A research team at the University at Buffalo has demonstrated a plastics identification method that reads "three-dimensional transient thermal barcodes" — heat patterns generated when mid-infrared light strikes plastic waste on a conveyor. The work, published in Nature Communications Engineering under the journal's early access policy, comes from the New York State Center for Plastics Recycling Research and Innovation at UB, a center designated by the New York State Department of Environmental Conservation.

The technique uses six mid-infrared wavelengths to identify all six common plastic waste types: PET, PP, PS, HDPE, LDPE and PVC. Mid-infrared is known as the "molecular fingerprint regime" because every plastic type shows unique spectral peaks in that band. As the plastic absorbs the light, molecular bonds vibrate and generate temporary heat patterns, which a thermal camera measures from a distance. Each pattern reflects the plastic's molecular structure — effectively a barcode readable in standoff mode, a requirement for material recovery facility conditions.

Crucially for MRF operators, the system also identifies black plastics. All black plastics are based on one of the six commodity types but carry significant carbon additives that defeat conventional optical sorting. Black plastic detection remains one of the persistent technical gaps in NIR spectroscopy, Raman spectroscopy and laser-induced breakdown spectroscopy — methods that suffer, according to the researchers, from poor sensitivity, subpar selectivity, slow speeds, black plastic blindness or the inability to operate in standoff mode.

The scale of the opportunity is material. Household and business plastic waste arrives at material recovery facilities where workers often separate it by hand, producing false identifications and contaminated bales. The absence of cost-competitive, accurate resin-code identification is one reason plastic recycling rates remain low worldwide. The UB team frames its technology as directly targeting that bottleneck.

"Our goal was to develop a cost-effective, scalable and industrially relevant plastics sorting technique that addresses the key prevailing scientific gaps restricting the recycling of plastics," says corresponding author Amit Goyal, PhD, SUNY Distinguished Professor and SUNY Empire Innovation Professor in the UB Department of Chemical and Biological Engineering. Goyal directs the UB Initiative on Plastics Recycling and Innovation. The system aims to "improve the quality of sorted plastics by reducing contamination and, hence, increasing the recycling of these materials to help enable a circular economy."

The researchers cite an estimate that one ton of recycled plastic saves 5.7 megawatts of electricity, 685 gallons of oil and 30 cubic yards of landfill space.

The system is not ready for industrial deployment. The team — led by lead author Kunal Singh, PhD, a postdoctoral fellow mentored by Goyal and co-corresponding author Thomas Thundat, PhD, a SUNY Distinguished Professor at UB — is now working on three fronts: faster hardware capable of sorting on high-speed conveyor belts; cost-effective light sources that project multiple mid-infrared wavelengths simultaneously; and AI-based software to improve performance. All three authors are affiliated with the Department of Chemical and Biological Engineering in UB's School of Engineering and Applied Sciences.

Goyal and Thundat credit Singh not only with the scientific research but with custom instrumentation, saying his "rigorous and meticulous work identified key shortcomings in our initial approach and provided elegant technical solutions."

"We hope to further develop the technique so that it can be transferred to industry," Singh says.

For now, the finding stands as validated lab-scale proof: the thermal barcode concept works at bench level and identified all six resin types plus carbon-laden black variants. What decides whether it reaches MRF conveyor lines is the engineering roadmap the team itself has set — high-speed hardware, multi-wavelength light sources at commercial cost, and AI sorting software robust enough for retrofit into existing sorting machinery. Each of those milestones, not the published proof of concept, will determine whether New York's state-designated plastics research center delivers an industrially deployable sorter.