With ever-increasing global energy demands and allied to efforts to reduce the use of fossil fuels and eliminate air pollution, it is essential to provide efficient, cost-effective, and environmentally friendly energy storage devices like lithium-ion batteries, supercapacitors or fuel cells (Sengupta and Kundu, 2020). Lithium-ion batteries (LIBs) are predominant energy storage systems used in portable devices, electrical vehicles, and other applications because of their high energy density and long cycling performance (Li et al., 2022). Many researchers have focused on the development of various electrochemically active materials, including silicon, metal oxides, and transition metal disulfides and their composite materials as anode materials for LIBs (Bates et al., 2000; Ali et al., 2022). In the last decade, transition-metal dichalcogenides (TMDs) have received attention as electrode materials for Li-ion batteries owing to their high electronic conductivity, superior specific capacity, marvelous structural stability, and environment benignity (Shin et al., 2021). The interlayer structure of TMDs comprises three stacked atom layers (X–M–X) held together by van der Waals forces between the transition metal (M) and the chalcogenide anions (X), allowing easy Li⁺ ion insertion/de-insertion (Chen H et al., 2020). In particular, the WS₂ nanomaterials have attracted attention and have been utilized as solid lubricant photosensitive films and super shock absorbers. Additionally, WS₂ has a superior volumetric energy density (3,248 mAh cm⁻³) due to its large density (7.5 g cm⁻³); these advantages make WS₂ favorable for LIBs. However, the electron conductivity of tungsten disulfide is too low, which could lead to rapid capacity fading and poor rate performance when used as the anode material (Feng et al., 2007; Kartick et al., 2014). Therefore, conductive additives based on carbon materials such as graphene, reduced graphene oxide, single-/multi-walled C nanotubes, and pyrolytic C, which enhance the electrochemical performance without side reactions involving Li-ion-batteries, have been used to enrich WS₂-based electrodes. For example, Li et al. (2016) described a nanocomposite of WS₂ and Super P carbon black (WS₂/C) as an anode material for Na-ion batteries (NIBs) and LIBs. The resulting anode exhibited a significantly enhanced electrochemical performance compared to that of a pristine WS₂ anode, which could be attributed to the high conductivity of Super P carbon black in cycling. Ren et al. (2018) successfully fabricated the foam structure of WS₂/single-wall carbon nanotube nanocomposites, which delivered a reversible capacity (∼688 mAh g⁻¹) for 1,000 cycles at 0.1 A g⁻¹. Many researchers have focused on graphene, which is a perfect compound material with advanced electrochemical characteristics and mechanical strength. The similar 2D structures of graphene and layered WS₂ maximize their geometrical compatibility and the possibility of stronger component interactions, leading to favorable outcomes. Zhou et al. (2016a) presented a one-pot method for the synthesis of a WS₂/reduced graphene oxide (rGO) composite, which dramatically improved battery performance. The WS₂/rGO anode showed a stable discharge capacity of 431.2 mAh/g at a current density of 0.1 A/g after 100 cycles, which could be improved.

Herein, we propose a conceptually new and facile method to fabricate a novel three-dimensional composite with unique sandwich-type architecture based on exfoliated WS₂ and ultrafast self-expanded and reduced graphene oxide (userGO). This molecular nanostructure served as a non-active agent-free electrode in LIBs in the form of a film to promote its electrochemical response compared to the composite based on exfoliated WS₂ with conventionally thermally-reduced GO in an oven. The integration of WS₂ and userGO into a composite was accomplished using the layer-by-layer vacuum filtration technique (Cheng et al., 2019) with precise concentrations of the respective components followed by the ultrafast and self-expanded reduction (USER) reaction occurring in the glovebox under an inert gas atmosphere at 350°C (Chen Y. W et al., 2020). The highly exothermic character of the USER reaction immediately starts from the touchpoint and rapidly spreads across the whole film, changing its color from grey to black. The ultra-quick nature of the USER reaction allows it to propagate easily on the WS₂/userGO film from a single point of contact. These internally generated gases were rapidly released from the WS₂/GO film to create highly porous channels inside the film to increase the film thickness and expansion. The modified WS₂/userGO film showed superior electrochemical performance owing to the effective exothermal USER reaction method to generate increased distances between sandwich structures to improve ion transport in LIBs.

