Figure 6. a) Atomic configurations of ACP/RuSAC+C, where the distance between RuN4 and Ru cluster is 4.3 Å, 6.1 Å and 7.9 Å. The distance is measured between Ru hollow site of RuC and RuSAC. b) The binding Gibbs free energy of H* (ΔGH*) at the hollow site of RuC with varying distances. For comparison, ΔGH* in the absence of RuSAC is also presented. c) Atomic structures of ACP/RuC (left) and ACP/RuSAC+C (right), where the RuSAC and RuC correspond to Ru single atom catalyst and cluster, respectively. d) The Gibbs free energy diagram for HER. Note that ACP/RuSAC+C (H*) indicates the system with pre-adsorbed H* due to the favorable adsorption of H* on the Ru single atom site of ACP/RuSAC+C. e) Projected density of states (PDOS) for different Ru cluster models. Red dashed line indicates the d-band center of Ru atoms in the cluster of each model.
3. Conclusion
We developed a facile and cost-effective method to synthesize a unique self-standing electrode through a series of steps (acid treatment, immersion, and high-temperature pyrolysis). AIMD simulations revealed that the defects formed during acid treatment played a crucial role in accelerating the adsorption of Ru ions and extending their binding duration. This resulted in increased incorporation of Ru atoms into the support, even when utilizing an identical amount of Ru loading. The coexistence of Ru single atoms and Ru nanoclusters of ACP/RuSAC+C was further confirmed by AC HAADF-STEM, XPS, and XAS analysis. Moreover, DFT calculations confirmed the cooperative interaction between coupled Ru single atoms and nanoclusters, resulting in a synergistic effect that may contribute to attaining an optimal H* binding strength. Consequently, the ACP/RuSAC+C electrode exhibited relatively good HER performance in acidic electrolyte. Particularly, it demonstrated an ultrahigh TOF of 3.96, 9.7, and 18 s-1 at overpotentials of 100, 150, and 200 mV, respectively. Moreover, ACP/RuSAC+C achieved an enhanced mass activity, outperforming that of commercial Pt, demonstrating the intrinsic catalytic capability of ACP/RuSAC+C.
4. Experimental Section
Synthesis of ACP : The bare CP was cleaned by sequential sonication in acetone, DI water, and ethanol for 15 min each. Afterwards, it was immersed in a solution consisting of 45 mL H2SO4 and 15 mL HNO3 for 2 h at 60 ℃ with continuous stirring. Following this step, it was adequately soaked in deionized water and dried at 60 ℃ to obtain the ACP.
Synthesis of ACP/RuSAC+C : ACP (1 \(\times\) 0.5 cm2) was immersed in a well-dispersed 3.0 mg mL−1 Ru solution for 15 h. The ACP-Ru3+ was then rinsed multiple times with DI water and dried at room temperature. Subsequently, it was carbonized under a N2 atmosphere with 1.2 g dicyanamide as the nitrogen source. It was calcined at 800 ℃ for 1 h at a heating rate of 10 ℃/min. The final sample was designated as ACP/RuSAC+C.
Synthesis of ACP/RuC : ACP/RuC was prepared using the same method as that for ACP/RuSAC+C, with the only difference being the concentration of the Ru solution used. Instead of using a 3.0 mg mL−1 Ru solution, a 30 mg mL−1 Ru solution was employed for the preparation of ACP/RuC.
Synthesis of CP/Ru : CP/Ru was synthesized by the same route as that for ACP/RuSAC+C; however, bare CP was used instead of ACP.
Electrochemical Characterization: The HER performance measurements were performed using an Ivium potentiostat V55630 and a three-electrode electrolytic setup. The saturated calomel electrode (SCE) was used as the reference electrode and graphite rod was used as the counter electrode. The as-prepared electrodes were directly used as working electrodes (Geometric area: 0.25 cm2 ). The HER performance of the corresponding electrodes was tested in 0.5 M H2SO4 solutions. Linear sweep voltammetry (LSV) curves were performed at a scan rate 5 mV s−1 from 0 V to \(-\)0.84 V vs. SCE. Electrochemical surface area (ECSA) was determined from the double layer capacitance (Cdl) of the catalyst surface. The Cdlwas determined by measuring Cyclic Voltammetry (CV) in the non-Faradaic potential region at multiple scan rates ranging from 20 to 100 mV s−1. Electrochemical impedance spectroscopy (EIS) data were measured in the frequency range of 100 kHz\(-\)0.1 Hz. The stability of ACP/RuSAC+C was examined by Cyclic Voltammetry (CV) with a scan rate of 50 mV s−1 for 1000 cycles. Thereafter, the stability of the prepared ACP/RuSAC+C was continuously evaluated viachronoamperometry (CA) for 50 h. The Faradaic efficiency of electrode was measured in a typical H-type cell. Prior to the testing, the electrolyte was bubbled with N2 for at least 30 min. The gas product was analyzed using gas chromatography (GC, Agilent Technologies).
Computational details: We performed density functional theory (DFT) calculations using Vienna Ab initio Simulation Program code (VASP version 5.4.4.) with projector augmented wave (PAW) pseudopotentials and Perdew–Burke–Ernzerhof (PBE) exchange-correlation functional.[38-40] A plane wave cutoff energy of 400 eV was employed, and convergence tolerances for energy and force were set to 10-4 eV and 0.05 eV/Å, respectively. We modelled Ru nanocluster and Ru single-atom catalyst on a (4 \(\times\)8) supercell of graphene. To avoid an artificial interaction between periodic images in z -direction, a vacuum layer of 15\(\mathring{\mathrm{A}}\) was added.
The Gibbs free energy of HER was calculated considering two consecutive protonation steps.
\(*+\left(H^{+}+e^{-}\right)\leftrightarrow\ H*\) (1)
\(H*+\left(H^{+}+e^{-}\right)\leftrightarrow\ *+H_{2}\) (2)
where the asterisk (*) indicates the adsorption site. To convert DFT-calculated total energies into Gibbs free energy, correction values for adsorbates and H2 molecule were calculated using the harmonic oscillator approximation and the ideal gas approximation, respectively, as implemented in atomic simulation environment (ASE) (Table S1).[41] The effect of the electrode potential was included using the computational hydrogen electrode (CHE) method, which assumes the equivalent chemical potential of 0.5 H2 (g) and a proton-electron pair (H++ e-) at 0 V vs. RHE and standard conditions.[42] When the potential \(U\) is applied, the chemical potential of the electron is shifted by \(-eU\), i.e., \(G\left(H^{+}+e^{-}\right)=0.5G\left(H_{2}\right)-eU\), thus enabling the potential-dependent calculation of Gibbs free energies. We performed ab initio molecular dynamics (AIMD) simulations to investigate a reactivity between Ru precursors and defective/pristine graphene supports. The simulations were sampled by the canonical ensemble (NVT) employing the Nose–Hoover thermostat with a time step of 1.0 fs for 10 ps at 330 K.[43,44] Based on the experimental synthesis process and ID/IGratio results, we modelled pristine CP and defective CP (ACP) using a (4\(\times\) 3) supercell of graphene. To consider an aqueous solution containing RuCl3, 58 water molecules and 1 RuCl3 molecule were added to the computational cell, setting the water density to be approximately 1 g cm−3(Figures S2 and S3, Supporting Information).