Page 1
P a g e | 1
UNIT I
Lesson – 1
ELECTROMAGNETIC THEORY AND OPTICAL PHYSICS
1.0 Aims and Objectives
1.1 (a) Electric potential at a point due to an electric dipole:
(b) Electric field at any point due to an electric dipole
1.2 Dielectric polarization
1.3 Relative permitivity (dielectric constant) and displacement vector
1.4 External field of a dielectric medium
1.5 Relation between electric displacement, electric field and polarisation (D, E and P)
1.6 Molecular field in a dielectric (Clausius-Mossotti relation)
1.7 Polarizaiton of polar-molecules (Debye’s formula)
1.8 Electrostatic energy and energy density in free space and in dielectric
1.9 Let us sum up
1.10 Check your progress
1.11 Lesson-end activities
1.12 Points for discussion
1.13 References
1.0 Aims and Objectives
This lesson deals with potential and fields due to electric dipole. The relation
between electric susceptibility, polarization, displacement will be obtained. The molecular
field, derivation of claussion mossotti relation for non-polar molecules, Debge formula for
polar molecules are explained in detail. The derivation of electrostatic energy and energy
density who has has discussed.
1.1(a) ELECTRIC POTENTIAL AT A POINT DUE TO AN ELETRIC DIPOLE:
Two charges –q at A and +q at B separated by a small distance 2d constitutes an electric
dipole and its dipole moment is p (Fig 1.1a).
Let r1 and r2 be the distances of the point P from +q and –q charges respectively. Let P
be the point at a distance r from the midpoint of the dipole O and q be the angle between PO
and the axis of the dipole OB.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 2
P a g e | 2
Fig 1.1(a)
Potential at P due to charge (+q) = 104
1
r
q
pe
Potential at P due to charge (-q) = ÷÷ø
öççè
æ-
204
1
r
q
pe
Total Potential at P due to dipole is, V = 2010
4
1
4
1
r
q
r
q
pepe-
÷÷ø
öççè
æ-=
210
11
4 rr
qV
pe (1)
Applying cosine law,
÷÷ø
öççè
æ+-=
-+=
2
2
22
1
222
1
cos21
cos2
r
d
rdrr
rddrr
q
q
Since d is very smaller than r, d2/r2 can be neglected.
2/1
1
cos21 ÷
ø
öçè
æ-=\
rdrr
q
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 3
P a g e | 3
or
2/1
1
cos2
111
-
÷ø
öçè
æ-= q
r
d
rr
Using the binomial theorem and neglecting higher powers
÷ø
öçè
æ+= qcos1
11
1r
d
rr (2)
Similarly,
( )q
q
cos2
180cos2222
2
222
2
rddrr
rddrr
++=
--+=
2/1
2cos
21 ÷
ø
öçè
æ+= q
r
drr (Since d2 / r2 is negligible)
or
2/1
2cos
21
11-
÷ø
öçè
æ+= q
r
d
rr
Using the Binomial theorem and neglecting higher powers,
÷ø
öçè
æ-= qcos1
11
2r
d
rr (3)
Substituting equation (2) and (3) in equation (1) and simplifying
÷ø
öçè
æ+-+= qq
pecos1cos1
1
4 0 r
d
r
d
r
qV
2
0
2
0
cos.
4
1
.4
cos2
r
p
r
dqV
q
pepe
q== (4)
Special cases:
1. When the point P lies on the axial line of the dipole on the side of +q, then q = 0.
\ V = p / (4pe0r2)
2. When the point P lies on the axial line of the dipole on the side of –q, then q = 180.
V = -p / (4pe0r2)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 4
P a g e | 4
3. When the point P lies on the equatorial line of the dipole, then, q = 90,
V = 0.
1.1(b) Electric field at any point due to an Electric dipole
AB is an electric dipole of dipole moment p. O is its midpoint. R is a point with polar
coordinate (r, q) (Fig 1.1(b))
Fig. 1.1(b) Electric field at any point due to an Electric dipole
\ OR = r, LROX = q
The dipole moment p is resolved into two components along the horizontal and
vertical directions. The point R is on the axial line of a dipole of moment p cosq and on the
equatorial line of the dipole of moment p sinq.
Electric field at R along the axial line at a distance r from O
3
0
3
0
1
cos2
4
12
4
1
r
p
r
pE
q
pepe== (along RA) ---(1)
Electric field at R along the equatorial line at a distance r from O is
3
0
3
0
2
sin
4
1
4
1
r
p
r
pE
q
pepe== (along RB) ---(2)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 5
P a g e | 5
\ Magnitude of the resultant electric field is,
E = Ö (E12 + E2
2) = p / (4pe0r3) . Ö (4 cos2q + sin2q)
E = p / (4pe0r3) . Ö (1 + 3 cos2q) ---(3)
If f is the angle between E1 and resultant, then
qpeq
peqf tan
2
1
4cos2
4sintan
3
0
3
0 ==rp
rp ---(4)
This gives the direction of the resultant electric field.
The unit of potential difference is volt. The potential difference between two points is
1 volt if 1 joule of work is done in moving 1 Coulomb of charge from one point to another
against the electric force.
The electric potential in an electric field at a point is defined as the amount of work
done in moving a unit positive charge from infinity to that point against the electric forces.
Relation between electric field and potential
Let the small distance between A and B be dx. Work done in moving a unit positive
charge from A to B is dV = E.dx.
The work has to be done against the force of repulsion towards the charge +q. Hence,
dV = -E.dx
E = - (dV/dx)
The change of potential with distance is known as potential gradient, hence the
electric field is equal to the negative gradient of potential.
The negative sign indicates that the potential decreases in the direction of electric
field. The unit of electric intensity can also be expressed as Vm-1.
1.2 DIELECTRIC POLARIZATION In substances, called insulators or dielectrics, which do not have free electrons, or the
number of such electrons is too low, the electrons are tightly bound to the atom. When
potential difference is applied to insulators no electric current flows, even then their
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 6
P a g e | 6
behaviour in fields is very important because the presence of the field may change the
behaviour of an insulator. The insulators whose behaviour gets modified in the electric field
are called as dielectrics. When the change in the behaviour of the dielectric is independent of
the direction of applied field, the dielectric is called as isotropic. On the other hand, if the
change in behaviour of the dielectric depends on the direction of applied field then the
dielectric is called as anisotropic.
If we consider a dielectric in an electric field it exerts a force on each charged particle.
The positive particles are pushed in one direction (direction of field) while the negative
particles in the opposite direction. As a result the positive and negative parts of each molecule
are displaced from their equilibrium positions in opposite directions. The overall effect is a
displacement of the entire positive charge in the dielectric relative to the negative charge. The
relative displacement of the charges is called polarization, and the dielectric is said to be
polarized.
The molecules of dielectrics are classified into two classes:
(1) polar molecules, and
(2) non-polar molecules.
The positive charge of the nucleus may be supposed to be concentrated at a point, say
the centre of gravity of positive charge. Similarly, the negative charge due to electrons in
orbit may be supposed to act at a point, known as centre of gravity of negative charge. When
two centres of gravity coincide in a molecule, the molecule as a whole possesses no resultant
charge and it is said to be non-polar, e.g., carbon tera-chloride. If two centres of gravity are
displaced from each other, the molecule as a whole possesses polarity and has permanent
electric moment and the molecule is said to be a polar molecule, e.g., CHCL3 etc. A polar
molecule in an electric field exhibits both permanent and induced dipole moments while non-
polar exhibits induced dipole moment only.
When either of these types of molecules is placed in an external field, the small
displacements of the orbital electrons will cause the distance between the centres of gravity to
change. As a result non-polar molecules become induced dipoles whereas polar molecules
will have both permanent and induced dipoles. The orientation of the induced dipoles or of
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 7
P a g e | 7
permanent dipoles in an external electric field will be such as to align the dipole along the
field lines. Thus resultant dipole moment even in case of polar molecules gets modified.
Electric polarization effects in simple models of non-polar and polar dielectric materials are
shown in Fig 1.2
Fig 1.2 Electric polarization effect in simple modes of non-polar and polar
dielectric materials (a) Non-polar (b) Polar.
Definition of Dielectric Polarisation: In a homogeneous isotropic dielectric, the
polarisation (i.e., reorientation of electronic orbits) depends directly on the total electric fields
E in the medium. All the molecules are polarized to the same extent. The polarisation charges
within the body of the medium neutralize each other except at the surface of end faces
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 8
P a g e | 8
A
B
C
D
E
G
F
H
E
-σP +σP
(ABCD and EFGH Fig 1.2(a). The electric field due to induced charges is opposite to the
applied (external) electric field E0. Thus resultant field E is less than the applied field E0.
Now the average dipole moment p induced in each molecule of the dielectric is proportional
to E provided E is not too great i.e.,
p µ E or p = aE ---(1)
where a is a constant called molecular polarizability and E is the intensity of field. If there
are n molecules per unit volume in a dielectric and p of each molecule be independent, we
have electric moment per unit volume = naE, and is represented by a vector P, given by
P = n a E ---(2)
Electric moment per unit volume is known as polarization.
Consider a rectangular block of dielectric material of length l [Fig 1.2a]. Let a be the
area of cross-section of the block perpendicular to the direction of the field (face ABCD or
EFGH). Let the surface density of charge appearing on these faces be sp due to polarization.
If l be the length of the block, then the charge induced on each face = spa
Therefore total induced electric dipole moment
=spa.l………..(3)
Fig.1.2(a): Electric moment Calculation
Again the total induced dipole moment = P. a l
Because P is the polarization i.e., electric dipole moment per unit volume and volume of the
block is a l.
From equations (3) and (4) we have
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 9
P a g e | 9
sp.al = P. a l
Or P =sp
Thus the polarization may also be defined as the surface density of charge appearing at faces
perpendicular to the direction of applied field.
Electrical Susceptibility:
It has been found experimentally that in a large number of dielectrics, polarization P
is proportional to the applied electric field E, atleast for field strengths that are not too large.
Thus,
POOE
= ke0E,
Where k is dimensionless constant called electrical susceptibility of the dielectric and e0 the
permittivity of free space, has been included to make k dimensionless. Equation is true when
the polarization properties of medium are independent of direction of polarization which is
the case for liquids, gases, amorphous solids and cubic crystals.
1.3 RELATIVE PERMITTIVITY (DIELECTRIC CONSTANT) AND
DISPLACEMENT VECTOR
The capacity of a parallel plate capacitor increases if the gap between the plates is
filled with a dielectric. The increase in capacity of capacitor with a dielectric compared to
capacity of the same capacitor with vacuum between plates, measures a constant known as
dielectric constant, relative permittivity or specific inductive capacity (S.I.C.), i.e.,
,0e
ee ==
vacuumwithcapacity
presentdielectricwithcapacityr
where e and e0 are absolute permittivities of the dielectric and vacuum.
Or e = er . eo
is absolute permittivity of dielectric.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 10
P a g e | 10
The dielectric constant is defined in terms of electric intensity as follows: When a
dielectric is polarized, the charges appearing on the surfaces perpendicular to the field are
opposite to the real charges present on the plates. Hence the field decreases in presence of
dielectric. Therefore relative permittivity is also defined as the ratio of two fields; that is
E
E
dielectricinfield
vacuuminfieldr
0==e
e
e
pe
pe 0
2
2
0
4
4==
rq
rq
because 3
0
0.
4
1
r
qrE
pe= and 3
.4
1
r
qrE
pe=
Thus the dielectric constant may defined as ratio of the absolute permitivity of the dielectric
to that of the free space.
Since the electric intensity depends on the medium, we define another electrical
vector which depends only on the magnitude of the charge and its distribution but
independent of the nature of medium. This vector is termed as displacement, D, and is given
by
3.
4
1
r
qrD
p=
Er
qrD e
pee =÷
ø
öçè
æ=
3.
4
1
3.
4
1
r
qrE
pe=
vector D is analogous to E. Its unit is coul. /m2.
While E is defined as force experienced by a positive charge in any medium, the product
D.dS gives the flux of normal displacement through surface dS and remains unaltered in any
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 11
P a g e | 11
medium. The direction of the two vectors being the same in isotropic homogeneous
dielectric.
1.4 EXTERNAL FIELD OF A DIELECTRIC MEDIUM
In the presence of an electric field, a dielectric becomes polarized and aquires a dipole
moment P, per unit volume. The calculation of electric field intently produced by elementary
dipoles though out the space at an external point will be presented below.
Consider a finite piece of polarized dielectric. Let it be characterized at each point r
by a polarization P(r). The polarization gives rise to an electric field. The aim is to calculate
the potential and field at point r[denoted by Q(r’) in Fig.1.4] outside the dielectric. Let the
volume of the dielectric is dividd into small elementary volumes and
Fig.1.4 Dielectric body with external point of observations.
consider one such volume element dV of the dielectric.
The potential due to dielectric body occupying volume dV as observed
outside dV for single dipole potential can be expressed as,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 12
P a g e | 12
( ) ( ) ( )( )ò
-
-=
VdV
rr
rrrPr
3
0 '
'.
4
1'
pef
( )( ) ( )
( )( )ò
-
-
-
=V rr
dVrr
rrrP
r2
0 '
'
'.
4
1'
pef
( ) ( ) ( )( )ò
-
-=
VdV
rr
rrrPr
2
0
0 '
'.
4
1'
pef
where (r-r’)0 represents unit vector directed from Q towards dV. It is obviously equal
to
( ) ( ) ( )
( ) ( ) ( ){ }222'''
'''
zzyyxx
kzzjyyixx
-+-+-
-+-+-
where x’, y’,z’ are the Cartesian coordinates of point r’ and (x,y,z)those at r. Further,
( ) ( ) ( ) ( )úû
ùêë
é-
¶
¶+-
¶
¶+-
¶
¶-=- ''''
0rr
xkrr
xjrr
xirr
and ( ) ( ) ( ) ( )zyxkPzyxjPzyxiPrP zyx ...... ++=
substituing these values in equation (1), we get
( )( )
( )( )
( )( )
( )þýü
îíì
-¶
¶
-+-
¶
¶
-+-
¶
¶
-= ò òò V V
zy
V
x dVrrzrr
PdVrr
yrr
PdVrr
xrr
Pr '.
''.
''.
'4
1'
222
0pef (2)
321fff ++=
Now we shall consider the integrals separately. It is obvious that
( )( )
( ) 22
'
1'.
'
1
rrxx
rr
rr -¶
¶=
¶
-¶
--
and dxdydzdV =
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 13
P a g e | 13
Fig.1.4 (a) Integration with respect to ‘x’
Therefore
( )òòò-¶
¶= dxdydz
rrxPx 2
0
1
'
1
4
1
pef
Let us first integrate with respect to x. If we consider integration along x-axis only, it would
be simply a contribution of a straight rod of material with area of cross-section dydz. If such
an elements cuts the surface of the body at x1 and x2, coordinates of respective points, then x
integral would be
( )( )
( )( )
( ) ( )dx
x
P
rrrr
zyxp
rr
zyxpdx
rrxP x
x
xxx
X
xx
x
x
X
x¶
¶
--
úúû
ù
êêë
é
--
úúû
ù
êêë
é
-=
-¶
¶òò
==
2
112
2
2
1
2
'
1
'
,,
'
,,
'
1
( ) ( ) ( ) ïþ
ïý
ü
ïî
ïí
ì
¶
¶
--
úú
û
ù
êê
ë
é
--
úú
û
ù
êê
ë
é
-= òòòòòòò
==
dxdydzx
P
rrdydz
rr
Pdydz
rr
P x
xx
x
xx
x
222
0
1
'
1
''4
1
12
pef
if n is unit outward normal at dA then
dy dz = n.dA2 = nxdA2
and dy dz = n.dA1 = nxdA1
since at x1 which is on the part S1, the x-component of n, i.e., nx has negative value.
dydz
x2 x1
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 14
P a g e | 14
Now the integral of Φ1 can be written as
( ) ( ) ( ) ïþ
ïýü
ïî
ïíì
¶
¶
--
-+
-= ò òò
12 '
1
''4
112
0
1S
x
V
xx
S
xx dVx
P
rrdA
rr
nPdA
rr
nP
pef
Since the two integrals over S1 and S2 cover whole boundary S, we have
( ) ( )dV
x
P
rrdA
rr
nP x
VS
xx
¶
¶
--
-= òò '
1
4
1
'4
1
00
1pepe
f
similar integrals for Φ2 and Φ3 can be written and hence value of Φ will be
( )( ) ( ) ïþ
ïýü
ïî
ïíì
÷÷ø
öççè
æ
¶
¶+
¶
¶+
¶
¶
--
-
++= òò dV
z
P
y
P
x
P
rrdA
rr
nPnPnPr zyx
VS
zzyyxx
'
1
'4
1'
0pef
or ( )( )
( )( )òò -
--
-=
VS rr
dVdivPdA
rr
nr
'4
1
'
.
4
1'
00pepe
fP
(3)
Equation (3) obviously shows that the potential at point r’ due to dielectric body is the same as if it
were replaced by a system of bound charges in empty space. A part of these bound charges appears
on the dielectric surfaces as a surface density σP’ given by
σP’=P.n
while the remaining bound charge appears throughout the volume V as a volume density rp’ given
as
÷÷ø
öççè
æ
¶
¶+
¶
¶+
¶
¶-=
z
P
y
P
x
Pzyx
P
'r
in the case of homogeneous and isotropic dielectrics, where the polarisation of each molecule is to
the same extent, the volume distribution of polarized charge is zero because the positive and
negative charges of dipoles neutralize each other. Hence the second term does not contribute
anything to the potential, i.e., -div P = 0. The potential is due to the surface distribution of charges
only. The intensity of electric field due to the polarized volume may be expressed as
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 15
P a g e | 15
( )( ) ( )
úú
û
ù
êê
ë
é
-
-+
-
-= ò ò
S V
PP dVrr
rrdA
rr
rrrE
3
'
3
'
0 '
'
'
'
4
1'
rs
pe
this gives the outer field due to polarisation.
The total bound charge of a dielectric body of volume V and complete boundary surfaces S
is zero as the total charge is
ò ò ¢+¢=¢S V
total dVdAq rs
òò =-=VS
divPdVndAP 0.
since by divergence theorem of vector algebra,
òò =VS
dVPdivndAP.
Equation (3) is in M.K.S system units. If it is to be expressed in e.s.u then
( )( ) ( )( )
ò ò-
--
-= dVPdiv
rrdA
rr
nPr
rr '00 '
11
'
.1'
eef
1.5 RELATION BETWEEN ELECTRIC DISPLACEMENT, ELECTRIC FIELD AND
POLARISATION (D, E and P)
(1) Electric Intensity: The electric intensity E at a point in the electric field is
numerically equal to the force experienced by a unit positive charge placed at that point. The
direction of E being the same as that of the field.
(2) Dielectric Polarisation: When a dielectric is placed between the two charged
metallic plates, the dielectric is polarized. The distorted atom is called as an electric dipole or
an electric doublet. The electric dipole moment per unit volume is called the dielectric
polarisation P. It is a vector parallel to the electric field.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 16
P a g e | 16
(3) Electric Displacement: When no electric field is present, Gauss’ law gives
0
0.e
qAEdSE ==ò
A
qE
0
0e
=
where E0 is the electric field and q is the free charge.
When dielectric is present, induced surface charges say-q’ appears, so that Gaussian
surface encloses (q-q’) charge . Gaussian’s law then becomes
0
.e
qqAEdSE
¢-==ò
A
q
A
qE
00 ee
¢-=
Now the field when dielectric is present is given by
A
qEE
rr 0
0
eee==
A
q
A
q
A
q
r 000 eeee
¢-=
A
q
A
q
A
q
r 000 eeee
¢+=
A
q
A
q
A
q
r
¢-÷
÷ø
öççè
æ=
0
0ee
e
PEA
q+= 0e
As the quantity on left hand side is termed as electric displacement D, we write
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 17
P a g e | 17
PED +=0
e
It can be noted that D is connected with free charge, q only. P is connected with
polarisation charges, q¢ only and E is connected with all charges that are actually present
whether free or polarised.
The relationship between D, E and P can also be obtained as follows:
The differential form of Gauss law is
0
.e
r tE =Ñr
(1)
where rt is the total charge density. For dielectrics, rt must include charge densities of real
and polarized charges, that is,
rt = r + r¢,
where rt is volume density of real charge and r¢ that of polarized charges.
Therefore eq. (1) becomes
0
.e
rr ¢-=Ñ E
r
00
.e
r
e
r=
¢Ñ Er
(2)
We have shown that P.Ñ-=¢r
r which when substituted in eq.(2) gives
00
.e
r
e=
Ñ¢+Ñ¢
PE
( ) PPE =+Ñ 0. er
r=Ñ D.r
PED += 0e (3)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 18
P a g e | 18
and defines electric displacement vector D. Eq. (3) shows that D is responsible for the partial
field due to real charges, and that D and E are additive. This result is called differential form
of Gauss’ theorem in a dielectric. The curl of the displacement vector, on the other hand, can
be written as
( )PED +´Ñ=´Ñ0
err
(4)
PE ´Ñ+´Ñ=rr
0e (5)
in case of electric field of solenoidal type
E´Ñr
=0
in case of polarisation of isotropic homogeneous dielectric
0=´Ñ Pr
so that 0=´Ñ Dr
in such cases. But curl of displacement vector D does not vanish in the case of non-solenoidal
type of fields and anisotropic dielectric.
Further relation(4) can be put in the form,
PED += 0e
÷÷ø
öççè
æ+=
0
0e
eP
E
÷÷ø
öççè
æ+=
0
0
0e
ee
PkE
( )kE += 10e (6)
Eree 0=
where ( )rk e=+1
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 19
P a g e | 19
is the relative permitivity of medium defining dielectric constant by 0e
ee =
r , so that
D = ε E (7)
Now P is related to E by the relation
P = D- ε0E = ε0 εr E - ε0E
= ε0 (εr-1)E (8)
A relation can also be developed between r and r’ as follows:
P.Ñ-=¢r
r
( )Er 1. 0 -Ñ-= eer
÷÷ø
öççè
æ-Ñ-=
r
rE
eee
11.
0
r
( ) ÷÷ø
öççè
æ-Ñ=
r
De
11.
r by eq.(5)
( )Dr
.1
1 Ñ÷÷ø
öççè
æ--=
r
e
÷÷ø
öççè
æ--=
rer
11
As er > 1, r¢ < -r that is bound charge density is always opposite in sign and less in
magnitude than free charge density.
1.6 MOLECULAR FIELD IN A DIELECTRIC (CLAUSIUS-MOSSOTTI RELATION)
The aim of this section is to examine the molecular nature of the dielectric and to
consider how the electric field & responsible for polarizing the molecule is related to the
macroscopic electric filed. The electric field which is responsible for polarizing a molecule of
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 20
P a g e | 20
the dielectric is called the molecular field. This is the electric field at a molecular field or
local field or local field is produced by all external sources and by all polarized molecules in
the dielectric, except one molecule under consideration.
Fig.1.6(a) Fig.1.6(b)
The molecular field, Ein, may be calculated in the following way. Let us cut a
spherical cavity of radius r (such that its dimensions are very great as compared to the
molecular dimensions and very small as compared to the volume of the dielectric)
surrounding the point at which the molecular field is to be computed. The dielectric which is
now left, will be treated as a continuum. The cavity is put in its original position in the
dielectric (molecule by molecule) except the molecule where the molecular field is to be
computed. The molecules which are just replaced in the cavity are treated as individual
dipoles and not as a continuum.
Let the dielectric be placed in the uniform electric field between two parallel plates
(of a condenser) as shown in fig (1.6b). The dotted lines show the boundary of the dielectric.
Let the surface density of real charges on the capacitor plates be s. Again let the surface of
cavity has polarized charges of surface density sp.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 21
P a g e | 21
The field experienced by the molecule of the dielectric at the centre of the cavity C,
Ein, given by
Ein = E1 + E2 + E3 + E4
Where
(i) E1 is the field between two plates with no dielectric, so that E1 = s / e0,
(ii) E2 is the field at C due to polarized charges on the plane surfaces of the
dielectric facing the capacitor plates and is given by E2 = sp / e0.
(iii) E3 is the field at C due to polarized charges on the surface of cavity, to be
calculated.
(iv) E4 is the field at C due to permanent dipoles. But in present case for non-polar
isotropic dielectrics E4 = 0.
Thus, 3
00
EEp
in +-=e
s
e
s (1)
Evaluation of E3 : Consider a small elemental area ds on the surface of cavity of an
angular width dq and an angle q with the direction field E. The vector P shows the direction
of displacement at the centre of ds, Fig (1.6a). The normal component of displacement is
PN = P cos q.
By definition of polarization, it is the surface charge per unit area. Such a charge on ds to
provide flux normal to ds is,
PN = P cos q ds
and the electric intensity at C due to this charge is given by,
2
04
cos
r
dsP
pe
q=
where r is the radius of cavity. The field is directed along the radius CA. Resolving the
intensity along and perpendicular to the applied field, we have the components
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 22
P a g e | 22
qpe
qcos
4
cos2
0r
dsP= along the field
qpe
qsin
4
cos2
0r
dsP= perpendicular to field
If the area ds be taken round through 2p radians about LM, it will describe a ring, the
surface area of which is given by,
qqp drr ..sin2=
qqp dr sin2 2=
Therefore intensity at C due to ring in the field direction is
qqppe
qdr
r
dsPsin2
4
cos 2
2
0
2
=
qqqe
dr
Psincos
22
2
0
=
while the normal components of intensity due to the ring cancel each other.
Integrating the intensity at C due to charges on the surface of cavity, we have
3
2
2 0
´=e
P
therefore the total intensity or Ein at C is given by eq.(1) as
000 3ee
s
e
s PE p
in +-= (2)
The resultant field, E between the plates is
ò=p
qqqe 0
2
0
3 sincos2
dP
E
03e
P=
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 23
P a g e | 23
00e
s
e
s p-=
Therefore 0
3e
PEEin +=
Further we know that PEED +== 0ee
So that 0ee -
=P
E
Putting this value in eq.(2) for Ein, we get
00 3eee
PPE
in+
-=
÷÷ø
öççè
æ
-
+=
0
0
0
2
3 ee
ee
e
P (3)
if the number of the molecules per unit volume , a molecular polarizability then polarization,
P, is defined as electric moment per unit volume . That is
inEnP a=
Putting this value in eq,(3), we get
÷÷ø
öççè
æ
-
+=
0
0
0
2
3 ee
ee
ea
P
n
P
÷÷ø
öççè
æ
+
-=
0
0
0 23 ee
ee
e
an
2
1
0
0
+
-=
ee
ee
2
1
3 0 +
-=
r
rn
e
e
e
a
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 24
P a g e | 24
where er is the relative permittivity or dielectric constant. Equation (6) is well known
Claussius-Mossotti relation.
We know that er = mg2 where mg is the refractive index of gas. Hence equation (6) can
be expressed in terms of refractive index. The relation in that case is
2
1
3 2
2
0 -
-=
g
gn
m
m
e
a (4)
called Lorentz formula and is valid only as long as er is frequency independent.
Validity of Claussius-Mossotti Relation:
The number of molecules per unit volume is proportional to the density of gas. If
molecular polarizability a is taken to be a constant, the number n is proportional to 2
1
+
-
r
r
e
e.
Also the constant a is proportional to the cube of the radius of molecule. Hence finding e r
experimentally and calculating n from the density at definite temperature and pressure, the
value of the radius of the molecule of dielectric may be reckoned. The values, so obtained,
fairly agree with the values obtained by other methods and proves the validity of the relation.
It is true only in case of monatomic gases and weak solutions. The experimental and
theoretical values disagree in case of strong solutions and solids. It is due to the fact that in
these cases the interaction forces among molecules are sufficiently great, the account of
which has not been taken.
Molar Polarizability:
Multiplying Eq. (4) by M*/d on both sides, where M* is the molecular weight and d
the density of the dielectric, we have
d
nM
d
M
r
r
*
0
*
2
.3
.1
e
a
e
e=
-
+
But (M*n / d) = N and is called the Avogadro’s number. Therefore
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 25
P a g e | 25
d
MN
r
r
*
20
.1
3 +
-=
e
e
e
a
where Na is known as molar polarizability.
1.7 POLARIZAITON OF POLAR-MOLECULES,(DEBYE’S FORMULA):
Dependence of polarizability on temperature.
While discussing the polarization of non-polar molecules, it has been observed that
the molecular polarizability, a, does not depend on the temperature. But such is not the case
with polar molecules. If polar molecules are subjected to electric field then in addition to
induced polarization, the permanent polarization of molecules is also present. In absence of
electric field, the permanent dipoles point in random directions and the net polarization is
zero. On application of field, the permanent dipoles are forced to orient in the direction of the
field. But the orientations along the field direction are continuously disturbed by thermal
agitation. Therefore an equilibrium is attained and the dipoles make
Fig.1.7
zero to p angles with field direction. The polarization of such permanent molecules may
therefore be resolved in the direction of the field and the average value of Pp cos q may be
found (fig 1.7). Thus we have Pp the permanent polarization, and Pi the induced polarization.
Therefore total polarization is given by
PT = Pp + Pi
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 26
P a g e | 26
If Ein is the inner field acting on the dielectric molecules, the potential energy of
permanent dipoles per unit volume of moment Pp is given by
Ein . Pp = Ein . Pp cos q.
Calculation of the average value of Pp cos q:
Following the statistical method of Boltzmann used in kinetic theory of gases, the
probability that a molecule is found in an energy state W, at temperature T is proportional to
e-W/KT, where K = R/N is the Boltzmann constant, R and N are gas constant and the Avogadro
number respectively. In present case, the potential energy of dipoles per unit volume is given
by - Ein . Pp. Hence probability of finding dipoles at an angle q with the field direction is
given by
Pq µ eEin . Pp /KT
Or Pq µ eEin . Pp cos q /KT
The number of dipoles at an orientation will be proportional to its probability. Hence
if Nq is the number of electric dipoles per unit volume at an angle q and Np/2 at q/2, we have
Nq µ eEin . Pp cos q /KT
Np/2 µ 1
\ (Nq / Np/2) = eEin . Pp cos q /KT
or Nq = Np/2 eEin . Pp cos q /KT
Differentiating, we obtain the number of dipoles in an angular width dq as
qqq
pqqq dkT
PEeNdN
pinkTPE
d
pin sincos
2-=+®
The average moment per unit volume in the direction of field is given by
ò
ò=
p
q
p
qq
q
0
0
cos
cos
dN
dNP
Pp
p
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 27
P a g e | 27
Substituting the value of dNθ, we get
( )
( ) qq
qqqq
pq
p
pq
p
deNKT
EP
deNKT
EP
PkTPEinp
kTPEinp
p
pin
pin
ò
ò=
0
cos
2
0
cos
2
2
sin
sincoscos
( )( ) qq
qqqq
pq
pq
de
dePP
kTPE
kTPE
p
ppin
pin
ò
ò=
0
cos
0
cos
sin
sincoscos
KT
EPu inp=
qq
qqqq
pq
pq
du
duPP
p
p
ò
ò=
0
cos
0
cos
sin
sincoscos
qcosuKT
PEt in ==
qq dudt sin=
we get
( )
ò
ò
-
-=u
u
t
u
u
t
p
p
dte
dttuePP qcos
[ ][ ]u
u
t
u
u
tt
p
eu
eteP
-
--=
( )úû
ùêë
é
-
--
+
+=
-
-
-
-
uu
uu
uu
uup
ee
ee
ee
eeu
u
P
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 28
P a g e | 28
÷ø
öçè
æ-=
uuP
p
1coth
÷÷
ø
ö
çç
è
æ-=
inp
inp
pEP
KT
KT
EPP coth
which is called Langevin equation
when the inner field, Ein , is small and temperature T is large
i.e., 1ááKT
EP inp
then we can expand the exponentials in Langevin equation as
÷÷ø
öççè
æ-=
uu
uPP pp
1
sinh
coshcosq
úúúú
û
ù
êêêê
ë
é
-
÷÷ø
öççè
æ++
ççè
æ++
=uu
u
uu
Pp
1
.........!3
..........!4!2
1
3
42
úúû
ù
êêë
é-÷÷
ø
öççè
æ++÷÷
ø
öççè
æ++=
-
u
uu
uPp
1....
!3....
!21
132
3.
2u
u
Pp=
3
uPp
=
putting KT
EPu inp=
we get average polarisation of permanent molecules in the direction of the field as
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 29
P a g e | 29
KT
EPP inp
p3
cos2
=q
qcosPPPiT+=
KT
EPPP inp
iT3
2
+=
inii EP a=
KT
EPEP inp
iniT3
2
+= a
in
p
i EKT
P÷÷ø
öççè
æ+=
3
2
a
inT Ea=
where Ta is the total polarizability per unit volume is
KT
Pp
iT3
2
+= aa
where ia is induced polarizability per unit volume and
KT
Pp
3
2
is permanent polarizability per
unit volume.
Study of Molecular Structure and Debye’s Relation:
The Claussius-Mossotti relation, modified for polar molecules, shows that the
constant a is temperature dependent. The modification was utilized by Debye to study
molecular structure. The modified Claussius-Mossotti relation or Debye’s relation is
02 3
1
e
a
e
e T
r
r º-
+
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 30
P a g e | 30
úû
ùêë
é+º
-
+
i
p
r
r
KT
Pa
ee
e
33
112
02
00
2
3
1.
9 e
a
eip
TK
P+=
2
1
+
-=
r
rye
e
K
Pm
p
0
2
9e=
Tx
1=
03e
aic =
For non-polar molecules Pp = 0 and therefore m = 0, ie., the straight line is parallel to
x-axis. Therefore, if 2
1
+
-
r
r
e
e is plotted as ordinate against 1/T as abscissae, a straight line
parallel to x-axis is obtained.
The plot of 2
1
+
-
r
r
e
e against 1/T for CHCl3, the chloroform, and CCl4, the carbon
tetrachloride, is shown in fig 1.6a. We interpret that:
(i) The line for CCl4 is parallel to x-axis which shows that its molecules are
non-polar. It implies that the chlorine atoms in CCl4 are symmetrically
placed with respect to carbon atom, so that the center of gravity of negative
charges coincides with that of the positive charges and no permanent
dipoles are present.
(ii) The straight line for CHCl3 is inclined and shows the presence of polar
molecules. The polarity appears due to the charge asymmetry of the
molecule. The permanent dipole moment per unit volume is calculated
from the slope of the line. The dipole moment is then calculated on
dividing it by the number of dipoles per unit volume n. Constant ai is
determined from the intercept on y- axis.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 31
P a g e | 31
Fig.1.6(a)
The study has been extended to a great number of molecules of monatomic gases like
Argon and molecules like H2 which are made up of two similar atoms. They do not
exhibit charge asymmetry and are therefore non-polar. But for molecules made up of
different atoms the charge asymmetry comes in and the molecules act like permanent
dipoles. The examples are NH3, H2O etc.
1.8 ELECTROSTATIC ENERGY AND ENERGY DENSITY IN FREE SPACE AND
IN DIELECTRIC
We shall first calculate the potential energy of a group of n point charges. It will be
equal to the work done in assembling the charges, bringing in one at a time.
If in the field of a stationary charge q1, another charge q2 is brought from infinity to a
distance 2112 rrr -= then work done is
12
21
0
24
1
r
qqW
pe=D (1)
=If a third charge q3 is brought from infinity to a point distant r13 and r23 from charges q1 and
q2 respectively then work done is
23
21
013
21
0 4
1
4
1
r
qq
r
qqW
pepe+=D (2)
From eqs. (1) and (2), the total work done in assembling the three charges is
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 32
P a g e | 32
32 WWW D+D=D
úû
ùêë
é++=
23
21
13
21
12
21
04
1
r
qq
r
qq
r
qq
pe
which can be extended for an assembly of n charges as
úúû
ù
êêë
é+++++= ...............
