View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/D5TA04041H (Paper) J. Mater. Chem. A, 2025, 13, 24590-24598

Honeycomb graphite network confined in biphasic TiO2 homojunction nanotubes as the sulfur host for advanced lithium sulfur batteries†

Shidi Huang *, Xuan Zhao , Zheqian Yu , Weiye Tong and Yijie Zhang
School of Ecological Technology and Engineering, Shanghai Institute of Technology, Shanghai 201418, P. R. China. E-mail: sdhuang@sit.edu.cn

Received 19th May 2025 , Accepted 21st June 2025

First published on 4th July 2025

Abstract

We designed a biphasic TiO2 homojunction nanotube containing an N-doped honeycomb graphite network coated with by a carbon layer (HGN@TiO2@C) as the sulfur host to improve the Li–S battery performance. Theoretical calculations and experiments demonstrated that the design of a high-conductivity TiO2 homojunction with near-perfect lattice matching and rich interfaces/boundaries, accompanied by the conductive HGN inside the conductive carbon layer, effectively confines sulfur and offers additional buffer space. This design not only facilitates the effective conversion of LiPSs and enhances both electronic and ion conductivity but also improves the sulfur adsorption and catalytic properties, thereby boosting the kinetics of sulfur evolution reactions. Besides, the unique porous structure of HGN achieves an overall enhancement in S/HGN conductivity by providing a high specific surface area to load more sulfur and accommodate the volume changes occurring during the charge–discharge process. Benefiting from these synergistic effects, HGN@TiO2@C exhibits high coulombic efficiency, outstanding rate performance, and superior cycling stability (601.7 mA h g−1 for the fourth cycle and ≈626.8 mA h g−1 after 200 cycles at 1C, corresponding to a capacity retention of ≈104.2%).

Introduction

Lithium–sulfur (Li–S) batteries are considered a promising alternative to conventional lithium-ion batteries due to their low cost and natural abundance. Sulfur has an impressive theoretical capacity and specific energy that surpasses that of state-of-the-art lithium-ion batteries.1,2 However, the practical application of Li–S batteries is limited by low sulfur utilization, poor long-term cyclability, and inferior rate capability, primarily due to sluggish ion diffusion and reaction kinetics. Additionally, the dissolution of lithium polysulfide (LiPS) intermediates (Li2Sn, 2 ≤ n ≤ 8) in the electrolyte triggers a series of side effects and the “shuttle effect”.3–5 Therefore, developing an effective solution to these issues is crucial for the practical application of Li–S batteries.

In recent years, various strategies have been proposed to address the issues associated with Li–S batteries, including the construction of a nanocarbon host,6 the introduction of an electrolyte additive,7 and the fabrication of separators/interlayers.8,9 Among these, nanocarbon-based hosts, such as graphene,10 carbon fiber,11 and foam carbon,12 are particularly popular due to their high conductivity, large specific surface area, physical adsorption of LiPSs and inhibition of the shuttle effect. These properties can suppress the shuttle effect, expedite sulfur's electrochemical reaction kinetics and enhance the electrochemical performance of Li–S batteries. However, Li–S batteries based solely on carbon materials often experience significant capacity decay due to the loss of sulfur, which results from the weak interactions between the polar LiPSs and nonpolar carbon host materials.13 Therefore, modifying carbon with polar materials as host materials can significantly enhance sulfur utilization. Introducing polar materials into porous carbon hosts, such as metal oxides, metal sulfides, or metal nitrides, can significantly enhance the utilization of sulfur due to the polar–polar chemical interactions between the polar materials and LiPSs.14–16 However, the intrinsic poor conductivity of these introduced polar materials not only slows electron transport down and reduces the reaction kinetics but also hinders the adsorption of active sites on the surface of the nanocarbon host.

To improve the electron conductivity of the introduced polar materials, heterostructure engineering has been demonstrated to be a promising strategy to enhance the reaction kinetics in recent years. The constructed heterostructure can modify the electronic structure, thereby improving the electron conductivity of the electrodes.17 Additionally, the abundant phase boundaries in the heterostructure offer extensive migration paths and reduce the diffusion energy barrier for Li+ ions.18 Furthermore, heterostructure engineering with introducing varied polar materials can realize strong adsorption, high catalytic activity, bidirectional conversion promotion, and so on, which can further improve the cycling stability of the Li–S batteries.19 For example, heterojunctions such as TiO2–TiN,20 VC–VO,21 and Mo2C–MoS2 (ref. 22) exhibit good electronic conductivity, different for the single-component compounds, that can balance the trapping ability and catalytic activity toward the LiPSs simultaneously. These heterojunctions have been specifically engineered to optimize the kinetics of LiPS conversion and significantly enhance the overall performance. Although heterostructure engineering has been successful in improving the electrochemical performance of Li–S batteries, the varying material compositions obtained from their heterogeneous components in heterostructures lead to abundant side reactions in the heterojunction interface with Li+ reactions, which might destroy the heterostructure, resulting in electrolyte consumption and irreversible trapping of Li+ and sulfides, significantly decreasing the lifespan of the Li–S batteries. Also, the synthesis of heterostructures with rich phase boundaries as sulfur hosts typically results in an open composite structure, which cannot fully maximize the synergistic adsorption and catalytic effects, limiting the overall performance of Li–S batteries. It remains a significant challenge to develop a facile strategy to construct rich heterostructure boundaries with completely encased sulfur to obtain high-performance Li–S batteries.

