(385e) Fundamentals of the Removal of Sulfur Compounds from Gaseous Streams Via Reactive Sorption with Copper Oxide

Hydrocarbon streams derived from natural gas and oil processing are often contaminated by sulfur containing compounds, such as hydrogen sulfide and alkyl-mercaptans, which affect the quality and specifications of the product streams, pose a corrosion hazard in transfer pipelines and equipment, and conflict with health and environmental regulations. Several removal techniques have been studied and implemented in industrial purification processes ranging from gas-alkaline liquid absorption and physisorption on molecular sieve materials. Chemical reaction of sulfur molecules with copper oxide is an attractive purification strategy (especially for dilute concentrations) because the reactions are thermodynamically favorable at low temperature and the sulfur is sequestered as solid Cu-sulfur products. This study aims to elucidate the fundamentals and mechanistic details of these gas-solid reactions. Specifically, the results from this study led to semi-empirical correlations between reaction conditions (contaminant concentration, temperature, and presence of other contaminants) and the reaction rates and saturation capacities of the solid phase. Moreover, the effect of structural differences resulting from different preparation methods and starting precursors on the performance of copper oxide adsorbents in reactive sorption experiments was investigated to develop a structure-function relationship that can help in future design and tuning of adsorbent properties. A combination of thermodynamic calculations, fixed bed sorption experiments, and advanced characterization techniques are used to achieve this goal.

Apparent H₂S uptake kinetics and contaminant capacities were determined from contaminant breakthrough profiles collected from fixed beds of sorbents. The breakthrough curves were analyzed using irreversible adsorption models based on linear driving force approximations. This assumption leads to an analytical solution to the contaminant mass balance that consists of two parameters: the saturation concentration of contaminant in solid phase and an apparent sorption rate parameter. The rate parameter can be de-convoluted into contributions from bulk diffusion, pore diffusion, and reaction-diffusion phenomena within the solid. The effect of agglomerate size was probed using a commercially produced Cu-Zn-Al oxide material. As the agglomerate size increased from 600 to 1000 microns, the sorption rate parameter decreased from 1.04x10^-4 s^-1 to 6.05x10^-5 s^-1. The value of the sorption parameter for agglomerates larger than 550 microns were only weakly dependent on temperature and pressure and were in close agreement with theoretically calculated values of the contribution from diffusion within the agglomerate macropores. For agglomerate sizes less than 500 microns, the rate parameter is less sensitive to decreasing agglomerate size, and the rate parameter eventually becomes independent of particle size at 196 microns. The results from these agglomerate size studies indicate that bulk and pore diffusion are the dominant resistances to H₂S uptake for agglomerates that are larger than ~500 microns, however, as agglomerate size decreases, bulk and pore diffusion resistances become negligible compared to reaction-diffusion phenomena within the copper oxide and copper sulfide phases.

The effects of varying reaction conditions (temperature, pressure, inlet concentration of H₂S) on the rate parameter were also studied. Varying the total pressure (1-2 atm) while maintaining constant temperature and inlet concentration had negligible effect on the rate parameter while increasing the inlet concentration (930-2200 ppm) of H₂S led to an increase in the parameter from 3.74x10^-5 s^-1 to 1.1x10^-4 s^-1. Semi-empirical relations between concentration and the reaction rate parameter were developed. The rate parameter has an apparent 0.5 order with respect to inlet H₂S concentration. The saturation capacity remained constant at an average of 12.8 wt% (g of H₂S/100 g of sorbent) over a concentration range of 930-2200 ppm. Moreover, increasing temperature (21-150°C) increased the saturation capacity from 11 wt% to 42 wt%. The rate parameter exhibits an Arrhenius-type dependency on temperature with a pseudo-activation energy of 18 kJ/mol.

