Module I: Electromagnetic waves

Lecture 5: EM waves with boundaries

Amol Dighe
TIFR, Mumbai
Outline

1 EM waves at dielectric boundaries: reflection, transmission

2 EM waves in conductors: inside and at the boundary

Coming up...

1 EM waves at dielectric boundaries: reflection, transmission

2 EM waves in conductors: inside and at the boundary

Reflection and refraction

An EM wave is incident from one medium (1 , µ1 , n1 , c1 ) to another

medium (1 , µ1 , n1 , c1 ), at an angle θI with the normal to the boundary.
Incident, reflected and refracted waves
Incident wave

~I
E ~ I0 ei(~kI ·~r−ωt)
= E (1)
~kI
~I
B = ×E ~ I = 1 (k̂I × E
~ I) (2)
ω c1

Reflected wave

~R
E ~ R0 ei(~kR ·~r−ωt)
= E (3)
~kR
~R
B = ×E ~ R = 1 (k̂R × E
~ R) (4)
ω c1

Transmitted wave

~T
E = ~ T 0 ei(~kT ·~r−ωt)
E (5)
~kT
~T
B = ×E ~ T = 1 (k̂T × E
~T) (6)
ω c2
Boundary conditions on phases
~ ⊥ is continuous across the boundary
D

~ I⊥ + 1 E
1 E ~ R⊥ ~ T⊥
= 2 E (7)
~ I⊥0 e i(~kI ·~r−ωt) ~
~ R⊥0 ei(kR ·~r−ωt) ~ T ⊥0 ei(~kT ·~r−ωt)
1 E + 1 E = 2 E (8)

The equatity should be valid at all ~r on the boundary

~kI · ~r = ~kR · ~r = ~kT · ~r (9)

With origin at the point of incidence:

|~kI |r sin θI = |~kR |r sin θR = |~kT |r sin θT (10)

Using |kI | = |kR | and |kT |/|kI | = n2 /n1 ,

sin θI n2
sin θI = sin θR , = (11)
sin θT n1
The first is the law of reflection the second is the Snell’s law
Boundary conditions on amplitudes
Our discussion would have worked for any of the boundary
~ ⊥ as an example. Now we need not worry
conditions, we just took D
about the phases, since the laws of reflection and refraction derived
there guarantee that the phase conditions will be satisfied.
Boundary conditions

~ I⊥0 + 1 E
1 E ~ R⊥0 = ~ T ⊥0
2 E (12)
~ I⊥0 + B
B ~ R⊥0 = ~ T ⊥0
B (13)
~ Ik0 + E
E ~ Rk0 = ~ T k0
E (14)
1~ 1~ 1~
BIk0 + BRk0 = BT k0 (15)
µ1 µ1 µ2

For convenience we’ll divide the incident electric field into a

component in the plane of incidence (the plane that contains
~kI , ~kR , ~kT ) and a component normal to the plane of incidence. These
two clearly won’t interfere, and they can be added together at any
time, using the principle of superposition, to get the net electric field.
~ in the plane of incidence
E
Applying boundary conditions

~
Boundary conditions involving E

−1 EI0 sin θI + 1 ER0 sin θR = −2 ET 0 sin θT (16)

−EI0 cos θI + ER0 cos θR = ET 0 cos θT (17)
Solution:
   
α−β 2
ER0 = EI0 , ET 0 = EI0 (18)
α+β α+β

where
cos θT µ1 c1
α≡ , β= (19)
cos θI µ2 c2
~ give exactly the same conditions.
Boundary conditions involving B
Reflection and transmission coefficient
Rate of energy transported by incoming wave normal to the
boundary: (Correct this, right language, factors of c, connect
with N)
1 ~ I0 |2 cos θI
Incident wave : II = 1 c1 |E (20)
2
1 ~ R0 |2 cos θR
Reflected wave : IR = 1 c1 |E (21)
2
1 ~ T 0 |2 cos θT
Transmitted wave : IT = 2 c2 |E (22)
2
Reflection coefficient
 α − β 2
 
IR
R= =   (23)
II α +β

Transmission coefficient
IT 2 c2 cos θT 4Re(α∗ β)
T = = = (24)
II 1 c1 cos θI |α + β|2
Comments on reflection and transmission coefficients

R + T = 1, as expected
R = 1, T = 0 possible if α is purely imaginary.
q q
1 − sin2 θT 1 − (n2 /n1 )2 sin2 θT
α= = , (25)
cos θI cos θI

so if sin θI > (n1 /n2 ), there is no transmission.

This is the condition for Total Internal reflection.
R = 0, T = 1 possible if α = β. This condition takes a simple
form if µ1 = µ2 , since then

cos θT c1 sin θI c1
= = = (26)
cos θI c2 sin θT c2

This leads to sin 2θI = sin 2θT , that is θI + θT = π/2.

In such a case, θI is called the Brewster’s angle.
~ normal to the plane of incidence
E
Comments on this scenario

The values for R and T will in general be different. In particular,

R = 0 is not possible here.
If an unpolarized wave is incident on a dielectric surface, the
reflected and transmitted waves will therefore, in general, be
polarized.
Coming up...

1 EM waves at dielectric boundaries: reflection, transmission

2 EM waves in conductors: inside and at the boundary

Reflection from a conducting surface

No wave is transmitted inside the conductor; i.e. fields inside the

conductor are zero.
For a normal incidence, E ~ I = −E~R
I.e. there is a phase-shift by π.
~ I and E
For incidence at an angle, the components of E ~ R parallel
~ Ik = −E
to the boundary cancel, i.e. E ~ Rk

There will be charge oscillations at the metal surface

~ I⊥ + E
corresponding to 1 (E ~ I⊥ ) = σs , where σs is the surface
charge density
The movements of these charges along the surface correspond
~ Ik + H
to surface currents, which account for finite values of H ~ Rk
at the boundary.
The net B ~ normal to the surface vanishes, i.e. B
~ I⊥ + B
~ R⊥ = 0.
~
This follows automatically from the Ek conditions above.
Recap of topics covered in this lecture

Reflection and transmission at the surface of a dielectric

Boundary conditions at a conducting surface