Catalyzed carbon-based materials for CO₂-battery utilization

Electrochemistry of nonaqueous M-CO₂ batteries

A typical nonaqueous M-CO₂ battery consists of three main components: a metal anode, a porous CO₂-breathing cathode, and a separator containing a metal-ion-conducting electrolyte^[25]. Figure 1A provides a schematic illustration of this configuration. Unlike the reversible insertion of Li⁺ or Na⁺ ions between the electrodes seen in traditional Li-ion or Na-ion batteries, the energy storage and release mechanism in M-CO₂ batteries is based on conversion reactions^[26-29]. During discharge, the metal anode is oxidized, forming metal ions (Mⁿ⁺, where n represents the oxidation state of metal ions), which then migrate through the organic electrolyte to the cathode. The movement direction of Mⁿ⁺ is indicated by the green arrow. Concurrently, CO₂ is reduced at the cathode, and the resulting reduction products react with the migrating metal ions to generate the final discharge product. This is referred to as the CO₂ reduction reaction (CO₂RR). During charging, the reactions reverse. When the battery is charged, the reverse occurs: discharge products are decomposed, releasing Mⁿ⁺ and CO₂, and the Mⁿ⁺ re-moves from the cathode to the anode (in the direction of the orange arrows), with metal being deposited back onto the anode and CO₂ evolving at the cathode. This process is commonly referred to as the CO₂ evolution reaction (CO₂ER)^[30].

Figure 1. Schematic illustration of (A) Nonaqueous and (B) Aqueous M-CO₂ battery configuration.

In nonaqueous M-CO₂ batteries, the CO₂RR can proceed via either a two-electron or a four-electron transfer pathway, resulting in distinct discharge products^[7]. The two-electron transfer pathway typically leads to the formation of metal oxalate [M₂(C₂O₄)_n], while the four-electron transfer process yields a combination of metal carbonate and carbon as the discharge products. The mechanism involving the four-electron transfer pathway is now widely recognized, and the corresponding reaction equations can be summarized as follows: corresponding reactions can be expressed as follows:

Anodic reaction: M → Mⁿ⁺ + ne^-

Cathodic reaction: 2CO₂ + 2e^- → C₂O₄^2-

C₂O₄^2- → CO₂^2- + CO₂

C₂O₄^2- + CO₂^2- → 2CO₃^2- + C

CO₃^2- + 2 Mⁿ⁺ → M₂(CO₃)_n

Overall reaction: 4 M + 3CO₂ ↔2 M₂(CO₃)_n + C

Experimentally, significant progress has been achieved in understanding the electrochemistry of the four-electron transfer pathway in M-CO₂ batteries, and several physicochemical factors influencing the reaction mechanisms have been studied^[31-33]. When studying intermediate discharge products, the intermediate M₂(C₂O₄)_n is usually further converted to metal carbonate due to its poor thermal stability. Interestingly, it was discovered that under specific conditions, the intermediate product C₂O₄^2- can be stabilized, implying that CO₂RR proceeds via the two-electron transfer pathway, and the corresponding reaction process is as follows:

2 M + 2CO₂ ↔ M₂(C₂O₄)_n

For example, when using a molybdenum carbide (Mo₂C) catalyst, it was found that Mo₂C stabilizes the amorphous intermediate discharge product Li₂C₂O₄ through the delocalized electrons generated by Mo-O coupling^[34,35]. Zhou et al. demonstrated that Mo-O chemical bonds form between Mo₂C and C₂O₄^2- during discharge processes^[36]. This interaction allows delocalized electrons from Mo²⁺ and Mo³⁺ in Mo₂C to transfer to the electronegative oxygen atoms in C₂O₄^2-, thereby stabilizing Li₂C₂O₄ as a discharge product. Notably, Li₂C₂O₄ decomposes thermodynamically more easily during charging (E° = 3.01 V vs. Li/Li⁺) compared to Li₂CO₃ (E° = 3.82 V vs. Li/Li⁺), reducing the charge potential to approximately 3.4 V. In contrast, catalysts such as carbon nanotubes (CNTs) lack this stabilization capability, causing C₂O₄^2- to disproportionate into CO₃^2- and carbon, leading to the formation of Li₂CO₃, which is thermodynamically stable but highly insulating and difficult to decompose below 4.0 V. Furthermore, experimental results, including a precipitation test with CaCl₂ and H₃PO₄, confirm the presence of C₂O₄^2- in Mo₂C-based systems but not in CNT-based systems. Density functional theory (DFT) calculations demonstrated the superior adsorption energy of Mo₂C for Li and CO₂, as well as its metallic conductivity and ability to stabilize Li₂C₂O₄. Charge density difference analyses further revealed significant electron transfer from Mo₂C to Li₂C₂O₄, unlike CNTs, which lack this capability. As a result, Li₂C₂O₄ forms as a stable discharge product with Mo₂C, while Li₂CO₃ is the primary discharge product for CNTs. Currently, few catalysts, including Mo₂C and MoN, have shown the ability to stabilize M₂(C₂O₄)_n. Enhancing interactions between Li⁺ ions and electrolyte molecules also promotes the formation of stable solution-phase C₂O₄^2- while suppressing the direct reduction of CO₂ to Li₂CO₃^[37]. Thermodynamically, M₂C₂O₄ decomposes more readily than M₂CO₃, reducing charging voltage, facilitating reversibility, and enabling stable M-CO₂ battery operation^[38-40]. DFT studies on Ru and Au catalysts in Li-CO₂ batteries further support this. Li₂C₂O₄ emerges as a kinetically favorable intermediate, with Li₂CO₃ as the thermodynamically stable product. Ru outperforms Au due to lower Gibbs free energy for nucleation and reduced activation barriers for rate-limiting steps, enabling efficient conversion of Li₂C₂O₄ into Li₂CO₃ and carbon along distinct reaction pathways^[41].

