PLANE WAVES AT MEDIA INTERFACE

In the previous chapter we studied the solution of the Maxwell's equations for time varying fields in an
unbound medium. We found that the time varying fields exist in the form of a plane transverse
electromagnetic wave in an infinite medium. For our convenience we oriented the co-ordinate system to
align with the direction of the wave motion and studied the behaviour of electric and magnetic fields in the
plane transverse to the direction of the wave propagation. The arbitrary orientation of the co-ordinate axis
was possible because in an infinite medium, there is no special direction as the medium looks same in all
directions. This situation is obviously an hypothetical situation. In practice we never see a medium which
is uniform in all directions. In the next few chapters we will gradually make the medium bound and try to
capture the EM wave in a completely closed space. The first step towards this would be to make the
medium semi-infinite i.e., divide the space into two semi-infinite regions with different medium properties
and study the behaviour of a plane wave at the interface of the two regions. Specifically we investigate
the transfer of fields and power from one medium to other. In this chapter we ask questions like- what
happens to the plane wave nature of the wave at the interface ? What fraction of the wave energy is
transported to the second medium ? What happens to the polarization of the wave etc. and try to find
their answers.
It is very clear that since the space is no more symmetric the coordinate axes can not be oriented
arbitrarily. If we orient the co-ordinate axes along the media interface, the wave will be travelling in an
arbitrary direction (not along any of the axes) and if the coordinate axes are oriented to get one of the
axes along the wave motion, the axes will have arbitrary orientation with respect to the media interface.
Generally the coordinate axes are aligned along the interface and the wave is assumed to travel at an
angle with respect to the media interface. It is therefore necessary to formulate the wave function for a
wave which is travelling at an arbitrary angle with respect to the co-ordinate axes.
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{Nfig5.1.eps}
\caption{Wave travelling in an arbitrary direction} \end{figure}
PLANE WAVE IN ARBITRARY DIRECTION
As we have seen earlier a plane wave is described by a phase front which is a plane perpendicular to the
direction of the wave motion. As shown in Figure 5.1, let us consider a wave travelling in some arbitrary
direction, and let the unit vector in the direction of wave motion be denoted by . If the unit vector
makes angles , , respectively with the three axes we have

where , , are called the direction cosines of the vector .

The planes perpendicular to are then the phase fronts or the constant phase planes. Let us consider
one of the phase fronts as shown in Figure 5.1, and let any point P on the plane has coordinates .
The vector can be written as

From Figure 5.1 we can see that the dot product

Eqn. (5.3) is valid for any point on the phase front and hence we can write the equation of the phase front
as

If we assume that phase of the plane which passes through the origin is zero and if the phase constant of
the wave is , the phase change over a distance of OA will be

The electric field of a plane wave travelling in direction can then be written as

where is a constant vector( MATH ).

For a transverse electromagnetic wave, lies in the constant phase plane and hence is perpendicular to
. We therefore have
If we now define the vector propagation constant (also called the wave vector) as

The field can be written as

with

where we have defined MATH , MATH , and MATH .

From eqns (5.9) and (5.10) we can verify that if the wave travels in direction ( as taken in the previous
chapter) = = and = 0, and the wave vector becomes

The corresponding electric field can be written as

Since is perpendicular to , it lies in the xy-plane. Expression (5.12) is identical to the plane wave
expression (4.54).
The magnetic field of a wave can be obtained by substituting in the Maxwell's equation

where MATH , MATH and MATH , and MATH

Since , , are constants,

Similarly, we can get

Substituting for partial derivative operators in eqn (5.13), the magnetic field can be written as

From eqn (5.15) we can note that is perpendicular to both and and since is perpendicular to ,
the three vectors , and are perpendicular to each other (Transverse electromagnetic wave).
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{Nfig5.2.eps}
\caption{Transverse nature of electromagnetic wave} \end{figure}
Since the magnitude of the wave vector MATH , and MATH ( intrinsic impedance of the
medium), eqn (5.15) yields

Since MATH , the magnitude of the magnetic field is equal to .

EXAMPLE 5.1 : A MHz uniform plane wave is travelling in free space along a direction which makes
angle with the -axis and angle with the -axis. Find the expression for the vector electric and
magnetic field . The electric field is linearly polarized and has a peak amplitude of V/m. Also, the -
component of the field is twice the -component of field.

Solution:

In free space the velocity of the wave is MATH

The direction cosines of wave normal are

Since,
The unit vector along the direction of the wave normal is

The electric field is given by

is a constant vector perpendicular to the wave normal.

Since, the wave is linearly polarized , and should be in phase. Therefore, without losing
generality, let us assume them to be real.
Since, for uniform plane wave is perpendicular to , we have

It is given that,

MATH
and the peak amplitude
Solving A, B and C, we get,

and the vector electric field is given as

MATH

Using eqn (5.15), the vector magnetic field is given as

MATH

Phase Velocity and Wavelength

As seen above, the choice of coordinate system is generally guided by the media boundaries etc., and
the wave travels in an arbitrary direction. It is however useful to find the velocity of the wave along the
principal co-ordinate axes. As will be seen in the following, the phase velocity of a wave along the
principal axes is not a simple vector resolution of the phase velocity of the wave. The phase velocity along
an axis is the velocity of the constant phase point along that axis.
Let us consider a wave with wave vector . The electric field for this wave is given by (see eqn (5.11)).

Let us now find the phase velocity along the z-axis. Rewriting (5.18) we get

Two things should be noted from eqn (5.19).

(1) In the plane (plane perpendicular to -direction) the phase is not constant. So -plane is not a
constant phase plane.

(2) The phase constant along z-direction is MATH .

Since is always less than or equal to unity, is always . The phase velocity in the -direction
therefore is

where is the phase velocity of the wave in the direction i.e., actual phase velocity of the wave.
Similarly we can get the phase velocities along the and directions as

It is interesting to see that since , , , the velocities , , are always

greater than or equal to . In fact when any of the angles , , MATH , the cosines of these angles
tend to 0 and the corresponding velocities approach infinity. The bounds of the phase velocity therefore
are
The wavelength of the wave in directions respectively are

where = is the wavelength of the wave.

From eqns (5.23)-(5.24) we should note that the phase velocity and the wavelength are smallest in the
direction of the wave motion.
If we consider the unbound medium as the free-space, the phase velocity of the wave is ( velocity
of light in vaccum). Equations (5.23)-(5.24) then yield

\begin{figure}[h]
\includegraphics[width=1.2\textwidth]{Nfig5.3a.eps}
\caption{Plane wave in yz plane} \end{figure}
One would then wonder how come the velocity is greater than the velocity of light, and if we can achieve
infinite velocity can we send an information with infinite speed? The answer to this question can be
obtained from Figure 5.3.
For geometrical simplicity let us consider a plane wave travelling in the -plane. Let a wavefront intersect
the -axis at point A at . After a time the point A moves to A'. The phase front now intersects the
-axis at B' which corresponds to point B at the wavefront at . Similarly, the wave front intersects the
-axis at point C' which corresponds to point C on the wavefront at . While calculating the phase
velocity along -direction or -direction, we do not actually find the distance travelled by a particular point
on the wavefront but find distance between two different points. Consequently we get a velocity which
appears to violate the physical laws (like velocity ). The velocity of energy corresponds to the distance
travelled by point A from A to A'. It is interesting to note from Figure 5.3 that as the angle increases the
distance BB' increases but the distance CC' decreases. Consequently, the phase velocity increases
but the phase velocity decreases and for MATH , . For MATH the distance CC' and
consequently but in this case also MATH if MATH . We can therefore conclude that the phase
velocity is not a velocity of a physical point on the wavefront and hence is not guided by the physical laws.
Obviously then we can not send information with infinite speed or for that matter with a speed greater
than the velocity of light.
It is worthwhile to ask a question at this juncture that with what velocity the point A moves along the -
axis? Since the point A moves with velocity in the direction of , its velocity in -direction is

is called the group velocity of the wave and denotes the group velocity in -direction. The group
velocity corresponds to the physical motion of a point on the wavefront and therefore gives the speed with
which the energy travels. As can be seen from eqns (5.26), is always less than . From (5.22) and
eqn (5.26) we also note that

In fact, eqn (5.27) is true for any arbitrary direction and hence we can make a general statement that `the
product of the group and phase velocity in any direction is equal to the square of the velocity of the wave
in the medium'. Figure 5.4 shows the exclusive domains of and . is always greater than or equal
to the velocity of the wave in the medium, whereas, is always less than or equal to the velocity of the
wave in the medium. As MATH , and as MATH also .
\begin{figure}[h]
\includegraphics[width=0.5\textwidth]{Nfig5.4.eps}
\caption{Domains of and } \end{figure}
EXAMPLE 5.2 : The complex vector electric and magnetic fields given by ideal dielectric medium are
given
Show that these fields represent fields of a uniform plane wave. Find the frequency and velocity of the
wave. Also find the phase velocities along the and and -directions. Find the state of polarization of
the wave.

