Electrocarboxylation of aromatic ketones: Influence of operative parameters on the competition between ketyl and ring carboxylation

Abstract

The purpose of this work is to investigate the effect of operational parameters on the competition between the formation of the target 2-hydroxy-2-arylpropanoic acid and ring carboxylation in the electrocarboxylation of aromatic ketones. For the investigated ketones, this competition has been found to be dramatically influenced by different parameters such as the water content and the ratio between the carbon dioxide and the ketone concentrations (q = [CO₂]/[ketone]). In particular, the target carboxylic acid formation can be favoured with respect to ring carboxylation by operating at high q ratios or by addition of small amounts of H₂O to the reaction medium. An increase of the water content leads, otherwise, also to an increase of the selectivity of other by-products such as alcohol and pinacol. The selectivity of the process was not appreciably influenced by other operative parameters such as the cathodic material or the presence of sacrificial anodes.

Introduction

The electrocarboxylation of ketones to the corresponding 2-hydroxy-2-arylpropanoic acids is considered to be a key step of a possible route to the production of some 2-arylpropanoic acids widely used as anti-inflammatory agents [1], [2], [3], [4], [5], [6], [7], [8], [9]. Usual by-products of these syntheses are pinacols and alcohols arising from dimerisation of the ketyl radical and hydrogenation of the carbonyl group, respectively. According to the literature, all these products derive from different evolutions of the same intermediate, i.e. the radical anion arising from the first electron uptake of the carbonyl group: [ArC(R)O]⁻. Recently, a theoretical and experimental study allowed to individuate the operative conditions to optimise the selectivity of the target carboxylic acid with respect to pinacol and alcohol [9].

Quite surprisingly, we have recently shown that the electrochemical carboxylation of numerous aromatic ketones gives rise also to the formation of substituted benzoic acids, apparently arising from substitution of an aromatic hydrogen by CO₂. In some cases, also small amounts of cyclohexene carboxylic acids were found [10].

In general, the electrocarboxylation of aromatic ketones can be accounted for by the following reaction sequence [9], [10], [11].

$$\text{ArC(R)O}+{\text{e}}^{-}\leftrightarrows [\text{ArC(R)O}{]}^{-}$$

$$[\text{ArC(R)O}{]}^{-}+{\text{CO}}_2\leftrightarrows [{\text{ArC(R)OCO}}_2{]}^{-}$$

$$[{\text{ArC(R)OCO}}_2{]}^{-}+{\text{e}}^{-}\leftrightarrows [{\text{ArC(R)OCO}}_2{]}^{2-}$$

$$[{\text{ArC(OCO}}_2)\text{R}{]}^{2-}+{\text{CO}}_2\to [{\text{ArC(R)(OCO}}_2){\text{CO}}_2{]}^{2-}$$

The ketyl radical anion reacts with CO₂ to give an adduct which is further reduced and carboxylated. Hydrolysis of the carbonate group in the work-up gives the α-hydroxyacid. Besides the heterogeneous electron transfer Eq. (3), reduction of the adduct can take place in solution through disproportionation reactions.

On the other hand, the formation of substituted benzoic acids and cyclohexene carboxylic acids, indicates that the reaction between the radical anion [ArC(R)O]⁻ and CO₂ can also be an attack of CO₂ at an aromatic carbon. The ketyl radical anion may have several resonant forms bearing a negative charge in the phenyl ring. Indeed, the presence of significant charge density in the phenyl ring has been confirmed by molecular orbital calculations on reduced aromatic ketones, which have evidenced that less than 50% of the total charge on the ketyl radical anion is associated with the carbonyl group [12], [13]. All possible resonant forms may be involved in the carboxylation reaction between the ketyl radical anion [ArC(R)O]⁻ and CO₂. Thus, besides the carboxylation reactions (2), (4) at the carbonyl group, an attack of CO₂ at the negatively charged carbons of the phenyl ring should be taken into consideration Eq. (5):

$$[\text{ArC(R)O}{]}^{-}+{\text{CO}}_2\to [{\text{R}}^{\prime }\text{C(R)O}{]}^{-}$$

where R′ stands for the carboxylated ring. Reaction (5) followed by further reduction, protonation and/or carboxylation steps leads to ring carboxylated products of general formula R″C(R)O, where R″ denotes the carboxylated moiety, which may be of aromatic or non aromatic nature.

It is clear that the desired carboxylation process Eqs. (2), (4) is in competition with the ring carboxylation Eq. (5), and the relative competition parameter p = k_c/k_r (k_c and k_r stand for the overall rate constants of carbonyl and ring carboxylation, respectively) seems, intuitively, strongly dependent on the nature of the substrate rather than on experimental conditions. In particular, a bulky and strong electron-donating group R, will increase the negative charge in the aromatic group Ar and will create a steric hindrance around the carbonyl moiety. This will result in an increase in k_r and a decrease in k_c, and hence in a decrease in the competition parameter p. Vice versa, if R is a less bulky, good electron-withdrawing group, carboxylation at the carbonyl moiety leading to α-hydroxyacid should be more favoured over ring carboxylation. These expectations have been met in our previous studies on the electrocarboxylation of aromatic ketones [10]. Indeed, no ring carboxylated products were detected (e.g., p ∼ ∝) when R was a phenyl ring or an aromatic group as in benzophenones, whereas with a methyl (acetophenone) and a t-butyl group (2,2-dimethylpropiophenone) values of the competition parameter of ca. 10 and 1.1, respectively, were found [10], in line with the fact that –C(CH₃)₃ exerts a strong steric hindrance around the carbonyl group.

From an applicative point of view the occurrence of ring carboxylation may cause separation problems, since a highly pure product is desirable in view of its destination as a pharmaceutical drug. In addition, in some cases (e.g. the electrocarboxylation of 2,2-dimethylpropiophenone), the total yield of ring carboxylated products was found to be only slightly lower than that of the target α-hydroxyacid, thus showing that the occurrence of ring carboxylation reactions can significantly affect the selectivity of the process. Hence, it is important to find experimental conditions which strongly favour formation of the target acid, in order to minimize, or preferably eliminate, the presence of ring carboxylated products in the reaction mixture.

To this end, we investigated the influence of numerous operative parameters on the selectivity of the main products and, in particular, on the competition between ring carboxylation and carboxylation of the ketyl group. This has been performed within the framework of the electrocarboxylation of 2,2-dimethylpropiophenone (1a) and acetophenone (1b). The investigated compounds, together with their main products, are shown in Scheme 1. These substrates were chosen as model compounds because they are particularly suited to investigate the influence of the operative parameters on the process, since both of them give quite low yields of the corresponding α-hydroxyacids in the usual electrocarboxylation conditions [10]. 2,2-Dimethylpropiophenone, in particular, is characterised by a very high selectivity for ring carboxylated products, which constitute the main by-products, whereas electrocarboxylation of acetophenone has a quite low selectivity for all carboxylated products as a result of massive formation of 1-phenylethanol and pinacol.

It is worth noting that the reaction environment may influence the distribution of the charge in the radical anion [ArC(R)O]⁻ and, as a consequence, the values of k_c and k_r and, in a final count, the competition parameter p. In particular, it is well known that radical anions generated by 1e⁻ reduction of aromatic ketones form adducts with electrophiles such as proton donors [14], [15], [16], [17], [18] or metallic cations [19]. The charge distribution in such adducts may be significantly different from that of the free ketyl radical anion. Hence, in the following we have focused on the effect of species such as H₂O that are capable of forming adducts with the ketones and/or their radical anions. Also the effect of other operative parameters, such as the substrate and the carbon dioxide concentrations, that strongly influence the competition between carboxylation, protonation and dimerisation reactions [9], has been investigated in detail.