Figure 1a illustrates the schematic synthesis procedure of Fe-TCPP@Cu-BTC and Cu-BTC by a conventional hydrothermal method31. The Fe-TCPP@Cu-BTC exhibits regular octahedral crystals, whose morphology is similar to Cu-BTC (Supplementary Fig. 1). Element mappings reveals uniform distribution of Cu and Fe atoms in Fe-TCPP@Cu-BTC, suggesting even dispersion of Fe-TCPP molecules within Cu-BTC channels (Supplementary Fig. 2). The powder X-ray diffraction (PXRD) reveals no obvious peak shifts from Cu-BTC to Fe-TCPP@Cu-BTC (Fig. 1b), reconfirming the encapsulation of Fe-TCPP in Cu-BTC frameworks. Based on these findings, Supplementary Fig. 3 provided the optimized Fe-TCPP@Cu-BTC molecular model according to the first principle simulations.
Furthermore, Fourier-transform infrared (FT-IR) and UV-vis absorption spectroscopy were used to investigate the chemical interactions between Fe-TCPP and Cu-BTC. As depicted in Fig. 1c, the featured C-H bonding vibration peak of pyrrole (~ 1001 cm− 1) in Fe-TCPP experienced a higher wavenumber (1003 cm− 1) shift upon combination Cu-BTC (Fe-TCPP@Cu-BTC)31, 32. This FT-IR finding implies the occurrence of chemical bonding between Fe-TCPP and Cu-BTC. UV-vis absorption spectra in Fig. 1d revealed a noticeable shift in the intense Soret band (~ 420 nm) compared to Fe-TCPP (~ 416 nm), likely due to depolymerization induced by Cu-BTC and subsequently self-assembly between Fe-TCPP and Cu-BTC33. Consequently, nitrogen adsorption-desorption tests yielded a specific surface area of 695.0 m2 g− 1 for Fe-TCPP@Cu-BTC, slightly lower than the 907.4 m2 g− 1 of Cu-BTC (Supplementary Fig. 4). In addition, X-ray photoelectron spectroscopy (XPS) was employed to evaluate the impact of Fe-TCPP on the electronic structure of Cu-BTC. Compared to Cu-BTC, the binding energies of Fe-TCPP@Cu-BTC for Cu 2p core levels were up-shifted from 952.2/932.5 to 954.1/934.4 eV and the O 1s core level shifted from 531.5 to 532.3 eV (Fig. 1f and Supplementary Fig. 5b)34, 35. On the contrary, the binding energies of Fe 2p core level were down-shifted from 724.8/711.5 to 722.4/710.1 eV, and the N 1s core level shifted from 399.9/397.7 to 400.6/398.8 eV (Fig. 1g and Supplementary Fig. 5a)36, 37. This means the existence of electron transfer from outer Cu-BTC to inner Fe-TCPP in Fe-TCPP@Cu-BTC, which can chemically anchor Fe-TCPP and thereby inhibit its dissolution. Supplementary Fig. 6 illustrates the stability of Fe-TCPP@Cu-BTC and Fe-TCPP in a standard LSB electrolyte. It demonstrates that the encapsulation of Fe-TCPP with Cu-BTC effectively prevents Fe-TCPP dissolution. The local solvent environment on the catalyst surface was probed using FT-IR spectroscopy (Fig. 1e), where Fe-TCPP@Cu-BTC was immersed in DME for 2 h followed by drying. In this scenario, the peaks at 1110 and 940 cm‒1 correspond to the C‒O and O‒Cu bonds of Cu-BTC31; while the peak at 2878 cm‒1 corresponds to the ‒CH3 group of DME38. Upon treatment with DME, the characteristic C‒O and O‒Cu signals of Fe-TCPP@Cu-BTC shifted to 1090 and 917 cm‒1, respectively; while the ‒CH3 signal of DME shifted to 2931 cm‒1. This observation likely originates from dipole-dipole interactions between the solvent and Cu-BTC. As previously reports, these interactions between the catalyst and solvent may facilitate the desolvation of LiPSs, thereby promoting their smooth conversion36.
