X-ray absorption spectroscopy investigations into the stability and activity of fuel cell electrocatalysts

Abstract

Despite the advancements in fuel cell technology over the years, many fundamental electrocatalyst activity and stability challenges remain. In particular, poisoning of precious metal catalysts and durability over the lifetime of fuel cells are paramount. Many of the effects caused by catalyst poisoning or instability are exhibited in fundamental electrochemical assays in relatively short periods of time. However, the electrochemical methods by themselves do not identify the mechanism for which electrocatalysts decay or deactivate. This information can only come from spectroscopic investigations that probe the material at an atomic level. In this thesis, in situ x-ray absorption spectroscopy (XAS) is coupled with electrochemical methods to investigate these phenomena. With detailed knowledge of catalyst decay mechanism(s), more active, resistant catalysts can be made to propel fuel cell technology to become the power source of the future. This thesis begins by introducing the electrochemical and spectroscopic methods which will be largely employed in Chapters 2 - 6 to assess the stability and poisoning of model fuel cell compounds. Chapter 2 illustrates the well known phenomenon of halide poisoning by the delta-μ-XANES method, which had never been employed previously on this system. Platinum electrocatalysts are shown being poisoned by chloride ions by different methods at two different concentrations. Water activation is slowed down at 10-3 M chloride and almost completely stopped at 10-2 M. Also presented is the significant effect that halides (including Br and I) have on the typically facile hydrogen oxidation reaction as it is relevant to several systems. Chapters 3 and 4 deal with both catalysts employed in the direct methanol fuel cell, more specifically, platinum-ruthenium alloys in Chapter 3 and carbon supported platinum in Chapter 4. First in Chapter 3, the stability of today's state-of-the-art PtRu anode materials is tested and ruthenium dissolution is found to occur in two commercially available materials. Both materials undergo ruthenium dissolution by different mechanisms as shown by in situ XAS. The formation of ruthenium surface islands occurs differently for the two materials and the effects are discussed. The ruthenium ions that leach out of the electrocatalyst are observed causing changes to the microviscosity of the Nafion membrane and the cathode electrocatalyst in Chapter 4. Ruthenium was found to spontaneously deposit on the platinum at open circuit potential and block surface sites needed for efficient oxygen reduction. The ruthenium ions prefer to block three-fold platinum sites but are also observed in one-fold sites under potential control conditions albeit at much lower concentrations. Trivalent ruthenium ions have a significant effect on the structure of perflourinated membranes and will likely affect the ability to transport protons. In Chapter 5 the blue copper oxidase enzyme laccase (T. versicolor) is investigated for applications in biological fuel cells. Laccase with its full complement of 4 divalent copper ions in its active site is shown to be an effective enzyme for reducing dioxygen. However, the complexity of the active site has made a precise mechanism elusive. Both direct and mediated electron transfer phenomena are investigated by electrochemical methods and in situ XAS. Though the reaction occurs in a sub-second time scale it is shown that XAS can reveal reaction intermediates and a new proposed mechanism is shown. Finally, the concluding chapter presents a new class of material being examined for fuel cell catalysis that excludes noble metals, which is essential for fuel cell commercialization. The material shown is iron based and synthesized from a relatively inexpensive iron acetate or porphyrin precursor. The material has been shown to be quite active for oxygen reduction, however, the exact structure is not known. Through a preliminary ex situ XAS investigation, a proposed structure is shown to involve an Fe/N/C macrocyclic moiety incorporated directly onto the carbon support. Though the material undergoes deactivation in acid relatively quickly, it is much more stable in alkaline solution. It is also shown that the deactivated material can be re-activated by a reactivation pyrolysis. The initial structure determination suggests the material is mixed-phase requiring further study.