Page 268 - Electrical Properties of Materials

250 Dielectric materials

THz source would lead to numerous applications in security, a consideration
not irrelevant today.
In order to have transfer of power from electrons to optical phonons, first
of all, we need electrons. In CdS, used in the acoustic amplifier, they were ob-
tained by optical excitation. That is certainly a possibility but there is no need
for it. There are a number of II–VI (e.g. InSb) and III–V (e.g. GaAs) materials
which display optical phonon resonances in the THz region (around 5.4 THz
for the former and 9 THz for the latter) and happen to be semiconductors with
the further advantage of high mobility. Will such oscillators come? There is a
chance.

10.15 Optical fibres
I have tried to show that dielectric properties have importance in optics as well
as at the more conventional electrical engineering frequencies. That there are
no sacred boundaries in the electromagnetic spectrum is shown very clearly by
a fairly recent development in communications engineering. This involves the
transmission (guiding) of electromagnetic waves. The principle of operation is
very simple. The optical power remains inside the fibre because the rays suffer
total internal reflection at the boundaries. This could be done at any frequency,
but dielectric waveguides have distinct advantages only in the region around
μm wavelengths. The particular configuration used is a fibre of rather small
diameter (say 5–50 μm) made of glass or silica. Whether this transmission line
is practical or not will clearly depend on the attenuation. Have we got the for-
mula for the attenuation of a dielectric waveguide? No, we have not performed
that specific calculation, but we do have a formula for the propagation coef-
ficient of a plane wave in a lossy medium, and that gives a sufficiently good
approximation.
Recall eqn (1.38),

2
k =(ω μ +iωμσ) 1/2 , (10.72)

and assuming this time that

ω σ, (10.73)

we get the attenuation coefficient

√ √

1 ω σ 1 ω
k imag = = tan δ. (10.74)
2 c ω 2 c
The usual measure is the attenuation in decibels for a length of one
For optical communications to be-
kilometre, which may be expressed from eqn (10.74) as follows:
come feasible, A should not ex-
–1
ceed 20 db km , first pointed out
by Kao and Hockham in 1966. A =20 log exp(1000 k imag ) = 8680 k imag
10
√
Kao received the Nobel Prize in ω –1
= 4340 tan δ db km . (10.75)
2009. c