Optical WG

The University of Texas at Dallas Erik Jonsson School of
Engineering & Computer Science
OPTICAL WAVEGUIDES
Notes prepared for EE 6310
by
Professor Cyrus D. Cantrell
August–December 2003

c C. D. Cantrell (06/2003)
The University of Texas at Dallas Erik Jonsson School
PhoTEC
GUIDED WAVES
• A waveguide is “A structure that causes a wave to propagate in a chosen

direction with some measure of confinement in the planes transverse to the
direction of propagation” (Ramo, Whinnery and Van Duzer)
Change of direction of waveguide ⇒ change of direction of wave with
negligible reflection (within limits)
Guiding is accomplished by either
◦ A relation between fields and currents/charges on the surfaces of a
metallic waveguide structure, or
◦ Total internal reflection at the interfaces of a dielectric structure
Guiding can be analyzed using either
◦ Geometrical-optics concepts
Leaves out interference, diffraction and polarization
◦ Field/wave concepts
◦ Circuit concepts (for transmission lines and plane waves in bulk media)

THE UNIVERSITY OF TEXAS AT DALLAS Erik Jonsson School of Engineering
and Computer Science
RECTANGULAR METALLIC WAVEGUIDE

ELECTROMAGNETIC MODES OF WAVEGUIDES
• A mode of a waveguide is a stable, propagating pattern of electric and mag-

netic fields that is periodic along the axis of the waveguide, apart from
attenuation
An ideal waveguide is shift-invariant in the direction of propagation
⇒ periodicity of the modes
Modes of symmetrical guiding structures can be studied analytically
◦ Rectangular and cylindrical conducting waveguides
Modes are either TE (transverse electric, Ez = 0) or
TM (transverse magnetic, Hz = 0)
◦ Layered plane dielectric waveguides
TE and TM modes
◦ Coaxial cylindrical dielectric waveguides
TE, TM and hybrid modes (both Ez = 0 and Hz = 0)
Modes of asymmetrical structures must be studied numerically

TRANSMISSION LINES
• A transmission line is a guiding structure consisting of two or more

parallel conductors and dielectrics
Guides waves without confining them inside a metal or dielectric structure
Guided waves are still confined to the vicinity of the transmission line in
planes that are transverse to the axis
• Transmission lines with low attenuation have a distinctive relation between
E and H that permits them to be analyzed either with field concepts or with
circuit concepts
They support TEM (transverse electromagnetic) modes
E and H are perpendicular to the axis of propagation (both Ez = 0 and
Hz = 0)

WAVE EQUATION FOR A WAVEGUIDE (1)
• Maxwell’s equations lead to the wave equation

∂E ∂ 2E
∇ E = µσ
2
+ µ 2
∂t ∂t
Let
jωt−γ̂z
E(x, y, z, t) = Re Ê(x, y)e
where
γ̂ 2 = jωµ(σ + jω)
Separate γ̂ into real and imaginary parts:
γ̂ = α + jβ
α = attenuation coefficient
β = propagation constant
E(x, y, z, t) is a wave propagating in the +z direction if β > 0

• Eliminating Êy from the curl equations leads to an equation for Ĥx in terms
of the z components only:
∂ Êz ∂ Ĥz
(γ̂ 2 + k 2)Ĥx = jω − γ̂
∂y ∂x
where
k 2 = ω 2µ
• Similarly,
∂ Êz ∂ Ĥz
−(γ̂ + k )Ĥy = jω
2 2
+ γ̂
∂x ∂y
∂ Ĥz ∂ Êz
−(γ̂ 2 + k 2)Êx = jωµ + γ̂
∂y ∂x
∂ Ĥz ∂ Êz
(γ̂ 2 + k 2)Êy = jωµ − γ̂
∂x ∂y

• The equations on the previous slide make it possible to find Êx, Êy , Ĥx, Ĥy
in terms of the z-components Êz , Ĥz
This works as long as at least one of Êz , Ĥz is non-zero
◦ In this case we can solve a wave equation for Êz or Ĥz (or both)
◦ Don’t have to compute all six components of E and H
If both Êz , Ĥz are zero, then we have a TEM (transverse electromagnetic)
mode (see next slide)

