Introduction

Emerging concerns regarding energy and the environment have prompted the development of large-scale energy storage devices to better use renewable energy sources such as solar and wind1. Redox flow batteries (RFBs) have emerged as a promising technology for large-scale grid energy storage, offering scalability in terms of independent power and energy density2,3,4,5. RFBs store energy in redox-active species that are strategically positioned in the anolyte and catholyte tanks. These batteries are broadly categorized into aqueous RFBs (ARFBs) and nonaqueous RFBs (NRFBs), depending on the electrolyte solvent employed6,7,8,9,10. ARFBs have the advantages of high safety and low cost; however, they generally suffer from a restricted electrochemical voltage window (< 1.6 V), which results in a relatively modest energy density of 20–50 Wh/L3. In contrast, NRFBs can present a broader electrochemical window of up to 5.5 V11, enabling the use of high-potential redox-active materials, thereby enhancing their energy density12,13. However, several factors impede the development of NRFBs, including the limited availability of ion-exchange membranes, low chemical compatibility with nonaqueous solvents, and high cost14,15. Membrane-free models have been successfully demonstrated in aqueous/aqueous and aqueous/nonaqueous liquid/liquid biphasic systems16,17,18,19,20,21,22. However, reports on membrane-free batteries based on nonaqueous/nonaqueous biphasic systems are limited, primarily because of the limited availability of immiscible nonaqueous electrolytes11,23. The development of cost-effective nonaqueous membrane-free systems remains a challenge. The development of liquid/solid or liquid/gel biphasic systems is a potential alternative to liquid/liquid biphasic strategies. Among the various types of electrolytes, polymer-based electrolytes are considered ideal candidates for developing the solid or gel phases of biphasic systems24,25,26,27. Polymer-based electrolytes offer numerous advantages over conventional solvents, including tunable ionic conductivity, a wide electrochemical window, nonflammability, and cost-effectiveness. Over the past few years, polymer-based electrolytes, available in both solid and gel forms, have been extensively explored across various battery chemistries28,29,30,31,32. Although numerous classes of polymers have been explored, including poly(ethylene oxide), poly(vinylidene fluoride), poly(vinylidene fluoride-co-hexafluoropropylene), and poly(methyl methacrylate)28,29,30,33,34, the application of polymer electrolytes has primarily focused on solid-state batteries.

In this study, we expanded the application of polymer-based solid and gel negolytes to the construction of liquid/solid and liquid/gel biphasic membrane-free batteries. Two liquid/solid biphasic batteries were fabricated; in the first, a 2,4,6-tri-(1-cyclohexyloxy-4-imino-2,2,6,6-tetramethylpiperidine)−1,3,5-triazine (Tri-TEMPO)/fluoroethylene carbonate (FEC) catholyte was combined with a poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF) solid polymer/Li anolyte (denoted as PVDF-Li); in the second, a Tri-TEMPO/tetra (ethylene glycol) dimethyl ether (TEGDME) catholyte was paired with a polypropylene carbonate (PPC) gel polymer/Li anolyte (denoted as PPC-Li). The PVDF-Li anolyte paired with the Tri-TEMPO/FEC catholyte (PVDF-Li | |Tri-TEMPO) produced a cell voltage of 3.43 V, whereas the pairing of the PPC-Li anolyte with the Tri-TEMPO/TEGDME catholyte (PPC-Li | |Tri-TEMPO) yielded a cell voltage of 3.45 V. Both PVDF-Li | |Tri-TEMPO and PPC-Li | |Tri-TEMPO biphasic battery configurations were evaluated for their charge/discharge performance for different concentrations of Tri-TEMPO (0.2 and 0.5 M). The PVDF-Li | |Tri-TEMPO battery exhibited capacity retentions of 90.7% and 81.78% and Coulombic efficiencies (CEs) of 95.4% and 96.7% under static and flow conditions, respectively. In contrast, the PPC-Li | |Tri-TEMPO battery exhibited capacity retentions of 96.8% and 78.8% and CEs of 97.8% and 98.4% under similar conditions. These findings demonstrate the feasibility and effectiveness of the proposed strategy for enhancing battery performance.

Results and discussion

Preparation of polymer electrolytes and biphasic system

Two widely explored polymers, PVDF and PPC, which exhibit wide voltage windows, good thermal stability, and high mechanical strength, were selected to prepare solid- and gel-based polymer electrolytes for Li negative electrodes. Two distinct phases of polymer electrolytes (solid and gel) were developed (Fig. 1a) as Li negolytes to study their impact on battery cycling performance.

Fig. 1: Electrochemical characterizations.
figure 1

a Schematic of a polymer-based anolyte (solid or gel) and organic catholyte membrane-free system. b Chemical structure of Tri-TEMPO. c Cyclic voltammograms of PVDF-Li and Tri-TEMPO (5 mM) in FEC/LiTFSI (0.1 M). d Cyclic voltammograms of PPC-Li and Tri-TEMPO (5 mM) in TEGDME/LiTFSI (0.1 M). e Cycling stability of PVDF-Li and PPC-Li anolytes. Cyclic voltammograms of Tri-TEMPO (5 mM) at different scan rates in (f) FEC/LiTFSI (0.1 M) and (g) TEGDME/LiTFSI (0.1 M). Cycling stability of Tri-TEMPO (5 mM) in (h) FEC/LiTFSI (0.1 M) and (i) TEGDME/LiTFSI (0.1 M) for 200 cycles at a scan rate of 50 mV/S. Source data are provided as a Source Data file.

Full size image

Development of solid polymer electrolyte

Initially, an N-methyl-2-pyrrolidone (NMP) solution of PVDF (0.2 M) was prepared. Then, different concentrations (0.1, 0.2, 0.3, and 0.4 wt.%) of the electrolyte salt lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) were added to the solution. The mixture was stirred at 50 °C for 12 h, resulting in a viscous, brown-colored solution. This solution was cast onto a glass surface and heat-dried to form a thick sheet (~160 µm, from vernier caliper). The ionic conductivity of the resulting film was measured using electrochemical impedance spectroscopy (EIS). The solid polymer electrolyte showed low ionic conductivity on the order of 10−4 mS/cm, regardless of the composition of LiTFSI (Supplementary Table 1). To further enhance ionic conductivity, the ionic liquid 1-butyl-3-methylimidazolium-tetrafluoroborate (BMIMBF4) was added to the electrolyte solution of PVDF (1.0 wt.%)/LiTFSI (0.3 wt.%). BMIMBF4 was selected as the electrolyte additive because it offers significant advantages, including high ionic conductivity, non-volatility, non-flammability, high chemical and thermal stability, and a wide potential window35. Room-temperature ionic liquids have the potential to function as conductive media and nonvolatile plasticizers in electrolytes33. The increase in ionic conductivity upon incorporating BMIMBF4 into PVDF is attributed to several different types of interactions. Electrostatic interactions occur between the charged species in BMIMBF4(BMIM+ and BF4−) and the dipoles generated by the C–F bonds in PVDF, leading to swelling of the polymer for enhanced ionic mobility36. Hydrogen bonding between the fluorine atoms in PVDF and the hydrogen atoms in the BMIM cation further stabilizes the ionic liquid within the matrix, reducing crystalline regions and increasing amorphousness, which promotes ion migration37,38,39. Further, the solution effects of BMIMBF4 disrupt the PVDF crystalline structure, lowering the glass transition temperature and enhancing ionic conductivity at room temperature36. Moreover, the favorable interactions between BMIMBF4 and PVDF can improve the mechanical stability and Li ion diffusion in electrolytes40. Different concentrations of BMIMBF4 (0.1, 0.2, 0.3, and 0.5 wt.%) with PVDF/LiTFSI (0.3 wt.%) were tested (Supplementary Table 1). The PVDF (1.0 wt.%)/LiTFSI (0.3 wt.%) electrolyte with 0.5 wt.% of BMIMBF4 showed the highest ionic conductivity (1.04 mS/cm). Therefore, the electrolyte solution of PVDF (1.0 wt.%)/LiTFSI (0.3 wt.%)/BMIMBF4 (0.5 wt.%) was used to prepare the solid polymer electrolyte for the Li negative electrode (denoted as PVDF-Li; see the experimental section of the Supplementary Information for details). To develop the biphasic system, the PVDF-Li anolyte was paired with the Tri-TEMPO/FEC (LiTFSI) catholyte. The chemical structure of Tri-TEMPO is shown in Fig. 1b. FEC was selected as the positive electrode solvent because it is chemically compatible with both PVDF and Li.

