How do npn and pnp transistors operate?

A bipolar transistor (bipolar junction transistor: BJT) consists of the collector, base, and emitter regions, with the very thin base region located between the collector and emitter regions. The base region has two pn junctions between collector and emitter.  To obtain a high current gain, the emitter region is orders of magnitude more heavily doped than the base region.
A BJT turns on the PN junction between the base and the emitter by applying a forward voltage VBE between the emitter and the base. Carriers flowing from the emitter to the base (emitter current) are controlled by raising and lowering the energy barrier between the BEs with the voltage between the BEs determined by the base current. 
The impurity concentration of the emitter is high, and a large amount of carriers flow in. Furthermore, since the base width is thin, it flows into the collector with almost no recombination. Therefore, the BJT operates as an element that amplifies the base current.　

Let’s consider the npn transistor where the collector potential is higher than the emitter potential and the base potential is roughly 0.7 V higher than the emitter potential. In other words, the base-emitter junction is forward-biased whereas the base-collector junction is reverse-biased.

When the base-emitter junction is forward-biased, a small current flow into the base, injecting holes into the p-doped base region. This lowers the energy barrier between the base and emitter, allowing some free electrons in the emitter to diffuse to the base. Due to the difference in impurity concentration between the emitter and the base, the number of holes in the base is small compared to the diffused electrons. Also, the base is very thin. Since the number of electrons is greater than that of holes and the time that they exist in the base is short, most of the diffused electrons reach the base-collector junction and flow into the collector without recombination. This is the collector current.

The DC current gain h_FE is mainly determined by the difference in impurity concentration between the base and the emitter and the thickness of the base width. BJT is a device whose collector current can be controlled by base current.

Next, let’s consider the pnp transistor. Suppose that the collector potential is lower than the emitter potential and that the base potential is roughly 0.7 V lower than the emitter potential. In the case of the pnp transistor, electrons are injected into the n-doped base region. This lowers the energy barrier and allows some of the holes to diffuse out of the emitter. Some of these holes recombine with the electrons injected into the base region. The remaining holes diffuse across the base region, reaching the collector.

The following documents also contain related information: