Rigid-flexible heptazine-biguanide frameworks enable fast electron delocalization and low-steric-hindrance ammonium-ion storage

Abstract

Polymer anodes solve the solubility issue of small molecules while offering structure-function merits compared with inorganics for superior ammonium-ion batteries (AIBs), but current research focuses either on rigid polymers for rapid ion transport or flexible ones for high active-site utilization. Here, we design polymeric heptazine-biguanide frameworks (HBFs) via integrating planar three-electron meleme and rotated four-electron chlorhexidine linkers, which harness the advantages of rigid heptazine and flexible biguanide while alleviating their respective shortcomings. Heptazines afford fast electron delocalization, and biguanide chains reduce steric hindrances, leading to ultrahigh utilization of imine sites (99.6%) and ultralow activation energy (0.15 electron volts) in HBFs. Septuple hydrogen-bonded NH₄⁺ coordination per heptazine-biguanide module enables a record capacity (314 milliampere hours per gram) and an exceptional rate capability (60 amperes per gram) among reported polymer anodes. The structural merits of HBFs also enable state-of-the-art all-polymer AIBs with unprecedented energy density (100.6 watt-hours per kilogram of cell) and long life (120,000 cycles). This work gives a previously unidentified paradigm for designing rigid-flexible organic materials toward better AIBs.

A rigid-flexible heptazine-biguanide framework anode enables fast electron delocalization and low-steric-hindrance NH₄⁺ storage.

INTRODUCTION

Aqueous batteries have emerged as promising candidates for grid-scale energy storage due to their resource richness, low cost, and environmental friendliness (1–3). Compared to prevalent high-charge metallic ions (e.g., Zn²⁺, Ca²⁺, Mg²⁺, and Al³⁺) (4–8), nonmetallic ions (e.g., H⁺ and NH₄⁺) show small hydrated structures, low weights, resource sustainability, and fast kinetics (9, 10), making them appealing shuttle ions for propelling aqueous batteries. Although the smallest and lightest H⁺ ions can achieve ultrafast kinetics but inevitably cause electrode corrosion, thereby degrading electrochemical stability (11–14). In contrast, nonmetallic NH₄⁺ ions have favorable tetrahedral shapes and strong preferential orientations, allowing for flexible H-bonding interactions with host materials (9, 15). NH₄⁺ ions thus demonstrate the strong vitality to synergistically enhance redox kinetics and structural stability for propelling stable aqueous ammonium-ion batteries (AIBs) (16–22).

A key task for advanced AIBs is to engineer well-compatible anode materials to fully unlock their electrochemcial properties (23–25). Several inorganic materials (e.g., Prussian blue and Mn/V/Mo-based metal oxides) are booming for AIBs but typically face slow kinetics and inferior stability (26–28). The alternative option was currently expanded to aromatic organic materials owing to their abundant resources and greater flexibility in structural and functional design at the molecular level, enabling systematic optimization of NH₄⁺ storage metrics (26, 29–33). Small molecules with high-mass-content ratio of active groups that allow more electron transfer (e.g., 4,9,10-perylenetetracarboxylic dianhydride and 3,4,9,10-perylenetetracarboxylic diimide) have shown promising NH₄⁺ storage performances (34, 35). However, their high dissolution in aqueous electrolytes result in structural instability, triggering rapid capacity attenuation and short life span (<2000 cycles) (36).

To address this problem, researchers resorted to polymerize soluble small molecules into rigid or flexible antidissolution polymers for stable AIBs (11, 31, 37). Present studies tend to focus either on rigid polymers for fast electron delocalization or on flexible ones for maximizing active site utilization. However, they are difficult to complement each other to realize both targets simultaneously. Specifically, rigid polymers (e.g., heptazine-based and triazine-based organic frameworks) with extended π-conjugated planar structures afford fast ion transport channels and electron delocalization paths to ensure high electrochemical activity (38–40). As an example, Alshareef group designed rigid quinone-pyrazine covalent organic frameworks to endow fast ion diffusion kinetics, which affords high capacity (220.4 mA·hour g⁻¹) and rate performance (10 A g⁻¹) (17). In contrast, flexible polymers (e.g., biguanide-based and spiral organics) with rotating molecular chains give a versatile platform to reduce the steric hindrance of electroactive aromatic motifs (41–44), thus achieving full utilization of redox-active groups toward high-capacity storage. Recently, Lai group proposed a flexible polymeric diaminophenazine with spiral architectures to give a high active-site utilization of 95% but an unsatisfied rate capacity of 88.36 mA·hour g⁻¹ (42). These advancements broaden the design perspective of polymers to boost battery metrics. Unfortunately, rigid polymers face limited capacity (typically <250 mA·hour g⁻¹) because of part accessible active sites, whereas flexile polymers suffer from unsatisfied large-current tolerance (<10 A g⁻¹) caused by slow electron delocalization.

