Overview on CO2 Valorization: Challenge of Molten Carbonates

Chery, Déborah; Lair, Virginie; Cassir, Michel

doi:10.3389/fenrg.2015.00043

REVIEW article

Front. Energy Res., 02 October 2015

Sec. Fuel Cells, Electrolyzers and Membrane Reactors

Volume 3 - 2015 | https://doi.org/10.3389/fenrg.2015.00043

Overview on CO₂ valorization: challenge of molten carbonates

$\r\nDborah Chery$ Déborah Chery

Virginie Lair $Michel Cassir*\r\n$ Michel Cassir*

Chimie ParisTech, CNRS, Institut de Recherche de Chimie Paris, PSL Research University, Paris, France

The capture and utilization of CO₂ is becoming progressively one of the significant challenges in the field of energetic resources. Whatever the energetic device, it is impossible to avoid completely the production of greenhouse gas, even parting from renewable energies. Transforming CO₂ into a valuable fuel, such as alcohols, CO, or even C, could constitute a conceptual revolution in the energetic bouquet offering a huge application domain. Although several routes have been tested for this purpose, on which a general panorama will be given here, molten carbonates are attracting a renewed interest aiming at dissolving and reducing carbon dioxide in such melts. Because of their unique properties, molten carbonates are already used as electrolytes in molten carbonate fuel cells; they can also provoke a breakthrough in a new economy considering CO₂ as an energetic source rather than a waste. Molten carbonates’ science and technology is becoming a strategic field of research for energy and environmental issues. Our aim in this review is to put in evidence the benefits of molten carbonates to valorize CO₂ and to show that it is one of the most interesting routes for such application.

The Problematic of CO₂ and Its Valorization

Global warming due to rising levels of greenhouse effect gases, such as atmospheric carbon dioxide, and energy consumption are the major challenges in the energetic field. Decreasing carbon dioxide emissions is a key issue but not the only one; thus, it is becoming compulsory to capture and reuse this gas in an efficient mode as a new source of fuels. There are different approaches to this stimulating problem.

Some Facts

Carbon dioxide is in the heart of fundamental processes over millions of years on Earth. It allows stocking solar energy through synthesis, and it is a key molecule for producing fossil fuels and, thus, it is an indicator of the consumption of energy from fossil fuels. Of course, the negative effect of CO₂ in the atmosphere is dramatic and contributes to a significant increase in the temperature of our planet, which is becoming alarming. Concentration of this gas is nowadays increasing by more than 2 ppm per year, reaching 400 ppm against 280 ppm in the preindustrial era. The challenge is to store and to transform the CO₂ into valuable chemicals and fuels along with a better understanding of the physicochemical properties and behavior of CO₂. This would be an extraordinary input in the development of sustainable energy.

Key Molecule in a New Energetic Economy

Carbon capture and storage (Sequestration) (CCS) is a potential method to allow the continued use of fossil-fueled power stations while preventing emissions of CO₂ from reaching the atmosphere. Gas and coal (and biomass)-fired power stations can respond to changes in demand more readily than many other sources of electricity production, hence the importance of retaining them as an option in the energy mix (Boot-Handford et al., 2014).

Various technologies for the capture of CO₂ from fossil fuel-fired power plants are available. The four different capture methods are post-combustion, pre-combustion, oxy-combustion, and chemical looping combustion. Each technology has its own advantages and disadvantages and is at different stages of development (Spigarelli and Komar Kawatra, 2013). Recently, molten carbonates are emerging as challenging electrolytes for CO₂ capture (Cassir et al., 2012). In effect, molten carbonate fuel cells (MCFCs) can be used as CO₂ separator and concentrator while producing electricity from hydrocarbons fuels. According to calculations and some tests, using MCFC for carbon capture, the resulting SPECCA (Specific, Primary Energy Consumption for CO₂ avoided), is about 0.5 MJ kg⁻¹ CO₂, which is well below the value obtained with conventional post-combustion removal based on amines (Suguira et al., 2003; Cassir et al., 2012).

A Large Panel of Transformation Media

In 2010, 12 carbon dioxide valorization recovery methods were identified by the Alcimed Company, whose activity lies at the interface of science and marketing (Figure 1).

