What is a semiconductor?

(upbeat music) (air whooshing) - In this video, we'll graze by the fundamental physics of semiconductors enough to give you a taste of what's involved, but at a high enough level to keep an eye on the big picture of how that physics translates to potential advances and innovations, to the challenges in scaling microchip performance and manufacturing, and to how different materials and architectures impact capacity, resilience, our competitive edge, our ability to infiltrate new markets, and our impact on the environment. It all starts with the atom, the basic building block of all matter, whether solid, liquid, or gas. Atoms are composed of a central nucleus, which is positively charged, and an orbiting cloud of negatively charged electrons surrounding that nucleus. Different elements have different number of electrons, and similarly different charged nuclei, which defines their chemical, electrical, and mechanical properties. The electrons fill up spots in their orbit around the nucleus, and depending on the number of electrons, there are some which may be very weakly bound to the atom. So when atoms are brought together to form a material, electrons might be shared between neighboring atoms forming bonds which hold the material together. But those weekly bound electrons with just a bit of heat or light may start freely floating through the material. Those electrons are free to conduct electricity. In conductors, there are enough of these free electrons to create this ocean of charge. In the case of insulators, the electrons are tightly bound to their nucleus and live happily in their location forming covalent bonds with neighboring atoms. A lot of energy, heat, or light, for example, would be required to knock them out of their location, so no electrical current flows. (bright music) Semiconductors are somewhere in between conductors and insulators. Elements which are used to form semiconductors have just the right number of electrons to form covalent bonds with their neighbors without any extra electrons left over. You may have heard of the group four elements in the periodic table, including silicon, carbon, and germanium, and others, which have this property. So a pure lattice of silicon is not a good conductor because it doesn't have free electrons. Same for silicon carbide or silicon germanium, since all these elements have the same number of outer electrons, four. But we can actually engineer how conductive the silicon is by introducing atoms of other elements into the lattice. This is known as doping. Depending on the number of electrons in these impurities embedded into the lattice, they may either donate an extra electron to the material, so these are called donors, or create a welcoming spot for any rogue free electron to fill, effectively stealing it away from the ocean of conducting electrons. These are called acceptors. Common donors for silicon are phosphorus and arsenic from group five with an extra electron. While boron from group three is most commonly used as the acceptor dopant for silicon. You don't need a lot of these impurities to wildly change the conductivity of the semiconductor, maybe one in a million. We can set the density of these impurities, how many of the silicon atoms are replaced with the donor or acceptor element per cubic centimeter, to control the conductivity by orders of magnitude. This feature is perhaps one of the most important that has allowed semiconductor-based microelectronics as we know it to succeed on such an epic scale, quickly, overthrowing vacuum tubes and mechanical switches for computing and signal processing. (upbeat music) (air whooshing)