Operando elucidation of all six reversible Li–Ga phase transitions, inverted hysteresis, and interfacial dynamics in a nanoconfined CMK-3/Ga anode†

To evaluate the phase transformation of gallium during lithiation/delithiation, in situ X-ray diffraction (XRD) techniques were performed during the electrochemical charge–discharge process. In the cell assembly, a side-by-side electrode arrangement was utilized into an ECC-Opto-Std-Aqu electrochemical test cell holder from EL-CELL test equipment GmbH, Hamburg, Germany, with a beryllium window top cover to pass the X-ray without interference, and 1 M LiPF₆ in EC:DEC (1:1 volume ratio) electrolyte was used. Detailed in situ cell preparation settings of real-time photos and video are shown in Fig. S1 of the ESI.† The CMK-3/Ga sample was charged and discharged with a potential window of 0.01–3 V at a constant current density of 100 mA g⁻¹. At the same time, XRD measurement was continuously recorded in the 2θ range from 10° to 80° at a scan rate of 0.5° min⁻¹ for the entire range, and the measurement was repeated for several hours.

In situ XRD patterns were collected using the D-8 Bruker (Bruker AXS GmbH, Karlsruhe, Germany) diffractometer with Cu Kα radiation (λ = 1.542 Å, operating at 40 kV and 25 mA); at the same time the galvanostatic charge/discharge process was performed with the Bio-Logic SP-300. The in situ technique allows for the direct characterization of phase changes in electrode materials during real-time electrochemical measurements.

Coin cells, fabricated as described in the general electrochemical performance section, were used for these measurements. The programmed sequence for each cycle was meticulously designed. It commenced with a charge (lithium deintercalation/de-alloying) via Linear Sweep Voltammetry (LSV) from 0.01 V to 3.0 V (versus Li/Li⁺). An EIS measurement was then performed immediately at the fully charged state (3.0 V). Following this, a discharge (lithium intercalation/alloying) was conducted via LSV from 3.0 V to 0.01 V. Finally, another EIS measurement was performed immediately at the fully discharged state (0.01 V).

This “Charge–EIS–Discharge–EIS” sequence constituted one full electrochemical cycle and was precisely programmed to repeat for a total of 15 cycles. All CV scans were conducted at a consistent scan rate of 0.1 mV s⁻¹ within the potential range of 0.01–3.0 V. EIS measurements were performed in the 1 MHz–10 mHz frequency range and with a 20 mV AC amplitude using a Bio-Logic VSP 300 potentiostat. This (LSV–PEIS–LSV–PEIS) sequence, looped for 15 cycles, provides an invaluable dataset for understanding the interplay between redox activity and impedance changes in real-time.

Conversely, during the heating process (Fig. 3e), as the temperature increased from 150 K to 300 K, a slow reversible phase transition from solid to liquid Ga was detected between 250 K and 280 K. This melting transition was characterized by the gradual broadening and eventual disappearance of crystalline peaks, signifying the complete liquefaction of gallium. The distinct temperature ranges for crystallization (200K–150 K) highlight a pronounced thermal hysteresis and the broad nature of the phase transitions, typical characteristics of nanoconfined materials.

These findings align with the results of Yarema et al., who reported that Ga nanoparticles (approximately 24 nm in size) remained liquid at room temperature and for several months of storage in the fridge (ca. −10 °C (263.15 K)).⁶¹ Our low-temperature analysis corroborates their findings, as we similarly observed Ga crystallite peaks between 200 K and 250 K during the cooling process.

The insights gained from this study are highly significant, as the CMK-3/Ga composite uniquely exhibits reversible amorphous-to-crystalline phase transition at low temperatures. This inherent thermomechanical property, where the nanoconfined gallium reversibly solidifies and melts over a specific temperature range, is crucial. It suggests that such Ga-infused mesoporous carbon could fundamentally enhance electrode structure stability, potentially reduce volume expansion through self-healing capability, prevent dendrite growth, and find novel applications in phase change devices.^44,52,61 This distinctly demonstrates that the CMK-3/Ga composite material itself holds beneficial thermomechanical features, extending beyond its electrochemical performance, making it a compelling material for advanced battery architecture and thermal management applications.

	Fig. 4 Electrochemical characterization of CMK-3/Ga 1:0.5 low loading composite anodes. (a) Nyquist EIS spectrum. (b) Second cycle cyclic voltammetry (CV) curves obtained at a scan rate of 0.1 mV s−1 within the potential window of 0.01 V to 3 V (vs. Li/Li+), with numbered peaks indicating identified Li–Ga alloying/de-alloying phase transitions: Lithiation (L#): L1: Ga → Li3Ga14 (∼0.7–0.71 V), L2: Li3Ga14 → Li5Ga9 (∼0.65 V), L3: Li5Ga9 → LiGa (∼0.59 V), L4: LiGa → Li5Ga4 (∼0.53 V), L5: Li5Ga4 → Li3Ga2 (∼0.44 V), L6: Li3Ga2 → Li2Ga (∼0.10–0.12 V); delithiation (D#): D1: Li2Ga → Li3Ga2 (initial), D2: Li3Ga2 → Li5Ga4 (∼0.30–0.31 V), D3: Li5Ga4 → LiGa (∼0.46–0.47 V), D4: LiGa → Li5Ga9 (∼0.75–0.77 V), D5: Li5Ga9 → Li3Ga14 (∼0.77–0.80 V), D6: Li3Ga14 → Ga (∼0.88–0.9 V) and the reversible A1 CuGa2 alloying interphase around 1.2 V. (c) Galvanostatic charge/discharge (GCD) cycling performance. (d) Rate capability.

The EIS spectrum of the pristine CMK-3-130 electrode (Fig. 4a, blue dot) displays a large semicircular arc in the high-frequency region, corresponding to a charge-transfer resistance of approximately 248 Ω. This elevated resistance suggests sluggish interfacial kinetics at the electrode–electrolyte interface, possibly arising from the formation of a resistive or unstable SEI layer. Additionally, the high surface area and defective pore architecture of CMK-3 may contribute to uneven current distribution and interfacial discontinuities that further hinder effective charge transfer. After the semicircle, the impedance response transitions into a short, nearly horizontal segment before gradually rising into a less steep Warburg slope. This two-step low-frequency behavior reflects the buildup of double-layer capacitance at pore entrances, followed by bulk ion diffusion within the mesoporous channels. The combination of high interfacial resistance and delayed ionic diffusion indicates that pristine CMK-3 suffered from limited charge transport kinetics, inefficient Li⁺ mobility, and a higher tendency toward side reactions, such as continuous SEI growth.

Although no equivalent circuit fitting was performed in this work, the observed impedance behavior aligns well with established models of mesoporous carbon-based electrodes.⁵ To support and validate our interpretations, we further refer to the work of Saikia et al.,⁶² where detailed Nyquist analysis of CMK-3 and CMK-8 electrodes was performed using equivalent circuit modeling. In that study, the authors attributed the depressed semicircles to the combined resistance of the SEI and charge-transfer processes, where the inclined Warburg line was linked to lithium-ion diffusion within mesoporous channels. Their results showed that differences in pore geometry significantly influenced both charge-transfer resistance and ion transport properties. Through correspondence, the large semicircle and shallow Warburg slope observed in our pristine CMK-3-130 sample are consistent with the diffusion and interfacial limitations described in their analysis, further confirming the interpretation of our general Nyquist features even in the absence of fitting.

The EIS spectrum reveals substantial changes in charge-transfer and ion transport characteristics upon the incorporation of Ga into CMK-3. For the CMK-3-130–Ga 1:0.5 composite (Fig. 4a, black dot), the Nyquist plot exhibits a significantly reduced semicircle diameter, corresponding to a lower combined resistance of the SEI layer and charge transfer (∼98 Ω). This marked decrease suggests that low Ga content facilitates enhanced charge-transfer kinetics, likely due to the optimal nanoconfinement of Ga within the mesoporous channels of CMK-3. Such confinement improves electronic connectivity across the carbon–Ga interface, forming more efficient electron transport pathways. In the low-frequency region, the impedance trace of the 1:0.5 composite displays a well-defined and steeper Warburg slope, accompanied by a leftward shift in the impedance response. This indicates a shorter efficient diffusion pathway of Li⁺, attributed to the homogeneous dispersion of Ga within the interconnected mesoporous matrix. The combination of reduced resistance and more favorable ion diffusion pathways reflects a highly synergistic interaction between Ga and CMK-3 at this optimal loading, enabling rapid and efficient lithium-ion transport throughout the electrode structure.

In contrast, the CMK-3-130–Ga 1:1.5 composite (Fig. 4a, red dots) presents a moderately sized semicircle, corresponding to a higher interfacial resistance of approximately 213 Ω. While this value remains significantly lower than that of pristine CMK-3 (∼248 Ω), it is notably higher than that observed for the 1:0.5 composite. The increased resistance at higher Ga loading suggests that excessive Ga may agglomerate or partially block the mesoporous channels of CMK-3, thereby diminishing electronic and ionic transport efficiency. Additionally, the low-frequency region of the 1:1.5 composite shows a rightward shift, and a steeper Warburg slope compared to the 1:0.5 sample, indicating increased ion diffusion resistance. This suggests that Li⁺ mobility was hindered due to reduced accessibility of the active surface area, likely caused by the partial pore obstruction or aggregation of Ga, which elongates the effective diffusion pathways and limits transport through the porous network.

The critical role of Ga nanoconfinement in enhancing charge transport and diffusion was further supported by recent findings from Guo et al., who demonstrated that in room-temperature liquid metal alloy systems, improved Li⁺ transport kinetics and reduced interfacial resistance were strongly correlated with the alloy's structural homogeneity and phase composition. Specifically, they reported that the Li diffusion coefficient in Ga-based systems reached up to ∼1.01 × 10⁻⁷ cm² s⁻¹ when in a liquid phase, several orders of magnitude higher than in typical solid-state electrodes, typically due to favorable chemical environments and improved phase interfaces within the alloy matrix.⁵² Although their study was on bulk LM systems, it directly underscores how phase structure, alloying behavior, and interface design influence ionic transport. Our findings revealed that nanoconfined Ga within CMK-3 channels can likewise promote fast ion movement by preserving conductive and accessible pathways during alloying/de-alloying transitions.

Overall, EIS results underscore the critical role of Ga loading in modulating the charge transport and diffusion behavior of CMK-3/Ga electrodes. Optimal low Ga content (CMK-3-130–Ga 1:0.5) leads to significantly improved electrochemical kinetics, manifesting as reduced charge-transfer and diffusion resistances, thereby facilitating more effective Li⁺ diffusion kinetics. This precise control over charge transport and diffusion is intrinsically linked to the enhanced phase resolution observed in the subsequent Cyclic Voltammetry (CV) profiles (Fig. 4b). Improved kinetics evidenced in EIS for lower Ga loadings directly correlate with the ability to resolve distinct Li–Ga alloying/de-alloying transitions in CV, reinforcing the conclusion that an optimal balance of Ga nanoconfinement offers the favorable pathways for both electron and ion transport within the composite anode.

Despite these advances, the electrochemical identification of the full spectrum of Li–Ga phases in carbon-based composites remains a significant challenge. Previous electrochemical analyses on similar Ga-containing carbon composites typically reported the formation of a limited number of phases, such as Li₃Ga₁₄, LiGa, and Li₂Ga.^41,42,44,45

To overcome these limitations and fully elucidate the redox reaction mechanisms and phase transformations occurring within the CMK-3/Ga anodes during lithiation and delithiation, operando Cyclic Voltammetry (CV) was conducted over a potential range of 0.01–3.0 V (vs. Li/Li⁺) at a scan rate of 0.1 mV s⁻¹. Representative second-cycle CV profiles are presented in Fig. 4b, with the first and fifth cycles shown in Fig. S4 a and b† for CMK-3-130–Ga1:0.5 and CMK-3-130–Ga-1:1.5, respectively. The second cycle was chosen as it typically provides a more stable and representative profile of the reversible redox response, reflecting the material's behavior after initial SEI layer formation and stabilization of the electrode surface. For comparison, the CV data of the pristine CMK-3-130 are also presented (Fig. 4b, blue line). It exhibits broad reduction features (below 0.4 V) and oxidation peaks (0.1–0.5 V) characteristic of carbonaceous materials, which emerge during lithium insertion and extraction.^11,62 However, CMK-3 itself possesses intrinsic drawbacks, such as a low initial reversible capacity ratio of approximately 34%¹¹ and poor cycling performance, which limit its practical applications. The incorporation of active liquid metals, such as gallium (which forms several compounds with lithium, including Li₃Ga₁₄, Li₅Ga₉, LiGa, Li₅Ga₄, Li₃Ga₂, and Li₂Ga⁵¹), presents a capable alternative for energy storage systems. The inherent phase transition from reversible liquid to solid in such metals can enhance the durability, cycle life, and stability of rechargeable alkali-ion batteries.

In stark contrast to pristine CMK-3, the successful incorporation of gallium into the CMK-3 mesopores is reflected in the distinct features of the CV profiles (CMK-3-130–Ga 1:0.5, black dashed line; CMK-3-130-–Ga 1:1.5, red line in Fig. 4b). The CV of the CMK-3/Ga composite reveals the formation of the SEI and clear electrochemical signatures corresponding to a complete set of Li–Ga alloying and de-alloying reactions observed across various Ga loadings (Fig. S4a and b†).

To investigate phase transitions during discharge/charge cycles (0.01–3 V, 100 mA g⁻¹), in situ XRD was conducted on pristine CMK-3-130 and CMK-3-130–Ga 1:1.5. The cell assembly for these measurements is illustrated in Fig. S1† and the video. Distinct phase evolution was observed for pristine CMK-3-130, with newly emerging peaks in the lower angle regions during lithiation (Fig. S11a–e†), highlighting irreversible changes and the potential versus time of this process indicated in Fig. S11f.† These peaks become more prominent and continue to increase in intensity over time, with insignificant reversible peak shifts observed during delithiation. Extracted patterns at 2θ ranges of 10 to 80° for the 0–11 and 11–22 h cycling periods (Fig. 6a and b) confirm the formation of compounds such as Cu₂O (PDF cards: 96-101-0942, 111-35-36, and 110-295-30), CuO (PDF cards 45-0937, 05-0661), Cu(OH)₂ (PDF cards 13-0420, 72-0140, 35-0505), Li₂O₂ (PDF cards 01-073-1640, 09-0355), LiOH (PDF cards 01-085-0777, 01-085-1064), Li₂CO₃ (PDF cards 87-0729, 22-1141), and Li_xC₆. These XRD peaks were referenced using multiple pdf cards providing additional verification of the observed diffraction patterns.^22,68–76 Peaks corresponding to the Cu collector cubic Cu (PDF card no. 003-1018), porous separator, and Be window were differentiated through control XRD measurements (Fig. S12†).

	Fig. 6 In situ XRD patterns with the 2θ ranges from 10° to 80° highlighting various compounds observed during GCD lithiation/delithiation of pristine CMK-3-130 at 100 mA g−1. (a) Phase evaluation for 0–11 hours. (b) Phase evaluation for 11–22 hours. (c) Comparison of in situ XRD patterns extracted at selected 2θ ranges from 34.75° to 35.3° and 43°–44° during various time frames (0, 11, 22, 33, 44, and 49 hours) of GCD cycling highlighting the phase transformation and providing insight into the redox process, corresponding to CuO, Cu(OH)2, Li2O2 and LiOH. (d) Corresponding to Cu, Cu2O, and Cu(OH)2.

Fig. 6c and d further isolates time-specific redox reactions showing Cu surface change alongside irreversible compound formation highlighting Cu surface dynamics. Fig. 6c displays a peak at approximately 2θ = 35°, which can be indexed to the [111] plane of CuO. The shoulder peak at 35° corresponds to the [111] plane and the peak at 34.9° to the [002] plane of Cu(OH)₂. Additionally, the peak at 35.14° can be indexed to the [101] plane of Li₂O₂, while the peak at 34.8° is associated with the [110] plane of LiOH. Fig. 6d shows a peak at around 2θ = 43.5°, which can be indexed to the [111] plane of Cu. The shoulder peaks at 43.5° can be attributed to the [200] plane of Cu₂O and the [131] plane of Cu(OH)₂.

In situ XRD measurements reveal that the interaction of the Cu surface with the electrolyte initiates a dynamic redox process. These processes suggest electrolyte-induced Cu degradation or oxidation initiated via microcracks, as supported by post-cycling SEM images (Fig. 5). This process can lead to the formation of various copper oxide and hydroxide phases, as demonstrated by XRD analysis. This lack of reversibility could be a sign of irreversible structural modification or the trapping of lithium within the carbon matrix limits the material's ability to return to its original after delithiation fully. Pristine CMK-3's low coulombic efficiency (34%) indicates irreversible reactions, including SEI formation, surface oxidation at [H], [O] sites, and Li trapping in the carbon matrix and corrosion-like reactions of Li_xC₆.¹¹ Findings by Zhang et al. demonstrate that Cu oxidation that led to Cu redeposition on the anode surface during cycling can result in performance degradation due to an increase in SEI layer thickness.⁷⁷ Our analysis of pristine CMK-3 revealed structural and textural features relevant to its performance as an anode material. Previously, direct confirmation of the mechanism behind copper current collector oxidation or degradation was challenging due to limitations in post-cycling surface analysis. However, in the current study, our comprehensive electrochemical analysis, particularly through operando CV-EIS measurements, provided direct evidence indicating surface oxidation of the copper current collector at higher potentials. This phenomenon aligns with observations by Wang et al.,²⁰ who elucidated the oxidation of Cu to Cu₂O and CuO during delithiation, manifested as characteristic peaks at 1.5 V, 2.5 V, and a shoulder at 2.7 V in their CV analysis. While direct evidence of redeposition remained elusive in our current surface characterization, the confirmed Cu oxidation in our system is a potential factor contributing to capacity fading. The specific high-potential features observed in our operando sequential CV-EIS analysis validate this Cu oxidation, along with their inferences for electrochemical performance, discussed in the CV-EIS section.

