Kinetics of catalyzed acid/acid and acid/aldehyde condensation reactions to non-symmetric ketones

Abstract

The kinetics and mechanism of acid and aldehyde condensations to produce non-symmetric ketones with CeO₂-based catalysts were studied using a combination of conventional and pulse microreactor tests. The effects of oxygen and water on the reactions were also studied.

Supported CeO₂ catalysts effectively catalyze the ketonization of acids at essentially complete conversion for extended periods, at weight hourly space velocities of 4–5. The optimal temperature range is 400–430 °C, depending on feed. Time on stream and number of regeneration cycles improved catalyst performance. Selectivities are improved by promotion with small amounts of potassium.

The acid/acid reaction to a typical methylketone proceeds roughly three times faster than the acid/aldehyde reaction, while the aldehyde/aldehyde initial reaction rate to desired methylketones is much slower; multiple aldol condensations predominate. When using acid/aldehyde feeds, water enhances ketone production, probably by supplying oxygen to the catalyst surface. While O₂ can fulfill a similar role, it also promotes combustion. Substitution of D₂O and CD₃COOH for water and acetic acid, respectively, led to kinetic isotope effects between 1.4 and 6.7, which is in the expected range for carboxylate decompositions.

Experiments at low conversion using CD₃COOH and either cyclopropanecarboxylic acid or its aldehyde showed that acetone and methylcyclopropylketone are formed preferentially as five D- and two D-atom isotopomers, respectively, for both acid/acid and acid/aldehyde feeds. This suggests the formation of a surface ketene intermediate, preferentially from acetic acid, which attacks a surface carboxylate to form the ketone, eliminating CO₂. The same conclusions could be drawn from ${}^{\mathrm{<mtext>13</mtext>}}\text{C}$ distributions in experiments using labeled acetic acid.

Introduction

Non-symmetric ketones are known to be produced by the decarboxylative condensation of carboxylic acids. Rajadurai [1] and Pestman et al. [2] have reviewed the process and possible reaction mechanisms. The general reaction is

$$\text{RCOOH}\text{+}\text{R′COOH}\text{→}\text{RCOR′}\text{+}\text{H}{}_2\text{O}\text{+}\text{CO}{}_2$$

Esters can also be used as feedstock. Such ketones are useful as intermediates in making pesticides, herbicides, and pharmaceuticals, and as solvents. In this work, catalysts to make two methylketones of industrial interest, methylcyclopropylketone (MCPK, or ethanone, 1-cyclopropyl) and methylnonylketone (MNK, or 2-undecanone), were studied. These can be made by the decarboxylative condensation of acetic acid (HOAc) with cyclopropanecarboxylic acid (CCA) and decanoic acid (DAc), respectively. It is shown here that certain aldehydes can be condensed catalytically to also produce methyl ketones.

Most previous work on acid condensation reactions involved experiments using supported mixed metal oxides with weakly acidic and basic sites, possibly along with smaller amounts of strongly basic oxide additives. Most catalysts exhibit either low or moderate reducibility at typical reaction conditions of >650 K and 0.1–2.0 MPa [1], [2], [3], [4], [5], [6], [7], [8]. Such materials include the amphoteric or weakly basic oxides such as CeO₂, MnO₂/Mn₃O₄, Fe₂O₃, ZnO, TiO₂ or ZrO₂ [9], [10], [11], [12], [13], [14], [15]. However, in previous work using supported CeO₂ catalysts, ketonization reaction selectivity was found to be low compared to that of aldol condensation or reductive coupling reactions [15], [16], [17]. In our previous work [18], we showed that, at high conversions and temperatures >670 K, supported rare-earth oxides are in fact excellent long-lived catalysts over a wide range of partial pressures and space velocities.

Our goal here is to better understand the condensation mechanism(s) for acid/acid, aldehyde/acid and aldehyde/aldehyde feeds, at conditions of process interest. Such understanding may lead to the design of more selective catalysts for non-symmetric ketonization. Past studies have led to several proposed mechanisms for ketonization of acids and aldehydes. Gonzalez et al. [19] and Pestman et al. [2] proposed that the key step is the coupling of a surface ketene intermediate and a carboxylate. A depiction is shown in Scheme 1. The surface ketene is formed by dehydrogenation of the acid, and then it reacts with an adsorbed carboxylate to ultimately form the ketone, eliminating CO₂. The details of the coupling step are not clear; it is likely that more than one distinct step is involved.

Some factors supporting this mechanism, on certain catalysts, are as follows. First, at least one of the acid reactants must contain an α-hydrogen atom [2]; as the number of α-hydrogen atoms decreases, the rate of formation of ketone does also. Second, ketene is formed at low conversions from acetic acid over many metal oxide catalysts [2], [20], [21]. Third, using acetic [${}^{13}\text{C}$] acid and trimethylacetic acid, it was found that the carbonyl in the resulting ketone arises not from the trimethylacetate only [2]. In some TPD studies, no carboxylates could even be found on oxide surfaces at ≳650 K [14], [20], where ketonization occurs. Finally, acetone does not exchange its H-atoms with a deuterated TiO₂ surface at ketonization conditions, but acetic acid does, in the α-position [2]. However, the materials used in these studies bear little relation to what appear to be the best ketonization catalysts [3], [4], [5], [6], [7], [8], [18].

Kuriacose and coworkers [22], [23] proposed a mechanism for the reaction of two adsorbed carboxylates to produce ketone, water, and carbon dioxide (Scheme 2), based on studies of HOAc ketonization. Okumura and Iwasawa [24] also concluded, from a study of HOAc on ZrO₂/SiO₂ catalysts, that one possibility for ketonization is the reaction of two carboxylates. Another possible mechanism is the reaction of an adsorbed acyl carbenium ion (RCO⁺) with an adsorbed carboxylate [23], [24], [25]. Acyl carbenium ions can be formed in the presence of strong Lewis acids. On a Pd/CeO₂ or Co/CeO₂ catalyst, the ketonization at lower temperatures has also been attributed to reaction between an acetyl intermediate and an alkyl group to produce a ketone [17]. Therefore, it is seen that the main feature distinguishing the proposed mechanisms is the identity of the electrophile attacking the adsorbed carboxylate. In this work we apply detailed product analysis from experiments using differently labeled (${}^2\text{H}$ and ${}^{13}\text{C}$) feeds to provide insight into this identity.

There have also been observed reactions of aldehydes and alcohols that produce ketones rather than simple aldol condensation products [15], [17], [26], [27], [28], [29], [30], [31], [32], [33]. Some evidence supports an aldol condensation mechanism, followed by decomposition of the primary aldol products [26], [27], [33]. Alternative mechanisms suppose that an alcohol is dehydrogenated to the aldehyde, which is oxidized to an adsorbed carboxylate. The carboxylates can then undergo ketonization as usual [15], [17], [30], [32]. Again, we will apply detailed product analysis from experiments using labeled feeds to better understand which pathways predominate for a selective catalyst at typical process conditions.