4
1
23
32
13
31
12
21
0 ij
ji
r
qq
r
qq
r
qq
r
qq
pe (3)
ij
jin
j
n
i r
qqåå ==
=1
01 4
1
2
1
pe
where the prime on the second summation implies that the terms i = j are to be excluded from
the sum. The factor 2
1 has been introduced because of the fact that each pair of charges
occurs twice in the summation.
Equation (3) then represents the total electrostatic energy U of the assembly of n charges and
if we put
i
n
jj
i
q
qj
pe=å =1
04
1
In eqn. (3), we get
i
n
i iqU jå ==
12
1
If the point charges have been assembled in a linear dielectric medium, instead of vacuum or
free space, then free space permittivity e0 is to be replaced by absolute permittivity e = e0 er ,
where er is the relative permittivity or dielectric constant of the medium.
Arbitrary Charge Distribution:
The expression for the field energy, U can be expressed with much advantage in terms of
volume and surface integrals involving the useful quantities electric field E and displacement
D when there is arbitrary charge distribution.
Let us consider a system of arbitrary charge distribution with density of initial charge
distribution asp. Let us remove the charge to infinity so that at any time the charge density is
ar where a is a parameter with values is ranging from unity to zero. At a point, where
density is ar, the charge dq’ in a volume element ‘dv’ will be
dq’ = (ar) dv (1)
and the potential f’ = af (2)
because f µ r and f is the initial potential at that point.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 33
P a g e | 33
As the charge is removed to infinity, charge density at every point in the distribution will fall,
say, from ar to (a - da) r so that decrease in the charge contained in the volume element dv
will be
dq” = (da) r dv (3)
And the energy withdrawn from the system in removing this charge from volume element dv
is
dU = f’ dq” (4)
= (af) (da) r dv (5)
On using Eqs. (2) and (3).
If the whole charge from the system is removed to infinity i.e., the charge density is
reduced to zero everywhere then the total energy withdrawn, equal energy of the system, will
be
òò=v
dvdU fraa1
0
ò
ò
=
÷÷ø
öççè
æ=
v
v
dv
dvU
fr
fra
2
1
2
1
0
2
We know that r = div D
So that field energy is ò= dvDdivU f2
1
Further
div (fD) = f div D + D grad f
so that
dvDD
dvgradDDdivU
v
v
ò
ò
Ñ-Ñ=
-=
ff
ff
rr.)(.
2
1
.)(2
1
But E = - Ñr
f,
so that òò +Ñ=vv
dvEDdvDU .2
1)(.
2
1f
r
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 34
P a g e | 34
Changing the first volume integral into surface integral by Gauss divergence theorem, we get
òò +=vs
dvDEdSDU .2
1
2
1f (6)
as v can be any volume which includes all the charges in the system, we can choose the
bounding surface S at large distance from the charge distribution. At large distances
r
1af
2
1
rD a
2rdS a
so that r
dSDS
1. =òf
¥®® rwhen0
Eq. (6) is then left with
ò=v
dvDEV .2
1 (7)
predicting that the energy is distributed with a density ).(2
1DE per unit volume. Hence the
energy density i.e., the ener gy per unit volume in an electrostatic field is
).(2
1DEu = (8)
Eq. (7) holds for linear dielectrics.
Form linear dielectric,
D = e E
So that energy density is 2
2
1Eu e= (9)
In free space e = e0 (since er = 1), we have energy density of electrostatic field in free
space as given by 20
2
1Eu e= (10)
Eqs. (8) and (9) hold good for isotropic homogeneous dielectric in which D and E are
proportional to each other. But if the medium is anisotropic, the permittivity e has different
values in different directions and consequently e which relates D and E is a tensor e0b . The
relation can be written as
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 35
P a g e | 35
baba e ED =
babee Er )(0= (11)
where (er)ab is the relative permittivity or the dielectric constant of the medium. Putting eq.
(11) in eq. (7), we get
dvEEU r ababee )(2
10ò= (12)
The dielectric constant (er)ab is a symmetric tensor which can be judged from the integrand of
eq. (12),
abbebaababab eee EEEEEE rrr )()()( == (13)
where the indices of the last step step have been interchanged. Therefore,
[ ] 0)()( =-ò dvEEv
rr babaab ee (14)
which shows that for arbitrary values of Ea and Eb , the tensor
(er)ab = (er)ba (15)
except when field energy vanishes identically. Physically it means that it does not matter
whether the displacement Da is produced by field Ea, the value of (er)ab or (er)ban is the same
for the specified components.
When Charge Distribution is Discrete:
From eqs. (10), (8) and (7), we get for free space
dvEUv
20
2
1ò= e (16)
If the distribution is discrete then
E2 = E.E = (E1 + E2 + E3 + …...) . (E1 + E2 + E3 ……)
ji
n
j
n
i
n
j j EEE .11
1
2 ååå==
=+= (17)
The term for which i = j are included in first term. Prime stands for which ji ¹ . Putting eq.
(17) into eq. (16), we get
dvEEdvEUv
ji
n
j
n
iv
n
jj ò ååò å ÷
÷ø
öççè
æ+÷
÷ø
öççè
æ=
===
.22 11
0
1
20 ee (18)
Obviously for the first term the position of other charges is immaterial for any charge i.e., this
term represents the work done in the creation of the charges and is known as the self energy
of the system, U0. Thus
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 36
P a g e | 36
dvEEUUv
ji
n
j
n
iò åå ÷
÷ø
öççè
æ+=
==
.2 11
00
e (19)
The second term represents the work done by one charge in the field of others and so on i.e.,
it is the work done in bringing the charges from infinity to the space to constitute the given
distribution of charge.
1.9 Let us sum up
In this unit we described the terms potential and fields due to an electric
dipole.
Given the relation between electric susceptibility, Polarization, Displacement.
Explained the molecular field and the derivation of Mosotti relation for non-polar
molecules and Debye formula for polar molecules.
Finally the derivation for electrostatic energy and energy density has been done.
1.10 Lesson end activities:-
(i) What is electric dipole and dielectric polarization?
(ii) Define the terms Electric Susceptibility,Polarization(P) and
Displacement(D).
1.11 Points for Discussion
(iii) Prove D = e0 E + P
(iv) Explain Gauss law in Dielectrics
(v) What are dielectrics under what conditions do they differ from conductors?
Is dielectric an insulator? Explain.
1.12 Check your progress
(vi) Derive Classius-Mosotti relation for Non-Polar molecules.
(vii) Derive Langevin Debye formulae for Polar molecules.
1.13 Sources/References
1. Electromagnetic Theory by K.K Chopra and G.C. Agrawal.
2. Electricity and Magnetism by R.Murugesan.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 37
P a g e | 37
UNIT – I Lesson – 2
2.0 aims and objectives
This lesson dealer with the discussion about Bio and savant law Lorenz force law, ampere’s
circuital law and its application in magne to statics. The divergence and civil of the magnetic
Induction B and the derivation for the scalar and vector potentiah is abo discussed.
2.1 Magnetic Induction
let us consider a +ve test chare q0, with a velocity v through a point
P. If this charge experiences a side-way deflecting force F, then a magnetic
field is said to exist at that point. This field is defined by means of a vector
quantity B and is called magnetic Induction, shown in Y axis.
If a charge moving through a point P in a magnetic field experiences
no side way deflecting force then the direction of motion of the charge is
defined as the direction of B. Conditions are
(1) when V is parallel to B (in the same direction) F is minimum. (2) when V
is perpendicular to B, F is maximum.
Definition: If a +ve test charge q0 moving with velocity v through point P in
a magnetic field experiences a deflecting force F, then magnetic induction B
at P is defined by
F = q0 V x B ---------(1)
The above relation defines both the direction and magnitude of B is F = q0 V B sinθ
B = F / q0 V sinθ ; θ is angle between V x B.
From the above V is parallel to B, θ = 0 ; F = 0 (min) v is perpendicular
θ = 90 ; F = Bq0 V (max)
Units; Newton / amp –turn
Z F
P
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 38
P a g e | 38
B +q0 θ Y
V
X
2.2 Ampere’s force law (Force on current element)
The concept of magnetic field is introduced by considering a test
charge q moving in a region of space with velocity V. Suppose the charge
experiences a force F, then the region is said to be having magnetic field B,
we write
F = qV x B -----------(1)
The above equation in terms of current i.e. the current crossing a surface is
defined as the rate at which charge flows across the surface,
ie I = dq / dt ---------(2)
ie, the force experienced (dF) by the charge dq moving with velocity V then
Eq. (1) becomes
dF = dq VxB
= I dt VxB.
Suppose in the time dt, charge dq travels along the length ‘dl’ of the
conductor then
V = dl / dt ie., dF = I dt (dl /dt) x B
dF = I dl x B --------(3)
this is Ampere’s force law
The total force experienced by the total volume containing the charge can be
calculated by integrating the above equation
F = v0l ( J xB ) dv [ie, I = J.ds and ds.dl = V]
2.3 Ampere’s circuital law
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 39
P a g e | 39
The law states that the line integral of the magnetic induction vector B
around a closed path is equal to µ0 times the total current crossing any
surface bounded by the line integral path, ie
c B . dl = µ0 s J.ds = µ0 I ---------(1)
we know curl of the magnetic induction
x B = µ0 J ( the differential form of Amperes’
Circuital law) -----(2)
Integrating over an open surface S, bounded by a closed curve C we get
s ( x B).ds = µ0 s J.ds ------(3)
But according to Stoke's theorem surface integral can be changed to line
integral as,
s ( x B) . ds = c B dl ---------(4)
Substituting Eq.[4] in Eq.[3] we get
c B dl = µ0 s J.ds =MOI
µ0 I - which is Amperes Circuital law. ( s J.ds = I ) this is integral form of
Ampere’s circuital law.
2.4 Biot & Savart law – Magnetic Induction.
Ampere performed series of experiments to find the forces between two
current elements. He observed that force between two current elements dl 1
and dl 2 carrying steady currents I1 and I2 depends on
(1) It varies directly as the magnitude of each current (I1, I2 )
(2) It varies inversely as the square of the distance between the two circuit
elements (r² 21)
(3) It depends upon the lengths and orientation of the two current elements
(dl 1, dl 2 )
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 40
P a g e | 40
(4) It depends upon the nature of the medium
- +
- + dl
r r1
dl 2
Fig. 2.4 Two currents I1 and I2
The force exerted on current element dl 2 by current element dl 1 is given by
dF21 = (µ0 / 4 ) (I1, I2) ( 1 / r²21) [dl 2 x (dl 1 x r21 /r21) ]
(µ0 / 4 ) ---------- arises due to (4) -------nature of the media
I1, I2 --------------- due to (1)
(1 / r²21) ------------due to (2)
[dl 2 x ((dl 1 x r21 /r21) ] arise due to (3) r2 1 / r21
represents unit vector along r21
The above equation dF21 can be written for whole lengths of the conductors
as
F21 = (µ0 / 4 ) (I1, I2) 1 2 [dl 1 x (dl 1 x r2) ] / r3 21 -------(2)
The above equations not does have much practical value because of r321 The
force cannot expressed as the interaction of current I1 with field current I2 .
However the above equation. can be written as,
F21 = I1 1 dl 1 x [ µ0 / 4 I2 2 (dl 2 x r2 1 )/ r³2 1 ------(3)
F21 I2 dl 2 x B1
Where B1 =( µ0 / 4 ) 2 dl 2 x r2 1 / r³2 1 --------(4)
B1 is called the magnetic induction, magnetic flux density or magnetic field
current. Unit is web / m2 or Tesla. In general the magnetic induction B at a
position r due to a current carrying circuit of element I dl will be
B =(µ0 / 4 ) I dl x r / r3
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 41
P a g e | 41
This is Biot savart law
Point (1) If the current I is distributed in space with a current density J
then I dl = Jd
Here B = µ0 / 4 J x r / r3 d
Point (2) If a single charge q moving with velocity v then
B = µ0 / 4 q ( v x r / r3)
or
B =(µ0 0 v x q r) / 4 0r3
But µ0 0 = 1 / c2 and E = qr/4 0r3
B = v x E / c2
The above gives the relation between electric (E) and magnetic fields (B) of a
uniformally moving charge as v<<c.
2.5 THE DIVERAGENCE OF THE MAGNETIC INDUCTION B
We had demonstrated that the magnetic field of moving charges were
such that . B = 0. It is also possible to arrive at this same result for
steady currents starting from the Biot – Savart law. We know that,
B = (µ0 / 4 ) r' (J1 x r1) / r3 d ' ---------(1)
. B =(µ0 / 4 ) r' (Jf r1) / r2 * dr' = (µ0 / 4 ) r' . ( Jf x r1/r2) d '
. ( Jf x r1/r2) r 1 / r2 .( x Jf ) - Jf. ( x r1/r2)
Where the first term on the right is zero because Jf is a function of the
source point x' , y' , z' while the del operator involves derivatives with respect
to the field point x , y , z. The second term on the right is also zero because
i j k
( x r1)/r2 =( x r )/ r3 = / x / y / z 0
(x - x')/r3 (y - y')/r3 (z - z')/r3
Then
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 42
P a g e | 42
.B = 0 -------------(2)
This equation follows form the definition of B given in Eq.(1) we also know
that it is a consequence of Coulomb’s law and of the Lorentz transformation.
The fact thatÑ. . B is zero means that these cannot be sources of B.
The net flux of magnetic induction through any closed surfaces is
equal to zero since
B.da = r . B dr = 0 › B.da = r . Bdr=0 ----------(3)
2.6 The Vector Potential A (Magnetic Vector Potential)
The calculation of electric fields was much simplified by the
introduction of the electrostatic potential. For an electrostatic field, the
relation between electrostatic field E and electrostatic potential V is given by
E = - V
Here V is a scalar quantity
In the case of a magnetic field,
div B = 0
Since the divergence of any curl is zero, it is reasonable to assume that the
magnetic induction may be written as,
B = Curl A = x A
'A' refers to magnetic potential and is called the magnetic vector potential.
Therefore the magnetic vector potential A can be defined as the vector,
whose curl at any point gives the vale of the magnetic fields B at that point.
The only other requirement placed on A is that
x B = x [ x A] = µ0 J
The unit of A is Wb / m
Derivation of the magnetic vector potential of a current loop:-
According to the Biot-Savart law, the magnetic induction at a distance
r from the element of length l carrying a current I (Fig. 2.6) is given by
B = µ0I / 4 . l x r / r3 -------(1)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 43
P a g e | 43
dI dB
›
r p(x,y,z)
Fig. 2.6
We have (1 / r) = - r /r3 ----(2)
B =(µ0I / 4 ) l x { - (1/r)}
= (µ0I / 4 ) { x1/r) x l } ------(3)
We have the vector identity.
x ( A) = ( ) x A + ( x A)
Where is a scalar and A is a vector
Or ( ) x A = x ( A ) - ( x A)
So (1/r) x l = x ( l / r) – 1/r ( x l ) ------(4)
Using Eq. (4) in Eq.(3) we get
B = (µ0I / 4 ) [ x ( l / r) – 1/r ( x l ) -----(5)
In this equation, x l = 0 because the operator is a function of (x,y,z)
and the current element is not a function of (x,y,z) as shown in the Fig. 2.6
Then Eq. (5) reduces to
B = (µ0I / 4 ) x ( l / r) ----- (6)
Therefore the total magnetic induction at the given point by a closed loop
carrying current is given by
B = (µ0I / 4 ) x (dI/r) -------(7)
The operation x is independent of the integration of dI /r around the
closed loop.
The Eq.(7) can be rewritten as
B = x [µ0I / 4 dI/r ] = (curl µ0I / 4 ) dI/r ------(8)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 44
P a g e | 44
Thus, we conclude that a vector exists such that by taking its curl, the
magnetic induction produced at any point by a closed loop carrying current
may be obtained. This vector is known as magnetic vector potential
A. Thus
B = curl A ------(9)
Where magnetic vector potential A = (µ0I / 4 ) dI/r ---------(10)
Eq. (9) is frequently used to derive magnetic induction B at any point from
the magnetic vector potential A at that point.
If the current is flowing through the length element is distributed over
a cross-sectional area a, we write l = Ja. Eq (10) is thus written as
A = µ0 / 4 v (J /r) x dV ----------(11)
The vector potential defined by Eq. (11) is not uniquely defined. We find
that we can add any term, whose curl is zero to the vector potential and it
still gives the same magnetic field. Unlike V, A does not have a physical
significance. It serves as a convenient intermediate step for the computation
of B.
2.7 THE CURL OF THE MAGNETIC INDUCTION B
We have shown that the magnetic induction is always equal to the
curl of the vector potential : B = x A. We shall now show that
x B = µ0 Jf ---------(1)
assuming a steady state and the absence of magnetic materials. In terms of
A,
x B = x x A = ( .A) - 2 A ----(2)
We have already shown that . A is proportional to the time derivative of
the electric potential V, then, with the above assumptions.
. A = 0 -----------(3)
For the second term we have from the definition of A that
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 45
P a g e | 45
2 A = (µ0 / 4 ) ' 2 (Jf / r) x d ' ---------(4)
Where we have interchanged the order of differentiation and integration.
Let us imagine At the field point P (x,y,z) where we wish to compute 2 A,
we form the vector J d '/ r, where Jf of and d ' are respectively the current
density and the volume element at the source point P', and where r is the
distance form P' to P. We compute the Laplacian of this vector at P by
taking the appropriate derivatives with respect to the coordinates x, y, z of P.
We then sum the contributions from all such sources in the volume ' which
includes all points at which Jf exists. The volume ' may include the field
point P, where r = 0.
Figure 2.7 Source point P' and field point P for the calculation of 2 A
Since Jf is not a function of the coordinates of P, we can write the
integral as
2 A = (µ0 / 4 ) ' Jf 2 (1/r) d ' -------(5)
Now, by differentiation of
(1/r) = 1/[ (x-x')2 + (y - y')2 + (z - z')2 ]½ -----(6)
we find that 2 (1/r) = 0 if r 0. There can thus be no contribution to the
integral from any element d ' except possibly if P and P' coincide and r is
zero
To investigate the integral at r = 0 we consider a small volume
enclosing the point P, where we wish to calculate 2 A, situated inside the
current distribution.
We take the volume so small that Jf does not change appreciably
within it; Jf may then be removed from the integral:
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 46
P a g e | 46
2 A = (µ0 Jf / 4 ) ' › 0 2 (1/r) d ' -------(7)
The meaning of this integral is as follows. For each element of volume d '
centered at the point P' within ' we calculate
2 (1/r) = ( 2/ x2 + 2/ y 2 + 2/ z2) * 1 /[ (x-x')2 + (y - y')2 + (z - z')2 ]½ -----(8)
Multiply by d ' and sum the results. Since 2 (1/r) = ' 2 (1/r)
2 A = (µ0 Jf / 4 ) ' › 0 2 (1/r) d ' ---------(9)
= (µ0 Jf / 4 ) r' › 0 ' . ' (1/r) d ' ---------(10)
(µ0 Jf / 4 ) s' › 0 ' (1/r) . da --------(11)
from the divergence theorem,
2 A = -(µ0 Jf / 4 ) s' › 0 (r1 . da) / r2 –----(12)
where r1 is the unit vector from the source point to the field point. In this
case r1 points inward toward the point P, Thus
2 A = - (µ0 Jf / 4 ) s' › 0 d -----(13)
where d is the element of solid angle subtended at the point P by the
element of area da. Since the surface S' completely surrounds P,
2 A = - µ0 Jf -------(14)
and x B = µ0 Jf ------(15)
This result is again valid only for static fields and in the absence of
magnetic materials.
2.8 THE FORCE ON A POINT CHARGE MOVING IN A MAGNETIC FIELD
The force on a current element I dl is I dl x B. Now, if the cross
sectional area of the wire is da
I = n (da v )Q -------(1)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 47
P a g e | 47
Where n is the number of carriers per unit volume, v is their average drift
velocity, and Q is the charge on one carrier. The reason for this relation is
that the total charge flowing per second is the charge on the carriers that
are contained in the length v of the wire
Then the force on the element dl is
η da d Qv X B
And the force on a single charge Q moving at a velocity v in a field B is
Qv x B
This force is perpendicular both to the velocity v and to the local magnetic
induction B.
More generally if there is also an electric field E, the force is
Q[E + (vxB)
This is the Lorentz force.
2.9 Magnetic Scalar Potential
Consider a closed current loop carrying current I (Fig 2.9). Consider a
point P (r) having position vector r relative to current element I dI. From
Biot - Savart law, the magnetic induction B at P due to whole loop is
B =(µ0/ 4 ) I dI x r / r3
Let the point of observation P(r) be moved through an infinitesimal distance
dr say from P (r) to Q (r+dr). then,
B.dr = (µ0/ 4 ) (I dI) x r / r3 .dr
(µ0I/ 4 ) dr. (dI x r) / r3
µ0I/ 4 (dr x dI) . r / r3
When the point P is shifted to Q, the solid angle subtended by the loop at P
changes by d . But we can also get the same change in solid angle d
keeping P fixed and giving every point of the loop the same but opposite
displacement (-dx). Then, the above equation becomes
B.dr = - µ0I/ 4 (- dr x dI ) . r / r3
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 48
P a g e | 48
But, -dx * dI = dS = area traced out by current element dI during the
displacement (-dx)
Therefore B. dr = -µ0I/4 dS . r / r3
But, (dS . r) / r 3 = d = Change in solid angle subtended by current loop
when point P is displaced to Q
Therefore B.dr = - (µ0I/4 ) d ---------(1)
Since is a scalar function of (x,y,z) ›
d = .dr
Hence Eq. (1) becomes
B.dr = - (µ0I/4 ) . d
Or
B = - (µ0I/4 ) = - [ µ0I / 4 ] ----(2)
The direction of B is that of - , so that B points away from the loop along
its positive normal.
Comparing Eq.(2) with B = - Vm we get
Magnetic Scalar Potential Vm = (µ0I )/ 4
= µ0 /4 x current x solid angle ------(3)
Negative gradient of Vm gives the magnetic induction B.
Q (r + dr)
dr
P(r) d › dl ds
dr
Fig. 2.9
dr
I
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 49
P a g e | 49
2.10 Equivalance of a small current loop and a magneticdipole
(Application of magnetic vector potential)
Consider a small current loop of area S carrying current I (Fig. 2.10).
Let P be a point having position vector r relative to centre of loop. Then
magnetic scalar potential at P = Vm = (µ0I / 4 ) * .
Here = solid angle subtended by current loop at P
= (S.r / r3 ) = S cos θ / r2
therefore Vm = (µ0I / 4 ) S cos θ / r2 = (µ0 /4 ) * (IS) cos θ / r2 -----(1)
The expressions similar to that for potential due to electric dipole
V = (1/4 0) þ cos θ / r2 -------(2)
Thus magnetic field due to a small current loop is similar to electric field of
an dipole
The product (IS) is called magnetic dipole moment,
P
m
› r S θ
s
Fig. 2.10
m = IS ----(3)
Thus a current loop of area S carrying current I is equivalent to a magnetic
dipole of dipole moment m = IS.
s
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 50
P a g e | 50
2.11 ELECTRIC FIELD VECTOR IN TERMS OF SCALAR AND VECTOR
POTENTIAL
In the case of magnetic field,
div. B = 0 ------------(1)
Since the divergence of any curl is zero
B = curl A = x A ---------(2)
Here A is called the magnetic vector potential. It has physical importance
where magnetic field varies with time
The time variation on magnetic field, from Eq. (2) is
B / t = / t ( x A) ---------(3)
Interchanging the space and time operator, we get
B/ t = x A/ t -----------(4)
The differential form of Faraday’s law
x E = - B / t --------(5)
therefore Eq. (4) becomes x ( E+ A / t) = 0 -----(6)
From vector identity, we find that the curl free field must be gradient of a
scalar potential φ. We write
E + A / t = - grad φ ------(7)
or E = - A/ t – grad φ -------(8)
Eq. (8) suggests that for time dependent magnetic field, we may think of –
grad φ as the contribution to E due coulomb field and - A / t as the
contribution due to electromagnetic induction
2.12 Let us sum up
From this unit you have learnt the Biot-savart law, Lorentz force
law,Ampere’s circuital law and its applications in Magnetostatics. Also it dealt
with divergence and curl of B and the derivation for scalar and vector potentials.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 51
P a g e | 51
2.13 Lesson end Activities
(i) Mention the importance of Scalar and Vector potentials in Magnetostatics
(ii) State Biot-Savart law
(iii) From Biot-Savart law derive Ampere’s circuital law
(iv) Write the equation for Lorentz force.
2.14 Points for discussion
(i) From Biot-Savart law derive B=m0 H
(ii) Find the magnetic induction at the centre of a square current loop of side
I metre carrying a current of 1 ampere.( Ans. 8Ö2 * 10-7 wb/m2)
(iii) In the Bohr model of the hydrogen atom, the electron revolves round
the nucleus in a path of radius 5.29 * 10-11 m at a frequency of 6.58 *
1015 Hz.
What is the dipole moment and Magnetic induction?
[Hint: m=Ai ; B =m0 i / Za ; i= e n ]
[ Ans: 9.266 * 10 -24 Am2 ; 12.52 T]
2.15 Check your progress
(i) Derive curl and divergence of B
(ii) Derive the equations for scalar and vector potentials
2.16 Source/Reference
(i) Electricity and Magnetism by R.Murugesan
(ii) Electrodynamics by Gupta, Kumar and Singh
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 52
P a g e | 52
UNIT – II - Field Equations
Lesson - 3
3.0. Aim and Objectives: In this unit you study the meaning of
continuity equation, displacement current and also the derivation of
Maxwell’s equations in differential and integral form also you study about
the scalar (Φ) and vector potential (A) and its relation to Lornetz and
Coulomb gauge.
3.1 Equation of Continuity
According to the principle of conservation of charge the net amount of
charge in an isolated system remains constant. The principle can be stated
as follows.
In the net charge crossing a surface bounding a closed volume is not
zero, then the charge density within the volume must change with time in a
manner that the time rate of decrease of charge within the volume equals
the net rate of flow of charge out of the volume. This statement can be
expressed by the equation of continuity.
Derivation: Let us consider that charge density ρ, is a function of time. The
transport of charge constitutes the current i.e.,
I = dq /dt = d/dt v ρ. dV ------------ (1)
Here we have considered that the current is extended in space of volume V
closed by the surface 'S'. The net amount of charge which crosses a unit
area (normal to the direction of charge flow) of a surface in unit time is
defined as the current density J. We know, if a net amount of current is
flowing outward closed surface the charge contained within that volume
should decrease at the rate
-dq / dt = I ----------- (2)
Where I is the total current flowing through surface S. if J is the current
density, then by definition, total current I will le
I = s J. ds------- (3)
From equations (2) and (3) we get
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 53
P a g e | 53
s J. ds = - dq/dt
= -d/dt v ρ. dV ---------- (4)
{Using equ. (1) }
Because it is ρ which is changing with time, we can write
d/dt v ρ.dV = v ρ./ t.dV
so that equ. (4) becomes
s J.ds = - v ( ρ/ t). dV ----------(5)
From divergence theorem, we have
s J.ds = - v (div J) dV
so that equ (5) becomes
v (div J) dv = - v ( ρ/ t). dV
(or)
v (div J + ρ/ t)dV = 0 ------(6)
Since Eq. (6) holds for any arbitrary volume, we can put integral equal to zero. i.e.,
div J + ρ/ t = 0 --------(7)
It is referred to as the equation of continuity. It is the mathematical
expression for the conservation of charge. It states that the “ total current
flowing out of some volume must be equal to the rate of decrease of charge
within the volume, assuming that charge cannot the created or destroyed.
i.e., no sources and sinks are present in that volume”. In case of stationary
currents, charge density at any point within the region remain constant
i.e.., ρ/ t = 0
› So that div J=0 or . J = 0 which express the fact – that there is no net
outward flux of current density J.
3.2 Displacement current (D)
Maxwell changed the definition of total current density to adapt the
equation of continuity to time dependent fields
Ampere’s circuital law is
s B. dl = µ0I
s H. dρ = I = s J.ds
Changing line integral into surface integral, by stoke`s theorem,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 54
P a g e | 54
s curl H. ds = s J.ds (or)
curl H = J ----------- (1)
Let us substitute it in equation of continuity, then
div J = - ρ / t
we get div ( curl H) = - ρ/ t
0 = - ρ / t
Here equ (1) leads to steady state condition in which charge density is not
changing. Therefore for time dependent (changing) fields, Eq (1) should be
modified. Maxwell suggested that the definition of total current density is
incomplete and advised to add something to it. Let it be J' then Eq (1)
becomes
Curl H = (J+ J') ---------- (2)
In order to identify J', we take divergence of Eq (2) That is
div (curl H) = div(J+ J')
0=div J+div J'
(or)
div J' = -div J = ρ/ t -----------(3)
We know that ›
ρ. = . D So that Eq. (3) becomes
› div J' = / t = ( .D )
› = . D / t
= div ( D / t)
(or)
div [J' - ( D / t)]= 0 --------- (4)
Eq. (4) is true for any arbitrary volume, we can have
J' = ( D / t) -------(5)
Therefore the modified form of the ampere's law is
Curl H= J+( D / t) --------- (6)
Note: 1. Since J' arises due to the variation of electric displacement D
with time, it is termed as displacement current density . According to
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 55
P a g e | 55
Maxwell it is just as effective as J, the conduction current density in
producing magnetic field. (2) The important inference that we get from Eq.
(6) is that, since displacement current J' is related to the electrtic field vector
D (as D = εE ) it is not possible in case of time varying fields to deal
separately with electric and magnetic fields but, instead the two fields are
interlinked giving rise to electromagnetic fields. Thus J' results into
unification of electric and magnetic phenomenon.
3.3 The Maxwell’s equations ( Differential form)
The four equations of Maxwell’s are.,
› (i) .D = ρ ---- obtained by the application of Gauss theorem in
electrostatics. D is the electric displacement in coulomb / meter2 and ρ is
the free charge density in coulomb/meter3.
› (ii) . . B = 0; obtained by the application of Gauss theorem to magnetic
field and B is the magnetic induction in Weber/mt2
› (iii) .E = - B/ t; obtained by Faraday’s and Lenz`s law in
electromagnetic induction and E is the electric intensity in volt /
meter
› (iv) x H = J + D/ t : obtained by Maxwell’s modification of Ampere’s
law in a circuital form for magnetic field accompanying an electric
current and H is magnetic field intensity in ampere/meter and. J is
the current density in ampere/meter2 .
(A) Derivation of Maxwell’s equations
(i) div D= ρ
Consider a surface S bounding a volume V in a dielectric medium. From
Gauss theorem the integral E.ds of the normal component of E over any
closed surface is equal to the total charge enclosed within the surface. Also
we know that the total charge must include both the free and the
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 56
P a g e | 56
polarisation charges or the total charge density ρp = -div p and ρ is the free
charge density at a point in a small volume element dV. Thus total charge
density at that point will be, ρ - (divp) then Gauss law can be expressed as
s E.ds = v div E.dV= 1/ 0 (ρ-div p) dV
(or)
div ( 0E+P)dV = ρdV
the quantity ( 0E+P) is D called electric displacement, so that
div D dV = ρdV
(or)
(div D - ρ)dV = 0
Since this equation is true for all volume, the integrand in this equation
must vanish i.e., Div D= ρ
when the medium is isotropic the three vectors D,E, P are in the same
direction and for small field, D is proportional to E, that is
D = εE
Where ε is called dielectric constant of the medium.
(ii)
Since the magnetic lines of force are either closed or go off to infinity, the
number of magnetic line of force entering any arbitrary closed surface is
exactly the same as leaving it. It means that the flux of magnetic induction
B across any closed surface is always zero i.e..,
B.ds = 0
Transforming the surface integral into volume integral, we have
div B dV = 0
The integrand should vanish for the surface boundary as the volume is
arbitrary, i.e.,
div B=0
(iii) curl E - B/ t
By Faraday law we know that emf induced in a closed loop is given by
e = - / t = B/ t . ds
div B= 0
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 57
P a g e | 57
Since the flux = s B.ds where S is any surface having the loop as
boundary. E.m.f `e` can also be found by calculating the work done in
carrying a unit charge completely around the loop. Thus e = E. dl where E
is the intensity of the electric field associated with induced em f
Therefore, equating above two equations, we get
E.dl = - s B/ t . ds
Applying stokes` theorem, the line integral can be transformed into surface
integral i.e.,
› s ( x E) . ds = - s ( B/ t) . ds
This equation must be true for any surface whether small or large in the
field. Therefore the two vectors in the integrands must be equal at every
point, i.e.,
› x E = - B/ t
Crul E = - B/ t
(IV) --- Curl H = J+ D/ t
Ampere’s law in the circuital form gives this equation. According to this law,
the work done in carrying a unit magnetic pole once round closed arbitrary
path linked with the current is expressed as
H. dl = I (or) = J. ds
where the integral on the right is taken over the surface through which the
charge flow corresponding to the current I take place. Now changing the
line integral into surface integral by stoke`s theorem
s curl H.ds = J.ds
Curl H= J
The above relation, derived on the basis of Amperes` law, stands only for
steady closed current. But for the changing electric fields, the current
density should be modified. The divergence of the above equation is
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 58
P a g e | 58
div (curl H ) = 0 (or)
div J=0 which conflicts with the equation of continuity div J = (- ρ/ t.)
Adding J we get curl H= (J+J'). Taking divergence of the above equation, we
get
div (curl H) = (divJ+divJ') (or)
0= div J+divJ' (or)
div J' = -divJ = + ρ / t.
› we know that ρ = . D
Substituting this value in the expression for Div J', we get
› Div J' = / t ( . D)
› › . J`= ( . D/ t)
therefore the Maxwell’s fourth relation can be written as
Curl H = J+ ( D/ t)
3.4 Maxwell’s equation in free space
In the free space, where the current density J and volume charge
density ρ are zero, Maxwell’s equations reduce to
› .D = 0
› .B = 0
› x E = - B/ t
› x H = D/ t
Maxwell’s equations in linear Isotropic Media
In linear isotropic media
D = E
and H= B/µ
Where is the dielectric constant, µ permeability of the medium.
The Maxwell’s equations become
› . E = ρ /
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 59
P a g e | 59
› . H = 0
› x E + µ( H/ t) = 0 › x H - ( E/ t) = J.
3.5 Energy in electromagnet fields: Poynting vector – ( Poynting
theorem)
Energy may be transported through space by means of e.m. waves.
Let the material inside S be isotropic homogenous and characterised by
permeability µ, permittivity and conductivity . For derivation, consider a
volume V bounded by a closed surface S.
Maxwell’s third and fourth relations are
Curl E = - B/ t
Curl H = J + D/ t
Taking scalar product of both sides of the above equations with H and E
respectively and subtracting, we get
E. curl H . H curl E = J. E + [(E.( D/ t) + H ( B/ t)] ---(1)
But we know that
H. curl E-E. curl H = div (ExH)
So that equ. (1) becomes
-div (E x H) = J. E[E.( / t) ( . E) + H( / t) (µH)]
(or) J.E + (1/2. / t ( E2) + 1/2. / t(µH2) + div (ExH) = 0
(or) J.E +(1/2. / t (E. E) +1/2. / t (H. µH) + div (ExH) = 0
(or) J.E +(1/2. / t (E.D) + 1/2. / t(H.B) + div (ExH) = 0 --(2)
Integrating over the volume V bounded by the surface S, we get
v (J.E) dv + v 1/2. / t(E.D+H. B) + vdiv (ExH) dv = 0 ----(3)
But as v div (ExH) dv = (ExH).ds
We write equation (3) as
v (J.B) dv + v 1/2 / t (E.D + H.B) = - s (ExH).ds ---- (4)
Integrating the second term of Equ. (4) we get
1/2 / t(E.D+H.B)dv = 1/2 . / t(E. E + H .µH )dv
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 60
P a g e | 60
= / t (1/2 . E2 + 1/2 µH2) dv -----(5)
The first and second term on right hand side represent the time rate of
increase of energy stored in the electric and magnetic fields respectively in
the volume V. Considering the Eq.(4), LHS of this represents the sum of the
power expended by the fields due to the motion of charge and the time rate
of increase of stored energy in the fields. On the other hand RHS of Eq. (4)
must represent the power flow into the volume V across the surface S, or the
power flow out of the volume V across the surface S
= s(ExH) . ds
= s P.ds
where P= (ExH) ----(6)
It then follows that the vector P has the meaning of power density associated
with the electromagnetic filed at that point. The statement represented by
Eq.(6) is known as poynting theorem and the vector P is known as the
poynting vector.
3.6 Electromagnetic potentials – Maxwell’s equations is terms of
electromagnetic Potentials
Consider the Maxwell’s equations
µCurl H = µJ+µ D/ t -------(1)
Curl B = µJ + µ E/ t -----(2)
Where and µ are permittivity and permeability, Substituting for B (i.e. B =
Curl A) and E(-grad φ- A/ t) (where A and φ are electromagnetic potentials),
we get
Curl (CurlA) = µJ +µ / t (-grad φ- A/ t)
i.e.., grad div A- 2A=µJ-µ / t (grad φ)- µ 2A/ t2)
i.e., 2A -µ 2A/ t2 –grad (divA+µ φ/ A) = -µJ ----(3)
Considering other Maxwell’s equations, namely
Div D = ρ
div E = ρ
i.e., div (- grad φ- A/ t) = ρ/
i.e., 2 φ = / t(div A) = - ρ/
Adding and subtracting µ φ 2/ t2 it becomes
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 61
P a g e | 61
2φ - µ 2φ/ t2 - / t(div A+µ φ/ t)= - ρ/ --(4)
Equations (3) and (4) are field equations in terms of electromagnetic
potentials. Here Maxwell’s equations are reduced from four to two by
electromagnetic potentials,
Note: Electromagnetic potential define the field vectors uniquely though
they themselves are non-unique. We get the same field vectors when we use
the set (A, φ) or (A', φ’). These transformations are called gauge
transformations.
3.7 Lorentz Gauge
Maxwell’s field equations in terms of electromagnetic potentials are
2A- µ 2A/ t2-grad (divA+µ φ/ t) = -µJ ---(1)
2 φ - µ 2φ/ t2 + / t(div A+µ φ/ t)= -ρ/ –(2)
The above equations may be simplified as
div A +µ φ/ t = 0 ----(3)
This requirement is called the Lorentz condition and when the vector and
scalar potential satisfy it, the gauge is known as Lorentz gauge.
So with Lorentz condition field equations reduce to
2A-µ ( 2A/ t2) = -µJ ----- (4)
and 2φ -µ ( 2φ/ t2) = -ρ/ -----(5)
But we know µ = 1/v2 (i.e., V= 1 µ )
Hence Equations (4) and (5) can be written as
2A = -µJ ------(6)
2 φ = - ρ/ -----(7)
and 2 = 2 – 1/v2 * ( 2 Λ / t2)
Equations (6) and (7) are inhomogenous wave equations and are known as
D' Alembertian equations and can be solved. The potentials obtained by
solving these equations are called retarded potentials.
To determine the requirement that Lorentz condition Λ, we substitute A' and
φ' from equations already given earlier.
Div (A' –grad Λ) + µ / t (φ' + Λ / t) = 0
i.e., div A' = µ ( φ'/ t) = 2 Λ - µ ( 2 Λ / t2)
Hence A' and φ' will satisfy the equations (3) i.e.., Lorentz condition provides
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 62
P a g e | 62
2A - µ ( 2 Λ / t2) = 0 ----- (8)
i.e., 2A = 0
Lorentz condition is invariant under those gauge transformations for which
the gauge functions are solutions of the homogeneous wave equations.
Advantages:
(1) It makes equations for A and φ independent of each other.
(2) It leads to the wave equation.
(3) It is a concept which is independent of Co – ordinate system.
3.8 Coulomb Gauge
Consider the field equations in terms of electromagnetic potentials we get,
2 φ + / t (div A) = - ρ/ -----(1)
If we assume div A = 0
The above equation (1) reduces to Poisson’s equation
2 φ(r,t) =ρ(r' t)/ ---(2)
Whose solution is
φ (r,t) = 1/4 (ρ (r' t)/R) d '----(3)
i.e., the scalar potential is just the instantaneour Colombian potential due to
charge ρ (x', y', z',t). This is the origin of the name coulomb gauge.
From equations (2) (3) we get
2 { 1/4 (ρ (r' t)/R) d ' } = -ρ ( r' , t)/ ---(4)
As Poisson’s equations holds good for both scalar and vectors replacing
ρ (r' t) by J' we get
2 [1/4 (J' /R) d ' ] = -J' / --------(5)
As J' is confined to the volume ' the surface contributions will vanish, so
(J' /R) d ' = ( 'J' /R) d ' --------(6)
(Since J' = -(1/4 ) . J' /R) d ' +1/4 x x (J' /R*) d ’]
We know x (J' /R) = 1/R ' x J' x ' (1/R)
= ( ' x J') /R) d 2 + (J' /R) x ds
(as x V d 2 = - sV x ds )
As J' confined to volume ' surface contribution will vanish so
x (J' /R) d ' = ( ' x J' /R) *d ' --------(7)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 63
P a g e | 63
Now as ' x J' = x [-1/4 ( '. J'/R) d ' ----------- (8)
i.e., ' x J' = 0 (as curl grad φ = 0)
and . J' = [ x ( ' x J' /R) d '
i.e. . J = 0 (as div curl V = 0)
In terms of vector potential
2 A – (1/v2) 2A/ t2 = -µJ
2A = -µJ
i.e.., the equation for A can be expressed entirely interms of the transverse
current.
The Coulomb gauge has a certain advantage. In it the scalar potential is
exactly the electrostatic potential and electric field is given by
E = - grad φ - A/ t
It is separable into an electrostatic field V = φ and a wave field given by -
A/ t. This gauge is used when no sources are present. If φ = 0 and A
satisfies the homogeneous wave equation, the field is given by
E = - A/ t and B = x A
3.9 Radiation produced by a low velocity accelerated charged particle
(Larmor`s formulae)
If a charge is accelerated but is observed in a reference frame where
its velocity is small compared to that of light, then in that coordinate frame
the acceleration field is given by
. Ea = e/c [n x (nx β) / R ]ret -----(1)
Where β = dβ / dt is the ordinary acceleration divided by C. n is the unit
vector, `ret` mean that the quantity in the brackets is to be evaluated at
retarded time. The instantaneous energy flux is given by the poynting vector
S = C/4 (ExB) = C/4 -Ea- 2 n---- (2)
It means that the power radiated per unit solid angle d
dP /d = C/ 4 -REa - 2 = [e2/4 c [nxnxβ2) ] ----(3)
If θ is the angle between the acceleration V and n, then the power radiated can be written as . dP/d = (e2 /4 c3) -V - 2 sin2θ ---- (4)
This exhibits the characteristic sin2θ is angular dependence. We note from
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 64
P a g e | 64
. equ.(1) that the radiation is polarised in the plane containing V and n. The
total instantaneous power radiated is obtained by integrating equ. (4) Over
all solid angle. Thus
. P= 2/3 (e2/c3) -V - 2 -----(5)
This is Larmor result for a non-relativistic accelerated charge
Larmor`s formulae can be generalised by arguments about covariance
under Lorentz transformations to yield a result which is valid for arbitrary
velocities of the charge. (Radiated electromagnetic energy behaves under
Lorntz transformation like the fourth component of a 4 –vector)
3.10 Radiating systems.
The charges are the ultimate source of electromagnetic fields. Bohr`s
theory of hydrogen spectrum and the allied theories of radiation emission by
atoms and molecules exhibit the basic radiating system. Here the radiant
component of electric and magnetic fields associated with accelerated charge
e are given by the relation
› › › Er = (e/4 0c) [rx(r-rβ) x β / (r-r. β) 3 ----- (1)
and Br = (nxEr)/C -----(2)
The flow of the electromagnetic energy is given by poynting vector
Pr = Er x Hr ----(3)
The radiating system is normally considered as an oscillating dipole. Now
we discuss of half-wave for antenna array systems employed for propagation
of electromagnetic energy into space.
3.10 (a) Radiation due to an Oscillating Electric dipole.
A pair of charges equal in magnitude and opposite in nature separated by a
small distance constitute an electric dipole. Such a dipole is shown below
dl
P
+ +q
-
-q
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 65
P a g e | 65
The dipole moment p is defined by p = q.dl --- (1)
Consider the dipole is acted upon by a field of frequency ω varying sinusoid
ally. The separation dl varies according to applied field and varies the
dipole moment. The instaneous value of dipole moment may be written as
þ = þ0e-iωt ----(2)
Where þ0 represent the amplitude of dipole moment. Considering volume
distribution of charge density ρ, the dipole contribution is
þ = (ρ dV') r --------(3)
Where dv' is the source volume separated by a distance r from the origin.
Considering time rate variation of dipole moment,
. dp/dt = p = d/dt (q.dl) = dq/dt x dl = Idl ----(4)
where dq/dt = I, the current. This implies that the time rate variation of
dipole moment is equivalent to a current element I.dl. An oscillatory motion
of a system is an accelerated type of motion and acceleration associated with
such a motion is responsible for radiation.
Poynting vector and Radiated power
The poynting vector P is given by
P= Er x Hr = Er x (Br /µ0) .. ..
þ › (þ)2sin2ø / 16 2 0c3r2 (nθ x n) ---(1)
Where nθ and n are the transverse components of unit normal vector
(þ)2sin2 θ /16 2 0c3r2 . nr ------(2)
Where nr is the radial components of unit normal vector
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 66
P a g e | 66
The figure shows P varies as sin2θ. Let us take [þ] = þ0 cosω (t-r/c)
..
[þ]= ω2 þ0 cosθ (t-r/c) ----(3)
Substituting this value in equation (2) we get
P = ω4 þ02sin2θ/16 2 0r2c3 x cos2ω (t-r/c)nr ----(4)
The average of cos2ω (t-r/c) over a period of oscillation is 1/2. Therefore
average power radiated per unit area over a time period is
-P- = ω4 þ02sin2θ/32 2 0c3r2 x nr -----(5)
The power radiated through area ds is
dP =-P-ds
but if ds subtends solid angle d at the source distant r from ds, then
d = ds/r2 = r2sinθ dθ dφ / r2 -----(6)
and therefore power radiated per unit solid angle is
dp/d =-P-ds / (ds/r2) =-P1-r2
= ω4 þ02sin2θ/32 2 0c3 -----(7)
the total radiated power is given by
PT = (dp/d ) x d 2
0 0 ω4 þ02sin2θ/32 2 0c3 (sinθdθ)dφ
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 67
P a g e | 67
= ω4þ02/16 0c3 0 (1-cos2θ) sinθdθ
= ω4þ02/16 0c3 (- cosθ + cos3θ/3]0
PT = ω4þ02/12 0c3 ----(7)
This relation shows that the total power depends upon fourth power of the
exciting wavelength and þ02 . The dipole moment amplitude þ0 i s a
characteristic of the substance under test. The theory of dipole radiation is
used in explaining the radiation by atoms, molecules and by the antenna.
3.11 Radiation due to a small current element.
An oscillating dipole may be treated as a small current element, hence the
radiation field equations developed for a dipole may be applied to the case of
linear small current element. The case of the linear small current element
may be further extended to the case of lengthy current elements like linear
antenna and antenna arrays. The radiation components of electric and
magnetic fields due to an oscillating dipole are given by
.. E = [(þ) sinθ/4 0c2r] x nθ
.. B =[µ0(þ) sinφ/4 cr] x nφ -------(1)
The dipole moment of an oscillating dipole, which is treated as small current
element, is given by
þ = q.dl .
dp/dt = þ = (dq/dt) x dl = Idl ..
Also d2þ/dt2 = þ = (dI/dt) x dl ------(2)
Taking Eq. (2) becomes ..
þ = -i ωI0e-i ωtdl
Whose retarded value for a field and source point distant r is written as ..
[þ] = - iωI0dleiω(t-r/c) -----(3)
with real components of current, the substitution of Eq.(3) in Eq. (1) gives
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 68
P a g e | 68
E = [iωI0dlsinθ/4 0c2r] x e-i ω (t-r/c)nθ ----(4)
B =[iµ0ωI0dlsinθ/4 cr] x e-i ω (t-r/c)nφ
= - iωI0dlsinθ/4 0c3r x e-i ω (t-r/c)nφ since µ0 0 = 1/c2
[ -E- /C ] x nφ -----(5)
Where -E-is the numerical value of electric field given by equ. (4). Knowing
the components of electric and magnetic radiation fields, we can compute
the poynting vector. The average value of poynting vector for complex fields
is given by
P = (ExH)/2 = (ExB)/2µ0
Where the real part of E and B have to be taken into account
-P-= [ω2I02dl 2sin2θ/2(16 2 0c3r2) x nθ x nφ
= [ω2I02dl 2sin2θ/32 2 0c3r2] x nr -----(6)
Total radiated power is given by
PT = -P-ds = φ=2 θ=
= φ=0 =0 [ω2I02dl 2sin2θ /32 2 0c3r2 ] x (r2sinθdθdφ)
= ω2I02dl 2/32 2 0c3[8 /3]
φ=2 θ = ( Since φ=0 =0 sin3θdθdφ = [8 /3] )
Using ω = 2 c/λ,I0 = Irms 2 and µ0 0=1/c2 we get
PT = [(2 c/λ)2 ( 2 Irms)2.dl 2 /12 c] x µ0
=(2 µ 0c/3) (dl/ λ) 2 I2rms
Substituting µ0 = 4 x10 -7met/sec and C = 3x108 met/sec we get
PT = 80 2(dl/λ)2 I2rms -----(7)
The radiation resistance may be calculated by comparing equ. (7) with the
equation for the loss of power i.e.,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 69
P a g e | 69
P = RI2rms ------(8)
It gives radiation resistance as
Rr = 80 2(dl/λ)2 ohms ---(9)
This shows that the radiation resistance is equal to the resistance in which
the power consumed and converted into heat is equal to the power radiated
in the form of E.M waves be a current element, provided the same current is
maintained.
Note: A short linear antenna is an example of linear current radiator. A
centre fed linear antenna is shown below.
Z F(x )
Coaxial θ r
X
φ
Y
Fig. Short linear antenna.
In case of short linear antenna dl<<λ. In such cases the radiation
resistance is usually quite low as compared to the ohmic resistance. For
efficient radiation of electromagnetic energy it is essential that the length of
the antenna is comparable with wavelength. A half-wave antenna i.e.
dl=λ/2 is commonly used to radiate electromagnetic energy into space.
3.12 Linear Half –wave antenna.
A linear half wave antenna is a straight conductor of length equal to
half the free space wavelength. If a linear conductor is set vertically on the
ground then only quarter free space wavelengths (λ/4) of the conductor
serves the purpose of half wave antenna as additional λ/4 length is
furnished by the ground reflection. Such antenna is used for radio
broadcast purpose and is shown below
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 70
P a g e | 70
z
dl F(r,θ)
λ/4 r' λ/4
l r
λ/4
λ/4
Fig. 3.12 Linear half wave antenna
Let the current fed to the antenna be I = I0cosωt and at centre I= I0 at t=0
consider an element dl at a distance l from the centre, then the amplitude
of current in dl is
I0cosωt ( /c) = I0cos (2 l / λ)
Since t = (l/c) and ω = 2 f = 2 c/ λ
The current in dl at any moment t is
I = [I0cos (2 l /λ)]cosωt
= I0/2 [cosωt - 2 l / λ) + cos(ωt + 2 / λ)
Writing this equation in exponential form we get
I = I0/2 [e-i(ωt – (2 l/λ) + e-i(ωt + (2 l/λ)] -----(1)
Now with this value of current in dl at the time t, we can calculate electric
intensity and it can be related to B (magnetic induction ). It can be written
as
E = -(µ0c/2 r) I0eiωI(t-r/c) . cos [ /2 cosθ] /sinθ * nθ ------(2)
The above equation shows that E is independent of frequency. Now B and
E are related as shown below.
B = -E-/C * nφ substituting -E-by equation (2) we get
B = (iµ0I0/2 r) * e-iω(t-r/c) . cos [ /2 cosθ] /sinθ * nφ web/m2 ----(3)
Poynting vector and Radiated power
Knowing E and B poynting vector and radiated power can be evaluated. The
average of Poynting vector or complex field is given by
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 71
P a g e | 71
P = ExH/2 = ExB/2µ0= ExE/2µ0c * (n θ x nφ ) = (E2/2µ0c) * nr
Substituting the value of E from (2) we get
P =(µ0c I20/8 2r2) * e-2 iω(t-r/c)] cos² [( /2 cosθ] / sin²θ * nr
Using I0 = 2 Irms we get
P = µ0c/4 2 *I2rms/r2 * cos² [( /2 cosθ] / sin²θ * nr -----(4)
Total radiated power is given by
φ=2 θ= PT = µ0c/4 2 *I2rms θ=0 θ=0 cos2 [( /2) cosθ]/r2 sin2θ * r2 sinθdθdφ
=µ0c/2 *I2rms θ=0 cos2 [( /2) cosθ/sinθ) .dθ -----(5)
Now let ( /2) cosθ = (β/2 - /2) so that cos2 [ /2 cosθ] = cos2θ [β/2 - /2]
=sin2 β/2 = 1-cos β/2
and ( /2) sinθdθ = 1/2dβ (or) dθ = d β/( s inθ) and µ0c/2 =60
Substituting these values in (5) we get
2 PT = 60 I2rms 0 (1-cos β)/β(4 - 2 β)] xd β
But 1/ β(4 - 2 β) = 1/4 [1/β + 1/2 - β]
2 2 PT = 15I2rms[ 0 (1-cos β/β) *dβ + 0 (1-cos β/2 - β) * dβ The integrals within the square brackets are evaluated graphically within
limits 0 to 2 . The plots of (1-cos β) against β and 2 - β are the same as
shown in figure below. The plot (1-cos β) is symmetrical about β = .
Therefore the areas expressed by two integrals within limits 0 to 2 . are
equal and
2 PT = 30I2rms 0 (1-cos β/β) *dβ
The value of the integral as obtained graphically is
2 0 (1-cos β/β) *dβ = 2.44
Therefore PT = 73.2 I2rms ------(6)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 72
P a g e | 72
2 -β
1-cos β
0 2
Fig- 3.13 Current Let sum up
This shows that the radiation resistance of half wave antenna is much
higher than the dipole antenna. This implies that in case of such antenna
most of the power is radiated and ohmic losses are negligibly small. Hence
the efficiency of such antenna is large. It is for this reason that such
antenna are commonly used for radio broadcasting.
-------
3.13 Let us sum up
In this unit we described the terms and derivation for equations of
continuity, displacement current, Lorentz gauge and Coulomb gauge.
Also explained the Maxwell’s equations along with the
derivations.
Finally you have learnt about the low velocity acceleration charged
particle and the radiation due to a small current element and its application
to linear half wave antenna.
3.14 Lesson end activities
(i) What is the displacement current?
(ii) Write the fundamental Maxwell’s filed equations in differential
form?
(iii) Define Poynting vector.
3.15 Points for discussion
(i) Find the magnitude of displacement current.
(ii) Assuming the total energy Ze of an atomic nucleus is uniformly
distributed within a sphere of radius a, show that the potential
energy of such a nucleus is
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 73
P a g e | 73
n = a
Ze2
0 5
3
4
1
pe
[Hint : U = ½ òe0 E2 dv ; Ein = 2
0
2
04
1
R
e
pe; Eout=
2
2
04
1
r
e
pe
Substituting q = Ze and Ro = a and solving the integral we
get the result.]
3.16 Check your progress
§ Derive Maxwell’s field equations.
§ Derive equations for
a) Lorentz gauge
b) Coulomb gauge
§ What is oscillating dipole? Derive an equation for radiation due to a
small current element.
§ Write a short note on linear half-antenna.
3.17 References
o Electrodynamics by Gupta ,Kumar Singh
o Electromagnetic Fields and waves by Pauli Lorrain and Dale Corson.
o Electromagnetic theory by K.K. Chopra and G.C.Agrawal.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 74
P a g e | 74
UNIT – II Lesson - 4
INTERACTION OF ELECTROMAGNETIC WAVES WITH MATTER
4.0 Aim and Objectives:
In this chapter, we shall consider the interaction of electromagnetic waves with
matter, i.e., the phenomena of reflection, refraction, scattering and dispersion. We sha ll
see that at the boundary between two dielectrics, the electromagnetic waves obey the
familiar laws of reflection and refraction. The derivations in these phenomena will be
b a s e d o n general electromagnetic equations. In order to discuss the behaviour of
electromagnetic waves at the boundary, we shall first discuss the boundary conditions
which the electric and magnetic fields must satisfy at the surface of discontinuity
between the two media.
4.1.Boundary Conditions For The Electromagnetic Field Vectors: B, E, D, And H, (At
The Interface Between Two Media)
We shall now investigate the boundary conditions which the time dependent
electromagnetic field vectors B, E, D and H sa tisfy at the interface between two
different media. Boundary conditions are:
(1) The normal component of magnetic induction B is continuous across
boundary, i.e.,
B1n=B 2n
(2) The tangential component of E is continuous across the i nterface, ie.,
E1t=E2t.
(3) The normal component of electric displacement D is discontinuous across
the interface, i.e.,
D1n - D2n = s.
(4) The tangential component of magnetic intensity H is continuous across the surface
separating two dielectrics, i.e.,
H1t =H2t ·
(1) Boundary condition for B : The magnetic induction B satisfies the Maxwell's equation
div B = O. .. (1)
At the interface between two media, we construct a pill box like surface as.
shown in figure 4.1.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 75
P a g e | 75
Fig.4.1. Pill box shaped surface at the interface of two media.
The surface is composed of four surfaces i.e., S1, S2 , S1 ’ and S2 ’. We apply the
divergence theorem to the divergence of B over the volume enclosed by this surface and
obtain
0 div =ò dvV
B …(2)
where dV is the infinitesimal volume element. Transforming volume integral into
surface integral by Gauss theorem, we get
ò =S
ndSB 0. …(3)
where n is an outwardly drawn unit vector normal to the infini tesimal area element dS of
the surface. Applying equation (3) to the whole surface of pill box, we have
0....'2
'121
22112211 =+++ òòòò dsnBdsnBdsnBdsnBSSSS
…(4)
The term( the third and fourth) in equation (4) give the contribution to the surface integral
from the walls of the pill box. If B is finite every where, then making the height of pill box to
approach zero. i.e, when S1 and S2 approach each other towards the interface, i.e, in the limit
0®hd , we have,
0)..( 2211 =+ò dAnBnBA
as S1 and S2 approach the area A.
Since the area A is quite arbitrary, the above equation becomes
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 76
P a g e | 76
2211 .. nBnB -= …(5)
If n12 is the unit normal vector pointing from the first medium into the second medium, then
n1=-n12 and n2=n12. Equation(5) gives
122121 .. nBnB -=-
or 122121 .. nBnB =
or nn BB 21 = …(6)
i.e., the normal component of magnetic induction is continuous across the boundary.
(2) Boundary condition for E:
The boundary condition, which the tangential component of electric field must
satisfy, may be obtained from Maxwell’s equation
t
BcurlE
¶
¶-= …(7)
Fig.4.1(a) Rectangular loop at the interface between two media.
At the interface, we construct a rectangular loop ABCD bounding a surface S as shown in
Figure 4.1(a). Integrating equation (7) over the surface bound by rectangular loop ABCD, we
have,
ò ò ¶
¶-=
S SdSn
t
BdS E.n curl …(8)
Transforming the surface integral of the left hand side into a line integral over the
path ABCD with the help of Stoke’s theorem, equation(8) yields,
ò ò ¶
¶-=
ABCD SdSn
t
BdlE 1
Or
ò òò ¶
¶-=++
AB SCDdSn
t
BdlEdlE DA and BC sides from onsContributi21 ..(9)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 77
P a g e | 77
If the loop is now shrunk ( letting 0®hd ), the contribution from sides BC and DA will
vanish. Provided tB ¶¶ is every where finite, the right hand side of equation (9) also vanishes
because the area of integration also vanishes i.e, no flux can be enclosed. Thus when 0®hd .
ò ò =+AB CD
dlEdlE 021
or E1AB+E2CD=0
or E1AB-E2AB=0
or E1t=E2t …(10)
where E1t and E2t are the tangential components of the electric field in the two media.
Equation(10) shows that the tangential components of E are continuous across the interface.
(3) Boundary Condition of electric Displacement:D
The boundary condition for the electric displacement D can be obtained from
Maxwell’s equation
divD=r …(11)
integrating equation(11) over the pill box shaped volume V,we obtain
ò ò=V V
dVdV rdivD …(12)
Transforming the volume integral of left hand side of equation(12) in to surface integral with
the help of Gauss divergence theorem, we get,
ò ò=S V
dVdS rD.n
or òòò =++VSS
dVdSndSn r wallsfrom onsContributi.D.D21
2211 …(13)
Now let us suppose that the surfaces S1 and S2 approach each other towards the
interface so that S1 = S2=A, i.e, 0®hd the contribution from walls tends to zero. Now
instead of volume charged density r, the concept of surface charge density s must be used,
i.e,
òò ==®VV
AdAdVhLim ssrd 0 …(14)
\ AdSndSnSS
s=+ òò .D.D21
2211 …(15)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 78
P a g e | 78
or D1.n1S1+D2.n2S2=sA
or D1.n1+D2.n2=s as S1=S2=A
or D1.n1+D2.n1=s
or D1n-D2n=s, …(16)
where D1n and D2n are components of the electric displacement in the two media normal to
the interface in the direction of n1. Evidently, the normal component of electric displacement
is not continuous across the interface and charges by an amount equal to the free surface
density of charge at the interface.
(4) Boundary Condition for H:
Finally, consider the behaviour of the tangential component of magnetic field
intensity. According to Maxwell’s equation, we have,
t
D
¶
¶+= JH curl …(17)
Now consider a small rectangular contour (as shown in figure 4.1(a)) enclosing an
area S and integrate equation(17) over this area we get,
òò ÷ø
öçè
æ
¶
¶+=
SSndS
t
DJdS .H.n curl …(18)
Transforming the surface integral into line integral with the help of Stoke’s theorem,
we have
ò ò ÷ø
öçè
æ
¶
¶+=
ABCD SndS
t
DJdlH ..
or òòò ÷ø
öçè
æ
¶
¶+=++
SCDABndS
t
DJdldl .DA and BC sides from onsContributiHH 21 (19)
If the loop is shrunk in the limit 0®h , we note that the contribution to line integral
along sides BC and DA vanishes. That is
0.DA and BC0
®ò®dlHLim
h
ò ®¶
¶®
0.0
ndSt
DLimh
i.e, t
D
¶
¶ is bound every where
and ò ^®
®S
Sh
lJdSLim J.0
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 79
P a g e | 79
where ^SJ represents the component of surface current density perpendicular to the
direction of H-component which is being matched. The idea of surface current density
is closely analogous to that of a surface charge density - it represents a finite current in
an infinitesimal layer. Then in the limit 0®h eq. (19) takes the form
lJdSdl SCDAB
^=+ òò J..H
or lJlHlH Stt ^=- 21
or ^=- Stt JHH 21 …(20)
The surface current density is zero unless the conductivity is infinite, for finite conductivity,
we write
tt HH 21 = …(21)
i.e, the tangential component is continuous across the surface separating the two dielectrics.
4.2. General Treatment of Reflection and Refraction:
Now we shall consider the reflection and refraction. Consider the case of a plane
interface between two different isotropic, homogeneous, stationary, charge free, linear and
non-conducting media of finite extensions. Let the medium of one side (medium-1) has the
permittivity ε1 and permeability m1 and the medium of other side ( medium-2) has the
permittivity ε2 and permeability m2. Consider that an electromagnetic wave is propagating in
medium-1 and is incident at a point P(figure 4.2) on the interface. This wave gives rise to
both a reflected and a transmitted wave. In the figure n i,nr and nt are unit vectors normal
to the respective wave- fronts and point in the direction of propagation. The angles qI,qr
and qt are the angles of incidence, reflection and refraction respectively. The electric field
and magnetic field intensities of the incident, reflected and transmitted waves at the point of
incidence P are described by the following equations:
( ){ }[ ]( ){ }[ ]ïî
ïíì
-=
-=
trnjHH
trnjEE
iiiioii
iiiioii
wmew
wmew
. exp
. exp
11
11
( ){ }[ ]( ){ }[ ]ïî
ïíì
+-=
+-=
'11
11
. exp
. exp
AtrnjHH
AtrnjEE
rirrorr
rirrorr
wmew
wmew
( ){ }[ ]( ){ }[ ]ïî
ïíì
+-=
+-=
'11
11
. exp
. exp
BtrnjHH
BtrnjEE
tittott
tittott
wmew
wmew
1(a)
1(b)
2(a)
2(b)
3(a)
3(b)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 80
P a g e | 80
where r i is the position vector of P and A, A', B and B' allow for possible phase
differences with the incident wave at interface.
Fig.4.2. Showing the phenomena of reflection and refraction.
We can investigate the characteristics of both the r eflected and transmitted waves
from the fact that the tangential components of E and the tangential components of H
must both be continuous across the interface. Thus (E i + Er ) must be equal to E t . The
reflected and the transmitted waves could also be obta ined from the continuity of the
normal components of D and B across the interface,
In order to have the continuity of tangential components of E and H at the interface,
it is essential that there must be some valid relation between E i E r , and Et , for all time t
and for all vectors r i which terminate on the interface. There are two possibilities to
have such a relation, (i) if all three vectors E i E r, and Et are identical function of the
time t and of position ri and (ii) there exist certain relations betwe en E 0 i E 0r, and E0t ,
Applying the condition (i), we have
( ){ }[ ] ( ){ }[ ]
( ){ }[ ]Btrn
Atrntrn
titt
rirriiii
+-=
+-=-
wmew
wmewwmew
.
. .
22
1111 (4)
which must be true for all time t and for all vectors ri. It follows that
tri www == …(5)
This shows that all three waves have the same frequency. Again from equations(4), at any
point ri on the interface, we must also have
( ) ( ) ( ) BrnArnrn itirii +=+= ... 221111 mewmewmew …(6)
Evidently, from equation (6) we get
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 81
P a g e | 81
( )( )11
.mew
Arnn iri =- …(7)
where the term ( )11mew
A is a constant for all ri. The projection on (ni-nr) of any vector ri
terminating on the surface must be a constant. The vector (ni-nr) is also a constant for any
given incident plane wave. Then (ni-nr) must be normal to the interface, i.e, parallel to the
unit vector n normal to the interface at the point P, hence ni,nr and n are coplanar. The plane
of the three vectors is called the plane of incidence.
Since (ni-nr) is normal to the interface, the tangential components of these two vectors
must be equal and opposite sign then
ri qq = …(8)
i.e., the angle of reflection is equal to the angle of incidence. These are the laws of reflection.
Again considering equation(6) we obtain,
( ) ( ) Brnn iti =- ].[ 2211 mewmew …(9)
Thus the vector ( ) ( ) ][ 2211 ti nn mewmew - must be normal to the interface, so that
ni,nr, nt and n are coplanar. In this way all the four vectors ni,nr, nt and n are in the plane of
incidence. Moreover, the tangential components of ( ) in11mew and ( ) tn22mew must be
equal but opposite in sign i.e.,
( ) ( ) ti qmewqmew sinsin 2211 =
( )( ) 1
2
0011
0022
11
22
)(
)(
sin
sin
n
n
t
i ===meme
meme
me
me
q
q
or ti nn qq sinsin 21 = …(10)
where n1,n2 are the indices of refraction of the two media. This is Snell’s law.
The constants A and B are related to the choice of the origin. At a given point of interface,
and at a given time, the reflected and transmitted waves have definite phases and if the origin
is displaced, A and B must be adjusted accordingly. If the interface is chosen as the origin
then from Equations (7) and (9) A=B=0, since the ri will be normal to
(ni-nr) and ( ) ( ) ][ 2211 ti nn mewmew -
Taking the axes as shown, we have,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 82
P a g e | 82
ttt
rrr
iii
kjn
kjn
kjn
qq
qq
qq
cossin
cossin
cossin
+-=
+-=
+-=
…(11)
The electric field of the incident wave at a point r =jy+kz in the plane of incidence is given by
{ }[ ]tzyjEE iiii wqqmew -+-= )cossin()(.exp 110 …(12a)
{ }[ ]tzyjEE rrrr wqqmew -+-= )cossin()(.exp 110 …(12b)
{ }[ ]tzyjEE tttt wqqmew -+-= )cossin()(.exp 110 …(12c)
4.3. Fresnel’s Equations(Dynamic Properties):
Let us find the relation between the equations iE0 rE0 and tE0 . The formulae relating
the amplitude of the reflected and transmitted waves with that of incident wave are known as
Fresnel formulae. We know that E and H vectors in a plane electromagnetic wave are always
perpendicular to the direction of propagation and to each other. The vector E does not fix the
direction uniquely and it is said that the wave is polarized in a particular direction of its
electric vector. The vector E of incident wave can be oriented in any direction perpendicular
to vector ni.
It is convenient to consider two cases: (i) in which the incident wave is polarized such
that its vector E is normal to the plane of incidence (ii) in which the vector E is parallel to the
plane of incidence.
Case I. Incident wave polarized with its vector E normal to the plane of incidence:
In this case the electric and magnetic field vectors E and H of the incident wave are
perpendicular to the direction of propagation ki, as shown in figure 4.3(a).
The pictorial diagram of this case is shown in figure 4.3(b).Since the media are
isotropic, the electric field vectors of the reflected and transmitted waves will also be
perpendicular to the plane of incidence.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 83
P a g e | 83
Fig.4.3(a). Reflection and Refraction with Fig 4.3(b). Showing incident, reflected Polarization perpendicular to the plane of and transmitted waves when the incident incidence. wave is polarized with the E vector normal to the plane of incidence.
