Solar-powered synthesis of hydrocarbons from carbon dioxide and water

The synthesis of hydrocarbons via electroreduction of CO₂ is an attractive approach to store energy generated from intermittent renewable sources of electricity (e.g., solar) through formation of the high-energy C–C and C–H bonds of reduced carbon compounds (1, 2). Establishment of such processes also represents a critical step toward the sustainable production of carbon-based commodity chemicals and energy-rich liquid fuels from nonpetroleum resources (3). Despite the promise that such strategies hold, facilitating the rapid, selective, and efficient electrosynthesis of multicarbon products from CO₂ is an inherently difficult proposition. Part of the challenge stems from the fact that CO₂ reduction half-reactions that generate value-added C₁ and multicarbon products take place within a narrow potential window that is less than 0.5 V wide (Fig. 1).

Fig. 1.

Equilibrium potentials for reductive and oxidative half-reactions relevant to the synthesis of hydrocarbons from carbon dioxide, water, and sunlight. Electron and proton equivalents generated via water oxidation at the anode of an EC can be utilized for fuel-forming cathodic processes to generate hydrogen gas or a broad array of reduced carbon products. Promoting the efficient and selective formation of hydrocarbons of high value and utility requires a highly integrated approach in which electrocatalysts, electrolytes, cell parameters, and light-harvesting assemblies are optimized to work together.

As a result of the reaction landscape illustrated in Fig. 1, it is virtually impossible to target a given CO₂ reduction product based purely on thermodynamic considerations. For instance, electrochemical reduction of CO₂ to ethylene occurs at 0.08 V vs. reversible hydrogen electrode (RHE). Accordingly, any electrochemical process that targets this product must be run at a potential at which production of other species such as ethane (E° = 0.14 V) and methane (E° = 0.17 V) is also thermodynamically feasible. Moreover, kinetics associated with CO₂ reduction reactions (CO₂RRs) can often be sluggish. This is particularly true for CO₂RR processes that generate reduced products with C–C bonds, which require application of modest overpotentials of at least 400 to 500 mV. In practice, electrochemical hydrocarbon evolution ultimately requires cathodic potentials that are more negative than –0.5 V vs. RHE, which ultimately brings all of the CO₂ couples of Fig. 1 into play. Matters are further complicated by the fact that each CO₂ reduction couple requires both e^– and H⁺ equivalents. As a result, the kinetically facile reduction of protons to hydrogen gas (E° = 0.0 V) represents a competitive cathodic process that must be suppressed for efficient and selective hydrocarbon evolution to be realized.

Each CO₂ reduction half-reaction requires multiple e^– and H⁺ equivalents that are most logically provided by the oxidation of water (E° = 1.23 V). Accordingly, electrochemical cells (ECs) for sustainable hydrocarbon synthesis must juxtapose cathode and anode catalysts that can manage the demanding multielectron proton-coupled electron transfer reactions attendant to CO₂ reduction and H₂O oxidation, respectively. Moreover, since production of hydrocarbons from CO₂ and H₂O is energetically uphill, an external bias such as that provided by a photovoltaic (PV) assembly is needed to generate electrons and holes of sufficient potential to power the cathodic CO₂ reduction and anodic H₂O oxidation reactions illustrated in Fig. 1.

Development of improved systems for CO₂ conversion has been an area of intense activity, with significant emphasis placed on EC design and on discovery of improved CO₂ reduction (4, 5) and H₂O oxidation (6) catalysts based on nonprecious earth-abundant elements. PV-EC systems for solar-driven CO₂RRs have also been reported (7–9); however, only two such systems have provided appreciable selectivity for reduced carbon compounds containing C–C bonds (10, 11). Among the PV-EC systems for CO₂ conversion, the most efficient one generates the C₁ feedstock CO (not hydrocarbons) using a GaInP/GaInAs/Ge PV cell (9). A 3% solar-to-hydrocarbon efficiency has been achieved using an iridium oxide anode coupled to a four-terminal III-V/Si tandem PV cell (11). Despite the significant advances that prior CO₂ solar-electrolyzers have provided, these systems either do not produce hydrocarbon products, require expensive electrocatalysts (i.e., anodes or cathodes containing Ir, Au, Ag), or utilize designer PV cells that are poorly suited for widespread use. Accordingly, the efficient production of hydrocarbons via reduction of CO₂ using integrated PV-EC systems remains as a major challenge in the field of molecular energy conversion.

In PNAS, Huan et al. (12) describe a comparatively low-cost PV-EC assembly that evolves C₂ hydrocarbons (ethylene and ethane) and O₂, from CO₂, water, and simulated sunlight. Unlike other photoelectrochemical assemblies for CO₂ valorization that employ photocathodes and/or photoanodes (13), the solar electrolyzer that Huan et al. describe separates the jobs of light capture and catalysis, permitting each task to be independently optimized. The PV-EC system leverages a dendritic nanostructured copper oxide material (DN-CuO) for CO₂ reduction, which was chosen since Cu-based cathodes have previously shown the ability to generate hydrocarbons and oxygenates. Notably, the DN-CuO also catalyzes the oxygen evolution reaction (OER), which conveniently permits the same material to be used for the anodic H₂O oxidation half-reaction that provides the e^– and H⁺ equivalents needed for cathodic CO₂ reduction.

The DN-CuO is based on inexpensive earth-abundant elements and promotes both CO₂ reduction and OER with manageable overpotentials, enabling construction of an EC for sustainable and efficient hydrocarbon synthesis. Huan et al. (12) use an integrated approach to minimize energy losses in their EC device and optimize electrochemical performance. By employing Cs⁺-based electrolytes, the authors enhance CO₂ reduction kinetics while promoting C–C bond-forming electroreduction processes. These efforts are in line with previous studies demonstrating how the size and electronic properties of electrolyte additives can impact the selectivity and kinetics of CO₂RR processes (14, 15). Moreover, by tailoring the electrolyte concentration and pH of the catholyte (pH 6.8) and anolyte (pH 11), Huan et al. could achieve rapid reduction of CO₂ to hydrocarbons.

At cathodic potentials of 0.95 V vs. RHE, the electrochemical reduction of CO₂ to the C₂ products ethylene and ethane proceeded with Faradaic yield (FY) values of 37% and 13%, respectively. CO (FY of 5%) and HCOOH (FY of 7%) were observed as minor products, as the DN-CuO–based EC favored formation of C–C bond-coupled products over reduced C₁ species by an impressive 3:1 margin. These product distributions were realized with stable current densities of ∼25 mA⋅cm⁻² at cell potentials of ∼3 V. The ability to achieve such high selectivities and production rates for the electrochemical conversion of CO₂ to hydrocarbon products and O₂ without the need for concentrated alkaline electrolytes and at relatively low cell potential represents a significant achievement, as the above metrics correspond to a record cell energy efficiency for hydrocarbon formation of ∼20%.

In PNAS, Huan et al. describe a comparatively low-cost PV-EC assembly that evolves C₂ hydrocarbons (ethylene and ethane) and O₂, from CO₂, water, and simulated sunlight.

With an effective, relatively simple, and low-cost hydrocarbon-evolving electrolyzer in place, Huan et al. (12) proceeded to power their EC with sunlight. Rather than opt for more expensive multijunction silicon- or III/V semiconductor-based PVs, a light-harvesting array comprising inexpensive and easily processable triple-cation perovskite solar cells (16) was used to construct the PV-EC device. The perovskite PV cell was configured such that its power output (generated under simulated sunlight) was well matched to the voltage and current requirements for optimal EC performance. Ensuring that a PV cell’s output is tailored to a given EC’s requirements can often be a tricky proposition, and is particularly important for CO₂ reduction schemes in which small deviations in cell potential can lead to changes in electrolysis kinetics and hydrocarbon vs. C₁ product selectivities. By slightly adjusting electrode surface areas such that the EC would be optimized for the perovskite array’s output, the PV-EC could be powered solely by sunlight with a current density of ∼20 mA⋅cm⁻² at a cell potential of 2.8 V. CO₂ reduction product distributions under the solar-driven conditions leaned heavily toward hydrocarbon production, with ethylene (FY of 34%) and ethane (FY of 7%) predominating over C₁ products by nearly a 4:1 margin. When taken together, these results correspond to a solar-to-hydrocarbon efficiency of 2.3%, which represents a new benchmark for the direct solar-driven synthesis of hydrocarbons using a PV-EC that does not rely on precious elements.