Our in situ XRD findings link the observed poor electrochemical efficiency and irreversible capacity loss in the pristine CMK-3 system to the surface instability of the carbon host. This instability leads to the degradation and subsequent oxidation of the copper current collector, resulting in the formation of various copper oxide and hydroxide phases. Furthermore, the emergence of irreversible compounds such as Li₂O₂ and Li₂CO₃ is observed. These cumulative factors collectively hinder lithium-ion cycling and significantly contribute to the system's capacity fading.

Building upon this understanding of the pristine CMK-3, the subsequent section presents the detailed in situ XRD analysis of the CMK-3-130–Ga 1:1.5 composite. This investigation thoroughly elucidates the sequential formation of all six Li–Ga alloying phases (L1–L6) during lithiation, their corresponding dealloying behavior (D1–D6) during delithiation, the unique inverted hysteresis features observed, and the concurrent formation of CuGa₂.

3.5.2. In situ XRD of CMK-3-130–Ga 1:1.5.

Our in situ XRD analysis reveals significant insights into the phase transitions of the CMK-3-130–Ga 1:1.5 composite during lithiation and delithiation processes. We observed the appearance of new diffraction peaks which are marked with a solid black circle at the onset of lithiation, and it progressively shifts towards lower angles over time, indicating the expansion of the lattice as lithium ions are intercalated into the Ga-infused CMK-3 structure (Fig. 7(a–d)). In contrast, during delithiation, these peaks shift back towards higher angles, reflecting the contraction of the lattice as lithium is deintercalated. By the end of the delithiation process (solid dashed line), the XRD peaks return to their original positions, consistent with the material's initial state before cycling. These reversible processes were observed over prolonged periods of time of up to 65 h (Fig. S13a–d†).

	Fig. 7 (a–c) In situ XRD patterns with the 2θ ranges from figure left to right (2θ = 24°–30°, 40°–45°, and 45°–50°) during GCD lithiation (solid black circle)/delithiation (black dashed line) of CMK-3-130–Ga 1:1.5 at 100 mA g−1. (d) In situ XRD measurement discharge/charge potential vs. time (h) for 0–43.5 h.

In situ XRD measurements were conducted on the 1:1.5 Ga-loaded sample due to the insufficient signal intensity observed at the lower 1:0.5 loading. The higher Ga content produced well-defined diffraction peaks, enabling clear tracking of lattice expansion and contraction during lithiation/delithiation, and unambiguous identification of Li–Ga phase transitions. This provided critical insight into alloying–dealloying mechanisms but revealed peak intensity decreases associated with capacity fading.

Although in situ data focused on the 1:1.5 sample, operando CV (Fig. 4b) and sequential CV-EIS analysis offer complementary mechanistic insights for both loadings. The 1:0.5 sample showed stable and reproducible redox features, while the 1:1.5 sample exhibited progressive peak reduction. Together, these results capture the structural dynamics, diffusion kinetics, phase evolution (L1–L6, D1–D6), unique inverted hysteresis features, and interfacial behavior across different Ga concentrations.

The appearance of these new XRD peaks and their corresponding voltage plateaus aligns well with recent ex situ XRD reports on carbon black confined Ga anodes.⁴¹ In such a case, Ga reacts with Li from the initial to the end of the lithiation, forming three electrochemically stable phases, including Li₃Ga₁₄ (rhombohedral), LiGa (cubic), and Li₂Ga (orthorhombic). Deshpande et al. also demonstrated the Ga solid–liquid transition state of the self-healing electrode surface.⁴⁰ Our in situ XRD outcomes prove these dynamic changes in the real-time lithiation process and reveal the appearance of the intermetallic phase through the gradual peak intensity changes over an extended time of discharging. The linear shifting of XRD peaks in both lithiation and delithiation suggests a strong correlation with the molar fraction of lithium within the lattice as discussed in the previous literature on similar systems^{40–42,51,78,79} and is consistent with our in situ XRD findings. This indicates that a complete alloying (liquid to solid)–dealloying (solid to liquid) transition occurs during the discharge/charge multi-step reaction process.⁴¹ The consistent, linear shift to lower angles during lithiation implies that the lattice parameter increases proportionally as lithium concentration rises, likely due to the incorporation of lithium into the Ga lattice.

Notably, Guo et al.⁵² identified several lithiation and delithiation phases in their ex situ study at specific voltages (0.23 V, 0.42 V, 0.7 V, 0.83 V, 0.97 V during charge and 0.8 V, 0.6 V, 0.4 V, 0.2 V during discharge), corresponding to Li₂Ga, LiGa, and Li₂Ga₇. Their interpretations, based on the Li–Ga phase diagram by Saint et al.⁵¹ distinguished two structural families: Li-rich phases (Li₂Ga, Li₃Ga₂, Li₅Ga₄) with layered 2D structures and Ga-rich phases (LiGa, Li₅Ga₉, Li₂Ga₇) with 3D frameworks. While these ex situ investigations provided valuable insights into peak evolution and some phase transitions, our in situ XRD uniquely offers direct and continuous identification of all six Li–Ga alloying and dealloying phases (L1–L6, D1–D6) throughout the electrochemical process, providing a comprehensive understanding beyond previously reported observations.

Extracted patterns at 2θ ranges of 10 to 80° for the 0–11 (lithiation), and 11–22 h (lithiation and delithiation) GCD cycling periods (Fig. 8a and b) confirm all six Li–Ga binary phase formation, and the copper electrode surface contact with Ga might form a surface level alloy of CuGa₂.^23,80,81 It reveals a significant improvement in the structural and electrochemical stability during the lithiation/delithiation process. The formation of Li₂Ga alloy was evident, suggesting that a robust interaction between Ga and lithium and CuGa₂ proves Ga contact with the copper current collector. During the initial lithiation (Fig. 8a) 0–11 hours, the patterns indicate the suppression of Cu oxidation, as no significant CuO or Cu(OH)₂ peaks emerge compared to pristine CMK-3. This result underscores Ga's ability to mitigate copper oxidation by reducing direct Cu surface exposure to the electrolyte. Unlike the pristine sample, which exhibits peaks corresponding to irreversible compounds such as Li₂O₂ and Li₂CO₃ (Fig. 6b), the CMK-3/Ga sample shows no such signals during the 11–22 hour cycling period (Fig. 8b), highlighting the stabilizing effect of Ga in mitigating their formation. These findings indicate that the CMK-3/Ga coating stabilizes the electrode's structural integrity and prevents the capacity loss associated with irreversible side reactions. The observed improvements highlight the role of Ga in enhancing the efficiency and lifespan of CMK-3-130–Ga electrodes. This contrasts with the pristine CMK-3-130, where Cu oxidation and irreversible phase formation dominate, leading to poor electrochemical performance.

	Fig. 8 In situ XRD patterns of CMK-3-130–Ga 1:1.5 recorded at 100 mA g−1 showing 2θ ranges from 10° to 80° during lithiation/delithiation. (a) Phase evaluation during 0–11 hours; the inset shows the partially enlarged view of phase transition from amorphous Ga to Li3Ga14 (L1), Li5Ga9 (L2), and the initial new peaks' appearance. (b) Phase evolution during 11–22 hours; the inset shows the partially enlarged view of progressive lithiation and its lithium-rich transition phases LiGa (L3), Li5Ga4 (L4), Li3Ga2 (L5), and Li2Ga (L6) shown with a green line, for delithiation reverse phase transformation is shown with a red line Li2Ga (D1), Li3Ga2 (D2), Li5Ga4 (D3), LiGa (D4), Li5Ga9 (D5), and Li3Ga14 (D6). Comparison of in situ XRD patterns of CMK-3-130–Ga 1:1.5 in the 2θ range of 43°–44° during GCD cycling (0–66 hours), CuGa2 phase evolution characteristics. (c) Dynamic CuGa2 peak shape changes during lithiation (black line) and delithiation (red line) for 0–33 hours, intensity shifted. (d) Without peak intensity shift for 0–66 hours.