Solution:
\begin{figure}[h]
\includegraphics[width=.7\textwidth]{wave5.2.eps}
\end{figure}For uniform plane wave , and the wave normal are perpendicular to each other.
Therefore, , and should be zero.
Now

The fields therefore represent fields of a uniform plane wave.

Now taking the coefficient of in the exponent, we get,

The components of and are not in time phase and consequently the fields are not linearly polarized.
For the fields the -component leads the -component by and as the -component lags the -
component by . At some instant when the -componentis at ats positive peak, the -component and
the -component are zero. At that instant we have MATH
and MATH
Now for a uniform plane wave

The velocity of the wave

Now we know that MATH

frequency of the wave

The phase velocity along the -direction

and the phase velocity along -direction

and the phase velocity along -direction

To find the state of polarization of the wave, let us trace the tip of the electric field vector. The three
components of the electric fields are

At , , ,

MATH
At , ,MATH ,MATH MATH

At , , , MATH
Since, the is constant as a function time, it is the circular polarization. To get the sense of rotation
draw the fields at different times as shown in Fig. E5.2. The vector rotation is towards the left hand if we
face in the direction of wave. The wave polarization therefore is LHC.
PLANE WAVE AT DIELECTRIC INTERFACE
Up till now we discussed the propagation of a plane wave in an unbound medium. Let us now consider
the propagation of a plane wave across a dielectric interface. Let the space be divided into two semi-
infinite regions with different medium parameters like the permeability and permittivity. Let us also
assume that both the regions are loss-less i.e., the conductivity for both regions is zero. Without losing
generality let us orient the co-ordinate system such that the dielectric interface is along the -plane
passing through as shown in Figure 5.5.
\begin{figure}[t]
\includegraphics[width=0.6\textwidth]{Nfig5.5a.eps}
\caption{Electromagnetic wave at dielectric interface} \end{figure}
Let us denote the region on the left of the interface ( ) as region 1, and let all its parameters be
denoted by suffix 1. Similarly let the region on the right of the interface ( ) be denoted as region 2 and
let its parameters be denoted by suffix 2. Then , represent permeability and permittivity of medium 1
respectively and , represent corresponding parameters for medium 2.
Let us now consider a plane wave incident on the interface from medium 1 side. For simplicity let us
assume that the wave vector lies in the -plane and makes an angle with the normal to the interface
( in this case -direction). The angle is called the angle of incidence. The plane which contains the
wave vector and the normal to the interface is called the plane of incidence. The angles which the wave
vector makes with the three axes are therefore

We can then write the field (electric or magnetic) for this wave as

The suffix indicates the incident field, is a constant vector and is the phase constant of the wave
in medium 1, MATH .
Substituting for , , into eqn (5.30) we get

From eqn (5.31) it is clear that the field is independent of .

At the interface i.e., at , the phase is constant along -direction and it linearly increases along the -
direction. The field magnitude in the -plane therefore is

The field variation is like a sinusoidally corrugated sheet with corrugations oriented along the -axis as
shown in Figure 5.6.
\begin{figure}[h]
\includegraphics[width=0.5\textwidth]{Nfig5.6a.eps}
\caption{Field variation} \end{figure}
When the sinusoidally corrugated field is incident at the interface a field with similar corrugated variation is
induced on the other side of the interface. This is due to the boundary conditions which require continuity
of the fields at the dielectric boundary. One can additionally show that both magnetic field and electric
field in general, can not satisfy the boundary conditions without modifying the incident fields. In other
words we can say that when a plane wave is incident on an interface, the fields with similar phase
variations are induced on both side of the interface. In region 1 then the total field is the combination of
the incident field and the induced field, where as in region 2 there is only the induced field. Since the
induced fields are also time varying, they also constitute waves. The induced fields in region 1 then form a
wave moving away from the interface i.e., back in region 1, whereas the induced fields in region 2 form a
wave going away from the interface in region 2 as shown in Figure 5.7. Since both induced waves see
infinite medium ahead of them they are in the form of plane waves. These waves are called the reflected
and transmitted (or refracted) waves respectively. We can now note that since the phase variation for the
induced fields is same as that of the incident field at the interface (constant along -direction), the wave
vectors of the reflected and transmitted waves also lie in the plane containing the incident wave vector
and the interface normal i.e., the plane of incidence. One can then conclude that the wave vectors for the
incident, reflected and refracted (transmitted) waves, and the normal to the interface lie in the same
plane. In bulk optics the wave vectors are called the light rays. The reader would then note that the above
conclusion is nothing but the first law of reflection of light which states that `the reflected and refracted
rays lie in the plane of incidence'.
\begin{figure}[h]
\includegraphics[width=0.80\textwidth]{5.7.eps}
\caption{Incident, reflected and refracted waves} \end{figure}
We have seen above that the incident, reflected and refracted waves travel in the same plane, the plane
of incidence. We however do not know in what directions they would be travelling. Let us say that the
reflected wave travels back in medium 1 at an angle with respect to the interface normal and the
refracted (transmitted) wave travels in the medium 2 at an angle with respect to the interface normal,
as shown in Figure 5.8.
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{Nfig5.8.eps}
\caption{Incident, transmitted and reflected waves} \end{figure}
For reflected wave MATH , and MATH , and the field for the reflected wave can be
written as

Similarly for the transmitted wave MATH , , and and its field can be written as

where MATH is the phase constant in medium 2. At the interface i.e., at the fields must
satisfy the boundary conditions. If fields , , represent electric fields, their tangential components
must be continuous across the interface. On the other hand, if we take , , as magnetic fields,
generally their normal components must be continuous. However in the present case, their tangential
components also must be continuous at the interface since for ideal dielectrics there is no surface current.
So without worrying about whether represents or , we can say in general that their tangential
components must be continuous at the interface. We therefore have

Suffix ' ' represents the tangential component of the vector.

Equation (5.35) is true for every point on the interface i.e., for every value of and . This can happen
only if the phases of the three terms are same i.e.,

and

From eqn (5.37) we get the second law of reflection that is, `the angle of reflection is equal to the angle of
incidence'.
For ideal dielectrics MATH (free space permeability), MATH and MATH where and
are the dielectric constants of the two media. Noting from eqn (4.140) that MATH (refractive index of
medium 1) and MATH (refractive index of medium 2 ), eqn (5.39) can be written as

Equation (5.40) is the well known Snell's law of refraction. One should however note that (5.40) is a
special case of (5.39) which represents the generalized Snell's law for any loss-less medium.
Once the phases of the three terms in eqn (5.35) are equated the equation reduces to

That is, the problem of wave propagation across an interface reduces to finding the fields of the three
waves at the interface using eqn (5.41). From the knowledge of , and we can write the fields
at any point in the space. If the point is in region 1 the field is a superposition of and whereas if the
point is in region 2 the field is .

