Complementary acid site mechanisms in hydrogen-free polyethylene upcycling: elucidating the distinct roles of Brønsted and Lewis sites in Ce-modified zeolites

Abstract

The environmental burden of petrochemical-derived plastics, particularly polyolefins such as polyethylene and polypropylene, has spurred the search for sustainable upcycling strategies. Conventional hydrogenolysis and hydrocracking processes rely on external H₂, over 98% of which is produced via steam methane reforming or coal gasification; both methods yield significant CO₂ emissions. In this study, we demonstrate a hydrogen-free approach to PE upcycling using zeolite Y ion-exchanged with various cations (Na⁺, Li⁺, K⁺, H⁺, La³⁺, and Ce³⁺). Among these, the Ce-exchanged mesoporous zeolite (Ce_meso_Y) achieved complete PE conversion with an 88.7% yield of naphtha-range hydrocarbons (C₅–C₁₂). NH₃-TPD and pyridine-DRIFTS analyses revealed that Brønsted acid sites (BASs) drive C–C bond cleavage, while strong Lewis acid sites (LASs) promote intramolecular hydrogen transfer from the polymer backbone, thereby eliminating the need for external H₂. Extending this approach to post-consumer polyolefin waste (including HDPE bottles, LDPE film, and PP cases) delivered 70.5–82.6% conversion and 77.8–84.8% naphtha selectivity. Our findings establish a sustainable, hydrogen-free route for plastic upcycling by harnessing intrinsic polymer hydrogen and fine-tuning acid site functionality.

---

Insoo Ro

Professor Insoo Ro earned his B.S. magna cum laude in Chemical Engineering from Rice University in 2012 and his Ph.D. from the University of Wisconsin–Madison in 2017 under Professors James Dumesic and George Huber. After postdoctoral research at the University of California–Santa Barbara with Professor Phillip Christopher, he joined Seoul National University of Science and Technology in 2020 and is now an Associate Professor. He has published over 30 papers in journals such as Nature, Nat. Commun., and Angew. Chem., and received the Science & Technology Outstanding Paper Award from the Korean Federation of Science and Technology Societies.

---

Green foundation

1. Our work addresses a previously overlooked role of Lewis acid sites in polyethylene upcycling. While past research has largely focused on Brønsted acid sites for C–C bond cleavage, our study demonstrates that Lewis acid sites can enhance hydrogen transfer under hydrogen-free conditions. This discovery enables a sustainable catalytic system that directly converts polyethylene into naphtha without requiring fossil–derived hydrogen.

2. We achieved an 88.7% naphtha yield from polyethylene without external hydrogen or noble metals. The catalyst is highly reusable and processes diverse post-consumer plastics—LDPE film, HDPE bottles, and PP cases—demonstrating a promising and greener alternative for the plastic recycling industry.

3. The clarified roles of Brønsted and strong Lewis acid sites in our system offer a rational design framework that can be applied to a wider range of plastics. This could enable hydrogen-free upcycling of polymers beyond conventional polyolefins, including those that have traditionally shown poor reactivity without added hydrogen, by optimizing acid site distributions for specific substrate characteristics.

Introduction

The widespread consumption of petrochemical-derived plastics has become a pressing global environmental concern. According to Plastics Europe, global plastic production reached 413.8 million tonnes in 2023, with 90.4% derived from fossil fuels.¹ Despite this scale, an OECD report indicates only 14% of plastic waste undergoes recycling, while the remainder accumulates in landfills (50%), is incinerated (17%), or is lost to the environment (19%).² Polyolefin-based materials—particularly polyethylene (PE) and polypropylene (PP)—constitute 45.2% of plastic production and dominate the packaging sector with their notably short lifecycle.^1,2 The environmental challenge is intensified by the inherent resistance of these plastics to natural degradation due to the high bond energy of carbon–carbon (C–C) bonds (348 kJ mol⁻¹).^3,4 To address this issue, researchers are actively developing advanced depolymerization approaches, including pyrolysis and catalytic upcycling, for sustainable polyolefin waste management.^5–8

Traditional pyrolysis requires extremely high temperatures (500 °C to 1300 °C) to overcome the high activation energy of C–C bonds, rendering the process both energy-intensive and non-selective for valuable products.^9–11 In contrast, catalytic methods—particularly hydrogenolysis and hydrocracking—can depolymerize PE under milder conditions (200–300 °C).^6,12,13 However, these methods often rely on expensive noble metal catalysts, such as Pt or Ru, which hampers their scalability and economic feasibility.^14–26 While recent advances have explored more cost-effective alternatives like Zr- or Ni-based catalysts with promising activity,^3,27,28 these approaches still face a significant limitation: their reliance on molecular hydrogen. With global hydrogen production heavily dependent on fossil fuels, the associated CO₂ emissions make hydrogen-driven processes less sustainable.²⁹

To overcome the dependence on external hydrogen in PE upcycling, researchers have investigated alternative hydrogen sources, including methanol and tetralin. Our group previously demonstrated a tandem approach combining PE hydrogenolysis with methanol aqueous-phase reforming (APR).³⁰ While methanol APR successfully generates hydrogen in situ for PE hydrogenolysis, it produces CO₂ as a by-product, diminishing its overall environmental benefits.³⁰ Similarly, when others employed tetralin as a hydrogen donor, the resulting dehydrogenation product, naphthalene, poisoned catalytic active sites, reducing performance compared to systems utilizing molecular hydrogen.³¹ These limitations have prompted a strategic shift toward utilizing the intrinsic hydrogen contained within PE molecules themselves. While novel catalytic systems employing Pt/γ-Al₂O₃³² and Ru/HZSM-5³³ have successfully converted waste plastics into valuable products without requiring external hydrogen, their relatively high noble metal content (up to 6.9%) and low production rates limit industrial implementation. In response, acid catalysts such as zeolites have gained attention for their ability to crack PE without noble metals. The acidity characteristics of zeolites—specifically Brønsted acid sites (BASs) and Lewis acid sites (LASs)—play decisive roles in determining catalytic performance. BASs function as proton donors that facilitate C–C bond cleavage, thereby producing smaller hydrocarbons and light aromatics.^34,35 Research has demonstrated that increasing BAS concentration enhances catalytic activity and improves liquid product yield in ZSM-5 catalysts,³⁶ as these sites significantly lower the activation energy barrier for C–C bond cleavage compared to LASs.³⁷ While LASs facilitate paraffin dehydrogenation and olefinic intermediates formation, their direct contribution to C–C bond cleavage remains debated.³⁸ Notably, PE cracking activity per external surface area shows a nearly linear correlation with BAS density but exhibits inconsistent patterns with LAS density,³⁹ highlighting the need for more detailed mechanistic investigation.

