Catalytic Hydrocarboxylation of Olefins with CO₂ and H₂ – a DFT Computational Analysis

Abstract

One of the major challenges for the utilization of carbon dioxide as a chemical feedstock is to devise thermodynamically feasible transformations to valuable chemicals through efficient catalytic processes. In this paper, we examine the feasibility of the experimentally not yet realized direct hydrocarboxylation of alkenes, using computational methods. A conceivable catalytic cycle was devised for the addition of H2 and CO2 to ethene, which affords propionic acid. The corresponding energy profiles were calculated with state‐of‐the‐art DFT methods for three rhodium pincer complexes as potential catalyst. Several junctions within the productive catalytic cycle were identified, leading to competing hydrogenation reactions of ethene or CO2. A profound analysis of the reaction network by means of the energetic span model allowed us to identify parameters that facilitate the carboxylation and that favor it over the hydrogenation reaction. A comparison of the relevant activation energies revealed that two of the three investigated complexes slightly favour the hydrogenation of ethene, but one complex preferentially stays within the carboxylation cycle.

Carbon-carbon bonds with CO2: Insights from computational studies

“…Rhodium: In 2012, Ostapowitz et al reported a computational investigation of a hypothetical Rh-catalyzed hydrocarboxylation of ethene with CO 2 and H 2 [43]. From their previous investigation [37], three Rh-pincer complexes with different electronic properties were chosen as potential catalysts (Fig.…”

“…12) [32]. Insertion into a Rh-C sp3 bond takes place through a three-membered cyclic TS, where CO 2 interacts with Rh only via the carbon atom [37,43] (Figs. 19 and 25).…”

“…Despite the good performance of NHC ligands, the interaction of CO 2 with free carbenes can lead to energetically low-lying off-cycle intermediates [45]. For rhodium, bidentate phosphine ligands appear to give lower barriers than COD [32], and anionic pincer ligands appear better than neutral ligands [43].…”

“…Most studies added dispersion corrections (via a Grimme correction [23] or via a functional that is parametrized to reproduce dispersion, e.g. M06 [54]) [26,28,30,39,43,50], but it can be noted that two studies concluded that dispersion corrections gave results that are in disagreement with experiment [22,45]. This is not in line with the general experience in the field, which shows that dispersion corrections provide more accurate reaction energies [55].…”

Carbon Dioxide as a C1 Building Block for the Formation of Carboxylic Acids by Formal Catalytic Hydrocarboxylation

“…This transformation-formally referred to as hydrocarboxylation [15] -is thermodynamically feasible and provides an attractive direct route to free carboxylic acids through the use of CO 2 as a C 1 building block.…”