Introduction

Polyolefin thermoplastics have dominated many manufacturing sectors of our economy owing to their low cost, excellent physicochemical properties, and facile processability. [1] Polystyrene (PS), in particular, is produced 25 million tons annually and utilized for single-use plastic cutlery, packaging, synthetic rubber, and more. [2][3][4] However, PS natural degradation is one of the slowest, [5] resulting in accretion in landfills and water bodies, thus threatening the ecosystem. To confront the plastic crisis and to achieve a circular economy, [6][7][8][9] conventional large-scale recycling operations rely on mechanical recycling to repurpose and redeploy plastics. [10] However, the mechanical recycling of PS, especially expanded polystyrene (EPS), encounters difficulties such as low density, foam bulkiness, potential contamination, deteriorated quality, and low value of regenerated plastics. [3] Thus, PS is rarely recycled, at the lowest recycling rate of 0.9 % among commodity plastics, and several municipalities even consider banning the use of EPS. [11][12] Therefore, it is imperative to develop alternative strategies for recycling PS to achieve complete carbon circularity, especially considering that PS waste represents an excellent source of low-cost aromatics and phenyl makes up 74 % of the polymer's total weight.

Recently, chemical upcycling has attracted much interest due to its potential to increase product value and address the financial viability of the recycling industry. Through thermo-and photochemical strategies, plastics are converted to fuels, [13][14] adhesives, [15][16] fine chemicals, [17] and monomers. [18] With or without catalysts, [19][20][21][22][23][24] PS can be depolymerized to styrene, which can be used for repolymerization and copolymerization. But the monomer is of low value, restricting the scalability of the depolymerizationrepolymerization strategy. Alternatively, our group has reported a photochemical degradation-upcycling ("Deg-Up") strategy to extract aromatics, the most valuable part of PS, for downstream conversion to high-value chemicals. [25] Others have targeted oxygenated value-added aromatic compounds, including ketones, aldehydes, carboxylic acids, and nitrogen-containing arylamines. [26][27][28][29] Qin and co-workers applied Fe II -catalyzed nitrogenation of alkylarenes to cleave CÀC bonds in the PS backbone and generated Nnonylaniline. [27] Hu [30] and Stache's groups [31] explored Fecatalyzed hydrogen atom transfer (HAT) for oxidative cleavage of CÀC bonds in PS. In both works, benzoic acid was the main product, but the oxidative reactions yielded many by-products, such as acetophenone and benzaldehyde, giving poor selectivity. Similarly, Huang et al. developed a metal-free photodegradation pathway catalyzed by p-TsOH, producing low-cost formic acid and a mixture of benzoic acid and acetophenone. [32] Das's group reported another metal-free HAT photodegradation utilizing N-bromosuccinimide and sodium triflinate photocatalysts, [33] but the aerobic reaction conditions led to a partial loss of carbon as CO and CO 2 . Cao et al. employed graphitic carbon nitride to upcycle PS into a mixture of benzoic acid and other aromatic oxygenates at 150 °C, along with large quantities of CO 2 . [34] Most of the existing chemical recycling and upcycling methodologies are tailored to a specific single product (i.e., benzoic acid) and thus not versatile. Herein we report a versatile and general "Deg-Up" strategy to produce a wide range of high-value chemicals (Figure 1). The "Deg-Up" process consists of two cascade reactions: the first reaction selectively degrades polymers to a dominant intermediate using a variety of aromatic and non-aromatic solvents. The second reaction derivatizes the intermediate to high-value chemicals utilizing acyl, alkyl and sulfinyl chlorides. The key is that the intermediate must be a platform chemical easily adoptable by the existing chemical industry. For aromatic plastics such as PS, benzene is naturally an ideal intermediate because it makes up 74 % of the polymer's total weight and is the precursor to almost all aromatic compounds. This design broadens the product scope by simply adjusting the second reaction, which can be designed to proceed in the same reactor using the same catalysts. Following this concept, we upcycle PS to a wide range of aryl ketones and diphenyl sulfide that are important to the pharmaceutical and fragrance industries via a low-temperature process.

Results and Discussion

To obtain the benzene intermediate, we first tested the thermochemical degradation of commercial PS pellets (M n � 120 kDa) using AlCl 3 in deuterated benzene (Figure S1). The choice of C 6 D 6 is critical to allow for quantification of benzene yield though 1 H NMR analysis (Figures S2 and S3). The benzene yield increased with AlCl 3 loading, reaching � 97.6 % of the phenyl groups in PS when the mass of AlCl 3 amounted to that of PS waste (entry 6, Table S1). With AlCl 3 , the PS molecular weight dropped rapidly from 120 kDa to 1.5 kDa in 30 min (Figure 2d). Without AlCl 3 , however, PS did not appreciably degrade, and the molecular weight only decreased by � 20 kDa after refluxing in C 6 D 6 for 5 h. The slight decline is likely induced by the cleavage of alkene "weak links" commonly found in polyolefins. [35] Thermal heating substantially accelerated the degradation. An investigation into the degradation kinetics revealed a random-chain scission behavior, as supported by the linear relationship between the inverse molecular weight and reaction time (Figure 2e). The degradation rate constant (2.5 × 10 À2 s À1 ) is higher than other reported values for Lewis acid catalyzed PS degradation (Table S2). [36] Gas chromatography-mass spectrometry (GC-MS) revealed a small amount of side products in the liquid phase, including ethylbenzene, cumene, 1-methylindan, and diphenylmethane (DPM) (Figure S4). Interestingly, these alkylbenzenes were also present in the photochemical degradation of PS, [25] suggesting analogous fragmentation pathways. Additionally, cyclic alkylbenzenes such as tetralin and indanes were generated, likely from intramolecular alkylation of phenyl groups after the initial dephenylation (Scheme S1). [37] Nonetheless, the side products made up < 3 wt.% of the original PS phenyl content. The PS backbone decomposed to light products, mainly gaseous isobutane and highly volatile isopentane (Figure S5). After 5 h of degradation, no high-molecular-weight polymers remained, and most products were < 650 Da, as confirmed by SEC and high-resolution mass spectrometer (HR-MS) (Figures 2d and S6). The solid phase was extracted by warm benzene and characterized via GC-MS, showing polycyclic naphthalene, phenanthrene, and dihydropyrene derivatives (Figure S7). The solid by-products did not influence the subsequent upcycling, except the sulfinylation reaction. Therefore, separating polycyclic compounds from the de-graded mixture before upcycling was deemed unnecessary. As the degradation was carried out under air-free conditions, no oxygenates were detected in any phase of the degraded mixture. Overall, the thermochemical degradation of PS produced a mixture of hydrocarbons, with benzene as the dominant product.

To determine the efficacy of PS degradation in alternative organic solvents, we set up a series of gram-scale experiments in toluene, carbon disulfide, and hexanes. In all cases (even in a poor solvent of hexanes), benzene was the predominant product (Figures 2c and S8-S10). Not surprisingly, PS degradation in aromatic hydrocarbons (e.g., toluene and benzene) was the most effective and bolstered the dephenylation to high yields (entries 1 and 2, Table 1). Due to the poor solubility of PS in hexanes, degradation was performed in a Schlenk tube at � 108 °C and � 3 bar to accelerate the reaction, and the benzene yield was moderate (entry 3, Table 1). PS degradation in a good solvent of CS 2 at a lower temperature of � 85 °C afforded a comparable yield to that in hexanes at � 108 °C (entry 4, Table 1). Regardless, SEC chromatograms showed no detectable polymer peaks after degradation (Figures S9 and S10), illustrating that PS underwent effective chain cleavage in all tested solvents. Because benzene is the product from PS degradation, it becomes an excellent solvent for the process and removes the need for additional solvent-product separation.

To test the resilience of our method to impurities, a comparative study was conducted using real-world expanded polystyrene (EPS) in C 6 D 6 (Figure 2d). Remarkably, contaminants such as food particles did not substantially affect the selectivity towards benzene, eliminating the necessity to pretreat the EPS wastes before degradation. The benzene yield was still above 90 % for EPS, which is promising for large-scale recycling of municipal wastes.

To obtain a broad range of value-added chemicals (Table 2), PS "Deg-Ups" were conducted in a one-pot manner, using various acyl and alkyl chlorides for the upcycling reactions. Through Friedel-Crafts reactions, nine representative high-value chemicals for cosmetic and pharmaceutical applications [38][39][40][41][42] were obtained in moderate to high yields, as confirmed by GC-MS (Table 2, Figures S11 and S12). Especially, oxalyl chloride afforded benzophenone as the only product in accordance with literature, [43] corresponding to the incorporation of one carbonyl group ) Benzene yield (%) a AlCl 3 Pressure Temperature 1 d-benzene 1 g 1 bar 80 °C 5 97.6 2 toluene 1 g 1 bar 100 °C 5 70.5 3 hexane 1 g 3 bar 108 °C 5 51.9 4 CS 2 1 g 3 bar 85 °C 5 53.1 [a] NMR yield. Table 2: Upcycling of the intermediate product of benzene from PS (1 g) to nine representative high-value chemicals for various applications.

* All yields are determined using GC and an internal standard. ** For entry 1-9, an excess of 7.5 mmol AlCl3 was added to the upcycling reaction. For entry 10, the remaining AlCl 3 from the degradation step was sufficient for the upcycling reaction and thus no additional AlCl 3 was added.

(entry 7, Table 2 and Figure S11 product 6). Benzophenone is a large-volume chemical with a market price of � $15,000/ ton [25] and its application ranges from cosmetics to pharmaceuticals. [44] Therefore, the synthesis of benzophenone from PS waste presents an attractive alternative that does not rely on expensive diphenylmethane or extremely toxic phosgene. [45] Notably, the "Deg-Up" strategy can selectively provide multiple products via stepwise upcycling. We investigated the multi-step upcycling using difunctional 3-chloropropionyl chloride (CPC). CPC is interesting because it features both acyl and alkyl carbons and thus allows for controlled sequential upcycling due to the different electrophilicities of the two sites. At room temperature, Friedel-Crafts acylation of benzene took precedence because the interaction of acyl chloride and AlCl 3 often generates an acylium ion intermediate, a strong electrophile. [46] On the contrary, the alkylation by a primary alkyl chloride may not involve a free carbocation as active electrophile, and consequently necessitates a higher energy to activate the reaction. Indeed, the room temperature upcycling yielded 3chloropropiophenone exclusively in quantitative yield (entry 5, Table 2); no alkylation product was detected by GC-MS (Figure 3). Via an addition step of simply refluxing at 80 °C, 3-chloropropiophenone was further converted to dihydrochalcone (i.e., 3-phenylpropiophenone) over the same catalyst.

Sulfur-containing compounds, especially arylsulfides, exhibit strong biological activities and are important scaffolds for pharmaceutical products. [47][48][49] In particular, diphenyl sulfide is a precursor chemical to diphenyl sulfone, [50][51] a special high-temperature solvent for processing polyether ether ketone. [52] Typically, diphenyl sulfide is synthesized by reacting benzene and sulfur monochloride in the presence of aluminum chloride, but the reaction generates sulfur and thianthrene by-products. [53] Alternative approaches utilize thionyl chloride as a sulfur source for benzene sulfinylation. [54] It is widely understood that the sulfinylation goes through two electrophilic aromatic substitution (EAS) after the acid catalyst activates the SÀCl bond. [55] But the reaction has poor selectivity and typically yields a complex mixture of sulfides, sulfoxides, sulfones, etc. [54] Here we investigated the PS upcycling utilizing thionyl chloride (SOCl 2 ). Surprisingly, the "Deg-Up" of PS waste with thionyl chloride exhibited an unusually high selectivity for diphenyl sulfide (Figure 4a). Remarkably, even with sulfuryl chloride (SO 2 Cl 2 ), the PS upcycling furnished the same sulfide product, indicating that S(VI) was reduced to S(II) in the reaction.

To unveil the cause for high selectivity towards sulfides during upcycling, we set up a series of control experiments in the absence of PS. Starting from pure benzene, the reaction with SOCl 2 at 70 °C predominantly afforded diphenyl sulfoxide in a high yield of 98 % (Figure 4b, top Scheme and Figure S13); no sulfide product was observed in the GC and GC-MS (Figure 4d, blue chromatogram). Thus, we speculate that the side-products from PS degradation are responsible for the apparent selectivity shift from sulfoxides to sulfides. After surveying various additives (Figure 4b, bottom scheme) featuring the basic structural motifs of PS degradation by-products, phenanthrene and naphthalene (products 11 and 12) substantially changed the selectivity towards the sulfide. Although undetected in our products, anthracene also displayed a high selectivity towards the sulfide. Researchers have extensively investigated Scholl-type coupling of aromatic compounds in the presence of Lewis acids with or without electron acceptors. [56][57][58][59] Electron-rich arenes, such as polycyclic compounds, undergo dehydrogenative aryl-aryl coupling promoted by AlCl 3 and release hydrogen gas. [60] In this context, naphthalene and phenanthrene likely follow the Scholl reaction and provide the necessary reductant to reduce the high-valent sulfur to S(II). GC-MS spectra of the control reaction with naphthalene revealed two phenylnaphthalene isomers, likely generated by the Scholl coupling of naphthalene and benzene (Figure S14). Potentially, heavier dehydrogenation products, such as 1,1'-binaphthalene and perylene could be produced, [61] but they could not be detected by our GC due to the low vaporization potential (1, 1'-binaphthalene boiling point = 360 °C; predicted vapor pressure = 1.03 × 10 À11 MPa). Indeed, high molecular weight polyaromatic hydrocarbons (PAHs) were detected by high resolution Mass Spectrome-try (Figure S15) in a control reaction of phenanthrene in benzene catalyzed by AlCl 3 (Scheme S2). Despite the low concentration of PAHs, the Scholl coupling of phenanthrene led to cascading sp 2 CÀC bond formations, producing highmolecular-weight PAHs and sufficient hydrogen for reducing diphenyl sulfoxide to diphenyl sulfide. UV/Vis spectra of phenanthrene with and without AlCl 3 validated the presence of electrophilic arene-Lewis acid complexes, which could induce subsequent Scholl-type dehydrogenative aromatic coupling. [62] Normally, phenanthrene does not absorb visible light, but the phenanthrene-AlCl 3 complex exhibits significant absorbance in the blue-green region of the spectrum, displaying a yellow color (Figure S16). Thus, in the AlCl 3catalyzed sulfinylation reaction, PAHs reduced the sulfoxide and directed the selectivity towards the sulfide, likely through the in situ generation of a reducing agent of hydrogen (Scheme S3 and Figure S17). The degradation process tolerated contaminants, such as food and dirt and was scaled up to 50 grams in the laboratory (Figure S18 and S19). The design of benzene, instead of oxygenated aromatics, as a platform intermediate has several advantages. First, benzene is the precursor to almost all aromatic compounds, and it has the versatility to be adapted by a wide range of chemical industries. Second, phenyl groups constitute the majority of the carbon content within PS. The dephenylation reaction extracts the most valuable part of PS waste. Third, because benzene is an excellent solvent for PS, the reaction can be conducted in benzene, removing the need for separating products from the reaction solvent. It is crucial to note that the upcycling step only consumes the volume of benzene harvested from PS. For a large-scale process, the unused solvent can be recycled to dissolve more PS feed. Lastly, all carbon atoms in benzene are chemically equivalent, and thus the benzene intermediate ensures that no unwanted isomers are formed during upcycling.

Conclusion

Herein, we report an effective "Deg-Up" strategy to chemically recycle PS waste under mild temperatures. This thermochemical approach is more adaptable into existing industrial processes than the recently reported photochemical methods. The thermochemical process starts with PS dephenylation to generate a platform intermediate chemical of benzene, followed by derivatization of the intermediate to various aromatic compounds via Friedel-Crafts reactions. The one-pot process is designed with a great emphasis on versatility and generalizability to achieve a library of highvalue chemicals. As a proof of concept, diverse acyl/alkyl chlorides and sulfinyl chlorides are tested, which affords the synthesis of 9 ketones and sulfide. Importantly, temperature controlled chemoselective upcycling with difunctional additives enables the synthesis of high-value chemicals in a stepwise fashion. Furthermore, we have shown that the presence of polyaromatic compounds controls the selectivity of CÀS bond formation reactions, which represents a new approach to synthesizing symmetric aryl sulfides without the need for sophisticated catalysts. Overall, our "Deg-Up" method is energy efficient and scalable. The conversion of PS waste into value-added chemicals of diverse functions opens new chemical feedstock to the cosmetic and pharmaceutical industries. The strategy of cascading reactions can be adopted by the chemical recycling and upcycling of other plastic wastes.

Supporting Information includes supplementary discussion on dephenylation and dephenylation mechanism, hydrodemethylation of toluene under the degradation conditions, comparative experiments on benzene recovery from real-world EPS waste vs. clean PS pellets, characterization of solid, liquid, and gaseous side products of degradation, characterization of upcycled products, and future perspectives.