Logically, the next key step is to leverage the structural advantages of rigid and flexible motifs and compensate for their respective shortcomings. Accordingly, the elaborate selection of rigid and flexible aromatic motifs with customized redox structures and functions as alternative precursors to design rigid-flexible combined polymeric frameworks is an important and ongoing work aimed at maximizing their applications value in the energy storage field (45). Benefiting from highly stable imine covalent bonds, the heptazine unit acts as a rigid and strong electroactive building block, providing a large number of electroactive moieties (38, 39). Its rigid plane can bring rapid electron migration to boost redox kinetics during the (de)charge process. Meanwhile, the biguanide unit, featured with flexible rotated structures, can bring lower steric hindrances to fully expose imine active sites for ion coordination (41, 44). Of note, heptazine/biguanide derivatives have been well documented in medicinal chemistry (46–48), which provides a promising design platform for rigid-flexible organic materials. On the basis of the lessons learned from preliminary studies on the relationship between structures and functions of rigid or flexible systems, we envision integrating rigid heptazine and flexible biguanide motifs into rigid-flexible polymer frameworks via strategic structural engineering. This can be expected to leverage the advantages of rigid and flexible units and compensate for their respective flaws, thereby unlocking fast and stable electron transfer and superior NH₄⁺ storage to reform AIBs, but this has not yet been achieved.

Building on these considerations, rigid-flexible polymeric heptazine-biguanide frameworks (HBFs) are developed, which act as a highly active and stable organic anode for better AIBs. HBFs are designed via the favorable integration of three-electron-accepting planar meleme units and four-electron-accepting rotated 1,6-bis[5-(p-chlorophenyl)biguandino]hexane [chlorhexidine (CH)] blocks. Rigid heptazine units of HBFs afford rapid electron delocalization pathways with an ultralow activation energy (0.15 eV). Meanwhile, flexible biguanide molecular chains decrease structural steric hindrances to ensure full utilization of imine redox-active sites (99.6%). As a consequence, the HBF anode initiates a fast and stable 7 e⁻ NH₄⁺ storage process per heptazine-biguanide unit, affording high capacity, high-rate performance, and desirable cycling stability. Besides, the HBF anode can be further expanded to design state-of-the-art all-polymer AIBs with superior battery-level energy density and cycling life. These findings mark substantial progress in rigid-flexible organic framework materials toward better aqueous batteries.

RESULTS

Figure 1A illustrates the design concept of rigid-flexible HBFs, which integrate three-electron melem units of planar 2,4,6-triazine symmheptazine (TS) and four-electron rotated CH into an extended multiple redox-active polymeric skeleton. Generally, because of the extended π-conjugated rigid planar structure, TS delivers rapid ion transport channels and electron delocalization paths to support high-kinetics redox reactions and enhanced electrochemical activity. However, the low density of accessible redox sites and severe dissolution in aqueous electrolytes of TS result in a low capacity. In contrast, flexible CH with rotated amine linkers allows for high utilization of redox-active sites by reduced steric hindrance within confined spaces but, in turn, leads to slow electron delocalization and unsatisfactory rate performance.

Fig. 1. Schematic fabrication of HBFs.

(A) Schematic fabrication route and (B) MEP distribution of HBFs.