In this study, we investigated the catalytic activity of zeolite Y ion-exchanged with monovalent (Na⁺, Li⁺, K⁺), protonic (H⁺), and trivalent (La³⁺, Ce³⁺) cations for hydrogen-free PE cracking. Monovalent-exchanged zeolites exhibited low reactivity (16.6–23.4%), while H_Y reached 56.4%, and La_meso_Y achieved 64.8%. Remarkably, Ce_meso_Y delivered 100% conversion and an 88.7% yield of naphtha (C₅–C₁₂). Through comprehensive characterization techniques including Brunauer–Emmett–Teller (BET), Ammonia-Temperature Programmed Desorption (NH₃-TPD), Pyridine-Diffuse Reflectance Infrared Fourier Transform Spectroscopy (pyridine-DRIFTS), and X-ray Photoelectron Spectroscopy (XPS), we correlate catalyst properties with reactivity. Our results indicate that BASs are crucial for C–C bond cleavage, while strong LASs, facilitate effective hydrogen supply, contributing to the superior reactivity of Ce_meso_Y. Furthermore, comparison with catalysts synthesized via impregnation revealed that only ion-exchanged catalysts retain strong LASs, resulting in enhanced performance. Ce_meso_Y demonstrates excellent reusability and achieves high conversion (70.5–85.2%) across various post-consumer plastics, including HDPE bottles, commercial LDPE, and PP file cases. This study provides valuable insights into the roles of BAS and LASs in hydrogen-free PE cracking and establishes a foundation for sustainable plastic upcycling.

Experiments

Chemicals and materials

Polyethylene (M_n: ∼1700, M_w: ∼4000, cat. no. 427772, lot: MKCT5408), polypropylene (M_n: ∼67 000, M_w: ∼250 000, cat. no. 427888, lot: MKCQ4298), pyridine (≥99.9%, cat. no. 270407, lot: SHBN7324), alkane standard solution C₈–C₂₀ (cat. no. 04070, lot: BCCD8495), BTEX solution (a mixture of benzene, toluene, ethylbenzene, o-, m-, p-xylene in methanol; cat. no. CRM47993, lot: LRAD1192), ethylcyclohexane (≥99%, code: E19154, lot: SHBQ3496), n-Octadecane (99%, code: O652, lot: MKCM5285), lithium nitrate (cat. no. 227986, lot: BCCC9191), potassium nitrate (99.99%, cat. no. 204110, lot: 0000067250), ethylenediamine-tetraacetic acid (cat. no. E9884, lot: BCCJ0200), and γ-Al₂O₃(cat. no. 544833, lot: BCBK5286 V) were purchased from Sigma-Aldrich. H-form zeolite Y (SiO₂/Al₂O₃ ratio = 5.1, cat. no. 45866, lot: X12H041), H-form zeolite Y (SiO₂/Al₂O₃ ratio = 80, cat. no. 45872, lot: Q26H045), Na-form zeolite Y (SiO₂/Al₂O₃ ratio = 5.1, cat. no. 45862, lot: R19I004), n-hexane (99%, cat. no. L09938, lot: 10226628), toluene (99.7%, cat. no. 022903, lot: R30J756), n-heptane (99%, cat. no. 19894, lot: 10229122), n-octane (≥98%, cat. no. A13181, lot: 10230530), n-nonane (99%, cat. no. A16177, lot: 10228094), n-docosane (99%, cat. no. A18050, lot: 10203774), lanthanum(III) nitrate hexahydrate (99.9%, cat. no. 012915, lot: N07I056), and 2,6-di-tert-butylpyridine (≥97%, cat. no. H55761, lot: Q19H078) were purchased from Alfa Aesar. n-Pentane (≥99%, cat. no. 16787, lot: A0405284), n-dodecane (99%, cat. no. 43459, lot: A0445744), and cerium(III) nitrate hexahydrate (99.5%, cat. no. 21869, lot: A0459009) were purchased from Acros Organics. Benzene (99.5%, cat. no. B0020, lot: IOAEK-PR), methylcyclopentane (96.0%, cat. no. M0203, lot: RD8QA-MI), ethylbenzene (99.0%, cat. no. E0064, lot: T2AXM-DR), and 1,3,5-tri-tert-butyl-benzene (98.0%, cat. no. T1633, lot: E654I-SE) were purchased from Tokyo Chemical Industry (TCI). Cyclohexane (99.9%, cat. no. 2606-7100, lot: C3312UC1) and Sodium hydroxide (98%, cat. no. 7571-4100, lot: S3370UG1) were purchased from Daejung Chemical. Methylcyclohexane (≥99%, cat. no. 74180-0430, lot: 2021E2178) was purchased from Junsei Chemical. Methylene chloride (99.9%, cat. no. 000D1602, lot: 092324) was purchased from Samchun Chemical. Commercial low-density polyethylene was purchased from Hanwha Total (product code: 530G), a high-density polyethylene (HDPE) bottle was obtained from Seoul Milk, and a polypropylene (PP) file case was obtained from Office DEPOT.

Catalysis preparation

Na_meso_Y (SiO₂/Al₂O₃ ratio = 5.1) was prepared by treating Na_micro_Y.⁴⁰ Typically, 6.7 g of Na_micro_Y was dispersed in 100 mL of an aqueous ethylenediamine-tetraacetic acid (EDTA) solution (0.07 mol L⁻¹) in a flask. The mixture was stirred at 500 rpm and refluxed at 100 °C for 6 hours. The resulting solid product was then filtered using filter paper (F2142 Grade, cat. no. F2142-110, batch: da0042A) and dried in an oven at 100 °C overnight. For the subsequent NaOH treatment, 1.7 g of the dried solid was added to 50 mL of an aqueous NaOH solution (0.4 mol L⁻¹) and stirred at 65 °C for 30 min. The formed Na_meso_Y was then filtered and dried. To functionalize Na_meso_Y, ion exchange was performed using 100 mL of nitrate solutions (0.1–0.3 mol L⁻¹, typically 0.2 mol L⁻¹) of Ce³⁺, La³⁺, Li⁺, and K⁺, respectively. In each case, 1.0 g of Na_meso_Y was treated at 80 °C for 12 hours. The resulting solids were filtered and subsequently calcined in air at 550 °C for 6 hours. The obtained materials are denoted as X_meso_Y–Z, where X represents the exchanged cation (Li, K, La, or Ce), and Z refers to the molar concentration of the precursor solutions. For the representative case where Z = 0.2, the value is omitted for simplicity.

The elemental composition of the prepared samples was determined using inductively coupled plasma optical emission spectroscopy (ICP-OES, Agilent 5110(SVDV)). Approximately 10 mg of each sample was dissolved in a mixture of HNO₃ (69%, 4 mL) and HF (45%, 2 mL), and the resulting solution was analyzed to quantify the metal content. The results are summarized in Table S1.†

For comparison with the ion exchange method, Ce/HY-WI catalysts were synthesized using the wet impregnation (WI) method. To avoid interference from metal ions in the Na-form, the catalysts were prepared using HY support without further treatment. The Ce loading was adjusted to match the content determined by ICP-OES analysis. In a typical synthesis, 0.56 g of HY support and 0.22 g of Ce(NO₃)₃·6H₂O were dispersed in 100 mL of deionized water. The suspension was stirred at 600 rpm for 2 hours to ensure thorough mixing, then heated to 50 °C until completely dried. The resulting solid was further dried at 100 °C for 12 hours and subsequently calcined under an airflow of 80 mL min⁻¹ at 400 °C for 3 hours, with a ramping rate of 2.5 °C min⁻¹.

Reaction procedure

A glass-coated magnetic stir bar, catalyst, and PE were added to a glass liner, which was then placed inside a 50 mL 316L stainless-steel high-pressure batch reactor fitted with a spiral wound gasket (Hanwoul Engineering CO., Ltd). Reactions were typically performed using 0.1 g of catalyst and 0.2–0.3 g of PE without any added solvent. To promote homogeneous mixing, the solid components were stirred on a magnetic stir plate at 300 rpm for 5 minutes prior to heating. The reactor was purged three times with Ar (30 bar) before being charged with argon at ambient pressure. It was then placed in a heater and gradually heated to the reaction temperature (260–320 °C) over approximately 1 hour. Specifically, the heating profile was as follows: ramp to 210 °C over 23 min, to 293 °C over 14 min, and finally to 300 °C over another 14 min, avoiding overshooting. The reaction temperature was monitored using a thermocouple inserted into the reactor. Once the temperature reached 150 °C, stirring at 500 rpm was initiated and maintained for the designated reaction time. A stirring rate of 500 rpm was selected to minimize mass transfer limitations, as confirmed by control experiments with our most active catalyst. This condition ensures that observed catalytic activity reflects intrinsic reactivity rather than diffusion constraints. After completion, the reactor was rapidly cooled to room temperature using an ice bath.

For gas analysis, since the gas product pressure was below 1 bar, additional argon was introduced to bring the pressure to 10 bar, and the gas was transferred to a 2 L PVC gas bag for analysis. For liquid analysis, 1,3,5-tri-tert-butylbenzene (about 7 mmol L⁻¹) was used as an internal standard, and methylene chloride was employed as a washing solvent to collect the liquid phase. The resulting solution was transferred to a centrifuge tube and centrifuged at 10 000 rpm for 15 min to separate the solid components (catalyst and unreacted PE) from the liquid products. As prolonged storage of methylene chloride in centrifuge tubes can cause degradation, the liquid was immediately transferred to a glass vial. The unreacted PE and catalyst were separated and subsequently dried in a vacuum oven at 80 °C for 48 hours. The dried centrifuge tube and glass vial—containing the reaction products and catalyst—were weighed, and the weights of the initial empty containers and the added catalyst were subtracted to determine the exact amount of unreacted reactant. To ensure accuracy, the weight loss of 1,3,5-tri-tert-butylbenzene during the drying process was confirmed through a controlled experiment and excluded from the weight calculations (Table S2†).

Reaction products were analyzed using a gas chromatograph (Agilent 8890 GC System) equipped with a flame ionization detector (FID) and a thermal conductivity detector (TCD). A GS-Carbon PLOT (Agilent) column was used for gas-phase analysis, while an HP-1 column (Agilent) was used for liquid-phase analysis. Gas-phase analysis was performed with a temperature ramping rate of 20 °C min⁻¹ from 38 to 325 °C, followed by a 10 min hold. The retention times of C₁–C₆ hydrocarbons were identified, and the FID area signals were calibrated using a reference gas mixture (CH₄, C₂H₄, C₂H₆, C₃H₈, and n-C₄H₁₀ in Ar) from Union Gas. The TCD area signals were calibrated using high-purity hydrogen (99.999%) from Union Gas. Liquid-phase analysis was conducted using the Agilent 8890 GC System with a temperature ramping rate of 5 °C min⁻¹ from 50 to 300 °C. Initially, the reaction products were identified by a gas chromatograph-time of flight mass spectrometry (GC/TOF-MS) system (Pegasus BT 4D, LECO) equipped with an RXI-5MS column (Restek), and the compounds with the highest similarity were calibrated using GC-FID. The corresponding retention times of hydrocarbons and FID area signals for the identified hydrocarbons are summarized in Table S3.† Conversion, yield, and H/C ratio were calculated as follows:

To evaluate the catalyst reusability, the spent catalyst was calcined at 550 °C under an air flow of 80 mL min⁻¹. To compensate for the approximately 5% catalyst loss during recovery, as reported in previous work,²² the reaction was repeated multiple times to obtain a sufficient quantity of catalyst for subsequent experiments.

For acid site poisoning, 680 mg of pyridine was added dropwise via pipette to 100 mg of Ce_meso_Y, corresponding to 40 times the estimated acid site density as determined by NH₃-TPD.²⁴ The mixture was dried under vacuum in a desiccator at room temperature for 12 hours and used immediately in the reaction.

Characterization

Nitrogen adsorption–desorption isotherm measurements were performed using a BELSORP Mini II (BEL Japan, Inc.) analyzer. Prior to the measurement, the sample was degassed at 100 °C for 1 h and subsequently at 300 °C for 6 h to eliminate the moisture. The pore volume was obtained using the single-point analysis (P/P₀ = 0.995), and the BET surface area was evaluated using the standard multipoint method at P/P₀ = 0.05–0.3.

Ammonia-temperature programmed desorption (NH₃-TPD) spectra were obtained using a BELCAT II (Microtrac MRB) instrument to measure the acidity of the zeolites. Typically, 0.1 g of the zeolite sample was pretreated in a helium flow (50 mL min⁻¹) at 300 °C for 1 h, then cooled to 50 °C. The adsorption of NH₃ was performed at 50 °C for 1 h by flowing a 10% NH₃/He mixture (50 mL min⁻¹) at 50 °C. Then, the zeolite samples were flushed with flowing helium (50 mL min⁻¹) for 1 h to remove the physically adsorbed NH₃. Desorption was measured using a TCD while heating from 50 °C to 600 °C at a ramp rate of 10 °C min⁻¹. Acid site density was calculated based on the area of the desorption peak.

The Brønsted to Lewis acid site density ratio (BAS/LAS) of the zeolite sample was evaluated by pyridine adsorbed Infrared (py-IR) spectroscopy. Spectra of pyridine adsorption were acquired using a Nicolet iS20 FTIR spectrometer equipped with a Hg/Cd/telluride detector cooled using liquid nitrogen. Prepared catalysts were loaded onto a Harrick Praying Mantis reaction chamber with ZnSe windows. Before measurements, the zeolite sample were pretreated in an argon flow (50 mL min⁻¹) at 450 °C for 1 h, then cooled to 50 °C A baseline spectrum was recorded before introducing pyridine. Subsequently, the zeolite samples were exposed to pyridine vapor in argon for 30 min, generated by passing argon flow through a bubbler filled with liquid pyridine.¹⁶ Physically adsorbed pyridine was then removed using pure argon for 30 min. All spectra were obtained using 128 scans with a resolution of 2 cm⁻¹. The samples were outgassed by heating at 450 °C (ramp 5 °C min⁻¹). After which the spectra were recorded. The BAS/LAS was estimated by using the relative surface areas of the bands around 1540 cm⁻¹ (BASs) and around 1450 cm⁻¹ (LASs) bands.

Scanning transmission electron microscopy (STEM) combined with energy dispersive X-ray spectroscopy (EDS) was performed using a JEOL NEOARM microscope, equipped with an SDD-EDS detector and a Cs probe aberration corrector. The instrument operated at 200 kV, achieving a spatial resolution of less than 0.1 nm.

X-ray photoelectron spectroscopy (XPS) analysis was carried out with a Thermo Scientific NEXSA instrument, utilizing monochromatic Al Kα radiation as the X-ray source. XPS spectra were collected with scan parameters set to five scans and a dwell time of 50 ms.

Thermogravimetric analysis (TGA) was conducted using a Shimadzu DTG-60H to assess the coke content in spent catalysts. Prior to measurement, the catalysts were preheated at 200 °C for one hour under a nitrogen flow of 50 mL min⁻¹ to eliminate physically adsorbed water. The samples were then heated to 800 °C (ramp 10 °C min⁻¹) in an alumina cell under an air flow of 50 mL min⁻¹.

For experiments using deuterium-exchanged n-hexane (hexane-d14), gas-phase products were analyzed by mass spectrometry (HPR-20 R&D, Hiden Analytical Ltd). The mass spectrometer operated under a secondary vacuum (10⁻⁶–10⁻⁵ mbar) using a turbomolecular pump, and the QIC (quartz inert capillary) was maintained at 200 °C to prevent condensation of water vapor.

Results and discussion

The catalytic activity was assessed using various exchanged ions (Na⁺, Li⁺, K⁺, H⁺, La³⁺, and Ce³⁺). The results showed that monovalent metal cations (Na⁺, Li⁺, K⁺) exhibited relatively low PE conversions of 16.6–23.4%, while the protonated form (H_Y) achieved a moderate conversion of 56.4% (Fig. 1a). In contrast, zeolite exchanged with trivalent metal cations (La³⁺, Ce³⁺) displayed significantly higher conversions of 64.8–100%. Moreover, the product distribution varied with the type of exchanged ion (Fig. S1†); in particular, the Ce-exchanged mesoporous zeolite (Ce_meso_Y) delivered an exceptionally high naphtha (C₅–C₁₂) yield of 88.7%, outperforming the other ion-exchanged zeolites (Fig. 1b).

Fig. 1 PE catalytic cracking results and catalytic characterization results of different ion-exchanged zeolites. (a) PE conversion as a function of exchanged ion type, (b) product distribution of PE catalytic cracking over Ce_meso_Y. Reaction conditions: 300 °C, 4 h, atmospheric Ar, 0.1 g catalyst, and 0.2 g PE, (c and d) N₂ adsorption-desorption isotherms and BJH pore size distribution plots of zeolites, and (e) NH₃-TPD profiles of various ion-exchanged zeolites.

To decipher the factors governing these differences in reactivity, we first characterized the textural properties of the catalysts. BET analysis confirmed that all mesoporous zeolites possessed similar mesoporosity, regardless of the exchanged ion (Table 1 and Fig. 1c, d),⁴⁰ thereby ruling out pore size differences as the primary factor. Our focus then shifted to the acid site properties, which are known to influence catalytic performance.³⁶ NH₃-TPD analysis (Fig. 1e and Table 1) revealed that monovalent cation-exchanged zeolites (Na⁺, Li⁺, K⁺) predominantly exhibit weak acid sites, whereas H- and trivalent cation-exchanged zeolites (La³⁺, Ce³⁺) feature a mixture of medium and strong acid sites. Weak acid sites primarily facilitate initial adsorption but are less effective at protonating substrates.⁴¹ In contrast, strong acid sites are crucial for substrate protonation, formation of carbenium ions, β-scission, and the skeletal isomerization process essential for producing short-chain alkanes, such as naphtha.⁴¹ Although this trend explains the superior performance of H- and trivalent metal cations-exchanged zeolites over those exchanged with monovalent ions, it does not fully account for the superior performance of Ce_meso_Y compare to both H- and La-exchanged zeolites (Fig. 1a). This discrepancy arises from the inherent limitation of NH₃-TPD analysis, which cannot distinguish between Brønsted and Lewis acid sites (BASs and LASs), both of which play distinct roles in PE cracking.^24,42

Table 1 Catalytic characterization results of different ion-exchanged zeolites

Catalyst	Surface area^a (m^2 g^−1)	Pore volume^b (cm^3 g^−1)	Acid site density^c (%)			Total acid site density^c (mmol NH3 g^−1)
			Weak (50–250 °C)	Medium (250–400 °C)	Strong (400–600 °C)
a BET surface area evaluated using the standard multipoint method at P/P0 = 0.05–0.3. b Total pore volume at P/P0 = 0.995. c Acid site density determined by NH3-TPD analysis.
Na_micro_Y	649	0.344	80.5	19.5	0.0	1.75
Na_meso_Y	616	0.442	83.4	16.6	0.0	2.04
Li_meso_Y	651	0.472	78.2	21.8	0.0	2.02
K_meso_Y	599	0.436	92.5	7.5	0.0	1.39
H_Y	536	0.339	57.0	31.1	11.8	1.77
La_meso_Y	492	0.383	52.4	34.2	13.4	0.99
Ce_meso_Y	514	0.400	52.2	34.1	13.7	1.90
Ce/HY-WI	455	0.300	51.2	30.5	18.3	1.92
Regenerated Ce_meso_Y	458	0.371	62.6	27.4	10.1	1.93

---

To elucidate the acid site functionality in PE cracking, we performed pyridine-DRIFTS analysis to distinguish the roles of BAS and LAS (Fig. 2 and Table 2). In the PE catalytic cracking reaction, BASs act as proton donors to generate carbenium ions—key intermediates in PE depolymerization.^35,36,43 These ions undergo skeletal isomerization to stabilize intermediates, facilitating iso-alkane production, followed by β-scission to cleave C–C bonds into smaller hydrocarbons.^35,36,43 In contrast, LASs enhance hydride transfer and activate inert C–H bonds in PE,⁴¹ with strong LAS (>450 °C) being particularly vital for hydrogen transfer in the self-supplied hydrogen mechanism under hydrogen-free conditions.⁴¹ Notably, monovalent metal-containing catalysts exhibited no detectable BAS at any temperature (Fig. 2a–d). The absence of BAS—key drivers of C–C bond cleavage³⁶—explains their low reactivity. For H_Y, NH₃-TPD and pyridine-DRIFTS revealed an inverse relationship between BASs and LASs: BASs increased while LASs diminished with rising temperature (Fig. 2e), resulting in an imbalance that inhibits effective hydrogen transfer. For La_meso_Y, BASs remained relatively stable with increasing temperature, but LASs were absent, similar to the behaviour of H_Y (Fig. 2f). Despite the presence of sufficient BASs, the lack of strong LASs limits the conversion of the PE catalytic cracking process under hydrogen-free conditions, as strong LASs (>450 °C) are essential for hydrogen-free operation.⁴¹ In contrast, Ce_meso_Y demonstrated a higher concentration of strong LASs (Fig. 2g). The introduction of Ce³⁺ ions into the Y zeolite appears to promote the formation of non-framework or defective aluminium species, which in turn generates strong LASs.⁴⁴ These findings suggest that while BASs are crucial for producing short-chain alkanes, strong LASs are essential for achieving higher overall conversion under hydrogen-free conditions. The dual presence of sufficient BASs and strong LASs in Ce_meso_Y underpins its enhanced catalytic performance. To clarify the unexpected absence of BASs in H_Y and Ce/HY-WI at low temperatures, additional pyridine-DRIFTS analyses were conducted after calcination. These confirmed the presence of BASs and indicated minimal change in catalytic activity (Fig. S2†).

Fig. 2 Pyridine-DRIFTS spectra of various ion-exchanged zeolites and Ce/HY-WI catalysts at different desorption temperatures. (a) Na_micro_Y, (b) Na_meso_Y, (c) Li_meso_Y, (d) K_meso_Y, (e) H_Y, (f) La_meso_Y, (g) Ce_meso_Y, and (h) Ce/HY-WI.

