Hydrocarbon streams derived from natural gas and petroleum processing contain high concentrations of hydrogen sulfide (H2S) which is corrosive to process equipment, detrimental to product quality, and harmful to human health and the environment. Reaction of trace concentrations (<100 ppm) of sulfur compounds in natural gas streams with copper oxide (CuO) to form environmentally benign copper sulfides is an effective method to purify these streams. Such reactive sorption processes are advantageous because they are thermodynamically more favorable than physisorption, leading to higher purity streams and higher solid phase contaminant capacities. However, the primary drawbacks to reactive sorption systems are incomplete conversion of the copper oxide phase and slow kinetics, which detrimentally impact their industrial use by necessitating continuous process shutdowns to reload the reactors or/and over-sized beds. Moreover, most current predictive models lack the detail to fully capture the important reaction-diffusion phenomena associated with reactive sorption at the molecular level. Thus, the extrapolation of these models across a broad range of conditions is difficult. Motivated by these considerations, the overall goal of this project is to combine advanced characterization experiments at the atomic level with intrinsic reaction kinetic data at the reactor level to create microkinetic descriptions that relate reaction mechanisms of sulfur-containing molecules to the chemical and structural changes of CuO-based materials. This multi-scale research approach (atoms, crystals, pellets, reactor) exhibits a significant departure from traditional approaches that focus on either iterative materials screening or simplified reactor modeling.

This project has unraveled the following fundamental insights about the CuO-H2S reactions: i) CuO-based materials with decreasing extents of crystallinity contain high concentration of oxygen vacancies which results in higher extent of conversion due to the slower rates of molecule dissociation and increased diffusivities of sulfur atoms, oxygen atoms, and/or reactant molecules within the solid phases. ii) The Cu2-xSx products change over the course of the reaction due to changes in the electronic and coordination environments of the interfacial disulfide groups, with products resembling covellite (CuS) as the conversion increases. iii) The introduction of foreign atoms to CuO lattice (Zn and La atoms) enhances CuO sulfidation conversions by disrupting the CuO lattice and forming vacancies for the diffusion of sulfur and oxygen atoms, and by donating electron density to Cu2+ centers and consequently weakening the Cu-O bonds and making them more susceptible to rapture and reaction with sulfur moieties. Finally, the project highlights the utility of using fixed bed tests and simple linear driving force models to describe the surface reaction/diffusion phenomena contingent on operating at conditions where bulk and pore diffusion resistances are minimal. These insights, among others, are critical for the rational design of high capacity reactive sorbents and for the optimal operation of the desulfurization processes at conditions that approach their thermodynamic limits.