4). For the outer orbits of the Rydberg atoms, the

- 26 -
techniques of “sculpturing” the wave functions is predicted (Weinacht, Ahn, and Bucksbaum 1998).
The existence of the non-dispersive “planetary states”, stabilized by the microwaves is also shown
(Buchleitner, Delande and Zakrzewski 2002; Maeda and Gallagher 2004). In the laboratory terms, the
problem of how to generate the desired unitary transformations, however, is far from concluded. One of
most promising methods is to stimulate the transitions between the selected pairs of bound states |n ,
|m of a certain given energy operator H0, with H0|n = En|n , H0|m = Em|m by applying a
monochromatic external field of resonant frequency : = Em En, with the time dependent
Hamiltonian:

H(t) = H0 + sin t (16.1)

If has a non-vanishing transition amplitude n| |m 0, (16.1) should cause the controllable Rabi
rotation between two eigenstates |n , |m (Rabi 1937; Rabi, Ramsey, and Schwinger 1954; Scully and
Zubairy 2002; Cohen-Tannoudji, Dupond-Roc, and Grynberg 1992), permitting to achieve any
superposed state |n + |m . An empirical technique of Rabi rotations between two energy eigenstates
(“blue” and “red”) of a Rydberg atom permitting to obtain the superposed (“pink”) states was designed
for the first time by Brune et al (1996) in the Ecole Normale Superieure (Figure 15).

FIGURE 15. The experimental setup applied in Ecole Normale Superiore permitted to generate Rabi rotations obtaining the
superposition of two Rydberg energy eigenstates of the Rubidium atom. Above: O, an oven ejecting the Rb atoms to the box
B, where the circular Rydberg states are prepared; S is a source of the coherent photon field injected to the cavity C in
which the Rb atoms are affected by the field pulses of resonance frequency; D, a detector registering the superposed energy
states (following Brune et.al. 1996).

An essential property of the monochromatic, resonant waves is their selectivity: while causing the Rabi
rotations between any chosen pair of levels, they affect very slightly any other pair if only the energy
spacing of the Hamiltonian H0 differ significantly. Basing on that, it is principally possible to design
the algorithms of step by step operations to achieve an arbitrary dynamical transformation in the n-level
subspace, e.g., inverting completely the level populations in the atoms or ions (Schirmer et. al. 2002).
However, the precision of all control methods goes down as the complexity of the system increases.
Apart of technical applications, the methods of quantum control are systematically studied due to their
significance for the fundamental problems of QM and for the possibility of quantum computing (c.f.
Sec. 18,22,23).

17. MEASUREMENT THEORY AND ITS CONCEPTUAL CONSEQUENCES

17.1. State reduction (collapse of the wave packet)

The interpretational problems of QM were in the center of attention of the Copenhagen School, headed
by N. Bohr, W. Heisenberg, W. Pauli, V.A. Fock, L. Rosenfeld et al (c.f. Jammer 1966, 1974; Primas
1983). The known mathematician, John von Neumann was on the threshold of the school, differing in

- 27 -
some important ideas, though his mathematical results contributed to the school philosophy. According
to the Copenhagen doctrine (shared at this point by P.A.M. Dirac), the pure quantum state describes
each single micro-particle and the measuring device is a macroscopic apparatus described in terms of
the classical theory. One of the most controversial points was the axiom about the reduction (collapse)
of quantum state suggested by the probabilistic interpretation of .

The laboratory experience shows that the measurement results in QM, though unpredictable are stable:
if the measurement of an observable A gave one of possible results (eigenvalues) n, and if the same
measurement is repeated immediately after, the result must be n again. In agreement with the
statistical interpretation (10.5.3), this certainty means that the micro-system state after the measurement
cannot be arbitrary; it is reduced (collapsed) to the corresponding eigenstate un (von Neumann 1932,
Dirac 1935). Some limitations of this argument are due to the existence of measurements of the
“second kind”, which cannot be immediately repeated (Landau and Peierls, and Pauli, see Primas 1983,
Ch 3.5) or cannot be repeated at all, since the micro-system does not survive the measurement
(Margenau 1936, c.f. also Jammer 1974, Ch 11.4). However, they could not change the central role of
the collapse postulate, which soon became one of fundamental QM axioms. Since its formulation, the
discussions continue, whether it is a real process, involving a physical change of the micro-system, or a
simple result of selection (nothing changes in the micro-systems, the only thing the apparatus does is to
select a new sub-ensemble; see also discussions by Jammer 1974, Ch 10.4; Haag 1996, Ch. VII).
According to the Copenhagen School the reduction is a physical process, caused by the macroscopic
apparatus, affecting each separate micro-individual (though the details of the doctrine are presented
differently by various authors; c.f. e.g. Jammer 1974). The most common explanation is that due to the
quantized (granular) character of the energy the apparatus cannot interact with a particle in an
arbitrarily subtle (non-invasive) way. The act of measurement thus involves an unpredictable and
impossible to eliminate perturbation (a jump) of the particle state described precisely as the ‘collapse of
the wave packet’. The most convincing argument that such an effect indeed occurs is the transmission
of the photon beams through the polarization filters (see Figure 9). Should the light beam a, polarized
in the direction a, fall into a polarization filter A selecting the polarization a perpendicular to a, no
photon could cross A . However, if the same beam a is first transmitted trough the filter B (measuring
apparatus) checking for the light polarization b, then a part of the photons emerging from B will cross
the subsequent filter A . Evidently, there was no sub-ensemble of such photons in the original beam a;
hence, the measuring device B not only selects but also affects the photons (see also Dirac 1935).

17.2. The strong uncertainty principle

The consequence of the collapse postulate was a stronger form of the uncertainty principle in QM. In
fact, if two observables (operators) A and B do not commute, then they do not have a basis of common
eigenstates in H: an eigenvector of A, in general, is not an eigenvector of B and vice versa. This
implies not only that both statistical errors (uncertainties) A and B cannot be simultaneously reduced
to zero, but also, that the simultaneous measurements of A and B are not possible at all: such a
measurement would have to produce the joint eigenstate of A and B which does not exist.

17.3. The idea of complementarity

The impossibility of the simultaneous measurement of non-commuting observables gave rise to an

important idea of complementarity. According to the interpretation developed by N. Bohr, W.
Heisenberg, Rosenfeld, and others, the quantum system possesses a variety of aspects (faces) which
can be revealed in the experiments; but some of them are complementary: if we observe one, we

- 28 -
unavoidably loose the other (Bohr 1948; Rosenfeld 1953; Heisenberg 1955, 1958; c.f. also Jammer
1966, 1974; Primas 1983). Some physicists assert that the corpuscular and wave-like behaviour are just
two complementary aspects of a particle: so either the particle shows its corpuscular or its wave nature.
Some philosophers and metaphysicists go still further, considering the body (material structure) and the
consciousness of a living organism as two complementary aspects (an opinion which still waits for a
verifiable formulation). A different interpretational school pays less attention to the metaphysical
problems, but more to the structural aspects: it notices that the quantum systems submitted to the
macroscopic tests seem to obey a new type of logic.

17.4. “Quantum logic”

The formal logical operations reflect the oldest activity of the human mind, which consists in
classifying the objects (phenomena) of the surrounding universe. The classification is done by the
elementary yes-no measurements which represent the statements (propositions) of the logic. The
proposition about an object is true if the answer of the experiment is yes. When applied to an ensemble
of classical objects, each yes-no experiment selects a subensemble. Hence, the propositions of the
logic correspond to the subsets of . The set theoretical relations in thus acquire a logical sense
illustrated by the Aristotelian circles: if a, b, c , then the inclusion a b is the implication (a b),
the complement a = \a is the negation (“not a”), the set theoretical sum and intersection are
interpretable as the alternative (“or”) and the conjunction (“and”) of the logic. The basic property
of the so obtained logic of subsets (Boolean logic) is its distributive law (see Figure 16.I):

(a b) c (a c) (b c). (17.4.1)

Departing from an analogous argument, Garret Birkhoff, John von Neumann, David Finkelstein,
Constantin Piron, and others developed the idea that the structure of the quantum yes-no measurements
reflects the “logic” of particles (“quantum logic”). Each yes-no measurement (filter) placed on the way
of a quantum beam is a proposition of the logic, describing a physical property of the beam individuals.
The proposition is true if the answer of the measurement is certainly yes. Differently than in case of the
Aristotelian logic, in the quantum case the positive answer is certain not on an unspecified subset, but
on a subspace of H, i.e., the eigensubspace of the filter corresponding to the eigenvalue 1 (As the
subspace of H one understands a subset which has all properties of H, is linear and complete, i.e.,
closed in H ). So, the propositions of the logic are now subspaces a, b, c, … H. The implication
and the conjunction (“and” of the logic) correspond to the set theoretical inclusion and intersection
of subspaces, the alternative (“or” of the logic) defines the smallest subspace containing both: a
b = a b, while the orthogonal complement a = a is the logic negation. Differently than the classical
logic, the quantum one is no longer distributive. An elementary example is any triple of nontrivial
subspaces a, b, c H, where a and b = a are orthogonal (interpretable as the yes and not
eigensubspaces of certain yes-no measurement) and c has trivial intersections with both, a c = b c =
{0}, but it is contained in their linear sum a b = H. Then (a c) (b c) = {0} {0} = {0}, whereas
(a b) c = H c = c {0} and the distributive law (17.4.1) is broken (Figure 16.II). Anytime
(17.4.1) fails for a, b = a , c, the propositions a and c are incompatible. The incompatible propositions
in quantum logic reflect the complementary properties of quantum systems in the sense of Copenhagen
School (Birkhoff and v. Neumann 1936; Finkelstein 1963; Piron 1964; Mittelstaedt 1972; Jammer
1974; Beltrametti and Casinelli 1981; Primas 1983).

- 29 -
FIGURE 16. I. The Aristotelian circles reflect the distributive property (17.4.1) of the classical (Boolean) logic. II. The
verifiable propositions about the quantum object are represented by the subspaces of the Hilbert space H. The logic breaks
the distributive law.

17.5. Quantum bit (qubit)

The simplest case of “quantum logic” is the structure of every 2-dimensional superposition subspace of
H, generated by any two linearly independent state vectors. The corresponding set of quantum states
| |, || || = 1, is isomorphic with the 2-dimensional surface of the Euclidean sphere S of the unit
diameter 1 in 3. The points of the sphere provide the unique images of all physically different
superposed states. Each pair of antipodes x = | | and x = | | represents an orthogonal pair of
states, corresponding to the outputs “0” and “1” of a certain yes-no measurement, and so, providing a
single bit of information. Yet, the whole sphere S is much more than one bit: it is formed by a
continuous family of orthogonal pairs (antipodes of S). Each one of them can be, in itself, a model of a
single information bit; though they are complementary (incompatible, linked by linear dependence and
nontrivial transition probabilities, see Figure 17): when one is checked, the others are lost of sight.
Their whole collection (the surface of the sphere) is what one calls the “quantum bit” (qubit). The
simple physical models of this structure are the states of the spin-1/2 fermions (electrons, nucleons,
etc.), or else, the polarization states of photons. The same geometry is shared by the states of an
arbitrary 2-level system. In fact, the state manifold of any quantum system can be viewed as an infinite
family of overlapping, intersecting or orthogonal qubits, an aspect essential for the quantum control and
for the recently developed ideas of quantum computing (see Sec. 23.2).

FIGURE 17. A continuous family of complementary (incompatible) bits (antipodes) accounts for a rich structure of the
quantum bit (qubit) represented by the surface of the sphere S of the diameter 1. The transition probability between any
state (point) a and a pair of antipodes b, b representing a complementary bit are given by p(a,b) = |ab |2 and p(a,b ) =|ab|2;
their sum is p(a,b) + p(a,b ) = 1 due to the Pythagoras theorem. The simplest physical models are the polarization states of
the photon or the spin states of a spin-1/2 particle. Above: the equator of S collects the linear polarization states of the
photon, poles are two opposite circular polarizations, and the remaining points on the surface represent elliptic polarizations.

- 30 -
18. INTERPRETATIONAL POLEMICS AND PARADOXES

18.1. Quest for alternative theories

The structure of quantum mechanics with the deterministic evolution (8.1, 13.1, 13.2) but non-
deterministic (statistical) interpretation including the uncertainty and complementarity, found the
outstanding opponents and is still the subject of fundamental discussions. Luis de Broglie believed that
the wave function obeying (8.1) is just a guiding physical field (“pilot wave”) on which the proper
particle propagates in form of the second, singular solution of (8.1), whose singularity (point-source)
obeys the well known Born’s statistics (de Broglie1960, Holland 1993). A tempting hypothesis was
also the existence of some invisible, classical state variables (hidden parameters) which would
determine with an infinite precision the results of all quantum measurements; so that the known
probabilistic results of QM would be just the results of the classical statistics. Given the hypothetical
set of the (hidden) classical states, the quantum states (represented by the ket vectors |a , |b , …)
should then correspond to the subsets a, b,… (see Figure 16.I), together with some probability
measures, describing the probability distributions of hidden variables “inside” of each quantum state.
The orthogonal quantum states |a |b would correspond to the disjoint subsets a b = ; the QM
transition probability would become the conditional probability p(a,b) = measure(a b), and the
statistical interpretation of any measurement for which the system in an initial state |a H chooses
between a sequence of orthogonal eigenstates |b1 , |b2 ,… H, would be given by the probability
measures of the corresponding intersections: a bk, k = 1,2,… The difficulty of this Boolean
representation was that the quantum filtering operations, in general, don’t commute, so they cannot be
reduced to a “passive” selection of the subsets (but see the ingenious models in Accardi 1981, and
Bohm and Bub 1966, where the classical variables are not only selected but also affected by the
measurements). In a more general attempt of David Bohm (in a sense, inverting the original heuristics
of Schrödinger) the classical trajectories of quantum particles emerge once more as legitimate partners
of the quantum mechanical wave packets (Bohm 1952a-b, Goldstein 1998), though the theory shows
some gaps and arbitrary assumptions. Indeed, as later proved, the theories based on hidden parameters
cannot reproduce the quantum mechanical probabilities for dim H 3 (Gleason 1957; Kochen and
Specker 1967, see also the anthology of Hooker 1975). The statistical analogue of Bohm’s idea was
proposed by Edward Nelson (the stochastic particle painting the picture of the wave packet, Nelson
1967; de la Peña and García-Colin 1967, 1968; de la Peña and Cetto 1969); though apparently, not all
interpretational questions could be answered (see e.g., Gilson 1968; Albeverio and Hoegh-Krohn 1974;
Kracklauer 1974; Mielnik and Tengstrand 1980). It might be that the real chance of the stochastic
approaches is not so much to reproduce, but rather to amend the quantum theories (see the earlier
remarks by J. v Neumann 1932, Ch III.2, p 210; though consult also the locality counter-arguments in
Sec. 22.1-2). Nonetheless, the field equivalents of the stochastic QM were also proposed (de la Peña
and Cetto 1996).

