Interface-Structure-Modulated CuF₂/CF_x Composites for High-Performance Lithium Primary Batteries

1 Introduction

The first prototype of the lithium battery was proposed in the middle of the last century. Since then, various primary batteries that use Li metal anodes in non-aqueous electrolytes, such as Li/(CF)_n, Li/SO₂, Li/FeS₂, Li/MnO_2, and Li/SOCl₂, have been commercialized.^[1] Thus far, lithium primary batteries (LPBs) have been studied extensively for military, aerospace, and medical devices because of their higher energy densities and longer shelf lives than those of commercial rechargeable lithium-ion batteries (LIBs).^[2-4] However, although LPBs with liquid cathodes (such as SOCl₂ and SO₂) are favorable for low-temperature operations, their corrosive and toxic nature and limited maneuverability have restricted them from further applications that require high safety and prolonged execution times.^{[5, 6]} Therefore, in LPBs, solid cathodes in the form of pouch cells have gradually substituted liquid ones because of the feasibility of electrode web coating and improved energy densities. Recently, a number of gaseous cathodes, such as NF₃^[7] and SF₆,^[8-10] have been proposed for high-energy-density LPBs, but their low solubility in conventional organic solvents demanded improvements in technical maturity before they may be used for practical applications. Therefore, LPBs that use solid cathodes have become predominant. Meanwhile, besides the aforementioned systems, Li/S^[11] and organic compounds^[12] have also been investigated for use in LPBs, although their corresponding batteries are accompanied by obstacles in terms of self-discharge and the necessity for additional electrolytes for extra capacity. Among batteries with solid cathodes, the Li/CF_x battery has attracted the most attention from researchers because of its high theoretical energy density, which indicates that it could be an ideal power source for low rates and at moderate temperatures.^{[13, 14]} In the past few years, our group has pursued several endeavors involving this system and demonstrated the manufacture and use of CF_x compounds with energy densities exceeding those of graphite fluorides.^[15-17]

Because of the great successes achieved in the use of CF_x compounds as cathodes of LPBs, the use of other fluorides as electrodes for batteries, to generate high electrochemical energy, is a promising prospect because of the high electronegativity of fluorine and the exceptional free energy for forming fluorides.^[18] Metal fluorides (MeF_x), which exhibit the highly ionic character of metal–fluoride bonds, have been investigated as cathode materials for lithium batteries because of the high theoretical energy densities of their conversion reactions, according to thermodynamic calculations.^[19] Among the various types of MeF_x, CuF₂ exhibits impressive theoretical energy density and is especially advantageous in terms of its theoretical volumetric energy density (7908 Wh L⁻¹). However, during the last century, there have been few developments on the CuF₂ cathode because of its high electrical resistance generated by an intrinsically high band gap and partial dissolution in electrolytes. Badway et al. were the first to realize efficient utilization of CuF₂ by embedding CuF₂ nanodomains in a MoO₃ matrix with the aid of high-energy ball milling. Its good performance was attributed to the decreased electron transport distance of nanosized CuF₂ and the crystallographic transformation of interfacial CuF₂ through fluorine–oxygen exchange.^{[20, 21]} It has also been demonstrated that the lithiation of CuF₂ is a direct conversion reaction and that Cu segregates into dispersed nanoparticles in an LiF matrix after the initial cathodic process.^{[22, 23]} However, the formation of a Cu¹⁺ intermediate during the subsequent anodic process leads inevitably to Cu dissolution and prevents transformation back to CuF₂ from Cu and LiF.^[24] Thus, CuF₂ exhibits deteriorated capacity retention in comparison with those of other MeF_x, which demonstrate reversibility when used as cathodes of LIBs.^[25-27] Although surface coating using NiO^[28] and incorporation with other metals to form ternary fluorides^{[29, 30]} could enhance the reversibility of CuF₂, the lack of long cycling performance, significant capacity loss during the initial cycles, and large voltage hysteresis between charge and discharge processes have prohibited their practical application in rechargeable batteries. Nonetheless, CuF₂ remains an appropriate cathode and, when coupled with Li anode, yields an impressive volumetric energy density among those of LPBs.