On the basis of these considerations, HBFs were designed via the nucleophilic substitution reaction between the electron-pushing effect of amine groups of TS (donor) and the electron-pulling ability of halogen bonds of CH (acceptor) into well-arranged π-conjugated aromatic structure with aid of H-bonds and π-π interactions. Molecular electrostatic potential (MEP) simulation (49, 50), suggests that electronegative C═N groups in heptazine and guanidinium units of HBFs are redox-active sites (red area) for NH₄⁺ chelation (Fig. 1B), whereas the remaining aromatic domains (blue area) maintain efficient electron localization. Overall, rigid-flexible HBFs leverage the merits of TS (fast electron delocalization) and CH (low steric hindrance) while compensating for their respective flaws, affording full accessible active sites and structural insolubility.

Fourier transform infrared (FTIR) spectroscopy was used to verify the chemical structure of HBFs (Fig. 2A). The absorption peak at 3470 cm⁻¹ corresponds to ─NH₂ bending/ring vibration of TS, whereas the peak at 919 cm⁻¹ is ascribed to the stretching vibration of aromatic C─Cl of CH. The remarkable consumption of C─Cl and ─NH₂ bands suggests the donor-acceptor interaction between TS and CH, confirming that their successful polymerization into HBFs with generated C─N species. The solution ¹H nuclear magnetic resonance (NMR) spectra show a signal at 9.29 parts per million (ppm), suggesting the formation of ─NH─ species in HBFs in contrast with TS and CH monomers (fig. S1, A to C). Moreover, the solid-state ¹³C NMR spectrum of HBFs shows the signal of C─N at 162.1 ppm (fig. S1D), which derives from the remarkable consumption of C─Cl in CH and ─NH₂ in TS. These results confirm the structure of HBFs.

Fig. 2. Structure characterization of HBFs.

(A) FTIR spectra. (B) E_g values. (C) RDG scattered points against sign(λ₂)ρ. (D) XRD patterns. (E) SEM image. (F) TEM image. (G) UV-Vis spectra of CH-, TS-, and HBF-soaked NH₄OTF/H₂O electrolytes for 1 month. (H) Electrical conductivities. (I) Nitrogen sorption isotherm and pore size distribution curve.

Thanks to the extended π-conjugated rigid-flexible aromatic structures, HBFs achieve a low optical energy gap (E_g) of 2.74 eV (Fig. 2B) compared to CH (3.12 eV) and TS (2.92 eV), liberating high inherent electronical conductivity to propel redox reaction with low energy barriers (39, 51). Reduced density gradient (RDG) analysis of HBFs (52, 53) as conducted to investigate the intramolecular forces in HBFs (Fig. 2C and fig. S2). Blue spike signals positioned at −0.04~−0.02 arbitrary units (a.u.) and green spike regions located at −0.02~0.01 a.u. of the sign(λ₂)ρ denote powerful intramolecular H-bonding and π-π stacking, due to the large π-conjugated rigid-flexible structure. The crystallinity of HBFs was validated by powder x-ray diffraction (XRD) analysis (Fig. 2D). Three distinguishable peaks at 6.1°, 12.3°, and 27.4°, assigned to the (100), (210), and (001) plane, agree with the simulated eclipsed AA stacking model (inset of Fig. 2D). The high crystallinity of HBFs is beneficial for achieving high conductivity (54). Of note, a distinguishable peak at 27.4°, corresponding to the (001) facet, can be attributed to strong π-π interactions within rigid-flexible HBFs (Fig. 2, E and F, and fig. S3).

Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images demonstrate the spherical morphology of HBFs with a d spacing of 0.32 nm (Fig. 2, E and F, and fig. S4). Compared with soluble CH and TS, HBFs have no ultraviolet-visible (UV-Vis) absorption signal in aqueous NH₄OTF solution (Fig. 2G), demonstrating their strong antidissolution ability and structural ruggedness. For conductivity test, HBFs were mixed with acetylene black and polytetrafluoroethylene in a mass ratio of 6:3:1. The extended π-conjugated rigid-flexible aromatic structures endows the HBF anode with an efficient electron delocalization path, thereby bringing a higher conductivity of 25.4 S cm⁻¹ (Fig. 2H and fig. S5) than those of TS/CH small molecules (15.4/6.5 S cm⁻¹). The markedly enhanced conductivity of the HBF anode favors a fast electron transfer process to facilitate high-kinetics and stable redox reactions (figs. S6 and S7). Besides, HBFs display a composite type I isotherm (Fig. 2I) with a high surface area of 185 m² g⁻¹ and a concentrated porous size of 3.25 nm (consistent with a pore diameter of 3.28 nm predicted by the XRD model; Fig. 2D). Overall, rigid-flexible HBFs integrate fast electron delocalization, multiredox accessible sites, low steric hindrance, and structural insolubility, which are desirable features for advanced AIBs.