Electrochemistry of aqueous M-CO₂ batteries

Aqueous M-CO₂ batteries are gaining particular attention due to their feature of simultaneously reducing CO₂ to value-added carbonaceous products and energy storage. As illustrated in Figure 1B, A typical aqueous M-CO₂ battery consists of a metal anode, a catalytically active cathode, and a separator (e.g., solid electrolyte for Li/Na-CO₂, or bipolar membrane for Zn/Al/Mg-CO₂ batteries) for separating the anolyte and catholyte. The main function of the anode chamber is to dissolve and deposit metals at the anode for energy release and storage during discharge and charging, respectively, similar to nonaqueous M-CO₂ batteries. The cathode chamber is similar to that of the electrolyzer of direct electrocatalytic CO₂ reduction, and thus offers unique advantages for CO₂ conversion in the CO₂RR process based on the proton-coupled electron transfer mechanism that produces a wide range of carbon-containing products (C_xH_yO_z, such as HCOOH, CO, CH₄, CH₃OH, etc.), and the corresponding discharge reaction is as follows:

Anode: M → Mⁿ⁺+ ne^-

Cathode: CO₂ + H⁺ + ne^- → C_xH_yO_z

Overall discharge reaction: M + CO₂ + H⁺ → C_xH_yO_z+ Mⁿ⁺

During the charging phase of aqueous M-CO₂ batteries, depending on the discharge product, the reaction may involve reversible conversion of the discharge product (mainly for HCOOH) or oxidation of H₂O [oxygen evolution reaction (OER)] (mainly for other difficult to oxidize products such as CO₂ or CH₄, etc.). The charging reaction can be expressed as follows:

Anode: Mⁿ⁺ + ne^-→ M

Cathode: HCOOH→ CO₂ + 2H⁺ + 2e^- or 2H₂O → O₂(g) + 4H⁺ + 4e^-

Overall charging reaction: Mⁿ⁺ + HCOOH → CO₂ + H⁺ +M

or Mⁿ⁺ + 2H₂O →O₂ + 2M+ 4H⁺

It is worth mentioning that (i) the fundamental difference between aqueous and nonaqueous M-CO₂ batteries lies in their reaction mechanisms, which are strongly influenced by the electrolyte environment. In nonaqueous M-CO₂ batteries, the reaction mechanism primarily involves the reduction of CO₂ to form solid discharge products such as M₂(CO₃)_n + C or M₂(C₂O₄)_n. These solid products are decomposed reversibly during charging, releasing CO₂ again. In contrast, in aqueous M-CO₂ batteries, the reaction is more similar to direct electrocatalytic CO₂RR to produce carbon-containing products (C_xH_yO_z) due to the aqueous medium. Upon charging, oxidation of HCOOH or H₂O occurs depending on the type of discharge product.

(ii) Nonaqueous systems, typically composed of metal salts and solvents such as tetraethylene glycol dimethyl ether (TEGDME), leverage the low volatility, wide electrochemical window, and high chemical stability of ether-based electrolytes, making TEGDME a representative solvent widely used in nonaqueous M-CO₂ batteries^[7,24]. In contrast, aqueous M-CO₂ batteries, such as Zn-CO₂ batteries, feature an alkaline anolyte [e.g., KOH mixed with Zn(Ac)₂] to promote Zn redox reactions and a near-neutral catholyte [e.g., KHCO₃ solution or KHCO₃ mixed with Zn(Ac)₂] to favor CO₂RR, as alkaline conditions conflict with CO₂RR pathways^[42]. The two electrolytes are separated by a bipolar membrane to maintain their respective chemical environments, ensuring optimal reaction conditions at both electrodes.