To elucidate the behavior of LiPSs within Fe-TCPP@Cu-BTC and to further substantiate our hypothesis, density functional theory (DFT) simulations were conducted to mimic the sequential sulfur transformations throughout the adsorption-catalysis-desorption processes, as illustrated in Fig. 2. The existence of abundant oxygen-containing functional groups within the Cu-BTC component of the Fe-TCPP@Cu-BTC catalyst facilitates the formation of Li‒O bonds with Li in LiPSs during the reduction process, which in turn attract LiPSs to congregate around the Fe-TCPP@Cu-BTC. Under the influence of an electric field, the Cu-BTC component catalyzes the transformation of Li2S8 into shorter-chain Li2S6, and their shrunk lengths and reduced volumes enable them to penetrate more readily the cavities of the Fe-TCPP@Cu-BTC39.
Upon entering the cave formed by Fe-TCPP@Cu-BTC, the terminal S (ST) atoms of Li2S6 interact with the Cu and Fe components of the Fe-TCPP@Cu-BTC, leading to the formation of Cu-S and Fe-S bonds. Note that this interaction is significantly exothermic by 27.0 kcal/mol, indicating that the encapsulation of Li2S6 in this cavity is thermodynamically favorable and likely to occur spontaneously. The subsequent cleavage of Li2S6, facilitated by the strain exerted on the sulfur skeleton by the Cu-S and Fe-S bonds (as evidenced by the difference in S-S bond lengths in state I and pristine Li2S6), may occur through three distinct pathways, denoted as IIa, IIb, and IIc, culminating in the formation of Li-Sx-M (where x = 2, 3, or 4) moieties, respectively. Obviously, the homolytic cleavage leading to the formation of Li-S3-M in IIb is thermodynamically favored over the heterolytic cleavage into Li-Sx-M (x = 2 or 4) (-9.1 kcal/mol vs. -1.2 and 11.0 kcal/mol, respectively, relative to state I), which aligns well with experimental observations and previous reports.
Continuing with a constant supply of Li+ in the environment, the S3 moiety within Li-S3-M tends to further divide into Sy and S3 − y-M (where y = 1 or 2), generating dissociative Li2S or Li2S2 species. The production of Li2S/LiS2-M in IVa and Li2Sz/LiS3 − z-M (where z = 1 or 2) in IVb involves energetically more demanding steps by 61.0 and 27.2 kcal/mol, as compared to the less energy-intensive formation of Li2S2/LiS-M in IVc and Li2Sz/LiS3 − z-M (where z = 1 or 2) in IVd by 16.7 and 17.2 kcal/mol, respectively, with reference to the Li-S3-M state denoted as III in this progression. The final cleavage of the remaining S-M bonds in Vc, facilitated by additional Li+, can proceed smoothly through a slightly endothermic step of 7.2 kcal/mol, completing the sulfur reduction process. Conversely, the transformation from IVd is thermodynamically quite disfavored. Therefore, the sulfur transformation process is envisioned to proceed in a stepwise manner through stages I – IIb – III – IVc – Vc – VIc, as illustrated in Fig. 2.
