English

• With the rapid advancement of energy storage systems, rechargeable Li-CO₂ battery has garnered considerable research interest due to the potential for CO₂ capture and energy conversion [1-6]. Furthermore, owing to its high theoretical specific capacity (1876 Wh/kg), Li-CO₂ battery is used as a promising strategy for reducing fossil fuel consumption in meeting high energy demands and alleviating the global greenhouse effect. Li₂CO₃ is an insulating material that accumulates on the electrode surface, which obstructs the active sites and gas diffusion channels and thus increases the impedance of rechargeable Li-CO₂ battery. However, the decomposition of Li₂CO₃ necessitates the operation at a high potential, which not only induces the electrolyte degradation but also affects the cycling life of battery [7-9]. Moreover, the Li dendrites growth as well as the volume expansion of Li metal caused from repeated Li plating/stripping during cycling, are also potential threats to cycle life and safety performance.

  To cope with the high charging potential resulting from the decomposition of Li₂CO₃, various solid catalysts have been developed in cathodes, including Ru, Cu nanoparticles, Mo₂C/CNT, N—CNTs@Ti, Ru-Cu-G, CFB@NCNT-Mo₂N, Mo₃N₂, and RuO₂@a-MWCNTs [9-16]. These cathode catalysts provide sufficient catalytic active sites, which effectively improving the slow dynamics of CO₂ reduction reaction (CO₂RR) and electrochemical reduction of CO₂ (CO₂ER), as well as facilitating the decomposition of Li₂CO₃ to improve the electrochemical performance. Nevertheless, the utilization of metal catalysts results in an unfavorable morphology of discharge products, as well as incurs high production costs due to the presence of metals and complex preparation processes [17-19].

  Electrolyte is considered to serve as a crucial factor in improving the electrochemical performance, but the incorporation of different additives can cause certain deficiencies that directly affect the longevity of batteries. The decomposition of electrolyte and degradation of cathode due to high overpotential severely affect the energy efficiency and the cycling performance of Li-CO₂ batteries. One widely employed strategy for minimizing the overpotential of batteries involves utilizing redox mediators (RMs), with halogens being extensively investigated as RMs in Li-O₂ batteries. It is illustrated that 1, 2-dimethyl-3-propylimidazole iodide (DMPII) can effectively mitigate the shuttle effect of I⁻ by generating a stable "self-defense" solid electrolyte interphase (SEI) in situ to protect Li anode [20]. Moreover, the organic iodide acetylthiocholine iodide (ATCl), containing N and S, is proposed as a defensive donor RM for Li anode, effectively addresses the high charge overpotential and unstable Li metal interface. Furthermore, the introduction of ATCl results in a prolonged lifespan up to 190 cycles, approximately six times that of the Li-O₂ battery without additive (~27 cycles), while achieving a charge potential of 3.49 V [21]. Additionally, Phenylbromosilane (TBMPSiBr) is introduced as a RM and film-forming agent for Li-O₂ batteries. Li dendrite growth is effectively inhibited by the SEI layer resulting from π coupling, while the overpotential can be reduced though the dissociated Br₃⁻/Br⁻. Consequently, the inclusion of TBMPSiBr enables stable cycling of Li-O₂ batteries for 220 cycles at the current density of 1000 mAh/g [22]. According to previous studies, the following reaction is feasible, as indicated by Eq. 1:

  $ 2 \mathrm{Li}_2 \mathrm{CO}_3+\mathrm{C}+2 \mathrm{Br}_2=3 \mathrm{CO}_2+4 \mathrm{LiBr} $

  (1)

  Moreover, the electromotive force value for this reaction stands at −0.39 V [23,24]. In other words, the reaction between electrochemically generated Br₂ and Li₂CO₃ demonstrates the viability of LiBr as the RM, primarily through the redox reaction involving Br₃⁻/Br⁻ electric pairs, which effectively mitigates the overpotential of Li-CO₂ batteries during cycling.

  Although RM has been proven effective in reducing overpotential, it is susceptible to diffuse from the separator to Li anode and react with Li anode subsequently, commonly known as the "shuttle effect", which not only triggers Li anode corrosion but also leads to the rapid failure of batteries [25-31]. Consequently, for further advancements in Li-CO₂ batteries, it becomes imperative to address both Li anode protection and overpotential reduction.

  To improve the interface instability of Li anode, diverse strategies have been employed to optimize its morphology. For instance, nanostructures have been utilized to mitigate volume fluctuations in Li anode, while electrolyte additives have been employed to form a robust SEI layer [32-43]. Additionally, manipulation of solvation structure within the electrolyte has been explored to modulate ion transport, and sacrificial coatings or other pre-treatments have been investigated for modifying the initial formation and deposition behavior of Li [40,44-51]. The common goal of these approaches is to modulate chemical, morphological or mechanical properties of the interface through engineered SEI layers, thereby enabling rapid and uniform ion transport as well as reversible volume changes [52-57]. Despite some progress achieved by the above-mentioned methods, the challenges related to interface instability still persist in Li anode due to a limited comprehension of the underlying physical and chemical mechanisms.

  To tackle the instability challenges at the Li anode interface, an efficient approach involves electrostatic shielding of heterogeneous surface regions where Li⁺accumulate and deposit under high electric fields, thereby fundamentally altering the pathway of Li⁺ deposition. This can be achieved experimentally by introducing electrically inactive cations at the interface, which effectively "shield" uneven protuberances on the electrode surface. Studies suggests that by adjusting the concentration of Cs⁺ based on the Nernst equation, a lower reduction potential than that of Li⁺ can be attained. Additionally, the result indicates that the introduction of Cs⁺ can improve the morphology of Li deposition [58,59]. In addition to Cs⁺, there are many other organics with similar mechanisms, such as the surfactant trimethylammonium chloride (CTAC), the ionic liquid (Pyr1(12)FSI) [60,61]. Among them, the cations exhibit an electrostatic shielding mechanism, which adsorb around the Li protuberances and repel Li⁺ by electrostatic repulsive forces during cycling. Therefore, Li⁺ is induced to deposit uniformly in the vicinity of protuberances, slowing down Li dendrites formation, thus improving the electrochemical performance of batteries.

  Recently, imidazole-based ionic liquids (ILs) have garnered significant attention due to the low viscosity, non-volatility, low flammability, and exceptional ionic conductivity. Moreover, imidazolium-based ILs are typically non-toxic or low-toxic and relatively low cost, making them highly appealing for industrial applications [62]. In addition, imidazolium based ILs have no harmful side effects on the components of battery and cannot affect the electrochemical performance. And the stability of imidazolium ILs and the properties of promoting CO₂ dissolution are crucial for the long-term lifespan of batteries [63,64]. During charging, Br⁻ is oxidized to Br₃⁻ and acts as a RM in the electrolyte, effectively reducing the charge overpotential [65]. To the best of our knowledge, an organic bromide containing an imidazole group has not been previously reported as RM for Li-CO₂ batteries.

  Keeping all these factors in mind, current study is proposed to introduce EMIBr as a bifunctional additive, synergistically combining the effects of EMI⁺ and Br⁻ to achieve homogeneous Li deposition and reduced overpotential simultaneously. During cycling, the RMs (Br₃⁻/Br⁻) can reduce the overpotential, while EMI⁺ can be adsorbed on the Li anode leading to Li deposition uniformly through electrostatic shielding, rather than at the dendritic tips or uneven surface. The electrostatic protective layer formed by EMIBr not only inhibits DMSO-induced corrosion on anode, but also reduces the side reactions between RMs and Li anode. Therefore, the incorporation of EMIBr is confirmed to suppress the formation of Li dendrites on the anode and enhance the cycling stability effectively. Consequently, EMIBr-containing Li-CO₂ battery shows a lower overpotential (~1.17 V), favorable cycling stability and good reversibility, surpassing KBr-based Li-CO₂ battery. In the current density of 500 mA/g, a longer lifespan of 200 cycles has been obtained for Li-CO₂ battery with EMIBr, which is five times than that of Li-CO₂ battery without additive. This work presents a feasible strategy for developing multifunctional RM materials for Li-CO₂ battery.

  The stability of EMIBr (the molecular structural formula as shown in Fig. 1a) within the operating voltage range of Li-CO₂ batteries has been evaluated though LSV. As depicted in Fig. S1a (Supporting information), the oxidation peak observed between 3.4–4.0 V (vs. Li⁺/Li) can be attributed to the presence of Br anion, based on the following redox reactions [66]:

  $ 3 \mathrm{Br}^{-} \rightarrow \mathrm{Br}_3^{-}+2 \mathrm{e}^{-} E=3.48 \mathrm{~V} \text{ }{ vs. \text{ }} \mathrm{Li}^{+} / \mathrm{Li} $

  (2)

  $ 2 \mathrm{Br}_3^{-} \rightarrow 3 \mathrm{Br}_2+2 \mathrm{e}^{-} E=4.02 \mathrm{~V} \text{ } { vs. }\text{ } \mathrm{Li}^{+} / \mathrm{Li} $

  (3)

  Figure 1

  

  Figure 1.  (a) The molecular structural formula of EMIBr. (b) DFT calculations of binding energy of DMSO and EMI⁺ on Li anode. (c) Schematic illustrations of the different working mechanism in the EMIBr- or KBr-containing Li-CO₂ batteries.

  DownLoad: Full-Size Img PowerPoint

  The other peak observed at 4.6 V (vs. Li⁺/Li) corresponds to the potential at which electrolyte oxidation and decomposition occurs. Additionally, no additional oxidation and reduction peak is detected in the EMIBr-containing electrolyte compared to the electrolyte without additive (Fig. S1b in Supporting information). Prior to Li⁺ deposition, EMI⁺ is not preferentially reduced and exhibits favorable stability, confirming the electrochemical stability of EMIBr within the operating voltage range of Li-CO₂ batteries (2.2–4.5 V), demonstrating that EMIBr can be used as a non-consumable additive for electrostatic shielding.