The continuity of the tangential component of electric field intensity at the interface requires
that
tri EEE 000 =+ …(1)
Similarly, the continuity of magnetic field intensity at the interface requires that
ttrrii HHH qqq coscoscos 000 =+ …(2)
at any given time at any given point on the interface. We know that,
11wmii
i
EkH
´= úû
ùêë
é ´==
wm
EkBandHBQ
or ( )[ ]
1
11
11 m
me
wm
iiiiii
EnEnKH
´=
´= ( )÷÷
ø
öççè
æ= 11me
wi
iKQ
Hence, ( )
ii EH 0
1
11
0m
me=
( )rr EH 0
1
11
0m
me= and
( )tt EH 0
2
22
0m
me=
Substituting these values in Equation (2), we have
( ) ( ) ( )ttirii EEE q
m
meq
m
meq
m
mecoscoscos 0
2
22
0
1
11
0
1
11 =-
or ttriii EEE 0
2
20
1
10
1
1 coscoscos qm
eq
m
eq
m
e÷÷ø
öççè
æ=÷÷
ø
öççè
æ-÷÷
ø
öççè
æ …(3)
Eliminating E0t from equations (1) and (3), we get
)(coscoscos 00
2
20
1
10
1
1ritriii EEEE +÷÷
ø
öççè
æ=÷÷
ø
öççè
æ-÷÷
ø
öççè
æq
m
eq
m
eq
m
e
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 84
P a g e | 84
or úú
û
ù
êê
ë
é
÷÷ø
öççè
æ-÷÷
ø
öççè
æ=
úú
û
ù
êê
ë
é
÷÷ø
öççè
æ+÷÷
ø
öççè
ætiitir EE q
m
eq
m
eq
m
eq
m
ecoscoscoscos
2
2
1
10
2
2
1
10
ti
ti
Ni
r
E
E
qm
eq
m
e
qm
eq
m
e
coscos
coscos
2
2
1
1
2
2
1
1
0
0
÷÷ø
öççè
æ+÷÷
ø
öççè
æ
÷÷ø
öççè
æ-÷÷
ø
öççè
æ
=÷÷ø
öççè
æ\ …(4)
The equation gives the amplitude of electric field of reflected wave w.r.t that of incident
wave. Similarly, eliminating E0r from equations(1) and (3), we have
ttitiii EEEE 0
2
200
1
10
1
1 cos][coscos qm
eq
m
eq
m
e÷÷ø
öççè
æ=-÷÷
ø
öççè
æ-÷÷
ø
öççè
æ
ttiii EE 0
2
2
1
1
1
10 coscoscos2
úú
û
ù
êê
ë
é
÷÷ø
öççè
æ+÷÷
ø
öççè
æ=÷÷
ø
öççè
æq
m
eq
m
eq
m
e
or
ti
i
i
t
E
E
qm
eq
m
e
qm
e
coscos
cos2
2
2
1
1
1
1
0
0
÷÷ø
öççè
æ+÷÷
ø
öççè
æ
÷÷ø
öççè
æ
=÷÷ø
öççè
æ
Therefore,
ti
i
Ni
t
E
E
qm
eq
m
e
qm
e
coscos
cos2
2
2
1
1
1
1
0
0
÷÷ø
öççè
æ+÷÷
ø
öççè
æ
÷÷ø
öççè
æ
=÷÷ø
öççè
æ …(5)
Where the index N indicates that iE0 is normal to the plane of incidence. Equations (4) and
(5) are the two Fresnel’s equations. Equation (5) gives the amplitude of electric field of
refractive wave w.r.t that of incident wave.
4.4.Brewster Angle and Degree of Polarization:
According to the Fresnel’s equation, we have
i
t
ttii
ttii
ti
ti
Pi
r
n
n
nn
nn
E
E
q
q
qqqq
qqqq
qq
qq
sin
sin as
cossinsincos
cossinsincos
cos)(cos
cos)(cos
2
1
21
21
0
0
=+
+-=
+
+-=÷÷
ø
öççè
æ
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 85
P a g e | 85
)tan(
)tan(
it
it
qq
qq
+
-= …(1)
Equation (1) shows that when ¥=+ )tan( it qq i .e., 2)( pqq =+ it , 00
0 =÷÷ø
öççè
æ
Pi
t
E
E. Thus,
when reflected and refracted rays are perpendicular to each other, there is no energy carried
by the reflected ray. The condition of continuity at the interface are then satisfied by two
waves only-the incident and transmitted waves- instead of the usual three. The angle of
incidence is then called the Brewster angle Bq . At the Brewster angle
( )
B
B
B
t
i
t
n
n
q
qp
q
q
q
q
sin
2sin
sin
sin
sin
sin
2
1 -===
Bqcot=
÷÷ø
öççè
æ=\ -
2
11cotn
nBq …(2)
If an unpolarized wave is incident on the interface at the Brewster angle, then the only
portion of the wave with electric vector perpendicular to the plane of incidence will be
reflected. Thus the reflected wave will be linearly polarized with vibrations perpendicular to
the plane of incidence and hence the Brewster angle is sometimes called as polarization
angle. For light incident on glass with refractive index 1.6, Brewster angle is about 58° and
at tq = 32°. Thus if l ight incident on glass at an angle 58°, the reflected light will be plane
polarized with vibrations perpendicular to the plane of incidence and the transmitted light is
also plane polarized with vibrations parallel to the plane of incidence as shown in fig.4.4.
Fig.4.4. Showing the reflection and transmission at Brewster angle.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 86
P a g e | 86
At radio frequencies the refractive index of water is 9 and the Brewster angle is 84°
and tq = 6°, so water will not reflect a vertically polarized radio wave when the angle of
incidence is 84°.
It is observed that even if the light is unpolarised and is incident at angles other than
Brewster’s angle, there is a tendency of reflected wave to be predominantly polarized with
vibrations perpendicular to the plane of incidence. In such a case we define the degree of
polarization as
Degree of Polarization =PN
PN
RR
RR
-
+ …(3)
Where NR and PR are reflection coefficients for the normal and parallel components
of reflected light. These coefficients are given by,
( )( ) )(sin
)(sin
coscos
coscos2
22
21
21
ti
ti
ti
tiN
nn
nnR
qq
qq
qq
qq
+
-=ú
û
ùêë
é
+
-= ….(4)
and ( )( ) )(tan
)(tan
coscos
coscos2
22
12
21
ti
ti
i
tiP
nn
nnR
qq
qq
qq
qq
+
-=ú
û
ùêë
é
+-
+-= …(5)
4.5. Total Internal Reflection:
Suppose a plane electromagnetic wave is incident on the interface from a denser
medium to a rarer medium. The angle of incidence iq is in denser medium and angle of
refraction tq is rarer medium. Here the word internal implies that the incidence and reflected
waves are in a medium of large refractive index than the refracted wave, i.e, 21 nn > .
Now according to Snell’s law,
ti
t
i
n
n
n
nqq
q
qsinsinor
sin
sin
1
2
1
2 == …(1)
ti qq < because 12 / nn is less than one. The value of tq becomes 90° for a particular value of
angle of incidence iq . This value of angle of incidence for which iq becomes 90° is known
as critical angle and is denoted by cq . Hence, from equation(1),
1
2
1
2 90sinsin n
n
n
nc =°=q …(2)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 87
P a g e | 87
Thus for waves incident at ci qq = , the refractive wave is propagated parallel to the
surface. Now the problem is that what will happen if the angle of incidence iq is greater than
tq . To answer it let us consider equation(1), i.e.,
)(1sin
sinsinsin
12
ci
c
iit as
nnqq
q
qqq >>== …(3)
As we know that the sine of any real angle can never be greater than one, equation(3)
suggests that tq is imaginary, i.e., tq is complex angle with a purely imaginary cosine.
( ) jbjc
itt =
ïþ
ïýü
ïî
ïíì
-÷÷ø
öççè
æ=-=
21
2
2 1sin
sinsin1cos
q
qqq …(4)
In order to find the meaning of imaginary tq , let us consider the transmitted electromagnetic
wave
( ){ }[ ]trnjEE ittt wmew -= .exp 220 …(5)
Consider that the transmitted wave is in x-z plane. Then,
tt
ttii
zx
kzjyixkirn
qq
qq
cossin
)).(cossin(.
+=
+++=
Now eq.(5) can be written as,
( ){ }[ ])cossin(exp 220 tttt zxtjEE qqmeww +--= …(6)
Using equations (3) and (4) equation (6) can be written as
( ) ( )úúû
ù
êêë
é
þýü
îíì
-÷÷ø
öççè
æ--= zjbxtjEE
c
itt 22220
sin
sinexp mew
q
qmeww
or ( )[ ] ( )úúû
ù
êêë
é
þýü
îíì
÷÷ø
öççè
æ--´=
c
itt xtjbzEE
q
qmewwmew
sin
sinexpexp 22220 …(7)
Equation (7) shows that for ci qq > , the transmitted wave is propagated only parallel to the
surface and is attenuated exponentially beyond the interface.
Energy flow in second medium:
The time-averaged normal component of Poynting vector just inside the surface gives
the energy flux carried by transmitted wave per unit area per second. Here we will show that
the lime averaged normal component of the Poynting vector just inside the surface is zero,
Hence the net energy flow through the surface into second medium is zero. Now, we have
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 88
P a g e | 88
{ }[ ]
part) real no hasit 0(as
).()(expRe. 2
1
.cos .Re. 2
1
.cos)( ofpart Real 2
1.
2
220
0
1
*
0
1
*
=
-÷÷ø
öççè
æ=
úú
û
ù
êê
ë
é
÷÷ø
öççè
æ=
´=Ù
njbbzE
nEE
nHEnP
t
ttt
tttt
mewm
e
qm
e
q
Intensity ratio of reflected wave to incident wave:
Now let us consider about the fields with the help of Fresnel's equation using the
values of sin tq and cos tq given by equations (3) and (4) respectively. In the first case when
E is normal to the plane of incidence, we have
2222
22
2
12
i
21
21
21
0
0
2
)(
)(
)(
)(
1sin
b
cos)( where
cos)(
coscos)(
ba
abj
ba
ba
jba
jba
jba
jba
nn
nnajba
jba
osnn
nn
E
E
i
ti
ti
Ni
r
+-
+
-=
-
-´
+
-=
úú
û
ù
êê
ë
é-÷÷
ø
öççè
æ=
=+
-=
+
-=÷÷
ø
öççè
æ
q
q
qq
qq
Let us consider
),sin(cos]exp[ NNN rjrjrjba
jbafff -==
+
- …(11)
Comparing equations (10) and (11), we have
22
22
22cos and
2sin
ba
bar
ba
abr NN
+
-=
+= ff
Which shows that r=1 and 2
tan22 ba
abN
+=f
Thus from equation(11), we have
a
be
jba
jba Nj N =÷ø
öçè
æ=
+
- -
2with tan
ff …(12)
Similarly when E is parallel to the plane of incidence, we have
…(10)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 89
P a g e | 89
cos)( where
coscos)(
coscos)(
cos)(cos
cos)(cos
12
12
12
21
21
0
0
i
ti
ti
ti
ti
Pi
r
nnajba
jba
nn
nn
nn
nn
E
E
q
qq
qq
qq
qq
=¢+¢
-¢=
+
-=
+
-=÷÷
ø
öççè
æ
or
a
b
2with tan P
0
0
¢=÷
ø
öçè
æ=÷÷
ø
öççè
æ ffPj
Pi
r eE
E
….(13)
It is clear from equation(12) and (13) that the amplitude of reflected wave is the same as that
of incident wave, i.e, the intensity of reflected wave is same as that of the incident wave,
showing thus that the wave is totally reflected. The phases of normal and parallel reflected
waves are denoted by Nf and Pf respectively, and both are the function of angle of incidence.
4.6. Reflection and Refraction at the boundary of a non-conducting and a conducting
media (i.e., metallic reflection) :
Let us consider the case of reflection and refraction at the boundary of non-
conducting and a conducting media. Here we shall modify the results of reflection and
refraction at the boundary of two non-conducting media by changing the second non-
conducting medium by a conducting medium. Let the non-conducting medium
(medium 1) has permittivity 1e and permeability 1m and the conducting medium bas
permittivity 2e permeability 2m and electric conductivity 2s . When a monochromatic
electromagnetic wave is propagating in medium 1, it is observed that there exists two more
waves (i) reflected wave propagating in medium 1 and (ii) a transmitted wave propagating in
medium 2. The electric and magnetic field intensities are described by the following
equations:
For non-conducting medium
( ){ }[ ]( ){ }[ ]ïî
ïíì
-=
-=
trnjHH
trnjEE
iiiioii
iiiioii
wmew
wmew
. exp
. exp
11
11
( ){ }[ ]( ){ }[ ]ïî
ïíì
+-=
+-=
AtrnjHH
AtrnjEE
rirrorr
rirrorr
wmew
wmew
. exp
. exp
11
11
For conducting medium
1(a)
1(b)
2(a)
2(b)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 90
P a g e | 90
( ){ }[ ]( ){ }[ ]ïî
ïíì
+-=
+-=
BtrnjHH
BtrnjEE
tittott
tittott
wmew
wmew
. exp
. exp
11
11
where ri is the position vector of the point of incidence, ni,nr and nt are the unit vectors in the
directions of incidence, reflection and transmission. A and B are the phase difference of
reflected and transmitted waves with incident wave.
We can investigate the properties of both the reflected and transmitted waves using
the fact that the tangential components of both electric field E and magnetic field intensity H
must be continuous across the interface. Following points are observed :
tri www == …(4)
i e., all the three waves must be of the same angular frequency.
ni,nr and nt are coplanar
ri qq = …(5)
ti qqmew sink sin( 2)11 =
or 1
2
)11
2
)11
2
((sin
sin
n
n
c
ckk
t
i ===me
w
mewq
q …(6)
where c is the phase velocity of the wave in free space, n2 is the refractive index of medium
2, and n1 refractive index of medium -1. Equation (6) is known as Snell's Jaw of refraction. It
should be noted that n2 is a complex quantity and is given by
)( 222 baw
jc
n += …(7)
Where
21
21
222
2222
2 112 ú
ú
û
ù
êê
ë
é+
þýü
îíì
+÷ø
öçè
æ=
we
smewa ….(8)
and
21
21
222
2222
2 112 ú
ú
û
ù
êê
ë
é-
þýü
îíì
+÷ø
öçè
æ=
we
smewb …(9)
The values of 22 and ba are taken from the articles of propagation of E.M.W. in conducting
media. Substituting these values in equation(7) we have
…(10)
In case of good conducting medium equation (10) reduces to
3(a)
3(b)
úúú
û
ù
êêê
ë
é
úú
û
ù
êê
ë
é-
þýü
îíì
+÷ø
öçè
æ+
ïþ
ïý
ü
ïî
ïí
ì
úú
û
ù
êê
ë
é+
þýü
îíì
+÷ø
öçè
æ=
21
21
222
2222
21
21
222
2222
2 112
112 we
smew
we
smew
wj
cn
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 91
P a g e | 91
úú
û
ù
êê
ë
é
÷÷ø
öççè
æ÷ø
öçè
æ+÷÷
ø
öççè
æ÷ø
öçè
æ=
we
smew
we
smew
w 2
222
2
2222
22j
cn
or )1(2
222 jcn +÷÷
ø
öççè
æ=
w
ms …(11)
From Snell’s law, we have
itn
nqq sinsin
2
1=
and ïþ
ïýü
ïî
ïíì
÷÷ø
öççè
æ-= it
n
nqq sin1cos
2
2
1 …(12)
since, the refractive index of a conducting medium is very large as compared to the refractive
index of a non conducting medium(n2>>n1) hence inn q2221 sin)( is negligible as compared
to unity and we obtain
1cos =tq …(13)
i.e, tqcos is real and is approximately equal to unity. Expression (13) shows that the angle of
refraction in a good conducting medium is very close to zero irrespective of the angle of
incidence. Thus in a good conducting medium the transmitted wave is propagated n the
normal direction.
The transmitted wave can be represented by
[ ] [ ][ ] [ ]î
íì
+--=
+--=
).(exp).(exp
).(exp).(exp
2
22
BtrnjrnHH
BtrnjrnEE
ititott
ititott
wab
wab
This shows that the transmitted wave is a highly damped wave.
The depth of penetration is given by
÷÷ø
öççè
æ=
ïþ
ïý
ü
ïî
ïí
ì-÷÷
ø
öççè
æ+÷÷
ø
öççè
æ=
wms
me
s
mewb
22
21
21
222
22
222
2
11211
and is independent of angle of incidence.
4.7.BASIC CONCEPTS OF WAVE GUIDES
In this section we seek the solution of Maxwell's equations in a cylindrical region of
space of arbitrary cross-section. Let a region of non-conducting space be bound by an
14(a)
14(b)
…(15)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 92
P a g e | 92
infinitely long perfectly conducting tube of an arbitrary cross- sectional shape, the cross
sectional area and shape of the tube being uniform throughout, The bounded region is
assumed to be lossless region. Such a structure which guides the transmission of
electromagnetic energy in the direction of the axis is called a waveguide. Thus by the term
waveguide we mean a cylindrical pipe bound by a conductor of high conductivity filled with
a dielectric (of low loss) having arbitrary cross section. The two most common waveguides
are those with rectangular and circular cross sections, but in this article we shall consider a
general case.
When an electromagnetic wave is introduced in the bound region, it undergoes
multiple reflections from the wall of the tube and propagates along its axis, say, along z-axis
as shown in fig. 4.7
Fig.4.7.Cylindrical Waveguide
The electric field E and magnetic field intensity H at a point in the bound region
satisfy the following Maxwell's equations
t
HcurlE
¶
¶-= m …(1)
t
EcurlH
¶
¶= e …(2)
Div E=0 …(3)
DivH=0 …(4)
Here E and H are the functions of position coordinates a 1, a2, z and time t.
In rectangular coordinates a 1, a2, are x and y and, in cylindrical coordinate system, they
are r and q respectively, Evidently a1, and a2, are the coordinates lying in a plane normal
to z direction, i.e., in the cross section of the tube. The time dependence of E and H is
often described by tje w- and hence Maxwell's equations reduce to
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 93
P a g e | 93
t
tzaaHtzaaCurlE
¶
¶-=
),,,( ),,,( 21
21 m
or tjtj et
zaaHezaaCurlE ww m --
¶
¶-=
),,,( ),,,( 21
21
or tjtj ezaaHjezaaCurlE ww wm -- --= ),,( )(),,( 2121
or ),,( ),,( 2121 zaaHjzaaCurlE wm= …(5)
Similarly,
),,( ),,( 2121 zaaEjzaaHCurl we-= …(6)
div 0),,( Div 21 =zaaE .(7)
div 0),,(H Div 21 =zaa …(8)
Taking the curl of equation (5) we have,
)],,( curl ),,(curl[ 2121 zaaHjzaaCurlE wm=
),,( )( 21 zaaEjj wewm -= (from Eq(6))
),,( 212 zaaEmew=
),,(),,(curl curl 212
21 zaaEzaaE mew=\
),,(),,(),,( div grad 212
212
21 zaaEzaaEzaaE mew=Ñ-
or 0),,( div as ),,(),,( 21212
212 ==Ñ- zaaEzaaEzaaE mew
0),,(),,( 212
212 =+Ñ\ zaaEzaaE mew ….(9)
Similarly, 0),,(),,( 212
212 =+Ñ zaaHzaaH mew ….(10)
It is again useful to single out the spatial variation of ),,( 21 zaaE and ),,( 21 zaaH in the z-
direction. We assume that
jkz
jkz
eaaHzaaH
eaaEzaaE
),(),,(
),(),,(
2121
2121
=
=
where k is guide propagating constant. Applying this equation to equation(9) we get,
0),(),( 212
212 =+Ñ jkzjkz eaaEeaaE mew
or 0),(),( 212
212
22 =+÷÷
ø
öççè
æ
¶
¶+Ñ jkzjkz
t eaaEeaaEz
mew
or 0),()(),( 2122
212 =-+Ñ jkzjkzt eaaEkeaaE mew
or 0),(),( 212
212 =¢+Ñ aaEkaaEt …(11)
where 2tÑ is transverse component of 2Ñ and )( 222 kk -=¢ mew .
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 94
P a g e | 94
The value of 2
2
2
22
yxt
¶
¶+
¶
¶=Ñ in rectangular coordinate system and
2
2
2
2 11
q¶
¶+÷
ø
öçè
æ
¶
¶
¶
¶=Ñ
rrr
rrt in cylindrical coordinate system.
It is useful to separate the fields ),( 21 aaE and ),( 21 aaH into components parallel to
and transverse to z-axis. Thus
),(),(),( 212121 aaEaaEaaE zt += …(12)
and ),(),(),( 212121 aaHaaHaaH zt += …(13)
where the parallel components are given by
[ ]maaEmaaE z ),(.),( 2121 = …(14)
and [ ]maaHmaaH z ),(.),( 2121 = …(15)
the transverse components are given by
[ ] maaEmaaEt ´´= ),(),( 2121 …(16)
and [ ] maaHmaaH t ´´= ),(),( 2121 …(17)
where m is unit vector in the z-direction. Manipulation of the curl equations (5) and (6) and
the use of the z-dependence jkze lead to the determination of the transverse component in
terms of the axial components, thus
[ ]),(),()(
1),( 21212221 aaEmjaaHjk
kaaH ztztt Ñ´+Ñ
-= we
mew …(18)
and [ ]),(),()(
1),( 21212221 aaHmjaaEjk
kaaE ztztt Ñ´-Ñ
-= we
mew …(19)
where tÑ is the transverse gradient operator. It is ÷÷ø
öççè
æ
¶
¶+
¶
¶
yj
xi in rectangular coordinate
system. Equations (18) and (19) show that transverse components can be calculated from the
axial components. Therefore, it is sufficient to determine the axial components
),( 21 aaE z and ),( 21 aaH z as approximate solutions of equations
0),(),( 212
212 =¢+Ñ aaEkaaE zzt …(A)
and 0),(),( 212
212 =¢+Ñ aaHkaaH zzt …(B)
At the perfectly conducting bounding surface of interest, the field ),,,( 21 tzaaE and
),,,( 21 tzaaH satisfy the boundary conditions:
(i) Tangential component is continuous i.e.,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 95
P a g e | 95
0),,,( 21 =´s
tzaaEn …(20)
(ii) Normal Component is continuous i.e.,
0),,,(. 21 =s
tzaaHn …(21)
Where n is a unit vector normal to the surface. The boundary condition (20) is equivalent to
0),( 21 =sz aaE …(C)
For normal component of ),,,( 21 tzaaH we have,
0),,(. 21 =-
s
tjezaaHn w
or 0),(. 21 =-
s
tjjkzeeaaHn w
or 0),(. 21 =s
aaHn
or 0),(),(. 2121 =+szt aaHaaHn
or 0),(.),(. 2121 =+szst aaHnaaHn
or [ ] 0),(),()(
1. 212122
=Ñ´++Ñ- sztszt aaEmjaaHjk
kn we
mew
because 0),(. 21 =sz aaHn
or 0),(. 21 =Ñszt aaHn
or 0),(. 21 =¶
¶
s
z aaHn
…(D)
where n¶¶ is the normal derivative at a point on the surface.
Because the boundary conditions ),( 21 aaE z and ),( 21 aaH z are different, they can not be
satisfied simultaneously. Consequently, the electromagnetic fields may be divided into
distinct categories.
(i) Transverse electric (TE) Mode:
From the boundary conditions it is clear that either ),( 21 aaE z is present
or ),( 21 aaH z or both. When ),( 21 aaH z is present and ),( 21 aaE z =0 everywhere, such waves
are termed as transverse electric waves(TE Mode). The boundary conditions as
0),( 21 =¶
¶aaH
nz …(22)
The transverse components are given by
),()(
),( 212221 aaHk
jkaaH ztt
®
Ñ-
=mew
…(23)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 96
P a g e | 96
and ),()(
),( 212221 aaHmk
jaaE ztt
®
Ñ´-
=mew
we (24)
(ii) Transverse magnetic TM Mode:
In this case, ),( 21 aaE z is present and ),( 21 aaH z =0 i.e., there is component of E in the
direction of propagation but no component of H. the boundary condition is
0),( 21 =- aaE z …(25)
The transverse components are given by,
),()(
),( 212221 aaEmk
jaaH ztt
®
Ñ´-
=mew
we …(26)
and ),()(
),( 212221 aaEk
jkaaE ztt
®
Ñ-
=mew
…(27)
(iii) Transverse electromagnetic (TEM) Mode:
There is a degenerate mode called the transverse electromagnetic (TEM) mode in
which both ),( 21 aaE z is present and ),( 21 aaH z vanish. We shall find that in case of hollow
conducting pipes either TE or TM waves can propagate but TEM waves do not propagate.
In case of waveguide the only problem is to solve the equations A and B for different
co-ordinate systems by applying the boundary conditions C and D. Now we shall consider
rectangular wave guide.
4.8.Rectangular Waveguide:
When the cross section of the metal tube is rectangular, it is called a rectangular
waveguide. The rectangular wave guide is shown in figure 4.8. It has cross sectional lengths a
and b, and the z-axis parallel to the axis of the tube. The conductivity of the metal is assumed
infinite, and the interior filled with a loss free dielectric.
Fig.4.8.Rectangular waveguide. (a). TE wave in rectangular waveguide:
In this case
0),( =yxE z everywhere …(1)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 97
P a g e | 97
),( yxH z satisfies the wave equation
( )22222 k where0),(),( kyxHkyxH zzt -=¢=¢+Ñ mew
0),(),( 2
2
2
2
2
=¢+÷÷ø
öççè
æ
¶
¶+
¶
¶yxHkyxH
yxzz …(2)
The boundary conditions for ),( yxH z are
úúúú
û
ù
êêêê
ë
é
===¶
¶
===¶
¶
bx
yxHand
axxx
yxH
z
z
y and 0yat 0),(
and 0at 0),(
…(3)
Now equation (2) is
0),(),(),( 2
2
2
2
2
=¢+¶
¶+
¶
¶yxHk
y
yxH
x
yxHz
zz …(4)
Let XYyYxXyxH z == )()(),( …(5)
Where X is a function of x and Y is a function of y,
Substituting equation(5) into equation(4) we have,
02
2
2
2
2
=¢++ XYkdy
YdX
dx
XdY
or 011 2
2
2
2
2
=¢++ kdy
Yd
Ydx
Xd
X
or 2
22
2
2 11
dy
Yd
Yk
dx
Xd
X-=¢+ …(6)
The left hand side is a function of x while right hand side is a function of y and both
are equal to each other. This is only possible when they are separately equal to a constant.
Hence,
2
2
22
2
2 11A
dy
Yd
Yk
dx
Xd
X=-=¢+ …(7)
Thus we should solve two equations
(i) 22
2
21Ak
dx
Xd
X=¢+
or 0)(1 22
2
2
=-¢+ Akdx
Xd
X
or 02
2
2
=+ XBdx
Xd …(8)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 98
P a g e | 98
where )( 222 AkB -¢=
The solution of equation (8) is
BxCBxCX sincos 21 += …(9)
Where C1 and C2 are two constants.
(ii) 2
2
21A
dy
Yd
Y=-
or 02
2
2
=+ YAdy
Yd …(10)
The solution of equation(10) is given by
AyCAyCY sincos 43 += …(11)
Where C3 and C4 are two constants.
Now we shall apply the boundary conditions (3) to eqs (9) and (11). Conditions(3) are
equivalent to
by
Yaxx
x
X===
¶
¶===
¶
¶y and 0yat 0 and and 0at 0
Differentiating eq.(9) w.r.to x we get,
BxBCBxBCx
Xcossin 21 +-=
¶
¶
When x=0, C2B=0. This gives C2=0
When x=a, -C1BsinBa+0=0 i.e., C1BsinBa =0
( )alBlBaC pp ===\¹ or or 0sinBa 01
where l =0,1,2,3…
xa
lCX
pcos1= …(12a)
Again differentiating eq.(11) w.r.to y, we get,
AyACAyACy
Ycossin 43 +-=
¶
¶
Where y=0, C4A=0. This gives C4=0
When y=b, -C3AsinAb+0=0 i.e., C3AsinAb =0
( )blAlC pp ¢=¢==\¹ or Abor 0sinAb 03
where l ¢=0,1,2,3…
yb
lCY
p¢= cos3 …(12b)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 99
P a g e | 99
Finally, the solution becomes
b
yl
a
xlCCyxH z
pp ¢= coscos),( 31
or b
yl
a
xlHyxH z
pp ¢= coscos),( 0 …(13)
hence, ll ¢, are two parameters and a set of given values ll ¢, defines a waveguide mode which
we call ll ¢, waveguide mode for TE wave designated as llTE ¢, mode.
Cut off frequency and Wavelength:
úû
ùêë
é ¢+=÷
ø
öçè
æ ¢+÷
ø
öçè
æ=+=¢
2
2
2
22
22
222
b
l
a
l
b
l
a
lABk p
pp …(14)
where the indices ll ¢ and specify the mode.
The cut off wavelength is given by
÷ø
öçè
æ=ú
û
ùêë
é ¢+=÷÷
ø
öççè
æ
¢l
p
l
2
2
11 21
2
2
2
2
kb
l
a
l
llc
Q …(15)
While cut off frequency will be
( ) ÷ø
öçè
æ=ú
û
ùêë
é ¢+=¢
l
pwpw
c
b
l
a
lcll
2
2
2
2
2
Q …(16)
The modes corresponding to ll ¢ and are designated as llTE ¢, mode. The case 0 =¢= ll gives a
static field which do not represent a wave propagation. So TE00 mode does not exist. If a<b,
the lowest cutoff frequency results for 1 and 0 =¢= ll i.e.,
( )b
kor 01
ppw =¢=
b
c ….(17)
The TE01 mode is called the principal or dominant mode.
Phase Velocity and Group velocity:
The phase velocity pv is given by
0
k
kc
kv p ==
w ÷
ø
öçè
æ=
ck
w0 Q
( ) ( )[ ]
1
c22
0222
0
0
kkkk
ck
¢-=
¢-= ( )22
02 kkk ¢-=Q
( )[ ]
1
2
0 c
p
cv
ll-=\ ÷
ø
öçè
æ=
l
p2 kQ ….(a)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 100
P a g e | 100
where ¥== pc v,0 ll , i.e, phase velocity becomes infinite at cut off. The group velocity
with which energy is propagated along the axis of the guide will be given by
( )
( )
úú
û
ù
êê
ë
é
÷÷ø
öççè
æ ¢-=
=¢+=
úûù
êëé ¢+
¶
¶=
¶
¶=
-
2
0
2
0
21
22
21
22
1
2 2
1
k
kc
ckkkkc
kkckk
vg
m
w
1
2
0
úú
û
ù
êê
ë
é
÷÷ø
öççè
æ-=\
c
g cvl
l …(b)
Transverse components of Electric field and Magnetic field intensity
The transverse components of electric field and magnetic field intensity may now be
calculated as follows:
coscos)(
),()(
),(
022
22
b
yl
a
xlH
yj
xim
k
j
yxHmk
jyxE ztt
pp
mew
wm
mew
wm
¢÷÷ø
öççè
æ
¶
¶+
¶
¶´
-
-=
Ñ´-
-=
®
úû
ùêë
é ¢+
¢¢
-
-=
¢÷÷ø
öççè
æ
¶
¶-
¶
¶
-
-=
b
xl
a
xl
a
lj
b
yl
a
xl
b
li
k
Hj
b
yl
a
xlH
yi
xj
k
j
pppppp
mew
wm
pp
mew
wm
cossinsincos)(
coscos)(
22
0
022
…(18)
From this equation, the components ),(E and ),( yxyxE yx can be calculated.
Similarly,
coscos)(
),( 022 b
yl
a
xlH
yj
xi
k
jkyxH t
pp
mew
¢÷÷ø
öççè
æ
¶
¶+
¶
¶
-=
úû
ùêë
é ¢÷ø
öçè
æ ¢+
¢÷ø
öçè
æ-
-=
b
yl
a
xl
b
lj
b
yl
a
xl
a
li
k
jkH pppppp
mewsincoscossin
)( 22
0 …(19)
with the help of equation(19) the components Hx ),(H and ),( yxyxH yz can be calculated.
(b) TM wave in rectangular wave guide:
In this case, 0),( =yxH z … (20)
),( yxE z satisfies the equation
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 101
P a g e | 101
0),(),( 2
2
2
2
2
=¢+÷÷ø
öççè
æ
¶
¶+
¶
¶yxEkyxE
yxzz …(21)
The boundary conditions for ),( yxE z are
úû
ùêë
é
===
===
byxE
axxyxE
z
z
y and 0yat 0),(
and 0or 0),( …(22)
This case can also be solved as the case of TE wave, hence we have,
b
yl
a
xlEyxE z
pp ¢= sinsin),( 0 …(23)
with úû
ùêë
é ¢+=¢
2
2
2
222
b
l
a
lk p
Cut off Wavelength: The cut off wavelength is given by
÷ø
öçè
æ=ú
û
ùêë
é ¢+=÷÷
ø
öççè
æ
¢l
p
l
2
2
11 21
2
2
2
2
kb
l
a
l
llc
Q …(24)
Cut off frequency: The cut off frequency is given by
( ) ÷ø
öçè
æ=ú
û
ùêë
é ¢+=¢
l
ppw
ck
b
l
a
lcll
2
2
2
2
2
Q …(25)
This implies that TE and TM modes of rectangular guide have the same set of cut off
frequencies. However, in this case 1or 0 and 1 =¢=¢= lll represents non-trivial solution since
this gives 0 ),( =yxE z and hence all the components of E and H will be zero; of course, this
mode i.e., 0 and 1 =¢= ll was dominant in TE mode. In TM mode the lowest mode has
1 =¢= ll and may be represented by TM11. The cut off frequency of lowest mode is given by
1a
11 2
1
2
221
2211 úû
ùêë
é+=úû
ùêë
é+=
b
ac
bac
ppw …(26)
Since a>b, the cut off frequency of lowest TM mode is greater than that of the lowest TE
mode by a factor ( ) 21221 ba+ .
4.9 Let us sum up
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 102
P a g e | 102
From this unit you have understood about boundary conditions at interfaces and
Reflection and Refraction at the surfaces which leads to Fresnel’s laws. Also we described
about the total internal reflection and critical angle. Finally you have learnt about the wave
guides with the particular reference to rectangular wave guide.
4.10 Lessons-end activities
(1) Write Fresnel’s equations
(2) What is meant by Brewster’s angle
(3) Define degree of Polarization
(4) Define reflection coefficient and Transmission coefficient between two non-
conducting media.
4.11 Points for discussion
(i) Prove that for glass-air interface (n1= 1.5 and n2= 1.0 ) for no0rmal incidence
the reflection and Transmission coefficient are 0.04 and 0.96 respectively.
[Hint : R = 2
21
221
)(
)(
nn
nn
+
-; T=
221
21
)(
4
nn
nn
+]
(ii) A hollow rectangular waveguide has a= 6 cm and b= 4cm. The frequency of
the impressed signal is 3GHz. Compute for TE10 mode (a) cut-off
wavelength (b) guide wavelength.
4.12 Check your progress
1. Explain total internal reflection and critical angle
2. Derive an equation for rectangular wave guide
4.13 Sources/References
1. Electrodynamics by Gupta, Kumar, Singh
2. Electromagnetic fields and waves by R.N.Singh
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 103
P a g e | 103
UNIT – III
Lesson – 5
Scattering Electromagnetic wave
5.0 Aims and Objectives
This unit deals with different types of scattering namely Thomson and
Rayleigh scattering and its importance in the interaction of electromagnetic
waves with matter. Also this unit deals with Normal and Anomalous
dispersion and its application to solids, liquids and gases.
5.1 Scattering and Scattering Parameters
If an electromagnetic wave is incident on a system of charged
particles, the electric and magnetic components of the wave will exert
Lorentz force on the charges and they will be set into motion. Since the
electromagnetic wave is periodic in time, so will be the motion of the
particles. Thus, there will be changes in the directions of motion and hence
there will be accelerations. The system will therefore radiate; that is energy
will be absorbed from the incident wave by the particles and will be re-
emitted into space in all directions. We describe such a process as
scattering of the electromagnetic wave by the system of charged particles. If
the energies of the incident and scattered radiations are equal the scattering
is called elastic otherwise inelastic.
