Sol–Gel-Derived Ge-Substituted LLZO Ceramic Coatings on Lithium-Rich Layered Oxide Cathodes for Improved Interfacial Stability

TL;DR

A sol-gel method is used to coat lithium-rich cathode particles with a ceramic layer that improves battery performance and stability.

Contribution

A citrate-assisted sol-gel process is introduced to create Ge-substituted LLZO coatings that enhance interfacial stability in lithium-rich oxide cathodes.

Findings

Ge-LLZO coatings are successfully deposited on LMLO cathodes using a sol-gel method.

Coated cathodes show improved electrochemical stability and slower capacity decay compared to uncoated ones.

Structural analysis confirms preservation of the LMLO host structure after coating.

Abstract

Gel-based routes, particularly sol–gel processes, offer a versatile pathway to generate uniform inorganic networks and gel-derived functional ceramics with controlled composition and interfacial coverage. In this study, we employ a citrate-assisted sol–gel coating strategy to form a precursor gel containing Li, La, Zr, and Ge species on lithium-rich manganese-based layered oxide (LMLO) cathode particles, followed by drying/thermal conversion to obtain a Ge-substituted garnet-type Li7La3Zr2O12 (Ge-LLZO) ceramic coating. Structural and surface analyses (FE-SEM/EDS, XPS, and FE-TEM) confirm the presence of surface-deposited coating-related species and coating-induced changes in surface chemistry, while bulk XRD is primarily used to verify that the layered LMLO host structure is preserved after the gel-to-ceramic treatment. Electrochemical testing indicates that the gel-derived Ge-LLZO coating can influence interfacial kinetics and resistance evolution, as reflected by differential capacity behavior, impedance responses, and rate capability trends, alongside microstructural observations suggesting reduced damage compared with bare LMLO after cycling. Overall, this work demonstrates that gelation-assisted deposition and gel-to-ceramic conversion enable Ge-LLZO surface coatings on LMLO cathodes that modulate interfacial kinetics and resistance evolution. Under the harsh 4.8–2.0 V/1C condition, the bare LMLO shows an abrupt capacity drop after ~60 cycles, while the coated LMLO exhibits a more gradual decay up to 100 cycles; further optimization is required for robust long-term stability.

Figures

Electrochemical potential profiles and electrochemical reaction equations of lithium-rich manganese-based layered oxide (LMLO) during the initial charge–discharge processes.

Full text

1. Introduction

In the contemporary era, marked by the proliferation of portable electronic devices and electric vehicles (EVs) and the rising adoption of renewable energy sources, there is an unprecedented global demand for high-energy storage solutions. Among these technologies, lithium-ion batteries (LIBs) have emerged as a cornerstone due to their high energy density, long cycle life, and relatively low weight. However, as the demand for portable devices continues to increase worldwide, the need for lithium batteries capable of meeting increasing energy storage requirements while maintaining high performance and durability has become increasingly urgent [1,2,3,4,5,6,7,8,9,10,11,12].

The active cathode material is a critical component influencing the electrochemical performance and overall efficiency of lithium batteries. Recent attention has turned toward Lithium-rich Manganese-based Layered Oxide (LMLO) cathode active materials as promising candidates for next-generation high-capacity cathode materials. Li-rich layered oxide cathodes have garnered significant interest due to their exceptional specific capacity and energy density. However, concerns persist regarding irreversible capacity loss during the initial charge cycles, leading to a degradation in Coulombic efficiency. This phenomenon, related to structural rearrangement caused by irreversible oxygen redox reactions, poses significant challenges for the practical implementation of these materials in high-performance lithium batteries [1,2,3,4,5,13,14,15,16,17,18,19,20]. The characteristic electrochemical features of LMLO during the initial activation, together with the associated reaction pathways and structural evolution, are summarized schematically in Figure 1.

