(731d) Direct Conversion of Ethylene to Propylene On Ni/MCM-41 – Insights Into Catalyst Structure

Propylene demand – mainly driven by polypropylene production – has been steadily growing in the past. The supply of light olefins like propylene is traditionally provided by either catalytic or steam cracking processes. These will, however, fail in the long run to cope with the changed demand situation since they have been historically designed to maximize ethylene output [1] and cannot be adjusted at will towards higher C₃ yields. Therefore, significant research activities have been aimed at the development of on-purpose propylene production processes.

One of the possible approaches in this direction is the recently found conversion of ethylene (or ethanol) to propylene on Ni/MCM-41 [2]. It has been proposed that the net transformation 3 C₂H₄ → 2 C₃H₆ proceeds via a reaction cascade comprised of ethylene dimerization, butene isomerization and finally cross-metathesis between ethylene and 2-butenes. A particularly fascinating aspect is the hypothesis of nickel-catalyzed olefin metathesis since nickel has not been found active for this kind of reaction in the past. One of the open issues for this catalytic system is the chemical identity of the nickel compound on the catalyst surface.

Catalysts with varying nickel contents (0.46 – 5.72 wt%) have been prepared following the template-ion exchange approach reported by Iwamoto and Kosugi [2]. The catalyst sample loaded with 3.18 wt% nickel gave a maximal ethylene conversion of 82 % combined with a propylene selectivity of 51 % (process conditions: packed-bed reactor, feed: 10 % ethylene in nitrogen, W/F = 0.7 kg h/m³, 400 °C, atmospheric pressure). This being said, we will focus in the present contribution on investigations regarding the nature of the supported nickel phase as well as the catalyst structure. Different techniques were used to elucidate the state of nickel on the catalyst. Increasing nickel loadings resulted in the formation of a lamellar phase at the expense of the mesoporous support as evidenced by nitrogen physisorption and XRD. TEM micrographs showed irregularly folded and crumpled sheets covering the MCM-41 particles. These foil-like structures could either present some kind of nickel phyllosilicate or turbostratic nickel hydroxide. However, XPS investigations proved the silicatic nature of the nickel surface compound. These findings are in line with the preliminary study of Ikeda et al. [3], who speculated about the existence of a nickel silicate based on their EXAFS results. Our own TPR studies revealed a single silicatic species in all samples, which turned out to be of the same kind regardless of nickel loading. Finally, FTIR analyses were employed to establish the specific type of nickel silicate: in all cases a badly crystallized 2:1 phyllosilicate was formed.

On the grounds of nickel silicate formation chemistry as well as some auxiliary arguments, it is proposed that nickel silicate is situated exclusively at the external surface of the MCM-41 particles. This is in contrast to Iwamoto [4], who assumed the nickel compound to be present within the pores of the mesoporous support.  Furthermore, we suggest that the template-ion exchange approach of Iwamoto and coworkers is more adequately described as a template-protected deposition-precipitation.

References

[1]    J. S. Plotkin, Catal. Today (2005), 106, 10-14.

[2]    M. Iwamoto, Y. Kosugi, J. Phys. Chem. C (2007), 111, 13-15.

[3]    K. Ikeda, Y. Kawamura, T. Yamamoto, M. Iwamoto, Catal. Commun. (2008), 9, 106-110.

[4]    M. Iwamoto, Catal. Surv. Asia (2008), 12, 28-37.