EXAMPLE 5.3 : A light beam is incident from air to a medium with a dielectric constant and relative
permeability . If the angle of incidence is . Find the angle of reflection and angle of refraction.

Solution :

The angle of reflection = angle of incidence =

From the Snell's law,

MATH

REFLECTION AND REFRACTION OF WAVES AT DIELECTRIC

INTERFACE
Uptill now we had a general discussion on the propagation of a plane wave across a loss-less dielectric
interface. Let us now analyse the specific cases of field orientation with respect to the plane of incidence.
As can be seen from Figure 5.9, (and consequently ) vector can make an arbitrary angle with respect
to the plane of incidence.
\begin{figure}[h]
\includegraphics[width=0.55\textwidth]{Nfig5.9.eps}
\caption{field in arbitrary direction} \end{figure}
Geometrically it is quite complicated to handle the fields with arbitrary orientation. Instead one can
elegantly handle the case by decomposing the field into its components, one in the plane of incidence
and other perpendicular to it and then combining the results for the two components. In the following,
therefore we discuss the two cases (i) Electric field in the plane of incidence, also called parallel
polarization (ii) Electric field normal to the plane of incidence also called perpendicular polarization .
Specifically we are interested in finding out the field amplitudes of the transmitted and the reflected waves
for a given incident wave. Normally we carry out the analysis for the E field and as discussed earlier we
can find the magnetic field as and when needed from the property of the uniform plane wave. We
therefore define two parameters of interest namely

and find their values for the parallel and perpendicular polarizations.
Reflection and Refraction with Perpendicular Polarization
Let us consider a plane wave with perpendicular ( ) polarization incident at a dielectric interface as shown
in Figure 5.10. The angle of incidence is say (angle which the wave vector makes with the interface
normal) and the plane of incidence is the plane of the paper. Without losing generality let us assume that
the incident electric field is pointing out of the plane of the paper (oriented in -direction). Then from
the property of the plane wave the magnetic field will be perpendicular to and will lie in the plane of
the paper perpendicular to the wave vector as shown in Figure 5.10. The therefore lies in the -plane.
Also the ratio of the magnitudes of and is , the intrinsic impedance of medium 1.
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{Nfig5.10.eps}
\caption{Perpendicular polarization} \end{figure}
As argued above, the wave vectors of reflected and transmitted waves also lie in the plane of incidence
(in this case the plane of the paper), and the angle of reflection is same as the angle of incidence. One
may further argue that since the tangential component of should be continuous at the interface and
since the incident wave has only -directed electric field, the reflected and transmitted wave also must
have -directed field only. Of course the fields might point into the plane of paper or point out of it.
However the direction reversal can easily be accommodated by assigning a +ve or -ve sign to a field.
Therefore without losing generality we assume that both fields and are along direction i.e.,
pointing outwards from the plane of the paper. If any of them or both of them were directed in the
opposite direction (along - -direction) their signs will come negative automatically.
Now for each wave, the Poynting vector should give the direction of the wave vector. Since the
directions of the incident, reflected and transmitted waves are known, the directions of the magnetic field
for the three waves , and can be easily obtained as shown in Figure 5.10. Note that the
incident and transmitted waves are travelling from left to right upwards and consequently the magnetic
fields point downwards. Whereas, the reflected wave is travelling from right to left upwards and hence its
magnetic field is oriented upwards.
We can write the electric fields for the three waves as (see eqns (5.31), (5.33), (5.34))

where MATH , MATH and , , and are the vector amplitudes of the three waves.
For all the fields, a time variation of is implicitly assumed. Since and are related through intrinsic
impedance of the medium we can write

where MATH and MATH are the intrinsic impedances of medium 1 and medium 2
respectively.
Now applying the boundary conditions i.e., tangential component of the electric field is continuous at the
interface, and since there is no surface current even the tangential component of magnetic field is
continuous at the interface, we get at

Note that the fields are in -direction and hence are tangential to the interface whereas for the magnetic
field we have to take the component along the interface. Substituting eqns (5.43)-(5.45) into (5.49)-(5.50)
we get

Solving eqns (5.51) and (5.52) we can obtain the reflection and transmission coefficients as

From eqns (5.53), (5.54) or dividing eqn (5.51) by we get relation between the transmission and the
reflection coefficients as

From (5.53) and (5.54) we can make following observations:

(1) The reflection and transmission coefficients are real. That means there is no arbitrary phase change in
the reflected or transmitted wave at the interface. Depending upon the sign of and , the phase
change could be either 0 or ( 0 for +ve sign and for negative sign).

(2) Magnitude of Reflection coefficient is always less than unity whereas the magnitude of the
transmission coefficient could be greater or less than unity. That means the amplitude of the transmitted
field could be greater than the amplitude of the incident electric field. Obviously this would happen
when is positive ( see eqn (5.56)), i.e., when

From Snell's law since MATH we get

Substitution of in eqn (5.57) yields

Therefore if eqn (5.61) is satisfied the transmitted electric field is greater than the incident electric field.
It should be noted here that higher transmitted electric field does not mean higher transmitted power. The
magnetic field reduces in the medium 2 appropriately to give the transmitted power less than or equal to
the incident power. After all there has to be conservation of power making sum of the transmitted and the
reflected power equal to the incident power i.e.,

EXAMPLE 5.4 : A uniform plane wave having power density is incident from air at air-dielectric
interface at the angle of incidence . The electric field vector for the wave lies perpendicular to the
plane of incidence. Find the power density of the transmitted and the reflected wave. The relative
permittivity of the dielectric medium is 25. If the frequency of the wave is MHz, find the amplitude of
the electric and magnetic field at a distance of m on either side of the interface.

Solution:

The wavelength in air

In air the power density

The transmission angle is

From (5.53) and (5.54), we have,

Following the coordinate system in Fig. 5.11, we have

MATH
Power density of the transmitted wave
Power density of the reflected wave = MATH
In region , the electric field is (at z = 1 m)

In region , the electric field is

Similarly, the magnetic fields for the three waves are(see eqns (5.46), (5.47) and (5.48)).

Hence total magnetic field magnitude in x direction in region 1 can be written as

-ve sign shows reversal of direction.

Similarly, total field in z direction in region 1 can be written as

Hence peak amplitude of magnetic field in region 1 is

Total magnetic field in region 2 is same as i.e. MATH A/m

Reflection and Refraction with Parallel polarization
On the lines similar to that for the perpendicular polarization, we can now analyze reflection and refraction
of a plane wave having parallel polarization. In this case the E-vector lies in the plane of incidence. Again
assuming that plane of incidence is same as the plane of the paper, both wave vector and the vector lie
in the plane of the paper. Since the magnetic field is perpendicular to both, it is now perpendicular to the
plane of the paper. Without losing generality therefore let us assume that the magnetic field is pointing
out of the paper (oriented along +y direction) ( see Fig. 5.11).
Using similar arguments as in previous case, we can assume that the magnetic fields for the reflected and
transmitted waves also are oriented along +y direction. Using the Poynting vector argument we can then
find the directions of the respective electric fields. The electric fields for the incident, reflected and
transmitted waves have exactly same form as given by (5.43)-(5.45) except that the vector , and
are no more oriented along the direction but lie in the -plane. The amplitudes of the magnetic
fields for the three waves are also given by (5.46)-(5.48), but their direction is perpendicular to the plane
of incidence ( direction).
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{5.11.eps}
\caption{Parallel polarization} \end{figure}
Applying continuity of the tangential components of the electric and magnetic fields at the interface we get

Again using (5.45)-(5.48) and solving (5.63) and (5.64) we get the reflection and the transmission
coefficients for the parallel polarization as
Reflection coefficient:

Transmission coefficient:

From (5.63) we note that

In the case of parallel polarization also we can make similar observations as that in the previous case i.e.,
(1) The magnitude of the reflection coefficient is always less than unity and there is a phase difference
of 0 or between the incident and the reflected wave.
(2) The transmission coefficient could be greater or less than unity i.e., the transmitted electric field
could be greater or less than the incident electric field.
Reflection and Refraction - Normal Incidence
The reflection and refraction for the normal incidence i.e., incidence along the normal to the interface can
be obtained by making the angle of incidence zero either for perpendicular or for parallel polarization.
Figure 5.12 shows the incident, reflected and transmitted waves with their respective fields for in
the two polarizations. \begin{figure}[h]
\includegraphics[width=1.2\textwidth]{Nfig5.12.eps}
\caption{Normal incidence} \end{figure}
Substituting and consequently due to the Snell' law, in (5.53) and (5.65) we get

Since for , the and polarizations represent same case one would wonder why the two reflection
coefficients are negative of each other! The reason for this lies in the original directions assumed for the
E-fields. In perpendicular polarization case as , and point in the same direction whereas for
parallel polarization when , and point in opposite direction (Figure 5.12(b)). It is then clear
that for normal incidence if we assume the and pointing in the same direction the reflection
coefficient is same as that given by eqn (5.68). We therefore have for normal incidence,

The case of normal incidence can also be analysed in a different way. For normal incidence since
MATH the electric fields for the three waves can be written as (put MATH in eqns (5.43)-(5.45))

Equations (5.72)-(5.77) are identical to the transmission line equations with E replacing V, H replacing I
and and replaced by the characteristic impedances respectively. One can therefore conveniently
use the normal incidence case using transmission line concept.
\begin{figure}[h]
 \includegraphics[width=1.2\textwidth]{Nfig5.13.eps}
\caption{Analogy between normal incidence case and the transmission line} \end{figure}
Figure 5.13 shows the equivalence of the two cases. Since line 2 is infinitely long towards the right, at the
junction it appears like an impedance ( input impedance of an infinitely long line is always the
characteristic impedance). The voltage reflection coefficient on line 1 then is

which is same as eqn (5.68) and the transmission coefficient is

The impact of transmission line analogy can not be realized until we analyse the propagation of a plane
wave through a multi-layer medium. In the following we analyze normal incidence of a plane wave in a
layered medium using wave as well as transmission line model. To develop the concepts clearly, we take
only a three media problem here and leave the multi-media problem to the interested readers.

EXAMPLE 5.5 : A uniform plane wave with V/m electric field is normally incident on an infinitely thick
slab of a material of dielectric constant . Find the electric and magnetic field just inside the slab surface.
How much power penetrates the material slab ?

Solution:
The wave is incident from air to the material. Hence, we have

The transmission coefficient

The electric field inside the slab MATH

The magnetic field inside the slab

The power transferred to the slab is =1.416 .

NORMAL INCIDENCE ON A LAYERED MEDIUM
Let us consider a layered three media geometry as shown in Figure 5.14. Media 1 and 2 are semi-infinite
in size whereas media 2 is of finite thickness . The parameters for media 1, 2 and 3 are denoted by
suffices respectively. Let us consider a plane wave with electric field normally incident on the
interface (1/2) from medium 1. Since the electric and magnetic fields are oriented along the co-ordinate
axes, without writing explicitely we will assume that all vectors are along direction and vectors are
along -directions depending upon the direction of the wave vector. If the wave is moving along
direction the H will be along direction and if wave is moving along direction the H vector will be in
direction. (Note that these signs are chosen to give correct direction of the Poynting vector for each
wave).
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{Nfig5.14.eps}
\caption{Layered medium} \end{figure}
In the following analysis we will drop the vector notation for the writing simplicity. The incident electric and
magnetic fields therefore can be written as

Our primary interest now is to find

(i) Fields for reflected wave in region 1.

(ii) Fields for transmitted wave in region 3.
As mentioned above the problem can be solved in two ways (i) by multiple reflection and transmission at
the two interfaces (1/2) and (2/3) (ii) by transmission line analogy.
Multiple Reflection/Transmission of waves
Let us first create a notation for the transmission and reflection coefficients at the interfaces in forward as
well as backward directions. Let represent the reflection coefficient at interface for a wave incident
from medium . Similarly let represent the transmission coefficient from medium to medium . For
example will represent reflection coefficient at interface (1/2) for a wave incident from medium 1 and
will represent transmission coefficient from medium 3 to medium 2 for a wave travelling backwards in
medium 3.
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{5.15.eps}
\caption{Multiple reflection} \end{figure}
From (5.78) and (5.79) we then have

and

From (5.82) it is clear that MATH . Let now a wave with electric field be normally incident on the
interface (1-2) from medium 1 side. At the interface (1/2) a part of the energy is reflected and a part is
transmitted. The amplitude of the reflected field is and the amplitude of the transmitted field is
. The transmitted field travels a distance in medium 2 and therefore undergoes a phase change of
. The phase shifted wave MATH gets partly reflected and partly transmitted at the interface
(2/3). The reflected wave at (2/3) interface travels backwards upto interface (1/2) and further undergoes a
phase change of . At the interface (1/2) a part of the energy is transmitted backward in medium 1
and a part is reflected back in medium 2. This process continues for infinite reflections and transmissions.
In steady state then we have a reflected wave in medium 1 which is a superposition of first reflected wave
and subsequent backward transmitted waves. In medium 2 we see two sets of waves one travelling
from left to right and other from right to left. The total field at any point is equal to the sum of all the
forward and backward travelling waves. In region 3 we have superposition of transmitted waves travelling
from left to right.
From Fig. 5.15 we can write reflected field in medium 1 as

i.e.,

The terms in {} form a geometric series with progression ratio

MATH . Since and are less than 1, and
MATH the series is convergent and we can write its sum. The reflected field and the reflection
coefficient therefore can be written as

giving

Similarly, from Figure 5.15 we can write the transmitted wave in medium 3 as

Again the terms in curly brackets form the same geometric series as that in (5.85), and we get the
transmission coefficient as

The procedure explained above can be extended to any number of layers with of course increased
algebraic complexity.
Transmission Line Analogy
Let us now try to solve the problem of wave propagation in a multi-layer medium using the transmission
line analogy. The geometry in Figure 5.15 is equivalent to three transmission lines of characteristic
impedances MATH respectively.
\begin{figure}[h]
\includegraphics[width=0.9\textwidth]{Nfig5.16.eps}
\caption{Transmission line analogy} \end{figure}
The transmission lines 1 and 3 are infinitely long and the transmission line 2 is of length . Also note that
the phase constants of the three lines are MATH respectively as shown in Figure 5.16.
Now, since line 3 is infinitely long, it sees an impedance (characteristic impedance of line 3 ) at the
junction (2/3). This impedance can be transformed to an impedance (say) Z at (1/2) junction as

Note that the characteristic impedance of the transforming line 2 is .