Development of gel polymer electrolyte

PPC was selected for the preparation of the gel polymer electrolytes because of its unique properties, including a wide voltage window and high mechanical strength. Derived from the copolymerization of CO2 and propylene oxide, PPC is an amorphous alternating copolymer. This distinctive origin not only offers a potential solution for mitigating CO2 emissions from various sources but also makes PPC biodegradable33,41,42. Compared to other polymers such as poly(ethylene oxide) and poly(methyl methacrylate), PPC has a glass transition temperature close to room temperature, enabling its transition into a rubbery state when a small amount of ionic liquid is incorporated33, facilitating Li ion movement within the polymer matrix.

The polymer gel electrolyte was prepared by homogeneously mixing PPC (1.0 wt.%)/LiTFSI (0.2 wt.%) in dimethyl carbonate (DMC), followed by heat drying. The resultant gel polymer electrolyte exhibited a low ionic conductivity of 5.4 × 10−3 S/cm. To enhance the ionic conductivity of the gel polymer electrolyte, the ionic liquid 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide (BMPTFSI) was added in various concentrations (0.5–2.0 wt.%) (see the experimental section of the Supplementary Information for details). The PPC (1.0 wt.%)/LiTFSI (0.2 wt.%) gel polymer electrolyte with 2.0 wt.% BMPTFSI exhibited an optimized ionic conductivity of 2.2 mS/cm (Supplementary Fig. 1, and Supplementary Table 2). This gel polymer electrolyte was coupled with Li to afford the PPC-Li anolyte and paired with the Tri-TEMPO/TEGDME (LiTFSI) catholyte to form a PPC-Li | |Tri-TEMPO membrane-free battery.

In addition, to assess the long-term stability of both gel and solid polymer electrolytes, CV (Supplementary Fig. 2a, b) was performed to evaluate the electrochemical stability window, while EIS (Supplementary Fig. 2c, d) and ionic conductivity (Supplementary Fig. 2e, f) measurements were conducted to monitor potential degradation over time. The results indicated stable electrochemical behavior and no significant degradation in both solid and gel polymer electrolytes, suggesting good long-term stability.

Electrochemical characterization

Cyclic voltammetry (CV) measurements were conducted to assess the redox behavior of Tri-TEMPO in the organic electrolytes (FEC and TEGDME) and to evaluate the electrochemical behavior of the PVDF-Li and PPC-Li anolytes. Tri-TEMPO demonstrated reversible redox couples at 0.37 V vs. Ag/Ag+ in FEC (0.1 M LiTFSI) and 0.39 V vs. Ag/Ag+ in TEGDME (0.1 M LiTFSI). Therefore, pairing the PVDF-Li and PPC-Li anolytes with Tri-TEMPO dissolved in FEC (0.1 M LiTFSI) and TEGDME (0.1 M LiTFSI), respectively, resulted in theoretical cell voltages of 3.43 and 3.45 V (Fig. 1c, d). These values exceeded some relevant previously reported biphasic membrane-free batteries (Supplementary Table 3)14,16,17,19,20,22,23,43,44,45,46. In addition to the high cell voltages, the long-term charge/discharge cycling of symmetric cells with a Li negative electrode in both solid and gel polymer electrolytes demonstrated good cycling stability (Fig. 1e). Further, Li | |Li symmetrical cell tests were performed using the solid polymer electrolyte for 300 hours (13 days), demonstrating stable cycling performance over the extended period (Supplementary Fig. 3). In addition, post-cycling analysis of the Li metal with the solid polymer electrolyte revealed no signs of dendrite formation (Supplementary Fig. 4), further confirming the compatibility of the polymer electrolyte with the Li negative electrode.

The peak potential separations of Tri-TEMPO dissolved in FEC (0.1 M LiTFSI) and TEGDME (0.1 M LiTFSI) at a scan rate of 5 mV/s were approximately 75 and 72 mV, respectively, indicating a one-electron redox process according to the Nernst equation. Subsequently, CV measurements were conducted at different scan rates (50–200 mV/s) to investigate the electrochemical kinetics of the catholyte compounds (Fig. 1f, g). The ratios of the cathodic peak current (ipc) to the anodic peak current (ipa) for Tri-TEMPO at the investigated scan rates were close to 1 in both systems (Supplementary Fig. 5a, b), suggesting the high electrochemical reversibility of the compounds in FEC and TEGDME. Moreover, both the anodic and cathodic peak currents exhibited a direct proportionality with the square root of the scan rate (ν1/2) in both the Tri-TEMPO/FEC and Tri-TEMPO/TEGDME systems (Supplementary Figs. 6 and 7), suggesting a diffusion-controlled redox process. In addition, the CV cycling results revealed the high electrochemical stability of the Tri-TEMPO/FEC and Tri-TEMPO/TEGDME systems, as indicated by the nearly superimposed cyclic voltammograms after one and 200 cycles (Fig. 1h, i).

To gain a deeper insight into the electrochemical kinetics of Tri-TEMPO, linear sweep voltammetry (LSV) was performed alongside rotating disk electrode (RDE) experiments, as depicted in Supplementary Figs. 8 and 9. Determination of the Koutecký–Lévich curves using Eq. 1 (see “Methods” section) enabled the estimation of the diffusion coefficients (D) of the redox materials across varying rotation rates. The diffusion coefficients of Tri-TEMPO in both FEC and TEGDME were found to be 2.1 × 10−5 and 1.7 × 10−5 cm2/s, respectively. Furthermore, the kinetic rate constants (ko) for the redox compounds were derived using the Butler–Volmer equation (Eq. 2, see “Methods” section) and were 1.6 × 10−4 cm/s in FEC and 1.2 × 10−4 cm/s in TEGDME. Notably, these diffusion coefficients and rate constants are comparable to those of selected redox-active organic materials used in aqueous flow batteries47,48. The rapid electrode kinetics of Tri-TEMPO suggest its potential for low activation polarization loss, which holds significant promise for applications in flow batteries.