Incident beam Electron Unpolarized Nucleus Scattered Radiation Fig. 5.1
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 104
P a g e | 104
Scattering is most conveniently characterized by following parameters
–
(A) Differential Scattering Cross-section: It is defined as the ratio of
the amount of energy scattered by the system per unit time per unit angle to
the energy flux density or intensity (i.e. energy per unit area per unit time in
a normal direction) of the incident radiations.
So if a solid angle d is subtended at the system of an area ds.
(fig 5.2 a) The mean power (i.e. energy per unit time) scattered by the system
will be given by
dPsr = Ssrds
where Ssr is the intensity of Scattered radiation. The mean energy scattered
per unit time per unit solid angle will therefore be
dPsr / d = Ssrds / d
or dPsr / d = Ssrr2 (as d = ds/ r2 ) -------(1)
If the incident energy flux density i.e. intensity (Poynting vector) is SIr, the
differential Scattering cross-section will be
d / d = dPsr/ d / SIr -------(A)
d / d = Ssrr2/ SIr
(Substituting the value dPsr/ d from Eq. 1)
From the above it is clear that the differential scattering cross section has
the dimensions of the area.
(B) Total Scattering Cross-section: We have defined the differential
scattering cross-section as
d /d = Ssrr2/ SIr
so d = Ssrr2 d / SIr
Therefore = (Ssr/ SIr) . ds (as d = ds/r2)
i.e. = Psr/ SIr [ as Psr = (Ssrds) ] --------(B)
is called the total scattering cross-section and is defined as the ratio
of the power scattered (total energy scattered per sec) to the intensity (energy
per unit are per unit time) of the incident radiations.
5.2 Scattering by a free electron (Thomson-Scattering)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 105
P a g e | 105
Let there be an electron of mass m and change q in the path of a
plane polarized monochromatic wave in vaccum. Both the electric and
magnetic vectors E and B will exert a force on q given by the Lorentz formula
F = q [E + v xB] --------(1)
Where v is the velocity of the particle produced by the wave, and assumed
throughout this treatment to be non-relativistic so that v << c. Because in
a plane wave
i.e B = n x E /c
so expression (1) becomes
F = q[E+(v/c) x (nxE) ]
F = qE (as v /c << 1 ) ---------(2)
Thus only action of the electric field on the charge need to be
considered. And so the equation of motion
F = ma = m (d2r / dt2 ) --------(3)
Substituting for F in (2) becomes m (d2r / dt2 ) = qE. -------(4)
And because in a plane wave
E = E0 e- i(ωt – k.r)
Equation (4) becomes
(d2r / dt2 ) = (q E0/m) e- i(ωt – k.r) ---------(5)
Equation (5) implies that the acceleration, velocity and displacement of the
particle are all in the same direction as E0 which itself is constant and that
the charge is oscillating sinusoidally.
Now if the incident electromagnetic wave in which the electric vector is
along x –axis (as it is plane polarised) is moving along z-axis, then the
acceleration in the x – direction will be given by
(d2x/dt2) = (q E0/mω2) e- i(ωt – k.z) --------(6)
So that the displacement x at time t will be given by
x = - (qE0 /mω2) e- i(ωt – k.z) --------(7)
Now as an oscillating charge behaves like an oscillating dipole with dipole
moment
p = qx
It follows from equation (7) that
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 106
P a g e | 106
p = (q2 E0 /m ω2) e- i(ωt – k.z)
or p0 = (q2 E0 /m ω2) -------(8)
But as the average energy radiated per sec. per unit area in a normal
direction by an oscillating dipole is given by
Ssr = (1/4 0) (ω4p02/ c 3r2) sin2θ
or Ssr = (1/4 0) (ω4/8 c 3r2) [q2 E0 /m ω2]² sin2 θ
(putting the value of p0 from eqn. 8)
i.e. Ssr = (1/4 0) (q4 E02/8 m2c3r2) sin2 θ --------(9)
Further as for a plane wave
SIr = E x H
So the average value of SIr will be given by
SIr = (½ E0H0) (as E is to H )
i.e SIr = (1/2 0cE02 ) (as H0 c 0 E0) ------(10)
so the differential scattering cross – section
(d /d ) = (Ssr r2/ SIr)
= (1/4 0) (q4 E02 sin2 θ/8 m2c3) x (2 / 0cE02)
= (q2/4 0mc2)2 sin2 θ
or (d /d ) = r02 sin2 θ = r02 cos2φ ------(A)
Where r0 = (q2/4 0mc2) and is called the classical radius of the electron.
The incident radiation here has been taken to be plane polarised.
For unpolarised or randomly polarised waves an average must be taken over
all x
ds
+ x θ d Y p
φ E0 o - x
(a) (b)
Fig. 5.2
orientations of the plane of E. Suppose in Fig. 5.2(b) AB is the direction of
E in another wave incident on the particle of the fig. (a), the angle between
AB and the plane of fig. (a), containing field point is ψ. It is now preferable
to express the scattering in terms of the angle φ which is common to all
B θ O A
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 107
P a g e | 107
azimuths. In fig. (b) the plane POB is drawn perpendicular to the plane
containing AB so that the length BQ is given both by r cos θ and by r sin φ
cos ψ.
Hence
cos θ = sin φ cos ψ .
i.e. cos2 θ = sin2φ cos2 ψ
i.e sin2 θ = 1 - sin2φ cos2 ψ
i.e sin2 θ = 1 - cos2 ψ (1 - cos2 φ) ------------------(11)
[For a plane polarised light as ψ = 0 )
sin2 θ = 1 – (1- cos2 φ) = cos2 φ
which is also evident from fig. 5.2a in which θ = (( / 2) –φ).
2 2
(cos2 ψ ) av = (1/2 ) 0 ( cos2 ψ dψ) = (1/2 ) 1/2 0 (1+ cos 2ψ ) dψ 2
= (1/2 ) 1/2 ψ + (sin2ψ/2) = 1/2 ] 0
Averaging equation (11) over all ψ, we get
sin2 θ = 1 – ½ (1- cos2 φ ) [as (cos2 ψ )av = ½]
i.e sin2 θ = ½ (1+ cos2 φ ) --------(12)
Substituting the value of sin2 θ from, equation (12) in (A), we get
d /d = r02 ½ (1+ cos2 φ) ------------------(B)
This is called Thomson formula for scattering of radiation and is
appropriate for the scattering of X – rays by electrons or γ (gamma) – rays by
protons. In it angle φ is called the scattering angle and the factor
½ (1+ cos2 φ) is called degree of dipolarisation. From expression (B) it is
clear that:
(i) Scattering of electromagnetic waves is independent of the nature
of incident wave, (i.e.ω).
(ii) Scattering occurs in all directions and is maximum when φ = 0
or (i.e. in the forward and backward directions) while minimum when
(φ = ) or (3 /2) (i.e. in the side way directions).
(iii) Scattering depends on the nature of the charged particle i.e.
scatterer and is symmetrical about the line given by (φ = /2).
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 108
P a g e | 108
A plot of dirrerential scattering cross-section as a function of
scattering angle φ for plane polarized and unpolarised incident radiations is
shown below.
1
(1/r02) d /d .5
0 /2
The total scattering cross-section will be
= (d /d ) d
[If the incident radiation plane polarised then
/2
= r02 sin2 θ d - /2 r02 sin2 θ 2 cos θ dθ
/2 /2
i.e = (4 r02) 0 (sin2 θ cos θ dθ) = (4 r02) [ (sin3θ/3)]0 = ((4/3) x r02 )
or = 0 (r02 cos2φ d ) (as (sin2 θ = cos2φ )
i.e = 0 (cos2φ 2 sinφ dφ ) = 2 r02 0 (cos2φ sinφ dφ )
= 2 r02[( - cos3φ/3)]0 = (4/3 x r02) ]
= r 02 ½ (1+ cos2φ ) d
= r02 0 ½ (1+ cos2φ ) 2 sinφ dφ ( as = 2 (1-cosφ)
= r02 0 (sinφ + cos2φ sinφ ) dφ
= r02 [ (- cosφ) - (cos3φ/3) ]0
i.e = (8 /3) x r02 -------------------------(C)
Result (C) was first derived by Thomson and so after his name it is
called Thomson scattering cross-section.
Unpolarised
Plane polarised
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 109
P a g e | 109
A quantum mechanical calculation carried out by Klein and Nishina
shows that deviations from Thomson result become significant for incident
photon energy hv which is comparable with or larger than the rest energies
of the scattering electron mc2. According to them
KN = r02 { (8 /3) (1- 2hv/mc2 + ..........)} for hv<< mc2
and KN = r02{ ( mc 2/hv [loge (2hv/ mc2 + ½ ) ]} for hv >> mc2
These results are shown graphically in figure given below
1
Hv/
1/r02 d /d .5
0
/2
From these curves it is clear that:
(i) The scattering depends on the nature of the incident radiations.
(ii) Quantum mechanical result approaches the classical one on the long
wavelength side as the frequency υ = ω/2 goes to zero.
(iii) The scattering is not symmetrical and radiations more concentrated in
the forward direction i.e. φ = 0).
Apart from these there is another feature of Thomson scattering which
is modified by quantum considerations. Classically the scattered radiations
has the same frequency as the incoming waves but quantum mechanical
calculations shows that the frequency of the scattered radiation is lesser
than that of incoming wave and depends on the angle of scattering. The
relation between the wavelength of the scattered radiation at an angle φ and
the incident radiation is
λs = λi + h/mc(1-cosφ)
hv/mc2 = 0 hv/mc2 = 0.2 hv/mc2 = 1
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 110
P a g e | 110
5.3 Scattering by a Bound Electron (Rayleigh Scattering).
Considering now the same charge as in the previous section but
elastically bound so that a restoring force – mω02x is brought into play when
it is displaced. We also assume that there is a small amount of damping
proportional to dx/dt which may be produced in particle by Collisions or
radiation. The equation of motion now becomes
md2x/dt2 = qE – mγ dx/dt - mω02x
where γ is the damping constant per unit mass. Thus
d2x/dt2 + γ (dx/dt)ω02x = (q/m) x E0 e-i(ωt – kz) -----------(1)
The solution of this differential equation consists of two parts-
(a) The complementary function: It is obtained by solving equation
d2x/dt2 + γ (dx/dt) + ω02x = 0 ---------------(2)
Let the solution of equation (2) be
x = A t -----------------(3)
so that dx/dt = A e t
and d2x/dt2 = A 2 e t
Substituting the value of x, dx/dt and d2x/dt2 from above in equation
(2) we get
2 + γ + ω02 = 0
i.e. = γ/2 ± (γ2/4) – ω02 )
or = γ/2 ± i (ω02 - γ2/4) (as ω0 >> γ)
So that = A1 e[ - (γ/2) + i (ω02 – (γ 2/4)] t + A2 [ - (γ/2) - i (ω02 – γ 2/4)] t
i.e x = e ( - (γ/2) [ A1eiβt + A2e-iβt ] ------(a)
with β = (ω02 - γ2/4). The constant A1 and A2 can be determined by
applying initial conditions.
(b) The perpendicular integral : It is obtained by solving
(d2x/dt2) + γ (dx/dt) + ω02 x = (q/m) E0e-i(ωt – kz) ---------(4)
Let the solution of equation (4) be
x = Be-i (ωt – kz) --------------------(5)
dx/dt = -iω Be-i (ωt – kz)
d2x/dt2 = – ω2 Be-i (ωt – kz)
Substituting these values of x,dx/dt and d2x/dt2 in equation (4) we get
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 111
P a g e | 111
- ω2 B + γ ( - 2ωB) + ω02B = (q/m) E0
or B = qE0/m[ω02 - ω2 – iγω]
so that x = q E0 / m [ω02 - ω2 – iγω] x e-i (ωt – kz) -----------------(a)
= q E0/[(ω02 - ω2) + iγω] /m [ω02 - ω2)2 + γ2 ω2] x e-i (ωt – kz)
(q E0 e-i (ωt – kz) )/ (m [ω02 - ω2)2 + γ2 ω2]1/2 ) { (ω02 - ω2) /[(ω02 - ω2)2 + γ2 ω2]1/2 )
+ i (γω) / [(ω02 - ω2)2 + γ2 ω2]1/2 }
= (q E0 e-i (ωt – kz) ) / [(ω02 - ω2)2 + γ2 ω2]1/2 ] x [ cos δ + isin δ ]
x = (q E0 e-i (ωt – kz) ) / (m [ω02 - ω2) + γ2 ω2]1/2 ) ----------(b)
with δ = tan-1 γω/ (ω02 - ω2)
So from expressions (a) and (b) we conclude that the general solution
of equation (1) will be
x = e-(γ/2) t [ A1eiβt + A2eiβt ] + q E0 / m(ω02 - ω2)2 + γ2 ω2]1/2 ] e-i (ωt – kz)
-------------(c)
In this solution first term on the R.HS represents the free damped
vibrations of the charge. These vibrations die out soon on account of the
factor e-(γ/2) t and hence the frist term can be neglected in considering the
final motion. So equation (c) reduces to
x = q E0 / m(ω02 - ω2)2 + γ2 ω2]1/2 ] e-i (ωt – kz) -----(d)
Now as an oscillating charge is equivalent to an induced electric dipole of
moment
p = qx
It follows from equation (d)
p= q2 E0 / m(ω02 - ω2)2 + γ2 ω2]1/2 ] x e-i (ωt – kz
or p0 = q2 E0 / m(ω02 - ω2)2 + γ2 ω2]1/2] -----------(6)
But as average energy radiated per sec per unit area in a normal
direction by an oscillating dipole is given by
S = (1/4 0) (ω4 p02/8 c 3r2) x sin2 θ
So for the present situation
Ssr = (1/4 0) (ω4 p02/8 c3r2) q4 E02 sin2 θ / m2 [(ω02 - ω2)2 + γ2 ω2]
and as SIr ½ 0c E02
dσ/d = (Ssrr2 / SIr)
= (q2/(4 0mc2))2 (ω4 sin2 θ)/(ω02 - ω2)2 + γ2 ω2]
i.e. dσ/d = r02 (ω4 sin2 θ) / [(ω02 - ω2)2 + γ2 ω2]-------(A)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 112
P a g e | 112
Now if is the angle of scattering and the incident radiations are
unpolarised then as in previous section
sin2θ = 1/2 ( 1 + cos2φ ) -----------(7)
So equation (A) in the light of (7) becomes
dσ /d = ½ ( 1 + cos2φ )r02ω4 / [(ω02 - ω2)2 + γ2 ω2] ------(B)
This is the required result. From this, it is clear that –
(i) Scattering depends on the nature of the incident radiations – i.e. ω
(ii) Scattering depends on the angle of scattering φ
(iii) Scattering depends on the nature of the scatterer i.e. ω0 and γ
The total scattering cross-section will be
σ = ( dσ /d ) d
= ½ ( 1 + cos 2φ )r02ω4 / [(ω02 - ω2)2 + γ2 ω2] d
= r02ω4/[(ω02 - ω2)2 + γ2 ω2] 0 ½ ( 1 + cos2φ) 2 sin φ dφ
σ = (8 /3) (r02ω4) / [[(ω02 - ω2)2 + γ2 ω2] --------------(C)
The expression (C) gives the total scattering cross-section for an
elastically bound electron. From this, it is evident that total scattering
cross-section is a function of the frequency of incident radiations. Fig. 5.3 is
a plot of σ against ω. From the curve of fig. 5.3 or from equation C it is
clear that
8 /3 r02(ω0/γ )2 Resonance
σ
Thomson
8 /3 r02
ω4 Rayleigh
0 ω0 ω
Fig. 5.3
(i) If ω >>ω0 then σ › (8 /3) r02 as γ>0
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 113
P a g e | 113
i.e σ › σT
i.e Thomson Scattering occurs and the charge behaves as if it were free.
This is likely to occur at the X-ray end of the electromagnetic spectrum,
(ii) if ω ~ ω0 then σ = (8 /3) r02 (ω0/ γ)2
which is generally very large compared to Thomson scattering cross-section.
The scattering is known as resonance scattering. One example of resonance
scattering occurs when sodium vapour is illuminated with the characteristic
yellow sodium radiation. The whole volume of the vapour then becomes
luminous because it is strongly scattering radiations of its natural
frequency.
(iii) if ω<< ω0 then σ › (8 r02/3) (ω/ ω0)4
i.e σR › K/λ4 (With K = (8 02/3) λ04)
i.e the amount of scattered light is proportional to 1/ λ4 where λ is the
wavelength of the incident radiation. This scattering is known as Rayleigh
scattering. This is likely to occur when ω corresponds to the frequencies of
visible light and ω0 to that of ultraviolet.
In case of Rayleigh scattering shorter the wave length the more
strongly is the incident light scattered. This is the result that is used to
account for the blue colour of sky; the light entering the eye-has been
strongly scattered away from its direct path from the sun by air molecules
and since shortest visible wavelength corresponds to blue and since shortest
visible wavelength corresponds to blue light-Similarly, the red colour of a
sun-set or sunrise arises because the light coming to the eye through a
thick layer of air has had virtually all the blue light scattered out of this
direct path leaving only the long wavelength (red) component of the light to
be seen. Further as red light has longest wavelength in the visible region it
is scattered least. So it can travel longer distances in atmosphere and can
be seen from larger distances in comparison to other wavelengths of visible
region. This is why red light is used for danger signals while the eye is most
sensitive to yellow green.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 114
P a g e | 114
5.4 Coherence and incoherence in scattered light.
Scattering from real materials does not always exhibit the simple
features because the total mean intensity is the result of adding
contributions from large number of particles. The resultant intensity of a
number of waves is always found by adding the amplitudes taking due
account of the phases, and squaring the resultant. If the phase difference
between any two waves at a point in space is constant for all time, they are
said to be coherent, while if the phase difference varies in a completely
random way they are said to be incoherent. In practice a state somewhere
between the two extremes is common and the waves are then said to be
partially coherent
As we are concerned more with waves from different radiating
particles, all being excited by the same incident radiation, we might except
some degree of coherence. In general the resultant intensity of a set of
incoherent waves at any waves at any point is obtained by adding the
intensities. Thus if A1,A2 etc., are the amplitudes of the waves for a set of
sources then intensity I is
I = (A1 + A2+A3 + ……)2 …..(a1) for coherent sources
and I = A12 + A22 + A32 + ….(b1) for incoherent sources
The expressions (a1) and (b1) for N identical sources each of amplitude A
reduce to
I = N2A2 -------(a2) for coherent sources
and I = N A2 -------(b2) for incoherent sources
Let us now apply these ideas of coherence to the problem of scattering
by gas molecules. The phase of the scattered waves at the point P shown in
fig 5.4 arising from a molecule located at 1 will depend on the distance the
wave has travelled from the sources to 1 and from 1 of P. Since the
molecules of a gas have no permanent locations with respect to each other
their path lengths will be randomly related, their phases will be incoherent,
and we shall be justified in adding intensities as stated by equation (b1) or
(b2)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 115
P a g e | 115
This procedure would not be permissible in discussing the scattering
of X-rays in crystals, where the various atoms have fixed positions in the
lattice so one observes coherent scattering (i.e equation a1or a2 apply)
P
2
1 p
2 3
1
(a) 3 (b)
Fig. 5.4
The use of equations (b1) and (b2) would also not be valid if we were looking
at the radiations which are scattered from a gas and are observed in a
direction close to the incident direction as shown in fig 5.4(b) If we are far
enough in front, the path lengths are all about the same, the phases are
approximately equal and we get coherent scattering (i.e. interference) rather
than incoherent scattering.
5.5 Normal and Anomalous, dispersion
If in a medium the index of refraction varies with frequency (i.e
wavelength) then the medium is said to be dispersive, the phenomenon itself
is called dispersion and the rate of change of refractive index with
wavelength i.e. dn/dλ is known as dispersive power.
Generally the variation of n is such that
(i) The index of refraction increases as the frequency increases.
(ii) The rate of increase dn/dω i.e. the slope of the n – ω curve is
greater at high frequencies.
However it is also found that over small frequency range there is often
decrease of index of refraction with-the increase in frequency.
104
N A N A
N ω
n = 1
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 116
P a g e | 116
n = 96
A – Regions in which the dispersion is anomalous N – Regions in which the dispersion is normal
Fig. 5.5
In these narrow spectral regions due to its abnormal behaviours the
dispersion is called anomalous. Normal and anomalous dispersion are
shown graphically in fig. 5.5.
5.6 Dispersion in Gases (Lorentz Theory)
In order to investigate the frequency dependence of refractive index n
or dielectric constant r and to discuss dispersive effects, Lorentz assumed
that in case of gases –
(i) There is no appreciable interaction between the atoms in case of
atomic gases or between the molecules in case of molecular gases.
(ii) As an electromagnetic wave passes through a gas the electric field
induces dipole moment in the gas molecules.
(iii) In polarisation the positions of the electrons are altered from their
equilibrium value while nuclei remain stationary.
(iv) The electrons are bound to the nucleus in an atom by linear
restoring force.
(v) There is a damping proportional to the velocity of the electron.
(vi) Over an atom or a molecule E is constant in space i.e.
E = E0e-i(ωt – k.r) E0e-iωt
In the light of above assumptions the equation of motion of an electron will
be
m (d2r/dt2) +( mγ0 dr/dt) + (mω02r) = eE
or (d2r/dt2) +γ0 (dr/dt) + (ω02r) = (e/m)E --------(1)
This equation is already discussed in the previous hint and its solution was
found to be
r = ((e/m) E0e-iωt )/ (ω02 – ω2) - i γ0ω -----------(2)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 117
P a g e | 117
So the dipole moment which results from the displacement of the
electron under consideration
P = er = [(e2/m) E0e-iωt ] / (ω02 – ω2) - i γ0ω ----------(3)
Now if there are N electrons per unit volume in the gas, the
polarisation vector
P = Np
In the light of equation (3) is given by
P = N(e2/m) E /(ω02 – ω2) - i γ0ω --------(4)
In deriving the above equation we have assumed that there is only one
type of charge which is characterised by the constants ω0 and γ0. It is quite
reasonable to expect that the electrons are not all in identical situations
within molecules and that there should be different pairs of characteristic
frequencies ω0j and associated damping factors γ0j each pair reflecting that
particular environment in which the given type of electron is found. So if we
define fj as the probability that an electron has characteristic frequency ω0j
and damping coefficient γ0j then the generalization of equation (4) yields
P= (Ne2E/m) (Σj) fj / (ω0j2 – ω2) – iγ0jω
So the electrical polarisability
= (e2/m) Σj fj / (ω0j2 – ω2) – iγ0jω ( as = Pm/Em=NPm/NEm = P /NE)
i.e r = 1 + (Ne2/ 0m) Σj fj / (ω0j2 – ω2) – iγ0jω
(as r = 1+x0 = 1+P// 0E = 1+N / 0)
n2 = 1 + 1/(4 0) .(4 Ne 2/m) Σj fj / (ω0j2 – ω2) – iγ0jω ------(A)
(as n = r)
This is the required result which expresses the frequency dependence of r
or n. To study what this equation implies we consider the following
situations –
(A) Static Case: If ω>i.e , the frequency of the incident wave is very small in
comparison to the natural frequency of the electrons, equation (A) reduces to
n2 = 1 + 1/(4 0) .(4 Ne 2/m) Σj fj / (ω0j2 ----(a)
Equation (a) clearly shows that the index of refraction is a constant greater
than unity and depends on the nature of the medium i.e. ω0j
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 118
P a g e | 118
(B) Normal dispersion : if γ0j› 0 and ω < ω0j i.e. the region is remote from
the natural frequencies of the electrons (which is true for optical
frequencies) equation (A) reduces to
n2 = [ 1+1/(4 0) .(4 Ne 2/m) Σ (fj / ω0j2 – ω2] --------(b)
From equation (b) it is clear that the refractive index is real and increase
with frequency of the incident waves i.e. for a given medium red light has
the lowest while violet largest index of refraction in the optical range of
frequencies.
(c) Anomalous dispersion : If ω ω0j i.e. in the extremely narrow spectral
region in which the impressed frequencies include one of the so many
natural frequencies of the electron. For simplicity we assume that there is
one natural frequency i.e., ω0j = ω0 so that equation (A) becomes
n*2 = 1 + [1/(4 0) .(4 Ne2/m) ] x [1/((ω02 – ω2) – iγ0ω]
Now as the index of refraction of gases under normal conditions is
approximately unity so that expression (1 + y)1/2 = 1+ ½ y for y < 1 may be
employed to obtain
n* = [1 + 1/(4 0) ] x[ (2 Ne2/m) ] x 1/((ω02 – ω2) – iγ0ω
Multiplying numerator and denominator of the second term on RHS by its
complex conjugate
n* = 1 + [1/(4 0)] x [ (2 Ne2/m) ] [ (ω02 – ω2)+ iγ0ω] /[ω02 – ω2)2+ γ02ω2]
or n* = n + ik ---------(1)
with n = [1 + 1/(4 0)] x [ (2 Ne 2/m) ]x [ (ω02 – ω2) /( ω02 – ω2)2+ γ02ω2]
------–(2)
and k = [1/(4 0) ] x (2 Ne 2/m) x [ γ0 ω] /( ω0 – ω2)2 + γ02ω2] ------------
(3)
(Here Eq. (1), (2), (3) = c )
Equation (c-1) shows that the index of refraction is a complex function of
frequency of the electromagnetic waves propagating through the gas.
The real part of equation (c-1) i.e.., n is plotted as function of
frequency ω in fig. 5.6. At very low frequencies n is slightly greater than
unity, n increases with increasing ω, reaching a maximum at ω1 (ω0 - γ0/2)
falling rapidly to unity at (ω - ω0), n continues to decrease rapidly until (ω0
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 119
P a g e | 119
+ γ0/2) where upon it increases again and approaches unity asymptotically
for large values of ω.
γ0
n
96
k
ωj ω0 ω ω Fig. 5.6
The imaginary part of n* i.e., k is also plotted in fig. 5.6. It
corresponds to absorption of the electromagnetic waves propagating through
the gas. The imaginary part of n* has a typical resonance shape. k is
maximum at (ω = ω0), where n is unity and a width at half maximum
approximately equal to γ0. Therefore in regions where n changes rapidly, the
gas is relatively high absorbing.
For any real gas there exist many resonant frequencies ω0j and
corresponding damping coefficients γ0j so that
n* 1+[1/(4 0) ] x[ (2 Ne 2/m) ] Σ fr/(ω0j2 – ω2)2 + iγ0j2ω2 -----(d)
The behaviour of the real and imaginary parts of the equation (C) is
illustrated in fig. 5.6 (a)
n
k
ω0 1 ω0 2 ω0 3
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 120
P a g e | 120
ω
Fig. 5.6 (a)
It is worthy to note here that this classical theory cannot
predict the values of the resonance frequencies ω0j ; only a correlation of
observations relating to optical properties of matter can be attempted.
Quantum theory must be used for complete description and even though a
quantum calculation can be in principle yield values for the resonance
frequencies, the computations can be carried out exactly for the most simple
cases.
5.7 Dispersion in Liquids and Solids.
In the materials the molecules are sufficiently close to each other so
the effect of interactions among the molecules can no longer be neglected.
Since the material is polarised, we expect that the actual electric field on a
given charge will have a contribution from the polarisation and hence it will
be different from the applied field. The usual way of approximating this is to
imagine a small sphere centered at the position of the electron in question to
be cut out of the material as shown in fig 5.7 . The sphere is to be large
enough microscopically and small enough macroscopically so that the
material outside it can be described in terms of the continuous polarisation
vector P. Then because of the discontinuity in P there are bound surface
charges on the surface of spherical volume with density
σ' = P.n = P cos θ
So the contribution to the local field E' by the change on an area ds
d E' = σ'ds / 4 0R2
The components normal to the direction of P cancel, so integrating over all
the surface we find that local field E' parallel to P is equal to
2
E'= dE' cos θ = (P/ 4 0 ) 0 0 (cos2 θ R2 sinθ dθ dφ/ R2) = (P/3 0) -----(1)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 121
P a g e | 121
E
Fig. 5.7
The result is the contribution to the local field from the material
outside the sphere. We still have to calculate the contribution due to the
molecules within the small sphere. It can be shown that this contribution
averages to zero in an isotropic material such as a liquid or when the
molecules are arranged in an cubic lattice. Therefore if E is the applied
electric field, the total electric field acting on the charge is
ET = E + E' = (E + P/3 0) (using equ. 1)-------(2)
So the equation of motion of an electron in this case will be
m(d2r/dt2) + mγ0 (dr/dt) +(m 02r)= e [E + P/3 0]
m(d2r/dt2) + mγ0 (dr/dt) +(m 02r)= e [E + (Ner/3 0] (as P = Ner)
or (d2r/dt2) + γ0 (dr/dt) + (ω02 – (Ne2/3 0m) r = (e/m). E -------(3)
The solution of equation (3)
r = (e/m). E / [(ω02 – (Ne2/3 0m) - ω2 –iγ0ω]
So the dipole moment which results from the displacement of the
electron under consideration will be
þ = er = (e2/m). E / [(ω02 – (Ne2/3 0m) - ω2 –iγ0ω]
Now as there are N electron per unit volume
P = Np = N (e2/m). E/[(ω02 – (Ne2/3 0m) - ω2 –iγ0ω] ----------(4)
However as electrons are no all in identical situations with the
molecules, there should be different pairs of characteristic frequencies ω0j
P
+ P – + – R + θ – + θ – dE` + + –
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 122
P a g e | 122
and associated damping factors γ0j, each pair reflecting the particular
environment in which the given type of electron is found. So if we define fj
as the probability that an electron has characteristic frequency ω0j, and
damping coefficient γ0j, then the generalisation of equation (4) yields
P =(Ne2E/m) Σ fj / [(ω0j2 – (Ne2 fj /3 0m) - ω2 –iγ0ω]
i.e = (p/E) =(Np/NE =(P/NE)= (e2/m) Σj fj /[(ω0j2 – (Ne2 fj /3 0m) - ω2 –iγ0ω
5.8 Let us sum up:
This lesson described various methods adopted in scattering of electro
magnetic waves with particular reference to Thomson scattering (free
electron) and Rayeigh scattering (bound electron). Also you have understood
about normal and anomalous dispersion in gases, liquids and solids.
5.9 Lesson end activities
(i) Write a note on basic concepts about scattering and scattering parameters. (ii) What is meant by total scattering cross – section
(iii) Write the equation for the force in a plane EM waves
5.10 Points for discussion
(i) What is resonance scattering? Explain
(ii) The intensity of X-rays scattered by an electron at distance r
from it and making an angle θ write the original direction is
given by
2
Ie = e2/(4 0mc2) I ½(1+cos2 θ) / r2 where I is the intensity
of the incident beam. Calculate (a) the intensity of scattered
radiation Is interms of Ic for n electrons (i) when they scatter
independently (ii) when all of them act as a single scattering
centre.
5.11 Check your progress
(i) Write short notes on
(a) Normal and anomalous dispersion
(b) Dispersion in liquids and solids
(ii) Explain in detail
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 123
P a g e | 123
(a) Scattering by a free electron
(b) Scattering by a bound electron.
5.12 References
(i) Electrodynamics by Gupta, Kumar, Singh
(ii) Electromagnetic fields and waves by R.N. Singh
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 124
P a g e | 124
UNIT – III
Lesson - 6
COHERENCE AND INTERFERENCE
6.0 Aims and Objectives
We have so far assumed that light sources emit perfect harmonic
waves. In an ideal harmonic wave there exists a definite relationship
between the phase of the wave at a given time and at a certain time later;
and also at given point and at certain distance away. In reality, light
sources do not emit perfectly harmonic waves. Even a very best practical
monochromatic source emits a finite range of wavelengths and the light
waves are quasi-monochromatic. If it were not so, the light waves would
have been ideally coherent and interference would be observed at all times.
In practice, light is emitted form a light source when excited atoms
pass from the upper excited state to lower energy state. The process of
transition from upper state to a lower state lasts for brief time of about
10-8s. It means that an atom starts emitting a light wave as it leaves the
excited state and ceases emission as soon as it reaches the lower energy
state. Therefore, an emission event produces a light burst. Each light burst
occurs over a period of 10 -8s only, during which period a train of finite
length having a certain limited number of wave oscillations is generated.
Such a light burst is known as wave train or a wave packet. After some time
the atom again receives energy and jumps into excited state and
subsequently emits another burst of light. These emission events occur
quite randomly. Each atom in the source acts independently and different
atoms emit wave trains at different instants and their combination in
millions and millions constitutes the light from a light source. In order to
appreciate some of the peculiarities of natural light, the following fact is to
be well understood. The light emitted by an ordinary light source is not an
infinitely long, simple harmonic wave but is composed of jumble of finite
wave trains. We therefore call a real monochromatic source as a quasi-
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 125
P a g e | 125
monochromatic source. The wave trains issuing out of a quasi-
monochromatic source are as shown in Fig. 6.0
E (t)
t
Fig. 6.0
6.1 WAVE TRAIN
Fig. 6.1 shows a wave train generated by an atom
x
l
Fig. 6.1
If such a wave train lasts for a time interval t, then the length of the wave
train in a vacuum is
l = c t ------(1)
Where c is the velocity of light in a vacuum.
For example, if t = 10-8s, and c = 3x108 m/s,
then l = (3x10 8 m/s)( 10-8s) = 3 m.
The number of oscillations present in the wave train is
N = l/λ ----(2)
Where λ is the wavelength. If we assume λ = 5000Å = 5x10-7m, then
N = 3m/5x10-7m = 6x106
Thus, a wave train contain about a million wave oscillations in it.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 126
P a g e | 126
Adding together the wave packets generated by all atoms in the light
source, one finds a succession of wave trains, as shown in Fig. 6.1(a). In
passing from one wave train to the next, there is an abrupt change in phase
and also in plane of polarization. It is not possible to relate the phase at a
point in wave train Q to a point in wave train P.
P Phase break Q
T (or x)
tcoh tcoh
(l coh) (l coh)
Fig. 6.1 (a)
Consequently there is no correlation between the phase of different
wave trains. Each wave train has a sustained phase for only about 10-8s,
after which a new wave train is emitted with a totally random phase which
also lasts only for about 10-8s. The phase of the wave train from one atom
will remain constant with respect to the phase of the wave train from
another atom for utmost 10-8s. It means that the wave trains can be
coherent for a maximum of 10-8s only. If two light waves overlap, sustained
interference is not observed since the phase relationship between the waves
changes rapidly, nearly at the rate of 108s times per second.
6.2 COHERENCE LENGTH AND COHERENCE TIME
The wave train, shown Fig. 6.1(a) appears fairy sinusoidal for some
number of oscillations between abrupt changes of frequency and phase.
The length of the wave train over which it may be assumed to have a fairly
sinusoidal character and predictable phase is known as coherence length.
We denote it by l coh. We may consider coherence length as approximately
equal to the length of the wave train, c t, over which its phase is
predictable. The time interval during which the phase of the wave train can
be predicted reliably is called coherence time. It is the time, t, during
which the phase of the wave train does not become randomized but
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 127
P a g e | 127
undergoes change in a regular systematic way. Coherence time is denoted
as tcoh. We can therefore write.
l coh = c t -------(3)
tcoh = t --------(4)
l coh = c tcoh ----------(5)
A wave train consists of a group of waves, which has a continuous
spread of wavelengths over a finite range λ0 centred on a wavelength λ0.
