Basic Semiconductor Crystal Structure

The crystal structure of pure silicon is of course 3-dimensional, but that is difficult to display or to see, so the image to the left is often used to represent the crystal structure of silicon. For you physics types, silicon (and germanium) falls in column IVa of the Periodic Table. This is the carbon family of elements. The essential characteristic of these elements is that each atom has four electrons to share with adjacent atoms in forming bonds.
While this is an oversimplified description, the nature of a bond between two silicon atoms is such that each atom provides one electron to share with the other. The two electrons thus shared are in fact shared equally between the two atoms. This type of sharing is known as a covalent bond. Such a bond is very stable, and holds the two atoms together very tightly, so that it requires a lot of energy to break this bond.
For those who are interested, the actual bonds in a 3-dimensional silicon crystal are arranged at equal angles from each other. If you visualize a tetrahedron (a pyramid with three points on the ground and a fourth point sticking straight up) with the atom centered inside, the four bonds will be directed towards the points of the tetrahedron.
Now we have our silicon crystal, but we still don't have a semiconductor. In the crystal we saw above, all of the outer electrons of all silicon atoms are used to make covalent bonds with other atoms. There are no electrons available to move from place to place as an electrical current. Thus, a pure silicon crystal is quite a good insulator. In fact, it is almost glass, which is silicon dioxide. A crystal of pure silicon is said to be an intrinsic crystal.
To allow our silicon crystal to conduct electricity, we must find a way to allow some electrons to move from place to place within the crystal, in spite of the covalent bonds between atoms. One way to accomplish this is to introduce an impurity such as Arsenic or Phosphorus into the crystal structure, as shown to the right. These elements are from column Va of the Periodic Table, and have five outer electrons to share with other atoms. In this application, four of these five electrons bond with adjacent silicon atoms as before, but the fifth electron cannot form a bond. This electron can easily be moved with only a small applied electrical voltage. Because the resulting crystal has an excess of current-carrying electrons, each with a negative charge, it is known as "N-type" silicon.
This construction does not conduct electricity as easily and readily as, say, copper or silver. It does exhibit some resistance to the flow of electricity. It cannot properly be called a conductor, but at the same time it is no longer an insulator. Therefore, it is known as a semiconductor.
While this effect is interesting, it still isn't particularly useful by itself. A plain carbon resistor is easier and cheaper to manufacture than a silicon semiconductor one. We still don't have any way to actually control an electrical current.
Experimentation shows that there is an empty place where an electron should logically go, and often an electron will try to move into that space to fill it. However, the electron filling the hole had to leave a covalent bond behind to fill this empty space, and therefore leaves another hole behind as it moves. Yet another electron may move into that hole, leaving another hole behind, and so forth. In this manner, holes appear to move as positive charges through the crystal. Therefore, this type of semiconductor material is designated "P-type" silicon.