FIGURE 1

Figure 1. Summary of the 12 CO₂ valorization recovery methods (Thybaud and Lebain, 2010).

CO₂ can be valorized with three main routes. The “No transformation” route will not be described in this review as it is out of the scope of this article. With a “Chemical CO₂ conversion” route, it is technically possible to use CO₂ as a carbon source for the synthesis of commodity products, from simple CO to liquid fuels and high-molecular-weight polymers (Hu et al., 2013). The electrochemical deposition route will be more detailed being our principal focus.

Biologic Transformation

Microalgae

Microalgae, due to their photosynthetic activity, could be used to capture and valorize the CO₂. Microalgae have the ability of fixing CO₂ directly from waste streams such as flue gas as well as using nitrogen from the gas as a nutrient. Before microalgae can be converted into fuels, the biomass content has to be harvested and dried. The conversion can be carried out through thermochemical or biochemical conversion. The products synthesized depend on the algae or bacteria variety chosen and also on the operating conditions. Pharmaceutical industry already uses this route to produce food supplements (spirulina, β-carotene, etc.). Biofuels (biodiesel precursors, ethanol) are emerging. The US start-up SOLIX¹ has initiated commercialization of biofuel from CO₂ proceeding from combustion smokes.

Biocatalysis

It is the use of biochemical agents, as enzymes, to stimulate a chemical reaction and, thus, obtain products by using less drastic operating conditions than in heterogeneous or homogenous catalysis. Inspired by nature, a large body of research has been devoted to new materials catalyzing CO₂ functionalization in either electrochemical or chemical ways. In 2009, the US start-up Carbon Science² has produced liquid fuels thanks to a prototype. The demonstration prototype uses Carbon Science innovative biocatalytic process to break down CO₂ and water and then combines carbon and hydrogen to form methanol, a liquid fuel. Enzymes from microorganisms transform CO₂ and H₂O to methane, methanol, and butanol. CO₂ can be captured by a biologic agent and then electroconverted to bicarbonate and formate (Karthikeyan et al., 2014).

Chemical Transformation

Energetic Value Product

Photonic energetic power source route. In the photonic energetic power source route, carbon dioxide reacts with photogenerated electrons in order to create an anion radical as intermediate species (Hu et al., 2013). Photoelectrochemical reduction of CO₂ requires photoelectrocatalysts combining a semiconductor and a molecular catalyst. Lim et al. reported the use of four types of semiconductors: p-Si, p-GaAs, p-InP, and p-GaP. Metallic molecular catalysts used are mostly based on nickel, cobalt, ruthenium, and rhenium, whereas enzymes (dehydrogenase and pyridine) are the non-metallic molecular catalysts (Lim et al., 2014). CO₂ photoelectrocatalysis is feasible in gaseous phase or in aqueous solvent. In gaseous phase, the photocatalysts directly react on carbon dioxide and steam to produce synthetic or oxygenated hydrocarbons. Water dissociation and carbon dioxide reduction are realized at the same time. On the contrary, in the other system, water dissociation is physically separated from CO₂ reduction because the photocatalysts are integrated in a photoelectrochemical reactor. CO₂ and steam direct photocatalysis produces hydrocarbons and alcohols such as methanol, methane, carbon monoxide, and other compounds of alkene type or paraffin. Different photocatalytic CO₂ reduction systems exist, one of them is the photoactive metal-organic frameworks (MOFs) (Wang et al., 2015). MOFs’ porosity facilitates CO₂ adsorption and, thus, increases the photocatalytic efficiency. Different photocatalysts have been investigated for their performance in photocatalytic CO₂ reduction (TiO₂, BiVO₄, BiWO₆, Zn₂GeO₄). Present investigations are developing less toxic catalysts. The challenge is to develop efficient photocatalysts that can reduce CO₂ under visible light, because currently most of them are only active in the UV region, and to design an efficient harnessing solar radiation reactor (optical fiber, monolith photoreactors, etc.) (Das and Wan Daud, 2014).

In atmospheric conditions, in a dielectric barrier discharge, syngas is produced, either by the conversion of a mixture of CO₂ and CH₄ or by CO₂ reduction in water (Ozkan et al., 2015).

