Mph08 Unit 1
Mph08 Unit 1
UNIT 1
SYMMETRY IN QUANTUM
MECHANICS-I
Structure
1.1 INTRODUCTION
In your undergraduate courses in mechanics you have studied the laws of
conservation of energy, linear momentum and angular momentum and derived
them from the laws of motion that govern the dynamics of the system. In your
first semester course in classical mechanics (MPH-002), however, you have
seen that these conservation laws follow from something far more basic, which
are the symmetries that govern the physical system. This semester, in
Classical mechanics-II (MPH-006) you will study Noether’s theorem which
puts on a firm footing the idea that every continuous symmetry (like
symmetry under the translation or rotation) of a physical system leads to a
conservation law.
One example of a symmetry that you may also be familiar with, from your
undergraduate courses is symmetry in molecules and crystals. For example, if
we take a molecule of benzene and rotate it by 120° about an axis passing
through the centre of the molecule it will look the same as before. It will also
look the same if we reflect it through a mirror plane passing through opposite
single and double bonds. Crystal symmetry, however, is an example of a
discrete symmetry.
A symmetry operation on any system can be defined as an action (some
change) on a physical system that leaves the system invariant. For example if
the force acting on a physical system does not depend upon the location of
the system, then the result of any experiment performed on the system would 9
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
be the same irrespective of which region of space the experiment is carried
out. So one could translate the experiment in space and get the same result.
This is an example of translational symmetry, a consequence of the
homogeneity of space and leads us to the conservation of linear momentum.
Similarly in the absence of a torque, the results of an experiment do not
change if we describe it in a coordinate system that is rotated with respect to
the initial frame. This is an example of rotational symmetry, a consequence of
the uniformity of space and it leads us to the conservation of angular
momentum. We also talk about the homogeneity of time or time-translation
which tells us that an experiment performed at two different times on identical
systems must give us the same result. This leads us to the law of
conservation of energy. Space and time translations and rotations which leave
the dynamics of an isolated physical system unchanged are examples of
symmetry operations. These are also examples of continuous symmetries.
In general, a symmetry transformation of a physical system is one that does
not change the result of any experiment performed on it. Mathematically
symmetry operations (like the rotation of a vector in three dimensional space)
are represented by groups and you have studied about these in Unit 19 of
MPH-001. In this unit we study how a symmetry operation is defined in
quantum mechanics and understand how this leads us to the theorems on the
conservation of various quantities in quantum mechanical systems.
Quantum mechanical systems are described by states and all transformations
of the quantum mechanical system are described by operators acting on
states. You must remember that in quantum mechanics (Unit 11, MPH-004),
the average measured value of an observable for a physical system, (say A) is
the expectation value of the operator ( Â ) in the state ( ) of the system
Eugene Paul "E.P." n
Wigner (1902-995),
Hungarian American
which is Aˆ ai i 2 where i are the eigenkets of  . So the
i 1
physicist and result of any measurement depends on the transition amplitudes between
mathematician, received
quantum states and therefore a symmetry transformation on a quantum
a share of the Nobel
Prize in Physics in 1963
system is one that leaves these transition amplitudes between states
"for his contributions to unchanged. This is the underlying principle of the mathematical formulation of
the theory of the atomic symmetries like translation, rotation, parity, time reversal etc.
nucleus and the
There is also a famous theorem by Eugene Wigner (read the margin remark)
elementary particles,
particularly through the called Wigner’s theorem which states that a symmetry transformation in the
discovery and Hilbert space of states could be represented by either a unitary or an anti-
application of unitary operator . We will be using both of these concepts in defining
fundamental symmetry symmetry transformations in quantum systems in this Unit and the next .
principles". His work laid
the foundation for the We first derive the operators that represent the symmetry operations on a
theory of symmetries in quantum mechanical system like space and time translation, rotation,
quantum mechanics. reflection etc and then we study the action of these operators on states and on
operators.
In Sec. 1.2 we study about space translation. We define the space translation
operator and show how linear momentum is the generator of space translation.