18.2. The measurement problem

The strongest doubts against the orthodox interpretation of QM were expressed by Erwin Schrödinger
(Moore 1990, p 473) and by Albert Einstein (relativistic case). The center of discussion from the very
beginning was the assumed reduction of the wave packet. If the general evolution law (8.1, 13.2) is
linear and if the eigenstates |u1 , |u2 , … of the measured observable are unchanged by the
measurement (Figure 10), then there is no way how the nontrivial superposition | = c1|u1 + c2 |u2
+… (with c1, c2, … 0) could collapse into one of the eigenstates |u1 , |u2 ,… Due to the fundamental

- 31 -
linearity of the evolution law, it must be transformed into a new superposition | | = c~1 |u1 + c~2
|u2 + , where none of the components |u1 , |u2 , … can be missing and the coefficients cannot change
their norms, | c~k |2 = |ck|2. The argument might seem oversimplified if the micro-system is in a feedback
interaction with the measuring apparatus: however, taking into account that the measuring device itself
is a huge quantum system, the linear evolution law (13.1) should be recovered for the joint state of the
micro-system and the apparatus, assuring once again the linear, deterministic state evolution, excluding
the possibility of an unpredictable collapse.

18.3. Borderline doctrine

The answer of the Copenhagen School was the borderline doctrine sustaining that the reduction of the
wave packet (quantum state) cannot be perceived on the purely microscopic level of the theory. To
account for it, the physicist must draw a borderline (also: the Heisenberg’s cut) between the quantum
system itself (the object) and the rest of the universe (the subject, including the external observers and
their measuring devices). The borderline is not uniquely defined but it must be always placed
somewhere. While the evolution of the quantum system is described by the Schrödinger equation, its
interaction with the sensitive but macroscopic surrounding is represented in a simplified way by the
reduction (collapse) axioms (Jammer 1974; Primas 1983). This is sometimes considered as one more
manifestation of the complementarity (if one wants to describe with all details the quantum evolution,
one looses of sight the measurement, and vice versa). As confirmed by the mathematical results of John
v. Neumann, the final statistical predictions of the theory are always the same, no matter where exactly
the borderline is placed (v. Neumann 1932, Ch VI). In the opinion of many physicists, though,
(including v. Neumann himself) the borderline masks only some unsolved problems.

18.4. The paradox of the Schrödinger’s cat

The most provocative illustration belongs to E. Schrödinger who considered a living organism: a cat in
a hermetically closed box, containing also a radioactive atom, a detector and an apparatus programmed
to kill the cat if the decay is registered (Schrödinger 1935). At the beginning, the atom is undecayed
and the cat is alive. Now, if the detector, the killing machine, and the cat are considered as a
macroscopic measuring apparatus, then at certain moment the atom decays, triggering the detector and
the cat is killed. If, however, the detector, apparatus, and the cat are considered a huge quantum
systems, then the only thing which can happen is that the system will find itself in a superposed state
dictated by the linear evolution: atom undecayed, cat alive = |ALIVE , and atom decayed, cat dead =
|DEAD . After the average life time of the atom, the difference between both predictions is remarkable:

According to the measurement axiom:

(18.4.1)

However, according to the quantum evolution law:

|ALIVE c1|ALIVE + c2|DEAD , | c1 | 2 = | c2 | 2 = ½ (18.4.2)

For a statistical ensemble of the Schrödinger’s cats the probabilistic predictions of both schemes are
the same. However, what happens to each single animal? Is it either definitely alive or definitely dead

- 32 -
according to (18.4.1)? Or is in a superposed (undecided) state (18.4.2) of being both alive and dead
consistently with the linear evolution law? (see Figure 18).

FIGURE 18. The peculiar situation of the Schrödinger’s cat.

The question does not disappear if the borderline expands still further, including the human observer
and his laboratory as a part of the quantum system. Will the experimentalist enter into a superposed
state of seeing simultaneously the measurement results 1 and 2? If so, why he thinks he had seen only
one of them? Trying to find a consistent answer, a group of authors (v Neumann, London and Bauer,
and Wigner) raised the hypothesis of consciousness as the only classical mechanism: it is, therefore, the
consciousness of an observer which reduces finally the quantum state of the system (London and Bauer
1939; Wigner 1962, 1963; see also Good 1962; Jammer 1974, Ch 11.4). The hypothesis, beyond the
pragmatic approach of the Copenhagen School, was not generally accepted (see e.g., the discussions in
Jammer 1974; Primas 1983; Lalöe 2001; polemic remarks by John Bell 1987, Ch 15).

18.5. The doctrine of decoherence

Many physicists believe that the collapse postulate makes no practical difference. While the linear
evolution (13.2) cannot describe the collapse of the quantum state into one of its components, it can
describe a different phenomenon with similar statistical consequences. It can be most simply illustrated
if the initial state | is the superposition of just two eigenstates of the measured observable, | =
c1|u1 + c2|u2 . Assuming that the eigenstates during the measurement remain physically unchanged, the
state vectors can be still multiplied by phase factors, |uk ei k
|uk , thus converting any superposed
state | = c1|u1 + c2|u2 , into a new superposition:

| ~ c~1 u1 c~2 u 2 , where c~k ck e i k

, k (18.5.1)

If the phases 1, 2, are out of control, the experiment performed on a sequence of the identical states
| will produce a mixture of different superposed states | ~ with the random phase factors in (18.5.1);
the phenomenon known as the decoherence (see e.g. Shimony 1963; Primas 1983, Ch 3.5, p 124-125;
Haroche 1998; Nielsen and Chuang 2000; Lalöe 2001; Hornberger et al 2003; Hackermüller et al
2004). Due to the quadratic character of the statistical averages (10.5.2), the random sequence of the
superposed states with the erratic phases (18.5.1) cannot be distinguished from the sequence of the
“reduced states” |u1 , |u2 ,…, occurring with the frequencies |c1|2, |c2|2. Thus, e.g., if the initial state is
| = |u+ = 1 (|u1 + |u2 ), and if it is transformed into a sequence of the superposed states ~ = |u
2
= 1 (|u1 |u2 ), with the randomly varying signs, then the “unreduced” and “reduced” state
2
sequences |u+ , |u , |u+ , |u ,…, and |u1 , |u2 , |u2 , |u1 ,…, cannot be operationally distinguished; the
fact reflected by their common density matrix:

- 33 -
 = 1 (|u1 u1| + |u2 u2| ) = 1 (|u+ u+| + |u u |) (18.5.2)
2 2

So, an ensemble of superposed cats (18.4.2) with variable superposition phases cannot be
distinguished from an ensemble of the reduced cats (18.4.1), either definitely |ALIVE or definitely
|DEAD . The above equivalence of mixtures accounts for a widely shared opinion that, for all practical
reasons, the only process important in the quantum measurement is the decoherence.

18.6. The ensemble interpretation

Some theoreticians adopt an even more radical ensemble interpretation of QM denying completely that
the quantum state can describe a single physical object. The idea was put forward already by Einstein
in 1936, followed by a group of authors (see, e.g., Blokhintzev 1968; Ballentine 1970). The simplest
argument is precisely the ambivalent representation of the quantum ensemble after the measurement.
Indeed, if the ensemble can be represented as a mixture of two eigenstates |u1 and |u2 , but also as a
mixture of two totally different (unreduced) states |u , it means that each single particle is neither in
one of |u1 , |u2 states nor in |u+ , |u . It is in no state at all (Park 1968; see also discussions in Jammer
1974, Ch. 10.4). The states describe only the statistical ensembles and not the single individuals (c.f.
Ballentine et al 1971). The measurement is just the transition from the pure to a mixed state due to the
decoherence process, and the whole problem of the state reduction disappears.

18.7. The realistic view

The above minimalist doctrine, however, applies if the ensemble survives the measurement but the
measurement information is lost. Yet, if somebody conserved the “detailed data” permitting to identify
the ensemble individuals for which the results were 1, 2, etc, he can repeat the measurement for each
one of them with an absolute certainty that the result will be repeated as well. Accepting the statistical
interpretation (10.5.3) (Dirac 1935), this certainty implies that the micro-system states after the
measurement are indeed |u1 , |u2 and not |u+ , |u (since otherwise the result of the second
measurement could be guessed only with probability ½). Basing on this, some authors believe that the
quantum states |ui (in agreement with the Copenhagen School) are in fact describing the single micro-
systems after the measurement and moreover, that the reduction (collapse) is a real physical
phenomenon (Penrose 1994,1996a,1997; Mermin 1998, 1989; Griffiths 2003).

18.8. Many worlds interpretation of Everett

An intriguing alternative of the reduction axiom (18.4.1) was proposed by Everett in 1957 (see also
Everett 1973), who assumed that the measurement performed on a superposed state of any observable
does not privilege one of the eigenvalues j (j = 1,2,…). All possible results indeed occur, since the
Universe is split into a number of branches. Each different measurement result (eigenvalue j) occurs
just in one branch (copy) of the divided Universe and is seen by the corresponding copy of the
macroscopic observer (which is split as well and looses immediately the contact with his other selfs,
trapped in distinct branches and contemplating different experiment results.) In this way, after each
new experiment our Universe would be split again and over again, forming an infinity of ramifications
in which any possible combinations of virtual events became real. The question of how quickly the
contact between the different branches is lost and whether it occurs as a sudden jump, was not indeed
solved (Greenberger and Yasin 1988,1989; Raimond, Brune and Haroche 1997; see also Lalöe 2001).
A disadvantage of this picturesque interpretation is its highly speculative character: till now there is no
empirical sign that the “parallel Universes” physically exists (Ballentine et al 1971).

- 34 -
18.9. Quantum jumps and quantum Zeno effect

Meanwhile, an empirical evidence of quantum jumps (tentatively interpreted as the collapse events)
appeared in the phenomenology of the 3 level optical systems. In one of possible scenarios, three
energy eigenstates |0 , |1 , |2 are affected by two laser beams. One of them induces the fast transitions
|0 |2 , visible as a strong, steady fluorescence; the other one causes weak transitions (in form of
slow Rabi rotations) between the ground state |0 and meta-stable state |1 . The weak Rabi rotations
cannot suddenly convert |1 into the ground state |0 , the starting point of the luminescent process.
However, they can slowly transform |1 into a superposition | = cos( t)|2 + sin( t)|0 (where is
the Rabi rotation velocity). The observed jumps of luminescence suggest the sudden spontaneous
reductions of | to the ground state |0 , making possible the fluorescent process (Cook and Kimble
1985; Schenzle, De-Voe and Brewer 1986; Nagourney, Sandberg and Dehmelt 1986; Sauter et al 1986;
Bergquist et al 1986; Javanainen 1986; Cook 1988).

A kind of limiting application of the same mechanism was noticed later, though it had a longer
prehistory under the name of the quantum Zeno effect. If the reduction of the wave packets is a real
process, the measuring devices should be applicable not only to extract the numerical data but also to
control the evolution of micro-systems. As predicted by Khalfin in 1958, and by Misra and Sudarshan
in 1977, the sequence of the repeated measurements of any observable should “intimidate” the quantum
system against escaping from the eigenstate of the measured quantity. The phenomenon (Zeno effect),
in principle, should block even the decay of the unstable states. A 3-level analogue of such process was
indeed discovered by Itano et al in 1990. In this case the low frequency waves cause the slow Rabi
rotation between |0 and |1 . The higher frequency wave induces simultaneously stronger transition
between |0 and |2 , observed as the continuous fluorescent light of frequency 02 = (E2 E0)/ . The
direct communication between |1 and |2 is precluded. For the system performing the Rabi rotations
between |0 and |1 , each application of the high frequency (blue) field (Figure 19) is interpreted as a
measurement checking whether the system is in the state |0 or |1 : if it is in |0 , the induced |0 |2
jumps will produce a burst of the fluorescent light, but if it is in |1 , no reaction will be detected (no
transition between |1 and |2 ). The experiments by Itano et al show that if the rest intervals between
the “blue” pulses shrink, the Rabi rotations cease and the system can be blocked in |1 , justifying the
hypothesis about the Zeno effect. However, the discussion is not indeed concluded (an alternative
explanation might be a stabilizing influence of high frequency fields, Ballentine 1991, see also
Pascazio and Namiki 1994). A contrary effect of accelerating the decay by properly separated reduction
acts is considered in (Kofman and Kurizki 2000, Fischer, Gutierrez-Medina and Raizen 2001). Even
though the deeper interpretational problems are still open, the control (manipulation) of micro-systems
by sequences of measurements is one of working laboratory techniques.

Figure 19. The slow Rabi rotations between a pair of state |0 , |1 , blocked by a dense sequence of higher frequency pulses
interpretable as the yes-no measurements applied to the 3-level system (Itano et al 1990).

- 35 -
In an independent investigation track, Roger Penrose supports the thesis that the collapse is a real,
spontaneous jump, catalyzed by gravity (Penrose 1996a). He also believes that it can take place, e.g., in
human neuronal cells (Penrose 1996b; Penrose and Hameroff 1996; Hameroff 1998), in a sense,
endorsing the former ideas about the role of consciousness (von Neumann 1932, London and Bauer
1939, Jammer 1974 Ch. 11.2, 11.3, Wigner 1967). The “psychological model” was questioned by John
Bell (1987, Ch 15) on the philosophical ground, in technical terms by Swiss group (Hepp 1998).
Penrose too admits half ironically that to treat the QM axioms to the letter involves traps (Penrose
1997). In turn, Serge Haroche reports the experiments unrelated to the human consciousness in which
the eigenstates are “robust” whereas the superposed ones are unstable and perform random jumps to the
“eigenstates”, though he acknowledges that the mechanism of each particular jump involves a mystery
(Haroche1998).

18.10. The macroscopic superposition

One of the keys to understand the hypothetical reduction mechanisms might be the observation of the
macroscopic superposition whenever it can exist. The capacity of the mesoscopic and macroscopic
bodies to enter into the superposed states is still an open problem. However, the ability of forming the
interference patterns is already checked for beams of Na, Be atoms and for much heavier fullerens C60
and C70 molecules (Arndt et al 1999, Hackermüller 2003). The superposed position states of heavy
and/or complex micro-systems are sometimes called the Schrödinger’s Kitten (see Gribbin 1984,
1996). The sequence of results showing the existence of the superposed states for ions and for the
Rydberg atoms is described in (Haroche 1998). The macroscopic states of the Josephson system
forming a superconducting qubit are reported in (Leggett 2002). The metamorphosis of the entangled
photons into the macroscopic plasmons and vice versa are described in (Attewischer et al 2002).
However, an increasing difficulty due to the decoherence phenomena is expected as the size of
quantum systems increases (Hepp 2006).

19. ENTANGLED STATES

19.1. General concepts

The symbolic language of QM permits to describe conveniently a coupling between the independent
properties (degrees of freedom) of a micro-system. If A and B are two mutually commuting and
independent families of observables (each A A commutes with every B B), and if A and B generate
all observables, then the system can be described in terms of “partial states” (partial wave functions)
|a , |b in two independent Hilbert Spaces HA and HB defining the probability distributions for the
observables A A and B B. The complete system states then form the tensor product space HA HB ,
in which the basis is defined by the ordered pairs (formal products) |a |b (denoted also |a |b ), while
the general states | are their (formal) linear combinations:

| = ij |ai |bj = ij |ai |bj (19.1.1)

ij ij

called entangled states and describing the special links between the “partial observables” which can
occur for the concrete states (19.1.1).

