3.2.

Uniaxial and Biaxial Media 61

3 E = x̂ Ex + ŷ Ey = ê+ E+ + ê− E− , where E± =

1
2
(Ex ± jEy ) (3.1.1)

The circular components E+ and E− represent right and left polarizations (in the
Propagation in Birefringent Media IEEE convention) if the wave is moving in the positive z-direction, but left and right if it
is moving in the negative z-direction.
Because the propagation medium is not isotropic, we need to start with the source-
free Maxwell’s equations before we assume any particular constitutive relationships:

∇ × E = −jωB , ∇ × H = jωD , ∇ · D = 0, ∇·B=0 (3.1.2)

For a uniform plane wave propagating in the z-direction, we may replace the gradient
by ∇ = ẑ ∂z . It follows that the curls ∇ × E = ẑ × ∂z E and ∇ × H = ẑ × ∂z H will be
transverse to the z-direction. Then, Faraday’s and Ampère’s laws imply that Dz = 0
and Bz = 0, and hence both of Gauss’ laws are satisfied. Thus, we are left only with:
3.1 Linear and Circular Birefringence
ẑ × ∂z E = −jωB
In this chapter, we discuss wave propagation in anisotropic media that are linearly or cir- (3.1.3)
ẑ × ∂z H = jωD
cularly birefringent. In such media, uniform plane waves can be decomposed in two or-
thogonal polarization states (linear or circular) that propagate with two different speeds. These equations do not “see” the components Ez , Hz . However, in all the cases that
The two states develop a phase difference as they propagate, which alters the total po- we consider here, the conditions Dz = Bz = 0 will imply also that Ez = Hz = 0. Thus,
larization of the wave. Such media are used in the construction of devices for generating all fields are transverse, for example, E = x̂ Ex + ŷ Ey = ê+ E+ + ê− E− . Equating x, y
different polarizations. components in the two sides of Eq. (3.1.3), we find in the linear basis:
Linearly birefringent materials can be used to change one polarization into another,
such as changing linear into circular. Examples are the so-called uniaxial crystals, such ∂z Ex = −jωBy , ∂z Ey = jωBx
(linear basis) (3.1.4)
as calcite, quartz, ice, tourmaline, and sapphire. ∂z Hy = −jωDx , ∂z Hx = jωDy
Optically active or chiral media are circularly birefringent. Examples are sugar solu-
tions, proteins, lipids, nucleic acids, amino acids, DNA, vitamins, hormones, and virtually Using the vector property ẑ × ê± = ±j ê± and equating circular components, we
most other natural substances. In such media, circularly polarized waves go through obtain the circular-basis version of Eq. (3.1.3) (after canceling some factors of j):
unchanged, with left- and right-circular polarizations propagating at different speeds.
This difference causes linearly polarized waves to have their polarization plane rotate ∂z E± = ∓ωB±
as they propagate—an effect known as natural optical rotation. (circular basis) (3.1.5)
∂z H± = ±ωD±
A similar but not identical effect—the Faraday rotation—takes place in gyroelec-
tric media, which are ordinary isotropic materials (glass, water, conductors, plasmas)
subjected to constant external magnetic fields that break their isotropy. Gyromagnetic 3.2 Uniaxial and Biaxial Media
media, such as ferrites subjected to magnetic fields, also become circularly birefringent.
We discuss all four birefringent cases (linear, chiral, gyroelectric, and gyromagnetic) In uniaxial and biaxial homogeneous anisotropic dielectrics, the D−E constitutive rela-
and the type of constitutive relationships that lead to the corresponding birefringent tionships are given by the following diagonal forms, where in the biaxial case all diagonal
behavior. We begin by casting Maxwell’s equations in different polarization bases. elements of the permittivity matrix are distinct:
An arbitrary polarization can be expressed uniquely as a linear combination of two ⎡ ⎤ ⎡ ⎤⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎡ ⎤
Dx e 0 0 Ex Dx 1 0 0 Ex
polarizations along two orthogonal directions.† For waves propagating in the z-direction, ⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥
⎣ Dy ⎦ = ⎣ 0 o 0 ⎦ ⎣ Ey ⎦ and ⎣ Dy ⎦ = ⎣ 0 2 0 ⎦ ⎣ Ey ⎦ (3.2.1)
we may use the two linear directions {x̂, ŷ}, or the two circular ones for right and left
Dz 0 0 o Ez Dz 0 0 3 Ez
polarizations {ê+ , ê− }, where ê+ = x̂ − jŷ and ê− = x̂ + jŷ.‡ Indeed, we have the
following identity relating the linear and circular bases: For the uniaxial case, the x-axis is taken to be the extraordinary axis with 1 = e ,
† For ∗ whereas the y and z axes are ordinary axes with permittivities 2 = 3 = o .
complex-valued vectors e1 , e2 , orthogonality is defined with conjugation: e1 · e2 = 0.
‡ Note that ê± satisfy: ê∗ ∗
± · ê± = 2, ê+ · ê− = 0, ê+ × ê− = 2j ẑ, and ẑ × ê± = ±j ê± .
The ordinary z-axis was chosen to be the propagation direction in order for the
transverse x, y axes to correspond to two different permittivities. In this respect, the

60
62 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.3. Chiral Media 63

uniaxial and biaxial cases are similar, and therefore, we will work with the biaxial case.
Setting Dx = 1 Ex and Dy = 2 Ey in Eq. (3.1.4) and assuming B = µ0 H, we have:

∂z Ex = −jωµ0 Hy , ∂z Ey = jωµ0 Hx
(3.2.2)
∂z Hy = −jω1 Ex , ∂z Hx = jω2 Ey

Differentiating these once more with respect to z, we obtain the decoupled Helmholtz
equations for the x-polarized and y-polarized components:

∂2z Ex = −ω2 µ0 1 Ex
(3.2.3)
∂2z Ey = −ω2 µ0 2 Ey
The forward-moving solutions are:
√
Ex (z)= Ae−jk1 z , k1 = ω µ0 1 = k0 n1
√ (3.2.4)
Ey (z)= Be−jk2 z , k2 = ω µ0 2 = k0 n2
√
where k0 = ω µ0 0 = ω/c0 is the free-space wavenumber and we defined the refractive
 
indices n1 = 1 /0 and n2 = 2 /0 . Therefore, the total transverse field at z = 0 and
at distance z = l inside the medium will be: Fig. 3.2.1 Linearly and circularly birefringent retarders.