We study about rotation in Sec.1.3 and derive the expression for the rotation
operator. We also use the properties of rotation groups that you have studied in
10
Unit 1 Symmetry in Quantum Mechanics-I
MPH-003 and show how the angular momentum operator is the generator of
rotations in quantum systems. In Sec. 1.4 we study time translation and derive
the expression for the time-translation operator. In Sec. 1.5 we study how the
general ideas of symmetry lead us to the conservation laws and degeneracies in
the eigenstates of the Hamiltonian.
the space translation operator for infinitesimal and finite space translation
and be able to establish its properties;
the matrix representation of the rotation operator for infinitesimal and finite
rotation about an axis and establish its properties;
(x )
(x )
x x x x x
(a) (b)
The system is now localized around the point x x (the maximum of the
probability distribution is now at x x ). So the wave function is displaced to
the right by a distance x . Let us say that the system is now located at the
point x where:
x x x (1.2)
T ( x ) ( x x ) ( x ) x terms in higher powers of x
x
(1.4b)
T ( x ) 1 x ( x ) 1 i x Kˆ x ( x ) (1.5a)
x
where Kˆ x i (1.5b)
x
12
Unit 1 Symmetry in Quantum Mechanics-I
So the translation of the wavefunction is brought about by the action of an
operator K̂ x acting on (x ) . The operator K̂ x is called the generator of
infinitesimal translation along the x-axis.
pˆ
So: Kˆ x x (1.5c)
pˆ
T ( x ) 1 i x x ( x ) (1.6)
We can carry out the same analysis for the transformation of the three
dimensional wave function (r ) under an infinitesimal space translation
defined by r r , where is a constant vector, to get:
pˆ
T ( r ) 1 i (r ) (1.7)
where p̂ i .
The two different functional forms of the wave function, (x ) and T (x ) of the
same state can also be taken to be the representation of the state in
two different bases in Hilbert space , namely x and x . So we can write:
( x ) x and T ( x ) x (1.8)
There could be an arbitrary phase factor in the RHS of Eq. (1.9), which we
take to be unity. A particle in the state x is located at x. When the operator
Tˆ ( ) acts on the state x , it is transformed to the state x . A particle in
x
the state x x is located at x x .
Tˆ ( x ) Tˆ ( x ) dx x x dx Tˆ( x ) x x dx x x x
(1.10)
Defining x x x , Eq. (1.10) reduces to
Tˆ ( x ) dx x x x (1.11)
Tˆ ( x ) dx x x x (1.12)
This tells us the translated wave function is just the original wavefunction
translated to the right by an amount x .
The next step is to deduce a form for the translation operator. We ask that the
translation operator defined in Eq. (1.9) have the following properties:
Tˆ ( x ) Tˆ ( x )1 (1.14a)
Tˆ ( x ) x x x ; Tˆ ( ) x x x x (1.14b)
Therefore:
Tˆ ( x )Tˆ ( x ) x x Tˆ ( x )Tˆ ( x ) ˆI Tˆ ( x ) Tˆ ( x ) 1
Tˆ ( x )† Tˆ ( x ) 1 Tˆ ( x )† Tˆ ( x ) Î
Tˆ ( x )† Tˆ ( x ) (1.16)
As 0 , Tˆ ( x ) Î (1.17)
14
Unit 1 Symmetry in Quantum Mechanics-I
SAQ 1
a) Show that Tˆ ( x )Tˆ ( ) Tˆ ( )Tˆ ( x ) .
Notice that we have not yet imposed any particular form for the translation
operator. We now show that if we choose the translation operator for an
infinitesimal translation x to be of the form:
Tˆ ( x ) ˆI i x Kˆ x (1.18)
where K̂ x is a Hermitian operator Kˆ x † Kˆ x , then it satisfies all the properties
listed above. Let us check a few of these.
Unitarity
Hence Tˆ † ( x )Tˆ ( x ) ˆI i x Kˆ x ˆI i x Kˆ x ˆI x 2Kˆ x 2 ˆI
Tˆ ( x )Tˆ ( x ) Tˆ ( x x )
Tˆ ( x )Tˆ ( x ) ˆI i x Kˆ x ˆI i x Kˆ x ˆI i x x Kˆ x
Tˆ ( x x ) (1.19)
SAQ 2
i) Tˆ ( x )† Tˆ ( x ) ; ii) Tˆ ( x ) Tˆ ( x ) 1 .
Using the form of the translation operator in Eq. (1.18), it is possible to prove
an important property of the Hermitian operator K̂ x .