(iii) The electrochemical performance of both nonaqueous and aqueous M-CO₂ batteries is heavily influenced by the properties of the cathode catalyst materials. In nonaqueous M-CO₂ batteries, M₂(CO₃)_n + C is widely recognized as the main discharge product. However, it is found that under the influence of some special catalysts, the intermediate M₂(C₂O₄)_n does not decompose further and can be stabilized while forming the final product. For example, when using a molybdenum carbide (Mo₂C) catalyst, it was found that Mo₂C stabilizes the amorphous intermediate discharge product Li₂C₂O₄ through the delocalized electrons generated by Mo-O coupling^[34]. Only a few reported catalysts (e.g., Mo₂C and MoN) have been able to stabilize the production of M₂(C₂O₄)_n. Therefore, catalysts that promote the stabilization of solid M₂(C₂O₄)_n shall be preferred due to their high catalytic efficiency, and ability to stabilize intermediates. In addition, it is essential to develop catalysts that promote the reversible generation and decomposition of discharge products, such as M₂(CO₃)_n + C. These catalysts can reduce charging polarization, thereby improving energy efficiency and prolonging battery life. For aqueous M-CO₂ batteries, catalysts that exhibit high selectivity and efficiency for CO₂ reduction to specific products are preferred. Examples include metal-based catalysts (e.g., Au and Ag) that can produce CO, and Sn-based catalysts that can produce HCOOH. Moreover, this catalyst needs to have OER catalytic activity to facilitate the charging reaction and reduce energy consumption.

Fundamental challenges of cathode in M-CO₂ batteries

CO₂-breathing cathodes are the primary site of CO₂ capture, conversion, product formation, decomposition or release, involving multi-phase interfaces and complex reactions^[18,43-46]. Typically, sufficiently efficient electron and mass (e.g., Mⁿ⁺ and CO₂) transfers and the abundance of catalytically active sites at the three-phase interface of the CO₂-breathing cathode are required to ensure that CO₂ is converted efficiently, so that the high performance of M-CO₂ batteries can be realized. Therefore, significant challenges arise and need to be addressed before their practical application [Figure 2].

Figure 2. Challenges, structural design principles and typical carbon-based materials of CO₂-breathing cathodes for M-CO₂ batteries.

(i) Limited CO₂ availability and utilization: ensuring a continuous and sufficient supply of CO₂ to the cathode is critical for the sustained operation of M-CO₂ batteries. However, the cathode structure (e.g., surface area, porosity, wettability, etc.) may be suboptimal, with uneven distribution of surface reaction sites, low reactivity, or a catalyst surface that is easily covered by products, making it difficult for CO₂ to adequately reach the reaction sites or to overflow before it has been reacted, leading to a lower efficiency of CO₂ utilization. Catalyst surfaces may also become blocked by discharge products, reducing access to active reaction sites. In aqueous systems, the hydrogen evolution reaction (HER) competes with CO₂RR^[47,48], consuming energy and further decreasing CO₂ conversion efficiency due to the favorable reduction of H₂O to H₂ gas.

(ii) Sluggish reaction kinetics of CO₂RR: CO₂ is chemically stable, with a strong C=O bond (750 kJ mol^-1) necessitating high overpotentials for activation. The CO₂RR process involves multi-step electron transfers and the formation, dissolution, and adsorption of intermediates. Suboptimal interactions between the cathode material and these intermediates often lead to kinetic bottlenecks. Ideal cathode materials should effectively adsorb CO₂ and intermediates, facilitating their reduction while maintaining optimal binding strength.

(iii) Clogging of CO₂-breathing cathodes: in aqueous M-CO₂ batteries, the discharge products are dissolved in the electrolyte. However, in nonaqueous M-CO₂ batteries, insoluble carbonate discharge products (e.g., Li₂CO₃ and Na₂CO₃) precipitate on the cathode surface or within its pores, clogging channels and blocking CO₂ and ion diffusion. The limited solubility and low diffusivity of CO₂ in nonaqueous electrolytes exacerbate uneven product deposition, concentrating discharge products near the CO₂ source. This spatial imbalance leads to increased charge transfer resistance, higher overpotentials, and significant capacity decline over time, ultimately reducing battery lifespan and practical viability.

(iv) Cathode corrosion and degradation: high charging voltages required to decompose insulating carbonate products can cause undesirable side reactions, including electrolyte decomposition and cathode corrosion. Carbon-based materials may corrode due to reactive oxygen species, while metal cathodes can form passivating oxides or hydroxides in the presence of H₂O, O₂ and CO₂. Repeated redox cycling can induce stress accumulation, microstructural changes, or crack formation, invalidating active sites and limiting performance. At the nanoscale, local electric fields and structural remodeling can further degrade the catalyst, diminishing system efficiency and durability.

Structural design of cathode in M-CO₂ batteries

To clearly address the aforementioned challenges, a well-designed CO₂-breathing cathode should fulfill several key functions, including facilitating the diffusion of Mⁿ⁺ ions and CO₂, catalyzing the formation and decomposition of the discharge products, providing sufficient storage space for these products, and guiding their morphological evolution. To meet these functional requirements, several critical performance metrics must be optimized: (i) High electrical conductivity and an optimized porous structure are key for CO₂ cathodes. Conductivity ensures fast electron transfer, reducing internal resistance and boosting power output. A well-designed porous structure helps diffuse Mn⁺ ions and CO₂ and stores discharge products. Mesopores (2-50 nm) can facilitate CO₂ and ion transport, while macropores (> 50 nm) provide ample space for the storage of insulating discharge products such as Li₂CO₃ or Na₂CO₃, reducing impedance and enhancing cycle performance^[49]. One study shows that combining micro- and mesopores increases surface area (2,003 vs. 1,813 m²g^-1), enhancing ion and gas movement. However, excessive activation reagents can collapse the structure, reducing surface area. A balance between pore size and volume is crucial^[50]. Therefore, a balance between pore size and volume is desirable; (ii) Favorable CO₂ bonding affinity. An ideal CO₂ cathode should possess a high binding affinity for CO₂ to facilitate its adsorption and activation, which helps reduce the energy barriers associated with CO₂RR, improving overall reaction kinetics and battery performance; and (iii) Excellent catalytic activity. To achieve high electrochemical performance, catalytic cathodes should possess abundant and accessible active sites for CO₂RR and CO₂ER, to lower the overpotential during both charge and discharge cycles, and to improve the overall efficiency of the battery. Whereas in aqueous M-CO₂ batteries, depending on their charging/discharging mechanism, effective CO₂RR/CO₂ER or CO₂RR/OER bifunctional electrocatalysts are required on the CO₂ cathode.

Importance and strategies of carbon-based cathode for M-CO₂ batteries

As mentioned earlier, the remarkable electrical conductivity, low mass density, affordability, significant specific surface area, customizable porous structure, and macroscopic morphology contribute to the extensive investigation of carbon materials in M-CO₂ batteries. For example, Super P and KB with inherent high conductivity have been widely recognized as conductive agents, which are essential for electron transfer in M-CO₂ batteries^[14,45,51]. In addition, carbon materials (e.g., CNTs and graphene) have a tunable porous structure allowing for optimized pore size distribution and a large specific surface area, which contributes to the efficient adsorption of CO₂, thus ensuring the efficient diffusion of CO₂ and Mⁿ⁺ ions and providing space for the storage of discharge products. Additionally, carbon paper and carbon cloth serve as current collectors or gas diffusion layers, providing a conductive framework that allows for the efficient electron transfer from the external circuit to the active sites, while facilitating the diffusion of CO₂ into the catalyst and enhancing the supply of CO₂ at the active sites^[52-54]. Consequently, the presence of carbon materials is indispensable in M-CO₂ batteries.

However, pure carbon materials have limited catalytic activity in promoting CO₂RR and CO₂ER processes, which may lead to higher overpotentials. Carbon materials alone may not be sufficient to achieve the high catalytic activity required for practical M-CO₂ batteries^[55-57]. To overcome this limitation, various integrated strategies have been reported: (a) Nanostructure engineering to design conductive networks with hierarchical porous structures to alleviate pore clogging and promote sustained Mⁿ⁺ and CO₂ transport; (b) constructing of CO₂ cathodes with sufficient electrocatalytic activity, e.g., doping (e.g., heteroatom doping such as N, S, B, etc.), metal/carbon composites, single atom (SA) catalysts, MOFs or, to enhance their CO₂ bonding affinity and catalytic activity; and (c) designing self-supporting cathodes for surface protection of the carbon layer to mitigate the degradation of the carbon cathode.

Ketjen black and super P

Commercial carbon materials, such as KB and Super P, are widely used as cathode materials for M-CO₂ batteries due to their favorable electronic conductivity, large surface area, chemical stability, low cost, and scalable manufacturing processes. While Super P was among the first carbon materials used as a cathode in M-CO₂ batteries, the electrolyte plays a critical role in determining overall battery performance, even when the electrode materials remain the same. For instance, Ci et al. found that Super P had negligible discharge capacity in ionic liquid (IL) electrolytes^[58]. In contrast, Yang et al. demonstrated significantly improved discharge capacity (6,062 mAh g^-1) in a Li-CO₂ battery using Super P with an ether-based electrolyte. This capacity further increased to 8,229 mAh g^-1 with the addition of Ru metal (Ru@Super P)^[59]. Similarly, Guo et al. developed a porous KB-supported Ru nanoparticle (Ru@KB) composite that enhances the decomposition of Na₂CO₃, lowering charge voltage and improving cycling performance in Na-CO₂ batteries^[60]. With the Ru@KB cathode, the batteries achieved a high discharge capacity of 11,537 mAh g^-1 and a Coulombic efficiency of 94.1%, and ran for over 130 cycles with stable performance^[60]. The disparity in performance is attributed to the electrolyte's effect on the stability of intermediate discharge products and the formation of final products. Early studies on Na-CO₂ batteries with Super P cathodes in different electrolytes confirmed this trend. However, their electrochemical performance is limited by inherent drawbacks, including low conductivity, small pore volume, and limited active sites.

