Electrosynthesis of chlorine from seawater-like solution through single-atom catalysts

Chlorine is central to water treatment, organic synthesis, disinfectants, and pharmaceuticals. Industrial chlorine is produced via electrocatalytic oxidation of saturated brine in chlor-alkali processes, commonly using dimensionally stable anodes (DSA) composed of RuO2/TiO2. However, the intrinsic low conductivity and surface area of metal oxides limit activity and mass transport, and RuO2 is highly active for oxygen evolution (OER), leading to a scaling relationship between CER and OER that reduces chlorine selectivity (~95% Cl2 at pH ~2). Moreover, only a small fraction of Ru is accessible in bulk DSA. Carbon-supported single-atom catalysts (CS-SACs) offer nearly 100% metal atom utilization, low coordination environments, and tunable local structures, and have excelled in diverse electrocatalytic reactions. While atomically dispersed or cluster-based catalysts have been explored for chloride oxidation, single-atom nanomaterials specifically designed for CER and their mechanisms remain scarcely reported. This study synthesizes 2D carbon-supported Ru-O4 single-atom moieties (Ru-O4 SAM) anchored on oxygen-rich MOF-derived nanosheets and demonstrates superior CER activity, selectivity, and durability, elucidating a direct Cl− adsorption pathway that enhances performance versus RuO2.

Prior work has established RuO2/TiO2 DSA as the industrial standard for CER due to reduced energy consumption, but with limitations in electron transfer, surface area, and selectivity arising from competing OER. A scaling relationship between CER and OER on RuO2 suggests common active sites or intermediates, constraining selectivity. Advances in SACs with varied coordination environments (M–N, M–P, M–S) have boosted activity/selectivity in HER/OER and CO2/ORR systems, and recent atomically dispersed Pt catalysts showed efficacy for CER with potential-dependent mechanisms. However, single-atom catalytic design for CER and mechanistic understanding, particularly discriminating direct Cl− adsorption versus OCl− intermediates on Ru-based sites, have been limited. This work builds on these insights by creating an oxygen-coordinated Ru single-atom site to modulate adsorption energetics and decouple CER from OER.

Synthesis: Oxygen-coordinated Ru single-atom catalysts (Ru-O4 SAM) were prepared by impregnating ultrathin MOF-derived nanosheets (MOFNDs) with Ru(acac)3 followed by pyrolysis at 750 °C under Ar, anchoring isolated Ru atoms on oxygen-defect-rich 2D carbon matrices. MOFNDs were obtained by pyrolyzing Zn-BDC nanosheets at 950 °C, acid leaching, washing, and drying. A nanoparticle control (CS-Ru NPs) was synthesized similarly with higher Ru precursor concentration. Characterization: Morphology and dispersion were examined by TEM, HRTEM, HAADF-STEM (including aberration-corrected), EDS mapping, XRD, XPS, ICP-AES, N2 sorption (BET surface area ~1320 m2 g−1). XANES/EXAFS at Ru K-edge and soft XANES (O and C K-edges) probed oxidation state and coordination, with WT-EXAFS distinguishing Ru–O versus Ru–Ru scattering. EXAFS fitting quantified Ru–O coordination number (~3.8–4) and bond length (~1.99 Å). DFT formation energy comparisons among RuOxC4−x models identified RuO4 as most stable. Electrochemistry (H-cell): CER was evaluated in Ar-saturated 1 M NaCl (pH = 1) using a three-electrode H-type cell (working: catalyst on glassy carbon; counter: Pt mesh; reference: Ag/AgCl, converted to RHE). Polarization curves (5 mV s−1, 1600 rpm) with 95% iR compensation assessed onset potential and overpotential at 10 mA cm−2. Tafel slopes were derived in 1 M NaCl. Selectivity was determined by RRDE (ring at 0.95 V) and iodometric titration. TOF was calculated from current, area, Faraday constant, and moles of Ru. Stability was examined via chronoamperometry. Flow cell testing: A Nafion 324-separated flow cell used a gas diffusion layer anode with 0.1 mg cm−2 catalyst loading and a Pt/C cathode (1 cm2 area). Anolyte (NaCl) flow was 10 mL min−1 with Ar purge to remove Cl2. Cell voltage versus current density, long-term operation at 100 and 1000 mA cm−2, selectivity by iodometry, and dissolved Ru by ICP-MS were measured. Operando spectroscopy: In situ Raman (532 nm) and SR-FTIR probed intermediates under potentials from OCP to 1.45 V vs RHE in 1 M NaCl (pH 1). Operando Ru K-edge XAS in a custom cell tracked oxidation state and coordination changes with potential; EXAFS fitting in k2 and k3 weighting identified Ru–Cl formation during CER and reversibility post-bias. Computations: First-principles DFT (CASTEP, PBE-GGA, norm-conserving pseudopotentials, U = 4.2 eV on Ru) modeled RuOxC4−xC10/C12 (001) and RuO2(110) surfaces with vacuum spacing of 18 Å. Adsorbates (Cl, O, OH, OOH) were placed per operando-derived configurations. Energies, ZPE, and entropic terms yielded Gibbs free energies. PDOS and d-band centers were analyzed. Thermodynamic overpotentials for CER and OER, potential-determining steps, and selectivity descriptor ΔGSelectivity = ηTD(OER) − ηTD(CER) − 0.13 were computed.

• Catalyst structure: Ru atoms are atomically dispersed on oxygen-functionalized carbon, forming Ru–O4 moieties with Ru valence ~+3.2 (XANES), Ru–O coordination only (no Ru–Ru in EXAFS/WT). DFT shows RuO4 is the most stable configuration among RuOxC4−x models.
• Activity (H-cell, 1 M NaCl, pH 1): Onset potential Eonset = 1.37 V vs RHE. Overpotential to reach 10 mA cm−2 is −30 mV (i.e., 1.33 V vs RHE), outperforming DSA (85 mV) and CS-Ru NPs (110 mV). Intrinsic activity (ECSA-normalized) at 50 mV overpotential: 0.02 mA cm−2 ECSA (vs DSA 0.01 and CS-Ru NPs 0.002). Tafel slope ~48.2 mV dec−1, similar to DSA and CS-Ru NPs, indicating higher exchange current density drives enhanced activity.
• Selectivity: RRDE detects immediate ring current corresponding to Cl2 reduction with 99% Cl2 selectivity at disk current density >10 mA cm−2. Iodometric titration corroborates high selectivity. Superior selectivity persists at higher pH values (supplementary).
• Turnover and production: TOF 17.8 s−1 per Ru at 50 mV overpotential; production rate 1.6 mmol cm−2 h−1 at 1.43 V.
• Stability (H-cell): At 10 mA cm−2, current retention 95% after 12 h (vs CS-Ru NPs 80%, DSA 86%). Post-reaction XAS/EXAFS and AC-HAADF-STEM confirm maintained Ru–O coordination, atomic dispersion, and only slight Ru oxidation state increase (+3.5 after 1000 h) without aggregation.
• Flow cell performance: At 100 mA cm−2, cell voltage 1.52 V (lower than CS-Ru NPs 1.64 V; DSA 1.58 V). Cl2 selectivity >97.5% across a range of cell voltages. At constant 100 mA cm−2, only 4.5% voltage increase after 100 h; morphology and electronic structure preserved.
• Industrial-scale test: At 1000 mA cm−2, initial cell voltage 2.32 V, with no significant decline over 1000 h; Cl2 selectivity >98% throughout. Dissolved Ru <1 ppb after 1000 h (DSA ~5 ppb and ~94% selectivity, with voltage rising from 2.9 to 3.1 V).
• Mechanism (operando and DFT): In situ Raman on Ru-O4 SAM reveals emergent bands at 142 and 325 cm−1 under bias, assigned to Cl adsorption on Ru (Cl–Ru–O4), absent on RuO2. SR-FTIR detects O=Cl stretching at 727 cm−1 only on RuO2, consistent with OCl− intermediate; no 727 cm−1 signal on Ru-O4 SAM. Operando XAS shows formation of Ru–Cl coordination on Ru-O4 SAM at ≥1.4 V, reversible upon bias removal; RuO2 shows only Ru–O. CV indicates direct Cl2/Cl− reduction near 1.36 V on Ru-O4 SAM versus oxygen-site-mediated reduction on RuO2. DFT free-energy diagrams identify direct Cl− adsorption (*Cl) as favored on RuO4Cl0 and OCl-type adsorption on RuO2(110). The potential-determining step is Cl2 formation (recombination) with ΔG2 = 0.06 eV for RuO4Cl0 (ηTD(CER) = 0.16 V), 0.32 eV for RuO2(110), and 0.74 eV for RuO4Cl2. PDOS shows more negative Ru d-band center for RuO4Cl0 (−3.83 eV) versus RuO2(−2.85 eV) and RuO4Cl2(−0.98 eV), indicating optimized binding and facile Cl2 release. OER overpotentials: 0.67 V (RuO4Cl0), 0.77 V (RuO4Cl2), 0.63 V (RuO2). Selectivity descriptor ΔGSelectivity is 0.38 V for RuO4Cl0, higher than RuO2 (0.08 V) and RuO4Cl2 (−0.2 V), indicating superior CER selectivity.