Similarly:
Tˆ ( x )xˆ x xˆ x x x Tˆ ( x ) x x (1.21)
Therefore 15
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
xˆTˆ ( x ) Tˆ ( x )xˆ x x x x x x
x x x (1.22)
x x x ˆI i x Kˆ x x x x i x 2 Kˆ x x x x (1.24)
In arriving at the result of Eq. (1.24) we have once again neglected terms of
the order of x 2 . Using the result of Eq. (1.24) in Eq. (1.22), we can write:
xˆTˆ( x ) xˆTˆ ( x ) x
x x xˆ ,Tˆ ( x ) x (1.25)
Remember that the only conditions that we have imposed on the translation
operator is that x is infinitesimal and K̂ is hermitian. We can extend the
analysis to a three dimensional infinitesimal translation to write the
translation operator as:
Tˆ ( ) ˆI i Kˆ (1.28)
Then, on substituting for K̂ x from Eq.(1.31) into Eq. (1.27) we get back the
commutation relations: xˆ , pˆ x i . So the definition is consistent with all the
properties of the translation operation and the basic commutation rules. So the
translation operator for an infinitesimal translation of x along the x-direction is
pˆ
Tˆ ( x ) ˆI i x x (1.32)
For a 3D translation, , the translation operator is
pˆ
Tˆ ( ) ˆI i (1.33)
Substituting from Eq. (1.31) into Eq. (1.29) we get back the familiar
commutation relations: xˆ , pˆ x i ; yˆ , pˆ y i ; zˆ , pˆ z i
We may write:
pˆ
x x exp i x x x (1.35a)
For a finite translation ( x , y , z ), Eq. (1.32) reduces to:
pˆ x pˆ x y pˆ y z pˆ z
Tˆ ( ) exp i exp i (1.35b)
SAQ 3
Use Eq. (1.32) to show pˆ x , pˆ y 0 .
1.3 ROTATION
As we have done for space translation, let us first look at how the wave
function of a particle changes with a rotation of the system. Consider a
quantum system represented by the wave function ( r ) ( x, y , z ) and after
rotation the wave function is R ( r ) R ( x , y , z ) .
Let us consider that the rotation of the system is by an angle about the z-
axis (Fig. 1.2). Let us say that the system is localized at ( x, y , z ) before the
rotation and at ( x , y , z ) after the rotation.
x
(r )
y
We can now assume that the value of the initial wave function of the system at
the point ( x, y , z ) , i.e. ( x, y , z ) must be equal to value of the final wave
function at ( x , y , z ) . i.e. R ( x , y , z ) . So :
R ( x , y , z ) ( x, y , z ) (1.38a)
cos sin 0
where the rotation matrix is R z () sin cos 0 (1.41b)
0 0 1
The subscript z in Eq. (1.41b) indicates that the rotation is about the z-axis
and is the magnitude of rotation. The one-parameter group of rotations
about an axis is SO(2) which is a sub group of SO(3).
Since rotations about a fixed axis add up, it is possible to construct a finite
rotation about a fixed axis by adding up a large number of infinitesimal
rotations. An infinitesimal rotation about the z-axis can be written as:
20
Unit 1 Symmetry in Quantum Mechanics-I
2 3
1 ... ... 0
2! 3!
3 2
R z ( ) 1 0
3! 2!
0 0 1
2
I A z Az 2 ... exp(Az ) (1.42a)
2!
where the matrix Az is said to be the generator of the infinitesimal rotation
about the z-axis. Similar rotation matrices R x and R y can be defined for
infinitesimal rotations about the x and y-axes respectively:
R x ( ) exp( Ax ) ; R y ( ) exp( Ay ) (1.42b)
system to be symmetric under rotation we must have (Eq. 1.38a): Rˆ1 Rˆ 2 Rˆ 3 Rˆ1Rˆ 2 Rˆ 3
R (r ) (r ) R (r ) (R 1r ) (1.44a)
where R 1r are the basis kets for the rotated wave function. You know that
the operator R̂ is represented by a matrix. If the Hilbert space is
N-dimensional, the order of the matrix corresponding to the operator R̂ is
N N . So if the Hilbert space is 2-dimensional, the matrix is 2 2 and so on.
In this respect the matrix corresponding to the rotation operator R̂ is different
from the rotation matrices R you have studied about in Unit 19 of MPH-001,
because they are always 3 3 and act on the three components of a three
dimensional classical vector.
