DISPERSED CARBONATE CATALYSTS FOR THE REVERSE WATER-

GAS SHIFT REACTION

GOVERNMENT RIGHTS

[0001] This invention was made with Government support under contract U.S. Department of Energy (DE-SC0020394) and NASA Grant Number 80NSSC2K0181. The Government has certain rights in the invention.

CROSS REFERENCE TO RELATED APPLICATIONS [0002] This application claims priority under 35 U.S.C. §119 from U.S. Application No. 63/177,091, filed April 20, 2021, entitled CARBONATE-CATALYZED REVERSE WATER GAS SHIFT TO PRODUCE GAS FERMENTATION FEEDSTOCKS FOR RENEWABLE LIQUID FUELS, by Kanan et al., which is incorporated herein by reference for all purposes.

BACKGROUND

[0003] The background description provided here is for the purpose of generally presenting the context of the disclosure. Information described in this background section, as well as aspects of the description that may not otherwise qualify as prior art at the time of filing, are neither expressly nor impliedly admitted as prior art against the present disclosure.

[0004] This invention relates to converting CO2 into renewable fuels and chemicals using ¾ gas. Specifically, the invention relates to CO2 to CO conversion by the reverse water gas shift reaction, which can be coupled with downstream processes to generate liquid fuels or chemicals. Conventional renewable fuel and chemical production using biomass has various limitations including low efficiency for natural photosynthesis, large land requirements, and competition with food generation.

SUMMARY

[0005] To achieve the foregoing and in accordance with the purpose of the present disclosure, a catalyst for performing the reverse-water-gas-shift (RWGS) reaction is provided comprising an alkali carbonate dispersed on a porous support.

[0006] In another manifestation, a method is provided comprising, providing an alkali carbonate catalyst on a porous support, supplying the catalyst with a flow of CO2 and ¾ gas, and heating the catalyst to >350 °C, resulting in an output containing CO.

[0007] These and other features of the present invention will be described in more detail below in the detailed description of the disclosure and in conjunction with the following figures. BRIEF DESCRIPTION OF THE DRAWINGS [0008] The present disclosure is illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:

[0009] FIG. 1 is a high level flow chart of a method used in some embodiments.

[0010] FIG. 2A is a schematic illustration of a carbonate-catalyzed reverse water gas shift cycle.

[0011] FIG. 2B is a schematic illustration of a reversible deprotonation of FFS.

[0012] FIG. 3 is a graph showing equilibrium-limited CO ₂ conversion percentage to CO versus temperature.

[0013] FIG. 4A shows SEMs of the dispersed carbonate catalysts on a porous substrate. [0014] FIG. 4B is a graph of absorbance versus wave number.

[0015] FIG. 5A is a graph that compares the steady-state CO yield at each temperature step for dispersed carbonates, unloaded oxide supports, and a commercial Pt catalyst.

[0016] FIG. 5B is a graph that compares the CO ₂ conversion and selectivity for CO vs temperature for Pt/Ti0 ₂ , CS ₂ CO ₃ /T1O ₂ , and CS ₂ CO ₃ /AI ₂ O ₃ .

[0017] FIG. 6A are graphs of CO and CH ₄ yield versus time.

[0018] FIG. 6B are graphs illustrating the stability of the catalysts at sub-equilibrium conversion.

[0019] FIG. 6C are graphs illustrating the effect of sulfur-containing impurities.

[0020] FIG. 7 is a graph illustrating the RWGS performance for CO yield at different temperatures for K ₂ CO ₃ .

[0021] FIG. 8 is a graph illustrating the RWGS performance for CO yield at different temperatures for CS ₂ CO ₃ .

[0022] FIG. 9 is a schematic view of a system for providing RWGS used in some embodiments.

DETAILED DESCRIPTION OF ILLUSTRATED EMBODIMENTS

[0023] This invention relates to the reaction of CO ₂ and FU to form CO and H ₂ O, which is known as the reverse water gas shift reaction (RWGS, as shown in Eq. 1).

[0024] C0 ₂ + H ₂ CO + H ₂ O (AH°298K = 42.1 kj mol ^-1 ) Equation 1

[0025] RWGS is typically performed by flowing a gas mixture containing CO ₂ and H ₂ over a catalyst in a reactor and heating the reactor to maintain a desired reaction temperature. Because

RWGS is endothermic, the equilibrium-limited single-pass conversion to the products CO and

H ₂ O increases as the reaction temperature is increased. [0026] In some embodiments, RWGS using the catalysts disclosed in this invention can be used as part of a process that produces liquid fuels or multi-carbon chemicals (chemicals whose molecular formula contains more than 1 carbon). In order to facilitate understanding, FIG. 1 is a high level flow chart of a method used in some embodiments. A feedstock for RWGS containing CO2 and H2 is provided (step 104). A RWGS catalyst disclosed in this invention is provided (step 108). The feedstock is flowed to the RWGS catalyst in a reactor (step 112) and heat is supplied to the reactor to convert the feedstock gas into the RWGS product gas (step 116). In some embodiments, the RWGS product gas is further converted into a liquid fuel or other chemical product using a chemical or biological process (step 120).

