Abstract:
The Raman effect is the inelastic scattering of photons by matter. When a
monochromatic light beam (pump beam) propagates in an optical fibre, spon-
taneous Raman scattering transfers some of the photons to new frequencies.
The probability of a photon scattering to a particular frequency shift, depends on that frequency shift, forming a characteristic spectrum. The scattered photons may lose energy (Stokes) or gain energy (anti-Stokes). If the pump beam is linearly polarized, then the polarization of scattered photons may be the same (parallel scattering) or orthogonal (perpendicular scattering). If photons are already present at other frequencies then the probability of scattering to those frequencies is enhanced (stimulated scattering).
In this thesis, the equations governing the Raman effect, both in bulk
glass and single mode fibre, are restated. The probability of spontaneous
scattering (parallel and perpendicular) in an optical fibre has been measured
for frequency shifts between 4THz and 58THz. Particular attention has been
given to those frequency shifts where the scattering is intrinsically weak. This
has identified those frequency shifts where it might be possible to eliminate
Raman noise from experiments designed to generate quantum states of light
in optical fibres using four wave mixing.
From the measured scattering probabilities, the Raman equations can be used to predict the scattering cross-section in bulk glass. The prediction is found to be in reasonable agreement with measurements of the bulk glass cross-section published by other authors.
By increasing the pump power, the transition to stimulated scattering has
been observed. In the case of Stokes scattering there is reasonable agreement
between the Raman equations and experimental observations. In the case of
anti-Stokes scattering, and in the limit of very high pump power, behaviour
is observed which is not predicted by the Raman equations, indicating that
other processes must be included to successfully describe such experiments.