(555d) Computational Study of Selective Oxidation of Ethane on Oligomeric VOx/SiO2 and Bulk Mixed Oxide Catalysts

Yilang Liu¹ and Prashant Deshlahra¹*

¹Tufts University, Medford, MA 02155 (USA)

*prashant.deshlahra@tufts.edu

Introduction

Selective oxidations of light alkanes are promising routes to produce more valuable chemicals.¹ However, selectivity to the desired products are limited by parallel and secondary reactions leading to undesired fully oxidized products. Primary alkane activations are typically limited by C-H activation, but mechanisms of parallel and secondary reactions are still unclear, while they are important for selectivity limitation. Here, to understand reaction steps limiting selectivity, we use hybrid-DFT calculations to probe all plausible paths of C₂H₆ oxidation reactions on different sites of VO_x/SiO₂catalyst.

Materials and Methods

To model oligomeric VO_x/SiO₂, silsesquioxane clusters (H₈Si₈O₁₂) with four of its SiO_x moieties replaced by VO_x are used. (Figure 1)^2,3DFT calculations were performed using Gaussian09 program with B3LYP functional and 6-31g(d,p) basis sets. Transition states were calculated using quadratic synchronous transit method. The optimized structures are characterized as minima points for intermediates and saddle points for transition states by frequency calculations. Intrinsic reaction coordinates (IRC) were calculated to confirm that transition states connect reactants and products. Activation energy barriers are calculated as differences between electronic energies of relevant intermediates or transition states to gaseous reactants and bare sites, and will be converted to free energies to determine reactivity trends at realistic temperatures and pressures and the effect of van der Waals interactions will be probed.

Results and Discussion

C₂H₆ oxidative dehydrogenation (ODH) on metal oxides occurs via Mars-van Krevelen redox cycles in which the C-H activation at lattice oxygens of oxides (O*) leads to reduced catalytic centers in the form of surface hydroxyls (OH*) or oxygen vacancies (*) (Figure1a). Re-oxidation of isolated reduced centers by O₂can lead to peroxo species (OO*). Therefore, plausible reaction paths for primary, parallel and secondary reactions at both types of species and steps accounting for the formation and consumption of these species were considered (Figure 1).

Figure 1. (a) Lattice oxygens (O*), vacancies (*), and peroxo species (OO*) involved in reduction-oxidation cycles of metal oxides during ethane-O₂ reactions. DFT derived structures and the metal-oxygen bond distances for (b) oxo (O*) and (c) peroxo (OO*) sites on VO_x/SiO₂.

On oxo site, the primary C₂H₆ ODH to form C₂H₄ is accomplished by two C-H activations on terminal O-atoms, with the highest barrier of 186 kJ/mol. C₂H₆ can also react in parallel to form C₂H₅OH via bonding between C₂H₅ radical and HO* after first C-H activation, with an activation barrier of 180 kJ/mol. C₂H₅OH can then be oxidized to CH₃CHO with activation barrier of 125 kJ/mol. The first step in secondary reactions after formation of C₂H₄ involves O-insertion to form C₂H₄O, which has a much lower activation barrier of 132 kJ/mol. Thus, both parallel and secondary paths have comparable or lower barriers and can contribute to C₂H₄ selectivity limitation. The removal of bound OH* or products of O-insertion is assisted by O₂ to form peroxo species in rapid steps with low barriers. On peroxo site, first C-H activation to form C₂H₄ or C₂H₅OH has a barrier of 122 kJ/mol while C₂H₄ can undergo O-insertion even more preferentially, with a 55 kJ/mol barrier. The calculated activation energies suggest that peorxo sites are more selective to undesired O-insertions than oxo sites, consistent with known tendency of peroxo sites to favor combustion reactions.⁴These steps lead to oxidized products and oxo sites to complete the catalytic turnovers. The overall reaction barriers for highest points along the reaction coordinate for primary, parallel and secondary reactions are listed in Table 1.

Table 1. Electronic activation energies relative to bare sites and gaseous reactants for ethane and ethylene oxidation at oxo and peroxo sites.

We further studied subsequent reactions of C₂H₄O on oxo sites. The most favorable reaction paths involve isomerization of C₂H₄O to form CH₃CHO, with an activation barrier of 103 kJ/mol. The CH₃CHO molecules bind strongly on oxo sites to form stable and relatively unreactive acetate species with 101 kJ/mol energy barrier and a -117 kJ/mol intermediate energy relative to gas phase CH₃CHO. On acetate covered surfaces, bridge O-atoms are more reactive than terminal O-atoms for C-H activation and oxygen insertion. Activation barriers of C₂H₄ formation and secondary oxidation of C₂H₄ to form C₂H₄O are 218 kJ/mol and 217 kJ/mol, respectively, shifted toward favoring C-H activation relative to oxo site.

These results lead us to conclude that selectivity to primary products is determined by O-insertion reactions parallel and sequential to desired C-H activations on oxo, peroxo and acetate covered surfaces. Preliminary computations of such reaction steps on microporous MoVTeNb mixed metal oxides suggest that undesired O-insertion reactions can be suppressed by steric hindrance in their small micropores while stabilizing the C-H activations by van der Waals interactions. Detailed comparisons of reaction steps in the two oxides help understand the role of electronic properties and size restrictions in determining selectivity limitations in oxidative conversion of small molecules.

Acknowledgements

Financial support from Tufts University and access to computing resources from Tufts Research Computing Cluster and the Extreme Science and Engineering Discovery Environment (XSEDE; NSF ACI-1053575; Proposal number TG-CTS150005) are gratefully acknowledged.

References

1. Labinger, J. A. and Bercaw J. E. Nature, 2002, 417, 507

2. Döbler, J., Pritzsche, M., and Sauer, J. J. Am. Chem. Soc., 2005, 127, 10861

3. Goodrow, A., and Bell, A. T. J. Phys. Chem. C, 2007, 111, 14753

4. Liu, J., Mohamed, F., Sauer, J. J. Catal., 2014, 317, 75