A 405 W h kg⁻¹ Ah-level lithium–sulfur pouch battery stabilized over 200 cycles by an electron-triode-like GeS₂–NiS₂ heterostructure

Abstract

Lithium–sulfur batteries (LSBs) form soluble polysulfides (LiPSs) during discharge, leading to decline in cycling performance, especially the failure of pouch batteries. The failure may be due to the fact that conventional sulfur hosts can only adsorb LiPSs and cannot rapidly inject and transfer electrons in electrochemical reactions. The sluggish electrochemical interconversion of LiPSs leads to continuous loss of active sulfur materials, which is a barrier to long-life commercial LSBs. Herein, an electron-triode-like GeS₂–NiS₂ heterostructure is successfully designed and synthesized to serve as a catalytic sulfur host. An Ohmic contact rather than a Schottky contact is formed between GeS₂ and NiS₂, which is proven using the ultraviolet photoelectron spectra and X-ray absorption fine structure spectra. Therefore, the LiPSs can be interconverted with an electron-triode-like model: NiS₂ acts as the emitter and injects a batch of electrons into the LiPSs (the collector) collectively through the GeS₂ base electrode, with a maximum reaction current amplification factor (β_R) of 105.87. In situ XRD and ex situ AFM indicate that the collective injection of electrons can achieve an earlier deposition of Li₂S as early as ∼80% of SOC. Ultimately, the S@GeS₂–NiS₂/rGO battery achieves a high specific capacity of 1007.8 mA h g⁻¹ at 0.5C. The 1.2 Ah pouch battery can achieve a high energy density of 405 W h kg⁻¹ and work stably for 200 cycles, highlighting its great potential for practical applications.

---

Broader context

The decarbonization of transportation and grid infrastructure urgently demands electrochemical energy storage systems transcending the 350 W h kg⁻¹ threshold. In addition, in order to reduce the dependence on heavy metals such as Co, a sulfur material, which is abundant in resources, has become a widely researched cathode. While lithium–sulfur batteries (LSBs) theoretically satisfy this demand through sulfur's ultrahigh capacity (1675 mA h g⁻¹) and global abundance (≈5% of terrestrial crust), their commercialization is hindered by two interlinked sustainability challenges, such as shuttling losses of lithium polysulfides (LiPSs) and excessive electrolyte usage (E/S ratio > 5 μL mg⁻¹), contradicting green chemistry principles. Addressing these challenges, we propose a GeS₂–NiS₂ heterostructured sulfur host mimicking solid-state triode operation: the Ohmic contact creates an electron-rich region at the interface of the heterostructure, to rapidly inject batch of electrons into soluble LiPSs. The electron injection effect enables a LiPS conversion reaction current amplification factor (β_R) of 105.87. Therefore, earlier Li₂S nucleation from 20% to 80% of SOC via the electron-triode-like GeS₂–NiS₂ heterostructure. The developed 1.2 Ah pouch batteries deliver a 405 W h kg⁻¹ energy density, which is one of the available methods to achieve these targets, and the state-of-the-art 200 cycles with 81.2% capacity retention under lean electrolyte conditions.

1. Introduction

The quest for low-cost and high-energy-density devices has prompted extensive research into high-quality rechargeable batteries beyond traditional lithium-ion batteries.^1–3 Among the prospective candidates, lithium–sulfur batteries (LSBs) have the advantages of high theoretical specific capacity (1675 mA h g⁻¹), abundant sulfur source, and inexpensive cost.^4–7 Nevertheless, the application of LSBs is still hindered by poor cycling stability due to their intrinsic defects.^8–10 The inherently low electronic conductivity of sulfur and its product Li₂S results in slow redox kinetics and low rate capacity. The intermediate products of the reaction, soluble lithium polysulfides (LiPSs), are prone to migrate from the cathode to the anode in the liquid electrolyte, thereby corroding lithium metal and leading to an irreversible loss of active materials.^11–14 Consequently, LSBs suffer from insufficient reaction kinetics, poor cycle life, and rapid capacity decay during cycling.

To tackle these issues, various catalysts have been developed as sulfur hosts.^15–19 These studies have successfully strengthened the ability to adsorb LiPSs, thereby extending the life-span of LSBs.^20–24 However, in a practical high-load lithium–sulfur pouch battery, the electrons will not transfer/distribute evenly in the electrodes: the electrons will concentrate at the edge regions and the tab-near regions (Scheme 1(a)).²⁵ The uneven electron distribution should be responsible for the failure of the pouch battery. Taking discharge as an example, S₈ will undergo multiple soluble LiPSs, including Li₂S₈, Li₂S₆, and Li₂S₄, and ultimately form the final solid product Li₂S. Therefore, LiPSs will generate unevenly, and will dissolve in the electrolyte and migrate to the anode side, eventually leading to rapid capacity degradation and long-life cycle failure of the pouch battery.^26–28

Scheme 1 (a) and (b) Schematic diagram of the lithium–sulfur pouch battery, and the current will result in an uneven electron distribution manner because of the continuous reaction of the transferred electrons. (a) LiPSs are generated unevenly due to the uneven distribution of electrons. (b) This work develops a method to inject a batch of electrons from the electrode to LiPSs, to accelerate the multi-electron experience through soluble LiPSs, achieving earlier Li₂S deposition. (c) Conventional heterostructures work separately: one component is used as the LiPS adsorbent and the other is used as the LiPS conversion catalyst. (d) In this work, an electron-triode-like GeS₂–NiS₂ heterostructure with an Ohmic contact forms an electron-rich region at the heterointerfaces to collectively inject into S₈ to obtain Li₂S without the accumulation of soluble LiPSs. (e) Electron-triode-like GeS₂–NiS₂ heterostructure achieves 105.87 times current amplification (β_R = 105.87) during cycling, greatly retarding the accumulation of soluble LiPSs.

As the unevenly distributed electrons are inevitable, it is important to develop methods to inject a batch of electrons from the electrode to the LiPSs, to accelerate the multi-electron experience through the soluble LiPSs, and realize an earlier Li₂S deposition (Scheme 1(b)). In this regard, we designed and synthesized an electron-triode-like GeS₂–NiS₂ heterostructure to realize the collective injection of electrons by amplifying the reaction current. Unlike conventional heterostructures, where one component works as the LiPS adsorbent and the other works as the LiPS conversion catalyst (Scheme 1(c)), the GeS₂–NiS₂ heterostructure possesses an “Ohmic contact” behavior. In the theory of semiconductor physics, the contacts for the two components in a heterostructure include Schottky contacts and Ohmic contacts.^29,30 For Schottky contacts, the interface forms a potential barrier with very low conductivity to electrons, called the electron barrier layer. For Ohmic contacts, the interface has a very low resistance and is a highly conductive region for electrons, called the anti-blocking layer. In the designed electron-triode-like GeS₂–NiS₂ heterostructure, the “Ohmic contact” between the metallic component (NiS₂) and the semiconductor (GeS₂), as shown in Scheme 1(d), will form a highly conductive heterointerface. In this electron-rich region, a large number of free-moving electrons are gathered.³¹ When LiPSs are close to the heterointerface, abundant S–S and Li–S bonds formed can effectively adsorb LiPSs through the orbital effect. The electrons accumulated at the interface can be rapidly transferred into the LiPSs, and the conversion of LiPSs is facilitated by the successive sequential action of electrons. During the electrochemical process, the base GeS₂ can effectively adsorb LiPSs, the emitter NiS₂ can rapidly inject a batch of electrons, and the collector LiPSs can receive these electrons to amplify the reaction current (β_R = 105.87) (Scheme 1(e)). In addition, reduced graphene oxide (rGO) is incorporated into the GeS₂–NiS₂ heterostructure. rGO has been reported to serve as an excellent electron conductor. Although it possesses a lower work function, its low polarity nature ensures that it does not interfere with the designed Ohmic contact between GeS₂ and NiS₂, which originates from the orbital interaction between GeS₂ and NiS₂.^32–34

Because of the above advantages, the GeS₂–NiS₂ heterostructure can support a long-term cycling and achieve robust LBSs. As expected, the S@GeS₂–NiS₂/rGO cathode battery exhibits a high specific capacity of 878.5 mA h g⁻¹ at a high current density of 2C, and an extremely low-capacity decay of 0.046% per cycle after 1000 long-term cycles. The 1.2 Ah pouch battery constructed with GeS₂–NiS₂/rGO can achieve a high energy density of 405 W h kg⁻¹ and provide a stable cycling capacity of 200 cycles.