On the other hand, hybrid cathodes for LPBs can utilize the advantages of each of its components. Therefore, active materials with superior rate capabilities have been added to cathode formulations to mediate the limitations of CF_x, such as voltage delays observed at low temperatures, severe heat generation at high rates, low tap densities, safety concerns, and high costs.^[31-33] For example, developmental Li/CF_x-MnO₂ D-size cells have already been assembled at a vendor facility (Eagle Picher). The fabrication of a hybrid cathode of CF_x and CuF₂ is expected to combine their respective advantages, and these compounds were chosen because of their high gravimetric and volumetric energy densities, respectively.

In this study, CF_x/CuF₂ hybrid cathodes for LPBs were manufactured, and changes in the valence state of Cu at the interface between these two components were discovered for the first time. The partial charge transfer from CuF₂ to CF_x resulted in unique electrochemical behaviors for the CF_x/CuF₂ hybrid cathodes. Moreover, the influence on the discharge performance of LPBs was systematically evaluated and analyzed through density functional theory (DFT). The charge redistribution in the hybrid cathodes enhanced the electrochemical activity of CuF₂ and facilitated charge transfer in this novel hybrid system, which exhibited its highest volumetric energy density of 4163.40 Wh L⁻¹ when the mass ratio of its components was optimized. The proposed formulation and manufacturing method provide a new avenue for the design of fluoride compounds with tunable electrochemical performance to improve the properties of LPBs.

2 Results and Discussion

The purchased anhydrous CuF₂ powder was characterized by a white color (Figure S1a). Its scanning electron microscopy (SEM) image (Figure S1b) shows an irregular particle morphology in the form of agglomerated particles with a particle size of 300–500 nm. The X-ray diffraction (XRD) pattern of CuF₂ (Figure S2) shows the typical monoclinic structure of CuF₂ (JCPDS card no. 42-1244); the average crystallite size calculated using the Scherrer formula is approximately 31.0 nm. In contrast, the change in color, from black to white powder, of the CF_x product (Figures S3a and S4a) indicates the effective fluorination of hard carbon;^[34] whereas the preservation of particle morphology before and after fluorination, as observed in the SEM images (Figures S3b and S4b) and in the amorphous structure shown by the XRD patterns (Figure S5), is consistent with findings from our previous study.^[16]

The CuF₂/CF_x composites were prepared via planetary ball milling (PBM) (Figure 1a). The prepared composites, with CuF₂-to-CF_x mass ratios of 4:1, 3:2, and 1:1, were labeled CuF₂/CF_x-4.0, CuF₂/CF_x-1.5, and CuF₂/CF_x-1.0, respectively. SEM images of the CuF₂/CF_x composites (Figure S6) reveal drastically reduced particle sizes after PBM, that is, without agglomerated particles and without the original crystal shapes. The corresponding elemental mapping of CuF₂/CF_x-1.0 (Figure S7), selected as representative, confirms the homogeneous distribution of components. It has been demonstrated that during PBM, CF_x acts as a pseudo-lubricant, and thus, ball-milled CuF₂ fine particles tend to disperse uniformly on the surface of CF_x.^[35] As a result, adjusting the ratio of components can change particle size and agglomeration, thus leading to a 3D interconnected network for further reaction.

