CO₂ utilization in energy storage and conversion

Abstract

The world is facing the pressing dual challenge of an escalating energy crisis and the excessive accumulation of greenhouse gases. Current researchers are now reimagining carbon dioxide (CO₂) not simply as a waste to be captured, but as a valuable energy carrier to be utilized in next-generation energy systems. However, a comprehensive understanding of how CO₂ participates in energy conversion and storage remains limited. This review addresses this critical knowledge gap by systematically examining four innovative types of CO₂-based energy technologies. These include metal-CO₂ batteries and CO₂ capture-integrated storage systems that leverage the electrochemical activity of CO₂ for efficient and sustainable energy storage, as well as molten carbonate fuel cells and CO₂-based electricity generators that generate electricity energy in innovative approaches. These emerging systems leverage the physicochemical properties of CO₂ to create transformative energy solutions. This review highlights the pivotal role of CO₂, examines critical challenges, and proposes forward-looking strategies for future energy development.

---

Gemeng Liang

Dr Gemeng Liang received his PhD in 2021 from the University of Wollongong, Australia, and is currently an Australian Research Council (ARC) DECRA Fellow at the University of Adelaide. His research centres on energy storage and conversion, with a focus on Li-ion, Na-ion, and K-ion batteries. He specialises in applying synchrotron and neutron-based characterisation techniques to probe structural evolution, phase transitions, and reaction mechanisms in functional materials. His work seeks to deepen the fundamental understanding of electrochemical processes and inform the rational design of next-generation energy storage devices with enhanced performance and durability.

---

Introduction

Human activities, particularly the large-scale combustion of fossil fuels, have significantly increased atmospheric carbon dioxide (CO₂) concentrations from approximately 280 ppm in pre-industrial times to over 425 ppm by 2024. This rapid rise has contributed to a global temperature increase of 1.5 °C.^1,2 In addition to global warming, elevated CO₂ levels have also resulted in more frequent extreme weather events, ocean acidification, and widespread ecological disruptions, posing serious threats to the sustainable development of human society and the natural environment.³ Consequently, reducing atmospheric CO₂ concentrations is imperative for mitigating climate change, preserving ecosystem stability, and achieving global carbon neutrality targets.⁴

The ongoing energy crisis poses a significant threat to modern society, affecting nearly every aspect of daily life. Driven by the depletion of fossil fuel resources, geopolitical instability, and the urgent need for decarbonization, the crisis has led to high energy prices, disrupted supply chains, and increased pressure on energy security. These challenges not only hinder economic growth but also limit access to affordable energy, and impede the global transition toward sustainable development. In response to the energy crisis, significant efforts have been devoted to advancing renewable energy sources, enhancing energy efficiency, and developing advanced energy producing and storage technologies.⁵

Among emerging solutions, the integration of CO₂ utilization into energy conversion systems has opened new paradigms, where CO₂ is captured and utilized as a functional reactant contributing to energy storage and producing.⁶ This approach leverages the electrochemical reactivity of CO₂ to participate in redox reactions, enabling the design of novel energy storage cells and electricity generation devices. Furthermore, coupling CO₂ conversion with renewable electricity provides a sustainable pathway to close the carbon loop, reduce reliance on fossil fuels, and enhance overall energy system efficiency.^7,8 As such, CO₂-based energy producing and storage technologies represent a promising intersection of carbon management and clean energy innovation. Despite rapid advancements in CO₂-utilizing energy technologies, the field still lacks a comprehensive and cohesive review that connects scattered breakthroughs into a clear and guiding framework, leaving a critical gap in the literature that must be addressed to drive future innovation.

This timely review critically examines the emerging strategy of integrating CO₂ utilization into energy producing and storage, an area at the nexus of carbon mitigation and clean energy innovation. We focus on four major approaches: metal–CO₂ batteries, CO₂ capture-based energy storage systems, molten carbonate fuel cells, and CO₂-based electricity generators (Fig. 1). Their working principles, material design strategies, technical challenges, and application prospects are systematically discussed. By consolidating recent progress and outlining future directions, this work provides essential insights to advance this rapidly evolving and impactful field.

Fig. 1 Summary of CO₂ utilization in energy storage and producing. Four systems including metal–CO₂ batteries, CO₂ capture-based energy storage, molten carbonate fuel cells, and CO₂-based electricity generators are illustrated. The mechanism schematic of CO₂-based electricity generators is reproduced with permission.⁹ Copyright 2024, Springer Nature.

CO₂ utilization in metal–CO₂ batteries

Metal–CO₂ batteries (MCBs) have emerged as a promising class of energy storage systems that simultaneously address carbon emissions and deliver high energy output.¹⁰ Depending on the metal anode and electrolyte environment, these systems can convert CO₂ into a diverse array of discharge products. It employs metals such as lithium (Li), sodium (Na), potassium (K), or zinc (Zn) as the anode, with CO₂ serving as the cathodic reactant.¹¹ During discharge, the battery generates carbonates, bicarbonates, or other carbon-containing compounds.¹² For clarity and systematic understanding of the CO₂ conversion, MCBs can be classified into four groups, including carbonate with carbon, oxalate, carbonate and CO, as well as value-added carbon-containing compounds such as formate and methane (Table 1).

Table 1 Performance summary of typical metal-CO₂ batteries

CO2 conversion products	Battery type	Cathode catalysts	Electrolyte	Discharge/charge voltage	Energy efficiency	Max discharge capacity	Stability	Ref.
Carbonate and carbon	Li–CO2	NiCo2S4	1 M lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) in tetraethylene glycol dimethyl ether (TEGDME)	2.67/3.82 V at 20 μA cm^−2	88.7%	5528 μA h cm^−2	600 h at 20 μA cm^−2	13
		Ru/CNTs	1 M LiTFSI in dimethyl sulfoxide (DMSO)	2.8/3.75 V at 100 mA g^−1	—	11 888 mA h g^−1 at 100 mA g^−1	2000 h at 100 mA g^−1	12
		Fe2O3@CoS	1 M LiTFSI in TEGDME	2.9/3.1 V 20 μA cm^−2	92.4%	4900 μA h cm^−2	550 h at 20 μA cm^−2	14
		NiFeCoCuRu high entropy alloys	1 M LiTFSI in DMSO	2.85/3.72 V at 100 mA g^−1	—	—	2900 h at 100 mA g^−1	15
		Hydrogen-bonded organic framework-FJU-1-Ru@CNT	1 M LiTFSI in TEGDME	2.79/3.89 V	65.77%	24 245.3 mA h g^−1 at 10 mA g^−1	200 cycles at 1 A g^−1 with 1000 mA h g^−1	16
	Na–CO2	Tetraethylene glycol dimethyl-treated multi-wall carbon nanotube	1 M NaClO4 in TEGDME	2.6/3.2 V at 1 A g^−1	81%	60 000 mA h g^−1 at 1 A g^−1	200 cycles with 2000 mA h g^−1	17
		Metal-free porous carbon	1 M NaClO4 in TEGDME	2.28/3.32 V	71.2% at 10 μA cm^−2	2422 μA h cm^−2	1000 h at 10 μA cm^−2	18
		Activated multiwall carbon nanotubes	[Poly(vinylidene fluoride-co-hexafluoropropylene)]-4% SiO2/NaClO4–TEGDME	2.75/3.49 V	—	5000 mA h g^−1 at 50 mA g^−1	400 cycles at 500 mA g^−1	19
	K–CO2	Bamboo-like N-doped carbon nanotube	1.0 M potassium bis(trifluoromethanesulfonyl)imide (KTFSI) in TEGDME	2.40/4.01 V	53–62%	9436 mA h g^−1 at 200 mA g^−1	450 cycles or 3100 h with capacity of 500 mA h g^−1	20
		Carboxyl-containing multi-walled carbon nanotubes	0.3 M KPF6 in TEGDME	2.2/3.0 V	—	3681 mA h g^−1 at 100 mA g^−1	400 cycles	21
	Mg–CO2	CNTs	MgCl2 and Mg(TFSI)2 in DME	Discharge at 1 V at 200 mA g^−1	—	—	250 h	22
	Al–CO2	NPG@Pd	AlCl3 and 1-ethyl-3-methylimidazolium chloride with a mole ratio of 1.3	0.72/0.74 V	87.7%	638 mA h g^−1	30 cycles at 333 mA g^−1	23
Oxalate	Li–CO2	Cu(II) MOF (benzene-1,3,5-tricarboxylic acid)/CNTs	1 M LiTFSI in TEGDME	2.8/3.7 V	—	9040 mA h g^−1 at 100 mA g^−1	400 cycles at 500 mA g^−1	24
		Mo2C/carbon nanotube	1 M LiCF3SO3/TEGDME	2.8/3.5 V	77%	1150 mA h at 20 mA	40 cycles at 20 mA	25
		MoN nanofibers on carbon-cloth	1 M LiTFSI/TEGDME	2.83/3.19 V	88%	6542.9 μA h cm^−2	86 cycles at 100 μA cm^−2	26
	Na–CO2	Co-encapsulated N-doped carbon framework	Gel electrolyte	Voltage gap: 1.7 V	—	3094 mA h g^−1	366 cycles (2200 h) at 0.1 mA cm^−2	27
		Ni0·95Fe0·05(OH)2	Sodium ion conducting solid-state electrolyte	3.35/3.5 V	—	1166 mA h g^−1	100 h	28
	Mg–CO2	Nitrogen-doped carbon nanotubes	0.25 M Mg(TFSI)2-TEGDME with 1,3-propylene amine additives	0.75/2.25 V	—	600 mA h g^−1 at 200 mA g^−1	70 cycles, more than 400 h at 200 mA g^−1	29
	Al–CO2	Ketjen black	AlCl3: 1-ethyl-3-methylimidazolium chloride ionic liquid electrolyte	1.12/1.17 V	—	9394 mA h gcarbon^−1	1000 h at 0.2 mA cm^−2	30
Carbonate and CO	Li–CO2	Porous fractal Zn	1 M LiTFSI/TEGDME	1.91 V/–	67%	—	1500 min during discharge	31
Carbon-containing compound products	Zn–CO2	BiOF	0.8 M KHCO3 saturated with CO2	1.2 V/–	70.74%	—	200 cycles and 68 h	32
		3D porous Pd interconnected nanosheets	Catholyte: 1 M NaCl; anolyte: 1 M KOH added with 0.02 M Zn(CH3COO)2	Charge voltage: 1 V	81.2 %	—	Over 100 cycles	33
		Ni Nanoclusters and single atom site	Catholyte: 1 M KOH; anolyte: 6 M KOH and 0.2 M Zn(CH3COO)2	0.65 V	—	—	1200 cycles and 420 h at 5 mA cm^−2	34
		Ir@Au bimetal nanomaterial	Catholyte: 0.8 M KHCO3; anolyte: 0.8 M KOH with 0.02 M Zn(CH3COO)2	Charge voltage: 2 V	68%	—	90 cycles at 0.9 mA (5 mA cm^−2)	35

---

MCBs with carbonate and carbon products

In MCBs, Li-, Na-, K-, Mg-, and Al-based systems are the primary configurations known to follow a discharge pathway that yields carbonate and carbon as the main products.³⁶ During discharge, the metal anode is oxidized, releasing electrons to form metal ions (Mⁿ⁺, n = 1, 2, or 3), while CO₂ molecules at the cathode gain electrons and react with Mⁿ⁺ to generate metal carbonate and carbon (see eqn (1) and (2)). The reactions are reversible, offering significant potential for efficient energy conversion and storage. This type of reaction is typically carried out in organic electrolytes, with common choices including metal TFSI salts dissolved in TEGDME or DMSO.

Anode: M → Mⁿ⁺ + e⁻ (1)

Cathode: Mⁿ⁺ + CO₂ + e⁻ → M₂(CO₃)_n + C (2)

Among these MCBs, Li–CO₂ battery has emerged as one of the most extensively studied configurations due to its large theoretical specific energy (∼1876 W h kg⁻¹), high cell voltage of 2.8 V, and the largest theoretical specific capacity of Li (3861 mA h g⁻¹). During discharge, the valence state of carbon in CO₂ shifts from +4 to 0 (eqn (3)), which proceeds through multiple intermediates and reaction steps. The CO₂ reduction starts with the adsorption of CO₂ on the catalytic sites at the cathodes. Upon accepting two electrons, the CO₂ molecules are converted into oxalate anions (C₂O₄²⁻), as described by eqn (4). The C₂O₄²⁻ is usually unstable and undergoes a disproportion reaction to form CO₂²⁻ and CO₂ through eqn (5). This is followed by the coupling step of C₂H₄²⁻ with CO₂²⁻ to produce CO₃²⁻ and amorphous carbon (C) via eqn (6). Finally, the CO₃²⁻ species readily combine with lithium ions (Li⁺) derived from the anode (eqn (7)), resulting in the formation of lithium carbonate (Li₂CO₃) products. The nucleation and growth of Li₂CO₃ and carbon in Li–CO₂ batteries are fundamentally governed by the interaction strength between reaction intermediates and the catalyst surface.¹⁰ There are two distinct growth mechanisms that critically influence the morphology of discharge products and the overall battery performance. When the binding energy between intermediates and the catalyst (e.g., Co₃O₄-based cathodes) is weak, product formation proceeds via a solvation-mediated growth mechanism.³⁷ In this case, Li₂CO₃ tends to nucleate in the electrolyte near the cathode surface, leading to the development of bulky, flower-like structures. These discharge products can deliver high capacity but often compromise charge efficiency due to their large size and sluggish decomposition. In contrast, strong adsorption between intermediates and the cathode (e.g., MnOOH arrays on stainless-steel (SS) mesh cathodes) drives a surface adsorption growth mechanism, resulting in thin, uniform layers of Li₂CO₃ directly on the electrode.³⁸ This morphology is more conducive to efficient charge reactions, enhancing reversibility and energy efficiency, though it may limit the achievable capacity. Balancing these two mechanisms is therefore key to optimizing the electrochemical performance of Li–CO₂ batteries.