[0027] In some embodiments, providing the feedstock (step 104) comprises providing water electrolysis to produce the ]¾ of the feedstock. Water electrolysis performs the reaction shown in Eq. 2.

[0028] 2 H2O 2 H ₂ + O2 Equation 2

[0029] Water electrolysis is a well-developed and rapidly growing commercial technology; proton exchange membrane (PEM) electrolyzers can operate at up to -80% energy efficiency at 2 A cm ^-2 . Alkaline water electrolysis is experiencing rapid scale-up and deployment, with numerous companies announcing plans for GW-scale electrolyzer manufacturing plants by 2030. Water electrolysis can be powered by renewable energy (wind or solar energy) or another low-carbon electricity source such as nuclear power.

[0030] Liquid fuels currently provide 30% of global energy demand and are likely to remain a major pillar of global energy supply for many decades. Achieving a carbon-neutral energy system will therefore likely require replacing fossil-derived liquid fuels with renewable alternatives made from CO2, H ₂ 0, and low-carbon energy. At present, the only significant source of renewable liquid fuels are biofuels produced from agricultural products (e.g., corn, sugarcane, palm oil). Biofuel production is limited by the low efficiency of natural photosynthesis, which has a global average of -0.2%. As they are currently produced, biofuels also compete with food for arable land, which can lead to environmentally devastating land use changes. The conversion of renewable electric power to liquid fuel is an attractive alternative because it could take advantage of the relatively high efficiency and falling costs of low-carbon electricity and it would not require arable land.

[0031] Previously reported RWGS catalysts include transition metal nanoparticles on catalyst supports, metal oxides, metal carbides, and metal sulfides. Most of these catalysts have been developed for high-temperature RWGS (>600 °C) and only a small subset have substantial activity at lower temperatures (<600 °C). While operating at high temperatures allows for high single-pass conversion, it necessitates the use of more expensive reactors, heaters, and insulating materials. It also makes it more difficult to use electric heating, which is essential for powering the system with renewable or low-carbon electricity. Performing RWGS at <600 °C is advantageous if it is combined with a subsequent step that can utilize gas mixtures containing CO2 and CO with a relatively high CO2 content. Alternatively, the gaseous products of a RWGS reactor can be separated and recycle loops can be used to achieve high CO2 conversion to CO over multiple passes.

[0032] A major limitation of previously reported RWGS catalysts is that many produce methane as a side product by catalyzing the exothermic methanation reaction (Eq. 3).

[0033] C0 ₂ + 4 H ₂ CH ₄ + 2 H2O (AH ⁰ ₂₉ 8K = -252.9 kj mol ^-1 ) Equation 3

[0034] Because methane formation requires 4 ¾ instead of 1 for RWGS, methane formation tends to increase as the C02:H ₂ gas ratio supplied to the reactor is decreased.

[0035] As an illustrative example, a catalyst consisting of Pt nanoparticles dispersed on Ti0 ₂ (Pt/Ti0 ₂ ) has been reported to achieve near-equilibrium RWGS conversion with a 1:1 C0 ₂ :H ₂ input ratio at a weight hourly space velocity (WHSV) of 12 h ^-1 from 300 °C - 600 °C (Kim, Lee & Hong, Appl. Catal. B, 119, 100-108 (2012)). However, as will be further discussed below, Pt/Ti0 ₂ produces significant amounts of CH ₄ with a 1:3 C0 ₂ :H ₂ feed.

[0036] Methane formation is undesired for any application of RWGS as part of a process to produce liquid fuels or multi-carbon chemicals because methane formation reduces the yield of the desired fuels or multi-carbon chemicals. Methane is also a highly potent greenhouse gas (GHG), at least 25x more potent than CO2. If it is released to the atmosphere as a side product in a CO2 conversion process, it can negate the GHG benefits of utilizing CO2.

[0037] Previously reported RWGS catalysts are susceptible to a number of deactivation pathways that reduce catalyst lifetime, which increases both the material and operating costs of a process. Common deactivation mechanisms include the formation of carbonaceous deposits on the catalyst surface (“coking”), sintering of the catalytically active component (e.g., a transition metal nanoparticle) to reduce its surface area, or reduction of the oxidation state of the catalyst that results in lower RWGS activity or a change in selectivity to favor an undesired product like methane.

