Quantum Mechanics
:
A graduate level course
Richard Fitzpatrick
Associate Professor of Physics
The University of Texas at Austin
Contents
1 Introduction 5
1.1 Major sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Fundamental concepts 6
2.1 The breakdown of classical physics . . . . . . . . . . . . . . . . . . 6
2.2 The polarization of photons . . . . . . . . . . . . . . . . . . . . . . 7
2.3 The fundamental principles of quantum mechanics . . . . . . . . . 9
2.4 Ket space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.5 Bra space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.6 Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.7 The outer product . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.8 Eigenvalues and eigenvectors . . . . . . . . . . . . . . . . . . . . . 20
2.9 Observables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.10 Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.11 Expectation values . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.12 Degeneracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.13 Compatible observables . . . . . . . . . . . . . . . . . . . . . . . . 27
2.14 The uncertainty relation . . . . . . . . . . . . . . . . . . . . . . . . 28
2.15 Continuous spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3 Position and momentum 33
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2 Poisson brackets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.3 Wavefunctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.4 Schr¨ odinger’s representation  I . . . . . . . . . . . . . . . . . . . . 39
3.5 Schr¨ odinger’s representation  II . . . . . . . . . . . . . . . . . . . 43
3.6 The momentum representation . . . . . . . . . . . . . . . . . . . . 46
3.7 The uncertainty relation . . . . . . . . . . . . . . . . . . . . . . . . 48
3.8 Displacement operators . . . . . . . . . . . . . . . . . . . . . . . . 50
4 Quantum dynamics 55
4.1 Schr¨ odinger’s equations of motion . . . . . . . . . . . . . . . . . . 55
4.2 Heisenberg’s equations of motion . . . . . . . . . . . . . . . . . . . 59
2
4.3 Ehrenfest’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.4 Schr¨ odinger’s waveequation . . . . . . . . . . . . . . . . . . . . . 65
5 Angular momentum 71
5.1 Orbital angular momentum . . . . . . . . . . . . . . . . . . . . . . 71
5.2 Eigenvalues of angular momentum . . . . . . . . . . . . . . . . . . 74
5.3 Rotation operators . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.4 Eigenfunctions of orbital angular momentum . . . . . . . . . . . . 81
5.5 Motion in a central ﬁeld . . . . . . . . . . . . . . . . . . . . . . . . 84
5.6 Energy levels of the hydrogen atom . . . . . . . . . . . . . . . . . . 86
5.7 Spin angular momentum . . . . . . . . . . . . . . . . . . . . . . . 89
5.8 Wavefunction of a spin onehalf particle . . . . . . . . . . . . . . . 91
5.9 Rotation operators in spin space . . . . . . . . . . . . . . . . . . . 93
5.10 Magnetic moments . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5.11 Spin precession . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.12 Pauli twocomponent formalism . . . . . . . . . . . . . . . . . . . . 99
5.13 Spin greater than onehalf systems . . . . . . . . . . . . . . . . . . 105
5.14 Addition of angular momentum . . . . . . . . . . . . . . . . . . . . 110
6 Approximation methods 120
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
6.2 The twostate system . . . . . . . . . . . . . . . . . . . . . . . . . . 120
6.3 Nondegenerate perturbation theory . . . . . . . . . . . . . . . . . 122
6.4 The quadratic Stark effect . . . . . . . . . . . . . . . . . . . . . . . 124
6.5 Degenerate perturbation theory . . . . . . . . . . . . . . . . . . . . 129
6.6 The linear Stark effect . . . . . . . . . . . . . . . . . . . . . . . . . 132
6.7 Fine structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6.8 The Zeeman effect . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
6.9 Timedependent perturbation theory . . . . . . . . . . . . . . . . . 144
6.10 The twostate system . . . . . . . . . . . . . . . . . . . . . . . . . . 146
6.11 Spin magnetic resonance . . . . . . . . . . . . . . . . . . . . . . . 149
6.12 The Dyson series . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.13 Constant perturbations . . . . . . . . . . . . . . . . . . . . . . . . . 154
6.14 Harmonic perturbations . . . . . . . . . . . . . . . . . . . . . . . . 158
6.15 Absorption and stimulated emission of radiation . . . . . . . . . . 159
3
6.16 The electric dipole approximation . . . . . . . . . . . . . . . . . . . 162
6.17 Energyshifts and decaywidths . . . . . . . . . . . . . . . . . . . . 165
7 Scattering theory 170
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
7.2 The LipmannSchwinger equation . . . . . . . . . . . . . . . . . . 170
7.3 The Born approximation . . . . . . . . . . . . . . . . . . . . . . . . 175
7.4 Partial waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
7.5 The optical theorem . . . . . . . . . . . . . . . . . . . . . . . . . . 181
7.6 Determination of phaseshifts . . . . . . . . . . . . . . . . . . . . . 182
7.7 Hard sphere scattering . . . . . . . . . . . . . . . . . . . . . . . . . 184
7.8 Low energy scattering . . . . . . . . . . . . . . . . . . . . . . . . . 186
7.9 Resonances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
4
1 INTRODUCTION
1 Introduction
1.1 Major sources
The textbooks which I have consulted most frequently while developing course
material are:
The principles of quantum mechanics, P.A.M. Dirac, 4th Edition (revised), (Ox
ford University Press, Oxford, UK, 1958).
The Feynman lectures on physics, R.P. Feynman, R.B. Leighton, and M. Sands,
Volume III (AddisonWesley, Reading MA, 1965).
Quantum mechanics, E. Merzbacher, 2nd Edition (John Wiley & Sons, New York
NY, 1970).
Modern quantum mechanics, J.J. Sakurai, (Benjamin/Cummings, Menlo Park
CA, 1985).
5
2 FUNDAMENTAL CONCEPTS
2 Fundamental concepts
2.1 The breakdown of classical physics
The necessity for a departure from classical mechanics is clearly demonstrated
by:
1. The anomalous stability of atoms and molecules: According to classical physics,
an electron orbiting a nucleus should lose energy by emission of synchrotron
radiation, and gradually spiral in towards the nucleus. Experimentally, this
is not observed to happen.
2. The anomalously low speciﬁc heats of atoms and molecules: According to the
equipartition theorem of classical physics, each degree of freedom of an
atomic or molecular system should contribute R/2 to its molar speciﬁc heat,
where R is the ideal gas constant. In fact, only the translational and some
rotational degrees of freedom seem to contribute. The vibrational degrees
of freedom appear to make no contribution at all (except at high temper
atures). Incidentally, this fundamental problem with classical physics was
known and appreciated in the middle of the nineteenth century. Stories that
physicists at the start of the twentieth century thought that classical physics
explained everything, and that there was nothing left to discover, are largely
apocryphal (see Feynman, Vol. I, Cha. 40).
3. The ultraviolet catastrophe: According to classical physics, the energy density
of an electromagnetic ﬁeld in vacuum is inﬁnite due to a divergence of en
ergy carried by short wavelength modes. Experimentally, there is no such
divergence, and the total energy density is ﬁnite.
4. Waveparticle duality: Classical physics can deal with waves or particles. How
ever, various experiments (e.g., light interference, the photoelectric effect,
electron diffraction) show quite clearly that waves sometimes act as if they
were streams of particles, and streams of particles sometimes act as if they
were waves. This is completely inexplicable within the framework of classi
cal physics.
6
2.2 The polarization of photons 2 FUNDAMENTAL CONCEPTS
2.2 The polarization of photons
It is known experimentally that when plane polarized light is used to eject photo
electrons there is a preferred direction of emission of the electrons. Clearly, the
polarization properties of light, which are more usually associated with its wave
like behaviour, also extend to its particlelike behaviour. In particular, a polariza
tion can be ascribed to each individual photon in a beam of light.
Consider the following wellknown experiment. A beam of plane polarized
light is passed through a polaroid ﬁlm, which has the property that it is only
transparent to light whose plane of polarization lies perpendicular to its optic
axis. Classical electromagnetic wave theory tells us that if the beam is polarized
perpendicular to the optic axis then all of the light is transmitted, if the beam is
polarized parallel to the optic axis then none of the light is transmitted, and if the
light is polarized at an angle α to the axis then a fraction sin
2
α of the beam is
transmitted. Let us try to account for these observations at the individual photon
level.
A beam of light which is plane polarized in a certain direction is made up of a
stream of photons which are each plane polarized in that direction. This picture
leads to no difﬁculty if the plane of polarization lies parallel or perpendicular
to the optic axis of the polaroid. In the former case, none of the photons are
transmitted, and, in the latter case, all of the photons are transmitted. But, what
happens in the case of an obliquely polarized incident beam?
The above question is not very precise. Let us reformulate it as a question
relating to the result of some experiment which we could perform. Suppose that
we were to ﬁre a single photon at a polaroid ﬁlm, and then look to see whether
or not it emerges from the other side. The possible results of the experiment are
that either a whole photon, whose energy is equal to the energy of the incident
photon, is observed, or no photon is observed. Any photon which is transmitted
though the ﬁlm must be polarized perpendicular to the optic axis. Furthermore,
it is impossible to imagine (in physics) ﬁnding part of a photon on the other side
of the ﬁlm. If we repeat the experiment a great number of times then, on average,
a fraction sin
2
α of the photons are transmitted through the ﬁlm, and a fraction
7
2.2 The polarization of photons 2 FUNDAMENTAL CONCEPTS
cos
2
α are absorbed. Thus, we conclude that a photon has a probability sin
2
α of
being transmitted as a photon polarized in the plane perpendicular to the optic
axis, and a probability cos
2
α of being absorbed. These values for the probabilities
lead to the correct classical limit for a beamcontaining a large number of photons.
Note that we have only been able to preserve the individuality of photons,
in all cases, by abandoning the determinacy of classical theory, and adopting a
fundamentally probabilistic approach. We have no way of knowing whether an
individual obliquely polarized photon is going to be absorbed by or transmitted
through a polaroid ﬁlm. We only know the probability of each event occurring.
This is a fairly sweeping statement, but recall that the state of a photon is fully
speciﬁed once its energy, direction of propagation, and polarization are known.
If we imagine performing experiments using monochromatic light, normally in
cident on a polaroid ﬁlm, with a particular oblique polarization, then the state of
each individual photon in the beam is completely speciﬁed, and there is nothing
left over to uniquely determine whether the photon is transmitted or absorbed by
the ﬁlm.
The above discussion about the results of an experiment with a single obliquely
polarized photon incident on a polaroid ﬁlm answers all that can be legitimately
asked about what happens to the photon when it reaches the ﬁlm. Questions as
to what decides whether the photon is transmitted or not, or how it changes its
direction of polarization, are illegitimate, since they do not relate to the outcome
of a possible experiment. Nevertheless, some further description is needed in
order to allow the results of this experiment to be correlated with the results of
other experiments which can be performed using photons.
The further description provided by quantum mechanics is as follows. It is
supposed that a photon polarized obliquely to the optic axis can be regarded as
being partly in a state of polarization parallel to the axis, and partly in a state of
polarization perpendicular to the axis. In other words, the oblique polarization
state is some sort of superposition of two states of parallel and perpendicular
polarization. Since there is nothing special about the orientation of the optic
axis in our experiment, we must conclude that any state of polarization can be
regarded as a superposition of two mutually perpendicular states of polarization.
8
2.3 The fundamental principles of quantum mechanics 2 FUNDAMENTAL CONCEPTS
When we make the photon encounter a polaroid ﬁlm, we are subjecting it
to an observation. In fact, we are observing whether it is polarized parallel or
perpendicular to the optic axis. The effect of making this observation is to force
the photon entirely into a state of parallel or perpendicular polarization. In other
words, the photon has to jump suddenly from being partly in each of these two
states to being entirely in one or the other of them. Which of the two states it will
jump into cannot be predicted, but is governed by probability laws. If the photon
jumps into a state of parallel polarization then it is absorbed. Otherwise, it is
transmitted. Note that, in this example, the introduction of indeterminacy into
the problem is clearly connected with the act of observation. In other words, the
indeterminacy is related to the inevitable disturbance of the system associated
with the act of observation.
2.3 The fundamental principles of quantum mechanics
There is nothing special about the transmission and absorption of photons through
a polaroid ﬁlm. Exactly the same conclusions as those outlined above are ob
tained by studying other simple experiments, such as the interference of photons
(see Dirac, Sect. I.3), and the SternGerlach experiment (see Sakurai, Cha. 1;
Feynman, Cha. 5). The study of these simple experiments leads us to formulate
the following fundamental principles of quantum mechanics:
1. Dirac’s razor: Quantum mechanics can only answer questions regarding the
outcome of possible experiments. Any other questions lie beyond the realms
of physics.
2. The principle of superposition of states: Any microscopic system (i.e., an atom,
molecule, or particle) in a given state can be regarded as being partly in
each of two or more other states. In other words, any state can be regarded
as a superposition of two or more other states. Such superpositions can be
performed in an inﬁnite number of different ways.
3. The principle of indeterminacy: An observation made on a microscopic system
causes it to jump into one or more particular states (which are related to
9
2.4 Ket space 2 FUNDAMENTAL CONCEPTS
the type of observation). It is impossible to predict into which ﬁnal state
a particular system will jump, however the probability of a given system
jumping into a given ﬁnal state can be predicted.
The ﬁrst of these principles was formulated by quantum physicists (such as Dirac)
in the 1920s to fend off awkward questions such as “How can a system suddenly
jump from one state into another?”, or “How does a system decide which state to
jump into?”. As we shall see, the second principle is the basis for the mathemat
ical formulation of quantum mechanics. The ﬁnal principle is still rather vague.
We need to extend it so that we can predict which possible states a system can
jump into after a particular type of observation, as well as the probability of the
system making a particular jump.
2.4 Ket space
Consider a microscopic system composed of particles or bodies with speciﬁc prop
erties (mass, moment of inertia, etc.) interacting according to speciﬁc laws of
force. There will be various possible motions of the particles or bodies consistent
with the laws of force. Let us term each such motion a state of the system. Accord
ing to the principle of superposition of states, any given state can be regarded as
a superposition of two or more other states. Thus, states must be related to math
ematical quantities of a kind which can be added together to give other quantities
of the same kind. The most obvious examples of such quantities are vectors.
Let us consider a particular microscopic system in a particular state, which we
label A: e.g., a photon with a particular energy, momentum, and polarization.
We can represent this state as a particular vector, which we also label A, residing
in some vector space, where the other elements of the space represent all of the
other possible states of the system. Such a space is called a ket space (after Dirac).
The state vector A is conventionally written
A). (2.1)
Suppose that state A is, in fact, the superposition of two different states, B and
10
2.4 Ket space 2 FUNDAMENTAL CONCEPTS
C. This interrelation is represented in ket space by writing
A) = B) + C), (2.2)
where B) is the vector relating to the state B, etc. For instance, state B) might
represent a photon propagating in the zdirection, and plane polarized in the x
direction, and state C) might represent a similar photon plane polarized in the
ydirection. In this case, the sum of these two states represents a photon whose
plane of polarization makes an angle of 45
◦
with both the x and ydirections (by
analogy with classical physics). This latter state is represented by B) + C) in ket
space.
Suppose that we want to construct a state whose plane of polarization makes
an arbitrary angle α with the xdirection. We can do this via a suitably weighted
superposition of states B and C. By analogy with classical physics, we require
cos α of state B, and sinα of state C. This new state is represented by
cos αB) + sinαC) (2.3)
in ket space. Note that we cannot form a new state by superposing a state with
itself. For instance, a photon polarized in the ydirection superposed with another
photon polarized in the ydirection (with the same energy and momentum) gives
the same photon. This implies that the ket vector
c
1
A) +c
2
A) = (c
1
+c
2
)A) (2.4)
corresponds to the same state that A) does. Thus, ket vectors differ from con
ventional vectors in that their magnitudes, or lengths, are physically irrelevant.
All the states of the system are in one to one correspondence with all the possi
ble directions of vectors in the ket space, no distinction being made between the
directions of the ket vectors A) and −A). There is, however, one caveat to the
above statements. If c
1
+ c
2
= 0 then the superposition process yields nothing at
all: i.e., no state. The absence of a state is represented by the null vector 0) in
ket space. The null vector has the fairly obvious property that
A) + 0) = A), (2.5)
for any vector A). The fact that ket vectors pointing in the same direction repre
sent the same state relates ultimately to the quantization of matter: i.e., the fact
11
2.4 Ket space 2 FUNDAMENTAL CONCEPTS
that it comes in irreducible packets called photons, electrons, atoms, etc. If we ob
serve a microscopic system then we either see a state (i.e., a photon, or an atom,
or a molecule, etc.) or we see nothing—we can never see a fraction or a multiple
of a state. In classical physics, if we observe a wave then the amplitude of the
wave can take any value between zero and inﬁnity. Thus, if we were to represent
a classical wave by a vector, then the magnitude, or length, of the vector would
correspond to the amplitude of the wave, and the direction would correspond to
the frequency and wavelength, so that two vectors of different lengths pointing
in the same direction would represent different wave states.
We have seen, in Eq. (2.3), that any plane polarized state of a photon can
be represented as a linear superposition of two orthogonal polarization states
in which the weights are real numbers. Suppose that we want to construct a
circularly polarized photon state. Well, we know from classical physics that a cir
cularly polarized wave is a superposition of two waves of equal amplitude, plane
polarized in orthogonal directions, which are in phase quadrature. This suggests
that a circularly polarized photon is the superposition of a photon polarized in
the xdirection (state B) and a photon polarized in the ydirection (state C), with
equal weights given to the two states, but with the proviso that state C is 90
◦
out of phase with state B. By analogy with classical physics, we can use complex
numbers to simultaneously represent the weighting and relative phase in a linear
superposition. Thus, a circularly polarized photon is represented by
B) + i C) (2.6)
in ket space. A general elliptically polarized photon is represented by
c
1
B) +c
2
C), (2.7)
where c
1
and c
2
are complex numbers. We conclude that a ket space must be
a complex vector space if it is to properly represent the mutual interrelations
between the possible states of a microscopic system.
Suppose that the ket R) is expressible linearly in terms of the kets A) and B),
so that
R) = c
1
A) +c
2
B). (2.8)
12
2.4 Ket space 2 FUNDAMENTAL CONCEPTS
We say that R) is dependent on A) and B). It follows that the state R can be
regarded as a linear superposition of the states A and B. So, we can also say that
state R is dependent on states A and B. In fact, any ket vector (or state) which
is expressible linearly in terms of certain others is said to be dependent on them.
Likewise, a set of ket vectors (or states) are termed independent if none of them
are expressible linearly in terms of the others.
The dimensionality of a conventional vector space is deﬁned as the number
of independent vectors contained in the space. Likewise, the dimensionality of
a ket space is equivalent to the number of independent ket vectors it contains.
Thus, the ket space which represents the possible polarization states of a photon
propagating in the zdirection is twodimensional (the two independent vectors
correspond to photons plane polarized in the x and ydirections, respectively).
Some microscopic systems have a ﬁnite number of independent states (e.g., the
spin states of an electron in a magnetic ﬁeld). If there are N independent states,
then the possible states of the system are represented as an Ndimensional ket
space. Some microscopic systems have a denumerably inﬁnite number of inde
pendent states (e.g., a particle in an inﬁnitely deep, onedimensional potential
well). The possible states of such a system are represented as a ket space whose
dimensions are denumerably inﬁnite. Such a space can be treated in more or less
the same manner as a ﬁnitedimensional space. Unfortunately, some microscopic
systems have a nondenumerably inﬁnite number of independent states (e.g., a
free particle). The possible states of such a system are represented as a ket space
whose dimensions are nondenumerably inﬁnite. This type of space requires a
slightly different treatment to spaces of ﬁnite, or denumerably inﬁnite, dimen
sions.
In conclusion, the states of a general microscopic system can be represented as
a complex vector space of (possibly) inﬁnite dimensions. Such a space is termed
a Hilbert space by mathematicians.
13
2.5 Bra space 2 FUNDAMENTAL CONCEPTS
2.5 Bra space
A snack machine inputs coins plus some code entered on a key pad, and (hope
fully) outputs a snack. It also does so in a deterministic manner: i.e., the same
money plus the same code produces the same snack (or the same error message)
time after time. Note that the input and output of the machine have completely
different natures. We can imagine building a rather abstract snack machine which
inputs ket vectors and outputs complex numbers in a deterministic fashion. Math
ematicians call such a machine a functional. Imagine a general functional, labeled
F, acting on a general ket vector, labeled A, and spitting out a general complex
number φ
A
. This process is represented mathematically by writing
¸F(A)) = φ
A
. (2.9)
Let us narrow our focus to those functionals which preserve the linear dependen
cies of the ket vectors upon which they operate. Not surprisingly, such functionals
are termed linear functionals. A general linear functional, labeled F, satisﬁes
¸F(A) + B)) = ¸F(A)) +¸F(B)), (2.10)
where A) and B) are any two kets in a given ket space.
Consider an Ndimensional ket space [i.e., a ﬁnitedimensional, or denumer
ably inﬁnite dimensional (i.e., N → ∞), space]. Let the i) (where i runs from 1
to N) represent N independent ket vectors in this space. A general ket vector can
be written
1
A) =
N
i=1
α
i
i), (2.11)
where the α
i
are an arbitrary set of complex numbers. The only way the func
tional F can satisfy Eq. (2.10) for all vectors in the ket space is if
¸F(A)) =
N
i=1
f
i
α
i
, (2.12)
1
Actually, this is only strictly true for ﬁnitedimensional spaces. Only a special subset of denumerably inﬁnite
dimensional spaces have this property (i.e., they are complete), but since a ket space must be complete if it is to
represent the states of a microscopic system, we need only consider this special subset.
14
2.5 Bra space 2 FUNDAMENTAL CONCEPTS
where the f
i
are a set of complex numbers relating to the functional.
Let us deﬁne N basis functionals ¸i which satisfy
¸i(j)) = δ
ij
. (2.13)
It follows from the previous three equations that
¸F =
N
i=1
f
i
¸i. (2.14)
But, this implies that the set of all possible linear functionals acting on an N
dimensional ket space is itself an Ndimensional vector space. This type of vector
space is called a bra space (after Dirac), and its constituent vectors (which are
actually functionals of the ket space) are called bra vectors. Note that bra vectors
are quite different in nature to ket vectors (hence, these vectors are written in
mirror image notation, ¸  and  ), so that they can never be confused). Bra
space is an example of what mathematicians call a dual vector space (i.e., it is
dual to the original ket space). There is a one to one correspondence between
the elements of the ket space and those of the related bra space. So, for every
element A of the ket space, there is a corresponding element, which it is also
convenient to label A, in the bra space. That is,
A)
DC
←→¸A, (2.15)
where DC stands for dual correspondence.
There are an inﬁnite number of ways of setting up the correspondence between
vectors in a ket space and those in the related bra space. However, only one
of these has any physical signiﬁcance. For a general ket vector A, speciﬁed by
Eq. (2.11), the corresponding bra vector is written
¸A =
N
i=1
α
∗
i
¸i, (2.16)
where the α
∗
i
are the complex conjugates of the α
i
. ¸A is termed the dual vector
to A). It follows, from the above, that the dual to c¸A is c
∗
A), where c is a
complex number. More generally,
c
1
A) +c
2
B)
DC
←→c
∗
1
¸A +c
∗
2
¸B. (2.17)
15
2.5 Bra space 2 FUNDAMENTAL CONCEPTS
Recall that a bra vector is a functional which acts on a general ket vector, and
spits out a complex number. Consider the functional which is dual to the ket
vector
B) =
N
i=1
β
i
i) (2.18)
acting on the ket vector A). This operation is denoted ¸B(A)). Note, however,
that we can omit the round brackets without causing any ambiguity, so the oper
ation can also be written ¸BA). This expression can be further simpliﬁed to give
¸BA). According to Eqs. (2.11), (2.12), (2.16), and (2.18),
¸BA) =
N
i=1
β
∗
i
α
i
. (2.19)
Mathematicians term ¸BA) the inner product of a bra and a ket.
2
An inner prod
uct is (almost) analogous to a scalar product between a covariant and contravari
ant vector in some curvilinear space. It is easily demonstrated that
¸BA) = ¸AB)
∗
. (2.20)
Consider the special case where B) →A). It follows from Eqs. (2.12) and (2.20)
that ¸AA) is a real number, and that
¸AA) ≥ 0. (2.21)
The equality sign only holds if A) is the null ket [i.e., if all of the α
i
are zero in
Eq. (2.11)]. This property of bra and ket vectors is essential for the probabilistic
interpretation of quantum mechanics, as will become apparent later.
Two kets A) and B) are said to be orthogonal if
¸AB) = 0, (2.22)
which also implies that ¸BA) = 0.
Given a ket A) which is not the null ket, we can deﬁne a normalized ket 
˜
A),
where

˜
A) =
_
_
1
_
¸AA)
_
_
A), (2.23)
2
We can now appreciate the elegance of Dirac’s notation. The combination of a bra and a ket yields a “bra(c)ket”
(which is just a number).
16
2.6 Operators 2 FUNDAMENTAL CONCEPTS
with the property
¸
˜
A
˜
A) = 1. (2.24)
Here,
_
¸AA) is known as the norm or “length” of A), and is analogous to the
length, or magnitude, of a conventional vector. Since A) and cA) represent
the same physical state, it makes sense to require that all kets corresponding to
physical states have unit norms.
It is possible to deﬁne a dual bra space for a ket space of nondenumerably
inﬁnite dimensions in much the same manner as that described above. The main
differences are that summations over discrete labels become integrations over
continuous labels, Kronecker deltafunctions become Dirac deltafunctions, com
pleteness must be assumed (it cannot be proved), and the normalization conven
tion is somewhat different. More of this later.
2.6 Operators
We have seen that a functional is a machine which inputs a ket vector and spits
out a complex number. Consider a somewhat different machine which inputs a
ket vector and spits out another ket vector in a deterministic fashion. Mathemati
cians call such a machine an operator. We are only interested in operators which
preserve the linear dependencies of the ket vectors upon which they act. Such
operators are termed linear operators. Consider an operator labeled X. Suppose
that when this operator acts on a general ket vector A) it spits out a new ket
vector which is denoted XA). Operator X is linear provided that
X(A) + B)) = XA) +XB), (2.25)
for all ket vectors A) and B), and
X(cA)) = cXA), (2.26)
for all complex numbers c. Operators X and Y are said to be equal if
XA) = YA) (2.27)
17
2.6 Operators 2 FUNDAMENTAL CONCEPTS
for all kets in the ket space in question. Operator X is termed the null operator if
XA) = 0) (2.28)
for all ket vectors in the space. Operators can be added together. Such addition
is deﬁned to obey a commutative and associate algebra:
X +Y = Y +X, (2.29)
X + (Y +Z) = (X +Y) +Z. (2.30)
Operators can also be multiplied. The multiplication is associative:
X(YA)) = (XY)A) = XYA), (2.31)
X(Y Z) = (XY)Z = XY Z. (2.32)
However, in general, it is noncommutative:
XY ,= Y X. (2.33)
So far, we have only considered linear operators acting on ket vectors. We can
also give a meaning to their operating on bra vectors. Consider the inner product
of a general bra ¸B with the ket XA). This product is a number which depends
linearly on A). Thus, it may be considered to be the inner product of A) with
some bra. This bra depends linearly on ¸B, so we may look on it as the result of
some linear operator applied to ¸B. This operator is uniquely determined by the
original operator X, so we might as well call it the same operator acting on B). A
suitable notation to use for the resulting bra when X operates on ¸B is ¸BX. The
equation which deﬁnes this vector is
(¸BX)A) = ¸B(XA)) (2.34)
for any A) and ¸B. The triple product of ¸B, X, and A) can be written ¸BXA)
without ambiguity, provided we adopt the convention that the bra vector always
goes on the left, the operator in the middle, and the ket vector on the right.
Consider the dual bra to XA). This bra depends antilinearly on A) and must
therefore depend linearly on ¸A. Thus, it may be regarded as the result of some
18
2.7 The outer product 2 FUNDAMENTAL CONCEPTS
linear operator applied to ¸A. This operator is termed the adjoint of X, and is
denoted X
†
. Thus,
XA)
DC
←→¸AX
†
. (2.35)
It is readily demonstrated that
¸BX
†
A) = ¸AXB)
∗
, (2.36)
plus
(XY)
†
= Y
†
X
†
. (2.37)
It is also easily seen that the adjoint of the adjoint of a linear operator is equiva
lent to the original operator. A Hermitian operator ξ has the special property that
it is its own adjoint: i.e.,
ξ = ξ
†
. (2.38)
2.7 The outer product
So far we have formed the following products: ¸BA), XA), ¸AX, XY, ¸BXA).
Are there any other products we are allowed to form? How about
B)¸A ? (2.39)
This clearly depends linearly on the ket A) and the bra B). Suppose that we
rightmultiply the above product by the general ket C). We obtain
B)¸AC) = ¸AC)B), (2.40)
since ¸AC) is just a number. Thus, B)¸A acting on a general ket C) yields
another ket. Clearly, the product B)¸A is a linear operator. This operator also
acts on bras, as is easily demonstrated by leftmultiplying the expression (2.39)
by a general bra ¸C. It is also easily demonstrated that
(B)¸A)
†
= A)¸B. (2.41)
Mathematicians term the operator B)¸A the outer product of B) and ¸A. The
outer product should not be confused with the inner product, ¸AB), which is just
a number.
19
2.8 Eigenvalues and eigenvectors 2 FUNDAMENTAL CONCEPTS
2.8 Eigenvalues and eigenvectors
In general, the ket XA) is not a constant multiple of A). However, there are
some special kets known as the eigenkets of operator X. These are denoted
x
), x
), x
) . . . , (2.42)
and have the property
Xx
) = x
x
), Xx
) = x
x
) . . . , (2.43)
where x
, x
, . . . are numbers called eigenvalues. Clearly, applying X to one of its
eigenkets yields the same eigenket multiplied by the associated eigenvalue.
Consider the eigenkets and eigenvalues of a Hermitian operator ξ. These are
denoted
ξξ
) = ξ
ξ
), (2.44)
where ξ
) is the eigenket associated with the eigenvalue ξ
. Three important
results are readily deduced:
(i) The eigenvalues are all real numbers, and the eigenkets corresponding to
different eigenvalues are orthogonal. Since ξ is Hermitian, the dual equation to
Eq. (2.44) (for the eigenvalue ξ
) reads
¸ξ
ξ = ξ
∗
¸ξ
. (2.45)
If we leftmultiply Eq. (2.44) by ¸ξ
, rightmultiply the above equation by ξ
),
and take the difference, we obtain
(ξ
−ξ
∗
)¸ξ
ξ
) = 0. (2.46)
Suppose that the eigenvalues ξ
and ξ
are the same. It follows from the above
that
ξ
= ξ
∗
, (2.47)
where we have used the fact that ξ
) is not the null ket. This proves that the
eigenvalues are real numbers. Suppose that the eigenvalues ξ
and ξ
are differ
ent. It follows that
¸ξ
ξ
) = 0, (2.48)
20
2.9 Observables 2 FUNDAMENTAL CONCEPTS
which demonstrates that eigenkets corresponding to different eigenvalues are
orthogonal.
(ii) The eigenvalues associated with eigenkets are the same as the eigenvalues
associated with eigenbras. An eigenbra of ξ corresponding to an eigenvalue ξ
is
deﬁned
¸ξ
ξ = ¸ξ
ξ
. (2.49)
(iii) The dual of any eigenket is an eigenbra belonging to the same eigenvalue,
and conversely.
2.9 Observables
We have developed a mathematical formalism which comprises three types of
objects—bras, kets, and linear operators. We have already seen that kets can be
used to represent the possible states of a microscopic system. However, there is
a one to one correspondence between the elements of a ket space and its dual
bra space, so we must conclude that bras could just as well be used to repre
sent the states of a microscopic system. What about the dynamical variables of
the system (e.g., its position, momentum, energy, spin, etc.)? How can these be
represented in our formalism? Well, the only objects we have left over are oper
ators. We, therefore, assume that the dynamical variables of a microscopic system
are represented as linear operators acting on the bras and kets which correspond to
the various possible states of the system. Note that the operators have to be linear,
otherwise they would, in general, spit out bras/kets pointing in different direc
tions when fed bras/kets pointing in the same direction but differing in length.
Since the lengths of bras and kets have no physical signiﬁcance, it is reasonable
to suppose that nonlinear operators are also without physical signiﬁcance.
We have seen that if we observe the polarization state of a photon, by placing
a polaroid ﬁlm in its path, the result is to cause the photon to jump into a state
of polarization parallel or perpendicular to the optic axis of the ﬁlm. The former
state is absorbed, and the latter state is transmitted (which is how we tell them
apart). In general, we cannot predict into which state a given photon will jump
21
2.9 Observables 2 FUNDAMENTAL CONCEPTS
(except in a statistical sense). However, we do know that if the photon is initially
polarized parallel to the optic axis then it will deﬁnitely be absorbed, and if it is
initially polarized perpendicular to the axis then it will deﬁnitely be transmitted.
We also known that after passing though the ﬁlm a photon must be in a state of
polarization perpendicular to the optic axis (otherwise it would not have been
transmitted). We can make a second observation of the polarization state of
such a photon by placing an identical polaroid ﬁlm (with the same orientation of
the optic axis) immediately behind the ﬁrst ﬁlm. It is clear that the photon will
deﬁnitely be transmitted through the second ﬁlm.
There is nothing special about the polarization states of a photon. So, more
generally, we can say that when a dynamical variable of a microscopic system
is measured the system is caused to jump into one of a number of independent
states (note that the perpendicular and parallel polarization states of our photon
are linearly independent). In general, each of these ﬁnal states is associated with
a different result of the measurement: i.e., a different value of the dynamical
variable. Note that the result of the measurement must be a real number (there
are no measurement machines which output complex numbers). Finally, if an
observation is made, and the system is found to be a one particular ﬁnal state,
with one particular value for the dynamical variable, then a second observation,
made immediately after the ﬁrst one, will deﬁnitely ﬁnd the system in the same
state, and yield the same value for the dynamical variable.
How can we represent all of these facts in our mathematical formalism? Well,
by a fairly nonobvious leap of intuition, we are going to assert that a measure
ment of a dynamical variable corresponding to an operator X in ket space causes
the system to jump into a state corresponding to one of the eigenkets of X. Not
surprisingly, such a state is termed an eigenstate. Furthermore, the result of the
measurement is the eigenvalue associated with the eigenket into which the system
jumps. The fact that the result of the measurement must be a real number implies
that dynamical variables can only be represented by Hermitian operators (since only
Hermitian operators are guaranteed to have real eigenvalues). The fact that the
eigenkets of a Hermitian operator corresponding to different eigenvalues (i.e., dif
ferent results of the measurement) are orthogonal is in accordance with our ear
lier requirement that the states into which the system jumps should be mutually
22
2.9 Observables 2 FUNDAMENTAL CONCEPTS
independent. We can conclude that the result of a measurement of a dynamical
variable represented by a Hermitian operator ξ must be one of the eigenvalues of
ξ. Conversely, every eigenvalue of ξ is a possible result of a measurement made
on the corresponding dynamical variable. This gives us the physical signiﬁcance
of the eigenvalues. (From now on, the distinction between a state and its rep
resentative ket vector, and a dynamical variable and its representative operator,
will be dropped, for the sake of simplicity.)
It is reasonable to suppose that if a certain dynamical variable ξ is measured
with the system in a particular state, then the states into which the system may
jump on account of the measurement are such that the original state is dependent
on them. This fairly innocuous statement has two very important corollaries.
First, immediately after an observation whose result is a particular eigenvalue ξ
,
the system is left in the associated eigenstate. However, this eigenstate is orthog
onal to (i.e., independent of) any other eigenstate corresponding to a different
eigenvalue. It follows that a second measurement made immediately after the
ﬁrst one must leave the system in an eigenstate corresponding to the eigenvalue
ξ
. In other words, the second measurement is bound to give the same result as
the ﬁrst. Furthermore, if the system is in an eigenstate of ξ, corresponding to an
eigenvalue ξ
, then a measurement of ξ is bound to give the result ξ
. This follows
because the system cannot jump into an eigenstate corresponding to a different
eigenvalue of ξ, since such a state is not dependent on the original state. Second,
it stands to reason that a measurement of ξ must always yield some result. It fol
lows that no matter what the initial state of the system, it must always be able to
jump into one of the eigenstates of ξ. In other words, a general ket must always
be dependent on the eigenkets of ξ. This can only be the case if the eigenkets
form a complete set (i.e., they span ket space). Thus, in order for a Hermitian oper
ator ξ to be observable its eigenkets must form a complete set. A Hermitian operator
which satisﬁes this condition is termed an observable. Conversely, any observable
quantity must be a Hermitian operator with a complete set of eigenstates.
23
2.10 Measurements 2 FUNDAMENTAL CONCEPTS
2.10 Measurements
We have seen that a measurement of some observable ξ of a microscopic system
causes the system to jump into one of the eigenstates of ξ. The result of the
measurement is the associated eigenvalue (or some function of this quantity). It
is impossible to determine into which eigenstate a given system will jump, but it is
possible to predict the probability of such a transition. So, what is the probability
that a system in some initial state A) makes a transition to an eigenstate ξ
) of an
observable ξ, as a result of a measurement made on the system? Let us start with
the simplest case. If the system is initially in an eigenstate ξ
) then the transition
probability to a eigenstate ξ
) corresponding to a different eigenvalue is zero,
and the transition probability to the same eigenstate ξ
) is unity. It is convenient
to normalize our eigenkets such that they all have unit norms. It follows from the
orthogonality property of the eigenkets that
¸ξ
ξ
) = δ
ξ
ξ
, (2.50)
where δ
ξ
ξ
is unity if ξ
= ξ
, and zero otherwise. For the moment, we are
assuming that the eigenvalues of ξ are all different.
Note that the probability of a transition from an initial eigenstate ξ
) to a ﬁ
nal eigenstate ξ
) is the same as the value of the inner product ¸ξ
ξ
). Can we
use this correspondence to obtain a general rule for calculating transition prob
abilities? Well, suppose that the system is initially in a state A) which is not an
eigenstate of ξ. Can we identify the transition probability to a ﬁnal eigenstate
ξ
) with the inner product ¸Aξ
)? The straight answer is “no”, since ¸Aξ
) is, in
general, a complex number, and complex probabilities do not make much sense.
Let us try again. How about if we identify the transition probability with the mod
ulus squared of the inner product, ¸Aξ
)
2
? This quantity is deﬁnitely a positive
number (so it could be a probability). This guess also gives the right answer for
the transition probabilities between eigenstates. In fact, it is the correct guess.
Since the eigenstates of an observable ξ form a complete set, we can express
any given state A) as a linear combination of them. It is easily demonstrated that
A) =
ξ
ξ
)¸ξ
A), (2.51)
24
2.11 Expectation values 2 FUNDAMENTAL CONCEPTS
¸A =
ξ
¸Aξ
)¸ξ
, (2.52)
¸AA) =
ξ
¸Aξ
)¸ξ
A) =
ξ
¸Aξ
)
2
, (2.53)
where the summation is over all the different eigenvalues of ξ, and use has been
made of Eq. (2.20), and the fact that the eigenstates are mutually orthogonal.
Note that all of the above results follow from the extremely useful (and easily
proved) result
ξ
ξ
)¸ξ
 = 1, (2.54)
where 1 denotes the identity operator. The relative probability of a transition to
an eigenstate ξ
), which is equivalent to the relative probability of a measure
ment of ξ yielding the result ξ
, is
P(ξ
) ∝ ¸Aξ
)
2
. (2.55)
The absolute probability is clearly
P(ξ
) =
¸Aξ
)
2
ξ
¸Aξ
)
2
=
¸Aξ
)
2
¸AA)
. (2.56)
If the ket A) is normalized such that its norm is unity, then this probability simply
reduces to
P(ξ
) = ¸Aξ
)
2
. (2.57)
2.11 Expectation values
Consider an ensemble of microscopic systems prepared in the same initial state
A). Suppose a measurement of the observable ξ is made on each system. We
know that each measurement yields the value ξ
with probability P(ξ
). What is
the mean value of the measurement? This quantity, which is generally referred
to as the expectation value of ξ, is given by
¸ξ) =
ξ
ξ
P(ξ
) =
ξ
ξ
¸Aξ
)
2
25
2.12 Degeneracy 2 FUNDAMENTAL CONCEPTS
=
ξ
ξ
¸Aξ
)¸ξ
A) =
ξ
¸Aξξ
)¸ξ
A), (2.58)
which reduces to
¸ξ) = ¸AξA) (2.59)
with the aid of Eq. (2.54).
Consider the identity operator, 1. All states are eigenstates of this operator
with the eigenvalue unity. Thus, the expectation value of this operator is always
unity: i.e.,
¸A1A) = ¸AA) = 1, (2.60)
for all A). Note that it is only possible to normalize a given ket A) such that
Eq. (2.60) is satisﬁed because of the more general property (2.21) of the norm.
This property depends on the particular correspondence (2.16), that we adopted
earlier, between the elements of a ket space and those of its dual bra space.
2.12 Degeneracy
Suppose that two different eigenstates ξ
a
) and ξ
b
) of ξ correspond to the same
eigenvalue ξ
. These are termed degenerate eigenstates. Degenerate eigenstates
are necessarily orthogonal to any eigenstates corresponding to different eigen
values, but, in general, they are not orthogonal to each other (i.e., the proof of
orthogonality given in Sect. 2.8 does not work in this case). This is unfortunate,
since much of the previous formalism depends crucially on the mutual orthogo
nality of the different eigenstates of an observable. Note, however, that any linear
combination of ξ
a
) and ξ
b
) is also an eigenstate corresponding to the eigenvalue
ξ
. It follows that we can always construct two mutually orthogonal degenerate
eigenstates. For instance,
ξ
1
) = ξ
a
), (2.61)
ξ
2
) =
ξ
b
) −¸ξ
a
ξ
b
)ξ
a
)
1 − ¸ξ
a
ξ
b
)
2
. (2.62)
26
2.13 Compatible observables 2 FUNDAMENTAL CONCEPTS
This result is easily generalized to the case of more than two degenerate eigen
states. We conclude that it is always possible to construct a complete set of mu
tually orthogonal eigenstates for any given observable.
2.13 Compatible observables
Suppose that we wish to simultaneously measure two observables, ξ and η, of
a microscopic system? Let us assume that we possess an apparatus which is ca
pable of measuring ξ, and another which can measure η. For instance, the two
observables in question might be the projection in the x and zdirections of the
spin angular momentum of a spin onehalf particle. These could be measured us
ing appropriate SternGerlach apparatuses (see Sakurai, Sect. 1.1). Suppose that
we make a measurement of ξ, and the system is consequently thrown into one
of the eigenstates of ξ, ξ
), with eigenvalue ξ
. What happens if we now make
a measurement of η? Well, suppose that the eigenstate ξ
) is also an eigenstate
of η, with eigenvalue η
. In this case, a measurement of η will deﬁnitely give the
result η
. A second measurement of ξ will deﬁnitely give the result ξ
, and so on.
In this sense, we can say that the observables ξ and η simultaneously have the
values ξ
and η
, respectively. Clearly, if all eigenstates of ξ are also eigenstates
of η then it is always possible to make a simultaneous measurement of ξ and η.
Such observables are termed compatible.
Suppose, however, that the eigenstates of ξ are not eigenstates of η. Is it
still possible to measure both observables simultaneously? Let us again make an
observation of ξ which throws the system into an eigenstate ξ
), with eigenvalue
ξ
. We can now make a second observation to determine η. This will throw
the system into one of the (many) eigenstates of η which depend on ξ
). In
principle, each of these eigenstates is associated with a different result of the
measurement. Suppose that the system is thrown into an eigenstate η
), with
the eigenvalue η
. Another measurement of ξ will throw the system into one
of the (many) eigenstates of ξ which depend on η
). Each eigenstate is again
associated with a different possible result of the measurement. It is clear that if
the observables ξ and η do not possess simultaneous eigenstates then if the value
27
2.14 The uncertainty relation 2 FUNDAMENTAL CONCEPTS
of ξ is known (i.e., the system is in an eigenstate of ξ) then the value of η is
uncertain (i.e., the system is not in an eigenstate of η), and vice versa. We say
that the two observables are incompatible.
We have seen that the condition for two observables ξ and η to be simultane
ously measurable is that they should possess simultaneous eigenstates (i.e., every
eigenstate of ξ should also be an eigenstate of η). Suppose that this is the case.
Let a general eigenstate of ξ, with eigenvalue ξ
, also be an eigenstate of η, with
eigenvalue η
. It is convenient to denote this simultaneous eigenstate ξ
η
). We
have
ξξ
η
) = ξ
ξ
η
), (2.63)
ηξ
η
) = η
ξ
η
). (2.64)
We can leftmultiply the ﬁrst equation by η, and the second equation by ξ, and
then take the difference. The result is
(ξ η −ηξ)ξ
η
) = 0) (2.65)
for each simultaneous eigenstate. Recall that the eigenstates of an observable
must form a complete set. It follows that the simultaneous eigenstates of two
observables must also form a complete set. Thus, the above equation implies that
(ξ η −ηξ)A) = 0), (2.66)
where A) is a general ket. The only way that this can be true is if
ξ η = ηξ. (2.67)
Thus, the condition for two observables ξ and η to be simultaneously measurable is
that they should commute.
2.14 The uncertainty relation
We have seen that if ξ and η are two noncommuting observables, then a deter
mination of the value of ξ leaves the value of η uncertain, and vice versa. It is
28
2.14 The uncertainty relation 2 FUNDAMENTAL CONCEPTS
possible to quantify this uncertainty. For a general observable ξ, we can deﬁne a
Hermitian operator
∆ξ = ξ −¸ξ), (2.68)
where the expectation value is taken over the particular physical state under con
sideration. It is obvious that the expectation value of ∆ξ is zero. The expectation
value of (∆ξ)
2
≡ ∆ξ ∆ξ is termed the variance of ξ, and is, in general, nonzero.
In fact, it is easily demonstrated that
¸(∆ξ)
2
) = ¸ξ
2
) −¸ξ)
2
. (2.69)
The variance of ξ is a measure of the uncertainty in the value of ξ for the particu
lar state in question (i.e., it is a measure of the width of the distribution of likely
values of ξ about the expectation value). If the variance is zero then there is no
uncertainty, and a measurement of ξ is bound to give the expectation value, ¸ξ).
Consider the Schwarz inequality
¸AA)¸BB) ≥ ¸AB)
2
, (2.70)
which is analogous to
a
2
 b
2
≥ a b
2
(2.71)
in Euclidian space. This inequality can be proved by noting that
(¸A +c
∗
¸B)(A) +cB)) ≥ 0, (2.72)
where c is any complex number. If c takes the special value −¸BA)/¸BB) then
the above inequality reduces to
¸AA)¸BB) − ¸AB)
2
≥ 0, (2.73)
which is the same as the Schwarz inequality.
Let us substitute
A) = ∆ξ ), (2.74)
B) = ∆η ), (2.75)
29
2.14 The uncertainty relation 2 FUNDAMENTAL CONCEPTS
into the Schwarz inequality, where the blank ket  ) stands for any general ket.
We ﬁnd
¸(∆ξ)
2
)¸(∆η)
2
) ≥ ¸∆ξ ∆η)
2
, (2.76)
where use has been made of the fact that ∆ξ and ∆η are Hermitian operators.
Note that
∆ξ ∆η =
1
2
[∆ξ, ∆η] +
1
2
{∆ξ, ∆η} , (2.77)
where the commutator, [∆ξ, ∆η], and the anticommutator, {∆ξ, ∆η}, are deﬁned
[∆ξ, ∆η] ≡ ∆ξ ∆η −∆η∆ξ, (2.78)
{∆ξ, ∆η} ≡ ∆ξ ∆η +∆η∆ξ. (2.79)
The commutator is clearly antiHermitian,
([∆ξ, ∆η])
†
= (∆ξ ∆η −∆η∆ξ)
†
= ∆η∆ξ −∆ξ ∆η = − [∆ξ, ∆η] , (2.80)
whereas the anticommutator is obviously Hermitian. Now, it is easily demon
strated that the expectation value of a Hermitian operator is a real number,
whereas the expectation value of an antiHermitian operator is a pure imaginary
number. It is clear that the right hand side of
¸∆ξ ∆η) =
1
2
¸[∆ξ, ∆η]) +
1
2
¸{∆ξ, ∆η}), (2.81)
consists of the sum of a purely real and a purely imaginary number. Taking the
modulus squared of both sides gives
¸∆ξ ∆η)
2
=
1
4
¸[ξ, η])
2
+
1
4
¸{∆ξ, ∆η})
2
, (2.82)
where use has been made of ¸∆ξ) = 0, etc. The ﬁnal term in the above expression
is positive deﬁnite, so we can write
¸(∆ξ)
2
)¸(∆η)
2
) ≥
1
4
¸[ξ, η])
2
, (2.83)
where use has been made of Eq. (2.76). The above expression is termed the
uncertainty relation. According to this relation, an exact knowledge of the value
of ξ implies no knowledge whatsoever of the value of η, and vice versa. The one
exception to this rule is when ξ and η commute, in which case exact knowledge
of ξ does not necessarily imply no knowledge of η.
30
2.15 Continuous spectra 2 FUNDAMENTAL CONCEPTS
2.15 Continuous spectra
Up to now, we have studiously avoided dealing with observables possessing eigen
values which lie in a continuous range, rather than having discrete values. The
reason for this is because continuous eigenvalues imply a ket space of nonde
numerably inﬁnite dimension. Unfortunately, continuous eigenvalues are un
avoidable in quantum mechanics. In fact, the most important observables of all,
namely position and momentum, generally have continuous eigenvalues. Fortu
nately, many of the results we obtained previously for a ﬁnitedimensional ket
space with discrete eigenvalues can be generalized to ket spaces of nondenumer
ably inﬁnite dimensions.
Suppose that ξ is an observable with continuous eigenvalues. We can still
write the eigenvalue equation as
ξξ
) = ξ
ξ
). (2.84)
But, ξ
can now take a continuous range of values. Let us assume, for the sake of
simplicity, that ξ
can take any value. The orthogonality condition (2.50) gener
alizes to
¸ξ
ξ
) = δ(ξ
−ξ
), (2.85)
where δ(x) denotes the famous Dirac deltafunction. Note that there are clearly a
nondenumerably inﬁnite number of mutually orthogonal eigenstates of ξ. Hence,
the dimensionality of ket space is nondenumerably inﬁnite. Note, also, that eigen
states corresponding to a continuous range of eigenvalues cannot be normalized
so that they have unit norms. In fact, these eigenstates have inﬁnite norms: i.e.,
they are inﬁnitely long. This is the major difference between eigenstates in a
ﬁnitedimensional and an inﬁnitedimensional ket space. The extremely useful
relation (2.54) generalizes to
_
dξ
ξ
)¸ξ
 = 1. (2.86)
Note that a summation over discrete eigenvalues goes over into an integral over
a continuous range of eigenvalues. The eigenstates ξ
) must form a complete set
if ξ is to be an observable. It follows that any general ket can be expanded in
31
2.15 Continuous spectra 2 FUNDAMENTAL CONCEPTS
terms of the ξ
). In fact, the expansions (2.51)–(2.53) generalize to
A) =
_
dξ
ξ
)¸ξ
A), (2.87)
¸A =
_
dξ
¸Aξ
)¸ξ
, (2.88)
¸AA) =
_
dξ
¸Aξ
)¸ξ
A) =
_
dξ
¸Aξ
)
2
. (2.89)
These results also follow simply from Eq. (2.86). We have seen that it is not possi
ble to normalize the eigenstates ξ
) such that they have unit norms. Fortunately,
this convenient normalization is still possible for a general state vector. In fact,
according to Eq. (2.89), the normalization condition can be written
¸AA) =
_
dξ
¸Aξ
)
2
= 1. (2.90)
We have now studied observables whose eigenvalues can take a discrete num
ber of values as well as those whose eigenvalues can take any value. There are
number of other cases we could look at. For instance, observables whose eigen
values can only take a ﬁnite range of values, or observables whose eigenvalues
take on a ﬁnite range of values plus a set of discrete values. Both of these cases
can be dealt with using a fairly straightforward generalization of the previous
analysis (see Dirac, Cha. II and III).
32
3 POSITION AND MOMENTUM
3 Position and momentum
3.1 Introduction
So far, we have considered general dynamical variables represented by general
linear operators acting in ket space. However, in classical mechanics the most
important dynamical variables are those involving position and momentum. Let
us investigate the role of such variables in quantum mechanics.
In classical mechanics, the position q and momentum p of some component
of a dynamical system are represented as real numbers which, by deﬁnition, com
mute. In quantum mechanics, these quantities are represented as noncommuting
linear Hermitian operators acting in a ket space which represents all of the pos
sible states of the system. Our ﬁrst task is to discover a quantum mechanical
replacement for the classical result qp − pq = 0. Do the position and momen
tum operators commute? If not, what is the value of qp −pq?
3.2 Poisson brackets
Consider a dynamic system whose state at a particular time t is fully speciﬁed
by N independent classical coordinates q
i
(where i runs from 1 to N). Associ
ated with each generalized coordinate q
i
is a classical canonical momentum p
i
.
For instance, a Cartesian coordinate has an associated linear momentum, an an
gular coordinate has an associated angular momentum, etc. As is wellknown,
the behaviour of a classical system can be speciﬁed in terms of Lagrangian or
Hamiltonian dynamics. For instance, in Hamiltonian dynamics,
dq
i
dt
=
∂H
∂p
i
, (3.1)
dp
i
dt
= −
∂H
∂q
i
, (3.2)
where the function H(q
i
, p
i
, t) is the energy of the system at time t expressed
in terms of the classical coordinates and canonical momenta. This function is
33
3.2 Poisson brackets 3 POSITION AND MOMENTUM
usually referred to as the Hamiltonian of the system.
We are interested in ﬁnding some construct of classical dynamics which con
sists of products of dynamical variables. If such a construct exists we hope to
generalize it somehow to obtain a rule describing how dynamical variables com
mute with one another in quantum mechanics. There is, indeed, one wellknown
construct in classical dynamics which involves products of dynamical variables.
The Poisson bracket of two dynamical variables u and v is deﬁned
[u, v] =
N
i=1
_
∂u
∂q
i
∂v
∂p
i
−
∂u
∂p
i
∂v
∂q
i
_
, (3.3)
where u and v are regarded as functions of the coordinates and momenta q
i
and
p
i
. It is easily demonstrated that
[q
i
, q
j
] = 0, (3.4)
[p
i
, p
j
] = 0, (3.5)
[q
i
, p
j
] = δ
ij
. (3.6)
The time evolution of a dynamical variable can also be written in terms of a
Poisson bracket by noting that
du
dt
=
N
i=1
_
∂u
∂q
i
dq
i
dt
+
∂u
∂p
i
dp
i
dt
_
=
N
i=1
_
∂u
∂q
i
∂H
∂p
i
−
∂u
∂p
i
∂H
∂q
i
_
= [u, H], (3.7)
where use has been made of Hamilton’s equations.
Can we construct a quantum mechanical Poisson bracket in which u and v are
noncommuting operators, instead of functions? Well, the main properties of the
classical Poisson bracket are as follows:
[u, v] = −[v, u], (3.8)
34
3.2 Poisson brackets 3 POSITION AND MOMENTUM
[u, c] = 0, (3.9)
[u
1
+u
2
, v] = [u
1
, v] + [u
2
, v], (3.10)
[u, v
1
+v
2
] = [u, v
1
] + [u, v
2
] (3.11)
[u
1
u
2
, v] = [u
1
, v]u
2
+u
1
[u
2
, v], (3.12)
[u, v
1
v
2
] = [u, v
1
]v
2
+v
1
[u, v
2
], (3.13)
and
[u, [v, w]] + [v, [w, u]] + [w, [u, v]] = 0. (3.14)
The last relation is known as the Jacobi identity. In the above, u, v, w, etc.,
represent dynamical variables, and c represents a number. Can we ﬁnd some
combination of noncommuting operators u and v, etc., which satisﬁes all of the
above relations?
Well, we can evaluate the Poisson bracket [u
1
u
2
, v
1
v
2
] in two different ways,
since we can use either of the formulae (3.12) or (3.13) ﬁrst. Thus,
[u
1
u
2
, v
1
v
2
] = [u
1
, v
1
v
2
]u
2
+u
1
[u
2
, v
1
v
2
] (3.15)
= {[u
1
, v
1
]v
2
+v
1
[u
1
, v
2
]} u
2
+u
1
{[u
2
, v
1
]v
2
+v
1
[u
2
, v
2
]}
= [u
1
, v
1
]v
2
u
2
+v
1
[u
1
, v
2
]u
2
+u
1
[u
2
, v
1
]v
2
+u
1
v
1
[u
2
, v
2
],
and
[u
1
u
2
, v
1
v
2
] = [u
1
u
2
, v
1
]v
2
+v
1
[u
1
u
2
, v
2
] (3.16)
= [u
1
, v
1
]u
2
v
2
+u
1
[u
2
, v
1
]v
2
+v
1
[u
1
, v
2
]u
2
+v
1
u
1
[u
2
, v
2
].
Note that the order of the various factors has been preserved, since they now
represent noncommuting operators. Equating the above two results yields
[u
1
, v
1
](u
2
v
2
−v
2
u
2
) = (u
1
v
1
−v
1
u
1
)[u
2
, v
2
]. (3.17)
Since this relation must hold for u
1
and v
1
quite independent of u
2
and v
2
, it
follows that
u
1
v
1
−v
1
u
1
= i ¯h[u
1
, v
1
], (3.18)
u
2
v
2
−v
2
u
2
= i ¯h[u
2
, v
2
], (3.19)
35
3.2 Poisson brackets 3 POSITION AND MOMENTUM
where ¯h does not depend on u
1
, v
1
, u
2
, v
2
, and also commutes with (u
1
v
1
−v
1
u
1
).
Since u
1
, etc., are quite general operators, it follows that ¯h is just a number. We
want the quantum mechanical Poisson bracket of two Hermitian operators to
be an Hermitian operator itself, since the classical Poisson bracket of two real
dynamical variables is real. This requirement is satisﬁed if ¯h is a real number.
Thus, the quantum mechanical Poisson bracket of two dynamical variables u and
v is given by
[u, v] =
uv −v u
i ¯h
, (3.20)
where ¯h is a new universal constant of nature. Quantum mechanics agrees with
experiments provided that ¯h takes the value h/2π, where
h = 6.6261 10
−34
J s (3.21)
is Planck’s constant. Somewhat confusingly, the notation [u, v] is convention
ally reserved for the commutator uv − v u in quantum mechanics. We will use
[u, v]
quantum
to denote the quantum Poisson bracket. Thus,
[u, v]
quantum
=
[u, v]
i ¯h
. (3.22)
It is easily demonstrated that the quantum mechanical Poisson bracket, as deﬁned
above, satisﬁes all of the relations (3.8)–(3.14).
The strong analogy we have found between the classical Poisson bracket,
deﬁned in Eq. (3.3), and the quantum mechanical Poisson bracket, deﬁned in
Eq. (3.22), leads us to make the assumption that the quantum mechanical bracket
has the same value as the corresponding classical bracket, at least for the simplest
cases. In other words, we are assuming that Eqs. (3.4)–(3.6) hold for quantum
mechanical as well as classical Poisson brackets. This argument yields the funda
mental commutation relations
[q
i
, q
j
] = 0, (3.23)
[p
i
, p
j
] = 0, (3.24)
[q
i
, p
j
] = i ¯hδ
ij
. (3.25)
These results provide us with the basis for calculating commutation relations be
tween general dynamical variables. For instance, if two dynamical variables, ξ
36
3.3 Wavefunctions 3 POSITION AND MOMENTUM
and η, can both be written as a power series in the q
i
and p
i
, then repeated
application of Eqs. (3.8)–(3.13) allows [ξ, η] to be expressed in terms of the fun
damental commutation relations (3.23)–(3.25).
Equations (3.23)–(3.25) provide the foundation for the analogy between quan
tum mechanics and classical mechanics. Note that the classical result (that every
thing commutes) is obtained in the limit ¯h → 0. Thus, classical mechanics can be
regarded as the limiting case of quantum mechanics when ¯h goes to zero. In classi
cal mechanics, each pair of generalized coordinate and its conjugate momentum,
q
i
and p
i
, correspond to a different classical degree of freedom of the system. It is
clear from Eqs. (3.23)–(3.25) that in quantum mechanics the dynamical variables
corresponding to different degrees of freedom all commute. It is only those variables
corresponding to the same degree of freedom which may fail to commute.
3.3 Wavefunctions
Consider a simple system with one classical degree of freedom, which corre
sponds to the Cartesian coordinate x. Suppose that x is free to take any value
(e.g., x could be the position of a free particle). The classical dynamical vari
able x is represented in quantum mechanics as a linear Hermitian operator which
is also called x. Moreover, the operator x possesses eigenvalues x
lying in the
continuous range −∞ < x
< +∞ (since the eigenvalues correspond to all the
possible results of a measurement of x). We can span ket space using the suit
ably normalized eigenkets of x. An eigenket corresponding to the eigenvalue x
is denoted x
). Moreover, [see Eq. (2.85)]
¸x
x
) = δ(x
−x
). (3.26)
The eigenkets satisfy the extremely useful relation [see Eq. (2.86)]
_
+∞
−∞
dx
x
)¸x
 = 1. (3.27)
This formula expresses the fact that the eigenkets are complete, mutually orthog
onal, and suitably normalized.
37
3.3 Wavefunctions 3 POSITION AND MOMENTUM
A state ket A) (which represents a general state A of the system) can be
expressed as a linear superposition of the eigenkets of the position operator using
Eq. (3.27). Thus,
A) =
_
+∞
−∞
dx
¸x
A)x
) (3.28)
The quantity ¸x
A) is a complex function of the position eigenvalue x
. We can
write
¸x
A) = ψ
A
(x
). (3.29)
Here, ψ
A
(x
) is the famous wavefunction of quantum mechanics. Note that state
A is completely speciﬁed by its wavefunction ψ
A
(x
) [since the wavefunction
can be used to reconstruct the state ket A) using Eq. (3.28)]. It is clear that
the wavefunction of state A is simply the collection of the weights of the cor
responding state ket A), when it is expanded in terms of the eigenkets of the
position operator. Recall, from Sect. 2.10, that the probability of a measurement
of a dynamical variable ξ yielding the result ξ
when the system is in state A is
given by ¸ξ
A)
2
, assuming that the eigenvalues of ξ are discrete. This result is
easily generalized to dynamical variables possessing continuous eigenvalues. In
fact, the probability of a measurement of x yielding a result lying in the range
x
to x
+ dx
when the system is in a state A) is ¸x
A)
2
dx
. In other words,
the probability of a measurement of position yielding a result in the range x
to
x
+dx
when the wavefunction of the system is ψ
A
(x
) is
P(x
, dx
) = ψ
A
(x
)
2
dx
. (3.30)
This formula is only valid if the state ket A) is properly normalized: i.e., if
¸AA) = 1. The corresponding normalization for the wavefunction is
_
+∞
−∞
ψ
A
(x
)
2
dx
= 1. (3.31)
Consider a second state B represented by a state ket B) and a wavefunction
ψ
B
(x
). The inner product ¸BA) can be written
¸BA) =
_
+∞
−∞
dx
¸Bx
)¸x
A) =
_
+∞
−∞
ψ
∗
B
(x
) ψ
A
(x
) dx
, (3.32)
38
3.4 Schr¨ odinger’s representation  I 3 POSITION AND MOMENTUM
where use has been made of Eqs. (3.27) and (3.29). Thus, the inner product of
two states is related to the overlap integral of their wavefunctions.
Consider a general function f(x) of the observable x [e.g., f(x) = x
2
]. If B) =
f(x)A) then it follows that
ψ
B
(x
) = ¸x
f(x)
_
+∞
−∞
dx
ψ
A
(x
)x
)
=
_
+∞
−∞
dx
f(x
) ψ
A
(x
)¸x
x
), (3.33)
giving
ψ
B
(x
) = f(x
) ψ
A
(x
), (3.34)
where use has been made of Eq. (3.26). Here, f(x
) is the same function of the
position eigenvalue x
that f(x) is of the position operator x: i.e., if f(x) = x
2
then
f(x
) = x
2
. It follows, from the above result, that a general state ket A) can be
written
A) = ψ
A
(x)), (3.35)
where ψ
A
(x) is the same function of the operator x that the wavefunction ψ
A
(x
)
is of the position eigenvalue x
, and the ket ) has the wavefunction ψ(x
) = 1.
The ket ) is termed the standard ket. The dual of the standard ket is termed the
standard bra, and is denoted ¸. It is easily seen that
¸ψ
∗
A
(x)
DC
←→ψ
A
(x)). (3.36)
Note, ﬁnally, that ψ
A
(x)) is often shortened to ψ
A
), leaving the dependence on
the position operator x tacitly understood.
3.4 Schr¨ odinger’s representation  I
Consider the simple system described in the previous section. A general state ket
can be written ψ(x)), where ψ(x) is a general function of the position operator x,
and ψ(x
) is the associated wavefunction. Consider the ket whose wavefunction
39
3.4 Schr¨ odinger’s representation  I 3 POSITION AND MOMENTUM
is dψ(x
)/dx
. This ket is denoted dψ/dx). The new ket is clearly a linear func
tion of the original ket, so we can think of it as the result of some linear operator
acting on ψ). Let us denote this operator d/dx. It follows that
d
dx
ψ) =
dψ
dx
). (3.37)
Any linear operator which acts on ket vectors can also act on bra vectors.
Consider d/dx acting on a general bra ¸φ(x). According to Eq. (2.34), the bra
¸φd/dx satisﬁes
_
¸φ
d
dx
_
ψ) = ¸φ
_
d
dx
ψ)
_
. (3.38)
Making use of Eqs. (3.27) and (3.29), we can write
_
+∞
−∞
¸φ
d
dx
x
) dx
ψ(x
) =
_
+∞
−∞
φ(x
) dx
dψ(x
)
dx
. (3.39)
The righthand side can be transformed via integration by parts to give
_
+∞
−∞
¸φ
d
dx
x
) dx
ψ(x
) = −
_
+∞
−∞
dφ(x
)
dx
dx
ψ(x
), (3.40)
assuming that the contributions from the limits of integration vanish. It follows
that
¸φ
d
dx
x
) = −
dφ(x
)
dx
, (3.41)
which implies
¸φ
d
dx
= −¸
dφ
dx
. (3.42)
The neglect of contributions from the limits of integration in Eq. (3.40) is rea
sonable because physical wavefunctions are squareintegrable [see Eq. (3.31)].
Note that
d
dx
ψ) =
dψ
dx
)
DC
←→¸
dψ
∗
dx
= −¸ψ
∗
d
dx
, (3.43)
where use has been made of Eq. (3.42). It follows, by comparison with Eqs. (2.35)
and (3.36), that
_
d
dx
_
†
= −
d
dx
. (3.44)
40
3.4 Schr¨ odinger’s representation  I 3 POSITION AND MOMENTUM
Thus, d/dx is an antiHermitian operator.
Let us evaluate the commutation relation between the operators x and d/dx.
We have
d
dx
x ψ) =
d(x ψ)
dx
) = x
d
dx
ψ) +ψ). (3.45)
Since this holds for any ket ψ), it follows that
d
dx
x −x
d
dx
= 1. (3.46)
Let p be the momentumconjugate to x (for the simple systemunder consideration
p is a straightforward linear momentum). According to Eq. (3.25), x and p
satisfy the commutation relation
x p −px = i ¯h. (3.47)
It can be seen, by comparison with Eq. (3.46), that the Hermitian operator
−i ¯hd/dx satisﬁes the same commutation relation with x that p does. The most
general conclusion which may be drawn from a comparison of Eqs. (3.46) and
(3.47) is that
p = −i ¯h
d
dx
+f(x), (3.48)
since (as is easily demonstrated) a general function f(x) of the position operator
automatically commutes with x.
We have chosen to normalize the eigenkets and eigenbras of the position oper
ator so that they satisfy the normalization condition (3.26). However, this choice
of normalization does not uniquely determine the eigenkets and eigenbras. Sup
pose that we transform to a new set of eigenbras which are related to the old set
via
¸x

new
= e
i γ
¸x

old
, (3.49)
where γ
≡ γ(x
) is a real function of x
. This transformation amounts to a
rearrangement of the relative phases of the eigenbras. The new normalization
condition is
¸x
x
)
new
= ¸x
e
i γ
e
−i γ
x
)
old
= e
i (γ
−γ
)
¸x
x
)
old
= e
i (γ
−γ
)
δ(x
−x
) = δ(x
−x
). (3.50)
41
3.4 Schr¨ odinger’s representation  I 3 POSITION AND MOMENTUM
Thus, the new eigenbras satisfy the same normalization condition as the old
eigenbras.
By deﬁnition, the standard ket ) satisﬁes ¸x
) = 1. It follows from Eq. (3.49)
that the new standard ket is related to the old standard ket via
)
new
= e
−i γ
)
old
, (3.51)
where γ ≡ γ(x) is a real function of the position operator x. The dual of the
above equation yields the transformation rule for the standard bra,
¸
new
= ¸
old
e
i γ
. (3.52)
The transformation rule for a general operator A follows from Eqs. (3.51) and
(3.52), plus the requirement that the triple product ¸A) remain invariant (this
must be the case, otherwise the probability of a measurement yielding a certain
result would depend on the choice of eigenbras). Thus,
A
new
= e
−i γ
A
old
e
i γ
. (3.53)
Of course, if A commutes with x then A is invariant under the transformation. In
fact, d/dx is the only operator (we know of) which does not commute with x, so
Eq. (3.53) yields
_
d
dx
_
new
= e
−i γ
d
dx
e
i γ
=
d
dx
+ i
dγ
dx
, (3.54)
where the subscript “old” is taken as read. It follows, from Eq. (3.48), that the
momentum operator p can be written
p = −i ¯h
_
d
dx
_
new
− ¯h
dγ
dx
+f(x). (3.55)
Thus, the special choice
¯hγ(x) =
_
x
f(x) dx (3.56)
yields
p = −i ¯h
_
d
dx
_
new
. (3.57)
42
3.5 Schr¨ odinger’s representation  II 3 POSITION AND MOMENTUM
Equation (3.56) ﬁxes γ to within an arbitrary additive constant: i.e., the special
eigenkets and eigenbras for which Eq. (3.57) is true are determined to within an
arbitrary common phasefactor.
In conclusion, it is possible to ﬁnd a set of basis eigenkets and eigenbras of
the position operator x which satisfy the normalization condition (3.26), and for
which the momentum conjugate to x can be represented as the operator
p = −i ¯h
d
dx
. (3.58)
A general state ket is written ψ(x)), where the standard ket ) satisﬁes ¸x
) =
1, and where ψ(x
) = ¸x
ψ(x)) is the wavefunction. This scheme of things is
known as Schr¨ odinger’s representation, and is the basis of wave mechanics.
3.5 Schr¨ odinger’s representation  II
In the preceding sections, we have developed Schr¨ odinger’s representation for
the case of a single operator x corresponding to a classical Cartesian coordinate.
However, this scheme can easily be extended. Consider a system with N general
ized coordinates, q
1
q
N
, which can all be simultaneously measured. These are
represented as N commuting operators, q
1
q
N
, each with a continuous range
of eigenvalues, q
1
q
N
. Ket space is conveniently spanned by the simultaneous
eigenkets of q
1
q
N
, which are denoted q
1
q
N
). These eigenkets must form
a complete set, otherwise the q
1
q
N
would not be simultaneously observable.
The orthogonality condition for the eigenkets [i.e., the generalization of Eq. (3.26)]
is
¸q
1
q
N
q
1
q
N
) = δ(q
1
−q
1
) δ(q
2
−q
2
) δ(q
N
−q
N
). (3.59)
The completeness condition [i.e., the generalization of Eq. (3.27)] is
_
+∞
−∞
_
+∞
−∞
dq
1
dq
N
q
1
q
N
)¸q
1
q
N
 = 1. (3.60)
The standard ket ) is deﬁned such that
¸q
1
q
N
) = 1. (3.61)
43
3.5 Schr¨ odinger’s representation  II 3 POSITION AND MOMENTUM
The standard bra ¸ is the dual of the standard ket. A general state ket is written
ψ(q
1
q
N
)). (3.62)
The associated wavefunction is
ψ(q
1
q
N
) = ¸q
1
q
N
ψ). (3.63)
Likewise, a general state bra is written
¸φ(q
1
q
N
), (3.64)
where
φ(q
1
q
N
) = ¸φq
1
q
N
). (3.65)
The probability of an observation of the system ﬁnding the ﬁrst coordinate in the
range q
1
to q
1
+dq
1
, the second coordinate in the range q
2
to q
2
+dq
2
, etc., is
P(q
1
q
N
; dq
1
dq
N
) = ψ(q
1
q
N
)
2
dq
1
dq
N
. (3.66)
Finally, the normalization condition for a physical wavefunction is
_
+∞
−∞
_
+∞
−∞
ψ(q
1
q
N
)
2
dq
1
dq
N
= 1. (3.67)
The N linear operators ∂/∂q
i
(where i runs from 1 to N) are deﬁned
∂
∂q
i
ψ) =
∂ψ
∂q
i
). (3.68)
These linear operators can also act on bras (provided the associated wavefunctions
are square integrable) in accordance with [see Eq. (3.42)]
¸φ
∂
∂q
i
= −¸
∂φ
∂q
i
. (3.69)
Corresponding to Eq. (3.46), we can derive the commutation relations
∂
∂q
i
q
j
−q
j
∂
∂q
i
= δ
ij
. (3.70)
44
3.5 Schr¨ odinger’s representation  II 3 POSITION AND MOMENTUM
It is also clear that
∂
∂q
i
∂
∂q
j
ψ) =
∂
2
ψ
∂q
i
∂q
j
) =
∂
∂q
j
∂
∂q
i
ψ), (3.71)
showing that
∂
∂q
i
∂
∂q
j
=
∂
∂q
j
∂
∂q
i
. (3.72)
It can be seen, by comparison with Eqs. (3.23)–(3.25), that the linear oper
ators −i ¯h∂/∂q
i
satisfy the same commutation relations with the q’s and with
each other that the p’s do. The most general conclusion we can draw from this
coincidence of commutation relations is (see Dirac)
p
i
= −i ¯h
∂
∂q
i
+
∂F(q
1
q
N
)
∂q
i
. (3.73)
However, the function F can be transformed away via a suitable readjustment of
the phases of the basis eigenkets (see Sect. 3.4, and Dirac). Thus, we can always
construct a set of simultaneous eigenkets of q
1
q
N
for which
p
i
= −i ¯h
∂
∂q
i
. (3.74)
This is the generalized Schr¨ odinger representation.
It follows from Eqs. (3.61), (3.68), and (3.74) that
p
i
) = 0. (3.75)
Thus, the standard ket in Schr¨ odinger’s representation is a simultaneous eigenket
of all the momentum operators belonging to the eigenvalue zero. Note that
¸q
1
q
N

∂
∂q
i
ψ) = ¸q
1
q
N

∂ψ
∂q
i
) =
∂ψ(q
1
q
N
)
∂q
i
=
∂
∂q
i
¸q
1
q
N
ψ). (3.76)
Hence,
¸q
1
q
N

∂
∂q
i
=
∂
∂q
i
¸q
1
q
N
, (3.77)
so that
¸q
1
q
N
p
i
= −i ¯h
∂
∂q
i
¸q
1
q
N
. (3.78)
45
3.6 The momentum representation 3 POSITION AND MOMENTUM
The dual of the above equation gives
p
i
q
1
q
N
) = i ¯h
∂
∂q
i
q
1
q
N
). (3.79)
3.6 The momentum representation
Consider a system with one degree of freedom, describable in terms of a coordi
nate x and its conjugate momentum p, both of which have a continuous range
of eigenvalues. We have seen that it is possible to represent the system in terms
of the eigenkets of x. This is termed Schr¨ odinger’s representation. However, it is
also possible to represent the system in terms of the eigenkets of p.
Consider the eigenkets of p which belong to the eigenvalues p
. These are
denoted p
). The orthogonality relation for the momentum eigenkets is
¸p
p
) = δ(p
−p
), (3.80)
and the corresponding completeness relation is
_
+∞
−∞
dp
p
)¸p
 = 1. (3.81)
A general state ket can be written
φ(p)) (3.82)
where the standard ket ) satisﬁes
¸p
) = 1. (3.83)
Note that the standard ket in this representation is quite different to that in
Schr¨ odinger’s representation. The momentum space wavefunction φ(p
) sat
isﬁes
φ(p
) = ¸p
φ). (3.84)
The probability that a measurement of the momentum yields a result lying in the
range p
to p
+dp
is given by
P(p
, dp
) = φ(p
)
2
dp
. (3.85)
46
3.6 The momentum representation 3 POSITION AND MOMENTUM
Finally, the normalization condition for a physical momentumspace wavefunction
is
_
+∞
−∞
φ(p
)
2
dp
= 1. (3.86)
The fundamental commutation relations (3.23)–(3.25) exhibit a particular
symmetry between coordinates and their conjugate momenta. If all the coor
dinates are transformed into their conjugate momenta, and vice versa, and i is
then replaced by −i, the commutation relations are unchanged. It follows from
this symmetry that we can always choose the eigenkets of p in such a manner
that the coordinate x can be represented as (see Sect. 3.4)
x = i ¯h
d
dp
. (3.87)
This is termed the momentum representation.
The above result is easily generalized to a system with more than one degree
of freedom. Suppose the system is speciﬁed by N coordinates, q
1
q
N
, and
N conjugate momenta, p
1
p
N
. Then, in the momentum representation, the
coordinates can be written as
q
i
= i ¯h
∂
∂p
i
. (3.88)
We also have
q
i
) = 0, (3.89)
and
¸p
1
p
N
q
i
= i ¯h
∂
∂p
i
¸p
1
p
N
. (3.90)
The momentum representation is less useful than Schr¨ odinger’s representa
tion for a very simple reason. The energy operator (i.e., the Hamiltonian) of
most simple systems takes the form of a sum of quadratic terms in the momenta
(i.e., the kinetic energy) plus a complicated function of the coordinates (i.e., the
potential energy). In Schr¨ odinger’s representation, the eigenvalue problem for
the energy translates into a secondorder differential equation in the coordinates,
with a complicated potential function. In the momentum representation, the
47
3.7 The uncertainty relation 3 POSITION AND MOMENTUM
problem transforms into a highorder differential equation in the momenta, with
a quadratic potential. With the mathematical tools at our disposal, we are far bet
ter able to solve the former type of problem than the latter. Hence, Schr¨ odinger’s
representation is generally more useful than the momentum representation.
3.7 The uncertainty relation
How is a momentum space wavefunction related to the corresponding coordi
nate space wavefunction? To answer this question, let us consider the represen
tative ¸x
p
) of the momentum eigenkets p
) in Schr¨ odinger’s representation for
a system with a single degree of freedom. This representative satisﬁes
p
¸x
p
) = ¸x
pp
) = −i ¯h
d
dx
¸x
p
), (3.91)
where use has been made of Eq. (3.78) (for the case of a system with one degree
of freedom). The solution of the above differential equation is
¸x
p
) = c
exp(i p
x
/¯h), (3.92)
where c
= c
(p
). It is easily demonstrated that
¸p
p
) =
_
+∞
−∞
¸p
x
) dx
¸x
p
) = c
∗
c
_
∞
−∞
exp[−i (p
−p
) x
/¯h] dx
. (3.93)
The wellknown mathematical result
_
+∞
−∞
exp(i ax) dx = 2πδ(a), (3.94)
yields
¸p
p
) = c

