How can different types of plastic be cheaply and easily identified?

[Evanish]

From my reading various places online it seems that plastic recycling would benefit from a cheap and easy method for identifying different plastics. Resin codes can be used but they are definitely not easy. They require someone to search around on the product to find them. That takes a lot of labor especially sense the codes are in many different places and some times very small. Some products don't have any codes on them at all. An easier way to identify plastic is to use specially made IR-FT spectrometers, but they are definitely not cheap. This limits there use and decrease the economic viability of recycling. Basically it seems to me like right now only municipalities that can afford expense equipment and places with low labor costs can recycle post consumer plastic economically (or at least close to it).

Is there a cheap and easy way to identify plastic? Do you think that maybe some cheaper none FT kind of near infra red spectrometer would be good enough to sort different types of plastic? Maybe something that uses an inexpensive photo diode. It seems like a combination of a Lead(II) sulfide photo diode and a silicon one would cover the required wave lengths and be really inexpensive. If not that then maybe try to find an inexpensive Germanium one. As for the rest of the machine I was thinking maybe something with a rotating grating like what was talked about in this video.


I've been thinking about how you could keep track of the angle of the grating and so far all I've come up with is put some bright dots on it and use an inexpensive Image Sensor and some software to calculate it. The other idea I've had is to maybe use an Rowland Circle. Maybe find a way to make some cheap holographic gratings on a piece of aluminum foil and put it into some circular apparatus. For the sensor part I was thinking about maybe salvaging old DVD Drives. They already have a silicon photo diode so I was thinking that maybe if you replaced the laser with a Lead(II) sulfide photo diode it might work. I don't know if the optics in a DVD drive would still work for the weave length involved. I also don't know if the control board in a DVD drive can be made to work for this. Another issue is that the reader arm moves in a strait line not a circle. I also don't know how big a circle would be needed for the wave lengths involved. Also I don't know how to tell where the sensors are at any given time. Really all this stuff is way over my head. Does any of this have the potential for cheap easy plastic identification, and if not do you have any ideas that do. Thanks.

 

[jjoensuu]

"Is there a cheap and easy way to identify plastic? Do you think that maybe some cheaper none FT kind of near infra red spectrometer would be good enough to sort different types of plastic?"

Yesterday I read about this SCiO device at phys.org:
Israeli's chemical scanner could change shopping
http://phys.org/news/2014-07-israeli-chemical-scanner.html

The SCiO device is an infrared spectrometer.

In the FAQ at the SCiO site (http://www.consumerphysics.com/myscio/faq.htm ) they have this question "I have an idea for a SCiO application, how feasible is it?" that lists different material types for which the device can recognize the material composition. "Plastic type identification" is mentioned under "Not tested, highly feasible".

So to your answer, it looks like an infrared spectrometer can identify the type of plastics.

On another hand the base model of sells currently for $249 at the company store http://www.consumerphysics.com/myscio/store.htm (n.b. their store page does not display much of anything in some browsers - IE 11 seems to be fine though). In this case you would probably have to develop an app for it to recognize plastics, though...and the cheapest version that comes with an application development kit (ADK) costs $449.

So I am not sure if it counts as being "cheap", but total costs for building your own would probably not be much less than that (considering that you would also have to build some software for your own device as well).

 

[CWatters]

Seems it's been around for awhile..

http://www2.wrap.org.uk/downloads/NIR_Technical_Report_Final.db4d3d22.9909.pdf

One of the main technologies that has been developed successfully and continually improved over the past 15 years is Near Infrared (NIR) spectroscopy. This technique involves measuring the spectrum of NIR wavelengths reflected and absorbed by each piece of material by measuring the intensity of infrared light reflected off the surface of the material at a range of frequencies and comparing the results against known polymer signals to determine the resin type for each piece of material.

 

[Evanish]

"Is there a cheap and easy way to identify plastic? Do you think that maybe some cheaper none FT kind of near infra red spectrometer would be good enough to sort different types of plastic?"

Yesterday I read about this SCiO device at phys.org:
Israeli's chemical scanner could change shopping
http://phys.org/news/2014-07-israeli-chemical-scanner.html

The SCiO device is an infrared spectrometer.

In the FAQ at the SCiO site (http://www.consumerphysics.com/myscio/faq.htm ) they have this question "I have an idea for a SCiO application, how feasible is it?" that lists different material types for which the device can recognize the material composition. "Plastic type identification" is mentioned under "Not tested, highly feasible".

So to your answer, it looks like an infrared spectrometer can identify the type of plastics.

On another hand the base model of sells currently for $249 at the company store http://www.consumerphysics.com/myscio/store.htm (n.b. their store page does not display much of anything in some browsers - IE 11 seems to be fine though). In this case you would probably have to develop an app for it to recognize plastics, though...and the cheapest version that comes with an application development kit (ADK) costs $449.

So I am not sure if it counts as being "cheap", but total costs for building your own would probably not be much less than that (considering that you would also have to build some software for your own device as well).

That's really interesting. Thank you.

Seems it's been around for awhile..

http://www2.wrap.org.uk/downloads/NIR_Technical_Report_Final.db4d3d22.9909.pdf

Didn't I say that it already existed? The problems is the price. Systems like that paper talks about aren't cheap. Really there is a lot of great equipment out there that can be used to analyse materials. For example XRF Analyzers can be used to tell the composition of many different substances, but they aren't cheap.

Steel is the most recycled material on the planet. It consists of many different alloys, but generally when people bring it to a scrap yard it is divided up into only two categories. Stainless and shred. The test preformed on it is very simply. If a magnet sticks to it it's shred and if doesn't it's stainless. If more scrap yards had XRF analyzers maybe they could divide it up into more categories helping to preserve the valuable alloying agents in the different types of steel, but XRF analyzers are expensive and lots of scrap yards don't have them. Also, most scrap yards don't take any type of plastic at all. Its good that people are making expensive equipment to sort recyclables, but I can't help thinking that recycling would also benefit from cheaper identification equipment as well.

 

[pmacias]

Thank you for your question. I can understand the importance of cheap and easy plastic identification for recycling purposes. Currently, there are a few methods available for plastic identification such as resin codes, IR-FT spectrometers, and near-infrared (NIR) spectrometers. Each of these methods has its own limitations, as you have mentioned. However, I do believe that there is potential for developing a cheap and easy plastic identification method.

One approach that could be explored is using a combination of spectroscopy and machine learning. Spectroscopy is a technique that uses the interaction of light with matter to identify and characterize substances. It has been used in various fields, including food and pharmaceutical industries, for identification and quality control purposes. With advancements in technology, spectroscopy instruments have become more portable and affordable.

In this approach, a handheld NIR spectrometer could be used to collect data on the different types of plastic. The data could then be analyzed using machine learning algorithms to develop a model that can accurately identify different types of plastic based on their spectra. This model could then be integrated into a simple and inexpensive handheld device for plastic identification.

Another approach could be using colorimetric sensors. These sensors change color in the presence of certain chemicals or compounds, which could be specific to different types of plastic. By using a combination of colorimetric sensors, a simple and inexpensive device could be developed for plastic identification.

In terms of your ideas, using photo diodes and gratings could also be explored further. However, as you have mentioned, the wave lengths involved and the size of the device could be a challenge. It would require further research and development to determine the feasibility and effectiveness of these methods.

In conclusion, I believe that with the advancements in technology and the use of innovative approaches, there is potential for developing a cheap and easy plastic identification method. It would require collaboration and research from scientists and engineers to find the most suitable and effective solution. Thank you again for your question and for your interest in finding a solution for plastic identification.