An Intuitive Approach To The Coherent and Squeezed States of The QHO

An intuitive approach to the coherent and
squeezed states of the quantum harmonic

oscillator (QHO)
Spiros Konstantogiannis
Physicist, M.Sc.
Copyright 2017 by Spiros Konstantogiannis. All rights reserved. No part of this
eBook may be reproduced, in any form or by any means, without the express written
permission of the writer.
2
Contents
An intuitive approach to the coherent and squeezed states of the quantum harmonic
oscillator (QHO)........................................................................................................ 1
Contents .................................................................................................................... 3
Preface ...................................................................................................................... 4
I. Preliminaries .......................................................................................................... 5
Applying a spatial translation to the ground state of the quantum harmonic
oscillator (QHO)................................................................................................ 5
More on spatial translation operators ................................................................13
Momentum translation operators.......................................................................21
The combined action of a spatial and a momentum translation operator ............29
Returning back to the QHO The displacement operator..................................31
II. The coherent states of the QHO ...........................................................................41
III. An intuitive introduction to the squeezed states of the QHO ...............................68
The coherent states as states of minimum energy expectation value ..................69
IV. General references..............................................................................................76
3
Preface
The purpose of the present eBook is to introduce, in a simple, intuitive way, the
coherent and squeezed states of the quantum harmonic oscillator (QHO), through a
series of exercises, which are solved in detail.
Starting from the application of a spatial translation to the ground state of the QHO,
we introduce the spatial and momentum translations, focusing on their application to
the QHO, which leads us to the displacement operator. We then introduce the
coherent states and examine their basic aspects. Next, we proceed to give a simple and
purely intuitive introduction to the squeezed states and we conclude by identifying the
coherent states as states of minimum energy expectation value compared to the
respective squeezed states.
The reader is assumed to have a basic knowledge of the postulates and the
mathematical formalism of quantum mechanics, including the Dirac notation and the
ladder operator method of the QHO.
4
Preliminaries
I. Preliminaries
Applying a spatial translation to the ground state of the quantum
harmonic oscillator (QHO)

ipx
1) At time t = 0 , the wave function of a QHO is y ( x ) = exp - 1 y 0 ( x ) ,
h
d
where y 0 ( x ) is the ground-state wave function, p = -ih is the momentum
dx
operator (in the position representation), and x1 is an eigenvalue of the
position operator (be careful, it is not the position operator).
i) Show that the wave function y ( x ) is normalized.
ii) Calculate the position and momentum expectation values of the QHO at
time t = 0 .
iii) Calculate the energy expectation value of the QHO at time t 0 .
It is given that
1
mw 4 mw 2
y 0 ( x) = exp - x
ph 2h
and
p b2
dx exp ( - ax 2 - 2bx ) =exp , a > 0.
-
a a
For the integral, see, for instance,
https://en.wikipedia.org/wiki/Gaussian_integral
Solution

ipx
i) Using the Taylor expansion of the operator exp - 1 , i.e.
h
n

ipx

- 1

ipx h
exp - 1 = ,
h n =0 n!
the wave function y ( x ) is written as
ipx 1
n
1
ipx
n
-
-
h y ( x ) = h y ( x )
y ( x) =
n =0 n! 0 n =0 n!
0

Since x1 is a number (not an operator), the momentum operator commutes with x1 ,
and with i too, and thus
n n
1 ix1 n
ipx
- = - p
h h
5
Preliminaries
Therefore, the wave function y ( x ) is written as

n n
ix1 ix1

-
- n
h n h d
y ( x) = p y 0 ( x ) = -i h y 0 ( x) =
n =0 n! n =0 n! dx
n
ix1
- ( -ih )
n

( i 2 x1 )
n

h dn
= n
y 0 ( x) = y 0( n ) ( x ) =
n= 0 n! dx n =0 n!
( - x1 ) y 0(n ) ( x )
n

= y 0 ( x) =
( n)
( - x1 ) = y 0 ( x - x1 )
n
n= 0 n! n =0 n!
Remember that the Taylor series of a (proper) function f ( x ) about x is written

f(
m)
( x )
as f ( x ) =
m =0 m!
( x - x )
That is
y ( x ) = y 0 ( x - x1 ) (1)
Using (1), we have

dx y ( x ) = dx y ( x - x )
2 2
- -
0 1 (2)
Changing the integration variable to

y = x - x1 ,
we have
dx = dy
and
lim y =
x
Thus, (2) becomes

dx y ( x ) = dy y ( y )
2 2
- -
0 = 1,
because the ground-state wave function is normalized.

Therefore, the wave function y ( x ) is also normalized.
Another way of showing that y ( x ) is normalized is by observing that the operator

ipx
exp - 1 is unitary.
h

ipx
Indeed, the Hermitian conjugate of exp - 1 is
h
6
Preliminaries

ipx 1
n

ipx
1
n
ipx

n

- 1
-
1 h

- h
=

ipx h
exp - h =
=
n =0 n! { n =0 n! n =0 n!
Ai = Ai
i i
That is

ipx
1
n
-
1
ipx

h
exp - h = (3)
n= 0 n!
But

ipx 1
n
ipx
ipx ipx
ipx
ipx
- = - 1 ... - 1 = - 1 ... - 1 - 1 =
h 144h42444 h3 144 h4 2444 h3 h

n times ( n -1) times

ipx n -1
1 ipx
1

ipx1
ipx1

ipx1
= - - ... - = - -
h 144 h4 2444 h3 h h

( n -1) times
That is

ipx 1 ipx
n
1 ipx

1
n -1
- = - -
h h h
Using repeatedly, n times, the previous relation, we obtain

ipx 1 ipx
n
1 ipx

1 ipx
n -1
1 ipx 1 ipx

1
n -2
- = - - = - - - =
h h h h
14442444 h
3 h
2 times
n -3
1 ipx

1 ipx
ipx 1 ipx 1
= - - - - = ... =
14444
h 4244444
h h 3 h
3 times

n -n n

1 ipx
ipx 1 ipx 1 ipx
1
= - ... - - = -
h h h h
1444 24443 14243
n times ipx 0
- 1 =1
h
That is
n
ipx 1 ipx
n
1

- = - (4)
h h
7
Preliminaries
By means of (4), (3) becomes

n
ipx
1
n
ipx
1

- -
1
ipx

h
h
exp - h = =
n= 0 n! n =0 n!
Since the momentum operator is Hermitian,

1 ipx
ipx 1
- =
h h
Thus
n
1
ipx

1
ipx 1
h = exp ipx
exp - =
h n =0 n ! h
That is

1
ipx 1
ipx
exp - h = exp h (5)

Using (5) we obtain

1
ipx 1
ipx 1
ipx 1
ipx
exp - h exp - h = exp h exp - h

1
ipx
ipx
Since the commutator of and - 1 is zero, applying the property
h h
( ) ( ) ( ) 1
2

exp A + B = exp A exp B exp - A , B

yields

ipx
ipx
ipx
ipx
exp 1 exp - 1 = exp 1 - 1 = exp 0 = 1
h h h h
In the same way,

ipx
ipx
ipx
ipx
ipx
ipx
exp - 1 exp - 1 = exp - 1 exp 1 = exp - 1 + 1 = 1
h h h h h h
Thus

1
ipx 1
ipx 1
ipx 1
ipx
exp - h exp - h = exp - h exp - h = 1

ipx
Therefore, the operator exp - 1 is unitary, and thus it preserves the norms of the
h
states on which it acts.
Then, in the state space, we have
8
Preliminaries

ipx
y = exp - 1 0 = 0
h
Thus
2 2
y = 0 yy = 00
Using the completeness of the position eigenstates, we have

y dx x x y = 0 dx x x 0
-
14243 -
14243
1 1

2 2
dx {
-
y x xy = dx {
-
0 x x0 dx
-
xy = dx
-
x0
* *
xy x0
Substituting x 0 = y 0 ( x ) and x y = y ( x ) , and using that the ground state is

dx y ( x )
2
normalized, i.e. 0 = 1 , we obtain
-
dx y ( x )
2
=1
-
ii) At time t = 0 , the wave function of the QHO is the wave function (1). Thus, its
position expectation value is written as

= dxy ( x - x ) xy ( x - x ) (6)
*
x 0 0 1 0 1
-
Changing again the integration variable to

y = x - x1 ,
we have
x = y + x1 ,
dx = dy
and
lim y =
x
Thus (6) becomes

dyy 0* ( y )( y + x1 )y 0 ( y ) = dyy 0* ( y ) yy 0 ( y ) + x1 dy y 0 ( y )
2
x 0
=
-

- -
dyy ( y ) yy ( y )
*
The integral 0 0 is the position expectation value of the QHO in the
-
ground state, which is zero.
9
Preliminaries
We remind that the position expectation value of the QHO in an energy eigenstate
is zero.
dy y ( y )
2
Also, 0 = 1 , since the ground-state wave function is normalized.
-
Thus
x 0
= x1 (7)
Likewise, the momentum expectation value of the QHO at t = 0 is

d
= dxy ( x - x ) -ih dx y ( x - x )
*
p 0 0 1 0 1
-
Doing again the variable change y = x - x1 , the previous equation becomes

d
= dyy ( y ) -ih dy y ( y )
*
p 0 0 0
-

d
dyy ( y ) -ih dy y ( y )
*
The integral 0 0 is the momentum expectation value of the
-
QHO in the ground state, which is zero.
We remind that the momentum expectation value of the QHO in an energy

eigenstate is also zero.
Thus
p 0
= 0 (8)
ipx
We see that the operator exp - 1 , acting on the ground state of the QHO,
h
changes (translates) the position expectation value by x1 while leaving unchanged the
momentum expectation value.
This is easily generalized to an arbitrary, but bound, state not necessarily an energy
eigenstate of a one-dimensional quantum system not necessarily of the QHO. The
ipx
constraint that the state on which the operator exp - 1 acts should be bound is
h
necessary so that it is normalizable and the respective integrals are finite.
Moreover, it can be shown see the following exercise that if the operator
ipx
exp - 1 acts on a position eigenstate x , it yields the position eigenstate
h
x + x1 , i.e.

ipx
exp - 1 x = x + x1
h
10
Preliminaries
Due to its property to translate the position, or the position expectation value, by its

ipx
argument x1 , the operator exp - 1 is a spatial translation operator or, simply, a
h
translation operator.
Obviously, for each position eigenvalue x1 , we can define a spatial translation
operator

ipx
Tx1 exp - 1
h
iii) Since the Hamiltonian of the QHO is time independent, the Ehrenfest theorem
gives for the time evolution of the QHO energy expectation value
d E i
t
= H , H = 0
dt h t
Thus
E t= E 0
(9)
At t = 0 , the state of the QHO is described by the wave function y ( x ) .

Thus, the energy expectation value of the QHO at t = 0 is

= dxy ( x ) H ( x )y ( x ) (10)
*
E 0
-
where H ( x ) is the Hamiltonian of the QHO in the position representation, i.e.

2
d
-ih 1
H ( x ) =
dx
+ mw 2 x 2
2m 2
Substituting (1) and the previous Hamiltonian into (10), we obtain
d
2

-i h 1
dx
E 0 = dxy 0* ( x - x1 ) + mw 2 x 2 y 0 ( x - x1 ) =
-
2m 2

2
d
-ih 1

dx
= dxy 0 ( x - x1 )
*
y 0 ( x - x1 ) + mw dxy 0* ( x - x1 ) x 2y 0 ( x - x1 )
2
- 2 m 2 -
That is
2
d
-i h 1

dx
E 0
= dxy 0 ( x - x1 )
*
y 0 ( x - x1 ) + mw dxy 0* ( x - x1 ) x 2y 0 ( x - x1 ) (11)
2
- 2m 2 -
Well calculate the two integrals separately.

Changing again the integration variable to
11
Preliminaries
y = x - x1 ,
we have
x = y + x1 ,
dx = dy
and
lim y =
x
Thus, the first integral on the right-hand side of (11) is written as

2
d
2
d
-i h -ih dy
dx
- dxy 0 ( x - x1 ) 2m y 0 ( x - x1 ) = - dyy 0 ( y ) 2m y 0 ( y ) (12)
* *
The second integral on the right-hand side of (11) is written as

dxy ( x - x ) x y ( x - x ) = dyy ( y )( y + x ) y ( y ) =
* 2 * 2
0 1 0 1 0 1 0
- -

= dyy ( y ) ( y + 2 x1 y + x12 )y 0 ( y ) =
* 2
0
-

= dyy ( y ) y y ( y ) + 2 x dyy ( y ) yy ( y ) + x dyy ( y )y ( y )

* 2 * 2 *
0 0 1 0 0 1 0 0
- - -
dyy ( y ) yy ( y )
*
The integral 0 0 is the position expectation value in the ground state
-
of the QHO, which is zero.
Also, since the ground-state wave function is normalized, the integral
dyy ( y )y ( y ) is 1.
*
0 0
-
Thus

dxy ( x - x ) x y ( x - x ) = dyy ( y ) y y ( y ) + x
* 2 * 2 2
0 1 0 1 0 0 1 (13)
- -
Substituting the integrals (12) and (13) into (11), we obtain

2
d
-ih dy
y y + 1 mw 2 dyy * y y 2y y + x 2 =

E 0
= dyy 0* ( y ) 0( ) 0 ( ) 0( ) 1
-
2m 2 -
2
d
-ih dy
1 1
= dyy 0 ( y )
*
y 0 ( y ) + dyy 0* ( y ) mw 2 y 2y 0 ( y ) + mw 2 x12 =
- 2m - 2 2
12
Preliminaries
d
2

- i h

dy 1 1
= dyy 0* ( y ) + mw 2 y 2 y 0 ( y ) + mw 2 x12 =
2m 2 2
-

1
= dyy 0* ( y ) H ( y )y 0 ( y ) + mw 2 x12
- 2

dyy ( y ) H ( y )y ( y )
*
The integral 0 0 is the energy expectation value of the QHO in
-
hw
the ground state, which, obviously, is equal to the ground-state energy, i.e. .
2
Thus
hw 1
E 0
= + mw 2 x12 (14)
2 2

ipx
We see that the action of the spatial translation operator exp - 1 on the ground
h
state of the QHO increases the energy expectation value by the positive amount
1
mw 2 x12 , which is the energy of a classical harmonic oscillator with amplitude x1 .
2
By means of (14), (9) is written as
hw 1
E t= + mw 2 x12 (15)
2 2
This is the energy expectation value of the QHO at time t 0 .
More on spatial translation operators

ipx
2) i) If Tx1 exp - 1 is a spatial translation operator, show that
h
x, Tx = x1Tx .
1 1
ii) Using the previous commutator, show that the action of Tx1 on a position
eigenstate yields a position eigenstate with eigenvalue translated by x1 , i.e.
T x = x + x .
x1 1
iii) Using the Baker-Campbell-Hausdorff formula

( ) ( ) 1
2
1
3!
1
exp A B exp - A = B + A , B + A , A , B + A , A , A , B + ... + A , A ,... A , B ...
n ! 144 424443
n A ' s or n commutators
show that, for an infinitely many times differentiable function f ,
Tx1 f ( x ) Tx1 = f ( x + x1 ) .
iv) A QHO is in the state Tx1 y , with y being an arbitrary state of the QHO.
Using the previous property of T , express the expectation values of the
x1
13
Preliminaries
position, momentum, and energy, as well as the position and momentum

uncertainties, and the position-momentum uncertainty product, in the
translated state Tx1 y in terms of the respective quantities in the untranslated
state y .
Solution
i) Using the Taylor expansion of Tx1 , the commutator x , Tx1 is written as
1
ipx
n
- n

x , Tx = x , exp - ipx1 = x , h = 1 x , - ipx1 =

1
h n =0 n ! n = 0 n ! h

n
ix1

-
h
= x, p n
n= 0 n !
ix1
In the last equality, we used that the quantity - is a number, and thus it commutes
h
with the momentum operator.
We remind that the argument x1 of the spatial translation operator Tx1 is a real
number, not an operator. You can think of it as being an eigenvalue of the position
operator.
Thus
n
ix1

-
x , Tx = h x, p n (1)
1
n =0 n!
We need to calculate the commutator x , p n .

