(224c) The Effect of Physiochemical Properties of Copper Oxide Sorbents on the Chemisorptive Removal of Sulfur | AIChE

(224c) The Effect of Physiochemical Properties of Copper Oxide Sorbents on the Chemisorptive Removal of Sulfur


Lopez-Ausens, T., University of California Los Angeles
Ghosh, R., UCLA
Alshafei, F., UCLA
The removal of sulfur contaminants from gaseous streams is an application relevant to many industries due to sulfur’s deleterious effects on the environment, public health, product specifications, and process equipment. The importance of this application justifies the continuous effort to develop technologies that enhance the efficiency of removal and the safe sequestration of contaminants. Chemical reaction of sulfur molecules with copper oxide (CuO) is an attractive purification strategy because the reactions are thermodynamically favorable at low temperature and the sulfur is sequestered in a benign copper sulfide (CuS) form. Previous research efforts have focused on evaluating fixed beds of sorbents to deduce bulk kinetics based on the reaction-diffusion phenomena at the pellet scale. In this work, we aim to establish a more comprehensive understanding of the process which relates sorbents physiochemical properties and reaction conditions to the removal capacity and kinetics at both atom and reactor scales. The study implements an unprecedented synergistic approach that combines computational modeling, fixed bed sorption experiments, and advanced synchrotron-based characterization techniques. The preliminary results of this work indicate that crystallite size is the most critical factor in determining the performance of pure unsupported CuO sorbents. It also shows that the introduction of foreign metal atoms (zinc and lanthanum) enhances the performance of CuO sorbents while the presence of trace carbon residues deteriorates it significantly.

The effect of microscopic size (crystallite size) and intraparticle void space on the removal capacity was explored by testing beds of CuO sorbents that are produced through a range of synthesis techniques, treatment conditions and starting precursors. Variations in preparation parameters yielded materials of different morphologies (spherical particles, flowerlike particles, nanofibers, and nanobelts) and a wide range of crystallite sizes (5-74 nm). Removal capacities of particles, belts and flowerlike plates showed an increasing trend, from 0.5 wt.% to 12.0 wt.%, with the decrease in crystallite size from 26 to 5 nm. A possible explanation, supported by DFT calculations, suggests that decreasing crystallite sizes exposes more reactive surface facets in CuO and increases O-vacancies. Open facets facilitate H-S bond cleavage which is necessary to produce S atoms that react with CuO moieties. It also enhances the diffusion of sulfur atoms to fresh substrate layers allowing for a deeper propagation of the reaction.

Nevertheless, this trend of increasing capacity with crystallite size decrease didn’t apply to nanofibrous materials despite their enhanced macro-porosity. For example, nanofibers (14 nm) crystallite size achieved a removal capacity of 2.2 wt.% compared to 7.3 wt.% achieved by flower-like particles (13 nm). XPS measurements confirmed the presence of trace amounts of carbon residues even after thermal treatment at elevated temperature (823 K). To confirm the role of carbon in dampening sulfur removal performance, different amounts of a polymeric surfactant (PVP) were added to the nanoparticles’ co-precipitation recipe. The PVP-particles performed consistently worse than their pure counterparts (2.7 wt.% versus 5.8 wt.% at 18 nm) confirming the negative effects of carbon contamination.

Moreover, the effect of doping of CuO nanoparticles with different metals (Zn, Ni, La, Co, Mn, Bi, Mg, and Al) was studied at different metal to copper ratios in sol-gel salt precursors (1%, 10%, 33%, 50%). At 1% metal to copper ratio, sorbents achieved comparable capacities (5.2-7.2 wt.%) and exhibited similar crystallite sizes (6.9-8.2 nm). At higher mixing ratio of 10%, the difference between dopants becomes more pronounced with CuO-La achieving the highest capacity of 26.7 wt.% at 3.3 nm and 11.7 wt.% at 8 nm followed by CuO-Zn (4.8 wt.% at 5.3 nm). The rest of dopants achieved negligible capacities ranging from 0.5 wt.% (CuO-Mg) to 2.7 wt.% (CuO-Co) within the same crystallite size range. At 33%, CuO-La sorbent achieved 33.5 wt.% which is 80% of stoichiometric saturation capacity followed by CuO-Zn at 15.7 wt.%.

Introducing a foreign atom to CuO lattice can influence the reactivity of CuO towards sulfur by disrupting the lattice structure which manifests as an increase in oxygen vacancies and enhancement in diffusivity of oxygen and sulfur atoms. It can also alter the electronic structure by changing the density of states at the fermi level which affects oxygen reactivity (and cations reactivity). A preliminary interpretation of these results suggests that Lanthanum causes the largest disruption to the CuO lattice (1.5 radii ratio of La+3+/Cu+2) compared to the other metals with cation radii ratio ranging from 0.6 for Mn+2 to1.0 for Zn+2. Advanced characterization (X-ray absorption spectrometry at Cu K, S K and La L edges) will be performed to further understand the difference in structural parameters between doped sorbents at different ratios and the difference in the formed CuySx species in the product phases. This knowledge of structure-function relationship can serve as a guide for rational design of sorbents instead of the common trial-and-error approach.