[0038] Sources of CO2 often contain impurities, including sulfur-containing gases such as ¾S, thiols, thioethers, and SO2, and nitrogen-containing gases such as ammonia, amines, and HCN. Most previously reported RWGS catalysts are poisoned by one or more of these impurities such that the impurity binds to the catalyst and reduces RWGS activity or alters selectivity. Removing impurities that poison catalysts from CO2 sources requires additional process equipment, energy, and cost, which decreases the overall efficiency of a CO2 to fuels or chemicals process. Catalysts that involve transition metals or transition metal oxides are likely to be poisoned by sulfur-containing impurities (Rodriguez & Hrbek, Acc. Chem. Res. 32, 719-728 (1999)).

[0039] This invention describes RWGS catalysts comprising alkali carbonates (M2CO3; M = Li, Na, K, Rb, Cs) dispersed in mesoporous materials. Mesoporous materials are materials with pores with diameters between 2 nm and 50 nm. Examples include mesoporous oxides such as mesoporous alumina (A1 ₂₍ ¾), silica (S1O2), and various mesoporous oxides of zirconium, cerium, tin, niobium, and tantalum. Other examples include mesoporous carbon materials such as activated carbon.

[0040] Dispersion of M2CO3 in a mesoporous material results in a high surface area of M2CO3 and renders a substantial portion of the M2CO3 amorphous, or non-crystalline. These dispersed carbonate materials are able to deprotonate very weakly acidic molecules at elevated temperature in the presence of CO2. For example, CS2CO3 or K2CO3 dispersed in mesoporous T1O2 or mesoporous carbon can deprotonate a benzene C-H bond in the presence of 10 bar CO2 at -400 °C, generating a putative carbanion intermediate that reacts with CO2 to form benzoate (Xiao, Chant, Frankhouser, Chen, Yau, Washton & Kanan, Nat. Chem., 11, 940-947 (2019)). Benzene has a gas phase acidity of 400 kcal/mol, which is the enthalpy change for the dissociation of benzene into a proton and C6H5 anion in the gas phase. Since ¾ has the same gas phase acidity as benzene, we hypothesized that dispersed carbonates would catalyze RWGS in a similar temperature regime via a cycle involving ¾ carboxylation to form formate (HCO2 ) and subsequent dehydroxylation to form CO, as shown in FIG. 2A.

[0041] Because alkali cations have very weak affinity for CO, dispersed carbonates should be incapable of reducing CO further to CH ₄ during RWGS. While M ₂ CO ₃ is likely to react with H2S through acid-base chemistry, this process should be completely reversible under RWGS conditions because alkali sulfides are not stable in the presence of H2O, as shown in FIG. 2B. Thus, we further hypothesized that dispersed carbonates would be highly selective and sulfur tolerant.

[0042] In some embodiments, RWGS using the catalysts disclosed in this invention can be combined with anaerobic gas fermentation to generate liquid fuels or chemicals from CO2 and

¾. Anaerobic gas fermentation is a process in which acetogenic microorganisms convert a Ci feedstock (CO and/or CO2) and an energy source (CO and/or ¾) into chemicals and fuels. Gas fermentation has been developed in industry over the past 20 years and recently commercialized as a technology to produce ethanol fuel from waste gas streams that are rich in CO, such as industrial off-gases or pyrolyzed biomass. Ethanol can be blended with gasoline for light vehicles or upgraded to jet fuel using established technologies.

[0043] Instead of using emission streams that are rich in CO, which are limited and can be difficult to access, the feedstock gas for anaerobic gas fermentation can be prepared from CO2,

H ₂ 0, and renewable power in a power-to-fuel process. The feedstock gas needs to contain CO and may contain H ₂ and C0 ₂ . While the opportunity to utilize gas fermentation in a power-to- fuel process has been recognized, the only technology that has received substantial attention for preparing the requisite CO component is low-temperature C0 ₂ electrolysis. Although significant progress has been made over the past decade, low-temperature C0 ₂ electrolysis is still in an early stage of development and faces a number of major science and engineering challenges for use in fuel production. The energy efficiency of low-temperature C0 ₂ electrolyzers that produce

CO has been limited to 40-45% at modest current densities (-0.2 A cm ^-2 ) because of problems caused by carbonate formation in the cell (Rabinowitz and Kanan, Nat. Commun., 11, 5231

(2020)). Low-temperature C0 ₂ electrolyzers utilize specialized components that have never been produced at a large scale (C0 ₂ reduction gas diffusion electrodes and anion exchange membranes) and precious metal anodic catalysts (e.g., Ir0 ₂ ) to minimize overpotential losses and avoid corrosion. Achieving efficient mass transport is uniquely challenging for a C0 ₂ electrolyzer because it requires delivering a gaseous reactant and removing both gaseous (CO, unreacted C0 ₂ ) and liquid (H ₂ 0) products. The coinage metals (typically Ag) that are used as the cathodic catalyst for selective C0 ₂ reduction to CO are irreversibly poisoned by low levels of H ₂ S. An alternative strategy is to use H ₂ 0 electrolysis to produce H ₂ and then convert C0 ₂ and H ₂ into a C0/C0 ₂ /H ₂ feedstock by performing RWGS (as shown in Eq. 1) in a fixed-bed reactor. The reactor design principles for heterogeneously catalyzed transformations of gaseous reactants are well established from the refining and chemicals industry such that with a suitable catalyst, RWGS could be rapidly scaled for use in power-to-fuel systems.

[0044] The selectivity and productivity of anaerobic gas fermentation depend on the composition of the feedstock gas supplied to the fermenter, and in particular on the CO:CO:H ₂ ratio. The stoichiometry for C0 ₂ hydrogenation to any saturated acyclic alcohol corresponds to a

1:3 C0 ₂ :H ₂ ratio, whereas CO hydrogenation to alcohol requires a 1:2 CO:H ₂ ratio. With C0 ₂ ,

H ₂ 0, and energy as the primary inputs, a 1:3 C0 ₂ :H ₂ feedstock is the easiest to prepare because it requires only water electrolysis. Unfortunately, a 1:3 C0 ₂ :H ₂ feedstock (with no CO) is not suitable for commercial alcohol production because most acetogens produce only acetate from this mixture and those that do produce alcohol have low growth rates, low productivity, and modest selectivity. A 1:2 CO:H ₂ feedstock supports high microbial growth and alcohol productivity, but this mixture is the most difficult to prepare from CO ₂ and H ₂ O. In between 1:3 CC> ₂ :H ₂ and 1:2 CO:H ₂ are any number of C0:C0 ₂ :H ₂ ratios with the appropriate stoichiometry for alcohol synthesis. Recent developments indicate that relatively low C0:C0 ₂ ratios are sufficient for high gas fermentation alcohol productivity. Commercially viable gas fermentation has been demonstrated using gasified agricultural residues as the feedstock, which typically has a C0:C0 ₂ ratio of ~1:1. Thus, gas fermentation can utilize relatively CC rich gas streams. [0045] The use of RWGS to prepare CO/CO ₂ /H ₂ gas fermentation feedstocks from CO ₂ and H ₂ requires a catalyst that meets a number of criteria. To minimize the process complexity and energy demand, the catalyst should operate at the equilibrium-limited single-pass conversion with an industrially relevant space velocity at the temperature needed to produce a suitable gas ratio to sustain robust gas fermentation of the desired fuel or chemical product. FIG. 3 is a graph showing equilibrium- limited CO ₂ conversion percentage to CO versus temperature in degrees centigrade and various output C0:C0 ₂ :H ₂ gas ratios obtained after H ₂ O removal for a RWGS reactor operating with an input of 1:3 C0 ₂ :H ₂ . For example, with an input of 1:3 C0 ₂ :H ₂ , the equilibrium-limited RWGS conversion corresponds to an output C0:C0 ₂ :H ₂ gas ratio of 1:1:5 at 504 °C. Given the apparent commercial viability of this ratio for ethanol production, some embodiments take 504 °C to be the maximum target temperature. Further developments in gas fermentation processes may make lower ratios and correspondingly lower temperatures viable, e.g., 1:2:8 C0:C0 ₂ :H ₂ at 375 °C.

[0046] The RWGS catalyst must also avoid the exothermic reaction to produce CH ₄ , as shown in Equation 3. Any CH ₄ produced by the catalyst would need to be vented or combusted to prevent its accumulation because acetogens cannot metabolize CH ₄ . Venting even a small percentage of CH ₄ negates the GHG benefit of utilizing CO ₂ because of the 25x higher global warming potential of CH ₄ compared to CO ₂ .

[0047] The RWGS will preferably need to operate in the presence of FFS impurities because H ₂ S is a significant impurity in some of the most accessible emissions sources for CO ₂ utilization, including off-gases from ammonia production and com ethanol fermentation, and H ₂ S is also produced in the gas fermentation process itself.

[0048] Some embodiments provide dispersed carbonate RWGS catalysts to convert FF and

CO ₂ to an appropriate ratio of CO/CO ₂ /H ₂ for anaerobic gas fermentation. In some embodiments, dispersed carbonates can produce the necessary output gas via equilibrium- limited RWGS at >430 °C at useful space velocities for at least days of operation with no apparent degradation and no FFS sensitivity. Additionally, such catalysts are trivial to synthesize and can be readily scaled for rapid deployment. These results provide a compelling case for utilizing carbonate-catalyzed RWGS to bridge the gap between H ₂ and gas fermentation to ethanol, other multi-carbon alcohols, or other valuable multi-carbon products as fermentation technology develops. By upgrading the alcohol products to jet fuel, this strategy could be used to produce sustainable aviation fuel without harvesting biomass.