- 36 -
19.2. Spin-configuration entanglement

For the non-relativistic particle possessing the Pauli’s spin, the QM wave function is constructed as a
functional pair (x,t) = ( +(x,t), (x,t)), i.e., the entangled spin-coordinate state = | + |+ + | |
(Dirac 1935, Ch VI). For the (non-relativistic) Hamiltonian of the spin-1/2 particle H = H0 + B,
where H0 depends only on the configuration variables and B is a magnetic field, the energy eigenstates
are the simple tensorial products | =| | = | | of the position | and spin | states, where
| is the eigenstate of the initial Hamiltonian H0 and |+ or | are the eigenstates of B contributing
with the corrections B to the spinless energy levels.
2mc

The most picturesque experiments in which the spin is entangled with the position were performed by
Gerlach and Stern in 1921-22 (without knowing still about the existence of the spin! The result at that
time was interpreted as the proof of a mysterious, “classically unexplainable bivaluedness”). Today we
know that it has shown the existence of a tiny s = ½ spin of the silver atoms entangled with the
“massive” coordinate packets. Thanks to the spin-coordinate entanglement, the particle spin could be
indirectly checked by measuring the position and reducing the coordinate packet to either | + or |
(see Figure 20). An analogy of the silver nucleus and the massive Schrödinger’s cat entangled with its
own tiny spin can be noticed.

FIGURE 20. The spin measurement due to the spin and position entanglement: the particles localized in the packets |
have the spin states reduced to | respectively (following Cohen-Tannoudji, Diu and Lalöe 1977 Ch. IV).

19.3. Many particle entangled states

The many particle configurations in QM are the most important example of the entangled states. For a
quantum system composed of n particles with pure states represented by vectors |a , |b , |c ,…, in
Hilbert spaces Ha, Hb, Hc, …, the basic states are the ordered sequences (simple products) |a |b |c
denoted also |a |b |c (meaning that the first particle is in state |a , the second one in |b , etc.)
The general pure states are the arbitrary linear combinations of the simple products, in the multiple
tensor product spaces Ha Hb Hc , defining the multi-particle correlations. Thus, e.g., for the two
particle system, the entangled state | = c1|a1 |b1 + c2|a2 |b2 , where |c1|2 + |c2|2 = 1, tells that if the
state of the first particle is |a1 then simultaneously the state of the second one must be |b1 , but if the
first one is |a2 then the second one is |b2 ; the probabilities of both options are |c1|2 and |c2|2
respectively; the phase factor (c1|c2|/c2|c1|) defining the tentative output of the interference experiments.

For the indistinguishable particles, the space of states is the tensor product of the single Hilbert space
by itself: H H H , and the states can be classified according to their behaviour under the particle
permutation. If P is the n-particle permutation group, then the totally symmetric state can be obtained
by summing up all permuted versions of any simple product:

- 37 -
| = 1 P(|a |a2 |a3 …|an ) (19.3.1)
sym 1
n!
P
whereas the totally antisymmetric state is an analogous sum with the coefficients ( 1)P = 1 for the
even and odd permutation respectively (see e.g. Gasiorowicz 1974):

| = 1 ( 1)P P(|a1 |a2 |a3 …|an ) (19.3.2)

anti
n!
P

According to the spin-statistics theorems (Corson 1953), the symmetric case (19.3.1) describes the
particles with an even spin (photons, mesons, some nucleons, etc), called generally bosons, from the
name of S.N. Bose who discovered their equilibrium density known later as Bose-Einstein distribution:

1
n , (19.3.3)
E
e 1

Here n is the number of particles in a quantum state of energy E, is a normalization constant

depending on the total particle number, =1/kT where k is the Boltzman’s constant and T the absolute
(Kelvin) temperature (see e.g. Merzbacher 1998 Ch. 22). The formula (19.3.3) takes into account that
many bosons can occupy the same quantum state. The dense clouds of bosons, known as bosonic
condensations, propagate according to the non-linear Schrödinger equation (see e.g. Alexandrov 1993).

The antisymmetric states (19.3.2), characterize the odd spin particles (electrons, nucleons, etc) known
as fermions and obeying the Pauli’s exclusion principle, automatically granted by the antisymmetric
structure of many fermion states (19.3.2). The Pauli’s principle is crucial to explain the energy shells of
many electron atoms (Schiff 1968); including the classification of the Mendeleev table. The clouds of
fermions obey the Fermi-Dirac statistical distribution:

1
n , (19.3.4)
E
e 1

The populations of electrons near 0 K, obeying the limiting case of (19.3.4), are known as the Fermi
gas. Its perturbations are important for the propagation of the Cooper pairs, explaining the
superconductivity phenomena (c.f. Bardeen, Cooper, and Schrieffer 1957a-b; see also Schrieffer 1964).

20. DIRAC’S THEORY OF THE ELECTRON AS THE SQUARE ROOT OF THE KLEIN-
GORDON LAW

20.1. The Klein-Gordon equation

The problem of the relativistic QM proved highly non-trivial. The simplest relativistic generalization of
(8.1) was the partial differential equation corresponding to the relativistic energy formula
E2 p2 m02 c 2 :
c2

2
1 m 2c 2
0 (20.1.1)
c2 t 2 2

- 38 -
Designed in 1926 by several physicists, including Schrödinger, de Broglie, Klein, and Fock, today
known as the Klein-Gordon equation due to the independent works by Oskar Klein and Walter Gordon
in 1927, the ‘many fathers’ equation (20.1.1) enjoyed the required relativistic invariance (for Klein
(1926) it was a part of the five-dimensional theory unifying the general relativity and electrodynamics).
However, to grant the correspondence with the Schrödinger equation (8.2), should be interpreted as
the wave function of a relativistic particle. This requires a 4-component, conservative probability
current J , with: J = 0. However, for the standard J = * * , the time
0 *
component J * t t is not positive-definite, and therefore cannot be a probability density.
The situation led some physicist including Pauli and Dirac to discard (20.1.1) and to look for some
alternatives.

20.2. Dirac’s equation

The idea of the spinning electron was accepted since 1926 (cf. Section 10.10). However, “the question
remained as to why Nature should have chosen this particular model for the electron instead of being
satisfied with the point charge” (Moyer 1981a-c). Looking for a new law, Dirac applied a symbolic
method, based on natural demands: (I) The equation should be linear in / t, so that the wave function
at any time determines the wave function at any later time (II) It should correspond to the classical
theory for c (III) It should be invariant under Lorentz transformation (IV) It should predict
observed quantum phenomena and (V) It should not predict unobserved ones (indeed, the initial hope
of Dirac was to replace (20.1.1) by some new equation with the positive probability density and
positive energy). In an act of almost magic inspiration, he postulated the new wave equation:

e
i H ; H c A mc 2 e (20.2.1)
t i c

where the k (k = 1,2,3) and were 4 operators acting on the hypothetical internal degrees of freedom
of the relativistic electron, e and m its charge and rest mass, A and the vector and scalar potentials of
the surrounding electromagnetic field. To deduce the Klein-Gordon eq. (20.1.1), Dirac had to assume
the anticommutation relations: k l + l k = 0 (k l), k + k = 0, as well as 12 = 22 = 32 = 2 = 1.
The demands could be satisfied by the 4 component wave functions = { A(x,t)}A=0,1,2,3 and a proper
choice of 4 matrices k, . Their simplest form, making evident the correspondence to the electron
of Pauli (Sec. 10.10) was:

k 1
k , (20.2.2)
k 1

The manifestly relativistic equivalent of (20.2.1-2) reads:

e
A imc 0 (20.2.3)
i c

where A is the relativistic 4-potential of the electromagnetic field, the repeated indices
mean the Einstein’s summation convention (see e.g., Corson 1953); ( = 0,1,2,3) are 4 4

- 39 -
matrices satisfying the identities 1 ( + ) = g , where g = diag( 1,1,1,1) is the metric tensor
2
of the special relativity. For A = 0 (20.2.3) permitted to deduce immediately the Klein-Gordon
equation (20.1.1), granting also the invariant character (frame independence) of the new law.

The transition between (20.2.1) and its manifestly relativistic form (20.2.3) is obtained e.g. by putting:
k =
0 k
and = i 0. By taking A = 0, = e and solving the eigenvalue problem for (20.2.1) one
r
obtains a modified sequence of the energy eigenvalues for the hydrogen bound states (see Dirac 1935,
Ch. 72, Schiff 1968, de la Peña 1991):
1
2
2
2
En,j = mc 1 2
(20.2.4)
n j2 2

where = e2/ c and j is an eigenvalue of the total angular momentum J.

20.3. The hypothesis of positron

The equations (20.2.1) and the result (20.2.4) passed several important tests: The most impressive
success of Dirac was that the spinning electron is a natural consequence of the symbolic formalism
(20.2.1-4). The exact magnetic moment of the electron and the electron bound states for the hydrogen
atom (including the fine structure constant) were derived with no additional assumptions (Dirac
1928a); the Zeemman splitting and the selection rules were also obtained (Dirac 1928b-c). However,
the relativistic equation (20.2.1-3) did not fulfill the initial expectation (V) since it predicted the
negative kinetic energies, a disturbing conclusion which seemed totally unacceptable. The forthcoming
avalanche of objections, by N. Bohr, G. Breit, C.G. Darwin, and N.F. Mott, and others gave impression
that the Dirac theory was “just an approximation”. However, the criticism stimulated only the next step
of the theory.

Already in October of 1929, in a letter addressed to Bohr, Dirac suggested that the majority (indeed, an
infinity) of electrons are in the negative energy, high velocity states (c.f. Moyer, 1981b, p 1057). The
distribution of these electrons should be uniform so that they will not produce any electromagnetic
field, forming a practically unobservable “sea”. However, if not all states of negative energy are
occupied, there might be some “holes” moving in the electromagnetic field as though they had positive
charges. Dirac believed that these holes were protons. When an electron of positive energy drops into a
hole, the electron and proton disappear and emit radiation. Bohr answered raising two objections: (1)
the effect of infinite electric density would be a problem, (2) the fatal transition from positive to
negative energies seemed to be a limitation in the applicability of the energy concept rather than an
indication as to the nature of protons. Nevertheless, Dirac published his hole interpretation supported
subsequently by Weyl.

By March 1930, a new problem was noticed: Robert Oppenheimer calculated the mean lifetime for the
Dirac’s holes. If they were protons this lifetime should be about 10-10 sec, inconsistently with the
stability of the matter. Soon, Bohr returned to the battle by criticizing the Dirac’s hole interpretation
and insisting on the limitations of quantum mechanics. The discussions continued between Bohr,
Gamow, Fermi, Pauli, Landau and Peierls, and others.

- 40 -
The turning point was the new 1931 hypothesis by Dirac. He recognized that the hole “necessarily has
the same mass as the electron” and modified its interpretation: “A hole, if there were one, would be a
new kind of particle, unknown to experimental physics, having the same mass and opposite charge to
an electron. We may call such a particle an antielectron”. Then, he added “Presumably the protons will
have their own negative-energy states, all of which normally are occupied, an unoccupied one
appearing as an antiproton.” (In the same paper, Dirac predicted the possible existence of a magnetic
monopole.) The ideas were not easily accepted. In the 1932 Volta Conference in Rome, Bohr repeated
the capital argument that the Dirac’s theory failed because it predicted absurd negative energies,
contradicting the correspondence principle. He strongly believed that this failure indicated the
‘breakdown of quantum mechanics’ at nuclear dimensions.

The discussions were interrupted by the letter of C.D. Anderson, published in Science on 9 September
1932, announcing the observation of “a positively charged particle having a mass comparable with that
of an electron”. Dirac obtained the Nobel prize in 1933 “for the discovery of new productive forms of
atomic theory”. Anderson was rewarded with the 1936 Nobel prize for “his discovery of the positron”.
The negative proton (antiproton) was detected experimentally in 1955 by E. Segré and O. Chamberlain.
In the subsequently developed particle physics the existence of antiparticles is a typical phenomenon.

21. FEYNMAN: THE INTERFERENCE OF VIRTUAL HISTORIES

As soon discovered, the search for the new wave equations is not the only tool of quantum theory. The
fact can be most easily illustrated for the traditional Schrödinger’s QM. An important aspect of the
unitary evolution (13.2,13.4) is that the free evolution H0 = p2/2m and the potential part V(x) of H do
not commute and so, do not contribute to the motion of the particle in a classical way. To examine the
phenomenon, Richard Feynman (1948) used the simple algorithm (mathematically elaborated by
Trotter) granting that the exponential function (13.4) of the operator sum H0 + V(x) can be
approximated by splitting the time interval [t0, t] into a great number of sufficiently small subintervals
t0 < t1 < t2 < < tn-1 < tn = t, where tj = tj – tj-1 0, and then by substituting the little evolution step
i t (H V) i t jH0 i t jV
e j 0
in each of them by a pair of two more elementary steps e e . He obtained
an operational sequence:

U(t,t0) e i t n H 0 e i t nV e i t1 H 0 e i t1V (21.1)

i t jH0
in which every free evolution step e affects the uncertainty of position and every “potential
i t jV
shock” e affects the momentum uncertainty. The final result brings a new insight into the
mechanism of quantum propagation. The value of the wave function (x, t) at the moment t = tn is the
sum (integral) of the exponential contributions from the sequence of all uncertain particle positions in
all previous time moments. Each single contribution has the form of a numerical phase factor:

e iS ( x 0 , x1 ,..., x n ) / (21.2)

where the function S(x0, x1, …,xn) is the classical action integral along one of the virtual trajectories
formed by n intermediate points x0 x1 xn 1 x n x randomly selected in the time
moments t1 t 2 t n t . An essential property of the virtual trajectories is that they do not need
to obey the classical motion equation and so, the action S(x0, x1, …,xn) in (21.2) is not limited to take
extreme or stationary values ( S 0). The unitary evolution law (13.4) acquires then a simple sense.

- 41 -
The values of the function (x0, t0) “propagate” along all possible virtual trajectories and arrive at the
point x in the moment t with additional phases defined by (21.2); the new (x, t) is created by the
interference (sum) of all contributions. The result:

(x, t) = N 1 (x0, t0) exp[- iS(x0, x1, …, xn)/ ] dx0 dx1 … dxn, (21.3)

where N is a normalization constant, is called the Feynman’s path integral (Figure 21). The limiting
form of this law, universally applicable to all areas of quantum physics (from the double slit
experiment, trough the unitary QM evolution (13.4), the graphs of the quantum field theory up to
quantum gravity and cosmology) is the Feynman quantization rule. It tells that the behaviour of any
physical system for 0 can be described by considering its classical analogue evolving
simultaneously along all possible virtual histories (some obeying and some disobeying the classical
equations). The system states in the future emerge from the interference of all virtual histories with
phase factors defined by the action integrals (cf. Feynman 1948; Feynman, Leighton, and Sands 1963).
The capacity of the quantum systems to run simultaneously over an infinity of alternative trajectories
was one of the principal inspirations for quantum computing.