Fig. S14† illustrates the in situ XRD patterns of CMK-3-130–Ga 1:1.5 during GCD cycling (0–66 h) in the 2θ range of 34.8–35.5°. Unlike the pristine CMK-3-130, which exhibits significant copper oxidation and unstable electrode response, CMK-3-130/Ga demonstrates the absence of Cu oxidation phases, attributed to the protective role of Ga. Ga's unique properties, such as its ability to form stable alloys and its inherent electrochemical activity, may provide a protective effect, preventing Cu from oxidizing. Such an action was proven by the initial coulombic efficiency improvement, stabilizing the Cu surface, reducing irreversible capacity losses, and enhancing system performance.

Insets in Fig. 8a and b, focusing on the 2θ ranges of 24.5–26.5° and 40.75–42.2°, display the full progression of Li–Ga phases. These insets highlight the evolution of all six Li–Ga alloying phases (L1–L6, solid green lines) during lithiation, corresponding to transitions from Ga-rich to lithium-rich compositions: Li₃Ga₁₄ (L1), Li₅Ga₉ (L2), LiGa (L3), Li₅Ga₄ (L4), Li₃Ga₂ (L5), and Li₂Ga (L6). Subsequent delithiation, presented as dealloying transition phases D1–D6 (solid red lines), corresponds to transition from lithium-rich back to Ga-rich compositions: Li₂Ga (D1), Li₃Ga₂ (D2), Li₅Ga₄ (D3), LiGa (D4), Li₅Ga₉ (D5), Li₃Ga₁₄ (D6). Table S3† provides a detailed summary of the lithiation/delithiation potentials and corresponding key structural observations.

3.5.2.1 Lithiation phase evolution. Fig. 8a illustrates the structural changes during the first lithiation cycle (0 to 11 hours). Initially, the pattern (black line) shows the characteristic amorphous signature of gallium, consistent with its nanoconfined state within the CMK-3 mesopores. As lithiation commences and progresses, we observe a gradual appearance of new diffraction peaks at around 2θ = ∼25°, ∼41.3°, and ∼49°. These initial peaks broaden and exhibit narrow intensity, indicating a major structural change in Ga as vast lithium insertion forms the initial alloying phases, Li₃Ga₁₄ (L1) and Li₅Ga₉ (L2). As lithiation further progresses, more sequential formation of lithium-rich Ga alloyed intermetallic phases LiGa (L3), Li₅Ga₄ (L4), Li₃Ga₂ (L5), and Li₂Ga (L6) shows a gradual increment in intensity and a characteristic shift towards lower 2θ angles (Fig. 8b), inset corresponding to lithiation. This shift signifies continuous lattice expansion as lithium is incorporated into the gallium host.

While individual phase transitions (L1–L6) compose a broad, single feature in the XRD pattern due to their closely spaced potential (Fig. 4b) and nanocrystalline nature, subtle fluctuations in peak intensity on top of this broad signal were observed, supporting the underlying, sequential phase transitions. At the final stage of lithiation, the overall broad peak appears to be slightly narrowed, with its composite nature becoming a more defined, ordered structure. This collectively demonstrates a well-defined sequential order of lithiation phases, evidenced by the gradual appearance of new peaks, their consistent increment in intensity, and a characteristic shift towards lower angles. As the second lithiation cycle commences, new L1 and L2 peaks emerge (dashed blue line, Fig. 8b). These phases exhibit similar structural dynamics to the first lithiation, with new peaks gradually emerging and shifting towards lower angles, affirming the high reversibility of the alloying process over multiple cycles. Prominently, although the individual differences in XRD peaks for these six phases (L1–L6) might be composed into one broad signal, conclusively distinguishable peaks detected for all six lithiation phases in our sequential CV-EIS operando analysis (as detailed in Fig. 9) strongly support the complete identification of all six Li–Ga lithiation phases. The similarity between these two powerful characterization techniques provides robust evidence for the full range of Li–Ga alloying events.

	Fig. 9 Operando sequential cyclic voltammograms (CVs) of CMK-3-150/Ga composites recorded over 15 cycles within a potential window of 0.01–3 V at a scan rate of 0.1 mV s−1. A partially enlarged view (0.01–1.5 V) highlights representative cycles 1, 2, 3, and 6 to capture key phase transitions. (a) Delithiation profiles for C:Ga = 1:0.5, the inset shows the single cycle D4 and D5 peaks partially highlighted with distinguishable humps. (b) Lithiation profiles for C:Ga = 1:0.5. (c) Delithiation profiles for C:Ga = 1:1.5. (d) Lithiation profiles for C:Ga = 1:1.5. Annotated peak positions correspond to Li–Ga phase transitions, where D1–D6 indicate delithiated states and L1–L6 denote lithiated states, in accordance with the Li–Ga binary phase diagram. Al represents the reversible alloying (lithiation) and dealloying (delithiation) of the interfacial CuGa2 phase.

3.5.2.2 Delithiation phase transition and inverted hysteresis. Fig. 8b captures the structural dynamics from the end of the first lithiation to the start of delithiation (11 to 22 hours). Upon initiation of delithiation, a rapid structural rearrangement occurs, with peaks emerging at approximately 2θ = ∼25.2°, ∼41.5°, and ∼49.2°. These initial delithiation phase transitions, particularly from the highly lithiated Li₂Ga, corresponding to D1, D2, and D3 (Li₂Ga, Li₃Ga₂, and Li₅Ga₄), are marked by the appearance of distinct, high-intensity peaks exhibiting split and relatively narrow features (red lines with white diamonds). This initial structural evolution is crucial, as it directly supports the inverted hysteresis observed in our electrochemical CV data (Fig. 4b) and further evidenced by our CV-EIS sequential data over 15 cycles (Fig. S19 and S20†). The sharp, high-intensity peaks in this region indicate significant structural rearrangements or electrochemical annealing that kinetically favor lithium extraction (dealloying) at lower potentials (0.3 V for D2, 0.46 V for D3), contrasting with their kinetically hindered formation (alloying) at higher lithiation potentials (0.44 V for L5, 0.53 V for L4). Crucially, our real-time in situ XRD analysis powerfully corroborates this intriguing inverted hysteresis, continuously capturing structural changes over prolonged cycling (up to 65 h for these initial delithiation phases, Fig. 7 and S13†). As delithiation continues, the subsequent dealloying of higher potential Li–Ga phases LiGa (D4), Li₅Ga₉ (D5), and Li₃Ga₁₄ (D6) exhibits a progressive broadening and a gradual decrement in intensity (red lines), indicating the sequential removal of lithium and the eventual transformation back towards amorphous gallium. Peaks consistently shift towards higher 2θ angles, reflecting lattice contraction. Upon complete delithiation, the peaks diminish, indicating a return towards the amorphous gallium initial state.
3.5.2.3 Dynamic evolution of CuGa₂ interphase at the current collector interface. Beyond the major Li–Ga alloying transformations, our in situ XRD patterns also provide critical insights into the dynamic behavior of the CuGa₂ phase at the current collector interface, a factor often overlooked but crucial for long-term battery performance. Fig. 8c and d specifically highlight the evolution of the CuGa₂ diffraction peak within the 2θ range of 43–44° during extended GCD cycling (0–66 hours).

Fig. 8c, which focuses on the 0–33 hour period with intensity shifts for clarity, reveals the subtle yet consistent changes in the CuGa₂ peak shape during lithiation (black line, annotated as A1 with green) and delithiation (red line, annotated as A1 with red). Upon lithiation (alloying), the CuGa₂ peak tends to become narrower, and its intensity shows a slight, gradual increase. This suggests that the formation of CuGa₂ leads to a more coherent or ordered crystalline structure. Conversely, during delithiation (de-alloying), the peak tends to become broader, and its intensity shows a slight decrement. This broadening, despite minimal overall intensity change, indicates that the decomposition process may lead to a comparatively more disordered, nanocrystalline, or strained interface.

The extended observation over 0–66 hours in Fig. 8d, without intensity shifts, illustrating long-term stability further emphasizes that these subtle shape and intensity fluctuations for the CuGa₂ phase are consistently reversible across multiple cycles. While the changes are minute, their reproducibility signifies that the CuGa₂ layer is dynamically forming and deforming at the interface in a way that effectively maintains its overall structural integrity and protective function. This ongoing yet subtle evolution of the CuGa₂ peak provides robust structural evidence for its role as a stable, buffering interphase layer, accommodating volume changes without undergoing bulk changes that would otherwise lead to significant pulverization or rapid capacity decay.²³

The formation and behavior of copper–gallium intermetallic compounds (IMCs) like Cu₉Ga₄ and CuGa₂ are well-established^21,82 in fields such as microelectronics and solders, where they're often associated with interfacial reactions and diffusion characteristics. For instance, studies have shown how Ga addition can even suppress detrimental IMC growth in other systems by forming dense protective layers.³⁶ However, our findings, particularly within the nanoconfined environment of CMK-3, reveal a remarkable transformation; the CuGa₂ phase exhibits dynamic stability and reversibility under electrochemical cycling, evolving from a potentially problematic material into a beneficial component of the battery anode.