The result of a commutator is independent of the representation we choose to
calculate it.
Since the momentum operator is to the nth power, we choose to calculate the
commutator in the momentum representation, where the momentum operator is a
scalar variable, and thus we can handle the exponential much easier than if it was an
operator.
d
We remind that in the momentum representation, x = ih and p = p .
dp
Choosing an arbitrary wave function f ( p ) (in the momentum representation), the
action of x , p n on f ( p ) yields
d n d n d d n df ( p )
ih dp , p f ( p ) = ih dp p - p ih dp f ( p ) = ih dp ( p f ( p ) ) - p =
n n
dp

= ih ( p n ) f ( p ) + p nf ( p ) - p nf ( p ) = ihnp n -1f ( p )

14
Preliminaries
Here, the prime denotes differentiation with respect to p .
That is
d n
ih dp , p f ( p ) = ihnp f ( p )
n -1

Since the wave function f ( p ) is arbitrary,
d n n -1
ih dp , p = ihnp

or, in representation-free form,
x, p n = ihnp n-1 (2)
with n 1 . If n = 0 , the commutator (2) is zero.

Substituting into (1) yields
n n n -1
ix1 ix1 ix1 -ix1

-
-
- ih
x , Tx = h h h h p n -1 =

1
ihnp =
n -1
ihp =
n -1
n =1 n ! n =1 ( n - 1) ! n =1 ( )
n - 1 !
n -1 n -1 n n
ix1 ix1 ix1
ipx

- x1
-
-
- 1
h h h p n = x h =
= p n -1 = x1 p n -1 ={ x1 1
n =1 ( n - 1)! n =1 ( n - 1) ! n= n -1 n = 0 n ! n = 0 n !

ipx
= x1 exp - 1 = x1Tx1
h
That is
x, Tx = x1Tx (3)
1 1
ii) Using the previous commutator, we obtain

x1 - Tx1 x = x1Tx1 xT
xT x1 = Tx1 x + x1Tx1
x1 on a position eigenstate x yields

Thus, the action of the operator xT
( )
x1 x = Tx1 x + x1Tx1 x = Tx1 x x + x1Tx1 x = Tx1 x x + x1Tx1 x
xT
Since x is a number an eigenvalue of the position operator it commutes with Tx1 ,

and thus
x1 x = xTx1 x + x1Tx1 x = ( x + x1 ) Tx1 x
xT
That is
x1 x = ( x + x1 ) Tx1 x
xT
Thus, the state Tx1 x is a position eigenstate with eigenvalue x + x1 .

Therefore
15
Preliminaries
Tx1 x = A x + x1 ,
where A is a complex constant.
In the previous exercise, we showed that the operator Tx1 is unitary, and thus it
preserves the norms of the states on which it acts.
Therefore, although the position eigenstates are not bound, we can assume that
A =1
Omitting the physically unimportant constant phase of A , we end up to

Tx1 x = x + x1 (4)

ipx
iii) For A = 1 and B = f ( x ) , the Baker-Campbell-Hausdorff formula is written as
h

ipx
ipx
ipx 1 ipx
ipx

exp 1 f ( x ) exp - 1 = f ( x ) + 1 , f ( x ) + 1 , 1 , f ( x ) +
h h h 2 h h
1 ipx
ipx
ipx 1 ipx
ipx
ipx
+ 1 , 1 , 1 , f ( x ) + ... + 1 , 1 ,... 1 , f ( x ) ... (5)
3! h h h n! h h h
We need to calculate the commutator p , f ( x ) . Since the function f is unknown, it

is suitable to do the calculation in the position representation, where the position
operator is a scalar variable, and thus the operator function f ( x ) becomes a scalar
function.
d
We remind that in the position representation, x = x and p = -ih .
dx
Choosing an arbitrary wave function f ( x ) (in the position representation), the action
of p , f ( x ) on f ( x ) yields
d d d
-ih dx , f ( x ) f ( x ) = -ih dx f ( x ) - f ( x ) -ih dx f ( x ) =

d df ( x )
= -ih ( f ( x ) f ( x ) ) - f ( x ) = -ih ( f ( x ) f ( x ) + f ( x ) f ( x ) - f ( x ) f ( x ) ) =
dx dx
= -ihf ( x ) f ( x )
Here, the prime denotes differentiation with respect to x .
That is
d
-ih dx , f ( x ) f ( x ) = -ihf ( x ) f ( x )
Since the wave function f ( x ) is arbitrary,
d
-ih dx , f ( x ) = -ihf ( x )
16
Preliminaries

p , f ( x ) = -ihf ( x ) (6)
Using (6), we have

1
ipx ix1 ix1
h , f (
x ) =
h
p , f (
x )
=
h
( -ihf ( x ) ) = x1 f ( x )
ipx
1 ipx
1 ipx
1 i 2 i 2
h , h , f ( x ) = h , x1 f ( x ) = h x1 p , f ( x ) = h x1 ( -ihf ( x ) ) = x1 f ( x )
2
Assuming that
ipx
1 ipx

ipx
, 1 ,... 1 , f ( x ) ... = x1k f ( k ) ( x )
h h
1444442444443 h
k commutators
then

ipx
1 ipx
1 ipx 1
ipx
ipx
ipx ipx

, ,... , f ( x ) ... = 1 , 1 ,... 1 , f ( x ) ... = 1 , x1k f ( k ) ( x ) =
h h
1444442444443 h h 1444 h 4h24444 3 h
( k +1) commutators
k commutators

i
h
k i
h
(k +1
)
= x1k +1 p , f ( ) ( x ) = x1k +1 -ihf ( ) ( x ) = x1k +1 f ( ) ( x )
k +1
Thus
ipx
1 ipx

ipx
, 1 ,... 1 , f ( x ) ... = x1n f ( n ) ( x ) ,
h h
1444442444443 h
n commutators
for every n = 1, 2,...

Thus, the right-hand side of (5) is written as
1 2 1 1
f ( x ) + x1 f ( x ) + x1 f ( x ) + x13 f ( ) ( x ) + ... + x1n f ( ) ( x ) =
3 n
2 3! n!
f ( x ) 2 f ( x ) 3
( 3)
f ( ) ( x ) n f ( ) ( x ) n
n n
= f ( x ) + f ( x ) x1 + x1 + x1 + ... + x1 = x1
2 3! n! n =0 n!

f(
n)
( x ) x n
The series n= 0 n!
1 is the Taylor expansion of f ( x + x1 ) about x , i.e.

f(
n)
( x ) x n =
n =0 n!
1 f ( x + x1 )
Thus, the right-hand side of (5) is equal to f ( x + x1 ) .

Besides, in the previous exercise, we showed that
17
Preliminaries

ipx
ipx
exp 1 = exp - 1
h h
Thus, the left-hand side of (5) is written as

1
ipx 1
ipx
exp - h f ( x ) exp - h = Tx1 f ( x ) Tx1

Therefore, (5) is written as
Tx1 f ( x ) Tx1 = f ( x + x1 ) (7)
iv) Denoting by O and O , respectively, the expectation values of an

y Tx1 y
operator O in the untranslated state y and in the translated state Tx1 y , we have
x Tx1 y (
= Tx1 y , xT
x1 y )=(y , Tx1 xT
x1 y ),
where, for more clarity, we use the general notation for inner products, since the
spatial translation operators are non-Hermitian.
Using (7), we have
Tx1 xT
x1 = x + x1 ,
since, in this case, f ( x ) = x .

Thus, the position expectation value in the translated state is written as
x Tx1 y
= ( y , ( x + x1 ) y ) = (14
y , x y ) + ( y , x1 y )=
243
x y
= x y
+ x1 ( y , y ) = x y
+ x1
1424 3
1
We assume that the initial, untranslated state y is normalized.
That is
x Tx1 y
= x y
+ x1 (8)
To calculate the position expectation value (8), we didnt use that the state y is a
QHO state, we only used that it is bound, so that its norm is 1.
Thus, the relation (8) holds for every bound, and normalized, state of any one-
dimensional quantum system.
The momentum expectation value in the translated state is written as
p Tx1 y (
= Tx1 y , pT
x1 y ) =(y , Tx1 pT
x1 y )
Since the spatial translation operators depend only on the momentum operator, they
commute with the momentum operator, and with every function of the momentum
operator, i.e. if g ( p ) is a function of the momentum operator, then
18
Preliminaries
g ( p ) , Tx = 0 (9)
1
Thus
Tx1 pT
x1 = Tx1 Tx1 p = p
{
1
Then, the momentum expectation value in the translated state is written as

p Tx1 y
= ( y , p y )= p y
That is
p Tx1 y
= p y
(10)
As in the case of the relation (8), the relation (10) also holds for every bound, and
normalized, state of any one-dimensional quantum system.
The energy expectation value in the translated state is written as
E Tx1 y (
= Tx1 y , HT
y
x1 )=(y , Tx1 HT
y
x1 )
The Hamiltonian of the QHO is
p 2 1
H = + mw 2 x 2
2m 2
Thus, we have
= T p + 1 mw 2 x 2 T = 1 T p 2T + 1 mw 2T x 2T
2
Tx1 HT x1 x1 x1 x x1 x1 x1
2m 2 2m 1 2
Using (7) and that the momentum operator squared commutes with Tx1 , we obtain
1 2 1 p 2 1
2 2
+ mw 2 ( x + x1 ) =
2
Tx1 HTx1 = Tx1 Tx1 p + mw Tx1 x Tx1 =

2m { 2 1
424 3 2m 2
1
( x + x1 )2
p 2 1 p 2 1 1
+ mw 2 ( x 2 + 2 x1 x + x12 ) =
2
= + mw 2 x 2 + mw 2 x1 x + mw 2 x12 =
2m 2 2m 2 2
1
= H + mw 2 x1 x + mw 2 x12
2
That is
= H + mw 2 x x + 1 mw 2 x 2
Tx1 HT x1 1 1
2
Then, the energy expectation value in the translated state is written as

E Tx1 y
=y

1
2

14243
( 14243
)
, H + mw 2 x1 x + mw 2 x12 y = y , H y + mw 2 x1 ( y , x y ) +
E x y
y
1 1
+ mw 2 x12 ( y , y ) = E y
+ mw 2 x1 x y
+ mw 2 x12
2 1424 3 2
1
19
Preliminaries
That is
1
E Tx1 y
= E y
+ mw 2 x1 x y
+ mw 2 x12 (11)
2
The relation (11) holds for every state of the QHO.
If the untranslated state y is an energy eigenstate n , then x y
= 0 and
1
E y
= En = n + hw . Then, (11) becomes
2
1 1
E Tx1 y
= n + hw + mw 2 x12 (12)
2 2
For n = 0 (ground state), (12) gives (14) or (15) of the previous exercise, as it should
do.
The position uncertainty in the translated state is
( )
2
( Dx )T x1 y
= x2 Tx1 y
- x Tx1 y
(13)
The expectation value of the position operator squared in the translated state is
x2 Tx1 y
(
= Tx1 y , x 2Tx1 y )=(y , Tx1 x 2Tx1 y )
Using (7), we obtain
Tx1 x 2Tx1 = ( x + x1 ) = x 2 + 2 x1 x + x12

2
Thus
x2 Tx1 y
(
= y , ( x 2 + 2 x1 x + x12 ) y )=(y , x 2 y ) + 2x ( y
1 , x y )+ x (y
1
2
,y )=
= x2 + 2 x1 x y
+ x12
y
That is
x2 = x2 + 2 x1 x y
+ x12 (14)
Tx1 y y
Substituting (14) and (8) into (13), we obtain
( )
2
( Dx )T x1 y
= x2 y
+ 2 x1 x y
+ x12 - x y
+ x1 =

( ) + 2x
2
= x2 + 2 x1 x y
+ x12 - x y 1 x y
+ x12 =
y

( )
2
= x2 y
- x y
= ( Dx ) y
That is
( Dx )T x1 y
= ( Dx ) y (15)
20
Preliminaries
Thus, the position uncertainty does not change, and this holds for every bound, and
Similarly, the momentum uncertainty in the translated state is
( )
2
( Dp )T x1 y
= p2 Tx1 y
- p Tx1 y
(16)
The expectation value of the momentum operator squared in the translated state is
p2 Tx1 y
(
= Tx1 y , p 2Tx1 y )=(y , Tx1 p 2Tx1 y )
Since the momentum operator squared commutes with Tx1 ,
Tx1 p 2Tx1 = Tx1 Tx1 p 2 = p 2

{
1
Thus
p2 = ( y , p 2 y )= p2
Tx1 y y
That is
p2 = p2 (17)
Tx1 y y
Substituting (17) and (10) into (16), we obtain

( Dp )T x1 y
= ( Dp ) y (18)
Thus, the momentum uncertainty does not change, and this also holds for every
bound, and normalized, state of any one-dimensional quantum system.
From (15) and (18), we derive that the position-momentum uncertainty product is the
same in the translated and untranslated state, and this also holds for every bound, and
Momentum translation operators
3) Similar to the case of spatial translation operators, a momentum translation

ip x
operator Tp1 is defined as Tp1 exp 1 , where x is the position operator
h
and p1 is an eigenvalue of the momentum operator.
i) Show that the operator T is unitary. p1
ii) Show that p , Tp1 = p1Tp1 .

iii) Show that the action of Tp1 on a momentum eigenstate yields a momentum
eigenstate with eigenvalue translated by p , i.e. T p = p + p . 1 p1 1
iv) Using the Baker-Campbell-Hausdorff formula show that, for an infinitely

many times differentiable function f , Tp1 f ( p ) Tp1 = f ( p + p1 ) .
v) A QHO is in the state T y , with y being an arbitrary state of the QHO.
p1
Using the previous property of Tp1 , express the expectation values of the
21
Preliminaries
position, momentum, and energy, as well as the position and momentum

uncertainties, and the position-momentum uncertainty product, in the
translated state Tp1 y in terms of the respective quantities in the untranslated
state y .
Solution
ip x
i) Using the Taylor expansion of exp 1 , the Hermitian conjugate of Tp1 is written
h
as

ip1 x n

ip1 x n ip1 x n

h

ip1 x h h
Tp1 = exp

=
h n =0 n !
=
n =0 n ! n =0
=
n!