The reflection coefficient on the line 1 then can be written as

Equations (5.87) and (5.92) although look algebraically different, are infact the same. It is only a matter of
algebraic manipulation to reduce one to the other.
The magnitude of the transmission coefficient can be obtained by applying conservation of power.
The electric field in medium 1 has a power density of . The power density of the reflected wave
will be MATH . Since all the media are lossless, the difference of the two power densities is equal to the
power density of the transmitted wave in medium 3, , giving

The above analysis clearly shows that the problem of normal incidence on a multi-layer medium can be
elegently solved using the transmission line concepts.
The analysis of layered medium finds many practical applications. For example, we may be interested in
sending electromagnetic energy efficiently from one medium to another, or we may be interested in
providing a protective sheet of a material around a radiating instrument. In both type of applications,
ofcourse the primary objective is the efficient power transfer with minimal reflections. One can then ask,
under what condition the reflected power reduces to zero i.e., the reflection coefficient goes to zero ?
From transmission line view point the answer to this question is rather simple. `Use a quarter wavelength
transformer to match two transmission lines'. We can get the same result by equating the reflection
coefficient in eqn (5.92) to zero, giving

Separation of real and imaginary parts yields

Since, both and can not simultaneously become zero for any angle, we have two conditions under
which there is no reflection.
(1) and

This implies

is the wavelength in medium 2. (MATH )

This condition suggests that if a slab of any material which is multiples of half wavelength thick is inserted
normally in the path of an electromagnetic wave, there is no reflection i.e., the propagation on either side
of the slab remains unaffected. This is interesting and useful, as it suggests that if an antenna is covered
with a half wavelength thick sheet, the flow of the radiated wave is unaffected. This is the principle of
radom design. Radom is a cover which protects an antenna from the weather. Obviously, since the
purpose of radom is mechanical, it should affect the wave propagation in the least possible fashion.
Multiple dielectric layers can be used for designing proper radoms.
(2) and MATH .
This is the famous quarter wavelength transformer. The conditions imply

This condition suggests that an electromagnetic wave can be fully transmitted across a media-interface
without reflection by inserting a sheet of matching medium which is quarter wavelength thick and has
intrinsic impedance equal to the geometric mean of the intrinsic impedances of the two media. This
technique is frequently used in realizing anti-reflecting coatings in optical components.
EXAMPLE 5.6 : A Uniform plane wave having power density is normally incident on a cm thick
dielectric sheet with = . If the frequency of the wave is GHz. Find the power density of the wave
transmitted through the sheet.

Solution:

Intrinsic impedance of the dielectric is

The transmission coefficient from eqn (5.89) is

Power density of the transmitted wave

TOTAL INTERNAL REFLECTION

Let us now investigate an interesting case of the wave propagation known as the `total internal reflection'.
From eqns. (5.38) and (5.58) we note that if

the angle of transmission does not exist, since and is imaginary. This means that if a
plane wave is launched at a media interface at an angle which satisfies eqn (5.102) there is no
transmitted wave. The angle for which the equality in eqn (5.102) holds is called the critical angle, .
The critical angle therefore is given by

For a wave launched at the critical angle, we get meaning that the transmitted wave travels along
the interface.
For , the reflection coefficients for the two polarizations (prependicular and parallel ) are re-written
such that the quantity in the square root sign is positive, giving

where we have written explicitly in terms of , and have taken common within the square root
sign. It is now easy to note that both reflection coefficients are of the form

and therefore their magnitudes are unity. That is, for MATH and . The reflection
coefficients however have different phase angles. The phase angles for the two polarizations are

Since the magnitudes of the reflection coefficients are unity, the entire incident power is reflected back
into medium 1. This phenomenon therefore is called the Total Internal Reflection (TIR).
There are certain important things which are worth mentioning about the total internal reflection.

(i) The total internal reflection can take place only if the wave is travelling from denser to rarer
medium.
From eqn (5.102) we can see that the angle of incidence for which the total internal reflection can occur is
given by

Since has to be , for TIR to take place we must have

i.e., medium 2 is rarer compared to medium 1.
A special case of this is an ideal dielectric interface, for which MATH , and

and are refractive indices of the two dielectric media. Condition (5.108) then reduces to the well
known relation

(ii) Wave undergoes a phase change during total internal reflection

From (5.105) and (5.106) it is clear that the reflection coefficient has a phase which implies that the
reflected wave lags with respect to the incident wave. The phase change depends upon the medium
parameters as well as the angle of incidence. The angle at which the equality holds in eqn (5.102) is
called the critical angle . This is the angle of incidence (MATH ) at which

meaning that the transmitted wave travels just along the interface.
At critical angle the phase lag in the reflected wave is zero (quantity in the square root sign goes to zero)
and it increases as the angle of incidence increases beyond . It is also important to note that the two
polarizations, parallel and perpendicular undergo different phase changes at toal internal reflection. As
will be discussed later, this differential phase change for the two polarizations can be exploited to alter the
state of polarization of an electromagnetic wave.
(iii) At TIR the fields do not vanish in the second medium.
Substituting (5.58) with TIR condition (5.102) in (5.54) and (5.66) we find that even when and
becomes unity, the transmission coefficients and do not become zero. In other words, total
reflection of power in medium 1 i.e., no transmission of power to medium 2, does not mean no
transmission of the fields to medium 2. The fields can very well exist in the second medium provided they
do not constitute any power flow. These fields are called the 'evanescent fields'. Distribution of these
fields in medium 2 can be readily obtained by substituting for in the expression for the transmitted field
(5.45). Substituting

and

in eqn (5.45) we get

In eqn (5.114), the exponential term I represents a travelling wave in x-direction. Term II however
represents only an exponentially varying field in the -direction and not a wave since there is no phase
variation along -direction. Since the field can not grow indefinitely away from the interface, we should
choose the -ve sign for the exponent which represents exponentially decaying fields away from the
interface in medium 2. The field in medium 2 therefore is correctly written as

Note that the exponential decay constant in -direction is a function of the angle of incidence besides the
media parameters. Figure 5.17 shows the field variation as function of distance from the interface for
different angles of incidence.
\begin{figure}[h]
\includegraphics[width=1.3\textwidth]{Nfig5.17a.eps}
\caption{Electric field variation} \end{figure}
At the critical angle, the decay constant is zero and the field is constant along the -direction in medium 2.
As the angle of incidence is increased beyond the critical angle, the field gets more confined to the
interface. However, it is important to note that in no circumstance the field in medium 2 goes to zero. The
boundary conditions demand that the fields must be continuous at the interface. Since the fields in
medium 1 at the interface are non-zero, the fields in medium 2 are also finite at the interface. For TIR,
where the energy is totally reflected in medium 1, one should not undermine the importance of the
evanescent fields in medium 2. Though these fields do not constitute any power flow, there presence is
as important as the fields in medium 1 which have power flow associated with them. Any disturbance to
the evanescent fields eventually disturbs the total internal reflection.
EXAMPLE 5.7 : A uniform plane wave travelling in a dielectric of refractive index 3 is incident at the
dielectric air interface. The angle of incidence is . The electric field of incident wave is V/m. Find
the phase velocity of the field and the magnitude of the field at a distance of cm on either side of the
dielectric air interface. Assume perpendicular polarization for the wave. Frequency of wave is GHz.

SOLUTION :

The critical angle at the interface is

Since the angle of incidence is , the wave is total internally reflected at the interface. In dielectric
medium then we have superposition of the incident and the reflected wave and in air we have
exponentially decaying fields.
From eqn (5.104) reflection coefficient can be written as

Hence total electric field in dielectric medium is given by

In the air, electrical field decays as we will go away from the interface. In air, is along the x-axis. New
will be given by

Electic field in air is given by (5.115)

WAVE POLARIZATION AT MEDIA INTERFACE

In section 4.5, we have shown that any arbitary polarization can be decomposed into any two orthogonal
states of polarization. We can therefore analyse the polarization of the reflected and the transmitted
waves at an interface by decomposing the incident electric field into its components, one in the plane of
incidence and other perpendicular to it. In other words, we decompose the polarization of the incident
wave into two orthogonal linear polarizations, one in the plane of incidence and other normal to it, find
reflection and transmission for the two polarizations separately and combine them. Let us therefore write
the electric field as

Note that locally the coordinate system is oriented such that is along -axis and is along -axis
and is the direction of the wave vector (see Fig. 5.18).
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{5.18.eps}
\caption{Polarization of a plane wave at medium interface} \end{figure}
If the reflection and transmission coefficients for parallel and perpendicular polarizations are denoted by
and respectively, the reflected and transmitted electric fields are

since the reflection and transmission coefficients are different for parallel and perpendicular polarizations,
we can say that in general the states of polarization of the reflected and transmitted waves are not same
as that of the incident wave. However, it is worthwhile to investigate here a few simple but important
cases.
Change in Polarization at Simple Reflection
(i) Linearly Polarized Incident Wave

If the incident wave is linearly polarized, and are in phase i.e., . For simple reflection ( not
total internal reflection) since the reflection coefficients and are real ( though they could be positive
or negative), the components of the reflected wave, and are either in phase or out of phase
(depending upon sign of and ). The polarization of the reflected wave hence remains linear. The
orientation of the linear polarization however will be different since

The argument is equally applicable to the transmitted wave as well since and are also real in this
case.
We therefore conclude that a linearly polarized wave remains linearly polarized after simple reflection but
the plane of polarization changes as shown in Figure 5.19.

