A Little Metathesis Catalyst Goes a Long Way

Amakawa, K.; Wrabetz, S.; Kröhnert, J.; Tzolova-Müller, G. Schlögl, R.; Trunschke, A. “In Situ Generation of Active Sites in Olefin Metathesis” J. Am. Chem. Soc. 2012, 134, 11462-11473.

Heterogeneous, silica-supported tungsten catalysts are used to conduct the most efficient and largest practiced metathesis reactions in industry (The Phillips Triolefin Process). Despite their prominence, these catalysts are less understood mechanistically than the solution phase metathesis processes with well-defined catalyst systems. Trunschke and coworkers have recently published new insights into solving this problem that has existed about as long as metathesis itself.

Traditional, heterogeneous catalysts for metathesis vary from Mo, W to Re and can be supported on alumina, silica or combinations of the two. Trunschke and coworkers focused on the traditional Mo on mesoporous silica (MoOx/SBA-15). Through characterization of the surfaces, they found that after oxidative pretreatment, the surface exhibited strong Lewis and Brønsted acidity and molybdenum was in the highest oxidation state possible. Raman spectroscopy of the catalyst surface indicated the presence of Mo(VI) and the distinct absence of molybdenum in lower oxidation states, while IR spectroscopy showed the presence of H-bonded groups that are distinct from those present in the mesoporous silica alone. Secondly, ammonia treatment of the MoOx/SBA-15 surface showed NH3 adsorption, whereas, the SBA-15 did not.

With a basic understanding of the varied chemical surface they were working on, the Trunschke group then began investigating the target reaction. Microcalorimetry demonstrated a modest exotherm upon adsorption of propene on the control SBA-15 and a larger exotherm on MoOx/SBA-15. It was found that the surfaces without molybdenum show reversible binding of propene whereas the surfaces with molybdenum showed irreversible absorption. It is inferred from these experiments that the irreversible propene adsorption on the MoOx/SBA-15 is correlated with the formation of active sites. When Trunschke ran the numbers, they found that a whopping 1.5% of the molybdenum atoms contain active sites!

Now it gets interesting. The IR of MoOx/SBA-15 with propene adsorbed showed stretching frequencies similar to those of isopropoxide. The control experiment of treating MoOx/SBA-15 directly with isopropanol showed an IR sprectrum that correlated directly. This supports the theory that the propene is hydrated with the Brønsted acid hydroxyls on the Mo(VI) surface to generate coordinated isoproxide. In addition, small bands correlating to carbonyls are observed, indicating the oxidation of IPA generating acetone and thus reducing Mo(VI) to Mo(IV). A lower intensity of the carbonyls implies, however, that most of the isopropanol is not actually being oxidized to acetone and the majority of the molybdenum remains in the VI oxidation state with IPA adsorbed.
Based on this evidence, Trunschke proposed the mechanism shown below in which a second equivalent of propene adds to the Mo(IV) sites that have been generated by the dissociation of acetone to form the active carbene that is responsible for olefin metathesis.

Carbonyl signals have been previously reported on related metathesis heterogeneous surfaces and have led to a proposed “pseudo-Wittig” mechanism for formation of the active, metathesizing carbene. However, this new evidence for the generation of Mo(IV) centers by isopropanol formation and oxidation to acetone is compelling and not consistent with the former “pseudo-Wittig” mechanism.

Although these heterogeneous metathesis reactions are among the most efficient metathesis reactions known, if only 1.5% of the molybdenum is involved in catalysis, it begs the question of how efficient could it really be? By inserting a pre-treatment involving propene adsorption-desorption, the authors found that the metathesis rate increased by about a factor of two but that the number of active sites did not increase.

It appears that increasing the number of active sites will be a challenge for future studies given the varied nature of the surface metal oxides. The studies by the Trunschke group present a significant stride forward towards improving these industrially relevant metathesis reactions.

Be Sociable, Share!


  1. says

    one alternative how to increase the content of active Mo would be to try to impregnate the silica support with a mixture of inorganic molybdate (VI) salt (i.e. ammonium pentamolybdate) and a silanizing agent like R-SiCl3

    Molybdates like to form hetero-poly anions like phosphomolybdenate. Silica-polymolybdenates are known also. It would be interesting to try those as a source of Mo(VI), it is entirely plausible that the active catalyst has more than one Mo atom

  2. Rosemary Kiser says

    Milkshake: It is entirely possible and quite likely that the active catalyst(s) has more than one Mo atom and that the observed structures are only representatives of a wide range.

    Other studies and catalysts (Re and W as well as Mo) that have been conducted with “desiccant-like” supports and your idea about added silanizing agents could be spot on for activation of catalyst centers and dessication. The authors mention prior work in this area but did not investigate that approach. It is often challenging to compare activities where many different catalysts and supports are evaluated.

Leave a Reply

Your email address will not be published. Required fields are marked *