You are on page 1of 6

# Electricity and Magnetism II - Homework Assignment 2

Alejandro G´omez Espinosa

February 17, 2013
Jackson, 7.5 A plane polarized electromagnetic wave E = E
i
e
ik·x−iωt
is incident normally on a ﬂat
uniform sheet of an excellent conductor (σ ωε
0
) having a thickness D. Assuming that in space
and in the conducting sheet µ/µ
0
= ε/ε = 1, discuss the reﬂection and transmission of the incident
wave.
(a) Show that the amplitudes of the reﬂected and transmitted waves, correct to the ﬁrst order in
(εω/σ)
1/2
, are:
E
r
E
i
=
−(1 −e
−2λ
)
(1 −e
−2λ
) + γ(1 + e
−2λ
)
(1)
E
t
E
i
=
2γe
−λ
(1 −e
−2λ
) + γ(1 + e
−2λ
)
(2)
where
γ =
_

0
ω
σ
(1 −i) =
ωδ
c
(1 −i) (3)
λ = (1 −i)D/δ (4)
and δ =
_
2/ωµσ is the penetration depth.
Lets deﬁne the electric ﬁeld vectors on the incident side:
E = E
i
exp (ik · x −iωt) ; E
R
= E
r
exp (ik · x −iωt)
where the index i represents the incident waves and r is the reﬂected wave. Inside the conductor:
E
c
i
= E
+
exp (ik
0
· x −iωt) ; E
c
r
= E

exp (ik
0
· x −iωt)
and for transmitted side:
E
T
= E
t
exp (ik · x −iωt)
Using the boundary conditions for ﬁelds perpendicular to the plane of incidence, we have for
the incidente side:
E
i
+ E
r
= E
+
+ E

(5)
(E
i
−E
r
) = n(E
+
−E

) (6)

gomez@physics.rutgers.edu
1
Since the polarized wave is incident normally, notice that all the angle dependency is gone. For
the transmited side:
E
+
e
ikD
+ E

e
−ikD
= E
t
(7)
n(E
+
e
ikD
−E
r
e
−ikD
) = E
t
(8)
In equations (6) and (10), n must be deﬁne as:
n =
_
ε
ε
0
=
_
1 +

ωε
0
≈ (1 + i)
_
σ
2ωε
0
=
2
γ
(9)
where n is the complex index of refraction, σ is the dielectric conductivity and, we approximate
this value to the conditions of (3) . We must introduce this deﬁnition because we want to treat
our system as a dielectric, as before, and we can do this as long as we handle our conductor as
a medium with complex dielectric constant. Using this deﬁnition, we can also deﬁne a phase
change:
φ = kD =
ωnD
c
≈ (1 + i)
ωD
c
_
σ
2ωε
0
= (1 + i)D
_
µ
0
σω
2
= iλ (10)
where the approximation is consistent with (4). Given this, lets used the boundary conditions
to found the coeﬃcients of reﬂected and transmited waves.
(5) + (6):
E
i
_
1 +
1
n
_
+ E
r
_
1 −
1
n
_
= 2E
+
(11)
(5) - (6):
E
i
_
1 −
1
n
_
+ E
r
_
1 +
1
n
_
= 2E

(12)
(9) + (10):
2E
+
e

= E
t
_
1 +
1
n
_
(13)
(9) - (10):
2E

e
−iφ
= E
t
_
1 −
1
n
_
(14)
Replacing (13) in (11):
E
i
_
1 +
1
n
_
+ E
r
_
1 −
1
n
_
= E
t
_
1 +
1
n
_
e
−iφ
(15)
Replacing (14) in (12):
E
i
_
1 −
1
n
_
+ E
r
_
1 +
1
n
_
= E
t
_
1 −
1
n
_
e

(16)
Finally, solving (15) for E
r
and replacing in (16), we found:
E
t
E
i
=
4
n
_
1 +
1
n
_
2
e
−iφ

_
1 −
1
n
_
2
e

=
4
ne

_
1 +
1
n
2
_
(1 −e
2iφ
) +
2
n
(1 + e
2iφ
)
(17)
2
Same for E
t
, replacing it in (15) and, rearrange terms:
E
r
E
i
=

_
1 −
1
n
2
_
(1 −e
2iφ
)
_
1 +
1
n
2
_
(1 −e
2iφ
) +
2
n
(1 + e
2iφ
)
(18)
Finally, keeping only the ﬁrst terms of
1
n
2
, and replacing the values of (9) and (10), the relations
are given by:
E
r
E
i
=
−(1 −e
−2λ
)
(1 −e
−2λ
) + γ(1 + e
−2λ
)
E
t
E
i
=
2γe
−λ
(1 −e
−2λ
) + γ(1 + e
−2λ
)
Results agree with (1) and (2).
(b) Verify that for zero thickness and inﬁnite thickness you obtain the proper limiting results.
Zero thickness corresponds to D →0 that, according to (10), corresponds to λ →0, therefore:
E
r
E
i
=
−(1 −e
−2λ
)
(1 −e
−2λ
) + γ(1 + e
−2λ
)
−−−→
λ→0
0
E
t
E
i
=
2γe
−λ
(1 −e
−2λ
) + γ(1 + e
−2λ
)
−−−→
λ→0
1
This results makes sense due to in the case of zero thickness there will be not reﬂected wave
and the transmited wave must be the incident wave.
On the other hand, in the case of inﬁnite thickness, λ →∞, hence:
E
r
E
i
=
−(1 −e
−2λ
)
(1 −e
−2λ
) + γ(1 + e
−2λ
)
−−−→
λ→∞

1
1 + γ
E
t
E
i
=
2γe
−λ
(1 −e
−2λ
) + γ(1 + e
−2λ
)
−−−→
λ→∞
0
Here there will be no transmited wave because of the inﬁnite thickness and only reﬂected waves
depending upon γ, i.e. the conductivity of the material.
(c) Show that, except for sheets of very small thickness, the transmission coeﬃcient is
T =
8(Reγ)
2
e
−2D/δ
1 −2e
−2D/δ
cos(2D/δ) + e
−4D/δ
(19)
Sketch logT as a function of (D/δ), assuming Reγ = 10
−2
. Deﬁne ”very small thickness”.
Lets deﬁne ﬁrst very small thickness for the transmitted wave, i.e. (2). In this equation, the second
term in the denominator must be approximate to zero, therefore:
0 ≈ γ(1 + e
−2λ
) ≈ γ (1 + 1 −2λ + ...) ≈ 2γ −2λ + ...
|2γ| ≈ |2λ|
D
δ

ωδ
c
3
where we can deﬁne small thickness as D <
ωδ
2
c
. Then, since the second term in the denominator
is approximate zero for small thickness, the ratio can be approximate as:
E
t
E
i

2γe
−2λ
1 −e
−2λ
Let us calculate now the transmitted coeﬃcient T:
T =
¸
¸
¸
¸
E
t
E
i
¸
¸
¸
¸
2
=
¸
¸
¸
¸
2γe
−2λ
1 −e
−2λ
¸
¸
¸
¸
2
= Re
_
4|γ|
2
e
−2λ
1 −2e
−2λ
+ e
−4λ
_
=
8(Reγ)
2
e
−2D/δ
1 −2e
−2D/δ
cos(2D/δ) + e
−4D/δ
Finally, Figure 1 sketches, in a logarithm scale, the dependence of the transmitted coeﬃcient as
function of D/δ.
δ D/
0 0.5 1 1.5 2 2.5 3
T

c
o
e
f
f
i
c
i
e
n
t
-4
10
-3
10
-2
10
-1
10
Figure 1: Depencende of the transmitted coeﬃcient T, in logarithm scale, as function of D/δ according
to equation (19).
4
Jackson, 7.12 The time dependence of electrical disturbances in good conductors is governed by the
frequency-dependent conductivity (7.58). Consider longitudinal electric ﬁelds in a conductor, using
Ohm’s law, the continuity equation, and the diﬀerential form of Coulomb’s law.
(a) Show that the time-Fourier-transformed charge density satisﬁes the equation
[σ(ω) −iωε
0
]ρ(x, ω) = 0 (20)
Starting with the continuity equation:
∇· J = −
∂ρ
∂t
(21)
Let us replace in (21) Ohm’s Law, J = σE:
∇· (σE) = −
∂ρ
∂t
where, if σ is uniform:
σ(∇· E) = −
∂ρ
∂t
σ
_
ρ
ε
0
_
= −
∂ρ
∂t
σρ + ε
0
∂ρ
∂t
= 0 (22)
Now, the time-Fourier-transformed charge density is given by:
ρ(t) =
1

_
ρ(ω)e
−iωt
dω (23)
Pluging (23) in (22):
σ

_
ρ(ω)e
−iωt
dω +
ε
0

_
ρ(ω)

∂t
e
−iωt
dω = 0
1

_
_
σρ(ω)e
−iωt
−iωε
0
ρ(ω)e
−iωt
_
dω = 0
σρ(ω)e
−iωt
−iωε
0
ρ(ω)e
−iωt
= 0
(σ −iωε
0
) ρ(ω)e
−iωt
= 0
where for every time t:
[σ −iωε
0
]ρ(ω) = 0 (24)
In agreement with (20).
5
(b) Using the representation
σ(ω) =
σ
0
1 −iωτ
(25)
where σ
0
= ε
0
ω
2
p
τ and τ is a damping time, show that in the approximation ω
p
τ 1 any
initial disturbance will oscillate with the plasma frequency and decay in amplitude with a decay
constant λ = 1/2τ. Note that if you use σ(ω) ≈ σ(0) = σ
0
in part a, you will ﬁnd no oscillations
and extremely rapid damping with the (wrong) deca constant λ
ω
= σ
0

0
.
Using (24), let us replace σ(ω) from (25):
σ −iε
0
ω = 0
σ
0
1 −iωτ
−iε
0
ω = 0
σ
0
−iε
0
ω(1 −iωτ)
1 −iωτ
= 0
ε
0
ω
2
p
τ −iε
0
ω −ε
0
τω
2
1 −iωτ
= 0
ε
0
ω
2
p
τ −iε
0
ω −ε
0
τω
2
= 0
ω
2
+

τ
−ω
2
p
= 0
ω =
1
2
_

i
τ
±
_

1
τ
2
+ 4ω
2
p
_
=
1
2
_
_

i
τ
±
¸

2
p
τ
2
−1
τ
2
_
_
where if ω
p
τ 1:
ω = −
i

±ω
p
which probes that any initial disturbance oscillates with the plasma frequency and decay in ampli-
tude with a decay constant λ =
1

.
6