-
Lithium metal batteries (LMBs) have attracted wide attention due to the high theoretical capacity (~3860 mAh·g−1) and low redox potential (−3.04 V vs. NHE) of the lithium metal anode [1−3]. However, notorious growth of lithium dendritic induced short-circuit triggers thermal runaway, posing serious safety hazards [4−6]. Compared to liquid electrolytes, solid polymer electrolytes (SPEs) have higher mechanical strength and thermal stability, leading to better compatibility with lithium metal anode [7, 8]. In addition, excellent plasticity and flexibility of SPEs allow for intimate physical contacts with both the lithium metal anode and the composite cathode, achieving stable cycling without the need for unpractical high stack pressure [9, 10]. However, the SPE is a dual-ion conductor, including free migrating Li+ and anions, and its Li+ transference number (
$ t_{\rm{Li^+}} $ ), i.e., the ratio of Li+ current to the total current, is usually low [11, 12]. The accumulated anions at the SPE|electrode interface cause concentrated polarization and reduce electrochemical reaction rate, thereby lowering capacity of the cell, especially at a high current density [13, 14].Metal fluorides can trap anions through the coordination of metal cations and anions, increasing
$ t_{\rm{Li^+}} $ without significantly changing the SPE network [15]. It is widely reported that the incorporation of CaF2, MgF2, CuF2, FeF3, AlF3, TiF4, etc. in SPEs has led to an increase in$ t_{\rm{Li^+}} $ [16−21]. Generally, the higher the charge quantity of the metal ion, the more significant the trapping of anions, and the more significant the increase in$ t_{\rm{Li^+}} $ . However, the addition of CuF2 and FeF3 resulted in an increase in$ t_{\rm{Li^+}} $ from 0.15 to 0.42 and 0.16 to 0.45, respectively, indicating that the difference is not significant as expected with the increase in the charge quantity of the metal ions. Differently, the incorporation of AlF3, which has the same metal ion charge as FeF3 but a smaller ionic radius, led to a substantial increase in$ t_{\rm{Li^+}} $ from 0.15 to 0.67. It is demonstrated that the effect of metal fluorides anion trap agents on the increase in$ t_{\rm{Li^+}} $ needs to take both the ion charge and ionic radius into consideration.In this work, we utilize the ionic potential (Z/r, the ratio of charge quantity to ionic radius) to generalize the anion trapping capability of metal fluorides. An SPE is prepared via ultraviolet (UV) light-initiated copolymerization of vinyl ethylene carbonate (VEC), 2,2,3,3,4,4,5,5-octafluoro-1,6-hexanediol diacrylate (OFHDODA) monomers, incorporating the fluorides anion trap agents of MnF2, FeF3, or NbF5 with metal ionic potential of 2.99 Å−1, 4.65 Å−1, and 7.81 Å−1, respectively. The resulted SPE exhibits
$ t_{\rm{Li^+}} $ increased from 0.35 to 0.52 for MnF2, 0.61 for FeF3, and 0.67 for NbF5 at 25 °C. As a result, the assembled LiCoO2 (LCO) ||Li batteries with a cutoff voltage of 4.45 V at 25 °C based on NbF5-SPE deliver discharge capacity of 153.81 mAh·g−1 at 5 C and capacity retention of 86% for 5 C/0.2 C, significantly higher than 146.59 mAh·g−1, 81.5% based on MnF2, and 143.95 mAh·g−1, 83.6% based on FeF3. The fluoride anion trap agent with high metal ionic potential is beneficial for significantly increasing$ t_{\rm{Li^+}} $ of SPE. In addition, Li3N/LiF-rich solid electrolyte interphase (SEI) facilitates rapid Li+ transport, which further enhances high-rate performance of LMBs. -
Lithium bis(trifluoromethane) sulfonamide (LiTFSI) was purchased from Dodo Chemical Co., Ltd. VEC (99%) was purchased from Sigma-Aldrich. Manganese(II) fluoride (MnF2, 98%), Iron(III) fluoride (FeF3, 97%), Niobium(V) fluoride (NbF5, 98%), and OFHDODA (99.0%) were purchased from Adamas-beta. Phenylbis(2,4,6-trimethylbenzoyl) phosphine oxide (IRGACURE819, 99%) was purchased from Badische Anilin-und-Soda-Fabrik (BASF).