The interactions between Fe-TCPP@Cu-BTC and LiPSs were further investigated through static adsorption tests. As depicted in Supplementary Fig. 7, the fading of the yellow Li2S6 solution after 12 h of treatment with Fe-TCPP@Cu-BTC, Cu-BTC, Fe-TCPP, and CNTs followed a color depth order of Li2S6 > CNTs > Fe-TCPP > Cu-BTC > Fe-TCPP@Cu-BTC, indicating the highest LiPS absorptivity of Fe-TCPP@Cu-BTC40. Subsequently, XPS measurements were conducted to analyze the compositions of the soaked Fe-TCPP@Cu-BTC, Cu-BTC, Fe-TCPP, or CNTs surfaces. As shown in Figs. 3a‒c, distinct Cu 2p and Fe 2p XPS peaks were observed in Li2S6-treated Fe-TCPP@Cu-BTC (Cu 2p: ~950.8/930.8 eV; Fe 2p: 723.9/710.0 eV), Li2S6-treated Cu-BTC (Cu 2p: ~953.0/933.1 eV), and Li2S6-treated Fe-TCPP (Fe 2p: ~724.5/710.9 eV), corresponding to Cu‒S and Fe‒S bond formations resulting from Cu atoms of Cu-BTC, Fe atoms of Fe-TCPP, and S atoms of LiPSs41–43. This observation was corroborated by the S 2p XPS spectra and corresponding deconvolution analysis (Fig. 3c). The presence of Fe and Cu atoms led to shifts in the binding energies of bridge S (SB) and ST in Li2S6-treated Fe-TCPP@Cu-BTC towards higher fields, confirming the formation of Fe‒S and Cu‒S bonds. Moreover, compared to Li2S6-treated CNTs, the higher binding energy shifts of SB and ST in Li2S6-treated Fe-TCPP@Cu-BTC, Cu-BTC, and Fe-TCPP indicated decreased electron density, suggesting the catalyst's mediation of sulfur conversions. Additionally, a lower binding energy shift of Li 1s XPS peaks was observed in Li2S6-treated Fe-TCPP@Cu-BTC compared to Li2S6-treated Cu-BTC, Li2S6-treated Fe-TCPP, and Li2S6-treated CNTs (Supplementary Fig. 8). Consequently, the O 1s and N 1s peaks shifted towards higher binding energy after Li2S6 treatment, indicating strong electron transfer from O and N atoms of Fe-TCPP@Cu-BTC to Li atoms of Li2S6, leading to Li···O and Li···N bond formation, known as Li-bonds, according to previous reports (Supplementary Fig. 9)36. These interactions, mainly through Li-bonds and Fe‒S/Cu‒S bonds, resulted in LiPS enrichment in Fe-TCPP@Cu-BTC and inhibited the shuttle effect of LiPSs.
The catalytic effect of Fe-TCPP@Cu-BTC on LiPSs was extensively investigated through cyclic voltammetry (CV) tests in a symmetrical battery employing Li2S6 electrolyte (Supplementary Fig. 10). Fe-TCPP@Cu-BTC cells exhibited a significantly higher current density compared to Cu-BTC, Fe-TCPP, and CNTs, indicative of substantially improved redox kinetics of LiPSs by Fe-TCPP@Cu-BTC. Electrochemical reactivity was further assessed via CV analysis. Supplementary Fig. 11a‒d illustrates the initial CV profiles of cells with Fe-TCPP@Cu-BTC, Cu-BTC, Fe-TCPP, and CNTs at a rate of 0.1 mV s‒1, revealing two major peaks at ~ 2.31 and 2.05 V, corresponding to the reduction of S8 to soluble long-chain LiPS (Li2Sn, 4 ≤ n ≤ 8), and further to solid Li2S2/Li2S, respectively. The oxidation peaks at ~ 2.42 V are attributed to the conversion from Li2S2/Li2S to Li2Sn and S8. Comparing with other control cells, the CV curve of the Fe-TCPP@Cu-BTC cell displays a distinct positive shift of the cathodic peak and negative shift of the anodic peak (Supplementary Fig. 12a‒b), indicating reduced voltage polarization and enhanced reaction kinetics influenced by Fe-TCPP@Cu-BTC.