Carbon nanotubes

Commercial CNTs, featuring a porous, cross-linked structure with 5 nm wall thickness and 0.34 nm interlayer spacing, are integral to M-CO₂ batteries. The outer CNT walls facilitate nucleation and anchoring of M₂(CO₃)_n products, while inner walls enhance conductivity and electron transfer. Xiao et al. examined the discharge and charge behavior of Li-CO₂ batteries with CNT electrodes under varying current densities [Figure 3A]^[61]. Voltage profiles revealed two stages: an initial low-voltage plateau decreasing during charging and a sharp rise to completion [Figure 3B]. |dV/dQ| analysis identified three phases: rapid polarization (L0), characteristic peaks (L1), and stabilization (L2). Lower currents prolonged L1 but increased crystalline Li₂CO₃ formation, raising resistance and irreversibility. Pure CNT cathodes often exhibit high charging voltages and electrolyte decomposition^[62]. However, multi-walled CNTs (MWCNTs) have demonstrated improved performance, such as a Li-CO₂ battery with Au nanoparticle-supported CNTs achieving 6,399 mAh g^-1 at 100 mA g^-1[63]. Further advancements will be discussed in the next chapter.

Figure 3. (A and B) Discharge/charge mechanisms and performances of Li-CO₂ batteries with CNT electrodes^[61]. (C) Preparation of NG and TDG and (D and E) long-term cycling performances and stability^[67]. These figures are reproduced with permission from American Chemical Society^[61] and Wiley^[67], respectively.

Graphene

Graphene, with its excellent conductivity, large surface area, and high stability, is an ideal cathode material for Li-CO₂ batteries^[64,65]. Its structure enables efficient oxygen diffusion, ample discharge product storage, and abundant active sites. Zhang et al. developed ultrathin, porous graphene nanosheets with enhanced electrolyte wettability and CO₂ diffusion, achieving discharge capacities of 14,722 and 6,600 mAh g^-1, with a stable 2.77 V plateau at 50 mA g^-1[66]. However, the efficiency and reaction kinetics of graphene-based cathodes remain moderate. Strategies such as heteroatom doping, defect engineering, and compositing with metals or non-metals are being explored to further boost the electrochemical performance of graphene in M-CO₂ batteries.

Defective carbon

Topological defects in carbon structures, such as pentagon and heptagon rings, create highly active sites for CO₂RR in Li-CO₂ batteries. Ye et al. developed defect-rich graphene (TDG) by removing nitrogen dopants from N-doped graphene at high temperatures, introducing topological defects [Figure 3C]^[67]. TDG-1000 exhibited a high surface area (410 m² g^-1) and porous structure, enhancing electrolyte and CO₂ contact, Li-ion transport, and product absorption. As an air cathode, TDG-1000 achieved a discharge capacity exceeding 69,000 mAh g^-1 at 0.5 A g^-1, low overpotential (1.87 V) at 2 A g^-1, and stable cycling for 600 cycles at 1 A g^-1 [Figure 3D and E]. Both DFT and experiments highlighted that topological defects outperform conventional N-doped sites in Li-CO₂ batteries^[67].

N-doped carbon

Nitrogen doping significantly enhances the CO₂RR in carbon-based materials by disrupting charge neutrality and redistributing electrons, imparting unique catalytic properties. The synthesis methods and precursor materials play a pivotal role in modulating heteroatom dopants. Common approaches for N-doped carbon catalysts include: (i) etching carbon with nitrogen sources such as NH₃; (ii) pyrolytic conversion of nitrogen-rich precursors such as MOFs or polymers; and (iii) in situ growth from nitrogen-containing gaseous species^[68-70]. Among nitrogen configurations, pyridinic-N and pyrrolic-N exhibit superior catalytic performance in CO₂RR and CO₂ER compared to graphitic-N, due to stronger interactions with CO₂ and Li that enhance adsorption and reaction kinetics^[71-73]. These active nitrogen species contribute to electronic affinity, wettability, and catalytic activity, while graphitic-N improves electrical conductivity, facilitating charge transport^[74]. Thus, optimizing catalysts with a high content of pyridinic- and pyrrolic-N alongside good conductivity is essential for enhanced performance^[75]. Chen et al. activated nitrogen (N)-doped graphene with CO₂, achieving 72.65% pyridinic- and pyrrolic-N content, which enabled reversible Li₂CO₃ transformation in Li-CO₂ batteries^[76]. This resulted in a low voltage gap of 2.13 V at 1,200 mA g^-1 and excellent cycling stability over 170 cycles at 500 mA g^-1. Additionally, K-CO₂ batteries with N-doped CNT (NCNT) cathodes and KF-rich solid electrolyte interphase (SEI)-protected K anodes achieved 9,436 mAh g^-1, capacity, a 0.81 V overpotential gap at 50 mA g^-1, and 450 cycles. Fiber-shaped batteries also demonstrated stable power output under bending conditions, highlighting their flexibility and potential for next-generation energy storage^[77]. These advancements emphasize nitrogen doping as a crucial strategy for optimizing M-CO₂ battery performance.