In the same way we proved the relation (4) of the exercise 1, we prove that
n
ip1 x n ip1 x
=
h h
Then, using also that the position operator is Hermitian, we obtain

n
ip1 x n
ip1 x

h

-
= = h = exp - ip1 x
Tp1
n =0 n!

n =0 n!

h

That is
ip x
Tp1 = exp - 1 (1)
h
Then, the product Tp1 Tp1 is written as
ip x ip x
Tp1 Tp1 = exp - 1 exp 1
h h
ip1 x ip x
Since the commutator of - and 1 is zero,
h h
ip x ip x ip x ip x
exp - 1 exp 1 = exp - 1 + 1 = exp 0 = 1
h h h h
That is
Tp1 Tp1 = 1
In the same way, we have
ip x ip x ip x ip x
Tp1Tp1 = exp 1 exp - 1 = exp 1 - 1 = 1
h h h h
22
Preliminaries
Thus
Tp1 Tp1 = Tp1Tp1 = 1
Therefore, Tp1 is unitary.

ip x
ii) Using again the Taylor expansion of exp 1 , the commutator p , Tp1 is
h
written as
ip1 x
n
ip1 x n

n
p , Tp = p , h = p , h = 1 p , ip1 x
1
n =0 n ! n = 0 n ! n =0 n ! h

ip1
Using that the position operator commutes with the number (remember that p1 is
h
a number), we obtain
n
p , Tp = 1 p , x n
1 ip
1
n= 0 n ! h
In the previous exercise, we showed that the commutator of the momentum operator
with a differentiable function f of the position operator is
p , f ( x ) = -ihf ( x )
Thus
p , x n = -ihnx n-1
with n 1 . If n = 0 , the commutator is zero.

Then, the commutator p , Tp1 is written as
n n -1
p , Tp = 1 1 ( -ihnx n -1 ) = 1 1 ip1
ip ip
h ( -ih ) x =
n -1
1
n =1 n ! h n =1 ( n - 1) ! h
n -1 n
ip1 x ip1 x
n -1
1 ip1 h h =
= p1 x n -1 = p1 { 1
= p
n =1 ( n - 1) ! h n =1 ( n - 1) ! n= n -1 n = 0 n !
ip x
= p1 exp 1 = p1Tp1
h
That is
p , Tp = p1Tp (2)
1 1
iii) Using the previous commutator, we obtain

p1 - Tp1 p = p1Tp1 pT
pT p1 = Tp1 p + p1Tp1
23
Preliminaries
p1 on a momentum eigenstate p yields

Thus, the action of pT
(
p1 p = Tp1 p + p1Tp1
pT )p = Tp1 p p + p1Tp1 p = Tp1 p p + p1Tp1 p
where p is the eigenvalue of the momentum eigenstate p , i.e. it is a number, and

thus it commutes with T . p1
Then
p1 p = pTp1 p + p1Tp1 p = ( p + p1 ) Tp1 p
pT
That is
p1 p = ( p + p1 ) Tp1 p
pT
The state Tp1 p is thus a momentum eigenstate with eigenvalue p + p1 .

Therefore
Tp1 p = B p + p1 ,
where B is a complex constant.
Since Tp1 is a unitary operator, it preserves the norms of the states on which it acts.
Therefore, although the momentum eigenstates are not bound, we can assume that
B =1
Then, omitting the physically unimportant constant phase of B , we obtain

Tp1 p = p + p1 (3)
ip x
iv) For A = - 1 and B = f ( p ) , the Baker-Campbell-Hausdorff formula,
h
( ) ( ) 1
2
1
3!
1
exp A B exp - A = B + A , B + A , A , B + A , A , A , B + ... + A , A ,... A , B ... ,
n ! 144 424443
n A ' s or n commutators
is written as
ip x ip x ip x 1 ip x ip x
exp - 1 f ( p ) exp 1 = f ( p ) + - 1 , f ( p ) + - 1 , - 1 , f ( p ) +
h h h 2 h h
1 ip x ip x ip x 1 ip x ip x ip x
+ - 1 , - 1 , - 1 , f ( p ) + ... + - 1 , - 1 ,... - 1 , f ( p ) ... (4)
3! h h h n! h h h
Well use the momentum representation to calculate the commutator x, f ( p ) , since

d
in the momentum representation, x = ih and p = p , and thus the operator function
dp
f ( p ) becomes a usual, and easy-to-handle, scalar function.
24
Preliminaries
Thus, choosing an arbitrary wave function f ( p ) (in the momentum representation),

we have
d d d d df ( p )
ih dp , f ( p ) f ( p ) = ih dp f ( p ) - f ( p ) ih dp f ( p ) = ih dp ( f ( p ) f ( p ) ) - f ( p ) dp =

= i h ( f ( p ) f ( p ) + f ( p ) f ( p ) - f ( p ) f ( p ) ) = i hf ( p ) f ( p )
Here, the prime denotes differentiation with respect to p .
That is
d
ih dp , f ( p ) f ( p ) = ihf ( p ) f ( p )

Since the wave function f ( p ) is arbitrary,
d
ih dp , f ( p ) = ihf ( p )

x, f ( p ) = ihf ( p ) (5)
Using (5), we have

ip1 x ip1 ip1
- h , f ( p ) = - h x, f ( p ) = - h ihf ( p ) = p1 f ( p )
ip1 x ip1 x ip1 x ip12

-
h h , - , f (
p ) =
h - , p f (
p ) = - x, f ( p ) =
h
1

ip 2
= - 1 ihf ( p ) = p12 f ( p )
h
Assuming that
ip1 x ip1 x ip1 x
- , - ,... - , f ( p ) ... = p1k f ( k ) ( p )
h h
1444444 h
424444444 3
k commutators
then

ip1 x ip1 x ip1 x ip x ip x ip x
- , - ,... - , f ( p ) ... = - 1 , - 1 ,... - 1 , f ( p ) ... =
h h
1444444 h h 14444 h h
4244444 3
424444444 3
( k +1) commutators k commutators
k +1 k +1
ip x ip ip
= - 1 , p1k f ( k ) ( p ) = - 1 x, f ( k ) ( p ) = - 1 ihf ( k +1) ( p ) = p1k +1 f ( k +1) ( p )
h h h
Thus
25
Preliminaries
ip1 x ip1 x ip1 x

- , - ,... - , f ( p ) ... = p1n f ( n ) ( p ) ,
h h
1444444 h
424444444 3
n commutators
for every n = 1, 2,...

Then, the right-hand side of (4) becomes
1 2 1 1
f ( p ) + p1 f ( p ) + p1 f ( p ) + p13 f ( ) ( p ) + ... + p1n f ( ) ( p ) =
3 n
2 3! n!
f ( p ) 2 f ( p ) 3
(3)
f ( ) ( p ) n f ( ) ( p ) n
n n
= f ( p ) + f ( p ) p1 + p1 + p1 + ... + p1 = p1
2 3! n! n =0 n!

f(
n)
( p ) p n
The series
n =0 n!
1 is the Taylor expansion of f ( p + p1 ) about p , i.e.

f(
n)
( p ) p n =
n= 0 n!
1 f ( p + p1 )
Thus, the right-hand side of (4) is equal to f ( p + p1 ) .

ip x ip x
Using (1), the left-hand side of (4), i.e. the term exp - 1 f ( p ) exp 1 , is
h h
written as Tp1 f ( p ) Tp1 .

Therefore, (4) is written as

Tp1 f ( p ) Tp1 = f ( p + p1 ) (6)
v) Using again, for more clarity, the general notation for inner products, the position
expectation value in the translated state Tp1 y is written as
x Tp1 y (
= Tp1 y , xT
p1 y ) =(y , Tp1 xT
p1 y )
The momentum translation operators depend only on the position operator, and thus
they commute with the position operator, and with every function of the position
operator, i.e. if g ( x ) is a function of the position operator, then
g ( x ) , Tp = 0 (7)
1
Thus
Tp1 xT
p1 = Tp1 Tp1 x = x
{
1
Then, the position expectation value in the translated state is written as

x Tp1 y
= ( y , x y )= x y
That is
x Tp1 y
= x y
(8)
26
Preliminaries
Since we didnt make use of the fact that the state y is a QHO state, the relation (8)
holds for every bound, and normalized, state of any one-dimensional quantum system.
The momentum expectation value in the translated state is written as
p Tp1 y (
= Tp1 y , pT
p1 y )=(y , Tp1 pT
p1 y )
Tp1 pT
p1 = p + p1
Thus
p Tp1 y
= ( y , ( p + p1 ) y )=(y , p y )+ p (y1 ,y )= p y
+ p1
That is
p Tp1 y
= p y
+ p1 (9)
The relation (9) also holds for every bound, and normalized, state of any one-
As we did in the case of the spatial translation see the previous exercise we can
directly calculate the energy expectation value in the translated state. Alternatively,
we can calculate first the expectation values of the position squared and momentum
squared, since well need them to calculate the respective uncertainties, and from
them well derive the energy expectation value.
The expectation value of the position squared in the translated state is written as
x2 Tp1 y
(
= Tp1 y , x 2Tp1 y )=(y , Tp1 x 2Tp1 y )
Tp1 x 2Tp1 = Tp1 Tp1 x 2 = x 2
{
1
Thus
x2 = ( y , x 2 y )= x2
Tp1 y y
That is
x2 = x2 (10)
Tp1 y y
The expectation value of the momentum squared in the translated state is written as
p2 Tp1 y
(
= Tp1 y , p 2Tp1 y )=(y , Tp1 p 2Tp1 y )
Tp1 p 2Tp1 = ( p + p1 ) = p 2 + 2 p1 p + p12

2
Thus
27
Preliminaries
p2 Tp1 y
(
= y , ( p 2 + 2 p1 p + p12 ) y )=(y , p 2 y )+ 2p ( y
1 , p y ) + p (1y424
1
2
,y )=
3
1
= p 2
+ 2 p1 p y
+ p1 2
y
We remind that weve assumed that the bound, and thus normalizable, state y is
normalized.
That is
p2 = p2 + 2 p1 p y
+ p12 (11)
Tp1 y y
Using the Hamiltonian of the QHO, the energy expectation value of the QHO in the
translated state is written as
p2
p 2 1 Tp1 y 1
E Tp1 y
= H = + mw 2 x 2 = + mw 2 x 2 Tp1 y
Tp1 y 2m 2 Tp1 y
2m 2
Substituting (10) and (11) yields

p2 + 2 p1 p y
+ p12 1 p2 1 p
y y
E Tp1 y
= + mw 2 x 2 y
= + mw 2 x 2 y + 1 p y
+
2m 2 2m 4224444
1444 3 m
E y
2 2
p p1 p
+ 1
= E y
+ p y
+ 1
2m m 2m
That is
p1 p12
E Tp1 y
= E y
+ p y
+ (12)
m 2m
Obviously, the relation (12) holds only for the states of the QHO.
If the untranslated state y is an energy eigenstate n , then p y
= 0 and
1
E y
= En = n + hw . Then, (12) becomes
2
1 p2
E Tp1 n
= n + hw + 1 (13)
2 2m
Using (8) and (10), we derive that the position uncertainty in the translated state is the
same as in the untranslated state, i.e.
( Dx )T p1 y
= ( Dx ) y (14)
Using (9) and (11), the momentum uncertainty in the translated state is written as
28
Preliminaries
( ) ( )
2 2
( Dp )T
p1 y
= p2 Tp1 y
- p Tp1 y
= p2 y
+ 2 p1 p y
+ p12 - p y
+ p1 =

( p ) + 2p
2
= p2 + 2 p1 p y
+ p12 - y 1 p y
+ p12 =
y

( )
2
= p2 y
- p y
= ( Dp ) y
That is
( Dp )T
p1 y
= ( Dp ) y (15)
The momentum uncertainty in the translated state is the same as the momentum
uncertainty in the untranslated state.
From (14) and (15), we derive that the position-momentum uncertainty product in the
translated state is the same as the position-momentum uncertainty product in the
untranslated state, and this also holds for every bound, and normalized, state of any
one-dimensional quantum system.
The combined action of a spatial and a momentum translation

operator

ipx
4) Let Tx1 be a spatial translation operator, i.e. Tx1 exp - 1 , and Tp1 be a
h
ip x
momentum translation operator, i.e. Tp1 exp 1 .
h
i ( p1 x - px
1)
We also define the operator Tp1 , x1 exp .
h
i ( p1 x - px
1) i ( - px
1 + p1 x )
Since exp = exp , the order of the two operators
h h

doesnt matter, and Tp1 , x1 = Tx1 , p1 .
Well show that the three operators Tx1Tp1 , Tp1Tx1 , and Tp1 , x1 differ only by a
constant phase, and thus they are physically equivalent.
This means that their action on a physical state not necessarily a QHO state
yields states that differ only by a constant phase, and thus they are
physically equivalent, i.e. they are the same state.
Solution
The physical equivalence of the three operators follows from the fact that the
ip x ipx

commutator 1 , - 1 is a constant.
h h
Indeed
ip1 x ipx 1 ip1 ix1 px ix p
h , - = - [ x , p ] = 1 2 1 ih = 1 1
h h h h h
29
Preliminaries
That is
ip1 x ipx 1 ix1 p1
h , - = (1)
h h
Since the previous commutator is a constant, we can use the identity
( ) ( ) ( ) 1
2

exp A + B = exp A exp B exp - A , B

which holds if the operators A and B commute with their commutator A , B .

ip x ipx

In our case, the commutator 1 , - 1 is a constant, and thus it commutes with
h h
ip x
ipx
both 1 and - 1 .
h h
Then, using the previous identity and the commutator (1), we obtain
i ( p1 x - px
1) ip1 x ipx
ip x
ipx 1 ix1 p1
exp = exp - 1 = exp 1 exp - 1 exp - =
h h h h h 2 h
ip x ipx ix p
= exp 1 exp - 1 exp - 1 1
h h 2h
That is
ix p
Tp1 , x1 = Tp1Tx1 exp - 1 1
2h
ix p
The exponential exp - 1 1 is a constant complex number, and thus it commutes
2h
with both Tx1 and Tp1 .
Thus
ix p
Tp1 , x1 = exp - 1 1 Tp1Tx1 (2)
2h
Therefore, the operators Tp1Tx1 and Tp1 , x1 differ only by a constant phase.
Besides, from (1) we obtain
ipx 1 ip1 x ip1 x ipx
1 ix1 p1
- h , h = - h , - h = - h ,
Thus, using again the previous identity, we obtain

ipx ip x
ipx ip x 1 ix p
exp - 1 + 1 = exp - 1 exp 1 exp - - 1 1 =
h h h h 2 h
ipx ip x ix p
= exp - 1 exp 1 exp 1 1
h h 2h
That is
30
Preliminaries

ipx ip x ix p
exp - 1 + 1 = Tx1Tp1 exp 1 1
h h 2h
ix p
The exponential exp 1 1 is a constant complex number, and thus it commutes
2h
with both Tx1 and Tp1 .
Thus

ipx ip x ix p
exp - 1 + 1 = exp 1 1 Tx1Tp1 (3)
h h 2h
Also

ipx ip x ip x ipx
i ( p1 x - px
1)
exp - 1 + 1 = exp 1 - 1 = exp = Tp1 , x1
h h h h h
Thus, (3) is written as
ix p
Tp1 , x1 = exp 1 1 Tx1Tp1 (4)
2h
Therefore, the operators Tx1Tp1 and Tp1 , x1 differ only by a constant phase.
Besides, comparing (2) and (4) yields
ix p ix p
exp - 1 1 Tp1Tx1 = exp 1 1 Tx1Tp1
2h 2h
ix p ix p ix p ix p
exp 1 1 exp - 1 1 Tp1Tx1 = exp 1 1 exp 1 1 Tx1Tp1
2h 2h 2h 2h
ix p
Tp1Tx1 = exp 1 1 Tx1Tp1
h
That is
ix p
Tp1Tx1 = exp 1 1 Tx1Tp1 (5)
h
Therefore, the operators Tx1Tp1 and Tp1Tx1 differ only by a constant phase.
This means that the spatial translations commute with the momentum
translations, or, in other words, they are independent.
Returning back to the QHO The displacement operator
i ( p1 x - px
1)
5) For the QHO, express the operator Tp1 , x1 exp in terms of the
h
ladder operators and show that it is written as exp ( l a - l *a ) , with
31
Preliminaries
1 x1 p1 h
l= + i , where x0 = is the length scale and p0 = mhw is the
2 x0 p0 mw
momentum scale of the QHO.
Solution
Solving the definition relations of the ladder operators for the position and momentum
operators, we obtain
h
x =
2mw
( a + a )
mhw
p = i
2
( a - a )
Introducing into the previous relations the length and momentum scales of the QHO,
we obtain
x0
x =
2
( a + a ) (1)

p0
p = i
2
( a
- a ) (2)
i ( p1 x - px
1)
By means of (1) and (2), the operator is written as
h
x p
i p1 0 ( a + a ) - i 0 ( a - a ) x1
i ( p1 x - px1 )

h
=
2
h
2
(
= 1 ix p a + a + p x a - a =
2h
0 1( ) 0 1( ) )
1 1
=
2h
( ix0 p1a + ix0 p1a + p0 x1a - p0 x1a ) =
2h
( ( p0 x1 + ix0 p1 ) a - ( p0 x1 - ix0 p1 ) a )
That is
i ( p1 x - px
1) 1
h
=
2h
( ( p0 x1 + ix0 p1 ) a - ( p0 x1 - ix0 p1 ) a )
Using that x0 p0 = h , the previous equation is written as
i ( p1 x - px
1) 1
h
=
2 x0 p0
( ( p0 x1 + ix0 p1 ) a - ( p0 x1 - ix0 p1 ) a ) =
1 x1 p1 x1 p1 1 x1 p1 1 x1 p1
= + i a - - i a = + i a - - i a =
2 x0 p0 x0 p0 2 x0 p0 2 x0 p0
1 x p1 1 x1 p1
*
=
1
+ i
a -
2x + i a
2 x0 p0 p0
0

That is
32
Preliminaries
1 ) 1 x1
*
i ( p1 x - px p 1 x1 p
= + i a -
1
+ i a
1
h 2 x0 p0 2 x0 p0

Setting
1 x1 p1
l= + i (3)
2 x0 p0
we end up to
i ( p1 x - px
1)
= ( l a - l *a ) (4)
h
Thus, the operator Tp1 , x1 is written as
Tp1 , x1 = exp ( l a - l *a ) (5)
In (5), the operator Tp1 , x1 is written in terms of one, but complex, parameter, the
parameter l , instead of the two real parameters x1 and p1 (the spatial and
momentum translations).
Since the parameters x1 and p1 can be any real number, from (3) we see that the
parameter l can be any complex number.
The operator exp ( l a - l *a ) is called the displacement operator and it is usually
denoted by D ( l ) , i.e.
D ( l ) = exp ( l a - l *a ) (6)
The parameter l is called the displacement parameter.

As well see below, for each value of the displacement parameter l , the displacement
operator, acting on the ground state of the QHO, yields an eigenstate of the
annihilation operator, a so-called coherent state.
From (5) and (6), we have
Tp1 , x1 = D ( l ) (7)
i ( p1 x - px
1)
where Tp1 , x1 exp .
h

That is, the operator Tp1 , x1 is the displacement operator.
Besides, in the previous exercise, we showed that the operator Tp1 , x1 differs from the
operators Tx1Tp1 and Tp1Tx1 only by a, physically unimportant, constant phase. Thus,
the action of the displacement operator is physically the same as the combined action
of a spatial and a momentum translation operator, or a momentum and a spatial
translation operator. In other words, the displacement operator, acting on an arbitrary
state of the QHO, yields a spatial and a momentum translation, or a momentum and a
spatial translation.
33
Preliminaries
i ( p1 x - px
1)
6) At t = 0 , a QHO is in the state Tp1 , x1 0 , where Tp1 , x1 exp .
h
i) Show that the state Tp1 , x1 0 is normalized.
ii) Expand the state T 0 in the basis of the energy eigenstates of the QHO.
p1 , x1
iii) Write the time evolution of the state Tp1 , x1 0 for t > 0 .
iv) Show that the probability that the QHO is found in an energy eigenstate at
time t 0 is given by a Poisson distribution. What is the parameter of the
distribution?
Solution
i) We showed in the exercise 4 that the operator Tp1 , x1 is written as
ix p
Tp1 , x1 = exp - 1 1 Tp1Tx1 (1)
2h
The Hermitian conjugate of Tp1 , x1 is then

ix p ix p
( ) ix p

Tp1 , x1 = exp - 1 1 Tp1Tx1 = exp 1 1 Tp1Tx1 = exp 1 1 Tx1 Tp1
2h 2h 2h
That is
ix p
Tp1 , x1 = exp 1 1 Tx1 Tp1 (2)
2h
Using (1) and (2), and that the operators Tx1 and Tp1 are unitary, we have
ix p ix p
Tp1 , x1 Tp1 , x1 = exp 1 1 Tx1 Tp1 exp - 1 1 Tp1Tx1 =
2h 2h
ix p ix p ix p ix p
= exp 1 1 exp - 1 1 Tx1 Tp1 Tp1 Tx1 = exp 1 1 - 1 1 Tx1 Tx1 = 1
2h 2h { 4 2h3 {
2h2444
1 144 1
exp0 =1
Also
ix p ix p
Tp1 , x1Tp1 , x1 = exp - 1 1 Tp1Tx1 exp 1 1 Tx1 Tp1 =
2h 2h
ix p ix p ix p ix p
= exp - 1 1 exp 1 1 Tp1 Tx1 Tx1 Tp1 = exp - 1 1 + 1 1 Tp1 Tp1 = 1
2h 2h { 1 1444 22444 h 2h3 {
1
exp0 =1
That is
Tp1 , x1 Tp1 , x1 = Tp1 , x1Tp1 , x1 = 1
Therefore, the operator Tp1 , x1 is unitary.
34
Preliminaries
ix1 p1 ix1 p1
We remind that the terms exp - and exp , as constant complex
2h 2h
numbers, commute with any of the above operators.
Since the operator Tp1 , x1 is unitary, it preserves the norms of the states on which it
acts, and thus
Tp1 , x1 0 = 0 = 1
Therefore, the state Tp1 , x1 0 is normalized.

ii) Using the completeness relation of the energy eigenstates, i.e. n
n =0
n = 1 , the
state Tp1 , x1 0 is written as

Tp1 , x1 0 = n n Tp1 , x1 0 = n n Tp1 , x1 0 = n Tp1 , x1 0 n
n =0 n =0 n =0
That is

Tp1 , x1 0 = n Tp1 , x1 0 n (3)
n= 0
This is the expansion of the state Tp1 , x1 0 in the energy basis of the QHO, but we
have to calculate the amplitude n T 0 .p1 , x1
To do that, we can use that see the previous exercise

Tp1 , x1 = exp ( l a - l *a )
1 x1 p1
where l = +i .
2 x0 p0
We observe that

l a , -l a = l ( -l ) a , a = - l - a , a = l

* * 2 2
123
1
That is
2
l a , -l *a = l (4)
Since the previous commutator is a constant, using the identity
( ) ( ) ( )
1
2

exp A + B = exp A exp B exp - A , B ,

for A = l a and B = -l * a , we obtain
1 2
exp ( l a - l *a ) = exp ( l a ) exp ( -l * a ) exp - l
2
35
Preliminaries
1 2
The term exp - l is a constant, and thus we can move it to the left and write
2
1 2
exp ( l a - l *a ) = exp - l exp ( l a ) exp ( -l * a )
2
Thus, the operator Tp1 , x1 is written as
1 2
Tp1 , x1 = exp - l exp ( l a ) exp ( -l *a ) (5)
2
Then, to calculate the action of Tp1 , x1 on the ground state, we must first calculate the
action of exp ( -l *a ) on the ground state.
Using the Taylor expansion of exp ( -l *a ) , its action on the ground state is written as
( -l *a )m
( -l * a )
m

( -l * ) m
m
exp ( -l * a ) 0 = 0 = 0 = a 0
m= 0 m ! m =0 m! m= 0 m!

In the last equality, we used that a commutes with the constant number -l * .
That is
exp ( -l * a ) 0 =

( -l ) * m
a m 0
m =0 m!
But, since a kills the ground state,
a m 0 = 0 if m = 1, 2,...

( -l )
* m
Then, in the series

m= 0 m!
a m 0 only the first term, with m = 0 , survives.
Thus
exp ( -l *a ) 0 = 0 (6)
Using (5) and (6), the action of Tp1 , x1 on the ground state is written as
1 2
Tp1 , x1 0 = exp - l exp ( l a ) 0
2
Using the Taylor expansion of exp ( l a ) , we obtain
36
Preliminaries
( l a ) m ( l a )m
1 2 1 2
Tp1 , x1 0 = exp - l 0 = exp - l
0 =
2 m =0 m ! 2 m= 0 m !

1 2 l m
m
= exp - l a 0
2 m =0 m !
In the last equality, we used that a commutes with the constant number l .
That is
1 2 l m
m
Tp1 , x1 0 = exp - l a 0 (7)
2 m =0 m !
Using that a n = n + 1 n + 1 , we have
a m 0 = a m-1 1 = 2a m- 2 2 = 2!a m -2 2 = 2! 3a m-3 3 =

= 3!a m -3 3 = ... = m ! a{
m- m m = m ! m
a 0 =1
That is
a m 0 = m ! m (8)
1 2 l
m
Tp1 , x1 0 = exp - l
m (9)
2 m =0 m !
Using (9), the amplitude n Tp1 , x1 0 is written as
1 2 l
m

n Tp1 , x1 0 = n exp - l
m =
2 m =0 m !
1 2
l m
1 2 l
m
= exp - l n m = exp - l nm
2 m= 0 m ! 2 m =0 m !
That is
1 2 l
m
n Tp1 , x1 0 = exp - l n m (10)
2 m =0 m !
Using the orthonormality of the energy eigenstates, i.e. n m = d nm , the series

lm
m =0 m!
n m becomes

lm
lm ln
m= 0 m!
n m =
m =0 m!
d nm =
n!
Substituting into (10), we obtain
37
Preliminaries
1 2 l
n

n Tp1 , x1 0 = exp - l (11)
2 n!
Substituting the amplitude (11) into the expansion (3), we obtain
1 2 l
n
Tp1 , x1 0 = exp - l n (12)
2 n =0 n !
1 x1 p1
where l = +i .
2 x0 p0
This is the expansion of the state Tp1 , x1 0 in the energy basis of the QHO.
Since T = D ( l ) , (12) is also written as
p1 , x1
1 2 l
n
D ( l ) 0 = exp - l
n (13)
2 n= 0 n !
iii) Denoting by n t
the time evolution of an energy eigenstate n , then
iE t
n t = exp - n n ,
h
1
with t 0 and En = n + hw .
2
Then, using (12), the time evolution of the state Tp1 , x1 0 let us denote it by y ( t )
is
l 1 2 l
n n
1 iE t
y ( t ) = exp - l n t = exp - l
2
exp - n n
2 n =0 n ! 2 n =0 n ! h
That is
1 2 l
n
iE t
y (t ) = exp - l exp - n n (14)
2 n =0 n ! h
with t 0 and y ( 0 ) Tp1 , x1 0 .
If the initial state is taken at t = t0 , instead of t = 0 , then the time evolution of an

energy eigenstate n is
iE ( t - t0 )
n t = exp - n n ,
h
and the time evolution of the state Tp1 , x1 0 is then
1 l
n
iE ( t - t0 )
y ( t ) = exp - l
2
exp - n n
2 n =0 n ! h
with t t0 and y ( t0 ) Tp1 , x1 0 .
38
Preliminaries
iv) The state of the QHO at t 0 is given by (14).

