Problem 1.

A conducting slab
A plane polarized electromagnetic wave E = EI eikz−ωt is incident normally on a flat uniform
sheet of an excellent conductor (σ  ω) having thickness D. Assume that in space and in
the conducting sheet µ =  = 1, discuss the reflection an transmission of the incident wave.

(a) Show that the amplitudes of the reflected and transmitted waves, corrrect to first order
in (ω/σ)1/2 , are:

ER −(1 − e−2λ )
= (1)
EI (1 − e−2λ ) + γ(1 + e−2λ )
ET 2γe−λ
= (2)
EI (1 − e−2λ ) + γ(1 + e−2λ )

where
r
2ω ωδ
γ= (1 − i) = (1 − i) (3)
σ c
λ =(1 − i)D/δ (4)
p
and δ = 2/ωµσ is the skin depth.

(b) Verify that for zero thickness and infinite skin depth you obtain the proper limiting
results.

(c) Optional: Show that, except for sheets of very small thickness, the transmission
coefficient is
8(Reγ)2 e−2D/δ
T = (5)
1 − 2e−2D/δ cos(2D/δ) + e−4D/δ
Sketch log T as a function of D/δ, assuming Reγ = 10−2 . Define “very small thickness”.

1
Problem 1. A conducting slab
A plane polarized electromagnetic wave E = EI eikz !t is incident normally on a flat uniform
sheet of an excellent conductor ( !) having thickness D. Assume that in space and in
the conducting sheet µ = ✏ = 1, discuss the reflection an transmission of the incident wave.

(a) Show that the amplitudes of the reflected and transmitted waves, corrrect to first order
in (!/ )1/2 , are:

ER (1 e 2 )
= (1)
EI (1 e 2 ) + (1 + e 2 )
ET 2 e
= 2 2
(2)
EI (1 e ) + (1 + e )

where
r
2! !
= (1 i) = (1 i) (3)
c
=(1 i)D/ (4)
p
and = 2/!µ is the skin depth.

(b) Verify that for zero thickness and infinite skin depth you obtain the proper limiting
results.

(c) Optional: Show that, except for sheets of very small thickness, the transmission
coefficient is
8(Re )2 e 2D/
T = (5)
1 2e 2D/ cos(2D/ ) + e 4D/
Sketch log T as a function of D/ , assuming Re = 10 2 . Define “very small thickness”.

1
Problem 2. In class problems
(a) Consider an incoming plane wave of light propagating in vacuum

EI = E0 eikz−iωt  . (6)

The light is normally incident (i.e. with angle of incidence θI = 0) upon a semi-infinite
slab of dielectric with µ = 1 and dielectric constant , which fills the half of space with
z > 0. Use the reflection and transmission coefficients discussed in class to show that
the (time-averaged) force per area on the front face of the dielectric is away from the
dielectric (i.e. in the −ẑ direction) and is equal in magnitude to

|F z |
 
1 2 n−1
= E0 (7)
A 2 n+1

(b) Consider an incoming plane wave of light propagating in vacuum

EI = E0 eikz−iωt  . (8)

The light is normally incident upon a slab of metal with conductivity in σ and µ =
 = 1. In class, we evaluated the (time-averaged) Poynting vector just inside the metal
and computed the (time-averaged) energy flux into the metal per area per time:
r
2ω
hS · ni = c |E0 |2 (9)
σ
Show that this energy flux is equal to (time-averaged) Joule heating in the metal.
(Hint: for ohmic material the energy dissipated as heat per volume per time is E · J
– I understand this result as qE · v/∆V = (force×velcoity)/(Volume).)

2
Problem 3. Snell’s law in a crystal
Consider light of frequency ω in vacuum incident upon a uniform dielectric material filling
the space y > 0. The light is polarized in plane (as shown below) and has incident angle θ1 .
The dielectric material has uniform permittivity  and µ = 1.

(a) Derive Snell’s law from the boundary conditions of electrodynamics.

Consider light propagating in a crystal with µ = 1 and dielectric tensor ij . Along the
principal crystalline axes ij is given by
 
1 0 0
ij =  0 2 0  , (10)
0 0 3

and thus, along the axes Di = i Ei (no sum over i).

