In situ and Operando X-ray Absorption Spectroscopy Investigations of Cathode Surface Reactions in Electrochemical Cells, Fuel Cells, and Batteries

X-ray Absorption Spectroscopy (XAS) has been increasingly used to study fuel cell catalysts through application of the Δ&mgr; XANES technique in combination with the more traditional X-ray Absorption Near Edge Structure (XANES) and Extended X-ray Absorption Fine Structure (EXAFS) techniques. Traditionally the most active and durable fuel cell catalyst has been Pt nanoparticles supported on carbon, and the Δ&mgr; technique is able to provide, under operating conditions, surface adsorbate coverages of relevant species such as H, OH, O and CO. This dissertation expands the applicability of the Δ&mgr; technique to very different systems, namely to oxide supports inserted in the fuel cell catalysts, as well as to other energy relevant materials such as the cathode in a battery. The lithium iron phosphate battery cathode was studied during operando cycling conditions to better understand the two phase intercalation mechanism with either a Li metal or a carbon anode. In the cathode, bulk electronic properties rather than adsorption properties are more important, requiring an entirely new Δ&mgr; interpretive approach.;A critical requirement for mass production of fuel cell powered automobiles is to develop more active, durable, and less expensive catalysts than the current standard of Pt nanoparticles supported on Vulcan carbon. This dissertation evaluates Pt nanoparticles supported on carbon but with a thin conducting layer of tantalum or niobium oxide or oxyphosphate inserted between the Pt nanoparticles and the C support utilized for the oxygen reduction reaction (ORR). This was studied by following both the XAS Pt L2/3 and tantalum/niobium K edges. The placement of Pt nanoparticles on tantalum oxide or oxyphosphate, supported on Vulcan carbon, alone does not improve the catalyst activity for the ORR but it does improve the durability. Further, high temperature heat treatment changes the way the oxyphosphate groups are associated around the Pt nanoparticles. Heating caused polyphosphate chains to form, some in direct contact with the Pt. Facile proton conduction along the chains helps remove OH poisons from the Pt surface, now increasing the ORR activity. These findings were established with the Δ&mgr; XANES and EXAFS techniques, but are fully corroborated with high-angle annular dark-field scanning transmission electron microscopy (HAADF) with high resolution energy-dispersive x-ray spectroscopy (EDS) capabilities.;In contrast to tantalum, the insertion of niobium oxide or oxyphosphate between the Pt and supported Vulcan carbon has completely different effects on the Pt surface and the ORR. There is very little improvement in the ORR activity, but a very interesting variable metal support interaction (MSI) is induced by an oxygen spillover/reverse spillover phenomenon. When oxygen spills over from Pt to the niobium support, the resulting covalent character of the niobium support weakens the Pt-O bond. Reverse spillover restores the ionic MSI from the niobium support and once again restores the Pt-O bond strength.;The second part of this dissertation focuses on lithium ion batteries, in particular lithium iron phosphate, a relatively new cathode material, shown to be safer, more environmentally friendly than lithium cobalt oxide and allowing many charge/discharge cycles. However several conflicting models for the two-phase transition mechanism during lithium (de)intercalation have been proposed. Application of the Δ&mgr; XANES technique, in combination with traditional EXAFS, allowed for the electronic structure and structural disorder to be tracked with changing lithium content during real battery cycling conditions. The structural disorder can be followed by the EXAFS (Debye-Waller factor), which is dominated by the extent of disorder in the transition region between the two solid solution end phases. The behavior with charge/discharge suggests either of two previously proposed models, the "collective mosaic" or the "sequential" unrelaxed single-phase kinetic model. Finally the Δ&mgr; data shows spectroscopically for the first time how the charge/discharge mechanism within a LiFePO4 cathode proceeds differently depending on the anode. With the Li anode the larger particles proceed first, but with the C anode the smaller particles proceed first. The nature of the anode determines the Li-ion chemical potential or concentration in the electrolyte, and this causes different rates of inter-particle vs. electrolyte-particle Li transfer.