E(0) = x̂ A + ŷ B
(3.2.5)
3.3 Chiral Media
−jk1 l −jk2 l
 j(k1 −k2 )l −jk1 l
E(l) = x̂ Ae + ŷ Be = x̂ A + ŷ Be e
Ever since the first experimental observations of optical activity by Arago and Biot in
The relative phase φ = (k1 − k2 )l between the x- and y-components introduced by the early 1800s and Fresnel’s explanation that optical rotation is due to circular bire-
the propagation is called retardance: fringence, there have been many attempts to explain it at the molecular level. Pasteur
was the first to postulate that optical activity is caused by the chirality of molecules.
2πl There exist several versions of constitutive relationships that lead to circular bire-
φ = (k1 − k2 )l = (n1 − n2 )k0 l = (n1 − n2 ) (3.2.6)
λ fringence [265–281]. For single-frequency waves, they are all equivalent to each other.
For our purposes, the following so-called Tellegen form is the most convenient [34]:
where λ is the free-space wavelength. Thus, the polarization nature of the field keeps
changing as it propagates.
D = E − jχH
In order to change linear into circular polarization, the wave may be launched into (chiral media) (3.3.1)
B = µH + jχE
the birefringent medium with a linear polarization having equal x- and y-components.
After it propagates a distance l such that φ = (n1 − n2 )k0 l = π/2, the wave will have where χ is a parameter describing the chirality properties of the medium.
changed into left-handed circular polarization: It can be shown that the reality (for a lossless medium) and positivity of the energy
density function (E ∗ · D + H ∗ · B)/2 requires that the constitutive matrix
E(0) = A x̂ + ŷ
(3.2.7)  −jχ
E(l) = A x̂ + ŷ ejφ e−jk1 l = A x̂ + j ŷ e−jk1 l jχ µ
Polarization-changing devices that employ this property are called retarders and are
be hermitian and positive definite. This implies that , µ, χ are real, and furthermore,
shown in Fig. 3.2.1. The above example is referred to as a quarter-wave retarder because √
that |χ| < µ. Using Eqs. (3.3.1) in Maxwell’s equations (3.1.5), we obtain:
the condition φ = π/2 may be written as (n1 − n2 )l = λ/4.
∂z E± = ∓ωB± = ∓ω(µH± + jχE± )
(3.3.2)
∂z H± = ±ωD± = ±ω(E± − jχH± )
√  √
Defining c = 1/ µ, η = µ/, k = ω/c = ω µ, and the following real-valued
√
dimensionless parameter a = cχ = χ / µ (so that |a| < 1), we may rewrite Eqs. (3.3.2)
64 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.3. Chiral Media 65

in the following matrix forms: both right-polarized and both propagate with the same wavenumber k+ . Similarly, the
left-polarized waves EL+ and EL− both propagate with k− .
∂ E± jka k E± Thus, a wave of given circular polarization (left or right) propagates with the same
=∓ (3.3.3)
∂z ηH± −k jka ηH± wavenumber regardless of its direction of propagation. This is a characteristic difference
of chiral versus gyrotropic media in external magnetic fields.
These matrix equations may be diagonalized by appropriate linear combinations. For
Consider, next, the effect of natural rotation. We start with a linearly polarized field
example, we define the right-polarized (forward-moving) and left-polarized (backward-
at z = 0 and decompose it into its circular components:
moving) waves for the {E+ , H+ } case:
1
1 E(0)= x̂ Ax + ŷ Ay = ê+ A+ + ê− A− , with A± = (Ax ± jAy )
ER+ = E+ − jηH+ E+ = ER+ + EL+ 2
2

1  (3.3.4)
1 H+ = − ER+ − EL+ where Ax , Ay must be real for linear polarization. Propagating the circular components
EL+ = E+ + jηH+ jη forward by a distance l according to Eq. (3.3.10), we find:
2

It then follows from Eq. (3.3.3) that {ER+ , EL+ } will satisfy the decoupled equations: E(l) = ê+ A+ e−jk+ l + ê− A− e−jk− l

∂ ER+ −jk+ 0 ER+ ER+ (z)= A+ e−jk+ z = ê+ A+ e−j(k+ −k− )l/2 + ê− A− ej(k+ −k− )l/2 e−j(k+ +k− )l/2 (3.3.11)
= ⇒ (3.3.5)
∂z EL+ 0 jk− EL+ EL+ (z)= B+ ejk− z 
= ê+ A+ e−jφ + ê− A− ejφ e−j(k+ +k− )l/2
where k+ , k− are defined as follows:
where we defined the angle of rotation:
√
k± = k(1 ± a)= ω µ ± χ (3.3.6) 1
φ= (k+ − k− )l = akl (natural rotation) (3.3.12)
2
We may also define circular refractive indices by n± = k± /k0 , where k0 is the free-
√ √ √ Going back to the linear basis, we find:
space wavenumber, k0 = ω µ0 0 . Setting also n = k/k0 = µ/ µ0 0 , we have:
1 1
k± = n± k0 , n± = n(1 ± a) (3.3.7) ê+ A+ e−jφ + ê− A− ejφ = (x̂ − jŷ) (Ax + jAy )e−jφ + (x̂ + jŷ) (Ax − jAy )ejφ
2 2
For the {E− , H− } circular components, we define the left-polarized (forward-moving)
 
= x̂ cos φ − ŷ sin φ Ax + ŷ cos φ + x̂ sin φ Ay
and right-polarized (backward-moving) fields by:
= x̂ Ax + ŷ Ay
1
EL− = E− + jηH− E− = EL− + ER−
2 Therefore, at z = 0 and z = l, we have:

1  (3.3.8)
1 H− = EL− − ER− 
ER− = E− − jηH− jη E(0)= x̂ Ax + ŷ Ay
2  (3.3.13)
E(l) = x̂ Ax + ŷ Ay e−j(k+ +k− )l/2
Then, {EL− , ER− } will satisfy:
The new unit vectors x̂ = x̂ cos φ−ŷ sin φ and ŷ = ŷ cos φ+x̂ sin φ are recognized
∂ EL− −jk− 0 EL− EL− (z)= A− e−jk− z
= ⇒ (3.3.9) as the unit vectors x̂, ŷ rotated clockwise (if φ > 0) by the angle φ, as shown in Fig. 3.2.1
∂z ER− 0 jk+ ER− ER− (z)= B− ejk+ z (for the case Ax = 0, Ay = 0.) Thus, the wave remains linearly polarized, but its
polarization plane rotates as it propagates.
In summary, we obtain the complete circular-basis fields E± (z):
If the propagation is in the negative z-direction, then as follows from Eq. (3.3.10), the
roles of k+ and k− are interchanged so that the rotation angle becomes φ = (k− −k+ )l/2,
E+ (z) = ER+ (z)+EL+ (z)= A+ e−jk+ z + B+ ejk− z
(3.3.10) which is the negative of that of Eq. (3.3.12).
E− (z) = EL− (z)+ER− (z)= A− e−jk− z + B− ejk+ z If a linearly polarized wave travels forward by a distance l, gets reflected, and travels
back to the starting point, the total angle of rotation will be zero. By contrast, in the
Thus, the E+ (z) circular component propagates forward with wavenumber k+ and Faraday rotation case, the angle keeps increasing so that it doubles after a round trip
backward with k− , and the reverse is true of the E− (z) component. The forward-moving (see Problem 3.10.)
component of E+ and the backward-moving component of E− , that is, ER+ and ER− , are
66 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.5. Linear and Circular Dichroism 67