-
First, VEC and OFHDODA were mixed at a molar ratio of 20:1 in a reaction vessel. Subsequently, LiTFSI and metal fluorides (MnF2/FeF3/NbF5) were sequentially added to the mixture. The solution was stirred for 30 min at 25 °C, with the molar ratio of LiTFSI to polymer monomers maintained at 1:10. Next, IRGACURE 819 was added at a molar ratio of 0.5% relative to the monomers and stirred for 10 min. The resultant solution was blade-cast onto a substrate, followed by a photocuring process via 30 W of 365-nm UV light. To ensure an inert atmosphere, all procedures were conducted within an argon-filled glove box.
-
FTIR spectra were collected using a Thermo Scientific Nicolet 6700 spectrometer. Raman spectra were measured using a Renishaw inVia Qontor confocal Raman microscope. SEM images were obtained by a FEI Quanta 400 FEG. XPS spectra were measured by a PHI 5000 Versaprobe Ⅱ in a Vacuum Interconnected Nanotech Workstation (Nano-X, SINANO).
-
Ionic conductivity was obtained using stainless steel||stainless steel symmetric cells via electrochemical impedance spectroscopy (EIS) in a frequency range from 7 MHz to 100 MHz with an applied amplitude of 10 mV. The
$ t_{\rm{Li^+}} $ was obtained from chronoamperometry (CA) conducted on Li||Li symmetrical cells under a constant 10 mV polarization until achieving steady state current. The electrochemical stability window (ESW) extracted by linear sweep voltammetry (LSV) was measured on an asymmetrical (Li|SPE|C) cell at a scan rate of 0.1 mV·s−1 from open circuit voltage (OCV) to 6 V. Also, EIS measurements were taken before and after polarization. The electrochemical performance of LCO||Li cells was evaluated using CR2032-type coin cells. For LCO cathode fabrication, a slurry comprising LCO powder, acetylene black, and polyvinylidene fluoride (PVDF) at a mass ratio of 8:1:1 was dispersed in N-methyl-pyrrolidone (NMP) and blade-coated onto carbon-coated aluminum foil with a mass loading of about 1.0 mg·cm−2. Electrochemical performance of LCO||Li cells were carried out on Neware BTS battery test system in a potential range of 2.8–4.45 V at 30 °C. -
A solid polymer electrolyte (SPE) film with a thickness of approximately 200 μm was prepared by blending NbF5, LiTFSI, OFHDODA, and VEC. The mixture was blade-cast onto a substrate and subsequently subjected to 365 nm-UV light curing as illustrated in FIG. 1(a). The mixture then turned from liquid to solid (FIG. S1 in Supplementary materials (SM)). As reported in our previous work, OFHDODA with enhanced oxidation resistance facilitates the transmission of electron-withdrawing effects via its polyfluorinated segments, thereby improving electrochemical stability [22]. In contrast to poly(ethylene oxide) (PEO)-based electrolytes with low ionic conductivity at room temperature (RT), poly(vinylene carbonate) (PVEC) with high ionic conductivity at RT was used here due to facilitated Li+ conduction by coordinating with C=O and O–C–O groups at multiple sites [23]. Fluorine ion was utilized to construct the LiF-rich interphase during charge‒discharge cycles, which is an electronic insulator and ionic conductor, benefiting improved cycling stability of the cell [18]. Fourier-transform infrared (FTIR) spectroscopy analysis revealed that the pre-polymerized VEC-OFHDODA-NbF5 exhibited a characteristic peak at 1640 cm−1, corresponding to the stretching vibration of C=C bonds in the monomers (FIG. 1(b)). In contrast, this peak was invisible in the spectra of the P(VEC-OFHDODA)-NbF5, indicating that the VEC monomer and the OFHDODA monomer were copolymerized via the C=C bond.
