Dissipation of the carbon resource embedded in end-of-life plastics: exploring the limits of chemical recycling to monomer

Purpose
Plastic represents an important category of carbon-based materials and a potential source of carbon resource in a circular economy. In life cycle impact assessment (LCIA), we are currently witnessing a shift from approaches that address resource depletion to those that focus on resource accessibility and dissipation. However, little is known about factors controlling the dissipation of the carbon resource. This study aims to present a new classification approach for identifying dissipative forms of carbon embedded in plastic polymers. 

Methods
We first define the carbon resource in a plastic context based on the thermodynamics of chemical recycling to monomers (i.e., depolymerization yield and the related ceiling temperature, above which depolymerization is favored over polymerization) and the chemistry of carbon in the plastic (i.e., the average oxidation state of carbon atoms in the polymer molecule). We then propose reference values for accessibility, reflecting the most accessible sources of carbon for use as a material feedstock at short- and medium- to long-term perspectives (25 and 100 years, respectively). Finally, we propose two dissipation criteria to differentiate dissipative from non-dissipative forms of carbon in plastics and apply them in practice to 51 polymers representing dominant plastics used globally. 

Results and discussion
For the short-term perspective of 25 years and a threshold ceiling temperature of 433 K, carbon in 12 out of 23 polymers is classified as dissipative, while in the remaining 11, it is classified as non-dissipative. For the medium- to long-term perspective of 100 years and a threshold ceiling temperature of 640 K, three out of 23 polymers contain carbon classified as dissipative, and in the remaining 20, non-dissipative. When the carbon oxidation state is used as the dissipation criterion, 44 polymers out of 51 contain carbon classified as dissipative in the short term, while 11 out of 51, contain carbon classified as dissipative in the medium–long term. The two criteria agree on carbon classification in 11 (short term) and 18 (medium–long term) out of 23 cases. 

Conclusions
Unlike depletion-oriented resource accounting approaches, which do not differentiate between different forms of carbon, this work shows that not all carbon forms may be equal from a resource dissipation perspective. These differences should be considered in the development of life cycle inventory procedures and matching characterization factors for the carbon resource.