3.4 Gyrotropic Media These satisfy the decoupled equations:

Gyrotropic† media are isotropic media in the presence of constant external magnetic ∂ ER+ −jk+ 0 ER+ ER+ (z)= A+ e−jk+ z
= ⇒
fields. A gyroelectric medium (at frequency ω) has constitutive relationships: ∂z EL+ 0 jk+ EL+ EL+ (z)= B+ ejk+ z
⎡ ⎤ ⎡ ⎤⎡ ⎤ (3.4.7)
Dx 1 j2 0 Ex ∂ EL− −jk− 0 EL− EL− (z)= A− e−jk− z
⎢ ⎥ ⎢ ⎥⎢ ⎥ = ⇒
⎣ Dy ⎦ = ⎣ −j2 1 0 ⎦ ⎣ Ey ⎦ , B = µH (3.4.1) ∂z ER− 0 jk− ER− ER− (z)= B− ejk− z
Dz 0 0 3 Ez
Thus, the complete circular-basis fields E± (z) are:
For a lossless medium, the positivity of the energy density function requires that the
permittivity matrix be hermitian and positive-definite, which implies that 1 , 2 , 3 are
E+ (z) = ER+ (z)+EL+ (z)= A+ e−jk+ z + B+ ejk+ z
real, and moreover, 1 > 0, |2 | ≤ 1 , and 3 > 0. The quantity 2 is proportional to the
(3.4.8)
external magnetic field and reverses sign with the direction of that field. E− (z) = EL− (z)+ER− (z)= A− e−jk− z + B− ejk− z
A gyromagnetic medium, such as a ferrite in the presence of a magnetic field, has
similar constitutive relationships, but with the roles of D and H interchanged: Now, the E+ (z) circular component propagates forward and backward with the same
⎡ ⎤ ⎡ ⎤⎡ ⎤ wavenumber k+ , while E− (z) propagates with k− . Eq. (3.3.13) and the steps leading to
Bx µ1 jµ2 0 Hx
⎢ ⎥ ⎢ ⎥⎢ ⎥ it remain valid here. The rotation of the polarization plane is referred to as the Faraday
⎣ By ⎦ = ⎣ −jµ2 µ1 0 ⎦ ⎣ Hy ⎦ , D = E (3.4.2)
rotation. If the propagation is in the negative z-direction, then the roles of k+ and k−
Bz 0 0 µ3 Hz
remain unchanged so that the rotation angle is still the same as that of Eq. (3.3.12).
where again µ1 > 0, |µ2 | ≤ µ1 , and µ3 > 0 for a lossless medium. If a linearly polarized wave travels forward by a distance l, gets reflected, and travels
In the circular basis of Eq. (3.1.1), the above gyrotropic constitutive relationships back to the starting point, the total angle of rotation will be double that of the single
take the simplified forms: trip, that is, 2φ = (k+ − k− )l.
D± = (1 ± 2 )E± , B± = µH± , (gyroelectric) Problems 1.9 and 3.12 discuss simple models of gyroelectric behavior for conduc-
(3.4.3) tors and plasmas in the presence of an external magnetic field. Problem 3.14 develops
B± = (µ1 ± µ2 )H± , D± = E± , (gyromagnetic)
the Appleton-Hartree formulas for plane waves propagating in plasmas, such as the
where we ignored the z-components, which are zero for a uniform plane wave propa- ionosphere [282–286].
gating in the z-direction. For example, The gyromagnetic case is essentially identical to the gyroelectric one. Eqs. (3.4.5) to
(3.4.8) remain the same, but with circular wavenumbers and impedances defined by:
Dx ± jDy = (1 Ex + j2 Ey )±j(1 Ey − j2 Ex )= (1 ± 2 )(Ex ± jEy )

Next, we solve Eqs. (3.1.5) for the forward and backward circular-basis waves. Con- √ µ±
µ± = µ1 ± µ2 , k± = ω µ± , η± = (3.4.9)
sidering the gyroelectric case first, we define the following quantities: 

√ µ Problem 3.13 discusses a model for magnetic resonance exhibiting gyromagnetic
± = 1 ± 2 , k± = ω µ± , η± = (3.4.4)
± behavior. Magnetic resonance has many applications—from NMR imaging to ferrite mi-
crowave devices [287–298]. Historical overviews may be found in [296,298].
Using these definitions and the constitutive relations D± = ± E± , Eqs. (3.1.5) may
be rearranged into the following matrix forms:

∂ E± 0 ∓k± E± 3.5 Linear and Circular Dichroism

= (3.4.5)
∂z η± H± ±k± 0 η± H±
Dichroic polarizers, such as polaroids, are linearly birefringent materials that have widely
These may be decoupled by defining forward- and backward-moving fields as in different attenuation coefficients along the two polarization directions. For a lossy ma-
Eqs. (3.3.4) and (3.3.8), but using the corresponding circular impedances η± : terial, the field solutions given in Eq. (3.2.4) are modified as follows:
1 1 √
ER+ = E+ − jη+ H+ EL− = E− + jη− H− Ex (z)= Ae−jk1 z = Ae−α1 z e−jβ1 z , k1 = ω µ1 = β1 − jα1
2 2 √ (3.5.1)
(3.4.6) Ey (z)= Be−jk2 z = Be−α2 z e−jβ1 2 , k2 = ω µ2 = β2 − jα2
1 1
EL+ = E+ + jη+ H+ ER− = E− − jη− H−
2 2 where α1 , α2 are the attenuation coefficients. Passing through a length l of such a
† The term “gyrotropic” is sometimes also used to mean “optically active.” material, the initial and output polarizations will be as follows:
68 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.6. Oblique Propagation in Birefringent Media 69