To test the applicability of linear driving force models for the reactive sorption of other sulfur containing compounds, reactive sorption experiments were conducted using methyl and ethyl mercaptan at varying temperatures (21-200°C) on the commercially produced Cu-Zn-Al material. Unlike the CuO-H₂S reactions that yielded the same products throughout the experiments’ duration and with varying conditions, CuO-mercaptan experiments resulted in various products. Examining the concentration profiles of different species identified the dominant reaction/s under given conditions. At room temperature, methyl and ethyl mercaptans were physically adsorbed by the aluminum base of the materials and breakthrough happened fast with no other sulfur containing compounds detected in effluent streams which suggests that, although thermodynamically favored, the rate of formation of dimethyl and diethyl sulfide at room temperature is too low for any reaction to be detected. At 120°C the methyl mercaptan reactive sorption experiment initially produces water and a high molecular weight sulfur containing compound which indicates the reaction: 2CH₃SH + 2 CuO à Cu₂O +H₂O + CH₃S-SCH₃. Over time, the water concentration in the effluent decreases with a corresponding increase in CH₃SCH₃ and methyl mercaptan which suggests that the reaction 2CH₃SH + CuO à CH₃SCH₃ + CuS is taking place at a low rate. The spent sample of this experiment showed a color change from black to orange which indicates the formation of Cu₂O via 2CH₃SH + 2 CuO à Cu₂O +H₂O + CH₃S-SCH₃. Increasing temperature to 200°C yielded CH₃SCH₃ and CH₃SH throughout the reaction duration which suggests that at 200°C the 2CH₃SH + 2 CuO à Cu₂O +H₂O + CH₃S-SCH₃reaction pathway is both kinetically and thermodynamically favored.

Fresh and spent nanostructures were analyzed via XRD and SEM to identify composition, crystallite agglomerate size, and crystallite size (using the Sherrer Equation). XRD patterns of fresh CuO nanostructures showed the characteristic peaks of tenorite, the monoclinic structure of CuO, for both nanofibers and nanoparticles regardless of precursor. Furthermore, nanofibers and nanoparticles consisted of CuO crystallites of similar size (16-18 nm). XRD analysis of the commercial materials also indicated monoclinic CuO with ~20 nm crystallites. SEM images revealed that the nanofibers exist as approximately rectangular solid shaped agglomerates of crystallites that are 250-300 nm in size and arranged in a one-dimensional stacking pattern. Nanoparticles and commercial Cu-Zn-Al oxide also consisted of 250-300 nm sized crystallite agglomerates, however, these agglomerates are packed into sphere like structures that are 5-10 microns in size.

To further probe the reaction between H₂S and CuO at the structural/morphological level, H₂S removal using CuO nanofibers and nanoparticles was compared. Nanofibers (230 nm diameter) and nanoparticles exhibited similar maximum capacities (2.4 wt%), however, sorption rate parameters for nanofibers (3.0 x 10^-3 s^-1) were 1.7 times larger than those for nanoparticles (1.7 x 10^-3 s^-1). This faster sorption rate is likely the result of the higher intra-agglomerate void fraction in the nanofibers compared to the nanoparticles. Moreover, the effects of changing nanofiber diameter on sorption rate and saturation capacity was studied. It was found that an increase of diameter (230-600 nm) resulted in a drop in the saturation capacity (2.4-0.3 wt%) and a decrease in the sorption rate parameter from 3.0 x 10^-3 s^-1 to 1.4 x 10^-4 s^-1. This decrease can be explained by the decrease in the intra-agglomerate void space which decreases accessibility to the reactive domains. The temperature effects were probed in CuO nanofibers and nanoparticles in similar fashion to the commercial Cu-Zn-Al. Increasing the temperature from 21°C to 200°C increased the saturation capacity for both 600 nm nanofibers (from 0.3 wt% to 38.8 wt%) and CuO nanoparticles (from 2.4 wt% to 40 wt%) very close to the stoichiometric limit of 42 wt% for complete conversion of CuO to CuS. Moreover, the concentration effects on rate parameter and saturation capacity were probed in CuO nanoparticles. In the range of 250-1000 ppm, the increase in inlet concentration increased the rate parameter from 6.95x10^-5 to 1.6x10^-3while for higher concentrations (1000-2000 ppm) the dependency becomes weaker and the rate parameter eventually becomes invariant to concentration.

The results from these studies indicate that equations derived from linear driving force approximations can be used as predictive models for the removal of sulfur compounds via reactive sorption with CuO. Mass transfer is the predominant resistance to contaminant removal for large particles of sorbent material, and thus, reaction rate parameters for linear force driving models can be calculated from theoretical relationships. Reaction-diffusion phenomena within the solid are more complex and require semi-empirical relationships to determine rate parameters for linear driving force models. Studies of CuO from different synthesis methods reveals that the fundamental kinetics and mechanism of these gas-solid reactions are highly sensitive to the nanostructure of CuO.