There is no question that the PV-EC assembly that Huan et al. (12) have developed is highly encouraging and represents a considerable improvement in our ability to promote the solar-driven conversion of CO₂ to ethylene and ethane. As importantly, this work should also spur intense efforts as the catalysis and energy-conversion communities continue to refine their approach to effect efficient hydrocarbon synthesis from CO₂, H₂O, and sunlight. For example, Cu-based cathodes are known to display poor selectivity for reduction of CO₂ over protons at low overpotentials. Although this PV-EC system sets a new standard for direct solar-to-hydrocarbon conversion efficiency, the DN-CuO catalyst directs more than 40% of the EC current to hydrogen evolution, as opposed to CO₂RR. Since H₂ is of far lower commodity value than ethylene, propylene, and nonvolatile CO₂RR products (HCO₂H, CH₃CO₂H, CH₃CH₂OH), this side reaction represents a parasitic process that must be mitigated if CO₂-to-hydrocarbon electrolysis is to become commercially viable.

In addition, most cathode materials that generate C–C coupled products from CO₂ electroreduction display broad product distributions. While the DN-CuO platform shows good selectivity for ethylene, a number of other volatile reduced carbon products are also generated. Each of these species ultimately needs to be removed from the product stream for the true commodity value of ethylene to be realized, potentially introducing energetic and financial impediments to solar-driven hydrocarbon synthetic schemes. These factors emphasize the need to improve and maximize FYs and energy efficiencies for the highest-value C–C coupled products and/or liquid fuels that are compatible with existing energy storage, distribution, and conversion technologies (17). Future efforts aimed at developing new catalyst/electrolyte combinations and EC designs that suppress H₂ evolution and favor CO₂RRs with high selectivity will be important in this regard.

Improved EC designs will also likely be important to improving the kinetics of direct CO₂-to-hydrocarbon conversion, which will be required for such processes to be of commercial value. The ∼25 mA⋅cm⁻² current densities that are obtained with the EC described by Huan et al. (12) are significant for a system operating near neutral pH, and are well-matched to the power characteristics of the perovskite PV used to drive solar hydrocarbon evolution. However, current densities on the order of 10² to 10³ mA⋅cm⁻² are needed for such technologies to be implemented in real-world energy storage and commodity chemical synthetic schemes. Adapting the DN-CuO EC by introduction of gas-diffusion electrodes or other advanced cell designs may enable even higher current densities for CO₂-to-hydrocarbon production as has been achieved with electrolyzers that convert carbon monoxide to hydrocarbons under alkaline conditions (18). Similarly, improved solar-to-fuel efficiencies will be important to the large-scale adoption of solar electrolyzers for CO₂ conversion. A more thorough understanding of PV-EC performance under real-world operating conditions should prove useful in devising improved systems for renewable hydrocarbon synthesis. Such efforts will also greatly facilitate holistic solar-electrolyzer design and scaling efforts, because the manner in which PV cells and ECs are coupled and the potential gains to be made through use of DC power optimizers and associated technologies should not be understated (19). Given all of the above, we should expect that efforts to design and map new catalyst materials, electrolyte additives, electrode geometries, and electrolyzer designs will continue to be critical to establishing selective and energy-efficient platforms for the electrochemical valorization of CO₂.