PhoTEC
TEM MODES (1)
• The z-components of Ê and Ĥ (the components along the axis of

propagation) can vanish if and only if
γ̂ 2 + k 2 = 0
This is the condition for the existence of a TEM mode
Since γ̂ 2 = −k 2, γ̂ = ±jk
But γ̂ = α + jβ; then
β = ±k
⇒ waves on a transmission line have the same phase velocity as
electromagnetic waves in bulk dielectric
Also, γ̂ 2 = jωµ(σ + jω) = −ω 2µ + jωµσ and γ̂ 2 = −k 2 imply that
σ=0
⇒ zero attenuation is a necessary condition for an ideal TEM mode

PhoTEC
TEM MODES (2)
• The equation that is satisfied by Ê(x, y) for a TEM mode is

2
∂ ∂2
+ Ê(x, y) = 0
∂x2 ∂y 2
There’s a similar equation for Ĥ(x, y)
This is Laplace’s equation
◦ The same equation is satisfied by the electrostatic potential
Mode field patterns (field lines of Ê(x, y) and Ĥ(x, y)) for a TEM mode
look like equipotentials
Because Ê(x, y) must vanish on perfectly conducting walls, and because
a solution of Laplace’s equation can’t have a maximum or a minimum
inside a closed region, there are no TEM modes inside closed
conducting waveguides
Another condition for the existence of a TEM mode is that the dielectric
must be homogeneous (same value of everywhere)

NON-TEM MODES (1)
• Maxwell’s equations lead to the wave equation

∂E ∂ 2E
∇ E = µσ
2
+ µ 2
∂t ∂t
Let
jωt−γ̂z
E(x, y, z, t) = Re Ê(x, y)e
where
γ̂ 2 = jωµ(σ + jω)
The equation satisfied by Êz (x, y) (and Ĥz (x, y)) for a general mode is
2
∂ ∂ 2 2 2

2
+ 2 Êz (x, y) + γ̂ + κ Êz (x, y) = 0
∂x ∂y
where
κ2 = k 2 − jωµσ
and
k 2 = ω 2µ

NON-TEM MODES (2)
• For now, specialize to zero attenuation (σ = 0):

2 2

∂ ∂ 2
+ Ê z (x, y) + k c Êz (x, y) = 0
∂x2 ∂y 2
where
kc2 = k 2 − β 2
Êz (x, y) and Ĥz (x, y) satisfy the Helmholtz equation
◦ Generalizes the 1-dimensional oscillator equation
d2f 2
+ k c f =0
dx2
kc is the wavenumber of transverse oscillations of Êz (x, y) if k 2 − β 2 > 0
Mathematically, kc is an eigenvalue
◦ Each value of kc that is permitted by the boundary condi-
tions defines a mode

NON-TEM MODES (3)
• The transverse field components Êx, Êy , Ĥx, Ĥy can be found from the
longitudinal components Êz , Ĥz using the equations
∂ Êz ∂ Ĥz
2 2
(γ̂ + k )Ĥx = jω − γ̂
∂y ∂x
∂ Êz ∂ Ĥz
−(γ̂ 2 + k 2)Ĥy = jω + γ̂
∂x ∂y
∂ Ĥz ∂ Êz
−(γ̂ + k )Êx = jωµ
2 2
+ γ̂
∂y ∂x
∂ Ĥz ∂ Êz
2 2
(γ̂ + k )Êy = jωµ − γ̂
∂x ∂y
A mode with Êz = 0, Ĥz = 0 is called transverse electric (TE)
A mode with Êz = 0, Ĥz = 0 is called transverse magnetic (TM)
A mode with Êz = 0, Ĥz = 0 is called hybrid

NON-TEM MODES (4)
• From the equation k 2 − β 2 = kc2 one gets the dispersion relation for a
non-TEM mode:
β = k 2 − kc2
• The cutoff frequency ωc is such that
µωc2 = kc2
A wave with frequency ω > ωc cannot propagate in the mode that
corresponds to kc
Subsequent slides illustrate TE and TM modes, and the concept of cutoff
frequency, for a planar conducting waveguide

PLANAR CONDUCTING WAVEGUIDES (1)
• In a wide planar conducting waveguide, E and H are independent of y

Êz and Ĥz depend on x but not on y
• For a TM mode, Êz = 0 and Ĥz = 0
Because the tangential components of E vanish on the waveguide’s
perfectly conducting walls,
Êz (0) = Êz (d) = 0 (see figure on next slide)
• The differential equation for Êz is
d2
2
Êz (x) = −k 2
c Êz (x)
dx
The solution is
Êz (x) = A cos(kcx) + B sin(kcx)
where A = 0 because of the boundary conditions
For the TMn mode, kcd = nπ where n is a non-negative integer