[0049] In other embodiments, RWGS using the catalysts disclosed in this invention can be combined with subsequent chemical or other abiotic process steps to generate liquid fuels or chemicals. In some embodiments, Fischer-Tropsch catalysis is used to convert the RWGS product gas into hydrocarbons for hydrocarbon fuels such as gasoline, diesel, or aviation fuels. More broadly, the RWGS product gas can be converted to other products using any process that converts syngas to liquid fuels or other multi-carbon chemical products. The RWGS product gas may be subjected to one or more separation steps to separate or enrich components of the product gas prior to other catalytic steps that convert the product gas or components thereof into liquid fuels or other multi-carbon chemical products.

Embodiments using Cs ₂ C0 ₃

[0050] In some embodiments, materials with a high degree of carbonate dispersion that are suitable for fixed bed reactors can be easily prepared from commercially available mesoporous Ti0 ₂ , g-A1 ₂ 0 ₃ , and granular activated carbon (GAC) supports using incipient wetness impregnation. For the examples described here, Ti0 ₂ pellets, g-A1 ₂ 0 ₃ pellets, and granular activated carbon catalyst supports were all purchased from Alfa Aesar of Tewksbury, MA. As- purchased g-A1 ₂ 0 ₃ or Ti0 ₂ pellets were ground lightly in a mortar and pestle and sieved to a size fraction of 250-500 pm (35-60 mesh). The GAC support was etched with 2 M KOH and washed with aqueous HC1 prior to loading. The surface areas of the support materials before Cs ₂ C0 ₃ loading determined by Brunauer-Emmett-Teller (BET) N ₂ sorption were 153 m ² /g, 251 m ² /g, and 608 m ² /g, respectively. Cs ₂ C0 ₃ was loaded into the supports via incipient wetness impregnation using an aqueous solution for g-A1 ₂ 0 ₃ and Ti0 ₂ and a methanolic solution for GAC. The final Cs ₂ C0 ₃ loading was 34.3 wt% for Cs CCVy-AFCF, 23.2 wt% for Cs ₂ C0 ₃ /Ti0 ₂ , and 10.4 wt% for Cs ₂ C0 ₃ /GAC. SEMs of the dispersed carbonate catalysts on a porous substrate are shown in FIG. 4A.

[0051] Infrared (IR) spectroscopy was performed to qualitatively assess the degree of carbonate surface dispersion and the nature of the C0 ₃ ^2- coordination environments. FIG. 4B is a graph of absorbance versus wave number. These measurements were performed after long RWGS runs to allow for any rearrangements that occur during operation. For both materials, the C-0 stretching region is dominated by a pair of broad peaks centered at -1570 and -1340 cm ^-1 , which is indicative of reduced symmetry in the CO ₃ ^2- coordination environment compared to the bulk crystalline lattice. The magnitude of the peak splitting is consistent with a bidentate and/or bridging coordination to the Cs ⁺ cation, as has been assigned previously based on IR spectra of K ₂ CO ₃ dispersed in g-Aΐ ₂₍ ¾. These results indicate that a dispersed, non-crystalline carbonate is stable on the CS ₂ CO ₃ /T1O ₂ and CS ₂ CO ₃ /AI ₂ O ₃ materials under RWGS conditions. Excessive scattering by the GAC support material prevented IR characterization of CS ₂ CO ₃ /GAC.

[0052] CO2 hydrogenation reactions were carried out in a 3/8” O.D. (0.305” I.D.) 316 stainless steel fixed bed reactor equipped with a vertical furnace. For each reaction, an amount of catalyst containing 500 mg of support material was loaded into the reactor and heated while exposing to a flowing mixture of ¾ and CO2 gas (10 bar pressure, 3:1 H2:C02) at a rate of 40 standard cubic centimeters per min (seem), which corresponds to a weight hourly space velocity (WHSV, mass of gas per mass of catalyst per hour) of 2.0 h ^-1 for CS2CO3/AI2O3 and 2.1 h ^-1 for Cs2C03/Ti02. This WHSV was chosen to be within the range of values that are used in industrial applications of heterogeneously catalyzed gas-phase reactions while enabling convenient evaluation of multiple materials over long-duration experiments. Water produced in the reaction was collected in a gravity trap, while the gaseous products were passed to a gas chromatograph (GC) for analysis at regular intervals. CO, CO2, and CH ₄ were quantified in each GC analysis based on independent calibration standards. Some embodiments provide a weight hourly space velocity (WHSV) of >2 h ^-1 .

[0053] An initial survey was performed by stepping the reactor temperature from 355 to 520 °C in 25-30 °C increments, with each temperature step held for 4 h. These experiments revealed that the dispersed carbonates are highly active and selective RWGS catalysts at temperatures greater than -400 °C. FIG. 5A compares the steady-state CO yield at each temperature step for dispersed carbonates, unloaded oxide supports, and a commercial Pt catalyst. A control experiment with the empty reactor showed a small amount of CO production across the temperature screen, which reflects low levels of RWGS activity catalyzed by the stainless steel reactor walls. The unloaded T1O ₂ support showed moderate CO production above this background level but was still far below the equilibrium- limited conversion even at the highest temperature evaluated. By contrast, the CS ₂ CO ₃ /T1O ₂ attained the RWGS equilibrium-limited CO yield at >490 °C (denoted by the black dashed line). Notably, T1O ₂ is partially reduced under the H2-rich atmosphere used in these experiments. A reducible support is not beneficial for carbonate-catalyzed RWGS, however, as seen by the results using an irreducible AI ₂ O ₃ support. The unloaded g-Aΐ2(¾ support itself showed considerably higher CO production than unloaded T1O2, but did not reach the equilibrium- limited yield even at 515 °C. The CS2CO3/AI2O3 catalyst was substantially more active than CS2CO3/T1O2, reaching the equilibrium-limited CO yield at >430 °C. Both CS2CO3/T1O2 and CS2CO3/AI2O3 showed essentially 100% selectivity for CO.

[0054] An oxide support is not essential for carbonate-catalyzed RWGS. Whereas a GAC support showed no increase in CO production relative to the empty reactor, the CS ₂ CO ₃ /GAC catalyst came close to the equilibrium-limited CO yield (within a few percentage points) at >450 °C. CS ₂ CO ₃ /GAC was less active than CS ₂ CO ₃ /T1O ₂ or CS ₂ CO ₃ /AI ₂ O ₃ , but the CS ₂ CO ₃ loading was substantially lower in CS ₂ CO ₃ /GAC (10 wt% vs 23-34 wt%) because of the need to use a methanolic CS2CO3 solution with this more hydrophobic support. Increasing the loading would likely improve the performance with this material.

[0055] For performance comparison with transition metal-catalyzed RWGS, a commercial platinum nanoparticle catalyst (1 wt.% Pt/Ti0 ₂ ) was examined (500 mg catalyst, corresponding to a WHSV of 2.7 hr ¹ ). To our knowledge, Pt/Ti0 ₂ is the most active previously reported RWGS catalyst (on a mol CO per g catalyst per s basis) in the 300-500 °C range. As seen in FIG. 5A, the commercial Pt/Ti0 ₂ catalyst showed considerable RWGS catalysis at all temperatures examined. In contrast to CS ₂ CO ₃ /T1O ₂ and CS ₂ CO ₃ /AI ₂ O ₃ , however, Pt/Ti0 ₂ never reaches the equilibrium-limited CO yield because of competitive CH ₄ formation. FIG. 5B compares the CO ₂ conversion and selectivity for CO vs temperature for Pt/Ti0 ₂ , CS ₂ CO ₃ /T1O ₂ , and CS ₂ CO ₃ /AI ₂ O ₃ . Whereas the dispersed CS ₂ CO ₃ catalysts are essentially 100% selectivity for CO across all temperatures, Pt/Ti0 ₂ shows increasing CH ₄ production as the temperature is increased, reaching -12% (24% of Ci products) at 520 °C. The exothermic methanation of CO ₂ /CO (Eq. 3) is a common competing reaction pathway for RWGS catalysts. The absence of CH ₄ with dispersed carbonate catalysts is consistent with the notion that the low CO affinity of Cs ⁺ precludes further reduction

[0056] To assess their stability, CS ₂ CO ₃ /AI ₂ O ₃ and Pt/Ti02 were compared in 40 h reactions performed under the same flowing conditions (3:1 H ₂ :C02, 10 bar, 40 seem) at 468 °C. FIG. 6A is a graph of CO and CH ₄ yield versus time for these experiments. CS ₂ CO ₃ /AI ₂ O ₃ maintained the equilibrium- limited CO yield without producing CH ₄ for the entire run, with no detectable degradation in performance. In order to demonstrate stability, in contrast, with Pt/Ti02 the CH ₄ remained above 6% after reaching the temperature setpoint but oscillated between 7.5% and 6.6% over the course of the run. The high CH ₄ yield was coupled with a slow reduction in CO yield from 37.9% to 34.4%. This result suggests that there are rearrangements of the Pt nanoparticles on the multi-hour timescale that change the distribution of active sites, which is a common phenomenon for transition metal nanoparticle -based catalysts.

[0057] The stability of the catalysts at sub-equilibrium conversion temperatures were also assessed in 40+ hr reactions performed at 410 °C and the same flow conditions (3:1 H ₂ :C0 ₂ , 10 bar, 40 seem, as shown in FIG. 6B). For the CS ₂ CO ₃ /AI ₂ O ₃ , the CO yield remained steady at 35.2% with negligible CH ₄ production. Cs ₂ C0 ₃ /Ti0 ₂ exhibited similar selectivity for CO but had a much lower CO yield over the course of the experiment, falling to 16.9%. The Pt/Ti0 ₂ also saw a loss in CO yield over the course of the experiment, but CH ₄ production slowly increased from 1.9% to 2.8%.

[0058] Finally, the effect of sulfur-containing impurities was assessed by performing a 40+ h stability test at 404 °C with a gas mixture in which 50 ppm of H2S gas was introduced into the 1:3 C0 ₂ :H ₂ feed, as shown in FIG. 6C. For Pt/Ti0 ₂ , the presence of H2S primarily attenuated CH ₄ formation such that 0.3% CH ₄ yield was maintained throughout the run. In addition, the CO yield showed a slow decline from 22.5% to 19.8% over the course of 40 h. Since the CH ₄ production is essentially constant, we attribute this decline to a slow poisoning of the RWGS sites on the Pt particles, which is consistent with other studies of H2S effects on Pt catalysts. In contrast, the CS ₂ CO ₃ /AI ₂ O ₃ showed essentially the same CO yield and 100% CO selectivity over 40 h.

[0059] The demonstration of robust carbonate-catalyzed RWGS activity under application relevant conditions motivates a preliminary assessment of the accessible energy conversion efficiency for a power-to-ethanol system that utilizes this technology of electrolysis and RWGS. An estimate can be made by examining the energy requirements for each of the primary process steps: 1) electrolytic H ₂ production, 2) RWGS, 3) fermentation, and 4) isolation of the product via distillation. Based on our results above, we assume a 1:1:5 C0:C0 ₂ :H ₂ gas ratio output from the RWGS unit operating at the equilibrium-limited CO2 conversion at -500 °C. Commercial PEM electrolyzers operate at 80% efficiency, requiring 364 kJ mol ^-1 FE generated. The RWGS unit must supply the reaction enthalpy of 37.1 kJ mol ^-1 CO produced at 500 °C. At the required process temperature, industrial electric heaters can approach 100% efficiency. Based on the reported emissions intensity of acetogenic fermentation of syngas to ethanol, estimates for utilities requirements and energy recovery place the total energy requirement for fermentation and distillation at 0.39 MJ mol ^-1 ethanol. Accounting for the carbon efficiency of fermentation (91%) and adjusting for the stoichiometry of ethanol production (i.e., 6 mol FE required per 1 mol CO produced via RWGS), the total baseline power requirement for the system is estimated at 2.82 MJ mol ^-1 ethanol, corresponding to a 48.5% efficiency for power to fuel grade ethanol. [0060] In contrast to CO ₂ electrolysis, the reactor design principles for large-scale, heterogeneously catalyzed gas-phase processes are very well-established from many decades of their use in the chemical and fuel industries. The dispersed carbonate RWGS catalysts described here are trivial to prepare from readily available, non-precious materials (e.g., CS ₂ CO ₃ and g- AI2O3). The H2S tolerance of carbonate-catalyzed RWGS could obviate the need for desulfurization, reducing the overall process complexity and energy demand.

Embodiments using K3CO3 or Na ₂ CQ ₃

[0061] To assess the performance of other alkali carbonates, K2CO3/AI2O3 (18.1 wt% K2CO3) and Na ₂ C0 ₃ /Al ₂ 0 ₃ (14.5 wt%) were prepared in the same way as CS2CO3/AI2O3 using the same g-Aΐ ₂ q ₃ powder such that molar alkali metal carbonate loading was consistent. An amount of catalyst containing 500 mg of support was loaded into the reactor and heated while exposing to a flowing mixture of ¾ and CO2 gas (10 bar pressure, 3:1 H ₂ :C0 ₂ ) at a rate of 40 seem. The RWGS performance was surveyed by stepping the reactor temperature from 355 to 520 °C in 25 °C increments, with each step held for 3 h. The results are shown in FIG. 7. All catalysts were able to reach equilibrium conversion during the course of the reaction while remaining essentially 100% selective for CO. CS2CO3/AI2O3 reached equilibrium conversion at the lowest temperature of -430 °C, followed by K2CO3/AI2O3 at -450 °C, and finally Na ₂ C0 ₃ /Al ₂ 0 ₃ at -480 °C.

[0062] Embodiments using non-powder catalyst morphologies

[0063] Various embodiments using dispersed carbonate RWGS catalysts will require using larger reactor tubes, such as tubes with I.D. greater than or equal to 2”. These larger tubes generally require using catalysts with mm or larger particle dimensions, such as pellets or spheres. To assess performance in a larger morphology, CS ₂ CO ₃ was loaded directly into the cylindrical g-Aΐ ₂ q ₃ pellets (4 mm length; 3 mm diameter) via incipient wetness (34.3 wt% CS2CO3). Two separate batches were prepared, one from a small 1.5 g sample of pellets and another from 100 g of pellets. The resulting Cs ₂ C0 ₃ /Al ₂ 0 ₃ (pellet) catalysts were evaluated in RWGS reactions. An amount of catalyst containing 500 mg of support was loaded into the reactor and heated while exposing to a flowing mixture of ¾ and CO2 gas (10 bar pressure, 3:1 H ₂ :C0 ₂ ) at a rate of 40 seem. The RWGS performance was surveyed by stepping the reactor temperature from 356 to 487 °C in 25 °C increments, with each step held for 3 h. The results are shown in FIG. 8. The performance of Cs ₂ C0 ₃ /Al ₂ 0 ₃ (pellet) catalysts was very similar to CS2CO3/AI2O3 catalysts with powder morphology, with the equilibrium-limited CO yield reached at >430 °C and essentially no CH4 detected. [0064] FIG. 9 is a schematic view of a system 900 for providing RWGS used in some embodiments. The dispersed alkali metal carbonate 904 is in pores 908 of a porous support 912 placed inside a fixed bed reactor 916. FF and CO2 gas are supplied to the catalyst by MFC(s)