FIGURE 21. The idea of path integration (21.3). The system state in the future (t > t0) is the result of the wave-like
propagation of the quantum state along an infinity of virtual trajectories. Blue: some trajectories bringing the exponential
contributions (21.2) to (x1, t). Green: some trajectories contributing to (x2, t).

22. THE LOCALITY PROBLEMS

22.1. The EPR paradox

Despite the success of Dirac’s heuristics, the coexistence between the quantum and relativistic theories
was never easy. Apart of Erwin Schrödinger, in opposition against the Copenhagen School was Albert
Einstein, who apparently did not accept all consequences of his 1905 discovery. “God does not play
dice” was his favorite argument (In his letter to Born, 4 December 1926, Einstein wrote “Quantum
Mechanics is certainly imposed. But an inner voice tells me that it is not yet the real thing… I, at any
rate, am convinced that He is not playing dice”, Born 1971, p 91). An apparently elementary two
particle propagation in the Minkowski space-time led Einstein, Podolsky and Rosen (EPR) to serious
doubts about the soundness of the entire QM interpretation. Their original paper exhibiting the peculiar
aspects of the entangled states (latter known as the EPR paradox) was published in 1935, though its
most challenging form (discussed below) was presented in 1957 by Bohm and Aharonov.

A pair of photons emitted from a common source (typically, by a down conversion mechanism, see
e.g., Greenberger et.al. 1990,1995; Ribeiro, Monken and Barbosa 1994; Banaszek, U'Ren and
Walmsley 2001) in two opposite directions (called “left” and “right”) with the total linear momentum

- 42 -
and polarization vanishing, forms an entangled, antisymmetric polarization state | . Its notable
property is the invariance under the simultaneous rotation of both (left and right) polarization qubits,
leading to an infinity of equivalent expressions:

| = 1 (| | | | ) = 1 (| | | | )=… (22.1.1)
2 2

encoding the correlations beyond the classical models. Indeed, suppose, both emitted photons arrive at
the detectors of two distant observers, Alice (left) and Bob (right). Bob can check a variety of the
polarization states of his photon in a variety of directions. He can use the polarization filter (Nicol
prism) oriented to distinguish e.g. the polarization states | and | : If his photon shows the
polarization | , then the entangled state (22.1.1) is reduced to | | , and the polarization state of the
Alice photon must be | ; but if the polarization of the Bob’s photon turns | , then the photon of
Alice must show the polarization | . Until now, there is no surprise from the purely classical point of
view. However, Bob can also rotate the axis of his analyzer to distinguish between the polarization
states | and | . Then, the reduction mechanism of | applies to the 2nd part of (22.1.1): if he finds
the polarization | then the Alice’s photon will be in the opposite state | , and vice versa. Hence, if
Bob keeps his analyzer in the first position, Alice must receive either | or | states, but if he shifts to
the second one, the random states received by Alice will be | and | . (An analogous behaviour is
shared by a pair of the spin-1/2 particles, e.g., electrons, emitted from a common source in two opposite
directions with a total linear momentum and total spin vanishing). The above possibility of
manipulating at distance is not limited to the spin and polarization states. Detected already in their
original 1935 paper, it awoke the criticism of Einstein, Podolsky and Rosen, who reproached the
unphysical character of quantum axioms.

FIGURE 22. The sequence of photon pairs in the singlet state (22.2.1) is sent to a pair of observers, Alice and Bob. If Bob
keeps the axis of his polarization detector vertical ( ), Alice receives a mixture of | and | photons, but when Bob turns
his detector axis to ( ), Alice must receive the sequence (mixture) of | and | . In spite of the non-local state
manipulation, Alice cannot tell the difference, since in the orthodox QM both mixtures are undistinguishable.

The answer of Bohr (1935) to the original 1935 EPR paper pointed out an intrinsic consistency of the
Copenhagen interpretation. It can be easily adapted to the 1957 version of Bohm and Aharonov. By
performing just one measurement of the spin or polarization of the particle one knows only the spin
after the measurement, but one cannot infer what was the spin state before. Hence, in order to know
whether Alice can observe Bob’s manipulations, the single EPR pair is insufficient. Alice and Bob
must receive a sequence (ensemble) of photon pairs, produced in the entangled state (22.1.1). If now

- 43 -
Bob maintains for some time the “vertical” position of his analyzer, Alice receives the equitable
mixture of the polarization states | and | , but if he turns his analyzer by /4, Alice will receive the
similar mixture of | and | . One might think that by using her polarization filters Alice can therefore
distinguish both mixtures, and decode Bob’s signals. However, it is not the case: due to the character of
QM observables, their average values (10.5.2) for both mixtures are exactly the same (compare
(18.5.2)). Henceforth, the output of the EPR experiment does not contradict the relativistic causality:
even if the state vectors can be manipulated at distance, the message remains unreadable.

22.2. Hidden variables and Bell’s inequalities

While consistent with the Copenhagen doctrine, the EPR correlations frustrated the intents of classical
interpretation. An intriguing question was, whether the absence of the classical state variables (“hidden
parameters”) can be proved on the ground of purely numerical results?

To check this, John Bell designed a tentative classical model for the EPR experiment in its spin
version. A pair of identical, classical particles is shot in two opposite directions, towards two
macroscopic observers A and B (Alice and Bob). The hypothetical, undetectable micro-states of both
particles (“hidden parameters”) form a certain classical phase space . According to Bell’s conjecture,
they define uniquely the results of any spin (or polarization) measurement applied to each one of the
escaping particles. The consequences are most easily described by using the spin observables A = n ,
B = m , C = l , … (the double spin components in the directions n, m, l, where |n| = |m| = |l| = 1). If
the classical model is true, to all these observables should correspond the subsets a, b, c,… of the
microstates for which the measurement of A, B, C, …, on the right particle yields the result +1, while
their complements a , b , c , …, gather all hidden variables for which the same measurement gives the
opposite result 1 (inversely for the left particle). The hypothetical subsets a, a , b, b , c, c , provide the
classical Boolean logic (Figure 16.I) for the spin experiments. Since the experiment is statistical there
should also exist a probability measure on the subsets of , ( ) = 1, defining the participation of the
“hidden variables” in the acts of emission. Due to the elementary EPR statistics, (a) = (b) = (c) =
(a ) = (b ) = (c ) = ½ (the probabilities of the values of any spin component for the left or right
EPR particle are ½). Now, permits to evaluate the overlaps a b, defining the classical equivalent of
the QM conditional probabilities (also transition probabilities) for a and b: p(a,b) = (a b) (a) =
2 (a b), so that the existence of the “hidden parameters” converts the spin states into a Boolean
probability space. It yields also the correlations C(a,b) defined as the statistical averages for the
products of the A and B values measured independently by Alice and Bob on both sides of the EPR
arrangement. In the Boolean model they would have to be: C(a,b) = (a b ) + (a b) (a b)
(a b ) = 1 2p(a,b).

However, the Boolean model has a natural geometry of a metrical space where the “distance” d(a,b)
for any pair of subsets a,b is given by the measure of their symmetric difference: d(a,b) = (a) +
(b) 2 (a b) = 1 p(a,b) = ½[1 + C(a,b)]. The triangle inequalities (i) d(a,c) d(a,b) + d(b,c) and
(ii) |d(a,b) d(a,c)| d(b,c), henceforth imply:

|C(a,b) C(a,c)| 1 + C(b,c) (22.2.1)

which is precisely the original form of the Bell’s inequality (Bell 1993). The iterated triangle
inequalities lead to the next generations of Bell’s theorem. Thus, e.g.: d(a,d) d(a,b) + d(b,c) + d(c,d)
imply the best known:

|C(a,c) + C(a,d) + C(b,d) C(b,c)| 2. (22.2.2)

- 44 -
The Bell’s inequalities (22.2.1-2) are the necessary condition for the classical mechanism behind the
QM correlations in the EPR experiments. However, they are not fulfilled by the experimentally verified
correlation functions. The experimental breaking of (22.2.1) and (22.2.2) is generally considered one of
the most serious arguments against the existence of hidden variables in QM. The less radical
conclusions expressed e.g. in the collection of Greenberger and Zeilinger (1995) are that breaking of
Bell’s inequalities strictly speaking implies only the non-locality but not the causality breaking
(Though, on the other hand, in a classical, but non-local theory the causality violation could be hardly
avoided).

Historical note: The inequalities (22.2.1-2) were found by John Bell (1966) by applying a different
argument. Their equivalence to the triangle inequality was shown by Daniel Fivel (1991), thus
revealing a new bridge between the abstract mathematics and concrete physical problems.

22.3. “Seeing in the dark”

The EPR correlations for some time seemed the principal locality challenge in QM. More recently,
though, an equally surprising experiment was designed by Elitzur and Vaidman (1991, 1993). The
photon wave function (forming a narrow light beam) is divided by a system of mirrors and beam
splitters (Mach-Zehnder interferometer) into two coherent branches (beams) propagating along two
widely separated space trajectories (Figure 23). Then, both branches join and interfere, reconstituting a
single beam parallel to the initial one. If, however, one of the trajectories is cut by an absorbing
obstacle (interpreted as the measuring apparatus) the photon state is reduced. The photon is either
absorbed in the obstructed branch or (more interesting) it collapses to the free branch, leaving the
obstructed branch empty. If reduced to the free branch, the photon has now the ½ probability to be
registered by the detector D0 to which it had no access previously. Should this happen, the detection by
D0 will reveal the existence of an obstacle which the photon never touched. In the original paper of
Elitzur and Vaidman the obstacle is a bomb which must explode if absorbing a photon. The bomb has
therefore the ½ probability to explode but if it doesn’t, the photon has still the probability ½ to detect
the bomb which was never illuminated. The bomb can be “seen in the dark” by the photon which could
pass light years away (see also Kwiat, Weinfurter and Zeilinger 1996). The effect indeed unifies the
traditional double slit experiment and the delayed choice experiment proposed by Wheeler (see
Wheeler and Zurek 1983).

FIGURE 23. The interaction free experiment of Elitzur and Vaidman. (a) The photon split into two propagation branches,
which subsequently unify, must fall into the detector D; (b) If one of the branches is blocked by an absorbing obstacle (e.g.
a bomb), then either the photon will be absorbed (the free branch reduced, bomb explodes), or the blocked branch will be
reduced (collapsed to non-existence). In this last case, the photon will have the ½ probability to fall into the detector D0,
thus revealing the existence of an obstacle with which it had no physical contact.

- 45 -
22.4. Teleportation

The EPR paradox and the state collapse by the “interaction free experiments” brought back the
question about the meaning of quantum states and the state reduction in QM. According to some
opinions, the impossibility of reading the EPR messages might mean that there is a certain redundancy
in the very concept of the entangled states: they are just our auxiliary mental constructions, and
therefore, there is no harm if they suffer the instantaneous “alterations at distance”, as long as this does
not lead to a readable, non-local communication.

FIGURE 24. INNSBRUCK EXPERIMENT begins with a short pulse of ultraviolet laser light. Traveling left to right
through a crystal, this pulse produces the entangled pair of photons A and B, which travel to Alice and Bob. Reflected back
trough the crystal, the pulse creates two more photons, C and D. A polarizer prepares photon D in a specific state, X. Photon
C is detected, confirming that photon X has been sent to Alice. Alice combines photons A and X with a beam splitter. If she
detects one photon in each detector (as occurs at most 25 percent of the time), she notifies Bob, who uses a polarizing beam
splitter to verify that his photon has acquired X’s polarization, thus demonstrating teleportation (taken from Zeilinger 2000,
as a courtesy of the author and the Scientific American Journal)

Yet, if the argument of Section 17.7 is accepted, the state correlations in the EPR experiment acquire a
realistic sense: they mean that the measurement performed by Bob not only reduces the state of Bob’s
(right) photon but the effect is instantaneously transmitted (out of the light cone), and defines the state
of the Alice’s (left) photon. Somehow, the polarization state eliminated by Bob is teleported to the
Alice’s side. Assuming that the exact EPR correlations won’t be disproved (which seems less and less
probable; c.f. Tittel et.al. 1998), the teleportation acquires an almost realistic sense. Alice still cannot
check it without knowing which polarization direction was measured by Bob and what was the result of
his measurement. Hence, the causality is not violated. Yet, once she receive the missing information
(which can be send by conventional channels) she will be able to confirm the fact a posteriori, guessing
with 100% probability her photon state. The absolute certainty of her guess makes the teleportation
model a legitimate working hypothesis. At the moment, the effect is verified with an increasing
precision by several research groups (Huttner et.al. 1996; Zeilinger 2000; N. Gisin et.al. 2002). A more
sophisticated method of teleporting the unknown quantum state by entangling it with the EPR pair in
the singlet state (22.1.1) was presented by the American group (Bennet, Brassard and Crépeau 1993).
One of the EPR observers (e.g. Bob) while waiting for his EPR particle, keeps also the third particle in
an unknown polarization state . When the EPR particle arrives at his laboratory (it serves as the
“ancilla”), he entangles it with an unknown by performing an elementary two particle measurement.
Depending on its result, on the opposite (Alice) side 4 possible transformed versions of the teleported

- 46 -
state can appear. Though is unknown, the transformations are simple and once the results of
Bob’s measurements are transmitted to Alice by normal channels, she is able to invert them for every
arriving particle, thus converting the “teleported debris” into a pure ensemble in the (still unknown)
state . As recently shown, the teleportation effects occur also in chemical molecules (Nielsen, Knill,
and Laflamme 1998). In this case, the quantum teleportation can be experimentally implemented over
inter-atomic distances (a few Angstroms) by using liquid-state nuclear magnetic resonance (Rabi
rotations, c.f. Ch 16) and exploiting the natural phase decoherence of the carbon nuclei in a
trichloroethylene molecule (see Figure 25). The state of one of these carbon nuclei is teleported to the
hydrogen nucleus (see also Ferguson et al 2002 and Havel et al 2002)

FIGURE 25. Trichloroethylene (TCE) molecule. Nielsen, Knill and Laflamme (1998) referred to the left C nucleus as the
data, the right C as the ancilla, and the hydrogen nucleus H as the target qubit. The circuit teleports the state of the data to
the target qubit. According with Nielsen et. al., the phase decoherence times (T2) for the C are approximately 0.4s and 0.3s,
and ~3s for H. The relaxation times (T1) of approximately 20-30s for the carbons, and 5s for the hydrogen (all other T2 and
T1 times for all three nuclei are much longer).

