Nanoarchitectured in situ pre-lithiated carbon anodes for high-power and long-life Li-ion capacitors

Abstract

Lithium-ion capacitors (LICs) combine high energy and power densities but often suffer from poor cycle stability (<10 000 cycles) due to uncontrolled Li⁺ ion losses during solid electrolyte interphase (SEI) layer formation and irreversible side reactions. From an industrial standpoint, achieving >20 000 cycles necessitates an adequate pre-lithiation strategy that efficiently replenishes ions to offset such losses. This work proposes a scalable pre-lithiation approach by adding a thin piece of lithium metal foil (46 μm) in direct contact with the anode while assembling LICs. The electrochemical potential difference between the Li foil and the carbon-coated porous current collector anode facilitates the lithiation process and promotes in situ pre-lithiation (ISP). After a 10 h pre-lithiation time, the resultant LiCx and SEI layer were verified by ex situ characterization studies. The optimized LIC demonstrated a best-in-class specific energy of 204 Wh kg⁻¹ and a specific power of 5.5 kW kg⁻¹. The device achieved a remarkable capacity retention of 87% after 40 000 full charge–discharge cycles, equivalent to 631 h. This structurally engineered strategy underscores the critical role of pre-lithiation in advancing next-generation, high-performance energy storage solutions.

---

1. Introduction

The rapid expansion of electric vehicles (EVs) and portable electronics has significantly increased the demand for advanced energy storage solutions that offer both high energy (≥50 Wh kg⁻¹) and power densities (∼10 kW kg⁻¹).¹ Lithium-ion capacitors (LICs) have emerged as promising candidates to meet these requirements by integrating lithium-ion batteries' high energy density with supercapacitors' high power density. This unique hybrid configuration, featuring battery-type anodes and capacitive cathodes, enables LICs to excel in applications such as regenerative braking in electric buses, industrial machinery, grid stabilization, and high-performance consumer electronics.^2,3 The typical assembly includes graphitic anode and activated carbon (AC) cathode materials. However, huge irreversible losses of active Li⁺ ions during the initial cycles caused by forming solid electrolyte interphase (SEI) layers across the anodes lead to poor initial coulombic efficiency (ICE) and capacity.⁴ Also, most of the high-capacity graphitic anodes attain large surface area and amorphous features that further lead to the intensified consumption of active Li⁺ ions from the electrolyte during the repeated charging-discharging process.⁵ This restrains the capacity and cyclability of LICs, ultimately limiting their applicability in the industry. Providing extra Li sources in the cell is imperative to offset these losses.

Consequently, pre-lithiation has become an appealing approach, compensating for these Li⁺ ion losses and allowing the full utilization of the electrode capacity.⁵ Pre-lithiation can be done in anodes, cathodes, and electrolytes. AC primarily stores charge at the surface, which can lead to self-discharge of ions, making it unsuitable for pre-lithiation. Conversely, if a Li source is added to electrolytes, this can increase the viscosity of the electrolyte and ultimately decrease the kinetics of ions.⁶ In this regard, the pre-lithiation in anodes becomes an effective approach that specifically targets the (1) SEI layer formation and (2) lithiation of negative electrodes to form LiC_x components. In response to this, several pre-lithiation strategies, including chemical routes,⁷ thermal vaporization,^8,9 electrochemical pre-lithiation (ECP),^10,11 melt-deposition,¹² and internal/external shorting by direct contact,^5,13,14 have been undertaken, predominantly focusing on the lithiation of anodes.

Among these, chemical and thermal vaporization techniques rely on a Li complex solution obtained by mixing Li metals and naphthalene with solvents of tetrahydrofuran, butyl methyl ether, and dimethoxyethane.^15,16 These complex solutions exhibit redox couples >0.3 V vs. Li⁺/Li, which implies that only the SEI layer can be formed across anodes, but this method does not achieve lithiation. Additionally, these methods require a large-sized assembly for commercialization, further restricting their use in industry.¹⁷ Furthermore, in the ECP method, the anodes undergo lithiation by a series of charge–discharge cycles in half-cell configurations, after which they are dismantled and reassembled into full cells against a capacitive cathode.¹⁸ Such a reassembling step demands much human time and almost doubles the manufacturing cost, prohibiting their large-scale application. In contrast, the direct contact approach can be employed while assembling the LICs, eliminating the need for a reassembly step.¹⁴ In such an approach, a sacrificial Li source (powder or chip) is meticulously added between the electrode and the separator to provide direct contact (internally or externally) in the presence of an electrolyte.^19,20 The difference in redox potential (μ) or Gibbs free energy (ΔG) between the Li source and the anode material spontaneously induces the flow of Li⁺ ions towards the anode. These movements result in the pre-lithiation of anodes along with the simultaneous SEI layer formation by the reduction of electrolyte organic components before the cycling of the cell. In this context, direct contact, also known as in situ pre-lithiation (ISP), becomes a prominent choice attributed to its easy process and time-saving benefits, making it favorable for large-scale implementation.

However, in the direct contact process, not all Li sources get effectively utilized, and the leftover dead Li sources across the surface of the anode deter proton/charge-transfer reactions.²¹ This increases polarization and Li plating during subsequent cell cycling, eventually causing cell failure. Additionally, the Li metal source may become surrounded by an SEI layer to isolate it from the anode surface, further restricting the lithiation process.²² This necessitates the development of a new strategy toward efficient and continuous lithiation during cell operation by eliminating the concern for unutilized metal. The progressive solutions to these have modified the anodic interphases by targeting electrolytes and Li sources. Replacing liquid electrolytes with highly concentrated and solid-state electrolytes can help reshape SEI layer formation across Li metal. For example, in a recent effort, Manthiram et al. utilized a localized high-concentration electrolyte that helps to artificially generate the mechanically robust SEI layer during the pre-lithiation process, which profoundly doubled the cycle stability of full cell LIBs.²³ In another study, the surface of the graphite anode was coated by a 30 nm thick Li⁺ ion conductive solid state layer (LIPON) as an electrolyte, and the Li source was deposited upon it by thermal evaporation.²² Due to the voids of the LIPON layer, the Li metal source comes in contact with the underlying anode to form an electron-conductive channel, which triggers the lithiation process. This method utilizes Li⁺ ions more efficiently and inhibits the decomposition of electrolytes, which again improves the stability of the battery. Nevertheless, in such approaches, the additive steps followed to modulate electrolytes further complicate the pre-lithiation process. Towards its solution, Zhang et al. pre-lithiated the carbon-based anodes by placing a piece of Li foil behind the porous Cu current collector to supply the Li⁺ ions during the performance of LICs.²⁴ From this assembly, they have impressively obtained 80.1% retention after 5000 cycles. However, little attention has been paid to the pre-lithiation mechanism and relevant characterization studies that can validate the insertion of Li⁺ ions. For the prolonged cycle life, this commercializable approach needs special attention.

In this work, we have strategically positioned the thin Li foil (∼46 μm) leftwards in direct contact with the anode. This prevents the dendritic growth towards the separator that usually shorts the cell. Using facile hydrothermal and vapor deposition methods, we have systematically grown the nanoporous carbon sheets for the anodes over 3D carbon cloth (NCS@CC). This robust adhesion of NCS with CC significantly improves the stability of anodes. Also, the soft and flexible nature of CC helps maintain a homogeneous attachment with Li foil, promoting uniform lithiation and SEI formation. When metallic Li comes into contact with the electrolyte, due to the difference in their electrochemical potential or Gibbs free energy, it oxidizes, releases Li⁺ ions in the electrolyte, and injects electrons into the anode.^20,25 The Li⁺ ions get solvated in electrolytes due to the solvent molecules. Meanwhile, the electrolyte also gets reduced at the surface of the anode, which leads to the deposition of several organic/inorganic components in the form of the SEI layer.²⁵ Furthermore, the solvated Li⁺ ions migrate toward the anode and lithiate it by accepting the electrons (Scheme 1a).²⁶ In this process, a thin SEI layer is formed across the anode, and a fraction of Li⁺ ions gets injected into the NCS sheets to form LiC_x. Nonetheless, there can be two mechanisms for the Li⁺ ion movement across the anode. In the first one, electrons can pass through the edge of the NCS, and the corresponding Li⁺ ions will be intercalated to neutralize the charge (Scheme 1a, path 1-solid line).²⁷ In the second case, when the edge is not in direct contact with the Li foil, electrons pass directly through CC and then move upwards (Scheme 1a, path 2-dotted line).²⁷ Such an approach has significantly pre-lithiated the anodes to prevent an initial capacity drop.

Scheme 1 (a) Schematic illustrating the lithiation mechanism. When Li foil comes into contact with the NCS@CC anode, the Li⁺ ions flow due to the difference in the electrochemical potential of Li foil (μ_Li) and the anode material (μ_C) and leads to the formation of an initial thin SEI layer and LiC_x components. During this process, electrons move from the foil either through the edges of NCS (Path 1-solid line) or from the point where the CC (Path 2-dotted line) directly touches the Li foil. (b) Logarithmic comparison plot for cycle stability and decay per cycle of our LIC device with that of previously reported different types of anodic materials, including carbons, metal oxides, organics, nitrides, sulfides, and others. This shows low decay per cycle and remarkable cycle stability of our NCS@CC sample. Note: the performance is compared by collecting the data under similar conditions, specifically targeting the weight loadings and current rates. More details can be found in Table S1.

To assess controlled pre-lithiation, we compared the lithiated anodes obtained from ISP and ECP methods with the support of various characterization techniques, such as Raman spectroscopy and Fourier transform infrared (FT-IR) spectroscopy. The formation of the SEI layer and LiC_x components was verified through ex situ X-ray photoelectron spectroscopy (XPS) and high-resolution transmission electron microscopy (HRTEM). Moreover, the reliability of our method was demonstrated by a comparative analysis of its electrochemical performance with ECP-derived LICs (ECP-C) and non-pre-lithiated cells (NP-C). Our ISP-derived capacitor (ISP-C, pre-lithiated for 10 h) provides impressive stability, maintaining 87% capacity at 2 A g⁻¹ after 40 000 cycles, surpassing that of the ECP-C cells, which retain only 76.3% capacity under similar conditions up to 10 000 cycles. The designed cell also outclasses the LICs obtained by using metal-based compounds, organic materials, nitrides, carbon-based materials, and several other composites in terms of decay per cycle and cyclability (Scheme 1b). These merits have allowed ISP-C to achieve a remarkable specific energy of 204 Wh kg⁻¹ and 5.5 kW kg⁻¹ specific power. Such improvements can be ascribed to the consistent supply of Li⁺ ions in ISP-C, which compensates for their continuous losses during cycling. These outcomes present a benchmark by introducing a straightforward and industrially applicable direct contact pre-lithiation method that can be generalized and employed for other metal ion batteries and capacitors.

2. Results and discussion

2.1 Physicochemical characterization of electrode materials

In this work, NCS@CC was synthesized by following our previous approach with certain modifications.²⁸ The detailed synthesis procedure is discussed in the Experimental section. Briefly, the methodology includes, first, the formation of MgO templates using the hydrothermal method and then their carbonization using chemical vapor deposition. During the hydrothermal reaction, the activated 3D CC was added along with the precursors of MgO templates. This results in the unique growth of layered Mg(OH)₂ templates over 3D CC(Mg@CC), as observed in field emission scanning electron microscopy (FESEM) images (Fig. S1a–d). The thin layered structure of sheets is attributed to the binding properties of PVP, and the hexagonal shape emerges as per the inherent characteristics of the Mg(OH)₂ crystal structure.²⁹ Subsequently, the carbonization of Mg@CC was performed utilizing the chemical vapor deposition integrated with acetonitrile vapors as the carbon source. The pyrolysis of the sample releases water and CO₂ molecules that lead to the formation of nanosized MgO particles embedded in the carbon matrix (MgO@CC). After carbonization, these inorganic MgO particles were leached out to achieve the porous NCS@CC product. The uniformly engraved porous 2D sheets of NCS can be observed in Fig. 1a–c, S1e and f. The interconnected channels of the porous NCS@CC facilitate fast ion/electron transfer, provide a short migration path for the Li⁺ ions, and promote electrolyte penetration. Furthermore, HRTEM was utilized to investigate the structure of NCS@CC. Fig. 1d and S1g show the formation of uniform pores in NCSs. Interestingly, in HRTEM images (Fig. 1e and S1h), it can be noted that the NCSs are composed of an onion-ring type structure with disordered graphitic planes at the boundary and completely amorphous content in the middle. Such a structure significantly improves the capacity at higher current rates, which is critical for kinetics matching in LICs.³⁰ Moreover, a selected area electron diffraction (SAED) pattern was also obtained in pursuit of more profound insights into such amorphous and graphitic planes. Across the edges of onion rings, two distinct rings corresponding to the (100) and (002) planes of graphitic carbon are obtained (Fig. 1e(I)), whereas no such rings were observed in its middle part (see the SAED pattern in the inset of Fig. S1h). Such a random arrangement of graphitic planes at boundaries creates numerous voids that serve as micropores, whereas the space between different onion-ring arrays forms mesoporous structures. The micropores provide access to the ions to reach the core of each onion ring. Overall, in such a hierarchical structure of NCS@CC, a large number of ions can be stored, which contributes to delivering excess storage capacity. Detailed information on the distribution of pores and associated surface area was obtained using a surface area analyzer. In Fig. S2a, the N₂ adsorption–desorption curve represents the co-existence of type-I and type-IV isotherms (evaluated at 77 K). This confirms that the NCS possesses a microporous and mesoporous structure, providing a high specific surface area of 590.7 m² g⁻¹ (modeled by multi-point Brunauer–Emmett–Teller (BET)). Moreover, the pore size distribution (Fig. S2b) also verifies the presence of micropores (5.82%) and mesopores (94.18%), with 4.58 nm average pore size. The abundant availability of the mesoporous structure can prevent the anode from sluggish kinetics of Li⁺ ions and clogging of pores to achieve stable capacity at higher rates.

Fig. 1 Structural and compositional characteristics of NCS@CC. (a–c) FESEM images at different magnifications; HRTEM images showing (d) the porous architecture of sheets, and (e) amorphous nature of NCS with small range distorted planes of carbon (inset shows the (I) SAED pattern with two rings confirming the (100) and (002) planes, (II) I-FFT pattern and (III) live profile providing an approximate interplanar spacing of 0.36 nm). (f) The PXRD spectra, (g) XPS survey spectrum, and deconvoluted narrow region XPS plots for (h) C 1s, and (i) N 1s for the NCS@CC sample.

The surface properties of MgO@CC and NCS@CC were analyzed by energy-dispersion spectroscopy (EDS). The spectrum plots and corresponding maps provide evidence for the uniform presence of Mg, C, O, and S elements in MgO@CC and C, N, and O elements in NCS@CC (Fig. S3 and S4). The results demonstrate that the MgO templates are completely etched out after acid treatment, resulting in the formation of porous carbon. The homogeneous distribution of N in NCS@CC highlights its doping into the graphitic planes, resulting from adding ammonium hydroxide during the hydrothermal reaction. Such doping is advantageous for increasing the interplanar spacing of graphitic planes. To verify this, the interplanar spacing was evaluated using the live profile of HRTEM, Inverse Fast Fourier transform (I-FFT), and the powder X-ray diffraction (PXRD) pattern. Live profile image (Fig. 1e(II)) and I-FFT (Fig. 1e(III)) reveal a value of 0.37 nm for interplanar spacing, which is higher than that of graphite (0.35 nm). Furthermore, the PXRD pattern (Fig. 1f) shows two broad peaks located at 24.05° and 44.72° corresponding to the (002) and (100) planes, respectively (for the PXRD pattern of MgO@CC, see Fig. S5). As per Bragg's law, the (002) graphitic plane also gives an interplanar spacing of about 0.37 nm, further corroborating the HRTEM findings. These structural properties highlight the benefits of N doping, which contributes to increased interplanar spacing and enhanced porosity, ultimately facilitating the formation of amorphous structures. The coexistence of amorphous and graphitic phases provides efficient channels for the reversible transport of Li⁺ ions.³¹

The compositional characteristics of NCS@CC were confirmed by XPS. Survey spectra (Fig. 1g) validated the presence of C (89.12%), N (6.37%), and O (4.51%) elements in NCS@CC. Moreover, the high-resolution spectra for C 1s and N 1s (Fig. 1h and i) were deconvoluted into different components as C 1s (C=C, C–C, C–N, C–O, C=O, C(O)–O and π–π* at 284.48, 284.98, 285.78, 286.38, 287.28, 288.68, and 290.68 eV) and N 1s (pyridinic N, pyrrolic N, graphitic N and oxidized N at 398.39, 400.98, 402.46, and 404.88 eV), respectively.³² The peak corresponding to C–N again confirms the N doping in the NCS@CC sample. Pyridinic N, located at the edge of graphene rings, has a larger electronegativity than C atoms and induces electron-donating properties in NCS.³³ Such properties increase the active sites for Li⁺ ions and hence improve the lithiophilicity of NCS, thereby promoting uniform pre-lithiation.^34,35 This underscores the potential of N doping in NCS to facilitate uniform pre-lithiation and reversible Li⁺ ion movements during the charge–discharge process. Additionally, Raman spectra (Fig. S6) present the existence of D (out-of-plane vibrations) and G (in-plane vibrations) bands at 1360 and 1587 cm⁻¹, respectively,²⁸ with an I_D/I_G ratio of 1.01. The largely disordered microstructure indicates the co-existence of defects and random graphitic domains in NCS.³⁶ Defects primarily arise from N doping and the porous structure of NCS.³⁷ The broad 2D band across 2820 cm⁻¹ corresponds to the simultaneous presence of sp²–sp³ hybridization along with a highly porous network in NCS.³⁸ Similarly, the structural characteristics of commercial AC were also analyzed using FESEM, PXRD, XPS, and a surface area analyzer (Fig. S7 and S8). The detailed information on each characterization is provided in Note S1.

2.2 Electrochemical characterization

2.2.1 Half-cell test. To explore the anodic behavior of NCS@CC, cyclic voltammetry (CV) was first conducted at different scan rates varying from 0.2 to 5 mV s⁻¹ (Fig. 2a and S9a). Initial 5 CV cycles at 0.2 mV s⁻¹ (Fig. S9b) display high stability attributed to the conductive network of the nanoporous structure. Nevertheless, a slight increase in current after the first cycle is ascribed to formation of a stable SEI layer and activation of the electrode material. A significant contribution of the current at lower voltages (<1 V vs. Li/Li⁺) is the evidence that it is suitable as an anodic material. Notably, the area enclosed by the CV curves increases with the scan rates, implying a direct relation of current with voltage (Fig. 2a). The broad peaks corresponding to lithiation (0.2 and 0.88 V) and delithiation (0.5 and 1.02 V) for 0.2 mV s⁻¹ depict the reversibility of Li⁺ ions. CV curves of other sweep rates follow a similar pattern. However, there is a shift of anodic and cathodic peaks toward positive and negative potentials, respectively, apparently due to the delay in reaction with the electrode bulk material and the increase in surface charge transfer.³⁹ Additionally, the peaks of CV curves are broadened with an increase in scan rates (Fig. S9a), which indicates the pseudocapacitive behavior increases at higher rates. Dunn's method (eqn (1), Note S2) was employed to calculate the diffusive/capacitive contribution at different voltage rates. In the bar graph (Fig. 2b), the capacitive contribution increases with the scan rates because of added pseudo-capacitive behavior due to a porous surface. The shaded region in Fig. 2c showcases the diffusive contribution at 0.2 mV s⁻¹; the large-shaded area validates its battery-type behavior. Moreover, the diffusion contributions for other scan rates (0.4, 0.6, 0.8, 1.0, 2, and 5 mV s⁻¹) are plotted in Fig. S10. Furthermore, the Power law (eqn (2), Note S2) was applied to conclude the diffusive/capacitive behavior of the charge stored at different voltage values. A linear plot for log (current, I) vs. log (scan rate) corresponding to two major peaks across the anodic and cathodic side (highlighted by the dotted line in Fig. 2a) is shown in Fig. 2d. The b values obtained after linear fitting the data for anodic and cathodic peaks are 0.51 and 0.86, respectively. These represent the simultaneous participation of diffusive and capacitive mechanisms during operation.

Fig. 2 Half-cell investigation of the NCS@CC anode. (a) CV curves at different scan rates (0.2 to 1 mV s⁻¹) with two prominent peaks highlighted by dashed lines. (b) Capacitive/diffusive contribution at different scan rates (note: CV for 2 and 5 mV s⁻¹ are plotted in Fig. S9a (SI)). (c) CV showing diffusive contribution (green shaded region) at 0.2 mV s⁻¹. (d) The log (current, I) vs. log (scan rate) plot illustrates the b-parameter corresponding to anodic and cathodic peaks of CV curves. (e) The GCD curves at different current rates, (f) the rate capability test, and (g) the cycle stability test for 200 cycles at 2 A g⁻¹.

Furthermore, gravimetric charge–discharge (GCD) analysis at current rates ranging from 0.05 to 2 A g⁻¹ was conducted (Fig. 2e), and a large capacity contribution below 1 V indicates its anodic-type features. The first five GCD cycles at 0.05 A g⁻¹ are plotted (Fig. S11a), and only a slight fall in specific capacity (SC) from 1003.39 mA h g⁻¹ (1st cycle) to 901.21 mA h g⁻¹ (2nd cycle) is observed, indicating reversibility of Li⁺ ions in NCS. After the second cycle, SC became almost constant, demonstrating the remarkable stability of our anode material. For similar reasons, GCD at higher current rates also shows stable capacities with a coulombic efficiency (CE) of more than 98% (Fig. 2f). To infer the kinetics matching of electrodes, an anodic stability test was performed at a high rate of 2 A g⁻¹, which maintained 82.3% retention (200 cycles), demonstrating its suitability for LICs (Fig. 2g). Electrochemical impedance spectroscopy (EIS) was further performed, and the obtained data were fitted using a Randles circuit (Fig. S11b). The values of each component corresponding to the fitted circuit are tabulated in the Table S2. A minimal charge transfer resistance (R_CT) with a value of 11.71 Ω is observed, attributed to the highly conductive porous network of the anode. The galvanostatic intermittent titration technique (GITT) characterization technique was conducted to understand the diffusion kinetics of Li⁺ ions (displayed in Fig. S12a). Fick's law was adopted to calculate the Li⁺ diffusion coefficient at different lithiation states, and its values lie in the range of 10⁻⁹ to 10⁻¹¹ cm² s⁻¹ (Fig. S12b and eqn (3), Note S3). Notably, there is an increase in diffusion coefficients at higher potentials corresponding to the plateau region. This behavior is due to the barrier faced by Li⁺ ions at lower potentials, as per the formation of LiC_x components.⁴⁰ The electrochemical half-cell investigations for AC against Li are also performed (Fig. S13 and Note S4) to confirm its capacitive behavior and suitability as a cathode.

2.2.2 Ex situ characterization of lithiated NCS@CC. After finding impressive performances of NCS@CC as an anode in the half-cell configuration, its potential needs to be evaluated for full-cell LICs. For the full-cell assembly, anodes are pre-lithiated by the ISP method (presented in Scheme 1a), where a piece of thin Li foil (∼46 μm) is placed behind the NCS@CC anode, along with a few drops of electrolyte in between them (details for complete assembly are in the Experimental section, and the full assembly is shown in Fig. S14). The corresponding findings are comprehensively compared with those of the anode obtained from the conventional ECP method (A_E). First, to optimize the degree of pre-lithiation, several ISP-C cells were soaked for different times (pristine, 4, 6, 8, and 10 h), and the corresponding anodes are named NCS@CC, A4, A6, A8, and A10, respectively. The optimizations aimed to reach a pre-lithiation degree comparable to that of ECP.

After successful pre-lithiation of NCS with each method, the corresponding cells were decrimped, and several ex situ characterization studies, such as Raman spectroscopy, XPS, and HRTEM, were performed to confirm the formation of the SEI layer and LiC_x components. Ex situ Raman spectroscopy was employed to compare the extent of pre-lithiation as shown by the stacked plot in Fig. 3a. The G band located at 1582 cm⁻¹ for NCS@CC is blue-shifted with values of 1591.8, 1594, 1596, and 1601 cm⁻¹ for A4, A6, A8, and A10 samples, respectively. These observations agree with the previous reports on graphite, graphene, and graphitic carbon.^41–43 Moreover, the G band for the A_E anode is located at 1602 cm⁻¹. The approximately same position of the G band for A10 and A_E anodes implies that both are lithiated to the same extent. However, the stage of lithiation (LiC_x) is not identified due to the amorphous nature of NCS, which has no uniform distribution of planes.⁴⁴ Thus, the optimized degree of pre-lithiation emphasizes that the A10 anode from ISP-C can be further used for other ex situ characterization studies and comparable full-cell electrochemical analysis.

Fig. 3 Ex situ structural analysis confirming pre-lithiation and SEI layer formation across the NCS@CC anode. (a) Raman spectrum comparison of NCS@CC anodes with different anodes obtained by ISP (A4, A6, A8, and A10) and ECP (A_E) methods. The XPS narrow region spectra of C 1s, Li 1s, O 1s, Li 1s, and P 2p for (b) A10 and (c) A_E anode samples at different sputtering times. The XPS plots confirm the presence of several components in the SEI layer and lithiated C in both samples. (d) Schematics show that Ar-ions hitting the surface gradually etch the SEI layer with time (0–360 s), and simultaneously, X-rays detect the chemical composition at different intervals.

Furthermore, XPS was performed to corroborate the formation of the SEI layer and LiC_x across the surface of A10 and A_E anodes. As discussed above, the SEI layer (∼15 nm) is formed due to the reduction of electrolyte (which contains EC, DMC solvent, and LiPF₆ salt), and it is composed of several organic and inorganic compounds. An ion beam (Ar ions) sputtering was performed for different intervals (0, 120, 240, and 360 s) to etch the surface. As per the instrumental parameters, the sputtering for 1 min. duration can etch the 2–3 nm layer from the surface. Therefore, a sputtering of 360 s leads to the etching of a 12–15 nm thick layer down the surface. Consequently, in the XPS survey (Fig. S15) of A10 and A_E anodes, Li 1s, F 1s, P 2p, C 1s, and O 1s elements are present, which are attributed to the components of the SEI layer. It is well established that the SEI layer comprises several organic (alkyl carbonates) and inorganic species such as LiF, Li₂O, and Li₂CO₃, which can be further validated by the narrow region spectra for each element (Fig. 3b and c). In the C 1s spectra, both electrodes distinctly exhibit commonly expected species, including C–C (284.58 eV), C–O (286.58 eV), and C(O)=O (289.8 eV).⁴⁵ In contrast, when the sputtering time increases to 360 s, the peak corresponding to LiC_x appears at 282.88 eV.^46,47 This demonstrates that the SEI layer is completely removed after 360 s sputtering time, and the surface of NCS is exposed to give an additional signal corresponding to LiC_x in both samples. For a clear understanding and comparison, we have plotted only four peaks showcasing C–C, C–O, C(O)=O, and LiC_x components. In addition, the main components of the SEI layer, such as LiF and Li₂O, can be confirmed from F 1s and O 1s spectra. Significantly, the peak representing Li–O (55.2 eV in Li 1s and 528.8 eV in O 1s) increases with etching, providing evidence that when the outer SEI layer is etched out, Li₂O is observed in the inner SEI layer.⁴⁸ The intense peak for LiF (55.76 eV in Li 1s and 685.05 eV in F 1s) covering a broad region shows that it is abundantly present across the anode surface. Also, the species such as POF₃, Li_xPOF_y, and OPF(OCH₃)₂ obtained due to the reduction of PF₆⁻ can be verified by using the peaks of P–F (687.18 eV in F 1s and 137.28 eV in P 2p), and P–O/P=O in the spectra of P 2p, O 1s and F 1s of both the samples (Fig. 3b and c).^49,50 The atomic percentages of each element at different etching times are compared by using a bar graph (Fig. S16). The atomic percentages of C 1s, O 1s, and P 2p decrease with sputtering, which is expected to be because of the removal of species of electrolyte. Conversely, it significantly increases for F 1s and Li 1s due to intensified SEI layer components. The schematic in Fig. 3d illustrates the complete process of depth profiling using ex situ XPS. Layer-by-layer removal of the surface by Ar ions at different intervals leads to the complete removal of the SEI layer (blue color) and finally allows access to LiC_x in NCS. Several Li alkyl carbonates, such as ROCOOLi and ROCOLi (R: alkyl group), are present in the SEI layer, which can be observed by performing ex situ FTIR. Consequently, Li ethyl dicarbonate (LEDC) or Li methyl dicarbonate (LMC) are the possible components at the surface of lithiated samples. First, to differentiate these features of the SEI layer from those of solvents and electrolytes, the FTIR spectra for NCS@CC dipped in ethylene carbonate (EC), dimethyl carbonate (DMC), and electrolyte (1 M LiPF₆ in EC/DMC) are also obtained. In the comparison plot (Fig. S17), no peak was identified for the pristine sample (NCS@CC), and when it was rinsed with DMC for 1 min. (red box). However, when NCS@CC was rinsed with EC, several peaks representing the carbonates were observed. A similar pattern was observed when it was dipped in the electrolyte with an additional peak at 838 cm⁻¹ as per the P(O)–F bond present in the salt.⁵¹ Notably, no feature for Li alkyl carbonates was identified for these samples. On the other hand, in the spectra for ISP samples (green box), the broad peaks corresponding to different components were observed and intensified as the pre-lithiation time increased from 4 to 10 h. This shows that the elevated soaking time increases the number of reduced electrolyte components across the surface; hence, the SEI layer's growth is also elevated. The peaks in the region 1500–1350 cm⁻¹ are characteristic of vibrational modes related to C–H bending in EC and symmetric stretches of O–CO₂ groups in LEDC (a component of the SEI layer).^49,52 Additionally, the characteristic peaks centered at 1630 cm⁻¹ and in the region from 1200 to 1000 cm⁻¹ are the signals of C=O and C–O, respectively, corresponding to the stretching present in LEDC or LMC components.^49,52 Also, the broad region from 900 to 780 cm⁻¹ is apparently due to the presence of stretching modes of the P–F bonds in the PF₆⁻ anion or by P–F or P–O–F obtained from POF₃, Li_xPOF_y, and OPF(OCH₃)₂ (reduction products of PF⁶⁻).^52,53 Similar peaks were observed for the A_E anode sample (blue box, Fig. S17). Therefore, collectively, XPS and FTIR have successfully identified the features of SEI layer components.

HRTEM was employed to further determine the structural composition of the SEI layer in A6 and A10 samples. The HRTEM images (Fig. 4a and b) show the change in thickness of the SEI layer with the soaking time. Noteably, the layer had a thickness of ∼5 nm when the cell de-crimped after a soaking time of 6 h, whereas it was ∼15 nm when the time increased to 10 h. Several crystalline planes corresponding to the different components of the SEI layer are visible in the HRTEM image (Fig. 4c) of the A10 anode sample. Live FFT and live height profiling were performed to examine these crystalline components carefully. The live FFT pattern (Fig. 4d) shows the rings and crystalline points corresponding to different planes of LiF (002) and Li₂O ((111) and (113)). These planes are also analyzed by measuring their interplanar spacing using the live height profile patterns (Fig. 4e). The interplanar spacing associated with the (002) plane of LiF (golden color), (111) plane of Li₂O (red color), and (113) plane of Li₂O (blue color) are 0.22 nm, 0.14 nm, and 0.28 nm, respectively.^50,54 The crystallographic planes are further drawn by operating the I-FFT and are plotted beside the respective live height profile pattern. The high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) image, in parallel with electron energy loss spectroscopy (EELS), helps monitor the structural properties of the SEI layer. The EELS observed at the edge of NCS (marked by the pink box in Fig. 4f) shows the Li–K edge spectra corresponding to Li₂O (red color) and LiF (golden color) in Fig. 4g, which again validates the formation of the SEI layer.^55,56 The ex situ investigations reveal that the inorganic components, such as LiF and Li₂O, are abundantly present throughout the SEI layer, which is consistent with the previous reports.^55,57 These components are highly beneficial for the fast, reversible ionic movements across the surface of the anode.

Fig. 4 Ex situ HRTEM analysis of the lithiated anode. (a) HRTEM image for A6 showing the SEI layer grown across the edge of NCS with a thickness of ∼5 nm, (b) HRTEM image for the A10 sample presents that the thickness of the SEI layer increased to ∼15 nm with time. (c) HRTEM image exhibits crystalline planes in the SEI layer across the anode and corresponding (d) live FFT pattern displays three planes (111) and (113) for Li₂O, and (002) for LiF, (e) live profile data at different positions from the HRTEM image show the interplanar spacing of these planes and corresponding I-FFT patterns are plotted beside each profile, (f) HAADF-STEM image of lithiated NCS from the A10 sample and (g) corresponding Li–K edge EELS spectra showing the presence of Li₂O and LiF.

2.2.3 Full cell test. After successfully optimizing pre-lithiation, the electrochemical performance of ISP-C and ECP-C was evaluated. The assembly of ISP-C is illustrated in Fig. 5a. First, CV and GCD studies were conducted for ISP-C with different mass loadings of NCS (anode) and AC (cathode) with 1 : 1, 1 : 2, 1 : 3, and 1 : 4 ratios. Analysis of the GCD curves for each assembly at 50 mA g⁻¹ indicates that the ISP-C with a 1 : 2 ratio exhibits the maximum charge–discharge duration along with a negligible IR drop (Fig. 5b). In contrast, the ISP-C cells with 1 : 1, 1 : 3, and 1 : 4 ratios depict IR drops of 0.16, 0.04, and 0.10 V, respectively. An increased potential drop results in a reduced effective voltage window, leading to diminished capacitive behavior of NCS and, consequently, a decrease in both specific energy and power.⁵⁸ GCD measurements for LICs with 1 : 1, 1 : 3, and 1 : 4 ratios were also performed at current rates of 50, 100, 200, and 500 mA g⁻¹ (Fig. S18b–d). Notably, the IR drop for each cell increases with currents that negatively impact the power performance of cells. Furthermore, their CV plots were also compared at 5 mV s⁻¹, where only the 1 : 2 configuration maintained the rectangular shape. This indicates the capacitive behavior, which aligns with its lowest IR drop (Fig. 5c and S19a). In addition, the CV curves for ISP-C at various sweep rates are displayed in Fig. S19b. Overall, the absence of a significant IR drop and well-maintained rectangular-shaped CV exhibits the optimal mass-capacity balancing in the 1 : 2 configuration. Consequently, a 1 : 2 ratio of NCS and AC was selected for further comparative analysis of ISP-C and ECP-C cells. The effectiveness of pre-lithiation was evaluated by simultaneously comparing the performance of ISP-C and ECP-C with that of NP-C. The initial 5 cycles of CV curves at 5 mV s⁻¹ for ISP-C, ECP-C, and NP-C are shown in Fig. S20a–c. It is noteworthy that the CV curves of ECP-C exhibit a slightly distorted rectangular shape, which is likely to be because of some anodic material loss while reassembling the cell. Moreover, the current for NP-C dropped significantly with each cycle, which is attributed to the consumption of electrolyte-active Li⁺ ions in the formation of the SEI layer. However, no such current drop was observed in ISP-C and ECP-C samples, showcasing the advantage of pre-lithiation in stabilizing the LICs. Furthermore, the ISP-C and ECP-C GCD curves show nearly the same charging-discharging time for the initial two cycles, affirming that ISP is an applicable approach for pre-lithiation (Fig. S20d and e). Importantly, in the second cycle, the ECP-C exhibits a more significant drop compared to ISP-C; this is expected to be because of the limited Li⁺ ions used in the pre-lithiation of anodes in ECP-C. The observation suggests that there might be a requirement for additional Li⁺ ions to sustain performance in subsequent cycles, which can be supplied by Li foil behind NCS@CC, reinforcing the advantages of the ISP approach for LICs. On the other hand, NP-C exhibited a sharp decline in discharging time even in the first cycle, emphasizing the necessity of pre-lithiation. By the second cycle, it experiences a substantial reduction in charge–discharge time due to the depletion of active ions in the electrolyte, as illustrated in Fig. S20f. Meanwhile, GCD cycles at various current rates of 0.05–3 A g⁻¹ were also recorded for ISP-C (Fig. 5d) and ECP-C (Fig. S21) cells.

Fig. 5 Electrochemical performance of ISP-C (green color) and ECP-C (blue color). (a) Schematic diagram of the LIC cell assembly for the ISP-C cell. (b) GCD curves of ISP-C cells with different anode/cathode weight ratios of 1 : 1, 1 : 2, 1 : 3, and 1 : 4 at 0.05 A g⁻¹ (inset shows the corresponding IR drop). (c) CV curve for ISP-C at 5 mV s⁻¹. (d) GCD curves at different current rates varying from 0.05 A g⁻¹ to 3 A g⁻¹. (e) Voltage leakage test for 60 h. (f) SCs (five cycles) of ISP-C and ECP-C at different current rates with the corresponding coulombic efficiencies. (g) EIS curves for freshly prepared and cycled cells (10 cycles) belonging to ISP-C and ECP-C, respectively.

At equilibrium, the open circuit voltage (OCV) of full cells, determined by the potential difference between the cathode and anode, is representative of how much electric quantity is held by a cell.⁵⁹ Additionally, the change in OCV for a freshly packed cell can provide insights into the dynamics of chemical reactions. In this regard, the OCV for freshly prepared ISP-C and ECP-C cells was compared (Fig. S22a). Immediately after packing the cell, the OCV of the ISP-C cell (red box) is 2.78 V, whereas, for ECP-C, it is 2.0 V, as supported by optical images of the multimeter in Fig. S22b. The higher OCV for ISP-C is ascribed to the increased potential difference between the cathode and anode, resulting from the placement of Li foil adjacent to the anode. After 10 h (yellow box), the OCV values increased to 3.04 V for ISP-C and 2.6 V for ECP-C (Fig. S22c). The voltage increase is expected to stem from the migration of Li⁺ ions from the electrolyte toward the anode and the reduction of electrolyte components at its surface.⁶⁰ Overall, the higher OCV for ISP-C is representative of the better specific energy in the cells. Interestingly, the OCV of ECP-C becomes stable earlier (<30 min) than that of ISP-C, which is assumed to be due to the pre-grown passivation layer in the ECP-C. The reduction process of electrolytes gets minimized for ECP-C, and the corresponding cell reaches the equilibrium condition quickly (<30 min). In contrast, the reduction process simultaneously occurs in ISP-C that gradually changes the OCV for a longer time (∼1 h).⁶¹ Therefore, the OCV indicates that chemical reactions occur exquisitely during the ISP process.

Furthermore, self-discharge is a critical factor in assessing the long-term durability of cells in industrial applications, as it determines voltage decay or energy loss in an open-circuit state.⁶² It is well known that supercapacitors exhibit a drawback of high self-discharge voltage due to their surface mechanisms. The high free energy in their charged state drives a force towards a discharged state with lower free energy by reducing the charge immediately.⁶³ On the other hand, the redox reactions in the anodes of batteries or LICs lead to the diffusion of ions within the atomic planes, thereby preventing their self-discharge. We performed a self-discharge test for the ISP-C cell by charging it at 1 A g⁻¹ till 4.2 V and recording its OCV for 60 h. It was observed that the cell maintained a 3.36 V voltage after 60 h (Fig. 5e), thereby representing only a 0.014 V h⁻¹ voltage drop, which is ten times better than that for supercapacitors and exhibits remarkable retention compared to several previously reported LICs (Table S3). This highlights the potential of our cells to retain their charge for extended periods.

The addition of Li foil across anodes can help to maintain a higher voltage window even for faster diffusion kinetics, exhibiting its tendency to perform with greater capacity at elevated rates. To corroborate the benefits of the ISP method, the rate capability test was conducted for various currents in the range of 0.05 to 3 A g⁻¹ (Fig. 5f). ISP-C and ECP-C presented almost the same capacity values till 0.4 A g⁻¹ current rate; however, beyond 0.5 A g⁻¹, ISP-C started dominating over ECP-C. At a 2 A g⁻¹ current rate, the SC values for ISP-C and ECP-C are 29.89 and 18.5 mA h g⁻¹, respectively. This high-rate performance of ISP-C is undoubtedly linked to its unique assembly. There is a continuous loss of a significant amount of Li⁺ ions with each cycle in ECP-C, which can be replenished in ISP-C due to the Li foil. The better capacity performance of ISP-C at higher current rates eventually helps increase the specific energies. Moreover, EIS spectra were recorded for each assembly in their fresh states and after 10 cycles (Fig. 5g). Fig. S23 displays the fitted Randles circuit used for each spectrum and Table S4 lists the corresponding values of different parameters. Uniform SEI layer formation reduces the electrochemical impedance and hence the charge transfer resistance (R_CT) for Li⁺ ions during charging and discharging. For fresh cells, ISP-C and ECP-C show an R_CT of 28.8 and 55.9 Ω, whereas, after the first cycle, these increased to 37.71 and 67.01 Ω, respectively. The higher R_CT for ECP-C demonstrates the resistance corresponding to the already lithiated A_E sample and the grown SEI layer. Also, the consumption of active Li⁺ ions is greater in ECP-C compared to ISP-C, which leads to a fall in the number of active ions in the electrolytes and decreases the diffusion kinetics of ions. Nonetheless, it is noteworthy that ohmic resistance (R_O) for ISP-C (26.96 Ω) is a little higher than that of ECP-C (11.97 Ω), which is expected to be due to the contact resistance offered between Li foil and the anode electrode. The electrochemical tests were also performed on NP-C cells (Fig. S24 and Note S5).

The stable cycle life of a cell is the key to estimating the potential for industrial applications substantially. Therefore, to further investigate the influence of the lithiation method on the cyclability, both cells were evaluated at different rates (1 A g⁻¹ and 2 A g⁻¹) for a longer period. In Fig. 6a, ISP-C shows a remarkable performance at 2 A g⁻¹ by retaining 87% of SC after 40 000 cycles, whereas that of ECP-C declined to 76.3% within just 10 000 cycles. The exceptional stability of ISP-C challenges the performance of several previously reported powered LICs containing carbons, organic, nitrides, chalcogenides, and many other anode materials, which can be visualized by using the logarithmic plot displayed in Scheme 1b. It includes examples of LICs obtained using different pre-lithiation methods, such as soft carbon by chemical,⁶⁴ Au@CCS by electrodeposition,⁶⁵ CdNb₂O₆ by direct contact,⁶⁶ and NDPC by ECP.⁶⁷ These samples yielded a stability of 88.6, 89.3, 75.8, and 68% after 10 000, 2500, 10 000, and 15 000 cycles, respectively, with >0.001 capacity decay per cycle (for more details, refer to Table S1). In contrast, our cell exhibited a very low decay per cycle of 0.0003, nearly 10 times less than that in these reports. The highly sustained performance is again attributed to the continuous Li⁺ ion supply by the Li foil during cycling. After the pre-lithiation step, the breakage and reformation of the SEI layer also occur during the cycling process, which again requires Li⁺ ions for compensation.⁵⁷ In all other methods, the initially performed lithiation provides limited ions in the anode. It does not supply enough ions to compensate for consecutive losses, and the electrolyte hinders the recovery of ions. This continuously drops the cell's capacity and leads to a high drop per cycle. A uniform SEI layer after 40 000 cycles can be seen in the HRTEM image (Fig. S25a). Notably, in traditional direct contact pre-lithiation, the Li source is placed between the anode and separator, which can also continuously provide Li⁺ ions for compensation.¹⁶ However, the leftover Li metal over the surface of the anode causes uncontrolled dendrite growth and compromises the cell's lifetime.⁶⁸ The strategic positioning of foil on the left side of the anode mitigates such a problem to enhance the service life of our cell. The ex situ FESEM images (Fig. S25b and c) obtained after 40 000 cycles do not show any dendrite branches, confirming that there is no deposition of unnecessary Li metal over NCS@CC. Also, the image of the dismantled cell shows leftover Li foil stuck to the case of the cell, and no deposition of metal is visible over the NCS@CC electrode (Fig. S26). Moreover, at 1 A g⁻¹, the ISP-C retained 85.6% of SC, much higher than that for ECP-C (40.8%) after 10 000 cycles due to similar reasons. The GCD curves of these 10 000 cycles at 1 A g⁻¹ were also compared, signifying that ISP-C worked for nearly 631 h, whereas ECP-C completed these cycles in just 421 h (Fig. 6b) due to the reasons mentioned above. Also, the initial and last four GCD cycles of each cell (inset of Fig. 6b) show that the charge–discharge time of ECP-C, as well as ISP-C, has decreased in the last cycles, but for ECP-C, the drop is higher and has fallen to almost half that of ISP-C. This again shows that the Li foil helped maintain large voltage windows by retaining lower redox potentials of anodes, which eventually results in lower IR drops and elevated SCs. Overall findings conclude that pre-lithiation is a prerequisite for the optimal performance of LICs, and ISP is more prominent than the ECP approach.

Fig. 6 Electrochemical performance comparison of ISP-C and ECP-C. (a) SC retention test for both cells at 1 and 2 A g⁻¹ current rates and the corresponding coulombic efficiencies (right y-axis). This shows that ISP-C outclasses ECP-C at both currents in terms of capacity retention and stable coulombic efficiencies. (b) Overlapped GCD cycles for 10 000 cycles at 1 A g⁻¹, and magnified plots in the inset show the initial and last four cycles for both cells. (c) Ragone plot comparing the specific energy and specific power for both cells with the previously reported aqueous and non-aqueous LICs (data and references tabulated in Table. S5 and S6, SI). The examples include several materials belonging to families of carbons, metal oxides, sulfides, nitrides, etc. Note: the comparison is made on the basis of the sum of active materials of both the electrodes.

As shown in Fig. 6c, the Ragone plot compares the specific energy and power of ISP-C and ECP-C LICs with previous reports based on aqueous and non-aqueous LICs (Tables S5 and S6). Consequently, our device exhibits specific energy/power values of 204 Wh kg⁻¹/152 W kg⁻¹, which are much higher than those in the recent reports (CoSe₂–SnSe‖AC: 131.03 Wh kg⁻¹/100 W kg⁻¹,⁶⁹ CdNb₂O₆‖AC: 150.3 Wh kg⁻¹/200.0 W kg⁻¹,⁶⁶ Sb@BP/C:1 74.3 Wh kg⁻¹/50.4 W kg⁻¹,⁷⁰ CoSe₂/N-rGO‖AC: 104.2 Wh kg⁻¹/270.6 W kg⁻¹,⁷¹ NDPC‖NDPC: 111.3 Wh kg⁻¹/78.1 W kg⁻¹,⁶⁷ and many more). The commendable performance is due to the inherent characteristic properties of the NCS@CC anode. Additionally, in Table S5, it can be inferred that some materials, such as B–BCN, illustrate a comparable specific energy and power of 200.3 Wh kg⁻¹ and 239.9 W kg⁻¹, respectively.⁷² But the obvious consumption of ions again leads to their early capacity fade, retaining only 80% after 7000 cycles. Also, at higher current rates, ISP-C shows more specific energy than ECP-C cells by attaining an 80.5 Wh kg⁻¹ value at a maximum power of 5.5 kW kg⁻¹ (calculated by employing eqn (S3)–(S5) in Note S2). This shows the remarkable approach of using ISP for LICs. The energy density of the ISP-C cell, considering the mass of current collectors, electrolyte, separators, Li foil, and electrode materials, is calculated to be ∼16.5 Wh kg⁻¹, representing the practical viability of our device (detailed calculations are in Note S6). In addition, a mini-calculator was successfully used to demonstrate the practical application of our ISP-C cell, as shown in the inset of Fig. 6c. The overall electrochemical performance of ISP-C represents its viability for several industrial applications requiring voltages >3 V.

3. Conclusions

This study comprehensively optimizes the direct contact pre-lithiation approach that substantially advances the state of the art in LIC technology. The experimental outcomes demonstrate the advantages of the ISP method compared to conventional ECP in enhancing LIC longevity. Ex situ analysis confirms that pre-lithiation facilitates the formation of the SEI layer and LiC_x compounds within the anode material. The continuous infusion of Li⁺ ions into the 3D NCS@CC anode, driven by electrochemical potential differences, achieves an exceptional cycle stability of 87% after 40 000 cycles at 2 A g⁻¹, along with a specific energy of 204 Wh kg⁻¹ and specific power of 5.5 W kg⁻¹. Unlike conventional pre-lithiation methods that necessitate cell reassembly, our ISP approach offers a scalable solution for large-scale LIC applications. These findings deepen the understanding of pre-lithiation mechanisms and pave the way for advancements in ultra-stable hybrid energy storage systems with high energy and power. This versatile and efficient approach invites pioneers to explore similar strategies for other anodic materials in different applications. Furthermore, it opens avenues to tune their performance, targeting the effect of different parameters such as foil thickness, type of electrolytes, additives, current collectors, and many more.

4. Experimental section

4.1 Chemicals

Acetonitrile (>99.9%), ammonium hydroxide (NH₄OH), polyvinylidene fluoride (PVDF), ethylene carbonate (EC; >99%), lithium hexafluorophosphate (LiPF₆), n-methyl-2-pyrroledone (NMP), and hydrochloric acid (HCl, 18%) were purchased from Sigma Aldrich. Polyvinyl pyrrolidone (PVP, M.W. = 40 000) and dimethyl carbonate (DMC; >98.0%) were purchased from TCI Chemicals Pvt. Ltd. Furthermore, magnesium sulfate heptahydrate (MgSO₄·7H₂O) and ethylene diamine tetra-acetic acid (EDTA) were purchased from Sd Fine Chem. Ltd. Activated carbon (ACS20) was obtained from China Steel Chemical Corporation. All chemicals were used without any further physical or chemical modification. Methanol and deionized (DI) water were used for washing samples at various steps.

4.2 Material preparation

4.2.1 Synthesis of Mg@CC. MgO templates were grown over a 3D CC matrix using a hydrothermal route. Initially, 1 g MgSO₄ was dissolved in 30 ml of DI, followed by the addition of 1.2 g PVP and 0.396 g EDTA. After that, 0.5 g of NH₄OH was added dropwise to the solution and further stirred for 1 h to achieve a white colloidal solution. The mixture was then transferred to a 50 ml Teflon-lined autoclave. Alongside, a piece (4 × 2 cm) of CC (pre-activated with a mixture of H₂SO₄ : HNO₃ = 1 : 1 solution) was placed inside it, which provides abundant nucleation sites for the growth of MgO sheets. The autoclave was further heated at 120 °C for 8 h with a ramp rate of 5 °C min⁻¹. After the natural cooling of the autoclaves, the resulting Mg@CC product was washed 3–4 times using DI and methanol until a clear solution with pH ∼ 7 was achieved.

4.2.2 Synthesis of NCS@CC. The obtained Mg@CC sample was carbonized in a tube furnace under an inert atmosphere (N₂). It was initially ramped at 2 °C min⁻¹ to reach 350 °C and maintained for 2 h. The temperature was further increased to 700 °C for 3 h. Meanwhile, acetonitrile (5 ml) as the carbon precursor was also added in the path of the N₂ gas flow. The vapors of acetonitrile reach the MgO templates and carbonize them. After carbonization, the sample was dipped in 4 M HCl for 6 h to etch out the MgO template. Finally, the sample was again washed using DI and methanol several times to obtain NCS@CC.

4.2.3 Material characterization. The detailed structural analysis of NCS@CC and AC was carried out using the PXRD method (Malvern Panalytical Empyrean; Cu Kα, λ = 1.5406 Å) and XPS (NEXSA, Thermofisher Scientific, Al Kα). Thermo Scientific Avantage software was used for deconvolution and peak fitting of all elemental narrow region spectra obtained by XPS. To analyze the chemical properties of lithiated samples, depth profiling was performed by sputtering over samples with an Ar ion source (1 kV). To obtain the surface area and pore-size distribution, BET analysis (N₂ adsorption–desorption isotherm and pore size distribution) of both materials (NCS and AC) was evaluated using an Autosorb-iQ (Quantachrome) at 77 K. To optimize the lithiation state, ex situ Raman spectroscopy was performed by using a Thermo Fisher Scientific Dxr3 Raman microscope. FT-IR spectroscopy, coupled with attenuated total reflection (ATR), was employed for ex situ analysis to verify the formation of SEI layers. The morphological study of all samples and corresponding elemental mapping was performed using an FESEM (JEOL, JSM 7900F) equipped with an EDS (OXFORD). A TEM (JEOL NEO ARM 200F) was employed to obtain the high-resolution images of NCS and for the ex situ investigation of lithiated anodes. Moreover, EELS was performed to analyze the SEI layer components of lithiated anode samples. Digital Micrograph 3.6 (DM, Gatan) software was used to analyze the SAED pattern, live profile, and HRTEM images of all samples.

4.3 Electrochemical measurements

4.3.1 Electrode preparation. For anodes, circular discs of diameter 1.5 cm were punched out from the parent NCS@CC (4 × 2 cm) sample. The circular discs were directly used as anodes for half-cell and full-cell investigations. To prepare the cathode electrodes, the slurry was prepared by finely mixing 90 wt% % AC with 10 wt% % PVDF, which was further dispersed uniformly in NMP. This slurry was coated over aluminium foil using a tape casting coater and then vacuum dried at 100 °C for 12 h. The coated foil was further passed through a hot roller. These cathodes were also punched into 1.5 cm diameter discs and used for further electrochemical testing.

4.3.2 Half-cell test. Electrochemical tests of all the cells were performed using the standard CR 2023 cell model. All cells were packed in an Ar-filled glove box (H₂O < 0.1, O₂ < 0.1). NCS@CC (∼2–3 mg) and AC (2–12 mg) electrodes were first packed in a half-cell configuration. Electrodes were placed against a Li metal chip (diameter: 1.5 cm) and were separated by a glass fiber separator with 90% porosity. The cell was finally assembled using electrolyte (30–35 μl) containing 1 M LiPF6 salt in EC : DMC (1 : 1 by volume) solvent. The cell containing NCS@CC‖Li was analyzed using CV (Biologic SAS, VSP electrochemical workstation) and GCD (Neware Battery Test System) tests in the voltage range of 0.01 to 3 V, whereas AC‖Li was analyzed in the 2 and 4.2 V window. EIS measurements for both cells were also performed on a Biologic SAS, VSP electrochemical workstation, and EC Lab11.60 software was used for modeling the corresponding Randles circuit. For each EIS measurement, the amplitude was 5 mV, and the frequency range was 10 mHz to 100 kHz.

4.3.3 Prelithiation by the ISP method. For the ISP, an optimized thin lithium foil (0.046 mm) was first obtained by pressing a Li metal chip (0.45 mm) with a roller. The obtained Li foil was placed adjacent to the NPC@CC anode while assembling the cell. Therefore, the order of overall full-cell assembly is Li foil, electrolyte (1 M LiPF₆ in EC : DMC = 1 : 1), the soaked NCS@CC anode, the separator, and the AC cathode. The cell is filled with 30–35 μl of electrolyte. For optimization of the pre-lithiation degree, the cells were given a rest for different times: 4, 6, 8, and 10 h. After soaking, to identify the degree of pre-lithiation, the cells were decrimped to perform several ex situ characterization techniques.

4.3.4 Prelithiation by the ECP method. For ECP, the NCP@CC anode was packed against the Li chip in the half-cell configuration. The charge–discharge was investigated five times at a current rate of 0.1 A g⁻¹, and in the fifth cycle, the discharging was terminated at 0.1 V (lithiated state). At this lithiation state, the cell was immediately decrimped inside the glove box, and the lithiated anode (AE) was extracted. The AE anode was further utilized for full assembly against the AC cathode. In the case of NL-C, the anodic and cathodic electrodes were directly packed into a full cell assembly without any pre-lithiation step.

4.3.5 Ex situ measurements. The respective cells corresponding to the ISP and ECP methods were decrimped in the glove box. The obtained anodes were then rinsed with ∼2 ml DMC solvent and then dried in the glove box for two days. The dried samples were transferred to a tightly sealed Ar-filled container and transferred out from the glove box for ex situ FTIR, Raman spectroscopy, and XPS. The samples were immediately utilized for analysis. For ex situ HRTEM, the NCS sample was scratched from the CC and then dispersed in the DMC solvent. After this, the dispersed solution was drop-cast over a lacey carbon TEM grid. The grid was again dried overnight (∼12 h) in the glove box. The grid was also transferred to an Ar-filled, tightly packed vial and then taken out to load immediately for the measurements.

4.3.6 Full cell test. Following the same assembly of ISP-C, the cells with different mass ratios of NCS and AC were packed. The mass ratios of NCS : AC are 1 : 1, 1 : 2, 1 : 3, and 1 : 4. CV and EIS tests for each cell were conducted in the 2 to 4.2 V voltage range again using the Biologic SAS, VSP electrochemical workstation. Moreover, GCD (2 to 4.2 V), self-leakage, and OCV tests were carried out on a Neware Battery Test System (CT-4008Q-5V100mA-124). For the self-leakage test, the ISP-C cell was first charged with 0.1 A g⁻¹ current and then rested for 60 h to observe the change in OCV. Cycle stability tests for both cells (ISP-C and ECP-C) were conducted at 1 and 2 A g⁻¹ current rates. All the calculations were done by considering the sum of the active materials of the anode and cathode. Specific energy and power were calculated by using eqn (S3)–(S5).

Author contributions

The manuscript was written through the contributions of all authors. All authors have approved the final version of the manuscript. Neetu Bansal-conceptualization, methodology, data curation, formal analysis, validation, writing-original draft; Anwar Hussain-investigation, data curation, formal analysis, writing-review & editing; Nitish Kumar-formal analysis, writing-review & editing; Changyong Park-formal analysis, data curation; Heejoon Ahn-supervision, validation, review & editing; Yusuke Yamauchi- supervision and validation; Rahul R. Salunkhe-conceptualization, supervision, validation, review & editing, funding acquisition, resources, and project administration.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the SI. Supplementary information: comparison tables, FESEM images, BET data, EDS mappings, XPS data, electrochemical data, actual device assembly images, ex situ FT-IR, postmortem analysis including ex situ FESEM, HRTEM, and optical image of device, etc. See DOI: https://doi.org/10.1039/d5ta05543a.

Acknowledgements

We acknowledge the following funding sources: Science and Engineering Research Board, Government of India, SB/S2/RJN-023/2017, Science and Engineering Research Board, Government of India, CRG/2020/003199, DST-FIST grant SR/FST/PS-I/2023/234 DST-NEST grant: DST/CEST/NEST/2024/2024/241, DRDO research grant PRAISE research grant, IIT Jammu, and Prime Minister Research Fellowship, SAPTARSHI Facility, Indian Institute of Technology Jammu. This work was partially supported by an ARC Laureate Fellowship (FL230100095) and JST-ERATO Yamauchi Materials Space Tectonics Project (JPMJER2003). The authors express their gratitude for English editing software, such as Grammarly and ChatGPT, for refining language and checking grammatical errors in this manuscript.

References

1. J. Liang and D. Wang, Adv. Energy Mater., 2022, 12, 2200920.
2. N. S. George, G. Singh, R. Bahadur, P. Kumar, K. Ramadass, C. Sathish, M. Benzigar, D. Sajan, A. Aravind and A. Vinu, Adv. Sci., 2024, 11, 2406235.
3. P. Cai, K. Wang, T. Wang, H. Li, M. Zhou, W. Wang and K. Jiang, Adv. Energy Mater., 2024, 14, 2401183.
4. D. Du, Q. Liu, J. Gao, Y. Qin, X. Jiang, Y. Shi, M. Hua, X. Lin, Z. Zhang, C. Wang, L. Yin and R. Wang, J. Mater. Chem. A, 2024, 12, 33958–33971.
5. F. Li, Y. Cao, W. Wu, G. Wang and D. Qu, Small Methods, 2022, 6, 2200411.
6. X. Liu, Z. Wu, L. Xie, L. Sheng, J. Liu, L. Wang, K. Wu and X. He, Energy Environ. Mater., 2023, 6, e12501.
7. J. Choi, H. Jeong, J. Jang, A. R. Jeon, I. Kang, M. Kwon, J. Hong and M. Lee, J. Am. Chem. Soc., 2021, 143, 9169–9176.
8. E. Adhitama, F. Dias Brandao, I. Dienwiebel, M. M. Bela, A. Javed, L. Haneke, M. C. Stan, M. Winter, A. Gomez-Martin and T. Placke, Adv. Funct. Mater., 2022, 32, 2201455.
9. E. Adhitama, M. M. Bela, F. Demelash, M. C. Stan, M. Winter, A. Gomez-Martin and T. Placke, Adv. Energy Mater., 2022, 13, 2203256.
10. Y. Qin, J. Gao, J. Lu, S. Zhang, Q. Li, C. Wang, Z. Zhang, L. Yin and R. Wang, Chem. Eng. J., 2024, 500, 157237.
11. S. Y. Ham, E. Sebti, A. Cronk, T. Pennebaker, G. Deysher, Y. T. Chen, J. A. S. Oh, J. B. Lee, M. S. Song, P. Ridley, D. H. S. Tan, R. J. Clément, J. Jang and Y. S. Meng, Nat. Commun., 2024, 15, 2991.
12. K. Schönherr, M. Pöthe, B. Schumm, H. Althues, C. Leyens and S. Kaskel, Batteries, 2023, 9, 53.
13. K. Zou, Z. Song, X. Gao, H. Liu, Z. Luo, J. Chen, X. Deng, L. Chen, G. Zou, H. Hou and X. Ji, Angew. Chem., Int. Ed., 2021, 60, 17070–17079.
14. Y. Yang, J. Wang, S. C. Kim, W. Zhang, Y. Peng, P. Zhang, R. A. Vilá, Y. Ma, Y. K. Jeong and Y. Cui, Nano Lett., 2023, 23, 5042–5047.
15. Y. Shen, J. Zhang, Y. Pu, H. Wang, B. Wang, J. Qian, Y. Cao, F. Zhong, X. Ai and H. Yang, ACS Energy Lett., 2019, 4, 1717–1724.
16. X. Min, G. Xu, B. Xie, P. Guan, M. Sun and G. Cui, Energy Storage Mater., 2022, 47, 297–318.
17. C. Yang, H. Ma, R. Yuan, K. Wang, K. Liu, Y. Long, F. Xu, L. Li, H. Zhang, Y. Zhang, X. Li and H. Wu, Nat. Energy, 2023, 8, 703–713.
18. C. Li, Y. An, L. Wang, K. Wang, X. Sun, H. Zhang, X. Zhang and Y. Ma, Chem. Eng. J., 2024, 485, 149880.
19. W. J. Cao and J. P. Zheng, J. Power Sources, 2012, 213, 180–185.
20. C. Wang, F. Yang, W. Wan, S. Wang, Y. Zhang, Y. Huang and J. Li, Energy Environ. Sci., 2023, 16, 4660–4669.
21. X. Yue, Y. Yao, J. Zhang, S. Yang, Z. Li, C. Yan and Q. Zhang, Adv. Mater., 2022, 34, 2110337.
22. S. Y. Yang, X. Y. Yue, W. Q. Dai, D. L. Wang, H. Y. Xia, Y. Qiao and Z. W. Fu, Chem. Commun., 2021, 57, 10371–10374.
23. M. Yi, Z. Cui and A. Manthiram, ACS Appl. Mater. Interfaces, 2024, 16, 42270–42282.
24. J. Jiang, P. Nie, B. Ding, Y. Zhang, G. Xu, L. Wu, H. Dou and X. Zhang, J. Mater. Chem. A, 2017, 5, 23283–23291.
25. H. J. Kim, S. Choi, S. J. Lee, M. W. Seo, J. G. Lee, E. Deniz, Y. J. Lee, E. K. Kim and J. W. Choi, Nano Lett., 2016, 16, 282–288.
26. B. Liu, J. Chen, B. Yang, L. Liu, Y. Sun, R. Hou, Z. Lin and X. Yan, J. Mater. Chem. A, 2021, 9, 10722–10730.
27. N. Liu, L. Hu, M. T. McDowell, A. Jackson and Y. Cui, ACS Nano, 2011, 5, 6487–6493.
28. N. Bansal, K. P. Agrim, A. M. Bohra, T. Ahamad, C. Park, H. Ahn and R. R. Salunkhe, ACS Appl. Nano Mater., 2024, 7, 17305–17315.
29. P. Liu, P. M. Abdala, G. Goubert, M. Willinger and C. Copéret, Angew. Chem., Int. Ed., 2021, 60, 3254–3260.
30. X. Peng, Y. Li, F. Kang, X. Li, Z. Zheng and L. Dong, Small, 2024, 20, 2305547.
31. J. Ding, M. Du, S. Wang, L. Zhang, Y. Yue and M. M. Smedskjaer, Energy Environ. Sci., 2025, 18, 1587–1611.
32. N. Kumar, N. Bansal, Y. Yamauchi and R. R. Salunkhe, Chem. Mater., 2022, 34, 4946–4954.
33. Y. J. Oh, J. H. Park, J. S. Park, S. S. Kim, S. J. Hong, Y. W. Na, J. H. Kim, S. Nam and S. J. Yang, Energy Storage Mater., 2022, 44, 416–424.
34. Y. Wu, P. Nie, J. Jiang, B. Ding, H. Dou and X. Zhang, ChemElectroChem, 2017, 4, 1560–1565.
35. L. Liu, Y. X. Yin, J. Y. Li, S. H. Wang, Y. G. Guo and L. J. Wan, Adv. Mater., 2018, 30, 1706216.
36. W. S. V. Lee, X. Huang, T. L. Tan and J. M. Xue, ACS Appl. Mater. Interfaces, 2018, 10, 1690–1700.
37. M. Bruna, A. K. Ott, M. Ijäs, D. Yoon, U. Sassi and A. C. Ferrari, ACS Nano, 2014, 8, 7432–7441.
38. J. Liu, Y. You, L. Huang, Q. Zheng, Z. Sun, K. Fang, L. Sha, M. Liu, X. Zhan, J. Zhao, Y. Han, Q. Zhang, Y. Chen, S. Wu and L. Zhang, Adv. Mater., 2024, 36, 2407369.
39. Y. K. Wang, M. C. Liu, J. Cao, H. J. Zhang, L. B. Kong, D. P. Trudgeon, X. Li and F. C. Walsh, ACS Appl. Mater. Interfaces, 2020, 12, 3709–3718.
40. X. Liu, Y. Wang, Y. Yang, W. Lv, G. Lian, D. Golberg, X. Wang, X. Zhao and Y. Ding, Nano Energy, 2020, 70, 104550.
41. A. P. Cohn, N. Muralidharan, R. Carter, K. Share, L. Oakes and C. L. Pint, J. Mater. Chem. A, 2016, 4, 14954–14959.
42. W. Bao, J. Wan, X. Han, X. Cai, H. Zhu, D. Kim, D. Ma, Y. Xu, J. N. Munday, H. D. Drew, M. S. Fuhrer and L. Hu, Nat. Commun., 2014, 5, 4224.
43. A. R. Neale, D. C. Milan, F. Braga, I. V. Sazanovich and L. J. Hardwick, ACS Energy Lett., 2022, 7, 2611–2618.
44. A. Shellikeri, V. Watson, D. Adams, E. E. Kalu, J. A. Read, T. R. Jow, J. S. Zheng and J. P. Zheng, J. Electrochem. Soc., 2017, 164, A3914–A3924.
45. D. Huang, Z. Wang, R. Han, S. Hu, J. Xue, Y. Wei, H. Song, Y. Liu, J. Xu, J. Ge and X. Wu, Adv. Sci., 2023, 10, 2302966.
46. C. Fang, J. Li, M. Zhang, Y. Zhang, F. Yang, J. Z. Lee, M. H. Lee, J. Alvarado, M. A. Schroeder, Y. Yang, B. Lu, N. Williams, M. Ceja, L. Yang, M. Cai, J. Gu, K. Xu, X. Wang and Y. S. Meng, Nature, 2019, 572, 511–515.
47. Q. Wang, Z. Yao, C. Zhao, T. Verhallen, D. P. Tabor, M. Liu, F. Ooms, F. Kang, A. Aspuru-Guzik, Y. S. Hu, M. Wagemaker and B. Li, Nat. Commun., 2020, 11, 4188.
48. X.-P. Zhang, Y. Y. Sun, Z. Sun, C. S. Yang and T. Zhang, Nat. Commun., 2019, 10, 3543.
49. B. S. Parimalam, A. D. MacIntosh, R. Kadam and B. L. Lucht, J. Phys. Chem. C, 2017, 121, 22733–22738.
50. Y. Gao, T. Rojas, K. Wang, S. Liu, D. Wang, T. Chen, H. Wang, A. T. Ngo and D. Wang, Nat. Energy, 2020, 5, 534–542.
51. R. T. Pekarek, A. Affolter, L. L. Baranowski, J. Coyle, T. Hou, E. Sivonxay, B. A. Smith, R. D. McAuliffe, K. A. Persson, B. Key, C. Apblett, G. M. Veith and N. R. Neale, J. Mater. Chem. A, 2020, 8, 7897–7906.
52. A. Dopilka, Y. Gu, J. M. Larson, V. Zorba and R. Kostecki, ACS Appl. Mater. Interfaces, 2023, 15, 6755–6767.
53. V. R. Rikka, S. R. Sahu, A. Chatterjee, P. V. Satyam, R. Prakash, M. S. R. Rao, R. Gopalan and G. Sundararajan, J. Phys. Chem. C, 2018, 122, 28717–28726.
54. B. Han, Z. Zhang, Y. Zou, K. Xu, G. Xu, H. Wang, H. Meng, Y. Deng, J. Li and M. Gu, Adv. Mater., 2021, 33, 2100404.
55. F. Wang, J. Graetz, M. S. Moreno, C. Ma, L. Wu, V. Volkov and Y. Zhu, ACS Nano, 2011, 5, 1190–1197.
56. S. Lee, K. Lee, S. Kim, K. Yoon, S. Han, M. H. Lee, Y. Ko, J. H. Noh, W. Kim and K. Kang, Sci. Adv., 2022, 8, eabq0153.
57. S. K. Heiskanen, J. Kim and B. L. Lucht, Joule, 2019, 3, 2322–2333.
58. H. Y. Wei, D. S. Tsai and C. L. Hsieh, RSC Adv., 2015, 5, 69176–69183.
59. Y. Xiao, X. Chen, J. Jian, Y. Cheng, Y. Zou, Y. Su, Q. Wu, C. Tang, Z. Zhang, M. Wang, J. Zheng and Y. Yang, Small, 2024, 20, 2308472.
60. Y. Li, K. Wu, H. Luo, M. Li, L. Wang, K. Sun and Y. Zheng, Phys. Rev. Appl., 2024, 21, 024038.
61. Y. Wang, D. Xiao, Y. Xu and X. Zhang, ACS Appl. Energy Mater., 2024, 7, 11525–11532.
62. L. Caizán-Juanarena, M. Arnaiz, E. Gucciardi, L. Oca, E. Bekaert, I. Gandiaga and J. Ajuria, Adv. Energy Mater., 2021, 21, 2100912.
63. T. Liang, Z. Mao, L. Li, R. Wang, B. He, Y. Gong, J. Jin, C. Yan and H. Wang, Small, 2022, 18, 2201792.
64. S. Jeon, S. Lm, I. Kang, D. Shin, S. Yu, M. Lee and J. Hong, Small, 2024, 20, 2401295.
65. B. Liu, J. Chen, B. Yang, L. Liu, Y. Sun, R. Hou, Z. Lin and X. Yan, J. Mater. Chem. A, 2021, 9, 10722–10730.
66. C. Cheng, D. Wu, T. Gong, Y. Yan, Y. Liu, W. Ji, L. Hou and C. Yuan, Adv. Energy Mater., 2023, 13, 2302107.
67. K. Liang, K. Zou, J. Liu, Y. Deng and G. Chen, ACS Energy Lett., 2023, 8, 3204–3213.
68. K. Su, Y. Wang, B. Yang, X. Zhang, W. Wu, J. Lang and X. Yan, Energy Environ. Mater., 2023, 6, e12506.
69. S. Tao, R. Momen, Z. Luo, Y. Zhu, X. Xiao, Z. Cao, D. Xiong, W. Deng, Y. Liu, H. Hou, G. Zou and X. Ji, Small, 2023, 19, 2207975.
70. Y. Ma, K. Wang, Y. Xu, X. Zhang, Q. Peng, Y. Guo, X. Sun, X. Zhang, Z. Wu and Y. Ma, Adv. Energy Mater., 2024, 14, 2304408.
71. W. Wei, L. Wang, C. Liang, W. Liu, C. Li, Y. An, L. Zhang, X. Sun, K. Wang, H. Zhang, X. Zhang and Y. Ma, Chem. Eng. J., 2023, 474, 145788.
72. H. Jiang, D. Shi, X. Sun, S. Wang, Y. Li, B. Chang, B. Zhang, Y. Shao, Y. Wu and X. Hao, ACS Appl. Mater. Interfaces, 2020, 12, 47425–47434.

---