Highly efficient recycling of polyester wastes to diols using Ru and Mo dual-atom catalyst

Plastic wastes pose substantial environmental challenges and demand circular-economy solutions. Chemical recycling offers routes to convert polymers into valuable chemicals, surpassing incineration and landfilling. Polyesters, comprising about one third of plastic markets, can be depolymerized by hydrolysis, alcoholysis, aminolysis, and hydrogenolysis, yet face limitations in degradation reactivity and selectivity to desired products. Diols (e.g., ethylene glycol, 1,2-propanediol) are important but energy-intensive to produce through conventional processes. Because polyesters form via condensation of diols and dicarboxylic or hydroxy-carboxylic acids, selective hydrogenolysis of ester carbonyls to CH2–OH could yield diols. However, exclusive production of diols is challenging due to formation of carboxylic acids and over-hydrogenation byproducts. Atomically dispersed metal catalysts, particularly dual-atom systems, can tune electronic structures and offer tandem active sites for selectivity control. This study asks whether Ru–Mo dual-atom catalysts can enable efficient, selective conversion of diverse polyesters to diols under mild, aqueous H2 conditions, while suppressing hydrodeoxygenation of the formed diols.

Recent advances highlight chemical upcycling of plastics including hydrocracking of polyolefins and multiple polyester depolymerization strategies (hydrolysis, alcoholysis, aminolysis, hydrogenolysis). Prior reports achieved hydrogenolysis of PLA to 1,2-propanediol at higher temperatures but struggled with selectivity due to carboxylic acid and byproduct formation. Ru-based catalysts can hydrogenate functionalized carboxylic acids to diols, yet over-hydrogenation to alkanes remains an issue. Single-atom catalysts (SACs) improve atom utilization and tunability; dual-atom catalysts (DACs) can exhibit synergistic effects via electronic interactions and tandem active sites, optimizing adsorption and lowering barriers. However, roles of partially unpaired metal single-atom sites are often overlooked. These insights motivate designing Ru–Mo DACs to decouple H2 activation from selective carbonyl hydrogenation while suppressing diol hydrodeoxygenation.

Catalyst synthesis: Anatase TiO2 nanoparticles (NPs) were dispersed in water (50 mg TiO2 in 50 mL). Aqueous precursors RuCl3 and ammonium molybdate were added at targeted Ru:Mo molar ratios, followed by dropwise excess NaBH4 under ultrasonication to generate Ru–Mo bimetal NPs on TiO2. The solids were filtered, washed (3×), dried (60 °C, vacuum), then treated in flowing O2 at 200 °C for 3 h to obtain dual-atom catalysts denoted RuxMoy/TiO2. A series with varying Ru:Mo were prepared; the optimal sample (Ru4Mo1/TiO2) had Ru:Mo ≈ 4:1 and Ru loading 3.76 wt%. Controls included Ru/TiO2 (6.0 wt% Ru) and Mo2/TiO2 (1.74 wt% Mo). Regeneration: Used catalysts were washed, dried, and re-treated in O2 at 200 °C for 3 h. Catalytic tests: Depolymerization of polyesters (e.g., PLA) conducted in a stainless-steel autoclave (15 mL Teflon liner) with polyester, catalyst, water, and H2 charged after degassing. Typical conditions: 160 °C, 4 MPa H2, 12 h for PLA; substrate 1 g, catalyst 100 mg, H2O 3 mL, H2 8 MPa, 48 h for larger-scale tests of various polyesters; solvent-free runs for molten PLA and ethyl acetate (3 mmol substrate, 30 mg catalyst, H2 8 MPa, 180 °C). Products were analyzed by 1H/13C NMR (paraformaldehyde internal standard) and GC-MS for gases. Conversions, yields, and selectivities were calculated from NMR peak areas and mass balances, with ≥3 repeats (mean SD <5%). Characterization: HR-TEM and AC HAADF-STEM to visualize atomic dispersion; EDS mapping and line scans for elemental distribution; XRD to confirm anatase TiO2 structure; XANES/EXAFS (Ru, Mo K-edges) and WT-XAFS to determine oxidation states and coordination (Ru–O and Mo–O paths, absence of Ru–Ru/Mo–Mo in optimal sample); soft XAS for surface oxygen states; EPR for paramagnetic species and electron environment; operando Raman (H2 dissociation via Ru–H at 867 cm−1; hydrogenation behavior); operando FTIR-SEIRAS and ex situ FTIR to monitor LA adsorption (C=O shift 1741→1727 cm−1) and conversion to alcohols (C–H bands ~2930 cm−1). Used and regenerated catalysts were likewise characterized to track structural changes (Ru SA to clusters under H2, re-dispersion after O2 treatment). Computations: DFT (VASP) with PAW and PBE-GGA, D3 dispersion, 12 Å vacuum, energy cutoff 450 eV, 2×2×1 k-mesh, convergence 1e−5 eV and 0.02 eV/Å. Models of Ru SA/TiO2, Mo SA/TiO2, and Ru–O–Mo dual-atom sites constructed. Adsorption geometries of lactic acid (LA) screened; PDOS and Bader charge analyses performed; free-energy profiles computed for hydrogenation and hydrodeoxygenation steps.