In an attempt to understand the internal strain and probable structural modulation at the interfaces of the components, morphological and interfacial studies were performed using transmission electron microscopy (TEM) to provide investigative insights into the CuF₂/CF_x composites. The TEM image of CF_x (Figure S8a) shows randomly stacked fluorographene layers, and the corresponding high-resolution TEM (HRTEM) image (Figure S8b) shows an amorphous structure, which is in agreement with its XRD pattern and the ambiguous rings in its selected area electron diffraction (SAED) pattern (Figure S8c). The TEM image of CuF₂ (Figure S9a) reveals the agglomeration of CuF₂ nanoparticles, and the crystallite observed in the HRTEM image (Figure S9b) shows clear and coherent stripes. The interplanar spacing of the lattice fringe was measured to be 2.40 Å, which is indexed to the (111) planes of CuF₂. The SAED pattern in this phase (Figure S9c) shows several rings composed of discrete spots, indicating the polycrystalline nature of the CuF₂ primary particles, which is also illustrated by the distinct domains of different crystalline orientations in the HRTEM image. The TEM images of the CuF₂/CF_x composites (Figure 1b–d I) are similar to each other; the observed large clusters were constructed from several small particles. The corresponding HRTEM images (Figure 1b–d II) reveal that the CuF₂ nanocrystalline regions were encapsulated in amorphous CF_x matrices, which indicate the uniform mixture and close interconnection of CuF₂ and CF_x after PBM. These characteristics are associated with similar grain sizes of approximately several nanometers. However, the interplanar spacings of the lattice fringes in CuF₂/CF_x composites are slightly larger than those in pristine CuF₂, which are also shown in their SAED patterns (Figure 1b–d III). The blurred and weak rings are ascribed to the introduction of the amorphous CF_x component. The d-spacings derived from the SAED patterns are slightly larger than the standard numbers in the JCPDS card for CuF₂, although the P21/n structure was maintained. These results have similarly been observed for CuF₂ nanocomposites that utilize conducting matrices of metal oxides, which induce a distinct crystallographic change to CuF₂ monoclinic crystallites and the formation of corresponding oxyfluorides at the interface of the CuF₂ and oxide materials.^[20] The contrast line profiles across the CuF₂/CF_x composites (Figure S1b–d IV) at the interior and edge locations have also been plotted using line segments. The distinct lattice fringes on the interior clearly became disorderly at the edge, which are obviously different from the observations for the pristine CuF₂ nanoparticles (Figure S9d). This phenomenon indicates that after cooperation with CF_x, the crystalline structure of the CuF₂ surface was severely changed. In addition, with regard to the fast Fourier-transform (FFT) patterns of the different framed areas in the HRTEM image of CuF₂/CF_x-1.0 (Figure S10), the FFT pattern of the interfacial region between CuF₂ and CF_x (labeled II) exhibits an overlap of symmetric lattice and amorphous halos, which is different from the observations for the FFT patterns derived for the pristine CuF₂ (labeled I) and CF_x (labeled III) regions. These results illustrate that through mechanical stresses in PBM, the crystalline structure at the surface of CuF₂ can be modulated using CF_x.

The crystal structures of the CuF₂/CF_x composites were examined via XRD, and the corresponding patterns (Figure 2a) reveal a dominant CuF₂ phase (P21/n) without CF_x diffraction signals because of the amorphous nature of CF_x. The average crystallite sizes of these composites, as calculated using the Scherrer formula and based on the diffraction peak of (011) reflection (2θ = 27.7°), were 28.2, 26.1, and 23.5 nm, respectively. The reduced crystallite sizes, accompanied by broadened Bragg peaks, also indicate nonhomogeneous internal strains induced by PBM.^[26]

X-ray photoelectron spectroscopy (XPS) was employed to determine the elemental compositions, nature of chemical bonds, and interactions between CF_x and CuF₂ in their composites. According to high-resolution F1s spectra (Figure 2b), the characteristic peaks at 689.8, 688.8, and 687.6 eV in the pristine CF_x correspond to perfluorinated –CF₂/CF₃, covalent (CF)_n, and semi-ionic (C_xF)_n, respectively,^{[36, 37]} which can be also observed in the high-resolution C1s spectrum of CF_x (Figure S11). The proportions of different components were summarized in Table S1. For the Fourier-transform infrared spectroscopy (FT-IR) patterns for CF_x (Figure S12), the peak at 1216 cm⁻¹ and the shoulder at 1312 cm⁻¹ are ascribed to the stretching vibrations of covalent C–F bonds and –CF₂ perfluorinated species, respectively. Meanwhile, there is only one dominant peak at 684.4 eV in the F1s spectrum of the purchased anhydrous CuF₂, and this peak is assigned to the Cu–F bond. Its FT-IR spectrum also reveals the stretching vibration of Cu–F bonds at 486 cm⁻¹.^{[38, 39]} According to the XPS F1s spectra of the CuF₂/CF_x composites, the proportion of perfluorinated –CF₂/CF₃ decreased as the CuF₂ content increased, whereas by contrast, the proportion of semi-ionic (C_xF)_n increased. Furthermore, the peak ascribed to Cu–F bonds in the CuF₂/CF_x composites gradually shifted to positions with higher bonding energies as the CF_x content was increased, which probably indicates charge redistribution at the interface of CF_x and CuF₂.^[40] Therefore, the corresponding high-resolution Cu 2P_3/2 spectra of the CuF₂/CF_x composites (Figure 2c) were subsequently obtained. For the purchased anhydrous CuF₂, three peaks can be observed at 935.9, 942.2, and 944.2 eV, which are attributed to the oxidation state of Cu²⁺ and two strong satellite peaks, respectively (Table S2). It was determined that after CuF₂ was incorporated with CF_x, the peak position of Cu 2P_3/2 was overall shifted by approximately 0.8 eV to a higher bonding energy. Simultaneously, the characteristic peak of Cu³⁺ at 938.9 eV^{[41, 42]} appears for all CuF₂/CF_x composites, and its fraction increased along with the proportion of CF_x, which implies a strong interaction between these two components.

To verify the occurrence of charge transfer on the surfaces of CuF₂ and CF_x, a model of CuF₂, exposing the (011) crystal face, and two fluorographene layers, representing CF_x, was constructed based on 3 × 3 × 1 Monkhorst–Pack k-points for a periodic interface model. The band structures and densities of states (DOS) of CuF₂ and monolayer fluorographene were calculated via DFT. It was determined that the electron density of Cu in CuF₂ decreased while the electron density around the F atoms increased (Figure S13). Similarly, the electron density around C in CF_x decreased while the electron density around the F atoms increased (Figure S14). This was mainly due to the electronegativities of Cu, F, and C, which are 1.9, 4.0, and 2.5, respectively; the electrons were transferred to atoms with higher electronegativities. The increased electron density between C and C was mainly due to the common electron pair in C–C. The differential charge density of CuF₂/CF_x (Figure S15) shows that the electron density of Cu in CuF₂ in the composite was lower than that in pure CuF₂ because of the transfer of Cu atom electrons on the contact surface to nearby F atoms, which caused the valence of Cu to increase, in accordance with the XPS measurements. The electron density of F atoms in CF_x in the composite was also lower than that in pure CF_x because of the transfer of electrons from the F atoms to the C atoms. The DOS diagram of CuF₂/CF_x (Figure S16a) can also show the electron transfer at their contact interface due to the greater number of electrons at the Fermi level, resulting in metallic properties and facilitating the transmission of electrons in the interface structures. However, the DOS diagrams of CF_x and CuF₂ (Figure S16b,c) reveal notable band gaps, which are unfavorable to the transmission of electrons.^[43]

