Structure sensitivity in catalysis

Some fundamental concepts of catalysis are as of yet not fully explained, but are of paramount importance for the development of improved supported metal catalysts for chemical industries and environmental remediation. An example is the concept of structure sensitive, and structure insensitive reactions, where surface-normalized activity changes, or does not change, with catalyst metal nanoparticle size, respectively. We have explored this concept, and its relation to surface reconstruction on a range of well-defined silica-supported Ni metal nanoparticles (in the range of 1-6 nm) for a structure sensitive (i.e., CO₂ hydrogenation) [1, 2] and for a structure insensitive (i.e., ethene hydrogenation) [3] reaction. Using state-of-the-art techniques, inter alia in-situ high-resolution STEM, and quick-X-ray absorption spectroscopy with sub-second time resolution, we compare the fundamental concepts behind structure sensitivity and insensitivity and find that the geometric and electronic effects [1, 2] that we show to cause structure sensitivity make it unlikely for structure insensitivity to exist (while we do observe it empirically). We then postulate that restructuring of metal nanoparticles may play a role in the empirical observation of structure insensitivity, which we again study by a range of state-of-the-art techniques. In the case of the structure insensitive ethene hydrogenation reaction, such size-dependent effects as the decrease of the reversibility of adsorbate-induced restructuring and the increase of carbon diffusion with increasing particle size are observed (see Figure 1). For the structure sensitive CO₂ hydrogenation no such perturbation was observed. We show that this particle size dependent restructuring induced by ethene hydrogenation can make a structure sensitive reaction structure insensitive. We postulate here that structure insensitive reactions should rather be termed apparently structure insensitive, which changes our fundamental understanding of the age-old empirical observation of structure insensitivity.

Figure 1: Analysis of ethene-pulsed alternating feedstock experiments. For each Ni metal nanoparticle size, 50.000-70.000 spectra were individually fit, fixing the coordination number but varying σ² which can be taken as a descriptor for the degree of restructuring of the samples under study. The values have been given an arbitrary offset.

References

[1] Vogt, C.; Monai, M.; Kramer, G. J.; Weckhuysen, B. M.; Nature Catalysis, 2019, 2, 188-197.

[2] Vogt, C.; Groeneveld, E.; Kamsma, G.; Nachtegaal, M.; Lu, L.; Kiely, C.J.; Berben, P.H.; Meirer, F.; Weckhuysen, B.M.; Nature Catalysis, 2018, 1, 127–134.

[3] Vogt, C.; Groeneveld, E.; Monai, M.; Ferri, D.; van Santen, R. A.; Nachtegaal, M.; Unocic, R. R.; Frenkel, A. I.; Meirer, F.; Weckhuysen, B. M.; manuscript under review.