In principle, the light to be detected strikes the surface of a photocathode, which emits individual electrons by photoemission. Since the resulting current is so extremely low, it is increased by secondary electron multipliers. In this process, the electrons knocked out of the photocathode are accelerated by an electrostatic field to the first of several electrodes. The multiplier electrodes, called dynodes, generate multiple secondary electrons upon electron impact, which in turn are accelerated to the next dynode. The secondary electron emission is further amplified via the downstream dynodes and tapped at the last dynode as a current flow. The principle of secondary electron multiplication is such that downstream dynodes always have a higher accelerating voltage than the previous one.
If each primary electron generates three secondary electrons when hitting a dynode, then the secondary current at 10 dynodes is about 60,000 times higher (`3^10`) than the primary current generated by the photocathode.
Photomultipliers are used in highly sensitive light sensors, in optical spectroscopy, in night vision devices and in drum scanners.