2. Results and discussion

2.1. Structural characterization and analysis

GeS₂–NiS₂/rGO was synthesized by high-temperature calcination followed by a multiple-step solvothermal method with GeO₂ as a Ge source, anhydrous NiCl₂ as a Ni source, and anhydrous ethanol as a solvent. The reference samples GeS₂/rGO and NiS₂/rGO were also synthesized using similar methods with GeO₂ as the Ge source and anhydrous NiCl₂ as the Ni source, respectively (For details, see the Experimental section in the ESI†). The material ratios were optimized based on the cyclic voltammetry (CV) of symmetric batteries and cycle capacity tests of the GeS₂–NiS₂/rGO heterostructure (Fig. S1, ESI†). Coupled plasma-optical emission spectroscopy (ICP-OES) was used to determine the ratios of GeS₂ and NiS₂ in the heterostructure (Table S1, ESI†). The heterostructure with three different ratios (GeS₂ : NiS₂ = 1 : 1, 1 : 2 and 1 : 3) were prepared in this work. As shown in Fig. S1a (ESI†), when GeS₂ : NiS₂ = 1 : 2, the GeS₂–NiS₂/rGO heterostructure exhibits the best catalytic performance for Li₂S₆. In addition, the S@GeS₂–NiS₂/rGO battery has the highest discharge capacity and the best capacity retention when GeS₂ : NiS₂ = 1 : 2 (Fig. S1b, ESI†). In the heterostructure, electrons are transferred from NiS₂ to GeS₂, forming an electron-rich region at the interface. NiS₂ has high electrical conductivity and can effectively enhance the electron transfer efficiency. However, a 1 : 1 ratio has an insufficient NiS₂ content, and a 1 : 3 ratio may result in too little GeS₂, which affects electron acceptance and structural stability. Therefore, a heterostructure with a 1 : 2 ratio of GeS₂ to NiS₂ was chosen as a sulfur host for the study, achieving improved catalytic performance and excellent electrochemical cycling performance.

The morphology and structure of GeS₂–NiS₂/rGO are characterized to understand the microstructure and fine heterointerfaces. From Fig. 1(a), we can confirm that the GeS₂–NiS₂/rGO exists as uniformly dispersed nanoparticles from scanning electron microscopy (SEM). The transmission electron microscopy (TEM) image in Fig. 1(b) further indicates that GeS₂–NiS₂ nanoparticles (70–90 nm) are firmly attached onto the rGO sheets. The corresponding energy dispersive X-ray (EDX) elemental mappings in Fig. 1(c) show that the elements S, Ni and Ge are distributed evenly. The high-resolution TEM (HRTEM) image (Fig. 1(d)) and corresponding fast Fourier transform (FFT) patterns (Fig. 1(e)) exhibit the crystal structure of the GeS₂–NiS₂/rGO material. The lattice distances of 0.58 and 0.28 nm can be attributed to the (111) plane of orthorhombic GeS₂ (Fdd2) and the (200) plane of cubic NiS₂ (Pa3̄), respectively.

Fig. 1 (a) SEM image of GeS₂–NiS₂/rGO. (b) TEM image and (c) corresponding EDX elemental mappings of GeS₂–NiS₂/rGO. (d) HRTEM image of GeS₂–NiS₂/rGO and (e) corresponding reduced FFT patterns. (f) XRD patterns of GeS₂–NiS₂/rGO, NiS₂/rGO and GeS₂/rGO. (g) XRD patterns of S@GeS₂–NiS₂/rGO, S@NiS₂/rGO and S@GeS₂/rGO. (h) Raman spectra of GeS₂–NiS₂/rGO, NiS₂/rGO and GeS₂/rGO. (i) The D-band and G-band of GeS₂–NiS₂/rGO, NiS₂/rGO, GeS₂/rGO and GO.

Powder X-ray diffraction (XRD) spectra provide further evidence of the successful synthesis of GeS₂–NiS₂/rGO. Fig. 1(f) clearly shows the coexistence of GeS₂ (JCPDS: 40-0443) and NiS₂ (JCPDS: 11-0099) in GeS₂–NiS₂/rGO.^35,36 Moreover, the Rietveld refinement result of the GeS₂–NiS₂/rGO heterostructure is given in Fig. S2 (ESI†). The calculated refinement reliability parameters of R_wp = 8.76% and R_p = 7.85% indicate that the refinement results are reliable, and the crystal structure is determined to be NiS₂ and GeS₂ phases. According to the Rietveld refinement results, the content of GeS₂ and NiS₂ in the heterostructure can be calculated to be 55.1% and 26.4%, respectively, which is consistent with the ICP results. The corresponding diffraction peaks of GeS₂ and NiS₂ are also exhibited in the XRD pattern after sulfur loading (Fig. 1(g)), which further indicates the successful synthesis of the S@GeS₂–NiS₂ material.

GeS₂–NiS₂/rGO, NiS₂/rGO and GeS₂/rGO are used as sulfur hosts by a melt impregnation method and tested with a thermogravimetric (TG) analyzer (Fig. S3, ESI†). The test results show that the sulfur content of three materials is 72.9%, 70.2% and 73.2%, respectively. The composition and molecular structure of GeS₂–NiS₂/rGO are further characterized by Raman spectroscopy (Fig. 1(h and i)). Fig. 1(h) shows two typical Raman peaks at 341 and 480 cm⁻¹, corresponding to GeS₂ and NiS₂, respectively.^37,38 Notably, in GeS₂–NiS₂/rGO, NiS₂/rGO and GeS₂/rGO, the intensity ratios of the D-band to G-band are all higher than that of GO, which confirms the reduction of GO (Fig. 1(i)).³⁹ The above results suggest the successful formation of the GeS₂–NiS₂/rGO material.

2.2. Heterointerface states of GeS₂–NiS₂

To understand the correlation of changes in the electronic structure, density functional theory (DFT) calculations were performed. Density of states (DOS) results are illustrated for GeS₂–NiS₂ (Fig. 2(a)), NiS₂ (Fig. 2(b)) and GeS₂ (Fig. 2(c)), respectively. The calculation results indicate that NiS₂ has no band gap and exhibits metallic properties, and GeS₂ has a significant band gap and is a semiconductor. The GeS₂–NiS₂ heterostructure exhibits a combination of GeS₂ and NiS₂, where a more contribution from Ni can be observed at the region near the Fermi level (E_f). The differences in DOS are evident throughout the E_f, where GeS₂ shows poor electron transfer ability due to the wide band gap, while NiS₂ and GeS₂–NiS₂ provide high electron conductivity without a band gap across the E_f. Thus, the calculations of DOS indicate that the formation of heterointerfaces in GeS₂–NiS₂ improves electrical conductivity during sulfur redox reactions.

Fig. 2 (a)–(c) Calculated pDOS near the Fermi level of GeS₂–NiS₂ (a), NiS₂ (b) and GeS₂ (c). The combination of the n-type semiconductor GeS₂ with high conductivity NiS₂ results in a highly conductive region with a very low contact resistance at the heterointerfaces. (d) Differential charge density and planar average density in the Z-direction of GeS₂–NiS₂ (yellow: electron accumulation; cyan: electron depletion). The strong bonding interactions of the GeS₂–NiS₂ heterostructure form rich catalytic heterointerfaces. (e) High-resolution Ni 2p XPS spectra of GeS₂–NiS₂/rGO and NiS₂/rGO. (f) High-resolution Ge 3d XPS spectra of GeS₂–NiS₂/rGO and GeS₂/rGO. The electron transfer from NiS₂ to GeS₂ in the GeS₂–NiS₂ heterostructure. (g) Schematic diagram of band structures of NiS₂ and GeS₂ before and after contacts (E_o: vacuum energy level, E_Fm: Fermi level, E_Fs: band gap, E_c: conduction band, and E_v: valence band). The Ohmic contact between GeS₂ and NiS₂ forms an electron-rich region at the GeS₂–NiS₂ heterointerfaces.

Based on experimental characterization, GeS₂ (111) and NiS₂ (200) models were conducted to further clarify the catalytic effect of heterointerfaces at the atomic level by DFT calculations. Fig. 2(d) shows the differential charge density and planar average density in the Z-direction of GeS₂–NiS₂. Both the interfacial charge gain/loss region and the plane-averaged density amplitude are significant, indicating the electron transfer between GeS₂ and NiS₂. In the GeS₂–NiS₂ heterostructure, electron transfer at the interface leads to a change in the center of charge distribution. Bader charge analysis calculated the charge redistribution at the heterointerfaces and quantified an average gain of 0.57 electrons for GeS₂ from NiS₂. An electron exchange region of approximately 10 Å was identified in the GeS₂–NiS₂ heterostructure. This finding indicates that the S–S bonds effectively bridge the GeS₂ and NiS₂ phases through orbital interactions, thereby facilitating the formation of the Ohmic contact.