The energy eigenstates are orthonormal, thus the probability amplitude that the QHO
is found in an energy eigenstate m , at time t 0 , is m y ( t ) , which, using (14), is
written as
1 2 l iE t
n
m y ( t ) = m exp - l exp - n n =
2 n =0 n ! h
1 2 ln iE t
= exp - l m exp - n n =
2 n= 0 n ! h
1 2 l 1 2 l
n m
iE t iE t
= exp - l exp - n m n = exp - l exp - m
2 n =0 n ! h { d
2 m! h
mn
That is
1 2 l
m
iE t
m y ( t ) = exp - m exp - l (15)
h 2 m!
Then, the probability that the QHO is found in an energy eigenstate m , at time
t 0 , is
2
1 2 l
m
2 iE t
Pm ( t ) = m y ( t ) = exp - m exp - l
h 2 m!
Using that for a complex number z , z = zz* , the probability Pm ( t ) becomes

2
*
1 2 l 1 2 l
m m
iE t iEm t
Pm ( t ) = exp - m exp - l exp -
h exp -
2 l =
h 2 m ! m!
1 2 (l )
m *
1 2 l
m
iE t iE t
= exp - m exp - l exp m exp - l
h 2 m! h 2 m!
Using that ( l m ) = ( l * ) , which follows from the property ( z1 z2 ...zm ) = z1* z2*...zm*
* m *
for z1 = z 2 = ... = zm = l , which, in turn, follows easily, by induction, from the

elementary property ( z1 z2 ) = z1* z2* , the probability Pm ( t ) becomes
*
1 2 l (l )
m * m
iE t iE t 1 2
Pm ( t ) = exp - m exp m exp - l exp - l =
h h 2 2 m ! m !
1444424444 3 14444 4244444 3
1
(
exp - l
2
)
( ll ) * m
(l )
2 m
(
= exp - l
2
) m!
(
= exp - l
2
) m!
That is
(l ) 2 m
(
Pm ( t ) = exp - l
2
) m!
(16)
39
Preliminaries
with m = 0,1,...
The probability (16) is time independent and it is given by a Poisson distribution with
2
parameter l .
Observe that
(l )
2 m
(l ) 2 m
P ( t ) = exp ( - l ) ( )

2 2
m = exp - l =1
m= 0 m= 0 m! m !3
1424
m =0
( )
exp l
2
That is, the probabilities (16) add up to 1, as they should.
1 x1 p1 2
Using that l = + i , we obtain the parameter l in terms of the spatial
2 x0 p0
and the momentum translation, x1 and p1 , respectively. Then, we have
1 x p
2 2
2
l = 1 + 1 (17)
2 x0 p0

40
The coherent states of the QHO
II. The coherent states of the QHO

7) The coherent states of the QHO are defined as the eigenstates of the
annihilation operator.
See, for instance, https://arxiv.org/pdf/quant-ph/9708012v1.pdf.
i) If l is an eigenstate of the annihilation operator, i.e. if a l = l l , with
l , show that D ( l ) 0 = l , i.e. the displacement operator, acting on the
ground state, generates the coherent states.
We remind that the eigenvalues of the annihilation operator are complex
numbers, because the annihilation operator is not Hermitian.
ii) Show that the coherent states are states of minimum position-momentum
uncertainty AND the two individual uncertainties, i.e. the position uncertainty
and the momentum uncertainty, are equally distributed, in the sense that
x p Dx Dp 1
Dx = 0 and Dp = 0 , or, in dimensionless form, = = , with x0 , p0
2 2 x0 p0 2
being, respectively, the length and momentum scales of the QHO. This
property, i.e. the equal distribution of the position uncertainty and the
momentum uncertainty, differentiates the coherent from the squeezed states,
which are also minimum position-momentum uncertainty states, but the two
individual uncertainties are NOT equally distributed.
iii) Calculate the energy expectation value and uncertainty in a coherent state.
iv) Using the expansion of the coherent state l in the energy basis of the
QHO, write its time evolution and show that although it remains a coherent
state, and thus a state of minimum position-momentum uncertainty, its
eigenvalue changes. What is the time evolution of the eigenvalue l ?
Solution
i) Well show that the states D ( l ) 0 and l have the same expansion in the energy
basis of the QHO, and thus they are the same state.
In the exercise 6, we showed that the expansion of the state D ( l ) 0 in the energy
basis of the QHO is
1 2 l
n
D ( l ) 0 = exp - l n (1)
2 n= 0 n !
Let us now find the expansion of the state l .
Using the completeness relation of the energy eigenstates, the state l is written as

l = n n l = n n l = n l n
n =0 n =0 n= 0
That is

l = n l n (2)
n= 0
Then, the action of the annihilation operator on the state l is written as
41

a l = a n l n = n l a n
n =0 n =0
Using that a n = n n - 1 , we obtain

a l = n l n n -1
n =1
Changing the summation index to n = n - 1 , we obtain

a l = n + 1 l n + 1 n
n= 0
Renaming the summation index to n , we end up to

a l = n + 1 l n + 1 n (3)
n= 0
Since the state l is eigenstate of a with eigenvalue l ,
a l = l l
Substituting (2) and (3) into the previous equation yields

n= 0
n +1 l n +1 n = l n l n n +1 n +1 l n = l n l n
n =0 n =0 n= 0
( )n

n +1 n +1 l - l n l =0
n =0
Since the energy eigenstates are linearly independent, from the last equation we
obtain
n +1 n +1 l - l n l = 0 n +1 n +1 l = l n l
Thus
l
n +1 l = n l (4)
n +1
with n = 0,1,...
Applying the recursive relation (4) repeatedly gives
l l l l l l
n +1 l = nl = n -1 l = n-2 l =
n +1 n +1 n n +1 n n -1
l l l l l l
= ... = ... n-n l = ... 0l =
n +1 n n -1 n - ( n - 1) n +1 1
l n+1
= 0l
( n + 1)!
That is
42
l n+1
n +1 l = 0l
( n + 1)!
Thus
ln
nl = 0 l (5)
n!
The constant 0 l can be calculated using that the state l is normalized, i.e.
l l = 1.
Substituting (5) into (2) yields

ln
ln
l = 0l n = 0l n
n =0 n! n =0 n!
That is

ln
l = 0l
n =0 n!
n (6)
Then, the bra l is

(l ) n *
(l )
* n

* *
l = 0l n = 0l n
n =0 n! n= 0 n!
where, in the last equality, we used that ( z1 z2 ...zn ) = z1* z2*...zn* , a complex number
*
property that follows easily, by induction, from the basic property ( z1 z2 ) = z1* z2* .
*
Thus, the bra l is

(l ) * n

*
l = 0l n (7)
n =0 n!
Using (6) and (7), we obtain

(l )
* m

ln
l l = 0 l m 0 l
*
n =
m =0 m! n =0 n!

(l )
* m
ln
(l )
* m
ln

* 2
= 0l 0l mn = 0l mn
m, n =0 m! n! m ,n = 0 m! n!
Since the two sums are independent, we use different summation indices.
Using the orthonormality of the energy eigenstates, i.e. m n = d mn , we obtain
43

(l ) * m
ln
(l )
* n
ln
( ll )
* n

2 2 2
l l = 0l d mn = 0 l = 0l =
m, n = 0 m! n! n =0 n! n! n =0 n!
(l ) 2 n
( )

2 2 2
= 0l = 0l exp l
n =0 n!
That is
l l = 0l
2
exp l ( ) 2
Since the state l is normalized,
1= 0 l
2
exp l( ) 2
0l
2
(
= exp - l
2
) 1 2
0 l = exp - l
2
Omitting the physically unimportant phase of 0 l , we end up to
1 2
0 l = exp - l (8)
2
Substituting (8) into (6) yields
1 2 l
n
l = exp - l n (9)
2 n =0 n !
Comparing (1) and (9), we obtain
D ( l ) 0 = l (10)
ii) In the exercise 4, we showed that the operator Tp1 , x1 is written as
ix p
Tp1 , x1 = exp - 1 1 Tp1Tx1
2h
ix p
Then, since the term exp - 1 1 is a constant phase, the action of Tp1 , x1 on a
2h
physical state is equivalent, i.e. it is physically the same, to the action of Tp1Tx1 .
Since the operator T is the displacement operator D ( l ) , as shown in the exercise
p1 , x1
5, the action of D ( l ) is physically equivalent to the action of Tp1Tx1 .

Then, omitting the physically unimportant constant phase, we can write
D ( l ) 0 = Tp1Tx1 0
By means of (10), the previous equation becomes
l = Tp Tx 0 (11)
1 1
1 x1 p1
with l = +i .
2 x0 p0
44
The equation (11) provides a nice intuitive picture of the coherent states, as it
tells us that they result from the application of spatial and momentum
translations to the ground state of the QHO.
We remind that, as also shown in the exercise 4, the order of the two translations
doesnt matter, i.e.
Tp1Tx1 0 = Tx1Tp1 0
In the ground state of the QHO, the position and momentum uncertainties are,
respectively,
x0
( Dx ) 0 = (12)
2
p0
( Dp ) 0 = (13)
2
The position-momentum uncertainty product is then
x0 p0
( DxDp ) 0 =
2
Using that x0 p0 = h , we obtain
h
( DxDp ) 0 = (14)
2
The ground state is then a state of minimum position-momentum uncertainty.
The ground state is also a coherent state, since it is eigenstate of the annihilation
operator with eigenvalue 0, i.e. a 0 = 0 0 .
In the exercise 2, we showed that the application of a spatial translation does not
change either the position or the momentum uncertainty.
Thus
( Dx )T
x1 0
= ( Dx ) 0
( Dp )T x1 0
= ( Dp ) 0
By means of (12) and (13), the previous two equations become, respectively,
x0
( Dx )T 0
= (15)
x1
2
p0
( Dp )T 0
= (16)
x1
2
Similarly, in the exercise 3, we showed that the application of a momentum
translation does not change either the position or the momentum uncertainty.
Thus, taking the state Tx1 0 as the initial state,
45
( Dx )T T
p1 x1 0
= ( Dx )T
x1 0
( Dp )T T
p1 x1 0
= ( Dp )T
x1 0
By means of (15) and (16), the previous two equations become, respectively,
x0
( Dx )T T 0
=
p1 x1
2
p0
( Dp )T T 0
=
p1 x1
2
By means of (11), the previous two uncertainties are written as
x0
( Dx ) l = (17)
2
p0
( Dp ) l = (18)
2
By means of (17), (18), and the relation x0 p0 = h , we obtain
h
( DxDp ) l = (19)
2
Therefore, the coherent states are states of minimum position-momentum

uncertainty and the two individual uncertainties are equally distributed.
iii) The energy expectation value of the QHO in the coherent state l is
E l
= l H l
The Hamiltonian of the QHO is written as

1
H = hw a a +
2
Thus
1 1
E l
= l H l = E = l hw a a + l = hw l a a + l =
l
2 2
1 1
= hw l a a l + l l = hw l l a l +
2 { 2
1
where, in the last equality, we used that a l = l l .
Thus
1
E l
= l l a l + hw (20)
2
46
Turning again for more clarity to the general notation for inner products, the inner
product l a l is written as ( l , a l ) , and using the definition of the Hermitian
conjugate of a , we have

( l , a l ) = ( a ) l , l = ( a l , l ) = ( l l , l ) = l * ( l , l )

{ 1424
3
a 1
Thus
l a l = l * (21)
Using (21) and that a l = l l , we have

2
l a a l = l l a l = ll * = l
That is
2
l a a l = l (22)
Also, since N a a , (22) is the expectation value of the number operator in the
coherent state l .

2 1
E l
= l + hw (23)
2
This is the energy expectation value in the coherent state l .
Observe that the energy expectation value depends only on the magnitude of l , not
on its phase.
The energy uncertainty in the coherent state l is
( )
2
( DE ) l = E2 l
- E l
Well calculate the expectation value of the energy squared in the state l , which is
E2 = l H 2 l
l
Using the previous expression of the QHO Hamiltonian, we have

1 1 2 1 1
H 2 = hw a a + hw a a + = ( hw ) a a + a a + =
2 2 2 2

2 1 1 1 2 1
= ( hw ) a aa a + a a + a a +

= ( hw ) a aa
a + a a +
2244 2 3 4 4
14 4
1
2 a a
2
That is
47
2
a + a a +
1
H 2 = ( hw ) a aa
4
Then, the expectation value of the energy squared is written as
2
a + a a + l =
1
E2 = l ( hw ) a aa
l
4
2
a l + l a a l + l l
1
= ( hw ) l a aa
4
Using (22) and that the state l is normalized, we obtain
2
a l + l + (24)
1
= ( hw ) l a aa
2
E2 l
4
Well now calculate the inner product l a aa
a l .
Using again the general notation for inner products, we have

(
a l = l l , a ( aa
a l l , a aa )) = l ( a ) l , aa

l a aa l l =
{ {
ll a

l = ll * ( l , aa
= l a l , aa
l )
{
ll
That is
l a aa
a l = l
2
(l
) (25) l
, aa
To calculate the inner product ( l , aa

l ) , we use the commutator a , a = 1 to
with 1+a a .
replace aa
Thus, we have
(l l
, aa ) = ( l , (1+a a ) l ) = ( l

, l ) + ( l , a a l )
Using (22) and that (l ,l ) = 1 (the state l is normalized), we obtain
(l l
, aa ) = 1+ l 2

4 2
l a aa
a l = l + l (26)
48

2 1 2 1
E l = ( hw ) l + l + l + = ( h w ) l + l + l + =
2 4 2 2 2 4 2
4 14 4244 4
3
2 1
l +
2

2
2
2 2 1
( )
2
2 1 2
= ( hw ) l + l + = ( l hw ) + l + hw = ( l hw ) + E
2 2 2
2 2 l

where, in the last equality, we used (23).
Thus
( )
2
= ( l hw ) + E
2
E2 l
(27)
l
Then, the energy uncertainty in the coherent state l is
( ) -( E )
2 2
( DE ) l ( l hw )
2
= + E l l
= l hw
That is
( DE ) l = l hw (28)
Another way of calculating the energy expectation value and uncertainty in the
coherent state l is by using its expansion in the energy basis of the QHO.
In i, we showed that the state l is generated by the action of the displacement
operator in the ground state, i.e. l = D ( l ) 0 . Since the displacement operator is
the operator T , we have
p1 , x1
l = Tp , x 0 (29)
1 1
1 x1 p1
with l = +i .
2 x0 p0
Also, in the exercise 6, we showed that the expansion of the state Tp1 , x1 0 in the
energy basis of the QHO is
1 2 l
n
Tp1 , x1 0 = exp - l n
2 n =0 n !
By means of (29), the previous expansion is written as
l
n
1
l = exp - l
2
n (30)
2 n =0 n !
This is the expansion of the coherent state l in the energy basis of the QHO.
Using (30), the action of the Hamiltonian on the state l yields
49
1 2 l 1 2 l
n n
H l = exp - l
H n = exp - l En n
2 n =0 n ! 2 n =0 n !
1
Using that En = n + hw , we obtain
2
1 2 l
n
1
H l = exp - l n + hw n =
2 n =0 n ! 2
1 2 ln 1 ln
= exp - l n n + n hw
2 n =1 ( n - 1) ! 2 n =0 n !

