Li LYK

Progress In Electromagnetics Research, PIER 26, 249291, 2000
ELECTROMAGNETIC SCATTERING BY MULTILAYERED

CHIRAL-MEDIA STRUCTURES:
A SCATTERING-TO-RADIATION TRANSFORM
L.-W. Li, D. You, M.-S. Leong, and T.-S. Yeo
Communications and Microwave Division
Department of Electrical Engineering
National University of Singapore
10 Kent Ridge Crescent, Singapore 119260
J. A. Kong
Research Laboratory of Electronics and
Department of Electrical Engineering
Massachusetts Institute of Technology
Cambridge, MA 02139, USA
1. Introduction
2. The Chiral-Medium Wave Equation
3. Electromagnetic Scattering by a Multilayered Chiral
Sphere
3.1 Scattering Theory Using Eigenfunction Expansion Method
3.1.1 The Vector Wave Function Expansion
3.1.2 Determination of Scattering Coecients by
Boundary Conditions
3.2 The Method by Dyadic Greens Function
3.2.1 Dyadic Greens Function for Multilayered
Spherically Chiral Media
3.2.2 Source Which Generates Plane Waves at Innity
3.2.3 Radiated Field
4. Scattering by a Multilayered Chiral Cylinder
4.1 The Method of Eigenfunction Expansion
4.1.1 The Eigenfunction Expansion
250 Li et al.
4.1.2 Determination of Scattering Coecients by
Boundary Conditions
4.2 The Dyadic Greens Function Method
4.2.1 Dyadic Greens Function for Multilayered
Cylindrically Chiral Media
4.2.2 Source Which Generates a Plane Waves at Innity
5. Scattering by Planarly Stratied Chiral Media
6. Conclusions
References
1. INTRODUCTION
In electromagnetic wave theory, there exist two well-known distinct
areas. One of them is the electromagnetic radiation due to an antenna
located in an environment where scatterers of dierent geometries are
present, and the other is the electromagnetic scattering of an electro-
magnetic plane wave by objects of dierent geometries. These two
areas are almost parallelly developed. This paper aims at relating the
scattering problem to the radiation problem by utilizing the volumetric
distribution of an electric current source that is assumed to be located
at innity so as to generate the plane wave. Therefore, the scattering
problem can be considered as a specic radiation problem where the
radiated source is located at innity, which forms the scattering-to-
radiation transform.
During the past few decades, much attention has been focused on the
interaction between chiral media and electromagnetic elds, as a result
of numerous applications in the electromagnetic scattering, antenna ra-
diation, and radio wave propagation. The historical background, and
a general description, of the subject of electromagnetic chirality and
its applications can be found, as examples, in literature [18]. Among
the research work about the electromagnetic wave scattering by chiral
objects, exact (or analytic) and asymptotic (or numerical) solutions to
electromagnetic scattering by chiral sphere(s) [913], cylinder(s) [14
17], spherical shells [18] and a spheroid [19] are available in the liter-
ature. In the present paper, the method of vector wave eigenfunction
expansion [10, 16, 19, 20] is utilized to calculate the electromagnetic
scattering by a planarly stratied chiral medium structure, a multilay-
Electromagnetic scattering by multilayered chiral-media structures 251
ered chiral cylinder, and a multilayered chiral sphere. Electromagnetic
elds in free space as well as in each layer of the multilayered chiral
media of three geometries are expanded in terms of the correspond-
ing vector eigenfunctions in their convenient coordinates systems, and
the coecients of the scattered electric eld are derived by matching
boundary conditions satised by the transmission coecient matrices
[21, 22] at all dielectric interfaces. By using the asymptotic forms for
large-argument Hankel functions and spherical Hankel functions, the
scattered electromagnetic elds in far-zone are obtained.
The dyadic Greens function (DGF) technique [2326], is a very
powerful and elegant method for solving boundary-value problems, and
has been extensively developed and used to solve the fundamental elec-
tromagnetic chirality problems of various geometries [1, 2, 5, 2729].
The developments and applications of dyadic Greens functions can
be found from the existing published work involving dierent physical
geometries, e.g., the unbounded chiral media [3, 3035], the planarly
stratied media [3642], the spherically multilayered chiral media [36,
4345], and the cylindrically multilayered media [46, 47].
Recently, Li et al. [48] presented an idea of making the plane wave in
electromagnetic scattering equivalent to an electromagnetic eld due
to a point source located at innity in a unbounded isotropic medium.
Also, the method for deriving the volumetric current distribution is
developed. Based on the dyadic Greens functions for chiral media of
dierent geometries and the assumed volumetric current distribution,
electromagnetic elds are formulated in terms of integrals consisting
the derived volumetric current distribution located at innity and the
dyadic Greens functions. Both scattering theory and radiation theory
are utilized in the paper for comparison. Good agreements between
the results obtained from the two methods are observed. Hence, the
scattering-to-radiation transform in three basic coordinates systems,
i.e., rectangular, cylindrical and spherical coordinates systems, are es-
tablished and veried.
2. THE CHIRAL-MEDIUM WAVE EQUATION
Consider an N-layered geometry of the chiral medium. The incident
waves of both parallel and perpendicular polarizations are assumed to
illuminate in the rst layer towards the (N 1)-layered medium. In
the following analysis the time dependence, e
it
, is assumed and is
suppressed throughout the paper. A chiral medium of a layer is usually
252 Li et al.
characterized by the following set of constitutive relations: [24, 35, 38,
47]
D
f
=
f
E+ i
f
H, (1a)
B
f
=
f
Hi
f
E, (1b)
where
f
,
f
, and
f
are the mediums permittivity, permeability
and chirality parameter, respectively, and f = 1, 2, , N denoting the
eld region. If
f
,
f
or
f
are complex, the medium is considered
to be lossy. If
f
=0, then (1a) and (1b) reduce to the constitutive
relations for an achiral medium.
In a chiral medium without source distributions, the wave equations
with constitutive relation are
2
_
E
f
H
f
_
+ [k]
2
_
E
f
H
f
_
= 0, (2a)
where
[k] =
_
f
i
f
i
f

f
_
. (2b)
By following Bohren or Lindell [9, 8], the coupling caused by [k] in
the wave equation can be removed by diagonalizing [k] such that
[A]
1
[k][A] =
_
k
(R)
f
0
0 k
(L)
f
_
. (3)
A simple form of [A] is found to be
[A] =
_
1 1
f
i
f
_
, (4)
with the achiral wave impedance given by :
f
=
_
f
. (5)
The propagation constant k
f
in each layer of the multilayered medium
is designated generally as [47]:
k
2
f
=
2
(
f
f

