Terahertz Waves for Production-Line Inspection?

http://www.eetimes.com/author.asp?section_id=36&doc_id=1330007

Bill Schweber, Engineer, Author, Editor

Using various slices of the RF spectrum for sensing rather than communications has fascinating potential and some impressive implementations, but there are still many significant challenges, especially in the terahertz (sub-mm) band.

The electromagnetic spectrum between 0.3 THz and 3 THz (1 THz = 10¹² Hz), which falls between microwaves and infrared radiation and commonly called terahertz or submillimeter waves, has long had a reputation of great potential for sensing situations. But, progress at these frequencies has had little real-world progress. (For a more-familiar frame of reference, note that 0.3 THz is 300 GHz.) To understand why this is so, see the excellent IEEE Spectrum article The Truth about Terahertz. The article explains that it's not just due to a lack of good sources and sensors in this region. Indeed, there are also some fundamental realities of physics that explain why these waves have so much promise and yet induce so much exasperation and frustration.

Still, physicists and engineers keep trying. After all, that's what progress is about, and there has been significant THz progress. Remember that what we now call the "short-wave" band (roughly 5 MHz to 20 MHz), and the high-frequency HF and very-high-frequency VHF bands (the 3-to-30 MHz and 30-to-300 MHz ranges, respectively), were once considered out-of-reach and unusable. That "short-wave" designation is a historical artifact rather than a modern reality, as those frequencies are now easily and casually used parts of the RF spectrum.

Still, I was surprised to see that some of the terahertz progress is now in production-line applications rather than high-end R&D test and measurement. A recent short article in Laser Focus World led me to a paper from a small Pennsylvania company, Applied Research & Photonics, which discussed the challenges of inspecting the aluminum-foil seals used for pill containers (Figure 1). The paper also discussed their experimental solution using THz-based reflectometry Figure 2, (see here). Their system uses a 10 mW dendrimer dipole excitation (DDE) THz source and yields a non-destructive, non-contact system which detects tiny changes in the reflection intensity, which is proportional to distance, for resolution sensitivity on the order of 25 nm.

Figure 1. The aluminum foil used for sealing jars shows tiny imperfections around the sealed circumference; seeing these on a production line is a major challenge. Source: Applied Research & Photonics.

Figure 2. The THz-based arrangement for inspecting the container seals uses the optical components and configuration of reflectometry, including a reference source, beam splitter, and comparison. Source: Applied Research & Photonics.

Bottle-seal inspection is a problem I never thought much about, frankly, but when you think about it, it certainly is important. It's a surprisingly difficult problem, given the type of defects and the dimensions involved. (I strongly urge you look at Laser Focus World and Photonics Spectra, either online or print, to keep up with optical and electro-optics, which are so closely tied in with both consumer and specialized electronic components, systems, and products.)

After I read the article and realized how little I know about terahertz-wave applications, I did some online searches to see what else was going out in this region. For example, one company I came across, U.K.-based Teraview (a spinout of Toshiba Research Europe), showed many fascinating applications with which they are involved. Obviously, from a web site alone, I can't determine how real or successful they are. Still, there's a lot worth thinking about in what they showed, and in how difficult-to-handle technologies can still be employed for diverse, apparently unrelated yet interesting and surprisingly difficult test issues in the real world.

In TV shows and movies, we often see forensics teams that casually use high-end equipment, easily to making all sorts of firm determinations. Some of what we see them doing is reality, some, between the product development lab and reality, and some is not really here yet. But the blurry boundaries between those zones keeps shifting, and what was science fiction one day is practical another day, as we well know.

Still, sometimes the largest test and measurement progress is where we normally don't see it. Those less-glamorous but high-speed, high-volume, 24/7 production lines can also be the drivers, where installation has to be solid and straightforward, up-time has to be very high, and consistency of results has to be a given. In these situations, the installed system must make binary pass/fail decisions quickly with near-zero false positives and negatives.

Have you been involved with innovating and implementing advanced sensing technologies and techniques, especially for production lines? What surprises did you encounter? Were there cases where you had to finally give up, because the solution was not working out or was impractical?