____________________________________________________________________________________________________

Material science

Paper No. : Quantum Mechanics

Module : Basic Postulates of Wave Mechanics

Development Team
Prof. Vinay Gupta, Department of Physics and Astrophysics,
Principal Investigator University of Delhi, Delhi

Prof. V. S. Bhasin (Retired) ,Department of Physics, University

Paper Coordinator
of Delhi, Delhi-110007

Content Writer Prof.V. S. Bhasin, (Retired),Department of Physics,

University ofDelhi, Delhi-110007

Content Reviewer

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

Subject Physics

Paper No and Title Quantum Mechanics I

Module No and Title 11 Basic Postulates of Wave Mechanics: Representaton of

States and Observables
Module Tag V S Bhasin

TABLE OF CONTENTS
1. Learning Outcomes
2. Introduction
3. . Postulates of Wave Mechanics: Repesentation of States and
Observables

3.1(a) Postulate 1: Representation of states; Postulate 2. Superposition

Principle

3.2 Dirac bracket notation:

3.3 Representation of dynamical variables and operators: Observables,

Expectation values; Postulate 3

3.3.1 Commutation Relations

3.4 Eigen values and Eigen functions: Postulate 4

3.4.1 Hermitian Operators

3.5 Expectation Values : Postulate 5

3.5.1 Adjoint Operator and Self-adjoint operators

3.5.2 Functions of operators

3.5.3 Inverse and Unitary Operators

3.5.4 Projection Operators

4. Summary

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

1. Learning Outcomes

After studying this module, you shall be able to

 Learn postulates 1 and 2 in terms of the representation of states and the

significance of superposition principle
 Know postulate 3 introducing operators and expectation values and commutation
relations
 Learn postulate 4 giving the definition of Eigen-values, Eigen-functions and
Hermitian operators
 Study postulate 5 introducing the definition of Expectation values, adjoint and
self-adjoint operators
 Learn the definition of functions of operators, inverse and unitary operators and
projection operators

2. Introduction

Having introduced the elementary ideas of quantum mechanics in the previous

modules, we have learned so far about the wave mechanical description of single particle
systems by solving the one-dimensional Schrodinger equation. The wave nature of matter
and the correspondence principle have led us to certain fundamental notions which we
may regard as postulates of wave mechanics. We have mow come to a stage where
mathematical structure of wave mechanics and its physical content in terms of these
postulates should be made more explicit.

3. Postulates of Wave Mechanics: Repesentation of States and Observables

3.1

(a) Postulate 1: Representation of states

The state of a quantum mechanical system is represented by a wave function


 (r , t ).

The wave function contains all the information that can be known about the
system at any instant of time. This function is, in general complex and is not directly
measurable. In fact, all constant multiples of a given 𝝍 (i.e., c 𝝍, for real or complex
numbers c) describe one and the same state. We often use the state 𝝍 which implies the

same physical concept as described by  (r , t ).

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

(b) Postulate 2: The Superposition Principle

According to this principle, which we have already described earlier, if

 1 and  2 are wave functions describing any two states of a given system, then
corresponding to every linear combination, ( c1 1  c2 2 ) , of the two functions there
exists a state of the system. This fundamental principle of quantum mechanics has no
analog in classical mechanics. We have already assumed this in constructing wave packet
states of a free particle. The observed phenomenon of interference is a direct consequence
of this principle.

From a mathematical stand point, this principle is a fundamental characteristic of

 
vectors representing positions in space. Recall that if v1 and v2 are the position vectors
 
of any two points, c1v1  c2 v2 is again a vector representing some point in space. We may
think of wave functions too as vectors of some kind. Ordinarily, the vectors which we are

familiar with, need only three real numbers for a complete specification; a vector v in
three dimensional space is specified completely by the three components, v1 , v2 , v3 .
However, to specify a wave function vector we would require an infinite set of complex

numbers, namely the values of  at all points r in configuration space. Thus the wave
function vectors are, in fact, infinite dimensional. The set of all wave functions of a given
system forms an infinite dimensional complex linear vector space.
 
Thus the wave functions  (r ) for different values of r are thought of as

different components of the wave function vector  ; the continuous variable r serves
as label distinguishing the various components and is a counter part of the index I

labeling the components vi of an ordinary vector v . Identifying this way, the norm
 2 
  (r ) dr is the analogue of  vi and may therefore be regarded as the square of the
2

i
  2
length of the wave function vector. Since  (r ) is complex, the absolute square  (r )

has to be taken to get the length real and, indeed, it has to be integrated over r in order
to take care of the summation. More generally, analogous to the scalar product of two
  
vectors  ui vi we can also define    (r ) (r ) dr as a scalar product of two wave
i
functions  and  . We often use the notation ( , ) for the scalar product as the

counterpart of the scalar product of two vectors, expressed as (u.v ) .

From the definition of scalar product of two wave functions

  
( , )     (r ) (r ) dr (11.1)

it follows that

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

( , )  ( , )  ,
( , c )  c( , ) and (c  , )  c  ( , ),
Also Norm of   ( , )  0 (11.2)

The equality sign holds if and only if   0 .

3.2 Dirac bracket notation:

We introduce a very convenient notation,due to Dirac, where the scalar product of

 
two square integrable functions  1 (r ) and  2 (r ) is denoted by the symbol  1  2 ,
explicitly written as:
  
 1  2   1 (r ) 2 (r ) dr (11.3)

The symbol  2 is known as a ket while  1 is known as a bra. From the definition
(11.3) it follows that


1  2   2 1 (11.4)

Also if c is a complex number and we have a third function  3 , we have the relations:

 1 c 2  c  1  2 (11.5a)
c 1  2  c   1  2 (11.5b)
 3  1  2   3  1   3  2 (11.5c)

Two functions  1 and  2 are said to be orthogonal if their scalar product vanishes:

 1  2  0. (11.6)

Using the Dirac bracket notation, the normalization condition is written as:

1  2  1 (11.7)

3.3 Representation of dynamical variables and operators: Observables,

Expectation values

Postulate 3

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

In quantum mechanics each dynamical variable,


A( r , p) , is represented by a linear operator
 
Aˆ  A( xˆ , pˆ )  A(r ,i).

The operators act on the wave functions of the system. An operator  acting on the
wave function converts it into another wave function Â .

Linearity of the operator means that a linear combination of two or more wave functions
 1 and  2 is converted into the same linear combination of A 1 and A 2 ‘

In general, Aˆ (c1 1  c2  2 )  c1 ( Aˆ  1 )  c2 ( Aˆ  2 ) (11.8)

ˆ ˆ ˆ ˆ
These operators do not in general commute with each other: A B  B A . This difference

Aˆ Bˆ  Bˆ Aˆ is called the commutator of Aˆ and Bˆ and is denoted as [ Aˆ , Bˆ ]  Aˆ Bˆ  Bˆ Aˆ .

3.3.1 Commutation Relations: As a simple example of deducing the values of the

commutators, we have seen earlier in the case of position and momentum variables in one
dimension ,

   
( xˆpˆ  pˆ xˆ )  i x   ( x )  i
 x x  (11.9)

It is thus clear that the operator [ xˆ, pˆ ] has the effect of simply multiplying any arbitrary
function  by i  . We can, therefore, write [ xˆ, pˆ ]  i. This relation can be extended to
three dimensions by writing the components as xi and p j , where (i,j=1,2,3 and
x1 , x2 and x3 represent the components x, y, z and similarly p1 , p 2 , p3 denote
p x , p y and p z respectively. It is a simple exercise to show that the commutator

[ xi , p j ]  xi p j  x j pi  i ij
(11.10)

where  ij is the Kronecker delta function defined as

 ij  1 if i  j and  ij  0 if i  j.
(11.11)

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

It is a trivial matter to verify that

[ xi , x j ]  0 , [ pi , p j ]  0 and [ x, p x ]  [ y, p y ]  [ z, p z ]  i . (11.12)

3.4 Eigen values and Eigen functions

Postulate 4

The only result of a precise measurement of the dynamical variable A is one of the
eigen values an of the linear operator  associated with A.