Table 2 Pyridine-DRIFTS analysis results of Brønsted and Lewis acid sites at different desorption temperatures

Catalyst	At 50 °C			At 300 °C			At 450 °C
	BAS (%)	LAS (%)	BAS/LAS	BAS (%)	LAS (%)	BAS/LAS	BAS (%)	LAS (%)	BAS/LAS
Na_micro_Y	0.0	100	0.0	0.0	100	0.0	0.0	100	0.0
Na_meso_Y	0.0	100	0.0	0.0	100	0.0	0.0	100	0.0
Li_meso_Y	0.0	100	0.0	0.0	100	0.0	0.0	100	0.0
K_meso_Y	0.0	100	0.0	0.0	100	0.0	0.0	100	0.0
H_Y	17.7	82.3	0.215	75.8	24.2	3.133	81.8	18.2	4.481
La_meso_Y	40.3	59.7	0.674	88.4	11.6	7.634	90.9	9.1	9.973
Ce_meso_Y	38.3	61.7	0.621	68.2	31.8	2.144	74.1	25.9	2.864
Ce/HY-WI	19.0	81.0	0.235	87.4	12.6	6.926	85.7	14.3	5.972
Regenerated Ce_meso_Y	38.4	61.6	0.623	69.3	30.7	2.258	65.0	35.0	1.861

---

To further understand the PE catalytic cracking reaction over the Ce_meso_Y catalyst, a series of experiments were conducted by varying reaction conditions. The effect of reaction temperature was first analyzed. In the absence of hydrogen and noble metals, activity was relatively low below 280 °C, with conversions of 41.3% at 260 °C and 60.4% at 280 °C (Fig. 3a and Fig. S3†). However, above 300 °C, the activity increased significantly, reaching complete conversion (100%), although temperatures exceeding 320 °C led to over-cracking and increased gas yields at the expense of valuable liquid products.

Fig. 3 PE catalytic cracking results over Ce_meso_Y as a function of (a) reaction temperature (0.2 g of PE) and (b) reaction time. (c) C and H contents and H/C ratio of product and residues as a function of reaction time. Reaction conditions: (unless otherwise mentioned) 300 °C, 4 h, atmospheric Ar, 0.1 g catalyst, and 0.3 g PE.

Given that the yield of the naphtha component also decreased above 300 °C, we selected 300 °C as the optimal temperature for subsequent reaction tests. Additionally, reaction time studies were conducted at a higher PE/catalyst ratio of 3, as complete conversion was already achieved at a ratio of 2, making it difficult to observe differences in reaction progression (Fig. S4†). At the early stage of the reaction (2 hours), the conversion was 47.5% and gradually increased with time, 73.1% at 4 hours, 84.6% at 8 hours, and stabilizing at approximately 95% from 16 hours onward (Fig. 3b and Fig. S5†). Notably, up to 16 hours, the yield of the higher-value-added naphtha fraction increased in parallel with the conversion, with the highest naphtha yield of 76.8%. These results underscore the high selectivity of Ce_meso_Y toward this valuable product fraction. We also monitored the distribution of carbon (C) and hydrogen (H) throughout the reaction (Fig. 3c). As the reaction time elapsed, the C and H content gradually shifted from solid residue to product, accompanied by a progressive decrease in the H/C ratio of the residue and a corresponding increase in that of the product. This observation suggests that hydrogen in PE was effectively transferred from residue to product as the reaction progressed, aligning with our hypothesis that the strong LASs in Ce_meso_Y facilitate hydrogen transfer in the self-supplied hydrogen mechanism, explaining its superior performance under hydrogen-free conditions.

Based on these reaction results and mechanisms previously reported,^{24,32,33,41,45} we propose a PE cracking mechanism using an acidic catalyst under hydrogen-free conditions (Fig. 4). Polymer chains initially undergo hydride abstraction at strong LASs, resulting in protonation and carbocation formation. These carbocations then migrate to BASs, where skeletal rearrangement occurs to form more branched stable intermediates that favor the production of iso-alkanes over n-alkanes. Following this rearrangement, β-scission cleaves the C–C bond, generating shorter hydrocarbon fragments that subsequently undergo deprotonation to facilitate the formation of C=C double bonds. These newly formed double bonds can re-enter the reaction cycle through hydride abstraction at strong LASs, leading to further protonation. When protonated species are positioned appropriately near adjacent C=C bonds, ring closure can form cycloalkanes,³³ which may further undergo dehydrogenation and aromatization to yield aromatics. Concurrently, hydrogen transfer reactions mediated by strong LASs convert olefins into alkanes,⁴¹ influencing overall product distribution.

Fig. 4 Proposed PE catalytic cracking mechanisms.

To validate the proposed hydrogen transfer mechanism, we conducted model reactions using hexane-d14 and n-pentane over both Na_meso_Y (prior to Ce exchange) and Ce_meso_Y catalysts, followed by mass spectrometric analysis of the reaction products (Fig. 5).⁴¹ Hydrogen–deuterium exchange was observed exclusively over Ce_meso_Y, which contains strong Lewis acid sites introduced via cerium exchange. Specifically, partially deuterated hexane (H-for-D exchange) and hydrogen-incorporated n-pentane (D-for-H exchange) were detected, indicating active hydrogen transfer. In contrast, Na_meso_Y—lacking such Lewis acidity—exhibited negligible exchange. These results indicate that Ce-induced Lewis acid sites facilitate the proposed hydrogen transfer mechanism.

Fig. 5 Mass spectra after reaction on Na_meso_Y and Ce_meso_Y of (a) pentane and (b) hexane.

To systematically examine the influence of acid site characteristics on catalytic performance, we analysed the relationship between conversion and acid site density. For BASs, conversion increased with increasing BAS density (Fig. 6a), suggesting their essential role in C–C bond cleavage, as reported previously.^38,39 This linear relationship provides strong evidence that BASs serve as primary active sites for the initial bond scission steps in polyethylene depolymerization. To further probe this correlation, we synthesized La-meso-Y and Ce-meso-Y catalysts with varying metal precursor concentrations (0.1–0.3 mol L⁻¹) and analyzed their acid site characteristics and catalytic performance (Table S1,† 4–5). NH₃-TPD and pyridine-DRIFTS analyses of La-meso-Y revealed an increase in Brønsted acidity with increasing La content, resulting in a conversion enhancement from 64.8% to 82.4% (Fig. S6–8†). In contrast, Ce-meso-Y exhibited a volcano-shaped trend in both acidity and conversion. Moderate Ce loading (Ce_meso_Y-0.2) led to a conversion increase from 81.8% to 100%, whereas excessive loading (Ce_meso_Y-0.3) reduced conversion to 70.7%, likely due to cerium oxide aggregation and a loss of Brønsted acidity.^46–48 These results demonstrate that catalytic performance is governed more directly by the density of active acid sites than by the overall elemental ratio. In contrast, no clear trend emerged when plotting strong LAS density against conversion (Fig. S9a†). While conversion generally increased with increasing the concentration of strong LASs, the correlation lacked the clear linearity observed with BASs. This observation aligns with previous studies reporting an unclear relationship between LASs and overall conversion.³⁹ The absence of a direct correlation suggests that strong LASs might contribute to the reaction through secondary mechanisms rather than direct C–C bond activation.