\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{5.19.eps}
\caption{Linearly polarized incident wave} \end{figure}
(ii) Circularly Polarized Incident wave

For a circularly polarized incident wave, we have

Since the reflection and transmission coefficients are real, the reflected (and also transmitted) wave has a
phase difference of between their components. However, their two components no more have same
magnitudes, since MATH and MATH . The reflected and transmitted waves therefore become
elliptically polarized with the major axis of each ellipse either lying in the plane of incidence or
perpendicular to it as shown in Figure 5.20.
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{5.20.eps}
\caption{Circularly polarized wave incident at a dielectric boundary} \end{figure}
Depending upon the magnitudes of the reflection coefficients, the polarization ellipse may degenerate into
a line. A circular polarization therefore may change even to the linear polarization after reflection. This
aspect will be discussed later in section 5.8.
Polarization Change at TIR
For the total internal reflection, there is no transmitted wave, so we need to discuss the polarization of the
reflected wave only. For the TIR the magnitudes of the two reflection coefficients and , are unity but
their phases are different. Consequently, the phase difference between the two reflected components is
not same as that between the incident components. We can then only say that the state of polarization
always changes at the total internal reflection, except when the incident polarization is either purely
parallel or perpendicular. We can also note that the change in polarization depends upon the angle of
incidence. The TIR can therefore be cleverly exploited to change the state of polarization of a wave
without affecting its magnitude.

EXAMPLE 5.8: A circularly polarized plane wave is incident on a thick glass slab at an angle of incidence
of . Find the state of polarization of the reflected and transmitted waves. The frequency of the wave is
Hz and the refractive index of glass is 1.5.
Solution :
Since, the incident wave is circularly polarized, the parallel and perpendicular components of the electric
field are equal in magnitude but out of phase. Let the perpendicular component be leading the
parallel component, that is

The reflection and transmission coefficients for parallel and perpendicular polarizations are (from eqns
(5.75), (5.76), (5.88), (5.89))

The reflected and transmitted fields are

The reflected wave is elliptically polarized and the transmitted wave is almost circularly polarized.

BREWSTER ANGLE
The Brewster angle is the angle of incidence for which there is no reflection, i.e., the reflection coefficient
is zero. Equating (5.53) and (5.65) to zero we get

where and are the Brewster angles for perpendicular and parallel polarizations respectively.
Substitution for from (5.58) in (5.119) and (5.120) yields

After a little algebraic manipulation we get

Substituting MATH , MATH , MATH , and MATH , the Brewster angles for the two
polarizations can be obtained as

In principle, the Brewster angle can exist for both polarizations. However, for non-magnetic materials
(dielectrics), MATH (free-space permeability) and consequently the Brewster angle does not exist
for the perpendicular polarization. This can be seen by substituting MATH in (5.125). The quantity
in the square bracket becomes negative and hence there is no real angle .
The Brewster angle for the parallel polarization exists and is given by,

Note: From (5.127) it might appear that when MATH i.e., when there is no medium discontinuity at the
interface (in fact in that case there is no interface as there is only one medium), the Brewster angle is
. But this is absurd because if there is no medium discontinuity, no-reflection condition should be true
for all angles and not only for . This absurdity however is purely mathematical. If MATH , the quantity
in square brackets of eqn (5.126) is i.e., is indeterminate and not .
From the above discussion it is clear that at Brewster angle, a particular polarization has no reflection.
Now if the incident wave has some arbitrary polarization ( which is a combination of parallel and
perpendicular polarizations), after reflection at the Brewster angle the wave will have only one polarization
component. If the angle of incidence is , the reflected wave has only perpendicular polarization, and if
the angle of incidence is , the reflected wave has only parallel polarization. In both the cases, the
polarization of the reflected wave is linear ( in the plane of incidence or perpendicular to it) irrespective of
the polarization of the incident wave. The Brewster angle therefore is also called the `Polarizing angle'.
This principle is used to obtain linearly polarized light from an unpolarized one. In many applications, like
optical interferometry, lasers etc where the light has to be linearly polarized, the Brester angle concept
can be conviniently used.
EXAMPLE 5.9 : An elliptically polarized wave is incident on an air dielectric interface at the Brewster
angle. The axial ratio of the polarization ellipse is 2 and the major axis of ellipse is perpendicular to the
plane of incidence. Find the axial ratio and the orientation of the major axis of the reflected and
transmitted waves.

Solution :
Since, the interface is dielectric interface, at Brewster angle, parallel polarization is completely
transmitted. The reflected wave has only perpendicular polarization. The reflected wave therefore has
linear polarization perpendicular to the plane of incidence.
For transmitted wave, we have to obtain the field components.
The angel of incidence

Using Snell's Law we get

For the ellipse since the AR is 2 we have,

the phase difference between and is , since, the major axis is along the normal to the plane of
incidence. Substituting and in expression of transmission coefficients we get,

The transmitted fields therefore are,

The phase between and is still as was in the original wave. The transmitted wave therefore
has same sense of rotation (LH or RH) as the incident wave. The ratio of the amplitudes of two
components is

Since, (because if , there is no interface) MATH and consequently the transmitted wave
also is elliptically polarized. If MATH i.e , if , the major axis of the ellipse will be to the
plane of incidence and AR will be .

However, if MATH i.e., if , the major axis of the ellipse will be in the plane of incidence
and AR will be .
For any medium is always greater than 1 and consequently always i.e., we have,
MATH . The transmitted wave therefore is elliptically polarized with major axis in the plane of
incidence.
FIELDS AND POWER FLOW AT MEDIA INTERFACE
From the discussions of the reflection and refraction of the plane waves at a media interface, it is clear
that in medium 1, we have superposition of the incident and the reflected waves, and in medium 2 we
have only the transmitted wave. It is then worthwhile to ask a question - what is the distribution of the
fields in the two media, and what is the direction of the net power flow. Let us investigate the fields in the
two media for a wave with perpendicular polarization. In the two media we have,

Substituting for and from (5.43)-(5.45), we get

For perpendicular polarization MATH and are in the same direction.

Substituting for and from eqns (5.53), (5.54) we can get the fields in the two media as

Medium 1
The expression inside the square bracket can be written in polar form ( magnitude and phase form) giving

where,

and
The amplitude of the total field in medium 1 is

and the phase of the total field, including the time phase, is

From eqn (5.138) we can note that the field amplitude is a function of now. Since MATH
varies between -1 and +1, the field amplitude varies from MATH to MATH . This is a partial
standing wave created by superposition of the incident and the reflected waves. A plot of the field
amplitude appears like a corrugated surface with corrugations running parallel to the interface as shown
in Figure 5.21.
\begin{figure}[h]
\includegraphics[width=0.8\textwidth]{5.21.eps}
\caption{Electric field amplitude} \end{figure}
The field amplitude is constant in a plane parallel to the interface ( ), but it varies from plane to
plane. For example, plane corresponds to the minimum field whereas corresponds to the
maximum field in Figure 5.21. We can then call these planes as `constant amplitude planes'.
The distance between the two adjacent constant amplitude planes, one passing through the maximum
amplitude points and other passing through the minimum amplitude points is given by

where is the wavelength in medium 1. Note that when , whereas when MATH ,
.
Like the constant amplitude planes, the constant phase surfaces are not planes. Depending upon the
angle of incidence, reflection coefficient etc, we may get different constant phase surfaces.
The phase constant of the composit wave in -direction is MATH and hence the phase velocity in -
direction is

In the -direction however, the phase constant cannot be written as explicitly as that in the -direction. We
therefore have to obtain the phase velocity in -direction from the first principles. Making the phase
stationary along -direction (and also making constant) we get

From eqns (5.139) and (5.142) two things are worth making note of:
(1) The constant phase surfaces are no more planes.
(2) The phase velocity in -direction is not constant but is a function of . Consequently, the total phase
velocity becomes a function of space.
\begin{figure}[h]
\includegraphics[width=1.1\textwidth]{5.22.eps}
\caption{Variation of phase} \end{figure}
\begin{figure}[h]
\includegraphics[width=0.5\textwidth]{5.22b.eps}
\caption{Variation of as a function of } \end{figure}
Figure 5.22 shows the phase as a function of space and Figure 5.23 shows the variation of the phase
velocity as function of space.