2
hδ(p
−p
). (3.95)
This is consistent with Eq. (3.80), provided that c
= h
−1/2
. Thus,
¸x
p
) = h
−1/2
exp(i p
x
/¯h). (3.96)
48
3.7 The uncertainty relation 3 POSITION AND MOMENTUM
Consider a general state ket A) whose coordinate wavefunction is ψ(x
), and
whose momentum wavefunction is Ψ(p
). In other words,
ψ(x
) = ¸x
A), (3.97)
Ψ(p
) = ¸p
A). (3.98)
It is easily demonstrated that
ψ(x
) =
_
+∞
−∞
dp
¸x
p
)¸p
A)
=
1
h
1/2
_
+∞
−∞
Ψ(p
) exp(i p
x
/¯h) dp
(3.99)
and
Ψ(p
) =
_
+∞
−∞
dx
¸p
x
)¸x
A)
=
1
h
1/2
_
+∞
−∞
ψ(x
) exp(−i p
x
/¯h) dx
, (3.100)
where use has been made of Eqs. (3.27), (3.81), (3.94), and (3.96). Clearly, the
momentum space wavefunction is the Fourier transform of the coordinate space
wavefunction.
Consider a state whose coordinate space wavefunction is a wavepacket. In
other words, the wavefunction only has nonnegligible amplitude in some spa
tially localized region of extent ∆x. As is wellknow, the Fourier transform of a
wavepacket ﬁlls up a wavenumber band of approximate extent δk ∼ 1/∆x. Note
that in Eq. (3.99) the role of the wavenumber k is played by the quantity p
/¯h. It
follows that the momentum space wavefunction corresponding to a wavepacket
in coordinate space extends over a range of momenta ∆p ∼ ¯h/∆x. Clearly, a mea
surement of x is almost certain to give a result lying in a range of width ∆x.
Likewise, measurement of p is almost certain to yield a result lying in a range of
width ∆p. The product of these two uncertainties is
∆x ∆p ∼ ¯h. (3.101)
This result is called Heisenberg’s uncertainty principle.
49
3.8 Displacement operators 3 POSITION AND MOMENTUM
Actually, it is possible to write Heisenberg’s uncertainty principle more exactly
by making use of Eq. (2.83) and the commutation relation (3.47). We obtain
¸(∆x)
2
)¸(∆p)
2
) ≥
¯h
2
4
(3.102)
for any general state. It is easily demonstrated that the minimum uncertainty
states, for which the equality sign holds in the above relation, correspond to
Gaussian wavepackets in both coordinate and momentum space.
3.8 Displacement operators
Consider a system with one degree of freedom corresponding to the Cartesian
coordinate x. Suppose that we displace this system some distance along the x
axis. We could imagine that the system is on wheels, and we just give it a little
push. The ﬁnal state of the system is completely determined by its initial state,
together with the direction and magnitude of the displacement. Note that the
type of displacement we are considering is one in which everything to do with the
system is displaced. So, if the system is subject to an external potential, then the
potential must be displaced.
The situation is not so clear with state kets. The ﬁnal state of the system
only determines the direction of the displaced state ket. Even if we adopt the
convention that all state kets have unit norms, the ﬁnal ket is still not completely
determined, since it can be multiplied by a constant phasefactor. However, we
know that the superposition relations between states remain invariant under the
displacement. This follows because the superposition relations have a physical
signiﬁcance which is unaffected by a displacement of the system. Thus, if
R) = A) + B) (3.103)
in the undisplaced system, and the displacement causes ket R) to transform to
ket Rd), etc., then in the displaced system we have
Rd) = Ad) + Bd). (3.104)
50
3.8 Displacement operators 3 POSITION AND MOMENTUM
Incidentally, this determines the displaced kets to within a single arbitrary phase
factor to be multiplied into all of them. The displaced kets cannot be multiplied
by individual phasefactors, because this would wreck the superposition relations.
Since Eq. (3.104) holds in the displaced system whenever Eq. (3.103) holds in
the undisplaced system, it follows that the displaced ket Rd) must be the result
of some linear operator acting on the undisplaced ket R). In other words,
Rd) = DR), (3.105)
where D an operator which depends only on the nature of the displacement. The
arbitrary phasefactor by which all displaced kets may be multiplied results in D
being undetermined to an arbitrary multiplicative constant of modulus unity.
We now adopt the ansatz that any combination of bras, kets, and dynamical
variables which possesses a physical signiﬁcance is invariant under a displace
ment of the system. The normalization condition
¸AA) = 1 (3.106)
for a state ket A) certainly has a physical signiﬁcance. Thus, we must have
¸AdAd) = 1. (3.107)
Now, Ad) = DA) and ¸Ad = ¸AD
†
, so
¸AD
†
DA) = 1. (3.108)
Since this must hold for any state ket A), it follows that
D
†
D = 1. (3.109)
Hence, the displacement operator is unitary. Note that the above relation implies
that
A) = D
†
Ad). (3.110)
The equation
v A) = B), (3.111)
51
3.8 Displacement operators 3 POSITION AND MOMENTUM
where the operator v represents a dynamical variable, has some physical signiﬁ
cance. Thus, we require that
v
d
Ad) = Bd), (3.112)
where v
d
is the displaced operator. It follows that
v
d
Ad) = DB) = Dv A) = Dv D
†
Ad). (3.113)
Since this is true for any ket Ad), we have
v
d
= Dv D
†
. (3.114)
Note that the arbitrary numerical factor in D does not affect either of the results
(3.109) and (3.114).
Suppose, now, that the system is displaced an inﬁnitesimal distance δx along
the xaxis. We expect that the displaced ket Ad) should approach the undisplaced
ket A) in the limit as δx →0. Thus, we expect the limit
lim
δx→0
Ad) − A)
δx
= lim
δx→0
D−1
δx
A) (3.115)
to exist. Let
d
x
= lim
δx→0
D−1
δx
, (3.116)
where d
x
is denoted the displacement operator along the xaxis. The fact that D
can be replaced by Dexp(i γ), where γ is a real phaseangle, implies that d
x
can
be replaced by
lim
δx→0
Dexp(i γ) −1
δx
= lim
δx→0
D−1 + i γ
δx
= d
x
+ i a
x
, (3.117)
where a
x
is the limit of γ/δx. We have assumed, as seems reasonable, that γ tends
to zero as δx → 0. It is clear that the displacement operator is undetermined to
an arbitrary imaginary additive constant.
For small δx, we have
D = 1 +δx d
x
. (3.118)
52
3.8 Displacement operators 3 POSITION AND MOMENTUM
It follows from Eq. (3.109) that
(1 +δx d
†
x
)(1 +δx d
x
) = 1. (3.119)
Neglecting order (δx)
2
, we obtain
d
†
x
+d
x
= 0. (3.120)
Thus, the displacement operator is antiHermitian. Substituting into Eq. (3.114),
and again neglecting order (δx)
2
, we ﬁnd that
v
d
= (1 +δx d
x
) v (1 −δx d
x
) = v +δx (d
x
v −v d
x
), (3.121)
which implies
lim
δx→0
v
d
−v
δx
= d
x
v −v d
x
. (3.122)
Let us consider a speciﬁc example. Suppose that a state has a wavefunction
ψ(x
). If the system is displaced a distance δx along the xaxis then the new
wavefunction is ψ(x
− δx) (i.e., the same shape shifted in the xdirection by a
distance δx). Actually, the new wavefunction can be multiplied by an arbitrary
number of modulus unity. It can be seen that the new wavefunction is obtained
from the old wavefunction according to the prescription x
→x
−δx. Thus,
x
d
= x −δx. (3.123)
A comparison with Eq. (3.122), using x = v, yields
d
x
x −x d
x
= −1. (3.124)
It follows that i ¯hd
x
obeys the same commutation relation with x that p
x
, the
momentum conjugate to x, does [see Eq. (3.25)]. The most general conclusion
we can draw from this observation is that
p
x
= i ¯hd
x
+f(x), (3.125)
where f is Hermitian (since p
x
is Hermitian). However, the fact that d
x
is unde
termined to an arbitrary additive imaginary constant (which could be a function
of x) enables us to transform the function f out of the above equation, leaving
p
x
= i ¯hd
x
. (3.126)
53
3.8 Displacement operators 3 POSITION AND MOMENTUM
Thus, the displacement operator in the xdirection is proportional to the momen
tum conjugate to x. We say that p
x
is the generator of translations along the
xaxis.
A ﬁnite translation along the xaxis can be constructed from a series of very
many inﬁnitesimal translations. Thus, the operator D(∆x) which translates the
system a distance ∆x along the xaxis is written
D(∆x) = lim
N→∞
_
1 − i
∆x
N
p
x
¯h
_
N
, (3.127)
where use has been made of Eqs. (3.118) and (3.126). It follows that
D(∆x) = exp(−i p
x
∆x/¯h) . (3.128)
The unitary nature of the operator is now clearly apparent.
We can also construct displacement operators which translate the system along
the y and zaxes. Note that a displacement a distance ∆x along the xaxis com
mutes with a displacement a distance ∆y along the yaxis. In other words, if the
system is moved ∆x along the xaxis, and then ∆y along the yaxis then it ends
up in the same state as if it were moved ∆y along the yaxis, and then ∆x along
the xaxis. The fact that translations in independent directions commute is clearly
associated with the fact that the conjugate momentum operators associated with
these directions also commute [see Eqs. (3.24) and (3.128)].
54
4 QUANTUM DYNAMICS
4 Quantum dynamics
4.1 Schr¨ odinger’s equations of motion
Up to now, we have only considered systems at one particular instant of time. Let
us now investigate how quantum mechanical systems evolve with time.
Consider a system in a state A which evolves in time. At time t the state of the
system is represented by the ket At). The label A is needed to distinguish the
ket from any other ket (Bt), say) which is evolving in time. The label t is needed
to distinguish the different states of the system at different times.
The ﬁnal state of the system at time t is completely determined by its initial
state at time t
0
plus the time interval t − t
0
(assuming that the system is left
undisturbed during this time interval). However, the ﬁnal state only determines
the direction of the ﬁnal state ket. Even if we adopt the convention that all state
kets have unit norms, the ﬁnal ket is still not completely determined, since it
can be multiplied by an arbitrary phasefactor. However, we expect that if a
superposition relation holds for certain states at time t
0
then the same relation
should hold between the corresponding timeevolved states at time t, assuming
that the system is left undisturbed between times t
0
and t. In other words, if
Rt
0
) = At
0
) + Bt
0
) (4.1)
for any three kets, then we should have
Rt) = At) + Bt). (4.2)
This rule determines the timeevolved kets to within a single arbitrary phase
factor to be multiplied into all of them. The evolved kets cannot be multiplied by
individual phasefactors, since this would invalidate the superposition relation at
later times.
According to Eqs. (4.1) and (4.2), the ﬁnal ket Rt) depends linearly on the
initial ket Rt
0
). Thus, the ﬁnal ket can be regarded as the result of some linear
operator acting on the initial ket: , i.e.,
Rt) = TRt
0
), (4.3)
55
4.1 Schr¨ odinger’s equations of motion 4 QUANTUM DYNAMICS
where T is a linear operator which depends only on the times t and t
0
. The
arbitrary phasefactor by which all time evolved kets may be multiplied results
in T(t, t
0
) being undetermined to an arbitrary multiplicative constant of modulus
unity.
Since we have adopted a convention in which the norm of any state ket is
unity, it make sense to deﬁne the time evolution operator T in such a manner
that it preserves the length of any ket upon which it acts (i.e., if a ket is prop
erly normalized at time t then it will remain normalized at all subsequent times
t > t
0
). This is always possible, since the length of a ket possesses no physical
signiﬁcance. Thus, we require that
¸At
0
At
0
) = ¸AtAt) (4.4)
for any ket A, which immediately yields
T
†
T = 1. (4.5)
Hence, the time evolution operator T is a unitary operator.
Up to now, the time evolution operator T looks very much like the spatial
displacement operator D introduced in the previous section. However, there are
some important differences between time evolution and spatial displacement.
In general, we do expect the expectation value of some observable ξ to evolve
with time, even if the system is left in a state of undisturbed motion (after all,
time evolution has no meaning unless something observable changes with time).
The triple product ¸AξA) can evolve either because the ket A) evolves and the
operator ξ stays constant, the ket A) stays constant and the operator ξ evolves,
or both the ket A) and the operator ξ evolve. Since we are already committed to
evolving state kets, according to Eq. (4.3), let us assume that the time evolution
operator T can be chosen in such a manner that the operators representing the
dynamical variables of the system do not evolve in time (unless they contain some
speciﬁc time dependence).
We expect, from physical continuity, that as t → t
0
then At) → At
0
) for any
ket A. Thus, the limit
lim
t→t
0
At) − At
0
)
t −t
0
= lim
t→t
0
T −1
t −t
0
At
0
) (4.6)
56
4.1 Schr¨ odinger’s equations of motion 4 QUANTUM DYNAMICS
should exist. Note that this limit is simply the derivative of At
0
) with respect to
t
0
. Let
τ(t
0
) = lim
t→t
0
T(t, t
0
) −1
t −t
0
. (4.7)
It is easily demonstrated from Eq. (4.5) that τ is antiHermitian: i.e.,
τ
†
+τ = 0. (4.8)
The fact that T can be replaced by T exp(i γ) (where γ is real) implies that τ is
undetermined to an arbitrary imaginary additive constant (see previous section).
Let us deﬁne the Hermitian operator H(t
0
) = i ¯hτ. This operator is undetermined
to an arbitrary real additive constant. It follows from Eqs. (4.6) and (4.7) that
i ¯h
dAt
0
)
dt
0
= i ¯h lim
t→t
0
At) − At
0
)
t −t
0
= i ¯hτ(t
0
)At
0
) = H(t
0
)At
0
). (4.9)
When written for general t this equation becomes
i ¯h
dAt)
dt
= H(t)At). (4.10)
Equation (4.10) gives the general law for the time evolution of a state ket in
a scheme in which the operators representing the dynamical variables remain
ﬁxed. This equation is denoted Schr¨ odinger’s equation of motion. It involves a
Hermitian operator H(t) which is, presumably, a characteristic of the dynamical
system under investigation.
We saw, in the previous section, that if the operator D(x, x
0
) displaces the
system along the xaxis from x
0
to x then
p
x
= i ¯h lim
x→x
0
D(x, x
0
) −1
x −x
0
, (4.11)
where p
x
is the operator representing the momentum conjugate to x. We now
have that if the operator T(t, t
0
) evolves the system in time from t
0
to t then
H(t
0
) = i ¯h lim
t→t
0
T(t, t
0
) −1
t −t
0
. (4.12)
57
4.1 Schr¨ odinger’s equations of motion 4 QUANTUM DYNAMICS
Thus, the dynamical variable corresponding to the operator H stands to time t as
the momentum p
x
stands to the coordinate x. By analogy with classical physics,
this suggests that H(t) is the operator representing the total energy of the system.
(Recall that, in classical physics, if the equations of motion of a system are in
variant under an xdisplacement of the system then this implies that the system
conserves momentum in the xdirection. Likewise, if the equations of motion are
invariant under a temporal displacement then this implies that the system con
serves energy.) The operator H(t) is usually called the Hamiltonian of the system.
The fact that the Hamiltonian is undetermined to an arbitrary real additive con
stant is related to the wellknown phenomenon that energy is undetermined to
an arbitrary additive constant in physics (i.e., the zero of potential energy is not
welldeﬁned).
Substituting At) = TAt
0
) into Eq. (4.10) yields
i ¯h
dT
dt
At
0
) = H(t) TAt
0
). (4.13)
Since this must hold for any initial state At
0
) we conclude that
i ¯h
dT
dt
= H(t) T. (4.14)
This equation can be integrated to give
T(t, t
0
) = exp
_
_
−i
_
t
t
0
H(t
) dt
/¯h
_
_
, (4.15)
where use has been made of Eqs. (4.5) and (4.6). (Here, we assume that Hamil
tonian operators evaluated at different times commute with one another). It is
now clear how the fact that H is undetermined to an arbitrary real additive con
stant leaves T undetermined to a phasefactor. Note that, in the above analysis,
time is not an operator (we cannot observe time, as such), it is just a parameter
(or, more accurately, a continuous label). Since we are only dealing with non
relativistic quantum mechanics, the fact that position is an operator, but time is
only a label, need not worry us unduly. In relativistic quantum mechanics, time
and space coordinates are treated on the same footing by relegating position from
being an operator to being just a label.
58
4.2 Heisenberg’s equations of motion 4 QUANTUM DYNAMICS
4.2 Heisenberg’s equations of motion
We have seen that in Schr¨ odinger’s scheme the dynamical variables of the system
remain ﬁxed during a period of undisturbed motion, whereas the state kets evolve
according to Eq. (4.10). However, this is not the only way in which to represent
the time evolution of the system.
Suppose that a general state ket A is subject to the transformation
A
t
) = T
†
(t, t
0
)A). (4.16)
This is a timedependent transformation, since the operator T(t, t
0
) obviously
depends on time. The subscript t is used to remind us that the transformation is
timedependent. The time evolution of the transformed state ket is given by
A
t
t) = T
†
(t, t
0
)At) = T
†
(t, t
0
) T(t, t
0
)At
0
) = A
t
t
0
), (4.17)
where use has been made of Eqs. (4.3), (4.5), and the fact that T(t
0
, t
0
) = 1.
Clearly, the transformed state ket does not evolve in time. Thus, the transforma
tion (4.16) has the effect of bringing all kets representing states of undisturbed
motion of the system to rest.
The transformation must also be applied to bras. The dual of Eq. (4.16) yields
¸A
t
 = ¸AT. (4.18)
The transformation rule for a general observable v is obtained from the require
ment that the expectation value ¸AvA) should remain invariant. It is easily seen
that
v
t
= T
†
v T. (4.19)
Thus, a dynamical variable, which corresponds to a ﬁxed linear operator in Schr¨ o
dinger’s scheme, corresponds to a moving linear operator in this new scheme. It
is clear that the transformation (4.16) leads us to a scenario in which the state
of the system is represented by a ﬁxed vector, and the dynamical variables are
represented by moving linear operators. This is termed the Heisenberg picture, as
opposed to the Schr¨ odinger picture, which is outlined in Sect. 4.1.
59
4.2 Heisenberg’s equations of motion 4 QUANTUM DYNAMICS
Consider a dynamical variable v corresponding to a ﬁxed linear operator in the
Schr¨ odinger picture. According to Eq. (4.19), we can write
T v
t
= v T. (4.20)
Differentiation with respect to time yields
dT
dt
v
t
+T
dv
t
dt
= v
dT
dt
. (4.21)
With the help of Eq. (4.14), this reduces to
HT v
t
+ i ¯hT
dv
t
dt
= v HT, (4.22)
or
i ¯h
dv
t
dt
= T
†
v HT −T
†
HT v
t
= v
t
H
t
−H
t
v
t
, (4.23)
where
H
t
= T
†
HT. (4.24)
Equation (4.23) can be written
i ¯h
dv
t
dt
= [v
t
, H
t
]. (4.25)
Equation (4.25) shows how the dynamical variables of the system evolve in
the Heisenberg picture. It is denoted Heisenberg’s equation of motion. Note that
the timevarying dynamical variables in the Heisenberg picture are usually called
Heisenberg dynamical variables to distinguish them from Schr¨ odinger dynamical
variables (i.e., the corresponding variables in the Schr¨ odinger picture), which do
not evolve in time.
According to Eq. (3.22), the Heisenberg equation of motion can be written
dv
t
dt
= [v
t
, H
t
]
quantum
, (4.26)
where [ ]
quantum
denotes the quantumPoisson bracket. Let us compare this equa
tion with the classical time evolution equation for a general dynamical variable
v, which can be written in the form [see Eq. (3.7)]
dv
dt
= [v, H]
classical
. (4.27)
60
4.3 Ehrenfest’s theorem 4 QUANTUM DYNAMICS
Here, [ ]
classical
is the classical Poisson bracket, and H denotes the classical
Hamiltonian. The strong resemblance between Eqs. (4.26) and (4.27) provides
us with further justiﬁcation for our identiﬁcation of the linear operator H with
the energy of the system in quantum mechanics.
Note that if the Hamiltonian does not explicitly depend on time (i.e., the sys
tem is not subject to some timedependent external force) then Eq. (4.15) yields
T(t, t
0
) = exp [−i H(t −t
0
)/¯h] . (4.28)
This operator manifestly commutes with H, so
H
t
= T
†
HT = H. (4.29)
Furthermore, Eq. (4.25) gives
i ¯h
dH
dt
= [H, H] = 0. (4.30)
Thus, if the energy of the system has no explicit timedependence then it is rep
resented by the same nontimevarying operator H in both the Schr¨ odinger and
Heisenberg pictures.
Suppose that v is an observable which commutes with the Hamiltonian (and,
hence, with the time evolution operator T). It follows from Eq. (4.19) that v
t
= v.
Heisenberg’s equation of motion yields
i ¯h
dv
dt
= [v, H] = 0. (4.31)
Thus, any observable which commutes with the Hamiltonian is a constant of the mo
tion (hence, it is represented by the same ﬁxed operator in both the Schr¨ odinger
and Heisenberg pictures). Only those observables which do not commute with
the Hamiltonian evolve in time in the Heisenberg picture.
4.3 Ehrenfest’s theorem
We have now derived all of the basic elements of quantum mechanics. The only
thing which is lacking is some rule to determine the form of the quantum mechan
ical Hamiltonian. For a physical system which possess a classical analogue, we
61
4.3 Ehrenfest’s theorem 4 QUANTUM DYNAMICS
generally assume that the Hamiltonian has the same form as in classical physics
(i.e., we replace the classical coordinates and conjugate momenta by the corre
sponding quantum mechanical operators). This scheme guarantees that quantum
mechanics yields the correct classical equations of motion in the classical limit.
Whenever an ambiguity arises because of noncommuting observables, this can
usually be resolved by requiring the Hamiltonian H to be an Hermitian oper
ator. For instance, we would write the quantum mechanical analogue of the
classical product x p, appearing in the Hamiltonian, as the Hermitian product
(1/2)(x p + px). When the system in question has no classical analogue then we
are reduced to guessing a form for H which reproduces the observed behaviour
of the system.
Consider a threedimensional systemcharacterized by three independent Carte
sian position coordinates x
i
(where i runs from 1 to 3), with three corresponding
conjugate momenta p
i
. These are represented by three commuting position op
erators x
i
, and three commuting momentum operators p
i
, respectively. The com
mutation relations satisﬁed by the position and momentum operators are [see
Eq. (3.25)]
[x
i
, p
j
] = i ¯hδ
ij
. (4.32)
It is helpful to denote (x
1
, x
2
, x
3
) as x and (p
1
, p
2
, p
3
) as p. The following useful
formulae,
[x
i
, F(p)] = i ¯h
∂F
∂p
i
, (4.33)
[p
i
, G(x)] = −i ¯h
∂G
∂x
i
, (4.34)
where F and G are functions which can be expanded as power series, are easily
proved using the fundamental commutation relations Eq. (4.32).
Let us now consider the threedimensional motion of a free particle of mass m
in the Heisenberg picture. The Hamiltonian is assumed to have the same form as
in classical physics:
H =
p
2
2 m
=
1
2 m
3
i=1
p
2
i
. (4.35)
62
4.3 Ehrenfest’s theorem 4 QUANTUM DYNAMICS
In the following, all dynamical variables are assumed to be Heisenberg dynamical
variables, although we will omit the subscript t for the sake of clarity. The time
evolution of the momentum operator p
i
follows from Heisenberg’s equation of
motion (4.25). We ﬁnd that
dp
i
dt
=
1
i ¯h
[p
i
, H] = 0, (4.36)
since p
i
automatically commutes with any function of the momentum operators.
Thus, for a free particle the momentum operators are constants of the motion,
which means that p
i
(t) = p
i
(0) at all times t (for i is 1 to 3). The time evolution
of the position operator x
i
is given by
dx
i
dt
=
1
i ¯h
[x
i
, H] =
1
i ¯h
1
2 m
i ¯h
∂
∂p
i
_
_
_
3
j=1
p
2
j
_
_
_ =
p
i
m
=
p
i
(0)
m
, (4.37)
where use has been made of Eq. (4.33). It follows that
x
i
(t) = x
i
(0) +
_
_
p
i
(0)
m
_
_
t, (4.38)
which is analogous to the equation of motion of a classical free particle. Note
that even though
[x
i
(0), x
j
(0)] = 0, (4.39)
where the position operators are evaluated at equal times, the x
i
do not commute
when evaluated at different times. For instance,
[x
i
(t), x
i
(0)] =
_
_
p
i
(0) t
m
, x
i
(0)
_
_
=
−i ¯ht
m
. (4.40)
Combining the above commutation relation with the uncertainty relation (2.83)
yields
¸(∆x
i
)
2
)
t
¸(∆x
i
)
2
)
t=0
≥
¯h
2
t
2
4 m
2
. (4.41)
This result implies that even if a particle is welllocalized at t = 0, its position
becomes progressively more uncertain with time. This conclusion can also be
obtained by studying the propagation of wavepackets in wave mechanics.
63
4.3 Ehrenfest’s theorem 4 QUANTUM DYNAMICS
Let us now add a potential V(x) to our free particle Hamiltonian:
H =
p
2
2 m
+V(x). (4.42)
Here, V is some function of the x
i
operators. Heisenberg’s equation of motion
gives
dp
i
dt
=
1
i ¯h
[p
i
, V(x)] = −
∂V(x)
∂x
i
, (4.43)
where use has been made of Eq. (4.34). On the other hand, the result
dx
i
dt
=
p
i
m
(4.44)
still holds, because the x
i
all commute with the new term V(x) in the Hamilto
nian. We can use the Heisenberg equation of motion a second time to deduce
that
d
2
x
i
dt
2
=
1
i ¯h
_
dx
i
dt
, H
_
=
1
i ¯h
_
p
i
m
, H
_
=
1
m
dp
i
dt
= −
1
m
∂V(x)
∂x
i
. (4.45)
In vectorial form, this equation becomes
m
d
2
x
dt
2
=
dp
dt
= −∇V(x). (4.46)
This is the quantum mechanical equivalent of Newton’s second law of motion.
Taking the expectation values of both sides with respect to a Heisenberg state ket
that does not move with time, we obtain
m
d
2
¸x)
dt
2
=
d¸p)
dt
= −¸∇V(x)). (4.47)
This is known as Ehrenfest’s theorem. When written in terms of expectation
values, this result is independent of whether we are using the Heisenberg or
Schr¨ odinger picture. In contrast, the operator equation (4.46) only holds if x and
p are understood to be Heisenberg dynamical variables. Note that Eq. (4.47) has
no dependence on ¯h. In fact, it guarantees to us that the centre of a wavepacket
always moves like a classical particle.
64
4.4 Schr¨ odinger’s waveequation 4 QUANTUM DYNAMICS
4.4 Schr¨ odinger’s waveequation
Let us nowconsider the motion of a particle in three dimensions in the Schr¨ odinger
picture. The ﬁxed dynamical variables of the system are the position operators
x ≡ (x
1
, x
2
, x
3
), and the momentum operators p ≡ (p
1
, p
2
, p
3
). The state of the
system is represented as some time evolving ket At).
Let x
) represent a simultaneous eigenket of the position operators belonging
to the eigenvalues x
≡ (x
1
, x
2
, x
3
). Note that, since the position operators are
ﬁxed in the Schr¨ odinger picture, we do not expect the x
) to evolve in time. The
wavefunction of the system at time t is deﬁned
ψ(x
, t) = ¸x
At). (4.48)
The Hamiltonian of the system is taken to be
H =
p
2
2 m
+V(x). (4.49)
Schr¨ odinger’s equation of motion (4.10) yields
i ¯h
∂¸x
At)
∂t
= ¸x
HAt), (4.50)
where use has been made of the time independence of the x
). We adopt Schr¨ od
inger’s representation in which the momentum conjugate to the position operator
x
i
is written [see Eq. (3.74)]
p
i
= −i ¯h
∂
∂x
i
. (4.51)
Thus,
_
x
¸
¸
¸
¸
¸
¸
p
2
2 m
¸
¸
¸
¸
¸
¸
At
_
= −
_
_
¯h
2
2 m
_
_
∇
2
¸x
At), (4.52)
where use has been made of Eq. (3.78). Here, ∇
≡ (∂/∂x
, ∂/∂y
, ∂/∂z
) denotes
the gradient operator written in terms of the position eigenvalues. We can also
write
¸x
V(x) = V(x
)¸x
, (4.53)
65
4.4 Schr¨ odinger’s waveequation 4 QUANTUM DYNAMICS
where V(x
) is a scalar function of the position eigenvalues. Combining Eqs. (4.49),
(4.50), (4.52), and (4.53), we obtain
i ¯h
∂¸x
At)
∂t
= −
_
_
¯h
2
2 m
_
_
∇
2
¸x
At) +V(x
)¸x
At), (4.54)
which can also be written
i ¯h
∂ψ(x
, t)
∂t
= −
_
_
¯h
2
2 m
_
_
∇
2
ψ(x
, t) +V(x
) ψ(x
, t). (4.55)
This is Schr¨ odinger’s famous waveequation, and is the basis of wave mechanics.
Note, however, that the waveequation is just one of many possible representa
tions of quantum mechanics. It just happens to give a type of equation which we
know how to solve. In deriving the waveequation, we have chosen to represent
the system in terms of the eigenkets of the position operators, instead of those
of the momentum operators. We have also ﬁxed the relative phases of the x
)
according to Schr¨ odinger’s representation, so that Eq. (4.51) is valid. Finally, we
have chosen to work in the Schr¨ odinger picture, in which state kets evolve and
dynamical variables are ﬁxed, instead of the Heisenberg picture, in which the
opposite is true.
Suppose that the ket At) is an eigenket of the Hamiltonian belonging to the
eigenvalue H
:
HAt) = H
At). (4.56)
Schr¨ odinger’s equation of motion (4.10) yields
i ¯h
dAt)
dt
= H
At). (4.57)
This can be integrated to give
At) = exp[−i H
(t −t
0
)/¯h]At
0
). (4.58)
Note that At) only differs from At
0
) by a phasefactor. The direction of the
vector remains ﬁxed in ket space. This suggests that if the system is initially in an
eigenstate of the Hamiltonian then it remains in this state for ever, as long as the
66
4.4 Schr¨ odinger’s waveequation 4 QUANTUM DYNAMICS
system is undisturbed. Such a state is called a stationary state. The wavefunction
of a stationary state satisﬁes
ψ(x
, t) = ψ(x
, t
0
) exp[−i H
(t −t
0
)/¯h]. (4.59)
Substituting the above relation into Schr¨ odinger’s wave equation (4.55), we
obtain
−
_
_
¯h
2
2 m
_
_
∇
2
ψ
0
(x
) + (V(x
) −E) ψ
0
(x
) = 0, (4.60)
where ψ
0
(x
) ≡ ψ(x
, t
0
), and E = H
is the energy of the system. This is
Schr¨ odinger’s timeindependent waveequation. A bound state solution of the
above equation, in which the particle is conﬁned within a ﬁnite region of space,
satisﬁes the boundary condition
ψ
0
(x
) →0 as x
 →∞. (4.61)
Such a solution is only possible if
E < lim
x
→∞
V(x
). (4.62)
Since it is conventional to set the potential at inﬁnity equal to zero, the above
relation implies that bound states are equivalent to negative energy states. The
boundary condition (4.61) is sufﬁcient to uniquely specify the solution of Eq. (4.60).
The quantity ρ(x
, t), deﬁned by
ρ(x
, t) = ψ(x
, t)
2
, (4.63)
is termed the probability density. Recall, from Eq. (3.30), that the probability of
observing the particle in some volume element d
3
x
around position x
is propor
tional to ρ(x
, t) d
3
x
. The probability is equal to ρ(x
, t) d
3
x
if the wavefunction
is properly normalized, so that
_
ρ(x
, t) d
3
x
= 1. (4.64)
Schr¨ odinger’s timedependent waveequation, (4.55), can easily be written in
the form of a conservation equation for the probability density:
∂ρ
∂t
+∇
j = 0. (4.65)
67
4.4 Schr¨ odinger’s waveequation 4 QUANTUM DYNAMICS
The probability current j takes the form
j(x
, t) = −
_
i ¯h
2 m
_
[ψ
∗
∇
ψ− (∇
ψ
∗
) ψ] =
_
¯h
m
_
Im(ψ
∗
∇
ψ). (4.66)
We can integrate Eq. (4.65) over all space, using the divergence theorem, and the
boundary condition ρ →0 as x
 →∞, to obtain
∂
∂t
_
ρ(x
, t) d
3
x
= 0. (4.67)
Thus, Schr¨ odinger’s waveequation conserves probability. In particular, if the
wavefunction starts off properly normalized, according to Eq. (4.64), then it
remains properly normalized at all subsequent times. It is easily demonstrated
that
_
j(x
, t) d
3
x
=
¸p)
t
m
, (4.68)
where ¸p)
t
denotes the expectation value of the momentum evaluated at time t.
Clearly, the probability current is indirectly related to the particle momentum.
In deriving Eq. (4.65) we have, naturally, assumed that the potential V(x
) is
real. Suppose, however, that the potential has an imaginary component. In this
case, Eq. (4.65) generalizes to
∂ρ
∂t
+∇
j =
2 Im(V)
¯h
ρ, (4.69)
giving
∂
∂t
_
ρ(x
, t) d
3
x
=
2
¯h
Im
_
V(x
) ρ(x
, t) d
3
x
. (4.70)
Thus, if Im(V) < 0 then the total probability of observing the particle anywhere
in space decreases monotonically with time. Thus, an imaginary potential can
be used to account for the disappearance of a particle. Such a potential is often
employed to model nuclear reactions in which incident particles can be absorbed
by nuclei.
The wavefunction can always be written in the form
ψ(x
, t) =
_
ρ(x
, t) exp
_
_
i S(x
, t)
¯h
_
_
, (4.71)
68
4.4 Schr¨ odinger’s waveequation 4 QUANTUM DYNAMICS
where ρ and S are both real functions. The interpretation of ρ as a probability
density has already been given. What is the interpretation of S? Note that
ψ
∗
∇
ψ =
√
ρ ∇
(
√
ρ) +
_
i
¯h
_
ρ∇
S. (4.72)
It follows from Eq. (4.66) that
j =
ρ ∇
S
m
. (4.73)
Thus, the gradient of the phase of the wavefunction determines the direction of
the probability current. In particular, the probability current is locally normal to
the contours of the phasefunction S.
Let us substitute Eq. (4.71) into Schr¨ odinger’s timedependent waveequation.
We obtain
−
1
2 m
_
¯h
2
∇
2
√
ρ +2i ¯h∇
(
√
ρ)∇
S −
√
ρ ∇
S
2
+ i ¯h
√
ρ ∇
2
S
_
+
√
ρ V
=
_
i ¯h
∂
√
ρ
∂t
−
√
ρ
∂S
∂t
_
. (4.74)
Let us treat ¯h as a small quantity. To lowest order, Eq. (4.74) yields
−
∂S(x
, t)
∂t
=
1
2 m
∇
S(x
, t)
2
+V(x
, t) = H(x
, ∇
S, t), (4.75)
where H(x, p, t) is the Hamiltonian operator. The above equation is known as the
HamiltonJacobi equation, and is one of the many forms in which we can write
the equations of classical mechanics. In classical mechanics, S is the action (i.e.,
the pathintegral of the Lagrangian). Thus, in the limit ¯h → 0, wave mechanics
reduces to classical mechanics. It is a good approximation to neglect the terms
involving ¯h in Eq. (4.74) provided that
¯h∇
2
S ¸∇
S
2
. (4.76)
Note that, according to Eq. (4.71),
¯ λ =
¯h
∇
S
, (4.77)
69
4.4 Schr¨ odinger’s waveequation 4 QUANTUM DYNAMICS
where ¯ λ is the de Broglie wavelength divided by 2π. The inequality (4.76) is
equivalent to
∇
¯ λ ¸1. (4.78)
In other words, quantum mechanics reduces to classical mechanics whenever
the de Broglie wavelength is small compared to the characteristic distance over
which things (other than the quantum phase) vary. This distance is usually set by
the variation scalelength of the potential.
70
5 ANGULAR MOMENTUM
5 Angular momentum
5.1 Orbital angular momentum
Consider a particle described by the Cartesian coordinates (x, y, z) ≡ r and their
conjugate momenta (p
x
, p
y
, p
z
) ≡ p. The classical deﬁnition of the orbital angular
momentum of such a particle about the origin is L = r p, giving
L
x
= yp
z
−z p
y
, (5.1)
L
y
= z p
x
−x p
z
, (5.2)
L
z
= x p
y
−yp
x
. (5.3)
Let us assume that the operators (L
x
, L
y
, L
z
) ≡ L which represent the compo
nents of orbital angular momentum in quantum mechanics can be deﬁned in an
analogous manner to the corresponding components of classical angular momen
tum. In other words, we are going to assume that the above equations specify
the angular momentum operators in terms of the position and linear momentum
operators. Note that L
x
, L
y
, and L
z
are Hermitian, so they represent things which
can, in principle, be measured. Note, also, that there is no ambiguity regard
ing the order in which operators appear in products on the righthand sides of
Eqs. (5.1)–(5.3), since all of the products consist of operators which commute.
The fundamental commutation relations satisﬁed by the position and linear
momentum operators are [see Eqs. (3.23)–(3.25)]
[x
i
, x
j
] = 0, (5.4)
[p
i
, p
j
] = 0, (5.5)
[x
i
, p
j
] = i ¯hδ
ij
, (5.6)
where i and j stand for either x, y, or z. Consider the commutator of the operators
L
x
and L
z
:
[L
x
, L
y
] = [(yp
z
−z p
y
), (z p
x
−x p
z
)] = y[p
z
, z] p
x
+x p
y
[z, p
z
]
= i ¯h(−yp
x
+x p
y
) = i ¯hL
z
. (5.7)
71
5.1 Orbital angular momentum 5 ANGULAR MOMENTUM
The cyclic permutations of the above result yield the fundamental commutation
relations satisﬁed by the components of an angular momentum:
[L
x
, L
y
] = i ¯hL
z
, (5.8)
[L
y
, L
z
] = i ¯hL
x
, (5.9)
[L
z
, L
x
] = i ¯hL
y
. (5.10)
These can be summed up more succinctly by writing
L L = i ¯hL. (5.11)
The three commutation relations (5.8)–(5.10) are the foundation for the whole
theory of angular momentum in quantum mechanics. Whenever we encounter
three operators having these commutation relations, we know that the dynamical
variables which they represent have identical properties to those of the compo
nents of an angular momentum (which we are about to derive). In fact, we shall
assume that any three operators which satisfy the commutation relations (5.8)–
(5.10) represent the components of an angular momentum.
Suppose that there are N particles in the system, with angular momentum
vectors L
i
(where i runs from 1 to N). Each of these vectors satisﬁes Eq. (5.11),
so that
L
i
L
i
= i ¯hL
i
. (5.12)
However, we expect the angular momentum operators belonging to different par
ticles to commute, since they represent different degrees of freedom of the sys
tem. So, we can write
L
i
L
j
+L
j
L
i
= 0, (5.13)
for i ,= j. Consider the total angular momentum of the system, L =
N
i=1
L
i
. It is
clear from Eqs. (5.12) and (5.13) that
L L =
N
i=1
L
i
N
j=1
L
j
=
N
i=1
L
i
L
i
+
1
2
N
i,j=1
(L
i
L
j
+L
j
L
i
)
= i ¯h
N
i=1
L
i
= i ¯hL. (5.14)
72
5.1 Orbital angular momentum 5 ANGULAR MOMENTUM
Thus, the sum of two or more angular momentum vectors satisﬁes the same
commutation relation as a primitive angular momentum vector. In particular,
the total angular momentum of the system satisﬁes the commutation relation
(5.11).
The immediate conclusion which can be drawn from the commutation rela
tions (5.8)–(5.10) is that the three components of an angular momentum vector
cannot be speciﬁed (or measured) simultaneously. In fact, once we have speci
ﬁed one component, the values of other two components become uncertain. It is
conventional to specify the zcomponent, L
z
.
Consider the magnitude squared of the angular momentum vector, L
2
≡ L
2
x
+
L
2
y
+L
2
z
. The commutator of L
2
and L
z
is written
[L
2
, L
z
] = [L
2
x
, L
z
] + [L
2
y
, L
z
] + [L
2
z
, L
z
]. (5.15)
It is easily demonstrated that
[L
2
x
, L
z
] = −i ¯h(L
x
L
y
+L
y
L
x
), (5.16)
[L
2
y
, L
z
] = +i ¯h(L
x
L
y
+L
y
L
x
), (5.17)
[L
2
z
, L
z
] = 0, (5.18)
so
[L
2
, L
z
] = 0. (5.19)
Since there is nothing special about the zaxis, we conclude that L
2
also commutes
with L
x
and L
y
. It is clear from Eqs. (5.8)–(5.10) and (5.19) that the best we can
do in quantum mechanics is to specify the magnitude of an angular momentum
vector along with one of its components (by convention, the zcomponent).
It is convenient to deﬁne the shift operators L
+
and L
−
:
L
+
= L
x
+ i L
y
, (5.20)
L
−
= L
x
− i L
y
. (5.21)
Note that
[L
+
, L
z
] = −¯hL
+
, (5.22)
73
5.2 Eigenvalues of angular momentum 5 ANGULAR MOMENTUM
[L
−
, L
z
] = +¯hL
−
, (5.23)
[L
+
, L
−
] = 2 ¯hL
z
. (5.24)
Note, also, that both shift operators commute with L
2
.
5.2 Eigenvalues of angular momentum
Suppose that the simultaneous eigenkets of L
2
and L
z
are completely speciﬁed by
two quantum numbers, l and m. These kets are denoted l, m). The quantum
number m is deﬁned by
L
z
l, m) = m¯hl, m). (5.25)
Thus, m is the eigenvalue of L
z
divided by ¯h. It is possible to write such an
equation because ¯h has the dimensions of angular momentum. Note that m is a
real number, since L
z
is an Hermitian operator.
We can write
L
2
l, m) = f(l, m) ¯h
2
l, m), (5.26)
without loss of generality, where f(l, m) is some real dimensionless function of l
and m. Later on, we will show that f(l, m) = l (l +1). Now,
¸l, mL
2
−L
2
z
l, m) = ¸l, mf(l, m) ¯h
2
−m
2
¯h
2
l, m) = [f(l, m) −m
2
]¯h
2
, (5.27)
assuming that the l, m) have unit norms. However,
¸l, mL
2
−L
2
z
l, m) = ¸l, mL
2
x
+L
2
y
l, m)
= ¸l, mL
2
x
l, m) +¸l, mL
2
y
l, m). (5.28)
It is easily demonstrated that
¸Aξ
2
A) ≥ 0, (5.29)
where A) is a general ket, and ξ is an Hermitian operator. The proof follows
from the observation that
¸Aξ
2
A) = ¸Aξ
†
ξA) = ¸BB), (5.30)
74
5.2 Eigenvalues of angular momentum 5 ANGULAR MOMENTUM
where B) = ξA), plus the fact that ¸BB) ≥ 0 for a general ket B) [see Eq. (2.21)].
It follows from Eqs. (5.27)–(5.29) that
m
2
≤ f(l, m). (5.31)
Consider the effect of the shift operator L
+
on the eigenket l, m). It is easily
demonstrated that
L
2
(L
+
l, m)) = ¯h
2
f(l, m) (L
+
l, m)), (5.32)
where use has been made of Eq. (5.26), plus the fact that L
2
and L
z
commute. It
follows that the ket L
+
l, m) has the same eigenvalue of L
2
as the ket l, m). Thus,
the shift operator L
+
does not affect the magnitude of the angular momentum of
any eigenket it acts upon. Note that
L
z
L
+
l, m) = (L
+
L
z
+ [L
z
, L
+
])l, m) = (L
+
L
z
+ ¯hL
+
)l, m)
= (m+1) ¯hL
+
l, m), (5.33)
where use has been made of Eq. (5.22). The above equation implies that L
+
l, m)
is proportional to l, m+1). We can write
L
+
l, m) = c
+
l,m
¯hl, m+1), (5.34)
where c
+
l,m
is a number. It is clear that when the operator L
+
acts on a simulta
neous eigenstate of L
2
and L
z
, the eigenvalue of L
2
remains unchanged, but the
eigenvalue of L
z
is increased by ¯h. For this reason, L
+
is called a raising operator.
Using similar arguments to those given above, it is possible to demonstrate
that
L
−
l, m) = c
−
l,m
¯hl, m−1). (5.35)
Hence, L
−
is called a lowering operator.
The shift operators step the value of m up and down by unity each time they
operate on one of the simultaneous eigenkets of L
2
and L
z
. It would appear, at
ﬁrst sight, that any value of m can be obtained by applying the shift operators a
sufﬁcient number of times. However, according to Eq. (5.31), there is a deﬁnite
upper bound to the values that m
2
can take. This bound is determined by the
75
5.2 Eigenvalues of angular momentum 5 ANGULAR MOMENTUM
eigenvalue of L
2
[see Eq. (5.26)]. It follows that there is a maximum and a
minimum possible value which m can take. Suppose that we attempt to raise
the value of m above its maximum value m
max
. Since there is no state with
m > m
max
, we must have
L
+
l, m
max
) = 0). (5.36)
This implies that
L
−
L
+
l, m
max
) = 0). (5.37)
However,
L
−
L
+
= L
2
x
+L
2
y
+ i [L
x
, L
y
] = L
2
−L
2
z
− ¯hL
z
, (5.38)
so Eq. (5.37) yields
(L
2
−L
2
z
− ¯hL
z
)l, m
max
) = 0). (5.39)
The above equation can be rearranged to give
L
2
l, m
max
) = (L
2
z
+ ¯hL
z
)l, m
max
) = m
max
(m
max
+1) ¯h
2
l, m
max
). (5.40)
Comparison of this equation with Eq. (5.26) yields the result
f(l, m
max
) = m
max
(m
max
+1). (5.41)
But, when L
−
operates on n, m
max
) it generates n, m
max
− 1), n, m
max
− 2), etc.
Since the lowering operator does not change the eigenvalue of L
2
, all of these
states must correspond to the same value of f, namely m
max
(m
max
+1). Thus,
L
2
l, m) = m
max
(m
max
+1) ¯h
2
l, m). (5.42)
At this stage, we can give the unknown quantum number l the value m
max
, with
out loss of generality. We can also write the above equation in the form
L
2
l, m) = l (l +1) ¯h
2
l, m). (5.43)
It is easily seen that
L
−
L
+
l, m) = (L
2
−L
2
z
− ¯hL
z
)l, m) = ¯h
2
[l (l +1) −m(m+1)]l, m). (5.44)
Thus,
¸l, mL
−
L
+
l, m) = ¯h
2
[l (l +1) −m(m+1)]. (5.45)
76
5.2 Eigenvalues of angular momentum 5 ANGULAR MOMENTUM
However, we also know that
¸l, mL
−
L
+
l, m) = ¸l, mL
−
¯hc
+
l,m
l, m+1) = ¯h
2
c
+
l,m
c
−
l,m+1
, (5.46)
where use has been made of Eqs. (5.34) and (5.35). It follows that
c
+
l,m
c
−
l,m+1
= [l (l +1) −m(m+1)]. (5.47)
Consider the following:
¸l, mL
−
l, m+1) = ¸l, mL
x
l, m+1) − i ¸l, mL
y
l, m+1)
= ¸l, m+1L
x
l, m)
∗
− i ¸l, m+1L
y
l, m)
∗
= (¸l, m+1L
x
l, m) + i ¸l, m+1L
y
l, m))
∗
= ¸l, m+1L
+
l, m)
∗
, (5.48)
where use has been made of the fact that L
x
and L
y
are Hermitian. The above
equation reduces to
c
−
l,m+1
= (c
+
l,m
)
∗
(5.49)
with the aid of Eqs. (5.34) and (5.35).
Equations (5.47) and (5.49) can be combined to give
c
+
l,m

2
= [l (l +1) −m(m+1)]. (5.50)
The solution of the above equation is
c
+
l,m
=
_
l (l +1) −m(m+1). (5.51)
Note that c
+
l,m
is undetermined to an arbitrary phasefactor [i.e., we can replace
c
+
l,m
, given above, by c
+
l,m
exp(i γ), where γ is real, and we still satisfy Eq. (5.50)].
We have made the arbitrary, but convenient, choice that c
+
l,m
is real and posi
tive. This is equivalent to choosing the relative phases of the eigenkets l, m).
According to Eq. (5.49),
c
−
l,m
= (c
+
l,m−1
)
∗
=
_
l (l +1) −m(m−1). (5.52)
We have already seen that the inequality (5.31) implies that there is a maxi
mum and a minimum possible value of m. The maximum value of m is denoted
77
5.3 Rotation operators 5 ANGULAR MOMENTUM
l. What is the minimum value? Suppose that we try to lower the value of m
below its minimum value m
min
. Since there is no state with m < m
min
, we must
have
L
−
l, m
min
) = 0. (5.53)
According to Eq. (5.35), this implies that
c
−
l,m
min
= 0. (5.54)
It can be seen from Eq. (5.52) that m
min
= −l. We conclude that m can take a
“ladder” of discrete values, each rung differing from its immediate neighbours by
unity. The top rung is l, and the bottom rung is −l. There are only two possible
choices for l. Either it is an integer (e.g., l = 2, which allows m to take the values
−2, −1, 0, 1, 2), or it is a halfinteger (e.g., l = 3/2, which allows m to take the
values −3/2, −1/2, 1/2, 3/2). We will prove in the next section that an orbital
angular momentum can only take integer values of l.
In summary, using just the fundamental commutation relations (5.8)–(5.10),
plus the fact that L
x
, L
y
, and L
z
are Hermitian operators, we have shown that the
eigenvalues of L
2
≡ L
2
x
+L
2
y
+L
2
z
can be written l (l +1) ¯h
2
, where l is an integer,
or a halfinteger. We have also demonstrated that the eigenvalues of L
z
can only
take the values m¯h, where m lies in the range −l, −l + 1, l − 1, l. Let l, m)
denote a properly normalized simultaneous eigenket of L
2
and L
z
, belonging to
the eigenvalues l (l +1) ¯h
2
and m¯h, respectively. We have shown that
L
+
l, m) =
_
l (l +1) −m(m+1) ¯hl, m+1) (5.55)
L
−
l, m) =
_
l (l +1) −m(m−1) ¯hl, m−1), (5.56)
where L
±
= L
x
±i L
y
are the socalled shift operators.
5.3 Rotation operators
Consider a particle described by the spherical polar coordinates (r, θ, ϕ). The
classical momentum conjugate to the azimuthal angle ϕ is the zcomponent of
angular momentum, L
z
. According to Sect. 3.5, in quantum mechanics we can
78
5.3 Rotation operators 5 ANGULAR MOMENTUM
always adopt Schr¨ odinger’s representation, for which ket space is spanned by the
simultaneous eigenkets of the position operators r, θ, and φ, and L
z
takes the
form
L
z
= −i ¯h
∂
∂ϕ
. (5.57)
We can do this because there is nothing in Sect. 3.5 which speciﬁes that we
have to use Cartesian coordinates—the representation (3.74) works for any well
deﬁned set of coordinates.
Consider an operator R(∆ϕ) which rotates the system an angle ∆ϕabout the z
axis. This operator is very similar to the operator D(∆x), introduced in Sect. 3.8,
which translates the system a distance ∆x along the x axis. We were able to
demonstrate in Sect. 3.8 that
p
x
= i ¯h lim
δx→0
D(δx) −1
δx
, (5.58)
where p
x
is the linear momentum conjugate to x. There is nothing in our deriva
tion of this result which speciﬁes that x has to be a Cartesian coordinate. Thus,
the result should apply just as well to an angular coordinate. We conclude that
L
z
= i ¯h lim
δϕ→0
R(δϕ) −1
δϕ
. (5.59)
According to Eq. (5.59), we can write
R(δϕ) = 1 − i L
z
δϕ/¯h (5.60)
in the limit δϕ → 0. In other words, the angular momentum operator L
z
can be
used to rotate the system about the zaxis by an inﬁnitesimal amount. We say
that L
z
is the generator of rotations about the zaxis. The above equation implies
that
R(∆ϕ) = lim
N→∞
_
1 − i
∆ϕ
N
L
z
¯h
_
N
, (5.61)
which reduces to
R(∆ϕ) = exp(−i L
z
∆ϕ/¯h). (5.62)
79
5.3 Rotation operators 5 ANGULAR MOMENTUM
Note that R(∆ϕ) has all of the properties we would expect of a rotation operator
R(0) = 1, (5.63)
R(∆ϕ) R(−∆ϕ) = 1, (5.64)
R(∆ϕ
1
) R(∆ϕ
2
) = R(∆ϕ
1
+∆ϕ
2
). (5.65)
Suppose that the system is in a simultaneous eigenstate of L
2
and L
z
. As before,
this state is represented by the eigenket l, m), where the eigenvalue of L
2
is
l (l + 1) ¯h
2
, and the eigenvalue of L
z
is m¯h. We expect the wavefunction to
remain unaltered if we rotate the system 2π degrees about the zaxis. Thus,
R(2π)l, m) = exp(−i L
z
2π/¯h)l, m) = exp(−i 2 πm)l, m) = l, m). (5.66)
We conclude that m must be an integer. This implies, from the previous section,
that l must also be an integer. Thus, orbital angular momentum can only take on
integer values of the quantum numbers l and m.
Consider the action of the rotation operator R(∆ϕ) on an eigenstate possessing
zero angular momentum about the zaxis (i.e., an m = 0 state). We have
R(∆ϕ)l, 0) = exp(0)l, 0) = l, 0). (5.67)
Thus, the eigenstate is invariant to rotations about the zaxis. Clearly, its wave
function must be symmetric about the zaxis.
There is nothing special about the zaxis, so we can write
R
x
(∆ϕ
x
) = exp(−i L
x
∆ϕ
x
/¯h), (5.68)
R
y
(∆ϕ
y
) = exp(−i L
y
∆ϕ
y
/¯h), (5.69)
R
z
(∆ϕ
y
) = exp(−i L
z
∆ϕ
z
/¯h), (5.70)
by analogy with Eq. (5.62). Here, R
x
(∆ϕ
x
) denotes an operator which rotates the
system by an angle ∆ϕ
x
about the xaxis, etc. Suppose that the system is in an
eigenstate of zero overall orbital angular momentum (i.e., an l = 0 state). We
know that the system is also in an eigenstate of zero orbital angular momentum
about any particular axis. This follows because l = 0 implies m = 0, according
80
5.4 Eigenfunctions of orbital angular momentum 5 ANGULAR MOMENTUM
to the previous section, and we can choose the zaxis to point in any direction.
Thus,
R
x
(∆ϕ
x
)0, 0) = exp(0)0, 0) = 0, 0), (5.71)
R
y
(∆ϕ
y
)0, 0) = exp(0)0, 0) = 0, 0), (5.72)
R
z
(∆ϕ
z
)0, 0) = exp(0)0, 0) = 0, 0). (5.73)
Clearly, a zero angular momentum state is invariant to rotations about any axis.
Such a state must possess a spherically symmetric wavefunction.
Note that a rotation about the xaxis does not commute with a rotation about
the yaxis. In other words, if the system is rotated an angle ∆ϕ
x
about the xaxis,
and then ∆ϕ
y
about the yaxis, it ends up in a different state to that obtained
by rotating an angle ∆ϕ
y
about the yaxis, and then ∆ϕ
x
about the xaxis. In
quantum mechanics, this implies that R
y
(∆ϕ
y
) R
x
(∆ϕ
x
) ,= R
x
(∆ϕ
x
) R
y
(∆ϕ
y
), or
L
y
L
x
,= L
x
L
y
, [see Eqs. (5.68)–(5.70)]. Thus, the noncommuting nature of the
angular momentum operators is a direct consequence of the fact that rotations
do not commute.
5.4 Eigenfunctions of orbital angular momentum
In Cartesian coordinates, the three components of orbital angular momentum can
be written
L
x
= −i ¯h
_
y
∂
∂z
−z
∂
∂y
_
, (5.74)
L
y
= −i ¯h
_
z
∂
∂x
−x
∂
∂z
_
, (5.75)
L
z
= −i ¯h
_
x
∂
∂y
−y
∂
∂x
_
, (5.76)
using the Schr¨ odinger representation. Transforming to standard spherical polar
coordinates,
x = r sinθ cos ϕ, (5.77)
81
5.4 Eigenfunctions of orbital angular momentum 5 ANGULAR MOMENTUM
y = r sinθ sinϕ, (5.78)
z = r cos θ, (5.79)
we obtain
L
x
= i ¯h
_
sinϕ
∂
∂θ
+ cot θcos ϕ
∂
∂ϕ
_
(5.80)
L
y
= −i ¯h
_
cos ϕ
∂
∂θ
− cot θsinϕ
∂
∂ϕ
_
(5.81)
L
z
= −i ¯h
∂
∂ϕ
. (5.82)
Note that Eq. (5.82) accords with Eq. (5.57). The shift operators L
±
= L
x
± i L
y
become
L
±
= ±¯h exp(±i ϕ)
_
∂
∂θ
±i cot θ
∂
∂ϕ
_
. (5.83)
Now,
L
2
= L
2
x
+L
2
y
+L
2
z
= L
2
z
+ (L
+
L
−
+L
−
L
+
)/2, (5.84)
so
L
2
= −¯h
2
_
_
1
sinθ
∂
∂θ
sinθ
∂
∂θ
+
1
sin
2
θ
∂
2
∂ϕ
2
_
_
. (5.85)
The eigenvalue problem for L
2
takes the form
L
2
ψ = λ ¯h
2
ψ, (5.86)
where ψ(r, θ, ϕ) is the wavefunction, and λ is a number. Let us write
ψ(r, θ, ϕ) = R(r) Y(θ, ϕ). (5.87)
Equation (5.86) reduces to
_
_
1
sinθ
∂
∂θ
sinθ
∂
∂θ
+
1
sin
2
θ
∂
2
∂ϕ
2
_
_
Y +λ Y = 0, (5.88)
where use has been made of Eq. (5.85). As is wellknown, square integrable
solutions to this equation only exist when λ takes the values l (l + 1), where l is
an integer. These solutions are known as spherical harmonics, and can be written
Y
m
l
(θ, ϕ) =
¸
¸
¸
¸
_
2 l +1
4π
(l −m)!
(l +m)!
(−1)
m
e
i mϕ
P
m
l
(cos ϕ), (5.89)
82
5.4 Eigenfunctions of orbital angular momentum 5 ANGULAR MOMENTUM
where m is a positive integer lying in the range 0 ≤ m ≤ l. Here, P
m
l
(ξ) is an
associated Legendre function satisfying the equation
d
dξ
_
(1 −ξ
2
)
dP
m
l
dξ
_
−
m
2
1 −ξ
2
P
m
l
+l (l +1) P
m
l
= 0. (5.90)
We deﬁne
Y
−m
l
= (−1)
m
(Y
m
l
)
∗
, (5.91)
which allows m to take the negative values −l ≤ m < 0. The spherical harmonics
are orthogonal functions, and are properly normalized with respect to integration
over the entire solid angle:
_
π
0
_
2π
0
Y
m∗
l
(θ, ϕ) Y
m
l
(θ, ϕ) sinθdθdϕ = δ
ll
δ
mm
. (5.92)
The spherical harmonics also form a complete set for representing general func
tions of θ and ϕ.
By deﬁnition,
L
2
Y
m
l
= l (l +1) ¯h
2
Y
m
l
, (5.93)
where l is an integer. It follows from Eqs. (5.82) and (5.89) that
L
z
Y
m
l
= m¯hY
m
l
, (5.94)
where m is an integer lying in the range −l ≤ m ≤ l. Thus, the wavefunction
ψ(r, θ, ϕ) = R(r) Y
m
l
(θ, φ), where R is a general function, has all of the expected
features of the wavefunction of a simultaneous eigenstate of L
2
and L
z
belonging
to the quantum numbers l and m. The wellknown formula
dP
m
l
dξ
=
1
√
1 −ξ
2
P
m+1
l
−
mξ
1 −ξ
2
P
m
l
= −
(l +m)(l −m+1)
√
1 −ξ
2
P
m−1
l
+
mξ
1 −ξ
2
P
m
l
(5.95)
can be combined with Eqs. (5.83) and (5.89) to give
L
+
Y
m
l
=
_
l (l +1) −m(m+1) ¯hY
m+1
l
, (5.96)
L
−
Y
m
l
=
_
l (l +1) −m(m−1) ¯hY
m−1
l
. (5.97)
83
5.5 Motion in a central ﬁeld 5 ANGULAR MOMENTUM
These equations are equivalent to Eqs. (5.55)–(5.56). Note that a spherical har
monic wavefunction is symmetric about the zaxis (i.e., independent of ϕ) when
ever m = 0, and is spherically symmetric whenever l = 0 (since Y
0
0
= 1/
√
4π).
In summary, by solving directly for the eigenfunctions of L
2
and L
z
in Schr¨ od
inger’s representation, we have been able to reproduce all of the results of Sect. 5.2.
Nevertheless, the results of Sect. 5.2 are more general than those obtained in this
section, because they still apply when the quantum number l takes on halfinteger
values.
5.5 Motion in a central ﬁeld
Consider a particle of mass M moving in a spherically symmetric potential. The
Hamiltonian takes the form
H =
p
2
2 M
+V(r). (5.98)
Adopting Schr¨ odinger’s representation, we can write p = −(i/¯h)∇. Hence,
H = −
¯h
2
2 M
∇
2
+V(r). (5.99)
When written in spherical polar coordinates, the above equation becomes
H = −
¯h
2
2 M
_
_
1
r
2
∂
∂r
r
2
∂
∂r
+
1
r
2
sinθ
∂
∂θ
sinθ
∂
∂θ
+
1
r
2
sin
2
θ
∂
2
∂ϕ
2
_
_
+V(r). (5.100)
Comparing this equation with Eq. (5.85), we ﬁnd that
H =
¯h
2
2 M
_
_
−
1
r
2
∂
∂r
r
2
∂
∂r
+
L
2
¯h
2
r
2
_
_
+V(r). (5.101)
Now, we know that the three components of angular momentum commute
with L
2
(see Sect. 5.1). We also know, from Eqs. (5.80)–(5.82), that L
x
, L
y
,
and L
z
take the form of partial derivative operators of the angular coordinates,
84
5.5 Motion in a central ﬁeld 5 ANGULAR MOMENTUM
when written in terms of spherical polar coordinates using Schr¨ odinger’s repre
sentation. It follows from Eq. (5.101) that all three components of the angular
momentum commute with the Hamiltonian:
[L, H] = 0. (5.102)
It is also easily seen that L
2
commutes with the Hamiltonian:
[L
2
, H] = 0. (5.103)
According to Sect. 4.2, the previous two equations ensure that the angular mo
mentum L and its magnitude squared L
2
are both constants of the motion. This
is as expected for a spherically symmetric potential.
Consider the energy eigenvalue problem
Hψ = Eψ, (5.104)
where E is a number. Since L
2
and L
z
commute with each other and the Hamil
tonian, it is always possible to represent the state of the system in terms of the
simultaneous eigenstates of L
2
, L
z
, and H. But, we already know that the most
general form for the wavefunction of a simultaneous eigenstate of L
2
and L
z
is
(see previous section)
ψ(r, θ, ϕ) = R(r) Y
m
l
(θ, ϕ). (5.105)
Substituting Eq. (5.105) into Eq. (5.101), and making use of Eq. (5.93), we ob
tain
_
_
¯h
2
2 M
_
_
−
1
r
2
d
dr
r
2
d
dr
+
l (l +1)
r
2
_
_
+V(r) −E
_
_
R = 0. (5.106)
This is a SturmLiouville equation for the function R(r). We know, from the gen
eral properties of this type of equation, that if R(r) is required to be wellbehaved
at r = 0 and as r → ∞ then solutions only exist for a discrete set of values of E.
These are the energy eigenvalues. In general, the energy eigenvalues depend on
the quantum number l, but are independent of the quantum number m.
85
5.6 Energy levels of the hydrogen atom 5 ANGULAR MOMENTUM
5.6 Energy levels of the hydrogen atom
Consider a hydrogen atom, for which the potential takes the speciﬁc form
V(r) = −
e
2
4π
0
r
. (5.107)
The radial eigenfunction R(r) satisﬁes Eq. (5.106), which can be written
_
_
¯h
2
2 µ
_
_
−
1
r
2
d
dr
r
2
d
dr
+
l (l +1)
r
2
_
_
−
e
2
4π
0
r
−E
_
_
R = 0. (5.108)
Here, µ = m
e
m
p
/(m
e
+m
p
) is the reduced mass, which takes into account the fact
that the electron (of mass m
e
) and the proton (of mass m
p
) both rotate about a
common centre, which is equivalent to a particle of mass µ rotating about a ﬁxed
point. Let us write the product r R(r) as the function P(r). The above equation
transforms to
d
2
P
dr
2
−
2 µ
¯h
2
_
_
l (l +1)¯h
2
2 µr
2
−
e
2
4π
0
r
−E
_
_
P = 0, (5.109)
which is the onedimensional Schr¨ odinger equation for a particle of mass µ mov
ing in the effective potential
V
eff
(r) = −
e
2
4π
0
r
+
l (l +1) ¯h
2
2 µr
2
. (5.110)
The effective potential has a simple physical interpretation. The ﬁrst part is the
attractive Coulomb potential, and the second part corresponds to the repulsive
centrifugal force.
Let
a =
¸
¸
¸
¸
_
−¯h
2
2 µE
, (5.111)
and y = r/a, with
P(r) = f(y) exp(−y). (5.112)
Here, it is assumed that the energy eigenvalue E is negative. Equation (5.109)
transforms to
_
_
d
2
dy
2
−2
d
dy
−
l (l +1)
y
2
+
2 µe
2
a
4π
0
¯h
2
y
_
_
f = 0. (5.113)
86
5.6 Energy levels of the hydrogen atom 5 ANGULAR MOMENTUM
Let us look for a powerlaw solution of the form
f(y) =
n
c
n
y
n
. (5.114)
Substituting this solution into Eq. (5.113), we obtain
n
c
n
_
n(n −1) y
n−2
−2 ny
n−1
−l (l +1) y
n−2
+
2 µe
2
a
4π
0
¯h
2
y
n−1
_
= 0. (5.115)
Equating the coefﬁcients of y
n−2
gives
c
n
[n(n −1) −l (l +1)] = c
n−1
_
_
2 (n −1) −
2 µe
2
a
4π
0
¯h
2
_
_
. (5.116)
Now, the power law series (5.114) must terminate at small n, at some positive
value of n, otherwise f(y) behaves unphysically as y →0. This is only possible if
[n
min
(n
min
−1)−l (l+1)] = 0, where the ﬁrst term in the series is c
n
min
y
n
min
. There
are two possibilities: n
min
= −l or n
min
= l + 1. The former predicts unphysical
behaviour of the wavefunction at y = 0. Thus, we conclude that n
min
= l + 1.
Note that for an l = 0 state there is a ﬁnite probability of ﬁnding the electron at
the nucleus, whereas for an l > 0 state there is zero probability of ﬁnding the
electron at the nucleus (i.e., ψ
2
= 0 at r = 0, except when l = 0). Note, also,
that it is only possible to obtain sensible behaviour of the wavefunction as r →0
if l is an integer.
For large values of y, the ratio of successive terms in the series (5.114) is
c
n
y
c
n−1
=
2 y
n
, (5.117)
according to Eq. (5.116). This is the same as the ratio of successive terms in the
series
n
(2 y)
n
n!
, (5.118)
which converges to exp(2 y). We conclude that f(y) → exp(2 y) as y → ∞. It
follows from Eq. (5.112) that R(r) → exp(r/a)/r as r → ∞. This does not cor
respond to physically acceptable behaviour of the wavefunction, since
_
ψ
2
dV
must be ﬁnite. The only way in which we can avoid this unphysical behaviour is
87
5.6 Energy levels of the hydrogen atom 5 ANGULAR MOMENTUM
if the series (5.114) terminates at some maximum value of n. According to the
recursion relation (5.116), this is only possible if
µe
2
a
4π
0
¯h
2
= n, (5.119)
where the last term in the series is c
n
y
n
. It follows from Eq. (5.111) that the
energy eigenvalues are quantized, and can only take the values
E = −
µe
4
32π
2
2
0
¯h
2
n
2
. (5.120)
Here, n is a positive integer which must exceed the quantum number l, otherwise
there would be no terms in the series (5.114).
It is clear that the wavefunction for a hydrogen atom can be written
ψ(r, θ, ϕ) = R(r/a) Y
m
l
(θ, ϕ), (5.121)
where
a =
n4π
0
¯h
2
µe
2
= 5.3 10
−11
n meters, (5.122)
and R(x) is a wellbehaved solution of the differential equation
_
_
1
x
2
d
dx
x
2
d
dx
−
l (l +1)
x
2
+
2 n
x
−1
_
_
R = 0. (5.123)
Finally, the Y
m
l
are spherical harmonics. The restrictions on the quantum numbers
are m ≤ l < n. Here, n is a positive integer, l is a nonnegative integer, and m
is an integer.
The ground state of hydrogen corresponds to n = 1. The only permissible
values of the other quantum numbers are l = 0 and m = 0. Thus, the ground
state is a spherically symmetric, zero angular momentum state. The energy of
the ground state is
E
0
= −
µe
4
32π
2
2
0
¯h
2
= −13.6 electron volts. (5.124)
88
5.7 Spin angular momentum 5 ANGULAR MOMENTUM
The next energy level corresponds to n = 2. The other quantum numbers are
allowed to take the values l = 0, m = 0 or l = 1, m = −1, 0, 1. Thus, there are
n = 2 states with nonzero angular momentum. Note that the energy levels given
in Eq. (5.120) are independent of the quantum number l, despite the fact that l
appears in the radial eigenfunction equation (5.123). This is a special property
of a 1/r Coulomb potential.
In addition to the quantized negative energy state of the hydrogen atom, which
we have just found, there is also a continuum of unbound positive energy states.
5.7 Spin angular momentum
Up to now, we have tacitly assumed that the state of a particle in quantum me
chanics can be completely speciﬁed by giving the wavefunction ψ as a function
of the spatial coordinates x, y, and z. Unfortunately, there is a wealth of experi
mental evidence which suggests that this simplistic approach is incomplete.
Consider an isolated system at rest, and let the eigenvalue of its total angular
momentum be j (j +1) ¯h
2
. According to the theory of orbital angular momentum
outlined in Sects. 5.4 and 5.5, there are two possibilities. For a system consisting
of a single particle, j = 0. For a system consisting of two (or more) particles, j is
a nonnegative integer. However, this does not agree with observations, because
we often ﬁnd systems which appear to be structureless, and yet have j ,= 0. Even
worse, systems where j has halfinteger values abound in nature. In order to
explain this apparent discrepancy between theory and experiments, Gouldsmit
and Uhlenbeck (in 1925) introduced the concept of an internal, purely quantum
mechanical, angular momentum called spin. For a particle with spin, the total
angular momentum in the rest frame is nonvanishing.
Let us denote the three components of the spin angular momentum of a par
ticle by the Hermitian operators (S
x
, S
y
, S
z
) ≡ S. We assume that these operators
obey the fundamental commutation relations (5.8)–(5.10) for the components of
an angular momentum. Thus, we can write
S S = i ¯hS. (5.125)
89
5.7 Spin angular momentum 5 ANGULAR MOMENTUM
We can also deﬁne the operator
S
2
= S
2
x
+S
2
y
+S
2
z
. (5.126)
According to the quite general analysis of Sect. 5.1,
[S, S
2
] = 0. (5.127)
Thus, it is possible to ﬁnd simultaneous eigenstates of S
2
and S
z
. These are
denoted s, s
z
), where
S
z
s, s
z
) = s
z
¯hs, s
z
), (5.128)
S
2
s, s
z
) = s (s +1) ¯h
2
s, s
z
). (5.129)
According to the equally general analysis of Sect. 5.2, the quantum number s can,
in principle, take integer or halfinteger values, and the quantum number s
z
can
only take the values s, s −1 −s +1, −s.
Spin angular momentum clearly has many properties in common with orbital
angular momentum. However, there is one vitally important difference. Spin
angular momentum operators cannot be expressed in terms of position and mo
mentum operators, like in Eqs. (5.1)–(5.3), since this identiﬁcation depends on
an analogy with classical mechanics, and the concept of spin is purely quantum
mechanical: i.e., it has no analogy in classical physics. Consequently, the re
striction that the quantum number of the overall angular momentum must take
integer values is lifted for spin angular momentum, since this restriction (found
in Sects. 5.3 and 5.4) depends on Eqs. (5.1)–(5.3). In other words, the quantum
number s is allowed to take halfinteger values.
Consider a spin onehalf particle, for which
S
z
±) = ±
¯h
2
±), (5.130)
S
2
±) =
3 ¯h
2
4
±). (5.131)
Here, the ±) denote eigenkets of the S
z
operator corresponding to the eigen
values ±¯h/2. These kets are orthonormal (since S
z
is an Hermitian operator),
so
¸+−) = 0. (5.132)
90
5.8 Wavefunction of a spin onehalf particle 5 ANGULAR MOMENTUM
They are also properly normalized and complete, so that
¸++) = ¸−−) = 1, (5.133)
and
+)¸+ + −)¸− = 1. (5.134)
It is easily veriﬁed that the Hermitian operators deﬁned by
S
x
=
¯h
2
( +)¸− + −)¸+ ) , (5.135)
S
y
=
i ¯h
2
( − +)¸− + −)¸+ ) , (5.136)
S
z
=
¯h
2
( +)¸+ − −)¸− ) , (5.137)
satisfy the commutation relations (5.8)–(5.10) (with the L
j
replaced by the S
j
).
The operator S
2
takes the form
S
2
=
3 ¯h
2
4
. (5.138)
It is also easily demonstrated that S
2
and S
z
, deﬁned in this manner, satisfy the
eigenvalue relations (5.130)–(5.131). Equations (5.135)–(5.138) constitute a re
alization of the spin operators S and S
2
(for a spin onehalf particle) in spin space
(i.e., that Hilbert subspace consisting of kets which correspond to the different
spin states of the particle).
5.8 Wavefunction of a spin onehalf particle
The state of a spin onehalf particle is represented as a vector in ket space.
Let us suppose that this space is spanned by the basis kets x
, y
, z
, ±). Here,
x
, y
, z
, ±) denotes a simultaneous eigenstate of the position operators x, y, z,
and the spin operator S
z
, corresponding to the eigenvalues x
, y
, z
, and ±¯h/2,
respectively. The basis kets are assumed to satisfy the completeness relation
___
( x
, y
, z
, +)¸x
, y
, z
, + + x
, y
, z
, −)¸x
, y
, z
, − ) dx
dy
dz
= 1.
(5.139)
91
5.8 Wavefunction of a spin onehalf particle 5 ANGULAR MOMENTUM
It is helpful to think of the ket x
, y
, z
, +) as the product of two kets—a
position space ket x
, y
, z
), and a spin space ket +). We assume that such a
product obeys the commutative and distributive axioms of multiplication:
x
, y
, z
)+) = +)x
, y
, z
), (5.140)
(c
x
, y
, z
) +c
x
, y
, z
)) +) = c
x
, y
, z
)+)
+c
x
, y
, z
)+) (5.141)
x
, y
, z
) (c
+
+) +c
−
−)) = c
+
x
, y
, z
)+)
+c
−
x
, y
, z
)−), (5.142)
where the c’s are numbers. We can give meaning to any position space operator
(such as L
z
) acting on the product x
, y
, z
)+) by assuming that it operates only
on the x
, y
, z
) factor, and commutes with the +) factor. Similarly, we can give
a meaning to any spin operator (such as S
z
) acting on x
, y
, z
)+) by assuming
that it operates only on +), and commutes with x
, y
, z
). This implies that every
position space operator commutes with every spin operator. In this manner, we
can give meaning to the equation
x
, y
, z
, ±) = x
, y
, z
)±) = ±)x
, y
, z
). (5.143)
The multiplication in the above equation is of quite a different type to any
which we have encountered previously. The ket vectors x
, y
, z
) and ±) are
in two quite separate vector spaces, and their product x
, y
, z
)±) is in a third
vector space. In mathematics, the latter space is termed the product space of the
former spaces, which are termed factor spaces. The number of dimensions of a
product space is equal to the product of the number of dimensions of each of the
factor spaces. A general ket of the product space is not of the form (5.143), but
is instead a sum or integral of kets of this form.
A general state A of a spin onehalf particle is represented as a ket A)) in the
product of the spin and position spaces. This state can be completely speciﬁed by
two wavefunctions:
ψ
+
(x
, y
, z
) = ¸x
, y
, z
¸+A)), (5.144)
ψ
−
(x
, y
, z
) = ¸x
, y
, z
¸−A)). (5.145)
92
5.9 Rotation operators in spin space 5 ANGULAR MOMENTUM
The probability of observing the particle in the region x
to x
+dx
, y
to y
+dy
,
and z
to z
+ dz
, with s
z
= +1/2 is ψ
+
(x
, y
, z
)
2
dx
dy
dz
. Likewise, the
probability of observing the particle in the region x
to x
+ dx
, y
to y
+ dy
,
and z
to z
+dz
, with s
z
= −1/2 is ψ
−
(x
, y
, z
)
2
dx
dy
dz
. The normalization
condition for the wavefunctions is
___
_
ψ
+