According to Fourier analysis the frequency bandwidth υ is given by
υ = 1/ t
where t is the average lifetime of the excited state of the atom. However, t
is the time during which a wave train is radiated by the atom and
corresponds to the coherence time, tcoh, of the wave train
Therefore υ = 1 / t = 1/ tcoh ---------(6)
Using the relation (5) into Eq. (6) , we get
υ = c/ tcoh -------------(7)
6.3 BANDWIDTH.
A wave packet is not a harmonic wave. Therefore, it cannot be
represented mathematically by simple sine l max
functions. The mathematical representation
of a wave packet is done in terms of Fourier l max/2 λ
integrals. If light emitted from a source is
analyzed with the help of a spectrograph,
it is known to be made up of discrete spectral λ(nm)
lines. Wave packets emitted by atoms form λ/2 λ 0 λ/2
these spectral lines. Therefore, spectral line
and a wave packet are equivalent descriptions. Fig 6.3
The wavelength of a wave packet or a spectral line
is not precisely defined There is a continuous spreads of wavelengths over
a finite range, λ, centered on a wavelength λ0. The maximum intensity of
the wave packet occurs at λ0 and the intensity falls off rapidly on either side
of λ 0, as shown in Fig. 6.3. The spread of wavelengths is called the
bandwidth.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 128
P a g e | 128
The bandwidth is the wavelength interval from λ0 – ( λ/2) to λ0 +
( λ/2) which contains the major portion of the energy of the wave packet.
In practice a source, which is said to produce line spectrum, produces a
number of sharp wavelength distributions.
6.4 RELATION BETWEEN COHERENCE LENGTH AND BANDWIDTH.
The frequency and wavelength of a light wave are related through the
equation
υ = c / λ --------(8)
Where λ0 is the vacuum wavelength.
Differentiating Eq. (8) on both sides, we get
u= - c/λ2 x λ --------(9)
Using the relation (7) into eau. (9) , we obtain
c/l coh = -( c/ λ2) * λ ------(10)
Rearranging the terms, we get
l coh = λ2/ λ --------(11)
The minus sign has no significance and hence is ignored. Eq. (11)
means that the coherence length ( the length of the wave packet) and the
bandwidth of the wave packet are related to each other. The longer the wave
packet, the narrower will be the bandwidth (see Fig. 6.4(a). In the limiting
case, when the wave is infinitely long, we obtain monochromatic radiation of
frequency v0 (wavelength λ0).
λ
λ
λ - +
λ0
Fig. 6.4 (a)
From the equ.(2), the coherence length may be designated as the
product of the number of wave oscillations N contained in the wave train
and of the wavelength, λ. Thus
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 129
P a g e | 129
l coh = N λ ---------(12)
Equating (11) and (12), we get
N = λ / λ
λ / λ = 1/N --------(13)
Eq. (13) shows that the larger the number of wave oscillations in a wave
packet, the smaller is the bandwidth. In the limiting case, when N is
infinitely large, that is when the wave packet is infinitely long, the wave will
be monochromatic having a precisely defined wavelength. The dependence
of bandwidth on the length of the wave packer is schematically shown in
Fig. 6.4 (a)
6.5 COHERENCE
Coherence is an important property of light. It refers to the
connection between the phase of light waves at one point and time, and the
phase of the light waves at another point and time. Coherence effects are
mainly divided into two categories: temporal and spatial. The temporal
coherence is related directly to the finite bandwidth of the source, whereas
the spatial coherence is related to the finite size of the source.
TEMPORAL COHERENCE
Temporal coherence is also known as longitudinal coherence. Let a
point source of quasi-monochromatic light S (Fig. 6.5) emit light in all
directions. Let us consider light traveling along the line SP1P2. The phase
relationship between the points P1 and P2 depends on the distance P1P2 and
the coherence length of the light beam.
The electric fields at P1 and P2 will be
correlated in phase when a single wave
train extends over greater length than P3
the distance P1P2; that is if the distance P2
P1P2 is less than the coherence length l coh. S P1
Then, the waves are correlated in their Fig. 6.5
rising and falling and they will preserve a
constant phase difference.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 130
P a g e | 130
The points P1 and P2 would not have any phase relationship if the
longitudinal distance P1P2 is greater than l coh since in such a case many
wave trains would span the distance. It means that different independent
wave trains would be at P1 and P2 at any instant and therefore the phase at
the two points would be independent of each other. The degree to which a
correlation exists is known as the amount of longitudinal coherence.
Monochromaticity
From Eq.(11) and Fig(6.4 (a)) we conclude that temporal coherence is
indicative of monochromaticity of source. An ideally monochromatic source
is an absolute coherence source. The degree of monochromaticity of a
source is given by
ξ = v /v0 -------(14)
when the ratio v /v0 = 0, the light wave is ideally monochromatic.
Purity of spectral line.
The width of a spectral line is given by λ (see Fig. 6.3). It is seen
from equ.(11) that it is related to the temporal coherence. Thus,
λ = λ2 / l coh ---------(15)
6.6 SPATIAL COHERENCE
Spatial coherence refers to the continuity and uniformity of a wave in
a direction perpendicular to the direction of propagation. If the phase
difference for any two fixed points in a plane normal to the wave propagation
does not vary with time, then the wave is said to exhibit spatial coherence.
It is also known as lateral coherence. Again looking at the point source S
(Fig. 6.5), SP1 = SP3 and therefore, the fields at points P1 and P3 would have
the same phase. Thus, an ideal point source exhibits spatial coherence, as
the waves produced by it are likely to have the same phase at points in
space, which are equidistant from the source. On the other hand, an
extended source is bound to exhibit lesser lateral spatial coherence. Two
points on the source separated by a lateral distance greater than one
wavelength will behave quite independently. Therefore, correlation is absent
between the phase of the waves emitted by them. The degree of contrast of
the interference fringes produced by a source is a measure of the degree of
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 131
P a g e | 131
the spatial coherence of its waves. The higher the contrast, the better is the
spatial coherence.
Fig. 6.6 Spatial coherence
6.7 DETERMINATION OF COHERENCE LENGTH
The coherence length can be measured by means of Michelson
interferometer. In a Michelson interferometer, a light beam from the source
S is incident on a semi-silvered glass plate G (see Fig.6.7) and gets divided
into two components; one component is reflected (1) and the other (2) is
transmitted. These two beams, 1 and 2, are reflected back at mirrors M1
and M2 respectively and are received by the telescope where interference
fringes are produced. It is obvious that the beams produce stationary
interference only if they are coherent.
Let M2' be the image of M2 formed by G. If the reflecting surfaces M1
and M2' (the image of M2) are separated by a distance d, then 2d will be the
path difference between the interfering waves. The condition of fixed phase
relationship between the two waves, 1 and 2, will be satisfied if
2d<<l coh
In such a case distinct interference fringes will be seen. If, however,
2d>>l coh
then phase of the two waves are not correlated and interference fringes will
not be seen. To determine the coherence length of wave emitted by a light
source, the distance d between the mirrors M1 and M2' (the image of M2) is
varied by moving one of the mirrors. As the distance varies, the contrast of
the fringes decreases and ultimately they disappear. The path difference 2d
at the particular stage where the fringes disappear gives us the coherence
length.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 132
P a g e | 132
The light from a sodium lamp has coherence length of the order of 1
mm, that of green mercury line is about 1 cm, neon red line 3 cm, red
cadmium line 30 cm, orange krypton line 80 cm, and that of a commercial
He – Ne laser is about 15m. The coherence length of light from some of the
lasers goes up to a few km.
M1
M'2
1
S 2
G 1 M2
2
T
Fig. 6.7
6.8 CONDITION FOR SPATIAL COHERENCE
The degree of spatial coherence of a beam of light can be deduced
from the contrast of the fringes produced by it. The broader the source of
the light, the lesser is the degree of coherence. In Young’s` double slit
experiment, if the slits S1 and S2 are directly illuminated by a source,
interference fringes are not observed. Instead the screen is uniformly
illuminated. The absence of fringes indicates that the light issuing from the
slits do not posses spatial coherence. If a narrow slit is introduced before
the double slit, light passing through the narrow slit S illuminates S1 and
S2. The waves emerging from them, having been derived through wave front
division, are coherent and stationary interference pattern will be observed
on the screen. If the width of the slit S is gradually increased, the contrast
of fringe pattern decreases and fringes disappear. When the slit S is wider,
S1 and S2 receive waves from different parts of S which do not maintain
coherence. When S is narrow it ensures that the wave trains incident on
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 133
P a g e | 133
slits S1 and S2 originate from a small region of the source and hence they
have spatial coherence.
A S1
B'
S 22 θ
B 2α 2θ d A'
S2
B
Fig. 6.8
Formation of distinct fringe pattern depends on two parameters in the
double slit experiment. One is the size of the slit S and second is the
separation, d, between the two slits S1 and S2 . From Fig. 6.8(a), it is seen
that the path difference between the two waves passing at the edges A and B
of the slit S is
= b sin a
Distinct fringes will be obtained when
<<(1/2 ) λ
Therefore, it requires that
b sin a = <<(1/2)* λ ---------(16)
This equation determines the size of the source, which is spatially
coherence to produce fringes of satisfactory contrast. Equ.(16) shows that
the smaller is the size of the source, the better is the spatial coherence.
Spatial coherence of waves decrease with increasing size of the light source.
From 6.8 it is seen that the edges A and B of the slit S subtend an
angle 2θ at the centre of the double slit and the double slit subtends an
angle 2 a at the centre of S. Waves from A produce interference pattern
which has its centre at A'. Similarly waves from B produce fringe Pattern
with its centre at B'. The patterns cancel each other if the maximum of one
falls on the minimum of the other. The condition for no fringes is that
2θ = (λ / 2d) --------(17)
when the slit S is made wider, the edges A and B move away from each
other. As a result the patterns at A’ and B' also move away. When the
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 134
P a g e | 134
maxima of one pattern fall on the maxima of the other, the fringes appear.
Therefore, the condition for fringes is that
2θ = λ / d
or d = λ /2θ ------(18)
In other words, there is restriction imposed on the separation of the slits
S1 and S2. d must be very small compared to λ /2θ in order to maintain
spatial coherence and obtain fringe pattern on the screen.
Light produced by lasers is highly coherent. Therefore, when a laser
is used as the source, interference fringes can be obtained without the aid of
slit S.
P
A
α
b
B
Fig. 6.8 (a)
6.9 THE LINE WIDTH
In the Michelson interferometer experiment discussed in the previous
section, the decrease in the contrast of the fringes can also be interrupted as
being due to the fact that the source is not emitting at a single frequency
but over a narrow band of frequencies. When the path difference between
the two interfering beams is zero or very small, the different wavelength
components produce fringes superimposed on one another and the fringe
contrast is good. On the other hand, when the path difference is increased,
different wavelength components produce fringe patterns which are slightly
displaced with respect to one another, and the fringe contrast becomes
poorer. One can say that the poor fringe visibility for a large optical path
difference is due to then non-monochromaticity of the light source.
The equivalence of the above two approaches can be easily understood
if we consider the Michelson interferometer experiment using two closely
spaced wavelengths λ1 and λ2 we had shown that for two closely spaced
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 135
P a g e | 135
wavelengths λ1 and λ2 (Like D1 and D2 lines of sodium), the interference
pattern will disappear if
(2d/ λ2) – (2d/ λ1) = 1/2 --------(1)
where 2d represents the path difference between the two beams. Thus
2d = (λ1 λ2)/2(λ1 - λ2) λ2 / 2(λ1 - λ2) ------(2)
Instead of two discrete wavelengths, if we assume that the beam consists of
all wavelengths lying between λ and λ + λ, then the interference pattern
produced by the wavelengths λ and λ +1/2 λ will disappear if
2d = λ2/2 ( 1/2 λ ) = λ2 / λ -------(3)
Further for each wavelength lying between λ and λ +1/2 ( λ), there will be
a corresponding wavelength (lying between λ and λ + (1/ λ) and λ+ λ)
such that the minima of one falls on the maxima of the other, making the
fringes disappear. Thus for
2d λ 2 / λ ---------(4)
the contrast of the interference fringes will be extremely poor. We may
rewrite the above equation in the form
λ λ2/2d -------(5)
implying that if the contrast of the interference fringes becomes very poor
when the path difference is ~d, then the spectral width of the source would
be ~ λ2/2d.
Now we had observed that if the path difference exceeds the coherence
length L, the fringes are not observed. From the above discussion therefore
follows that the spectral width of the source
λ, will be given by
λ ~ (λ2 / L) = λ2 / c c -------(6)
Thus the temporal coherence c of the beam is directly related to the spectral
width λ. For example, for the red cadmium line, λ = 6438 Å,
L 30 cm ( c 10-9 sec)
( λ - λ2) / c c = (6.438x10-5)2 / 3x 1010 x 10-9
~ 0.01 Å
for the sodium line, λ 5890 Å, L 3 cm ( c 10-10 sec) and λ ~ 0.1 Å.
Further since u = c/λ, the frequency spread v of a line would be
u ~ (c/ λ2 ) λ ~ (c/L) ---------(7)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 136
P a g e | 136
where we have disregarded the sign. Since c = L/c, we obtain
u ~ 1 / c ------------(8)
Thus the frequency spread of spectral line is of the order of the inverse of
the coherence time. For example, for the yellow line of sodium
(λ = 5890 Å),
c ~ 10-10 s › v ~ 10 10 Hz
u= c/ λ = 3x1010 / 5.89 x 10-5 5x1014 Hz
we get
u /u~ 1010 / 5 x1014 = 2 x 10-5
The quantity u /u represents the monochromaticity (or the spectral
purity) of the source and one can see that even for an ordinary light source
it is very small. For a commercially available laser beam, c 1~ 50 nsec
implying v /v ~ 4 x10 -8. The fact that the finite coherence time directly
related to the spectral width of the source can also be seen using Fourier
analysis.
6.10 THE SPATIAL COHERENCE (Experiment)
Till now we have considered the coherence of two fields arriving at
particular point in space from a point source through two different optical
paths. In this section we will discuss the coherence properties of the field
associated with the finite dimension of the source.
We consider the Young’s double-hole experiment with the point source
S being equidistant from S1 and S2 (see fig.6.10 (a)). We assume S to be
monochromatic so that it produces interference fringes of good contrast on
the screen PP'. The point O on the screen is such that S1O = S2O. Clearly
the point source S will produce an intensity maximum around the point O.
We next consider another similar source S' at a distance l from S. We
assume that
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 137
P a g e | 137
P
S' S1
l α d
α Q Q
S2
a1 a2
P'
(a)
S1
l S O' O
θ S2
a
(b)
Fig. 6.10 (a) Young’s double – hole interference experiment with two independent point sources S and S'. (b) The same experiment with an extended source.
Waves from S and S' have no definite phase relationship. Thus the
interference pattern observed on the screen PP ' will be a superposition of
the intensity distributions of the interference patterns formed due to S and
S'. If the separation l is slowly increased from zero, the contrast of the
fringes on the screen PP ' becomes poorer because of the fact that the
interference pattern produced by S' is slightly shifted from that produced by
S. Clearly, if
S'S2 - S'S1 = λ/2 --------(1)
the minima of the interference pattern produced by S will fall on the maxima
of the interference pattern produced by S' and no fringe pattern will be
observed. It can be easily seen that
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 138
P a g e | 138
½
S'S2 = a2 + (d /2 + l)2 a + (1 / 2a) (d/2 + l)2
and ½
S'S1 = a2 + (d/2 - l)2 a + (1/2a) (d/2 - l)2
Where a = a1 + a2
And we have assumed a >> d, l, Thus
S'S2 - S'S1 (l d)/a
Thus for the fringes to disappear, we must have
λ/2 = S'S2 - S'S1 (l d)/a
or
l (λa) /2d
Now, if we have an extended incoherent source whose linear dimensions is ~
λa /d then for every point on the source, there is a point at distance of (λa)
/2d which produces fringes which are shifted by half a fringe width.
Therefore the interference pattern will not be observed. Thus for an
extended incoherence source, interference fringes of good contrast will be
observed only when
l << λa /d --------(2)
Now, if θ is the angle subtended by the source at the slits
(see fig. 6.10 (b)] then θ /a and the above condition for obtaining fringes of
good contrast takes the form
d <<λ /θ --------(3)
On the other hand, if
d ~ λ/θ -------(4)
the fringes will be of very poor contrast. Indeed, a more rigorous diffraction
theory tells us that the interference pattern disappears when
d = 1.22 λ/θ, (2.25 λ)/θ, (3.24 λ)/θ ----(5)
Thus as the separation of the pinholes is increased from zero, the
interference fringes disappear when d = (1.22 λ)/θ; if d is further increased
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 139
P a g e | 139
the fringes reappear with relatively poor contrast and they are washed out
again when d = 2.25 λ/θ, and so on. The distance
l w = λ/θ (Lateral coherence width)
gives the distance over which the beam may be assumed to be spatially
coherent and is referred to as the lateral coherence width.
6.11 (I) MICHELSON STELLAR INTERFEROMETER - Theory
Using the concept of spatial coherence, `Michelson developed an ingenious
method for determining the angular diameter of stars. The method is based
on the result that for a distant circular source, the interference fringes will
disappear if the distance between the pinholes S1 and S2 (fig. 6.11 (a) is
given by (see equ. 5)
d = (1.22 λ)/θ --------(1)
Where θ is the angle subtended by circular source. For a star whose
angular diameter is 10-7 radians the distance d for which the fringes will
disappear would be
d ~ (1.22 x 5 x 10-5) / 10-7 600 cm
where we have assumed λ 5000Å. Obviously, for such a large value of d,
the fringe width will becomes extremely small. Further one has to use a big
lens, which is not only difficult to make, but only a small portion of which
will be used. In order overcome this difficulty, Michelson used
L
S1
S d
S2
f
Fig. 6.11 (a) S is source of certain spatial extend; S1 and S2 are two slits separated by a distance d which can be varied. The fringes are observed on the focal plane of the lens L
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 140
P a g e | 140
two movable mirrors M1 and M2 as shown in Fig. 6.11(b) thus he effectively
got a large value of d. The apparatus is known as Michelson’s stellar
interferometer. In a typical experiment the first disappearance occurred
when the distance M1M2 was about 24 feet, which gave
θ (1.22 x 5x 10-5) / (24 x 12 x 254) radians 0.02”
for the angular diameter of the star. This star is known as Arctures. From
the known distance of the star, one can estimate that the diameter of the
star is about 27 times that of the sun.
M1
M2 L
Fig. 6.11(b) Michelson’s stellar interferometer
We should point out that a laser beam is spatially coherent across
the entire beam. Thus, if laser beam is allowed to fall directly on double-slit
arrangement ( Fig. 3.2.11 (c) ), then as long as the beam falls on both the
slits, a clear interference pattern is observed on the screen. This shows that
the laser beam is spatially coherent across the entire wave front.
Laser beam
Screen
Fig. 611(c)- if a laser beam falls on a double – slit arrangement, interference are observed on the screen.
This shows that the laser beam is spatially coherent across
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 141
P a g e | 141
the entire wave front.
Fig. 6.11(b) shows the interference pattern obtained by Nelson and Collins
by placing a pair of slit of width 7.5µm separated by distance 54.1µm on the
end of the ruby rod in a ruby laser the interference pattern aggress with the
theoretical calculation to with in 20%. To show that the spatial coherencies
is indeed due to laser action, they showed that below, threshold (of the laser)
no regular interference pattern was observed; only a uniform darkening of
the photographic plate was obtained.
6.11 (II) Michelson Stellar Interferometer (Another Method)
Using the concept of spatial coherence, Michelson developed an
ingenious method for determining the angular diameters of stars. In this
interferometer, the separation of effective slits for receiving start light may
be increased considerably while the separation of slits for producing fringes
was small and fixed. He accomplished this by mounting four mirrors M1, M2
, M3, M4 on a long iron girder in front of the telescope objective (fig. 6.11(c) )
S1 and S2 are slits in front of the telescope objective L. Light from the star is
incident on the mirrors M1 and M2 which reflect into M3 and M4 and which
reflect into M3and M4 and which, in their turn, pass light into telescope
thorough the apertures of double slit. When the mirrors have been adjusted
so that M1 is parallel to M3 and M2 to M4 and the optical paths M1M3S1 and
M2M4S2 are equal, the beams overlap in the focal plane of the objective and
the Airy`s disc is seen crossed by a number of fine fringes. Airy`s disc is
elongated and much larger than would be formed by the telescope alone
because of its restricted aperture in the present case. Starting with M1 and
M2 close to the fixed mirrors M3 and M4, they moved out symmetrically
maintaining optical paths M1M3S1 and M2M4S2 equal until the fringes
disappeared. Let a be the least separation of the mirrors M1 and M2 for
disappearance of fringes. Then, the angular diameter of the star is
Ф = (1.22 λ) / a
λ is the effective wavelength for the star.
The fringe spacing, which depends on the separation b of S1 and S2
remains constant as the separation a of the movable apertures M1 and M2 is
varied.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 142
P a g e | 142
Ф
B A M1
Ф S1
Light from star M3
A b
M4 F
B P S2 A
Ф M2
Airy`s
Fig. 6.11 (c) Michelson stellar Interferometer
6.12 Let us sum up
In this lesson we described the theory of partial coherence, coherence time and
coherence length and derivation of each. Also we have described about the spatial
coherence for extended sources.
6.13 Lesson end activities
(i) What is partial coherence?
(ii) Explain coherence time and coherence length
(iii) What is spatial coherence?
6.14 Points for discussion
(i) Derive an expression for coherence length and time.
(ii) Write a note on Multi layer films
6.15 Check your progress
(i) Write a neat diagram explaining the measurement of stellar diameter using
Michelson interferometer.
6.16 References
(i) Text book of optics by Brijlal & Subramaniyam
(ii) Optics by Ajoy Ghatak
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 143
P a g e | 143
UNIT – IV
Lesson – 7
Optics of Solids
7.0 Aims and Objectives
This lesson deals with optics of solids, in which the wave equation in a
conducting medium in derived. Also you study about to theory of optical
activity with special reference to Quartz. Finally you study the principle
involved in Faraday rotation.
7.1 THE WAVE EQUATION IN A CONDUCTING MEDIUM For a conducting medium
J = E -----------(1)
Where represents the conductivity of the medium. Thus, Maxwell’s
equation become
div E = 0 -------(2)
div H = 0 ----------(3)
curl E = - µ ( H / t) -------(4)
curl H = E + ( E / t) ------(5)
Taking the curl of Eq. (4), we get
Curl curl E = - µ ( / t) curl H
or
grad div E - 2E = - µ ( E / t) - µ ( 2E / t2)
Using Eq. (2), we get 2 E - µ ( E / t) - µ ( 2E / t2) ------(6)
Which is the wave equation for a conducting medium. For a plane wave of
the type
E = E0 exp [ i (kz – ωt ] -------(7)
We obtain
-k2 + iω΅ +ω2 µ = 0 -------(8)
Which shows that k must be a complex number. If we write
K = + i β ------------(9)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 144
P a g e | 144
Then
- ( 2 + 2i β) + iω΅ + ω2 µ = 0
Equating real and imaginary parts, we get
2 – β2 = ω2 µ ---------(10)
and
β = ω /2 --------(11)
Substituting for β in Eq. (10) and solving for , we get
= Ö ω µ ½ ± ½ (1 + ( 2 / ω2 2)1/2 1/2 -----(12)
We must choose the positive sign, the negative sign, would make complex.
Thus
= ω Ö µ ½ ± ½ (1 + ( 2 / ω2 2)1/2 1/2
β = ω / 2 --------(13)
Now, when k is complex, Eq. (7) becomes
E = E0 exp ( - βz) exp [ i ( z – ωt) -------(14)
Which represents an attenuated wave. The attenuation is due to the Joule-
loss. For a good conductor
/ ω >>1 ----------(15) and one obtains
β (ω΅ /2 )1/2 -------(16)
Indeed of / ω <<1, the medium can be classified as s dielectric and if /
ω >>1 (say 100), the medium can be classified as a conductor. For
0.01 / ω 100,
the medium is said to be a quasi-conductor. Thus depending on the
frequency, a particular material can behave as a dielectric or a conductor.
For example, for fresh water / 0 80 and 10-3 mhos/m. (Both and
can be assumed to be constants at low frequencies.) Thus
/ 10-3 / (80 x 8.55 x 10-2) 1.4 x 10 6 sec-1
For ω = 2 x 1010 sec-1
/ω 2 x 10-5
Thus fresh water behaves as a good conductor for υ 103 sec-1 and as
dielectric for υ 107 sec-1. On the other hand, for copper one may assume
0 and
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 145
P a g e | 145
5.8 x 107 mhos/m and for ω 2 x 1010 sec-1
/ω (5.8 x 107) / (2 x 1010 x 8.9 x 10-12) 108
Thus even for such frequencies it behaves as an excellent conductor.
From Eq. (14) it can be easily seen that the field decrease by a factor e
in traversing a distance
= 1 / β
which is known as the penetration depth. For copper, µ µ 0 = 4 x 10-7 N / amp2
and [ (2/ω΅ ) ]½ ( 2 / 2 v x 4 x 10-7 x 5.8 x 107 )½
0.065 / v x m
Thus for v 100 sec-1, 0.0065 m = 0.65 cm whereas for v 10 8 sec-1,
6.5 x 10-6 m showing that the penetration decreases with increase in
frequency.
7.2 FARADY ROTATION
Consider linearly polarized light propagating through a medium. If a
magnetic field is applied along the direction of propagation of the polarized
wave, then the plane of polarization gets rotated – this rotation is usually
referred to as Faraday rotation after the famous Physicist Michel Faraday
who discovered this phenomenon in 1845. In the presence of (longitudinal)
magnetic field, the modes of propagation are the left circularly polarized
(LCP) wave and the right circularly polarized (RCP) wave. Thus the situation
is somewhat similar to the phenomenon of optical activity. The angle θ by
which the plane of polarization rotates is given by the empirical formula
θ = VH
where H is the applied magnetic field, l is the length of the medium and V is
the constant. For silica V 2.64 x 10-4 deg / Ampere 4.6 x 10-6 radians /
Ampere.
The Faraday rotation has a very important application in measuring
large currents using single mode optical fibers. We consider a large length
of a single mode fiber wound in many turns in the form of a loop around a
current carrying conductor (see Fig.7.2 ). If a current I is passing through
the conductor then by ampere’s law
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 146
P a g e | 146
H .dI = NI
where N represents the number of loops of the fiber around the conductor.
Thus if a linearly polarized light is incident on the fiber, then its plane of
polarization will get rotated by the angle.
Fig. 7.2 A single mode fiber wound helically around a current
carrying conductor. The rotation of the plane of polarization
is detected by passing the light through a Wollaston prism
and then an electronic processor
------
θ = VNI
The rotation θ does not depend on the shape of the loop. As an
example, for I = 200 Amperes and N = 50, θ 0.26 degree. The light from
the fiber is allowed to fall on a Wollaston prism and the outputs are
measured separately; the Faraday rotation θ is given by
θ = constant (I1 – I2) / (I1 + I2)
where I1 and I2 are the currents in the electronic processor due to the two
beams coming out of the Wollaston prism. We can determine the actual
variation of the output with the current passing through the conductor.
Such a set up can be used to measure very high currents (10000 Amperes).
7.2 (a) General idea of Polarization
A simple method for eliminating one of the beams is through selective
absorption; this property of selective absorption is known as dichroism. A
crystal like tourmaline has different coefficient of absorption for the two
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 147
P a g e | 147
linearly polarized beams into which the incident beams splits up.
Consequently, one of the beams gets absorbed quickly and the other
component passes through without much attenuation. Thus if an
unpolarized beam is passed through a tourmaline crystal, the emergent
beam will be linearly polarized
Y Incident unpolarized
z
Linearly polarized wave x
Tourmaline crystal
Fig. 7.2 (a) when an unpolarized beam enters a diachroic
crystal like tourmaline, it splits up into two linearly polarized
components. One of the components gets absorbed quickly and
the other component passes through without much attenuation.
------- Another method for eliminating one of the polarized beam is through
total internal reflection. We will show that the two beams have different ray
velocities and as such the corresponding refractive indices will be different.
If one can sandwich a layer of material whose refractive index lies between
the two, then for one of the beams, the incidence will be at a rarer medium
and for the other it will be at a denser medium. This principle is used in a
Nicol prism which consist of a calcite crystal cut in such a way that for the
beam, for which the sandwiched materials is a rarer medium, the angle of
incidence is greater than the critical angle. Thus this particular beam will
be eliminated by total internal reflection. Figure (7.2 (b) ) shows a properly
cut calcite crystal in which layer of Canada Balsam has been introduced so
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 148
P a g e | 148
that the ordinary ray undergoes total internal reflection. The extra-ordinary
component passes through and the beam emerging from the crystal is
linearly polarized.
Calcite 90º
68º
Fig 7.2 (b) The Nicol prism. The dashed outline corresponds
to the natural crystal which is cut in such a way that the
ordinary ray undergoes total internal reflection at the Canada
Balsam layer
7.3 THEORY OF OPTICAL ACTIVITY
As mentioned earlier, in an isotropic dielectric, the D vector is in the
same direction as E and we have
D = E = 0n2E ---------(1)
Where 0 ( = 8.854 x 10-12 MKS units) is the permittivity of free space and
n = / 0 is the refractive index of the medium. Now if we dissolve cane
sugar in water, the medium is still isotropic; however because of the spiral
like structure of sugar molecules, the relation between D and E is given by
the following relation
D = 0 n2E + ig k x E
0 n2 (E + i k x E ) ----------(2)
where
= g / 0 n2
and k is the vector along the direction of propagation of the wave. The
parameter can be positive or negative but it is usually an extremely small
48º e -ray
Axis Optic o -ray axis
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 149
P a g e | 149
number (<<1). Without any loss of the generality, we may assume
propagation along the z- axis so that kx = ky = 0 and k1 = 1 giving
x y z
K x E 0 0 1 = x Ey + y Ex
Ex Ey Ez
Thus Dx 0n2 -ig 0 Ex = Dy ig 0n2 0 Ey -------(3)
Dz 0 0 0n2 Ez
The matrix is still Hermitian but there is small off diagonal imaginary
element. The presence of these off - diagonal terms give rise to optical
activity.
n2w / c2 µ0 [ E – ( k . E) K = D
We write the x and y components of the above equation and since kx = 0 = ky
and kz =1, we get
(n2w / c2 µ0) Ex = Dx = 0n2 Ex - igEy and
(n2w / c2 µ0 ) Ey = Dy = igEx + 0n2 Ey
Thus
( (n2w / n2) – 1) ) Ex = -i Ey and
( (n2w / n2) – 1) ) Ey = -i Ex
where, we have used that fact that c = (1 / 0µ0.) For nontrivial solutions.
(n2w / n2 – 1)2 = 2 giving
nw = n 1 ± --------------(4) and
Ey = ± i Ex ------------(5)
We write the two solutions as nr ( = n 1 + and n1 ( = n 1 - ;
the corresponding propagation constants will be given by
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 150
P a g e | 150
k = kr = (ω/c) nr = (ω/c) n 1 + ---------(6)
and
k = kl = (ω/c) nl = (ω/c) n 1 - ---------(7)
For nw = nr, if
Ex = E0ei(krz – ωt)
Then
Ey = + iEx = E0eikrz – (ωt + /2)
Which would represent an RCP (Right Circularly Polarized) wave and hence
the subscript r. Similarly For nw = nl , if Ex = E0ei(klz – ωt)
Then Ey = -iEx = E0 eiklz – (ωt - /2)
Which would represent an LCP (Left Circularly Polarized ) wave and hence
the subscript l. The RCP and LCP waves are the two “modes” of the
“optically active” substance and for an arbitrary incident state of
polarization, we must write it as a superposition of the two modes and
study the independent propagation of the two modes. Hence,
nr - nl = n [ 1+ - 1 - ] -------(8)
n
7.4 OPTICAL ACTIVITY IN QUARTZ
One observes optical activity for a plane polarised wave propagation
along the optic axis of quartz crystal. The general theory of propagation of
electromagnetic wave in such crystals is quite difficult; if the propagation is
not along the optic axis the “modes” are very nearly linearly polarized and
one may use the analysis discussed already. If the propagation is along the
z-axis then we may write (Eq.3)
Dx 0n20 -ig 0 Ex
Dy = ig 0n20 0 Ey
Dz 0 0 0n2e Ez
Where n0 and ne are constant of the crystal. Carrying out an identical
analysis, we would get
nr n0 [ 1 + ½ ]
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 151
P a g e | 151
and
nl n0 [ 1 – ½ ] giving
nr - nl n0 and
ρ = / λ0 (nl - nr ) ( n0 ) / λ0
where λ0 is measured centimetres. For quartz
ρ ± 8.54 radians /cm at λ0 = 4046.56 Å
± 3.79 radians /cm at λ0 = 5892.90 Å
± 2.43 radians / cm at λ0 = 7281.35 Å
In quartz, we can have nr > nl or nr < nl . For λ0 = 4046.56 Å, we readily get
- n l - nr - 1.1 x 10-4
We may compare this with the value ne – n0 0.9 x 10-2. At higher
wavelengths, the value of - n r - nl - is much less.
7.5 THEORY OF FARADAY ROTATION
The equation of motion for the electron, in the presence of an external
electric field E is given by
(d2r / dt2) +(ω2 r) = - q/m E --------(1)
In the presence of static magnetic field B, we would have an additional (VxB)
term
. (d2r / dt2 )+ω2 r + (q/m) r x B = - (q/m) E -----(2)
where r = x x + y y + z z represents the position vector of the electron, x, y,
and z, are the unit vectors and q ( = + 1.6 x 10-19 C) is the magnitude of the
electronic charge. We assume the magnetic field to be in the z- direction.
Bx = 0 = By and Bz = B0 -------(3)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 152
P a g e | 152
Thus
. x y z r x B dx/dt dy/dt dz/dt = [ x (dy/dt) – y (dx/dt) ] B0
0 0 B0
Now for a circular polarized light wave propagating along the z – direction
E± = ( x ± i y) E0ei( kz –ωt ) ---------(5)
Where the upper and lower signs correspond to RCP and LCP respectively.
If we now write the x and y components of (Equ. (2), we would get
(d2x / dt2) + (ω02x) + (qB0 / m) ( dy/dt) = ( - q/m) E0ei( kz –ωt) ------(6)
and
(d2y / dt2) + (ω02y ) - (qB0 / m) ( dx/dt) = ± (i q/m) E0ei( kz –ωt) -----(7)
where the upper and lower signs corresponding to RCP ( Right Circularly
Polarization) and LCP ( Left Circularly Polarization) respectively. Writing
x = x0 e i( kz –ωt) and y = y0 e i( kz –ωt)
we get x (ω2 – ω02) (ω2 – ω20) x0 + iωc iω y0 = + (q/m) E0
(ω2 – ω20) y0 + iωc iω x0 = ± (q/m) E0
x - iωcω
--------------(8 , 9)
where
ωc = (qB0 / m)
is the electron frequency. If we multiply Eq. (8) by (ω2 – ω02)and Eq. (9) -
iωcω and add the two equation we should get
[ (ω2 – ω02)2 – ω2cω2 ]x0 = (q/m)E0 [ (ω2 – ω02)2 ± ωcω
giving
x0 = q E0 / m [(ω2 – ω02)2 ± ωcω ]
similarly
y0 = ±i q E0 / m [(ω2 – ω02)2 ± ωcω ] = ±ix0
Thus the polarization is given by
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 153
P a g e | 153
P = - Nqr
= -Nq . q E0 ( x ± i y ) / m [(ω2 – ω02)2 ± ωcω] e i(kz –ωt)
= Χ = E±
where the susceptibility Χ is given by
Χ =( Nq2 /m) . 1 / [(ω02 – ω02)2 ± ωcω]
Thus the modes are circularly polarized and the corresponding refractive
indices are given
n2 ± = 1 + ( Nq2 /m 0) . 1 / [(ω2 – ω02)2 ± ωcω]
Where the upper and lower signs correspond to RCP and LCP respectively.
7.6 Let us sum up
From this lesson, you have learnt the propagation of light in conducting media
and also the propagation of light in crystals. Further you have learnt about the optical
activity, Faraday rotation in solids.
7.7 Lesson end activities
(i) What is Faraday rotation?
(ii) What is optical activity?
7.8 Points for discussion
(i) Write a note on reflection by a conducting medium?
(ii) Write a note on Magneto and electro optic effects.?
7.9 Check your progress
(i) Explain the theory of optical activity?
(ii) Explain the theory of Faraday rotation?
7.10 References
(i) Text book of optics by Brijlal & Subramaniyam
(ii) Optics by Ajoy Ghatak
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 154
P a g e | 154
UNIT – IV
Lesson – 8
Optical Fibres
8.0 Aims and Objectives
The refractive index `n` or index of refraction is related to the speed of
light (c) in free space and the speed of light in the medium (v) i.e.., n = c/v
and the value of C = 3x108 m/s. The value of `n` for air is 1.00, for water
1.33 and for glasses 1.50.
When light ray travelling in medium of refractive index (n1) is incident
on a dielectric material of lower refractive index (n2) a part of the light is
reflected and a part is refracted at the loundary. According to Snell’s law
n1 Sinφ1 = n2 Sinφ2
Incident ray
Reflected ray
Lower index n1 φ1
(air)
Higher index n2
(glass) φ2 Refracted ray.
Fig. 8.0 (a) Incident light ray on glass – air interface
Here φ1 is the angle of incidence and φ2 is the angle of refraction. As the
angle of incidence φ1 increases, the angle of refraction φ2 also increases and
approaches 90º. When the angle refraction is 90º, the angle of incidence φ 1
is called the critical angle (φc). When the angle of incidence is greater than
the critical angle (φc), the ray is completely reflected back into the
originating medium fig. 8.0 (b). This phenomenon is known as Total
Internal Reflection.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 155
P a g e | 155
n2
n1
φ φ
Fig. 8.0 (b) Total internal reflection
8.1 OPTICAL FIBRES
An optical fibre is a dielectric waveguide and is normally cylindrical in
form. It consists of bundle of very thin individual fibres. A beam of light
enters at one end of the bundle of fibres and reaches the other end with less
amount of energy loss in a direction parallel to its direction, as shown is fig.
8.1 (c). Optical fibres can also transmit an image from one end to the other
end.
φ Core φ
8.1 (c) Fig. Total internal reflection
8.1.1 Propagation of Light in Optical Fibres
The light rays propagating into the fibre core are of two types namely;
meridional rays and skew rays. The light rays which pass through the axis
of the core are known as meridional rays. The passage of such rays in a step
index fibre is shown in fig. 8.0(a)
Cladding
Cladding
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 156
P a g e | 156
a) Meridional (b) Skew rays
Fig. (8.1d) Propagation optical fibre
The rays which never pass through the axis of the core are known as
Skew rays. A typical passage of skew rays in a step index fibre is shown in
fig 8.0 (b). The skew rays will not utilise the full area of the core and they
travel farther than meridional rays and undergo higher attenuation.
8.1.2 Acceptance Angle
The fibre core will propagate the incident light, which are incident at
an angle less than the critical angle (φc) will continue to be propagated down
its length. The geometry of the launching of light rays into an optical fibre is
shown in fig 8.1.2.
A Eventually lost by
radiations Φc Acceptance cone Φc Core
B Cladding
8.1.2 Acceptance angle
Let a meridional ray be incident at an angle θa in the core – cladding
interface of the fibre. Any ray A which is incident at an angle greater than θa
will be transmitted into the core cladding interface at an angle less than the
critical angle θc and will not undergo total internal reflection.
A ray B which is incident at an angle greater than θa is eventually lost.
Incident rays which are incident on fibre core within conical half angle φa
will be refracted into fibre core and propagate into the core by total internal
reflection. φa is called the angle of acceptance or minimum acceptance angle
or total acceptance angle.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 157
P a g e | 157
8.1.3 Structure of optical fibres
An optical fibre consists of an inner glass cylinder of high refractive
index (n1) which is known as the core. The core is surrounded by a
cylindrical shell of glass or plastic of lower refractive index (n2) called
cladding. Since n1 > n2 total internal reflection takes place.
The thickness of the core is 50µm and that of the cladding is 100µm
to 200µm. The overall thickness of an optical fibre is nearly 125µm to
200µm. Thus an optical cable is small in size and of light weight.
Φc core n1 Φmax
Φ
Incident ray
8.1.3 Configuration of optical fibre
8.1.4 Numerical Aperture:
Numerical aperture (NA) is a measure of the light rays that can be
accepted by the fibre. Consider a ray of light entering the optical fibre at an
angle φ1. Let φ1 be less than the angle of acceptance φa. The ray enters from
air of refractive index n0 to the core of refractive index n1 . n1 > n2, the
refractive index of the cladding. By Snell’s law
n0 Sin φ1 = n 1 φ2 ------(1)
Air (n0)
n2
φ1 A C n1
φ2 φ φ
B n2
Fig. 8.1.4 Numerical aperture
From the above fig, in triangle ABC
φ2 = ( /2 – φ) -----(2)
Cladding n2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 158
P a g e | 158
Substituting in (1)
n0 Sin φ 1 = n 1 Sin( /2 – φ )
n 0 Sin φ1 = n 1cos φ -----(3)
n0 Sin φ 1 = n1(1-sin2 φ)1/2 -----(4)
When total internal reflection takes place φ = φc and φ1 = φa Equ. (4)
becomes
n0 Sin φ a = (n 12 - n 22)1/2
i.e., NA = n0Sin φ a = (n 12 - n 22) 1/2 ------(5)
Refractive index of air n0 = 1, NA is equal to sin φa.
The NA is terms of refractive index difference ( ) between the core and
cladding in given by the relation
NA = (n 12 - n 22)1/2 = n1 (2 ) 1/2 where = (n 21 - n 22 ) / 2n12 = (n1 – n2) / n1
NA is the Numerical Aperture passing through an optical fibre is given by
NA is the Numerical Aperture passing through the number of modes optical
fibre is given by
N 4.9 (N Ad/λ)2
Where d is he diameter of the core and λ is the wavelength of light used.
When d 0.76 λ/NA the fibre propagates only a single mode.
Fibre optic materials and their properties. Optical fibres are made up
of silica (glass) and plastic. The optical fibre materials have the following
properties.
i) Low scattering cross section
ii) Low optical (absorption, attenuation, dispersion) energy loss
iii) Efficient guide for light waves.
8.2 Classification of optical fibres
Optical fibres are classified based on
i) Number of modes and
ii) Refractive index profiles.
Based on number of modes
Depending on the number of modes, optical fibres are classified as
single mode fibre and multimode fibres.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 159
P a g e | 159
Single mode fibre: If only one mode s transmitted through a fibre, then it
is a single –mode fibre. It can be excited with LASER diodes. The core
diameter is of the order of a few wavelengths. The difference between the
refractive indices of core and cladding is small.. A typical single mode fibre
may have a core radius of 3µm and numerical aperture of 0.1 at a
wavelength of 0.8µm.
Multimode fibre: If more than one mode is transmitted through a fibre,
then it is a multimode fibre.
Advantages of multimode fibre: The larger core radii of multimode fibres
make it easier to launch optical power into the fibre and facilitates end to
end connection of similar fibres. Another advantage is that light can be
launched into multimode fibre using a light emitting diode (LED) source.
The disadvantage of multimode fibres is that they suffer from intermodal
dispersion.
Based on Refractive index profile:
The optical fibres are classified in the following manner:
1) Step index fibre
2) Graded index fibre
Step index fibre: It is based on the comparison of the refractive indices of
air, cladding and core. The refractive index (n) profile with respect to the
radial distance (r) from the fibre axis is given by
r r
a
Refractive Index (nr)
n1 Core
n2 Cladding
(a) Multimode
r
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 160
P a g e | 160
Refractive index a Core
(nr) n1 Cladding
n2
(b) Single mode
Fig. 8.2 Step index fibre
For core n(r) = n 1 ; r < a For cladding n (r) = n 2 ; r a
The step index fibres propagate both single and multimode signals
within the core. A typical step index fibre which propagates multimode
signals is shown is fig. 8.2(a). The diameter of the core is 50µm or greater.
A typical single mode or monomode step index fibre is shown is fig. 8.2(b).
The core diameter of the single mode fibre is of the order of 2 to 10µm.
A typical structure of the single mode step index fibre is shown in fig.
8.2(c) The core and cladding are made of glasses with high purity to avoid
losses due to absorption. The outer layer is of less refractive index and
made of glass of lesser purity. The thickness of the cladding is at least 10
times as that of the core radius. The same can be increased for easy
handling and for reducing the damage due to micro bending.
Buffer jacket
Primary coating Refractive Index
Cladding n1
n2
Core
Fig. 8.2 (c) Single mode fibre waveguide structure
In practical step index fibres the core (of radius a ) refractive index
n1 = 1.48. The refractive index of the cladding n 2 is slightly less than η1.
n 2 = n 1(1- )
The parameter is the core –cladding index difference or simply the
core index difference. is of the order of 0.01 since the core refractive index
.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 161
P a g e | 161
is larger than the cladding refractive index., electromagnetic energy at
optical frequency is made to propagate along the fibre wave guide by total
internal reflection at the core-cladding interface.
ii) Graded index fibre: In graded index fibre, the refractive index of the core
varies with the radial distance from fibre axis. The refractive index at the
fibre axis is maximum at the core – cladding interface. The index (n) profile
of the radial distance (r) from the fibre axis is given as
for core n (r) = n 1[1-2 (r/a)α]1/2 ; r < a
for cladding n (r) = n 1[1-2 ) 1/2 = n 2 ; r a
Where α is the profile parameter which gives the characteristic refractive
index profile of the optic fibre core. The index difference for the graded –
indeed fibre is given by
= (n 12 - n 22)/2 n 1 (n 1 - n 2) / n 1
The approximation on the right –hand side of the above expression gives the
index difference of the step –index fibre.
When α is infinite the fibre will be step index. When α = 2 the fibre
will be parabolic and when α = 1 the fibre will be triangular as shown in the
figure Refractive index
α2 α1
n2 n2
-a Core axis b Radial distance
Fig. 8.2(d) Refractive index profile against values
A parabolic refractive index profile core with α=2 is used for
multimode propagation as shown 8.2 (e) . In the graded-index fibre, the
refractive indeed of the core is maximum at the core axis and a gradual
decrease from the centre of the core results in the creation of many
refractions of the incident rays as shown in Fig. 8.2 (e). The above
mechanism can be easily explained by considering the Fig. 8.2 (f). The
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 162
P a g e | 162
expanded diagram shows the various refractive index values from high to
low at different interfaces within a graded-index fibre. When the ray is
incident at the air core medium, the refracted ray shows a gradual curve
due to the ever increasing a angle of incidence until the condition for total
internal reflection at met and then the ray travel back to the core axis
gradually be continuous refraction.
r
r
n2 a
Refractive Index n1
core
Cladding
Fig. 8.2 (e) Ray transmission multimode graded fibre
The intermodal dispersion less in the case of multimode graded-index fibre
than multimode step index fibre due to their refractive index profile. The
multimode graded index fibre with relatively high numerical aperture finds
application in high bandwidth and medium half.
r Refraction Total int.reflection
n6 n5
n4 n3 n2 Core n1 Cladding
Fig 8.2(f) Curved ray path in a graded fibre
A typical multimode graded indexed wave guide structure is shown
Fig. 8.2 (g). Similar to single mode wave guide, in order to avoid the
absorption loses high purity core and cladding martial were selected. The
features of outer layer ae to provided and improved rigidity to resist bending.
The outer layer may be of same material with different refractive index. The
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 163
P a g e | 163
thickness of the cladding will be normally very thin as it is used only as a
barrier to prevent the impurity ions from outer layer to migrate into the core
region. For higher refractive index of the outer layer, the thickness of the
cladding will be of 10 wavelengths. The diameter of the core will be in the
order of 50 -60µm with NA of 0.2 – 0.3
Buffer jacket
Primary coating Refractive Index
Cladding
n1
Core nn
Fig. 8.2 (g) Multimode grated index wave guide structure
8.2.1 TRANMISSION CHARACTERISTICS OF OPTICAL FIBRES
The study of the transmission characterises of the optical fibres helps
in selecting the fibre for optical communication. Following are the two
factors which affect the transmission of light waves in optical fibres.
(i) Attenuation
(ii) Dispersion
(i) Attenuation:
The attenuation or transmission loss of light signals when it
propagates through an optical fibre is an important factor to be considered
in the design of an optical fibre communication. The main sources, which
are responsible for attenuation are namely electron absorption. Raleigh
scattering, material absorption and impurity absorption. The first three
absorption are known as intrinsic absorption mechanism since these
depend on the glass fibre. The absorption due to impurity is known as
extrinsic absorption mechanism. The attenuation/wavelength
characteristics of a glass fibre is shown in Fig. (4.3.1)
.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 164
P a g e | 164
Attenuation (dB/Km)
Impurity absorption
100-
10-
1- Rayleigh scattering
Electron Material absorption absorption
0.1
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
Wavelength (µm)
Fig.8.2.1 Attenuation/wavelength of a silica based glass fibre
(i) Intrinsic absorption
In intrinsic absorption attenuation is due to Rayleigh scattering
material absorption and electron absorption. In the case of Rayleigh
scattering, the scattering of light takes place due to the small irregularities
in the structure of the core. The irregularities in the core are due to the
density fluctuation during the glass manufacturing. The fundamental
Rayleigh scattering is significant only when the wavelength of the incident
light is in the same order as the dimension of the scattering mechanism.
The above losses can be reduced by having the operation at longer
wavelength.
It is clear from Rayleigh scattering that the operational long
wavelength will produce low losses. However, the atomic bonds associated
with the core material will absorb long wavelength light and is known as
material absorption. Therefore, the operation at wavelength greater than
1.55 mm will result in a significant drop in attenuation.
The absorption of photons by exciting the electrons from core of the
atom to higher energy states due to incidence of flight in the ultraviolet
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 165
P a g e | 165
region is known as electron absorption. This type of absorption takes place
only in the lower wavelength region and is shown in the figure.
(ii) Dispersion
In optical fibres communication due to digital modulation, broadening
of the transmitted light pulse as they travel for a long distance take space.
This to the dispersion mechanisms involved within the fibre during
transmission. The broadening of light pulse using the digital bit pattern
1011 as they are transmitted along a fibre is shown in Fig. 8.2.1 The initial
pulses in the fibre input are shown in Fig 8.2.1 (a). It is clear from the Fig.
8.2.1(b) and Fig. 8.2.1(c) each pulse broadens and overlaps with is
neighbours and finally becomes indistinguishable at the receiver input. This
effect is known as intersymbol interference (ISI). The ISI is more effective in
digital communication than analog.
1 0 1 1
Amplitude Amplitude 1 0 1 1
Time Time
8.2.1 (a) Fibre input 8.2.1 (b) Fibre output at a distance L1
No zero level Composite pattern
Amplitude
ISI
Time
8.2.1 (c) Fibre output at a distance L1 > L2
Fig. – Broadening of light pulse
8.3 FABRICATION OF OPTICAL FIBRES:
Generally, fibres are prepared by using glass and plastic materials. Glass
fibres are fabricated mainly on the basis of Sio2 and doping is done with
oxides such as Geo2, PbO, B2O3 etc., Similarly the plastic fibres are
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 166
P a g e | 166
commonly made with polystyrene as core and methyl-methacrylate as
cladding materials. The different methods used for the fabrication of optical
fibres are the following.
I) modified chemical vapour deposition (MCVD)
ii) Fibre drawing from preform
iii) Fibre drawing from a double crucible.
8.3 (a) Modified chemical vapour deposition (MCVD)
A block diagram of the experimental setup used for the formation of
silica cladding tube and deposition of core glasses is shown in figure 8.3 (a)
a) Formation of Silica Vitrifying flame
CLADDING TUBE Sio2 Glass
Deposition flame
Si,O2 Soot
Mandrel
Sicl4 O2
Fig. 8.3 (a) Formation of silica cladding tube
(b) DEPOSITION OF CORE GLASSES:
Silica tube
Core Reaction area Exhaust gases
material glass
Rotating joint Rotating joint
Reaction flame
Fig:8.3 (b) Schematic diagram used in MCVD
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 167
P a g e | 167
The MCVD is the method used to produce a glass preform. A glass
preform is a final fibre which has refractive index profile as that of final
fibre. The optical fibres are drawn from the glass preform by heating and
pulling a thin stand. The first step in the preparation of glass perform is the
process of producing a dio2 tube or substrate. This forms the cladding of
the final fibre and hence it needs a dopant to maintain a difference in
refractive index between core and cladding. The tube or substrate is made
by depositing a layer of d io2 particles and dopants known as soot on rotating
ceramic former or mandrel. When the soot reaches the required
temperature say 1400ºC it is vitrified into a clear glass. Then, the rotating
ceramic former is withdrawn. In other words the complete glass preform
can be made by depositing the glass first and then the cladding using the
deposition flame as shown in fig (a). After this process, the rotating ceramic
former can be withdrawn. Then the tube is collapsed to get the required
glass preform. This process is known as outside vapour phase oxidation.
In MCVD process, the tube is then fitted with the lathe and the core
particle as gaseous constituents is passed into it as shown in Fig. 8.3(b).
A clear glass obtained by sintering of core particles. The vapour is shut off
after achieving the required core depth. Then, the necessary heating is
given to the tube get the tube to collapse resulting in the required glass
preform. The glass preform is then placed in a pulling tower where the fibre
is drawn out.
8.4 DISPERSION:
The term dispersion describes the pulse broadening effect in fibres.
The dispersions produced by the properties of the core material and the line
width of the light signal passing through the fibre. As shown in Fig. 8.4 (a)
the pulse that appears at the output of a fibre is wider than the input pulse.
Dispersion may, therefore be defined as the output light pulse width
produced by an input pulse of zero width i.e.., a monochromatic light which
has single wavelength.. In that case, Pulse width available at the output
would be totally a result of fibre dispersion. Let us take a practical case of
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 168
P a g e | 168
an input pulse of with W1 and the output pulse of width with W2 > W1. The
fibre dispersion dT is defined as
dT = W22 – W21
W1 Output pulse
(a)
Input pulse W2
Fibre
Fig. 8.4 (a)
Dispersions measured in units of time either in nanoseconds (10-9s)
or picosecs. (10-12). The total dispersion produced by an fibre depends
directly on its length. Here, manufactures give the dispersion per unit
length of the fibre either in nanoseconds per kilometre (ns/km) or
picoseconds per kilometre(ps/Km). This permits the users to calculate the
expected total dispersion for a given length of fibre. If L is the length of the
fibre in kilometre, then the total fibre dispersion is given by
dT = L x (dispersion /Km)
The dispersion can be divided into two categories
i) Intermodal dispersion
ii) ) Intramodal dispersion
8.4 (a) Intermodal dispersion:
--- ---------
This type of dispersion result from the fact that the light ware
propagates through a fibre in different modes as explained in figure. There
is difference in the propagation times for the different modes which
produces dispersion between these different modes. For full understanding
of this dispersion, consider two extreme modes as shown in figure.
--------------------------------
n2
------θc------------------------------------- n1 Zero mode
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 169
P a g e | 169
The critical mode propagating at angle θc and another mode where the
propagation angle is zero i.e., zero mode. A pulse of light launched into the
fibre will propagate along the fibre in both modes.
For zero modes, travel time will be minimum and is given by
Tdo = L/V1 = L/C/n1 = L x n1/C
Where V1 is the velocity of light through the core material, C is the velocity of
light in vacuum and L is the length of the fibre. For the light rays travelling
at the critical angle θc, the delay would be maximum and is given by
Idc = (L/Cos θc )/ V1 = (L/Cos θc)/ (C/ n1)
= (L x n 1) / (Cos θcxC)
The time difference (Tdc - Td0) represents the pulse width dT at the output
dT = Tdc - Td0
= (L x n1)/Cos θcxC – (L x n1/C)
Using the relation Cos θc = n 2 / n 1; we get
dT = (Lx n1/C) x { (n1 – n 2) / n 2}
Where n1 and n2 are the refractive indices of the core and cladding
respectively. The expression (n 1 – n 2) /η2 = . Hence, the pulse width can
be expressed as
dT = = (Lx n1/C) x
Substituting the value of in terms of numerical aperture we get
dT = [Lx(N.A)2] /(2x n1xC) (or)
dT/L = (N.A)2 /(2x n1xC)
Where dT is the amount of pulse broadening and dT/L is the broadening per
unit length of the fibre.
8.4. (b) Intramodal dispersion:
It is due to the fact that he light signal propagating in fiber does not
consist of a single frequency but a group of frequency. It is also called
chromatic dispersion. It is related to the line width of the light pulse. It is
often expressed in terms of Pico seconds per Kilometre per nanometre of line
width.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 170
P a g e | 170
8.5 Fibre losses:
The three major causes of light losses in an fibre are as under:
i) Material loss: It is due to the absorption of light by the fibre material. It
includes absorption due to light interacting with the molecular structure of
the material as well as loss due to material impurities. The loss due to the
atomic structure of the material itself is relatively small. Losses due to
impurities can be reduced to by better manufacturing.
ii) Light Scattering:
Cladding
Light is scattered by the molecules of the material due to structural
imperfections and impurities. The scattered light does not propagate down
the fibre, it is lost.
As shown in figure when light is scattered by an obstruction, it
produces power loss. In optical fibres, obstruction refer to density
variations in the material which produce changes in the refractive index.
The index variations behave like point sources scattering light in all
directions. Scattering can also be due to micobends i.e.., fibre deformation.
This type of loss has been made negligible with the help of improved
manufacturing techniques.
iii) Waveguide and Bend losses:
These are caused by imperfections and deformations of the fibre
structure which cause radiation of light away from the fibre. Fig. Shows
radiation loss caused by a change in diameter. As seen in the absence of
deformation, the light ray would have been confined inside the fibre. Very
small changes in the core size (i.e.., micobends) also cause radiation of light
as well as back scatter shown in figure, obviously, in both instances, the
loss results from scattering of light.
Core
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 171
P a g e | 171
------------------------ ------------------------ Ray leaves fibre
Back Scattered Radiated light
All the above losses are dependant on the wavelength of light used. By
carefully choosing the operating wavelength, the above losses can be
minimized.
Calculation of losses:
Cladding
Pi P0
As shown in figure, if Pi is the power input to the fibre and P0 the
power output the fibre loss is given by the ration P0/ Pi . The decibel loss is
given by
Loss = 10log P0/ Pi decibel
Since the loss increase with fibre length, the loss is usually given by decibels
per Kilometre (dB/Km).
8.6 APPLICATIONS
1) The small size and large information-carrying capacity of optic fibres is
much higher as compared to copper twisted-par cables in telephone
systems.
Fiber Core
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 172
P a g e | 172
2) Continuous passive links (no repeaters) more than 100Km long have been
produced.
3) The low weight of fibre cable a compared to coaxial lines gives them
distinct advantages for submerged cable application because of the relative
ease of transporting and laying the fibres.
4) Optic fibre communication are compatible with electrified railways
because they are not effected by EMI
5) Fibres are used in application that ae primarily video i.e.., broadcast
television and cable T.V remote monitoring and surveillance.
6) Fibre systems are particularly suited for transmission of digital data such
as that generated by computers
7) Military application of fibre optics include communication, command and
control links on ships and aircraft, data links for satellite earth stations and
transmission lines for tactical command post communication
8.7 Fibre Endoscope
In medicine, a fibre endoscope is used to study the interior of the
lungs and the other parts of the human body that cannot be viewed directly.
It can also be used to study the tissues and blood vessels far below the skin.
Te fibre endoscope uses bundle pf flexible fibres.
The schematic representation of flexible fibre endoscope is shown in
Fig. 8.7 It consists of two fibres namely inner and outer fibre. The inner
fibre illuminates the inner structure of the object under study, while the
outer fibre is used to collect the reflected light from that area. Thus, one can
view the inner structure of the object.
The optical light source is attached at the viewing end of the
endoscope. The optical arrangements, which are used to collect and view
the object, are arranged at the respective fibre ends. One can also achieve a
better imagine of the object by interfacing a telescope system at the internal
part of the endoscope.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 173
P a g e | 173
Outer fibre conduct light to object
`
``
Object
Inner fibre conduct image to observer
`
Fig. 8.7 Flexible fibre endoscope
8.7.1 Advantages of optical fibres in communication
(i) A large number of telephone signals, nearly 15,000 signals can be passed
through the optical fibres in particular time without any interface, where as
in an ordinary copper cable, only 48 signals can be passed in particular
time.
(ii) The signal leakage is nil due to total internal reflection and hence there is
no cross-talk.
(iii) It is one of the most ideal means for communication in explosive
environments.
(iv) It gives a nearly foolproof communication during the wartime
(v) The cost of the cables is very low compared to metal cables.
(vi) Optical fibres have immunity to adverse temperature, moisture and
chemical reactions.
8.7.2 Other Applications
(a) In computers
(i) To exchange the information between different terminals in a network
(ii) The optical fibres are used to exchange information in cable television,
space vehicle submarine, etc.,
(b) In industry
The optical fibres are used industrial automation, security alarm system and
process control.
(c) Optical applications.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 174
P a g e | 174
(i) Fibre optic delay lines the fibre optic cables are widely used interethnic
fields to produce required delay
(ii) Fibre optic sensor it is used as a sensor for electric field, magnetic field,
temperature, mechanical force, etc.,
8.8 Waveguide Dispersion Equations:
We had discussed material dispersion which results from the
dependence of the refractive index of the fibre on wavelength. Even if we
assume that the refractive indices n1 and n2 to be independent of λ0; the
group velocity of each mode does depend on the wavelength; physically, this
is due to the fact that the spot – size of the mode depends on the wavelength
this leads to what is known as waveguide dispersion. The detailed theory is
rather involved* but a convenient empirical formula for a step-index single-
mode fiber is given by
w = L/c n2 [0.080 + 0.549(2.834 – v)2 λ0/ λ0
for 1.4<v<2.6
If we assume L = 1 Km = 103m, λ 0 = 1nm and c=3x10-4 m/ps we get
Dw = - n2 /3 λ0 x 107 [0.080+0.549(2.834 –v)2] ps/Km.nm
Where λ0 is measured in manometers. As before, the negative sign indicates
that longer wavelengths travel faster. The total dispersions given by the
sum of material and waveguide dispersions:
Dtot = Dm+Dw
Let us consider the two single – mode fibers discussed earlier.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 175
P a g e | 175
30
20
10
0
-10
-20
-30
1.1 1.2 1.3 1.4 1.5 1.6 λ0 (µm)
The wavelength dependence of Dm, Dw and Dtot
for a typical conventional single-mode fiber (CSF) with parameters as given in example 24.6. The total dispersion passes through zero around λ0 1300nm which is known as zero dispersion wavelength.
8.9 Let us sum up
On going through this lesson you would have understood about the
propagation of light in optical fibre. Also about the meaning of Acceptance angle and
Numerical aperture. Finally about the fabrication technique and the types of optical fibre
and their application.
8.10 Lesson end activities
(i) What is optical fibre?
(ii) Explain Acceptance angle?
(iii) What is meant by Numerical aperature?
(iv) What is intermodal dispersion
8.11 Points for discussion
(i) Explain the wave propagation through optical fibre?
(ii) In an optical fibre, the core material has refractive index 1.6 and refractive
index of clad material is 1.3. What is the value of critical angle? Also calculate
the value of angle of acceptance.
CSF Dm
Dtot
Dw
Dis
per
sion (
ps/
nm
-Km
)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 176
P a g e | 176
[ Hint: Critical angle sin fc =1
2
n
n ; Acceptance angle qa = sin -1 Ö n1
2-n22]
(iii) The numerical aperture of an optical fibre is 0.5 and the core
refractive index is 1.54. Find the refractive index of the cladding.
[Hint: n2 = Ö n12-(NA)2 ]
(iv) A step index fibre cable has an acceptance angle of 30° and a core index of
refraction is 1.4. Calculate the refractive index of cladding.
[Hint : sin2 qa = n12-n2
2 find n2 ]
8.12 Check your progress
(i) Write a note on Fibre losses
(ii) Explain different types of Fabrication technique of optical fibre
(iii) Write any two applications of optical fibres.
8.13 References
(i) Optics and Spectroscopy by R. Murugesan
(ii) Text book of optics by Brijlal and Subramaniyam and M.N. Avadhanulu.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 177
P a g e | 177
UNIT – V
Lesson – 9
LASERS
9.0 Aims and Objectives
this lesson provides a concise account of principles and operation of LASER. It covers the
characteristics of a LASER beam which make it distinct from the light beam.
Also it deals with the type of lasers. Some of the laser commonly used in practical
applications are described in detail. Semiconductor diode laser is given special attention and
the relevant theory and principles are discussed at length.
INTRODUCTION
It is one of the most important inventions of the 20th century. It is a light source but
differs vastly from common light sources and is not used in illumination purposes. Lasers are
useful in radio and microwave transmitters which produce coherent, monochromatic and
polarized beam of electromagnetic radiation. It occupies a unique place in modern
technology. It has wide applications in metalworking, printing, entertainment,
communications, medical diagnosis, surgery, ophthalmology, weapon guidance etc.
9.1 Characteristics of LASER
The discovery of laser made enormous impact in the scientific world and showed that the
function of optics is very much alive.
(i)Directionality
The light beam can travel as a parallel beam up to a distance d2/l, where d is the
diameter of the aperture through which the light is passing and l the wavelength of the
light used. After traveling a distance d2/l, the light beam spreads radially. In ordinary
light the angular spread is given by Dq = l/d. For a typical laser beam, the angular spread
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 178
P a g e | 178
is 1mm per1m, but for ordinary source of light, the angular spread of light is1m per1m.
This shows the directionality of the laser beam. For example the laser beam can be
focused to moon from the earth with an angular spread of a few kilometers.
(ii) Intensity
The intensity of the laser beam is very high. If a person is allowed to observe the
laser beam from the same distance, the entire laser beam penetrates through his eye and
will damage the eye of the observer. This shows the high intensity of the laser beam.
(iii) Monochromaticity
The band width (Dn=0) of the laser beam is narrow, while ordinary light spreads
over wide range of frequencies.
(iv) Coherence
The degree of coherence of the laser beam is very high than the other sources. The
coherence of laser emission results in extremely high power (5´1012 WCm-2). The height
from the laser source consists of wave trains that are identical in phase and directions of
propagation. Also it can be focused to a very small area of 0.7 mm thickness.
9.2 Principle pf LASER (Atomic basis for LASER action)
We can get an idea from the theory of interaction of radiation with matter
regarding the working of laser. Consider an atom that has only two energy levels E1 and
E2.when it is exposed to radiation having stream of photons, each with energy hn , three
distinct processes takes place:
(i) Absorption.
(ii) Spontaneous emission and
(iii) Stimulated emission.
(i) Absorption
An atom or molecule in the ground state E1 can absorbs a photon of energy hn
and go to the higher energy state E2.This process is known as absorption and is shown in fig
9.2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 179
P a g e | 179
Fig. 9.2
The rate of upward transition R12 from ground state E1 to excited state E2 is proportional
to the population of lower energy level N1 (number of atoms per unit volume) and to the
energy density of radiation rn
i.e. R12µ rn
µN1
Thus R12= B12rn N1 ……………………. (1)
Where B12 is the probability of absorption per unit time.
Normally, the higher energy state is an unstable state and hence, the atoms will make a
transition back to the lower energy state with the emission of photon. Such a emission can
take place by one of the following two methods.
(ii) Spontaneous emission
In this the atoms or molecule in the higher energy state E2 return to the ground
state by emitting their excess energy spontaneously. This process is independent of external
radiation .the rate of the spontaneous emission is directly proportional to the population of the
energy level E2.
i.e. R21µN2
R21 = A21 N2……………… (2)
Where A21 is the probability per unit time that the atoms will
spontaneously fall to the ground state and N2 the number of atoms per unit volume in the
state E2.This process is shown in the Fig. 9.2 (i)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 180
P a g e | 180
Fig. 9.2 (i)
Stimulated emission
In this a photon having energy E1 equal to the difference in energy between the two
levels E2 and E1, stimulate an atom in the higher state to make transition to the lower
state with the creation of second photon as shown in figure 9.2(ii)
Fig. 9.2 (ii)
The rate of stimulated emission R21 is given as
R21 = B21rn N2…………. (3)
Where B21 is the probability per unit time that the atoms undergo transition from higher
energy state to lower energy state by stimulated emission. Under conditions of thermal
equilibrium, the population of the energy levels obey the Boltzmann’s distribution
formulae.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 181
P a g e | 181
9.3 EINSTEIN’S THEORY OF STIMULATED EMISSION
In 1917, Einstein proposed a mathematical expression for the existence of stimulated
emission of light. It is known as Einstein’s expression.
Consider two level energy systems (E1 and E2). Let N1 and N2 be the number of atoms
in the ground state and excited state. Assume that only the spontaneous emission is present
and there is no stimulated emission of light.
At thermal equilibrium condition,
The rate of absorption= the rate of emission of light
From the equation (1) and (2)
rn´B12 N1 = A21N2
112
221
NB
NA=nr (4)
according to Boltzmann’s distribution function, the population of atoms in the upper and
lower energy levels are related by
kTE
kTE
e
e
N
N/
/
1
2
1
2
-
-
= (5)
where k is the Boltzmann’s constant and T is the absolute temperature.
Substituting 1
2
N
N in equation (4)
( ) kTEEeB
A /
21
21 12 --=n
r (6)
kTheB
A/
21
21 1nnr
-=
according to blackbody radiation, the energy density
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 182
P a g e | 182
1
18/3
3
-=
- kThec
hnn
npr (7)
Where h is the Plank’s constant and c is the velocity of light. Comparing (6) and (7) one can
observe that they are not in agreement. To rectify this discrepancy, Einstein proposed another
kind of emission known as stimulated emission of radiation. Therefore, the total emission is
the sum of the spontaneous and stimulated radiation.
At thermal equilibrium condition
The rate of absorption= the rate of emission
From the equation (1), (2) and (3) we get
B21N1rn = A2N2 + B21N2rn
221112
221
NBNB
NA
-=nr
(8)
dividing each and every term on the R.H.S of equation (8) by N2, we get
21
2
112
21
BN
NB
A
-÷øö
çèæ
=nr
substituting for 2
1
N
Nfrom the equation (5)
( )( )21
/
12
21
21 BeB
AkTEE -
=-nr
we know that E2-E1= hn. Hence
( )21
/
12
21
BeB
AkTh -
=nnr
(9)
the coefficient A21,B12,B21 are known as Einstein’s coefficients.
Comparing equation (9) with (7) we get B12=B21 and
3
3
21
21 8
c
h
B
A np=
(10)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 183
P a g e | 183
Equation (7) and (9), the ratio of the stimulated emission to spontaneous emission is
given by
( )( ) 1
1/
221
221
21
21
-==
- kTheNA
NB
SPR
STRn
nr
(11)
SP = spontaneous emission
ST = stimulated emission
From the equation (11)
Einstein proved the existence of the stimulated emission of radiation.