The charge transfer was also confirmed by comparing the independent X-ray photoelectron spectroscopy (XPS) spectra of NiS₂, GeS₂, and GeS₂–NiS₂ (Fig. 2(e, f) and Fig. S4, ESI†). The band corresponding to adventitious C–C at 284.8 eV is used as a reference. The peaks located at 284.8, 286.4 and 288.9 eV are attributed to C–C/C=C, C–O and O–C=C, respectively (Fig. S4a, c and e, ESI†).⁴⁰ For the S 2p high resolution spectrum, the peaks located at 163.0 and 164.3 eV are assigned to the 2p_3/2 and 2p_1/2 of S²⁻ and S₂²⁻, respectively (Fig. S4b, d and f, ESI†).^41,42 In the Ni 2p spectrum of GeS₂–NiS₂/rGO, the fitted XPS peaks at 854.5 and 856.9 eV are Ni 2p_3/2, and the peaks at 871.6 and 875.4 eV are Ni 2p_1/2, accompanied by three shake-up satellite peaks centered at 858.9, 862.1 and 88.0.6 eV (Fig. 2(e)).^43,44 Similar peaks are also observed in the NiS₂/rGO material. As expected, compare with NiS₂/rGO, significant positive shifts of Ni 2p are observed in the GeS₂–NiS₂/rGO material. In addition, the Ge 3d spectrum in Fig. 2(f) shows that the peak at 31.6 eV is typical of the bonding of the Ge⁴⁺ ion in GeS₂.⁴⁵ Similarly, the Ge 3d spectrum of GeS₂–NiS₂/rGO shows an obvious negative shift compared to that of GeS₂/rGO. These shifts indicate the electron loss of NiS₂ when in contact with GeS₂ at the heterointerfaces, suggesting the electron gain of GeS₂ and the charge transfer at the heterointerfaces, which agrees with the DFT results.

Ultraviolet photoelectron spectra (UPS) were used to further characterize the surface electron transfer of GeS₂–NiS₂/rGO (Fig. 2(g) and Fig. S5, ESI†). According to the UPS results shown in Fig. S5 (ESI†), the work function (E_Φ) of NiS₂/rGO (6.51 eV) is lower than that of GeS₂/rGO (7.92 eV). The Mott–Schottky plot was measured in the dark from an alternating current frequency of 1 KHz to determine the semiconductor type of GeS₂ (Fig. S6, ESI†). As shown in Fig. S6 (ESI†), a positive slope of the plot around 0 V can be observed, indicating that GeS₂ is an n-type semiconductor.⁴⁶ According to the theory of semiconductor physics, based on the combination of the n-type semiconductor GeS₂ and higher conductivity NiS₂, there exists an Ohmic contact between GeS₂ and NiS₂ without a Schottky barrier.⁴⁷ The contact resistance of the heterointerfaces is very small, creating an electron-rich region, which reduces the power dissipation of GeS₂–NiS₂ and increases the electron mobility.⁴⁸

According to the DFT calculation of adsorption energy (Fig. S7–S10, ESI†), all the sulfur species (Li₂S, Li₂S₂, Li₂S₄, Li₂S₆, Li₂S₈, and S₈) have the strongest adsorption capacities (E_b) when adsorbed onto GeS₂–NiS₂, compared to those when adsorbed onto NiS₂ and GeS₂, which suggests that the heterointerfaces of GeS₂–NiS₂ are effective cooperative sites for LiPSs. In the GeS₂–NiS₂ heterostructure, the electron-rich region is formed at the interfaces by the orbital effect of S–S bonds, which can accumulate a large number of free-moving electrons. In addition, when LiPSs are close to the GeS₂–NiS₂ heterostructure, the orbital effect of S and Li in LiPSs can form abundant S–S and Li–S bonds with S in the heterostructure (Fig. S7, ESI†). Therefore, the abundant S–S and Li–S bonds at the heterointerface can effectively adsorb LiPSs through the orbital effect. As a result, the electrons accumulated at the heterointerface can be rapidly transferred to LiPSs to promote their catalytic conversion and achieve the rapid generation of Li₂S. Therefore, compared to NiS₂ and GeS₂, the GeS₂–NiS₂ heterostructure exhibits enhanced adsorption ability for LiPSs (Figure S10, ESI†).

X-ray absorption fine structure (XAFS) spectroscopy is used to further demonstrate the electron transfer behavior in the electron-triode-like GeS₂–NiS₂ heterostructure. Fig. 3(a) exhibits the X-ray absorption near-edge structure (XANES) spectra of the Ge K-edge in GeO₂, Ge foil, GeS₂/rGO and GeS₂–NiS₂/rGO, respectively. Compared with GeS₂/rGO, the K-edge of Ge in GeS₂–NiS₂/rGO shifts towards a lower energy value, suggesting that Ge is reduced to a lower oxidation state, which is consistent with the XPS results for Ge 3d (Fig. 2f).⁴⁹ In the meantime, as shown in Fig. 3(b), after the formation of the GeS₂–NiS₂/rGO heterostructure, the Ni K-edge shifts to a higher energy value compared to NiS₂/rGO, indicating that Ni loses electrons and is oxidized to a higher state, corresponding to the XPS results of Ni 2p (Fig. 2(e)).⁵⁰ These results demonstrate that electrons transfer from NiS₂ to GeS₂ in the heterostructure, forming highly conductive heterointerfaces with free-moving electrons.

Fig. 3 (a) Ge K-edge XANES spectra of GeO₂, Ge foil, GeS₂/rGO and GeS₂–NiS₂/rGO. (b) Ni K-edge XANES spectra of NiO, Ni foil, NiS₂/rGO and GeS₂–NiS₂/rGO. Electron transfer from NiS₂ to GeS₂ in the GeS₂–NiS₂ heterostructure. (c) Ge R-space EXAFS analysis of GeO₂, Ge foil, GeS₂/rGO and GeS₂–NiS₂/rGO. (d) Ni R-space EXAFS analysis of NiO, Ni foil, NiS₂/rGO and GeS₂–NiS₂/rGO. (e) WT contour plots of Ge for Ge foil, GeS₂/rGO and GeS₂–NiS₂/rGO at R-space. (f) WT contour plots of Ni for Ni foil, NiS₂/rGO and GeS₂–NiS₂/rGO at R-space. The successfully formed electron-triode-like GeS₂–NiS₂/rGO heterostructure forms an electron-rich region, which will facilitate the collective injection of electrons during reaction processes.

According to the R-space of the extended X-ray absorption fine structure (EXAFS) for Ge in Fig. 3(c), the main peak appears near 2.0 Å for GeS₂/rGO and GeS₂–NiS₂/rGO, corresponding to the Ge–S bond.⁵¹ The Ge–S peak intensity of GeS₂–NiS₂/rGO is lower than that of GeS₂/rGO, which can be attributed to the formation of S–Ni and Ge–Ni bonds in the GeS₂–NiS₂/rGO heterostructure. For the R-space EXAFS spectrum of Ni (Fig. 3(d)), the main peak near 2.0 Å corresponds to the Ni–S bond in NiS₂/rGO and GeS₂–NiS₂/rGO.⁵² As the peak intensity change in Ge–S, the intensity of the Ni–S bond in GeS₂–NiS₂/rGO is also weaker than that in NiS₂/rGO, because of the formation of S–Ge and Ni–Ge bonds in the heterostructure interface. These results indicate the abundant formation of the GeS₂–NiS₂/rGO heterostructure and the existence of new coordination structures.

The wavelet transform (WT) further proves the spatial structural information in the GeS₂–NiS₂/rGO heterostructure.^53,54 As shown in the WT contours of Ge (Fig. 3(e)), the intensity maximum of the Ge–Ge bond in Ge foil can be observed at k = 9.1 Å⁻¹. In the GeS₂/rGO and GeS₂–NiS₂/rGO samples, the intensity maximum coordination structure can be observed at k = 7.7 Å⁻¹, which can be assigned to the Ge–S bond. Moreover, the GeS₂–NiS₂/rGO sample presents a new coordination structure at k = 13.8 Å⁻¹, referred to as the formation Ge–Ni bond. Interestingly, compared to the Ge–S bond, the intensity of the newly formed Ge–Ni bond is not significant, indicating that the scattering signal of Ge–Ni is relatively weak, which is due to the electron transfer from NiS₂ to GeS₂ in the heterostructure. The WT contours of Ni show a maximum intensity at k = 5.3 Å⁻¹ (Fig. 3(f)), corresponding to Ni–S in NiS₂/rGO and GeS₂–NiS₂/rGO. In addition, a new coordination structure, referred to as the Ni–Ge bond, is formed at k = 11.7 Å⁻¹ in the GeS₂–NiS₂/rGO heterostructure. Unlike the newly formed Ge–Ni bond peak of Ge (Fig. 3(e)) for GeS₂–NiS₂/rGO, the intensity of the Ni–Ge bond is significant, even comparable to the existing Ni–S bond. This is identical to our electron-triode-like material design, where the GeS₂ obtains electrons from NiS₂ in the heterostructure; therefore, the scattering ability of Ge in GeS₂ is relatively high to produce a higher Ni–Ge signal intensity. These results show that in the electron-triode-like GeS₂–NiS₂/rGO heterostructure, electrons are transferred from NiS₂ to GeS₂, and electron-rich heterointerfaces are formed, which will facilitate the collective injection of electrons during electrochemical reactions.