From Eq. (1.44b), since the state vector is arbitrary we can write:
r Rˆ R 1r (1.45a)
which has the Hermitian conjugate: Rˆ † r R 1r (1.45b)
However we also know that while r represents a system localised at r ,
Rˆ r represents a rotated system which is now localised at Rr , so we can
As you know, the write:
rotation matrices
satisfy the property of Rˆ r Rr (1.45c)
closure, therefore for
any two rotation
Now let us consider the action of Rˆ Rˆ † on a state r . From Eq. (1.45b) we
matrices, R1 and R 2
which are elements of get
the rotation group, we
can write: R1R 2 R3
Rˆ Rˆ † r Rˆ R 1r Rˆ Rˆ † r RR 1r Rˆ Rˆ † Î (1.46)
where R 3 is also an
element of the rotation The rotation operators also have another important property.
group.
The set of rotation operators R̂ form a representation of the rotation
group
Rˆ 1 Rˆ 2 r Rˆ 1 R2 r Rˆ 1Rˆ 2 r R1R2r
22
Unit 1 Symmetry in Quantum Mechanics-I
Since we know that because R1R 2 R3 (see margin remark), Eq. (1.45a)
can also be written as
Rˆ 1Rˆ 2 r R3 r Rˆ 1Rˆ 2 Rˆ 3 (1.47b)
Rˆ ˆI Rˆ (1.47c)
Rˆ Rˆ 1 Î (1.47d)
Rˆ 1 Rˆ 2 Rˆ 3 Rˆ 1Rˆ 2 Rˆ 3 (1.47e) Contd.....
Similarly
You can prove some of the properties of the rotation operators in the following
SAQ. Rˆ Rˆ 1 r Rˆ R 1 r
R R 1 r r
SAQ 4
Rˆ Rˆ 1 Î
a) Show that the rotation operator R̂ is linear.
Also
b) Show that the rotation operator conserves the scalar product, i.e. for any
two states and , is
Rˆ 1 Rˆ 2 Rˆ \ 3 r
R Rˆ and R Rˆ we have R . Rˆ 1 R 2 R3 r
R1 R 2 R 3 r
Let us now construct the general operator R̂ for an infinitesimal rotation
Rˆ 1 Rˆ 2 Rˆ 3 r
about a fixed axis, characterized by the unit vector n̂ . In analogy with
Eq. (1.18) for the translation operator, let us assume that the infinitesimal Similarly
rotation operator has the form: Rˆ 1 Rˆ 2 Rˆ 3 r
Rˆ 1Rˆ 2 R3 r
Rˆ n () ˆI iKˆ (1.48)
R1 R 2 R 3 r
where K̂ is a Hermitian operator, which is the generator of the rotation. Just
Rˆ 1 Rˆ 2 Rˆ 3 r
as we have shown that the generator of translation in any direction is related
to the corresponding linear momentum in that direction, let us start with the
assumption that the orbital angular momentum operator L̂ is related to the
generator of infinitesimal rotation and so:
Rˆ n ( ) ˆI i
Lˆ nˆ (1.49)
where Rˆ n () is the rotation operator for an infinitesimal rotation about the
direction specified by the unit vector n̂ . You can check for yourself that the
operator Rˆ (nˆ ) is unitary.
Lˆ
Rˆ z ( z ) ˆI i z z (1.50a)
And for the rotations x and y about the x and y axes respectively we have:
Lˆ
Rˆ x ( x ) ˆI i x x (1.50b)
23
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
y Lˆy
Rˆ y ( y ) ˆI i (1.50c)
Let us now use the definition of the rotation operators and the one-to-one
correspondence between the rotation operators and the rotation group to
derive the angular momentum commutation relations.
From Eq. (1.42c) we can write for infinitesimal rotations x , y and z about
the x , y and z axes respectively:
Expanding the functions in Eq. (1.51) up to the first term only we get:
Lˆ ˆ ˆ y Lˆ y ˆ ˆ x y Lˆ z
ˆI i y y ˆI i x L x I i ˆI i x L x I i (1.52)
which can be written as:
Lˆ ˆ Lˆ ˆ
ˆI i y y i x L x y x Lˆ y Lˆ x ˆI i y y i x L x y x Lˆ y Lˆ x
2 2
x y Lˆ z
Î i (1.53a)
Dropping terms of order higher than y x , we get from Eq. (1.53a):
ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ
ˆI i y Ly i x Lx i y Ly i x L x x y Ly L x x y Lx Ly 2 x y Ly L x
2 2 2
x y Lˆz
ˆI i
which is:
ˆ ˆ ˆ ˆ ˆ
ˆI x y Ly Lx x y Lx Ly ˆI i x y Lz (1.53b)
2 2
Thus we get:
Lˆx ,Lˆy iLˆz (1.54)
SAQ 5
Using the properties of the rotation operator show that Lˆy , Lˆz iLˆ x .