0
Z
Y

• For a TE mode, Êz = 0 and Ĥz = 0

Because the tangential components of E vanish on the waveguide’s
perfectly conducting walls,
Êy (0) = Êy (d) = 0
To get Êy in terms of Ĥz , we use the equation
jωµ d
Êy (x) = 2 Ĥz (x)
kc dx
• The differential equation for Ĥz is
d2
2
Ĥ z (x) = −k c Êz (x) ⇒ Ĥz (x) = A cos(kc x) + B sin(kc x)
2
dx
The tangential electric field is given by
kcÊy (x) = jωµ (−A sin(kcx) + B cos(kcx))
where B = 0 because of the boundary conditions
For the TEn mode, kcd = nπ where n is a non-negative integer

• Dispersion relation:

nπ
2 √ 2
n2ωc,min
β= k 2 − kc2 = µω 2 −= (ω µ) 1−
d ω2
The cutoff frequency for the TEn and TMn modes is
nπ
ωc,n = √
d µ
The minimum cutoff frequency ωc,min occurs for n = 1:
π
ωc,min = ωc,1 = √
d µ
Waves with a frequency less than ωc,min cannot propagate in a planar
waveguide except in the TEM mode
• The TEn and TMn modes propagate at an angle θ to the walls of the
waveguide, where
kc
cos θ =
k
Dispersion relations for modes of a
planar conducting waveguide
β/ω√µε
1.0
TEM
0.8
0.6
0.4
TE1, TM1 TE2, TM2 TE3, TM3 TE4, TM4
0.2
ω/ωc
1 2 3 4 5
• The TEn and TMn modes propagate at an angle θ to the walls of the
waveguide, where
kc
cos θ =
k
k 2 = kc2 + β 2 ⇒ kc, β and k are the sides of a right triangle
◦ k is the actual wavenumber of a wave propagating in the dielectric
◦ kc is the wavenumber measured perpendicular to the waveguide axis
◦ β is the wavenumber measured along the axis of the waveguide
The dispersion relation kcd = nπ is equivalent to the condition
2kd cos θ = 2nπ
◦ 2kcd = 2nπ is the condition for constructive interference of waves
successively reflected from the bottom wall of the waveguide
◦ Also the condition for destructive interference (at the position of the
top wall) of an “incident” wave at angle θ with the wave reflected from
the bottom wall of the waveguide

PLANAR DIELECTRIC WAVEGUIDES (1)
• A simple planar dielectric waveguide has 3 layers:

The substrate (s)
The film (f ), in which a guided lightwave propagates
The cladding or cover (c)
• We assume that the waveguide is very wide (in the y direction)
Ê and Ĥ depend only on x
The E fields in the 3 layers of the waveguide are

Ec(x, y, z, t) = Re Êc(x)ej(ωct−βcz)

Ef (x, y, z, t) = Re Êf (x)ej(ωf t−βf z)

j(ωs t−βs z)
Es(x, y, z, t) = Re Ês(x)e
Similar equations apply to the H fields

}c
}f
X
}s
0
Z

FABRICATION OF PLANAR DIELECTRIC WAVEGUIDES
http://www.lightwavemicro.com/PDFs/Lamwhitepaper.pdf
• Assume a TE mode in this and subsequent slides

Êy = 0, Êz = 0
Continuity of Êy at the s − f and f − c interfaces leads to
ωc = ωf = ωc = ω
βc = βf = βc = β
(same frequency and propagation constant in all 3 layers)
Let
√ nc ω
k c = ω µc c =
c
(where nc is the refractive index of c), and similarly for s and f
kc is the propagation constant of a plane wave in bulk cladding material
In a useful waveguide,
nf > nc and nf > ns

• The equations for Êy in the 3 layers are

d2
2
Ê c,y (x) + (k 2
c − β 2
)Êc,y (x) = 0
dx
d2
Êf,y (x) + (k 2
f − β 2
)Êf,y (x) = 0
dx2
d2
2
Ê s,y (x) + (k 2
s − β 2
)Ês,y (x) = 0
dx
Guided mode: Transverse oscillations in f , evanescent (decaying)
waves in c and s
kf2 − β 2 > 0, kc2 − β 2 < 0, ks2 − β 2 < 0
Radiation mode: Strong radiative coupling outside of c
kf2 − β 2 > 0, kc2 − β 2 > 0, ks2 − β 2 > 0
Leaky mode: Transversely damped waves in all 3 layers
kf2 − β 2 < 0, kc2 − β 2 < 0, ks2 − β 2 < 0