4Li⁺ + 3CO₂ + 4e⁻ → 2Li₂CO₃ + C (3)

2CO₂ + 2e⁻ → C₂O₄²⁻ (4)

C₂O₄²⁻ → CO₂²⁻ + CO₂ (5)

C₂O₄²⁻ + CO₂²⁻ → 2CO₃²⁻ + C (6)

CO₃²⁻ + 2Li⁺ → 2Li₂CO₃ (7)

A critical limitation to the rechargeability of Li–CO₂ batteries lies in the decomposition of discharge products during the charging process. Li₂CO₃, a wide-bandgap insulating compound with inherently low solubility in typical organic electrolytes, limit the battery reversibility and rechargeability.³⁹ Its decomposition typically demands a high overpotential, largely determined by the electrochemical properties at the CO₂/electrolyte/catalyst interface. Under optimized conditions, where a catalytically active interface is established, Li₂CO₃ and the accompanying amorphous carbon can be decomposed simultaneously, thereby closely reversing the discharge reaction (eqn (7)). Achieving such favorable conditions typically requires the integration of highly active catalysts, such as Ru, Ir, Ni, or Mn, with specially designed electrolytes (e.g., water-in-salt systems or ionic liquids) that support interfacial reactivity and Li⁺ ion transport.⁴⁰ However, when the interfacial environment is suboptimal with inactive catalysts, incompatible electrolytes, or ineffective surface contact between solid catalysts and discharge products, carbon fails to participate in the decomposition, and only Li₂CO₃ undergoes direct electrochemical decomposition at elevated voltages. This alternative pathway often triggers the formation of reactive oxygen species such as singlet oxygen (¹O₂), as described in eqn (8) and (9).^36,41,42 These species are known to aggressively degrade organic electrolytes and carbon cathode structures, accelerating parasitic reactions and compromising cell stability. Although Li–CO₂ batteries are technically rechargeable along both pathways, the latter route is plagued by irreversible side reactions, accumulation of unwanted by-products, and poor cycling performance. Therefore, elucidating the fundamental discharge–charge pathways, particularly the formation and decomposition mechanisms of Li₂CO₃, carbon, and associated intermediates, is critical for the rational design of cathode materials and electrolytes.

2Li₂CO₃ → 2CO₂ + O₂˙⁻ + 4Li⁺ + 3e⁻ (8)

2Li₂CO₃ → 2CO₂ + ¹O₂ + 4Li⁺ + 4e⁻ (9)

The formation and deposition of carbonate and carbon as discharge products have been confirmed through a range of advanced characterization techniques. Since Li₂CO₃ typically grows partially or entirely in crystalline form on the cathode surface, its characteristic peaks are prominently observed in the XRD patterns.¹² On the other hand, the carbonate ion is Raman- and IR-active, making both spectroscopic techniques effective tools for confirming the presence of lithium carbonate. However, one main challenge lies in verifying carbon (C) as a genuine discharge product, as this requires ruling out interference from the commonly used carbon-based gas diffusion electrodes, which can contribute background carbon signals. To address this, Yang et al. employed a Ru catalyst nanoparticle-decorated Ni foam (Ru@Ni) cathode to specifically identify carbon products.⁴³ Their ex situ Raman measurements revealed clear D and G bands of carbon on the fully discharged cathode, thereby providing direct evidence of carbon formation during discharge. Qiao et al. employed in situ surface-enhanced Raman spectroscopy (SERS) to study the CO₂ conversion pathways in an in situ Raman Li–CO₂ cells (0.5 M LiClO₄ in DMSO, saturated with CO₂) (Fig. 2a).⁴⁴ To construct the catalytic cathode for SERS analysis, a thin Ru film was deposited onto the Au substrate supported on the Al mesh via radio frequency magnetron sputtering. By replacing the conventional carbon-based gas diffusion layers, this electrode configuration effectively eliminates interference from pre-existing carbon species, thereby ensuring in situ identification of newly formed carbonaceous products. The Raman spectra exhibited a distinct, sharp peak at 1085 cm⁻¹, corresponding to the symmetric stretching vibration of Li₂CO₃, alongside two broader bands at 1319 cm⁻¹ and 1587 cm⁻¹, attributable to the disordered (D-band) and graphitic (G-band) vibrations of the amorphous carbon, respectively, confirming the co-formation of Li₂CO₃ and amorphous C. The two discharge products show synchronously increasing trend (Fig. 2b). To provide more quantitative information on the discharge products, the H₂SO₄ is injected to the cathode after discharge with 0.6 mA h capacity, followed by gas analysis using gas chromatography-mass spectrometry (GC-MS) (Fig. 2c). The inset pie chart presents the proportion of formed Li₂CO₃, determined from the integrated area under the CO₂ evolution rate relative to the total amount of CO₂ reduced, as estimated from the discharge capacity. Moreover, the in situ differential electrochemical mass spectrometry (DEMS) serves as another powerful technique to monitor real-time CO₂ consumption and gaseous product evolution, providing critical insights into reaction stoichiometry and product formation. In the proposed reaction pathway of eqn (3), the theoretical charge-to-mass ratio (e⁻/CO₂) is 1.33, reflecting the number of electrons required per CO₂ molecule involved. Wang et al. also employed DEMS to quantify both electron transfer and CO₂ conversion during the operation of a Li–CO₂ cell.⁴⁵ Their results revealed that a discharge and charge capacity of 0.2 mA h corresponded to approximately 5.6 μmol of CO₂ consumption or release, coupled with 7.48 μmol of electron transfer. Specifically, during the discharge process, 5.3 μmol of CO₂ was consumed, yielding a calculated e⁻/CO₂ ratio of 1.41. Upon charging, 5.1 μmol of CO₂ was released, corresponding to a slightly higher e⁻/CO₂ value of 1.47. These results align well with the theoretical prediction for the reaction pathway yielding Li₂CO₃ and carbon as the main discharge products, thereby offering strong empirical support for the proposed mechanism. Additionally, the integration of in situ GC-MS with isotope tracing techniques enables precise monitoring of gas evolution and unambiguous identification of the carbon origin in the evolved CO₂ during the charging process. Zhou et al. preloaded the cathode with Li₂CO₃ and ¹³C for subsequent charging, employing a carbon-only cathode and a LiCF₃SO₃/tetraglyme electrolyte (molar ratio 1 : 4).⁴⁶ During the charge process, CO₂ with a mass-to-charge ratio (m/z) of 44 was identified as the dominant gaseous product, indicating that carbon did not participate in the charge reaction under these conditions. However, the information from in situ experiment is still limited by current instrumentation and analytical techniques, such as inadequate temporal and spatial resolution. Furthermore, the typically low concentrations of intermediates and products in Li–CO₂ batteries result in weak signal intensities and significant spectral overlap, which further complicate the identification of CO₂-involved reaction pathways.

Fig. 2 Overview of MCBs with carbonate and carbon as discharge products. (a) Conceptual illustration of an energy storage system based on Li–CO₂ electrochemistry technology. (b) Galvanostatic discharge profiles and corresponding Raman signal intensity of a Li–CO₂ cell comprising a sputtered gold cathode, a CO₂-saturated 0.5 M LiClO₄-DMSO electrolyte, and a lithium metal foil. The cell was tested under a current of 5 μA with a capacity cutoff of 10 μAh. (c) CO₂ gas evolution rate during the quantification of carbonate species on a cathode previously discharged to 0.6 mA h. The blue arrow denotes the injection of H₂SO₄, which decomposes Li₂CO₃ to release CO₂. The inset pie chart shows the yield of fixed Li₂CO₃ relative to the total reduced CO₂, as determined by integrating the CO₂ evolution curve and correlating it with the discharge capacity. (a)–(c) Are reproduced with permission.⁴⁴ Copyright 2017, Cell Press. (d) Schematic of p-band center modulation in nitrogen and sulfur co-doped carbon catalysts, along with the catalyst synthesis strategy for Na–CO₂ batteries. (e) Rate performance of batteries with the four cathodes, including N, S co-doped porous carbon on carbon paper (NS-PC/CP), N-doped porous carbon on carbon paper (N-PC/CP), S-doped porous carbon on carbon paper (S-PC/CP), and the porous carbon on carbon paper (PC/CP). (d) and (e) Are reproduced with permission.¹⁸ Copyright 2024, Elsevier. (f) Ex situ Raman spectra of a K–CO₂ battery during discharge and charge. (g) In situ DEMS results for the K–CO₂ battery with a B-NCNT cathodes under discharge current of 200 μA within a capacity limit of 1000 μAh. The green, orange, and purple dotted lines represent theoretical gas evolution rates for CO₂, CO, and O₂, respectively, corresponding to potential reaction pathways. (f) and (g) Are adapted with permission.²⁰ Copyright 2024, Wiley.

In summary, Li–CO₂ batteries exhibit several notable advantages: (1) a high theoretical energy density (∼1876 W h kg⁻¹), which makes them highly attractive for long-duration and high-capacity energy storage applications; (2) the direct utilization of CO₂ as a reactant, offering a promising route for carbon mitigation and contributing to carbon neutrality goals; and (3) the use of a lightweight lithium metal anode, which helps reduce overall system weight while enhancing the energy output per unit mass. However, despite these merits, the practical application of Li–CO₂ batteries is currently hindered by several critical challenges: (1) poor rechargeability arising from the formation of electrochemically inert and insulating Li₂CO₃ discharge products, which impede reversible reactions; (2) low energy efficiency and short cycle life, mainly due to sluggish reaction kinetics and side reactions; and (3) significant safety concerns, including dendrite growth on the lithium anode and its interaction with electrolytes and impurities (such as moisture or oxygen). Therefore, overcoming these limitations through material innovation, electrolyte engineering, and system design will be essential for enabling Li–CO₂ batteries to emerge as a viable and sustainable alternative to conventional lithium-ion battery technologies.

In comparison to Li–CO₂ batteries, Na–CO₂ batteries offer a compelling alternative due to the abundance and low cost of Na metal. In Na–CO₂ batteries, the formation and decomposition of carbonate and carbon represent the predominant and most widely reported electrochemical reactions.^17,47 A representative reaction, 4Na + 3CO₂ → 2Na₂CO₃ + C (ΔrG° = −905.6 kJ mol⁻¹), exhibits a lower Gibbs free energy compared to the analogous reaction in Li–CO₂ batteries (eqn (3), ΔrG° = −1081 kJ mol⁻¹). This reduced reaction Gibbs free energy in Na–CO₂ systems suggests the potential for achieving lower charge voltages, thereby minimizing the risk of electrolyte decomposition during cycling. Hu et al. developed a rechargeable room-temperature Na–CO₂ battery, composed of a sodium metal anode, a glass fiber separator, an ether-based electrolyte, and a tetraethylene glycol dimethyl ether-treated multi-walled carbon nanotube (t-MWCNT) cathode supported on a binder-free Ni mesh.¹⁷ The battery demonstrated an impressive reversible capacity of 60 000 mA h g⁻¹ at 1 A g⁻¹ (corresponding to approximately 1000 W h kg⁻¹), and maintained stable cycling over 200 cycles with a controlled capacity of 2000 mA h g⁻¹ and a charge voltage of below 3.7 V. Recently, Wang et al. developed a nitrogen and sulfur co-doped porous carbon cathode featuring high heteroatom content and a moderately positioned p-band center for application in Na–CO₂ batteries.¹⁸ The active sites of the carbon catalyst are engineered through the incorporation of nitrogen or sulfur heteroatoms, which modulate the p-band center to tailor orbital hybridization, thereby enhancing the kinetics of both CO₂ reduction and evolution reactions (Fig. 2d and e). This system demonstrated exceptional cycling performance of over 1000 hours with a small voltage gap of just 1.04 V and a high energy efficiency of 71.2% at a current density of 10 μA cm⁻². These results represent a remarkable improvement over most Li–CO₂ batteries that operate via the carbonate and carbon formation pathway, underscoring the potential of Na–CO₂ systems for CO₂ utilization in energy storage. Despite a relatively lower theoretical energy density of 1125 W h kg⁻¹, their economic viability and material sustainability have drawn significant interest from both scientific and industrial communities.

K–CO₂ batteries have recently emerged as a promising alternative to Li–CO₂ and Na–CO₂ systems. Similarly, K–CO₂ batteries utilize the electrochemical conversion of CO₂ to K₂CO₃ and carbon.²¹ The standard redox potential of K (K/K⁺ = −2.92 V vs. SHE) enables a high theoretical discharge voltage of 2.48 V (based on K₂CO₃ formation), surpassing that of Na–CO₂ batteries (2.35 V). Beyond its favorable thermodynamics, K⁺ offers additional intrinsic advantages. In comparision to Li⁺ and Na⁺, K⁺ exhibits weaker Lewis acidity and a smaller Stokes radius (3.6 Å in propylene carbonate), which facilitating faster ion migration in the electrolyte and across the electrode–electrolyte interface.⁴⁸ This kinetic advantage, coupled with the earth abundance and low cost of K, positions K–CO₂ batteries as a promising solution for scalable and economically viable CO₂-based energy storage system. Zhang et al. developed a K–CO₂ battery using bifunctional metal-free 3D porous carbon electrocatalysts (N doped carbon nanotubes (N-CNTs)) for CO₂ reduction and evolution.⁴⁹ The 3D carbon network cathode effectively prevents graphene restacking, enhances active site exposure, offers space for discharge product deposition, and facilitates efficient electron, ion, and CO₂ transport. As a result, the battery exhibits high reversibility over 250 cycles at the limited specific capacity of 300 mA h g⁻¹. Li et al. reported a K–CO₂ battery involving passivated K anodes with artificial surface film and N-doped carbon nanotubes cathodes.²⁰ As shown in the Raman spectra (Fig. 2f), the D and G bands of carbon at 1335 and 1560 cm⁻¹ are consistently observed throughout the redox process. A distinct peak at 1030 cm⁻¹, attributed to the vibrational modes of K₂CO₃, appears during discharge. In situ DEMS analysis (Fig. 2g) confirms that CO₂ consumption begins with discharge and closely follows the theoretical reaction: 4K⁺ + 3CO₂ + 4e⁻ → 2K₂CO₃ + C. Notably, CO evolution is negligible, indicating that CO₂ reduction proceeds predominantly via K₂CO₃ and C formation. This battery achieves a high specific discharge capacity (9436 mA h g⁻¹), maintains excellent rate performance with a small voltage gap (0.81 V at 50 mA g⁻¹), and demonstrates excellent durability, lasting 450 cycles or 3100 hours under a limited capacity of 500 mA h g⁻¹. Despite these achievements, K–CO₂ systems are still in the early stages of development. Limited by the lack of efficient bifunctional catalysts, their electrochemical performance remains unsatisfactory, and fundamental reaction mechanisms are yet to be fully understood. Nevertheless, the K–CO₂ battery shows strong potential as a future low-cost and sustainable energy storage system.

