Introduction

The massive global production and utilization of single-use plastics has resulted in a unique predicament where plastic waste continues to both accumulate and persist in the environment, leading to various impacts on human and environmental health. [1][2][3][4] With more than 460 million metric tons of plastic generated annually and a projected increase to 1.2-1.6 billion metric tons by 2050, 5,6 many strategic approaches have been implemented toward a sustainable plastics economy, including developing efficient recycling methods, [7][8][9] reducing single-use School of Polymer Science and Engineering, The University of Southern Mississippi, Hattiesburg, MS, 39406, USA. E-mail: zhe.qiang@usm.edu plastics, 10,11 and building infrastructure for improved waste management. 12,13 Particularly, an enormous amount of effort has been focused over the past decade on developing scalable methods of converting plastic waste into reprocessed/reengineered products or valuable raw materials, enabling their extended use and lifetime. 14,15 Historically, landfilling and incineration have been the most prevalent waste management methods for end-of-life plastics, however, they directly lead to microplastic generation/contamination or toxic gas emissions. 8 Mechanical recycling offers a simple approach to recycling a wide range of commercial thermoplastics, though this practice is limited to single-sourced and high-quality waste feedstocks. 16 If waste feedstocks are not properly separated or compatibilized, phase separation between differing plastics leads to poor interfacial adhesion and reprocessed plastics exhibit significantly reduced mechanical properties. 17 Moreover, mechanical reprocessing can cause polymer chain scission and/or crosslinking, which reduces material properties and processability, restricting this method to only a few life cycles of plastics. 18 Additionally, a consistent limitation with mechanical recycling is its inability to address thermoset polymers, excluding a significant portion of total plastic waste. 19,20 While chemical recycling has been demonstrated as an effective route to convert various commodity polymers to chemical building blocks or fuels, sequential use of new products could often result in additional cost and energy use, potentially offsetting environmental benefits. 14,21 Another important consideration is that secondary raw materials may not be cost competitive compared to their inexpensive virgin analogs. 22,23 Chemical upcycling of complex plastic wastes into valueadded materials is a promising route to simultaneously address issues related to both recycling and economic feasibility. 24,25 Pyrolysis, or a thermochemical process involving heating under an inert environment, has been demonstrated to produce various advanced materials and/or chemical building blocks from a wide range of plastic precursors. 24,26,27 While toxic offgases are produced during pyrolysis, these byproducts can be sequestered as well as repurposed as energy sources. [28][29][30] We note that the processing parameters in pyrolysis can be tuned to modulate the formation of upcycled products; in some literature precedents, slow heating rates can result in increased amounts of char while fast heating rates can favor gaseous or wax-based molecules. 31,32 Recently, several industrial-scale pyrolysis plants have been developed for plastic waste recycling aimed at producing pyrolysis oil, or mixtures of small molecules, which could be used feedstocks for ethylene cracking. 33,34 Alternatively, conversion of plastic waste into carbon materials can be carried out through pyrolysis. 25,35 We note that advanced carbon materials, ranging from carbon fiber, carbon nanotube, graphene, to porous carbon, could provide enhanced value due to their combination of physical properties that confer excellent performance in various applications. [36][37][38][39] For instance, porous carbons, due to their chemical stability, high surface area, and tunable pore texture, 40,41 are a versatile family of materials suitable for various applications, including water remediation, [42][43][44] biomedicine, 45,46 CO 2 capture sorbents, 40,47,48 catalysts, [49][50][51][52] and catalyst supports. [53][54][55] Moreover, due to their high electrical conductivities and structural stability, carbon materials can be used in energy storage, acting as both supercapacitors and electrodes in high-performance devices. [56][57][58] Owing to their excellent mechanical properties and high strength-to-weight ratio, carbon fibers are often utilized in high-performance composites ranging from aerospace to construction applications. 59 This broad application space is primarily due to the wide range of carbon material properties. Specifically, carbon-based sorbents exhibit promising performance properties due to tunable pore textures, high surface areas, chemical stability, and their ability to be chemically functionalized such as doping. Electrocatalysts and capacitors can also benefit from high surface areas and tailorable surface chemistries, while leveraging carbon's high electrical conductivity. In comparison, carbon fibers used in structural applications primarily rely on their excellent mechanical properties to maintain structural integrity, and therefore, the presence of pores can result in defective sits which may reduce material strength. The extensive application space for carbon materials and their wide range of critical material properties demonstrates the importance of understanding how processes and different feedstocks impact final carbon structure and performance. One route of upcycling waste to carbon materials involves mixing plastics with catalyst species followed by catalytic decomposition into gas-or liquid-phase sources, which then interact with catalysts to produce carbon nanomaterials. 60 This concurrent waste degradation and carbon formation pathway has been utilized for preparing carbon nanotubes, graphene, and carbon nanofibers from polyolefins (POs). 35 Specifically, linear low-density polyethylene (LLDPE) can be converted to carbon nanotubes, carbon nanofibers, and amorphous carbon through catalytic carbonization with a nickel(III) oxide catalyst and polyvinyl chloride (PVC) resin, which acts as a source of chlorine radicals to promote LLDPE dehydrogenation and aromatization. 61 In an example of using a sequential approach, plastic waste is first stabilized through a crosslinking reaction to form an efficient carbon precursor. Once plastic waste is thermally stabilized, carbon products are then formed through pyrolysis under an inert environment. Crosslinking of plastic waste can be achieved through thermal treatment under air or through the use of chemical agents to facilitate crosslinking. 25 For instance, Lee et al. utilized thermal oxidation to crosslink LLDPE, allowing its subsequent conversion to graphitic carbon with up to 50 wt% carbon yield. 62 Following thermal oxidation at 330 1C and pyrolysis up to 1200 1C, wastes derived from commercial LLDPE products achieved comparable carbon yields to virgin materials. While several studies have demonstrated heat treatment is capable of stabilizing plastic waste, [63][64][65][66] most commercial polymers do not have optimal molecular structures to undergo thermal oxidation-induced crosslinking. 25 To overcome this, chemical agents have been used to promote crosslinking of various commodity plastics. 67 For instance, exposing polyolefin waste to concentrated sulfuric acid at elevated temperatures allows for high degrees of crosslinking and conversion of polymers into efficient carbon precursors. 68,69 Briefly, the polymer This journal is © The Royal Society of Chemistry 2025 backbone is functionalized with sulfonic acid groups which then dissociate forming double bonds, followed by a series of additions and rearrangements resulting in intermolecular crosslinking. Thermal stabilization of polystyrene (PS) has also been achieved using carbon tetrachloride (CCl 4 ) acting as a crosslinker, resulting in hierarchically porous carbons with relatively high surface areas (B679 m 2 g À1 ) after pyrolysis. 70 Often, carbonized materials from plastic wastes can be post-functionalized through steps such as surface modification or activation to further improve their performance and functionality for target applications. 71,72 Postcarbonization modification involves introducing various species, ranging from metals, small molecules, or polymers, onto the carbon surface to impart varied functionality, while activation uses reactive agents to encourage the formation of micropores within the carbon matrix, and thus increasing surface areas. 73 While several reviews delve into converting plastic waste into specific carbon nanomaterials for target applications, 32,35,67,73 the intent of this review is to discuss the core aspects of various commodity plastics, and how these distinct polymer structures influence their respective optimal carbonization processes and resulting carbon material properties (Fig. 1). Specifically, we will discuss the predominant classes of commercial thermoplastic systems, including from POs, polyethylene terephthalate (PET), and PS, with an emphasis on relations between processing conditions, carbon structure, and performance properties. Following discussion of thermoplastic systems, upcycling of more complex waste feedstocks is covered, including engineering plastics and thermosets. A summary of key upcycling parameters (precursor, carbonization methodology, carbon yield, and target applications) for works discussed is provided in Table 1. Moreover, we provide a brief perspective on the outlook of pyrolysis-based processes for upcycling plastics wastes into carbon materials highlighting future challenges and opportunities.t

Commodity thermoplastic waste

Polyolefins

Constituting more than half of global plastic production, polyolefins (POs), including high-density polyethylene (HDPE), low-density PE (LDPE), and isotactic polypropylene (iPP), can exhibit tailorable mechanical properties due to decades of advances in synthetic and processing mthods, allowing for precise control of molecular structure and polymer crystallinity. 21,74,75 Briefly, HDPE is a linear polymer capable of efficiently packing into zigzag-like orthorhombic crystals granting high degrees of crystallinity, hardness, and tensile strength, whereas LDPE is a type of branched polymer, with lower crystallinity, higher optical transparency, and increased mechanical flexibility. These materials possess significant longevity in the environment after service life that can last at least a few decades. Currently, only a small amount of PO waste is recycled, with the vast majority still being incinerated or landfilled. 17 Current industrial recycling processes of POs largely rely on mechanical recycling, while a substantial amount of effort is focused on developing scalable chemical recycling routes. 76 Conversion of POs to carbon structures has been studied for decades, which in general requires an intermediate step of chemical crosslinking. For example, Pennings et al. reacted linear low-density PE (LLDPE) with chlorosulphonic acid at room temperature to achieve high degrees of crosslinking. 68 Upon pyrolysis under tension between 600-1100 1C in an inert atmosphere, carbon fibers with a Young's modulus of 60 GPa and a tensile strength of 1.15 GPa were obtained. Expanding on this approach, Naskar et al. formed hollow carbon fibers by fabricating a bicomponent fiber containing PE and polylactic acid (PLA), which led to patterned structures following PLA removal. 77 To upcycle PO waste streams to carbon materials, Qiang et al. used the sulfonation-based crosslinking method to convert PP-based disposable masks to multifunctional carbon fibers. 78 Specifically, PP fibers were exposed to sulfuric acid at 155 1C for 4 h and then carbonized at 800 1C, resulting in sulfur-doped carbon fibers which exhibited excellent sorbent capabilities for both adsorption of oil and organic dye (basic blue 17). In a follow-up study, 48 PP-derived carbon fibers were utilized for the capture of CO 2 molecules, reaching high adsorption capacities (3.11 mmol g À1 ) and CO 2 /N 2 selectivity (445) primarily due to high affinity of CO 2 molecules to the sulfur heteroatom within the carbon matrix. Furthermore, the conversion of macrostructured POs into carbon materials was investigated using PP-based textile wastes as illustrated in Fig. 2a. 79 Through sulfonation-induced crosslinking at 150 1C, fabric wastes with fiber diameters of B50 mm were able to reach a carbon yield of 59 wt% once carbonized at 800 1C (Fig. 2b). Fig. 2c shows the intricate woven architectures of the initial fabrics were retained following upcycling to woven carbon fiber mats, with a dimensional shrinkage of B30%, and the retention of complex microstructure was further confirmed through scanning electron microscopy (SEM) in Fig. 2d. These PP-derived carbon mat exhibited excellent Joule heating performance as seen in Fig. 2e and f. Moreover, Joule heating performance was enhanced, and surface functionality was tuned, by introducing a simple copper electroplating step. Through a similar process, three-dimensional carbon structures have been achieved by fused filament fabrication (FFF) 3D-printing of PP, 80 PE, 81 This journal is © The Royal Society of Chemistry 2025 Chem. Commun., 2025, 61, 10470-10483 | 10473 and their mixture prior to thermal stabilization and pyrolysis. 53 Additionally, LDPE-based plastic bags can be upcycled to porous sulfur-doped carbon for lithium-sulfur batteries through microwave-assisted sulfonation, 82 where it was found that the microwave process promotes the sulfonation reaction while also inducing a large number of micropores. Another PE upcycling strategy involved grinding PE plastic bags with basic magnesium carbonate pentahydrate, 83 a flame-retardant agent which acts as a template and also increases PE thermal stability. Following carbonization to 900 1C under a nitrogen environment, hierarchically porous carbons (HPCs) can be obtained after a sequential step of ammonia treatment for simultaneous activation and nitrogen doping. The resulting carbon can serve as electrodes for symmetric supercapacitors, displaying high energy densities (43 W h kg À1 ) and cycle stability (97.1%) at 2 A g À1 after 10 000 cycles. The fabrication of straight and helical carbon nanotubes has also been demonstrated through catalytic pyrolysis of PE with maleated PP and ferrocene at 700 1C with a 12 h hold. 84 In this study, PE-derived carbon nanotubes exhibited a total yield of 80 wt%, specifically containing 5 wt% for helical nanotubes, with diameters ranging from 20-60 nm.

Polystyrene

Polystyrene (PS) is primarily utilized in single-use applications such as food packaging, construction, and insulation materials. 85 Specifically, general purpose PS is a glassy, transparent material largely used in products prepared through extrusion or injection molding processes. In contrast to the rigid PS, expanded or foamed PS is produced by incorporating blowing agents to expand PS materials to form a closed cell structure consisting predominantly of air (B95%). Through this process, expanded PS becomes highly insulative and lightweight. Moreover, high impact PS (HIPS) with superior toughness can be manufactured by introducing a small amount of polybutadiene (5-10 wt%) rubber that grants higher impact resistance. Despite some recent work demonstrating the upcycling of PS into highvalue small molecules through thermochemical degradation, 86,87 recycling of PS is generally challenging due to high recycling costs as well as low density of expanded PS (which can lead to high transportation cost due to increased volume). 85,88 However, the upcycling of PS waste through pyrolysis can create carbon materials with diverse nanostructures, which might be economically competitive.

In a study by Fu et al., 70 HPCs were fabricated from PS through a stabilization reaction followed by pyrolysis. By introducing CCl 4 , which acted as both a solvent and crosslinker, to PS in the presence of an aluminum chloride (AlCl 3 ) Friedel-Crafts catalyst, a high degree of crosslink density was achieved via carbonyl crosslinking bridges. Upon carbonization, carbon samples with macropores (up to 400 nm), mesopores, and micropores were obtained with a pore volume of 0.66 cm 3 g À1 and a surface area of 679 m 2 g À1 . They found that the PS nanoparticles formed interconnected three-dimensional networks after crosslinking, with this loose aggregation forming macro-and mesopores, and this morphology was retained after converted to carbon. A similar approach using CCl 4 /AlCl 3 was employed to prepare PS-derived HPCs as anode materials for high-rate lithium-ion batteries, 89 exhibiting stable capacity of 410 mA h g À1 over 100 cycles. In comparison to amorphous, non-porous carbons, the improved performance from PSderived HPCs was directly attributed to the porosity derived from the crosslinked PS precursor. In another work, Tang et al. investigated the upcycling of PS waste into 3D-HPCs, as shown in Fig. 3a. 90 PS waste was first blended with iron(III) oxide (Fe 2 O 3 ), which acts as both a template and catalyst, carbonized at 700 1C for 1 h, and activated with KOH at 800 1C for 1 h, resulting in HPC with a surface area of 2100 m 2 g À1 and pore volume of 3.03 cm 3 g À1 ; their microstructure was further investigated through SEM and transmission electron microscopy (TEM) imaging (Fig. 3b and c). Additionally, PS foam has been demonstrated to form activated carbon through pyrolysis and chemical activation with the absence of intermediate stage of stabilization treatment. 91 Specifically, by dissolving waste PS foam in acetone, drying, and heating to 530 1C with a 10 1C min À1 heating rate, a carbon char can be formed. Activated carbons with surface area up to 2712 m 2 g À1 and a pore volume of 1.2 cm 3 g À1 were prepared by activation with KOH, leading to sorbents capable of uptaking methylene blue with a capacity greater than 1 g g À1 . Another strategy involved upcycling PS microplastics through stabilization via nitrification which were then pyrolyzed with iron (Fe) and molybdenum (Mo) precursors to form bimetallic Fe/Mo-doped sponge carbons for Fenton reaction catalysts with enhanced electron transfer. 92 The synthesis of advanced carbon nanomaterials, including carbon nanotubes, 93 carbon dots, 94 carbon microspheres, and graphene, from PS waste have also been demonstrated. For instance, Jou et al. prepared carbon nanotubes from the catalytic pyrolysis of polystyrene at 700 1C in the presence of iron nanoparticles, 93 where it was found that nanotubes with outer diameters from 7.5 to 25 nm can be attained. Specifically for the 25 nm nanotube, an inner diameter of 3 nm was observed indicating a much higher wall thickness than nanotubes prepared through an identical procedure from PP (outer diameters of 16.5-40 nm with wall thicknesses ranging from 3.8-14 nm). The thicker nanotube walls were attributed to the generation of aromatic hydrocarbons during polymer decomposition, in contrast to the aliphatic hydrocarbons from polyolefins, yielding secondary deposition on nanotube surfaces. It is important to note that carbon dots can also be synthesized with high yields (60-85 wt%) from waste PS foams by a solvothermal method in dichloromethane (DCM) and nitric acid at 200 1C for 5 h. 94 Waste-derived carbon dots displayed tunable photoluminescence ranging from white to orange and yellow, allowing for their use as multi-color light-emitting diodes. Furthermore, waste PS cups and commercial PS were converted to solid carbon microspheres (diameters between 4-10 mm) through thermal pyrolysis at 700 1C under autogenic pressure (B1000 psi), 95 where both feedstocks produced carbon microspheres; however, the waste PS foam cups required a slightly lower time hold (1 h) compared to commercial PS (2 h).

Polyethylene terephthalate

PET, the most abundant polyester consisting of ethylene glycol and terephthalic acid units, has been extensively utilized for various applications, primarily for single-use packaging and synthetic fibers. 96 Though PET waste has been mechanically recycled for decades, the U.S. recycling rate for PET bottles continues to remain relatively low. 97 In addition, less than 10% of recycled PET is considered commercial-grade usable material due to significant chain scission that occurs during reprocessing. 98 While several examples of chemical recycling of PET down to monomeric species have been demonstrated, 99 upcycling them to carbon nanomaterials can present an economically competitive product.

Various PET waste-derived carbon materials have been developed, 25 the most common of which being porous carbons. For example, waste PET bottles were converted to microporous carbon by first dissolving them in a mixture of DCM and trifluoroacetic acid, 100 electrospinning to PET fibers, stabilizing fibers by heating to 220 1C with a heating rate of 5 1C min À1 and a 2 h hold, and then carbonizing at 750 1C. The final carbon exhibited a surface area of 457 m 2 g À1 . Ok et al. fabricated porous carbon as CO 2 capture sorbents from PET waste, which was collected from several sources, washed and cut into small pieces followed by pyrolyzing to 600 1C with a 1 h hold. 101 To functionalize the PET-derived carbons, activation was carried out, resulting in high CO 2 adsorption capacities up to 3.28 mmol g À1 . Through life-cycle and techno-economic assessment, it was determined that this upcycling method can display both economic feasibility and environmental benefits, particularly when using CO 2 as the carbon activation agent. Furthermore, porous carbon cuboids (PCCs) were manufactured from waste PET bottles by first preparing a calcium metal organic framework (Ca-MOF) through solvothermal synthesis of PET waste with calcium chloride at 180 1C for 24 h (Fig. 4a). 102 The Ca-MOFs were then pyrolyzed up to 700-900 1C to form the final PCCs. Fig. 4b shows the nitrogen physisorption isotherms for the PCCs carbonized at 700 1C, 800 1C, and 900 1C, which all exhibit a type IV isotherm. Pore size distributions derived from density functional theory (DFT) are shown in Fig. 4c, where carbons pyrolyzed at all temperatures exhibit pore sizes centered at 0.5 nm, 1.3 nm, and B12-14 nm. The surface areas corresponding to micropores, mesopores, and macropores are summarized in Fig. 4d, where the microporous surface area decreases with higher carbonization temperature, potentially due to pore shrinkage. Additionally, Lee et al. demonstrated the synthesis of porous carbon from PET waste through autogenic pressure carbonization at 600 1C and followed by KOH activation at 800 1C. 103 By decomposing PET bottles in a confined reactor, a series of secondary reactions can take place eventually forming pyrolytic carbon. With a high surface area of 1931 m 2 g À1 , these PET-derived carbons were used as CO 2 capture sorbents, exhibiting an adsorption capacity of 4.44 mol kg À1 and a CO 2 /N 2 selectivity of 7.99.

Carbon nanomaterials have also been synthesized from PET waste through a molten salt approach, 104 where PET waste was blended with sodium chloride (NaCl) and heated to 1300 1C under air at 10 1C min À1 . After removal of salt, graphitic nanosheets were observed with a thickness less than 10 nm, a bulk density of 1 g cm À3 , and a high electrical conductivity of 1150 S m À1 . Hassan et al. studied the synthesis of graphene from PET waste by heating to 800 1C for 2 h under nitrogen with a catalyst prepared from citric acid combustion of anhydrous iron(III) chloride and aluminum nitrate nonahydrate. 105 The nanostructure of derived graphene was probed using TEM, where single, or a few, layers of sheets were all observed. Furthermore, PET waste from used water bottles were exposed to a rotating cathode arc discharge technique to prepare nanosized, ultrafine solid carbon spheres with an average diameter of 221 nm at lower anode temperatures (1700 1C), whereas elevated temperatures (2600 1C) led to carbon tubules and multiwalled carbon nanotubes (MWCNTs) with an averaged diameter of 364 nm and 20 nm, respectively. 106 Additionally, carbon dots can be prepared from PET textile waste by blending PET with ethylene glycol and exposing to microwave irradiation (540 W) for 20 min. 107 Carbon dots were then fabricated through a hydrothermal approach by blending the unpurified mixture with urea and pyromellitic acid in an autoclave at 260 1C for 24 h. PET-derived carbon dots were found to have an average particle size of 2.8 nm with excitationindependent emission ranging from 340-440 nm and a fluorescence quantum yield of 97.3%.

Complex waste streams

Mixed plastics

Compounding the many issues of thermoplastic recycling, mixed plastics, containing contaminated and comingled waste streams represent an even greater challenge, as conventional mechanical recycling can result in substantially downgraded properties due to weak interfaces from phase separation. 108,109 Various compatibilizers have been developed to improve mechanical recycling feasibility, 17,[110][111][112] while several chemical recycling methods are focused on selectively decomposing specific components down to building blocks. 14,113,114 Recent studies indicate the potential of enabling collective upcycling of comingled plastics to carbon materials. 73,115,116 The feasibility of pyrolysis-based recycling of mixed plastics has been investigated, where studies suggest these processes can be cost-competitive while also reducing environmental impacts and energy use when compared to energy recovery recycling methods. 117,118 In a recent work, Qiang et al. collected single-use polyolefin wastes, prepared mixed polyolefin blends at various PE and PP compositions, and upcycled them to structured carbons (Fig. 5a). 53 Three-dimensionally structured materials were first prepared from mixed plastic wastes through fused filament fabrication (FFF) 3D-printing and converted to carbon through sulfonation-induced crosslinking at 150 1C and pyrolysis at 800 1C. Fig. 5b displays the complex structures attainable through this mixed waste-derived carbon monolith fabrication method. Mixed polyolefin blends with varied PE/PP compositions (20 : 80, 40 : 60, and 80 : 20 wt% PP : PE) were investigated, with blends requiring reduced crosslinking reaction times compared to homopolymers, while no impact on resulting carbons were observed. The accelerated crosslinking kinetics were facilitated by the lower melting temperature of PE (compared to PP) coupled with phase separation between polyolefin components promoting sulfuric acid diffusion. As discussed in previous sections, structured carbons can function as highly efficient Joule heating materials, which can then be implemented as catalyst supports for electrified thermo-chemical synthesis. When utilized for ammonia decomposition for hydrogen production, it was found that conversion increased significantly compared to a conventionally heated (convection heating) system (between 50-250% increase) as shown in Fig. 5c. The enhanced catalytic activity was attributed to potential electric field impacts, promoting associative desorption of nitrogen atoms as well as localized heating of catalytic nanoparticles. Importantly, life cycle analysis found that global warming impact and life cycle energy consumption associated with Joule-heated hydrogen production powered from renewable energy sources could be decreased by 400% and 200%, respectively, compared to current industrial practices (Fig. 5d and e). Moreover, Wang et al. compared homopolymer waste with two blend compositions consisting of 0.66 : 0.21 : 0.13 PP : LDPE : PS (mass ratio, blend 1) and 0.35 : 0.4 : 0.2 : 0.05 PP : LDPE : PS : PET (mass ratio, blend 2). 119 Following catalytic pyrolysis at 500 1C for 50 min with the presence of bimetallic catalysts (MgO-supported Co-Ni, Fe-Ni, and Co-Fe), mixed plastics exhibited comparable carbon yields (32.6 wt% for blend 1 and 31.2 wt% for blend 2) to bulk PP (30.2 wt%), LDPE (32.0 wt%), and PS (38.3 wt%) while having similar morphologies. Additionally, in-line catalytic decomposition of a mixed plastic waste blend (0.8 : 0.15 : 0.04 : 0.01 PE : PP : PC/PET : metal/wood, mass ratio) was carried out with a nickel-upgraded slug oxide (Ni-UGSO) catalyst at 700 1C with a feed rate of 0.33 g min À1 with a reaction duration of 2 h. 120 Mixed wastes were found to form carbon nanofilaments, with diameters ranging from 8-90 nm, though the carbon yield decreased by B10 wt% compared to a bulk HDPE feedstock (56 wt%). Additionally, the carbon nanofilaments were covered by a thin layer of amorphous carbon due to the presence of PET, PS, as well as other organic contaminants.

Fabrication of graphene from mixed plastic wastes has also been demonstrated through flash Joule heating (FJH) where mixed waste (0.4 : 0.2 : 0.2 : 0.1 : 0.08 : 0.02 HDPE : PP : PET : LDPE : PS : PVC, mass ratio) was ground with 5 wt% carbon black to obtain a conductive material. 121 To provide a brief background, FJH involves passing a direct current between two electrodes across a packed bed of precursor material. The resistance of the material allows the conversion of electrical current directly to heat, capable of achieving very high temperatures (B3000 1C) within milliseconds. With a carbon content of 81 wt%, the mixed waste blend formed 22 wt% high quality graphene following FJH, only slightly lower than bulk HDPE (27 wt%). Upcycling of mixed plastics (PP : LDPE : HDPE : PS : PET : PVC 0.24 : 0.3 : 0.26 : 0.13 : 0.04 : 0.03, mass ratio) to porous carbon sheets has been demonstrated through pressurized carbonization at 500 1C with magnesium oxide. 122 The majority of carbon growth was attributed to aromatic constituents, with the mixed waste-derived sheets displaying a 33.4 wt% carbon yield, a maximum surface area of 713 m 2 g À1 , and a pore volume of 5.27 cm 3 g À1 . Moreover, porous carbon nanosheets can be produced from mixed plastics through catalytic carbonization on organically-modified montmorillonite (OMMT) followed by KOH activation. 123 Specifically, a blend of 40 wt% PE, 35 wt% PP, 18 wt% PS, 3 wt% PET, and 3 wt% polyvinyl chloride (PVC) containing trace amounts of moisture and ash was blended with OMMT and carbonized to 700 1C for 10 min. The waste-derived carbon nanosheets had high surface areas (up to 2198 m 2 g À1 ) and pore volumes (up to 3.02 cm 3 g À1 ) with hierarchical porosity, containing micropores and mesopores (with an average pore size of 3.6 nm). These carbons were then utilized as supercapacitor electrodes, which can show high specific capacitances, such as 207 F g À1 in aqueous electrolytes and 120 F g À1 in organic electrolytes at a current density of 0.2 A g À1 . Furthermore, autogenic pressure pyrolysis of a plastic waste blend, containing 31.6 wt% PP, 31.4 wt% LDPE, 22.1 wt% HDPE, and 11.9 wt% PS, at 700 1C for 30 min with a reactor pressure up to 7 MPa can result in high CH 4 content pyrolysis gas and carbon microspheres (2-8 mm diameters) containing semicrystalline graphitic structures. 124 Noteworthily, this process resulted in negative carbon emissions (À449 kg CO 2 -eq. per ton of plastic waste), while having 294 kg CO 2 -eq. in carbon emissions associated with converting wastes into fuels and carbon nanomaterials. Moreover, carbon nanotubes have also been fabricated from mixed plastics (LDPE : PP : PS : PET 0.4 : 0.4 : 0.1 : 0.1, mass ratio) through pyrolysis between 600-800 1C and followed by chemical vapor deposition via the use of nickel-based catalysts. 125 Though carbon nanocages and MWCNTs could be manufactured, a lower carbon yield was observed at 800 1C for the waste blend (22 wt%) compared to PP (32 wt%) and LDPE (31 wt%) due to PET decreasing hydrocarbon gas reactivity. At lower pyrolysis temperatures (500 1C), carbon yield was much lower for the mixed waste blend (3 wt%) than PP (32 wt%) and LDPE (21 wt%). Altogether, these findings suggest that upcycling mixed wastes to value-added carbons with varied structures and material properties is attainable, while additional fundamental understanding into recycling strategies would be required to address the complexity from blend compositions in real waste systems.

Thermosets and engineering plastics

Thermoset polymers, with 65 million metric tons generated annually, account for a large fraction of total polymer production (B20%). 126 While permanent crosslinking typically imparts chemical resistance and dimensional stability, its presence makes thermosets particularly challenging to recycle. 127 With demand anticipated to increase in the coming decades and a broad application space, including aerospace, construction, automotive, and electronics, developing waste management strategies for thermoset materials remains a critical challenge and need. 20 While recycling methods like solvolysis and hydrogenolysis as well as the development of reversible networks have been extensively studied, 19 the vast extent of chemistries, material properties, and compositions within all thermosets and engineering plastics necessitate a robust upcycling approach. 127 It is worth discussing the role of polymer structure on the resulting carbon material formed. For the majority of thermoplastic waste precursors which exhibit minimal carbon yield without additional intermediate steps of chemical treatment, significant variation on final carbon yield and carbon architecture is observed for the same polymer feedstock, indicating these important characteristics are primarily dependent on the crosslinking and carbonization strategy (see Table 1). Though carbonization processes and the feedstock's elemental composition are key aspects to consider when upcycling to carbon materials, it should also be noted that initial macrostructures of plastic product could being retained (such as foamed morphology). When considering engineering plastics or thermosets that can be directly pyrolyzed to carbon materials via a one-step process, the polymeric backbone plays a key role in dictating carbon yield in addition to the carbonization method. For example, carbon yields of common precursors include B20 wt% for crosslinked PS, [128][129][130] B45% for polyfurfuryl alcohol, 131 and 40-65 wt% for polyacrylonitrile. 132 Several model thermosets and engineering plastics, with varied polymer backbones, degrees of crosslinking, and resulting carbon yields are discussed below.

Crosslinked polyurethane (PU) foams, due to their nitrogen content and interconnected porous structure, are a promising waste stream for upcycling to doped carbon materials with open porosity. 133 Wang et al. demonstrated the upcycling of PU foam waste to nitrogen-doped porous carbons and their implementation as polysulfide reservoirs for lithium-sulfur batteries. 134 Foam fragments, mixed with potassium carbonate (K 2 CO 3 ), were directly carbonized at 800 1C for 1 h to form carbons with an interconnected porous, sheet-like morphology and a high surface area (1315 m 2 g À1 ) and nitrogen content (7.2 at%). When employed as a sulfur host for preparing lithium-sulfur batteries, the device achieved high capacities (766 mA h g À1 ) at 0.1C after 100 cycles and high-rate performance (466 mA h g À1 ) at 5C after 100 cycles. Production of graphitic 1D materials from PU waste has also been reported through a FJH technique. 135 Ferrocene and iron(III) chloridebased catalysts were mixed with polymer feedstocks and underwent FJH (0.05-3 s) up to 2726 1C. Compared to HDPE-derived graphitic 1D materials, PU-derived carbons exhibited similar morphologies and properties while also retaining nitrogen heteroatoms in the framework (B1.4 at%). PU foam can be upcycled to porous carbons through autogenic atmosphere pyrolysis under ambient condition at 700 1C for 30 min with a 5 1C min À1 heating rate, followed by a step of KOH activation at 700 1C for 1 h. 136 Furthermore, upcycling of crosslinked PE foams into porous carbons for CO 2 sorbents has been achieved by crosslinking with fuming sulfuric acid, reaching high degrees of crosslinking (resulting in B50 wt% carbon yield) within 30 min. 137 Phenolic resins, a mature class of thermoset prepared from the condensation of phenol with formaldehyde under basic conditions, have long been used as carbon precursors due to their ability to reach relatively high char yields (30-55 wt%), and their processability prior to curing. [138][139][140] In general, phenolic resin is a recognized class of carbon precursor, with broad uses in electronics, coatings, and composites. 141 Upcycling of phenolic resin waste into activated carbon was demonstrated through a microwave-assisted carbonization method, 142 where raw waste was blended with either KOH or phosphoric acid (H 3 PO 4 ) prior to microwave irradiation (500 W) for 30 min. The KOH activated phenol waste-derived carbon was found to have a surface area of 924 m 2 g À1 and relatively low potassium content (0.21 wt%), whereas the H 3 PO 4 activated sample exhibited a lower surface area 272 m 2 g À1 and a high phosphorus content (6.1 wt%). Interestingly, activated carbon prepared through pyrolysis using an electric furnace exhibited much lower phosphorus content (1.2 wt%) while no change was observed in potassium content. This was attributed to the high dielectric constant and dielectric loss of the H 3 PO 4 activated sample leading to greater microwave irradiation catalyzed H 3 PO 4 interactions, and in turn the more rapid carbonization process can kinetically trap a greater content of heteroatom in the final carbon framework.

Additionally, polybenzoxazines represent another family of advanced thermoset resins, prepared through oxazine ring-opening polymerization, which can have fairly high char yields (35-75 wt%) due to their extensive hydrogen bonding throughout their backbone coupled with high aromatic contents. 140,143,144 Nebhani et al. investigated the fabrication of nitrogen-doped carbons for high-performance supercapacitors directly from biosourced polybenzoxazine thermosets. 145 Samples were first cured under air at 150-200 1C between 4-8 h followed by heating to 600 1C, with 30 min holds for every 100 1C interval. Polybenzoxazine-derived carbon was found to have nitrogen loading B5 at% and a specific capacitance of 700 F g À1 at a current density of 10 A g À1 . Moreover, poly(ether ether ketones) (PEEKs) are a class of engineering polymers utilized in high performance applications due to their high mechanical strength, chemical resistance, and thermal stability. 146 As these high performance polymers typically require high processing temperatures, conventional recycling methods are often ill-suited to address these engineering plastics. Though the upcycling of PEEK waste has not been directly studied, PEEK has been demonstrated as an efficient carbon precursor. For example, Liu et al. investigated the synthesis of microporous carbons by carbonizing PEEK under argon gas up to 900 1C with a 10 1C min À1 heating rate followed by changing atmosphere to steam or CO 2 gas for activation, 147 resulting in carbons with high surface areas (B3000 m 2 g À1 ) and pore volumes (1.7 cm 3 g À1 ). The PEEKderived carbons were then used for hydrogen storage devices where they displayed volumetric and gravimetric hydrogen storage capacities of 35 g L À1 and 5 wt%, respectively at 20 bar and À196 1C. In a similar approach, high performance supercapacitors with a capacitance of 237 F g À1 and excellent cycling stability after 6000 cycles (B93% retention) were developed from using two-layered films of PEEK-derived microporous carbons and reduced graphene oxide, acting as both a binder and active electrode material. 148 It was found that the layered structure, compared to a blended composite film, improved ion accessibility to the hydrophilic microporous carbons and established conductive paths through the graphene oxide network. Furthermore, carbon fiber-reinforced silicon carbide (C/C-SiC) can be fabricated through a liquid infiltration method with polyetherketoneketone (PEKK), 149 another high-performance thermoplastic, 150 and PEEK powders acting as carbon precursors. Though the carbon yield of PEKK (65 wt%) was found to be slightly higher than PEEK (54 wt%), primarily attributed to PEKK's greater crosslinking potential, both PEKK-and PEEK-derived C/C-SiC displayed comparable Young's modulus (40 GPa), strain to failure (40.55%), and mean flexural strength (4200 MPa). While the conversion of many more high-performance engineering plastics to functional carbon materials have been reported, [151][152][153][154] upcycling their waste streams and the underlying impacts of conversion processes on carbon yield and formation still need to be further investigated.

Polymer composites

Polymer composites, containing functional or reinforcing agents within the polymer matrix, are an essential material class across various industries, such as electronics, construction, and water remediation. 59,155 Significant efforts have been focused on enhancing performance and properties of various composite systems, often by optimizing filler orientation, distribution, and adhesion to the matrix; 156 although, the presence of multiple components make composite recycling quite challenging. 157,158 With increasing demand for polymer composites, these difficultto-recycle waste streams possess environmental concern, though they also hold economic potential stemming from their valuable and high-performance filler materials. 59 From carbon fiberreinforced polymer (CFRP) composites alone, it is expected that waste from end-of-life wind turbine blades and commercial aircraft will reach 483 000 tons and 500 000 tons by 2050, respectively. 159,160 The majority of polymer composite recycling methods primarily focus on extracting fillers (typical high value components) from the lower-value polymer matrices. For example, direct pyrolysis of polymer composites between 300-750 1C has been frequently shown to decompose matrices and allow for recycling of filler materials. [161][162][163] A drawback of this strategy is not only can the matrix material not be recycled, but recovered fillers often have reduced performance properties, due to carbon residues remaining on filler surfaces which require chemical decomposition or burnout under aerobic conditions. For instance, a carbon fiber-reinforced epoxy-amine network was recycled by exposing the composite to monoethanolamine (MEA) with 10 wt% KOH at 160 1C for 90 min. 164 In a binary alkaline solution, the epoxy-amine networks were degraded up to 98 wt% while the carbon fibers retained 93% of their initial tensile strength and elastic modulus. Additionally, carbon fiberreinforced polyester composites have been recycled by hydrolysis with 80% hydrazine hydrate with NaOH along with a Fenton reaction at 100 1C, where reclaimed fibers kept the morphology, chemical structure, and tensile strength of their virgin analogs. 165 Recycling of epoxy-based composites have also been investigated through several pathways, including thermal decomposition with molten KOH at B300 1C for 1 h, 166 phosphoric acid catalytic degradation at 250 1C for 1 h, 167 and treatment with peracetic acid at 65 1C for 4 h. 168 While these strategies result in filler extraction from polymer composites with retention of mechanical properties, the bulk of the composite is still not recycled. Aiming at simultaneously recovering carbon fiber and upcycling bismaleimide (BMI) resin from a composite material, Hou et al. developed a chemical degradation approach with acetic acid/calcium nitrate (Ca(NO 3 ) 2 catalytic system at 240 1C for 10 h. 169 Fig. 6a shows images of recycled carbon fiber and nitrogen-doped carbon derived from BMI degradation products. Carbon fiber extracted from BMI composites with Ca(NO 3 ) 3 /acetic acid (CF-3) was found to have comparable tensile strengths to virgin CF prior to treatment (CF) and following treatment with Ca(NO 3 ) 3 /acetic acid (CF-1) and Al(NO 3 ) 3 /acetic acid (CF-2) (Fig. 6b). While the virgin CF's tensile strength declined dramatically following treatment (from 12.42 cN dtex À1 to 7.42 cN dtex À1 ), the CF extracted from BMI composites had minimal decline (12.32 cN dtex À1 ) due to BMI resin acting as a barrier and reducing CF contact time with the catalytic system. BMI resin degradation products were then separated from carbon fibers and characterized through 13 C Nuclear Magnetic Resonance spectroscopy (NMR) as depicted in Fig. 6c, where the obtained oligomers' carbon skeletal structure maintained integrity. These results suggested the C-N bonds were cleaved from the catalytic system. Though the reclaimed oligomers lost CQC functionality, they retained nitrogen heteroatoms and could function as nitrogen-doped carbon precursors. BMI resin degradation products were then blended with potassium carbonate, and carbonized to 800 1C for 2 h. Subsequently, BMIderived porous carbons were used as catalysts for the aerobic oxidation of 5-hydroxymethylfurfural (HMF) to 2,5 diformylfuran (DFF), which displayed up to an 89 wt% yield at 120 1C following 4 h (Fig. 6d). Upcycling of carbon fiber-reinforced epoxy composites has also been demonstrated through plasma treatment, 170 through exposing graphene supported on an alumina fiber porous material to microwave irradiation. It was found that the high temperature generated from plasma converted the polymeric network to combustible gas while also improving the graphitization degree of reclaimed carbon fibers.

Outlook and conclusions

Though the upcycling of plastic waste into carbon materials has been studied for a few decades, its industrial implementation is relatively recent and still an ongoing effort. Looking forward, there remain several challenges that must be addressed to expand the utility of this approach and enable its large-scale application. A critical issue to address is the sequestration of toxic off-gasses during pyrolysis. While these byproducts can be used as energy sources, efficient capture and storage technologies would further enhance the efficacy of this strategy by increasing recycling and decarbonization impacts. Another major challenge is the necessity to enable energy efficiency in pyrolysis processes, as high temperature treatment is often involved. Specifically, similar to industrial chemical syntheses of commodity chemicals, including ammonia, hydrogen, and methanol, which require high reaction temperatures, pyrolysis of plastic waste can be a highly energy-intensive process that can result in large CO 2 emissions as well as costs associated with process heat. 171 Significant efforts have been focused on developing processes that are economically viable while simultaneously reducing emissions. Several life cycle analyses (LCA) and techno-economic assessments (TEA) have investigated the thermochemical recycling of plastic wastes into small molecules feedstocks and oils, where pyrolysis has been found to potentially be the most profitable strategy compared to gasification, hydrothermal liquefaction, and catalytic processes (hydrocracking and hydrogenolysis). 172 When considering CO 2 emissions, pyrolysis can be more environmentally friendly than a few processes (gasification and hydrothermal liquefaction) though other processes (such as hydrogenolysis and hydrocracking) could have further reduced environmental impacts. Key barriers that hinder pyrolysis upcyling include use of nitrogen gas, energy costs, and inadequate carbon capture technologies. Similar LCA and TEA investigations into the upcycling of PET waste into porous carbon CO 2 sorbents have found that these processes hold potential for financial viability and the possibility of carbon neutrality when scaled up to industrial levels. 101 As mentioned previously, mixed plastics wastes (PP:LDPE:HDPE:PS) were converted to a combination of pyrolysis gas and carbon microspheres through autogenic pressure pyrolysis. It was found that this process exhibited negative carbon emissions (À449 kg CO 2 -eq. per ton of plastic waste) that outweighted the carbon emissions from the pyrolysis process (294 kg CO 2 -eq.), 124 the majority of which stemming from energy losses. One strategy that circumvents this issue is the electrification of heating processes, which can be integrated with renewable energy sources to potentially produce a zero-carbon heating solution. As an example, a Jouleheated approach that utilizes a resistive heating element, can allow for the direct conversion of electricity to heat. 173,174 Through this efficient heating method, high reaction temperatures can be reached quickly, resulting in process intensification with smaller required reactor volumes, reduced environmental impacts, and lower process heat inputs. Another advantage of this electrified process is rapid reactor start-up and shutdown speeds, granting adaptability to fluctuating waste feedstock and electricity availability. 175,176 Life cycle analyses of plastic waste upcycling to graphene through FJH have indicated noticeable environmental improvements and reduction in energy usage compared to conventional recycling methods. 177 Microwaveassisted pyrolysis has also been investigated as an alternative for plastic waste upcycling method that can significantly reduce energy consumption and global warming impacts. 178 Furthermore, development of pyrolysis strategies that can readily incorporate a broad range of mixed plastics in addition to highthroughput and continuous material sorting, pretreatment, and feeding pathways will be critical. Upcycling methods for intricate waste streams like copolymers, including both random copolymers, such as acrylonitrile butadiene styrene (ABS), and block copolymers, such as styrenic thermoplastic elastomers, must also be developed due to their high production volumes as well as their valuable morphological features. As demonstrated with several foam waste feedstocks, structured wastes can be valuable as they can be converted to unique carbon materials, if macroscopic geometries can be retained. Additionally, nanostructured feedstocks, such as polystyrene-block-poly(ethylene-ran-butylene)block-polystyrene (SEBS), presents a distinct precursor that can directly generate carbon materials with mesoporous morphologies. 179,180 In conclusion, increased generation of plastic waste remains a global issue that threatens both the environment and biological life. To properly confront this grand challenge, significant technological development is required involving robust recycling strategies that can incorporate complex plastic waste feedstocks at scale. Moreover, to ensure economic viability and scaled implementation, recycled products must match, if not exceed, the market value of virgin materials (i.e. waste precursors). Upcycling of plastic waste into value-added materials has been demonstrated through several approaches, enabling the production of carbon materials with versatile nanostructures, including carbon nanotubes, graphene, and carbon fiber. This review summarizes the conversion of commodity polymers into carbon materials with tailored structures and properties using different processing methods, and offers a brief outlook on future technologies to enable industrial-scale plastic waste upcycling to carbon materials.