Fig. 6 Correlations between catalytic properties and reactivity. (a) Conversion as a function of Brønsted acid sites (BASs) density. (b) H/C ratio difference between products and residues as a function of strong Lewis acid sites (LASs) density.

Given this ambiguity, we sought to identify an alternative pathway through which LASs might influence catalytic performance. Our earlier analysis of Ce_meso_Y revealed that increases in the difference in the H/C ratio over reaction time were associated with hydrogen transfer processes (Fig. 3c). This observation raised an important question: could strong LASs facilitate hydrogen transfer reactions that indirectly enhance conversion? To investigate this hypothesis, we implemented an alternative analytical approach focusing on hydrogen transfer reactions as a potential mechanism for LAS-mediate effects. Upon plotting the H/C ratios of both products and residues against the strong LAS density (Fig. S9b†), we discovered a compelling trend. The difference in H/C ratio between products and residues increased proportionally with strong LAS density (Fig. 6b), providing clear evidence that strong LASs facilitate hydrogen transfer reactions. Because HY, La-meso-Y, and Ce-meso-Y exhibit similar product distributions, we compared this trend at similar conversion levels and obtained consistent results, which can be attributed to differences in hydrogen yield (Fig. S10†). This finding explains the previously unclear relationship between LASs and conversion: rather than directly catalyzing C–C bond cleavage, strong LASs enhance overall conversion by promoting more efficient hydrogen redistribution during the reaction.

To determine whether the acidity arises from the introduction of Ce species or the ion-exchange method itself, we synthesized Ce/HY-WI catalysts with the same Ce content (as measured by ICP) and compared their reactivity (Table S1†). Reaction results revealed that Ce introduction via wet impregnation led to a conversion of 73.3%, which was higher than that of H_Y (56.4%) but lower than that of Ce_meso_Y (100%) (Fig. 7a and Fig. S11a†). BET analysis showed that the pore structure of Ce/HY-WI closely resembled that of H_Y, indicating minimal impact on the pore structure from the impregnation method (Fig. S12a†). NH₃-TPD analysis indicated that while the overall acidity increased with Ce impregnation (1.92 mmol NH₃ g⁻¹ for Ce/HY-WI vs. 1.77 mmol NH₃ g⁻¹ for H_Y; Table 2 and Fig. S12b†), consistent with previous studies.^49–52 However, pyridine-DRIFTS analysis revealed that the Ce/HY-WI catalyst lacked strong LASs, similar to H_Y (Fig. 2h). These results suggest that although Ce impregnation enhances overall acidity and reactivity, its effectiveness is limited compared to Ce_meso_Y. This is primarily due to the absence of strong LASs, which are essential for hydrogen transfer in PE cracking under hydrogen-free conditions.

Fig. 7 PE catalytic cracking results and characterization of Ce_meso_Y zeolite. (a) Product yield over Ce_meso_Y, Ce/HY-WI, H_Y and pyridine-poisoned Ce_meso_Y. Reaction conditions: 300 °C, 4 h, atmospheric Ar, 0.1 g catalyst, and 0.2 g PE. (b) Ce 3d XPS spectra of Ce/HY-WI and Ce_meso_Y catalyst. (c) Schematic illustration of ion-exchange and wet impregnation methods.

To further clarify the role of Ce, we conducted cerium 3d X-ray photoelectron spectroscopy (XPS) analysis. In Ce/HY-WI, a pronounced Ce(IV) peak at 916.6 eV was observed, whereas this peak was significantly weaker in Ce_meso_Y^{40,47,48,53,54} (Fig. 7b). This difference stems from the synthesis methods: the ion-exchange process employed for Ce_meso_Y preferentially generates Ce(III) by replacing Na⁺ ions, while the wet impregnation method predominantly forms CeO₂ in Ce/HY-WI⁴⁰ (Fig. 7c). Consistent with previous reports, the XPS spectra of Ce_meso_Y closely align with those of Ce(NO₃)₃, confirming that cerium remains mainly in the trivalent state (Ce³⁺).⁵³ This trend was observed even with used and regenerated catalysts, suggesting that their oxidation state remained stable throughout the reaction cycle (Fig. S13†). The predominance of Ce³⁺ enhances the formation of strong LASs,^44,55 thereby improving catalytic performance. TEM and energy-dispersive X-ray spectroscopy (EDX) further confirmed a uniform Ce distribution throughout the internal cross-section of Ce_meso_Y (Fig. 8 and Fig. S14†). Finally, acid site poisoning experiments using pyridine resulted in a significant drop in conversion (to 19.1%; Fig. 3a and Fig. S11b†), underscoring the essential role of BASs and LASs in PE catalytic cracking. To investigate the spatial proximity between Lewis and Brønsted acid sites in facilitating hydrogen transfer, we evaluated catalytic activity using HY zeolites with varying Si/Al ratios (Fig. S15†). An increase in Si/Al ratio leads to lower Brønsted acid site density and greater inter-site distance,^56–58 both of which contribute to reduced conversion. This highlights the importance of acid site proximity within a single framework (Fig. S15 and Table S6–7†).

Fig. 8 STEM-EDX mappings of Al, Si, and Ce elements in Ce_meso_Y catalyst.

During the PE catalytic cracking process, coke deposition primarily occurs on the zeolite. TG-DTA analysis showed that coke accumulation was reduced by 31.8% for Ce_meso_Y and by 49.7% for Ce/HY-WI relative to their parent zeolites (Fig. 9a). Since the weight loss of the fresh catalysts was only ∼3% (Fig. S16†), the observed weight loss can be mainly attributed to coke deposition. This difference in coke formation can be attributed to the enhanced hydrogen transfer efficiency facilitated by the Ce_meso_Y catalyst. Coke formation typically results from the condensation of unsaturated hydrocarbons;⁵⁹ however, the strong LASs in Ce_meso_Y effectively promote hydrogen transfer reactions. These reactions convert unsaturated compounds into their saturated counterparts, thereby substantially reducing coke accumulation on the catalyst surface.

Fig. 9 Analysis of spent and regenerated catalyst characterization and results of stability and versatility tests. (a) Thermogravimetric analysis (TGA) of spent Ce_meso_Y and spent Ce/HY-WI catalysts. (b) N₂ adsorption-desorption isotherms and BJH pore size distribution plots of regenerated Ce_meso_Y. (c) Pyridine-DRIFTS spectra of regenerated Ce_meso_Y catalysts at different desorption temperatures. (d) Reusability test results. Reaction conditions: 300 °C, 4 h, atmospheric Ar, 0.1 g catalyst, and 0.3 g reactant, and (e) versatility test results for various post-consumer waste plastics. Reaction conditions: 320 °C, 4 h, atmospheric Ar, 0.1 g catalyst, and 0.2 g reactant.

Based on DTA analysis (Fig. S17a†), the calcination temperature for catalyst regeneration was set to 550 °C. Catalytic characterization of the regenerated catalyst indicated minimal changes in catalyst properties (Table 1). BET analysis of the regenerated catalyst revealed a slight decrease in pore volume and surface area while maintaining comparable performance to the fresh catalyst (Fig. 9b). NH₃-TPD and pyridine-DRIFTS analyses confirmed consistent acidity (Fig. S17b†) and retention of strong LASs in both the regenerated and original catalysts (Fig. 9c and Table 2). Reusability tests under mild reaction conditions were conducted to assess catalyst stability (Fig. 9d). The catalyst maintained consistent performance over three cycles, with only a slight decrease in PE conversion (73.1% → 71.3% → 67.7%) and no significant alteration in product distribution (Fig. S18†). Notably, naphtha selectivity also remained consistently high (90.0% → 86.1% → 84.3%). To demonstrate practical applicability, we tested Ce_meso_Y with various post-consumer plastics (Fig. 9e and Fig. S19†). The reaction with HDPE bottles achieved 70.5% conversion and 84.8% naphtha selectivity, while commercial LDPE yielded 73.7% conversion and 79.0% naphtha selectivity. Interestingly, PP exhibited higher conversion than PE, achieving 85.2% conversion with 85.9% naphtha selectivity and 82.6% conversion with 77.8% naphtha selectivity for a PP file case. To further explore the potential applicability of upcycling waste automobile shredder residue plastics, a mixture of PP and acrylonitrile butadiene styrene (ABS) was used as a model feedstock,⁶⁰ achieving a high conversion of 83.8% and a naphtha selectivity of 82.1% (Fig. S20†). These findings highlight the robustness and versatility of Ce_meso_Y across diverse feedstocks, reinforcing its promise for industrial applications in plastic waste upcycling.

Conclusions

We have successfully developed a hydrogen-free catalytic upcycling process for polyolefins using Ce_meso_Y. This approach eliminates the need for external hydrogen sources by utilizing strong LASs that facilitate hydrogen transfer from polyolefins, enabling efficient C–C bond cleavage at BASs. Among various ion-exchanged zeolites, Ce_meso_Y demonstrated the highest catalytic activity, achieving 100% PE conversion with 88.7% selectivity toward valuable naphtha products. Our mechanistic studies revealed that the synergistic effect between BASs and strong LASs is crucial for catalytic performance—BASs facilitate C–C bond cleavage while strong LASs enable efficient hydrogen redistribution. Comparative studies between ion-exchanged and impregnated catalysts confirmed that the ion-exchange method generates predominantly Ce³⁺ species, which enhance the formation of strong LASs essential for hydrogen transfer in the absence of external hydrogen. The catalyst demonstrated excellent stability over multiple reaction cycles and maintained high performance across various post-consumer polyolefins, achieving conversions of 70.5–85.2% with 77.8–85.9% selectivity toward naphtha. This study clarifies the previously ambiguous role of LASs in polyolefin depolymerization and provides a framework for designing more effective catalysts for hydrogen-free plastic upcycling. By eliminating dependence on fossil-based hydrogen sources, this approach represents a significant step toward more sustainable plastic recycling technologies with reduced environmental impact.

Author contributions

Taeeun Kwon: formal analysis, investigation, methodology, validation, visualization, writing – original draft, writing – review & editing; Jonghyun Park: investigation; Ki Hyuk Kang: investigation; Dae Sung Jung: investigation; Wangyun Won: supervision; Insoo Ro: project administration, conceptualization, supervision, funding acquisition, writing – original draft, writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the ESI.†

Acknowledgements

This work was supported by Hyundai Motor Company. This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (RS-2025-24523348 to I. R.).

References

1. P. Europe, Plastics - the fast Facts 2024, 2024.
2. OECD, Policy Scenarios for Eliminating Plastic Pollution by 2040, OECD Publishing, Paris, 2024.
3. S. Chen, A. Tennakoon, K.-E. You, A. L. Paterson, R. Yappert, S. Alayoglu, L. Fang, X. Wu, T. Y. Zhao, M. P. Lapak, M. Saravanan, R. A. Hackler, Y.-Y. Wang, L. Qi, M. Delferro, T. Li, B. Lee, B. Peters, K. R. Poeppelmeier, S. C. Ammal, C. R. Bowers, F. A. Perras, A. Heyden, A. D. Sadow and W. Huang, Nat. Catal., 2023, 6, 161–173.
4. V. D. Knyazev, J. Phys. Chem. A, 2007, 111, 3875–3883.
5. S. Jung and I. Ro, Korean J. Chem. Eng., 2023, 40, 693–705.
6. J. Sun, J. Dong, L. Gao, Y.-Q. Zhao, H. Moon and S. L. Scott, Chem. Rev., 2024, 124, 9457–9579.
7. T. Kwon and I. Ro, Korean Chem. Eng. Res., 2023, 61, 505–516.
8. T. Kwon, H. Jeong, M. Kim, S. Jung and I. Ro, Langmuir, 2024, 40, 17212–17238.
9. S. Al-Salem, A. Antelava, A. Constantinou, G. Manos and A. Dutta, J. Environ. Manage., 2017, 197, 177–198.
10. S. D. A. Sharuddin, F. Abnisa, W. M. A. W. Daud and M. K. Aroua, Energy Convers. Manage., 2016, 115, 308–326.
11. Y. Peng, Y. Wang, L. Ke, L. Dai, Q. Wu, K. Cobb, Y. Zeng, R. Zou, Y. Liu and R. Ruan, Energy Convers. Manage., 2022, 254, 115243.
12. J. Kim, S. Sun, D. Kim, B. G. Park, H. Lee, W. Huang and K. An, Chem. Catal., 2024, 4, 101076.
13. A. Musa, E. Jaseer, S. Barman and N. Garcia, Energy Fuels, 2024, 38, 1676–1691.
14. T. Kim, H. Nguyen-Phu, T. Kwon, K. H. Kang and I. Ro, Environ. Pollut., 2023, 121876.
15. J. A. Sun, P. A. Kots, Z. R. Hinton, N. S. Marinkovic, L. Ma, S. N. Ehrlich, W. Zheng, T. H. Epps III, L. T. Korley and D. G. Vlachos, ACS Catal., 2024, 14, 3228–3240.
16. B. C. Vance, P. A. Kots, C. Wang, Z. R. Hinton, C. M. Quinn, T. H. Epps III, L. T. Korley and D. G. Vlachos, Appl. Catal., B, 2021, 299, 120483.
17. P. A. Kots, S. Liu, B. C. Vance, C. Wang, J. D. Sheehan and D. G. Vlachos, ACS Catal., 2021, 11, 8104–8115.
18. S. Liu, P. A. Kots, B. C. Vance, A. Danielson and D. G. Vlachos, Sci. Adv., 2021, 7, eabf8283.
19. J. E. Rorrer, A. M. Ebrahim, Y. Questell-Santiago, J. Zhu, C. Troyano-Valls, A. S. Asundi, A. E. Brenner, S. R. Bare, C. J. Tassone and G. T. Beckham, ACS Catal., 2022, 12, 13969–13979.
20. J. E. Rorrer, C. Troyano-Valls, G. T. Beckham and Y. Román-Leshkov, ACS Sustainable Chem. Eng., 2021, 9, 11661–11666.
21. J. E. Rorrer, G. T. Beckham and Y. Román-Leshkov, JACS Au, 2020, 1, 8–12.
22. M. Tamura, S. Miyaoka, Y. Nakaji, M. Tanji, S. Kumagai, Y. Nakagawa, T. Yoshioka and K. Tomishige, Appl. Catal., B, 2022, 318, 121870.
23. Y. Nakaji, M. Tamura, S. Miyaoka, S. Kumagai, M. Tanji, Y. Nakagawa, T. Yoshioka and K. Tomishige, Appl. Catal., B, 2021, 285, 119805.
24. T. Kwon, B. Ahn, K. H. Kang, W. Won and I. Ro, Nat. Commun., 2024, 15, 1–15.
25. H. Nguyen-Phu, T. Kwon, T. Kim, L. T. Do, K. H. Kang and I. Ro, Chem. Eng. J., 2023, 475, 146076.
26. W. T. Lee, A. van Muyden, F. D. Bobbink, M. D. Mensi, J. R. Carullo and P. J. Dyson, Nat. Commun., 2022, 13, 4850.
27. B. C. Vance, P. A. Kots, C. Wang, J. E. Granite and D. G. Vlachos, Appl. Catal., B, 2023, 322, 122138.
28. B. C. Vance, Z. Yuliu, S. Najmi, E. Selvam, J. E. Granite, K. Yu, M. G. Ierapetritou and D. G. Vlachos, Chem. Eng. J., 2024, 487, 150468.
29. J. N. Hancock and J. E. Rorrer, Appl. Catal., B, 2023, 338, 123071.
30. J. Lee, T. Kwon, K. Hyuk Kang, W. Won and I. Ro, Angew. Chem., Int. Ed., 2025, 64, e202420748.
31. Z. Chen, B. J. Erwin and L. Che, J. Cleaner Prod., 2023, 424, 138861.
32. F. Zhang, M. Zeng, R. D. Yappert, J. Sun, Y.-H. Lee, A. M. LaPointe, B. Peters, M. M. Abu-Omar and S. L. Scott, Science, 2020, 370, 437–441.
33. J. Du, L. Zeng, T. Yan, C. Wang, M. Wang, L. Luo, W. Wu, Z. Peng, H. Li and J. Zeng, Nat. Nanotechnol., 2023, 18, 772–779.
34. J. Duan, W. Chen, C. Wang, L. Wang, Z. Liu, X. Yi, W. Fang, H. Wang, H. Wei, S. Xu, Y. Yang, Q. Yang, Z. Bao, Z. Zhang, Q. Ren, H. Zhou, X. Qin, A. Zheng and F.-S. Xiao, J. Am. Chem. Soc., 2022, 144, 14269–14277.
35. Z. Zhang, H. Chen, G. Li, W. Hu, B. Niu, D. Long and Y. Zhang, ACS Catal., 2024, 14, 2552–2561.
36. M. L. Pennel, A. K. Maurya, A. M. Ebrahim, C. J. Tassone and M. Cargnello, ACS Sustainable Chem. Eng., 2023, 11, 12623–12630.
37. Q. Li, K. C. Hunter and A. L. East, J. Phys. Chem. A, 2005, 109, 6223–6231.
38. V. Mortezaeikia, O. Tavakoli and M. S. Khodaparasti, J. Anal. Appl. Pyrolysis, 2021, 160, 105340.
39. A. Peral, J. Escola, D. Serrano, J. Přech, C. Ochoa-Hernández and J. Čejka, Catal. Sci. Technol., 2016, 6, 2754–2765.
40. J. Li, Y. He, L. Tan, P. Zhang, X. Peng, A. Oruganti, G. Yang, H. Abe, Y. Wang and N. Tsubaki, Nat. Catal., 2018, 1, 787–793.
41. Z. Cen, X. Han, L. Lin, S. Yang, W. Han, W. Wen, W. Yuan, M. Dong, Z. Ma, F. Li, Y. Ke, J. Dong, J. Zhang, S. Liu, J. Li, Q. Li, N. Wu, J. Xiang, H. Wu, L. Cai, Y. Hou, Y. Cheng, L. L. Daemen, A. J. Ramirez-Cuesta, P. Ferrer, D. C. Grinter, G. Held, Y. Liu and B. Han, Nat. Chem., 2024, 16, 871–880.
42. R. Liu, Y. Bian and W. Dai, Chin. J. Struct. Chem., 2024, 43, 100250.
43. J. Weitkamp, ChemCatChem, 2012, 4, 292–306.
44. X. Guo, Z. Li, A. Zhou, Y. Shi, Y. Wang and R. Zhou, J. Rare Earths, 2024, 42, 2068–2077.
45. E. Selvam, P. A. Kots, B. Hernandez, A. Malhotra, W. Chen, J. M. Catala-Civera, J. Santamaria, M. Ierapetritou and D. G. Vlachos, Chem. Eng. J., 2023, 454, 140332.
46. D. R. Araújo, J. Silva, J. L. d. Macedo, G. F. Ghesti, S. C. Dias, J. A. Dias and G. N. R. Filho, J. Braz. Chem. Soc., 2011, 22, 1894–1902.
47. J. Li, P. Zeng, L. Zhao, S. Ren, Q. Guo, H. Zhao, B. Wang, H. Liu, X. Pang and X. Gao, J. Catal., 2015, 329, 441–448.
48. Z. Mo, Y. Qin, Y. Zu, H. Wang, X. Zhang and L. Song, ChemistrySelect, 2019, 4, 13034–13044.
49. P. Zhao, W. Guo, Z. Gui, J. Jiang, Z. Zhu, J.-J. Li, L. Zhao, J. Zhou and Z. Xi, ACS Sustainable Chem. Eng., 2024, 12, 5738–5752.
50. X. Wu, X. Wang, L. Zhang, X. Wang, S. Song and H. Zhang, Angew. Chem., 2024, 136, e202317594.
51. J. Liu, L. Ma, H. Zhang, Y. Zhao, P. Cheng and Z. Zhao, J. Porous Mater., 2024, 1–11.
52. J. Sun, E. Selvam, A. Bregvadze, W. Zheng and D. G. Vlachos, Green Chem., 2025, 27, 3905–3915.
53. M. Xue, R. Chitrakar, K. Sakane, T. Hirotsu, K. Ooi, Y. Yoshimura, M. Toba and Q. Feng, J. Colloid Interface Sci., 2006, 298, 535–542.
54. J. Jiao, Y. Qin, J. Zheng, Y. Hui, L. Zhang, X. Gao and L. Song, J. Rare Earths, 2020, 38, 912–920.
55. L. Zinner and A. Araujo, J. Alloys Compd., 1992, 180, 289–294.
56. A. N. Mlinar, P. M. Zimmerman, F. E. Celik, M. Head-Gordon and A. T. Bell, J. Catal., 2012, 288, 65–73.
57. D. Zhai, L. Zhao, Y. Liu, J. Xu, B. Shen and J. Gao, Chem. Mater., 2015, 27, 67–74.
58. J. Zhou, F. Zheng, H. Li, J. Wang, N. Bu, P. Hu, J.-m. Gao, Q. Zhen, S. Bashir and J. L. Liu, Chem. Eng. J., 2020, 381, 122698.
59. V. Kapustin and V. Glagoleva, Pet. Chem., 2016, 56, 1–9.
60. H.-j. Kim, J.-s. Kim, S.-y. Kim, S.-k. Chun, D.-S. Jung and S.-h. Jeong, Clean Technol., 2024, 30, 322–329.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5gc01799h

---