TiO₂/Porous Carbon Coated Celgard 2400 Separators Enhance Lithium/Sulfur Battery Performance

Abstract

The practical use of lithium/sulfur batteries is limited by the migration of soluble polysulfides (Li₂S_n, 4 ≤ n ≤ 8) from cathode to anode, which undermines cell stability. In this study, we fabricated a TiO₂/porous carbon (TiO₂/PC) composite coating on a Celgard 2400 separator, acting as a robust polysulfide barrier. The highly conductive PC, with a three‑dimensional ordered porous framework, physically confines polysulfides while also serving as an auxiliary current collector. TiO₂ nanoparticles anchored on the PC surface chemically adsorb polysulfides during charge–discharge cycles. This dual physical‑chemical strategy delivers an initial discharge capacity of 926 mAh g⁻¹ at 0.1 C and retains 75 % of the capacity after 150 cycles. Rate‑capability tests show the modified cell recovers 728 mAh g⁻¹ at 0.1 C after high‑rate cycling, maintaining ~88 % of the initial reversible capacity.

Background

Lithium/sulfur (Li/S) batteries have attracted attention as next‑generation energy storage devices due to their exceptionally high theoretical energy density (2,600 Wh kg⁻¹) and specific capacity (1,675 mAh g⁻¹)¹. Their additional advantages include low toxicity, inexpensive raw materials, and abundant natural sulfur². Nonetheless, several intrinsic challenges hinder commercial deployment: (i) the insulating nature of elemental sulfur (σ₂₉₈ = 5 × 10⁻³⁰ S cm⁻¹) leads to poor active‑material utilization; (ii) the volumetric expansion when Li₂S forms from S₈ causes capacity fade; and (iii) the dissolution and migration of polysulfides in the electrolyte reduce Coulombic efficiency and accelerate capacity decay^3,4. Extensive research has focused on confining sulfur within cathode hosts such as porous carbon, inorganic oxides, and polymers^5–13, yet high loading of trapping materials inevitably diminishes overall energy density. Consequently, strategies that modify the internal architecture of the cell, particularly the separator, have gained traction. Coating interlayers on separators have been shown to impede polysulfide shuttling through physical absorption^14,15 and chemical interactions with polar metal oxides like CeO₂ or layered double hydroxides^18–22. The TiO₂ surface, in particular, has demonstrated strong polar–polysulfide interactions both experimentally and theoretically^23,24. Here, we present a TiO₂/PC composite coating on Celgard 2400 that synergistically combines physical confinement and chemical binding to suppress the shuttle effect.

Methods

Preparation of Li/S Battery with TiO₂/PC-Modified Separator

Preparation of Porous Carbon

Figure 1 illustrates the fabrication route of the TiO₂/PC‑modified Celgard 2400 separator. Monodisperse silica microspheres were synthesized via TEOS hydrolysis in ammonia, followed by ethanol centrifugation. After natural drying, the silica opal was impregnated with a resol solution and pyrolyzed at 600 °C for 2 h under Ar (ramp 2 °C min⁻¹). The process yielded a carbonized template with ~11 % weight loss. Subsequent HF etching removed the silica framework, producing a three‑dimensional ordered porous carbon (PC) scaffold.

TiO₂/Porous Carbon Coated Celgard 2400 Separators Enhance Lithium/Sulfur Battery Performance

Deposition of TiO₂ on PC

The TiO₂ precursor was prepared via a sol–gel route. TTIP (2.84 g, 0.1 mol), HCl (2.4 g), and ethyl alcohol (4.0 g) were mixed and stirred for 1.5 h, forming a transparent gel. The PC template was soaked in this solution for 24 h, then dried for 3 days and calcined at 450 °C for 1 h under N₂ to anchor TiO₂ nanoparticles onto the carbon matrix.

Preparation of the TiO₂/PC-Modified Separator

A slurry of 0.7 g TiO₂/PC, 0.2 g carbon black, and 0.1 g PVDF in NMP was coated onto commercial Celgard 2400 and dried at 50 °C overnight in vacuum. The resulting TiO₂/PC layer was 37 µm thick, with an areal loading of ~0.5 mg cm⁻². Disks of 1 cm diameter were punched for cell assembly.

Material Characterizations

Crystalline phases were identified by powder XRD (Cu–Kα, λ = 1.5406 Å, 2θ = 10–90°). Morphology was examined via SEM (JSM‑7100F) and TEM (JEM‑2100F, 200 kV). Contact angles were measured with a JGW‑360Y meter. XPS (Kratos AXIS Ultra DLD, Al–Kα) assessed surface chemistry before and after cycling.

Electrochemical Measurements

Sulfur cathodes were prepared by mixing 0.8 g S, 0.1 g carbon black, and 0.1 g PVDF in NMP, coated onto Al foil, and dried at 60 °C under vacuum. Sulfur loading was ~2.0 mg cm⁻². Coin cells (CR2025) were assembled in an Ar glove box (≥ 99.9995 %) with 1 M LiTFSI in a 1:1 (v/v) DOL/DME electrolyte. Charge–discharge cycling was conducted between 1.5 and 3 V using a Neware BTS‑5V5mA tester at room temperature.

Results and Discussion

Figure 2 displays the XRD pattern of the TiO₂/PC-modified separator, confirming anatase TiO₂ (JCPDS No.21‑1272) and characteristic carbon peaks at ~23° (002) and ~44° (100).

SEM and TEM images (Figure 3) reveal a uniform, ordered porous network (~110 nm pores) with TiO₂ nanoparticles evenly dispersed on the PC matrix. A lattice spacing of 0.35 nm, matching the (101) facet of anatase TiO₂, confirms the crystalline nature of the embedded nanoparticles.

Figure 4a shows the nitrogen adsorption–desorption isotherm of TiO₂/PC, yielding a BET surface area of 263 m² g⁻¹. The pore size distribution (Figure 4b) highlights micropores around 1 nm and a broad mesoporous range, reinforcing the physical confinement capability.

Figure 5a presents the XPS survey spectrum after cycling, confirming the presence of O, Ti, C, and S. High‑resolution spectra (Figure 5b–d) identify C–C/C=C (284.6 eV), O–C=O (290.4 eV), S–Ti (162.9 eV), S 2p_2/3 (163.9 eV), S 2p_1/2 (165.0 eV), sulfate (170.4 eV), –SO₃ (167.0 eV), C–S (169.0 eV), Ti–S (458.25 eV), Ti 2p_2/3 (459 eV), and Ti 2p_1/2 (464.7 eV). The Ti–S bond indicates a strong chemical interaction between TiO₂ and polysulfides.

Figure 6a demonstrates the mechanical flexibility of the TiO₂/PC-coated separator. Contact‑angle measurements (Figure 6b) show a complete wetting (0°) on the modified surface versus 38° on the pristine separator, reflecting the enhanced electrolyte infiltration afforded by the polar TiO₂/PC composite.

CV curves (Figure 7) for cells with and without the TiO₂/PC separator were recorded at 0.1 mV s⁻¹. The modified separator shifts the cathodic peaks to higher potentials (2.27 V and 1.97 V) and reduces the anodic peak (2.44 V), indicating suppressed polarization and enhanced reaction kinetics.

Galvanostatic charge/discharge curves at 0.1 C (Figure 8) display two plateaus at 2.27 V and 1.97 V, corresponding to the two‑step reduction of S₈ and subsequent conversion to Li₂S₂ / Li₂S. The first cycle delivers 1,060 mAh g⁻¹; subsequent cycles maintain 926 mAh g⁻¹ (second cycle) and 853 mAh g⁻¹ (third cycle), evidencing robust cycling stability.

Cycling performance (Figure 9) shows the modified cell retains ~75 % (708 mAh g⁻¹) of its initial reversible capacity after 150 cycles at 0.1 C, whereas the pristine separator degrades rapidly. At 1 C (Figure 10), the modified separator delivers 788 mAh g⁻¹ initially and 564 mAh g⁻¹ after 300 cycles, underscoring its superior long‑term stability.

Rate capability (Figure 11) demonstrates reversible capacities of 823, 672, 578, and 455 mAh g⁻¹ at 0.1, 0.5, 1, and 2 C, respectively. After high‑rate cycling, the capacity recovers to 728 mAh g⁻¹ at 0.1 C, preserving ~88 % of the initial capacity. The pristine separator exhibits markedly lower capacities, confirming the effectiveness of the TiO₂/PC layer in enhancing sulfur utilization and curbing polysulfide migration.

Self‑discharge behavior was examined after 72 h rest following the initial three cycles at 0.1 C. The unmodified separator shows a 0.21 V drop (2.28→2.07 V), indicating significant polysulfide migration. In contrast, the TiO₂/PC-modified separator experiences only a 2.6 % voltage decline (2.30→2.24 V), reflecting its superior ability to trap polysulfides and mitigate self‑discharge.

Conclusions

We have demonstrated that a TiO₂/PC composite coating on Celgard 2400 effectively suppresses the polysulfide shuttle in Li/S batteries. The TiO₂ nanoparticles provide strong electrostatic attraction (S–Ti–O), while the ordered porous carbon matrix offers high electrical conductivity and physical confinement. The resulting cell delivers an initial specific capacity of 926 mAh g⁻¹ and retains 75 % after 150 cycles, achieving superior cycling stability and rate performance. This separator‑modification strategy presents a practical route toward high‑performance Li/S energy storage devices.