23. THE IDEA OF QUANTUM COMPUTING AND FUTURE PERSPECTIVES

23.1. Quantum cryptography

The ideas of quantum cryptography were historically earlier than the concept of teleportation, but only
after this last one was conceived, they exhibited their operatorial sense. In the measurements of EPR
type one of the entangled state components affects the other ones in such a way that the effects are not
directly observable. They may encode a message which propagates instantly (out of the causal cone),
but is hidden in the statistical noise. The message can be decoded only with the help of another
(macroscopic) information (“the key”) which propagates causally (inside of the light cones). Who has
no key cannot read the message which is protected by the fundamental laws of QM. The simplest case
is again illustrated by the Bohm and Aharonov version of the EPR experiment. Alice and Bob receive a
sequence of photon pairs in the antisymmetric, entangled polarization state | . They have a previous
agreement defining consistently the rotation angles of their polarization filters. Now, Bob applies his
polarization filter (yes-no measurement) giving the answers 1 and 0 if the photon polarization is along
or is perpendicular to the axis of his filter (fixed at will in each single measurement). Once the
sequence of Bob’s experiment is performed, he sends to Alice a conventional, unprotected message by
radio, phone or even by a postcard. He does not tell her what was the sequence of his measurement
results for each of the received photons. He just tell how should she keep her analyzer to obtain the
correct 1 or 0 answers from her photons. The existence of an eavesdropper (Eve) who might hijack
some photons will be easily discovered by interrupting Bob’s message by some short test sequences.
An ample family of more sophisticated coding methods is already elaborated in (Bennett 1992; Bennett
et al 1992; Bennett, Brassard, and Ekert 1992; Bennett, Brassard, and Mermin 1992; Huttner et al 1995;
Bouwmeester, Ekert, and Zeilinger 2001; Scully and Zubairy 2002, Ch. 18.6; )

- 47 -
23.2. Quantum computing

In the classical theory, any information could be reduced to simple sequences of 0 and 1 numbers
(words), e.g., 0000…0, 0100…0, 0010…0, 1100…0, 1010…0, which were then elaborated step by step
by the classical computers, changing at every step just a single bit in the processed sequence, according
to the principles of classical logic (e.g. Turing machine, see Penrose 1989; Nielsen and Chuang 2000).
An idea arises that a more efficient information processing could be obtained by using the rich
informational structure of quantum states, with the sequences of “quantum bits” (qubits) substituting
the classical digits (see e.g. Mermin 2003). As the simplest material model one might consider a system
of n spin-1/2 particles endowed with the magnetic moments and kept in a strong external magnetic
field. The numbers 0 and 1 could then mean the “spin down” | and “spin up” | states of each qubit
(which, however, are only two points on the qubit sphere; c.f. Figure 17). The qubits could be
heterogeneous (with different magnetic moments) to assure their separate controllability and could be
selectively manipulated by Rabi rotations induced by adequate magnetic field frequencies. The
quantum equivalents of the classical “words” would be the simple multi-particle states: | ,
| , | ,… An important novelty, though would be that now the words could form a
variety of the entangled (superposed) states:

|W = 0000 0| + 1000 0 | + 0100 0| + (23.2.1)

When introduced to the computing algorithm, each word (23.2.1) evolves along a whole variety of
virtual histories, every entangled component checking separately one of the computation tracks. An
idea thus arise that the computation algorithm could become equivalent to an infinity of the
simultaneous classical processes programmed so that at the very end the false options cancel due to the
destructive interference. The most elementary example of such process were described in (Benioff
1982; Feynman 1982; Deutsch 1985; Shor 1994; Barrenco et al 1995; Deutsch, Barrenco, and Ekert
1995; Cerf, Adami, and Kwiat 1998; Leonard 1998; Chuang et al 1998; Brennen et al 1999; Ferguson
et al 2002, Havel et al 2002)

The basic technical challenge, though, consists in the selective quantum control which should create the
arbitrary entangled states (23.2.1). At the moment, the microsystems for which this is possible contain
at most several qubits, and the time needed to create an efficient quantum computer cannot be easily
estimated (Haroche and Raimond 1996, but see also Rauschenbeutel et al 2000).

24. OPEN QUESTIONS

The QM laws explained ample domains of physical phenomena at the cost of leaving some basic
problems unsolved. In the subsequently developed particle physics, the QM formalism of linearly
added complex amplitudes and operator observables remains unchanged, though implemented at the
new levels of the theory. Translated into the graphs of quantum fields it acquires almost the
universality of the modern Ptolemean scheme. The new experimental results are obtained by applying
the increasingly powerful particle accelerators, expanding the energy range of the collision
experiments. However, the progress in this direction might be limited for technical and/or economic
reasons. It is therefore comforting that the fundamental ideas of QM are far from exhausted and seem
to show some new alternatives.

While the QM can be interpreted as a deformation of classical theory, the deformations of the quantum
scheme are also studied. The best known cases are the non-commuting coordinates [xk, xm] = km (Majid
1994), q-deformations (Klimyk and Schmudgen 1997), non-commutative geometry (Connes 1994), as

- 48 -
well as the fuzzy spaces (Madore 1995), some of them conceived with the hope of the space-time
quantization. An independent vision of the quantum space-time operators is the loop quantization
(Ashtekar and Geroch 1974; Ashtekar 1986, 1987; Rovelli 2004). The combinatorial topology, general
relativity and quantum oscillators meet in an ample, multidimensional scenario of the string theory (see
e.g. Witten 1995; Gubser, Klebanov and Polyakov 1998; Seiberg and Witten 1999). All these trends still
belong to the linear, Hilbert space paradigm, with the operator observables. The possibility of more
general statistical theories, where the Hilbert space formalism is not assumed, and the quantum logic
could be deformed, was also considered (Ludwig 1964, Mielnik 1974, Bell and Hallet 1982). They are
interrelated to the hypothetical nonlinear generalizations of QM, based on nonlinear wave equations
(Bya ynicki-Birula and Mycielski 1976; Haag and Bannier 1978; Shimony 1979; Kibble and Randjbar-
Daemi 1980; Weinberg 1989; Goldin 1992; Doebner and Goldin 1992; Goldin and Svetlichny 1994;
Dodonov and Mizrahi 1995; Mosna, Hamilton and Delle-Site 2005), where the superposition principle
is broken and the statistical interpretation leads to non-quadratic average values. A notable outsider in
this family is the nonlinear gravitation of Roger Penrose, a hypothetical entity which admits a nice
analytic description but exist “in itself”, without external measuring devices, and henceforth, still waits
for a statistical interpretation (Penrose 2004). Yet, until now, intents of formulating the nonlinear QM
are blocked by the mechanism which produces the superluminal messages if the pair of the EPR
particles forms the entangled state (22.1.1) and if at least one of them obeys the nonlinear QM (Gisin
1990, Gisin and Rigo 1995).

The older interpretational challenges too, remain alive. Will the future QM see the revival of
fundamental discussions? Should the quantum states be interpreted in a minimalist way, just as the
code symbols used to predict the statistical effects (c.f. the event enhanced quantum theory, Haag 1996)
or will they recover some realistic sense? Will the paradox of the Schrödinger’s cat be solved by a new
generation of QM laws? Should they remain linear forever, with the “collapse of the wave packet”
meaning simply that our probabilistic predictions must be from time to time reset? The technical
problems are no less intriguing. Can one achieve a precise dynamical control of many qubit (many
level) systems? How long might it take before we shall have a credible quantum computer? No answers
as yet exist. This report belongs to a turbulent epoch of an early prehistory when the solutions were still
hardly emerging. What it might show is that the unsolved problems, even if they look secondary, can
be more important than the established scientific theories and their results.

- 49 -
50
Bibliography

Abbe E. (1873) Beiträge zur theorie des mikroskops und der mikroskopischen wahrnehmung, Arch.
F. Microsc. Anat. 9, 413-468 [The historical paper explaining essential optical concepts for the
resolution of microscopic images. Used subsequently to illustrate the quantum uncertainty principle
in the "Heisenberg's microscope"]

Accardi L. (1981) Topics in quantum probability, Phys. Rep. 77, 169-192 [An outstanding review of
the hidden parameter theory. For readers interested in the subject all works of Accardi are relevant]

Arndt M., Nairz O., vos-Andrae J., Séller C., van der Zouw G., and Zeillinger A. (1999) Wave particle
duality of C60 molecules, Nature 401, 680-682 [The interference pictures obtained for beams of
massive corpuscles C60 are one of the most convincing proofs that the particle-wave duality is not
limited to photons, electrons and other tiny particles. Indeed, it was verified independently for nucleons
(neutron interferometry) and for some atoms (e.g. sodium). Most recently it is also observed for C70
fullerenes]

Ashtekar A. and Geroch R. (1974) Quantum theory of gravitation, Rep. Prog. Phys. 37, 1211-1256 [An
essay explaining the ideas of the loop quantum gravity, a quickly developing domain, originated by
works of Abhay Ashtekar (Center for Gravitational Physics and Geometry, USA). In his vision the
space points are substituted by closed curves (loops), the surfaces and volumes are quantized, an effect
which becomes essential below the Planck distance. The interested readers might consult the works of
Ashtekar A., Rovelli C., Lewandowski J., Smolin L. and others. In particular, Ashtekar A, Rovelli C,
(1991) Gravitations and loops, Phys. Rev. D 44, 1740-1755, Rovelli C.(2004); furthermore, the study
of the Big Bang from this point of view is found in Ashtekar et al., "Quantum nature of the big-bang:
an analytical and numerical investigation", Phys.Rev. D 73, 124038-124071 (2006). The critical
discussions of the black hole entropy are found, e.g. in (Domagala M and Lewandowski J (2004)
"Black hole entropy from quantum geometry", Class .Quant. Grav. 21, 5233-5244]

Attewischer E., van Exter M.P., and Woerdmann J.P. (2002) Plasmon-assisted transmission of
entangled photons, Letters to Nature, Nature 418, 304-306 [One of examples that the photon pure
states are not necessarily destroyed by a temporary absorption by a macroscopic environment
(plasmons). C.f. also the ideas of Leggett A.J. (2002)]

Avron J.E., and Simon B. (1981) Almost periodic Hill’s equation and the rings of Saturn, Phys. Rev.
Lett. 46, 1166-1168 [The description of the Saturn rings proposed by the authors was not accepted by
the astronomers; yet the paper is a surprising example of how the Schrodinger's equation can become a
model for purely macroscopic phenomena]

Ballentine L.E. (1970) The statistical interpretation of quantum mechanics, Rev. Mod. Phys. 42, 358-
381 [The author sticks to the "ensemble interpretation" of QM, denying that the state vector ("wave
function") can represent the single particle state and negating the reality of the state collapse
(reduction)]

Ballentine L.E., Pearle Ph., Walker E.H., Sachs M., Koga T., and Gerver J. (1971, April) Quantum-
mechanics debate, Phys. Today 24, 36-44; (1971, October) Still more quantum mechanics, ibid 11-15

Ballentine L.E. (1991) Comment on “Quantum Zeno effect”, Phys Rev A 43, 5165-5167 [Following
the ensemble interpretation, L.E. Ballentine questions the need of using the reduction postulate in the
explanation of the 'quantum Zeno effect'. In his parallel polemic paper, "Limitations of the projection
postulate" (1990) Found. Phys. 20, 1329-1343, he concludes: "...the projection postulate may
accurately be described as either redundant or wrong". See, however, different opinions e.g. in (von

51
Neumann 1932, Dirac 1935, Wheeler and Zurek 1983, Itano et al. 1990, Bennet et al. 1993, Penrose
1994, Greenberger and Zeilinger 1995, Penrose 2004)]

Bell J.S. (1966) On the problem of hidden variables in quantum theory, Rev. Mod. Phys. 38, 447-452
[The historical paper in which the simplest form of John S. Bell inequality was derived. An equally
relevant case known as CHSH inequality (Clauser J.F., Horne M.A., Shimony A. and Holt R.A. (1969)
Phys. Rev. Lett. 23, 880-884) is frequently evoked to show the non-locality of the quantum EPR
correlations; see, e.g. (Leggett A.J. 2002) The still more general forms of Bell's inequalities due to
Mermin N.D., are collected e.g. in (Cereseda J.L. (2001) Phys. Lett. A 286, 376-382)]

Bell J.S., and Hallet M. (1982) Logic, quantum logic and empiricism, Phil. Sci. 49, 355-79 [The
fundamental paper questioning Putnam's arguments on the validity of quantum logical axioms and the
necessity of the Hilbert space formalism in quantum theory]

Bell J.S. (1987) Speakable and unspeakable in quantum mechanics, Cambridge: Cambridge University
Press [The summa of John S. Bell ideas on the interpretation of quantum theory. The reader will enjoy
his humorous comments on some generally accepted concepts and opinions]

Beltrametti E.B., and Casinelli G. (1981) The logic of quantum mechanics, in Encyclopedia of
Mathematics, Rota G.C. (Ed) London, UK: Addison Wesley

Bennett C.H., Bessette F., Brassard G., Salvail L., and Smolin J. (1992) Experimental quantum
cryptography, J. Cryptology 5, 3-28

Bennett C.H., Brassard G., and Ekert A.K. (1992) Quantum cryptography, Sci. Am. 267, 50-57

Bennett C.H., Brassard G., and Crépeau C. (1993) Teleporting an unknown quantum state via dual
classical and Einstein-Podolski-Rosen channels, Phys. Rev. Lett., 70, 1895-1899 [A classical paper
explaining in simple terms the mechanism of teleportation of an arbitrary, unknown state between two
spatially separated detectors in the EPR experiment. The effect was experimentally verified by works
of Insbruck group, C.f. (Zeilinger 2000)]

Bergquist J.C., Hulet R.G., Itano W., and Wineland D.J. (1986) Observation of quantum jumps in a
single atom, Phys. Rev. Lett. 57, 1699-1702 [One of the well known papers on the sudden luminosity
jumps observed in the radiation of the 3-level system, affected by the resonant 2-frequency
electromagnetic waves. A suggestive interpretation is that the jumps correspond to the spontaneous
reductions of the 3-level states]

Bialynicki-Birula I., and Mycielski J. (1976) Nonlinear-wave mechanics, Ann. Phys. 100, 62-93 [The
scheme of quantum mechanics based on the Schrodinger's wave equation with the logarithmic non-
linearity. The theory makes possible a natural formulation of the collapse hypothesis for packets
composed of several space separated parts]

Biberman L., Sushkin N., and Fabricant V. (1949) Difraktsiya poocheredno letyashchikh
elektronov, Docl. Acad. Sci. USSR 66, 185-186 [i.e., "The diffraction of separately flying
electrons"; the paper reports the first experiment in which the wave behavior is observed for an
electron beam of so low intensity that it can hardly be attributed to a collective effect of many
electron interactions]

Birkhoff G., and von Neumann J. (1936) The logic of quantum mechanics, Ann. Math. 37, 823-843
[An emblematic paper suggesting that the complementarity considered by the Copenhagen School is
indeed the manifestation of a new type of the logic obeyed by quantum phenomena]

52
Blokhintsev D.I. (1968) The philosophy of quantum mechanics, Dordrecht, Holland: Reidel [Dimitri
Ivanovich Blokhintsev, an outstanding follower of the "ensemble interpretation" of QM; a doctrine
opposed to the Copenhagen School, derived from Einstein's ideas, popular also in the former Soviet
Union for ideological reasons]

Bohm D., and Aharonov Y. (1957) Discussion of experimental proof for the paradox of Einstein,
Rosen, and Podolsky, Phys. Rev. 108, 1070-1076 [The spin variant of the EPR thought-experiment
illustrates most transparently the non-local aspects of the reduction (collapse) axiom for the
antisymmetric singlet state]

Bohm D., and Bub J. (1966) A proposed solution of the measurement problem in quantum mechanics
by a hidden variable theory, Rev. Mod. Phys. 38, 453-469; A refutation of the proof by Jauch and Piron
that hidden variables can be excluded in quantum mechanics, ibid 470-475 [One of the most interesting
intents of constructing the hidden variables for the spin measurements. Even if the generalization to the
higher dimension is questionable (Kochen S. and Specker E.P., 1967) and even if it could not explain
the EPR statistics for the singlet state, the paper is worth of attention for rising some wider questions
as, e.g., whether the decoherence is indeed sufficient to describe the mechanism of the measurement.
The authors doubt: '...When this packet arrives at the film (...) it is "broken up" into many very small
packets, which cease to interfere coherently. What is still unexplained, however, is that only one of
these packets contains an electron...' (p.459, col.1, lines 13-8 from the bottom). C.f. also (Haroche S.
1998, Leggett A.J. 2002)]

Bohr N. (1913) On the constitution of atoms and molecules, Phil. Mag. 26, 1-25, 476-502, 857-875
[The observation that the atomic spectral frequencies are differences between some fixed spectral
terms lead Bohr to his historical discovery of the distinguished, stable orbits of the atomic electrons]

Bohr N. (1928) The quantum postulate and the recent development of atomic theory, Nature 121
(Suppl.) 580-590

Bohr N. (1935) Can quantum-mechanical description of physical reality be considered complete?,

Phys. Rev. 48, 696-702 [Niels Bohr's famous answer to the equally famous objections in (Einstein A.,
Podolsky B. and Rosen N., 1935) about the non-locality of quantum measurements]

Bohr N. (1948) On the notions of causality and complementarity, Dialectica 2, 312-319 (reprinted in:
Science 111 (1950) 51-54)

Born M. (1926a) Zur quantenmechanik der stossvorgänge (The quantum mechanics of the impact
process) Z. Physik 37, 863-867. Translated in Wheeler and Zurek (1983) [In his first, intuitive intent to
formulate the statistical interpretation of the wave packet Max Born (Nobel Prize in Physics, 1954)
postulated the probabilities proportional to the transition amplitudes. An improved, though still
imprecise version (probabilities = amplitude squares) was added in the last moment, in his note added
in proof. The correct formula was finally published in (Born 1926b)]

Born M. (1926b) Quantenmechanik der stossvorgänge, Z. Physik 38, 803-827

Born M. (1969) Atomic physics, New York: Dover Pub. Inc.

Born M (1971) The Born-Einstein letters, Correspondence between Albert Einstein and Max and
Hedwig Born from 1916 to 1955, with commentaries by Max Born, New York: Walker and Company

53
Bouwmeester D., Ekert A., and Zeilinger A. (2001) The physics of quantum information, New York:
Springer [The book by known leaders of quantum information, of notable interest to the advanced
students (C.f. also Nielsen M.A. and Chuang I.L, 2000)]

Brumer P., and Shapiro M. (1995) Laser control of chemical-reactions, Sci. Am. 272, 56-58

Brune M, Schmidt-Kaler F., Maali A., Dreyer J., Hagley E., Raymond J.M., and Haroche S. (1996)
Quantum Rabi oscillations: a direct test of field quantization in a cavity, Phys. Rev. Lett. 76, 1800-
18003 [The patient research of the experimental/theoretical group in the Ecole Normale Superieure
(Paris) shows the reality of the superposed energy states in a single atom. The superposed "pink state"
is created by generating the Rabi rotations between the "red" and "blue" energy eigenstates of an atom
passing through an optical cavity. Perhaps, it can answer a part of the doubts expressed in (Primas
1983)?]

Cervantes-Cota J.L., Galindo S., Klapp J., and Rodríguez-Meza M.A. (2005) Las Mejores Historias
del Joven Einstein (in Spanish), México: Ediciones del Milenio [A sequence of historical studies on the
origin of quantum and relativistic ideas at the beginning of XX c. The reader will find lot of unknown
facts, amazing anecdotes and polemic disputes relevant for the shape of the present day physical
theories]

Chu S. (1992) Laser trapping of neutral particles, Sci. Am. 266, Num 2, 48-54

Cohen-Tannoudji C., Diu B., and Lalöe F. (1977) Quantum mechanics, Paris, France: Hermann and
Wiley

Cohen-Tannoudji C., and Phillips W.D. (1990, October) New mechanisms for laser cooling, Phys.
Today, 33-40 [Two of the 1997 Nobel Prize winners about one of the most ingenious methods of
controlling the atomic motion]

Cohen-Tannoudji C., Dupond-Roc J., and Grynberg G. (1992) Atom-photon interactions, New
York: Wiley [A careful review of the radiation theory, Floquet spectra and the principal methods of
quantum optics. Transparently written, of high relevance for the advanced students]

Cohen-Tannoudji C. (1998) Manipulating atoms with photons, Rev. Mod. Phys. 70, 707-719

Compton A.H. (1921) The magnetic electron, J. of the Franklin Institute 192, 145-155

Compton A.H. (1961) Scattering of x-rays as particles, Am. J. Phys. 29, 817-818 [One of crucial
experiments confirming the granular nature of the light, described retrospectively by the author
himself. When Arthur H. Compton (Nobel Prize in Physics, 1927) reported his result in 1923, his
colleagues were reluctant to believe. Compton had even some difficulties with publishing his paper.
Today, one of obligatory chapters in the quantum mechanics textbooks]

Corson E.M. (1953) Introduction to tensors, spinors and relativistic wave equations, London:
Blackie [A classical though forgotten monograph, still worth consulting]

Davisson C., and Germer L.H. (1927) Diffraction of electrons by a crystal of nickel, Phys. Rev. 30,
705-740 [The interference effects for an electron beam reflected by the crystalline lattice confirm the
strange idea of Luis de Broglie]

54
de Broglie L. (1924) Recherches sur le théorie des quanta, Paris: Masson [The historical hypothesis of
Luis de Broglie (Nobel Prize in Physics, 1929) that the beam of any micro-particles must show the
wave behavior. In his times, it seemed almost a science-fiction story]

de Broglie L. (1960) Non-linear wave mechanics: A causal interpretation, Amsterdam: Elsevier

de la Peña L. (1991) Introducción a la mecánica cuántica, México D.F.: Fondo de Cultura Económica

Deutsch D., Barenco A., and Ekert A. (1995) Universality in quantum computation, Proc. R. Soc.
Lond. A 449, 669-677

Deutsch D. (1985) Quantum theory, the Church-Turing principle and the universal quantum computer,
Proc. R. Soc. London A 400, 97-117 [The fundamental ideas by one of the pioneers of quantum
computing. Recommended as the first (and second) contact with the subject]

Dirac P.A.M. (1928a) The quantum theory of the electron part I, Proc. R. Soc. London 117, 610-
624; part II, Proc. R. Soc. London 118, 351-361

Dirac P.A.M. (1928b) On the quantum theory of the electron, Phys. Z. 29, 561-563 [The 1928
papers of Paul A.M. Dirac (Nobel Prize in Physics, 1933) presenting the theory of the relativistic
electron were for a long time under a persistent criticism of Niels Bohr and other authors. The
situation changed after the discovery of positron. See also the historical studies of Moyer D.F.
(1981a-c)]

Dirac P.A.M. (1935) The principles of quantum mechanics (2nd Ed.), Oxford: The Clarendon Press
[The bestseller textbook of P.A.M. Dirac, explaining the concepts, notation and elementary algebraic
methods of quantum mechanics. Based on the informational interpretation of quantum states (the pure
state is a maximal set of non contradictory information about the quantum system), it yields the most
transparent interpretation of quantum theory. In spite of the difference in epochs, the heuristics and
formalism of Dirac are highly recommended for the present day students of physics, mathematics and
physical chemistry]

Dirac P.A.M. (1938) Classical theory of radiating electrons, Proc. R. Soc. London A 167 (A929), 148-
169

Dodonov V.V. (2002) Nonclassical states in quantum optics: a 'squeezed' review of the first 75 years,
J. Opt. B-Quantum and Semiclassical Optics 4, R1-R33

Einstein A. (1905) Über einen die erzeugung und verwandlung des lichtes betreffenden heuristichen
gesichtspunkt, Annalen der Physik 17, 132-148 [The fundamental paper of Albert Einstein (Nobel
Prize in Physics, 1921) on the quantum nature of the light revealed by the photoelectric effects.
Curiously, opposed for a long time by Planck and Bohr in spite of their crucial contributions to
quantum theory]

Einstein A., Rosen N., and Podolsky B. (1935) Can quantum-mechanical description of physical
reality be considered complete?, Phys. Rev. 47, 777-780 [The thought experiment on the entangled
two particle state (shortly, the EPR experiment), with predictable non-local aspects leads Albert
Einstein, Boris Podolsky and Natan Rosen to question the interpretation of quantum mechanics. The
effect is up to now the subject of fundamental discussions, and simultaneously, the theoretical basis of
the control operations essential for quantum computing. The non-local correlations for the spin and
polarization variants of the EPR experiment are recently checked with an increasing precision (see,
e.g., Tittel W., et al. 1998)]

55
Einstein A., and Infeld L. (1966) The evolution of physics, Mass. USA: Murray Printing Co. [A
fascinating book outlining the successes and defeats of physical ideas during several centuries. The
error of the idea of "caloric", the non-existence of ether and other problems should be also remembered
by the present day vanguard theories (as e.g. the "dark energy”)]

Ekstrom P., and Wineland D. (1980) The isolated electron, Sci. Am. 243, Num 2, 90-101

Elitzur A.C., and Vaidman L. (1993) Quantum mechanical interaction-free measurements, Found.
Phys. 23, 987-997 [One of the most provocative papers in the last decade of XXc. Due to the editor's
indecision it circulated unpublished (as the preprint TAUP 1865-91) until the reinterpretation of Hardy
L. (1992) Phys. Lett. A 167, 11-16 opened a world wide discussion. The problem as to, whether the
reduction of the quantum state can affect the past of quantum systems is imminent. An attempt to solve
this problem was presented by Cramer in the frame of the "Transactional Quantum Mechanics".
However, the recent discussion in (Elitzur A.C. and Dolev Z. (2007) Multiple interaction-free
measurement as a challenge to the transactional interpretation of quantum mechanics, in Frontiers of
Time: Retrocausation - Experiments and Theory, AIP Conf. Proc. 863, pp. 27-43) implies that even the
"transactional interpretation" fails. A visible challenge for the old and new interpretations of QM might
indicate the need of a new theory]

Ferguson A.J., Cain P.A., Williams D.A., and Briggs A.D. (2002) Ammonia-based quantum computer,
Phys. Rev. A 65, 034303

Feynman R.P. (1948) Space-time approach to non-relativistic quantum mechanics, Rev. Mod. Phys. 20,
367-387 [The idea of path integrals presented for the first time by Richard Feynman at the level of the
non relativistic quantum mechanics. Today, one of the principal formulations of all quantum theories.
It permits to avoid the difficulties affecting the canonical quantization, introducing some unified
background into the different quantum field theories]

Feynman R.P., Leighton R.B., and Sands M. (1963) The Feynman Lectures on Physics Vol. 1, New
York: Addison-Wesley.

Feynman R.P. (1982) Simulating physics with computers, Int. J. Theor. Phys. 21, 467-488

Finkelstein D. (1963) The logic of quantum physics, Trans. NY Acad. Sci. 25, 621-637

Fivel D.I. (1991) Geometry underlying no-hidden-variable theorems, Phys. Rev. Lett. 67, 285-289
[The John S. Bell's theorem is shown equivalent to the triangle inequality in an abstract metrical space
defined by the probability measures on a Boolean algebra]

Gasiorowicz S. (1974) Quantum physics, Singapore: John Wiley and Sons

Gerlach W., and Stern O. (1921) The experimental proof of magnetic moment of silver atoms, Z.
Physik 8, 110-11 (Sp. Iss. Yrs 1921/2 Dec-Feb 1921) [The historical experiments in which the spin
spectral effects were observed still before the existence of the spin was accepted]

Gerlach W., and Stern O. (1922) The experimental evidence of direction quantization in the magnetic
field, Z. Physik 9, 349-352

Gilson J.G. (1968) On stochastic theories of quantum mechanics, Proc. Cambridge Phil. Soc. 64,
1061-1070

56
Gisin N., and Rigo M. (1995) Relevant and irrelevant nonlinear Schrödinger equations, J. Phys. A:
Math. and Gen. 28, 7375-7390 [A fundamental theorem showing that the non-linear quantum
mechanics for one particle states and the conventional QM for the entangled many particle states,
together with the reduction axiom, would imply the possibility of reading the superluminal messages in
the EPR-type experiments]

Gisin N., Ribordy G.G., Tittel W., and Zbinden H. (2002) Quantum cryptography, Rev. Mod. Phys. 74,
145-195

Glauber R.J. (1963) Coherent and incoherent states for the radiation field, Phys. Rev. 131, 2766-2788

Glauber R.J. (1965) Optical coherence and photon statistics, in DeWitt C., Blandin A., and Cohen-
Tannoudji C. (Eds), Quantum optics and electronics, New York: Gordon and Breach [A nicely written
article, offers a transparent guide to the ideas of Roy J. Glauber (Nobel Prize in Physics 2005)]

Golub R., Mampe W., Pendlebury J.M., and Ageron P. (1979) Ultracold neutrons, Sci. Am. 240, 106-
119

Greenberger D.M. (1983) The neutron interferometers as a device for illustrating the strange
behavior of quantum systems, Rev. Mod. Phys. 55, 875-905

Greenberger D.M., and Yasin A. (1989) Haunted measurements in quantum-theory, Found. Phys. 19,
679-704

Greenberger D.M., Horne M.A., Shimony A., and Zeilinger A. (1990) Bell’s theorem without
inequalities, Am. J. Phys. 58 (12), 1131-1143

Greenberger D.M., and Zeilinger A., Eds. (1995) Fundamental Problems in Quantum Theory, Ann. N.
Y. Acad. Sci., Vol. 755, New York, N.Y. [One of the most interesting Conference Proceedings,
dedicated to the entangled states, non-locality, hidden variables and other fundamental problems of
quantum theory]

Gribbin J. (1996) Schrodinger's kittens and the search for reality: Solving the quantum mysteries,
New York: Orion Publishing Co. [An excellent, humorous book about the models of the
Schrodinger's cat "in little"]

Haag R. (1996) Local quantum physics, fields, particles, algebras, 2nd Ed., Berlin: Springer Verlag
(See Chapter VII) [A profound mathematical presentation of the idea of locality, local observables and
measurements by Rudolf Haag, one of the principal leaders of the C*-algebraic approach to quantum
field theories. In the introduction the reader will find an ample conceptual study including the "event
enhanced quantum mechanics"]