Now a finite rotation can be thought of as the result of a large number of
successive infinitesimal rotations. For example, writing a finite rotation z
24
Unit 1 Symmetry in Quantum Mechanics-I
z
about the z-axis as N successive infinitesimal rotations each equal to , in
N
the limit of N , we can write:
N
Lˆ Lˆ
Rˆ z ( z ) limN ˆI i z z exp i z z (1.55a)
N
Similarly:
Lˆ ˆ y Lˆ y
Rˆ x ( x ) exp i x x ; R y ( y ) exp i (1.55b)
In analogy with Eq. (1.49) we can write the rotation operator for a finite rotation
about the direction n̂ as
Lˆ nˆ
Rˆ n () exp i (1.56)
which is
x x
Lˆ nˆ Lˆ nˆ Lˆ nˆ
y y z z
R n ( ) exp i
ˆ (1.57)
n̂ x , n̂ y and n̂ z are the components of the vector n̂ along the x, y and z
axes, respectively. However the components of the angular momentum
operators do not commute so:
Lˆ nˆ Lˆ nˆ Lˆ nˆ
x x y y z z
Lˆ x nˆ x
Lˆ y nˆ y
Lˆz nˆ z
exp i exp i exp i exp i
You can check for yourself that the rotation operators for finite rotations
defined in Eqs. (1.55 and 1.57) are also unitary.
At this you should also note that we have not yet explicitly used the definition
of the angular momentum operators in terms of the Cartesian operators, for
example Lˆ z xˆpˆ y yˆpˆ x .
SAQ 6
a) Show that the rotation operators for finite rotations are unitary.
b) Rˆ z (2) Î
The discussion of this section for spinless particles can be extended to particle
with spin by replacing the orbital angular momentum L̂ by the total angular
momentum Jˆ Lˆ Sˆ . So we can write for infinitesimal rotations x , x and
y about the x , y and z axes respectively :
Jˆ y Jˆy Jˆ
Rˆ x ( x ) ˆI i x x ; Rˆ y ( y ) ˆI i ; Rˆ z ( z ) ˆI i z z (1.58a)
And for the finite rotations x , x and y about the x , y and z axes
respectively we have : 25
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
Jˆ Jˆ ˆ
Rˆ x ( x ) exp i x x ; Rˆ y ( y ) exp i y y ; Rˆ ( ) exp i z J z
z z
(1.58b)
Next let us study how an observable O changes under the rotation of the
system.
1.3.3 Rotation of an Observable
Consider an operator Ô which describes an observable O of the physical
system. Let 1 , 2 …. n be the eigenkets corresponding to the discrete
non-degenerate eigenvalues of Ô :
Oˆ i ai i (1.59a)
Yet:
Oˆ R i R ai i R OˆRRˆ i ai Rˆ i (1.59c)
Rˆ †Oˆ R Rˆ i ai Rˆ † Rˆ i (1.59d)
SAQ 7
iHˆ (t t0 )
(t ) Uˆ (t , t 0 ) (t 0 ) ; Uˆ (t , t 0 ) e ( t t0 ) (1.62)
We also know that the condition that the inner product of the state vectors is
conserved, that is : (t (t (t 0 ) | (t 0 ) requires the unitarity of the time
evolution operator. Another important property of the time evolution operator,
is that the time evolution of a state from t 0 to t 2 is equivalent to a time
evolution of the state from t 0 to t1 followed by a time evolution from t1 to t 2 :
(t ) Uˆ (t 2 , t 0 ) (t 0 )
iHˆ ( t 2 t0 )
e (t 0 )
So
Uˆ (t 2 , t 0 ) Uˆ (t 2 , t1 )Uˆ (t1, t 0 ) (1.64)
Let us derive the time translation operator in the same way that we derive the
space translation and rotation operators. We first talk about an infinitesimal
time translation as we have in the previous two sections. Consider the
evolution of a state vector (t 0 ) over an infinitesimal time interval . So:
lim 0 Uˆ (t 0 , t 0 ) Î (1.66)
As before we assume that the time translation operator for an infinitesimal time
can be written as (Eq. 1.48)
27
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
Hˆ
the generator of time translation. With this idea, let us write: Kˆ and then
Eq. (1.66a) reduces to:
Hˆ
Uˆ ( ) ˆI i (1.67b)
You can check for yourself that the time translation operator for infinitesimal
time translation is unitary:
Uˆ ()† Uˆ () Î (1.68a)
And for infinitesimal translations 1 and 2 :
SAQ 8
Show that i) Uˆ ()† Uˆ () Î ; and ii) Uˆ (1 2 ) Uˆ (1 )Uˆ (2 ).