Mg–CO₂ batteries offer a competitive advantage due to the high theoretical volumetric capacity of Mg (3833 mA h cm⁻³) and abundant resources, allowing for more CO₂ fixation per cubic meter at a lower cost. Additionally, the dendrite-free nature of magnesium alleviates safety concerns. In Mg–CO₂ batteries, CO₂ can be converted into carbonate and carbon in organic electrolyte systems. For example, Zhang et al. developed a moisture-assisted Mg–CO₂ battery that showed over 250 hours of cycle life and maintained a discharge voltage above 1 V at 200 mA g⁻¹.²² Theoretical calculations indicate that MgCO₃·3H₂O forms and decomposes more easily than MgCO₃, supported by negative adsorption energy and stretched bonds. Raman spectroscopy confirmed the presence of CO₃²⁻ at 1120 cm⁻¹ and carbon D and G bands at 1350 cm⁻¹ and 1580 cm⁻¹, respectively, validating the reaction pathway in eqn (10). However, research on Mg–CO₂ batteries is still limited, and improving the cycling efficiency and energy density remains one of the key challenges in designing high-performance Mg–CO₂ batteries.

Reversible Al–CO₂ batteries based on the reaction in eqn (11) have demonstrated the feasibility of converting CO₂ into carbonate and carbon. Al not only exhibits a much higher intrinsic safety than Li, Na, and K, but also offers a high theoretical specific capacity of 2978 mA h g⁻¹, positioning it as a promising candidate for metal-CO₂ battery systems. Ma et al. designed a rechargeable Al–CO₂ battery employing Al foil as the anode, an ionic liquid electrolyte, and Pd-coated nanoporous gold (NPG@Pd) as the cathode catalyst.²³ This system delivered an exceptionally low voltage gap of 0.091 V and a high energy efficiency of 87.7%, indicating remarkable electrochemical performance. Nonetheless, its practical application remains hindered by limited cycling stability and an unclear reaction mechanism, with insufficient mechanistic insight from in situ or operando characterizations. Therefore, the rational design of high-performance bifunctional catalysts and the elucidation of fundamental reaction pathways will be critical to unlocking the full potential of Al–CO₂ batteries.

2Mg + 3CO₂ + 6H₂O ↔ 2MgCO₃·3H₂O + C (10)

4Al + 9CO₂ ↔ 2Al₂(CO₃)₃ + 3C (11)

MCBs with oxalate products

In MCBs, the formation of carbonate and carbon is commonly attributed to the thermal instability of the intermediate oxalate species (C₂O₄²⁻), which undergoes disproportionation reactions. However, if C₂O₄²⁻ can be stabilized at the CO₂/catalyst/electrolyte interface, the following disproportionation reactions can be effectively suppressed, enabling oxalate to be the final discharge product.⁵⁰ Compared to carbonate, which is typically insulating and insoluble, oxalate is more easily decomposed (+0.81 V vs. NHE), offering a potentially reversible and energy–efficient reaction pathway.

In Li–CO₂ batteries, since the initial report by Hou et al. in 2017, which demonstrated that Mo-based catalysts enabled the formation of Li₂C₂O₄ as the sole discharge product, significant efforts have been devoted to stabilizing oxalate intermediates to modulate the reaction pathway and enhance battery performance.²⁵ To date, researchers have successfully employed strategies involving catalyst design, redox mediators (RMs), and electrolyte engineering to benefit the selective formation of Li₂C₂O₄, thereby achieving substantial improvements in energy efficiency, reversibility, and cycling stability.

Recent studies have demonstrated that Mo-based catalysts, such as Mo₂C, MoN, and Mo₃P, are highly effective in stabilizing Li₂C₂O₄ as the primary discharge product in Li–CO₂ batteries. The formation of a Mo–O coupling bridge plays a critical role in preserving Li₂C₂O₄ and enhancing reaction kinetics. For instance, Mo₂C catalysts have been extensively investigated through experimental and computational approaches, showing exclusive Li₂C₂O₄ generation, a reduced charging voltage of 3.5 V, and an energy efficiency of 77%. DFT calculations by Yang et al. revealed that the disproportionation of Li₂C₂O₄ is thermodynamically unfavorable on Mo₂C, ensuring Li₂C₂O₄ stability.⁵¹ Similarly, Cheng et al. developed MoN nanofibers on carbon cloth (CC@MoN NFs), where the delocalized electrons of Mo³⁺ stabilize Li₂C₂O₄ through weak Mo–O bonds, achieving a low charge potential of 3.19 V, 88% energy efficiency, and >500 hours of cycle life.²⁶ Moreover, a Mo₃P/Mo Mott–Schottky heterojunction catalyst reported by Wang et al. delivered a high specific capacity of 10 577 mA h g⁻¹, a low overpotential of 0.15 V, and high energy efficiency of 94.7% by promoting reversible Li₂C₂O₄ conversion and accelerating interface kinetics.⁵² These findings collectively highlight the critical role of Mo-based catalysts in effectively stabilizing oxalate intermediates and advancing the development of high-performance Li–CO₂ batteries.

The use of RMs has emerged as another effective strategy to promote Li₂C₂O₄ formation and enhance the electrochemical performance of Li–CO₂ batteries. In a typical RM-assisted discharge process, the RM first reacts with CO₂ to form an RM-CO₂ adduct, which subsequently undergoes reduction to regenerate the RM and yield Li₂C₂O₄. Acting as a liquid catalyst, the RM enables multidirectional interaction with CO₂, thereby significantly accelerating cathode reaction kinetics. Typical RMs, such as tris(2,2′-bipyridyl)-dichloro-ruthenium(II) (Ru-(bpy)₃Cl₂), binuclear copper(I) complexes, cupric chloride (CuCl₂), and 2,2,6,6-tetramethylpiperidoxyl, have been demonstrated to stabilize Li₂C₂O₄.^53–56 The advantages of RM-assisted discharge/charge process include ultra-low overpotentials and exceptionally high energy efficiencies. For instance, Wang et al. demonstrated that the 2,2,6,6-tetramethylpiperidoxyl RM enabled the preferential formation of Li₂C₂O₄, delivering an outstanding discharge voltage of 2.97 V and a low charge voltage of approximately 3.15 V across a wide current range (100 to 800 mA g⁻¹), indicating excellent rate capability.⁵⁶ However, the application of soluble RMs also faces challenges, such as the shuttle effect and parasitic reactions with the anode, which can deteriorate battery stability over time. To overcome the shuttle effect of soluble RMs and enhance battery stability, Li et al. developed a solid-state redox mediator anchored onto the cathode via a Cu(II) coordination compound with benzene-1,3,5-tricarboxylic acid (Fig. 3a).²⁴ This design enables efficient CO₂ capture and promotes Li₂C₂O₄ formation through a dimerization mechanism (Fig. 3b). Because of the solid nature of the RMs, no shuttle effect occurs in this system. The resulting Li–CO₂ battery exhibits low overpotential (0.9 V), and significantly improved cycling stability of over 400 cycles, advancing metal–CO₂ batteries toward practical application.

Fig. 3 Overview of oxalate-based discharge products in MCBs. (a) Schematic of a Li–CO₂ battery employing a solid RM(II)-BTC cathode. The solid-state nature of RM(II)-BTC suppresses the shuttle effect, while its high CO₂ adsorption capacity, fast redox kinetics, and reversibility enable efficient battery performance. (b) Proposed reaction pathway for oxalate (Li₂C₂O₄) formation mediated by RM(II)-BTC. Upon discharge, RM(II)-BTC is first reduced to Li₂RM(I)-BTC, which subsequently reacts with CO₂ to form Li₂RM(II)-BTC-(CO₂)₂⁻. This intermediate then decomposes to yield Li₂C₂O₄ and regenerate RM(II)-BTC. The dashed frame illustrates the transition of RM(II)-BTC into its reduced lithium-containing form. Atom colours: gray (C), red (O), cyan (Cu²⁺), green (Cu⁺), purple (Li⁺), blue (e⁻). (a) and (b) Are adapted with permission.²⁴ Copyright 2024, Springer Nature. (c) Electrochemical reaction mechanism of CO₂RR in a quasi-solid-state Na–CO₂ battery with Co-encapsulated N-doped carbon framework (Co-NCF) catalysts and gel electrolyte. (d) In situ Raman spectra of the quasi-solid-state Na–CO₂ batteries, revealing the generation of C₂O₄²⁻. (c) and (d) Are adapted with permission.²⁷ Copyright 2022, Cell Press. (e) Schematic of a Mg–CO₂ battery utilizing 1,3-propylene diamine (PDA) as a bifunctional additive, facilitating both CO₂ capture and conversion. (f) Raman spectra of the pristine and discharged Ni@Ru cathodes. (e) and (f) Are adapted with permission.²⁹ Copyright 2024, Wiley.

The stabilization of oxalate can also be achieved by electrolyte engineering. In a high-concentration LiTFSI/DMSO-based electrolyte, solid Li₂CO₃ and carbon products are formed. Wang et al. optimized the electrolyte composition to a low-concentration dimethyl sulfoxide (LC-DMSO) electrolyte (LiTFSI : DMSO = 1 : 12), resulting in switching the discharge products to oxalate.⁵⁷ Spectroscopic and electrochemical analyses demonstrated that the LC-DMSO electrolyte facilitates the formation of C₂O₄²⁻ intermediates through interactions with dissolved CO₂ molecules. Owing to the high donor number of DMSO (29.8), a stronger solvation sheath around Li⁺ ions is constructed, which effectively prevents direct association between Li⁺ and the soluble C₂O₄²⁻ intermediates. As a result, a reduced charge potential of 3.4 V and sustained cycling of over 260 hours are achieved. This study provides valuable insights for electrolyte design in regulating the discharge products in metal–CO₂ batteries.

Oxalate discharge products have also been reported in Na–CO₂ batteries, although relatively less frequently. This is likely due to the limited and still-developing research on Na–CO₂ batteries. Compared to conventional discharge products such as sodium carbonate and carbon, sodium oxalate is more readily decomposed, leading to a significant reduction in the charging voltage. Amin et al. employed a bimetallic layered double hydroxide catalyst on the cathode of Na–CO₂ batteries and observed the formation of sodium oxalate after discharge.²⁸ The cell demonstrated stable operation for up to 100 hours before degradation occurred. Furthermore, by adopting either a prolonged charging strategy or an asymmetric cycling protocol, the cell could be effectively revived from its degraded state. Xu et al. further advanced the development of Na–CO₂ batteries by constructing a reversible quasi-solid-state system capable of forming stable Na₂C₂O₄ discharge products.²⁷ Using a Co-encapsulated N-doped carbon framework catalyst and a gel electrolyte containing an imidazole-based organic cation, the system effectively stabilized the thermodynamically unstable Na₂C₂O₄ (Fig. 3c). The electron-agglomeration effect of Co nanoparticles enhanced CO₂ adsorption and promoted Na₂C₂O₄ formation, which is confirmed by the in situ Raman results (Fig. 3d). As a result, the battery delivered a high discharge capacity of 3094 mA h g⁻¹, maintained 1777 mA h g⁻¹ at 0.5 mA cm⁻², and achieved excellent cycling stability of over 366 cycles (2200 hours). These works highlight the critical value of designing Na–CO₂ battery systems capable of producing sodium oxalate, offering new avenues for improving battery reversibility and efficiency.

Oxalates have also been reported in Mg–CO₂ and Al–CO₂ batteries, though their formation requires specific conditions. For instance, Peng et al. designed a highly reversible and high-rate Mg–CO₂ battery using a liquid 1,3-propylene diamine (PDA) additive in conventional electrolytes(Fig. 3e and f).²⁹ At the cathode, PDA enhances Mg²⁺ de-solvation, CO₂RR kinetics, and the formation of the decomposable discharge product MgC₂O₄. At the anode, it promotes the formation of a Mg²⁺-conductive solid electrolyte interphase (SEI), enabling highly reversible Mg plating/stripping. This approach results in a Mg–CO₂ battery with over 400 hours lifespan at 200 mA g⁻¹ and excellent rate performance from 100 to 2000 mA g⁻¹. Fetrow et al. developed a rechargeable Al–CO₂ battery that produces oxalate discharge products by using aluminium iodide as a redox mediator.³⁰ This innovation enables the battery to achieve an ultralow overpotential of 0.05 V. Additionally, the Al–CO₂ cell maintains a high discharge voltage of 1.12 V and achieves a capacity of 9394 mA h g⁻¹. The aluminium oxalate, confirmed by the nuclear magnetic resonance (NMR) analysis, contributes to the improvement of battery reversibility. These findings highlight the potential of oxalate-based discharge products in improving the performance and reversibility of Mg–CO₂ and Al–CO₂ batteries.

MCBs with carbonate and CO discharge products

In MCBs, only Li–CO₂ batteries have been reported to generate carbonate and CO as discharge products. Xie et al. first reported the detection of CO gas in Li–CO₂ batteries by gas chromatography (GC), accompanied by the identification of carbonate species via Raman spectroscopy.³¹ Subsequently, Iputera et al. employed a more precise soft X-ray absorption spectroscopy (SXAS) technique to detect trace carbon signals, confirming the exclusive formation of Li₂CO₃ without evidence of other carbon-based solid byproducts.⁵⁸ The evolution of CO gas was further validated by a GC-MS signal of m/z = 28. Based on these findings, a mechanistic pathway was proposed: CO₂ initially undergoes a one-electron reduction to form CO₂²⁻ intermediates (eqn (12)), which subsequently react with additional CO₂ molecules to generate C₂O₄²⁻ species (eqn (13)). The resulting C₂O₄²⁻ species then react with lithium ions and electrons to produce Li₂CO₃ and release CO gas (eqn (14)). These studies establish a clear mechanistic understanding of the selective formation of carbonate and CO in Li–CO₂ batteries.

CO₂ + e⁻ → CO₂⁻ (12)

CO₂⁻ + CO₂ → C₂O₄²⁻ (13)

2Li⁺ + 2CO₂ + 2e⁻ → Li₂CO₃ + CO(g) (14)

MCBs with value-added carbon-containing compound products

In aqueous metal–CO₂ batteries, the discharge reactions typically involve the electrochemical reduction of CO₂ into value-added chemicals or fuels, such as CO, formate, and CH₄. These processes closely resemble those reported in the broader field of electrochemical CO₂ reduction. However, to function effectively as a rechargeable battery, it is essential that the discharge reactions be coupled with suitable oxidation reactions during the charging process to meet energy storage requirements. Among various systems, the Zn–CO₂ battery stands out as the most widely studied aqueous metal–CO₂ battery capable of producing value-added carbon-containing compounds.