Hackermüller L., Uttenthaler S., Hornberger K., Reiger E., Brezger B., Zeilinger A., and Arndt M.
(2003) The wave nature of biomolecules and fluorofullerenes, Phys. Rev. Lett. 91, 090408

Haroche S., and Raimond J.M. (1996) Quantum computing: Dream or nightmare?, Phys. Today 48 (8),
51-52

Haroche S. (1998) Entanglement, decoherence and the quantum/classical boundary, Phys. Today 51
(7) 36-42 [Serge Haroche, one of the best known leaders in the foundations and phenomenology of
QM. Together with the team of Ecole Normale Superieure (Paris) he is developing a systematic study
to check the reality of the quantum mechanical "thought experiments" decisive for the meaning of

57
quantum theory. One of the most surprising results of the ENS-group is the laboratory observation of a
single photon in an optical cavity. The atom crossing the cavity "borrows" the photon which causes the
Rabi rotation of the atomic states. When the atom leaves the cavity, the photon is honestly turned back
without any energy loss: however the transaction is documented in form of a changed phase of the
atomic state... Haroche reports also the robust behavior (stability) of the eigenstates of a measured
observable and the volatility of the superposed states; though (as he states) the mechanism of the
choice still involves a mystery]

Hau L.V. (2004) Frozen light, Sci. Am. Special 13, Number 1, p 44

Havel T.F., Cory D.G., Lloyd S., Boulant N., Fortunato E.M., Pravia M.A., Teklemariam G.,
Weinstein Y.S., Bhattacharyya A., and Hou J. (2002) Quantum information processing by nuclear
magnetic resonance spectroscopy, Am. J. Phys. 70, 345-362

Heisenberg W. (1927) Über den ansclauchichen Inhalt der quantentheoretischen Kinematik und
Mechanik, Zeitschrift für Physik 43, 172-198 (English translation, Wheeler and Zurek 1983, p 62-84)

Heisenberg W. (1930) The physical principles of quantum theory, Chicago: University of Chicago
Press [The formulation of QM by Werner Heisenberg was parallel to the wave mechanics of de Broglie
and Schrödinger but obeyed a different idea of an "empirical realism" (indeed, a neopositivist
philosophy in QM). According to this approach, the physical quantity "in itself" makes no sense unless
associated with a well defined measuring prescription. However, the experimental prescriptions
defining the physical observables might be incompatible, which means that some properties or
concepts ("faces of reality") are complementary i.e. not simultaneously meaningful. "It makes as little
sense to speak of the frequency of the light wave at a definite instant as of the energy of an atom at a
definite moment" (translated in Wheeler J. and Zurek W. 1983, p.67). Following the idea, Heisenberg
proposed an ample collection of the "uncertainty principles" including position-momentum, phase-
action, time-energy, etc. His description of observables and the formulation of the dynamic theory
obeyed the rule of the canonical quantization exploring the analogy between the Poisson brackets and
commutators in a linear algebra]

Heisenberg W. (1955) The development of the interpretation of the quantum theory, in W. Pauli (Ed),
Niels Bohr and the development of physics, London: Pergamon Press, pp 12-29

Heisenberg W. (1958) Physics and philosophy, New York: Harper and Brothers

Hepp K. (1998) Toward the demolition of a computational quantum brain, in Quantum Future, P.
Blanchard, and A. Jadczyk (Eds), Lecture Notes in Physics, 517, 92-104 [Evaluating the orders of
magnitude, Klaus Hepp arrives at a skeptical conclusion about the collapse of the wave packet as a
mechanism of our thinking. Our brain is macroscopic and so are our thoughts, he says. See also a
sharp criticism of the "quantum computing" and its inadequacy to describe the working of our brain
(Koch Ch and Hepp K (2006), Quantum mechanics in the brain, Nature 440, 611-612)]

Holland P.R. (1993) The quantum theory of motion, Cambridge, UK: Cambridge University Press [An
ample monograph dedicated to the possibility of the classical background of quantum mechanics. A
particular attention is dedicated to the hidden variable scheme of David Bohm (1952), Phys. Rev. 85,
166-193 (reprinted also in Wheeler J. and Zurek W. 1983, Ch.III.3, pp. 369-396) presented by Holland
in a wider perspective, including the mathematical and philosophical aspects, application to many
particle systems, non-locality and other fundamental problems]

Huttner B., Muller A., Gautier J.D., Zbinden H., and Gisin N. (1996) Unambiguous quantum
measurement of non-orthogonal states, Phys. Rev. A 54, 3783-3789 [A surprising result about the

58
techniques of distinguishing unmistakably the non-orthogonal pure states of a quantum ensemble, at
the cost of loosing a part of the ensemble individuals. The effect illustrates the influence of an
absorbing medium onto the quantum states of micro-systems even if they apparently escaped
interaction. The authors discuss its possible advantages for quantum cryptography]

Itano W.M., Heinzen D.G., Bollinger J.J., and Wineland D.G. (1990) Quantum Zeno effect, Phys. Rev.
A, 41, 2295-2300 [A widely commented article about the three level optical system under the influence
of two competing, resonant electromagnetic fields. One of them causes a slow Rabi rotation between a
pair of levels, while the other one, in form of short pulses and rest intervals, is supposed to act as a
sequence of measurements inducing the collapse phenomena. If frequent enough, they should block the
Rabi rotations; an effect indeed observed, taken by the authors as a proof of the physical reality of the
state collapse (however, see the strong objections of other authors; e.g., Ballentine L.E. 1991)]

Jammer M. (1966) The conceptual development of quantum mechanics, New York: McGraw-Hill

Jammer M. (1974) The philosophy of quantum mechanics; the interpretations of quantum mechanics
in historical perspective, New York: Wiley [Both books of Max Jammer are one of the most complete
sources of historical and conceptual information. Who will read them carefully, without avoiding the
labyrinth of footnotes, can find a lot of forgotten truth about the origin and development of physical
ideas without excluding amazing anecdotes (e.g., Max Planck did not believe in the existence of
photons, Wolfgang Pauli was an antagonist of the electron spin, etc.)]

Joffe A. (1913) Sitz. Ber. Bayer. Akad. Wiss. (München) Math. Phys. Kl. 43, 19 [One of crucial
experiments confirming the existence of photons, absent in most textbooks and monographs. See also
(Meyer E. and Gerlach W. 1914)]

Kim J., Benson O, Kan H., and Yamamoto Y. (1999) A single-photon turnstile device, Nature 397 (2)
500-503 [Thanks to new methods of cavity quantum electrodynamics, an explicit proof of the existence
of a single photon interacting with an atomic system could be obtained; see also (Haroche 1998)]

Klein F. (1893) Vergleichende Betrachtungen über neuere geometrische Forschungen,

Mathematische Annalen 43, 63-100 [The famous Verlangen program of Felix Klein considers
groups as fundamental, geometries as secondary. Klein classifies the different geometries according
to their invariance groups. The point of view which might seem inverting the natural order of
concepts, turns useful to explain the origin of some physical geometries where the group was indeed
first. Thus, e.g., the simplectic geometry of the classical phase space appears as the invariant of the
classical motion group, while the Hilbert space geometry in QM is the product of the (unitary) state
transformations generated by the Schrodinger (or Dirac) evolution equations. Quite similarly, in the
special relativity, the Lorentz group was first, the geometry of the Minkowski space-time appeared
later]

Klein O. (1926) The quantum theory and five-dimensional relativity theory, Z. Phys. 37, 895-906

Kochen S., and Specker E.P. (1967) The problem of hidden variables in quantum mechanics, J. Math.
Mech. 17, 59-87

Kofman A.G., and Kurizki G. (2000) Acceleration of quantum decay processes by frequent
observations, Letters to Nature, Nature 405, 546-549

Kwiat P., Weinfurter H., and Zeilinger A. (1996, November) Quantum Seeing in the Dark, Sci. Am.
275, 52-58 [The consequences of the interaction free experiment of Avshalom C. Elitzur and Lev

59
Vaidman are analyzed. It makes possible to detect a physical object by a photon which passed far away
without ever illuminating it]

Lalöe F. (2001) Do we really understand quantum mechanics? Strange correlations, paradoxes, and
theorems, Am. J. Phys. 69, 655-701 (erratum in Am. J. Phys. 70 (2002) 556) [The Author reconsiders
the interpretation, paradoxes and unsolved conceptual problems of quantum mechanics, collecting
carefully all arguments on favor or against the widely shared convictions. In spite of all disquieting
questions, he finds some important advantages of the orthodox Copenhagen interpretation. Interesting
to read, recommended to all students of the exact sciences]

Leggett A.J. (2002) Testing the limits of quantum mechanics: motivation, state of play, prospects, J.
Phys.: Condens. Matter 14, R415-R451 [The possible steps of QM towards a new quantum theory are
considered. The principal challenge on this way is the measurement paradox (known also as the
paradox of the Schrodinger's cat). Against the belief of some other authors, Anthony J. Leggett (Nobel
Prize in Physics, 2003) considers the state reduction as a real process and the question about its
mechanism as a key to further development of the theory. He dedicates a systematic study to the
mesoscopic systems (such as the currents crossing the Josephson junction) which can produce the
superpositions of macroscopically different states]

Leonard M.A. (1998) Computing with DNA, Sci. Am. 279, 54-61

London F., and Bauer E. (1939) La Théorie de l’observation en mécanique quantique, Paris: Hermann
et Cie. [The metaphysical hypothesis that the consciousness of an intelligent observer is what indeed
reduces the wave packet of a quantum system. (English version see, Wheeler J. and Zurek W. 1983,
Ch.II.1) The idea is supported in (Wigner E.P. 1962); in a more realistic sense in (Penrose R. 1996),
objected in (Bell J.S. 1987, Ch.15, p.117), then in (Hepp K. 1998)]

Madore J. (1995) An introduction to noncommutative differential geometry and its applications,

Cambridge, UK: Cambridge University Press [The readers willing to focus their attention on the
simplest physical ideas, might also like Madore J (1992) Fuzzy Physics, Ann. Phys. 219, 187-198. For
more mathematically minded: Majid S (1995) Foundations of Quantum Group Theory, Cambridge,
UK, Cambridge University Press. Highly recommended as well one of the pioneer papers:
Woronowicz S.L. (1987) Compact matrix pseudo groups, Commun. Math. Phys. 111 613-665]

Maeda H., and Gallagher T.F. (2004) Nondispersing wave packets, Phys. Rev. Lett. 92 133004 [An
experimental research showing the existence and stability of the circulating "planetary" states in the
Rydberg atoms. (For the mathematically minded readers, worth reading are also: Delande D,
Zakrzewski J, Phys. Rev. A 58, 466 (1998); Bialynicki-Birula I, Phys. Rev. Lett. 93, 020402 (2004) and
the literature cited there)]

Magnus W., and Winkler S. (1973) Hill’s Equation, New York: Dover

Margenau H. (1936) Quantum-mechanical description, Phys. Rev. 49, 240-242

Marshall T.W. and Santos E. (1989) Stochastic optics: A local realistic analysis of optical tests of Bell
inequalities, Phys. Rev. A 39, 6271-6283 [The paper represents the still existing opposition against the
reality of the light quanta. The authors defend the original ideas of Max Planck and Niels Bohr that the
electromagnetic wave is continuous. If it is absorbed in portions, this is only because of the mechanism
of absorption (relaxation phenomena), not due to the quantum structure of the light itself]

Mehra J. (1974) The quantum principle: Its interpretation and epistemology, Boston, USA: D. Reidel
Publishing Company

60
Mermin N.D. (1998) What is quantum mechanics trying to tell us? Am. J. Phys. 66, 753-767 [The
points of view of Mermin are not easy to classify. In his widely discussed papers he insists that most of
the quantum mechanical concepts are secondary if not illusionary. The only phenomenon described by
quantum theory are correlations (coincidence probabilities) between quantum events. "Correlations and
only correlations!" he says. This would be close to the ensemble interpretation, but simultaneously
Mermin considers the probabilities well defined for each single micro-system. His point of view is
questioned by several other authors, e.g. (Griffiths R.B. (2003) Found. Phys. 33, 1423-1459), but yet,
worth attention for the fans of quantum information]

Mermin N.D. (2003) From cbits to qbits: Teaching computer scientists quantum mechanics, Am. J.
Phys. 71, 23-30

Meyer E., and Gerlach W. (1914) Über den Photoelektrischen Effekt an ultra microskopischen
Metallteilchen (The Photoelectric effect on ultra microscopic metallic particles), Ann. der Phys 45,
177-236 [A crucial (and forgotten) experiment showing that the energy quanta absorbed in photo-
effects cannot be explained by a gradual accumulation of the continuous field energy (see also Joffe
1913)]

Mielnik B. (1974) Generalized quantum mechanics, Commun. Math. Phys. 31, 221-256 [An outline of
a statistical theory in which the pure states do not need to correspond to the unit vectors, the transition
probabilities are not the moduli squared of the scalar products and the "quantum logic" is not
necessarily orthocomplemented. The existence of such structures might mean that the formalism of
Hilbert spaces is not the only option for quantum theories. Supported in (Bell J.S. and Hallett M.
1982), discussed also in (Penrose R. 1997)]

Mielnik B., and Tengstrand G (1980) Nelson-Brown motion: Some question marks, Int. J. Theor.
Phys. 19, 239-250 [It is argued that the model of Brownian motion used by E. Nelson to explain the
Schrodinger's wave equation has some interpretational gaps; if analyzed carefully, it becomes just a
statistical variant of the pilot wave theories]

Millikan R.A. (1948) Franklin’s discovery of the electron, Am. J. Phys. 16, 319

Misra B., and Sudarshan E.C.G. (1977) The Zeno's paradox in quantum theory, J. Math. Phys. 18,
756-763 [A famous result showing that the frequently repeated measurement can inhibit the evolution
of quantum systems. A mathematically equivalent theorem was proved earlier in (Khalfin L.A. (1958)
JETP 6, 1053), though the work of Misra and Sudarshan turned better known and more inspiring in the
fundamental research]

Moyer D.F. (1981a) Origins of Dirac’s electron, 1925-1928, Am. J. Phys. 49, 944-949

Moyer D.F. (1981b) Evaluations of Dirac’s electron, 1928-1932, Am. J. Phys. 49, 1055-1062

Moyer D.F. (1981c) Vindications of Dirac’s electron, 1932-1934, Am. J. Phys. 49, 1120-1125 [In
his three articles of 1981 D.F. Moyer describes the historical polemic in which the arguments of
P.A.M. Dirac (Nobel Prize in Physics, 1933), at the beginning, seemed in disadvantage (the
negative energy levels of the free electron, the hypothetical "electron sea" with an infinite number
of electrons occupying the continuum of negative levels, etc., all this under the persistent attacks of
Bohr and other physicists) But then, one of his most risky conjectures, about the "positive electron",
turned real. This started a fast development of quantum field theories, though against the skepticism
of Dirac himself]

61
Nagourney W., Sandberg J., and Dehmelt H. (1986) Shelved optical electron amplifier: Observation of
quantum jumps, Phys. Rev. Lett. 56, 2797-2799 [One of widely discussed 1986 papers showing the
instability and sudden transitions (quantum jumps) in a single atom. Tentatively, illustrating the
reduction of the wave packet]

Nelson E. (1967) Dynamical theories of Brownian motion, Princeton University Press, Princeton, USA
[A polemic hypothesis of Edward Nelson (known as the stochastic quantum mechanics). that the
wave-like behavior of micro-particles described by the Schrodinger's equation is indeed the
consequence of a fundamentally statistical but classical mechanics based on Newton's equations of
motion]

Newburgh R., Peidle J., and Rueckner W. (2006) Einstein, Perrin, and the reality of atoms: 1905
revisited, Am. J. Phys., 74, 478-481 [An interesting historical and fundamental essay analyzing the true
sources of our knowledge]

Nielsen M.A., Knill E., and Laflamme R. (1998) Complete quantum teleportation using nuclear
magnetic resonance, Letters to Nature, Nature 396, 52-55

Nielsen M.A., and Chuang I.L. (2000) Quantum computation and quantum information, Cambridge:
Cambridge University Press

Nogues G., Rauschenbeutel A., Osnaghi S., Brune M., Raimond J.M., and Haroche S. (1999) Seeing a
single photon without destroying it, Nature 400, 239-242 [One of the most ingenious experiments at
the end of XX c: the existence of a photon is documented without absorbing its energy]

Pais A. (1991) Niels Bohr's times, in physics, philosophy, and polity, Oxford, UK: Clarendon Press

Park J.L. (1968) Nature of quantum states, Am. J. Phys. 36, 211-225 [The Ph.D. thesis of James L.
Park contains one of the strongest arguments against the reduction of the wave packet, on favor of the
ensemble interpretation: due to the non-unique mixture decomposition into the pure states, the single
particle after the measurement is in no state at all (so, neither it is in an eigenstate!)]

