ИСТИНА

Oxygen reduction reaction (ORR) is a key reaction that drives the operation of fuel cells and metal-air batteries. ORR is known to have sluggish kinetics, due to its complex multistep character, thus a catalyst is strongly required to use ORR for practical applications. Noble metals and alloys are the most active catalysts for ORR, however, their high cost forces the search for non-noble materials with comparable catalytic activity [1]. Carbon doped with light elements (especially N and B) is considered as promising non-noble electrode material that exhibits high electrocatalytic activity towards ORR. When dopant substitutes a carbon atom in sp2 graphite lattice, active sites with high electron and spin density are generated either on the dopant atom itself or on the neighboring carbon atoms [2]. Such electrode materials are usually prepared starting from porous graphite oxide, by treating it with dopant precursors at high temperatures [3]. It’s often challenging to characterize such samples in terms of doping level, as dopant atoms are normally present in various forms, including surface contaminations, thus it’s hard to correlate the catalytic activity with the amount of dopant in graphite lattice. From the other hand, employed synthetic routes usually lead to the changes in morphology (porosity, specific surface area, a number structural of defects). That hinders the possibility to evaluate the electrocatalytic effects coming solely from substitutional doping, thus a lot of discrepancies could be found in the literature regarding the measured ORR kinetics on doped carbon materials. Having been initially introduced for the ORR in aqueous media, doped carbon materials now attract a lot of attention as catalysts also for ORR in aprotic electrolytes. A lot of works, both experimental and theoretical, shows that doped carbon catalysts provide higher ORR currents and low overpotential for such important practical systems as metal-oxygen batteries [4,5]. The main difference between ORR in aprotic and non-aqueous media is the relatively stable superoxide intermediate in the former case. It implies that the key ORR step in aprotic media is a quasireversible electron transfer to oxygen molecule: O2 + e- → O2-, and the overall rate of ORR will be determined by the rate constant of this reaction [6]. We use epitaxial graphene as a model carbon electrode with perfect planar morphology and composition easily controlled by the XPS and NEXAFS techniques, to evaluate the ORR kinetics in aprotic media on graphene with various types of doping. We compared the ORR kinetics on N-, B- and undoped epitaxial graphene grown by CVD on Ni(111)/W(110) single crystals, designed for two types of electrochemical cells: with hanging disk electrode, and disk electrode mounted in the cell bottom. The graphene structure, composition and type of substitution was precisely characterised by LEED, XPS and NEXAFS techniques. For the evaluation of ORR kinetics, oxygen-saturated 0.1 M solution of tetrabutylammonium perchlorate (TBAP) in dimethylsulfoxide (DMSO) was chosen as electrolyte, as DMSO is common solvent for metal-air batteries due to its relative chemical stability and high solvation ability. To determine the ORR rate constant on the graphene electrodes, the Nicholson approach for quasireversible electron transfer reactions was utilized [7]. It implies recording cyclic voltammograms of oxygen reduction/oxidation process at various potential sweep rates. By comparing the variation of peak-to-peak separation with the sweep rate with known tabulated values [7], electron transfer rate constants were determined for doped graphene electrodes and compared to undoped one. The stability issues of doped graphene upon ORR are also discussed.