The electrochemical performances of CuF₂, CF_x, and CuF₂/CF_x composites in coin cells with lithium metal anode and 1 M LiClO₄ in ethylene carbonate (EC) / dimethyl carbonate (DMC) electrolyte were tested using galvanostatic discharge at different current densities. At a current density of 10 mA g⁻¹, the discharge profile of the CuF₂ cathode (Figure S17) prepared via PBM with conductive carbon exhibited a short plateau with potential at approximately 2.83 V, corresponding to a capacity of approximately 80 mAh g⁻¹, and then a plateau at approximately 2.65 V, due to the continuous reduction in Cu²⁺, until a cutoff potential with a specific capacity of 538.15 mAh g⁻¹, which is similar to the findings of previous studies.^{[20, 24]} At the same current density (Figure S18), the CF_x cathode exhibited the typical two-phase nature of the discharge reaction, with a discharge capacity of 749.79 mAh g⁻¹ and stable plateaus of 3.10 V, which is in agreement with the findings of our previous studies.^{[15, 16]} However, three distinct discharge plateaus can be observed from the galvanostatic discharge profiles of the CuF₂/CF_x hybrid cathodes (Figure 3a–c), which are totally different from the discharge profiles of pristine CF_x and CuF₂ cathodes. At a current of 10 mA g⁻¹, three consecutive discharge plateaus at approximately 3.1, 2.9, and 2.7 V occurred for all CuF₂/CF_x hybrid cathodes. The specific capacities of CuF₂/CF_x-4.0, CuF₂/CF_x-1.5, and CuF₂/CF_x-1.0 cathodes were 565.61, 628.89, and 670.45 mAh g⁻¹, respectively, which were proportional to the amount of CF_x. Furthermore, the discharge capacities of the CuF₂/CF_x cathodes delivered at the first and third plateaus correspond to the amounts of CF_x and CuF₂ in the hybrid cathodes, respectively. The increase in discharge current density led to a clear decrease in energy density due to decreases in the output potential and discharge capacity. Because of the inherently insulating property of CuF₂ and CF_x, their cathodes suffer significant polarizations at high current densities. Nevertheless, the rate capabilities of CuF₂/CF_x hybrid cathodes at high current densities were improved remarkably, as indicated by the discharge plateau above 2.5 V at current densities as high as 5 A g⁻¹. According to the Ragone plots of the CuF₂, CF_x, and CuF₂/CF_x hybrid cathodes (Figure 3d) calculated from their discharge profiles at different current densities, the CuF₂/CF_x-1.0 cathode exhibited the best rate capability, delivering a maximum power density of 11.16 kW kg⁻¹, associated with a maximum energy density of 1818.08 Wh kg⁻¹. The improved rate capabilities of CuF₂/CF_x cathodes can also be verified using the impedance spectra of pristine cathodes (Figure 3e). The symbols R_s, R_ct, Z_w, and CPE represent the ohmic resistance, the charge-transfer resistance, Warburg impedance, and the double layer capacitance, respectively. Semicircles in Nyquist plots represent charge-transfer resistance (R_ct), and the CuF₂/CF_x hybrid cathodes resulted in smaller semicircles than those for pristine CuF₂ and CF_x cathodes, which confirm the improved reaction kinetics in these hybrid cathodes. In addition, the value of R_ct gradually decreases as the amount of CF_x in the composite increases, indicating the modified electrochemical activity due to their interfacial interaction.^[44] Moreover, the diffusion coefficient of lithium ions (D_Li⁺) of different cathodes was measured by galvanostatic intermittent titration technique (GITT) (Figure S19). The D_Li⁺ values of CuF₂/CF_x composites are slightly higher than the D_Li⁺ values of CuF₂ and CF_x, especially at high potential region, which is caused by the changed valence state of Cu and the optimized electrical structure at their interface. The much higher D_Li⁺ values of CuF₂/CF_x composites further demonstrate the improved rate capability. From the discharge profiles of hybrid cathodes at different current densities, these three distinct discharge plateaus merges into one with the increase of current densities, which should be associated with the more sluggish Li⁺ transport kinetics at the lower plateaus than the upper one revealed by the GITT test,^[45] and the similar behavior has also been observed in CF_x and sulfur hybrid cathode of LPBs.^[46] The theoretical calculation also confirms the improved rate capabilities of CuF₂/CF_x hybrid cathodes because their vanished band gaps guarantee excellent electronic conductance.^{[47, 48]}

