Lithium Ion‐Conducting Triazacoronene‐Based Neutral 2D Covalent Organic Framework as a Solid‐State Electrolyte

ABSTRACT

Solid‐state electrolytes (SSEs) are critical for advancing lithium‐ion battery (LIB) technology, as traditional liquid electrolytes compromise their safety and lifespan. Polymer electrolytes often lack adequate ionic conductivity, Li⁺ transference number, and the ability to prevent dendrite formation, whereas inorganic electrolytes usually suffer from poor processability and electrode wettability despite high ionic conductivity and mechanical strength, necessitating new SSE materials. Composed of tailor‐made organic building blocks that can bind and transport Li⁺ ions, crystalline, porous covalent organic frameworks (COFs) offer the best of both worlds. Here, we present a new, neutral hexagonal 2D COF, HMTAC‐COF1, constructed from a heteroatom‐rich hexamethoxy‐triazacoronene (HMTAC) building block with multiple Li⁺ binding sites, which, upon Li⁺‐doping exhibits impressive ionic conductivity (4.95 × 10⁻⁵ S/cm at 20°C and 1.95 × 10⁻⁴ S/cm at 120°C), low activation energy (0.25 eV), and a high transference number (0.57). When employed as an SSE in a Li || Li⁺@HMTAC‐COF1 || LiFePO₄ coin cell, it displayed a promising specific capacity (181 mAh/g) with >99% Coulombic efficiency over multiple cycles. These results demonstrate the potential of heteroatom‐rich neutral COFs as SSEs for next‐generation LIBs.

A new 2D COF based on hexamethoxy triazacoronene (HMTAC) exhibits impressive ionic conductivity (10⁻⁴ S/cm), activation energy (0.25 eV), Li⁺ transference number (0.57), as well as promising specific capacity of 181 mAh/g with >99% Coulombic efficiency when employed as a solid‐state electrolyte in a Li || Li⁺@HMTAC‐COF1 || LiFePO₄ cells, outperforming most neutral COFs developed to date.

1. Introduction

The widespread use of rechargeable lithium‐ion batteries (LIBs) in portable electronics, electric vehicles, and grid‐scale energy storage systems has heightened the need to improve their safety, performance, and longevity. While there is room for improvement in all components of LIBs, i.e., anode, cathode, and electrolyte, the widespread use of liquid electrolytes (LEs) that are flammable, corrosive, volatile, and leakage‐prone organic solvents poses the greatest safety hazards and performance limitations [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]. Therefore, it is imperative to develop efficient solid‐state electrolytes (SSEs) with high ionic conductivity (≥10⁻⁴ S/cm) and transference number (t _Li+ ≥ 0.5), adequate mechanical strength and thermal stability, straightforward processability, and good electrode wettability (i.e., the ability to form good interfacial contact with the electrodes) that can replace unsafe LEs from LIBs [11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21]. Although solid inorganic electrolytes (SIEs) based on hard metal oxides, ceramics, sulfides, and halides often possess high Li⁺ conductivity and transference number, excellent mechanical strength, and resistance to dendrite formation, they also suffer from complex processing and poor electrode wettability [22, 23, 24]. On the other hand, solid polymer electrolytes (SPEs) are easier to process and provide better electrode wettability, but their lower Li⁺ conductivity and transference number, poor mechanical stability, and most critically, the inability to prevent lithium dendrite formation, which can lead to an internal short circuit, diminish the battery performance, safety, and cyclability [25, 26, 27, 28].

Composed of covalently linked organic building blocks with diverse physical, chemical, and molecular recognition properties, covalent organic frameworks (COFs) [29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42] are known for their robust, periodic, porous structures capable of binding, storing, and delivering various guest entities, including alkali and alkaline earth metal cations. These features render COFs a versatile platform for advanced technologies, such as energy storage systems and rechargeable batteries [43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53]. While the electrode materials of rechargeable batteries must have high electrical conductivity, the electrolytes and separators sandwiched between the electrodes must be electrical insulators but ionic conductors. Pristine COFs typically exhibit poor intrinsic electrical conductivity due to the lack of efficient electron transport pathways, but they can bind Li⁺ (and other) cations at designated recognition sites on the organic building blocks and facilitate cation transport through the ordered pores and channels adorned with these binding sites, thus exhibiting useful ionic conductivity. Furthermore, endowed with excellent thermal stability, mechanical strength, malleability, and the ability to create good electrode contact, inhibit lithium dendrite formation, and operate over a wide temperature range, COFs present the best features of both worlds as solid‐state electrolytes and separators [54, 55].

