The technical challenges with terahertz imaging don't stop with sources. Detectors, particularly the types of dense array detectors suitable for imaging, are still lacking. That is a need that Heinz-Wilhelm Hübers, head of the department of terahertz and IR sensors at the German aerospace center DLR (Berlin, Germany), aims to address. Hübers and colleagues are working to develop array-based superconducting microbolometers. "Our main application right now is security," says Hübers, though he adds that the technology will also benefit astronomical and atmospheric sensing applications.
DLR coordinates the project Active Terahertz Imaging for Security (Terasec) which is funded by the European Commission. A joint effort of 14 partners from industry, universities, and public research institutions, Terasec puts the focus on short-range and stand-off systems for terahertz imaging in security applications; in particular, its mandate is to develop hardware, software, ethical standards, and a roadmap for further R&D activities. Project members are working on a standoff detection system able to image objects up to 20 m with a heterodyne detection scheme.
"For this, we are developing superconducting mixers that are very fast bolometers," Hübers says. Made of niobium nitride, the bolometers are 1 to 2 µm wide and 0.2 µm long, with thickness on the order of nanometers (see figure 1). The bolometer is embedded in the center of a planar antenna, basically derived from classical radio technology. At frequencies above 1 THz, the fabrication of waveguides and horn antennas becomes increasingly difficult because of size constraints.
Fig. 1 Atomic-force microscope images show superconducting hot-electron bolometer integrated in a planar logarithmic-spiral antenna. The bolometer is located in the center of the two arms of the antenna. It is 2mm wide and 0.2 mm long (=distance between the ends of the two arms). Photo credit: DLR and University of Karlsruhe
To solve this problem, designers frequently use quasi-optical or hybrid antennas. These antennas consist of two parts: a lens and a planar feed antenna (see figure 2). The substrate of the feed antenna is mounted to the back side of the lens. The refractive index of both lens and substrate is the same or very close in order to minimize reflection losses; the planar antenna is located at the focus of the lens. The group designs the device, which is fabricated by another Terasec partner with a combination of electron-beam lithography and RF sputtering.
Fig. 2 The hot-electron bolometer (logarithmic-spiral antenna is the circular area in the center) is formed on a silicon chip which is glued to a hyper-hemispheric lens. Photo credit: DLR and University of Karlsruhe
"We have single-point mixers now," says Hübers. "We are also developing a 19-element array."
The development of an array detector is important. Single-point detectors require raster scanning, which increases system cost and complexity while reducing resolution. Most important, a raster scan approach significantly increases image acquisition time, a serious drawback when you're trying to image travelers in an airport or shopping mall, for example. An array would simplify the imaging process.
A key advantage to the microbolometer technology is the sensitivity. According to Hübers, the sensitivity of the superconducting microbolometers is, "very close to the quantum limit. It's very likely that we can image relatively fast," he says. "Not at video rates, but close."
The group is currently still in early design and development. "We have a simple passive imager that can image at 1 cm resolution and 7 m standoff distance," he says. "The goal is 1 cm resolution and 20 m, which is, at least at this stage, quite ambitious."
Then again, setting ambitious goals--and meeting them--is what Terasec is all about.
Kristin Lewotsky is a freelance technology writer based in Amherst, NH.