Although the gravimetric specific capacities of the CuF₂/CF_x hybrid cathodes are lower than that of pristine CF_x cathode, CuF₂ exhibits superiority in terms of its volumetric specific capacity because of its high mass density. Therefore, according to the different electrode loading masses and their thicknesses, as measured using cross-sectional SEM images (Figure S20), CuF₂/CF_x cathodes have an obvious advantage over pristine CF_x and CuF₂ cathodes in terms of volumetric energy density (Figure 3f). As shown in Ragone plot of volumetric energy density vs power density (Figure S21), among these hybrid cathodes, CuF₂/CF_x-1.0 exhibits an exceptional volumetric energy density of 4163.40 Wh L⁻¹ and a maximum power density of 25.56 kW L⁻¹. The electrochemical performances of CuF₂/CF_x-1.0 were also superior to most reported high volumetric energy density cathodes (Table S3). Therefore, CuF₂/CF_x cathodes are highly suitable for applications with high demands for volumetric energy density.

To investigate the origin of the additional discharge plateau and the reaction mechanisms of the CuF₂/CF_x cathodes during the discharge process, the CuF₂/CF_x-1.0 cathode, which demonstrated the highest volumetric energy density, was selected for further analysis. Specifically, the component evolutions at different charge cutoff voltages, for discharge at 10 mA g⁻¹ (Figure 4a), were analyzed via ex situ XPS. The resulting high-resolution Cu 2p_3/2 spectra of the CuF₂/CF_x-1.0 cathode discharged at different voltages (Figure 4b) illustrate the consecutive reduction in Cu cations to metallic Cu during the whole discharge process, in particular, in the emergence of a signal peak at 293.2 eV for Cu⁰ (Table S4). However, according to the Cu 2p_3/2 spectrum for the discharge to 2.9 V, the intensity of the peak at 937.2 eV, attributed to Cu³⁺, drastically reduced, associated with a slight shrinkage in peak width at 234.8 eV for Cu²⁺. The Cu 2p_3/2 spectrum for the discharge to 2.7 V exhibits significant decreased peak intensity for Cu²⁺, accompanied by the complete disappearance of the Cu³⁺ signal. The changes in these peaks during the discharge process illustrate the successive reduction in Cu³⁺ and Cu²⁺ at the first and second discharge plateaus based on the valence states. At the end of the discharge process, the remaining Cu²⁺ ions in the hybrid cathode were reduced to Cu⁰. On the other hand, to investigate the change in the CF_x component during the discharge process, the corresponding high-resolution F1s spectra of the CuF₂/CF_x-1.0 hybrid cathode discharged at different voltages were also obtained (Figure 4c). In addition to the typical peaks at 689.8, 688.8, and 687.6 eV for perfluorinated species, covalent, and semi-ionic C–F bonds of CF_x, respectively, the peak at 685.9 eV is ascribed to the Cu–F bond of CuF₂, associated with a weak peak at 690.2 eV corresponding to adhesive poly(vinylidene fluoride) (PVDF) in the cathode (Table S5). The F1s spectrum of the CuF₂/CF_x-1.0 hybrid cathode discharged to 2.9 V shows the consumption of the semi-ionic C–F bond and the partial of covalent C–F bond during the first discharge plateau. Meanwhile, the relatively increased peak intensity at 685.9 eV, compared with that of the F1s spectrum of pristine composite cathode, is ascribed to LiF crystals formed based on the electrode reaction, which overlapped with the Cu–F bond. The gradually decreased peak intensity of the covalent C–F bond in the hybrid cathode indicates that this conversion reaction occurred in the whole reaction process, and the sole remaining peak at 685.9 eV in the F 1s spectrum of the hybrid cathode discharged to 1.5 V further confirms the full conversion reaction of active materials. As a result, it can be concluded that the sequence of cathode reactions starts with the reduction in Cu³⁺, formed at the interface of CuF₂ and CF_x, accompanied by the conversion of the semi-ionic C–F bond, consumption of the covalent C–F bond, and subsequent reduction in Cu²⁺, although the latter two reactions are involved throughout the entire discharge process.