Solar energy could be stored as formic acid which is electrochemically generated from CO₂; it is a viable solar fuel pathway. White et al. accomplished this transformation by semi-optimized indium-based electrolyser stack powered by a photovoltaic panel. They separated light absorption and CO₂ reduction through the use of a commercial solar panel (White et al., 2014).

Another route to valorize the CO₂ is the use of plasma. Plasma permits to reform CO₂ by methane, providing a highly efficient fuel compared to thermocatalytic methods. But it is not commercially viable due to the higher energy plasma input. Carbon nanofibers along with syngas (H₂ + CO), traces of methanol, and some hydrocarbons (C₂H₂, allene) were produced by CO₂ reduction by plasma-assisted in situ decomposition of water. Mahammadunnisa et al. reported, for the first time, the simultaneous activation of carbon dioxide and water in a catalytic non-thermal plasma dielectric barrier discharge reactor operated under ambient conditions. Three different conditions were applied (plasma alone without catalyst, plasma with NiO/γ-Al₂O₃, or plasma with Ni/γ-Al₂O₃). The higher CO₂ conversion was due to synergy between plasma excitation of CO₂ molecule and catalytic action of NiO catalyst. Nickel-based catalysts are promising for H₂ and syngas production. Plasma alone is good for syngas formation. But for the reduction of CO to methane, NiO catalyst facilitates the conversion of CO into methane (Mahammadunnisa et al., 2013).

Ionic Liquids

Ionic liquids’ (ILs) physicochemical properties make them very attractive for the CCS. Once separated from the atmosphere or exhaust flue, there are two options for the carbon dioxide: chemical transformation into useful products or into a form suitable for long-term storage. Long-term CO₂ storage is not perfect because the CO₂ absorption in ILs is reversible (increase in the viscosity leading to CO₂ and absorption rate diffusion restriction). Electrochemical transformation of carbon dioxide is achieved by direct electroreduction at noble metals, or at carbon electrodes (with a substantial overpotential for the last one); thus, yielding carbon dioxide radical anion ( ${CO}_{2}^{• -}$ ), which then typically dimerizes in $C_{2} O_{4}^{2 -}$ or disproportionates in CO and ${CO}_{3}^{2 -} .$ Also electroreduction could be achieved in more active metals, such as transition-based metal centers (Ni, Fe, Pd, Ru, Re, Rh) complexed with various ligands such as porphyrins, bipyridines, or phosphines; thus, yielding a variety of reduction products, most commonly carbon monoxide and onward products such as formic and oxalic acid but also extending to methanol and methane (hydrogen evolves at the same potentials required for CO₂ reduction). CO₂ electroreduction in RTILs (room temperature ionic liquids) allows long-term storage and produces valuable useful chemicals. The choice of RTILs and reaction conditions greatly influences the reduction products. Also, lower potential routes are available for CCS, which are indirect electrochemical reduction routes. The formation of peroxodicarbonate $C_{2} O_{6}^{2 -}$ by oxygen reduction to superoxide in RTILs is an example. Among hydrogen, nitrogen, and carbon monoxide, carbon dioxide and water have the highest solubility in an ionic liquid (C₄C₁-iPF₆) at 25°C. CO₂ is the most soluble in fluorinated IL. CO₂ can be collected from the ILs by pressure decreasing. Optimizing the design of the appropriate RTILs, as well as the electrode material, is compulsory to enhance the absorption and the reduction of CO₂ (Rees and Compton, 2011).

CO₂ Electrochemical Reduction in Aqueous Solutions

This process has been studied since the late 90s on metallic electrodes in aqueous solutions. It is depicted in Figure 2. The main products of CO₂ reduction are methane, ethylene, formate, carbon monoxide, and some alcohols (methanol, ethanol, and propanol). Copper is the most studied electrode, able to give a large panel of products (hydrocarbons and other products such as alcohols at a high current efficiency) when compared with other metals. Indeed, metals are divided into four groups for the CO₂ aqueous electrochemical reduction. Group 1 (Hg, Pb, Bi, In, Sn, Cd, Tl) binds the ${CO}_{2}^{• -}$ intermediates and gives formate (or formic acid) as a product. Group 2 (Au, Ag, Zn, Ga, Pd) yields CO as the major product. Group 2 binds the ${CO}_{2}^{• -}$ intermediate to varying degrees, but cannot reduce CO. Copper is the only commonly studied (for the CO₂ reduction) metal that falls into group 3. Copper binds the ${CO}_{2}^{• -}$ intermediate and reduces the CO to higher reduction products such as hydrocarbons (methane, ethylene, etc.) and alcohols. The last group (Ni, Fe, Pt, and Ti) does not reduce CO₂ directly, strongly binding hydrogen to produce hydrogen only.