$$[eqn]$$ $$[eqn]$$

During the discharge process, lithium ions engage in intercalation reactions at three distinct locations: within the layered, monoclinic, and rock-salt structures. Remarkably, the rock-salt structure exhibits relative stability, prohibiting its reversion to the monoclinic structure. Specifically, at voltages exceeding 4.4 V, Li ions intercalate into the monoclinic structure, albeit with minimal capacity utilization. In contrast, the majority of the capacity is observed when lithium ions intercalate into the layered structure, predominantly at voltages below 4.4 V. Additionally, within the low-voltage range (3.6 V and below), lithium ions preferentially intercalate into the spinel-like structure [3,21,22,23,24,25].

$$[eqn]$$ $$[eqn]$$ $$[eqn]$$

Electrochemical potential profiles and electrochemical reaction equations of lithium-rich manganese-based layered oxide (LMLO) during the initial charge–discharge processes.

Among various mitigation strategies, surface and interfacial engineering, including protective coatings and ion-conducting interlayers, has been actively explored to reduce electrolyte decomposition at high voltage, regulate interfacial impedance growth, and buffer mechanically/chemically driven surface reconstruction [2,26]. In particular, integrating a lithium super-ion conductor such as garnet-type Li_7_La_3_Zr_2_O_12_ (LLZO) has attracted attention as an interface-modification concept for Li-rich cathodes, because LLZO can provide a chemically robust Li^+^ transport pathway and potentially decouple fast charge transfer from direct cathode-electrolyte contact [2]. This approach has also been linked to improved rate capability and moderated resistance evolution in cathode systems where interfacial transport becomes limiting at high current density [2,27,28]. In addition, the lithium-rich nature of LLZO and its interfacial Li^+^ conductivity have been discussed as potentially beneficial for mitigating phase transformations associated with cation disordering in layered cathodes [27,29,30].

A practical challenge, however, is that realizing a thin and uniform LLZO-type ceramic layer on oxide cathode particles is nontrivial, because conventional ceramic processing typically requires high-temperature densification and can lead to discontinuous coverage or undesired side reactions [31,32,33]. To address this processing barrier, we adopt a citrate-assisted sol–gel coating route in which metal–citrate complexes form a processable precursor gel on LMLO particle surfaces. Subsequent drying and calcination convert the gel into a thin, gel-derived ceramic layer intended for cathode-interface engineering [3,4]. This wet-chemical “gel-to-ceramic” strategy is designed to improve coating conformity and scalability while retaining the advantages of an oxide-based, lithium-ion–conducting interphase.

In this context, Ge substitution is selected because Ge incorporation in garnet-type LLZO has been reported to stabilize the high-conductivity cubic phase and enhance ionic conductivity within specific composition windows, while also influencing microstructural densification behavior [30,34,35,36,37,38]. Prior studies on Li_7_La_3_Zr_2-x_Ge_x_O_12_ report cubic garnet formation across x = 0.1–0.4, supporting Ge on Zr substitution as a reproducible composition window; in this work, we employ x = 0.10 as a representative design point for a thin interphase []. Notably, securing sufficient Li^+^ transport through a thin, gel-derived ceramic interphase is critical for avoiding transport penalties; thus, Ge substitution is selected to help ensure adequate ionic conductivity even when the coating is thin and potentially partially crystallized after gel-to-ceramic conversion. Therefore, a Ge-substituted LLZO (Ge-LLZO) coating is expected to serve as a Li^+^ conductive and chemically robust interphase that can reduce interfacial polarization and modulate resistance evolution, especially during early-stage cycling, where interfacial impedance growth can be pronounced.

Unlike conventional ceramic-coating approaches that often rely on high-temperature densification and can yield discontinuous coverage or unwanted interfacial reactions, this work introduces a citrate-assisted, gelation-enabled deposition route that forms a conformal precursor network directly on LMLO particle surfaces and converts it into a thin garnet-type ceramic interphase via gel-to-ceramic processing. In addition, we deliberately employ Ge substitution to help maintain sufficient Li^+^ transport even when the coating is thin and potentially partially crystallized, thereby minimizing transport penalties while targeting interfacial stabilization. Finally, we correlate coating formation and surface chemistry (FE-SEM/EDS and XPS) with electrochemical proxies of interfacial kinetics and resistance evolution (CV/dQ/dV and EIS) and post-cycling cross-sectional observations, providing a practical framework for evaluating gel-derived ceramic interphases on Li-rich cathodes.

