Silicon is great for many things. Light isn't one of them. Silicon can't emit light and does a poor job of absorbing lt. By adding layers of GeSi, we overcame the absorption problem. Alone, GeSi layers don't absorb ligth much better than pure Si. However, GeSi does have a different, slightly higher, index of refraction than Si. That means that light will tend to get trapped in such layers. The familiar analogy is the way light is reflected off the surface of a swimming pool - when viewed from below! To trap the light in our GeSi layers, we illuminated them from the side with a glass fiber (which uses the same principle to trap light). Then, because the light encountered the GeSi to Si layer interfaces at a shallow angle, it became trapped, ricocheting down the length of the layer. This long "optical path" made it highly likely that light would be converted to electrons. These could then be collected by the P-N diode that had been built around the GeSi layers. The net result was one of the first both fast and efficient Si-based photodetectors. A later, more sophisticated device is shown in the next image.