FIGURE 2

Figure 2. Mechanism for electrochemical CO₂ reduction on metal surfaces in water (Jones et al., 2014).

The pH of the solution impacts the CO₂ solubility as well as the first CO₂ reduction competition reaction, the hydrogen evolution reaction. The selection of the electrolyte is also a very important parameter for electrochemical reduction of carbon dioxide (Gatrell et al., 2007; Hori, 2008; Oloman and Li, 2008). Many researchers found that lower temperatures suppress hydrogen evolution and increase CO₂ reduction efficiency (Lote, 2014). High-pressure CO₂ electrolysis (till 60 atm) could be the most feasible method for achieving a commercial electrochemical process. The high current density and efficiency observed in this case on different electrodes are comparable with those obtained with high-temperature solid oxide electrolysers.

Molten Carbonates: Strategic Electrolytes

Physicochemical Properties

The interest of molten carbonates proceeds from their unique chemical and physical properties described by many authors (Yuh et al., 1995; Janz and Lorenz, 1961; Ward and Janz, 1965; Kojima et al., 2008; Lair et al., 2012). First of all, they cover a large domain of temperatures according to their nature and combinations, roughly from 450°C to more than 1,000°C. They are constituted either by single alkali carbonates, such as Li₂CO₃, Na₂CO₃, and K₂CO₃, with melting points of 730, 901, and 858°C, respectively, or more commonly by eutectic mixtures, Li₂CO₃–K₂CO₃ (62–38 mol% and 42.8–57.2 mol%), Li₂CO₃–Na₂CO₃ (52–48 mol%), Li₂CO₃–Na₂CO₃–K₂CO₃ (43.5–31.5–25 mol%), or Na₂CO₃–K₂CO₃ (58–42 mol%) with lower melting points, respectively, 488, 499, 501, 397, and 710°C (Janz and Lorenz, 1961; Sangster and Pelton, 1987). Their conductivities are another major feature depending on the nature of the alkali cation, knowing that smaller ions imply higher conductivity: for Li₂CO₃, Na₂CO₃, and K₂CO₃, the values are, respectively, 5.4, 2.8, and 2 S cm⁻¹ at 900°C (Tanase et al., 1987).

The self-ionization constant of a molten carbonate is characterized by the equilibrium:

M_{2} {CO}_{3} (1) \Leftrightarrow M_{2} O(s) + {CO}_{2} (g)

(1)

With (l), (s), and (g) as the liquid, solid, and gas phase, respectively.

This equilibrium can be simplified in terms of a simple ionic form, and considering a molten carbonate as a strong electrolyte, the equilibrium can be given in an ionic form:

{CO}_{3}^{2 -} \Leftrightarrow O^{2 -} + {CO}_{2}

(2)

The oxoacidic concept derives from this equilibrium, where O²⁻ is an electron pair donor which is associated with an oxoacid CO₂ and forms an oxobase $C O_{3}^{2 -}$ (Flood and Forland, 1947; Yamada and Uchida, 1994). The oxoacidity can be fixed by imposing a carbon dioxide partial pressure or by adding oxides in the melt. High partial pressures of CO₂ represent, for instance, highly oxoacidic media.

Present Applications

Fuel Cells

Molten Carbonate Fuel Cells

If molten carbonate media have gained their reputation, it is surely due to their use in MCFCs that have nowadays reached an advanced state of maturity and a first step toward commercial market entry (Cassir et al., 2012). In the last 3 years, more power has been produced by MCFCs, between 70 and 100 MW/year, than with any other fuel cells, including polymer electrolyte membrane fuel cells (PEMFCs), used for transport and, in particular, electric vehicles. The global cell reaction in an MCFC is the following:

H_{2} + \frac{1}{2} O_{2} + {CO}_{2} (cathode) \to H_{2} O + {CO}_{2} (anode)

(3)

