(679j) Understanding Facet Dependent Selectivity Trends in Catalytic Hydrogenation and Hydrodeoxygenation of Lignin-Derived Aromatic Compounds

Transition metal catalysts are observed to be effective for the hydrogenation and/or hydrodeoxygenation (HDO) of lignin-derived aromatic platform molecules. [1–5] For example, guaiacol is effectively hydrogenated over Pd/C to 2-methoxycyclohexanone and 2-methoxycyclohexanol. Similarly, cinnamaldehyde (CAL), a representative of lignin derived aldehydes, is hydrogenated to hydrocinnamaldehyde (HCAL) and hydrocinnamylalcohol (HCOL) over Pd/C catalyst. However, obtaining the desired selectivity in such reactions is often challenging. It is therefore fundamental to understand the interactions of these molecules with the catalyst particles. Herein, hydrogenation reactions of CAL and guaiacol over synthesized spheres, octahedra Pd (111) and cubes Pd (100) are explored. Density functional theory (DFT) simulations are employed to capture the energetics of the possible reaction routes to unravel pathways leading to specific reaction products.

Hydrogenation of CAL is performed using Pd octahedra (26 nm) and Pd cubes (22 nm), and Pd spheres (28 nm). Spherical particles of Pd provide 100% conversion at 303 K and 5 bar H₂ in 1 hr with 64% selectivity to HCAL, which is formed by the selective activation of the C=C bond on the CAL side chain. The fully hydrogenated product HCOL (hydrogenating both C=C as and C=O bonds) is observed to increase slowly with reaction time. Similarly, Pd octahedra particles show 100% conversion in 45 min with high selectivity to HCAL, which attains a maximum of 84% in 6h (Figure 1A). Simultaneously, the selectivity to HCOL is intriguingly low (Figure 1A). Even with prolonged reaction time, the selectivity remained unaltered. In contrast, on Pd cubes, the reaction is slow that it reached 100% in 120 mins, and the selectivity to HCOL is observed to be high (65%) over HCAL (26%) (Figure 1B). A detailed mechanistic study is therefore carried out using DFT simulations to understand the observed shift in selectivities. On both surfaces, CAL prefers to bind in a flat configuration through the aromatic ring as well as the side chain (Figure 1C). CAL is observed to bind more strongly on the Pd (100) surface (-311 kJ/mol) than on the Pd (111) surface (-231 kJ/mol), which in turn affects the bond activation towards hydrogenation. The activation of the C=C bond is found to be favorable on Pd (111) surface (activation barriers for the stepwise hydrogenation of C=C are, E_a1=51 kJ/mol, and E_a2=58 kJ/mol, Figure 1C). However, from the experiments, its is observed that further hydrogenation of HCAL to HCOL on the Pd (111) surface is not feasible. In contrast, on Pd (100) surface, the C=O bond hydrogenation is preferred over C=C hydrogenation, and the cinnamylalcohol (COL) intermediate easily undergo hydrogenation to form HCOL. Therefore, we may conclude that the HCOL formation is mainly from the Pd (100) facet, and the overall selectivity to HCOL observed on the spherical particle is primarily determined by the Pd (100) sites. In addition, the overall selectivity to HCAL observed on the spherical particle originates from the Pd (111) surface.

The HDO of guaiacol is examined over Pd (111) and Pd (100) surfaces. It is observed to interact via the aromatic ring in a flat configuration, with binding energy -176 kJ/mol on Pd (111) and -205 kJ/mol on Pd (100) surfaces, respectively. The barriers for the dissociation reactions (C-O, C-H, O-H, OH-Ph, OCH₃, CH₃) are compared with the hydrogenation barriers, out of which O-H dissociation (82-85 kJ/mol) is observed to be most favorable irrespective of the facets (Figures 1D,E). Interestingly, the direct deoxygenation routes are accompanied by high activation barriers on Pd (111) surface (E_a=258 kJ/mol) and Pd (100) surface (E_a=212 kJ/mol) (Figures 1D,F). Therefore, the HDO of guaiacol on Pd surfaces may essentially proceed via hydrogenation of the aromatic ring. Excitingly, the experimental results show high selectivity to the hydrogenated products, which are in agreement with the theoretical findings.

References

1. Porwal, G., Gupta, S., Sreedhala, S., Elizabeth, J., Khan, T. S., Haider, M. A., Vinod, C. P, ACS Sustainable Chem. Eng 7, 17126, (2019).
2. Gupta, S., Alam, Md. I., Khan, T. S., Haider, M. A ACS Sustainable Chem. Eng 7, 10165, (2019)
3. Gupta, S., Khan, T. S., Saha, B., Haider, M. A Eng. Chem. Res 58, 16153, (2019).
4. Jalid, F., Khan, T. S., Mir, F. Q., Haider, M. A Catal, 353, 265, (2017).