## Box 121 Dielectric waveguides

How to guide the wave. The phenomenon of light refraction has probably been known since the start of human civilization. Put a stick in water and it appears as if it is broken. It is a consequence of the fact that when light moves from a medium of lower refractive index (like air) into a medium of higher refractive index (like water) the rays will change direction at the boundary as shown in Fig. Bi2.l(a). When light is incident

Fig. B12.1 (a) A ray of light refracts when moving from a medium of lower refractive index into a medium of higher refractive index. (b) If the angle of incidence is small relative to the boundary, then there is no transmitted ray in the medium of higher refractive index and the ray is completely reflected. (c) Light may travel in a glass rod by subsequent reflections at the boundaries.

from a higher refractive index material it is refracted towards the boundary. As the angle of incidence increases, there is an angle at which the refracted ray is directed just along the boundary. This is when the refraction turns into reflection, a phenomenon known as total internal reflection (see Fig. Bi2.i(b)). It follows then that if light is incident at the surface of a glass rod (which has an index of refraction higher than that of the surrounding air) at a shallow angle, as shown in Fig. Bi2.i(c), it will be trapped inside by subsequent reflections. Such guidance of light waves was probably known to the ancient Egyptians who used glass tubes for ornamental purposes.

If the aim is to guide the light along the dielectric waveguide then it is strictly necessary for the index of refraction inside to exceed that outside. Since the guide cannot levitate in air and is in need of some support, this condition will not apply at the points of support where the electromagnetic energy may leak out. The way to overcome the problem is to clad the dielectric waveguide in a suitable material. It turns out to be advantageous to choose a cladding material whose index of refraction is just slightly smaller than that inside. Such an arrangement would allow single mode operation with a reasonably sized, not too small, inner core. Single mode operation is preferable because, as mentioned for hollow metal waveguides, it can avoid distortion of the signal.

History. The first demonstration of light trapping was done more or less simultaneously by Daniel Colladon in Geneva and Jacques Babinet in Paris. In Colladon's experiment, conducted in 1841, light was made to move along water jets and break free when the jets themselves broke, providing some spectacular pictures to the great enjoyment of the public.

One of the ways of producing a dielectric waveguide is to draw molten glass fast enough so that it turns into a thin glass fibre. This was done by Charles Vernon Boys towards the end of the last century at the Royal College of Science in South Kensington. His method was original, to say the least. He stuck an arrow to one end of a glass rod, which he heated until it was soft enough, and then shot the arrow to a distance of 90 ft. He was rewarded by a fibre 90 ft long and having a diameter of one-tenthousandth of an inch.

A more modern embodiment of a dielectric waveguide came in the 1950s with the endoscope, which consists of a bundle of glass fibres no longer produced with the aid of bows and arrows.

0 0