Furthermore, sulfur redox kinetics were quantitatively evaluated using Tafel slopes based on each individual process (Fig. 3a‒b and Supplementary Fig. 13). The Fe-TCPP@Cu-BTC cell also demonstrated the smallest value of 28.3, 33.4, and 43.7 mV dec‒1 at S8→Li2Sn, Li2Sn→Li2S, and Li2S→Li2Sn conversions, respectively, compared to Cu-BTC (32.3, 52.6, and 85.1 mV dec‒1), Fe-TCPP (38.4, 62.3, and 111.7 mV dec‒1), and CNTs (41.9, 70.9, and 132.7 mV dec‒1). These results highlight the excellent redox reaction kinetics behaviors and electrochemical reversibility of Fe-TCPP@Cu-BTC. Additionally, the Li+ diffusion coefficient in each cathode was quantified using CV tests under various scan rates ranging from 0.10 to 0.30 mV s− 1. Plots in Fig. 3d, Supplementary Fig. 14, and Supplementary Fig. 15a‒c indicate that the peaks for cells with Fe-TCPP@Cu-BTC are much more intense than those for the other three cells, suggesting promoted LiPS conversions by Fe-TCPP@Cu-BTC. The Li ion diffusion properties were estimated using the classical Randles-Sevcik equation. The cells with Fe-TCPP@Cu-BTC exhibited larger slopes of lithium-ion diffusion coefficient (PA = 3.24 × 10‒9; PB = 1.30 × 10‒8; PC = 2.65 × 10‒8 cm2 s‒1) compared to Cu-BTC (2.95 × 10‒9; 6.39 × 10‒9; 1.83 × 10‒8 cm2 s‒1), Fe-TCPP (2.06 × 10‒9; 5.97 × 10‒9; 1.51 × 10‒8 cm2 s‒1), and CNTs (1.40 × 10‒9; 5.52 × 10‒9; 7.94 × 10‒9 cm2 s‒1), demonstrating enhanced LiPS conversion kinetics of Fe-TCPP@Cu-BTC throughout the charge/discharge processes. Nucleation Transformation Ratio (NTR) was used to assess the kinetics behaviors of the cathode reactions44. All NTR values for Fe-TCPP@Cu-BTC cells at different scan rates were close to 3, indicating rapid transformation of LiPSs to Li2S facilitated by Fe-TCPP@Cu-BTC (Fig. 3g).
The process of converting LiPSs to Li2S during discharge accounts for three-quarters of the theoretical capacity, underscoring the importance of enhancing kinetics in this phase. To investigate the impact of Fe-TCPP@Cu-BTC on this conversion, we conducted potentiostatic intermittent titration technique (PITT) tests on cells, focusing on liquid–solid conversion kinetics (see Supplementary Fig. 16). Generally, PITT discharge curves exhibit two distinct regions: the first involves liquid-liquid conversion from long-chain to short-chain LiPS, while the second region entails liquid-solid conversion, corresponding to Li2S deposition. In the initial process, the addition of Fe-TCPP@Cu-BTC led to notable enhancements in initial current responses, with increases of 55% (Cu-BTC), 134% (Fe-TCPP), and 228% (CNTs) at each potentiostatic step, indicating improved liquid-liquid conversion kinetics. Subsequently, during the second process, the Li2S deposition peak time was earliest for the Fe-TCPP@Cu-BTC cell (1970 s), indicative of a faster deposition rate45–47. Further Li2S nucleation tests were conducted, with results detailed in Supplementary Fig. 16. According to Faraday’s law, deposition capacities on Fe-TCPP@Cu-BTC, Cu-BTC, Fe-TCPP, and CNTs electrodes were measured at 236.9, 182.9, 175.7, and 142.6 mAh g− 1, respectively48. These findings highlight the ability of Fe-TCPP@Cu-BTC to inhibit shuttle effects and enhance sulfur utilization, thereby enabling high-capacity Li2S precipitations.
Dynamic monitoring techniques were employed to trace the sulfur evolution features. In-situ ultraviolet‒visible (UV‒Vis) absorption spectra of four cells: Fe-TCPP@Cu-BTC, Cu-BTC, Fe-TCPP, and CNTs are depicted in Supplementary Fig. 18‒19. Each cell exhibits characteristic absorption peaks at approximately 492, 475, 420, and 617 nm, corresponding to S82−, S62−, S42−, and S32−/S3*− anions, respectively. During discharge, the Fe-TCPP@Cu-BTC cell demonstrated the most rapid decrease in S82− intensity compared to other cells in the Li2S8 electrolyte (Fig. 4a)40. Moreover, the intensity of S32‒/S3*‒ initially increased followed by a decrease, indicating its role as an intermediate bridge for Li2S2/Li2S formation. A similar trend was observed in the Li2S4 electrolyte (Fig. 4b), suggesting the capability of Fe-TCPP@Cu-BTC catalyst to promote both long-chain and short-chain LiPS conversion, along with Li2S2/Li2S generation. Notably, unlike other samples, the Fe-TCPP sample exhibited an additional absorption peak at 430 nm, which is the characteristic of porphyrin iron, further confirming Fe-TCPP's tendency to dissolve in the electrolyte.
To gain a deeper insight into the catalytic effect of Fe-TCPP@Cu-BTC on sulfur species transformation, in-situ Raman spectra were measured (Figs. 4c‒f, Supplementary Fig. 20). Without a catalyst (Fig. 4f, Supplementary Fig. 20d), only the S8/Li2S8 peak (152, 246, 437, 472 cm‒1) was observed, persisting even after discharge to 1.6 V, indicating incomplete sulfur reduction, which leads to the poor CNT cell performance. However, with Fe-TCPP@Cu-BTC, the intensity of the S8/Li2S8 peak gradually decreased during discharge. By 2.5 V discharge, peaks of S62‒ (325 cm‒1) and S42‒ (496 cm‒1) emerged, while the S8/Li2S8 peak vanished, indicating decomposition into Li2S6 and Li2S4. When discharging to 2.0 V, Li2S6 vanished, replaced by a Li2S4 peak (496 cm‒1), along with a small amount of S32‒/S3*‒ (534 cm‒1) and Li2S2 (510 cm‒1). Upon discharge to 1.6 V, Li2S4 and S32‒/S3*‒ were consumed, yielding Li2S2 and Li2S (375 cm‒1), indicating complete conversion of S8/Li2S8 to Li2S2/Li2S49–51. The Cu-BTC cell (Fig. 4d, Supplementary Fig. 20b) exhibited slower disappearance or generation of sulfur species compared to the Fe-TCPP@Cu-BTC cell. Notably, the S8/Li2S8 peak vanished at 2.3 V, while complete consumption and generation of S8/Li2S6, Li2S3/Li2S4 occurred at 2.4 V, 2.1 V, 1.8 V, and 1.9 V, respectively, along with a small production of Li2S2. Moreover, the solubility of Fe-TCPP in the electrolyte could cause catalyst loss, reducing sulfur reaction kinetics. In the Fe-TCPP cell (Fig. 4e, Supplementary Fig. 20c), only the disappearance of S8/Li2S8 (2.1 V) and generation of Li2S6 were detected. Thus, the addition of Fe-TCPP@Cu-BTC catalyst facilitated sulfur species reduction via homolysis, increasing the sulfur reduction pathways and ensuring rapid and complete sulfur conversion.
The semi-in-situ analysis of Cu 2p and Fe 2p XPS spectra further elucidated the catalytic mechanisms, as depicted in Figs. 4g‒h. Evidently, the Cu 2p1/2, Cu 2p3/2, and Cu‒S peaks associated with Cu-BTC for the Fe-TCPP@Cu-BTC cell gradually shifted to lower binding energies (Cu 2p1/2/Cu 2p3/2 and Cu‒S; 2.8 V: 954.1/933.5 eV and 952.0/932.0 eV; 2.1 V: 953.4/933.1 eV and 951.6/931.8 eV; 1.6 V: 953.7/932.6 eV and 951.3/930.7 eV, respectively) compared to the Cu-BTC cell (Cu 2p1/2/Cu 2p3/2 and Cu‒S; 2.8 V: 954.4/934.4 eV and 952.8/933.7 eV; 2.1 V: 954.3/934.4 eV and 952.8/933.3 eV; 1.6 V: 953.9/933.9 eV and 952.6/933.0 eV, Supplementary Fig. 21a). Similar observations were made for the Fe 2p1/2, Fe 2p3/2, and Fe‒S peaks in Fe-TCPP@Cu-BTC (Fe 2p1/2/ Fe 2p3/2 and Fe‒S; 2.8 V: 727.7/714.0 eV and 725.7/712.1 eV; 2.1 V: 727.4/713.9 eV and 724.3/711.4 eV; 1.6 V: 727.0/712.9 eV and 724.0/710.8 eV, respectively) and Fe-TCPP cell (Fe 2p1/2/ Fe 2p3/2 and Fe‒S; 2.8 V: 728.3/714.3 eV and 726.7/712.3 eV; 2.1 V: 728.1/714.3 eV and 725.9/712.0 eV; 1.6 V: 727.9/714.3 eV and 725.8/711.9 eV, Supplementary Fig. 21b)52, 53. The lower binding energy shifts of Fe‒S and Cu‒S bonds during the discharge process suggest enhanced interactions between Cu and S, and Fe and S atoms, facilitating the breakage of S‒S bonds from long-chain to short-chain LiPSs. Moreover, the Cu‒S and Fe‒S bonds between LiPSs and Fe-TCPP@Cu-BTC led to a gradual increase in the electron cloud density around Cu and Fe atoms, which contributing to S‒S bond breaking.
