Physics 506 Midterm exam March 4, 2004

1.
20 points
Consider a waveguide with a square cross section of sidelength a. A monochromatic field with
angular frequency is present the guide.

a) What are the three basic types of waveguide modes? Which ones occur in the given
guide?
b) For the given guide, determine the lowest cutoff frequency. What is the degree of
degeneracy of the lowest cutoff frequency, i.e. how many guide modes have that cutoff
frequency? Determine all E and H-field components of these modes. Include all spatial
and temporal dependences in your answers for the fields.

c) We assume cm a 5 = and s rad / 10 4 . 2
10
= . A small experimental setup that has no
effect of the waveguide field is located at the middle of the guide. How much power
needs to be injected to obtain an amplitude of the transverse electric field of 100 V/cm
in the setup? If you found degeneracy in part b, assume that only one of the degenerate
modes is excited.
d) Assume that the wave travels through 1000m of guide before it reaches the experiment.
How much power needs to be injected in order to achieve the condition of part c?
(wall conductivity
1 8
) 10 2 (

= m ).

e) The shape perturbation indicated in the drawing is applied to the guide. Using
perturbation theory and the mode(s) of part b, determine the k-numbers and polarization
angles of the transverse electric fields of the perturbed modes (all walls have = ).



You may use all applicable results stated in the homework solutions or in the textbook
without losing credit. You may receive partial credit for derivations in case an answer is
wrong. Derivations are required for results not found in the textbook/homework.
2.
20 points

A current
t
o
I I
i
e Re

= is flowing in the x-y plane on a circle with radius a centered to the
origin.
a) Find all spherical-multipole radiation coefficients ) , ( m l a
E
and ) , ( m l a
M
without
using any small-source approximation.

Which coefficients are zero, and which ones are not?

b) Knowing the coefficients, how would you obtain the E and H-fields in the radiation
zone, the angular distribution of the radiated power,
d
dP
, and the total radiated power
P ? It is not required to work out the expressions for the fields, d dP/ and P.

c) Using the small-source approximation, ka<<1, and assuming that the observation point
is in the radiation zone, determine the r ,

- and - components of the E and H-fields

of the lowest-order non-vanishing spherical-multipole radiation field. What are the
distribution of the radiated power,
d
dP
, and the total radiated power P?



3.
20 points
Scattering of light with wavelength nm 632 = is used to study spherical non-absorbing and
non-permeable nanoparticles with 20nm diameter and refractive index n=1.5. The incident light
is a plane wave propagating in z-direction.

a) Find the differential scattering cross section of a single particle for the case that the
incident light is linearly polarized ( x
0
= , no selectivity with respect to outgoing
polarization). Provide a numerical result.
b) Find the differential scattering cross section for the case that the incident light is
circularly polarized with negative helicity (

=
0
, no selectivity with respect to
outgoing polarization). Provide a numerical result.
c) A small-diameter laser beam with circular polarization and 10W power is traversing
through a 1mm thick dust layer filled with such particles at a density of
3 11
10

cm . The
particles are in a state of total spatial disorder. A detector with an area of
2
1cm is
located at a distance of 50cm from the intersection volume between the laser beam and
the dust layer, at a scattering angle of . What is the power measured by the detector?
Make reasonable approximations and provide a numerical result.
Phys. 506 Electricity and Magnetism Winter 2004
Prof. G. Raithel
Midterm exam
50 Points = 100%
1. Problem 20 Points (4 on each part)
a): Waveguides with multiple surfaces have TE, TM and TEM modes. The given one has only one surface
and therefore only supports TE and TM modes.
b): According to Section 8.4 of Jackson, the lowest mode is the TE
10
-mode. Since the cross section is square,
the TE
10
mode is degenerate with TE
01
(i.e. the degree of degeneracy is two). The cuto frequency is

10
=
01
=

a
=
c
a
and k =
1
c

2
01
(not required). The elds are given by Eq. 8.46 of Jackson,
TE
10
:
H
z
= H
0
cos

x
a

exp(ikz it)
H
x
=
ika

H
0
sin

x
a

exp(ikz it)
E
y
=
ia

H
0
sin

x
a

exp(ikz it)
TE
01
:
H
z
= H
0
cos

y
a

exp(ikz it)
H
y
=
ika

H
0
sin

y
a

exp(ikz it)
E
x
=
ia

H
0
sin

y
a

exp(ikz it)
where the extra in the last line follows from the fact that S =
1
2
E
t
H

t
must point in the positive
z-direction.
c): Pick, for instance, TE
10
. It is
P =
1
2

A
Re [ z E
t
H

t
] da =
1
2

A
Re [E
y
H

x
] da =
a
2
4
ka
2

2
|H
0
|
2
Since also the amplitude of the transverse electric eld, E
y
= H
0
a

, it is
P = E
2
0
a
2
k
4
We also use k =
1
c

c
a

2
, which gives k = 49.61m
1
. The numerical result for P is
P = 102.8W
Alternate method: Use Eq. 8.51 with = H
0
cos

x
a

with H
0
=
E
0
a
= 20.83
A
m
.
d): We need to nd the damping constant, which we can calculate, for instance, for the TE
10
-mode:
=
1
2P
1
2

|H|
2
dl =
1
2P
1
2

|H
z
|
2
+|H
x
|
2

dl
Along the ysides, which have x = 0 or x = a, it is H
x
= 0 and H
z
= H
0
, and thus

x=0 or a
|H
z
|
2
dl = 2a|H
0
|
2
Along the xsides, which have y = 0 or y = a, it is

y=0 or a

|H
z
|
2
+|H
x
|
2

dl = 2

a
x=0
|H
0
|
2

cos
2

x
a

+
k
2
a
2

2
sin
2

x
a

dx = a|H
0
|
2

1 +
k
2
a
2

and the sum over all sides,

|H|
2
dl = a|H
0
|
2

3 +
k
2
a
2

With the result from c), P =

a
2
4
ka
2

2
|H
0
|
2
, we nd
=

1
2
3
2
+ k
2
a
2
ka
3
= 3.32 10
3
m
1
The injected power is the result of c) times exp(2 1000m) = 765,
P = 78.8kW
Alternate method to nd : use Eq. 8.63 with the information in the next two paragraphs.
e): Since there are two degenerate TE-modes, we use the result of Problem 8.13a. Since the modes will be
normalized during the process, we can use just

10
= cos

x
a

and
01
= cos

y
a

The normalization integrals are then

N
10
= N
01
=

A
|
any
|
2
da =
a
2
2
The deformation is = a/16 on the lower side for a/2 < x < a and on the upper side for 0 < x < a/2.
Otherwise = 0. Also, since we are dealing with TE-modes, all single-normal-derivative terms in the
equation for
ij
are zero. On the lower side,

10
= cos

x
a

and
01
= 1 and

2

10
n
2
=

2

10
y
2

y=0
= 0 and

2

01
n
2
=

2

01
y
2

y=0
=

2
a
2
On the upper side,

10
= cos

x
a

and
01
= 1 and

2

10
n
2
=

2

10
(y)
2

y=a
= 0 and

2

01
n
2
=

2

01
(y)
2

y=a
=

2
a
2
Thus,

10,10
=

(x)

10

10
n
2
dl = 0

01,10
=

(x)

01

10
n
2
dl = 0

01,01
=

(x)

01

01
n
2
dl = 2

a/2
x=0
a
16

2
a
2

dx =

2
16

10,01
=

(x)

10

01
n
2
dl = 2

a/2
x=0
a
16

2
a
2
cos

x
a

dx =

2
8a

sin

x
a

a/2
0
=

8
Thus, the equation to be solved is

x
a
2
2

8
0 x
a
2
2


2
16

a
10
a
01

= 0
where we use x =
2

2
0
. The determinant is zero if x = 0 or x =

2
8a
2
. The perturbed solutions then are
For x = 0: a
10
= 1 and a
01
= 0. This is the original unperturbed TE
10
-mode. Since x = 0, this mode also
retains its unperturbed values of and k, and it is k = k
0
= 49.61m
1
. The electric-eld polarization is
= /2 (i.e. polarized in y-direction).
For x =

2
8a
2
: Since x =
2

2
0
= k
2
0
k
2
=

2
8a
2
, the perturbed k-value is
k =

k
2
0

2
8a
2
= 44.36m
1
Inserting x =

2
8a
2
into the above equation for a
10
and a
01
, we nd a
10
=
2

a
01
and an unnormalized
perturbed solution =
2

10
+
01
. Noting further that the unperturbed eld modes dier by a minus in
the electric-eld components, we see that for an electric eld of 1 in some unit in the x-direction, coming
from the TE
01
-part, the electric eld in the y direction, coming from the TE
10
-part, is +
2

. The polarization
angle thus is = arctan

= +32
o
.
2. Problem 20 Points (8 on a, 6 on b, 6 on c)
a): Use Eqs. 9.167f with = 0, M = 0 and
J(x) =
I
0
a
(r a)(cos )

No proof required, but to see that this is correct integrate over a plane of constant :

r=0

J(x)rddr =

r=0
I
0
r
a
(r a)(cos )ddr = I
0
Since in the present case r J, it is a
E
(l, m)=0.
To determine a
M
(l, m), we require (r J):
r J =
I
0
a
(r a)(cos )r(r

) = I
0
(r a)(cos )

(r J) =
I
0
r sin

(sin (r a)(cos ))
=
I
0
(r a)
r sin

cos (cos ) + sin

(cos )

=
I
0
(r a)
r sin

0 + sin

d cos
d

d
d cos
(cos )

=
I
0
(r a)
r
(cos )
d
d cos
sin
This is to be read as a distribution, i.e. the derivative needs to be applied on all -dependent functions that
will show up under the integral. Thus, following Eq. 9.168 of Jackson
a
M
(l, m) =
I
0
k
2
i

l(l + 1)

(r a)
r
j
l
(kr)(cos )
d
d cos
(sin Y

lm
(, )) r
2
d cos d
=
iI
0
k
2
aj
l
(ka)

l(l + 1)

(cos )
d
d cos
(sin Y

lm
(, )) d cos d
=
iI
0
k
2
aj
l
(ka)

l(l + 1)
2
m,0

(cos )
d
d cos
(sin Y

l0
(, 0)) d cos
=
iI
0
k
2
aj
l
(ka)

l(l + 1)
2
m,0

(cos )
d
d cos
(sin Y

l0
(, 0)) d cos
Denoting x = cos , we have
a
M
(l, m) =
iI
0
k
2
aj
l
(ka)

l(l + 1)
2
m,0

2l + 1
4

1
1
(x)
d
dx

1 x
2
P
l
(x)

dx
=
iI
0
k
2
aj
l
(ka)

(2l + 1)

l(l + 1)

m,0

1
1
(x)

d
dx
P
l
(x)

dx
=
iI
0
k
2
aj
l
(ka)

(2l + 1)

l(l + 1)

m,0
d
dx
P
l
(x)

x=0
(1)
It is a
M
(l, m) = 0 for m = 0 and l odd.
b): Copy Eqs. 9.149, 9.150 and 9.155 of Jackson for our case,
H =
exp(ikr it)
kr

l odd
(i)
l+1
a
M
(l, 0) n X
l,0
E = Z
0
H n
dP
d
=
Z
0
2k
2

l odd
(i)
l+1
a
M
(l, 0)X
l,0

2
P =
Z
0
2k
2

l odd
|a
M
(l, 0)|
2
Note n = r.
c): Use result of a) and Eq. 9.88, which says j
1
(ka) =
ka
3
for ka 0. Also,
d
dx
P
1
(x)