Changing the summation index of the first series to n = n - 1 , we obtain

ln
l n+1
n = n + 1 n + 1
n =1 ( n - 1)! n = 0 n !
Changing again the summation index to n , we end up to

ln
l n+1
n = n +1 n +1
n =1 ( n - 1)! n= 0 n!
Thus, the action of the Hamiltonian on l yields
1 2 l 1 ln
n +1
H l = exp - l
n +1 n +1 + n hw (31)
2 n= 0 n ! 2 n =0 n !
From (30), the bra l is
1 2 (l )
m *
l = exp - l m (32)
2 m=0 m !
Using (31) and (32), the energy expectation value of the QHO in the state l is
E l
= l H l =
( lm )
*
1 2

1 2 l
n +1
1 ln
= exp - l
m exp - l n +1 n +1 + n hw =
2 m =0 m ! 2 n =0 n ! 2 n =0 n !

( l m )* n +1 ( lm ) ln
*

( ) l
1
= exp - l n +1 m n +1 + m n hw
2
m ,n = 0 m ! n ! 2 m, n =0 m ! n !

Using the orthonormality of the energy eigenstates, i.e.
m n + 1 = d m, n +1 and m n = d mn
we obtain
50
( l m )* n +1 ( lm ) ln
*

( ) l
1
m
n + 1d m ,n +1 + d mn hw =
2
E l = exp - l
,n= 0 m! n! 2 m, n =0 m ! n !

( l n +1 )
1 (l ) l n
* n *

= exp - l
2
(

n= 0
)(
l n +1
n + 1) ! n !
n +1 +
2 n= 0 n! n!
hw =

( l n +1 ) n +1
1 (l ) ln
* n *

2
(
= exp - l
n= 0
) n!
l
+
n ! 2 n =0 n ! n !
hw =

(ln ) l* n
l l 1 (l ) l
* n * n
= exp - l
2
(

n= 0
) n!
+
n ! 2 n =0 n!
hw =

( l n ) l n 1 (ln ) l n
* *
( )

= exp - l l l + hw =
2 *
n =0 n! 2 n= 0 n!

2

1 l
n
(
2
= exp - l l +

2
) 2 n =0 n !
hw

But
ln
2
( )
l
n 2
l
2n
( )
l
2 n
( )

= = =
2
= exp l
n= 0 n! n =0 n! n =0 n! n =0 n!
Thus
E l
= exp - l( 2
) l 2 1
2
2
( )
2 1
+ exp l hw = l + hw
2
That is
2 1
E l
= l + hw
2
which is the relation (23).
Working in a similar way, we have
1 2 l 2 1 2 l
n n
H 2 l = exp - l H n = exp - l En 2 n =
2 n =0 n ! 2 n =0 n !
2
1 2 l 1 1 2 l 2 1
n n
= exp - l n + hw n = exp - l n + n + n ( hw ) =
2

2 n =0 n ! 2 2 n= 0 n ! 4
1 2 l 2 1
n
= exp - l n + n + n ( hw )
2
2 n= 0 n ! 4
Using that n 2 = n ( n - 1) + n , we obtain
51
1 2 l 1
n
H 2 l = exp - l n ( n - 1) + 2n + n ( hw ) =
2
2 n =0 n ! 4
1 2 ln
ln 1 ln
= exp - l n ( n - 1) n + 2 n n + n ( hw ) =
2
2 n = 2 ( n - 2 ) ! n =1 ( n - 1) ! 4 n =0 n !

1 2 l l ln
n+ 2 n +1
1
= exp - l ( n + 2 )( n + 1) n + 2 + 2 n +1 n +1 + n ( hw )
2
2 n= 0 n ! n =0 n! 4 n= 0 n !
where, in the first series, we changed the summation index, first to n = n - 2 , and
then again to n , and in the second series, we changed the summation index, first to
n = n - 1 , and then again to n .
Thus
H 2 l =
1 2 l l n+1 1 ln
n+ 2
= exp - l ( n + 2 )( n + 1) n + 2 + 2 n ( hw ) (33)
2
n +1 n +1 +
2 n =0 n ! n =0 n! 4 n =0 n !
The expectation value of the energy squared in the state l is
E2 = l H 2 l
l
By means of (32) and (33), the previous expression becomes
E2
l ( (
= exp - l
2
))
( l m )* n + 2 (l ) 1 (l ) ln
m * m *
l
l n+1
( n + 2 )( n + 1)d m,n+ 2 + 2 n + 1d m, n +1 + d m , n ( hw )
2
m ,n = 0 m ! n ! m, n =0 m! n! 4 m, n =0 m ! n !

But

(l )m *
l n +2
(l )m *
l n+1 1 (l ) ln
m *

m, n =0 m! n!
( n + 2 )( n + 1)d m,n+ 2 + 2
m ,n = 0 m! n!
n + 1d m, n +1 +
4 m, n =0 m ! n !
d m ,n =
=

(l )
n+2 *
l n+ 2
( n + 2)( n + 1) + 2

(l ) n +1 *
1 (l ) l n
l n+1
n +1 + =
n *
n= 0 ( n + 2)! n! n =0 ( n + 1)! n! 4 n =0 n ! n !
=

(l ) n+ 2 *
l n+ 2
+ 2

(l )
n +1 *
l n+1 1 (l ) ln
+ =
n *
n= 0 n! n! n= 0 n! n ! 4 n =0 n ! n !
=

(l ) (l )
n * 2 *
l nl 2
+ 2

(l ) l
n * *
l nl 1 (l ) ln
+ =
n *
n= 0 n! n! n =0 n! n ! 4 n =0 n ! n !
= l2 (l

(l ) n *
ln
(l ) n *
ln 1 (l ) l n
n *
)
2 *
n! n!
+ 2ll *
n!
+
n ! 4 n= 0 n! n!
=
n= 0 n =0
52
* 1
= l 2 ( l 2 ) + 2ll * +
ln 1
= l 2 ( l* ) + 2 l +
2 2
2
( )
l
n 2
=
4 n =0 n ! 4 n =0 n!
l
2n
( )
l
2 n
( ) +2 l

1 1
= ( ll * ) + 2 l +
2 2
+
2 2 2
= l =
4 n= 0 n ! 41 n
424
n =0 !
3
( )
exp l
2
4 2 1
= l + 2 l + exp l
4
2
( )
Thus
E2 l ( (
= exp - l
2
)) l 4 2 1
4
2 2 4

2
( )
1
+ 2 l + exp l ( hw ) = l + 2 l + ( hw ) =
4
2
2
2 1 2 2 2 1
= l + + l ( hw ) = l + hw + ( l hw )
2 2
2 2

That is
2
2 1
= ( l hw )
2
E 2
+ l + hw
l
2
or, using (23),
( )
2
= ( l hw ) + E
2
E2 l
l
Therefore, the energy uncertainty in the coherent state l is
( ) ( )
2 2
( DE ) l = ( l hw ) + E
2
l
- E l
= l hw
That is
( DE ) l = l hw
which is the relation (28).

iv) As we saw, the expansion of the state l in the energy basis is given by (30).
Also, the time evolution of the energy eigenstate n is
iE t
n t = exp - n n
h
Thus, the time evolution of the state l let us denote it by l ( t ) is
1 2 l 1 2 l
n n
iE t
l (t ) = exp - l n t = exp - l exp - n n
2 n =0 n ! 2 n =0 n ! h
That is
53
1 2 l
n
iE t
l (t ) = exp - l exp - n n (34)
2 n =0 n ! h
Then, the action of the annihilation operator on the time evolution of the coherent
state l yields
1 2 l
n
iE t
a l ( t ) = exp - l exp - n a n
2 n =0 n ! h
Using that a n = n n - 1 , we obtain
1 2 l
n
iE t
a l ( t ) = exp - l exp - n n n - 1 =
2 n =1 n ! h
1 2 l n
iE t
= exp - l exp - n n - 1
2 n =1 ( n - 1) ! h
Changing the summation index to n = n - 1 , we obtain

1 2 l
n+1
iE t
a l ( t ) = exp - l exp - n+1 n =
2 n=0 n! h
1 2
l n
iE t
= l exp - l exp - n+1 n
2 n=0 n! h
1
Besides, using that En = n + hw , we derive that
2
En+1 = En + hw
Substituting into the expression of a l ( t ) , we obtain
1 2 l
n
i ( En + hw ) t
a l ( t ) = l exp - l exp - n =
2 n= 0 n! h
1 2 l
n
iE t
= l exp ( -iwt ) exp - l exp - n n = l exp ( -iwt ) l ( t )
2
1444444 n!
n=0424444444 h 3
l (t )
That is
a l ( t ) = l exp ( -iwt ) l ( t ) (35)
Therefore, the state l ( t ) is an eigenstate of the annihilation operator, and thus it

remains a coherent state, but its eigenvalue, l exp ( -iw t ) , changes periodically with
time.
We observe that the magnitude of the eigenvalue of l ( t ) is
l exp ( -iwt ) = l exp ( -iwt ) = l

14243
1
54
That is, the magnitude of the eigenvalue of l ( t ) is constant.
To summarize, the time evolution of a coherent state is a coherent state with

eigenvalue having constant magnitude l and time-dependent phase exp ( -iwt ) .
8) Express the energy expectation value in a coherent state l of the QHO in

terms of the expectation values of the position and momentum.
Solution
In the previous exercise, we showed that the energy expectation value in a coherent
state l is
2 1
E l
= l + hw (1)
2
Since the state l is an eigenstate of a with eigenvalue l , we have
a l = l l
Then, the expectation value of a in the state l is
a = l a l = l l l = l l l
l {
1
That is
a l
= l (2)
In terms of the length and momentum scales, the annihilation operator is written as
1 x p
a = +i
2 x0 p0
Thus
1 x l p l
a = +i
l
2 x0 p0

Comparing the previous equation with (2) yields
1 x l p l
+i = l = Re l + i Im l
2 x0 p0

or
1 x l 1 p l
+i = Re l + i Im l
2 x0 2 p0
Since the expectation values of the position and momentum are real, the previous
equation gives
55
1 x l
Re l = (3)
2 x0
1 p l
Im l = (4)
2 p0
Using (3) and (4), the square of the magnitude of l is
2 2

1 x l p l

l = ( Re l ) + ( Im l )
2 2 2
= +
2 x0 p0

That is
2 2

1 x l
2
p l
(5)
l = +
2 x0 p0

2 2

1 x l p l
E = + + 1 hw
l
2 x0 p0

h
Substituting x0 = and p0 = mhw into the previous expression, we obtain
mw

( x ) +( p )
( p )

2 2 2
1 1
(x )
l l 2 l
E = + 1 hw = mw 2 + + hw =
l
2 h mhw 2 l
m
mw
( )
2
p 1
( x ) + h2w
l 2
= + mw 2 l
2m 2
That is
( p )
2
1
( x ) + h2w (6)
l 2
E l
= + mw 2 l
2m 2
hw hw
Observe that E
, as expected, since
l
is the ground-state energy of the
2 2
QHO, and the equality holds if and only if the expectation values of the position and
momentum are both zero in the coherent state l . There is only one coherent state in
which the position and momentum expectation values are both zero, and that coherent
state is the ground state.
In the previous exercise (question ii), we showed that
l = Tp Tx 0 (7)
1 1
56
Using (7), it is easily shown that x l

= x1 and p l
= p1 .
Then, (6) is written as
p12 1 hw
E l
= + mw 2 x12 + (8)
2m 2 2
hw
We see that E l
= if and only if x1 = 0 and p1 = 0 , thus if and only if
2
T 0 = 0
l = T{0 {0
1 1
hw
Therefore, E l
= if and only if the coherent state is the ground state of the
2
QHO.
9) Calculate the wave functions describing a coherent state l in the position

and momentum representations, respectively. Express the results in terms of
the expectation values of the position and momentum.
Solution
The wave function of the state l in the position representation is
y l ( x) = x l
In the exercise 7 (question ii), we showed that
l = Tp Tx 01 1
Then, the wave function y l ( x ) is written as
y l ( x ) = x Tp Tx 0 = Tp ( x ) Tx ( x ) x 0 = Tp ( x ) Tx ( x )y 0 ( x )
1 1 1 1 { 1 1
y 0 ( x)
That is
y l ( x ) = Tp ( x ) Tx ( x )y 0 ( x ) (1)
1 1
where, by Tp1 ( x ) we denote the momentum translation operator Tp1 in the position
representation, by Tx1 ( x ) we denote the spatial translation operator Tx1 in the position
representation, and y 0 ( x ) is the ground-state wave function of the QHO in the
position representation.
d
In the position representation, x = x and p = -ih .
dx
Thus
ip x
Tp1 ( x ) = exp 1
h
and
57
d
i -ih dx x1
Tx1 ( x ) = exp - = exp - x d
1
h dx

ip1 x d
y l ( x ) = exp exp - x1 y 0 ( x ) (2)
h dx
d
Using the Taylor expansion of exp - x1 , i.e.
dx
n
d

- x1
( - x1 ) d n
n

d dx
exp - x1 = =
dx n =0 n! n =0 n ! dx n
we obtain
d ( - x1 )n d n
( - x1 ) d n
n
exp - x1 y 0 ( x ) = y ( x ) = y 0 ( x) =
dx n = 0 n ! dx n 0 n ! dx n
n =0
( - x1 ) y 0(n ) ( x )
n

= y 0 ( x) =
( n)
( - x1 ) = y 0 ( x - x1 )
n
n= 0 n! n =0 n!
We remind that the Taylor series of a (proper) function f ( x ) about x is
f ( x) =

f(
m)
( x )
( x - x )
m =0 m!
Thus
d
exp - x1 y 0 ( x ) = y 0 ( x - x1 ) (3)
dx
Obviously, (3) also holds for an arbitrary function y ( x ) that has derivatives of all
orders.

ip1 x
y l ( x ) = exp y 0 ( x - x1 ) (4)
h
Using that x1 = x l
and p1 = p l
, (4) becomes
i p x
y l ( x ) = exp
h
l
y 0 x - x
( l ) (5)

58
This is the wave function of the coherent state l in the position representation,
expressed in terms of the position and momentum expectation values.
1 1 1 x 2
Using that y 0 ( x ) = exp - , the wave function y l ( x ) takes the
x0 4
1
2 x0
p
form
x- x i p l x
2
1 1 - 1 (6)
y l ( x) =
l
exp +
x0 4
1
2 x0 h
p
h
where x0 = is the length scale of the QHO.
mw
The wave function y% l ( p ) , i.e. the wave function describing the coherent state l in
the momentum representation, can be calculated by taking the Fourier transform of
the wave function y l ( x ) , i.e.
1