-
The changes in ionic transport property of SPEs by incorporating metal fluorides with different metal ionic potential, such as MnF2 (2.99 Å−1), FeF3 (4.65 Å−1), and NbF5 (7.81 Å−1), were investigated. Here, the molar ratio of NbF5 to LiTFSI was set to an optimal 1:15 (FIG. S2 in SM). The Li||Li symmetric cells were assembled using P(VEC-OFHDODA) with and without fluorides at 25 °C. FIG. 2(a) shows that the
$ t_{\rm{Li^+}} $ of P(VEC-OFHDODA) is 0.35, whereas those modified with MnF2, FeF3 and NbF5 exhibit significantly enhanced values of 0.52, 0.61 and 0.67, respectively (testing details in FIG. S3 (SM)). FIG. 2(b) shows ionic conductivity σ of P(VEC-OFHDODA) is 0.66 mS·cm−1. After incorporating MnF2, FeF3 and NbF5, σ is 0.62 mS·cm−1, 0.64 mS·cm−1, and 0.68 mS·cm−1, respectively. Moreover, ESW of P(VEC-OFHDODA) before and after incorporating MnF2, FeF3, and NbF5 is 4.69, 4.71, 4.75, and 4.76 V (vs. Li+/Li), respectively (FIG. S4 in SM). These results demonstrate that$ t_{\rm{Li^+}} $ of the SPE is sensitive to the metal ionic potential, but ionic conductivity and ESW of the SPE exhibit minor difference.To understand the effect of metal ionic potential on
$ t_{\rm{Li^+}} $ , Raman spectroscopy was employed to elucidate the bonding states of TFSI− [24]. The proportion of free TFSI− (725 cm−1) decreased from 57.03% to 42.48% for MnF2, 37.10% for FeF3 and 35.67% for NbF5. Differently, the combined contribution of contact ion pairs (CIPs, i.e. TFSI− interacting with Li+, 742 cm−1) and aggregate ion pairs (AGGs, i.e. TFSI− interacting with more Li+, 747 cm−1) increased from 42.97% to 57.52% for MnF2, 62.90% for FeF3 and 64.34% for NbF5 (FIG. 2(c–f)). Meanwhile, the peak position of the stretching vibration of O=S=O red-shifted from 1133 cm−1 to 1132 cm−1 for MnF2, 1129 cm−1 for FeF3 and 1128 cm−1 for NbF5, respectively (FIG. S5 in SM). These results confirm that a higher metal ion potential strengthens coordination between the metal cation and the oxygen atom in TFSI−, thereby more significantly suppressing anion migration. -
The interfacial property between the SPE and the lithium metal anode is critical for achieving high-rate performance. Li||Li symmetric cells were assembled using P(VEC-OFHDODA) without and with the addition of MnF2, FeF3, NbF5. As shown in FIG. 3(a), cells based on P(VEC-OFHDODA)-NbF5 and P(VEC-OFHDODA)-FeF3 exhibit flat polarization profiles with average constant polarization voltages of 33 mV and 34 mV for 600 h, respectively. In contrast, cells based on P(VEC-OFHDODA) and P(VEC-OFHDODA)-MnF2 soft-short after 100 h and show significant variations of polarization voltage after 300 h, respectively. In addition, cells based on P(VEC-OFHDODA)-NbF5 and P(VEC-OFHDODA)-FeF3 were further evaluated at higher current densities (FIG. 3(b)). The P(VEC-OFHDODA)-NbF5 cell exhibited stable and low average constant polarization voltages of 31 mV, 70 mV, and 121 mV under 0.1, 0.25, and 0.5 mA·cm−2, respectively. In contrast, the P(VEC-OFHDODA)-FeF3 cell displayed unstable and larger polarization over 0.25 mA·cm−2.
To understand the superior polarization of Li||Li symmetric cells based on NbF5, scanning electron microscopy (SEM) was conducted to investigate lithium deposition morphology on Li before and after the P(VEC-OFHDODA)-NbF5 cell cycling (FIG. 3(c)). It is shown that lithium deposition is uniform and growth of lithium dendrites is not significant. X-ray photoelectron spectroscopy (XPS) was employed to further investigate the evolution of the SEI on Li anode. As shown in FIG. 3(d), the F 1s spectra exhibit no obvious fluorinated species before cycling, but a significant Li-F signal at 685.5 eV is seen after cycling. Meanwhile, the N 1s spectra also show no obvious nitrides before cycling, but two significant peaks at 398.4 eV (Li3N) and 401.1 eV (Li2N-SO2−) are visible after cycling (FIG. 3(e)).
These results demonstrated that enhanced lithium compatibility based on NbF5-SPE stems from formation of Li3N and LiF in SEI, which mainly originates from decomposition of TFSI−. The Li3N/LiF-rich SEI facilitates rapid Li+ transport while suppressing SPE degradation by blocking electrochemical reaction. Even under a high current density of 0.5 mA·cm−2, the NbF5-SPE maintains outstanding interfacial compatibility, enabling the course of uniform lithium plating and stripping.