Successful preparation of few-layered WS₂ and GO samples as shown in Scheme 1 were confirmed by TEM (Figures 1A, B) and AFM (Figures 1C, D). The TEM images of both film components exhibited flake-like structures typical for 2D materials. Moreover, the analysis indicated that the samples contained no impurities. The number of WS₂ and GO layers was determined through high-profile analysis with AFM (Figures 1E, F). The flakes were typically 14 nm in height for WS₂ and 5 nm for GO, which corresponded to approximately 20 and 7 layers of WS₂ (Xu et al., 2017) and GO (You et al., 2013), respectively.

The morphologies of the cross-sections of the obtained films before and after the USER reaction with different ex-WS₂ ratios were studied by SEM, as shown in Figures 2A–F. Before the USER reaction, the films were characterized by condensed and compact phases of GO (dark gray areas) and ex-WS₂ (light gray areas). The ex-WS₂ sheets were located between GO layers. After the USER reaction, the structure expanded. The release of the gases and internal expansion of the film structures created cavities in the GO structure. Additionally, larger distances were observed between the layers in both ex-WS₂ and GO. Therefore, a more efficient lithiation/delithiation process was expected.

The crystal structure and phase purity of the synthesized molecular hybrids were verified and compared to exfoliated WS₂ by XRD (Figure 3). The diffraction patterns of the three prepared samples were similar to those for ex-WS₂. The diffraction peaks of the films with different components ratio located at 14.7°, 29.2°, 33.2°, 34°, 39.9°, 44.3°, 50°, 58.8°, 60.2°, 61° were assigned to the (002), (004), (100), (101), (103), (006), (005), (106), (108), and (112) planes of WS₂, respectively (ICDD card no. 04-004-4224). Moreover, an intense and sharp (002) peak observed in the XRD pattern indicated that the WS₂/userGO films were stacked together with a highly ordered packing (Vattikuti et al., 2016). Additionally, a clear shift of WS₂ peaks to lower 2θ angles was observed in all WS₂/userGO films compared to ex-WS₂, indicating increased d-spacing in the ex-WS₂ after the USER reaction. The USER reaction removes most of the oxygenated functionalities from the surface of the graphene oxide sheet, reducing the structure to graphene, which was detected in diffraction angles at 12°, 23°, 42° corresponding to (001), (002), and (100) planes (Supplementary Figure S1).

The Raman spectra for ex-WS₂ and WS₂/userGO films are shown in Figure 4. The Raman spectra for the samples were similar in range, from 150 to 800 cm⁻¹. Two Raman scattering peaks at 417 cm⁻¹ and 349 cm⁻¹ were also detected. These two peaks [A_1g(Γ) and E¹ _2g (Γ), respectively] are usually observed in the back-scattering configuration. In a backscattering geometry, these spectra include first-order modes at the Brillouin zone center [A_1g(Γ) and E¹ _2g (Γ)] (Berkdemir et al., 2013). The remaining Raman bands were correlated with the second-order modes at the M point and were a combination or difference of bands coupled with the longitudinal acoustic mode LA(M). The LA(M) mode indicates the in-plane collective movements of the atoms in the lattice. The WS₂/userGO-based films displayed several Raman modes characteristic of WS₂. However, the relative intensities of the first and second-order modes of the films differed from those of the ex-WS₂. The decreased band intensity may be associated with more efficient exfoliation, which might have occurred in the ex-WS₂ layered structure during the USER reaction. The Raman spectra of the films showed two additional peaks characteristic of carbon materials. The D peak arising from the doubly resonant disorder-induced mode (∼1,323 cm⁻¹) due to the stretching of the C–C bond and the G peak, a double phonon mode occurring due to the first-order scattering of the of the sp (Li et al., 2022) C atoms at the Brillouin zone center (∼1,600 cm⁻¹). The films showed a prominent D peak, indicative of significant structural disorder due to the incorporation of WS₂ flakes and the occurrence of USER reaction and is related to the sizes of the in-plane sp (Li et al., 2022) domains and defects (Dresselhaus et al., 2005).

The electrochemical performances of the WS₂/userGO-based films and reference ex-WS₂ were investigated using half-cells. In this configuration, metallic lithium acted as reference and counter electrodes, while the prepared materials acted as working electrodes. The five cycles of cyclic voltammetry (CV) of WS₂/userGO-based films are presented in Figures 5A–C. All these plots showed three redox peaks: 1) at ∼ 2 V and 2) at ∼ 1.8 V, indicating the two-step reduction of W⁴⁺ to metallic W, as shown in Eq. 1, 2, respectively; 3) a peak at ∼2.5 V, confirming the oxidation of W to W⁴⁺ and, thus, the reformation of WS₂ (Eq. 3). These findings verified the reversibility of the redox process in the charge/discharge tests (Shiva et al., 2013). Another signal at ∼0.5 V corresponded to irreversible SEI layer formation. Due to ex-WS₂ pulverization upon charging/discharging, a higher amount of CB was added to an electrode to improve the stability of the reference ex-WS₂ electrode (Figure 5D). However, the CV response of this reference film was less stable than that observed in the composite films, as shown by the better overlapping of the curves in the following cycles. The CV tests proved the excellent stability and electrochemical reversibility of the WS₂/userGO-based films as lithium-ion storage. The CV test results showed that WS₂/userGO1 had the highest potential for electrochemical energy conversion. W S 2 + x L i − + e − → L i x W S 2   x = 3 − 4 , (1) L i x W S 2 + 4 − x   L i + + 4 − x   e − → 2 L i 2 S + W , (2) x L i 2 S + W → x L i + W S 2 . (3)

Figure 6 shows the galvanostatic discharge/charge curves (GCD) of the WS₂/userGO-based films and reference ex-WS₂ electrode executed over 0.05 A/g. The initial charge/discharge specific capacities were 714/762, 624/645, 798/851, and 584/622 mAh/g for WS₂/userGO1, WS₂/userGO2, WS₂/userGO3, and ex-WS₂, respectively. Additionally, the initial specific capacities of all electrodes are shown in Table 1. The highest and lowest capacities were observed for WS₂/userGO3 and the reference ex-WS₂, respectively. These results suggested that the unique structure of userGO plays a greater role in lithiation yield compared to ex-WS₂. This occurred mostly due to the expanded structure of userGO, which allowed a smoother de-/lithiation process. In the case of WS₂/userGO-based electrodes (Figures 6A–C), the following cycles showed insignificant capacity fading due to the excellent stability of the userGO framework. The films exhibited plateaus at 2 and 1.8 V (in the discharge region), matching the two-step reduction of ex-WS₂ into metallic tungsten. An oxidation plateau at ∼2.5 V corresponded to the reformation of ex-WS₂ during charging. These plateaus were most detectable in the pristine ex-WS₂ electrode (Figure 6D) and this analysis is analogous to the CV measurements. The presence of carbon black in the ex-WS₂ electrode did not protect the device against a huge capacity drop. However, userGO not only protected the ex-WS₂ from structure degradation upon charging/discharging but also further increased the capacity than when CB was introduced to the ex-WS₂ reference electrode.