The electrochemical performance of HBFs as the anode was evaluated in a three-electrode Swagelok cell with Ag/AgCl as the reference electrode, activated carbon as the counter electrode, and 3 M NH₄OTF/H₂O solution as the electrolyte (OTF⁻ = CF₃SO₃⁻). Galvanostatic charge/discharge (GCD) curves (Fig. 3A) display a high capacity of 314 mA·hour g⁻¹ at 0.2 A g⁻¹ for the HBF anode, surpassing TS (121 mA·hour g⁻¹) and CH (151 mA·hour g⁻¹) anodes (fig. S8). Compared with soluble CH and TS molecules showing inferior cycling stability (fig. S8), the HBF anode with extended π-conjugated rigid-flexible aromatic skeleton deliver excellent antidissolution ability in the NH₄OTF/H₂O electrolyte, which thus contributes to ultradurable redox activity with a high-capacity retention after persistent cycling (Fig. 3F). When the current density is increased to 60 A g⁻¹, the HBF anode still maintains a high capacity of 101 mA·hour g⁻¹ (Fig. 3B), emphasizing its outstanding large-current performance. Notably, as the current density goes back to 0.2 A g⁻¹, the capacity can still be restored to 313 mA·hour g⁻¹, underscoring the structural robustness and flexibility of HBFs. Cyclic voltammetry (CV) profiles of the HBF anode exhibits two pairs of redox peaks (Fig. 3C), suggesting a two-step redox reaction. A comparison of the rate performance with other reported NH₄⁺-host materials is summarized in Fig. 3D (2, 9, 16–18, 30, 38, 42). Obviously, the HBF anode exhibits the highest rate performance, which is highly competitive with other organic materials. Furthermore, the HBF anode shows an outstanding life span with 99% capacity retention after 1000 cycles at 0.2 A g⁻¹ (Fig. 3E) and exhibits an extraordinary life span of 60,000 times with 86.7% capacity retention at 10 A g⁻¹ (Fig. 3F), surpassing recently reported organic materials (Fig. 3G and table S1) (2, 9, 16, 18, 31, 37, 55–63).

Fig. 3. Electrochemical performances of the HBF anode.

(A) GCD profiles. (B) Rate performance. (C) CV profiles. (D) Rate performance comparison with reported organic materials. (E) Cyclic stability. (F) Long-term cycling performance. (G) Life comparison of HBFs and reported organic materials. (H) Calculated E_a values.

In addition, postcycling SEM image characterization confirms the superior structural and functional stability of the HBF anode (fig. S9), which can be attributed to its rigid-flexible nature that effectively accommodates volume changes during cycling. Moreover, electrochemical impedance spectroscopy (EIS) of the HBF anode at different temperatures (T) were performed to determine the interfacial charge transfer resistance (R_ct) and to estimate the activation energy (E_a) based on the Arrhenius equation (fig. S10). A lower E_a value signifies a lower energy barrier, facilitating fast redox reactions and efficient charge storage (64–67). The E_a value for charge storage process in the HBF anode is calculated to be 0.15 eV (Fig. 3H), which is much lower than CH (0.36 eV) and TS (0.34 eV) anodes, indicating superior reaction kinetics. The self-discharge characteristics of the HBF anode was evaluated at the fully charged state of 0.4 V at 0.2 A g⁻¹, displaying a high-capacity retention of 95% after resting 7 days (fig. S11). This result suggests the desirable structural stability and practical applicability of the HBF anode in the NH₄OTF/H₂O electrolyte. Overall, the HBF anode with rigid-flexible HBFs leverages the merits of fast electron delocalization, low steric hindrance, and structural insolubility, allowing for high accessibility of multiple redox-active sites to enable state-of-the-art capacity, rate performance, and long-term cycling stability.