• For simplicity, assume that the c and s layers are very thick; then
lim Êc,y (x) = 0, lim Ês,y (x) = 0
x→∞ x→−∞
• From the differential equations for Êc,y and Ês,y ,

Êc,y (x) = Êc,y (d)e−(x−d)/δc (x > d)
Ês,y (x) = Ês,y (0)ex/δs (x < 0)
where
1 1
δc = , δs =
β 2 − kc2 β 2 − ks2
are the penetration distances of the evanescent waves into c and s
• From the differential equation for Êf,y ,
Êf,y (x) = Êf cos(κf x − φs)

where κf = kf2 − β 2, and φs is determined by the boundary conditions

• Boundary conditions on Êy :

Êy (x) must be continuous at x = 0 and x = d; then
Ês,y (0) = Êf cos φs
Êc,y (d) = Êf cos(κf d − φs)
Ĥz must be continuous at x = 0 and x = d ⇒ dÊy /dx must be continuous
at x = 0 and x = d; then
1
Ês,y (0) = κf Êf sin φs
δs
1
− Êc,y (d) = −κf Êf sin(κf d − φs)
δc
Then
1 1
tan φs = and tan(κf d − φs) =
κf δs κf δc

• Derivation of the TE dispersion relation for a planar dielectric waveguide:

The boundary conditions imply that
1 1
tan φs = and tan(κf d − φs) =
κf δs κf δc
Let
1
φc = κf d − φs − νπ ⇒ tan φc =
κf δc
where ν is a non-negative integer
Since
1 1
κf = kf2 − β 2, δs = , δc =
β − ks
2 2 β 2 − kc2
where kf , ks and kc are known, the relation κf d − φc − φs = νπ
constrains β to have one of a discrete set of values

• The dispersion relation for TE modes in a planar dielectric waveguide is

κf d − φc − φs = νπ
2κf d − 2φc − 2φs = 2νπ is the condition for constructive interference of
waves successively reflected from the s − f interface
For the purpose of making a numerical evaluation, define the
normalized frequency

√
V = ωd µf f n2f − n2s
and the square of the normalized waveguide refractive index
n2ef f − n2s
b= 2
nf − n2s
where the effective refractive index is
β
nef f = √
ω µf f

• Rewrite the TE dispersion relation in terms of V and b:

κf in terms of the effective refractive index:

√ √ √
β = nef f ω µf f , kf = nf ω µf f ⇒ κf = ω µf f n2f − n2ef f
n2f − n2s , etc., in terms of V and b:

2 2
V κf
nf − ns =
2 2
√ , n2f − n2ef f = √
ωd µf f ω µf f
2
n2ef f − n2s n2f − n2ef f κf d √
1−b=1− 2 = 2 = ⇒ V 1 − b = κf d
nf − ns 2 nf − ns 2 V
Similarly,

b 1 b+a 1 n2s − n2c
= and = where a = 2
1 − b κf δs 1 − b κf δc nf − n2s
(a is the asymmetry parameter)

• Dispersion relation for TE modes in terms of V , b and a:

√ b b+a
V 1 − b = νπ + tan−1 + tan−1
1−b 1−b
At cutoff (b = 0) for the lowest TE mode (ν = 0), the value of V is
−1
√
V0 = tan a
Note that V0 = 0 for a symmetric slab waveguide (nc = ns ⇒ a = 0)
The effective thickness of the waveguide is
def f = d + δs + δc
because the electromagnetic energy is essentially confined to a region of
thickness def f
The plot of b vs. V on the subsequent slide is taken from H. Kogelnik,
“Theory of Dielectric Waveguides”, in Integrated Optics (Topics in Ap-
plied Physics, Vol. 7), edited by T. Tamir (Springer-Verlag, 1982)


• Dispersion relation for TM modes in terms of V , b and a:

√ −1
2
nf b −1
2
nf b + a
V 1 − b = νπ + tan + tan
ns 1 − b
2 n2c 1 − b
At cutoff (b = 0) for the lowest TM mode (ν = 0), the value of V is

−1
nf √
2
V0 = tan 2
a
nc
Note that V0 = 0 for a symmetric slab waveguide (nc = ns ⇒ a = 0)

