ABSTRACT
Fiber optics is well established today due to the large capacity and reliability it provides, but it has a clear disadvantage in cost. Furthermore, components are typically large and expensive due to bulky ˜ber elements. ™is, linked to the time consuming assembly and packaging required for precision alignment of micro-optics with dimensions of the order of micron, has led to increasing interest in the possibility of building devices integrated on a single chip of transparent material and, consequently, bene˜tting from the much smaller footprint in integrated photonics. Ever since the earliest research on optical circuits, dating back to the 1970s, there have been visions of an optical superchip [1,2], containing a variety of integrated optical components to carry out light manipulation (generation, modulation, detection, switching, ˜ltering, and ampli˜cation). In the early days of integrated photonics the work was associated with ferroelectric materials such as lithium niobate, and III-V semiconductors such as the gallium arsenide and indium phosphide based systems. ™is was mainly due to a large electro-optic eŸect enabling high modulation speed in the case of lithium niobate, and the possibility to integrate a source and other critical components in the case of III-V semiconductors.
