Innovative photomultiplier-tube design lowers dark current
Andrew Wilson
In applications such as emission spectroscopy, beta and gamma spectroscopy, and bio- and chemiluminescence, the counting of low levels of electrons is mandatory. Despite recent advances in solid-state detectors, this task is predominantly performed with photomultiplier tubes (PMTs).
"Since 1965 the design of photomultiplier tubes has changed little," says Don Lake, business-development manager at EG&G Imaging (Santa Clara, CA). "During operation, photons arriving at the photocathode of a conventional photomultiplier tube generate electrons that are focused through an electrode onto a series of dynodes. Because the potential at each dynode decreases as the electrons flow toward the photomultiplier`s anode, each electron that strikes a dynode produces about four secondary electrons," he explains.
Depending on the number of dynodes used in the PMT, a large gain can be obtained using such devices. "Although relatively expensive to make," says Lake, "standardizing on conventional designs has allowed manufacturers such as Hamamatsu Corp. (Bridgewater, NJ) to build these devices in the Far East at relatively low cost." For suppliers of spectroscopy systems, however, using such devices requires the design of power supplies that generate multiple high voltages, shock-resistant assemblies, and noise-compensation circuits.
To overcome these limitations and to decrease dark current to the low pico- ampere range, Ruimond Bardon, division manager at EG&G Heimann GmbH (Wiesbaden, Germany), and his colleagues have developed a patented monolithic-photomultiplier technology for a line of channel-photomultiplier devices designed to replace most of the PMTs available today. Marketed by EG&G Imaging, the new ultrahigh-sensitive optosensor differs radically in construction and postprocessing electronics from conventional PMT designs (see figure). Primarily, it eliminates the hundreds of elements required to construct a regular PMT, thereby decreasing noise and increasing dynamic range.
Unlike existing PMTs that use many independent dynodes to amplify generated electrons, the optosensor uses a single, wavy, elongated, continuously resistive glass, called a channel photomultiplier, which is housed in a 75 x 10.5-mm cylindrical package containing a proprietary encapsulation material. The four input/output wires require no "vacuum-tube socket" for installation. The detector is a small, head-on type. A choice of window materials allows coverage of the spectrum from 115 nm (ultraviolet) to 850 nm (near-infrared) in six ranges. In this design approach, only the entrance window differs in material to accommodate the different spectral ranges. The overall size of the optosensor and the contents of its postprocessing electronics remain the same for all ranges.
Rather than applying a range of different voltages to independent dynodes, as done for conventional units, a single high-voltage potential is set up between the top detector and the bottom anode. Because the glass is composed of special rare elements, the length of the wavy glass from the channel electron multiplier entrance to end emulates the effects of the independent dynodes in a conventional PMT.
"However," says Lake, "because discrete biasing is not used, the resulting effect is a self-aligned, self-biasing PMT." When compared with cesium iodide (CeI) PMTs from Hamamatsu, for example, the performance of the ultrahigh-sensitive optosensor with an applied 2-kV potential exhibits greater current amplification (5 x 106) and three orders of magnitude less anode dark current (see table). When a 3-kV potential is applied, the device exhibits another order of magnitude of current amplification with relatively the same dark-current level," claims Lake.
Whereas many vendors have abandoned PMT development in favor of developing solid-state sensors, EG&G has used its expertise to re-invent the PMT by developing a more efficient device with a greater sensitivity and the same price as conventional PMT designs.
