Equilibrium concentrations of dissolved platinum species from a Pt/C electrocatalyst sample in 0.5 M H 2 SO 4 at 80°C were found to increase with applied potential from 0.9 to 1.1 V vs reversible hydrogen electrode. In addition, platinum surface area loss for a short-stack of proton exchange membrane fuel cells ͑PEMFCs͒ operated at open-circuit voltage ͑ϳ0.95 V͒ was shown to be higher than another operated under load ͑ϳ0.75 V͒. Both findings suggest that the formation of soluble platinum species ͑such as Pt 2+ ͒ plays an important role in platinum surface loss in PEMFC electrodes. As accelerated platinum surface area loss in the cathode ͑from 63 to 23 m 2 /g Pt in ϳ100 h͒ was observed upon potential cycling, a cycled membrane electrode assembly ͑MEA͒ cathode was examined in detail by incidence angle X-ray diffraction and transmission electron microscopy ͑TEM͒ to reveal processes responsible for observed platinum loss. In this study, TEM data and analyses of Pt/C catalyst and cross-sectional MEA cathode samples unambiguously confirmed that coarsening of platinum particles occurred via two different processes: ͑i͒ Ostwald ripening on carbon at the nanometer scale, which is responsible for platinum particle coarsening from ϳ3 to ϳ6 nm on carbon, and ͑ii͒ migration of soluble platinum species in the ionomer phase at the micrometer scale, chemical reduction of these species by crossover H 2 molecules, and precipitation of platinum particles in the cathode ionomer phase, which reduces the weight of platinum on carbon. It was estimated that each process contributed to ϳ50% of the overall platinum area loss of the potential cycled electrode.

In this paper, electrochemical impedance spectroscopy ͑EIS͒ is used to resolve various sources of polarization loss in a pure hydrogen-fueled polymer electrolyte fuel cell ͑PEFC͒. EIS data are fitted to a fuel cell model in which the catalyst layer physics are accurately represented by a transmission line model. Extracted parameters include cell ohmic resistance, catalyst layer electrolyte resistance, and double-layer capacitance. The results showed that the catalyst layer electrolyte resistance for a stateof-the-art electrode ͑47 wt % Pt on Vulcan XC-72 carbon, 0.8 Nafion ͑1100EW͒-to-carbon weight ratio, 13 µm thick͒ at 80°C and fully humidified conditions was approximately 100 m⍀-cm 2 ; this translates to a dc voltage loss of about 33 mV at a current density of 1 A/cm 2 . Similar results were obtained for two experimental methods, one using H 2 ͑anode͒ and O 2 ͑cathode gas feed͒ and another with H 2 and N 2 supplies, and for two cell active areas, 5 and 50 cm 2 . The measured catalyst layer electrolyte resistance increased with decreasing ionomer concentration in the electrode, as expected. We also observed that the real impedance measured at 1 kHz, often interpreted as the ohmic resistance in the cell, can include contributions from the electrolyte in the catalyst layer.

* These authors contributed equally to this work.The thousandfold increase in data-collection speed enabled by aberration-corrected optics allows us to overcome an electron microscopy paradox -how to obtain atomic-resolution chemical structure in individual nanoparticles, yet record a statistically significant sample from an inhomogeneous population. This allowed us to map hundreds of Pt-Co nanoparticles to show atomic-scale elemental distributions across different stages of the catalyst aging in a proton-exchange-membrane fuel cell, and relate Pt-shell thickness to treatment, particle size, surface orientation, and ordering. 11/28/11 8:49 PM 2 Bulk and reciprocal space measurements provide accurate ensemble averages of nanoparticle systems, yet in doing so lose the connections between microscopic degrees of freedom when integrating over the myriad of different particles in any representative sample.Out of necessity, nanoscale chemical imaging to date has relied on a handful of spectra collected from a few selected particles, as it often takes a few hours to record a spectral map of a single nanoparticle. However, nanoparticle systems-especially during electrocatalysis-are heterogeneous and have multiple competing processes running in parallel. Thus, identifying and quantifying dominant mechanisms requires statistics on scores to hundreds of particles in order to reliably connect the microstructure to the bulk properties. With the development of aberrationcorrected scanning transmission electron microscopy (STEM) 1, 2 and efficient electron energy loss spectra (EELS) collection systems, elemental concentrations and chemical bonding information can now be collected roughly a thousand times faster than on a conventional microscope, allowing rapid and reliable 2-D mapping of chemical distributions at atomic resolution 3 . While much of the focus of aberration correction has been on producing increasingly small sub-Angstrom electron beams, here we instead stop at an atomic-sized beam and increase the usable beam current. This enabled us to collect over one million EELS spectra and map out the concentrations of all atomic species in hundreds of Pt-Co nanoparticles used as fuel cell electrocataysts. We can thus quantify and correlate internal ordering, facet termination and surface structure-nanoparticle by nanoparticle-to identify the dominant degradation chemistries that limit the catalyst's efficiency. These measurements that would have taken years to record, and thus be too slow to provide feedback in a rapidly evolving field, were now collected in sessions of a few hours to days. ; a reduction in surface area for the remaining material as the average particle size increases; and a reduction in the specific activity from the remaining surface area as the particle composition and structure are altered 18,23 . While the first two issues are better understood and common to both Pt and Pt-M alloys, our focus here is on the final two issues which are determined by the less-well understood microscopic underpinnings and ...

In 1928, U.S. presidential candidate Herbert Hoover promised growing prosperity represented by “a chicken in every pot and two cars in every garage.” We now find ourselves at a point in history wondering if and when the power for those cars will come from fuel cells instead of internal combustion engines.

Oxygen Reduction Reaction (ORR) currents have been measured under the potential range of an operating fuel cell (0.72-0.9 Volts) while keeping proton and oxygen transport-related overpotentials insignificant by testing with pure O 2 , at 100%RH and at low current densities. Low potential points are achieved by reducing the platinum loading on the cathode and operating under sub-ambient (<101.3 kPa abs ) pressure. The resulting experimental data are fit to Pt-oxide-coverage-dependent kinetics and the kinetic parameters extracted. The impact of Pt-oxide on apparent Tafel-slope transition (i.e., a function of potential) and on ORR performance is discussed.Tremendous progress has been made in demonstrating required performance and durability of Polymer Electrolyte Membrane Fuel Cells (PEMFCs) for automotive application. The development focus is now shifting to cost reduction with a high emphasis on reducing the cathode platinum loading. Current demonstration levels typically are greater than 0.5 gm/kW which translates to >$3600 platinum/vehicle at today's Pt prices. Initial commercialization requirements are in the neighborhood of <0.1 gm/kW, whereas reaching <0.05 gm/kW would achieve the cost equivalent of the precious metal in modern catalytic convertor technology. The reaction rate for the hydrogen oxidation reaction (HOR) 1 on the anode is several orders of magnitude larger than that for the oxygen reduction reaction (ORR) 2 on the cathode. Hence very little Pt catalyst is needed to facilitate the anode reaction and the fuel cell kinetic losses are mainly controlled by the Pt loading used in the cathode. Over the years, there has been a significant amount of work focused on developing highly active ORR catalysts to reduce cathode Pt loading. Research effort on developing highly active cathode catalysts is incomplete, however, without a full mechanistic understanding of the ORR. In spite of decades of work, on oxygen reduction reaction using Pt and Pt alloys, there is still a disturbing lack of consensus on this mechanism.ORR Rates are conventionally represented using simple Tafel kinetics:The Tafel slope (slope of the plot of cell potential vs. the logarithm of kinetic current density) is given by 2.303RT αF in units of volts/decade, where α is the transfer coefficient. Rotating disk electrode [RDE] experiments for ORR have shown a transition in Tafel slope from low values (a transfer coefficient of α∼1) at high potentials >0.85 V to high values (a transfer coefficient of α∼0.5) at low potentials <0.85 V. [3][4][5] In these RDE experiments, however, the observed currents at low potential (or high current density) are convoluted by the diffusion limited currents. The diffusion and kinetic processes are in series such that the kinetic currents can only be obtained by correcting the observed currents for the diffusion limited currents -a process which can introduce significant uncertainty. A cleaner approach is to measure kinetics without limitations from mass transport. Microelectrode studies, where the O 2 mass tr...