Cutting edge technology for sorting plastic waste

We know plastic waste is a problem – the question is, how big is it? In 2017-18, the Pollution Control Board put out a study, estimating India’s annual plastic waste generation at 9.4 MT (million tons), of which approximately 60% is processed. While this looks impressive, especially when compared to the global average of 20%, it belies an uncomfortable truth about how waste is handled in India.

Nothing depicts it better than the dehumanizing, yet depressingly common visuals of young children & women scavenging through garbage dumps to collect plastic. Yet, despite putting life & limb at risk, rag pickers earn just a pittance. Most of the plastic discarded is in the form of thin plastic bags & therefore they need to collect a large number of them to make a kilo – because scrap sells by weight. Recent regulations like increasing minimum thickness of plastic bags to 50 microns, banning single use plastics & making declaration of plastic grade as well as scrap rate on the pack mandatory are all steps in the right direction & as compliance improves, things will get better. Further, . all players in the plastics value chain are now required to collect equal amount of plastic as they have generated through sales of their products & report the details to PCB authorities on monthly / quarterly basis. While Reduce-Reuse-Recycle remains the mantra, what is a simple habit change at individual level, becomes a huge challenge when you scale it up to municipality. Take sorting of plastics, for example.

My interest in this subject was piqued when I came across this news item in the Chemical Engineering journal which highlighted how Canon is using Raman Spectroscopy to overcome the limitations of Near Infrared technology in large scale sorting of plastics! I have often wondered how some of the Nobel prize winning discoveries affect our lives so you can understand I was excited to see this reference to Raman spectroscopy because this is what got the great physicist the Nobel in 1930! I did not know much about IR spectroscopy either so I had to back track and first find out what that was and how the two technologies could help solve what is probably one of the most intractable challenges in recycling: sorting of plastics. Firing a canon to kill a mosquito? No way!

Tomra Systems, Norway of course is a world leader on reprocessing technology and they do everything at massive scale & we will see as we go along why Near Infrared (NIR) is their technology of choice. The innovation of using Raman spectroscopy came from Canon & it is amazing how it has leveraged its core competency in imaging across a wide variety of industries. While we will talk about both the technologies in some detail, the schematic below shows the fundamental difference between the two. NIR is based on reflection, Raman spectroscopy on scattering of light.

The main reason why sorting of plastics is necessary is because the grades vary vastly in chemical composition which rules out co-processing. But unfortunately due to the wide variety of uses, their physical properties are often quite similar so standard segregation techniques like gravity separation don’t work. Also it is very rarely that a product uses just one type of plastic. Say a family going on a road trip for a day. On the way they pick up a few bottles of mineral water & some disposable cups. When they come back in the evening, all the plastic waste is dropped into the bin. Let us say it includes one of the kid’s water bottles as well as the lid cracked when he accidentally dropped it! Now consider the following – the kid’s water bottle is HDPE but the cap (which is usually more robust but brittle) was mostly PVC. The mineral water bottle is PET, but the cap is either HDPE or PP. The disposable cups are polystyrene and the plastic bag in which all the plastic waste was collected was LDPE. So you have as many as five different types of plastics in that trash can!

Of late, in order to reduce requirement of virgin plastic grades, governments are stipulating minimum amount of recycled content in the recipe. In the US for example, PET bottles have to have minimum 10% recycled material (R-PET) & you would be surprised that it is actually quite difficult to find. As the circular economy picks up pace, the need for such high purity recycled material will increase manifold. The challenge therefore, is not just about sorting at scale, but also about a high level of specificity.

The NIR system is installed on the conveyor carrying the unsegregated plastic chips & as they travel they are exposed to the infrared rays. Some of it gets absorbed and the rest is collected in the detector. Inter-molecular bonds within the polymer have their characteristic frequency of vibration & since the incident beam contains a wide range of wavelengths (750-2500 nm), there is a fairly likelihood some the frequencies will match. If that happens, energy from the incident beam corresponding to those frequencies gets transferred to the polymer molecule while the rest of the beam reflects off the surface & is collected at the detector. To visualise this, imagine you are walking down a street when you see a familiar face across the road. She spots you at the same time & gives you an enthusiastic wave which you reciprocate & continue on your way with a smile on our face. In a way she has transferred some of her ‘energy’ to someone whose ‘wavelength’ matched her own!

By studying the wavelengths absorbed and matching it with the built-in library, the IR spectroscope identifies the grade of plastic and a jet of air directs it in to the appropriate bin. This technology is quick and accurate but it has a limitation. Since it works on the principle of reflection, colour (especially black) interferes with the measurement & hence it can be used only for transparent plastics. Before you think it is rather niche, let me hasten to add that it is actually the most common type that is encountered and since all the plastic grades can give transparent or near transparent products, it is impossible to classify them by visual inspection.

Unlike NIR, Raman spectroscopy works on the principle of light scattering. Since this involves some of the concepts of quantum physics, I will do a quick recap of the basics. Matter is made of molecules – which in turn are made of atoms. At a very basic level we can compare the atomic structure with our solar system – where the sun represents the nucleus and the electrons, the planets. Obviously beyond the superficial, there is nothing else in common – obviously unlike planets all electrons are similar in nature & most importantly electrons can move between orbits or energy levels as they are called.

To move an electron to a higher level, energy has to supplied – often through a laser which is a stream of high energy particles called photons. As the schematic below illustrates, when this energy beam hits the atom, some of the electrons in the atoms get excited from their ground state (1) to temporarily move to a higher energy state (2). When the beam is withdrawn, the electrons release energy & fall back to a lower energy level (3). In the process, the photons absorbed earlier are released – scattering uniformly in all directions. When the frequency of the scattered light is of the same frequency as the incident beam – which is most of the time - it is referred to as Rayleigh scattering. Very rarely, of the order of a one-in-a-million, the beam of scattered light has a different frequency from the incident beam. This is called Raman scattering.

In case of Rayleigh’s since the scattered rays are of the same frequency as the incident beam, they don’t yield any new information – whereas in Raman scattering, since the rays are of different frequency, it is possible to deduce what caused it & by extension what the substance is made of! On a lighter note, it reminds me of Corporate Communication meetings over Zoom during Covid. Often, the only voice that the speaker would hear for a long time would be his own - so much so, that he could be excused for wondering if anyone was attending till unexpectedly somebody piped up and asked a question! That is the difference between Rayleigh & Raman scattering!

The difference between the frequency of the incident beam & scattered beam is called the Raman shift and it is unique for each substance & type of bond. Since the Raman shifts for the molecules & bonds being investigated are known, matching that with the peak heights gives a fairly accurate estimate of the composition of the polymer. If I were to oversimplify with an analogy - imagine that schools, offices & malls in your city have designated spots – so that they are quite far from each other. Since traffic pattern at these locations is fairly predictable, given a snapshot of the number of vehicles at each of the locations, you can probably guess the time of the day with reasonable accuracy.

IR is cheap, lightning quick and can be adapted at scale but it cannot handle coloured plastics. Raman spectroscopy on the other hand is unaffected by the colour of plastic but has much lower throughput. That doesnot make one better than the other – just that, depending on what your priority is, you choose the technology or maybe a combination of technologies to go with. Irrespective of whether or not you work in any part of the extended plastics value chain, I hope you found this interesting. ???

That is all for this week. Continuing on the theme of waste disposal, next week we will discuss how sintered glass is becoming the preferred option for disposing off heavy metal waste!