Medium 2
In medium 2 the fields are rather simple as there is only one travelling wave. The amplitude of the field is

and the phase of the wave (including the time phase) is

The amplitude of the field is constant every where in medium 2 ( see Figure 5.21) and the phase varies
linearly with and . Consequently, the phase velocities in and directions
are constant in space.
It should be clear that although the above analysis has been carried out for the perpendicular polarization,
the discussion is valid for any polarization. The things to note are that due to interference of the incident
and the reflected waves, we do not have constant field strength in medium 1. For a field receiver, like a
radio or a mobile phone, moving in medium 1, the receiver output therefore varies since it encounters
stronger field at some place and weaker field at other. This phenomenon is called `signal fading'. For a
single media interface, as taken here, the field pattern is quite regular and predictable. In an environment
where there are many reflecting boundaries, the field patterns are quite complex and the fading becomes
more or less random. In modern communication systems, the fading is one of the prime issues. This
aspect will be discussed in detail in the chapter on `propagation of electromagnetic waves'.
LOSSY MEDIA INTERFACE
As discussed earlier, if the conductivity of a dielectric medium is non-zero, the medium becomes a lossy
medium. The propagation constant and the intrinsic impedance of a lossy medium are

If we replace by complex , in the formulation carried out in the previous section, we get the general
formulation for any lossy media interface. The formulation is valid for any arbitrary but finite value of
conductivity. For (ideal conductor) however we have to reformulate the problem. This is due to the
fact that for an ideal conductor, one has to include surface current in the boundary conditions. The
continuity of the tangential component of the magnetic field is not valid for an ideal conductor as we have
taken in our boundary conditions. Therefore the case of ideal conducting boundary will be analyzed
separately.
Although the extension of results from a loss-less media to lossy media appears straight forward, one
may encounter certain conceptual difficulties regarding origin of the incident wave in medium 1. In the
loss-less case, we never asked, where the incident wave was originated and how much distance it had
travelled before reaching the interface. This question was rather unimportant because the amplitude of
the wave was constant, and no matter how much the wave travelled (even infinite), it had the same
amplitude. This however is not true for a lossy medium. If we assume that the incident wave was
originated at an infinite distance from the interface, it would attenuate to zero for any finite value of (that
is any finite value of ). On the other hand if we assume that an incident wave of finite amplitude is
present at the interface, and if it was originated at an infinite distance from the interface, its energy would
be infinite at its origin. Well, we need not worry about this as in practice neither the waves would have
travelled infinite distance, nor they will be ideal plane waves. In our analysis therefore we start with an
incident wave whose amplitude is known at the interface and find the fields in the two media in the vicinity
of the interface with proper incorporation of the attenuation constant .
EXAMPLE 5.10 : A uniform plane is normally incident on an infinitely thick dielectric slab, having dielectric
constant 10 and loss tangent at rad/sec. If the power density of the incident wave is
. Find the power density of the wave in the dielectric at a distance of 10 m from the surface.

Solution :

For the incident wave the power density is

(All the quantities are assumed to be RMS here.)

The transmitted field at the surface is

Since, the dielectric slab is lossy, the field amplitude reduces exponentially as a function of
distance where,
The first magnitude at 10 m distance from the surface is

The power density of the wave is

EXAMPLE 5.11 : A plane wave is normally incident on an infinitely thick dielectric slab of refractive index
2. Find the phase velocities of the total waves in air and in the dielectric . If the velocity varies with
distance, find the upper and lower bound on the phase velocity.

Solution :

We can note that in air we have incident as well as the reflected waves whereas in the dielectric we have
only transmitted wave. In air therefore we get standing wave and in the dielectric we have only a traveling
wave.
Let the phase constant of a wave in air be . Then the phase constant in the dielectric is MATH .
In the dielectric the phase velocity

(Note: , the velocity of light in air ). In the air, we have to find the phase velocity from the first
principles. Writing the total electric field for the standing wave in the air we have

Here the time variation is written explicitly and is the reflection coefficient.

Substituting for we get,

The phase of the standing wave therefore is

The phase velocity can be obtained by making the phase stationary as a function of time i.e.,

The phase velocity in air therefore varies as a function z. The maximum and minimum value of phase
velocity are and .
REFLECTION FROM A CONDUCTING BOUNDARY
Let us investigate here the reflection of a plane wave from an ideal conducting boundary ( ). In this
case naturally we have a dielectric-conductor interface and the plane wave is incident from the dielectric
side. Since no fields can exist inside an ideal conductor, there is no question of transmission (i.e.
refraction) of the wave to medium 2. We therefore have only reflection at the dielectric-conductor
interface. As in previous case, here also we analyse reflection of two polarizations, namely parallel and
perpendicular.
Perpendicular Polarization
Let us consider a plane wave with polarization perpendicular to the plane of incidence as shown in Figure
5.24.
\begin{figure}[h]
\includegraphics[width=0.70\textwidth]{Nfig5.24a.eps}
\caption{Reflection from a conducting boundary: Prependicular polarization.} \end{figure}
Let the angle of incidence be , and let the -fields for both incident and reflected waves be oriented
along -direction. The directions of the -fields can be obtained using the Poynting theorem, i.e.,
should give the direction of the wave propagation. The fields for the incident wave hence can be written
as

and the reflected fields can be written as

Applying boundary conditions at the interface i.e., continuity of the tangential components of the electric
field and the normal components of the magnetic field, and noting that the fields are zero inside an ideal
conductor, we get at ,

Noting that the fields are tangential to the interface, we get from (5.154),
y-component of the elctric field field:

For normal component of the magnetic field,

z-component of the magnetic field:

The -field is in -direction and hence is intrinsically parallel to the boundary. (5.156) and (5.157) both
give the same condition i.e.