SUMMARY OF PLANAR DIELECTRIC WAVEGUIDES
• Points to remember after you forget the mathematics:

For a symmetric waveguide (cladding index = substrate index), the cutoff
frequency of the lowest TE and TM modes is ω = 0
At cutoff (b = 0) for the next lowest TE or TM mode (ν = 1), and for a
symmetric waveguide, the value of V is
V1 = π
For frequencies such that 0 < V < π, a symmetric waveguide is
two-mode (only the ν = 0 TE and TM modes can propagate)
Electric field direction (polarization) in a planar dielectric waveguide:
◦ TE mode: Parallel to the interfaces
◦ TM mode: Perpendicular to the interfaces

OPTICAL FIBER WAVEGUIDES (1)
• Analysis involves Bessel functions and a lot of vector calculus

For a thorough treatment, take EE 6314 or read “Fiberoptic Mode Func-
tions: A Tutorial”, by Cantrell and Hollenbeck
In a step-index fiber there are only two media, the core (refractive index
n1) and the cladding (refractive index n2)
The mode that has a cutoff frequency equal to 0 is a hybrid electric
mode (HE11), in which both Êz and Ĥz are non-zero
◦ There are two polarizations with the same frequency and propagation
constant (degenerate HE11 modes with different polarizations)
◦ For 0 < V < 2.405... the HE11 mode is the only mode that can
propagate (all others are below the cutoff frequency)
◦ This is called “single-mode” operation, despite the fact that there
are two degenerate modes...

OPTICAL FIBER WAVEGUIDES (2)
• The normalized frequency and normalized propagation constant are

β n2
−
n k n
V = k0a n21 − n22 and b = 1 0 n2 1
1−
n1
where a is the core radius
• The plot of b vs. V on the subsequent slide is taken from Govind P. Agrawal,
Fiber-Optic Communication Systems, Second Edition (New York,
Wiley, 1997), p. 35
• The field line plots on later slides are taken from Takanori Okoshi, Optical
Fibers (New York, Academic Press, 1982), pp. 68–71

HE11
EH11
HE21
HE31
TE01
TM01
PhoTEC
ARRAYED WAVEGUIDE “GRATINGS”
• Arrayed waveguide gratings (AWGs) are both

Versatile imaging devices
◦ Belonging to a class of devices called planar lightwave circuits
Multiple-beam inteferometers
• AWGs provide many functions necessary for wavelength-division
multiplexing (WDM)
• AWGs can be fabricated on wafers
Economies of large-scale manufacturing and integration with other devices
• AWGs were proposed by M. K. Smit, “New focusing and dispersive planar
component based on an optical phased array,” Electronics Letters 24, 385–
386 (1988)

PhoTEC
BASIC PRINCIPLES OF AWGs
• An AWG consists of:

An input slab waveguide that couples the incoming light into an array of
waveguides
An array of planar waveguides in which adjacent waveguides’ lengths
increase by a constant increment ∆L in order to induce a constant phase
difference between adjacent paths
◦ At a chosen central wavelength, ∆L corresponds to a phase difference
of 2πm, where m = integer
◦ Where the waveguide array meets the output slab, there is a 2-d array of
images of the input waveguide with a constant phase difference between
adjacent images — as occurs at the surface of a diffraction grating
An output slab waveguide that images the outgoing light in much the
same way as in a bulk-optics diffraction-grating spectrometer
• Refer to optics-review.pdf for further details on diffraction gratings

PhoTEC
MATERIALS AND PROCESSES FOR AWGs
• SiO2-based
Low propagation and fiber-coupling losses
Low index difference ⇒ large device size
Fabrication using standard silicon processes ⇒ low cost, high volume
• InP-based
High propagation and fiber-coupling losses
High index difference ⇒ small device size
Fabrication using compound-semiconductor processes ⇒ high cost, low
volume
Possibility of integration with active devices

PhoTEC
SILICA-BASED AWGs
6-inch silicon wafer with thirty-four 16-channel, 100-GHz AWGs
http://www.lightwavemicro.com/PDFs/Lamwhitepaper.pdf
PhoTEC
PACKAGING CHALLENGES FOR AWGs
• Fiber alignment
• Stress-induced birefringence
Intrinsic stress due to different thermal expansion coefficients for
waveguide and substrate or cladding
Extrinsic stress due to different thermal expansion coefficients for different
packaging materials
Stress-induced birefringence gives an AWG different transmission
properties for different states of polarization
• Maintaining thermal stability of device
Lack of thermal stability causes problems of fiber alignment and stress-
induced birefringence, and changes the device’s transmission properties