CO₂ formed at the anode is recycled and consumed at the cathode. The fuel introduced in the anode compartment is H₂, resulting from natural gas conversion usually by reforming. The anode reaction is:

H_{2} + {CO}_{3}^{2 -} \to H_{2} O + {CO}_{2} + {2e}^{-}

(4)

The oxidant introduced at the cathode side is constituted by a mixture of air and CO₂. The cathode reaction is:

\frac{1}{2} O_{2} + {CO}_{2} + 2 e^{-} \to {CO}_{3}^{2 -}

(5)

Hybrid Direct Carbon Fuel Cell

Direct oxidation of carbon in a fuel cell is an old dream of the industrial revolution where cities were polluted by CO₂ emission. Nevertheless, due to advances in electrochemical kinetics, it is only in the recent years that impressive progress has been realized in direct carbon fuel cells (DCFCs), which is an important issue the important resources in carbon spread all over the world. In a DCFC, a carbon fuel (pyrolysis of biofuels, farming products, etc.) is oxidized at the anode with an energy produced per unit volume four times higher than with methane (Zecevic et al., 2004). In such devices, molten carbonates may have two roles, either directly as the electrolyte or as a carrier of carbon fuel, acting also as an electrochemical mediator for the oxidation process, in the anode reservoir of a hybrid direct carbon fuel cell (HDFC), with a solid oxide electrolyte (Irvine et al., 2006; Pointon et al., 2006; Nabae et al., 2008).

Hybrid Fuel Cells (MCFC/SOFC)

Another kind of hybrid fuel cells presenting a growing interest is constituted by an electrolyte combining solid oxides with molten carbonates (Benamira et al., 2011, 2012). Oxide ions are mostly responsible for the conductivity in the oxide phase and carbonates in the carbonate phase. These composites form highly disordered interfacial regions between the oxide phase and the carbonate phase (Zhu et al., 2010).

Carbon, Capture and Storage, CCS in Molten Carbonates

The molten carbonate fuel cell system can be considered as a CO₂ separator and concentrator because CO₂ can be transferred from the cathode to the anode stream while producing electricity from fuels at the anode side. CO₂ can be extracted from the flue gas of a combined cycle power plant while generating electricity, avoiding loss in plant efficiency and consequent increase in primary energy consumption. It is clear that MCFC is an interesting device for CCS, but its effectiveness still has to be proven against the other present technologies for CO₂ emission mitigation (Spigarelli and Komar Kawatra, 2013). Theoretically, MCFCs can reach 90% CO₂ separation efficiency while producing electricity at high generation efficiency. In fact, some systems are under investigation and seem to have interesting features (Suguira et al., 2003; Wade et al., 2007). Small power plants of 10–20 Mton CO₂ per year are feasible and can find more immediate implementation. Nevertheless, although the potential market is enormous, the present capacity for producing MCFCs at the level of larger power plants of more than 100 Mton CO₂ per year is not yet possible.

High-Temperature CO₂ Electrolysis

General Features

It is well known that high-temperature electrolysis requires less electrical energy but higher thermal energy; however, as thermal energy is cheaper, electrolysis processes are favored at high temperature (Chery et al., 2015). This question is crucial in water steam electrolysis, which is thoroughly investigated in solid oxide electrolysers (Bradley et al. 2004). Recently, it has been shown that CO₂ electrolysis is also feasible in such devices. Bidrawn et al. have shown that an SOE system operating under CO–CO₂ atmosphere exhibited a total cell impedance of 0.36 Ω cm², its efficiency is similar to that of water electrolysis (Bidrawn et al., 2008). The feasibility of such processes in ceramic electrolysers has been investigated in view of NASA’s future exploration of Mars, aiming at converting CO₂ from the Mars atmosphere into life-supporting oxygen and oxidant/propellant fuel (Washman, 2003). The reduction of CO₂ into CO, at typically 800–900°C, can be associated to water electrolysis into H₂ in a so-called co-electrolysis process producing syngas: H₂ + CO. The durability of these types of electrolysis devices is still under study. In any case, it would be interesting to explore the same electrochemical approach at lower temperatures to avoid the drastic corrosion of the materials. Molten carbonates with lower operation temperatures, from 600 to 650°C, and allowing a high solubility of CO₂ within the molten electrolyte, are very interesting candidates to combine the benefits of high temperature, but not too high, with the reactivity of CO₂ in a liquid medium.

Electrolysis in Molten Carbonates

Before introducing molten carbonates, it should be outlined that electroreduction of CO₂ into C has been evidenced in molten chlorides, i.e., NaCl–KCl eutectic; the reduction mechanism has been studied by cyclic voltammetry at a gold electrode, showing a three-step reaction, the first one involving a radical intermediary species, ${CO}_{2}^{2 -}$ ; the second formation of CO; and the third elemental carbon (Novoselova et al., 2007, 2008; Ijije et al., 2014). The production of C by electroreduction of CO₂ in LiCl–Li₂O or CaCl₂–CaO has also been depicted by other authors (Otake et al., 2013).

In the last decade, there is a growing and significant attention on electrolysis processes in molten carbonates with different goals, from water electrolytic transformation into hydrogen, which can also be extended to water and CO₂ coelectrolysis yielding syngas, to CO₂ electroreduction into C or CO in view of producing CH₄ as a second step. The main attractiveness of such melts is their capability of solubilizing carbon dioxide, from single alkali molten carbonates to a variety of carbonate eutectics (Claes et al., 1996, 1999; Chery et al., 2014). Table 1 shows some solubility data extracted from the literature, showing a relatively high solubility. Nevertheless, it is clear that the values obtained depend on the technique used and on the experimental procedure. In view of the important applications of CO₂ in molten media (capture and valorization), a new set of reliable data is urgently required.

TABLE 1

Table 1. Solubility of CO₂ in molten carbonates.

CO₂ properties in molten carbonates have been analyzed early in the literature. Peelen et al. have given some useful characteristics on the CO₂/CO redox system in Li–K carbonate eutectic, finding a simple charge transfer and evaluating the square root of the diffusion constant and the solubility S√D (Peelen et al., 1997). Claes et al. have found by a manometric technique that CO₂ solubility was higher than predicted, attributing this fact to the production of $C_{2} O_{4}^{2 -}$ by the reaction of CO₂ with the carbonate melt (Claes et al., 1996, 1999). Other authors have also detected the reduction of CO₂ directly or as a rate-limiting species in oxygen reduction (Yamada et al., 1995). Nevertheless, processing this molecule into valuable fuels by electrolysis is a more recent challenge. The literature pointing out the feasibility and interest of electroreduction of CO₂ in molten carbonates into elemental C or CO is recent and significant, but is lacking a systematic approach and rigorous data (Groult et al., 2003, 2006; Kaplan et al., 2002, 2010; Le Van et al. 2009; Yin et al., 2013; Chery et al., 2014, 2015; Ijije et al., 2014). Kaplan et al. and Groult et al. were the first in conceiving the conversion of CO₂ in molten carbonates into energy-storage materials (Groult et al., 2003; Kaplan et al., 2002). These authors obtained carbon nanopowders by electrochemical reduction of CO₂ in Li–Na–K carbonates at 450°C, a relatively low temperature, on nickel and glassy carbon. The amorphous carbon obtained had a high specific surface area (450–850 m² g⁻¹) and was tested for lithium-ion intercalation in view of its application in batteries (Groult et al., 2006). Ge et al. used an inert platinum anode and a tungsten cathode CO₂ into amorphous carbon and oxygen in molten LiCl–Li₂CO₃ salt at 700°C (Ge et al., 2015). Yin et al. also used the ternary carbonate eutectic at 500°C, with a Ni cathode and a SnO₂ anode, captured and converted carbon dioxide into a carbon material, exhibiting high BET surface areas of more than 400 m² g⁻¹; SnO₂ anode was found efficient for oxygen production (Yin et al., 2013). In similar conditions, but optimizing the electrolysis cell voltage and over a temperature range of 450–650°C, Tang et al. obtained carbon powder at 450°C with a low energy consumption of 35.6 kWh, a current efficiency of 87.86%, and under a cell voltage of 3.5 V (Tang et al., 2013). Over the production of carbon in molten salts, Ijije et al. developed a review including electrolysis in molten chlorides and molten carbonates (Ijije et al., 2014).

