In this Review, we have described CO₂ hydrogenation as an alternative method for so-called artificial photosynthesis to produce fuels, such as formate/formic acid and methanol, with good selectivity and high efficiency. Recent progress in CO₂ hydrogenation using homogeneous catalysts has been remarkable and, combined with formic acid dehydrogenation without producing detectable CO, contributes greatly to the realization of a hydrogen economy.

For CO₂ hydrogenation, an inexpensive and green source of H₂ is needed in contrast to the industrial reforming of natural gas. While H₂ could be produced by electrolysis of water in the presence of a catalyst using solar-generated electricity, obtaining high pressure H₂ for storage and transportation requires additional energy input and engineering considerations. Thus, the storage of H₂ in formic acid as a transportable liquid is attractive provided the technology for clean conversion of CO₂/H₂ and formic acid under mild conditions is developed. For this purpose, scientists have been applying various strategies including utilization of solvents, additives, and the design of sophisticated catalysts. With basic amine additives, solvent-free systems can be achieved, although in most cases an organic solvent such as DMSO or DMF is required. The formate/amine system generally exhibits higher catalytic performance than an aqueous system with the one drawback of losing volatile amines during the hydrogen release process. A few aqueous systems with Cp*Ir catalysts in high formic acid concentrations (4 M to almost pure FA) show promising results. Such a system is ecofriendly and easier to operate. Therefore, the development of efficient and water-soluble catalysts for aqueous FA systems is highly desirable. Water-soluble phosphine ligands and PNP-type pincer complexes have been applied in aqueous systems with considerable success, although the latter require a small amount of organic cosolvent. As for additives, the addition of bases can greatly promote the reaction, but it also results in a product separation problem. To obtain formic acid for the regeneration of H₂ or use in fuel cells, additional acid must be added to neutralize the formate. Application of ionic liquids to facilitate the evaporation of formic acid is a promising solution to this problem. Moreover, the separation of formic acid is not necessary because the direct use of the formate-containing system for hydrogen regeneration has also been achieved. As an alternative to the use of a base in CO₂ hydrogenation, a Lewis acid is a useful additive for H2release from formic acid. To reduce the cost, catalysis with earth-abundant metals such as Fe or Co is highly desirable, and considerable progress has been achieved. However, the high stability of certain hydride species sometimes requires large (e.g., stoichiometric) amounts of expensive additives such as Verkade’s base to produce a vacant coordination site for catalytic hydrogenation reactions. It is still questionable whether the real cost is reduced considering the relatively lower activity and durability of such catalysts as compared to the currently widely used platinum-group catalysts. Development of catalysts with non-noble metals is an important subject for further research; nevertheless, exploration of highly active and durable platinum-group catalysts is still worth pursuing.

The most important aspect of future research is the design of efficient and durable catalysts for CO₂ transforming systems. Innovative ligands with functional groups that contribute to improved activity through metal–ligand cooperation have shown promise. These noninnocent ligands include electro-responsive ligands capable of gaining or losing one or more electrons, ligands having a hydrogen-bonding function, proton-responsive ligands capable of gaining or losing one or more protons, and photoresponsive ligands capable of undergoing a useful change in properties upon irradiation. Theoretical calculations are frequently used to predict and explain these noninnocent ligand effects because the specific contributions of the ligand are often difficult to resolve experimentally. Pincer complexes efficiently activate small molecules such as H₂ and CO₂ via unique dearomatization reactions and/or hydrogen-bonding interactions. Catalysts designed to mimic enzymes such as hydrogenases are also very successful. The bioinspired proton-responsive complexes bearing hydroxy groups at ortho positions to the coordinating N atoms in aromatic N-heterocyclic ligands exhibited extraordinary activity for the catalytic transformation of H₂/CO₂ in aqueous solution under mild conditions. While electronic effects of the substituents are important, the pendant bases in the second coordination sphere of such ligands greatly improve the catalytic activity through the smooth movement of protons. Kinetic isotope effects and computational studies provide clear evidence for the involvement of a water molecule in the rate-determining heterolysis of H₂ in CO₂ hydrogenation that accelerates proton transfer through the formation of a water bridge. Solution pH alters the rate-determining step for H₂ generation from formic acid with these bioinspired complexes. These unique properties, similar to those of enzymes, demonstrated the remarkable success of learning from nature. We believe innovative explorations to improve metal catalysts via the rational design of ligands (i.e., electronic and geometric effects, proton-responsive properties, pendant bases in the second-coordination sphere, etc.) to promote reactions under mild conditions and to optimize the use of water as a solvent are essential for creating carbon-neutral energy sources and for avoiding catastrophic global warming.

Although homogeneously catalyzed hydrogenation of CO₂ to MeOH is rather difficult to achieve, significant progress has been made through indirect hydrogenation of formate, carbonate, or urea, disproportionation of formic acid, multiple-step synthesis, and, most recently, direct CO₂ hydrogenation by a Ru triphos complex. The development of efficient non-noble metal catalysts and metal-free organo-catalysts such as frustrated Lewis pairs will be important directions. Great attention is devoted to the transformation of CO₂ to fuels, and the likelihood of significant success in the near future is quite high.
The authors declare no competing financial interest.

http://pubs.acs.org/doi/abs/10.1021/acs.chemrev.5b00197