  The molecular structure formula of EMIBr is presented in Fig. 1a. Subsequently, the binding energy of DMSO and EMI⁺ on Li anode is calculated (Fig. 1b). In previous theoretical studies, it indicates that the (110) crystal plane of Li metal exhibits the lowest crystal plane energy and diffusion barrier for Li⁺ in the electrolyte. Therefore, the Li (110) crystal plane is selected as the basis for DFT calculations [67,68]. The results reveal a positive binding energy value of 0.155 eV for DMSO and a negative binding energy of −3.09 eV for EMI⁺, indicating that EMI⁺ can be preferentially adsorbed on the Li anode surface during electrodeposition, thereby generating an electrostatic shielding effect to mitigate Li dendrites formation.

  The effect of EMIBr on Li deposition is illustrated in Fig. 1c. During the initial Li deposition, uneven tips are inevitably grown on the anode, leading to stronger electric fields and dendrite formation in the Li-CO₂ batteries without additive. Both in KBr-based and EMIBr-based Li-CO₂ batteries, Br^– can act as RM to reduce the overpotential during cycling. However, Br^– and Br₃^– will act with Li anode due to the "shuttle effect" of bromine ions, thereby attacking and causing corrosion of Li anode during cycling. However, when incorporating EMIBr into electrolyte, EMI⁺ can be adsorbed surrounding the raised tips through electrostatic interactions and form protective layers on these tips. Consequently, approaching Li⁺ are repelled and transferred to adjacent flat areas with minimal EMI⁺ coverage, until a homogeneous Li deposition is achieved (Fig. 1c). Overall, EMIBr can be used as a bifunctional electrolyte additive to reduce overpotential while inhibiting Li dendrite growth and promoting uniform Li deposition.

  To directly investigate the effect of EMI⁺-derived Li repellent protective layer on the stability of Li anode, the cycling performance of Li||Li symmetric cells assembled with different electrolytes have been studied. Incorporating EMIBr in the Li||Li symmetric cell results in lower overpotential, as depicted in Fig. 2a. Li||Li symmetric cell without additive shows a poor cycling stability (< 250 h) at a current density of 0.1 mA/cm² and a cut-off capacity of 0.05 mAh/cm², with a larger overpotential (63 mV at 20 h, 318 mV at 200 h). Although the cycle life improved after incorporating KBr into electrolyte, the overpotential increased sharply after 300 h, indicating certain limitations. Conversely, the stable Li plating/stripping behavior (> 450 h) under identical test condition is obtained in the battery with EMIBr, accompanied by small overpotentials (40 mV at 20 h, 60 mV at 200 h). The above results demonstrate that exceptional cycling performance and lower overpotential for approximately 480 h can be obtained in Li||Li symmetric cell with EMIBr. The decent cycling stability confirms the formation of Li repellent protective layers, which effectively suppresses side reactions and consequently reduces by-product accumulation as well as "dead Li".

  Figure 2

  

  Figure 2.  (a) Cycling performances of Li||Li symmetric cells with various electrolytes at current density of 0.1 mA/cm² and limited capacity of 0.05 mAh/cm² and enlarged voltage profiles with periods of 20–50 h and 150–200 h. SEM images (top) together with the corresponding cross-sectional views (bottom) of Li anodes in batteries with (b, c) pure, (d, e) KBr-, and (f, g) EMIBr-containing electrolytes after 10 cycles at 500 mA/g.

  DownLoad: Full-Size Img PowerPoint

  To further manifest the protective benefits of EMIBr on Li anode, the surface morphologies of Li anode with/without additive after cycling have been characterized via SEM images (Figs. 2b-g). In the absence of any additive, cracks and holes are detected on Li anode after only 10 cycles (Figs. 2b and c). In addition, many mossy cracks are noticed on the Li anode in additive-free and KBr-containing batteries after 40 cycles (Figs. S2a and c in Supporting information), which result from the attack of oxygen species and Br₃^–. Furthermore, it shows that Li anodes become loose and porous in KBr-containing and additive-free batteries after cycling, indicating that Li anodes are eroded from surface to inside (Figs. 2c and e). In contrast, the presence of EMIBr results in a flat surface for Li anode even after 10 and 40 cycles (Fig. 2f and Fig. S2e in Supporting information). The surface of Li anode in EMIBr-containing battery is smoother and maintains a tight cross-section, effectively alleviating corrosion problems, compared to the batteries without additive or with KBr (Figs. 2b and d, Figs. S2a and c). Based on the above findings, it can be concluded that Li dendrite growth can be suppressed via the addition of EMIBr to produce a smoother and more stable interface, allowing for decent cycling performance. The electrochemical performance of Li||Li symmetric cells have been evaluated based on different concentrations of EMIBr ranging from 10 mmol/L to 80 mmol/L (Fig. S3 in Supporting information). As shown in Fig. S3a, the lifespan of the Li||Li symmetric cell increases with higher EMIBr concentration. Specifically, the Li||Li symmetric cell containing 50 mmol/L EMIBr exhibits a stable voltage curve and a lifespan exceeding 450 h, given at a limited capacity of 0.05 mAh/cm² and a current density of 0.1 mA/cm². However, the lifespan of the cell with 80 mmol/L EMIBr decreases due to the exacerbation of shuttle effect from high bromide ion concentration. Figs. S3b and c illustrate the voltage profiles at 100–150 h and 300–350 h, respectively. The presented results show an overpotential within less than 120 mV can be achieved in the Li||Li symmetric cell based on 50 mmol/L EMIBr even after cycling for 300 h, indicating the favorable cycling stability.

  To unveil the role of EMIBr in reducing overpotential, galvanostatic discharge/charge measurements have been conducted across various current densities. Fig. 3a shows the initial discharge/charge voltage profiles of Li-CO₂ batteries without additive, with KBr or EMIBr at a current density of 200 mA/g and a fixed capacity of 1000 mAh/g. The charge plateau for the additive-free Li-CO₂ battery is observed to be 4.22 V, as anticipated, while the discharge plateau measures at 2.65 V. In contrast, the KBr-based Li-CO₂ battery exhibits a reduced charge plateau of 4.02 V and a discharge plateau of 2.72 V, indicating that a mitigating effect on polarization can be obtained by KBr. It is worth noting that the EMIBr-containing Li-CO₂ battery shows a significantly lower charging potential (~3.9 V, overpotential: ~1.17 V), compared to the Li-CO₂ battery without additive (~4.2 V, overpotential: ~1.57 V), as shown in Fig. 3b. The remarkable reduction in overpotential can be attributed to the enhanced oxidation kinetics of Li₂CO₃ facilitated by Br anion redox couples and the repellent protective layers formed by EMI cations [65].

  Figure 3

  

  Figure 3.  (a) Voltage curves and (b) the corresponding overpotentials obtained at the first discharge/charge process for the Li-CO₂ batteries without additive, with KBr or EMIBr. (c) The discharge/charge terminal voltage versus cycle number of the Li-CO₂ batteries without additive, with KBr or EMIBr. Cycling stability of Li-CO₂ batteries (d) without additive, with (e) KBr and (f) EMIBr at a current density of 500 mA/g.

  DownLoad: Full-Size Img PowerPoint

  The corresponding discharge/charge potential versus cycle numbers for Li-CO₂ batteries with EMIBr, KBr or without additive have been studied at a limited capacity of 1000 mAh/g and a current density of 500 mA/g (Fig. 3c), The EMIBr-containing Li-CO₂ battery exhibits decent cycling stability with an extended cycle life of approximately 200 cycles, which is about five times that of additive-free Li-CO₂ battery (44 cycles). The results demonstrate that the utilization of EMIBr not only leads to a reduction in overpotential, but also significantly enhances the cycling stability of Li-CO₂ batteries.

  To further prove the advantage of defense-donor RM, the electrochemical performance of Li-CO₂ batteries without additive, with KBr and EMIBr have been investigated (Figs. 3d-f). The results reveal that the KBr-containing Li-CO₂ battery exhibits high polarization and a significant increase in overpotential during 72 cycles. As a more intense contrast, a relatively low and stable charging platform below 4.3 V for up to 200 cycles have been obtained for Li-CO₂ battery with EMIBr. The results prove that a protective layer is effectively established through the electrostatic incorporation of EMI⁺, thereby safeguarding Li anode against soluble Br₃⁻ attacks. Moreover, the protective layer avoids excessive consumption of Br₃⁻/Br⁻, inhibits Li dendrites growth, facilitates uniform Li deposition, and significantly enhances the electrochemical performance. Furthermore, the superior interface stability of EMIBr-based Li-CO₂ batteries have been further confirmed via EIS results. Before cycling, the charge transfer impedance of EMIBr-containing Li-CO₂ battery is larger than that of additive-free Li-CO₂ battery (Fig. S4 in Supporting information). The increase in impedance values is mainly attributed to the adsorption of additive on the surface of Li anode, which hinders the transport of Li ions. However, the interface resistance of EMIBr-containing Li-CO₂ batteries is significantly lower than that of additive-free Li-CO₂ batteries after cycling. The presence of Li dendrites and dead Li trigger more by-products on the surface of Li anode, which increases the ion diffusion resistance near the anode. With the inclusion of EMIBr, the accumulation of Li dendrites and dead Li are effectively suppressed by EMI-derived protective layer, thereby reducing the interface resistance.

  In addition, the effect of varying concentrations of EMIBr on the cycling lifespan of Li-CO₂ batteries have been further investigated (Fig. S5 in Supporting information). To prevent Li dendrite growth, the cations need to possess a sufficient concentration to establish an effective electrostatic barrier. As illustrated in Figs. S5b–d, a low concentration of EMIBr (for example, 10 mmol/L in this work) exhibits an inability to meet the concentration requirements for the dynamic shield, and there is no enough EMI⁺ to effectively cover the rapidly growing protrusion tips, resulting in shortened cycle life and reduced energy efficiency [59]. Conversely, employing a high concentration of EMIBr (for example, 80 mmol/L in this work) leads to an evident shuttle effect caused by increased Br ion concentration. The severe shuttling effect results in increased polarization during cycling, reaching the cut-off voltage of charging platform and consequently shortening the cycle life. The Li-CO₂ battery based on 50 mmol/L EMIBr exhibits an optimal concentration of EMI⁺ for effective electrostatic shielding, while mitigating the shuttle effect of bromide ion. Therefore, an optimal concentration of 50 mmol/L for EMIBr is selected for this work. Moreover, a significant decrease in overpotential and prolonged cycle life of 200 cycles can be obtained for the Li-CO₂ battery with 50 mmol/L EMIBr, indicating the improved electrochemical performance (Fig. 3f).

  To explore whether EMI⁺ is reduced, XPS analysis have been employed to examine the differences in SEI formed in Li-CO₂ batteries without additive and with 50 mmol/L EMIBr. As depicted in Fig. 4, there is no significant disparity observed in the composition of C 1s and N 1s spectra. Specifically, peaks corresponding to CO₃^2- (289.8 eV), C=O (288.4 eV), C-O (286.4 eV) and C-C (284.8 eV) are detected, which may be attributed to solvent reduction [69]. Notably, no peak related to N-C is observed in both C 1s and N 1s spectra, indicating the composition of SEI film is not affected though the introduction of EMI⁺. Moreover, no reduction products associated with EMI⁺ are observed in the SEI film, which prove that EMIBr is stable and not reduced. Additionally, significant differences are noted in the peak area and height for CO₃²⁻ and C-O compounds within the C 1s spectra. This finding suggests that the electrolyte containing EMI⁺ leads to a lower presence of C-O compounds but higher levels of CO₃²⁻-containing compounds within the SEI film, which corresponds well with fitting results from O 1s and Li 1s spectra (Fig. 4e and Fig. S6c in Supporting information). The stability of SEI film can be significantly enhanced by incorporating Li₂CO_3, owing to its thermodynamic stability that resists further decomposition or dissolution into electrolyte. Moreover, the inclusion of Li₂CO₃ facilitates the formation of a cross-linked structure with other organic components, thereby augmenting the strength and resilience of SEI film against electrode volume changes. Therefore, the introduction of EMI⁺ simultaneously enhances the mechanical properties and Br₃⁻ corrosion resistance of SEI film, thus improving the cycling stability of Li anode.

  Figure 4

  

  Figure 4.  XPS spectra of C 1s, O 1s and N 1s of the Li anodes disassembled from Li-CO₂ batteries (a-c) without additive and (d-f) with EMIBr, respectively.

  DownLoad: Full-Size Img PowerPoint

  To gain insights into the reversibility of Li-CO₂ batteries containing EMIBr, the generation and decomposition of discharge products have been investigated via SEM images. As shown in Fig. 5, the surface morphologies of CNT-based cathode have been examined during discharge/charge processes, respectively. Unlike the bulk particles of Li-CO₂ batteries without additive (Fig. 5d), film-like discharge products are observed on the cathode surface in Li-CO₂ batteries containing EMIBr (Fig. 5b), suggesting that incorporating EMIBr into electrolyte may influence the morphology of discharge products. After charging, almost discharge products are decomposed in both the Li-CO₂ batteries containing EMIBr or without additive (Figs. 5c and e), and the morphologies of cathode are restored to similar as the pristine CNT-based cathode (Fig. 5a), which prove that the favorable reversibility of EMIBr-containing Li-CO₂ batteries.

  Figure 5

  

  Figure 5.  SEM micrographs of CNT-based cathodes with different electrolytes after the discharge and charge process: (a) Pristine; (b) with 50 mmol/L EMIBr, discharge; (c) with 50 mmol/L EMIBr, charge; (d) without additive, discharge; (e) without additive, charge. (f) XRD patterns of discharged and charged CNT-based cathodes in EMIBr-containing batteries after 5 cycles at a current density of 500 mA/g.

  DownLoad: Full-Size Img PowerPoint

  To characterize the composition of discharge products, X-ray diffraction (XRD) patterns have been further analyzed. As depicted in Fig. 5f, it shows that Li₂CO₃ is the primary component of discharge products in EMIBr-containing Li-CO₂ batteries after the initial discharge. After charging, the XRD pattern of CNT-based cathode is returned to the original state, indicating the decomposition of Li₂CO₃. The discharge/charge mechanism in Li-CO₂ batteries is generally governed by the reversible formation and decomposition of Li₂CO₃, and the electrochemical reaction can be expressed as Eq. 4:

  $ 4 \mathrm{Li}^{+}+3 \mathrm{CO}_2+4 \mathrm{e}^{-} \rightarrow 2 \mathrm{Li}_2 \mathrm{CO}_3+\mathrm{C} $

  (4)

  The XRD results strongly support the decent reversibility on the cathode side for EMIBr-containing Li-CO₂ batteries, consistent with the SEM images.

  In accordance with the above findings, the specific mechanism of EMIBr in the Li plating process can be categorized into the following aspects. Self-healing electrostatic shielding (SHES) mechanism, which fundamentally regulates the deposition pathway of Li ions by adding organic cations with lower adsorption energy. As Li ions are deposited, the cations form a charge barrier around Li protuberance, repelling Li ions and preventing more Li ions from being deposited on the protuberance, instead allowing Li to be deposited in depressions or other flatter areas, which prevents dendrites from self-growing and spreading. A key feature is that the additive cation is an adsorption behavior rather than a chemical reaction at the Li anode, and the cation is not reduced or consumed during the adsorption process. Ideally, the additive is "reused" like a catalyst during cycling and not consumed, resulting in a longer cycle life than other consumable additives.

  Electrochemical stability and preferential adsorption at the Li anode surface are keys to the selection of non-consumable additives, and this can be illustrated by the LSV curves in Fig. S1 (Supporting information), the DFT calculations in Fig. 1b, the XPS spectrum in Fig. S7 (Supporting information) and the electrochemical cycling curves in Figs. 2a and 3f. The results presented in Fig. 1b demonstrate the preferential adsorption of EMI cations on the surface of Li anode over the electrolyte solvent (DMSO). Furthermore, there is no signal of elemental N detected on the SEI in the XPS spectrum, indicating that EMI⁺ is not involved in the reaction process of SEI formation, but only exhibits physisorption behavior, at least within a limited current density. The longer cycle life of EMIBr-based battery proves that EMIBr is not reduced or consumed during cycling. Consistent with previous studies (Table S1 in Supporting information), this study further confirms the important role of RM in enhancing the discharge capacity and improving the overall performance of Li-CO₂ batteries (Fig. S8 in Supporting information).

  In summary, it has been proved that the key challenges associated with Li-CO₂ batteries, including high charge overpotential and Li dendrite growth, can be skillfully solved by employing EMIBr as a defense donor RM. On one hand, the bromine anion exhibits a significant effect on reducing overpotential as RM. Meanwhile, owing to the electrostatic adsorption, imidazolium cations facilitate the formation of a Li-repellent protective layer with anti-bromine anion corrosion properties, resulting in more uniform Li deposition and significantly improves the stability of Li anode. Therefore, the EMIBr-containing Li-CO₂ batteries show remarkable performance with lower overpotential (~1.17 V) and longer cycle life (~200 cycles). Additionally, the impact of ionic liquid cations and halogen anions on Li nucleation/growth is systematically investigated, as well as a simple and effective strategy for reducing overpotential and simultaneously stabilizing Li anode in Li-CO₂ batteries is proposed.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Aonan Wang: Writing – review & editing, Writing – original draft, Investigation, Data curation, Conceptualization. Jingwen Dai: Writing – review & editing, Writing – original draft, Methodology, Data curation. Yiming Guo: Writing – review & editing, Writing – original draft, Formal analysis. Fanghua Ning: Writing – review & editing, Writing – original draft, Software, Methodology, Formal analysis. Xiaoyu Liu: Formal analysis, Methodology, Writing – original draft. Sidra Subhan: Writing – review & editing, Writing – original draft, Methodology. Jiaqian Qin: Writing – review & editing, Writing – original draft, Methodology, Formal analysis. Shigang Lu: Writing – review & editing, Writing – original draft, Resources, Methodology. Jin Yi: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization.

  Acknowledgments

  This work was financially supported by National Natural Science Foundation of China (No. 22075171). We gratefully acknowledge HZWTECH for providing computation facilities.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110186.

  
  
• 1. [1]

     S. Min, X. Liu, A. Wang, et al., Chin. Chem. Lett. 34 (2023) 108586.

  2. [2]

     H. Ouyang, S. Min, J. Yi, et al., ACS Appl. Mater. Interfaces 14 (2022) 53648–53657. doi: 10.1021/acsami.2c13807

  3. [3]

     H. Ouyang, S. Min, J. Yi, et al., Green Energy Environ 8 (2023) 1195–1204.

  4. [4]

     G. Qi, J. Zhang, J. Cheng, et al., Small 20 (2024) 2309064.

  5. [5]

     G. Qi, J. Zhang, L. Chen, B. Wang, J. Cheng, Adv. Funct. Mater. 32 (2022) 2112501.

  6. [6]

     Y. Liu, J. Cai, J. Zhou, et al., eScience 2 (2022) 389–398.

  7. [7]

     S.R. Gowda, A. Brunet, G.M. Wallraff, B.D. McCloskey, J. Phys. Chem. Lett. 4 (2013) 276–279. doi: 10.1021/jz301902h

  8. [8]

     J.M. Garcia-Lastra, J.S.G. Myrdal, R. Christensen, K.S. Thygesen, T. Vegge, J. Phys. Chem. C 117 (2013) 5568–5577. doi: 10.1021/jp3107809

  9. [9]

     S. Yang, Y. Qiao, P. He, et al., Energy Environ. Sci. 10 (2017) 972–978.

  10. [10]

      Z. Zhang, Z. Zhang, P. Liu, et al., J. Mater. Chem. A 6 (2018) 3218–3223. doi: 10.1039/c7ta10497a

  11. [11]

      Y. Hou, J. Wang, L. Liu, et al., Adv. Funct. Mater. 27 (2017) 1700564.

  12. [12]

      Y. Li, J. Zhou, T. Zhang, et al., Adv. Funct. Mater. 29 (2019) 1808117.

  13. [13]

      Z. Zhang, C. Yang, S. Wu, et al., Adv. Energy Mater. 9 (2019) 1802805.

  14. [14]

      Z. Wang, Y. Cai, Y. Ni, et al., Chin. Chem. Lett. 34 (2023) 107405.

  15. [15]

      L. Chen, J. Zhou, Y. Wang, et al., Adv. Energy Mater. 13 (2023) 2202933.

  16. [16]

      G. Qi, J. Zhang, J. Cheng, B. Wang, SusMat 3 (2023) 276–288. doi: 10.1002/sus2.123

  17. [17]

      T. Jian, W. Ma, C. Xu, H. Liu, J. Wang, eScience 3 (2023) 100114.

  18. [18]

      Y. Zhou, S. Guo, eScience 3 (2023) 100123.

  19. [19]

      A.M. Mansour, J. Metal Mater. Minerals 30 (2020) 1–18. doi: 10.55713/jmmm.v30i4.871

  20. [20]

      D. Li, C. Zhu, M. Zhang, et al., Chem. Eng. J. 408 (2021) 127335.

  21. [21]

      Z. Sun, Y. Tan, Y. Zhan, et al., Small Methods 7 (2023) e2201289.

  22. [22]

      C.Y. Li, M.S. Wu, W.R. Chen, et al., J. Mater. Chem. A 11 (2023) 937–942. doi: 10.1039/d2ta07201g

  23. [23]

      X.G. Wang, C. Wang, Z. Xie, et al., ChemElectroChem 4 (2017) 2145–2149. doi: 10.1002/celc.201700539

  24. [24]

      A. Jain, S.P. Ong, G. Hautier, et al., APL Mater. 1 (2013) 011002.

  25. [25]

      I. Landa-Medrano, I. Lozano, N. Ortiz-Vitoriano, I. Ruiz de Larramendi, T. Rojo, J. Mater. Chem. A 7 (2019) 8746–8764. doi: 10.1039/c8ta12487f

  26. [26]

      Z.F. Chen, X. Lin, H. Xia, et al., J. Mater. Chem. A 7 (2019) 14260–14270. doi: 10.1039/c9ta03133b

  27. [27]

      A.G. Tamirat, X. Guan, J. Liu, J. Luo, Y. Xia, Chem. Soc. Rev. 49 (2020) 7454–7478. doi: 10.1039/d0cs00489h

  28. [28]

      X. Bi, K. Amine, J. Lu, J. Mater. Chem. A 8 (2020) 3563–3573. doi: 10.1039/c9ta12414d

  29. [29]

      X. Liu, X. Song, Z. Guo, et al., Angew. Chem., Int. Ed. 60 (2021) 16360–16365. doi: 10.1002/anie.202104003

  30. [30]

      X. Wu, H. Wu, S. Guan, et al., Nano Res 16 (2023) 9453–9460. doi: 10.1007/s12274-023-5663-5

  31. [31]

      D. Zhou, J. Zhang, T. Bian, et al., Adv. Energy Mater. 14 (2024) 2303192.

  32. [32]

      Y. Liu, D. Lin, Y. Jin, et al., Sci. Adv. 3 (2017) eaao0713.

  33. [33]

      Y. Zhang, W. Luo, C. Wang, et al., Proc. Natl. Acad. Sci 114 (2017) 3584–3589. doi: 10.1073/pnas.1618871114

  34. [34]

      C. Niu, H. Pan, W. Xu, et al., Nat. Nanotechnol. 14 (2019) 594–601. doi: 10.1038/s41565-019-0427-9

  35. [35]

      H. Zhang, X. Liao, Y. Guan, et al., Nat. Commun. 9 (2018) 3729.

  36. [36]

      C.P. Yang, Y.X. Yin, S.F. Zhang, N.W. Li, Y.G. Guo, Nat. Commun. 6 (2015) 8058.

  37. [37]

      R. Zhang, X. Chen, X. Shen, et al., Joule 2 (2018) 764–777.

  38. [38]

      K. Yan, B. Sun, P. Munroe, G. Wang, Energy Storage Mater. 11 (2018) 127–133.

  39. [39]

      S. Choudhury, C.T.C. Wan, W.I. Al Sadat, et al., Sci. Adv. 3 (2017) e1602809.

  40. [40]

      X. Ren, S. Chen, H. Lee, et al., Chem 4 (2018) 1877–1892.

  41. [41]

      S. Choudhury, Z. Tu, S. Stalin, et al., Angew. Chem. Int. Ed. 56 (2017) 13070–13077. doi: 10.1002/anie.201707754

  42. [42]

      S. Chen, J. Zheng, D. Mei, et al., Adv. Mater. 30 (2018) 1706102.

  43. [43]

      S. Choudhury, L.A. Archer, Adv. Electron. Mater. 2 (2016) 1500246.

  44. [44]

      X. Cao, X. Ren, L. Zou, et al., Nat. Energy 4 (2019) 796–805. doi: 10.1038/s41560-019-0464-5

  45. [45]

      C.V. Amanchukwu, X. Kong, J. Qin, Y. Cui, Z. Bao, Adv. Energy Mater. 9 (2019) 1902116.

  46. [46]

      Z. Yu, H. Wang, X. Kong, et al., Nat. Energy 5 (2020) 526–533. doi: 10.1038/s41560-020-0634-5

  47. [47]

      A.C. Kozen, C.F. Lin, A.J. Pearse, et al., ACS Nano 9 (2015) 5884–5892. doi: 10.1021/acsnano.5b02166

  48. [48]

      N.W. Li, Y.X. Yin, C.P. Yang, Y.G. Guo, Adv. Mater. 28 (2016) 1853–1858. doi: 10.1002/adma.201504526

  49. [49]

      Z. Tu, S. Choudhury, M.J. Zachman, et al., Nat. Energy 3 (2018) 310–316. doi: 10.1038/s41560-018-0096-1

  50. [50]

      G. Zheng, S.W. Lee, Z. Liang, et al., Nat. Nanotechnol. 9 (2014) 618–623. doi: 10.1038/nnano.2014.152

  51. [51]

      D. Lin, Y. Liu, W. Chen, et al., Nano Lett. 17 (2017) 3731–3737. doi: 10.1021/acs.nanolett.7b01020

  52. [52]

      G. Zheng, C. Wang, A. Pei, et al., ACS Energy Lett. 1 (2016) 1247–1255. doi: 10.1021/acsenergylett.6b00456

  53. [53]

      Z. Yu, D.G. Mackanic, W. Michaels, et al., Joule 3 (2019) 2761–2776.

  54. [54]

      Y. Gao, Z. Yan, J.L. Gray, et al., Nat. Mater. 18 (2019) 384–389. doi: 10.1038/s41563-019-0305-8

  55. [55]

      Z. Tu, S. Choudhury, M.J. Zachman, et al., Joule 1 (2017) 394–406.

  56. [56]

      S. Choudhury, S. Stalin, D. Vu, et al., Nat. Commun. 10 (2019) 4398.

  57. [57]

      S. Stalin, M. Tikekar, P. Biswal, et al., Nano Lett. 20 (2020) 5749–5758. doi: 10.1021/acs.nanolett.0c01501

  58. [58]

      F. Ding, W. Xu, G.L. Graff, et al., J. Am. Chem. Soc. 135 (2013) 4450–4456. doi: 10.1021/ja312241y

  59. [59]

      F. Ding, W. Xu, X. Chen, et al., J. Phys. Chem. C 118 (2014) 4043–4049. doi: 10.1021/jp4127754

  60. [60]

      H. Dai, K. Xi, X. Liu, C. Lai, S. Zhang, J. Am. Chem. Soc. 140 (2018) 17515–17521. doi: 10.1021/jacs.8b08963

  61. [61]

      D.J. Yoo, K.J. Kim, J.W. Choi, Adv. Energy Mater. 8 (2018) 1702744.

  62. [62]

      P.T. Nguyen, T.N. Nguyen, V.S. Nguyen, et al., RSC Adv. 8 (2018) 13142–13147. doi: 10.1039/c7ra12904a

  63. [63]

      H.H. Lin, J.D. Peng, V. Suryanarayanan, D. Velayutham, K.C. Ho, J. Power Sources 311 (2016) 167–174.

  64. [64]

      M. Xiao, H. Liu, H. Gao, W. Olson, Z. Liang, Appl. Energy 235 (2019) 311–319.

  65. [65]

      F. Marques Mota, J.H. Kang, Y. Jung, et al., Adv. Energy Mater. 10 (2020) 1903486.

  66. [66]

      Z. Liang, Y.C. Lu, J. Am. Chem. Soc. 138 (2016) 7574–7583. doi: 10.1021/jacs.6b01821

  67. [67]

      X. Hu, Y. Gao, B. Zhang, L. Shi, Q. Li, EcoMat 4 (2022) e12264.

  68. [68]

      F. Shi, A. Pei, A. Vailionis, et al., Proc. Natl. Acad. Sci. U. S. A. 114 (2017) 12138–12143. doi: 10.1073/pnas.1708224114

  69. [69]

      F. Chu, J. Liu, Z. Guan, R. Deng, L. Mei, F. Wu, Adv. Mater. 35 (2023) 2305470.

• 

  Figure 1  (a) The molecular structural formula of EMIBr. (b) DFT calculations of binding energy of DMSO and EMI⁺ on Li anode. (c) Schematic illustrations of the different working mechanism in the EMIBr- or KBr-containing Li-CO₂ batteries.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Cycling performances of Li||Li symmetric cells with various electrolytes at current density of 0.1 mA/cm² and limited capacity of 0.05 mAh/cm² and enlarged voltage profiles with periods of 20–50 h and 150–200 h. SEM images (top) together with the corresponding cross-sectional views (bottom) of Li anodes in batteries with (b, c) pure, (d, e) KBr-, and (f, g) EMIBr-containing electrolytes after 10 cycles at 500 mA/g.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Voltage curves and (b) the corresponding overpotentials obtained at the first discharge/charge process for the Li-CO₂ batteries without additive, with KBr or EMIBr. (c) The discharge/charge terminal voltage versus cycle number of the Li-CO₂ batteries without additive, with KBr or EMIBr. Cycling stability of Li-CO₂ batteries (d) without additive, with (e) KBr and (f) EMIBr at a current density of 500 mA/g.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  XPS spectra of C 1s, O 1s and N 1s of the Li anodes disassembled from Li-CO₂ batteries (a-c) without additive and (d-f) with EMIBr, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  SEM micrographs of CNT-based cathodes with different electrolytes after the discharge and charge process: (a) Pristine; (b) with 50 mmol/L EMIBr, discharge; (c) with 50 mmol/L EMIBr, charge; (d) without additive, discharge; (e) without additive, charge. (f) XRD patterns of discharged and charged CNT-based cathodes in EMIBr-containing batteries after 5 cycles at a current density of 500 mA/g.

  下载: 全尺寸图片 幻灯片
•