2
+ ψ
−

2
_
dx
dy
dz
= 1. (5.146)
5.9 Rotation operators in spin space
Let us, for the moment, forget about the spatial position of the particle, and
concentrate on its spin state. A general spin state A is represented by the ket
A) = ¸+A)+) +¸−A)−) (5.147)
in spin space. In Sect. 5.3, we were able to construct an operator R
z
(∆ϕ) which
rotates the system by an angle ∆ϕ about the zaxis in position space. Can we also
construct an operator T
z
(∆ϕ) which rotates the system by an angle ∆ϕ about
the zaxis in spin space? By analogy with Eq. (5.62), we would expect such an
operator to take the form
T
z
(∆ϕ) = exp(−i S
z
∆ϕ/¯h). (5.148)
Thus, after rotation, the ket A) becomes
A
R
) = T
z
(∆ϕ)A). (5.149)
To demonstrate that the operator (5.148) really does rotate the spin of the
system, let us consider its effect on ¸S
x
). Under rotation, this expectation value
changes as follows:
¸S
x
) →¸A
R
S
x
A
R
) = ¸AT
†
z
S
x
T
z
A). (5.150)
Thus, we need to compute
exp( i S
z
∆ϕ/¯h) S
x
exp(−i S
z
∆ϕ/¯h). (5.151)
93
5.9 Rotation operators in spin space 5 ANGULAR MOMENTUM
This can be achieved in two different ways.
First, we can use the explicit formula for S
x
given in Eq. (5.135). We ﬁnd that
Eq. (5.151) becomes
¯h
2
exp( i S
z
∆ϕ/¯h) ( +)¸− + −)¸+ ) exp(−i S
z
∆ϕ/¯h), (5.152)
or
¯h
2
_
e
i ∆ϕ/2
+)¸− e
i ∆ϕ/2
+ e
−i ∆ϕ/2
−)¸+ e
−i ∆ϕ/2
_
, (5.153)
which reduces to
S
x
cos ∆ϕ−S
y
sin∆ϕ, (5.154)
where use has been made of Eqs. (5.135)–(5.137).
A second approach is to use the so called BakerHausdorff lemma. This takes
the form
exp( i Gλ) A exp(−i Gλ) = A+ i λ[G, A] +
_
_
i
2
λ
2
2!
_
_
[G, [G, A]] + (5.155)
+
_
i
n
λ
n
n!
_
[G, [G, [G, [G, A]]] ],
where G is a Hermitian operator, and λ is a real parameter. The proof of this
lemma is left as an exercise. Applying the BakerHausdorff lemma to Eq. (5.151),
we obtain
S
x
+
_
i ∆ϕ
¯h
_
[S
z
, S
x
] +
_
1
2!
_ _
i ∆ϕ
¯h
_
2
[S
z
, [S
z
, S
x
]] + , (5.156)
which reduces to
S
x
_
_
1 −
∆ϕ
2
2!
+
_
_
−S
y
_
_
ϕ−
∆ϕ
3
3!
+
_
_
, (5.157)
or
S
x
cos ∆ϕ−S
y
sin∆ϕ, (5.158)
where use has been made of Eq. (5.125). The second proof is more general than
the ﬁrst, since it only uses the fundamental commutation relation (5.125), and is,
therefore, valid for systems with spin angular momentum higher than onehalf.
94
5.9 Rotation operators in spin space 5 ANGULAR MOMENTUM
For a spin onehalf system, both methods imply that
¸S
x
) →¸S
x
) cos ∆ϕ−¸S
y
) sin∆ϕ (5.159)
under the action of the rotation operator (5.148). It is straightforward to show
that
¸S
y
) →¸S
y
) cos ∆ϕ+¸S
x
) sin∆ϕ. (5.160)
Furthermore,
¸S
z
) →¸S
z
), (5.161)
since S
z
commutes with the rotation operator. Equations (5.159)–(5.161) demon
strate that the operator (5.148) rotates the expectation value of S by an angle ∆ϕ
about the zaxis. In fact, the expectation value of the spin operator behaves like
a classical vector under rotation:
¸S
k
) →
l
R
kl
¸S
l
), (5.162)
where the R
kl
are the elements of the conventional rotation matrix for the rotation
in question. It is clear, from our second derivation of the result (5.159), that this
property is not restricted to the spin operators of a spin onehalf system. In fact,
we have effectively demonstrated that
¸J
k
) →
l
R
kl
¸J
l
), (5.163)
where the J
k
are the generators of rotation, satisfying the fundamental commuta
tion relation J J = i ¯hJ, and the rotation operator about the kth axis is written
R
k
(∆ϕ) = exp(−i J
k
∆ϕ/¯h).
Consider the effect of the rotation operator (5.148) on the state ket (5.147).
It is easily seen that
T
z
(∆ϕ)A) = e
−i ∆ϕ/2
¸+A)+) + e
i ∆ϕ/2
¸−A)−). (5.164)
Consider a rotation by 2π radians. We ﬁnd that
A) →T
z
(2π)A) = −A). (5.165)
95
5.10 Magnetic moments 5 ANGULAR MOMENTUM
Note that a ket rotated by 2π radians differs from the original ket by a minus
sign. In fact, a rotation by 4π radians is needed to transform a ket into itself.
The minus sign does not affect the expectation value of S, since S is sandwiched
between ¸A and A), both of which change sign. Nevertheless, the minus sign
does give rise to observable consequences, as we shall see presently.
5.10 Magnetic moments
Consider a particle of charge q and velocity v performing a circular orbit of radius
r in the xy plane. The charge is equivalent to a current loop of radius r in the
xy plane carrying current I = qv/2πr. The magnetic moment µ of the loop is of
magnitude πr
2
I and is directed along the zaxis. Thus, we can write
µ =
q
2
r v, (5.166)
where r and v are the vector position and velocity of the particle, respectively.
However, we know that p = v/m, where p is the vector momentum of the parti
cle, and m is its mass. We also know that L = rp, where L is the orbital angular
momentum. It follows that
µ =
q
2 m
L. (5.167)
Using the usual analogy between classical and quantum mechanics, we expect the
above relation to also hold between the quantum mechanical operators, µ and L,
which represent magnetic moment and orbital angular momentum, respectively.
This is indeed found to the the case.
Does spin angular momentum also give rise to a contribution to the magnetic
moment of a charged particle? The answer is “yes”. In fact, relativistic quantum
mechanics actually predicts that a charged particle possessing spin should also
possess a magnetic moment (this was ﬁrst demonstrated by Dirac). We can write
µ =
q
2 m
(L +gS) , (5.168)
where g is called the gyromagnetic ratio. For an electron this ratio is found to be
g
e
= 2
_
_
1 +
1
2π
e
2
4π
0
¯hc
_
_
. (5.169)
96
5.11 Spin precession 5 ANGULAR MOMENTUM
The factor 2 is correctly predicted by Dirac’s relativistic theory of the electron.
The small correction 1/(2π137), derived originally by Schwinger, is due to quan
tum ﬁeld effects. We shall ignore this correction in the following, so
µ · −
e
2 m
e
(L +2 S) (5.170)
for an electron (here, e > 0).
5.11 Spin precession
The Hamiltonian for an electron at rest in a zdirected magnetic ﬁeld, B = B ^z, is
H = −µB =
_
e
m
e
_
SB = ωS
z
, (5.171)
where
ω =
e B
m
e
. (5.172)
According to Eq. (4.28), the time evolution operator for this system is
T(t, 0) = exp(−i Ht/¯h) = exp(−i S
z
ωt/¯h). (5.173)
It can be seen, by comparison with Eq. (5.148), that the time evolution operator
is precisely the same as the rotation operator for spin, with ∆ϕ set equal to ωt.
It is immediately clear that the Hamiltonian (5.171) causes the electron spin to
precess about the zaxis with angular frequency ω. In fact, Eqs. (5.159)–(5.161)
imply that
¸S
x
)
t
= ¸S
x
)
t=0
cos ωt −¸S
y
)
t=0
sinωt, (5.174)
¸S
y
)
t
= ¸S
y
)
t=0
cos ωt +¸S
x
)
t=0
sinωt, (5.175)
¸S
z
)
t
= ¸S
z
)
t=0
. (5.176)
The time evolution of the state ket is given by analogy with Eq. (5.164):
A, t) = e
−i ωt/2
¸+A, 0)+) + e
i ωt/2
¸−A, 0)−). (5.177)
97
5.11 Spin precession 5 ANGULAR MOMENTUM
Note that it takes time t = 4π/ω for the state ket to return to its original state. By
contrast, it only takes times t = 2π/ω for the spin vector to point in its original
direction.
We now describe an experiment to detect the minus sign in Eq. (5.165). An
almost monoenergetic beam of neutrons is split in two, sent along two different
paths, A and B, and then recombined. Path A goes through a magnetic ﬁeld free
region. However, path B enters a small region where a static magnetic ﬁeld is
present. As a result, a neutron state ket going along path B acquires a phaseshift
exp(∓i ωT/2) (the ∓ signs correspond to s
z
= ±1/2 states). Here, T is the time
spent in the magnetic ﬁeld, and ω is the spin precession frequency
ω =
g
n
e B
m
p
. (5.178)
This frequency is deﬁned in an analogous manner to Eq. (5.172). The gyro
magnetic ratio for a neutron is found experimentally to be g
n
= −1.91. (The
magnetic moment of a neutron is entirely a quantum ﬁeld effect). When neu
trons from path A and path B meet they undergo interference. We expect the
observed neutron intensity in the interference region to exhibit a cos(±ωT/2+δ)
variation, where δ is the phase difference between paths A and B in the absence
of a magnetic ﬁeld. In experiments, the time of ﬂight T through the magnetic
ﬁeld region is kept constant, while the ﬁeldstrength B is varied. It follows that
the change in magnetic ﬁeld required to produce successive maxima is
∆B =
4π¯h
e g
n
¯ λ l
, (5.179)
where l is the pathlength through the magnetic ﬁeld region, and ¯ λ is the de
Broglie wavelength over 2π of the neutrons. The above prediction has been ver
iﬁed experimentally to within a fraction of a percent. This prediction depends
crucially on the fact that it takes a 4π rotation to return a state ket to its original
state. If it only took a 2π rotation then ∆B would be half of the value given above,
which does not agree with the experimental data.
98
5.12 Pauli twocomponent formalism 5 ANGULAR MOMENTUM
5.12 Pauli twocomponent formalism
We have seen, in Sect. 5.4, that the eigenstates of orbital angular momentum
can be conveniently represented as spherical harmonics. In this representation,
the orbital angular momentum operators take the form of differential operators
involving only angular coordinates. It is conventional to represent the eigenstates
of spin angular momentum as column (or row) matrices. In this representation,
the spin angular momentum operators take the form of matrices.
The matrix representation of a spin onehalf system was introduced by Pauli in
1926. Recall, from Sect. 5.9, that a general spin ket can be expressed as a linear
combination of the two eigenkets of S
z
belonging to the eigenvalues ±¯h/2. These
are denoted ±). Let us represent these basis eigenkets as column matrices:
+) →
_
_
1
0
_
_
≡ χ
+
, (5.180)
−) →
_
_
0
1
_
_
≡ χ
−
. (5.181)
The corresponding eigenbras are represented as row matrices:
¸+ → (1, 0) ≡ χ
†
+
, (5.182)
¸− → (0, 1) ≡ χ
†
−
. (5.183)
In this scheme, a general bra takes the form
A) = ¸+A)+) +¸−A)−) →
_
_
¸+A)
¸−A)
_
_
, (5.184)
and a general ket becomes
¸A = ¸A+)¸+ +¸A−)¸− →(¸A+), ¸A−)). (5.185)
The column matrix (5.184) is called a twocomponent spinor, and can be written
χ ≡
_
_
¸+A)
¸−A)
_
_
=
_
_
c
+
c
−
_
_
= c
+
χ
+
+c
−
χ
−
, (5.186)
99
5.12 Pauli twocomponent formalism 5 ANGULAR MOMENTUM
where the c
±
are complex numbers. The row matrix (5.185) becomes
χ
†
≡ (¸A+), ¸A−)) = (c
∗
+
, c
∗
−
) = c
∗
+
χ
†
+
+c
∗
−
χ
†
−
. (5.187)
Consider the ket obtained by the action of a spin operator on ket A:
A
) = S
k
A). (5.188)
This ket is represented as
A
) →
_
_
¸+A
)
¸−A
)
_
_
≡ χ
. (5.189)
However,
¸+A
) = ¸+S
k
+)¸+A) +¸+S
k
−)¸−A), (5.190)
¸−A
) = ¸−S
k
+)¸+A) +¸−S
k
−)¸−A), (5.191)
or
_
_
¸+A
)
¸−A
)
_
_
=
_
_
¸+S
k
+) ¸+S
k
−)
¸−S
k
+) ¸−S
k
−)
_
_
_
_
¸+A)
¸−A)
_
_
. (5.192)
It follows that we can represent the operator/ket relation (5.188) as the matrix
relation
χ
=
_
¯h
2
_
σ
k
χ, (5.193)
where the σ
k
are the matrices of the ¸±S
k
±) values divided by ¯h/2. These
matrices, which are called the Pauli matrices, can easily be evaluated using the
explicit forms for the spin operators given in Eqs. (5.135)–(5.137). We ﬁnd that
σ
1
=
_
_
0 1
1 0
_
_
, (5.194)
σ
2
=
_
_
0 −i
i 0
_
_
, (5.195)
σ
3
=
_
_
1 0
0 −1
_
_
. (5.196)
100
5.12 Pauli twocomponent formalism 5 ANGULAR MOMENTUM
Here, 1, 2, and 3 refer to x, y, and z, respectively. Note that, in this scheme, we
are effectively representing the spin operators in terms of the Pauli matrices:
S
k
→
_
¯h
2
_
σ
k
. (5.197)
The expectation value of S
k
can be written in terms of spinors and the Pauli
matrices:
¸S
k
) = ¸AS
k
A) =
±
¸A±)¸±S
k
±)¸±A) =
_
¯h
2
_
χ
†
σ
k
χ. (5.198)
The fundamental commutation relation for angular momentum, Eq. (5.125),
can be combined with (5.197) to give the following commutation relation for the
Pauli matrices:
σ σ = 2 i σ. (5.199)
It is easily seen that the matrices (5.194)–(5.196) actually satisfy these relations
(i.e., σ
1
σ
2
−σ
2
σ
1
= 2 i σ
3
, plus all cyclic permutations). It is also easily seen that
the Pauli matrices satisfy the anticommutation relations
{σ
i
, σ
j
} = 2 δ
ij
. (5.200)
Let us examine how the Pauli scheme can be extended to take into account the
position of a spin onehalf particle. Recall, from Sect. 5.8, that we can represent
a general basis ket as the product of basis kets in position space and spin space:
x
, y
, z
, ±) = x
, y
, z
)±) = ±)x
, y
, z
). (5.201)
The ket corresponding to state A is denoted A)), and resides in the product
space of the position and spin ket spaces. State A is completely speciﬁed by the
two wavefunctions
ψ
+
(x
, y
, z
) = ¸x
, y
, z
¸+A)), (5.202)
ψ
−
(x
, y
, z
) = ¸x
, y
, z
¸−A)). (5.203)
Consider the operator relation
A
)) = S
k
A)). (5.204)
101
5.12 Pauli twocomponent formalism 5 ANGULAR MOMENTUM
It is easily seen that
¸x
, y
, z
¸+A
)) = ¸+S
k
+)¸x
, y
, z
¸+A))
+¸+S
k
−)¸x
, y
, z
¸−A)), (5.205)
¸x
, y
, z
¸−A
)) = ¸−S
k
+)¸x
, y
, z
¸+A))
+¸−S
k
−)¸x
, y
, z
¸−A)), (5.206)
where use has been made of the fact that the spin operator S
k
commutes with
the eigenbras ¸x
, y
, z
. It is fairly obvious that we can represent the operator
relation (5.204) as a matrix relation if we generalize our deﬁnition of a spinor by
writing
A)) →
_
_
ψ
+
(r
)
ψ
−
(r
)
_
_
≡ χ, (5.207)
and so on. The components of a spinor are now wavefunctions, instead of com
plex numbers. In this scheme, the operator equation (5.204) becomes simply
χ
=
_
¯h
2
_
σ
k
χ. (5.208)
Consider the operator relation
A
)) = p
k
A)). (5.209)
In the Schr¨ odinger representation, we have
¸x
, y
, z
¸+A
)) = ¸x
, y
, z
p
k
¸+A))
= −i ¯h
∂
∂x
k
¸x
, y
, z
¸+A)), (5.210)
¸x
, y
, z
¸−A
)) = ¸x
, y
, z
p
k
¸−A))
= −i ¯h
∂
∂x
k
¸x
, y
, z
¸−A)), (5.211)
where use has been made of Eq. (3.78). The above equation reduces to
_
_
ψ
+
(r
)
ψ
−
(r
)
_
_
=
_
_
−i ¯h∂ψ
+
(r
)/∂x
k
−i ¯h∂ψ
−
(r
)/∂x
k
_
_
. (5.212)
102
5.12 Pauli twocomponent formalism 5 ANGULAR MOMENTUM
Thus, the operator equation (5.209) can be written
χ
= p
k
χ, (5.213)
where
p
k
→−i ¯h
∂
∂x
k
I. (5.214)
Here, I is the 2 2 unit matrix. In fact, any position operator (e.g., p
k
or L
k
)
is represented in the Pauli scheme as some differential operator of the position
eigenvalues multiplied by the 2 2 unit matrix.
What about combinations of position and spin operators? The most commonly
occurring combination is a dot product: e.g., S L = (¯h/2) σ L. Consider the
hybrid operator σa, where a ≡ (a
x
, a
y
, a
z
) is some vector position operator. This
quantity is represented as a 2 2 matrix:
σa ≡
k
a
k
σ
k
=
_
_
+a
3
a
1
− i a
2
a
1
+ i a
2
−a
3
_
_
. (5.215)
Since, in the Schr¨ odinger representation, a general position operator takes the
form of a differential operator in x
, y
, or z
, it is clear that the above quantity
must be regarded as a matrix differential operator which acts on spinors of the
general form (5.207). The important identity
(σa) (σb) = ab + i σ(a b) (5.216)
follows fromthe commutation and anticommutation relations (5.199) and (5.200).
Thus,
j
σ
j
a
j
k
σ
k
b
k
=
j
k
_
1
2
{σ
j
, σ
k
} +
1
2
[σ
j
, σ
k
]
_
a
j
b
k
=
j
k
(σ
jk
+ i
jkl
σ
l
) a
j
b
k
= ab + i σ(a b). (5.217)
A general rotation operator in spin space is written
T(∆φ) = exp(−i Sn∆ϕ/¯h) , (5.218)
103
5.12 Pauli twocomponent formalism 5 ANGULAR MOMENTUM
by analogy with Eq. (5.148), where n is a unit vector pointing along the axis of
rotation, and ∆ϕ is the angle of rotation. Here, n can be regarded as a trivial
position operator. The rotation operator is represented
exp(−i Sn∆ϕ/¯h) →exp(−i σn∆ϕ/2) (5.219)
in the Pauli scheme. The term on the righthand side of the above expression is
the exponential of a matrix. This can easily be evaluated using the Taylor series
for an exponential, plus the rules
(σn)
n
= 1 for n even, (5.220)
(σn)
n
= (σn) for n odd. (5.221)
These rules follow trivially from the identity (5.216). Thus, we can write
exp (−i σn∆ϕ/2) =
_
_
1 −
(σn)
2
2!
_
∆ϕ
2
_
2
+
(σn)
4
4!
_
∆ϕ
2
_
4
+
_
_
−i
_
_
(σ n)
_
∆ϕ
2
_
−
(σn)
3
3!
_
∆ϕ
2
_
3
+
_
_
= cos(∆ϕ/2) I − i sin(∆ϕ/2) σn. (5.222)
The explicit 2 2 form of this matrix is
_
_
cos(∆ϕ/2) − i n
z
sin(∆ϕ/2) (−i n
x
−n
y
) sin(∆ϕ/2)
(−i n
x
+n
y
) sin(∆ϕ/2) cos(∆ϕ/2) + i n
z
sin(∆ϕ/2)
_
_
. (5.223)
Rotation matrices act on spinors in much the same manner as the corresponding
rotation operators act on state kets. Thus,
χ
= exp(−i σn∆ϕ/2) χ, (5.224)
where χ
denotes the spinor obtained after rotating the spinor χ an angle ∆ϕ
about the naxis. The Pauli matrices remain unchanged under rotations. How
ever, the quantity χ
†
σ
k
χ is proportional to the expectation value of S
k
[see
Eq. (5.198)], so we would expect it to transform like a vector under rotation
(see Sect. 5.9). In fact, we require
(χ
†
σ
k
χ)
≡ (χ
†
)
σ
k
χ
=
l
R
kl
(χ
†
σ
l
χ), (5.225)
104
5.13 Spin greater than onehalf systems 5 ANGULAR MOMENTUM
where the R
kl
are the elements of a conventional rotation matrix. This is easily
demonstrated, since
exp
_
i σ
3
∆ϕ
2
_
σ
1
exp
_
−i σ
3
∆ϕ
2
_
= σ
1
cos ∆ϕ−σ
2
sin∆ϕ (5.226)
plus all cyclic permutations. The above expression is the 2 2 matrix analogue
of (see Sect. 5.9)
exp
_
i S
z
∆ϕ
¯h
_
S
x
exp
_
−i S
z
∆ϕ
¯h
_
= S
x
cos ∆ϕ−S
y
sin∆ϕ. (5.227)
The previous two formulae can both be validated using the BakerHausdorff
lemma, (5.156), which holds for Hermitian matrices, in addition to Hermitian
operators.
5.13 Spin greater than onehalf systems
In the absence of spin, the Hamiltonian can be written as some function of the
position and momentum operators. Using the Schr¨ odinger representation, in
which p →−i ¯h∇, the energy eigenvalue problem,
HE) = EE), (5.228)
can be transformed into a partial differential equation for the wavefunction
ψ(r
) ≡ ¸r
E). This function speciﬁes the probability density for observing the
particle at a given position, r
. In general, we ﬁnd
Hψ = Eψ, (5.229)
where H is now a partial differential operator. The boundary conditions (for a
bound state) are obtained from the normalization constraint
_
ψ
2
dV = 1. (5.230)
This is all very familiar. However, we now know how to generalize this scheme
to deal with a spin onehalf particle. Instead of representing the state of the par
ticle by a single wavefunction, we use two wavefunctions. The ﬁrst, ψ
+
(r
),
105
5.13 Spin greater than onehalf systems 5 ANGULAR MOMENTUM
speciﬁes the probability density of observing the particle at position r
with spin
angular momentum +¯h/2 in the zdirection. The second, ψ
−
(r
), speciﬁes the
probability density of observing the particle at position r
with spin angular mo
mentum −¯h/2 in the zdirection. In the Pauli scheme, these wavefunctions are
combined into a spinor, χ, which is simply the row vector of ψ
+
and ψ
−
. In gen
eral, the Hamiltonian is a function of the position, momentum, and spin opera
tors. Adopting the Schr¨ odinger representation, and the Pauli scheme, the energy
eigenvalue problem reduces to
Hχ = Eχ, (5.231)
where χ is a spinor (i.e., a 1 2 matrix of wavefunctions) and H is a 2 2 matrix
partial differential operator [see Eq. (5.215)]. The above spinor equation can
always be written out explicitly as two coupled partial differential equations for
ψ
+
and ψ
−
.
Suppose that the Hamiltonian has no dependence on the spin operators. In this
case, the Hamiltonian is represented as diagonal 2 2 matrix partial differential
operator in the Schr¨ odinger/Pauli scheme [see Eq. (5.214)]. In other words, the
partial differential equation for ψ
+
decouples from that for ψ
−
. In fact, both
equations have the same form, so there is only really one differential equation.
In this situation, the most general solution to Eq. (5.231) can be written
χ = ψ(r
)
_
_
c
+
c
−
_
_
. (5.232)
Here, ψ(r
) is determined by the solution of the differential equation, and the
c
±
are arbitrary complex numbers. The physical signiﬁcance of the above ex
pression is clear. The Hamiltonian determines the relative probabilities of ﬁnding
the particle at various different positions, but the direction of its spin angular
momentum remains undetermined.
Suppose that the Hamiltonian depends only on the spin operators. In this
case, the Hamiltonian is represented as a 2 2 matrix of complex numbers in
the Schr¨ odinger/Pauli scheme [see Eq. (5.197)], and the spinor eigenvalue equa
tion (5.231) reduces to a straightforward matrix eigenvalue problem. The most
106
5.13 Spin greater than onehalf systems 5 ANGULAR MOMENTUM
general solution can again be written
χ = ψ(r
)
_
_
c
+
c
−
_
_
. (5.233)
Here, the ratio c
+
/c
−
is determined by the matrix eigenvalue problem, and the
wavefunction ψ(r
) is arbitrary. Clearly, the Hamiltonian determines the direc
tion of the particle’s spin angular momentum, but leaves its position undeter
mined.
In general, of course, the Hamiltonian is a function of both position and
spin operators. In this case, it is not possible to decompose the spinor as in
Eqs. (5.232) and (5.233). In other words, a general Hamiltonian causes the di
rection of the particle’s spin angular momentum to vary with position in some
speciﬁed manner. This can only be represented as a spinor involving different
wavefunctions, ψ
+
and ψ
−
.
But, what happens if we have a spin one or a spin threehalves particle? It
turns out that we can generalize the Pauli twocomponent scheme in a fairly
straightforward manner. Consider a spins particle: i.e., a particle for which the
eigenvalue of S
2
is s (s + 1) ¯h
2
. Here, s is either an integer, or a halfinteger.
The eigenvalues of S
z
are written s
z
¯h, where s
z
is allowed to take the values
s, s − 1, , −s + 1, −s. In fact, there are 2 s + 1 distinct allowed values of s
z
.
Not surprisingly, we can represent the state of the particle by 2 s + 1 different
wavefunctions, denoted ψ
s
z
(r
). Here, ψ
s
z
(r
) speciﬁes the probability density
for observing the particle at position r
with spin angular momentum s
z
¯h in the
zdirection. More exactly,
ψ
s
z
(r
) = ¸r
¸s, s
z
A)), (5.234)
where A)) denotes a state ket in the product space of the position and spin op
erators. The state of the particle can be represented more succinctly by a spinor,
χ, which is simply the 2 s + 1 component row vector of the ψ
s
z
(r
). Thus, a spin
onehalf particle is represented by a twocomponent spinor, a spin one particle
by a threecomponent spinor, a spin threehalves particle by a fourcomponent
spinor, and so on.
107
5.13 Spin greater than onehalf systems 5 ANGULAR MOMENTUM
In this extended Schr¨ odinger/Pauli scheme, position space operators take the
form of diagonal (2 s + 1) (2 s + 1) matrix differential operators. Thus, we can
represent the momentum operators as [see Eq. (5.214)]
p
k
→−i ¯h
∂
∂x
k
I, (5.235)
where I is the (2 s +1) (2 s +1) unit matrix. We represent the spin operators as
S
k
→s ¯hσ
k
, (5.236)
where the (2 s +1) (2 s +1) extended Pauli matrix σ
k
has elements
(σ
k
)
jl
=
¸s, jS
k
s, l)
s ¯h
. (5.237)
Here, j, l are integers, or halfintegers, lying in the range −s to +s. But, how can
we evaluate the brackets ¸s, jS
k
s, l) and, thereby, construct the extended Pauli
matrices? In fact, it is trivial to construct the σ
z
matrix. By deﬁnition,
S
z
s, j) = j ¯hs, j). (5.238)
Hence,
(σ
z
)
jl
=
¸s, jS
z
s, l)
s ¯h
=
j
s
δ
jl
, (5.239)
where use has been made of the orthonormality property of the s, j). Thus, σ
z
is the suitably normalized diagonal matrix of the eigenvalues of S
z
. The matrix
elements of σ
x
and σ
y
are most easily obtained by considering the shift operators,
S
±
= S
x
±i S
y
. (5.240)
We know, from Eqs. (5.55)–(5.56), that
S
+
s, j) =
_
s (s +1) −j (j +1) ¯hs, j +1), (5.241)
S
−
s, j) =
_
s (s +1) −j (j −1) ¯hs, j −1). (5.242)
It follows from Eqs. (5.237), and (5.240)–(5.242), that
(σ
x
)
jl
=
_
s (s +1) −j (j −1) δ
j,l+1
2 s
108
5.13 Spin greater than onehalf systems 5 ANGULAR MOMENTUM
+
_
s (s +1) −j (j +1) δ
j,l−1
2 s
, (5.243)
(σ
y
)
jl
=
_
s (s +1) −j (j −1) δ
j,l+1
2 i s
−
_
s (s +1) −j (j +1) δ
j,l−1
2 i s
. (5.244)
According to Eqs. (5.239) and (5.243)–(5.244), the Pauli matrices for a spin one
half (s = 1/2) particle are
σ
x
=
_
_
0 1
1 0
_
_
, (5.245)
σ
y
=
_
_
0 −i
i 0
_
_
, (5.246)
σ
z
=
_
_
1 0
0 −1
_
_
, (5.247)
as we have seen previously. For a spin one (s = 1) particle, we ﬁnd that
σ
x
=
1
√
2
_
_
_
_
_
0 1 0
1 0 1
0 1 0
_
_
_
_
_
, (5.248)
σ
y
=
1
√
2
_
_
_
_
_
0 −i 0
i 0 −i
0 i 0
_
_
_
_
_
, (5.249)
σ
z
=
_
_
_
_
_
1 0 0
0 0 0
0 0 −1
_
_
_
_
_
. (5.250)
In fact, we can now construct the Pauli matrices for a spin anything particle.
This means that we can convert the general energy eigenvalue problem for a
spins particle, where the Hamiltonian is some function of position and spin op
erators, into 2 s + 1 coupled partial differential equations involving the 2 s + 1
wavefunctions ψ
s
z
(r
). Unfortunately, such a system of equations is generally
too complicated to solve exactly.
109
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
5.14 Addition of angular momentum
Consider a hydrogen atom in an l = 1 state. The electron possesses orbital angu
lar momentum of magnitude ¯h, and spin angular momentum of magnitude ¯h/2.
So, what is the total angular momentum of the system?
In order to answer this question, we are going to have to learn how to add
angular momentum operators. Let us consider the most general case. Suppose
that we have two sets of angular momentum operators, J
1
and J
2
. By deﬁnition,
these operators are Hermitian, and obey the fundamental commutation relations
J
1
J
1
= i ¯hJ
1
, (5.251)
J
2
J
2
= i ¯hJ
2
. (5.252)
We assume that the two groups of operators correspond to different degrees of
freedom of the system, so that
[J
1i
, J
2j
] = 0, (5.253)
where i, j stand for either x, y, or z. For instance, J
1
could be an orbital angular
momentum operator, and J
2
a spin angular momentum operator. Alternatively, J
1
and J
2
could be the orbital angular momentum operators of two different parti
cles in a multiparticle system. We know, from the general properties of angular
momentum, that the eigenvalues of J
2
1
and J
2
2
can be written j
1
(j
1
+ 1) ¯h
2
and
j
2
(j
2
+1) ¯h
2
, respectively, where j
1
and j
2
are either integers, or halfintegers. We
also know that the eigenvalues of J
1z
and J
2z
take the form m
1
¯h and m
2
¯h, respec
tively, where m
1
and m
2
are numbers lying in the ranges j
1
, j
1
−1, , −j
1
+1, −j
1
and j
2
, j
2
−1, , −j
2
+1, −j
2
, respectively.
Let us deﬁne the total angular momentum operator
J = J
1
+J
2
. (5.254)
Now J is an Hermitian operator, since it is the sum of Hermitian operators. Ac
cording to Eqs. (5.11) and (5.14), J satisﬁes the fundamental commutation rela
tion
J J = i ¯hJ. (5.255)
110
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
Thus, J possesses all of the expected properties of an angular momentum opera
tor. It follows that the eigenvalue of J
2
can be written j (j + 1) ¯h
2
, where j is an
integer, or a halfinteger. The eigenvalue of J
z
takes the form m¯h, where m lies
in the range j, j −1, , −j +1, −j. At this stage, we do not know the relationship
between the quantum numbers of the total angular momentum, j and m, and
those of the individual angular momenta, j
1
, j
2
, m
1
, and m
2
.
Now
J
2
= J
2
1
+J
2
2
+2 J
1
J
2
. (5.256)
We know that
[J
2
1
, J
1i
] = 0, (5.257)
[J
2
2
, J
2i
] = 0, (5.258)
and also that all of the J
1i
operators commute with the J
2i
operators. It follows
from Eq. (5.256) that
[J
2
, J
2
1
] = [J
2
, J
2
2
] = 0. (5.259)
This implies that the quantum numbers j
1
, j
2
, and j can all be measured simulta
neously. In other words, we can know the magnitude of the total angular momen
tum together with the magnitudes of the component angular momenta. However,
it is clear from Eq. (5.256) that
[J
2
, J
1z
] ,= 0, (5.260)
[J
2
, J
2z
] ,= 0. (5.261)
This suggests that it is not possible to measure the quantum numbers m
1
and
m
2
simultaneously with the quantum number j. Thus, we cannot determine the
projections of the individual angular momenta along the zaxis at the same time
as the magnitude of the total angular momentum.
It is clear, from the preceding discussion, that we can form two alternate
groups of mutually commuting operators. The ﬁrst group is J
2
1
, J
2
2
, J
1z
, and J
2z
.
The second group is J
2
1
, J
2
2
, J
2
, and J
z
. These two groups of operators are in
compatible with one another. We can deﬁne simultaneous eigenkets of each
operator group. The simultaneous eigenkets of J
2
1
, J
2
2
, J
1z
, and J
2z
are denoted
111
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
j
1
, j
2
; m
1
, m
2
), where
J
2
1
j
1
, j
2
; m
1
, m
2
) = j
1
(j
1
+1) ¯h
2
j
1
, j
2
; m
1
, m
2
), (5.262)
J
2
2
j
1
, j
2
; m
1
, m
2
) = j
2
(j
2
+1) ¯h
2
j
1
, j
2
; m
1
, m
2
), (5.263)
J
1z
j
1
, j
2
; m
1
, m
2
) = m
1
¯hj
1
, j
2
; m
1
, m
2
), (5.264)
J
2z
j
1
, j
2
; m
1
, m
2
) = m
2
¯hj
1
, j
2
; m
1
, m
2
). (5.265)
The simultaneous eigenkets of J
2
1
, J
2
2
, J
2
and J
z
are denoted j
1
, j
2
; j, m), where
J
2
1
j
1
, j
2
; j, m) = j
1
(j
1
+1) ¯h
2
j
1
, j
2
; j, m), (5.266)
J
2
2
j
1
, j
2
; j, m) = j
2
(j
2
+1) ¯h
2
j
1
, j
2
; j, m), (5.267)
J
2
j
1
, j
2
; j, m) = j (j +1) ¯h
2
j
1
, j
2
; j, m), (5.268)
J
z
j
1
, j
2
; j, m) = m¯hj
1
, j
2
; j, m). (5.269)
Each set of eigenkets are complete, mutually orthogonal (for eigenkets corre
sponding to different sets of eigenvalues), and have unit norms. Since the op
erators J
2
1
and J
2
2
are common to both operator groups, we can assume that the
quantum numbers j
1
and j
2
are known. In other words, we can always determine
the magnitudes of the individual angular momenta. In addition, we can either
know the quantum numbers m
1
and m
2
, or the quantum numbers j and m, but
we cannot know both pairs of quantum numbers at the same time. We can write
a conventional completeness relation for both sets of eigenkets:
m
1
m
2
j
1
, j
2
; m
1
, m
2
)¸j
1
, j
2
; m
1
, m
2
 = 1, (5.270)
j
m
j
1
, j
2
; j, m)¸j
1
, j
2
; j, m = 1, (5.271)
where the righthand sides denote the identity operator in the ket space corre
sponding to states of given j
1
and j
2
. The summation is over all allowed values of
m
1
, m
2
, j, and m.
The operator group J
2
1
, J
2
2
, J
2
, and J
z
is incompatible with the group J
2
1
, J
2
2
,
J
1z
, and J
2z
. This means that if the system is in a simultaneous eigenstate of the
former group then, in general, it is not in an eigenstate of the latter. In other
112
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
words, if the quantum numbers j
1
, j
2
, j, and m are known with certainty, then a
measurement of the quantum numbers m
1
and m
2
will give a range of possible
values. We can use the completeness relation (5.270) to write
j
1
, j
2
; j, m) =
m
1
m
2
¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m)j
1
, j
2
; m
1
, m
2
). (5.272)
Thus, we can write the eigenkets of the ﬁrst group of operators as a weighted
sum of the eigenkets of the second set. The weights, ¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m),
are called the ClebschGordon coefﬁcients. If the system is in a state where a
measurement of J
2
1
, J
2
2
, J
2
, and J
z
is bound to give the results j
1
(j
1
+1) ¯h
2
, j
2
(j
2
+
1) ¯h
2
, j (j + 1) ¯h
2
, and j
z
¯h, respectively, then a measurement of J
1z
and J
2z
will
give the results m
1
¯h and m
2
¯h with probability ¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m)
2
.
The ClebschGordon coefﬁcients possess a number of very important proper
ties. First, the coefﬁcients are zero unless
m = m
1
+m
2
. (5.273)
To prove this, we note that
(J
z
−J
1z
−J
2z
)j
1
, j
2
; j, m) = 0. (5.274)
Forming the inner product with ¸j
1
, j
2
; m
1
, m
2
, we obtain
(m−m
1
−m
2
)¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m) = 0, (5.275)
which proves the assertion. Thus, the zcomponents of different angular mo
menta add algebraically. So, an electron in an l = 1 state, with orbital angular
momentum ¯h, and spin angular momentum ¯h/2, projected along the zaxis, con
stitutes a state whose total angular momentum projected along the zaxis is 3¯h/2.
What is uncertain is the magnitude of the total angular momentum.
Second, the coefﬁcients vanish unless
j
1
−j
2
 ≤ j ≤ j
1
+j
2
. (5.276)
We can assume, without loss of generality, that j
1
≥ j
2
. We know, fromEq. (5.273),
that for given j
1
and j
2
the largest possible value of m is j
1
+ j
2
(since j
1
is the
113
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
largest possible value of m
1
, etc.). This implies that the largest possible value of j
is j
1
+j
2
(since, by deﬁnition, the largest value of m is equal to j). Now, there are
(2 j
1
+1) allowable values of m
1
and (2 j
2
+1) allowable values of m
2
. Thus, there
are (2 j
1
+ 1) (2 j
2
+ 1) independent eigenkets, j
1
, j
2
; m
1
, m
2
), needed to span the
ket space corresponding to ﬁxed j
1
and j
2
. Since the eigenkets j
1
, j
2
; j, m) span
the same space, they must also form a set of (2 j
1
+1) (2 j
2
+1) independent kets.
In other words, there can only be (2 j
1
+ 1) (2 j
2
+ 1) distinct allowable values of
the quantum numbers j and m. For each allowed value of j, there are 2 j + 1
allowed values of m. We have already seen that the maximum allowed value of
j is j
1
+ j
2
. It is easily seen that if the minimum allowed value of j is j
1
− j
2
then
the total number of allowed values of j and m is (2 j
1
+1) (2 j
2
+1): i.e.,
j
1
+j
2
j=j
1
−j
2
(2 j +1) ≡ (2 j
1
+1) (2 j
2
+1). (5.277)
This proves our assertion.
Third, the sum of the modulus squared of all of the ClebschGordon coefﬁ
cients is unity: i.e.,
m
1
m
2
¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m)
2
= 1. (5.278)
This assertion is proved as follows:
¸j
1
, j
2
; j, mj
1
, j
2
; j, m) =
m
1
m
2
¸j
1
, j
2
; j, mj
1
, j
2
; m
1
, m
2
)¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m)
=
m
1
m
2
¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m)
2
= 1, (5.279)
where use has been made of the completeness relation (5.270).
Finally, the ClebschGordon coefﬁcients obey two recursion relations. To ob
tain these relations we start from
J
±
j
1
, j
2
; j, m) = (J
±
1
+J
±
2
) (5.280)
m
1
m
2
¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m)j
1
, j
2
; m
1
, m
2
).
114
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
Making use of the wellknown properties of the shift operators, which are speci
ﬁed by Eqs. (5.55)–(5.56), we obtain
_
j (j +1) −m(m±1) j
1
, j
2
; j, m±1) =
m
1
m
2
_
_
j
1
(j
1
+1) −m
1
(m
1
±1) j
1
, j
2
; m
1
±1, m
2
)
+
_
j
2
(j
2
+1) −m
2
(m
2
±1) j
1
, j
2
; m
1
, m
2
±1)
_
¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m). (5.281)
Taking the inner product with ¸j
1
, j
2
; m
1
, m
2
, and making use of the orthonormal
ity property of the basis eigenkets, we obtain the desired recursion relations:
_
j (j +1) −m(m±1) ¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m±1) =
_
j
1
(j
1
+1) −m
1
(m
1
∓1) ¸j
1
, j
2
; m
1
∓1, m
2
j
1
, j
2
; j, m)
+
_
j
2
(j
2
+1) −m
2
(m
2
∓1) ¸j
1
, j
2
; m
1
, m
2
∓1j
1
, j
2
; j, m). (5.282)
It is clear, from the absence of complex coupling coefﬁcients in the above re
lations, that we can always choose the ClebschGordon coefﬁcients to be real
numbers. This is a convenient choice, since it ensures that the inverse Clebsch
Gordon coefﬁcients, ¸j
1
, j
2
; j, mj
1
, j
2
; m
1
, m
2
), are identical to the ClebschGordon
coefﬁcients. In other words,
¸j
1
, j
2
; j, mj
1
, j
2
; m
1
, m
2
) = ¸j
1
, j
2
; m
1
, m
2
j
1
, j
2
; j, m). (5.283)
The inverse ClebschGordon coefﬁcients are the weights in the expansion of the
j
1
, j
2
; m
1
, m
2
) in terms of the j
1
, j
2
; j, m):
j
1
, j
2
; m
1
, m
2
) =
j
m
¸j
1
, j
2
; j, mj
1
, j
2
; m
1
, m
2
)j
1
, j
2
; j, m). (5.284)
It turns out that the recursion relations (5.282), together with the normal
ization condition (5.278), are sufﬁcient to completely determine the Clebsch
Gordon coefﬁcients to within an arbitrary sign (multiplied into all of the coefﬁ
cients). This sign is ﬁxed by convention. The easiest way of demonstrating this
assertion is by considering some speciﬁc examples.
115
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
Let us add the angular momentum of two spin onehalf systems: e.g., two
electrons at rest. So, j
1
= j
2
= 1/2. We know, from general principles, that
m
1
 ≤ 1/2 and m
2
 ≤ 1/2. We also know, from Eq. (5.276), that 0 ≤ j ≤ 1,
where the allowed values of j differ by integer amounts. It follows that either
j = 0 or j = 1. Thus, two spin onehalf systems can be combined to form either a
spin zero system or a spin one system. It is helpful to arrange all of the possibly
nonzero ClebschGordon coefﬁcients in a table:
m
1
m
2
1/2 1/2 ? ? ? ?
1/2 1/2 ? ? ? ?
1/2 1/2 ? ? ? ?
1/2 1/2 ? ? ? ?
j
1
=1/2 j 1 1 1 0
j
2
=1/2 m 1 0 1 0
The box in this table corresponding to m
1
= 1/2, m
2
= 1/2, j = 1, m = 1
gives the ClebschGordon coefﬁcient ¸1/2, 1/2; 1/2, 1/21/2, 1/2; 1, 1), or the in
verse ClebschGordon coefﬁcient ¸1/2, 1/2; 1, 11/2, 1/2; 1/2, 1/2). All the boxes
contain question marks because we do not know any ClebschGordon coefﬁcients
at the moment.
A ClebschGordon coefﬁcient is automatically zero unless m
1
+ m
2
= m. In
other words, the zcomponents of angular momentum have to add algebraically.
Many of the boxes in the above table correspond to m
1
+m
2
,= m. We immediately
conclude that these boxes must contain zeroes: i.e.,
m
1
m
2
1/2 1/2 ? 0 0 0
1/2 1/2 0 ? 0 ?
1/2 1/2 0 ? 0 ?
1/2 1/2 0 0 ? 0
j
1
=1/2 j 1 1 1 0
j
2
=1/2 m 1 0 1 0
The normalization condition (5.278) implies that the sum of the squares of all
116
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
the rows and columns of the above table must be unity. There are two rows and
two columns which only contain a single nonzero entry. We conclude that these
entries must be ±1, but we have no way of determining the signs at present.
Thus,
m
1
m
2
1/2 1/2 ±1 0 0 0
1/2 1/2 0 ? 0 ?
1/2 1/2 0 ? 0 ?
1/2 1/2 0 0 ±1 0
j
1
=1/2 j 1 1 1 0
j
2
=1/2 m 1 0 1 0
Let us evaluate the recursion relation (5.282) for j
1
= j
2
= 1/2, with j = 1,
m = 0, m
1
= m
2
= ±1/2, taking the upper/lower sign. We ﬁnd that
¸1/2, −1/21, 0) +¸−1/2, 1/21, 0) =
√
2 ¸1/2, 1/21, 1) = ±
√
2, (5.285)
and
¸1/2, −1/21, 0) +¸−1/2, 1/21, 0) =
√
2 ¸−1/2, −1/21, −1) = ±
√
2. (5.286)
Here, the j
1
and j
2
labels have been suppressed for ease of notation. We also
know that
¸1/2, −1/21, 0)
2
+¸−1/2, 1/21, 0)
2
= 1, (5.287)
from the normalization condition. The only real solutions to the above set of
equations are
√
2 ¸1/2, −1/21, 0) =
√
2 ¸−1/2, 1/21, 0)
= ¸1/2, 1/21, 1) = ¸1/2, 1/21, −1) = ±1. (5.288)
The choice of sign is arbitrary—the conventional choice is a positive sign. Thus,
117
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
our table now reads
m
1
m
2
1/2 1/2 1 0 0 0
1/2 1/2 0 1/
√
2 0 ?
1/2 1/2 0 1/
√
2 0 ?
1/2 1/2 0 0 1 0
j
1
=1/2 j 1 1 1 0
j
2
=1/2 m 1 0 1 0
We could ﬁll in the remaining unknown entries of our table by using the re
cursion relation again. However, an easier method is to observe that the rows
and columns of the table must all be mutually orthogonal. That is, the dot prod
uct of a row with any other row must be zero. Likewise, for the dot product of
a column with any other column. This follows because the entries in the table
give the expansion coefﬁcients of one of our alternative sets of eigenkets in terms
of the other set, and each set of eigenkets contains mutually orthogonal vectors
with unit norms. The normalization condition tells us that the dot product of a
row or column with itself must be unity. The only way that the dot product of the
fourth column with the second column can be zero is if the unknown entries are
equal and opposite. The requirement that the dot product of the fourth column
with itself is unity tells us that the magnitudes of the unknown entries have to
be 1/
√
2. The unknown entries are undetermined to an arbitrary sign multiplied
into them both. Thus, the ﬁnal form of our table (with the conventional choice
of arbitrary signs) is
m
1
m
2
1/2 1/2 1 0 0 0
1/2 1/2 0 1/
√
2 0 1/
√
2
1/2 1/2 0 1/
√
2 0 1/
√
2
1/2 1/2 0 0 1 0
j
1
=1/2 j 1 1 1 0
j
2
=1/2 m 1 0 1 0
The table can be read in one of two ways. The columns give the expansions of
118
5.14 Addition of angular momentum 5 ANGULAR MOMENTUM
the eigenstates of overall angular momentum in terms of the eigenstates of the
individual angular momenta of the two component systems. Thus, the second
column tells us that
1, 0) =
1
√
2
( 1/2, −1/2) +  −1/2, 1/2) ) . (5.289)
The ket on the lefthand side is a j, m) ket, whereas those on the righthand side
are m
1
, m
2
) kets. The rows give the expansions of the eigenstates of individual
angular momentum in terms of those of overall angular momentum. Thus, the
second row tells us that
1/2, −1/2) =
1
√
2
( 1, 0) + 0, 0) ) . (5.290)
Here, the ket on the lefthand side is a m
1
, m
2
) ket, whereas those on the right
hand side are j, m) kets.
Note that our table is really a combination of two subtables, one involving
j = 0 states, and one involving j = 1 states. The ClebschGordon coefﬁcients cor
responding to two different choices of j are completely independent: i.e., there is
no recursion relation linking ClebschGordon coefﬁcients corresponding to differ
ent values of j. Thus, for every choice of j
1
, j
2
, and j we can construct a table of
ClebschGordon coefﬁcients corresponding to the different allowed values of m
1
,
m
2
, and m (subject to the constraint that m
1
+m
2
= m). A complete knowledge
of angular momentum addition is equivalent to a knowing all possible tables of
ClebschGordon coefﬁcients. These tables are listed (for moderate values of j
1
, j
2
and j) in many standard reference books.
119
6 APPROXIMATION METHODS
6 Approximation methods
6.1 Introduction
We have developed techniques by which the general energy eigenvalue problem
can be reduced to a set of coupled partial differential equations involving various
wavefunctions. Unfortunately, the number of such problems which yield exactly
soluble equations is comparatively small. Clearly, we need to develop some tech
niques for ﬁnding approximate solutions to otherwise intractable problems.
Consider the following problem, which is very common. The Hamiltonian of a
system is written
H = H
0
+H
1
. (6.1)
Here, H
0
is a simple Hamiltonian for which we know the exact eigenvalues and
eigenstates. H
1
introduces some interesting additional physics into the problem,
but it is sufﬁciently complicated that when we add it to H
0
we can no longer
ﬁnd the exact energy eigenvalues and eigenstates. However, H
1
can, in some
sense (which we shall specify more exactly later on), be regarded as being small
compared to H
0
. Can we ﬁnd the approximate eigenvalues and eigenstates of the
modiﬁed Hamiltonian, H
0
+H
1
, by performing some sort of perturbation analysis
about the eigenvalues and eigenstates of the original Hamiltonian, H
0
? Let us
investigate.
6.2 The twostate system
Let us begin by considering timeindependent perturbation theory, in which the
modiﬁcation to the Hamiltonian, H
1
, has no explicit dependence on time. It is
usually assumed that the unperturbed Hamiltonian, H
0
, is also timeindependent.
Consider the simplest nontrivial system, in which there are only two indepen
dent eigenkets of the unperturbed Hamiltonian. These are denoted
H
0
1) = E
1
1), (6.2)
120
6.2 The twostate system 6 APPROXIMATION METHODS
H
0
2) = E
2
2). (6.3)
It is assumed that these states, and their associated eigenvalues, are known. Since
H
0
is, by deﬁnition, an Hermitian operator, its two eigenkets are orthonormal and
form a complete set. The lengths of these eigenkets are both normalized to unity.
Let us now try to solve the modiﬁed energy eigenvalue problem
(H
0
+H
1
)E) = EE). (6.4)
In fact, we can solve this problem exactly. Since, the eigenkets of H
0
form a
complete set, we can write
E) = ¸1E)1) +¸2E)2). (6.5)
Rightmultiplication of Eq. (6.4) by ¸1 and ¸2 yields two coupled equations,
which can be written in matrix form:
_
_
E
1
−E +e
11
e
12
e
∗
12
E
2
−E +e
22
_
_
_
_
¸1E)
¸2E)
_
_
=
_
_
0
0
_
_
. (6.6)
Here,
e
11
= ¸1H
1
1), (6.7)
e
22
= ¸2H
1
2), (6.8)
e
12
= ¸1H
1
2). (6.9)
In the special (but common) case of a perturbing Hamiltonian whose diagonal
matrix elements (in the unperturbed eigenstates) are zero, so that
e
11
= e
22
= 0, (6.10)
the solution of Eq. (6.6) (obtained by setting the determinant of the matrix equal
to zero) is
E =
(E
1
+E
2
) ±
_
(E
1
−E
2
)
2
+4 e
12

2
2
. (6.11)
Let us expand in the supposedly small parameter
=
e
12

E
1
−E
2

. (6.12)
121
6.3 Nondegenerate perturbation theory 6 APPROXIMATION METHODS
We obtain
E ·
1
2
(E
1
+E
2
) ±
1
2
(E
1
−E
2
)(1 +2
2
+ ). (6.13)
The above expression yields the modiﬁcations to the energy eigenvalues due to
the perturbing Hamiltonian:
E
1
= E
1
+
e
12

2
E
1
−E
2
+ , (6.14)
E
2
= E
2
−
e
12

2
E
1
−E
2
+ . (6.15)
Note that H
1
causes the upper eigenvalue to rise, and the lower eigenvalue to
fall. It is easily demonstrated that the modiﬁed eigenkets take the form
1)
= 1) +
e
∗
12
E
1
−E
2
2) + , (6.16)
2)
= 2) −
e
12
E
1
−E
2
1) + . (6.17)
Thus, the modiﬁed energy eigenstates consist of one of the unperturbed eigen
states with a slight admixture of the other. Note that the series expansion in
Eq. (6.13) only converges if 2  < 1. This suggests that the condition for the
validity of the perturbation expansion is
e
12
 <
E
1
−E
2

2
. (6.18)
In other words, when we say that H
1
needs to be small compared to H
0
, what we
really mean is that the above inequality needs to be satisﬁed.
6.3 Nondegenerate perturbation theory
Let us now generalize our perturbation analysis to deal with systems possessing
more than two energy eigenstates. The energy eigenstates of the unperturbed
Hamiltonian, H
0
, are denoted
H
0
n) = E
n
n), (6.19)
122
6.3 Nondegenerate perturbation theory 6 APPROXIMATION METHODS
where n runs from 1 to N. The eigenkets n) are orthonormal, form a complete
set, and have their lengths normalized to unity. Let us now try to solve the energy
eigenvalue problem for the perturbed Hamiltonian:
(H
0
+H
1
)E) = EE). (6.20)
We can express E) as a linear superposition of the unperturbed energy eigenkets,
E) =
k
¸kE)k), (6.21)
where the summation is from k = 1 to N. Substituting the above equation into
Eq. (6.20), and rightmultiplying by ¸m, we obtain
(E
m
+e
mm
−E)¸mE) +
k=m
e
mk
¸kE) = 0, (6.22)
where
e
mk
= ¸mH
1
k). (6.23)
Let us now develop our perturbation expansion. We assume that
e
mk

E
m
−E
k
∼ O(), (6.24)
for all m ,= k, where ¸1 is our expansion parameter. We also assume that
e
mm

E
m
∼ O(), (6.25)
for all m. Let us search for a modiﬁed version of the nth unperturbed energy
eigenstate, for which
E = E
n
+O(), (6.26)
and
¸nE) = 1, (6.27)
¸mE) ∼ O(), (6.28)
for m ,= n. Suppose that we write out Eq. (6.22) for m ,= n, neglecting terms
which are O(
2
) according to our expansion scheme. We ﬁnd that
(E
m
−E
n
)¸mE) +e
mn
· 0, (6.29)
123
6.4 The quadratic Stark effect 6 APPROXIMATION METHODS
giving
¸mE) · −
e
mn
E
m
−E
n
. (6.30)
Substituting the above expression into Eq. (6.22), evaluated for m = n, and
neglecting O(
3
) terms, we obtain
(E
n
+e
nn
−E) −
k=n
e
nk

2
E
k
−E
n
= 0. (6.31)
Thus, the modiﬁed nth energy eigenstate possesses an eigenvalue
E
n
= E
n
+e
nn
+
k=n
e
nk

2
E
n
−E
k
+O(
3
), (6.32)
and a eigenket
n)
= n) +
k=n
e
kn
E
n
−E
k
k) +O(
2
). (6.33)
Note that
¸mn)
= δ
mn
+
e
∗
nm
E
m
−E
n
+
e
mn
E
n
−E
m
+O(
2
) = δ
mn
+O(
2
). (6.34)
Thus, the modiﬁed eigenkets remain orthonormal and properly normalized to
O(
2
).
6.4 The quadratic Stark effect
Suppose that a oneelectron atom [i.e., either a hydrogen atom, or an alkali metal
atom (which possesses one valance electron orbiting outside a closed, spher
ically symmetric shell)] is subjected to a uniform electric ﬁeld in the positive
zdirection. The Hamiltonian of the system can be split into two parts. The un
perturbed Hamiltonian,
H
0
=
p
2
2 m
e
+V(r), (6.35)
and the perturbing Hamiltonian,
H
1
= e E z. (6.36)
124
6.4 The quadratic Stark effect 6 APPROXIMATION METHODS
It is assumed that the unperturbed energy eigenvalues and eigenstates are
completely known. The electron spin is irrelevant in this problem (since the spin
operators all commute with H
1
), so we can ignore the spin degrees of freedom of
the system. This implies that the system possesses no degenerate energy eigen
values. This is not true for the n ,= 1 energy levels of the hydrogen atom, due to
the special properties of a pure Coulomb potential. It is necessary to deal with
this case separately, because the perturbation theory presented in Sect. 6.3 breaks
down for degenerate unperturbed energy levels.
An energy eigenket of the unperturbed Hamiltonian is characterized by three
quantum numbers—the radial quantum number n, and the two angular quantum
numbers l and m (see Sect. 5.6). Let us denote such a ket n, l, m), and let its
energy level be E
nlm
. According to Eq. (6.32), the change in this energy level
induced by a small electric ﬁeld is given by
∆E
nlm
= e E ¸n, l, mzn, l, m)
+e
2
E
2
n
,l
,m
=n,l,m
¸n, l, mzn,
l
, m
)
2
E
nlm
−E
n
l
m
. (6.37)
Now, since
L
z
= x p
y
−yp
x
, (6.38)
it follows that
[L
z
, z] = 0. (6.39)
Thus,
¸n, l, m[L
z
, z]n
, l
, m
) = 0, (6.40)
giving
(m−m
)¸n, l, mzn
, l
, m
) = 0, (6.41)
since n, l, m) is, by deﬁnition, an eigenstate of L
z
with eigenvalue m¯h. It is
clear, from the above relation, that the matrix element ¸n, l, mzn
, l
, m
) is zero
unless m
= m. This is termed the selection rule for the quantum number m.
Let us now determine the selection rule for l. We have
[L
2
, z] = [L
2
x
, z] + [L
2
y
, z]
125
6.4 The quadratic Stark effect 6 APPROXIMATION METHODS
= L
x
[L
x
, z] + [L
x
, z]L
x
+L
y
[L
y
, z] + [L
y
, z]L
y
= i ¯h(−L
x
y −yL
x
+L
y
x +x L
y
)
= 2 i ¯h(L
y
x −L
x
y + i ¯hz)
= 2 i ¯h(L
y
x −yL
x
) = 2 i ¯h(x L
y
−L
x
y), (6.42)
where use has been made of Eqs. (5.1)–(5.6). Similarly,
[L
2
, y] = 2 i ¯h(L
x
z −x L
z
), (6.43)
[L
2
, x] = 2 i ¯h(yL
z
−L
y
z). (6.44)
Thus,
[L
2
, [L
2
, z]] = 2 i ¯h
_
L
2
, L
y
x −L
x
y + i ¯hz
_
= 2 i ¯h
_
L
y
[L
2
, x] −L
x
[L
2
, y] + i ¯h[L
2
, z]
_
,
= −4 ¯h
2
L
y
(yL
z
−L
y
z) +4 ¯h
2
L
x
(L
x
z −x L
z
)
−2 ¯h
2
(L
2
z −z L
2
). (6.45)
This reduces to
[L
2
, [L
2
, z]] = −¯h
2
_
4 (L
x
x +L
y
y +L
z
z)L
z
−4 (L
2
x
+L
2
y
+L
2
z
) z
+2 (L
2
z −z L
2
)
_
. (6.46)
However, it is clear from Eqs. (5.1)–(5.3) that
L
x
x +L
y
y +L
z
z = 0. (6.47)
Hence, we obtain
[L
2
, [L
2
, z]] = 2 ¯h
2
(L
2
z +z L
2
). (6.48)
Finally, the above expression expands to give
L
4
z −2 L
2
z L
2
+z L
4
−2 ¯h
2
(L
2
z +z L
2
) = 0. (6.49)
Equation (6.49) implies that
¸n, l, mL
4
z −2 L
2
z L
2
+z L
4
−2 ¯h
2
(L
2
z +z L
2
)n
, l
, m
) = 0. (6.50)
126
6.4 The quadratic Stark effect 6 APPROXIMATION METHODS
This expression yields
_
l
2
(l +1)
2
−2 l (l +1) l
(l
+1) +l
2
(l
+1)
2
−2 l (l +1) −2 l
(l
+1)] ¸n, l, mzn
, l
, m
) = 0, (6.51)
which reduces to
(l +l
+2) (l +l
) (l −l
+1) (l −l
−1)¸n, l, mzn
, l
, m
) = 0. (6.52)
According to the above formula, the matrix element ¸n, l, mzn
, l
, m
) vanishes
unless l = l
= 0 or l
= l ±1. This matrix element can be written
¸n, l, mzn
, l
, m
) =
___
ψ
∗
nlm
(r
, θ
, ϕ
) r
cos θ
ψ
n
m
l
(r
, θ
, ϕ
) dV
, (6.53)
where ψ
nlm
(r
) = ¸r
n, l, m). Recall, however, that the wavefunction of an l = 0
state is spherically symmetric (see Sect. 5.3): i.e., ψ
n00
(r
) = ψ
n00
(r
). It follows
from Eq. (6.53) that the matrix element vanishes by symmetry when l = l
= 0.
In conclusion, the matrix element ¸n, l, mzn
, l
, m
) is zero unless l
= l ± 1.
This is the selection rule for the quantum number l.
Application of the selection rules to Eq. (6.37) yields
∆E
nlm
= e
2
E
2
n
l
=l±1
¸n, l, mzn
, l
, m)
2
E
nlm
−E
n
l
m
. (6.54)
Note that all of the terms in Eq. (6.37) which vary linearly with the electric ﬁeld
strength vanish by symmetry, according to the selection rules. Only those terms
which vary quadratically with the ﬁeldstrength survive. The polarizability of an
atom is deﬁned in terms of the energyshift of the atomic state as follows:
∆E = −
1
2
αE
2
. (6.55)
Consider the ground state of a hydrogen atom. (Recall, that we cannot address
the n > 1 excited states because they are degenerate, and our theory cannot
handle this at present). The polarizability of this state is given by
α = 2 e
2
n>1
¸1, 0, 0zn, 1, 0)
2
E
n00
−E
100
. (6.56)
127
6.4 The quadratic Stark effect 6 APPROXIMATION METHODS
Here, we have made use of the fact that E
n10
= E
n00
for a hydrogen atom.
The sum in the above expression can be evaluated approximately by noting
that [see Eq. (5.120)]
E
n00
= −
e
2
8π
0
a
0
n
2
(6.57)
for a hydrogen atom, where
a
0
=
4π
0
¯h
2
µe
2
= 5.3 10
−11
meters (6.58)
is the Bohr radius. We can write
E
n00
−E
100
≥ E
200
−E
100
=
3
4
e
2
8π
0
a
0
. (6.59)
Thus,
α <
16
3
4π
0
a
0
n>1
¸1, 0, 0zn, 1, 0)
2
. (6.60)
However,
n>1
¸1, 0, 0zn, 1, 0)
2
=
n
,l
,m
¸1, 0, 0zn
, l
, m
)¸n
, m
, l
z1, 0, 0)
= ¸1, 0, 0z
2
1, 0, 0), (6.61)
where we have made use of the fact that the wavefunctions of a hydrogen atom
form a complete set. It is easily demonstrated from the actual form of the ground
state wavefunction that
¸1, 0, 0z
2
1, 0, 0) = a
2
0
. (6.62)
Thus, we conclude that
α <
16
3
4π
0
a
3
0
· 5.3 4π
0
a
3
0
. (6.63)
The true result is
α =
9
2
4π
0
a
3
0
= 4.5 4π
0
a
3
0
. (6.64)
It is actually possible to obtain this answer, without recourse to perturbation the
ory, by solving Schr¨ odinger’s equation exactly in parabolic coordinates.
128
6.5 Degenerate perturbation theory 6 APPROXIMATION METHODS
6.5 Degenerate perturbation theory
Let us now consider systems in which the eigenstates of the unperturbed Hamil
tonian, H
0
, possess degenerate energy levels. It is always possible to represent
degenerate energy eigenstates as the simultaneous eigenstates of the Hamilto
nian and some other Hermitian operator (or group of operators). Let us denote
this operator (or group of operators) L. We can write
H
0
n, l) = E
n
n, l), (6.65)
and
Ln, l) = L
nl
n, l), (6.66)
where [H
0
, L] = 0. Here, the E
n
and the L
nl
are real numbers which depend on
the quantum numbers n, and n and l, respectively. It is always possible to ﬁnd a
sufﬁcient number of operators which commute with the Hamiltonian in order to
ensure that the L
nl
are all different. In other words, we can choose L such that
the quantum numbers n and l uniquely specify each eigenstate. Suppose that for
each value of n there are N
n
different values of l: i.e., the nth energy eigenstate
is N
n
fold degenerate.
In general, L does not commute with the perturbing Hamiltonian, H
1
. This
implies that the modiﬁed energy eigenstates are not eigenstates of L. In this situ
ation, we expect the perturbation to split the degeneracy of the energy levels, so
that each modiﬁed eigenstate n, l)
acquires a unique energy eigenvalue E
nl
. Let
us naively attempt to use the standard perturbation theory of Sect. 6.3 to evalu
ate the modiﬁed energy eigenstates and energy levels. A direct generalization of
Eqs. (6.32) and (6.33) yields
E
nl
= E
n
+e
nlnl
+
n
,l
=n,l
e
n
l
nl

2
E
n
−E
n
+O(
3
), (6.67)
and
n, l)
= n, l) +
n
,l
=n,l
e
n
l
nl
E
n
−E
n
n
, l
) +O(
2
), (6.68)
where
e
n
l
nl
= ¸n
, l
H
1
n, l). (6.69)
129
6.5 Degenerate perturbation theory 6 APPROXIMATION METHODS
It is fairly obvious that the summations in Eqs. (6.67) and (6.68) are not well
behaved if the nth energy level is degenerate. The problem terms are those
involving unperturbed eigenstates labeled by the same value of n, but different
values of l: i.e., those states whose unperturbed energies are E
n
. These terms
give rise to singular factors 1/(E
n
− E
n
) in the summations. Note, however, that
this problem would not exist if the matrix elements, e
nl
nl
, of the perturbing
Hamiltonian between distinct, degenerate, unperturbed energy eigenstates cor
responding to the eigenvalue E
n
were zero. In other words, if
¸n, l
H
1
n, l) = λ
nl
δ
ll
, (6.70)
then all of the singular terms in Eqs. (6.67) and (6.68) would vanish.
In general, Eq. (6.70) is not satisﬁed. Fortunately, we can always redeﬁne the
unperturbed energy eigenstates belonging to the eigenvalue E
n
in such a manner
that Eq. (6.70) is satisﬁed. Let us deﬁne N
n
new states which are linear combina
tions of the N
n
original degenerate eigenstates corresponding to the eigenvalue
E
n
:
n, l
(1)
) =
N
n
k=1
¸n, kn, l
(1)
)n, k). (6.71)
Note that these new states are also degenerate energy eigenstates of the unper
turbed Hamiltonian corresponding to the eigenvalue E
n
. The n, l
(1)
) are chosen
in such a manner that they are eigenstates of the perturbing Hamiltonian, H
1
.
Thus,
H
1
n, l
(1)
) = λ
nl
n, l
(1)
). (6.72)
The n, l
(1)
) are also chosen so that they are orthonormal, and have unit lengths.
It follows that
¸n, l
(1)
H
1
n, l
(1)
) = λ
nl
δ
ll
. (6.73)
Thus, if we use the new eigenstates, instead of the old ones, then we can employ
Eqs. (6.67) and (6.68) directly, since all of the singular terms vanish. The only
remaining difﬁculty is to determine the new eigenstates in terms of the original
ones.
130
6.5 Degenerate perturbation theory 6 APPROXIMATION METHODS
Now
N
n
l=1
n, l)¸n, l = 1, (6.74)
where 1 denotes the identity operator in the subspace of all unperturbed energy
eigenkets corresponding to the eigenvalue E
n
. Using this completeness relation,
the operator eigenvalue equation (6.72) can be transformed into a straightfor
ward matrix eigenvalue equation:
N
n
l
=1
¸n, l
H
1
n, l
)¸n, l
n, l
(1)
) = λ
nl
¸n, l
n, l
(1)
). (6.75)
This can be written more transparently as
Ux = λ x, (6.76)
where the elements of the N
n
N
n
Hermitian matrix U are
U
jk
= ¸n, jH
1
n, k). (6.77)
Provided that the determinant of U is nonzero, Eq. (6.76) can always be solved
to give N
n
eigenvalues λ
nl
(for l = 1 to N
n
), with N
n
corresponding eigenvectors
x
nl
. The eigenvectors specify the weights of the new eigenstates in terms of the
original eigenstates: i.e.,
(x
nl
)
k
= ¸n, kn, l
(1)
), (6.78)
for k = 1 to N
n
. In our new scheme, Eqs. (6.67) and (6.68) yield
E
nl
= E
n
+λ
nl
+
n
=n,l
e
n
l
nl

2
E
n
−E
n
+O(
3
), (6.79)
and
n, l
(1)
)
= n, l
(1)
) +
n
=n,l
e
n
l
nl
E
n
−E
n
n
, l
) +O(
2
). (6.80)
There are no singular terms in these expressions, since the summations are over
n
,= n: i.e., they speciﬁcally exclude the problematic, degenerate, unperturbed
energy eigenstates corresponding to the eigenvalue E
n
. Note that the ﬁrstorder
energy shifts are equivalent to the eigenvalues of the matrix equation (6.76).
131
6.6 The linear Stark effect 6 APPROXIMATION METHODS
6.6 The linear Stark effect
Let us examine the effect of an electric ﬁeld on the excited energy levels of a
hydrogen atom. For instance, consider the n = 2 states. There is a single l =
0 state, usually referred to as 2s, and three l = 1 states (with m = −1, 0, 1),
usually referred to as 2p. All of these states possess the same energy, E
200
=
−e
2
/(32π
0
a
0
). As in Sect. 6.4, the perturbing Hamiltonian is
H
1
= e E z. (6.81)
In order to apply perturbation theory, we have to solve the matrix eigenvalue
equation
Ux = λ x, (6.82)
where U is the array of the matrix elements of H
1
between the degenerate 2s and
2p states. Thus,
U = e E
_
_
_
_
_
_
_
_
0 ¸2, 0, 0z2, 1, 0) 0 0
¸2, 1, 0z2, 0, 0) 0 0 0
0 0 0 0
0 0 0 0
_
_
_
_
_
_
_
_
, (6.83)
where the rows and columns correspond to the 2, 0, 0), 2, 1, 0), 2, 1, 1), and
2, 1, −1) states, respectively. Here, we have made use of the selection rules,
which tell us that the matrix element of z between two hydrogen atom states is
zero unless the states possess the same m quantum number, and l quantum num
bers which differ by unity. It is easily demonstrated, from the exact forms of the
2s and 2p wavefunctions, that
¸2, 0, 0z2, 1, 0) = ¸2, 1, 0z2, 0, 0) = 3 a
0
. (6.84)
It can be seen, by inspection, that the eigenvalues of U are λ
1
= 3 e a
0
E,
λ
2
= −3 e a
0
E, λ
3
= 0, and λ
4
= 0. The corresponding eigenvectors are
x
1
=
_
_
_
_
_
_
_
_
1/
√
2
1/
√
2
0
0
_
_
_
_
_
_
_
_
, (6.85)
132
6.6 The linear Stark effect 6 APPROXIMATION METHODS
x
2
=
_
_
_
_
_
_
_
_
1/
√
2
−1/
√
2
0
0
_
_
_
_
_
_
_
_
, (6.86)
x
3
=
_
_
_
_
_
_
_
_
0
0
1
0
_
_
_
_
_
_
_
_
, (6.87)
x
4
=
_
_
_
_
_
_
_
_
0
0
0
1
_
_
_
_
_
_
_
_
. (6.88)
It follows from Sect. 6.5 that the simultaneous eigenstates of the unperturbed
Hamiltonian and the perturbing Hamiltonian take the form
1) =
2, 0, 0) + 2, 1, 0)
√
2
, (6.89)
2) =
2, 0, 0) − 2, 1, 0)
√
2
, (6.90)
3) = 2, 1, 1), (6.91)
4) = 2, 1, −1). (6.92)
In the absence of an electric ﬁeld, all of these states possess the same energy, E
200
.
The ﬁrstorder energy shifts induced by an electric ﬁeld are given by
∆E
1
= +3 e a
0
E, (6.93)
∆E
2
= −3 e a
0
E, (6.94)
∆E
3
= 0, (6.95)
∆E
4
= 0. (6.96)
Thus, the energies of states 1 and 2 are shifted upwards and downwards, respec
tively, by an amount 3 e a
0
E in the presence of an electric ﬁeld. States 1 and
2 are orthogonal linear combinations of the original 2s and 2p(m = 0) states.
133
6.7 Fine structure 6 APPROXIMATION METHODS
Note that the energy shifts are linear in the electric ﬁeldstrength, so this is a
much larger effect that the quadratic effect described in Sect. 6.4. The energies
of states 3 and 4 (which are equivalent to the original 2p(m = 1) and 2p(m = −1)
states, respectively) are not affected to ﬁrstorder. Of course, to secondorder the
energies of these states are shifted by an amount which depends on the square of
the electric ﬁeldstrength.
Note that the linear Stark effect depends crucially on the degeneracy of the 2s
and 2p states. This degeneracy is a special property of a pure Coulomb potential,
and, therefore, only applies to a hydrogen atom. Thus, alkali metal atoms do not
exhibit the linear Stark effect.
6.7 Fine structure
Let us now consider the energy levels of hydrogenlike atoms (i.e., alkali metal
atoms) in more detail. The outermost electron moves in a spherically symmetric
potential V(r) due to the nuclear charge and the charges of the other electrons
(which occupy spherically symmetric closed shells). The shielding effect of the
inner electrons causes V(r) to depart from the pure Coulomb form. This splits the
degeneracy of states characterized by the same value of n, but different values of
l. In fact, higher l states have higher energies.
Let us examine a phenomenon known as ﬁne structure, which is due to inter
action between the spin and orbital angular momenta of the outermost electron.
This electron experiences an electric ﬁeld
E =
∇V
e
. (6.97)
However, a charge moving in an electric ﬁeld also experiences an effective mag
netic ﬁeld
B = −v E. (6.98)
Now, an electron possesses a spin magnetic moment [see Eq. (5.170)]
µ = −
e S
m
e
. (6.99)
134
6.7 Fine structure 6 APPROXIMATION METHODS
We, therefore, expect a spinorbit contribution to the Hamiltonian of the form
H
LS
= −µB
= −
e S
m
e
v
_
1
e
r
r
dV
dr
_
=
1
m
2
e
r
dV
dr
LS, (6.100)
where L = m
e
rv is the orbital angular momentum. When the above expression
is compared to the observed spinorbit interaction, it is found to be too large by
a factor of two. There is a classical explanation for this, due to spin precession,
which we need not go into. The correct quantummechanical explanation requires
a relativistically covariant treatment of electron dynamics (this is achieved using
the socalled Dirac equation).
Let us now apply perturbation theory to a hydrogenlike atom, using H
LS
as
the perturbation (with H
LS
taking one half of the value given above), and
H
0
=
p
2
2 m
e
+V(r) (6.101)
as the unperturbed Hamiltonian. We have two choices for the energy eigenstates
of H
0
. We can adopt the simultaneous eigenstates of H
0
, L
2
, S
2
, L
z
and S
z
, or the
simultaneous eigenstates of H
0
, L
2
, S
2
, J
2
, and J
z
, where J = L + S is the total
angular momentum. Although the departure of V(r) from a pure 1/r form splits
the degeneracy of same n, different l, states, those states characterized by the
same values of n and l, but different values of m
l
, are still degenerate. (Here,
m
l
, m
s
, and m
j
are the quantum numbers corresponding to L
z
, S
z
, and J
z
, respec
tively.) Moreover, with the addition of spin degrees of freedom, each state is
doubly degenerate due to the two possible orientations of the electron spin (i.e.,
m
s
= ±1/2). Thus, we are still dealing with a highly degenerate system. We
know, from Sect. 6.6, that the application of perturbation theory to a degenerate
system is greatly simpliﬁed if the basis eigenstates of the unperturbed Hamilto
nian are also eigenstates of the perturbing Hamiltonian. Now, the perturbing
Hamiltonian, H
LS
, is proportional to LS, where
LS =
J
2
−L
2
−S
2
2
. (6.102)
135
6.7 Fine structure 6 APPROXIMATION METHODS
It is fairly obvious that the ﬁrst group of operators (H
0
, L
2
, S
2
, L
z
and S
z
) does not
commute with H
LS
, whereas the second group (H
0
, L
2
, S
2
, J
2
, and J
z
) does. In
fact, LS is just a combination of operators appearing in the second group. Thus,
it is advantageous to work in terms of the eigenstates of the second group of
operators, rather than those of the ﬁrst group.
We now need to ﬁnd the simultaneous eigenstates of H
0
, L
2
, S
2
, J
2
, and J
z
. This
is equivalent to ﬁnding the eigenstates of the total angular momentum resulting
from the addition of two angular momenta: j
1
= l, and j
2
= s = 1/2. According
to Eq. (5.276), the allowed values of the total angular momentum are j = l +1/2
and j = l −1/2. We can write
l +1/2, m) = cos αm−1/2, 1/2) + sinαm+1/2, −1/2), (6.103)
l −1/2, m) = −sinαm−1/2, 1/2) + cos αm+1/2, −1/2). (6.104)
Here, the kets on the lefthand side are j, m
j
) kets, whereas those on the right
hand side are m
l
, m
s
) kets (the j
1
, j
2
labels have been dropped, for the sake of
clarity). We have made use of the fact that the ClebschGordon coefﬁcients are
automatically zero unless m
j
= m
l
+ m
s
. We have also made use of the fact that
both the j, m
j
) and m
l
, m
s
) kets are orthonormal, and have unit lengths. We
now need to determine
cos α = ¸m−1/2, 1/2l +1/2, m), (6.105)
where the ClebschGordon coefﬁcient is written in ¸m
l
, m
s
j, m
j
) form.
Let us now employ the recursion relation for ClebschGordon coefﬁcients,
Eq. (5.282), with j
1
= l, j
2
= 1/2, j = l + 1/2, m
1
= m − 1/2, m
2
= 1/2 (lower
sign). We obtain
_
(l +1/2) (l +3/2) −m(m+1) ¸m−1/2, 1/2l +1/2, m)
=
_
l (l +1) − (m−1/2) (m+1/2) ¸m+1/2, 1/2l +1/2, m+1), (6.106)
which reduces to
¸m−1/2, 1/2l +1/2, m) =
¸
¸
¸
¸
_
l +m+1/2
l +m+3/2
¸m+1/2, 1/2l +1/2, m+1). (6.107)
136
6.7 Fine structure 6 APPROXIMATION METHODS
We can use this formula to successively increase the value of m
l
. For instance,
¸m−1/2, 1/2l +1/2, m) =
¸
¸
¸
¸
_
l +m+1/2
l +m+3/2
¸
¸
¸
¸
_
l +m+3/2
l +m+5/2
¸m+3/2, 1/2l +1/2, m+2). (6.108)
This procedure can be continued until m
l
attains its maximum possible value, l.
Thus,
¸m−1/2, 1/2l +1/2, m) =
¸
¸
¸
_
l +m+1/2
2 l +1
¸l, 1/2l +1/2, l +1/2). (6.109)
Consider the situation in which m
l
and m both take their maximum values,
l and 1/2, respectively. The corresponding value of m
j
is l + 1/2. This value is
possible when j = l +1/2, but not when j = l −1/2. Thus, the m
l
, m
s
) ket l, 1/2)
must be equal to the j, m
j
) ket l + 1/2, l + 1/2), up to an arbitrary phasefactor.
By convention, this factor is taken to be unity, giving
¸l, 1/2l +1/2, l +1/2) = 1. (6.110)
It follows from Eq. (6.109) that
cos α = ¸m−1/2, 1/2l +1/2, m) =
¸
¸
¸
_
l +m+1/2
2 l +1
. (6.111)
Now,
sin
2
α = 1 −
l +m+1/2
2 l +1
=
l −m+1/2
2 l +1
. (6.112)
We now need to determine the sign of sinα. A careful examination of the
recursion relation, Eq. (5.282), shows that the plus sign is appropriate. Thus,
l +1/2, m) =
¸
¸
¸
_
l +m+1/2
2 l +1
m−1/2, 1/2)
+
¸
¸
¸
_
l −m+1/2
2 l +1
m+1/2, −1/2), (6.113)
137
6.7 Fine structure 6 APPROXIMATION METHODS
l −1/2, m) = −
¸
¸
¸
_
l −m+1/2
2 l +1
m−1/2, 1/2)
+
¸
¸
¸
_
l +m+1/2
2 l +1
m+1/2, −1/2). (6.114)
It is convenient to deﬁne so called spinangular functions using the Pauli two
component formalism:
¸
j=l±1/2,m
l
= ±
¸
¸
¸
_
l ±m+1/2
2 l +1
Y
m−1/2
l
(θ, ϕ) χ
+
+
¸
¸
¸
_
l ∓m+1/2
2 l +1
Y
m+1/2
l
(θ, ϕ) χ
−
=
1
√
2 l +1
_
_
_
±
_
l ±m+1/2 Y
m−1/2
l
(θ, ϕ)
_
l ∓m+1/2 Y
m+1/2
l
(θ, ϕ)
_
_
_. (6.115)
These functions are eigenfunctions of the total angular momentum for spin one
half particles, just as the spherical harmonics are eigenfunctions of the orbital
angular momentum. A general wavefunction for an energy eigenstate in a
hydrogenlike atom is written
ψ
nlm±
= R
nl
(r) ¸
j=l±1/2,m
. (6.116)
The radial part of the wavefunction, R
nl
(r), depends on the radial quantum num
ber n and the angular quantum number l. The wavefunction is also labeled by
m, which is the quantum number associated with J
z
. For a given choice of l, the
quantum number j (i.e., the quantum number associated with J
2
) can take the
values l ±1/2.
The l ±1/2, m) kets are eigenstates of LS, according to Eq. (6.102). Thus,
LSj = l ±1/2, m
j
= m) =
¯h
2
2
[j (j +1) −l (l +1) −3/4] j, m), (6.117)
giving
LSl +1/2, m) =
l ¯h
2
2
l +1/2, m), (6.118)
LSl −1/2, m) = −
(l +1) ¯h
2
2
l −1/2, m). (6.119)
138
6.7 Fine structure 6 APPROXIMATION METHODS
It follows that
_
(¸
l+1/2,m
)
†
LS¸
l+1/2,m
dΩ =
l ¯h
2
2
, (6.120)
_
(¸
l−1/2,m
)
†
LS¸
l−1/2,m
dΩ = −
(l +1) ¯h
2
2
, (6.121)
where the integrals are over all solid angle.
Let us nowapply degenerate perturbation theory to evaluate the shift in energy
of a state whose wavefunction is ψ
nlm±
due to the spinorbit Hamiltonian H
LS
.
To ﬁrstorder, the energyshift is given by
∆E
nlm±
=
_
(ψ
nlm±
)
†
H
LS
ψ
nlm±
dV, (6.122)
where the integral is over all space. Equations (6.100) (remember the factor of
two), (6.116), and (6.120)–(6.121) yield
∆E
nlm+
= +
1
2 m
2
e
_
1
r
dV
dr
_
l ¯h
2
2
, (6.123)
∆E
nlm−
= −
1
2 m
2
e
_
1
r
dV
dr
_
(l +1) ¯h
2
2
, (6.124)
where
_
1
r
dV
dr
_
=
_
(R
nl
)
∗
1
r
dV
dr
R
nl
r
2
dr. (6.125)
Equations (6.123)–(6.124) are known as Lande’s interval rule.
Let us now apply the above result to the case of a sodium atom. In chemist’s
notation, the ground state is written
(1s)
2
(2s)
2
(2p)
6
(3s). (6.126)
The inner ten electrons effectively form a spherically symmetric electron cloud.
We are interested in the excitation of the eleventh electron from 3s to some
higher energy state. The closest (in energy) unoccupied state is 3p. This state
has a higher energy than 3s due to the deviations of the potential from the pure
139
6.8 The Zeeman effect 6 APPROXIMATION METHODS
Coulomb form. In the absence of spinorbit interaction, there are six degenerate
3p states. The spinorbit interaction breaks the degeneracy of these states. The
modiﬁed states are labeled (3p)
1/2
and (3p)
3/2
, where the subscript refers to the
value of j. The four (3p)
3/2
states lie at a slightly higher energy level than the two
(3p)
1/2
states, because the radial integral (6.125) is positive. The splitting of the
(3p) energy levels of the sodium atom can be observed using a spectroscope. The
wellknown sodium D line is associated with transitions between the 3p and 3s
states. The fact that there are two slightly different 3p energy levels (note that
spinorbit coupling does not split the 3s energy levels) means that the sodium D
line actually consists of two very closely spaced spectroscopic lines. It is easily
demonstrated that the ratio of the typical spacing of Balmer lines to the splitting
brought about by spinorbit interaction is about 1 : α
2
, where
α =
e
2
2
0
hc
=
1
137
(6.127)
is the ﬁne structure constant. Note that Eqs. (6.123)–(6.124) are not entirely
correct, since we have neglected an effect (namely, the relativistic mass correction
of the electron) which is the same order of magnitude as spinorbit coupling.
6.8 The Zeeman effect
Consider a hydrogenlike atom placed in a uniform zdirected magnetic ﬁeld. The
change in energy of the outermost electron is
H
B
= −µB, (6.128)
where
µ = −
e
2 m
e
(L +2 S) (6.129)
is its magnetic moment, including both the spin and orbital contributions. Thus,
H
B
=
e B
2 m
e
(L
z
+2 S
z
). (6.130)
Suppose that the energyshifts induced by the magnetic ﬁeld are much smaller
than those induced by spinorbit interaction. In this situation, we can treat H
B
as
140
6.8 The Zeeman effect 6 APPROXIMATION METHODS
a small perturbation acting on the eigenstates of H
0
+H
LS
. Of course, these states
are the simultaneous eigenstates of J
2
and J
z
. Let us consider one of these states,
labeled by the quantum numbers j and m, where j = l ± 1/2. From standard
perturbation theory, the ﬁrstorder energyshift in the presence of a magnetic
ﬁeld is
∆E
nlm±
= ¸l ±1/2, mH
B
l ±1/2, m). (6.131)
Since
L
z
+2 S
z
= J
z
+S
z
, (6.132)
we ﬁnd that
∆E
nlm±
=
e B
2 m
e
(m¯h +¸l ±1/2, mS
z
l ±1/2, m) ) . (6.133)
Now, from Eqs. (6.113)–(6.114),
l ±1/2, m) = ±
¸
¸
¸
_
l ±m+1/2
2 l +1
m−1/2, 1/2)
+
¸
¸
¸
_
l ∓m+1/2
2 l +1
m+1/2, −1/2). (6.134)
It follows that
¸l ±1/2, mS
z
l ±1/2, m) =
¯h
2 (2 l +1)
[(l ±m+1/2) − (l ∓m+1/2)]
= ±
m¯h
2 l +1
. (6.135)
Thus, we obtain Lande’s formula for the energyshift induced by a weak magnetic
ﬁeld:
∆E
nlm±
=
e ¯hB
2 m
e
m
_
1 ±
1
2 l +1
_
. (6.136)
Let us apply this theory to the sodium atom. We have already seen that the
nonCoulomb potential splits the degeneracy of the 3s and 3p states, the latter
states acquiring a higher energy. The spinorbit interaction splits the six 3p states
into two groups, with four j = 3/2 states lying at a slightly higher energy than
two j = 1/2 states. According to Eq. (6.136), a magnetic ﬁeld splits the (3p)
3/2
141
6.8 The Zeeman effect 6 APPROXIMATION METHODS
quadruplet of states, each state acquiring a different energy. In fact, the energy of
each state becomes dependent on the quantum number m, which measures the
projection of the total angular momentum along the zaxis. States with higher
m values have higher energies. A magnetic ﬁeld also splits the (3p)
1/2
doublet
of states. However, it is evident from Eq. (6.136) that these states are split by a
lesser amount than the j = 3/2 states.
Suppose that we increase the strength of the magnetic ﬁeld, so that the energy
shift due to the magnetic ﬁeld becomes comparable to the energyshift induced
by spinorbit interaction. Clearly, in this situation, it does not make much sense to
think of H
B
as a small interaction term operating on the eigenstates of H
0
+H
LS
.
In fact, this intermediate case is very difﬁcult to analyze. Let us consider the
extreme limit in which the energyshift due to the magnetic ﬁeld greatly exceeds
that induced by spinorbit effects. This is called the PaschenBack limit.
In the PaschenBack limit we can think of the spinorbit Hamiltonian, H
LS
, as
a small interaction term operating on the eigenstates of H
0
+ H
B
. Note that the
magnetic Hamiltonian, H
B
, commutes with L
2
, S
2
, L
z
, S
z
, but does not commute
with L
2
, S
2
, J
2
, J
z
. Thus, in an intense magnetic ﬁeld, the energy eigenstates of a
hydrogenlike atom are approximate eigenstates of the spin and orbital angular
momenta, but are not eigenstates of the total angular momentum. We can label
each state by the quantum numbers n (the energy quantum number), l, m
l
,
and m
s
. Thus, our energy eigenkets are written n, l, m
l
, m
s
). The unperturbed
Hamiltonian, H
0
, causes states with different values of the quantum numbers n
and l to have different energies. However, states with the same value of n and l,
but different values of m
l
and m
s
, are degenerate. The shift in energy due to the
magnetic ﬁeld is simply
∆E
nlm
l
m
s
= ¸n, l, m
l
, m
s
H
B
n, l, m
l
, m
s
)
=
e ¯hB
2 m
e
(m
l
+2 m
s
). (6.137)
Thus, states with different values of m
l
+2 m
s
acquire different energies.
Let us apply this result to a sodium atom. In the absence of a magnetic ﬁeld,
the six 3p states form two groups of four and two states, depending on the values
142
6.8 The Zeeman effect 6 APPROXIMATION METHODS
of their total angular momentum. In the presence of an intense magnetic ﬁeld
the 3p states are split into ﬁve groups. There is a state with m
l
+2 m
s
= 2, a state
with m
l
+2 m
s
= 1, two states with m
l
+2 m
s
= 0, a state with m
l
+2 m
s
= −1,
and a state with m
l
+2 m
s
= −2. These groups are equally spaced in energy, the
energy difference between adjacent groups being e ¯hB/2 m
e
.
The energyshift induced by the spinorbit Hamiltonian is given by
∆E
nl m
l
m
s
= ¸n, l, m
l
, m
s
H
LS
n, l, m
l
, m
s
), (6.138)
where
H
LS
=
1
2 m
2
e
1
r
dV
dr
LS. (6.139)
Now,
¸LS) = ¸ L
z
S
z
+ (L
+
S
−
+L
−
S
+
)/2 )
= ¯h
2
m
l
m
s
, (6.140)
since
¸L
±
) = ¸S
±
) = 0 (6.141)
for expectation values taken between the simultaneous eigenkets of L
z
and S
z
.
Thus,
∆E
nlm
l
m
s
=
¯h
2
m
l
m
s
2 m
2
e
_
1
r
dV
dr
_
. (6.142)
Let us apply the above result to a sodium atom. In the presence of an intense
magnetic ﬁeld, the 3p states are split into ﬁve groups with (m
l
, m
s
) quantum
numbers (1, 1/2), (0, 1/2), (1, −1/2), or (−1, 1/2), (0, −1/2), and (−1, −1/2), re
spectively, in order of decreasing energy. The spinorbit term increases the en
ergy of the highest energy state, does not affect the next highest energy state,
decreases, but does not split, the energy of the doublet, does not affect the next
lowest energy state, and increases the energy of the lowest energy state. The net
result is that the ﬁve groups of states are no longer equally spaced in energy.
The sort of magnetic ﬁeldstrength needed to get into the PaschenBach limit
is given by
B
PB
∼ α
2
e m
e
0
ha
0
· 25 tesla. (6.143)
143
6.9 Timedependent perturbation theory 6 APPROXIMATION METHODS
Obviuously, this is an extremely large ﬁeldstrength.
6.9 Timedependent perturbation theory
Suppose that the Hamiltonian of the system under consideration can be written
H = H
0
+H
1
(t), (6.144)
where H
0
does not contain time explicitly, and H
1
is a small timedependent
perturbation. It is assumed that we are able to calculate the eigenkets of the
unperturbed Hamiltonian:
H
0
n) = E
n
n). (6.145)
We know that if the system is in one of the eigenstates of H
0
then, in the ab
sence of the external perturbation, it remains in this state for ever. However,
the presence of a small timedependent perturbation can, in principle, give rise
to a ﬁnite probability that a system initially in some eigenstate i) of the unper
turbed Hamiltonian is found in some other eigenstate at a subsequent time (since
i) is no longer an exact eigenstate of the total Hamiltonian), In other words, a
timedependent perturbation causes the system to make transitions between its
unperturbed energy eigenstates. Let us investigate this effect.
Suppose that at t = t
0
the state of the system is represented by
A) =
n
c
n
n), (6.146)
where the c
n
are complex numbers. Thus, the initial state is some linear su
perposition of the unperturbed energy eigenstates. In the absence of the time
dependent perturbation, the time evolution of the system is given by
A, t
0
, t) =
n
c
n
exp([−i E
n
(t −t
0
)/¯h] n). (6.147)
Now, the probability of ﬁnding the system in state n) at time t is
P
n
(t) = c
n
exp[−i E
n
(t −t
0
)/¯h]
2
= c
n

2
= P
n
(t
0
). (6.148)
144
6.9 Timedependent perturbation theory 6 APPROXIMATION METHODS
Clearly, with H
1
= 0, the probability of ﬁnding the system in state n) at time t is
exactly the same as the probability of ﬁnding the system in this state at the initial
time t
0
. However, with H
1
,= 0, we expect P
n
(t) to vary with time. Thus, we can
write
A, t
0
, t) =
n
c
n
(t) exp[−i E
n
(t −t
0
)/¯h] n), (6.149)
where P
n
(t) = c
n
(t)
2
. Here, we have carefully separated the fast phase oscilla
tion of the eigenkets, which depends on the unperturbed Hamiltonian, from the
slow variation of the amplitudes c
n
(t), which depends entirely on the perturba
tion (i.e., c
n
is constant if H
1
= 0). Note that in Eq. (6.149) the eigenkets n) are
timeindependent (they are actually the eigenkets of H
0
evaluated at the time t
0
).
Schr¨ odinger’s time evolution equation yields
i ¯h
∂
∂t
A, t
0
, t) = HA, t
0
, t) = (H
0
+H
1
) A, t
0
, t). (6.150)
It follows from Eq. (6.149) that
(H
0
+H
1
)A, t
0
, t) =
m
c
m
(t) exp[−i E
m
(t −t
0
)/¯h] (E
m
+H
1
) m). (6.151)
We also have
i ¯h
∂
∂t
A, t
0
, t) =
m
_
i ¯h
dc
m
dt
+c
m
(t) E
m
_
exp[−i E
m
(t −t
0
)/¯h] m), (6.152)
where use has been made of the timeindependence of the kets m). According
to Eq. (6.150), we can equate the righthand sides of the previous two equations
to obtain
m
i ¯h
dc
m
dt
exp[−i E
m
(t −t
0
)/¯h]m) =
m
c
m
(t) exp[−i E
m
(t −t
0
)/¯h] H
1
m).
(6.153)
Leftmultiplication by ¸n yields
i ¯h
dc
n
dt
=
m
H
nm
(t) exp[i ω
nm
(t −t
0
)] c
m
(t), (6.154)
where
H
nm
(t) = ¸nH
1
(t)m), (6.155)
145
6.10 The twostate system 6 APPROXIMATION METHODS
and
ω
nm
=
E
n
−E
m
¯h
. (6.156)
Here, we have made use of the standard orthonormality result, ¸nm) = δ
nm
.
Suppose that there are N linearly independent eigenkets of the unperturbed
Hamiltonian. According to Eq. (6.154), the time variation of the coefﬁcients
c
n
, which specify the probability of ﬁnding the system in state n) at time t, is de
termined by N coupled ﬁrstorder differential equations. Note that Eq. (6.154) is
exact—we have made no approximations at this stage. Unfortunately, we cannot
generally ﬁnd exact solutions to this equation, so we have to obtain approximate
solutions via suitable expansions in small quantities. However, for the particu
larly simple case of a twostate system (i.e., N = 2), it is actually possible to
solve Eq. (6.154) without approximation. This solution is of enormous practical
importance.
6.10 The twostate system
Consider a system in which the timeindependent Hamiltonian possesses two
eigenstates, denoted
H
0
1) = E
1
1), (6.157)
H
0
2) = E
2
2). (6.158)
Suppose, for the sake of simplicity, that the diagonal matrix elements of the in
teraction Hamiltonian, H
1
, are zero:
¸1H
1
1) = ¸2H
1
2) = 0. (6.159)
The offdiagonal matrix elements are assumed to oscillate sinusoidally at some
frequency ω:
¸1H
1
2) = ¸2H
1
1)
∗
= γexp(i ωt), (6.160)
where γ and ω are real. Note that it is only the offdiagonal matrix elements
which give rise to the effect which we are interested in—namely, transitions be
tween states 1 and 2.
146
6.10 The twostate system 6 APPROXIMATION METHODS
For a twostate system, Eq. (6.154) reduces to
i ¯h
dc
1
dt
= γexp[+i (ω−ω
21
) t ] c
2
, (6.161)
i ¯h
dc
2
dt
= γexp[−i (ω−ω
21
) t ] c
1
, (6.162)
where ω
21
= (E
2
− E
1
)/¯h, and assuming that t
0
= 0. Equations (6.161) and
(6.162) can be combined to give a secondorder differential equation for the
time variation of the amplitude c
2
:
d
2
c
2
dt
2
+ i (ω−ω
21
)
dc
2
dt
+
γ
2
¯h
2
c
2
= 0. (6.163)
Once we have solved for c
2
, we can use Eq. (6.162) to obtain the amplitude c
1
.
Let us look for a solution in which the system is certain to be in state 1 at time
t = 0. Thus, our boundary conditions are c
1
(0) = 1 and c
2
(0) = 0. It is easily
demonstrated that the appropriate solutions are
c
2
(t) =
−i γ/¯h
_
γ
2
/¯h
2
+ (ω−ω
21
)
2
/4
exp[−i (ω−ω
21
) t/2]
sin
__
γ
2
/¯h
2
+ (ω−ω
21
)
2
/4 t
_
, (6.164)
c
1
(t) = exp[ i (ω−ω
21
) t/2] cos
__
γ
2
/¯h
2
+ (ω−ω
21
)
2
/4 t
_
−
i (ω−ω
21
)/2
_
γ
2
/¯h
2
+ (ω−ω
21
)
2
/4
exp[ i (ω−ω
21
) t/2]
sin
__
γ
2
/¯h
2
+ (ω−ω
21
)
2
/4 t
_
. (6.165)
Now, the probability of ﬁnding the system in state 1 at time t is simply P
1
(t) =
c
1

2
. Likewise, the probability of ﬁnding the system in state 2 at time t is P
2
(t) =
c
2

2
. It follows that
P
2
(t) =
γ
2
/¯h
2
γ
2
/¯h
2
+ (ω−ω
21
)
2
/4
147
6.10 The twostate system 6 APPROXIMATION METHODS
sin
2
__
γ
2
/¯h
2
+ (ω−ω
21
)
2
/4 t
_
, (6.166)
P
1
(t) = 1 −P
2
(t). (6.167)
This result is known as Rabi’s formula.
Equation (6.166) exhibits all the features of a classic resonance. At resonance,
when the oscillation frequency of the perturbation, ω, matches the frequency
ω
21
, we ﬁnd that
P
1
(t) = cos
2
(γt/¯h), (6.168)
P
2
(t) = sin
2
(γt/¯h). (6.169)
According to the above result, the system starts off at t = 0 in state 1. After a
time interval π¯h/2 γ it is certain to be in state 2. After a further time interval
π¯h/2 γ it is certain to be in state 1, and so on. Thus, the system periodically
ﬂipﬂops between states 1 and 2 under the inﬂuence of the timedependent per
turbation. This implies that the system alternatively absorbs and emits energy
from the source of the perturbation.
The absorptionemission cycle also take place away from the resonance, when
ω ,= ω
21
. However, the amplitude of oscillation of the coefﬁcient c
2
is reduced.
This means that the maximum value of P
2
(t) is no longer unity, nor is the mini
mum value of P
1
(t) zero. In fact, if we plot the maximum value of P
2
(t) as a func
tion of the applied frequency, ω, we obtain a resonance curve whose maximum
(unity) lies at the resonance, and whose fullwidth halfmaximum (in frequency)
is 4 γ/¯h. Thus, if the applied frequency differs from the resonant frequency by
substantially more than 2 γ/¯h then the probability of the system jumping from
state 1 to state 2 is very small. In other words, the timedependent perturbation
is only effective at causing transitions between states 1 and 2 if its frequency of
oscillation lies in the approximate range ω
21
± 2 γ/¯h. Clearly, the weaker the
perturbation (i.e., the smaller γ becomes), the narrower the resonance.
148
6.11 Spin magnetic resonance 6 APPROXIMATION METHODS
6.11 Spin magnetic resonance
Consider a spin onehalf system (e.g., a bound electron) placed in a uniform z
directed magnetic ﬁeld, and then subjected to a small timedependent magnetic
ﬁeld rotating in the xy plane. Thus,
B = B
0
^z +B
1
(cos ωt ^ x + sinωt ^ y), (6.170)
where B
0
and B
1
are constants, with B
1
¸ B
0
. The rotating magnetic ﬁeld usu
ally represents the magnetic component of an electromagnetic wave propagating
along the zaxis. In this system, the electric component of the wave has no effect.
The Hamiltonian is written
H = −µB = H
0
+H
1
, (6.171)
where
H
0
=
e B
0
m
e
S
z
, (6.172)
and
H
1
=
e B
1
m
e
(cos ωt S
x
+ sinωt S
y
) . (6.173)
The eigenstates of the unperturbed Hamiltonian are the ‘spin up’ and ‘spin
down’ states, denoted +) and −), respectively. Thus,
H
0
±) = ±
e ¯hB
0
2 m
e
±). (6.174)
The timedependent Hamiltonian can be written
H
1
=
e B
1
2 m
e
_
exp( i ωt) S
−
+ exp(−i ωt) S
+
_
, (6.175)
where S
+
and S
−
are the conventional raising and lowering operators for the spin
angular momentum. It follows that
¸+H
1
+) = ¸−H
1
−) = 0, (6.176)
and
¸−H
1
+) = ¸+H
1
−)
∗
=
e ¯hB
1
2 m
e
exp( i ωt). (6.177)
149
6.12 The Dyson series 6 APPROXIMATION METHODS
It can be seen that this system is exactly the same as the twostate system
discussed in the previous section, provided that we make the identiﬁcations
1) → −), (6.178)
2) → +), (6.179)
ω
21
→
e B
0
m
e
, (6.180)
γ →
e ¯hB
1
2 m
e
. (6.181)
The resonant frequency, ω
21
, is simply the spin precession frequency for an elec
tron in a uniform magnetic ﬁeld of strength B
0
. In the absence of the perturba
tion, the expectation values of S
x
and S
y
oscillate because of the spin precession,
but the expectation value of S
z
remains invariant. If we now apply a magnetic
perturbation rotating at the resonant frequency then, according to the analysis of
the previous section, the system undergoes a succession of spinﬂops, +)
÷
÷−),
in addition to the spin precession. We also know that if the oscillation frequency
of the applied ﬁeld is very different from the resonant frequency then there is
virtually zero probability of the ﬁeld triggering a spinﬂop. The width of the
resonance (in frequency) is determined by the strength of the oscillating mag
netic perturbation. Experimentalist are able to measure the magnetic moments
of electrons, and other spin onehalf particles, to a high degree of accuracy by
placing the particles in a magnetic ﬁeld, and subjecting them to an oscillating
magnetic ﬁeld whose frequency is gradually scanned. By determining the reso
nant frequency (i.e., the frequency at which the particles absorb energy from the
oscillating ﬁeld), it is possible to calculate the magnetic moment.
6.12 The Dyson series
Let us now try to ﬁnd approximate solutions of Eq. (6.154) for a general system.
It is convenient to work in terms of the time evolution operator, U(t
0
, t), which is
deﬁned
A, t
0
, t) = U(t
0
, t) A). (6.182)
150
6.12 The Dyson series 6 APPROXIMATION METHODS
Here, A, t
0
, t) is the state ket of the system at time t, given that the state ket at
the initial time t
0
is A). It is easily seen that the time evolution operator satisﬁes
the differential equation
i ¯h
∂U(t
0
, t)
∂t
= (H
0
+H
1
) U(t
0
, t), (6.183)
subject to the boundary condition
U(t
0
, t
0
) = 1. (6.184)
In the absence of the external perturbation, the time evolution operator re
duces to
U(t
0
, t) = exp[−i H
0
(t −t
0
)/¯h]. (6.185)
Let us switch on the perturbation and look for a solution of the form
U(t
0
, t) = exp[−i H
0
(t −t
0
)/¯h] U
I
(t
0
, t). (6.186)
It is readily demonstrated that U
I
satisﬁes the differential equation
i ¯h
∂U
I
(t
0
, t)
∂t
= H
I
(t
0
, t) U
I
(t
0
, t), (6.187)
where
H
I
(t
0
, t) = exp[+i H
0
(t −t
0
)/¯h] H
1
exp[−i H
0
(t −t
0
)/¯h], (6.188)
subject to the boundary condition
U
I
(t
0
, t
0
) = 1. (6.189)
Note that U
I
speciﬁes that component of the time evolution operator which is due
to the timedependent perturbation. Thus, we would expect U
I
to contain all of
the information regarding transitions between different eigenstates of H
0
caused
by the perturbation.
Suppose that the system starts off at time t
0
in the eigenstate i) of the un
perturbed Hamiltonian. The subsequent evolution of the state ket is given by
Eq. (6.149),
i, t
0
, t) =
m
c
m
(t) exp[−i E
m
(t −t
0
)/¯h] m). (6.190)
151
6.12 The Dyson series 6 APPROXIMATION METHODS
However, we also have
i, t
0
, t) = exp[−i H
0
(t −t
0
)/¯h] U
I
(t
0
, t) i). (6.191)
It follows that
c
n
(t) = ¸nU
I
(t
0
, t)i), (6.192)
where use has been made of ¸nm) = δ
nm
. Thus, the probability that the system
is found in state n) at time t, given that it is deﬁnitely in state i) at time t
0
, is
simply
P
i→n
(t
0
, t) = ¸nU
I
(t
0
, t)i)
2
. (6.193)
This quantity is usually termed the transition probability between states i) and
n).
Note that the differential equation (6.187), plus the boundary condition (6.189),
are equivalent to the following integral equation,
U
I
(t
0
, t) = 1 −
i
¯h
_
t
t
0
H
I
(t
0
, t
) U
I
(t
0
, t
) dt
. (6.194)
We can obtain an approximate solution to this equation by iteration:
U
I
(t
0
, t) · 1 −
i
¯h
_
t
t
0
H
I
(t
0
, t
)
_
_
1 −
i
¯h
_
t
t
0
H
I
(t
0
, t
) U
I
(t
0
, t
)
_
_
dt
· 1 −
i
¯h
_
t
t
0
H
I
(t
0
, t
) dt
+
_
−i
¯h
_
2
_
t
t
0
dt
_
t
t
0
H
I
(t
0
, t
) H
I
(t
0
, t
) dt
+ . (6.195)
This expansion is known as the Dyson series. Let
c
n
= c
(0)
n
+c
(1)
n
+c
(2)
n
+ , (6.196)
where the superscript
(1)
refers to a ﬁrstorder term in the expansion, etc. It
follows from Eqs. (6.192) and (6.195) that
c
(0)
n
(t) = δ
in
, (6.197)
152
6.12 The Dyson series 6 APPROXIMATION METHODS
c
(1)
n
(t) = −
i
¯h
_
t
t
0
¸nH
I
(t
0
, t
)i) dt
, (6.198)
c
(2)
n
(t) =
_
−i
¯h
_
2
_
t
t
0
dt
_
t
t
0
¸nH
I
(t
0
, t
) H
I
(t
0
, t
)i) dt
. (6.199)
These expressions simplify to
c
(0)
n
(t) = δ
in
, (6.200)
c
(1)
n
(t) = −
i
¯h
_
t
t
0
exp[ i ω
ni
(t
−t
0
)] H
ni
(t
) dt
, (6.201)
c
(2)
n
(t) =
_
−i
¯h
_
2
m
_
t
t
0
dt
_
t
t
0
dt
exp[ i ω
nm
(t
−t
0
)]
H
nm
(t
) exp[ i ω
mi
(t
−t
0
)] H
mi
(t
), (6.202)
where
ω
nm
=
E
n
−E
m
¯h
, (6.203)
and
H
nm
(t) = ¸nH
1
(t)m). (6.204)
The transition probability between states i and n is simply
P
i→n
(t
0
, t) = c
(0)
n
+c
(1)
n
+c
(2)
n
+ 
2
. (6.205)
According to the above analysis, there is no chance of a transition between
states i) and n) (i ,= n) to zerothorder (i.e., in the absence of the perturbation).
To ﬁrstorder, the transition probability is proportional to the time integral of the
matrix element ¸nH
1
i), weighted by some oscillatory phasefactor. Thus, if the
matrix element is zero, then there is no chance of a ﬁrstorder transition between
states i) and n). However, to secondorder, a transition between states i) and
n) is possible even when the matrix element ¸nH
1
i) is zero.
153
6.13 Constant perturbations 6 APPROXIMATION METHODS
6.13 Constant perturbations
Consider a constant perturbation which is suddenly switched on at time t = 0:
H
1
(t) = 0 for t < 0
H
1
(t) = H
1
for t ≥ 0, (6.206)
where H
1
is timeindependent, but is generally a function of the position, mo
mentum, and spin operators. Suppose that the system is deﬁnitely in state i) at
time t = 0. According to Eqs. (6.200)–(6.202) (with t
0
= 0),
c
(0)
n
(t) = δ
in
, (6.207)
c
(1)
n
(t) = −
i
¯h
H
ni
_
t
0
exp[ i ω
ni
(t
−t)] dt
=
H
ni
E
n
−E
i
[1 − exp( i ω
ni
t)], (6.208)
giving
P
i→n
(t) · c
(1)
n

2
=
4 H
ni

2
E
n
−E
i

2
sin
2
_
_
(E
n
−E
i
) t
2 ¯h
_
_
, (6.209)
for i ,= n. The transition probability between states i) and n) can be written
P
i→n
(t) =
H
ni

2
t
2
¯h
2
sinc
2
_
_
(E
n
−E
i
) t
2 ¯h
_
_
, (6.210)
where
sinc(x) =
sinx
x
. (6.211)
The sinc function is highly oscillatory, and decays like 1/x at large x. It is a
good approximation to say that sinc(x) is small except when x
<
∼
π. It follows
that the transition probability, P
i→n
, is small except when
E
n
−E
i

<
∼
2π¯h
t
. (6.212)
Note that in the limit t → ∞ only those transitions which conserve energy (i.e.,
E
n
= E
i
) have an appreciable probability of occurrence. At ﬁnite t, is is possible
to have transitions which do not exactly conserve energy, provided that
∆E∆t
<
∼
¯h, (6.213)
154
6.13 Constant perturbations 6 APPROXIMATION METHODS
where ∆E = E
n
− E
i
 is change in energy of the system associated with the
transition, and ∆t = t is the time elapsed since the perturbation was switched
on. Clearly, this result is just a manifestation of the wellknown uncertainty re
lation for energy and time. This uncertainty relation is fundamentally different
to the positionmomentum uncertainty relation, since in nonrelativistic quan
tum mechanics position and momentum are operators, whereas time is merely a
parameter.
The probability of a transition which conserves energy (i.e., E
n
= E
i
) is
P
i→n
(t) =
H
in

2
t
2
¯h
2
, (6.214)
where use has been made of sinc(0) = 1. Note that this probability grows quadrat
ically with time. This result is somewhat surprising, since it implies that the prob
ability of a transition occurring in a ﬁxed time interval, t to t +dt, grows linearly
with t, despite the fact that H
1
is constant for t > 0. In practice, there is usually
a group of ﬁnal states, all possessing nearly the same energy as the energy of the
initial state i). It is helpful to deﬁne the density of states, ρ(E), where the num
ber of ﬁnal states lying in the energy range E to E+dE is given by ρ(E) dE. Thus,
the probability of a transition from the initial state i to any of the continuum of
possible ﬁnal states is
P
i→
(t) =
_
P
i→n
(t) ρ(E
n
) dE
n
, (6.215)
giving
P
i→
(t) =
2 t
¯h
_
H
ni

2
ρ(E
n
) sinc
2
(x) dx, (6.216)
where
x = (E
n
−E
i
) t/2 ¯h, (6.217)
and use has been made of Eq. (6.210). We know that in the limit t → ∞ the
function sinc(x) is only nonzero in an inﬁnitesimally narrow range of ﬁnal ener
gies centred on E
n
= E
i
. It follows that, in this limit, we can take ρ(E
n
) and H
ni

2
out of the integral in the above formula to obtain
P
i→[n]
(t) =
2π
¯h
H
ni

2
ρ(E
n
) t
¸
¸
¸
¸
¸
E
n
E
i
, (6.218)
155
6.13 Constant perturbations 6 APPROXIMATION METHODS
where P
i→[n]
denotes the transition probability between the initial state i) and
all ﬁnal states n) which have approximately the same energy as the initial state.
Here, H
ni

2
is the average of H
ni

2
over all ﬁnal states with approximately the
same energy as the initial state. In deriving the above formula, we have made
use of the result
_
∞
−∞
sinc
2
(x) dx = π. (6.219)
Note that the transition probability, P
i→[n]
, is now proportional to t, instead of t
2
.
It is convenient to deﬁne the transition rate, which is simply the transition
probability per unit time. Thus,
w
i→[n]
=
dP
i→[n]
dt
, (6.220)
giving
w
i→[n]
=
2π
¯h
H
ni

2
ρ(E
n
)
¸
¸
¸
¸
¸
E
n
E
i
. (6.221)
This appealingly simple result is known as Fermi’s golden rule. Note that the
transition rate is constant in time (for t > 0): i.e., the probability of a transition
occurring in the time interval t to t + dt is independent of t for ﬁxed dt. Fermi’s
golden rule is sometimes written
w
i→n
=
2π
¯h
H
ni

2
δ(E
n
−E), (6.222)
where it is understood that this formula must be integrated with
_
ρ(E
n
) dE
n
to
obtain the actual transition rate.
Let us now calculate the secondorder term in the Dyson series, using the
constant perturbation (6.206). From Eq. (6.202) we ﬁnd that
c
(2)
n
(t) =
_
−i
¯h
_
2
m
H
nm
H
mi
_
t
0
dt
exp( i ω
nm
t
)
_
t
0
dt
exp( i ω
mi
t )
=
i
¯h
m
H
nm
H
mi
E
m
−E
i
_
t
0
[exp( i ω
ni
t
) − exp( i ω
nm
t
] ) dt
=
i t
¯h
m
H
nm
H
mi
E
m
−E
i
[exp( i ω
ni
t/2) sinc(ω
ni
t/2)
156
6.13 Constant perturbations 6 APPROXIMATION METHODS
−exp( i ω
nm
t/2) sinc(ω
nm
t/2)] . (6.223)
Thus,
c
n
(t) = c
(1)
n
+c
(2)
n
=
i t
¯h
exp( i ω
ni
t/2)
_
_
_
_
H
ni
+
m
H
nm
H
mi
E
m
−E
i
_
_
sinc(ω
ni
t/2)
−
m
H
nm
H
mi
E
m
−E
i
exp( i ω
im
t/2) sinc(ω
nm
t/2)
_
_
, (6.224)
where use has been made of Eq. (6.208). It follows, by analogy with the previous
analysis, that
w
i→[n]
=
2π
¯h
¸
¸
¸
¸
¸
¸
H
ni
+
m
H
nm
H
mi
E
m
−E
i
¸
¸
¸
¸
¸
¸
2
ρ(E
n
)
¸
¸
¸
¸
¸
¸
¸
E
n
E
i
, (6.225)
where the transition rate is calculated for all ﬁnal states, n), with approximately
the same energy as the initial state, i), and for intermediate states, m) whose
energies differ from that of the initial state. The fact that E
m
,= E
i
causes the
last term on the righthand side of Eq. (6.224) to average to zero (due to the
oscillatory phasefactor) during the evaluation of the transition probability.
According to Eq. (6.225), a secondorder transition takes place in two steps.
First, the system makes a nonenergyconserving transition to some intermedi
ate state m). Subsequently, the system makes another nonenergyconserving
transition to the ﬁnal state n). The net transition, from i) to n), conserves en
ergy. The nonenergyconserving transitions are generally termed virtual transi
tions, whereas the energy conserving ﬁrstorder transition is termed a real transi
tion. The above formula clearly breaks down if H
nm
H
mi
,= 0 when E
m
= E
i
.
This problem can be avoided by gradually turning on the perturbation: i.e.,
H
1
→ exp(ηt) H
1
(where η is very small). The net result is to change the en
ergy denominator in Eq. (6.225) from E
i
−E
m
to E
i
−E
m
+ i ¯hη.
157
6.14 Harmonic perturbations 6 APPROXIMATION METHODS
6.14 Harmonic perturbations
Consider a perturbation which oscillates sinusoidally in time. This is usually
called a harmonic perturbation. Thus,
H
1
(t) = V exp( i ωt) +V
†
exp(−i ωt), (6.226)
where V is, in general, a function of position, momentum, and spin operators.
Let us initiate the system in the eigenstate i) of the unperturbed Hamiltonian,
H
0
, and switch on the harmonic perturbation at t = 0. It follows from Eq. (6.201)
that
c
(1)
n
=
−i
¯h
_
t
0
_
V
ni
exp(i ωt
) +V
†
ni
exp(−i ωt
)
_
exp( i ω
ni
t
) dt
, (6.227)
=
1
¯h
_
_
1 − exp[ i (ω
ni
+ω) t]
ω
ni
+ω
V
ni
+
1 − exp[ i (ω
ni
−ω) t]
ω
ni
−ω
V
†
ni
_
_
,
where
V
ni
= ¸nVi), (6.228)
V
†
ni
= ¸nV
†
i) = ¸iVn)
∗
. (6.229)
This formula is analogous to Eq. (6.208), provided that
ω
ni
=
E
n
−E
i
¯h
→ω
ni
±ω. (6.230)
Thus, it follows fromthe previous analysis that the transition probability P
i→n
(t) =
c
(1)
n

2
is only appreciable in the limit t →∞if
ω
ni
+ω · 0 or E
n
· E
i
− ¯hω, (6.231)
ω
ni
−ω · 0 or E
n
· E
i
+ ¯hω. (6.232)
Clearly, (6.231) corresponds to the ﬁrst termon the righthand side of Eq. (6.227),
and (6.232) corresponds to the second term. The former term describes a process
by which the system gives up energy ¯hω to the perturbing ﬁeld, whilst making a
transition to a ﬁnal state whose energy level is less than that of the initial state
158
6.15 Absorption and stimulated emission of radiation 6 APPROXIMATION METHODS
by ¯hω. This process is known as stimulated emission. The latter term describes
a process by which the system gains energy ¯hω from the perturbing ﬁeld, whilst
making a transition to a ﬁnal state whose energy level exceeds that of the initial
state by ¯hω. This process is known as absorption. In both cases, the total energy
(i.e., that of the system plus the perturbing ﬁeld) is conserved.
By analogy with Eq. (6.221),
w
i→[n]
=
2π
¯h
V
ni

2
ρ(E
n
)
¸
¸
¸
¸
¸
E
n
=E
i
−¯hω
, (6.233)
w
i→[n]
=
2π
¯h
V
†
ni

2
ρ(E
n
)
¸
¸
¸
¸
¸
E
n
=E
i
+¯hω
. (6.234)
Equation (6.233) speciﬁes the transition rate for stimulated emission, whereas
Eq. (6.234) gives the transition rate for absorption. These equations are more
usually written
w
i→n
=
2π
¯h
V
ni

2
δ(E
n
−E
i
+ ¯hω), (6.235)
w
i→n
=
2π
¯h
V
†
ni

2
δ(E
n
−E
i
− ¯hω). (6.236)
It is clear from Eqs. (6.228)(6.229) that V
†
ni

2
= V
ni

2
. It follows from
Eqs. (6.233)–(6.234) that
w
i→[n]
ρ(E
n
)
=
w
n→[i]
ρ(E
i
)
. (6.237)
In other words, the rate of stimulated emission, divided by the density of ﬁnal
states for stimulated emission, equals the rate of absorption, divided by the den
sity of ﬁnal states for absorption. This result, which expresses a fundamental
symmetry between absorption and stimulated emission, is known as detailed bal
ancing, and is very important in statistical mechanics.
6.15 Absorption and stimulated emission of radiation
Let us use some of the results of timedependent perturbation theory to inves
tigate the interaction of an atomic electron with classical (i.e., nonquantized)
159
6.15 Absorption and stimulated emission of radiation 6 APPROXIMATION METHODS
electromagnetic radiation.
The unperturbed Hamiltonian is
H
0
=
p
2
2 m
e
+V
0
(r). (6.238)
The standard classical prescription for obtaining the Hamiltonian of a particle of
charge q in the presence of an electromagnetic ﬁeld is
p → p +qA, (6.239)
H → H−qφ, (6.240)
where A(r) is the vector potential and φ(r) is the scalar potential. Note that
E = −∇φ −
∂A
∂t
, (6.241)
B = ∇A. (6.242)
This prescription also works in quantum mechanics. Thus, the Hamiltonian of an
atomic electron placed in an electromagnetic ﬁeld is
H =
(p −e A)
2
2 m
e
+e φ +V
0
(r), (6.243)
where A and φ are functions of the position operators. The above equation can
be written
H =
_
p
2
−e Ap −e pA +e
2
A
2
_
2 m
e
+e φ +V
0
(r). (6.244)
Now,
pA = Ap, (6.245)
provided that we adopt the gauge ∇A = 0. Hence,
H =
p
2
2 m
e
−
e Ap
m
e
+
e
2
A
2
2 m
e
+e φ +V
0
(r). (6.246)
Suppose that the perturbation corresponds to a monochromatic planewave,
for which
φ = 0, (6.247)
A = 2 A
0
cos
_
ω
c
nr −ωt
_
, (6.248)
160
6.15 Absorption and stimulated emission of radiation 6 APPROXIMATION METHODS
where and n are unit vectors which specify the direction of polarization and
the direction of propagation, respectively. Note that n = 0. The Hamiltonian
becomes
H = H
0
+H
1
(t), (6.249)
with
H
0
=
p
2
2 m
e
+V(r), (6.250)
and
H
1
· −
e Ap
m
e
, (6.251)
where the A
2
term, which is second order in A
0
, has been neglected.
The perturbing Hamiltonian can be written
H
1
= −
e A
0
p
m
e
(exp[ i (ω/c) nr − i ωt] + exp[−i (ω/c) nr + i ωt]) . (6.252)
This has the same form as Eq. (6.226), provided that
V = −
e A
0
p
m
e
exp[−i (ω/c) nr ] (6.253)
It is clear, by analogy with the previous analysis, that the ﬁrst term on the right
hand side of Eq. (6.252) describes the absorption of a photon of energy ¯hω,
whereas the second term describes the stimulated emission of a photon of energy
¯hω. It follows from Eq. (6.236) that the rate of absorption is
w
i→n
=
2π
¯h
e
2
m
2
e
A
0

2
¸n exp[ i (ω/c) nr] pi)
2
δ(E
n
−E
i
− ¯hω). (6.254)
The absorption crosssection is deﬁned as the ratio of the power absorbed by
the atom to the incident power per unit area in the electromagnetic ﬁeld. Now
the energy density of an electromagnetic ﬁeld is
U =
1
2
_
_
0
E
2
0
2
+
B
2
0
2 µ
0
_
_
, (6.255)
161
6.16 The electric dipole approximation 6 APPROXIMATION METHODS
where E
0
and B
0
= E
0
/c = 2 A
0
ω/c are the peak electric and magnetic ﬁeld
strengths, respectively. The incident power per unit area of the electromagnetic
ﬁeld is
c U = 2
0
c ω
2
A
0

2
. (6.256)
Now,
σ
abs
=
¯hωw
i→n
c U
, (6.257)
so
σ
abs
=
πe
2
0
m
2
e
ωc
¸n exp[ i (ω/c) nr] pi)
2
δ(E
n
−E
i
− ¯hω). (6.258)
6.16 The electric dipole approximation
In general, the wavelength of the type of electromagnetic radiation which in
duces, or is emitted during, transitions between different atomic energy levels is
much larger than the typical size of a light atom. Thus,
exp[ i (ω/c) nr] = 1 + i
ω
c
nr + , (6.259)
can be approximated by its ﬁrst term, unity (remember that ω/c = 2π/λ). This
approximation is known as the electric dipole approximation. It follows that
¸n exp[ i (ω/c) nr] pi) · ¸npi). (6.260)
It is readily demonstrated that
[r, H
0
] =
i ¯hp
m
e
, (6.261)
so
¸npi) = −i
m
e
¯h
¸n[r, H
0
]i) = i m
e
ω
ni
¸nri). (6.262)
Using Eq. (6.258), we obtain
σ
abs
= 4π
2
αω
ni
¸nri)
2
δ(ω−ω
ni
), (6.263)
162
6.16 The electric dipole approximation 6 APPROXIMATION METHODS
where α = e
2
/(2
0
hc) = 1/137 is the ﬁne structure constant. It is clear that if
the absorption crosssection is regarded as a function of the applied frequency,
ω, then it exhibits a sharp maximum at ω = ω
ni
= (E
n
−E
i
)/¯h.
Suppose that the radiation is polarized in the zdirection, so that = ^z. We
have already seen, from Sect. 6.4, that ¸nzi) = 0 unless the initial and ﬁnal
states satisfy
∆l = ±1, (6.264)
∆m = 0. (6.265)
Here, l is the quantum number describing the total orbital angular momentum
of the electron, and m is the quantum number describing the projection of the
orbital angular momentum along the zaxis. It is easily demonstrated that ¸nxi)
and ¸nyi) are only nonzero if
∆l = ±1, (6.266)
∆m = ±1. (6.267)
Thus, for generally directed radiation ¸nri) is only nonzero if
∆l = ±1, (6.268)
∆m = 0, ±1. (6.269)
These are termed the selection rules for electric dipole transitions. It is clear, for
instance, that the electric dipole approximation allows a transition from a 2p
state to a 1s state, but disallows a transition from a 2s to a 1s state. The latter
transition is called a forbidden transition.
Forbidden transitions are not strictly forbidden. Instead, they take place at a
far lower rate than transitions which are allowed according to the electric dipole
approximation. After electric dipole transitions, the next most likely type of tran
sition is a magnetic dipole transition, which is due to the interaction between the
electron spin and the oscillating magnetic ﬁeld of the incident electromagnetic
radiation. Magnetic dipole transitions are typically about 10
5
times more unlikely
than similar electric dipole transitions. The ﬁrstorder term in Eq. (6.259) yields
163
6.16 The electric dipole approximation 6 APPROXIMATION METHODS
socalled electric quadrupole transitions. These are typically about 10
8
times more
unlikely than electric dipole transitions. Magnetic dipole and electric quadrupole
transitions satisfy different selection rules than electric dipole transitions: for
instance, the selection rules for electric quadrupole transitions are ∆l = 0, ±2.
Thus, transitions which are forbidden as electric dipole transitions may well be
allowed as magnetic dipole or electric quadrupole transitions.
Integrating Eq. (6.263) over all possible frequencies of the incident radiation
yields
_
σ
abs
(ω) dω =
n
4π
2
αω
ni
¸nri)
2
. (6.270)
Suppose, for the sake of deﬁniteness, that the incident radiation is polarized in
the xdirection. It is easily demonstrated that
[x, [x, H
0
] ] = −
¯h
2
m
e
. (6.271)
Thus,
¸i[x, [x, H
0
] ]i) = ¸ix
2
H
0
+H
0
x
2
−2 x H
0
xi) = −
¯h
2
m
e
, (6.272)
giving
2
n
(¸ixn)E
i
¸nxi) −¸ixn)E
n
¸nxi)) = −
¯h
2
m
e
. (6.273)
It follows that
2 m
e
¯h
n
ω
ni
¸nxi)
2
= 1. (6.274)
This is known as the ThomasReicheKuhn sum rule. According to this rule,
Eq. (6.270) reduces to
_
σ
abs
(ω) dω =
2π
2
α¯h
m
e
=
πe
2
2
0
m
e
c
. (6.275)
Note that ¯h has dropped out of the ﬁnal result. In fact, the above formula is ex
actly the same as that obtained classically by treating the electron as an oscillator.
164
6.17 Energyshifts and decaywidths 6 APPROXIMATION METHODS
6.17 Energyshifts and decaywidths
We have examined how a state n), other than the initial state i), becomes popu
lated as a result of some timedependent perturbation applied to the system. Let
us now consider how the initial state becomes depopulated.
It is convenient to gradually turn on the perturbation from zero at t = −∞.
Thus,
H
1
(t) = exp(ηt) H
1
, (6.276)
where η is small and positive, and H
1
is a constant.
In the remote past, t →−∞, the system is assumed to be in the initial state i).
Thus, c
i
(t → −∞) = 1, and c
n=i
(t → −∞) = 0. Basically, we want to calculate
the time evolution of the coefﬁcient c
i
(t). First, however, let us check that our
previous Fermi golden rule result still applies when the perturbing potential is
turned on slowly, instead of very suddenly. For c
n=i
(t) we have fromEqs. (6.200)–
(6.201) that
c
(0)
n
(t) = 0, (6.277)
c
(1)
n
(t) = −
i
¯h
H
ni
_
t
−∞
exp[ (η + i ω
ni
)t
] dt
= −
i
¯h
H
ni
exp[ (η + i ω
ni
)t ]
η + i ω
ni
, (6.278)
where H
ni
= ¸nH
1
i). It follows that, to ﬁrstorder, the transition probability
from state i) to state n) is
P
i→n
(t) = c
(1)
n

2
=
H
ni

2
¯h
2
exp(2 ηt)
η
2
+ω
2
ni
. (6.279)
The transition rate is given by
w
i→n
(t) =
dP
i→n
dt
=
2 H
ni

2
¯h
2
ηexp(2 ηt)
η
2
+ω
2
ni
. (6.280)
Consider the limit η →0. In this limit, exp(ηt) →1, but
lim
η→0
η
η
2
+ω
2
ni
= πδ(ω
ni
) = π¯hδ(E
n
−E
i
). (6.281)
165
6.17 Energyshifts and decaywidths 6 APPROXIMATION METHODS
Thus, Eq. (6.280) yields the standard Fermi golden rule result
w
i→n
=
2π
¯h
H
ni

2
δ(E
n
−E
i
). (6.282)
It is clear that the deltafunction in the above formula actually represents a func
tion which is highly peaked at some particular energy. The width of the peak is
determined by how fast the perturbation is switched on.
Let us now calculate c
i
(t) using Eqs. (6.200)–(6.202). We have
c
(0)
i
(t) = 1, (6.283)
c
(1)
i
(t) = −
i
¯h
H
ii
_
t
−∞
exp(ηt
) dt
= −
i
¯h
H
ii
exp(ηt)
η
, (6.284)
c
(2)
i
(t) =
_
−i
¯h
_
2
m
H
mi

2
_
t
−∞
dt
_
t
−∞
dt
exp[ (η + i ω
im
)t
] exp[ (η + i ω
mi
)t
],
=
_
−i
¯h
_
2
m
H
mi

2
exp(2 ηt)
2 η(η + i ω
mi
)
. (6.285)
Thus, to secondorder we have
c
i
(t) · 1 +
_
−i
¯h
_
H
ii
exp(ηt)
η
+
_
−i
¯h
_
2
H
ii

2
exp(2 ηt)
2 η
2
+
_
−i
¯h
_
m=i
H
mi

2
exp(2 ηt)
2 η(E
i
−E
m
+ i ¯hη)
. (6.286)
Let us now consider the ratio ˙ c
i
/c
i
, where ˙ c
i
≡ dc
i
/dt. Using Eq. (6.286), we
can evaluate this ratio in the limit η →0. We obtain
˙ c
i
c
i
·
_
¸
_
_
−i
¯h
_
H
ii
+
_
−i
¯h
_
2
H
ii

2
η
+
_
−i
¯h
_
m=i
H
mi

2
E
i
−E
m
+ i ¯hη
_
¸
_
__
1 −
i
¯h
H
ii
η
_
·
_
−i
¯h
_
H
ii
+ lim
η→0
_
−i
¯h
_
m=i
H
mi

2
E
i
−E
m
+ i ¯hη
. (6.287)
166
6.17 Energyshifts and decaywidths 6 APPROXIMATION METHODS
This result is formally correct to secondorder in perturbed quantities. Note that
the righthand side of Eq. (6.287) is independent of time. We can write
˙ c
i
c
i
=
_
−i
¯h
_
∆
i
, (6.288)
where
∆
i
= H
ii
+ lim
η→0
m=i
H
mi

2
E
i
−E
m
+ i ¯hη
(6.289)
is a constant. According to a wellknown result in pure mathematics,
lim
→0
1
x + i
= P
1
x
− i πδ(x), (6.290)
where > 0, and P denotes the principle part. It follows that
∆
i
= H
ii
+P
m=i
H
mi

2
E
i
−E
m
− i π
m=i
H
mi

2
δ(E
i
−E
m
). (6.291)
It is convenient to normalize the solution of Eq. (6.288) so that c
i
(0) = 1.
Thus, we obtain
c
i
(t) = exp
_
−i ∆
i
t
¯h
_
. (6.292)
According to Eq. (6.149), the time evolution of the initial state ket i) is given by
i, t) = exp[−i (∆
i
+E
i
) t/¯h] i). (6.293)
We can rewrite this result as
i, t) = exp(−i [E
i
+ Re(∆
i
) ] t/¯h) exp[ Im(∆
i
) t/¯h] i). (6.294)
It is clear that the real part of ∆
i
gives rise to a simple shift in energy of state i),
whereas the imaginary part of ∆
i
governs the growth or decay of this state. Thus,
i, t) = exp[−i (E
i
+∆E
i
) t/¯h] exp(−Γ
i
t/2¯h) i), (6.295)
where
∆E
i
= Re(∆
i
) = H
ii
+P
m=i
H
mi

2
E
i
−E
m
, (6.296)
167
6.17 Energyshifts and decaywidths 6 APPROXIMATION METHODS
and
Γ
i
¯h
= −
2 Im(∆
i
)
¯h
=
2π
¯h
m=i
H
mi

2
δ(E
i
−E
m
). (6.297)
Note that the energyshift ∆E
i
is the same as that predicted by standard time
independent perturbation theory.
The probability of observing the system in state i) at time t > 0, given that it
is deﬁnately in state i) at time t = 0, is given by
P
i→i
(t) = c
i

2
= exp(−Γ
i
t/¯h), (6.298)
where
Γ
i
¯h
=
m=i
w
i→m
. (6.299)
Here, use has been made of Eq. (6.222). Clearly, the rate of decay of the initial
state is a simple function of the transition rates to the other states. Note that the
system conserves probability up to secondorder in perturbed quantities, since
c
i

2
+
m=i
c
m

2
· (1 −Γ
i
t/¯h) +
m=i
w
i→m
t = 1. (6.300)
The quantity ∆
i
is called the decaywidth of state i). It is closely related to the
mean lifetime of this state,
τ
i
=
¯h
Γ
i
, (6.301)
where
P
i→i
= exp(−t/τ
i
). (6.302)
According to Eq. (6.294), the amplitude of state i) both oscillates and decays as
time progresses. Clearly, state i) is not a stationary state in the presence of the
timedependent perturbation. However, we can still represent it as a superposi
tion of stationary states (whose amplitudes simply oscillate in time). Thus,
exp[−i (E
i
+∆E
i
) t/¯h] exp(−Γ
i
t/2¯h) =
_
f(E) exp(−i Et/¯h) dE, (6.303)
168
6.17 Energyshifts and decaywidths 6 APPROXIMATION METHODS
where f(E) is the weight of the stationary state with energy E in the superposition.
The Fourier inversion theorem yields
f(E)
2
∝
1
(E − [E
i
+ Re(∆
i
)])
2
+Γ
2
i
/4
. (6.304)
In the absence of the perturbation, f(E)
2
is basically a deltafunction centred on
the unperturbed energy E
i
of state i). In other words, state i) is a stationary state
whose energy is completely determined. In the presence of the perturbation, the
energy of state i) is shifted by Re(∆
i
). The fact that the state is no longer station
ary (i.e., it decays in time) implies that its energy cannot be exactly determined.
Indeed, the energy of the state is smeared over some region of width (in energy)
Γ
i
centred around the shifted energy E
i
+Re(∆
i
). The faster the decay of the state
(i.e., the larger Γ
i
), the more its energy is spread out. This effect is clearly a man
ifestation of the energytime uncertainty relation ∆E∆t ∼ ¯h. One consequence of
this effect is the existence of a natural width of spectral lines associated with the
decay of some excited state to the ground state (or any other lower energy state).
The uncertainty in energy of the excited state, due to its propensity to decay, gives
rise to a slight smearing (in wavelength) of the spectral line associated with the
transition. Strong lines, which correspond to fast transitions, are smeared out
more that weak lines. For this reason, spectroscopists generally favour forbid
den lines for Doppler shift measurements. Such lines are not as bright as those
corresponding to allowed transitions, but they are a lot sharper.
169
7 SCATTERING THEORY
7 Scattering theory
7.1 Introduction
Historically, data regarding quantum phenomena has been obtained from two
main sources—the study of spectroscopic lines, and scattering experiments. We
have already developed theories which account for some aspects of the spectra of
hydrogenlike atoms. Let us now examine the quantum theory of scattering.
7.2 The LipmannSchwinger equation
Consider timeindependent scattering theory, for which the Hamiltonian of the
system is written
H = H
0
+H
1
, (7.1)
where H
0
is the Hamiltonian of a free particle of mass m,
H
0
=
p
2
2 m
, (7.2)
and H
1
represents the nontimevarying source of the scattering. Let φ) be an
energy eigenket of H
0
,
H
0
φ) = Eφ), (7.3)
whose wavefunction ¸r
φ) is φ(r
). This state is a planewave state or, possibly,
a sphericalwave state. Schr¨ odinger’s equation for the scattering problem is
(H
0
+H
1
)ψ) = Eψ), (7.4)
where ψ) is an energy eigenstate of the total Hamiltonian whose wavefunction
¸r
ψ) is ψ(r
). In general, both H
0
and H
0
+ H
1
have continuous energy spectra:
i.e., their energy eigenstates are unbound. We require a solution of Eq. (7.4)
which satisﬁes the boundary condition ψ) → φ) as H
1
→ 0. Here, φ) is a
solution of the free particle Schr¨ odinger equation, (7.3), corresponding to the
same energy eigenvalue.
170
7.2 The LipmannSchwinger equation 7 SCATTERING THEORY
Formally, the desired solution can be written
ψ) = φ) +
1
E −H
0
H
1
ψ). (7.5)
Note that we can recover Eq. (7.4) by operating on the above equation with
E − H
0
, and making use of Eq. (7.3). Furthermore, the solution satisﬁes the
boundary condition ψ) →φ) as H
1
→0. Unfortunately, the operator (E −H
0
)
−1
is singular: i.e., it produces inﬁnities when it operates on an eigenstate of H
0
corresponding to the eigenvalue E. We need a prescription for dealing with these
inﬁnities, otherwise the above solution is useless. The standard prescription is to
make the energy eigenvalue E slightly complex. Thus,
ψ
±
) = φ) +
1
E −H
0
±i
H
1
ψ
±
), (7.6)
where is real, positive, and small. Equation (7.6) is called the LipmannSchwinger
equation, and is nonsingular as long as > 0. The physical signiﬁcance of the ±
signs will become apparent later on.
The LipmannSchwinger equation can be converted into an integral equation
via left multiplication by ¸r. Thus,
ψ
±
(r) = φ(r) +
_
_
r
¸
¸
¸
¸
¸
1
E −H
0
±i
¸
¸
¸
¸
¸
r
_
¸r
H
1
ψ
±
) d
3
r
. (7.7)
Adopting the Schr¨ odinger representation, we can write the scattering problem
(7.4) in the form
(∇
2
+k
2
) ψ(r) =
2 m
¯h
2
¸rH
1
ψ), (7.8)
where
E =
¯h
2
k
2
2 m
. (7.9)
This equation is called Helmholtz’s equation, and can be inverted using standard
Green’s function techniques. Thus,
ψ(r) = φ(r) +
2 m
¯h
2
_
G(r, r
)¸r
H
1
ψ) d
3
r
, (7.10)
171
7.2 The LipmannSchwinger equation 7 SCATTERING THEORY
where
(∇
2
+k
2
) G(r, r
) = δ(r −r
). (7.11)
Note that the solution (7.10) satisﬁes the boundary condition ψ) →φ) as H
1
→
0. As is wellknown, the Green’s function for the Helmholtz problem is given by
G(r, r
) = −
exp(±i k r −r
 )
4πr −r

. (7.12)
Thus, Eq. (7.10) becomes
ψ
±
(r) = φ(r) −
2 m
¯h
2
_
exp(±i k r −r
 )
4πr −r

¸r
H
1
ψ) d
3
r
. (7.13)
A comparison of Eqs. (7.7) and (7.13) suggests that the kernel to Eq. (7.7) takes
the form
_
r
¸
¸
¸
¸
¸
1
E −H
0
±i
¸
¸
¸
¸
¸
r
_
= −
2 m
¯h
2
exp(±i k r −r
 )
4πr −r

. (7.14)
It is not entirely clear that the ± signs correspond on both sides of this equation.
In fact, they do, as is easily proved by a more rigorous derivation of this result.
Let us suppose that the scattering Hamiltonian, H
1
, is only a function of the
position operators. This implies that
¸r
H
1
r) = V(r) δ(r −r
). (7.15)
We can write
¸r
H
1
ψ
±
) =
_
¸r
H
1
r
)¸r
ψ
±
) d
3
r
= V(r
) ψ
±
(r
). (7.16)
Thus, the integral equation (7.13) simpliﬁes to
ψ
±
(r) = φ(r) −
2 m
¯h
2
_
exp(±i k r −r
)
4πr −r

V(r
) ψ
±
(r
) d
3
r
. (7.17)
Suppose that the initial state φ) is a planewave with wavevector k (i.e., a
stream of particles of deﬁnite momentum p = ¯hk). The ket corresponding to
this state is denoted k). The associated wavefunction takes the form
¸rk) =
exp( i kr)
(2π)
3/2
. (7.18)
172
7.2 The LipmannSchwinger equation 7 SCATTERING THEORY
The wavefunction is normalized such that
¸kk
) =
_
¸kr)¸rk
) d
3
r
=
_
exp[−i r(k −k
)]
(2π)
3
d
3
r = δ(k −k
). (7.19)
Suppose that the scattering potential V(r) is only nonzero in some relatively
localized region centred on the origin (r = 0). Let us calculate the wavefunction
ψ(r) a long way from the scattering region. In other words, let us adopt the
ordering r ¸r
. It is easily demonstrated that
r −r
 · r − ^rr
(7.20)
to ﬁrstorder in r
/r, where
^r =
r
r
(7.21)
is a unit vector which points from the scattering region to the observation point.
Let us deﬁne
k
= k^r. (7.22)
Clearly, k
is the wavevector for particles which possess the same energy as the
incoming particles (i.e., k
= k), but propagate from the scattering region to the
observation point. Note that
exp(±i k r −r
 ) · exp(±i k r) exp(∓i k
r
). (7.23)
In the larger limit, Eq. (7.17) reduces to
ψ(r)
±
·
exp( i kr)
(2π)
3/2
−
m
2π¯h
2
exp(±i k r)
r
_
exp(∓i k
r
) V(r
) ψ
±
(r
) d
3
r
. (7.24)
The ﬁrst term on the righthand side is the incident wave. The second term
represents a spherical wave centred on the scattering region. The plus sign (on
ψ
±
) corresponds to a wave propagating away from the scattering region, whereas
the minus sign corresponds to a wave propagating towards the scattering region.
173
7.2 The LipmannSchwinger equation 7 SCATTERING THEORY
It is obvious that the former represents the physical solution. Thus, the wave
function a long way from the scattering region can be written
ψ(r) =
1
(2π)
3/2
_
_
exp( i kr) +
exp( i kr)
r
f(k
, k)
_
_
, (7.25)
where
f(k
, k) = −
(2π)
2
m
¯h
2
_
exp(−i k
r
)
(2π)
3/2
V(r
) ψ(r
) d
3
r
= −
(2π)
2
m
¯h
2
¸k
H
1
ψ). (7.26)
Let us deﬁne the differential crosssection dσ/dΩ as the number of particles
per unit time scattered into an element of solid angle dΩ, divided by the incident
ﬂux of particles. Recall, from Sect. 4, that the probability ﬂux (i.e., the particle
ﬂux) associated with a wavefunction ψ is
j =
¯h
m
Im(ψ
∗
∇ψ). (7.27)
Thus, the probability ﬂux associated with the incident wavefunction,
exp( i kr)
(2π)
3/2
, (7.28)
is
j
inci
=
¯h
(2π)
3
m
k. (7.29)
Likewise, the probability ﬂux associated with the scattered wavefunction,
exp( i k r)
(2π)
3/2
f(k
, k)
r
, (7.30)
is
j
scat
=
¯h
(2π)
3
m
f(k
, k)
2
r
2
k^r. (7.31)
Now,
dσ
dΩ
dΩ =
r
2
dΩj
scat

j
inci

, (7.32)
174
7.3 The Born approximation 7 SCATTERING THEORY
giving
dσ
dΩ
= f(k
, k)
2
. (7.33)
Thus, f(k
, k)
2
gives the differential crosssection for particles with incident mo
mentum ¯hk to be scattered into states whose momentum vectors are directed in
a range of solid angles dΩ about ¯hk
. Note that the scattered particles possess
the same energy as the incoming particles (i.e., k
= k). This is always the case
for scattering Hamiltonians of the form shown in Eq. (7.15).
7.3 The Born approximation
Equation (7.33) is not particularly useful, as it stands, because the quantity
f(k
, k) depends on the unknown ket ψ). Recall that ψ(r) = ¸rψ) is the solu
tion of the integral equation
ψ(r) = φ(r) −
m
2π¯h
2
exp( i k r)
r
_
exp(−i k
r
) V(r
) ψ(r
) d
3
r
, (7.34)
where φ(r) is the wavefunction of the incident state. According to the above
equation the total wavefunction is a superposition of the incident wavefunction
and lots of sphericalwaves emitted from the scattering region. The strength of
the sphericalwave emitted at a given point is proportional to the local value of
the scattering potential, V, as well as the local value of the wavefunction, ψ.
Suppose that the scattering is not particularly strong. In this case, it is reason
able to suppose that the total wavefunction, ψ(r), does not differ substantially
from the incident wavefunction, φ(r). Thus, we can obtain an expression for
f(k
, k) by making the substitution
ψ(r) →φ(r) =
exp( i kr)
(2π)
3/2
. (7.35)
This is called the Born approximation.
The Born approximation yields
f(k
, k) · −
m
2π¯h
2
_
exp[ i (k −k
)r
] V(r
) d
3
r
. (7.36)
175
7.3 The Born approximation 7 SCATTERING THEORY
Thus, f(k
, k) is proportional to the Fourier transform of the scattering potential
V(r) with respect to the wavevector q ≡ k −k
.
For a spherically symmetric potential,
f(k
, k) · −
m
2π¯h
2
___
exp( i qr
cos θ
) V(r
) r
2
dr
sinθ
dθ
dφ
, (7.37)
giving
f(k
, k) · −
2 m
¯h
2
q
_
∞
0
r
V(r
) sin(qr
) dr
. (7.38)
Note that f(k
, k) is just a function of q for a spherically symmetric potential. It is
easily demonstrated that
q ≡ k −k
 = 2 k sin(θ/2), (7.39)
where θ is the angle subtended between the vectors k and k
. In other words, θ
is the angle of scattering. Recall that the vectors k and k
have the same length
by energy conservation.
Consider scattering by a Yukawa potential
V(r) =
V
0
exp(−µr)
µr
, (7.40)
where V
0
is a constant and 1/µ measures the “range” of the potential. It follows
from Eq. (7.38) that
f(θ) = −
2 mV
0
¯h
2
µ
1
q
2
+µ
2
, (7.41)
since
_
∞
0
exp(−µr
) sin(qr
) dr
=
q
µ
2
+q
2
. (7.42)
Thus, in the Born approximation, the differential crosssection for scattering by a
Yukawa potential is
dσ
dΩ
·
_
_
2 mV
0
¯h
2
µ
_
_
2
1
[2 k
2
(1 − cos θ) +µ
2
]
2
, (7.43)
given that
q
2
= 4 k
2
sin
2
(θ/2) = 2 k
2
(1 − cos θ). (7.44)
176
7.3 The Born approximation 7 SCATTERING THEORY
The Yukawa potential reduces to the familiar Coulomb potential as µ → 0,
provided that V
0
/µ → ZZ
e
2
/4π
0
. In this limit the Born differential cross
section becomes
dσ
dΩ
·
_
_
2 mZZ
e
2
4π
0
¯h
2
_
_
2
1
16 k
4
sin
4
(θ/2)
. (7.45)
Recall that ¯hk is equivalent to p, so the above equation can be rewritten
dσ
dΩ
·
_
_
ZZ
e
2
16π
0
E
_
_
2
1
sin
4
(θ/2)
, (7.46)
where E = p
2
/2 m is the kinetic energy of the incident particles. Equation (7.46)
is the classical Rutherford scattering crosssection formula.
The Born approximation is valid provided that ψ(r) is not too different from
φ(r) in the scattering region. It follows, from Eq. (7.17), that the condition for
ψ(r) · φ(r) in the vicinity of r = 0 is
¸
¸
¸
¸
¸
¸
m
2π¯h
2
_
exp( i k r
)
r
V(r
) d
3
r
¸
¸
¸
¸
¸
¸
¸1. (7.47)
Consider the special case of the Yukawa potential. At low energies, (i.e., k ¸ µ)
we can replace exp( i k r
) by unity, giving
2 m
¯h
2
V
0

µ
2
¸1 (7.48)
as the condition for the validity of the Born approximation. The condition for the
Yukawa potential to develop a bound state is
2 m
¯h
2
V
0

µ
2
≥ 2.7, (7.49)
where V
0
is negative. Thus, if the potential is strong enough to form a bound
state then the Born approximation is likely to break down. In the highk limit,
Eq. (7.47) yields
2 m
¯h
2
V
0

µk
¸1. (7.50)
This inequality becomes progressively easier to satisfy as k increases, implying
that the Born approximation is more accurate at high incident particle energies.
177
7.4 Partial waves 7 SCATTERING THEORY
7.4 Partial waves
We can assume, without loss of generality, that the incident wavefunction is
characterized by a wavevector k which is aligned parallel to the zaxis. The
scattered wavefunction is characterized by a wavevector k
which has the same
magnitude as k, but, in general, points in a different direction. The direction of k
is speciﬁed by the polar angle θ (i.e., the angle subtended between the two wave
vectors), and an azimuthal angle ϕ about the zaxis. Equation (7.38) strongly
suggests that for a spherically symmetric scattering potential [i.e., V(r) = V(r)]
the scattering amplitude is a function of θ only:
f(θ, ϕ) = f(θ). (7.51)
It follows that neither the incident wavefunction,
φ(r) =
exp( i k z)
(2π)
3/2
=
exp( i k r cos θ)
(2π)
3/2
, (7.52)
nor the total wavefunction,
ψ(r) =
1
(2π)
3/2
_
_
exp( i k r cos θ) +
exp( i k r) f(θ)
r
_
_
, (7.53)
depend on the azimuthal angle ϕ.
Outside the range of the scattering potential, both φ(r) and ψ(r) satisfy the
free space Schr¨ odinger equation
(∇
2
+k
2
) ψ = 0. (7.54)
What is the most general solution to this equation in spherical polar coordinates
which does not depend on the azimuthal angle ϕ? Separation of variables yields
ψ(r, θ) =
l
R
l
(r) P
l
(cos θ), (7.55)
since the Legendre functions P
l
(cos θ) form a complete set in θspace. The Leg
endre functions are related to the spherical harmonics introduced in Sect. 5 via
P
l
(cos θ) =
¸
¸
¸
_
4π
2 l +1
Y
0
l
(θ, ϕ). (7.56)
178
7.4 Partial waves 7 SCATTERING THEORY
Equations (7.54) and (7.55) can be combined to give
r
2
d
2
R
l
dr
2
+2 r
dR
l
dr
+ [k
2
r
2
−l (l +1)]R
l
= 0. (7.57)
The two independent solutions to this equation are called a spherical Bessel func
tion, j
l
(k r), and a Neumann function, η
l
(k r). It is easily demonstrated that
j
l
(y) = y
l
_
−
1
y
d
dy
_
l
siny
y
, (7.58)
η
l
(y) = −y
l
_
−
1
y
d
dy
_
l
cos y
y
. (7.59)
Note that spherical Bessel functions are wellbehaved in the limit y →0 , whereas
Neumann functions become singular. The asymptotic behaviour of these func
tions in the limit y →∞is
j
l
(y) →
sin(y −l π/2)
y
, (7.60)
η
l
(y) → −
cos(y −l π/2)
y
. (7.61)
We can write
exp( i k r cos θ) =
l
a
l
j
l
(k r) P
l
(cos θ), (7.62)
where the a
l
are constants. Note there are no Neumann functions in this expan
sion, because they are not wellbehaved as r → 0. The Legendre functions are
orthonormal,
_
1
−1
P
n
(µ) P
m
(µ) dµ =
δ
nm
n +1/2
, (7.63)
so we can invert the above expansion to give
a
l
j
l
(k r) = (l +1/2)
_
1
−1
exp( i k r µ) P
l
(µ) dµ. (7.64)
It is wellknown that
j
l
(y) =
(−i)
l
2
_
1
−1
exp( i yµ) P
l
(µ) dµ, (7.65)
179
7.4 Partial waves 7 SCATTERING THEORY
where l = 0, 1, 2, [see Abramowitz and Stegun (Dover, New York NY, 1965),
Eq. 10.1.14]. Thus,
a
l
= i
l
(2 l +1), (7.66)
giving
exp( i k r cos θ) =
l
i
l
(2 l +1) j
l
(k r) P
l
(cos θ). (7.67)
The above expression tells us how to decompose a planewave into a series of
sphericalwaves (or “partial waves”).
The most general solution for the total wavefunction outside the scattering
region is
ψ(r) =
1
(2π)
3/2
l
[A
l
j
l
(k r) +B
l
η
l
(k r)] P
l
(cos θ), (7.68)
where the A
l
and B
l
are constants. Note that the Neumann functions are allowed
to appear in this expansion, because its region of validity does not include the
origin. In the larger limit, the total wavefunction reduces to
ψ(r) ·
1
(2π)
3/2
l
_
_
A
l
sin(k r −l π/2)
k r
−B
l
cos(k r −l π/2)
k r
_
_
P
l
(cos θ), (7.69)
where use has been made of Eqs. (7.60)–(7.61). The above expression can also
be written
ψ(r) ·
1
(2π)
3/2
l
C
l
sin(k r −l π/2 +δ
l
)
k r
P
l
(cos θ), (7.70)
where the sine and cosine functions have been combined to give a sine function
which is phaseshifted by δ
l
.
Equation (7.70) yields
ψ(r) ·
1
(2π)
3/2
l
C
l
exp[ i (k r −l π/2 +δ
l
)] − exp[−i (k r −l π/2 +δ
l
)]
2 i k r
P
l
(cos θ), (7.71)
which contains both incoming and outgoing sphericalwaves. What is the source
of the incoming waves? Obviously, they must be part of the larger asymptotic
180
7.5 The optical theorem 7 SCATTERING THEORY
expansion of the incident wavefunction. In fact, it is easily seen that
φ(r) ·
1
(2π)
3/2
l
i
l
(2l +1)
exp[ i (k r −l π/2)] − exp[−i (k r −l π/2)]
2 i k r
P
l
(cos θ) (7.72)
in the larger limit. Now, Eqs. (7.52) and (7.53) give
(2π)
3/2
[ψ(r) −φ(r)] =
exp( i k r)
r
f(θ). (7.73)
Note that the righthand side consists only of an outgoing spherical wave. This
implies that the coefﬁcients of the incoming spherical waves in the larger ex
pansions of ψ(r) and φ(r) must be equal. It follows from Eqs. (7.71) and (7.72)
that
C
l
= (2 l +1) exp[ i (δ
l
+l π/2)]. (7.74)
Thus, Eqs. (7.71)–(7.73) yield
f(θ) =
∞
l=0
(2 l +1)
exp( i δ
l
)
k
sinδ
l
P
l
(cos θ). (7.75)
Clearly, determining the scattering amplitude f(θ) via a decomposition into par
tial waves (i.e., sphericalwaves) is equivalent to determining the phaseshifts δ
l
.
7.5 The optical theorem
The differential scattering crosssection dσ/dΩ is simply the modulus squared of
the scattering amplitude f(θ). The total crosssection is given by
σ
total
=
_
f(θ)
2
dΩ
=
1
k
2
_
dϕ
_
1
−1
dµ
l
l
(2 l +1) (2 l
+1) exp[ i (δ
l
−δ
l
]
sinδ
l
sinδ
l
P
l
(µ) P
l
(µ), (7.76)
181
7.6 Determination of phaseshifts 7 SCATTERING THEORY
where µ = cos θ. It follows that
σ
total
=
4π
k
2
l
(2 l +1) sin
2
δ
l
, (7.77)
where use has been made of Eq. (7.63). A comparison of this result with Eq. (7.75)
yields
σ
total
=
4π
k
Im[f(0)] , (7.78)
since P
l
(1) = 1. This result is known as the optical theorem. It is a reﬂection
of the fact that the very existence of scattering requires scattering in the forward
(θ = 0) direction in order to interfere with the incident wave, and thereby reduce
the probability current in this direction.
It is usual to write
σ
total
=
∞
l=0
σ
l
, (7.79)
where
σ
l
=
4π
k
2
(2 l +1) sin
2
δ
l
(7.80)
is the lth partial crosssection: i.e., the contribution to the total crosssection from
the lth partial wave. Note that the maximumvalue for the lth partial crosssection
occurs when the phaseshift δ
l
takes the value π/2.
7.6 Determination of phaseshifts
Let us now consider how the phaseshifts δ
l
can be evaluated. Consider a spher
ically symmetric potential V(r) which vanishes for r > a, where a is termed
the range of the potential. In the region r > a, the wavefunction ψ(r) satisﬁes
the freespace Schr¨ odinger equation (7.54). The most general solution which is
consistent with no incoming sphericalwaves is
ψ(r) =
1
(2π)
3/2
∞
l=0
i
l
(2 l +1) A
l
(r) P
l
(cos θ), (7.81)
182
7.6 Determination of phaseshifts 7 SCATTERING THEORY
where
A
l
(r) = exp( i δ
l
) [ cos δ
l
j
l
(k r) − sinδ
l
η
l
(k r) ] . (7.82)
Note that Neumann functions are allowed to appear in the above expression,
because its region of validity does not include the origin (where V ,= 0). The
logarithmic derivative of the lth radial wavefunction A
l
(r) just outside the range
of the potential is given by
β
l+
= k a
_
_
cos δ
l
j
l
(k a) − sinδ
l
η
l
(k a)
cos δ
l
j
l
(k a) − sinδ
l
η
l
(k a)
_
_
, (7.83)
where j
l
(x) denotes dj
l
(x)/dx, etc. The above equation can be inverted to give
tanδ
l
=
k aj
l
(k a) −β
l+
j
l
(k a)
k aη
l
(k a) −β
l+
η
l
(k a)
. (7.84)
Thus, the problem of determining the phaseshift δ
l
is equivalent to that of ob
taining β
l+
.
The most general solution to Schr¨ odinger’s equation inside the range of the
potential (r < a) which does not depend on the azimuthal angle ϕ is
ψ(r) =
1
(2π)
3/2
∞
l=0
i
l
(2 l +1) R
l
(r) P
l
(cos θ), (7.85)
where
R
l
(r) =
u
l
(r)
r
, (7.86)
and
d
2
u
l
dr
2
+
_
_
k
2
−
2m
¯h
2
V −
l (l +1)
r
2
_
_
u
l
= 0. (7.87)
The boundary condition
u
l
(0) = 0 (7.88)
ensures that the radial wavefunction is wellbehaved at the origin. We can
launch a wellbehaved solution of the above equation from r = 0, integrate out
to r = a, and form the logarithmic derivative
β
l−
=
1
(u
l
/r)
d(u
l
/r)
dr
¸
¸
¸
¸
¸
¸
r=a
. (7.89)
183
7.7 Hard sphere scattering 7 SCATTERING THEORY
Since ψ(r) and its ﬁrst derivatives are necessarily continuous for physically ac
ceptible wavefunctions, it follows that
β
l+
= β
l−
. (7.90)
The phaseshift δ
l
is obtainable from Eq. (7.84).
7.7 Hard sphere scattering
Let us test out this scheme using a particularly simple example. Consider scatter
ing by a hard sphere, for which the potential is inﬁnite for r < a, and zero for
r > a. It follows that ψ(r) is zero in the region r < a, which implies that u
l
= 0
for all l. Thus,
β
l−
= β
l+
= ∞, (7.91)
for all l. It follows from Eq. (7.84) that
tanδ
l
=
j
l
(k a)
η
l
(k a)
. (7.92)
Consider the l = 0 partial wave, which is usually referred to as the swave.
Equation (7.92) yields
tanδ
0
=
sin(k a)/k a
−cos(k a)/ka
= −tank a, (7.93)
where use has been made of Eqs. (7.58)–(7.59). It follows that
δ
0
= −k a. (7.94)
The swave radial wave function is
A
0
(r) = exp(−i k a)
[cos k a sink r − sink a cos k r]
k r
= exp(−i k a)
sin[k (r −a)]
k r
. (7.95)
The corresponding radial wavefunction for the incident wave takes the form
˜
A
0
(r) =
sink r
k r
. (7.96)
184
7.7 Hard sphere scattering 7 SCATTERING THEORY
It is clear that the actual l = 0 radial wavefunction is similar to the incident l = 0
wavefunction, except that it is phaseshifted by k a.
Let us consider the low and high energy asymptotic limits of tanδ
l
. Low energy
means k a ¸ 1. In this regime, the spherical Bessel functions and Neumann
functions reduce to:
j
l
(k r) ·
(k r)
l
(2 l +1)!!
, (7.97)
η
l
(k r) · −
(2 l −1)!!
(k r)
l+1
, (7.98)
where n!! = n(n −2) (n −4) 1. It follows that
tanδ
l
=
−(k a)
2 l+1
(2 l +1) [(2 l −1)!!]
2
. (7.99)
It is clear that we can neglect δ
l
, with l > 0, with respect to δ
0
. In other words,
at low energy only swave scattering (i.e., spherically symmetric scattering) is
important. It follows from Eqs. (7.33), (7.75), and (7.94) that
dσ
dΩ
=
sin
2
k a
k
2
· a
2
(7.100)
for k a ¸1. Note that the total crosssection
σ
total
=
_
dσ
dΩ
dΩ = 4πa
2
(7.101)
is four times the geometric crosssection πa
2
(i.e., the crosssection for classical
particles bouncing off a hard sphere of radius a). However, low energy scattering
implies relatively long wavelengths, so we do not expect to obtain the classical
result in this limit.
Consider the high energy limit k a ¸ 1. At high energies, all partial waves
up to l
max
= k a contribute signiﬁcantly to the scattering crosssection. It follows
from Eq. (7.77) that
σ
total
=
4π
k
2
l
max
l=0
(2 l +1) sin
2
δ
l
. (7.102)
185
7.8 Low energy scattering 7 SCATTERING THEORY
With so many l values contributing, it is legitimate to replace sin
2
δ
l
by its average
value 1/2. Thus,
σ
total
=
ka
l=0
2π
k
2
(2 l +1) · 2πa
2
. (7.103)
This is twice the classical result, which is somewhat surprizing, since we might ex
pect to obtain the classical result in the short wavelength limit. For hard sphere
scattering, incident waves with impact parameters less than a must be deﬂected.
However, in order to produce a “shadow” behind the sphere, there must be scat
tering in the forward direction (recall the optical theorem) to produce destruc
tive interference with the incident planewave. In fact, the interference is not
completely destructive, and the shadow has a bright spot in the forward direc
tion. The effective crosssection associated with this bright spot is πa
2
which,
when combined with the crosssection for classical reﬂection, πa
2
, gives the ac
tual crosssection of 2πa
2
.
7.8 Low energy scattering
At low energies (i.e., when 1/k is much larger than the range of the potential)
partial waves with l > 0, in general, make a negligible contribution to the scatter
ing crosssection. It follows that, at these energies, with a ﬁnite range potential,
only swave scattering is important.
As a speciﬁc example, let us consider scattering by a ﬁnite potential well, char
acterized by V = V
0
for r < a, and V = 0 for r ≥ a. Here, V
0
is a constant.
The potential is repulsive for V
0
> 0, and attractive for V
0
< 0. The outside
wavefunction is given by [see Eq. (7.82)]
A
0
(r) = exp( i δ
0
) [j
0
(k r) cos δ
0
−η
0
(k r) sinδ
0
] (7.104)
=
exp( i δ
0
) sin(k r +δ
0
)
k r
, (7.105)
where use has been made of Eqs. (7.58)–(7.59). The inside wavefunction follows
186
7.8 Low energy scattering 7 SCATTERING THEORY
from Eq. (7.87). We obtain
A
0
(r) = B
sink
r
r
, (7.106)
where use has been made of the boundary condition (7.88). Here, B is a constant,
and
E −V
0
=
¯h
2
k
2
2 m
. (7.107)
Note that Eq. (7.106) only applies when E > V
0
. For E < V
0
, we have
A
0
(r) = B
sinhκ r
r
, (7.108)
where
V
0
−E =
¯h
2
κ
2
2 m
. (7.109)
Matching A
0
(r), and its radial derivative at r = a, yields
tan(k a +δ
0
) =
k
k
tank
a (7.110)
for E > V
0
, and
tan(k a +δ
0
) =
k
κ
tanhκ a (7.111)
for E < V
0
.
Consider an attractive potential, for which E > V
0
. Suppose that V
0
 ¸ E
(i.e., the depth of the potential well is much larger than the energy of the inci
dent particles), so that k
¸ k. It follows from Eq. (7.110) that, unless tank
a
becomes extremely large, the righthand side is much less that unity, so replacing
the tangent of a small quantity with the quantity itself, we obtain
k a +δ
0
·
k
k
tank
a. (7.112)
This yields
δ
0
· k a
_
_
tank
a
k
a
−1
_
_
. (7.113)
187
7.9 Resonances 7 SCATTERING THEORY
According to Eq. (7.102), the scattering crosssection is given by
σ
total
·
4π
k
2
sin
2
δ
0
= 4πa
2
_
_
tank
a
k
a
−1
_
_
2
. (7.114)
Now
k
a =
¸
¸
¸
_
k
2
a
2
+
2 mV
0
 a
2
¯h
2
, (7.115)
so for sufﬁciently small values of k a,
k
a ·
¸
¸
¸
_
2 mV
0
 a
2
¯h
2
. (7.116)
It follows that the total (swave) scattering crosssection is independent of the
energy of the incident particles (provided that this energy is sufﬁciently small).
Note that there are values of k
a (e.g., k
a · 4.49) at which δ
0
→ π, and the
scattering crosssection (7.114) vanishes, despite the very strong attraction of the
potential. In reality, the crosssection is not exactly zero, because of contributions
from l > 0 partial waves. But, at low incident energies, these contributions are
small. It follows that there are certain values of V
0
and k which give rise to almost
perfect transmission of the incident wave. This is called the RamsauerTownsend
effect, and has been observed experimentally.
7.9 Resonances
There is a signiﬁcant exception to the independence of the crosssection on en
ergy. Suppose that the quantity
_
2 mV
0
 a
2
/¯h
2
is slightly less than π/2. As the
incident energy increases, k
a, which is given by Eq. (7.115), can reach the value
π/2. In this case, tank
a becomes inﬁnite, so we can no longer assume that the
righthand side of Eq. (7.110) is small. In fact, at the value of the incident energy
when k
a = π/2, it follows from Eq. (7.110) that k a + δ
0
= π/2, or δ
0
· π/2
(since we are assuming that k a ¸1). This implies that
σ
total
=
4π
k
2
sin
2
δ
0
= 4πa
2
_
1
k
2
a
2
_
. (7.117)
188
7.9 Resonances 7 SCATTERING THEORY
Note that the crosssection now depends on the energy. Furthermore, the mag
nitude of the crosssection is much larger than that given in Eq. (7.114) for
k
a ,= π/2 (since k a ¸1).
The origin of this rather strange behaviour is quite simple. The condition
¸
¸
¸
_
2 mV
0
 a
2
¯h
2
=
π
2
(7.118)
is equivalent to the condition that a spherical well of depth V
0
possesses a bound
state at zero energy. Thus, for a potential well which satisﬁes the above equation,
the energy of the scattering system is essentially the same as the energy of the
bound state. In this situation, an incident particle would like to form a bound
state in the potential well. However, the bound state is not stable, since the
system has a small positive energy. Nevertheless, this sort of resonance scattering
is best understood as the capture of an incident particle to form a metastable
bound state, and the subsequent decay of the bound state and release of the
particle. The crosssection for resonance scattering is generally far higher than
that for nonresonance scattering.
We have seen that there is a resonant effect when the phaseshift of the swave
takes the value π/2. There is nothing special about the l = 0 partial wave, so it
is reasonable to assume that there is a similar resonance when the phaseshift of
the lth partial wave is π/2. Suppose that δ
l
attains the value π/2 at the incident
energy E
0
, so that
δ
l
(E
0
) =
π
2
. (7.119)
Let us expand cot δ
l
in the vicinity of the resonant energy:
cot δ
l
(E) = cot δ
l
(E
0
) +
_
dcot δ
l
dE
_
E=E
0
(E −E
0
) + (7.120)
= −
_
_
1
sin
2
δ
l
dδ
l
dE
_
_
E=E
0
(E −E
0
) + . (7.121)
Deﬁning
_
_
dδ
l
(E)
dE
_
_
E=E
0
=
2
Γ
, (7.122)
189
7.9 Resonances 7 SCATTERING THEORY
we obtain
cot δ
l
(E) = −
2
Γ
(E −E
0
) + . (7.123)
Recall, from Eq. (7.80), that the contribution of the lth partial wave to the scat
tering crosssection is
σ
l
=
4π
k
2
(2 l +1) sin
2
δ
l
=
4π
k
2
(2 l +1)
1
1 + cot
2
δ
l
. (7.124)
Thus,
σ
l
·
4π
k
2
(2 l +1)
Γ
2
/4
(E −E
0
)
2
+Γ
2
/4
. (7.125)
This is the famous BreitWigner formula. The variation of the partial crosssection
σ
l
with the incident energy has the form of a classical resonance curve. The
quantity Γ is the width of the resonance (in energy). We can interpret the Breit
Wigner formula as describing the absorption of an incident particle to form a
metastable state, of energy E
0
, and lifetime τ = ¯h/Γ (see Sect. 6.17).
190
Contents
1 Introduction 1.1 Major sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Fundamental concepts 2.1 The breakdown of classical physics . . . . . . . . . 2.2 The polarization of photons . . . . . . . . . . . . . 2.3 The fundamental principles of quantum mechanics 2.4 Ket space . . . . . . . . . . . . . . . . . . . . . . . 2.5 Bra space . . . . . . . . . . . . . . . . . . . . . . . 2.6 Operators . . . . . . . . . . . . . . . . . . . . . . . 2.7 The outer product . . . . . . . . . . . . . . . . . . 2.8 Eigenvalues and eigenvectors . . . . . . . . . . . . 2.9 Observables . . . . . . . . . . . . . . . . . . . . . . 2.10 Measurements . . . . . . . . . . . . . . . . . . . . 2.11 Expectation values . . . . . . . . . . . . . . . . . . 2.12 Degeneracy . . . . . . . . . . . . . . . . . . . . . . 2.13 Compatible observables . . . . . . . . . . . . . . . 2.14 The uncertainty relation . . . . . . . . . . . . . . . 2.15 Continuous spectra . . . . . . . . . . . . . . . . . . 3 Position and momentum 3.1 Introduction . . . . . . . . . . . 3.2 Poisson brackets . . . . . . . . . 3.3 Wavefunctions . . . . . . . . . . 3.4 Schr¨dinger’s representation  I . o 3.5 Schr¨dinger’s representation  II o 3.6 The momentum representation . 3.7 The uncertainty relation . . . . . 3.8 Displacement operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5 6 6 7 9 10 14 17 19 20 21 24 25 26 27 28 31 33 33 33 37 39 43 46 48 50
4 Quantum dynamics 55 4.1 Schr¨dinger’s equations of motion . . . . . . . . . . . . . . . . . . 55 o 4.2 Heisenberg’s equations of motion . . . . . . . . . . . . . . . . . . . 59
2
4.3 4.4
Ehrenfest’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Schr¨dinger’s waveequation . . . . . . . . . . . . . . . . . . . . . 65 o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 71 74 78 81 84 86 89 91 93 96 97 99 105 110 120 120 120 122 124 129 132 134 140 144 146 149 150 154 158 159
5 Angular momentum 5.1 Orbital angular momentum . . . . . . . . . . 5.2 Eigenvalues of angular momentum . . . . . . 5.3 Rotation operators . . . . . . . . . . . . . . . 5.4 Eigenfunctions of orbital angular momentum 5.5 Motion in a central ﬁeld . . . . . . . . . . . . 5.6 Energy levels of the hydrogen atom . . . . . . 5.7 Spin angular momentum . . . . . . . . . . . 5.8 Wavefunction of a spin onehalf particle . . . 5.9 Rotation operators in spin space . . . . . . . 5.10 Magnetic moments . . . . . . . . . . . . . . . 5.11 Spin precession . . . . . . . . . . . . . . . . . 5.12 Pauli twocomponent formalism . . . . . . . . 5.13 Spin greater than onehalf systems . . . . . . 5.14 Addition of angular momentum . . . . . . . .
6 Approximation methods 6.1 Introduction . . . . . . . . . . . . . . . . . . . . 6.2 The twostate system . . . . . . . . . . . . . . . . 6.3 Nondegenerate perturbation theory . . . . . . . 6.4 The quadratic Stark effect . . . . . . . . . . . . . 6.5 Degenerate perturbation theory . . . . . . . . . . 6.6 The linear Stark effect . . . . . . . . . . . . . . . 6.7 Fine structure . . . . . . . . . . . . . . . . . . . . 6.8 The Zeeman effect . . . . . . . . . . . . . . . . . 6.9 Timedependent perturbation theory . . . . . . . 6.10 The twostate system . . . . . . . . . . . . . . . . 6.11 Spin magnetic resonance . . . . . . . . . . . . . 6.12 The Dyson series . . . . . . . . . . . . . . . . . . 6.13 Constant perturbations . . . . . . . . . . . . . . . 6.14 Harmonic perturbations . . . . . . . . . . . . . . 6.15 Absorption and stimulated emission of radiation
3
. . . . . .6. . 7. . . . . . . .8 Low energy scattering . . . . . . . . . . . . . . . . . . . . . . .5 The optical theorem . . . . . . . . . . . . . . . . . . .9 Resonances . . . . . . 7. . .2 The LipmannSchwinger equation 7. . .6 Determination of phaseshifts . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Partial waves . .17 Energyshifts and decaywidths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Introduction . 7. . . . .16 The electric dipole approximation . . . . . . .3 The Born approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 7 Scattering theory 7. .7 Hard sphere scattering . . . . . . . . 170 170 170 175 178 181 182 184 186 188
4
. . . . . . . . 7. . . . . . . . . . . . . . . . . . . . . . 7. . . . . . . . . . . . . . . . . . . . . . . . . . . 162 6. . . . . . . . . . . 7. . . . . . . . . . . . . . . . . . 7. .
A.B. Oxford. 2nd Edition (John Wiley & Sons.1 INTRODUCTION
1
Introduction
1. J. P ford University Press. Quantum mechanics. 1985). Dirac. . 1970). 1958). and M. Reading MA.J. R.
5
.M. Merzbacher. Modern quantum mechanics. E. Sands. R. (OxThe principles of quantum mechanics. Sakurai. Volume III (AddisonWesley. New York NY. (Benjamin/Cummings. UK.1 Major sources The textbooks which I have consulted most frequently while developing course material are: . 4th Edition (revised). Menlo Park CA.P Feynman. 1965). The Feynman lectures on physics. Leighton.
various experiments (e. Incidentally. 4. light interference.
6
. However. The anomalously low speciﬁc heats of atoms and molecules: According to the equipartition theorem of classical physics. The ultraviolet catastrophe: According to classical physics.. I.1 The breakdown of classical physics The necessity for a departure from classical mechanics is clearly demonstrated by: 1. each degree of freedom of an atomic or molecular system should contribute R/2 to its molar speciﬁc heat. Experimentally. Waveparticle duality: Classical physics can deal with waves or particles.g. and the total energy density is ﬁnite. there is no such divergence. are largely apocryphal (see Feynman. this fundamental problem with classical physics was known and appreciated in the middle of the nineteenth century. In fact. this is not observed to happen. only the translational and some rotational degrees of freedom seem to contribute. 40). Vol. 2. This is completely inexplicable within the framework of classical physics. and gradually spiral in towards the nucleus. The anomalous stability of atoms and molecules: According to classical physics. the energy density of an electromagnetic ﬁeld in vacuum is inﬁnite due to a divergence of energy carried by short wavelength modes. electron diffraction) show quite clearly that waves sometimes act as if they were streams of particles. Cha. Stories that physicists at the start of the twentieth century thought that classical physics explained everything. The vibrational degrees of freedom appear to make no contribution at all (except at high temperatures). where R is the ideal gas constant. an electron orbiting a nucleus should lose energy by emission of synchrotron radiation. Experimentally. the photoelectric effect.2 FUNDAMENTAL CONCEPTS
2
Fundamental concepts
2. and that there was nothing left to discover. and streams of particles sometimes act as if they were waves. 3.
Classical electromagnetic wave theory tells us that if the beam is polarized perpendicular to the optic axis then all of the light is transmitted. Furthermore. in the latter case. which has the property that it is only transparent to light whose plane of polarization lies perpendicular to its optic axis. But. what happens in the case of an obliquely polarized incident beam? The above question is not very precise. and if the light is polarized at an angle α to the axis then a fraction sin 2 α of the beam is transmitted.2. or no photon is observed. also extend to its particlelike behaviour. and then look to see whether or not it emerges from the other side. on average. it is impossible to imagine (in physics) ﬁnding part of a photon on the other side of the ﬁlm. If we repeat the experiment a great number of times then. and a fraction
7
. Any photon which is transmitted though the ﬁlm must be polarized perpendicular to the optic axis. all of the photons are transmitted. Consider the following wellknown experiment. and. A beam of plane polarized light is passed through a polaroid ﬁlm. none of the photons are transmitted. whose energy is equal to the energy of the incident photon.2 The polarization of photons
2 FUNDAMENTAL CONCEPTS
2. Let us try to account for these observations at the individual photon level. a polarization can be ascribed to each individual photon in a beam of light. This picture leads to no difﬁculty if the plane of polarization lies parallel or perpendicular to the optic axis of the polaroid. a fraction sin2 α of the photons are transmitted through the ﬁlm. the polarization properties of light. is observed. In particular.2 The polarization of photons It is known experimentally that when plane polarized light is used to eject photoelectrons there is a preferred direction of emission of the electrons. Suppose that we were to ﬁre a single photon at a polaroid ﬁlm. In the former case. which are more usually associated with its wavelike behaviour. Clearly. A beam of light which is plane polarized in a certain direction is made up of a stream of photons which are each plane polarized in that direction. The possible results of the experiment are that either a whole photon. Let us reformulate it as a question relating to the result of some experiment which we could perform. if the beam is polarized parallel to the optic axis then none of the light is transmitted.
we must conclude that any state of polarization can be regarded as a superposition of two mutually perpendicular states of polarization. If we imagine performing experiments using monochromatic light. the oblique polarization state is some sort of superposition of two states of parallel and perpendicular polarization. We only know the probability of each event occurring. We have no way of knowing whether an individual obliquely polarized photon is going to be absorbed by or transmitted through a polaroid ﬁlm.
8
. we conclude that a photon has a probability sin 2 α of being transmitted as a photon polarized in the plane perpendicular to the optic axis. since they do not relate to the outcome of a possible experiment. are illegitimate. The further description provided by quantum mechanics is as follows. in all cases. These values for the probabilities lead to the correct classical limit for a beam containing a large number of photons. and partly in a state of polarization perpendicular to the axis. and adopting a fundamentally probabilistic approach. Thus. some further description is needed in order to allow the results of this experiment to be correlated with the results of other experiments which can be performed using photons. but recall that the state of a photon is fully speciﬁed once its energy. The above discussion about the results of an experiment with a single obliquely polarized photon incident on a polaroid ﬁlm answers all that can be legitimately asked about what happens to the photon when it reaches the ﬁlm. and there is nothing left over to uniquely determine whether the photon is transmitted or absorbed by the ﬁlm. Since there is nothing special about the orientation of the optic axis in our experiment. Questions as to what decides whether the photon is transmitted or not. or how it changes its direction of polarization.2 The polarization of photons
2 FUNDAMENTAL CONCEPTS
cos2 α are absorbed. direction of propagation.2. and polarization are known. then the state of each individual photon in the beam is completely speciﬁed. and a probability cos2 α of being absorbed. It is supposed that a photon polarized obliquely to the optic axis can be regarded as being partly in a state of polarization parallel to the axis. Note that we have only been able to preserve the individuality of photons. Nevertheless. normally incident on a polaroid ﬁlm. by abandoning the determinacy of classical theory. This is a fairly sweeping statement. with a particular oblique polarization. In other words.
3 The fundamental principles of quantum mechanics
2 FUNDAMENTAL CONCEPTS
When we make the photon encounter a polaroid ﬁlm. Any other questions lie beyond the realms of physics. the introduction of indeterminacy into the problem is clearly connected with the act of observation. we are observing whether it is polarized parallel or perpendicular to the optic axis. the indeterminacy is related to the inevitable disturbance of the system associated with the act of observation. Feynman. and the SternGerlach experiment (see Sakurai.3 The fundamental principles of quantum mechanics There is nothing special about the transmission and absorption of photons through a polaroid ﬁlm.e. the photon has to jump suddenly from being partly in each of these two states to being entirely in one or the other of them. The principle of indeterminacy: An observation made on a microscopic system causes it to jump into one or more particular states (which are related to
9
. Dirac’s razor: Quantum mechanics can only answer questions regarding the outcome of possible experiments. I. Cha. Which of the two states it will jump into cannot be predicted. such as the interference of photons (see Dirac.. Note that. molecule. 1. but is governed by probability laws. 5).
2. it is transmitted. an atom. 2. Exactly the same conclusions as those outlined above are obtained by studying other simple experiments. The principle of superposition of states: Any microscopic system (i. or particle) in a given state can be regarded as being partly in each of two or more other states. 3. The effect of making this observation is to force the photon entirely into a state of parallel or perpendicular polarization. we are subjecting it to an observation. any state can be regarded as a superposition of two or more other states. In fact. in this example. Sect.2. In other words. If the photon jumps into a state of parallel polarization then it is absorbed. In other words. Cha. The study of these simple experiments leads us to formulate the following fundamental principles of quantum mechanics: 1.3). In other words. Such superpositions can be performed in an inﬁnite number of different ways. Otherwise.
We need to extend it so that we can predict which possible states a system can jump into after a particular type of observation. any given state can be regarded as a superposition of two or more other states. The state vector A is conventionally written A . B and
10
. Thus. There will be various possible motions of the particles or bodies consistent with the laws of force. the superposition of two different states. states must be related to mathematical quantities of a kind which can be added together to give other quantities of the same kind. The ﬁrst of these principles was formulated by quantum physicists (such as Dirac) in the 1920s to fend off awkward questions such as “How can a system suddenly jump from one state into another?”.4 Ket space Consider a microscopic system composed of particles or bodies with speciﬁc properties (mass. which we also label A.1)
Suppose that state A is. as well as the probability of the system making a particular jump. Such a space is called a ket space (after Dirac).g. the second principle is the basis for the mathematical formulation of quantum mechanics..) interacting according to speciﬁc laws of force. in fact. where the other elements of the space represent all of the other possible states of the system. or “How does a system decide which state to jump into?”.
2. According to the principle of superposition of states. The most obvious examples of such quantities are vectors. We can represent this state as a particular vector.4 Ket space
2 FUNDAMENTAL CONCEPTS
the type of observation). residing in some vector space. momentum. moment of inertia. Let us consider a particular microscopic system in a particular state. which we label A: e.2. etc. and polarization. The ﬁnal principle is still rather vague. Let us term each such motion a state of the system. It is impossible to predict into which ﬁnal state a particular system will jump. however the probability of a given system jumping into a given ﬁnal state can be predicted. (2. As we shall see. a photon with a particular energy.
If c1 + c2 = 0 then the superposition process yields nothing at all: i. Note that we cannot form a new state by superposing a state with itself. we require cos α of state B. no distinction being made between the directions of the ket vectors A and −A . This new state is represented by cos α B + sin α C (2. The absence of a state is represented by the null vector 0 in ket space. one caveat to the above statements. and state C might represent a similar photon plane polarized in the ydirection. the sum of these two states represents a photon whose plane of polarization makes an angle of 45◦ with both the x. The fact that ket vectors pointing in the same direction represent the same state relates ultimately to the quantization of matter: i. All the states of the system are in one to one correspondence with all the possible directions of vectors in the ket space. and sin α of state C.2. The null vector has the fairly obvious property that A + 0 = A . (2. For instance. (2.3)
in ket space. are physically irrelevant.e..4)
corresponds to the same state that A does. This latter state is represented by B + C in ket space. By analogy with classical physics. and plane polarized in the xdirection. no state. etc. For instance. Thus. ket vectors differ from conventional vectors in that their magnitudes. In this case. however.4 Ket space
2 FUNDAMENTAL CONCEPTS
C. There is. or lengths.5)
for any vector A .. a photon polarized in the ydirection superposed with another photon polarized in the ydirection (with the same energy and momentum) gives the same photon. This implies that the ket vector c1 A + c2 A = (c1 + c2 )A (2.2)
where B is the vector relating to the state B. state B might represent a photon propagating in the zdirection. Suppose that we want to construct a state whose plane of polarization makes an arbitrary angle α with the xdirection. This interrelation is represented in ket space by writing A = B + C . the fact
11
. We can do this via a suitably weighted superposition of states B and C.and ydirections (by analogy with classical physics).e.
4 Ket space
2 FUNDAMENTAL CONCEPTS
that it comes in irreducible packets called photons.3).. in Eq. then the magnitude. This suggests that a circularly polarized photon is the superposition of a photon polarized in the xdirection (state B) and a photon polarized in the ydirection (state C).) or we see nothing—we can never see a fraction or a multiple of a state. etc.2. etc. but with the proviso that state C is 90 ◦ out of phase with state B. or an atom. with equal weights given to the two states. If we observe a microscopic system then we either see a state (i. Suppose that we want to construct a circularly polarized photon state. or a molecule. so that two vectors of different lengths pointing in the same direction would represent different wave states. (2. we know from classical physics that a circularly polarized wave is a superposition of two waves of equal amplitude. that any plane polarized state of a photon can be represented as a linear superposition of two orthogonal polarization states in which the weights are real numbers. Thus.8)
12
. Thus. electrons. which are in phase quadrature. A general elliptically polarized photon is represented by c1 B + c2 C . Suppose that the ket R is expressible linearly in terms of the kets A and B . Well. and the direction would correspond to the frequency and wavelength. we can use complex numbers to simultaneously represent the weighting and relative phase in a linear superposition. a photon.6)
where c1 and c2 are complex numbers. if we were to represent a classical wave by a vector.e. so that R = c1 A + c2 B . (2. or length. We have seen. (2. a circularly polarized photon is represented by B + i C in ket space. atoms.7) (2. of the vector would correspond to the amplitude of the wave. We conclude that a ket space must be a complex vector space if it is to properly represent the mutual interrelations between the possible states of a microscopic system. plane polarized in orthogonal directions. In classical physics. By analogy with classical physics. if we observe a wave then the amplitude of the wave can take any value between zero and inﬁnity.
and ydirections.g. the dimensionality of a ket space is equivalent to the number of independent ket vectors it contains. a particle in an inﬁnitely deep.g. Some microscopic systems have a denumerably inﬁnite number of independent states (e.. the ket space which represents the possible polarization states of a photon propagating in the zdirection is twodimensional (the two independent vectors correspond to photons plane polarized in the x. In conclusion. The possible states of such a system are represented as a ket space whose dimensions are nondenumerably inﬁnite. dimensions. Some microscopic systems have a ﬁnite number of independent states (e. a set of ket vectors (or states) are termed independent if none of them are expressible linearly in terms of the others.. If there are N independent states. This type of space requires a slightly different treatment to spaces of ﬁnite.4 Ket space
2 FUNDAMENTAL CONCEPTS
We say that R is dependent on A and B .2. In fact. we can also say that state R is dependent on states A and B. the states of a general microscopic system can be represented as a complex vector space of (possibly) inﬁnite dimensions. It follows that the state R can be regarded as a linear superposition of the states A and B. Unfortunately.g. or denumerably inﬁnite. Likewise.. Likewise. The dimensionality of a conventional vector space is deﬁned as the number of independent vectors contained in the space. Thus. then the possible states of the system are represented as an Ndimensional ket space. So.
13
. the spin states of an electron in a magnetic ﬁeld). Such a space can be treated in more or less the same manner as a ﬁnitedimensional space. respectively). some microscopic systems have a nondenumerably inﬁnite number of independent states (e. The possible states of such a system are represented as a ket space whose dimensions are denumerably inﬁnite. Such a space is termed a Hilbert space by mathematicians. any ket vector (or state) which is expressible linearly in terms of certain others is said to be dependent on them. onedimensional potential well). a free particle).
. such functionals are termed linear functionals. a ﬁnitedimensional. This process is represented mathematically by writing F(A ) = φA . (2. we need only consider this special subset. space].
14
. Let the i (where i runs from 1 to N) represent N independent ket vectors in this space.9)
Let us narrow our focus to those functionals which preserve the linear dependencies of the ket vectors upon which they operate. A general linear functional.5 Bra space A snack machine inputs coins plus some code entered on a key pad. N → ∞). We can imagine building a rather abstract snack machine which inputs ket vectors and outputs complex numbers in a deterministic fashion. Note that the input and output of the machine have completely different natures. (2. labeled A.2. Mathematicians call such a machine a functional. they are complete). labeled F. It also does so in a deterministic manner: i.. satisﬁes F(A + B ) = F(A ) + F(B ).11)
where the αi are an arbitrary set of complex numbers.e.e.
(2. The only way the functional F can satisfy Eq.e.5 Bra space
2 FUNDAMENTAL CONCEPTS
2. Only a special subset of denumerably inﬁnite dimensional spaces have this property (i. this is only strictly true for ﬁnitedimensional spaces. Not surprisingly.10)
A =
i=1
αi i . or denumerably inﬁnite dimensional (i. where A and B are any two kets in a given ket space. and spitting out a general complex number φA . labeled F.e. but since a ket space must be complete if it is to represent the states of a microscopic system. Consider an Ndimensional ket space [i.
(2. Imagine a general functional. and (hopefully) outputs a snack. the same money plus the same code produces the same snack (or the same error message) time after time. A general ket vector can be written1
N
(2..10) for all vectors in the ket space is if
N
F(A ) =
i=1
1
fi αi .12)
Actually.. acting on a general ket vector.
It follows. Let us deﬁne N basis functionals i which satisfy i(j ) = δij . There is a one to one correspondence between the elements of the ket space and those of the related bra space. This type of vector space is called a bra space (after Dirac). So.17)
.5 Bra space
2 FUNDAMENTAL CONCEPTS
where the fi are a set of complex numbers relating to the functional. · · ·  and  · · · . this implies that the set of all possible linear functionals acting on an Ndimensional ket space is itself an Ndimensional vector space. and its constituent vectors (which are actually functionals of the ket space) are called bra vectors. these vectors are written in mirror image notation.
DC
(2. speciﬁed by Eq. More generally.11). it is dual to the original ket space). Bra space is an example of what mathematicians call a dual vector space (i. that the dual to c A is c∗ A . the corresponding bra vector is written
N
A =
i=1
α∗ i. 1 2
15
DC
α∗ i
(2.13)
F =
i=1
fi i. For a general ket vector A. However.
(2. That is. It follows from the previous three equations that
N
(2. A ←→ A. (2. A is termed the dual vector to A . i
(2. there is a corresponding element.2. where c is a complex number.16)
where the are the complex conjugates of the αi . c1 A + c2 B ←→ c∗ A + c∗ B.. from the above.
There are an inﬁnite number of ways of setting up the correspondence between vectors in a ket space and those in the related bra space.14)
But.e.15)
where DC stands for dual correspondence. only one of these has any physical signiﬁcance. so that they can never be confused). which it is also convenient to label A. in the bra space. Note that bra vectors are quite different in nature to ket vectors (hence. for every element A of the ket space.
. we can deﬁne a normalized ket  A . This property of bra and ket vectors is essential for the probabilistic interpretation of quantum mechanics. This expression can be further simpliﬁed to give BA . so the operation can also be written BA .20) Consider the special case where B → A .18). 2 An inner product is (almost) analogous to a scalar product between a covariant and contravariant vector in some curvilinear space.11)]. which also implies that BA = 0. Consider the functional which is dual to the ket vector
N
B =
i=1
βi i
(2.5 Bra space
2 FUNDAMENTAL CONCEPTS
Recall that a bra vector is a functional which acts on a general ket vector. (2. According to Eqs. if all of the α i are zero in Eq. The combination of a bra and a ket yields a “bra(c)ket” (which is just a number).
(2.
2
AA ≥ 0.12).19)
Mathematicians term BA the inner product of a bra and a ket.2.23) A = AA
We can now appreciate the elegance of Dirac’s notation. ˜ Given a ket A which is not the null ket. that we can omit the round brackets without causing any ambiguity.20) that AA is a real number. i
(2. This operation is denoted B(A ). and (2. (2.12) and (2. where 1 ˜ A . (2.18)
acting on the ket vector A . (2. It is easily demonstrated that BA = AB ∗ .16).21)
(2. (2.
N
BA =
i=1
β∗ α i . (2.11). and spits out a complex number. Note.e. as will become apparent later.22)
16
. Two kets A and B are said to be orthogonal if AB = 0. (2. and that The equality sign only holds if A is the null ket [i. however. It follows from Eqs.
of a conventional vector. Operators X and Y are said to be equal if XA = YA (2.25)
17
. It is possible to deﬁne a dual bra space for a ket space of nondenumerably inﬁnite dimensions in much the same manner as that described above. More of this later.2. and is analogous to the length. Kronecker deltafunctions become Dirac deltafunctions. and X(cA ) = cXA . or magnitude.
2. Suppose that when this operator acts on a general ket vector A it spits out a new ket vector which is denoted XA .27) (2.6 Operators
2 FUNDAMENTAL CONCEPTS
with the property ˜ ˜ AA = 1. it makes sense to require that all kets corresponding to physical states have unit norms. Operator X is linear provided that X(A + B ) = XA + XB . Since A and cA represent the same physical state. The main differences are that summations over discrete labels become integrations over continuous labels. Consider a somewhat different machine which inputs a ket vector and spits out another ket vector in a deterministic fashion.26) (2.6 Operators We have seen that a functional is a machine which inputs a ket vector and spits out a complex number. Such operators are termed linear operators. (2. Consider an operator labeled X. completeness must be assumed (it cannot be proved).24) Here. for all ket vectors A and B . AA is known as the norm or “length” of A . for all complex numbers c. and the normalization convention is somewhat different. We are only interested in operators which preserve the linear dependencies of the ket vectors upon which they act. Mathematicians call such a machine an operator.
X + (Y + Z) = (X + Y) + Z. Operator X is termed the null operator if XA = 0 (2.34)
for any A and B.29) (2. This operator is uniquely determined by the original operator X.30)
So far. and the ket vector on the right. and A can be written BXA without ambiguity. A suitable notation to use for the resulting bra when X operates on B is BX.33) (2. Consider the dual bra to XA . This product is a number which depends linearly on A . The equation which deﬁnes this vector is ( BX)A = B(XA ) (2. We can also give a meaning to their operating on bra vectors.32) (2. X. it is noncommutative: X Y = Y X. The multiplication is associative: X(YA ) = (X Y)A = X YA . (2. Such addition is deﬁned to obey a commutative and associate algebra: X + Y = Y + X.31) (2.2. This bra depends antilinearly on A and must therefore depend linearly on A. it may be considered to be the inner product of A with some bra.6 Operators
2 FUNDAMENTAL CONCEPTS
for all kets in the ket space in question. Operators can also be multiplied. the operator in the middle. so we may look on it as the result of some linear operator applied to B. so we might as well call it the same operator acting on B . we have only considered linear operators acting on ket vectors. Operators can be added together. Thus. it may be regarded as the result of some
18
. This bra depends linearly on B. in general. Consider the inner product of a general bra B with the ket XA . However. provided we adopt the convention that the bra vector always goes on the left. Thus. X(Y Z) = (X Y)Z = X Y Z.28)
for all ket vectors in the space. The triple product of B.
the product B A is a linear operator. This operator is termed the adjoint of X. Suppose that we rightmultiply the above product by the general ket C . Thus. We obtain B AC = AC B .7 The outer product So far we have formed the following products: BA . which is just a number. X Y.38) (2.36)
2.2. Clearly. B A acting on a general ket C yields another ket. Are there any other products we are allowed to form? How about B A ? (2. A Hermitian operator ξ has the special property that it is its own adjoint: i. and is denoted X† . This operator also acts on bras. (2.39)
This clearly depends linearly on the ket A and the bra B .39) by a general bra C. ξ = ξ† . as is easily demonstrated by leftmultiplying the expression (2. BXA .7 The outer product
2 FUNDAMENTAL CONCEPTS
linear operator applied to A. AB .41)
Mathematicians term the operator B A the outer product of B and A.
19
. Thus. (2.37) It is also easily seen that the adjoint of the adjoint of a linear operator is equivalent to the original operator. XA . (2. It is also easily demonstrated that (B A)† = A B. DC XA ←→ AX† . The outer product should not be confused with the inner product.40)
since AC is just a number. plus (X Y)† = Y † X† . (2.35) It is readily demonstrated that BX† A = AXB ∗ .e. AX. (2..
and the eigenkets corresponding to different eigenvalues are orthogonal. . These are denoted ξξ = ξ ξ . the dual equation to Eq. are numbers called eigenvalues. Three important results are readily deduced: (i) The eigenvalues are all real numbers.2. (2. These are denoted x .47) where we have used the fact that ξ is not the null ket. (2. (2. . the ket XA is not a constant multiple of A .8 Eigenvalues and eigenvectors
2 FUNDAMENTAL CONCEPTS
2. (2. Since ξ is Hermitian.48)
20
. This proves that the eigenvalues are real numbers. Xx = x x .45)
If we leftmultiply Eq.42)
where x .44) where ξ is the eigenket associated with the eigenvalue ξ . (2. there are some special kets known as the eigenkets of operator X. Suppose that the eigenvalues ξ and ξ are different. Consider the eigenkets and eigenvalues of a Hermitian operator ξ..44) (for the eigenvalue ξ ) reads ξ ξ = ξ
∗
ξ . It follows from the above that ξ = ξ ∗.. (2.43) . Clearly.. and take the difference.. applying X to one of its eigenkets yields the same eigenket multiplied by the associated eigenvalue.46)
Suppose that the eigenvalues ξ and ξ are the same... rightmultiply the above equation by ξ .44) by ξ . x .
(2. However. (2. x . we obtain (ξ − ξ ∗ ) ξ ξ = 0. (2. It follows that ξ ξ = 0. x and have the property Xx = x x . .8 Eigenvalues and eigenvectors In general.
49) (iii) The dual of any eigenket is an eigenbra belonging to the same eigenvalue. the only objects we have left over are operators. energy.2. spin.9 Observables
2 FUNDAMENTAL CONCEPTS
which demonstrates that eigenkets corresponding to different eigenvalues are orthogonal. otherwise they would. by placing a polaroid ﬁlm in its path. assume that the dynamical variables of a microscopic system are represented as linear operators acting on the bras and kets which correspond to the various possible states of the system. kets.9 Observables We have developed a mathematical formalism which comprises three types of objects—bras. in general. Since the lengths of bras and kets have no physical signiﬁcance. (2. We have seen that if we observe the polarization state of a photon. An eigenbra of ξ corresponding to an eigenvalue ξ is deﬁned ξ ξ = ξ ξ . However. (ii) The eigenvalues associated with eigenkets are the same as the eigenvalues associated with eigenbras. and linear operators. We have already seen that kets can be used to represent the possible states of a microscopic system. the result is to cause the photon to jump into a state of polarization parallel or perpendicular to the optic axis of the ﬁlm.)? How can these be represented in our formalism? Well. and conversely. momentum. and the latter state is transmitted (which is how we tell them apart). The former state is absorbed. so we must conclude that bras could just as well be used to represent the states of a microscopic system. it is reasonable to suppose that nonlinear operators are also without physical signiﬁcance. We. therefore. In general. its position. What about the dynamical variables of the system (e. spit out bras/kets pointing in different directions when fed bras/kets pointing in the same direction but differing in length. Note that the operators have to be linear.
2. we cannot predict into which state a given photon will jump
21
. etc..g. there is a one to one correspondence between the elements of a ket space and its dual bra space.
then a second observation. each of these ﬁnal states is associated with a different result of the measurement: i.2. we can say that when a dynamical variable of a microscopic system is measured the system is caused to jump into one of a number of independent states (note that the perpendicular and parallel polarization states of our photon are linearly independent). There is nothing special about the polarization states of a photon.. by a fairly nonobvious leap of intuition. However. Furthermore. if an observation is made. we do know that if the photon is initially polarized parallel to the optic axis then it will deﬁnitely be absorbed. we are going to assert that a measurement of a dynamical variable corresponding to an operator X in ket space causes the system to jump into a state corresponding to one of the eigenkets of X. Note that the result of the measurement must be a real number (there are no measurement machines which output complex numbers).e. the result of the measurement is the eigenvalue associated with the eigenket into which the system jumps. and the system is found to be a one particular ﬁnal state.. We can make a second observation of the polarization state of such a photon by placing an identical polaroid ﬁlm (with the same orientation of the optic axis) immediately behind the ﬁrst ﬁlm. with one particular value for the dynamical variable.9 Observables
2 FUNDAMENTAL CONCEPTS
(except in a statistical sense). So. will deﬁnitely ﬁnd the system in the same state. The fact that the eigenkets of a Hermitian operator corresponding to different eigenvalues (i. It is clear that the photon will deﬁnitely be transmitted through the second ﬁlm. Finally. a different value of the dynamical variable. and yield the same value for the dynamical variable.e. such a state is termed an eigenstate. In general. and if it is initially polarized perpendicular to the axis then it will deﬁnitely be transmitted. Not surprisingly. We also known that after passing though the ﬁlm a photon must be in a state of polarization perpendicular to the optic axis (otherwise it would not have been transmitted). The fact that the result of the measurement must be a real number implies that dynamical variables can only be represented by Hermitian operators (since only Hermitian operators are guaranteed to have real eigenvalues). How can we represent all of these facts in our mathematical formalism? Well. different results of the measurement) are orthogonal is in accordance with our earlier requirement that the states into which the system jumps should be mutually
22
. more generally. made immediately after the ﬁrst one.
Conversely. A Hermitian operator which satisﬁes this condition is termed an observable. This can only be the case if the eigenkets form a complete set (i. the distinction between a state and its representative ket vector. We can conclude that the result of a measurement of a dynamical variable represented by a Hermitian operator ξ must be one of the eigenvalues of ξ. corresponding to an eigenvalue ξ . This gives us the physical signiﬁcance of the eigenvalues. it must always be able to jump into one of the eigenstates of ξ.e. However. It follows that no matter what the initial state of the system. In other words. then the states into which the system may jump on account of the measurement are such that the original state is dependent on them. This fairly innocuous statement has two very important corollaries. a general ket must always be dependent on the eigenkets of ξ.e. Thus.. for the sake of simplicity. if the system is in an eigenstate of ξ. Second. then a measurement of ξ is bound to give the result ξ .. in order for a Hermitian operator ξ to be observable its eigenkets must form a complete set. (From now on.) It is reasonable to suppose that if a certain dynamical variable ξ is measured with the system in a particular state. In other words. First. it stands to reason that a measurement of ξ must always yield some result. It follows that a second measurement made immediately after the ﬁrst one must leave the system in an eigenstate corresponding to the eigenvalue ξ . will be dropped. any observable quantity must be a Hermitian operator with a complete set of eigenstates. This follows because the system cannot jump into an eigenstate corresponding to a different eigenvalue of ξ. the system is left in the associated eigenstate. and a dynamical variable and its representative operator. Furthermore. since such a state is not dependent on the original state.9 Observables
2 FUNDAMENTAL CONCEPTS
independent. immediately after an observation whose result is a particular eigenvalue ξ . independent of) any other eigenstate corresponding to a different eigenvalue.2. the second measurement is bound to give the same result as the ﬁrst. this eigenstate is orthogonal to (i. they span ket space). Conversely.
23
. every eigenvalue of ξ is a possible result of a measurement made on the corresponding dynamical variable.
as a result of a measurement made on the system? Let us start with the simplest case. a complex number. How about if we identify the transition probability with the modulus squared of the inner product.
24
(2. This guess also gives the right answer for the transition probabilities between eigenstates. So. Since the eigenstates of an observable ξ form a complete set.50)
where δξ ξ is unity if ξ = ξ . If the system is initially in an eigenstate ξ then the transition probability to a eigenstate ξ corresponding to a different eigenvalue is zero.10 Measurements
2 FUNDAMENTAL CONCEPTS
2. It is easily demonstrated that A =
ξ
ξ
ξ A . (2. suppose that the system is initially in a state A which is not an eigenstate of ξ. In fact. since Aξ is.2.10 Measurements We have seen that a measurement of some observable ξ of a microscopic system causes the system to jump into one of the eigenstates of ξ. For the moment.  Aξ 2 ? This quantity is deﬁnitely a positive number (so it could be a probability). Note that the probability of a transition from an initial eigenstate ξ to a ﬁnal eigenstate ξ is the same as the value of the inner product ξ ξ . we are assuming that the eigenvalues of ξ are all different. It is impossible to determine into which eigenstate a given system will jump. It is convenient to normalize our eigenkets such that they all have unit norms. we can express any given state A as a linear combination of them.51)
. and the transition probability to the same eigenstate ξ is unity. but it is possible to predict the probability of such a transition. what is the probability that a system in some initial state A makes a transition to an eigenstate ξ of an observable ξ. and complex probabilities do not make much sense. it is the correct guess. Can we use this correspondence to obtain a general rule for calculating transition probabilities? Well. Can we identify the transition probability to a ﬁnal eigenstate ξ with the inner product Aξ ? The straight answer is “no”. The result of the measurement is the associated eigenvalue (or some function of this quantity). and zero otherwise. in general. Let us try again. It follows from the orthogonality property of the eigenkets that ξ ξ = δξ ξ .
The relative probability of a transition to an eigenstate ξ . (2. Note that all of the above results follow from the extremely useful (and easily proved) result ξ ξ  = 1. We know that each measurement yields the value ξ with probability P(ξ ).20). The absolute probability is clearly P(ξ ) =  Aξ 2  Aξ 2 . What is the mean value of the measurement? This quantity.11 Expectation values
2 FUNDAMENTAL CONCEPTS
A =
ξ
Aξ Aξ
ξ
ξ .57)
2.55)
If the ket A is normalized such that its norm is unity.54)
ξ
where 1 denotes the identity operator. which is equivalent to the relative probability of a measurement of ξ yielding the result ξ .52)  Aξ 2 . (2. = 2 AA ξ  Aξ  (2.56) (2. and use has been made of Eq. Suppose a measurement of the observable ξ is made on each system.11 Expectation values Consider an ensemble of microscopic systems prepared in the same initial state A . which is generally referred to as the expectation value of ξ. (2.2. ξ A =
ξ
(2. is given by ξ =
ξ
ξ P(ξ ) =
ξ
ξ  Aξ 2
25
. is P(ξ ) ∝  Aξ 2 . (2. and the fact that the eigenstates are mutually orthogonal.53)
AA
=
where the summation is over all the different eigenvalues of ξ. then this probability simply reduces to P(ξ ) =  Aξ 2 .
that any linear combination of ξa and ξb is also an eigenstate corresponding to the eigenvalue ξ ... (2. Degenerate eigenstates are necessarily orthogonal to any eigenstates corresponding to different eigenvalues. (2. These are termed degenerate eigenstates. between the elements of a ket space and those of its dual bra space. however. that we adopted earlier.8 does not work in this case).16).60) is satisﬁed because of the more general property (2. the expectation value of this operator is always unity: i.e.59)
2. This is unfortunate. 1. in general. This property depends on the particular correspondence (2. All states are eigenstates of this operator with the eigenvalue unity.21) of the norm. Note. It follows that we can always construct two mutually orthogonal degenerate eigenstates. (2. they are not orthogonal to each other (i. 1 −  ξa ξb 2 (2. Consider the identity operator. since much of the previous formalism depends crucially on the mutual orthogonality of the different eigenstates of an observable. = ξb − ξa ξb ξa .
(2.58)
which reduces to ξ = AξA with the aid of Eq. For instance. 2. ξ1 ξ2 = ξa . but.62)
26
.54).e.12 Degeneracy
2 FUNDAMENTAL CONCEPTS
=
ξ
ξ Aξ
ξ A =
ξ
Aξξ
ξ A . Thus.2.61) (2. A1A = AA = 1. (2. the proof of orthogonality given in Sect.12 Degeneracy Suppose that two different eigenstates ξa and ξb of ξ correspond to the same eigenvalue ξ .60) for all A . Note that it is only possible to normalize a given ket A such that Eq.
if all eigenstates of ξ are also eigenstates of η then it is always possible to make a simultaneous measurement of ξ and η.13 Compatible observables Suppose that we wish to simultaneously measure two observables. In this sense. suppose that the eigenstate ξ is also an eigenstate of η. Another measurement of ξ will throw the system into one of the (many) eigenstates of ξ which depend on η . Suppose that we make a measurement of ξ. 1. of a microscopic system? Let us assume that we possess an apparatus which is capable of measuring ξ. Each eigenstate is again associated with a different possible result of the measurement. we can say that the observables ξ and η simultaneously have the values ξ and η . These could be measured using appropriate SternGerlach apparatuses (see Sakurai.13 Compatible observables
2 FUNDAMENTAL CONCEPTS
This result is easily generalized to the case of more than two degenerate eigenstates. Such observables are termed compatible. Suppose that the system is thrown into an eigenstate η . and another which can measure η. the two observables in question might be the projection in the x.and zdirections of the spin angular momentum of a spin onehalf particle. and the system is consequently thrown into one of the eigenstates of ξ. however. each of these eigenstates is associated with a different result of the measurement. ξ . A second measurement of ξ will deﬁnitely give the result ξ . ξ and η. with eigenvalue ξ . with eigenvalue ξ . In principle. We can now make a second observation to determine η. that the eigenstates of ξ are not eigenstates of η.
