Wanted: recycling inventions

Remember Popular Mechanics magazine? Way back in the Stone Age (before the Internet) that publication used to have a department called “Inventions Wanted”. It featured cartoon depictions of great ideas that you couldn’t buy, like self-cleaning paintbrushes and glow-in-the-dark lug nuts. This bit of trivia comes to mind because recently I’ve been inundated with press releases regarding the recyclability of plastics. Automotive OEMs are especially keen to be recognized as “green” with part coding and even vehicle “disassembly” lines. Ford has even purchased Ontario and Quebec auto “recyclers” to streamline the process.

All this “product stewardship” stuff is fine, except that it requires someone in the supply chain to separate the waste streams into the various classes of thermoplastics, and that’s the problem, as I see it anyway. Blue box is fine, but there’s just no way to split the PET stream from the PS from the HDPE, etc., etc.

And that’s the point of this month’s column.

Turning recycled thermoplastics into a serious, reliable raw material is going to require some way to separate the waste materials without human intervention. And to do that, we’re going to need a fast, cheap method of polymer characterization that doesn’t need a chemist to operate. Something portable, maybe the size of a paperback book, that characterizes the materials in a thermoplastic part non-destructively. It would also have to be cheap, both to own and operate. I’m guessing around $1000 to start, for a simple “polymer identifier”, with higher priced models for copolymers and alloys. A device like that could also be used at the molding plant’s receiving dock, allowing high production molders to make a quick accept/reject judgement before the gaylord of repro is even off the truck.

The ability of recyclers to test their grind, and molders to test their incoming recycled material is crucial if the industry is going to leave the high capital, government-supported initiatives behind and become truly entrepreneurial. For “QA person in a box” to become a reality, at least at reasonable cost, is going to take some engineering. That’s because current techniques require fairly invasive methods to crack the chemistry of polymeric materials.

Spectroscopy might work, where the plastic is bombarded with some wavelength or wavelengths of radiation, and the frequencies absorbed and reflected leave a sort of fingerprint unique to each chemical compound. That’s great, except that even with current levels of automation, the equipment is big and expensive, and takes skilled operating personnel.

Nuclear magnetic resonance, as used by the medical community, is another possibility. It’s very sensitive, but again expensive.

Mass spectroscopy is my favorite, because it works like a mechanical device: fling the stuff into a curved magnetic field, and the different elements spin off to the outside of the curve, like cars off an icy curve. Position detectors at various points along the curve, and you collect the different elements at different points as they bend more or less in the magnetic field. Drawbacks? You need to accelerate the stuff in a vacuum, and as you suspected, it’s expensive.

I believe the technology to identify recycled materials may already exist in some proprietary databank, but when it does come, I’ll bet it will use solid-state lasers in some form of spectroscopy. We’ll just lay the probe on the part or resin, press a button, and read the composition in percentages right off a screen. The data will then go via wireless Internet into the QC computer for SPC and lot traceability.

Real recycling will have to mean more than molding mud flaps and floor mats. Keep the costs in line, and a technology such as this could open the door to smaller niche players both in recycling of thermoplastics and specialty molders of recycled materials. It’s an invention that’s not just wanted, but necessary for the future of plastics processing.