The study demonstrates that tailoring the local coordination of Ru at the single-atom level enables a distinct CER pathway. Ru-O4 SAM provides isolated Ru centers that directly adsorb chloride (forming Ru–Cl intermediates) instead of forming OCl− species characteristic of RuO2 surfaces. This direct adsorption lowers the Gibbs free-energy barrier for Cl2 formation and optimizes intermediate binding, as supported by operando Raman, SR-FTIR, and XAS. DFT quantifies the thermodynamics, identifying RuO4Cl0 as the most effective configuration with minimal ΔG for the potential-determining recombination step and a favorable d-band center that balances adsorption and desorption. Concurrently, OER is sufficiently inhibited, widening the thermodynamic selectivity window for CER and thereby enhancing Cl2 selectivity. These mechanistic insights explain the markedly lower overpotential, higher TOF, and exceptional selectivity and stability observed in both H-type cells and flow cells at practically relevant current densities, positioning Ru-O4 SAM as a promising replacement or complement to DSA in chlorine production from seawater-like media.

This work introduces a two-dimensional carbon-supported Ru-O4 single-atom catalyst that achieves state-of-the-art CER performance in seawater-like, acidic NaCl electrolytes. Key contributions include: (1) synthesis of atomically dispersed Ru–O4 moieties with high surface area and robust electronic/geometric stability; (2) record-low overpotential (−30 mV at 10 mA cm−2), high TOF, and ~99% Cl2 selectivity in H-cells; (3) scalable flow-cell operation with low cell voltages and sustained >98% selectivity over 1000 h at 1000 mA cm−2 with negligible Ru leaching; and (4) operando spectroscopic and DFT evidence for a direct Cl− adsorption mechanism on Ru sites that lowers the CER free-energy barrier and enhances selectivity over OER. Future work could explore operation in real seawater with complex impurities (e.g., Br−, organics), optimization of membrane/reactor architectures for further energy savings, and extension of coordination engineering strategies to other single-atom centers for CER and related halogen electrosyntheses.

Experiments were primarily conducted in 1 M NaCl at pH = 1 and simulated seawater-like conditions; performance and selectivity in real seawater containing impurities such as bromide, iodide, organics, and particulates were not comprehensively evaluated. The mechanism is supported by operando spectroscopy and DFT thermodynamics; full kinetic modeling under mass-transport-limited industrial conditions was not provided. Long-term stability was demonstrated up to 1000 h; potential degradation pathways beyond this duration, and robustness under variable pH or temperature, remain to be assessed. Reactor-scale system considerations (membrane durability, gas handling, corrosion, and byproduct management) were outside the study’s scope.