Article pubs.acs.org/Langmuir
Rewriting Electron-Transfer Kinetics at Pyrolytic Carbon Electrodes Decorated with Nanometric Ruthenium Oxide Joseph F. Parker,† Gabrielle E. Kamm,‡ Ashlee D. McGovern,‡ Paul A. DeSario,† Debra R. Rolison,† Justin C. Lytle,*,‡ and Jeffrey W. Long*,† †
Surface Chemistry Branch, Code 6170, U.S. Naval Research Laboratory, Washington, DC 20375, United States Chemistry Department, Pacific Lutheran University, Tacoma, Washington 98447, United States
‡
S Supporting Information *
ABSTRACT: Platinum is state-of-the-art for fast electron transfer whereas carbon electrodes, which have semimetal electronic character, typically exhibit slow electron-transfer kinetics. But when we turn to practical electrochemical devices, we turn to carbon. To move energy devices and electro(bio)analytical measurements to a new performance curve requires improved electron-transfer rates at carbon. We approach this challenge with electroless deposition of disordered, nanoscopic anhydrous ruthenium oxide at pyrolytic carbon prepared by thermal decomposition of benzene (RuOx@CVD-C). We assessed traditionally fast, chloride-assisted ([Fe(CN)6]3−/4−) and notoriously slow ([Fe(H2O)6]3+/2+) electron-transfer redox probes at CVD-C and RuOx@CVD-C electrodes and calculated standard heterogeneous rate constants as a function of heat treatment to crystallize the disordered RuOx domains to their rutile form. For the fast electron-transfer probe, [Fe(CN)6]3−/4−, the rate increases by 34× over CVD-C once the RuOx is calcined to form crystalline rutile RuO2. For the classically outer-sphere [Fe(H2O)6]3+/2+, electron-transfer rates increase by an even greater degree over CVD-C (55×). The standard heterogeneous rate constant for each probe approaches that observed at Pt but does so using only minimal loadings of RuOx.
■
INTRODUCTION Nanostructured carbons are critical components in the electrodes that empower a wide range of electrochemical applications, from energy storage (e.g., Li-ion batteries,1 Li−S batteries,2 redox flow batteries,3 electrochemical capacitors4,5) and energy conversion (fuel cells6) to capacitive deionization7 and electroanalytical sensing.8 Because of the morphological diversity of nanostructured carbons, they can be designed to provide many functions in an electrochemical device, such as (i) storing charge by ion-insertion or double-layer capacitance; (ii) catalyzing electrochemical reactions directly; (iii) acting as a support for other electrocatalysts; or more simply, (iv) providing enhanced electronic conductivity to poorly conducting active materials (e.g., transition-metal oxides). One shortcoming of carbon-based electrodes has traditionally been suboptimal electron-transfer kinetics at the carbon surface, as observed when oxidizing and reducing simple redox probes in solution.9 Electron-transfer rates depend strongly on the particular type of surface expressed at the carbon: basal plane versus edge-plane sites, the presence of heteroatom dopants or other electronic defects, and carbon−oxygen surface moieties.10 Though many recent advances have been made in understanding electrochemical reactions at carbon surfaces11 and enhancing their electron-transfer rates,12 carbon electrodes still fall short of the activity achieved at noble metal electrodes (e.g., © XXXX American Chemical Society
Pt) with the notable exception of the pristine basal plane surfaces expressed by freshly cleaved highly oriented pyrolytic graphite (HOPG).13,14 Sluggish electron-transfer rates are particularly problematic when carbon electrodes are used for electroanalytical sensing15 but may also undercut performance in carbon-containing electrodes designed for energy-storage applications, for example, when electron handoff between the carbon conductor and active material (e.g., LiFePO4) in a powder-composite electrode becomes a bottleneck under highpower demands.16 While Pt is an ideal electrode material for fast electron transfer in many applications, its cost and susceptibility to poisoning limit practical applications. Ruthenium dioxide (RuO2), while still an oxide of a platinum-group metal, has been used extensively for industrial-scale electrochemical applications,17 where it exhibits outstanding electrocatalytic activity for such reactions as the chlor-alkali process.18 Inspired by the innately attractive electrochemical properties of RuO2,19 Special Issue: Fundamental Interfacial Science for Energy Applications Received: March 31, 2017 Revised: May 24, 2017
A
DOI: 10.1021/acs.langmuir.7b01107 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
dilute benzene vapor that was carried through separate plumbing into a trap. At 1000 °C, the slide was exposed to flowing benzene vapor for 3 min and then heated in flowing Ar for 10 min. This cycle was repeated once more and the furnace was turned off. All glassware for the cryogenic electroless deposition of ruthenia was washed sequentially in a base bath, an acid bath, and water before drying in an oven at T > 100 °C. A 10 mL ampule of RuO4 (0.5% aq solution Polysciences Inc.) was pipetted into a 125 mL separatory funnel and immediately washed for 30 s with 10 mL of petroleum ether (ACS reagent, boiling range 40−60 °C, ACROS Organics) that had been prechilled for 1 min in a dry ice/acetone bath. The bottom aqueous layer (pale yellow) was drained into a vial and the top layer was drained into a separate 20 mL vial that contained a thin layer of anhydrous magnesium sulfate (≥97% Sigma-Aldrich). The petroleum ether solution was swirled with the desiccant for several seconds and then decanted through a funnel lined with glass filter paper (90 mm type A/E Gelman Sciences) into an Erlenmeyer flask that was submerged in a dry ice/acetone bath. This process was repeated to extract additional RuO4 from the first aqueous layer. After collecting the second filtrate, the Erlenmeyer flask was gently warmed in an ice water bath for 2 min followed by pouring the solution over the CVD-C slides in a glass basin. The container holding the samples was then sealed and left undisturbed overnight. The following day, all slides were successively sonicated for 60 s each in three separate 150 mL baths of petroleum ether to remove any residual RuOx agglomerates. Materials Characterization. Grazing-incidence X-ray diffraction (GIXRD; Rigaku Smartlab, Cu Kα radiation) was collected in parallel beam mode at 40 kV and 44 mA with an incident beam angle of 0.4° and a 0.5° min−1 scan rate. Diffraction peak fitting and phase identification was performed with PDXL software (Rigaku). The chemical speciation of Ru and O was assessed by X-ray photoelectron spectroscopy (XPS; Thermo Scientific K-Alpha X-ray photoelectron spectrometer) at a chamber pressure of
