FIGURE 6 Proposed rection path and the reaction energy barrier diagrams: (a) Proposed reaction path of PPE hydrogenolysis. (b) Reaction energy barrier of PPE hydrogenolysis by Ni@N-C SAC and Ni@NC.
To explore the range of the Ni@N-C SAC catalyst, two more β-O-4 model compounds and one α-O-4 model compound were evaluated as the substrates. It is found that benzyl phenyl ether (α-O-4) and 2-phenylethyl phenyl ether (β-O-4) over the Ni@N-C SAC are nearly completely converted, providing the corresponding phenols and arenes in high yields (≥ 80%, Table S3, entries 1-2). It is known that β-O-4 dimeric model compound with Cγ-OH is structurally closer to the true lignin. Gratefully, such a model compound was also completely converted (Table S3, entry 3), affording guaiacol and 4-propylguaiacol in yields of 80% and 53%, respectively. These data demonstrated the versatility of Ni@N-C SAC in selective cleavage of the C-O bond in different lignin model compounds.
The recyclability of the Ni@N-C SAC was explored by carrying out the hydrogenolysis of PPE as listed in Figure 7a. After each run, the used catalyst was washed with ethanol following by vacuum drying. It was found that the conversion of PPE in Ni@N-C SAC-catalyzed hydrogenolysis remains unchanged after three runs. Characterizations of the used catalyst were conducted by XRD, XPS, HRTEM and aberration-corrected HAADF-STEM. There is no Ni metal diffraction peak in XRD pattern (Figure 7b). The binding energy of Ni 2p3/2 peak (855.13 eV) of used catalyst is almost the same as that of original Ni@N-C SAC (Figure 7c). No Ni clusters or particles are observed on the surface of the used catalysts (Figure 7d, e). Aberration-corrected HAADF-STEM characterizations show that Ni atoms were still atomically dispersed with no aggregation observed (Figure 7 f). These results demonstrate the excellent stability of Ni@N-C SAC in the hydrogenolysis systems.