To further investigate the reaction mechanism of the hybrid cathode, the lowest-energy ordering sequence for the Li_yCuF₂/CF_x model was obtained (Figure S22) to reflect the sequence of reactions between lithium ions and the CuF₂/CF_x composite. The voltage curve for the conversion reaction involving Li can be calculated based on the lowest-energy ordering sequence obtained for Li_yCuF₂/CF_x. In the calculated voltage curve (Figure 5a), there are three voltage stages,^[49] corresponding to reactions between the intercalation of lithium ions and the composite structure (Figure 5b). The first stage occurs when y (Li_yCuF₂/CF_x) is ~1, and lithium ions react with F atoms in the CuF₂/CF_x interfaces and CF_x. The structure then proceeds with a two-phase reaction. In other words, the second stage may mainly correspond to lithium ions reacting with F atoms in CuF₂ in the bulk phase to generate LiF. The third stage corresponds to further reactions with CuF₂ and CF_x in the bulk phase. Meanwhile, the calculated voltage curve is in sufficient agreement with the discharge profile measured in the experiment. In addition, the parts of the CuF₂/CF_x structure corresponding to the three discharge plateaus are explained theoretically.

Based on the calculation results, it is concluded that the fluorine atom of CF_x in CuF₂/CF_x attracts electrons around the Cu atom in CuF₂, resulting in a decrease in the electron density around the Cu atom, which increases the valence of the Cu atom and the electronic conductivity. Furthermore, through XPS and HRTEM analyses, the presence of Cu³⁺ was characterized, and the lattice structure of CuF₂ was modulated. Based on these results, combined with the analyses of discharge cutoff voltage-dependent XPS spectra and theoretical voltage curves, the three discharge plateaus in the discharge curves were inferred to correspond to CF_x moieties that are prone to take up lithium through a two-phase reaction, including F atoms in the CuF₂/CF_x interfaces and CF_x. At that point, lithium reacts further with F atoms on Cu³⁺, which exist in the interface of CuF₂/CF_x, after which the bulk of CuF₂ reacts with lithium. The LiF produced on the surface of the cathode material alters and hinders the rate of lithium diffusion to the reaction sites of CuF₂/CF_x, which leads to the plateaus gradually transitioning to sloping discharging profiles until the end of the reaction. Future developments of fluoride cathodes should focus particularly on the exploration of changes in the fluoride valence state and electron/ion conductive pathways both in crystal and amorphous structures, which are expected to further improve the discharge reaction potential and energy conversion efficiency.

3 Conclusion

A series of CuF₂/CF_x composites were prepared via PBM with the objective of enhancing the volumetric energy densities of LPBs. The maximum volumetric energy density, 4163.40 Wh L⁻¹, was achieved through the optimization of component proportions in the hybrid cathode. More importantly, the occurrence of Cu³⁺ and the modulated electronic structure at the interface of CF_x and CuF₂, due to charge transfer on their contact surface caused by the mechanical stresses in PBM, result in unique discharge profiles with three distinct stages and superior rate capabilities in the hybrid cathodes, with a maximum power density of 11.16 kW kg⁻¹. The XPS spectra and DFT calculations for the CuF₂/CF_x hybrid cathode at different depths of discharge reveal the high electrochemical activity of Cu³⁺ at the interface and the stepwise reaction mechanism of the hybrid cathode. This novel hybrid system provides an insightful and useful strategy for optimizing the electrochemical performance of hybrid fluoride cathodes through interfacial modification, for applications that require high volumetric energies and power densities.

4 Experimental Section

Conflict of Interest

The authors declare no conflict of interest.