Two are better than one

From DNA repair and natural photosynthesis to biological nitrogen fixation, redox-active metalloenzymes enjoy high reactivity thanks to the delicate local environments around their active sites. When electron transfer takes place in such systems, the energetics of both the transition and final states can be stabilized with the transfer of a charged species, usually a proton, through interactions in the second coordination sphere¹. Such a synergistic proton-coupled electron transfer leads to low activation energy for catalysis as well as long-range charge transfer in enzymes¹. Inspired by nature, chemists seek to develop molecular catalysts with ingenious ligand design to similarly fine-tune the interactions occurring in the second coordination environment. Using characterization techniques such as crystallography, chemists have been able to determine structures and fundamental mechanistic details of enzymes and model catalysts that have enabled them to design kinetically efficient catalysts with high turnover frequencies^2,3.

The success of fine-tuning interactions in the second coordination sphere of catalytic molecular systems invites us to wonder whether we can apply this concept to catalytic materials. Two challenges present themselves. The first one involves sample preparation: although molecular catalysts are universally uniform as chemical compounds, a large degree of heterogeneity may exist on a material’s surface. Yet the recent rise of single-atom catalysts⁴ — which possess individual atoms as active sites distributed on a presumably reaction-innocent surface — provides a platform to study such interactions. Then comes the second challenge: obtaining structural information at the atomic level for catalytic materials so as to establish a structure–function correlation. Thankfully, the development of characterization techniques such as electron microscopy and synchrotron radiation-based X-ray methods has significantly advanced our understanding of materials at the atomic level. Mastering both the synthesis and characterization of single-atom catalysts offers a unique opportunity to study the synergistic effect in the second coordination sphere for heterogeneous catalysis.

Now, writing in Nature Chemistry, Chen and co-workers have described how pairs of Cu atoms on a material’s surface can exhibit a synergistic effect for electrochemical CO₂ fixation (Fig. 1)⁵. Built upon their strong expertise of single-atom catalysts^4,6, the researchers developed a method to synthesize Pd₁₀Te₃ alloy nanowires loaded with a controllable amount of Cu, ranging between 0 and 0.20 wt%. The extremely low loading amount of Cu ensures sparsely distributed Cu atoms, which is desirable for mechanistic investigation. They tested the catalytic properties of the prepared materials for electrochemical CO₂fixation and measured both the selectivity of generating CO from CO₂(versus proton reduction to H₂) and the rate of CO formation. While Pd₁₀Te₃ nanowires are not known to be very reactive, the added Cu is well known for its catalytic activity in CO₂ fixation. Thus, one may intuitively postulate that higher loading amounts of Cu should translate to higher values of both selectivity and reaction rate. Instead, an optimal loading amount of Cu was discovered, at roughly 0.10 wt%, which exhibited distinctively high selectivity and reaction rate. This intriguing observation led Chen and colleagues to investigate the underlying mechanism.

A metallic Cu⁰ atom activates CO₂ and facilitates electron transfer, the product of which may be stabilized by hydrogen bonding from the H₂O bound to a proximal Cu^x+ atom (x = 0, 1, 2). Figure adapted from ref. ⁵, Springer Nature Ltd.

Full size image

They characterized the catalysts and obtained detailed structural information using electron microscopy and synchrotron radiation-based X-ray techniques. They concluded that at low loading amounts the surface possesses clusters of four Cu atoms in which a pair of Cu atoms is exposed as the possible active site with two additional atoms sitting underneath. One atom of the exposed pair is metallic (Cu⁰), whereas the other is less well defined (Cu^x+, x = 0, 1, 2) (Fig. 1). By contrast, at loading amounts higher than 0.10 wt%, the average oxidation state of the Cu atoms approaches +2 and a Cu species similar to that in bulk CuO is generated. Therefore, Chen and colleagues propose that it is the pairs of exposed Cu atoms (Cu⁰–Cu^x+) on the Pd₁₀Te₃ nanowires that are predominantly responsible for the observed high reactivity of CO₂ fixation.

The Cu⁰–Cu^x+ pair is proposed to exhibit a synergistic effect that involves hydrogen bonding. Computational results suggest that the metallic Cu⁰ acts as an active site for CO₂ binding and electron transfer, while the proximal Cu^x+ acts as a Lewis acid, binding H₂O. The proton of the bound H₂O in the second coordination sphere of Cu⁰ is acidic enough to potentially form a hydrogen bond with the terminal oxygen in CO₂ and promote the formation of CO (Fig. 1). Comparing with control scenarios, the interaction in Cu⁰–Cu^x+ is reported to lower the energy of the transition state during charge transfer. Furthermore, the strong tendency for hydrogen bonding seems to raise the energetics of both hydride formation and the binding of CO, which suppresses the competing proton reduction to H₂ as well as the further reduction of CO into other organic species, respectively.

The computational results are consistent with observed high selectivity and reaction rate when the Cu⁰–Cu^x+ pair is most abundant, corroborating the claim that the Cu⁰–Cu^x+ pair is more reactive and can serve as the basis for advanced design of CO₂ reduction catalysts. In the future, additional characterization, especially using in-situ electrochemistry techniques, should provide more information about the catalytic process and help illustrate the proposed mechanism. Also, one feature of the Cu⁰–Cu^x+ pair is that the reaction was observed to involve a rate-limiting, single-electron transfer to CO₂, which is most likely followed by a subsequent proton transfer. It will be very interesting if the rate-limiting step can be switched to proton transfer or even a concerted process², with the use of other atom pairs that modulate the redox chemistry, the binding strength of CO₂ and the acidity of bound H₂O.

Overall, the reported Cu⁰–Cu^x+ pair, dubbed by Chen and colleagues as an atom-pair catalyst, showcases a model system that translates the interactions in the second coordination sphere into heterogeneous catalytic materials. The synthetic and characterization methods demonstrated by the researchers can be applied to other catalytic systems and the concept of pairing atoms for a better catalyst will open up new opportunities for atomically dispersed catalysts. When marrying atoms and creating synergy, two are better than one.