2
f
), (6)
and a symbol
f
is dened by
f
=
f
. (7)
Hence, there are two circularly polarized modes present in the un-
bounded medium, i.e., the right- and left-handed circularly polarized
(RCP and LCP) waves. Their corresponding wave numbers in (3) are
given by
k
(R)
f
=
f
+
f
, (8a)
k
(L)
f
=
f
+
f
. (8b)
Dene ( E
f
, H
f
) in terms of ( E
R
f
, E
L
f
), i.e.,
_
E
f
H
f
_
= [A]
_
E
(R)
f
E
(L)
f
_
(9)
where E
(R)
f
and E
(L)
f
are the electric elds of right and left circularly
polarized waves with propagation constants k
(R)
f
and k
(L)
f
. Thus the
decoupled source-free wave equations in chiral media can be written as
[16]:
2
_
E
(R)
f
E
(L)
f
_
+
_
k
R
2
E
(R)
f
k
L
2
E
(L)
f
_
= 0.
Subsequently, we will consider three general cases for the development
and application of the scattering-to-radiation transform, namely spher-
ical, cylindrical, and planar structures. Because the spherical coordi-
nates system has the highest-degree symmetry, the plane wave scat-
tering by a multilayered chiral sphere is considered rstly. The second
case is the scattering by a multilayered chiral cylinder and the last one
is the plane wave scattering by stratied planarly chiral media.
3. ELECTROMAGNETIC SCATTERING BY A
MULTILAYERED CHIRAL SPHERE
In this section, two techniques are to be developed and compared. The
rst is the scattering theory and the other is the radiation theory.
254 Li et al.
3.1 Scattering Theory Using Eigenfunction Expansion
Method
3.1.1 The Vector Wave Function Expansion
Consider two pairs of incident electromagnetic waves: parallel (I)
and perpendicular (II) polarizations incident at an arbitrary angle on
a sphere whose center o is at the origin of the Cartesian coordinates
system. The incident wave elds are expressed by:
E
i
I
= E
I
(cos xsin z)e
ik
0
(xsin +z cos )
, (11a)
H
i
I
=
k
0
E
I
0
ye
ik
0
(xsin +z cos )
, (11b)
and
E
i
II
= E
II
ye
ik
0
(xsin +z cos )
, (12a)
H
i
II
=
k
0
E
II
0
(cos xsin z)e
ik
0
(xsin +z cos )
, (12b)
where E
I
and E
II
are the amplitude of the incident electric eld,
is the incident angle with respect to the z-axis. It is assumed for con-
venience that the incident wave lies on the xz-plane, i.e.,
= 0 . The
incident electromagnetic elds can be expanded in terms of spherical
vector wave eigenfunctions dened in the spherical coordinates system
as follows [23]:
Me
o
mn
(k) =
mz
n
(kr)
sin
P
m
n
(cos )
sin
cos
m
z
n
(kr)
P
m
n
(cos )
cos
sin
m
, (13a)
Ne
o
mn
(k) =
n(n + 1)z
n
(kr)
kr
P
m
n
(cos )
cos
sin
mr
+
[rz
n
(kr)]
krr
P
m
n
(cos )
cos
sin
m
m
sin
[rz
n
(kr)]
krr
P
m
n
(cos )
sin
cos
m
, (13b)
where z
n
(kr) represents the spherical Bessel functions of order n , and
P
m
n
(cos ) identies the associated Legendre function of the rst kind
with the order (n, m) .
The incident waves under the two polarizations have, as introduced
by Morrison and Cross [49], the following forms:
E
i
I
II
=
n=1
n
m=0
_
P
i
o
e
mn
M
(1)
o
e
mn
(k
0
) + Q
i
e
o
mn
N
(1)
e
o
mn
(k
0
)
_
, (14a)
H
i
I
II
=
ik
0
n=1
n
m=0
_
P
i
o
e
mn
N
(1)
o
e
mn
(k
0
) + Q
i
e
o
mn
M
(1)
e
o
mn
(k
0
)
_
, (14b)
where the spherical Bessel functions of the rst kind, i.e., z
n
(k
0
r) =
j
n
(k
0
r) , are used in the above vector wave functions, the orthogonal
properties of Me
o
mn
(k
0
) and Ne
o
mn
(k
0
) are considered, and the co-
ecients of the expanded incident electromagnetic wave, P
i
e
o
mn
and
Q
i
e
o
mn
, are given by [50, 51]:
P
i
o
e
mn
= i
n
(2
m0
)N
mn
_
_
mP
m
n
(cos )
sin
E
I
P
m
n
(cos )
E
II
_
_
, (15a)
Q
i
e
o
mn
= i
n+1
(2
m0
)N
mn
_
_
P
m
n
(cos )
E
I
mP
m
n
(cos )
sin
E
II
_
_
, (15b)
with
mn
(= 1 for m = n ; and 0 for m = n) denoting the Kronecker
symbol and N
mn
being the normalization coecient given by
N
mn
=
(2n + 1)
n(n + 1)
(n m)!
(n + m)!
. (16)
For simplicity, we make abbreviation as follows:
a =
mP
m
n
(cos )
sin
, (17a)
b =
P
m
n
(cos )
. (17b)
The electric eld in a multilayered chiral sphere in the eigenfunction
expansion form can be written as a superposition of right-handed and
256 Li et al.
left-handed circularly polarized elds. And the right-handed and left-
handed circularly polarized eld modes can be expressed using vector
wave functions as follows [16]:
E
(p)
Rmn
= M
(p)
e
o
mn
(k
(R)
) +N
(p)
e
o
mn
(k
(R)
), (18a)
E
(p)
Lmn
= M
(p)
e
o
mn
(k
(L)
) N
(p)
e
o
mn
(k
(L)
). (18b)
where the superscript p equals 1 or 3 representing the rst type of
spherical Bessel function and the rst kind of spherical Hankel function.
The electric eld and magnetic eld in f-th layer can be expressed
in the following form:
E
(p)
mn,f
= E
(p)
Lmn,f
+E
(p)
Rmn,f
, (19a)
H
(p)
mn,f
=
i
f
_
E
(p)
Lmn,f
+E
(p)
Rmn,f
_
. (19b)
Under the spherical coordinates, the electromagnetic elds usually con-
sist of the radially outgoing- and incoming-propagation wave modes.
Hence, the electric and magnetic elds in the layer f ranging from the
second to the (N 1)-th region are expressed as:
E
f
=
m,n
_
Ce
o
1f
_
M
(3)
e
o
mn
(k
(R)
f
) +N
(3)
e
o
mn
(k
(R)
f
)
_
+ Ce
o
2f
_
M
(3)
e
o
mn
(k
(L)
f
) N
(3)
e
o
mn
(k
(L)
f
)
_
+ Ce
o
3f
_
M
(1)
e
o
mn
(k
(R)
f
) +N
(1)
e
o
mn
(k
(R)
f
)
_
+ Ce
o
4f
_
M
(1)
e
o
mn
(k
(L)
f
) N
(1)
e
o
mn
(k
(L)
f
)
__
, (20a)
H
f
=
i
m,n
_
Ce
o
1f
_
M
(3)
e
o
mn
(k
(R)
f
) +N
(3)
e
o
mn
(k
(R)
f
)
_
Ce
o
2f
_
M
(3)
e
o
mn
(k
(L)
f
) N
(3)
e
o
mn
(k
(L)
f
)
_
+ Ce
o
3f
_
M
(1)
e
o
mn
(k
(R)
f
) +N
(1)
e
o
mn
(k
(R)
f
)
_
Ce
o
4f
_
M
(1)
e
o
mn
(k
(L)
f
) N
(1)
e
o
mn
(k
(L)
f
)
__
. (20b)
While only the inward waves exist in the inner-most layer, the outward
waves exist in the outer-most layer. Therefore, the scattering coe-
cients corresponding to the outgoing waves in the inner-most layer and
to the incoming waves in the outer-most layer must vanish. The elec-
tric eld in the out-most and inner-most layer are written, respectively,
as follows:
E
1
=E
i
+E
s
=E
i
+
m,n
_
Ce
o
11
_
M
(3)
e
o
mn
(k
0
) +N
(3)
e
o
mn
(k
0
)
_
+ Ce
o
21
_
M
(3)
e
o
mn
(k
0
) N
(3)
e
o
mn
(k
0
)
__
, (21a)
E
N
=
m,n
_
Ce
o
3N
_
M
(1)
e
o
mn
(k
(R)
N
) +N
(1)
e
o
mn
(k
(R)
N
)
_
Ce
o
4N
_
M
(1)
e
o
mn
(k
(L)
N
) N
(1)
e
o
mn
(k
(L)
N
)
__
, (21b)
where k
0
is the wave number outside the multilayered sphere, given
by
k
0
=
0
. (22)
3.1.2 Determination of Scattering Coecients by Boundary
Conditions
The electric and magnetic elds satisfy the following boundary con-
ditions at the spherical interfaces r = a
j
(j = 1, 2, , N 1) :
r E
f
= r E
(f+1)
, (23a)
r H
f
= r H
(f+1)
. (23b)
Without any loss of generality of the problem, we extend (23a)(23b)
into a linear equation system. To simplify the complicated algebraic
calculations, let us introduce the following operators
(r,l)
im
=j
n
(k
(r,l)
i
a
m
), (24a)
h
(r,l)
im
=h
(1)
n
(k
(r,l)
i
a
m
), (24b)
(r,l)
im
=
1
d[j
n
()]
d
=k
(r,l)
i
a
m
, (24c)
h
(r,l)
im
=
1
d[h
(1)
n
()]
d
=k
(r,l)
i
a
m
; (24d)
i = 1, 2, , N,
m = i 1 and i.
258 Li et al.
Writing the linear equation system of the coecients in matrix form,
we have the following equation
F
f
C
f
= F
f+1
C
f+1
, (25)
where the parameter and coecient matrices are dened as:
F
f
=
_
_
h
(r)
ff
h
(l)
ff

(r)
ff

(l)
ff
h
(r)
ff
h
(l)
ff

(r)
ff

(l)
ff
1
f
h
(r)
ff

1
f
h
(l)
ff

1
f

(r)
ff

1
f

(l)
ff
1
f
h
(r)
ff

1
f
h
(l)
ff

1
f

(r)
ff

1
f

(l)
ff
_
_
, (26a)
C
f
=
_
_
Ce
o
1f
Ce
o
2f
Ce
o
3f
Ce
o
4f
_
_
, (26b)
F
f+1
=
_
_
h
(r)
(f+1)f
h
(l)
(f+1)f
h
(r)
(f+1)f
h
(l)
(f+1)f
1
(f+1)
h
(r)
(f+1)f

