The chemical reduction of CO2 will be one of the most important reactions in the 21st century. As a greenhouse gas, CO2 contributes to global climate change, and therefore anthropogenic emission of CO2 is resulting in increasing global concern. In order to limit the global mean temperature increase by 2.0-2.4 °C, the global CO2 emission would have to be reduced by 50-80% by 2050  (based on the emission level in 2000). The most effective, and perhaps unlikely, route to decreasing anthropogenic CO2 emissions would be the replacement of all current (13 Terawatt) and future (23 Terawatt by 2050) global energy requirements with non-emissive and renewable resources , however, the abundance and low monetary cost of coal and natural gas cannot be discounted from the global future energy. With foreseeable new policies gearing towards the reduction of carbon emissions, enabling technologies capable of efficient chemical reduction of CO2 to fuels and materials is crucial. The management of large-scale CO2 emissions, such as from coal- and gas-fired power plants, can be achieved in many different ways. At present, CO2 capture and sequestration (CCS) is perhaps the closest to practical application . Despite being the most straightforward solution, the extremely large scale coupled with the energy requirements and the potential for environmental consequences as well as CO2 leakage result in significant obstacles and concern for the implementation of CCS. On the other hand, the reduction of CO2 to fuels using non-carbon based energy sources (such as solar, wind, nuclear, or geothermal), although highly challenging, is expected to be a truly sustainable alternative to CO2 emission. Among the leading synthetic approaches for CO2 reduction include electrochemical reduction , solar driven photochemical reduction , and thermal catalysis such as hydrogenation . Despite their potential, much fundamental work is still required for each approach to improve the overall energy efficiencies and product selectivities to practical and implementable levels.
While the need to overcome the overpotential or kinetic limitation of initial CO2 reduction step is a common critical reaction across the different technologies, the grand challenge lies in the ability to convert CO2 to valuable products without increasing anthropogenic CO2 emissions. Although technologies for the sequential conversion of CO2 to CO and further to higher hydrocarbons exist (through Fischer-Tropsch synthesis), the efficiencies of these processes are low, and most of these processes are currently powered by fossil energy sources. . The development of energy efficient technologies for the reduction of CO2¬ to high energy density fuels without a net increase in anthropogenic CO2 emissions is essential and will require interdisciplinary synergy.
To date, the communication and cooperativity between disciplines and between different approaches has been limited in spite of the many common features. In particular, the need to reduce CO2 while avoiding high energy intermediates would suggest that interplay between theoretical and experimental research for the many different approaches should be highly synergistic in that the understanding gained from one approach should compliment others. Hence, a breakthrough platform to promote interdisciplinary exchange and collaborations is deemed essential, if not of the highest urgency.
The initial complexation/adsorption of CO2 is thought to define the subsequent reactivity both for heterogeneous and homogeneous catalysts. The number of interactions and the nature of the interactions is expected to influence the reactivity due to the activation or deactivation of either oxygen or carbon or both. For catalysts that bind through carbon, reactivity at oxygen may be expected, and thereby the formation of CO as an initial product or intermediate. For catalysts that bind through oxygen, reactivity at carbon may be preferred, such as the hydride transfer to carbon to produce formate. These trends are expected to be general for the different approaches and may be best elucidated through a combination of theoretical and experimental techniques.
The difficulties in cleaving the CO2, on the one hand, necessitate high reaction temperature to overcome the activation energy barrier for C-O bond cleavage. On the other hand, the high temperature reaction favors the formation of C1 molecules such as carbon monoxide due to higher kinetic energy, preventing the formation of longer chain molecules . To overcome this problem, it is crucial to understand the characteristics of CO2 adsorption, the interactions with catalyst surface and reductants, as well as the molecular energetics so that catalysts that are active at lower temperatures can be developed . Effects such as surface defects/kinks/steps, oxygen mobility and ionicity, surface acidity/basicity, shape confinement, hydrogen and oxygen spillover are among the important parameters that can potentially influence the catalytic CO2 reduction. This requires robust and multistage first principles calculations, for example geometry optimisation by generalized gradient approximation (GGA) [11,12], which to date is lacking. Establishing these fundamental appreciations of surface molecular catalysis is highly critical in the designing of lower temperature catalytic surfaces with improved reactivity and selectivity for more complex molecules.
The cornerstone in heterogeneous electrochemical reduction of CO2 was recently discussed by Hori . The challenge is that no effective and selective electrocatalyst material for CO2 reduction is known. In fact, only copper and its alloys have been shown to be capable of producing significant quantities of hydrocarbons from CO2, but they do so very inefficiently. Even on Cu, reduction of CO2 takes place only at very low potentials and with a large fraction of the current being wasted in hydrogen evolution. Hydrocarbons like methane and ethylene are only produced at very reducing potentials. Recently elements of the reaction mechanism on Cu have been studied applying DFT simulations . However, the detailed reaction mechanism is still not known. Furthermore, an understanding of how to lower the overpotential on the electrocatalyst and increase the selectivity compared to hydrogen evolution is still completely missing.
The use of molecular catalysis can improve selectivity and lower overpotentials while increasing mechanistic understanding of CO2 reduction . Promising examples of molecular electrocatalysts include transition metal catalysts such as palladium triphosphine complexes, rhenium bipyridine complexes, and bimetallic copper complexes [16-18]. Additionally, non-metal-based molecular systems have also been reported, including the recent work of Bocarsly . Molecular complexes can serve multiple functions for the development of electrocatalysts for CO2 reduction: these catalysts could be used directly in solution or immobilized at an electrode surface, or the insight gained from molecular electrocatalysts can be extended to development of heterogeneous electrocatalysts. Once understood, the principles governing catalysis should be transferrable to other approaches.
One of the largest motivations for the photochemical conversion of CO2 is the ability to directly harness the abundance of solar energy. In a typical photochemical reaction, irradiation with a light source of energy equal to or greater than the bandgap of a semiconductor photocatalyst is required to generate electron-hole pairs. The energy of these extractable charge carriers is determined predominantly by the conduction and valence band potentials of the photocatalysts. Typically, the direct reduction of linear CO2 to form bent CO2•- is difficult (E° = -1.90 VNHE) and results in a large overpotential in the overall reduction reaction . A multi-electron reduction step, although challenging, requires considerably less energy due to more positive reduction potentials .
In terms of suitable materials, there are in general a number of metal d0 and d10 oxide simple heterogeneous photocatalytic materials [21,22] with suitable conduction band potential capable of CO2 reduction, for example ZrO2 (5.0 eV), Nb2O5 (3.8 eV), Ta2O5 (4.0 eV), Ga2O3 (4.6 eV) and GeO2 (5.5 eV). However the bandgap of these materials are far too wide for the efficient utilization of solar light. Hence, band engineering of the parent materials is being actively adopted to narrow the bandgap of the material without affecting its conduction band potential . Such effort is commonly accompanied by calculations by density of state (DOS) or density functional theory (DFT), to ascertain the exact LUMO and HOMO positions of the new materials .
Like that of thermal catalytic conversion, the adsorption configuration of CO2 on the photocatalyst surface is an important parameter that dictates the extent of the reaction overpotential. Quite recently, the photocatalytic reduction of CO2 in nanotubular structure was shown to demonstrate good activity . The composite of photocatalyst and organometallic catalyst is another promising system for efficient solar photoconversion . For efficient interfacial electrons transfer, the DFT studies on the conjugation and the resultant configuration of organometallic adsorption is essential , while prediction of other properties such as interfacial energetics and distribution of photoelectrons are of equal importance and deserve rigorous investigation.