A finite time translation can now be constructed by dividing the time interval
into a large number (N) of infinitesimal time intervals to write:
N
N
ˆ ˆ Hˆ Hˆ
U () limN I i exp i (1.69)
N
which is identical to the time evolution operator of Eq. (1.62) with t t 0 .
Also:
iHˆ t iHˆ t iEi t
(t ) e (0) ci e Ei ci e Ei (1.72b)
i i
iHˆ t iEi t
(t ) (0) e ci Ei e (1.72c)
i
28
Unit 1 Symmetry in Quantum Mechanics-I
iHˆ t iHˆ t
Aˆ (t ) (0) e Aˆ e (0)
iE j t iE t
i
cj Ej e
j i
Aˆ c i e E i
i Ei E j t
c i c j E j Aˆ E i e (1.73)
i j
E i Ej
If we write ij then we get:
iij t
Aˆ (t ) c i c j E j Aˆ E i e (1.74)
i j
With this we come to the end of our description of the transformation of the
wavefunction / state ket describing the physical state of the system and
observables under space and time translation and rotation of the physical
system.
Note that all these operations on the physical system, involve changes that
can be looked upon as successive infinitesimal changes (translation or
rotation). So these are continuous transformations. However all
transformations in quantum mechanics are not continuous transformations.
They can involve discrete changes. We will look at these in the next Unit.
We now discuss what how symmetry operations lead us to the conservation
principles.
d L L
0
dt q i q i
L
When L does not contain q i ( q i is a cyclic coordinate), then 0 and
q i
d L
0 pi constant . If L is not a function of q i , this also means
dt q i
that L is invariant under an infinitesimal translation defined by q i q i q i .
So the invariance of L under a translation of the coordinate q i implies that the
canonical momentum p i corresponding to the generalized coordinate q i is a
constant of motion or as we usually say: p i is a conserved quantity.
29
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
So whenever the Lagrangian or the Hamiltonian of a system are invariant
under a translation or have a symmetry corresponding to the translation
q i q i q i , there is a conserved quantity, which is the momentum.
Now we have already defined the operators for space and time translation and
rotation in Secs. 1.2 to 1.4. We know that the unitary operator for infinitesimal
translation or rotation is:
Sˆ ( ) ˆI iKˆ (1.75)
Sˆ () S Sˆ † ( ) S (1.76)
Let us now say that invariance of the system under the symmetry
operation Ŝ implies that the following is true:
Hˆ S Hˆ S Hˆ Sˆ † Hˆ Sˆ (1.77a)
Hˆ i Kˆ , Hˆ (1.78)
From this we get that for the Hamiltonian Ĥ of the system to be invariant under
the symmetry operation Ŝ , we must have:
Kˆ , Hˆ 0 (1.79)
pˆ
Tˆ † ( k )Hˆ Tˆ ( k ) Hˆ Tˆ ( k ), Hˆ 0 where Tˆ () ˆI - i k . Therefore
30
pˆ k , Hˆ 0 and
d ˆk
p
dt
0 . So the momentum pk is a constant of motion.
Unit 1 Symmetry in Quantum Mechanics-I
The translational invariance of the Hamiltonian implies that the same
experiment performed at two different places in the earth will give the
same result (with respect to the local observers).
Every interaction known to man is translationally invariant, hence an
experiment repeated at any site will always give the same result. This is
why momentum conservation applies to any system irrespective of the
nature of the interactions in the system and reflects the homogeneity of
space.
If the Hamiltonian of the system is invariant under rotation, then
Rˆ k † ( )Hˆ Rˆ k ( ) Hˆ where Rˆ k ( ) ˆI iLˆk , so Rˆ k ( ),, Hˆ Lˆk , Hˆ 0 and
d Lˆk
so 0 . The angular momentum Lk is a constant of motion and L̂k
dt
and Ĥ have a common set of eigenvectors.
L̂k can be written in terms of the operators xˆ i , pˆ i which transform under
rotation in the same way as their classical counterparts x i , pi as shown in
Eqs. (1.36). So if the operator Ĥ is invariant under rotation, it means that
the corresponding classical Hamiltonian is invariant too and vice-versa.
