(622a) Use of in-Situ XAS and TXM Techniques for the Simultaneous Determination of Reaction Kinetics and Structural Evolution of CuO during Sulfidation Reactions.

Natural gas and oil-based streams contain a range of sulfur contaminants that require removal to meet environmental regulations and product specifications. One attractive yet not fully studied technique involves the reaction of copper oxide (CuO) with hydrogen sulfide (H₂S) to form copper sulfide. This reaction is thermodynamically favorable and results in environmentally benign products. However, the performance of CuO sorbents can be extremely sensitive to changes in the structure of CuO and process conditions. Developing fundamental insight for this gas-solid reaction requires detailed knowledge of the atomistic characteristics of the solid and how these characteristics influence reaction mechanism and kinetics. In this work, we present the use of fixed bed reactors, X-ray absorption spectroscopy (XAS), and transmission X-ray microscopy (TXM) to simultaneously probe the fundamental kinetics of the reaction of CuO with H2S to form CuS, and thereby probe spatial-temporal chemical and structural changes of copper during this reaction.

In-situ XAS experiments at the Cu K-edge were performed at the Stanford Synchrotron Radiation Laboratory using a custom designed reactor in which samples of CuO were reacted in a gas flow of 1000 ppm-vol H₂S in helium under controlled conditions. Analysis of this XAS data produced quantitative uptake kinetics under gradientless conditions. In addition, uptake kinetics were determined from fixed bed experiments carried out under a flow of 1000 ppm-vol H₂S in nitrogen. These experiments were conducted using CuO nanoparticles with crystallites ranging in size from 7 to 47 nm and using a commercial CuO-based material containg CuO (63.5 wt%), ZnO (25 wt%), and Al₂O₃ (10 wt%).

H₂S removal reaction kinetics show similar trends in both fixed bed reactors and in individual particles of 10-20 µm diameter. Experimentally determined linear driving force model reaction rate parameters derived from plug-flow reactors were a factor of 10 larger for the CuO nanoparticles (4.3x10^-4 s^-1) compared to the commercial material (4.0x10^-5 s^-1), however, the commercial material exhibited higher conversion to CuS (44% compared to 14%). Conversion of CuO was also measured using in-situ XAS of a fixed bed of material and using in-situ TXM-XAS of individual particles. Analysis of XAS data revealed similar conversions of CuO nanoparticles (13%) and commercial samples (38%) compared to plug-flow experiments, however, random pore model reaction rate parameters were higher for the nanoparticle sample (8.4 x 10^-3 cm⁴ mol^-1 s^-1) than for the commercial sample (5.6 x 10^-3 cm⁴ mol^-1 s^-1). Analysis of TXM images also revealed that reaction fronts proceed through the entire diameter of particles heterogeneously, indicating the presence of pore diffusion resistance even at very small length scales. Furthermore, values for the reaction rate constant decrease as the radial distance from the particle surface to the center increase. These differences in reaction kinetics and conversion indicate the critical impact of possible atomic scale differences and the formation of different copper sulfide products. Furthermore, the changes in rate parameter with distance suggest that a single reaction assumption (i.e., formation of only CuS) is insufficient to completely describe the reaction kinetics. Thus, detailed characterization of all possible Cu_xS_y products is required to construct a detailed mechanistic model that incorporates the possible formation of these products.

In-situ XAS was also used to probe temperature change effects. The bulk removal capacity doubled as temperature increased from 323 K to 353 K (9.5 wt% to 21 wt% and 12 wt% to 22 wt% based on XAS and fixed bed results, respectively). Sulfur uptake was also observed to be inversely proportional to crystallite size. Removal capacity of 12 wt% was achieved with 7 nm CuO nanoparticles compared to 5 wt% for 47 nm crystallite sizes. Moreover, the analysis of XAFS data showed that as temperature increases, the product copper sulfide phase resembled copper (II) sulfide rather than copper (I) sulfide. This phase change with temperature can be used to elucidate mechanistic aspects of the reaction that cannot be accomplished via conventional techniques.