From (5.158) it is clear that the reflected electric field is equal in magnitude to the incident field but
out of phase. In other words, the reflection coefficient

We may recall that on a transmission line, the voltage reflection coefficient becomes for the short
circuit load. The ideal conducting boundary therefore is analogous to the short circuit on the transmission
line.
Substituting for from eqn (5.158), the total field in medium 1 is

Similarly, H-field can be written as:

Equations (5.161) and (5.164) represent a travelling wave in -direction and a standing wave iz -
direction. Figure 5.25 shows the amplitude of , and components as a function of distance in
-direction.
\begin{figure}[t]
\includegraphics[width=1.0\textwidth]{5.24.eps}
\caption{Various field amplitudes} \end{figure}
From Figure 5.25 following things can be noted:
(1) and standing wave patterns are aligned in space whereas, the standing wave pattern is
shifted by quardrature with respect to the (or ) pattern. That is, where is maximum, is
zero and vice versa.
(2) and are zero at the interface as required by the boundary conditions.
(3) The electric field becomes zero when

Here we have used MATH and only negative sign has been chosen since is negative in medium 1.
Above equation (5.166) is the equation of a plane parallel to the -plane, i.e., parallel to the interface.
We therefore see that the electric field is zero in all planes which are parallel to the interface and are at
distances MATH , MATH , MATH , and so on. Also from (5.164) the -component of the
magnetic field i.e., the component of the magnetic field which is normal to the planes, is also zero in these
planes. In other words, the magnetic field becomes tangential on these planes.
(4) The magnetic field component is zero when

(5) The tangential component of the magnetic field, is maximum at the interface. Now since there is no
magnetic field inside the conductor, there has to be a surface current at the interface
where unit outward normal to the interface. For our coordinate system . We therefore have

The surface current has a magnitude of MATH and it flows along direction.
(6) The net power flow is in -direction, that is along the interface. This can be seen in two ways:
(a) The fields represent a travelling wave along -direction with a phase constant MATH .
MATH
(b) If we calculate the average Poynting vector , we find that the fields and are out of
phase, giving average power flow zero along -direction. The fields and are however in phase, and
therefore constitute an average power flow along -direction ( MATH ). The average power flow is

(7) The Wavelength along the interface is

(8) This case is analogous to a short-circuited transmission line. If we take for voltage on the
transmission line, the variation of is same as the voltage standing wave pattern on the line.
(9) For , i.e., for grazing incidence, becomes zero. We can therefore conclude that no wave
with polarization can be launched at a conducting interface at the grazing angle.
Parallel Polarization
Let us now carry out an analysis for a wave which has its electric field in the plane of incidence as shown
in Figure 5.26. Let the angle of incidence be .
Since and the incident wave vector lie in the plane of the paper, the magnetic filed vector has to be
perpendicular to the plane of the paper. Without losing generality, let us assume that the magnetic fields
for both, incident and reflected waves are pointing out of the plane of the paper,i.e., along -axis. The
direction of the electric fields can be appropriately chosen to give correct directions of the Poynting
vectors for the two waves as shown in Figure 5.26.
\begin{figure}[h]
\includegraphics[width=1.0\textwidth]{Nfig5.26.eps}
\caption{Variation of field amplitudes} \end{figure}
In vector form the fields for the incident and the reflected waves can be written as
Incident Wave:

Reflected Wave:

Continuity of the tangential components of the electric field at the interface yields

That is, the reflection coefficient for the field direction assumed in Figure 5.26 is

Substituting for from (5.178) and carrying out some algebric manipulations we get

and

The interpretations of the above equations are similar to that for the perpendicular polarization except the
field components are different in two cases. In this case the electric field has two components and
and the magnetic field has only component. The fields have standing wave nature in -direction and a
travelling wave nature in the -direction. The standing wave patterns for the three components are shown
in Figure 5.27.
\begin{figure}[t]
\includegraphics[width=0.70\textwidth]{Nfig5.27b.eps}
\caption{Variation of field amplitudes} \end{figure}
Following observations can be made from eqns (5.180) and (5.181):
(1) and patterns are aligned in space and pattern is in quardrature with respect to them.
(2) At the interface the magnetic field is enhanced by factor of 2 i.e., the magnetic field at the interface is
double of the incident field.
(3) The surface current at the ineterface is MATH and is independent of the angle of incidence.
(4) The wave travels along -direction with a phase constant of

giving guided wavelength along the interface

(5) The average power flow is along -direction and the average power flow density is

The power flow ceases for , i.e., for the normal incidence.
(6) The -component of the electric field goes to zero in planes parallel to the interface and at locations
given by

Again only -ve sign is taken as is negative in medium 1.

(7) For i.e., at the grazing incidence,

Note that unlike polarization, the reflection at grazing angle is possible for the parallel polarization. This
is due to the fact that for the grazing incidence of a parallely polarized wave, the electric field is
perpendicular to the conducting boundary, and the magnetic field is tangential to the conducting
boundary. In other words, in this case the tangential component of the electric field and normal
component of the magnetic field are intrinsically zero satisfying the natural boundary conditions at the
interface. The conducting boundary consequently does not have any effect on the wave propagation and
the wave keeps propagating as it would have propagated in the absence of the boundary. There is a little
difference however and that is, in the absence of the boundary the wave fronts will be over the infinite
space (MATH ), whereas in the presence of the boundary the phase fronts is only over the semi-
infinite space (MATH ).
Normal Incidence
For normal incidence, and and become parallel to the conducting boundary. Taking and
oriented along the -direction, and and along -direction the fields can be written as

Incident Wave:

Reflected Wave:

At the interface (i.e., ), the electric field, which is intrinsically tangential to the boundary in this case,
has to be zero, giving

The reflection coefficient therefore is

It can be observed that this case is identical to a transmission line where the wave propagation is in one
dimension ( along -direction), and the line is terminated in a short circuit load. Since the magnitude of the
reflection coefficient is unity, there is no net average power flow in any direction.
The above analysis clearly indicates that irrespective of the polarization of the incident wave, the electric
field component parallel to the boundary goes to zero at a distance MATH from the boundary. This
means if we place an infinite conducting plane parallel to the interface at this distance, the fields would
inherently satisfy the boundary conditions and hence would not be disturbed. The presence of another
conducting plane however, creates a bound structure in the -direction. In other words, the
electromagnetic fields are confined between two parallel conducting planes and there is a net flow of
power along the planes (along -direction).
We therefore find that a parallel conducting plane geometry is capable of guiding electromagnetic waves.
This structure hence is called the `parallel plane waveguide'. The parallel plane waveguide, although not
a very practical structure, forms the foundation of a more practical structure, the `rectangular waveguide'.
It is interesting to note that even though no energy penetrates the conducting boundary, the boundary
controls the flow of power. Ofcourse there are surface charges and surface currents which are
responsible for the guiding of the wave.
In the next chapter we will analyse different metallic waveguides and later the dielectric waveguides.
REVIEW QUESTIONS
1. What is a media interface?
2. What is the equation of a phase front of a unifrom plane wave travelling in an arbitrary direction?
3. What is phase velocity? and what is its range?
4. What is group velocity?
5. In what situation the phase and group velocities are equal?
6. What is plane of incidence?
7. Give arguments to show that when a uniform plane wave is incident on a plane media interface, the
reflected and refracted waves are also uniform plane waves.
8. What are the Snells laws of reflection and refraction?
9. Define angles of incidence, reflection and refraction.
10. What does perpendicular and parallel polarisation mean?
11. What is reflection and transmission coefficient?
12. If the magnitude of reflection coefficient is one, should the transmission coefficient be zero?why?
13. Can the transmitted electric field be greater than the incident electric field ?
14. Can the reflected electric field be greater than the incident electric field.?
15. Under what condition the reflection coefficient is real?
16 If the transmitted electric field is greater than the incident electric field should the transmitted magnetic
field be necessarily smaller than the incident magnetic field?
17. On a media interface can one get reflectionless transmission?
18. What is Brewster angle?
19. Why the Brewster angle is also called the polarizing angle?
20. A infinitely thick dielectric sheet is coated with a dielectric layer. What is the dielectric constant of the
coated layer so as not to get any reflection for a normally incident wave?
21. Why a linearly polarized wave remains linearly polarized after reflection from a dielectric interface but
the circularly polarised wave becomes elliptically polarised?
22. What is total internal reflection?
23. What is the difference between TIR and ordinary reflection?
24. Will a linearly polarized wave remain linearly polarized after TIR.
25. What is the propagation constant of the transmitted fields at TIR.
26. How the transmitted fields vary as a function of distance away from the interface?
27. Draw the amplitude variation of net travelling wave at TIR.
28. For a dielectric interface does the Brewster angle always exist?
29. At TIR what is the direction of net power flow?
30. At a conducting interface why the reflection coefficient is ?
31. A uniform plane wave is obliquely incident on an ideal conducting interface. What is the direction of
the surface current at the interface? See for both perpendicular and parallel polarizations.