Hence the result of any experiment and its rotated version will be the
same.
If the Hamiltonian Ĥ of the system is invariant under time-translation
Hˆ
Uˆ ( ) ˆI - i , then we must have that the result of an experiment
performed on the system at any instant of time must be identical.
Consider a system prepared in an initial state (t1 ) at time t1 which
undergoes an infinitesimal time-translation to the state (t1 ) . So
Hˆ (t1 )
(t1 ) ˆI i ( t 1 )
Hˆ (t 2 ) Hˆ (t 2 )
(t 2 ) ˆI i (t 2 ) ˆI i ( t 1 )
SAQ 9
Show that for a symmetry transformation Ŝ , if E i is an eigenstate of Ĥ with
an eigenvalue Ei , then Ŝ E i is also an eigenstate of Ĥ with the same
eigenvalue Ei .
1.6 SUMMARY
A symmetry operation on any system can be defined as an action (
some change) that leaves the physical system invariant. In quantum
mechanical systems, all transformations are described by operators
acting on the quantum mechanical states. A symmetry transformation
on a quantum system is a transformation that leaves the transition
amplitudes between quantum mechanical states unchanged.
Wigner’s theorem states that a symmetry transformation in the Hilbert
space of states could be represented by either a unitary or an anti-
unitary operator.
Space Translation
The wavefunction of a one-dimensional system following an
pˆ
infinitesimal translation x along x is: T ( x ) 1 i x x ( x )
32
Unit 1 Symmetry in Quantum Mechanics-I
For a infinitesimal translation ( x , y , z ) the translation operator
pˆ
is Tˆ ( ) ˆI i .
pˆ
For a finite translation exp i .
Rotation
The wavefunction of a quantum system following an infinitesimal
rotation by an angle about the z-axis is:
Lˆ
R ( x, y , z ) 1 i z ( x, y , z )
Lˆ y Lˆy Lˆ
Rˆ x ( x ) ˆI i x x ; Rˆ y ( y ) ˆI i ; Rˆ z ( z ) ˆI i z z
The angular momentum commutation relations follow from the
properties of the rotation operator since rotations along different
directions do not commute.
For a finite rotation about the direction n̂ the rotation operator is:
Lˆ nˆ
Rˆ n () exp i
Conservation Laws
If the Hamiltonian of a physical system is unchanged under a
translation or rotation of the system defined by an operator Ŝ , Ŝ is a
symmetry operation and the generator of the symmetry operator is a
constant of motion.
33
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
If the Hamiltonian of the system is invariant under translation
pˆ
defined by the operator Tˆ ( ) ˆI i k the linear momentum
momentum pk is a constant of motion.
4. Show that for a finite rotation of 2 about the z-axis for a spin-half system,
Rz ( 2)
Tˆ ( )Tˆ ( x ) Tˆ ( x )Tˆ ( )
†
2. i) Tˆ † ( x ) ˆI i x Kˆ x ˆI i x Kˆ x Tˆ ( x )
Kˆ Kˆ †
x x
ii) Tˆ † ( x )Tˆ ( x ) Î Tˆ † ( x ) T 1( x )
34 Tˆ 1( x ) ( ˆI i x Kˆ x )† ˆI i x Kˆ x Tˆ ( x )
Unit 1 Symmetry in Quantum Mechanics-I
3. Since Tˆx ( x )Tˆ ( y ) Tˆ ( y )Tˆ ( x ) we can write
ˆ i x pˆ x ˆ i y pˆ y ˆ i y pˆ y
I
ˆ i x pˆ x
I
I I
x y i pˆ i y pˆ y ˆ y x i y pˆ y i x pˆ y
ˆI pˆ x pˆ y x x I pˆ y pˆ x
2 2
pˆ x pˆ y pˆ y pˆ x 0 pˆ x , pˆ y 0
R̂ is a linear operator.
b) R Rˆ R Rˆ †
R Rˆ † Rˆ since Rˆ † Rˆ Î .