The redox charge storage kinetics of the HBF anode were further investigated. CV curves at various scan rates ranging from 1 to 5 mV s⁻¹ display two pairs of redox peaks with well-maintained shape and minimal polarization (Fig. 4A), indicating stable electrochemical behavior and low ion transport barriers. The kinetics were analyzed using Dunn’s method (36, 68) based on the power-law relationship i = kv^b, where k is a constant and the b value (typically 0.5 to 1.0) reflects the charge storage mechanism. The calculated b values for the oxidation (P_O1/P_O2) and reduction (P_R1/P_R2) peaks of HBFs are determined to be 0.97/0.98 and 0.96/0.95 (Fig. 4B), implying a mixed charge storage process involving both diffusion-controlled and capacitive-controlled mechanisms. The capacitive contribution was quantitatively assessed, revealing a high proportion of 86% at 1 mV s⁻¹ for the HBF anode (fig. S12). With the increase in the scan rate, the capacitive contribution increasingly dominates from 86 to 94% (Fig. 4C), surpassing the diffusion-controlled process. These results corroborate the high-kinetics and stable energy storage behavior of the HBF anode, which originates from its rigid-flexible π-conjugated structure to provide fast electron delocalization and low steric hindrance, making for full utilization of active motifs and rapid ion migration. Besides, the rigid-flexible framework design can be extended to triazine-biguanide polymers with desirable electrochemical advantages (fig. S13), which widens the philosophy of versatile rigid-flexible organic materials.

Fig. 4. Charge storage mechanism of the HBF anode in AIBs.

(A) CV profiles at different scan rates. (B) Calculated b values. (C) Ratios of capacitive and diffusion-controlled contribution at various scan rates. (D) GCD curve. (E) Overview of FTIR spectra. (F) XPS spectra of N 1s. (G) Discharge storage mechanism of the HBF anode.

To elucidate the charge mechanism of the HBF anode, FTIR spectroscopy and x-ray photoelectron spectroscopy (XPS) were conducted to investigate its structural evolution at various (dis)charge voltages. The two distinct voltage plateaus observed in the GCD profile correspond to the two-step sequential energy storage process of the HBF anode (Fig. 4D). Regarding FTIR spectra (Fig. 4E), the peak intensity of the C═N bond in the guanidinium unit (1438 cm⁻¹) decreases (state A→C) and remains unchanged (state C→E) during discharging, accompanied by the increase in C─N vibration (1232 cm⁻¹), signifying the high electrochemical activity of C═N motifs in HBFs. In addition, a new peak emerges at 2923 cm⁻¹ during discharging, assigned to N─H species from NH₄⁺ charge carriers, confirming that the coordination between C═N sites of the HBF anode and NH₄⁺ ions. In contrast, the peak intensity of C═N species of heptazine in TS (1529 cm⁻¹) is nearly unchanged during the initial discharge (state A→C) and gradually decreases upon further charging (state C→E). Notably, a new signal appears at 2956 cm⁻¹ (state B→C→E), attributed to the generation of N─H bands of NH₄⁺ ions, unraveling the coordination between NH₄⁺ ions and C═N sites. After charging (state E→I), all redox-active species return to their original levels due to the decoordination of NH₄⁺. These results demonstrate that C═N groups of both heptazine and guanidinium motifs in HBFs participate in the reversible electrochemical redox reactions with NH₄⁺ ions.

Furthermore, high-resolution XPS of the N 1s region was conducted to study the H-bonding behavior between NH₄⁺ charge carriers and C═N sites in the HBF anode during cell operation (Fig. 4F). The deconvoluted peaks at 400.1 and 399.0 eV are assigned to C═N and C─N bonds, respectively. The concentration of C═N moieties decreases during discharging (state A →E), accompanied by the increase in C─N species, confirming the reversible transformation between C═N and C─N during discharging-charging. Simultaneously, a newly emerged signal at 400.3 eV is associated with the formation of H-bonding positive center of (N─H···)⁺ corresponds to the coordination of NH₄⁺ ions. Clearly, the concentration of C═N decreases during the discharge process (state A→E) and increases during the charge process (state E→I), whereas the opposite concentration trend is observed for the C─N bond. These variations in concentration align well with FTIR results, confirming the interaction between NH₄⁺ ions and C═N sites. After charging (state E→I), all species revert to their initial levels due to the decoordination of NH₄⁺ from redox-active sites. This indicates the reversible electrochemical redox reactions between C═N groups in the HBF anode and NH₄⁺ ions. Besides, the structural robustness and antidissolution properties of the HBF anode are evidenced by the lack of detectable UV-Vis signals. The NH₄OTF/H₂O electrolyte remains colorless after being soaked with the HBF anode at different voltage states (figs. S14 to S16).

Considering the mild acidity of the 3 M NH₄OTF/H₂O electrolyte (pH ≈ 4.85), the contribution of H⁺ ions to the energy storage of the HBF anode was studied in the HOTF/H₂O electrolyte (with the same pH value as the 3 M NH₄OTF/H₂O electrolyte). The HBF anode using the HOTF/H₂O electrolyte delivers different electrochemical behaviors with a very low capacity (20 mA·hour g⁻¹; fig. S17), which can be negligible compared to the total capacity of 314 mA·hour g⁻¹ (Fig. 2A). These results confirm that NH₄⁺ is the dominant and effective charge carrier, which is responsible for the electrochemical redox reaction of the HBF anode in the NH₄OTF/H₂O electrolyte. When deducting the insignificant capacity from the H⁺ storage (20 mA·hour g⁻¹; fig. S17) and conductive agent (17 mA·hour g⁻¹; fig. S18) from the experimental capacity (314 mA·hour g⁻¹; Fig. 3A), the actual capacity of HBFs is determined to be 277 mA·hour g⁻¹, corresponding to 7 e⁻ redox reactions per heptazine-biguanide motifs, with an actual use efficiency of 99.6% for redox-active sites. Electrochemical analysis and spectral results confirm that the HBF anode initiates a fast and stable 7 e⁻ NH₄⁺ coordination mechanism per heptazine-biguanide module, involving four C═N sites of biguanide coordinating with four NH₄⁺ ions first, followed by three C═N sites of heptazine coupling with three NH₄⁺ ions (Fig. 4G). These advantages are attributed to the favorable rigid-flexible structure of HBFs, which enables a low energy barrier and high active-site utilization, underscoring its potential as a robust anode for advanced energy storage.

The electronic structure of HBFs was examined via density functional theory (DFT) calculation to elucidate its energy levels and redox behavior. It is well known that the extended π-π stacking conjugated structure of organic materials typically leads to reduced levels of both the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO) (67). The calculated LUMO-HOMO gap (ΔE) of HBFs is 2.97 eV (Fig. 5A), much lower than those of TS/CH (4.24/6.48 eV; fig. S19), which indicates a low-energy-barrier pathway for rapid electron transfer, stemming from a highly π-conjugated electronic configuration. Furthermore, the π-electron localization function (ELF-π) confirms a highly π-conjugated structure throughout the aromatic backbone in HBFs, promoting efficient electron delocalization (Fig. 5B).

Fig. 5. Theoretical calculation of electrochemical reaction behaviors of the HBF anode.

(A) Frontier molecular orbitals and energy levels. (B) ELF-π map of HBFs. (C) Optimized geometries and corresponding ΔE values after NH₄⁺ uptake. (D) ACID plots. (E) Charge density difference isosurfaces and corresponding Bader charges.

The optimized structures of the HBF anode upon two-step NH₄⁺ (de)coordination were investigated. In the preliminary discharge process (step 1), the coordination of four NH₄⁺ ions to four C═N sites of a biguanide unit is spontaneous, as indicated by a negative Gibbs free energy (ΔG) value of −17.91 eV (Fig. 5C and fig. S20). In the following discharge process (step 2), the coordination of three additional NH₄⁺ ions to the C═N motifs of heptazine units is also spontaneous with a negative ΔG value (−11.85 eV). The more negative ΔG value in step 1 compared to step 2 indicates that the two steps proceed sequentially. According to the minimum energy principle, a steady two-stage 7 e⁻ coordination reaction occurs in the HBF anode, belonging to two distinct discharge plateaus (Fig. 3C). The global π-aromaticity of HBFs is evidenced via anisotropy of the induced current density (ACID) calculation (63, 68), revealing a clockwise diamagnetic current circulating the throughout molecular framework (Fig. 5D). Bader charge analysis monitors the charge accumulation/consumption between NH₄⁺ charge carriers and C═N sites (Fig. 5E), confirming their favorable interactions to generate stable structures with notable charge transfer (2.17 e of state I; 4.83 e of state II). Moreover, the fast kinetics of NH₄⁺ charge carriers maximizes the utilization of C═N sites in HBFs, thereby enhancing the highly active and durable 7 e⁻ NH₄⁺ charge storage process in AIBs.

Considering the advantages of high capacity, desirable rate performance, and cycling stability of HBFs, HBFs as the anode, polyaniline (PANI) as the cathode (69), and 3 M NH₄OTF/H₂O as the electrolyte were coupled to construct an HBFs||PANI full battery (Fig. 6A). PANI was chosen as the cathode material mainly because of its desirable working potential, high capacity, and good conductivity (figs. S21 and S22). As shown in Fig. 6B, the HBF anode demonstrates a stable operating potential window of −1.0 ~ 0.3 V and the PANI cathode exhibits a complementary potential range of 0.0 ~ 0.9 V (versus Ag/AgCl). GCD curves of the HBFs||PANI battery displays discharge capacities of 300 mA·hour g⁻¹_HBFs (Fig. 6C; based on the mass loading of HBFs in the anode) and 117 mA·hour g⁻¹_HBFs+PANI (based on the total mass loading of the HBF anode and the PANI cathode) at 0.2 A g⁻¹. Even at a high current density of 60 A g⁻¹, a capacity of 145 mA·hour g⁻¹_HBFs (56 mA·hour g⁻¹_HBFs+PANI) is still retained, demonstrating excellent rate capability and electrochemical reversibility over 120 cycles (Fig. 6D).

Fig. 6. Electrochemical performances of the HBFs||PANI battery.

(A) Schematic diagram of the HBFs||PANI battery. (B) CV profiles of the HBF anode and the PANI anode. (C) GCD curves. (D) Rate metrics. (E) Voltage-capacity contour plots. (F) Cycling stability. Inset: A pouch cell powers colored lights. (G) Comparison of electrochemical performance indexes of the full battery with recently related reports.

The high capacity (117 mA·hour g⁻¹) and high average output voltage (0.86 V) bring a state-of-the-art battery-level energy density of 100.6 watt-hours (Wh) kg⁻¹ (based on the total mass loading of the HBF anode and the PANI cathode), which is the highest value among reported AIBs (Fig. 6E and table S2) (9, 18, 27, 30, 38, 46, 55, 70–75). Notably, the HBFs||PANI battery exhibits extraordinary cycle life of 120,0000 cycles with ~80.1% capacity retention at 10 A g⁻¹ (Fig. 6F and fig. S21). The Coulombic efficiency remains consistently near 100% throughout cycling, without no noticeable decay. To assess practical applicability, the HBFs||PANI pouch cell was fabricated (Fig. 6F), which can power multiple small colored lights, demonstrating real-world functionality. The radar chart further highlights the performance superiority of the HBFs||PANI full battery (Fig. 6G), including high capacity, Coulombic efficiency, superior rate performance, and ultralong cycle life.

DISCUSSION

In conclusion, rigid-flexible HBFs are designed as a high-performance anode for AIBs. Rigid heptazine units of HBFs afford fast electron delocalization paths with an ultralow activation energy (0.15 eV), whereas flexible biguanide molecular chains reduce steric hindrances to allow for ultrahigh utilization of imine active sites (99.6%). The HBF anode thus starts high-kinetics and stable multielectron NH₄⁺ coordination mechanism per heptazine-biguanide module, delivering ultrahigh capacity, large-current survivability, and excellent cycling stability. The structural merits of the HBF electrode are further highlighted in state-of-the-art all-organic AIBs with superior battery-level energy density and unprecedented cycle life. This study paves a previously unknown path for developing various rigid-flexible organic framework materials toward advanced energy storage.

MATERIALS AND METHODS

Material synthesis

CH (101 mg, 0.2 mmol) and TS (220 mg, 0.8 mmol) were dissolved in 4.0 ml of dioxane, 4.0 ml of toluene, and 1.2 ml of water to form a transparent and uniform mixture solution. The mixture was subsequently stirred for 30 min under N₂ atmosphere and sealed under vacuum and heated at 120°C for 72 hours. After filtration, washing, and vacuum drying, the precipitate products were obtained (denoted as HBFs). The resulting HBF precipitate was isolated by vacuum filtration and washed three times with ethanol and methanol and then dried at 60°C under vacuum for 24 hours to give a pale yellow product.

Characterizations

FTIR spectroscopy and NMR spectroscopy were carried out to study the sample structures. SEM (Hitachi S-4800) and TEM (JEM-2100) were carried out to study the microstructures. XRD analysis was performed to observe the chemical structures using a Bruker D8 advance powder diffractometer (Cu Kα radiation source). The nitrogen sorption isotherm was collected on a Micromeritics ASAP 2460 analyzer at −196°C. The surface area and pore size distribution were estimated by the Brunauer-Emmett-Teller method and a nonlocal DFT equilibrium model. The UV-Vis spectra were collected using a UV-Vis spectrometer (JASCO V-750). The thermal stability was monitored on a STA409 PC thermogravimetric analyzer in nitrogen atmosphere at a heating speed of 10°C min⁻¹.

The electrical conductivity of the sample was measured using an RTS-8 four-point probe. The four-point probe method involves placing four probes on the sample surface, two of which are used to apply current and the other two are used to measure voltage drop. This method can eliminate the influence of contact resistance, thereby more accurately measuring the intrinsic conductivity of materials. Specifically, the HBF anode was prepared by mixing HBFs, acetylene black conductive additive, and polytetrafluoroethylene with a mass ratio of 6:3:1. It was placed in the round mold (1 cm in diameter) and then was pressed into a thin sheet (diameter: 1 cm; thickness: 0.1 cm; bulk density: ~0.75 g cm⁻³) for electrical conductivity test on the RTS-8 four-point probe. The elemental maps were observed on a JEM-F200 instrument equipped with an energy diffraction system. XPS (AXIS Ultra DLD) was performed to study the surface chemistries of the sample. For ex situ spectroscopic tests including FTIR, XPS, and SEM, organic anodes were collected by disassembling batteries at specific (dis)charging voltages.

Electrochemical tests

To fabricate organic anodes, a mixture of HBFs (or CH and TS), acetylene black, and polytetrafluoroethylene (with a mass ratio of 6:3:1) was pressed onto the stainless steel mesh. Then, organic anodes were dried at 60°C for 12 hours under a vacuum condition. The mass loading of HBFs on the anode is 3.5 mg cm⁻². Electrochemical measurements of the HBF anode were performed in a three-electrode Swagelok cell at room temperature, using Ag/AgCl as the reference electrode, activated carbon as the counter electrode, and 3 M NH₄CF₃SO₃/H₂O as the electrolyte. To assemble the ammonium-ion full battery, HBFs (mass loading: 3.5 mg cm⁻²) and PANI were used as the anode and cathode (negative/positive ratio: 1.56) by using the 3 M NH₄CF₃SO₃/H₂O electrolyte (135 μl) and glass fiber separator (Whatman). The theoretical capacity (C_m, mA·hour g⁻¹) of HBFs was determined on the basis of the equation of C_m = nF/(3.6 × M), where n is the electron transfer number (18) during the redox process, F is a constant (96,485 C mol⁻¹), and M is the molar weight (1731 g mol⁻¹). The theoretical capacity of HBFs was calculated to be 278 mA·hour g⁻¹.

Molecular property simulation

DFT calculations were carried out with the Gaussian 16 software package (76–81). Geometrical optimizations were performed at the B3LYP-D3/def2-SVP level of theory. The water solvent was included in the calculations using the solvation model based on the density model. The ELF-π calculation was obtained via the Multiwfn 3.8 program.

Supplementary Materials

This PDF file includes:

Sections S1 to S4

Figs. S1 to S23

Tables S1 and S2