Another route was tested by Kaplan et al. focusing on the conversion of carbon dioxide into carbon monoxide by continuous electrolysis in Li₂CO₃ at 900°C using a titanium cathode and graphite anode; these authors obtained current densities superior to 100 mA cm⁻² (Kaplan et al., 2010). Licht and his coworkers (Licht, 2009, 2011; Licht et al., 2011, 2013) explored the concept of a large-scale development of CO₂ electrolysers combining the production of CO or C with solar thermal energy. Evidence was given on the feasibility of producing valuable molecules, such as CO and C, with a solar efficiency reaching 50% using the so-called STEP (Solar Thermal Electrochemical Production) process. Figure 3 shows the capability of driving three in series molten carbonate electrolysis (Li₂CO₃) with only the maximum power point photovoltage of one Spectrolab CPV (Licht et al., 2011). According to the electrolysis temperature, either C (T < 900°C) or CO (T > 950°C) is formed.

FIGURE 3

Figure 3. STEP carbon capture in which three molten carbonate electrolysis in series are driven by a concentrator photovoltaic. The maximum power point photovoltage of one Spectrolab CPV is sufficient to drive three in series carbon dioxide splitting 950°C molten Li₂CO₃ electrolysis cells (Licht et al., 2011).

Apart from CO₂ electrochemical reduction, another important rising application of molten carbonates is water electrolysis yielding hydrogen and oxygen. Hu et al. have very recently given the proof of concept of such approach. CO₂ valorization can also be combined with water transformation into H₂ through coelectrolysis of CO₂ and CO in order to produce syngas (Hu et al., 2014). Another possible goal with a significant importance in the present energetic field would also be the production of CH₄ through methanation reactions (CO₂ + 4H₂ → CH₄ + 2H₂O; CO + 3H₂ → CH₄ + H₂O). High-temperature electrolysis of CO₂ is probably one of the very promising routes for valorizing this molecule into CO. This benefit could be enhanced by water electrolysis leading to the combination between CO and H₂. The possibility of producing graphite, used also as a fuel in the direct carbon fuel cell (DCFC), is then possible as well.

Although the proof of concept of high-temperature electrolysis processes in molten carbonates has been proven up to a certain level, chemistry and electrochemistry of molten carbonates is complex and requires a deeper understanding, scarcely existing in the present literature, in order to optimize the materials and operation conditions, from the nature and conductivity of the electrolyte to the electrocatalytic properties of the electrodes. Thus, it is compulsory to select the carbonate melt and the precise running conditions according to the required process. For example, succeeding in preparing CO or C, at lower potentials and temperatures (avoiding deterioration of the materials and the cell), is beneficial from an energetic point of view and requires a thorough comparison of the main molten carbonate eutectics. In this sense, we have focused our efforts in two recent articles on a thorough analysis of the electrochemical behavior of CO₂ in different molten carbonate eutectics, from a thermodynamic predictive approach and an experimental one (Chery et al., 2014, 2015). Some significant results concerning this approach will be given.