Here, we report a sol–gel-derived Ge-LLZO coating strategy for LMLO cathode particles and evaluate how the gel-derived coating affects coating formation and surface chemistry, electrochemical kinetics and polarization, and the evolution of interfacial resistance during cycling [4,5,27]. To further assess mechanically induced degradation, we perform cross-sectional observations to compare particle-level cracking and structural deformation in LMLO electrodes with and without the coating [14,39].

2. Results and Discussion

2.1. Structural and Chemical Properties

FE-SEM confirmed the uniform morphology and spherical particle distribution of the synthesized materials in Figure 2. In addition, Surface analyses using FE-SEM/EDS provide direct evidence of coating-related elemental distribution on the LMLO particle surface as shown in Figure 3. The EDS maps show the spatial presence of La and Zr together with Ge on the coated particles, supporting successful deposition of coating-derived species on the LMLO surface. Because the coating is thin and the interaction volume in EDS can include the underlying LMLO, the EDS maps are used to support elemental distribution trends rather than to report an absolute coating composition. Bulk XRD patterns of both bare and coated samples are dominated by the reflections of the LMLO host, including the characteristic layered and monoclinic features in Figure 4, indicating that the gel-to-ceramic treatment does not disrupt the bulk crystal structure. No distinct garnet-type reflections attributable to the Ge-LLZO coating are observed, which is expected because the coating is intentionally formed as a thin, low-loading surface layer and can be below the detection limit of laboratory XRD. Rietveld refinement was not performed because the present laboratory XRD dataset does not provide the resolution and instrumental calibration required for reliable refinement; thus, XRD is used here to confirm preservation of the LMLO host structure. The slight peak-position shifts after the coating/heat-treatment step can arise from thermal history–induced residual strain and near-surface defect/compositional variations, rather than a bulk structural transformation of LMLO. The dominant layered/monoclinic reflections remain unchanged, supporting preservation of the host framework.

The XPS analysis provides surface-sensitive information on the elemental states and bonding environments. In the coated sample, clear La 3d, Zr 3d, and Ge 3d (and Al 2p) signals are observed together with an O 1s shift, indicating oxygen-bonded surface modification associated with the gel-derived coating layer (Figure 5). These XPS results are interpreted as surface-sensitive indicators of coating-related element presence, and a rigorous quantitative stoichiometry of the thin coating is not claimed within this work.

FE-TEM observations show additional island-like nanoparticles on the surface of the coated sample (Figure 6). Moreover, FE-TEM directly shows a conformal surface interphase on the coated particle edge, with an estimated thickness of ~2–4 nm (Figure 6c). FE-TEM/EDS suggests that these localized islands are enriched in La and Zr (and may contain Ge), while the oxygen signal from such nanoscale features can be affected by thickness/overlap and is not used here to assign a definitive oxygen stoichiometry. Accordingly, we discuss these islands as localized secondary products formed during gel-to-ceramic conversion and distinguish them from the primary thin coating layer distributed over the LMLO surface. The measured lattice spacing of the (003) planes in the layered structure, determined by FE-TEM, is consistent with the d003 value estimated from the (003) reflection position in XRD (≈0.47 nm). Monoclinic structural planes such as (020) and (112) were also detected with spacings of 0.425 and 0.183 nm, respectively, in Figure 6.

In summary, the combined structural and chemical property analyses provided comprehensive insights into the composition, morphology, and surface characteristics of the synthesized materials (see Figure 2, Figure 3 and Figure 6) [4,14,16,27,39,40]. These findings deepen our understanding of the properties of these materials and their potential applications in various fields, including lithium-ion battery technology. Further investigations are warranted to elucidate the role of surface modifications and impurities in the performance and stability of the synthesized materials.