Charge/discharge performance of the PVDF-Li | |Tri-TEMPO battery

The performance of the PVDF-Li | |Tri-TEMPO battery was explored for charge/discharge under static and flow conditions.

Under static conditions

To assess the viability of the developed PVDF-Li | |Tri-TEMPO membrane-free battery, schematically illustrated in Supplementary Fig. 10, an initial evaluation was conducted under static conditions in an argon filled glove box at 27 °C. PVDF-Li | |Tri-TEMPO (0.2 M) was subjected to charging/discharging at a current density of 0.3 mA/cm2. Although the current density could be increased further given the low overpotential, it is important to highlight that the operational current density was higher than some of previously reported aqueous and nonaqueous biphasic membrane-free systems23,24. PVDF-Li | |Tri-TEMPO (0.2 M) (Fig. 2a) demonstrated good cycling performance with a capacity retention of 94% (99.94% per cycle, 99.64% per day) over 100 cycles (9 days), with a theoretical capacity utilization of ~23%. This confirms the high stability of the redox-active materials and supporting electrolytes under actual battery operating conditions. Moreover, no significant change in the charge transfer resistance (Rct) (Supplementary Fig. 11) after battery cycling indicates the high stability of the Li negative electrode in the solid polymer electrolyte.

Fig. 2: Charge/discharge performance of PVDF-Li | |Tri-TEMPO static batteries.
figure 2

a Capacity retention and Coulombic efficiency (CE) of PVDF-Li | |Tri-TEMPO (0.2 and 0.5 M) under static conditions. b Cyclic voltammograms of the PVDF-Li | |Tri-TEMPO (0.5 M) static battery catholyte before and after cycling. c Charge/discharge performance of the PVDF-Li | |Tri-TEMPO (0.5 M) static battery at different applied current densities. d Variation in the CE, voltage efficiency (VE), and energy efficiency (EE) of the PVDF-Li | |Tri-TEMPO (0.5 M) static battery at different applied current densities. Source data are provided as a Source Data file.

Full size image

Based on these results, a static battery containing a higher concentration of catholyte (0.5 M) was assembled and subjected to long-term charging/discharging at a current density of 0.3 mA/cm2 (Fig. 2a). The PVDF-Li | |Tri-TEMPO (0.5 M) battery exhibited a capacity retention of 91% (99.91% per cycle, 99.47% per day) over 100 cycles (21 days), with a theoretical capacity utilization of ~22%. Moreover, PVDF-Li | |Tri-TEMPO (0.5 M) exhibited CE, VE, and EE values of 96%, 85%, and 81%, respectively. The PVDF-Li | |Tri-TEMPO static battery exhibited a discharge volumetric energy density of 10 Wh/L (Supplementary Table 3). Furthermore, the PVDF-Li | |Tri-TEMPO (0.5 M) static battery exhibited a high CE ( >99%) and energy efficiency (EE) (81%). A CE of 99% suggests negligible side reactions. However, increasing the concentration of the Tri-TEMPO leads to lower coulombic efficiency and reduced capacity retention. This could be due to the following reasons. Higher concentrations of Tri-TEMPO (0.5 M) lead to increased viscosity and reduced ionic conductivity in the electrolyte solution49. The decrease in Coulombic efficiency at higher Tri-TEMPO concentrations is driven by factors, including transport limitations, kinetic barriers, and parasitic reactions. Increased Tri-TEMPO concentration raises electrolyte viscosity, slowing diffusion, and causing concentration polarization near the electrode. This necessitates higher overpotentials to sustain charge transfer, which in turn exacerbates parasitic reaction pathways and contributes to a decline in Coulombic efficiency. In addition, side reactions towards high concentrations, such as disproportionation and solvent interactions, lead to the irreversible decomposition of active species, consuming charge carriers, and degrading the electrolyte. Increased ion pairing or aggregation at high concentrations also reduces the effective concentration of free Tri-TEMPO, further reducing the Coulombic efficiency. These combined factors result in higher charge losses and lower Coulombic efficiency. On the other hand, the PVDF-Li | |Tri-TEMPO static battery demonstrates a capacity retention of approximately 91%, indicating a 9% loss in capacity, the interfacial self-discharge (Supplementary Fig. 12) in the PVDF-Li | |Tri-TEMPO (0.5) static batteries (1.19 mV/h for 0.5 M static battery) could be the primary reason for the capacity loss. For further explanation, while Tri-TEMPO species have low solubility in the PVDF-based solid polymer electrolyte, they may still migrate or diffuse into the electrolyte during prolonged cycling. Electrochemical gradients and mechanical stress can induce microcracks in the electrolyte, facilitating this migration, which contributes to capacity loss. Cyclic voltammograms show a slight decay in peak current values (Fig. 2b), accounting for approximately 50% of the capacity fade. Secondly, dendrite formation on the Li metal negative electrode, even at microscopic levels, can cause overpotentials (Supplementary Fig. 13), resulting in less efficient charge/discharge cycles and further capacity fade. The EIS analysis of the battery before and after cycling (Supplementary Fig. 14) indicated no significant change in the overall battery resistance, confirming the high stability of the battery components throughout the cycling process. Therefore, the high cycling stability of the PVDF-Li | |Tri-TEMPO static batteries with different Tri-TEMPO concentrations (0.2 and 0.5 M) suggests the high compatibility of the Li negative electrode with the developed solid polymer electrolyte and that of graphite felt with the catholyte solvent.

To investigate the impact of high current density on the performance of the PVDF-Li | |Tri-TEMPO (0.5 M) membrane-free batteries under static conditions, the charge/discharge behavior was examined at current densities of 0.3, 0.5, and 0.6 mA/cm2, with the average CE, VE, and EE values measured for three cycles (Fig. 2c). As the current density increased from 0.3 to 0.6 mA/cm2, the PVDF-Li | |Tri-TEMPO (0.5 M) static battery exhibited increased overpotential of approximately 161 mV, which can be attributed to the limitations associated with mass diffusion. The PVDF-Li | |Tri-TEMPO (0.5 M) battery displayed CE values of 99.6%, 99.7%, and 99.9% at current densities of 0.3, 0.5, and 0.6 mA/cm2, respectively, indicating that fast charging/discharging did not induce noticeable side reactions (Fig. 2d). The EE values of the PVDF-Li | |Tri-TEMPO (0.5 M) battery decreased from 88% to 84% and 75% as the applied current density increased from 0.3 to 0.6 mA/cm2, respectively. This decrease in EE is attributed to an increase in overpotential due to mass transport loss. Notably, the battery regained its original efficiency when cycled back to 0.3 mA/cm2, indicating its high charge rate performance.

The polarization of the PVDF-Li | |Tri-TEMPO (0.5 M) battery under static conditions was investigated at different states of charge (SOC). The peak power density of the PVDF-Li | |Tri-TEMPO (0.5 M) battery under static conditions was 34 mW/cm2 at 100% SOC (Supplementary Fig. 15a). Correspondingly, the area-specific resistance (ASR) data for both the resistance of the electrolyte and the whole cell under static conditions were examined. Both the ASR of the electrolyte and that of the whole cell increased with increasing operating current density (Supplementary Fig. 15b). Notably, the electrolyte resistance accounted for over 57% of the whole cell resistance.

Under flow conditions

While membrane-free batteries perform well under static conditions, their energy and power are interdependent, restricting their scalability. Therefore, the charge/discharge performance of the PVDF-Li | |Tri-TEMPO membrane-free battery was examined under flow-catholyte conditions. Under flow configuration, the anolyte side coupled with a solid Li metal negative electrode was kept under static conditions, while the catholyte was circulated at the optimized flow rate (Supplementary Fig. 16a, b).

Optimizing the flow rate is critical for redox flow batteries, as it directly influences mass-transport kinetics, efficiency, and energy output. Moreover, it ensures effective electrolyte circulation, prolongs battery life, and enhances battery performance under varying operational conditions. The catholyte flow rate was evaluated by performing charge/discharge cycles at flow rates ranging from 1 to 4 mL/min. The battery exhibited optimal performance at a flow rate of 3 mL/min (Supplementary Fig. 17). Under optimized catholyte flow conditions, the PVDF-Li | |Tri-TEMPO (0.5 M) hybrid battery was cycled at 0.6 mA/cm2, retaining 82% of its capacity (98.81% per cycle, 99.34% per day) over 100 cycles (22 days). The PVDF-Li | |Tri-TEMPO (0.5) hybrid flow battery exhibits a theoretical capacity utilization of 49%, which is nearly 2.5 times higher than the theoretical capacity utilization of 0.5 M battery under static condition. In addition, the battery exhibits CE, VE, and EE values of 96%, 88%, and 85%, respectively (Fig. 3a). The consistent CV peak current of the catholyte after cycling indicated the stability of the Tri-TEMPO molecules under prolonged battery operation with a flowing catholyte (Fig. 3b). Therefore, the interfacial self-discharge of 1.24 mV/h in the PVDF-Li | |Tri-TEMPO (0.5 M) hybrid flow battery (Supplementary Fig. 18) could be the primary reason for capacity loss. Furthermore, the EIS spectra (Supplementary Fig. 19) revealed no significant increase in cell resistance, suggesting the sustained integrity of the Li throughout the cycling period. In addition, the post-mortem characterizations Li metal negative electrode of PVDF-Li | |Tri-TEMPO (0.5 M) hybrid flow battery conducted using scanning electron microscopy (SEM), indicate no evidence of dendrite formation or degradation of the Li metal negative electrode over an extended operational period of approximately 22 days (Supplementary Fig. 20). These findings underscore the practical viability of polymer-coated Li-based membrane-free flow batteries, which exhibit enhanced scalability and stable performance under dynamic flow conditions.

Fig. 3: Charge/discharge performance of PVDF-Li | |Tri-TEMPO hybrid flow batteries.
figure 3

a Capacity retention and Coulombic efficiency (CE), voltage efficiency (VE), the energy efficiency (EE) of the PVDF-Li | |Tri-TEMPO (0.5 M) battery under flow conditions for 100 charge/discharge cycles at an applied current density of 0.6 mA/cm2. b Cyclic voltammograms of the catholyte of the PVDF-Li | |Tri-TEMPO (0.5 M) hybrid flow battery before and after 100 charge/discharge cycles. c Charge/discharge performance of PVDF-Li | |Tri-TEMPO (0.5 M) hybrid flow battery at different applied current densities. d Variation in the CE, VE, and EE of the PVDF-Li | |Tri-TEMPO (0.5 M) hybrid flow battery at different applied current densities. Source data are provided as a Source Data file.

Full size image

Moreover, the charge-rate capabilities of the PVDF-Li | |Tri-TEMPO (0.5 M) hybrid flow battery were examined at current densities of 0.6, 0.8, and 1.0 mA/cm2. Figure 3c shows the cycling profiles corresponding to these densities, with the CE, VE, and EE values averaged over three cycles at 0.6 mA/cm2 shown in Fig. 3d. At 1.0 mA/cm2, the battery exhibited a discharge overpotential of 92 mV, which is slightly higher than that observed at 0.8 mA/cm2. This increase in overpotential and subsequent decrease in the VE at higher current densities are attributed to mass transport constraints. However, upon reverting to the applied current density of 0.8 mA/cm2, the battery regained its original performance, indicating good rate cyclability of the membrane-free battery under flow conditions. Moreover, a detailed analysis of the ASR results for the electrolyte and entire cell under flow conditions revealed intriguing insights. The electrolyte resistance contributes over 62% of the whole cell resistance (Supplementary Fig. 21a). However, there was a discernible decrease in both the electrolyte and whole cell ASR under flow conditions compared with those under static conditions. This decrease in resistance enables the battery to achieve higher current and power densities without compromising its performance. Consequently, the peak power density of the PVDF-Li | |Tri-TEMPO (0.5 M) battery under flow conditions increased significantly to 55 mW/cm2 (Supplementary Fig. 21b) compared to that of the PVDF-Li | |Tri-TEMPO static battery (34 mW/cm2). This improvement in peak power density can be attributed to the improved mass transport achieved under flow conditions.

Charge/discharge performance of the PPC-Li | |Tri-TEMPO battery

The charge/discharge performance of the PPC-Li | |Tri-TEMPO membrane-free battery was investigated under static and flow conditions.

Under static conditions

Initially, two static membrane-free batteries were assembled by pairing the PPC-Li anolyte with the TEGDME/Tri-TEMPO (0.2 M and 0.5 M) catholyte (Supplementary Fig. 22). To assess the battery cycling performances, the cells underwent charging and discharging within the voltage range of 3.8–2.6 V, adhering to cut-off limits. All the charge/discharge experiments were conducted in an argon gas glove box (H2O: < 0.1 ppm and O2: < 0.1 ppm) at a temperature of 27 °C. To evaluate the long-term cycling performance, the 0.2 and 0.5 M batteries were operated at a current density of 0.5 mA/cm2. The PPC-Li | |Tri-TEMPO (0.2 M) cell exhibited a capacity retention of 97.86% (99.97% per cycle, 99.78% per day, Fig. 4a) over 100 cycles (10 days). Similarly, the PPC-Li | |Tri-TEMPO (0.5 M) cell exhibited a capacity retention of 96.75% (99.96% per cycle, 99.78% per day) over 100 cycles (25 days). Further, the PPC-Li | |Tri-TEMPO batteries under the static configurations exhibit capacity utilizations of approximately 29% and 27% at the 0.2 M and 0.5 M concentrations, respectively. In addition, both static batteries displayed high CE values of 98.6% (0.2 M) and 97.8% (0.5 M). Furthermore, the CV results of the pre- and post-cycling electrolytes for both static cells showed negligible changes in the peak current densities (Fig. 4b), indicating the electrochemical stability of Tri-TEMPO throughout the extended cycling period. Furthermore, the EIS analysis of PPC-Li | |Tri-TEMPO (0.5 M) showed no significant increase in either the solution or charge transfer resistance, suggesting the good stability of the active battery components under prolonged battery cycling (Supplementary Fig. 23).

Fig. 4: Charge/discharge performance of PPC-Li | |Tri-TEMPO static batteries.
figure 4

a Capacity retention and Coulombic efficiency (CE) of PPC-Li | |Tri-TEMPO (0.2 M and 0.5 M) under static conditions at a current density of 0.5 mA/cm2. b Cyclic voltammograms of the PPC-Li | |Tri-TEMPO (0.5 M) static battery catholyte before and after 100 charge/discharge cycles. c Charge/discharge performance of the PPC-Li | |Tri-TEMPO (0.5 M) static battery at different applied current densities. d Variation in the CE, voltage efficiency (VE) and energy efficiency (EE) of the PPC-Li | |Tri-TEMPO (0.5 M) static battery at different applied current densities. Source data are provided as a Source Data file.

Full size image

The charge/discharge behavior of the PPC-Li | |Tri-TEMPO (0.5 M) static battery was investigated at current densities of 0.5, 0.8, and 1.0 mA/cm2, and the average CE, VE, and EE values were measured over four cycles. As the current density increased from 0.5 to 0.8 and from 0.8 to 1.0 mA/cm2, the PPC-Li | |Tri-TEMPO (0.5 M) static battery exhibited an increase in charge voltages of 47 and 52 mV, respectively, accompanied by a decrease in discharge voltage of 52 and 67 mV (Fig. 4c). The increase in overpotential at higher current densities is attributed to the limitations associated with mass diffusion. The PPC-Li | |Tri-TEMPO static battery exhibited negligible changes in CE at current densities of 0.5, 0.8, and 1.0 mA/cm2 (Fig. 4d), indicating that fast charging/discharging did not induce significant side reactions. The EE values of the PPC-Li | |Tri-TEMPO (0.5 M) static battery decreased from 88% to 84% and 75% as the applied current density was increased from 0.5 to 0.8 and 1.0 mA/cm2, respectively. The decrease in EE was attributed to an increase in the overpotential due to mass transport losses. Notably, all three batteries regained their original efficiencies when cycled back to 0.5 mA/cm2, confirming their improved charge-rate performance.

The polarization of the PPC-Li | |Tri-TEMPO (0.5 M) batteries under static conditions was explored at different SOCs. The peak power density of the PPC-Li | |Tri-TEMPO (0.5 M) battery under static conditions was 64 mW/cm2 at 100% SOC (Supplementary Fig. 24a). The ASR data for the resistance of both the electrolyte and the whole cell under static conditions were analyzed. The ASR of both the electrolyte and the whole cell increased with increasing operating current density. Notably, the electrolyte resistance accounted for over 50% of the whole cell resistance (Supplementary Fig. 24b). Furthermore, the energy density of the static PPC-Li | |Tri-TEMPO (0.5 M) battery was 13.2 Wh/L.

In addition, we investigated the impedance behavior of each component in the system using a three-electrode setup (Supplementary Fig. 25). For the positive electrode, the equivalent circuit consists of bulk impedance in series with a parallel system (comprising a constant phase element and impedance). For the negative electrode, the equivalent circuit includes bulk impedance in series with two parallel systems, representing the SEI layer and the gel electrolyte, respectively. We separately measured the impedance of the positive electrode (ab), the negative electrode (bc), and the entire system (ac), and performed fitting analyses. The results show that the bulk impedance of the positive electrode and negative electrode are similar, with the positive electrode being 10.3 Ω, the negative electrode 8.47 Ω, and the entire system 18.76 Ω. The impedances of the negative electrode’s SEI layer and gel electrolyte are approximately 72.72 Ω and 93.75 Ω, respectively, which are the main limiting factors of this system.

Under flow conditions

To scale its energy and power independently, a PPC-Li | |Tri-TEMPO biphasic battery was fabricated under flowable conditions (Supplementary Fig. 16a, b). The optimal catholyte flow rates were evaluated through charge/discharge cycles at various rates of 1–4 mL/min, with optimal performance observed at 4 mL/min (Supplementary Fig. 26). Consequently, a PPC-Li | |Tri-TEMPO flow battery with an optimized flow rate was constructed, and its long-term cycling performance was assessed under an inert argon atmosphere at a current density of 0.8 mA/cm2.

The PPC-Li | |Tri-TEMPO hybrid flow battery was cycled for 100 charge/discharge cycles (37 days), retaining 78% of its capacity (99.8% per cycle, or 99.4% per day). In addition, it exhibited CE, VE, and EE values of approximately 96%, 89%, and 86%, respectively (Fig. 5a). Notably, the PPC-Li | |Tri-TEMPO hybrid flow battery at a 0.5 M concentration delivers a good capacity retention of ~86%, which is nearly ~3.2 times higher than the PPC-Li | |Tri-TEMPO (0.5 M) static battery and ~1.8 times higher than the PVDF-Li | |Tri-TEMPO (0.5 M) hybrid flow battery. Subsequent CV tests on the battery catholyte after charge/discharge cycling revealed a slight decrease in current density (Fig. 5b), suggesting capacity fading over prolonged cycling. To investigate whether Tri-TEMPO crosses over into the anolyte phase, a CV test was conducted on the gel anolyte dissolved in dichloromethane (Supplementary Fig. 27). The resulting CV trace of the anolyte showed traces of Tri-TEMPO, confirming its crossover from the catholyte to the anolyte phase due to convective mass-transport-induced disturbances at the liquid–gel electrolyte interface under flow conditions. The CV qualification analysis results (Supplementary Fig. 28) suggest that 89% of the total capacity fade is caused by the crossover of Tri-TEMPO molecules due to convective diffusion from the flowing catholyte to the gel anolyte. Hence, the PPC-Li | |Tri-TEMPO (0.5 M) hybrid battery under flow configuration exhibits a slightly higher voltage loss, with a rate of 1.27 mV/h (Supplementary Fig. 29), compared to the static configuration (0.98 mV/h) (Supplementary Fig. 30). However, the EIS analysis revealed no significant change in Rct after battery cycling (Supplementary Fig. 31), suggesting a healthy state of the Li negative electrode in the gel polymer electrolyte. In addition, post-mortem analysis of the Li metal negative electrode in the PPC-Li | |Tri-TEMPO (0.5 M) hybrid flow battery, performed using SEM, reveals no evidence of dendrite formation or deterioration of the Li metal negative electrode over an extended operational period of approximately 37 days (Supplementary Fig. 32).

Fig. 5: Charge/discharge performance of PPC-Li | |Tri-TEMPO hybrid flow batteries.
figure 5

a Capacity retention and Coulombic efficiency (CE), voltage efficiency (VE), and energy efficiency (EE) of PPC-Li | |Tri-TEMPO (0.5 M) under flow conditions for 100 charge/discharge cycles at an applied current density of 0.5 mA/cm2. b Cyclic voltammograms of the catholyte of the PPC-Li | |Tri-TEMPO (0.5 M) flow battery before and after 100 charge/discharge cycles. c Charge/discharge performance of PPC-Li | |Tri-TEMPO (0.5 M) flow battery at different applied current densities. d Variation in CE, VE, and EE of the PPC-Li | |Tri-TEMPO (0.5 M) flow battery at different applied current densities. Source data are provided as a Source Data file.

