From Recycled Plastic to Pyrolysis Oils: A Renewed Need for a Broad Range of Gas Chromatography (GC) Detector Technologies

The global plastics recycling market is currently $50 billion dollars and growing rapidly. Although thermal processing is still part of the procedure, much effort has gone into developing new technologies for catalytically cracking polymers back into constituent monomers. These constituent monomers can then serve as feedstocks for fuels and other chemical products. There are several challenges in characterizing these monomer mixtures. They vary in content with the variability of input material and cracking conditions. The resultant monomer mixtures, or pyrolysis oils, are complex and require careful characterization for further processing. They are high in olefin content, and it is desirable to know their aromatic content as you would for a traditional refined hydrocarbon stream, but there is also a large need to characterize heteroatom-containing species in the mixture. Excessive heteroatom content can further compromise processing the pyrolysis oils if not appropriately accounted. Gas chromatography (GC) is an indispensable tool, but the traditional pantheon of detectors do not provide satisfactory performance to accomplish the task. There is a resurging need for detectors that can detect and characterize heteroatom-containing species. Many were developed long ago and exist in textbooks but are unfamiliar to those who work in analytical laboratories. It is time to revisit these detection technologies and involve them more in analysis to help advance the growing plastics recycling market. Some of these detector types are sulfur chemiluminescence, flame photometric, nitrogen- phosphorous, and vacuum ultraviolet (VUV) detectors.

As a child, I was fascinated by recycling plastic. I learned to connect the numbers on plastic bottles and containers with different chemical names, even if I did not understand what the names meant. I remember that, initially, some plastics were acceptable to put in the recycling bin, whereas others were not. Where I grew up, that approach changed over time to one where they at least collected all plastics, even if maybe they did not process them all. I remember visiting different places and noticing that the recycling instructions and capabilities often varied considerably with location.

Plastics production has increased almost 10% every year since 1950. Today, the global production of plastics is around 300 metric tons (MT) annually. Plastics can have variable life spans, but roughly 40% is estimated to have a usage life of less than one month. More than three-quarters of the plastics produced are recyclable thermoplastics. These thermoplastics include polyolefins, such as polyethylene, polypropylene, polystyrene, and polyvinylchloride, among others. Once waste plastics are collected, they are initially separated based on density, before they are subjected to a cracking process to depolymerize the material for refining and reusing. There are three predominant cracking processes used, namely hydrocracking, thermal cracking, and catalytic cracking. The latter is currently an area of intense research and development.

As the name connotes, catalytic cracking involves the use of a suitable catalyst to carry out the depolymerization process. The use of catalysts reduces both the time and temperature needed for cracking. The resulting products, often termed pyrolysis oils, are characterized by a narrower distribution of carbon atom number, centered more around lighter carbon chain lengths, and are attained at much lower temperatures compared to the products of other cracking processes. Once plastics are cracked and characterized, the resulting pyrolysis oils can be further refined into fuels and feedstocks for chemical production. Anecdotally, I have heard that diesel fuel could be obtained from refined pyrolysis oils that is cleaner than that generated from a traditional petroleum stream. Catalysts and their application conditions are constantly under development. Clearly, this development must be aided by capable analytical measurements to characterize the pyrolysis oils produced.

Currently, there appears to be an almost limitless feedstock for recycled plastics. On the one hand, the global plastics demand is expected to continue to increase over the next 30 years. On the other hand, by 2050, more than half of plastic production could be based on plastic reuse and recycling. The share of reused plastics from recovered monomer or recovered feedstock from catalytic cracking is expected to increase by more than 15% annually in the same time frame. Environmental and social corporate governance (ESG) considerations will continue to drive the use of recycled products and feed stocks over virgin sources (1).

The large variability in the plastics feedstocks, catalytic processing, and the resulting pyrolysis oils presents a number of significant challenges to the analytical chemist. Generated pyrolysis oils can be characterized in much the same way as hydrocarbon streams, but their content is markedly different. For example, it is not uncommon for pyrolysis oils to have as much as 50% or more olefin content. Further, heteroatom content (such as sulfur, nitrogen, and oxygen) can be highly variable and much higher than in hydrocarbon streams. These contents need to be well known and understood for pyrolysis oils to be further refined. For example, excessive sulfur content can poison many catalysts, if not appropriately accounted for or removed.

Homing in on the heteroatom content in a mixture of hydrocarbons ranging from approximately C6 to C40 requires more than just the standard gas chromatography (GC) detectors. This field of hydrocarbon analysis is the first place where I have seen a clear application for many GC detection systems that were only known to me previously as an acronym. The comprehensive analysis of pyrolysis oils will require the use of multiple separation and detection systems.

Without ultrahigh resolution, mass spectrometry (MS) will get quickly bogged down in the extreme number of isomers and mixed class species, as the carbon number increases. Ultrahigh-resolution MS is a powerful, but expensive, tool that has been applied to the characterization of pyrolysis oils from a wide range of feedstocks (2). One can often delineate heteroatom content as part of generated elemental formulae, but it can still be difficult to completely define to what chemical compound these heteroatoms are connected, without the aid of other tools, including some selective separations. A recent publication from Focant and coworkers demonstrated the power of on-line comprehensive two-dimensional GC (GCxGC), in combination with photoionization–time-of-flight (TOF) MS, for characterizing commercial dodecene (primarily, C12 olefin) mixtures (3). This complexity already provides a significant challenge for one of the highest resolution analytical approaches we have available.

Vacuum ultraviolet (VUV) spectroscopic detection offers considerable power for overall classification of species as paraffins, isoparaffins, olefins, naphthenes, or aromatics (PIONA) (4). Chemical classification based on gas phase absorption detection from 120 to 430 nm has been shown to be highly complementary to MS measurements, especially delineating isomeric species, but its application to high carbon number species still needs further development. Although VUV may have significant ability to delineate mixed class species (such as a chemical compound that possesses both olefin and aromatic character), its ability to delineate heteroatom attachment to that species is generally lacking. A recent publication by Dunkle and coworkers is a good example of the state-of-the-art in the application of GC-VUV for the characterization of pyrolysis oils in comparison with standard hydrocarbon streams (5).

Inevitably, the conversation around characterizing heteroatom content in pyrolysis oils turns to some acronyms that you have probably heard when first learning about GC detection but never had a chance to use. Flame ionization detection (FID) is quite ubiquitous in GC, but its usefulness in the current application is minimal, given its response to any carbon-containing molecule and its inability to provide any further qualitative information. The same could be said for thermal conductivity detection (TCD).

Instead, in this instance, more selective approaches, based on technologies like sulfur chemiluminescence detection (SCD), nitrogen-phosphorous detection (NPD), flame photometric detection (FPD), and flame thermionic detection (FTD) come to mind.

As the name connotes, SCD is a highly selective and quite sensitive (low picogram detection limits) detector for sulfur. A notable advantage of SCD is its ability to respond uniformly to sulfur-containing compounds; in other words, a calibration curve prepared for one sulfur-containing compound is often sufficient for the analysis of another if both compounds contain the same number of sulfur atoms. In the detector, sulfur species are first oxidized under extremely high temperature (around 1000 ^oC). They are then made to react with ozone, where excited-state sulfur species will chemiluminesce, releasing light that can be detected. The SCD is generally considered to be a more complex GC detector to operate.

Another detector with obvious applications is the NPD. It provides selective detection for nitrogen- and phosphorous-containing compounds. Nitrogen-containing compounds can be enriched in pyrolysis oils generated from plastics containing dye molecules. The NPD is a type of FTD that relies on heating the effluent in the absence of a significant hydrogen:air ratio. This approach minimizes hydrocarbon ionization, but facilitates ionization of the nitrogen- and phosphorous- containing compounds, the signal for which is collected by an electrometer. Because of its excellent sensitivity (low picogram detection limits), high purity gases for operating the detector is a must.

The FTD also responds selectively to nitrogen- and phosphorus-containing compounds. It relies on the passage of the effluent across a plasma created around a coiled alkali ion source (a rubidium salt). Rubidium radicals react with organic nitrogen species (this detector does not detect inorganic nitrogen species) and phosphorus-compounds to create ionic species, which are registered using a collector electrode.

FPD provides selective detection of phosphorus-, sulfur-, and organic tin compounds. It relies on the combustion of these compounds and their release of light at specific wavelengths. Using a filter, only the wavelengths of light of interest are passed to the detector to be registered as a signal. The FPD is generally considered easier to operate than some of the other element-specific detectors, but it is also less selective.

Of course, the list of detectors presented here is not exhaustive, and other detectors may also find their place in this field. Regardless, the notion of pairing data from highly selective detectors for specific element–containing compounds with detectors like MS and VUV that can elaborate a broader picture and classify different compounds, seems like a powerful combination. One can imagine post-column splits to different detectors, but this is not necessarily ideal in terms of maximizing sensitivity. Some detectors offer the potential to flow through, such as the case for VUV, so that a second detector could be placed in series. However, this approach has not been well developed, to date. Of course, linking data acquired from two different systems with different detectors would be possible, but aligning the data also presents its own challenge.

In the end, there is room to innovate to create more powerful analytical configurations that are applicable for use in a market that is growing by leaps and bounds. Rarely a day goes by that you cannot find the announcement of another initiative or partnership that expands plastics recycling and processing worldwide. Just as the use of multiple detectors can offer significant benefit to delineating heteroatom content in addition to classification of molecules in the mixture, so will increasing use of multidimensional separations. Given the ultimate complexity of compounds desired to be speciated in pyrolysis oils, analysis of recycled hydrocarbons is a rich area for the further development of multifaceted GC technology, including advanced data treatment algorithms.

Disclaimer

The author is a member of the scientific advisory board for VUV Analytics, Inc.

References

(1) K.A. Schug, The LCGC Blog: Environmental and Social Governance (ESG) in the Hydrocarbon Processing Industry (HPI) is an Opportunity for Industrial–Academic Partnerships. LCGC, [online] 2021. https://www.chromatographyonline. com/view/the-lcgc-blog-environmental-and-social-governance-esg-in-the-hydrocarbon-processing-industry-hpi-is-an-opportunity-for-industrial-academic- partnerships.

(2) R.L. Ware, S.M. Rowland, R.P. Rodgers, and A.G. Marshall, Energy Fuels 31, 8210– 8216 (2017).

(3) Y. Zou, P.-H. Stefanuto, M. Maimone, M. Janssen, and J.-F. Focant, J. Chromatogr. A 1645, 462103 (2021). DOI: 10.1016/j.chroma.2021.462103.

(4) P. Walsh, M. Garbalena, and K.A. Schug, Anal. Chem. 88, 11130–11138 (2016).

(5) M.N. Dunkle, P. Pijcke, W.L. Winniford, M. Ruitenbeek, and G. Bellos, J. Chromatogr. A 1637, 461837 (2021). DOI:10.1016/j.chroma.2020.461837.

Kevin A. Schug is with the Department of Chemistry & Biochemistry at the University of Texas at Arlington, in Arlington, Texas. Direct correspondence to: kschug@uta.edu.