2.2. Surface and Morphological Characterization

Surface analyses using FE-SEM/EDS (Figure 4) and XPS (Figure 5) confirm the deposition of coating-related elements on the particle surfaces. Because the coating is thin and the interaction volume in EDS can include the underlying LMLO, the EDS maps are used to support elemental distribution trends rather than to report an absolute coating composition. The island-like particles observed on the coated surface show La/Zr-rich signals, while oxygen mapping/quantification for such nanoscale features can be less reliable. In contrast, the XPS survey and O 1s spectra show a distinct peak shift without a decrease in oxygen intensity, suggesting the presence of an oxygen-bonded surface layer distributed on the LMLO surface.

Taken together, the FE-SEM/EDS and XPS evidence, along with the electrochemical trends, is consistent with the formation of a thin, gel-derived garnet-type (Ge-LLZO) ceramic interphase on the LMLO surface. Because the coating is intentionally thin and low-loading, its contribution can be below the detection limit of bulk XRD; therefore, our conclusions rely primarily on surface-sensitive evidence and the observed modulation of interfacial resistance evolution.

2.3. Electrochemical Properties

The electrochemical property analysis revealed key insights into the performance of the synthesized materials. Figure 7 illustrates the initial charge–discharge profiles and dQ/dV curves, revealing comparable behavior between the bare and coated samples, albeit with a slight increase in the discharge capacity of the coated sample. Notably, the coated sample exhibits higher dQ/dV peaks during the initial cycle, indicating enhanced lithiation and delithiation kinetics. In Figure 8c, the bare LMLO exhibits a pronounced retention drop after ~60 cycles at 1 C, whereas the Ge-LLZO-coated LMLO shows a more gradual capacity decay up to 100 cycles, suggesting that the coating delays the onset of severe fading under high-rate cycling. Additionally, the resistance values obtained from Table 1 highlight a significant increase in the resistance of the bare sample after 100 cycles, further underscoring the detrimental effects of prolonged cycling on electrode performance. To contextualize our cycling results, we compared representative surface-coating strategies reported for Li-rich layered oxide cathodes under similar high-voltage operation (Table 2). While some studies report higher capacity retention, direct ranking is not straightforward because the reported values depend strongly on cell configuration and electrode conditions (e.g., areal loading, electrolyte formulation, cycle number, and C-rate). Therefore, Table 2 is provided to benchmark the present performance within a realistic range and to highlight the practical contribution of our approach. Notably, many high-performance interphases rely on vacuum-based processes such as ALD/PVD, whereas our citrate-assisted sol–gel route enables a solution-based, batch-scalable coating formation followed by moderate-temperature gel-to-ceramic conversion, offering a straightforward pathway to form an ultrathin interphase at controlled nominal loading. Figure 9 compares the cyclic voltammograms of bare and Ge-LLZO-coated LMLO electrodes over repeated scans. Both electrodes exhibit a pronounced high-voltage anodic feature near ~4.5 V during the initial scan, which is commonly associated with the activation of the Li-rich component and a strong anionic-redox contribution. Upon subsequent cycling, the redox response evolves toward a more distributed set of cathodic/anodic features in the mid-voltage region (highlighted box), reflecting the combined transition-metal redox and the progressive electrochemical/structural evolution that is typical of Li-rich layered oxides. Notably, the Ge-LLZO-coated electrode shows improved cycle-to-cycle overlap of the CV traces compared to the bare electrode. In the highlighted mid-voltage region, the coated sample exhibits reduced peak broadening and smaller cycle-dependent peak migration, indicating mitigated polarization growth upon repeated cycling. In contrast, the bare electrode displays more evident distortion and shift in the redox features with cycling, consistent with increasing kinetic hindrance and interfacial degradation. These observations support that the gel-derived Ge-LLZO interphase stabilizes interfacial charge transfer by providing a Li^+^-conductive and chemically robust transport pathway, thereby moderating resistance build-up and polarization during cycling. Additionally, as depicted in Figure 10, the discharge voltage gradually decreases over cycling, which is indicative of a degradation in the electrochemical performance. Figure 11 shows the rate capability measurements, revealing a noticeable decrease in the discharge capacity with increasing rate, which is attributed to inadequate lithium-ion diffusion due to the low ionic conductivity of the electrode. Remarkably, the coated sample showed some capacity recovery at lower rates, in sharp contrast to the failure of the bare sample to deliver any capacity at 2 C.

Post-cell measurements confirmed performance degradation, with cross-sectional FE-SEM imaging (Figure 12) revealing cracks in the bare sample. These cracks likely resulted from physical distortions owing to conductivity and structural issues, emphasizing the importance of structural stability in electrode materials [14,27].

3. Conclusions

This study demonstrates a gel-based (citrate sol–gel) coating approach to fabricate a gel-derived Ge-LLZO garnet ceramic layer on LMLO cathode particles. The formation of a precursor gel and its subsequent gel-to-ceramic conversion provide a practical route to distribute coating-forming species across the cathode surface and modify surface chemistry, as supported by combined FE-SEM/EDS, XRD, XPS, and FE-TEM analyses. Electrochemical evaluations further suggest that the gel-derived coating affects interfacial processes, evidenced by changes in differential capacity features, impedance behavior, and rate capability responses, together with post-cycling cross-sectional observations indicating mitigated mechanical damage relative to bare LMLO. While the bare LMLO exhibits a pronounced retention drop after ~60 cycles at 1 C, the Ge-LLZO-coated LMLO shows a delayed onset of severe fading and a more gradual decay up to 100 cycles. Further optimization of coating coverage/thickness and conversion conditions is still required for robust long-term stability. Quantitative composition and stoichiometry verification of the gel-derived interphase will require dedicated survey-based quantification and calibration, and this remains a subject for future work. Compared with representative coating approaches (Table 2), the key advantage of this work is the scalable, solution-based gel-to-ceramic processing that enables an ultrathin, low-loading ceramic interphase without vacuum-based deposition.

4. Materials and Methods

4.1. Synthesis of Precursors for LMLO Materials

The preparation of precursors for LMLO cathodes involves a meticulous two-step procedure. Initially, a precise amount of transition metal sulfate with a stoichiometry of Mn/Co/Ni at 0.54:0.13:0.13 was dissolved alongside ammonium bicarbonate in a solution containing distilled water and polyethylene glycol (PEG). After 30 min of vigorous stirring, the resulting brownish-red homogeneous transparent solution was carefully transferred to a Teflon-lined stainless-steel autoclave. Subsequently, the reactor was heated to 180 °C and maintained at that temperature for 15 h in a convection drying oven, followed by gradual cooling to room temperature. The resulting product was washed and dried to yield pink carbonate precursor powder.

4.2. Synthesis of LMLO Cathode Materials

The synthesis of the final LMLO cathode materials began by blending the prepared precursor with an appropriate lithium source and incorporating a 5 wt.% excess of Li_2_CO_3_ to offset lithium salt evaporation during subsequent high-temperature calcination. Thorough grinding with a mortar and pestle ensured homogeneity, after which the mixture was transferred to an alumina crucible. The calcination process was conducted in a tube furnace under ambient conditions, starting with an initial step at 500 °C for 5 h, followed by a subsequent treatment at 900 °C for 12 h. Both calcination stages adhered to a heating rate of 5 °C min^−1^. The resultant material derived from the precursors was designated as bare LMLO.

