Hydrodesulfurization On Re Metal and Sulfide Catalysts: Effects of Catalyst Structure and Metal-Sulfur Binding Energy On Reactivity

Hydrodesulfurization on Re Metal and Sulfide Catalysts:

Effects of Catalyst Structure and Metal-Sulfur Binding Energy on Reactivity

Edwin Yik^a, Huamin Wang^b, and Enrique Iglesia^a,*

^aDepartment of Chemical and Biomolecular Engineering, University of California, Berkeley, CA 94720

^b Pacific Northwest National Laboratory, Energy and Efficiency Division, Richland, WA 99352

*iglesia@berkeley.edu

Recent mechanistic studies of thiophene hydrodesulfurization (HDS) [1,2] on silica-supported Ru and Pt metal clusters, which avoid the structural anisotropy of layered metal sulfides encountered in commercial Co(Ni)Mo(W) sulfides [3], and on the effects of their cluster size have shown that HDS turnover rates are lower on surfaces with higher S-metal binding energies, due to a lower availability of vacant sites required for reaction at a given sulfur chemical potential, which is set by varying H₂S:H₂ ratios [4]. We extend those studies to silica-supported Re metal clusters and layered Re sulfides (ReS_x) and observe that, despite differences in the catalyst identity and structure, HDS of thiophene occurs via elementary steps that are similar to those prevalent on Ru and Pt.  Therefore, the differences in turnover rates on these materials may be attributed to the inherent properties of the catalysts rather than to a change in the reaction mechanism. Reactivity decreases from Pt to Ru to Re metal clusters, consistent with an increase in the S-metal bond strength [5] and consequent decrease in the density of empty sites as a result of the higher S-coverage.

HDS turnover rates on ReS_x/SiO₂ samples were roughly 100 times higher than those on Re/SiO₂, even when the former were normalized by the total number of Re atoms (because transmission electron microscopy (TEM) images showed well-dispersed lamellar features) and the latter by its surface atoms. Both samples were synthesized from different treatments of a common ReO_x/SiO₂ precursor and remained kinetically stable during the course of the experiments (1-3 MPa of H₂, 1-10 kPa of thiophene, and 0.2-3.0 kPa of H₂S) over a period of more than 100 h. X-ray absorption (XAS), TEM, and S-contents (obtained through temperature programmed reduction (TPR) in H₂) suggested that Re metal retained its metallic bulk during HDS and exhibited an amount of sulfur corresponding to near-monolayer surface coverages after HDS. TEM images also showed that Re clusters did not form lamellar ReS_x structures and that ReS_x samples did not form isotropic Re clusters during HDS, further indicating an absence of interconversion between the two phases despite their exposure to identical experimental conditions. The marked differences in reactivity on surfaces of ReS_x and on those of S-covered Re metal likely reflect weaker S-Re bond strengths in the former [4], which lead to a concomitant increase in S-vacancy surface concentrations. In addition, the structural buckling that may exist in ReS_x could allow higher accessibility to the vacant sites on which the reaction may proceed. Thus, the inhibition of turnover rates as a result of increasing H₂S:H₂ ratios is more pronounced on Re/SiO₂ than on ReS_x/SiO₂.

References

1.      Wang, H., and Iglesia, E. J. Catal. 270, 245 (2010).

2.      Wang, H., and Iglesia, E. ChemCatChem. 3, 1166 (2011).

3.      Topsøe, H., Clausen, B.S., and Massoth F.E., in “Catalysis Science and Technology” (J.R. Anderson and M. Boudart, eds.) Vol. 11. Springer-Verlag, New York, 1996.

4.      Bartholomew, C.H., Agrawal, P.K., and Katzer, J.R. Adv. in Catalysis. 31, 135 (1982).

5.      Alfonso, D.R. J. Phys. Chem. C. 115, 17077 (2011).