New materials and configurations will allow photomultipliers to become more sensitive and rugged or smaller and less expensive. Technology and economics drive photodetector innovations. Technology determines what can be developed, while economics governs both what is developed and the rate of development. However, economics varies from market to market and over time. Semiconductor and optical communications are two markets that have been declining over the last year. Nonetheless, these markets are so large and so much driven by innovation that photodetector technology continues to advance. Until recently, the needs of the military and law enforcement did not influence detectors very much. With the events of Sept. 11, this has changed, generating a large increase in demand for the detection of explosives and nuclear material. The aging population continues to fuel advances in medical diagnostics, and detector technology has changed to suit the needs of both existing instruments and of those undergoing development. Authorization of reimbursement for positron emission tomography (PET) scans of patients who have a variety of cancers has doubled sales of PET scanners over the last two years. Because these scanners use hundreds of photomultiplier tubes, there is a strong motivation to find a way to reduce their cost. However, because of the high skill level required, workers in developed countries make these tubes. Therefore, the manufacturers must rely on technological innovation and not lower labor costs to improve their price point. A microPET scanner developed at the University of California in Davis uses 90 photomultiplier tubes. Conventional scanners use many more, encouraging photomultiplier tube suppliers to produce more inexpensive devices. Courtesy of Simon Cherry, University of California. Although the innovations are not public knowledge, they probably involve the use of electron trajectory simulations to find dynode configurations that are easier to assemble and cheaper to manufacture. At Hamamatsu, we are also developing a multielement flat panel photomultiplier that can be used in a high-throughput PET scanner. Drug discovery has also influenced the development of optical detection devices. Animal PET scanners provide an excellent method to noninvasively test a drug candidate on laboratory animals such as mice. A scanner optimized for such applications requires an array of detectors placed close together to provide resolution sufficient to locate the site of activity of a drug. Manufacturers are developing compact multielement photomultipliers for small-animal scanners, and arrays of avalanche photodiodes will compete hard for this segment. Because the total detection area of the PET scanner is small, the higher cost-per-unit area of an avalanche photodiode detector is not a big concern. The solid-state detector’s lack of sensitivity to magnetic fields makes it easier to combine animal PET scanning with magnetic resonance imaging (MRI). PET registers the location of the drug, while MRI measures the anatomy. For drug discovery in cells and medical diagnostics, manufacturers are trying to increase throughput with devices that simultaneously measure more components. This requires the use of more dyes and, in turn, requires extending the range of the dyes into the blue and red. It also has generated efforts to improve sensitivity at 300 and 800 nm. Look for new photocathodes based on GaN and extended red multialkali. Multi-element detectors will also increase throughput or the amount of information from a single measurement. Other biomedical applications such as gene chip scanners and flow cytometers have identical requirements for extended spectral coverage and multiple elements. These instruments will see accelerated sales as research proves their capability to detect and analyze the presence of biological and chemical agents. Telecommunications will continue to grow. Most people believe that detector sales will be the greatest for metropolitan networks and, eventually for applications in the local loop. Thus, emphasis will be on lower-cost, higher-volume devices. High-gain devices such as avalanche photodiodes with solid-state amplifiers will soon be available for this market. Initially, they will be hybrid devices, but monolithic detectors will follow to help lower costs. For short distances, it is more economical to transmit data at a low bit rate over multiple fibers because it incurs less cost for electronics. We can expect to see multielement detector arrays. In silicon, they will be in the form of optoelectronic integrated circuits in which processing electronics and the detector will occupy one chip. For InGaAs, a hybrid array will be available, but suppliers will soon be placing the detector array and the electronics on one device. Finally, as photolithography implements 157-nm F2 lasers, we will see photodiodes able to withstand long-term exposure to this wavelength without damage. One such approach is to use a phototube with a diamond photocathode. The next year will yield evolutionary changes in photodetectors, some driven by the need to reduce costs, others by performance requirements. Meet the author Kenneth J. Kaufmann is in marketing at Hamamatsu Corp. in Bridgewater, N.J.