Increasingly in physics and astrophysics research the challenge lies in extracting very rare events from a tangle of unwanted but inescapable background processes.

In the Borexino experiment (currently under construction), one of the objectives is to measure the rate of neutrino events from the very low-energy beryllium-7 process, originating in the Sun. The energies of interest are lower than those of other neutrino experiments requiring a detection threshold of less than 1 MeV in a scintillator medium of some 300 t.

An important aspect of this experiment is the continuous purification of the scintillator, which involves the formidable industrial process of reverse osmosis on a grand scale. Needless to say, those parts of the experiment that cannot be continuously refined, such as the photomultiplier, need to be intrinsically pure.

In the dark-matter experiments of DAMA at Gran Sasso, Italy, and UKDMC at Boulby, UK, WIMP and other rare events from weakly interacting particles are being sought in scintillator-based detectors. Experiments using massive NaI(Tl) crystals, with elaborate anticoincidence, operating deep underground, use highly refined scintil-lator material and demand special photomultipliers.To match the advances achieved in reducing levels of radioactivity in the scintillators, manufacturers have had to scrutinize the materials used in the internal metal and ceramic parts of the photomultiplier, as well as in the glass envelope, where major sources of activity lie.

Fused silica (quartz) has very low radioactivity and is commercially available as an optional window material in photomultipliers. However, the highly desirable all-quartz photomultiplier is unattainable - essentially because of the mismatch in the expansion coefficients of quartz, and also the metal pins at the base of the photomultiplier making it impossible to manufacture.

Potassium is an important constituent of all commercial glass because its addition to the melt facilitates the working of the glass and ensures the optical quality. However, it contains a small proportion of the long-lived radioactive isotope potassium-40, making its presence in glass unacceptable from an experimental physics perspective.

The long-lived decay products of thorium and uranium are ubiquitous, particularly in sand, a major constituent of all glass. Radioactive decays from these sources produce a spectrum of gammas, ranging from 100 to 2700 keV, which encroach on the low-energy region of interest to low background experiments.

The practicalities of manufacturing low-background photomultipliers ultimately rely on the use of glass. Electron Tubes Ltd has located a source of low-activity sand and has worked with specialist glass companies to devise suitable manufacturing processes.

The glass mixture has to be maintained at extremely high temperatures, making it very corrosive. The design and material of the melting pot are critical to avoid contributing radioactive isotopes to the glass mixture or cracking within the first few days of operation. Glassblowing lasts only a few hours of each day. The pot is filled with material, and only when the melt is judged to be of acceptable quality and optimal working temperature does blowing begin.

Glassblowers work rapidly, producing a bulb every 1 or 2 min. When the level of the melt drops below some desirable level, or it ceases to work properly, the process is abandoned until the next day, when the cycle is repeated. A good day may produce 20 envelopes, a bad day nothing and the mechanical properties of the pot are continually degrading under continuous heat cycling. Replacing the pot and commissioning the furnace takes up to one month with a complete break in output.

The progress that has been made in the attainable radiopurity levels of low-background glass is shown below. Levels of potassium, thorium and uranium are measured in parts per million or billion.

Standard glass contains <60>Low background glass contains 300 ppm K, 250 ppb Th and 100 ppb U
Ultra-low background glass contains 60 ppm K, 30 ppb Th and 30 ppb U
Fused silica (quartz) contains <10>
In conclusion, current experiments at the forefront of science rely on a technology that has hardly changed over the centuries. Glassblowing demands the skills of a dwindling group of craftsmen, working by feel and by eye, unsung and largely unappreciated. How different the world of physics - or is it?

Photon detection for biomedical applications

Instrumentation for clinical analysis increasingly depends on bioluminescence, chemiluminescence and fluorescence. The amounts of light are small and often require photon counting, so highly sensitive, low-noise detectors are needed. The most flexible and cost-effective detector for this purpose is a photomultiplier tube (PMT), which in most cases is the only option.

Electron Tubes has developed a range of integrated light-detection assemblies specifically aimed at the needs of the biomedical instrument designer. These include a PMT and all associated electronics and signal processing, so the customer needs to provide only a low-voltage input and interface to a digital (normally TTL) or analogue current or voltage output.

Compact Peltier coolers are available that maintain the PMT at a constant temperature up to 40 °C below ambient.

The mechanical design can be customized to simplify final assembly of the instrument by the customer.

More information is available at