The results of measurement are real numbers. Therefore the eigenvalue

spectrum, which represents the totality of eigenvalues of the operator must be real. In
some cases, this spectrum may consist only of discrete eigenvalues, in other cases, it may
be continuous or even a mixture of both- continuous and discrete eigenvalues.

3.4.1 Hermitian Operators

This requirement that the linear operators representing dynamical variables must
have real eigenvalues imposes the condition that these must be Hermitian operators,
defined as:

 Aˆ   ( Aˆ  )  (11.13)
If  n is an eigenfunction of the operator  corresponding to the eigenvalue an , then we
have

Aˆ  n  an n (11.14)

It follows that

 n Aˆ  n  an  n  n (11.14)

Also, since

( A n )   a n n (11.15)

We also have

( Aˆ  n )  n  a n  n  n (11.16)

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

Clearly, if the operator  is Hermitian , left hand sides of

Eqs.(11.14) and (11.16) are equal and hence a n  a n . Thus a sufficient condition for an
operator to have real eigenvalues is to be Hermitian.

Example:


Let us take the example of linear momentum operator pˆ x  i in one dimension.
x
Using Eq.(11.13), the matrix element


d d 
 pˆ x    (i)  
 
dx   i      i  dx
dx  dx  (11.17)

In Eq.(11.17), we simplified the right hand side by partial integration. Since the wave
 and 
functions are normalizable wavefunctions and therefore they must vanish at
x  
both the limits ( ) , the first term on the right hand side vanishes under these
limits. The second term reduces to

 ( pˆ  ) dx  pˆ x 

x
(11.18)

p̂ x
showing thereby that is Hemitian.


pˆ x  i
Now to find the eigenfunctions of the operaor x , suppose that the operator
 (x) px
acting on the function has the eigenvalue . Thus

 ( x)
 i  p x ( x)
x (11.19)

which can be written as

d i
 p x dx
  (11.20)

On integrating both sides of this equation, we have

 ( x)  C exp(ip x x / )  C exp(ik x x)
, (11.21)
Physics PAPER No. : TITLE
MODULE No. : TITLE
____________________________________________________________________________________________________

C is a normalization constant.We in this example

considered the simple one-dimensional case, which can be
easily extended for a three dimensional case, where we can write
  
 p (r )  C exp(ip..r / )
(11.22)

  
p̂  i p
as the eigenfunction of the operator having momentum vector as an
eigenvalue.

3.5 Expectation Values

Postulate 5

If a large number of measurements of a dynamical variable are made on a system


representing the same state , the results of the individual measurements will, in
general, be different but the expectation or average value of the dynamical variable is
expressed as

 Aˆ 
Aˆ 
 (11.23)

 Â
Since is a Hermitian operator, the is real. In case the wavefunction is normalized
  1
to unity, we have , we write Eq.(11.23) as

Aˆ   Aˆ 
(11.24)


Note that since each system is identical and is described by the wavefunction ,
Â
does not represent the average of a classical statistical distribution of the dynamical
Â
variable A. The actual value of the operator obtained in an experiment on a single
 Â
system is inherently unpredictable, unless, of course, is an eigenfunction of . For
instance, in a Stern-Gerlach experiment, measurements of the component in a certain
direction of the magnetic moment of the atom can be immediately repeated by passing
the beam through a second apparatus. In such a case, it is reasonable to expect that if a

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

an
particular result is obtained in the first measurement,
the same result will be obtained if the measurement is repeated immediately. Since the
result of the second measurement can be predicted, we infer that after the first
measurement, the state of the system is described by the eigenfunction
 n of Aˆ an
belonging to the eigenvalue .

3.5.1 Adjoint Operator and Self-adjoint operators

Â
It is useful to introduce an operator , called the adjoint or Hermirian conjugate
Â
of the linear operator by the relation


 Aˆ    ( Aˆ  )    Aˆ 
(11.25)

 and 
where are any pair of square integrable functions.

X
If we define a bra by the relation

X   Aˆ 
, (11.26)

Â 
where the operator acts to the left on the bra , then it follows from Eq.(11.25) that
X and 
kets are given by the relation

X  Aˆ 
(11.27)

Â
A linear operator satisfying

Aˆ  Aˆ 
(11.28)

is aid to be self-adjoint. From Eq.(11.13), we see that a self-adjoint operator is Hermitian.

Note that as in the case of a complex number, we have its complex conjugate,in a similar
way we define the adjoint of an operator. A self-adjoint operator is the operator
Â
generalization of a real number. The operator is not, in genera, equal to the operator

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________


Â pˆ x  i
. For example, the operator x is Hermitian

pˆ x  pˆ x pˆ x  i   pˆ x pˆ x  pˆ x
so that , but x . Thus . The adjoint operator has the
following three important properties:

(i) (cAˆ )   c  Aˆ  (11.29)

(ii) ( Aˆ  Bˆ )   Aˆ   Bˆ  (11.30)
(iii) ( Aˆ Bˆ )   Bˆ  Aˆ  (11.31)

3.5.2 Functions of operators

If a f(z) can be expanded in power series,


f ( z )   ci z i
i 0 (11.32)

f ( Aˆ )
Then the operator function can be defined as


f ( Aˆ )   ci Aˆ i
i 0 (11.33)

n  an
As a result, if is one of the eigenfunctions of corresponding to the value ,
Aˆ  n  (an )  n
i i
, we thus have

f ( Aˆ )  n  f (an ) n
(11.34)

f ( Aˆ )
The adjoint operator to , can be obtained as

 f ( Aˆ )   c ( Aˆ )   c ( Aˆ )
  
 i    i
i i
i 0 i 0
 
(11.35)
 f (A )

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

where we have used the properties given above. In

Â
particular, if is a self-adjoint operator, we get


)]  f  [ Aˆ ]
[ f (A
(11.36)

3.5.3 Inverse and Unitary Operators


The unit operaor is the operator that leaves the function unchanged. Thus

Iˆ  
(11.37)

 B̂
If for an operator , there exists another operator such that

Aˆ Bˆ  Bˆ Aˆ  I
(11.38)

B̂ Â Bˆ  Aˆ 1
Then is said to be the inverse of , i.e., . (11.39)

Û
A linear operator is said to be unitary if

Uˆ 1  U  or Uˆ Uˆ   Uˆ Uˆ  I
(11.40)

Uˆ  exp(iAˆ ),
A unitary operator can be expressed in the form: (11.41)

Â
where is a Hermitian operator. Now

Uˆ   exp(iAˆ )   exp( iAˆ )

(11.42)

Eqn(11.42) clearly satisfies the condition Eq.(11.40).

3.5.4 Projection Operators


Consider the ket state , which is normalized to unity:

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

  1
(11.43)

Let there be an operator defined as

P̂   
(11.44)


which is applied to an arbitrary ket state to give

P̂     
(11.45)

 
The projection operator acting on the ket state projects out the ket
with the
  by 
coefficient of proportionality , which, in fact, is the scalar product of .

P̂
The projection operator can thus be interpreted as the orthogonal projection operator

on to the ket . This interpretation is confirmed by the fact that

( Pˆ ) 2  Pˆ
(11.46)

To see this

( Pˆ ) 2  Pˆ Pˆ     

(11.47)


In this expression is simply a number which is equal to one, proving the assertion:

( Pˆ ) 2     P

4. Summary

In this module, you have been able to

 Learn postulates 1 and 2 in terms of the representation of states and the

significance of superposition principle

Physics PAPER No. : TITLE

MODULE No. : TITLE
____________________________________________________________________________________________________

 Know postulate 3 introducing operators and

expectation values and commutation relations
 Learn postulate 4 giving the definition of Eigen-values, Eigen-functions and
Hermitian operators
 Study postulate 5 introducing the definition of Expectation values, adjoint and
self-adjoint operators
 Learn the definition of functions of operators, inverse and unitary operators and
projection operators

Physics PAPER No. : TITLE

MODULE No. : TITLE