Based on the nature of the charge–discharge reactions on the cathode, Zn–CO₂ batteries can be categorized into reversible and non-reversible types. In reversible Zn–CO₂ batteries, the discharge reaction involves the reduction of CO₂ into value-added chemicals, with formic acid (HCOOH) being one of the most successfully demonstrated products to date. During discharge, HCOOH is generated at the cathode and dissolves into the aqueous electrolyte, following eqn (15). During the subsequent charging process, HCOOH undergoes oxidation back to CO₂ (the reverse reaction of eqn (15)), thereby completing a reversible cycle. Wang et al. developed a Zn–CO₂ battery using a dual-anion regulated Bi electrocatalyst (Fig. 3a).⁵⁹ The BiOF catalysts achieved highest FE_HCOO- value at −1.8 V is 97% (Fig. 3b) and exhibited stability against high and low current densities (Fig. 3c). Moreover, Xie et al. developed a reversible aqueous Zn–CO₂ battery using a bifunctional Pd cathode with a 3D porous nanosheet structure.³³ The system achieved 90% FE for CO₂ reduction to HCOOH and stable battery performance, delivering high energy efficiency (81.2%), a low charge voltage (1 V), and excellent cycling stability over 100 cycles, demonstrating the feasibility and outstanding of this Zn–CO₂ system.

CO₂ + 2H⁺ + 2e⁻ → HCOOH (15)

In non-reversible Zn–CO₂ batteries, the discharge process still involves the CO₂ reduction to valuable chemicals or fuels, but these chemicals and fuels can be separated timely to avoid being consumed during charging process. In this circumstance, the water molecules can be oxidized to form oxygen gas at the cathode surface (eqn (16)).³⁵ This reaction pathway design removes the need for re-oxidizing discharge products, allowing the generation and separation of value-added products, thereby positioning the Zn–CO₂ cell as both an energy storage device and a chemical synthesis platform. Miao et al. developed bifunctional electrocatalysts comprising Ni clusters (Ni_x) coupled with single Ni sites (Ni–N₄/Ni_x), for CO₂ reduction and oxygen evolution in aqueous rechargeable Zn–CO₂ batteries (Fig. 4d).³⁴ Theoretical calculations (Fig. 4e) and in situ FTIR results (Fig. 4f) reveal that the *COOH intermediate of the electrochemical CO₂ reduction exhibits a synergistic binding with both Ni_x clusters and Ni–N₄–C single-atom sites. This dual–site interaction facilitates CO₂ activation and effectively lowers the energy barrier of the potential-determining step. This catalyst achieved nearly 100% FE for CO production from −0.4 to −0.8 V vs. RHE, outperforming Ni–N₄–C (55.0%). In addition, this bifunctional catalyst exhibited OER activities superior to commercial RuO₂, while exceeding that of Ni–N₄–C. This Zn–CO₂ battery delivered a peak power density of 11.7 mW cm⁻² and stable cycling over 1200 cycles (420 h), highlighting the importance of rational bifunctional catalyst design for enhancing non-reversible Zn–CO₂ battery performance.

H₂O → 1/2O₂ + 2H⁺ + 2e⁻ (16)

Fig. 4 Summary of the MCBs producing value-added carbon-containing compounds. (a) Schematic of a rechargeable aqueous Zn–CO₂ batteries with a BiOF-based cathode. (b) Faradaic efficiencies for formate production on various catalysts, including BiOF, BiF₃, and Bi₂O₃ catalysts. (c) Galvanostatic charge–discharge profiles at various current densities. (a)–(c) Are adapted with permission.⁵⁹ Copyright 2024, Wiley. (d) Schematic of an alkaline Zn–CO₂ battery employing a Ni–N₄/Ni_x cathode and 1 M KOH catholyte. (e) Calculated free energy diagrams for electrochemical CO₂ reduction to CO reaction on Ni–N₄, Ni–N₄/Ni₅, Ni–N₄/Ni₈, and N₄/Ni₅ catalyst at U = 0 V vs. RHE. (f) In situ FTIR spectra of Ni–N₄/Ni₅ catalysts collected across a potential range from open circuit potential (OCP) to −0.8 V vs. RHE. (d)–(f) Are adapted with permission.³⁴ Copyright 2023, Wiley.

At the anode, the anodic reactions during both discharge and charge are governed by the Zn/Zn²⁺ redox couple, with the specific processes influenced by the electrolyte pH. In near-neutral electrolytes, Zn is oxidized to Zn²⁺ (eqn (17)), where Zn²⁺ acts as the charge carrier. In alkaline electrolytes, Zn is first oxidized to Zn(OH)₄²⁻ (eqn (18)), followed by the decomposition of Zn(OH)₄²⁻ into ZnO, OH⁻, and H₂O. These anodic processes are critical for maintaining charge balance and enabling reversible energy storage.

Zn → Zn²⁺ + 2e⁻, E⁰ = −0.76 V vs. SHE (17)

Zn + 4OH⁻ → Zn(OH)₄²⁻ + 2e⁻, E⁰ = −1.22 V vs. SHE (18)

Potential applications

MCBs are a class of emerging energy storage systems that directly utilize carbon dioxide as a reactive component, offering a unique opportunity to couple energy storage with carbon management. Their application potential is largely determined by the behaviours of CO₂ conversion and the discharge products.

In systems that generate solid-state discharge products, such as Li–CO₂ batteries, the formation of carbonate and carbon results in exceptionally high theoretical energy density (1876 W h kg⁻¹). These systems typically exhibit notable discharge overpotentials, but also possess the potential for reversible cycling, especially when supported by advanced catalysts and tailored electrolytes. The combination of high energy output and rechargeability positions these batteries as strong candidates for electric vehicles (EVs), stationary grid storage, military power supplies, and long-duration applications such as space exploration or underwater systems. Crucially, the ability to store energy while consuming CO₂ in these systems highlights their dual value: CO₂ utilization and enabling high-performance energy solutions.

In contrast, MCBs that produce liquid-phase or gas-phase discharge products, such as Zn–CO₂ batteries, often follow irreversible reaction pathways, yet offer considerable advantages in terms of safety in aqueous electrolyte and effective CO₂ conversion. Their operational characteristics make them especially well-suited for low-power and intermittent-use applications, including wearable electronics, remote communication devices, environmental monitoring sensors, and portable control systems. Importantly, the CO₂ consumed in these batteries is not only utilized but also transformed into valuable chemical or fuel products, contributing directly to the development of technologies that reduce more CO₂ than they produce. By designing systems that allow the efficient separation of liquid products from electrolytes or the capture of gaseous products from flowing streams, these batteries can function as both energy reservoirs and chemical production units. This dual functionality significantly enhances the economic value proposition, enabling the recovery of value-added fuels or industrial feedstocks directly from waste CO₂.

In summary, MCBs not only offer a versatile platform for energy storage but also present a transformative pathway for harnessing CO₂ as a resource, aligning with global efforts toward net-zero carbon technologies and sustainable energy systems.

Challenges and outlooks

Despite their promising potential in coupling energy storage with carbon dioxide utilization, MCBs still face several critical challenges that hinder their practical deployment. Addressing these issues is essential for unlocking the full potential of MCBs in next-generation energy and carbon management systems.

Limited electrocatalytic activity and the need for advanced bifunctional catalysts. A major limitation of current MCBs lies in the low catalytic activity of cathode for both CO₂ reduction and the reverse oxidation of discharge products, leading to sluggish kinetics and large overpotentials. To overcome this, there is an urgent need to develop highly active and durable bifunctional catalysts that can efficiently facilitate both the discharge and charge processes. Recent advances in artificial intelligence (AI), particularly machine learning (ML), offer a powerful approach to accelerate catalyst discovery. By integrating computational screening, data-driven modeling, and experimental validation, machine learning can help identify optimal catalyst compositions, structures, and surface properties, greatly enhancing the design efficiency and performance predictability of next-generation electrocatalysts.

Interface issues associated with solid-state discharge products. In systems like Li–CO₂ batteries, the discharge products (e.g., Li₂CO₃) are typically solid and electronically insulating. Their deposition on the cathode surface leads to poor solid–solid interfacial contact, severely impeding reaction kinetics and reversibility. To address this, effective mobile species, such as redox mediators should be developed. These mediators can transfer electrons between the electrode and the discharge product, facilitating decomposition at lower overpotentials. Moreover, they can help alleviate interfacial resistance and improve charge transport in systems dominated by solid-phase reactions, offering a promising route toward more reversible and efficient MCBs.

Low productivity of chemical/fuel-producing MCBs and product separation challenges. In MCBs that can convert CO₂ into valuable chemicals or fuels (such as formate, methane, or syngas), the inherently low current densities limit the total amount of products generated. This increases the relative cost of product separation and collection, particularly in flow systems. To address this bottleneck, it is critical to develop flow-type cells with optimized design parameters, including current density, electrolyte flow rate, and gas flow rate, to ensure high conversion efficiency and selective separation of products without affecting the stable energy storage role.

CO₂ capture-based energy storage

Reducing CO₂ emissions and ensuring clean energy production/supply through conventional CO₂ capture and advanced energy storage, respectively, hold promise for a sustainable future. With growing research interest in this sector, integrated CO₂ capture-based energy storage stands at the intersection of environmental sustainability and energy innovation, which can accomplish both the reduced CO₂ footprint and clean energy demands.⁶⁰ This coupling would enable the simultaneous CO₂ management and storage of electrochemical energy, that not only allows carbon sequestration but also offers a way to store and dispatch energy on demand. It can also be combined with renewable energy sources, allowing flexible CO₂ capture and energy storage while compensating the intermittency of renewable energy.⁶¹ In addition, the isolated CO₂ capture is energy intensive. Integrating it with energy storage and renewable energy systems offers several advantages, including reduced capital costs through shared infrastructure and the provision of grid-scale storage. This approach could significantly enhance the economic viability compared to conventional technologies. Table 2 provides the recent progress in this technology for CO₂ capture coupled energy storage.

Table 2 Performance summary of CO₂ capture-based energy storage systems

Energy storage system	Electrodes	Mechanism	CO2 adsorption capacity	Energy consumption	Performance			Ref.
					No. of cycles	Current/voltage	Coulombic efficiency
Battery	Cathode: CO2-binding polyanthraquinone-carbon nanotube (Q-CNT) coated substrate	Electro-swing for CO2 capture/release during battery charging/discharging	7.4 μmol of CO2 (at 1.3 V)	43 kJ mol^−1 (at 60% bed utilization)	7000 (<30% capacity loss)	−1.8 V	Faradaic efficiency ∼90%	62
	Anode: ferrocene-CNT (Fc-CNT) coated substrate sandwiched between two cathodes
Supercapacitor	YP80F, a biowaste-derived activated carbon electrode	pH swing for CO2 capture/release	170 mmol CO2 per kg of electrode (30 mA g^−1, pure CO2)	18 kJ molCO2^−1 (300 mA g^−1)	12 000 (>2500 h, pure CO2)	150 mA g^−1	> 99.8% (under 15% O2)	63
		Positive charging mode for supercapacitor operation
Aqueous flow-battery	Negolyte: sodium 2,2′-(phenazine-1,8-diyl)bis(ethane-1-sulfonate) (1,8-ESP)	pH swing for CO2 capture/release	1.41 mol L^−1 (10 mA cm^−2)	55 kJ molCO2^−1 (10 mA cm^−2)	220 cycles for CO2 capture/release (18 days, no capacity decay)	20 mA cm^−2	95–82% (under 3–20% O2)	64
		Battery charging: CO2 capture and electricity storage
	Posolyte: ferricyanide/ferrocyanide redox couple (Fe(CN)6^3−/Fe(CN)6^4−)	Battery discharging: CO2 release and electricity delivery
Vanadium redox-flow battery	Posolyte: vanadium redox couple (VO^2+/VO2^+)	pH swing for CO2 capture/release	—	54 kJ mol CO^2−1	10 day–night cycles for CO2 capture/release	0.5 V (daytime) and −0.5 V (night time)	—	61
	Negolyte: ferricyanide/ferrocyanide redox couple (Fe(CN)6^3−/Fe(CN)6^4−)	CO2 release: oxidation of VO^2+ to VO2^+ in the posolyte, reduction of Fe(CN)6^3− to Fe(CN)6^4− in the negolyte and electricity storage (daytime)
	CO2 sorbent: potassium carbonate (K2CO3) in a separate absorption column	CO2 capture: reduction of VO2^+ back to VO^2+ in the posolyte, oxidation of Fe(CN)6^4− to Fe(CN)6^3− in the negolyte and electricity delivery (nighttime)

---

Working mechanism

A primary requirement of this integrated system to work lies in identifying CO₂ capture molecules with redox properties to allow energy storage and release in addition to CO₂ absorption and desorption. By integrating these CO₂-binding mechanisms with redox-active materials, the capture-release cycle can be directly coupled with electron transfer processes required for energy storage. Most common examples include amines, carbonates or redox-active sorbents, that not only reversibly bind with small molecules (e.g., CO₂) but can also participate in electron transfer.^62,63,65,66 This dual functionality assists in electrochemically mediated CO₂ capture, where redox cycling drives the uptake and release of CO₂ while simultaneously storing or delivering electrical energy. Integrated electrochemical CO₂ capture and energy storage systems generally include pH swing- and electro swing-based systems.

pH swing-based systems. Electrochemically driven pH swing systems involve reversible proton generation and consumption to modulate electrolyte pH, thereby controlling CO₂ solubility and formation of dissolved species. CO₂ can be converted to bicarbonate (HCO₃⁻) or carbonate (CO₃²⁻) ions under alkaline conditions (pH > 8) to accelerate the capture process, while CO₂ can be released upon acidification.⁶⁷ This mechanism is mediated by redox-active molecules through a proton-coupled electron transfer (PCET), aiding the reversible H⁺ insertion/removal depending on the applied potential.⁶⁸ As the pH swings, the energy required to change the pH can be stored as electrical energy in redox-active materials, such as, quinones, phenazines, etc.^64,69 This system provides several advantages such as high CO₂ selectivity and binding affinity, compatibility with safe aqueous electrolytes, lower environmental impact and efficient electrochemical reversibility for reversible CO₂ capture and release.

The pH-swing CO₂ capture can be integrated with existing energy storage technologies to improve both the carbon capture and grid storage (Fig. 5a). For example, the process of charging and discharging the redox-flow battery can align with the pH swing cycle i.e. storing energy during the CO₂ release phase and releasing energy during the CO₂ capture phase.⁶⁹ Also, supercapacitors, which store energy via electrostatic charge, could be used in combination with pH-swing systems to serve this purpose.⁶³ In this configuration, a supercapacitor could be used to control the electrochemical reactions driving the pH swing, while simultaneously providing fast energy storage or supply to the grid.

Fig. 5 (a) CO₂-capture coupled energy storage in batteries and supercapacitors based on pH-swing mechanism. (b) Electro-swing-based CO₂ capture integrated energy storage following electrochemical sorbent regeneration mechanism. Application of CO₂ capture-based energy storage system in (c) energy-intensive CO₂ emitting industries, (d) HVAC equipped smart buildings and (e) renewable energy storage for grid applications.