1 2 ipx
y% l ( p ) = dxy l ( x ) exp -
2p h - h
Since the wave function y l ( x ) is Gaussian, the previous integral can be calculated
using that

p b2
dx exp ( - ax 2 + bx + c ) = exp + c , with a > 0
-
a 4a
See, for instance, https://en.wikipedia.org/wiki/Gaussian_integral.
Alternatively, we may again use that
l = Tp Tx 0
1 1
or better, that
l = Tx Tp 0
1 1
We remind that, as shown in the exercise 4, the operators Tp1 Tx1 and Tx1 Tp1 differ
only by a constant phase, and thus their action on an arbitrary state is physically the
same.
Using the previous equation, the wave function y% l ( p ) = p l is written as
y% l ( p ) = p Tx Tp 0 = Tx ( p ) Tp ( p ) p 0 = Tx ( p ) Tp ( p )y% 0 ( p )
1 1 1 1 { 1 1
y% 0 ( p )
That is
y% l ( p ) = Tx ( p ) Tp ( p )y% 0 ( p ) (7)
1 1
59
where, by Tx1 ( p ) we denote the spatial translation operator Tx1 in the momentum
representation, by T ( p ) we denote the momentum translation operator T in the
p1 p1
momentum representation, and y% 0 ( p ) is the ground-state wave function of the QHO

in the momentum representation.
d
In the momentum representation, x = ih and p = p .
dp
Thus
ipx
Tx1 ( p ) = exp - 1
h
d
ip1ih dp d
Tp1 ( p ) = exp = exp - p1
h dp

ipx1 d
y% l ( p ) = exp - exp - p1 y% 0 ( p )
h dp
From (3), it is obvious that
d
exp - p1 y% 0 ( p ) = y% 0 ( p - p1 )
dp
Then, the wave function y% l ( p ) is written as
ipx1 %
y% l ( p ) = exp - y 0 ( p - p1 )
h
Using again that x1 = x l
and p1 = p l
, we end up to
ip x
y% l ( p ) = exp -
h
l
(
y% 0 p - p
l ) (8)

This is the wave function of the coherent state l in the momentum representation,
expressed in terms of the position and momentum expectation values.
1 1 1 p 2
Using that y% 0 ( p ) = exp - , the wave function y% l ( p ) takes the
p0 4
1
2 p0
p
form
p- p
2
ip x
1 1 -1 (9)
y% l ( p ) =
l l
exp -
p0 4
1
2 p0 h
p
Note
60
The energy eigenfunctions of the QHO in the position representation, i.e. the
functions y n ( x ) , and the energy eigenfunctions of the QHO in the momentum
x p
x0 p0
}
representation, i.e. the functions y% n ( p ) , are related by y n ( x ) y% n ( p ) and
p x
p0 x0
}
y% n ( p ) y n ( x ) .
In other words, the functions y% n ( p ) are derived from y n ( x ) by replacing
position with momentum and the length scale with the momentum scale, and,
likewise, the functions y n ( x ) are derived from y% n ( p ) by replacing momentum
with position and the momentum scale with the length scale.
This is a unique property of the QHO, which is due to the form of the harmonic
oscillator potential.
10) Find the time evolution of the wave functions y l ( x ) and y% l ( p ) of the
previous exercise.
Solution
In the exercise 7, we showed that the state l ( t ) , which is the time evolution of the
coherent state l , is also a coherent state, but with time-varying eigenvalue
l ( t ) = l exp ( -iwt ) (1)
Since the state l ( t ) is a coherent state, it can be written as
l ( t ) = Tp (t )Tx (t ) 0 (2)
1 1
1 x1 ( t ) p (t )
with l ( t ) = +i 1 .
2 x0 p0
We remind that, as shown in the exercise 7 (question ii), the coherent state l is
1 x1 p1
written as l = Tp1Tx1 0 , with l = +i .
2 x0 p0
Then, the time evolution of the wave function y l ( x ) , which describes the state l ,
is the wave function
y l ( x, t ) = x l ( t ) (3)
which describes the state l ( t ) .

Using (2), (3) becomes
y l ( x, t ) = x Tp (t )Tx (t ) 0 = Tp (t ) ( x ) Tx (t ) ( x ) x 0 = Tp (t ) ( x ) Tx (t ) ( x )y 0 ( x )
1 1 1 1 { 1 1
y 0 ( x)
61
That is
y l ( x, t ) = Tp (t ) ( x ) Tx (t ) ( x )y 0 ( x ) (4)
1 1
where, by Tp1 ( t ) ( x ) we denote the momentum translation operator Tp1 ( t ) in the position
representation, by Tx1 ( t ) ( x ) we denote the spatial translation operator Tx1 ( t ) in the
position representation, and y 0 ( x ) is the ground-state wave function of the QHO in
the position representation, i.e.
1 1 1 x 2
y 0 ( x) = exp -
x0 4
1
2 x0
p
d
In the position representation, x = x and p = -ih .
dx
Thus
ip ( t ) x
Tp1 ( t ) ( x ) = exp 1
h
d
i -ih dx x1 ( t )
= exp - x1 ( t )
d
Tx1 ( t ) ( x ) = exp -
h dx

Since the function x1 ( t ) depends only on time, it commutes with the operator
d
.
dx
Thus
ip1 ( t ) x d
y l ( x, t ) = exp exp - x1 ( t ) y 0 ( x ) (5)
h dx
In the previous exercise, we proved that
d
exp - x1 y 0 ( x ) = y 0 ( x - x1 )
dx
Then, obviously,
d
exp - x1 ( t ) y 0 ( x ) = y 0 ( x - x1 ( t ) )
dx
ip1 ( t ) x
y l ( x, t ) = exp y 0 ( x - x1 ( t ) ) (6)
h
62
In the exercise 8, we explained that for the coherent state l = Tp1Tx1 0 , it holds that
x l
= x1 and p l
= p1 .
Then, for the state (2) we have x l (t )
= x1 ( t ) and p l (t )
= p1 ( t ) .
Thus, (6) becomes
i p x
y l ( x, t ) = exp

l (t )
h
(
y 0 x - x
l (t ) ) (7)

Since the state l ( t ) is the time evolution of the state l , the expectation values
x l (t )
and p l (t )
are the time evolution of the expectation values x l
and p l
.
1 x1 p1 1 x1 ( t ) p (t )
Besides, substituting l = + i and l ( t ) = + i 1 into (1),
2 x0 p0 2 x0 p0
we obtain
1 x1 ( t ) p ( t ) 1 x1 p1
+i 1 = + i exp ( -iwt ) =
2 x0 p0 2 x0 p0
1 x1 p1
= + i ( cos wt - i sin wt ) =
2 x0 p0
1 x1 x1 p1 p1
= cos w t - i sin wt + i cos wt + sin wt =
2 x0 x0 p0 p0
1 x1 p1 p1 x1
= cos wt + sin w t + i cos wt - sin wt
2 x0 p0 p0 x0
That is
1 x1 ( t ) p ( t ) 1 x1 p1 p1 x1
+i 1 = cos w t + sin wt + i cos wt - sin wt
2 x0 p0 2 x0 p0 p0 x0
or
x1 ( t ) p ( t ) x1 p p x
+i 1 = cos w t + 1 sin wt + i 1 cos wt - 1 sin wt
x0 p0 x0 p0 p0 x0
Since the translations x1 ( t ) and p1 ( t ) are real, the previous equation gives
x1 ( t ) x1 p
= cos wt + 1 sin wt (8)
x0 x0 p0
p1 ( t ) p1 x
= cos wt - 1 sin wt (9)
p0 p0 x0
From (8) we obtain
x0
x1 ( t ) = x1 cos wt + p1 sin wt
p0
63
Substituting the length and momentum scales into the previous equation yields
p1
x1 ( t ) = x1 cos wt + sin wt
mw
Since x l (t )
= x1 ( t ) , the previous equation gives
p1
x l (t )
= x1 cos wt + sin w t (10)
mw
Similarly, from (9) we obtain
p1 ( t ) = -mw x1 sin wt + p1 cos wt
Since p l (t )
= p1 ( t ) , the previous equation gives
p l (t )
= - mw x1 sin wt + p1 cos wt (11)
Thus, to summarize, the time evolution of the wave function y l ( x ) is given by (7),
where the expectation values x l (t )
and p l (t )
are given by (10) and (11),
respectively.
In the same way, we find that the time evolution of the wave function y% l ( p ) is the
wave function
ip x
y% l ( p, t ) = exp -
h
l (t )
(
y% 0 p - p
l (t ) ) (12)

1 1 1 p 2
where y% 0 ( p ) = exp - is the ground-state wave function of the
p0 4
1
2 p0
p
QHO in the momentum representation.
We leave to the reader to verify (12).
11) Overlap and overcompleteness of the coherent states.

i) Calculate the overlap between two coherent states l and l . What do
you observe?
ii) Show that the set of all coherent states satisfy a completeness relation.
Solution
i) Using the expansion of a coherent state in the energy basis of the QHO, which we
proved in the exercise 7, we have
l
n
1
l = exp - l
2
n
2 n =0 n !
Thus
1 2 (l )
m *
l = exp - l
m
2 m= 0 m !
64
Then, the inner product l l is written as
( l m )
*

1 2 l
n
1 2
l l = exp - l
m exp - l n =
2 m =0 m ! 2 n =0 n !

1 2 1 2 (l ) l
m *
n
= exp - l - l mn
2 2 m ,n = 0 m ! n !
Using the orthonormality of the energy eigenstates, i.e. m n = d mn , we obtain
1 2 1 2 ( l ) l
n m *
l l = exp - l - l d mn =
2 2 m ,n = 0 m ! n !
1 2 1 2 (l ) l
n * n
= exp - l - l
2 2 n =0 n!
Now, using that ( l n ) = ( l * ) , we obtain

* n
1 2 1 2 (l ) l 1 2 1 2 ( l l )
* n n * n
l l = exp - l - l
= exp - l - l =
2 2 n =0 n! 2 2 14
n =0 n !3
24
(
exp l *l )
1 2 1 2 1 2 1 2
= exp - l - l exp ( l *l ) = exp - l - l + l *l =
2 2 2 2
1 2
( 2
= exp - l + l - 2ll *
2

)
That is
l l = exp -
1
2
(2 2
l + l - 2ll * (1)

)
If l l , (1) gives l l 0 , i.e. the states are not orthogonal, they overlap.
The annihilation operator is not Hermitian, thus two eigenstates with different
eigenvalues, i.e. two different coherent states, are not orthogonal.
2 2
Well now write the term l + l - 2ll * as a square plus or minus something.
We have
2 2
l + l - 2ll * = ll * + l l * - ll * - ll * = ll *
1444 + l l4 - ll * - l *3
*24444 l + l *l - ll * =
( l - l )( l - l ) * *
= ( l - l ) ( l * - l * ) + l *l - ( l *l ) = ( l - l )( l - l ) + 2i Im ( l *l ) =
* *
14 4244 3
(
2 i Im l *l )
= l - l + 2i Im ( l *l )
2
That is
65
l + l - 2ll * = l - l + 2i Im ( l *l )
2 2 2

1
l l = exp -
2
2
(

1
2
2
)
l - l + 2i Im ( l *l ) = exp - l - l - i Im ( l *l ) =

1
2
2
= exp - l - l exp -i Im ( l *l )

( )
That is
( )
l l = exp -i Im ( l *l ) exp - l - l (2)
2
1 2
The overlap between the two coherent states is the absolute value of l l , i.e.
1 2
l l = exp - l - l (3)
2
Since l - l 0 , the maximum overlap happens when l = l , i.e. when the two
eigenvalues are equal, i.e. when the two coherent states coincide, and it is equal to the
norm of the state to the square, i.e. 1.
If the distance between the eigenvalues of the two coherent states is large, then the
term l - l is large, and the overlap between the two states is very small.
On the contrary, if the distance between the two eigenvalues is small, i.e. if the
eigenvalues are close to each other in the complex plane, the overlap tends to 1.
ii) Using again the expansion of a coherent state in the energy basis of the QHO, the
integral d l l l , where l , is written as
1 2 (l )
m *
1 2 l
n
d l l l = 2
d l exp - l
n =0
n
n!
exp -
2
l
m= 0 m !
m =

(l )* m
ln
( )

= d l exp - l
2
n m
m ,n = 0 m !n !
In polar coordinates, the complex number l is written as l = reij . Thus, its complex
conjugate l * is l * = reij . In polar coordinates, the differential d l is written as
d l = rdrdj , where r is from 0 to , while the polar angle j is from 0 to
2p (excluded). Thus
l = rdrdj exp ( -r

( re ) ( re )
- ij m ij n
dl l
2
) m !n !
n m =
m ,n = 0
r m + ne (
i n - m )j
= rdrdj exp ( - r 2 ) n m =
m ,n = 0 m !n !
2p i ( n - m )j

n m
= d j e drr
m + n +1
exp ( - r 2 )
m, n =0 0 0 m !n !
66
2p
i ( n - m )j
The integral dj e
0
is zero when m n and 2p when m = n , i.e.
2p
i ( n - m )j
dj e
0
= 2pd mn
Thus, the integral d l l l becomes

n m
d l l l = 2pd mn drr m + n +1
exp ( - r 2
) =
m ,n = 0 0 m !n !
n n
= 2p drr 2 n +1exp ( -r 2 )
n =0 0 n!
But

1 1
exp ( - r ) = dr 2 r 2 n exp ( - r 2 ) = dr 2 ( r 2 ) exp ( - r 2 ) =
n
drr
2 n +1 2
0
20 20
} 1
r2 =s
1 1
= dss n exp ( - s ) = G ( n + 1) = n !
20 2 2
where we made use of the property G ( n ) = ( n - 1) ! of the gamma function

G ( t ) = dxx t -1 exp ( - x ) , Re t > 0 .
0
Substituting into the expression of the integral d l l l , we obtain

1 n n
d l l l = 2p
n =0 2
n !
n!
= p n n =p
1 424
n =0
3
1
where we used the completeness relation of the energy eigenstates.

