The research market is not usually profitable, but it is still very dear to detector manufacturers. While it can cost a company as much to serve a research customer placing one $500 order as it does an original equipment manufacturer who annually orders detectors costing in excess of $100,000, the information learned from research projects often proves quite valuable in terms of new product development.
Most of the time our newest detectors will first be used in a laboratory. Researchers in both academic and industrial settings are always looking for an edge, such as higher speed, more sensitivity, higher dynamic range, or access to a new portion of the spectrum. They are also more willing to take a risk than someone developing a new commercial instrument. If the detector has just been developed, they will often test it in ways that we might not think of.
One such example is a photon-counting module we introduced about a year ago. The module is based on gallium arsenide phosphide, and has a quantum efficiency in excess of 40%. Initially our engineers designed it with a photocathode that was 5 mm in diameter since many of our customers wanted to have as large an area as possible. The 5-mm photocathode generated several hundred dark counts at room temperature and the use of a Peltier cooler seemed to solve this problem. When we started talking with some research customers doing confocal microscopy, however, we found that the vibrations from the cooling fan would interfere with the measurements. So our engineers developed a photomultiplier tube (PMT) with a 2-mm diameter photocathode. It has a much lower dark current without requiring a cooling fan.
Another type of research need extended the market for a PMT with sensitivity to 1700 nm, which we introduced several years ago in Japan. As it turned out, several researchers in the United States needed either the sensitivity or time resolution of this new detector to demonstrate that the signal they wanted to detect was present, so that they could get funding for a near-infrared detection system, which is quite expensive. The majority of their research involved semiconductor materials studies such as photoluminescence from quantum dots, nanoparticles, and optical fiber amplifiers, or chemical reaction dynamics studies using fluorescence or singlet oxygen luminescence.
Occasionally, researchers will come to us with a request for a new detector or a complete detection system, which can give us important marketing input and help us decide which development projects to pursue. Our president, Teruo Hiruma, is especially interested in customers' requirements for detectors or light sources that do not exist. His interests are not in extensions of existing technologies, but in radically new devices that might be as much as 10 years away from commercialization.
We developed detectors for the Super Kamiokande neutrino detector, for example, which required a new half-meter diameter photomultiplier. This tube had to operate under water, so the packaging for the high-voltage connection and dynode resistors had to be water tight down to a few atmospheres of pressure. The experiment required about 11,000 tubes (see Laser Focus World, October 2000, p. 77). So, in addition to the tube development we constructed a new building, and designed new machines for semiautomatic device construction.
Some experiments such as LIDAR generate a lot of background during a laser pulse. Based on feedback from our customers, we developed a gated version of the microchannel plate PMT to eliminate the laser background.
We also have a long history, starting with our streak cameras in the mid 1970s, of working with scientists doing picosecond and femtosecond spectroscopy. In the United States and in Europe, many projects with limited budgets were looking for lower cost alternatives. For instance, using the microchannel plate photomultiplier in the time-correlated photon-counting mode measures the time of detection of a single pulse with respect to an optical or electrical stimulation. The main factor that determines the time resolution is the temporal time spread (TTS) of the photomultiplier because the electrons can take many different trajectories, and each trajectory has its own transit time. By working with several universities we have, over the years, reduced the TTS to as short as 20 ps. To accomplish this we had to reduce the pore size of the microchannel plate as well as optimize the electronics. We concentrated on improvements in the tube while our academic collaborators worked on tweaking the electronics.
Researchers provide us with great ideas for new devices, are often the first to try a detector, and give us insight into the future. So we encourage people reading this article to e-mail the author with their ideas. Of course, we do not want any confidential information. In fact, if there is one barrier to developing new detectors it is the increasing requirement, even at a university, for nondisclosure agreements.