1
(f+1)
h
(l)
(f+1)f
1
(f+1)
h
(r)
(f+1)f

1
(f+1)
h
(l)
(f+1)f
(r)
(f+1)f

(l)
(f+1)f
(r)
(f+1)f

(l)
(f+1)f
1
(f+1)
(r)
(f+1)f

1
(f+1)
(l)
(f+1)f
1
(f+1)
(r)
(f+1)f

1
(f+1)c
(l)
(f+1)f
_
_
, (26c)
Ce
o
f+1
=
_
_
Ce
o
1(f+1)
Ce
o
2(f+1)
Ce
o
3(f+1)
Ce
o
4(f+1)
_
_
, (26d)
with
1
f
=
_
f
, (27a)
1
(f+1)
=
(f+1)
(f+1)
. (27b)
To simplify the derivation of the coecients, we found the inverse of
F
f+1
by using commercially available softwares with symbolic calcu-
lations such as Mathematica.
We can rewrite the linear equation (25) into the following form
C
f+1
= T
f
C
f
(28)
where the transmission matrix in the eigen-expansion domain is given
by:
T
f
= F
1
f+1
F
f
=
_
T
f
j
_
44
. (29)
For convenience and simplicity, let us assume that
T
(k)
=
_
T
(k)
j
_
44
= [T
N1
][T
N2
] [T
k+1
][T
k
]. (30)
According to equation (28), we have the relation
C
N
= [T
N1
T
N2
T
2
T
1
] C
1
(31)
which can be re-written as:
C
N
= T
(1)
C
1
(32)
where
C
N
=
_
_
0
0
Ce
o
3N
Ce
o
4N
_
_
, (33a)
T
(1)
=
_
_
T
(1)
11
T
(1)
12
T
(1)
13
T
(1)
14
T
(1)
21
T
(1)
22
T
(1)
23
T
(1)
24
T
(1)
31
T
(1)
32
T
(1)
33
T
(1)
34
T
(1)
41
T
(1)
42
T
(1)
43
T
(1)
44
_
_
, (33b)
C
1
=
_
_
Ce
o
11
Ce
o
21
Ce
o
31
Ce
o
41
_
_
. (33c)
260 Li et al.
Based on the equations (14) and (20a), we obtain :
C
I
o31
=
P
i
omn
2
, C
I
e31
=
Q
i
emn
2
, (34a)
C
I
o41
=
P
i
omn
2
, C
I
e41
=
Q
i
emn
2
, (34b)
C
II
o31
=
Q
i
omn
2
, C
II
e31
=
P
i
emn
2
, (34c)
C
II
o41
=
Q
i
omn
2
, C
II
e41
=
P
i
emn
2
, (34d)
From Eqs. (29) to (33), we can derive the scattering coecients, Ce
o
11
and Ce
o
12
, as follows:
C
I
e
o
11
=
_
Q
i
emn
P
i
omn
_
T
(1)
13
T
(1)
22
T
(1)
23
T
(1)
12
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
+
_
Q
i
emn
P
i
omn
_
T
(1)
14
T
(1)
22
T
(1)
24
T
(1)
12
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
, (35a)
C
I
e
o
21
=
_
Q
i
emn
P
i
omn
_
T
(1)
11
T
(1)
23
T
(1)
13
T
(1)
21
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
+
_
Q
i
emn
P
i
omn
_
T
(1)
11
T
(1)
24
T
(1)
14
T
(1)
21
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
, (35b)
C
II
e
o
11
=
_
P
i
emn
Q
i
omn
_
T
(1)
13
T
(1)
22
T
(1)
23
T
(1)
12
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
+
_
P
i
emn
Q
i
omn
_
T
(1)
14
T
(1)
22
T
(1)
24
T
(1)
12
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
, (35c)
C
II
e
o
21
=
_
P
i
emn
Q
i
omn
_
T
(1)
11
T
(1)
23
T
(1)
13
T
(1)
21
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
+
_
P
i
emn
Q
i
omn
_
T
(1)
11
T
(1)
24
T
(1)
14
T
(1)
21
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
. (35d)
Hence, the scattered electric eld can be written as follows:
E
s
I
=
n=1
n
m=0
C
I
e
o
M
M
(3)
e
o
mn
(k
0
) + C
I
e
o
N
N
(3)
e
o
mn
(k
0
), (36a)
E
s
II
=
n=1
n
m=0
C
II
e
o
M
M
(3)
e
o
mn
(k
0
) + C
II
e
o
N
N
(3)
e
o
mn
(k
0
). (36b)
where
C
I,II
e
o
M
= C
I,II
e
o
11
+ C
I,II
e
o
21
, (37a)
C
I,II
e
o
N
= C
I,II
e
o
11
C
I,II
e
o
21
. (37b)
So far, we have derived the scattered electric eld using the conven-
tional scattering theory where the vector wave function expansion tech-
nique and boundary condition matching are applied. In the next sec-
tion, we shall use the radiation theory to reconsider this same problem
and compare the results by the dierent techniques.
3.2 The Method by Dyadic Greens Function
The total electric eld E(r) of an electromagnetic wave everywhere
is governed by the Maxwells equations. The vector wave equation can
be expressed as follows if suppressing the time factor e
it
:
E(r) 2
f
E(r) k
2
E(r)
= iJ(r
)(r r
), (38a)
H(r) 2
f
H(r) k
2
H(r)
=
_
J(r
)(r r
, (38b)
where k
2
= ( i) , and , and are the permeability, the
permittivity and the conductivity, respectively. In free space, we have
k
2
0
=
2
0
.
The solution to Eq. (38) may be represented in terms of the integral
involving the electric type of dyadic Greens function as follows [23]:
E(r) = i
___
V
G
e
(r, r
) J(r
)dV
, (39)
262 Li et al.
where E(r) represents the total electric eld and G
e
(r, r
) stands
for the electric type of dyadic Greens functions.
To solve for the unknown scattered eld, the following well-known
integrals of scattered elds excited by the current distribution J(r
)
in free space are recommended:
E
s
(r) = i
___
V
G
es
(r, r
) J(r
)dV
, (40a)
H
s
(r) =
___
V
G
es
(r, r
) J(r
)dV
. (40b)
Dierent from that in (39), the Greens function G
es
(r, r
) used in
(40) identies the scattering dyadic Greens function exclusive of con-
tributions due to direct waves.
3.2.1 Dyadic Greens Function for Multilayered Spherically Chiral
Media
A general expression of spatial-domain dyadic Greens function for
dening the electromagnetic radiation elds in spherically arbitrary
multilayered and chiral media is available [45]. For the case of scat-
tering in the present paper, an equivalent current source is assumed
to be located at innity and to be in the rst region where the eld
is concerned. According to Part A of Section V in [45], the scattering
dyadic Greens function is formulated as follows:
G
(fs)
es
(r, r
)
=
i
2(k
(r)
s
+ k
(l)
s
)
m,n
(2
0
)
(2n + 1)(n m)!
n(n + 1)(n + m)!
_
C
fs
12
V
(3)
e
o
mn
(k
(r)
f
)V
(3)
e
o
mn
(k
(r)
s
) +C
fs
22
W
(3)
e
o
mn
(k
(l)
f
)V
(3)
e
o
mn
(k
(r)
s
)
+ C
fs
14
V
(3)
e
o
mn
(k
(r)
f
)W
(3)
e
o
mn
(k
(l)
s
) +C
fs
24
W
(3)
e
o
mn
(k
(l)
f
)W
(3)
e
o
mn
(k
(l)
s
)
_
,(41)
where the normalized spherical vector wave functions are dened as
V e
o
mn
(k) =
Me
o
mn
(k) +Ne
o
mn
(k)
2
, (42a)
We
o
mn
(k) =
Me
o
mn
(k) Ne
o
mn
(k)
2
. (42b)
x"
y
x
z
r"
y"
z"
"
o
Two coordinate systems
with notations:
(r,,) and (r",",")
o
o"
r
r
o
Figure 1. Coordinate translation.

The outmost layer is assumed to be free space, so the scattering dyadic
Greens function is simplied as follows:
G
(11)
es
(r, r
)
=
i
4k
0
m,n
(2
0
)
(2n + 1)(n m)!
n(n + 1)(n + m)!
_
C
11
12
V
(3)
e
o
mn
(k
0
)V
(3)
e
o
mn
(k
0
) +C
11
22
W
(3)
e
o
mn
(k
0
)V
(3)
e
o
mn
(k
0
)
+ C
11
14
V
(3)
e
o
mn
(k
0
)W
(3)
e
o
mn
(k
0
) +C
11
24
W
(3)
e
o
mn
(k
0
)W
(3)
e
o
mn
(k
0
)
_
, (43)
3.2.2 Source Which Generates Plane Waves at Innity
The coordinates translation is shown in Fig. 1. The elds are ob-
served under the coordinates system (r, , ) and the point source at
innity is located in the coordinates system (r
) . r
0
denotes
the position of the point source with the distance r
0
, at the po-
lar angle = , and at the azimuth angle = 0 (where the sphere is
symmetrical with respect to the z -axis so that a simple azimuth angle
can be chosen). r
represents the position of the dierential element

264 Li et al.
in the source region. It can be seen from the gure that the position
r
0
of the point-source in the eld coordinates, the position r
of a
dierential element in the eld coordinates, and the position r
of the
dierential element in the source coordinates are related by
r
= r
0
+r
. (44)
The source at innity can be regarded as a point source, therefore a
delta function can be used to describe the source. For the dierent
polarizations, the source can be expressed by
J
I
(r
) = E
I
f(r
0
)(r
r
0
)
, (45a)
J
II
(r
) = E
II
f(r
0
)(r
r
0
)
. (45b)
where f(r
0
) to be determined is a function of the distance of the point
source. The plane wave can be considered as the one excited by the
point source that is located at innity (i.e., r
0
) . Since r
0
is
a constant vector, the vectors r
and r
and their derivatives in the

two coordinates systems are related by
dr
= dr
, (46a)
r
r
0
= r
, (46b)
so that Eq. (45) can be rewritten as
J
I
(r
) = E
I
f(r
0
)(r
= E
I
f(r
0
)
(r
)(
)(
)
[r
]
2
sin
, (47a)
J
II
(r
) = E
II
f(r
0
)(r
= E
II
f(r
0
)
(r
)(
)(
)
[r
]
2
sin
. (47b)
Furthermore, we have the equivalent current distributions as follows:
_
J
I
(r
)
J
II
(r
)
_
dr
= f(r
0
)
_
E
I
E
II
_
(r
)(
)(
)dr
. (48)
For simplicity, we dene the following intermediates resulted in from
substituting Eq. (41) and Eq. (48) into Eq. (39):
H
I,II
e
o
=
___
v
V
e
o
mn
(k
0
) J
I,II
(r
)dr
, (49a)
K
I,II
e
o
=
___
v
W
e
o
mn
(k
0
) J
I,II
(r
)dr
. (49b)
Substitution of Eq. (48) into Eq. (39) leads to r
r
0
,
, and
0 in the integral containing the vector eigenfunctions, V
e
o
mn
(k
0
)
and W
e
o
mn
(k
0
) . Thus, we nally obtain the following expressions:
H
I
e
o
=
f(r
0
)
2
E
I

_
_
P
m
n
(cos )
d[(k
0
r
0
)h
(1)
n
(k
0
r
0
)]
(k
0
r
0
)d(k
0
r
0
)
mP
m
n
(cos )
sin
h
(1)
n
(k
0
r
0
)
_
_
, (50a)
H
II
e
o
=
f(r
0
)
2
E
II

_
P
m
n
(cos )
h
(1)
n
(k
0
r
0
)
mP
m
n
(cos )
sin
d[(k
0
r
0
)h
(1)
n
(k
0
r
0
)]
(k
0
r
0
)d(k
0
r
0
)
_
_
, (50b)
K
I
e
o
=
f(r
0
)
2
E
I

_
P
m
n
(cos )
d[(k
0
r
0
)h
(1)
n
(k
0
r
0
)]
(k
0
r
0
)d(k
0
r
0
)
mP
m
n
(cos )
sin
h
(1)
n
(k
0
r
0
)
_
_
, (50c)
K
II
e
o
=
f(r
0
)
2
E
II

_
P
m
n
(cos )
h
(1)
n
(k
0
r
0
)
mP
m
n
(cos )
sin
d[(k
0
r
0
)h
(1)
n
(k
0
r
0
)]
(k
0
r
0
)d(k
0
r
0
)
_
_
, (50d)
where the function f( r
0
) is constructed as in [48]:
f(r
0
) = i
4
0
r
0
e
ik
0
r
0
. (51)
According to Eq. (39), the scattered eld of the multilayered chiral
sphere is written as follows:
E
s
I,II
=i
___
A
G
es
(r, r
) J
I,II
(r
)dS
,
=