- Ru–Mo dual-atom catalyst Ru4Mo1/TiO2 converts diverse polyesters to corresponding diols with 100% selectivity under mild aqueous H2 conditions (e.g., 160 °C, 4 MPa). - PLA: At 160 °C and 4 MPa H2, lactic acid (LA) intermediate disappears and 1,2-propanediol yield approaches 100%; larger-scale runs give 97% yield; PLA straw and lid waste converted to 1,2-propanediol in 98% and 97% yields, respectively. - Other polyesters: PGA → ethylene glycol (100% selectivity); PCL → 1,6-hexanediol (100% selectivity); PBS → 1,4-butanediol (sole product); PBA → 1,6-hexanediol and 1,4-butanediol. PBAT: only adipic acid portion hydrogenated; terephthalic acid moiety not hydrogenated under tested conditions. - Solvent-free tests: Ethyl acetate fully hydrogenated to ethanol; molten PLA to 1,2-propanediol in 95% yield. - Structure–function: AC HAADF-STEM and EDS show Ru and Mo as single atoms on TiO2 with O-bridged Ru–Mo dual-atom (Ru–O–Mo) sites; Ru/TiO2 contains ~3 nm Ru nanoparticles. XANES indicates Ru valence between 0 and +4, higher in presence of Mo; Mo valence decreases in Ru4Mo1/TiO2 vs Mo2/TiO2, evidencing strong Ru–Mo electronic interplay. EXAFS shows Ru–O and Mo–O first-shell peaks without Ru–Ru or Mo–Mo in optimal sample, confirming atomic dispersion. - Mechanism: Operando Raman detects Ru–H vibration at 867 cm−1 from 25–140 °C, indicating strong H2 activation by Ru SAs. Operando Raman/FTIR show Ru4Mo1/TiO2 inhibits hydrodeoxygenation of 1,2-propanediol, unlike Ru/TiO2. FTIR-SEIRAS shows LA adsorbs via C=O on Mo sites (C=O shift 1741→1727 cm−1) and converts to –C–OH. - DFT: LA preferentially adsorbs at Mo of Ru–O–Mo (adsorption energy −1.76 eV). PDOS reveals d–d interaction within Ru–O–Mo lowers orbital energy and enhances delocalization, raising barriers for hydrodeoxygenation. Bader analysis shows electron transfer from Mo to LA carbonyl O upon adsorption. Free-energy diagrams show higher barrier (0.66 eV) for 1,2-propanediol → isopropanol step on Ru4Mo1/TiO2, explaining suppressed HDO; Ru SAs primarily activate H2; Ru/TiO2 favors further hydrogenolysis to alkanes. - Stability: Catalyst deactivates via Ru SA clustering under H2 but is fully regenerated by 200 °C O2 treatment; maintains activity over 10 reuse cycles.

The study addresses the challenge of selectively converting polyester wastes to diols without over-hydrogenation. By engineering a TiO2-supported Ru–Mo dual-atom catalyst comprising Ru single-atom sites and O-bridged Ru–Mo sites, the system decouples functions: Ru SAs efficiently activate H2 and provide active hydrogen for carboxyl reduction, while Ru–O–Mo sites preferentially adsorb carbonyls (via Mo) and disfavor hydrodeoxygenation of the resulting diols due to weak diol adsorption and higher energy barriers. This synergy yields near-quantitative conversion of PLA and other aliphatic polyesters to their diols at 160 °C and moderate H2 pressure in water, with complete selectivity. Spectroscopy (operando Raman and FTIR-SEIRAS), XAFS/HAADF-STEM, and DFT consistently support the proposed mechanism: LA formation via TiO2-acidic hydrolysis, H2 activation at Ru SAs with spillover, LA adsorption/activation at Ru–O–Mo, and selective hydrogenation to diols while suppressing HDO. The findings demonstrate an effective platform for chemical recycling that aligns with circular-economy objectives by directly producing high-value diols under relatively mild and regenerable conditions.

This work introduces a robust Ru–Mo dual-atom catalyst on anatase TiO2 that selectively recycles diverse aliphatic polyesters into corresponding diols with up to 100% selectivity under mild aqueous H2 conditions. Structural characterization reveals coexistence of Ru single-atom and Ru–O–Mo dual-atom sites with strong electronic interplay. Mechanistic studies and DFT show Ru SAs activate H2, while Ru–O–Mo sites adsorb and reduce carbonyls yet suppress diol hydrodeoxygenation, ensuring selectivity. The catalyst is regenerable by low-temperature O2 treatment and retains performance over multiple cycles. Future work could extend this design to aromatic polyesters (e.g., enhancing activity toward terephthalic units), optimize operation at lower H2 pressures/temperatures, scale continuous processes, and explore other DAC compositions/supports for broader plastic upcycling.

- The catalyst exhibits deactivation under H2 due to Ru single-atom aggregation into clusters; activity is restored by O2 regeneration at 200 °C, implying a regeneration step is needed for long-term operation. - Limited activity toward aromatic carboxyl groups: in PBAT, the terephthalic acid moiety was not hydrogenated under tested conditions, indicating constraints for PET-like substrates. - The process requires elevated H2 pressure (typically 4–8 MPa) and temperatures (up to 160–180 °C), which may impact energy and safety considerations. - Ru is a precious metal; although atom efficiency is high (single-atom usage), cost and resource considerations remain.