N, S co-doped carbon

Carbon and sulfur (S), both p-block elements, share similar electronic structures with p-orbital outer electrons. However, the higher electronegativity of sulfur enables stronger electron attraction and interactions with atoms and functional groups, making S-doped carbon nanomaterials highly attractive for catalysis. Sulfur also acts as a bridge, inducing coupling effects that form robust carbon-based composites^[78-80]. Recently, introducing N and S heteroatoms in carbon catalysts has emerged as a strategy to modulate p-band centers, optimize orbital hybridization, and accelerate CO₂RR and CO₂ER kinetics^[81-84]. Wang et al. demonstrated the vertical growth of an N, S co-doped porous carbon (NS-PC) network on self-supporting carbon paper using salt-template and vacuum-sealed sulfurization methods, guided by DFT calculations^[85]. This two-step synthesis reduced formation energy and yielded a high concentration of dopants [Figure 4A and B]. The study revealed that the NS-PC exhibited a moderate p-band center, resulting in superior CO₂RR and CO₂ER kinetics. The N, S co-doped graphene (NS-G) catalyst, with its moderate p-band center, exhibited the highest CO₂ER catalytic activity [Figure 4C]. Weak adsorption causes premature Na₂CO₃ desorption, while overly strong adsorption increases the energy barrier for Na-O bond dissociation [Figure 4D]. This balance enabled highly reversible formation and decomposition of Na₂CO₃ and carbon, resulting in excellent cycling stability over 1,000 h at 10 μA cm^-2, a low 1.04 V voltage gap, and 71.2% energy efficiency. Moreover, N, S-doped CNTs combined with flexible gel electrolytes have demonstrated superior performance in quasi-solid, flexible planar Li-CO₂ batteries. These systems omit traditional separators, enhancing the reversible formation and decomposition of Na₂CO₃ for improved efficiency and stability^[86].

Figure 4. (A) Synthesis of NS-PC/CP, (B) Formation energies of NS-G. (C) Illustration of p-band center design and (D) the CO₂RR kinetics on NS-G catalysts^[85]. (E) FE-SEM and (F) HR-TEM images of C-BN@600. (G) Reaction coordinates for CO₂ reduction on BN^[88]. (H) Electrochemical performance. These figures are reproduced with permission from Elsevier^[85,88].

N, B co-doped carbon

Additionally, N and boron (B) co-doped porous graphene has been investigated as a bifunctional metal-free cathode catalyst for rechargeable Li-CO₂ batteries^[87]. For example, Qie et al. developed a novel B, N co-doped porous graphene (BN-hG), which serves as an efficient bifunctional cathode catalyst for rechargeable Li-CO₂ batteries^[87]. Due to the unique porous nanostructure and high catalytic activity of the BN-hG catalyst, the newly developed Li-CO₂ battery exhibits low polarization, excellent rate performance, and exceptional long-term cycling stability, achieving 200 cycles at a current density of 1.0 A g^-1. Recently, Kaur et al. developed a tubular B, N co-doped nanotubes-like material (C-BN@600), as shown in Figure 4E and F, from an IL and MOF composite^[88]. The C-BN@600 catalyst efficiently reduces CO₂ to methanol with a 74% Faradaic efficiency and a production rate of 2,665 μg h^-1 mg^-1, driven by the synergistic effects of B and N co-doping that enhance catalytic sites. DFT calculations revealed that B and N play distinct roles: B acts as an electron-withdrawing site, and N as an electron-donating site, together improving the adsorption of intermediates necessary for methanol formation, Figure 4G^[88]. When used in a Zn-CO₂ battery, the system delivered a power density of 5.42 mW cm^-2 and an energy density of 330 Wh kg^-1, while operating stably for over 12 days and 800 cycles [Figure 4H].

Noble metal/carbon composites

Ru catalysts offer a cost advantage among platinum group metals, being the least expensive. Their highly tunable electronic structure and unfilled 4d⁷ orbitals enable superior electrocatalytic activity for CO₂RR^[40,91,92]. For instance, Qiao et al. demonstrated that in Li-CO₂ batteries, Ru enables both Li₂CO₃ and carbon to undergo reversible decomposition (2Li₂CO₃ +C → 2CO₂ + 4Li⁺ + 4e^-, E° = 2.8 V vs. Li/Li⁺), overcoming the 3.82 V thermodynamic limit observed with Au catalysts (2Li₂CO₃ → 2CO₂ + O₂ + 4Li⁺ + 4e^-, E° = 3.82 V vs. Li/Li⁺), where carbon is nearly impossible to decompose^[93]. This is attributed to the catalytic activity of specific Ru crystal facets, which lower reaction barriers and promote reversibility. While the precise catalytic mechanisms of Ru remain partially unclear, its superior performance and economic feasibility have made it a focus of research. Ru AC + SA@NCB integrates Ru clusters (Ru AC) and Ru-N₄ sites, improving carbon desorption, preventing poisoning, and accelerating discharge voltage^[94]. This synergy reduces energy barriers, achieving high rates with low overpotentials. In M-CO₂ batteries, Ru catalysts are often combined with or supported by carbon materials^[7,95,96]. Atomically dispersed Ru anchored on Co₃O₄ nanosheets supported by carbon cloth (SA Ru-Co₃O₄/CC) forms bifunctional catalysts^[97]. This combination provides high electrical conductivity, large surface areas, and well-structured porosity, which are critical for enhancing CO₂ diffusion, ion transport, and uniform Ru distribution. These properties maximize catalytic efficiency while functional groups on the carbon surface interact with Ru, further stabilizing active catalytic sites and enhancing the catalyst's performance. In addition to the excellent catalytic activity of metallic Ru, its oxide, RuO₂, also demonstrates significant catalytic capabilities for reversible M-CO₂ batteries. In our previous work, as depicted in Figure 5A, we developed RuO₂ nanoparticles coated on carbon paper (RuCP) for Na-CO₂ batteries^[98]. Hydrous RuO_xN_yS_z was deposited onto pretreated carbon paper and annealed in nitrogen, producing highly dispersed RuO₂ nanoparticles within a 3D carbon matrix. This design enhances structural stability, promotes sodium nucleation, and improves CO₂RR and CO₂ER kinetics^[98].

Figure 5. (A) Fabrication of RuCP^[98]. (B) CO₂ conversion in Pt-based Li-CO₂ batteries. (C) Adsorption behavior of CO₂, Li and Li₂CO₃ on different orientations of Pt surface^[99]. (D) Stacked Li-CO₂ pouch cell structure. (E) Cycling performance. (F) Energy efficiency comparison with other work^[99]. These figures are reproduced with permission from John Wiley & Sons Australia^[98] and Elsevier^[99], respectively.

Other noble metals, such as Pt, Ir, Au, and Pd, along with their oxides, have also shown promise as effective bifunctional catalysts for M-CO₂ batteries, warranting further exploration^[99-101]. Chen et al. advanced this further by engineering a porous Pt-(111)@CC catalyst using electro-joule heating. Its (111) crystal orientation improved CO₂ conversion kinetics and catalytic activity [Figure 5B and C]^[99]. The catalyst exhibited an exceptionally low overpotential (0.45 V) and high stability over 200 cycles, doubling area-specific capacity. A stacked Li-CO₂ pouch cell incorporating this catalyst demonstrated stable cycling at 20 µA cm^-2 with a low overpotential (0.5 V) and 82.2% energy efficiency, as shown in Figure 5D-F. Compared to prior designs, this porous Pt-based cell showcased superior stability and scalability, highlighting its potential for large-scale applications^[99]. A Pt-N-doped polypyrrole (NPPy)-CNT composite by synthesizing NCNTs (NPPy-CNTs) and depositing Pt nanoparticles via magnetron sputtering was developed by Chen et al.^[102]. This cathode catalyst delivered a high discharge capacity of 29,614 mAh g^-1 with a low overpotential (0.75 V) and stable cycling over 30 cycles. N-doping enhanced conductivity and catalytic activity by creating defects and active sites, while uniform Pt dispersion facilitated Li₂CO₃ decomposition during charging, boosting performance and reversibility^[102]. Zhang et al. designed a Pt/FeNC catalyst for Li-CO₂/O₂ batteries by dispersing 2.4 nm Pt nanoparticles onto porous FeNC microcubic supports^[103]. This catalyst achieved a low overpotential (0.54 V) and stable cycling over 142 cycles, outperforming FeNC and NC catalysts. The synergy between FeNC supports and Pt active sites constrained Li₂CO₃ particle size, reducing decomposition potential and enhancing performance^[103].

Iridium (Ir) and its oxides have shown excellent catalytic performance in CO₂RR, reducing overpotentials, increasing discharge capacity, and influencing discharge product formation. For example, Xing et al. synthesized ultrathin Ir nanosheets supported on N-doped carbon nanofibers (Ir NSs-CNFs), which stabilized intermediate products such as Li₂C₂O₄ during discharge^[104]. This stabilization delayed the formation of plate-like Li₂CO₃, enabling easier decomposition of discharge products at lower charging voltages and improving the cycling performance of Li-CO₂ batteries. IrO₂, in particular, is highly stable and active in oxygen evolution and CO₂-related reactions. IrO₂/carbon fibers derived from kapok (IrO₂/CK) were developed using simple carbonization and hydrothermal processes^[105]. These cathodes demonstrated excellent CO₂ catalytic activity, enhanced electron transfer, and efficient space for Li₂CO₃ deposition, enabling Li-CO₂ batteries to operate for over 4,000 h. Another example is NCNTs modified with ultrafine IrO₂ nanoparticles (IrO₂-N/CNT)^[106]. Batteries using these cathodes achieved high capacities (4,634 mAh g^-1), low overpotentials (3.95/1.34 V), and extended cycling lives (over 2,500 h or 316 cycles). These studies highlight the potential of IrO₂ as a key material for advancing efficient and long-lasting Li-CO₂ batteries. Beyond Ir, noble metals such as Pd have shown promise, with Pd-coated nanoporous gold (NPG@Pd) cathodes for Al-CO₂ batteries achieving high energy efficiency (up to 87.7%) and small discharge-charge voltage gaps (as low as 0.091 V)^[107]. Similarly, PdO has shown potential in enhancing CO₂ reduction efficiencies, particularly in hybrid or composite configurations^[108].

Non-noble metal/carbon composites

Non-noble metals such as Ni, Co, Mo, Cu, Fe, and Mn, along with their composites, are among the most promising catalysts due to their abundance, cost-effectiveness, environmental benefits, and, importantly, their tunable structures and multivalence states. These metals can form a wide variety of oxides, sulfides, nitrides, and phosphides with distinct crystal structures, leading to excellent electrochemical catalytic performance. Numerous nanomaterials, including nanoparticles, nanosheets, and nanorods, have been developed for catalysts such as MoS₂, Mo₂C, WSe₂, and VS₂, all of which demonstrate high catalytic activity^[109,110].

A major challenge in using MoS₂ as a CO₂RR catalyst is its highest catalytic activity being confined to edge sites, coupled with limited electron transfer due to its semiconducting nature. To address this, heterostructures generating localized electric fields can enhance intermediate interactions and overall performance. Naik et al. developed a NiFe₂O₄/MoS₂/MWCNTs heterostructure using a two-step hydrothermal process for Li-CO₂ batteries under simulated Martian conditions^[111]. NiFe₂O₄, a p-type inverse spinel oxide, forms heterojunctions with MoS₂, modifying its electronic properties and enhancing CO₂RR through interfacial interactions. MWCNTs ensure efficient electron flow, reducing charge transfer resistance. This cathode achieved a high discharge capacity of 26,533.5 mAh g^-1 with strong cycling stability, demonstrating its potential for high-performance M-CO₂ batteries, as shown in Figure 6A and B^[111]. Peng et al. synthesized Cu₃P/C nanocomposites through phosphatization of a copper-based MOF precursor for use in Zn-CO₂ batteries, achieving a strong performance with an open-circuit voltage of 1.5 V^[112].

Figure 6. (A) Discharge profiles for Li-CO_{2 Mars} batteries. (B) 1-D charge density profile for Li₂CO₃ adsorption on the NiFe₂O₄/MoS₂^[111]. (C) MOC@NCNF electrochemical reconstruction mechanisms. (D) Discharge and charge profiles, and (E) Cycle performance of MOC@NCNF. (F) Schematic of flexible Li-CO₂ battery. (G) LEDs powered by flexible Li-CO₂ batteries at various bending states^[119]. These figures are reproduced with permission from Elsevier^[111] and Wiley^[119], respectively.

Transition metal catalysts coordinated with nitrogen on carbon supports, particularly those involving Fe, Co, Ni, Cu, and Bi, have shown great promise for CO₂ reduction in M-CO₂ batteries. For example, ultrathin Cu-N₂/GN nanosheets achieved a peak power density of 0.6 mW cm^-2 in Zn-CO₂ batteries, while Fe₁NC/S₁-1000 catalysts with atomic Fe-N₃ sites reached 0.5 mW cm^-2[113]. However, low power densities, with CO as the primary CO₂RR product, remain a challenge. To address this, Bi clusters (BiC/HCS) on hollow carbon spheres achieved a peak power density of 7.2 mW cm^-2, exceeding 200 recharge cycles and demonstrating 68.9% energy efficiency for formate production, paving the way for practical advancements^[110]. Carbon-based transition metal oxides and sulfides in composite materials offer excellent catalytic properties, making them ideal for M-CO₂ battery applications. Liu et al. demonstrated the pivotal role of octahedral Co sites in Co₃O₄, where low eg orbital filling enhanced CO₂-catalyst bonding, leading to a low overpotential of 1.0 V and a 600-h cycle life for Li-CO₂ batteries^[114]. Co/Co₉S₈ nanoparticles anchored on S, N co-doped porous carbon (Co/Co₉S₈@SNHC) exhibited superior catalytic activity in Na-CO₂ batteries, achieving an areal capacity of 18.9 mAh cm^-2 at 0.5 mA cm^-2[115]. Xu et al. developed Fe-Cu-N-C catalysts with synergistic Fe/Fe₃C nanocrystals, Fe-N_x, and Cu-N_x sites, achieving a low voltage gap of 0.44 V and over 1,550 cycles, showcasing excellent durability and performance^[75].

Manganese-based catalysts excel in CO₂RR and CO₂ER due to their diverse crystal structures, valence states, and electronic configurations^[116-118]. Liu et al. developed a 3D nanofiber framework with dual Mn active sites (MOC) supported on N-doped carbon nanofibers (MOC@NCNF) via electrospinning^[119]. They observed in situ electrochemical reconstruction between Mn(II) and Mn(III) during cycling [Figure 6C]. During charging, Mn(II) oxidizes to Mn(III), generating abundant Mn(III) active sites, which revert to Mn(II) during discharge. DFT calculations revealed that Mn(II) lowers energy barriers for CO₂RR intermediates, while Mn(III) activates Li₂C₂O₄ during CO₂ER, facilitating Li-O bond cleavage. Li-CO₂ batteries with MOC@NCNF cathodes achieved high discharge capacity (10.31 mAh cm^-2), energy efficiency (64.94%), and prolonged cycle life (327 cycles, 1,308 h) at 50 µA cm^-2 [Figure 6D and E]. The flexible MOC@NCNF electrode also supports wearable electronics, maintaining performance under bending and folding [Figure 6F and G].