8k
0
f(r
0
)
n=1
n
m=0
(2
m0
)N
mn
__
C
11
12
V
(3)
e
o
mn
(k
0
)
+ C
11
22
W
(3)
e
o
mn
(k
0
)
_
H
I,II
e
o
+
_
C
11
14
V
(3)
e
o
mn
(k
0
)
+ C
11
24
W
(3)
e
o
mn
(k
0
)
_
K
I,II
e
o
_
. (52)
266 Li et al.
By substituting the expression of f(r
0
) into Eq. (52) and using the
asymptotic form of large-argument spherical Hankel function given as
follows:
h
(1)
n
(k
0
r
0
) = (i)
n+1
e
ik
0
r
0
k
0
r
0
, (53a)
d[(k
0
r
0
)h
(1)
n
(k
0
r
0
)]
(k
0
r
0
)d(k
0
r
0
)
= (i)
n
e
ik
0
r
0
k
0
r
0
; (53b)
Eq. (52) can thus be further reduced to
E
s
I
II
=
1
2k
2
0
_
E
I
E
II
_

n=1
n
m=0
(i)
n
(2
m0
)N
mn
_
M
(3)
o
e
mn
(k
0
)
_
(C
11
12
+C
11
22
+C
11
14
+C
11
24
)
_
a
b
_
+M
(3)
e
o
mn
(k
0
)
_
(C
11
12
+C
11
22
C
11
14
C
11
24
)
_
ib
ia
_
+N
(3)
o
e
mn
(k
0
)
_
(C
11
12
C
11
22
+C
11
14
C
11
24
)
_
a
b
_
+ N
(3)
e
o
mn
(k
0
)
_
(C
11
12
C
11
22
C
11
14
+C
11
24
)
_
ib
ia
__
(54)
Substituting the expression for scattering coecients [45] into the equa-
tion, we can easily obtain:
E
s
I
II
=
_
P
i
o
e
mn
M
(3)
o
e
mn
(k
0
) +Q
i
e
o
mn
M
(3)
e
o
mn
(k
0
)
_
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
_
T
(1)
13
T
(1)
22
T
(1)
23
T
(1)
12
+T
(1)
11
T
(1)
23
T
(1)
13
T
(1)
21
_
+
_
P
i
o
e
mn
M
(3)
o
e
mn
(k
0
) Q
i
e
o
mn
M
(3)
e
o
mn
(k
0
)
_
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
_
T
(1)
14
T
(1)
22
T
(1)
24
T
(1)
12
+T
(1)
11
T
(1)
24
T
(1)
14
T
(1)
21
_
+
_
P
i
o
e
mn
N
(3)
o
e
mn
(k
0
) +Q
i
e
o
mn
N
(3)
e
o
mn
(k
0
)
_
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
_
T
(1)
13
T
(1)
22
T
(1)
23
T
(1)
12
T
(1)
11
T
(1)
23
+T
(1)
13
T
(1)
21
_
+
_
P
i
o
e
mn
N
(3)
o
e
mn
(k
0
) Q
i
e
o
mn
N
(3)
e
o
mn
(k
0
)
_
2
_
T
(1)
12
T
(1)
21
T
(1)
11
T
(1)
22
_
_
T
(1)
14
T
(1)
22
T
(1)
24
T
(1)
12
T
(1)
11
T
(1)
24
+T
(1)
14
T
(1)
21
_
(55)
with
Po
e
mn
= (i)
n
(2
m0
)N
mn
_
_
mP
m
n
(cos )
sin
E
I
P
m
n
(cos )
E
II
_
_
, (56a)
Qe
o
mn
= (i)
n+1
(2
m0
)N
mn
_
_
P
m
n
(cos )
E
I
mP
m
n
(cos )
sin
E
II
_
_
, (56b)
We note, however, that the direction of propagation of the plane wave
created here is opposite to the one considered by eigenfunction expan-
sion [23]. In order to interpret this problem, we refer to Mie scattering.
According to Stratton [51], the incident wave of Mie scattering is
in positive z-axis direction. Under the time dependence e
it
, the
incident electric eld is expanded as:
E
i
I
=
n=0
_
a
I
n
M
(1)
o1n
(k
0
) + b
I
n
N
(1)
e1n
(k
0
)
_
, (57a)
E
i
II
=
n=0
_
a
II
n
M
(1)
e1n
(k
0
) + b
II
n
N
(1)
o1n
(k
0
)
_
. (57b)
with
_
a
I
n
a
II
n
_
= i
n
2n + 1
n(n + 1)
_
1
1
_
, (58a)
_
b
I
n
b
II
n
_
= i
n+1
2n + 1
n(n + 1)
_
1
1
_
, (58b)
In Eq. (15), we let m = 1 and = 0 . Following Ferrers denition of
associate Legendre function in [51], we easily obtain:
P
1
n
(cos )
sin
=0
=
n(n + 1)
2
, (59a)
P
1
n
(cos )
=0
=
n(n + 1)
2
. (59b)
268 Li et al.
By substituting the above equation, Eq. (15) can be reduced to Eq.
(57).
In the method of dyadic Greens function, we also can obtain the
expansion of incident wave by applying the following expressions:
E(r) = i
___
V
G
0
(r, r
) J(r
)dV
, (60)
G
0
expanded by vector spherical wave function is available in [51]. In
order to match the wave propagation direction, substitution Eq. (48)
into the above equation leads to r
r
0
,
, and
0 in
the integral containing the vector eigenfunctions. When = , we
borrow the following equations:
P
1
n
(cos )
sin
=
= (1)
n
n(n + 1)
2
, (61a)
P
1
n
(cos )
=
= (1)
n
n(n + 1)
2
. (61b)
Apply the expression for source and f(r
0
) and after integration, the
same expression for coecients as Eq. (57) is obtained.
As to the scattered eld of Mie scattering, Stratton obtained the
result by applying method of eigenfunction expansion in [51] (p. 564).
Tai also found the result by using the method of dyadic Greens func-
tion in [23] (p. 217). The dierence between them is dierent wave
propagation direction. To calculate Mie scattering using the method
of dyadic Greens function, if we let
0 , we can obtain the same

result as in [23]. That means the incident wave is in negative z-axis
direction which is dierent from that considered by Stratton [51]. This
can be easily matched by let
and applying Eq. (61) in calcu-

lation [52] (p. 1466).
In the present paper, when applying the method of dyadic Greens
function, we let
which is the angle between the incident wave

direction and positive z-axis.(see Fig. 1). While in the method of
eigenfunction expansion, we still follow the procedure of Stratton [51]
in which the incident angle is between the incident wave direction and
negative z-axis. The dierent wave propagation direction results in
the dierence between two methods. To solve this problem, we only
need to let
+ and apply the following relation:

P
m
n
(cos )
sin( + )
= (1)
nm
P
m
n
(cos )
sin
, (62a)
P
m
n
(cos )
= (1)
nm
P
m
n
(cos )
. (62b)
It should be point out there are two dierent denition of Associated
Legenfre functions. We may write such relation
P
m(H)
n
(cos ) = (1)
m
P
m(F)
n
(cos ) (63)
with P
m(H)
n
represents Hobsons denition of associated Legendre
functions which can be found in some books such as [53] (p. 951) and
in Matheatica software. While P
m(F)
n
denotes associated Legendre
functions without the factor (1)
m
dened by Ferrer which is utilized
by Stratton [51] (p. 401). Hence, the associate Legendre functions in
Eq. (15) is referring to the denition with superscript (F).
After the manipulation, it is easily veried that the coecients of the
spherical vector wave functions M
(3)
omn
(k
0
) , M
(3)
emn
(k
0
) , N
(3)
omn
(k
0
)
and N
(3)
emn
(k
0
) are the same as the C
I,II
e
o
M
and C
I,II
e
o
N
in the Eq. (37).
Hence, the scattered elds due to the multilayer chiral sphere obtained
by the two methods are identical.
From the above procedure, it is seen that both the scattering theory
and the specic radiation theory give the identical results of the scat-
tered electromagnetic elds. In other words, the scattering problem
can be transformed into the radiation problem for which the dyadic
Greens functions have been derived and the current distribution at
innity has been known.
4. SCATTERING BY A MULTILAYERED CHIRAL
CYLINDER
Assume a plane electromagnetic wave is obliquely incident upon the
cylinder which is innitely long. Also two pairs of incident electro-
magnetic waves, i.e., parallel (I) (TM) and perpendicular (II) (TE)
polarizations incident at an arbitrary angle on a cylinder whose center
O is at the origin of the Cartesian coordinates system, are considered.
The plane wave is specied by its propagation vector k with associ-
ated spherical components k , , and . The angle is the angle
between the propagation k and the axis of the cylinder and is the
angle between the projection of the vector k onto the transverse plane
and the x-axis. Let
i
be the angle between the vectors r and k .
270 Li et al.
Then we have
cos
i
= cos cos + sin sin cos( ). (64)
The incident waves of the two polarizations can be expressed by:
E
i
I
=
i
k
(
0
)
1
2
H
i
I
, (65a)
H
i
I
=
kE
I
0
(sin xcos y)e
ik
0
r cos
i
; (65b)
and
E
i
II
= E
II
(sin xcos y)e
ik
0
r cos
i
, (66a)
H
i
II
=
1
ik
(
0
)
1
2
E
i
II
; (66b)
where E
I
and E
II
are the amplitude of the incident electric elds
with parallel and vertical polarizations.
4.1.1 The Eigenfunction Expansion
Since the problem dened is now changed into cylindrical coordi-
nates, the circular cylindrical coordinates system is the most conve-
nient one.
The incident electromagnetic elds can be expanded in terms of
cylindrical vector wave eigenfunctions dened in the cylindrical coor-
dinates system as follows:
Me
o
n
(h) =
_
nZ
n
()
sin
cos
n
dZ
n
()
d
cos
sin
n
_
e
ihz
, (67a)
Ne
o
n
(h) =
1
k
_
ih
dZ
n
()
d
cos
sin
n
inh
Z
n
()
sin
cos
n
+
2
Z
n
()
cos
sin
nz
_
e
ihz
, (67b)
where Z
n
(k) represents the cylindrical Bessel functions of order n
and h is the longitudinal wave number. Also, M
(3)
e
o
n
(h) and
N
(3)
e
o
n
(h) are dened in terms of the cylindrical Hankel function of
the rst kind, H
(1)
n
() .
The incident waves under the two polarizations have, as introduced
by Rao and Barakat [54, 55], the following forms:
E
i
I
=
E
I
k sin
n=0
(2
n0
)i
n
_
cos(n)N
(1)
en
(h
0
, )
+sin(n)N
(1)
on
(h
0
, )
_
, (68a)
H
i
I
=(
0
)
1
2
E
I
ik sin
n=0
(2
n0
)i
n
_
cos(n)M
(1)
en
(h
0
, )
+sin(n)M
(1)
on
(h
0
, )
_
, (68b)
and
E
i
II
=
E
II
ik sin
n=0
(2
n0
)i
n
_
cos(n)M
(1)
em
(h
0
, )
+sin(n)M
(1)
on
(h
0
, )
_
, (69a)
H
i
II
= (
0
)
1
2
E
II
k sin
n=0
(2
n0
)i
n
_
cos(n)N
(1)
en
(h
0
, )
+sin(n)N
(1)
on
(h
0
, )
_
, (69b)
and where the cylindrical Bessel functions of the rst kind, i.e., Z
n
(x)
= J
n
(x) , are used in the above cylindrical vector wave functions. The
tangential wave number = k sin and the longitudinal one is h
0
=
k cos . For generality, combining the above equations, we rewrite
them as the follows:
E
i
I
II
=
n=0
_
Q
i
en
N
(1)
en
(h
0
, ) + Q
i
on
N
(1)
on
(h
0
, )
P
i
en
M
(1)
en
(h
0
, ) + P
i
on
M
(1)
on
(h
0
, )
_
, (70a)
H
i
I
II
=
i
n=0
_
Q
i
en
M
(1)
en
(h
0
, ) + Q
i
on
M
(1)
on
(h
0
, )
P
i
en
N
(1)
en
(h
0
, ) + P
i
on
N
(1)
on
(h
0
, )
_
, (70b)
Rewrite the coecients of the expanded incident electromagnetic elds,
P
i
e
o
n
and Q
i
e
o
n
, as follows:
P
i
e
o
n
= (2
n0
)i
n+1
E
I
k sin
_
cos(n)
sin(n)
_
, (71a)
272 Li et al.
Q
i
e
o
n
= (2
n0
)i
n
E
II
k sin
_
cos(n)
sin(n)
_
. (71b)
Again,
mn
(= 1 for m = n ; and 0 for m = n) here denotes the Kro-
necker symbol. It should be noted that the form of the expansion of
incident wave for cylindrical systems is dierent from that for spherical
systems. The polarization-I and polarization-II incident waves corre-
spond to N and M terms respectively. Furthermore, the notation
o
e
n
and
e
o
n here have a dierent meaning from those in Eq. (14) where the
upper (or lower) notation denotes the parallel (or the perpendicular)
polarization. While in the above equation, it means the summation of
both upper and lower modes should be taken into account.
The vector wave eigenfunction expansion in multilayered chiral
cylinder can be written as a superposition of right-handed and left-
handed circularly polarized elds. Since the coupling between the
even and odd modes exit in vector cylindrical wave functions, both the
even and odd modes should be taken into account. The right-handed
and left-handed circularly polarized elds can be expressed using wave
functions as follows:
E
(p)
Rn
= M
(p)
e
o
n
(h
(R)
) +N
(p)
e
o
n
(h
(R)
) +M
(p)
o
e
n
(h
(R)
) +N
(p)
o
e
n
(h
(R)
), (72a)
E
(p)
Ln
= M
(p)
e
o
n
(h
(L)
) N
(p)
e
o
n
(h
(L)
) +M
(p)
o
e
n
(h
(L)
) N
(p)
o
e
n
(h
(L)
), (72b)
where p equals 1 or 3. The superscript (1) represents the rst type
of cylindrical Bessel function and the superscript (3) denotes the rst
kind of cylindrical Hankel function used in the vector wave functions.
Following the procedure in the scattering by multilayered chiral sphere,
we can expand the electric eld in the layer from 2nd to (N 1)th
layer of the cylinder as follows:
E
f
=
n=0
_
Ae
o
f
_
M
(3)
e
o
n
(h
(R)
f
) +N
(3)
e
o
n
(h
(R)
f
)
_
+Bo
e
f
_
M
(3)
o
e
n
(h
(R)
f
) +N
(3)
o
e
n
(h
(R)
f
)
_
+Ce
o
f
_
M
(3)
e
o
n
(h
(L)
f
) N
(3)
e
o
n
(h
(L)
f
)
_
+Do
e
f
_
M
(3)
o
e
n
(h
(L)
f
) N
(3)
o
e
n
(h
(L)
f
)
_
+A
e
o
f
_
M
(1)
e
o
n
(h
(R)
f
) +N
(1)
e
o
n
(h
(R)
f
)
_
+B
o
e
f
_
M
(1)
o
e
n
(h
(R)
f
) +N
(1)
o
e
n
(h
(R)
f
)
_
+C
e
o
f
_
M
(1)
e
o
n
(h
(L)
f
) N
(1)
e
o
n
(h
(L)
f
)
_
+D
o
e
f
_
M
(1)
o
e
n
(h
(L)
f
) N
(1)
o
e
n
(h
(L)
f
)
__
, (73a)
H
f
=
i
f
[E
f,R
E
f,L
] . (73b)
The electric elds in the out-most and inner-most regions are written
as follows, respectively:
E
1
=E
i
+E
s
=E
i
+
n=0
_
Ae
o
1
_
M
(3)
e
o
n
(h
0
, ) +N
(3)
e
o
n
(h
0
, )
_
+Bo
e
1
_
M
(3)
o
e
n
(h
0
, ) +N
(3)
o
e
n
(h
0
, )
_
+Ce
o
1
_
M
(3)
e
o
n
(h
0
, ) N
(3)
e
o
n
(h
0
, )
_
+Do
e
1
_
M
(3)
o
e
n
(h
0
, ) N
(3)
o
e
n
(h
0
, )
__
, (74a)
E
N
=
n=0
_
A
e
o
N
_
M
(1)
e
o
n
(h
0
, ) +N
(1)
e
o
n
(h
0
, )
_
+B
o
e
N
_
M
(1)
o
e
n
(h
0
, ) +N
(1)
o
e
n
(h
0
, )
_
+C
e
o
N
_
M
(1)
e
o
n
(h
0
, ) N
(1)
e
o
n
(h
0
, )
_
+D
o
e
N
_
M
(1)
o
e
n
(h
0
, ) N
(1)
o
e
n
(h
0
, )
__
, (74b)
4.1.2 Determination of Scattering Coecients by Boundary
Conditions
The electric and magnetic eld satises the following boundary con-
ditions at the cylindrical interfaces = a
j
(where j = 1, 2, , N1) :
E
f
= E
(f+1)
, (75a)
H
f
= H
(f+1)
. (75b)
Without any loss of generality of the problem, we extend Eq. (75) into
a linear equation system and replace it by coecient matrix equation
274 Li et al.
system, we have the following equation:
F
f
C
f
= F
f+1
C
f+1
, (76)
where
[F
jm
] =
_
_
_
U
h
jm
_ _
U
jm
_
_
D
h
jm
_ _
D
jm
_
_
_
, (77a)
[C
f
] =
_
_
Ae
o
f
Bo
e
f
Ce
o
f
Do
e
f
A
e
o
f
B
o
e
f
C
e
o
f
D
o
e
f
_
_
; (77b)
with
_
U
jm
=
_
_
_
P
jm
_
_
Q
jm
_
R
,r
jm
_

_
R
,l
jm
_
_
_
, (78a)
_
D
jm
=
1
f
_
_
S
jm
_
L
jm
(r)
f+
R
,r
jm
_ _
(l)
f+
R
,l
jm
_
_
. (78b)
To simplify the complicated and tedious representation, the following
operators are introduced:
(r,l)
im
=
J
n
()
=
(r,l)
j
a
m
, (79a)
h
(r,l)
im
=
H
(1)
n
()
=
(r,l)
j
a
m
, (79b)
(r,l)
im
=
d[J
n
()]
=
(r,l)
j
a
m
, (79c)
h
(r,l)
im
=
d[H
(1)
n
()]
=
(r,l)
j
a
m
, (79d)
(r,l)
j
=
ih
(r,l)
j
k
(r,l)
j
, (79e)
(r,l)
jm
=
a
m
k
(r,l)
j
_
(r,l)
j
_
3
, (79f)
(r,l)
f
=
_
k
(r,l)
f

