Appendix: Electromagnetism in Dielectrics

Appendix Electromagnetism in
dielectrics
A
This appendix summarizes the principal results of electromagnetism that
are used throughout the book. It is hoped that the reader will be famil-
A.1 Electromagnetic fields
iar with this material. The main purpose of the appendix is to collect
and Maxwell’s
equations 330 together the equations in a concise form for quick reference, and also to
A.2 Electromagnetic
define the notation. SI units are used throughout. A short bibliography
waves 333 of suitable supplementary texts is given under Further Reading.
Further reading 339
A.1 Electromagnetic fields and Maxwell’s

equations
The response of a dielectric material to an external electric field is char-
acterized by three macroscopic vectors:
• the electric field strength E;

• the polarization P ;
Copyright © 2010. Oxford University Press, Incorporated. All rights reserved.
• the electric displacement D.
The microscopic response of the material is determined primarily by the

polarization. For this reason, the first task in all the examples treated by
electromagnetism in this book is to calculate P . The dielectric constant
r is then determined from P , and the optical properties are deduced
from r .
The polarization is defined as the net dipole moment per unit volume.
The application of a field produces a polarization by the forces exerted
on the positive and negative charges of the atoms that are contained
within the medium. If the molecules have permanent dipole moments,
the field will apply a torque to these randomly orientated dipoles and
tend to align them along the field direction. If there are no permanent
dipoles, the field will push the positive and negative charges in opposite
directions and induce a dipole parallel to the field. In either case, the end
result is the same: the application of the field tends to produce many
microscopic dipoles aligned parallel to the direction of the external field.
This generates a net dipole moment within the dielectric, and hence a
polarization.
The microscopic dipoles will all tend to align along the field direction,
and so the polarization vector will be parallel to E. This allows us to
Fox, Mark. <i>Optical Properties of Solids : Optical Properties of Solids</i>, Oxford University Press, Incorporated, 2010. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/vcu/detail.action?docID=3056080.
Created from vcu on 2019-09-24 07:59:07.
A.1 Electromagnetic fields and Maxwell’s equations 331
write:
P = 0 χE, (A.1)
where 0 is the electric permittivity of free space and χ is the electric
susceptibility of the medium. The value of 0 is 8.854 × 10−12 Fm−1 in
SI units.
Equation A.1 makes two assumptions that need a brief word of expla-
nation.
(1) We have assumed that the medium is isotropic, even though we
know that some materials are anisotropic. In particular, anisotropic
crystals have preferred non-equivalent axes, and P will not neces-
sarily be parallel to E. A discussion of how to treat non-
isotropic materials may be found in
(2) We have assumed that P varies linearly with E. This will not Section 2.5, while nonlinear optics is
always be the case. In particular, if the optical intensity is very the subject of Chapter 11.
large, we can enter the realm of nonlinear optics, in which eqn A.1
is not valid.
Both of these qualifications introduce unnecessary complications at this
stage, and are not considered further in this appendix.
The electric displacement D of the medium is related to the electric
field E and polarization P through:
D = 0 E + P . (A.2)
This may be considered to be the definition of D. By combining eqns A.1

and A.2, we can write:
D = 0 r E, (A.3)
where
r = 1 + χ . (A.4)
r is the relative dielectric constant of the medium, and is an ex-
tremely important parameter in the understanding of the propagation
of light through dielectrics.
In electrostatic problems we are frequently interested in calculating
the spatial dependence of the electric field, and hence the electric poten-
tial V , from the free charge density . This calculation can be performed
by using the Poisson equation:

∇2 V = − . (A.5)
r 0
Poisson’s equation is derived from Gauss’s law of electrostatics:

∇·E = . (A.6)
r 0
We recall that the electric field strength is the gradient of the potential:
E = −∇V. (A.7)
Equation A.5 follows directly by substituting for E in eqn A.6 using

eqn A.7. Once we know V , we can then calculate E from eqn A.7. This
332 Electromagnetism in dielectrics
approach is useful for treating devices that have a fixed potential across
them determined by an external voltage source.
The response of a material to external magnetic fields is treated in
a similar way to the response of dielectrics to electric fields. The mag-
netization M of the medium is proportional to the magnetic field
strength H through the magnetic susceptibility χM :
M = χM H. (A.8)
The magnetic flux density B is related to H and M through:

B = µ0 (H + M )
= µ0 (1 + χM )H (A.9)
= µ0 µr H,
where µ0 is the magnetic permeability of the vacuum and µr = 1 + χM
is the relative magnetic permeability, of the medium. The value of
µ0 is 4π × 10−7 H m−1 in SI units.
The laws that describe the combined electric and magnetic response
of a medium are summarized in Maxwell’s equations of electromag-
netism:
∇·D = (A.10)
∇·B = 0 (A.11)
∂B
∇×E = − (A.12)
∂t
∂D
∇×H = j+ , (A.13)
∂t
where is the free charge density, and j is the current density. The
first of these four equations is Gauss’s law of electrostatics (eqn A.6)

written in terms of D rather than E. The second is the equivalent of
Gauss’s law for magnetostatics with the assumption that free magnetic
monopoles do not exist. The third equation combines the Faraday and
Lenz laws of electromagnetic induction. The fourth is a statement of
Ampere’s law, with the second term on the right-hand side to account
for the displacement current.
The second Maxwell equation naturally leads to the concept of the
vector potential. This is defined through the equation
B = ∇×A . (A.14)
It is apparent that the vector potential A automatically satisfies eqn A.11,

because ∇ · (∇×A) = 0 for all A. By substituting for B in the third
Maxwell equation (eqn A.12) using eqn A.14, we see that:

∂ ∂A
∇×E = − (∇×A) = ∇× − . (A.15)
∂t ∂t
The solution is
∂A
E=− + constant , (A.16)
∂t
A.2 Electromagnetic waves 333
where the constant is any vector whose curl is zero. If the scalar potential
is V , then we can combine eqn A.16 with eqn A.7 by writing:
∂A
E =− − ∇V . (A.17)
∂t
This works because ∇×∇V = 0. The more general definition of E given
in eqn A.17 reduces to eqn A.7 when the magnetic field does not vary
with time, and to
∂A
E =− (A.18)
∂t
when the scalar potential is constant throughout space.
It is important to note that the definition of A in eqn A.14 does not
define the vector potential uniquely. We can add any vector of the form
∇ϕ to A without changing B because
∇×(A + ∇ϕ) = ∇×A + ∇×(∇ϕ) = ∇×A . (A.19)
ϕ(r) can be any scalar function of r. For this reason, we have to give an
additional definition, which specifies the gauge in which we are working.
The Coulomb gauge is defined by
∇ · A = 0. (A.20)
This gauge is convenient because it allows us to recover Poisson’s equa-

tion (A.5) by taking the divergence of eqn A.17. The vector potential in
the Coulomb gauge is used in the semi-classical treatment of the inter-
action of light with atoms discussed in Section B.2 of Appendix B.
A.2 Electromagnetic waves

Maxwell was able to show that eqns A.10–A.13 were consistent with
wave-like solutions in a medium with no free charges or currents. To see
this we first simplify eqns A.12 and A.13 by setting j = 0 and eliminating
B and D using eqns A.3 and A.9. This gives:
∂H
∇×E = −µ0 µr , (A.21)
∂t
and
∂E
∇×H = 0 r
. (A.22)
∂t
We then take the curl of eqn A.21 and eliminate ∇×H using eqn A.22.
This gives:
∂2E
∇×(∇×E) = − µ0 µr 0 r . (A.23)
∂t2
The left-hand side can be simplified by using the vector identity
∇×(∇×E) = ∇(∇ · E) − ∇2 E . (A.24)
Equation A.6 with = 0 tells us that ∇ · E = 0. Therefore we obtain

the final result:
∂2E
∇2 E = µ0 µr 0 r . (A.25)
∂t2
Equation A.25 is of the same form as the wave equation:
∂2y 1 ∂2y
2
= 2 2, (A.26)
∂x v ∂t
where v is the velocity of the wave. We therefore identify eqn A.25 as
describing electromagnetic waves with a phase velocity v given by
1
= µ0 µr 0 r . (A.27)
v2
In free space r = µr = 1 and the velocity of the wave is c, so we have:
1
c= √ = 2.998 × 108 m s−1 . (A.28)
µ0 0
At the same time, we see from eqns A.27 and A.28 that the velocity in
a medium is given by
1
v=√ c. (A.29)
r µr
We define the refractive index n of the medium as the ratio of the
velocity of light in free space to the velocity in the medium:
c
n= . (A.30)
v
At optical frequencies we can set µr = 1, and thus conclude:
√
n= r . (A.31)
This allows us to relate the propagation constants of electromagnetic

waves in a medium to the dielectric constant.
The use of complex solutions of the The solutions to eqn A.25 are of the form:
type given in eqn A.32 simplifies the
mathematics and is used extensively E(z, t) = E 0 ei(kz−ωt) , (A.32)
throughout this book. Physically mea-
surable quantities are obtained by tak-
ing the real part of the complex wave. where E 0 is the amplitude of the wave, z is the direction of propagation,
In some texts i is replaced by −j, but k is the wave vector, and ω is the angular frequency. The magnitude of
this makes no physical difference. the wave vector k is given by:
2π ω nω
k= = = , (A.33)
λ v c
where λ is the wavelength inside the medium. The first equality is the
Note that the word ‘polarization’ is definition of k, the second follows by substitution of eqn A.32 into
used both for the dielectric polarization eqn A.25 with v given by eqn A.27, and the third follows from the
2 and for the direction of the electric definition of n given in eqn A.30.
field in an electromagnetic wave. It is
usually obvious from the context which The direction of the electric field in an electromagnetic wave is called
meaning is intended. the polarization. Several different types of polarization are possible.
• Linear: the electric field vector points along a constant direction.

• Circular: the electric field vector rotates as the wave propagates,
mapping out a circle for each cycle of the wave.
(a) x
• Elliptical: this is similar to circular polarization, except that the
rotating electric field vector maps out an ellipse rather than a circle
as the wave propagates. E
• Unpolarized: the light is randomly polarized.
z
In free space the polarization of a wave is constant as it propagates.
However, in anisotropic or chiral materials, the polarization can change k
as the wave propagates. (See Sections 2.5 and 2.6.) y H
Figure A.1 depicts a linearly polarized wave propagating along the
z axis with the polarization along the x axis. If the beam is travelling
parallel to a horizontal optical bench with the x axis perpendicular to
the surface, the x-polarized wave shown in Fig. A.1(a) is said to be
(b)
Ex ,Hy
vertically polarized. A y-polarized beam would similarly be called
horizontally polarized. It is apparent from eqn A.21 or A.22 that
the magnetic field must be perpendicular to the electric field. E, H,
and k therefore form a right-handed system as depicted in Fig. A.1(a). z
The orthogonal electric and magnetic fields vary sinusoidally in space,
as shown in Figure A.1(b).
For the x-polarized wave propagating along the z direction shown in
Fig. A.1(a), the components of the complex fields are of the form: Fig. A.1 (a) The electric and magnetic
fields of an electromagnetic wave form a
E x (z, t) = E x0 ei(kz−ωt) right-handed system. The figure shows
E y (z, t) = 0 the directions of the fields in a wave po-
(A.34) larized along the x axis and propagat-
Hx (z, t) = 0 ing in the z direction. (b) Spatial varia-
tion of the electric and magnetic fields.
Hy (z, t) = Hy0 ei(kz−ωt) ,
where k is the magnitude of the wave vector defined in eqn A.33 and ω
is the angular frequency. On substituting the fields from eqn A.34 into
eqn A.21, we find that:
k E x0 = µ0 µr ω Hy0 , (A.35)
and hence that:
E x0
Hy0 =
, (A.36)
Z
where
k µ0 µr 1
Z= = = . (A.37)
µ0 µr ω 0 r c0 n
The second equality in eqn A.37 follows from eqns A.33 and A.27, while
the third follows from eqns A.28 and A.31 with µr = 1. The quantity
Z is called the wave impedance, and takes the value of 377 Ω in free
space.
In general, an electromagnetic wave propagating in the z direction will
have electric field components along both the x and y axes, with:
E x (z, t) = E x0 ei(kz−ωt) ,
(A.38)
E y (z, t) = E y0 ei(kz−ωt+∆φ) .
Table A.1 Relative amplitudes and phases for the orthogonal compo-
nents of the electric field for various types of polarization. E x0 , E y0 , and
∆φ are defined in eqn A.38.
Polarization Relative amplitude Relative phase ∆φ
Linear any 0, π
Positive circular + E x0 = E y0 +π/2
Negative circular E x0 = E y0 −π/2
Elliptical E x0 = E y0 ±π/2
E x0 = E y0 = 0, ±π/2, or π
Unpolarized random random
This can be represented in short-hand form as:
E 0 = E x0 x̂ + E y0 ei∆φ ŷ . (A.39)
Table A.1 summarizes the effect of varying the relative amplitudes and
phases of the two orthogonal components.
For the case of linearly polarized light, the direction of the polarization
is given by the resultant of (E x0 x̂±E y0 ŷ), where the + sign is used when
∆φ = 0 and the − sign when ∆φ = π.
For circularly polarized light, we can either have right circular po-
larization or left circular polarization depending on whether the
electric field vector rotates to the right (clockwise) or left (anti-clockwise)
in a fixed plane as the observer looks towards the light source. In σ +
circular polarization, the electric field rotates clockwise as seen from
the source, making it equivalent to left circular light, and vice versa for
σ − polarization. Since the phase difference between the two orthogonal
linear components is ±π/2, we can represent σ + and σ − light in the

following form:
√
σ + = E 0 (x̂ + iŷ)/ 2 ,
√ (A.40)
σ − = E 0 (x̂ − iŷ)/ 2 .
These can easily be inverted to demonstrate that linearly polarized light

can be considered to consist of two opposite circular polarizations. For
example, for x-polarized light we have:
√
E x ≡ E 0 x̂ = (σ + + σ − )/ 2 . (A.41)
The energy flow in an electromagnetic wave can be calculated from

In evaluating the Poynting vector for the Poynting vector:
complex fields of the form given in I = E×H . (A.42)
eqn A.32, the real parts must be taken
before the multiplication is performed. This gives the power flow per unit area in W m−2 , which is equal to
the intensity of the light wave. The intensity is defined as the energy
crossing a unit area in unit time, and is therefore given by:
I = vuν , (A.43)
where v is the velocity of the wave and uν is the energy density per unit
volume of the beam. For a linearly polarized wave, the Poynting vector
can easily be evaluated by substituting eqns A.34–A.37 into eqn A.42 to
obtain:
E(t)2 rms 1
I= = c0 nE 20 , (A.44)
Z 2
where E(t)2 rms represents the root-mean-square time average. This
shows that the intensity of a light wave is proportional to the square
of the amplitude of the electric field. The relationship can be general-
ized for all light waves irrespective of the particular polarization of the
beam.
In many topics covered in this book, it will be necessary to treat
the refractive index as a complex number. A well-known example of
how such a situation arises occurs when treating the propagation of
electromagnetic waves through a conducting medium such as a metal.
In a conductor, the current density is related to the electric field through
the electrical conductivity σ according to:
j = σE. (A.45)
By using this relationship to substitute for j in eqn A.13, and eliminating

D, B, and H in the same way that led to eqn A.25, we obtain:
∂E ∂2E
∇2 E = σµ0 µr + µ0 µr 0 r 2 . (A.46)
∂t ∂t
We now look for plane wave solutions of the type given by eqn A.32.
Substitution of eqn A.32 into eqn A.46 gives:
k 2 = σµ0 µr ω i + µ0 µr 0 r ω 2 . (A.47)
This can be made compatible with the usual relationship between ω and
k given in eqn A.33 by allowing n to be a complex number. The complex
refractive index is usually written ñ, and is defined by
ω
k = ñ . (A.48)
c
By combining eqns A.47 and A.48 we obtain: It is shown in Chapter 7 that eqn A.49
is only valid at low frequencies. This
µr σ is because the AC conductivity at high
ñ2 = i + µr r , (A.49) frequencies is not the same as the DC
0 ω
conductivity that enters eqn A.45.
where we have made use of eqn A.28. This of course reduces to eqn A.31
if we set σ = 0 and µr = 1. The physical significance of the complex
refractive index implied by eqn A.49 is developed in more detail in Sec-
tion 1.3.
The Maxwell equations also allow us to treat the transmission and
reflection of light at an interface between two materials. This situation
is depicted in Fig. A.2. Part of the beam is transmitted into the medium
and the rest is reflected. The solution for an arbitrary angle of incidence
air or vacuum optical medium

n=1 refractive index ñ
Exi
Fig. A.2 Transmission and reflection
of light at an interface between air and
a medium of refractive index ñ. The in- incident Ext
cident, transmitted and reflected rays Hyi
are shown displaced from each other for
clarity. All rays are normal to the in- Exr transmitted
terface. The symbol for the magnetic Hyt
fields of the incident and transmitted
reflected
rays indicates that the field direction
is out of the page, while the symbol ⊗ Hyr
for the reflected wave indicates that the
field is pointing in to the page.
was treated by Fresnel, and the resulting formulæ are known as Fres-
nel’s equations. We restrict ourselves here to the simpler case when
the angle of incidence is zero: that is, normal incidence.
We consider again an x-polarized light beam propagating in the z
direction, with the field directions as shown in Fig. A.1(a). The electric
and magnetic fields are given by eqn A.34. The beam is incident on a
medium with complex refractive index ñ. The fields are related to each
other through eqn A.36, with Z given by eqn A.37, although we now
have to allow for the possibility that n may be complex.
The boundary conditions at the interface between two dielectrics tell
us that the tangential components of the electric and magnetic fields are
continuous. On applying this to the situation shown in Fig. A.2, we must
have that both E x and Hy are conserved across the boundary. Hence we
can write:
E ix + E rx = E tx , (A.50)
and
Hyi − Hyr = Hyt , (A.51)
where the superscript labels i, r, and t refer to the incident, reflected,
and transmitted beams respectively. By making use of the relationship
between the magnetic and electric fields given in eqns A.36–A.37, we
can rewrite eqn A.51 as:
E ix − E rx = ñE tx , (A.52)
where we have assumed that the light is incident from air with ñ = 1 and
that µr = 1 at the optical frequencies of interest here. Equations A.50
and A.52 can be solved together to obtain
E rx ñ − 1
= . (A.53)
Ex i ñ + 1
This can be re-arranged to obtain the required result for the reflectivity
R: r 2
Ex ñ − 1 2

R= i = . (A.54)
Ex ñ + 1
Further reading 339
For the more general case where the light is reflected at the interface
between two materials with complex refractive indices of ñ1 and ñ2 re-
spectively, this becomes:

ñ2 − ñ1 2
R= . (A.55)
ñ2 + ñ1
These formulæ are used in many examples throughout the book.
Further reading
The subject matter of this appendix is standard electro- (1990), or Lorrain, Corson, & Lorrain (2000). The sub-
magnetism, and there are numerous books on the market ject is also covered in many optics texts such as Born
that cover the material, for example: Bleaney & Bleaney & Wolf (1999), Hecht (2001), or Smith, King, & Wilkins
(1976), Duffin (1990), Good (1999), Grant & Phillips (2007).

Appendix: Electromagnetism in Dielectrics

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Appendix: Electromagnetism in Dielectrics

Uploaded by

Copyright:

Available Formats

Appendix Electromagnetism in

A.1 Electromagnetic ﬁelds and Maxwell’s

• the electric ﬁeld strength E;

• the electric displacement D.

The microscopic response of the material is determined primarily by the

This may be considered to be the deﬁnition of D. By combining eqns A.1

Equation A.5 follows directly by substituting for E in eqn A.6 using

The magnetic ﬂux density B is related to H and M through:

ﬁrst of these four equations is Gauss’s law of electrostatics (eqn A.6)

It is apparent that the vector potential A automatically satisﬁes eqn A.11,

∇×(A + ∇ϕ) = ∇×A + ∇×(∇ϕ) = ∇×A . (A.19)

This gauge is convenient because it allows us to recover Poisson’s equa-

A.2 Electromagnetic waves

∇×(∇×E) = ∇(∇ · E) − ∇2 E . (A.24)

Equation A.6 with  = 0 tells us that ∇ · E = 0. Therefore we obtain

This allows us to relate the propagation constants of electromagnetic

• Linear: the electric ﬁeld vector points along a constant direction.

Polarization Relative amplitude Relative phase ∆φ

This can be represented in short-hand form as:

linear components is ±π/2, we can represent σ + and σ − light in the

These can easily be inverted to demonstrate that linearly polarized light

The energy ﬂow in an electromagnetic wave can be calculated from

By using this relationship to substitute for j in eqn A.13, and eliminating

air or vacuum optical medium

You might also like

Equation A.6 with = 0 tells us that ∇ · E = 0. Therefore we obtain