We recall from Sec. 2.9 that a uniform plane wave propagating in a lossless isotropic
E(0)= x̂ A + ŷ B dielectric in the direction of a wave vector k is given by:
(3.5.2)
E(l) = x̂ Ae−jk1 l + ŷ Be−jk2 l = x̂ Ae−α1 l + ŷ Be−α2 l ejφ e−jβ1 l n
E(r)= E e−j k·r , H(r)= H e−j k·r , with k̂ · E = 0 , H= k̂ × E (3.6.1)
In addition to the phase change φ = (β1 −β2 )l, the field amplitudes have attenuated η0

by the unequal factors a1 = e−α1 l and a2 = e−α2 l . The resulting polarization will be where n is the refractive index of the medium n = /0 , η0 the free-space impedance,
elliptic with unequal semi-axes. If α2 α1 , then a2 a1 and the y-component can be and k̂ the unit-vector in the direction of k, so that,
ignored in favor of the x-component. √ ω √
This is the basic principle by which a polaroid material lets through only a preferred k = k k̂ , k = |k| = ω µ0  = nk0 , k0 = = ω µ0 0 (3.6.2)
c0
linear polarization. An ideal linear polarizer would have a1 = 1 and a2 = 0, correspond-
ing to α1 = 0 and α2 = ∞. Typical values of the attenuations for commercially available and k0 is the free-space wavenumber. Thus, E, H, k̂ form a right-handed system.
polaroids are of the order of a1 = 0.9 and a2 = 10−2 , or 0.9 dB and 40 dB, respectively. In particular, following the notation of Fig. 2.9.1, if k is chosen to lie in the xz plane
Chiral media may exhibit circular dichroism [267,280], in which the circular wavenum- at an angle θ from the z-axis, that is, k̂ = x̂ sin θ + ẑ cos θ, then there will be two inde-
bers become complex, k± = β± − jα± . Eq. (3.3.11) reads now: pendent polarization solutions: TM, parallel, or p-polarization, and TE, perpendicular,
or s-polarization, with fields given by
E(l) = ê+ A+ e−jk+ l + ê− A− e−jk− l n
(TM or p-polarization): E = E0 (x̂ cos θ − ẑ sin θ) , H=E0 ŷ
 −j(k+ −k− )l/2 j(k+ −k− )l/2 −j(k+ +k− )l/2 η0
= ê+ A+ e + ê− A− e e (3.5.3) n (3.6.3)
(TE or s-polarization): E = E0 ŷ , H= E0 (−x̂ cos θ + ẑ sin θ)
 η0
= ê+ A+ e−ψ−jφ + ê− A− eψ+jφ e−j(k+ +k− )l/2
where, in both the TE and TM cases, the propagation phase factor e−j k·r is:
where we defined the complex rotation angle:

1 1 1 e−j k·r = e−j(kz z+kx x) = e−jk0 n(z cos θ+x sin θ) (3.6.4)
φ − jψ = (k+ − k− )l = (β+ − β− )l − j (α+ − α− )l (3.5.4)
2 2 2 The designation as parallel or perpendicular is completely arbitrary here and is taken
Going back to the linear basis as in Eq. (3.3.13), we obtain: with respect to the xz plane. In the reflection and refraction problems discussed in
Chap. 6, the dielectric interface is taken to be the xy plane and the xz plane becomes

E(0)= x̂ Ax + ŷ Ay the plane of incidence.
 (3.5.5)
E(l) = x̂ Ax + ŷ Ay e−j(k+ +k− )l/2 In a birefringent medium, the propagation of a uniform plane wave with arbitrary
wave vector k is much more difficult to describe. For example, the direction of the
where {x̂ , ŷ } are the same rotated (by φ) unit vectors of Eq. (3.3.13), and Poynting vector is not towards k, the electric field E is not orthogonal to k, the simple
dispersion relationship k = nω/c0 is not valid, and so on.
Ax = Ax cosh ψ − jAy sinh ψ In the previous section, we considered the special case of propagation along an ordi-
(3.5.6)
Ay = Ay cosh ψ + jAx sinh ψ nary optic axis in a birefringent medium. Here, we discuss the somewhat more general
case in which the xyz coordinate axes coincide with the principal dielectric axes (so that
Because the amplitudes Ax , Ay are now complex-valued, the resulting polarization the permittivity tensor is diagonal,) and we take the wave vector k to lie in the xz plane
will be elliptical. at an angle θ from the z-axis. The geometry is depicted in Fig. 3.6.1.
Although this case is still not the most general one with a completely arbitrary direc-
tion for k, it does contain most of the essential features of propagation in birefringent
3.6 Oblique Propagation in Birefringent Media media. The 3M multilayer films mentioned above have similar orientations for their
optic axes [246].
Here, we discuss TE and TM waves propagating in oblique directions in linearly birefrin-
The constitutive relations are assumed to be B = µ0 H and a diagonal permittivity
gent media. We will use these results in Chap. 7 to discuss reflection and refraction in
tensor for D. Let 1 , 2 , 3 be the permittivity values along the three principal axes and
such media, and to characterize the properties of birefringent multilayer structures. 
define the corresponding refractive indices ni = i /0 , i = 1, 2, 3. Then, the D -E
Applications include the recently manufactured (by 3M, Inc.) multilayer birefrin-
relationship becomes:
gent polymer mirrors that have remarkable and unusual optical properties, collectively
⎡ ⎤ ⎡ ⎤⎡ ⎤ ⎡ ⎤⎡ ⎤
referred to as giant birefringent optics (GBO) [246]. Dx 1 0 0 Ex n21 0 0 Ex
⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥⎢ ⎥
Oblique propagation in chiral and gyrotropic media is discussed in the problems. ⎣ Dy ⎦ = ⎣ 0 2 0 ⎦ ⎣ Ey ⎦ = 0 ⎣ 0 n22 0 ⎦ ⎣ Ey ⎦ (3.6.5)
Further discussions of wave propagation in anisotropic media may be found in [31–33]. Dz 0 0 3 Ez 0 0 n23 Ez
70 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.6. Oblique Propagation in Birefringent Media 71

Thus, the TE mode propagates as if the medium were isotropic with index n = n2 ,
whereas the TM mode propagates in a more complicated fashion. If the wave vector k
is along the ordinary x-axis (θ = 90o ), then k = kx = n3 ω/c0 (this was the result of
the previous section), and if k is along the extraordinary z-axis (θ = 0o ), then we have
k = kz = n1 ω/c0 .
For TM modes, the group velocity is not along k. In general, the group velocity
depends on the ω-k relationship and is computed as v = ∂ω/∂k. From Eq. (3.6.10), we
find the x- and z-components:

∂ω kx c20 N
vx = = = c0 2 sin θ
∂kx ωn23 n3
Fig. 3.6.1 Uniform plane waves in a birefringent medium. (3.6.12)
∂ω kz c20 N
vz = = = c0 2 cos θ
∂kz ωn21 n1
For a biaxial medium, the three ni are all different. For a uniaxial medium, we take
the xy-axes to be ordinary, with n1 = n2 = no , and the z-axis to be extraordinary, with The velocity vector v is not parallel to k. The angle θ̄ that v forms with the z-axis is
n3 = ne .† The wave vector k can be resolved along the z and x directions as follows: given by tan θ̄ = vx /vz . It follows from (3.6.12) that:

k = k k̂ = k(x̂ sin θ + ẑ cos θ)= x̂kx + ẑkz (3.6.6)

n21
tan θ̄ = tan θ (group velocity direction) (3.6.13)
n23
The ω-k relationship is determined from the solution of Maxwell’s equations. By
analogy with the isotropic case that has k = nk0 = nω/c0 , we may define an effective
Clearly, θ̄ = θ if n1 = n3 . The relative directions of k and v are shown in Fig. 3.6.2.
refractive index N such that:
The group velocity is also equal to the energy transport velocity defined in terms of the
Poynting vector P and energy density w as v = P /w. Thus, v and P have the same
ω
k = Nk0 = N (effective refractive index) (3.6.7) direction. Moreover, with the electric field being orthogonal to the Poynting vector, the
c0
angle θ̄ is also equal to the angle the E-field forms with the x-axis.
We will see in Eq. (3.6.22) by solving Maxwell’s equations that N depends on the
chosen polarization (according to Fig. 3.6.1) and on the wave vector direction θ:
⎧
⎪
⎪ n1 n3
⎨  , (TM or p-polarization)
2
N= n sin 2
θ + n23 cos2 θ (3.6.8)
⎪
⎪
1
⎩ n , (TE or s-polarization)
2

For the TM case, we may rewrite the N-θ relationship in the form:

1 cos2 θ sin2 θ
= + (effective TM index) (3.6.9)
N2 n21 n23

Multiplying by k2 and using k0 = k/N, and kx = k sin θ, kz = k cos θ, we obtain the Fig. 3.6.2 Directions of group velocity, Poynting vector, wave vector, and electric field.
ω-k relationship for the TM case:
Next, we derive Eqs. (3.6.8) for N and solve for the field components in the TM
ω2 k2z k2x
2 = 2 + 2 (TM or p-polarization) (3.6.10) and TE cases. We look for propagating solutions of Maxwell’s equations of the type
c0 n1 n3 E(r)= E e−j k·r and H(r)= H e−j k·r . Replacing the gradient operator by ∇ → −jk and
Similarly, we have for the TE case: canceling some factors of j, Maxwell’s equations take the form:

ω2 k2
2 = (TE or s-polarization) (3.6.11)
c0 n22
† In Sec. 3.2, the extraordinary axis was the x-axis.
72 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.6. Oblique Propagation in Birefringent Media 73

∇ × E = −jωµ0 H k × E = ωµ0 H E(r) = E0 ŷ e−j k·r

∇ × H = jωD k × H = −ωD n2 (TE) (3.6.18)
⇒ (3.6.14) H(r) = E0 (−x̂ cos θ + ẑ sin θ)e−j k·r
η0
∇·D=0 k·D=0
where the TE propagation phase factor is:
∇·H=0 k·H=0

The last two equations are implied by the first two, as can be seen by dotting both e−j k·r = e−jk0 n2 (z cos θ+x sin θ) (TE propagation factor) (3.6.19)
sides of the first two with k. Replacing k = k k̂ = Nk0 k̂, where N is still to be determined,
we may solve Faraday’s law for H in terms of E : The TM case requires a little more work. The linear system (3.6.17) becomes now:
 
ω N n21
N k̂ × E = ωµ0 H ⇒ H= k̂ × E (3.6.15) 1− Ex = (Ex sin θ + Ez cos θ)sin θ
c0 η0 N2
  (3.6.20)
where we used η0 = c0 µ0 . Then, Ampère’s law gives: n2
1 − 32 Ez = (Ex sin θ + Ez cos θ)cos θ
N
2
1 1 ω N 1
D=− k×H=− N k̂ × H = k̂ × (E × k̂) ⇒ k̂ × (E × k̂)= D Using the identity sin2 θ + cos2 θ = 1, we may rewrite Eq. (3.6.20) in the matrix form:
ω ω c0 η0 c0 0 N2
⎡ ⎤
n21
where we used c0 η0 = 1/0 . The quantity k̂ × (E × k̂) is recognized as the component of ⎢ cos θ −
2
− sin θ cos θ ⎥ Ex
⎢ N2 ⎥ =0 (3.6.21)
E that is transverse to the propagation unit vector k̂. Using the BAC-CAB vector identity, ⎣ n2 ⎦ Ez
we have k̂ × (E × k̂)= E − k̂(k̂ · E). Rearranging terms, we obtain: − sin θ cos θ sin2 θ − 32
N

1 Setting the determinant of the coefficient matrix to zero, we obtain the desired con-
E− D = k̂(k̂ · E) (3.6.16)
0 N2 dition on N in order that a non-zero solution Ex , Ez exist:
  
Because D is linear in E, this is a homogeneous linear equation. Therefore, in order n2 n2
to have a nonzero solution, its determinant must be zero. This provides a condition cos θ − 12
2
sin θ − 32
2
− sin2 θ cos2 θ = 0 (3.6.22)
N N
from which N can be determined.
To obtain both the TE and TM solutions, we assume initially that E has all its three This can be solved for N2 to give Eq. (3.6.9). From it, we may also derive the following
components and rewrite Eq. (3.6.16) component-wise. Using Eq. (3.6.5) and noting that relationship, which will prove useful in applying Snell’s law in birefringent media:
k̂ · E = Ex sin θ + Ez cos θ, we obtain the homogeneous linear system:

   
n1  N2 sin2 θ
n2 N cos θ = n3 − N sin θ = n1 1 −
2 2 2
(3.6.23)
1 − 12 Ex = (Ex sin θ + Ez cos θ)sin θ n3 n23
N
 
n22 With the help of the relationships given in Problem 3.16, the solution of the homo-
1− Ey = 0 (3.6.17) geneous system (3.6.20) is found to be, up to a proportionality constant:
N2
  n3 n1
n23 Ex = A cos θ , Ez = −A sin θ (3.6.24)
1− Ez = (Ex sin θ + Ez cos θ)cos θ n1 n3
N2
The TE case has Ey = 0 and Ex = Ez = 0, whereas the TM case has Ex = 0, Ez = 0,  constant A can be expressed in terms of the total magnitude of the field E0 =
The
|E| = |Ex |2 + |Ez |2 . Using the relationship (3.7.11), we find (assuming A > 0):
and Ey = 0. Thus, the two cases decouple.
In the TE case, the second of Eqs. (3.6.17) immediately implies that N = n2 . Setting N
A = E0  (3.6.25)
E = E0 ŷ and using k̂ × ŷ = −x̂ cos θ + ẑ sin θ, we obtain the TE solution: n1 + n23 − N2
2

The magnetic field H can also be expressed in terms of the constant A. We have:
74 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.7. Problems 75

