AVS 53rd International Symposium
    Applied Surface Science Monday Sessions
       Session AS-MoA

Paper AS-MoA7
Structure Elucidation of Nano-Composite Catalysts by Multivariate Analysis and Regression Modeling of XPS Data

Monday, November 13, 2006, 4:00 pm, Room 2005

Session: Developing Methods for Data Analysis
Presenter: K. Artyushkova, University of New Mexico
Authors: K. Artyushkova, University of New Mexico
J.E. Fulghum, University of New Mexico
T.S. Olson, University of New Mexico
P. Atanassov, University of New Mexico
Correspondent: Click to Email
A new class of non-platinum electrocatalysts, based on pyrolysed porphyrins), was studied using X-ray photoelectron spectroscopy (XPS) in combination with multivariate analysis. Understanding the CoTMPP electrocatalyst structure in combination with surface medications occurring during pyrolysis requires identification of subtle changes in a very complex system. Deconvoluting numerous overlapping photoelectron peaks is a particular challenge in this case, as new species formed during pyrolysis may appear at the same binding energy as existing species. High-resolution spectra acquired from the precursor and electrocatalysts pyrolyzed at various experimental conditions were curve-fit using a) individual peaks of constrained width and shape as well as b) experimentally obtained photopeaks from the precursor and additional peaks required for a complete curve fit. Principal Component Analysis (PCA) was applied to quantitative results from the curve-fits of both types of spectra of pure elements, and various combinations, to identify types of species both formed and destroyed during the pyrolysis process and to find correlations between them. It was established that the catalyst presents a nano-composite of highly dispersed pyropolymer with remaining Nx-centers inserted in a graphite-like matrix. Approximately 50% of the metal is Co2+ associated with remaining N4-centers. The remaining cobalt is present in crystallites of metallic Co, coated with thin layer of CoO. The distribution of these types of moieties is directly related to the efficiency of oxygen reduction. A Spectra-to-property relationship was developed by applying multivariate regression models correlating XPS data with rotating-ring disk electrode (RRDE) data for the CoTMPP catalyst treated in various acids which selectively remove chemical moieties. These models have a potential for predicting the chemical composition of the CoTMPP catalyst which optimizes electrochemical performance.