Electronic metal-support interaction enhanced oxygen reduction activity and stability of boron carbide supported platinum

Abstract

Catalysing the reduction of oxygen in acidic media is a standing challenge. Although activity of platinum, the most active metal, can be substantially improved by alloying, alloy stability remains a concern. Here we report that platinum nanoparticles supported on graphite-rich boron carbide show a 50-100% increase in activity in acidic media and improved cycle stability compared to commercial carbon supported platinum nanoparticles. Transmission electron microscopy and x-ray absorption fine structure analysis confirm similar platinum nanoparticle shapes, sizes, lattice parameters, and cluster packing on both supports, while x-ray photoelectron and absorption spectroscopy demonstrate a change in electronic structure. This shows that purely electronic metal-support interactions can significantly improve oxygen reduction activity without inducing shape, alloying or strain effects and without compromising stability. Optimizing the electronic interaction between the catalyst and support is, therefore, a promising approach for advanced electrocatalysts where optimizing the catalytic nanoparticles themselves is constrained by other concerns.

Figure 1. Catalyst particle size and dispersion.

TEM images taken at 270 k magnification showing Pt particle size and dispersion on the C and BC supports, TEM scale bars represent 20 nm; histogram scale bars represent 20%; histograms show Pt size distribution obtained from 100 particles across three images each.

Figure 2. Electronic equilibration across the catalyst support interface.

(a) XPS spectra of the C 1 s and Pt 4f region of the Pt/C and Pt/BC catalysts, spectra are aligned to the graphitic C (sp²) peak at 284 eV; (b) Schematic of charge transfer across the support-catalyst interface due to Fermi level equilibration rationalizing relative shifts in the XPS Pt 4f and C 1 s core levels; (c) In situ XANES L₂ and L₃ edge measured at 744 mV versus NHE probing d-band occupancy in electrochemical environment.

Figure 3. Electrochemical characterization.

(a) Shows cyclic voltammograms and Tafel plots for 10, 20 and 40 eq wt% Pt/BC, and (b) compares 40 eq wt% Pt/BC with 40 wt% Pt; CVs recorded at 20 mVs⁻¹ between 0.05 and 1.2 V versus RHE and normalized for surface area specific current, Tafel plots obtained from anodic scan recorded at 1,600 r.p.m. between 0.05 and 1.2 V versus RHE and normalized for surface area specific current; all experiments performed in 0.1 M HClO₄ at room temperature.

Figure 4. Catalyst cycle stability.

(a) Compares CVs between the first and 6,000th cycle, and (b) plots the change in ECSA after cycling at 50 mVs⁻¹ between 0.6 and 1.0 V versus RHE in 0.1 M HClO₄ at room temperature.

Figure 5. Catalyst particle shape.

Co-ordination numbers N₁ through N₄ for icosahedra and cuboctahedra with superimposed calculated co-ordination numbers from the EXAFS for the 20 eq wt% Pt/BC and 20 wt% Pt/C against average particle size as suggested by TEM. Error bars indicate s.d. of TEM particle size distribution and confidence intervals from EXAFS fit. Shape models adapted from Glasner and Frenkel.

Figure 6. Comparison of surface area specific ORR activities.

Specific current densities at 0.9 V versus RHE for the Pt/BC and Pt/C catalysts are compared with reference data for ORR activity of similar Pt/C catalysts from Garsany et al. adjusted for temperature and O₂ mass transport correction.