Thus
1
p dl l l = 1 (4)
This is the completeness relation of the set of the coherent states. Moreover, since the
coherent states overlap, the set is said to be overcomplete.
67
An intuitive introduction to the squeezed states of the QHO
III. An intuitive introduction to the squeezed states of

the QHO
We know that the ground-state wave function of a QHO having length scale x0 is
1 1 1 x 2
y 0 ( x) = exp -
x0 4
1
2 x0
p
Let us now consider the wave function
1 1 1 x 2
y ( x; x ) = exp -
x x0 4
1
2 x x0
p
where x is a dimensionless, positive real parameter.
If x 1 , the wave function y ( x; x ) is not the ground-state wave function of the QHO
having length scale x0 , but it can be thought of as the ground-state wave function of
another QHO, having length scale x x0 .
Moreover, for every value of x , y ( x; x ) describes the ground state of a, different
each time, QHO.
Since y ( x; x ) always describes the ground state of a QHO, the position-momentum
uncertainty product will be minimum in the state described by y ( x; x ) .
If x0 = x x0 is the length scale of the new QHO, with ground-state wave function
p
y ( x; x ) , then its momentum scale will be p0 = 0 , where p0 is the momentum scale
x
of the first QHO.
Indeed, since the product of the two scales must be equal to h , we have
x0 p0 = h = x0 p0
Thus
p0
x0 p0 = x0 p0 x x0 p0 = x0 p0 p0 =
x
Since the wave function y ( x; x ) describes the ground state of a QHO with scales x0
and p0 , the position and momentum uncertainties in the state described by y ( x; x )
will be
x0 p0
Dx = and Dp = .
2 2
In terms of the scales x0 and p0 of the first QHO, the previous two uncertainties are
respectively written as
x x0 p0
Dx = and Dp = .
2 2x
68
If x 1 , we have
Dx x 1 Dp
= =
x0 2 2x p0
That is
Dx Dp

x0 p0
Thus, with respect to the first QHO, in the state described by y ( x; x ) , which is a state
of minimum position-momentum uncertainty product, the two individual uncertainties
are not equally distributed.
Therefore, for the first QHO, the state described by y ( x; x ) is always, i.e. for every
value of the parameter x , a state of minimum position-momentum uncertainty
h
product, i.e. DxDp = , but the uncertainties of the position and momentum are not
2
equally distributed, taking different values each time the parameter x changes.
For each value of the parameter x , the state described by y ( x; x ) is called a
squeezed state of the first QHO, and particularly, it is a squeezed state of the ground
state of the first QHO.
In the same way, from each coherent state of the first QHO, we construct squeezed
states of the first QHO.
Thus, making the change x0 x x0 in each coherent state of a QHO having

length scale x0 , we construct squeezed states of that QHO.
We defined the squeezed states in the position representation, as this provides a better
intuitive picture of the squeezed states. However, working in the same way, we may
well define the squeezed states in the momentum representation too.
The parameter x determines the squeezing of the position and momentum

uncertainties, and thus we may call it squeezing parameter.
The coherent states as states of minimum energy expectation

value
12) Show that the energy expectation value of a squeezed state is always
greater than the energy expectation value of its respective coherent state, and
only when the squeezed state coincides with the respective coherent state,
i.e. only when the squeezing parameter is 1, the two energy expectation
values are equal.
Solution
The expectation values do not depend on the representation we may use to calculate
them they are representation free and thus we choose to work in the position
representation.
69
In the exercise 9, we showed that, in the position representation, the coherent state
l is described by the wave function
x- x i p l x
2
1 1 - 1 (1)
y l ( x) =
l
exp +
x0 4
1
2 x0 h
p
To construct a squeezed state of the coherent state l , we make in (1) the change
x0 x x0 , with x > 0 .
Thus, in the position representation, a squeezed state is described by the wave
function
x- x i p l x
2
1 1 - 1 (2)
y l ( x; x ) =
l
exp +
x x0 4
1
2 x x0 h
p
As shown in the exercise 8, the energy expectation value of the QHO in the coherent
state l is
(p )
2
1
( ) + h2w (3)
l 2
E l
= + mw 2 x l
2m 2
Denoting by l ; x the squeezed state that is described by the wave function
y l ( x; x ) , the energy expectation value of the QHO in the state l ; x is
E l ;x
= l ; x H l ; x (4)
We may calculate the previous energy expectation value directly in the position
representation, and we urge the reader to do the relevant calculations.
Alternatively, we may use that the state l ; x is a coherent state of a second QHO,
p0
having length and momentum scales x x0 and , respectively, where x0 and p0 are,
x
respectively, the length and momentum scales of the first QHO.
Then, as shown in the exercise 7, with respect to the second QHO, the position and
momentum uncertainties in the state l ; x are equally distributed, i.e.
x x0
( Dx ) l ;x =
2
and
p0
( Dp ) l ;x =
2x
Thus
( ) x x0 x 2 x0 2
( )
2 2
x2 - x l ;x
= x2 = + x l ;x
(5)
l ;x l ;x 2
2
and
70
( ) p0 p0 2
( )
2 2
p2 - p = p2 = + p (6)
l ;x l ;x
2x l ;x 2x 2 l ;x
Well use the wave function y l ( x; x ) to calculate the position and momentum
expectation values. Then, from (5) and (6), well calculate the expectation values of
the position squared and momentum squared, and then, well use them to calculate the
energy expectation value.
In the position representation, the position expectation value in the state l ; x is
written as

= dxy l ( x;x ) xy l ( x;x )

*
x l ;x
-
Substituting into the integral the wave function y l ( x; x ) from (2), we obtain
x- x
2

1 1
- dxx exp - x x0 (7)
l
x =
l ;x
x x0 12
p
Changing the integration variable to x = x - x l
, the previous integral becomes
x- x
2
x 2
( )
- dxx exp - x x0 = dx x + x
l
-
l
exp - =
x x0

x2
x2
= dxx exp - + x l
dx exp -
( x x )
2
( x x )
2

- 0 - 0
x2
The first integral is zero, because the function x exp - is odd, as product of
(x x ) 2
0
x2
the odd function x with the even function exp - , and the integration
( x x )2
0
interval is symmetric.

p
dx exp ( -ax ) = , where a > 0 , the second integral is
2
Using that
- a
x2 1
- ( 0)
2
dx exp - = p x x = p 2
x x0
(x x )2
0
Thus
x- x
2
1
- l = p 2x x x
- dxx exp
x x0

0 l

71
x l ;x
= x l
(8)
Therefore, the position expectation value in a squeezed state is equal to the position
expectation value in the respective coherent state.
In other words, the squeezing of a coherent state does not change the position
expectation value.
In the same way, we calculate the momentum expectation value.

In the position representation, we have

d

dy ( x; x )
p l ;x
= - l
dxy *
( x; x ) -i h y
l ( x; x ) = -i h dxy l * ( x; x ) l
dx - dx
dy l ( x; x )
Using (2), the derivative is
dx
dy l ( x; x ) 1 x - x i p x l -x i p
= -
l
+
l
y l ( x; x ) = +
l
y l ( x; x )
dx x x0 x x0 h ( x x0 ) 2 h

Thus, the momentum expectation value is written as
x l -x i p
= -ih dxy l * ( x; x ) y l ( x; x ) =
l
p +
l ;x ( )
2
- x x0
h

( )

ih
dxy l * ( x; x ) x - x y l ( x; x ) + p dxy l ( x; x )y l ( x;x ) =
(x x )
= 2 l l
*
0 - -

ih
= dxy l ( x; x ) xy l ( x; x ) - x l dxy l ( x; x )y l ( x; x ) +
* *

(x x0 ) -
2

1444424444 3 1444
-
424444 3
x l ;x 1

( )

ih
+ p l dxy l * ( x; x )y l ( x; x ) = x l ;x - x l + p l
( )
2
1444
-
424444 3 x x0
1
Using (8), we end up to

p l ;x
= p l
(9)
Therefore, as in the case of the position expectation value, the squeezing of a

coherent state does not change the momentum expectation value.
dxy l ( x; x )y l ( x; x ) is 1, as the state l ; x is normalized. This

*
The integral
-
follows from the fact that the state l ; x is a coherent state of the second QHO,
72
p0
which has scales x x0 and , and thus it is generated by the action of a
x
displacement operator, which is unitary, on the ground state of the second QHO.
Substituting (8) and (9) into (5) and (6), respectively, we obtain
x 2 x02
( )
2
x 2
l ;x
= + x l
2
p0 2
( )
2
p2 = + p
l ;x 2x 2 l
Substituting into the previous two equations the length and momentum scales,
h
x0 = and p0 = mhw , we obtain, respectively,
mw
x 2h
( )
2
x2 = + x (10)
l ;x 2mw l
mhw
(p )
2
p2 = + (11)
l ;x 2x 2 l
By means of (10) and (11), the energy expectation value of the QHO in the squeezed
state l ; x is
mhw
( )
2
p 2 + p
2x 2 x 2h
( ) =
l
l ;x 1 1 2
E = + mw 2 x 2 = + mw 2 + x
2mw
l ;x l
2m 2 l ;x 2m 2
( )
2
p
1 hw 1
( )
l 2
= 2 +x2 + + mw 2 x
x
l
4 2m 2
That is
(p )
2
1
( ) + x1 + x hw
l 2
E l ;x
= + mw 2 x l 2
2
(12)
2m 2 4
By means of (3), (12) is written as
hw 1 hw 1 hw
E = E - + 2 +x 2 = E + 2 + x 2 - 2 =
l ;x l
2 x 4
l
x 4
2
1 hw
= E + x -
l
x 4
That is
2
1 hw
E = E + x - (13)
l ;x l
x 4
73
2
1 hw
Since x - 0 , (13) gives
x 4
E l ;x
E l
and the equality holds only when

1
x- = 0 x 2 = 1
{x =1
x x >0
Therefore, the energy expectation value in a squeezed state is always greater

than the energy expectation value in the respective coherent state, and only when
the squeezing parameter is 1, i.e. only when there is no squeezing, and thus the
squeezed state coincides with the coherent state, the two energy expectation
values are equal.
It is also worth noting that, as seen from the equation (13), the energy expectation
1
value in the states l ; x and l ; is the same, i.e.
x
E l ;x
= E l;
1
x
13) What is the energy expectation value in the squeezed state l ; x when
x 0+ and x ? Comment on the results.
Solution
We showed in the previous exercise that the energy expectation value in the squeezed
state l ; x is
2
1 hw
E = E + x -
l ;x l
x 4
where E l
is the energy expectation value in the respective coherent state, and it is
(p )
2
1
( x ) + h2w
l 2
E l
= + mw 2 l
2m 2
We observe that
2
2
1 1
lim+ x - = 0 - + = ( - ) =
2
x 0
x { 0

and
74
2

2
1 1
lim x - = - =
x x
{
0
Thus, in both cases, the energy expectation value tends to infinity, i.e.
lim E l ;x

x 0+ ,
In the previous exercise, we showed that the position and momentum uncertainties in
the squeezed state l ; x are, respectively,
x x0 p0
( Dx ) l ;x = and ( Dp ) l ;x = .
2 2x
We see that, when x 0+ , the position uncertainty tends to zero and the momentum
uncertainty tends to infinity. Then, the squeezed state l ; x tends to become a
position eigenstate, i.e. lim+ l ; x = x - x l
, and the position eigenstates are states
x 0
of infinite energy for the QHO.

Similarly, when x , the position uncertainty tends to infinity and the momentum
uncertainty tends to zero. Then, the squeezed state l ; x tends to become a
momentum eigenstate, i.e. lim l ; x = p - p l
, and the momentum eigenstates are
x
also states of infinite energy for the QHO.
75
General references
IV. General references

1. David J. Griffiths, Introduction to Quantum Mechanics, Prentice Hall, 1995
2. Stephen Gasiorowicz, Quantum Physics, Wiley, 1974
3. J. J. Sakurai, Jim J. Napolitano, Modern Quantum Mechanics, Second Edition,

Addison-Wesley, 2011
4. Spiros Konstantogiannis, Special Topics In One-Dimensional Quantum Mechanics:

Selected Exercises In Spatial and Momentum Translations, Lulu, 2017, available at:
http://www.lulu.com/shop/spiros-konstantogiannis/special-topics-in-one-dimensional-
quantum-mechanics-selected-exercises-in-spatial-and-momentum-
translations/ebook/product-23048614.html
76

An Intuitive Approach To The Coherent and Squeezed States of The QHO

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

An Intuitive Approach To The Coherent and Squeezed States of The QHO

Încărcat de

Drepturi de autor:

Formate disponibile

An intuitive approach to the coherent and

squeezed states of the quantum harmonic

Therefore, the wave function y ( x ) is written as

Remember that the Taylor series of a (proper) function f ( x ) about x is written

Changing the integration variable to

Thus, (2) becomes

because the ground-state wave function is normalized.

By means of (4), (3) becomes

Using the completeness of the position eigenstates, we have

Substituting x 0 = y 0 ( x ) and x y = y ( x ) , and using that the ground state is

Changing again the integration variable to

Thus (6) becomes

Likewise, the momentum expectation value of the QHO at t = 0 is

Doing again the variable change y = x - x1 , the previous equation becomes

We remind that the momentum expectation value of the QHO in an energy

At t = 0 , the state of the QHO is described by the wave function y ( x ) .

where H ( x ) is the Hamiltonian of the QHO in the position representation, i.e.

Well calculate the two integrals separately.

Thus, the first integral on the right-hand side of (11) is written as

The second integral on the right-hand side of (11) is written as

= dyy ( y ) y y ( y ) + 2 x dyy ( y ) yy ( y ) + x dyy ( y )y ( y )

Substituting the integrals (12) and (13) into (11), we obtain

More on spatial translation operators

iii) Using the Baker-Campbell-Hausdorff formula

position, momentum, and energy, as well as the position and momentum

We need to calculate the commutator x , p n .

Here, the prime denotes differentiation with respect to p .

with n 1 . If n = 0 , the commutator (2) is zero.

ii) Using the previous commutator, we obtain

x1 on a position eigenstate x yields

Since x is a number an eigenvalue of the position operator it commutes with Tx1 ,

Thus, the state Tx1 x is a position eigenstate with eigenvalue x + x1 .

Omitting the physically unimportant constant phase of A , we end up to

We need to calculate the commutator p , f ( x ) . Since the function f is unknown, it

Here, the prime denotes differentiation with respect to x .

Since the wave function f ( x ) is arbitrary,

or, in representation-free form,

Using (6), we have

for every n = 1, 2,...

Thus, the right-hand side of (5) is equal to f ( x + x1 ) .

iv) Denoting by O and O , respectively, the expectation values of an

since, in this case, f ( x ) = x .

We assume that the initial, untranslated state y is normalized.

Then, the momentum expectation value in the translated state is written as

Tx1 x 2Tx1 = ( x + x1 ) = x 2 + 2 x1 x + x12

Substituting (14) and (8) into (13), we obtain

Tx1 p 2Tx1 = Tx1 Tx1 p 2 = p 2

Substituting (17) and (10) into (16), we obtain

Momentum translation operators

3) Similar to the case of spatial translation operators, a momentum translation

ii) Show that p , Tp1 = p1Tp1 .

iv) Using the Baker-Campbell-Hausdorff formula show that, for an infinitely

position, momentum, and energy, as well as the position and momentum

Then, using also that the position operator is Hermitian, we obtain

Therefore, Tp1 is unitary.

with n 1 . If n = 0 , the commutator is zero.

iii) Using the previous commutator, we obtain

p1 on a momentum eigenstate p yields

where p is the eigenvalue of the momentum eigenstate p , i.e. it is a number, and

The state Tp1 p is thus a momentum eigenstate with eigenvalue p + p1 .

Then, omitting the physically unimportant constant phase of B , we obtain

Well use the momentum representation to calculate the commutator x, f ( p ) , since

Thus, choosing an arbitrary wave function f ( p ) (in the momentum representation),