Recently, compared to heterojunctions made of different materials, homojunctions composed of the same chemical constituents with near-perfect lattice matching have been shown to provide an additional driving force for promoting charge transfer and regulating the catalytic activity, thereby enhancing the performance of Li–S batteries.23,24 TiO2 in both its rutile and anatase forms has been shown to exhibit high binding energy for long-chain LiPSs (Li2S4–8) in the rutile phase and a higher adsorption capacity for short-chain Li2S2 in the anatase phase. Additionally, TiO2 in a crystalline phase at heterointerfaces can generate a built-in electric field, resulting in accelerated electron transport and improved reaction kinetics.25 However, TiO2 has poor conductivity, which can lead to low electron transfer and sluggish redox kinetics.26 The chemical adsorption of TiO2/S/C complexes can stably capture the soluble LiPSs and enhance the conductivity. Nevertheless, the LiPSs are bound physically or chemically and cannot participate in subsequent electrochemical reactions, leading to the formation of “dead sulfur” and a decrease in battery capacity.

Herein, we designed and fabricated biphasic TiO2 homojunction nanotubes containing an N-doped honeycomb graphite network with a covering carbon layer (HGN@TiO2@C) as a sulfur host to improve the Li–S battery performance. Detailed electrochemical analysis and theoretical calculations demonstrated that the high-conductivity TiO2 homojunction with near-perfect lattice matching and rich interfaces/boundaries, accompanied by conductive HGN inside the conductive carbon layer, which could effectively confine the sulfur and provide additional buffer space for the electrochemical reaction, could not only realize the effective and fluent conversion of LiPSs and promote the electronic and ion conductivity, it could also effectively enhance the adsorption ability and catalytic ability for the transformation of Li2S. Furthermore, the unique porous structure of HGN facilitated an overall enhancement in S/HGN conductivity, providing a high specific surface area for sulfur loading and for accommodating volume changes during the charge–discharge process. As a result, HGN@TiO2@C demonstrated superior electrochemical performance in Li–S batteries, including a high reversible capacity, remarkable rate capability, and outstanding cycling stability.

Experimental

Material preparation

Synthesis of MIL-88A hexagonal nanorods. MIL-88A hexagonal nanorods were synthesized according to a previous report.27 Typically, 25 mL deionized water was preheated to 70 °C and then 140 mg fumaric acid was added into the water and stirred for 10 min. Then, 525 mg Fe(NO3)3·9H2O was added into the solution. After 10 min stirring, the mixed solution was transferred to a 50 mL Teflon-lined stainless-steel autoclave and kept at 110 °C for 6 h. After cooling down to room temperature, the orange precipitate was collected by centrifugation before being dried at 60 °C for 12 h to obtain MIL-88A hexagonal nanorods.
Synthesis of hollow Fe2O3@TiO2 nanotubes. To obtain the hollow Fe2O3@TiO2 nanotubes, TiO2-coated MIL-88A (MIL88@TiO2) was first prepared as follows: 100 mg MIL-88A powder was dispersed in 50 mL ethanol followed by adding 0.5 mL ammonium hydroxide and stirring for 5 min. Then, 0.1 ml titanium isopropoxide was dropwise injected into the dispersion under stirring at 100 rpm. After reaction for 1 h, MIL88@TiO2 was obtained by centrifugation and washed several times with ethanol. Next, the dried MIL88@TiO2 powder was calcined at 500 °C in air for 2 h at a heating rate of 1 °C min−1. After cooling down to room temperature, hollow Fe2O3@TiO2 nanotubes were obtained. Additionally, hollow TiO2 nanotubes were obtained by immersing Fe2O3@TiO2 nanotubes into 4 M HCl. The hollow Fe2O3 polyhedral material was prepared by directly calcining MIL88.
Synthesis of HGN@TiO2@C. The honeycomb graphite network confined in biphasic TiO2 homojunction nanotube@carbon (HGN@ TiO2@C) was synthesized by a modified CVD method. Typically, 50 mg Fe2O3@TiO2 nanotube powder was annealed under a flow of Ar/H2 (90[thin space (1/6-em)]:[thin space (1/6-em)]10) for 2 h at 700 °C at a heating ramp rate of 2 °C min−1. Excessive 2-methylimidazole was placed in the upper flow of the Fe2O3@TiO2 nanotube powder. The obtained black product was immersed into 4 M HCl solution for 24 h to remove the excess iron oxides. Then the HGN@TiO2@C was collected by centrifugation and dried at 60 °C for 12 h. TiO2@C and NCNT were prepared as control samples by replacing the Fe2O3@TiO2 nanotubes with hollow TiO2 nanotubes and the hollow Fe2O3 polyhedral material, respectively.
Synthesis of S/HGN@TiO2@C. To obtain the uniform S/HGN@TiO2@C, a mixture of HGN@TiO2@C and sulfur powder (1[thin space (1/6-em)]:[thin space (1/6-em)]5, weight ratio) was sealed in a glass vessel under argon protection, and heated at 300 °C for 5 h.
Adsorption tests of the LiPSs. The Li2S8 catholyte was prepared by the chemical reaction between sulfur and lithium sulfide (Li2S + 7S → Li2S8). In a typical process, 4.48 g sulfur and 0.92 g Li2S were dissolved in 20.0 mL DOL/DME solution (VDOL[thin space (1/6-em)]:[thin space (1/6-em)]VDME = 1[thin space (1/6-em)]:[thin space (1/6-em)]1) with 2.0 wt% LiNO3 additives in a 50.0 mL bottle and kept stirring overnight in an Ar-filled glove box. Then, this suspension was heated at 80 °C in a vacuum oven inside the glove box for one day to yield the Li2S8 catholyte (1.0 M) with a red–brown color. The Li2S8 solution (1.0 M) was first diluted to 2.0 mM for further use. After that, powders of different samples with the same mass (20.0 mg) were added to 2.0 mM Li2S8 solution (4.0 mL), respectively. After aging for 3.0 h inside the glove box, the supernatant liquid was sealed in cylinder quartz for the UV-vis absorption spectroscopy analysis.
Kinetics of Li2S precipitation on the host materials. For the Li2S precipitation test, the electrode was prepared by casting slurries of the samples, conductive carbon, and PVDF (7[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]1 weight ratio) on carbon paper by the doctor-blade technique. After drying at 50 °C under vacuum overnight, the electrode was cut into wafers with a diameter of 12.7 mm. A coin cell was assembled with the active materials as the cathodes, respectively, lithium as the anode, and a Celgard membrane as the separator. The loading of Li2S6 catholyte for the Li2S precipitation test was 1.0 mg cm−2. The cathode and the anode sides were supplemented with 15.0 μL electrolytes, respectively. All the assembled coin cells were aged at room temperature for 12.0 h. After that, the cell was first discharged galvanostatically at 0.1C to 2.12 V and then discharged potentiostatically at 2.05 V for Li2S nucleation and growth. The current vs. time curve was collected for the kinetic analysis.

Material characterization

The morphology of the obtained samples was investigated by a LEO 1530 field emission SEM instrument and a JEOL-2100 TEM instrument (JEOL, GmbH, Eching, Germany) at 200 kV. XRD patterns were collected in the Bragg–Brentano geometry on a Bruker D8 Advance diffractometer with Cu Kα radiation using a zero-background holder and a step size of 0.03°/step and a measuring time of 1 s/step. The chemical states of the elements in the samples were characterized using X-ray photoelectron spectroscopy (XPS) with an ESCA-Lab-220i-XL X-ray photoelectron spectrometer (Thermo Fisher Scientific) with Al Kα sources (hν = 1486.6 eV).

Electrochemical measurements

CR2032 coin cells were assembled with the cathode, the Li foil as the anode, and a piece of Celgard membrane as the separator in an Ar-filled glove box (UNIlab plus, M. BRAUN) with a H2O content <0.5 ppm and O2 content <0.5 ppm. Also, 1.0 M lithium bis(trifluormethylsulfonyl)amid in a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 volume ratio of 1,3-dioxolane (DOL)/1,2-dimethoxyethane (DME) with 2.0 wt% LiNO3 was used as the electrolyte. The cathode and the anode side were supplemented with 15.0 μL electrolyte, respectively. The electrolyte/sulfur ratio was maintained at ∼15 μL mg−1. Before the electrochemical tests, all the cells were aged at room temperature under the open circuit potential for 12.0 h to let the electrolyte wet the electrode. In this work, the current density of 1.0C equals 1675.0 mA g−1. The specific capacity was calculated based on the mass of sulfur. The galvanostatic charge and discharge were conducted on a LAND battery tester (CT2001A) at room temperature. The CV curves of the assembled coin cells were measured using an Autolab electrochemical workstation (PGSTAT302N potentiostat).

DFT calculations

The density functional theory (DFT) calculations were performed using the Vienna Ab initio Simulation Package (VASP), with the generalized gradient approximation (GGA) Perdew–Burke–Ernzerhof (PBE) functional to describe electron exchange and correlation. The projector-augmented plane wave (PAW) potentials were used to describe the core–valence electron interaction and valence electrons were considered using a plane wave basis set with a kinetic energy cut-off of 500 eV. The vdWs interaction was included using the empirical DFT-D3 method. Partial occupancies of the Kohn–Sham orbitals were allowed using the Gaussian smearing method and a width of 0.05 eV. The electronic energy was considered self-consistent when the energy change was smaller than 10−5 eV. Geometry optimization was considered convergent when the force change was smaller than 0.02 eV Å−1. A k-points sampling of 2 × 2 × 1 with the Monkhorst–Pack scheme was used in all calculations and all the calculations considered the spin-polarization effect.

Results and discussions

The synthesis process for S/HGN@TiO2@C followed a multi-step reaction process, as illustrated in Fig. 1a. Uniform hexagonal MIL88 nanorods were prepared using a simple hydrothermal method, as reported previously.27 The scanning electron microscopy (SEM) images in Fig. S1a and S1b (ESI†) show that the MIL88 nanorods had a smooth surface, were highly uniform, and had a diameter of ∼450 nm and a length of ∼2.5 μm. Subsequently, the surface of the MIL88 nanorods was uniformly coated with a thin layer of amorphous TiO2, resulting in MIL88@TiO2 nanorods, as shown in Fig. S1c and S1d (ESI†). The rough surface and the larger diameters (about 500 nm) of MIL88@TiO2 compared with that of pure MIL88 indicated that the TiO2 layer had been successfully coated on with a thickness of about 25 nm. Additionally, the transmission electron microscopy (TEM) images and TEM-energy dispersive X-ray spectroscopy (EDS) mapping of MIL88@TiO2 further confirmed the existence of TiO2, as shown in Fig. S2.† In order to obtain hollow Fe2O3@TiO2 nanotubes, a calcination process was utilized for MIL88@TiO2 in air to transform MIL88 into hollow Fe2O3 nanorods inside TiO2 through a Kikendall effect. As shown in Fig. S3 (ESI†), the SEM and TEM images clearly revealed the hollow and core–shell structure of the Fe2O3@TiO2 nanotubes. Also, the X-ray diffraction (XRD) patterns of Fe2O3@TiO2 in Fig. S3d† demonstrated the presence of Fe2O3 and TiO2. For preparing the N-doped HGN inside the TiO2 nanotubes and the transformation of TiO2 into a biphasic anatase-rutile homojunction, a CVD strategy was used to realize the synthesis. As shown in Fig. S4 (ESI†), TiO2 was transformed into the anatase and rutile phases and Fe2O3 was reduced to Fe3O4 nanoparticles. Then, the CVD product was etched with HCl solution to remove the extra Fe3O4 catalysts, and the final sulfur host comprising biphasic TiO2 homojunction nanotubes containing an N-doped honeycomb graphite network with a covering carbon layer (HGN@TiO2@C) was obtained. Fig. S5 (ESI†) shows the XRD pattern of HGN@TiO2@C, confirming the presence of anatase and rutile TiO2, with no other materials detected. The SEM images at different magnifications in Fig. 2b–d clearly reveal that HGN@TiO2@C had a uniform morphology of hexagonal nanorods (∼300 nm in diameter) with a core–shell structure, consisting of a thin TiO2 layer on the outside and HGN inside. Fig. 2e shows a low-magnification TEM image of HGN@TiO2@C, without Fe3O4 nanoparticles inside, indicating the successful removal of the catalyst. The high-resolution TEM (HRTEM) image in Fig. 2f revealed the coexistence of anatase and rutile TiO2, forming homojunctions with perfect lattice matching. Moreover, a thin carbon layer shell with a thickness of 2–3 nm covered the surface of the TiO2 homojunctions. The TiO2 homojunction maintains close contact with the thin carbon layer, providing efficient pathways for fast electron/ion transfer during charging/discharging, thereby improving the electrochemical behavior, reaction mechanisms, and kinetics. First, forming the rich homojunction interfaces/boundaries at the nanoscale can offer faster Li+ pathways and decrease the Li+-diffusion barriers during the cycle process, which can significantly improve the reaction kinetics for Li–S batteries. Second, the thin TiO2 homojunction layer on the surface of the thin carbon layer greatly increases the electronic conductivity between the TiO2 homojunction and thin carbon layer, which can facilitate the electrochemical reactions between Li+ and the active materials of sulfur. Third, the TiO2 homojunction with different Fermi level positions can generate a strong built-in electric field, inducing the fluent diffusion of LiPSs, which can promote the rate capacity, and cycle life during cycling processes.28Fig. 2g displays a fractured HGN@TiO2@C, revealing the exposed inner structure consisting of highly graphitized, porous graphite nanoflakes, which was consistent with the SEM images. Therefore, it was confirmed that we had successfully synthesized the HGN@TiO2@C composite, featuring honeycomb graphite on the inside and TiO2 homojunctions on the outside, as a sulfur host. The sulfur active material, anchored closely on the graphitized porous graphite nanoflakes, ensures structural stability, a large specific surface area, and improved electrical conductivity and electrochemical kinetics, while accommodating the volume expansion of sulfur. Additionally, hollow TiO2@C, without graphite nanoflakes inside, and pure N-doped carbon nanotubes (NCNT), without TiO2, were prepared for the control experiments. The detailed characterizations of the hollow TiO2@C and NCNT are shown in Fig. S6 and S7 (ESI†).
image file: d5ta04041h-f1.tif
Fig. 1 (a) Schematic of the synthesis process of the S/HGN@TiO2@C composite. (b–d) SEM images of the HGN@TiO2@C composite. (e–g) TEM images of the HGN@TiO2@C composite.

image file: d5ta04041h-f2.tif
Fig. 2 (a–d) High-resolution XPS spectra of C 1s, N 1s, O 1s and Ti 2p in the HGN@TiO2@C composite. (e) Raman spectrum of the HGN@TiO2@C composite. (f) BET isotherm plots of the HGN@TiO2@C composite.

In order to aid a further comprehension of the advantages of different structures and compositions, X-ray photoelectron spectroscopy (XPS) was used to examine the chemical bonding states of the HGN@TiO2@C composite. The high-resolution C 1s spectrum in Fig. 2a exhibited four subpeaks, with the two peaks at around 284.6 and 285.1 eV assigned to C–C sp2 and C[double bond, length as m-dash]N,29 verifying the heteroatom doping of N in the carbon matrix. The other two peaks at 286.2 and 290.5 eV corresponded to C–O and C[double bond, length as m-dash]C bonds, respectively.30 The N-doping originated from the decomposition and carbonization of 2-methylimidazole, and the chemical status of N could be further uncovered by the high-resolution N 1s spectrum in Fig. 2b, where the three peaks located at 397.6, 399.6 and 400.7 eV could be attributed to pyridinic N, pyrrolic N and graphitic N, respectively.31 The O 1s spectrum in Fig. 2c showed three different peaks at around 530.6, 531.6 and 533.1 eV, in accordance with O–O, C–O and Ti–O bonds, respectively.32Fig. 2d shows the three characteristic peaks of TiO2. Next, the Raman spectrum of HGN@TiO2@C was analyzed to confirm the status of the graphite in Fig. 2e. The two peaks at approximately 1376 and 1594 cm−1 originated from the D band and G band of carbon. The ratio of ID/IG ≈0.82 (less than 1) and the presence of a 2D band at approximately 2800 cm−1 suggested a high degree of graphitization.33 The Eg peak at around 634 cm−1 was attributed to TiO2.34 Next, nitrogen-adsorption isotherm plots were measured to explore the specific surface property, as shown in Fig. 2f. HGN@TiO2@C displayed a typical type IV isotherm with complex type H2 + H3 hysteresis loops and possessed a high specific surface area of 125.6 m2 g−1, which mainly resulted from the unique structure of the honeycomb graphite network.

Sulfur was loaded into the HGN@TiO2@C host using a modified vapor-phase infusion method. It could be observed from the SEM images in Fig. 3a–c that the S/HGN@TiO2@C composite retained its original hexagonal rod-like structures with a smooth surface, indicating there was no significant sulfur particle aggregation. The XRD patterns in Fig. 3d confirmed the presence of elemental sulfur. The thermogravimetry (TG) curve in Fig. 3e clearly revealed the high loading of 82.1% sulfur. TEM images of the S/HGN@TiO2@C composites were also collected, which showed no distinct changes compared to HGN@TiO2@C (Fig. 3f and g). The TEM-EDS mapping further confirmed the presence of Ti, O, N, and S, also showing that sulfur and N-doped graphite nanoflakes were both present within the S/HGN@TiO2@C composite. Moreover, the HGN@TiO2@C structures retained their morphology well after the melt–diffusion reaction, demonstrating their structural stability. HGN with a narrow pore channel structure could not only prevent irregular sulfur agglomeration but also enhances the conductivity of the electrode and provides a buffer for sulfur expansion during cycling.


image file: d5ta04041h-f3.tif
Fig. 3 (a–c) SEM images, (d) XRD patterns, (e) TG curve, (f and g) TEM images and (h) EDS mapping of the S/HGN@TiO2@C composite.

To evaluate the adsorption capability of HGN@TiO2@C toward lithium polysulfides, DFT calculations were performed to investigate the chemisorption mechanism between LiPSs and the as-synthesized samples. Fig. 4a and S8–10 (ESI†) show the optimized adsorption configurations of different sulfur species on the surfaces of anatase/rutile TiO2, anatase TiO2 and rutile TiO2. The atomic model configuration clearly shows the effective interaction between anatase/rutile-TiO2 and the polysulfides based on the chemical bonding of Ti–S and Li–O. The corresponding binding energies of various sulfur species on these surfaces are shown in Fig. 4a. It was evident that the binding energies of anatase/rutile TiO2/LiPSs were more negative than those of anatase TiO2/LiPSs and rutile TiO2/LiPSs, indicating that anatase/rutile TiO2 had the strongest adsorption capability for LiPSs. Moreover, stronger chemical interaction between LiPSs and anatase/rutile TiO2 could form, further enhancing the chemisorption mechanism between LiPSs and the synthesized anatase/rutile TiO2.35 Meanwhile, a visualized adsorption test was carried out by adding different samples with the same weight (5 mg) into sealed vials containing Li2S6, as shown in Fig. 4b. After 6 h (Fig. 4c), the solution containing HGN@TiO2@C became nearly colorless, demonstrating its strongest adsorption ability for Li2S6, while the pure Li2S6 showed no significant color change. Additionally, the related UV-visible absorption spectra of the different Li2S6 solutions after the experiments were also collected to allow a direct comparison among them, as shown in Fig. 4d. It was evident that all the solutions exhibited a reduced intensity of the characteristic Li2S6 peak around 450 nm, but the peak had nearly completely disappeared in the spectrum for the solution containing HGN@TiO2@C, indicating its strong chemisorption capability for lithium polysulfides.


image file: d5ta04041h-f4.tif
Fig. 4 (a) Binding energies of various sulfur species adsorbed on anatase/rutile TiO2, anatase TiO2 and rutile TiO2. (b and c) Adsorption experiments of different lithium polysulfide products. (d) Related UV-visible absorption spectra. (e–g) Potentiostatic discharge curves at 2.05 V for Li2S nucleation on the HGN@TiO2@C, anatase/rutile TiO2, and HGN electrodes. (h) Nyquist plots of S/HGN@TiO2@C, S/NCNT and S/TiO2@C after the first cycle. (i) Schematic of the Li+ migration pathways and band diagrams for anatase TiO2/rutile TiO2.

The catalytic ability of polysulfide conversion in the electrode is crucial for achieving fast redox-reaction kinetics, which can enhance the electrochemical performance. In order to clearly understand the positive catalytic effect of HGN@TiO2@C on promoting the polysulfide redox kinetics in Li–S batteries, Li2S precipitation experiments were performed. Fig. 4e–g shows the time-dependent reduction profiles of the different electrodes. The HGN@TiO2@C electrode delivered the highest capacity for Li2S conversion (140.3 mA h g−1), compared to the TiO2 heterostructure (76.2 mA h g−1) and NCNT (60.1 mA h g−1). Furthermore, the nucleation peaks of Li2S on the HGN@TiO2@C electrode were much sharper than those of the TiO2 heterostructure and the NCNT electrodes. These results demonstrated that HGN@TiO2@C could significantly reduce electrochemical polarization and facilitate the reduction of LiPSs through its catalytic effect, thereby accelerating the conversion kinetics in Li–S batteries.36,37

Finally, electrochemical impedance spectroscopy (EIS) was used to evaluate the conductivity of the HGN@TiO2@C, TiO2@C, and NCNT electrodes, as shown in Fig. 4h. It was evident that the HGN@TiO2@C electrode had a significantly lower charge-transfer resistance than the NCNT and TiO2@C electrodes, indicating its superior conductivity. This could be attributed to the strong built-in electric field arising from the different intrinsic bandgaps of anatase and rutile TiO2, created by the rich TiO2 homojunction interfaces at the nanoscale, which facilitate excellent charge transfer through the O–O band at the homojunction interface,38 as shown in Fig. 4i. Additionally, the rich TiO2 homojunction interfaces/boundaries at the nanoscale can also offer faster Li+ pathways and decrease the Li+-diffusion barriers during the cycling process,39–42 which can further improve the conductivity of the HGN@TiO2@C electrode, as shown in Fig. 4i. Moreover, conductive HGN within the conductive carbon layer provides a pathway for electron transport to the TiO2–polysulfide interface, triggering the redox reaction of soluble polysulfides, which could significantly improve the electrochemical performance of lithium–sulfur batteries.

Next, the electrochemical performance of the S/HGN@TiO2@C cathodes for LSBs was evaluated in coin-type cells. Cyclic voltammetry (CV) measurements were performed in the voltage window ranging from 1.6–2.8 V at a scan rate of 0.1 mV s−1, as shown in Fig. 5a. Two distinct reduction peaks could be located at ∼2.32 and ∼2.02 V in the cathodic scans, corresponding to sulfur reduction. The peak at ∼2.32 V corresponded to the polymerization of sulfur into long-chain lithium polysulfide (Li2Sn, 4 < n < 8), while the second reduction peak at ∼2.02 V indicated the further reduction of the as-formed LiPSs to short-chain LiPSs. In the anodic curve, the strong peak located at ∼2.38 V corresponded to a rearrangement of cyclo-S8. The nearly overlapping cathodic and anodic peak positions at different scan rates suggested the robust kinetics of S/HGN@TiO2@C. Next, galvanostatic charge/discharge voltage profiles at different current densities of S/HGN@TiO2@C were investigated (Fig. 5b). The galvanostatic charge/discharge voltage profiles showed two distinct discharge plateaus (2.4–2.1 V) and a long charge plateau (2.2–2.4 V), which closely matched the peaks observed in the CV curves. Furthermore, the GCD curves of S/HGN@TiO2@C at high current rates still well maintained the featured shape with the typical charge–discharge plateaus, implying rapid LiPSs conversion reactions. To evaluate the rate capability, the S/HGN@TiO2@C electrode was tested at various current densities, as shown in Fig. 5c. The S/HGN@TiO2@C electrode delivered specific capacities of 826.9, 659.4, 577.6, 497.5, and 651.3 mA h g−1 (average of 10 cycles) at current densities of 0.5, 1, 2, 5, and 0.5C, respectively. Additionally, long-cycle testing is crucial for evaluating the lifespan and stability of Li–S batteries. As shown in Fig. 5d, the cycling performances of S/HGN@TiO2@C, S/TiO2@C, and S/NCNT were evaluated at 1C. Note that the test was performed at a current density of 0.5C for the first 3 cycles to facilitate slow activation of the S/HGN@TiO2@C electrodes. The capacity of the S/HGN@TiO2@C electrode was ≈601.7 mA h g−1 in the fourth cycle and ≈626.8 mA h g−1 after 200 cycles at a high current rate of 1C, corresponding to a capacity retention of ≈104.2%. In addition, the coulombic efficiency remained at ≈99.1% over 200 cycles. For comparison, the control samples, S/TiO2@C and S/NCNT, retained capacities of ≈639.8 and 577.3 mA h g−1 in the fourth cycle and ≈443.5 and 325.8 mA h g−1 after 200 cycles, with coulombic efficiencies of 98.3% and 96.7%, corresponding to 71.2% and 54.2% capacity retention, respectively, which were much lower than those of the S/HGN@TiO2@C electrode. The morphologies of the HGN@TiO2@C/S composite structures before and after 200 cycles at 1.0C were examined by SEM, as shown in Fig. S11 (ESI†). It was found that the HGN@TiO2@C/S electrode was well preserved without any apparent destruction, cracks or joint points, demonstrating the advantages of rapid charge transfer among the active materials, conductive carbon and binders during the electrochemical lithium insertion/extraction processes. These results highlight the superiority of S/HGN@TiO2@C as a sulfur host material for advanced Li–S batteries.


image file: d5ta04041h-f5.tif
Fig. 5 (a) CV curves of the S/HGN@TiO2@C composite. (b) Galvanostatic charge–discharge voltage profiles of the S/HGN@TiO2@C composite at different rates. (c) Rate performance of the S/HGN@TiO2@C composite. (d) Cycling performance of S/HGN@TiO2@C, S/NCNT and S/TiO2@C at 1C.

Conclusions

In summary, we successfully designed biphasic TiO2 homojunction nanotubes containing an N-doped honeycomb graphite network with a covering carbon layer (HGN@TiO2@C) as an efficient multifunctional sulfur host material for use as a separator modifier for Li–S batteries. Experimental characterizations and theoretical calculations demonstrated that the TiO2 homojunction, with near-perfect lattice matching and rich interfaces/boundaries, along with the conductive HGN embedded in the carbon layer, could effectively confine sulfur and provide additional buffer space for the electrochemical reactions. This structure not only enhances the electronic and ionic conductivity but also improves the adsorption and catalytic abilities, thereby facilitating the sulfur redox kinetics. Additionally, the unique porous structure of HGN enhances the overall conductivity of the S/HGN composite, providing a high specific surface area for sulfur loading and for accommodating volume changes during the charge–discharge process. Consequently, the HGN@TiO2@C/S cathode delivered a remarkable discharge capacity of 601.7 mA h g−1 in the fourth cycle and ≈626.8 mA h g−1 after 200 cycles at 1C, with a capacity retention of ≈104.2%. We expect that this work will promote the development of highly adsorptive, catalytic, and robust sulfur host materials to achieve high-performance Li–S batteries.

Data availability

All relevant data are within the manuscript and its additional files.

Author contributions

Shidi Huang: conceptualization, methodology, supervision, writing – original draft, writing – review & editing, and funding acquisition. Xuan Zhao and Zheqian Yu: methodology, formal analysis, data curation, and visualization. Weiye Tong and Yijie Zhang: investigation, formal analysis, and validation.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.

Acknowledgements

This work was supported by the State Key Laboratory for Advanced Fiber Materials (Donghua University) (KF2517) and the Research Foundation of Shanghai Institute of Technology (grant no. YJ2022-37).

References

  1. Y. Guo, Q. Niu, F. Pei, Q. Wang, Y. Zhang, L. Y. Du, Y. Zhang, Y. S. Zhang, Y. Y. Zhang, L. Fan, Q. Y. Zhang, L. X. Yuan and Y. H. Huang, Energy Environ. Sci., 2024, 17, 1330–1367 RSC .
  2. T. Wang, J. R. He, Z. Zhu, X. B. Cheng, J. Zhu, B. G. Lu and Y. P. Wu, Adv. Mater., 2023, 35, 2303520 CrossRef CAS .
  3. G. Zhou, H. Chen and Y. Cui, Nat. Energy, 2022, 7, 312–319 CrossRef CAS .
  4. Z. Shen, X. Jin, J. Tian, M. Li, Y. Yuan, S. Zhang, S. Fang, X. Fan, W. Xu and H. Lu, Nat. Catal., 2022, 5, 555–563 CrossRef CAS .
  5. S. Y. Zhou, J. Shi, S. G. Liu, G. Li, F. Pei, Y. H. Chen, J. X. Deng, Q. Z. Zheng, J. Y. Li, C. Zhao, I. H. Hwang, C. J. Sun, Y. Z. Liu, Y. Deng, L. Huang, Y. Qiao, G. L. Xu, J. F. Chen, K. Amine, S. G. Sun and H. G. Liao, Nature, 2023, 621, 75–81 CrossRef CAS .
  6. C. X. Xian, S. H. Shen, T. Q. Yang, Z. Qiu, Y. Q. Zhang, F. G. Cao, X. Q. Liang, M. H. Chen, X. P. He, Y. Xia, C. Wang, W. J. Wan, W. K. Zhang, X. H. Xia, J. P. Tu and J. C. Zhou, ACS Nano, 2024, 18, 27451–27464 CrossRef .
  7. F. L. Chu, M. Wang, J. M. Liu, Z. Q. Guan, H. Y. Yu, B. Liu and F. X. Wu, Adv. Funct. Mater., 2022, 32, 2205393 CrossRef CAS .
  8. W. Q. Wang, H. Li, C. Ding, Z. H. Guo and G. C. Wang, J. Energy Storage., 2023, 71, 108072 CrossRef .
  9. X. Guan, Y. J. Zhao, H. J. Pei, M. J. Zhao, Y. Wang, X. P. Zhou, M. G. Mohamed, S. W. Kuo and Y. Ye, Chem. Eng. J., 2023, 473, 144733 CrossRef CAS .
  10. A. Eftekhari and D. W. Kim, J. Mater. Chem. A, 2017, 5, 17734–17776 RSC .
  11. L. Zhou, D. L. Danilov, R. A. Eichel and P. H. L. Notten, Adv. Energy Mater., 2011, 11, 2001304 CrossRef .
  12. Y. Zhu, J. Wang, C. Xie, M. Yang, Z. Zheng and R. Yu, Mater. Chem. Front., 2020, 4, 2971–2975 RSC .
  13. J. Yu, J. Xiao, A. Li, Z. Yang, L. Zeng, Q. Zhang, Y. Zhu and L. Guo, Angew. Chem., Int. Ed., 2020, 59, 13071–13078 CrossRef CAS .
  14. S. Ng, M. Y. L. Lau and W. Ong, Adv. Mater., 2021, 33, 2008654 CrossRef CAS .
  15. H. C. Li, X. Y. Wang, H. W. Ma, D. Y. Guo, L. H. Wu, H. L. Jin, X. A. Chen and S. Wang, ChemElectroChem, 2024, 11, e202300696 CrossRef CAS .
  16. J. L. Cheong, C. Hu, W. Liu, M. F. Ng, M. B. Sullivan and J. Y. Ying, Nano Energy, 2022, 102, 107659 CrossRef CAS .
  17. K. Y. Zhang, Z. Q. Zhao, H. Chen, Y. K. Pan, B. Niu, D. H. Long and Y. Y. Zhang, Small, 2024, 21, 2409674 CrossRef .
  18. T. Wang, J. R. He, Z. Zhu, X. B. Cheng, J. Zhu, B. G. Lu and Y. P. Wu, Adv. Mater., 2023, 35, 2303520 CrossRef CAS .
  19. S. Z. Huang, Z. H. Wang, Y. V. Lim, Y. Wang, Y. Li, D. H. Zhang and H. Y. Yang, Adv. Energy Mater., 2021, 11, 2003689 CrossRef CAS .
  20. T. Zhou, W. Lv, J. Li, G. Zhou, Y. Zhao, S. Fan, B. Liu, B. Li, F. Kang and Q. H. Yang, Energy Environ. Sci., 2017, 10, 1694–1703 RSC .
  21. X. Dong, X. Liu, P. K. Shen and J. Zhu, Adv. Funct. Mater., 2023, 33, 2210987 CrossRef CAS .
  22. X. Li, Y. Z. Zuo, Y. Z. Zhang, J. Wang, Y. L. Wang, H. M. Yu, L. Zhan, L. C. Ling, Z. G. Du and S. B. Yang, Adv. Energy Mater., 2024, 14, 2303389 CrossRef CAS .
  23. X. J. Guan, S. C. Zong and S. H. Shen, Nano Res., 2022, 15, 10171–10184 CrossRef CAS .
  24. L. Ma, Y. Q. Zhang, S. A. Zhang, L. Wang, C. X. Zhang, Y. J. Chen, Q. Wu, L. B. Chen, L. J. Zhou and W. F. Wei, Adv. Funct. Mater., 2023, 33, 2305788 CrossRef CAS .
  25. Z. Yang, C. X. Peng, R. I. Meng, L. H. Zu, Y. T. Feng, B. J. Chen, Y. L. Mi, C. Zhang and J. H. Yang, ACS Cent. Sci., 2019, 5, 1876–1883 CrossRef CAS .
  26. Y. Guo, Q. Niu, F. Pei, Q. Wang, Y. Zhang, L. Du, Y. Zhang, Y. Zhang, Y. Zhang, L. Fan, Q. Zhang, L. Yuan and Y. Huang, Energy Environ. Sci., 2024, 17, 1330–1367 RSC .
  27. Z. Cui, Q. Liu, J. Zhu, H. Wang, M. Gao, W. Wang, M. F. Yuen, J. Hu, H. Chen and R. J. Zou, Nano Lett., 2023, 23, 1600–1607 CrossRef CAS .
  28. H. Li, C. Chen, Y. Yan, T. Yan, C. Cheng, D. Sun and L. Zhang, Adv. Mater., 2021, 33, 2105067 CrossRef CAS .
  29. Y. Yamada, H. Tanaka, S. Kubo and S. Sato, Carbon, 2021, 185, 342–367 CrossRef CAS .
  30. D. M. Itkis, D. A. Semenenko, E. Y. Kataev, A. I. Belova, V. S. Neudachina, A. P. Sirotina, M. H. vecker, D. Teschner, A. K. Gericke, P. Dudin, A. Barinov, E. A. Goodilin, Y. Shao-Horn and L. V. Yashina, Nano Lett., 2013, 13, 4697–4701 CrossRef CAS .
  31. Y. Zhou, L. Zhang, W. Huang, Q. Kong, X. Fan, M. Wang and J. Shi, Carbon, 2016, 99, 111–117 CrossRef CAS .
  32. C. Li, M. Gu, M. Gao, K. Liu, X. Zhao, N. Cao, J. Feng, Y. Ren, T. Wei and M. Zhang, J. Colloid Interface Sci., 2022, 609, 341–352 CrossRef CAS .
  33. S. X. Yang, D. B. Jiang, Q. Su, S. X. Yuan, Y. Guo, K. J. Duan, M. W. Xiang, J. M. Guo, W. Bai and S. L. Chou, Adv. Energy Mater., 2024, 14, 2400648 CrossRef CAS .
  34. J. R. Jokisaari, D. Bayerl, K. Zhang, L. Xie, Y. Nie, D. G. Schlom, E. Kioupakis, G. W. Graham and X. Pan, Chem. Mater., 2015, 27, 7896–7902 CrossRef CAS .
  35. X. Liang, Y. Rangom, C. Y. Kwok, Q. Pang and L. F. Nazar, Adv. Mater., 2017, 29, 1603040 CrossRef .
  36. J. Zhang, Z. Xie, W. Xi, Y. Zhang, R. Wang, Y. Gong, B. He, H. Wang and J. Jin, Adv. Energy Mater., 2024, 14, 2401792 CrossRef CAS .
  37. J. Cai, Z. Sun, W. Cai, N. Wei, Y. Fan, Z. Liu, Q. Zhang and J. Sun, Adv. Funct. Mater., 2021, 31, 2100586 CrossRef CAS .
  38. H. Wang, Y. Wei, G. Wang, Y. Pu, L. Yuan, C. Liu, Q. Wang, Y. Zhang and H. Wu, Adv. Sci., 2022, 9, 2201823 CrossRef CAS .
  39. X. R. Shi, J. L. Li, X. J. Zhang, M. J. Li, Q. Jing, G. Z. Fang and M. Q. Long, J. Phys. Chem. C, 2024, 128, 6189–6197 CrossRef CAS .
  40. K. Chen, Y. Y. Zhang, Z. X. Fan, J. F. Zhu, Q. C. Xu and J. Xu, Adv. Funct. Mater., 2025, 35, 2500574 CrossRef .
  41. X. F. Leng, K. C. Yang, L. P. Sun, J. Weng and J. Xu, Angew. Chem., Int. Ed., 2025, 64, e202416992 CrossRef CAS .
  42. K. Chen, Y. X. Zhu, Z. J. Huang, B. Han, Q. C. Xu, X. L. Fang and J. Xu, ACS Nano, 2024, 18, 34791–34802 CrossRef CAS .

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta04041h
This journal is © The Royal Society of Chemistry 2025
Click here to see how this site uses Cookies. View our privacy policy here.