Electro swing-based systems. Electro-swing-based CO₂ capture and energy storage involves CO₂ capture from gaseous streams and energy storage through reversible electrochemical reactions.⁶² An electrochemical sorbent regeneration mechanism is followed to couple CO₂ capture/release with energy storage using electrochemically active sorbents that can undergo reversible redox reactions to absorb/desorb CO₂.^70,71 During capture process, CO₂ is adsorbed by the sorbent material at the electrode surface through an electrochemical reduction reaction. While during release process, CO₂ is desorbed from the sorbent on application of an oxidation potential (Fig. 5b). When CO₂ is released, the electrochemical energy stored in the sorbent during capture can be recovered or used for energy storage applications.⁶² The choice of sorbent materials will highly influence the efficiency, capacity, and operational viability of an electro-swing-based CO₂ capture coupled energy storage system. Organic materials like quinones can capture CO₂ during their reduction phase by undergoing reversible electron transfer.⁷² Upon oxidation, CO₂ is released, and the original redox-active species is regenerated, serving dual roles as charge carriers and CO₂ sorbents. Conductive polymers⁷³ and metal oxides,⁷⁴ which can undergo redox reactions and interact with CO₂, can also serve as sorbents.^75–77 The CO₂ binding strength and reversibility can be tuned through modification strategies, including functional group substitution, backbone rigidity, and solvent interactions for enhanced performance.^78–81

The electrochemical cell used for CO₂ capture can serve as an energy buffer, storing renewable energy during off-peak hours and releasing energy during peak demand.⁶⁶ This buffering function enables the decoupling of CO₂ capture from the grid and allows the system to operate in a more flexible and efficient manner. For example, integrating the electro-swing CO₂ capture system with redox flow batteries presents an innovative approach. In this system, the electrochemical sorbents could be used as the electrolyte in the flow battery, allowing the system to function as both a CO₂ capture device and an energy storage system. This would allow for continuous operation where energy is stored and released in a closed loop. Supercapacitors can also be used in combination with the electrochemical sorbent regeneration system.

Potential applications

Integrating CO₂ capture with energy storage presents a promising pathway to simultaneously address climate and energy crises. The potential applications which can be realized using this coupled system are discussed below. Fossil-fuels based energy-intensive industries like product manufacturing, chemical processing, cement production, fuel generation, etc. are major contributors towards global CO₂ emissions. CO₂-capture coupled energy storage presents a viable solution to capture excess CO₂ emissions from these industries (Fig. 5c), where the stored energy can further be utilized (a) to release and supply high-purity CO₂ for different applications such as oil recovery, carbonated drinks, chemical synthesis, etc.⁸² or (b) to share the power requirements in industrial processes, thereby reducing the dependence on fossil-fuels. Further utilization of renewable energy with such integrated system would provide additional environmental benefits.⁸³ This integrated technology can also be implemented into smart buildings to cut down the CO₂ footprint and at the same time improving energy efficiency. In modern buildings equipped with HVAC (heating, ventilation and air conditioning) systems, the management of both energy consumption and indoor air quality has become critical. Despite providing comfortable indoor environments, HVAC systems are energy-intensive and significantly contribute to CO₂ emissions. Integrating HVAC system with CO₂ capture coupled energy storage technology can potentially address these challenges.^84,85 This integrated system can absorb CO₂ during air circulation and improve air quality along with renewable energy storage, whilst release the stored energy in reverse cycle to power the HVAC system (Fig. 5d). The utilization of stored renewable energy to power HVAC would considerably reduce energy consumption and thus the operational costs in smart buildings. Besides, the CO₂-capture unit would improve the building's air quality, thus providing health benefits. Although this system has not been investigated yet but holds immense potential to be explored in future. Integration of CO₂ capture/release and energy storage also shows promise to store surplus renewable electricity while simultaneously capturing and managing CO₂ emissions.^61,64 Electrochemical CO₂ capture, which allows the release of high-purity CO₂ from different emission streams with lower energy losses than traditional thermal and pressure-induced processes, can play a crucial role in decarbonization of energy technologies. Decarbonizing energy sector depends not only on the deployment of intermittent renewable energy but also on the parallel development of grid-scale energy storage. A reversible, redox-mediated CO₂ capture/release cycle that is electrically driven, lies at the core of this integrated system. During peak production hours of renewable energy (e.g., midday solar), clean electricity can be used to (a) drive redox reactions to enable CO₂ capture from flue gas or ambient air and (b) store electrical energy in the form of chemical potential. In contrast, when the grid demands energy (when renewable energy is unavailable), reverse operation is realized, where pure CO₂ is released electrochemically from the sorbent and the stored electricity is delivered back into the grid (Fig. 5e). Notably, during periods of low electricity demand or excess renewable generation, surplus energy can be utilized to manage the CO₂ capture process.

The CO₂-capture integrated energy storage systems have already started to progress in grid-scale storage applications. For instance, pH swing-based CO₂ capture integrated energy storage is reported using an aqueous flow cell system.⁶⁴ A redox-active phenazine derivative (negolyte) on the negative side of cell (sodium 2,2′-(phenazine-1,8-diyl)bis(ethane-1-sulfonate) (1,8-ESP)) allowed PCET for pH swing-based CO₂ capture. When paired with ferrocyanide (posolyte) on the positive side, this cell enabled energy storage function. The negolyte pH was increased during charging, resulting in reaction between CO₂ and OH⁻ ions, enabling CO₂ capture (1.4 mol_CO2 L⁻¹ at 10 mA cm⁻²) and energy storage. This process is reversed during discharging due to the decreased pH, aiding CO₂ release and energy delivery. In simple terms, deacidification during charging results in CO₂ capture and energy storage, while acidification during discharging releases pure CO₂ and stored energy (Fig. 6a and b). This system could flexibly work as a pure energy storage device (open-circuit voltage: 1.1 V; coulombic efficiency: 99.9%; round-trip energy efficiency: 78.5%; capacity retention: 95% for 170 days (fade rate ∼0.05% per day)) by isolating the electrolyte from CO₂, depending on the market requirements.

Fig. 6 (a) CO₂ capture-energy storage and (b) CO₂ release-energy delivery in phenazine-based electrochemical cell during charging and discharging process respectively. Reproduced with permission.⁶⁴ Copyright 2023, Springer Nature. Schematic representation of vanadium-based redox flow battery for integrated CO₂-capture and renewable energy storage showing (c) energy storage and CO₂ release during daytime and (d) energy supply to grid and CO₂ capture during night. Reproduced with permission.⁶¹ Copyright 2025, American Chemical Society. (e) Challenges for current performance and projected applications in CO₂-capture integrated energy storage technology.

However, improving the stability of organic redox molecules in presence of O₂ (present in flue gas ∼3–5%) or (air ∼20%) remains crucial, since O₂ can oxidise these molecules during prolonged cycling. In this case, the chemical oxidation of negolyte (1,8-ESP) due to O₂ was reversed by applying reduction potential. But it caused the oxidation of posolyte (ferrocyanide), requiring electrochemical rebalancing by applying reverse potentials.⁸⁶

Lately, vanadium redox flow batteries have also been employed for CO₂ capture-based energy storage (Fig. 6c and d).⁶¹ It establishes (a) absorbent and pH-based CO₂ capture/release, (b) battery charging during daylight hours (when renewable electricity is available) with CO₂ release and (c) CO₂ capture and electricity delivery to the grid during night (when solar energy is inaccessible). The membrane-separated electrochemical cell contained vanadium redox (VO²⁺/VO₂⁺) posolyte and ferricyanide/ferrocyanide redox (Fe(CN)₆³⁻/Fe(CN)₆⁴⁻) negolyte. An absorption column containing potassium carbonate (K₂CO₃) solution can capture CO₂ in the form of dissolved HCO₃⁻ or CO₃²⁻ ions, which is fed to negolyte. During daytime i.e. CO₂ release and battery charging, VO²⁺ and H₂O undergo oxidation to generate VO₂⁺, H⁺ and e⁻ which migrate towards negolyte (reduction), decreasing its pH and thereby releasing absorbed CO₂. Upon reversing the potential, VO₂⁺ is reduced to VO²⁺ in posolyte due to negolyte oxidation, while simultaneous increase in pH results in CO₂ capture and delivery of the stored energy to the grid during night. It is noteworthy to mention that the vanadium redox couple in this system is oxygen insensitive while the decoupling of negolyte and CO₂ absorption column suppressed the adverse effects of O₂ impurities.

Challenges and strategies

Integrated CO₂-capture and energy storage system can substantially reduce CO₂ footprint while enhancing energy storage capabilities. However, to unlock their full potential, a number of shortcomings need to be overcome (Fig. 6e). To begin with the performance bottlenecks, the redox-active materials must withstand extended capture-release cycling process. For pH swing-based systems, CO₂ capture systems must be optimized to reduce the energy required for the desorption process by employing robust redox molecules.⁸⁷ Further, suppressing the side reactions with oxygen under high oxidative potentials is essential for the success of pH-swing based CO₂ capture coupled energy storage.⁸⁷ This necessitates the development of less O₂-sensitive organic redox molecules to improve their performance and durability. Electro-swing systems also demonstrate scalability potential, encouraging their use in both small-scale and large-scale applications (e.g., renewable energy-based grid-storage). Their compatibility with organic solvents increases the ability to achieve high energy density and fast kinetics.⁶² Though, improving the long-term stability and CO₂ absorption capacity of materials along with the lifespan and energy efficiency of electrochemical device, is the need of the hour. Most of the sorbent materials are still in the early stages of development and there is a huge room for improvement. In view of their real-time application, rigorous research towards minimized degradation and enhanced cycling is a prerequisite. Understanding the complex solid–gas–liquid interfaces and optimizing the stability of sorbents is critical through combined experimental and theoretical investigations.⁷⁸ Moreover, the energy requirements to regenerate the sorbent and release CO₂ are high, requiring optimization of the electrochemical processes to make the system more energy efficient. Integrating CO₂-capture with advanced energy storage technologies and hybrid systems hold potential to minimize energy losses and increase the overall system efficiency.

For real-time applications, such as grid-scale energy storage, the main challenge lies in the intermittent nature of renewables (e.g., solar, wind) which needs to be balanced for energy demand and supply.⁸⁸ Further, the CO₂ release process also require continuous energy supply.⁸⁹ Both challenges necessitate an efficient and long-cycling energy storage device which can provide grid stability. On the other hand, application in smart buildings equipped with HVAC systems require sophisticated control systems, large space, high efficiency and durability sustainability and economic feasibility of such systems. As far as the energy-intensive industries are concerned, practical concerns regarding infrastructure, space requirements and effective storage and transport of CO₂ needs to be considered. Overall, the integrated CO₂-capture and energy storage system represents a highly complex system with high projected capital costs, necessitating improved energy efficiency and CO₂ capture-release performance with minimal energy consumption. Continued research in materials development and performance engineering is the need of the hour.⁹⁰ The use of solid-state CO₂ capture materials with improved electrochemical stability over a wide range of applied potentials show promise while extending the research for other potential energy storage technologies (e.g., batteries, supercapacitors) can improve the commercial viability. Technical challenges including CO₂ management and possible safety issues also require advanced monitoring and safety protocols, but at a later stage.

Molten carbonate fuel cells

Molten carbonate fuel cells (MCFCs) represent a highly promising technology that not only enables efficient capture of CO₂ from combustion flue gases, but also simultaneously generates electricity and hydrogen. This dual functionality facilitates both energy conversion and greenhouse gas mitigation in an integrated process. Compared to other energy storage or generation systems, MCFCs can more effectively utilize CO₂ present in flue gases, thereby significantly reducing carbon emissions. Unlike conventional carbon capture methods such as amine scrubbing, MCFCs offer the additional advantage of producing power during CO₂ capture, resulting in substantially lower overall energy consumption. As such, MCFCs hold great potential for leveraging CO₂ as a valuable resource for clean energy production.

Mechanisms

In MCFCs, CO₂ is not a waste byproduct but an active participant in the electrochemical process. Through a unique mechanism involving carbonate ion transport, CO₂ from flue gas is absorbed at the cathode, transferred across the electrolyte, and ultimately released for caption at the anode during power generation.⁹¹

Anode chemistry and fuel reforming. In MCFCs, methane (CH₄) undergoes internal reforming at the anode, enabling the simultaneous production of hydrogen (H₂) and preparation of CO₂ for capture. This process involves two key steps: (1) steam reforming (eqn (19)), where CH₄ reacts with water to form carbon monoxide (CO) and H₂; (2) the water-gas shift reaction (eqn (20)), which converts CO and steam into CO₂ and additional H₂.

CH₄ + H₂O → CO + 3H₂ (19)

CO + H₂O → CO₂ + H₂ (20)

These reactions are endothermic, particularly in the steam reforming step (+206 kJ mol_CH4⁻¹), but the required heat is partially offset by the exothermic electrochemical oxidation of H₂. This thermal balance helps maintain the high operating temperature of the cell (600–650 °C) without external heating. The H₂ produced is partly used in the electrochemical reaction to generate electricity, while the CO₂ remains in the anode stream for downstream capture. This coupling of reforming and electrochemical processes allows MCFCs to generate power and provide enrich CO₂ simultaneously. By adjusting conditions such as fuel utilization and temperature, the anode acts as an efficient zone for both fuel conversion and CO₂ management, giving MCFCs a distinct advantage over conventional power or capture systems.

Electrochemical reactions. The electrochemical operation of MCFCs is driven by coupled redox reactions occurring at the anode and cathode, with carbonate ions (CO₃²⁻) serving as the charge carrier through the molten electrolyte. At the anode, hydrogen produced from fuel reforming undergoes oxidation, reacting with CO₃²⁻ to form water and CO₂ while releasing electrons (eqn (21)). At the cathode, carbon dioxide from the flue gas combines with oxygen and incoming electrons to regenerate carbonate ions (eqn (22)). Therefore, CO₂ is actively transported from the cathode to the anode in the form of CO₃²⁻ ions, enabling integrated CO₂ capture within the power generation process. The overall electrochemical reaction within the cell can be found in eqn (23).

H₂ + CO₃²⁻ → H₂O + CO₂ + 2e⁻ (21)

CO₂ + ½O₂ + 2e⁻ → CO₃²⁻ (22)

CO₂ (cathode) + H₂ (anode) + ½O₂ → H₂O + CO₂ (anode) + electricity + heat (23)

MCFCs capture CO₂ as an intrinsic part of the electricity generation process. For each methane molecule consumed in the anode, four CO₂ molecules are drawn into the system at the cathode, and five are released at the anode.⁹³ This results in a net transfer of CO₂ from the external flue gas into a concentrated stream suitable for sequestration or utilization. Different from traditional carbon capture methods that separate CO₂ post-combustion with a parasitic energy penalty, MCFCs utilize CO₂ as an electrochemical reactant, enabling both energy production and CO₂ concentration in a single step.