2.3. Electrocatalysis for polysulfide redox

Rapid injection of a batch of electrons can amplify the reaction current during the electrochemical processes, thereby achieving rapid catalytic conversion of LiPSs. CV tests with symmetric batteries were performed using 0.5 M Li₂S₆ electrolyte within a voltage window of −1.0 to 1.0 V, to analyze the redox kinetics of LiPSs in different sulfur hosts (Fig. 4(a)). The GeS₂–NiS₂/rGO material shows the highest peak currents and the lowest polarization voltage compared with the NiS₂/rGO and GeS₂/rGO materials, which indicates that GeS₂–NiS₂/rGO can significantly promote the catalytic conversion of LiPSs.

Fig. 4 (a) CV tests for the symmetrical batteries based on GeS₂–NiS₂/rGO, NiS₂/rGO and GeS₂/rGO materials with a scan rate of 18 mV s⁻¹. The GeS₂–NiS₂/rGO material exhibits the highest peak currents and the lowest polarization voltages, promoting the catalytic conversion of LiPSs. (b) Amplified current plot of GeS₂–NiS₂/rGO versus NiS₂/rGO from the CV results of symmetrical batteries. The electron-triode-like GeS₂–NiS₂/rGO realizes the collective injection of electrons and amplifies the reaction current (β_R = 105.87 and β_O = 30.24). (c) UV-vis absorption spectra of the adsorbed solutions. (d) Li₂S precipitation experiments with the Li₂S₈ cathode catholyte at 2.05 V of GeS₂–NiS₂/rGO, NiS₂/rGO and GeS₂/rGO. (e) Li₂S dissolution experiments of three materials at 2.40 V. (f) CV profiles of S@GeS₂–NiS₂/rGO, S@NiS₂/rGO and S@GeS₂/rGO cathode batteries in the range of 1.7–2.8 V at a scan rate of 0.1 mV s⁻¹. (g) NTR values of three batteries at different current scanning rates. (h) Galvanostatic discharge/charge profiles of three batteries at a current rate of 0.1C and the corresponding (i) R_ct values of the different sample's electrodes.

As the electron injection process from GeS₂–NiS₂/rGO to LiPSs can be treated like an electron triode, current amplification factors (β) for LiPS reduction and oxidation can be quantitatively calculated with eqn (1) and (2), respectively:

where β_R and β_O are the current amplification factors for LiPS reduction and oxidation processes. I_R,h and I_O,h are the reduction and oxidation currents of the GeS₂–NiS₂/rGO heterostructure system (electron triode) from the symmetric battery CV tests, respectively. I_R,e and I_O,e are the reduction and oxidation currents of the NiS₂/rGO system (only the emitter) from the symmetric battery CV tests, respectively. In the GeS₂–NiS₂/rGO heterostructure, both the accumulation of electrons at heterointerfaces and the subsequent directional movement contribute to collective electron injection into LiPSs. As shown in Fig. 4(a), if only the accumulated electron should be responsible for the catalytic LiPS conversion,³¹ the redox currents should be amplified equally for GeS₂–NiS₂/rGO compared with the NiS₂/rGO and GeS₂/rGO batteries. In the designed electron-triode-like GeS₂–NiS₂/rGO, a maximum reduction current of over 100 times (β_R = 105.87) and a maximum oxidation current of over 30 times (β_O = 30.24) can be achieved at specific voltages (Fig. 4(b)). Therefore, the directional movement of the electrons into LiPSs is critical for the electron injection process. This collective electron injection and extraction behavior between GeS₂–NiS₂/rGO and LiPSs can rapidly realize the interconversion between S₈ and Li₂S without detention of the soluble LiPSs, which is important for a practical LSB.

The ultraviolet-visible (UV-vis) absorption spectra of adsorbed solutions can reflect the adsorption capacity of polysulfides.⁵⁵ As can be seen in Fig. 4(c), the Li₂S₆ absorption peak intensity of GeS₂–NiS₂/rGO is the weakest among different materials (GeS₂–NiS₂/rGO < NiS₂/rGO < GeS₂/rGO < rGO). Fig. S11 (ESI†) exhibits the optical images of the adsorption ability for the above materials after 1 and 8 h, respectively. The absorption wavelengths of S₆²⁻ species range from 300 to 500 nm (around 315, 355, and 475 nm).⁵⁶ In the Li₂S₆ solution of rGO, the color and the absorbance associated with Li₂S₆ remain almost unchanged. In the Li₂S₆ solutions of NiS₂/rGO and GeS₂/rGO, the adsorption peaks decrease and the solutions become transparent after the adsorption for 8 h. Compared with other samples, GeS₂–NiS₂/rGO has the lightest color after adsorption, accompany with the weakest adsorption peak from the UV-vis absorption spectra, indicating the strongest adsorption ability for LiPSs.

Except for the adsorption of soluble LiPSs, the ability to convert soluble LiPSs to insoluble Li₂S in LSBs is a more important factor for a sulfur host. To study the advantage of the GeS₂–NiS₂/rGO material on the liquid–solid conversion of LiPSs, the potentiostatic discharge curves at 2.05 V were recorded. The curves were fitted to obtain the deposition of Li₂S (Fig. 4(d)) for GeS₂–NiS₂/rGO, GeS₂/rGO, and NiS₂/rGO, respectively.⁵⁷ The peak current of Li₂S deposition for GeS₂–NiS₂/rGO is much higher and appears earlier (0.65 mA at 3151 s) compared with GeS₂/rGO (0.18 mA at 5022 s) and NiS₂/rGO (0.14 mA at 5145 s). Also, the GeS₂–NiS₂/rGO material shows the highest Li₂S deposition capacity (198.21 mA h g⁻¹), compared with the GeS₂/rGO (167.58 mA h g⁻¹) and NiS₂/rGO (135.12 mA h g⁻¹) materials, indicating rapid LiPS conversion and effective sulfur utilization by GeS₂–NiS₂/rGO. In our previous work, it was confirmed that due to the partial lattice matching nature between Fdd2 GeS₂ and Fm3̄m Li₂S, it can guide the multi-site nucleation and rapid deposition of Li₂S.³⁵ Therefore, the Li₂S deposition and dissolution capacity of GeS₂/rGO is higher than that of NiS₂/rGO. Furthermore, the Li₂S dissociation is another critical indicator to demonstrate the advantages of GeS₂–NiS₂/rGO in promoting the reverse conversion of Li₂S to S₈. As shown in Fig. 4(e), the Li₂S dissolution capacity of GeS₂–NiS₂/rGO (535.92 mA h g⁻¹) is much higher than that of GeS₂/rGO (366.17 mA h g⁻¹) and NiS₂/rGO (286.44 mA h g⁻¹). In general, we can confirm that GeS₂–NiS₂/rGO exhibits excellent catalytic activity and fast redox kinetics for both charge and discharge processes.

The S@GeS₂–NiS₂/rGO, S@NiS₂/rGO and S@GeS₂/rGO cathodes were prepared by a simple melt-diffusion method. Fig. 4(f) exhibits the CV profiles of S@GeS₂–NiS₂/rGO, S@NiS₂/rGO and S@GeS₂/rGO cathode coin batteries with lithium metal as the anode, to demonstrate the redox reaction process in a LSB. The two distinct cathodic peaks at about 2.30 and 2.05 V are attributed to the reduction of solid S₈ to long-chain soluble LiPSs and the subsequent conversion to insoluble Li₂S₂/Li₂S, respectively. The anodic peaks are the oxidation of Li₂S₂/Li₂S to LiPSs and then to S₈.⁵⁸ The CV profile of the S@GeS₂–NiS₂/rGO battery shows the highest peak currents and the lowest overpotentials than those of the S@NiS₂/rGO and S@GeS₂/rGO batteries, suggesting accelerated redox kinetics of LiPS conversion.