To reveal the enzymatic catalysis mechanism of Fe-TCPP@Cu-BTC, a steady-state kinetic analysis was conducted by varying the concentration of S3*‒ within a fixed concentration of Fe-TCPP@Cu-BTC by in-situ UV‒Vis spectra, which exhibited conformity with the classic Michaelis‒Menten kinetics throughout all stages of discharge processes (Figs. 5c‒d). The Michaelis‒Menten plots provided values for the enzyme kinetic parameters of Michaelis‒Menten constants (Km), where Km reflects the affinity of the biomimetic enzyme towards the substrate (lower Km indicates a higher affinity) and Vmax indicates the catalytic activity of the biomimetic enzyme. Compared to Cu-BTC (fitting the equation within 2.5–2.2 V with Km = 1.02×10− 3 mM) and Fe-TCPP (fitting the equation within 2.1–1.6 V with Km = 3.89×10− 2 mM), Fe-TCPP@Cu-BTC exhibits consistently smaller Km values throughout the entire sulfur conversion range (2.8–2.2 V, Km = 5.92×10− 3 mM; 2.1–1.6 V, Km = 7.79×10− 4 mM) (see Supplementary Table 1). In contrast, CNTs did not exhibit conformity with this kinetics (Supplementary Fig. 22). These findings indicate that the Fe-TCPP@Cu-BTC biomimetic enzyme exhibits higher affinity, which is consistent with LiPS adsorption findings (see Supplementary Fig. 7). Furthermore, Vmax was determined at different sulfur conversion segments. The Fe-TCPP@Cu-BTC (2.8–2.2 V: Vmax = 5.87×10− 4 mM min− 1; 2.1–1.6 V: Vmax = 1.87×10− 4 mM min− 1) is about two orders of magnitude higher than that of Fe-TCPP (2.5–2.2 V: Vmax = 8.47×10− 6 mM min− 1) and Cu-BTC (2.1–1.6 V: Vmax = 1.24×10− 6 mM min− 1) (see Supplementary Table 2). This suggests that the homolytic reaction of LiPSs under the influence of this biomimetic enzyme can increase the sulfur conversion rate by nearly 100 times.
To disclose the kinetic improvements in sulfur redox reactions, the activation barrier at specific voltages was experimentally determined. This was achieved by fitting the charge transfer resistance measured at various temperatures using electrochemical impedance spectroscopy (EIS), to the activation energy (Ea) (Fig. 5e and Supplementary Fig. 23–28). The Fe-TCPP@Cu-BTC cell exhibited a significantly lower Ea value (0.29–1.03 eV) compared to Cu-BTC (0.34–1.20 eV), Fe-TCPP (0.38–1.25 eV), and CNTs (0.55–1.42 eV) cells within the voltage range of 2.4–1.6 V. These results demonstrate that Fe-TCPP@Cu-BTC can catalyze the conversion of both long-chain and short-chain LiPSs more effectively54.