f
_
, (79g)
= h, or, (79h)
_
P
jm
=
_
n
(r)
j

jm
(r)
(r)
j

(r)
jm
(r)
j

(r)
jm
n
(r)
j

jm
(r)
_
, (79i)
_
Q
jm
=
_
n
(l)
j

jm
(l)
(l)
j

(l)
jm
(l)
j

(l)
jm
n
(l)
j

jm
(l)
_
, (79j)
_
R
,(r,l)
jm
_
=
_
(r,l)
jm

jm
(l) 0
0
(r,l)
jm

jm
(l)
_
, (79k)
_
S
jm
=
_
n
(r)
j

(r)
jm
(r)
f

(r)
j

(r)
jm
(r)
f
(r)
j

(r)
jm
(r)
f
n
(r)
j

(r)
jm
(r)
f
_
, (79l)
_
L
jm
=
_
n
(l)
j

(l)
jm
(l)
f+

(l)
j

(l)
jm
(l)
f+
(l)
j

(l)
jm
(l)
f+
n
(l)
j

(l)
jm
(l)
f+
_
, (79m)
j = 1, 2, , N,
m = j 1 or j.
It should be pointed out that the symbols in (79a)(79d) look identical
to, but are dened dierently from, those in (24a)(24d). Following the
same procedure in the problem of scattering by a multilayered chiral
sphere, we obtain:
C
N
= [T
N1
T
N2
T
2
T
1
] C
1
(80)
which can be written as:
C
N
= T
(1)
C
1
(81)
where
T
(k)
=
_
T
(k)
j
_
88
= [T
N1
][T
N2
] [T
k+1
][T
k
], (82)
with the transmission matrix in the eigen-expansion domain is dened
by:
T
f
= F
1
f+1
F
f
=
_
T
f
j
_
88
. (83)
276 Li et al.
We now rewrite Eq. (81) as:
C
N
=
_
_
0
0
0
0
A
e
o
N
B
o
e
N
C
e
o
N
D
e
o
N
_
_
, (84a)
T
(1)
= [T
N1
][T
N2
] [T
2
][T
1
]
=
_
_
_
T
(1)
(11)
_
44
_
T
(1)
(12)
_
44
_
T
(1)
(21)
_
44
_
T
(1)
(22)
_
44
_
_
, (84b)
C
1
=
_
_
Ae
o
1
Bo
e
1
Ce
o
1
Do
e
1
A
e
o
1
B
o
e
1
C
e
o
1
D
o
e
1
_
_
. (84c)
After some simple matrix algebraic manipulations, we easily obtain the
following matrix equation:
_
_
Ae
o
1
Bo
e
1
Ce
o
1
Do
e
1
_
_
=
_
T
(1)
11
_
1
44
_
T
(1)
11
_
44
_
A
e
o
1
B
o
e
1
C
e
o
1
D
o
e
1
_
_
. (85)
According to Eq. (73a), we obtain that B
o
e
1
and D
o
e
1
both equal 0.
Thus, we further assume:
_
X
(1)
_
44
=
_
T
(1)
11
_
1
44
_
T
(1)
11
_
44
=
_
_
X
(1)
11
X
(1)
12
X
(1)
13
X
(1)
14
X
(1)
21
X
(1)
22
X
(1)
23
X
(1)
24
X
(1)
31
X
(1)
32
X
(1)
33
X
(1)
34
X
(1)
41
X
(1)
42
X
(1)
43
X
(1)
44
_
_
. (86)
Finally we obtain:
_
_
A
I
e
o
1
B
I
o
e
1
C
I
e
o
1
D
I
o
e
1
_
_
=
_
_
X
(1)
11
X
(1)
12
X
(1)
13
X
(1)
14
X
(1)
21
X
(1)
22
X
(1)
23
X
(1)
24
X
(1)
31
X
(1)
32
X
(1)
33
X
(1)
34
X
(1)
41
X
(1)
42
X
(1)
43
X
(1)
44
_
_
_
_
Q
i
e
o
n
2
0
Q
i
e
o
n
2
0
_
_
, (87a)
_
_
A
II
e
o
1
B
II
o
e
1
C
II
e
o
1
D
II
o
e
1
_
_
=
_
_
X
(1)
11
X
(1)
12
X
(1)
13
X
(1)
14
X
(1)
21
X
(1)
22
X
(1)
23
X
(1)
24
X
(1)
31
X
(1)
32
X
(1)
33
X
(1)
34
X
(1)
41
X
(1)
42
X
(1)
43
X
(1)
44
_
_
_
_
P
i
e
o
n
2
0
P
i
e
o
n
2
0
_
_
. (87b)
Therefore, the scattering coecients are derived. The scattered eld
can then be written as follows:
E
s
I
=
n=0
_
_
_
_
_
_
X
(1)
11
X
(1)
13
+X
(1)
31
X
(1)
33
_ Q
i
e
o
n
2
_
_
M
(3)
e
o
n
(h
0
, )
+
_
_
_
X
(1)
21
X
(1)
23
+X
(1)
41
X
(1)
43
_ Q
i
e
o
n
2
_
_
M
(3)
o
e
n
(h
0
, )
+
_
_
_
X
(1)
11
X
(1)
13
X
(1)
31
+X
(1)
33
_ Q
i
e
o
n
2
_
_
N
(3)
e
o
n
(h
0
, )
+
_
_
_
X
(1)
21
X
(1)
23
X
(1)
41
+X
(1)
43
_ Q
i
e
o
n
2
_
_
N
(3)
o
e
n
(h
0
, )
_
_
_
, (88a)
E
s
II
=
n=0
_
_
_
_
_
_
X
(1)
11
+X
(1)
13
+X
(1)
31
+X
(1)
33
_ P
i
e
o
n
2
_
_
M
(3)
e
o
n
(h
0
, )
278 Li et al.
+
_
_
_
X
(1)
21
+X
(1)
23
+X
(1)
41
+X
(1)
43
_ P
i
e
o
n
2
_
_
M
(3)
o
e
n
(h
0
, )
+
_
_
_
X
(1)
11
+X
(1)
13
X
(1)
31
X
(1)
33
_ P
i
e
o
n
2
_
_
N
(3)
e
o
n
(h
0
, )
+
_
_
_
X
(1)
21
+X
(1)
23
X
(1)
41
X
(1)
43
_ P
i
e
o
n
2
_
_
N
(3)
o
e
n
(h
0
, )
_
_
_
. (88b)
According to the dyadic Greens function method in Section 3.2,
the scattered eld can be obtained by solving the following well-known
integrals consisting of the dyadic Greens function G
es
(r, r
) and the
current distribution J(r
) in free space:
E
s
(r) = i
___
V
G
es
(r, r
) J(r
)dV
, (89a)
H
s
(r) =
___
V
G
es
(r, r
) J(r
)dV
. (89b)
4.2.1 Dyadic Greens Function for Multilayered Cylindrically Chiral
Media
The dyadic Greens function for a cylindrically arbitrary multilay-
ered medium was constructed by Yin and Wang [46] and the gener-
alized coecients of the dyadic Greens function were provided by Li
et al. [47]. For the present case, a current source is assumed to be
located at innity and to be in the rst region where the eld is of
our interest. According to Part A of Section IV in [47], the scattering
dyadic Greens function is formulated as follows:
G
(11)
es
(r, r
)
=
i
4(k
(r)
s
+ k
(l)
s
)
_

dh(2
0n
)
_
V
(3)
e
o
n
(r)
f
(h)
_
C
11
12
V
(3)
e
o
n
(r)
s
(h) +C
11
14
W
(3)
e
o
n
(l)
s
(h)
_
+V
(3)
o
e
n
(r)
f
(h)
_
C
11
12
V
(3)
e
o
n
(r)
s
(h) +C
11
14
W
(3)
e
o
n
(l)
s
(h)
_
+W
(3)
e
o
n
(l)
f
(h)
_
C
11
22
V
(3)
e
o
n
(r)
s
(h) +C
11
24
W
(3)
e
o
n
(l)
s
(h)
_
+ W
(3)
o
e
n
(l)
f
(h)
_
C
11
22
V
(3)
e
o
n
(r)
s
(h) +C
11
24
W
(3)
e
o
n
(l)
s
(h)
__
, (90)
where the same denitions as in (42) but for dierent (cylindrical)
vector wave functions are given below:
Ve
o
n
(h) =
Me
o
n
(h) +Ne
o
n
(h)
2
, (91a)
We
o
n
(h) =
Me
o
n
(h) Ne
o
n
(h)
2
. (91b)
The outmost region is assumed to be free space, the scattering dyadic
Greens function is simplied as:
G
(11)
es
(r, r
)
=
i
8k
0
_

dh(2
0n
)
_
V
(3)
e
o
n
0
(h
0
)
_
C
11
12
V
(3)
e
o
n
0
(h
0
) +C
11
14
W
(3)
e
o
n
0
(h
0
)
_
+V
(3)
o
e
n
0
(h
0
)
_
C
11
12
V
(3)
e
o
n
0
(h
0
) +C
11
14
W
(3)
e
o
n
0
(h
0
)
_
+W
(3)
e
o
n
0
(h
0
)
_
C
11
22
V
(3)
e
o
n
0
(h
0
) +C
11
24
W
(3)
e
o
n
0
(h
0
)
_
+ W
(3)
o
e
n
0
(h
0
)
_
C
11
22
V
(3)
e
o
n
0
(h
0
) +C
11
24
W
(3)
e
o
n
0
(h
0
)
__
. (92)
4.2.2 Source Which Generates a Plane Wave at Innity
Following the procedure in [48], we can derive the current distri-
bution for the source at innity by applying the cylindrical systems.
It has the same form as in Eq. (47). For dierent polarizations, the
sources can be expressed by:
_
J
I
(r
)
J
II
(r
)
_
= f(r
0
)
_
E
I
E
II
_
(r
)(
)(
). (93)
280 Li et al.
Substituting Eq. (92) and Eq. (93) into Eq. (39), we can nd the ra-
diated eld. First, we introduce the approximation for large argument
for the Hankel function in vector cylindrical wave function, that is:
H
(1)
n
(r) =
_
2
r
_1
2
(i)
n+
1
2
e
ir
. (94)
The functions M
(3)
e
o
n
(h) and N
(3)
e
o
n
(h) , therefore, become:
M
(3)
e
o
n
(h) = (i)
n+
3
2
(
2
r
)
1
2
e
i(r
hz
)
cos
sin
(n)
, (95a)
N
(3)
e
o
n
(h) = (i)
n+
1
2
k
(
2
r
)
1
2
e
i(r
hz
)
cos
sin
(n)(hr + z). (95b)
We now change the cylindrical variables into spherical variables. In
order to obtain the same propagation direction wave, we let
= k sin( + ), (96a)
h = k cos( + ), (96b)
r
= Rsin , (96c)
z
= Rcos . (96d)
Hence, we have
r
hz
= kR
cos[ ( + )], (97a)

hr + z = k(r cos( + ) z sin( + )) = k
, (97b)
Substitution of Eq. (95) into Eq. (92) and Eq. (39) leads to R
r
0
,
+ , and
+ in the integral containing the vector

wave eigenfunctions V
e
o
n
(h) and W
e
o
n
(h) . According to [47],
the coecients can be expressed as:
C
11
12
=
k
0
[
0
]
2
X
(1)
11
, (98a)
C
11
14
=
k
0
[
0
]
2
X
(1)
13
, (98b)
C
11
12
=
k
0
[
0
]
2
X
(1)
21
, (98c)
C
11
14
=
k
0
[
0
]
2
X
(1)
23
, (98d)
C
11
22
=
k
0
[
0
]
2
X
(1)
31
, (98e)
C
11
24
=
k
0
[
0
]
2
X
(1)
33
, (98f)
C
11
22
=
k
0
[
0
]
2
X
(1)
41
, (98g)
C
11
24
=
k
0
[
0
]
2
X
(1)
43
. (98h)
Substituting the coecients and f(r
0
) into Eq. (88). Meanwhile notice
that:
cos(n( + )) = (1)
n
cos(n), (99a)
sin(n( + )) = (1)
n
sin(n). (99b)
By using the method of saddle-point integration [23], we obtain the
scattered elds as follows:
E
s
I
=
E
I
2k
0
sin
n=0
_
cos(n)
sin(n)
_
(2
0n
)i
n
__
X
(1)
11
X
(1)
13
X
(1)
31
+X
(1)
33
_
N
(3)
e
o
n
(k
0
cos )
+
_
X
(1)
11
X
(1)
13
+X
(1)
31
X
(1)
33
_
M
(3)
e
o
n
(k
0
cos )
+
_
X
(1)
21
X
(1)
23
X
(1)
41
+X
(1)
43
_
N
(3)
o
e
n
(k
0
cos )
+
_
X
(1)
21
X
(1)
23
+X
(1)
41
X
(1)
33
_
M
(3)
o
e
n
(k
0
cos )
_
, (100a)
E
s
II
=
E
II
2k
0
sin
n=0
_
cos(n)
sin(n)
_
(2
0n
)i
n+1
__
X
(1)
11
+X
(1)
13
+X
(1)
31
+X
(1)
33
_
M
(3)
e
o
n
(k
0
cos )
+
_
X
(1)
11
+X
(1)
13
X
(1)
31
X
(1)
33
_
N
(3)
e
o
n
(k
0
cos )
282 Li et al.
+
_
X
(1)
21
+X
(1)
23
+X
(1)
41
+X
(1)
43
_
M
(3)
o
e
n
(k
0
cos )
+
_
X
(1)
21
+X
(1)
23
X
(1)
41
X
(1)
43
_
N
(3)
o
e
n
(k
0
cos )
_
. (100b)
It is observed that the above result is the same as that obtained by the
method of eigenfunction expansion in Eq. (88).
5. SCATTERING BY PLANARLY STRATIFIED
CHIRAL MEDIA
Spectral-domain dyadic Greens function in layered chiral media has
been well documented by Ali [38]. For simplicity, the main procedure
and expressions are omitted here. Following the notations of Ali [38],
the eld in region n can be written as:
E
n
(z
n
)
= f
(r)
n
e
r
[
(r)
n
] exp[i
(r)
n
z
n
] + f
(l)
n
e
l
[
(l)
n
] exp[i
(l)
n
z
n
]
+ g
(r)
n
e
r
[
(r)
n
]exp[i
(r)
n
z
n
] + g
(l)
n
e
l
[
(l)
n
]exp[i
(l)
n
z
n
] (101)
which can be written in a matrix form as:
E
n
(z
n
) =
_
e
r
(
(r)
n
) e
l
(
(l)
n
)
_
A
n
(z
n
)
+
_
e
r
(
(r)
n
) e
l
[
(l)
n
]
_
B
n
(z
n
) (102)
where
e
r
(
(r)
n
) = i
h(
(r)
n
) + v(
(r)
n
), (103a)
e
l
(
(l)
n
) = i
h(
(l)
n
) v(
(l)
n
); (103b)
with
h(
(r,l)
n
) =
1
k
s
( xk
y
yk
s
), (104a)
v(
(r,l)
n
) =
1
k
r,l
n
[
r,l
n