3.7 Problems
N N
H= k̂ × E = (x̂ sin θ + ẑ cos θ)×(x̂Ex + ẑEz ) 3.1 For the circular-polarization basis of Eq. (3.1.1), show
η0 η0
 
N N n3 n1 E = ê+ E+ + ê− E− ⇒ ẑ × E = j ê+ E+ − j ê− E− ⇒ ẑ × E± = ±jE±
= ŷ(Ex cos θ − Ez sin θ)= ŷ A cos2 θ + sin2 θ (3.6.26)
η0 η0 n1 n3
3.2 Show the component-wise Maxwell equations, Eqs. (3.1.4) and (3.1.5), with respect to the
N n1 n3 A n1 n3 linear and circular polarization bases.
= ŷ A = ŷ
η0 N2 η0 N
3.3 Suppose that the two unit vectors {x̂, ŷ} are rotated about the z-axis by an angle φ resulting
where we used Eq. (3.7.10). In summary, the complete TM solution is: in x̂ = x̂ cos φ + ŷ sin φ and ŷ = ŷ cos φ − x̂ sin φ. Show that the corresponding circular
  basis vectors ê± = x̂ ∓ j ŷ and ê± = x̂ ∓ j ŷ change by the phase factors: ê± = e±jφ ê± .
N n3 n1
E(r) = E0  x̂ cos θ − ẑ sin θ e−j k·r 3.4 Consider a linearly birefringent 90o quarter-wave retarder. Show that the following input
n21 + n23 − N2 n1 n3
polarizations change into the indicated output ones:
(TM) (3.6.27)
E0 n1 n3
H(r) =  ŷ e−j k·r x̂ ± ŷ → x̂ ± j ŷ
η0 n2 + n2 − N2 x̂ ± j ŷ → x̂ ± ŷ
1 3

where the TM propagation phase factor is: What are the output polarizations if the same input polarizations go through a 180o half-
wave retarder?
e−j k·r = e−jk0 N(z cos θ+x sin θ) (TM propagation factor) (3.6.28) 3.5 A polarizer lets through linearly polarized light in the direction of the unit vector êp =
The solution has been put in a form that exhibits the proper limits at θ = 0o and x̂ cos θp + ŷ sin θp , as shown in Fig. 3.7.1. The output of the polarizer propagates in the
o z-direction through a linearly birefringent retarder of length l, with birefringent refractive
90 . It agrees with Eq. (3.6.3) in the isotropic case. The angle that E forms with the x-axis
indices n1 , n2 , and retardance φ = (n1 − n2 )k0 l.
in Fig. 3.6.2 is given by tan θ̄ = −Ez /Ex and agrees with Eq. (3.6.13).
Next, we derive expressions for the Poynting vector and energy densities. It turns
out—as is common in propagation and waveguide problems—that the magnetic energy
density is equal to the electric one. Using Eq. (3.6.27), we find:
 
1 E2 n1 n3 N n1 n3
P= Re(E × H ∗ )= 0 x̂ sin θ + ẑ cos θ (3.6.29)
2 2η0 n21 + n23 − N2 n3 n1
and for the electric, magnetic, and total energy densities:
1 1 1 n21 n23 Fig. 3.7.1 Polarizer-analyzer measurement of birefringence.
we = Re(D · E ∗ )= 0 n21 |Ex |2 + n23 |Ez |2 = 0 E02 2
2 4 4 n1 + n23 − N2
The output E(l) of the birefringent sample goes through an analyzing linear polarizer that
1 1 1 n21 n23 lets through polarizations along the unit vector êa = x̂ cos θa + ŷ sin θa . Show that the light
wm = Re(B · H ∗ )= µ0 |Hy |2 = 0 E02 2 = we (3.6.30)
2 4 4 n1 + n23 − N2 intensity at the output of the analyzer is given by:

1 n21 n23  2  2
w = we + wm = 2we = 0 E02 2 Ia = êa · E(l) = cos θa cos θp + ejφ sin θa sin θp 
2 n1 + n23 − N2
For a circularly birefringent sample that introduces a natural or Faraday rotation of φ =
The vector P is orthogonal to E and its direction is θ̄ given by Eq. (3.6.13), as can be
(k+ − k− )l/2, show that the output light intensity will be:
verified by taking the ratio tan θ̄ = Px /Pz . The energy transport velocity is the ratio of
 2
the energy flux to the energy density—it agrees with the group velocity (3.6.12): Ia = êa · E(l) = cos2 (θp − θa − φ)
 
P N N
v= = c0 x̂ 2 sin θ + ẑ 2 cos θ (3.6.31) For both the linear and circular cases, what are some convenient choices for θa and θp ?
w n3 n1
3.6 A linearly polarized wave with polarization direction at an angle θ with the x-axis goes
To summarize, the TE and TM uniform plane wave solutions are given by Eqs. (3.6.18) through a circularly birefringent retarder that introduces an optical rotation by the angle
and (3.6.27). We will use these results in Sects. 7.6 and 7.8 to discuss reflection and re- φ = (k+ − k− )l/2. Show that the input and output polarization directions will be:
fraction in birefringent media and multilayer birefringent dielectric structures. Further
discussion of propagation in birefringent media can be found in [187] and [246–264]. x̂ cos θ + ŷ sin θ → x̂ cos(θ − φ)+ŷ sin(θ − φ)
76 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.7. Problems 77

3.7 Show that an arbitrary polarization vector can be expressed as follows with respect to a
linear basis {x̂, ŷ} and its rotated version {x̂ , ŷ }:
EL+ (0)= EL+ (l)e−jk− l = ER+ (l)e−jk− l = ER+ (0)e−2jk+ l
 
E = A x̂ + B ŷ = A x̂ + B ŷ   ER− (0)= ER− (l)e−jk+ l = EL− (l)e−jk+ l = EL− (0)e−2jk− l