920. The entire system is pressurized with pressure being maintained by a back pressure regulator downstream of the reactor. The reactor is heated by a proportional-integral-derivative (PID) heater 924. Internal and external K-type thermocouples are used to measure the temperature. The product water is captured in a gravity trap 928 downstream of the reactor. The gaseous products continue to flow to a gas chromatography machine (GC) 932 for quantification. The downstream MFCs 936 maintain a regular flow rate and allow for dilution which are practical considerations for calculating yields but do not affect performance. A check valve 940 is also included as a safety measure, but also does not affect performance.

Alternative Embodiments

[0065] Other embodiments may use other inorganic bases as well as other cations to achieve robust and selective RWGS catalysis provided the materials are adequately dispersed in a support that is compatible with these conditions. Tailoring the support pore volume and surface area to enable maximal loading of dispersed bases is a straightforward strategy to maximizing

RWGS activity. The ability to utilize different support materials also provides flexibility to tailor other catalyst properties relevant to industrial application such as thermal conductivity. In various embodiments, the alkali carbonate is in the form of M2CO3, where M ⁺ = lithium (Li ⁺ ), sodium (Na ⁺ ), potassium (K ⁺ ), rubidium (Rb ⁺ ), and/or cesium (Cs ⁺ ). In some embodiments, the catalyst is substantially free of elements in groups 5 through 12 of the periodic table, wherein group 5 is the column of the periodic table beginning with vanadium (V) and group 12 is the column of the periodic table beginning with zinc (Zn). The lack of these elements in groups 5 through 12 renders the catalyst more tolerant to impurities found in CO2 gas sources. In some embodiments, alkali carbonate is substantially amorphous. The lack of crystallinity increases the surface area and engenders greater reactivity for the M2CO3. The substrate is a porous substrate.

In some embodiments, the substrate is mesoporous with an average pore size in the range of 2 nm to 50 nm. In some embodiments, the substrate has an open porosity in the range of 5% to

50% determined by B arret- Joyner-Halenda (BJH) analysis of N2 gas sorption isotherms. In some embodiments, the substrate has an open porosity in the range of 15% to 40%/ In some embodiments, the substrate is an oxide material. In some embodiments, the substrate is a metal or metalloid oxide. In some embodiments, the substrate is one or more of substrates commonly used as catalyst carriers in chemical processes, such as titania (T1O ₂ ), alumina (AI ₂ O ₃ ), zirconia (Zr0 ₂ ), or a carbon material, such as granular activated carbon. In some embodiments, the catalyst is heated to a temperature above 350 °C during the RWGS process. In some embodiments, the catalyst is heated to a temperature above 400 °C during the RWGS process. In some embodiments, the catalyst is heated to a temperature in the range of 300 °C to 600 °C in order to enable the use of ordinary materials of construction for the reactor. In some embodiments, the output of the RWGS process is CO and H ₂ 0 and is free of CH ₄ . In some embodiments, the output of the RWG process is CO and H ₂ 0 and less than 2% CH4. In other embodiments, the output of the RWG process is CO and H ₂ 0 and less than 1% CH4. In other embodiments, the output of the RWG process is CO and H ₂ 0 and less than 0.5% CH ₄ . In other embodiments, the output of the RWG process is CO and H ₂ 0 and less than 0.1% CH4.

[0066] While this invention has been described in terms of several preferred embodiments, there are alterations, permutations, modifications, and various substitute equivalents, which fall within the scope of this invention. It should also be noted that there are many alternative ways of implementing the methods and apparatuses of the present invention. It is therefore intended that the following appended claims be interpreted as including all such alterations, permutations, modifications, and various substitute equivalents as fall within the true spirit and scope of the present invention. As used herein, the phrase “A, B, or C” should be construed to mean a logical (“A OR B OR C”), using a non-exclusive logical “OR,” and should not be construed to mean ‘only one of A or B or C. Each step within a process may be an optional step and is not required. Different embodiments may have one or more steps removed or may provide steps in a different order. In addition, various embodiments may provide different steps simultaneously instead of sequentially.