2. a measurement of η will deﬁnitely give the result η . What happens if we now make a measurement of η? Well. In this case. Sect. with eigenvalue η . respectively. Clearly. Suppose. with the eigenvalue η . This will throw the system into one of the (many) eigenstates of η which depend on ξ . and so on. Is it still possible to measure both observables simultaneously? Let us again make an observation of ξ which throws the system into an eigenstate ξ . We conclude that it is always possible to construct a complete set of mutually orthogonal eigenstates for any given observable. For instance.1). It is clear that if the observables ξ and η do not possess simultaneous eigenstates then if the value
27
.2.
2.64)
We can leftmultiply the ﬁrst equation by η.63) (2. with eigenvalue η . the condition for two observables ξ and η to be simultaneously measurable is that they should commute.14 The uncertainty relation We have seen that if ξ and η are two noncommuting observables.66)
Thus.. (2. Suppose that this is the case. We have ξξ η ηξ η = ξ ξ η . It is convenient to denote this simultaneous eigenstate ξ η . It is
28
.
2. also be an eigenstate of η. the system is in an eigenstate of ξ) then the value of η is uncertain (i. We say that the two observables are incompatible. the system is not in an eigenstate of η). The result is (ξ η − η ξ)ξ η = 0 (2. every eigenstate of ξ should also be an eigenstate of η)..67) (2. = η ξ η .e. and the second equation by ξ. We have seen that the condition for two observables ξ and η to be simultaneously measurable is that they should possess simultaneous eigenstates (i. Thus. (2. It follows that the simultaneous eigenstates of two observables must also form a complete set. then a determination of the value of ξ leaves the value of η uncertain. where A is a general ket.65)
for each simultaneous eigenstate.. the above equation implies that (ξ η − η ξ)A = 0 . and vice versa. Recall that the eigenstates of an observable must form a complete set.e. and vice versa. Let a general eigenstate of ξ. with eigenvalue ξ .e. and then take the difference.14 The uncertainty relation
2 FUNDAMENTAL CONCEPTS
of ξ is known (i. The only way that this can be true is if ξ η = η ξ.
(2.71) (2. nonzero. and a measurement of ξ is bound to give the expectation value.69)
The variance of ξ is a measure of the uncertainty in the value of ξ for the particular state in question (i. ξ . which is the same as the Schwarz inequality. Let us substitute A B = ∆ξ . If the variance is zero then there is no uncertainty.e.. (2.70)
where c is any complex number.72) a2  b2 ≥ a · b2 (2. Consider the Schwarz inequality AA BB ≥  AB 2 .68) where the expectation value is taken over the particular physical state under consideration. For a general observable ξ. in general. The expectation value of (∆ξ)2 ≡ ∆ξ ∆ξ is termed the variance of ξ. it is easily demonstrated that (∆ξ)2 = ξ2 − ξ 2 . In fact. we can deﬁne a Hermitian operator ∆ξ = ξ − ξ .75) (2. it is a measure of the width of the distribution of likely values of ξ about the expectation value).14 The uncertainty relation
2 FUNDAMENTAL CONCEPTS
possible to quantify this uncertainty.73)
29
. If c takes the special value − BA / BB then the above inequality reduces to AA BB −  AB 2 ≥ 0. It is obvious that the expectation value of ∆ξ is zero. (2.2.74) (2. (2. and is. This inequality can be proved by noting that ( A + c∗ B)(A + cB ) ≥ 0. which is analogous to in Euclidian space. = ∆η .
∆η])† = (∆ξ ∆η − ∆η ∆ξ)† = ∆η ∆ξ − ∆ξ ∆η = − [∆ξ.14 The uncertainty relation
2 FUNDAMENTAL CONCEPTS
where use has been made of the fact that ∆ξ and ∆η are Hermitian operators. ∆η} 2 .77) ∆ξ ∆η = [∆ξ. ∆η} . and the anticommutator. (2.
([∆ξ. in which case exact knowledge of ξ does not necessarily imply no knowledge of η. Now. an exact knowledge of the value of ξ implies no knowledge whatsoever of the value of η.81) 2 2 consists of the sum of a purely real and a purely imaginary number.2. whereas the expectation value of an antiHermitian operator is a pure imaginary number.
(2. ∆η] + {∆ξ. {∆ξ. [∆ξ. and vice versa.
30
. (2.76)
{∆ξ. (2. ∆η}. We ﬁnd (∆ξ)2 (∆η)2 ≥  ∆ξ ∆η 2 . (2. it is easily demonstrated that the expectation value of a Hermitian operator is a real number. The above expression is termed the uncertainty relation. The commutator is clearly antiHermitian. Taking the modulus squared of both sides gives ∆ξ ∆η = 1 1  ∆ξ ∆η 2 =  [ξ. so we can write 1 (∆ξ)2 (∆η)2 ≥  [ξ. According to this relation.78) (2.83) 4 where use has been made of Eq. ∆η] ≡ ∆ξ ∆η − ∆η ∆ξ. (2. ∆η] + {∆ξ. ∆η} .79)
into the Schwarz inequality. etc.82) 4 4 where use has been made of ∆ξ = 0. ∆η] .76). It is clear that the right hand side of 1 1 [∆ξ.80)
whereas the anticommutator is obviously Hermitian. η] 2 . The ﬁnal term in the above expression is positive deﬁnite. 2 2 where the commutator. η] 2 +  {∆ξ. (2. The one exception to this rule is when ξ and η commute. where the blank ket  stands for any general ket. Note that 1 1 (2. are deﬁned [∆ξ. ∆η} ≡ ∆ξ ∆η + ∆η ∆ξ. ∆η].
54) generalizes to dξ ξ ξ  = 1. continuous eigenvalues are unavoidable in quantum mechanics.e. (2. they are inﬁnitely long. Unfortunately. The extremely useful relation (2.. Note. these eigenstates have inﬁnite norms: i. Fortunately. that ξ can take any value. rather than having discrete values. we have studiously avoided dealing with observables possessing eigenvalues which lie in a continuous range.15 Continuous spectra Up to now. the dimensionality of ket space is nondenumerably inﬁnite. The reason for this is because continuous eigenvalues imply a ket space of nondenumerably inﬁnite dimension.2.84)
But. (2.50) generalizes to ξ ξ = δ(ξ − ξ ). In fact. many of the results we obtained previously for a ﬁnitedimensional ket space with discrete eigenvalues can be generalized to ket spaces of nondenumerably inﬁnite dimensions. The orthogonality condition (2. (2. This is the major difference between eigenstates in a ﬁnitedimensional and an inﬁnitedimensional ket space. for the sake of simplicity.85) where δ(x) denotes the famous Dirac deltafunction. Let us assume. generally have continuous eigenvalues. also. We can still write the eigenvalue equation as ξξ = ξ ξ .86)
Note that a summation over discrete eigenvalues goes over into an integral over a continuous range of eigenvalues. In fact. that eigenstates corresponding to a continuous range of eigenvalues cannot be normalized so that they have unit norms.15 Continuous spectra
2 FUNDAMENTAL CONCEPTS
2. the most important observables of all. It follows that any general ket can be expanded in
31
. ξ can now take a continuous range of values. The eigenstates ξ must form a complete set if ξ is to be an observable. Note that there are clearly a nondenumerably inﬁnite number of mutually orthogonal eigenstates of ξ. Hence. Suppose that ξ is an observable with continuous eigenvalues. namely position and momentum.
II and III). ξ A = dξ  Aξ 2 .87) (2. (2.15 Continuous spectra
2 FUNDAMENTAL CONCEPTS
terms of the ξ . or observables whose eigenvalues take on a ﬁnite range of values plus a set of discrete values.86).88) (2. according to Eq. For instance. Both of these cases can be dealt with using a fairly straightforward generalization of the previous analysis (see Dirac.90)
We have now studied observables whose eigenvalues can take a discrete number of values as well as those whose eigenvalues can take any value. this convenient normalization is still possible for a general state vector.89)
A = AA =
dξ Aξ dξ Aξ
These results also follow simply from Eq.
32
. There are number of other cases we could look at.2. In fact. the normalization condition can be written AA = dξ  Aξ 2 = 1. observables whose eigenvalues can only take a ﬁnite range of values. (2. Fortunately. (2. We have seen that it is not possible to normalize the eigenstates ξ such that they have unit norms.51)–(2. the expansions (2.53) generalize to A = dξ ξ ξ A . (2.89). Cha. ξ . In fact.
what is the value of q p − p q?
3. the behaviour of a classical system can be speciﬁed in terms of Lagrangian or Hamiltonian dynamics. a Cartesian coordinate has an associated linear momentum. This function is
33
. Our ﬁrst task is to discover a quantum mechanical replacement for the classical result q p − p q = 0. the position q and momentum p of some component of a dynamical system are represented as real numbers which. we have considered general dynamical variables represented by general linear operators acting in ket space. etc. in Hamiltonian dynamics. dt ∂pi dpi ∂H = − . dqi ∂H = . For instance. these quantities are represented as noncommuting linear Hermitian operators acting in a ket space which represents all of the possible states of the system. commute. Do the position and momentum operators commute? If not. However.2 Poisson brackets Consider a dynamic system whose state at a particular time t is fully speciﬁed by N independent classical coordinates qi (where i runs from 1 to N). pi . Associated with each generalized coordinate qi is a classical canonical momentum pi .1) (3. t) is the energy of the system at time t expressed in terms of the classical coordinates and canonical momenta. an angular coordinate has an associated angular momentum. As is wellknown. Let us investigate the role of such variables in quantum mechanics.2)
where the function H(qi . by deﬁnition. in classical mechanics the most important dynamical variables are those involving position and momentum. In quantum mechanics. In classical mechanics. For instance. dt ∂qi (3.3 POSITION AND MOMENTUM
3
Position and momentum
3.1 Introduction So far.
5) (3. pj ] = δij . The Poisson bracket of two dynamical variables u and v is deﬁned
N
[u. It is easily demonstrated that [qi . v] = −[v. We are interested in ﬁnding some construct of classical dynamics which consists of products of dynamical variables. the main properties of the classical Poisson bracket are as follows: [u. pj ] = 0. where use has been made of Hamilton’s equations.6)
The time evolution of a dynamical variable can also be written in terms of a Poisson bracket by noting that du = dt =
i=1 N i=1 N
∂u dpi ∂u dqi + ∂qi dt ∂pi dt ∂u ∂H ∂u ∂H − ∂qi ∂pi ∂pi ∂qi (3. (3. indeed.7)
= [u. [qi . If such a construct exists we hope to generalize it somehow to obtain a rule describing how dynamical variables commute with one another in quantum mechanics.
34
(3. [pi . ∂qi ∂pi ∂pi ∂qi
(3. instead of functions? Well. H].2 Poisson brackets
3 POSITION AND MOMENTUM
usually referred to as the Hamiltonian of the system.3)
where u and v are regarded as functions of the coordinates and momenta qi and pi . v] =
i=1
∂u ∂v ∂u ∂v − . one wellknown construct in classical dynamics which involves products of dynamical variables.4) (3. There is.3. qj ] = 0. u].
Can we construct a quantum mechanical Poisson bracket in which u and v are noncommuting operators.8)
.
etc.15)
= [u1 . v1 ]v2 + v1 [u1 . v2 ]u2 + u1 [u2 . v] = [u1 . v2 ] (3. v] = [u1 . (3. [u. v1 v2 ]u2 + u1 [u2 . v2 ]} = [u1 .16) (3. v]. w. represent dynamical variables. v2 ]. Equating the above two results yields [u1 .
(3. v1 ] + [u.12) (3. v1 ]v2 + u1 v1 [u2 . [u1 + u2 . v1 ]v2 u2 + v1 [u1 .9) (3. v] + [u2 . we can evaluate the Poisson bracket [u1 u2 . v1 ]v2 + v1 [u1 u2 . and c represents a number. since they now represent noncommuting operators. Thus. v2 ] [u1 u2 . Can we ﬁnd some combination of noncommuting operators u and v. v1 ]u2 v2 + u1 [u2 .13) (3. Note that the order of the various factors has been preserved.. [u1 u2 . [u. it follows that u1 v1 − v1 u1 = i ¯ [u1 . v1 v2 ] = {[u1 . v2 ]. v]u2 + u1 [u2 . h
35
(3.2 Poisson brackets
3 POSITION AND MOMENTUM
[u. c] = 0. v2 ]u2 + v1 u1 [u2 . v1 v2 ] = [u1 u2 . v2 ]. v1 v2 ] in two different ways. v1 v2 ] = [u.12) or (3. v1 ]. v]] = 0.11) (3. and [u. [u. u. In the above. and [u1 u2 .10) (3. which satisﬁes all of the above relations? Well. v.19)
.18) (3. v1 ](u2 v2 − v2 u2 ) = (u1 v1 − v1 u1 )[u2 . v1 ]v2 + v1 [u. h u2 v2 − v2 u2 = i ¯ [u2 . v1 ]v2 + v1 [u2 .3. v1 + v2 ] = [u. [v. since we can use either of the formulae (3. u]] + [w. v]. v2 ]} u2 + u1 {[u2 .13) ﬁrst. w]] + [v. v2 ].. v2 ]. etc.17)
Since this relation must hold for u1 and v1 quite independent of u2 and v2 .14)
The last relation is known as the Jacobi identity. [w. v1 v2 ] = [u1 . v1 ]v2 + v1 [u1 .
3.24) (3. and the quantum mechanical Poisson bracket.25)
These results provide us with the basis for calculating commutation relations between general dynamical variables. where h h = 6. (3. Quantum mechanics agrees with h experiments provided that ¯ takes the value h/2π. h (3.14). h Thus. pj ] = i ¯ δij . etc. This requirement is satisﬁed if ¯ is a real number. Thus. v] is conventionally reserved for the commutator u v − v u in quantum mechanics. satisﬁes all of the relations (3.22). [pi .22) i¯ h It is easily demonstrated that the quantum mechanical Poisson bracket. (3.3). (3. (3. the quantum mechanical Poisson bracket of two dynamical variables u and v is given by uv − vu [u.6261 × 10−34 J s (3. u2 . deﬁned in Eq. v]quantum = The strong analogy we have found between the classical Poisson bracket. v] = . we are assuming that Eqs. at least for the simplest cases. This argument yields the fundamental commutation relations [qi .6) hold for quantum mechanical as well as classical Poisson brackets. if two dynamical variables. In other words. (3.23) (3. h Since u1 . For instance. [qi . ξ
36
. We will use [u. it follows that ¯ is just a number. Somewhat confusingly. the notation [u.4)–(3. v]quantum to denote the quantum Poisson bracket.21)
is Planck’s constant. [u. since the classical Poisson bracket of two real dynamical variables is real. and also commutes with (u1 v1 −v1 u1 ). are quite general operators.. v] . v2 . deﬁned in Eq. qj ] = 0. We h want the quantum mechanical Poisson bracket of two Hermitian operators to be an Hermitian operator itself. as deﬁned above.20) i¯ h where ¯ is a new universal constant of nature.2 Poisson brackets
3 POSITION AND MOMENTUM
where ¯ does not depend on u1 . [u. pj ] = 0. leads us to make the assumption that the quantum mechanical bracket has the same value as the corresponding classical bracket.8)–(3. v1 .
3 Wavefunctions
3 POSITION AND MOMENTUM
and η.
37
.27)
This formula expresses the fact that the eigenkets are complete. qi and pi . correspond to a different classical degree of freedom of the system.86)]
+∞
dx x
−∞
x  = 1. each pair of generalized coordinate and its conjugate momentum. which corresponds to the Cartesian coordinate x.
3.3.
(3.13) allows [ξ. (3. x could be the position of a free particle). and suitably normalized.25) provide the foundation for the analogy between quantum mechanics and classical mechanics. It is clear from Eqs. (2. can both be written as a power series in the qi and pi . mutually orthogonal.g. classical mechanics can be h regarded as the limiting case of quantum mechanics when ¯ goes to zero. [see Eq.25). Note that the classical result (that everything commutes) is obtained in the limit ¯ → 0. the operator x possesses eigenvalues x lying in the continuous range −∞ < x < +∞ (since the eigenvalues correspond to all the possible results of a measurement of x). (2. An eigenket corresponding to the eigenvalue x is denoted x . In classih cal mechanics. (3. Moreover. Moreover.26)
The eigenkets satisfy the extremely useful relation [see Eq..8)–(3. Suppose that x is free to take any value (e.25) that in quantum mechanics the dynamical variables corresponding to different degrees of freedom all commute.23)–(3.23)–(3.85)] x x = δ(x − x ).3 Wavefunctions Consider a simple system with one classical degree of freedom. We can span ket space using the suitably normalized eigenkets of x. (3. The classical dynamical variable x is represented in quantum mechanics as a linear Hermitian operator which is also called x. η] to be expressed in terms of the fundamental commutation relations (3. Equations (3. Thus.23)–(3. then repeated application of Eqs. It is only those variables corresponding to the same degree of freedom which may fail to commute.
28)].30)
This formula is only valid if the state ket A is properly normalized: i.28)
The quantity x A is a complex function of the position eigenvalue x . We can write x A = ψA (x ). Note that state A is completely speciﬁed by its wavefunction ψA (x ) [since the wavefunction can be used to reconstruct the state ket A using Eq. It is clear that the wavefunction of state A is simply the collection of the weights of the corresponding state ket A . Thus. the probability of a measurement of position yielding a result in the range x to x + dx when the wavefunction of the system is ψA (x ) is P(x . The corresponding normalization for the wavefunction is
+∞
ψA (x )2 dx = 1. if AA = 1. (3. In fact.29) Here. This result is easily generalized to dynamical variables possessing continuous eigenvalues. Recall. 2. that the probability of a measurement of a dynamical variable ξ yielding the result ξ when the system is in state A is given by  ξ A 2 . (3. B
(3. when it is expanded in terms of the eigenkets of the position operator. In other words. ψA (x ) is the famous wavefunction of quantum mechanics. from Sect. (3.10..27).31)
Consider a second state B represented by a state ket B and a wavefunction ψB (x ). The inner product BA can be written
+∞ +∞
BA =
−∞
dx Bx
x A =
−∞
ψ∗ (x ) ψA (x ) dx . the probability of a measurement of x yielding a result lying in the range x to x + dx when the system is in a state A is  x A 2 dx .3.32)
38
.
+∞
A =
−∞
dx x A x
(3. dx ) = ψA (x )2 dx . assuming that the eigenvalues of ξ are discrete.
−∞
(3.3 Wavefunctions
3 POSITION AND MOMENTUM
A state ket A (which represents a general state A of the system) can be expressed as a linear superposition of the eigenkets of the position operator using Eq. (3.e.
If B = f(x)A then it follows that
+∞
ψB (x ) = =
x f(x)
−∞ +∞
dx ψA (x )x (3. The ket is termed the standard ket.g.e.29).3. Consider a general function f(x) of the observable x [e. the inner product of two states is related to the overlap integral of their wavefunctions.34) where use has been made of Eq.36)
Note. and is denoted .33)
dx f(x ) ψA (x ) x x .27) and (3. The dual of the standard ket is termed the standard bra.I o
3 POSITION AND MOMENTUM
where use has been made of Eqs. Here. where ψ(x) is a general function of the position operator x. It follows. ﬁnally.4 Schr¨dinger’s representation . (3.4 Schr¨ dinger’s representation . (3. A general state ket can be written ψ(x) . and the ket has the wavefunction ψ(x ) = 1. (3. from the above result..
−∞
giving ψB (x ) = f(x ) ψA (x ). if f(x) = x 2 then f(x ) = x 2 .26). that ψA (x) is often shortened to ψA .
3..35) where ψA (x) is the same function of the operator x that the wavefunction ψ A (x ) is of the position eigenvalue x . leaving the dependence on the position operator x tacitly understood. It is easily seen that
∗ ψA (x) ←→ ψA (x) . and ψ(x ) is the associated wavefunction. that a general state ket A can be written A = ψA (x) . f(x) = x 2 ]. (3. Thus. Consider the ket whose wavefunction
39
. f(x ) is the same function of the position eigenvalue x that f(x) is of the position operator x: i. DC
(3.I o Consider the simple system described in the previous section.
This ket is denoted dψ/dx . The new ket is clearly a linear function of the original ket. (3.3. It follows.4 Schr¨dinger’s representation . dx
(3. (3. (3. we can write
+∞ −∞
d φ x dx ψ(x ) = dx
+∞
φ(x ) dx
−∞
dψ(x ) .43) dx dx dx dx where use has been made of Eq.I o
3 POSITION AND MOMENTUM
is dψ(x )/dx . dx dx (3. (3. dx
(3.27) and (3.40)
assuming that the contributions from the limits of integration vanish.37)
Any linear operator which acts on ket vectors can also act on bra vectors.35) and (3. by comparison with Eqs. (3. (3. Let us denote this operator d/dx. Note that d dψ DC dψ∗ d ψ = ←→ = − ψ∗ .42) dx dx The neglect of contributions from the limits of integration in Eq. (3.36). that d † d =− .44) dx dx
40
. (2.31)].39)
The righthand side can be transformed via integration by parts to give
+∞ −∞
d φ x dx ψ(x ) = − dx
+∞ −∞
dφ(x ) dx ψ(x ). (3.40) is reasonable because physical wavefunctions are squareintegrable [see Eq.29). It follows that dφ(x ) d . (3.38) φ dx dx Making use of Eqs. It follows that d dψ ψ = .42). the bra φ d/dx satisﬁes d d ψ = φ ψ . (2.41) φ x = − dx dx which implies d dφ φ =− . According to Eq. Consider d/dx acting on a general bra φ(x).34). so we can think of it as the result of some linear operator acting on ψ .
3.4 Schr¨dinger’s representation  I o
3 POSITION AND MOMENTUM
Thus, d/dx is an antiHermitian operator. Let us evaluate the commutation relation between the operators x and d/dx. We have d d(x ψ) d xψ = =x ψ +ψ . (3.45) dx dx dx Since this holds for any ket ψ , it follows that d d x−x = 1. (3.46) dx dx Let p be the momentum conjugate to x (for the simple system under consideration p is a straightforward linear momentum). According to Eq. (3.25), x and p satisfy the commutation relation xp − px = i¯. h (3.47)
It can be seen, by comparison with Eq. (3.46), that the Hermitian operator −i ¯ d/dx satisﬁes the same commutation relation with x that p does. The most h general conclusion which may be drawn from a comparison of Eqs. (3.46) and (3.47) is that d p = −i ¯ h + f(x), (3.48) dx since (as is easily demonstrated) a general function f(x) of the position operator automatically commutes with x. We have chosen to normalize the eigenkets and eigenbras of the position operator so that they satisfy the normalization condition (3.26). However, this choice of normalization does not uniquely determine the eigenkets and eigenbras. Suppose that we transform to a new set of eigenbras which are related to the old set via x new = e i γ x old , (3.49) where γ ≡ γ(x ) is a real function of x . This transformation amounts to a rearrangement of the relative phases of the eigenbras. The new normalization condition is x x
new
=
x e i γ e−i γ x
old
= e i (γ −γ
)
x x
old
= e i (γ −γ ) δ(x − x ) = δ(x − x ).
41
(3.50)
3.4 Schr¨dinger’s representation  I o
3 POSITION AND MOMENTUM
Thus, the new eigenbras satisfy the same normalization condition as the old eigenbras. By deﬁnition, the standard ket satisﬁes x  = 1. It follows from Eq. (3.49) that the new standard ket is related to the old standard ket via
new
= e−i γ
old ,
(3.51)
where γ ≡ γ(x) is a real function of the position operator x. The dual of the above equation yields the transformation rule for the standard bra,
new = old e iγ
.
(3.52)
The transformation rule for a general operator A follows from Eqs. (3.51) and (3.52), plus the requirement that the triple product A remain invariant (this must be the case, otherwise the probability of a measurement yielding a certain result would depend on the choice of eigenbras). Thus, Anew = e−i γ Aold e i γ . (3.53)
Of course, if A commutes with x then A is invariant under the transformation. In fact, d/dx is the only operator (we know of) which does not commute with x, so Eq. (3.53) yields d dγ d d +i , (3.54) = e−i γ e i γ = dx new dx dx dx where the subscript “old” is taken as read. It follows, from Eq. (3.48), that the momentum operator p can be written p = −i ¯ h Thus, the special choice ¯ γ(x) = h yields p = −i ¯ h d dx .
new
d dx
−¯ h
new x
dγ + f(x). dx
(3.55)
f(x) dx
(3.56)
(3.57)
42
3.5 Schr¨dinger’s representation  II o
3 POSITION AND MOMENTUM
Equation (3.56) ﬁxes γ to within an arbitrary additive constant: i.e., the special eigenkets and eigenbras for which Eq. (3.57) is true are determined to within an arbitrary common phasefactor. In conclusion, it is possible to ﬁnd a set of basis eigenkets and eigenbras of the position operator x which satisfy the normalization condition (3.26), and for which the momentum conjugate to x can be represented as the operator d . (3.58) dx A general state ket is written ψ(x) , where the standard ket satisﬁes x  = 1, and where ψ(x ) = x ψ(x) is the wavefunction. This scheme of things is known as Schr¨dinger’s representation, and is the basis of wave mechanics. o p = −i ¯ h
3.5 Schr¨ dinger’s representation  II o o In the preceding sections, we have developed Schr¨dinger’s representation for the case of a single operator x corresponding to a classical Cartesian coordinate. However, this scheme can easily be extended. Consider a system with N generalized coordinates, q1 · · · qN , which can all be simultaneously measured. These are represented as N commuting operators, q1 · · · qN , each with a continuous range of eigenvalues, q1 · · · qN . Ket space is conveniently spanned by the simultaneous eigenkets of q1 · · · qN , which are denoted q1 · · · qN . These eigenkets must form a complete set, otherwise the q1 · · · qN would not be simultaneously observable. The orthogonality condition for the eigenkets [i.e., the generalization of Eq. (3.26)] is The completeness condition [i.e., the generalization of Eq. (3.27)] is
+∞ −∞ +∞
q1 · · · qN q1 · · · qN = δ(q1 − q1 ) δ(q2 − q2 ) · · · δ(qN − qN ). ··· dq1 · · · dqN q1 · · · qN q1 · · · qN  = 1.
(3.59)
(3.60)
−∞
The standard ket is deﬁned such that q1 · · · qN  = 1.
43
(3.61)
∂qi ∂qi (3.42)] φ ∂ ∂φ =− . The associated wavefunction is ψ(q1 · · · qN ) = q1 · · · qN ψ .46).3.69)
Corresponding to Eq.67)
The N linear operators ∂/∂qi (where i runs from 1 to N) are deﬁned ∂ψ ∂ ψ = . ∂qi ∂qi (3.70)
. dq1 · · · dqN ) = ψ(q1 · · · qN )2 dq1 · · · dqN . ∂qi ∂qi
44
(3. (3.
(3. Finally..II o
3 POSITION AND MOMENTUM
The standard bra is the dual of the standard ket.64) (3. where The probability of an observation of the system ﬁnding the ﬁrst coordinate in the range q1 to q1 + dq1 . the normalization condition for a physical wavefunction is
+∞ −∞ +∞
(3.63)
(3. (3. we can derive the commutation relations ∂ ∂ qj − q j = δij .68)
These linear operators can also act on bras (provided the associated wavefunctions are square integrable) in accordance with [see Eq.66)
···
−∞
ψ(q1 · · · qN )2 dq1 · · · dqN = 1.5 Schr¨dinger’s representation .62)
(3.65)
φ(q1 · · · qN ) = φq1 · · · qN . A general state ket is written ψ(q1 · · · qN ) . etc.
(3. Likewise. the second coordinate in the range q2 to q2 + dq2 . a general state bra is written φ(q1 · · · qN ). is P(q1 · · · qN .
61). ∂qi (3.75)
Thus.71)
It can be seen.4. The most general conclusion we can draw from this coincidence of commutation relations is (see Dirac) pi = −i ¯ h ∂F(q1 · · · qN ) ∂ + . we can always construct a set of simultaneous eigenkets of q1 · · · qN for which pi = −i ¯ h ∂ . ∂qi 1 (3. that the linear operators −i ¯ ∂/∂qi satisfy the same commutation relations with the q’s and with h each other that the p’s do. by comparison with Eqs. q1 · · · q N  so that h q1 · · · qN pi = −i ¯
45
∂ψ ∂ψ(q1 · · · qN ) ∂ ∂ ψ = q1 · · · qN  = = q · · · qN ψ . (3. and Dirac). Thus.74)
This is the generalized Schr¨dinger representation. Note that q1 · · · q N  Hence.72) (3.73)
However.68).76) ∂qi ∂qi ∂qi ∂qi 1 ∂ ∂ = q · · · qN .3. ∂qi ∂qj ∂qi ∂qj ∂qj ∂qi showing that ∂ ∂ ∂ ∂ = . ∂qi ∂qj ∂qj ∂qi (3.23)–(3. 3. the standard ket in Schr¨dinger’s representation is a simultaneous eigenket o of all the momentum operators belonging to the eigenvalue zero.77)
(3. o It follows from Eqs.74) that pi = 0. and (3.5 Schr¨dinger’s representation . (3. the function F can be transformed away via a suitable readjustment of the phases of the basis eigenkets (see Sect.78)
.25). ∂qi ∂qi (3. (3.II o
3 POSITION AND MOMENTUM
It is also clear that ∂2 ψ ∂ ∂ ∂ ∂ ψ = = ψ. (3. (3. ∂qi ∂qi 1 ∂ q · · · qN .
84) The probability that a measurement of the momentum yields a result lying in the range p to p + dp is given by P(p .
(3.83) (3. describable in terms of a coordinate x and its conjugate momentum p. (3. The orthogonality relation for the momentum eigenkets is p p
+∞
∂ q · · · qN .85)
.6 The momentum representation Consider a system with one degree of freedom.
46
(3. However.81)
A general state ket can be written φ(p) where the standard ket satisﬁes p  = 1.
(3. both of which have a continuous range of eigenvalues.6 The momentum representation
3 POSITION AND MOMENTUM
The dual of the above equation gives pi q1 · · · qN = i ¯ h 3. it is o also possible to represent the system in terms of the eigenkets of p. dp ) = φ(p )2 dp . This is termed Schr¨dinger’s representation. We have seen that it is possible to represent the system in terms of the eigenkets of x.80)
and the corresponding completeness relation is dp p
−∞
p  = 1. (3. The momentum space wavefunction φ(p ) sato isﬁes φ(p ) = p φ . These are denoted p .79)
= δ(p − p ). Consider the eigenkets of p which belong to the eigenvalues p .82)
Note that the standard ket in this representation is quite different to that in Schr¨dinger’s representation.3. ∂qi 1
(3.
3.6 The momentum representation
3 POSITION AND MOMENTUM
Finally, the normalization condition for a physical momentum space wavefunction is +∞ φ(p )2 dp = 1.
−∞
(3.86)
The fundamental commutation relations (3.23)–(3.25) exhibit a particular symmetry between coordinates and their conjugate momenta. If all the coordinates are transformed into their conjugate momenta, and vice versa, and i is then replaced by −i, the commutation relations are unchanged. It follows from this symmetry that we can always choose the eigenkets of p in such a manner that the coordinate x can be represented as (see Sect. 3.4) x = i¯ h d . dp (3.87)
This is termed the momentum representation. The above result is easily generalized to a system with more than one degree of freedom. Suppose the system is speciﬁed by N coordinates, q1 · · · qN , and N conjugate momenta, p1 · · · pN . Then, in the momentum representation, the coordinates can be written as ∂ qi = i ¯ h . (3.88) ∂pi We also have qi = 0, (3.89) and h p1 · · · pN qi = i ¯ ∂ p · · · pN . ∂pi 1 (3.90)
The momentum representation is less useful than Schr¨dinger’s representao tion for a very simple reason. The energy operator (i.e., the Hamiltonian) of most simple systems takes the form of a sum of quadratic terms in the momenta (i.e., the kinetic energy) plus a complicated function of the coordinates (i.e., the potential energy). In Schr¨dinger’s representation, the eigenvalue problem for o the energy translates into a secondorder differential equation in the coordinates, with a complicated potential function. In the momentum representation, the
47
3.7 The uncertainty relation
3 POSITION AND MOMENTUM
problem transforms into a highorder differential equation in the momenta, with a quadratic potential. With the mathematical tools at our disposal, we are far better able to solve the former type of problem than the latter. Hence, Schr¨dinger’s o representation is generally more useful than the momentum representation.
3.7 The uncertainty relation How is a momentum space wavefunction related to the corresponding coordinate space wavefunction? To answer this question, let us consider the representative x p of the momentum eigenkets p in Schr¨dinger’s representation for o a system with a single degree of freedom. This representative satisﬁes p x p = x pp = −i ¯ h d x p , dx (3.91)
where use has been made of Eq. (3.78) (for the case of a system with one degree of freedom). The solution of the above differential equation is x p = c exp(i p x /¯ ), h where c = c (p ). It is easily demonstrated that
+∞
(3.92)
p p
=
−∞
p x dx x p
=c c
∗
−∞
∞
exp[−i (p − p ) x /¯ ] dx . h
(3.93)
The wellknown mathematical result
+∞
exp(i a x) dx = 2π δ(a),
−∞
(3.94)
yields p p = c 2 h δ(p − p ). (3.95) This is consistent with Eq. (3.80), provided that c = h−1/2 . Thus, x p = h−1/2 exp(i p x /¯ ). h (3.96)
48
3.7 The uncertainty relation
3 POSITION AND MOMENTUM
Consider a general state ket A whose coordinate wavefunction is ψ(x ), and whose momentum wavefunction is Ψ(p ). In other words, ψ(x ) = Ψ(p ) = It is easily demonstrated that
+∞
x A , p A .
(3.97) (3.98)
ψ(x ) =
−∞
dp x p
+∞
p A (3.99)
1 = 1/2 h and
+∞
Ψ(p ) exp(i p x /¯ ) dp h
−∞
Ψ(p ) =
−∞
dx p x
+∞
x A (3.100)
1 = 1/2 h
ψ(x ) exp(−i p x /¯ ) dx , h
−∞
where use has been made of Eqs. (3.27), (3.81), (3.94), and (3.96). Clearly, the momentum space wavefunction is the Fourier transform of the coordinate space wavefunction. Consider a state whose coordinate space wavefunction is a wavepacket. In other words, the wavefunction only has nonnegligible amplitude in some spatially localized region of extent ∆x. As is wellknow, the Fourier transform of a wavepacket ﬁlls up a wavenumber band of approximate extent δk ∼ 1/∆x. Note that in Eq. (3.99) the role of the wavenumber k is played by the quantity p /¯ . It h follows that the momentum space wavefunction corresponding to a wavepacket in coordinate space extends over a range of momenta ∆p ∼ ¯ /∆x. Clearly, a meah surement of x is almost certain to give a result lying in a range of width ∆x. Likewise, measurement of p is almost certain to yield a result lying in a range of width ∆p. The product of these two uncertainties is ∆x ∆p ∼ ¯ . h This result is called Heisenberg’s uncertainty principle.
49
(3.101)
if the system is subject to an external potential. However.8 Displacement operators Consider a system with one degree of freedom corresponding to the Cartesian coordinate x.104)
.. We obtain (∆x)
2
(∆p)
2
¯2 h ≥ 4
(3.
3. The situation is not so clear with state kets. correspond to Gaussian wavepackets in both coordinate and momentum space. for which the equality sign holds in the above relation.102)
for any general state. The ﬁnal state of the system is completely determined by its initial state.103)
in the undisplaced system. (2. then in the displaced system we have Rd = Ad + Bd . the ﬁnal ket is still not completely determined. together with the direction and magnitude of the displacement. The ﬁnal state of the system only determines the direction of the displaced state ket. It is easily demonstrated that the minimum uncertainty states. if R = A + B (3.3.8 Displacement operators
3 POSITION AND MOMENTUM
Actually. Even if we adopt the convention that all state kets have unit norms.47). Suppose that we displace this system some distance along the xaxis.83) and the commutation relation (3. This follows because the superposition relations have a physical signiﬁcance which is unaffected by a displacement of the system. it is possible to write Heisenberg’s uncertainty principle more exactly by making use of Eq. Note that the type of displacement we are considering is one in which everything to do with the system is displaced. then the potential must be displaced. etc. and the displacement causes ket R to transform to ket Rd . So. and we just give it a little push. since it can be multiplied by a constant phasefactor.
50
(3. Thus. we know that the superposition relations between states remain invariant under the displacement. We could imagine that the system is on wheels.
Note that the above relation implies that A = D† Ad . The displaced kets cannot be multiplied by individual phasefactors. The normalization condition AA = 1 (3. the displacement operator is unitary.106)
for a state ket A certainly has a physical signiﬁcance.103) holds in the undisplaced system. Thus. because this would wreck the superposition relations. (3. (3.110) The equation v A = B . this determines the displaced kets to within a single arbitrary phasefactor to be multiplied into all of them.111)
.108) (3. so AD† DA = 1. Since this must hold for any state ket A . kets. Now. it follows that the displaced ket Rd must be the result of some linear operator acting on the undisplaced ket R . In other words. (3. (3. The arbitrary phasefactor by which all displaced kets may be multiplied results in D being undetermined to an arbitrary multiplicative constant of modulus unity. Since Eq. We now adopt the ansatz that any combination of bras.3. and dynamical variables which possesses a physical signiﬁcance is invariant under a displacement of the system. Rd = DR . we must have AdAd = 1.8 Displacement operators
3 POSITION AND MOMENTUM
Incidentally. (3. Ad = DA and Ad = AD† .
51
(3.109) (3.104) holds in the displaced system whenever Eq.105)
where D an operator which depends only on the nature of the displacement. it follows that D† D = 1.107)
Hence.
Let
(3. (3.113) (3. has some physical signiﬁcance. as seems reasonable. We have assumed. we require that vd Ad = Bd . The fact that D can be replaced by D exp(i γ). we expect the limit Ad − A D−1 = lim A δx→0 δx→0 δx δx lim dx = lim to exist.118)
. that the system is displaced an inﬁnitesimal distance δx along the xaxis.117)
where ax is the limit of γ/δx.3.
52
(3. we have vd = D v D † . implies that d x can be replaced by D exp(i γ) − 1 D − 1 + iγ = lim = dx + i a x . Since this is true for any ket Ad .114). We expect that the displaced ket Ad should approach the undisplaced ket A in the limit as δx → 0. where vd is the displaced operator. that γ tends to zero as δx → 0. (3.114) (3. For small δx. It is clear that the displacement operator is undetermined to an arbitrary imaginary additive constant. now. δx→0 δx→0 δx δx lim (3. Thus. It follows that vd Ad = DB = D v A = D v D† Ad .8 Displacement operators
3 POSITION AND MOMENTUM
where the operator v represents a dynamical variable.109) and (3. Suppose. we have D = 1 + δx dx .112)
Note that the arbitrary numerical factor in D does not affect either of the results (3.116) δx→0 δx where dx is denoted the displacement operator along the xaxis.115)
D−1 . Thus. where γ is a real phaseangle.
(3. we ﬁnd that vd = (1 + δx dx ) v (1 − δx dx ) = v + δx (dx v − v dx ). (3.123)
(3. Substituting into Eq.8 Displacement operators
3 POSITION AND MOMENTUM
It follows from Eq.
(3. The most general conclusion we can draw from this observation is that px = i ¯ dx + f(x). the h momentum conjugate to x.122)
Let us consider a speciﬁc example. (3. the same shape shifted in the xdirection by a distance δx). h (3.3. Actually. does [see Eq. which implies vd − v = dx v − v d x . we obtain dx† + dx = 0. However. A comparison with Eq. If the system is displaced a distance δx along the xaxis then the new wavefunction is ψ(x − δx) (i. Neglecting order (δx)2 .124)
It follows that i ¯ dx obeys the same commutation relation with x that px .125)
where f is Hermitian (since px is Hermitian).126)
. the displacement operator is antiHermitian. It can be seen that the new wavefunction is obtained from the old wavefunction according to the prescription x → x − δx. xd = x − δx. Thus.121)
(3. using x = v.e.119)
Thus. δx→0 δx lim (3.120) (3. the new wavefunction can be multiplied by an arbitrary number of modulus unity. yields dx x − x dx = −1. (3.25)]. Suppose that a state has a wavefunction ψ(x ). leaving px = i ¯ d x .114)..109) that (1 + δx dx† )(1 + δx dx ) = 1. (3. h
53
(3.122). and again neglecting order (δx)2 . the fact that dx is undetermined to an arbitrary additive imaginary constant (which could be a function of x) enables us to transform the function f out of the above equation.
The fact that translations in independent directions commute is clearly associated with the fact that the conjugate momentum operators associated with these directions also commute [see Eqs. (3. if the system is moved ∆x along the xaxis.and zaxes.126).3. We can also construct displacement operators which translate the system along the y. Thus.8 Displacement operators
3 POSITION AND MOMENTUM
Thus. the operator D(∆x) which translates the system a distance ∆x along the xaxis is written D(∆x) = lim ∆x px 1−i N ¯ h
N
N→∞
. (3. (3. In other words.24) and (3. A ﬁnite translation along the xaxis can be constructed from a series of very many inﬁnitesimal translations.128)]. h The unitary nature of the operator is now clearly apparent.118) and (3.128)
54
. It follows that D(∆x) = exp (−i px ∆x/¯ ) .
(3. the displacement operator in the xdirection is proportional to the momentum conjugate to x. Note that a displacement a distance ∆x along the xaxis commutes with a displacement a distance ∆y along the yaxis. We say that px is the generator of translations along the xaxis. and then ∆x along the xaxis. and then ∆y along the yaxis then it ends up in the same state as if it were moved ∆y along the yaxis.127)
where use has been made of Eqs.
the ﬁnal ket can be regarded as the result of some linear operator acting on the initial ket: .e. the ﬁnal ket Rt depends linearly on the initial ket Rt0 . However. assuming that the system is left undisturbed between times t0 and t. the ﬁnal state only determines the direction of the ﬁnal state ket. The label t is needed to distinguish the different states of the system at different times. At time t the state of the system is represented by the ket At . (4.3)
. we have only considered systems at one particular instant of time..
55
(4.4 QUANTUM DYNAMICS
4
Quantum dynamics
4. Even if we adopt the convention that all state kets have unit norms. since this would invalidate the superposition relation at later times. if Rt0 = At0 + Bt0 for any three kets.1) and (4.2) (4. i. The evolved kets cannot be multiplied by individual phasefactors. then we should have Rt = At + Bt . According to Eqs. since it can be multiplied by an arbitrary phasefactor. The ﬁnal state of the system at time t is completely determined by its initial state at time t0 plus the time interval t − t0 (assuming that the system is left undisturbed during this time interval). we expect that if a superposition relation holds for certain states at time t0 then the same relation should hold between the corresponding timeevolved states at time t. (4. However. Consider a system in a state A which evolves in time.2).1)
This rule determines the timeevolved kets to within a single arbitrary phasefactor to be multiplied into all of them. The label A is needed to distinguish the ket from any other ket (Bt . the ﬁnal ket is still not completely determined. In other words. Thus. say) which is evolving in time. Let us now investigate how quantum mechanical systems evolve with time. Rt = T Rt0 .1 Schr¨ dinger’s equations of motion o Up to now.
However. that as t → t0 then At → At0 for any ket A.4. Hence.4)
(4. We expect. This is always possible.6) lim t→t0 t − t0 t→t0 t − t0
56
(4.5)
. according to Eq.. Thus. The arbitrary phasefactor by which all time evolved kets may be multiplied results in T (t. time evolution has no meaning unless something observable changes with time). Since we are already committed to evolving state kets. Up to now. there are some important differences between time evolution and spatial displacement. since the length of a ket possesses no physical signiﬁcance. we require that At0 At0 = AtAt for any ket A. t0 ) being undetermined to an arbitrary multiplicative constant of modulus unity. let us assume that the time evolution operator T can be chosen in such a manner that the operators representing the dynamical variables of the system do not evolve in time (unless they contain some speciﬁc time dependence). The triple product AξA can evolve either because the ket A evolves and the operator ξ stays constant.3). if a ket is properly normalized at time t then it will remain normalized at all subsequent times t > t0 ). the time evolution operator T is a unitary operator.e. Since we have adopted a convention in which the norm of any state ket is unity. from physical continuity. the ket A stays constant and the operator ξ evolves. (4. the time evolution operator T looks very much like the spatial displacement operator D introduced in the previous section. or both the ket A and the operator ξ evolve. the limit T −1 At − At0 = lim At0 (4. we do expect the expectation value of some observable ξ to evolve with time. which immediately yields T † T = 1. Thus.1 Schr¨dinger’s equations of motion o
4 QUANTUM DYNAMICS
where T is a linear operator which depends only on the times t and t0 . even if the system is left in a state of undisturbed motion (after all. it make sense to deﬁne the time evolution operator T in such a manner that it preserves the length of any ket upon which it acts (i. In general.
(4. that if the operator D(x.7) that i¯ h dAt0 At − At0 = i ¯ lim h = i ¯ τ(t0 )At0 = H(t0 )At0 .11)
where px is the operator representing the momentum conjugate to x. x0 ) − 1 . This equation is denoted Schr¨dinger’s equation of motion.9)
When written for general t this equation becomes i¯ h dAt = H(t)At . It follows from Eqs.8)
The fact that T can be replaced by T exp(i γ) (where γ is real) implies that τ is undetermined to an arbitrary imaginary additive constant (see previous section).4. h t→t0 dt0 t − t0 (4.e. (4.12)
. We saw. τ† + τ = 0. x0 ) displaces the system along the xaxis from x0 to x then px = i ¯ x→x h lim D(x. It involves a o Hermitian operator H(t) which is. a characteristic of the dynamical system under investigation. Let T (t.10) gives the general law for the time evolution of a state ket in a scheme in which the operators representing the dynamical variables remain ﬁxed.5) that τ is antiHermitian: i. t→t0 t − t0
(4. t0 ) − 1 .1 Schr¨dinger’s equations of motion o
4 QUANTUM DYNAMICS
should exist. (4.10)
Equation (4. t0 ) evolves the system in time from t0 to t then H(t0 ) = i ¯ lim h
57
T (t.6) and (4.. (4. in the previous section. We now have that if the operator T (t. dt (4. Let us deﬁne the Hermitian operator H(t0 ) = i ¯ τ. 0 x − x0 (4. t0 ) − 1 . Note that this limit is simply the derivative of At 0 with respect to t0 .7) τ(t0 ) = lim t→t0 t − t0 It is easily demonstrated from Eq. This operator is undetermined h to an arbitrary real additive constant. presumably.
time is not an operator (we cannot observe time. but time is only a label. the dynamical variable corresponding to the operator H stands to time t as the momentum px stands to the coordinate x. Likewise.5) and (4.. the zero of potential energy is not welldeﬁned). it is just a parameter (or. (4. a continuous label). Since we are only dealing with nonrelativistic quantum mechanics. the fact that position is an operator. Note that. in the above analysis. In relativistic quantum mechanics. in classical physics.e. Substituting At = T At0 into Eq. t0 ) = exp −i
H(t ) dt /¯ . h
(4.1 Schr¨dinger’s equations of motion o
4 QUANTUM DYNAMICS
Thus.
58
. (Here. more accurately. dt (4.13)
Since this must hold for any initial state At0 we conclude that i¯ h dT = H(t) T.15)
where use has been made of Eqs.) The operator H(t) is usually called the Hamiltonian of the system. we assume that Hamiltonian operators evaluated at different times commute with one another).14)
This equation can be integrated to give T (t. The fact that the Hamiltonian is undetermined to an arbitrary real additive constant is related to the wellknown phenomenon that energy is undetermined to an arbitrary additive constant in physics (i.6). time and space coordinates are treated on the same footing by relegating position from being an operator to being just a label. this suggests that H(t) is the operator representing the total energy of the system. By analogy with classical physics. It is now clear how the fact that H is undetermined to an arbitrary real additive constant leaves T undetermined to a phasefactor. if the equations of motion of a system are invariant under an xdisplacement of the system then this implies that the system conserves momentum in the xdirection. if the equations of motion are invariant under a temporal displacement then this implies that the system conserves energy. (Recall that. need not worry us unduly. as such).4. dt
t t0
(4. (4.10) yields i¯ h dT At0 = H(t) T At0 .
16)
This is a timedependent transformation.5).16) yields At  = AT. t0 ) = 1. Thus.16) leads us to a scenario in which the state of the system is represented by a ﬁxed vector. (4.16) has the effect of bringing all kets representing states of undisturbed motion of the system to rest. since the operator T (t.17)
where use has been made of Eqs. (4. o
59
.2 Heisenberg’s equations of motion We have seen that in Schr¨dinger’s scheme the dynamical variables of the system o remain ﬁxed during a period of undisturbed motion. which is outlined in Sect.3). The dual of Eq. the transformation (4. It is easily seen that vt = T † v T. (4. this is not the only way in which to represent the time evolution of the system. The subscript t is used to remind us that the transformation is timedependent.18)
The transformation rule for a general observable v is obtained from the requirement that the expectation value AvA should remain invariant. a dynamical variable. (4. This is termed the Heisenberg picture. Suppose that a general state ket A is subject to the transformation At = T † (t. as opposed to the Schr¨dinger picture. (4. t0 )At = T † (t.19) Thus.1. However. t0 ) T (t. (4. t0 )At0 = At t0 .2 Heisenberg’s equations of motion
4 QUANTUM DYNAMICS
4. corresponds to a moving linear operator in this new scheme. The transformation must also be applied to bras. 4. Clearly.10). t 0 ) obviously depends on time. and the fact that T (t0 . (4.4. whereas the state kets evolve according to Eq. The time evolution of the transformed state ket is given by At t = T † (t. which corresponds to a ﬁxed linear operator in Schr¨o dinger’s scheme. and the dynamical variables are represented by moving linear operators. It is clear that the transformation (4. t0 )A . (4. the transformed state ket does not evolve in time.
(3. Let us compare this equation with the classical time evolution equation for a general dynamical variable v. dt (4. Equation (4. Ht ]. the corresponding variables in the Schr¨dinger picture).26) dt where [· · ·]quantum denotes the quantum Poisson bracket. which do o not evolve in time.22). (4. this reduces to H T vt + i ¯ T h or i¯ h where Ht = T † H T. Differentiation with respect to time yields dvt dT dT vt + T =v . According to Eq.14).21) (4. According to Eq. which can be written in the form [see Eq.7)] dv = [v. (4.e.23) can be written i¯ h dvt = [vt .24) dvt = v H T. It is denoted Heisenberg’s equation of motion.19).25) (4. dt (4. dt dt dt With the help of Eq.20)
(4.23)
dvt = T † v H T − T † H T v t = v t H t − H t vt .27)
. dt
Equation (4.2 Heisenberg’s equations of motion
4 QUANTUM DYNAMICS
Consider a dynamical variable v corresponding to a ﬁxed linear operator in the Schr¨dinger picture. (3.22) (4. (4. H]classical . Ht ]quantum . the Heisenberg equation of motion can be written dvt = [vt . we can write o T vt = v T. Note that the timevarying dynamical variables in the Heisenberg picture are usually called Heisenberg dynamical variables to distinguish them from Schr¨dinger dynamical o variables (i.. dt
60
(4.25) shows how the dynamical variables of the system evolve in the Heisenberg picture.4.
19) that vt = v. (4. hence. if the energy of the system has no explicit timedependence then it is represented by the same nontimevarying operator H in both the Schr¨dinger and o Heisenberg pictures. For a physical system which possess a classical analogue. with the time evolution operator T ).30) dt Thus. Only those observables which do not commute with the Hamiltonian evolve in time in the Heisenberg picture. Eq. Note that if the Hamiltonian does not explicitly depend on time (i. it is represented by the same ﬁxed operator in both the Schr¨dinger o and Heisenberg pictures). so Ht = T † H T = H.3 Ehrenfest’s theorem We have now derived all of the basic elements of quantum mechanics. (4. we
61
.. (4.31) dt Thus. i¯ h Suppose that v is an observable which commutes with the Hamiltonian (and. t0 ) = exp [−i H (t − t0 )/¯ ] . H] = 0.27) provides us with further justiﬁcation for our identiﬁcation of the linear operator H with the energy of the system in quantum mechanics. (4. H] = 0. (4. (4.15) yields T (t. It follows from Eq. and H denotes the classical Hamiltonian. The strong resemblance between Eqs. [· · ·]classical is the classical Poisson bracket. The only thing which is lacking is some rule to determine the form of the quantum mechanical Hamiltonian.25) gives dH = [H.29) (4. any observable which commutes with the Hamiltonian is a constant of the motion (hence.e. Furthermore.3 Ehrenfest’s theorem
4 QUANTUM DYNAMICS
Here.28)
4.4. the system is not subject to some timedependent external force) then Eq. Heisenberg’s equation of motion yields dv i¯ h = [v.26) and (4. (4. h This operator manifestly commutes with H.
For instance. (4. are easily proved using the fundamental commutation relations Eq. p3 ) as p.35) H= pi 2 .32) It is helpful to denote (x1 . Whenever an ambiguity arises because of noncommuting observables. This scheme guarantees that quantum mechanics yields the correct classical equations of motion in the classical limit. 2 m 2 m i=1
62
.25)] [xi .4. we would write the quantum mechanical analogue of the classical product x p. ∂F . p2 . as the Hermitian product (1/2)(x p + p x). x2 . this can usually be resolved by requiring the Hamiltonian H to be an Hermitian operator.34)
where F and G are functions which can be expanded as power series.32). When the system in question has no classical analogue then we are reduced to guessing a form for H which reproduces the observed behaviour of the system. h (4. The following useful formulae. respectively.e. we replace the classical coordinates and conjugate momenta by the corresponding quantum mechanical operators). ∂pi ∂G [pi . Consider a threedimensional system characterized by three independent Cartesian position coordinates xi (where i runs from 1 to 3). pj ] = i ¯ δij . (3.33) (4. These are represented by three commuting position operators xi . Let us now consider the threedimensional motion of a free particle of mass m in the Heisenberg picture. with three corresponding conjugate momenta pi . G(x)] = −i ¯ h .3 Ehrenfest’s theorem
4 QUANTUM DYNAMICS
generally assume that the Hamiltonian has the same form as in classical physics (i. The commutation relations satisﬁed by the position and momentum operators are [see Eq. ∂xi [xi . The Hamiltonian is assumed to have the same form as in classical physics: 3 1 p2 = (4. F(p)] = i ¯ h (4. x3 ) as x and (p1 . appearing in the Hamiltonian.. and three commuting momentum operators pi .
although we will omit the subscript t for the sake of clarity. For instance. pj 2 = dt i¯ h i¯ 2m h ∂pi j=1 m m where use has been made of Eq. m m
(4.40)
Combining the above commutation relation with the uncertainty relation (2. Thus. (4. (4. Note that even though [xi (0).37)
(4. H] = 0. It follows that pi (0) t. which means that pi (t) = pi (0) at all times t (for i is 1 to 3). xj (0)] = 0. We ﬁnd that dpi 1 = [pi . (4. for a free particle the momentum operators are constants of the motion. H] = i¯ h = .41) 4 m2 This result implies that even if a particle is welllocalized at t = 0. xi (t) = xi (0) + m
3
(4. dt i¯ h (4. The time evolution of the momentum operator pi follows from Heisenberg’s equation of motion (4. the xi do not commute when evaluated at different times.33).4. xi (0)] = .3 Ehrenfest’s theorem
4 QUANTUM DYNAMICS
In the following. xi (0) = .83) yields ¯ 2 t2 h 2 2 (∆xi ) t (∆xi ) t=0 ≥ . This conclusion can also be obtained by studying the propagation of wavepackets in wave mechanics. The time evolution of the position operator xi is given by 1 1 1 ∂ pi (0) pi dxi = [xi .36)
since pi automatically commutes with any function of the momentum operators. all dynamical variables are assumed to be Heisenberg dynamical variables. pi (0) t −i ¯ t h [xi (t). its position becomes progressively more uncertain with time.39) where the position operators are evaluated at equal times.25).
63
.38)
which is analogous to the equation of motion of a classical free particle.
43) dt i¯ h ∂xi where use has been made of Eq. (4.46) only holds if x and o p are understood to be Heisenberg dynamical variables.
64
. V(x)] = − . because the xi all commute with the new term V(x) in the Hamiltonian.3 Ehrenfest’s theorem
4 QUANTUM DYNAMICS
Let us now add a potential V(x) to our free particle Hamiltonian: p2 + V(x).45) dt2 i ¯ dt h i¯ m h m dt m ∂xi In vectorial form. m 2 = dt dt (4. (4. (4. When written in terms of expectation values.4. we obtain d2 x dp m = =− dt2 dt V(x) . We can use the Heisenberg equation of motion a second time to deduce that 1 dxi 1 pi 1 dpi 1 ∂V(x) d2 x i = . In contrast. this equation becomes d2 x dp = − V(x). On the other hand. (4. (4. Heisenberg’s equation of motion gives dpi 1 ∂V(x) = [pi .44)
still holds. V is some function of the xi operators.H = .46)
This is the quantum mechanical equivalent of Newton’s second law of motion. H= 2m (4. Note that Eq.47)
This is known as Ehrenfest’s theorem. In fact.34).42)
Here.H = =− .47) has no dependence on ¯ . the result dxi pi = dt m (4. Taking the expectation values of both sides with respect to a Heisenberg state ket that does not move with time. this result is independent of whether we are using the Heisenberg or Schr¨dinger picture. the operator equation (4. it guarantees to us that the centre of a wavepacket h always moves like a classical particle.
We can also write x V(x) = V(x ) x .53)
65
.74)] ∂ pi = −i ¯ h . x2 .10) yields o i¯ h ∂ x At = x HAt . p3 ). ∂t (4. p2 . We adopt Schr¨do inger’s representation in which the momentum conjugate to the position operator xi is written [see Eq.49) (4. ∂/∂y . x3 ). x3 ). (3. The o wavefunction of the system at time t is deﬁned ψ(x . The Hamiltonian of the system is taken to be H= p2 + V(x). Here. Note that. 2m (4. ¯2 2 h p2 At = − x At . The state of the system is represented as some time evolving ket At .4 Schr¨ dinger’s waveequation o Let us now consider the motion of a particle in three dimensions in the Schr¨dinger o picture.51) ∂xi Thus. t) = x At . x2 .52) x 2m 2m where use has been made of Eq. since the position operators are ﬁxed in the Schr¨dinger picture.4 Schr¨dinger’s waveequation o
4 QUANTUM DYNAMICS
4. and the momentum operators p ≡ (p1 . we do not expect the x to evolve in time. ∂/∂z ) denotes the gradient operator written in terms of the position eigenvalues. The ﬁxed dynamical variables of the system are the position operators x ≡ (x1 .48)
Schr¨dinger’s equation of motion (4. ≡ (∂/∂x . Let x represent a simultaneous eigenket of the position operators belonging to the eigenvalues x ≡ (x1 .50)
where use has been made of the time independence of the x .4. (4.78). (3. (4. (4.
h (4. Suppose that the ket At is an eigenket of the Hamiltonian belonging to the eigenvalue H : HAt = H At .4.10) yields o i¯ h This can be integrated to give At = exp[−i H (t − t0 )/¯ ]At0 . This suggests that if the system is initially in an eigenstate of the Hamiltonian then it remains in this state for ever. (4.
(4. dt (4. (4. t).54)
which can also be written ∂ψ(x . as long as the
66
. instead of the Heisenberg picture. It just happens to give a type of equation which we know how to solve.49). (4. The direction of the vector remains ﬁxed in ket space. we have chosen to work in the Schr¨dinger picture.58) dAt = H At . (4. in which the opposite is true. We have also ﬁxed the relative phases of the x o according to Schr¨dinger’s representation. we obtain ¯2 h ∂ x At i¯ h =− ∂t 2m
2
x At + V(x ) x At . in which state kets evolve and o dynamical variables are ﬁxed.55)
This is Schr¨dinger’s famous waveequation. Finally. Combining Eqs.51) is valid. however.50).57)
Note that At only differs from At0 by a phasefactor.
(4. (4. that the waveequation is just one of many possible representations of quantum mechanics. instead of those of the momentum operators.53). t) ¯2 h i¯ h =− ∂t 2m
2
ψ(x . o Note. we have chosen to represent the system in terms of the eigenkets of the position operators.4 Schr¨dinger’s waveequation o
4 QUANTUM DYNAMICS
where V(x ) is a scalar function of the position eigenvalues. t) + V(x ) ψ(x .52). In deriving the waveequation. and is the basis of wave mechanics.56) Schr¨dinger’s equation of motion (4. so that Eq. and (4.
can easily be written in o the form of a conservation equation for the probability density: ∂ρ + · j = 0.55). t). t) = ψ(x . (4. we o obtain ¯2 2 h ψ0 (x ) + (V(x ) − E) ψ0 (x ) = 0.
Since it is conventional to set the potential at inﬁnity equal to zero.61) is sufﬁcient to uniquely specify the solution of Eq. t)2 .64)
Schr¨dinger’s timedependent waveequation.60). The probability is equal to ρ(x . the above relation implies that bound states are equivalent to negative energy states.63) is termed the probability density. (4. t) d3 x if the wavefunction is properly normalized. (4. t0 ) exp[−i H (t − t0 )/¯ ]. t0 ). (3.30).60) − 2m where ψ0 (x ) ≡ ψ(x . h (4. Recall.
(4. and E = H is the energy of the system.62)
E < lim V(x ). t) = ψ(x . so that ρ(x . (4.61) (4. (4.59)
Substituting the above relation into Schr¨dinger’s wave equation (4.55). in which the particle is conﬁned within a ﬁnite region of space. The boundary condition (4. The wavefunction of a stationary state satisﬁes ψ(x . This is Schr¨dinger’s timeindependent waveequation. Such a state is called a stationary state. t) d3 x . deﬁned by ρ(x . (4.4.4 Schr¨dinger’s waveequation o
4 QUANTUM DYNAMICS
system is undisturbed. t) d3 x = 1. A bound state solution of the o above equation. satisﬁes the boundary condition Such a solution is only possible if ψ0 (x ) → 0
x →∞
as x  → ∞. from Eq.65) ∂t
67
. that the probability of observing the particle in some volume element d3 x around position x is proportional to ρ(x . The quantity ρ(x .
71)
. ¯ h
(4. ∂t
(4. The wavefunction can always be written in the form i S(x . t) d3 x = . ∂t ¯ h Thus. if the o wavefunction starts off properly normalized.4. then it remains properly normalized at all subsequent times. t) d3 x = Im V(x ) ρ(x .68) m where p t denotes the expectation value of the momentum evaluated at time t. according to Eq. Eq. t) = ρ(x .64).65) generalizes to ∂ρ + ∂t giving ∂ 2 (4.69)
(4. to obtain ∂ ρ(x . t) d3 x . and the boundary condition ρ → 0 as x  → ∞. (4. (4.66)
We can integrate Eq. t) = − i¯ h [ψ∗ 2m ψ−( ψ∗ ) ψ] = ¯ h Im(ψ∗ m ψ). (4. In this case.4 Schr¨dinger’s waveequation o
4 QUANTUM DYNAMICS
The probability current j takes the form j(x . Clearly. Schr¨dinger’s waveequation conserves probability. ψ(x . (4. t) exp ¯ h
68
·j=
2 Im(V) ρ. however. Such a potential is often employed to model nuclear reactions in which incident particles can be absorbed by nuclei. using the divergence theorem. In particular.70) ρ(x .67)
Thus. (4. It is easily demonstrated that pt j(x . t) . In deriving Eq. (4. Suppose.65) we have. an imaginary potential can be used to account for the disappearance of a particle. assumed that the potential V(x ) is real. t) d3 x = 0. the probability current is indirectly related to the particle momentum. naturally. if Im(V) < 0 then the total probability of observing the particle anywhere in space decreases monotonically with time. Thus.65) over all space. that the potential has an imaginary component.
t). t) is the Hamiltonian operator.4 Schr¨dinger’s waveequation o
4 QUANTUM DYNAMICS
where ρ and S are both real functions. In classical mechanics. It is a good approximation to neglect the terms involving ¯ in Eq. t) =  ∂t 2m S(x .e.74) yields h − 1 ∂S(x . p. Eq. To lowest order.73) m Thus.. The interpretation of ρ as a probability density has already been given. Let us substitute Eq. the pathintegral of the Lagrangian). (4.76)
(4. (4. in the limit ¯ → 0. the probability current is locally normal to the contours of the phasefunction S. and is one of the many forms in which we can write the equations of classical mechanics.66) that . the gradient of the phase of the wavefunction determines the direction of the probability current.4. t)2 + V(x .74) provided that h ¯ h Note that. wave mechanics h reduces to classical mechanics. In particular. Thus.71) into Schr¨dinger’s timedependent waveequation. The above equation is known as the HamiltonJacobi equation.  S
69
2
S