The rate performances of WS₂/userGO-based films and ex-WS₂ are depicted in Figure 7. The tests were performed at 0.05, 0.1, 0.2, 0.4, 0.6, 1, 2, 5, and 10 A/g current densities, with reversible capacity at 0.05 A/g. Although WS₂/userGO1 did not show the highest capacity, it was the most electrochemically durable material; its capacity remained the highest at high current densities among all the samples. Moreover, it showed excellent reversible capacity, keeping 86% of the initial capacity at 0.05 A/g. WS₂/userGO2 showed decreased capacity in low-to-high performance, although the reverse capacity retained 89% of the initial capacity. WS₂/userGO3 shows the cyclic performance of WS₂/userGO3, which showed the highest specific capacity among the tested materials at low current densities but the lowest specific capacity at high current densities. It also maintained the lowest extent in reverse capacity among studied films (85%). Therefore, userGO abundance provided the highest capacity at low current and ex-WS₂ maintained a higher reversible capacity. Therefore, equal loading of the film components (WS₂/userGO1) is the most optimal composition for overall electrochemical performance. Additionally, ex-WS₂ presents the rate performance of the reference ex-WS₂. Due to irreversible structure changes during the charge/discharge process, the electrochemical performance of ex-WS₂ worsened significantly. The electrode with the reference ex-WS₂ was prepared traditionally with other non-active agents. It was physically impossible to prepare a free-standing film using the proposed strategy based only on ex-WS_2. The ratio of ex-WS₂ to CB to PVDF was 4.5:4.5:1, respectively. The specific capacity of ex-WS₂ at 0.05 A/g was almost as high as that of the WS₂/userGO-based electrodes; however, it deteriorated drastically in the following cycles. Therefore, the presence of carbon buffer in the form of CB was not sufficient to fully protect ex-WS₂ from pulverization. These results demonstrated that userGO was an excellent shield for ex-WS₂ boosting both stability and capacity. Moreover, assembling userGO and ex-WS₂ composites as free-standing films reduced the electrode mass and made it more electrochemically and mechanically durable.

Additionally, the WS₂/userGO1 film was subjected to a long-term stability test (1,000 cycles, current density of 0.1 A/g) (Figure 8). The capacity faded for the first 130 cycles and stabilized at 51% of the initial capacity. After 500 stable cycles, the capacity started to fade again, losing 72% of the original capacity. However, the film displayed excellent Coulombic efficiency (101%), indicating that the cell generated more current than was used for charging. This may be related to structure collapse, leading to the fading capacity.

Figure 9 presents the Nyquist plots of the WS₂/userGO-based films and the ex-WS₂ sample obtained using the EIS technique. All electrodes possessed two semicircles and one slope. The first semicircle (at high frequencies range) correlated with SEI layer formation. The second semicircle (at the middle-frequencies range) corresponded to Li⁺ adsorption and charge transfer on the active material surface. It also indicated adequate device resistance. The steep slope, or Warburg impedance, confirmed the smooth Li⁺ diffusion throughout the electrodes. The values of each interphase resistance are shown in Table 2. Here, three resistances can be observed: R_1, or R_b—bulk solution resistance (e.g., electrolyte, separator), R₂, or R_ct—charge transfer resistance, composed of ionic and electronic resistances, and R₃, or R_int—resistance within the cell (Lei et al., 2013). Table 2 and Figure 9 demonstrate that WS₂/userGO-based films possess much higher resistance compared to the reference ex-WS₂. This occurs due to two reasons: i) the USER reaction expanded the GO structure, leaving many empty cavities, thus reducing the conductivity of the graphene sheets; and ii) the ex-WS₂ electrode was composed of 45 wt% of carbon black, which significantly increased the conductivity. However, the empty spaces of userGO are critical for the higher stability and accelerated performance of the composite materials.

Table 3 compares the results presented so far in current state-of-the-art anode materials based on WS₂ structure with respect to this work. Only the present work considers light free-standing structure based on pristine reduced graphene oxide and exfoliated WS₂ without any non-active materials. Its electrochemical performance is among the most promising anode materials in this field.

This study applied a facile method to fabricate WS₂/userGO film using a vacuum-assisted filtration technique. The synthesized three-dimensional composite film with sandwich-type architecture was treated via ultra-quick exothermal expansion and reduction reaction which allowed the rapid release of internally generated gases to create highly porous cavities inside the film. Additionally, this reaction improved ion transport in LIBs to promote the electrochemical performance of both film components. WS₂/userGO films show excellent stability and electrochemical reversibility over a wide voltage range. However, the optimal electrode composed of WS₂:userGO at a 1:1 ratio was the most electrochemically durable, with high capacity at high current densities among all samples. Furthermore, we believe that this strategy provides a universal ultra-fast route to fabricate free-standing electrodes with low mass and non-active free agents to promote both stability and capacity in energy devices.