-
The LCO||Li cells based on P(VEC-OFHDODA) without and with MnF2, FeF3 and NbF5 were assembled to evaluate rate performance (1 C=185 mAh·g−1). FIG. 4(a–d) show galvanostatic charge‒discharge curves of LCO|SPE|Li cells in a potential range of 2.8–4.45 V at 30 °C. As the current density increased from 0.2 C to 5 C, the LCO||Li cell based on P(VEC-OFHDODA) before and after incorporating MnF2, FeF3 and NbF5 exhibited the change of charging voltage platform successively decreased. The specific capacity obtained from galvanostatic charge‒discharge curves exhibit a higher discharge capacity of 179.04 mAh·g−1 at 0.2 C and ~153.81 mAh·g−1 at 5 C for P(VEC-OFHDODA)-NbF5 cell compared with those for other cells (FIG. 4(e)). The capacity retention of 5 C/0.2 C is 86% for P(VEC-OFHDODA)-NbF5, which is also higher than that of 61.5% for P(VEC-OFHDODA) cell, 81.5% for P(VEC-OFHDODA)-MnF2 cell and 83.6% for P(VEC-OFHDODA)-FeF3 cell. Cycling performance of P(VEC-OFHDODA)-NbF5 cell at 2 C was further evaluated as shown in FIG. 4(f), which shows an initial discharge capacity of 167.05 mAh·g−1 as well as a stable cycling with a discharge capacity of 138.68 mAh·g−1 after 400 cycles, corresponding to a capacity retention rate of ~ 83%. FIG. S6 in SM shows the corresponding differential capacity (dQ/dV) curves of the P(VEC-OFHDODA)-NbF5 cell, which exhibits invisible side reaction peaks and no obvious shifts of LCO redox peaks at the charging voltage of 2.8–4.45 V after the 1st, 200th, and 400th cycles.
It is demonstrated that high ionic potential inhibits TFSI− migration and suppresses concentrated polarization at the SPE|electrode interface, which maintains a high electrochemical reaction rate, resulting in superior discharge capacity of the cell even at a high current density. Also, the mechano-electrical coupling is alleviated and the interfacial side reaction is self-terminated, allowing for stable long-cycling [25].
-
In summary, we establish a descriptor for anion trapping capability of metal fluorides via the metal ionic potential (Z/r, the ratio of the charge quantity to the ionic radius). By incorporating the fluorides anion trap agents of MnF2, FeF3, or NbF5 with metal ionic potentials of 2.99 Å−1, 4.65 Å−1, and 7.81 Å−1, respectively.
$ t_{\rm{Li^+}} $ of the SPE composed of VEC, OFHDODA monomers increased from 0.35 to 0.52 for MnF2, 0.61 for FeF3, and 0.67 for NbF5 at 25 °C. Correspondingly, the assembled LCO||Li cells with a cutoff voltage of 4.45 V at 25 °C based on NbF5-SPE deliver a discharge capacity of 153.81 mAh·g−1 at 5 C and a capacity retention of 86% for 5 C/0.2 C, significantly higher than 146.59 mAh·g−1, 81.5% based on MnF2, and 143.95 mAh·g−1, 83.6% based on FeF3. Meanwhile, the assembled 4.45 V-LCO|NbF5-SPE|Li deliver stable cycling performance with a high capacity retention of ~83% after 400 cycles at 2 C. On one hand, the high metal ionic potential enhances coordination between metal fluorides and anions, which suppresses free migration of anions, thereby increasing$ t_{\rm{Li^+}} $ and mitigating concentrated polarization, especially at a high current density. On the other hand, the high metal ionic potential induced high$ t_{\rm{Li^+}} $ facilitates uniform Li deposition on the Li anode and suppresses dendritic growth, leading to stable cycling and improved capacity retention.Supplementary materials: FIG. S1 exhibits photo- graphs of SPE-NbF5 before and after polymerization; FIG. S2 exhibits
$ t_{\rm{Li^+}} $ of SPE with different NbF5 molar ratios; FIG. S3 exhibits CA and EIS curves of Li||Li cells based on SPE with and without metal fluorides; FIG. S4 exhibits LSV curves of Li||C cells based on SPE with and without metal fluorides; FIG. S5 exhibits FTIR spectra of SPE with and without metal fluorides; FIG. S6 exhibits corresponding dQ/dV curves of the LCO|SPE-NbF5|Li cell.
Tuning Ionic Potential of Metal Fluorides in Solid Polymer Electrolytes for High-Rate Lithium Batteries
- Received Date: 26/05/2025
- Available Online: 27/08/2026
-
Key words:
- High-rate /
- Lithium metal /
- LiCoO2 /
- Solid-state electrolytes /
- NbF5 /
- Coordination
Abstract: Incorporating metal fluorides is an effective strategy for increasing Li+ transference number (
首页
登录
注册


DownLoad: