Realizing Environmentally Scalable Pre-Lithiation via Protective Coating of LiSi Alloys to Promote High-Energy-Density Lithium-Ion Batteries

Abstract

Pre-lithiation using Li–Si alloy-type additives is a promising technical approach to address the drawbacks of Si-based anodes, such as a low initial Coulombic efficiency (ICE) and inevitable capacity decay during cycling. However, its commercial application is limited by the air sensitivity of the highly reactive Li–Si alloys, which demands improved environmental stability. In this work, a protective membrane is constructed on Li₁₃Si₄ alloys using low-surface-energy paraffin and highly conductive carbon nanotubes through liquid-phase deposition, exhibiting enhanced hydrophobicity and improved Li⁺/e⁻ conductivity. The Li₁₃Si₄@Paraffin/carbon nanotubes (Li₁₃Si₄@P-CNTs) composite achieves a high pre-lithiation capacity of 970 mAh g⁻¹ and superb environmental stability, retaining 92.2% capacity after exposure to ambient air with 45% relative humidity. DFT calculations and in situ XRD measurements reveal that the paraffin-dominated coating membrane, featuring weak dipole–dipole interactions with water molecules, effectively reduces the moisture-induced oxidation kinetics of Li₁₃Si₄@P-CNTs in air. Electrochemical kinetic analysis and XPS depth profiling reveal the enhancement in charge transfer dynamics and surface Li⁺ transport kinetics (SEI rich in inorganic lithium salts) in P-SiO@C pre-lithiated by Li₁₃Si₄@P-CNTs pre-lithiation additives. Benefitting from pre-lithiation via Li₁₃Si₄@P-CNTs, the pre-lithiated SiO@C(P-SiO@C) delivers high ICE (103.7%), stable cycling performance (981 mAh g⁻¹ at 200 cycles) and superior rate performance (474.5 mAh g⁻¹ at 3C) in a half-cell system. The LFP||P-Gr pouch-type full cell exhibits a capacity retention of 83.2% (2500 cycles) and an energy density of 381 Wh kg⁻¹ after 2500 cycles. The Li₁₃Si₄@P-CNTs additives provide valuable design concepts for the development of pre-lithiation materials.

1. Introduction

Li-ion batteries (LIBs) are considered ideal alternatives to fossil fuels due to their extensive application in electric vehicles (EVs), consumer electronics and energy storage systems [1,2,3]. Despite the rapid development of LIBs, their intrinsic limitation of low energy density cannot meet the requirement of the long driving range and high power output of EVs [4]. A high commercial standard of 350 Wh kg⁻¹ at the cell level is provided by the Department of Energy and Advanced Battery Consortium LLC in the United States [5]. Presently, graphite is the main commercial anode material due to its high abundance, low lithiation potential (0.1 V vs. Li/Li⁺) [6] and small volume expansion (10%) [7]. However, the low theoretical capacity (372 m Ah g⁻¹, LiC₆) [8] restricts the energy density of LIBs. Silicon (Si)-based materials are potential alternatives to graphite due to their high theoretical capacity (3579 mAh g⁻¹, Li₁₅Si₄) [9]. Nevertheless, Si-based materials suffer from limited conductivity, significant volume expansion (180~300%) and fracture during battery operation [10], thereby leading to poor initial Coulombic efficiency (ICE) and rate performance [11,12]. It is well known that the ICE ranges of graphite materials are 80–92% [13,14,15], while Si or SiO_x materials have relatively low ICEs of 65–85% or 65–82% [16,17,18,19,20,21] due to the irreversible side reactions from SEI formation and huge volume variations during cycling [22,23]. It is critical to promote advanced pre-lithiation technology to solve the problems of low ICE, reversible capacity, rate performance and energy density of Si-based anodes and their corresponding LIBs.

Pre-lithiation is an effective method for introducing extra active Li into batteries before cycling. To expand the commercial application of high-theoretical-capacity Si-based anode materials, various pre-lithiation strategies have been developed and divided into several categories: powder pre-lithiation (using Li_xSi, LiC_y and Li metal powders) and Li foil pre-lithiation (including round-shape/roll-to-roll press Li foil contact, Li deposition, electrochemical pre-lithiation and chemical pre-lithiation), based on the type of lithium sources [24,25,26]. It is worth noting that lithium foil pre-lithiation, achieved using Li metal with high intrinsic chemical activity, suffers from significant pre-lithiation capacity fade in the air (H₂O + Li → LiOH + H₂), which hinders its potential for commercial anode pre-lithiation [27]. On the other hand, a series of powder pre-lithiation methods suitable for the anode manufacturing environment have been proposed based on environmentally stable Li-containing powders. This makes it promising for large-scale commercialization of anode pre-lithiation to a certain extent. Generally, various air-stable Li-containing pre-lithiation powders have been synthesized, primarily through chemical modifications, such as surface oxidation, fluorination and acid self-assembly passivation [28,29]. Zhao et al. prepared core–shell Li₂₁Si₅@Li₂O nanoparticles via oxygen-induced surface oxidization of Li₂₁Si₅ nanoparticles. The pre-lithiation capacity retention of Li₂₁Si₅@Li₂O decreased to 70% after exposure to dry air for 120 h [30]. Additionally, Cui’s group proposed a fluorinated alkane (1-fluorodecane) to achieve artificial SEI formation on the surface of Li₂₁Si₅. An inorganic/organic coating layer (LiF and Li alkyl carbonate) was constructed on Li₂₁Si₅ through interface modification. Although the modified Li₂₁Si₅ only showed slight capacity degradation in humid air with 10% relative humidity(RH), it cannot retain stability in high-level-humidity air and polar solvents [31]. Zhao et al. synthesized Li₂₂Si₅@LiF nanoparticles via a chemical vapor deposition with F₂. The pre-lithiation capacity of Li₂₂Si₅@LiF nanoparticles slightly decreased from 2879 to 2328.9 mAh g⁻¹ after exposure to air with 40% RH for 24 h [32]. In summary, current studies about surface-modified Li-containing powder pre-lithiation additives are still difficult to apply for industrial pre-lithiation in an ambient air environment (>40% RH). Therefore, developing highly environmentally stable pre-lithiation additives suitable for industrial anode fabrication is essential and extensive for the widespread commercialization of LIBs with high energy densities.

In this work, hydrophobic paraffin and highly conductive carbon nanotubes (CNTs) were collaboratively employed to construct a coating membrane on Li₁₃Si₄ microparticles to prepare Li₁₃Si₄@P-CNTs pre-lithiation additives via liquid phase deposition. The paraffin/carbon nanotubes composite membrane achieved excellent environmental stability and strong Li⁺/e⁻ transfer kinetics. In situ, X-ray diffraction (XRD) and theoretical calculation reveal the mechanism of the composite membrane decreasing the moisture-induced oxidization kinetics of Li₁₃Si₄@P-CNTs in the air. Consequently, the highly environmentally stable Li₁₃Si₄@P-CNTs deliver an excellent pre-lithiation capacity of 895 mAh g⁻¹ and capacity retention of 92.2% after exposure in 45% RH air for 6 h. Electrochemical kinetic analysis and XPS depth etching data indicate that the P-SiO@C pre-lithiated by Li₁₃Si₄@P-CNTs greatly boosts the charge transfer dynamics and surface Li⁺ transport kinetics of the P-SiO@C electrode. Furthermore, the P-SiO@C half cell achieves a high ICE of 103.7% and the LFP||P-Gr pouch cell demonstrates a superior energy density of 381 Wh kg⁻¹ over 2500 cycles. This enables large-scale commercial pre-lithiation with effective and industrially compatible pre-lithiation additives.

2. Results and Discussion

2.1. Synthesis and Morphology of Li₁₃Si₄@P-CNTs

Li–Si alloys are easily oxidized by moisture or oxygen in the air due to the enrichment of highly electropositive Li atoms, which makes them unable to compensate active lithium for anode materials [33,34]. Paraffin molecules are composed of saturated hydrocarbon chains. Their repeated nonpolar structural segments (–CH₂–CH₂–) and low surface energy result in the superior hydrophobicity of paraffin [35]. However, considering the poor Li⁺/e⁻ conductivity of paraffin, ball-milled CNTs were utilized to in situ crosslink on the surface of Li₁₃Si₄ microparticles in a melted paraffin/DOL solution. Furthermore, paraffin/CNTs composites enclosing Li₁₃Si₄ microparticles were obtained via liquid phase deposition, followed by cooling (Figure 1a). The morphological features and microstructures of the Li₁₃Si₄@P-CNTs and Li₁₃Si₄ precursors were investigated via scanning electron microscope (SEM). As shown in Figure 1b and Figure S1a, pristine Li₁₃Si₄ microparticles exhibit an elliptical morphology, with particle dimensions ranging from 2 to 10 μm. The blurred and rough surface of particles illustrates the poor conductivity of pristine Li₁₃Si₄. Paraffin presented a solid-state irregular block structure (10 μm) in observation via SEM (Figure 1c). It can be seen in Figure 1d and Figure S1b that plenty of wire-like CNTs wrap the Li₁₃Si₄ microparticles, exhibiting a uniform crosslinked network coating of the Li₁₃Si₄@P-CNTs. Bright wires following the shape of the CNT bundles indicate the enhancement of conductivity of Li₁₃Si₄@P-CNTs after coating. The distribution of elements of pristine Li₁₃Si₄ and Li₁₃Si₄@P-CNTs was characterized using SEM-energy dispersive spectrometer (EDS) element mapping. The EDS mappings reveal that the Li₁₃Si₄ precursors and paraffin show higher abundances of Si and C, respectively (Figure 1e,f and Figure S2). However, the appearance of C in pristine Li₁₃Si₄ may be ascribed to the Li₂CO₃ formed due to the oxidation of Li–Si alloys (Li_xSi → Li₂CO₃) during the SEM sample preparation process in air. The C element is evenly coated on the surface of Li₁₃Si₄@P-CNTs (Figure 1g), indicating the homogeneous 3D spatial combination of paraffin/CNTs composite in Li₁₃Si₄@P-CNTs. CNTs in the paraffin-CNT layer provide lithium ion transport paths and facilitate charge transfer. Under an applied external voltage, lithium ions migrate from the electrolyte to the inner electrode via CNTs and initiate the alloying reaction (Figure S3).

2.2. Crystal Properties and Surface Chemistry Characteristics of Li₁₃Si₄@P-CNTs

The characterization of crystal structure and chemical composition are used to further analyze the Li₁₃Si₄@P-CNTs properties. XRD profiles (Figure 2a) show diffraction peaks for pristine Li₁₃Si₄ at 20.8°, 23.1°, 23.6°, 40.4°, 41.1°, 41.3° and 41.6°, corresponding to the (130), (210), (111), (002), (061), (321), (340) and (222) lattice planes of Li₁₃Si₄ (PDF#223-8940). Meanwhile, the XRD pattern of Li₁₃Si₄@P-CNTs is basically consistent with the standard Li₁₃Si₄ phase, while the presence of the additional peak at 21.3° corresponds to the (021) lattice plane of Li_xC [36], indicating the lithiation of CNTs after paraffin/CNTs composite coating the Li₁₃Si₄ microparticles(C + Li₁₃Si₄ → Li_xC + Li_xSi). The Raman spectrum in Figure 2b presents three peaks: 508 cm⁻¹ (Si-Si stretching vibrations), 1340 cm⁻¹ (D band of CNTs) and 1596 cm⁻¹ (G band of CNTs), respectively [37,38]. The characteristic peak at 508 cm⁻¹ appears in both the Raman spectrum of the Li₁₃Si₄ precursor and Li₁₃Si₄@P-CNTs, reflecting the relaxation of Si-Si bonds in the c-Li₁₃Si₄ lattice. Meanwhile, peaks at 1340 cm⁻¹ (D band) and 1596 cm⁻¹ (G band) are only observed in the Raman spectrum of Li₁₃Si₄@P-CNTs, illustrating that the carbon nanotubes adhere to the surface of the Li₁₃Si₄ microparticles after liquid phase coating. The components of the paraffin mixture were further investigated using pyrolysis gas chromatography–mass spectrometry (Py-GCMS). The multi-component paraffin primarily consists of Octacosane and Tetratetracontane (Figure 2c), with straight–chain saturated alkane Octacosane being the most abundant component.

The surface chemistry characteristics of Li₁₃Si₄ and Li₁₃Si₄@P-CNTs were further investigated using X-ray photoelectron spectroscopy (XPS). As shown in Figure 2d, the intense peaks of Li₁₃Si₄ and Li₁₃Si₄@P-CNTs XPS Li 1s spectrum at a binding energy of 54.7 eV are ascribed to the Li-Si bond, while the intense peak at 54.2 eV can be observed on the spectrum of Li₁₃Si₄@P-CNTs, attributed to the Li_xC bond [39,40]. For the XPS Si 2p spectrum (Figure 2e), Li₁₃Si₄ exhibits an intense peak at 97 eV (Li-Si bond), which is significantly weaker in Li₁₃Si₄@P-CNTs [41]. The C 1s spectrum (Figure 2f) shows peaks at 282.8, 284.8, 286.5 and 289.5 eV, corresponding to Li-C (lithiation of CNTs), C-C/C-H (paraffin, CNTs and C in conductive adhesive base), C-O (O in conductive adhesive base) and C-F (F in conductive adhesive base), respectively [42,43]. Based on the results of XPS analysis, the Li-C (lithiated CNTs) bonds, increasing intensity of C-C/C-H (paraffin/CNTs composite coating) bonds and decreasing intensity of Li-Si bonds after coating further illustrates that a uniform paraffin/CNTs composite interface is successfully constructed on the surface of Li₁₃Si₄ microparticles based on liquid phase deposition. Figure S4 displays thermogravimetry (TG) curves of three samples (Li₁₃Si₄@P-CNTs, Li₁₃Si₄@CNTs and Li₁₃Si₄) heating from 30 to 350 °C. Only Li₁₃Si₄@P-CNTs shows a 12% mass loss between 160 and 300 °C, corresponding to the volatilization of the paraffin in the coating. On the other hand, a slight weight increment is observed in both Li₁₃Si₄@CNTs and Li₁₃Si₄, which is assigned to the reaction between Li_xSi and N₂. These results illustrate that paraffin successfully coats the surface of Li₁₃Si₄ microparticles. Li₁₃Si₄@P-CNTs, Li₁₃Si₄@CNTs (obtained by ball milling Li₁₃Si₄ and CNTs composites), paraffin and CNTs are 0.1 mV, −0.1 mV, 0.7 mV and −0.3 mV, respectively (Figure S5). The Zeta potential shifts positively from −0.1 (Li₁₃Si₄@CNTs) to 0.1 mV due to the introduction of paraffin with positive surface charge (0.7 mV), illustrating that paraffin successfully coated on Li₁₃Si₄ microparticles.

2.3. Environmental Stability and Structural Evolution of Li₁₃Si₄@P-CNTs in Air

The coating membrane is predominantly composed of nonpolar straight-chain saturated alkane molecules (rich in repeating -CH₂- units). This spatial arrangement of the chain decreases polar interactions with H₂O molecules. The environmental stability and structural evolution of the Li₁₃Si₄@P-CNTs and pristine Li₁₃Si₄ pre-lithiation additives in air (45% RH) for 120 min are further systematically investigated through in situ XRD characterization. In the in situ XRD spectra of Li₁₃Si₄@P-CNTs pre-lithiation additives with air exposure (Figure 3a), the characteristic peaks are at 20.8°, 23.1°, 23.6°, 40.4°, 41.1°, 41.3°, 41.6° and 21.3°, corresponding to the (130), (210), (111), (002), (061), (321), (340) and (222) lattice planes of Li₁₃Si₄ and the (021) lattice plane of Li_xC. The intensities and positions of all the above peaks remain constant before and after the long-term air exposure process, illustrating that Li₁₃Si₄@P-CNTs exhibit enhanced environmental stability in humid air (45% RH). On the other hand, as shown in Figure 3b, the attenuation of Li₁₃Si₄/Li_xC characteristic peaks and enhancement of the LiOH characteristic peak represent the crystalline phase transition from Li₁₃Si₄ to LiOH during the air exposure of pristine Li₁₃Si₄, suggesting the high air sensitivity of pristine Li₁₃Si₄ without hydrophobic paraffin/CNTs composite coating membrane.

2.4. Theoretical Study on the Hydrophobic Mechanism of Paraffin-Dominant Coating Membrane

To further understand the hydrophobicity and environmental stability exhibited by the Li₁₃Si₄@P-CNTs after paraffin coating, density functional theory (DFT) and molecular dynamics simulations were performed and analyzed (paraffin was replaced by its main component, Octacosane, C₂₈H₅₈). As shown in the electrostatic potential mapping image of the water molecule (Figure 4a and Figure S6a), the local electrostatic potential of H₂O varies from −0.479 eV to 0.240 eV. The H atoms and O atoms separately exhibit strong positive potential and negative potential, indicating that the H₂O molecule exhibits strong polarity. However, the low difference in surface electrostatic potential (−0.084 eV~0.042 eV) and weak polarity can be observed in paraffin (C₂₈H₅₈) in Figure 4b and Figure S6b. Furthermore, a charge density difference was measured to qualitatively analyze the charge density distribution at the hetero-interface of H₂O-C₂₈H₅₈. Figure 4c illustrates the charge depletion in cyan and charge accumulation in yellow. Meanwhile, the charge density distribution (Figure 4c,d) further illustrates that the hetero-interface zones exhibit a small concentration of both positive and negative charge densities. The positive charges exhibit a low tendency to accumulate on H atoms of C₂₈H₅₈ and the negative ones are slightly probable to gather on the O atom of H₂O. This illustrates that paraffin coating exhibits weak dipole–dipole interaction with negatively charged O atoms of H₂O (van der Waals force). Therefore, the paraffin-dominant hydrophobic coating membrane of Li₁₃Si₄@P-CNTs improves their environmental stability in humid air.

To theoretically understand the water molecule transport behavior through the paraffin-dominant membrane around Li₁₃Si₄ microparticles, molecular dynamics simulations were performed. The canonical ensemble (NVT) and the unrestrained micro-canonical ensemble (NVE) were used in the simulation. The model for the H₂O-paraffin system, containing eight C₂₈H₅₈ and five H₂O molecules, is shown in Figure 4e. As shown in Figure 4f, the diffusion coefficient of water in the H₂O-paraffin system is 6.2 × 10⁻⁵ cm²·s⁻¹, much lower than that in the air (0.26 cm² s⁻¹) [44]. Thus, it can be concluded that the external paraffin-dominant membrane of Li₁₃Si₄@P-CNTs inhibits the diffusion of water molecules. FFV is a characteristic used to quantify molecular transport performance in a transport medium [45]. The pink surface in Figure 4g illustrates the free volume of molecules in the H₂O-paraffin system. The FFV is 20.3%, reflecting the low water permeability of the dense and uniform paraffin-dominant membrane on Li₁₃Si₄. These results indicate that the paraffin-dominant membrane, exhibiting weak interaction with water molecules, imparts Li₁₃Si₄@P-CNTs with superior hydrophobicity and enhances its environmental stability.

2.5. Electrochemical Performance of Li₁₃Si₄@PFPE/LiF Pre-Lithiated Anode

To verify the intrinsic electrochemical performance of Li₁₃Si₄@P-CNTs and its beneficial effect on improving the performance of half cells, a galvanostatic charge–discharge test and rate performance analysis are investigated in the section below. As shown in Figure 5a, the pre-lithiation capacities of Li₁₃Si₄@P-CNTs anodes before and after exposure to 45% RH air for 6 h are 970 and 895 mAh g⁻¹ (based on the total mass of Li₁₃Si₄, paraffin and CNT), respectively.

Furthermore, the pre-lithiation capacity retention is 92.2%, indicating that the Li₁₃Si₄@P-CNTs, with a hydrophobic paraffin-dominant coating, retains a superior pre-lithiation capacity after exposure to air. The protective effect of the paraffin/CNTs composite membrane on Li₁₃Si₄@P-CNTs is further validated by their electrochemical performance after long-term air exposure and polar solvent immersion tests. Li₁₃Si₄@P-CNTs still achieves a specific capacity of 413 mAh g⁻¹ after being exposed to 45% RH air for 360 h (Figure S7a) and shows a capacity of 401 mAh g⁻¹ after immersion in THF for 720 h (Figure S7b), illustrating that Li₁₃Si₄@P-CNTs retains a chemical structure and electrochemical activity after exposure to air or polar solvent. Half cells with different addition amounts of Li₁₃Si₄@P-CNTs were used to further investigate the impact of the pre-lithiation degree on the ICEs and cyclability of SiO@C or Gr anodes. As shown in Figure 5b, the introduction of 15% and 30% Li₁₃Si₄@P-CNTs to SiO@C anodes results in a significant increase in ICEs, from 82.8% for the pristine anode to 103.7% and 127.8%, respectively. A similar trend is observed for pre-lithiated Gr (P-Gr) anodes in Figure 5c, where the ICE increases from 88.1% (pristine) to 98.5% (5% addition) and 109.1% (10% addition). These results demonstrate that the Li₁₃Si₄@P-CNTs effectively facilitates the pre-lithiation of both Si-based and graphite anodes, with the pre-lithiation degree precisely regulated by the amount of Li₁₃Si₄@P-CNTs. Furthermore, for the pristine SiO@C anode and P-SiO@C anodes with various amounts of 15% and 30% Li₁₃Si₄@P-CNTs, the capacity retentions after 200 cycles are 43.5%, 66.8% and 53.7%, respectively (Figure 5d). Specifically, the P-SiO@C anode with a moderate pre-lithiation level (15%) exhibits the best cycling stability and is designated as P-SiO@C for subsequent research.

The rate performance of the pristine SiO@C and P-SiO@C anodes was further evaluated. As shown in Figure 5e, the P-SiO@C anode exhibits a superior rate capability with discharge capacities of 1009, 686 and 474.5 mAh g⁻¹ at 1 C, 2 C and 3 C, respectively. On the other hand, the SiO@C anode retains only 559, 273 and 99 mAh g⁻¹ discharge capacities at corresponding rates. EIS analysis was performed to investigate the effect of Pre-SEI after pre-lithiation on the impedance of the SiO@C anodes. As shown in Figure 5f, Nyquist plots usually show up as sloping straight lines in the low-frequency region and half circles in the high-frequency range. The kinetic parameters of both anodes were fitted based on the equivalent circuits in Figure S8. R_s represents the bulk resistance of the battery components, while R_ct corresponds to the charge-transfer resistance and CPE denotes the double-layer capacitance [46]. Charge-transfer resistance is associated with the mid- to low-frequency range and the high-frequency region represents the impedance of Li⁺ passing through SEI [47]. The cycled P-SiO@C exhibits an R_ct of 98.2 and R_surf of 23.7 Ω, which are lower than that of SiO@C (R_ct = 105.2 Ω and R_surf = 30.5 Ω) after 200 cycles. Meanwhile, in comparison to pristine SiO@C, a steeper slope tail of the EIS profile of P-SiO@C indicates improved Li-ion diffusion after pre-lithiation. The results further illustrate that the kinetics of the cycled SiO@C enhance with the pre-lithiation treatment.

2.6. SEI Regulation of Pre-Lithiation SiO@C via Li₁₃Si₄@P-CNTs

The SEI formation evolution and volume expansion of pre-lithiated anodes after cycling are systematically investigated in this section. To further analyze the chemical components of the anode surface before and after pre-lithiation of SiO@C at different thicknesses, the electrode surface is deeply etched at 5-second intervals with etching times of 0, 5 and 10 s. As shown in the XPS C 1s spectrum of SiO@C/P-SiO@C anodes (Figure 6a,d), the peaks with binding energies of 281.5 eV, 283.2 eV, 284.8 eV, 286.5 eV, 289 eV and 290 eV correspond to characteristic peaks associated with bonds of Si-C, Li-C, C-C/C-H, C-O, C-C and C=O. As shown in Figure 6b,e, the deconvolution of XPS Li 1s peaks can be divided into Li-O (51.5 eV), Li-C (52.8 eV), ROCOOLi (55 eV) and LiF (55.7 eV) [48,49,50,51]. The intense peaks near 682.5 eV, 684.7 eV and 687 eV originate from F⁻, LiF and LiPOF_z, respectively [52,53]. These results indicate that surface components of SiO@C and P-SiO@C are composed of Li_xC, SiC, free fluoride ion, organic lithium salts (ROCOOLi) and inorganic lithium salts (LiF, Li₂O and LiPOF_z).

Furthermore, the variation in SEI components at different etching levels is clarified through the XPS data of SiO@C and P-SiO@C. As the etch deepens, there are obvious decreases in the proportion and peak intensity of C-C/C-H/C-O organics (C 1s in Figure 6a,d), while the intensity of diffraction peaks of inorganic Li salts (LiF and Li₂O) increases in the P-SiO@C anode (Li 1s and F 1s in Figure 6b,c,e,f). In contrast to P-SiO@C, with the increased sputtering time, Figure 6d exhibits a decrease in organics (ROCOOLi, C-C, C-H and C-O) in SEI of SiO@C, while Figure 6e,f show no consistent trend of lithium salts (LiF and Li₂O). Meanwhile, P-SiO@C has significantly fewer organic SEI components and more inorganic Li salts (e.g., LiF and Li₂O) on the surface of the electrode in comparison to pristine SiO@C, illustrating that pre-lithiation may regulate components of SEI and enhance its Li⁺ transport kinetics. To evaluate the structural variation in the SiO@C/P-SiO@C anode during galvanostatic cycling, two anodes with equal thickness were fully lithiated in half cells and then observed via SEM. The cross-sectional SEM figures of fully lithiated P-SiO@C and SiO@C electrodes after cycling are presented in Figure 6g and Figure 6h, respectively. The cross-sectional image of the cycled P-SiO@C exhibits a basically even, stable and intact interface with thickness from 22 μm to 36 μm, whereas the cross-section of the cycled SiO@C shows an inhomogeneous electrode (28 μm to 57 μm) with many internal cracks/ruptures and active materials separation, attributed to stress failures caused by the repeated significant volume change in SiO@C during cycling. Therefore, the absence of cracks and decreasing thickness of the cross-section of cycled anodes illustrate that pre-lithiation of SiO@C via Li₁₃Si₄@P-CNTs alleviates the volume variation in SiO@C and enhances the mechanical stability of anodes.

2.7. Full Cell Model Evaluation Based on Li₁₃Si₄@PFPE/LiF Pre-Lithiation Strategy

The commercial pre-lithiation potential of Li₁₃Si₄@P-CNTs was further investigated in full-cell and pouch-cell configurations with the pristine/pre-lithiated Gr or SiO@C anodes. The ICE increases from 66.5% to 90.8% in NCM811||SiO@C full cells before and after pre-lithiation (Figure 7a). This verifies that the Li₁₃Si₄@P-CNTs powder pre-lithiation compensates for the active Li loss induced by SEI formation during the initial cycle. The cyclability of pristine/pre-lithiated full cells was further studied, as shown in Figure 7b. The specific capacity of NCM811||SiO@C battery rapidly declines from 121.2 to 90.8 mAh g ⁻¹ after 150 cycles with a low capacity retention (74.2%). On the other hand, the pre-lithiated cell exhibits a high discharge capacity of 149.1 mAh g⁻¹ and capacity retention of 86.8% during 150 cycles. As shown in Figure 7c, the NCM811||P-SiO@C exhibits an average discharge capacity of 141.19, 129.66, 109.9 and 91.5 mAh g⁻¹, with various rates at 0.5, 1, 2 and 3 C, respectively. Meanwhile, the NCM811||P-SiO@C can still return to a high discharge capacity of 139.8 mAh g⁻¹ after 35 cycles. In contrast, the pristine SiO@C shows a low rate performance, only maintaining discharge capacities of 117.8, 102.6, 78.9 and 56.5 mAh g⁻¹ at 0.5 C, 1 C, 2 C and 3 C, respectively.

To further assess the practicality of Li₁₃Si₄@P-CNTs in anode pre-lithiation, the pouch cell test was performed. The assembled LFP||P-Gr pouch cell achieves a superior ICE (93.8%), favorable capacity retention (83.2% at 2500 cycles) and ultrahigh energy density (381 Wh kg⁻¹) in Figure 7d,e (the energy density was calculated based on the mass of the cathode). Furthermore, in SiO@C/Gr composite systems, the LFP||P-SiO@C/Gr pouch full cell delivers a superior energy density of 408 Wh kg⁻¹ after 2000 cycles (Figure S9). These results demonstrate the outstanding potential of Li₁₃Si₄@P-CNTs additives for practical application in anode pre-lithiation of high-energy-density LIBs.

3. Materials and Methods

3.1. Preparation and Synthesis of Li₁₃Si₄ and Li₁₃Si₄@P-CNTs

The Li₁₃Si₄ microparticles (50 mesh) were purchased from Shandong Chongshan Optoelectronic Materials Co., Ltd. (Zibo, China). To synthesize the Li₁₃Si₄ precursors, 1 g of pristine Li₁₃Si₄ particles and 3 g of Zirconia balls were placed inside a stainless-steel milling jar (45 mL). Ball-milling was performed in a planetary mill (P7, Fritsch, Idar-Oberstein, Germany) for 4 h at 400 rpm. Then, 1 g multi-walled carbon nanotubes (Canrd, Dongguan, China) were ball-milled with Zirconia balls at 100 rpm. The paraffin (Macklin, Shanghai, China), 1,3-Dioxolane (Dol, Aladdin, Shanghai, China), ball-milled Li₁₃Si₄ microparticles (2–10 μm) and CNTs were utilized to prepare the Li₁₃Si₄@P-CNTs. A total of 10 mL of Dol was used to dissolve 200 mg melted paraffin at 400 rpm stirring rate and 65 °C heating temperature. Then, ball-milled carbon nanotubes (5 mg) and Li₁₃Si₄ microparticles (400 mg) were mechanically ground and subsequently added to paraffin-Dol solution. The solution was stirred for 2 h and further cooled to room temperature for 1 h. The well-mixed solution was vacuum-filtered, followed by vacuum drying to obtain the uniformly coated Li₁₃Si₄@P-CNTs.

3.2. Materials Characterization

SEM (Zeiss, Zigma, Oberkochen, Germany) was utilized to study the morphology and microstructure of the pristine Li₁₃Si₄ and Li₁₃Si₄@P-CNTs. The crystal structures of the pristine Li₁₃Si₄ and Li₁₃Si₄@P-CNTs were investigated through a powder XRD (Rigaku SmartLab, Tokyo, Japan) and XPS was performed using Thermo ESCALAB Xi⁺ XPS Microprobe (Thermo Fisher Scientific ESCALAB Xi⁺, Waltham, MA, USA). Based on a laser with a wavelength of 532 nm, Horiba Evolution was used to generate Raman spectra (Horiba LABHRev-UV, Kyoto, Japan). Pyrolysis-GC/MS was used to confirm the composition distribution of the long-chain alkanes in paraffin (Thermo Scientific TRACE 1310, Waltham, MA, USA, Thermo Scientific ISQ LT, Waltham, MA, USA).

3.3. Electrochemical Measurement

Using a weight ratio of 6.5:2:1.5, Li₁₃Si₄/Li₁₃Si₄@P-CNTs, Super P and PVDF were mixed with Dol solvent and stirred to obtain a homogeneous slurry. To prepare the Li₁₃Si₄/Li₁₃Si₄@P-CNTs electrodes, the slurry was casted on copper foil, dried and cut into disks of 12 mm in diameter. The mass loadings of Li₁₃Si₄/Li₁₃Si₄@P-CNTs electrodes were generally between 1.2 and 1.8 mg cm⁻². Celgard 2340 Polypropylene (PP) and lithium metal were employed as the separator and anode, respectively. The electrolyte was 1.0 M LiPF₆ in EC and DEC (V/V = 1:1) with a 5 wt% FEC additive. The assembled cells were measured using NEWARE battery test system. The cells for the EIS experiment were assembled identically and measured using a CHI 760 electrochemical workstation, with a frequency range spanning from 0.01 to 10⁵ Hz. The SiO@C electrode was prepared by mixing SiO@C (M33-3, Macau Aolixin Technology Co., Ltd., Macau), Super P and polyacrylic acid (PAA) in 95% ethanol with a weight ratio of 70:20:10, followed by casting and vacuum dried. And P-SiO@C was fabricated by casting the Dol dispersion of Li₁₃Si₄@P-CNTs, Super P and PVDF (mass ratio 45:5:10) on the low-loading SiO@C electrode via the pre-lithiation layer strategy. The Gr and P-Gr electrodes were completed using a similar method for the SiO@C and P-SiO@C electrodes, respectively.

The full cells were assembled with the NCM811 cathodes and the SiO@C/P-SiO@C anodes. The pouch cells were assembled with the LFP cathodes and the Gr/P-Gr/SiO@C-Gr/P-SiO@C-Gr anodes.

3.4. DFT and Molecular Dynamics Simulation

The first principles calculations from DFT were carried out using the Vienna Ab Initio simulation package (VASP) and the Dmol3 module of Materials Studio. To further investigate their molecular electrostatic potential and interaction, the geometry optimization and electronic energy state calculation were conducted under the generalized gradient approximation (GGA) with Perdew−Burke–Ernzerhof (PBE). The convergence tolerance of energy and force was set to 1 × 10⁻⁵ eV/atom and 2 × 10⁻² eV/Å, respectively. The differential charge density distribution was post-processed using Vesta software (ver. 3.5.8). Molecular dynamics (MD) simulations were performed to analyze the difference in diffusion kinetics of H₂O in the paraffin system through the Forcite module in Materials Studio. The H₂O-paraffin system was constructed using the Amorphous Cell module. The COMPASS force field was used in all the calculations. Specifically, the system was subject to a 0.5 ps canonical ensemble (NVT). After the equilibration stage, the 4.5 ps unrestrained micro-canonical ensemble (NVE) simulation was performed at 298 K.

The NVE production step was sampled for analysis of the diffusion coefficient (D) and the FFV, given by Equations (1) and (2), respectively [54,55]:

$$D={\displaystyle \frac{1}{6Na}}\underset{t\to \infty }{\lim }{\displaystyle \frac{d}{dt}}{\sum }_i^{Na}<{\left|{\gamma }_i\left(t\right)-{\gamma }_i\left(0\right)\right|}^2>$$

(1)

$$\eta ={\displaystyle \frac{V_{free}}{{V_{occupy}+V}_{free}}}\text{×}100\%$$

(2)

where Na represents the number of atoms to be tested and r_i(t) and r_i(0) are the coordinates of the ith particle at time t and time 0, respectively. η represents the free volume fraction, V_free represents the volume available for detection and V_occupy is the remaining volume.

4. Conclusions

In conclusion, an effective anode pre-lithiation strategy Via protective coating additives is proposed to resolve the issue of high air sensitivity of Li–Si alloy pre-lithiation additives. We designed a composite membrane on the basis of hydrophobic paraffin and conductive CNTs to coat the Li₁₃Si₄ microparticles via liquid phase deposition, which integrates superior environmental stability and high Li⁺/e⁻ transport kinetics. The Li₁₃Si₄@P-CNTs shows a superior pre-lithiation capacity of 895 mAh g⁻¹ and capacity retention of 92.2% after exposure to 45% RH air for 6 h, which is attributed to the hydrophobic paraffin protection membrane based on the weak dipole–dipole interaction between C₂₈H₅₈ and H₂O. Furthermore, the Li₁₃Si₄@P-CNTs significantly enhances the charge transfer dynamics and SEI Li⁺ transport kinetics of the P-SiO@C electrode, as illustrated by EIS and XPS depth etching results. Using Li₁₃Si₄@P-CNTs pre-lithiation additives, the NCM811||P-SiO@C full cell delivers an enhanced cyclability (149.1 mAh g⁻¹ at 150 cycles) and high rate of performance (91.5 mAh g⁻¹ at 3C). Furthermore, the LFP||P-Gr pouch cell achieves a maximum energy density of 381 Wh kg⁻¹ after 2500 cycles, demonstrating its potential for practical applications. This environmentally stable and scalable Li–Si alloy additive enclosed by a protective membrane is an important guideline for the industrial pre-lithiation of Si-based materials for LIBs.