Since the advent of practical silicon photo-detectors more than 20 years ago, many people in the industry have been predicting with increasing regularity the imminent demise of the photomultiplier. Seen as being based on outdated technology, difficult to use, and requiring high voltages to operate, the days of the photomultiplier (PMT) are numbered. Or are they?
It is certainly true that the PMT is based on old technology. Being a vacuum device, it is seen as a close relation to vacuum triodes and pentodes used in old radios and sound reproduction systems. However, what is often overlooked is the fact that the principle of operation and performance capability of the photomultiplier was way ahead of its time when first commercialised in the 1950s.
One of the first commercial applications of photomultipliers was for detecting the pulses of light emitted by scintillators when exposed to nuclear radiation, which is the principle still used today for monitoring environmental radiation in association with nuclear power stations. To meet the exacting safety requirements, it is necessary to be able to detect extremely low levels of radiation – whether it be in the atmosphere or present on clothing, and the PMT/scintillator combination can do this easily.
One of the distinguishing features of a PMT, which is not applicable to silicon devices, is its relatively large active input area – typically, the light-sensitive surface is 30-50mm diameter, but it can be up to 200mm diameter or even bigger. This is due to the fundamental quantum nature of how the PMT works by converting detected photons to electrons, which are then multiplied within the device itself before the signal is extracted. This multiplication process is relatively noise-free and can produce measurable signals from light sources as low as a few photons.
Such a low detection threshold is still not possible with silicon devices that have similarly large sensitive areas. However, it is true to say that certain types of small-area silicon detectors (few mm diameter or less) are capable of detecting down to single photon levels, but this is more likely to be the case when the light input is at the IR end of the spectrum where photomultipliers are much less sensitive and silicon devices more sensitive. So there is certainly a degree of substitution taking place in the near IR, and where the light is focused to a small area.
Another criticism of PMTs is that they require a high voltage to operate (typically 700 to 1,000V), which makes them relatively userunfriendly. However, PMTs are now available as integrated modules in which the HV is internally generated, and the only external voltage required is typically 5V DC. These modules are effectively ‘plug and play’.
There is often a reference to PMTs not being mechanically robust, and this stems from the fact that glass is used in their construction. This is actually quite easily overcome by using suitable mounting arrangements and, again, the integrated module approach makes life much easier for the PMT user.
One of the reasons that PMTs have survived, and arguably, prospered in the face of new technology is that the disadvantages, perceived or real, have been addressed by the PMT manufacturers over the years. When first introduced all those years ago, PMTs were made in an almost laboratory environment with all of the associated uncertainties. But in the past 20 years, great strides have been made in not only increasing their performance in all of the critical parameters, but more modern manufacturing techniques have brought a much higher level of consistency. This has helped to keep the market for PMTs very active – and, with it, a healthy level of competition to keep prices down. It would appear that it is a real case of product improvement stimulating continuing demand.
As with other key system elements, it is important to select the correct product for the application, and this has been made much easier by the availability of online parameterised search facilities as well as real-life application and technical support.
The largest industrial market for PMTs is medical systems such as PET scanners, but there is also a significant potential market in astrophysics, mostly in relation to research into the origins of the universe. There is at least one project that will require the equivalent of a cubic km of detectors. This would be some challenge for detectors that are less than 5mm diameter. Even with large-area detectors, such as PMTs, it is still a challenge that will hopefully keep manufacturers busy for many years to come.