To harness their potential benefits as SSEs, several anionic, cationic, and neutral COFs have been developed recently [56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73], which exhibit a wide range of Li⁺‐ion conductivity and transference numbers commensurate with their Li⁺ ion binding and transport capabilities. For instance, anionic COFs furnished with sulfonate, carboxylate, and imidazolate groups can bind Li⁺ cations and enable Li⁺ hopping between these sites [59, 62, 63, 71], whereas cationic frameworks based on imidazolium, guanidinium, and pyridinium groups containing free counter‐anions can putatively better solvate the infiltrated Li⁺ cations. As a result, the ionic COFs usually display high ionic conductivity (10⁻⁵—10⁻⁴ S/cm) and transference numbers (t _Li+ ≥ 0.5) than neutral frameworks [64, 74, 75]. However, the ionic COFs have limited chemical diversity and tunability because only a few charged groups can be incorporated into COFs through either direct solvothermal synthesis or post‐synthetic modifications without adversely affecting the COF formation and structural stability. On the other hand, neutral COFs are chemically and structurally more versatile, but they usually lack sufficient Li⁺ binding sites and transport pathways, which cause lower ionic conductivity and transference numbers compared to the ionic COFs [56, 57, 67, 69, 70, 73]. One approach to overcome these limitations of neutral COFs is to imbue them with Li⁺‐doped polyethylene oxide (PEO) to create Li⁺‐COF@PEO composites [57]. Another strategy involves building neutral COFs with heteroaromatic building blocks containing Lewis basic sites (N, O, and S) that can facilitate Li⁺ binding and transport. Although the PEO‐functionalized neutral COFs display reasonable ionic conductivity at elevated temperatures (≥ 10⁻⁵ S/cm above 100°C), they usually have poor room temperature ionic conductivity (≤10⁻⁸ S/cm) and transference numbers [70], due to strong Li⁺ chelation by multidentate PEO chains and restricted dynamics of long PEO chains, which hinder Li⁺ hopping through the COF pores.

Since the materials' ionic conductivity depends on their charge carrier concentration and ion mobility [17, 22, 27, 76, 77, 78], and the Li⁺ transference number indicates the fraction of total migration current carried by mobile Li⁺ ions, a proper balance must be struck between these two parameters. For instance, while strong Li⁺ binding to the COF walls is necessary for sufficient ion uptake, too tight binding could slow Li⁺ movement through the channels, diminishing the ionic conductivity and transference number. Therefore, for SSE applications, it is essential to construct COFs that can not only bind Li⁺ ions strongly enough to attain sufficient ionic concentration but also facilitate the captured Li⁺ ions to hop from one binding site to another. In addition, the COF should be resistant to oxidation and reduction to ensure a wide electrochemical stability window (≥3 V vs. Li/Li⁺) and meet another key requirement for SSEs [43, 46, 50].

Taking these criteria into account and building on our experience on electrically and ionic conducting metal–organic frameworks [79, 80, 81, 82, 83], herein, we introduced a new trigonal planar hexamethoxy triazacoronene building block equipped with three 4‐aminophenyl arms (TAHMTAC), which yielded a new, neutral, hexagonal 2D COF, namely, HMTAC‐COF1 (Figure 1), via Sc(III)‐catalyzed imine condensation with a linear terephthalaldehyde (TPA) linker (Scheme 1) [84, 85, 86]. Although triazacoronene (TAC)‐based COF formation via one‐step, acid‐catalyzed Pictet‐Spengler cyclization reactions had been claimed in the literature [87, 88, 89], no concrete experimental evidence (e.g., NMR and X‐ray data) was found to confirm the formation of TAC corners in those frameworks, as incomplete cyclization, which is quite common during this reaction, likely produced many defect sites instead of the desired TAC corners (vide infra). Thus, our HMTAC‐COF1 with a verified structure and composition represents the first of its kind. Each HMTAC corner contains three sets of ortho‐dimethoxy groups (a total of 6 OMe groups) that can help capture and transport Li⁺ through the hexagonal channels of the resulting COF. Consequently, upon soaking in a Li‐bistriflamide (LiTFSI) solution, the Li⁺‐doped HMTAC‐COF1, hereafter called Li⁺@HMTAC‐COF1, displays promising ionic conductivity across a wide temperature range (4.85 × 10⁻⁵ S/cm at 20°C, 1.95 × 10⁻⁴ S/cm at 120°C), a low thermal activation energy (E _a = 0.25 eV), and an impressive transference number (t _Li+ = 0.57). All these parameters are superior to most neutral COFs developed to date and almost on par with many ionic COFs (Table S1). When incorporated as an SSE into CR2032‐type Li || Li⁺@HMTAC‐COF1 || LiFePO₄ coin cells, it exhibits electrochemical stability up to 3.7 V, as well as a high specific capacity of approximately 181 mAh/g with Coulombic efficiency over 99%. Thus, this study presents the design, synthesis, characterization, ionic conductivity, and electrochemical performance of a new, neutral hexagonal 2D COF based on a HMTAC building block equipped with multiple Li⁺ ion binding sites, illustrating the potential of neutral COFs to serve as SSEs in next‐generation LIBs.

