For CO2 Reduction, Hydrogen-Bond Donors Do the Trick.

Carbon dioxide is the major pollutant responsible for climate change, an unfortunate byproduct of powering our society with fossil fuels. There is currently great interest in applying renewable energy to capture and reduce CO2 to provide carbon-neutral fuels. This approach mimics natural photosynthesis, which utilizes CO2 for energy storage and as a structural building block. Achieving this goal requires catalysts that can reduce CO2 to higher energy products or fuel precursors. A popular target reaction is the two-electron reduction of CO2 to CO. Along with H2, CO can be used in Fischer–Tropsch processes to generate liquid fuels. In this issue of ACS Central Science, Chapovetsky, Welborn, and co-workers describe an integrated experimental and computational investigation into mechanistic pathways for a series of cobalt complexes that catalyze this reaction.[1] The most active catalyst contains architectural features that are known to facilitate CO2 reduction in other molecular catalysts. However, their detailed analysis contained some surprises — and illuminates new ways in which the secondary structure can be harnessed to promote high catalytic activity.

The initial cobalt catalyst reported by Chapovetsky and co-workers contains four secondary amines along the ligand backbone (1 in Figure ).[2] The pendant N–H groups appear to be poised to assist CO2 binding through hydrogen-bonding interactions; this binding motif has been observed in a structurally similar Ni(cyclam)2+ catalyst.[3] The N–H assist hypothesis was buoyed by the strong positive dependence on the number of groups and catalytic rate. However, the calculated energies for CO2 binding in this fashion contained an unexpected result: the ring flip required to position the N–H for CO2 binding came at a prohibitively high energetic cost, and is inaccessible under catalytic conditions. Instead, the reason analogues with sequentially methylated amines (pendant R2N–H groups replaced by R2N–CH3) have lower catalytic activity is their increased steric profile, which inhibits CO2 binding.

Figure 1

Adapted from the catalytic cycle proposed by Chapovetsky et al. for the hydrogen-bond assisted reduction of CO2.[2] The Co(II) catalyst is reduced at the electrode (i), followed by formation of a CO2 preassociation complex (ii) and rapid initial protonation (iii). The rate-determining second protonation step is facilitated by a hydrogen-bonding interaction with the TFE acid (iv) to yield CO and H2O.

The presence of proximal hydrogen-bond donors is also known to assist CO2 reduction by mediating proton movement. (For a prominent example, see phenol modified Fe(porphyrin), Costentin et al.)[4] However, the computational results again indicate that direct protonation from the N–H group would require an energetically challenging conformational change. Instead, the N–H groups play a role in orienting the acid for the rate-determining second protonation step (Figure , step iv).

The detailed analysis of the catalytic cycle results in a rate law that quantifies the contribution of installing each N–H group in the ligand backbone on overall activity. Achieving this thorough understanding of the relationship between structure and function would likely have been impossible without the cooperative use of experimental observations and theoretical calculations. Ultimately, the unexpected role the N–H groups play in the catalytic cycle underscore the importance of considering structural dynamics in secondary coordination sphere interactions.[5]

The most active variant, 1, with four pendant N–H groups, is both fast and selective for the product CO, with few electron equivalents going toward side reactions (such as direct proton reduction to H2). However, it operates with a high overpotential (η), a reflection of energetic inefficiency. With these properties, it joins an exclusive group of CO2 reduction catalysts that score high on key reactive metrics (selectivity and activity), but require improvement in others (η). A collective examination of mechanistic studies for molecular CO2 reduction catalysts may not appear to coalesce around any single structural feature or challenging catalytic step that would provide cohesive guidelines for catalyst design. This complexity is a consequence of the multivariable challenges associated with optimizing this reaction. Many catalysts rely on highly reducing metal centers to activate relatively inert CO2, but at the price of high overpotential. In other cases, the release of the product CO, a good ligand for electron rich metal centers, limits the catalytic rate. Achieving high product selectivity also requires inhibiting proton reduction pathways. Additionally, as in most multi-electron and -proton reactions, uncoupled movement of the latter can contribute to sluggish kinetics or high energy barriers in the catalytic cycle.[6] Studies on catalysts that have overcome these challenges reveal insight into circumventing these troublesome catalytic steps. In Ni(cyclam)2+ and the phenol-modified Fe(porphyrin) catalysts mentioned previously, cooperative CO2 binding through hydrogen-bonding interactions enables catalysis at milder potentials while also serving as proton shuttles. Additionally, insight into the electronic structure of reduced metal centers provides valuable information on how to inhibit competitive proton reduction pathways and ease CO release.[7] Many other strategies to improve catalyst performance, including incorporation of electrostatic interactions, are also emerging.[8] Can all of these considerations be pieced together to achieve an optimal catalyst?

Many of the aforementioned considerations are actually already assembled in a single – albeit non-synthetic – catalyst. Electrochemical studies on the enzyme Ch CODH I from the anaerobic thermophile Carboxydothermus hydrogenoformans indicate fast and reversible interconversion of CO2 and CO with almost no overpotential indicating near-perfect energy efficiency.[9] Structural studies on the active site point to many of the elements described above, including cooperative CO2 binding and utilizing the local microenvironment to mediate proton movement.[10] Investigating the importance and impact of these interactions using both experiment and theory is crucial to achieving a synthetic catalyst with high rates, selectivity, and thermodynamic efficiency. In this way, we have the potential to mimic nature and master the art of CO2 chemistry.