Concerns about the environmental consequences of continued fossil fuel use for the synthesis of commodity chemicals has led to a search for alternative carbon sources which are sustainable. CO2 is an attractive feedstock due to its high abundance, low cost and toxicity, and relative ease of transport. However, the catalytic conversion of CO2 is complicated by its high kinetic and thermodynamic stabilities. A potential solution to this problem is to utilize transition metal complexes, which can interact with CO2 and weaken the strong C=O double bonds. This can provide a low energy pathway for the conversion of CO2 into value-added products. Nevertheless, at this stage the number of chemicals that can be produced using CO2 as a feedstock is small. Surprisingly, late transition metal complexes, which form weak M-O bonds, have not been extensively utilized for catalytic applications. Therefore, a major focus of our research is the development of homogeneous late transition metal catalysts for CO2 conversion.
The insertion of CO2 into a metal element σ-bond is relevant to many catalytic cycles for the conversion of CO2 into fine chemicals. Accordingly, we have explored CO2 insertion into Group 10 alkyl, allyl, alkynyl, hydroxyl, amido, oxazole and imidazole bonds. In almost all cases the reaction proceeds through an outersphere mechanism, with no direct interaction between CO2 and the metal in the transition state. In fact, we have established that in the majority of cases pre-coordination of CO2 to the metal center is not required for insertion into late transition metal element s-bonds. Subsequently, this fundamental information was used to design new and improved catalytic systems. We discovered two Pd based systems, which at the time of their development were the most efficient catalysts for the carboxylation of allylstannanes and allylboranes with CO2 (Eq 1). In one case, the active catalyst is a PdII complex with one η1- and one η3-allyl ligand, while in the other case it is an unusual PdI dimer with two bridging allyl ligands. The products from these reactions can be easily converted into β,γ-unsaturated carboxylic acids, which are versatile building blocks for further functionalization and are often relatively difficult to prepare.
The insertion of CO2 into a pincer supported Pd complex with an η1-allyl ligand was proposed as a key step in the catalytic carboxylation of allenes with CO2, first described by Iwasawa and co-workers. This reaction represents a potentially important method for the synthesis of unsaturated carboxylic acids. The original Pd catalyst featured an unusual PhPSiP (PhPSiP = Si(Me)(2-PPh2-C6H4)2) pincer ligand, but the mechanism of carboxylation was unclear. Using a slightly different system, we have identified, characterized and isolated all of the proposed intermediates in the catalytic cycle and shown that they are kinetically competent catalysts. In addition, several off-cycle species were isolated and shown to be in equilibrium with complexes in the catalytic cycle, as well as being active precatalysts. The major catalyst deactivation pathway was identified and species with an unusual five coordinate hypervalent Si atom bridging between two Pd centers identified. These are unusual examples of species with four center two electron bonds. Furthermore, this detailed mechanistic study has allowed us to develop a new catalyst for the hydroboration of CO2, which gives a maximum TON of greater than 60,000; the highest reported to date (Eq 2). The product of CO2 hydroboration can be used as a reactive formate source in synthetic chemistry. It is expected that by modifying the catalyst and the borane, it will be possible to change selectivity of the reaction to generate products such as formaldehyde or methoxyboranes. This will be the focus of future research.
In addition, we have explored the mechanism of CO2 insertion into a variety of pincer supported Ni, Pd and Ir complexes and established that the ligand trans to the hydride is crucial. In particular, we developed a simple density functional theory (DFT) model for determining the thermodynamics of CO2 insertion into 6-coordinate Ir hydrides and showed that it is unfavorable in most cases. However, by incorporating an H-bond donor into the secondary coordination sphere, we demonstrated the first isolated example of CO2 insertion into an Ir-H bond (Eq 3). Although, secondary coordination sphere interactions are now relatively common for activating small molecules, to the best of our knowledge this was the first time this strategy was used for CO2 activation and it represents a new approach over prior systems. Computational studies on CO2 insertion revealed that the mechanism involves an unusual nucleophilic attack of the hydride on electrophilic CO2 and that the presence of a strong trans-influence ligand opposite the hydride increases the thermodynamic driving force. The H-bond donor stabilizes the transition state and lowers the kinetic barrier for the reaction. Furthermore, the CO2 insertion product 2 is the most active water soluble catalyst reported to date for the hydrogenation of CO2 to formate, giving approximately 350,000 TON (Eq 4). We suggest that the H-bond donor can be crucial for both stoichiometric and catalytic activity and that secondary coordination sphere interactions are viable design elements for improved CO2 reduction. Future extension to other substrates seems plausible.