The stability of the current collector interface is a critical, long-standing challenge in high-performance Li-ion battery anodes, primarily due to the significant volume changes of anode materials during cycling, which lead to capacity fading and reduced cycle life. While existing solutions range from surface engineering (e.g., CuO nanowires, Cu foams)²⁰ to self-healing liquid metal interfaces,^23,43,80,81 and extensive research has explored architecture (e.g., 3D porous) or surface-modified copper current collectors to enhance interfacial stability and control Li deposition in high-performance anodes,^83,84 the broader concept of in situ interphase engineering offers a powerful alternative to address diverse interfacial challenges across various battery chemistries. For instance, a recent ab initio study demonstrated in situ interphase engineering in an oxide-based all-solid-state Li battery cathode, where a computationally identified dopant (Al) enabled the formation of a stable protective layer that simultaneously improved sinterability and interfacial stability.⁸⁵ Analogously, our work introduces an up-and-coming alternative, the spontaneous electrochemical formation of a reversible, protective CuGa₂ layer, uniquely enabled by nanoconfined Ga.

Our comprehensive in situ XRD analysis provides fundamental structural evidence for this phenomenon. It details both the primary Li–Ga alloying/de-alloying sequences and the subtle, yet crucial, stable behavior of the interfacial CuGa₂ phase. As shown in Fig. 8c and d, the reversible narrowing and broadening of the CuGa₂ peak over extended cycling indicate its function as a flexible, self-adjusting buffer layer that maintains interfacial contact. This exceptional flexibility, unlike the typically brittle nature of bulk CuGa₂, is attributed to nanoconfinement, which likely restricts macroscopic crack propagation and promotes localized, reversible atomic rearrangements.

Further support for this dynamic and reversible interphase comes from our sequential CV-EIS analyses. Our CV data (Fig. S15†) show small, reversible alloying/de-alloying features around 1.2 V, consistent with CuGa₂ formation and dissolution.⁶³ EIS measurements, in turn, reveal stable interfacial charge transfer resistance, confirming a robust and continuously active interface. This electrochemical signature, along with evidence from the pristine sample surface evolution showing that the Ga surface is intrinsically mitigated by this protective layer, highlights the self-regulating nature of our system.

By providing strong structural and electrochemical evidence for a stable, dynamically reconfigurable interphase at the Cu surface, our results lay critical groundwork for next-generation current collector strategies. This in situ electrochemically formed CuGa₂ interface, with its inherent flexibility and self-adjusting properties, holds significant promise for extending cycle life, minimizing interfacial resistance, and inspiring future anode designs that integrate self-regulating intermetallic protection layers. This innovative approach offers a powerful strategy to overcome a key limitation in battery technology, paving the way for more robust and long-lasting energy storage devices.

The operando in situ XRD analysis, extended to 65 hours of continuous cycling (Fig. 7 and S13†), was instrumental in directly capturing the dynamic structural evolution of the CMK-3/Ga composite. Unlike prior ex situ reports that inferred or partially resolved Li–Ga phase transitions, our real-time tracking identifies all six Li–Ga alloying (L1–L6) and dealloying (D1–D6) phases. This unprecedented resolution is largely attributed to the nanoconfinement effect of mesoporous CMK-3, which effectively accommodates volume changes, restricts bulk crystal growth, and enhances surface-driven reaction kinetics. Notably, the in situ XRD results also provide structural validation for the ‘inverted hysteresis’ observed in the electrochemical profiles (Fig. 8b), particularly during the early dealloying transitions (D1–D3), suggesting a possible electrochemical annealing or rearrangement effect. These comprehensive structural insights offer a strong foundation for understanding the complex redox behavior and long-term stability of nanoconfined gallium-based anodes.

While in situ XRD reveals the crystalline phase transitions with high reliability, a complete understanding of the electrode's electrochemical behavior and kinetic pathways requires complementary time-resolved electrochemical techniques. In the following section, we present extensive operando sequential cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) analyses over 15 cycles (Fig. 9 and 10). This integrated approach enables high-resolution differentiation of subtle and overlapping electrochemical features, providing deeper insights into phase-specific redox transitions, kinetic reversibility, and the influence of Ga loading on charge transport dynamics. Additionally, these techniques allow us to evaluate interfacial stability, identify CuGa₂ formation at the Cu interface, and compare the performance of pristine and Ga-infused CMK-3 under different synthetic conditions and compositions. Together, these findings further elucidate the mechanisms governing electrochemical stability and efficiency in nanoconfined alloy anodes.

	Fig. 10 Operando sequential cyclic voltammograms (CVs) and corresponding electrochemical impedance spectra (EIS) for CMK-3-130/Ga (1:0.5 mass ratio) within a potential window of 0.01–3 V at a scan rate of 0.1 mV s−1. (a) Delithiation, inset arrow demonstrates absence of Cu2O/CuO peaks, bottom shows the corresponding EIS spectrum of (a), inset shows the partially enlarged view of 1, 2, 5, and 6 cycles. (b) Lithiation, the bottom shows the corresponding EIS spectrum of (b).

This electrochemical signature is attributed to the reversible alloying/dealloying of Ga with the copper current collector, which forms the intermetallic CuGa₂ phase at the interface. The confined Ga within the CMK-3 matrix likely promotes localized alloying while simultaneously suppressing copper oxide formation. The consistent potential position and reversible nature of this A1 peak over multiple cycles suggest the dynamic but stable evolution of a protective CuGa₂ layer, which buffers mechanical stress and maintains electrical contact.

This interpretation is strongly supported by in situ XRD data (Fig. 8c and d), where CuGa₂ diffraction peaks are observed and remain stable over extended cycling. Together, these findings confirm the role of nanoconfined Ga in facilitating the reversible formation of a structurally robust interfacial phase, critical for current collector protection and long-term cycling performance.

The operando cyclic voltammetry (CV) profiles of CMK-3-150/Ga composites at both mass ratios (C:Ga = 1:0.5 and 1:1.5) closely match those of CMK-3-130/Ga samples (Fig. 9a–d and S15a–d†), confirming the robustness and reproducibility of the six-step Li–Ga alloying sequence (L1–L6) over 15 cycles. This consistency highlights the effectiveness of CMK-3 nanoconfinement in stabilizing distinct Li–Ga phases otherwise challenging to resolve due to overlapping potentials and fast kinetics.

During lithiation (Fig. 9b, d, S15b and d†), six well-defined cathodic peaks corresponding to Li₃Ga₁₄ (L1), Li₅Ga₉ (L2), LiGa (L3), Li₅Ga₄ (L4), Li₃Ga₂ (L5), and Li₂Ga (L6) appear sequentially. Notably, the L1 peak, corresponding to the initial formation of the Li₃Ga₁₄ phase, emerges as a narrow, high-intensity signal. A subtle, transient shift of this peak to slightly higher potentials is observed in the earliest cycles, characteristic of an initial activation or structural re-arrangement within the nanoconfined system, before consistently returning to stable cycling characteristics for subsequent cycles. In situ XRD confirms this phase formation through a broad peak that progressively narrows and shifts to lower angles, indicative of lattice expansion. For low Ga loading, the L1 peak remains stable in both intensity and position over cycling, correlating with superior coulombic efficiency and long-term performance. In contrast, at higher loadings, L1 shows greater initial intensity but rapidly declines with cycling, indicating poorer phase stability and increased irreversibility. The L2 and L3 peaks are closely spaced but distinguishable, with L2 being smaller and broader in low-loading samples. The L4 peak becomes sharper and more intense with increased Ga content, while the L5 and L6 phases associated with deeper lithiation exhibit contrasting trends. In low-loading systems, L5 is broad, and L6 grows sharper with continued cycling, reflecting improved deep lithiation. In high-loading samples, however, L6 becomes less distinct and often overlaps, indicating hindered phase formation and lithium transport. These findings align with EIS results, where low-loading samples display enhanced diffusion and reduced interfacial resistance.

While CV reveals distinct phase transitions, in situ XRD provides structural confirmation. The alloying-induced XRD peaks broaden initially and shift to lower angles with progressive lithiation, reflecting lattice expansion and nanocrystalline phase formation (Fig. 7, 8a and b). Despite overlap in some XRD patterns due to nanoconfinement effects, the high resolution of CV allows clear electrochemical identification of each Li–Ga intermetallic phase.

Fig. S16† CV demonstrates the reproducibility of the D1 delithiation transition and the presence of potential ‘pinning points’ in the CMK-3/Ga composites over 15 cycles. Specifically, the CMK-3-130–Ga 1:0.5 sample shows a highly consistent D1 transition at approximately 0.05 V, with strong convergence across cycles (marked in circle), indicating a stable and confined reaction environment (Fig. S16a†). Similarly, the CMK-3-150–Ga 1:0.5 sample exhibits a sharp and reproducible D1 transition at around 0.04 V, also demonstrating marked convergence throughout the cycles (Fig. S16b†). Conversely, for samples with a higher loading (1:1.5) of both CMK-3-130 and CMK-3-150, the D1 transitions occur near 0.05 V but show reduced convergence (Fig. S16c and d†). This suggests increased structural variability and a less confined phase evolution in these higher loading samples.

This striking feature is the distinct convergence of all 15 delithiation curves at approximately 0.05 V. From this point, a consistent formation of a ‘thick line’ across all cycles highlights its role as a highly robust and reproducible thermodynamic potential for an initial delithiation step. This effectively acts as a ‘pinning point’ that the system consistently achieves regardless of subtle kinetic fluctuations in previous cycles, indicating a kinetically favorable and highly stable delithiation sequence. For the higher Ga loading composites, such convergence at 0.05 V is notably absent, with individual cycle lines remaining distinguishable, suggesting increased kinetic hindrance or structural variability.

This remarkable electrochemical stability and the manifestation of a ‘pinning point’ in nanoconfined alloying materials find strong parallels in other well-studied high-performance systems. For instance, Chan et al.'s seminal work on silicon nanowires demonstrated how nanoscale architecture can effectively accommodate the massive volume changes inherent to Si during lithiation/delithiation, leading to significantly improved cycle life and stable electrochemical cycling. Their foundational insights underscored the vital role of nanoscale confinement in mitigating pulverization and preserving structural and electrical integrity.⁶⁴ Building upon this understanding of structural stability, our work further demonstrates that nanoconfinement enables striking voltage profile convergence and highly reproducible electrochemical behavior, even under substantial volume changes. In the CMK-3 confined Ga system, the superior nanoconfinement at low loading composites provides a similar buffering effect, allowing the Ga particles to relax and move within the confined environment without losing electrical contact or suffering irreversible pulverization. This sustained structural integrity is pivotal for the observed capacity recovery after initial cycles and the overall robust electrochemical response. We hypothesize that during cycling, the continuous volume changes and electrochemical cycling lead to the dynamic emergence of smaller, more active Ga particles from the initially confined material. These newly formed, highly dispersed active sites, benefiting from the cushioned nanoconfinement provided by the CMK-3 matrix, can then more effectively and consistently participate in the subsequent electrochemical reactions, contributing to the observed long-term stability and the characteristic 0.05 V delithiation pinning point. Conversely, higher Ga loadings likely hinder this dynamic particle refinement and effective buffering action, resulting in less consistent electrochemical features and limiting the capacity recovery.

Our observation of inverted hysteresis for the D2 (∼0.30 V) and D3 (∼0.46 V) delithiation peaks, being lower than their lithiation counterparts (L5 and L4, respectively), offers a critical insight into the electrochemical behavior of these phases. As shown in Fig. 9c, d, S15c and d,† these specific delithiation peaks exhibit distinct characteristics compared to their lithiation. This phenomenon suggests a structural reordering or “electrochemical annealing” process occurring during the delithiation cycle within the nanoconfined environment. The initial formation (lithiation) of Li₃Ga₂ and Li₅Ga₄ may lead to a more disordered or strained structure, requiring a higher overpotential. However, during delithiation, the confined environment could facilitate a more relaxed or ordered structural rearrangement, thereby lowering the kinetic barrier for Li extraction and shifting the de-alloying potential to a seemingly ‘lower’ (more favorable) value than expected based on lithiation. This finding highlights the complex interplay between electrochemistry, structural dynamics, and nanoconfinement in determining overall battery performance. Understanding these precise kinetic and structural details is paramount for designing next-generation Li-ion battery anodes with improved stability and performance.

For these phases, the delithiation peak potentials are found to be lower than their corresponding lithiation potentials (e.g., Li₃Ga₂ delithiates at ∼0.31 V vs. ∼0.44 V for its lithiation; Li₅Ga₄ delithiates at ∼0.46 V vs. ∼0.53 V for its lithiation). This inverted hysteresis is a powerful indicator of unique kinetic and structural factors at play. Our in situ XRD analysis directly corroborates this finding: upon initiation of delithiation, the corresponding XRD peaks (for Li₃Ga₂ and Li₅Ga₄) exhibit a notable increase in intensity and a slight shift towards higher 2-theta angles (Fig. 8b), inset. This suggests that the lithiated phases (Li₃Ga₂ and Li₅Ga₄) undergo a favorable structural rearrangement or electrochemical annealing after formation, allowing for a more facile and kinetically advantageous lithium extraction, despite the higher kinetic barriers encountered during their initial formation via lithiation. The D2 peak (Li₃Ga₂) appears with high intensity and a narrow shape, while D3 (Li₅Ga₄) exhibits shared higher intensity and a smooth, narrower profile, indicating a progressively more distinct electrochemical signature as lithium is extracted from the more Li-rich states.

However, a subtle yet significant difference emerges in the delithiation profiles of CMK-3-150/Ga, particularly for the low loading (1:0.5). While the D4 and D5 peaks often appear as a broad, merged feature in the CMK-3-130/Ga samples (especially in initial cycles), the CMK-3-150/Ga (1:0.5) distinctively reveals a more pronounced and distinguishable hump on top of this broadened peak (Fig. 9a), inset. This observation provides solid evidence for the clear resolution of both the D4 and D5 phases, which might otherwise appear merged, and this phenomenon becomes even more evident in progressive cycles. The subtle increase in pore size (as CMK-3-150 typically has slightly larger pores than CMK-3-130, though still within the 4–5 nm range) likely led to a more effective nanoconfinement effect. This refined nanoconfinement may subtly refine the local environment for these phase transformations, contributing to better separation or stability of the intermetallic phases over cycling and allowing these distinct humps to emerge more clearly. While peak separation is most visibly pronounced in the low-loading CMK-3-150 sample, strong evidence for the complete phase resolution, albeit less distinct, is also present across other loadings and in the CMK-3-130 series upon close examination of progressive cycles. The final delithiation D6 peak appears broader but with sharper intensity compared to its nearest D5 phase. Furthermore, the intensity of this D6 peak decreases with progressive cycling, which suggests a smoother transition back to its initial Ga states. The reversible appearance of the A1 peak during de-alloying further confirms the dynamic nature of the CuGa₂ interfacial layer.

This distinct behavior is further validated by the first-cycle GCD profile of CMK-3-130–Ga 1:0.5 shown in Fig. S17.† The discharge (lithiation) segment displays a sloping curve beyond the initial L1 plateau (∼0.7 V), reflecting kinetically hindered alloying transitions that result in merged signals rather than well-separated plateaus. Conversely, the charge (delithiation) profile exhibits exceptionally well-resolved, stepwise plateaus (D1–D6). These clear electrochemical fingerprints correspond to the sequential dealloying of all six Li–Ga binary intermetallic phases (from Li₂Ga to Li₃Ga₁₄), providing compelling evidence of their distinct electrochemical potentials. This striking asymmetry merged lithiation slope versus precisely resolved delithiation plateaus is characteristic of advanced alloying systems and strongly reinforces our sequential CV and real-time in situ XRD findings, which also distinguish all six redox pairs. Collectively, these GCD results profoundly underscore the high degree of phase reversibility and structural stability facilitated by nanoconfined Ga, especially under optimal loading, offering critical mechanistic insight into the performance of alloy-type anodes.

This detailed electrochemical analysis, driven by high-resolution operando CVs and powerfully corroborated by real-time in situ XRD, unequivocally demonstrates the successful formation and remarkable electrochemical reversibility of all six predicted Li–Ga binary intermetallic phases within the nanoconfined CMK-3 host (Table S3†). Our work represents the first comprehensive electrochemical resolution of these intricate Li–Ga phase transitions, providing critical mechanistic insights into the alloying processes that govern the performance of Ga-based anode materials. Crucially, the Ga-engineered composites effectively suppress copper current collector oxidation, thereby mitigating the irreversible capacity loss commonly observed in pristine CMK-3 electrodes. The observed phase-specific kinetics, including intriguing inverted hysteresis and distinct pinning points, highlight the profound impact of nanoconfinement. Specifically, optimal nanoconfinement (e.g., in CMK-3-130 and CMK-3-150 with ∼4–5 nm pores) at lower Ga loadings enables kinetically favorable delithiation processes, leading to superior long-term capacity retention and unique capacity recovery. Conversely, higher Ga loadings introduce increased kinetic hindrance and structural variability, limiting their long-term electrochemical performance. These comprehensive findings collectively lay a robust foundation for the rational design of highly stable and efficient Li-ion battery anodes with broader implications for next-generation electrochemical energy storage devices.

The CV profiles for both the 1:0.5 and 1:1.5 CMK-3/Ga mass ratios confirm the excellent reversibility of the Li–Ga alloying/de-alloying process over 15 cycles. The consistent appearance and relative positions of the lithiation (L1–L6) and delithiation (D1–D6) peaks across multiple cycles underscore the structural stability of the nanoconfined Ga.

However, a closer examination reveals the critical and differential role of Ga loading in the overall electrochemical performance. For instance, while the higher loading CMK-3-130/Ga (1:1.5) composite exhibits a notably higher initial coulombic efficiency (ICE) of approximately 57% for its initial cycles, compared to the lower ICE of around 50–52% for the 1:0.5 loading, this initial advantage does not translate to superior long-term stability for higher loading samples.

In terms of capacity retention, the 1:0.5 mass ratio composite demonstrates superior long-term cycling stability, retaining approximately 56% of its capacity after 100 cycles. Strikingly, this capacity retention value is even higher than its initial ICE, indicating a unique capacity recovery phenomenon over prolonged cycling. This is a significant claim, suggesting that nanoconfined gallium becomes increasingly active and accessible with cycling. This capacity recovery is consistent with our CV observations for low loading, where the deep lithiation peak (L6, near 0.1 V) remains distinguishable, even showing a minute shift towards lower potentials and an increase in intensity over cycles, suggesting a smoother and more complete lithiation process over time. In stark contrast, the higher loading (1:1.5) composite suffers from significant capacity fading, retaining only about 16% of its capacity after 100 cycles, primarily due to the diminished intensity of its deep lithiation peaks and increased irreversible reactions.

This beneficial behavior of low Ga loading, including the phenomenon of capacity recovery, is consistently observed across different CMK-3 synthesis temperatures, notably for the CMK-3-150/Ga low loading as well. Both CMK-3-130 and CMK-3-150 share similar pore sizes, generally ranging between 4 and 5 nm, which appears to be crucial for providing the optimal nanoconfinement necessary for these stable and recoverable electrochemical processes. The nanoconfinement ensures Ga remains highly active and accessible, preventing detrimental agglomeration and facilitating the observed long-term stability and capacity recovery.

The inherently high initial irreversible capacity loss of pristine mesoporous carbons like CMK-3 has been widely attributed to their large surface area, leading to extensive Solid Electrolyte Interphase (SEI) formation and associated irreversible faradaic reactions during the first lithiation cycle.¹¹ Our operando CV-EIS analyses of the pristine CMK-3-130 (Fig. S18†) directly demonstrate the severity of this issue and highlight how gallium infusion in CMK-3 significantly mitigates it.

In the pristine CMK-3-130 CV profile (Fig. S18a†), representing delithiation cycles 1–15, the initial cycle exhibits a pronounced irreversible capacity, characterized by a large voltage hysteresis between the first and second cycles (marked as 1 and 2). This large irreversible capacity is primarily associated with initial lithium consumption for SEI formation, which tends to be thicker and uneven due to the carbon's large surface area. Furthermore, the inset of Fig. S18a† clearly shows the appearance of distinct Cu oxidation peaks (e.g., at ∼1.89 V, 2.31 V, 2.5 V, and 2.7 V) in subsequent delithiation scans after the first cycle. This suggests that the significant volume changes experienced by the pristine carbon during initial lithiation lead to electrode cracking, allowing electrolyte penetration and subsequent oxidation of the copper current collector surface to complex phases such as Cu₂O/CuO and other oxides. A small intensity hump observed around 1.2 V during delithiation is consistent with partial SEI decomposition, as reported by other studies on CMK-3.⁶²

The corresponding EIS spectra for pristine CMK-3-130 (Fig. S18a†) provide direct evidence for the instability and progressive impedance increase over cycles. The Nyquist plots display a large, often ill-defined semicircle in the high-to-mid frequency range, where the R_SEI and R_ct largely overlap (indicated by the arrow with the circle mark), signifying a highly resistive and intertwined interfacial impedance during cycling. The inset of Fig. S18a,† showing an enlarged view of extracted cycles (1, 2, 5, and 6), further demonstrates that the first delithiation cycle exhibits a large semicircle, and for the second cycle the corresponding EIS shows an even higher and wider semicircle than the first, indicating an immediate increase in resistance. Beyond this, while some reduction in semicircle size is noted (e.g., cycle 5), inconsistent behavior with subsequent increases (e.g., cycle 6) is observed. Moreover, the low-frequency Warburg region during delithiation is poorly defined, and the observed two-step low-frequency behavior with a shallow Warburg slope indicates severely hindered ion diffusion within the pristine carbon. Collectively, these impedance trends, along with the electrochemical features, indicate significant structural irregularities and instability within the electrode, further supported by post-cycling surface morphology observations that reveal a cracked and uneven surface (Fig. 5e, i and j).

For the pristine CMK-3-130 lithiation (Fig. S18b†) and its corresponding EIS spectrum, fewer distinct features are observed in the CV, apart from a small nuisance peak around 0.15 V, likely associated with initial Li insertion into the carbon matrix. In the EIS (circled and arrowed mark), the R_s shows a slight shift towards the left, and the semicircle generally exhibits a decreasing trend. Warburg region was not included due to its noisy and inconsistent nature, reflecting complex diffusion dynamics during lithium insertion into this unstable pristine host.

3.6.5.1 CMK-3-130/Ga (1:0.5 mass ratio): enhanced reversibility and kinetics. Compared to pristine CMK-3-130, the CMK-3-130–Ga (1:0.5 mass ratio) composite exhibits markedly improved electrochemical behavior, as demonstrated through the sequential CV-EIS analysis (Fig. 10a and b). The cyclic voltammograms (CVs) over 15 cycles reveal a substantial reduction in first-cycle irreversible capacity, as evidenced by the minimized hysteresis between the first two cycles. Notably, the absence of parasitic Cu₂O/CuO oxidation peaks indicated in Fig. 10a persists throughout the entire cycling sequence. This suppression of side reactions is consistent with previous in situ XRD results, which confirmed the formation of a protective CuGa₂ interphase at the electrode interface. Although in situ validation focused on higher loadings, it is reasonable to infer similar behavior in the low-loading system due to analogous structural features.

In the delithiation profile (Fig. 10a), a fluctuating peak around 1.2 V initially absent emerges in later cycles, suggesting a reversible deformation of the CuGa₂ interphase, consistent with the A1 peak discussed in Section 3.6.3. The lithiation CV (Fig. 10b) shows overlapping, stable peak profiles with increasing or constant intensities, indicating enhanced reversibility. All six Li–Ga alloying phases (L1–L6), as outlined in Section 3.6.2, are resolved and maintain stability throughout the cycles. A prominent peak near 0.1 V further signifies stable deep lithiation behavior.

The corresponding EIS analysis during delithiation (Fig. 10a, bottom) reveals a progressive reduction in both solid-electrolyte interphase (SEI) resistance (R_SEI) and charge transfer resistance (R_ct), as evidenced by the decreasing diameters of the corresponding semicircles in Nyquist plots. These semicircles are distinctly separated, indicating well-resolved contributions from SEI and charge transfer processes. Concurrently, the Warburg region exhibits a leftward shift and a steeper slope, reflecting enhanced lithium-ion diffusion kinetics and improved interface stability, consistent with observations in Fig. 4a. The inset in Fig. 10a, displaying impedance spectra for cycles 1, 2, 5, and 6, further supports this trend, showing a consistent decrease in impedance and increasingly pronounced Warburg behavior over cycling. These findings align with the electrochemical convergence and pinning effect observed in Fig. S16,† suggesting improved interfacial stability and charge transport efficiency. Collectively, these results highlight the CMK-3/Ga's potential for enhanced cycling performance and longevity in lithium-ion batteries.

During lithiation (Fig. 10b bottom), the EIS reveals a progressive decrease in the diameter of the semicircle, indicating a reduction in R_ct and R_SEI. These overlapping R_ct and R_SEI contributions suggest the formation of a uniform and highly conductive electrode–electrolyte interface, which remains stable despite repeated volume changes during cycling. The solution resistance (R_s) also remains consistent, underscoring the structural integrity of the Ga-loaded electrode. In contrast, the pristine CMK-3 electrode exhibits large and erratic impedance fluctuations, likely due to its heterogeneous surface and unstable SEI formation. The Warburg region, associated with lithium-ion diffusion, was excluded from the spectra due to significant noise, which obscured a reliable interpretation; however, this does not detract from the observed trends in R_ct, R_SEI, and R_s. These results highlight the superior electrochemical stability of the Ga-loaded electrode compared to pristine CMK-3, attributed to gallium's role in enhancing interface uniformity and conductivity.

These findings are corroborated by prior studies on alloy-based anodes, which reported enhanced electrochemical performance through reduced charge-transfer resistance and increased electrochemically active surface area. For instance, Yu et al.⁸⁶ observed a significant reduction in impedance after the initial lithiation cycle in Cu–Ga alloy-based anodes, attributed to stabilized charge-transfer processes. Similarly, Li et al.⁸³ demonstrated that 3D Cu/Li electrodes with high surface area exhibit enhanced lithium-ion diffusion kinetics, leading to lower R_ct. Guo et al.⁵² further reported that removing surface oxides in liquid Ga–In anodes reduces impedance and improves lithium-ion diffusion, aligning with the stable R_s and decreasing R_ct observed in our CMK-3/Ga system. The progressive reduction in impedance in our Ga-loaded CMK-3 electrode supports the hypothesis that well-dispersed Ga nanoparticles increase the electrochemically active surface area, as evidenced by post-cycling SEM (Fig. 5k), thereby enhancing lithium-ion diffusion and cycling stability. This carbon-hosted nanoconfinement principle is mirrored in systems like carbon-encapsulated CeO₂–Co catalysts for zinc–air batteries, where nanostructured carbon frameworks form a corrosion-resistant interface, mitigating oxidative degradation thus improving structural reversibility.⁸⁷ These comparisons underscore the role of gallium nanoconfinement in achieving a robust and conductive electrode–electrolyte interface in our system.

3.6.5.2 CMK-3-130/Ga (1:1.5 mass ratio): interfacial instability at higher loading. The CV and EIS profiles of the CMK-3-130–Ga electrode with a 1:1.5 mass ratio (Fig. S19a and b†) reveal distinct electrochemical behavior compared to the optimized 1:0.5 counterpart. In delithiation (Fig. S19a†), the CV shows reduced first-cycle hysteresis and no Cu₂O/CuO peaks, with a minor 1.2 V peak confirming CuGa₂ formation, consistent with the low-loading electrode. However, lithiation (Fig. S19b†) exhibits high-intensity initial peaks that reduce with cycling, indicating capacity degradation. The 0.12 V peak's reduced resolution suggests lower surface activity for lithium storage, correlating with decreased long-term reversibility. We attribute the capacity degradation in high-loading CMK-3/Ga to larger Ga nanoparticles, which diminish carbon nanoconfinement and promote electrolyte decomposition, as evidenced by post-cycling SEM analysis (Fig. 5h).

EIS spectra for delithiation (Fig. S19a†) show an uncertain semicircle after some cycles, also the Warburg region shows scattered data due to high signal-to-noise ratios, reflecting unstable lithium-ion diffusion. During lithiation (Fig. S19b†), the semicircle diameter decreases slightly, yet R_SEI and R_ct resistances remain elevated and poorly resolved. The Warburg component was excluded due to excessive noise, suggesting a kinetic barrier, potentially exacerbated by high state-of-charge (SoC) or surface heterogeneity at elevated Ga loading. These findings highlight the limitations of excessive Ga incorporation, contrasting with the stable interface of the low-loading system.

3.6.5.3 CMK-3-150/Ga series: nanoconfinement enhances reversibility and kinetics. The electrochemical behavior of CMK-3-150/Ga composites (Fig. S20a–h†) further validates the robustness of the nanoconfinement strategy across different synthesis temperatures. For the 1:0.5 Ga loading sample (Fig. S20a and c†), the CVs for both delithiation and lithiation closely mirror the superior performance observed in the CMK-3-130–Ga (1:0.5) system. The delithiation CVs exhibit highly overlapping features and significantly reduced first-cycle hysteresis, along with the appearance of subtle humps corresponding to the D4 and D5 phases, as previously discussed in Fig. 9a. The comprehensive 15-cycle dataset further underscores the long-term reproducibility of phase evolution and the consistent suppression of detrimental Cu₂O/CuO-related side reactions. Correspondingly, lithiation CVs exhibited nearly overlapping cathodic peaks and stable fluctuations near 1.2 V, indicative of stable CuGa₂ interphase formation and transformation. This observation reinforces that the Ga content plays a pivotal role in current collector protection.

The EIS spectra during delithiation for the 1:0.5 Ga loading (Fig. S20b†) consistently exhibit stable solution resistance (R_s), and the progressive reductions in SEI resistance (R_SEI), and charge transfer resistance (R_ct). The continuous shrinkage of the high-frequency semicircle and a clear leftward shift of the Warburg region signify significantly enhanced lithium-ion diffusion kinetics and stabilized electrode–electrolyte interfaces. These impedance features, particularly the distinct reduction in semicircle width and the well-defined Warburg line displayed in Fig. S20b inset,† collectively support kinetically reversible processes, improved charge transfer, the formation of a stable and progressively thinner SEI layer, and optimized lithium-ion pathways. These findings closely match those observed in CMK-3-130–Ga (1:0.5), confirming that the optimized Ga dispersion strategy yields comparable kinetic and interfacial performance across both synthesis temperatures. During lithiation (Fig. S20d†), EIS also shows a reduction in semicircle diameter, although the Warburg region appears noisy and undefined, manifesting as circling behavior. This likely stems from dynamic structural strain at high states-of-charge (SoC), suggesting a kinetic barrier, which was further clarified through a magnified view in Fig. S20d inset.†

For the higher Ga loading (1:1.5) sample (Fig. S20e–h†), CV responses are consistent with the CMK-3-130–Ga (1:1.5) system. While the initial CV peaks appear with greater intensity, they gradually fade over continued cycling, indicating reduced long-term stability. The CuGa₂-related peak at 1.2 V remains apparent but becomes more prominent compared to low-loading samples. EIS analysis (Fig. S20f and h†) further confirms increased interfacial resistance and structural instability at this higher loading. Scattered and unstable Warburg behavior, coupled with poorly defined semicircle features, reflect compromised lithium-ion transport and interfacial control, which is consistent with the capacity fading trends observed in electrochemical testing.

Overall, the comparative analysis between CMK-3-150/Ga and CMK-3-130/Ga systems shows that Ga nanoconfinement plays a crucial role in achieving stable, reproducible, and kinetically favorable electrode behavior. Specifically, the 1:0.5 Ga loading consistently maintained optimal phase evolution, interfacial protection, and diffusion characteristics across both temperature regimes, demonstrating its superior reversibility and high structural stability. These findings highlight the kinetically driven pathways observed at optimal loadings (1:0.5), contrasting with the more significant kinetic hindrance observed in their high-loading counterparts.

3.6.5.4 Operando CV-EIS analysis summary and reaction pathway mapping. Taken together, our operando CV-EIS data combined with in situ XRD structural insights allow for the full mapping of lithiation and delithiation reaction pathways for the Ga-infused CMK-3 anode system. For the first time, all six Li–Ga alloy phases are both electrochemically resolved and structurally confirmed in sequence:

Lithiation: Ga(l) → Li3Ga14 → Li5Ga9 → LiGa → Li5Ga4 → Li3Ga2 → Li2Ga

Delithiation: Li2Ga → Li3Ga2 → Li5Ga4 → LiGa → Li5Ga9 → Li3Ga14 → Ga(l)

This proposed reaction sequence, supported by the reproducibility of sequential CV cycling and in situ XRD validation, can be incorporated into Fig. 4b to contextualize the phase evolution. Importantly, the nanoconfined Ga environment enables kinetic stabilization and clear resolution of otherwise transient or overlapping intermediate phases, particularly Li₃Ga₂, Li₅Ga₄, and Li₅Ga₉, which are often indistinguishable in bulk systems due to phase overlap and rapid transformation kinetics.

Additionally, a comparative overview of the CMK-3/Ga system with other advanced anode materials is summarized in Table S4† highlighting synthesis methods, structural properties, low-temperature behavior, and key electrochemical metrics. This comprehensive comparison allows us to contextualize insights within the broader field.

Across both CMK-3-130 and CMK-3-150 systems, the 1:0.5 mass ratio consistently exhibits superior electrochemical performance, attributed to optimal Ga dispersion within the carbon framework. Enhanced CV resolution, minimal impedance, and suppressed side reactions confirm the advantage of nanoconfined Ga at lower loadings. Conversely, higher Ga content compromises structural integrity and hinders long-term stability, despite initial improvements in ICE. These findings reaffirm the critical role of nanoconfinement and precise compositional tuning for advancing alloy-type anodes.