A typical molten carbonate fuel cell (MCFC) consists of a porous anode, a porous cathode, and a porous matrix that contains the molten carbonate electrolyte. The system relies on capillary pressure equilibrium among the three porous components. As illustrated in Fig. 7b, this equilibrium condition requires that the capillary pressures in the anode, cathode, and matrix be equal.⁹² When this balance is achieved, the distribution of the electrolyte becomes stable for a given total volume of melt. This distribution is governed by the total pore volume and pore size distribution of each component. Moreover, the most electrochemically active region within the anode and cathode of an MCFC is the three-phase boundary (TPB), where the gas, electrolyte, and solid phase electrodes coexist.⁹² For example, in porous nickel anodes in Fig. 7c, for instance, the TPB is located near the top of the electrolyte meniscus within the pores. The contact angle (θ) at this interface is a critical parameter, as it determines the extent to which reactant gases can penetrate into the pores and participate in electrochemical reactions. Ideally, the entire pore wall would be uniformly covered by a thin film of molten carbonate, allowing full utilization of the internal surface for reaction. Achieving this requires the contact angle θ to approach zero. Consequently, under practical conditions, only a portion of the internal surface area of the MCFC anode is electrochemically active. Realizing such optimal wetting conditions calls for careful interfacial engineering of the electrode materials to ensure a stable TPB and consistent electrochemical performance under high-temperature operating conditions.

Fig. 7 Summary of the MCFC systems. (a) Schematic of MCFC system, adapted with permission.⁹¹ Copyright 2020, Elsevier. (b) Schematic illustration of in-pore meniscus formation and wetting behavior, highlighting the capillary equilibrium condition characterized by the balance of interfacial tensions among the three porous components. (c) Details of the three phase boundary (TPB). (b) and (c) Are adapted with permission.⁹² Copyright 2020, Elsevier. Sankey-style diagrams illustrating the molar flow distributions for high (d) and low (e) fuel utilization conditions. The width of each arrow corresponds to the respective molar flow rate, with the exception of hydrogen (H₂), whose flow width has been scaled down by a factor of 3.3 to place it on a comparable chemical energy basis with methane (CH₄), enabling clearer visual comparison. The data in (d) and (e) are reproduced with permission.⁹³ Copyright 2022 Elsevier.

The energy consumption is an important component of the overall value proposition of the MCFC systems. Fig. 7d and e compared the fuel consumption and the corresponding CO₂ for sequestered and H₂ for recycling.⁹³ In Fig. 7d, the MCFC captures CO₂ through two sequential processes, with the co-produced H₂ recycled back into the fuel system. In the first stage, 100 mol of CH₄ is consumed, generating 100 mol of CO₂ for capture. The MCFC is assumed to capture 90% of this CO₂, although the capture rate can be adjusted depending on system design. To achieve this, an additional 32 mol of CH₄ is processed, all of which is fully captured.

By contrast, the configuration in Fig. 7e adopts a low fuel utilization strategy during the initial stage to maximize H₂ production (300 mol) while simultaneously capturing CO₂. After separation of H₂ and CO₂, the resulting hydrogen, characterized by minimal CO₂ emissions, is employed as a clean fuel in a secondary process. This approach offers a significant advantage in terms of energy efficiency. The integrated design not only enhances overall hydrogen output but also consolidates CO₂ capture in a single intensified unit, comprising the MCFC and H₂/CO₂ separation modules, thereby reducing system complexity, capital investment, and operational overhead. While the energy penalty may not be lower than that of the configuration in Fig. 7d, the ability to produce low-carbon H₂ and streamline process integration positions the design in Fig. 7e as a more energy- and cost-effective solution for sustainable fuel and CO₂ management.

Benefits. Integrated CO₂ capture capability and low CO₂ emission coefficients. MCFCs possess a unique capability to inherently capture CO₂ during normal electrochemical operation, thereby avoiding the need for separate energy-intensive post-combustion capture units. In MCFCs, CO₂ is used as a reactant at the cathode, and the cell facilitates its transfer from the cathode to the anode via carbonate ions in the electrolyte, enabling effective CO₂ separation and concentration. When MCFCs are integrated with traditional power plants or operated as stand-alone systems, the CO₂ concentration in the exhaust can be increased from typical flue gas levels (3–15%) to as high as 60–70% for further storage or utilization, depending on the system design and combustion strategy employed (e.g., pure oxygen combustion of anode exhaust).⁹⁴ Consequently, the CO₂ emission coefficient (CEC) of MCFC-based systems can be reduced to as low as 236 kg MW h⁻¹, significantly outperforming conventional systems such as coal-fired (736–1290 kg MW h⁻¹) or natural gas plants (344–713 kg MW h⁻¹).⁹⁵ By combining power generation with CO₂ capture, MCFCs offer a promising solution for developing cleaner, low-carbon energy systems.

Integration potential. MCFC systems are modular in nature and can be scaled to meet a range of power demands by integration within large-scale thermal power plants. Specifically, they can be retrofitted or co-located with existing infrastructure (Table 3), such as natural gas combined cycle (NGCC) plants, coal-fired power plants (CFPPs), and gas turbines (GTs).^96,97 By utilizing the existing plant exhaust gases as cathode feedstock, these hybrid configurations enable MCFCs to serve both as energy producers and CO₂ separators. The modular MCFC units can process these streams and recover CO₂ without significantly impacting the host plant's overall efficiency. For example, when combined with NGCC systems, MCFCs have been shown to achieve CO₂ capture efficiencies to 80% while maintaining or even improving net electric efficiency.⁹⁸ Moreover, the high-temperature exhaust from MCFCs (typically ∼340–650 °C) can be used for combined heat and power (CHP) applications, such as steam generation or district heating, enhancing overall energy utilization and reducing waste.⁹⁹ This flexibility in integration and application makes MCFCs attractive for decarbonizing of power infrastructure.

Table 3 Performance summary of molten carbon fuel cells

Systems	Capacity (MW)	Efficiency	CO2 capture ratio	CO2 emission coefficient (kg of CO2 kW h^−1)	Ref.
Coal-fired power plant (CFPP)-MCFC	CFPP (450)-MCFC (198)	45.77%	76.9%	0.148	102
CFPP-MCFC	—	40%	39%	0.253	103
Direct internal reforming (DIR)-MCFC-steam turbine (ST)	DIR-MCFC (0.5)-MGT (0.1)-ST	0.52%	—	0.101	104
Reciprocating engine (RE)-MCFC	Reciprocating engine (6)-MCFC (1.8)	41%	76%	0.121	105
Integrated gasification (IG)-MCFC	IG-MCFC (0.2)	44.7%	—	0.751	106
Natural gas combined cycle (NGCC)-MCFC (89)	NGCC (GT: 271, ST: 191)-MCFC (89)	58%	80%	0.071	98
NGCC-MCFC	NGCC (800)-MCFC (212)	52.5%	58.1%	0.149	107
MCFC-steam gas turbine (SGT)	MCFC (2.4)-SGT (0.33)	48.2%	583 t CO2/y	0.751	108

---

High energy efficiency. Unlike low-temperature fuel cells that rely solely on external fuel processing, MCFCs can internally convert methane or other hydrocarbons into hydrogen and carbon monoxide via reforming reactions at the anode, enabling more efficient fuel use. Stand-alone MCFC systems typically achieve electrical efficiencies of around 47%.⁹⁵ When integrated with effective waste heat recovery (e.g., in CHP or MCFC-GT hybrid systems), overall system efficiencies can reach or even exceed 75%.¹⁰⁰ This performance is comparable to that of the most advanced natural GTs and surpasses the efficiency of most conventional fossil fuel power systems. Furthermore, MCFC technology can utilize a wide range of fuels, such as natural gas, biogas, and coal-derived syngas, offering enhanced adaptability and flexibility across various application scenarios.¹⁰¹ These efficiency advantages not only improve energy output but also reduce fuel consumption and CO₂ emissions per unit of electricity generated, aligning with both economic and environmental goals in the transition toward cleaner energy systems.

Progress

Over the past decades, MCFC technology has undergone significant advancement.^103,106 When integrated with NGCCs, CFPPs, or GTs, MCFC systems can achieve significantly enhanced installed capacity (Table 3).

In recent years, the development of MCFCs using green ammonia as fuel has attracted attention due to its potential to reduce fossil fuel dependence. Lu et al. built a single-cell MCFC system to study the electrochemical performance of direct ammonia-fed MCFCs, focusing on key factors such as the NH₃/H₂ volume flow ratio and the operation temperature.¹⁰⁹ Operating at 680 °C, the cell attained its best performance, with a current density of 0.059 A cm⁻² and a high ammonia conversion rate of 91.6%. When using NH₃/H₂ mixed gas, the best results (0.057 A cm⁻², 84.9% conversion) were achieved at a volume ratio of 9, indicating that higher H₂ content can lower ammonia conversion. SEM analysis confirmed that ammonia does not damage the porous structure of the electrodes, supporting its direct use as a reliable fuel in MCFCs. Another promising fuel for MCFCs is the deployment of bio-based fuels. Dybiński et al. developed a molten carbonate fuel cell (MCFC) directly fuelled by raw bioethanol derived from fruit fermentation without pre-treatment.¹¹⁰ However, solid particle contamination in the fermentation product prevents stable long-term operation of the fuel cell, making purification of the bioethanol necessary before use.

MCFCs can also be integrated with other technologies such as biomass gasification. Tawakoli et al. proposed using sewage sludge as a biomass feedstock to address waste management and reduce CO₂ emissions, while supporting sustainable energy production.¹¹¹ They simulated a biomass gasification system combined with a MCFC in Aspen Plus software, using air and steam as gasifying agents. The study optimized key parameters including the equivalence ratio, steam to biomass ratio, gasification temperature, fuel cell temperature, inlet air flow rate, and system pressure to minimize the levelized cost of electricity and maximize power output and efficiency. The genetic algorithm-based optimization showed that the system performed best at an equivalence ratio of 0.102, steam to biomass ratio of 0.24, gasification temperature of 894 °C, fuel cell temperature of 776 °C, air flow rate of 24.8 kmol h⁻¹, and system pressure of 1.26 bar. Under these parameters, the system delivered a levelized cost of electricity of 0.42 USD kW h⁻¹, a power output of 0.224 W cm⁻², and an overall efficiency of 60.1%.

Moreover, a successful example of the potential application of MCFCs is their integration into a hybrid renewable energy system for a four-story building in Bandar Dayyer, southern Iran.¹¹² In this system, biogas produced from sewage digestion was used to fuel a MCFC, which operated at 650 °C and generated 98 kW of electricity. The system also included water desalination, producing over 1100 kg of fresh water per day, meeting both domestic and process demands. With overall energy and exergy efficiencies of 54% and 52.58% respectively, and relatively low CO₂ emissions, this application highlights the capability of MCFCs to provide clean electricity, water, heating, and cooling in a sustainable and environmentally friendly manner.

Challenges and strategies

Despite their advantages in efficiency and fuel flexibility, several technical and economic challenges hinder their widespread application. Addressing these challenges with targeted strategies is essential to advance the practical deployment of MCFC technology.

The first challenge in MCFCs is the material degradation. MCFCs operate at elevated temperatures (typically 600–700 °C), which accelerates the degradation of cell components, including electrodes, current collectors, and the electrolyte/electrode interface. Corrosion, thermal stress, and chemical instability significantly limit the system's lifespan and reliability. To address this, the development of advanced materials is essential. Research focuses on corrosion-resistant alloys, thermally stable ceramic matrices, and robust electrode coatings that can withstand harsh conditions. Implementing protective layers and optimizing material selection can significantly enhance the durability and operational life of MCFCs.

When using bio-derived or unprocessed fuels such as biogas or raw bioethanol, impurities like sulfur compounds, tars, or particulates can poison the anode catalyst or clog system components. This leads to rapid performance degradation and operational instability. Incorporating pre-treatment units to remove harmful contaminants and designing reformers that convert complex fuels into clean hydrogen-rich gas ensure stable cell performance. In cases where raw fuels are used, minimal purification steps must be introduced to maintain system feasibility without excessive cost.

Maintaining the integrity of the molten carbonate electrolyte is critical. Issues such as electrolyte leakage, evaporation, or compositional changes can disrupt ionic conductivity and reduce overall efficiency. Stable electrolyte matrices and improved sealing techniques can mitigate these problems. Using lithium/sodium/potassium carbonate mixtures with optimized ratios enhances thermal stability. Additionally, advanced cell designs that confine the electrolyte and reduce evaporation can improve electrolyte retention and system performance.

In summary, while MCFCs offer compelling advantages in terms of efficiency, fuel flexibility, and integrated CO₂ capture, their large-scale deployment is still constrained by several interrelated technical and economic challenges. A multidisciplinary approach that combines materials science, electrochemical engineering, and system integration will be essential to unlock the full potential of MCFCs as a cornerstone of future low-carbon energy infrastructure.

CO₂ based electricity generator

In recent years, researchers have developed innovative strategies to directly harvest electricity from CO₂, transforming this greenhouse gas from an environmental burden into a valuable energy source. Two representative systems demonstrate fundamentally different but highly promising mechanisms for CO₂-driven electricity generation (Table 4).

Table 4 Performance summary of CO₂-based electricity generators

Systems	Maximum output voltage	Energy conversion efficiency	Cycling life	Mechanism	Ref.
Photosynthesis-based electricity generator	2.4 V	38%	2 h	Coupling natural photosynthesis with mechanically induced electromagnetic generation	113
Nanosheet-agarose hydrogel (NAH)-based electricity generator	5 V (by connecting 50 of the generator)	0.6%	60 h	A spatial nanoconfinement-based ion separation strategy to convert CO2 adsorption into electrical energy	9

---

The first system achieves CO₂-to-electricity conversion by coupling natural photosynthesis with mechanically induced electromagnetic generation.¹¹³ In this design, CO₂ is dissoved in water at the first step. Then the aquatic plant, Ceratophyllum demersum, performs photosynthesis under illumination, converting dissolved CO₂ into O₂ bubbles (Fig. 8a). These bubbles accumulate beneath a superhydrophobic cone-shaped cap affixed to a buoyant (Fig. 8b). As buoyancy increases, the device ascends through a surrounding magnetic field. Upon reaching the air–water interface, gas release leads to a rapid density increase and initiates a downward motion. This cyclic surfacing-diving behavior, sustained by continuous photosynthetic activity, produces a time-varying magnetic flux that drives electricity generation via Faraday's law.

Fig. 8 Summary of CO₂-based electricity generation systems. (a) Atmospheric CO₂ dissolves in water under ambient conditions, serving as a carbon source for subsequent energy conversion. (b) A sunlight-driven mini-generator utilizes the harvested CO₂ to power multiple small electronic devices. (c)–(e) Photographs of the sunlight-powered generator system and its operation. (f) Time-dependent potential output curve of the solar-powered mini-generator. (a)–(f) Are adapted with permission.¹¹³ Copyright 2023, Wiley. (g) Schematic diagram of a prototype NAH (nano-adsorbent hydrogel) electricity generator, designed with a symmetric architecture between two electrodes to maintain balanced initial chemical potentials. (h) Mechanism of ion-separation-driven electricity generation from CO₂ adsorption: (i) polyethyleneimine (PEI) is grafted onto boron nitride nanosheets to form amine-functionalized adsorbents (h-BN-NH₂); (ii) upon CO₂ adsorption, charged species including cations (h-BN-NH₃⁺) and anions (HCO₃⁻) are released; (iii) the cations are spatially confined within nanostructured domains, while the anions diffuse through the hydrogel matrix, enabling efficient ion separation and electricity generation. (g) and (h) Are adapted with permission.⁹ Copyright 2024, Springer Nature.

The system achieves a maximum output voltage of 2.34 V and an energy conversion efficiency of up to 38%, sufficient to operate six LEDs for over two hours under ambient sunlight and using only naturally dissolved CO₂ (Fig. 8c–f). Notably, the process does not interfere with the intrinsic CO₂-recycling function of photosynthesis, as it simply harnesses the byproduct oxygen bubbles without introducing external energy inputs. This represents a rare example of energy harvesting from the thermodynamic minimum state of carbon. By integrating living biological components with engineered functional materials, the device offers a compelling strategy for sustainable power generation in environmentally benign, self-sustaining systems.

Overall, this technology enables CO₂-powered electricity generation without the need for external energy input. Its potential applications can be extended to weather monitoring, environmental sensing, and powering microelectronic devices. However, a key challenge is the strong dependence of photosynthesis on environmental conditions such as weather, which causes fluctuations in bubble generation and leads to unstable power output. To overcome this issue, it is necessary to optimize factors including light intensity and temperature to maintain a steady rate of oxygen production. In addition, the limited efficiency of electromagnetic induction results in relatively low voltage output. For example, a single surfacing-diving cycle generates only 0.32 to 2.34 V, with power output in the microwatt range restricting the system's practical applicability. Therefore, increasing the number of coil turns and adjusting the density of the movable device along with the amount of collected gas bubbles may help optimize the surfacing-diving motion, thereby improving voltage output and energy conversion efficiency.

The second system employs a spatial nanoconfinement-based ion separation strategy to convert CO₂ adsorption into electrical energy.⁹ Central to this approach is a nanosheet-agarose hydrogel (NAH) composite generator, which enables highly selective ion transport within its internal channels, thus facilitating direct electricity generation from CO₂ uptake (Fig. 8g). To address the inherent challenge of ion size similarity encountered with conventional adsorbents, a nanocomposite adsorbent was synthesized by grafting polyethyleneimine (PEI) with rich –NH/–NH₂ functional groups, onto two-dimensional hexagonal boron nitride (h-BN) nanosheets, forming PEI-functionalized h-BN (h-BN-NH₂). Upon exposure to CO₂, these nanocomposites undergo chemisorption and subsequently release large NH₃⁺ cations (100–300 nm) and angstrom-scale HCO₃⁻ anions, establishing a size disparity exceeding two orders of magnitude (Fig. 8h).

This significant size difference enables effective ion discrimination: the bulky NH₃⁺ ions, associated with the extended h-BN skeleton, are sterically confined within the agarose hydrogel matrix, while the smaller HCO₃⁻ anions freely diffuse through interconnected water-filled channels. As a result, a highly asymmetric ion transport is achieved, with a reported diffusion coefficient ratio (D⁻/D⁺) greater than 10⁶. This imbalance in ion migration creates a charge separation across the generator, giving rise to a measurable voltage. By connecting 50 of these generators in a vertically stacked configuration, the system delivers a stable output voltage of 5 V, which can power standard low-power devices such as light-emitting diodes.

Despite its promising performance and novel mechanism, low power density and limited output stability remain key challenges for NAH-based electricity generators. Although integrating multiple units can achieve open-circuit voltages up to 5 V, the output current (∼1 mA) is insufficient for high-power applications. Unlike conventional batteries that rely on rapid redox reactions, these generators produce electricity through ion migration driven by CO₂-induced chemical gradients. As a result, their voltage response is relatively slow, with noticeable time lag. Under continuous load, the voltage tends to drop and only recovers after the load is removed, indicating poor real-time performance. To address these challenges, it is essential to develop hydrogels with improved ion mobility and conductivity. Incorporating capacitive elements into hybrid designs can help buffer power fluctuations, while the introduction of micro- and nano-scale ion channels is suggested to enable more precise control over ion diffusion, thereby enhancing both response time and power output.

Together, these two systems highlight the diversity and potential of CO₂-based electricity generation. Despite being in the early stages of development, both systems underscore the transformative potential of converting CO₂ into a direct energy source, providing a foundation for future carbon-negative technologies.

Summary and perspectives

We are currently confronted with the dual challenges of a global energy crisis and escalating CO₂ emissions. In this context, utilizing CO₂ as a resource to generate or store electricity emerges as a promising bridge between carbon neutrality and energy innovation. This approach offers the potential to transform waste CO₂ into a valuable energy resource. Through the discussions presented in this work, we have explored four representative systems: metal-CO₂ batteries (MCBs), CO₂-capture-based energy storage, molten carbonate fuel cells (MCFCs), and CO₂-based electricity generators, all of which demonstrate considerable promise in converting CO₂ into usable energy. However, several critical challenges remain (Table 5).

Table 5 Comparison of the four technologies in mechanisms, advantages, and current limitations

Key technologies	Mechanisms	Advantages	Limitations
Metal-CO2 batteries	Cathode: CO2 reduction to various products during discharge and CO2/O2 evolution during charge	High theoretical energy density, particularly for Li–CO2 and Na–CO2 systems	Poor cycle life & stability because of the carbonate deposition, side reactions, and anode dendrite growth
		The CO2 utilization contributing to carbon neutrality	Scalability & safety concerns because of the gas management
	Anode: metal oxidation during discharge and reduction during charge	Generation of value-added chemicals/fuels, particularly in Zn–CO2 batteries	High cost of the anode metals
CO2 capture-integrated storage systems	CO2 capture and release mediated by energy storage device (e.g., supercapacitors, batteries) involving electro- and pH-swing mechanisms	Dual advantage of CO2 capture/release and energy storage	Early-stage and limited understanding
			Low cell voltage
		Extended applications in grid-storage	Poor cycling stability mainly due to reaction with atmospheric O2
			Complex system design and high maintenance
Molten carbonate fuel cells	Electricity generation through electrochemical reactions between fuel and oxidant, with carbonate ions serving as the charge carriers	Direct internal reforming of hydrocarbon fuels	High operating temperature accelerating material degradation
		Suitable for combined heat and power	High system cost and maintenance demand
		Integration with large stationary power applications
CO2-based electricity generators	Photosynthesis-generated O2 drives device motion to induce electricity via electromagnetic induction	Zero or low external energy input	Low and discontinuous voltage output
		Green and sustainable energy conversion	Slow energy conversion rate
	Carbon dioxide adsorption by spatially nanoconfined ion separation	Potential for integration with other devices for specific functions	Limited scalability
			Stability and durability challenges

---

For energy storage systems including MCBs and capture integrated energy storage, performance issues such as low energy efficiency, poor reversibility, and limited scalability hinder practical deployment. For CO₂ based electricity generation devices, the efficiency of CO₂ utilization, long-term operational stability, and cost-effectiveness remain significant barriers. To address these challenges and advance CO₂ utilization in energy systems, coordinated efforts are required across the following key areas.

• Performance engineering: materials design and optimization should prioritize the development of highly active, selective, and durable components, including electrodes, electrolytes, and separators. In MCBs, the development of bifunctional cathode catalysts remains critical to improve CO₂ conversion/utilization and energy storage. Besides, the challenges including poor interfacial contact due to solid-state discharge products necessitate the use of redox mediators to enhance charge transfer and battery reversibility. For integrated CO₂ capture-energy storage systems, redox active materials must endure prolonged cycling and resist oxidative degradation. Further, flow-cell designs must also be optimized to overcome low product yield and separation inefficiencies in MCBs and CO₂-capture integrated energy storage devices. The designing of oxygen insensitive and stable organic redox-active materials is the need of the hour. In MCFCs, material degradation is the main drawback caused by high-temperature operation resulting in corrosion and excessive thermal stress. This necessitates corrosion-resistant electrodes, thermally stable electrolytes and protective coatings to ensure enhanced durability and practical performance.

• Fundamental insights: complex interfacial mechanisms in MCBs need major emphasis to develop high-performance MCBs by employing advanced operando-techniques and correlating the inferences with theoretical and experimental analysis. Further, the exploration of new CO₂ conversion pathways is particularly important for CO₂-capture-based energy storage and electricity generation. The currently reported systems involve a variety of physical and chemical mechanisms by which CO₂ participates in energy-related reactions. However, these mechanisms remain largely underexplored or poorly understood, indicating substantial room for improvement and innovation. Interdisciplinary collaboration among researchers in physics, chemistry, biology, and materials science is essential to refine these pathways. By deepening our mechanistic understanding and optimizing reaction processes, it becomes possible to harness CO₂ more efficiently for power generation and energy storage. Lastly, in MCFCs, optimizing operation conditions and electrochemical activity require interfacial engineering at this stage to ensure practical viability.

• Device engineering: device architecture optimization plays a critical role in enhancing system stability, scalability, and real-world applicability. For example, CO₂-utilizing energy storage systems can be integrated with renewable energy and CO₂ capture technologies to build more sustainable and cost-effective platforms. Electricity generation devices should be scaled up through modular stacking designs to meet practical energy demands. MCFCs should be designed to efficiently integrate with power plants to enhance their practical significance.

• Furthermore, the current diversity of CO₂-utilizing systems is accompanied by a lack of standardized performance metrics, making it difficult to compare and evaluate different technologies. Therefore, the development of standardized testing protocols for system benchmarking is urgently needed to guide future research and enable fair comparisons across platforms.

By addressing these challenges through interdisciplinary innovation, CO₂-utilizing energy systems have the potential to evolve from conceptual technologies into practical solutions that support the global clean energy transition. This transformation could redefine CO₂ not merely as a greenhouse gas, but as a key resource for carbon mitigation and sustainable power generation.

Conflicts of interest

The authors declare no competing interests.

Data availability

No primary research results, software or code have been included and no new data were generated or analyzed as part of this review.

Acknowledgements

This work is supported by the Australian Research Council under grants DE250100603 and CE230100017, as well as Australia's Economic Accelerator Seed Program (Grant Number AE230100120). This research is supported by Sandland Bequest_Zou and an AINSE Ltd Early Career Researcher Grant (ECRG-J. Zou).

References

1. E. T. Vogt and B. M. Weckhuysen, Nature, 2024, 629, 295–306.
2. B. Soukharev, in Global Warming and Mass Migration: Climate Change and its Impact on Migration to the North, Springer, 2025, pp. 113–173.
3. T. A. Jacobson, J. S. Kler, M. T. Hernke, R. K. Braun, K. C. Meyer and W. E. Funk, Nat. Sustainability, 2019, 2, 691–701.
4. J. Cai, X. Li, B. Su, B. Guo, X. Lin, W. Xing, X. F. Lu and S. Wang, J. Mater. Sci., 2025, 234, 82–89.
5. J. Xiong, J. Ji, Q. Lei, X. Yang, Y. Bai, X. Zhang and H.-M. Cheng, eScience, 2025, 5, 100306.
6. L.-L. Wu, L.-Q. Yang, W.-X. Liu, T.-Y. Hang and X.-F. Yang, Rare Met., 2025, 1–35.
7. H. Liu, Q. Zhou, J. Yu, M. Nakabayashi, Y.-T. Lee, N. Shibata, Y. Li and J.-J. Delaunay, ACS Catal., 2025, 15, 8511–8521.
8. Y. Li, J. Zou, L. Sun, S. Liu, H. Li, Z. Song, J. Yu, L. Zhang and Z. Guo, Adv. Funct. Mater., 2025, 2509899.
9. Z. Wang, T. Hu, M. Tebyetekerwa, X. Zeng, F. Du, Y. Kang, X. Li, H. Zhang, H. Wang and X. Zhang, Nat. Commun., 2024, 15, 2672.
10. J. Zou, G. Liang, F. Zhang, S. Zhang, K. Davey and Z. Guo, Adv. Mater., 2023, 35, 2210671.
11. D. Gupta, J. Zou, J. Mao and Z. Guo, Energy Environ. Sci., 2025, 18(11), 5215–5249.
12. J. Zou, G. Liang, J. A. Yuwono, F. Zhang, Y. Fan, S. Zhang, B. Johannessen, L. Sun and Z. Guo, ACS Energy Lett., 2024, 9, 5145–5155.
13. Y. Chen, J. Li, B. Lu, Y. Liu, R. Mao, Y. Song, H. Li, X. Yu, Y. Gao and Q. Peng, Adv. Mater., 2024, 36, 2406856.
14. B. Lu, Z. Min, X. Xiao, B. Wang, B. Chen, G. Lu, Y. Liu, R. Mao, Y. Song and X. X. Zeng, Adv. Mater., 2024, 36, 2309264.
15. L. Sun, J. A. Yuwono, S. Zhang, B. Chen, G. Li, H. Jin, B. Johannessen, J. Mao, C. Zhang and M. Zubair, Adv. Mater., 2024, 36, 2401288.
16. Z. Cheng, Y. Fang, Y. Yang, H. Zhang, Z. Fan, J. Zhang, S. Xiang, B. Chen and Z. Zhang, Angew. Chem., Int. Ed., 2023, 135, e202311480.
17. X. Hu, J. Sun, Z. Li, Q. Zhao, C. Chen and J. Chen, Angew. Chem., Int. Ed., 2016, 55, 6482–6486.
18. Y. Wang, Y. Cheng, B. Chen, J. Zhou, H. Xie, Y. Fan, J. Sha, E. Liu, F. He and C. He, Energy Storage Mater., 2024, 71, 103655.
19. X. Hu, Z. Li, Y. Zhao, J. Sun, Q. Zhao, J. Wang, Z. Tao and J. Chen, Sci. Adv., 2017, 3, e1602396.
20. X. Li, G. Qi, J. Zhang, J. Cheng and B. Wang, Adv. Funct. Mater., 2022, 32, 2105029.
21. Y. Lu, Y. Cai, Q. Zhang, Y. Ni, K. Zhang and J. Chen, Angew. Chem., Int. Ed., 2021, 60, 9540–9545.
22. C. Zhang, A. Wang, L. Guo, J. Yi and J. Luo, Angew. Chem. Int. Ed., 2022, 134, e202200181.
23. W. Ma, X. Liu, C. Li, H. Yin, W. Xi, R. Liu, G. He, X. Zhao, J. Luo and Y. Ding, Adv. Mater., 2018, 30, 1801152.
24. W. Li, M. Zhang, X. Sun, C. Sheng, X. Mu, L. Wang, P. He and H. Zhou, Nat. Commun., 2024, 15, 803.
25. Y. Hou, J. Wang, L. Liu, Y. Liu, S. Chou, D. Shi, H. Liu, Y. Wu, W. Zhang and J. Chen, Adv. Funct. Mater., 2017, 27, 1700564.
26. G. Qi, J. Zhang, L. Chen, B. Wang and J. Cheng, Adv. Funct. Mater., 2022, 32, 2112501.
27. B. Xu, D. Zhang, S. Chang, M. Hou, C. Peng, D. Xue, B. Yang, Y. Lei and F. Liang, Cell Rep. Phys. Sci., 2022, 3(7), 100973.
28. R. Amin, M. Dixit, M. Li, R. Essehli, S. Neumayer, Y. Bai, A. Bisht, Y. Guang and I. Belharouak, J. Power Sources, 2024, 609, 234643.
29. C. Peng, L. Xue, Z. Zhao, L. Guo, C. Zhang, A. Wang, J. Mao, S. Dou and Z. Guo, Angew. Chem., Int. Ed., 2024, 136, e202313264.
30. C. Fetrow, C. Carugati, X. Yu, X.-D. Zhou and S. Wei, ACS Appl. Mater. Interfaces, 2023, 15, 12908–12914.
31. J. Xie, Q. Liu, Y. Huang, M. Wu and Y. Wang, J. Mater. Chem. A, 2018, 6, 13952–13958.
32. H. Wang, M. K. Aslam, Z. Nie, K. Yang, X. Li, S. Chen, Q. Li, D. Chao and J. Duan, Small Methods, 2024, 8, 2300867.
33. J. Xie, X. Wang, J. Lv, Y. Huang, M. Wu, Y. Wang and J. Yao, Angew. Chem., Int. Ed., 2018, 57, 16996–17001.
34. K. Miao, J. Qin, J. Yang and X. Kang, Adv. Funct. Mater., 2024, 34, 2316824.
35. X. Wang, J. Xie, M. A. Ghausi, J. Lv, Y. Huang, M. Wu, Y. Wang and J. Yao, Adv. Mater., 2019, 31, 1807807.
36. X. Mu, P. He and H. Zhou, Acc. Mater. Res., 2024, 5, 467–478.
37. Z. Lian, Y. Lu, C. Wang, X. Zhu, S. Ma, Z. Li, Q. Liu and S. Zang, Adv. Sci., 2021, 8, 2102550.
38. S. Ma, Y. Lu, H. Yao, Y. Si, Q. Liu and Z. Li, J. Energy Chem., 2022, 65, 472–479.
39. X. Zhang, T. Wang, Y. Yang, X. Zhang, Z. Lu, J. Wang, C. Sun, Y. Diao, X. Wang and J. Yao, ACS Energy Lett., 2021, 6, 3503–3510.
40. X. Mu, P. He and H. Zhou, Acc. Mater. Res., 2024, 5, 467–478.
41. N. Mahne, S. E. Renfrew, B. D. McCloskey and S. A. Freunberger, Angew. Chem., Int. Ed., 2018, 57, 5529–5533.
42. D. Cao, C. Tan and Y. Chen, Nat. Commun., 2022, 13, 4908.
43. S. Yang, Y. Qiao, P. He, Y. Liu, Z. Cheng, J.-j. Zhu and H. Zhou, Energy Environ. Sci., 2017, 10, 972–978.
44. Y. Qiao, J. Yi, S. Wu, Y. Liu, S. Yang, P. He and H. Zhou, Joule, 2017, 1, 359–370.
45. C. Wang, Y. Lu, S. Lu, S. Ma, X. Zhu, Z. Li and Q. Liu, J. Power Sources, 2021, 495, 229782.
46. S. Yang, P. He and H. Zhou, Energy Environ. Sci., 2016, 9, 1650–1654.
47. C. Kim, J. Kim, S. Joo, Y. Bu, M. Liu, J. Cho and G. Kim, iScience, 2018, 9, 278–285.
48. J. Ding, H. Zhang, W. Fan, C. Zhong, W. Hu and D. Mitlin, Adv. Mater., 2020, 32, 1908007.
49. W. Zhang, C. Hu, Z. Guo and L. Dai, Angew. Chem., Int. Ed., 2020, 59, 3470–3474.
50. X. Xiao, Z. Zhang, A. Yan, Y. Hao, K. Sun and P. Tan, Curr. Opin. Electrochem., 2024, 101599.
51. C. Yang, K. Guo, D. Yuan, J. Cheng and B. Wang, J. Am. Chem. Soc., 2020, 142, 6983–6990.
52. C. Wu, G. Qi, J. Zhang, J. Cheng and B. Wang, Small, 2023, 19, 2302078.
53. Z. Zhang, W. L. Bai, Z. P. Cai, J. H. Cheng, H. Y. Kuang, B. X. Dong, Y. B. Wang, K. X. Wang and J. S. Chen, Angew. Chem., Int. Ed., 2021, 133, 16540–16544.
54. X. Sun, X. Mu, W. Zheng, L. Wang, S. Yang, C. Sheng, H. Pan, W. Li, C.-H. Li and P. He, Nat. Commun., 2023, 14, 536.
55. A. Bharti, G. Achutharao and A. J. Bhattacharyya, ACS Appl. Mater. Interfaces, 2023, 15, 53342–53350.
56. Y. F. Wang, L. N. Song, L. J. Zheng, Y. Wang, J. Y. Wu and J. J. Xu, Angew. Chem., Int. Ed., 2024, 136, e202400132.
57. Y.-F. Wang, G.-J. Ji, L.-N. Song, X.-X. Wang and J.-J. Xu, ACS Energy Lett., 2023, 8, 1026–1034.
58. K. Iputera, J.-Y. Huang, S.-C. Haw, J.-M. Chen, S.-F. Hu and R.-S. Liu, J. Mater. Chem. A, 2022, 10, 3460–3468.
59. H. Wang, M. K. Aslam, Z. Nie, K. Yang, X. Li, S. Chen, Q. Li, D. Chao and J. Duan, Small Methods, 2024, 8, 2300867.
60. D. A. Vermaas and R. Kortlever, Joule, 2023, 7, 2426–2429.
61. M. Afshari, A. Refaie, P. Aleta, A. Hassan and M. Rahimi, ACS EST Eng., 2024, 5, 1099–1110.
62. S. Voskian and T. A. Hatton, Energy Environ. Sci., 2019, 12, 3530–3547.
63. Z. Xu, G. Mapstone, Z. Coady, M. Wang, T. L. Spreng, X. Liu, D. Molino and A. C. Forse, Nat. Commun., 2024, 15, 7851.
64. S. Pang, S. Jin, F. Yang, M. Alberts, L. Li, D. Xi, R. G. Gordon, P. Wang, M. J. Aziz and Y. Ji, Nat. Energy, 2023, 8, 1126–1136.
65. A. O. Mousa, C.-H. Chuang, S.-W. Kuo and M. G. Mohamed, Int. J. Mol. Sci., 2023, 24, 12371.
66. H. Seo and T. A. Hatton, Nat. Commun., 2023, 14, 313.
67. R. Sharifian, R. Wagterveld, I. Digdaya, C.-x. Xiang and D. Vermaas, Energy Environ. Sci., 2021, 14, 781–814.
68. S. Jin, M. Wu, R. G. Gordon, M. J. Aziz and D. G. Kwabi, Energy Environ. Sci., 2020, 13, 3706–3722.
69. L. Åkerlund, R. Emanuelsson, S. Renault, H. Huang, D. Brandell, M. Strømme and M. Sjödin, Adv. Energy Mater., 2017, 7, 1700259.
70. W. Ward III, Nature, 1970, 227, 162–163.
71. T. Harada, F. Simeon, E. Z. Hamad and T. A. Hatton, Chem. Mater., 2015, 27, 1943–1949.
72. S. Liu and D. Sinton, Nat. Chem. Eng., 2024, 1, 726–727.
73. K. S. Song, P. W. Fritz and A. Coskun, Chem. Soc. Rev., 2022, 51, 9831–9852.
74. S. Razavi, V. Rahmanian, R. Cai, M. Rukh, S. A. Khan and F. Li, J. Mater. Chem. A, 2025, 13, 9211–9221.
75. M. Massen-Hane, K. M. Diederichsen and T. A. Hatton, Nat. Chem. Eng., 2024, 1, 35–44.
76. G. H. Choi, H. J. Song, S. Lee, J. Y. Kim, M.-W. Moon and P. J. Yoo, Nano Energy, 2023, 112, 108512.
77. J. S. Kang, S. Kim and T. A. Hatton, Curr. Opin. Green Sustainable Chem., 2021, 31, 100504.
78. G. Iijima, J. Naruse, H. Shingai, K. Usami, T. Kajino, H. Yoto, Y. Morimoto, R. Nakajima, T. Inomata and H. Masuda, Energy Fuels, 2023, 37, 2164–2177.
79. X. Shi, H. Xiao, H. Azarabadi, J. Song, X. Wu, X. Chen and K. S. Lackner, Angew. Chem., Int. Ed., 2020, 59, 6984–7006.
80. P. Singh, J. H. Rheinhardt, J. Z. Olson, P. Tarakeshwar, V. Mujica and D. A. Buttry, J. Am. Chem. Soc., 2017, 139, 1033–1036.
81. B. Gurkan, F. Simeon and T. A. Hatton, ACS Sustain. Chem. Eng., 2015, 3, 1394–1405.
82. T. Kuramochi, A. Ramírez, W. Turkenburg and A. Faaij, Prog. Energy Combust. Sci., 2012, 38, 87–112.
83. D. P. Hanak and V. Manovic, Front. Chem. Sci. Eng., 2020, 14, 453–459.
84. Y. Shen and H. Yang, J. Clean. Prod., 2022, 379, 134869.
85. D. Luo, J. Liu, H. Wu, G. Zhang, Z. Pan and J. Huang, J. Energy Storage, 2025, 114, 115850.
86. S. Jin, M. Wu, Y. Jing, R. G. Gordon and M. J. Aziz, Nat. Commun., 2022, 13, 2140.
87. H. Seo, M. P. Nitzsche and T. A. Hatton, Acc. Chem. Res., 2023, 56, 3153–3164.
88. H. Jafarizadeh, E. Yamini, S. M. Zolfaghari, F. Esmaeilion, M. E. H. Assad and M. Soltani, Energy Rep., 2024, 12, 2179–2192.
89. A. Al-Mamoori, A. Krishnamurthy, A. A. Rownaghi and F. Rezaei, Energy Technol., 2017, 5, 834–849.
90. K. Sun, M. Tebyetekerwa, H. Zhang, X. Zeng, Z. Wang, Z. Xu, T. E. Rufford and X. Zhang, Adv. Energy Mater., 2024, 14, 2400625.
91. M. A. Abdelkareem, M. A. Lootah, E. T. Sayed, T. Wilberforce, H. Alawadhi, B. A. Yousef and A. Olabi, Sci. Total Environ., 2021, 769, 144243.
92. J. Selman, Int. J. Hydrogen Energy, 2021, 46, 15078–15094.
93. T. A. Barckholtz, K. M. Taylor, S. Narayanan, S. Jolly and H. Ghezel-Ayagh, Appl. Energy, 2022, 313, 118553.
94. R. Cooper, D. Bove, E. Audasso, M. Ferrari and B. Bosio, Int. J. Hydrogen Energy, 2021, 46, 15024–15031.
95. J.-H. Wee, Renewable Sustainable Energy Rev., 2014, 32, 178–191.
96. S. Campanari, P. Chiesa, G. Manzolini and S. Bedogni, Appl. Energy, 2014, 130, 562–573.
97. A. Amorelli, M. Wilkinson, P. Bedont, P. Capobianco, B. Marcenaro, F. Parodi and A. Torazza, Greenhouse Gas Control Technologies-6th International Conference, Pergamon, 2003, pp. 1325–1330.
98. S. Campanari, P. Chiesa and G. Manzolini, Int. J. Greenhouse Gas Control, 2010, 4, 441–451.
99. D. W. Hengeveld and S. T. Revankar, J. Power Sources, 2007, 165, 300–306.
100. P. Lunghi, S. Ubertini and U. Desideri, Energy Convers. Manage., 2001, 42, 1657–1672.
101. P. Lunghi, R. Bove and U. Desideri, J. Power Sources, 2003, 118, 108–117.
102. S. Campanari, J. Power Sources, 2002, 112, 273–289.
103. J. Milewski, J. Lewandowski and A. Miller, Turbo Expo: Power for Land, Sea, and Air, 2008, vol. 43123, pp. 389–395.
104. G. De Lorenzo and P. Fragiacomo, Fuel Cells, 2010, 10, 194–208.
105. D. Sánchez, S. Ubertini, J. M. De Escalona and R. Chacartegui, Int. J. Hydrogen Energy, 2014, 39, 4081–4088.
106. Y. Wang, Y. Huang, D. McIlveen-Wright, N. Hewitt, P. Eames, S. Rezvani and J. McMullan, Fuel, 2006, 85, 2133–2140.
107. P. Greppi, B. Bosio and E. Arato, Ind. Eng. Chem. Res., 2013, 52, 8755–8764.
108. P. S. Pak, Y. D. Lee and K. Y. Ahn, Energy, 2009, 34, 1903–1909.
109. Z. Lu, H. Zhang, L. Duan, R. Zhang, C. Lu and U. Desideri, Fuel, 2024, 357, 130032.
110. O. Dybiński, J. Milewski, A. Szczęśniak, A. Martsinchyk and Ł. Szabłowski, Int. J. Hydrogen Energy, 2024, 52, 889–901.
111. N. Tavakoli, F. Pourfayaz and M. Mehrpooya, Energy, 2025, 314, 134319.
112. A. Zahedi, E. Rasmi, A. Ahmadi and B. Kanani, Renewable Energy, 2024, 237, 121749.
113. D. Wang, J. Wang, R. Ming, C. Liu, P. Tian, M. Cheng, M. Xiao and F. Shi, Adv. Funct. Mater., 2023, 33, 2300019.

---

Footnote

† Equal contribution.

---