(b) Starting directly from the Maxwell equations in the dielectric medium, show that the
frequency and wave numbers of the plane wave solutions E(t, r) = Eeik·r−iωt in the
crystal are related by

ω 2 i
 
2
det ki kj − k δij + 2 δij = 0 (no sum over i). (11)
c

Now consider light of frequency ω in vacuum incident upon a dielectric crystal. The light
has incident angle θ1 , and propagates in the x − y plane, i.e. kz = 0. The incident light is
also polarized in x − y plane, and the axes of the dielectric crystal are aligned with the x, y, z
axes (see below). Only the y axis of the crystal has a slightly larger dielectric constant than
the remaining two axes,  
 0 0
ij = 0  (1 + δ) 0 , (12)
0 0 
with δ  1.

(c) Determine angle of refraction (or sin θ2 ) including the first order in δ correction to
Snell’s law.

(d) Is the refracted angle smaller or larger than in the isotropic case? Explain physi-
cally. Does the angular dependence of your correction makes physical sense? Explain
physically.

(e) If the incident light is polarized along the z axis (out of the x − y plane), what is the
deviation from Snell’s law? Explain physically.

3
Figure 1: Snell’s law geometry

4
Problem 4. Analysis of the Good-Hänchen effect
A “ribbon” beam1 of in plane polarized radiation of wavelength λ is totally internally reflected
at a plane boundary between a non-permeable (i.e. µ = 1) dielectric media with index of
refraction n and vacuum (see below). The critical angle for total internal reflection is θIo ,
where sin θIo = 1/n. First assume that the incident wave takes the form E(t, r) = EI eik·r−iωt
of a pure plane wave polarized in plane and study the transmitted and reflected waves.

(a) Starting from the Maxwell equations, show that for z > 0 (i.e. in vacuum) the electric
field takes the form:
√ ωn sin θI
− ωc ( n2 sin θI2 −1)z i
E2 (x, z) = E2 e e c x (13)

(b) Starting from the Maxwell equations, show that for θI > θI0 the ratio of the reflected
amplitude to the incident amplitude is a pure phase
ER o
= eiφ(θI ,θI ) (14)
EI
and determine the phase angle. Thus the reflection coefficient R = |ER /EI |2 = 1
However, phase has consequences.
(c) Show that for a monochromatic (i.e. constant ω = ck) ribbon beam of radiation in
the z direction with a transverse electric field amplitude, E(x)eikz z−iωt , where E(x) is
smooth and finite in the transverse extent (but many wavelengths broad), the lowest
order approximation in terms of plane waves is
Z
dκ
E(x, z, t) =  A(κ)eiκx+ikz−iωt (15)
(2π)
where k = ω/c. Thus, the finite beam consists of a sum plane waves with a small range
of angles of incidence, centered around the geometrical optics value.
1
By a “ribbon” beam I mean a beam which has finite transverse extent in the direction perperndicular
kI lying in the x-z place as drawn above. But, the beam is infinite in extent in the y direction (coming out
of page in the figure above) . Thus the incoming and outgoing “ribbion” beams form a kind of wedge.

5
(d) Consider a reflected ribbon beam and show that for θI > θIo the electric field can be
expressed approximately as
o
ER = R E(x00 − δx)eikR ·r−iωt+iφ(θI ,θI ) (16)

where R is a polarization vector, x00 is the coordinate perpendicular to the reflected

dφ
wave vector kR , and the displacement δx = − k1 dθI
is determined by phase shift.

(e) Using the phase shift you computed, show that the lateral shift of the reflected in plane
polarized beam is

λ sin θI sin2 θIo

Dk = (17)
π sin θI − sin2 θIo sin θI2 − cos θI2 sin2 θIo
p 2

6
Problem 5. Reflection of a Gaussian Wave Packet Off a Metal
Surface:
In class we showed that the amplitude reflection coefficient from a good conductor (ω  σ)
for a plane wave of wavenumber k = ω/c is
r r !
HR (k) 2µω 2µω
=1− (1 − i) ' 1 − eiφ(ω) , (18)
HI (k) σ σ

where the phase is for ω  σ: r

2µω
φ(ω) ' . (19)
σ
Consider a Gaussian wave packet with average wave number ko centered at z = −L at
time t = −L/c which travels towards a metal plane located at z = 0 and reflects. Show that
the time at which the center of the packet returns to z = −L is given by
L µδo
t= + (20)
c 2c
p
where the time delay is due to the phase shift dφ(ωo )/dω, and δo = 2c/σµko is the skin
depth.