x=0
= 1. Thus,
a
M
(1, 0) =
iI
0
k
3
a
2

6
and the power
P =
Z
0
2k
2
|a
M
(1, 0)|
2
=
Z
0
I
2
0
k
4
a
4

12
Also, using Table 9.1 of Jackson, it is
dP
d
=
Z
0
I
2
0
k
4
a
4

12
3
8
sin
2

The magnetic eld

H =
exp(ikr it)
kr
(i)
2
a
M
(1, 0) n X
1,0
=
exp(ikr it)
kr
iI
0
k
3
a
2

6
r

L
1

2
Y
1,0
=
exp(ikr it)
kr
iI
0
k
3
a
2

12
r

1
i

1
sin

3
4
cos
=
exp(ikr it)
kr
I
0
k
3
a
2
4
(r

) sin
=

exp(ikr it)
r
I
0
k
2
a
2
4
sin
and the electric eld
E = Z
0
Hr =

exp(ikr it)
r
Z
0
I
0
k
2
a
2
4
sin
Alternate method 1: Use Eqs. 9.171f of Jackson.
Alternate method 2: Use results of Chapter 9.3 of Jackson for a magnetic dipole m = I
0
a
2
z.
3. Problem 20 Points (7 on a, 7 on b, 6 on c)
a): The spheres have =
0
and =
r

0
with
r
= n
2
= 2.25. Since further the radius of the spheres
a = d/2 = 10nm = 632nm, this is a case of electric-dipole scattering. We can use Eq. 10.6 of Jackson,
d
d
= k
4
a
6

r
1

r
+ 2

2
|

0
|
2
with incident polarization
0
= x =

1
0
0

. Two orthonormal vectors of linear exit polarization are

1
=

=

cos cos
cos sin
sin

and
2
=

=

sin
cos
0

For the given incident polarization, the dierential scattering cross section summed over the exit polarizations
is
d
d
= k
4
a
6

r
1

r
+ 2

2
|

1

0
|
2
+|

2

0
|
2

= k
4
a
6

r
1

r
+ 2

cos
2
cos
2
+ sin
2

= 8.45 10
22
m
2
sterad

cos
2
cos
2
+ sin
2

The angular part can be written in various other forms, such as cos
2
cos
2
+sin
2
= cos
2
+sin
2
sin
2
=
1 sin
2
cos
2
.
b): Use
0
=
1

2
( x iy) =
1

1
i
0

. Then,
|

1

0
|
2
=
1
2
(cos cos i cos sin )(cos cos + i cos sin ) =
1
2
cos
2

2

0
|
2
=
1
2
(sin i cos )(sin + i cos ) =
1
2
d
d
= k
4
a
6

r
1

r
+ 2

2
|

1

0
|
2
+|

2

0
|
2

= k
4
a
6

r
1

r
+ 2

2
1
2

1 + cos
2

= 8.45 10
22
m
2
sterad
1
2

1 + cos
2

Alternate method for a and b: With incident eld E

0
= E
0

0
, the induced dipole is
p = 4
0

r
1

r
+ 2

a
3
E
0

0
In the far-eld, it produces a scattered electric eld
E
sc
= Z
0
ck
2
4
exp(ikr)
r
[(r p) r] = Z
0

0
ca
3
E
0
k
2

r
1

r
+ 2

exp(ikr)
r
[(r
0
) r]
The scattering cross section, summed over exit polarizations, is
d
d
=
r
2
E

sc
E
sc
E

0
E
0
= k
4
a
6

r
1

r
+ 2

2
|(r
0
) r|
2
= k
4
a
6

r
1

r
+ 2

2
|r
0
|
2
Insert r =

sin cos
sin sin
cos

and the incident polarizations

0
to obtain the already given results.
c): Since the laser beam illuminates only a small volume and the diode is quite far away, all scattering
detected occurs at a practically constant scattering angle . Also, due to the disorder of the scatterers, the
scattering is incoherent (i.e. scattering cross sections of multiple particles just add up).
Since the scattering is linear, we may assume an incident beam with constant intensity I
in
over an area
A
in
= P
in
/I
in
. There, P
in
is the incident power.
The number of illuminated particles is N
P
= A
in
dn, where d is the layer thickness and n the particle volume
density.
The detector covers a solid angle =
A
r
2
, where A is the detector area and r the detector distance.
The detected power P then is
P =
d
d
(I
in
N
P
)
=
d
d

P
in
A
in
A
in
d n
A
r
2

=
d
d

P
in
d n
A
r
2

Inserting the result of b) for

d
d
, and the given numbers (P
in
= 10W, d = 10
3
mm, n = 10
17
m
3
,
A = 10
4
m
2
, and r = 0.5m), we nd
P = 1.70 10
10
W(1 + cos
2
)