S2 . o We obtain − 1 ¯2 h 2m
2√
j=
ρ + 2i ¯ h
√ ( ρ)·
S−
√
ρ
√ √ S2 + i ¯ ρ 2 S + ρ V h √ ∂ ρ √ ∂S = i¯ h . (4. λ= ¯ ¯ h . (4.74) − ρ ∂t ∂t
Let us treat ¯ as a small quantity.
(4. (4.77)
. S is the action (i.72)
It follows from Eq. t) = H(x . What is the interpretation of S? Note that ψ∗ ψ= √ ρ √ ( ρ) + ρ S i ρ ¯ h S. according to Eq. (4. (4. (4. (4. S.75)
where H(x.71).
The inequality (4.4 Schr¨dinger’s waveequation o
4 QUANTUM DYNAMICS
where λ is the de Broglie wavelength divided by 2π.4.
70
. This distance is usually set by the variation scalelength of the potential. (4.76) is ¯ equivalent to  λ ¯ 1. quantum mechanics reduces to classical mechanics whenever the de Broglie wavelength is small compared to the characteristic distance over which things (other than the quantum phase) vary.78) In other words.
y. y. (5. (z px − x pz )] = y [pz .2) (5.5 ANGULAR MOMENTUM
5
Angular momentum
5. h (5.1 Orbital angular momentum Consider a particle described by the Cartesian coordinates (x. Lz = x p y − y p x . pz ] = i ¯ (−y px + x py ) = i ¯ Lz . Ly .3). Consider the commutator of the operators Lx and Lz : [Lx . h h
71
(5.6)
where i and j stand for either x. (5.23)–(3. that there is no ambiguity regarding the order in which operators appear in products on the righthand sides of Eqs.4) (5. Ly = z p x − x p z . In other words. pj ] = 0. xj ] = 0. Lz ) ≡ L which represent the components of orbital angular momentum in quantum mechanics can be deﬁned in an analogous manner to the corresponding components of classical angular momentum.3)
Let us assume that the operators (Lx . z] px + x py [z.7)
.5) (5. in principle. also.25)] [xi . Note that Lx .1) (5. py . and Lz are Hermitian.1)–(5. Note. The classical deﬁnition of the orbital angular momentum of such a particle about the origin is L = r × p. z) ≡ r and their conjugate momenta (px . so they represent things which can. [xi . be measured. (3. or z. giving Lx = y p z − z p y . Ly . we are going to assume that the above equations specify the angular momentum operators in terms of the position and linear momentum operators. Ly ] = [(y pz − z py ). The fundamental commutation relations satisﬁed by the position and linear momentum operators are [see Eqs. since all of the products consist of operators which commute. pz ) ≡ p. [pi . pj ] = i ¯ δij .
9) (5. h (5. Consider the total angular momentum of the system.11). Each of these vectors satisﬁes Eq.12)
for i = j.8)–(5. L = clear from Eqs. h (5. Suppose that there are N particles in the system. Whenever we encounter three operators having these commutation relations. h [Ly .13)
N i=1 Li . we know that the dynamical variables which they represent have identical properties to those of the components of an angular momentum (which we are about to derive).j=1 (5.5. we shall assume that any three operators which satisfy the commutation relations (5. (5.14)
72
= i¯ h
i=1
Li = i ¯ L. Ly ] = i ¯ Lz .12) and (5.1 Orbital angular momentum
5 ANGULAR MOMENTUM
The cyclic permutations of the above result yield the fundamental commutation relations satisﬁed by the components of an angular momentum: [Lx . So. Lx ] = i ¯ Ly .13) that
N N N N
However. h These can be summed up more succinctly by writing L × L = i ¯ L.8) (5.11) (5. we can write Li × Lj + Lj × Li = 0. In fact. (5. Lz ] = i ¯ Lx . h [Lz . with angular momentum vectors Li (where i runs from 1 to N). since they represent different degrees of freedom of the system.
It is
L×L =
i=1
Li ×
N
Lj =
j=1 i=1
1 Li × L i + (Li × Lj + Lj × Li ) 2 i. so that Li × L i = i ¯ L i . (5. we expect the angular momentum operators belonging to different particles to commute.10) are the foundation for the whole theory of angular momentum in quantum mechanics.8)– (5. h
.10)
The three commutation relations (5.10) represent the components of an angular momentum.
Lz ] = −i ¯ (Lx Ly + Ly Lx ).15)
(5. Lz ] = [Lx2 . the sum of two or more angular momentum vectors satisﬁes the same commutation relation as a primitive angular momentum vector. (5.20) (5.8)–(5. It is convenient to deﬁne the shift operators L+ and L− : L+ = L x + i L y .17) (5.1 Orbital angular momentum
5 ANGULAR MOMENTUM
Thus.19) Since there is nothing special about the zaxis.10) is that the three components of an angular momentum vector cannot be speciﬁed (or measured) simultaneously.11). Lz ] + [Ly2 . In fact.19) that the best we can do in quantum mechanics is to specify the magnitude of an angular momentum vector along with one of its components (by convention. It is conventional to specify the zcomponent.16) (5.5. Lz ] = 0. [Lz2 . It is easily demonstrated that h [Lx2 . (5. The immediate conclusion which can be drawn from the commutation relations (5. Note that [L+ . Lz . Lz ] = +i ¯ (Lx Ly + Ly Lx ). Lz ]. h [Ly2 . Lz ] + [Lz2 . In particular. once we have speciﬁed one component. Lz ] = 0. L2 ≡ Lx2 + Ly2 + Lz2 . so [L2 . the zcomponent). the values of other two components become uncertain. we conclude that L 2 also commutes with Lx and Ly .8)–(5. The commutator of L2 and Lz is written [L2 . It is clear from Eqs. Lz ] = −¯ L+ . the total angular momentum of the system satisﬁes the commutation relation (5. L− = L x − i L y .22)
.10) and (5.21)
(5. Consider the magnitude squared of the angular momentum vector.18)
(5. h
73
(5.
m) ¯ 2 − m2 ¯ 2 l. also. m have unit norms.28) (5. h (5. However. Now.23) (5. m . mLx2 l. m) = l (l + 1). and ξ is an Hermitian operator. It is possible to write such an h equation because ¯ has the dimensions of angular momentum. m = m ¯ l. The quantum number m is deﬁned by Lz l. h [L+ .2 Eigenvalues of angular momentum
5 ANGULAR MOMENTUM
[L− . mf(l. m + l. m is the eigenvalue of Lz divided by ¯ . (5. m = [f(l. Lz ] = +¯ L− . mLx2 + Ly2 l.25) Thus.5. mLy2 l. These kets are denoted l.29) l. m) ¯ 2 l. h (5. l and m. l. m . we will show that f(l.26) without loss of generality. m = l. m = f(l. h h h l. L− ] = 2 ¯ Lz . Later on.
74
(5. mL2 − Lz2 l. m . that both shift operators commute with L2 . h Note. m) − m2 ]¯ 2 . m = = It is easily demonstrated that Aξ2 A ≥ 0.2 Eigenvalues of angular momentum Suppose that the simultaneous eigenkets of L2 and Lz are completely speciﬁed by two quantum numbers. Note that m is a h real number. mL2 − Lz2 l.30)
.24)
5. since Lz is an Hermitian operator. m l. The proof follows from the observation that Aξ2 A = Aξ† ξA = BB . where f(l. m) is some real dimensionless function of l and m.27)
where A is a general ket. (5.
(5. assuming that the l. m . We can write L2 l.
Note that Lz L+ l.m h Hence. (5. m .31). m = (L+ Lz + [Lz . m is proportional to l. We can write L+ l.22). l. according to Eq. at ﬁrst sight. m − 1 . m + 1 . (5. L+ ])l. Thus. m . the shift operator L+ does not affect the magnitude of the angular momentum of any eigenket it acts upon. m + 1 . (5. However. m = c− ¯ l. m = c+ ¯ l. it is possible to demonstrate that (5. h (5. (2. For this reason. m . The above equation implies that L+ l. m = (L+ Lz + ¯ L+ )l.32)
where use has been made of Eq. The shift operators step the value of m up and down by unity each time they operate on one of the simultaneous eigenkets of L2 and Lz . It is easily demonstrated that L2 (L+ l. (5.29) that m2 ≤ f(l.35) L− l. (5. h Using similar arguments to those given above. m ) = ¯ 2 f(l. It follows from Eqs. L+ is called a raising operator. m ).m neous eigenstate of L2 and Lz .26).21)]. m h = (m + 1) ¯ L+ l. h (5.m h (5. m). This bound is determined by the
75
.34)
where c+ is a number. plus the fact that BB ≥ 0 for a general ket B [see Eq. there is a deﬁnite upper bound to the values that m2 can take.2 Eigenvalues of angular momentum
5 ANGULAR MOMENTUM
where B = ξA .31)
Consider the effect of the shift operator L+ on the eigenket l. l.27)–(5. the eigenvalue of L2 remains unchanged. plus the fact that L2 and Lz commute. that any value of m can be obtained by applying the shift operators a sufﬁcient number of times. but the eigenvalue of Lz is increased by ¯ .33)
where use has been made of Eq. m has the same eigenvalue of L2 as the ket l. It is clear that when the operator L+ acts on a simultal. m) (L+ l. It follows that the ket L+ l. L− is called a lowering operator. It would appear.5.
It follows that there is a maximum and a minimum possible value which m can take.40) (5.45)
. Since there is no state with m > mmax . etc. mmax = 0 . (5. (5. h (5. n. mmax = mmax (mmax + 1) ¯ 2 l. h so Eq. m .39) (5. mmax − 2 . mmax = 0 . h h Thus.37) yields (L2 − Lz2 − ¯ Lz )l.5. L− L+ = Lx2 + Ly2 + i [Lx .26)]. mmax = (Lz2 + ¯ Lz )l.36) This implies that L− L+ l. Since the lowering operator does not change the eigenvalue of L2 . mmax it generates n. (5.37)
But. We can also write the above equation in the form L2 l.38) (5. namely mmax (mmax + 1). mmax ) = mmax (mmax + 1). when L− operates on n. m = ¯ 2 [l (l + 1) − m (m + 1)]. m = (L2 − Lz2 − ¯ Lz )l. However. (5. (5. m = ¯ 2 [l (l + 1) − m (m + 1)]l. h h Comparison of this equation with Eq.26) yields the result f(l. m = l (l + 1) ¯ 2 l. all of these states must correspond to the same value of f. we can give the unknown quantum number l the value mmax . mmax . Thus.2 Eigenvalues of angular momentum
5 ANGULAR MOMENTUM
eigenvalue of L2 [see Eq. l. mL− L+ l. Suppose that we attempt to raise the value of m above its maximum value mmax . m . h The above equation can be rearranged to give L2 l. we must have L+ l.43)
(5. L2 l. h It is easily seen that L− L+ l. without loss of generality. mmax − 1 . Ly ] = L2 − Lz2 − ¯ Lz . mmax = 0 . m .42)
At this stage. m = mmax (mmax + 1) ¯ 2 l. h
76
(5.44) (5.41) (5.
46)
(5. mL− l.50)].34) and (5.m tive.m .48)
where use has been made of the fact that Lx and Ly are Hermitian. The maximum value of m is denoted
77
.m+1 = [l (l + 1) − m (m + 1)].2 Eigenvalues of angular momentum
5 ANGULAR MOMENTUM
However. Consider the following: l.m h l.m The solution of the above equation is c+ = l (l + 1) − m (m + 1). m + 1 − i l. by c+ exp(i γ). m + 1Ly l. l. m + 1Ly l. mL− L+ l. m = l.m We have made the arbitrary.m l.50)
Note that c+ is undetermined to an arbitrary phasefactor [i. m + 1Lx l.47) and (5. and we still satisfy Eq. m + 1 = ¯ 2 c+ c− . h l.49).49) c− l.35).31) implies that there is a maximum and a minimum possible value of m.. (5. m .47)
= ( l. This is equivalent to choosing the relative phases of the eigenkets l. mL− ¯ c+ l. m ∗ − i l. (5. The above equation reduces to + ∗ (5. m + 1L+ l.35). mLx l.m ) with the aid of Eqs.m−1 l. choice that c+ is real and posil. l.5. m l. l. given above. m + i l. l. m )∗ (5. mLy l. m + 1 l. Equations (5. m + 1 = = = l.e. c− = (c+ )∗ = l (l + 1) − m (m − 1). (5.
∗
(5.m+1 where use has been made of Eqs.51) (5.52)
We have already seen that the inequality (5.m l.m (5. It follows that c+ c− l. but convenient. m ∗ . where γ is real. m + 1Lx l. According to Eq.m+1 = (cl.m + cl. we also know that l. we can replace l.49) can be combined to give c+ 2 = [l (l + 1) − m (m + 1)].m (5.34) and (5. (5.
m (5.5.55) (5. each rung differing from its immediate neighbours by unity. 3. using just the fundamental commutation relations (5.52) that mmin = −l. There are only two possible choices for l. where m lies in the range −l. which allows m to take the values −2.3 Rotation operators Consider a particle described by the spherical polar coordinates (r.5. respectively. 1. We have shown that L+ l..8)–(5. We have also demonstrated that the eigenvalues of L z can only take the values m ¯ . · · · l − 1. m − 1 . this implies that c− min = 0. which allows m to take the values −3/2. What is the minimum value? Suppose that we try to lower the value of m below its minimum value mmin . According to Sect. 2). We will prove in the next section that an orbital angular momentum can only take integer values of l. θ. l. we must have L− l. and Lz are Hermitian operators. belonging to h h the eigenvalues l (l + 1) ¯ 2 and m ¯ . and the bottom rung is −l. h (5. m h denote a properly normalized simultaneous eigenket of L2 and Lz . Ly . mmin = 0. plus the fact that Lx . m = = l (l + 1) − m (m + 1) ¯ l. in quantum mechanics we can
78
.35). 5.g. −1.54)
It can be seen from Eq. ϕ). Since there is no state with m < mmin . (5. l = 2. Let l. where l is an integer. (5. In summary.53) According to Eq. The top rung is l. 3/2).10). Either it is an integer (e. The classical momentum conjugate to the azimuthal angle ϕ is the zcomponent of angular momentum. m L− l. m + 1 h l (l + 1) − m (m − 1) ¯ l. −1/2.g. 1/2. We conclude that m can take a “ladder” of discrete values. or it is a halfinteger (e. we have shown that the eigenvalues of L2 ≡ Lx2 + Ly2 + Lz2 can be written l (l + 1) ¯ 2 .. l = 3/2. 0.56)
where L± = Lx ± i Ly are the socalled shift operators. (5. Lz . −l + 1. l. h or a halfinteger.3 Rotation operators
5 ANGULAR MOMENTUM
l.
8.58)
where px is the linear momentum conjugate to x. (5. We say that Lz is the generator of rotations about the zaxis. and L z takes the form ∂ Lz = −i ¯ h . which translates the system a distance ∆x along the x axis. h (5. 3. The above equation implies that ∆ϕ Lz N R(∆ϕ) = lim 1 − i . (5. We were able to demonstrate in Sect.8 that px = i ¯ lim h D(δx) − 1 .57) ∂ϕ We can do this because there is nothing in Sect.62)
79
. Thus. introduced in Sect. We conclude that Lz = i ¯ lim h R(δϕ) − 1 . δx→0 δx (5. 3. There is nothing in our derivation of this result which speciﬁes that x has to be a Cartesian coordinate. (5. we can write R(δϕ) = 1 − i Lz δϕ/¯ h (5.3 Rotation operators
5 ANGULAR MOMENTUM
always adopt Schr¨dinger’s representation. and φ. 3. In other words.60)
in the limit δϕ → 0.5.5 which speciﬁes that we have to use Cartesian coordinates—the representation (3. for which ket space is spanned by the o simultaneous eigenkets of the position operators r.59).61) N→∞ N ¯ h which reduces to R(∆ϕ) = exp(−i Lz ∆ϕ/¯ ).74) works for any welldeﬁned set of coordinates. Consider an operator R(∆ϕ) which rotates the system an angle ∆ϕ about the zaxis. δϕ→0 δϕ (5.59)
According to Eq. the result should apply just as well to an angular coordinate. θ. This operator is very similar to the operator D(∆x). the angular momentum operator Lz can be used to rotate the system about the zaxis by an inﬁnitesimal amount.
an l = 0 state).5.e. Suppose that the system is in an eigenstate of zero overall orbital angular momentum (i. Thus. m = exp(−i Lz 2π/¯ )l. according
80
.63) (5. (5. an m = 0 state).67)
Thus. 0 = l. m = l. R(∆ϕ) R(−∆ϕ) = 1. this state is represented by the eigenket l. from the previous section.68) (5. This follows because l = 0 implies m = 0. This implies. h Rz (∆ϕy ) = exp(−i Lz ∆ϕz /¯ ). Thus. where the eigenvalue of L 2 is l (l + 1) ¯ 2 . so we can write Rx (∆ϕx ) = exp(−i Lx ∆ϕx /¯ ). m . We know that the system is also in an eigenstate of zero orbital angular momentum about any particular axis. m = exp(−i 2 π m)l. 0 . As before. We have R(∆ϕ)l. Rx (∆ϕx ) denotes an operator which rotates the system by an angle ∆ϕx about the xaxis. orbital angular momentum can only take on integer values of the quantum numbers l and m. (5.62). Here. Clearly.64) (5.. We expect the wavefunction to h h remain unaltered if we rotate the system 2π degrees about the zaxis. 0 = exp(0)l. (5. etc. R(2π)l.69) (5. the eigenstate is invariant to rotations about the zaxis. that l must also be an integer. R(∆ϕ1 ) R(∆ϕ2 ) = R(∆ϕ1 + ∆ϕ2 ). and the eigenvalue of Lz is m ¯ .66)
We conclude that m must be an integer.. Consider the action of the rotation operator R(∆ϕ) on an eigenstate possessing zero angular momentum about the zaxis (i.70)
by analogy with Eq. h Ry (∆ϕy ) = exp(−i Ly ∆ϕy /¯ ). h (5.3 Rotation operators
5 ANGULAR MOMENTUM
Note that R(∆ϕ) has all of the properties we would expect of a rotation operator R(0) = 1. m .e.65)
Suppose that the system is in a simultaneous eigenstate of L2 and Lz . There is nothing special about the zaxis. h (5. its wavefunction must be symmetric about the zaxis.
the noncommuting nature of the angular momentum operators is a direct consequence of the fact that rotations do not commute.76)
using the Schr¨dinger representation. Transforming to standard spherical polar o coordinates.
81
(5. a zero angular momentum state is invariant to rotations about any axis.4 Eigenfunctions of orbital angular momentum
5 ANGULAR MOMENTUM
to the previous section. Note that a rotation about the xaxis does not commute with a rotation about the yaxis. 0 Ry (∆ϕy )0.75) (5. = exp(0)0. if the system is rotated an angle ∆ϕx about the xaxis. (5. 0 Rz (∆ϕz )0. 0 = 0. x = r sin θ cos ϕ. Rx (∆ϕx )0.74) (5. 0 = 0. ∂z ∂y ∂ ∂ = −i ¯ z h . Thus.
5.68)–(5.5. Thus. In other words.72) (5.73)
Clearly. 0 = exp(0)0. this implies that Ry (∆ϕy ) Rx (∆ϕx ) = Rx (∆ϕx ) Ry (∆ϕy ). and then ∆ϕy about the yaxis. and then ∆ϕx about the xaxis. (5. and we can choose the zaxis to point in any direction.4 Eigenfunctions of orbital angular momentum In Cartesian coordinates.70)]. it ends up in a different state to that obtained by rotating an angle ∆ϕy about the yaxis. 0 . 0 .77)
. Such a state must possess a spherically symmetric wavefunction. In quantum mechanics. 0 = 0. the three components of orbital angular momentum can be written Lx = −i ¯ y h Ly Lz ∂ ∂ −z . [see Eqs. ∂y ∂x (5. 0 . = exp(0)0. or Ly Lx = Lx Ly .71) (5. −x ∂x ∂z ∂ ∂ = −i ¯ x h −y .
5.4 Eigenfunctions of orbital angular momentum
5 ANGULAR MOMENTUM
y = r sin θ sin ϕ, z = r cos θ, we obtain
(5.78) (5.79)
∂ ∂ + cot θ cos ϕ (5.80) ∂θ ∂ϕ ∂ ∂ − cot θ sin ϕ Ly = −i ¯ cos ϕ h (5.81) ∂θ ∂ϕ ∂ . (5.82) Lz = −i ¯ h ∂ϕ Note that Eq. (5.82) accords with Eq. (5.57). The shift operators L± = Lx ± i Ly become ∂ ∂ L± = ±¯ exp(±i ϕ) h ± i cot θ . (5.83) ∂θ ∂ϕ Now, L2 = Lx2 + Ly2 + Lz2 = Lz2 + (L+ L− + L− L+ )/2, (5.84) Lx = i ¯ sin ϕ h so ∂ ∂ 1 ∂2 2 1 2 L = −¯ h sin θ + . sin θ ∂θ ∂θ sin2 θ ∂ϕ2
(5.85)
The eigenvalue problem for L2 takes the form L2 ψ = λ ¯ 2 ψ, h where ψ(r, θ, ϕ) is the wavefunction, and λ is a number. Let us write ψ(r, θ, ϕ) = R(r) Y(θ, ϕ). Equation (5.86) reduces to ∂ 1 ∂2 1 ∂ sin θ + Y + λ Y = 0, sin θ ∂θ ∂θ sin2 θ ∂ϕ2
(5.86) (5.87)
(5.88)
where use has been made of Eq. (5.85). As is wellknown, square integrable solutions to this equation only exist when λ takes the values l (l + 1), where l is an integer. These solutions are known as spherical harmonics, and can be written Ylm (θ, ϕ) = 2 l + 1 (l − m)! m (−1)m e i m ϕ Pl (cos ϕ), 4π (l + m)!
82
(5.89)
5.4 Eigenfunctions of orbital angular momentum
5 ANGULAR MOMENTUM
m where m is a positive integer lying in the range 0 ≤ m ≤ l. Here, Pl (ξ) is an associated Legendre function satisfying the equation m m2 m d 2 dPl m (1 − ξ ) − P + l (l + 1) Pl = 0. 2 l dξ dξ 1−ξ
(5.90)
We deﬁne Yl−m = (−1)m (Ylm )∗ , (5.91) which allows m to take the negative values −l ≤ m < 0. The spherical harmonics are orthogonal functions, and are properly normalized with respect to integration over the entire solid angle:
π 0 2π
Ylm∗ (θ, ϕ) Ylm (θ, ϕ) sin θ dθ dϕ = δll δmm .
0
(5.92)
The spherical harmonics also form a complete set for representing general functions of θ and ϕ. By deﬁnition, L2 Ylm = l (l + 1) ¯ 2 Ylm , h where l is an integer. It follows from Eqs. (5.82) and (5.89) that h Lz Ylm = m ¯ Ylm , (5.94) (5.93)
where m is an integer lying in the range −l ≤ m ≤ l. Thus, the wavefunction ψ(r, θ, ϕ) = R(r) Ylm (θ, φ), where R is a general function, has all of the expected features of the wavefunction of a simultaneous eigenstate of L2 and Lz belonging to the quantum numbers l and m. The wellknown formula
m dPl 1 mξ m m+1 = √ P Pl − dξ 1 − ξ2 l 1 − ξ2 mξ m (l + m)(l − m + 1) m−1 √ + Pl = − P 1 − ξ2 l 1 − ξ2
(5.95)
can be combined with Eqs. (5.83) and (5.89) to give h L+ Ylm = l (l + 1) − m (m + 1) ¯ Ylm+1 , h L− Ylm = l (l + 1) − m (m − 1) ¯ Ylm−1 .
83
(5.96) (5.97)
5.5 Motion in a central ﬁeld
5 ANGULAR MOMENTUM
These equations are equivalent to Eqs. (5.55)–(5.56). Note that a spherical harmonic wavefunction is symmetric about the zaxis (i.e., independent of ϕ) when√ 0 ever m = 0, and is spherically symmetric whenever l = 0 (since Y0 = 1/ 4π). In summary, by solving directly for the eigenfunctions of L2 and Lz in Schr¨do inger’s representation, we have been able to reproduce all of the results of Sect. 5.2. Nevertheless, the results of Sect. 5.2 are more general than those obtained in this section, because they still apply when the quantum number l takes on halfinteger values.
5.5 Motion in a central ﬁeld Consider a particle of mass M moving in a spherically symmetric potential. The Hamiltonian takes the form p2 + V(r). H= 2M (5.98)
Adopting Schr¨dinger’s representation, we can write p = −(i/¯ ) . Hence, o h ¯2 h H=− 2M
2
+ V(r).
(5.99)
When written in spherical polar coordinates, the above equation becomes 1 ∂ ∂ 1 ∂2 ¯2 1 ∂ 2 ∂ h + V(r). r + sin θ + H=− 2 M r2 ∂r ∂r r2 sin θ ∂θ ∂θ r2 sin2 θ ∂ϕ2
(5.100)
Comparing this equation with Eq. (5.85), we ﬁnd that ¯2 1 ∂ 2 ∂ h L2 H= − 2 r + 2 2 + V(r). 2M r ∂r ∂r ¯ r h
(5.101)
Now, we know that the three components of angular momentum commute with L2 (see Sect. 5.1). We also know, from Eqs. (5.80)–(5.82), that Lx , Ly , and Lz take the form of partial derivative operators of the angular coordinates,
84
we already know that the most general form for the wavefunction of a simultaneous eigenstate of L2 and Lz is (see previous section) ψ(r. 4. (5. (5. But.102)
According to Sect. ϕ). and H. This is as expected for a spherically symmetric potential. (5. ϕ) = R(r) Ylm (θ. the previous two equations ensure that the angular momentum L and its magnitude squared L2 are both constants of the motion. (5.101). (5. 2M r dr dr r2
This is a SturmLiouville equation for the function R(r). (5. We know.101) that all three components of the angular momentum commute with the Hamiltonian: [L. Since L2 and Lz commute with each other and the Hamiltonian.105) Substituting Eq. and making use of Eq.106) − 2 r + V(r) − E R = 0. Consider the energy eigenvalue problem H ψ = E ψ. H] = 0. it is always possible to represent the state of the system in terms of the simultaneous eigenstates of L2 .104)
where E is a number. It follows from Eq. Lz .
85
. These are the energy eigenvalues. but are independent of the quantum number m. from the general properties of this type of equation.103) (5. we obtain ¯2 1 d 2 d h l (l + 1) + (5. H] = 0. In general.2.5. the energy eigenvalues depend on the quantum number l.105) into Eq. that if R(r) is required to be wellbehaved at r = 0 and as r → ∞ then solutions only exist for a discrete set of values of E.93).5 Motion in a central ﬁeld
5 ANGULAR MOMENTUM
when written in terms of spherical polar coordinates using Schr¨dinger’s repreo sentation. It is also easily seen that L2 commutes with the Hamiltonian: [L2 . (5. θ.
− 2 r 2 2µ r dr dr r 4π 0 r
(5. (5.109) − 2 − − E P = 0. 4π 0 r
(5.113) dy2 dy y2 4π 0 ¯ 2 y h
86
−¯ 2 h . µ = me mp /(me +mp ) is the reduced mass.107)
The radial eigenfunction R(r) satisﬁes Eq. 4π 0 r 2 µ r2 (5. which is equivalent to a particle of mass µ rotating about a ﬁxed point. The ﬁrst part is the attractive Coulomb potential. with P(r) = f(y) exp(−y). dr2 2 µ r2 4π 0 r ¯ h which is the onedimensional Schr¨dinger equation for a particle of mass µ movo ing in the effective potential e2 l (l + 1) ¯ 2 h Veff (r) = − + . Equation (5.106). which takes into account the fact that the electron (of mass me ) and the proton (of mass mp ) both rotate about a common centre. (5.111)
. which can be written e2 l (l + 1) ¯2 1 d 2 d h − + − E R = 0. The above equation transforms to e2 d2 P 2 µ l (l + 1)¯ 2 h (5. Let us write the product r R(r) as the function P(r).112) Here. (5. Let a= and y = r/a. for which the potential takes the speciﬁc form V(r) = − e2 .6 Energy levels of the hydrogen atom
5 ANGULAR MOMENTUM
5.108)
Here.5. it is assumed that the energy eigenvalue E is negative.6 Energy levels of the hydrogen atom Consider a hydrogen atom. and the second part corresponds to the repulsive centrifugal force.109) transforms to d l (l + 1) 2 µ e2 a d2 −2 − + f = 0. 2µE
(5.110)
The effective potential has a simple physical interpretation.
the power law series (5. cn [n (n − 1) − l (l + 1)] = cn−1 4π 0 ¯ 2 h
(5. ψ2 = 0 at r = 0. (5.114)
Substituting this solution into Eq. The only way in which we can avoid this unphysical behaviour is
87
. whereas for an l > 0 state there is zero probability of ﬁnding the electron at the nucleus (i. Note. We conclude that f(y) → exp(2 y) as y → ∞.116). that it is only possible to obtain sensible behaviour of the wavefunction as r → 0 if l is an integer. the ratio of successive terms in the series (5. where the ﬁrst term in the series is cnmin ynmin . (5. There are two possibilities: nmin = −l or nmin = l + 1.116)
Now.5.117)
according to Eq. since ψ2 dV must be ﬁnite. we obtain cn n (n − 1) y
n n−2
− 2ny
n−1
− l (l + 1) y
n−2
2 µ e2 a n−1 y + 4π 0 ¯ 2 h
= 0. at some positive value of n.115)
Equating the coefﬁcients of yn−2 gives 2 µ e2 a 2 (n − 1) − . Note that for an l = 0 state there is a ﬁnite probability of ﬁnding the electron at the nucleus. except when l = 0). Thus. This is the same as the ratio of successive terms in the series (2 y)n .113). cn−1 n (5.114) must terminate at small n.118) n! n which converges to exp(2 y).e. The former predicts unphysical behaviour of the wavefunction at y = 0. It follows from Eq.114) is cn y 2y = ..6 Energy levels of the hydrogen atom
5 ANGULAR MOMENTUM
Let us look for a powerlaw solution of the form f(y) =
n
c n yn . also. (5. For large values of y.112) that R(r) → exp(r/a)/r as r → ∞. This is only possible if [nmin (nmin −1)−l (l+1)] = 0. otherwise f(y) behaves unphysically as y → 0. we conclude that nmin = l + 1. (5.
(5. This does not correspond to physically acceptable behaviour of the wavefunction. (5.
32π2 02 ¯ 2 h
88
l (l + 1) 2 n 1 d 2d x − + − 1 R = 0.5.6 Energy levels of the hydrogen atom
5 ANGULAR MOMENTUM
if the series (5. The restrictions on the quantum numbers are m ≤ l < n. (5. The only permissible values of the other quantum numbers are l = 0 and m = 0. this is only possible if µ e2 a = n. Thus. n is a positive integer which must exceed the quantum number l. zero angular momentum state. θ. It is clear that the wavefunction for a hydrogen atom can be written ψ(r. ϕ) = R(r/a) Ylm (θ. E=− h 32π2 02 ¯ 2 n2 (5. the Ylm are spherical harmonics. and m is an integer.121)
(5. According to the recursion relation (5. l is a nonnegative integer.114). the ground state is a spherically symmetric.119)
where the last term in the series is cn yn . x2 dx dx x2 x
(5.123)
(5.114) terminates at some maximum value of n. The ground state of hydrogen corresponds to n = 1. It follows from Eq.3 × 10−11 n meters. 2 µe and R(x) is a wellbehaved solution of the differential equation
(5. ϕ).116). Here. otherwise there would be no terms in the series (5. where n 4π 0 ¯ 2 h a= = 5.120)
Here. and can only take the values µ e4 . 4π 0 ¯ 2 h (5.124)
.111) that the energy eigenvalues are quantized. The energy of the ground state is µ e4 E0 = − = −13. n is a positive integer.6 electron volts.122)
Finally.
this does not agree with observations. Gouldsmit and Uhlenbeck (in 1925) introduced the concept of an internal. we can write S × S = i ¯ S. j = 0.5. Even worse. According to the theory of orbital angular momentum h outlined in Sects. The other quantum numbers are allowed to take the values l = 0.4 and 5. there are n = 2 states with nonzero angular momentum. This is a special property of a 1/r Coulomb potential.8)–(5. and yet have j = 0. However. Thus.7 Spin angular momentum Up to now. Note that the energy levels given in Eq.10) for the components of an angular momentum. m = 0 or l = 1. Let us denote the three components of the spin angular momentum of a particle by the Hermitian operators (Sx .123). the total angular momentum in the rest frame is nonvanishing. Thus. j is a nonnegative integer. Unfortunately. Sy . and let the eigenvalue of its total angular momentum be j (j + 1) ¯ 2 . and z.120) are independent of the quantum number l. 1. there are two possibilities. For a system consisting of two (or more) particles.
5. 5. despite the fact that l appears in the radial eigenfunction equation (5. Sz ) ≡ S. Consider an isolated system at rest. (5. h
89
(5. 0.5. We assume that these operators obey the fundamental commutation relations (5. m = −1. In order to explain this apparent discrepancy between theory and experiments.125)
. In addition to the quantized negative energy state of the hydrogen atom. there is a wealth of experimental evidence which suggests that this simplistic approach is incomplete. we have tacitly assumed that the state of a particle in quantum mechanics can be completely speciﬁed by giving the wavefunction ψ as a function of the spatial coordinates x. systems where j has halfinteger values abound in nature. purely quantum mechanical. which we have just found. because we often ﬁnd systems which appear to be structureless. y. For a system consisting of a single particle.7 Spin angular momentum
5 ANGULAR MOMENTUM
The next energy level corresponds to n = 2. angular momentum called spin. For a particle with spin. there is also a continuum of unbound positive energy states.
sz = sz ¯ s.1.128) (5. (5. Spin angular momentum operators cannot be expressed in terms of position and momentum operators. take integer or halfinteger values. the quantum number s is allowed to take halfinteger values.2. like in Eqs.127) Thus. However. S2 ] = 0.132)
90
.3 and 5. sz . where Sz s. there is one vitally important difference. the ± denote eigenkets of the Sz operator corresponding to the eigenvalues ±¯ /2.4) depends on Eqs.5. for which ¯ h Sz ± = ± ± . the restriction that the quantum number of the overall angular momentum must take integer values is lifted for spin angular momentum. since this identiﬁcation depends on an analogy with classical mechanics. Spin angular momentum clearly has many properties in common with orbital angular momentum. since this restriction (found in Sects. it has no analogy in classical physics. s − 1 · · · − s + 1.3). h (5. Consider a spin onehalf particle. h = s (s + 1) ¯ 2 s.7 Spin angular momentum
5 ANGULAR MOMENTUM
We can also deﬁne the operator S2 = Sx2 + Sy2 + Sz2 .130) 2 3¯2 h S2 ± = ± . and the concept of spin is purely quantum mechanical: i.131) 4 Here. it is possible to ﬁnd simultaneous eigenstates of S2 and Sz . (5. 5. the quantum number s can.3). sz . 5.1)–(5. in principle. sz S2 s.e. In other words. −s. (5. These kets are orthonormal (since Sz is an Hermitian operator). These are denoted s. 5. (5.. and the quantum number s z can only take the values s. (5. Consequently.1)–(5.126)
According to the equally general analysis of Sect. h so +− = 0. sz . (5. According to the quite general analysis of Sect. [S.129) (5.
− ) dx dy dz = 1. Here. y .e.
2
5. + x . y . that Hilbert subspace consisting of kets which correspond to the different spin states of the particle).8 Wavefunction of a spin onehalf particle
5 ANGULAR MOMENTUM
They are also properly normalized and complete. z . (5.133)
It is easily veriﬁed that the Hermitian operators deﬁned by ¯ h ( + − + − + ) .135) Sx = 2 i¯ h ( − + − + − + ) .134) (5. y. y . − x .10) (with the Lj replaced by the Sj ).138) constitute a realization of the spin operators S and S2 (for a spin onehalf particle) in spin space (i.139)
91
.. z . x . z .135)–(5. z . The basis kets are assumed to satisfy the completeness relation ( x . y . (5. z . satisfy the eigenvalue relations (5. (5. ± . y . Equations (5. (5. so that ++ = −− = 1.130)–(5. deﬁned in this manner. (5.138) 4 It is also easily demonstrated that S2 and Sz . Let us suppose that this space is spanned by the basis kets x . ± denotes a simultaneous eigenstate of the position operators x.136) Sy = 2 ¯ h ( + + − − − ) . h respectively. y . z .131). y . z. and the spin operator Sz .137) Sz = 2 satisfy the commutation relations (5. and ±¯ /2.5.8 Wavefunction of a spin onehalf particle The state of a spin onehalf particle is represented as a vector in ket space. and + + + − − = 1. The operator S2 takes the form 3¯2 h S = .8)–(5. corresponding to the eigenvalues x . + + x . (5. z .
+ as the product of two kets—a position space ket x . y . y . y . A general state A of a spin onehalf particle is represented as a ket A in the product of the spin and position spaces. y . z ) = x . We can give meaning to any position space operator (such as Lz ) acting on the product x . z and ± are in two quite separate vector spaces. y .142) (5. y . y . the latter space is termed the product space of the former spaces. y . ± = x . y . y . y . z + +c x . (5.143). The ket vectors x . z + x . and their product x . y . y . x . z . z . z . z ) + = + x . and commutes with x . z + c x . z . This implies that every position space operator commutes with every spin operator. z ) = ψ− (x . z  −A . but is instead a sum or integral of kets of this form. In this manner.141) (5. This state can be completely speciﬁed by two wavefunctions: ψ+ (x . z . The number of dimensions of a product space is equal to the product of the number of dimensions of each of the factor spaces. y . z ± is in a third vector space. and commutes with the + factor. (5. we can give a meaning to any spin operator (such as Sz ) acting on x . y . y . z . y .8 Wavefunction of a spin onehalf particle
5 ANGULAR MOMENTUM
It is helpful to think of the ket x . z ± = ± x . which are termed factor spaces. = c x . A general ket of the product space is not of the form (5. z + (c x . y . y . y . z + by assuming that it operates only on + . y . y .5.144) (5. z − .143)
The multiplication in the above equation is of quite a different type to any which we have encountered previously. z (c+ + + c− − ) = c+ x . z + by assuming that it operates only on the x . In mathematics.145)
.
92
(5.140)
where the c’s are numbers. We assume that such a product obeys the commutative and distributive axioms of multiplication: x . z factor. we can give meaning to the equation x . z + +c− x . and a spin space ket + . y . Similarly. z  +A . y .
(5. h h
93
(5. the ket A becomes AR = Tz (∆ϕ)A . y to y + dy . Under rotation. y . z )2 dx dy dz . forget about the spatial position of the particle. A general spin state A is represented by the ket A = +A + + −A − (5. and z to z + dz . the probability of observing the particle in the region x to x + dx . let us consider its effect on Sx . The normalization condition for the wavefunctions is ψ+ 2 + ψ− 2 dx dy dz = 1.62). y .146)
5.150)
Thus.151)
.5. with sz = −1/2 is ψ− (x . h Thus. z )2 dx dy dz .147)
in spin space. we would expect such an operator to take the form Tz (∆ϕ) = exp(−i Sz ∆ϕ/¯ ). (5.148)
To demonstrate that the operator (5.9 Rotation operators in spin space
5 ANGULAR MOMENTUM
The probability of observing the particle in the region x to x + dx . In Sect. after rotation. (5. we need to compute
exp( i Sz ∆ϕ/¯ ) Sx exp(−i Sz ∆ϕ/¯ ). Likewise. and z to z + dz . and concentrate on its spin state.9 Rotation operators in spin space Let us. Can we also construct an operator Tz (∆ϕ) which rotates the system by an angle ∆ϕ about the zaxis in spin space? By analogy with Eq. y to y + dy . for the moment.149) (5.3.
(5. 5.148) really does rotate the spin of the system. we were able to construct an operator Rz (∆ϕ) which rotates the system by an angle ∆ϕ about the zaxis in position space. with sz = +1/2 is ψ+ (x . this expectation value changes as follows:
† Sx → AR Sx AR = ATz Sx Tz A .
and is.156)
(5. where use has been made of Eqs.9 Rotation operators in spin space
5 ANGULAR MOMENTUM
This can be achieved in two different ways.151). [G.125). [G. First.152) (5. [Sz .151) becomes ¯ h exp( i Sz ∆ϕ/¯ ) ( + − + − + ) exp(−i Sz ∆ϕ/¯ ). This takes the form i 2 λ2 [G. we obtain i ∆ϕ 1 Sx + [Sz .135)–(5. A second approach is to use the so called BakerHausdorff lemma. y 2! 3!
i ∆ϕ ¯ h
2
[Sz . (5. Sx ] + ¯ h 2! which reduces to ∆ϕ2 ∆ϕ3 1 − − S ϕ − Sx + ··· + · · · .153) (5.155) exp( i G λ) A exp(−i G λ) = A + i λ[G. (5.137). 2 which reduces to Sx cos ∆ϕ − Sy sin ∆ϕ. (5. · · · [G. A]]] · · ·].125). valid for systems with spin angular momentum higher than onehalf. [G.
(5. therefore. We ﬁnd that Eq. and λ is a real parameter. The second proof is more general than the ﬁrst.5.154)
i n λn ··· + [G. A] + 2!
(5.157)
or
Sx cos ∆ϕ − Sy sin ∆ϕ.135). (5.158)
where use has been made of Eq. h h 2 or ¯ i ∆ϕ/2 h e + − e i ∆ϕ/2 + e −i ∆ϕ/2 − + e −i ∆ϕ/2 . A]] + (5. n!
where G is a Hermitian operator. Applying the BakerHausdorff lemma to Eq. Sx ]] + · · · .
(5. (5. since it only uses the fundamental commutation relation (5. The proof of this lemma is left as an exercise.
94
. we can use the explicit formula for Sx given in Eq.
both methods imply that Sx → Sx cos ∆ϕ − Sy sin ∆ϕ (5.164)
(5. Consider a rotation by 2π radians.163)
where the Jk are the generators of rotation. We ﬁnd that A → Tz (2π)A = −A .5.
l
(5.148) rotates the expectation value of S by an angle ∆ϕ about the zaxis.161)
since Sz commutes with the rotation operator.159).159)
under the action of the rotation operator (5. we have effectively demonstrated that Jk → Rkl Jl . It is clear. It is easily seen that Tz (∆ϕ)A = e−i ∆ϕ/2 +A + + e i ∆ϕ/2 −A − .148) on the state ket (5.147).160) Furthermore. and the rotation operator about the kth axis is written h Rk (∆ϕ) = exp(−i Jk ∆ϕ/¯ ).
95
(5.9 Rotation operators in spin space
5 ANGULAR MOMENTUM
For a spin onehalf system.159)–(5.
(5. It is straightforward to show that Sy → Sy cos ∆ϕ + Sx sin ∆ϕ. Equations (5. satisfying the fundamental commutation relation J × J = i ¯ J. h Consider the effect of the rotation operator (5. from our second derivation of the result (5. (5.165)
.162)
where the Rkl are the elements of the conventional rotation matrix for the rotation in question. In fact. the expectation value of the spin operator behaves like a classical vector under rotation: Sk → Rkl Sl .161) demonstrate that the operator (5. Sz → S z .148).
l
(5. In fact. that this property is not restricted to the spin operators of a spin onehalf system.
It follows that q L. µ and L. Does spin angular momentum also give rise to a contribution to the magnetic moment of a charged particle? The answer is “yes”. the minus sign does give rise to observable consequences. In fact. Thus. The charge is equivalent to a current loop of radius r in the xy plane carrying current I = q v/2π r. In fact. For an electron this ratio is found to be e2 1 1 + .168) 2m where g is called the gyromagnetic ratio. relativistic quantum mechanics actually predicts that a charged particle possessing spin should also possess a magnetic moment (this was ﬁrst demonstrated by Dirac). However. respectively. since S is sandwiched between A and A .
5.166) 2 where r and v are the vector position and velocity of the particle. This is indeed found to the the case. respectively. we can write q µ = r × v. and m is its mass. where p is the vector momentum of the particle. ge = 2 2π 4π 0 ¯ c h
(5. we expect the above relation to also hold between the quantum mechanical operators. Nevertheless.5. We can write q µ= (L + g S) . We also know that L = r × p.167) µ= 2m Using the usual analogy between classical and quantum mechanics. The magnetic moment µ of the loop is of magnitude π r2 I and is directed along the zaxis. which represent magnetic moment and orbital angular momentum. as we shall see presently.169)
96
. The minus sign does not affect the expectation value of S. where L is the orbital angular momentum. we know that p = v/m. (5. (5.10 Magnetic moments
5 ANGULAR MOMENTUM
Note that a ket rotated by 2π radians differs from the original ket by a minus sign. (5. a rotation by 4π radians is needed to transform a ket into itself.10 Magnetic moments Consider a particle of charge q and velocity v performing a circular orbit of radius r in the xy plane. both of which change sign.
97
(5. is due to quantum ﬁeld effects.176)
The time evolution of the state ket is given by analogy with Eq.148).170)
5. by comparison with Eq. It is immediately clear that the Hamiltonian (5.159)–(5.174) (5. (5. 0 + + e i ω t/2 −A. that the time evolution operator is precisely the same as the rotation operator for spin. e > 0).177)
. me According to Eq. B = B ^.164): A.
− Sy + Sx
t=0 sin ωt. is z H = −µ·B = where e S·B = ω Sz .11 Spin precession The Hamiltonian for an electron at rest in a zdirected magnetic ﬁeld.175) (5. me (5. with ∆ϕ set equal to ω t. t=0 sin ωt. (5. In fact. We shall ignore this correction in the following.
(5. (4.5. so µ for an electron (here.28). derived originally by Schwinger.172)
(5. 0) = exp(−i Ht/¯ ) = exp(−i Sz ω t/¯ ).171)
eB . the time evolution operator for this system is ω= T (t. t = e−i ω t/2 +A. Eqs. The small correction 1/(2π 137). (5.173)
It can be seen.161) imply that Sx Sy Sz
t t t
= = =
Sx Sy Sz
t=0 cos ωt t=0 cos ωt t=0 .11 Spin precession
5 ANGULAR MOMENTUM
The factor 2 is correctly predicted by Dirac’s relativistic theory of the electron. 0 − .171) causes the electron spin to precess about the zaxis with angular frequency ω. − e (L + 2 S) 2 me (5. h h
(5.
while the ﬁeldstrength B is varied.11 Spin precession
5 ANGULAR MOMENTUM
Note that it takes time t = 4π/ω for the state ket to return to its original state. It follows that the change in magnetic ﬁeld required to produce successive maxima is ∆B = 4π ¯ h . mp (5. and then recombined. path B enters a small region where a static magnetic ﬁeld is present.165). the time of ﬂight T through the magnetic ﬁeld region is kept constant. T is the time spent in the magnetic ﬁeld. We now describe an experiment to detect the minus sign in Eq. where δ is the phase difference between paths A and B in the absence of a magnetic ﬁeld. When neutrons from path A and path B meet they undergo interference.5. and λ is the de ¯ Broglie wavelength over 2π of the neutrons. The above prediction has been veriﬁed experimentally to within a fraction of a percent.
98
. By contrast. (5. However. which does not agree with the experimental data. (5. In experiments. a neutron state ket going along path B acquires a phaseshift exp( i ω T/2) (the signs correspond to sz = ±1/2 states). We expect the observed neutron intensity in the interference region to exhibit a cos(±ω T/2 + δ) variation. and ω is the spin precession frequency ω= gn e B . sent along two different paths.172).179)
where l is the pathlength through the magnetic ﬁeld region. An almost monoenergetic beam of neutrons is split in two. e gn λ l ¯ (5. This prediction depends crucially on the fact that it takes a 4π rotation to return a state ket to its original state. Path A goes through a magnetic ﬁeld free region. (The magnetic moment of a neutron is entirely a quantum ﬁeld effect). As a result.91.178)
This frequency is deﬁned in an analogous manner to Eq. If it only took a 2π rotation then ∆B would be half of the value given above. A and B. it only takes times t = 2π/ω for the spin vector to point in its original direction. The gyromagnetic ratio for a neutron is found experimentally to be gn = −1. Here.
+ (5. in Sect. that a general spin ket can be expressed as a linear combination of the two eigenkets of Sz belonging to the eigenvalues ±¯ /2. + + − − c−
99
(5.181)
The corresponding eigenbras are represented as row matrices: + → (1. A− ).184) is called a twocomponent spinor. the orbital angular momentum operators take the form of differential operators involving only angular coordinates. 1
In this scheme. 0) ≡ χ† . 1) ≡ χ† . In this representation. 0
(5. −
and a general ket becomes
A = +A + + −A − →
+A −A
. a general bra takes the form
− → (0. from Sect. and can be written χ≡
A = A+ + + A− − → ( A+ . +A −A
(5. Recall.184)
The column matrix (5.12 Pauli twocomponent formalism
5 ANGULAR MOMENTUM
5.4. 5.186)
. In this representation.180) (5. Let us represent these basis eigenkets as column matrices: + − 1 → ≡ χ+ .185)
c = + = c χ +c χ .183)
0 → ≡ χ− . 5. that the eigenstates of orbital angular momentum can be conveniently represented as spherical harmonics. These h are denoted ± .
(5. The matrix representation of a spin onehalf system was introduced by Pauli in 1926.9. the spin angular momentum operators take the form of matrices.12 Pauli twocomponent formalism We have seen.182) (5.5. It is conventional to represent the eigenstates of spin angular momentum as column (or row) matrices.
187)
Consider the ket obtained by the action of a spin operator on ket A: A = Sk A .196)
1 0 = . The row matrix (5. −Sk + +A + −Sk − −A .
(5. = =
+Sk + +A + +Sk − −A . (5.194) (5.
(5. 0 −1
100
0 −i .5.135)–(5.195) (5. can easily be evaluated using the explicit forms for the spin operators given in Eqs.188)
+A −A
≡χ.193) 2 where the σk are the matrices of the ±Sk ± values divided by ¯ /2. A− ) = (c+ . (5. This ket is represented as A → +A −A or
(5. which are called the Pauli matrices. These h matrices.
(5. We ﬁnd that σ1 σ2 σ3 0 1 = .12 Pauli twocomponent formalism
5 ANGULAR MOMENTUM
where the c± are complex numbers.137).192)
(5.185) becomes
∗ ∗ ∗ ∗ χ† ≡ ( A+ . + −
(5. = i 0
. c− ) = c+ χ† + c− χ† .191)
+A −A
It follows that we can represent the operator/ket relation (5.188) as the matrix relation ¯ h χ = σk χ. 1 0
=
+Sk + −Sk +
+Sk − −Sk −
+A −A
.190) (5.189)
However.
. y . σ1 σ2 − σ2 σ1 = 2 i σ3 . y .12 Pauli twocomponent formalism
5 ANGULAR MOMENTUM
Here.125). z ) = Consider the operator relation A = Sk A . z . 2. y. 1. (5. Recall.201)
The ket corresponding to state A is denoted A .197)
The expectation value of Sk can be written in terms of spinors and the Pauli matrices: ¯ h † Sk = ASk A = A± ±Sk ± ±A = χ σk χ. (5.
101
x . Note that. we are effectively representing the spin operators in terms of the Pauli matrices: Sk → ¯ h σk . in this scheme. x .199)
It is easily seen that the matrices (5. 2 (5.198) 2 ± The fundamental commutation relation for angular momentum.8. State A is completely speciﬁed by the two wavefunctions ψ+ (x . y . z . Eq. σj } = 2 δij .e. 5. z  +A . can be combined with (5. from Sect. (5. plus all cyclic permutations).
(5. (5.200)
Let us examine how the Pauli scheme can be extended to take into account the position of a spin onehalf particle.197) to give the following commutation relation for the Pauli matrices: σ × σ = 2 i σ. z ) = ψ− (x . z ± = ± x . ± = x . y . and z.202) (5.5.
(5.203)
(5. It is also easily seen that the Pauli matrices satisfy the anticommutation relations {σi . y . that we can represent a general basis ket as the product of basis kets in position space and spin space: x . z  −A . y . respectively. and 3 refer to x.196) actually satisfy these relations (i.194)–(5.204)
. y . and resides in the product space of the position and spin ket spaces.
(5. y . z  −A = = +Sk + x . z pk −A ∂ = −i ¯ h x . z  −A . The components of a spinor are now wavefunctions. y .210)
x .78). z  −A . = −i ¯ ∂ψ− (r )/∂xk h ψ− (r )
102
(5. y .208)
In the Schr¨dinger representation. z  −A . ∂xk
(5. z  +A . z . y . the operator equation (5. It is fairly obvious that we can represent the operator relation (5. The above equation reduces to
h ψ (r ) −i ¯ ∂ψ+ (r )/∂xk + .204) becomes simply χ = Consider the operator relation A = pk A . = −i ¯ h ∂xk x .209) ¯ h σk χ. 2
(5.206) (5. y . y . In this scheme. y . z  +A x . y . y .207) ψ− (r ) and so on. (3.204) as a matrix relation if we generalize our deﬁnition of a spinor by writing ψ+ (r ) A → ≡ χ. −Sk + x . z  +A = x .5. y . instead of complex numbers. y .12 Pauli twocomponent formalism
5 ANGULAR MOMENTUM
It is easily seen that x . z  +A + −Sk − x . (5. y . (5. z pk +A ∂ x .211)
where use has been made of Eq. z  −A
=
(5.205)
where use has been made of the fact that the spin operator Sk commutes with the eigenbras x . z  +A + +Sk − x . y . we have o x .212)
.
σk ] aj bk 2 2 (σjk + i
jkl σl ) aj bk
=
j k
= a·b + i σ·(a × b). where a ≡ (ax .12 Pauli twocomponent formalism
5 ANGULAR MOMENTUM
Thus. This quantity is represented as a 2 × 2 matrix: σ·a ≡
k
+a3 a1 − i a 2 . any position operator (e.218)
. or z . a general position operator takes the form of a differential operator in x . ay .199) and (5. σ j aj
j k
σk bk =
j k
1 1 {σj .207). the operator equation (5. p k or Lk ) is represented in the Pauli scheme as some differential operator of the position eigenvalues multiplied by the 2 × 2 unit matrix. ak σ k = a1 + i a 2 −a3
(5. σk } + [σj .200).215)
o Since..
What about combinations of position and spin operators? The most commonly occurring combination is a dot product: e. y .g. The important identity (σ·a) (σ·b) = a·b + i σ·(a × b) (5.209) can be written χ = pk χ.216)
follows from the commutation and anticommutation relations (5. az ) is some vector position operator.214) ∂xk Here.g.5. I is the 2 × 2 unit matrix. it is clear that the above quantity must be regarded as a matrix differential operator which acts on spinors of the general form (5. where pk → −i ¯ h (5. in the Schr¨dinger representation. A general rotation operator in spin space is written T (∆φ) = exp (−i S·n ∆ϕ/¯ ) .. h
103
(5.213)
∂ I.217)
(5. (5. In fact. Thus. S · L = (¯ /2) σ · L. Consider the h hybrid operator σ·a.
224)
.216).225)
.
l
(5. (5.198)]. plus the rules (σ·n)n = 1 (σ·n)n = (σ·n) for n even. n can be regarded as a trivial position operator.5. Here.219)
in the Pauli scheme. we require (χ† σk χ) ≡ (χ† ) σk χ =
104
Rkl (χ† σl χ). Thus.148). for n odd.12 Pauli twocomponent formalism
5 ANGULAR MOMENTUM
by analogy with Eq. so we would expect it to transform like a vector under rotation (see Sect. The rotation operator is represented exp (−i S·n ∆ϕ/¯ ) → exp (−i σ·n ∆ϕ/2) h (5. the quantity χ† σk χ is proportional to the expectation value of Sk [see Eq. χ = exp (−i σ·n ∆ϕ/2) χ. and ∆ϕ is the angle of rotation.9). 5. The term on the righthand side of the above expression is the exponential of a matrix. This can easily be evaluated using the Taylor series for an exponential. (5. we can write (σ·n)2 ∆ϕ 1 − exp (−i σ·n ∆ϕ/2) = 2! 2
2
(σ·n)4 ∆ϕ + 4! 2
3
4
= cos(∆ϕ/2) I − i sin(∆ϕ/2) σ·n.222)
cos(∆ϕ/2) − i nz sin(∆ϕ/2) (−i nx − ny ) sin(∆ϕ/2) (−i nx + ny ) sin(∆ϕ/2) cos(∆ϕ/2) + i nz sin(∆ϕ/2)
Rotation matrices act on spinors in much the same manner as the corresponding rotation operators act on state kets.223)
where χ denotes the spinor obtained after rotating the spinor χ an angle ∆ϕ about the naxis. In fact.220) (5. (5.
(5. (5. Thus. However.221)
These rules follow trivially from the identity (5. The explicit 2 × 2 form of this matrix is
(σ·n)3 ∆ϕ ∆ϕ (σ· n) − −i 2 3! 2
+ · · ·
+ · · ·
(5. The Pauli matrices remain unchanged under rotations. where n is a unit vector pointing along the axis of rotation.
in addition to Hermitian operators. In general. This function speciﬁes the probability density for observing the particle at a given position. we now know how to generalize this scheme to deal with a spin onehalf particle. we ﬁnd H ψ = E ψ. since exp i σ3 ∆ϕ −i σ3 ∆ϕ σ1 exp = σ1 cos ∆ϕ − σ2 sin ∆ϕ 2 2 (5.
105
. The boundary conditions (for a bound state) are obtained from the normalization constraint ψ2 dV = 1. r . The ﬁrst. we use two wavefunctions.5. (5. h H E = E E .13 Spin greater than onehalf systems
5 ANGULAR MOMENTUM
where the Rkl are the elements of a conventional rotation matrix. can be transformed into a partial differential equation for the wavefunction ψ(r ) ≡ r E . ψ+ (r ). the Hamiltonian can be written as some function of the o position and momentum operators. Instead of representing the state of the particle by a single wavefunction.
5.229)
(5. ¯ h ¯ h (5.227)
The previous two formulae can both be validated using the BakerHausdorff lemma.9) exp −i Sz ∆ϕ i Sz ∆ϕ Sx exp = Sx cos ∆ϕ − Sy sin ∆ϕ. This is easily demonstrated.230)
This is all very familiar. 5. However. the energy eigenvalue problem. in which p → −i ¯ .156).13 Spin greater than onehalf systems In the absence of spin.226)
plus all cyclic permutations. (5. Using the Schr¨dinger representation.228)
where H is now a partial differential operator. (5. The above expression is the 2 × 2 matrix analogue of (see Sect. which holds for Hermitian matrices.
The second. In fact. (5. The physical signiﬁcance of the above expression is clear. the o partial differential equation for ψ+ decouples from that for ψ− . ψ(r ) is determined by the solution of the differential equation. In this situation.13 Spin greater than onehalf systems
5 ANGULAR MOMENTUM
speciﬁes the probability density of observing the particle at position r with spin angular momentum +¯ /2 in the zdirection. the most general solution to Eq. and the spinor eigenvalue equao tion (5. a 1 × 2 matrix of wavefunctions) and H is a 2 × 2 matrix partial differential operator [see Eq. and the c± are arbitrary complex numbers.231)
where χ is a spinor (i.231) reduces to a straightforward matrix eigenvalue problem. these wavefunctions are h combined into a spinor. (5. The most
106
. speciﬁes the h probability density of observing the particle at position r with spin angular momentum −¯ /2 in the zdirection. ψ− (r ).e. (5.215)]. The above spinor equation can always be written out explicitly as two coupled partial differential equations for ψ+ and ψ− . momentum.5. Adopting the Schr¨dinger representation.231) can be written c χ = ψ(r ) + . the Hamiltonian is represented as a 2 × 2 matrix of complex numbers in the Schr¨dinger/Pauli scheme [see Eq. the Hamiltonian is a function of the position. χ. and spin operao tors.214)]. and the Pauli scheme. Suppose that the Hamiltonian depends only on the spin operators. In this case.. (5. both equations have the same form. which is simply the row vector of ψ+ and ψ− . In general. so there is only really one differential equation. In other words. the energy eigenvalue problem reduces to H χ = E χ. Suppose that the Hamiltonian has no dependence on the spin operators. (5. In the Pauli scheme. but the direction of its spin angular momentum remains undetermined. In this case.232)
Here.197)]. The Hamiltonian determines the relative probabilities of ﬁnding the particle at various different positions. the Hamiltonian is represented as diagonal 2 × 2 matrix partial differential operator in the Schr¨dinger/Pauli scheme [see Eq. c−
(5.
a spin threehalves particle by a fourcomponent spinor. the ratio c+ /c− is determined by the matrix eigenvalue problem. a particle for which the eigenvalue of S2 is s (s + 1) ¯ 2 .
107
. ψ+ and ψ− . −s + 1. Not surprisingly. In other words. Consider a spins particle: i. the Hamiltonian is a function of both position and spin operators.233).e. a spin onehalf particle is represented by a twocomponent spinor. Here. the Hamiltonian determines the direction of the particle’s spin angular momentum. In this case.5. s − 1. a general Hamiltonian causes the direction of the particle’s spin angular momentum to vary with position in some speciﬁed manner. In general. and the wavefunction ψ(r ) is arbitrary. h The eigenvalues of Sz are written sz ¯ . χ. c−
(5. This can only be represented as a spinor involving different wavefunctions. we can represent the state of the particle by 2 s + 1 different wavefunctions. (5.233)
Here. The state of the particle can be represented more succinctly by a spinor. there are 2 s + 1 distinct allowed values of s z . ψsz (r ) = r  s. it is not possible to decompose the spinor as in Eqs.234)
where A denotes a state ket in the product space of the position and spin operators. Here.13 Spin greater than onehalf systems
5 ANGULAR MOMENTUM
general solution can again be written c χ = ψ(r ) + . Thus. denoted ψsz (r ). where sz is allowed to take the values h s. Clearly.232) and (5. which is simply the 2 s + 1 component row vector of the ψsz (r ). More exactly. of course. sz A .. But. s is either an integer. a spin one particle by a threecomponent spinor. but leaves its position undetermined. ψsz (r ) speciﬁes the probability density for observing the particle at position r with spin angular momentum sz ¯ in the h zdirection. and so on. · · · . what happens if we have a spin one or a spin threehalves particle? It turns out that we can generalize the Pauli twocomponent scheme in a fairly straightforward manner. −s. or a halfinteger. In fact. (5.
240)–(5. (5.242) (5. and (5.237) Sk → s ¯ σ k . j S− s.l+1 2s
108
.236)
Here. l and. (5. h (5. We know. The matrix elements of σx and σy are most easily obtained by considering the shift operators.235)
where I is the (2 s + 1) × (2 s + 1) unit matrix. h Hence. from Eqs. We represent the spin operators as where the (2 s + 1) × (2 s + 1) extended Pauli matrix σk has elements (σk )jl = s. j = = s (s + 1) − j (j + 1) ¯ s.242).55)–(5.13 Spin greater than onehalf systems
5 ANGULAR MOMENTUM
In this extended Schr¨dinger/Pauli scheme. Thus. j + 1 . (5. (σz )jl = S± = S x ± i S y .214)] pk → −i ¯ h ∂ I. h (5. j . or halfintegers. lying in the range −s to +s. that (σx )jl = s (s + 1) − j (j − 1) δj. (5. l .239) s¯ h s where use has been made of the orthonormality property of the s. s. we can represent the momentum operators as [see Eq. construct the extended Pauli matrices? In fact.238)
It follows from Eqs. how can we evaluate the brackets s. it is trivial to construct the σz matrix.237). l are integers. j. Thus. thereby. l j = δjl .5. By deﬁnition. jSk s. j − 1 . Sz s. σ z is the suitably normalized diagonal matrix of the eigenvalues of Sz . jSz s. s¯ h (5. j .240) (5. that S+ s. ∂xk (5.241) (5. But.56). jSk s. h s (s + 1) − j (j − 1) ¯ s. j = j ¯ s. position space operators take the o form of diagonal (2 s + 1) × (2 s + 1) matrix differential operators.
1 0
(5.
109
. 0 0 −1
(5. where the Hamiltonian is some function of position and spin operators. 2 0 i 0 1 0 0 = 0 0 0 . (5. (5.13 Spin greater than onehalf systems
5 ANGULAR MOMENTUM
s (s + 1) − j (j + 1) δj. This means that we can convert the general energy eigenvalue problem for a spins particle.246) (5.l−1 .247)
as we have seen previously. into 2 s + 1 coupled partial differential equations involving the 2 s + 1 wavefunctions ψsz (r ).243)–(5. we ﬁnd that σx 0 1 0 1 = √ 1 0 1.249)
σz
(5. = 0 −1
0 −i .248)
σy
0 −i 0 1 = √ i 0 −i . (5. Unfortunately.244) − 2is According to Eqs. For a spin one (s = 1) particle. we can now construct the Pauli matrices for a spin anything particle.l+1 (σy )jl = 2is s (s + 1) − j (j + 1) δj.243) 2s s (s + 1) − j (j − 1) δj.250)
In fact. 2 0 1 0
1 0 .245) (5. the Pauli matrices for a spin onehalf (s = 1/2) particle are + σx σy σz 0 1 = .239) and (5.l−1 .5. = i 0
(5. such a system of equations is generally too complicated to solve exactly.244).
since it is the sum of Hermitian operators. J1 could be an orbital angular momentum operator. −j1 + 1. that the eigenvalues of J12 and J22 can be written j1 (j1 + 1) ¯ 2 and h 2 j2 (j2 + 1) ¯ . · · · . these operators are Hermitian. respectively. j2 − 1. we are going to have to learn how to add angular momentum operators. and spin angular momentum of magnitude ¯ /2. J2j ] = 0.254)
Now J is an Hermitian operator. (5. h
We assume that the two groups of operators correspond to different degrees of freedom of the system. respectively. or z. where m1 and m2 are numbers lying in the ranges j1 . h h So. −j2 + 1. respech h tively. or halfintegers. what is the total angular momentum of the system? In order to answer this question. By deﬁnition. Alternatively. (5.5. J1 and J2 could be the orbital angular momentum operators of two different particles in a multiparticle system.14 Addition of angular momentum Consider a hydrogen atom in an l = 1 state. from the general properties of angular momentum.251) (5. We h also know that the eigenvalues of J1z and J2z take the form m1 ¯ and m2 ¯ . and J2 a spin angular momentum operator.252)
J2 × J 2 = i ¯ J 2 . J satisﬁes the fundamental commutation relation J × J = i ¯ J. For instance. (5. j1 − 1. h (5.253)
where i. The electron possesses orbital angular momentum of magnitude ¯ . · · · . y.14 Addition of angular momentum
5 ANGULAR MOMENTUM
5. Let us consider the most general case.14). −j2 . so that [J1i . According to Eqs. J1 and J2 . Let us deﬁne the total angular momentum operator J = J 1 + J2 .11) and (5. We know. Suppose that we have two sets of angular momentum operators. and obey the fundamental commutation relations J1 × J 1 = i ¯ J 1 . −j1 and j2 . where j1 and j2 are either integers. h (5. j stand for either x.255)
110
.
257) (5. J1z . it is clear from Eq. j and m.256) that [J 2 . (5. J possesses all of the expected properties of an angular momentum operator. We can deﬁne simultaneous eigenkets of each operator group. we cannot determine the projections of the individual angular momenta along the zaxis at the same time as the magnitude of the total angular momentum. (5. J2i ] = 0. J1z . [J2 . J22 ] = 0. j1 . or a halfinteger. and j can all be measured simultaneously. It follows that the eigenvalue of J 2 can be written j (j + 1) ¯ 2 . [J12 . It is clear. (5. J2z ] = 0.5. −j + 1. Thus. where m lies h in the range j. [J22 .261)
This suggests that it is not possible to measure the quantum numbers m1 and m2 simultaneously with the quantum number j. The simultaneous eigenkets of J12 . J12 ] = [J 2 . The ﬁrst group is J12 .258)
and also that all of the J1i operators commute with the J2i operators. The second group is J12 . j2 . J22 . It follows from Eq. J1i ] = 0.259) This implies that the quantum numbers j1 . we can know the magnitude of the total angular momentum together with the magnitudes of the component angular momenta. that we can form two alternate groups of mutually commuting operators. and Jz . The eigenvalue of Jz takes the form m ¯ . and those of the individual angular momenta. · · · . (5. These two groups of operators are incompatible with one another. and J2z are denoted
111
. we do not know the relationship between the quantum numbers of the total angular momentum. J22 . In other words. −j. and m2 .260) (5. At this stage. where j is an h integer. Now We know that J 2 = J12 + J22 + 2 J1 ·J2 . However. j − 1.256) that [J2 . J1z ] = 0.14 Addition of angular momentum
5 ANGULAR MOMENTUM
Thus. J 2 . m1 . and J2z . j2 . (5. from the preceding discussion. J22 .256)
(5.
m1 . j2 . j. j2 . m2 .270) (5. J 2 . in general.264) (5.14 Addition of angular momentum
5 ANGULAR MOMENTUM
j1 . and m. j2 .269)
Each set of eigenkets are complete. m2 . In other
112
. m . m2 . j2 . m2 J1z j1 . j2 .266) (5. h = j2 (j2 + 1) ¯ 2 j1 .265)
The simultaneous eigenkets of J12 . j. and Jz is incompatible with the group J12 . m . h = m2 ¯ j1 . but we cannot know both pairs of quantum numbers at the same time. m2 j1 . m J22 j1 . The operator group J12 . j2 . h (5. In addition. mutually orthogonal (for eigenkets corresponding to different sets of eigenvalues). m2 . j. j2 . m1 . j. and J2z . j2 . J 2 and Jz are denoted j1 . j. j. it is not in an eigenstate of the latter. m J2 j1 . m . In other words. m2 = j1 (j1 + 1) ¯ 2 j1 .
m1 m2
(5. The summation is over all allowed values of m1 . m1 .5. J1z . j2 .271)
j1 . m1 . m2 J22 j1 . J22 . j. m1 . h = m1 ¯ j1 . we can always determine the magnitudes of the individual angular momenta. J22 . and have unit norms. j2 . h (5. m = 1. j2 .267) (5. where J12 j1 . j2 . j. m2 J2z j1 . or the quantum numbers j and m.268) (5. m1 . J22 . m j1 . m1 . j2 . j. m2  = 1. Since the operators J12 and J22 are common to both operator groups. m2 . j2 . m1 . m1 . j2 . we can assume that the quantum numbers j1 and j2 are known. j. j. h = j2 (j2 + 1) ¯ 2 j1 .
j m
where the righthand sides denote the identity operator in the ket space corresponding to states of given j1 and j2 . h = j (j + 1) ¯ 2 j1 . We can write a conventional completeness relation for both sets of eigenkets: j1 . m1 . m . h = m ¯ j1 .263) (5. j2 . j2 . This means that if the system is in a simultaneous eigenstate of the former group then. j2 . m2 . where J12 j1 . we can either know the quantum numbers m1 and m2 . j2 . j2 . m . m Jz j1 . j2 . j. m = j1 (j1 + 1) ¯ 2 j1 . j2 . m1 .262) (5.
j1 . m1 . projected along the zaxis. j2 . If the system is in a state where a measurement of J12 . an electron in an l = 1 state.273)
which proves the assertion. j2 .276)
We can assume. the coefﬁcients are zero unless m = m 1 + m2 . m j1 . To prove this. Forming the inner product with j1 . m2 . j2 .274) (5. j2 (j2 + h 2 2 1) ¯ . m2 j1 .275) (5.
(5.14 Addition of angular momentum
5 ANGULAR MOMENTUM
words. conh h stitutes a state whose total angular momentum projected along the zaxis is 3¯ /2.273). we obtain (m − m1 − m2 ) j1 . We can use the completeness relation (5. m =
m1 m2
j1 . The weights.270) to write j1 . we can write the eigenkets of the ﬁrst group of operators as a weighted sum of the eigenkets of the second set. are called the ClebschGordon coefﬁcients. m1 .5. j. m1 . m = 0. j2 . m2 . m . that j1 ≥ j2 . (5. j. m 2 . Second. h h The ClebschGordon coefﬁcients possess a number of very important properties. m1 . m = 0. m1 . j. j2 . Thus. we note that (Jz − J1z − J2z )j1 . with orbital angular momentum ¯ . First.272)
Thus. without loss of generality. m2 j1 . m2 j1 . that for given j1 and j2 the largest possible value of m is j1 + j2 (since j1 is the
113
. j2 . m2 j1 . and Jz is bound to give the results j1 (j1 + 1) ¯ 2 . j2 . and spin angular momentum ¯ /2. and jz ¯ . J22 . from Eq. j2 . the zcomponents of different angular momenta add algebraically. j. j2 . j. j2 . h What is uncertain is the magnitude of the total angular momentum. We know. j. J 2 . and m are known with certainty. (5. So. m1 . then a measurement of the quantum numbers m1 and m2 will give a range of possible values. (5. j (j + 1) ¯ . j2 . respectively. the coefﬁcients vanish unless j1 − j2  ≤ j ≤ j1 + j2 . if the quantum numbers j1 . j. then a measurement of J1z and J2z will h h h give the results m1 ¯ and m2 ¯ with probability  j1 . j2 . j2 .
We have already seen that the maximum allowed value of j is j1 + j2 . m1 . m1 .5.278)
This assertion is proved as follows: j1 . there are (2 j1 + 1) allowable values of m1 and (2 j2 + 1) allowable values of m2 . Finally. j. j2 . j2 . j2 . j. m1 . j. j2 . In other words. This implies that the largest possible value of j is j1 + j2 (since. j2 .
(5.. m = (J± + J± ) 2 1 × j1 . For each allowed value of j. mj1 .14 Addition of angular momentum
5 ANGULAR MOMENTUM
largest possible value of m1 . Now. j2 . needed to span the ket space corresponding to ﬁxed j1 and j2 . m2 j1 . m2 j1 . m2 j1 . j1 . mj1 . m2 j1 . there are 2 j + 1 allowed values of m.e.). j2 .277)
This proves our assertion. Thus. m
m1 m2
=
m1 m2
 j1 . m1 . j2 . etc. m2 .
m1 m2
(5.279)
where use has been made of the completeness relation (5. the largest value of m is equal to j).
(5. j. It is easily seen that if the minimum allowed value of j is j1 − j2 then the total number of allowed values of j and m is (2 j1 + 1) (2 j2 + 1): i. j2 . j2 . j. m j1 . they must also form a set of (2 j1 + 1) (2 j2 + 1) independent kets. j2 .
j1 +j2
j=j1 −j2
(2 j + 1) ≡ (2 j1 + 1) (2 j2 + 1). m2 j1 . by deﬁnition.270). the sum of the modulus squared of all of the ClebschGordon coefﬁcients is unity: i. j2 . the ClebschGordon coefﬁcients obey two recursion relations. j2 . Since the eigenkets j1 . m1 . there are (2 j1 + 1) (2 j2 + 1) independent eigenkets. m = j1 . m1 .280)
114
.
m1 m2
(5.e. j2 . Third. m 2 = 1. there can only be (2 j1 + 1) (2 j2 + 1) distinct allowable values of the quantum numbers j and m. m 2 = 1. To obtain these relations we start from J± j1 .  j1 . j2 . m2 . m1 . m span the same space. j. j. j. j.. j2 .
m2
It is clear. j. m2 j1 .282). m . m1 . j. j2 . j2 . j1 . m1 . (5. m1 . m1 ± 1. mj1 . j.55)–(5. (5. we obtain the desired recursion relations: j (j + 1) − m (m ± 1) j1 . This sign is ﬁxed by convention. mj1 . m2 in terms of the j1 .
(5. j2 .282) 1) j1 . The easiest way of demonstrating this assertion is by considering some speciﬁc examples. (5. m : j1 .14 Addition of angular momentum
5 ANGULAR MOMENTUM
Making use of the wellknown properties of the shift operators. j.284)
It turns out that the recursion relations (5. m2 . j. m2 j1 . j2 .56).281) Taking the inner product with j1 . j2 . m .283)
The inverse ClebschGordon coefﬁcients are the weights in the expansion of the j1 . which are speciﬁed by Eqs. m . j2 . j. j. j2 . are sufﬁcient to completely determine the ClebschGordon coefﬁcients to within an arbitrary sign (multiplied into all of the coefﬁcients). j2 . m ± 1 = j1 (j1 + 1) − m1 (m1 + j2 (j2 + 1) − m2 (m2 1) j1 . j2 . j2 . In other words. j2 .278). j. m1 . m2 =
j m
j1 . since it ensures that the inverse ClebschGordon coefﬁcients.5. j2 . j2 . m1 . m2 . m2 = j1 . and making use of the orthonormality property of the basis eigenkets.
115
. m1 . j2 . j2 . mj1 . that we can always choose the ClebschGordon coefﬁcients to be real numbers. j. j2 . m1 . j2 . j2 . j2 . j2 . m ± 1 =
m1 m2
j1 (j1 + 1) − m1 (m1 ± 1) j1 . are identical to the ClebschGordon coefﬁcients. j2 . m1 . j. j2 . j1 . m2 ± 1 × j1 . m1 . m2 j1 . we obtain j (j + 1) − m (m ± 1) j1 . m2 j1 . m1 . m2 j1 . m 1j1 . m1 . m1 1. j. j2 . j2 . together with the normalization condition (5. m . (5. from the absence of complex coupling coefﬁcients in the above relations. This is a convenient choice. m2
+ j2 (j2 + 1) − m2 (m2 ± 1) j1 .
We also know. or the inverse ClebschGordon coefﬁcient 1/2. m2 = 1/2. 1. In other words. that m1  ≤ 1/2 and m2  ≤ 1/2. 1/2.5.g. A ClebschGordon coefﬁcient is automatically zero unless m1 + m2 = m. 1/2 . Thus. the zcomponents of angular momentum have to add algebraically. We immediately conclude that these boxes must contain zeroes: i.14 Addition of angular momentum
5 ANGULAR MOMENTUM
Let us add the angular momentum of two spin onehalf systems: e. two electrons at rest. m1 m2 1/2 1/2 ? 0 1/2 1/2 0 ? 1/2 1/2 0 ? 1/2 1/2 0 0 j1 =1/2 j 1 1 j2 =1/2 m 1 0 0 0 0 ? 1 1 0 ? ? 0 0 0
The normalization condition (5. Many of the boxes in the above table correspond to m1 +m2 = m. 1 .. 1/2. where the allowed values of j differ by integer amounts. from Eq. that 0 ≤ j ≤ 1. (5. 1/21/2. j = 1. It is helpful to arrange all of the possibly nonzero ClebschGordon coefﬁcients in a table: m1 m2 1/2 1/2 ? ? ? ? 1/2 1/2 ? ? ? ? 1/2 1/2 ? ? ? ? 1/2 1/2 ? ? ? ? j1 =1/2 j 1 1 1 0 j2 =1/2 m 1 0 1 0 The box in this table corresponding to m1 = 1/2. from general principles. 1/2. It follows that either j = 0 or j = 1.276). two spin onehalf systems can be combined to form either a spin zero system or a spin one system. 1/2.. 1/2. m = 1 gives the ClebschGordon coefﬁcient 1/2. All the boxes contain question marks because we do not know any ClebschGordon coefﬁcients at the moment. 1/2. 1. We know. 11/2.e. j1 = j2 = 1/2.278) implies that the sum of the squares of all
116
. So.
−1/21. 1/21. 1/21. 1/21. −1 = ±1. −1/21. 0 = 2 1/2. −1/21. 0 = 1/2. 0 2 = 1. −1 = ± 2.
117
. the j1 and j2 labels have been suppressed for ease of notation. m1 = m2 = ±1/2. taking the upper/lower sign. Thus.286)
Here. (5. 1/21. and 1/2. We also know that 1/2. 1/21. Thus. 0 = √ √ 2 −1/2. with j = 1. (5.287) from the normalization condition. m1 m2 1/2 1/2 ±1 0 0 0 1/2 1/2 0 ? 0 ? 1/2 1/2 0 ? 0 ? 1/2 1/2 0 0 ±1 0 j1 =1/2 j 1 1 1 0 j2 =1/2 m 1 0 1 0 Let us evaluate the recursion relation (5. The only real solutions to the above set of equations are √ √ 2 1/2. 0 2 + −1/2. −1/21. but we have no way of determining the signs at present. There are two rows and two columns which only contain a single nonzero entry.5.282) for j1 = j2 = 1/2.285) 1/2. 1/21.288)
The choice of sign is arbitrary—the conventional choice is a positive sign. −1/21. We ﬁnd that √ √ (5.14 Addition of angular momentum
5 ANGULAR MOMENTUM
the rows and columns of the above table must be unity. 0 + −1/2. 0 + −1/2. 0 = 2 −1/2. 1 = 1/2. (5. 1/21. We conclude that these entries must be ±1. m = 0. 1 = ± 2.
That is. The normalization condition tells us that the dot product of a row or column with itself must be unity. The requirement that the dot product of the fourth column with itself is unity tells us that the magnitudes of the unknown entries have to √ be 1/ 2.5. the ﬁnal form of our table (with the conventional choice of arbitrary signs) is m1 m2 1/2 1/2 1 0 0 √ 0 √ 1/2 1/2 0 1/√2 0 1/ √2 1/2 1/2 0 1/ 2 0 1/ 2 1/2 1/2 0 0 1 0 j1 =1/2 j 1 1 1 0 j2 =1/2 m 1 0 1 0 The table can be read in one of two ways. This follows because the entries in the table give the expansion coefﬁcients of one of our alternative sets of eigenkets in terms of the other set.14 Addition of angular momentum
5 ANGULAR MOMENTUM
our table now reads m1 m2 1/2 1/2 1 0 √ 1/2 1/2 0 1/√2 1/2 1/2 0 1/ 2 1/2 1/2 0 0 j1 =1/2 j 1 1 j2 =1/2 m 1 0 0 0 0 1 1 1 0 ? ? 0 0 0
We could ﬁll in the remaining unknown entries of our table by using the recursion relation again. Thus. The only way that the dot product of the fourth column with the second column can be zero is if the unknown entries are equal and opposite. Likewise. for the dot product of a column with any other column. the dot product of a row with any other row must be zero. The columns give the expansions of
118
. The unknown entries are undetermined to an arbitrary sign multiplied into them both. an easier method is to observe that the rows and columns of the table must all be mutually orthogonal. and each set of eigenkets contains mutually orthogonal vectors with unit norms. However.
0 = √ ( 1/2. whereas those on the righthand side are j. there is no recursion relation linking ClebschGordon coefﬁcients corresponding to different values of j. Note that our table is really a combination of two subtables. the second column tells us that 1 1. the ket on the lefthand side is a m1 . j2 and j) in many standard reference books.14 Addition of angular momentum
5 ANGULAR MOMENTUM
the eigenstates of overall angular momentum in terms of the eigenstates of the individual angular momenta of the two component systems. whereas those on the righthand side are m1 . the second row tells us that 1 1/2. j2 . The ClebschGordon coefﬁcients corresponding to two different choices of j are completely independent: i. −1/2 = √ ( 1. m ket. 1/2 ) . m2 ket. m kets. The rows give the expansions of the eigenstates of individual angular momentum in terms of those of overall angular momentum.e.290)
Here. These tables are listed (for moderate values of j 1 .5. one involving j = 0 states. and j we can construct a table of ClebschGordon coefﬁcients corresponding to the different allowed values of m 1 .
119
.289)
The ket on the lefthand side is a j. Thus. 0 ) . Thus. and m (subject to the constraint that m1 + m2 = m). for every choice of j1 . −1/2 +  − 1/2. m2 . 0 + 0.. 2 (5. m2 kets. A complete knowledge of angular momentum addition is equivalent to a knowing all possible tables of ClebschGordon coefﬁcients. 2 (5. Thus. and one involving j = 1 states.
The Hamiltonian of a system is written H = H 0 + H1 . in some sense (which we shall specify more exactly later on). we need to develop some techniques for ﬁnding approximate solutions to otherwise intractable problems. H1 .
120
(6. H0 + H1 . These are denoted H0 1 = E1 1 . Can we ﬁnd the approximate eigenvalues and eigenstates of the modiﬁed Hamiltonian. H0 . Consider the simplest nontrivial system.6 APPROXIMATION METHODS
6
Approximation methods
6. has no explicit dependence on time. It is usually assumed that the unperturbed Hamiltonian.1 Introduction We have developed techniques by which the general energy eigenvalue problem can be reduced to a set of coupled partial differential equations involving various wavefunctions. (6.2)
. but it is sufﬁciently complicated that when we add it to H0 we can no longer ﬁnd the exact energy eigenvalues and eigenstates. in which the modiﬁcation to the Hamiltonian. in which there are only two independent eigenkets of the unperturbed Hamiltonian. Consider the following problem. the number of such problems which yield exactly soluble equations is comparatively small. by performing some sort of perturbation analysis about the eigenvalues and eigenstates of the original Hamiltonian. H0 is a simple Hamiltonian for which we know the exact eigenvalues and eigenstates.
6.1) Here. However. H1 can. which is very common. H 0 ? Let us investigate. is also timeindependent. H1 introduces some interesting additional physics into the problem.2 The twostate system Let us begin by considering timeindependent perturbation theory. Clearly. be regarded as being small compared to H0 . Unfortunately.
6.6)
e11 = e22 = e12 =
1H1 1 .6) (obtained by setting the determinant of the matrix equal to zero) is (E1 + E2 ) ± (E1 − E2 )2 + 4 e12 2 E= . (6. (6.12)
.2 The twostate system
6 APPROXIMATION METHODS
H0 2
= E2 2 . E1 − E2 
121
(6.
(6. (6. The lengths of these eigenkets are both normalized to unity. by deﬁnition. the eigenkets of H 0 form a complete set. (6. an Hermitian operator. Let us now try to solve the modiﬁed energy eigenvalue problem (H0 + H1 )E = E E .10)
the solution of Eq. and their associated eigenvalues. 0
(6. its two eigenkets are orthonormal and form a complete set. we can write E = 1E 1 + 2E 2 .9)
In the special (but common) case of a perturbing Hamiltonian whose diagonal matrix elements (in the unperturbed eigenstates) are zero. so that e11 = e22 = 0.8) (6. (6.3)
It is assumed that these states. Since H0 is. Since. which can be written in matrix form:
E1 − E + e11 e12 ∗ e12 E2 − E + e22
1E 2E
Here. we can solve this problem exactly.4)
In fact. (6. 2H1 2 . 1H1 2 .4) by 1 and 2 yields two coupled equations. are known.7) (6.11) 2 Let us expand in the supposedly small parameter = e12  .5)
Rightmultiplication of Eq.
0 = .
(6.
19)
.3 Nondegenerate perturbation theory Let us now generalize our perturbation analysis to deal with systems possessing more than two energy eigenstates. The energy eigenstates of the unperturbed Hamiltonian. Note that the series expansion in Eq. 2 (6.18)
In other words. the modiﬁed energy eigenstates consist of one of the unperturbed eigenstates with a slight admixture of the other. E1 − E 2 e12 2 + ···. E1 − E 2 (6. H0 .6. E1 − E 2
(6. are denoted H0 n = En n .13) only converges if 2   < 1. (6. It is easily demonstrated that the modiﬁed eigenkets take the form 1 2 = 1 +
∗ e12 2 + · · · . This suggests that the condition for the validity of the perturbation expansion is e12  < E1 − E2  .3 Nondegenerate perturbation theory
6 APPROXIMATION METHODS
1 1 (E1 + E2 ) ± (E1 − E2 )(1 + 2 2 + · · ·).
6.17)
Thus. what we really mean is that the above inequality needs to be satisﬁed. and the lower eigenvalue to fall.13) 2 2 The above expression yields the modiﬁcations to the energy eigenvalues due to the perturbing Hamiltonian: E E1 e12 2 = E1 + + ···.14) (6.15)
We obtain
E2 = E 2 −
Note that H1 causes the upper eigenvalue to rise. (6. E1 − E 2 e12 = 2 − 1 + · · · .
122
(6.16) (6. when we say that H1 needs to be small compared to H0 .
29)
. We assume that emk  ∼ O( ).22) for m = n.21)
where the summation is from k = 1 to N.
(6.3 Nondegenerate perturbation theory
6 APPROXIMATION METHODS
where n runs from 1 to N. The eigenkets n are orthonormal. ∼ O( ). We also assume that (6. and have their lengths normalized to unity. neglecting terms which are O( 2 ) according to our expansion scheme. Em (6. for which E = En + O( ). (6. (6. (6. Em − E k for all m = k. Let us now try to solve the energy eigenvalue problem for the perturbed Hamiltonian: (H0 + H1 )E = E E .26) and nE mE = 1. Substituting the above equation into Eq. form a complete set.27) (6. we obtain (Em + emm − E) mE +
k=m
emk kE = 0.24) (6. (6.
(6. Let us search for a modiﬁed version of the nth unperturbed energy eigenstate.20)
We can express E as a linear superposition of the unperturbed energy eigenkets.
(6.22)
where emk = mH1 k . Suppose that we write out Eq. We ﬁnd that (Em − En ) mE + emn
123
0.25)
for all m. where emm  ∼ O( ).20). Let us now develop our perturbation expansion. E =
k
kE k . (6.23)
1 is our expansion parameter.28)
for m = n. and rightmultiplying by m.6.
En = En + enn + E − Ek k=n n and a eigenket n = n + Note that mn = δmn +
∗ emn enm + + O( 2 ) = δmn + O( 2 ).32)
ekn k + O( 2 ). the modiﬁed eigenkets remain orthonormal and properly normalized to O( 2 ). either a hydrogen atom.33)
(6. Em − E n En − E m
giving
(6. E − Ek k=n n
(6. (6.4 The quadratic Stark effect Suppose that a oneelectron atom [i. the modiﬁed nth energy eigenstate possesses an eigenvalue enk 2 + O( 3 ).34)
Thus. and neglecting O( 3 ) terms.35) H0 = 2 me and the perturbing Hamiltonian. (6..30) Em − E n Substituting the above expression into Eq. or an alkali metal atom (which possesses one valance electron orbiting outside a closed. spherically symmetric shell)] is subjected to a uniform electric ﬁeld in the positive zdirection.
124
(6.22).31)
(6. we obtain mE − enk 2 (En + enn − E) − = 0. The unperturbed Hamiltonian.6. H1 = e E z. p2 + V(r).4 The quadratic Stark effect
6 APPROXIMATION METHODS
emn .
6.e. Ek − E n k=n Thus.36)
. (6. The Hamiltonian of the system can be split into two parts. evaluated for m = n.
According to Eq. since Lz = x p y − y p x . z] = 0. the change in this energy level induced by a small electric ﬁeld is given by ∆Enlm = e E n. l. m 2 +e E .41) since n.m
2 2
(6. z] + [Ly2 . This implies that the system possesses no degenerate energy eigenvalues. n. giving (m − m ) n.4 The quadratic Stark effect
6 APPROXIMATION METHODS
It is assumed that the unperturbed energy eigenvalues and eigenstates are completely known. mzn.l. An energy eigenket of the unperturbed Hamiltonian is characterized by three quantum numbers—the radial quantum number n. l . mzn . It is necessary to deal with this case separately. The electron spin is irrelevant in this problem (since the spin operators all commute with H1 ).37)
Now. and let its energy level be Enlm . It is h clear. l.32). by deﬁnition. 6.6). Let us denote such a ket n. (6.3 breaks down for degenerate unperturbed energy levels. m is zero unless m = m. Thus. m = 0. an eigenstate of Lz with eigenvalue m ¯ . because the perturbation theory presented in Sect.39) (6. l. l . it follows that [Lz . l. We have [L2 . z]
125
(6. l.6. Let us now determine the selection rule for l. z] = [Lx2 . m = 0. l. so we can ignore the spin degrees of freedom of the system. m  n. from the above relation. due to the special properties of a pure Coulomb potential. l . (6.38) (6. This is not true for the n = 1 energy levels of the hydrogen atom. This is termed the selection rule for the quantum number m. mzn . 5. m[Lz . l .40)
. z]n . l. and the two angular quantum numbers l and m (see Sect.m =n. mzn.l . m . l. m is. that the matrix element n. Enlm − En l m n .
[L2 . we obtain [L2 . mL4 z − 2 L2 z L2 + z L4 − 2 ¯ 2 (L2 z + z L2 )n .42)
(6.4 The quadratic Stark effect
6 APPROXIMATION METHODS
= Lx [Lx .6). l. x] = 2 i ¯ (y Lz − Ly z). Similarly. z]] = 2 ¯ 2 (L2 z + z L2 ).48)
(6. h This reduces to [L2 . m = 0. h Finally. Hence. z] + [Lx . [L2 . (5. [L2 . z] .44)
(6. Ly x − Lx y + i ¯ z h h = 2 i ¯ Ly [L2 . z]Ly = i ¯ (−Lx y − y Lx + Ly x + x Ly ) h = 2 i ¯ (Ly x − Lx y + i ¯ z) h h = 2 i ¯ (Ly x − y Lx ) = 2 i ¯ (x Ly − Lx y). h [L2 .50)
. z]Lx + Ly [Ly .3) that Lx x + Ly y + Lz z = 0. z]] = 2 i ¯ L2 . (5. z]] = −¯ 2 4 (Lx x + Ly y + Lz z)Lz − 4 (Lx2 + Ly2 + Lz2 ) z h +2 (L2 z − z L2 ) . [L2 .49) implies that n. the above expression expands to give L4 z − 2 L2 z L2 + z L4 − 2 ¯ 2 (L2 z + z L2 ) = 0. However.49)
(6.6. x] − Lx [L2 . h Equation (6.43) (6. it is clear from Eqs. h Thus.46)
(6. y] + i ¯ [L2 . y] = 2 i ¯ (Lx z − x Lz ).1)–(5. z] + [Ly . h h where use has been made of Eqs. l . h h = −4 ¯ 2 Ly (y Lz − Ly z) + 4 ¯ 2 Lx (Lx z − x Lz ) h h −2 ¯ 2 (L2 z − z L2 ).45)
(6. h
126
(6.1)–(5. [L2 .47) (6.
that the wavefunction of an l = 0 state is spherically symmetric (see Sect.4 The quadratic Stark effect
6 APPROXIMATION METHODS
This expression yields l2 (l + 1)2 − 2 l (l + 1) l (l + 1) + l 2 (l + 1)2 −2 l (l + 1) − 2 l (l + 1)] n. the matrix element n. (Recall. mzn . l.56)
. This matrix element can be written n. 2 (6. 5.37) yields ∆Enlm = e2 E2
n
 n. mzn . mzn .53) that the matrix element vanishes by symmetry when l = l = 0. m vanishes unless l = l = 0 or l = l ± 1. the matrix element n. l. The polarizability of an atom is deﬁned in terms of the energyshift of the atomic state as follows: 1 ∆E = − α E2 . (6. mzn . Only those terms which vary quadratically with the ﬁeldstrength survive.55)
Consider the ground state of a hydrogen atom. m = ψ∗ (r . The polarizability of this state is given by α = 2e
2 n>1
 1. θ . 0. En00 − E100
127
(6. Recall. l .3): i.37) which vary linearly with the electric ﬁeldstrength vanish by symmetry. m 2 . This is the selection rule for the quantum number l. l. In conclusion.53) nlm
where ψnlm (r ) = r n. m which reduces to (l + l + 2) (l + l ) (l − l + 1) (l − l − 1) n. l . l. (6. that we cannot address the n > 1 excited states because they are degenerate.54)
Note that all of the terms in Eq. mzn .. however. (6. m . m is zero unless l = l ± 1. l. θ . l . (6. l . ϕ ) r cos θ ψn m l (r .6.e. m = 0. and our theory cannot handle this at present). l. (6. 1.52) = 0. l . 0 2 . mzn . Application of the selection rules to Eq.51)
According to the above formula. 0zn. Enlm − En l m l =l±1
(6. (6. It follows from Eq. ϕ ) dV . according to the selection rules. ψn00 (r ) = ψn00 (r ). l . l.
64) 4π 0 a03 = 4.59)
16 4π 3
0 a0 n>1
 1.3 4π
3 0 a0 .62) 1. 0.
(6.
n . we have made use of the fact that En10 = En00 for a hydrogen atom. We can write En00 − E100 ≥ E200 − E100 Thus. The sum in the above expression can be evaluated approximately by noting that [see Eq. 1.58)
3 e2 = . 0 2 . 1.  1.4 The quadratic Stark effect
6 APPROXIMATION METHODS
Here.61)
=
where we have made use of the fact that the wavefunctions of a hydrogen atom form a complete set.
(6. o α=
128
. where 4π 0 ¯ 2 h = 5. (5. 0.63)
9 (6. It is easily demonstrated from the actual form of the ground state wavefunction that (6. we conclude that α< The true result is 16 4π 3
3 0 a0
5. l .120)] e2 En00 = − (6. 0.3 × 10−11 meters a0 = 2 µe is the Bohr radius. 0z2 1. by solving Schr¨dinger’s equation exactly in parabolic coordinates. 0. 0 = a02 . 2 It is actually possible to obtain this answer. 0zn. 0. without recourse to perturbation theory. m 1.5 4π 0 a03 .m
(6.l . m . 0 2 =
n>1 n . l z1. α< However. 0zn. 0 (6. 0 . 0. 0. 4 8π 0 a0
(6. Thus. 0zn .6.57) 8π 0 a0 n2 for a hydrogen atom.60)
1. 0z2 1. 0.
66) where [H0 . l acquires a unique energy eigenvalue Enl . l = En n. the En and the Lnl are real numbers which depend on the quantum numbers n. (6. It is always possible to represent degenerate energy eigenstates as the simultaneous eigenstates of the Hamiltonian and some other Hermitian operator (or group of operators). we expect the perturbation to split the degeneracy of the energy levels. In this situation.l n en l nl n . This implies that the modiﬁed energy eigenstates are not eigenstates of L. In general. we can choose L such that the quantum numbers n and l uniquely specify each eigenstate.5 Degenerate perturbation theory
6 APPROXIMATION METHODS
6.69)
.68)