4.3. Citrate Sol–Gel (Gelation-Assisted) Synthesis of Ge-LLZO Coating on LMLO

The synthesis of the Ge-LLZO@LMLO composite involved several steps to achieve optimal composition and characteristics. Initially, bare LMLO was dispersed in anhydrous ethanol via sonication to yield suspension A. Subsequently, LiNO_3_ (10 wt.% excess), La(NO_3_)3·6H_2_O, Zr(OC_4_H_9_)4, and GeO_2_ were combined in anhydrous ethanol with a stoichiometric ratio of 6.6:3:2:0.10. The introduction of citric acid into this mixture was followed by a 2 h stirring period, resulting in solution B. Citric acid served as a chelating/gelation agent, enabling homogeneous distribution of Li–La–Zr–Ge precursor species and the formation of a precursor gel upon solvent evaporation. Solution B was then carefully introduced into suspension A, forming solution C. After 2 h of stirring at 60 °C, solution C underwent evaporation at 80 °C to yield the precursor gel. Consequently, the precursor gel was subjected to calcination at 750 °C for 5 h under oxygen conditions, resulting in the synthesis of the Ge-LLZO@LMLO composite. The precursor to the LMLO feed ratio was fixed to provide a nominal Ge-LLZO coating loading of ~3 wt.% relative to LMLO for reproducibility. This meticulous synthesis protocol ensured the formation of the desired composite material with tailored properties, facilitating its subsequent characterization and application within lithium-ion battery technology. Ge was introduced as an isovalent dopant targeting Zr-site substitution in the garnet framework, corresponding to a nominal Li_7_La_3_Zr_2-x_Ge_x_O_12_ composition with x = 0.10. This x value was selected within the reported cubic compositional window for Li_7_La_3_Zr_2-x_Ge_x_O_12_ [14].

4.4. Physical and Chemical Characterizations

The X-ray diffractometer used to determine the crystallinity of the prepared materials was the Rigaku Miniflex 600 (Rigaku Corporation, Tokyo, Japan), using Cu-Kα radiation (λ = 1.5406 Å) in a 2θ range of 10–80°. The particle morphology and size were characterized using field-emission scanning electron microscopy (FE-SEM, Leo Supra 55, Carl Zeiss Microscopy GmbH, Oberkochen, Germany) equipped with energy-dispersive X-ray spectroscopy (EDS, Oxford Instruments, Abingdon, UK) for elemental mapping. The morphology was observed using field-emission transmission electron microscopy (FE-TEM, JEM-2100F, JEOL Ltd., Tokyo, Japan). The surface elemental states and oxidation numbers were recorded using X-ray photoelectron spectroscopy (XPS, Multilab 2000, Thermo VG Scientific, East Grinstead, UK), and the results were deconvoluted using XPSpeak41 (Version 4.1) software.

4.5. Fabrication of Coin Half-Cell

The electrochemical performance of the as-prepared Ge-LLZO@LMLO cathode active material was studied using CR2032-type half cells. The synthesized active materials were combined with Denka black carbon and a PVDF binder at a ratio of 85:10:5 (wt.%) and mixed using a mortar and pestle in N-methyl-2-pyrrolidone (NMP) solvent to form a homogeneous slurry. The slurry was spread on aluminum foil and dried overnight at room temperature, followed by drying at 120 °C in a convection drying oven to prepare the working electrode. The as-prepared electrode was fabricated into a half-cell in an Ar-filled glove box (KIYON, Seoul, Republic of Korea) with lithium metal as the counter electrode and a Celgard 2340 film as the separator. A 1 M LiPF_6_ salt dissolved in ethylene carbonate (EC) and dimethyl carbonate (DMC) (1:1 vol.%) with 2 wt.% vinylene carbonate (VC) was used as the electrolyte.

4.6. Electrochemical Characterizations

Galvanostatic charge/discharge (GCD) tests were evaluated using a Battery Tester 05001 from HTC instruments, a Korean battery testing system, within a potential window of 2.0–4.8 V. Electrochemical impedance spectroscopy (EIS) and cyclic voltammetry (CV) were performed using an electrochemical workstation (COMPACTSTAT, IVIUM technologies, Eindhoven, The Netherlands).