where the new coefficients and the new basis vectors are related to the old ones by a rotation Show that the total Faraday rotation angle will be 2φ = (k+ − k− )l.
by an angle φ: 3.11 Show that the x, y components of the gyroelectric and gyromagnetic constitutive relation-
ships (3.4.1) and (3.4.2) may be written in the compact forms:
A cos φ sin φ A x̂ cos φ sin φ x̂
= , =
B − sin φ cos φ B ŷ − sin φ cos φ ŷ DT = 1 ET − j2 ẑ × ET (gyroelectric)
BT = µ1 HT − jµ2 ẑ × HT (gyromagnetic)
3.8 Show that the source-free Maxwell’s equations (3.1.2) for a chiral medium characterized by
√  √
(3.3.1), may be cast in the matrix form, where k = ω µ, η = µ/, and a = χ/ µ:
where the subscript T indicates the transverse (with respect to z) part of a vector, for exam-
E ka −jk E ple, DT = x̂ Dx + ŷ Dy .
∇× =
ηH jk ka ηH 3.12 Conductors and plasmas exhibit gyroelectric behavior when they are in the presence of an
external magnetic field. The equation of motion of conduction electrons in a constant mag-
Show that these may be decoupled by forming the “right” and “left” polarized fields: netic field is mv̇ = e(E + v × B)−mαv, with the collisional damping term included. The
magnetic field is in the z-direction, B = ẑ B0 .
ER k+ 0 ER 1 1
∇× = , where ER = (E − jηH) , EL = (E + jηH) Assuming ejωt time dependence and decomposing all vectors in the circular basis (3.1.1),
EL 0 −k− EL 2 2
for example, v = ê+ v+ + ê− v− + ẑ vz , show that the solution of the equation of motion is:
where k± = k(1 ± a). Using these results, show that the possible plane-wave solutions e e
propagating in the direction of a unit-vector k̂ are given by: E± Ez
v± = m , vz = m
α + j(ω ± ωB ) α + jω
E(r)= E0 (p̂ − j ŝ)e−j k+ ·r and E(r)= E0 (p̂ + j ŝ)e−j k− ·r
where ωB = eB0 /m is the cyclotron frequency. Then, show that the D−E constitutive
where k± = k± k̂ and {p̂, ŝ, k̂} form a right-handed system of unit vectors, such as {x̂ , ŷ , ẑ } relationship takes the form of Eq. (3.4.1) with:
of Fig. 2.9.1. Determine expressions for the corresponding magnetic fields. What freedom
do we have in selecting {p̂, ŝ} for a given direction k̂ ? jω2p jω2p
± = 1 ± 2 = 0 1 −  ,  3 = 0 1 −
3.9 Using Maxwell’s equations (3.1.2), show the following Poynting-vector relationships for an ω α + j(ω ± ωB ) ω(α + jω)
arbitrary source-free medium:
where ω2p = Ne2 /m0 is the plasma frequency and N, the number of conduction electrons
∇ · E × H ∗ = jω D ∗ · E − B · H ∗ per unit volume. (See Problem 1.9 for some helpful hints.)

∇ · Re E × H ∗ = −ω Im D ∗ · E + B ∗ · H 3.13 If the magnetic field Htot = ẑ H0 + Hejωt is applied to a magnetizable sample, the in-
duced magnetic moment per unit volume (the magnetization) will have the form Mtot =
Explain why a lossless medium must satisfy the condition ∇ · Re E × H ∗ = 0. Show that ẑ M0 + Mejωt , where ẑ M0 is the saturation magnetization due to ẑ H0 acting alone. The
this condition requires that the energy function w = (D ∗ · E + B ∗ · H)/2 be real-valued. phenomenological equations governing Mtot , including a so-called Landau-Lifshitz damping
For a lossless chiral medium characterized by (3.3.1), show that the parameters , µ, χ are term, are given by [295]:
required to be real. Moreover, show that the positivity of the energy function w > 0 requires
√ dMtot α
that |χ| < µ, as well as  > 0 and µ > 0. = γ(Mtot × Htot )− Mtot × (Mtot × Htot )
dt M0 H0
3.10 In a chiral medium, at z = 0 we lauch the fields ER+ (0) and EL− (0), which propagate by a
distance l, get reflected, and come back to the starting point. Assume that at the point of where γ is the gyromagnetic ratio and τ = 1/α, a relaxation time constant. Assuming that
reversal the fields remain unchanged, that is, ER+ (l)= EL+ (l) and EL− (l)= ER− (l). Using |H| H0 and |M| M0 , show that the linearized version of this equation obtained by
the propagation results (3.3.5) and (3.3.9), show that fields returned back at z = 0 will be: keeping only first order terms in H and M is:


EL+ (0)= EL+ (l)e−jk− l = ER+ (l)e−jk− l = ER+ (0)e−j(k+ +k− )l jωM = ωM (ẑ × H)−ωH (ẑ × M)−α ẑ × (M − χ0 H)×ẑ
ER− (0)= ER− (l)e−jk+ l = EL− (l)e−jk+ l = EL− (0)e−j(k+ +k− )l
where ωM = γM0 , ωH = γH0 , and χ0 = M0 /H0 . Working in the circular basis (3.1.1), show
Show that the overall natural rotation angle will be zero. For a gyrotropic medium, show that the solution of this equation is:
that the corresponding rountrip fields will be:
78 Electromagnetic Waves & Antennas – S. J. Orfanidis 3.7. Problems 79

For the case k̂ = x̂ (θ = 90o ), show that:

α ± jωH √
M± = χ 0 H ± ≡ χ ± H± and Mz = 0 0 N 2 =  3 , k = k3 = ω µ3 , E + = 0, E− = 0, Ez = 0
α + j(ω ± ωH ) √ +
0 N =  e ,
2
k = ke = ω µe , E+ = 0, E− = − E+ , Ez = 0
−
Writing B = µ0 (H + M), show that the permeability matrix has the gyromagnetic form of
Eq. (3.4.2) with µ1 ± µ2 = µ± = µ0 (1 + χ± ) and µ3 = µ0 . Show that the real and imaginary For each of the above four special solutions, derive the corresponding magnetic fields H .
parts of µ1 are given by [295]: Justify the four values of N2 on the basis of Eq. (3.7.3). Discuss the polarization properties of
the four cases. For the “extraordinary” wave k = ke , show that Dx = 0 and Ex /Ey = −j2 /1 .
α2 + ωH (ω + ωH )
µ 0 χ0 α2 − ωH (ω − ωH ) Eq. (3.7.4) and the results of Problem 3.14 lead to the so-called Appleton-Hartree equations
Re(µ1 ) = µ0 + +
2 α2 + (ω + ωH )2 α2 + (ω − ωH )2 for describing plasma waves in a magnetic field [282–286].
 
µ 0 χ0 αω αω 3.15 A uniform plane wave, Ee−j k·r and He−j k·r , is propagating in the direction of the unit vector
Im(µ1 ) = − + 2
2 α2 + (ω + ωH )2 α + (ω − ωH )2 k̂ = ẑ = ẑ cos θ + ẑ sin θ shown in Fig. 2.9.1 in a gyromagnetic medium with constitutive
relationships (3.4.2). Using Maxwell’s equations, show that:
Derive similar expressions for Re(µ2 ) and Im(µ2 ).
k × E = ωB , k·B=0 1
3.14 A uniform plane wave, Ee−j k·r and He−j k·r , is propagating in the direction of the unit vector ⇒ H− B = k̂(k̂ · H) (3.7.5)
k × H = −ωE , k·E=0 µ0 N 2
k̂ = ẑ = ẑ cos θ + ẑ sin θ shown in Fig. 2.9.1 in a gyroelectric medium with constitutive
relationships (3.4.1). where the effective refractive index N is defined through the wavevector k = k k̂, where
Show that Eqs. (3.6.14)–(3.6.16) remain valid provided we define the effective refractive index √
√ k = Nk0 , k0 = ω µ0 . Working in the circular polarization basis H = ê+ H+ + ê− H− + ẑ Hz ,
N through the wavevector k = k k̂, where k = Nk0 , k0 = ω µ0 . where H± = (Hx ± jHy )/2, show that Eq. (3.7.5) leads to the homogeneous system:
Working in the circular-polarization basis (3.1.1), that is, E = ê+ E+ + ê− E− + ẑ Ez , where ⎡ ⎤
1 µ+ 1 1
E± = (Ex ± jEy )/2, show that Eq. (3.6.16) leads to the homogeneous system: ⎢ 1 − 2 sin θ − µ N2 − −
2
sin2 θ sin θ cos θ ⎥⎡ ⎤
⎢ 0 2 2 ⎥ H+
⎡ ⎤ ⎢ ⎥⎢
1 + 1 1 ⎢ 1
− sin2 θ
1
1 − sin2 θ −
µ− 1
− sin θ cos θ ⎥ ⎣ H− ⎥
⎦=0 (3.7.6)
⎢ ⎥
⎢ 1 − 2 sin θ −  N2 − −
2
sin2 θ sin θ cos θ ⎥ ⎡ ⎤ ⎢ 2 2 µ0 N 2 2 ⎥
⎢ 0 2 2 ⎥ E+ ⎣ µ3 ⎦ Hz
⎢ 1 1 − 1 ⎥⎢ ⎥ − sin θ cos θ − sin θ cos θ sin θ −
2
⎢ − sin θ cos θ ⎥
⎢ − sin2 θ 1− sin2 θ − ⎥ ⎣ E− ⎦ = 0 (3.7.1) µ0 N 2
⎢ 2 2 0 N 2 2 ⎥
⎣  3 ⎦ Ez
− sin θ cos θ − sin θ cos θ sin2 θ − 2
where µ± = µ1 ± µ2 . Alternatively, show that in the linear-polarization basis:
0 N
⎡ ⎤⎡⎤
µ1 − µ0 N2 cos2 θ jµ2 µ0 N2 sin θ cos θ Hx
where ± = 1 ± 2 . Alternatively, show that in the linear-polarization basis: ⎢ ⎥⎢ ⎥
⎣ −jµ2 µ1 − µ0 N2 0 ⎦ ⎣ Hy ⎦ = 0 (3.7.7)
⎡ ⎤⎡ ⎤ µ0 N2 sin θ cos θ 0 µ3 − µ0 N2 sin2 θ Hz
1 − 0 N2 cos2 θ j2 0 N2 sin θ cos θ Ex
⎢ ⎥⎢ ⎥
⎣ −j2 1 −  0 N 2 0 ⎦ ⎣ Ey ⎦ = 0 (3.7.2)
For either basis, setting the determinant of the coefficient matrix to zero, show that a non-
0 N2 sin θ cos θ 0 3 − 0 N2 sin2 θ Ez
zero E solution exists provided that N2 is one of the two solutions of:
For either basis, setting the determinant of the coefficient matrix to zero, show that a non- µ3 (µ0 N2 − µ+ )(µ0 N2 − µ− ) 2µ + µ − µ21 − µ22
zero E solution exists provided that N2 is one of the two solutions of: tan2 θ = − , where µe = = (3.7.8)
µ1 (µ0 N2 − µ3 )(µ0 N2 − µe ) µ+ + µ − µ1
3 (0 N2 − + )(0 N2 − − ) 2+ − 2 − 22 Show that the two solutions for N2 are:
tan2 θ = − , where e = = 1 (3.7.3)
1 (0 N2 − 3 )(0 N2 − e ) + +  − 1 
(µ21 − µ22 − µ1 µ3 )sin2 θ + 2µ1 µ3 ± (µ21 − µ22 − µ1 µ3 )2 sin4 θ + 4µ22 µ23 cos2 θ
Show that the two solutions for N2 are: N =2
2µ0 (µ1 sin2 θ + µ3 cos2 θ)

(21 − 22 − 1 3 )sin2 θ + 21 3 ± (21 − 22 − 1 3 )2 sin4 θ + 422 23 cos2 θ For the special case θ = 0o , show that the two possible solutions of Eq. (3.7.6) are:
N2 = (3.7.4)
20 (1 sin2 θ + 3 cos2 θ) √
µ0 N 2 = µ + , k = k+ = ω µ+ , H+ = 0, H − = 0, Hz = 0
√
For the special case k̂ = ẑ (θ = 0 ), show that the two possible solutions of Eq. (3.7.1) are:
o µ0 N 2 = µ − , k = k+ = ω µ− , H + = 0, H− = 0, Hz = 0

√ For the special case θ = 90o , show that:

0 N 2 =  + , k = k+ = ω µ+ , E+ = 0, E − = 0, Ez = 0
√ √
0 N 2 =  − , k = k− = ω µ− , E + = 0, E− = 0, Ez = 0 µ0 N 2 = µ3 , k = k3 = ω µ3 , H + = 0, H − = 0, Hz = 0
√ µ+
µ0 N = µe ,
2
k = ke = ω µe , H+ = 0, H− = − H+ , Hz = 0
µ−
80 Electromagnetic Waves & Antennas – S. J. Orfanidis

For each of the above four special solutions, derive the corresponding electric fields E . Justify
the four values of N2 on the basis of Eq. (3.7.8). Discuss the polarization properties of the
four cases. This problem is the dual of Problem 3.14.
3.16 Using Eq. (3.6.9) for the effective TM refractive index in a birefringent medium, show the
following additional relationships:

sin2 θ cos2 θ
+ =1 (3.7.9)
n21 n23
1− 1−
N2 N2
n3 n1 n1 n3
cos2 θ + sin2 θ = (3.7.10)
n1 n3 N2
n21 n2 n2 + n23 − N2
sin2 θ + 32 cos2 θ = 1 (3.7.11)
n23 n1 N2
n21 n23
1− 1−
sin2 θ = N2 , cos2 θ = N2 (3.7.12)
n2 n23
1 − 12 1−
n3 n21
n21 n2 n23 n2
cos2 θ − 2
= − 21 sin2 θ , sin2 θ − 2
= − 32 cos2 θ (3.7.13)
N n3 N n1
Using these relationships, show that the homogeneous linear system (3.6.20) can be simpli-
fied into the form:
n1 n3 n3 n1
Ex sin θ = −Ez cos θ , Ez cos θ = −Ex sin θ
n3 n1 n1 n3