(6. E − En n . L does not commute with the perturbing Hamiltonian. so that each modiﬁed eigenstate n. A direct generalization of Eqs. respectively. l = n.5 Degenerate perturbation theory Let us now consider systems in which the eigenstates of the unperturbed Hamiltonian.32) and (6. = En + enlnl + E − En n . Suppose that for each value of n there are Nn different values of l: i. l .e. possess degenerate energy levels. H0 . H1 .33) yields Enl and n. l + where en l nl = n .l =n. the nth energy eigenstate is Nn fold degenerate. l .
129
(6.65)
en l nl 2 + O( 3 ). l + O( 2 ). Let us denote this operator (or group of operators) L.l n
(6. l H1 n. l = Lnl n. We can write H0 n. Here. In other words. It is always possible to ﬁnd a sufﬁcient number of operators which commute with the Hamiltonian in order to ensure that the Lnl are all different.l =n.67)
(6. L] = 0.. l . Let us naively attempt to use the standard perturbation theory of Sect. 6. and n and l.3 to evaluate the modiﬁed energy eigenstates and energy levels. (6. and L n.6.
70) is not satisﬁed. of the perturbing Hamiltonian between distinct. Fortunately. l(1) = λnl n. then we can employ Eqs. These terms give rise to singular factors 1/(En − En ) in the summations. (6. The problem terms are those involving unperturbed eigenstates labeled by the same value of n. l(1) = λnl δll . we can always redeﬁne the unperturbed energy eigenstates belonging to the eigenvalue En in such a manner that Eq. H1 n.68) directly. instead of the old ones. Thus. (6. (6.5 Degenerate perturbation theory
6 APPROXIMATION METHODS
It is fairly obvious that the summations in Eqs. enl nl . In general. but different values of l: i. H 1 . unperturbed energy eigenstates corresponding to the eigenvalue En were zero. l(1) n. Note. Eq. and have unit lengths.68) would vanish. Thus. l H1 n. those states whose unperturbed energies are En . kn. then all of the singular terms in Eqs..67) and (6.68) are not wellbehaved if the nth energy level is degenerate. The n. The only remaining difﬁculty is to determine the new eigenstates in terms of the original ones. k . (6.70)
n. however. l = λnl δll .70) is satisﬁed.73) n. (6. l(1) . l (1) H1 n. Let us deﬁne Nn new states which are linear combinations of the Nn original degenerate eigenstates corresponding to the eigenvalue En :
Nn
(6. It follows that (6.e. In other words. l
(1)
=
k=1
n. that this problem would not exist if the matrix elements. degenerate.67) and (6.6. l(1) are also chosen so that they are orthonormal. since all of the singular terms vanish.
(6. (6.71)
Note that these new states are also degenerate energy eigenstates of the unperturbed Hamiltonian corresponding to the eigenvalue En .67) and (6.
130
. if we use the new eigenstates. if n.72) The n. l(1) are chosen in such a manner that they are eigenstates of the perturbing Hamiltonian.
k . Note that the ﬁrstorder energy shifts are equivalent to the eigenvalues of the matrix equation (6. unperturbed energy eigenstates corresponding to the eigenvalue En .6. they speciﬁcally exclude the problematic. (6. l n.5 Degenerate perturbation theory
6 APPROXIMATION METHODS
Now
Nn
n. l
l =1
n. (6. Using this completeness relation. (xnl )k = n.. Eq. the operator eigenvalue equation (6.76).78) for k = 1 to Nn . l(1) +
n =n. En − E n
(6.79)
and n. kn. degenerate. since the summations are over n = n: i. with Nn corresponding eigenvectors xnl .74)
where 1 denotes the identity operator in the subspace of all unperturbed energy eigenkets corresponding to the eigenvalue En .e. where the elements of the Nn × Nn Hermitian matrix U are Ujk = n.76) can always be solved to give Nn eigenvalues λnl (for l = 1 to Nn ). (6. l = 1. l n. l + O( 2 ). In our new scheme. l H1 n.75)
This can be written more transparently as U x = λ x..80)
There are no singular terms in these expressions.
(6. l n. (6.l
(6. jH1 n.77) Provided that the determinant of U is nonzero.76)
en l nl 2 + O( 3 ).72) can be transformed into a straightforward matrix eigenvalue equation:
Nn
n. l(1) . l(1) = λnl n.68) yield Enl = En + λnl +
n =n.
131
. The eigenvectors specify the weights of the new eigenstates in terms of the original eigenstates: i. l(1) . En − E n
(6.l
en l nl n .e.
l=1
(6.67) and (6. Eqs. l(1) = n.
(6. 1. 1 . consider the n = 2 states.6. 1. 0 = 3 a0 . Here. and l quantum numbers which differ by unity.81)
In order to apply perturbation theory.4. λ3 = 0. 0z2. 6. λ2 = −3 e a0 E. we have to solve the matrix eigenvalue equation U x = λ x. E200 = −e2 /(32π 0 a0 ). 2. For instance. the perturbing Hamiltonian is H1 = e E z. 0 . usually referred to as 2p.
e E
U=
0 2. we have made use of the selection rules. 0 0 0 0
0 0 0 0
0 0 0 0
where the rows and columns correspond to the 2. and 2. Thus. (6. 0 = 2. 0. by inspection. (6.85) x1 = 0 0
132
.84)
. All of these states possess the same energy. 0. It is easily demonstrated. (6. The corresponding eigenvectors are √ 1/√2 1/ 2 . 0. 1. 2. usually referred to as 2s. respectively. 0. 0z2. 0z2. 0 . and λ4 = 0. 0. 0. 1. 0 0 0
2. 1. There is a single l = 0 state. 1.82) where U is the array of the matrix elements of H1 between the degenerate 2s and 2p states.
(6. As in Sect. 1. 0z2. −1 states. and three l = 1 states (with m = −1. 1). from the exact forms of the 2s and 2p wavefunctions.6 The linear Stark effect Let us examine the effect of an electric ﬁeld on the excited energy levels of a hydrogen atom. that the eigenvalues of U are λ1 = 3 e a0 E.6 The linear Stark effect
6 APPROXIMATION METHODS
6. that 2.83)
It can be seen. which tell us that the matrix element of z between two hydrogen atom states is zero unless the states possess the same m quantum number.
−1 . 1.91) (6.96)
Thus. 2 2. 0 √ = .6.5 that the simultaneous eigenstates of the unperturbed Hamiltonian and the perturbing Hamiltonian take the form 1 2 3 4 = 2. by an amount 3 e a0 E in the presence of an electric ﬁeld. 1 . ∆E4 = 0. (6.93) (6.92)
In the absence of an electric ﬁeld. 0 √ . 1.
(6.
133
. 0 + 2. respectively. the energies of states 1 and 2 are shifted upwards and downwards.87)
x4 =
(6.94) (6.89) (6. (6. 0 − 2. 2 = 2. The ﬁrstorder energy shifts induced by an electric ﬁeld are given by ∆E1 = +3 e a0 E. 6. 1.6 The linear Stark effect
6 APPROXIMATION METHODS
x2 =
√ 1/ √ 2 −1/ 2 0 0 0 0 1 0 0 0 0 1
. 0.86)
x3 =
(6. States 1 and 2 are orthogonal linear combinations of the original 2s and 2p(m = 0) states.95) (6. all of these states possess the same energy. = 2. 1. 0. ∆E3 = 0. .88)
It follows from Sect.90) (6. E 200 . ∆E2 = −3 e a0 E.
.
Note that the linear Stark effect depends crucially on the degeneracy of the 2s and 2p states.7 Fine structure
6 APPROXIMATION METHODS
Note that the energy shifts are linear in the electric ﬁeldstrength. The energies of states 3 and 4 (which are equivalent to the original 2p(m = 1) and 2p(m = −1) states. Let us examine a phenomenon known as ﬁne structure. higher l states have higher energies. to secondorder the energies of these states are shifted by an amount which depends on the square of the electric ﬁeldstrength.97)
Now. alkali metal atoms do not exhibit the linear Stark effect. e (6. which is due to interaction between the spin and orbital angular momenta of the outermost electron.6. so this is a much larger effect that the quadratic effect described in Sect. alkali metal atoms) in more detail. (6. a charge moving in an electric ﬁeld also experiences an effective magnetic ﬁeld B = −v × E. Thus. The outermost electron moves in a spherically symmetric potential V(r) due to the nuclear charge and the charges of the other electrons (which occupy spherically symmetric closed shells).98) (6.7 Fine structure Let us now consider the energy levels of hydrogenlike atoms (i. and.99)
134
.4. This splits the degeneracy of states characterized by the same value of n. an electron possesses a spin magnetic moment [see Eq. 6. The shielding effect of the inner electrons causes V(r) to depart from the pure Coulomb form. only applies to a hydrogen atom. In fact. This degeneracy is a special property of a pure Coulomb potential. Of course.170)] µ=− eS . but different values of l.e. me
However. therefore.
6. respectively) are not affected to ﬁrstorder. This electron experiences an electric ﬁeld E= V . (5..
S2 . ms = ±1/2). Now. We can adopt the simultaneous eigenstates of H0 . Sz . using H LS as the perturbation (with HLS taking one half of the value given above). and Jz .102)
. it is found to be too large by a factor of two. Although the departure of V(r) from a pure 1/r form splits the degeneracy of same n. different l. There is a classical explanation for this.) Moreover. ml . each state is doubly degenerate due to the two possible orientations of the electron spin (i. and mj are the quantum numbers corresponding to Lz . that the application of perturbation theory to a degenerate system is greatly simpliﬁed if the basis eigenstates of the unperturbed Hamiltonian are also eigenstates of the perturbing Hamiltonian. with the addition of spin degrees of freedom. are still degenerate. (Here. We know. Thus. we are still dealing with a highly degenerate system. ms . the perturbing Hamiltonian.101)
as the unperturbed Hamiltonian. J 2 . from Sect. expect a spinorbit contribution to the Hamiltonian of the form HLS = −µ·B 1 r dV eS ·v× = − me e r dr 1 dV L·S. 6. The correct quantum mechanical explanation requires a relativistically covariant treatment of electron dynamics (this is achieved using the socalled Dirac equation).6.6. L2 . but different values of ml . L2 . states. or the simultaneous eigenstates of H0 . and p2 + V(r) H0 = 2 me (6. and Jz . those states characterized by the same values of n and l. = me2 r dr
(6. L·S = 2
135
(6. where J = L + S is the total angular momentum. HLS .7 Fine structure
6 APPROXIMATION METHODS
We. which we need not go into. Let us now apply perturbation theory to a hydrogenlike atom. is proportional to L·S. We have two choices for the energy eigenstates of H0 . therefore. where J 2 − L2 − S 2 . respectively.100)
where L = me r × v is the orbital angular momentum. S 2 . Lz and Sz . due to spin precession..e. When the above expression is compared to the observed spinorbit interaction.
= − sin α m − 1/2. mj form. and j2 = s = 1/2. m l − 1/2. and have unit lengths. L2 .282). with j1 = l. and Jz ) does. m2 = 1/2 (lower sign).104)
Here.7 Fine structure
6 APPROXIMATION METHODS
It is fairly obvious that the ﬁrst group of operators (H0 . We can write l + 1/2. L2 .105)
.107) l + m + 3/2
136
(6. Thus. (6. 1/2l + 1/2. and Jz . J 2 . j2 labels have been dropped. We now need to ﬁnd the simultaneous eigenstates of H0 . (6. In fact.103) (6. it is advantageous to work in terms of the eigenstates of the second group of operators. −1/2 . We have also made use of the fact that both the j. ms kets are orthonormal. Let us now employ the recursion relation for ClebschGordon coefﬁcients.276). J 2 .6. whereas the second group (H0 . Eq. whereas those on the righthand side are ml . S 2 . j = l + 1/2. (5. m + 1 . Lz and Sz ) does not commute with HLS . the kets on the lefthand side are j. 1/2 + sin α m + 1/2. According to Eq. We have made use of the fact that the ClebschGordon coefﬁcients are automatically zero unless mj = ml + ms .106) which reduces to m − 1/2. 1/2l + 1/2. m = cos α m − 1/2. j2 = 1/2. m = l (l + 1) − (m − 1/2) (m + 1/2) m + 1/2. −1/2 . S2 . where the ClebschGordon coefﬁcient is written in ml . S 2 . ms kets (the j1 . for the sake of clarity). 1/2l + 1/2. This is equivalent to ﬁnding the eigenstates of the total angular momentum resulting from the addition of two angular momenta: j1 = l. m1 = m − 1/2. (5. 1/2l + 1/2. m . ms j. (6. mj and ml . L·S is just a combination of operators appearing in the second group. L2 . We now need to determine cos α = m − 1/2. rather than those of the ﬁrst group. m = l + m + 1/2 m + 1/2. the allowed values of the total angular momentum are j = l + 1/2 and j = l − 1/2. 1/2 + cos α m + 1/2. 1/2l + 1/2. We obtain (l + 1/2) (l + 3/2) − m (m + 1) m − 1/2. m + 1 . mj kets.
l + 1/2 = 1. (6.7 Fine structure
6 APPROXIMATION METHODS
We can use this formula to successively increase the value of ml . up to an arbitrary phasefactor. The corresponding value of mj is l + 1/2. m = Now. m = + l + m + 1/2 m − 1/2. 2l + 1
137
(6.109)
Consider the situation in which ml and m both take their maximum values.110)
We now need to determine the sign of sin α. mj ket l + 1/2. sin2 α = 1 − l + m + 1/2 l − m + 1/2 = . giving l. 1/2l + 1/2.108) This procedure can be continued until ml attains its maximum possible value.109) that cos α = m − 1/2. m + 2 .111) (6. m = l + m + 1/2 l + m + 3/2 l + m + 3/2 l + m + 5/2 × m + 3/2. ms ket l. 2l + 1 (6. −1/2 . l + 1/2 . (5. m − 1/2. but not when j = l − 1/2. It follows from Eq. l. this factor is taken to be unity. This value is possible when j = l + 1/2. 1/2l + 1/2. Thus. l and 1/2. 1/2l + 1/2. 2l + 1 (6. 1/2l + 1/2.113)
. For instance. respectively. Thus.6. the ml . l + 1/2.282). m = l + m + 1/2 l. shows that the plus sign is appropriate. Thus. l + 1/2 . A careful examination of the recursion relation. Eq. 1/2 must be equal to the j. 1/2 2l + 1 l − m + 1/2 m + 1/2.112) l + m + 1/2 . 2l + 1 2l + 1 (6. (6. m − 1/2. 1/2l + 1/2. 1/2l + 1/2. By convention.
ϕ) χ− 2l + 1
m−1/2 . Rnl (r). m .114) 2l + 1 It is convenient to deﬁne so called spinangular functions using the Pauli twocomponent formalism: Yl
j=l±1/2. 2 giving L·S l + 1/2.m
= ± +
l ± m + 1/2 m−1/2 Yl (θ. ϕ) χ+ 2l + 1 l m + 1/2 m+1/2 Yl (θ. mj = m = [j (j + 1) − l (l + 1) − 3/4] j. the quantum number j (i.
These functions are eigenfunctions of the total angular momentum for spin onehalf particles. = 2 (l + 1) ¯ 2 h l − 1/2.102). ¯2 h L·S j = l ± 1/2. 1/2 2l + 1
l + m + 1/2 m + 1/2.m . just as the spherical harmonics are eigenfunctions of the orbital angular momentum..118) (6. A general wavefunction for an energy eigenstate in a hydrogenlike atom is written The radial part of the wavefunction. −1/2 . m L·S l − 1/2. For a given choice of l. (6. m . m l¯2 h l + 1/2.116)
The l ± 1/2. m
= − +
l − m + 1/2 m − 1/2. m . ϕ) 1 ± l ± m + 1/2 Yl = √ m+1/2 2l + 1 l m + 1/2 Yl (θ.115)
ψnlm± = Rnl (r) Y j=l±1/2.117)
(6.7 Fine structure
6 APPROXIMATION METHODS
l − 1/2. (6. which is the quantum number associated with Jz . ϕ)
(6.6. m kets are eigenstates of L·S. = − 2
138
(θ.
(6. Thus.e. depends on the radial quantum number n and the angular quantum number l. (6. the quantum number associated with J 2 ) can take the values l ± 1/2. The wavefunction is also labeled by m.119)
. according to Eq.
121)
l−1/2. 2
(6. (6. The closest (in energy) unoccupied state is 3p. (6.m
l¯2 h . dΩ = 2 (l + 1) ¯ 2 h dΩ = − .125)
Let us now apply the above result to the case of a sodium atom. r dr r dr Equations (6. (6. In chemist’s notation.m †
l−1/2. Let us now apply degenerate perturbation theory to evaluate the shift in energy of a state whose wavefunction is ψnlm± due to the spinorbit Hamiltonian HLS . This state has a higher energy than 3s due to the deviations of the potential from the pure
139
.121) yield ∆Enlm+ ∆Enlm− where 1 dV l ¯ 2 h 1 .124) are known as Lande’s interval rule.m †
) L·S Y ) L·S Y
l+1/2. To ﬁrstorder. Equations (6.m
where the integrals are over all solid angle.123)–(6. and (6. We are interested in the excitation of the eleventh electron from 3s to some higher energy state.116).126)
The inner ten electrons effectively form a spherically symmetric electron cloud.
(6.123) (6. the ground state is written (1s)2 (2s)2 (2p)6 (3s). 2 me2 r dr 2 (6.124)
1 dV 1 dV = (Rnl )∗ Rnl r2 dr. the energyshift is given by ∆Enlm± = (ψnlm± )† HLS ψnlm± dV.6.120) (6.100) (remember the factor of two).122)
where the integral is over all space.120)–(6.7 Fine structure
6 APPROXIMATION METHODS
It follows that (Y (Y
l+1/2. = + 2 me2 r dr 2 1 1 dV (l + 1) ¯ 2 h = − .
127)
is the ﬁne structure constant. we can treat H B as
140
.125) is positive.129) 2 me is its magnetic moment.123)–(6. The modiﬁed states are labeled (3p)1/2 and (3p)3/2 . The wellknown sodium D line is associated with transitions between the 3p and 3s states.8 The Zeeman effect Consider a hydrogenlike atom placed in a uniform zdirected magnetic ﬁeld. The four (3p)3/2 states lie at a slightly higher energy level than the two (3p)1/2 states.6. including both the spin and orbital contributions. where 1 e2 = α= 2 0 h c 137 (6. The fact that there are two slightly different 3p energy levels (note that spinorbit coupling does not split the 3s energy levels) means that the sodium D line actually consists of two very closely spaced spectroscopic lines.
6.128)
e (L + 2 S) (6. where the subscript refers to the value of j. since we have neglected an effect (namely. In the absence of spinorbit interaction.8 The Zeeman effect
6 APPROXIMATION METHODS
Coulomb form. 2 me (6. there are six degenerate 3p states. The splitting of the (3p) energy levels of the sodium atom can be observed using a spectroscope. The change in energy of the outermost electron is HB = −µ·B. (6. HB = eB (Lz + 2 Sz ). In this situation. It is easily demonstrated that the ratio of the typical spacing of Balmer lines to the splitting brought about by spinorbit interaction is about 1 : α2 . Note that Eqs. Thus.130)
Suppose that the energyshifts induced by the magnetic ﬁeld are much smaller than those induced by spinorbit interaction. where µ=− (6. because the radial integral (6. The spinorbit interaction breaks the degeneracy of these states.124) are not entirely correct. the relativistic mass correction of the electron) which is the same order of magnitude as spinorbit coupling.
136). 1/2 2l + 1 l m + 1/2 m + 1/2. −1/2 .6. h 2 me (6. 2l + 1 (6. with four j = 3/2 states lying at a slightly higher energy than two j = 1/2 states. According to Eq. where j = l ± 1/2. m ) . mHB l ± 1/2. (6.135) l ± m + 1/2 m − 1/2.133)
Now.8 The Zeeman effect
6 APPROXIMATION METHODS
Since
a small perturbation acting on the eigenstates of H0 + HLS . m . these states are the simultaneous eigenstates of J 2 and Jz . we obtain Lande’s formula for the energyshift induced by a weak magnetic ﬁeld: e¯ B h 1 ∆Enlm± = m 1± . (6. The spinorbit interaction splits the six 3p states into two groups. mSz l ± 1/2. a magnetic ﬁeld splits the (3p)3/2
141
.134)
Thus. m = ¯ h [(l ± m + 1/2) − (l 2 (2 l + 1) m¯ h . We have already seen that the nonCoulomb potential splits the degeneracy of the 3s and 3p states. l ± 1/2.113)–(6. the latter states acquiring a higher energy. from Eqs. Let us consider one of these states. From standard perturbation theory. labeled by the quantum numbers j and m.132)
we ﬁnd that ∆Enlm± = eB (m ¯ + l ± 1/2.131) Lz + 2 S z = Jz + Sz . (6. mSz l ± 1/2. = ± 2l + 1 m + 1/2)] (6. (6. Of course. m = ± + It follows that l ± 1/2. (6. the ﬁrstorder energyshift in the presence of a magnetic ﬁeld is ∆Enlm± = l ± 1/2.136) 2 me 2l + 1 Let us apply this theory to the sodium atom.114).
Thus. Note that the magnetic Hamiltonian. this intermediate case is very difﬁcult to analyze. This is called the PaschenBack limit. l. = 2 me
(6. the six 3p states form two groups of four and two states. ms e¯ B h (ml + 2 ms ). l. Suppose that we increase the strength of the magnetic ﬁeld. J 2 . depending on the values
142
. S 2 . ms . ml . We can label each state by the quantum numbers n (the energy quantum number). The unperturbed Hamiltonian. so that the energyshift due to the magnetic ﬁeld becomes comparable to the energyshift induced by spinorbit interaction. as a small interaction term operating on the eigenstates of H0 + HB . our energy eigenkets are written n. H0 . A magnetic ﬁeld also splits the (3p)1/2 doublet of states. ml . l. Let us consider the extreme limit in which the energyshift due to the magnetic ﬁeld greatly exceeds that induced by spinorbit effects. HB . ml . The shift in energy due to the magnetic ﬁeld is simply ∆Enlml ms = n. the energy of each state becomes dependent on the quantum number m. it is evident from Eq.8 The Zeeman effect
6 APPROXIMATION METHODS
quadruplet of states. and ms . the energy eigenstates of a hydrogenlike atom are approximate eigenstates of the spin and orbital angular momenta. Thus.136) that these states are split by a lesser amount than the j = 3/2 states. in an intense magnetic ﬁeld. However. which measures the projection of the total angular momentum along the zaxis. Let us apply this result to a sodium atom. ml . In fact. but different values of ml and ms . are degenerate. in this situation. In the PaschenBack limit we can think of the spinorbit Hamiltonian.6. but does not commute with L2 . Lz . (6.137)
Thus. S 2 . each state acquiring a different energy. H LS . states with the same value of n and l. ms HB n. States with higher m values have higher energies. but are not eigenstates of the total angular momentum. Sz . causes states with different values of the quantum numbers n and l to have different energies. l. In the absence of a magnetic ﬁeld. In fact. commutes with L2 . states with different values of ml + 2 ms acquire different energies. Jz . it does not make much sense to think of HB as a small interaction term operating on the eigenstates of H0 + HLS . Clearly. However.
the energy of the doublet. the 3p states are split into ﬁve groups with (ml . respectively.142) 2 me2 r dr Let us apply the above result to a sodium atom. L·S = Lz Sz + (L+ S− + L− S+ )/2 (6. −1/2).141) = ¯ 2 ml ms . h The energyshift induced by the spinorbit Hamiltonian is given by ∆Enl ml ms = n. ml . ms ) quantum numbers (1. the energy difference between adjacent groups being e ¯ B/2 me . There is a state with ml + 2 ms = 2. in order of decreasing energy. h since L± = S ± = 0 for expectation values taken between the simultaneous eigenkets of Lz and Sz .8 The Zeeman effect
6 APPROXIMATION METHODS
of their total angular momentum.139)
1 1 dV L·S. and a state with ml + 2 ms = −2. (0. ml . (6. −1/2). 1/2). does not affect the next lowest energy state. Thus. The sort of magnetic ﬁeldstrength needed to get into the PaschenBach limit is given by e me 25 tesla. or (−1. ms . The spinorbit term increases the energy of the highest energy state. (6.138) (6. In the presence of an intense magnetic ﬁeld the 3p states are split into ﬁve groups. decreases. The net result is that the ﬁve groups of states are no longer equally spaced in energy. where HLS = Now. and increases the energy of the lowest energy state. but does not split. In the presence of an intense magnetic ﬁeld. (1. and (−1. ¯ 2 ml ms 1 dV h ∆Enlml ms = . a state with ml + 2 ms = −1. two states with ml + 2 ms = 0. 1/2).140) (6. (0. ms HLS n. l. does not affect the next highest energy state. These groups are equally spaced in energy. a state with ml + 2 ms = 1. 2 me2 r dr
. l.6. −1/2).143) BPB ∼ α2 0 h a0
143
(6. 1/2).
6.9 Timedependent perturbation theory
6 APPROXIMATION METHODS
Obviuously, this is an extremely large ﬁeldstrength.
6.9 Timedependent perturbation theory Suppose that the Hamiltonian of the system under consideration can be written H = H0 + H1 (t), (6.144)
where H0 does not contain time explicitly, and H1 is a small timedependent perturbation. It is assumed that we are able to calculate the eigenkets of the unperturbed Hamiltonian: H0 n = En n . (6.145) We know that if the system is in one of the eigenstates of H0 then, in the absence of the external perturbation, it remains in this state for ever. However, the presence of a small timedependent perturbation can, in principle, give rise to a ﬁnite probability that a system initially in some eigenstate i of the unperturbed Hamiltonian is found in some other eigenstate at a subsequent time (since i is no longer an exact eigenstate of the total Hamiltonian), In other words, a timedependent perturbation causes the system to make transitions between its unperturbed energy eigenstates. Let us investigate this effect. Suppose that at t = t0 the state of the system is represented by A =
n
cn n ,
(6.146)
where the cn are complex numbers. Thus, the initial state is some linear superposition of the unperturbed energy eigenstates. In the absence of the timedependent perturbation, the time evolution of the system is given by A, t0 , t =
n
cn exp([−i En (t − t0 )/¯ ] n . h
(6.147)
Now, the probability of ﬁnding the system in state n at time t is Pn (t) = cn exp[−i En (t − t0 )/¯ ]2 = cn 2 = Pn (t0 ). h
144
(6.148)
6.9 Timedependent perturbation theory
6 APPROXIMATION METHODS
Clearly, with H1 = 0, the probability of ﬁnding the system in state n at time t is exactly the same as the probability of ﬁnding the system in this state at the initial time t0 . However, with H1 = 0, we expect Pn (t) to vary with time. Thus, we can write A, t0 , t = cn (t) exp[−i En (t − t0 )/¯ ] n , h (6.149)
n
where Pn (t) = cn (t) . Here, we have carefully separated the fast phase oscillation of the eigenkets, which depends on the unperturbed Hamiltonian, from the slow variation of the amplitudes cn (t), which depends entirely on the perturbation (i.e., cn is constant if H1 = 0). Note that in Eq. (6.149) the eigenkets n are timeindependent (they are actually the eigenkets of H0 evaluated at the time t0 ). Schr¨dinger’s time evolution equation yields o ∂ A, t0 , t = H A, t0 , t = (H0 + H1 ) A, t0 , t . ∂t It follows from Eq. (6.149) that i¯ h (H0 + H1 )A, t0 , t =
m
2
(6.150)
cm (t) exp[−i Em (t − t0 )/¯ ] (Em + H1 ) m . h
(6.151)
We also have ∂ i ¯ A, t0 , t = h ∂t
i¯ h
m
dcm + cm (t) Em exp[−i Em (t − t0 )/¯ ] m , h dt
(6.152)
where use has been made of the timeindependence of the kets m . According to Eq. (6.150), we can equate the righthand sides of the previous two equations to obtain dcm exp[−i Em (t − t0 )/¯ ]m = h cm (t) exp[−i Em (t − t0 )/¯ ] H1 m . h i¯ h dt m m (6.153) Leftmultiplication by n yields i¯ h where Hnm (t) = nH1 (t)m ,
145
dcn = dt
Hnm (t) exp[i ωnm (t − t0 )] cm (t),
m
(6.154)
(6.155)
6.10 The twostate system
6 APPROXIMATION METHODS
En − E m . (6.156) ¯ h Here, we have made use of the standard orthonormality result, nm = δ nm . Suppose that there are N linearly independent eigenkets of the unperturbed Hamiltonian. According to Eq. (6.154), the time variation of the coefﬁcients cn , which specify the probability of ﬁnding the system in state n at time t, is determined by N coupled ﬁrstorder differential equations. Note that Eq. (6.154) is exact—we have made no approximations at this stage. Unfortunately, we cannot generally ﬁnd exact solutions to this equation, so we have to obtain approximate solutions via suitable expansions in small quantities. However, for the particularly simple case of a twostate system (i.e., N = 2), it is actually possible to solve Eq. (6.154) without approximation. This solution is of enormous practical importance. ωnm =
and
6.10 The twostate system Consider a system in which the timeindependent Hamiltonian possesses two eigenstates, denoted H0 1 H0 2 = E1 1 , = E2 2 . (6.157) (6.158)
Suppose, for the sake of simplicity, that the diagonal matrix elements of the interaction Hamiltonian, H1 , are zero: 1H1 1 = 2H1 2 = 0. (6.159)
The offdiagonal matrix elements are assumed to oscillate sinusoidally at some frequency ω: 1H1 2 = 2H1 1 ∗ = γ exp(i ω t), (6.160) where γ and ω are real. Note that it is only the offdiagonal matrix elements which give rise to the effect which we are interested in—namely, transitions between states 1 and 2.
146
165)
Now.162) can be combined to give a secondorder differential equation for the time variation of the amplitude c2 : dc2 γ2 d2 c 2 + i (ω − ω21 ) + 2 c2 = 0. i¯ h dt i¯ h (6. we can use Eq. h
(6. and assuming that t0 = 0.162)
where ω21 = (E2 − E1 )/¯ . It follows that γ2 /¯ 2 h P2 (t) = 2 2 γ /¯ + (ω − ω21 )2 /4 h
147
.161) (6. Likewise. (6. (6. Equations (6. the probability of ﬁnding the system in state 1 at time t is simply P 1 (t) = c1 2 .10 The twostate system
6 APPROXIMATION METHODS
For a twostate system. the probability of ﬁnding the system in state 2 at time t is P 2 (t) = c2 2 . h γ2 /¯ 2 + (ω − ω21 )2 /4 t h exp[ i (ω − ω21 ) t/2]
c1 (t) = exp[ i (ω − ω21 ) t/2] cos − i (ω − ω21 )/2 γ2 /¯ 2 + (ω − ω21 )2 /4 h
× sin
γ2 /¯ 2 + (ω − ω21 )2 /4 t . our boundary conditions are c1 (0) = 1 and c2 (0) = 0. Thus. It is easily demonstrated that the appropriate solutions are c2 (t) = −i γ/¯ h γ2 /¯ 2 + (ω − ω21 )2 /4 h × sin exp[−i (ω − ω21 ) t/2] (6.164)
γ2 /¯ 2 + (ω − ω21 )2 /4 t . dt dc2 = γ exp[−i (ω − ω21 ) t ] c1 .154) reduces to dc1 = γ exp[+i (ω − ω21 ) t ] c2 . dt2 dt ¯ h (6.161) and h (6. Eq.163)
Once we have solved for c2 .6.162) to obtain the amplitude c1 . Let us look for a solution in which the system is certain to be in state 1 at time t = 0.
ω. After a further time interval h π ¯ /2 γ it is certain to be in state 1.166) (6. This implies that the system alternatively absorbs and emits energy from the source of the perturbation. At resonance. h P2 (t) = sin2 (γ t/¯ ). the system periodically h ﬂipﬂops between states 1 and 2 under the inﬂuence of the timedependent perturbation. and so on. Thus.
148
. if the applied frequency differs from the resonant frequency by h substantially more than 2 γ/¯ then the probability of the system jumping from h state 1 to state 2 is very small. matches the frequency ω21 . if we plot the maximum value of P2 (t) as a function of the applied frequency. This result is known as Rabi’s formula. However.167)
P1 (t) = 1 − P2 (t). The absorptionemission cycle also take place away from the resonance. the timedependent perturbation is only effective at causing transitions between states 1 and 2 if its frequency of oscillation lies in the approximate range ω21 ± 2 γ/¯ .166) exhibits all the features of a classic resonance.e.6. the narrower the resonance.168) (6. h (6. the weaker the h perturbation (i. This means that the maximum value of P2 (t) is no longer unity. In fact. Clearly. the system starts off at t = 0 in state 1. ω. when ω = ω21 . After a time interval π ¯ /2 γ it is certain to be in state 2.10 The twostate system
6 APPROXIMATION METHODS
× sin2
γ2 /¯ 2 + (ω − ω21 )2 /4 t .169)
According to the above result. the amplitude of oscillation of the coefﬁcient c2 is reduced.
Equation (6. h
(6. In other words. when the oscillation frequency of the perturbation. Thus. nor is the minimum value of P1 (t) zero. and whose fullwidth halfmaximum (in frequency) is 4 γ/¯ . we obtain a resonance curve whose maximum (unity) lies at the resonance. we ﬁnd that P1 (t) = cos2 (γ t/¯ ). the smaller γ becomes)..
H0 ± = ± e ¯ B0 h ± . Thus.6.176) (6. and −H1 + = +H1 −
∗
(6.11 Spin magnetic resonance
6 APPROXIMATION METHODS
6. with B1 B0 . respectively.g. 2 me (6. (6. ^ ^ z B = B0 ^ + B1 (cos ωt x + sin ωt y). In this system.175)
where S+ and S− are the conventional raising and lowering operators for the spin angular momentum.173)
The eigenstates of the unperturbed Hamiltonian are the ‘spin up’ and ‘spin down’ states.171) (6.170)
where B0 and B1 are constants. The rotating magnetic ﬁeld usually represents the magnetic component of an electromagnetic wave propagating along the zaxis. It follows that +H1 + = −H1 − = 0. me
(6. The Hamiltonian is written H = −µ·B = H0 + H1 . and then subjected to a small timedependent magnetic ﬁeld rotating in the xy plane.174)
The timedependent Hamiltonian can be written H1 = e B1 exp( i ωt) S− + exp(−i ωt) S+ . 2 me
149
. Thus. me (6.177)
=
e ¯ B1 h exp( i ωt). 2 me (6. where H0 = and H1 = e B0 Sz .172)
e B1 (cos ωt Sx + sin ωt Sy ) .. a bound electron) placed in a uniform zdirected magnetic ﬁeld. the electric component of the wave has no effect. denoted + and − .11 Spin magnetic resonance Consider a spin onehalf system (e.
By determining the resonant frequency (i. in addition to the spin precession. (6. 2 me
6.181)
The resonant frequency. e B0 .154) for a general system. In the absence of the perturbation. is simply the spin precession frequency for an electron in a uniform magnetic ﬁeld of strength B0 . ω21 . which is deﬁned A. t = U(t0 .12 The Dyson series Let us now try to ﬁnd approximate solutions of Eq. (6. ω21 → me e ¯ B1 h γ → . it is possible to calculate the magnetic moment. and subjecting them to an oscillating magnetic ﬁeld whose frequency is gradually scanned. U(t 0 . It is convenient to work in terms of the time evolution operator. + − .182)
150
.
→ + . and other spin onehalf particles.179) (6.180) (6. the frequency at which the particles absorb energy from the oscillating ﬁeld). but the expectation value of Sz remains invariant. the system undergoes a succession of spinﬂops. t) A .. We also know that if the oscillation frequency of the applied ﬁeld is very different from the resonant frequency then there is virtually zero probability of the ﬁeld triggering a spinﬂop. Experimentalist are able to measure the magnetic moments of electrons. provided that we make the identiﬁcations 1 2 → − . (6. If we now apply a magnetic perturbation rotating at the resonant frequency then. to a high degree of accuracy by placing the particles in a magnetic ﬁeld.6. t). the expectation values of Sx and Sy oscillate because of the spin precession. The width of the resonance (in frequency) is determined by the strength of the oscillating magnetic perturbation. t0 .178) (6.12 The Dyson series
6 APPROXIMATION METHODS
It can be seen that this system is exactly the same as the twostate system discussed in the previous section.e. according to the analysis of the previous section.
(6. The subsequent evolution of the state ket is given by Eq. Thus.6.183)
(6. (6. ∂t subject to the boundary condition i¯ h U(t0 .187) (6. t). t is the state ket of the system at time t.184)
In the absence of the external perturbation. t) = HI (t0 . t0 . h (6.186)
Note that UI speciﬁes that component of the time evolution operator which is due to the timedependent perturbation.189) (6.190)
m
151
. h h subject to the boundary condition UI (t0 . t) = exp[−i H0 (t − t0 )/¯ ] UI (t0 . t) = exp[+i H0 (t − t0 )/¯ ] H1 exp[−i H0 (t − t0 )/¯ ]. t) UI (t0 . h (6. t0 .188) ∂UI (t0 . t = cm (t) exp[−i Em (t − t0 )/¯ ] m . A. h It is readily demonstrated that UI satisﬁes the differential equation i¯ h where HI (t0 . given that the state ket at the initial time t0 is A . t) = (H0 + H1 ) U(t0 .185) Let us switch on the perturbation and look for a solution of the form U(t0 . It is easily seen that the time evolution operator satisﬁes the differential equation ∂U(t0 . ∂t (6. i. t). we would expect UI to contain all of the information regarding transitions between different eigenstates of H 0 caused by the perturbation. t0 ) = 1. t0 ) = 1. Suppose that the system starts off at time t0 in the eigenstate i of the unperturbed Hamiltonian. t). t) = exp[−i H0 (t − t0 )/¯ ].12 The Dyson series
6 APPROXIMATION METHODS
Here.149). (6. the time evolution operator reduces to U(t0 .
Note that the differential equation (6.187). It follows from Eqs. t0 . t ) dt + · · · . etc.197)
. Thus. i UI (t0 . h It follows that cn (t) = nUI (t0 .191)
HI (t0 .6. t ) dt
HI (t0 . are equivalent to the following integral equation. is simply Pi→n (t0 . t ) UI (t0 . t ) HI (t0 . n n n (6.194)
We can obtain an approximate solution to this equation by iteration: UI (t0 . t) i . t) =  nUI (t0 . plus the boundary condition (6.192) and (6. t ) dt . t ) 1 − ¯ h t0
t t t t0
HI (t0 .195) that c(0) (t) = δin . n
152
(6. t ) dt
t0 2 t t
dt
t0 t0
HI (t0 . (6. t = exp[−i H0 (t − t0 )/¯ ] UI (t0 . given that it is deﬁnitely in state i at time t 0 . the probability that the system is found in state n at time t. t)i 2 .195)
This expansion is known as the Dyson series.196)
where the superscript (1) refers to a ﬁrstorder term in the expansion.
t0
(6. t) = 1 − ¯ h
t
(6. Let cn = c(0) + c(1) + c(2) + · · · .12 The Dyson series
6 APPROXIMATION METHODS
However.192) where use has been made of nm = δnm .
(6.189).193) This quantity is usually termed the transition probability between states i and n . we also have i. t) i 1− ¯ h i 1− ¯ h −i + ¯ h i HI (t0 . t ) UI (t0 . (6. (6. t)i .
there is no chance of a transition between states i and n (i = n) to zerothorder (i. Thus.6.
These expressions simplify to
(0) cn (t) = δin . ¯ h
(6.200)
t
i = − ¯ h = −i ¯ h
exp[ i ωni (t − t0 )] Hni (t ) dt .12 The Dyson series
6 APPROXIMATION METHODS
c(1) (t) n c(2) (t) n
i = − ¯ h = −i ¯ h
t
nHI (t0 . weighted by some oscillatory phasefactor.202)
×Hnm (t ) exp[ i ωmi (t − t0 )] Hmi (t ).
t0 2 t t
(6.205)
According to the above analysis.
153
. n n n En − E m . To ﬁrstorder.201)
(2) cn (t)
dt
t0 t0
dt exp[ i ωnm (t − t0 )] (6. t ) HI (t0 .199)
dt
t0 t0
nHI (t0 .204)
(6. (1) cn (t)
(6.e.203) (6. t )i dt . in the absence of the perturbation). t )i dt .198) (6. then there is no chance of a ﬁrstorder transition between states i and n . The transition probability between states i and n is simply Pi→n (t0 .. where ωnm = and Hnm (t) = nH1 (t)m . a transition between states i and n is possible even when the matrix element nH1 i is zero. t) = c(0) + c(1) + c(2) + · · · 2 . to secondorder. However. the transition probability is proportional to the time integral of the matrix element nH1 i .
t0 2 m t t
(6. if the matrix element is zero.
211) x The sinc function is highly oscillatory.202) (with t0 = 0). sinc2 2¯ h ¯2 h
(6.213)
. En − E i
(6. is small except when 2π ¯ h . According to Eqs. (6.209) Pi→n (t) cn  = En − Ei 2 2¯ h for i = n.207)
(6. provided that sinc(x) = ∆E ∆t < ¯ . Pi→n .208)
giving (En − Ei ) t 4 Hni 2 (1) 2 sin2 . (6.206)
where H1 is timeindependent. n c(1) (t) n
t i = − Hni exp[ i ωni (t − t)] dt ¯ h 0 Hni = [1 − exp( i ωni t)]. En = Ei ) have an appreciable probability of occurrence. but is generally a function of the position.6. (6.210)
sin x . The transition probability between states i and n can be written Hni 2 t2 (En − Ei ) t Pi→n (t) = . c(0) (t) = δin . is is possible to have transitions which do not exactly conserve energy. At ﬁnite t. ∼ h
154
where
(6. and decays like 1/x at large x.e. momentum. (6. and spin operators.13 Constant perturbations
6 APPROXIMATION METHODS
6.13 Constant perturbations Consider a constant perturbation which is suddenly switched on at time t = 0: H1 (t) = 0 H1 (t) = H1 for t < 0 for t ≥ 0.200)–(6..212) En − Ei  < ∼ t Note that in the limit t → ∞ only those transitions which conserve energy (i. Suppose that the system is deﬁnitely in state i at time t = 0. (6. It is a good approximation to say that sinc(x) is small except when x < π. It follows ∼ that the transition probability.
we can take ρ(En ) and Hni 2 out of the integral in the above formula to obtain Pi→[n] (t) = 2π Hni 2 ρ(En ) t ¯ h En
155
Pi→n (t) ρ(En ) dEn . and ∆t = t is the time elapsed since the perturbation was switched on. the probability of a transition from the initial state i to any of the continuum of possible ﬁnal states is Pi→ (t) = giving Pi→ (t) = where x = (En − Ei ) t/2 ¯ .210). despite the fact that H1 is constant for t > 0. where the number of ﬁnal states lying in the energy range E to E + dE is given by ρ(E) dE.
(6. It is helpful to deﬁne the density of states.. since it implies that the probability of a transition occurring in a ﬁxed time interval.216)
. t to t + dt. Note that this probability grows quadratically with time. there is usually a group of ﬁnal states.e. The probability of a transition which conserves energy (i. in this limit.218)
. this result is just a manifestation of the wellknown uncertainty relation for energy and time. grows linearly with t.214) ¯2 h where use has been made of sinc(0) = 1. since in nonrelativistic quantum mechanics position and momentum are operators. It follows that. all possessing nearly the same energy as the energy of the initial state i . Clearly. This uncertainty relation is fundamentally different to the positionmomentum uncertainty relation. whereas time is merely a parameter.6. ¯ h
(6.
Ei
(6. We know that in the limit t → ∞ the function sinc(x) is only nonzero in an inﬁnitesimally narrow range of ﬁnal energies centred on En = Ei .13 Constant perturbations
6 APPROXIMATION METHODS
where ∆E = En − Ei  is change in energy of the system associated with the transition. Thus. In practice.217) and use has been made of Eq. (6. This result is somewhat surprising. h (6. (6. ρ(E). E n = Ei ) is Hin 2 t2 Pi→n (t) = .215)
2t Hni 2 ρ(En ) sinc2 (x) dx.
Hni 2 is the average of Hni 2 over all ﬁnal states with approximately the same energy as the initial state.e. (6. using the constant perturbation (6. we have made use of the result ∞ sinc2 (x) dx = π. From Eq. wi→n = (6. In deriving the above formula. Note that the transition rate is constant in time (for t > 0): i. (6. dt (6. is now proportional to t..13 Constant perturbations
6 APPROXIMATION METHODS
where Pi→[n] denotes the transition probability between the initial state i and all ﬁnal states n which have approximately the same energy as the initial state.202) we ﬁnd that
(2) cn (t)
=
−i ¯ h
2
t
t
Hnm Hmi
m 0 t
dt exp( i ωnm t )
0
dt exp( i ωmi t )
i = ¯ h = it ¯ h
m
Hnm Hmi Em − E i
[exp( i ωni t ) − exp( i ωnm t ] ) dt
0
m
Hnm Hmi [exp( i ωni t/2) sinc(ωni t/2) Em − E i
156
. ¯ h where it is understood that this formula must be integrated with obtain the actual transition rate.220)
Let us now calculate the secondorder term in the Dyson series. wi→[n] = giving 2π Hni 2 ρ(En ) . instead of t2 . which is simply the transition probability per unit time. Pi→[n] . the probability of a transition occurring in the time interval t to t + dt is independent of t for ﬁxed dt. It is convenient to deﬁne the transition rate. (6.206). Fermi’s golden rule is sometimes written wi→[n] = 2π Hni 2 δ(En − E). Here. Thus.222) ρ(En ) dEn to dPi→[n] .221) ¯ h En E i This appealingly simple result is known as Fermi’s golden rule.6.219)
−∞
Note that the transition probability.
223)
+
m
Hnm Hmi exp( i ωim t/2) sinc(ωnm t/2) .225) from Ei − Em to Ei − Em + i ¯ η. It follows. conserves energy. cn (t) = c(1) n + c(2) n it exp( i ωni t/2) = ¯ h −
m H ni
(6. This problem can be avoided by gradually turning on the perturbation: i. The net result is to change the energy denominator in Eq. with approximately the same energy as the initial state. Em − E i
Hnm Hmi sinc(ωni t/2) Em − E i
(6. and for intermediate states.224)
where use has been made of Eq. The nonenergyconserving transitions are generally termed virtual transitions. by analogy with the previous analysis. H1 → exp(η t) H1 (where η is very small). whereas the energy conserving ﬁrstorder transition is termed a real transition. First. the system makes another nonenergyconserving transition to the ﬁnal state n . that wi→[n] 2π = Hni + ¯ h Hnm Hmi ρ(En ) Em − E i
2
.225)
m
where the transition rate is calculated for all ﬁnal states. from i to n . h
157
.e. the system makes a nonenergyconserving transition to some intermediate state m .13 Constant perturbations
6 APPROXIMATION METHODS
− exp( i ωnm t/2) sinc(ωnm t/2)] . Subsequently. (6. n . Thus.224) to average to zero (due to the oscillatory phasefactor) during the evaluation of the transition probability. (6. According to Eq. a secondorder transition takes place in two steps. i . The above formula clearly breaks down if Hnm Hmi = 0 when Em = Ei .
En E i
(6.208). (6. (6.225). m whose energies differ from that of the initial state.. The fact that Em = Ei causes the last term on the righthand side of Eq. The net transition.6.
in general. ¯ h 0 or En (6.227). whilst making a h transition to a ﬁnal state whose energy level is less than that of the initial state
158
.231) corresponds to the ﬁrst term on the righthand side of Eq. ¯ h ωni + ω ωni − ω where Vni =
† Vni =
nVi . (6. H0 . Let us initiate the system in the eigenstate i of the unperturbed Hamiltonian. it follows from the previous analysis that the transition probability P i→n (t) = c(1) 2 is only appreciable in the limit t → ∞ if n ωni + ω Ei − ¯ ω. (6.231) ωni − ω 0 or En Ei + ¯ ω. nV † i = iVn ∗ . (6.
(6. provided that ωni = En − E i → ωni ± ω. h (6.14 Harmonic perturbations Consider a perturbation which oscillates sinusoidally in time.232)
Clearly.232) corresponds to the second term. It follows from Eq.201) that c(1) n −i = ¯ h
t 0 † Vni exp(i ωt ) + Vni exp(−i ωt ) exp( i ωni t ) dt . H1 (t) = V exp( i ωt) + V † exp(−i ωt).6.208). This is usually called a harmonic perturbation.226)
where V is.228) (6. a function of position. and (6.14 Harmonic perturbations
6 APPROXIMATION METHODS
6. h (6. Thus.230)
Thus. (6. and switch on the harmonic perturbation at t = 0.227)
1 1 − exp[ i (ωni + ω) t] 1 − exp[ i (ωni − ω) t] † = Vni + Vni . The former term describes a process by which the system gives up energy ¯ ω to the perturbing ﬁeld.229)
This formula is analogous to Eq. momentum. and spin operators. (6.
(6.
234)
Equation (6.237) ρ(En ) ρ(Ei ) In other words. wi→[n] = wi→[n] 2π Vni 2 ρ(En ) .e. h ¯ h 2π † 2 = V  δ(En − Ei − ¯ ω). This result. (6.228)(6. This process is known as absorption.6. divided by the density of ﬁnal states for absorption.233)–(6.233) speciﬁes the transition rate for stimulated emission. the total energy h (i. V  ρ(En ) = ¯ h ni En =Ei +¯ ω h (6. (6. h ¯ h ni (6..e.234) gives the transition rate for absorption.
6. and is very important in statistical mechanics.15 Absorption and stimulated emission of radiation
6 APPROXIMATION METHODS
by ¯ ω. divided by the density of ﬁnal states for stimulated emission. (6. These equations are more usually written wi→n = wi→n 2π Vni 2 δ(En − Ei + ¯ ω). the rate of stimulated emission. that of the system plus the perturbing ﬁeld) is conserved. equals the rate of absorption.233) (6. is known as detailed balancing. In both cases.236)
† It is clear from Eqs.15 Absorption and stimulated emission of radiation Let us use some of the results of timedependent perturbation theory to investigate the interaction of an atomic electron with classical (i.229) that Vni 2 = Vni 2 . (6. which expresses a fundamental symmetry between absorption and stimulated emission.234) that wi→[n] wn→[i] = . It follows from Eqs. ¯ h En =Ei −¯ ω h 2π † 2 .235) (6. (6.. The latter term describes h a process by which the system gains energy ¯ ω from the perturbing ﬁeld. nonquantized)
159
. This process is known as stimulated emission. By analogy with Eq.221). whilst h making a transition to a ﬁnal state whose energy level exceeds that of the initial state by ¯ ω. whereas Eq.
2 me me 2 me ·A = 0. The above equation can be written p2 − e A·p − e p·A + e2 A2 H= + e φ + V0 (r).241) E = − φ− ∂t B = × A.239) (6. (6. (6. for which φ = 0. A = 2 A0 cos
160
(6. (6. (6. Hence.15 Absorption and stimulated emission of radiation
6 APPROXIMATION METHODS
electromagnetic radiation.244) 2 me Now.240)
This prescription also works in quantum mechanics. (6.243) 2 me where A and φ are functions of the position operators.248)
.245) provided that we adopt the gauge p2 e A·p e2 A2 H= − + + e φ + V0 (r). (6. p → p + q A.247) ω n·r − ωt .
(6. p·A = A·p. Thus. (6.246)
Suppose that the perturbation corresponds to a monochromatic planewave.238) 2 me The standard classical prescription for obtaining the Hamiltonian of a particle of charge q in the presence of an electromagnetic ﬁeld is H0 = H → H − q φ. c (6. the Hamiltonian of an atomic electron placed in an electromagnetic ﬁeld is
where A(r) is the vector potential and φ(r) is the scalar potential.242)
(p − e A)2 H= + e φ + V0 (r).6. Note that ∂A . The unperturbed Hamiltonian is p2 + V0 (r).
that the ﬁrst term on the righth hand side of Eq. me (6.15 Absorption and stimulated emission of radiation
6 APPROXIMATION METHODS
where and n are unit vectors which specify the direction of polarization and the direction of propagation. (6. Now the energy density of an electromagnetic ﬁeld is 1 U= 2
2 0 E0
2
B02 + . whereas the second term describes the stimulated emission of a photon of energy ¯ ω. by analogy with the previous analysis. Note that ·n = 0.251)
where the A2 term.6. h 2 ¯ me h (6.249) with H0 = and H1 − p2 + V(r). The Hamiltonian becomes H = H0 + H1 (t). has been neglected.253)
It is clear.252) describes the absorption of a photon of energy ¯ ω.236) that the rate of absorption is h wi→n 2π e2 = A0 2  n exp[ i (ω/c) n·r] ·p i 2 δ(En − Ei − ¯ ω). provided that V =− e A0 ·p exp[−i (ω/c) n·r ] me (6. It follows from Eq.250)
(6.254)
The absorption crosssection is deﬁned as the ratio of the power absorbed by the atom to the incident power per unit area in the electromagnetic ﬁeld. me (6. 2 µ0
(6. (6. 2 me e A·p . (6.255)
161
. which is second order in A0 . The perturbing Hamiltonian can be written H1 = − e A0 ·p (exp[ i (ω/c) n·r − i ωt] + exp[−i (ω/c) n·r + i ωt]) . (6.252)
This has the same form as Eq. respectively.226).
This approximation is known as the electric dipole approximation. (6. Thus.262)
(6. (6. H0 ] = so npi = −i Using Eq.6.16 The electric dipole approximation In general. we obtain σabs = 4π2 α ωni  n ·ri 2 δ(ω − ωni ). H0 ]i = i me ωni nri . ¯ h
(6.256) Now.258). σabs = so σabs = π e2  n exp[ i (ω/c) n·r] ·p i 2 δ(En − Ei − ¯ ω).
162
· npi .261)
me n[r. The incident power per unit area of the electromagnetic ﬁeld is c U = 2 0 c ω2 A0 2 . exp[ i (ω/c) n·r] = 1 + i ω n·r + · · · . c (6. It follows that n exp[ i (ω/c) n·r] ·p i It is readily demonstrated that [r.260)
i¯ p h .
(6. respectively. the wavelength of the type of electromagnetic radiation which induces.263)
.258) ¯ ω wi→n h . transitions between different atomic energy levels is much larger than the typical size of a light atom.257)
6. unity (remember that ω/c = 2π/λ). me
(6. cU (6. h 2ωc 0 me (6.16 The electric dipole approximation
6 APPROXIMATION METHODS
where E0 and B0 = E0 /c = 2 A0 ω/c are the peak electric and magnetic ﬁeldstrengths.259)
can be approximated by its ﬁrst term. or is emitted during.
the next most likely type of transition is a magnetic dipole transition. so that = ^. ω.265)
∆m = 0.
These are termed the selection rules for electric dipole transitions. Magnetic dipole transitions are typically about 10 5 times more unlikely than similar electric dipole transitions. which is due to the interaction between the electron spin and the oscillating magnetic ﬁeld of the incident electromagnetic radiation. they take place at a far lower rate than transitions which are allowed according to the electric dipole approximation. Instead.268) (6. It is clear that if the absorption crosssection is regarded as a function of the applied frequency.
Thus. 6. then it exhibits a sharp maximum at ω = ωni = (En − Ei )/¯ .269) ∆m = 0.4. (6. from Sect. (6.
Here. for generally directed radiation n ·ri is only nonzero if (6.264) (6. ∆l = ±1. l is the quantum number describing the total orbital angular momentum of the electron. (6. and m is the quantum number describing the projection of the orbital angular momentum along the zaxis. h Suppose that the radiation is polarized in the zdirection.266) (6. It is easily demonstrated that nxi and nyi are only nonzero if ∆l = ±1. Forbidden transitions are not strictly forbidden.267)
∆m = ±1. We z have already seen. for instance. ±1. It is clear.6. The ﬁrstorder term in Eq. After electric dipole transitions. The latter transition is called a forbidden transition. but disallows a transition from a 2s to a 1s state. that the electric dipole approximation allows a transition from a 2p state to a 1s state.259) yields
163
.16 The electric dipole approximation
6 APPROXIMATION METHODS
where α = e2 /(2 0 h c) = 1/137 is the ﬁne structure constant. that nzi = 0 unless the initial and ﬁnal states satisfy ∆l = ±1.
(6. h ¯2 i[x.274)
This is known as the ThomasReicheKuhn sum rule.6. me Thus. It is easily demonstrated that ¯2 h [x. transitions which are forbidden as electric dipole transitions may well be allowed as magnetic dipole or electric quadrupole transitions. = me 2 0 me c (6. that the incident radiation is polarized in the xdirection. (6. [x.263) over all possible frequencies of the incident radiation yields σabs (ω) dω =
n
4π2 α ωni  n ·ri 2 . In fact.270)
Suppose. H0 ] ] = − .273)
It follows that
(6. [x.270) reduces to σabs (ω) dω = 2π2 α ¯ h π e2 . Integrating Eq. me 2 me ¯ h ωni  nxi 2 = 1. Magnetic dipole and electric quadrupole transitions satisfy different selection rules than electric dipole transitions: for instance.
164
. ±2. Eq. H0 ] ]i = ix H0 + H0 x − 2 x H0 xi = − .272)
giving 2
n
¯2 h ( ixn Ei nxi − ixn En nxi ) = − .271)
(6. the selection rules for electric quadrupole transitions are ∆l = 0. These are typically about 10 8 times more unlikely than electric dipole transitions.16 The electric dipole approximation
6 APPROXIMATION METHODS
socalled electric quadrupole transitions.
n
(6. According to this rule. Thus. the above formula is exh actly the same as that obtained classically by treating the electron as an oscillator.
(6. for the sake of deﬁniteness.275)
Note that ¯ has dropped out of the ﬁnal result. me
2 2
(6.
Let us now consider how the initial state becomes depopulated. Thus.6.201) that
(0) cn (t) = 0. In this limit. and cn=i (t → −∞) = 0. In the remote past.200)– (6. the system is assumed to be in the initial state i . (6. First.17 Energyshifts and decaywidths
6 APPROXIMATION METHODS
6. ¯ 2 η2 + ωni h
(6.277)
t
i = − Hni ¯ h
exp[ (η + i ωni )t ] dt
−∞
i exp[ (η + i ωni )t ] = − Hni . Basically.276) where η is small and positive. It is convenient to gradually turn on the perturbation from zero at t = −∞. becomes populated as a result of some timedependent perturbation applied to the system. dt η2 + ωni ¯2 h Consider the limit η → 0. however. It follows that. other than the initial state i . we want to calculate the time evolution of the coefﬁcient ci (t). η→0 η + ωni
165
Hni 2 exp(2 η t) 2 . ci (t → −∞) = 1. exp(η t) → 1.279)
(6.281)
.17 Energyshifts and decaywidths We have examined how a state n . ¯ h η + i ωni
(6. t → −∞. the transition probability from state i to state n is Pi→n (t) = c(1) 2 = n The transition rate is given by 2 Hni 2 η exp(2 η t) dPi→n = wi→n (t) = 2 . let us check that our previous Fermi golden rule result still applies when the perturbing potential is turned on slowly. For cn=i (t) we have from Eqs. instead of very suddenly. (6. (1) cn (t)
(6. and H1 is a constant.280)
(6.278)
where Hni = nH1 i . to ﬁrstorder. but η h lim 2 2 = π δ(ωni ) = π ¯ δ(En − Ei ). H1 (t) = exp(η t) H1 . Thus.
2 η (Ei − Em + i ¯ η) h m=i
Let us now consider the ratio ci /ci . to secondorder we have ci (t) exp(η t) −i −i Hii + 1+ ¯ h η ¯ h −i + ¯ h
2
Hii 2
exp(2 η t) 2 η2 (6. The width of the peak is determined by how fast the perturbation is switched on. Let us now calculate ci (t) using Eqs. Using Eq. where ci ≡ dci /dt. Hmi 2 exp(2 η t) . (6. (6. We obtain ci ˙ ci
−i −i Hii + ¯ h ¯ h 1−
2
Hii 2 −i + η ¯ h
i Hii ¯ η h
Hmi 2 Ei − E m + i ¯ η h m=i
−i −i Hii + lim η→0 ¯ ¯ h h
Hmi 2 .282) wi→n = ¯ h It is clear that the deltafunction in the above formula actually represents a function which is highly peaked at some particular energy. (6. ¯ h η
(6.202).286)
Hmi 2 exp(2 η t) . we ˙ ˙ can evaluate this ratio in the limit η → 0.280) yields the standard Fermi golden rule result 2π Hni 2 δ(En − Ei ). We have ci (t) = 1. (6.200)–(6. Eq.286).283)
t
i = − Hii ¯ h = −i ¯ h −i ¯ h
2
exp(η t ) dt = −
−∞ t t
i exp(η t) Hii .17 Energyshifts and decaywidths
6 APPROXIMATION METHODS
Thus. 2 η (η + i ωmi ) (6.287)
.285)
=
Thus. Ei − E m + i ¯ η h m=i
166
(6.6.284)
Hmi 
m
2 −∞
dt
−∞
dt
2 m
× exp[ (η + i ωim )t ] exp[ (η + i ωmi )t ].
(1) ci (t) (2) ci (t) (0)
(6.
Thus. We can write ci ˙ −i ∆i . lim where =P (6. whereas the imaginary part of ∆i governs the growth or decay of this state.288)
(6.291)
It is convenient to normalize the solution of Eq. Thus. h We can rewrite this result as i.293)
It is clear that the real part of ∆i gives rise to a simple shift in energy of state i . x (6. (6. = ci ¯ h where Hmi 2 ∆i = Hii + lim η→0 E − Em + i ¯ η h m=i i 1 →0 x + i 1 − i π δ(x). It follows that Hmi 2 ∆i = Hii + P − iπ Hmi 2 δ(Ei − Em ). Ei − E m m=i
167
(6. (6.149).289)
is a constant.290)
> 0. h h (6.17 Energyshifts and decaywidths
6 APPROXIMATION METHODS
This result is formally correct to secondorder in perturbed quantities. h h where Hmi 2 ∆Ei = Re(∆i ) = Hii + P . t = exp(−i [Ei + Re(∆i ) ] t/¯ ) exp[ Im(∆i ) t/¯ ] i . i. (6.288) so that ci (0) = 1. we obtain −i ∆i t ci (t) = exp . t = exp[−i (Ei + ∆Ei ) t/¯ ] exp(−Γi t/2¯ ) i .287) is independent of time.292) ¯ h According to Eq. and P denotes the principle part.294) (6. E − Em m=i i m=i (6. t = exp[−i (∆i + Ei ) t/¯ ] i . According to a wellknown result in pure mathematics. (6. Note that the righthand side of Eq.6.296)
.295)
(6. the time evolution of the initial state ket i is given by i.
exp[−i (Ei + ∆Ei ) t/¯ ] exp(−Γi t/2¯ ) = h h f(E) exp(−i E t/¯ ) dE.17 Energyshifts and decaywidths
6 APPROXIMATION METHODS
and
Γi 2 Im(∆i ) 2π =− = Hmi 2 δ(Ei − Em ). Clearly. ¯ h ¯ h ¯ m=i h
(6. The probability of observing the system in state i at time t > 0.294). Clearly.303)
168
. (6. (6.300)
The quantity ∆i is called the decaywidth of state i . the rate of decay of the initial state is a simple function of the transition rates to the other states.297)
Note that the energyshift ∆Ei is the same as that predicted by standard timeindependent perturbation theory. is given by Pi→i (t) = ci 2 = exp(−Γi t/¯ ).
(6. since ci 2 +
m=i
cm 2
(1 − Γi t/¯ ) + h
m=i
wi→m t = 1. state i is not a stationary state in the presence of the timedependent perturbation.299)
Here. Note that the system conserves probability up to secondorder in perturbed quantities. (6. we can still represent it as a superposition of stationary states (whose amplitudes simply oscillate in time). (6.302) According to Eq. It is closely related to the mean lifetime of this state. use has been made of Eq. the amplitude of state i both oscillates and decays as time progresses. ¯ h τi = .222).6. However.301) Γi where Pi→i = exp(−t/τi ). ¯ h m=i (6. Thus. h where Γi = wi→m .298)
(6. h (6. given that it is deﬁnately in state i at time t = 0.
Indeed. due to its propensity to decay. One consequence of h this effect is the existence of a natural width of spectral lines associated with the decay of some excited state to the ground state (or any other lower energy state). which correspond to fast transitions.6. f(E)2 is basically a deltafunction centred on the unperturbed energy Ei of state i . the energy of state i is shifted by Re(∆i ).. spectroscopists generally favour forbidden lines for Doppler shift measurements. it decays in time) implies that its energy cannot be exactly determined.e. In the presence of the perturbation. are smeared out more that weak lines.
169
. The fact that the state is no longer stationary (i.. gives rise to a slight smearing (in wavelength) of the spectral line associated with the transition.304)
In the absence of the perturbation. (E − [Ei + Re(∆i )])2 + Γi 2 /4 (6. but they are a lot sharper. For this reason. state i is a stationary state whose energy is completely determined. the larger Γi ). Strong lines.e. In other words. the energy of the state is smeared over some region of width (in energy) Γi centred around the shifted energy Ei + Re(∆i ). Such lines are not as bright as those corresponding to allowed transitions. The faster the decay of the state (i.17 Energyshifts and decaywidths
6 APPROXIMATION METHODS
where f(E) is the weight of the stationary state with energy E in the superposition. This effect is clearly a manifestation of the energytime uncertainty relation ∆E ∆t ∼ ¯ . the more its energy is spread out. The Fourier inversion theorem yields f(E)2 ∝ 1 . The uncertainty in energy of the excited state.
(7. (7. Here. p2 . Let us now examine the quantum theory of scattering.
7. (7. In general. (7.e.4)
where ψ is an energy eigenstate of the total Hamiltonian whose wavefunction r ψ is ψ(r ). and scattering experiments. Schr¨dinger’s equation for the scattering problem is o (H0 + H1 )ψ = E ψ . their energy eigenstates are unbound. data regarding quantum phenomena has been obtained from two main sources—the study of spectroscopic lines. This state is a planewave state or. possibly.2 The LipmannSchwinger equation Consider timeindependent scattering theory. We have already developed theories which account for some aspects of the spectra of hydrogenlike atoms. H0 = 2m (7.3) whose wavefunction r φ is φ(r ). corresponding to the o same energy eigenvalue.1) where H0 is the Hamiltonian of a free particle of mass m. (7. both H0 and H0 + H1 have continuous energy spectra: i.2)
and H1 represents the nontimevarying source of the scattering.1 Introduction Historically. for which the Hamiltonian of the system is written H = H 0 + H1 . a sphericalwave state.3). φ is a solution of the free particle Schr¨dinger equation. H0 φ = E φ . Let φ be an energy eigenket of H0 .4) which satisﬁes the boundary condition ψ → φ as H1 → 0..
170
.7 SCATTERING THEORY
7
Scattering theory
7. We require a solution of Eq.
.e. Thus. and is nonsingular as long as > 0. The LipmannSchwinger equation can be converted into an integral equation via left multiplication by r. the operator (E − H0 )−1 is singular: i. E − H0 ± i (7. ¯ h where ¯ 2 k2 h E= . (7. otherwise the above solution is useless. (7.4) in the form 2m (7. and making use of Eq. The physical signiﬁcance of the ± signs will become apparent later on. ψ± = φ + 1 H1 ψ± .9) 2m This equation is called Helmholtz’s equation. Furthermore. (7. and can be inverted using standard Green’s function techniques. E − H0 (7.3). it produces inﬁnities when it operates on an eigenstate of H 0 corresponding to the eigenvalue E.7. and small. we can write the scattering problem o (7. We need a prescription for dealing with these inﬁnities.7)
Adopting the Schr¨dinger representation. Thus.4) by operating on the above equation with E − H0 . 2 ¯ h
171
(7. ψ(r) = φ(r) + 2m G(r. Unfortunately.10)
.5)
Note that we can recover Eq.8) ( 2 + k2 ) ψ(r) = 2 rH1 ψ . (7. The standard prescription is to make the energy eigenvalue E slightly complex.6)
where is real. positive. Thus.2 The LipmannSchwinger equation
7 SCATTERING THEORY
Formally. ψ± (r) = φ(r) + r 1 E − H0 ± i r r H1 ψ± d3 r .6) is called the LipmannSchwinger equation. the solution satisﬁes the boundary condition ψ → φ as H1 → 0. r ) r H1 ψ d3 r . the desired solution can be written ψ = φ + 1 H1 ψ . Equation (7.
(7.11)
Note that the solution (7. We can write r H1 ψ± = r H1 r r ψ± d3 r (7. (7. r ) = − Thus. 4π r − r 
(7.7) and (7. as is easily proved by a more rigorous derivation of this result.17)
Suppose that the initial state φ is a planewave with wavevector k (i. Eq. (7. r ) = δ(r − r ). a stream of particles of deﬁnite momentum p = ¯ k). ψ± (r) = φ(r) − Let us suppose that the scattering Hamiltonian. 4π r − r  (7.2 The LipmannSchwinger equation
7 SCATTERING THEORY
where (
2
+ k2 ) G(r. In fact.7) takes the form 2 m exp(±i k r − r  ) 1 r =− 2 . H1 . the Green’s function for the Helmholtz problem is given by G(r.14) r E − H0 ± i 4π r − r  ¯ h It is not entirely clear that the ± signs correspond on both sides of this equation. the integral equation (7. The ket corresponding to h this state is denoted k .18)
. (7. is only a function of the position operators. (7. This implies that r H1 r = V(r) δ(r − r ).e.13) 2 4π r − r  ¯ h A comparison of Eqs.12)
= V(r ) ψ± (r ). they do.16) (7. The associated wavefunction takes the form rk = exp( i k·r) . (2π)3/2
172
(7.13) suggests that the kernel to Eq.10) satisﬁes the boundary condition ψ → φ as H1 → 0..13) simpliﬁes to ψ± (r) = φ(r) − 2m ¯2 h exp(±i k r − r ) V(r ) ψ± (r ) d3 r .
(7. As is wellknown. Thus.7.15) exp(±i k r − r  ) .10) becomes 2 m exp(±i k r − r  ) r H1 ψ d3 r .
Eq.7.2 The LipmannSchwinger equation
7 SCATTERING THEORY
The wavefunction is normalized such that kk = = kr rk d3 r exp[−i r·(k − k )] 3 d r = δ(k − k ). whereas the minus sign corresponds to a wave propagating towards the scattering region..21) r is a unit vector which points from the scattering region to the observation point. The plus sign (on ψ± ) corresponds to a wave propagating away from the scattering region. where r − ^·r r (7.20)
r (7.e. (7. Let us calculate the wavefunction ψ(r) a long way from the scattering region. k is the wavevector for particles which possess the same energy as the incoming particles (i. It is easily demonstrated that r − r  to ﬁrstorder in r /r. Note that exp(±i k r − r  ) exp(±i k r) exp( i k ·r ). but propagate from the scattering region to the observation point.
173
.24) − 2 3/2 (2π) r 2π ¯ h
The ﬁrst term on the righthand side is the incident wave. let us adopt the ordering r r .19)
Suppose that the scattering potential V(r) is only nonzero in some relatively localized region centred on the origin (r = 0). Let us deﬁne k = k ^. k = k). r (7. (7. (7.17) reduces to ψ(r)± exp( i k·r) m exp(±i k r) exp( i k ·r ) V(r ) ψ± (r ) d3 r .22) ^= r
Clearly.23)
In the larger limit. In other words. The second term represents a spherical wave centred on the scattering region. (2π)3 (7.
e.31)
dσ r2 dΩ jscat  dΩ = .25)
where
(2π)2 m = − k H1 ψ . that the probability ﬂux (i. the wavefunction a long way from the scattering region can be written 1 exp( i kr) f(k .2 The LipmannSchwinger equation
7 SCATTERING THEORY
It is obvious that the former represents the physical solution. k)2 ¯ h k ^. Thus. 4.. divided by the incident ﬂux of particles. the probability ﬂux associated with the incident wavefunction. k) .7.27)
(7. exp( i k r) f(k . (2π)3/2 r (7. m Thus. k) . dΩ jinci 
174
(7. the particle ﬂux) associated with a wavefunction ψ is ¯ h Im(ψ∗ ψ). r = (2π)3 m r2
(7.
f(k . from Sect. ¯2 h
Let us deﬁne the differential crosssection dσ/dΩ as the number of particles per unit time scattered into an element of solid angle dΩ. ψ(r) = exp( i k·r) + (2π)3/2 r (2π)2 m f(k . the probability ﬂux associated with the scattered wavefunction. (2π)3/2 is jinci = (7. Recall. j= exp( i k·r) .29) (2π)3 m Likewise.32)
.30)
is jscat Now. (7. k) = − ¯2 h exp(−i k ·r ) V(r ) ψ(r ) d3 r 3/2 (2π) (7.28)
¯ h k.26)
(7.
as well as the local value of the wavefunction.3 The Born approximation
7 SCATTERING THEORY
giving dσ = f(k . f(k . ψ(r). (7. it is reasonable to suppose that the total wavefunction.e. (7.7. as it stands.15).34) 2 r 2π ¯ h where φ(r) is the wavefunction of the incident state. ψ. k)2 gives the differential crosssection for particles with incident momentum ¯ k to be scattered into states whose momentum vectors are directed in h a range of solid angles dΩ about ¯ k . ψ(r) = φ(r) − Suppose that the scattering is not particularly strong. V. According to the above equation the total wavefunction is a superposition of the incident wavefunction and lots of sphericalwaves emitted from the scattering region. k = k). k) depends on the unknown ket ψ . (7. k)2 ..33) dΩ Thus. The strength of the sphericalwave emitted at a given point is proportional to the local value of the scattering potential.
7.36)
. φ(r).3 The Born approximation Equation (7.35)
This is called the Born approximation. we can obtain an expression for f(k .33) is not particularly useful. Note that the scattered particles possess h the same energy as the incoming particles (i. because the quantity f(k . (2π)3/2 (7. k) by making the substitution ψ(r) → φ(r) = exp( i k·r) . Thus. In this case. The Born approximation yields f(k . This is always the case for scattering Hamiltonians of the form shown in Eq. k) − m 2π ¯ 2 h
exp [ i (k − k )·r ] V(r ) d3 r . Recall that ψ(r) = rψ is the solution of the integral equation m exp( i k r) exp(−i k ·r ) V(r ) ψ(r ) d3 r . does not differ substantially from the incident wavefunction.
175
(7.
It follows from Eq. the differential crosssection for scattering by a Yukawa potential is dσ dΩ given that
2 m V0 1 .38) ¯ q 0 h Note that f(k . k) − 2 r V(r ) sin(q r ) dr . k) giving 2m ∞ f(k . in the Born approximation.43)
q2 = 4 k2 sin2 (θ/2) = 2 k2 (1 − cos θ). (7.41) ¯ µ q2 + µ 2 h since ∞ q exp(−µ r ) sin(q r ) dr = 2 .42) µ + q2 0 Thus. f(k . (7.44)
. (7.3 The Born approximation
7 SCATTERING THEORY
Thus.
176
(7.40)
where V0 is a constant and 1/µ measures the “range” of the potential. Recall that the vectors k and k have the same length by energy conservation.37)
where θ is the angle subtended between the vectors k and k .39) − m 2π ¯ 2 h exp( i q r cos θ ) V(r ) r 2 dr sin θ dθ dφ . k) is just a function of q for a spherically symmetric potential.38) that 2 m V0 1 f(θ) = − 2 .7. It is easily demonstrated that q ≡ k − k  = 2 k sin(θ/2). θ is the angle of scattering. Consider scattering by a Yukawa potential V(r) = V0 exp(−µ r) . k) is proportional to the Fourier transform of the scattering potential V(r) with respect to the wavevector q ≡ k − k . [2 k2 (1 − cos θ) + µ2 ]2 ¯2µ h
2
(7. (7. µr (7. For a spherically symmetric potential. f(k . In other words. (7. (7.
if the potential is strong enough to form a bound state then the Born approximation is likely to break down. from Eq.47) µ)
Consider the special case of the Yukawa potential. provided that V0 /µ → Z Z e2 /4π 0 . In this limit the Born differential crosssection becomes 2 2 m Z Z e2 dσ 1 . Equation (7. Eq. (7. 4 16π 0 E sin (θ/2)
2
(7.. Thus.45) dΩ 4π 0 ¯ 2 h 16 k4 sin4 (θ/2) Recall that ¯ k is equivalent to p. that the condition for ψ(r) φ(r) in the vicinity of r = 0 is m 2π ¯ 2 h exp( i k r ) V(r ) d3 r r 1. It follows.47) yields 2 m V0  1.17). implying that the Born approximation is more accurate at high incident particle energies. so the above equation can be rewritten h dσ dΩ 1 Z Z e2 .49)
where V0 is negative.7.48)
as the condition for the validity of the Born approximation. giving 2 m V0  ¯ 2 µ2 h 1
(7. The Born approximation is valid provided that ψ(r) is not too different from φ(r) in the scattering region.e. ¯ 2 µ2 h (7. (i.46)
where E = p2 /2 m is the kinetic energy of the incident particles.46) is the classical Rutherford scattering crosssection formula. (7. At low energies. (7. In the highk limit. k we can replace exp( i k r ) by unity.7. (7. The condition for the Yukawa potential to develop a bound state is 2 m V0  ≥ 2. (7.
177
.50) ¯2 µk h This inequality becomes progressively easier to satisfy as k increases.3 The Born approximation
7 SCATTERING THEORY
The Yukawa potential reduces to the familiar Coulomb potential as µ → 0.
.53)
depend on the azimuthal angle ϕ. ϕ) = f(θ). The scattered wavefunction is characterized by a wavevector k which has the same magnitude as k. 5 via Pl (cos θ) = 4π Yl0 (θ. It follows that neither the incident wavefunction. that the incident wavefunction is characterized by a wavevector k which is aligned parallel to the zaxis. (2π)3/2 (2π)3/2
(7. and an azimuthal angle ϕ about the zaxis.e. 1 exp( i k r) f(θ) ψ(r) = exp( i k r cos θ) + .e.38) strongly suggests that for a spherically symmetric scattering potential [i. both φ(r) and ψ(r) satisfy the free space Schr¨dinger equation o (
2
+ k2 ) ψ = 0. V(r) = V(r)] the scattering amplitude is a function of θ only: f(θ. 2l + 1
178
(7.54)
What is the most general solution to this equation in spherical polar coordinates which does not depend on the azimuthal angle ϕ? Separation of variables yields ψ(r. φ(r) = nor the total wavefunction. The direction of k is speciﬁed by the polar angle θ (i. The Legendre functions are related to the spherical harmonics introduced in Sect. but. ϕ). (2π)3/2 r
(7. Equation (7.
(7.4 Partial waves
7 SCATTERING THEORY
7. θ) =
l
Rl (r) Pl (cos θ). without loss of generality.4 Partial waves We can assume.55)
since the Legendre functions Pl (cos θ) form a complete set in θspace.7.
Outside the range of the scattering potential. points in a different direction. in general.51)
exp( i k z) exp( i k r cos θ) = . the angle subtended between the two wavevectors)..
(7.56)
.52)
(7.
1 δnm . (7. whereas Neumann functions become singular.
(7.58) (7.59)
ηl (y) = −y
1 d − y dy
cos y .60) (7. It is easily demonstrated that r2 jl (y) = y
l
1 d − y dy
l
l
sin y . because they are not wellbehaved as r → 0. and a Neumann function.65)
179
. Note there are no Neumann functions in this expansion. jl (k r). The Legendre functions are orthonormal. ηl (k r).55) can be combined to give dRl d2 Rl + [k2 r2 − l (l + 1)]Rl = 0. The asymptotic behaviour of these functions in the limit y → ∞ is sin(y − l π/2) .
(7.54) and (7.4 Partial waves
7 SCATTERING THEORY
Equations (7. y jl (y) →
(7.64)
It is wellknown that (−i)l jl (y) = 2
1
exp( i y µ) Pl (µ) dµ.62)
where the al are constants. y cos(y − l π/2) ηl (y) → − .57) + 2r 2 dr dr The two independent solutions to this equation are called a spherical Bessel function. (7. y
Note that spherical Bessel functions are wellbehaved in the limit y → 0 .7.61)
We can write
exp( i k r cos θ) =
l
al jl (k r) Pl (cos θ).63) Pn (µ) Pm (µ) dµ = n + 1/2 −1 so we can invert the above expansion to give
1
al jl (k r) = (l + 1/2)
−1
exp( i k r µ) Pl (µ) dµ.
−1
(7. y
l
(7.
(7.60)–(7. they must be part of the larger asymptotic
180
.68) ψ(r) = (2π)3/2 l where the Al and Bl are constants. What is the source of the incoming waves? Obviously. New York NY. the total wavefunction reduces to ψ(r) 1 (2π)3/2
l
cos(k r − l π/2) sin(k r − l π/2) A − Bl Pl (cos θ).69)
where use has been made of Eqs.14]. Thus.7. The most general solution for the total wavefunction outside the scattering region is 1 [Al jl (k r) + Bl ηl (k r)] Pl (cos θ). 1. 2. Note that the Neumann functions are allowed to appear in this expansion.71)
× Pl (cos θ). because its region of validity does not include the origin.67)
The above expression tells us how to decompose a planewave into a series of sphericalwaves (or “partial waves”). Eq. In the larger limit. · · · [see Abramowitz and Stegun (Dover. l kr kr
(7. (7.70)
where the sine and cosine functions have been combined to give a sine function which is phaseshifted by δl . kr
(7. 10.4 Partial waves
7 SCATTERING THEORY
where l = 0.66) giving exp( i k r cos θ) =
l
il (2 l + 1) jl (k r) Pl (cos θ).61).
which contains both incoming and outgoing sphericalwaves. The above expression can also be written ψ(r) 1 (2π)3/2 Cl
l
sin(k r − l π/2 + δl ) Pl (cos θ). 1965). (7.1. al = il (2 l + 1). Equation (7.
(7.70) yields ψ(r) 1 (2π)3/2 Cl
l
exp[ i (k r − l π/2 + δl )] − exp[−i (k r − l π/2 + δl )] 2ikr (7.
.53) give (2π)3/2 [ψ(r) − φ(r)] = exp( i k r) f(θ). It follows from Eqs.73)
Note that the righthand side consists only of an outgoing spherical wave.e. (7.76)
× sin δl sin δl Pl (µ) Pl (µ). In fact.
181
.71)–(7. (7. it is easily seen that φ(r) 1 (2π)3/2 il (2l + 1)
l
exp[ i (k r − l π/2)] − exp[−i (k r − l π/2)] 2ikr (7. This implies that the coefﬁcients of the incoming spherical waves in the larger expansions of ψ(r) and φ(r) must be equal.5 The optical theorem
7 SCATTERING THEORY
expansion of the incident wavefunction. The total crosssection is given by σtotal = f(θ)2 dΩ
1
1 = 2 k
dϕ
−1
dµ
l l
(2 l + 1) (2 l + 1) exp[ i (δl − δl ] (7.52) and (7.72)
× Pl (cos θ) in the larger limit. r
(7.73) yield f(θ) =
∞ l=0
(2 l + 1)
exp( i δl ) sin δl Pl (cos θ).7. Eqs.
7.74) Thus. determining the scattering amplitude f(θ) via a decomposition into partial waves (i.71) and (7. (7. Now. sphericalwaves) is equivalent to determining the phaseshifts δ l . (7.72) that Cl = (2 l + 1) exp[ i (δl + l π/2)].5 The optical theorem The differential scattering crosssection dσ/dΩ is simply the modulus squared of the scattering amplitude f(θ). k
(7.75)
Clearly. Eqs.
e. σl =
7. the contribution to the total crosssection from the lth partial wave.77)
where use has been made of Eq. Consider a spherically symmetric potential V(r) which vanishes for r > a. where a is termed the range of the potential. It is usual to write σtotal = where
∞ l=0
σl . In the region r > a. (7.6 Determination of phaseshifts
7 SCATTERING THEORY
where µ = cos θ.54). and thereby reduce the probability current in this direction. Note that the maximum value for the lth partial crosssection occurs when the phaseshift δl takes the value π/2.
(7.. (7.63).7. It follows that σtotal = 4π k2 (2 l + 1) sin2 δl . This result is known as the optical theorem.80) 2 k is the lth partial crosssection: i.6 Determination of phaseshifts Let us now consider how the phaseshifts δl can be evaluated.79)
4π (2 l + 1) sin2 δl (7. A comparison of this result with Eq. (7.81)
182
. It is a reﬂection of the fact that the very existence of scattering requires scattering in the forward (θ = 0) direction in order to interfere with the incident wave.75) yields 4π σtotal = Im [f(0)] . the wavefunction ψ(r) satisﬁes the freespace Schr¨dinger equation (7. The most general solution which is o consistent with no incoming sphericalwaves is 1 ψ(r) = (2π)3/2
∞ l=0
il (2 l + 1) Al (r) Pl (cos θ).
l
(7.
(7.78) k since Pl (1) = 1.
88)
ensures that the radial wavefunction is wellbehaved at the origin. (7.7. The most general solution to Schr¨dinger’s equation inside the range of the o potential (r < a) which does not depend on the azimuthal angle ϕ is 1 ψ(r) = (2π)3/2 where Rl (r) = and
∞ l=0
il (2 l + 1) Rl (r) Pl (cos θ). k a ηl (k a) − βl+ ηl (k a)
(7.86)
d2 ul 2 2m l (l + 1) + k − 2 V− ul = 0.85)
(7. integrate out to r = a. etc. ul (r) . The above equation can be inverted to give (7. and form the logarithmic derivative βl− = 1 d(ul /r) (ul /r) dr
183
. r
(7.87) (7. = ka cos δl jl (k a) − sin δl ηl (k a) tan δl = k a jl (k a) − βl+ jl (k a) .89)
.83)
where jl (x) denotes djl (x)/dx.
r=a
(7.6 Determination of phaseshifts
7 SCATTERING THEORY
where Al (r) = exp( i δl ) [ cos δl jl (k r) − sin δl ηl (k r) ] .84)
Thus. dr2 r2 ¯ h The boundary condition ul (0) = 0
(7. the problem of determining the phaseshift δl is equivalent to that of obtaining βl+ . We can launch a wellbehaved solution of the above equation from r = 0. because its region of validity does not include the origin (where V = 0). The logarithmic derivative of the lth radial wavefunction Al (r) just outside the range of the potential is given by βl+ cos δl jl (k a) − sin δl ηl (k a) .82) Note that Neumann functions are allowed to appear in the above expression.
for which the potential is inﬁnite for r < a. It follows from Eq. Thus. It follows that (7. (7.91) for all l.94)
(7. The phaseshift δl is obtainable from Eq.7 Hard sphere scattering
7 SCATTERING THEORY
Since ψ(r) and its ﬁrst derivatives are necessarily continuous for physically acceptible wavefunctions. A0 (r) = kr
184
(7. 7.95) kr The corresponding radial wavefunction for the incident wave takes the form A0 (r) = exp(−i k a) sin k r ˜ . and zero for r > a. − cos(k a)/ka δ0 = −k a.93)
where use has been made of Eqs. Equation (7. which implies that u l = 0 for all l.58)–(7. It follows that ψ(r) is zero in the region r < a. (7. it follows that βl+ = βl− .7.90)
Consider the l = 0 partial wave.84) that tan δl = jl (k a) . ηl (k a) (7.7 Hard sphere scattering Let us test out this scheme using a particularly simple example. βl− = βl+ = ∞. (7. which is usually referred to as the swave. Consider scattering by a hard sphere. The swave radial wave function is [cos k a sin k r − sin k a cos k r] kr sin[k (r − a)] = exp(−i k a) . (7.84).96)
.92) yields tan δ0 = sin(k a)/k a = − tan k a.92) (7.59). (7.
Low energy means k a 1. the crosssection for classical particles bouncing off a hard sphere of radius a).. with respect to δ0 . Note that the total crosssection σtotal = dσ dΩ = 4π a2 dΩ (7. the spherical Bessel functions and Neumann functions reduce to: jl (k r) ηl (k r) (k r)l . Consider the high energy limit k a 1.98)
where n!! = n (n − 2) (n − 4) · · · 1. with l > 0.7 Hard sphere scattering
7 SCATTERING THEORY
It is clear that the actual l = 0 radial wavefunction is similar to the incident l = 0 wavefunction. It follows from Eq. spherically symmetric scattering) is important.75).e. low energy scattering implies relatively long wavelengths.100)
is four times the geometric crosssection π a2 (i. except that it is phaseshifted by k a.33).99)
It is clear that we can neglect δl .7. (2 l + 1) [(2 l − 1)!!]2 (7.102) k l=0
185
. At high energies.101) a2 (7. Let us consider the low and high energy asymptotic limits of tan δl .. (7. (7. (7. (7.94) that dσ sin2 k a = dΩ k2 for k a 1. It follows from Eqs. (k r)l+1 (7.e. all partial waves up to lmax = k a contribute signiﬁcantly to the scattering crosssection. at low energy only swave scattering (i.97) (7. However. so we do not expect to obtain the classical result in this limit. It follows that −(k a)2 l+1 tan δl = . In this regime. (2 l + 1)!! − (2 l − 1)!! .77) that l 4π max σtotal = 2 (2 l + 1) sin2 δl . In other words. and (7.
The potential is repulsive for V0 > 0. Thus.103) σtotal = 2 k l=0 This is twice the classical result.8 Low energy scattering
7 SCATTERING THEORY
With so many l values contributing.59).58)–(7. characterized by V = V0 for r < a.7. and V = 0 for r ≥ a. (7. when 1/k is much larger than the range of the potential) partial waves with l > 0. For hard sphere scattering. However. only swave scattering is important. (7. ka 2π (2 l + 1) 2π a2 . with a ﬁnite range potential. The outside wavefunction is given by [see Eq.. it is legitimate to replace sin 2 δl by its average value 1/2. and attractive for V0 < 0. there must be scattering in the forward direction (recall the optical theorem) to produce destructive interference with the incident planewave.104) (7. which is somewhat surprizing.8 Low energy scattering At low energies (i. It follows that. the interference is not completely destructive. since we might expect to obtain the classical result in the short wavelength limit.
7. The effective crosssection associated with this bright spot is π a 2 which.82)] A0 (r) = exp( i δ0 ) [j0 (k r) cos δ0 − η0 (k r) sin δ0 ] = exp( i δ0 ) sin(k r + δ0 ) . (7. The inside wavefunction follows
186
. in order to produce a “shadow” behind the sphere. make a negligible contribution to the scattering crosssection. Here.e. when combined with the crosssection for classical reﬂection. at these energies. let us consider scattering by a ﬁnite potential well. incident waves with impact parameters less than a must be deﬂected. and the shadow has a bright spot in the forward direction. V0 is a constant. gives the actual crosssection of 2π a2 . in general. kr (7. π a2 . As a speciﬁc example.105)
where use has been made of Eqs. In fact.
and ¯2k 2 h . (7.8 Low energy scattering
7 SCATTERING THEORY
from Eq. r (7. For E < V0 . we have A0 (r) = B where sinh κ r . the righthand side is much less that unity.88). r (7.111)
Consider an attractive potential. V0 − E = 2m Matching A0 (r).87). and its radial derivative at r = a.106)
where use has been made of the boundary condition (7. for which E > V0 . and tan(k a + δ0 ) = for E < V0 . so that k becomes extremely large.108)
¯ 2 κ2 h . Here. We obtain A0 (r) = B sin k r .112)
(7.106) only applies when E > V0 .7. the depth of the potential well is much larger than the energy of the incik. yields tan(k a + δ0 ) = k tan k a k k tanh κ a κ
(7. (7.110) that.110)
for E > V0 . so replacing the tangent of a small quantity with the quantity itself. Suppose that V0  E (i. It follows from Eq. (7.113)
.
(7..107) E − V0 = 2m Note that Eq. B is a constant.e. ka
187
k tan k a. unless tan k a dent particles).109)
(7. (7. k
(7. we obtain k a + δ0 This yields δ0 tan k a ka − 1 .
and the scattering crosssection (7.
7. despite the very strong attraction of the potential.7. In this case. (7.115)
so for sufﬁciently small values of k a. It follows that there are certain values of V0 and k which give rise to almost perfect transmission of the incident wave. can reach the value π/2.117)
. so we can no longer assume that the righthand side of Eq. it follows from Eq.114) vanishes.110) that k a + δ0 = π/2. (7. which is given by Eq.. This implies that σtotal = 4π 2 1 sin δ0 = 4π a2 2 2 . at low incident energies.9 Resonances There is a signiﬁcant exception to the independence of the crosssection on energy. (7. Note that there are values of k a (e. Suppose that the quantity 2 m V0  a2 /¯ 2 is slightly less than π/2.114)
(7. (7.49) at which δ0 → π.102).g. As the h incident energy increases.115). at the value of the incident energy π/2 when k a = π/2. ka 2 m V0  a2 . because of contributions from l > 0 partial waves. ¯2 h (7.9 Resonances
7 SCATTERING THEORY
According to Eq. or δ0 (since we are assuming that k a 1). the scattering crosssection is given by σtotal Now tan k a 4π 2 sin δ0 = 4π a2 − 1 . 2 k ka ka= k 2 a2 2 m V0  a2 . and has been observed experimentally. + ¯2 h
2
(7. the crosssection is not exactly zero. This is called the RamsauerTownsend effect.116)
It follows that the total (swave) scattering crosssection is independent of the energy of the incident particles (provided that this energy is sufﬁciently small). k a 4. tan k a becomes inﬁnite. these contributions are small. But. In fact.110) is small. In reality. k a. k2 k a
188
(7.
7.122)
. (7.121)
1 dδl = − 2 (E − E0 ) + · · · .9 Resonances
7 SCATTERING THEORY
Note that the crosssection now depends on the energy. However. The condition π 2 m V0  a2 = 2 2 ¯ h (7. We have seen that there is a resonant effect when the phaseshift of the swave takes the value π/2. Furthermore. The origin of this rather strange behaviour is quite simple. In this situation. (7. this sort of resonance scattering is best understood as the capture of an incident particle to form a metastable bound state. the bound state is not stable.114) for k a = π/2 (since k a 1). Suppose that δl attains the value π/2 at the incident energy E0 . the energy of the scattering system is essentially the same as the energy of the bound state.120) (7. Nevertheless. an incident particle would like to form a bound state in the potential well. sin δl dE E=E0 Deﬁning
2 dδl (E) = . There is nothing special about the l = 0 partial wave. dE E=E0 Γ
189
(7. and the subsequent decay of the bound state and release of the particle. The crosssection for resonance scattering is generally far higher than that for nonresonance scattering. so that π δl (E0 ) = . so it is reasonable to assume that there is a similar resonance when the phaseshift of the lth partial wave is π/2. the magnitude of the crosssection is much larger than that given in Eq. for a potential well which satisﬁes the above equation. since the system has a small positive energy. Thus.118)
is equivalent to the condition that a spherical well of depth V0 possesses a bound state at zero energy.119) 2 Let us expand cot δl in the vicinity of the resonant energy: cot δl (E) = cot δl (E0 ) +
d cot δl dE
E=E0
(E − E0 ) + · · ·
(7.
125)
This is the famous BreitWigner formula. The variation of the partial crosssection σl with the incident energy has the form of a classical resonance curve. that the contribution of the lth partial wave to the scattering crosssection is σl = 4π 1 4π . from Eq.17).80).7. of energy E0 . (2 l + 1) k2 (E − E0 )2 + Γ 2 /4 (7. Γ Recall. (7.9 Resonances
7 SCATTERING THEORY
we obtain
2 (7. The quantity Γ is the width of the resonance (in energy). 6. (2 l + 1) sin2 δl = 2 (2 l + 1) k2 k 1 + cot2 δl 4π Γ 2 /4 .123) cot δl (E) = − (E − E0 ) + · · · . σl (7.124)
Thus. h
190
. We can interpret the BreitWigner formula as describing the absorption of an incident particle to form a metastable state. and lifetime τ = ¯ /Γ (see Sect.