The rate performance demonstrates that the Fe-TCPP@Cu-BTC cell delivers superior discharge capacities of 1522, 1162, 1079, 1024, and 970 mAh g− 1 at 0.2, 0.5, 1.0, 1.5, and 2.0 C, respectively (Fig. 6a). When the rate returns to 1.5, 1.0, 0.5, and 0.2 C, the reversible capacities are restored to 1000, 1023, 1054, and 1189 mAh g− 1, respectively. In contrast, the control cells exhibit significant fluctuation and a general attenuation trend with changing rates. The improved performance of the Fe-TCPP@Cu-BTC cell indicates a kinetically efficient reaction process with fast electron transfer. Additionally, the galvanostatic charge-discharge curves of the Fe-TCPP@Cu-BTC electrode at various C-rates (0.2–2.0 C) are shown in Fig. 6b, revealing that the polarization of the charge-discharge curves increases slightly with higher C-rates. The decreased overpotential, superior rate capability, and high reversibility of the Fe-TCPP@Cu-BTC cell result from improved electrochemical kinetics, as further revealed by electrochemical impedance spectroscopy (EIS). Supplementary Fig. 29 illustrates the impedance for both control and Fe-TCPP@Cu-BTC cells after 0, 3, and 10 cycles. After 10 cycles, the Re (resistance of electrolyte), Rct (resistance of charge transfer), and Rmt (resistance of mass transfer) of the Fe-TCPP@Cu-BTC cell remain similar to those at 3 cycles, indicating continuous cycling stability of electrons/ions and improved sulfur utilization. It is reasonable to speculate that a solid electrolyte interphase (SEI) film forms after three cycles in the Fe-TCPP@Cu-BTC cell. Detailed impedance data is provided in Supplementary Tables 1‒4.
Moreover, the Fe-TCPP@Cu-BTC catalyst exhibits excellent long-term cycling performance for the Li–S cell. Figure 6c and Supplementary Fig. 30 shows that the Fe-TCPP@Cu-BTC cell delivers a high initial discharge capacity of 1548 mAh g− 1 at 0.2 C, maintaining 935 mAh g− 1 after 150 cycles. In contrast, the control cells (Cu-BTC, Fe-TCPP, and CNTs) deliver initial capacities of 1406, 1395, and 1298 mAh g− 1, maintaining 733, 574, and 448 mAh g− 1 after 150 cycles, respectively. The cycling stability of these cells was further measured at 1.0 C. Compared with Cu-BTC (0.056%), Fe-TCPP (0.075%), and CNTs (0.111%), the Fe-TCPP@Cu-BTC cell shows the highest capacity retention (Fig. 6d, Supplementary Fig. 31) and the lowest capacity decay (0.043% per cycle) with an initial discharge capacity of over 1016 mAh g− 1 and maintaining 751 mAh g− 1 after 600 cycles. Furthermore, higher sulfur loadings of 4.5, 5.6, and 6.2 mg cm− 2 were tested under an electrolyte/sulfur (E/S) ratio of 6.5 µL mg− 1, with initial discharge capacities of 1350, 1296, and 1280 mAh g− 1, fading to 976, 885, and 682 mAh g− 1 after 100 cycles (Fig. 6e). The superior stabilization effect of Fe-TCPP@Cu-BTC is confirmed by the pouch cell (Fig. 6f), which exhibited a high initial discharge capacity of 5.2 mAh cm− 2 (1435 mAh g− 1) with a sulfur loading of 4.0 mg cm− 2 and an E/S ratio controlled at 4 µL mg− 1. After 50 cycles, the capacity still maintains above 3.7 mAh cm− 2 (986 mAh g− 1). The pouch cell performance parameters for the first discharge cycle are summarized in Fig. 6g, showing a high specific discharge capacity of 1299 mA h g− 1, providing a total discharge capacity of 0.104 Ah. The average voltage was 2.19 V, and the discharge energy of the pouch cell was 0.228 Wh. Based on the total mass, the actual energy density of the pouch cell was calculated to be 300.4 Wh kg− 1.