Photomultipliers (sometimes called photon multipliers) are vacuum tubes, where light absorbed on a photocathode generates free electrons, which are subsequently accelerated with a high voltage (at least hundreds of volts), generate secondary electrons on other electrodes, and finally a usable photocurrent. Due to this avalanche process, the photocurrent can be orders of magnitude higher than from, e.g., a photodiode. Therefore, photomultipliers can be used for, e.g., single single photon counting. Photomultiplier tubes can be highly sensitive detectors with high bandwidth (>1 GHz), and good linearity in a wide dynamic range.
Note that there are also ordinary phototubes with only two electrodes and therefore much lower responsivity. They can be operated with a much lower voltage of e.g. 15 V. Some of these tubes, however, are gas-filled, exploiting a moderate amount of photocurrent amplification via ionization processes in the gas.
Disadvantages of photomultipliers are the large size (compared with e.g. photodiodes), the higher cost, the requirement for a high-voltage supply (often providing well above 1 kV), and in some cases the typically lower quantum efficiency (normally <25%, sometimes even <1%).
A very compact form of photomultiplier tubes are microchannel photomultipliers (MCPs = microchannel plates) based on a monolithic tube in doped glass. The small size allows one to assemble many channels to a 2D array and to achieve a very high detection bandwidth.
In some cases, photomultipliers can be replaced with avalanche photodiodes, which also exhibit an amplification mechanism, but in that case one which occurs within a solid-state (semiconductor) material, rather than in a vacuum tube. A large active area can be obtained with an array containing may avalanche diodes; such devices are sometimes called silicon photomultipliers. Compared with photomultiplier tubes, silicon photomultipliers can be cheaper and much more compact and robust. They typically exhibit a higher quantum efficiency, but also a higher amplification noise.
There are also hybrid photomultipliers, where electrons from a photocathode are accelerated with several kilovolts to a semiconductor chip similar to that of an avalanche diode. One first obtains an electron-bombardment gain from the fast electrons hitting the semiconductor, and subsequently an avalanche gain. Such devices are similar fast as conventional photomultipliers while also offering an improved pulse height resolution. Also, they have a small latency and are rather compact. They can be used in similar areas a photomultipliers, including single photon counting.
A larger active area of an avalanche photodiode device can be obtained by arranging multiple (even thousands) of pixels on one chip in close proximity. Such devices are sometimes called solid-state photomultipliers or silicon photomultipliers (when the diodes are based on silicon).
If you like this article, share it with your friends and colleagues, e.g. via social media: