The Wave Equation and

the Speed of Light

Chapter 1
Physics 208, Electro-optics
Peter Beyersdorf

Document info Ch 1
 Class Outline
Maxwells equations

Boundary conditions

Poyntings theorem and conservation laws

Complex function formalism

time average of sinusoidal products

Wave equation

Ch 1, 2
 Maxwells Equations

Electric field (E) and magnetic field (H) in free-
space can be generalized to the electric

displacement (D) and the magnetic induction (B)
that include the effects of matter. Maxwells
equations relate these vectors
!
B
! !
E+ = 0 Faradays law
t
D!
! !
H = J! Amperes law
t
! D
! = Gauss law (for magnetism)
! B
! = 0 Gauss law (for electricity)

What do each of these mean? Ch 1, 3

 Faradays Law
!
B
! E
!+ = 0
t
The Curl of the electric field is caused by
changing magnetic fields

A changing magnetic field can produce electric

fields with field lines that close on themselves

Ch 1, 4
 Amperes Law
!
D
! H
! = J!
t
The Curl of the magnetic field is caused by
current of charged particles (J) or of the field
they produce (dD/dt)

A changing electric field can produce magnetic

fields (with field lines that close on themselves)

For all cases considered in this class, J=0

Ch 1, 5
 Gauss Law
! D
! =
Electrical charges are the source of the electric field
! = "E
D ! = "0 E
! + P!

For all cases considered in this class, =0

is a 3x3 tensor not a scalar (unless the material is

isotropic)!

may be a function of E and H! (giving rise to non-

linear optics)

can be determined via measurements on a parallel

plate capacitor filled with a given material using the 6

equation C=A/d Ch 1,
Gauss Law for Magnetism

There are no source of magnetic fields

No magnetic monopoles
Magnetic field lines can only circulate

! = H
B ! = 0 H
! +M
!
is a 3x3 tensor not a scalar (unless the material is
isotropic)!

may be a function of E and H! (giving rise to non-

linear optics)

can be measured using the Biot-Savart law

Ch 1, 7
Waves and Maxwells Equations

A charged particle is a source of an electric field

When that particle moves it changes the (spatial

distribution of) the electric field

When the electric field changes it produces a

circulating magnetic field

If the particle accelerates this circulating magnetic

field will change

A changing magnetic field produces a circulating

electric field

The circulating electric field becomes the source of

a circulating magnetic field Ch 1, 8
 Boundary Conditions
When an EM wave propagates across an
interface, Maxwells equations must be satisfied
at the interface as well as in the bulk
materials. The constraints necessary for this to
occur are called the boundary conditions

! " !1 , 1 !2 , 2
! dA
D != dA
! "
! d!s = d ! dA
!
E B
! dt
! dA
B !=0
! " "
! d!s = J! dA!+ d ! dA
!
H D
dt

Ch 1, 9
 Boundary Conditions
Gauss law can be used to find the boundary
conditions on the component of the electric field
that is perpendicular to the interface.

If the materials are dielectrics there will be no

free charge on the surface (=0)
! " !1 , 1 !2 , 2
! dA
D != dA
! "
! d!s = d ! dA
!
E B
! dt
! dA
B !=0
! " "
! d!s = J! dA!+ d ! dA
!
H D
! dt
0
D1 D2 = dA D1 = D2

Ch 1, 10
 Boundary Conditions
Faradays law can be applied at the interface. If
the loop around which the electric field is
computed is made to have an infintesimal area
the right side will go to zero giving a
relationship between the parallel components of
the electric field
! " !1 , 1 !2 , 2
! dA
D != dA
! "
! d!s = d ! dA
!
E B
! dt
! dA
B !=0
! " "
! d!s = J! dA !+ d ! dA
!
H D
d
! dt
0
= E1! = E2!

E2! E1! B dA
dt Ch 1, 11
 Boundary Conditions
Gauss law for magnetism gives a relationship
between the perpendicular components of the
magnetic field at the interface

! " !1 , 1 !2 , 2
! dA
D != dA
! "
! d!s = d ! dA
!
E B
! dt
! dA
B !=0
! " "
! d!s = J! dA!+ d ! dA
!
H D
dt
B1 A B2 A = 0 B1 = B2 Ch 1, 12
 Boundary Conditions
Amperes law applied to a loop at the interface
that has an infintesimal area gives a relationship
between the parallel components of the
magnetic field. (Note that in most common
materials =o) In the absence of currents J=0
so
! " !1 , 1 !2 , 2
! dA
D != dA
! "
! d!s = d ! dA
!
E B
! dt
! dA
B !=0
! " "
! d!s = J! dA!+ d ! dA
!
H D
! 0 dt !
0!
H1! L H2! L = !+ d
J! dA ! H1! = H2!

D dA

Ch 1, 13
dt
 Poyntings Theorem
The flow of electromagnetic energy is given by
the Poynting vector
!=E
S ! H
!
which has a magnitude that is the power per
unit area carried by an electromagnetic wave in
the direction of S.

S [W/m2] E [v/m] H [A/m]

Ch 1, 14
Complex-Function Formalism
Steady-state sinusoidal functions of the form

a(t) = A cos(t + )

can be treated as having a complex amplitude

! = Aei
A

such that the function can be written as

! #
" it
a(t) = Re Ae

! !it! where it is understood

or in shorthand! !Ae
that the real part of this complex expression
represents the original sinusoidal function
Ch 1, 15
 Phasors
The complex amplitude of a sinusoidal function
can be represented graphically by a point (often
an arrow from the origin to a point) in the
complex plane

Im Im Im

Re Re Re

a(t) = cos t a(t) = sin t a(t) = cos (t) + sin (t)

i/4
!=1
A ! = i
A ! = 2e
A Ch 1,
16
 Phasors
Addition of same-frequency sinusoidal functions
involves factoring out the time dependance and
simply adding the phasor amplitudes.

Addition of difference frequency sinusoidal

function is often simplified by factoring out a
sinusoidal component at the average frequency.

Multiplication of sinusoidal functions can not be

done by multiplying phasors since

Re [x] Re [y] != Re [xy]

Ch 1, 17
 Phasor Example
For electric field amplitudes described by
E1 = E10 cos (1 t)
and
E2 = E20 cos (2 t)
Use the phasor representation to find a
representation of E1+E2 as a slow_ modulation of a
field at the average frequency =(1+2)/2

Ch 1, 18
 Phasor Example
5

E 1 + E2 = E10 ei1 t + E20 ei2 t

! " ! "
E E
2.5

= Eavg + ei1 t + Eavg ei2 t

2 2
! " ! "
E E
-4.8 -4 -3.2 -2.4 -1.6 -0.8 0 0.8 1.6 2.4 3.2 4 4.8

+
= Eavg + e i( 2 )t
+ Eavg ei( 2 )t

Example # 2 2
! " ! "$ -2.5

t t
E=cos(2t)+3cos(2.5t)
= 2Eavg cos + iE sin eit
2 2
# ! " ! "$ 12 -5

t t
= 4Eavg cos
2 2
+ (E) sin
2 2
eit+i
2 2
! " #$
E10 + E20 1 + 2 E t
Eavg
arctan tan
2 2 2Eavg 2
E E10 E20 1 2 Ch 1, 19
 Time Averages
Optical fields vary too fast to be directly
detected, instead it is the irradiance averaged
over many cycles that is detected as light.
!
1 T
!a(t)b(t)" = A cos(t + )B cos(t + )dt
T 0

In terms of the phasor amplitudes this is

1 ! " "#
!a(t)b(t)" = Re AB
2

Ch 1, 20
Poynting Vector Example
For electric and magnetic fields given by
E = E0 eit+
E0 it+
H = e
Z0
where !
0
Z0 = 0 377
"0

is the impedance of free space, what is the

irradiance of the wave? How much power is
measured by a detector of area A?

Ch 1, 21
Poynting Vector Example
For electric and magnetic fields given by
E = E0 eit+
E0 it+
H = e
0
where !
0
0 377
"0

is the impedance of free space, what is the

irradiance of the wave?

! "
1 # % 1 & E
'
E 2
! H
! = Re E$H
$ = Re E0 e 0
Savg = E i
ei
= 0
2 2 0 20

This is analogous to Pavg=V2/2R for AC circuits

2
!avg A
!A E
Pavg =S 0
20 Ch 1, 22
Wave Equation in Isotropic Materials

Starting with Faradays law

take the curl of both sides

use vector calculus relationship to get

Use Amperes law (in free space where J=0)

and Gauss law (in free space where =0) ! D
! =0
in an isotropic medium
Ch 1,23
 Wave Equation in Crystals

In an anisotropic medium

does not simplify as much since

!! ! !! !D
!! =! 0 ! does not imply
but rather
! D
! = ! "E
! = "
! E! +E! "
where0. In this case it is usually easiest to write the
wave equation as
2!
E ^ ^ E
D=E ^^ k ^^
! ! !
E + " 2 = 0
t E k
^ ^^
^kxE
kxE
or
! + " 2 E
!k !k E !=0
^kxkxE
^ ^^ ^^
kxkxE Ch 1,24
Spherical Solutions to the Wave Equation

Consider solutions for E such that 2E and 2E/dt2

are both proportional to E - allowing the two sides
to differ only by a constant term.

is one such solution in spherical coordinates. Using

the relationship for the Laplacian of a spherically
symmetric function:

Show that !! ! given above is a solution to the

wave equation Ch 1, 26
 Solutions to the Wave Equation

Recall the meaning of k and (k=2/, =2/T) we

can express this as

Since is the distance travelled by the wave in one

cycle, and T is the time to travel one cycle, /T is the
velocity of the wave, which can be determined from
electrostatics and magnetostatics!

Ch 1, 27
 Solutions to the Wave Equation

From our solution in free space

and Gauss law in free space (=0)

We find that since!! ! ! only has a spatial dependance on

r its divergence, given by

must be

implying ! ! ! ! ! ! meaning this must be a transverse

wave (and isnt a solution in anisotropic media) Ch 1, 28
 Speed of Light
In free-space where =0 and =0 the speed of
light is defined to be c 299792458 m/s. In this sense
any measurement of the speed of light in a vacuum is
really a measurement of the length of a meter (the unit of
time is also a defined quantity)

In material where =0 and =r0 the speed of light

is v=c/(r)1/2. We let n (r)1/2 and call n the
index of refraction for a material.

What is the physical interpretation of n?

If it is complex, what do the real and imaginary

parts represent? Ch 1, 28
 Index of Refraction
From our expression for the velocity of the wave
v=c/(r)1/2 we can substitute n=(r)1/2 to get v=c/n

Thus n represents how much slower light travels in

a material compared to free space.

Given the relation c=nv=/k. If a wave travels

from free space into a material causing it to slow
down, does change, or does k change (or both)?

Ch 1, 29
 Index of Refraction
Consider a wave in free space entering a material.
Doe the wavelength change, does the frequency
change or both?

The frequency cannot change (or else the boundary

would be discontinuous) so the wavelength (and
hence k) must change so that =0/n and k=nk0
Ch 1, 30
 Index of Refraction
Going back to the solution to the wave equation, we
can express it explicitly for propagation in a
material with index of refraction n

If n is complex such that n=n+i n

we have

and we see n is related to the absorption

coefficient used in Beers law:
by =2nk0. Ch 1, 31
 Phase Velocity
For a sinusoidal wave, or a waveform comprised of
many sinusoidal components that all propagate at
the same velocity, the waveform will move at the
phase velocity of the sinusoidal components

Weve seen already that the phase velocity is

vp=/k

What happens if the different components of the

wave have different phase velocities (i.e. because
of dispersion)?

Ch 1, 32
 Phase and Group Velocity

No dispersion (vp=vg)
E1
E2
E1+E2

Dispersion (vpvg)
E1
E2
E1+E2

Ch 1, 33
 Group Velocity
When the various frequency components of a
waveform have different phase velocities, the
phase velocity of the waveform is an average of
these velocities (the phase velocity of the
carrier wave), but the waveform itself moves at
a different speed than the underlying carrier
wave called the group velocity.

Ch 1, 34
Group vs Phase velocity
An analogy that may be useful for understanding
the difference comes from velodrome cycling:

Riders race as a team

and take turns as
leader with the old
leader peeling away
and going to the back
of the pack

As riders make their way from the rear of the

pack to the front they are moving faster than the
group that they are in
Ch 1, 35
 Group Velocity
The phase velocity of a wave is

v=
k
and comes from the change in the position of the
wavefronts as a function of time

The waveform moves at a rate that depends on the

relative position of the component wavefronts as a
function of time. This is the group velocity and is
d
vg =
dk
which can be found if you have ! "
c k dn
! != !vk! =! ! ! k! ! giving vg = v 1
n(k) n dk Ch 1, 36
 Slow Light
How slow can light be made to go?

In a Bose-Einstein Condensate light tuned to the

atomic resonance tremendous dispersion and has
been slowed to a speed of See Hau, et al. Light speed reduction to 17
metres per second in an ultracold atomic
gas, Nature 397, 594 - 598 (18 February
1999)

normal
normaldispersion
dispersion
!(")

normal dispersion

n(")
anomalous dispersion

Ch 1, 37
 Example
Given the dispersion equation
" #
N e2 ! fj
n () = 1 +
2
2 2
"0 me j 0j

where fj is the fraction of electrons that have a

resonant frequency of 0j, find the phase
velocity and group velocity of high frequency
electromagnetic waves (>>oj)

Ch 1, 38
 Example
" #
2 !
N e fj
n2 () = 1 + 2 2
"0 me j 0j
The phase velocity is v=c/n so
% 2
&
c Ne
v=! $ c 1 + 2! m 2
N e2
" # fj 0 e
1+ !0 me j 2 2
0j

The group velocity can be found from

d
vg =
dk
Ch 1, 39
 Example
" #
2 !
N e fj d
n2 () = 1 + 2 2 vg =
"0 me j 0j dk
! n
using! ! ! f!j =
! 1! and k=
j c
! 2
"
n Ne
k= 1
c c 2"0 me 2
! "
dk 1 d(n) 1 Ne 2
= +
d c d c 2"0 me 2
d c
vg = =
dk 1 + N e2 /2me "0 2 Ch 1, 40
 Warning
For the analysis so far we have treated and
as being scalars meaning the waves are
propagating through isotropic materials.

What changes if the materials are not isotropic?

Ch 1, 41
 References
Yariv & Yeh Optical Waves in Crystals chapter
1.1-1.5

Hecht Optics chapter 2.3-2.9, 7.2, 7.3, 7.6

Ch 1,42