Full size image

The charge-rate performance of the PPC-Li | |Tri-TEMPO (0.5 M) hybrid flow battery was evaluated at current densities of 0.8, 1.0, and 1.2 mA/cm2 (Fig. 5c). The corresponding cycling profiles (Fig. 5c) and CE, VE, and EE measurements (Fig. 5d) were recorded over an average of three cycles. At a current density of 1.2 mA/cm2, the battery displayed a discharge overpotential of 145 mV, which is nearly double that observed at 1.0 mA/cm2. This increase in the overpotential and the subsequent reduction in VE at higher current densities are attributed to mass transport limitations. However, upon cycling again at the applied current density of 0.8 mA/cm2, the battery recovered its original performance, indicating great rate cyclability under flow conditions. To better understand the specific sources of resistance, electrolyte vs. cell resistance, we analyze the contributions of both in the PVDF-Li | |Tri-TEMPO and PPC-Li | |Tri-TEMPO battery systems. The PVDF-Li | |Tri-TEMPO (0.5 M) battery exhibited electrolyte resistance contributing 57% (Supplementary Fig. 15b) and 52% (Supplementary Fig. 21a) of the total cell resistance for the static and flow battery configurations, respectively. On the other hand, the PPC-Li battery displayed electrolyte resistance of 50% (Supplementary Fig. 24a) and 62% (Supplementary Fig. 33a) of the whole cell under static and flow conditions, respectively. Nevertheless, there was a notable decrease in both the electrolyte and whole cell ASR under flow conditions compared to the static battery. This reduction in the ASR under flow conditions enables the battery to achieve higher current and power densities without compromising its performance. Consequently, the peak power density of the PPC-Li | |Tri-TEMPO (0.5 M) hybrid battery under flow conditions significantly surpassed that of the static battery (65 mW/cm2), reaching 126 mW/cm2 (Supplementary Fig. 33b). The enhanced peak power density can be attributed to the improved mass transport under flow conditions. The better performance of the high-concentration battery highlights its potential for fabricating systems with high catholyte loadings.

Comparison between the two designs

Membrane-free batteries based on PVDF-Li and PPC-Li polymer anolytes exhibited distinct attributes and performance characteristics owing to their different polymer matrices, physical states, and electrolyte additives. Future advancements in both systems should focus on optimizing the ionic conductivity for higher ion transport rates, potentially through the selection of more suitable electrolyte salts and additives. For solid polymer electrolytes, improving the interaction between solid polymers and ionic liquids, as well as exploring polymer blends, can enhance performance. For gel polymer electrolytes, tailoring the concentrations of electrolyte solvents, salts, and additives may further boost conductivity and stability, potentially leading to more efficient and durable batteries.

Sizing Analysis and Economical perspective

We acknowledge that the coupling of energy and power due to the use of static Li negative electrode may introduce sizing constraints that affect scalability and cost-effectiveness. We have addressed this concern through a comparative sizing analysis between our hybrid system and a conventional vanadium redox flow battery To compare the energy and power sizing of the PVDF-Li | |Tri-TEMPO and PPC-Li | |Tri-TEMPO hybrid flow batteries with a fully flowable VRFB system, each targeting 10 kWh of energy and 5 kW of power over 2 h, we have observed through detailed analysis (see Supplementary Information) that the VRFB achieves an efficient storage volume of 500 L, comprising balanced 250 L anolyte and catholyte tanks at 25 Wh/L, complemented by a 5 m2 electrode area with a power density of 1 kW/m². The PVDF-Li | |Tri-TEMPO hybrid requires 760.90 L of Tri-TEMPO catholyte at 22 Wh/L to deliver 10 kWh, surpassing VRFB’s volume due to lower energy density, yet benefits from the Li negative electrode’s high capacity (1891 mAh/g at 49% utilization), which minimizes its contribution. Conversely, the PPC-Li | |Tri-TEMPO hybrid significantly reduces the catholyte volume to 387.65 L at 42.7 Wh/L, 45% less than VRFB, owing to enhanced energy density and 86% Li utilization (3319 mAh/g), showcasing high efficiency. Both hybrids meet the energy and power demands, with positive electrode areas of 2.591 m2 (PVDF, 1.93 kW/m2) and 1.171 m2 (PPC, 4.27 kW/m2) driving performance. It is worth noting that the volume of the polymer electrolyte has not been optimized. Theoretically, the required volume should be just sufficient to encapsulate the lithium metal. These findings demonstrate the PPC hybrid’s compactness, establishing semi-flow systems as viable, space-efficient alternatives to VRFB’s fully flow design.

The key components of a typical NRFB include an ion-exchange membrane, a porous carbon electrode, a solvent, an electrolyte salt, a bipolar plate, a current collector plate, and additional elements such as pumps, gaskets, pipes, bolts, and endplates. The major cost components of these systems are the ion-exchange membrane (20–30%), solvent and salt (30–70%), porous carbon electrode (5–10%), and carbon bipolar plate (10–15%)50,51. Despite advancements, the capital cost of NRFB systems remains prohibitively high, ranging from $800/kWh to $2000/kWh51,52. The high cost can be attributed to several factors, including the use of low concentrations of redox materials, necessitating the use of large volumes of expensive nonaqueous electrolytes, which can cost up to $3/L or $9400/kWh52. In addition, the restricted availability and high cost of ion-exchange membranes increase the overall cost of NRFB systems. Therefore, minimizing the use of nonaqueous electrolytes and ion-exchange membranes can effectively reduce the overall cost of NRFB systems. The present study demonstrates an all-organic NRFB utilizing liquid/solid and liquid/gel biphasic systems. This unique configuration requires significantly less electrolyte solvent volume than conventional NRFB systems, potentially significantly decreasing the capital cost. In addition, this configuration eliminates the need for expensive ion-exchange membranes, further reducing the cost of the battery systems. Moreover, the hybrid flow system significantly impacts the economic evaluation compared to a battery with fully flow configurations. Since only the catholyte is flowing and the anolyte remains static, pumping costs for the negative electrode side are eliminated, reducing overall operational costs. These factors contribute to lower pumping and material costs, enhancing the system’s economic viability. Therefore, the developed configuration not only offers an efficient method for constructing all-nonaqueous biphasic systems with flow capabilities but also presents a potential pathway for developing comparatively more affordable, safe, and high-voltage NRFB systems. Furthermore, we recognize the importance of scalability, especially for grid-level energy storage. While our study focused on optimizing battery performance under controlled conditions, further steps are needed to evaluate its commercial viability. The all-nonaqueous biphasic system offers promise in reducing capital costs by minimizing the need for expensive ion-exchange membranes and electrolytes. However, scaling up will require optimization of parameters like electrolyte volume, material costs, and system stability. Future work will include developing prototypes, assessing performance under realistic conditions, and evaluating the economic and practical aspects of transitioning to commercial production.

In summary, two types of all-nonaqueous biphasic systems were successfully developed, featuring solid and gel polymer electrolytes as anolytes and organic solvents as catholytes. The ionic conductivities of both the polymer electrolytes were meticulously investigated using lithium salts and ionic liquids. Two distinct biphasic systems were created: a liquid/solid biphasic system utilizing FEC as the catholyte and a PVDF-based solid electrolyte as the anolyte, and a liquid/gel system utilizing TEGDME as the catholyte and a PPC-based gel polymer electrolyte as the anolyte. For the catholyte, PVDF-Li and PPC-Li anolytes were coupled with Tri-TEMPO in FEC/LiTFSI and TEGDME/LiTFSI electrolytes, respectively, resulting in cell voltages of 3.43 and 3.45 V. The charge/discharge performances of the PVDF-Li | |Tri-TEMPO and PPC-Li | |Tri-TEMPO cells were investigated under static and flow conditions. At 0.5 M, the PVDF-Li | |Tri-TEMPO battery demonstrated capacity retentions of 90.7% and 81.78% and CEs of 95.4% and 95.7% under static and flow conditions, respectively, while the PPC-Li | |Tri-TEMPO battery exhibited capacity retentions of 96.8% and 78% and CEs of 97.8% and 96% under similar conditions. The performance of membrane-free batteries based on gel polymer electrolytes, as demonstrated in this study under both static and flow conditions, is in good agreement with similar state-of-the-art battery systems reported in the literature (Supplementary Table 3). These findings underscore the feasibility and effectiveness of the proposed strategy for enhancing battery performance and provide insights into the development of advanced nonaqueous biphasic systems for energy storage applications. Moreover, while the developed batteries offer high cell voltage operation and stable long-term cycling; however, their performance at high current densities remains a challenge. This issue arises due to the lower ionic conductivity and slower kinetics of nonaqueous electrolyte solvents. Hence, at extremely high current densities, the batteries struggle to maintain a reasonable charge/discharge profile due to excessive overpotentials. Therefore, further research will focus on improving the electrolyte charge transport properties to address this limitation.

Methods

Materials

PVDF (MW ~ 400,000) and PPC (MW ~ 50,000) were purchased from Sigma-Aldrich®. LiTFSI (99.94%), DMC (99%), FEC (99%), and TEGDME (98.0%) were purchased from Fisher Scientific and used without further purification. The ionic liquids, BMIMBF4 (98%) and BMPTFSI (98%) were purchased from TCI Chemicals and used as received. Tri-TEMPO was prepared as per the reported procedures11.

Posolyte preparation

To make the nonaqueous catholyte, 3 mL of either FEC or TEGDME was poured into a clean glass vial. LiTFSI was then added to reach a final concentration of 1.0 M. All of this was done inside an argon-filled glove box (H2O and O2 levels kept below 0.1 ppm) at about 27 °C. The mixture was gently shaken and then placed in an ultrasonic bath for 10 minutes. This simple combination of shaking and sonication gave a clear, well-mixed solution that was ready for subsequent use.

Electrochemical characterization

Cyclic voltammetry (CV) measurements of the redox-active positive electrode material, Tri-TEMPO (5 mM), were performed in FEC/LiTFSI (0.1 M) and TEGDME/LiTFSI (0.1 M) electrolyte solutions. The glassy carbon disk working electrode (3 mm diameter, CHI Instruments) was polished before testing using a aluminum oxide slurry (Pine Research Instrumentation, Inc., 5 µm) with water as the medium. A platinum wire (0.5 mm diameter, 99.99% purity) served as the counter electrode, and an Ag/Ag⁺ reference electrode (Fisher Scientific) was used. CV experiments for PVDF-Li were conducted using a 0.1 M FEC/LiTFSI electrolyte solution, with PVDF-Li configured as the working electrode, a platinum wire (0.5 mm diameter, 99.99% purity) worked as the counter electrode and an Ag/Ag⁺ electrode (Fisher Scientific) was used as the reference electrode. For CV studies of Li in a PPC gel polymer electrolyte, Li metal served as the working electrode, paired with same counter and reference electrodes as in the PVDF-Li experiments. The CV measurements were performed at a scan rate of 50 mV/s, spanning a potential range of 0.01 V to 0.72 V vs. Ag/Ag⁺. EIS was carried out over a frequency range of 200 kHz to 100 mHz using a 10 mV amplitude perturbation under open-circuit voltage (OCV) conditions. Each EIS measurement generated 54 data points. Three replicates were tested for both CV and EIS to ensure reproducibility. All the CV and EIS experiments were conducted using a Bio-Logic potentiostat.

LSV was conducted using a rotating disk electrode (RDE) setup, where a 5 mm diameter glassy carbon electrode served as the working electrode. The counter and reference electrodes were Pt wire and Ag/Ag⁺, respectively. Prior to each measurement, the electrolyte was purged with ultra-pure argon (99.999%) for 20 minutes to remove dissolved oxygen. LSV scans were performed at 50 mV/s within a potential range of 0.02 V to 0.72 V versus Ag/Ag⁺, testing three samples at varying rotation speeds between 100 and 1600 rpm. The diffusion coefficient (D₀) of the redox-active materials was derived from the Levich equation, correlating limiting current density (i) with angular rotation rate (ω), as expressed below in Eq. 1;

$${{\rm{i}}}=0.62{{\rm{nFA}}}{{{\rm{C}}}}_{{{\rm{o}}}}{{{\rm{D}}}}^{2/3}{{{\rm{\omega }}}}^{1/2}{{{\rm{\upsilon }}}}^{-1/6}$$
(1)

where n is the number of electrons involved in the redox process, F is Faraday’s constant, A is the electrode surface area, C₀ is the concentration of active species, ω is the angular velocity, and υ is the kinematic viscosity of the electrolyte solutions.

In addition, the kinetic rate constant (k₀, cm/s) was calculated using the exchange current (io) obtained from Butler-Volmer fitting, according to Eq. 2;

$${{\mbox{i}}}_{{\mbox{o}}}={\mbox{FA}}{{{\rm{C}}}}_{{{\rm{o}}}}{{{\rm{k}}}}_{{{\rm{o}}}}$$
(2)

Here, io corresponds to the intercept of the fitted curve (0.0003 A), with A = 0.196 cm2, and C₀ = 0.5 × 10⁻⁶ mol/cm3.

Battery testing

The electrochemical performance of PVDF | |Tri-TEMPO and PPC-Li | |Tri-TEMPO batteries was assessed in both static and flow configurations using Land and Bio-Logic potentiostats. Testing was conducted over 100 cycles with charge and discharge cutoff voltages set at 3.9 V and 2.6 V, respectively. The PVDF-Li | |Tri-TEMPO batteries were operated at current densities of 0.3 mA/cm² for static configurations and 0.6 mA/cm² for hybrid flow systems, while PPC-Li | |Tri-TEMPO batteries were tested at 0.5 mA/cm² (static) and 0.8 mA/cm² (flow). Li metal chips with an active surface area of 0.3 cm² served as the negative electrode. The positive electrode was a circular graphite felt disc (diameter ~0.3 cm, thickness ~0.3 cm for static and ~0.6 cm for flow setups). Both electrodes were housed in a cylindrical glass cell designed for electrochemical experiments in static and flow configurations. For these tests, catholyte volumes of 0.15 mL (static) and 0.6 mL (flow) were used for PVDF-Li | |Tri-TEMPO batteries, and 0.15 mL (static) and 0.5 mL (flow) for PPC-Li | |Tri-TEMPO batteries. To investigate charge rate performance, PVDF-Li | |Tri-TEMPO batteries were cycled at current densities of 0.3, 0.5, and 0.6 mA/cm2 in static mode and 0.6, 0.8, and 1.0 mA/cm2 in flow mode. Similarly, PPC-Li | |Tri-TEMPO batteries were cycled at 0.5, 0.8, and 1.0 mA/cm2 (static) and 0.8, 1.0, and 1.2 mA/cm2 (flow). To ensure reproducibility, duplicate cells were tested for each battery concentration.

The Coulombic efficiency (CE), voltage efficiency (VE), and energy efficiency (EE) were calculated using the following Equations, as adapted from established methodologies53.

$${{\rm{CE}}}=\left[\frac{{{\rm{discharge}}}\; {{\rm{capacity}}}}{{{\rm{charge}}}\; {{\rm{capacity}}}}\right]\times 100$$
(3)
$${{\rm{VE}}}=\left[\frac{{{\rm{average}}}\; {{\rm{discharge}}}\; {{\rm{voltage}}}}{{{\rm{average}}}\; {{\rm{charge}}}\; {{\rm{voltage}}}}\right]\times 100\%$$
(4)

and

$${{\rm{EE}}}=[{{\rm{CE}}}\times {{\rm{VE}}}]$$
(5)

Area-specific resistance (ASR) for both static and flow cells was determined via linear sweep voltammetry (LSV). Polarization curves were generated by measuring current (i) as a function of applied voltage (V). The differential resistance (dV/di) was derived from the linear region of the polarization curve, typically at low overpotentials, where electrochemical kinetics dominate. The ASR was calculated as the inverse of the slope (S) of this linear region (ASR = 1/S), normalized to the electrode area. Duplicate cell tests were conducted to confirm result consistency. All experiments were performed in an argon-filled glove box (H2O < 0.1 ppm, O2 < 0.1 ppm) at 27 °C. Power density was evaluated for PVDF-Li | |Tri-TEMPO and PPC-Li | |Tri-TEMPO batteries at 0.5 M Tri-TEMPO concentration under various states of charge (SOC) in both static and flow configurations.

To assess the stability of Li metal electrodes with solid or gel polymer electrolytes, symmetric Li-Li cells were assembled using stainless-steel cells with either a solid or gel polymer electrolyte sandwiched between two lithium metal foils. These cells were cycled at 0.5 mA/cm2 for 30-minute intervals per cycle.

Measurement of capacity utilization and retention

Battery performance was evaluated by quantifying capacity utilization and retention. Capacity utilization was determined as the percentage ratio of the operational capacity (Cop) to theoretical capacity (Cth) at a defined state of charge (SOC), as expressed in Eq. 6. The SOC was established by monitoring the charge-discharge cycle durations.

$${{\rm{Capacity}}}\;{{\rm{utilization}}}\,(\%)=\left[\frac{{{{\rm{C}}}}_{{{\rm{op}}}}}{{{{\rm{C}}}}_{{{\rm{th}}}}}\right]\times 100$$
(6)

Here, Cop denotes the measured capacity during charge or discharge, and Cth represents the maximum theoretical capacity of the battery.

Capacity retention, reflecting the battery’s ability to maintain performance over multiple cycles, was calculated using Eq. 7. This metric compares the discharge capacity in subsequent cycles (Cn) to the initial cycle’s discharge capacity (C0).

$${{\rm{Capacity}}}\; {{\rm{retention}}}\,\left(\%\right)=\left[\frac{{{{\rm{C}}}}_{{{\rm{n}}}}\,}{{{{\rm{C}}}}_{0}}\right]\times 100$$
(7)

In this equation, C0 is the discharge capacity in the first cycle, and Cn is the discharge capacity in the nth cycle. These metrics provide critical insights into the battery’s operational efficiency and long-term stability.

Solubility test

To evaluate the solubility of Tri-TEMPO, three electrolyte systems were examined: FEC/LiTFSI (1 M), TEGDME/LiTFSI (1 M), and a PPC gel polymer. In each case, the electrolyte was gradually added to a vial containing a known amount of Tri-TEMPO until a clear, fully dissolved solution was obtained. The volume of electrolyte required was measured using an analytical pipette. Tri-TEMPO solubility was then calculated as the ratio of the molar amount of Tri-TEMPO to the corresponding electrolyte volume. Each measurement was performed twice to ensure accuracy, and the average values were reported. The maximum solubility of Tri-TEMPO was found to be 0.66 M in FEC/LiTFSI (1 M) and 0.71 M in TEGDME/LiTFSI (1 M). In contrast, the solubility in the gel or solid polymer electrolyte was negligible.

Electrolyte preparation

Solid polymer electrolyte

To prepare the solid polymer electrolyte, PVDF pellets (0.2 M) were dissolved in NMP solution and stirred for 6 hours at 50 °C to achieve a homogeneous solution. LiTFSI electrolyte salt was then added to the solution at wt.% of 0.1, 0.2, 3, and 0.4, relative to PVDF. The solution was stirred for 12 hours, resulting in a viscous, brown-colored solution at 50 °C. This solution was cast on a glass surface and dried to form thick solid polymer electrolyte films, after which the ionic conductivity of the films was measured using EIS. All solid polymer electrolyte films exhibited low ionic conductivity on the order of 10−4 S/cm, regardless of the LiTFSI compositions. To enhance ionic conductivity, BMIMBF4 ionic liquid was added to the PVDF/LiTFSI electrolyte as an additive. Initially, a low amount of BMIMBF4 (0.1 wt.% relative to PVDF) was added to the PVDF/LiTFSI (0.3 wt.%) system, resulting in a significant increase in ionic conductivity (from 1.04 × 10−4 mS/cm to 0.21 mS/cm). Therefore, three different concentrations of BMIMBF4 (0.2, 0.3, and 0.5 wt%) were tested with PVDF/LiTFSI (0.3 wt.%), and the results are listed in Supplementary Table 1. A solution of PVDF (1.0 wt%)/LiTFSI (0.3 wt%)/BMIMBF4 (0.5 wt%) in NMP was cast onto the Li metal negative electrode and heated for 10 hours at 50 °C to ensure complete evaporation of NMP solvent and the formation of a uniform solid layer on the Li metal surface, referred to as PVDF-Li anolyte. This PVDF-Li anolyte was used in subsequent investigations. All preparations were conducted within an argon-filled glove box (H2O: < 0.1 ppm and O2: < 0.1 ppm) at a temperature of 27 °C.

Gel polymer electrolyte

To prepare the polymer electrolyte gel, PPC/LiTFSI (1:0.3, wt.%/wt.%) were initially dissolved in DMC solvent and stirred the mixture for 8 hours. Following this, various amounts of BMPTFSI were added to the PPC/LiTFSI solution while stirring. Various weight ratios of BMPTFSI, ranging from 0.5 wt.% to 2.0 wt.%, were studied (see Supplementary Table 2). The mixtures were thoroughly stirred at 27 oC for at least 2 hours. Afterward, the solvent was allowed to gradually evaporate under vacuum drying at 50 °C overnight. The conductivity of the gel polymer electrolyte increased with higher weight ratios of the BMPTFSI ionic liquid. The gel polymer electrolyte exhibited an optimized ionic conductivity of 2.2 mS/cm at 2 wt.%. The gel polymer electrolyte designated as PPC/LiTFSI (0.3 wt.%)/BMPTFSI (2 wt.%) was selected for subsequent studies.