FIGURE 1.

The DFT‐optimized structure of the Li⁺‐doped HMTAC‐COF1 shows that Li⁺ ions are preferentially coordinated to the OMe groups.

SCHEME 1.

Design and synthesis of HMTAC‐COF1 from pre‐synthesized TAHMTAC and terephthalaldehyde via Sc(OTf)₃ catalyzed imine condensation. Inset: Photographs of HMTAC‐COF1 precursor solution without the catalyst (left) and orange precipitate (right) 2 h. after the catalyst addition. The SCXRD structures of TAHMTAC (top and bottom right corners) show large dihedral angles (∼60°) between the HMTAC core and phenyl sidearms projected at 120° angles with respect to each other.

2. Results and Discussion

In principle, acid‐catalyzed Pictet‐Spengler cyclization reactions between aromatic triamines such as 1,5,9‐triamino‐triphenylene (TATP) and TAHMTP, and linear dialdehydes can produce triazacoronene (TAC)‐based COFs in one step [87, 88, 89]. However, model Pictet‐Spengler reactions between triamino‐triphenylenes (such as TATP and TAHMTP) and monoaldehydes yield the corresponding triazacoronene derivatives with only 50%–75% yield [90, 91], with the rest being incomplete, mono‐ and di‐cyclized compounds and other unknown impurities. Since similar incomplete cyclization is inevitable during one‐step TAC‐based COF synthesis via this route, the resulting amorphous materials are likely to contain a large number of defect sites instead of the desired TAC corners, which can diminish their structural order and integrity. Therefore, like previous reports, which claimed TAC‐based COF formation via one‐step acid‐catalyzed Pictet‐Spengler cyclization reactions, but failed to present any concrete evidence (e.g., NMR and X‐ray) confirming the presence of TAC corners [87, 88, 89], our attempts to synthesize HMTAC‐based hexagonal 2D COFs directly from TAHMTP and terephthalaldehyde via this route yielded amorphous solids that lacked concrete evidence of TAC formation. Therefore, to eliminate the uncertainties regarding the structure and composition of our target COF, we introduced a preformed and fully characterized tri‐(4‐aminophenyl)‐HMTAC (TAHMTAC) building block to synthesize the first confirmed HMTAC‐based COF.

The TAHMTAC building block was synthesized (Scheme 1) via a Pictet‐Spengler cyclization reaction between triamino‐hexamethoxytriphenylene (TAHMTP) and 4‐nitrobenzaldehyde precursors, which yielded 2,3,6,7,10,11‐hexamethoxy‐4,8,12‐tri(4‐nitrophenyl)‐1,5,9‐triazacoronene (TNHMTAC) [90, 91], followed by a Pd‐catalyzed reduction of TNHMTAC to obtain TAHMTAC as a bright yellow solid. The ¹H and ¹³C NMR spectroscopies (Figures S1 and S2) and MALDI‐ToF analyses confirmed the formation of the TAHMTAC, while single‐crystal X‐ray diffraction (SCXRD) analysis revealed its structure (Scheme 1) ¹ featuring a planar triazacoronene core furnished with six OMe groups that can bind Li⁺ ions and three aminophenyl side‐arms projected at ∼120° angles relative to each other that can form the imine bonds with linear dialdehyde linkers to afford hexagonal 2D COFs. The dihedral angle between the triazacoronene core and phenyl sidearms is ∼60°, which played an important role in defining the preferred stacking pattern of the resulting HMTAC‐COF1.

To synthesize the desired HMTAC‐COF1 (Scheme 1) via imine bond formation between the 3‐fold symmetric TAHMTAC corner and a linear terephthalaldehyde (TPA) linker in a 2:3 molar ratio, we tested different reaction conditions using CH₃COOH, CF₃COOH, and Sc(TfO)₃ catalysts that are commonly used for imine‐linked COF formation [30, 92, 93]. Although protic acid‐catalyzed reactions between TAHMTAC and TPA at various temperatures (20–100°C) failed to produce any crystalline COF, upon addition of a catalytic amount of Sc(TfO)₃ (0.02 equiv. per aldehyde group) [84] to a TAHMTAC and TPA solution (2:3 molar ratio in 4:1 dioxane/mesitylene), the clear yellow reaction mixture quickly turned turbid and produced a large amount of orange solids within 2 h, indicating the COF formation (Scheme 1, inset). The resulting precipitate was collected by centrifugation, washed thoroughly with dioxane and MeCN, and dried under vacuum, to obtain HMTAC‐COF1 as a microcrystalline orange powder. The structure and composition of this material were characterized by powder X‐ray diffraction (PXRD), FT–IR, TGA, N₂ sorption, SEM‐EDX, high‐resolution TEM (HRTEM), and elemental analysis, while density functional theory (DFT) calculations provided valuable insights into its optimized structures (i.e., the preferred stacking pattern) and Li⁺ ion binding modes. Unlike previous attempts to synthesize triazacoronene‐based COFs via one‐pot Pictet‐Spengler reaction directly from TAHMTP precursor that yielded amorphous materials with unverifiable composition due to incomplete cyclization, which introduces structural defects [87, 88, 89], our synthetic approach employing a pre‐made TAHMTAC building block ensured the formation of a new COF with confirmed, defect‐free HMTAC corners.

The first experimental evidence of HMTAC‐COF1 formation came from FT–IR spectroscopy (Figure 2a), which showed the disappearance of characteristic amine (N─H stretch: 3460 cm⁻¹) and aldehyde (C═O stretch: 1682 cm⁻¹ and carbonyl C─H stretch: 2800 cm⁻¹) signals of TAHMTAC and TPA, respectively, along with the emergence of new imine groups (C═N stretching: 1697 cm⁻¹) of the product. The composition of HMTAC‐COF1, i.e., the relative populations of the two building blocks in the framework, was further confirmed by ¹H NMR spectrum (Figure S3) of the digested material (dissolved in DMSO‐d ₆ in the presence of a drop of CF₃CO₂D, which hydrolyzed the imine bonds, regenerating the building blocks), which displayed the characteristic signals of TAHMTAC and PTA with peak integrals consistent with a 2:3 molar ratio, as expected for a 2D hexagonal HMTAC‐COF1 structure. Furthermore, the elemental analysis data of HMTAC‐COF1 (C = 72.66 (obs.), 75.74 (calc.); H = 4.41 (obs.), 4.35 (calc.); N = 9.25 (obs.), 9.30 (calc.) wt.%) were consistent with a 2D hexagonal structure having a unit cell formula of C₁₁₄H₇₈N₁₂O₁₂.

FIGURE 2.

(a) FT–IR and (b) PXRD (baseline corrected) of HMTAC‐COF1. The energy‐optimized (c) eclipsed AA‐ and (d) staggered AB stacking patterns of hexagonal 2D HMTAC‐COF1, with the latter being 1.6 eV per formula unit more stable than the former due to less steric congestion around the HMTAC corners. (c) SEM and (d) HR‐TEM images of HMTAC‐COF1.

The PXRD profile of HMTAC‐COF1 shows prominent, albeit broad, peaks at 2θ = 5.3°, 7.9°, and 22.4° (Figure 2b), revealing its microcrystalline nature and periodic structure. Since the broad experimental PXRD profile of HMTAC‐COF1 was not suitable for structural refinement, to gain deeper insights into its structure and preferred stacking pattern, we calculated the DFT‐optimized structures of both AA (eclipsed) and AB (staggered) arrangements (Figure 2c,d) and compared their normalized formation energies to determine the energetically preferred stacking pattern. In the eclipsed AA pattern (a = 44.69 Å, b = 44.65 Å, c = 4.11 Å, α = 90°, β = 90°, γ = 120°), all hexagonal layers of HMTAC‐COF1 have the same orientation with a uniform interplanar distance of 4.11 Å (from the c‐parameter), whereas in the staggered AB staking pattern (a = 43.89 Å, b = 44.41 Å, c = 8.22 Å, α = 90°, β = 90°, γ = 120°), the alternate layers have the same orientation, as only the alternate HMTAC corners of successive layers overlap forming continuous π‐stacks, while the other three alternate HMTAC corners of each layer sit at the center of the hexagons of the neighboring layers (Figure 2d). Thus, the distance between the repeating layers of the AB pattern with the same orientation is doubled (c = 8.22 Å), although the distance between the adjacent layers with staggered alignment remains the same (4.11 Å). Furthermore, DFT calculations also indicate that the AB pattern (contains two formula units C₂₂₈H₁₅₆N₂₄O₂₄ per unit cell) is 1.6 eV per formula unit more stable than the AA pattern (one formula unit C₁₁₄H₇₈N₁₂O₁₂ per unit cell), making the former energetically preferred configuration, as it more effectively alleviates steric congestions at the corners caused by the large dihedral angle (∼60°) between the HMTAC core and phenyl sidearms. Although it was not possible to directly compare the broad experimental PXRD profile of HMTAC‐COF1 (Figure 2b) with the calculated PXRD profiles of its AA and AB patterns (Figure S4), which differed only in relative peak intensities, not in their positions, guided by computational results indicating that AB stacking pattern of HMTAC‐COF1 is energetically favored over AA, we attributed the observed PXRD peak at 22.4° to [002] plane's reflection. This corresponds to an interlayer distance of ca. 8 Å, which is in good agreement with c = 8.22 Å of the calculated AB pattern. The SEM image of HMTAC‐COF1 showed its morphology (Figure 2e), while its HR‐TEM image revealed its crystalline facets (Figure 2f). Notably, the d‐spacing (8.75 Å) measured from the distance between the lattice fringes observed in the HRTEM image of HMTAC‐COF1 (Figure 2f) is also in good agreement with the AB stacking pattern (c = 8.22 Å), further supporting this structural assignment.

The N₂ sorption isotherms of activated HMTAC‐COF1 (Figure S5) revealed a Brunauer–Emmett–Teller (BET) surface area (S_BET) of 12 m²/g, which is comparable to that of other Li⁺ ion conducting hexagonal 2D COFs with pores and channels lined with cation‐binding sites [70, 73]. The small surface area of HMTAC‐COF1 can be attributed to its preferred AB stacking pattern [73], which lacks large hexagonal channels; the presence of OMe groups at the HMTAC corners, which partially block the pores; and poor crystallinity. Thermogravimetric analysis (TGA) revealed that HMTAC‐COF1 is thermally stable up to 375°C (Figure S6), and it decomposes into volatile gases above 650°C.

Next, we turned our attention to the optical and electronic properties of HMTAC‐COF1, especially its electronic and ionic conductivity and the ability to serve as an SSE in LIBs. To this end, pristine HMTAC‐COF1 was doped with LiTFSI to obtain Li⁺@HMTAC‐COF1 (see Supporting Information for details). First, the ⁷Li NMR spectrum of digested Li⁺@HMTAC‐COF1 (dissolved in DMSO with a drop of TFA) displayed the characteristic Li⁺ signal at −0.85 ppm (Figure 3a), which was in good agreement with the free LiTFSI signal at −0.93 ppm observed in the same solvent medium, confirming the presence of Li⁺ ions in the COF [94]. A slight downfield shift of the Li⁺ signal in the COF with cation binding sites is consistent with Li⁺ ion coordination [94]. The solid‐state ⁷Li NMR spectrum of Li⁺@HMTAC‐COF1 (Figure S7) also featured a sharp signal at −0.27 ppm, confirming the presence of Li⁺ ions. Furthermore, the solid‐state 2D ¹H–⁷Li heteronuclear correlation (HETCOR) NMR experiment shows (Figure 3b) cross‐coupling between the Li⁺ ion (−0.27 ppm, x‐axis) and OCH₃ proton (4 ppm, y‐axis) signals, indicating close proximity between them (within a few Å) [95] due to Li⁺ ion coordination with the ─OMe groups, which was further corroborated by XPS and DFT results (vide infra).

FIGURE 3.

(a) The ⁷Li NMR (500 MHz, DMSO‐d ₆/CF₃CO₂D) of digested Li⁺@HMTAC‐COF1, LiTFSI, and pristine HMTAC‐COF1. (b) Solid‐state 2D ¹H–⁷Li HETCOR NMR spectrum, exhibiting cross‐coupling between ─OCH₃ protons and Li⁺ ion. (c) The FT–IR and (d) PXRD data of pristine and LiTFSI‐doped HMTAC‐COF1.

In addition, the FT‐IR spectrum of Li⁺@HMTAC‐COF1 (Figure 3c) displayed the characteristic signals of the framework (C═N stretching: 1697 cm⁻¹) as well as those of TFSI⁻ anion (CF₃ stretching: 1200 cm⁻¹ and SO₂ stretching: 610 cm⁻¹) [70], confirming the presence of LiTFSI in this material. The PXRD profile of Li⁺@HMTAC‐COF1 (Figure 3d) retained the characteristic signals of the COF, although the peak intensity diminished, which was not unexpected, as the insertion of bulky TFSI⁻ anions is known to reduce the crystallinity and periodicity of Li⁺‐doped COFs [70].

XPS analysis of pristine and LiTFSI‐doped HMTAC‐COF1 (Figure 4) provided further insights into Li⁺ ion binding within the COF. The deconvoluted high‐resolution O‐1s XPS signal of pristine HMTAC‐COF1 (Figure 4a) features a single characteristic peak at 532.73 eV corresponding to the C─O bond, whereas that of Li⁺@HMTAC‐COF1 (Figure 4b) consists of two distinct peaks—a new 530.32 eV peak corresponding to Li─O bonds and a 527.67 eV peak corresponding to C─O bond, which shifted to a lower energy, indicating Li⁺ coordination with the OMe groups. These XPS data are consistent with solid‐state 2D ¹H–⁷Li HETCOR NMR results (Figure 3b). In contrast, the deconvoluted N‐1s peaks of HMTAC‐COF1 remained essentially unchanged after LiTFSI doping (Figure 4c,d), indicating that the N‐atoms of the HMTAC core and imine linkers were not involved in Li⁺ coordination.

FIGURE 4.

(a–d) The deconvoluted high‐resolution O‐1s and N‐1s XPS peaks of pristine and Li⁺‐doped HMTAC‐COF1 indicate Li⁺ coordination with the OMe groups, but not with the N‐sites. (e) The DFT‐optimized structures of Li⁺@HMTAC‐COF1 also suggest energetically favored Li⁺ ion binding with the hard O‐sites over softer N‐sites.

In addition, DFT calculations were performed to assess Li⁺ ion binding at various sites of HMTAC‐COF1, including the OMe groups, the N‐atoms on the HMTAC corners, and the imine‐N atoms on the linkers (Figure 4e). These calculations suggested that the Li⁺ ions bind most preferentially to the OMe groups, with an average binding energy of ca. −2.0 eV per Li⁺ ion, which is slightly favored over Li⁺ binding with the imine‐N (−1.9 eV per Li⁺) but significantly preferred over Li⁺ binding with the HMTAC core‐N atoms (−0.8 eV per Li⁺). Simultaneous Li⁺ ion binding to the OMe groups and imine‐N sites remains energetically favorable (−1.8 eV per Li⁺ on average), but attempts to saturate all available O and N sites with Li⁺ led to structural instability under periodic boundary conditions, indicating a theoretical limit to Li⁺ loading that excluded the HMTAC core‐N sites. Notably, the theoretically predicted preferential coordination of Li⁺ to the hard OMe groups is corroborated by the experimental XPS data and consistent with the preferred hard–hard Li⁺···O interaction over hard–soft Li⁺···N mismatch.

The Li⁺ ion content in Li⁺@HMTAC‐COF1 was quantified by elemental analysis (see Supporting Information for details). Based on the S content (9.74 wt.% ≡ 0.3 mol%), which originated exclusively from the TFSI⁻ anion, the Li⁺ ion content was estimated to be 0.15 mol% (in LiTFSI, the S:Li molar ratio is 2). This accounted for roughly three Li⁺ ions per HMTAC corner (based on N mol% (0.49), each HMTAC unit contains six N atoms). Based on this estimation, we postulated that each set of ortho‐dimethoxy groups of the HMTAC corner possibly chelated one Li⁺ ion in a bidentate fashion, which is fully consistent with the experimental XPS data and DFT calculations (Figure 4). TGA analysis of Li⁺@HMTAC‐COF1 (Figure S6) showed that after initial solvent loss, it remained stable up to ca. 400°C. Unlike pristine HMTAC‐COF1, which left essentially no residual weight above 600°C (it decomposed into volatile gases), Li⁺@HMTAC‐COF1 left ∼2 wt.% residual weight, possibly corresponding to Li₂O. Based on the TGA data, we estimated that Li⁺@HMTAC‐COF1 contained 0.13 mol% of Li⁺, which was in good agreement with elemental analysis data and accounted for roughly three Li⁺ ions per HMTAC corner. The presence of the LiTFSI in Li⁺@HMTAC‐COF1 was also evident from scanning‐electron microscopy coupled with energy dispersive X‐ray analysis (SEM‐EDX), which exhibited the signature element peaks, especially S and F stemming from the TFSI⁻ anion (Figure S8).

The UV–vis–NIR diffuse reflectance spectrum (DRS) of orange HMTAC‐COF1 powder featured prominent peaks below 500 nm, and the corresponding Tauc plot revealed its optical band gap of 2.4 eV (Figure 5a,b). The cyclic voltammetric (CV) analysis of TAHMTAC (Figure S9) revealed a quasi‐reversible first oxidation peak at +1.2 V and a first reduction peak at −1.5 V (vs. Ag/AgCl), indicating a relatively wide electrochemical stability window. Furthermore, the current‐voltage (I/V) relationship of the pristine HMTAC‐COF1 pellet, measured by the two‐probe method (Figure S10), demonstrates its electrically insulating nature (σ ≤ 10⁻¹² S/cm), which bodes well for its potential application as an SSE material placed between two electrodes.

FIGURE 5.

(a) UV–vis–NIR DRS and (b) Tauc plots of pristine HMTAC‐COF1 and Li⁺@HMTAC‐COF1. (c) The temperature‐dependent Nyquist plots of Li⁺@HMTAC‐COF1. (d) The Arrhenius plot of temperature‐dependent ionic conductivity of Li⁺@HMTAC‐COF1.

The ionic conductivity of Li⁺@HMTAC‐COF1 was measured by electrochemical impedance spectroscopy (EIS) using its pellets placed between two stainless steel electrodes (Figure 5c) [79]. The Nyquist plots of Li⁺@HMTAC‐COF1 pellets demonstrated impressive ionic conductivity across a wide temperature range (4.95 × 10⁻⁵ S/cm at 20°C and 1.95 × 10⁻⁴ S/cm at 120°C). Notably, its room‐temperature ionic conductivity is significantly higher than that of other reported neutral COFs furnished with PEO‐chains (Table S1) and essentially on par with that of highly conducting ionic COFs, making it an attractive SSE material. The Arrhenius plot of temperature‐dependent ionic conductivities of Li⁺@HMTAC‐COF1 (Figure 5d) showed that it has a low activation energy (E _a = 0.25 eV), which was lower than most other Li⁺‐conducting COFs, suggesting that it can effectively facilitate Li‐salt dissociation as well as Li⁺ ion transport.

Next, to quantify the Li⁺ ions transport capability of Li⁺@HMTAC‐COF1, we measured its Li⁺ transference number (t _Li+ = 0.57) by the Bruce–Vincent method using a Li || Li⁺@HMTAC‐COF1 || Li symmetric cell [96, 97], which suggested that the mobile Li⁺ cations transported the majority (57%) of the total current (Figure 6a). The Li⁺@HMTAC‐COF1 transference number is among the highest values reported for Li⁺ ion‐conducting neutral COFs developed to date (Table S1), which can be attributed to facile Li⁺ ion binding and transport within HMTAC‐COF1, rendering it a promising SSE material. Encouraged by the excellent Li⁺ ion transport properties of Li⁺@HMTAC‐COF1, we investigated its ability to serve as SSE in LIBs. The electrochemical stability window was examined by linear sweep voltammetry (LSV) using a Li || Li⁺@HMTAC‐COF1 || stainless‐steel asymmetric coin cell (Figure 6b), which demonstrated its electrochemical stability in the voltage range of ∼2.0–3.7 V (vs. Li⁺/Li) and compatibility with the LiFePO₄ (LFP) cathode.

FIGURE 6.

(a) Chronoamperometry plot of a Li || Li⁺@HMTAC‐COF1 || Li symmetric cell under 10 mV polarization. Inset: Nyquist plots of Li⁺@HMTAC‐COF1 SSE before and after polarization at 10 mV. (b) LSV plot of Li⁺@HMTAC‐COF1 SSE shows the electrochemical stability window at a scan rate of 0.1 mV/s. (c) A photograph of the Li⁺@HMTAC‐COF1 SSE pellet and a graphical representation of Li || Li⁺@HMTAC‐COF1 || LFP asymmetric coin cell. (d) Cycling performance and (e) charge/discharge voltage profiles of Li || Li⁺@HMTAC‐COF1 || LFP asymmetric coin cell at 0.1C (17 mA/g) at 100°C. (f) Galvanostatic Li plating/stripping profile of Li⁺@HMTAC‐COF1 SSE in a Li || SSE || Li symmetric cell at 100°C shows a steady voltage with a minimal, gradual increase in overpotential over 250 h. (500 cycles) at a current density of 0.05 mA/cm², confirming electrode/electrolyte interfacial stability and ruling out short‐circuiting caused by Li dendrite formation.

To study its battery performance, we assembled CR2032‐type Li || Li⁺@HMTAC‐COF1 || LFP coin cells using a Li foil anode, an LFP cathode, and Li⁺@HMTAC‐COF1 SSE (Figure 6c), and measured their charge/discharge performance in the 2.5–3.7 V range at 0.1C and 100°C, which displayed a discharge specific capacity of 181 mAh/g and a Coulombic efficiency (CE) of 99.2% for the first cycle (Figure 6d,e). While CE held steady (>99%) for at least 20 cycles, the specific capacity gradually decreased, but still maintained ∼75% of the initial value (i.e., 135 mAh/g) after 20 cycles (Figure 6d,e), indicating its potential for long‐term cyclability. This behavior is not uncommon for COF‐based SSEs [70], especially for those using ionic liquids in SSE pellets and measured at elevated temperatures. Further engineering of battery construction, including optimization of Li⁺@HMTAC‐COF1 SSEs using different binder materials, artificial solid‐electrolyte interphase (SEI) formation, and/or replacing LFP with a cathode material with a cutoff voltage much lower than the SSE decomposition voltage (<3.7 V based on LSV), could further improve their long‐term electrochemical performance. Nevertheless, the CE of Li⁺@HMTAC‐COF1 (> 99%) at 100°C was particularly impressive, as elevated temperatures often accelerate side reactions at the electrode/electrolyte interfaces, thereby diminishing battery performance; however, this was not the case with our SSE. The initial discharge capacity of our Li || Li⁺@HMTAC‐COF1 || LFP cells slightly exceeded the reported theoretical limit of the LFP cathode (170 mAh/g at room temperature) [70], which can be attributed to potential mass loading errors in the commercially procured LFP/aluminum foil and/or because our battery measurements were conducted at 100°C, where electrochemical reaction kinetics are faster. Nevertheless, these results indicate that the rate of ion transport through the Li⁺@HMTAC‐COF1 SSE is sufficient to achieve high capacities, which bodes well for its potential practical application. Notably, both the discharge capacity and CE of Li || Li⁺@HMTAC‐COF1 || LFP cells were higher than those of SSEs based on PEO‐laced neutral and ionic COFs (Table S1), indicating that the higher ionic conductivity of Li⁺@HMTAC‐COF1 contributed to its superior performance as an SSE. This can be attributed to the fact that, unlike COFs decorated with long PEG chains, which tightly bind Li⁺ ions and whose restricted dynamics could hinder Li⁺ transport [70], HMTAC‐COF1, equipped with 1,2‐dimethxy groups, binds Li⁺ ions less tightly, which helps facilitate ion transport by allowing them to hop from one binding site to another inside the HMTAC‐COF1 pores and channels.

Finally, the interfacial stability between the Li electrode and Li⁺@HMTAC‐COF1 SSE was evaluated by galvanostatic Li‐plating/stripping experiment at 100°C under a current density of 0.05 mA/cm² using a Li || Li⁺@HMTAC‐COF1 || Li symmetric cell (Figure 6f), which showed a steady voltage for at least 250 h (500 cycles), with a slight, gradual increase in potential (from 20 mV at the beginning to 80 mV after 250 h). This indicated a slight thickening of the SEI layer but ruled out Li dendrite formation and an internal short circuit, as no abrupt change in voltage was observed. This result demonstrated the interfacial stability between the Li electrodes and HMTAC‐COF1 SSE, uniformity, and passivation efficiency [63, 65], which bodes well for its practical application.

3. Conclusion

In summary, we have designed and constructed a new, neutral, Li⁺ ion‐conducting hexagonal 2D covalent organic framework, HMTAC‐COF1, via a Sc(III)‐catalyzed imine reaction between a new 3‐fold symmetric TAHMTAC building block and a linear TPA linker, which can serve as an effective solid‐state electrolyte in LIBs. By introducing a preformed, fully characterized HMTAC building block with six bidentate OMe groups that bind Li⁺ ions, we not only ensured the formation of HMTAC‐COF1 with a well‐defined and verifiable structure and composition, but also overcame key limitations of neutral COFs, i.e., limited Li⁺ ion uptake capacity and transport efficiency. The XPS, ⁷Li NMR, and computational studies demonstrated Li⁺ ion binding to the OMe groups of HMTAC corners, while elemental analysis indicated that each HMTAC corner binds three Li⁺ ions (possibly one per bidentate ortho‐dimethoxy group). As a result, while pristine HMTAC‐COF1 behaves as an electrical insulator due to the lack of efficient charge‐transport pathways, the Li⁺‐doped material, Li⁺@HMTAC‐COF1, achieves adequate ionic concentration and mobility, exhibiting a high ionic conductivity of 4.95 × 10⁻⁵ S/cm at 20°C, a low activation energy of 0.25 eV, and a Li⁺ transference number of 0.57, making it a promising SSE for LIBs. Beyond these ion transport metrics, Li⁺@HMTAC‐COF1 exhibits electrochemical attributes required of a solid‐state electrolyte, including a stability window up to 3.7 V, excellent compatibility with Li anode and LFP cathode, and stable electrochemical performance in Li||Li+@HMTAC‐COF1||LiFePO₄ cells with a discharge capacity of 181 mAh/g and Coulombic efficiency above 99%. The sustained Li plating/stripping behavior further confirms robust interfacial stability and suppression of dendrite‐induced short circuiting. Together, these results establish HMTAC‐COF1 as the first confirmed HMTAC‐based COF and highlight a powerful molecular design strategy for neutral COFs that can rival, and in key respects approach, ionic COFs in solid‐state electrolyte performance. This work opens a promising route toward chemically versatile, processable, and high‐performing COF‐based electrolytes for safer next‐generation lithium batteries.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: smll73759‐sup‐0001‐SuppMat.pdf.

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.