PhoTEC
APPLICATIONS OF AWGs
• WDM/DWDM multiplexing/demultiplexing
One input fiber, N output fibers
Scalable to large number of outputs
• Wavelength routers
N input fibers, N output fibers
The N wavelengths on input port 1 are distributed, one to each output
port 1 to N
The N wavelengths on input port 2 are distributed similarly, but rotated
by one port
• Multiwavelength lasers
• Multiwavelength receivers
Monolithic integration on InP
Hybrid integration with SiO2-based waveguides

PhoTEC
AWG-BASED WDM DEMULTIPLEXER
01/2001

Optical WG

Uploaded by

Copyright:

Available Formats

Optical WG

Uploaded by

Document Information

Original Description:

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

Optical WG

Uploaded by

Copyright:

Available Formats

The University of Texas at Dallas Erik Jonsson School of

Engineering & Computer Science

Notes prepared for EE 6310

Professor Cyrus D. Cantrell

• A waveguide is “A structure that causes a wave to propagate in a chosen

RECTANGULAR METALLIC WAVEGUIDE

ELECTROMAGNETIC MODES OF WAVEGUIDES

• A mode of a waveguide is a stable, propagating pattern of electric and mag-

• A transmission line is a guiding structure consisting of two or more

WAVE EQUATION FOR A WAVEGUIDE (1)

• Maxwell’s equations lead to the wave equation

WAVE EQUATION FOR A WAVEGUIDE (2)

WAVE EQUATION FOR A WAVEGUIDE (3)

TEM MODES (1)

• The z-components of Ê and Ĥ (the components along the axis of

TEM MODES (2)

• The equation that is satisﬁed by Ê(x, y) for a TEM mode is

NON-TEM MODES (1)

• Maxwell’s equations lead to the wave equation

NON-TEM MODES (2)

• For now, specialize to zero attenuation (σ = 0):

NON-TEM MODES (3)

NON-TEM MODES (4)

PLANAR CONDUCTING WAVEGUIDES (1)

• In a wide planar conducting waveguide, E and H are independent of y

PLANAR CONDUCTING WAVEGUIDES (2)

PLANAR CONDUCTING WAVEGUIDES (3)

• For a TE mode, Êz = 0 and Ĥz = 0

PLANAR CONDUCTING WAVEGUIDES (4)

PLANAR CONDUCTING WAVEGUIDES (5)

PLANAR DIELECTRIC WAVEGUIDES (1)

• A simple planar dielectric waveguide has 3 layers:

PLANAR DIELECTRIC WAVEGUIDES (2)

FABRICATION OF PLANAR DIELECTRIC WAVEGUIDES

PLANAR DIELECTRIC WAVEGUIDES (3)

• Assume a TE mode in this and subsequent slides

PLANAR DIELECTRIC WAVEGUIDES (4)

• The equations for Êy in the 3 layers are

PLANAR DIELECTRIC WAVEGUIDES (5)

• From the diﬀerential equations for Êc,y and Ês,y ,

PLANAR DIELECTRIC WAVEGUIDES (6)

• Boundary conditions on Êy :

PLANAR DIELECTRIC WAVEGUIDES (7)

• Derivation of the TE dispersion relation for a planar dielectric waveguide:

PLANAR DIELECTRIC WAVEGUIDES (8)

• The dispersion relation for TE modes in a planar dielectric waveguide is

PLANAR DIELECTRIC WAVEGUIDES (9)

• Rewrite the TE dispersion relation in terms of V and b:

n2f − n2s , etc., in terms of V and b:

PLANAR DIELECTRIC WAVEGUIDES (10)

• Dispersion relation for TE modes in terms of V , b and a:

PLANAR DIELECTRIC WAVEGUIDES (11)

PLANAR DIELECTRIC WAVEGUIDES (12)

• Dispersion relation for TM modes in terms of V , b and a:

At cutoﬀ (b = 0) for the lowest TM mode (ν = 0), the value of V is

SUMMARY OF PLANAR DIELECTRIC WAVEGUIDES

• Points to remember after you forget the mathematics:

OPTICAL FIBER WAVEGUIDES (1)

• For a TE mode, Êz = 0 and Ĥz = 0