Note: the spontaneous emission produces incoherent light, while the stimulated emission
produces coherent light. In ordinary light source the spontaneous emission is dominant. For
laser action, stimulated emission should be predominant over spontaneous emission and
absorption. To achieve this, an artificial condition known as population inversion is required.
9.4 POPULATION INVERSION
Consider a two level energy system (E1 and E2). A photon of energy equal to the
energy difference between the two levels is incident on the system; Einstein showed that
under normal circumstances both the process absorption and stimulated emission are equally
probable.
In a system containing a very large number of atoms, the dominant process will
depend on the virtual number of atoms in the upper and lower state. A large population in the
upper level (N1< N 2) will result in stimulated emission dominating over the absorption. If
there are more number of atoms in the lower level i.e. ground state (N1> N2) there will be
more absorption than stimulated emission.
Under conditions of thermal equilibrium, Boltzmann’s distribution function relating
N1 and N2 given by equation (5) is obeyed
( ) kTEEe
N
N /
1
2 12 --= (12)
where k is the Boltzmann’s constant and T is the absolute temperature. For stimulated
emission to be dominant, it is necessary to increase the population of the upper energy level,
so that it is greater than that of the lower energy level. This is known as population inversion.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 184
P a g e | 184
In the above equation, if T is negative, then stimulated emission would dominate over
absorption. Such a condition is not possible. Different methods to increase population
inversion are given below
9.4.1 METHODS OF ACHIEVING POPULATION INVERSION
(i) Optical pumping
In this method, an external source like xenon flash lamp is employed to produce a
high population in the higher energy level of the laser medium. This is illustrated in
fig9.4.1(a). This method is used in solid state laser.
(ii) Direct electron excitation
The direct electron excitation in a gaseous discharge may be used to produce the
desired inversion. This method is used in some gaseous ion laser. In this type shown
in fig9.4.1(b), laser medium itself carries the discharge current under suitable
conditions of pressure and temperature. In this method the electrons directly excite the
active atoms to achieve higher population in certain higher energy levels compared to
the lower energy levels.
(iii) Inelastic atom-atom collisions
Here, electric discharge method is employed to cause collision and excitation of
the atom. In this method, a combination of two types of gasses are used say A and B, both
having the same excited state A* and B* that coincide. In the first step, during electric
discharge, A gets excited to A* (metastable) due to collision with electrons
A + e ® A* +e1 (13)
A* + B ®A + B*
The excited A* atoms now collide with B atoms so that the latter atoms gets excited to
higher energy B*. This type of transition is used in He-Ne laser.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 185
P a g e | 185
(iv) Chemical reactions
In this method, the molecules undergo chemical changes in which one of the product of
the reaction is a molecule or an atom that is left in an excited state under appropriate
conditions. Now population can occur. An example of this type of laser, in which
hydrogen fluoride molecule in the excited state result from the following chemical
reaction
H2 + F2 ® 2HF
9.4.2 METASTABLE STATES
Normally excited states have short life times and release their excess energyin a
matter of nanoseconds (10-9sec) by spontaneous emission. Atoms do not stay at such
excited states long enough to be stimulated to emit their energy. Though, the pumping
agent continuously raises the atoms to the excited level, many of them rapidly undergo
spontaneous transitions to the lower energy level. Population inversion can not be
therefore established. For population inversion the excited atom are required to wait at
upper level till a large number of atoms accumulate at that level. Thus, what is needed is
an excited state with a longer life time. Such a longer- lived upper levels from where an
excited atom does not return to lower level at once, but remains excited for an appropriate
time, are known as metastable states. Phosphors are an example of materials having
metastable states. They emit persistent- light called phosphorescence because of
metastable states excistent in them. Atoms stay in metastable states for about 10-6- 10-3s.
This is 103 to 106 times longer than the time of stay of atoms at excited levels. Therefore,
Fig. 9.4.1
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 186
P a g e | 186
it is possible for a large atoms to accumulate at metastable level. The metastable state
population can exceed the population of the lower level and lead to the state of population
inversion.
If the metastable states do not exists there could be no population inversion, no
stimulated emission and hence no laser action.
9.4.3 OPTICAL RESONATOR
In laser, feedback is obtained by placing the active medium between pair of mirrors
which are facing each other. The photons triggered many stimulated emissions and these
photons at the end are reflected back into the medium by the mirror. The photon reverse
their direction and travel through the medium till they reach the mirror at the other end.
Hence the pair of mirrors contributes an optical resonator. This structure is actually an
open resonator as the sides open.
As the light bounces back and forth in the optical resonator, it undergoes
amplification as well as it suffer various losses. These losses occur mainly due to the
scattering and diffraction of light within the active medium. For the proper build up of
oscillations, it is essential that the amplification between two consecutive reflections of
light from rear end mirror can balance the losses. The mirrors could be either plane or
curved and are designed such that one of them reflects all the light that reaches it while
the other one reflects most of the light incident on it. A small portion of light is
transmitted through this mirror as laser output beam. The basic components of laser are
shown in fig. 9.4.3
Fig. 9.4.3
(a) Closed cavity used at microwave frequencies
(a) (b) (c)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 187
P a g e | 187
(b) Open cavity used in optical frequencies
A laser consist of an active medium (AM) placed between two mirrors (M and OM)
and is pumped by pumping agent (P). An additional element (IE) may be located within
the cavity to achieve certain desired characteristics in the laser outputs.
9.4.4 ACTION OF OPTICAL RESONATOR (LASER ACTION)
In fig.9.4.4 shows the action of an optical resonator. In this the active centres in the
laser medium are in ground state (Fig.9.4.4 (a)). Through a suitable pumping mechanism,
the medium is taken into the state of population inversion (Fig.9.4.4 (b)). Here the
possibility is that some of the large number of excited atoms decays spontaneously. They
emit photons in various directions. Each spontaneous photon can trigger many stimulated
transitions along the direction of its propagation and stimulated photon will travel in
different directions. In the absence of end mirrors, the net effect would have been that
incoherent light is produced without a specific direction. In the presence of end mirrors a
specific direction is imposed on the photons. Photons propagating along the optic axis of
the pair of mirrors are restricted whereas photons emitted in the other direction will pass
out of the sides of the resonator and are lost, shown in (Fig.9.4.4 (c)).
A majority of photons travelling along the axis are reflected back on reaching the end
mirror. They propagate towards the opposite mirror and on their way they stimulate more
and more atoms and build up their strength as shown in the fig.9.4.4(d). Thus the mirror
provide positive feedback of light into the medium so that stimulated emission acts are
sustained and the laser operates as an oscillator.
The photons that strike the opposite mirror are reflected once more in the medium fig
9.4.4(e). The photons will travel along the axis generating more photons and cause more
and more amplification. Hence stimulated emission increased. At each reflection at the
front end mirror, light in partially transmitted through it. The transmitted component
constitutes a loss of energy from the resonator. When the losses at the mirrors and within
the medium balance the gain, the laser oscillations build up and laser output beam will
become bright as in fig 9.4.4(f). A highly collimated intense beam emerges from the laser.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 188
P a g e | 188
As the end mirrors reflected strongly into laser medium, the light travels are higher
within the resonator. Consequently, the radiation density r(n) is large within the medium
and sustains the stimulated emission. It is desirable that the laser oscillations at one
frequency produce monochromatic light. This is achieved by coating the mirror with
suitable material which is highly reflect photons of the desired frequency and at the same
time absorb the photons of the undesired frequency.
Fig. 9.4.4 Light amplification and oscillation due to action of optical inversion
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 189
P a g e | 189
Threshold condition
For the proper build up of oscillations, it is essential that the amplification between
two consecutive reflections of the light from rear end mirror can balance the losses.
Threshold gain can be calculated by considering the change in intensity of a beam of light
undergoing a round trip within the resonator.
Let us assume that the laser medium fills the space between the mirrors M1 and M2
which have reflectivity r1 and r2. Let L be the distance between mirrors. Let the intensity
of the light beam be the I0 at M1. Then in travelling from M1 to mirror M2, the beam
intensity increases from I0 to I (L), which is given by
I (L) = I0 e (g-as)L (1)
Where g(=-a) is referred to as the gain coefficient per unit length and is a positive
quantity. After reflection at M2, the beam intensity will be
I (2L) = r1r2I0e (g-as) 2L (2)
The amplification obtained during the round trip is
( ) ( ) LserrI
LIG 2
21
0
2 ag -== (3)
The product r1r2 represents to losses at the mirrors whereas as includes all the distributed
losses such as scattering, diffraction and absorption occurring in the medium. These
losses are balanced by gain, when 1³G or I(2L)=I0. it leads to
( )
12
21 ³- Lsrr
ag
( )
21
2 1
rre Ls ³-ag
taking log on both sides, we get
( ) 21ln2 rrL s -³-ag
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 190
P a g e | 190
( ) 21ln2
1rr
Ls -³-ag
21ln2
1rr
Ls -³ ag
21
1ln
2
1
rrLs +³ ag
The above equation is the condition for the lasing. It shows that the initial gain must
exceed the sum of the losses in the cavity. This condition is used to determine the
threshold value of pumping energy necessary for the laser action g, the amplification of
the laser will be dependent on how hard the laser medium is pumped. As the pump power
is slowly increased, a value of gth called threshold value will be reached and the laser
starts oscillating. The threshold value is
21
1ln
2
1
rrLsth += ag (5)
Therefore for the laser to oscillate,
thgg ³
the above equation is the threshold condition for lasing and states the criterion when the
net gain would be able to counteract the effect of losses in the cavity.
9.5 SOLID STATE LASER.
The term solid state has different meanings in the field of electronics and laser. In this
type the active centres are fixed in a crystal or glassy material. They are electrically non-
conducting. Also called doped insulator laser.
The basic principles that underline the operation of solid state laser is shown in
Fig 9.5.5. The active centres are dispersed in a dielectric crystal or a piece of glass. The
crystal atoms do not participate directly in the lasing action but acting as a host lattice to
the active centres which are present in small concentrations. The crystal is usually shaped
into a rod with reflecting mirrors placed at each end. Light from an external source such
as flash lamp excites the active centres in the rod and linear in shape. The linear lamp and
laser rod are placed close to each other in a reflective cylinder, which focusses the pump
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 191
P a g e | 191
light on the rod. The mirrors on the sides of the laser rod form a resonant cavity and
provide necessary feedback to generate laser beam.
It is required tha the host material should be transparent to the pump light and should not
absorb light at the laser wavelength. Most of the excitation energy ends up as heat rather
than light. Excess heat damages the laser crystal and hence has good thermal
conductivity. Normally, cooling can be done by circulating water.
Fig. 9.5 Solid State Laser
The active centres are ions of the metallic elements like Chromium, Neodymium (Nd),
erbium (Er), cerium (Ce) and titanium (Ti). These lasers are simple to maintain and
capable of generating high peak powers.
9.5.1 SEMICONDUCTOR LASER.
A semiconductor diode laser is a specially fabricated pn junction device which emits
light when it is forward biased. It is made from gallium arsenide (Ga-As) which operated
at low temperatures and emitted light is in the near IR region.
It is a pn junction diode with a p-type and n-type regions heavily doped. Under large
applied forward bias, electrons and holes are injected into and across the transition region
in considerable concentration. As a result, the region around the junction contains a large
number of electrons within the conduction band and a large number of holes within the
valence band. When the population density is high enough, a condition of population
inversion is achieved and recombination may be stimulated resulting in laser action.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 192
P a g e | 192
In the case of semiconductor laser there is no need of external mirrors. In a
germanium or silicon semi-conductor, due to such a recombination, only heat is
generated.
A typical Ga-As laser is shown in fig.9.5.1
9.5.1 Ga-As laser schematic representation
The Ga-As laser convert electricity into light. The efficiency can go up to 100% when the
temperature is reduced to 100K. The operating current is supplied from a pulse generator
upto 5-20ms. The energy separation between the conduction band and valence band is 1.4
eV and hence, the wavelength of the light emitted is 8874A° at room temperature.
Advantages
1) The efficiency is more and can be increased by decreasing the temperature
alone.
2) It can have a continues wave output or pulsed output.
3) The modulation of the output is possible.
4) It is highly economical and the arrangement is compact.
Disadvantage
The spatial and temporal coherence are very poor.
9.5.2 GAS LASER- HELIUM-NEON LASER
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 193
P a g e | 193
The advantage of gas laser is that they can be operated continuously. The gas laser
show exceptionally high monochromaticity and high stability of frequency. The output of the
laser can be tuned to a certain available wavelength. Hence the gas lasers are widely used in
the industries.
Helium-Neon LASER
Fig. 9.5.2 He-Ne laser
The two important parts of the laser are as follows.
Active medium:
The active medium used in this type of laser is a mixture of helium and neon gas.
These two gases are mixed under pressure of 1mm Hg of helium and 0.1mm Hg of neon in
the ratio of 10:1 the mixture of these gases is filled within the discharge tube for laser action.
Gas discharge tube
It is made up of fused quartz tube with a diameter of 11.5 cm and length 80 cm. The
end faces of the discharge tube is tilted at the Brewster angle known as Brewster window,
since the window are transparent to the preferred direction of polarisation. A fully reflecting
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 194
P a g e | 194
concave mirror is placed at one end of the discharge tube and a partially reflecting concave
mirror at the other end.
Working Principle
The population inversion for laser action was achieved by inelastic atom-atom
collision. The collisions of the atoms are made using any one of the following methods.
(i) Direct current discharge
(ii) Alternate current discharge
(iii) Electrodeless high frequency discharge
(iv) High voltage pulses
Due to electric discharge in the gas, an energetic electron interacts with the ground state
helium atoms. The impact of the electron results in exchange of some of its energy to the
helium atom. As a result, helium atoms are excited to higher levels1S0 and 3S1 known as
metastable state. The life time of these levels are relatively low. The collision of the first kind
is represented as
He + e 1 ® He* +e2 (1)
These two energy levels are very close to 2s and 3s levels of the neon atom and the collision
of the second kind takes place between the He and the Ne atoms, the neon atom goes to the
excited state,
He* +Ne ® He +Ne* (2)
Explanation
Therefore there are three types of transitions one from 2S to 2P the other from 3S to
3P and another from 3S to 2P levels of the laser action. These transitions constitute laser
beam in the infrared region (11,523 A° , 33912 A° ) and 6328 A° in the visible region. The
transition from 3S to 3P at 3.39mm will have adverse effect on the laser emission. The line is
suppressed in order to get a maxim um power output at 6328 A° .
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 195
P a g e | 195
In the fig 9.5.2(a), the energy levels are represented in terms of Paschen’s notation
and their corresponding electronic configuration are given in parentheses. The neon atom in
the terminal level 2P decay very rapidly to the 1S metastable state in 10-8 s, much faster than
spontaneous rate of decay from 2S to 2P level. Thus, the lower is relatively kept empty and
population inversion achieved between 3S and 2P.
9.5.3 MOLECULAR GAS LASERS-CORBON DIOXIDE LASERS(CO2)
This laser was first developed by C.K.N.Patel
Principle: To understand this, one has to recall the rotational and vibration spectrum of CO2
molecules. The three atoms can be considered as a ring over a straight line, the outer atom
being O with a carbon atom C at the centre. There are three models of vibrations and in each
mode the centre of gravity remains fixed.
Fig. 9.5.3 Co2 molecules- different modes of vibration
1) As shown in above fig 9.5.3(a), the carbon atom is fixed in its position and each oxygen
atom vibrate in the opposite direction symmetrical to the carbon atom with each other along a
straight line and is known as symmetric mode of vibration. The corresponding frequency is
called symmetric stretching frequency.
2) as shown in fig 9.5.3(b), the oxygen atom and the carbon atom may vibrate at right angles
to the line passing through centre of gravity. This is known as the bending mode and the
corresponding frequency is bending frequency.
3) as shown in fig 9.5.3(c), in asymmetric mode of operation, two oxygen atoms may vibrate
about the central C atom asymmetrically, and at the same time the carbon atom also vibrates
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 196
P a g e | 196
from its mean position. The corresponding frequency is called asymmetric stretching
frequency.
In addition to the above three vibrational modes the molecule can rotate and the quantized
rotational energy levels are also possible. A series rotational levels is associated with each
vibrational levels are denoted by J values.
Construction of CO2 laser
This laser depends on radiation power, i.e. output-power on the diameter of the tube.
It can be raised by increasing the diameter. In a powerful CO2 laser, the length of the
discharge tube is several metres and its diameter will be several centimeters as shown in the
fig 9.5.3(d)
Fig. 9.5.3(d) Co2 laser
The laser is powered with an a.c supply of frequency 50 cycles or d.c supply. In order to get a
high output power, a metallic mirror of gold is employed for proper reflection, the laser are
either water cooled or air cooled. The efficiency of CO2 laser is about 30%. The gas mixture
can be pumped either longitudinally or transversely into the gas discharge tube.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 197
P a g e | 197
Working principle
The CO2 laser uses two additional gasses, N2 and He. The nitrogen plays a similar
role as that of the He in the He-Ne laser. The N2 molecules go into an excited state by
collision of first kind with the electrons.
N2 +e1 ® N2* + e2 (1)
The excited atom undergoes a collision of the second kind and makes the CO2 molecule to be
excited.
N2* + CO2 ® CO2* + N2 (2)
The frequency of the CO2 laser in an energy level diagram is shown in fig 9.5.3(e)
Due to various factors, the most powerful transition in CO2 laser at normal operating
temperature occurs at 10.6mm. The operating temperature plays an important role in
determining the output power of the laser. The contamination of carbon monoxide and
oxygen will also have some effect in the laser action. The unused gasses can be
Fig. 9.5.3(e) Co2 laser- Energy level diagram
pumped out and fresh Co2 must be pumped. The temperature can be reduced by restricting the
tube diameter and also the addition of helium to the mixture of N2 and He. The helium serves
not only to improve the conductivity of heat to the walls of the tube, but also decreasing the
population in the lower levels. The power output coming from this laser is 10KW. It is a four-
level molecular laser and operates at 10.6 mm in far IR region.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 198
P a g e | 198
9.5.4 Ion gas lasers à Argon laser
Helium, Neon, Argon, Xenon and Krypton are noble gases and they have electronic states
capable of laser transitions. If these noble gases are first ionized by electron collisions then
they are easy to pump.
Argon- ion laser:
General description: The argon laser belongs to the group of ion lasers. It is a four- level laser
which operates in visible region over a wavelength range from 3510 to 5200 . It is used in
laser light shows. The Argon-ion laser can provide 25 visible wavelengths ranging from
4089 to 6861 and more than 10 UV wavelengths ranging from 2750 to 3538 .
Structure: In Argon lasers, the active medium is argon gas and the active centres are ionized
Argon atoms. It is shown in fig 9.5.4.
Fig. 9.5.4 ion laser tube
It consists of a narrow-water-cooled ceramic tube in which an arc discharge takes
place. The electrodes are arranged at ends of the capillary tube. The anode and cathode
spaces communicated through a return gas path which ensures free circulation of the gas. A
magnet surrounds the discharge tube. Its function is to reduce the discharge area and increase
the concentration of ions along the axis of tube. This increases the output power and
efficiency.
Working: The initial high voltage pulse ionizes the gas so that it conducts current Electrons in
the current transfer energy directly to Argon atoms, ionizes them and raises the ions to a
group high energy levels, shown in the fig 9.5.4(i). It is about 35eV above the ground state of
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 199
P a g e | 199
neutral Argon atom. Different processes populate the metastable upper laser level. They are
(a) electron collisions with Ar positive ions in the ground state, (b) collisions with ions in
metastable states and (c) radiative transitions from higher states. The life time of upper level
is 10-8sec while that of the lower level is of the order of 10-9sec. Therefore, the condition for
population inversion is satisfied. Transitions can occur between many pairs of upper and
lower lasing levels and therefore many laser wavelengths are emitted.
But of the large number of wavelengths emitted, the most important and the most
intense lines are 4881 (blue) and 5145 (green). Argons ions quickly drop from the
Fig. 9.5.4(i) Energy level for an argon atom
lower level to the ground state of the ion by emitting UV light at 740 . The ground state ion
either recaptures an electron and becomes a neutral atom or it is again excited to the upper
levels.
During the operation the positive ions tend to collect at cathode where they are
neutralized and diffuse slowly back into the discharge. However it leads to a pressure
gradient. A gas return path is provided between anode and cathode to equalize the pressure.
Without such return path, the discharge may eventually be extinguished.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 200
P a g e | 200
This laser needs active cooling. Any desired wavelength can be selected through
proper cavity optics. For example, a small prism may be inserted into the optical cavity and
the position of the end mirror is rotated such that it comes into a position normal to the path
of the light of desired wavelength. Then, only the desired wavelength is sustained in to and
fro reflections while the other wavelengths are lost from the cavity after a few reflections.
9.5.5. SOLID STATE LASERS
RUBY LASER:
In solid state laser, the active medium is a crystalline substance. It is the first
successful laser achieved by Maiman in 1960. The experimental setup is shown in figure
9.5.5. It has three parts
Fig. 9.5.5 Ruby Laser
(i) Ruby rod: Ruby is a crystalline substance of aluminium oxide doped with 0.05% by
weight of chromium oxide. The resultant pink colour is due to the presence of Cr3+ ions in the
appropriate concentration which replace aluminium atoms in the crystal lattice.
(ii) Resonating cavity: For this, a pink rod of 4 cm length and 0.5 diameter is used. The end
faces of the rod are grounded so that they are parallel and polished to a high degree. Further,
the end faces are silvered in such a way that one end face becomes fully reflecting while the
other end partially reflecting. In some cases, separate glass pieces are attached at the end
faces.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 201
P a g e | 201
(iii) Xenon flash tube: The laser action is achieved by using optical pumping. Helical xenon
flash tube that surrounds the ruby rod provides the pumping light to raise the chromium ions
to the upper energy levels. The flash of the xenon tube lasts for several milliseconds and the
tube consumes several thousand joules of energy. Only a part of the energy is used to excite
the Cr3+ ions, while the rest heats up the apparatus. A separate cooling arrangement is used to
reduce the temperature.
Ruby Laser-Energy level diagram
Working principle: The energy level diagram illustrating the operation of ruby laser is
shown above. The pumping light from the flash tube is absorbed by the Cr3+ ions raising them
from the ground state E0 to the excited state E1 or E2. From these levels, a rapid radiationless
transition to the level E, which is a metastable state, takes place. Decay from E is relatively
slow so that with sufficient excitation, population inversion between E and the ground state
E0 can occur. The photons are allowed to pass back and forth millions of times in the active
medium with the help of mirrors at the ends. When the condition for laser action is satisfied,
an intense pulse of light of the wavelength 6934 corresponding to the transition E to E0 is
obtained.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 202
P a g e | 202
9.6 Q-SWITCHING
There are basically three different techniques to obtain high pulse from a laser. They
are known as Q-switching, cavity dumping and mode locking. For many of the practical
applications, we require large powers even for a short time. Q-switching and cavity dumping
make large peak powers available for short time.
The method of controlling the laser output power is called Q-switching. Evaluation of
giant- laser pulse through Q-switching is shown in fig.9.6. Q-switching is widely used, but not
all laser can be Q-switched.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 203
P a g e | 203
Fig. 9.6 Evolution of a giant pulse through Q-switching
Method of Q-Switching –Electro Optic Shutter:
It can serve as voltage controlled gate which rapidly switches the cavity from high
loss to a low loss condition. It consists of a crystal that becomes double refracting when an
electric field is applied across the crystal. It is shown in fig. 9.6(i)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 204
P a g e | 204
9.6 (i) Electro-optic cell used as Q-switch
A voltage applied to the crystal during the pumping of the laser by the light from the
flash lamp. The magnitude of the voltage is chosen such that it transforms the electro-optic
crystal into a quarter wave plot. The light emitted by the laser becomes linearly polarized
light on passing through the polarizer. The linearly polarized light incident on the crystal
splits into mutually orthogonal components. As the two component travel through the electro-
optic crystal, a phase retardation of 900 is produced between them. On emerging from the
crystal, the two components combine to produce circularly polarized light. The light beams
reflect from the mirror and returns to the cavity travelling in the opposite direction. On
reflection the sense of rotation of circularly polarized light reverses in and on re-passing
through the crystal, the two components of circularly polarized light experience a further
retardation of 900 with respect to each other. Coming out of the crystal at the other end, the
components recombine to produce linearly polarized light. The direction of polarization of
light is now at 900 with respect to its original direction of polarization and transmission axis
of the polarizer. This light is not allowed to pass through the polarizer. Hence, light does not
comeback to the laser rod and the cavity is switched off. Thus cavity Q is reduced to a low
value. When the voltage applied to the crystal turned off, double refraction is absent in the
crystal and the state of polarization of the light passing through the crystal is unaffected.
Light now freely travels in both direction and returns to the laser rod to be further amplified.
Now the cavity is switched on and the Q regains its high value. The Q-switching is
synchronized with pumping mechanism such that the voltage applied to the crystal drops to
zero value at the time when the population inversion in the laser medium attains its peak
value. Two types of electro-optic shutters known as Kerr cell and Pockels cell are available
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 205
P a g e | 205
for Q-switching. In this Pockel cell is preferred over the Kerr cell because of the lower
voltage needed to produce desired effect.
9.6.1 MODE LOCKING FOR ULTRA SHORT PULSES
Q-switching produces giant pulses of duration of about 10-7 to 10-8 sec. The technique
of mode locking allows the generation of pulses of lesser duration of about 10-11 to10-12 sec.
The matching of phase of different modes are called mode locking. An essential requirement
for mode locking is existence of multimode operation in the active medium. The broadened
laser cavity in general supports oscillation in many axial modes simultaneously. The axial
modes are uniformly spread out on the frequency scale and are separated by ν given by
c/2mL.The resultant output of an ordinary laser (laser without mode locking) depends on the
relative frequencies, phases, phases and amplitudes of these modes. All these parameters are
time-varying, because each of the modes oscillates independently on the other mode.
The following figures illustrates
(i) output of a non-mode locked laser consisting of random fluctuations in intensity
(ii) Output of a mode locked laser, consisting of a narrow intensive pulse.
(iii) All modes are in phase
(iv) Time spacing and duration of pulses are produced by a mode- locked laser.
Thus we find the output of a mode- locked laser consist of a sequence of short pulses
at time intervals of 2µL/c; which is equal to the round trip transits time for light within
the cavity. Each pulse has a peak power equal to N times the average power. The
mode-locked condition can be viewed as a condition in which a pulse of light is
bouncing back and forth inside the cavity and every time it hits the mirror, a fraction
of it is transmitted as the output pulse
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 206
P a g e | 206
Output of a non-mode laser consisting of a random fluctuations in intensity
Output of a mode locked laser consisting of a narrow intense pulses
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 207
P a g e | 207
Time spacing and duration of pulses produced by a mode- locked laser
9.6.2 TECHNIQUES FOR MODE LOCKING
The most common methods are
1. Mode locking by modulating the loss of the cavity externally
2. Mode locking by means of a saturable absorber placed inside the laser cavity.
In the first an acoustic or electro-optic modulator is driven by an external signal and hence it
is called active mode locking. In the second, employs a saturable absorber whose absorption
coefficient varies non-linearly with light intensity and is called passive model locking
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 208
P a g e | 208
1. Active mode locking:
Kerr cell and Pockels cells or the acousto–optic shutter can be used as modulators for
achieving mode locking. If we consider electro-optic modulator (EOM) is kept inside the
cavity. Let the frequency of one of the axial modes be νm. Let the frequency of modulation be
equal to ∆ν, the intermode frequency separation. Since, the signal with frequency ∆ν is
applied to the electro-optic modulator, the loss of the laser activity is modulated at the same
frequency ∆ν. Consequently, the amplitude of the mode corresponding to ν m is also
modulated at ∆ν. The amplitude modulated mode at a frequency ν m generates two side bands
at frequencies ν m+ ∆ν and νm- ∆ν. Since ∆ν is chosen such that it is equal to intermode
spacing, the waves having new frequencies νm+ ∆ν and νm- ∆ν coincide with the two modes
located on either side of νm. Three modes become locked in phase. Since, the amplitudes of
these new modes are also modulated at the frequency ∆ν, they generate new side bands which
in turn correspond to some other axial modes. The process continues and at the end all modes
are forced to oscillate with a definite phase relationship. As a result, mode locking is
achieved.
2. Passive mode locking
Here a saturable absorber (an organic dye) placed within the laser activity. The solution has
the property of becoming more and more transparent as the intensity of the light falling on it
increases. The saturable absorber placed adjacent to one of the resorant mirrors. Initially, the
laser medium emits spontaneous radiation which forms within the cavity a radiation field
consisting of fluctuations of intensity having a noise like structure. Some of the intensity
peeks within the fluctuations bleach the saturable absorber more than the other components
and pass through. They get amplified in its round trip through the active medium. If the cell
recovery time is too short, the cell becomes opaque (absorbing) immediately after this pulse
passed through. As other fluctuations are weak, they are attenuated in the cell. Thus, the
strongest pulse will grow faster while the other less intense fluctuations are suppressed. At
regular intervals of time T, this pulse arrives at the output mirror and it is partially emitted.
Therefore, the laser output consists of a regular sequence of pulses with a pulse repetition
time of T.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 209
P a g e | 209
9.7 APPLICATION
LASER is bringing revolutionary changes in our lives. They are applied in
entertainment electronics, communications, surgery and related medical fields, information
processing and even in warefare. The list is unending and continue to grow. Laser
applications are broadly divided into two groups, one involving beams of high power and the
other involving beams of low power. High power, gas and solid laser are used in material
processing, nuclear fusion, medical field, defence etc. the low power semi-conductor laser are
used in CD players, optical floppy discs, optical memory cards, data processing and
information processing devices, hologram preparation and viewing, range finders, optical
communications etc.
9.7.1 LASER IN MECHANICAL ENGINEERING
It is used in cutting, drilling welding etc. jobs to be done on both metals and non-
metals. It requires transfer of energy from the laser beam to workpiece. It can happen only if
the material has high absorption at the wavelength corresponding to the laser beam. Once the
surface of the material absorbs energy, the material starts to melt and then vaporize. At high
intensity of radiation the vapour will be ionized to produce plasma. The plasma layer formed
between the laser and the workpiece absorbs light and prevents the laser beam from reaching
the workpiece. Therefore it is essential that the plasma should be removed to increase energy
coupling.
(i) Drilling: Drilling holes by a laser beam is based on the intense evaporation of materials
heated by a series of powerful light pulses of short duration of 10 -4 to 10 -3 sec. The energy
supplied for drilling should be such that rapid evaporation of material takes place before
radial distribution of heat into the work piece occurs. Use of short pulses minimises the
energy diffused laterally into the work piece and assists in controlling the size and shape of
the hole. Laser drilling is a non-contact process and does not require a physical drill bit. The
problems of wear and broken drills do not arise. The process becomes faster and drilling
operation can be done with extremely high precision and in any desired direction. CO2 laser
and Nd-YAG laser are used, the former laser is suitable for drilling in metallic and non-
metallic materials, the latter is used for drilling holes in metals only.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 210
P a g e | 210
(ii) Cutting: A wide range of materials can be cut by CO2 lasers. The materials include
paper, wood, cloth, glass, quartz, ceramics, steel etc. Laser cutting is done with the assistance
of air, oxygen or dry nitrogen gas. The advantage of laser cutting is that it is fine and precise.
The process does not introduce any contamination. It is automatised and high production
rates can be achieved.
(iii) Welding: It is joining of two or more pieces into a single unit. The metal plates are held
in contact at their edges and a laser beam is made to move along the line of contact of the
plates. The laser beam heats the edges of the two plates to their melting points and causes
them to fuse together where they are in contact. The advantage in this is that there is no
possibility of introduction of impurities into the joint. The work pieces do not get distorted as
the total amount of power input is very small compared to the conventional welding
processes. The most common laser used in welding is the CO2 laser. Gases like He, Ar, N2
are often used with laser welding for protection against oxidation of metal surfaces.
9.7.2 Lasers in Defence
Here it involves mainly ranging , guiding weapons to the intended target and laser
beam itself acting as a weapon. The optical radars, using laser beams, are employed in
detecting distant objects and collecting information about them. This is called range finding.
Nowadays, low flying aircrafts are used in ground attacks. They are fitted with laser
instrumentation (CO2 laser) for measuring the range of the target and guiding the bomb. Also
used in intercontinental Ballistic Missiles, which can hit the target within a few minutes. It is
required that the missile should be detected and destroyed before it can reach the target.
9.7.3 Lasers in Medicine
(i) They are used in destroying stones and gallstones. An optical fibre is threaded through
until it faces the stone directly. Laser pulses launched through the fibre shatters the stone into
small pieces that can pass through without pain.
(2) At times Tumors develop in brain and spinal cord. They cannot be operated in
conventional way because of their delicate nature. Laser surgery in such areas is possible and
safer.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 211
P a g e | 211
(3) Laser welding of human blood vessels is more advantageous than conventional suturing.
Argon laser with 1mm spot size is used for this purpose.
(4) It is possible with laser to selectively destroy cancer. In the treatment, a dye called
hemotopophyrin derivative (HPD) is injected into the patient body. The healthy cells flush
out the dye and the dye in the concentrated in cancer cells. The suspect areas are illuminated
with red light (628-632nm) and cells having HPD absorb light strongly. As a result, the dye
reacts chemically with oxygen in cancer cells and produce a highly reactive form of oxygen
that kill the cancer cells.
9.8 Let us sum up
From this unit you have learnt about the principles and operation of lasers. It dealt
with different types of lasers, its relevant theory and principles. Also it dealt with the
characteristics of laser beam which makes distinct from the light beam generated by
common light sources. Finally Q-switching and mode locking and their advantages
are discussed which will be useful. Finally laser application in mechanical, defence
and medical field are discussed.
9.9 Check your progress
(i) What is population inversion and spontaneous emission?
(ii) What are the characteristics of LASER light?
(iii) What is metastable state and mention its importance?
(iv) What is the working principle of He-Ne Laser?
(v)What is the reason for monochromaticity of laser beam?
9.10 Lesson end activities
(i) Find the ratio of population of the two states in He-Ne laser that produces light
of wavelength 6328 at 27°C .(Ans: 1.1*10-33 )
[Hint: N2 / N1 = e- (E2-E
1)/KT
E2-E1 = 12400/6328 eV = 1.96 eV]
This watermark does not appear in the registered version - http://www.clicktoconvert.com
Page 212
P a g e | 212
(ii) The half-width of the gain profile of a He-Ne laser material is about 2 * 10-3 nm
(∆λ). What should be the length of the cavity in order to have a single longitudinal
mode oscillation? [λ = 6328*10-10m and µ = 1] (Ans: 10cm)
[Hint: L = λ2/2µ∆λ]
9.11 Points for discussion
(i) What do you understand by an optical resonant cavity? Explain.
(ii) What is a semi-conductor laser? Discuss its main features and condition for action.
What are its merits and de-merits.
(iii) (a) Describe the working of solid state laser.
(b) Explain the principle and working of a He-Ne laser.
(iv) Write short notes on (a) Q-Switching
(b) Mode locking
(c) CO2 laser
9.12 References
(i) An introduction to LASERS
by M. N. Avadhanulu (S. Chand & Co)
(ii) Optics
by Brijlal & Subramaniam and M. N. Avadhanulu
This watermark does not appear in the registered version - http://www.clicktoconvert.com