(M,Ru)O2 (M = Mg, Zn, Cu, Ni, Co) Rutiles and Their Use as Oxygen Evolution Electrocatalysts in Membrane Electrode Assemblies under Acidic Conditions

David L Burnett,Jamie G Lok,Christopher M Zalitis,Martin R Lees,Jonathan D B Sharman,Reza J Kashtiban,Richard I Walton,Katie M Rigg,Enrico Petrucco

doi:10.1021/acs.chemmater.0c01884

Abstract

The rutiles (M,Ru)O2 (M = Mg, Zn, Co, Ni, Cu) are formed directly under hydrothermal conditions at 240 °C from potassium perruthenate and either peroxides of zinc or magnesium, or poorly crystalline oxides of cobalt, nickel or copper. The polycrystalline powders consist of lath-shaped crystallites, tens of nanometres in maximum dimension. Powder neutron diffraction shows that the materials have expanded a axis and contracted c axis compared to the parent RuO2, but there is no evidence of lowering of symmetry to other AO2-type structures, supported by Raman spectroscopy. Rietveld refinement shows no evidence for oxide non-stoichiometry and provides a formula (MxRu1-x)O2 with 0.14 < x < 0.2, depending on the substituent metal. This is supported by energy-dispersive X-ray analysis on the transmission electron microscope, while Ru K-edge XANES spectroscopy shows that upon inclusion of the substituent the average Ru oxidation state is increased to balance charge. Variable temperature magnetic measurements provide evidence for atomic homogeneity of the mixed metal materials, with suppression of the high temperature antiferromagnetism of RuO2 and increased magnetic moment. The new rutiles all show enhanced electrocatalysis compared to reference RuO2 materials for oxygen evolution in 1 M H2SO4 electrolyte at 60 °C, with higher specific and mass activity (per Ru) than a low surface area crystalline RuO2, and with less Ru dissolution over 1000 cycles compared to an RuO2 with a similar surface area. Magnesium substitution provides the optimum balance between stability and activity, despite leaching of the Mg2+ into solution, and this was proved in membrane electrode assemblies.

Highlights

Ruthenium dioxide adopts the rutile structure and is a material that has been well-studied because of various functional properties, including its magnetism,[1] use in electrocatalysis,[2] and as a host for lithium insertion.[3]
Note that an excess of the divalent metal precursor was used in the synthesis and hydroxide byproduct removed by acid washing so the samples we have studied represent the highest substitution level by this synthesis method
This occurs to the greatest extent for Co, followed by Mg, and Cu, Zn, and Ni, which all show similar losses of the substituent metals. This mirrors the activity data already discussed, which may imply that the loss of the base metal creates an active form of Ru-rich oxide, or at least a surface enriched in Ru. This is a noteworthy observation, since the electrochemical activity of a number of ruthenium and iridium oxides has recently been linked to the leaching of base metal ions into solution to yield the catalytically active material, such as in Y2Ir2O7,43 A2Ru2O7 (A = Yb, Gd, Nd),[44] and SrRuO3.45,46 Of the materials that we have studied, CoxRu1−xO2 is the only one that shows an increase in activity at end of life, which is probably due to its increase in surface area, as seen from the change in capacitance (Supporting Information) from beginning to end of life

Summary

■ INTRODUCTION

Ruthenium dioxide adopts the rutile structure and is a material that has been well-studied because of various functional properties, including its magnetism,[1] use in electrocatalysis,[2] and as a host for lithium insertion.[3]. Powder X-ray diffraction of the zinc and magnesium peroxides revealed phase-pure materials adopting cubic structures with lattice parameters close to those reported in the literature (Supporting Information). In the case of the cobalt, nickel, and copper materials, there is no evidence that these are peroxides: powder X-ray diffraction shows the samples are poorly crystalline oxides For liquid cell electrochemical testing, inks were spray coated onto a Toray paper (hydrophobic gas diffusion layer 60) at 0.2 mg cm−2 loading, verified using X-ray fluorescence (XRF) measurements. These samples were injected directly into an ICP-MS (Agilent Technologies 7700 series) to obtain the concentration of metal leached into solution

■ RESULTS AND DISCUSSION

■ CONCLUSIONS

■ ACKNOWLEDGMENTS

■ REFERENCES