The CV profiles of different batteries at various scan rates were carried out to quantitatively evaluate the reaction kinetics of LiPSs and Li₂S according to the nucleation transformation ratio (NTR) (Fig. 4(g) and Fig. S12, ESI†).^59,60

where A is the integral area (A V) of peak I and peak II, respectively. v and C are the scan rate (V s⁻¹) and electron transfer per gram (A s), respectively. The theoretical value of NTR is 3, which implies a complete conversion from the generated soluble LiPSs to Li₂S. Thus, the closer the NTR is to 3, the greater the efficiency on LiPS conversion. All NTR values were calculated for the three batteries at different scanning rates and are exhibited in Fig. 4(g). At all scanning rates, the NTR value of the S@GeS₂–NiS₂/rGO battery is the highest than those of the S@NiS₂/rGO and S@GeS₂/rGO batteries, which indicates a fast catalytic conversion of LiPSs to Li₂S.

Fig. 4(h) displays the galvanostatic discharge and charge curves of S@GeS₂–NiS₂/rGO, S@NiS₂/rGO and S@GeS₂/rGO batteries at 0.1C (1C = 1675 mA g⁻¹). The initial discharge capacity of S@GeS₂–NiS₂/rGO is 1325.4 mA h g⁻¹, which is higher than those of S@NiS₂/rGO (1135.8 mA h g⁻¹) and S@GeS₂/rGO (1013.6 mA h g⁻¹). The polarization voltage between the discharge and charge curves of the battery (ΔE, taken at 50% depth of discharge) is reported in Fig. 4(h) and Fig. S13, ESI.†⁶¹ A smaller ΔE value corresponds to a faster redox kinetics and a smaller Ohmic polarization of a battery. C₁ and C₂ are the capacities of the two discharge platforms, respectively. The ratio of C₂ to C₁ is used as another key indicator to analyze the catalytic activity of different sulfur hosts.⁶² As displayed in Fig. S13 (ESI†), the S@GeS₂–NiS₂/rGO battery exhibits the lowest ΔE and the highest C₂/C₁ compared to those of the S@GeS₂/rGO and S@NiS₂/rGO batteries.

The electron conduction of the materials and electrodes was tested using direct current (DC) resistance and electrochemical impedance spectroscopy (EIS) tests to demonstrate the enhanced conductivity of the GeS₂–NiS₂/rGO heterostructure (Fig. 4(i) and Fig. S14, ESI†). Electronic conductivity of different samples was investigated using a CHI1140C workstation with a constant voltage of 1.0 V over 1500 s (Fig. S14a and Table S2, ESI†). Fig. S14a and Table S2 (ESI†) show that GeS₂–NiS₂/rGO has the largest σ value, demonstrating the enhanced electrical conductivity of the heterostructure. The electronic conductivity of different electrodes, with or without sulfur loading, was tested using EIS. The GeS₂–NiS₂/rGO, NiS₂/rGO, GeS₂/rGO, S@GeS₂–NiS₂/rGO, S@NiS₂/rGO, and S@GeS₂/rGO electrodes were assembled into symmetric batteries (Fig. S14b and c, ESI†). Fig. 4(i) summarizes the charge transfer resistance (R_ct) values for the different electrodes. Among GeS₂–NiS₂/rGO, NiS₂/rGO, and GeS₂/rGO electrodes, the GeS₂–NiS₂/rGO electrode has the smallest R_ct (10.4 Ω), indicating improved electron conduction. What's more, the S@GeS₂–NiS₂/rGO electrode (20.1 Ω) has a smaller R_ct compared to the S@NiS₂/rGO (33.2 Ω) and S@GeS₂/rGO (57.6 Ω) electrodes. These results demonstrate that the GeS₂–NiS₂/rGO heterostructure has excellent electrical conductivity with the best LiPS catalytic and reaction kinetics.

To devote into the understanding of electron injection functions in advancing Li₂S deposition by the electron-triode-like GeS₂–NiS₂/rGO, in situ XRD spectra were recorded during the first discharge/charge processes of the S@GeS₂–NiS₂/rGO (Fig. 5(a)), S@GeS₂/rGO (Fig. 5(b)) and S@NiS₂/rGO (Fig. 5(c)) coin batteries. The major peaks detected at 22.7°, 25.4°, 26.3°, and 27.33° are indexed to be α-S₈.⁶³ The intensity of the characteristic peaks of α-S₈ diffraction decreases gradually as the discharge process proceeds for all the batteries. In the meantime, a distinct broad peak appears at 24–25.5°, corresponding to the phase transition from α-S₈ to LiPSs.^64,65 Then, the characteristic peak of Li₂S appeared at 26.5–27°, indicating the conversion of LiPSs into the final discharge product Li₂S. The abundant S–S and Li–S bonds formed between LiPSs and GeS₂–NiS₂/rGO realized the collective electron injection. As expected, the Li₂S characteristic peak appears earliest (prior ∼ 80% SOC) in S@GeS₂–NiS₂/rGO among the three batteries (∼40% SOC for S@GeS₂/rGO and ∼20% SOC for S@NiS₂/rGO). The earlier Li₂S deposition as well as the extremely high current amplification factors (β_R and β_O) verified that the GeS₂–NiS₂/rGO heterostructure can boost the LiPS conversion to Li₂S via the directional movement of electrons. All these results indicate that the electron-triode-like GeS₂–NiS₂ heterostructure can greatly accelerate electron injection/extraction and promote the catalytic conversion of LiPSs.

Fig. 5 (a)–(c) The first discharge/charge curves and corresponding in situ XRD contour plots of S@GeS₂–NiS₂/rGO (a), S@GeS₂/rGO (b) and S@NiS₂/rGO (c) batteries. The electron-triode-like GeS₂–NiS₂ heterostructure can inject a batch of electrons into LiPSs and achieve earlier Li₂S deposition (∼80% SOC). (d)–(f) Ex situ AFM images of S@GeS₂–NiS₂/rGO (d), S@GeS₂/rGO (e) and S@NiS₂/rGO (f) electrodes cycled to different voltages. The S@GeS₂–NiS₂/rGO electrode consistently maintains a flat and homogeneous surface, indicating non-accumulation and rapid conversion of LiPSs.

The catalytic conversion of LiPSs by the electron-triode-like GeS₂–NiS₂ can greatly maintain the stability of the electrode. We utilize ex situ atomic force microscopy (AFM) to understand the interface evolution of different electrodes in the z-direction during the discharge and charge processes (Fig. 5(d and e) and Fig. S15–S17, ESI†). AFM is a powerful tool for monitoring the surface changes in an electrode, especially for a surface with material conversion and volume changes.⁶⁶ We have introduced the ex situ AFM to monitor the surface evolution for the cathode of a coin LSB for the first time, to understand the GeS₂–NiS₂ function on morphological stability. For a sulfur cathode, when S₈ is first converted to LiPSs, which are soluble to the electrolyte, soluble LiPSs can travel on the electrode surface or shuttle to the electrolyte. Also, the S₈ to 8Li₂S material conversion will experience a ∼79% volume expansion, resulting in a dimension change, especially in the z-direction.⁶⁷ An ideal collective electron injection can realize an earlier Li₂S deposition rapidly and smoothly, by making a dent in the detainment of soluble LiPSs, thereby obtaining a uniform electrode surface during the discharge/charge process. As shown in Fig. 5(d and e), the fresh electrode surfaces of all three materials are uniform and flat, with no obvious particles. During the discharge process, there is no significant change on the surface of the S@GeS₂–NiS₂/rGO electrode (Fig. 5(d)). After discharge to 1.7 V, the surface of the S@GeS₂–NiS₂/rGO electrode remains dense and homogeneous. This is due to the rapid injection of free-moving electrons in the electron-triode-like GeS₂–NiS₂ heterostructure, which results in rapid conversion of S₈ to LiPSs, and finally to the uniformly deposited Li₂S on the surface of the S@GeS₂–NiS₂/rGO electrode. By comparison, the surfaces of S@GeS₂/rGO (Fig. 5(e)) and S@NiS₂/rGO (Fig. 5(f)) electrodes appear as large areas of fluctuation in the z-direction at 2.1 and 2.4 V, respectively, indicating the inhomogeneous S₈ and LiPS conversion on the electrode surfaces. After discharge to 1.7 V, both the S@GeS₂/rGO and S@NiS₂/rGO electrodes form rough surfaces with large peaks and valleys in the z-direction. These results imply that the LiPSs are not evenly generated and converted to Li₂S, resulting in rough electrode surfaces. Among them, the S@NiS₂/rGO electrode exhibits the most aggregated structure, suggesting its maximum coverage of LiPSs, which corresponds to the results of in situ XRD. Similar results are observed during the subsequent charge process (Fig. S17, ESI†), where the S@GeS₂–NiS₂/rGO electrode maintains a relatively flat and homogeneous surface until fully charge to 2.8 V. The surface of the S@GeS₂/rGO and S@NiS₂/rGO electrodes is consistently covered with a large number of inhomogeneous particles, demonstrating the loss and accumulation of LiPSs during cycling. These results indicate that a batch of electrons in the electron-triode-like GeS₂–NiS₂ heterostructure can rapidly inject into soluble LiPSs without accumulation, achieving rapid conversion of LiPSs and earlier deposition of Li₂S in the S@GeS₂–NiS₂/rGO battery.

2.4. Electrochemical performances of LSBs

To study in-depth the charge-carrier kinetics of the different hosts, ex situ EIS measurements were carried out during various states for the S@GeS₂–NiS₂/rGO, S@NiS₂/rGO and S@GeS₂/rGO batteries after 20 cycles (Fig. 6(a, b) and Fig. S18, ESI†). Fig. 6(a) presents the galvanostatic discharge/charge profile with the specified points at 0.1C. The corresponding ex situ EIS curves of S@GeS₂–NiS₂/rGO are given in Fig. 6(b). The fitting EIS results for the S@GeS₂–NiS₂/rGO, S@NiS₂/rGO and S@GeS₂/rGO batteries are summarized in Table S3 (ESI†). During the discharge process, the R_ct value of the S@GeS₂–NiS₂/rGO battery first increases slightly from 14.5 Ω (A point) to 23.6 Ω (C point), then decreases significantly and finally falls to 5.6 Ω (F point). After full charging, the R_ct value of the S@GeS₂–NiS₂/rGO battery decreases to 3.1 Ω. These small R_ct values correspond to the relatively complete transformation of sulfur species during cycling. In contrast, the S@NiS₂/rGO and S@GeS₂/rGO batteries show higher R_ct values (For details, see Table S3, ESI†) and change drastically throughout the discharge/charge processes, suggesting continuous accumulation and severe migration of LiPSs.

Fig. 6 (a) The galvanostatic discharge/charge profile of S@GeS₂–NiS₂/rGO with the specified points for ex situ EIS tests. (b) Ex situ EIS profiles of the S@GeS₂–NiS₂/rGO battery. (c) 40% and (d) 60% SOC profiles of the galvanostatic discharge of the S@GeS₂–NiS₂/rGO battery at 0.5C. (e) Cycling stability of S@GeS₂–NiS₂/rGO, S@GeS₂/rGO and S@NiS₂/rGO batteries at 0.5C for 300 cycles. (f) Rate performances of different batteries at current densities from 0.2 to 4C. (g) Long-term cycling performances of different batteries at 2C over 1000 cycles. Among the three batteries, S@GeS₂–NiS₂/rGO has the highest discharge capacity and the most stable cycling performance, suggesting that LiPS loss during cycling is effectively suppressed.

Due to the inevitable production of soluble LiPSs within different states of charge (SOC) ranges, evaluating the self-discharge performances of batteries is essential (Fig. 6(c, d) and Fig. S19, ESI†). In LSBs, about 60% of SOC corresponds to the conversion of Li₂S₆ to Li₂S₄, and about 40% of SOC corresponds to the reaction of Li₂S₄ to Li₂S₂/Li₂S.⁶⁸ Batteries are discharged to 60% or 40% of SOC for 48 h of standing, after 10 normal cycles. The capacity retention of S@GeS₂–NiS₂/rGO is 98.21% and 98.48% at 40% of SOC (Fig. 6(c)) and 60% of SOC (Fig. 6(d)), respectively, which is higher than that of S@NiS₂/rGO (96.22% at 40% of SOC and 96.64% at 60% of SOC) and S@GeS₂/rGO (94.65% at 40% of SOC and 95.94% at 60% of SOC). These results indicate that the GeS₂–NiS₂/rGO material can effectively reduce the loss of LiPSs and avoid the self-discharge of LSBs.

The effect of the GeS₂–NiS₂ heterostructure on the capacity and cycling performance of LSBs was further evaluated. The initial discharge capacity of S@GeS₂–NiS₂/rGO is 1007.8 mA h g⁻¹ at 0.5C, which is superior to that of S@GeS₂/rGO (951 mA h g⁻¹) and S@NiS₂/rGO (932.5 mA h g⁻¹) (Fig. 6(e)). The S@GeS₂–NiS₂/rGO battery also maintains a high-capacity retention rate of 89.06% after 300 cycles. The rate capabilities of different cathode batteries were tested at current densities ranging from 0.2 to 4C (Fig. 6(f) and Fig. S20, ESI†). The S@GeS₂–NiS₂/rGO battery displays the best rate capacities of 1235, 1089.1, 957.4, 871.4, 817.6, 720.8 and 657.1 mA h g⁻¹ at 0.2, 0.5, 1, 1.5, 2, 3 and 4 C, respectively. What's more, an impressive discharge capacity of 1170.4 mA h g⁻¹ can be achieved in the S@GeS₂–NiS₂/rGO battery when the current density returns to 0.2C. Even after 100 consecutive cycles, the capacity of S@GeS₂–NiS₂/rGO can remain at 1120.8 mA h g⁻¹, achieving an ultra-low-capacity decay. Compared with S@GeS₂/rGO and S@NiS₂/rGO, the S@GeS₂–NiS₂/rGO battery exhibits excellent rate performance. The galvanostatic discharge/charge voltage curves of S@GeS₂–NiS₂/rGO, S@GeS₂/rGO and S@NiS₂/rGO batteries at various current densities are further investigated (Fig. S20, ESI†). Two typical discharge plateaus of S@GeS₂–NiS₂/rGO are observed even at 4C, indicating the superior redox catalytic ability for LiPS conversion. The discharge/charge voltage curves of the S@GeS₂–NiS₂/rGO battery exhibits the smallest polarization potential compared to those of S@GeS₂/rGO and S@NiS₂/rGO batteries, indicating boosted redox kinetics of the GeS₂–NiS₂/rGO material.

The long-term cycling capabilities were measured at a high current rate of 2C to study the cycling stability of the S@GeS₂–NiS₂/rGO battery. As illustrated in Fig. 6(g), the S@GeS₂–NiS₂/rGO battery can maintain a high specific capacity of 474.6 mA h g⁻¹ over 1000 cycles, with a low-capacity decay rate of 0.046% per cycle and coulombic efficiency over 99.5%. In comparison, the S@GeS₂/rGO and S@NiS₂/rGO batteries can only deliver capacities of 300.8 and 354.8 mA h g⁻¹, respectively. These results demonstrate that the designed GeS₂–NiS₂/rGO material can realize a long-cycle LSB via the electron-triode-like approach.

To evaluate the practical feasibility of the GeS₂–NiS₂/rGO material, a lithium–sulfur pouch battery was fabricated using S@GeS₂–NiS₂/rGO as the cathode and a lithium–copper strip (80 m lithium) as the anode. In this lithium–sulfur pouch battery, the sulfur loading is 6.2 mg cm⁻², the electrolyte/sulfur ratio (E/S) is 4 L mg⁻¹, and the negative/positive ratio (N/P) is 1.6 : 1. Detailed information of the pouch cell is provided in the “Pouch battery assembly and Measurements” section of the ESI† and Table S4. The cycling performance (Fig. 7(a)) and the corresponding discharge/charge curves (Fig. 7(b)) are exhibited. The pouch battery attains a high specific capacity and maintains good cycling stability at 0.1C after 200 cycles. In addition, the 1.2 Ah pouch battery with an energy density of up to 405 W h kg⁻¹ can power an electric fun (Fig. 7(c)). These above results indicate that the electron-triode-like GeS₂–NiS₂ heterostructure can successfully increase the life-span of a practical pouch LSB, under the uneven electron distribution nature of from the tab to the electrode. Overall, the GeS₂–NiS₂ heterostructure is highly competitive compared to previously reported pouch batteries (Fig. 7(d)).^69–73

Fig. 7 (a) Cycling performance of the S@GeS₂–NiS₂/rGO pouch battery at 0.1C and (b) corresponding galvanostatic discharge/charge curves. (c) A electric fan is powered by the S@GeS₂–NiS₂/rGO pouch battery. (d) Comparison of the energy density of different lithium–sulfur pouch batteries. The S@GeS₂–NiS₂/rGO battery maintains a high capacity after 200 cycles, indicating its potential electrochemical performance in practical applications.^69–73

3. Conclusions

In this work, an electron-triode-like GeS₂–NiS₂/rGO heterostructure is successfully designed as a sulfur host to improve the electrochemical performance of LSBs. Theoretical calculations and experimental results show that the n-type semiconductor GeS₂ forms an Ohmic contact with the metallic component NiS₂, creating an electron-rich region at the heterointerfaces, which is verified by UPS and XAFS. The GeS₂–NiS₂ heterostructure can inject a batch of electrons into the LiPSs, accelerating the multi-electron experience through the LiPSs without accumulation of soluble LiPSs. Thus, the GeS₂–NiS₂ heterostructure can amplify the reaction current (up to β_R = 105.87) of LSBs during cycling and enable earlier deposition of Li₂S (from ∼20% to ∼80% SOC), which is verified by in situ XRD and ex situ AFM. Because of these outstanding advantages, the specific capacity of the S@GeS₂–NiS₂/rGO battery reaches 878.5 mA h g⁻¹ at a high current density of 2C, and the capacity decay rate is as low as 0.046% after 1000 long-term cycles. Moreover, a 405 W h kg⁻¹ 1.2 Ah pouch battery under high sulfur loading (6.2 mg cm⁻²), lean-electrolyte (E/S = 4 L mg⁻¹) and low N/P ratio (1.6 : 1) conditions is realized. This study elucidates the effectiveness of electron-triode-like GeS₂–NiS₂ in promoting catalytic conversion of LiPSs, providing inspiration for the design and development of heterostructure catalysts for high-performance lithium–sulfur pouch batteries.

Author contributions

All authors have given approval to the final version of the manuscript.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

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

Acknowledgements

The authors thank the National Natural Science Foundation of China (Grant No. 92372202, 22478043 and 22075033) and the Guangzhou Science and Technology Bureau (2024A03J0609).

Notes and references

1. C. D. Quilty, D. Wu, W. Li, D. C. Bock, L. Wang, L. M. Housel, A. Abraham, K. J. Takeuchi, A. C. Marschilok and E. S. Takeuchi, Chem. Rev., 2023, 123, 1327–1363.
2. F. Degen, M. Winter, D. Bendig and J. Tübke, Nat. Energy, 2023, 8, 1284–1295.
3. Y. Che, X. Hu, X. Lin, J. Guo and R. Teodorescu, Energy Environ. Sci., 2023, 16, 338–371.
4. M. Zhao, H.-J. Peng, B.-Q. Li and J.-Q. Huang, Acc. Chem. Res., 2024, 57, 545–557.
5. H. Li, Y. Li and L. Zhang, SusMat, 2022, 2, 34–64.
6. R. Chu, T. T. Nguyen, Y. Bai, N. H. Kim and J. H. Lee, Adv. Energy Mater., 2022, 12, 2102805.
7. K. Liu, H. Yuan, X. Wang, P. Ye, B. Lu, J. Zhang, W. Lu, F. Jiang, S. Gu, J. Chen, C. Yan, Y. Li, Z. Xu and Z. Lu, ACS Appl. Mater. Interfaces, 2023, 15, 31478–31490.
8. F. Zhao, J. Xue, W. Shao, H. Yu, W. Huang and J. Xiao, J. Energy Chem., 2023, 80, 625–657.
9. Z.-X. Chen, Q. Cheng, X.-Y. Li, Z. Li, Y.-W. Song, F. Sun, M. Zhao, X.-Q. Zhang, B.-Q. Li and J.-Q. Huang, J. Am. Chem. Soc., 2023, 145, 16449–16457.
10. R. Fang, K. Chen, Z. Sun, G. Hu, D.-W. Wang and F. Li, Interdiscip. Mater., 2023, 2, 761–770.
11. Z. Li, L.-P. Hou, N. Yao, X.-Y. Li, Z.-X. Chen, X. Chen, X.-Q. Zhang, B.-Q. Li and Q. Zhang, Angew. Chem., Int. Ed., 2023, 135, e202309968.
12. R. Deng, M. Wang, H. Yu, S. Luo, J. Li, F. Chu, B. Liu and F. Wu, Energy Environ. Mater., 2022, 5, 777–799.
13. X. Chen, H. Ji, Z. Rao, L. Yuan, Y. Shen, H. Xu, Z. Li and Y. Huang, Adv. Energy Mater., 2022, 12, 2102774.
14. Y. Liu, C. Cui, Y. Liu, W. Liu and J. Wei, RSC Adv., 2020, 10, 7384–7395.
15. Y. Li, Y. Deng, J.-L. Yang, W. Tang, B. Ge and R. Liu, Adv. Funct. Mater., 2023, 33, 2302267.
16. L.-P. Hou, Y. Li, Z. Li, Q.-K. Zhang, B.-Q. Li, C.-X. Bi, Z.-X. Chen, L.-L. Su, J.-Q. Huang, R. Wen, X.-Q. Zhang and Q. Zhang, Angew. Chem., Int. Ed., 2023, 62, e202305466.
17. T. Wang, J. He, Z. Zhu, X.-B. Cheng, J. Zhu, B. Lu and Y. Wu, Adv. Mater., 2023, 35, 2303520.
18. Y. Kong, L. Wang, M. Mamoor, B. Wang, G. Qu, Z. Jing, Y. Pang, F. Wang, X. Yang, D. Wang and L. Xu, Adv. Mater., 2024, 36, 2310143.
19. A. Eftekhari and D.-W. Kim, J. Mater. Chem. A, 2017, 5, 17734–17776.
20. A. I. Kamisan, T. I. T. Kudin, A. S. Kamisan, A. F. C. Omar, M. F. M. Taib, O. H. Hassan, A. M. M. Ali and M. Z. A. Yahya, Int. J. Hydrogen Energy, 2022, 47, 8630–8657.
21. B. Zhang, C. Luo, Y. Deng, Z. Huang, G. Zhou, W. Lv, Y.-B. He, Y. Wan, F. Kang and Q.-H. Yang, Adv. Energy Mater., 2020, 10, 2000091.
22. W. Yao, W. Zheng, J. Xu, C. Tian, K. Han, W. Sun and S. Xiao, ACS Nano, 2021, 15, 7114–7130.
23. S. Wang, S. Feng, J. Liang, Q. Su, F. Zhao, H. Song, M. Zheng, Q. Sun, Z. Song, X. Jia, J. Yang, Y. Li, J. Liao, R. Li and X. Sun, Adv. Energy Mater., 2021, 11, 2003314.
24. J. Liu, C. Lin, Q. Xie, D.-L. Peng and R.-J. Xie, Chem. Eng. J., 2022, 430, 133099.
25. C. Zhao, H. Jeong, I. Hwang, T. Li, Y. Wang, J. Bai, L. Li, S. Zhou, C. C. Su, W. Xu, Z. Yang, M. Almazrouei, C.-J. Sun, L. Cheng, G.-L. Xu and K. Amine, Joule, 2024, 8, 3397–3411.
26. G. Chen, L. Yan, H. Luo and S. Guo, Adv. Mater., 2016, 28, 7580–7602.
27. X. Long, Z.-H. Luo, W.-H. Zhou, S.-K. Zhu, Y. Song, H. Li, C.-N. Geng, B. Shi, Z.-Y. Han, G.-M. Zhou, W. Lv and J.-J. Shao, Energy Storage Mater., 2022, 52, 120–129.
28. Y. Li, J. Zhang, Q. Chen, X. Xia and M. Chen, Adv. Mater., 2021, 33, 2100855.
29. Y. Zheng, J. Gao, C. Han and W. Chen, Cell Rep. Phys. Sci., 2021, 2, 1–27.
30. G. Jiang, X. Hu and L. Sun, ACS Appl. Electron. Mater., 2023, 5, 3071–3077.
31. H. Chen, Y. Qiu, Z. Cai, W. Liang, L. Liu, M. Li, X. Hou, F. Chen, X. Zhou, T. Cheng, L. He, J. Wang, X. Zhang, S. Dou and L. Li, Angew. Chem., Int. Ed., 2025, e202423357.
32. Q. Su, W. Wang, J. Chen, J. Ji, W. Wang, W. Ren, L. Zhang, J. Xie and Q. An, Adv. Funct. Mater., 2024, 2419594.
33. W.-G. Lim, S. Kim, C. Jo and J. Lee, Angew. Chem., Int. Ed., 2019, 58, 18746–18757.
34. Y. Zhu, S. Murali, M. D. Stoller, K. J. Ganesh, W. Cai, P. J. Ferreira, A. Pirkle, R. M. Wallace, K. A. Cychosz, M. Thommes, D. Su, E. A. Stach and R. S. Ruoff, Science, 2011, 332, 1537–1541.
35. X. Jiao, X. Tang, J. Li, Y. Xiang, C. Li, C. Tong, M. Shao and Z. Wei, Chem. Sci., 2024, 15, 7949–7964.
36. Y. Jin, Y. Liu, L. Song, J. Yu, K. Li, M. Zhang and J. Wang, Chem. Eng. J., 2022, 430, 133029.
37. D. Feng, Q. Liu, Z. Li and T. Zeng, ACS Appl. Energy Mater., 2021, 4, 6096–6105.
38. X. Lou, P. Yan, B. Jiao, Q. Li, P. Xu, L. Wang, L. Zhang, M. Cao, G. Wang, Z. Chen, Q. Zhang and J. Chen, Nat. Commun., 2024, 15, 2730.
39. X. Ou, Z. Xiao, J.-F. Zhang, C. Wang, D. Wang, B. Zhang and Y. Wu, ACS Nano, 2020, 14, 13952–13963.
40. S. Luo, J. Shang, Y. Xu, H. Cheng, L. Zhang and Y. Tang, Adv. Funct. Mater., 2024, 34, 2403166.
41. Z. Xiao, C. Wang, B. Luo, L. Cao, R. Huang, J. Zhang, B. Zhang and X. Ou, Appl. Surf. Sci., 2021, 554, 149596.
42. K. Liang, K. Marcus, S. Zhang, L. Zhou, Y. Li, S. T. De Oliveira, N. Orlovskaya, Y.-H. Sohn and Y. Yang, Adv. Energy Mater., 2017, 7, 1701309.
43. S.-A. He, Z. Cui, Q. Liu, G. He, D. J. Brett, W. Luo, R. Zou and M. Zhu, Small, 2021, 17, 2104186.
44. Y. Lin, Z. Qiu, D. Li, S. Ullah, Y. Hai, H. Xin, W. Liao, B. Yang, H. Fan, J. Xu and C. Zhu, Energy Storage Mater., 2018, 11, 67–74.
45. H. Chen, C. Keiser, S. Du, H.-J. Gao, P. Sutter and E. Sutter, Phys. Chem. Chem. Phys., 2017, 19, 32473–32480.
46. W. Xiong, J. Lin, H. Wang, S. Li, J. Wang, Y. Mao, X. Zhan, D.-Y. Wu and L. Zhang, J. Energy Chem., 2023, 81, 492–501.
47. W. Song, K. C. Chong, G. Qi, Y. Xiao, G. Chen, B. Li, Y. Tang, X. Zhang, Y. Yao, Z. Lin, Z. Zou and B. Liu, J. Am. Chem. Soc., 2024, 146, 3303–3314.
48. N. Zhang, Y. Hu, L. An, Q. Li, J. Yin, J. Li, R. Yang, M. Lu, S. Zhang, P. Xi and C. Yan, Angew. Chem., Int. Ed., 2022, 61, e202207217.
49. Q. Huang, G.-J. Xia, B. Huang, D. Xie, J. Wang, D. Wen, D. Lin, C. Xu, L. Gao and Z. Wu, et al. , Energy Environ. Sci., 2024, 17, 5260–5272.
50. B. Zhang, J. Wang, G. Liu, C. M. Weiss, D. Liu, Y. Chen, L. Xia, P. Zhou, M. Gao and Y. Liu, et al., Nature, Catalysis, 2024, 7, 441–451.
51. I. Pethes, V. Nazabal, R. Chahal, B. Bureau, I. Kaban, S. Belin and P. Jóvári, J. Alloys Compd., 2016, 673, 149–157.
52. J. Yin, J. Jin, H. Zhang, M. Lu, Y. Peng, B. Huang, P. Xi and C.-H. Yan, Angew. Chem., Int. Ed., 2019, 131, 18849–18855.
53. H. Niu, H. Liu, L. Yang, T. Kang, T. Shen, B. Jiang, W.-H. Huang, C.-C. Chang, Y. Pei and G. Cao, et al. , Nat. Commun., 2024, 15, 9421.
54. P. Wang, X. Bai, H. Jin, X. Gao, K. Davey, Y. Zheng, Y. Jiao and S.-Z. Qiao, Adv. Funct. Mater., 2023, 33, 2300687.
55. Y. Cui, X. Zhou, X. Huang, L. Xu and S. Tang, ACS Appl. Mater. Interfaces, 2023, 15, 49223–49232.
56. J. Häcker, D. H. Nguyen, T. Rommel, Z. Zhao-Karger, N. Wagner and K. A. Friedrich, ACS Energy Lett., 2021, 7, 1–9.
57. S. Tian, G. Liu, S. Xu, C. Han, K. Tao, J. Huang and S. Peng, Adv. Funct. Mater., 2024, 34, 2309437.
58. S. Haldar, M. Wang, P. Bhauriyal, A. Hazra, A. H. Khan, V. Bon, M. A. Isaacs, A. De, L. Shupletsov, T. Boenke, J. Grothe, T. Heine, E. Brunner, X. Feng, R. Dong, A. Schneemann and S. Kaskel, J. Am. Chem. Soc., 2022, 144, 9101–9112.
59. R. Xu, H. Tang, Y. Zhou, F. Wang, H. Wang, M. Shao, C. Li and Z. Wei, Chem. Sci., 2022, 13, 6224–6232.
60. T. Wang, Q. Dong, C. Li and Z. Wei, Acta Phys.-Chim. Sin., 2024, 40, 2303061.
61. C. Zhang, R. Du, J. J. Biendicho, M. Yi, K. Xiao, D. Yang, T. Zhang, X. Wang, J. Arbiol, J. Llorca, Y. Zhou, J. R. Morante and A. Cabot, Adv. Energy Mater., 2021, 11, 2100432.
62. Y. Wu, D. Li, J. Pan, Y. Sun, W. Huang, M. Wu, B. Zhang, F. Pan, K. Shi and Q. Liu, J. Mater. Chem. A, 2022, 10, 16309–16318.
63. Y. Luo, Z. Fang, S. Duan, H. Wu, H. Liu, Y. Zhao, K. Wang, Q. Li, S. Fan, Z. Zheng, W. Duan, Y. Zhang and J. Wang, Angew. Chem., Int. Ed., 2023, 62, e202215802.
64. X. Wang, X. Zhang, Y. Zhao, D. Luo, L. Shui, Y. Li, G. Ma, Y. Zhu, Y. Zhang, G. Zhou, A. Yu and Z. Chen, Angew. Chem., Int. Ed., 2023, 135, e202306901.
65. B. Li, P. Wang, B. Xi, N. Song, X. An, W. Chen, J. Feng and S. Xiong, Nano Res., 2022, 15, 8972–8982.
66. S.-Y. Lang, R.-J. Xiao, L. Gu, Y.-G. Guo, R. Wen and L.-J. Wan, J. Am. Chem. Soc., 2018, 140, 8147–8155.
67. X. Jiao, X. Tang, J. Li, C. Li, Q. Liu and Z. Wei, Small, 2023, 19, 2304780.
68. H. Zhang, M. Zhang, R. Liu, T. He, L. Xiang, X. Wu, Z. Piao, Y. Jia, C. Zhang, H. Li, G. Xu, F. Zhou and Y. Mai, Nat. Commun., 2024, 15, 5451.
69. Z. Huang, Y. Zhu, Y. Kong, Z. Wang, K. He, J. Qin, Q. Zhang, C. Su, Y. L. Zhong and H. Chen, Adv. Funct. Mater., 2023, 33, 2303422.
70. H. Lu, Q. Zeng, L. Xu, Y. Xiao, L. Xie, J. Yang, J. Rong, J. Weng, C. Zheng, Q. Zhang and S. Huang, Angew. Chem., Int. Ed., 2024, 136, e202318859.
71. W. Chen, Y. Hu, Y. Liu, S. Wang, A. Hu, T. Lei, Y. Li, P. Li, D. Chen and L. Xia, et al. , Adv. Mater., 2024, 2312880.
72. L.-L. Su, N. Yao, Z. Li, C.-X. Bi, Z.-X. Chen, X. Chen, B.-Q. Li, X.-Q. Zhang and J.-Q. Huang, Angew. Chem., Int. Ed., 2024, 136, e202318785.
73. M. Liao, Y. Xu, M. M. Rahman, S. Tan, D. Wang, K. Wang, N. K. Dandu, Q. Lu, G. Li and L. Le, et al. , Nat. Sustainability, 2024, 1–10.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ee00615e

---