Expanding the functions in Eq. (i) up to the first term only we get:
ˆ z Lˆ z ˆ y Lˆ y ˆ ˆ y Lˆ y Lˆ
I i I i ˆI i z Lz I i ˆI i y z x
(ii)
which can be written as:
ˆ Lˆ ˆ Lˆ
ˆI i z Lz i y y z x Lˆz Lˆ y ˆI i z Lz i y y z x Lˆ z Lˆ y
2 2
y z Lˆ x
Î i (iii)
Dropping terms of order higher than y x , we get from Eq. (iii):
ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ
ˆI i z Lz i y L y i z Lz i y L y y z Lz L y y z Ly Lz 2 y z Lz Ly
2 2 2
y z Lˆ x
ˆI i
y zLˆzLˆy y zLˆy Lˆz y zLˆx
which is: ˆI
2
2
ˆI i
Lˆy , Lˆz iLˆx
Lˆ.nˆ ˆ † i Lˆ.nˆ
6. a) Rˆ n () exp i ; R n () exp
i Lˆ.nˆ i Lˆ. nˆ
So, Rˆ n () Rˆ n () e ˆI Rˆ n† () Rˆ n ()
†
e 35
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
Hence Rn () is a unitary operator.
Lˆ z Lˆz ml C l ,m l
ml n, l , ml
l , ml
i ( 2 )Lˆz
Rˆ z ( 2) Cn, l ,m e e
n, l , ml
n,l ,ml
Cn, l ,m e im 2 n, l, ml
l
l Cm l
n, l , ml
nl ml nl ml
Sˆ x Sˆ x Rˆ z† Sˆ x Rˆ z
R R
z i Jˆz i z JˆZ
Rˆ z† Sˆ x Rˆ z e e
2
Jˆ z ; Jˆ z
2 2
i z i
z i z Jˆz
ˆ † ˆ ˆ
Rz Sx Rz e 2 e 2 e 2
2
i z i z i i
z z
e 2 e 2 e 2 e 2
2
e e i z
i z
2
(cos z i sin z ) (cos z i sin z )
2
ˆ
2
S x cos z Sˆ y sin z
S x R S x cos S y sin
i (1 2 ) Hˆ
ii) Uˆ (1 2 ) ˆI
i Hˆ i Hˆ i 1 2 Hˆ 1 2 Hˆ 2
Uˆ 1 Uˆ 2 ˆI 1 ˆI 2 ˆI
2
ˆI
i 1 2 Hˆ
Uˆ 1 2 (dropping terms of order 1 2 )
Hˆ Sˆ E1 Sˆ Hˆ E i E i Sˆ Ei
Terminal Questions
1. Using Eq. (1.33)
pˆ i y pˆ y i z pˆ z
Tˆ ( ) ˆI i x x
ˆ
xˆ,Tˆ ( )
xˆ, pˆ x
[ xˆ, pˆ y ] [ zˆ, pˆ z ]
x, I i x i y i z
ˆ
i i y i .0
0 i x .0 z x
2. xˆ xˆ
After translation:
xˆ Tˆ † ( x ) xˆ Tˆ ( x )
T
xˆ Tˆ ( x ) Tˆ ( x ) xˆ x we have xˆ Tˆ Tˆ xˆ x
xˆ T Tˆ † Tˆ xˆ x xˆ x
Similarly, since pˆ x Tˆ ( x ) Tˆ ( x ) pˆ x
pˆ x T pˆ x T T † pˆ xTˆ T † ( x )Tˆ ( x ) pˆ x
p̂ x p̂x
i
3. i) We use Sˆ y and follow the steps of SAQ 7.
2
i Sˆ z z iSˆ
With Sˆ y e Sˆ y e we get
R
i Sˆ z iSˆ z
i
e i Sz Sˆ y e iSz e e
ˆ ˆ
2
37
Block 1 Symmetry and Conservation Laws in Quantum Mechanics
i i i i
i
e 2 e 2 e 2 e 2
2
i
(cos i sin ) (cos i sin )
2
cos Sˆ y sin Sˆ x
Sˆ y cos Sˆ y sin Sˆ x
R
ii) Sˆ z
2
iSˆz i Sˆ z i i i i
e Sˆ z e e 2 e 2 e 2 e 2 Ŝ z
2
Sˆ z Sˆ z
R
c1 c 2
i 2 Sˆ z
e e i e i
Now, e i e i 1
R z ( 2)
5. We have j, j and
i Jˆ y
R R y ( ) j , j 1 j, j
1
Jˆ y ( Jˆ Jˆ ) (Eq. 13.59, MPH-004)
2i
Jˆ j , j 0 ; Jˆ j , j 2 j j, j 1
i ˆ ˆ ˆ
R 1 (J Jˆ ) j , j j , j J j, j J j, j
2i 2 2
j, j
2
2 j j, j 1
2j
j, j 1 j, j R 2j
2 2
2 j
The probability of being in the state j , j 1 is .
2
38