From Single Atom to Five-Atom Cluster Catalysts onBoron-Doped Diamond - Interface Engineering and Dynamic Active SitesExploration for Acidic OER
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Figshare |
| Published | 2026-06-17 |
| Authors | Wenjie Cheng, Yutong Wang (3852589), Nan Gao (196665), Hongdong Li (137373) |
| DOI | 10.1021/acs.jpclett.6c01016.s001 |
The computed acidic OER overpotential for Ni5@BDD-β was 0.86 V at pH 0, the lowest among the three nickel structures examined. Its reported energy barrier was 2.09 eV. The five-atom β cluster improved on single-atom Ni@BDD by 0.03 V in overpotential and 0.03 eV in barrier. These are calculated values, not experimental OER measurements.
Performance against reference catalysts
Section titled “Performance against reference catalysts”The supporting information compares three nickel structures on boron-doped diamond with catalysts from prior studies. The table reproduces the reported overpotential, energy barrier, and stability condition.
| Catalyst | Overpotential (V) | Energy barrier (eV) | Stability condition | Source |
|---|---|---|---|---|
| Ni@ND | 1.55 | 2.78 | 500 K, 5 ps | Prior study 1 |
| Ni4@ND | 0.65 | 1.88 | 500 K, 5 ps | Prior study 1 |
| Sc@N2B2 | 0.79 | 2.02 | 500 K, 10 ps | Prior study 2 |
| Mn@ND | 1.15 | 2.38 | 300 K, 5 ps | Prior study 3 |
| RuO2 | 0.69 | 1.92 | Not reported | Prior study 4 |
| IrO2 | 0.56 | 1.63 | Not reported | Prior study 5 |
| Co/C2N2 | 0.30 | 1.53 | Not reported | Prior study 6 |
| Co/C3N | 0.43 | 1.66 | 300 K, 10 ps | Prior study 7 |
| Ni@BDD | 0.89 | 2.12 | 500 K, 10 ps | This work |
| Ni5@BDD-α | 1.20 | 2.43 | 500 K, 10 ps | This work |
| Ni5@BDD-β | 0.86 | 2.09 | 500 K, 10 ps | This work |
Within this work, Ni5@BDD-β had the lowest overpotential and energy barrier. Ni@BDD followed at 0.89 V and 2.12 eV. Ni5@BDD-α reached 1.20 V and 2.43 eV. The β cluster therefore did not establish a general five-atom advantage because its α counterpart performed worse than the single-atom model.
The β result was below Ni@ND and Mn@ND for both metrics. It remained above Ni4@ND, Sc@N2B2, RuO2, IrO2, Co/C2N2, and Co/C3N in both metrics. The paper’s strongest own result was therefore not the best value in the full comparison set. The stability entries describe temperature and simulation time, not experimental catalyst lifetime.
Acidic OER calculation
Section titled “Acidic OER calculation”The authors used the Nørskov electrochemical framework and the computational hydrogen electrode model. This model sets the chemical potential of the proton and electron pair to half that of gaseous H2 under standard hydrogen electrode conditions.
The acidic OER path follows four-electron water oxidation through adsorbed *OH, *O, and *OOH intermediates. The free energy at 298.15 K includes changes in reaction energy, zero-point energy, and entropy.
Applied electrode potential enters as ΔG_U = -neU, where n is the number of transferred electrons. The pH correction follows ΔG_pH = kBT ln 10 pH and is zero at pH 0.
Figure S1 reports the Ni@BDD Gibbs free-energy diagram under DFT+U with U = 6.0 eV. Supporting analyses include a bare BDD Gibbs diagram and a surface Pourbaix diagram for Ni@BDD at pH 0. The authors also compare Ni@BDD under 1H, 2H, and 3H coverages.
Interface search and cluster structures
Section titled “Interface search and cluster structures”The global structure search used particle swarm optimization through CALYPSO for M5@BDD, with M set to Fe, Co, Ni, Cu, and Pt. A 3 × 3 supercell contained an eight-layer oxygen-terminated BDD (100) surface.
The slab had bulk, unreconstructed surface, and reconstructed surface regions. The top three layers and clusters relaxed, while the bottom three layers remained fixed. Each generation produced 10 structures over 20 generations under point group symmetry constraints. The most stable and metastable structures were retained as the α and β cluster configurations.
Ni5@BDD-α had a reported average Ni-Ni distance of 2.016 Å, compared with 2.104 Å for β. The Ni120-Ni121 bond was 2.282 Å in α and 2.816 Å in β. The Ni121-Ni122 bond was 2.264 Å in α and 2.902 Å in β.
These values define distinct cluster geometries. They do not establish which geometry persists during experimental OER operation.
Adsorption and dissolution at the interface
Section titled “Adsorption and dissolution at the interface”Top-site adsorption was more negative than bridge-site adsorption for every cluster listed in Table S3. For Ni5@BDD-α, the values were -2.35 eV at the top and -1.99 eV at the bridge, a difference of 0.36 eV.
For Ni5@BDD-β, the adsorption energies were -2.46 eV and -2.18 eV, a difference of 0.28 eV. For Pt5@BDD-β, they were -3.01 eV and -2.83 eV, a difference of 0.18 eV.
The calculated dissolution potentials also varied with geometry. For Fe, Co, and Ni, the α and β values were 0.56/0.51 V, 0.55/0.49 V, and 0.77/0.74 V, respectively. For Cu and Pt, they were -0.75/-0.81 V and -0.39/-0.45 V.
The β value was lower in all five comparisons, by 0.03 V to 0.06 V. These calculations connect interface geometry with calculated adsorption and dissolution behavior rather than measured corrosion.
Figure S12 provides free-energy diagrams for Fe, Co, Cu, and Pt catalysts at U = 0 V. Figure S13 relates calculated OER overpotential to the d-band center and charge transfer.
Dynamic active sites and stability
Section titled “Dynamic active sites and stability”AIMD calculations followed total-energy evolution for Ni5@BDD-α and Ni5@BDD-β. The reported structures were optimized at 0 K and examined at 500 K for 10 ps with a 1 fs time step.
The analysis also considered cluster dissociation and surface migration. Figure S6 gives ΔE_rem for dissociation of five-atom clusters into four-atom clusters and single-atom adsorption systems. Figure S7 gives ΔE_a for diffusion in M5@BDD-β for Fe, Co, Ni, Cu, and Pt.
Potential energy surfaces examine the α and β configurations, including routes under *O for Fe, Co, Cu, and Pt. The supplied captions do not provide numerical ΔE_rem or ΔE_a values.
Figure S18 applies differential charge and Bader charge analysis to Cu5 clusters under different intermediate adsorption conditions. This analysis treats active sites as structure-dependent because adsorption, cluster geometry, and charge distribution are evaluated together. These are simulation results, so they do not show that the clusters rearrange experimentally under operating OER conditions.
Limits and engineering implications
Section titled “Limits and engineering implications”The supplied supporting information reports calculations but no catalyst synthesis or experimental electrochemical test. It gives no measured current density, Tafel slope, long-term durability, or dissolution rate for the nickel clusters.
The AIMD observation period is 10 ps, which cannot establish long-term catalyst lifetime. The free-energy model fixes pH 0 and 298.15 K. The interface model fixes oxygen termination and the BDD (100) orientation.
The supplied text does not specify boron concentration, metal loading, cluster preparation, or an experimental morphology for either nickel structure. An experimental program should therefore compare both Ni5 cluster geometries and track whether OER intermediates change the interface.
The results suggest this comparison because α and β have different adsorption, dissolution, and OER energetics. For diamond material selection, the paper supports boron-doped diamond as a catalyst interface whose behavior depends on metal nuclearity and local geometry.
The calculations do not show that larger clusters always improve acidic OER. Ni5@BDD-α was worse than Ni@BDD in both computed metrics. The 0.86 V value is a screening result for catalyst development, not a demonstrated operating voltage.
Materials and support
Section titled “Materials and support”The model requires an oxygen-terminated boron-doped diamond (100) surface carrying isolated metal atoms or five-atom clusters for acidic OER catalyst development. 6CCVD’s diamond materials catalog lists boron-doped diamond plates and wafers up to 125 mm, which matches the required substrate family. Researchers can contact the engineering support service to define the surface termination, geometry, and validation plan before fabrication.
View Original Abstract
The exceptional electrochemical stability of boron-doped diamond (BDD) makes it a promising oxygen evolution reaction (OER) anode in acidic media for proton exchange membrane water electrolyzers; however, its inertness leads to high overpotentials (η). To overcome this activity-stability conflict, we employ atomic-scale interface engineering via single-atom catalysts (SACs) and single-cluster catalysts (SCCs) anchored on BDD. Employing structure prediction and density functional theory (DFT) framework, we screen 28 SACs and 16 SCCs (including α and β isomers). Through stability assessments (formation energies, dissolution potentials, and diffusion barriers), M@BDD and M<sub>5</sub>@BDD-α/β (M = Fe, Co, Ni, Cu, and Pt) are identified as promising catalysts following the adsorbate evolution mechanism. Conventional descriptor analysis (η<sup>OER</sup> vs Δ<i>G</i><sub>*O</sub>-Δ<i>G</i><sub>*OH</sub>) reveals a volcano-type activity trend, and η<sup>OER</sup> exhibits a strong linear correlation with Δ<i>G</i><sub>*OOH</sub>. Crucially, a dynamic reaction pathway is unveiled where proton-coupled electron transfer and *O adsorption on Ni<sub>5</sub>@BDD-α/β trigger a dramatic reduction in the isomerization barrier, driving a thermodynamically favorable symmetry breaking and establishing a new multicenter bonding-mode. This active site evolution thereby circumvents the rate-determining step identified in static models, achieving a lower η of 0.56 V. This work establishes a design principle for BDD-based catalysts and provides fundamental insight into dynamic active sites.
Boron Doped Diamond On Titanium Bdd Electrode
View Specs →