Figure 4 depicts a potential oxoacidity diagram of the ternary molten carbonate eutectic at two temperatures. The full comprehension of such diagrams with an explanation of all the selected conditions can be found in the literature (Chery et al., 2015). All the potentials are referred to Li₂O/O₂ system. The oxoacidity domain is limited on the oxoacidic side by a P(CO₂) pressure, arbitrarily taken as 1 atm. (higher pressures can be also selected) and on the oxobasic side by the precipitation of the less soluble oxide among the three alkali cations of the melt: Li₂O. The oxidation limit of the diagram is always due to the oxidation of oxide ions into molecular oxygen in this specific case (or other species such as peroxide or superoxide ions under different melts and conditions). The direct reduction of CO₂ into elemental carbon may occur only at relatively oxobasic media, with P(CO₂) < 10⁻³ atm, which is not realistic under high partial pressures of CO₂ required for the desired application. At higher values of P(CO₂), thermodynamically the most favorable phenomenon is the electroreduction of CO₂ into CO, this tendency being increased at P(CO₂) = 1 atm. The second reduction step, occurring at lower potentials, is the reduction of CO into C. This is only one example of how thermodynamic diagrams allow predicting the conditions in which the electrochemical processes might occur. Of course, useful in high-temperature electrochemistry (where reactions are favored thermodynamically), this vision does not take into account kinetics, which can be analyzed through cyclic voltammetry, electrochemical impedance spectroscopy. Figure 5 shows cyclic voltammograms relative to CO₂ in the same ternary eutectic at a gold flag electrode at 600°C, under P(CO₂) = 1 atm. The three-electrode setup associating the gold working electrode with a gold auxiliary electrode and an Ag⁺/Ag reference was fully described in the literature (Chery et al., 2014). A slight pre-electrolysis, carried out during 360 s at −1.1 V vs. Ag⁺/Ag, was found to be important for conditioning the working gold electrode. The potential scan was initiated from the pre-electrolysis potential up to −0.2 V vs. Ag⁺/Ag and then reverses until reaching −1.35 V vs. Ag⁺/Ag. The reduction peak around −1.1 V vs. Ag⁺/Ag depicts the global reduction of CO₂ into CO ${(2CO}_{2} {+2 e}^{-} {= CO + CO}_{3}^{2 -}) .$ This peak is at least a combination of two reactions: the first forming an unstable species which is up to now undetermined: ${CO}_{2}^{-} {: 2 CO}_{2} {+ 2 e}^{-} {= 2 CO}_{2}^{-}$ (Peelen et al., 1997), ${CO}_{2}^{2 -}$ (Novoselova et al., 2008), $C_{2} O_{5}^{2 -}$ (Sangster and Pelton, 1987), or $C_{2} O_{4}^{2 -} ?$ The second step is the reduction of the unstable species into CO, i.e., ${2 CO}_{2}^{-} {= CO + CO}_{3}^{2 -} .$ The oxidation roughly shows a peak around −0.8 V vs. Ag⁺/Ag, which is probably the reoxidation of CO, and another around −0.5 V vs. Ag⁺/Ag, which might be attributed to the reoxidation of adsorbed CO. More insight is given in the literature (Chery et al., 2014), but no sharp conclusion can be given on the mechanistic process. According to the evolution of the CO₂ peak with the scan rate, the reduction potential is slightly moved toward negative potentials, showing a rapid to quasi-rapid system. It should be noted that the same system when using another electrolyte, for instance Li–K, appears slower (Chery et al., 2014). Thus, the properties of the system depend on the nature of the electrolyte and, of course, the electrocatalytic properties of the electrode. Analysis of the reduction peak also shows that it is diffusion-limited.

FIGURE 4

Figure 4. Influence of the temperature on the Li–Na–K (43.5–31.5–25 mol%) eutectic at 600°C for a H₂O/H₂ ratio in pressure of 1 bar and a CO pressure of 10⁻¹ bar.

FIGURE 5

Figure 5. Cyclic voltammograms of CO₂ reduction at a gold electrode in Li–Na–K at 600°C after a short pre-electrolysis of 360 s at −1.1 V vs. Ag⁺/Ag under 100% CO₂. Surface area of 2 cm².

Conclusion

Among the different techniques used for the valorization of carbon dioxide, electrochemical reduction through electrolysis processes has an important position because it is a relatively low-cost technique that is beginning to be associated with renewable energies, such as solar. Moreover, the variety of electrolytic media plays in favor of this approach. High-temperature electrolysis, either with solid oxides or molten salts, appears as interesting from an electrocatalytic and energetic viewpoint. As the MCFCs electrolytes state of art, molten carbonates are able to capture and dissolve CO₂; thus, they are very favourable media for transforming this greenhouse effect gas into valuable chemicals and/or fuels. A rational use of molten carbonates requires a perfect control of the properties and reactivity on CO₂ in such melts, in order to optimize the electrolysis process, making it more efficient and less expensive. This field is very attractive from both scientific and technological sides. It is one of the significant challenges in the energetic and environmental field.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

The authors acknowledge the French Program “Défi ENRS – Projets Exploratoires Émergence CO₂” for financial support. This work was also supported by the French program PLENEX ANR-11-EQPX-0-01.

Footnotes

References

Benamira, M., Ringuedé, A., Albin, V., Vannier, R.-N., Hildebrandt, L., Lagergren, C., et al. (2011). Gadolinia-doped ceria mixed with alkali carbonates for SOFC applications: I- a thermal, structural and morphological insight. J. Power Sources 196, 5546–5554. doi: 10.1016/j.jpowsour.2011.02.004