Paul W. (1993) in Nobel lectures in physics 1981-1990, Singapore: World Scientific, p 610

Pauli W. (1925a) Concerning the influence of the equilibrium ability of the electron mass on the
Zeeman effect, Z. Physik 31, 373-385 [In an amazing historical quid pro quo, Wolfgang Pauli was
one of the most outstanding adversaries of the electron spin. Once forced to recognize its existence,
he contributed to the formal description. Paradoxically, the spin 1/2 of the electrons is called today
the spin of Pauli]

Pauli W. (1925b) On the connection of the arrangement of electron groups in atoms with the
complex structure of spectra, Z. Physik 31, 765-783

Pauli W. (1927) Über gasentartung und paramagnetismus, Z. Physik 43, 81-102

Pauli W. (1964) Nobel Lecture, December 13, in Nobel Lectures, Physics 1942-1962, Amsterdam:
Elsevier Publishing Co.

Penning F.M. (1936) The spark discharge in low pressure between coaxial cylinders in an axial magnet
field, Physica 3, 873-894

Penrose R. (1989) The Emperor’s New Mind, Oxford, UK: Oxford University Press

62
Penrose R. (1994) Shadows of the Mind, Oxford, UK: Oxford University Press

Penrose R. (1996) Mechanism, microtubules, and the mind, Journal of Consciousness Studies 1,
241-249

Penrose R. (1997) The large, the small and the human mind, Cambridge, UK: Cambridge University
Press

Penrose R. (2004) The road to reality, London: Jonathan Cape. [Roger Penrose, one of the most
talented authors and philosophers of physics, approaching the questions on the frontier of quantum and
relativistic theories. In his four monographs he describes the origin, puzzles and the possible future of
physical theories. Analyzing the links between the quantum and classical information, he is convinced
about the reality of quantum states and state reduction for the single particles. He thinks that the
reduction (collapse) events might be precipitated by the gravitational interactions, and he also believes
that they may be essential in our brain activity (in a way, supporting the "psychological reduction
mechanism" of London F. and Bauer E. 1939, Wigner E.P. 1962) In his search for the realistic
mechanism of the state collapse, Penrose is aware of the numerous quid pro quo and unavoidable
paradoxes. Citing Bob Wald, he writes "If you truly believe in quantum mechanics, then you cannot
take it seriously" (Penrose 1997, p.72). His interest is also dedicated to the possibility of the non-linear
graviton, for which he proposes a plausible analytical description, though the problem of the statistical
interpretation remains open, see e.g. his most recent book (Penrose 2004). In the same monograph he
comments an excess of arbitrary parameters and ad hoc models in the present day physics, making the
vanguard theories hard to falsify, and hence, vulnerable to degenerate into never ending stories with
little chances to be either rejected or verified. A fascinating literature for the young and old readers]

Phillips W.D., and Metcalf H.J. (1987) Cooling and trapping atoms, Sci. Am. 256, 36-42 [As a
surprising consequence of the Doppler effect, the electromagnetic waves can efficiently slow down the
running atoms. William D. Phillips together with Steven Shu and Claude Cohen-Tannoudji received
the Nobel Prize in Physics in 1997]

Piron C. (1964) Axiomatique quantique, Helv. Phys. Acta 31, 439-468 [One of the most careful
axiomatic studies deducing the Hilbert space formalism of quantum theories from the
orthocomplementarity and weak modularity of quantum logic (for mathematically minded readers)]

Primas H. (1983) Chemistry, quantum mechanics and reductionism, Berlin: Springer-Verlag [One of
exceptional monographs presenting the fundamental problems, interpretation and evolution of quantum
ideas on the abstract and applied level. Truly priceless as a source of systematically presented problems
and abundant, transparent bibliography. Apart of the encyclopedic achievements, the author discusses
some uneasy questions about the applicability limits of quantum theory, e.g.: "Why so many stationary
states not exist?", or "Why can approximations be better than the exact solutions?" (Ch.1.4, pp 12-13).
He also offers unexpected interpretations of some traditional concepts, helping to understand the
essence of quantum theory, e.g. "Pure quantum states are objective but not real" (Ch.3.4, p.103). It can
be read almost as a novel.]

Rabi I.I. (1937) Space quantization in gyrating magnetic field, Phys. Rev. 51, 652-654

Rabi I.I., Ramsey N.F., and Schwinger J. (1954) Use of rotating coordinates in magnetic resonance
problems, Rev. Mod. Phys. 26, 167-171 [More than 50 years after the Rabi's 1937 paper, the results of
Rabi and collaborators proved crucial for the selective manipulations of the bound states of the spin
and/or atomic systems. C.f. the present day techniques of creating the superposed energy states in
optical cavities (Brune M. et al. 1996))]

63
Rauschenbeutel A, Nogues G, Osnaghi S, Bertet P, Brune M, Raimond JM, and Haroche S. (2000)
Step-by-step engineered multiparticle entanglement, Science 288, 2024-2028

Redhead M. (1987) Incompleteness nonlocality and realism: A prolegomenon to the philosophy of

quantum mechanics, New York: Oxford University Press [A transparent, compactly written
monograph, explains the physical background of the mathematically advanced ideas (for
mathematically minded readers)]

Reed M., and Simon B. (1978) Methods of Modern Mathematical Physics IV, New York: Academic
Press [An extensive "river-encyclopedia" painting the mathematical aspects of modern physics. Not
very easy, but extremely useful]

Riesz F., and Nagy B. (1990) Functional analysis, New York: Dover Pub Inc.

Romer A. (1948) Zeeman’s discovery of the electron, Am. J. Phys. 16, 216-218

Rosenfeld L (1953) Strife about complementarity, Science Progress 41, 393-410 (Revised version of
“L’évidence de la complementarité” in Louis de Broglie, physician et penseur, Paris: Albin Michel
(1953), pp 43-65)

Rovelli C. (2004) Quantum gravity, Cambridge monographs on mathematical physics, London:

Cambridge University Press

Sauter Th., Neuhauser W., Blatt R., and Toschel P.E. (1986) Observation of quantum jumps, Phys.
Rev. Lett. 57, 1696-1698

Schirmer S.G., Greentre A.D., Ramakrishna V., and Rabitz H. (2002) Constructive control of quantum
systems using factorization of unitary operators, J. Phys. A: Math. Gen. 35, 8315-8339 [The authors
present efficient methods of the step by step manipulation of many level quantum systems, including
the total population inversion. Mathematically elegant, some parts based on exact solutions]

Schrieffer J.R. (1964) Theory of superconductivity, Reading Mass.: W.A. Benjamin Inc.

Schrödinger E. (1978) Collected papers on wave mechanics, translated from the second 1928
German edition, New York: Chelsea Publishing Co.

Schrödinger E. (1935) Die gegenwrtige situation in der quantenmechanik, Naturwiss 23 807-812

(English translation, see Wheeler and Zurek 1983, Ch. I.11) [One of the most relevant works of
Erwin Schrodinger in which the informatics character of the wave function and its uncertainties are
recognized. "The psi-function as Expectation catalog" (Sec.7). While one information is sharp, the
other must be blurred (uncertainty principle). Schrödinger wondered, whether the blurring can
affect also the macroscopic systems and he illustrated his doubts in a little, humorous tale about a
living organism (a cat) coupled with an unstable atom, so that the animal must enter into a
superposed state of "being alive" and "being dead". While most of the world difficult papers end up
rather quickly in the dead archives, the problem of the Schrödinger's cat, presented in a simple
hearted language (Schrödinger E 1935, translated in Wheeler J. and Zurek W. 1983, p.157, col.1), is
still alive, subject of polemic discussions, which might turn decisive for the future of quantum
theory (see also Leggett A.J. 2002)]

Scully M.O., and Zubairy M.S. (2002) Quantum optics, Cambridge, UK: Cambridge University Press

64
Shimony A. (1963) Role of the observer in quantum theory, Am. J. Phys. 31, 755-773

Sommerfeld A. (1916) The quantum theory of spectral lines, Ann. der Phys. 51, 1-94; ibid III, 125-167
[The Nils Bohr's quantization rules in old QM are generalized for the relativistic electrons, yielding the
Bohr-Sommerfeld atomic levels]

Tittel W., Brendel J., Zbinden H., and Gisin N. (1998) Violation of Bell inequalities by photons more
than 10 Km apart, Phys. Rev. Lett. 81, 3563-3566

Thomson G.P. and Reid. A (1927) Diffraction of cathode rays by a thin film, Nature 119, 890-890

Thomson G.P. (1927) Diffraction of cathode rays by thin films of platinum, Nature 120, 802-802 [The
historical experiments of Thomson confirmed the hypothesis of Luis de Broglie, marking the
beginning of the modern quantum theory]

Uhlenbeck G.E., and Goudsmit S. (1925) Ersetzung der Hypothese von unmechanischen Zwang
durch eine Forderung bezuglich des inneren Verhaltens jedes einzelnen Elektrons,
Naturwissenschaften 13, 953-954

Uhlenbeck G.E., and Goudsmit S. (1926) Spinning Electrons and the Structure of Spectra, Nature
117, 264-265 [G.E. Uhlenbeck and S. Goudsmit, the authors of the hypothesis about the electron
spin, faced negative reactions of the scientific community around 1925. Intimidated by the criticism
of Wolfgand Pauli, Werner Heisenberg and others, they were at the point to give up their idea, but
finally, supported by Ehrenfest, decided to publish it in their 1925, 1926 papers]

von Neumann J. (1932) Mathematische Grundlagen der Quanten-mechanik, Berlin: Springer-

Verlag (English translation: 1955, Princeton: Princeton University Press) [The classical
mathematical monograph introducing the formalism of Hilbert spaces in Quantum Mechanics. The
reader may also find a study of the uncertainly principles with the detailed presentation of the
Heisenberg’s microscope, and the discussion of the collapse axiom with its intrinsic paradoxes]

Watson E.C. (1947) Jubilee of the electron, Am. J. Phys. 15, 458-464

Weinacht T.C., Ahn J., and Bucksbaum P.H. (1998) Measurement of the amplitude and phase of a
sculpted Rydberg wave packet, Phys. Rev. Lett. 80, 5508-5511 (erratum ibid 81, 3050-3050)

Weinberg S. (1989) Testing quantum-mechanics, Ann. Phys. N.Y. 194, 336-386 [In order to test
experimentally the linearity of quantum mechanics Stephen Weinberg assumed a non-linear model of
the Schrodinger's wave equation (see also Bialynicki-Birula I. and Mycielski J. 1976) and evaluated the
admissible non-linearity by calculating corrections to the energy spectra. Some general problems, like
e.g. non-existence of the orthodox measurement theory remained open. The possibilities of formulating
the non-linear QM for one particle states were soon limited by the theorem of Nicolas Gisin (c.f. Gisin
N. and Rigo M. 1995) showing the basic difficulty of extending the theory to many particle systems]

Weyl H. (1950) The theory of groups and quantum mechanics, New York: Dover Publications Inc.

Wheeler J., and Zurek W. (1983) Quantum theory and measurement, Princeton: Princeton University
Press [One of the most useful anthologies on the foundations, measurements and interpretation of
quantum theory. Apart of the original ideas of Wheeler and Zurek (on the "delayed choice
experiment") it contains the English translations of some classical papers, like e.g. (Schrodinger 1935,
Heisenberg 1927), reprinted versions of Bohm, Zeh and others]

65
Wigner E.P. (1962) Remarks on the mind-body question, in I.J. Good (Ed), The scientist speculates,
New York: Capricorn Books, Ch. 7.98 [A brilliant but perturbing paper in which Eugene P. Wigner
(Nobel Prize in Physics, 1963) returns to the hypothesis (London and Bauer 1939) that the reduction of
quantum states is caused by the consciousness of an intelligent observer. He then considers several
observers watching the same quantum effect, trying to distinguish between the subjective but collective
and solipsist "constructions of reality". He concludes at the end: "The present writer is well aware (...)
that he is not the first one to discuss the questions (...) and that the surmises of his predecessors were
either found to be wrong or improvable (...) He feels, however, that many of the earlier speculations on
the subject, even if they could not be justified, have stimulated and helped our thinking (...) in the
question which is, perhaps, the most fundamental question at all". (Reprinted also in Wheeler J. and
Zurek W. 1983)]

Wigner E.P. (1963) The problem of measurement, Am. J. Phys. 31, 6-15

Wineland D.J., and Itano W.M. (1979) Laser cooling of atoms, Phys. Rev. A 20, 1521-1540

Witten E (1995) String theory dynamics in various dimensions, Nucl. Phys. B 443, 85-126

Zeilinger A (2000, April) Quantum teleportation, Sci. Am., 32-41 [Anton Zeilinger, one of the best
known leaders of the experiments on quantum teleportation, an effect challenging our intuitions about
quantum states, confirmed in the Insbruck labs. The Insbruck-Viena group, founded in Nov.2003,
counts with a team of the world known specialists, headed by R. Blatt and R. Grimm (experimental
physics), H. Briegel and P. Zoller (theoretical physics as scientific directors), and a laboratory in
Vienna headed by A. Zeilinger (experimental physics). To be noticed is also their collaboration with
Juan Ignacio Cirac (Principe de Asturias Prize on Investigación Científica y Técnica 2006), C.f. e.g. I.
Cirac and P. Zoller (1995) Phys. Rev. Lett. 47, 4091]

66