K
s
+ k
s
z], (104b)
K
s
=
( xk
x
+ yk
y
)
k
s
; (104c)
and A
n
(z
n
) and B
n
(z
n
) are, respectively, the column matrices repre-
senting the amplitudes of the upgoing and the downgoing waves dened
by
A
n
(z
n
) =
_
A
(r)
n
A
(l)
n
_
=

G
n
(+z
n
) f
n
, (105a)
B
n
(z
n
) =
_
B
(r)
n
B
(l)
n
_
=

G
n
(z
n
) g
n
; (105b)
with
G
n
(z
n
) =
_
exp[i
(r)
n
z
n
] 0
0 exp[i
(l)
n
z
n
]
_
, (105c)
f
n
=
_
f
(r)
n
f
(l)
n
_
, (105d)
g
n
=
_
g
(r)
n
g
(l)
n
_
. (105e)
In Eq. (105a), A
(r)
n
and A
(l)
n
are, respectively, the amplitudes of the
upgoing waves of the right- and the left-handed polarizations while
B
(r)
n
and B
(l)
n
are those of the down-going waves of the right- and
the left-handed polarizations. In present case, the incident waves exist
in the upmost layer which is assumed to be free space, thus only the
downward global reection and the downward global transmission are
present. Utilizing the notation

R
n
in [38], we easily obtain:
A
1
=

R
n
B
1
=

G
1
(+z
1
) f
1
, (106)
Thus, the scattered eld is:
E
s
= [e
r
(
0
) e
l
(
0
)]

R
n

G
1
(+z
1
) g
1
(107)
where
0
is the longitudinal wave number in the upmost layer. We
can easily nd:
g
1
=
_
E
0
E
0
_
(108)
for TE waves; and
g
1
=
_
E
0
E
0
_
(109)
284 Li et al.
for TM waves with E
0
as the amplitude of the indent wave.
According to [38],
u
()
22
=

R
n
=
_
u
()
11
u
()
12
u
()
21
u
()
22
_
(110)
We can therefore write the scattered eld as:
E
s
I
=
_
u
()
11
u
()
12
_
E
0
e
r
(
0
) exp(i
0
z)
+
_
u
()
21
u
()
22
_
E
0
e
l
(
0
) exp(i
0
z) (111)
for the rst polarized (TM) incident waves; and
E
s
II
=
_
u
()
11
+ u
()
12
_
E
0
e
r
(
0
) exp(i
0
z)
+
_
u
()
21
+ u
()
22
_
E
0
e
l
(
0
) exp(i
0
z) (112)
for the second polarized (TE) incident waves.
The technique of fast Fourier transform (FTT) [56] suggests that
an alternative representation of the Green functions for planar strat-
ied media is to cast the eigenfunction expansion in the form of two-
dimensional Fourier transform. Tai [23] has used the free-space dyadic
Greens function to illustrate the formulation. The desirable vector
wave functions to used to represent the free-space dyadic Green func-
tions are dened by:
M(k) =
_
e
ikr
z
_
= i(k
2
x k
1
y)e
ikr
, (113a)
N(k) =
1
k
M(k)
=
1
k
_
k
1
k
3
x k
2
k
3
y + (k
2
1
+ k
2
2
)z
e
ikr
; (113b)
where
k = k
1
x + k
2
y + k
3
z, (114a)
k =
_
k
2
1
+ k
2
2
+ k
2
3
, (114b)
r = x x + y y + zz. (114c)
Ali [38] has derived the dyadic Greens function for a chiral medium
and a planarly layered chiral media in Cartesian coordinates. Ren
[57] also obtained the dyadic Greens function for a chiral medium in
Cartesian coordinates written by aforementioned wave function. By
using the integral representations of wave functions [51], the dyadic
Greens function expanded by vector cylindrical wave function for pla-
narly stratied chiral medium is obtained [57]. In our case, however,
the scattering dyadic Greens functions can be further reduced to:
G
es
(r, r
)
=
i
2
_

0
d
n=0
(2
n0
)
_
1
(0)
R
e
o
n
(
(0)
)
_
u
()
11
Re
o
n
(
(0)
) + u
()
21
Le
o
n
(
(0)
)
_
+
1
(0)
L
e
o
n
(
(0)
)
_
u
()
12
Re
o
n
(
(0)
) + u
()
22
Le
o
n
(
(0)
)
_
_
. (115)
The same two kinds of sources located at innity used in the prob-
lem of electromagnetic scattering by cylindrical multilayered chiral are
employed here to calculate the radiated eld. By applying the saddle
point method and following the same procedure in [23] (p. 213), the
radiated eld can be written as:
E
s
I
=
E
0
k
0
sin
n=0
_
cos(n)
sin(n)
_
(2
0n
)i
n
__
u
()
11
+ u
()
21
u
()
12
u
()
22
_
M
(3)
e
o
n
(k cos )
+
_
u
()
11
u
()
21
u
()
12
+ u
()
22
_
N
(3)
e
o
n
(k cos )
_
, (116a)
E
s
II
=
E
0
k
0
sin
n=0
_
cos(n)
sin(n)
_
(2
0n
)i
n+1
__
u
()
11
+ u
()
21
+ u
()
12
+ u
()
22
_
M
(3)
e
o
n
(k cos )
+
_
u
()
11
u
()
21
+ u
()
12
u
()
22
_
N
(3)
e
o
n
(k cos )
_
. (116b)
We can easily nd Eq. (104) satisfy the following relation:
i
he
iz
=
M(k)
k
s
, (117a)
286 Li et al.
ve
iz
=
N(k)
k
s
, (117b)
where
k
s
= k
0
sin (118)
with the M(k) and N(k) dened in Eq. (113). It should be noted
that the scattering coecients obtained by the two methods are same.
In 1981, Kong [58] expressed the electric eld in each layer of multilay-
ered isotropic media in term of an integration of Hankel functions. In
1994, Li et al. [42] developed the dyadic Green functions for planarly
stratied media expanded in cylindrical wave function. In the present
paper, the vector cylindrical wave function is used to present the prob-
lem of scattering by plannarly stratied chiral medium. According to
the expansion [54]:
e
ikr
=
n=0
(2
n0
)i
n
J
n
(k sin ) cos[n( )]e
ikz sin
(119)
and the denition of vector cylindrical wave function and vector Carte-
sian wave function, the relation between them can be easily found.
Thus, the results obtained from the two methods are identical.
6. CONCLUSIONS
In this paper, the electromagnetic wave scattering by a multilayered
chiral sphere, a multilayered chiral cylinder and a planarly stratied
chiral structure have been studied comparatively. Vector wave function
expansion technique is applied in the comparative analysis. Boundary
conditions are matched and a series of recursive transmission matrix
equations is obtained and so are the scattering coecients. Carte-
sian coordinates system, cylindrical coordinates system and spherical
coordinates system are applied respectively to study the scattering
problem; and correspondingly dierent vector wave functions are em-
ployed. Furthermore, the dyadic Greens functions for multilayered
chiral media of dierent structures are utilized to calculate the radi-
ated elds. The plane waves of perpendicular and parallel polarizations
are equivalent to two dierent point sources located at innity. The
electromagnetic elds are formulated in terms of integrals consisting
a volumetric current distribution and a dyadic Greens function. It
is thus found from the comparative studies that the scattering prob-
lem can be considered as a specic radiation problem where the point
radiated source is located at innity. The principle used behind is
very straightforward, elegant, and quite standard especially when the
dyadic Greens functions are known.
REFERENCES
1. Jaggard, D. L., A. R. Mickelson, and C. H. Papas, On electro-
magnetic waves in chiral media, Appl. Phys., Vol. 18, 211216,
1978.
2. Engheta, N. and D. L. Jaggard, Electromagnetic chirality and
its applications, IEEE Antennas Propagat. Soc. Newslett., Vol.
AP-30, 612, 1988.
3. Lakhtakia, A., V. K. Varadan, and V. V. Varadan, Field equa-
tions, huygenss principle, integral equations, and theorems for
radiation and scattering of electromagnetic waves in isotropic chi-
ral media, J. Opt. Soc. Am. A, Vol. 5, 175184, 1988.
4. Jaggard, D. L., X. Sun, and N. Engheta, Canonical sources and
duality in chiral media, IEEE Trans. Antennas Propag., Vol. 36,
10071013, 1988.
5. Lakhtakia, A., V. K. Varadan, and V. V. Varadan, Eds., Time-
Harmonic Electromagnetic Fields in Chiral Media, Springer,
Berlin, 1989.
6. Bassiri, S., Ed., Electromagnetic Waves in Chiral Media. In Re-
cent Advances in Electromagnetic Theory, H. N. Kritikos and D.
L. Jaggard, eds., Springer-Verlag, New York, 1990.
7. Engheta,N., Special issue on wave interaction with chiral and
complex media, J. Electromag. Waves Applic., Vol. 5/6, 537
793, 1992.
8. Lindell, I. V., A. H. Sihvola, S. A. Tretyakov, and A. J. Viitanen,
Electromagnetic Waves in Chiral and Bi-Isotropic Media, Artech
House, MA, 1994.
9. Bohren, C. F., Light scattering by an optically active sphere,
Chem. Phys. Lett., Vol. 29, 458462, 1974.
10. Lakhtakia, A., V. K. Varadan, and V. V. Varadan, Scatter-
ing and absorption characteristics of lossy dielectric, chiral,non-
spherical objects, Appl. Opt., Vol. 24, 41464154, 1985.
11. Lindell, I. V. and A. H. Sihvola, Quasi-static analysis of scat-
tering from a chiral sphere, J. Electromag. Waves Applic., Vol.
4(12), 12231231, 1990.
288 Li et al.
12. Rojas, R. G., Integral equations for the scattering by a three
dimensional inhomogeneous chiral body, J. Electromagn. Waves
Applic., Vol. 6(5/6), 733750, 1992.
13. Li, L. W., Y. Dan, and M. S. Leong, Electromagnetic waves
scattering by inhomogeneous and multilayered chiral media, in
Progress in Electromagnetics Research Symposium 1999, Taipei,
Taiwan, Mar. 1999.
14. Bohren, C. F., Scattering of electromagnetic waves by an opti-
cally active cylinder, J. Colloid Interface Sci., Vol. 66, 101109,
1974.
15. Kluskens, M. S. and E. H. Newman, Scattering by a chiral cylin-
der of arbitrary cross section, IEEE Trans. Antennas Propag.,
Vol. 39, 9196, 1991.
16. Kluskens, M. S. and E. H. Newman, Scattering by a multilayer
chiral cylinder, IEEE Trans. Antennas Propag., Vol. 39, 9196,
1991.
17. Chen, Z.-N., W. Hong, and W.-X. Zhang, Electromagnetic scat-
tering from a chiral cylinder-general case, IEEE Trans. Anten-
nas Propag., Vol. 44(7), 912917, 1996.
18. Bohren, C. F., Scattering of electromagnetic waves by an opti-
cally active spherical shell, J. Chem. Phys., Vol. 62, 15661571,
1975.
19. Cooray, M. F. and I. R. Ciric, Wave scattering by a chiral
spheroid, J. Opt. Soc. Am. A, Vol. 10(6), 11971203, June 1993.
20. Hamid, A. K., I. R. Ciric, and M. Hamid, Iterative solution of
the scattering by an arbitrary conguration of conducting of di-
electric spheres, IEE Proceedings-H, Vol. 138(6), 565572, 1991.
21. Waterman, P. C., Scattering by dielectric obstacles, Alta Freq.,
Vol. 38, 348352, 1969.
22. Barber, P. W. and C. Yeh, Scattering of electromagnetic waves
by arbitrarily shaped bodies, Appl. Opt., Vol. 14, 28642872,
1975.
23. Tai, C. T., Dyadic Greens Functions in Electromagnetic Theory,
the 2nd edition, IEEE press, New York, 1994.
24. Kong, J. A., Electromagnetic Wave Theory, the 3rd edition, John
Wiley & Sons, New York, 1990.
25. Collin, R. E., Ed., Field Theory of Guided Waves, IEEE Press,
New Jesey, 1991.
26. Chew, W. C., Waves and Fields in Inhomogeneous Media, Van
Nostrand, New York, 1994.
27. Pelet, P. and N. Enghtea, The theory of chirowaveguides, IEEE
Antennas Propagat., Vol. AP-38(1), 9098, 1990.
28. Graglia, R. D., P. L. E. Uslenghi, and C. L. Yu, Electromag-
netic oblique scattering by a cylinder coated with chiral layers
and anisotropic jump-immittance sheets, J. Electromagn. Waves
Applic., Vol. 6(5/6), 695719, 1992.
29. Kluskens, M. S. and E. H. Newman, Scattering by a chiral cylin-
der of arbitrary cross section in the presence of a half-plane, J.
Electromagn. Waves Applic., Vol. 6(5/6), 721731, 1992.
30. Bassiri, S., N. Engheta, and C. H. Papas, Dyadic Greens func-
tion and dipole radiation in chiral media, Alta Freq., Vol. LV(2),
8388, 1986.
31. Weiglhofer, W. S., A simple and straightforward derivation of
the dyadic Greens function of an isotropic chiral media, Archiv
f ur Electronik und

Ubertragungstechnik., Vol. 43(1), 427429,
1989.
32. Lindell, I. V., Simple derivation of various green dyadics for
chiral media, Archiv f ur Elektronik und

Ubertragungstechnik,
Vol. 44(5), 427429, 1990.
33. Weiglhofer, W. S., Analytic method and free-space dyadic
Greens functions, Radio Sci., Vol. 28(5), 847857, 1993.
34. Lakhtakia, A., Alternative derivation of the innite-medium
dyadic Greens function for isotropic chiral media, AEU, Archiv
f ur Elektronik und

Ubertragungstechnik: Electronics and Com-
munication, Vol. 45(5), 323324, Sep. 1991.
35. Lindell, I. V., Methods of Electromagnetic Field Analysis, Claren-
don Press, Oxford, 1992.
36. Engheta, N. and M. W. Kowarz, Antenna radiation in the pres-
ence of a chiral sphere, J. Appl. Phys., Vol. 67(2), 10571065,
1990.
37. Lakhtakia, A., Dyadic Greens functions for an isotropic chiral
half-space bounded by a perfectly conducting plane, Interna-
tional Journal of Electronics, Vol. 71(1), 139144, 1991.
38. Ali, S. M., T. M. Habathy, and J. A. Kong, Spectral-domain
dyadic Greens functions in layered chiral media, J. Opt. Soc.
Am. A, Vol. 9(3), 413423, March 1992.
39. Vegni Toscano, A. L., Spectral dyadic Greens function formula-
tion for planar integrated structures with a grounded chiral slab,
J. Electromagn. Waves Applic., Vol. 6, 751769, 1992.
40. Yin, W.-Y. and W.-B. Wang , Analyzing the radiated char-
acteristics of a two-layered chirostrip dipole antenna using the
dyadic Greens function, Microwave and Optical Technology Let-
ters, Vol. 6(4), 221223, March 1993.
290 Li et al.
41. A. Lakhtakia, Stratied planar chiral shields, in IEEE Interna-
tional Symposium on Electromagnetic Compatibility 1993, Dallas,
TX, USA, 1993.
42. Li, L. W., P. S. Kooi, M. S. Leong, and T. S. Yeo, On the
eigenfunctions of dyadic Greens function in planarly stratied
media, J. Electromagn. Waves Applic., Vol. 8(6), 663678, 1994.
43. Yin, W.-Y. and P. Li, Theory of multilayered spherical chi-
rodome, Microwave and Optical Technology Letters, Vol. 6(3),
179182, March 1993.
44. Yin, W.-Y. and P. Li, The focal points of radially stratied
spherical chirolenses, Microwave and Optical Technology Letters,
Vol. 6(9), 561564, 1993.
45. Li, L.-W., P.-S. Kooi, M.-S. Leong, and T.-S. Yeo, A general
expression of dyadic Greens functions in radially multilayered
chiral media, IEEE Trans. Antennas Propag., Vol. 43(3), 232
238, 1995.
46. Yin, W.-Y. and P. Li, Dyadic Greens function of cylindrical
multilayered chiral media and its applications, J. Electromagn.
Waves Applic., Vol. 7(7), 10051027, 1993.
47. Li, L. W., P. S. Kooi, M. S. Leong, and T. S. Yeo, Analytic rep-
resentation of scattering dyadic Greens functions coecients for
cylindrically multilayered chiral media, J. Electromagn. Waves
Applic., Vol. 9(9), 12071201, 1995.
48. Li, L. W. T. S. Yeo, P. S. Kooi, and M. S. Leong, An ecient
and simple approximation to microwave specic attention by dis-
torted raindrops, IEEE Trans. Antennas Propag., submitted in
1999.
49. Morrison, J. A. and M. J. Cross, Scattering of a plane electro-
magnetic wave by axisymmetric raindrops, Bell Syst. Tech. J.,
Vol. 53, 9551019, 1974.
50. Harrington, R. F., Ed., Time-Harmonic Electromagnetic Fields,
Mac-Graw Hill, 1961.
51. Stratton, J. A., Electromagnetic Theory, McGraw-Hill, New
York, 1972.
52. Morse, P. M. and H. Feshbach, Eds., Methods of theoretical
physics, McGraw-Hill, New York, 1953.
53. Balanis, C. A., Advanced engineering electromagnetics, John Wi-
ley & Sons, 1989.
54. Rao, T. C. and R. Barakat, Plane-wave scattering by a conduct-
ing cylinder partially buried in a conducting plane. 1. tm case,
J. Opt. Soc. Am. A, Vol. 6(9), 12701280, September 1989.
55. Rao, T. C. and R. Barakat, Plane-wave scattering by a conduct-
ing cylinder partially buried in a conducting plane. 2. te case,
J. Opt. Soc. Am. A, Vol. 8(12), 19861990, December 1991.
56. Nussbaumer, H. J., Ed., Fast Fourier Transform and Convolution
Algorithms, Springer Verlag, New York, 1982.
57. Ren, W., Dyadic Greens functions and dipole radiations in lay-
ered chiral media, J. Appl. Phys., Vol. 75(1), 3035, January 1,
1994.
58. Kong, J. A., Ed., Research Topics in Electromagnetic Waves The-
ory, Wiley, New York, 1981.

Li LYK

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Li LYK

Caricato da

Copyright:

Formati disponibili

Progress In Electromagnetics Research, PIER 26, 249291, 2000

ELECTROMAGNETIC SCATTERING BY MULTILAYERED

Figure 1. Coordinate translation.

represents the position of the dierential element

and their derivatives in the

0 in the integral containing the vector eigenfunctions, V

0 , we can obtain the same

and applying Eq. (61) in calcu-

which is the angle between the incident wave

+ and apply the following relation:

cos[ ( + )], (97a)

+ in the integral containing the vector

Potrebbero piacerti anche