The history of the development of the semiconductor is synonymous with the history of creating transistors that are smaller, faster and that consume less electricity. From the left, Planar Transistor, Fully Depleted (Fin) Transistor, and GAA Transistor
The most widely used transistor in the current semiconductor industry is the Metal-Oxide-Semiconductor (MOS). It consists of a metal electrode, an oxide insulator and a semiconductor channel. The first MOS transistor was of a planar architecture and was structured so that the gate and the channel made contact on one plane. But, as transistors become smaller, the distance between the source and the drain gets smaller, making it difficult for the gates to work as a switch. This is called a ‘short-channel effect’, and along with limiting voltage reduction, it means that planar transistors can only be applied to 20 or above nanometer nodes (or generations)1
In order to overcome the short-channel effect, the Fully Depleted transistor emerged as the next generation of transistor. This transistor uses a thin silicon (Si) channel to avoid the short-channel effect by enhancing the ability of the gate to adjust the channel. Its structure format evolved out of that of the conventional transistor (a gate on a plane channel) to become a thin, rugged structure with a standing rectangular channel that interlocks with gates on three sides. As this thin, standing channel somewhat resembles a fish’s dorsal fin, it is also called the ‘fin transistor’. Samsung has been manufacturing fin transistors since 2012 in a range of sizes, starting at just 14 nanometers.
Whereas a planar transistor only allows the channel and the gate to contact in just one plane, a fin transistor has a 3-dimensional structure that allows three sides of a channel (excluding its bottom) to come into contact with the gates. This increased contact with the gates improves semiconductor performance as well as increasing the reduction of operating voltage, solving the problems brought about by the short-channel effect.
Nevertheless, the fin transistor is now facing limitations after several generations of developments and process transitions. Nowadays, the semiconductor industry is increasingly requiring transistors that can reduce operating voltage even further. Despite the fin transistor’s 3-dimensional structure, that only three of the four sides are in contact with gates is now becoming a limitation, as transistors themselves continue to progress and subsequently get smaller.